Google Professional Machine Learning Engineer Practice Test 34
Question 330
During a production ML inference release on Google Cloud, the following condition occurs: A document classification service sends huge media files through synchronous prediction calls whose processing exceeds clients' timeout limits. The service must meet its agreed quality and latency requirements. Which result is the strongest postchange evidence of success?
- Large documents finish without client timeout and consumers retrieve signed results using recorded job IDs.
- A test client on the approved network succeeds while an external probe remains blocked.
- The nightly job completes within the agreed SLA without idle real-time serving charges.
- Only a limited portion initially uses the candidate and a simulated regression triggers a tested rollback.
Explanation
Correct Answer
A:Large documents finish without client timeout and consumers retrieve signed results using recorded job IDs. This is specific acceptance evidence for online prediction timeout and batch fallback. Synchronous request limits cannot be solved merely by expanding the model context or retrying indefinitely. It confirms the controlled correction: Use asynchronous or batch processing with durable job status for oversized workloads, reserving online calls for bounded requests.
Incorrect Answers
B:A test client on the approved network succeeds while an external probe remains blocked. This result could demonstrate success for private vpc routing failure for serving but not for the stated online prediction timeout and batch fallback issue. The direct test is: Large documents finish without client timeout and consumers retrieve signed results using recorded job IDs.
C:The nightly job completes within the agreed SLA without idle real-time serving charges. This result could demonstrate success for batch versus online inference schedule but not for the stated online prediction timeout and batch fallback issue. The direct test is: Large documents finish without client timeout and consumers retrieve signed results using recorded job IDs.
D:Only a limited portion initially uses the candidate and a simulated regression triggers a tested rollback. This result could demonstrate success for canary gradual traffic rollout but not for the stated online prediction timeout and batch fallback issue. The direct test is: Large documents finish without client timeout and consumers retrieve signed results using recorded job IDs.
Question 331
During a staged restoration test at a secondary serving deployment, the following condition occurs: A lightweight CPU model needs a custom HTTP handler with sporadic traffic, but the team provisions accelerator-bound infrastructure despite no accelerator operations. The team must prove an actual user-visible improvement without weakening access controls. Which result is the strongest postchange evidence of success?
- The selected service meets p95 latency and payload requirements with lower idle cost and reliable scaling.
- Large documents finish without client timeout and consumers retrieve signed results using recorded job IDs.
- Every controlled sample's returned label matches the model's trained class index with no misordered classes.
- Latency and memory decline on target hardware while validation degradation stays within the approved threshold.
Explanation
Correct Answer
A:The selected service meets p95 latency and payload requirements with lower idle cost and reliable scaling. This is specific acceptance evidence for cloud run versus managed inference endpoint. Containerized CPU inference can be appropriate on serverless compute; specialized ML endpoints are not required for every model. It confirms the controlled correction: Benchmark a Cloud Run custom container against a managed Agent Platform endpoint and choose for workload fit. This matters here because during a staged restoration test at a secondary serving deployment, the following condition occurs: A lightweight….
Incorrect Answers
B:Large documents finish without client timeout and consumers retrieve signed results using recorded job IDs. This result could demonstrate success for online prediction timeout and batch fallback but not for the stated cloud run versus managed inference endpoint issue. The direct test is: The selected service meets p95 latency and payload requirements with lower idle cost and reliable scaling.
C:Every controlled sample's returned label matches the model's trained class index with no misordered classes. This result could demonstrate success for custom postprocessing label mapping drift but not for the stated cloud run versus managed inference endpoint issue. The direct test is: The selected service meets p95 latency and payload requirements with lower idle cost and reliable scaling.
D:Latency and memory decline on target hardware while validation degradation stays within the approved threshold. This result could demonstrate success for inference optimization using quantization but not for the stated cloud run versus managed inference endpoint issue. The direct test is: The selected service meets p95 latency and payload requirements with lower idle cost and reliable scaling.
Question 332
During a staged restoration test at a secondary serving deployment, the following condition occurs: A PyTorch model container builds successfully but the serving platform marks its deployed revision unhealthy due to a missing health-check route and incorrect listener port. The team must prove an actual user-visible improvement without weakening access controls. Which outcome most directly demonstrates a successful, measurable fix?
