PDE Practice Test 2 — 25 Questions

Question 1

A user named Charlie needs to deploy a model to a Vertex AI Endpoint and also create training jobs. Which role should be assigned to Charlie?

Accepted Answer

roles/aiplatform.user. Charlie needs to deploy a model to a Vertex AI Endpoint and create training jobs. The `roles/aiplatform.user` role grants the necessary permissions to use all Vertex AI resources, including creating and managing endpoints, training jobs, models, and predictions. This role is the minimum required for a user to interact with Vertex AI services without granting broader project-level permissions.

Answer

roles/owner

Answer

roles/aiplatform.modelUser

Answer

roles/editor

Question 2

An MLOps team wants to implement continuous deployment of ML models using Cloud Build and Vertex AI. They have a GitHub repository with training code. What should they use?

Accepted Answer

Cloud Build trigger with a custom step to run Vertex AI Training job and deploy. Option D is correct because it directly addresses the requirement for continuous deployment of ML models using Cloud Build and Vertex AI. A Cloud Build trigger can be configured to fire on GitHub commits, and a custom step in the Cloud Build pipeline can invoke a Vertex AI Training job, followed by deploying the trained model to Vertex AI Endpoints. This provides a fully automated CI/CD pipeline for ML models without additional orchestration overhead.

Answer

Deploy using Cloud Run

Answer

Vertex AI Pipelines integrated with Cloud Build

Answer

Cloud Functions to monitor GitHub

Question 3

A data engineer is setting up CI/CD for a machine learning model using Cloud Build and Vertex AI. Which two components are essential? (Select 2)

Accepted Answer

Container Registry for model images. Container Registry (option B) is essential because it stores the Docker container images that encapsulate the trained model and its dependencies, which Cloud Build builds and pushes to the registry. Vertex AI Endpoints (option D) is essential because it provides the managed serving infrastructure to deploy the model image and expose it as a REST API for online predictions, enabling the CI/CD pipeline to automatically update the endpoint with new model versions.

Answer

Cloud Storage for datasets

Answer

Cloud Source Repositories

Answer

Cloud Functions for triggers

Question 4

A team has multiple versions of a model and wants to manage them centrally, including tracking metadata and promoting versions to production. Which tool should they use?

Accepted Answer

Vertex AI Model Registry. Vertex AI Model Registry is the correct tool because it is purpose-built for centrally managing multiple model versions, tracking metadata (such as training parameters, evaluation metrics, and lineage), and promoting versions through stages like staging to production. Unlike generic storage or version control systems, it provides native integration with Vertex AI Pipelines and endpoints for controlled rollout and rollback.

Answer

Cloud Storage

Answer

BigQuery

Answer

GitHub

Question 5

You manage a team that deploys multiple versions of a computer vision model for A/B testing on Vertex AI Endpoints. You need to route a small percentage of traffic to a canary version while the rest goes to the stable version. You also need to gradually increase the canary traffic over time based on performance metrics. Which approach should you take?

Accepted Answer

Deploy both models to the same endpoint and configure traffic splitting percentages using the Vertex AI console or API.. Vertex AI Endpoints natively support traffic splitting between model versions deployed to the same endpoint. This allows you to assign a percentage of traffic (e.g., 5%) to a canary version and the remainder to the stable version, and then adjust the split over time via the console or API as performance metrics dictate. This approach avoids the complexity and latency of external load balancers or application-level routing.

Answer

Create two separate endpoints, one for each version, and use a separate load balancer to route a percentage of requests to the canary endpoint.

Answer

Use Cloud Armor with weighted backend services to route a portion of requests to the canary version.

Answer

Implement feature flags in the application code to randomly select the model version for each prediction request.

Question 6

Your team uses Vertex AI Feature Store to serve features for online predictions. A feature value changes frequently (e.g., user session clicks). Which type of feature should you use to ensure low-latency writes and reads?

