I. Overview:

Learn how to operate machine learning solutions at cloud scale using Azure Machine Learning. This course teaches you to leverage your existing knowledge of Python and machine learning to manage data ingestion and preparation, model training and deployment, and machine learning solution monitoring with Azure Machine Learning and MLflow.

II. Duration: 04 days (32 hours)

III. Intended Audience:

This course is designed for data scientists with existing knowledge of Python and machine learning frameworks like Scikit-Learn, PyTorch, and Tensorflow, who want to build and operate machine learning solutions in the cloud.

IV. Course outlines:

1. Explore and configure the Azure Machine Learning workspace

Throughout this learning path you explore and configure the Azure Machine Learning workspace. Learn how you can create a workspace and what you can do with it. Explore the various developer tools you can use to interact with the workspace. Configure the workspace for machine learning workloads by creating data assets and compute resources.

• Explore Azure Machine Learning workspace resources and assets: As a data scientist, you can use Azure Machine Learning to train and manage your machine learning models. Learn what Azure Machine Learning is, and get familiar with all its resources and assets.

• Explore developer tools for workspace interaction: Learn how you can interact with the Azure Machine Learning workspace. You can use the Azure Machine Learning studio, the Python SDK (v2), or the Azure CLI (v2).

• Make data available in Azure Machine Learning: Learn about how to connect to data from the Azure Machine Learning workspace. You're introduced to datastores and data assets.
• Work with compute targets in Azure Machine Learning: Learn how to work with compute targets in Azure Machine Learning. Compute targets allow you to run your machine learning workloads. Explore how and when you can use a compute instance or compute cluster.
• Work with environments in Azure Machine Learning: Learn how to use environments in Azure Machine Learning to run scripts on any compute target.

2. Experiment with Azure Machine Learning

Learn how to find the best model with automated machine learning (AutoML) and by experimenting in notebooks.

• Find the best classification model with Automated Machine Learning: Learn how to find the best classification model with automated machine learning (AutoML). You'll use the Python SDK (v2) to configure and run an AutoML job.
• Track model training in Jupyter notebooks with MLflow: Learn how to use MLflow for model tracking when experimenting in notebooks.

3. Optimize model training with Azure Machine Learning

Learn how to optimize model training in Azure Machine Learning by using scripts, jobs, components and pipelines.

• Run a training script as a command job in Azure Machine Learning: Learn how to convert your code to a script and run it as a command job in Azure Machine Learning.

• Track model training with MLflow in jobs: Learn how to track model training with MLflow in jobs when running scripts.

• Perform hyperparameter tuning with Azure Machine Learning: Learn how to perform hyperparameter tuning with a sweep job in Azure Machine Learning.

• Run pipelines in Azure Machine Learning: Learn how to create and use components to build pipeline in Azure Machine Learning. Run and schedule Azure Machine Learning pipelines to automate machine learning workflows.

4. Manage and review models in Azure Machine Learning

Learn how to manage and review models in Azure Machine Learning by using MLflow to store your model files and using responsible AI features to evaluate your models.

• Register an MLflow model in Azure Machine Learning: Learn how to log and register an MLflow model in Azure Machine Learning.

• Create and explore the Responsible AI dashboard for a model in Azure Machine Learning: Explore model explanations, error analysis, counterfactuals, and causal analysis by creating a Responsible AI dashboard. You'll create and run the pipeline in Azure Machine Learning using the Python SDK v2 to generate the dashboard.

5. Deploy and consume models with Azure Machine Learning

Learn how to deploy a model to an endpoint. When you deploy a model, you can get real-time or batch predictions by calling the endpoint.

• Deploy a model to a managed online endpoint: Learn how to deploy models to a managed online endpoint for real-time inferencing.

• Deploy a model to a batch endpoint: Learn how to deploy models to a batch endpoint. When you invoke a batch endpoint, you trigger a batch scoring job.

6. Develop generative AI apps in Azure

Generative Artificial Intelligence (AI) is becoming more accessible through comprehensive development platforms like Microsoft Foundry. Learn how to build generative AI applications that use language models to chat with your users.

• Plan and prepare to develop AI solutions on Azure: Microsoft Azure offers multiple services that enable developers to build amazing AI-powered solutions. Proper planning and preparation involves identifying the services you'll use and creating an optimal working environment for your development team.
• Choose and deploy models from the model catalog in Microsoft Foundry portal: Choose the various language models that are available through the Microsoft Foundry's model catalog. Understand how to select, deploy, and test a model, and to improve its performance.
• Develop an AI app with the Microsoft Foundry SDK: Use the Microsoft Foundry SDK to develop AI applications with Microsoft Foundry projects.
• Get started with prompt flow to develop language model apps in the Microsoft Foundry: Learn about how to use prompt flow to develop applications that leverage language models in the Microsoft Foundry.
• Develop a RAG-based solution with your own data using Microsoft Foundry: Retrieval Augmented Generation (RAG) is a common pattern used in generative AI solutions to ground prompts with your data. Microsoft Foundry provides support for adding data, creating indexes, and integrating them with generative AI models to help you build RAG-based solutions.
• Fine-tune a language model with Microsoft Foundry: Train a base language model on a chat-completion task. The model catalog in Microsoft Foundry offers many open-source models that can be fine-tuned for your specific model behavior needs.
• Implement a responsible generative AI solution in Microsoft Foundry: Generative AI enables amazing creative solutions, but must be implemented responsibly to minimize the risk of harmful content generation.
• Evaluate generative AI performance in Microsoft Foundry portal: Evaluating copilots is essential to ensure your generative AI applications meet user needs, provide accurate responses, and continuously improve over time. Discover how to assess and optimize the performance of your generative AI applications using the tools and features available in the Azure AI Studio.