Learn how to build an end-to-end data to AI solution on Google Cloud using Vertex AI
Key Features
● Harness the power of AutoML capabilities to build machine learning models.
● Learn how to train custom machine learning models on the Google Cloud Platform.
● Accelerate your career in data analytics by leveraging the capabilities of GCP.
Description
Google Cloud Vertex AI is a platform for machine learning (ML) offered by Google Cloud, with the objective of making the creation, deployment, and administration of ML models on a large scale easier. If you are seeking a unified and collaborative environment for your ML projects, this book is a valuable resource for you.
This comprehensive guide is designed to help data enthusiasts effectively utilize Google Cloud Platform's Vertex AI for a wide range of machine learning operations. It covers the basics of the Google Cloud Platform, encompassing cloud storage, big query, and IAM. Subsequently, it delves into the specifics of Vertex AI, including AutoML, custom model training, model deployment on endpoints, development of Vertex AI pipelines, and the Explainable AI feature store.
By the time you finish reading this book, you will be able to navigate Vertex AI proficiently, even if you lack prior experience with cloud platforms. With the inclusion of numerous code examples throughout the book, you will be equipped with the necessary skills and confidence to create machine learning solutions using Vertex AI.
What you will learn
● Learn how to create projects, store data in GCP, and manage access permissions effectively.
● Discover how AutoML can be utilized for streamlining workflows.
● Learn how to construct pipelines using TFX (TensorFlow Extended) and Kubeflow components.
● Gain an overview of the purpose and significance of the Feature Store.
● Explore the concept of explainable AI and its role in understanding machine learning models.
Who this book is for
This book is designed for data scientists and advanced AI practitioners who are interested in learning how to perform machine learning tasks on the Google Cloud Platform. Having prior knowledge of machine learning concepts and proficiency in Python programming would greatly benefit readers.
Author(s): Hemanth Kumar K
Publisher: BPB Online
Year: 2023
Language: English
Pages: 387
Cover
Title Page
Copyright Page
Dedication Page
About the Author
About the Reviewers
Acknowledgements
Preface
Table of Contents
1. Basics of Google Cloud Platform
Introduction
Structure
Objectives
Introduction to Cloud
Advantages of Cloud
Importance of Cloud for data scientist
Types of Cloud
Introduction to Google Cloud Platform
Account creation on Google Cloud Platform
Footprint of Google Cloud Platform
Cloud Service Model
Services offered by GCP
Hierarchy of GCP
New Project Creation
Deletion of Project
Interacting with GCP services
Google Cloud Platform Console
Command Line Interface
APIs
Storage
Working with Google Cloud Storage
Deletion of bucket
Compute
Summary of Compute services of GCP
Creation of compute Engine (VM instances)
Accessing the VM instance
Deletion of VM instance
BigQuery
Working with BigQuery
Identity and access management
Conclusion
Questions
2. Introduction to Vertex AI and AutoML Tabular
Introduction
Structure
Objectives
Introduction to Vertex AI
Key features
Vertex AI prediction service
Working with Vertex AI
Vertex AI AutoML
Creation of tabular datasets
Model training
Model evaluation
Batch predictions
Model deployment for online predictions
Service account creation
Serving online predictions
Model un-deployment and deleting end point
Model deletion
Dataset deletion
Conclusion
Questions
3. AutoML Image, Text, and Pre-built Models
Introduction
Structure
Objectives
Vertex AI AutoML for image data
Image dataset creation
Model training image
Model evaluation image
Batch Predictions image
Model deployment for online predictions image
Serving online predictions image
Vertex AI AutoML for text data
Text dataset creation
Model training text
Model evaluation text
Batch Predictions text
Model deployment for online predictions text
Serving online predictions text
Pre-built models in GCP
Benefits of AutoML
Limitations of AutoML
Conclusion
Questions
4. Vertex AI Workbench and Custom Model Training
Introduction
Structure
Objectives
Vertex AI workbench
Vertex AI Workbench creation and working
Data for building custom model
Introduction to Containers and Dockers
Creation of Dockerfile
Model building
Image creation
Pushing image to container registry
Submitting the custom model training job
Completion of custom model training job
Deletion of resources
Conclusion
Questions
5. Vertex AI Custom Model Hyperparameter and Deployment
Introduction
Structure
Objectives
Hyperparameter in machine learning
Working of hyperparameters tuning
Vertex AI Vizier
Data for building custom model
Creation of workbench
Creation of Dockerfile
Model building code
Image creation
Submitting the custom model training job
Completion of custom model training job
Model importing
Model deployment and predictions
Submitting training job with Python SDK
Deletion of resources
Conclusion
Questions
6. Introduction to Pipelines and Kubeflow
Introduction
Structure
Objectives
What is machine learning pipeline
Vertex AI pipelines
Benefits of machine learning pipelines
Execution
Model versioning and tracking
Troubleshooting
Resource utilization
Introduction to Kubeflow
Components of Kubeflow
Tasks of Kubeflow
Data for model training
API enablement
Additional permission for compute engine
Pipeline code walk through
Execution of Pipeline
Deleting resources
Conclusion
Questions
7. Pipelines using Kubeflow for Custom Models
Introduction
Structure
Objectives
Data for model training
Additional permissions
Creation of Workbench
Pipeline code walk through
Pipeline
Pipeline comparison
Deletion of resources
Differences between Vertex AI and Kubeflow pipelines
Conclusion
Questions
8. Pipelines using TensorFlow Extended
Introduction
Structure
Objectives
What is TensorFlow Extended
TFX Pipelines
Components of TFX
Types of custom components
Functionalities of custom components
Data for pipeline building
Pipeline code walk through
Deletion of resources
Conclusion
Questions
9. Vertex AI Feature Store
Introduction
Structure
Objectives
Knowing Vertex AI feature store
Hierarchy of feature store
Advantages of feature store
Disadvantages of feature store
Data for feature store exercise
Working on feature store using GUI
Working on feature store using Python
Deleting resources
Best practices for feature store
Conclusion
Questions
10. Explainable AI
Introduction
Structure
Objectives
What is Explainable AI
Need of Explainable AI
XAI on Vertex AI
Example-based explanations
Feature-based explanations
Feature attribution methods
Data for Explainable AI exercise
Model training for image data
Image classification model deployment
Explanations for image classification
Tabular classification model deployment
Explanations for tabular data (classification)
Deletion of resources
Limitations of Explainable AI
Conclusion
Questions
Index
