Hello! I am experimenting with typesense vector searches, created a collection with some documents using the OAI

openai/text-embedding-ada-002

model. Now I would like to perform a similarity search of a document (so, like explained here) But it appears for that to work, one can’t create the document(s) using the OAI integration as in that case, these documents do not contain a valid vector field (

vec

, generally) Yes, of course I can query the collection by

embedding

by passing lets say a summary, or even the whole document content I want to find similars of. It would likely return a similar response like the similarity search. I might misunderstand something else though? It would be truly great to be able to pass the Document ID to the embedding search and find similes thereof, without first having to retrieve that document, eventually summarise its contents and do a search on that