600% memory footprint reduction on Vector Embeddings Quantization(github.com)4 points by betto 318 days ago | 1 comment