AlphaBia

A Makruk (Thai chess) engine trained from zero by AlphaZero-style self-play, on a single RTX 4060 at home.

This is a hobby project. I wanted to see how far a normal gaming PC can push a variant that has almost no neural engines.

Credits

The whole codebase is built on KataGomo by hzyhhzy, specifically the AnimalChess2025 branch, which itself is a fork of KataGo by lightvector. I did not write the engine core, the search, or the training pipeline - those people did, and they did it well. What I added is the Makruk game logic, rules, tests, and the training runs.

Btw Claude helped me a lot throughout this project :)

What's implemented

• Full Makruk rules: piece movement (Khun, Rua, Ma, Khon, Met, Bia), directional pieces, Bia promotion to Bia Gae, check, checkmate, and stalemate

• Thai draw counting rules:

  • 2 rooks: 8
  • 1 rook: 16
  • 2 bishops: 22
  • 2 knights: 32
  • 1 bishop: 44
  • 1 knight: 64
  • Only Met / promoted Bia Gae remaining: 64

• 64-move board counting rule

• Threefold repetition draw

• One move = two engine plies: first selecting the from-square, then selecting the to-square. Keep this in mind if you script against GTP.

• Test harness that plays real matches against Fairy-Stockfish over UCI, with strict rule enforcement and PGN output: test/engine_match.py

The model

There is one model in the releases: model_192 b15c192. It was trained on my RTX 4060 for roughly 50-70 hours of self-play. It beats Fairy-Stockfish makruk at low skill levels and holds its own a bit above that. Endgame conversion under the counting rule is the hard part and it is still improving.

If you want to continue training, my honest recommendation: don't keep growing this net - start a b20c256 and train it on data generated with model_192 (or its own self-play). The bigger net has a much higher ceiling. Expect around 2-3 days on similar hardware before it reaches the level of the model I'm giving you, and it climbs past it from there.

Building

Same as KataGo/KataGomo. Linux or WSL:

cd cpp
cmake -B build -DUSE_BACKEND=CUDA -DCMAKE_BUILD_TYPE=Release
cmake --build build -j$(nproc)

Use -DUSE_BACKEND=EIGEN for CPU-only. You need cuDNN for the CUDA backend.

Running

Play/analyze over GTP:

./cpp/build/katago gtp -config <your gtp config> -model model.bin.gz

Self-play training loop (selfplay -> shuffle -> train -> export) is in scripts_makruk/ubuntu_alphazero_train.sh with the config in scripts_makruk/makruk_selfplay_alphazero.cfg. The python training side is under scripts/alphabia/trainsgd/.

Rule unit tests build standalone without any GPU:

bash scripts_makruk/build_selftest.sh
./build_selftest/makruk_selftest.exe

To benchmark against Fairy-Stockfish, drop a fairy-stockfish-largeboard binary into test/ and run test/engine_match.py.

License

This repository is released under the MIT License. See LICENSE.

---