CapsNet with multiple layers of Conv.layer Generate capsules from higher-level features with more abstract representations.
$ pip install -r requirements.txt
Download the AffNIST dataset from the website(https://www.cs.toronto.edu/~tijmen/affNIST/) and save it. Clone git.
$ git clone https://github.com/Koko-zep/Stacked_CapsNet.git
$ cd Stacked_CapsNet
Train the model
$ python main.py
Check the results with Tensorboad.
$ cd capsnet_pytorch
$ tensorboard --logdir ./runs

