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

楼主您好,请问3500常用汉字的验证码识别,该模型大小够吗? #52

Open
niujiahe opened this issue Feb 25, 2020 · 5 comments

Comments

@niujiahe
Copy link

如题,非常感谢,请问您有什么建议吗?

@ypwhs
Copy link
Owner

ypwhs commented Feb 26, 2020

应该能达到95%以上的准确率,如果不够,可以调节层数、卷积核个数等。

@niujiahe
Copy link
Author

我训练了两天 结果完全发散了。。。可能是因为我训练机不够?每个epoch我采用了100万张图片训练,结果train的loss一直在降低,val的反而在上升。。。我觉得问题可能出现最后的dense,加了多几层dense层也没有用

@niujiahe
Copy link
Author

基本就是百分之0的准确率了,训练的时候还非常非常慢23333

@niujiahe
Copy link
Author

最终val的loss先降低到十几,然后就开始往上升高了

@ypwhs
Copy link
Owner

ypwhs commented Feb 28, 2020

100万张图肯定是非常够用了,你可以检查一下是哪里写错了

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

2 participants