train_gan.py

#To train the Discriminator
output_d_real = discriminator(real_images)
d_real_loss = criterion(output_d_real, real_labels)

z = torch.randn(batch_size, random_size).to(device)
fake_images = generator(z)
output_d_fake = discriminator(fake_images)
d_fake_loss = criterion(output_d_fake, fake_labels)
d_loss = d_real_loss + d_fake_loss

optimizer_d.zero_grad()
d_loss.backward()
optimizer_d.step()#this is going to update only parameters of discriminator

#to train the generator
# Input to generator is a noise of size random_size
z = torch.randn(batch_size, random_size) 
output_image = generator(z)
output_discriminator = discriminator(output_image)

#to train the generator the output of this should be compared with real_labels. 
#so we compare the output by real label. 

g_loss = criterion(outputs, real_labels)

optimizer_g.zero_grad()
g_loss.backward()
optimizer_g.step()