Accelerating Large-Scale Inference with Anisotropic Vector Quantization · HackerPapers discussions

Accelerating Large-Scale Inference with Anisotropic Vector Quantization — mentioned by peter_l_downs

Wed, 29 May 2024 19:55:20 GMT

> JVector, the library that powers DataStax Astra vector search, now supports indexing larger-than-memory datasets by performing construction-related searches with compressed vectors. This means that the edge lists need to fit in memory, but the uncompressed vectors do not, which gives us enough headroom to index Wikipedia-en on a laptop. It's interesting to note that JVector accomplishes this differently than how DiskANN described doing it. My understanding (based on the links below, but I didn't read the full diff in #244) is that JVector will incrementally compress the vectors it is using…

Paper and discussion history. A mention is not an endorsement.

Accelerating Large-Scale Inference with Anisotropic Vector Quantization — mentioned by contravariant

Thu, 30 Jul 2020 10:44:46 GMT

In case anyone else was confused why they didn't just project onto q first if they want to have <q,x'> as close as possible to the original <q,x>. Apparently q is a random query vector and their research shows how to define an alternative quantization loss where you weigh quantization loss more heavily when |<q,x>| is high [1]. It's worth noting that without this additional weighting you end up with the euclidean distance under the assumption that q is distributed symmetrically. [1]: https://arxiv.org/abs/1908.10396

Paper and discussion history. A mention is not an endorsement.

Accelerating Large-Scale Inference with Anisotropic Vector Quantization — mentioned by greenyoda

Tue, 07 Jul 2020 00:34:51 GMT

If someone is looking for the referenced paper (Accelerating Large-Scale Inference with Anisotropic Vector Quantization), it can be found here: https://arxiv.org/abs/1908.10396 (It's also linked in the repository's docs/algorithms.md, to which there's a link at the bottom of the main page.)

Paper and discussion history. A mention is not an endorsement.