Andrew Kane

Hi @nathanwilk7, the most helpful thing would be sharing specific use cases for either data type (how the vectors are generated and what specifically they're used for). These may be...

Thanks @nathanwilk7, this is really great context (and a great explanation)! I really appreciate it. I agree that sparse vectors are probably a better option for 1. I'd like to...

There's now a new [halfvec branch](https://github.com/pgvector/pgvector/compare/halfvec) for this for anyone who wants to try it (in a non-production environment). ```sql CREATE TABLE items (id bigserial PRIMARY KEY, embedding halfvec(3)); INSERT...

Yeah, think that could be a common use case. Pushed an update for casting between `vector` and `halfvec` (meant to include it earlier). This allows you to do: ```sql CREATE...

Support for half vectors is now merged (32a502c838b111078ee8d25995446afb00ba38e6) and will be included in 0.7.0 (#508) 🎉

Support for sparse vectors (abac7a3f776d4edbb423a000ba5234d3e8eab465) and bit vectors (94a444f02958f04946fc2dfccf0e10fe3718b660) is merged as well

Just fyi for anyone testing on x86-64: some key `halfvec` functions now use intrinsics, so you should see much better performance 🚀

Hey @pashkinelfe, thanks for testing/sharing. Which commit hash are you using?

Hi @ataudt, there's not a function for this right now, but I've added support in the [normalize_l2 branch](https://github.com/pgvector/pgvector/compare/normalize_l2). Can you share more about the use case for having this in...

Quick update: there's a good chance this function will be needed to make indexes work with more data types, so it may be included in 0.6.0.