Yuxian Meng

Hi，感谢CLUE团队对中文NLP社区做的贡献~ 我下载了你们提供的预训练模型，发现Masked LM的预测似乎比较随机，我猜测原因是MLM model中的一些参数没有被保存下来？不知道你们是否方便将这部分参数也release出来呢？

A suggestion on stable training

2

Hi, I am the author of [pytorch implementation of Matrix Capsules.](https://github.com/shzygmyx/Matrix-Capsules-pytorch) I find that you are facing the problem of unstable training, which did not happen in my project. I...

Hi, thanks for open-sourcing the inspiring knn-LM model! I'm trying to reproduce your results on Book Corpus dataset, but I found there is no standard train/valid/test splits. Could you please...