From the course: Azure for Developers: Retrieval-Augmented Generation (RAG) with Azure AI
Join today to access over 24,600 courses taught by industry experts.
Azure OpenAI embeddings model - Azure AI Services Tutorial
From the course: Azure for Developers: Retrieval-Augmented Generation (RAG) with Azure AI
Azure OpenAI embeddings model
- [Instructor] In the previous videos, we discussed how embeddings are important in RAG. Our documents are stored in vector form in our database, and user queries need to be converted into vector form to search for documents in the knowledge base. An embedding model is needed for text to be converted into numerical vectors. In this course, we'll be using the Azure OpenAI embed model to convert both our documents and user queries into vector form. As of this recording, we currently have three models. Text-embedding-3-large, text-embedding-3-small, and text-embedding-ada-002. The newer embedding three models perform better than the older ada-002 model, so it's recommended that you use the newer ones. They also have different dimensions. The older version, ada-002, has a size of 1,536 dimensions, and they cannot be modified. Text-embedding-3-small will have 1,536 dimensions while text-embedding-3-large will have 3,072. Both the new models will have the option to lower their dimension…
Download courses and learn on the go
Watch courses on your mobile device without an internet connection. Download courses using your iOS or Android LinkedIn Learning app.