Название: MLOps Engineering at Scale
Автор: Carl Osipov
Издательство: Manning Publications
Год: 2022
Страниц: 344
Язык: английский
Формат: epub
Размер: 10.2 MB

MLOps Engineering at Scale teaches you how to implement efficient machine learning systems using pre-built services from AWS and other cloud vendors. This easy-to-follow book guides you step-by-step as you set up your serverless ML infrastructure, even if you’ve never used a cloud platform before. You’ll also explore tools like PyTorch Lightning, Optuna, and MLFlow that make it easy to build pipelines and scale your deep learning models in production.

Who should read this book:

To get the most value from this book, you’ll want to have existing skills in data analysis with Python and SQL, as well as have some experience with machine learning. I expect that if you are reading this book, you are interested in developing your expertise as a machine learning engineer, and you are planning to deploy your machine learning—based prototypes to production.

This book is for information technology professionals or those in academia who have had some exposure to Machine Learning and are working on or are interested in launching a machine learning system in production. There is a refresher on machine learning prerequisites for this book in appendix A. Keep in mind that if you are brand new to machine learning you may find that studying both machine learning and cloud-based infrastructure for machine learning at the same time can be overwhelming.

If you are a software or a data engineer, and you are planning on starting a machine learning project, this book can help you gain a deeper understanding of the machine learning project life cycle. You will see that although the practice of machine learning depends on traditional information technologies (i.e., computing, storage, and networking), it is different from the traditional information technology in practice. The former is significantly more experimental and more iterative than you may have experienced as a software or a data professional, and you should be prepared for the outcomes to be less known in advance. When working with data, the machine learning practice is more like the scientific process, including forming hypotheses about data, testing alternative models to answer questions about the hypothesis, and ranking and choosing the best performing models to launch atop your machine learning platform.

If you are a machine learning engineer or practitioner, or a data scientist, keep in mind that this book is not about making you a better researcher. The book is not written to educate you about the frontiers of science in machine learning. This book also will not attempt to reteach you the machine learning basics, although you may find the material in appendix A, targeted at information technology professionals, a useful reference. Instead, you should expect to use this book to become a more valuable collaborator on your machine learning team. The book will help you do more with what you already know about data science and machine learning so that you can deliver ready-to-use contributions to your project or your organization. For example, you will learn how to implement your insights about improving machine learning model accuracy and turn them into production-ready capabilities.