MIT License

Copyright (c) 2026 David Litman

Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.

---

THIRD-PARTY NOTICES

The MIT license above covers the source code in this repository. It does not,
and cannot, relicense third-party work.

BUNDLED ARTWORK — the cburnett chess piece set by Colin M.L. Burnett, obtained
via python-chess's chess.svg.PIECES and redistributed here as the base64 PNGs in
chessformer_lens/piece_art.py. Triple-licensed by its author under GFDL / BSD /
GPL; this project uses it under the BSD option, which is compatible with MIT.
Attribution to Colin M.L. Burnett is required, and is given here and in
chessformer_lens/ui.py, which reads chess.svg.PIECES at import time.

PORTED CODE — chessformer_lens/adapters/searchless.py reproduces the FEN
tokenizer and the move list of DeepMind's searchless_chess
(https://github.com/google-deepmind/searchless_chess, Copyright 2025 DeepMind
Technologies Limited, Apache License 2.0). Those two functions are used under
Apache-2.0, whose notice and attribution requirements this paragraph meets;
the rest of the file is an independent PyTorch reimplementation.

DEPENDENCIES — python-chess (GPL-3.0-or-later) and maia3 (AGPL-3.0) are
installed by pip and no code from either is copied into this repository, so
using this project, and redistributing its source under MIT, are unaffected.
Redistributing a BUNDLE that ships them alongside it — a Docker image, a conda
package, a frozen binary, a vendored copy — makes that bundle a combined work,
which must then satisfy GPL-3.0, and AGPL-3.0 including its network-use clause
if maia3 is in it. Get a combined artifact or a hosted service reviewed rather
than relying on this summary.

MODEL WEIGHTS — Maia-3 checkpoints are downloaded at runtime from
https://huggingface.co/UofTCSSLab and are not covered by this license. Check the
model card before redistributing weights. DeepMind's searchless_chess
checkpoints are downloaded at runtime from
https://storage.googleapis.com/searchless_chess/checkpoints/ and are licensed
CC-BY 4.0 (Ruoss et al. 2024, arXiv:2402.04494); the converted .pt this package
caches is a derivative of them and carries the same attribution.
