/ChessWarrior

Final Project of Artificial Intelligence

Primary LanguagePython

ChessWarrior

Final Project of 2018 Artificial Intelligence at Tongji Univ.

Usage

There are three modes: data train and play

  • data: generate processed training data
  • train: the main process to train neural network
  • play: use the best model to play chess

python run.py -mode data

python run.py -mode train -lr 0.01

python run.py -mode play --no-cuda

Requirements

environment windows python 3.x

  • tensorflow-gpu
  • keras==2.1.5
  • h5py
  • numpy
  • python-chess

just

pip install -r requirements.txt

pretrained models

https://pan.baidu.com/s/1C2FVG9vQds-odrfgw5bUGQ passwd: imhk