Worth knowing
How researchers are adapting olmOCR 2 for historical and archaeological archives
It demonstrates the practical value of open-weights vision-language models for academic researchers who need to adapt models to specialized domains or process sensitive archival data locally without third-party APIs.
Reported by Ai2 ·
What happened
Historians and archaeologists are using Ai2's open-weights olmOCR 2 model to transcribe handwritten records and make archival document libraries searchable on local hardware. In one project, fine-tuning olmOCR 2 on just 57 annotated examples reduced transcription error on Roman pottery records from 15.58% to 1.11%.
Who it’s about
- Ai2
- olmOCR 2
- ArtificeOCR
- CENTURIA
- TU Wien
Source
More on open weights
Get a brief picked for you
Sign in to follow topics and companies, save stories and search the archive.
Sign in