What's a good embedding model and search to run locally? something fast and lightweight.

FAISS is no longer close to SoTA:

https://ann-benchmarks.com/index.html https://vector-index-bench.github.io/ https://big-ann-benchmarks.com/neurips23.html

Wow! 4GB for 10 million documents. This means one could build a reverse index much faster than before and devx processes like debugging, performance testing would become much smoother. Can't wait for the sqlite bindings to come out!

It would be nice to have the README be a little more human written for a project where you actually want people to adopt it

This looks perfect for local, privacy first search, but since it's built in Rust, has anyone tried compiling it to WASM to run directly inside a browser extension?

Why not just use Qdrant? They've been integrating TurboQuant for months, works well.

people should read turboquant's open review comments: https://openreview.net/forum?id=tO3ASKZlok

If anyone is looking to retrofit to an existing pipeline, I use similar ideas to compress vectors for job search, getting roughly 8x compression with about a 3.5% drop in quality. My experiment: https://corvi.careers/blog/vector-search-embedding-compressi...

Well. That is insane. O_O Fantastic job!

Who is this co-author called t <t@t>?

what could i use this for as part of my agentic workflow? codebase indexing? docs ?

lancedb and duckdb integrations would be great...

Another vibe coded slop where they can't even spend time on Readme or documentation around code...

Bloviating nonsense, 3rd time I’ve seen something like this in HN since TurboQuant came out. You don’t need float32, never did. Source: I’ve been writing on device embedding code for 4 years.