1 What Is Generative AI and Why PyTorch?

This chapter covers

Understanding generative AI vs. non-generative AI
Why PyTorch is ideal for deep learning and generative AI
Concept of Generative Adversarial Networks (GANs)
Benefits of attention mechanism and Transformers
Learning advantages of creating generative AI models from scratch

Generative AI has significantly impacted the global landscape, capturing widespread attention and becoming a focal point since the advent of ChatGPT in November 2022. This technological advancement has revolutionized numerous aspects of everyday life, ushering in a new era in technology and inspiring a host of startups to explore the extensive possibilities offered by various generative models.

Consider the advancements made by Midjourney, a pioneering company, which now creates high-resolution, realistic images from brief text inputs. Similarly, Freshworks, a software company, has accelerated application development dramatically, reducing the time required from an average of ten weeks to mere days, a feat achieved through the capabilities of ChatGPT. [1] To add a case in point, elements of this very introduction have been enhanced by generative AI, demonstrating its capability to refine content to be more engaging. [2]

1.1 Introducing Generative AI and PyTorch

1.1.1 What Is Generative AI?

1.1.2 The Python Programming Language

1.1.3 Using PyTorch as Our AI Framework

1.2 Generative Adversarial Networks (GANs)

1.2.1 A High-Level Overview of GANs

1.2.2 An Illustrating Example: Generating Anime Faces

1.2.3 Why Should You Care About GANs?

1.3 Transformers

1.3.1 The Attention Mechanism

1.3.2 The Transformer Architecture

1.3.3 Multimodal Transformers and Pre-Trained LLMs

1.4 Why Build Generative Models from Scratch?

1.5 Summary