# google-deepmind/acme

A library of reinforcement learning components and agents

Repository: https://github.com/google-deepmind/acme
Canonical: https://ross.abutalabs.com/products/acme
Language: Python
License: Apache-2.0
License Family: permissive
Topics: reinforcement-learning, agents, research
Last push: 2026-08-25T11:52:57+00:00

## Health v2 (maintenance only)
Score: 67/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 99, release rhythm 8, longevity 100
- inputs: {"age_days": 2315, "days_push": 8, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 4050, forks 550 (observed 2026-08-28T04:08:33.816206+00:00)

## What it is
Acme is a Python library of reinforcement learning building blocks and agent implementations from DeepMind. It provides readable reference agents and flexible components that scale from single-stream to distributed training.

## Use cases
- implement reinforcement learning agents in python
- run distributed RL training experiments
- get reference implementations of RL algorithms
- build novel RL research on reusable components
- benchmark RL algorithm baselines
- train agents in gym and other environments

## When to choose
- you are doing RL research and want strong, flexible baseline agents
- you need reference implementations of RL algorithms
- you want agents that scale from single-machine to distributed setups

## When to avoid
- you need a production-ready, turnkey RL product rather than a research framework
- you want a beginner-friendly tutorial library with guaranteed API stability
- you are not working in Python

## Facets
- artifact type: library
- maturity: active
- function: reinforcement-learning, machine-learning, agent-framework
- domain: reinforcement-learning, machine-learning
- platform: python
- tags: rl-agents, research-framework, distributed-rl, deepmind, research, linux, macos

## Member repositories
- google-deepmind/acme (main) score 67

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:08:33.816206+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 2026-08-29T18:23:40.731981+00:00, confidence not recorded.
  - readme: https://github.com/google-deepmind/acme (fetched 2026-08-28T04:08:33.816206+00:00, sha 5e2ed9736abb)
- Data as of 2026-08-30T08:39:29.467469+00:00.