The F1-Score decrease after the last update. about r-bert HOT 10 OPEN

This is a good job. Thank you.
And
I remember the f1-score > 88 some days ago, but the f1-score < 88 I ran last night.
Because the entities fully-connected layer use the same weight after the update?

Could you share how to adjust parameters such as the learning rate to get more results?
Because I want to use R-BERT in my dataset, but the result is not very well.
Thanks.

你好，请问你有复现到89.25吗 我现在最高调试到88.39，你的最高能到多少呢

from r-bert.

This is a good job. Thank you.
And
I remember the f1-score > 88 some days ago, but the f1-score < 88 I ran last night.
Because the entities fully-connected layer use the same weight after the update?
Could you share how to adjust parameters such as the learning rate to get more results?
Because I want to use R-BERT in my dataset, but the result is not very well.
Thanks.

你好，请问你有复现到89.25吗 我现在最高调试到88.39，你的最高能到多少呢

你好，我没有对这个进行调参，我们可以交流+Q 276431860

from r-bert.

This is a good job. Thank you.
And
I remember the f1-score > 88 some days ago, but the f1-score < 88 I ran last night.
Because the entities fully-connected layer use the same weight after the update?

Could you share how to adjust parameters such as the learning rate to get more results?
Because I want to use R-BERT in my dataset, but the result is not very well.
Thanks.

请问你是用中文训练的嘛

from r-bert.

This is a good job. Thank you.
And
I remember the f1-score > 88 some days ago, but the f1-score < 88 I ran last night.
Because the entities fully-connected layer use the same weight after the update?
Could you share how to adjust parameters such as the learning rate to get more results?
Because I want to use R-BERT in my dataset, but the result is not very well.
Thanks.

请问你是用中文训练的嘛

不是，只是跑了一下这个代码，当然用的是Sem2010数据集的。

from r-bert.

This is a good job. Thank you.
And
I remember the f1-score > 88 some days ago, but the f1-score < 88 I ran last night.
Because the entities fully-connected layer use the same weight after the update?
Could you share how to adjust parameters such as the learning rate to get more results?
Because I want to use R-BERT in my dataset, but the result is not very well.
Thanks.

请问你是用中文训练的嘛

不是，只是跑了一下这个代码，当然用的是Sem2010数据集的。

嘿嘿。请问下有没有类似的中文数据集啊？谢谢

from r-bert.

This is a good job. Thank you.
And
I remember the f1-score > 88 some days ago, but the f1-score < 88 I ran last night.
Because the entities fully-connected layer use the same weight after the update?
Could you share how to adjust parameters such as the learning rate to get more results?
Because I want to use R-BERT in my dataset, but the result is not very well.
Thanks.

请问你是用中文训练的嘛

不是，只是跑了一下这个代码，当然用的是Sem2010数据集的。

嘿嘿。请问下有没有类似的中文数据集啊？谢谢

我之前整理过中医数据集方面的资料，不过没有细看，你可以看一下。
http://cips-chip.org.cn/2020/eval2
https://github.com/GanjinZero/awesome_Chinese_medical_NLP
https://kns.cnki.net/kcms/detail/detail.aspx?dbcode=CMFD&dbname=CMFD202001&filename=1019668145.nh&v=PoGN%25mmd2FoPjZYSA%25mmd2Bk7VryERaP9ChrEabWA0giTnbdc%25mmd2FfyZZ5hDJj1Jw9gwEErKppD5Y
https://github.com/xiaopangxia/TCM-Ancient-Books
https://github.com/yao8839836/PTM
https://github.com/yao8839836/CEMRClass

from r-bert.

This is a good job. Thank you.
And
I remember the f1-score > 88 some days ago, but the f1-score < 88 I ran last night.
Because the entities fully-connected layer use the same weight after the update?
Could you share how to adjust parameters such as the learning rate to get more results?
Because I want to use R-BERT in my dataset, but the result is not very well.
Thanks.

你好，请问你有复现到89.25吗 我现在最高调试到88.39，你的最高能到多少呢

你好，我跑了一下这个代码，结果只得到87.95，请问你是通过什么调试到88.39的呢，方便分享下方法吗，谢谢~

from r-bert.

This is a good job. Thank you.
And
I remember the f1-score > 88 some days ago, but the f1-score < 88 I ran last night.
Because the entities fully-connected layer use the same weight after the update?
Could you share how to adjust parameters such as the learning rate to get more results?
Because I want to use R-BERT in my dataset, but the result is not very well.
Thanks.

你好，请问你有复现到89.25吗 我现在最高调试到88.39，你的最高能到多少呢

你好，我跑了一下这个代码，结果只得到87.95，请问你是通过什么调试到88.39的呢，方便分享下方法吗，谢谢~

就是修改不同的种子，我当时把seed设置为2333就可以得到，你可以尝试一下

from r-bert.

from r-bert.

Mul-BERT 在 the SemEval 2010 Task 8 dataset 到达 90.72 (Macro-F1）方法非常简单，推荐一波：https://github.com/DongPoLI/Mul-BERT，

from r-bert.