Open jasonjmcghee opened 5 months ago
So to clarify, do you mean adding a unified search that combines searching over both:
Potentially, definitely. Maybe there’s specifically a text embedding and/or an image embedding and/or a combination of both (like the model mentioned).
Depends what works.
And the medium of search might be text and/or some sort of image similarity interface.
Lots of cool stuff that we can do here...