Introducing Deploying ML models with AzureML

Introducing Deploying ML models with AzureML

Objectives

In the following Excercise you will learn:

  • Deploy an AzueML model to Azure Container Instance

Prerequisites

To run through below instructions, you need an Azure subscription, an AzureML workspace and an AzureML compute target (i.e. cpu-cluster). See instructions on how to create a workspace here and create an AzureML compute target here. In addition you require the registered model you have trained here here

Exercise: Deploy your model on Azure Container Instance

In the following exercise you will:

  • Register the model or load the registered model
  • Prepare to deploy. (Specify assets, usage, compute target.)
  • Deploy the model to the compute target.
  • Test the deployed model, also called a web service.
  1. Run the notebook to deploy to ACI Python SDK - deploy-credit-model.ipynb
  2. Check for your Endpoint in the Azure ML Workspace
  3. Test your new Web Service