Textbook in PDF format

This book offers a unique, hands-on, and practical approach while balancing theory and concepts. It introduces the latest trends in natural language processing (NLP), primarily large language models (LLMs), providing insights into how these networks work. Additionally, it includes projects that demonstrate the application of these models in creating retrieval-augmented generation (RAG) pipelines. These concepts represent cutting-edge developments in the field, allowing us to process written text and interact with it on a contextual level.
Much like most books related to LLMs, we begin with the foundation by exploring the details of transformer architecture to understand how these models are trained and how to interact with them using prompting techniques. We then dive into the industry-focused sections, first covering two well-known frameworks that can be used to leverage these models to create RAG-enabled applications (LlamaIndex and LangChain). This includes a variety of projects that provide hands-on experience, helping to deeply understand and apply these concepts. We also explore advanced techniques, such as using autonomous agents or incorporating vision capabilities to enhance question-answering. Finally, we explore deployment options for hosting the application and tips to make the process more efficient.
Introduction to Large Language Models
LLM Architectures & Landscape
LLMs in Practice
Introduction to Prompting
Introduction to LangChain & LlamaIndex
Prompting with LangChain
Retrieval-Augmented Generation
Advanced RAG
Agents
Fine-Tuning
Deployment