Numpy-MLP-MNIST

Mnist MLP using Numpy

The aim of this notebook is to write forward and backward propagation using numpy to learn more about the gradients computation. The notebook is well documented

Mnist Convnet Pytorch

using the same Mnist dataset we create a simple convnet showing the difference between the MLP model performance and the convnet which is better

TODO

• try to make a simple convnet using numpy
• adding momentum in the update step
• trying to implement other optimizers using numpy