# ShangtongZhang/DeepRL

Modularized Implementation of Deep RL Algorithms in PyTorch

Repository: https://github.com/ShangtongZhang/DeepRL
Canonical: https://ross.abutalabs.com/products/deeprl
Language: Python
License: MIT
License Family: permissive
Topics: pytorch, deep-reinforcement-learning, dqn, dueling-network-architecture, double-dqn, deeprl, ddpg, ppo, categorical-dqn, option-critic, quantile-regression, td3, a2c, option-critic-architecture, prioritized-experience-replay, rainbow
Last push: 2024-04-16T19:43:16+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": 3422, "days_push": 869, "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 3448, forks 697 (observed 2026-08-28T04:08:05.207084+00:00)

## What it is
A modularized PyTorch implementation of popular deep reinforcement learning algorithms including DQN variants, PPO, DDPG, TD3, A2C, and Option-Critic. It supports switching between toy tasks and Atari/Mujoco benchmark environments.

## Use cases
- implement deep reinforcement learning algorithms in pytorch
- train a DQN agent on Atari games
- run PPO on Mujoco continuous control tasks
- study reference implementations of Rainbow-style DQN variants
- benchmark deep RL algorithms on a single GPU

## When to choose
- you want readable, modular PyTorch code for classic deep RL algorithms
- you need a research baseline implementation of DQN, PPO, DDPG, or TD3
- you want to reproduce paper results on Atari or Mujoco

## When to avoid
- you need a production RL training platform with distributed scaling
- you want the latest maintained framework with broad environment support
- you need algorithms beyond the implemented set or newer PyTorch versions

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, reinforcement-learning, deep-learning
- domain: reinforcement-learning, machine-learning, deep-learning
- platform: python
- tags: pytorch, dqn, ppo, ddpg, td3, a2c, option-critic, atari, mujoco, research-code, gpu, docker

## Member repositories
- ShangtongZhang/DeepRL (main) score 23

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:08:05.207084+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 2026-08-29T18:37:35.724568+00:00, confidence not recorded.
  - readme: https://github.com/ShangtongZhang/DeepRL (fetched 2026-08-28T04:08:05.207084+00:00, sha 13d283c95a81)
- Data as of 2026-08-30T08:39:29.467469+00:00.