1Z0-1127-25 OCI Generative AI Service Practice Test 2 — 15 Questions

Question 1

A data scientist is using the OCI Generative AI Playground to test a summarization model. They want to generate shorter summaries and avoid repetitive phrasing. Which parameter adjustments should they make?

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

Accepted Answer

Decrease temperature and increase frequency penalty. Lowering temperature reduces randomness, increasing frequency penalty penalizes repeated tokens, and setting appropriate max tokens controls output length. Stop sequences define when generation stops, and presence penalty penalizes token presence overall.

Answer

Increase temperature and increase presence penalty

Answer

Set stop sequences and disable frequency penalty

Answer

Increase temperature and decrease frequency penalty

Question 2

A developer is fine-tuning a Cohere Command R model using OCI Data Science and the T-Few technique. They have prepared a dataset. What is the required format for the training data?

Accepted Answer

A JSONL file where each line contains a 'prompt' and a 'completion' field. OCI Data Science fine-tuning with the T-Few technique expects training data in JSONL format, where each line is a self-contained JSON object containing a 'prompt' and a 'completion' field. This structure lets the T-Few parameter-efficient fine-tuning pipeline pair each instruction/input with its expected output during few-shot adaptation. CSV, Parquet, and plain-text conversation formats are not parsed by the OCI fine-tuning job for Cohere Command R.

Answer

A CSV file with columns 'input' and 'output'

Answer

A Parquet file with 'text' and 'label' columns

Answer

A plain text file with one conversation per line

Question 3

A security administrator needs to grant a group of data scientists the ability to use OCI Generative AI service resources in the compartment 'genai-dev'. They want to allow the group to create endpoints and run inference, but not to manage IAM policies. Which policy statement is correct?

Accepted Answer

Allow group DataScientists to use generative-ai-family in compartment genai-dev. The 'use' verb in OCI IAM policies grants the ability to perform operational actions on resources without allowing management or administrative changes. For the OCI Generative AI service, 'use generative-ai-family' permits data scientists to create endpoints and run inference (the core actions needed), while explicitly preventing them from managing IAM policies or altering resource configurations. This aligns with the principle of least privilege required by the security administrator.

Answer

Allow group DataScientists to manage generative-ai-family in compartment genai-dev

Answer

Allow group DataScientists to inspect generative-ai-family in compartment genai-dev

Answer

Allow group DataScientists to read generative-ai-family in compartment genai-dev

Question 4

An organization wants to use OCI Generative AI for a high-volume summarization workload. They estimate 10 million tokens per month and need consistent low latency. Which pricing model is most cost-effective?

Accepted Answer

Provision a Dedicated AI Cluster with model units. On-demand pricing can be expensive at high volumes. Dedicated clusters offer predictable cost per model unit and low-latency dedicated inference, making them more cost-effective for high-volume, latency-sensitive workloads. Pay-as-you-go (on-demand) is suitable for low or variable usage.

Answer

Use the free tier

Answer

Use OCI Data Science notebook sessions

Answer

On-demand token-based pricing

Question 5

A company wants to implement a retrieval-augmented generation (RAG) chatbot using OCI Generative AI Agents. Which TWO services or components are required for this solution?

Accepted Answer

Knowledge Base. In OCI Generative AI Agents, a Knowledge Base (option C) is a required component because it stores and indexes the ingested documents that the agent retrieves from to ground its responses in your enterprise data. OCI Object Storage (option D) is also required because it serves as the data source where the source documents (for example PDFs, text, or HTML files) are stored before being ingested into the Knowledge Base. A Cohere Embed model (option A) is not a separate required selection here, since the embedding model is configured automatically as part of the Knowledge Base ingestion process rather than being a distinct service the customer must provision. A fine-tuned model (option B) is unnecessary because RAG relies on retrieval plus a foundation model, not on fine-tuning. A Dedicated AI Cluster (option E) is not required because Generative AI Agents and the underlying models can run on on-demand infrastructure without a dedicated cluster.

