DDPG

Reproduce DDPG with PARL

Based on PARL, the DDPG algorithm of deep reinforcement learning has been reproduced, reaching the same level of indicators as the paper in Mujoco benchmarks.

Paper: DDPG in Continuous control with deep reinforcement learning

Mujoco games introduction

PARL currently supports the open-source version of Mujoco provided by DeepMind, so users do not need to download binaries of Mujoco as well as install mujoco-py and get license. For more details, please visit Mujoco.

Benchmark result

+ Each experiment was run three times with different seeds

How to use

Dependencies:

• python3.7+
• parl>=2.1.1
• paddlepaddle>=2.0.0
• gym>=0.26.0
• mujoco>=2.2.2

Start Training:

# To train an agent for HalfCheetah-v4 game
# python train.py

# To train for other game
# python train.py --env [ENV_NAME]