# pfnet/pfrl

PFRL: a PyTorch-based deep reinforcement learning library

Repository: https://github.com/pfnet/pfrl
Canonical: https://ross.abutalabs.com/products/pfrl
Language: Python
License: MIT
License Family: permissive
Last push: 2026-03-02T04:48:45+00:00

## Health v2 (maintenance only)
Score: 54/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 70, release rhythm 8, longevity 100
- inputs: {"age_days": 2261, "days_push": 184, "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 1274, forks 158 (observed 2026-08-28T04:04:12.778374+00:00)

## What it is
PFRL is a PyTorch-based deep reinforcement learning library implementing state-of-the-art algorithms such as DQN, Rainbow, PPO, SAC, and A3C. It includes reproducibility scripts and pretrained model zoos for Atari and MuJoCo benchmarks.

## Use cases
- train deep RL agents on Atari games
- train continuous control policies with SAC or TD3 on MuJoCo
- reproduce published reinforcement learning benchmark results
- experiment with Rainbow and IQN algorithms in PyTorch
- use pretrained RL models as a starting point
- run asynchronous A3C training on CPU

## When to choose
- you want a PyTorch RL library with many classic algorithms implemented
- you need reproducible Atari or MuJoCo benchmark scripts and pretrained models
- you want both discrete and continuous action support with recurrent model options

## When to avoid
- you need the newest RL algorithms or active ecosystem like Stable-Baselines3 or RLlib
- you want CPU-asynchronous training for most algorithms (only A3C/ACER support it)
- you need non-PyTorch frameworks such as TensorFlow

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, reinforcement-learning, sdk
- domain: reinforcement-learning, machine-learning, deep-learning
- platform: python
- tags: pytorch, deep-reinforcement-learning, dqn, ppo, sac, a3c, atari, gym, model-zoo, linux, macos, gpu

## Member repositories
- pfnet/pfrl (main) score 54

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:12.778374+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 2026-08-30T05:03:08.694730+00:00, confidence not recorded.
  - readme: https://github.com/pfnet/pfrl (fetched 2026-08-28T04:04:12.778374+00:00, sha fb6c2bad6b82)
- Data as of 2026-08-30T08:39:29.467469+00:00.