marl 开源项目推荐（203 个）

pytorch /rl

一个模块化、原语优先、Python 优先的 PyTorch 强化学习库。

instadeepai /Mava

🦁 A research-friendly codebase for fast experimentation of multi-agent reinforcement learning in JAX

bold-lab-ai /JaxMARL

Multi-Agent Reinforcement Learning with JAX

hijkzzz /pymarl2

Fine-tuned MARL algorithms on SMAC (100% win rates on most scenarios)

facebookresearch /BenchMARL

BenchMARL is a library for benchmarking Multi-Agent Reinforcement Learning (MARL). BenchMARL allows to quickly compare different MARL algorithms, tasks, and models while being systematically grounded in its two core tenets: reproducibility and standardization.

proroklab /VectorizedMultiAgentSimulator

VMAS is a vectorized differentiable simulator designed for efficient Multi-Agent Reinforcement Learning benchmarking. It is comprised of a vectorized 2D physics engine written in PyTorch and a set of challenging multi-robot scenarios. Additional scenarios can be implemented through a simple and modular interface.

Ronchy2000 /Multi-agent-RL

Projects from basic algorithms to MARL. Implements MADDPG,MATD3,MA/HAPPO in Predator-Prey pursuit games with PettingZoo MPE environments.

TimeBreaker /MARL-papers-with-code

Multi-Agent Reinforcement Learning (MARL) papers with code

TimeBreaker /Multi-Agent-Reinforcement-Learning-papers

Multi-Agent Reinforcement Learning (MARL) papers

Cognitive-AI-Systems /pogema

[ICLR-2025]POGEMA 代表多代理的部分可观测网格环境。这是一个基于网格的环境，专门设计为灵活、可调和可扩展。它可以根据各种PO-MAPF或MAPF设置进行定制。

TimeBreaker /MARL-resources-collection

A Collection of Multi-Agent Reinforcement Learning (MARL) Resources

OpenMLRL /CoMLRL

Open-Source Library for Cooperative Multi-LLM Reinforcement Learning

mikelma /craftium

A framework for creating rich, 3D, Minecraft-like single and multi-agent environments for AI research. (Accepted at ICML 2025).

ChuaCheowHuan /gym-continuousDoubleAuction

A custom MARL (multi-agent reinforcement learning) environment where multiple agents trade against one another (self-play) in a zero-sum continuous double auction. Ray [RLlib] is used for training.

hsvgbkhgbv /SQDDPG

这是一个用于多智能体强化学习研究及实验实施的框架，题为《Shapley Q值：局部奖励方法解决全局奖励博弈》。

Lauqz /Drone-Swarm-RL-airsim-sb3

Training of Drone Swarms using StableBaselines3, PettingZoo, AirSim and UE4

zhaozijie2022 /dynamic-coverage-control

无人机动态覆盖控制；1. 实现了一个无人机点覆盖环境；2. 给出了无人机连通保持规则；3. 给出了基于MARL的控制算法

lich14 /CDS

[NeurIPS 2021] CDS achieves remarkable success in challenging benchmarks SMAC and GRF by balancing sharing and diversity.

instadeepai /marl-eval

A tool for aggregating and plotting MARL experiment data.

AlirezaShamsoshoara /Reinforcement_Learning_Team_Q_learnig_MARL_Multi_Agent_UAV_Spectrum_task

A solution for Dynamic Spectrum Management in Mission-Critical UAV Networks using Team Q learning as a Multi-Agent Reinforcement Learning Approach