Hi all, I'm building a small search tool for my documents. Currently just using the simple retriever with VectorStoreIndex. I'm curious how is everyone handling pagination? I'm able to get top k, but what should I do if I want more? Should I just increase the number of retrieved nodes and do a post pagination on the python side? or is there a better way to handle this? (but if I also want to use cohere reranker, I guess it'll be expensive/slow to give it tons of nodes?)
oh, I mean if I have tons of documents, each time it'll only return, say top 10, what if user want to load more? I could do top 20, and only return the 11th-20th, but curious if there are better way to do this.
