Deep Learning with PyTorch

Defining loss and optimizer

We will define a binary cross-entropy loss and two optimizers, one for the generator and another one for the discriminator, in the following code:

criterion = nn.BCELoss()# setup optimizeroptimizerD = optim.Adam(netD.parameters(), lr, betas=(beta1, 0.999))optimizerG = optim.Adam(netG.parameters(), lr, betas=(beta1, 0.999))

Up to this point, it is very similar to what we have seen in all our previous examples. Let's explore how we can train the generator and discriminator.