mteb-sgpt-bloom-7b-parameters

IN premise — summaries/2026/08/24/muennighoff-2022-mteb-s4-additional-modalities-text-embeddings-are-chunk-2.md

Created 2026-08-25T02:58:20+00:00

SGPT-BLOOM-7.1B-msmarco is a 7.1B-parameter BLOOM-based retrieval model evaluated in MTEB multilingual benchmarks.

Summary

This identifies SGPT-BLOOM-7.1B-msmarco as a medium-sized (7.1 billion parameter) retrieval model built on the BLOOM architecture and tested across multiple languages in the MTEB benchmark suite. It matters because every downstream claim about its performance, capabilities, or architectural properties in this system rests on correctly understanding what kind of model it is and where it has been measured.