Starstruk Logo MarkStarstruk

v0.16.0

datalab-to/suryav0.16.0Aug 29, 2025by tarun-menta

AI Summary

Updates the OCR model to use a better vision encoder and smaller vocab size.

Key Highlights

  • New OCR model with better vocab size
  • Improved vision encoder
  • Improved math performance
  • Fixed table model for MPS devices

New Features

  • New OCR model

Full Release Notes

## General OCR Improvements
- Update to a better OCR model. Uses less vocab size, and a more performant vision encoder. 
- Improved math performance, with less spurious use of math tags when not needed

## Misc
- Fix table model by pinning to CPU on MPS devices

## What's Changed
* update commercial license description in README and update LICENSE to… by @sandy0kwon in https://github.com/datalab-to/surya/pull/434
* Dev by @VikParuchuri in https://github.com/datalab-to/surya/pull/436
* Model Update: New Tokenizer and Encoder by @tarun-menta in https://github.com/datalab-to/surya/pull/440
* Dev by @tarun-menta in https://github.com/datalab-to/surya/pull/441

## New Contributors
* @sandy0kwon made their first contribution in https://github.com/datalab-to/surya/pull/434

**Full Changelog**: https://github.com/datalab-to/surya/compare/v0.15.4...v0.16.0