Reinforcement Learning to Play Doom

A project to apply reinforcement learning to play Doom. Uses StableBaselines3 implementations of RL algorithms on ViZDoom using the ViZDoomGym wrapper. Key takeaway: reward shaping is key!