Andy Wong

神他妈年薪百万

> > 你好，因科研需要，我们向你们的contact邮箱发送了全量数据的申请，但是一直未得到回复。这个数据是免费开放的吗，还是说全量数据需要额外付费？ > > 请问你最后拿到了数据集么？ 没有回复

Hi~, [CopyR](https://github.com/xiangrongzeng/copy_re) uses the version only annotating the last word, do you also follow this preprocessing setting? Or do you preprocessing on the original dataset released by [CopyR](https://github.com/xiangrongzeng/copy_re) and annotating...

@hjing100 -1取的是最后一个hidden_size

看我这个工作[Discontinuous Named Entity Recognition as Maximal Clique Discovery](https://github.com/131250208/InfExtraction)

@hemmatan Hi, thanks for your interest. We just followed the previous SoTA and used the same datasets. For all I know, they are not labeled manually and no manually labeled...

这种情况就直接在出错那行代码打断点查

正常，函数里定义的是整条向量所有输出标签全部预测正确，你可以自己修改

1. It brought a slight improvement. For more details, see [Extend "softmax+cross entropy" to multi-label classification problem](https://spaces.ac.cn/archives/7359) 2. For all I know, no. But you can refer to this blog...

用作negative samples