Docs: describe whether /learn is multi-modal

[nick-youngblut]

trafficstars

Problem

The learning about local data do not state whether /learn is multi-modal and will use text + images. For instance, including plots generated in Jupyter notebooks can be extremely useful. Of course, multi-modal input would only be possible for the models that support the feature.

Proposed Solution

It would be helpful if the docs clearly stated whether multi-modal input is used for /learn, at least for some models.

Additional context

If /learn does not support multi-modal input, this would be a great feature to have.