1Z0-1127-25 OCI Generative AI Service Practice Test 7 — 15 Questions

Question 1

Which statement accurately describes the T-Few fine-tuning technique used in OCI Generative AI?

Accepted Answer

It is a parameter-efficient fine-tuning method that updates only a fraction of the model parameters.. The T-Few fine-tuning technique is a parameter-efficient fine-tuning (PEFT) method that updates only a small fraction of the model's parameters, typically by introducing and training adapter layers or using low-rank updates. This approach significantly reduces computational and memory requirements compared to full fine-tuning, making it suitable for adapting large language models with limited resources. In OCI Generative AI, T-Few enables efficient customization without retraining the entire model.

Answer

It automatically adjusts hyperparameters during inference.

Answer

It does not require any training data and works by prompting only.

Answer

It updates all model parameters, requiring substantial compute resources.

Question 2

An OCI Generative AI user wants to host a fine-tuned model with guaranteed low latency for a production application. Which option should they choose?

Accepted Answer

Provision a Dedicated AI Cluster with model units for the fine-tuned model. A Dedicated AI Cluster with model units is the only OCI Generative AI hosting option that provides guaranteed, isolated compute capacity for a fine-tuned model, ensuring consistent low latency and throughput for production workloads. Model units let you size the cluster to your performance requirements, and the fine-tuned model is hosted on that dedicated infrastructure rather than sharing capacity with other tenants.

Answer

Use the Chat API with a base model

Answer

Use the on-demand shared inference endpoint

Answer

Use the OCI Generative AI Playground

Question 3

A company needs to generate vector embeddings for a multilingual document set to support semantic search across English and French documents. Which embedding model should they use?

Accepted Answer

Cohere embed-multilingual-v3.0. Cohere embed-multilingual-v3.0 is specifically trained to produce embeddings that align semantically across more than 100 languages, including English and French, so semantically equivalent sentences in different languages map to nearby vectors. This enables cross-lingual semantic search where an English query can retrieve relevant French documents and vice versa. The other options are either English-only embedding models or generative LLMs not designed for embedding generation.

Answer

Cohere embed-english-v3.0

Answer

Cohere Command R

Answer

Meta Llama 3 8B

Question 4

A developer is using the OCI Generative AI Chat API to build a conversational assistant. They want the assistant to adopt a formal tone regardless of user input. Which parameter should they set in the API request?

Accepted Answer

Set the system prompt to 'You are a formal assistant that responds in a professional tone.'. The system prompt (or preamble) sets the assistant's behavior consistently. Temperature affects randomness but not tone directly.

Answer

Set a high temperature (e.g., 0.9)

Answer

Set the frequency penalty to a high value

Answer

Set max tokens to a low value

Question 5

A user wants to quickly test different prompts and parameters (temperature, max tokens) with various OCI Generative AI models without writing any code. Which tool should they use?

Accepted Answer

The OCI Generative AI Playground. The OCI Generative AI Playground is a no-code web interface that lets users experiment with different models, prompts, and parameters such as temperature and max tokens. It is designed for quick testing and comparison without writing any code. The SDK, CLI, and Cloud Shell all require coding or command-line interaction.

Answer

The InferenceClient SDK in Python

Answer

OCI CLI with generative-ai commands

Answer

OCI Cloud Shell

Question 6

A company wants to use OCI Generative AI Agents to build a RAG application over documents stored in OCI Object Storage. What must they create first?

Accepted Answer

A knowledge base linked to the Object Storage bucket. OCI Generative AI Agents require a knowledge base to index data sources before creating an agent.

Answer

A Dedicated AI Cluster

Answer

An embedding endpoint

Answer

A fine-tuning job for the base model

Question 7

A developer needs to generate embeddings for a set of search queries to be used in a semantic search system. Which input type should they specify when calling the Embedding API?

Accepted Answer

search_query. The Embedding API in OCI Generative AI requires specifying the input type as 'search_query' when generating embeddings for search queries in a semantic search system. This input type optimizes the embedding model to produce vectors that are specifically tuned for query-side representation, ensuring better alignment with document embeddings during similarity search.

Answer

clustering

Answer

search_document

Answer

classification

Question 8

Which OCI Generative AI model is designed to rerank and improve the relevance of documents retrieved by a search system?

Accepted Answer

Cohere Rerank. Cohere Rerank is specifically designed to take a set of retrieved documents and reorder them by relevance to a query, improving the quality of search and RAG results. It uses a cross-encoder approach to score query-document pairs more accurately than embedding similarity alone. The other models serve different purposes: embeddings, text generation, or general chat.

Answer

Cohere Embed

Answer

Cohere Command R

Answer

Meta Llama 3

Question 9

A team fine-tuned a Cohere Command R model in OCI GenAI and validated it. They now need to deploy it for production inference with a dedicated endpoint. What is the correct sequence of steps?

