100% Latest & Updated Google Professional Machine Learning Engineer Practice Test Questions, Exam Dumps & Verified Answers!
30 Days Free Updates, Instant Download!
Question 340
An ML platform team is resolving 3 independent workstreams. Workstream A: A fraud team maintains labeled feature tables in BigQuery but training needs distributed custom PyTorch code. Workstream B: A boosted-tree model needs tuning but the team manually tries one parameter at a time and repeatedly inspects the final test set. Workstream C: A supervised model's validation accuracy is suspiciously high because a source field contains post-outcome status and several duplicate entity records cross splits. Select exactly THREE diagnostic observations, one directly addressing each separate workstream.
- Inspect table row counts and network and schema access from the training environment.
- Review the entrypoint, dependencies, runtime image and output path contract.
- Inspect feature timestamps, entity IDs and duplicate hashes around the train-test boundary.
- Define search ranges, a single validation objective and trial budget.
- Inspect per-step GPU memory, sequence length and peak batch allocation.
- Compare image digests, framework versions, nondeterministic operations and input order.
Explanation
Correct Answers
A:The evidence for this case shows situation: Inspect table row counts and network and schema access from the training environment. It directly examines the reported situation: A fraud team maintains labeled feature tables in BigQuery but training needs distributed custom PyTorch code. The corresponding corrective action is: Export or load authorized BigQuery examples through a supported managed training data path.
C:The evidence for this case shows situation: Inspect feature timestamps, entity IDs and duplicate hashes around the train-test boundary. It directly examines the reported situation: A supervised model's validation accuracy is suspiciously high because a source field contains post-outcome status and several duplicate entity records cross splits. The corresponding corrective action is: Remove post-label information and group related entities before splitting.
D:The evidence for this case shows situation: Define search ranges, a single validation objective and trial budget. It directly examines the reported situation: A boosted-tree model needs tuning but the team manually tries one parameter at a time and repeatedly inspects the final test set. The corresponding corrective action is: Use bounded managed hyperparameter tuning on training and validation data only.
Incorrect Answers
B:Reviewing the entrypoint, dependencies, runtime image and output path contract is relevant when a research team needs a custom PyTorch loss and optimizer but wants repeatable training jobs rather than manual notebook sessions; it does not establish the result required by this workstream.
E:Inspecting per-step GPU memory, sequence length and peak batch allocation is relevant when a large transformer job on GPU fails when sequence length doubles and memory allocation spikes; it does not establish the result required by this workstream.
F:Comparing image digests, framework versions, nondeterministic operations and input order is relevant when a custom model changes predictions after a new container image is pulled despite the same source dataset and random seed; it does not establish the result required by this workstream.
Question 341
Before changing the system, the team investigates the following behavior: A research team needs a custom PyTorch loss and optimizer but wants repeatable training jobs rather than manual notebook sessions. Which observation most directly identifies the cause or requirements before a change?
- Inspect Cloud Storage object generations and exact training manifest membership.
- Review the entrypoint, dependencies, runtime image and output path contract.
- Inspect dataset size, GPU budget and validation risk against adapter tuning.
- Inspect supported base-model adapters and memory requirements for fine-tuning.
Explanation
Correct Answer
B:The evidence for this case shows situation: Review the entrypoint, dependencies, runtime image and output path contract. It directly examines the reported situation: A research team needs a custom PyTorch loss and optimizer but wants repeatable training jobs rather than manual notebook sessions. The corresponding corrective action is: Package the training program into a supported custom-training container and submit an Agent Platform job.
Incorrect Answers
A:This evidence concerns a separate issue: A training pipeline uses an S3-style moving folder of CSV files and repeated runs receive different examples each week. It does not establish what is happening in the present case: A research team needs a custom PyTorch loss and optimizer but wants repeatable training jobs rather than manual notebook sessions. The diagnostic evidence to examine here is Review the entrypoint, dependencies, runtime image and output path contract.
C:This evidence concerns a separate issue: A foundation-model team has a small curated dataset but proposes unrestricted full-parameter updates to learn a narrow command format. It does not establish what is happening in the present case: A research team needs a custom PyTorch loss and optimizer but wants repeatable training jobs rather than manual notebook sessions. The diagnostic evidence to examine here is Review the entrypoint, dependencies, runtime image and output path contract.
