1Z0-1127-25 Practice Test 32 — 15 Questions

Question 1

A developer is building a RAG pipeline with LangChain. They have loaded PDF documents with PDFLoader. Which TWO steps must they perform before indexing the documents into a vector store?

Accepted Answer

Split the documents into chunks using a text splitter. Option B is correct because PDFLoader returns whole Document objects, and a text splitter (e.g., RecursiveCharacterTextSplitter) must break them into smaller chunks so each chunk fits the embedding model's token limit and retrieval returns focused context. Option D is correct because a vector store indexes vectors, not raw text, so each chunk must be passed through an embedding model (e.g., OpenAIEmbeddings) to produce the numeric vectors that are stored and later compared via similarity search. Option A is not required, since RAG pipelines work with documents in any language the embedding model supports. Option C is unnecessary because encryption is a security concern, not a prerequisite for indexing. Option E is wrong because gzip compression is irrelevant to vector indexing and would not produce embeddings.

Answer

Translate the documents into English

Answer

Encrypt the documents with AES-256

Answer

Compress the documents with gzip

Question 2

An application mixes RAG with other data sources. The vector search returns too many irrelevant chunks. What is the best approach to filter them?

Accepted Answer

Use a reranker model. A reranker model (Option A) is the best approach because it takes the initial set of retrieved chunks and re-orders them based on semantic relevance to the query, effectively filtering out irrelevant chunks. Unlike simple vector similarity, a reranker uses cross-encoding to evaluate the query-chunk pair as a whole, which significantly improves precision when mixing RAG with other data sources.

Answer

Use exact search instead of ANN

Answer

Reduce the number of retrieved chunks

Answer

Increase chunk size

Question 3

A developer is using LangChain with Oracle AI Vector Search (OracleVS) to store embeddings. They notice that similarity search queries are slow. Which THREE actions could improve query performance?

Accepted Answer

Use a smaller embedding model (e.g., 384 dimensions instead of 1536). Option C is correct because using a smaller embedding model (e.g., 384 dimensions instead of 1536) reduces the vector dimensionality, which directly shrinks the distance-computation cost and memory footprint of each similarity search, making queries faster. Option D is correct because creating an HNSW vector index on the VECTOR column enables Oracle AI Vector Search to perform approximate nearest-neighbor (ANN) search instead of a full exact scan, which is the primary way to accelerate similarity queries in OracleVS. Option A is not correct because reducing chunk_size only changes how documents are split into smaller pieces; it increases the number of stored vectors and does not inherently speed up similarity search. Option B is not correct because switching to IVF with a low number of centroids reduces index granularity and can degrade recall and performance rather than reliably improving query speed. Option E is not correct because increasing efSearch raises the HNSW search breadth, which improves recall at the cost of more distance computations and therefore makes queries slower, not faster.

Answer

Reduce the chunk_size parameter in the text splitter

Answer

Switch from HNSW to IVF with a low number of centroids

Answer

Increase the efSearch parameter in the HNSW index

Question 4

A company wants to process a large batch of documents to generate summaries using OCI Generative AI. They need the most cost-effective option without compromising on summary quality. Which approach should they use?

Accepted Answer

Use the on-demand (shared) inference endpoint with the Summarisation API. On-demand pricing is pay-per-token, suitable for batch workloads without dedicated capacity; it's cost-effective for sporadic or large batch jobs.

Answer

Use OCI Generative AI Agents with a knowledge base

Answer

Provision a dedicated AI cluster for the batch job

Answer

Fine-tune a model on summarization and deploy it on a dedicated cluster

Question 5

Refer to the exhibit. An administrator receives the error shown when attempting to deploy a custom model. What is the most likely cause?

Accepted Answer

The user or service does not have permission to read the model artifact from Object Storage. The error indicates that the deployment process cannot access the model artifact stored in Object Storage. In OCI Generative AI, the service must have read permission on the bucket and object to download the artifact. If the user or service principal lacks the necessary IAM policy (e.g., `allow service generative-ai to read objects in compartment X where target.bucket.name='Y'`), the deployment fails with this access-denied error.

Answer

The compartment ID is incorrect

Answer

The model artifact file is corrupted

Answer

The dedicated AI cluster ID is invalid

Question 6

A team wants to evaluate an LLM's performance on a text classification task. Which metric is most appropriate for a balanced dataset?

Accepted Answer

Accuracy. Accuracy is the most appropriate metric for evaluating an LLM on a text classification task with a balanced dataset because it directly measures the proportion of correctly predicted labels out of total predictions. For balanced classes, accuracy provides a reliable and intuitive performance indicator without the distortion caused by class imbalance.

Answer

BLEU score

Answer

Perplexity

Answer

ROUGE score

Question 7

Which of the following best describes the role of a Retriever in a LangChain RAG pipeline?

Accepted Answer

It fetches relevant document chunks from a vector store based on a query. In a LangChain RAG pipeline, the Retriever's job is to accept a natural-language query, embed it, and perform a similarity search against a vector store to return the most relevant document chunks. It sits between the vector store and the LLM, acting as the 'fetch' stage of retrieve-augmented generation. It does not transform or ingest documents — it only retrieves already-indexed content.

