vbp/VBPLayer.py at master · manuelhaussmann/vbp · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
import torch as th
import torch.nn as nn
import torch.nn.functional as F
from torch.nn.modules.utils import _pair
import math
import numpy as np


class VBPLinear(nn.Module):
    def __init__(self, in_features, out_features, prior_prec=10, isoutput=False):
        super(VBPLinear, self).__init__()
        self.n_in = in_features
        self.n_out = out_features

        self.prior_prec = prior_prec
        self.isoutput = isoutput


        self.bias = nn.Parameter(th.Tensor(out_features))
        self.mu_w = nn.Parameter(th.Tensor(out_features, in_features))
        self.logsig2_w = nn.Parameter(th.Tensor(out_features, in_features))
        self.reset_parameters()

        self.normal = False
        self.reset_parameters()

    def reset_parameters(self):
        stdv = 1. / math.sqrt(self.mu_w.size(1))
        self.mu_w.data.normal_(0, stdv)
        self.logsig2_w.data.zero_().normal_(-9, 0.001)
        self.bias.data.zero_()

    def forward(self, data):
        return F.linear(data, self.mu_w, self.bias)

    def KL(self, loguniform=False):
        if loguniform:
            k1 = 0.63576; k2 = 1.87320; k3 = 1.48695
            log_alpha = self.logsig2_w - 2 * th.log(self.mu_w.abs() + 1e-8)
            kl = -th.sum(k1 * F.sigmoid(k2 + k3 * log_alpha) - 0.5 * F.softplus(-log_alpha) - k1)
        else:
            logsig2_w = self.logsig2_w.clamp(-11, 11)
            kl = 0.5 * (self.prior_prec * (self.mu_w.pow(2) + logsig2_w.exp()) - logsig2_w - 1 - np.log(self.prior_prec)).sum()
        return kl

    def var(self, prev_mean, prev_var=None, C=100):
        if self.isoutput:
            m2s2_w = self.mu_w.pow(2) + self.logsig2_w.exp()
            term1 = F.linear(prev_var, m2s2_w)
            term2 = F.linear(prev_mean.pow(2), self.logsig2_w.exp())
            return term1 + term2

        else:
            pZ = th.sigmoid(C * F.linear(prev_mean, self.mu_w, self.bias))

            # Compute var[h]
            if prev_var is None:
                term1 = 0
            else:
                m2s2_w = self.mu_w.pow(2) + self.logsig2_w.exp()
                term1 = F.linear(prev_var, m2s2_w)
            term2 = F.linear(prev_mean.pow(2), self.logsig2_w.exp())
            varh = term1 + term2

            # Compute E[h]^2
            term3 = F.linear(prev_mean, self.mu_w, self.bias).pow(2)

            return pZ * varh + pZ * (1 - pZ) * term3

    def __repr__(self):
        return self.__class__.__name__ + ' (' \
               + str(self.n_in) + ' -> ' \
               + str(self.n_out) \
               + f', isoutput={self.isoutput})'


class VBPConv(VBPLinear):
    def __init__(self, in_channels, out_channels, kernel_size, prior_prec=10, stride=1,
                 padding=0, dilation=1, groups=1, isoutput=False):
        super(VBPLinear, self).__init__()
        self.n_in = in_channels
        self.n_out = out_channels

        self.kernel_size = kernel_size
        self.stride = _pair(stride)
        self.padding = _pair(padding)
        self.dilation = _pair(dilation)
        self.groups = groups

        self.prior_prec = prior_prec
        self.isoutput = isoutput

        self.bias = nn.Parameter(th.Tensor(out_channels))
        self.mu_w = nn.Parameter(th.Tensor(out_channels, in_channels, kernel_size, kernel_size))
        self.logsig2_w = nn.Parameter(th.Tensor(out_channels, in_channels, kernel_size, kernel_size))
        self.reset_parameters()


    def reset_parameters(self):
        n = self.n_in
        for k in range(1, self.kernel_size):
            n *= k
        self.mu_w.data.normal_(0, 1. / math.sqrt(n))
        self.logsig2_w.data.zero_().normal_(-9, 0.001)
        self.bias.data.zero_()

    def forward(self, data):
        return F.conv2d(data, self.mu_w, self.bias, self.stride, self.padding, self.dilation, self.groups)

    def var(self, prev_mean, prev_var=None, C=100):
        if self.isoutput:
            m2s2_w = self.mu_w.pow(2) + self.logsig2_w.exp()
            term1 = F.conv2d(prev_var, m2s2_w, None, self.stride, self.padding, self.dilation, self.groups)
            term2 = F.conv2d(prev_mean.pow(2), self.logsig2_w.exp(), None, self.stride, self.padding, self.dilation, self.groups)
            return term1 + term2

        else:
            pZ = th.sigmoid(C*F.conv2d(prev_mean, self.mu_w, self.bias, self.stride, self.padding, self.dilation, self.groups))

            # Compute var[h]
            if prev_var is None:
                term1 = 0
            else:
                m2s2_w = self.mu_w.pow(2) + self.logsig2_w.exp()
                term1 = F.conv2d(prev_var, m2s2_w, None, self.stride, self.padding, self.dilation, self.groups)
            term2 = F.conv2d(prev_mean.pow(2), self.logsig2_w.exp(), None, self.stride, self.padding, self.dilation, self.groups)
            varh = term1 + term2


            # Compute E[h]^2
            term3 = F.conv2d(prev_mean, self.mu_w, self.bias, self.stride, self.padding, self.dilation, self.groups).pow(2)

            return pZ * varh + pZ * (1 - pZ) * term3

    def __repr__(self):
        s = ('{name}({n_in}, {n_out}, kernel_size={kernel_size}'
             ', stride={stride}')
        if self.padding != (0,) * len(self.padding):
            s += ', padding={padding}'
        if self.dilation != (1,) * len(self.dilation):
            s += ', dilation={dilation}'
        if self.groups != 1:
            s += ', groups={groups}'
        s += ', bias=True'
        s += ')'

        return s.format(name=self.__class__.__name__, **self.__dict__)