Answer

Cohere Embed model

Answer

Fine-tuned model

Answer

Dedicated AI Cluster

Question 6

A developer is building a multi-turn chatbot using the OCI Generative AI Chat API. Which THREE parameters or features should they configure to maintain coherent conversation history?

Accepted Answer

System prompt. The system prompt (C) is correct because it establishes the chatbot's persona, role, and behavioral instructions that persist across all turns, anchoring the model's responses so the conversation stays coherent and on-topic. Multi-turn conversation history (D) is correct because the OCI Generative AI Chat API is stateless, so the developer must resend prior user and assistant messages in each request to give the model the context needed to resolve references and maintain continuity. Preamble override (E) is correct because it lets the developer replace the model's default preamble text, which is where system-level instructions and conversational framing are injected, thereby shaping how the model interprets and continues the dialogue. Temperature (A) only controls randomness/creativity of token sampling and does not preserve or convey conversation context. Stop sequences (B) merely tell the model when to halt generation at specified strings and have no role in retaining or supplying dialogue history.

Answer

Temperature

Answer

Stop sequences

Question 7

A data scientist is using the OCI Generative AI Embeddings API to generate vectors for a classification task. Which TWO input types are appropriate for this use case?

Accepted Answer

clustering. Option B (clustering) is correct because the OCI Generative AI Embeddings API accepts an input type of clustering, which produces embeddings optimized for grouping similar texts, exactly matching the data scientist's classification task. Option D (classification) is correct because the API explicitly supports a classification input type designed to generate embeddings tuned for training and running classification models. The unmarked options do not belong: search_query and search_document are input types intended for semantic search retrieval scenarios (query vs. document asymmetry), not classification, and text is not a valid input type value for the OCI Generative AI Embeddings API.

Answer

search_query

Answer

text

Answer

search_document

Question 8

A data scientist is fine-tuning a model using T-Few in OCI Generative AI. They have prepared a dataset with prompt/completion pairs. Which file format is required for the training data upload?

Accepted Answer

JSONL with each line containing 'prompt' and 'completion' keys. OCI Generative AI fine-tuning expects training data in JSONL format with specific fields.

Answer

Plain text file with one prompt-completion pair per line separated by a tab

Answer

Parquet file with a 'text' column containing concatenated prompt and completion

Answer

CSV with columns 'input' and 'output'

Question 9

An administrator wants to grant a group of data scientists permission to use OCI Generative AI resources in a specific compartment, but prevent them from creating Dedicated AI Clusters. Which IAM policy statement achieves this?

Accepted Answer

Allow group data-scientists to use generative-ai-family in compartment genai-dev where request.operation != 'CreateDedicatedAiCluster'. It uses the 'use' verb to grant the data scientists access to OCI Generative AI resources while adding a condition with 'request.operation != 'CreateDedicatedAiCluster'' to explicitly deny the ability to create Dedicated AI Clusters. In OCI IAM, the 'use' verb includes read and update capabilities but not create or delete, and the condition further restricts the specific create operation, aligning with the requirement to prevent cluster creation.

Answer

Allow group data-scientists to read generative-ai-family in compartment genai-dev

Answer

Allow group data-scientists to manage generative-ai-family in compartment genai-dev

Answer

Allow group data-scientists to use generative-ai-models in compartment genai-dev

Question 10

During fine-tuning a model using T-Few in OCI Generative AI, the job fails with a 'dataset format error'. The training dataset is a JSONL file. Which of the following is the MOST likely cause?

Accepted Answer

Some lines have an extra 'metadata' key. The most likely cause is that some lines have an extra 'metadata' key because OCI Generative AI fine-tuning with T-Few expects a specific JSONL format with only 'prompt' and 'completion' keys (or similar). Extra keys like 'metadata' are not allowed and cause a dataset format error. The other options are either not format errors or are within acceptable limits.

