GA3C: Reinforcement Learning through Asynchronous Advantage Actor-Critic on a GPU

A hybrid CPU/GPU version of the Asynchronous Advantage Actor-Critic (A3C) algorithm, currently the state-of-the-art method in reinforcement learning for various gaming tasks. This CPU/GPU implementation, based on TensorFlow, achieves a significant speed up compared to a similar CPU implementation.

How do I get set up?

• Install Python > 3.0
• Install TensorFlow 1.0
• Install OpenAI Gym
• Clone the repo.
• That's it folks!

How to Train a model from scratch?

Run sh _clean.sh first, and then sh _train.sh. The script _clean.sh cleans the checkpoints folder, which contains the network models saved during the training process, as well as removing results.txt, which is a log of the scores achieved during training.

Remember to save your trained models and scores in a different folder if needed before cleaning.

_train.sh launches the training procedure, following the parameters in Config.py. You can modify the training parameters directly in Config.py, or pass them as argument to _train.sh. E.g., launching sh _train.sh LEARNING_RATE_START=0.001 overwrites the starting value of the learning rate in Config.py with the one passed as argument (see below). You may want to modify _train.sh for your particular needs.

The output should look like below:

...
[Time: 33] [Episode: 26 Score: -19.0000] [RScore: -20.5000 RPPS: 822] [PPS: 823 TPS: 183] [NT: 2 NP: 2 NA: 32]
[Time: 33] [Episode: 27 Score: -20.0000] [RScore: -20.4815 RPPS: 855] [PPS: 856 TPS: 183] [NT: 2 NP: 2 NA: 32]
[Time: 35] [Episode: 28 Score: -20.0000] [RScore: -20.4643 RPPS: 854] [PPS: 855 TPS: 185] [NT: 2 NP: 2 NA: 32]
[Time: 35] [Episode: 29 Score: -19.0000] [RScore: -20.4138 RPPS: 877] [PPS: 878 TPS: 185] [NT: 2 NP: 2 NA: 32]
[Time: 36] [Episode: 30 Score: -20.0000] [RScore: -20.4000 RPPS: 899] [PPS: 900 TPS: 186] [NT: 2 NP: 2 NA: 32]
...

PPS (predictions per second) demonstrates the speed of processing frames, while Score shows the achieved score.
RPPS and RScore are the rolling average of the above values.

To stop the training procedure, adjuts EPISODES in Config.py propoerly, or simply use ctrl + c.

How to continue training a model?

If you want to continue training a model, set LOAD_CHECKPOINTS=True in Config.py, and set LOAD_EPISODE to the episode number you want to load. Be sure that the corresponding model has been saved in the checkpoints folder (the model name includes the number of the episode).

Be sure not to use _clean.sh if you want to stop and then continue training!

How to play a game with a trained agent?

Run _play.sh You may want to modify this script for your particular needs.

How to change the game, configurations, etc.?

All the configurations are in Config.py
As mentioned before, one useful way of modifying a config is to pass it as an argument to _train.sh. For example, to save the models while training, just run: train.sh TRAINERS=4.

Sample learning curves

Typical learning curves for Pong and Boxing are shown here. These are easily obtained from the results.txt file.

References

If you use this code, please refer to our ICLR 2017 paper:

@conference{babaeizadeh2017ga3c,
  title={Reinforcement Learning thorugh Asynchronous Advantage Actor-Critic on a GPU},
  author={Babaeizadeh, Mohammad and Frosio, Iuri and Tyree, Stephen and Clemons, Jason and Kautz, Jan},
  booktitle={ICLR},
  biurl={https://openreview.net/forum?id=r1VGvBcxl},
  year={2017}
}

This work was first presented in an oral talk at the The 1st International Workshop on Efficient Methods for Deep Neural Networks, NIPS Workshop, Barcelona (Spain), Dec. 9, 2016:

@article{babaeizadeh2016ga3c,
  title={{GA3C:} {GPU}-based {A3C} for Deep Reinforcement Learning},
  author={Babaeizadeh, Mohammad and Frosio, Iuri and Tyree, Stephen and Clemons, Jason and Kautz, Jan},
  journal={NIPS Workshop},
  biurl={arXiv preprint arXiv:1611.06256},
  year={2016}
}