AI-102 Question 475
Select 2You are designing an Azure OpenAI Service-based chatbot to assist customers with product inquiries. The chatbot frequently generates overly verbose responses that overwhelm users. Which parameter(s) should you configure to reduce the verbosity of the responses while maintaining relevance?
- A
Reduce the 'max_tokens' parameter
- B
Increase the 'temperature' parameter
- C
Reduce the 'temperature' parameter
- D
Adjust the 'top_p' parameter to a lower value
- E
Increase the 'frequency_penalty' parameter
Show answer and explanation
Correct answers: A, D
Explanation
To control verbosity, reducing 'max_tokens' limits the response length, ensuring it stays concise. Lowering 'top_p' ensures the model focuses on the most likely responses, indirectly contributing to shorter outputs. These parameters work together to create relevant yet concise responses for the chatbot.
- A. Correct.
Reducing the 'max_tokens' parameter limits the total length of the response, directly addressing the issue of verbosity.
- B. Incorrect.
Increasing the 'temperature' parameter makes the model's responses more creative and diverse, which does not help in reducing verbosity.
- C. Incorrect.
Reducing the 'temperature' parameter makes the responses more focused and deterministic, but it does not directly control verbosity.
- D. Correct.
Adjusting the 'top_p' parameter to a lower value reduces the range of possible responses by focusing on the most likely outcomes, which can indirectly help in controlling verbosity.
- E. Incorrect.
Increasing the 'frequency_penalty' parameter reduces repetitive usage of the same words but does not directly address verbosity or length.