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

loss下降问题 #35

Open
yyfanxing opened this issue Nov 28, 2018 · 5 comments
Open

loss下降问题 #35

yyfanxing opened this issue Nov 28, 2018 · 5 comments

Comments

@yyfanxing
Copy link

您好,我用这个代码训练synth90k的英文数据,数据量比较大,训练的时候一开始loss是正常下降的,但是后来经过每个epoch后loss竟然又上升了,这个不适合大数据量吗,无法收敛啊。我还试过小数据集进行训练,loss是正常下降了,但是到0.2就不下降了,不知道您有没有遇到类似情况。

2018-11-28 09-35-28

@wushilian
Copy link
Owner

可以把学习率调小一点,我已经很久没用这个代码了

@yyfanxing
Copy link
Author

yyfanxing commented Nov 28, 2018

@wushilian 我已经把学习率调到1e-5了,还要再调小吗,但是loss一直不下降,一开始不断上升,几个epoch后开始不再上升了,但是训练了一晚上,loss还是在1作用。

@wushilian
Copy link
Owner

@yyfanxing 我记得很久之前在syn90k数据上训练过,学习率是1e-4,优化器是adam,可以收敛

@yyfanxing
Copy link
Author

我现在用50万的数据进行训练,仍然很难收敛,是不是数据集大的话attention很难收敛?我感觉训练个几天都无法收敛。

@HansRen1024
Copy link

@yyfanxing 请问您解决这个问题了嘛?我最近在用这个模型训练,也发现不收敛。

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

3 participants