0% completed

All LessonsFree Lessons (3)

Introduction

Course Goals and Structure

Getting Started

Motivation: Why Jupyter Notebooks Aren’t Enough

The ML Training Pipeline

Quiz: Getting Started

Structuring the ML Pipeline

System Architecture Directory Structure Code Organization Code Style Managing Dependencies Quiz: Structuring the ML Pipeline

Directed Acyclic Graphs (DAGs)

What Is a DAG?Topological Sorting Quiz: DAGs

The ML Library

Object-Oriented Programming Configuration Files: OmegaConf Abstract Base Classes Mixins Datasets Models Report Generation Quiz: The ML Library

Project

Create Your First Data Pipeline with a Dashboard

The Pipeline Core

Structuring the Pipeline Core Parsing Command-Line Arguments: Argparse Experiment Tracking Logging The Complete Pipeline Docstrings Quiz: The Pipeline Core

Extending the Pipeline

Supporting a New Dataset Supporting a New Model

Project

Build a News ETL Data Pipeline Using Python and SQLite

Testing

Introduction to Unit Testing Pytest Introduction to System Testing Quiz: Testing

Deployment

Packaging the ML Library Inference

Other Considerations

Data Quality and Model Performance Monitoring Reproducibility Off-the-Shelf Solutions Quiz: Observability and Reproducibility

Wrapping Up

Appendix

Useful Resources

Course Assessment

Final Assessment

Building a Machine Learning Pipeline from Scratch/

...

/

Quiz: Getting Started

Quiz: Getting Started

Test your understanding of the lessons in this chapter.

We'll cover the following...

Technical Quiz

1.

(Select all that apply.) What is true regarding Notebooks? Multi-select

A.

Notebooks are well suited for EDA.

B.

Notebooks can’t be run programmatically.

C.

Notebook cells can be run in any order.

D.

Notebooks can leak sensitive data.

1 / 4

...