from Hacker News

OpenSpiel: A Framework for Reinforcement Learning in Games

by stablemap on 8/27/19, 5:29 PM with 1 comments

  • by falsedan on 8/30/19, 3:14 PM

    the tl;dr:

    * C++

    * with python bindings

    * implement your game actions and state in C++

    * then can pump the state and legal actions into an algorithm (that you write)

    * has some implementations for some algorithms (minimax, monte carlo) & games (tic-tac-toe, chess, backgammon)