Episodes Preview hsr Visualizer
23.8k episodes · 30 fps · 2 cameras · 640×480 av1

This is a fork of AIRoA-MoMa, ready for VLA training. The action vector is created by aggregating the absolute joint position (action.absolute) with base velocities (action.base).

Downloads last month
259

Models trained or fine-tuned on paulprt/airoa-moma-absolute