HORA

HORA research team

In-hand object rotation using reinforcement learning and rapid motor adaptation.

What you can explore

Study stable-grasp initialization, domain randomization and Allegro deployment.

Before you use it

Developed with a modified Allegro Hand. Public v4 examples and paper-reproduction versions differ; the README identifies a specific historical release.

Original contributors

Haozhi Qi and collaborators

License / access: See source terms; permissions for code, models and data may differ

Original code & documentation

Editorial reference · checked 2026-09-22