# rlworkgroup/garage

A toolkit for reproducible reinforcement learning research.

Repository: https://github.com/rlworkgroup/garage
Canonical: https://ross.abutalabs.com/products/rlworkgroup-garage
Language: Python
License: MIT
License Family: permissive
Topics: rl-algorithms, reproducibility, pytorch, tensorflow
Last push: 2023-05-04T14:44:22+00:00

## Health v2 (maintenance only)
Score: 23/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 8, longevity 100
- inputs: {"age_days": 3006, "days_push": 1217, "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 2126, forks 328 (observed 2026-08-28T04:06:16.880104+00:00)

## What it is
garage is a Python toolkit for developing and evaluating reinforcement learning algorithms, with a library of state-of-the-art RL algorithm implementations. It provides modular components like replay buffers, samplers, experiment checkpointing, and reproducibility tools supporting both PyTorch and TensorFlow.

## Use cases
- implement and benchmark reinforcement learning algorithms
- run reproducible RL experiments with fixed random seeds
- train PPO, DDPG, DQN, SAC and other RL agents
- evaluate RL algorithms on Gym and other benchmark environments
- checkpoint and resume long-running RL training runs
- log RL experiment metrics to TensorBoard

## When to choose
- you need reproducible RL research with experiment checkpointing and seeding
- you want ready implementations of many standard RL algorithms in one library
- you want framework flexibility between PyTorch and TensorFlow
- you need high-performance samplers and environment interfaces for RL benchmarks

## When to avoid
- you need cutting-edge algorithms actively updated with the latest research
- you want a lightweight single-algorithm library rather than a full toolkit
- your project requires frequent releases or active maintenance
- you need non-Python or production RL deployment tooling

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, reinforcement-learning, benchmarking, logging
- domain: reinforcement-learning, machine-learning, artificial-intelligence
- platform: python
- tags: reinforcement-learning, pytorch, tensorflow, reproducibility, experiment-tracking, rl-algorithms, research, linux, macos, docker

## Member repositories
- rlworkgroup/garage (main) score 23

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:16.880104+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 2026-08-30T02:52:32.126442+00:00, confidence not recorded.
  - readme: https://github.com/rlworkgroup/garage (fetched 2026-08-28T04:06:16.880104+00:00, sha 7f69dfc2671d)
  - registry_pypi: https://pypi.org/pypi/garage/json (fetched 2026-08-29T10:32:35.955204+00:00, sha 80efa7b3add1)
- Data as of 2026-08-30T08:39:29.467469+00:00.