Train Agents to Play 2048
Build up the game
The highest score I’ve got is this (5644) hahaha
Make a Gym environment
Design Neural Networks
Compare RL Algorithms
Performance comparison of PPO, DQN with the baseline ExpectiMax agent.
The highest score I’ve got is this (5644) hahaha
Performance comparison of PPO, DQN with the baseline ExpectiMax agent.