D:This evidence concerns a separate issue: A team needs a domain-specific output style from a large foundation model but lacks GPU memory for full weight updates. It does not establish what is happening in the present case: A research team needs a custom PyTorch loss and optimizer but wants repeatable training jobs rather than manual notebook sessions. The diagnostic evidence to examine here is Review the entrypoint, dependencies, runtime image and output path contract.
Question 342
An ML platform team is resolving 3 independent workstreams. Workstream A: An image classifier fits on one GPU but takes days; several GPU devices and a sharded dataset are available. Workstream B: A large transformer job on GPU fails when sequence length doubles and memory allocation spikes. Workstream C: A foundation-model team has a small curated dataset but proposes unrestricted full-parameter updates to learn a narrow command format. Select exactly THREE diagnostic observations, one directly addressing each separate workstream.
- Inspect dataset size, GPU budget and validation risk against adapter tuning.
- Inspect supported base-model adapters and memory requirements for fine-tuning.
- Inspect per-step GPU memory, sequence length and peak batch allocation.
- Review the entrypoint, dependencies, runtime image and output path contract.
- Inspect GPU utilization, gradient synchronization cost and data loading per worker.
- Compare image digests, framework versions, nondeterministic operations and input order.
Explanation
Correct Answers
A:The evidence for this case shows situation: Inspect dataset size, GPU budget and validation risk against adapter tuning. It directly examines the reported situation: A foundation-model team has a small curated dataset but proposes unrestricted full-parameter updates to learn a narrow command format. The corresponding corrective action is: Prefer prompt or parameter-efficient tuning first and justify full fine-tuning only by evidence.
C:The evidence for this case shows situation: Inspect per-step GPU memory, sequence length and peak batch allocation. It directly examines the reported situation: A large transformer job on GPU fails when sequence length doubles and memory allocation spikes. The corresponding corrective action is: Reduce microbatch size or apply supported gradient accumulation and memory-saving options.
E:The evidence for this case shows situation: Inspect GPU utilization, gradient synchronization cost and data loading per worker. It directly examines the reported situation: An image classifier fits on one GPU but takes days; several GPU devices and a sharded dataset are available. The corresponding corrective action is: Train with a supported distributed data-parallel strategy and synchronized optimizer updates.
Incorrect Answers
B:Inspecting supported base-model adapters and memory requirements for fine-tuning is relevant when a team needs a domain-specific output style from a large foundation model but lacks GPU memory for full weight updates; it does not establish the result required by this workstream.
D:Reviewing the entrypoint, dependencies, runtime image and output path contract is relevant when a research team needs a custom PyTorch loss and optimizer but wants repeatable training jobs rather than manual notebook sessions; it does not establish the result required by this workstream.
F:Comparing image digests, framework versions, nondeterministic operations and input order is relevant when a custom model changes predictions after a new container image is pulled despite the same source dataset and random seed; it does not establish the result required by this workstream.
Question 343
A platform team needs diagnostic evidence for this condition: A boosted-tree model needs tuning but the team manually tries one parameter at a time and repeatedly inspects the final test set. Which observation most directly identifies the cause or requirements before a change?
- Inspect supported base-model adapters and memory requirements for fine-tuning.
- Define search ranges, a single validation objective and trial budget.
- Inspect table row counts and network and schema access from the training environment.
- Inspect per-step GPU memory, sequence length and peak batch allocation.
Explanation
Correct Answer
B:The evidence for this case shows situation: Define search ranges, a single validation objective and trial budget. It directly examines the reported situation: A boosted-tree model needs tuning but the team manually tries one parameter at a time and repeatedly inspects the final test set. The corresponding corrective action is: Use bounded managed hyperparameter tuning on training and validation data only.
Incorrect Answers
A:This evidence concerns a separate issue: A team needs a domain-specific output style from a large foundation model but lacks GPU memory for full weight updates. It does not establish what is happening in the present case: A boosted-tree model needs tuning but the team manually tries one parameter at a time and repeatedly inspects the final test set. The diagnostic evidence to examine here is Define search ranges, a single validation objective and trial budget.
