Asynchronous Method for Deep Reinforcement Learning

Google DeepMind's Paper: Asynchronous Method for Deep Reinforcement Learning.

This implementation contains:

1. Deep Q-network and Q-learning
2. Varying epsilon for improving exploration vs exploitation
   • to reduce the correlations between consecutive parameter updates
3. Network for Q-learning targets are fixed for intervals
   • to reduce the correlations between target and predicted Q-values
4. Separate actor-learner threads
   • To improve system performance

Requirements

• Python 2.7 or Python 3.3+
• gym 0.7.0
• tqdm
• SciPy or OpenCV2
• TensorFlow 1.3.0
• Atari-py 0.0.21

Usage

First, install prerequisites with:

$ pip install tqdm gym[all]
$ pip install atari-py==0.0.21
$ pip install gym==0.7.0

To train a model for Breakout:

$ python main.py --env_name=Breakout-v0 --is_train=True
$ python main.py --env_name=Breakout-v0 --is_train=True --display=True

To test and record the screen with gym:

$ python main.py --is_train=False
$ python main.py --is_train=False --display=True

References

• simple_dqn
• Code for Human-level control through deep reinforcement learning

License

MIT License.