En Ouyang

为什么baseline.md里multitask的performance基本上都要比single task更差

4

如题，谢谢

手机上相册界面不显示

13

**I certify that I have first consulted** (check all with "x") **我确认我已经查看了** (使用 "x" 选择) - [x] [Hexo documentation | Hexo 文档](https://hexo.io/docs/) - [x] [Material theme documentation | Material 主题文档](https://material.viosey.com/)...

which performance is better when use crf loss or softmax loss?

12

Hello, thanks for your job. I have a question about different loss function, is there any difference for performance when using different loss function, like crf_loss and softmax_loss? thanks.

pre-trained embedding not used in input layer?

10

Hi, I noticed that the pre-trained embedding file was not used in "embedding layer", just used a lookup function to generation character embedding and seg embedding. The pre-trained embedding only...

about corpus format

4

I have a question about the format of corpus, I noticed the text8 corpus in data folder write in just one line, I want to know there is no influence...