Is this a VLM???
#4
by
zbbhhh
- opened
I use ollama to load and run this model, but it seems that the model is not able to recognize image.
And I ask the model like this: " Are you a LLM or a VLM?", the answer is LLM
Hoping for your help, thanks!!!
I recommend just pulling the model from Ollama directly: ollama run llama3.2-vision. It should work with the modelfile that was published in the pinned issue, but I haven't tested it yet. The answer is probably just a hallucination, considering the vision part of the model is not used for text only inputs.