AWS Certified AI Practitioner AIF-C01 — Question 96

A company has fine-tuned a large language model (LLM) to answer questions for a help desk. The company wants to determine if the fine-tuning has enhanced the model's accuracy. Which metric should the company use for the evaluation?