# Kaixhin/Rainbow

Rainbow: Combining Improvements in Deep Reinforcement Learning

Repository: https://github.com/Kaixhin/Rainbow
Canonical: https://ross.abutalabs.com/products/kaixhin-rainbow
Language: Python
License: MIT
License Family: permissive
Topics: deep-learning, deep-reinforcement-learning
Last push: 2022-01-13T01:24:38+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": 3250, "days_push": 1694, "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 1673, forks 294 (observed 2026-08-28T04:05:20.271039+00:00)

## What it is
A PyTorch implementation of the Rainbow deep reinforcement learning agent, combining DQN improvements like prioritized experience replay, dueling networks, distributional RL, and noisy nets. It trains on Atari games and ships pretrained models in its releases.

## Use cases
- train a Rainbow DQN agent on Atari games
- reproduce deep reinforcement learning research results
- compare DQN algorithm variants like double DQN and dueling networks
- run data-efficient reinforcement learning experiments
- download and evaluate pretrained Atari RL models
- learn how distributional RL and noisy nets are implemented in PyTorch

## When to choose
- you want a reference implementation of the Rainbow paper in PyTorch
- you need pretrained Atari RL agents for benchmarking
- you are studying or extending DQN-family algorithms

## When to avoid
- you need a maintained general-purpose RL library with many environments
- you want production RL training infrastructure
- you need environments beyond Atari without modification

## Facets
- artifact type: library
- maturity: maintenance
- function: reinforcement-learning, deep-learning, machine-learning
- domain: reinforcement-learning, machine-learning, deep-learning
- platform: python
- tags: dqn, atari, pytorch, rainbow, research-code, pretrained-models, linux, macos

## Member repositories
- Kaixhin/Rainbow (main) score 23

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:20.271039+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 2026-08-30T03:42:34.556265+00:00, confidence not recorded.
  - readme: https://github.com/Kaixhin/Rainbow (fetched 2026-08-28T04:05:20.271039+00:00, sha bffb7d11e176)
- Data as of 2026-08-30T08:39:29.467469+00:00.