Accepted Answer

Streaming feature. A is correct because streaming features in Vertex AI Feature Store are designed for low-latency writes and reads, making them ideal for frequently changing values like user session clicks. They use an online serving infrastructure (typically backed by Bigtable) that supports real-time updates and sub-millisecond retrieval, ensuring predictions are based on the latest data without batch delays.

Answer

Batch feature

Answer

Feature view

Answer

Bigtable-backed feature

Question 7

A user gets the above error when trying to get online predictions. The model was created and the endpoint exists. What is the most likely reason?

Accepted Answer

No version of the model is deployed to the endpoint.. Option C is correct because the error 'No version of the model is deployed to the endpoint' occurs when the endpoint exists but has no active model version assigned to it. In Amazon SageMaker, an endpoint must have at least one production variant with a deployed model version to serve predictions. Without a deployed version, the endpoint is essentially empty and cannot handle inference requests, even though the endpoint resource itself is created.

Answer

The endpoint does not exist.

Answer

The endpoint is in a different region than the model.

Answer

The model does not exist.

Question 8

A team has trained a scikit-learn model and wants to deploy it to AI Platform Prediction for online predictions. What is the required format for the model artifact?

Accepted Answer

A model.joblib file (or model.pkl) along with any custom code.. AI Platform Prediction (now Vertex AI) supports scikit-learn models natively. The required artifact format is a serialized model file (model.joblib or model.pkl) optionally accompanied by any custom code dependencies. This is because scikit-learn models are pickled objects, and the platform deserializes them using the same Python environment specified in the runtime version.

Answer

A single .h5 file containing the model weights.

Answer

A SavedModel directory containing the model for TensorFlow.

Answer

A model.pt file for PyTorch models.

Question 9

A data engineering team is building a CI/CD pipeline for machine learning models using Cloud Build and AI Platform. Which TWO practices are essential for ensuring reproducible and safe model deployments?

Accepted Answer

Tag each model version with the Git commit hash of the training code.. Tagging each model version with the Git commit hash of the training code (Option B) ensures full traceability from code to deployed model. This practice allows the team to exactly reproduce the training environment and code state, which is critical for debugging, auditing, and rolling back to a known-good version. Without this link, the model becomes a black box, and any attempt to recreate it relies on undocumented assumptions.

Answer

Use Cloud Functions to trigger retraining on new data arrival.

Answer

Use the same environment for training and serving, possibly via custom containers.

Answer

Directly deploy from the development environment using gcloud commands.

Question 10

A company has a model that requires GPU for inference and has strict latency requirements. They deployed on Vertex AI Endpoint with autoscaling but observe cold start latency when scaling up. What is the best solution?

Accepted Answer

Set a higher min_replica_count to keep instances warm. Setting a higher min_replica_count ensures that a baseline number of GPU instances are always running and ready to serve inference requests, eliminating cold start latency because new instances do not need to be provisioned and loaded from scratch when traffic spikes. This directly addresses the autoscaling-induced cold start issue by maintaining a warm pool of replicas.

Answer

Pre-compile the model with TensorRT

Answer

Use a larger GPU instance

Answer

Switch to batch prediction

Question 11

A company wants to implement model monitoring for a deployed classification model. Which three types of monitoring should they set up? (Select 3)

Accepted Answer

Training-serving skew. Training-serving skew (B) is correct because it detects discrepancies between the data used for training and the data the model sees in production, which can cause performance degradation. This is a critical monitoring type for classification models to ensure the model's assumptions remain valid in the live environment.

Answer

Infrastructure cost monitoring

Answer

Model version comparison

Question 12

A production model deployed on Vertex AI Endpoint is experiencing high latency during traffic spikes. The current configuration uses a single replica. What is the most efficient solution?

