Developing Apps with GPT–4 and ChatGPT

Preface
Conventions Used in This Book
Using Code Examples
O’Reilly Online Learning
How to Contact Us
Acknowledgments

1. GPT-4 and ChatGPT Essentials
Introducing Large Language Models
Exploring the Foundations of Language Models and NLP
Understanding the Transformer Architecture and Its Role in LLMs
Demystifying the Tokenization and Prediction Steps in GPT Models
A Brief History: From GPT-1 to GPT-4
GPT-1
GPT-2
GPT-3
From GPT-3 to InstructGPT
GPT-3.5, Codex, and ChatGPT
LLM Use Cases and Example Products
Be My Eyes
Morgan Stanley
Khan Academy
Duolingo
Yabble
Waymark
Inworld AI
Beware of AI Hallucinations: Limitations and Considerations
Optimizing GPT Models with Plug-ins and Fine-Tuning
Summary

2. A Deep Dive into the GPT-4 and ChatGPT APIs
Essential Concepts
Models Available in the OpenAI API
Trying GPT Models with the OpenAI Playground
Getting Started: The OpenAI Python Library
OpenAI Access and API Key
“Hello World” Example
Using ChatGPT and GPT-4
Input Options for the Chat Completion Endpoint
Output Result Format for the Chat Completion Endpoint
From Text Completions to Functions
Using Other Text Completion Models
Input Options for the Text Completion Endpoint
Output Result Format for the Text Completion Endpoint
Considerations
Pricing and Token Limitations
Security and Privacy: Caution!
Other OpenAI APIs and Functionalities
Embeddings
Moderation Model
Whisper and DALL-E
Summary (and Cheat Sheet)

3. Building Apps with GPT-4 and ChatGPT
App Development Overview
API Key Management
Security and Data Privacy
Software Architecture Design Principles
LLM-Powered App Vulnerabilities
Analyzing Inputs and Outputs
The Inevitability of Prompt Injection
Example Projects
Project 1: Building a News Generator Solution
Project 2: Summarizing YouTube Videos
Project 3: Creating an Expert for Zelda BOTW
Project 4: Voice Control
Summary

4. Advanced GPT-4 and ChatGPT Techniques
Prompt Engineering
Designing Effective Prompts
Thinking Step by Step
Implementing Few-Shot Learning
Improving Prompt Effectiveness
Fine-Tuning
Getting Started
Fine-Tuning with the OpenAI API
Fine-Tuning Applications
Generating and Fine-Tuning Synthetic Data for an Email Marketing Campaign
Cost of Fine-Tuning
Summary

5. Advancing LLM Capabilities with the LangChain Framework and Plug-ins
The LangChain Framework
Dynamic Prompts
Agents and Tools
Memory
Embeddings
GPT-4 Plug-ins
Overview
The API
The Plug-in Manifest
The OpenAPI Specification
Descriptions
Summary
Conclusion
Glossary of Key Terms

Index
About the Authors