Nomic vision embeddings

[italianconcerto]

Hi,
I see you integrated the nomic text models. I think it would be great to also add the vision models, so that it's possible to perform both text and vision queries on qdrant.

Thanks!