- The model loads at target concurrency without OOM and meets latency and cost targets.
- Every controlled sample's returned label matches the model's trained class index with no misordered classes.
- A known prediction request succeeds through the deployed endpoint and readiness reflects successful initialization.
- The selected service meets p95 latency and payload requirements with lower idle cost and reliable scaling.
Explanation
Correct Answer
C:A known prediction request succeeds through the deployed endpoint and readiness reflects successful initialization. This is specific acceptance evidence for inference container health interface mismatch. A valid container image is not necessarily compatible with an inference service's HTTP startup and health contract. It confirms the controlled correction: Implement the required serving contract and make the container report readiness only after loading model assets.
Incorrect Answers
A:The model loads at target concurrency without OOM and meets latency and cost targets. This result could demonstrate success for gpu cost and memory fit for large embedding model but not for the stated inference container health interface mismatch issue. The direct test is: A known prediction request succeeds through the deployed endpoint and readiness reflects successful initialization.
B:Every controlled sample's returned label matches the model's trained class index with no misordered classes. This result could demonstrate success for custom postprocessing label mapping drift but not for the stated inference container health interface mismatch issue. The direct test is: A known prediction request succeeds through the deployed endpoint and readiness reflects successful initialization.
D:The selected service meets p95 latency and payload requirements with lower idle cost and reliable scaling. This result could demonstrate success for cloud run versus managed inference endpoint but not for the stated inference container health interface mismatch issue. The direct test is: A known prediction request succeeds through the deployed endpoint and readiness reflects successful initialization.
Question 333
At a Google Cloud analytics team preparing a governed pilot, the following condition occurs: A financial regression model reports astonishing test accuracy because its validation query includes a post-settlement field derived from the target. The team must satisfy a reproducible low-code AI assessment. Which targeted engineering change most directly addresses the failure?
- Train an appropriately scaled BigQuery ML KMEANS model and assess cluster cohesion, size and business interpretability.
- Move deterministic feature transformations into the supported BigQuery ML TRANSFORM workflow and use the same model at inference.
- Exclude post-outcome fields, create point-in-time valid training and evaluation partitions, and rerun ML.EVALUATE.
- Evaluate Cloud Vision API for supported generic image detection and use custom training only for missing domain-specific classes.
Explanation
Correct Answer
C:Exclude post-outcome fields, create point-in-time valid training and evaluation partitions, and rerun ML.EVALUATE. This is the targeted corrective response for bigquery ml ml.evaluate leakage detection. A label-derived future field leaks the outcome into input features and makes offline metrics deceptively strong. Verify the diagnosis using: Inspect feature timestamps, label generation time, feature attributions, and whether the evaluation partition is independent. Confirm success using: Held-out metrics reflect only features available before the prediction timestamp and remain credible on later periods.
Incorrect Answers
A:Train an appropriately scaled BigQuery ML KMEANS model and assess cluster cohesion, size and business interpretability. This action addresses bigquery ml kmeans customer segmentation, not the failure in the described bigquery ml ml.evaluate leakage detection case. The targeted correction is: Exclude post-outcome fields, create point-in-time valid training and evaluation partitions, and rerun ML.EVALUATE.
B:Move deterministic feature transformations into the supported BigQuery ML TRANSFORM workflow and use the same model at inference. This action addresses bigquery ml transform serving parity, not the failure in the described bigquery ml ml.evaluate leakage detection case. The targeted correction is: Exclude post-outcome fields, create point-in-time valid training and evaluation partitions, and rerun ML.EVALUATE.
D:Evaluate Cloud Vision API for supported generic image detection and use custom training only for missing domain-specific classes. This action addresses vision api pretrained image labels versus custom detection, not the failure in the described bigquery ml ml.evaluate leakage detection case. The targeted correction is: Exclude post-outcome fields, create point-in-time valid training and evaluation partitions, and rerun ML.EVALUATE.
Question 334
While reproducing a training orchestration incident in isolated staging, the following condition occurs: A continuous-training pipeline reads a mutable all-data table each run and never logs which rows produced its newest artifact. The team requires an idempotent pipeline behavior and traceable evidence. Which targeted engineering change most directly addresses the failure?
- Configure failure and stale-model alerts tied to actionable run-status and freshness service objectives.
