Hello,
Thank you for sharing the version of the two datasets that you used.
However, I have a question about the CoINCo dataset.
Whenever the target word is "." or "," it seems that you have updated the target sentence where you have duplicated the previous word
for example:
..N 14925 18 until thirty seconds ago , i didn n't believe in magic or any of that kind of ...weirdness . "
you have updated to
..N 14925 18 until thirty seconds ago , i didn n't believe in magic or any of that kind of ...weirdness weirdness "
could you let me know why you have done this update?
Hello,
Thank you for sharing the version of the two datasets that you used.
However, I have a question about the CoINCo dataset.
Whenever the target word is "." or "," it seems that you have updated the target sentence where you have duplicated the previous word
for example:
..N 14925 18 until thirty seconds ago , i didn n't believe in magic or any of that kind of ...weirdness . "
you have updated to
..N 14925 18 until thirty seconds ago , i didn n't believe in magic or any of that kind of ...weirdness weirdness "
could you let me know why you have done this update?