Reranker models

#145
by GuillaumeGrosjean - opened

When trying to filter cross-encoder models (rerankers), it seems that none of the top models are actually cross-encoders. Is there a way to get a ranking of best cross-encoder models only ?

Seconded! This feels like a huge omission. I see there is a related github issue https://github.com/embeddings-benchmark/mteb/issues/1214

Sign up or log in to comment