Granite Embedding Models
Granite Team · 2025
Read Granite Embedding Models by Granite Team (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2502-20204.txt.
Description
Members of IBM's Granite team introduce, in a 2025 report, a family of text-code models built for search. Such models turn text into number lists that a search tool can match, and the family spans tight and loose match styles, in English and in many tongues. Slim six-layer forms sit beside fuller twelve-layer ones.
The models were raised with search-first early training, hard pair drills, knowledge poured from big kin, and blends of trained forms. In tests they beat like-sized open models on the team's own search tasks and held pace on public test sets. All are free under an open licence for study and trade use. The team built the data to a grade fit for firms.
Catalog record
| CREATOR | Granite Team |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 46 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2502-20204