search

0hq / tinyvector

A tiny nearest-neighbor embedding database built with SQLite and Pytorch. (In development!)
MIT License
772 stars 24 forks source link

Benchmark indexes + all operations #4

Open 0hq opened 1 year ago

0hq commented 1 year ago

We need to test the speed of the database on a range of vector dimensions and # of vectors, on all indexes. Adding some indexing functionality directly to the codebase would be best and integrating benchmarking as part of the testing library.