Generative AI Leader Google Cloud's Generative AI Offerings Practice Test 2 — 15 Questions

Question 1

During a load test, a Vertex AI endpoint serving a large language model experiences high latency and increased error rates. The endpoint is configured with autoscaling. What is the most likely cause?

Accepted Answer

The autoscaling metric is based on CPU utilization but the model is GPU-bound. When a model is GPU-bound, CPU utilization remains low even under heavy inference load, so autoscaling based on CPU metrics fails to trigger additional replicas. This leads to queued requests, increased latency, and eventual error rates as the existing GPU instances become saturated. Vertex AI endpoints default to CPU-based autoscaling unless explicitly configured with GPU metrics like 'gpu_utilization' or custom metrics.

Answer

There is a network bottleneck

Answer

The model size is too large for the machine type

Answer

The endpoint is using a global load balancer

Question 2

A data scientist is using Vertex AI Model Registry to manage multiple versions of a custom text classification model. They need to ensure that only the version that passes all evaluation metrics can be deployed to a Vertex AI Endpoint for online predictions. What deployment strategy should they use?

Accepted Answer

Set up continuous evaluation with model monitoring to auto-promote versions that meet thresholds. Vertex AI Model Registry supports continuous evaluation pipelines that automatically promote a model version to a 'default' or 'deployed' status only when it meets predefined evaluation thresholds (e.g., accuracy, F1 score). This ensures that only the version passing all metrics is eligible for deployment to a Vertex AI Endpoint, enabling automated, gated rollouts without manual intervention.

Answer

Use A/B testing between versions and manually select the best performer

Answer

Manually track version IDs and deploy the latest version

Answer

Deploy all versions to a single endpoint and route traffic manually

Question 3

Which TWO actions can help reduce latency for an online prediction endpoint served by a large language model on Vertex AI? (Select TWO.)

Accepted Answer

Enable response caching for frequent similar queries. Option A is correct because enabling response caching lets the endpoint return previously computed answers for frequent similar queries immediately, bypassing the model's forward-pass computation and thereby cutting inference latency. Option D is correct because a smaller distilled version of the model has fewer parameters and requires less compute per token, so it generates predictions faster than the large teacher model while retaining much of its quality. Option B is not correct because raising the max input token limit only allows longer prompts; it does not reduce latency and can actually increase it by processing more tokens. Option C is not correct because automatic scaling adds replicas to handle higher request volume or throughput, not to lower the per-request latency of a single prediction. Option E is not correct because preemptible (Spot) instances can be reclaimed at any time, which hurts availability and does not improve inference latency.

Answer

Increase the max input token limit to capture more context

Answer

Use automatic scaling to add more replicas

Answer

Set the model to preemptible instances

Question 4

Which THREE capabilities does Vertex AI Agent Builder provide out of the box? (Select THREE.)

Accepted Answer

Automatic escalation to human agents when confidence is low. Vertex AI Agent Builder provides out-of-the-box capabilities including automatic escalation to human agents when confidence is low (A), which is a built-in feature of its conversational agents to hand off to live representatives. It also supports grounding with enterprise data sources such as Cloud Storage and BigQuery (D), allowing agents to retrieve and ground responses in an organization's own data. Additionally, it offers pre-built customer service and sales agents (E), which are ready-to-use templates for common business scenarios. Fine-tuning of foundation models using custom datasets (B) is a capability of Vertex AI's model tuning services, not a specific out-of-the-box feature of Agent Builder. Enterprise search over internal documents (C) is provided by Vertex AI Search (formerly Enterprise Search), which is a related but separate product from Agent Builder.

Answer

Fine-tuning of foundation models using custom datasets

Answer

Enterprise search over internal documents

Question 5

A machine learning engineer submits the above batch prediction job for a large language model. The job is expected to process 100,000 instances. The job takes much longer than expected. Which change would most likely reduce the execution time?

Accepted Answer

Increase maxReplicaCount to 10. Increasing maxReplicaCount to 10 allows Vertex AI Batch Prediction to scale out to more worker replicas, processing the 100,000 instances in parallel. The default maxReplicaCount is often 1 or a low number, which forces sequential or limited parallel processing, causing long execution times. By raising this limit, the job can leverage horizontal scaling to distribute the workload across multiple machines, significantly reducing wall-clock time.

