Expose in the UI usage of custom models running locally (ollama, llamafile, vllm...)

[ividal]

Motivation

#770 adds support (via API) of users launching any model via ollama, llamafile or vllm. They would then have to launch experiments via curl.
Instead, ideally we should also be able to select them via UI.

Alternatives

Access via curl, requests and SDK are possible after #770 .

Contribution

Extending the UI guide with the example, once the feature is available.

Have you searched for similar issues before submitting this one?

• Yes, I have searched for similar issues