I know there are notebooks that show you how to do hybrid search. My question is about getting them into production. We can use the BM25 model, and fit it using the .fit() method, but is there a way I can save the fitted model so that I can use it again?
To save the BM25 params you can call .dump("file_path.json"), see example here. While BM25 is great in many cases, we would like to note that we recently released an advanced sparse model called pinecone-sparse-english-v0 that potentially could perform better, see docs here. Please let us know if you have any further questions.
Amnon
Thanks for the reply. Quick question, I thought the pinecone sparse model was not recmmened for production use cases yet