hugpadhugpad

lightonai / LightOnOCR-2-1B-bbox-soup

Vision1B2.3K18Updated Jun 29, 2026
Hugging Face

Downloads

2.3K

Likes

18

Last updated

Jun 29, 2026

License

apache-2.0

Merged bbox variant (advanced). This model combines OCR-improving RLVR signals with bounding-box-focused RLVR updates via joint merging, preserving OCR quality while providing image localization.

Tasks & library

  • image text to text
  • transformers

Tags

  • transformers
  • mistral3
  • text-generation
  • ocr
  • document-understanding
  • vision-language

Technical specifications

Parameters1B
Librarytransformers
Languages11 languages
Licenseapache-2.0