Introduction

In the first two lessons of this module, you have learned how to interact with Foundation Models in a SwiftUI app, starting with the most basic response and progressing to a streamed response that provides the user with a better experience and faster feedback. You then learned the strengths and weaknesses of LLMs, along with the additional limitations of a local model, such as Apple Foundation Models, and ways to mitigate and manage these limitations. You also learned about the methods the framework provides for tweaking and adjusting the model’s output, including temperature, sampling methods, and instructions.

In this lesson, you will learn about two of the most useful and powerful features of Apple Foundation Models. First, you’ll explore guided generation, which provides a way to easily generate entire Swift data structures with Foundation Models. You will then explore Tools which provide a way to extend the model and assist with answering a prompt with information otherwise unavailable to the model.