Awesome

About

Chinese Chess reinforcement learning by AlphaGo Zero methods.

This project is modified from two projects:

1. https://github.com/Zeta36/chess-alpha-zero.git
2. https://github.com/timebug/harmless.git And I attached their original version as ZIPs in the root folder.

Note: This project is still under construction!! The <b>new-threads</b> branch reconstrusted the original multi-threading logic and increased the performance reliably

Environment required for training

• Python 3.6.3
• tensorflow-gpu: 1.3.0
• Keras: 2.0.8

Environment required for playing/replaying

• python-2.7.x
• pygame-1.9.x

How to use

For training

cd chess-alpha-zero vim readme.md

For playing/replaying

cd harmless-master vim README.md