The dataset viewer is not available because its heuristics could not detect any supported data files. You can try uploading some data files, or configuring the data files location manually.
Prepficient engines and networks
Files the Prepficient app downloads on request. Each group keeps its own licence.
| Files | What | Licence |
|---|---|---|
stockfish-*, nn-*.nnue |
Stockfish and its network | GPL-3.0 (LICENSE) |
fairy-stockfish-*, *-<hash>.nnue variant networks, 3check-* |
Fairy-Stockfish and its networks | GPL-3.0 (LICENSE) |
atomic-stockfish-*, atomic_* |
Atomic-Stockfish and its network | GPL-3.0 (LICENSE) |
lc0, 11258-112x9-se.pb.gz |
Leela Chess Zero and a network | GPL-3.0 (LICENSE) |
maia-1100.pb.gz … maia-1900.pb.gz |
Maia (CSSLab) | GPL-3.0 (LICENSE) |
maia3-*.tflite |
Maia 3 by the University of Toronto Computational Social Science Lab | AGPL-3.0 (LICENSE-AGPL-3.0) |
Maia 3
maia3-5m.tflite, maia3-23m.tflite and maia3-79m.tflite are the Maia 3 checkpoints
maia3-5m.pt, maia3-23m.pt and maia3-79m.pt from
UofTCSSLab/Maia3-5M,
Maia3-23M and
Maia3-79M
(code: https://github.com/CSSLab/maia3, AGPL-3.0), converted to LiteRT on 2026-10-29
by Prepficient: batch size 1, int32 inputs, the input encoding and output layout changed
as documented in the export script. The -w8 files also store the weights as int8.
No training or other change to the weights.
They are licensed under the AGPL-3.0, like the original. Corresponding source: the
original checkpoints (links above), the Maia 3 code, and the export script
sources/maia3-litert-export.tar.gz in this dataset.
Maia 3 is described in: Daniel Monroe, George Eilender, Philip Chalmers, Zhenwei Tang, Ashton Anderson. Chessformer: A Unified Architecture for Chess Modeling. ICLR 2026. Prepficient is not affiliated with or endorsed by the Maia authors.
Sources
The GPL/AGPL source of every program here is in sources/.
- Downloads last month
- 94