Home › GCP ML Engineer › cloudcomputing › mleops › Which Vertex AI capability should an ML Engineer…
Which Vertex AI capability should an ML Engineer use to evaluate a GENERATIVE-AI solution (LLM-based RAG or fine-tuned LLM) on benchmarks + custom evaluators — referenced in PMLE §2.3?
ACloud Memorystore
BVertex AI gen-AI evaluation services (model-based and rubric-based evaluators in Vertex AI)
CManual eyeballing of a few outputs
DCloud DNS
Answer & Solution
Correct answer: B. Vertex AI gen-AI evaluation services (model-based and rubric-based evaluators in Vertex AI)
Vertex AI gen-AI evaluation is the canonical evaluator (per PMLE §2.3). The other options aren't systematic.
Related questions
Which GCP service should an ML Engineer use to schedule recurring ML pipeline runs (e.g. rWhich CI/CD pattern should an ML Engineer adopt for automated MODEL RETRAINING when new daWhich GCP service should an ML Engineer use to DEPLOY a fine-tuned LLM as a serverless conWhich Responsible AI consideration should an ML Engineer evaluate BEFORE deploying a hirinWhich Vertex AI Model Monitoring signal should an ML Engineer track to detect when INPUT FWhich Vertex AI capability should an ML Engineer use to PRODUCE feature-level explanationsWhich Vertex AI service should an ML Engineer use to MONITOR a deployed model for drift beWhich Vertex AI service should an ML Engineer use to TRACK lineage of ML artifacts across