# Molmo2
Molmo2 locates objects as points using a text vocabulary.
Tasks: Point detection. Install: pip install "libreyolo[molmo2]".
Verified against LibreYOLO v1.6.0.

## Install

```bash
pip install "libreyolo[molmo2]"
```

## Predict

**Python**

```python
from libreyolo import LibreVLM, SAMPLE_IMAGE

model = LibreVLM("molmo2-4b", device="cpu")
model.set_classes(["person", "building"])
result = model(SAMPLE_IMAGE)
print(result.points)
```

`set_classes()` defines the pointing vocabulary. A custom pointing template must contain `{label}`. The factory also accepts `molmo2-8b` and `molmo2-o-7b`. Training is not supported.

## Licensing

Check the license on the Hugging Face repository of the specific weights you download. That repository is authoritative and licenses are not always uniform across a family. This is a description of the licenses involved, not legal advice.

- Original work: Molmo2, Allen Institute for AI
- Upstream license: Apache-2.0
- Upstream source: https://huggingface.co/allenai
- LibreYOLO code: MIT
- Weights: Apache-2.0, republished at https://huggingface.co/LibreYOLO
- Interpretation: The model publisher declares Apache-2.0.