cohere-embed-english-v3.0 as reranker
#3
by
austinmw
- opened
Hi, I noticed that this model is listed on the MTEB Reranker leaderboard. Does this mean that this embedding model can also be repurposed as a reranking model? And if so, are there any examples of how to do this?
My assumption is that you would:
- embed the query
- embed each retrieved document
- calculate the cosine similarity between the embedded query and each embedded document
- sort the retrieved documents according to these similarity scores
Or am I misunderstanding something? Thanks!
Your understanding is correct, this is how MTEB measures the reranking task