Azure ML: Pipelines and Experiment Tracking
Build Azure ML pipelines, track experiments with MLflow and register and version the best model.
Without a pipeline, an ML workflow looks like this: you manually execute each step, in the right order, on the right machine. If a step fails at 3 AM, nobody knows until the next morning. If you want to re-run the same workflow next month, you have to start from scratch.
Sign in to read this course
A free account unlocks all 514 courses. 20 are readable without one.
What's inside
15 sections- 1 Table of Contents
- 2 Azure ML Pipelines – Core Concepts
- 3 Pipeline Components
- 4 Creating a Pipeline with the Designer
- 5 Creating a Pipeline with the Python SDK v2
- 6 Submitting and Monitoring Pipelines
- 7 Handling Failures and Retries
- 8 Experiment Tracking with MLflow
- 9 Comparing Runs and Selecting the Best Model
- 10 Model Registration and Versioning
- 11 Advanced Patterns – Hyperparameter Tuning
- 12 MLOps – CI/CD Integration
- 13 Summary and Key Takeaways
- 14 Glossary
- 15 Appendix: Quick Reference
More ML Platforms & Deployment courses
View all 10Introduction to Azure Machine Learning
Where Azure ML fits in the modern ML lifecycle, its common use cases and user workflows.
Azure ML Workspace Fundamentals
Azure ML workspace architecture, compute, datasets, environments, governance, security and cost.
Azure ML Studio and SDK – Overview
Navigate Azure ML Studio, notebooks, the Python SDK v2 and CLI v2 with a first end-to-end job.
Azure ML: Practical Use Cases
Choose the right technique and run classification, clustering and batch inference with AutoML and the Designer.
Deploying Models with Azure Machine Learning
Online and batch endpoints, scoring scripts, blue/green deployment, AKS and model monitoring on Azure ML.
Deploying Machine Learning Solutions
Deploy models with Flask, on serverless, on Google AI Platform and to AWS SageMaker.
Interested in this course?
Contact us to book it or get a custom training plan for your team.