# Datasets

I will be using a combination of

-   CoQA (unmodified)
-   Natural Questions (modified to make context from answer)