An ML engineer needs to develop an AI assistant by using Retrieval Augmented Generation (RAG) in Amazon Bedrock. The company stores a collection of PDF text files in an Amazon S3 bucket. The ML…

Question

Options

• AUse the latest version of Amazon Titan Text Premier to perform document embedding.
• BUse the latest version of Mistral 7B Instruct to perform instruction tuning.
• CUse the latest version of Anthropic Claude Sonnet to perform instruction tuning.
• DUse Cohere Embed Multilingual to perform document embedding.

• A
  3% (1)
• B
  3% (1)
• C
  10% (3)
• D
  83% (25)

Explanation

In a Retrieval Augmented Generation (RAG) pipeline, PDF documents must be converted into numerical vector representations (embeddings) before they can be stored in a vector store and searched semantically. This requires a dedicated embedding model, not a generative LLM. Cohere Embed Multilingual is an embedding model - its sole purpose is to transform text into dense vectors. Options B and C (Mistral 7B Instruct and Anthropic Claude Sonnet) are instruction-following generative LLMs, not embedding models; 'instruction tuning' is a fine-tuning technique, not a data ingestion step. Option A (Amazon Titan Text Premier) is also a text generation model, not an embedding model. Only Cohere Embed produces the vector representations required by the knowledge base's vector store.

No community discussion yet for this question.