This repo shows the CNN implementation based in pytorch for the fashion mnist dataset. The basic structure of the CNN is : Conv(1,16) ==>> BatchNorm ==>> MaxPool ==>> Conv(16,32) ==> BatchNorm ==>> MaxPool ==>> Fully Connected Layers
The Data set could be downloaded from the : https://github.com/zalandoresearch/fashion-mnist The Overall accuracy achieved in 10 epochs for train dataset is 93% while for test dataset it was 91%.