pettingzoo 0.1.5

Gym for multi-agent reinforcement learning

PettingZoo

PettingZoo is Python library for conducting research in multi-agent reinforcement learning. It's akin to a multi-agent version of OpenAI's Gym library.

We model environments as Agent Environment Cycle (AEC) games, in order to be able to support all types of multi-agent RL environments under one API.

Environment Types and Installation

PettingZoo includes the following sets of games:

• atari: Multi-player Atari 2600 games (both cooperative and competitive)
• classic: Classical, nongraphical, competitive games (i.e. chess, Texas hold 'em, and go)
• gamma: Cooperative graphical games developed by us. Policies for these must learn very coordinated behaviors.
• magent: Environments with massive numbers of particle agents, originally from https://github.com/geek-ai/MAgent
• mpe: A set of simple nongraphical communication tasks, originally from https://github.com/openai/multiagent-particle-envs
• sisl: 3 cooperative environments, originally from https://github.com/sisl/MADRL

To install, use pip install pettingzoo

We support Python 3.6, 3.7 and 3.8