- Retain the last healthy model and require canary validation before switching all prediction traffic.
- Create immutable versioned input snapshots or partition manifests and record them in training metadata.
- Package the fitted preprocessing component with the model and apply the same validated transform in both paths.
Explanation
Correct Answer
C:Create immutable versioned input snapshots or partition manifests and record them in training metadata. This is the targeted corrective response for data version reproducibility in ct jobs. Model retraining reproducibility requires stable inputs rather than unversioned reads of an evolving table. Verify the diagnosis using: Inspect snapshot identifiers, SQL filters, table partitions and the model version's recorded input manifest. Confirm success using: A previous model can be rebuilt from its exact documented training dataset and code revision.
Incorrect Answers
A:Configure failure and stale-model alerts tied to actionable run-status and freshness service objectives. This action addresses pipeline notification alert on silent failure, not the failure in the described data version reproducibility in ct jobs case. The targeted correction is: Create immutable versioned input snapshots or partition manifests and record them in training metadata.
B:Retain the last healthy model and require canary validation before switching all prediction traffic. This action addresses continuous training rollback protection, not the failure in the described data version reproducibility in ct jobs case. The targeted correction is: Create immutable versioned input snapshots or partition manifests and record them in training metadata.
D:Package the fitted preprocessing component with the model and apply the same validated transform in both paths. This action addresses training serving preprocessing artifact sharing, not the failure in the described data version reproducibility in ct jobs case. The targeted correction is: Create immutable versioned input snapshots or partition manifests and record them in training metadata.
Question 335
During a regulated MLOps pipeline deployment, the following condition occurs: A freshly trained PyTorch model deploys but rejects the production client's JSON keys because only training unit tests ran in CI. The workflow must be repeatable and protect the last approved production model. Select TWO: the decisive diagnostic step and the targeted repair for this specific issue.
- Inspect training data freshness, drift signals, label availability, business cadence and automated promotion rules.
- Inspect request schema, container route behavior, test fixtures and CI stage dependency graph.
- Retain the last healthy model and require canary validation before switching all prediction traffic.
- Make the load idempotent with keyed merge or overwrite-per-version semantics and safe retry handling.
- Require inference contract and integration tests with representative requests before promoting each candidate artifact.
Explanation
Correct Answers
B:Inspect request schema, container route behavior, test fixtures and CI stage dependency graph. This inspects evidence relevant to continuous delivery integration contract tests. Training success does not guarantee compatibility with the deployed model's consumer API. After identifying the cause, the matching correction is: Require inference contract and integration tests with representative requests before promoting each candidate artifact. This matters here because a freshly trained PyTorch model deploys but rejects the production client's JSON keys because only training unit….
E:Require inference contract and integration tests with representative requests before promoting each candidate artifact. This is the targeted corrective response for continuous delivery integration contract tests. Training success does not guarantee compatibility with the deployed model's consumer API. Verify the diagnosis using: Inspect request schema, container route behavior, test fixtures and CI stage dependency graph. Confirm success using: A deliberately incompatible payload prevents the release and a compatible fixture completes an end-to-end prediction. This matters here because a freshly trained PyTorch model deploys but rejects the production client's JSON keys because only training unit….
Incorrect Answers
A:Inspect training data freshness, drift signals, label availability, business cadence and automated promotion rules. This is an action or diagnostic relevant to data drift triggered retraining policy, not one of the required two steps. In this scenario verify: Inspect request schema, container route behavior, test fixtures and CI stage dependency graph. Then correct: Require inference contract and integration tests with representative requests before promoting each candidate artifact.
C:Retain the last healthy model and require canary validation before switching all prediction traffic. This is an action or diagnostic relevant to continuous training rollback protection, not one of the required two steps. In this scenario verify: Inspect request schema, container route behavior, test fixtures and CI stage dependency graph. Then correct: Require inference contract and integration tests with representative requests before promoting each candidate artifact.
D:Make the load idempotent with keyed merge or overwrite-per-version semantics and safe retry handling. This is an action or diagnostic relevant to airflow task idempotency on retry, not one of the required two steps. In this scenario verify: Inspect request schema, container route behavior, test fixtures and CI stage dependency graph. Then correct: Require inference contract and integration tests with representative requests before promoting each candidate artifact.
