Hey, how can I exclude metadata with None values when sending the node content to GPT? For example, I have a "name" attribute as metadata, and nodes without a specified name have an empty value for this attribute (None by default).
So, sometimes, when I ask about the name and the retriever extracts nodes with name = None, GPT returns None instead of "I don't know," for example.
By the way, I'm using version "0.6.31."

Does anyone gets embedding=None with the new versions when outputing the retriever.retrieve(question)? with 0.6.15 I used to get the list of embeddings but not anymore:
[NodeWithScore(node=Node(text='....', embedding=None, ....), score=None)]

Hey, can we know which nodes are used by the model to create the answer, or do we need a custom prompt for that?

Hey, when I add response_synthesizer to RetrieverQueryEngine to filter chunks by similarity_cutoff, I always get None, and checked the results of retriever.retrieve(question) and there are always chunks returned

Hey, I think the doc_ids parameter isn't functioning correctly in the VectorIndexRetriever. When I send doc_ids=['doc_id1'], it returns chunks with different doc_ids. Did anyone use it ?

Hey, what is the default prompt we're using for search ( exp in here : https://gpt-index.readthedocs.io/en/latest/guides/primer/usage_pattern.html#low-level-api )

Hey, is it possible to insert documents one by one into a ComposableGraph?