Play "Five or More" game with Artificial intelligence
Go to file
Yang Long aef9044e0e test 2019-06-12 12:00:38 +08:00
alphafiveormore fix game bugs and update README 2018-12-04 16:35:28 +08:00
pics basic UI structure 2018-11-10 14:42:38 +08:00
.gitignore Fix UI bugs 2018-12-03 20:59:02 +08:00
LICENSE Initial commit 2018-10-26 16:46:26 +08:00 test 2019-06-12 12:00:38 +08:00


Play "Five or More" game with Artificial intelligence


  • Home-made "Five or More" game
  • Curiosity-driven Exploration by Self-supervised Prediction [1,2]

How to use

  • Play "Five or more" game by yourself

      cd alphafiveormore
      python --playai
  • Re-train a new AI model

      cd alphafiveormore
      python --retrain --verbose
  • Continue to train the AI model

      cd alphafiveormore
      python --train --verbose
  • View the AI model to play game

      cd alphafiveormore
      python --playai --verbose

Press key Space to evaluate the step


GameEngine is currently inadequate. Working on it


  1. Pathak, Deepak, et al. "Curiosity-driven exploration by self-supervised prediction." International Conference on Machine Learning (ICML). Vol. 2017. 2017.
  2. Achiam, Joshua, and Shankar Sastry. "Surprise-based intrinsic motivation for deep reinforcement learning." arXiv preprint arXiv:1703.01732 (2017).

You're most welcome to contact with me to discuss any details about this project