Question 336
During a controlled replay of the data and model handoff to a second engineering team, the following condition occurs: A shared prototype notebook can read entire sensitive production BigQuery datasets because its attached credentials are inherited from a broadly privileged user. The handoff must preserve privacy, lineage and testable behavior. What acceptance result verifies the intended behavior after remediation?
- The equivalent daily query scans only the eligible partitions and returns the same number of matched entities.
- A held-out clean sample preserves meaningful ranges while invalid spikes trigger visible data-quality exceptions.
- The notebook still trains on approved sample tables while restricted production datasets deny unauthorized reads.
- A newly started Workbench or Colab Enterprise runtime reproduces the complete experiment without undocumented manual cell order.
Explanation
Correct Answer
C:The notebook still trains on approved sample tables while restricted production datasets deny unauthorized reads. This is specific acceptance evidence for colab enterprise notebook data access scope. Collaborative notebooks execute with real identities; editor convenience does not justify unrestricted production-data access. It confirms the controlled correction: Use a narrowly scoped service identity or approved user access, least-privilege dataset views and separated development data.
Incorrect Answers
A:The equivalent daily query scans only the eligible partitions and returns the same number of matched entities. This result could demonstrate success for bigquery partition filter cost optimization but not for the stated colab enterprise notebook data access scope issue. The direct test is: The notebook still trains on approved sample tables while restricted production datasets deny unauthorized reads.
B:A held-out clean sample preserves meaningful ranges while invalid spikes trigger visible data-quality exceptions. This result could demonstrate success for outlier handling before feature standardization but not for the stated colab enterprise notebook data access scope issue. The direct test is: The notebook still trains on approved sample tables while restricted production datasets deny unauthorized reads.
D:A newly started Workbench or Colab Enterprise runtime reproduces the complete experiment without undocumented manual cell order. This result could demonstrate success for notebook transient state and execution order but not for the stated colab enterprise notebook data access scope issue. The direct test is: The notebook still trains on approved sample tables while restricted production datasets deny unauthorized reads.
Question 337
During model-training validation on a Google Cloud experimental project, the following condition occurs: A regulated loan approval team chooses a large neural network despite needing explanations for which borrower attributes affected decisions. The change should preserve a meaningful independent model-quality comparison. What implementation change resolves the specific issue without substituting an unrelated fix?
- Choose a decision threshold using acceptable recall and alert burden with a validated independent cohort.
- Compare an interpretable model or validated feature attribution approach with the black box and document decision tradeoffs.
- Use licensed and permitted training images with appropriate governance reviews before any supported tuning job.
- Select a regression architecture and matching numeric loss rather than using a binary classifier for amounts.
Explanation
Correct Answer
B:Compare an interpretable model or validated feature attribution approach with the black box and document decision tradeoffs. This is the targeted corrective response for interpretability-constrained model baseline. Interpretability is an engineering requirement that must be considered with predictive power, not appended as a label. Verify the diagnosis using: Examine model governance constraints, feature sensitivity, explanation fidelity and performance of simpler interpretable baselines. Confirm success using: Auditors can reproduce feature influence evidence while holdout quality remains within the approved threshold.
Incorrect Answers
A:Choose a decision threshold using acceptable recall and alert burden with a validated independent cohort. This action addresses cost-sensitive threshold for rare positives, not the failure in the described interpretability-constrained model baseline case. The targeted correction is: Compare an interpretable model or validated feature attribution approach with the black box and document decision tradeoffs.
C:Use licensed and permitted training images with appropriate governance reviews before any supported tuning job. This action addresses fine-tuning image generator data rights, not the failure in the described interpretability-constrained model baseline case. The targeted correction is: Compare an interpretable model or validated feature attribution approach with the black box and document decision tradeoffs.
D:Select a regression architecture and matching numeric loss rather than using a binary classifier for amounts. This action addresses regression versus classification product choice, not the failure in the described interpretability-constrained model baseline case. The targeted correction is: Compare an interpretable model or validated feature attribution approach with the black box and document decision tradeoffs.
