fp.
newest
Open in hackernews
Compress Vectors by 4x by using 8-bit Rotational Quantization
https://weaviate.io/blog/8-bit-rotational-quantization
3
•
bobvanluijt
•
6h ago