Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Why is my eval_accuracy very low? #6

Open
wangyunxiangsha opened this issue Jan 8, 2019 · 10 comments
Open

Why is my eval_accuracy very low? #6

wangyunxiangsha opened this issue Jan 8, 2019 · 10 comments

Comments

@wangyunxiangsha
Copy link

eval_accuracy = 0.05
eval_loss = 0.1677554
global_step = 78750
loss = 0.16775554

@wangyunxiangsha wangyunxiangsha changed the title 为什么最后实验的准确率很低 Why is my eval_accuracy very low? Jan 8, 2019
@leejason
Copy link

Me too. Any advice would be appreciated.

@minmummax
Copy link

me too

@bbingy
Copy link

bbingy commented Feb 15, 2019

Me too. Does anyone find the problem?

@XSilverBullet
Copy link

Me too. Can anyone answer this question?

@netsesame2
Copy link

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?

@YuMiaoTHU
Copy link

Me too!!!!anyone can help? Thanks!!!

@qiu-nian
Copy link

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

@YuMiaoTHU
Copy link

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

@qiu-nian
Copy link

qiu-nian commented Oct 16, 2019 via email

@qiu-nian
Copy link

qiu-nian commented Oct 17, 2019

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。。。。

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

8 participants