From the course: LLM Foundations: Vector Databases for Caching and Retrieval Augmented Generation (RAG)

Introduction to Milvus DB

From the course: LLM Foundations: Vector Databases for Caching and Retrieval Augmented Generation (RAG)

Start my 1-month free trial Buy for my team

Introduction to Milvus DB

“

Having discussed the concepts of vector databases in the previous section, we will now explore Milvus, a popular vector database. We will review key features and capabilities as well as set up the database and exercise files. What is Milvus? Milvus is a vector database. It is built for storing, indexing, and searching vectors. It is one of the top databases that is used today for generative AI. Its features are similar to other popular vector databases. Let's review some of the key features of Milvus. Milvus is available in both open source and commercial versions. We can download and install Milvus from open-source repositories, as well as purchase with additional support capabilities. Milvus can be deployed in several configurations. It can be installed as a standalone service for simple deployments. It can be set up as a cluster for scalability and fault tolerance. It's also available as a managed cloud service called Zilliz cloud. This database is highly scalable for storing and searching vectors. Clustering helps it to scale horizontally. To help with vector search, it supports Euclidean distance, inner product, and cosine matrix. This is essential for effective semantic search implementations. It provides hybrid data storage with support for scalar data types in addition to vectors. It also provides search capabilities combining both vector and scalar datatypes. Milvus has SDKs in popular languages like Python, Node.js, Go, and Java to help in building applications.

Contents