Implementation of AlphaZero in PyTorch. In particular, we want to make it play checkers well.
Based on publicly available pseudocode from the publication A general reinforcement learning algorithm that masters chess, shogi, and Go through self-play by Silver et al. http://science.sciencemag.org/highwire/filestream/719481/field_highwire_adjunct_files/1/aar6404_DataS1.zip