Does Pinecone support multiple embeddings per ID? It would appear that given the upsert commands of:
index.upsert([
("A", [0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1]),
("A", [0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2]),
])
The second embedding would overwrite the first one.
The task I’m trying here is embedding multiple paragraph vectors for a single document (“A” being the document, in the example). If multiple embeddings per ID isn’t possible, what’s the recommended way to approach this? I can thing of encode the paragraph in the id, e.g. “A-0, A-1, A-2” and possibly in the metadata, but this doesn’t seem ideal, like if I want to delete or change all paragraphs for a document I wouldn’t have a consistent index of the IDs associated with a given doc.