A toolkit for developing and comparing reinforcement learning algorithms.
Issues are used to track tasks, bugs, and feature requests.