# TetrisRL
- Implement a reinforcement learning model which can play Tetris

# Version 0
- Policy Gradient Algorithm
- Reward Delay
- Bad Performance

# TODO
- change reward function
- reward baseline
- DQN