tree: 18cbf985a5469158e59bce580de1ea635a9b9582 [path history] [tgz]
  1. README.md
  2. ddpg.py
  3. policies.py
  4. qfuncs.py
  5. replay_mem.py
  6. run.py
  7. strategies.py
  8. utils.py
example/reinforcement-learning/ddpg/README.md

mx-DDPG

MXNet Implementation of DDPG

Introduction

This is the MXNet implementation of DDPG. It is tested in the rllab cart pole environment against rllab‘s native implementation and achieves comparably similar results. You can substitute with this anywhere you use rllab’s DDPG with minor modifications.

Dependency

  • rllab

Usage

To run the algorithm,

python run.py

The implementation relies on rllab for environments and logging and the hyperparameters could be set in run.py.

References

  1. Lillicrap, Timothy P., et al. Continuous control with deep reinforcement learning. arXiv preprint arXiv:1509.02971 (2015).
  2. rllab. URL: https://github.com/openai/rllab