Generative_Adversarial_Networks_PyTorch

Generative_Adversarial_Networks_PyTorch copied to clipboard

can you add more comments on LAPGAN of class CondiGAN_Discriminator,i could not understand that part according to the paper.

xuzhongyou

The two for loops in the minibatch discrimination part of the improvedGAN increase the running time by several magnitudes - is there a faster way? Thanks!

Priya22

Question about n_discrete

1

hello @AaronYALai in run_InfoGAN.py, I can't figure out the parameter n_discrete 's meaning.I'm looking forward to your reply.Thanks.

zhou-frank

hello @AaronYALai , in run_InfoGAN.py, 'D_loss=D_criterion(outputs[:,0],labels)' is the mean of 2*batchSize loss between outputs and labels, but I think D_loss=truth_input_loss+fake_input_loss, that should be the mean of 1*batchSize loss. I'm looking...

ran337287