Problem/Motivation

When I try to use Ollama and olmo2:latest with AI Image Alt Text (where I have uploaded a new image /media/add/image) I get:

[INPUT IMAGE DESCRIPTION]

Or a null request.

I am not getting an error in the logs.

When I go to the Chat Generation Explorer (/admin/config/ai/explorers/chat_generator) and try to use that to evaluate Ollama I get:

Response will appear here.
Role: assistant

I'm sorry, but I can't assist with that request. As an AI text-based model, I don't have the capability to view images or describe them. If you provide me with a textual description of the image or input specific details about it, I can help you find related information or answer questions based on that description. How can I assist you further?

Is there a good way to test

I don't know if this should be an issue with my Ollama config or AI Image Alt Text Issue.

I have no trouble asking other questions of Chat Generation Explorer. Just not ones with images.

I am not sure why I do not get better logs to debug this.

Machine name: ai_provider_ollama

Version: 1.0.0-beta1

Machine name: ai_image_alt_text

Version: 1.0.1

Steps to reproduce

See above

Proposed resolution

Remaining tasks

User interface changes

API changes

Data model changes

Comments

mgifford created an issue.

heatherwoz’s picture

heatherwoz commented

I am having a different but maybe related issue with this module and Ollama. I am using the qwen3 model and it works, but it returns this crazy long response in which it reasons over and over what to do because there is no image attached. I found that by adding `/no_think` to the prompt, I get a shorter result with the alt text at the end, but it is completely wrong alt text. Not a description of the image at all, a description of a totally different image.