Price: $59.99 - $35.46
(as of Apr 21,2025 19:48:20 UTC – Details)
Buy Now
From the Publisher
“Covers all the essential aspects of how to build and deploy LLMs. It goes into the deep and fascinating areas that most other books gloss over.”
Andrew Carr, Cartwheel
“A must-read for anyone looking to harness the potential of LLMs in production environments.”
Jepson Taylor, VEOX Inc.
“An exceptional guide that simplifies the building and deployment of complex LLMs.”
Arunkumar Gopalan, Microsoft UK
about the book
LLMs in Production teaches you how to develop an LLMOps plan that can take an AI app smoothly from design to delivery. You’ll learn techniques for preparing an LLM dataset, cost-efficient training hacks like LORA and RLHF, and industry benchmarks for model evaluation. Along the way, you’ll put your new skills to use in three exciting example projects: creating and training a custom LLM, building a VSCode AI coding extension, and deploying a small model to a Raspberry Pi.
about the authors
Christopher Brousseau and Matt Sharp are experienced engineers who have led numerous successful large scale LLM deployments.
Chapter 1: Words’ Awakening: Why Large Language Models Have Captured Attention Takes you through the historical significance of how humans share information—drawing an analogy to Gutenberg’s printing press—and explains why these models that can generate and interpret text so well have captured everyone’s imagination. It also weighs the pros and cons of deploying your own LLMs versus using existing services.
Chapter 2: Large Language Models: A Deep Dive into Language Modeling Provides the foundation for how computers handle language—from counting words (like in simple bag-of-words) all the way to sophisticated transformer-based methods. By the end of it, you’ll see how modern LLMs stitch together meaning from massive amounts of data and why that leap was such a big deal in NLP.
Chapter 3: Large Language Model Operations: Building a Platform for LLMs Pulls back the curtain on the practicalities of running LLMs at scale—covering infrastructure, hardware needs, and the nuts and bolts of large-model “DevOps.” It highlights challenges like high latency and GPU management, and shows how modern data pipelines evolve to handle these demands.
Chapter 4: Data Engineering for Large Language Models: Setting Up for Success Explores how high-quality data is just as critical as fancy model architectures. You’ll learn about available datasets, how to clean and prep them, and which industry benchmarks help you figure out if a model is doing well. It also covers specialized tools like text processors.
Chapter 5: Training Large Language Models: How to Generate the Generator Dives into the details of training, from multi-GPU environments to transfer learning. Different approaches—like finetuning with knowledge distillation and using LoRA or parameter-efficient finetuning—come to life. This chapter reveals how an LLM improves its capabilities through data.
Chapter 6: Large Language Model Services: A Practical Guide Shows how to wrap LLMs in a service so you can interact with them over APIs. You’ll learn how to store and load huge models efficiently, handle incoming requests without overloading resources, and augment the model’s responses by retrieving external context. It’s about building a stable environment so the LLM can do real work.
Chapter 7: Prompt Engineering: Becoming an LLM Whisperer Covers how to formulate requests in a way that gets better, more precise answers from an LLM. Basic prompt techniques come first, then more advanced ideas like using ReAct-style prompts and letting the model call external tools. You’ll see prompt design as an interplay between language, context, and software.
Chapter 8: Large Language Model Applications: Building an Interactive Experience Walks you through designing end-user applications—like chatbots that remember conversation state, or knowledge-extraction interfaces that deal with large amounts of text. It introduces token-tracking strategies, streaming responses, and RAG – methods for weaving user queries together with relevant data.
Chapter 9: Creating an LLM Project: Reimplementing Llama 3 Offers a hands-on experience: you build a simplified version of a state-of-the-art model from scratch. You’ll set up its architecture, tokenization, and basic evaluation, then get tips on improving it with techniques like quantization. It’s a peek under the hood of what an industrial research lab might do.
Chapter 10: Creating a Coding Copilot Project: This Would Have Helped You Earlier Focuses on turning an LLM into a programming assistant that can generate code snippets or explain logic. You’ll combine retrieval-augmented generation with embeddings, stitch it all together in an IDE plugin, and see how an AI model can become a companion for software development.
Chapter 11: Deploying an LLM on a Raspberry Pi: How Low Can You Go? Explores edge computing: shrinking a big model down so it runs on minuscule hardware. It walks through installing necessary libraries on a Raspberry Pi, quantizing the model to reduce its size, and optimizing performance so that the Pi can handle local inference at small scale.
Chapter 12: Production, an Ever-Changing Landscape: Things Are Just Getting Started Wraps up by surveying where the field might go next—larger models, clever new architectures, or advanced regulation. It posits that though LLMs have come a long way, future breakthroughs—like fully solving hallucinations or developing “agents” that handle tasks autonomously—are likely just around the corner.
Publisher : Manning (February 11, 2025)
Language : English
Paperback : 456 pages
ISBN-10 : 1633437205
ISBN-13 : 978-1633437203
Item Weight : 1.65 pounds
Dimensions : 7.38 x 1 x 9.25 inches
Buy Now