Redis As A Vector Database: Billion-Scale Fast Vector Similarity Search with RediSearch

Подписаться 7 тыс.

Просмотров 3,4 тыс.

50% 1

In this exciting video, we explore the power of Redis, a versatile key-value store, and its remarkable RediSearch module. We dive into the world of high-dimensional data and witness the incredible speed and accuracy of the Hierarchical Navigable Small World (HNSW) algorithm, enabling us to perform approximate nearest neighbor searches in the blink of an eye. Redis takes center stage as a vector database, efficiently storing and retrieving vast amounts of vector data. We embark on a journey to uncover the magic behind Redis' lightning-fast search capabilities, enabling us to locate the nearest neighbors of a given query vector among billions of others in just milliseconds.
But that's not all! Our exploration wouldn't be complete without the assistance of the deepface library for Python. We witness the transformation of facial images into vector embeddings, opening up a whole new realm of possibilities for face recognition and analysis. Join us on this captivating adventure as we unravel the mysteries of Redis, RediSearch, HNSW, and deepface, showcasing their incredible synergy and pushing the boundaries of speed and efficiency in vector-based search and analysis. Get ready to be amazed by the power of technology and its limitless potential in the world of data exploration.
Tutorial: sefiks.com/202...
If you like this video, you may like these too:
Fast Vector Similarity Search with Facebook Faiss: • Billion-scale Fast Vec...
Fast Vector Similarity Search with Spotify Annoy: • Billion-scale Fast Vec...
Fast Vector Similarity Search with ElasticSearch: • Billion-scale Fast Vec...
Fast Vector Similarity Search with NMSLIB: • Billion-scale Fast Vec...
Please Subscribe! That's what keeps me going ► bit.ly/40NfIS7
Want more? Connect with me here:
Blog: sefiks.com/
Twitter: / serengil
Threads: www.threads.ne...
Instagram: / serengil
Facebook: / sefikscom
Linkedin: / serengil
If you do like my videos, you can support my effort with your financial contributions on
Patreon: www.patreon.co...
GitHub Sponsors: github.com/spo...
Buy Me a Coffee: buymeacoffee.c...

Опубликовано:

21 окт 2024

Ссылка:

Скачать:

Готовим ссылку...

Добавить в:

Мой плейлист

Посмотреть позже

Комментарии : 13

@BulentSiyah Год назад

Thank you for this amazing video! Redis and RediSearch, along with the HNSW algorithm and deepface library, create an impressive synergy, unleashing lightning-fast search capabilities and opening new possibilities in facial recognition and data analysis. Mind-blowing technology with limitless potential!

@nhunter77018 Месяц назад

Thank you for this very informative video. I have a question. Once I have created the index in Redis, will Redis maintain the index in the background, or is it necessary to rebuild the index periodically?

@sefiks Месяц назад

Recreating the index is required

@chenjus 9 месяцев назад

I keep getting unknown command 'FT.SEARCH', with args beginning with:... Did you also get this error?

@sefiks 9 месяцев назад

May be related to your redis version

@mertcobanov Год назад

❤

@omarm.k.shehada3959 8 месяцев назад

What if we want to make the k dynamic? To find all the matching

@sefiks 8 месяцев назад

then set k to the n where n is the number of instances.

@abdulwahabkhan1086 5 месяцев назад

But wouldn't it take linear time?

@TheAwedExplorer Год назад

Hi sefik. I want to use deepface library to do face verification. I think my database has about 500 differrent faces in it. Can deepface do this job in a 1 second? i mean the speed is very imortant for the job.

@sefiks Год назад

Yes, that can be done in milliseconds. You should store vector embeddings for your database beforehand. Running its find function handles this. Then, you can search even in larger database in milliseconds.

@TheAwedExplorer Год назад

@@sefiks Thank you very much sir