Upload README.md with huggingface_hub
Browse files
README.md
ADDED
|
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: agpl-3.0
|
| 3 |
+
tags:
|
| 4 |
+
- grasping
|
| 5 |
+
- jax
|
| 6 |
+
- equivariance
|
| 7 |
+
- multi-embodiment
|
| 8 |
+
---
|
| 9 |
+
|
| 10 |
+
# Kinematics Flow checkpoints
|
| 11 |
+
|
| 12 |
+
Single-embodiment checkpoints for [Kinematics Flow](https://github.com/boschresearch/kinematics-flow), from ["Towards a Multi-Embodied Grasping Agent"](https://arxiv.org/abs/2510.27420). These are **mid-training checkpoints**, not final converged models.
|
| 13 |
+
|
| 14 |
+
## Checkpoints
|
| 15 |
+
|
| 16 |
+
| Folder | Gripper | DOF | Epoch | Test success rate | Test NJD |
|
| 17 |
+
|---|---|---|---|---|---|
|
| 18 |
+
| `se-panda_5000_170` | Panda | 2 | 170 / 500 | 97.8% | 0.293 |
|
| 19 |
+
| `se-shadow_5000_40` | Shadow Hand | 22 | 40 / 500 | 75.9% | 0.232 |
|
| 20 |
+
|
| 21 |
+
Both trained with `num_scenes=5000` on a single RTX 6000 Ada (48GB). Eval numbers are simulation-based grasp success rate and normalized joint diversity (NJD), computed via `kin_flow.cli.bench` on 10 held-out test scenes (100 sampled grasps/scene each).
|
| 22 |
+
|
| 23 |
+
Format: orbax/OCDBT checkpoint directories, loadable via `kin_flow.ctrl.trainer.Trainer.get_model_from_checkpoint`.
|