Vectorize and Parallelize RL Environments with JAX: Q-learning at the Speed of Light⚡ | Towards Data Science

Learn to vectorize a GridWorld environment and train 30 Q-learning agents in parallel on a CPU, at 1.8 million step per seconds!