Closed Jongchan closed 6 years ago
I wonder if you have successfully reproduced the numbers (accuracy) reported in the paper
These networks are too large for me to train so cannot be trained by me.
Did you try with smaller dataset. Cifar10 for example, does it brings any improvement?
I wonder if you have successfully reproduced the numbers (accuracy) reported in the paper