This repo contains labs that show how to use the Azure Machine Learning service with the Python SDK. These labs will teach you how to perform the training locally within a Deep Learning Virtual Machine (DLVM) as well scale out model training by using Azure Batch AI workspaces or Azure Databricks. Each lab provides instructions for you to peform them using the environment of your choice- Visual Studio (with the Visual Studio Tools for AI), PyCharm or Azure Databricks.

The following labs are available:

• Lab 0: Setting up your environment. If a lab environment has not be provided for you, this lab provides the instructions to get started in your own Azure Subscription.
• Lab 1: Setup the Azure Machine Learning service from code and create a classical machine learning model that logs metrics collected during model training.
• Lab 2: Use the capabilities of the Azure Machine Learning service to collect model performance metrics and to capture model version, as well as query the experimentation run history to retrieve captured metrics.
• Lab 3: Deploying a trained model to containers using an Azure Container Instance and and Azure Kubernetes Service using Azure Machine Learning.
• Lab 4: Using the automated machine learning (Auto ML) capabilities within the Azure Machine Learning service to automatically train multiple models with varying algorithms and hyperparameters and then select the best performing model.
• Lab 5: Training deep learning models built with Keras and a Tensorflow backend that utilize GPUs with the Azure Machine Learning service.
• Lab 6: Deploy a trained model container to an IoT Edge device via the Azure Machine Learning service.