# pytorch/ELF

ELF: a platform for game research with AlphaGoZero/AlphaZero reimplementation

Repository: https://github.com/pytorch/ELF
Canonical: https://ross.abutalabs.com/products/elf
Language: C++
License: NOASSERTION
License Family: other
Topics: reinforcement-learning, alphago-zero, rl, rl-environment, alpha-zero, go
Archived: true
Last push: 2019-06-21T01:50:03+00:00

## Health v2 (maintenance only)
Score: 10/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 8, longevity 100
- inputs: {"age_days": 3051, "days_push": 2631, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: archived, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 3411, forks 571 (observed 2026-08-28T04:08:03.379675+00:00)

## What it is
ELF is an extensive, lightweight, and flexible C++/Python platform for game research, best known for ELF OpenGo, a full AlphaGoZero/AlphaZero reimplementation that beat professional Go players. It provides RL environments and self-play training infrastructure for games like Go and real-time strategy games.

## Use cases
- train an AlphaZero-style Go bot
- run self-play reinforcement learning for board games
- reproduce AlphaGoZero research results
- build RL environments for real-time strategy games
- analyze human Go games with a strong AI model
- research distributed RL training with GPU acceleration

## When to choose
- you need a proven AlphaZero reimplementation with pretrained Go models
- you are doing RL game research on Linux with CUDA GPUs
- you want a high-performance C++ game environment for RL experiments

## When to avoid
- you need production-quality, well-documented software
- you develop on Windows or macOS
- you want a maintained project with active releases
- you need a general-purpose RL library with broad algorithm support

## Facets
- artifact type: framework
- maturity: maintenance
- function: reinforcement-learning, machine-learning, game-engine, simulation
- domain: reinforcement-learning, machine-learning
- platform: cpp, python
- tags: alphago-zero, alpha-zero, go, board-games, self-play, research-code, facebook-research, game-development, research, linux, gpu

## Member repositories
- pytorch/ELF (main) score 10

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:08:03.379675+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 2026-08-29T18:38:31.468960+00:00, confidence not recorded.
  - readme: https://github.com/pytorch/ELF (fetched 2026-08-28T04:08:03.379675+00:00, sha 993b9a9d40cf)
- Data as of 2026-08-30T08:39:29.467469+00:00.