Some reinforcement learning algorithms I'm (re)-implementing, all in one place. Also is a dumping ground for other ML work.