Skip to content

HybridRobotics/isaac_berkeley_humanoid

Repository files navigation

Berkeley Humanoid Traning Code

IsaacSim Isaac Lab Python Linux platform

Overview

This repository shows the training code for Berkeley Humanoid with IsaacLab.

Publications

If you use this work in an academic context, please consider citing the following publications:

@misc{2407.21781,
Author = {Qiayuan Liao and Bike Zhang and Xuanyu Huang and Xiaoyu Huang and Zhongyu Li and Koushil Sreenath},
Title = {Berkeley Humanoid: A Research Platform for Learning-based Control},
Year = {2024},
Eprint = {arXiv:2407.21781},
}

Installation

cd exts/berkeley_humanoid
python -m pip install -e .

Run

Training an agent with RSL-RL on Velocity-Rough-Berkeley-Humanoid-v0:

# run script for training
${ISAAC_LAB_PATH}/isaaclab.sh -p scripts/rsl_rl/train.py --task Velocity-Rough-Berkeley-Humanoid-v0
# run script for playing
${ISAAC_LAB_PATH}/isaaclab.sh -p scripts/rsl_rl/play.py --task Velocity-Rough-Berkeley-Humanoid-Play-v0

FAQ

Q: Why doesn't the maximum torque of each joint match the values in the paper?

A: The maximum torque is limited for safety reasons.

Q: Where is the joint armature from?

A: From CAD system.

Q: Why does the friction of each joint so large?

A: The motor we used has large cogging torque, we include it in the friction of the actuator model.

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages