Large Language Models with Azure
Offered By: Pragmatic AI Labs via edX
Course Description
Overview
Master Large Language Model Operations on Azure
- Unlock Azure's full potential for deploying & optimizing Large Language Models (LLMs)
- Build robust LLM applications leveraging Azure Machine Learning & OpenAI Service
- Implement architectural patterns & GitHub Actions workflows for streamlined MLOps
Course Highlights:
- Explore Azure AI services and LLM capabilities
- Mitigate risks with foundational strategies
- Leverage Azure ML for model deployment & management
- Optimize GPU quotas for performance & cost-efficiency
- Craft advanced queries for enriched LLM interactions
- Implement Semantic Kernel for enhanced query results
- Dive into architectural patterns like RAG for scalable architectures
- Build end-to-end LLM apps using Azure services & GitHub Actions
Ideal for data professionals, AI enthusiasts & Azure users looking to harness cutting-edge language AI capabilities. Gain practical MLOps skills through tailored modules & hands-on projects.
Syllabus
Week 1: Introduction to LLMOps with Azure
\\- Discover pre-trained LLMs in Azure and deploy basic LLM endpoints
\\- Identify strategies for mitigating risks when using LLMs
\\- Explain how large language models work and their potential benefits and risks
\\- Describe the core Azure services and tools for working with AI solutions like Azure ML and the Azure OpenAI Service
****
Week 2: LLMs with Azure
- Use Azure Machine Learning, including GPU quota management, compute resource creation, model deployment, and utilization of the inference API
- Use the Azure OpenAI Service and its playground by deploying models and creating required resources
- Apply your comprehension of keys, endpoints, and Python examples to integrate Azure OpenAI APIs, monitor usage, and ensure proper resource cleanup
****
Week 3: Extending with Functions and Plugins
- Use Semantic Kernel to create advanced, context-aware prompts for large language models
- Define custom functions to extend system capabilities
- Build a microservice for reusable functions to streamline system extensions
- Implement functions using external APIs and microservices to customize model behavior
****
Week 4: Building an End-to-End LLM application in Azure
- Understand architectural patterns like RAG for building LLM applications
- Use Azure AI Search to create search indexes and embeddings to power RAG
- Build GitHub Actions workflows to automate testing and deployment of LLM apps
- Deploy an end-to-end LLM application leveraging RAG, Azure, and GitHub Actions
Taught by
Noah Gift and Alfredo Deza
Related Courses
Orchestrate Your AI with Semantic KernelDevoxx via YouTube Semantic Kernel - AI Orchestration for Intelligent Apps
Devoxx via YouTube Operationalizing LLMs on Azure
Duke University via Coursera Python Data Science Day: Comprehensive Workshop on Data Analysis and AI Technologies
Visual Studio Code via YouTube RAG Using Semantic Kernel with Azure OpenAI and Azure Cosmos DB - Python Data Science
Visual Studio Code via YouTube