PMLE Scaling Prototypes into ML Models Practice Question

You are scaling a prototype ML model to production on Vertex AI. The model is trained with a custom training job and you want to ensure reproducibility and traceability of each training run. Which two practices should you implement? (Choose two.)

⚠ Common exam trap

The trap here is focusing on operational practices like hardware consistency or artifact organization, which do not capture the code and parameter metadata required for true reproducibility and traceability.

Answer choices

Why each option matters

Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.

Correct answer & explanation

✓

Store the training code in a Git repository and pass the commit hash as a hyperparameter or label to the Vertex AI training job.

To ensure reproducibility and traceability, you need to record the exact code version and the training parameters and metrics. Storing the Git commit hash with the training job links the model to the code, while Vertex AI Experiments logs parameters, metrics, and artifacts for each run. Together, these practices allow you to reproduce any model and trace its lineage. Model monitoring, hardware consistency, and timestamped folders do not provide the necessary metadata linkage.

Answer analysis

Option-by-option breakdown

For each option: why learners choose it and why it is or isn't the right answer here.

✗
Store the trained model artifacts in a Cloud Storage bucket with a unique timestamped folder for each run.
Why it's wrong here
Timestamped folders help organize artifacts but do not capture the training code, parameters, or environment. Without metadata linking the artifacts to the exact code and configuration, you cannot reliably reproduce the model. This practice is useful for organization but does not fulfill the need for reproducibility and traceability on its own.
✓
Store the training code in a Git repository and pass the commit hash as a hyperparameter or label to the Vertex AI training job.
Why this is correct
Recording the Git commit hash with the training job creates a direct link between the code version and the resulting model. This enables reproducibility because you can check out the exact code that produced a model, and it provides traceability for audits. Vertex AI training jobs support labels and hyperparameters, so you can include the commit hash as metadata without affecting the training logic.
✗
Use the same machine type and accelerator for every training run to ensure identical hardware.
Why it's wrong here
While consistent hardware can reduce variability, it does not guarantee reproducibility because other factors such as software versions, random seeds, and data order also matter. Moreover, it does not provide traceability of the training run. Hardware consistency alone is insufficient and does not address the requirement to track code and parameters.
✗
Enable Vertex AI Model Monitoring on the endpoint to detect training-serving skew.
Why it's wrong here
Model Monitoring detects skew and drift in deployed models by comparing serving data to training data. While valuable for production monitoring, it does not provide reproducibility or traceability of training runs. It operates after deployment and focuses on data distribution changes, not on capturing the training configuration or code version.
✓
Use Vertex AI Experiments to log parameters, metrics, and artifacts for each training run.
Why this is correct
Vertex AI Experiments automatically tracks parameters, metrics, and artifacts for each run, providing a centralized record. This makes it easy to compare runs, reproduce a specific model by reusing its parameters, and trace which dataset and code version were used. It integrates with Vertex AI Pipelines and custom training jobs, so you can log metadata programmatically or automatically.

Go deeper

PMLE Scaling Prototypes into ML Models Practice Question

Why each option matters

Option-by-option breakdown

Related to this question