AILab/bbit_ai/test/milvus/chainMain.ipynb

In [21]:

from langchain_milvus import BM25BuiltInFunction, Milvus
URI = "http://10.10.10.9:19530"
tongyiKey = "sk-9464b2498c184982a9fe9d2c2e725ab5"
from langchain_community.embeddings import DashScopeEmbeddings
embeddings = DashScopeEmbeddings(
    model="text-embedding-v3",
    dashscope_api_key= tongyiKey, 
)
vectorstore = Milvus(
    embedding_function=embeddings,
    connection_args={"uri": URI, "token": "root:Milvus", "db_name": "bbit_ai_lab"},
    collection_name="knowledge",
    index_params={"index_type": "FLAT", "metric_type": "L2"},
    consistency_level="Strong",
    auto_id=True,

    primary_field = "id",
    text_field="text",
    vector_field="vector",
    partition_key_field = "kn_id",
    enable_dynamic_field = True,
    drop_old=False,  # set to True if seeking to drop the collection with that name if it exists
)

from langchain.schema import Document

docs = [
    Document(
        page_content="这是第一条文本",
        metadata={
            "kn_id": "8ecd1179-4194-4b80-bc39-5addc678df4b",
            "is_active": True,
        }
    ),
    Document(
        page_content="这是第二条文本",
        metadata={
            "kn_id": "8ecd1179-4194-4b80-bc39-5addc678df4b",
            "is_active": True,
        }
    )
]

vectorstore.add_documents(docs)

Out [21]:

[460823023525530114, 460823023525530115]

In [ ]:

results = vectorstore.similarity_search(
    "",
    k=2,
    expr='kn_id == "8ecd1179-4194-4b80-bc39-5addc678df4b"',
)
for res in results:
    print(f"*{res.page_content} [{res.metadata}]")

*这是第一条文本 [{'kn_id': '8ecd1179-4194-4b80-bc39-5addc678df4b', 'id': 460823023525530108, 'is_active': True}]
*这是第一条文本 [{'kn_id': '8ecd1179-4194-4b80-bc39-5addc678df4b', 'id': 460823023525530110, 'is_active': True}]