---
license: mit
language:
- en
- fr
- es
base_model:
- unsloth/medgemma-4b-it-bnb-4bit
tags:
- medical
---

## MOSAIC 4B

<img src="https://raw.githubusercontent.com/aliswh/mosaic/refs/heads/main/mosaic-icon.png" alt="Mosaic logo" width="200"/>
MOSAIC is a framework for efficient radiological report classification that is:

- 🌐 Multilingual: Works across different languages
- 🎯 Taxonomy-Agnostic: Adapts to various classification schemes
- ⚡ Computationally Efficient: Optimized for resource usage

## Training code

Code available on [GitHub](https://github.com/aliswh/mosaic).

## References

[Model Paper](https://arxiv.org/abs/2509.04471)

A 12B model version is available at [AliceSch/mosaic-12b](https://huggingface.co/AliceSch/mosaic-12b).

Please cite this model as:
```
@article{schiavone2025mosaic,
  title={MOSAIC: A Multilingual, Taxonomy-Agnostic, and Computationally Efficient Approach for Radiological Report Classification},
  author={Schiavone, Alice and Fraccaro, Marco and Pehrson, Lea Marie and Ingala, Silvia and Bonnevie, Rasmus and Nielsen, Michael Bachmann and Beliveau, Vincent and Ganz, Melanie and Elliott, Desmond},
  journal={arXiv preprint arXiv:2509.04471},
  year={2025}
}
```