View Code? Open in Web Editor
NEW
This project forked from suragnair/alpha-zero-general
A clean implementation based on AlphaZero for any game in any framework + tutorial + Othello/Gobang/TicTacToe/Connect4
License: MIT License
Python 99.54%
Shell 0.46%
alpha-zero-general's Introduction
alpha-zero-general's People
Watchers
alpha-zero-general's Issues
Let some configuration arguments(numIters
, numEps
, load_folder_file
...) configurable via cli.
Multithreading is crucial to accelerate training procedure as cpu is bottleneck.
Multithreading target:
- Self Play
- Pitting against previous version
- Training
In the case of none of a player's pieces are left, game must terminated.
Actual agent interface that fits with official rule is required.