Dear Authors,
Thank you for this great work!
I have a question regarding the Wikipedia corpus used for PopQA in the paper.
Did you use the one you provide here? Which seems to have only the passages corresponding to intro sections (~6M passages).
Or did you use the full corpus provided in this Atlas corpora?
Thank you!
Dear Authors,
Thank you for this great work!
I have a question regarding the Wikipedia corpus used for PopQA in the paper.
Did you use the one you provide here? Which seems to have only the passages corresponding to intro sections (~6M passages).
Or did you use the full corpus provided in this Atlas corpora?
Thank you!