The Ettin Reranker Family
The Ettin Reranker Family has been introduced, featuring six new state-of-the-art Sentence Transformers CrossEncoder rerankers. These models are built on the Ettin ModernBERT encoders and are designed for improved relevance scoring in information retrieval tasks. The release includes training recipes and usage instructions for those interested in implementing or training their own models.
- ▪The new rerankers include models with sizes ranging from 17 million to 1 billion parameters.
- ▪They utilize a distillation recipe for training, enhancing their performance on retrieval tasks.
- ▪The models can be easily integrated into existing workflows with minimal code.
Opening excerpt (first ~120 words) tap to expand
Back to Articles Introducing the Ettin Reranker Family Published May 19, 2026 Update on GitHub Upvote 38 +32 Tom Aarsen tomaarsen Follow TL;DR Table of contents What is a reranker, and why pair one with an embedder? Usage End-to-end retrieve-then-rerank pipeline Architecture Details Results MTEB(eng, v2) Retrieval Speed Training Distillation recipe Dataset Training Arguments Evaluation Overall Training Script Conclusion Acknowledgements Citation TL;DR Today I'm releasing six new Sentence Transformers CrossEncoder rerankers, state-of-the-art at their respective sizes, built on top of the Ettin ModernBERT encoders, together with the data and full training recipe that produced them: cross-encoder/ettin-reranker-17m-v1 cross-encoder/ettin-reranker-32m-v1 cross-encoder/ettin-reranker-68m-v1…
Excerpt limited to ~120 words for fair-use compliance. The full article is at Huggingface.
Discussion
0 commentsMore from Huggingface
