Get started with Azure Machine Learning

Overview

Azure Machine Learning is a cloud‑based service for building, training, and deploying machine‑learning models at scale. This guide walks you through the initial steps to start creating intelligent solutions.

Prerequisites

• An active Azure subscription
• Python 3.8+ installed locally
• Azure CLI (v2.30+) installed
• Basic knowledge of machine‑learning concepts

Install the required SDKs:

pip install azure-ai-ml[notebooks] --upgradeCopy

Quickstart: Build and deploy a model

1. Set up your workspace

az login
az ml workspace create -w my-ml-workspace -g my-resource-groupCopy

2. Prepare a training script

# train.py
from azure.ai.ml import MLClient
from azure.identity import DefaultAzureCredential

def main():
    # Dummy training code
    print("Training complete!")

if __name__ == "__main__":
    main()Copy

3. Submit the training job

az ml job create --file job.ymlCopy

Example job.yml:

name: iris-classifier
experiment_name: quickstart
environment: azureml:Python3.8-ubuntu20.04
code: .
command: >-
  python train.py
compute: azureml:cpu-cluster
distribution:
  type: tensorflowCopy

4. Deploy the model as a real‑time endpoint

az ml online-endpoint create -n iris-endpoint -f endpoint.yml
az ml online-deployment create -n default -e iris-endpoint -f deployment.ymlCopy

After deployment, you can invoke the endpoint:

curl -X POST -H "Content-Type: application/json" \
  -d '{"data": [[5.1, 3.5, 1.4, 0.2]]}' \
  https://.azurewebsites.net/scoreCopy

Next steps

• Explore step‑by‑step tutorials
• Learn about data preparation and feature engineering
• Integrate with MLOps pipelines
• Read the best practices guide