_______ __ _______
| | |.---.-..----.| |--..-----..----. | | |.-----..--.--.--..-----.
| || _ || __|| < | -__|| _| | || -__|| | | ||__ --|
|___|___||___._||____||__|__||_____||__| |__|____||_____||________||_____|
on Gopher (inofficial)
(HTM) Visit Hacker News on the Web
COMMENT PAGE FOR:
(HTM) Show HN: DuckDB community extension for prefiltered HNSW using ACORN-1
edinetdb wrote 6 hours 31 min ago:
This is great for analytical workloads. I work with financial time
series data (Japanese company filings) and have been using BigQuery
with in-memory caching for the hot path. Curious whether DuckDB
extensions like this could replace the BQ dependency for smaller
datasets â the cold start + query cost model of serverless warehouses
can be painful for API-serving use cases.
jarpineh wrote 21 hours 29 min ago:
As an aside, there's now Lance data format support in DuckDB through
their extension. It has Lance's vector search support available among
other things: [1] I just noticed this, and your post, and haven't yet
checked neither (sorry). I'm however doing some vector search
benchmarking soon, with DuckDB's options alongside others. So your work
caught my attention here.
(HTM) [1]: https://github.com/lance-format/lance-duckdb/tree/main?tab=rea...
swaminarayan wrote 23 hours 16 min ago:
Does your method work better than standard ANN when filters are very
strictâand how does it affect speed vs accuracy?
wolfgangK wrote 1 day ago:
Nice !
My most pressing request for VSS would be efficient binary vectors : is
this on the table ?
cigrainger wrote 1 day ago:
I haven't given binary vectors a lot of thought, but I'm exploring
RaBitQ[1]
(HTM) [1]: https://arxiv.org/abs/2405.12497
esafak wrote 1 day ago:
Please upstream it.
(DIR) <- back to front page