Quick Start Guide to Large Language Models (LLMs): Unit 1

Overview

This course is designed to guide you through the evolution of natural language processing (NLP), from its historical roots to the cutting-edge advancements of today. You'll delve into the mechanics of modern deep learning architectures, exploring ground breaking concepts like attention and alignment mechanisms. Gain hands-on experience with leading LLMs such as ChatGPT, Llama, and T5, and discover how these models are revolutionizing AI-driven solutions. Through practical lessons, you'll learn about semantic search and build your own retrieval-augmented generation systems. Additionally, you'll gain experience in prompt engineering, enabling you to communicate effectively with LLMs. By the end of this course, you'll be equipped with the skills to effectively leverage the power of LLMs.

Syllabus

Introduction to Large Language Models

This course beings with the evolution of modern Natural Language Processing (NLP) and the significant advancements made in this technology over the past few decades. You'll delve into the mechanics of contemporary deep learning architectures, addressing questions such as how machines learn to read and write text and covering the key topics of attention and alignment mechanisms. Next, you'll explore large language models (LLMs) like ChatGPT, Llama, and T5, and learn about their underlying mechanisms. Building on this foundational knowledge, you'll transition to practical applications, such as performing semantic searches across extensive databases and building retrieval augmented generation (RAG) systems using both closed and open-source components. You will also learn how to craft effective prompts for LLMs. The course culminates in your constructing basic RAG systems and developing an AI agent capable of executing multiple tasks sequentially in a natural conversational manner. This module serves as your introduction to the expansive world of LLMs, catering to both beginners and enthusiasts eager to deepen their understanding.