ANN Search Operator: MilvusClient

author: junjie.jiangjjj

Desription

Search embedding in Milvus, please make sure you have inserted data to Milvus Collection.

Code Example

Please make sure you have inserted data into Milvus, refer to ann_insert.milvus_client and load the collection to memory.

from towhee import pipe, ops, DataCollection

p = pipe.input('text')  \
    .map('text', 'vec', ops.sentence_embedding.transformers(model_name='all-MiniLM-L12-v2'))  \
    .flat_map('vec', 'rows', ops.ann_search.milvus_client(host='127.0.0.1', port='19530', collection_name='text_db2', **{'output_fields': ['text']}))  \
    .map('rows', ('id', 'score', 'text'), lambda x: (x[0], x[1], x[2]))  \
    .output('id', 'score', 'text')

DataCollection(p('cat')).show()

# result:

from towhee import pipe, ops

# search additional info url:
from towhee import pipe, ops, DataCollection

p = pipe.input('text')  \
    .map('text', 'vec', ops.sentence_embedding.transformers(model_name='all-MiniLM-L12-v2'))  \
    .map('vec', 'rows', ops.ann_search.milvus_client(host='127.0.0.1', port='19530', collection_name='text_db2', **{'output_fields': ['text']}))  \
    .output('rows')

DataCollection(p('cat')).show()

Factory Constructor

Create the operator via the following factory method:

ann_search.milvus_client(host='127.0.0.1', port='19530', collection_name='textdb')

More Resources

• Tokopedia Achieved a 10x Smarter Search with Milvus: Indonesia's largest e-commerce platform, Tokopedia's quest for superior search functionality, led them to Milvus, a game-changer in semantic search.
• Spring AI and Milvus: Using Milvus as a Spring AI Vector Store - Zilliz blog: A comprehensive guide on how to use Milvus as a Spring AI vector store
• ArXiv Papers Vector Similarity Search with Milvus 2.1 - Zilliz blog: Run semantic search queries on ~640K papers in <50ms using Dask, SBERT SPECTRE, and Milvus Vector database
• Unlock Advanced Recommendation Engines with Milvus' New Range Search - Zilliz blog: Exploring Milvus’s newly released range search feature, how it differs from the traditional KNN search, and when to use it.
• Milvus 2.4 Unveils Game-Changing Features for Enhanced Vector Search - Zilliz Newsroom; Milvus 2.4 Unveils Game-Changing Features for Enhanced Vector Search: Milvus 2.4 Unveils Game-Changing Features for Enhanced Vector Search