MarkWuNLP/ResponseEdit

have questions about the input format

DaisyZHL opened this issue · 1 comments

According to what you said in the README, the input format of your code is "prototype context \t ins words \t del words", but after reading projects/train.src, shouldn't it be "prototype response \t ins words \t del words"?

Yes. I write a script to parse the input format to "prototype response \t ins words \t del words" but did not upload to this repo.