AI-102 Question 491
Single answerYou are tasked with fine-tuning an Azure OpenAI GPT model for a customer support chatbot. The chatbot must provide tailored responses based on your company’s product catalog. Which step is required to ensure the fine-tuned model performs effectively?
- A
Prepare a custom dataset formatted in JSONL with labeled examples relevant to the company’s product catalog.
- B
Directly upload the pre-trained GPT model to Azure OpenAI Service without any modifications.
- C
Use the Azure OpenAI Playground to train the model without providing any sample data.
- D
Set the temperature parameter to 0 to ensure deterministic responses during fine-tuning.
Show answer and explanation
Correct answer: A
Explanation
Fine-tuning an Azure OpenAI model requires preparing a high-quality dataset in JSONL format with labeled examples. This dataset enables the model to learn task-specific behaviors, such as responding accurately based on a product catalog. Other options either misunderstand the fine-tuning process or refer to unrelated features of the Azure OpenAI Service.
- A. Correct.
This is correct. Fine-tuning requires a well-prepared dataset in JSONL format with labeled examples, as this provides the necessary context for the model to learn specific patterns related to the task.
- B. Incorrect.
This is incorrect. You cannot upload a pre-trained model directly; fine-tuning requires using the Azure OpenAI Service with a properly formatted dataset.
- C. Incorrect.
This is incorrect. The Azure OpenAI Playground is primarily used for experimentation and testing, not for fine-tuning a model.
- D. Incorrect.
This is incorrect. The temperature parameter controls the randomness of the outputs during inference, not during the fine-tuning process.