Hi, Currently I have a collection with about 50M documents with some text on them. Solr fields are DocumentType, DocumentText and DocumentDate. Sometimes DocumentText has long information so I need to split it into chunks for better search experience.
I would like to start indexing embeddings for those documents so my question is about how to do it. Should I create different documents or keep the orifginal document and create nested documents with the chunks. Thanks a lot. Sergio