Answer

It splits documents into smaller chunks

Answer

It loads documents from various sources like PDF or HTML

Answer

It encodes documents into dense vector representations

Question 8

A data scientist is designing a RAG system using OCI Data Science and OCI Generative AI. Which two considerations are critical for optimal retrieval quality? (Choose 2.)

Accepted Answer

Use the same embedding model for both indexing and querying.. Option A is correct because retrieval quality in a RAG pipeline depends on the query embedding and the document embeddings living in the same vector space; using the same embedding model (e.g., the same OCI Generative AI embedding model and version) for both indexing and querying ensures cosine similarity scores are meaningful and comparable. Option D is correct because applying metadata filtering (for example, on attributes such as document type, date, or department stored alongside vectors in a vector store) narrows the candidate set before the vector similarity search, which reduces irrelevant matches and improves precision, especially in large corpora. Option B is not correct because maximizing chunk size can dilute the embedding with unrelated content and exceed the model's context window, hurting retrieval precision rather than helping it. Option C is not correct because fine-tuning the generation model affects answer synthesis, not retrieval quality, and is not a critical retrieval consideration. Option E is not correct because a hierarchical index structure primarily addresses search speed and scalability, not the relevance or quality of retrieved results.

Answer

Increase the chunk size to the maximum allowed to capture more context.

Answer

Fine-tune the generation model on domain-specific data.

Answer

Use a hierarchical index structure for faster search.

Question 9

During a fine-tuning job for a text generation model, the loss curve shows that the training loss decreases steadily, but the evaluation loss increases after a few epochs. Which action is most likely to improve the model's generalization?

Accepted Answer

Implement early stopping based on evaluation loss. The described pattern — training loss decreasing while evaluation loss increases — is classic overfitting, where the model memorizes training data and loses generalization. Early stopping based on evaluation loss halts training at the point of best validation performance, preventing further overfitting. This is the most direct and standard remedy for the observed divergence.

Answer

Increase the number of training epochs

Answer

Reduce the size of the training dataset

Answer

Increase the temperature parameter during generation

Question 10

A data scientist wants to quickly test a prompt with different parameters like temperature and max tokens without writing code. Which OCI GenAI feature should they use?

Accepted Answer

OCI Generative AI Playground.. The OCI Generative AI Playground is a web-based, no-code interface that allows data scientists to interactively test prompts and adjust parameters like temperature and max tokens without writing any code. This directly matches the user's requirement for quick, code-free experimentation.

Answer

OCI CLI.

Answer

OCI SDK.

Answer

OCI Data Science Notebooks.

Question 11

A developer is using a prompt template that includes placeholders like {context} and {question}. They want to version these templates for A/B testing. Which practice is BEST for managing prompt templates?

Accepted Answer

Use a dedicated prompt library with versioning, such as a database table with version numbers. A dedicated prompt library with versioning — such as a database table storing template text, version numbers, and metadata — is the best practice because it decouples prompt content from application code, supports A/B testing by version, and provides an auditable history of changes. This lets teams roll back, compare, and promote prompt versions without redeploying the application. It is the most scalable and operationally sound approach for managing prompt templates.

Answer

Save each template as a separate Python file in a Git repository

Answer

Store templates only in the application code as string constants

Answer

Use a spreadsheet to track template versions

Question 12

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?

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 13

Which of the following is a key limitation of large language models that RAG (Retrieval-Augmented Generation) aims to address?

Accepted Answer

Hallucinations (factual errors). RAG addresses hallucinations by grounding the model's output in retrieved documents that contain factual information. Knowledge cutoff, bias, and context length constraints are separate issues that RAG may partially help with, but its primary purpose is to reduce factual errors.

Answer

Context length constraints

Answer

Bias in training data

Answer

Knowledge cutoff date

Question 14

Which OCI Generative AI model would you use to reorder search results to improve relevance ranking?

Accepted Answer

Cohere Rerank. Cohere Rerank is specifically designed to take a list of candidate documents (e.g., from a vector search) and reorder them by relevance to the query using a cross-encoder model. It is the OCI Generative AI service's dedicated reranking model, improving precision over embedding-only retrieval. The other models generate text or embeddings, not relevance rankings.

Answer

Cohere Command R+

Answer

Cohere embed-english-v3.0

Answer

Meta Llama 3 70B

Question 15

A company has multiple teams sharing an OCI Generative AI Dedicated AI Cluster. They need to ensure that each team can only access their own fine-tuned models and cannot see or invoke models from other teams. What is the best approach?

Accepted Answer

Use OCI compartments and IAM policies with resource-level permissions for models. OCI compartments and IAM policies with resource-level permissions allow you to grant granular access to specific models within a Dedicated AI Cluster. By placing each team's fine-tuned models in separate compartments and writing policies that restrict access to those compartments, you ensure teams can only see and invoke their own models. This approach leverages OCI's native identity and access management without requiring separate clusters or network-level isolation.

Answer

Train separate models for each team

Answer

Encrypt model artifacts with different keys for each team

Answer

Use network security lists to isolate traffic