Weaviate Podcast cover image

Charles Frye on Full Stack Deep Learning - Weaviate Podcast #57!

Weaviate Podcast

CHAPTER

How to Design a Language Model for Different Sources

When you have multiple sources if you there's like a certain scale at which you can no longer be very precious about your handling of sources and you have to be very generic. I think you end up getting like worse results if all you're doing is saying like documents are going to be like annotated with what are their h1 and h2 headings but that means something very different in different websites. vector storage as an index into this information so an index in a database is something like we've already stored the data and then we add an index right  so we might add an index that's prefix reg X based, for example. This is a pre-compute a bunch of

00:00
Transcript
Play full episode

Remember Everything You Learn from Podcasts

Save insights instantly, chat with episodes, and build lasting knowledge - all powered by AI.
App store bannerPlay store banner