Use Azure Machine Learning in a notebook
Azure Notebooks comes pre-configured with the necessary environment to work with Azure Machine Learning. You can easily clone a sample project into your Notebooks account to explore a variety of Machine Learning scenarios.
Clone the sample into your account
Sign into Azure Notebooks.
Select My Projects to go to the projects dashboard.
Select the Upload GitHub Repo (the up arrow) button to open the Upload GitHub Repository popup.
In the popup, enter
Azure/MachineLearningNotebooksin GitHub repository, provide a name for the project in Project Name like "Azure Machine Learning", provide an identifier in Project ID, clear Public if you want, then select Import.
After a minute or two, Azure Notebooks automatically takes you to the new project's dashboard.
Run a sample notebook
Select 00 - configuration.ipynb to start the configuration section of the notebook, and follow its instructions to create an Azure Machine Learning Workspace.
- Because Azure Notebooks already contains the necessary Python packages, you can just run the code snippet in step 2 of the Prerequisites to verify the Azure ML SDK version.
Once configuration is complete, select 01.getting-started to open the folder containing thirteen different sample notebooks, each of which is self-explanatory.
The Azure Machine Learning documentation contains a variety of other resources that guide you through working with Machine Learning within notebooks:
- Quickstart: Use Python to get started with Azure Machine Learning
- Tutorial #1: Train an image classification model with Azure Machine Learning
- Tutorial #2: Deploy an image classification model in Azure Container Instance (ACI)
- Tutorial: Train a classification model with automated machine learning in Azure Machine Learning
Also see the documentation for the Azure Machine Learning SDK for Python.