PyGame-Learning-Environment

Why we included this project

PLE mirrors the Arcade Learning Environment interface but stays lightweight: just Python, Pygame, and a handful of small arcade games like Pong and Flappy Bird. The point is to skip environment plumbing so you can plug in your own agent and keep the focus on model design and experiments. It exposes the standard reset, step, and reward calls, so swapping in a custom agent takes minutes rather than a week of wiring. Because the project accepts community-submitted games and the codebase is small enough to read through, it works well for coursework, research prototypes, and simple benchmark harnesses. It is no longer actively maintained, so treat it as a stable learning tool rather than production infrastructure.