Classic Reinforcement Algorithms

Classic Reinforcement Algorithms (CRA) is a collection of classic reinforcement learning algorithms implementation, mainly based on this repository. The difference is this reposoitory used Gymnasium (a maintained fork of OpenAI’s Gym library) instead of Gym.

Supported Algorithms

• DQN (paper)

• CQL (paper)

• MADDPG (paper)

Installation

To install Gymnasium, use this:

  pip install gymnasium==0.28.1

Or check its website.

License

MIT