New encoder
Added new encoder and deleted the earlier branch to prevent Mac artifacts like MacOSX and .DS_Store
Got you, thanks @rosasalberto , I'll work on these over the weekend and resubmit
so Nvidia labs will be using pytorch from now on - there's never going to be an official tensorflow 2 - stylegan2 library. https://github.com/NVlabs/stylegan2-ada/issues/32
when I look at their code and derivative works - there's a lot of people who have implemented things on top of the dnn folder. Is it possible to just auto upgrade it automagically - or is your approach / complete rewrite better? I'm stuck on latest 3090 card and can't get code to run without docker.
I guess - what I'm wanting to say is - can you help port the nvidia labs code to tensorflow2? I want to use existing trained pickels - this code does a lot to convert tensorflow pickels - to pytorch. maybe helpful - https://github.com/GreenLimeSia/GenEdi
UPDATE I created / cherry picked a v2 compatibility branch here https://github.com/johndpope/stylegan2-ada
https://github.com/johndpope/stylegan2-ada/tree/digressions