why discriminator is designed like this?

aitorzip / PyTorch-CycleGAN

A clean and readable Pytorch implementation of CycleGAN

https://arxiv.org/abs/1703.10593

GNU General Public License v3.0

1.22k stars 283 forks source link

why discriminator is designed like this? #10

Open YunLord opened 5 years ago

YunLord commented 5 years ago

 loss_D_real = criterion_GAN(pred_real, torch.ones_like(pred_real));i think this should be      loss_D_real = criterion_GAN(pred_real, torch.zeros_like(pred_real))

FloCF commented 5 years ago

Not sure why you think that. At least that is the usual way to train GANs, check out https://pytorch.org/tutorials/beginner/dcgan_faces_tutorial.html