Skip to content

Files

Latest commit

saachigopalsaachigopal
Update how-to-manage-environments-in-studio.md
May 16, 2022
3e00901 · May 16, 2022

History

History
75 lines (50 loc) · 3.78 KB

how-to-manage-environments-in-studio.md

File metadata and controls

75 lines (50 loc) · 3.78 KB
title titleSuffix description services ms.service ms.subservice author ms.author ms.date ms.topic ms.custom
Manage environments in the studio
Azure Machine Learning
Learn how to create and manage environments in the Azure Machine Learning studio.
machine-learning
machine-learning
core
saachigopal
sagopal
10/21/2021
how-to
devx-track-python

Manage software environments in Azure Machine Learning studio

In this article, learn how to create and manage Azure Machine Learning environments in the Azure Machine Learning studio. Use the environments to track and reproduce your projects' software dependencies as they evolve.

The examples in this article show how to:

  • Browse curated environments.
  • Create an environment and specify package dependencies.
  • Edit an existing environment specification and its properties.
  • Rebuild an environment and view image build logs.

For a high-level overview of how environments work in Azure Machine Learning, see What are ML environments? For information, see How to set up a development environment for Azure Machine Learning.

Prerequisites

Browse curated environments

Curated environments contain collections of Python packages and are available in your workspace by default. These environments are backed by cached Docker images which reduces the run preparation cost and support training and inferencing scenarios.

Click on an environment to see detailed information about its contents. For more information, see Azure Machine Learning curated environments.

Create an environment

To create an environment:

  1. Open your workspace in Azure Machine Learning studio.
  2. On the left side, select Environments.
  3. Select the Custom environments tab.
  4. Select the Create button.

Create an environment by specifying one of the following:

  • Create a new docker context
  • Start from an existing custom or curated environment
  • Upload existing docker context
  • Use existing docker image with conda

:::image type="content" source="media/how-to-manage-environments-in-studio/create-page.jpg" alt-text="Environment creation wizard":::

You can customize the configuration file, add tags and descriptions, and review the properties before creating the entity.

If a new environment is given the same name as an existing environment in the workspace, a new version of the existing one will be created.

View and edit environment details

Once an environment has been created, view its details by clicking on the name. Use the dropdown menu to select different versions of the environment. Here you can view metadata and the contents of the environment through its various dependencies.

Click on the pencil icons to edit tags and descriptions as well as the configuration files under the Context tab.

Keep in mind that any changes to the Docker or Conda sections will create a new version of the environment.

:::image type="content" source="media/how-to-manage-environments-in-studio/details-page.jpg" alt-text="Environments details page":::

View image build logs

Click on the Build log tab within the details page to view the image build logs of an environment version.

Rebuild an environment

In the details page, click on the rebuild button to rebuild the environment. Any unpinned package versions in your configuration files may be updated to the most recent version with this action.