---
license: odc-by
features:
  - name: image_urls
    sequence:
      dtype: string
  - name: image_sha256s
    sequence:
      dtype: string
  - name: points
    sequence:
      sequence:
        struct:
          x:
            dtype: float32
          y:
            dtype: float32
  - name: counts
    sequence:
      dtype: int64
  - name: labels
    sequence:
      dtype: string
  - name: collection_method
    sequence:
      dtype: string

splits:
  - name: train
    num_examples: 472943

configs:
  - config_name: default
    data_files:
      - split: train
        path: "parquet/train-*"
---

# Molmo2 Multi-Image Pointing

This dataset contains **multi-image pointing/counting** metadata.
This dataset is generated by extending PixMo-Points using a semantic grouping algorithm designed to maximize coverage.

Molmo2-MultiImagePoint is a part of the [Molmo2 dataset collection](https://huggingface.co/collections/allenai/molmo2-data)  and was used to
provide the multi-image pointing capabilities of the [Molmo2 family of models](https://huggingface.co/collections/allenai/molmo2).

Quick links:
- 📃 [Paper](https://allenai.org/papers/molmo2)
- 🎥 [Blog with Videos](https://allenai.org/blog/molmo2)

## Columns

- `image_urls`: list of image URLs (original source URLs)
- `image_sha256s`: sha256 of each URL string, aligned with `image_urls`
- `points`: list (per image) of point lists; each point is a dict `{"x": float, "y": float}`
- `counts`: list of counts aligned with `image_urls`
- `labels`: list of labels aligned with `image_urls`
- `collection_method`: list of collection_method aligned with `image_urls`

## License
This dataset is licensed under ODC-BY. It is intended for research and educational use in accordance with Ai2’s [Responsible Use Guidelines](https://allenai.org/responsible-use).