Loading…
Loading…
A robotics policy fine-tuned with Servo and published to the Hub.
Provenance below is drawn from the model's Servo runtime manifest.
| Field | Value |
|---|---|
| Base model | gr00t-n1.7 |
| Vendor | nvidia |
| Runtime family | gr00t |
| Embodiment | yam |
| Morphology | bimanual |
| Action space | joint_position (dim 14) |
| Normalization | None |
| License tier | commercial_ok |
| Seed dataset | allenai/MolmoAct2-BimanualYAM-Dataset |
| Dataset URI | hf://allenai/MolmoAct2-BimanualYAM-Dataset |
These weights are the fine-tuned checkpoint produced by a Servo fine-tune job. Load them with the runtime for the base model family above; the action space, normalization, and embodiment contract in the table define how observations and actions must be shaped.