https://github.com/hijkzzz/alpha-zero-gomoku

A Multi-threaded Implementation of AlphaZero
https://github.com/hijkzzz/alpha-zero-gomoku

alphazero cpp gomoku-game libtorch multithreading

Last synced: 1 day ago
JSON representation

A Multi-threaded Implementation of AlphaZero

• Host: GitHub
• URL: https://github.com/hijkzzz/alpha-zero-gomoku
• Owner: hijkzzz
• Created: 2018-12-29T11:55:38.000Z (almost 6 years ago)
• Default Branch: master
• Last Pushed: 2023-01-07T07:25:29.000Z (almost 2 years ago)
• Last Synced: 2023-11-07T18:30:08.985Z (about 1 year ago)
• Topics: alphazero, cpp, gomoku-game, libtorch, multithreading
• Language: Python
• Homepage:
• Size: 7.83 MB
• Stars: 324
• Watchers: 10
• Forks: 53
• Open Issues: 1
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# AlphaZero Gomoku

A multi-threaded implementation of AlphaZero

## Features

* Easy Free-style Gomoku

* Multi-threading Tree/Root Parallelization with Virtual Loss and LibTorch

* Gomoku, MCTS and Network Infer are written in C++

* SWIG for Python C++ extension

* Update 2019.7.10: Supporting Ubuntu and Windows

* Update 2022.4.4: Re-compile with CUDA 11.6/PyTorch 1.10/LibTorch 1.10(Pre-cxx11 ABI)/SWIG 4.0.2

## Args

Edit config.py

## Packages

* Python 3.6+

* PyGame 1.9+

* CUDA 10+

* [PyTorch 1.1+](https://pytorch.org/get-started/locally/)

* [LibTorch 1.1+ (Pre-cxx11 ABI)](https://pytorch.org/get-started/locally/)

* [SWIG 3.0.12+](https://sourceforge.net/projects/swig/files/)

* CMake 3.8+

* MSVC14.0+ / GCC6.0+

## Run

```

# Compile Python extension

mkdir build

cd build

cmake .. -DCMAKE_PREFIX_PATH=path/to/libtorch -DPYTHON_EXECUTABLE=path/to/python -DCMAKE_BUILD_TYPE=Release

make -j10

# Run

cd ../test

python learner_test.py train # train model

python learner_test.py play  # play with human

```

## Pre-trained models

> Trained 2 days on GTX TITAN X (similar to GTX1070)

See GitHub Release: https://github.com/hijkzzz/alpha-zero-gomoku/releases

## GUI

![](https://github.com/hijkzzz/alpha-zero-gomoku/blob/master/assets/gomoku_gui.png)

## References

1. Mastering the Game of Go without Human Knowledge

2. Mastering Chess and Shogi by Self-Play with a General Reinforcement Learning Algorithm

3. Parallel Monte-Carlo Tree Search

4. An Analysis of Virtual Loss in Parallel MCTS

5. A Lock-free Multithreaded Monte-Carlo Tree Search Algorithm

6. github.com/suragnair/alpha-zero-general