课程: Building in Azure AI Foundry

免费学习该课程！

今天就开通帐号，24,700 门业界名师课程任您挑！

Apply evaluation metrics

Apply evaluation metrics

课程: Building in Azure AI Foundry

开始一个月试用 购买团队方案

Apply evaluation metrics

“

- [Instructor] Let's look at how to use industry standard metrics to evaluate your AI application in Azure AI foundry. Here's my demo project in the actual AI foundry portal. Click evaluation in the menu. On the automated evaluations, click create a new evaluation. I can create an evaluation with some industry standard metrics. For this demo, I'll evaluate the prompt flow of California tours, which we created in the previous video. Select the prompt flow. Enter an evaluation name under select flow to evaluate, I will choose the prompt flow we built previously. Click next. To configure test data, you can use the existing data set or add your data set. I can download a data template, open the template. It's a JSON lines file for entering my test data. Here are the sample data sets for our demo, choose add your dataset, then click upload file. Open my test data file. On the dataset mapping for prompt flow, I can configure the required inputs for my prompt flow. For the chat history, set…

内容