AWS Certified AI Practitioner AIF-C01 — Question 106

A company is introducing a mobile app that helps users learn foreign languages. The app makes text more coherent by calling a large language model (LLM). The company collected a diverse dataset of text and supplemented the dataset with examples of more readable versions. The company wants the LLM output to resemble the provided examples. Which metric should the company use to assess whether the LLM meets these requirements?