Accepted Answer

Enable autoscaling with minReplicaCount=1 and maxReplicaCount=10. Option B is correct because enabling autoscaling with minReplicaCount=1 and maxReplicaCount=10 allows Vertex AI Endpoint to dynamically add replicas during traffic spikes, distributing inference requests across multiple instances and reducing latency. This is the most efficient solution as it scales resources up only when needed, avoiding over-provisioning and minimizing cost during low traffic periods.

Answer

Set a higher min replica count (e.g., 3)

Answer

Use a larger machine type (e.g., n1-highmem-8)

Answer

Switch to batch prediction to handle spikes

Question 13

Refer to the exhibit. What is the cause of this error?

Accepted Answer

The machine type flag is only used during model deployment, not endpoint creation. The error occurs because the `machine_type` flag is only valid during model deployment (when creating a deployment in Vertex AI), not during endpoint creation. When creating an endpoint, you specify the endpoint name and region, but the machine type is configured later when deploying a model to that endpoint. Attempting to set `machine_type` during endpoint creation causes a validation error because the API does not accept that parameter at that stage.

Answer

The endpoint name already exists

Answer

The user must specify a model name

Answer

The region is missing

Question 14

You are designing a system to serve predictions from a large language model (LLM) with a latency SLO of 500ms. The model does not fit on a single GPU and requires model parallelism. You are considering using Vertex AI Endpoints with a custom container. What additional setup is required to achieve the latency target?

Accepted Answer

Use a machine type with multiple GPUs and configure the container to use tensor parallelism.. Option D is correct because the model does not fit on a single GPU and requires model parallelism. Using a machine type with multiple GPUs and configuring the container to use tensor parallelism allows the model to be split across GPUs within a single instance, enabling efficient parallel computation to meet the 500ms latency SLO. Tensor parallelism distributes individual tensor operations across GPUs, reducing communication overhead compared to pipeline parallelism and is a standard approach for large models on multi-GPU instances in Vertex AI.

Answer

Compile the model using TensorFlow XLA to optimize for single GPU execution.

Answer

Deploy the model across multiple endpoints and use a load balancer to send requests to different parts of the model.

Answer

Use Vertex AI Prediction as a service for LLMs, which automatically handles hardware selection.

Question 15

A financial services company deploys a fraud detection model on Vertex AI using a custom prediction container that runs a PyTorch model. The model requires GPU acceleration. The deployment succeeds but predictions return an error: 'CUDA error: out of memory'. What should the team do to resolve this issue?

Accepted Answer

Increase the GPU machine type to one with more memory (e.g., from NVIDIA T4 to A100). Option B is correct because the CUDA out-of-memory error indicates that the GPU's VRAM is insufficient to load the PyTorch model or process the inference batch. Increasing the GPU machine type to one with more memory, such as from an NVIDIA T4 (16 GB) to an A100 (40 or 80 GB), directly resolves the capacity issue. Vertex AI prediction endpoints allow you to select different accelerator types and sizes, and this change ensures the model fits within GPU memory.

Answer

Change the container to use a CPU-only image to avoid CUDA errors

Answer

Enable Vertex AI Model Monitoring to automatically scale the endpoint

Answer

Add CPU replicas to distribute the inferencing load

Question 16

An MLOps team manages a pipeline that retrains an XGBoost classifier weekly using BigQuery data. The pipeline is orchestrated with Cloud Composer and deploys the new model to Vertex AI Endpoint if validation metrics (AUC > 0.9) are met. Over the past month, the deployed model's AUC has dropped from 0.95 to 0.88, despite the training pipeline consistently reporting AUC > 0.9. Which THREE steps should the team take to diagnose and fix this issue?

Accepted Answer

Add a canary deployment step where new model version receives a small percentage of traffic before full rollout.. Option B is correct because a canary deployment allows the team to gradually roll out the new model to a small percentage of traffic, enabling early detection of performance degradation in production before a full rollout. This step directly addresses the discrepancy between training metrics and live performance by exposing the model to real-world data patterns that may differ from the training set. In Cloud Composer and Vertex AI, canary deployments can be implemented by routing a fraction of requests to the new model version and monitoring its AUC in real time.

