JaxMARL

Installation | Quick Start | Environments | Algorithms | Citation

Multi-Agent Reinforcement Learning in JAX

🎉 Update: JaxMARL was accepted at NeurIPS 2024 on Datasets and Benchmarks Track. See you in Vacouver!

JaxMARL combines ease-of-use with GPU-enabled efficiency, and supports a wide range of commonly used MARL environments as well as popular baseline algorithms. Our aim is for one library that enables thorough evaluation of MARL methods across a wide range of tasks and against relevant baselines. We also introduce SMAX, a vectorised, simplified version of the popular StarCraft Multi-Agent Challenge, which removes the need to run the StarCraft II game engine.

For more details, take a look at our blog post or our Colab notebook, which walks through the basic usage.

Environments 🌍

Environment	Reference	README	Summary
🔴 MPE	Paper	Source	Communication orientated tasks in a multi-agent particle world
🍲 Overcooked	Paper	Source	Fully-cooperative human-AI coordination tasks based on the homonyms video game
🦾 Multi-Agent Brax	Paper	Source	Continuous multi-agent robotic control based on Brax, analogous to Multi-Agent MuJoCo
🎆 Hanabi	Paper	Source	Fully-cooperative partially-observable multiplayer card game
👾 SMAX	Novel	Source	Simplified cooperative StarCraft micro-management environment
🧮 STORM: Spatial-Temporal Representations of Matrix Games	Paper	Source	Matrix games represented as grid world scenarios
🧭 JaxNav	Paper	Source	2D geometric navigation for differential drive robots
🪙 Coin Game	Paper	Source	Two-player grid world environment which emulates social dilemmas
💡 Switch Riddle	Paper	Source	Simple cooperative communication game included for debugging

Baseline Algorithms 🦉

We follow CleanRL's philosophy of providing single file implementations which can be found within the baselines directory. We use Hydra to manage our config files, with specifics explained in each algorithm's README. Most files include wandb logging code, this is disabled by default but can be enabled within the file's config.

Algorithm	Reference	README
IPPO	Paper	Source
MAPPO	Paper	Source
IQL	Paper	Source
VDN	Paper	Source
QMIX	Paper	Source
TransfQMIX	Paper	Source
SHAQ	Paper	Source
PQN-VDN	Paper	Source

Installation 🧗

Environments - Before installing, ensure you have the correct JAX installation for your hardware accelerator. We have tested up to JAX version 0.4.25. The JaxMARL environments can be installed directly from PyPi:

pip install jaxmarl

Algorithms - If you would like to also run the algorithms, install the source code as follows:

Clone the repository:

git clone https://github.com/FLAIROx/JaxMARL.git && cd JaxMARL

Install requirements:

pip install -e .[algs]
export PYTHONPATH=./JaxMARL:$PYTHONPATH

For the fastest start, we reccoment using our Dockerfile, the usage of which is outlined below.

Development - If you would like to run our test suite, install the additonal dependencies with: pip install -e .[dev], after cloning the repository.

Quick Start 🚀

We take inspiration from the PettingZoo and Gymnax interfaces. You can try out training an agent in our Colab notebook. Further introduction scripts can be found here.

Basic JaxMARL API Usage 🖥️

Actions, observations, rewards and done values are passed as dictionaries keyed by agent name, allowing for differing action and observation spaces. The done dictionary contains an additional "__all__" key, specifying whether the episode has ended. We follow a parallel structure, with each agent passing an action at each timestep. For asynchronous games, such as Hanabi, a dummy action is passed for agents not acting at a given timestep.

import jax
from jaxmarl import make

key = jax.random.PRNGKey(0)
key, key_reset, key_act, key_step = jax.random.split(key, 4)

# Initialise environment.
env = make('MPE_simple_world_comm_v3')

# Reset the environment.
obs, state = env.reset(key_reset)

# Sample random actions.
key_act = jax.random.split(key_act, env.num_agents)
actions = {agent: env.action_space(agent).sample(key_act[i]) for i, agent in enumerate(env.agents)}

# Perform the step transition.
obs, state, reward, done, infos = env.step(key_step, state, actions)

Dockerfile 🐋

To help get experiments up and running we include a Dockerfile and its corresponding Makefile. With Docker and the Nvidia Container Toolkit installed, the container can be built with:

make build

The built container can then be run:

make run

Contributing 🔨

Please contribute! Please take a look at our contributing guide for how to add an environment/algorithm or submit a bug report. Our roadmap also lives there.

Citing JaxMARL 📜

If you use JaxMARL in your work, please cite us as follows:

@article{flair2023jaxmarl,
      title={JaxMARL: Multi-Agent RL Environments in JAX},
      author={Alexander Rutherford and Benjamin Ellis and Matteo Gallici and Jonathan Cook and Andrei Lupu and Gardar Ingvarsson and Timon Willi and Akbir Khan and Christian Schroeder de Witt and Alexandra Souly and Saptarashmi Bandyopadhyay and Mikayel Samvelyan and Minqi Jiang and Robert Tjarko Lange and Shimon Whiteson and Bruno Lacerda and Nick Hawes and Tim Rocktaschel and Chris Lu and Jakob Nicolaus Foerster},
      journal={arXiv preprint arXiv:2311.10090},
      year={2023}
    }

FLAIROx/JaxMARL