C:This evidence concerns a separate issue: A fraud team maintains labeled feature tables in BigQuery but training needs distributed custom PyTorch code. It does not establish what is happening in the present case: A boosted-tree model needs tuning but the team manually tries one parameter at a time and repeatedly inspects the final test set. The diagnostic evidence to examine here is Define search ranges, a single validation objective and trial budget.
D:This evidence concerns a separate issue: A large transformer job on GPU fails when sequence length doubles and memory allocation spikes. It does not establish what is happening in the present case: A boosted-tree model needs tuning but the team manually tries one parameter at a time and repeatedly inspects the final test set. The diagnostic evidence to examine here is Define search ranges, a single validation objective and trial budget.
Question 344
An ML platform team is resolving 3 independent workstreams. Workstream A: A deep sequence model's training error falls every epoch while validation error climbs after a specific checkpoint. Workstream B: A team needs a domain-specific output style from a large foundation model but lacks GPU memory for full weight updates. Workstream C: A TPU training workload shows very low accelerator utilization while CPU parsing threads block on small files. Select exactly THREE diagnostic observations, one directly addressing each separate workstream.
- Inspect supported base-model adapters and memory requirements for fine-tuning.
- Inspect Cloud Storage object generations and exact training manifest membership.
- Review the entrypoint, dependencies, runtime image and output path contract.
- Compare image digests, framework versions, nondeterministic operations and input order.
- Inspect training and validation curves with identical metrics and clean splits.
- Inspect tf.data profiling and accelerator idle time versus input latency.
Explanation
Correct Answers
A:The evidence for this case shows situation: Inspect supported base-model adapters and memory requirements for fine-tuning. It directly examines the reported situation: A team needs a domain-specific output style from a large foundation model but lacks GPU memory for full weight updates. The corresponding corrective action is: Use a compatible LoRA or parameter-efficient tuning method on curated examples.
E:The evidence for this case shows situation: Inspect training and validation curves with identical metrics and clean splits. It directly examines the reported situation: A deep sequence model's training error falls every epoch while validation error climbs after a specific checkpoint. The corresponding corrective action is: Add validation-based early stopping and retain the best validated checkpoint.
F:The evidence for this case shows situation: Inspect tf.data profiling and accelerator idle time versus input latency. It directly examines the reported situation: A TPU training workload shows very low accelerator utilization while CPU parsing threads block on small files. The corresponding corrective action is: Optimize input sharding, parallel decoding, prefetching and efficient file formats.
Incorrect Answers
B:Inspecting Cloud Storage object generations and exact training manifest membership is relevant when a training pipeline uses an S3-style moving folder of CSV files and repeated runs receive different examples each week; it does not establish the result required by this workstream.
C:Reviewing the entrypoint, dependencies, runtime image and output path contract is relevant when a research team needs a custom PyTorch loss and optimizer but wants repeatable training jobs rather than manual notebook sessions; it does not establish the result required by this workstream.
D:Comparing image digests, framework versions, nondeterministic operations and input order is relevant when a custom model changes predictions after a new container image is pulled despite the same source dataset and random seed; it does not establish the result required by this workstream.
Question 345
An incident review is assessing this observed failure: A large transformer job on GPU fails when sequence length doubles and memory allocation spikes. Which observation most directly identifies the cause or requirements before a change?
- Inspect per-step GPU memory, sequence length and peak batch allocation.
- Define search ranges, a single validation objective and trial budget.
- Inspect training and validation curves with identical metrics and clean splits.
- Compare image digests, framework versions, nondeterministic operations and input order.
Explanation
Correct Answer
A:The evidence for this case shows situation: Inspect per-step GPU memory, sequence length and peak batch allocation. It directly examines the reported situation: A large transformer job on GPU fails when sequence length doubles and memory allocation spikes. The corresponding corrective action is: Reduce microbatch size or apply supported gradient accumulation and memory-saving options.
