ML environment setup

In Chapter 4, Developing and Deploying ML Models, in the Preparing the platform section, we learned about the ML platform in the Cloud. Then, in Chapter 7, Exploring Google Cloud Vertex AI, we introduced the Vertex AI services. For a customer-trained model development platform, we recommend Vertex AI Workbench user-managed notebooks. Let’s look at the details from the prospects of performance, cost, and security.

With Vertex AI Workbench user-managed notebooks, you have the flexibility and options to implement performance excellency. You can create an instance with the existing deep learning VM images that have the latest ML and data science libraries preinstalled, along with the latest accelerator drivers. Depending on your data, model, and workloads, you can choose the right VM instance type to fit your environment and optimize performance, from general-purpose compute (E2, N1, N2, and N2D), to memory-optimized (M1 and M2), to compute-optimized (C2...