From the course: Azure for Developers: Retrieval-Augmented Generation (RAG) with Azure AI
Join today to access over 25,200 courses taught by industry experts.
Evaluation metrics in generative AI - Azure AI Services Tutorial
From the course: Azure for Developers: Retrieval-Augmented Generation (RAG) with Azure AI
Evaluation metrics in generative AI
- [Instructor] Evaluation plays a very important role in developing a RAG solution as it builds trust and confidence in our application and ensures that responses are correct and behaving properly. Azure has a comprehensive approach to evaluation, which includes three categories, risk and safety evaluators which evaluate potential risks associated with AI-generated content with varying degrees of severity, performance and quality evaluators which involves assessing the accuracy, groundedness, and relevance of generated content using robust AI-assisted and natural language processing metrics, and custom evaluators tailored to meet specific needs and goals. This document shows the definition of each different evaluator under each category. Scrolling down further, you will then see the definition of how each evaluator is scored. You may want to read this document further to determine which one is appropriate for your business need. We will be using the Azure AI evaluation SDK to evaluate…
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.