Closed uselessai closed 2 years ago
Hi @uselessai You will first need to convert your pickle to a pt file using the code snippet provided above. Please try this out and let me know if this solves your issue. I'll add a comment in the README to make this clearer.
Hi @yuval-alaluf . Thanks for the quick response.
I tried to convert the stylegan3 pkl filo to pt using the links provided in ReStyle repository but I got an error, It only works for Stylegan2 models. I can not find the .pt generators in your repository.
Thanks in advance. Lau.
Converting the pkls to pt for SG3 is different than what was used for SG2. Try using the snippet I provided here: https://github.com/yuval-alaluf/stylegan3-editing/issues/16#issuecomment-1101533474
@uselessai Hello,I am also training my own restyle_e4e model, but when I use the SG3-R-FFHQ-102.pt model provided by the author, which is trained using the ffhq data set, the code is reporting an error. https://github.com/yuval-alaluf/stylegan3-editing/issues/53 I would like to ask if you successfully solved the error after converting pkl file to pt file? Because I now want to try to use the Stylegan3-related.PKL model file provided by NVIDIA officially, instead of the.pt file converted by the author, I would like to ask if it is meaningful for me to try this?
I am trying to train the encoder "train_restyle_psp" with my own stylegan3 pkl model and I've got this error.
Loading StyleGAN3 generator from path: pretrained_models/network-snapshot-002160Stylegan3.pkl Done! Traceback (most recent call last): File "inversion/scripts/train_restyle_psp.py", line 29, in <module> main() File "/home/laura/Escritorio/Anaconda/lib/python3.8/site-packages/pyrallis/argparsing.py", line 158, in wrapper_inner response = fn(cfg, *args, **kwargs) File "inversion/scripts/train_restyle_psp.py", line 24, in main coach = Coach(opts) File "./inversion/training/coach_restyle_psp.py", line 44, in __init__ self.avg_image = self.net(self.net.latent_avg.repeat(16, 1).unsqueeze(0), File "/home/laura/.local/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1102, in _call_impl return forward_call(*input, **kwargs) File "./inversion/models/psp3.py", line 66, in forward images = self.decoder.synthesis(codes, noise_mode='const', force_fp32=True) File "/home/laura/.local/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1102, in _call_impl return forward_call(*input, **kwargs) File "<string>", line 469, in forward File "/home/laura/.local/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1102, in _call_impl return forward_call(*input, **kwargs) File "<string>", line 218, in forward RuntimeError: batch1 dim 2 must match batch2 dim 1
I am working with a pkl model.