LLMs on Microsoft Azure
Published by O'Reilly Media, Inc.
Working with foundation models in the cloud
Course outcomes
- Understand how to develop LLM models on Azure cloud
- Learn to evaluate, assess, and validate models with private data
- Understand how to deploy LLMs with APIs for public consumption
- Learn the intricacies of LLMs and the deployment architectures on Azure
- Explore and experiment with developing AI-driven applications using pre-trained LLMs
- Understand Azure cloud tools and utilities to support LLM integration and development
Course description
Developing, training, and deploying large language models typically require massive amounts of computing resources and highly skilled machine learning professionals, but now cloud platforms are making it possible for a wider range of users to accomplish these tasks, with just a few clicks of a button. This is where Azure cloud comes in. The Azure AI Studio suite streamlines the entire process of working with LLMs, from inception to deployment.
Join expert Madhu Konda to explore how you can access pre-trained models, customize them to your specific needs, and deploy them seamlessly within your infrastructure, while exposing APIs for further consumption. This course works with Meta’s open source model, as well as Llama 2, and OpenAI’s GPT-3.5 and GPT-4 models.
What you’ll learn and how you can apply it
- Develop LLM models on Azure cloud
- Evaluate, assess, and validate models with private data
- Deploy LLMs with APIs for public consumption
This live event is for you because...
- You’re a cloud architect who wants to architect the LLM deployments for your organization.
- You’re a software engineer who’s looking to integrate applications with fine-tuned cloud-based LLMs.
- You’re an engineering manager who wants to provide guidance to your teams on how to integrate LLMs into their applications.
- You’re a DevOps engineer who provides guidance and support to application development teams.
Prerequisites
- An Azure account
- Some familiarity with Azure cloud
- Basic understanding of ChatGPT, generative AI, and LLMs
- Python programming experience (preferred but not essential)
Recommended follow-up:
- Read Hands-On Large Language Models (book)
- Read Mastering Azure Machine Learning (book)
- Read Generative AI on AWS (book)
Schedule
The time frames are only estimates and may vary according to how the class is progressing.
Introducing GenAI (25 minutes)
- Presentation: Setting the scene of LLMs and generative AI; dissecting large language models
- Group discussion: Understanding GenAI, ChatGPT, LLMs, and machine learning
- Q&A
Introduction to Azure Machine Learning (35 minutes)
- Presentation: Azure’s support for LLMs; understanding ML studio components
- Demonstration: Azure Machine Learning studio; setting up workspace
- Break
Foundation models (30 minutes)
- Presentation: Introduction to model catalog; learning foundational models on Azure
- Demonstration: Creating compute capacity for LLM development
- Q&A
Introduction to prompt flow (30 minutes)
- Presentation: Overview of prompt flow
- Demonstration: Creating visual prompt flows
- Q&A
- Break
Working with LLMs on Azure (60 minutes)
- Presentation: Developing Llama 2 model; understanding evaluation, testing, and validation of the LLM models
- Hands-on exercise: Create a chat application using OpenAI and prompt flow
- Q&A
Your Instructor
Madhusudhan Konda
Madhusudhan Konda is a passionate technologist and a lifelong tech learner who loves distilling complex problems into simpler solutions, looking at the big picture and providing technical direction, and experimenting with newer programming languages and shiny frameworks. Over his career, Madhusudhan has held roles such as solution architect, development lead, lead developer, and others, always with a strong inclination toward teaching his fellow techies programming languages, frameworks, and new technologies. He’s been instrumental in delivering high-quality solutions to major clients such as EBRD, Credit Suisse, UBS, Mizuho, Deutsche Bank, Halifax, British Petroleum, British Airways, and Lloyd’s of London, to name a few. His core competencies lie in not only creating simple architectures for complex business problems and designing and developing the software projects from the ground up but also providing strategic road maps, cost-effective architectures, and product designs, leading teams, mentoring, and providing thought leadership. He’s written well-received books and produced video courses on Java, Spring, and the Hibernate ecosystem.