Introduction

This package provides a Theano-based implementation of the deep Q-learning algorithm described in:

Playing Atari with Deep Reinforcement Learning Volodymyr Mnih, Koray Kavukcuoglu, David Silver, Alex Graves, Ioannis Antonoglou, Daan Wierstra, Martin Riedmiller

The neural network code is largely borrowed from Sander Dieleman's solution for the Galaxy Zoo Kaggle challenge.

Here is a video showing a trained network playing breakout:

http://youtu.be/SZ88F82KLX4

Dependencies

• A reasonably modern NVIDIA GPU
• Cython
• OpenCV
• Each of the following should be installed from the master branches on github:
  • Theano (https://github.com/Theano/Theano)
  • Pylearn2 (https://github.com/lisa-lab/pylearn2)
  • Arcade Learning Environment (https://github.com/mgbellemare/Arcade-Learning-Environment)
    • Make sure that you set the USE_RLGLUE flag to 1 in the makefile.
• RL-Glue
• RL-Glue Python Codec

Running

Use the script ale_run.py to start all the necessary processes:

$ python ale_run.py --exp_pref data

This will store output files in a folder prefixed with data in the current directory. Pickled version of the network objects are stored after every epoch. The file results.csv will contain the testing output. You can plot the progress by executing plot_results.py:

$ python plot_results.py data_09-29-15-46_0p0001_0p9/results.csv

After a couple of days, you can watch the trained network play using the ale_run_watch.py script:

$ python ale_run_watch.py data_09-29-15-46_0p0001_0p9/network_file_99.pkl

Getting Help

The deep Q-learning web-forum can be used for discussion and advice related to deep Q-learning in general and this package in particular.

See Also

• https://github.com/kristjankorjus/Replicating-DeepMind

  This is a work in progress. Their project is described here: http://robohub.org/artificial-general-intelligence-that-plays-atari-video-games-how-did-deepmind-do-it/

• https://github.com/muupan/dqn-in-the-caffe

  Working Caffe-based implementation. (I haven't tried it, but there is a video of the agent playing Pong successfully.)

• https://github.com/brian473/neural_rl

  This is an almost-working implementation developed last spring by my student Brian Brown. I haven't reused his code, but Brian and I worked together to puzzle through some of the blank areas of the original paper.