Do InternVL3-1B-i1-GGUF support image to text

#1
by romyull - opened

Do you know how can I use this model for image to text inference?
I want to use this quantized version with llama.cpp
Do I need to have mmproj file for that?

Do you know how can I use this model for image to text inference?

You just download and specify the provided mmproj file when loading the model with llama.cpp

I want to use this quantized version with llama.cpp

Yes you can do so. You can choose between any of the static or weighted/imatrix quants for the LLM layers and either use Q8 or F16 for the vision layers.

Do I need to have mmproj file for that?

Yes you have to use one of the following mmproj files:

Thank you for your reply.
Again, thanks for showing me the mmproj file. I missed it totally.

Sign up or log in to comment