ymcui / cmrc2018

A Span-Extraction Dataset for Chinese Machine Reading Comprehension (CMRC 2018)
https://ymcui.github.io/cmrc2018/
Creative Commons Attribution Share Alike 4.0 International
411 stars 87 forks source link

请问squad_style数据集中的answer_idx计算,有没有考虑context里存在多个与answer字符串相同的情况?是否是人工检查过确实是答案的那个substring? #23

Open demi6od opened 1 year ago