real image in save results looks strange

[uyo9ko]

in the train_vqgan.py, the code real_fake_images = torch.cat((imgs[:4], decoded_images.add(1).mul(0.5)[:4])) should be revised real_fake_images = torch.cat((imgs.add(1).mul(0.5)[:4], decoded_images.add(1).mul(0.5)[:4]))

[Ukuer]

change

real_fake_images = torch.cat((imgs[:4], decoded_images.add(1).mul(0.5)[:4]))

to

real_fake_images = torch.cat((imgs.add(1).mul(0.5)[:4], decoded_images.add(1).mul(0.5)[:4]))

[clearlyzero]

Hello, I have also been researching vqgan recently. May I ask if your network has good convergence performance.