Incorrect Answers
B:This evidence concerns a separate issue: A boosted-tree model needs tuning but the team manually tries one parameter at a time and repeatedly inspects the final test set. It does not establish what is happening in the present case: A large transformer job on GPU fails when sequence length doubles and memory allocation spikes. The diagnostic evidence to examine here is Inspect per-step GPU memory, sequence length and peak batch allocation.
C:This evidence concerns a separate issue: A deep sequence model's training error falls every epoch while validation error climbs after a specific checkpoint. It does not establish what is happening in the present case: A large transformer job on GPU fails when sequence length doubles and memory allocation spikes. The diagnostic evidence to examine here is Inspect per-step GPU memory, sequence length and peak batch allocation.
D:This evidence concerns a separate issue: A custom model changes predictions after a new container image is pulled despite the same source dataset and random seed. It does not establish what is happening in the present case: A large transformer job on GPU fails when sequence length doubles and memory allocation spikes. The diagnostic evidence to examine here is Inspect per-step GPU memory, sequence length and peak batch allocation.
Question 346
An ML platform team is resolving 3 independent workstreams. Workstream A: A supervised model's validation accuracy is suspiciously high because a source field contains post-outcome status and several duplicate entity records cross splits. Workstream B: A multi-hour training run restarts from epoch zero whenever a worker is preempted, losing expensive accelerator time. Workstream C: A training pipeline uses an S3-style moving folder of CSV files and repeated runs receive different examples each week. Select exactly THREE diagnostic observations, one directly addressing each separate workstream.
- Inspect Cloud Storage object generations and exact training manifest membership.
- Define search ranges, a single validation objective and trial budget.
- Compare image digests, framework versions, nondeterministic operations and input order.
- Inspect checkpoint write frequency and reliable object storage destinations.
- Inspect feature timestamps, entity IDs and duplicate hashes around the train-test boundary.
- Review the entrypoint, dependencies, runtime image and output path contract.
Explanation
Correct Answers
A:The evidence for this case shows situation: Inspect Cloud Storage object generations and exact training manifest membership. It directly examines the reported situation: A training pipeline uses an S3-style moving folder of CSV files and repeated runs receive different examples each week. The corresponding corrective action is: Freeze a Cloud Storage input manifest with object generations and use it for each reproducible run.
D:The evidence for this case shows situation: Inspect checkpoint write frequency and reliable object storage destinations. It directly examines the reported situation: A multi-hour training run restarts from epoch zero whenever a worker is preempted, losing expensive accelerator time. The corresponding corrective action is: Write periodic framework checkpoints to durable Cloud Storage and restore on restart.
E:The evidence for this case shows situation: Inspect feature timestamps, entity IDs and duplicate hashes around the train-test boundary. It directly examines the reported situation: A supervised model's validation accuracy is suspiciously high because a source field contains post-outcome status and several duplicate entity records cross splits. The corresponding corrective action is: Remove post-label information and group related entities before splitting.
Incorrect Answers
B:The approach “Define search ranges, a single validation objective and trial budget” is relevant when a boosted-tree model needs tuning but the team manually tries one parameter at a time and repeatedly inspects the final test set; it does not establish the result required by this workstream.
C:Comparing image digests, framework versions, nondeterministic operations and input order is relevant when a custom model changes predictions after a new container image is pulled despite the same source dataset and random seed; it does not establish the result required by this workstream.
F:Reviewing the entrypoint, dependencies, runtime image and output path contract is relevant when a research team needs a custom PyTorch loss and optimizer but wants repeatable training jobs rather than manual notebook sessions; it does not establish the result required by this workstream.
Question 347
Before changing the system, the team investigates the following behavior: A team needs a domain-specific output style from a large foundation model but lacks GPU memory for full weight updates. Which observation most directly identifies the cause or requirements before a change?
- Inspect GPU utilization, gradient synchronization cost and data loading per worker.
- Review the entrypoint, dependencies, runtime image and output path contract.
- Inspect supported base-model adapters and memory requirements for fine-tuning.
- Compare image digests, framework versions, nondeterministic operations and input order.
