Building Data Science Applications with FastAPI - Second Edition: Develop, Manage, and Deploy Efficient Machine Learning Applications with Python

Learn all the features and best practices of FastAPI to build, deploy, and monitor powerful data science and AI apps, like object detection or image generation. Purchase of the print or Kindle book includes a free PDF eBook Key Features Uncover the secrets of FastAPI, including async I/O, type hinting, and dependency injection Learn to add authentication, authorization, and interaction with databases in a FastAPI backend Develop real-world projects using pre-trained AI models Book Description Building Data Science Applications with FastAPI is the go-to resource for creating efficient and dependable data science API backends. This second edition incorporates the latest Python and FastAPI advancements, along with two new AI projects – a real-time object detection system and a text-to-image generation platform using Stable Diffusion. The book starts with the basics of FastAPI and modern Python programming. You'll grasp FastAPI's robust dependency injection system, which facilitates seamless database communication, authentication implementation, and ML model integration. As you progress, you'll learn testing and deployment best practices, guaranteeing high-quality, resilient applications. Throughout the book, you'll build data science applications using FastAPI with the help of projects covering common AI use cases, such as object detection and text-to-image generation. These hands-on experiences will deepen your understanding of using FastAPI in real-world scenarios. By the end of this book, you'll be well equipped to maintain, design, and monitor applications to meet the highest programming standards using FastAPI, empowering you to create fast and reliable data science API backends with ease while keeping up with the latest advancements. What you will learn Explore the basics of modern Python and async I/O programming Get to grips with basic and advanced concepts of the FastAPI framework Deploy a performant and reliable web backend for a data science application Integrate common Python data science libraries into a web backend Integrate an object detection algorithm into a FastAPI backend Build a distributed text-to-image AI system with Stable Diffusion Add metrics and logging and learn how to monitor them Who this book is for This book is for data scientists and software developers interested in gaining knowledge of FastAPI and its ecosystem to build data science applications. Basic knowledge of data science and machine learning concepts and how to apply them in Python is recommended. Contributors About the author About the reviewers Preface Who this book is for What this book covers To get the most out of this book Download the example code files Conventions used Get in touch Share Your Thoughts Download a free PDF copy of this book Part 1: Introduction to Python and FastAPI Chapter 1: Python Development Environment Setup Technical requirements Installing a Python distribution using pyenv Creating a Python virtual environment Installing Python packages with pip Installing the HTTPie command-line utility Summary Chapter 2: Python Programming Specificities Technical requirements Basics of Python programming Running Python scripts Indentation matters Working with built-in types Working with data structures – lists, tuples, dictionaries, and sets Performing Boolean logic and a few other operators Controlling the flow of a program Defining functions Writing and using packages and modules Operating over sequences – list comprehensions and generators List comprehensions Generators Writing object-oriented programs Defining a class Implementing magic methods Reusing logic and avoiding repetition with inheritance Type hinting and type checking with mypy Getting started Type data structures Type function signatures with Callable Any and cast Working with asynchronous I/O Summary Chapter 3: Developing a RESTful API with FastAPI Technical requirements Creating a first endpoint and running it locally Handling request parameters Path parameters Query parameters The request body Form data and file uploads Headers and cookies The request object Customizing the response Path operation parameters The response parameter Raising HTTP errors Building a custom response Structuring a bigger project with multiple routers Summary Chapter 4: Managing Pydantic Data Models in FastAPI Technical requirements Defining models and their field types with Pydantic Standard field types Optional fields and default values Validating email addresses and URLs with Pydantic types Creating model variations with class inheritance Adding custom data validation with Pydantic Applying validation at the field level Applying validation at the object level Applying validation before Pydantic parsing Working with Pydantic objects Converting an object into a dictionary Creating an instance from a sub-class object Updating an instance partially Summary Chapter 5: Dependency Injection in FastAPI Technical requirements What is dependency injection? Creating and using a function dependency Getting an object or raising a 404 error Creating and using a parameterized dependency with a class Using class methods