Metadata based filtering for similarity search (any matches)

[NaveenKinnal]

Hello,
I am using FAISS similarity search using metadata filtering option to retrieve the best matching documents.

The legacy way is to retrieve a non-calculated number of documents and filter them manually against the metadata value.

However, I came across the in-built metadata based search option which does this

if filter is not None: filter = { key: [value] if not isinstance(value, list) else value for key, value in filter.items() } if all(doc.metadata.get(key) in value for key, value in filter.items()): docs.append((doc, scores[0][j])) else: docs.append((doc, scores[0][j]))

which matches the document key-value metadata pairs against the input filter key-value pairs and only returns if all the metadata values are matching - (all condition).

I am looking for a feature that can retrieve a document even if there is at least one key-value match - (any condition). Is it possible to develop this?