Compress Vectors by 4x by using 8-bit Rotational Quantization

3 points| bobvanluijt | 6 months ago |weaviate.io