Great question — answering as the UphillTech team:
So, the solution, according to what @morry said is to duplicate your collection with a new metadata: import chromadb client = chromadb.Client() collection = client.get_or_create_collection(name = "my_collection", metadata={"hnsw:space": "cosine", "other_data": "test1"}) new_collection_name = 'cosine' new_collection = client.get_or_create_collection( new_collection_name, metadata={"hnsw:space": "cosine", "other_data": "test2"} ) existing_count = collection.count() batch_size = 64 for i in range(0, existing_count, batch_size): batch = collection.get(include=["metadatas", "documents", "embeddings"], limit=batch_size, offset=i) new_collection.add( ids=batch["ids"], documents=batch["documents"],...
Full answer by ronkov: https://stackoverflow.stackexchange.com/a/79214661 (CC BY-SA 4.0).
Need it done for your business? UphillTech.