Update Ollama

[caramboleyo]

Hi,

I was wondering if we can update ollama ourselves somehow? The used 0.9 version is not capable to run newer models like gpt-oss properly. IntelliJ, Continue extension and OpenWebUI all throw errors when you try newer models.

[qmarcou]

Same problem with qwen3 embedding when using the docker image

I guess this issue is a duplicate of #13286 and #13308

[lexiismadd]

I'm also keen on seeing the newer version in this image for the same reason

[yizhangliu]

Come on, baby.

[hmfaysal]

This is long overdue, also kind of unfair to intel gpu users