[QUESTION]: 请问hybrid_search中如果有sparse vector的检索，支持partition key吗

[lvjg]

Is there an existing issue for this?

• [X] I have searched the existing issues

What is your question?

请求异常，返回内容如下： pymilvus.exceptions.MilvusException: <MilvusException: (code=1100, message=Invalid expr: id in ['2e65e219-48d8-44b2-8c8b-a528076b5c54', ...]: invalid parameter)>

Anything else?

No response

[lvjg]

res = self._collection.hybrid_search( [dense_req, sparse_req], rerank=ranker, limit=query.hybrid_top_k, output_fields=output_fields, group_by_field=group_by_field )

[lvjg]

在hybrid_search中也没有使用expr这个参数

[XuanYang-cn]

https://milvus.io/docs/multi-vector-search.md @lvjg Please refer to this doc. The params of your hybrid_search AIP seems wrong, we need a list of AnnSearchRequest

[XuanYang-cn]

/assign @lvjg

[lvjg]

https://milvus.io/docs/multi-vector-search.md @lvjg Please refer to this doc. The params of your hybrid_search AIP seems wrong, we need a list of AnnSearchRequest

        sparse_emb = self.sparse_embedding_function.encode_queries(
            [query.query_str]
        )[0]
        sparse_search_params = {
            "metric_type": "IP",     
            "params": {
                "radius": 0.1,
                "range_filter": 1.0
            }
        }
        sparse_top_k = query.sparse_top_k or 20
        sparse_req = AnnSearchRequest(
            [sparse_emb],
            self.sparse_embedding_field,
            sparse_search_params,
            limit=sparse_top_k,
        )            

        dense_emb = query.query_embedding
        dense_req = AnnSearchRequest(
            [dense_emb],
            self.embedding_field,
            dense_search_params,
            limit=similarity_top_k,
        )

[lvjg]

使用的是一组AnnSearchRequest