Introduction

Welcome to this course on the science (and art) of fine-tuning LLMs! Over the subsequent lessons, we'll learn all about customizing LLMs on your own data. We'll teach LLMs to learn new tasks, exhibit new behaviors, generate new content, and much more.

In this course, I won't waste your time on complex theory or speculative BS. Instead, we'll dive into the exact tools and techniques that you'll need to build your own LLMs. We'll cover multiple solutions, from low-code, hosted services to open-source frameworks for fine-tuning LLMs using your own compute resources. By the end of this course, you'll not only understand in detail how fine-tuning works but will have the skill set to do so using a number of different approaches.

In this first lesson, we'll jump head-first into the theory and practice of fine-tuning LLMs, when to use fine-tuning and when not to, and how fine-tuning works under the hood. Let's get started!