---
library_name: lerobot
tags:
- molmoact2
- robotics
- lerobot
- vla
base_model: allenai/MolmoAct2
---

# molmoact2_cable_clip_norm_fix

Fine-tuned [MolmoAct2](https://huggingface.co/allenai/MolmoAct2) (action-expert-only) on SO101 data, with SO101 gripper normalization fix.

| | |
|---|---|
| **Base model** | [allenai/MolmoAct2](https://huggingface.co/allenai/MolmoAct2) |
| **Dataset** | [pravsels/cable_clip_remote_v2](https://huggingface.co/datasets/pravsels/cable_clip_remote_v2) |
| **Task** | `cable_clip` |
| **Action dim** | 6 (single-arm) |
| **Cameras** | `top`, `wrist`, `front` |
| **Normalization** | MEAN_STD (action + state + gripper), IDENTITY (visual) |
| **Training** | Isambard GH200, batch 64, bf16, gradient checkpointing |
| **W&B project** | [molmoact2_cable_clip](https://wandb.ai/pravsels/molmoact2_cable_clip) |
| **W&B run** | [dw5rxtxp](https://wandb.ai/pravsels/molmoact2_cable_clip/runs/dw5rxtxp) |

## Checkpoints

| Step | Path |
|------|------|
| 002000 | `checkpoints/002000/pretrained_model/` |

The latest step (`002000`) is also copied to the **repository root** for direct loading.

## Usage

```python
from lerobot.policies.molmoact2.modeling_molmoact2 import MolmoAct2Policy
policy = MolmoAct2Policy.from_pretrained("pravsels/molmoact2_cable_clip_norm_fix")
```