Discriminator

Overall Progress 0%

Discriminator expert vs agent; use as reward for policy gradient.

Go to Chapter 77: Generative Adversarial Imitation Learning (GAIL) →