Vector Database for Rapid Prototyping - Milvus Lite

Vector databases are used widely in AI enabled applications, but can also be used for general purposes applications.
One of the most prominent use of vector database is to do a similarity search using cosine similarity or by using euclidian or hamming distance.

Now vector databases are not the only way to do a similarity search, it can also be done by using K nearest neighbour or using data structures.

However if you'd like to consider a vector database for similarity search and want to do some prototyping, Milvus Lite vector database is one such database which allows you to do a rapid prototyping and see the results.

This video is all about understanding the vector database and how we can solve the problem of similarity search using the vector database. The example shown here is by using the Milvus Lite database but the same can be used with any other vector database.

The use of Milvus Lite database demonstrates the usage of vector database. For vector embedding, I've used the sentence transformers with the dimension of 384 for demonstration purposes.

Here is the chapter of this particular video.

Timecodes
00:00 Vector Database Usage
00: 13 : Vector database similarity search
00:58 : Milvus Lite Vector Database
01:25 : The Problem Statement
02:33 : Vector embedding using Sentence Transformers
04:09 : Vector dimensions
05:34 : Creating a Milvus Lite Vector Database and Schema
06:21 : Vector Database Primary Key and other fields
07:50 : Storing vector embeddings in the vector database
08:26 : Creating Index parameters of the vector database
09:46 : Creating a collection in the vector database
10:21 : Similarity Search in Vector Database
11:59 : Reusing the vector database for similarity search


#vectordatabase #milvus #artificialintelligence

#TheTheoryOfCode