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

Question for config.yaml #81

Open
ramnoob opened this issue Sep 6, 2020 · 4 comments
Open

Question for config.yaml #81

ramnoob opened this issue Sep 6, 2020 · 4 comments

Comments

@ramnoob
Copy link

ramnoob commented Sep 6, 2020

Thx great GAN !
I have two questions .

・What do parameters line8 : size: &SIZE 256 and line59 : image_size: [256, 256] in config.yaml mean ?
Maybe those are size of the image , I want to know more . Are line8 & line59 same value is better ?

・Are there other parts that can be changed as learning parameters ? epochs , batch size , lr , ...

@TheLostIn
Copy link

Have you solve your problem?
I searched image_size in the whole project, but image_szie only exists in config.yaml.

In the default training config, the size was 256,but the resolution of the best result in the paper is 1280720.
Is the training SIZE of the best model equal to 1280
720 ?

@thesby
Copy link

thesby commented Feb 2, 2021

I'm also want to know.

@shawnricecake
Copy link

Have you solve your problem? I searched image_size in the whole project, but image_szie only exists in config.yaml.

In the default training config, the size was 256,but the resolution of the best result in the paper is 1280_720. Is the training SIZE of the best model equal to 1280_720 ?

Hi,

Have u solved the problem with image size when training the model?

@shawnricecake
Copy link

Have you solve your problem? I searched image_size in the whole project, but image_szie only exists in config.yaml.

In the default training config, the size was 256,but the resolution of the best result in the paper is 1280_720. Is the training SIZE of the best model equal to 1280_720 ?

Hi,

I found that in this project, he transform the size into 1280x736 so that it can avoid the error of mismatch when adding upsampled feature maps.

Thanks

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

4 participants