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

How's the ImageNet training #5

Open
kakusikun opened this issue Nov 21, 2017 · 8 comments
Open

How's the ImageNet training #5

kakusikun opened this issue Nov 21, 2017 · 8 comments

Comments

@kakusikun
Copy link

I'm trying to train this model on ImageNet, but the loss seems to converge slowly after 60k iterations and the loss value is approximately 2.5.

Do you have lower loss value or the similar phenomena above?
I just want to check that I train this model correctly.
Thanks.

@MG2033
Copy link
Owner

MG2033 commented Nov 22, 2017

Sorry for the late reply. I've trained on TinyImageNet-200 and the loss was very small after converging. On ImageNet, I stopped the training because I was very busy doing my exams. However, I think the loss should have a smaller value than 2.5. When I finish training on ImageNet, I will upload the weights ASAP.

@kakusikun
Copy link
Author

The loss looks fine now and keeps decreasing. Thanks.

@juzhitao
Copy link

hi what is your val_accuracy based on TinyImageNet-200?

@AojunZhou
Copy link

@makitann Hi, how about the imagenet result ?

@tensorflowt
Copy link

@MG2033 Hi,I want to know how to train my own data set. Is there any reference code?

@tensorflowt
Copy link

@makitann Have you trained your own data set?

@tensorflowt
Copy link

@MG2033 When I run your code, the following question arises: what is the reason?
FileNotFoundError: [Errno 2] No such file or directory: 'weights.pkl'

@vba34520
Copy link

ASAP.

Hi, Thank you for your work.
Could you please share the weights on ImageNet?
Thanks again.

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

6 participants