Answer

Increase startingReplicaCount to 10 without changing maxReplicaCount

Answer

Increase the machine type to n1-standard-16

Answer

Decrease the batch size to 1

Question 6

A company wants to build a chatbot that can answer questions about its internal knowledge base using natural language. Which Google Cloud Generative AI offering should they use to quickly prototype and deploy this chatbot with minimal coding?

Accepted Answer

Generative AI Studio. Generative AI Studio is the correct choice because it provides a no-code/low-code interface for prototyping and deploying generative AI applications, including chatbots. It allows users to experiment with prompts, tune models, and deploy them with minimal coding, directly addressing the need for quick prototyping and deployment. The other options are either infrastructure components or pre-built APIs that do not offer the same ease of use for building a custom chatbot.

Answer

Vertex AI Endpoints

Answer

Cloud Natural Language API

Answer

Vertex AI Model Garden

Question 7

A team wants to fine-tune a PaLM 2 model with their own data on Vertex AI. What is the recommended way to prepare the training data?

Accepted Answer

JSON Lines file with 'input_text' and 'output_text' keys. The recommended format for fine-tuning PaLM 2 on Vertex AI is a JSON Lines file with 'input_text' and 'output_text' keys. This structure directly maps to the model's expected supervised fine-tuning (SFT) input, where each line represents a single training example with a prompt and its corresponding response. Vertex AI's fine-tuning pipeline is optimized to parse this schema, ensuring correct tokenization and alignment during training.

Answer

TFRecord files

Answer

CSV file with prompt and completion columns

Answer

Pickle serialized objects

Question 8

A data scientist is comparing two fine-tuned models on Vertex AI Model Evaluation. They want to choose the model with better factual accuracy for a medical Q&A task. Which evaluation metric should they prioritize?

Accepted Answer

exact_match. Exact Match (EM) is the correct metric because it measures whether the model's output exactly matches the ground truth answer, which is critical for factual accuracy in medical Q&A where even minor deviations (e.g., 'aspirin' vs. 'acetylsalicylic acid') could indicate incorrect or incomplete knowledge. Vertex AI Model Evaluation supports EM as a binary metric that penalizes any variation, making it ideal for high-stakes domains requiring precise factual recall.

Answer

pairwise_rouge

Answer

ROUGE-L

Answer

BLEU

Question 9

An organization is building a search application using Vertex AI Vector Search. They have encoded their documents into embeddings and want to retrieve the most similar documents for a query. Which TWO actions are required to set up a Vector Search index?

Accepted Answer

Deploy the index to the IndexEndpoint.. Option B is correct because a Vector Search index must be deployed to an IndexEndpoint before it can serve online queries; deployment associates the index with the endpoint and creates the resources needed for low-latency similarity search. Option E is correct because the IndexEndpoint resource is the serving container that hosts one or more deployed indexes, and it must exist before the index can be deployed to it. Option A is not required as a separate setup action here because the embedding dimension is specified when creating the index, but the question asks for the actions needed to set up the index for serving, and dimension is part of the index creation parameters rather than a distinct required action in this list. Option C is not required because the organization has already encoded documents into embeddings, so they do not need to train a custom embedding model. Option D is not required because query embeddings are generated and sent at query time through the matching API, not downloaded to local storage as part of index setup.

Answer

Specify embedding dimension size in the index config.

Answer

Train a custom embedding model.

Answer

Download the query embeddings to local storage.

Question 10

A company is using Vertex AI Generative AI Studio to iterate on a prompt template. They want to save and organize multiple versions of prompts. Which TWO features should they use?

Accepted Answer

Version history. Version history (B) is correct because Generative AI Studio automatically tracks changes made to a prompt, letting the team review, compare, and restore earlier iterations of the same prompt template. Prompt library (C) is correct because it lets users save prompts as reusable, named, organized assets that can be shared and managed across projects, which directly supports saving and organizing multiple prompt versions. Model Garden (A) is a catalog for discovering and deploying foundation models, not for storing prompt iterations. Parameter sliders (D) only adjust generation settings such as temperature, top-p, and token limits at runtime, and the Run button (E) merely executes a prompt, so neither provides versioning or organization of prompts.

Answer

Model Garden

Answer

Parameter sliders

Answer

Run button

Question 11

The exhibit shows a command to deploy a model to a Vertex AI endpoint with GPU. The deployment fails due to a resource constraint. What is the most likely reason?