Answer

The file contains more than 1000 lines

Answer

The completion field contains more than 1024 tokens

Answer

The file uses Unix line endings

Question 11

An organization wants to use OCI Generative AI for summarizing long legal documents. Which OCI Generative AI service component is specifically designed for this task?

Accepted Answer

Summarisation. The Summarisation API in OCI Generative AI is a dedicated endpoint optimized for condensing long texts into concise summaries. It uses specialized model configurations and prompt engineering to handle the context window and extraction requirements of legal documents, unlike general-purpose generation endpoints.

Answer

Chat API

Answer

Embedding API

Answer

Generate API

Question 12

An administrator is creating IAM policies for OCI Generative AI. They want to allow a group of developers to use (invoke) models and manage endpoints, but NOT create or delete Dedicated AI Clusters. Which TWO policy statements should be combined?

Accepted Answer

Allow group developers to use generative-ai-family in compartment dev. Option C is correct because 'use generative-ai-family' grants the ability to invoke/use the models (the generative-ai-family resource covers model inference actions) without granting cluster management rights. Option E is correct because 'manage generative-ai-endpoints' grants full control over endpoints, which is exactly what the developers need to create, update, and delete endpoints. Option A is incorrect because 'use generative-ai-clusters' would grant cluster usage permissions, which the scenario explicitly wants to avoid. Option B is incorrect because 'manage generative-ai-family' would grant full management of the model family, exceeding the intended invoke-only access. Option D is incorrect because 'manage generative-ai-clusters' would allow creating and deleting Dedicated AI Clusters, which the scenario explicitly prohibits.

Answer

Allow group developers to use generative-ai-clusters in compartment dev

Answer

Allow group developers to manage generative-ai-family in compartment dev

Answer

Allow group developers to manage generative-ai-clusters in compartment dev

Question 13

A company wants to build a customer service chatbot that answers questions about their internal policy documents. The documents are updated monthly, and the team cannot afford to retrain a model each time. Which approach is MOST appropriate?

Accepted Answer

Use Retrieval-Augmented Generation (RAG) with the policy documents indexed in a vector store. Retrieval-Augmented Generation (RAG) is the most appropriate approach because it allows the chatbot to answer questions based on the latest policy documents without retraining the model. By indexing the documents in a vector store, the system retrieves relevant chunks at query time and passes them as context to a frozen LLM, ensuring the model always uses the most current information.

Answer

Use a larger foundation model with a longer context window and paste all documents into each prompt

Answer

Train a custom model from scratch on the policy documents each month

Answer

Fine-tune a base LLM on the policy documents monthly

Question 14

A developer is using the OCI GenAI Playground to test a summarisation model. They want the summary to be concise and less creative. Which combination of parameter adjustments would best achieve this?

Accepted Answer

Decrease temperature to 0.1 and increase frequency penalty. Decreasing temperature to 0.1 makes the model more deterministic and less creative, while increasing the frequency penalty discourages repetition, both of which help produce a concise, less creative summary. This combination directly aligns with the goal of reducing randomness and controlling output length.

Answer

Set presence penalty to high and temperature to 0.5

Answer

Increase temperature to 0.9 and decrease frequency penalty to 0

Answer

Increase max tokens and set stop sequences

Question 15

A team wants to use OCI Generative AI Agents to build a question-answering system over documents stored in OCI Object Storage. They have created a knowledge base and are ready to test. Which API should they use to interact with the agent for multi-turn conversations?

Accepted Answer

Sessions API. The Sessions API is the correct choice because OCI Generative AI Agents uses sessions to maintain conversation state and context across multiple turns. When you create a session, the service persists the dialogue history so the agent can reference prior messages, which is essential for coherent multi-turn question-answering. The Chat API is used within a session to send individual prompts and receive responses, but the session itself is the container that enables multi-turn continuity.

Answer

Chat API

Answer

Embedding API

Answer

Generate API