1Z0-1127-25 Deploying and Managing Generative AI on OCI Practice Test 8 — 15 Questions

Question 1

A developer wants to invoke an OCI Generative AI model from an application running on a compute instance in OCI. The instance is in a private subnet. What is the most secure method to access the model endpoint?

Choose an answer to begin — your selection is scored in the full session.

Accepted Answer

Use a Service Gateway to access the endpoint privately.. A Service Gateway allows resources in a private subnet to access OCI services, including the Generative AI model endpoint, over the OCI private network without traversing the internet. This is the most secure method because traffic stays within the OCI backbone, avoiding exposure to public IPs and reducing the attack surface.

Answer

Use an Internet Gateway and public endpoint.

Answer

Use a VPN Connect to connect to the model's public IP.

Answer

Use a NAT Gateway to access the endpoint.

Question 2

A company is deploying a chatbot powered by OCI Generative AI. They want to inject the conversation history into the model prompt to maintain context. However, they notice that after a long conversation, the model starts to ignore earlier messages. What is the most likely cause?

Accepted Answer

The model has a limited context window size.. The model's context window size limits the total number of tokens (input + output) it can process at once. When the conversation history grows beyond this limit, older messages are truncated or dropped, causing the model to lose context from earlier parts of the conversation. This is a fundamental constraint of transformer-based models like those used in OCI Generative AI.

Answer

The model's max_tokens limit is too low, truncating the prompt.

Answer

The top_p parameter is set to 1, causing deterministic output.

Answer

The temperature setting is too high, causing randomness.

Question 3

An organization needs to ensure that all inference requests to OCI Generative AI are logged for compliance. Which OCI feature should be enabled?

Accepted Answer

OCI Logging for the AI service. OCI Logging for the AI service captures detailed request and response data for inference calls to OCI Generative AI, including payloads, timestamps, and user identities. This feature must be explicitly enabled per service endpoint to meet compliance requirements for logging all inference requests. Unlike Audit logs, which record control-plane operations, OCI Logging provides data-plane logging for the AI service itself.

Answer

OCI Cloud Guard

Answer

OCI Vault

Answer

OCI Audit logs

Question 4

A company deploys a fine-tuned model on an OCI Generative AI dedicated AI cluster. After deployment, they observe high latency during peak hours. The cluster has only one replica. Which action would most effectively reduce latency without increasing cost unnecessarily?

Accepted Answer

Enable auto-scaling with a maximum of 3 replicas.. Enabling auto-scaling with a maximum of 3 replicas (Option B) is the most effective action because it dynamically adds replicas during peak hours to handle increased load, reducing latency, while limiting the maximum to 3 prevents unnecessary cost overruns. This balances performance and cost, unlike a fixed large replica count or switching models, which either wastes resources or fails to address the root cause of insufficient compute capacity.

Answer

Increase the number of replicas to 10.

Answer

Switch to a larger base model.

Answer

Move to a serverless deployment model.

Question 5

An OCI administrator is configuring access control for OCI Generative AI. Which three IAM components are required to allow a group of data scientists to call the GenerateText API? (Choose three.)

Accepted Answer

An IAM group for the data scientists. Option A is correct because IAM authorization in OCI is always granted to groups, not directly to individual users, so the data scientists must be placed in an IAM group that the policy can reference. Option C is correct because a policy statement must grant the verb (for example manage or use) on the ai-services-generative-ai-family resource in the relevant compartment, which is what authorizes calls to the GenerateText API. Option E is correct because OCI policies are scoped to a compartment, so the Generative AI resources must reside in (or be covered by) a compartment for the policy to apply. Option B is incorrect because a local peering gateway is a VCN networking component for connecting VCNs and has nothing to do with IAM API authorization. Option D is incorrect because dynamic groups are used to authorize OCI compute instances and other resources as principals, not human data scientists who authenticate as users.

Answer

A local peering gateway

Answer

A dynamic group

Question 6

A developer is troubleshooting an OCI Generative AI inference request that returns a 400 Bad Request error. Which three common causes could result in this error? (Choose three.)

Accepted Answer

Incorrect endpoint URL. Option A is correct because a 400 Bad Request is returned when the request is sent to an incorrect or malformed endpoint URL, such as a wrong region-specific hostname or a mistyped path, so the service cannot route or parse the request. Option C is correct because OCI Generative AI requires specific fields in the JSON body (for example modelId, compartmentId, and the chat or generateText payload); omitting required parameters causes the service to reject the request with 400. Option D is correct because exceeding the model's maximum token limit (input plus output tokens) violates the request constraints and produces a 400 Bad Request rather than a server-side 5xx error. Option B is not correct because an invalid API key typically yields 401 Unauthorized (or 403 Forbidden), not 400 Bad Request. Option E is not correct because network connectivity issues generally manifest as timeouts or connection errors, not an HTTP 400 response from the service.

