Microsoft · AI-300
Training and experiment tracking
AI-300 · Lesson 4 of 15 · 26:47
YouTube loads only after you press play. This chapter opens at its timestamp in the full course. Watch on YouTube ↗
In this lesson
Prepare for the AI-300 exam (Microsoft Certified: Machine Learning Operations Engineer Associate). In this lesson: notebooks and command jobs, MLflow tracking with autolog, log_param, and log_metric, automated machine learning with mltable data, guardrails, and limits, and comparing runs in the studio or with search_runs. Includes three original practice questions.
Preview the topics
- A training script must record its regularization rate and its final accuracy, so runs can be compared later. Which MLflow calls fit?
- An AutoML run warns that one class makes up only two percent of the training rows. Which part of AutoML raised this warning?
- Across forty runs of one experiment, a team needs those with an accuracy above 0.85, the best first. What should it use?