Description
Teaching a bipedal humanoid to walk with hand-written gait controllers is complex and fiddly. This project trains humanoid locomotion with reinforcement learning using MuJoCo and PPO.
It supports robots like Cassie and JVRC-1 with complete training environments and policy code.
RL:PPO.
MuJoCo:Physics.
Robots:Cassie and JVRC.
Environments:Ready.
It supports robots like Cassie and JVRC-1 with complete training environments and policy code.
Features
RL:PPO.
MuJoCo:Physics.
Robots:Cassie and JVRC.
Environments:Ready.
