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

# molmoact2_tool_insert

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

| | |
|---|---|
| **Base model** | [allenai/MolmoAct2](https://huggingface.co/allenai/MolmoAct2) |
| **Dataset** | [villekuosmanen/armnetbench_tool_insert](https://huggingface.co/datasets/villekuosmanen/armnetbench_tool_insert) |
| **Task** | `tool_insert` |
| **Training** | Vast.ai H100 SXM, batch 64, bf16, gradient checkpointing |
| **W&B project** | [molmoact2_tool_insert](https://wandb.ai/pravsels/molmoact2_tool_insert) |
| **W&B run** | [8m94f03c](https://wandb.ai/pravsels/molmoact2_tool_insert/runs/8m94f03c) |

## Checkpoints

| Step | Path |
|------|------|
| 006000 | `checkpoints/006000/pretrained_model/` |
| 004000 | `checkpoints/004000/pretrained_model/` |

## Usage

```python
from lerobot.policies.molmoact2.modeling_molmoact2 import MolmoAct2Policy
policy = MolmoAct2Policy.from_pretrained("pravsels/molmoact2_tool_insert", revision="main")
```