Question 338
During an independent acceptance rehearsal of a customer-facing AI capability, the following condition occurs: A financial regression model reports astonishing test accuracy because its validation query includes a post-settlement field derived from the target. The owners require measurable quality and controlled operating cost. Which evidence most directly establishes what is wrong with this AI or ML workflow?
- Inspect the timestamp ordering, seasonal patterns, horizon requirement, and the BigQuery ML model_type and time-series options.
- Inspect feature timestamps, label generation time, feature attributions, and whether the evaluation partition is independent.
- Inspect label availability, feature types, tuning needs, governance requirements, and supported AutoML model interfaces.
- Inspect class definitions, sample error costs, throughput, p95 latency, pricing and published model capability constraints.
Explanation
Correct Answer
B:Inspect feature timestamps, label generation time, feature attributions, and whether the evaluation partition is independent. This inspects evidence relevant to bigquery ml ml.evaluate leakage detection. A label-derived future field leaks the outcome into input features and makes offline metrics deceptively strong. After identifying the cause, the matching correction is: Exclude post-outcome fields, create point-in-time valid training and evaluation partitions, and rerun ML.EVALUATE.
Incorrect Answers
A:Inspect the timestamp ordering, seasonal patterns, horizon requirement, and the BigQuery ML model_type and time-series options. This examines bigquery arima_plus forecast horizon rather than the specific mechanism reported here. It is not the decisive first check for bigquery ml ml.evaluate leakage detection. The relevant evidence is: Inspect feature timestamps, label generation time, feature attributions, and whether the evaluation partition is independent.
C:Inspect label availability, feature types, tuning needs, governance requirements, and supported AutoML model interfaces. This examines automl tabular supervised model selection rather than the specific mechanism reported here. It is not the decisive first check for bigquery ml ml.evaluate leakage detection. The relevant evidence is: Inspect feature timestamps, label generation time, feature attributions, and whether the evaluation partition is independent.
D:Inspect class definitions, sample error costs, throughput, p95 latency, pricing and published model capability constraints. This examines gemini model choice for latency versus complexity rather than the specific mechanism reported here. It is not the decisive first check for bigquery ml ml.evaluate leakage detection. The relevant evidence is: Inspect feature timestamps, label generation time, feature attributions, and whether the evaluation partition is independent.
Question 339
At a Google Cloud analytics team preparing a governed pilot, the following condition occurs: A high-volume classification API uses a slow reasoning-optimized foundation model for simple short labels even though a smaller supported model may suffice. The team must satisfy a reproducible low-code AI assessment. Which TWO selections represent the correct investigation and remediation together? Select TWO.
- Benchmark a supported lower-latency Gemini variant against the current model under representative traffic before routing requests.
- Evaluate Agent Platform AutoML tabular training against a simple baseline and select it if quality and operating cost justify the managed path.
- Inspect prompt token usage, retrieval relevance, model context requirements, answer grounding and p95 latency across queries.
- Use Cloud Translation API for supported text translation, then separately validate downstream classification against localized content.
- Inspect class definitions, sample error costs, throughput, p95 latency, pricing and published model capability constraints.
Explanation
Correct Answers
A:Benchmark a supported lower-latency Gemini variant against the current model under representative traffic before routing requests. This is the targeted corrective response for gemini model choice for latency versus complexity. Production model selection is a task-quality and cost tradeoff; using the largest model by default is unnecessary. Verify the diagnosis using: Inspect class definitions, sample error costs, throughput, p95 latency, pricing and published model capability constraints. Confirm success using: Classification F1 remains above the acceptance threshold while end-to-end latency and cost per thousand requests fall. Here, the relevant situation is: the following condition occurs: A high-volume classification API uses a slow reasoning-optimized foundation model….
E:Inspect class definitions, sample error costs, throughput, p95 latency, pricing and published model capability constraints. This inspects evidence relevant to gemini model choice for latency versus complexity. Production model selection is a task-quality and cost tradeoff; using the largest model by default is unnecessary. After identifying the cause, the matching correction is: Benchmark a supported lower-latency Gemini variant against the current model under representative traffic before routing requests. Here, the relevant situation is: the following condition occurs: A high-volume classification API uses a slow reasoning-optimized foundation model….
