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

我发现你读取数据做归一化的部分是有问题的 #7

Open
ltpjob opened this issue Jul 16, 2019 · 3 comments
Open

我发现你读取数据做归一化的部分是有问题的 #7

ltpjob opened this issue Jul 16, 2019 · 3 comments

Comments

@ltpjob
Copy link

ltpjob commented Jul 16, 2019

align/utils.py 第195行 归一化的的地方是不是有问题啊? 少加了括号。
image=tf.cast(image,tf.float32)-127.5/128.0

@LeslieZhoa
Copy link
Owner

LeslieZhoa commented Jul 18, 2019

确实是有问题的,我的模型应该就是基于此训练的,所以如果要改正需要重新训练模型,不过我想效果应该会更好!相应问题已添加到readme,再次感谢!

@ltpjob
Copy link
Author

ltpjob commented Jul 18, 2019

还有一处归一化错误在test/embedding.py 第115行 正好训练和测试的归一化都错了所以你在测试的时候发现不了问题。。 网络第一层conv就有bn,所以其实影响应该不大。
还有readme最后那个动图的识别效果是训练15epoch的效果嘛?

@LeslieZhoa
Copy link
Owner

设定是150epoch但我应该提前结束了,具体哪个epoch我也记不清了

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