Ericonaldo / visual_wholebody

Train a loco-manipulation dog with RL
https://wholebody-b1.github.io/
Other
138 stars 13 forks source link

Low-level policy urdf choice #15

Open wzx16 opened 2 months ago

wzx16 commented 2 months ago

Hi, thanks for sharing this amazing project. I found in the low-level policy folder there is a b1z1.urdf and a b1z1-new.urdf, inside which some of the parameters are different. I found if I load b1z1.urdf to isaac-sim, the structure of the z1 arm is incorrect as the following images: Screenshot from 2024-09-19 00-06-02 This issue doesn't exist in b1z1-new.urdf. However, I found the low-level policy use b1z1.urdf instead. Could you kindly explain the difference and give some choice suggestions? Thank you!