The base model is Qwen2.5-VL-3B and the announcement says a limitation is "Model... | Hacker News

Hacker Newsnew | past | comments | ask | show | jobs | submit

		nattaylor 74 days ago \| parent \| context \| favorite \| on: Nanonets-OCR-s – OCR model that transforms documen... The base model is Qwen2.5-VL-3B and the announcement says a limitation is "Model can suffer from hallucination"

gibsonf1 74 days ago [–]

Seems a bit scary that the "source" text from the pdfs could actually be hallucinated.

prats226 74 days ago | [–]

Given that input is image and not raw pdf, its not completely unexpected

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact