discussion: Should we implement hybrid search in llama index integration?

[gaocegege]

https://github.com/run-llama/llama_index/blob/main/llama_index/vector_stores/postgres.py#L56

pgvector implements a hard-coded hybrid search approach.

            text_search_tsv = Column(  # type: ignore
                TSVector(),
                Computed(
                    "to_tsvector('%s', text)" % text_search_config, persisted=True
                ),
            )

I am not sure if we should support it too in https://github.com/run-llama/llama_index/blob/main/llama_index/vector_stores/pgvecto_rs.py