bug: The multimodal model does not support image file uploads.

[ahs95]

Jan version

v0.5.6

Describe the Bug

I am experiencing an issue uploading image files to the multimodal model "llava-v1.5-13b-Q2_K.gguf". The model only accepts PDF documents for upload, preventing me from uploading images.
Operating System: Windows 11

Steps to Reproduce

1. Import model from Hugging Face to Jan app.
2. Download the quantization (2-bit) model named "llava-v1.5-13b-Q2_K.gguf".
3. Then start to use the model by clicking the use button.

Screenshots / Logs

What is your OS?

• MacOS
• Windows
• Linux

[imtuyethan]

Related to janhq/models#46

[louis-menlo]

Hi @ahs95 Vision model import is not supported yet. It requires two GGUF files to work. I see that you imported just the chat model, but there is still the clip model.