Accepted Answer

The machine type n1-standard-4 does not support GPU accelerators.. The n1-standard-4 machine type does not support attaching GPUs. In Vertex AI, GPU accelerators require specific machine series (e.g., n1-highmem-* or n1-highcpu-* for NVIDIA Tesla GPUs, or newer machine families like a2-highgpu-* for A100 GPUs). The n1-standard-4 is a general-purpose machine type that lacks the necessary PCIe lanes and power delivery to accommodate a GPU accelerator, causing a resource constraint failure during deployment.

Answer

The --model flag points to an autoML model.

Answer

The accelerator type is misspelled.

Answer

The min-replica-count is greater than the max-replica-count.

Question 12

A financial institution wants to deploy a custom fine-tuned model for loan approval recommendations. They must ensure compliance with regulatory requirements, including explainability and bias monitoring. Which combination of Google Cloud services and practices best addresses these needs?

Accepted Answer

Fine-tune a foundation model using a custom training pipeline, then deploy with Vertex AI Model Monitoring and Vertex AI Explainable AI. It combines custom fine-tuning of a foundation model (allowing domain-specific adaptation for loan approval logic) with Vertex AI Model Monitoring (for bias detection and drift monitoring) and Vertex AI Explainable AI (for feature attribution and regulatory explainability). This combination directly addresses the dual regulatory requirements of explainability and bias monitoring while leveraging generative AI capabilities.

Answer

Use Vertex AI Search with grounding on internal policies and enable AutoML for model training

Answer

Deploy a pre-built model from Model Garden and use Vertex AI Model Registry

Answer

Use Vertex AI AutoML for tabular data to train the model and enable Vertex AI Model Monitoring for bias

Question 13

A large enterprise is migrating their on-premise ML workloads to Vertex AI. They have a custom PyTorch model for text classification that they want to serve with minimal code changes. Which Vertex AI capability should they use for model serving?

Accepted Answer

Vertex AI Prediction with a custom container. Vertex AI Prediction with a custom container allows the enterprise to package their existing PyTorch model with any custom dependencies or runtime configurations into a Docker container, enabling deployment with minimal code changes. This approach provides full control over the serving environment while leveraging Vertex AI's managed infrastructure for scaling, monitoring, and endpoint management.

Answer

Vertex AI Endpoints with a pre-built PyTorch runtime

Answer

Vertex AI Model Garden

Answer

Vertex AI Vector Search for approximate nearest neighbor

Question 14

A global e-commerce company uses Vertex AI Gemini API for real-time product description generation. They observe that sometimes the model generates text in a language other than the user's language, despite being prompted in English. They need to ensure output language consistency. Which approach is most effective?

Accepted Answer

Run a language detection model on the output and regenerate if not English. The most effective approach because Vertex AI Gemini API does not have a built-in language parameter or a safety filter that can block non-English text. Safety filters are limited to predefined harmful content categories. Therefore, implementing a language detection model to verify the output and regenerating if it is not in English is a practical, albeit post-processing, solution. This ensures consistency without relying on unavailable API features.

Answer

Set the language parameter in the generation config to 'en'

Answer

Fine-tune the model on a dataset of English-only product descriptions

Answer

Configure a safety filter that blocks non-English text

Question 15

Which TWO factors are most important when choosing a base foundation model for fine-tuning on a domain-specific task?

Accepted Answer

Model size and architecture. Option A (Model size and architecture) is correct because the parameter count and architecture (e.g., transformer variant, context length, dense vs. mixture-of-experts) directly determine the model's capacity to absorb domain-specific patterns and the compute/memory budget required for fine-tuning. Option C (Relevance of the model's training data to the target domain) is correct because a base model pretrained on data close to the target domain already encodes relevant vocabulary, terminology, and semantics, so fine-tuning converges faster and yields higher accuracy than adapting an unrelated model. Option B is not a primary factor: community popularity may ease access to tooling or examples, but it does not determine how well the model fits the domain task. Option D is not a primary factor: licensing affects legal and commercial usability, not the technical suitability for fine-tuning. Option E is not a primary factor: inference latency matters for deployment, but fine-tuning quality depends on the base model's capacity and domain alignment, not its serving speed.

Answer

Model popularity in the developer community

Answer

Model license (open-source vs. proprietary)

Answer

Inference latency of the base model