pvnieo/beating-atari

Implementation of RL algorithms to beat Atari 2600 games

beating-atari

Modularized implementation of popular deep RL algorithms in PyTorch to beat Atari 2600 games. Easy switch between algorithms and challenging games.

Implemented algorithms:

• Nips DQN [1]
• Nature DQN [2]
• Double DQN [3]
• Prioritised Experience Replay [4]
• Dueling Network Architecture [5]

Requirement

• gym
• PyTorch
• OpenCV
• tensorboard

This project runs on python >= 3.6, use pip to install dependencies:

pip3 install -r requirements.txt

Project report

See project report here.

References

[1] Playing Atari with Deep Reinforcement Learning
[2] Human-level control through deep reinforcement learning
[3] Deep Reinforcement Learning with Double Q-learning
[4] Prioritized Experience Replay
[5] Dueling Network Architectures for Deep Reinforcement Learning