Creating Vector Embeddings: TensorFlow universal-sentence-encoder and Node.js
Learn how to build a movie recommender with TensorFlow, Postgres, PGVector, Javascript & Next.js. This is a series of videos where we build a project together step by step. Check the complete playlist here - • TensorFlow, Postgres, PGVector & Next... .
In this chapter you’ll learn how to use TensorFlow universal-sentence-encoder to create vector search embeddings in Node.js with Javascript. By following the steps in this video you’ll create a script that loads a model, takes a movie plot example and encodes it into a vector embedding.
You can find helpful links and code snippets in this accompanying article - https://aiven.io/developer/building-a....
Register with Aiven to host Postgres for free and get extra credits: https://go.aiven.io/get-pgvector
Chapters:
00:00 About the dataset that we'll use
00:28 Where to find all links and the dataset
01:54 How vector search works
02:47 How to create embeddings using TensorFlow model
03:36 Installing libraries
04:50 Writing code to create a vector from a single movie description
07:58 Running the code
08:48 Next steps
ABOUT AIVEN
Aiven’s cloud data platform helps your business reach its highest potential by making your data work for you.
It provides fully managed open source data infrastructure on all major clouds, helping developers focus on what they do best: innovate and create without worrying about the limitations of technology.
We like to think that Aiven is not only a cloud data platform but also an extension of your team. We are dedicated to helping you to succeed by removing barriers and finding the right solutions – with the help of the best data technology there is.
Sign up for a free Aiven account: https://go.aiven.io/youtube-signup
CONNECT WITH US
Website: http://aiven.io
LinkedIn: / aiven
GitHub: https://github.com/aiven
Twitter: / aiven_io