Closed brightmart closed 4 years ago
Thanks for your support.
There are several overlaps between our experiments and ChineseGLUE, such as THUCNEWS (We use a subset of the original dataset, DEV=97.66, TEST=97.64), LCQMC (DEV=90.20, TEST=87.95), XNLI (DEV=80.48, TEST=79.20), MSRANER (TEST=95.25). Please check the details in our paper (https://arxiv.org/pdf/1911.00720.pdf).
We will conduct experiments on other datasets in ChineseGLUE in the future.
Thanks very much!
Great! Thank for your feedback.
Great job!
What's the performance of ZEN on ChineseGLUE benchmark?
It has a wide range of tasks, now around 10 tasks with baselines available.
https://github.com/chineseGLUE/chineseGLUE