How to Store ML Pipeline Artifacts in a MLOps Platform?

You are building a MLOps platform to automate your company’s ML experiments and model retraining. You need to organize the artifacts for dozens of pipelines. How should you store the pipelines’ artifacts?

Store parameters in Cloud SQL, and store the models’ source code and binaries in GitHub.
Store parameters in Cloud SQL, store the models’ source code in GitHub, and store the models’ binaries in Cloud Storage.
Store parameters in Vertex ML Metadata, store the models’ source code in GitHub, and store the models’ binaries in Cloud Storage. Source Reference Answer
Store parameters in Vertex ML Metadata and store the models’ source code and binaries in GitHub.

Community Votes

100%

100% of anonymous learners picked answer C. Votes are pick records left by other test-takers — they are not the verified answer.

Community Insight

This question tests your knowledge of Google Cloud’s native MLOps storage patterns, where the trap is misplacing large model binaries into version control systems instead of object storage.

When building a MLOps platform on Google Cloud, correctly partitioning ML artifacts across specialized services ensures scalability and lineage tracking. The community consensus confirms that Vertex ML Metadata, Cloud Storage, and GitHub form the optimal architecture for parameters, binaries, and source code respectively.

Community Discussion (4 comments)

fitri001 👍 3 Selected: C

Vertex ML Metadata: This service is specifically designed to store and track metadata for ML pipelines, including parameters. It provides a centralized location to manage and query pipeline execution details, making it ideal for dozens of pipelines. Cloud Storage: This is a scalable and cost-effective storage solution for model binaries. It integrates well with Vertex AI and other cloud services. GitHub: While not a Google Cloud service, it's a popular version control system well-suited for storing and managing your models' source code, particularly for collaboration among team members.

pinimichele01 👍 1 Selected: C

shadz10

shadz10 👍 2 Selected: C

https://cloud.google.com/architecture/architecture-for-mlops-using-tfx-kubeflow-pipelines-and-cloud-build

pikachu007 👍 2 Selected: C

A. Cloud SQL and GitHub: Cloud SQL isn't designed for ML metadata management, potentially leading to challenges in tracking experiment details and lineage. B. Cloud SQL, GitHub, and Cloud Storage: While viable, this approach misses the benefits of Vertex ML Metadata for organized ML artifact management. D. Vertex ML Metadata and GitHub: Storing model binaries in GitHub can be inefficient for large files and might incur higher storage costs.

Comments & Corrections

No comments yet — spotted an error or have a note? Share it below.

Expert Analysis

Why the Answer Is Correct

Vertex ML Metadata is purpose-built for tracking experiment parameters, lineage, and dataset versions across multiple pipelines. Cloud Storage provides the scalable, cost-effective object storage required for large model binaries and datasets. GitHub remains the industry standard for managing and version-controlling machine learning source code. Together, these services align with Google Cloud’s recommended MLOps architecture for production environments.

Why the Other Options Are Wrong

Cloud SQL is a relational database unsuited for unstructured ML metadata or high-volume artifact storage. Storing model binaries in GitHub (Options A and D) bloats repositories, slows down cloning, and violates Git’s design principles. Option B incorrectly defaults to Cloud SQL for parameters, missing the automated lineage and query capabilities native to Vertex ML Metadata.

Community Comment Notes

Users consistently highlight that Vertex ML Metadata centralizes pipeline execution details, making it ideal for managing dozens of experiments [1]. The official reference architecture reinforces this triad pattern, emphasizing integration between Vertex AI, Cloud Build, and Cloud Storage [2]. Commenters also note that separating concerns across these three platforms prevents technical debt in scaling MLOps workflows [4].

Official Reference

Exam Strategy

Memorize the distinct roles of Google Cloud services in the ML lifecycle: use managed metadata stores for lineage, object storage for heavy artifacts, and VCS for code. When faced with MLOps architecture questions, always prioritize services explicitly labeled for ML workloads over general-purpose databases.

Related Analysis

Practice All PMLE Questions

Access 65 questions with complete answers and detailed explanations.

View Full PMLE Practice Test →

← Back to PMLE Study Guide