Find answers from the community

Updated last month

Troubleshooting Image Description Issues with Open-webui, Llama-index, and Ollama

At a glance

Hi guys, is there anyone who use open-webui, llama-index, ollama?
I want to receive a description about the uploaded image but the response is strange. like, it describe me about another image. this is the code what i use
thanks.

3 comments

WWhiteFang_Jr

LLM is doing the same for all the images or just for few specific ones?

Maybe image is not clear or maybe the LLM is hallucinating🤔

kkimkim2

Do the same for all the images.
I think llama-index try to search index about the description of image, so it doesn't give me the pure response.
Or, the open-webui give a wrong image_url to pipeline

WWhiteFang_Jr

If you just want to upload the image , read the image and have LLM just describe the image then you dont need to create an index for that.

Basically You can use Vision LLMs like GPT-4o that gives you multi-modal support allowing you to read the image and describe it directly.

you could do something: https://docs.llamaindex.ai/en/stable/examples/multi_modal/openai_multi_modal/#ask-the-model-to-describe-what-it-sees

Add a reply