Production and MLOps Section 096
Testing ML Code and CI
Automatic checks that catch a broken model before your users do.
10 of 10 lessons published Three reading levels on every lesson
Lessons in order
Work top to bottom. Each lesson assumes the one above it.
- Unit testing machine learning code
- Testing a training loop
- Behavioural tests for models
- Golden outputs and regression tests
- Testing output you cannot predict
- Recording and replaying model API calls
- Integration testing an inference server
- Evaluation gates in CI
- GitHub Actions for ML projects
- Automated retraining pipelines