Why is my eval_accuracy very low? about sentiment_analysis_fine_grain HOT 10 OPEN

Me too. Any advice would be appreciated.

from sentiment_analysis_fine_grain.

me too

from sentiment_analysis_fine_grain.

Me too. Does anyone find the problem?

from sentiment_analysis_fine_grain.

Me too. Can anyone answer this question?

from sentiment_analysis_fine_grain.

I got the similar result after nearly 16 hours run on GPU.
Can't accept both the accuracy and the long time, is it normally or not?

from sentiment_analysis_fine_grain.

Me too！！！！anyone can help? Thanks!!!

from sentiment_analysis_fine_grain.

我的accuracy也是0.05，有谁知道怎么解决这个问题吗？是不是代码或者计算的问题啊

from sentiment_analysis_fine_grain.

我的accuracy也是0.05，有谁知道怎么解决这个问题吗？是不是代码或者计算的问题啊

acc的计算不适用于多标签，低也是正常的，你可以自己写一个文件用来计算precision，recall，F1值

from sentiment_analysis_fine_grain.

from sentiment_analysis_fine_grain.

我的accuracy也是0.05，有谁知道怎么解决这个问题吗？是不是代码或者计算的问题啊

acc的计算不适用于多标签，低也是正常的，你可以自己写一个文件用来计算precision，recall，F1值

我后面又找到一个计算每个标签的acc的代码了，然后试了一下是可以的，https://www.kaggle.com/javaidnabi/toxic-comment-classification-using-bert/ 就是这个项目。
所以大家的acc都是0.05应该是因为作者的acc那一块的代码写得有问题。。。。

from sentiment_analysis_fine_grain.