Accepted Answer

Provision a Dedicated AI Cluster, deploy the model to the cluster, then create an endpoint. In OCI Generative AI, the correct sequence for deploying a fine-tuned model to a dedicated endpoint is: first provision a Dedicated AI Cluster (which provides the isolated compute infrastructure), then deploy the model to that cluster, and finally create an endpoint that exposes the deployed model for inference. This ensures the model is hosted on dedicated resources before the endpoint is created.

Answer

Create an endpoint, deploy the model, then provision a Dedicated AI Cluster

Answer

Deploy the model to a shared endpoint, then provision a Dedicated AI Cluster for scaling

Answer

Create an endpoint, then provision a Dedicated AI Cluster, then deploy the model

Question 10

A data scientist is configuring a fine-tuning job in OCI Generative AI. Which TWO of the following are required inputs for creating the job?

Accepted Answer

Training dataset (JSONL file). Option B is correct because a fine-tuning job in OCI Generative AI requires a training dataset, which must be supplied as a JSONL file stored in OCI Object Storage, containing the prompt-completion examples used to adapt the model. Option D is correct because the job must specify a supported base model (for example, a Cohere Command R model) that will be fine-tuned, since fine-tuning always starts from an existing pretrained model rather than from scratch. Option A is incorrect because temperature is a runtime inference parameter that controls randomness during text generation, not an input to the fine-tuning training process. Option C is incorrect because max tokens is an inference-time generation limit, and validation configuration in fine-tuning uses a validation dataset rather than a max tokens setting. Option E is incorrect because an inference endpoint is created after fine-tuning to serve the resulting custom model, and is not an input required when creating the fine-tuning job itself.

Answer

Temperature setting for fine-tuning

Answer

Max tokens limit for validation

Answer

Inference endpoint name

Question 11

A company wants to use OCI Generative AI Agents to create a RAG-powered customer support system. Which THREE components are essential for the agent to work?

Accepted Answer

A knowledge base. Option A (a knowledge base) is essential because OCI Generative AI Agents uses the knowledge base to store and retrieve the vectorized chunks of enterprise data that ground the RAG responses. Option C (a data source such as an OCI Object Storage bucket) is required because the agent ingests the source documents from that data source to build and populate the knowledge base. Option D (a base LLM such as Cohere Command R) is necessary because the agent relies on an underlying foundation model to reason over the retrieved context and generate the final natural-language answer. Option B is incorrect because a Dedicated AI Cluster is only needed for hosting custom fine-tuned or dedicated model deployments, not for a standard RAG agent using an on-demand base model. Option E is incorrect because OCI Generative AI Agents automatically handles embedding generation with managed embedding models, so a customer-supplied fine-tuned embedding model is not a required component.

Answer

A Dedicated AI Cluster

Answer

A fine-tuned embedding model

Question 12

Which OCI Generative AI service component is specifically designed to convert text into vector representations for use in semantic search and clustering?

Accepted Answer

Embedding. Embedding is the correct answer because it is the OCI Generative AI service component that converts text into dense vector representations (embeddings). These numerical vectors capture semantic meaning, enabling similarity comparisons for tasks like semantic search and clustering, where documents with similar vectors are considered contextually related.

Answer

Chat

Answer

Summarisation

Answer

Agents

Question 13

A data scientist needs to fine-tune a model using OCI Generative AI. They have prepared a dataset in JSONL format with prompt/completion pairs. The fine-tuning job is configured with the T-Few technique. What is a key characteristic of T-Few fine-tuning?

Accepted Answer

It modifies a small subset of parameters using lightweight adapter layers. T-Few is a parameter-efficient fine-tuning method that updates only a small fraction of model parameters, making it faster and more resource-efficient than full fine-tuning. It does not train all parameters, add new layers, or require unlabeled data.

Answer

It requires unlabeled data for unsupervised pre-training before fine-tuning

Answer

It updates all model parameters, requiring significant compute resources

Answer

It only trains a new classification head on top of a frozen base model

Question 14

An organization is deploying a custom fine-tuned model for a real-time fraud detection application. The model must respond within 200ms and cannot share infrastructure with other customers. Which OCI GenAI infrastructure option should they choose?

Accepted Answer

A dedicated AI cluster with provisioned model units. A dedicated AI cluster with provisioned model units ensures exclusive infrastructure, meeting the requirement to not share resources with other customers. This option also provides predictable, low-latency inference (under 200ms) by allocating dedicated compute capacity for the fine-tuned model, which is critical for real-time fraud detection.

Answer

On-demand token-based inference

Answer

Shared infrastructure with reserved capacity

Answer

OCI Functions using the GenAI SDK

Question 15

Which of the following is NOT an available model in OCI Generative AI service?

Accepted Answer

OpenAI GPT-4. OCI Generative AI service offers a curated catalog of models from partners such as Cohere (Command, Embed, Rerank) and Meta (Llama 2/3), plus Oracle's own models. OpenAI GPT-4 is not part of the OCI Generative AI model catalog; OpenAI models are accessed through Azure OpenAI or OpenAI's own API, not OCI. Therefore GPT-4 is the model that is NOT available in OCI Generative AI.

Answer

Cohere Embed English v3.0

Answer

Meta Llama 3

Answer

Cohere Command R+