Answer

Invalid API key in the request header

Answer

Network connectivity issues

Question 7

A company deploys a fine-tuned Llama 2 model using OCI Generative AI service. They want to ensure low-latency inference for a real-time chat application. Which deployment option should they use?

Accepted Answer

Dedicated AI cluster. A dedicated AI cluster provides reserved compute resources (GPUs) for low-latency, real-time inference by eliminating resource contention. This is essential for a fine-tuned Llama 2 model in a chat application where consistent sub-second response times are required, unlike shared or serverless options that introduce cold starts or queuing delays.

Answer

Batch inference job

Answer

OCI Functions

Answer

Serverless endpoint (standard)

Question 8

A team deploys a generative AI model endpoint and notices intermittent 429 Too Many Requests errors. The endpoint is configured with auto-scaling using a dedicated AI cluster. What is the most likely cause?

Accepted Answer

The auto-scaling policy is not aggressive enough. The 429 Too Many Requests error indicates that the endpoint is receiving more requests than it can handle. With auto-scaling enabled on a dedicated AI cluster, the most likely cause is that the auto-scaling policy is not aggressive enough to keep up with the request rate, meaning it scales up too slowly or has insufficient maximum instance limits to handle the traffic spike.

Answer

The model's context window exceeded

Answer

Insufficient storage on the cluster

Answer

Rate limiting at the OCI API Gateway

Question 9

A team uses OCI Generative AI’s fine-tuning capability to adapt a base model. After fine-tuning, they evaluate the model but see degraded performance on certain edge cases. What is the most likely cause?

Accepted Answer

Overfitting on the training data. Fine-tuning adapts a base model to a specific dataset, but if the training data is too narrow or the model is trained for too many epochs, it can memorize the training examples rather than learning generalizable patterns. This overfitting causes the model to perform well on training-like inputs but poorly on edge cases that deviate from the training distribution. In OCI Generative AI, overfitting is a common pitfall when fine-tuning hyperparameters like the number of epochs or learning rate are not properly validated.

Answer

Validation data leakage

Answer

Learning rate too high

Answer

Insufficient training epochs

Question 10

An administrator notices that a dedicated AI cluster is not scaling down after a period of low traffic. What could be the cause?

Accepted Answer

The cluster has a minimum size set to the current number of nodes. A dedicated AI cluster in OCI has a minimum size configuration that prevents the autoscaler from reducing the node count below that threshold. If the current number of nodes equals the configured minimum, the cluster will not scale down even during low traffic, as the autoscaler respects this lower bound. This ensures baseline capacity is always available for inference workloads.

Answer

There are pending inference requests

Answer

The cluster is in a compartment without permissions

Answer

The autoscaling policy uses a cooldown period that is too short

Question 11

Which TWO actions should be taken to monitor model drift in a deployed generative AI model? (Select TWO)

Accepted Answer

Compare inference statistics over time. Option A (Compare inference statistics over time) is correct because monitoring drift requires tracking changes in the statistical distribution of inputs and outputs (e.g., feature distributions, prompt/response patterns) across time windows, which is the core mechanism for detecting data or concept drift in a deployed generative AI model. Option D (Set up alerts on accuracy metrics) is correct because drift is only actionable when it is surfaced through automated alerting on relevant quality/accuracy metrics (e.g., evaluation scores, hallucination or groundedness rates), enabling timely remediation before degradation impacts users. Option B (Retrain the model weekly) is not a monitoring action; retraining is a remediation step and a fixed weekly cadence does not detect drift. Option C (Use OCI Data Labeling for new data) supports data preparation and evaluation but is not itself a drift monitoring mechanism. Option E (Deploy multiple model versions) is a deployment strategy (e.g., canary or A/B) and does not by itself monitor drift.

Answer

Retrain the model weekly

Answer

Use OCI Data Labeling for new data

Answer

Deploy multiple model versions

Question 12

A company has deployed a fine-tuned GPT model on OCI Generative AI using a dedicated AI cluster with 2 nodes. The endpoint is used by an internal application that generates product descriptions. Recently, the application started receiving timeouts and slow responses. The monitoring dashboard shows that the cluster's CPU utilization is consistently above 90%, and the request queue is growing. The team has verified that the model and code have not changed. The application traffic has increased by 20% over the past month. What should the team do to resolve the issue?

