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

ResNeXt (2017, Facebook AI Research) #58

Closed
taki0112 opened this issue Sep 6, 2017 · 2 comments
Closed

ResNeXt (2017, Facebook AI Research) #58

taki0112 opened this issue Sep 6, 2017 · 2 comments

Comments

@taki0112
Copy link

taki0112 commented Sep 6, 2017

HI
When I tested the fashion mnist dataset with 50 epochs, the accuracy was up to 90%. Maybe it will be higher if you increase the epoch and adjust the hyperparameter.

Data augmentation has not been done.

https://github.com/taki0112/ResNeXt-Tensorflow

thank you

@kashif
Copy link
Collaborator

kashif commented Sep 7, 2017

@taki0112 I suspect something is wrong with your ResNeXt implementation, since a simple 3 layer ConvNet will give about 91% accuracy after 10 or 12 epochs... So therefore kindly try to debug since I suspect ResNeXt should give you much more...

@taki0112
Copy link
Author

taki0112 commented Sep 7, 2017

Yes, thank you for your advice.
The residual layer was incorrectly coded and corrected. Let's try again.

@hanxiao hanxiao closed this as completed Sep 7, 2017
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

No branches or pull requests

3 participants