PMLE Scaling Prototypes into ML Models Practice Question

You are preparing to scale a prototype ML model to production on Vertex AI. The model is trained with a custom training job, and you want to ensure that the training is reproducible and that you can compare different runs. Which two practices should you follow? (Choose two.)

⚠ Common exam trap

The trap here is thinking that using the latest container or training on all data guarantees reproducibility, when the real keys are controlling randomness and systematically logging run metadata.

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

✓

Use Vertex AI Experiments to log parameters, metrics, and artifacts for each training run.

To make training reproducible and comparable, you should control randomness by setting and recording a fixed seed, and you should use Vertex AI Experiments to log parameters, metrics, and artifacts for each run. These two practices together let you reproduce a run and compare it with others. The other options either harm reproducibility by not pinning the environment or do not directly address run comparison.

Answer analysis

Option-by-option breakdown

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

✓
Use Vertex AI Experiments to log parameters, metrics, and artifacts for each training run.
Why this is correct
Vertex AI Experiments provides a structured way to track parameters, metrics, and artifacts across training runs, making it easy to compare runs and reproduce results. It integrates with Vertex AI Training and helps you maintain a record of what was used for each model, which is a key practice for scaling prototypes to production.
✓
Specify a fixed random seed in your training code and record it along with the training configuration.
Why this is correct
Setting a fixed random seed and recording it with the training configuration helps ensure that runs with the same data and code produce the same results, which is essential for reproducibility. Vertex AI training jobs can log these parameters, and you can compare runs by their recorded seeds and configurations.
✗
Always use the latest pre-built training container without pinning the version so that you get the newest features.
Why it's wrong here
Using an unpinned container version makes runs non-reproducible because the environment can change between runs. For reproducibility, you should pin the container version or use a custom container with a fixed image digest so that the environment is consistent across experiments.
✗
Store the trained model artifacts in a versioned location such as a Cloud Storage bucket with object versioning or a Vertex AI Model Registry entry.
Why it's wrong here
Storing model artifacts in a versioned location is good practice for traceability, but it is not one of the two core practices for reproducibility and comparison of training runs. It helps with model management, but without logging parameters and metrics, you cannot easily compare runs or reproduce the training process itself.
✗
Train on the full dataset without any sampling to ensure that all runs see the same data.
Why it's wrong here
Training on the full dataset does not by itself guarantee reproducibility, and it can be impractical for large datasets. Reproducibility depends on controlling randomness, data splits, and environment, not simply using all data. In fact, you still need to record the exact data version and split to reproduce results.

Go deeper

PMLE Scaling Prototypes into ML Models Practice Question

Why each option matters

Option-by-option breakdown

Related to this question