official-stockfish/nnue-pytorch
Stockfish NNUE (Chess evaluation) trainer in Pytorch
This project helps chess engine developers train and manage new neural network-based evaluation functions (NNUEs) for Stockfish. It takes in raw chess game data and outputs optimized NNUE files that can be integrated into the Stockfish engine. Its primary users are engine developers and researchers looking to enhance chess engine strength.
452 stars.
Use this if you are a chess engine developer looking to train and refine NNUE evaluation networks using PyTorch for integration with Stockfish.
Not ideal if you are an end-user simply wanting to play or analyze chess with Stockfish, or if you need to train a traditional alpha-beta chess engine without neural networks.
Stars
452
Forks
131
Language
Python
License
GPL-3.0
Category
Last pushed
Mar 09, 2026
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/official-stockfish/nnue-pytorch"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Related frameworks
LeelaChessZero/lc0
Open source neural network chess engine with GPU acceleration and broad hardware support.
QueensGambit/CrazyAra
A Deep Learning UCI-Chess Variant Engine written in C++ & Python :parrot:
bagaturchess/Bagatur
Java Chess Engine (UCI compatible)
xqbase/eleeye
ElephantEye - a XiangQi (Chinese Chess) Engine for XQWizard with Strong AI
connormcmonigle/seer-nnue
UCI chess engine using neural networks for position evaluation