Answer

Review the training pipeline's hyperparameter tuning configuration to ensure it is not overfitting to stale data.

Answer

Retrain the model using a longer training history to include older data that may still be relevant.

Question 17

Which THREE steps are essential for implementing a continuous training pipeline with Vertex AI?

Accepted Answer

If the new model passes evaluation, deploy it to a production endpoint.. Option A is correct because a continuous training pipeline aims to automate model updates. After a new model is trained and evaluated, deploying it to a production endpoint (e.g., using Vertex AI Endpoints) is the essential final step to serve predictions from the improved model. This completes the automation loop without manual intervention, assuming the evaluation passes predefined thresholds.

Answer

Manually approve each new model version before deployment.

Answer

Deploy the original model once and set it to auto-update.

Question 18

Your company has deployed a machine learning model on Vertex AI Endpoint to serve real-time predictions for a mobile application. The model was trained using TensorFlow and the prediction requests include raw images that are preprocessed by the client before sending. Recently, the application developers reported that the predictions are becoming less accurate over time. They suspect the issue is related to changes in the client-side preprocessing code. You need to verify this hypothesis and monitor for future regressions. What should you do?

Accepted Answer

Enable Vertex AI Model Monitoring for feature attribution and set up alerting on skew detection.. Option D is correct because Vertex AI Model Monitoring can automatically detect skew between the training data distribution and the live prediction data distribution. By enabling feature attribution and alerting on skew detection, you can quantitatively verify whether changes in client-side preprocessing are causing prediction drift, without manual code comparison or disruptive rollbacks.

Answer

Retrain the model using the latest client data to adapt to any changes in preprocessing.

Answer

Roll back to a previous model version that was known to work well and disable automatic retraining.

Answer

Ask the developers to provide the exact preprocessing code and manually compare it with the training pipeline's preprocessing.

Question 19

A company uses Vertex AI Feature Store for serving features to both training and prediction. The team notices that predictions made shortly after training use different feature values, causing a training-serving skew. What is the most effective way to prevent this skew?

Accepted Answer

Configure the Feature Store to use point-in-time lookup using the training timestamp. Option A is correct because point-in-time lookup ensures that feature values used during training are exactly the same as those used during prediction by retrieving the feature value as it existed at the training timestamp. This directly addresses training-serving skew caused by time-dependent feature changes, which is a common issue in Vertex AI Feature Store when features are updated after training.

Answer

Retrain the model more frequently to adapt to the new feature distributions

Answer

Use batch prediction instead of online prediction to ensure consistent features

Answer

Ensure that the training and prediction environments use identical compute resources

Question 20

A data science team has deployed a custom TensorFlow model on Vertex AI Prediction. They notice increasing prediction latency and a growing number of 503 errors during peak traffic hours. The model is served using a single regional endpoint with min replica count of 2 and max replica count of 10. Which TWO actions should the team take to address these issues?

Accepted Answer

Enable autoscaling with a higher max replica count and configure a CPU utilization target of 60%.. Option B is correct because increasing the max replica count and setting a CPU utilization target of 60% allows the Vertex AI Prediction service to scale out more aggressively during traffic spikes, reducing both latency and 503 errors. This directly addresses the bottleneck of insufficient compute capacity under peak load, as the current max of 10 replicas may be too low to handle demand.

Answer

Use a larger machine type (e.g., n1-highmem-8) instead of the current n1-standard-4 to improve per-replica throughput.

Answer

Reduce the min replica count to 0 to allow the service to scale down to zero when not in use.

Answer

Deploy the model as a batch prediction job and move all online predictions to batch.

PDE Practice Test 2 — 25 Questions

Exhibit

PDE Practice Test 2 — 25 Questions

Exhibit