Explanation
Correct Answer
C:The evidence for this case shows situation: Inspect supported base-model adapters and memory requirements for fine-tuning. It directly examines the reported situation: A team needs a domain-specific output style from a large foundation model but lacks GPU memory for full weight updates. The corresponding corrective action is: Use a compatible LoRA or parameter-efficient tuning method on curated examples.
Incorrect Answers
A:This evidence concerns a separate issue: An image classifier fits on one GPU but takes days; several GPU devices and a sharded dataset are available. It does not establish what is happening in the present case: A team needs a domain-specific output style from a large foundation model but lacks GPU memory for full weight updates. The diagnostic evidence to examine here is Inspect supported base-model adapters and memory requirements for fine-tuning.
B:This evidence concerns a separate issue: A research team needs a custom PyTorch loss and optimizer but wants repeatable training jobs rather than manual notebook sessions. It does not establish what is happening in the present case: A team needs a domain-specific output style from a large foundation model but lacks GPU memory for full weight updates. The diagnostic evidence to examine here is Inspect supported base-model adapters and memory requirements for fine-tuning.
D:This evidence concerns a separate issue: A custom model changes predictions after a new container image is pulled despite the same source dataset and random seed. It does not establish what is happening in the present case: A team needs a domain-specific output style from a large foundation model but lacks GPU memory for full weight updates. The diagnostic evidence to examine here is Inspect supported base-model adapters and memory requirements for fine-tuning.
Question 348
An ML platform team is resolving 3 independent workstreams. Workstream A: A foundation-model team has a small curated dataset but proposes unrestricted full-parameter updates to learn a narrow command format. Workstream B: A custom model changes predictions after a new container image is pulled despite the same source dataset and random seed. Workstream C: A research team needs a custom PyTorch loss and optimizer but wants repeatable training jobs rather than manual notebook sessions. Select exactly THREE diagnostic observations, one directly addressing each separate workstream.
- Compare image digests, framework versions, nondeterministic operations and input order.
- Inspect dataset size, GPU budget and validation risk against adapter tuning.
- Inspect supported base-model adapters and memory requirements for fine-tuning.
- Review the entrypoint, dependencies, runtime image and output path contract.
- Inspect table row counts and network and schema access from the training environment.
- Inspect checkpoint write frequency and reliable object storage destinations.
Explanation
Correct Answers
A:The evidence for this case shows situation: Compare image digests, framework versions, nondeterministic operations and input order. It directly examines the reported situation: A custom model changes predictions after a new container image is pulled despite the same source dataset and random seed. The corresponding corrective action is: Pin container and library versions and record seeds, data and hardware settings.
B:The evidence for this case shows situation: Inspect dataset size, GPU budget and validation risk against adapter tuning. It directly examines the reported situation: A foundation-model team has a small curated dataset but proposes unrestricted full-parameter updates to learn a narrow command format. The corresponding corrective action is: Prefer prompt or parameter-efficient tuning first and justify full fine-tuning only by evidence.
D:The evidence for this case shows situation: Review the entrypoint, dependencies, runtime image and output path contract. It directly examines the reported situation: A research team needs a custom PyTorch loss and optimizer but wants repeatable training jobs rather than manual notebook sessions. The corresponding corrective action is: Package the training program into a supported custom-training container and submit an Agent Platform job.
Incorrect Answers
C:Inspecting supported base-model adapters and memory requirements for fine-tuning is relevant when a team needs a domain-specific output style from a large foundation model but lacks GPU memory for full weight updates; it does not establish the result required by this workstream.
E:Inspecting table row counts and network and schema access from the training environment is relevant when a fraud team maintains labeled feature tables in BigQuery but training needs distributed custom PyTorch code; it does not establish the result required by this workstream.
F:Inspecting checkpoint write frequency and reliable object storage destinations is relevant when a multi-hour training run restarts from epoch zero whenever a worker is preempted, losing expensive accelerator time; it does not establish the result required by this workstream.
Question 349
A platform team needs diagnostic evidence for this condition: A multi-hour training run restarts from epoch zero whenever a worker is preempted, losing expensive accelerator time. Which observation most directly identifies the cause or requirements before a change?
- Compare image digests, framework versions, nondeterministic operations and input order.
