Vector Database for Large Language Models in Production

preview_player
Показать описание
Large Language Models (LLMs) have lowered the bar for getting a product to market however they are "not (always) large enough" in that they don't always contain all the information they need to solve a certain task.
This is where Vector Databases come in!

In this talk Sam Partee, Principal Applied AI Engineer at Redis covers topics such as:
Vector Embeddings, Vector Similarity Search, what is a Vector Database, some common design patterns using Vector Databases and LLMs, what to consider when building a system using Vector Databases and some example uses cases for Vector Databases.

Speaker:
Sam Partee, Principal Applied AI Engineer, Redis

Рекомендации по теме
Комментарии
Автор

Storing embeddings is really great. I was doing serialization earlier and loading from disk on app startup so i can find cosine similarity easily. Another great thing is indexing which will make the search efficient. Great Presentaion.

trifforttv