Introduction to Machine Learning Deployment
Chapter 1: Getting Started with Model Deployment
What is Machine Learning Deployment?
Why Deploy Machine Learning Models?
The Machine Learning Workflow Overview
Types of Deployment Strategies (Introduction)
Challenges in Model Deployment
Quiz for Chapter 1
Chapter 2: Preparing Your Model for Deployment
Saving Trained Models
Introduction to Model Serialization
Using Pickle for Model Persistence
Using Joblib for Model Persistence
Handling Model Dependencies
Saving Preprocessing Steps
Hands-on Practical: Saving and Loading a Simple Model
Quiz for Chapter 2
Chapter 3: Creating a Prediction Service with Flask
What is an API?
Introduction to Web Frameworks
Setting Up Flask
Creating a Basic Flask Application
Loading Your Saved Model in Flask
Defining a Prediction Endpoint
Handling Input Data (JSON)
Returning Predictions
Testing Your API Locally
Hands-on Practical: Building a Simple Flask Prediction API
Quiz for Chapter 3
Chapter 4: Introduction to Containerization with Docker
What is Containerization?
Introduction to Docker
Docker Concepts: Images and Containers
Installing Docker
Writing a Simple Dockerfile
Building a Docker Image for the Flask App
Running the Application in a Docker Container
Hands-on Practical: Containerizing the Prediction Service
Quiz for Chapter 4
Challenges in Model Deployment
Was this section helpful?
- Designing Machine Learning Systems: An Iterative Approach for Production-Ready AI, Chip Huyen, 2022 (O'Reilly Media) - This book offers a comprehensive guide to designing and building machine learning systems for production, covering data management, model deployment, monitoring, scalability, and MLOps practices.
- Hidden Technical Debt in Machine Learning Systems, D. Sculley, Gary Holt, Daniel Golovin, Eugene Davydov, Todd Phillips, Dietmar Ebner, Vinay Chaudhary, Michael Young, Jean-François Crespo, Dan Dennison, 2015 Advances in Neural Information Processing Systems (NeurIPS), Vol. 28 (Neural Information Processing Systems Foundation, Inc. (NeurIPS)) - This seminal paper highlights the challenges of building and maintaining production machine learning systems, focusing on various sources of technical debt that accumulate over time.
- MLOps: A Taxonomy of Challenges and Best Practices, David Kreuzer, Janek Belling, Matthias Neumüller, Oliver Niggl, and Andreas Bögl, 2022 IEEE Access, Vol. 10 (IEEE) DOI: 10.1109/ACCESS.2022.3216391 - This paper provides a recent and systematic overview of MLOps challenges and best practices, offering a taxonomy that covers the operational aspects of machine learning development and deployment.