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

训练的时候 报错conv_24层不存在 #25

Open
970814 opened this issue Oct 9, 2021 · 3 comments
Open

训练的时候 报错conv_24层不存在 #25

970814 opened this issue Oct 9, 2021 · 3 comments

Comments

@970814
Copy link

970814 commented Oct 9, 2021

yolov4.weightsyolov4-tiny.weights 都不存在 conv_24 层, 下图的第二行代码报错,只存在concat24层而不是 conv_24
image

后来我换成 yolov3-tiny.weights ,可以正常训练,不会报这个错,但是在训练过程中报 loss is nan,
image

@970814
Copy link
Author

970814 commented Oct 9, 2021

大佬求助

@970814
Copy link
Author

970814 commented Oct 9, 2021

已经 解决,作者实现的代码是针对 yolov3 训练的,yolov4 要自己改,是替换 140,151,162 三个卷积层,但NaN问题依然没有解决

@970814
Copy link
Author

970814 commented Oct 11, 2021

NaN问题依然没有解决,求助,尝试了降低学习率也不行,出现NaN 参数主要是,Prediction Feature 全是NaN

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

1 participant