Transform your machine learning projects into successful deployments with this practical guide on how to build and scale solutions that solve real-world problems
Includes a new chapter on generative AI and large language models (LLMs) and building a pipeline that leverages LLMs using LangChain
Key Features
- This second edition delves deeper into key machine learning topics, CI/CD, and system design
- Explore core MLOps practices, such as model management and performance monitoring
- Build end-to-end examples of deployable ML microservices and pipelines using AWS and open-source tools
Book Description
Machine Learning Engineering with Python, 2nd Edition, is the practical guide that MLOps and ML engineers need to build robust solutions to solve real-world problems, providing you with the skills and knowledge you need to stay ahead in this rapidly evolving field.
The book takes a hands-on, examples-focused approach providing essential technical concepts, implementation patterns, and development methodologies. You’ll go from understanding the key steps of the machine learning development lifecycle to building and deploying robust machine learning solutions. Once you’ve mastered the basics, you’ll get hands-on with deployment architectures and discover methods for scaling up your solutions.
This edition goes deeper into ML engineering and MLOps, with a sharper focus on ML. You’ll take CI/CD further with continuous training and testing and go in-depth into data and concept drift.
With a new generative AI chapter, explore Hugging Face, PyTorch, and GitHub Copilot, and consume an LLM via an API using LangChain. You’ll also cover deep learning considerations regarding workflow, hardware, and scaling up workloads, as well as orchestrating workflows with Airlfow and Kafka. And take advantage of ZenML as an open-source option for pipelining dataflows, and take deployment further with canary, blue, and green deployments.
What you will learn
- Plan and manage stages of machine learning development projects
- Explore ANNs, DNNs, and LLMs, and get to grips with the rise of generative AI in MLOps
- Use Python to package your own ML tools and scale up solutions with Apache Spark, Kubernetes, and Apache Airflow
- Use AutoML for hyperparameter tuning
- Detect drift and build robust mechanisms into your solutions
- Supercharge your error handling with robust control flows and vulnerability scanning
- Host and build an ML microservice using AWS and Flask
Who This Book Is For
This book is designed for MLOps and ML engineers, data scientists, and software developers who want to build robust solutions that use machine learning to solve real-world problems. If you’re not a developer but want to manage or understand the product lifecycle of these systems, you’ll also find this book useful. It assumes a basic knowledge of machine learning concepts and intermediate programming experience in Python. With its focus on practical skills and real-world examples, this book is an essential resource for anyone looking to advance their machine learning engineering career.
Table of Contents
- Introduction to Machine Learning Engineering
- The ML Development Process
- From Model-to-Model Factory - Training Systems
- Packaging Up
- Deployment Patterns and Tools
- Scaling Up
- Deep Learning, Generative AI and LLMOps Learning Use Case
- Building an example ML microservice
- Building an Extract Transform Machine Learning Use Case
Andrew P. McMahon has spent years building high-impact ML products across a variety of industries. He is currently Head of MLOps for NatWest Group in the UK and has a PhD in theoretical condensed matter physics from Imperial College London. He is an active blogger, speaker, podcast guest, and leading voice in the MLOps community. He is co-host of the AI Right podcast and was named ‘Rising Star of the Year’ at the 2022 British Data Awards and ‘Data Scientist of the Year’ by the Data Science Foundation in 2019.
