google-research/xtreme

Bug in PawsxProcessor

amineabdaoui opened this issue · 2 comments

Hi @sebastianruder,

There is no need to remove the first line of each file (the header) in the PawsxProcessor because it's already done during the download by the pawsx_preprocess() function.

We are losing the first pair of sentences for Pawsx files with the current code.

Thanks for spotting this and sending the PR. I've approved it.