Octen Reranker models
#3
by vvekthkr - opened
I am curious about the missing reranker model, Should we use the respectively sized qwen3 rerankers, if we were to use octen generated embeddings for hybrid search?
Yep, that works. And the rerank model doesn’t have to be the same size as the embedding model — they can be chosen independently.
Thanks for the inputs
vvekthkr changed discussion status to closed