# oxwhirl/pymarl

Python Multi-Agent Reinforcement Learning framework

Repository: https://github.com/oxwhirl/pymarl
Canonical: https://ross.abutalabs.com/products/pymarl
Language: Python
License: Apache-2.0
License Family: permissive
Last push: 2022-12-08T02:58:39+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 2869, "days_push": 1364, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2216, forks 411 (observed 2026-08-28T04:06:27.176879+00:00)

## What it is
PyMARL is WhiRL's PyTorch framework for deep multi-agent reinforcement learning, implementing algorithms such as QMIX, COMA, VDN, IQL, and QTRAN. It uses the SMAC StarCraft II environment for cooperative multi-agent experiments and supports Docker-based setup, model saving/loading, and replay generation.

## Use cases
- run multi-agent reinforcement learning experiments
- reproduce QMIX results on SMAC
- train cooperative multi-agent policies in StarCraft II
- compare value-decomposition MARL algorithms
- benchmark deep RL algorithms for multi-agent settings
- load and evaluate pretrained MARL checkpoints
- watch StarCraft II replays of trained agents

## When to choose
- you need reference implementations of classic MARL algorithms like QMIX or COMA
- your research uses the SMAC StarCraft II benchmark
- you want a PyTorch codebase for multi-agent RL experiments

## When to avoid
- you need a maintained framework with recent updates or new algorithm support
- you want environments other than StarCraft II without extra integration work
- you need production deployment rather than research experimentation

## Facets
- artifact type: framework
- maturity: maintenance
- function: reinforcement-learning, machine-learning, agent-framework
- domain: reinforcement-learning, machine-learning, artificial-intelligence, gaming-tools
- platform: python
- tags: multi-agent-reinforcement-learning, marl, qmix, coma, vdn, qtran, smac, starcraft-ii, pytorch, research, linux, docker, gpu

## Member repositories
- oxwhirl/pymarl (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:27.176879+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T02:45:57.420114+00:00, confidence not recorded.
  - readme: https://github.com/oxwhirl/pymarl (fetched 2026-08-28T04:06:27.176879+00:00, sha cffb231b9748)
- Data as of 2026-08-30T08:39:29.467469+00:00.