---
license: mit
---

Dataset paired from here: 
https://github.com/Yale-LILY/SummEval

It's smaller than I thought. Perhaps one can squeeze out a few hundred comparisons for an llm.