Skip to content
ml-recall

#generator-loss

2 questions from Generative Adversarial Networks

Timed test

2 questions · 2 min · score and explanations at the end

Take a test
Q 1 / 2Score 0 / 0

Question 1: Single choice

mediumAI-checkedGenerative Adversarial Networks

A GAN is trained with these helpers, where real_loss compares the discriminator's outputs with label 1 and fake_loss with label 0:

fake_images = G(z)
D_fake = D(fake_images)
g_loss = ???(D_fake)
g_loss.backward(); g_optimizer.step()

Which call belongs in place of ??? for the generator step?

Choose one option