Explanation:
The correct answer is C. This option ensures an organized and efficient method for tracking artifacts and comparing models by creating a Vertex AI pipeline that includes parameters, metrics, and dataset artifacts as inputs and outputs. By associating the pipeline with your experiment when submitting the job, it simplifies transitioning successful experiments to production. Options A and B involve manual steps that can introduce inconsistencies and aren't as streamlined for rapid transitions to production. Option D, while also leveraging pipelines, doesn't provide the same level of tracking granularity during the initial experimentation phase. Hence, option C best meets the criteria for efficiently and rapidly transitioning experiments to production.
Ultimate access to all questions.
No comments yet.
You are tasked with developing a machine learning model within a Vertex AI Workbench notebook environment. Your goal is to effectively track artifacts, such as datasets and models, and to compare performance metrics during experimentation using various approaches. Additionally, you want to ensure that once an experiment yields successful results, you can rapidly and easily transition these experiments into a production environment during ongoing iterations of your model. What should you do?
A
B
C
D