
Explanation:
Option C is the correct answer because it provides a comprehensive solution that addresses all requirements: reliability, repeatability, version tracking, lineage, and weekly retraining. Using Vertex AI Pipelines with CustomTrainingJobOp ensures a managed, reproducible training process. ModelUploadOp automatically registers the model in Vertex AI Model Registry, enabling version tracking and lineage. Cloud Scheduler with Cloud Run provides a robust scheduling mechanism for weekly execution. Option A lacks model registration and version tracking. Option B uses Notebooks API for scheduling, which is less reliable than Cloud Scheduler and doesn't provide full pipeline management. Option D uses HyperParameterTuningJobRunOp, which is unnecessary for standard retraining without hyperparameter tuning mentioned in the requirements.
Ultimate access to all questions.
No comments yet.
You have developed a new ML model in a Jupyter notebook and want to create a reliable, repeatable training process that tracks model artifact versions and lineage. You plan to retrain the model weekly. How should you operationalize this training process?
A
B
C
D