"Large Language Models in Production" is a practical and applied guide for specialists who want to learn how to transition solutions based on LLMs from the prototype stage to reliable production systems. The updated content, emphasis on real cases, and detailed analysis of technologies make the book a valuable source of knowledge for engineers and teams working with generative AI.
About This Book
The publication helps understand key concepts of large language models and teaches how to apply them in real products. The authors explain how LLMs work, how they are trained, and what features need to be considered when integrating models into existing technological landscapes.
Main Topics
- architecture and training of LLMs;
- comparison of ready-made models and custom developments;
- scaling ML platforms for generative AI tasks;
- fine-tuning and retraining models (PEFT, LoRA, RLHF);
- deployment in the cloud and on edge devices;
- MLOps and LLMOps for corporate solutions.
Practical Projects
The book does not limit itself to theory — the reader gets the opportunity to apply knowledge in practice.
You will become familiar with three real projects
- Creating and training your own LLM with an analysis of data preparation stages and architecture selection.
- Developing an AI extension for VSCode using modern generative AI tools.
- Launching a compact model on a Raspberry Pi and optimizing for limited computing resources.
Modern Trends and Technologies
The authors cover current trends in the development of generative AI, which are already being applied in the industry.
The book examines
- RAG systems and working with external data sources;
- knowledge graphs and corporate knowledge management;
- increasing context windows and new approaches to processing long queries;
- instrumental agents and LLM interaction with external services;
- regulation of AI and its impact on product development.
Who the Book is For
The publication will be especially useful for data scientists, ML engineers, and technical specialists familiar with Python and the basic principles of cloud deployment. The book helps transition from experimentation to creating full-fledged, sustainable, and scalable production solutions based on LLMs.
