TD3_Pytorch_for_Walker

Twin Delayed Deep Deterministic Policy Gradient for Gait Learning (Walker2d-v2) by Pytorch

Usage

• To train a new TD3 agent:
  

python train.py

• To test trained TD3 agent:
  

python test.py

Dependancies

python 3.7
mujoco-py 2.0.2.13
gym 0.15.4

Result

References

• Official TD3 paper and code
• https://github.com/nikhilbarhate99/TD3-PyTorch-BipedalWalker-v2
• Stable baseline spinning up
• https://sylim2357.github.io/paper%20review/td3/