Deep RL for Motion Imitation

This project uses deep reinforcement learning to train a control policy for a physically simulated humanoid character. The character is simulated using the Isaac Gym physics engine. The goal is to train the character to imitate a reference motion, similar to DeepMimic.