muyoucun / ddpg

reinforcement learning ddpg code. follow deepmind papers.

ddpg

Implementation of DDPG - Deep Deterministic Policy Gradient - on gym-torcs. with tensorflow.

ddpg paper: http://arxiv.org/pdf/1509.02971v2.pdf

Author: kenneth yu

installation dependencies:

1. tensorflow r1.4
2. gym_torcs: https://github.com/ugo-nama-kun/gym_torcs

how to run:

1. training mode:

 python3 gym_torcs_train_low_dim.py

2. evaluate mode:

  python3 gym_torcs_eval_low_dim.py