missing mmproject

#1
by kalle07 - opened

missing mmproject

Is this actually a vision model?

it's based on Qwen3ForCausalLM, not Qwen3VLForConditionalGeneration

I know it's tagged in the original model as image-text-to-text, but i don't think it is

yes sorry ... the small line indicate it is not
"14B text model based on Qwen3-14B"

All good! They still have it mislabeled so I understand the confusion 😅

Sign up or log in to comment