CI/CD for ML
Automated pipelines that test, validate, and deploy a model on every change, extending ordinary software continuous integration with stages specific to ML: data validation, retraining, and evaluation against a held-out set before a new model version is promoted.
Why exams ask this
Tested as what makes an ML pipeline different from ordinary software CI/CD. The exam wants data and model validation named as an added stage, not just code tests, since a pipeline can pass every unit test and still ship a model trained on corrupted data.
Relevant to
Related concepts
Resources
No resources linked to this concept yet.