This is a simple yet efficient, highly customizable grid-world implementation to run reinforcement learning algorithms.
