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

Training loss does not decrease, training acc is equal to 0 #99

Open
Jia-Xueyu opened this issue Dec 14, 2021 · 2 comments
Open

Training loss does not decrease, training acc is equal to 0 #99

Jia-Xueyu opened this issue Dec 14, 2021 · 2 comments

Comments

@Jia-Xueyu
Copy link

Hello, I use the casia-webface to train the model. However, the train loss loss does not drop and remains at around 14. When I use the pre-train model Resnet18-110, the loss can decrease. Can you elaborate on the process of training data?Thank you.

@disguise6
Copy link

After setting s = 6, m = 2, loss decreased and acc increased to 1. But I use my own dataset, not casia-webface.

@abcde-bit
Copy link

Hello, I use the casia-webface to train the model. However, the train loss loss does not drop and remains at around 14. When I use the pre-train model Resnet18-110, the loss can decrease. Can you elaborate on the process of training data?Thank you.

Hi brother. I also have the problem. How did you solve the problem at last.
image

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