Testing ML Pipelines

4 hours of instruction

This practical, hands-on course dives into testing of an entire ML pipeline starting from data and model validation and ending with the integration tests of the pipeline as a whole.

OBJECTIVES

1. By the end of this course, participants will be able to test and troubleshoot components of ML pipeline

PREREQUISITES

Creating CI/CD Pipeline for Machine Learning (ML)

SYLLABUS & TOPICS COVERED

1. Foundations of software testing
   • Summarize the concept and uses of ML pipeline testing
   • Outline ML pipeline test approaches
2. Robust testing frameworks
   • Discuss assertion statements and their limitations
   • Compare more sophisticated testing frameworks
3. Setting up a testing framework
   • Understanding different types of tests for ML model pipelines
   • Configuring tests with pytest in Amazon SageMaker
4. Conducting tests on an ML model pipeline
   • Run test cases in a variety of testing and validation domains
   • Troubleshoot failed tests using reporting

SOFTWARE REQUIREMENTS

API Gateway, AWS Sagemaker, Access to AWS accounts, CodeBuild, CodePipeline, Lambda, S3