Skip to main content
Version: 1.42

Deeploy CLI

The Deeploy Python Client includes a CLI. This CLI is currently intended to help Deeploy users to create their own Custom Docker images that can be used for deploying a model, explainer and/or transformer on Deeploy. The focus of these templates is currently on the KServe deployment service. Support for Azure Machine Learning and Sagemaker deployments is intended in the near future.

Using the CLI​

You can test if the Deeploy package is configured correctly (and check your current Python Client) by running:

deeploy --version

This should return (where x.x.x corresponds to your version):

deeploy, version x.x.x

You can use the Deeploy CLI with the deeploy command. To view to list of commands that is available in your current Deeploy Python Client version run:

deeploy --help

Currently the main functionality of the CLI is around generating templates for custom Docker deployments:

deeploy generate-template --help

This will generate the new project in the current working directory, which you can initialize as a git repository as reference for Deeploy deployments. The project will contain a README.md file with instructions on how to get started.

Once you have generated a template project, follow the instructions in the README.md file to get started.

Troubleshooting​

The most likely reasons you experience problems with the CLI are:

  • You have a Deeploy Python Client Version < 1.2.1 installed
  • Your PATH to you Python environment is not configured correctly