/Reinforcement-learning-tetris

Cross entropy method to train an agent to play tetris. The agent finally can delete around 5000-10000 blocks

Primary LanguageJupyter Notebook

Stargazers