Policies trained from data instead of hand-written controllers.
GPU-accelerated robot learning framework built on NVIDIA Isaac Sim.
Fast and simple PPO implementation for legged robot learning.
