Explore best practices for DP-100: Designing and implementing a data science solution on Azure learning path 7

Completed

DP-100: Learning Path 7 Train models with scripts in Azure Machine Learning

Learning Path 7 Overview

As you begin to teach this learning path, get familiar with what the students will learn during the learning path. In this learning path students will learn how to run the training script as a command job to help prepare production-ready script. Students will also learn about common terminologies like the code, command, environment, and compute.

This learning path consists of four focus areas:

• Converting your code to a script, parameters
• Command job configuration and terminology
• Run a training script as a command job in Azure Machine Learning
• Track model training with MLflow in jobs

Learning Path 7 Tips and Tricks

• Illustrate a completed experiment to show status, logs, and outputs
• Emphasize the difference between logging modes and make sure it's clear before moving forward

Learning Path 7 Demos (optional)

• Show where outputs can be found
• Compare both outputs: in a notebook and in the user interface (UI)
• Show where saved artifacts are located in a local file folder and in Blob storage

Learning Path 7 Labs

• Lab 7.1: Run a training script as a command job in Azure Machine Learning
• Lab 7.2: Use MLflow to track training jobs