# starry-sky6688/MARL-Algorithms

Implementations of IQL, QMIX, VDN, COMA, QTRAN, MAVEN, CommNet, DyMA-CL, and G2ANet on SMAC,  the decentralised micromanagement scenario of StarCraft II

Repository: https://github.com/starry-sky6688/MARL-Algorithms
Canonical: https://ross.abutalabs.com/products/marl-algorithms
Language: Python
License Family: other
Topics: multi-agent-reinforcement-learning, deep-reinforcement-learning, reinforcement-learning
Last push: 2022-09-08T08:53:29+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": 2596, "days_push": 1455, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1760, forks 302 (observed 2026-08-28T04:05:32.576510+00:00)

## What it is
PyTorch implementations of classic multi-agent reinforcement learning algorithms (IQL, QMIX, VDN, COMA, QTRAN, MAVEN, CommNet, DyMA-CL, G2ANet) evaluated on SMAC, the StarCraft II micromanagement benchmark. It is research code for training and evaluating cooperative MARL algorithms.

## Use cases
- implement qmix multi-agent reinforcement learning
- benchmark marl algorithms on smac starcraft ii
- learn value decomposition methods like vdn and qmix
- reproduce multi-agent rl research paper results
- train cooperative multi-agent policies in pytorch
- compare communication-based marl algorithms like commnet and g2anet

## When to choose
- you want ready PyTorch implementations of many classic MARL algorithms in one codebase
- you need a benchmark on SMAC/StarCraft II micromanagement scenarios
- you are studying or reproducing cooperative multi-agent RL papers

## When to avoid
- you need a maintained, production-grade MARL framework with a stable API
- you require a permissive license - the repo has no license
- you want environments other than StarCraft II/SMAC
- you need the latest SOTA MARL algorithms - the TODO list is unfinished

## Facets
- artifact type: library
- maturity: maintenance
- function: reinforcement-learning, machine-learning, simulation
- domain: reinforcement-learning, machine-learning
- platform: python, windows
- tags: multi-agent-reinforcement-learning, marl, starcraft-ii, smac, pytorch, qmix, value-decomposition, research-code, game-development, linux, macos

## Member repositories
- starry-sky6688/MARL-Algorithms (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:32.576510+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 2026-08-30T03:27:15.083418+00:00, confidence not recorded.
  - readme: https://github.com/starry-sky6688/MARL-Algorithms (fetched 2026-08-28T04:05:32.576510+00:00, sha 9e09b0ce4aef)
- Data as of 2026-08-30T08:39:29.467469+00:00.