Polygames: Improved Zero Learning
Machine Learning
2020-01-28 v1 Machine Learning
Abstract
Since DeepMind's AlphaZero, Zero learning quickly became the state-of-the-art method for many board games. It can be improved using a fully convolutional structure (no fully connected layer). Using such an architecture plus global pooling, we can create bots independent of the board size. The training can be made more robust by keeping track of the best checkpoints during the training and by training against them. Using these features, we release Polygames, our framework for Zero learning, with its library of games and its checkpoints. We won against strong humans at the game of Hex in 19x19, which was often said to be untractable for zero learning; and in Havannah. We also won several first places at the TAAI competitions.
Cite
@article{arxiv.2001.09832,
title = {Polygames: Improved Zero Learning},
author = {Tristan Cazenave and Yen-Chi Chen and Guan-Wei Chen and Shi-Yu Chen and Xian-Dong Chiu and Julien Dehos and Maria Elsa and Qucheng Gong and Hengyuan Hu and Vasil Khalidov and Cheng-Ling Li and Hsin-I Lin and Yu-Jin Lin and Xavier Martinet and Vegard Mella and Jeremy Rapin and Baptiste Roziere and Gabriel Synnaeve and Fabien Teytaud and Olivier Teytaud and Shi-Cheng Ye and Yi-Jun Ye and Shi-Jim Yen and Sergey Zagoruyko},
journal= {arXiv preprint arXiv:2001.09832},
year = {2020}
}