Without bounding box output

#11
by jasonni2 - opened

Congrats for this efficient model!
I'm using llama.cpp server for inference. When the document have digram, the ocr output only contains description of the diagram. Is this by design?

LightOn AI org

Thanks for trying the model!
yes the current version only indicates when there are images/charts/diagrams, we are working on adding bounding box information for each detected image natively by the model.
Stay tuned!

Sign up or log in to comment