Multimodal
To use multimodal models, when adding a chat, select the appropriate text model, e.g. MobileVLM-3B-q3_K_S.gguf
, activate the CLIP
option and select the appropriate CLIP (mmproj) model
, e.g. MobileVLM-3B-mmproj-f16.gguf
.
If everything is done, a button will appear in the chat to add an image to the message.
If the model does not respond to the image, check if the text and clip models are selected.