If you’ve survived the training up to this point, congratulations! You’ve just learned many common patterns that can be used in real-world machine learning systems, as well as understanding the tradeoffs when deciding which patterns to apply to your system.
In the last part of the book, we will build an end-to-end machine learning system to apply what we learned previously. We will gain hands-on experience implementing many patterns previously learned in this project. We’ll learn how to solve problems at a larger scale and take what’s developed on our laptops to large distributed clusters.
In chapter 7, we’ll go through the project background and system components. Then, we’ll go through the challenges in each of these components and share the patterns that we will apply to address them. Chapter 8 covers the basic concepts of the four technologies (TensorFlow, Kubernetes, Kubeflow, and Argo Workflows) and provides an opportunity to gain hands-on experience in each one of them to prepare our implementation of the final project.
In the last chapter of the book, we’ll implement the end-to-end machine learning system with the architecture we designed in chapter 7. Our complete implementation of each of the components will incorporate the previously discussed patterns. We’ll use the technologies we learned in chapter 8 to build different components of a distributed machine learning workflow.