metaai·lightalo unofficial · independent
Universe / Games & Strategy / ELF OpenGo
Games & Strategy · 2018

ELF OpenGo

20-0 against Go professionals, on a single GPU — then open-sourced

open source archived
★ 3.4k ELF❝ 122 citationsread 2026-09-03

Latest: Final model + ICML 2019 analysis paper (v2, 2019)

FAIR's open reimplementation of AlphaZero for Go, built on the ELF reinforcement-learning platform. Trained on 2,000 GPUs over about two weeks, it went 20-0 against four top-30 professional players — while running on a single GPU — and released its code, trained models and analysis so anyone could reproduce a superhuman Go engine.

Why it matters

An open reimplementation of AlphaZero for Go that went 20-0 against four top-30 professionals while running on a single GPU, then released code, models, 20 million self-play games and a playable Windows binary so anyone could reproduce a superhuman engine. The repository was archived in December 2020.

Facts

Try it yourself

Sources

More in Games & Strategy

PluribusReBeLCICERO

Read the Games & Strategy story on the sky →

✦ Open on the map Explore Games & Strategy Quiz me