Reinforcement Learning Game

In this project an agent learns on its own to find its way through mazes and around obstacles.

The game is implemented as a Gymnasium environment in Python. Rewards, observations and level layouts can be adjusted, which makes it easy to compare how different settings affect the training.