Accepted Answer

Increase the number of nodes in the dedicated AI cluster from 2 to 4.. The dedicated AI cluster with 2 nodes is experiencing sustained CPU utilization above 90% and a growing request queue due to a 20% increase in traffic. Scaling out the cluster by adding more nodes (from 2 to 4) increases the available compute capacity, allowing the cluster to handle the higher inference load without timeouts. This directly addresses the resource bottleneck without requiring code or model changes.

Answer

Switch to a serverless endpoint to handle variable traffic.

Answer

Reduce the batch size in the inference requests to lower CPU usage.

Answer

Implement a caching layer for frequently requested descriptions.

Question 13

A multinational corporation uses OCI Generative AI to power a customer support chatbot. The chatbot uses a fine-tuned model deployed on a dedicated AI cluster in the us-ashburn-1 region. The application is used globally, and users in Europe are experiencing high latency (over 2 seconds) compared to users in North America (under 500 ms). The company has a requirement to keep all data within the US due to compliance, so they cannot deploy in Europe. The latency is not due to network bandwidth but due to the inference time. The monitoring shows that the cluster is at 80% utilization during peak hours. The team wants to reduce the latency for European users without violating data residency. What is the best course of action?

Accepted Answer

Optimize the model using techniques like quantization or pruning to reduce inference time.. The latency issue is explicitly due to inference time, not network bandwidth or cluster utilization. Model optimization techniques like quantization (reducing precision of weights from FP32 to INT8) and pruning (removing redundant neurons) directly reduce the computational cost per inference, thereby lowering the response time without moving data or changing the deployment region. This approach satisfies the data residency constraint while addressing the root cause of high latency for European users.

Answer

Implement an edge caching layer in Europe to serve common queries.

Answer

Increase the number of nodes in the cluster to distribute the load.

Answer

Deploy an additional endpoint in a European region and use a global load balancer.

Question 14

Your team has deployed a fine-tuned GPT-2 model on OCI Model Deployment for a simple text generation API. The model performs text completion for short prompts (e.g., 50 tokens). The endpoint is working but response times are over 10 seconds for these short prompts. The model size is approximately 500MB and you used a VM.Standard.E3.Flex shape (2 OCPU, 16GB RAM). The deployment is in a single replica with no autoscaling. You have verified that the network latency is minimal (<5ms). The model was trained in OCI Data Science using a GPU shape, but during deployment you selected a CPU shape to reduce cost. The model is a transformer-based neural network. You've also confirmed that the deployment is healthy and there are no errors in the logs. The memory usage is within limits. What is the most likely cause of the high latency?

Accepted Answer

Missing GPU acceleration for inference. GPT-2 is a transformer-based neural network that relies heavily on matrix multiplications, which are far more efficiently executed on GPUs due to their parallel architecture. Even though the model is only 500MB, CPU inference for transformer models is notoriously slow because CPUs process sequential operations, while GPUs can parallelize the attention mechanism and feed-forward layers. The 10-second latency for a 50-token prompt is a classic symptom of missing GPU acceleration, as the CPU shape (2 OCPU) lacks the specialized tensor cores needed for fast transformer inference.

Answer

High network latency between the client and the model endpoint

Answer

Model is too large for the VM.Standard.E3.Flex shape

Answer

Insufficient CPU resources for the model size

Question 15

A company is deploying a generative AI model for a real-time inference API. To ensure high availability and cost efficiency under variable load, which two configurations should they implement? (Choose two.)

Accepted Answer

Set the number of model deployment replicas to at least 2. Option D is correct because setting the number of model deployment replicas to at least 2 eliminates a single point of failure and allows the inference API to keep serving traffic if one replica or its underlying host fails, which directly satisfies the high availability requirement. Option E is correct because enabling autoscaling based on average CPU utilization lets the deployment add replicas when inference load rises and remove them when load falls, providing cost efficiency under variable traffic while maintaining performance. Option A is incorrect because a single replica, even with a larger GPU, remains a single point of failure and cannot scale horizontally, so it fails the high availability goal. Option B is incorrect because confining the deployment to a single availability domain means an AD-level outage takes down the entire inference API. Option C is incorrect because disabling connection draining causes in-flight inference requests to be dropped abruptly during scaling or maintenance events, degrading availability and reliability.

Answer

Use a single replica with a larger GPU to handle all traffic

Answer

Deploy the model in a single availability domain to simplify management

Answer

Disable connection draining on the load balancer