Azure Machine Learning Compute Builder

ComputeAzure

Build Azure ML compute instance and AmlCompute cluster configs with GPU VMs, autoscaling, idle shutdown, and VNet integration.

Azure Machine Learning Configuration

Build Azure ML compute instance and AmlCompute cluster configs with GPU VMs, autoscaling, idle shutdown, VNet integration, and schedules.

Required Fields

workspaceNameresourceGroupcomputeTargetscomputeTargets[0].namecomputeTargets[0].computeTypecomputeTargets[0].properties.vmSize

{
  "workspaceName": "mlw-contoso-prod",
  "resourceGroup": "rg-ml",
  "location": "eastus2",
  "computeTargets": [
    {
      "name": "ci-ds-notebook",
      "computeType": "ComputeInstance",
      "properties": {
        "vmSize": "Standard_DS3_v2",
        "sshSettings": {
          "sshPublicAccess": "Disabled"
        },
        "subnet": {
          "id": "/subscriptions/a1b2c3d4/resourceGroups/rg-networking/providers/Microsoft.Network/virtualNetworks/vnet-ml/subnets/snet-compute"
        },
        "enableNodePublicIp": false,
        "idleTimeBeforeShutdown": "PT30M",
        "setupScripts": {
          "scripts": [
            {
              "scriptSource": "inline",
              "scriptData": "pip install torch transformers datasets",
              "timeout": "PT15M"
            }
          ]
        },
        "schedules": {
          "computeStartStop": [
            {
              "action": "Stop",
              "triggerType": "Cron",
              "cron": {
                "expression": "0 20 * * 1-5",
                "timezone": "Eastern Standard Time"
              }
            },
            {
              "action": "Start",
              "triggerType": "Cron",
              "cron": {
                "expression": "0 8 * * 1-5",
                "timezone": "Eastern Standard Time"
              }
            }
          ]
        }
      }
    },
    {
      "name": "cc-training-gpu",
      "computeType": "AmlCompute",
      "properties": {
        "vmSize": "Standard_NC24ads_A100_v4",
        "vmPriority": "LowPriority",
        "scaleSettings": {
          "minNodeCount": 0,
          "maxNodeCount": 8,
          "nodeIdleTimeBeforeScaleDown": "PT10M"
        },
        "subnet": {
          "id": "/subscriptions/a1b2c3d4/resourceGroups/rg-networking/providers/Microsoft.Network/virtualNetworks/vnet-ml/subnets/snet-training"
        },
        "enableNodePublicIp": false,
        "remoteLoginPortPublicAccess": "Disabled",
        "osType": "Linux",
        "isolatedNetwork": false,
        "identity": {
          "type": "UserAssigned",
          "userAssignedIdentity": "/subscriptions/a1b2c3d4/resourceGroups/rg-identity/providers/Microsoft.ManagedIdentity/userAssignedIdentities/mi-ml-training"
        }
      }
    }
  ],
  "tags": {
    "environment": "production",
    "team": "data-science",
    "costCenter": "ml-001"
  }
}

Generated Output

Output will appear here...

Overview

Build Azure ML compute instance and AmlCompute cluster configs with GPU VMs, autoscaling, idle shutdown, and VNet integration. This tool helps Azure engineers generate valid configurations quickly without consulting documentation, reducing errors and accelerating infrastructure deployment. All processing runs in your browser with no data sent to external servers.

How Engineers Use This

•Replacing a hand-edited Machine Learning Compute resource with a clean, reviewable version that lives in source control going forward.
•Producing a Machine Learning Compute snapshot for an internal audit where the auditor wants the canonical config alongside the rendered template.
•Translating an Azure architect's whiteboard sketch into a concrete Machine Learning Compute spec engineers can apply.
•Onboarding a new engineer by giving them a Machine Learning Compute sandbox to iterate on without touching live subscriptions.

Questions & Answers

Does this Machine Learning Compute builder consider the RBAC role needed to apply the configuration?

No — generating a Machine Learning Compute configuration is independent of the role required to apply it. Apply the output with a principal that has the documented permissions for that service. For least-privilege scoping, Azure's built-in roles and the Privileged Identity Management 'just enough access' workflows give you a starting point.

How does this Machine Learning Compute builder keep up with new Azure features?

The Machine Learning Compute options surface what is currently documented in the Azure reference for that service. When Microsoft adds a new property or value, we add it here after verifying the schema in a real subscription. If a recently-announced feature is not yet selectable, treat that as a 'not yet supported' signal rather than an opinion that it should not be used.

Related Learning Guides

Azure VM Series Guide22 min read

Was this tool helpful?

Disclaimer: This tool runs entirely in your browser. No data is sent to our servers. Always verify outputs before using them in production. AWS, Azure, and GCP are trademarks of their respective owners.