as dependencies Using dependencies at the path, router, and global level Using a dependency on a path decorator Using a dependency on a whole router Using a dependency on a whole application Summary Part 2: Building and Deploying a Complete Web Backend with FastAPI Chapter 6: Databases and Asynchronous ORMs Technical requirements An overview of relational and NoSQL databases Relational databases NoSQL databases Which one should you choose? Communicating with a SQL database with SQLAlchemy ORM Creating ORM models Defining Pydantic models Connecting to a database Creating objects Getting and filtering objects Updating and deleting objects Adding relationships Setting up a database migration system with Alembic Communicating with a MongoDB database using Motor Creating models that are compatible with MongoDB ID Connecting to a database Inserting documents Getting documents Updating and deleting documents Nesting documents Summary Chapter 7: Managing Authentication and Security in FastAPI Technical requirements Security dependencies in FastAPI Storing a user and their password securely in a database Creating models Hashing passwords Implementing registration routes Retrieving a user and generating an access token Implementing a database access token Implementing a login endpoint Securing endpoints with access tokens Configuring CORS and protecting against CSRF attacks Understanding CORS and configuring it in FastAPI Implementing double-submit cookies to prevent CSRF attacks Summary Chapter 8: Defining WebSockets for Two-Way Interactive Communication in FastAPI Technical requirements Understanding the principles of two-way communication with WebSockets Creating a WebSocket with FastAPI Handling concurrency Using dependencies Handling multiple WebSocket connections and broadcasting messages Summary Chapter 9: Testing an API Asynchronously with pytest and HTTPX Technical requirements An introduction to unit testing with pytest Generating tests with parametrize Reusing test logic by creating fixtures Setting up testing tools for FastAPI with HTTPX Writing tests for REST API endpoints Writing tests for POST endpoints Testing with a database Writing tests for WebSocket endpoints Summary Chapter 10: Deploying a FastAPI Project Technical requirements Setting and using environment variables Using a .env file Managing Python dependencies Adding Gunicorn as a server process for deployment Deploying a FastAPI application on a serverless platform Adding database servers Deploying a FastAPI application with Docker Writing a Dockerfile Adding a prestart script Building a Docker image Running a Docker image locally Deploying a Docker image Deploying a FastAPI application on a traditional server Summary Part 3: Building Resilient and Distributed Data Science Systems with FastAPI Chapter 11: Introduction to Data Science in Python Technical requirements What is machine learning? Supervised versus unsupervised learning Model validation Manipulating arrays with NumPy and pandas Getting started with NumPy Manipulating arrays with NumPy – computation, aggregations, and comparisons Getting started with pandas Training models with scikit-learn Training models and predicting Chaining preprocessors and estimators with pipelines Validating the model with cross-validation Summary Chapter 12: Creating an Efficient Prediction API Endpoint with FastAPI Technical requirements Persisting a trained model with Joblib Dumping a trained model Loading a dumped model Implementing an efficient prediction endpoint Caching results with Joblib Choosing between standard or async functions Summary Chapter 13: Implementing a Real-Time Object Detection System Using WebSockets with FastAPI Technical requirements Using a computer vision model with Hugging Face Implementing a REST endpoint to perform object detection on a single image Implementing a WebSocket to perform object detection on a stream of images Sending a stream of images from the browser in a WebSocket Showing the object detection results in the browser Summary Chapter 14: Creating a Distributed Text-to-Image AI System Using the Stable Diffusion Model Technical requirements Generating images from text prompts with Stable Diffusion Implementing the model in a Python script Executing the Python script Creating a Dramatiq worker and defining an image-generation task Implementing a worker Implementing the REST API Storing results in a database and object storage Sharing data between the worker and the API Storing and serving files in object storage Summary Chapter 15: Monitoring the Health and Performance of a Data Science System Technical requirements Configuring and using a logging facility with Loguru Understanding log levels Adding logs with Loguru Understanding and configuring sinks Structuring logs and adding context Configuring Loguru as the central logger Adding Prometheus metrics Understanding Prometheus and the different metrics Measuring and exposing metrics Adding Prometheus metrics to FastAPI Adding Prometheus metrics to Dramatiq Monitoring metrics in Grafana Configuring Grafana to collect metrics Visualizing metrics in Grafana Summary Index Why subscribe? Other Books You May Enjoy Packt is searching for authors like you Share Your Thoughts Download a free PDF copy of this book