As a ML engineer, I want a single entry point for retraining the model, so that I can update the model as new interactions arrive.
Acceptance criteria:
- Pipeline script runs extraction, preprocessing, training, and evaluation
- Pipeline outputs trained model artifacts and evaluation report
- Steps and parameters are documented in code or README
Metadata
As a ML engineer, I want a single entry point for retraining the model, so that I can update the model as new interactions arrive.
Acceptance criteria:
Metadata
user story