Contents

Azure Machine Learning

Create a new Azure Machine Learning

I recommand you create a new AML workspace use an Azure Resource Manager Template from Azure CLI. The Azure Resource Manager template can be found in the 201-machine-learning-advanced in GitHub repository.

This template creates the following Azure services, all those various services are required by the Azure Machine Learning workspace:

  • Azure Storage Account
  • Azure Key Vault
  • Azure Application Insights
  • Azure Container Registry
  • Azure Machine Learning workspace

Firstly, we need create a resource group, there is a example at the below. The resource group is the container that holds the services.

  • Create a new resource group, name: aml and location: eastus
1
az group create -l eastus -n aml

NOTE: List all of Azure location command at the below.

1
az account list-locations

I suggest you deploy all dependent resources behind a virtual network.

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
az deployment group create \
    --name "boyyan-ml" \
    --resource-group "aml" \
    --template-uri "https://raw.githubusercontent.com/Azure/azure-quickstart-templates/master/201-machine-learning-advanced/azuredeploy.json" \
    --parameters workspaceName="boyyanworkspace" \
      location="eastus" \
      vnetOption="new" \
      vnetName="mlvnet" \
      storageAccountBehindVNet="true" \
      keyVaultBehindVNet="true" \
      containerRegistryBehindVNet="true" \
      containerRegistryOption="new" \
      containerRegistrySku="Premium" \
      privateEndpointType="AutoApproval"

NOTE: Total transitive private endpoint usage 0 is equal or greater than quota 0. Please increase quotaby following the doc here( https://docs.microsoft.com/azure/machine-learning/how-to-manage-quotas#private-endpoint-and-private-dns-quota-increases) .