AI PROMPT ENGINEERING

Get acquainted with your awesome instructor, understand what this course has in store for you, and get hyped about diving into AI Prompt Engineering. Let's kick off this journey!

• Instructor introduction
• Course objectives & flow
• Final project overview
• Applications of AI in various industries
• Q&A

The course begins – in this first session, you'll explore the basics of AI prompt engineering, trace the evolution of neural networks, and get a sneak peek at the latest AI tools and platforms.

• Generative AI overview: History of NNs/transformers in 
• AI prompt engineering principles
• Overview of AI tools: Amazon Bedrock, SageMaker, ChatGPT, Bard/Gemini, Anthropic Claude, Meta Llama
• Overview of AI platforms: Hugging Face, Microsoft, Google, Amazon, Meta, Mistral

Assignment #1: Explore research papers "Attention is All You Need" and "Large Language Models: A Survey"

Didn’t you know that linguistics plays a huge role in AI interactions? By the end of this class, you'll be able to craft killer AI prompts using linguistic principles, play with AWS tools, and make your interactions smarter and smoother through hands-on experimentation with Amazon Bedrock.

• Linguistics in AI interactions: HCI 
• Crafting effective prompts for Amazon Bedrock AI models
• Exploring prompting roles in depth
• Enhancing AI interactions with linguistic strategies
• Hands-on prompt engineering with AWS tooling

Assignment #2: Experiment with prompt engineering techniques using Amazon Bedrock playground to craft diverse prompts

We'll explore how Large Language & Large Vision Models work, from text generation to visual interpretation, and see how they can power up your AI projects. Plus, we'll look at some real-life applications to spark your creativity.

• Large Language Models (LLMs): Deep dive
• Large Vision Models (LVMs)
• Multimodal LLMs
• Leveraging LLMs for AI development
• Case study: Applications of LLMs & LVMs

Get hands-on with Python for AI prompt engineering as we dive into essential techniques, explore key libraries and frameworks, and integrate Python with AI models using Amazon Bedrock and SageMaker.

• Python techniques for AI development with Amazon Bedrock/SageMaker JS
• Exploring Python libraries & frameworks in AI development
• Workshop: Integrating Python with AI Models for prompt engineering using Amazon Bedrock APIs & SageMaker

Assignment #3: Utilize Amazon Bedrock APIs for code generation and validate prompt techniques with models like Meta Llama, Mistral, and Anthropic Claude

In Class 5, learn to tackle domain-specific tasks hands-on with SageMaker JumpStart. Get ready to fine-tune models, pre-train LLMs, and have some fun with practical workshops on Gemma 2B.

• Customizing AI models for specific tasks 
• Techniques for fine-tuning AI models
• Pre-training of LLMs
• Workshop: Fine-tuning Gemma 2B on SageMaker JumpStart

Assignment #4: Fine-tuning Gemma 2B on Amazon SageMaker JumpStart

Dive into practical model optimization techniques like pruning, compression (QLoRa and LoRa), and quantization (GPTQ, AWQ, and GGML), with hands-on assignments on Amazon SageMaker. Get ready to fine-tune your models for peak performance.

• General model optimization techniques
• Knowledge distillation, hyperparameter tuning 
• Pruning, compression, and quantization

Assignment #5: Fine-tune an LLM on Amazon SageMaker (QLoRa OR LoRa) / Quantization model deployment on Amazon SageMaker Lab

In this class, you'll master Retrieval-Augmented Generation (RAG) techniques, including architecting guidelines and Bedrock agents. Get ready for some hands-on fun with Amazon Bedrock as we explore how to build Q&A systems and conversational interfaces with Bedrock and LangChain.

• Introduction to RAG architecting guidelines
• Bedrock with LangChain
• Bedrock agents/RAG 
• Workshop: Q&A system/conversational interfaces using RAG approach in Amazon Bedrock

Assignment #6: RAG approach using Amazon SageMaker JumpStart model Lab

Let’s level up. In Class 8, we'll dive into advanced RAG architectures, explore re-ranking and hybrid search techniques, and tackle a hands-on case study with Amazon Bedrock/SageMaker.

• Advanced RAG architectures
• Re-ranking & hybrid search techniques
• Case study: Advanced RAG approach with code using Amazon Bedrock/SageMaker

Assignment #7: Advanced RAG approach using Amazon SageMaker JumpStart model Lab

Why are evaluation metrics crucial for AI models? Let’s explore their various types, and learn how to choose the right ones for AI prompt engineering. Plus, a case study with Amazon Bedrock will show these metrics in action.

• Importance of evaluation metrics for AI model performance
• Evaluation metrics for different AI model types
• Techniques for selecting appropriate evaluation metrics 
• Using MMLU, HELM, Bert Score, Rouge Score, and FMEval Libs
• Case study: Evaluating AI models with Amazon Bedrock (human & automatic evaluation)

Assignment #8: Evaluation metrics

In this class, we're gonna explore keeping an eye on our models using CloudWatch and CloudTrail, dig into the responsible use of AI & ethics, and set up some guardrails to tackle bias and fairness issues head-on. 

• Monitoring models with CloudWatch & CloudTrail
• Responsible use of AI in various contexts
• Ethical guidelines for AI development
• Adding guardrails & safety nets: handling bias & fairness considerations 
• Workshop: Monitoring & guardrails with Bedrock Lab

Assignment #9: Guardrails with Prompt Engineering Lab

Explore how AI prompt engineering works in finance, healthcare, E-commerce, manufacturing, and beyond. Get inspired by real-life case studies and discover how AI can drive innovation, improve efficiency, reduce costs, and enhance agility across diverse industries. 

• AI prompt engineering: optimizing processes & driving innovation
• Case study: Successful AI-driven personalization strategies implementation
• Operations & supply chain optimization

Master deploying AI models in real-world settings. From TGI to DeepSpeed and Containerization, we'll dig into strategies that make those models hum. Plus, we've got a hands-on workshop lined up to navigate monitoring and managing those AI models in real-world environments.

• Deploying AI models in production: TGI, DeepSpeed, Containerization
• Workshop: Monitoring AI models in real-world environments
• Best practices for AI model lifecycle management 

In Class 13, explore the latest trends and future forecasts, chat about how AI advancements are shaking things up, and brainstorm how prompt engineering can tackle tomorrow's societal and technological challenges.

• AI prompt engineering trends
• Future of AI technology & impact on prompt engineering
• AI prompt engineering for future challenges
• Final project presentation

Let's get real about AI career paths. Discover the latest scoop on AI prompt engineering gigs, ace your networking game within the AI community, and learn how to craft a killer portfolio to flaunt your projects & skills.

• Career opportunities in AI prompt engineering landscape 
• Networking strategies in the AI community
• Portfolio development for AI prompt engineering 
• Insights into the AI industry, key players, trends & forecasts