This data is a sample (n=200) of the full CDC Wonderstats dataset. The questions have been reworded using the following chatGPT prompt:
"Reword the below questions using more casual and succinct language."
The answer portions of each message in the data are identical to the full dataset.
This data can be used to evaluate a model fine tuned with the CDC Wonderstats dataset using BERTscore.