Incorrect Answers
B:Evaluate Agent Platform AutoML tabular training against a simple baseline and select it if quality and operating cost justify the managed path. This is an action or diagnostic relevant to automl tabular supervised model selection, not one of the required two steps. In this scenario verify: Inspect class definitions, sample error costs, throughput, p95 latency, pricing and published model capability constraints. Then correct: Benchmark a supported lower-latency Gemini variant against the current model under representative traffic before routing requests.
C:Inspect prompt token usage, retrieval relevance, model context requirements, answer grounding and p95 latency across queries. This is an action or diagnostic relevant to gemini prompt context token budget, not one of the required two steps. In this scenario verify: Inspect class definitions, sample error costs, throughput, p95 latency, pricing and published model capability constraints. Then correct: Benchmark a supported lower-latency Gemini variant against the current model under representative traffic before routing requests.
D:Use Cloud Translation API for supported text translation, then separately validate downstream classification against localized content. This is an action or diagnostic relevant to translate api for multilingual product descriptions, not one of the required two steps. In this scenario verify: Inspect class definitions, sample error costs, throughput, p95 latency, pricing and published model capability constraints. Then correct: Benchmark a supported lower-latency Gemini variant against the current model under representative traffic before routing requests.
Question 340
While preparing a production training workflow on Gemini Enterprise Agent Platform, the following condition occurs: A customer-service model correctly follows general instructions but lacks newly updated internal warranty details, so developers schedule daily weight fine-tuning. Engineers need an auditable, resource-aware correction to the issue. Select TWO: the decisive diagnostic step and the targeted repair for this specific issue.
- Choose a decision threshold using acceptable recall and alert burden with a validated independent cohort.
- Split raw records first and resample only inside training folds, leaving validation cohorts untouched.
- Determine whether failures arise from changing reference knowledge, stable behavior patterns or inadequate retrieval coverage.
- Inspect time stamps, autocorrelation, horizon, seasonal cycles and out-of-sample forecast metrics against traditional baselines.
- Use grounded retrieval for frequently changing policy facts and reserve fine-tuning for measured durable behavior adaptation.
Explanation
Correct Answers
C:Determine whether failures arise from changing reference knowledge, stable behavior patterns or inadequate retrieval coverage. This inspects evidence relevant to foundation model prompt versus fine-tuning. External reference updates need fresh context; changing foundation weights for every factual revision is inefficient. After identifying the cause, the matching correction is: Use grounded retrieval for frequently changing policy facts and reserve fine-tuning for measured durable behavior adaptation.
E:Use grounded retrieval for frequently changing policy facts and reserve fine-tuning for measured durable behavior adaptation. This is the targeted corrective response for foundation model prompt versus fine-tuning. External reference updates need fresh context; changing foundation weights for every factual revision is inefficient. Verify the diagnosis using: Determine whether failures arise from changing reference knowledge, stable behavior patterns or inadequate retrieval coverage. Confirm success using: New warranty rules appear correctly after the index update without unnecessary retraining and sources remain traceable.
Incorrect Answers
A:Choose a decision threshold using acceptable recall and alert burden with a validated independent cohort. This is an action or diagnostic relevant to cost-sensitive threshold for rare positives, not one of the required two steps. In this scenario verify: Determine whether failures arise from changing reference knowledge, stable behavior patterns or inadequate retrieval coverage. Then correct: Use grounded retrieval for frequently changing policy facts and reserve fine-tuning for measured durable behavior adaptation.
B:Split raw records first and resample only inside training folds, leaving validation cohorts untouched. This is an action or diagnostic relevant to training class imbalance fold leakage, not one of the required two steps. In this scenario verify: Determine whether failures arise from changing reference knowledge, stable behavior patterns or inadequate retrieval coverage. Then correct: Use grounded retrieval for frequently changing policy facts and reserve fine-tuning for measured durable behavior adaptation.
D:Inspect time stamps, autocorrelation, horizon, seasonal cycles and out-of-sample forecast metrics against traditional baselines. This is an action or diagnostic relevant to forecast trend and seasonality versus llm, not one of the required two steps. In this scenario verify: Determine whether failures arise from changing reference knowledge, stable behavior patterns or inadequate retrieval coverage. Then correct: Use grounded retrieval for frequently changing policy facts and reserve fine-tuning for measured durable behavior adaptation.