search

alpc91 / NICE-GAN-pytorch

Official PyTorch implementation of NICE-GAN: Reusing Discriminators for Encoding: Towards Unsupervised Image-to-Image Translation
https://alpc91.github.io/NICE-GAN-pytorch
MIT License
235 stars 41 forks source link

question about NAN #27

Open yan-yuchen opened 2 years ago

yan-yuchen commented 2 years ago

When I train the model with learning rate:0.0001,after several echos,I got "d_loss: nan, g_loss: nan" ,how can I deal with this situation? Thanks a lot!

TMDTom commented 2 months ago

When I train the model with learning rate:0.0001,after several echos,I got "d_loss: nan, g_loss: nan" ,how can I deal with this situation? Thanks a lot!

I meet the same question.I guess maybe the generator is ovefitting or something. You can use the saved checkpoint to evaluate your data.