Small Language Models in Action: Build Fast, Efficient and Private AI Systems - Couverture souple

Acklin, James

Couverture souple

ISBN 13 : 9798278978435

Editeur : Independently published, 2025

Synopsis

Small language models are changing how AI is built, deployed, and trusted. Instead of relying on massive, expensive models that demand cloud access and complex infrastructure, developers can now build fast, efficient, and private AI systems that run locally, scale predictably, and integrate cleanly into real software products.

Small Language Models in Action is a practical, engineering-focused guide for developers, architects, and technical decision-makers who want to move beyond hype and build AI systems that actually work in production. This book explains how small language models are designed, optimized, deployed, and maintained, with a strong emphasis on performance, cost control, privacy, and reliability.

You will learn how modern transformer-based models can be compressed, fine-tuned, and executed efficiently on CPUs, GPUs, edge devices, and private servers. The book walks through real architectural choices, deployment patterns, and optimization strategies, showing when small models are the right tool—and how to get the most out of them.

Rather than focusing on theory alone, this book connects model behavior to real constraints: memory limits, latency targets, infrastructure budgets, and long-term maintenance. You will see how small language models are used in on-device assistants, enterprise systems, privacy-sensitive environments, embedded platforms, and hybrid architectures that combine small and large models intelligently.

By the end of this book, you will understand how to evaluate models realistically, fine-tune them responsibly, deploy them confidently, and maintain AI-powered applications over time without sacrificing control or predictability.

If you want to build AI systems that are fast, affordable, secure, and production-ready, this book gives you the knowledge and confidence to do it right.

Détails de l'édition

• Éditeur: Independently published
• Date d'édition: 2025
• Langue: anglais
• ISBN 13: 9798278978435
• Reliure: Broché
• Nombre de pages: 263
• Coordonnées du fabricant: Manufactured by Amazon on behalf of the author; https://www.amazon.fr/hz/contact-us; ; c/o Amazon Media EU S.à.r.l., 38 Avenue John F. Kennedy; Luxembourg; L-1855; Luxembourg