two question invertor and FFHQ training

Hi, thanks for nice works.

I am really interesting your works. Did you have FFHQmask pretrained model instead of CelebA? If you had, plz update it. thanks.

For invertor, you are using PTI, right?, Instead of it. can I use other invertor? If I can, which info  do I need? 512x512 size, definitely, and latent code as well. w+ space? but PTI is not w+, right? 
