Vision support lost in gguf conversion?
#1
by iamrobotbear - opened
When trying to use the model via LMStudio it doesn't seem to support multimodal input.
Fixed, download mmproj-Apriel-1.5-15b-Thinker to the folder with the model.
I warn you that the vision may work terribly because it was probably not touched during training.
Disya changed discussion status to closed