You have an environment, a PyTorch model, and a reinforcement learning framework that are designed to work together but don’t. PufferLib is a wrapper layer that makes RL on complex game environments as simple as RL on Atari. You write a native PyTorch network and a short binding for your environment; PufferLib takes care of the rest.
All of our Documentation is hosted by github.io. @jsuarez5341 on Discord for support -- post here before opening issues. I am also looking for contributors interested in adding bindings for other environments and RL frameworks.