Comment by visarga
2 years ago
current day version - if you got less than 10K - 100K vectors use numpy.dot instead of the vector store databases
2 years ago
current day version - if you got less than 10K - 100K vectors use numpy.dot instead of the vector store databases
Even if you have a lot more than that you could easily use SQLite with FAISS. It works great.
Bingo. This combination is underappreciated.
Yes, I’ve used this approach in my Swiss Army Llama project with huge numbers of vectors and it can scale massively. Also it’s free! These vector db as a service companies charge insane prices for this! It really does feel like snake oil to me.
https://github.com/Dicklesworthstone/swiss_army_llama
You can easily add two zeros to that.