Azure Machine Learning documentation

Learn how to train and deploy models and manage the ML lifecycle (MLOps) with Azure Machine Learning. Tutorials, code examples, API references, and more.

Overview

Overview

• What is Azure Machine Learning?
• What is Azure Machine Learning studio?
• Key concepts (v2)
• What is CLI & SDK v2?

Get Started

Tutorial

• Set up resources
• Train with the v2 CLI
• Train with the v2 SDK (preview)
• Day 1: Get started with Python
• Migrate from ML Studio (classic)

Train models

How-To Guide

• Run training code in the cloud (v2 CLI)
• Train and deploy a model in Jupyter notebook
• Tune hyperparameters for model training
• Build pipelines from reuseable components
• AutoML: Train a classification model

Deploy models

Deploy

• Streamline model deployment with endpoints (preview)
• Real-time scoring with a managed endpoint (preview)
• Batch scoring with managed endpoints (preview)
• Options for deploying models without using endpoints

Manage the ML lifecycle (MLOps)

How-To Guide

• Track, monitor, analyze training runs
• Model management, deployment, lineage & monitoring
• DevOps for a data ingestion pipeline

Security for ML projects

How-To Guide

• Create a secure workspace
• Connect to data sources
• Enterprise security & governance

Reference docs

Reference

• Python SDK
• REST API
• V2 CLI (preview)
• Algorithm & component reference

Resources

Reference

• Python SDK code examples
• v2 CLI code examples (preview)
• ML Studio (classic) documentation