Milvus Milvus With Jina

Milvus With Jina

image-searchvector-databasesemantic-searchIntegrationmilvusembeddingsunstructured-dataquestion-answeringLLMmilvus-bootcampdeep-learningimage-recognitionimage-classificationaudio-searchPythonragNLP

alph-notebooks/milvus-bootcamp / milvus_with_Jina.ipynb

Export

Run Notebooks

Contents

No cells yet

Add cells to see them here

Integrate Milvus with Jina AI

This guide demonstrates how to use Jina AI embeddings and Milvus to conduct similarity search and retrieval tasks.

Who is Jina AI

Jina AI, founded in 2020 in Berlin, is a pioneering AI company focused on revolutionizing the future of artificial intelligence through its search foundation. Specializing in multimodal AI, Jina AI aims to empower businesses and developers to harness the power of multimodal data for value creation and cost savings through its integrated suite of components, including embeddings, rerankers, prompt ops, and core infrastructure. Jina AI's cutting-edge embeddings boast top-tier performance, featuring an 8192 token-length model ideal for comprehensive data representation. Offering multilingual support and seamless integration with leading platforms like OpenAI, these embeddings facilitate cross-lingual applications.

Milvus and Jina AI's Embedding

In order to store and search these embeddings efficiently for speed and scale, specific infrastructure designed for this purpose is required. Milvus is a widely known advanced open-source vector database capable of handling large-scale vector data. Milvus enables fast and accurate vector(embedding) search according plenty of metrics. Its scalability allows for seamless handling of massive volumes of image data, ensuring high-performance search operations even as datasets grow.

Examples

Jina embeddings have been integrated into the PyMilvus model library. Now, we will demonstrate code examples to show how to use Jina embeddings in action.

Before we start, we need to install model library for PyMilvus.

[ ]

If you are using Google Colab, to enable dependencies just installed, you may need to restart the runtime (click on the "Runtime" menu at the top of the screen, and select "Restart session" from the dropdown menu).

General-Purpose Embedding

Jina AI's core embedding model, excels in understanding detailed text, making it ideal for semantic search, content classification thus supports advanced sentiment analysis, text summarization, and personalized recommendation systems.

[ ]

Bilingual Embeddings

Jina AI's bilingual models enhance multilingual platforms, global support, and cross-lingual content discovery. Designed for German-English and Chinese-English translations, they foster understanding among diverse linguistic groups, simplifying interactions across languages.

[ ]

Code Embeddings

Jina AI's code embedding model provides searching ability through code and documentation. It supports English and 30 popular programming languages that can be used for enhancing code navigation, streamlined code review and automated documentation assistance.

[ ]

Semantic Search with Jina & Milvus

With the powerful vector embedding function, we can combine the embeddings retrieved by utilizing Jina AI models with Milvus Lite vector database to perform semantic search.

[ ]

As for the argument of MilvusClient:

Setting the uri as a local file, e.g../milvus.db, is the most convenient method, as it automatically utilizes Milvus Lite to store all data in this file.

If you have large scale of data, you can set up a more performant Milvus server on docker or kubernetes. In this setup, please use the server uri, e.g.http://localhost:19530, as your uri.

If you want to use Zilliz Cloud, the fully managed cloud service for Milvus, adjust the uri and token, which correspond to the Public Endpoint and Api key in Zilliz Cloud.

With all data in Milvus vector database, we can now perform semantic search by generating vector embedding for the query and conduct vector search.

[10]

{'id': 1, 'distance': 0.8802614808082581, 'entity': {'text': "The Dartmouth Conference in 1956 is considered the birthplace of artificial intelligence as a field; here, John McCarthy and others coined the term 'artificial intelligence' and laid out its basic goals.", 'subject': 'history'}}

Jina Reranker

Jina Ai also provides rerankers to further enhance retrieval quality after searching using embeddings.

[ ]

[RerankResult(text="The Dartmouth Conference in 1956 is considered the birthplace of artificial intelligence as a field; here, John McCarthy and others coined the term 'artificial intelligence' and laid out its basic goals.", score=0.9370958209037781, index=1),
, RerankResult(text='The invention of the Logic Theorist by Allen Newell, Herbert A. Simon, and Cliff Shaw in 1955 marked the creation of the first true AI program, which was capable of solving logic problems, akin to proving mathematical theorems.', score=0.35420963168144226, index=3),
, RerankResult(text="In 1950, Alan Turing published his seminal paper, 'Computing Machinery and Intelligence,' proposing the Turing Test as a criterion of intelligence, a foundational concept in the philosophy and development of artificial intelligence.", score=0.3498658835887909, index=0),
, RerankResult(text='In 1951, British mathematician and computer scientist Alan Turing also developed the first program designed to play chess, demonstrating an early example of AI in game strategy.', score=0.2728956639766693, index=2)]