BenchMark Quizzes by MDFT Pro

Choose Training Data Format For Azure OpenAI Fine-Tuning

MDFT Pro, a well-known training agency, wants to create a specialized AI model for generating personalized course recommendations and study guidance tailored to their specific educational methodology and course catalog. Claire, the AI Model Developer, has collected 500 prompt-completion pairs from successful student interactions, course advisor conversations, and effective study recommendations that represent the ideal responses MDFT Pro wants their AI system to provide.

These training examples include prompts like “What certification should I pursue for cloud computing?” paired with completions that recommend specific MDFT Pro courses, prerequisites, and career pathways. Claire needs to prepare this training data in the correct format for Azure OpenAI fine-tuning to create a custom model that understands MDFT Pro’s educational approach.

Which format should Claire use for the training data file containing the 500 prompt-completion pairs?

Choose the correct answer from the options below.

Explanations for each answer:

XML is incorrect. XML format is not supported for Azure OpenAI fine-tuning training data, which requires a more streamlined format for prompt-completion pairs.
CSV is incorrect. CSV format is not the standard format for Azure OpenAI fine-tuning, which needs a format that can handle complex text structures and formatting.
JSONL is correct. JSONL (JSON Lines) format is the required format for Azure OpenAI fine-tuning, with each line containing a JSON object representing a prompt-completion pair.
TSV is incorrect. TSV (Tab-Separated Values) format is not suitable for Azure OpenAI fine-tuning, which requires JSON-based formatting for training data.

Learn more about Azure OpenAI fine-tuning:

Azure OpenAI Fine-Tuning

Exam Preparation Quiz

Choose Training Data Format For Azure OpenAI Fine-Tuning