Generative AI and LLMs on AWS (edX)

Course Highlights:

- Choose optimal LLM architectures for your applications

- Optimize cost, performance and scalability with auto-scaling and orchestration

- Monitor LLM metrics and continuously improve model quality

- Build secure CI/CD pipelines to train, deploy and update LLMs

- Ensure regulatory compliance via differential privacy and controlled rollouts

- Real-world, hands-on training for production-ready generative AI

- Unlock the power of large language models on AWS. Master operationalization using cloud-native services through this comprehensive, practical training program.

This course is part of the Large Language Model Operations (LLMOps) Professional Certificate.

What you'll learn

- Deploying large language models on AWS

- Selecting optimal LLM architectures and models

- Optimizing LLM cost, performance, and scalability

- Monitoring and logging LLM metrics

- Building reliable LLM CI/CD pipelines

- Ensuring regulatory compliance for LLM deployment

- Hands-on LLM operationalization using Amazon Bedrock

Syllabus

Week 1: Getting Started with Developing on AWS for AI

Introduction to AWS Cloud Computing for AI, including the AWS Cloud Adoption Framework

Setting up AI-focused development environments using AWS services like Cloud9, SageMaker, and Lightsail

Developing serverless solutions for data, ML, and AI using AWS Bedrock and Rust

Week 2: AI Pair Programming from CodeWhisperer to Prompt Engineering

Learning prompt engineering techniques to guide large language models

Using AWS CodeWhisperer as an AI pair programming assistant

Leveraging CodeWhisperer CLI to automate tasks and build efficient Bash scripts

Week 3: Amazon Bedrock

Key capabilities and components of Amazon Bedrock

Accessing and invoking Bedrock foundation models using AWS CLI, Boto3 Python SDK, and Rust SDK

Prompt engineering and model evaluation to optimize Bedrock model performance

Customizing models with fine-tuning and knowledge bases

Week 4: Project Challenges

Applying course concepts to build an end-to-end AI workflow

Developing Rust functions for Bedrock agents and integrating into an orchestration flow

Debugging, benchmarking, and prompt engineering to optimize a deployed AI application on AWS

By the end of this course, you will have gained hands-on experience with cutting-edge AI/ML tools on AWS like Bedrock, CodeWhisperer, and Rust. You'll be able to build and deploy efficient, serverless AI applications in production.