Vector Similarity Search at Scale // Dave Bergstein // MLOps Coffee Sessions #52
MLOps.community
00:00
The Engineering Challenges of Vector Search
When you go to larger a the moment i went to thousands, it slowed down significantly. So that's when you have to move into a proximate nearest neighbor and you want to get good result within a fast amount of time. The most the top algritm is face s from a facebook. Facebook uses this technology. And they really innovated in right the development of face,. A, there are others, th h and s w is a popular alganom now. It's part of face, too. But the next challenge that is at case s - using these open sorce libraries. Even in a small case, even if you could do it on your one lap toper
Play episode from 20:53
Transcript
