- Look at workspace and view model
- Look at code to register model (for awareness)
- Look at model deployment code to package model with assets and deploy to webservice
- Look at deployed model, validate by scoring data
- Using AML SDK to register and collect assets
- Environments are portable definitions to ensure consitency between training and deployment - more here https://docs.microsoft.com/en-us/azure/machine-learning/concept-environments
- InferenceConfig and DeploymentConfig
- Curated environments available out of the box: https://docs.microsoft.com/en-us/azure/machine-learning/resource-curated-environments
- Most of the code was from here: https://github.com/Azure/MachineLearningNotebooks/blob/master/how-to-use-azureml/deployment/deploy-to-cloud/model-register-and-deploy.ipynb - I will package and send this code as well
- Details about deploying to ACI, and many other services here: https://docs.microsoft.com/en-us/azure/machine-learning/how-to-deploy-azure-container-instance
- Attach to existing AKS instance and deploy: https://docs.microsoft.com/en-us/azure/machine-learning/how-to-deploy-azure-kubernetes-service#attach-an-existing-aks-cluster
- Or, deploy to App Service (preview): https://docs.microsoft.com/en-us/azure/machine-learning/how-to-deploy-app-service