goal-conditioned-rl
Random maze environments with different size and complexity for reinforcement learning and planning research. This is in-particular to investigate generalization and planning ability in dynamically changing environment.