- Review the entrypoint, dependencies, runtime image and output path contract.
- Inspect checkpoint write frequency and reliable object storage destinations.
- Inspect GPU utilization, gradient synchronization cost and data loading per worker.
Explanation
Correct Answer
C:The evidence for this case shows situation: Inspect checkpoint write frequency and reliable object storage destinations. It directly examines the reported situation: A multi-hour training run restarts from epoch zero whenever a worker is preempted, losing expensive accelerator time. The corresponding corrective action is: Write periodic framework checkpoints to durable Cloud Storage and restore on restart.
Incorrect Answers
A:This evidence concerns a separate issue: A custom model changes predictions after a new container image is pulled despite the same source dataset and random seed. It does not establish what is happening in the present case: A multi-hour training run restarts from epoch zero whenever a worker is preempted, losing expensive accelerator time. The diagnostic evidence to examine here is Inspect checkpoint write frequency and reliable object storage destinations.
B:This evidence concerns a separate issue: A research team needs a custom PyTorch loss and optimizer but wants repeatable training jobs rather than manual notebook sessions. It does not establish what is happening in the present case: A multi-hour training run restarts from epoch zero whenever a worker is preempted, losing expensive accelerator time. The diagnostic evidence to examine here is Inspect checkpoint write frequency and reliable object storage destinations.
D:This evidence concerns a separate issue: An image classifier fits on one GPU but takes days; several GPU devices and a sharded dataset are available. It does not establish what is happening in the present case: A multi-hour training run restarts from epoch zero whenever a worker is preempted, losing expensive accelerator time. The diagnostic evidence to examine here is Inspect checkpoint write frequency and reliable object storage destinations.
Question 350
An ML platform team is resolving 3 independent workstreams. Workstream A: A TPU training workload shows very low accelerator utilization while CPU parsing threads block on small files. Workstream B: A fraud team maintains labeled feature tables in BigQuery but training needs distributed custom PyTorch code. Workstream C: A boosted-tree model needs tuning but the team manually tries one parameter at a time and repeatedly inspects the final test set. Select exactly THREE diagnostic observations, one directly addressing each separate workstream.
- Define search ranges, a single validation objective and trial budget.
- Inspect tf.data profiling and accelerator idle time versus input latency.
- Review the entrypoint, dependencies, runtime image and output path contract.
- Compare image digests, framework versions, nondeterministic operations and input order.
- Inspect table row counts and network and schema access from the training environment.
- Inspect per-step GPU memory, sequence length and peak batch allocation.
Explanation
Correct Answers
A:The evidence for this case shows situation: Define search ranges, a single validation objective and trial budget. It directly examines the reported situation: A boosted-tree model needs tuning but the team manually tries one parameter at a time and repeatedly inspects the final test set. The corresponding corrective action is: Use bounded managed hyperparameter tuning on training and validation data only.
B:The evidence for this case shows situation: Inspect tf.data profiling and accelerator idle time versus input latency. It directly examines the reported situation: A TPU training workload shows very low accelerator utilization while CPU parsing threads block on small files. The corresponding corrective action is: Optimize input sharding, parallel decoding, prefetching and efficient file formats.
E:The evidence for this case shows situation: Inspect table row counts and network and schema access from the training environment. It directly examines the reported situation: A fraud team maintains labeled feature tables in BigQuery but training needs distributed custom PyTorch code. The corresponding corrective action is: Export or load authorized BigQuery examples through a supported managed training data path.
Incorrect Answers
C:Reviewing the entrypoint, dependencies, runtime image and output path contract is relevant when a research team needs a custom PyTorch loss and optimizer but wants repeatable training jobs rather than manual notebook sessions; it does not establish the result required by this workstream.
D:Comparing image digests, framework versions, nondeterministic operations and input order is relevant when a custom model changes predictions after a new container image is pulled despite the same source dataset and random seed; it does not establish the result required by this workstream.
F:Inspecting per-step GPU memory, sequence length and peak batch allocation is relevant when a large transformer job on GPU fails when sequence length doubles and memory allocation spikes; it does not establish the result required by this workstream.
Professional Machine Learning Engineer Practice Tests