Hi everyone. When using weaviate, does llama index support custom vectorIndexConfig parameters such as efConstruction, either by specifying them directly, or by using an existing schema during index construction, rather than creating a new one?
Other than the graph, which indices, if any, need to be saved to and subsequently loaded from disk? Does the graph contain everything needed for querying the full structure?
Hi everyone. I'm just getting started with this, and I've got it up and running, however I'm having an issue with the "create and refine" process. Sometimes I get a response like "Return the original answer. The new context is not relevant..." I've looked through the documentation but was unable to find any way to capture the intermediate responses that are being generated.
