Curriculum

Generative AI with IBM Cloud

Unit 1: Introduction to Generative AI

  • What is Generative AI?

    • Definition of Generative AI

      • Defining Generative AI vs. Discriminative AI

      • Key characteristics and properties

    • How Generative AI Works

      • Underlying principles: Latent spaces, Probability distributions, Neural Networks

    • Capabilities of Generative AI

      • Text generation, Image synthesis, Audio generation, Video generation, Code generation

    • Use Cases of Generative AI

      • Content creation, Marketing, Drug discovery, Financial modeling, Entertainment, more.

    • Applications in Different Sectors

      • Healthcare, Finance, Entertainment, Manufacturing, Education, and others.

  • Prompt Engineering Fundamentals

    • Introduction to Prompt Engineering

      • Defining Prompt Engineering and its importance for LLMs.

      • The art and science of guiding generative models.

    • Components of a Prompt

      • Instructions, Context, Input Data, Output Indicators, Examples

    • Best Practices for Prompt Creation

      • Clarity and Specificity, Context and Background, Language and Tone.

    • Common Prompt Engineering Techniques

      • Zero-Shot, One-Shot, Few-Shot Learning

      • Chain-of-Thought Prompting, Role Playing, and more.

    • Examples of Effective and Ineffective Prompts

      • Analyzing examples and identifying key elements for success.

    • Iterative Prompt Refinement

      • Refining prompts through testing, analysis, and feedback.

Unit 2: Generative AI Core Models

  • GANs (Generative Adversarial Networks)

    • Architecture, Training, Applications, Limitations

  • VAEs (Variational Autoencoders)

    • Architecture, Training, Applications, Limitations

  • Transformers

    • High-level Overview and Applications

Unit 3: Fundamental of Langchain and RAGs

  • Fundamentals of  Langchain

    • Introduction to Langchain

      • What is Langchain and its purpose?

    • Working with Prompts in Langchain

      • PromptTemplates, FewShotPromptTemplates, and more.

    • Chains in Langchain

      • LLMChain, SimpleSequentialChain, etc.

    • Indexes in Langchain

      • Document Loaders, Text Splitters, Vectorstores (introduction).

    • Examples of LLM Applications with Langchain.

  • Retrieval-Augmented Generation (RAG)

    • What is Retrieval-Augmented Generation?

      • Definition of RAG

      • Benefits and limitations of RAG

    • RAG System Components

      • Knowledge Base: Data sources, indexing techniques

      • Retrieval Module: Embedding models, vector databases (e.g., ChromaDB, Pinecone), similarity search

      • Generation Module: LLM integration, prompt optimization

    • Building a RAG Pipeline

      • Step-by-step implementation of a RAG system

      • Loading and processing data

      • Creating embeddings and indexing the knowledge base

      • Retrieving relevant information

      • Generating responses using an LLM

    • Evaluating RAG Performance

      • Metrics for evaluating RAG systems

      • Techniques for improving performance

Unit 4: Cloud Foundations, Watsonx Platform, Watson Orchestrate, Watson Studio

  • Introduction to IBM Cloud and Cloud Computing

    • What is Cloud Computing?

      • Definition, Key Characteristics

      • Benefits for Enterprises and Developers

    • Cloud Service Models

      • IaaS, PaaS, SaaS – Concepts and Examples

    • Introduction to IBM Cloud

      • Overview of IBM Cloud ecosystem

      • Key Services and Industry Use Cases

    • Getting Started with IBM Cloud

      • Creating an IBM Cloud Account (Free Tier, Pricing Model)

      • Navigating the IBM Cloud Console and CLI

  • Exploring the Watsonx Platform (3 Hours)

    • Overview of Watsonx Platform

      • What is Watsonx?