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

could you actually train a new model? #2

Open
chevalfouk opened this issue Dec 8, 2017 · 0 comments
Open

could you actually train a new model? #2

chevalfouk opened this issue Dec 8, 2017 · 0 comments

Comments

@chevalfouk
Copy link

chevalfouk commented Dec 8, 2017

Hi, I am currently trying to train the cityscapes dataset with your code, but the result is miserable: still 0.5263158 (=1/19) class accuracy after 120 epochs. Apparently, the loss of training data is converged correctly, so it seems like a perfect over fitting. Since I used the same settings as yours, i am wondering how you managed to reproduce the results(maybe i need less learning rate?). thanks in advance!

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