An ML engineer wants to run a quick experiment to compare two prompt variations for a generative AI agent. They need minimal setup and want to track the results. Which Google Cloud environment is most suitable?

Experiments on Gemini Enterprise Agent Platform

What is the purpose of tracking model artifacts and versions?

To improve the accuracy of the model

To ensure that the best-performing model can be identified and reproduced

To reduce the storage cost of models

To automatically deploy the latest model to production

A team is using Gemini Enterprise Agent Platform Pipelines to train and deploy models. They need to ensure that the model version deployed to production can be traced back to the exact pipeline run that produced it. They also need to be able to compare the performance of the deployed model with previous versions. What is the best way to achieve this?

Use the model artifact's timestamp to infer the pipeline run

Use a naming convention for model versions that includes the pipeline run ID

Store the pipeline run ID in the model artifact's metadata, and use the model registry to track versions and metrics

Manually record the pipeline run ID in a spreadsheet after deployment

A team is using LLM-as-a-judge to evaluate the quality of responses generated by their chatbot. They are concerned about the judge model being biased toward longer responses or responses that contain certain keywords. What is the most effective way to mitigate this bias?

Provide a detailed rubric in the prompt that defines the evaluation criteria and instructs the judge to ignore length and style

Use only human evaluation and avoid LLM-as-a-judge entirely

Remove all punctuation from the responses before evaluation

Use a different LLM as the judge for each evaluation run to introduce randomness

Which of the following is a benefit of using Gemini Enterprise Agent Platform Pipelines for experimentation?

It automatically deploys models to production

It provides a serverless environment for running experiments

It eliminates the need for data validation

It enables repeatable and reproducible experiment workflows

Google CloudProfessional Machine Learning Engineer

Domain 2Objective 3

Tracking and Running ML Experiments PROFESSIONAL-MACHINE-LEARNING-ENGINEER Practice Questions (Page 3)

Part of the Collaborating within and across teams to manage data and models domain, which accounts for ~16% of the PROFESSIONAL-MACHINE-LEARNING-ENGINEER exam.

29questions here

6free pages

10concepts

~16%of the exam

10 Concepts in This Objective

Choosing Google Cloud environments for ML development
Using Experiments on Gemini Enterprise Agent Platform
Using Gemini Enterprise Agent Platform Pipelines
Using Kubeflow Pipelines
Evaluating predictive ML models
Evaluating generative AI solutions
Using LLM-as-a-judge for evaluation
Tracking model artifacts and versions
Comparing model experiments
Managing model lineage

Questions 11–15

11foundation · easy
An ML engineer wants to run a quick experiment to compare two prompt variations for a generative AI agent. They need minimal setup and want to track the results. Which Google Cloud environment is most suitable?
Select an answer first
12foundation · easy
What is the purpose of tracking model artifacts and versions?
Select an answer first
13expert · hard
A team is using Gemini Enterprise Agent Platform Pipelines to train and deploy models. They need to ensure that the model version deployed to production can be traced back to the exact pipeline run that produced it. They also need to be able to compare the performance of the deployed model with previous versions. What is the best way to achieve this?
Select an answer first
14application · medium
A team is using LLM-as-a-judge to evaluate the quality of responses generated by their chatbot. They are concerned about the judge model being biased toward longer responses or responses that contain certain keywords. What is the most effective way to mitigate this bias?
Select an answer first
15foundation · easy
Which of the following is a benefit of using Gemini Enterprise Agent Platform Pipelines for experimentation?
Select an answer first

Finished these 5 questions?

Review the revealed explanations, or continue through the curriculum.

Next 5 questions →Back to PROFESSIONAL-MACHINE-LEARNING-ENGINEER

Free Basic Practice is a study aid with revealable answers — not a scored exam. Examers.io is independent and not affiliated with or endorsed by Google Cloud. “PROFESSIONAL-MACHINE-LEARNING-ENGINEER” is a trademark of its owner, used for identification only.