maderix / pytorch-notebooks

A collection of notebooks implementing basic NN models in pytorch

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

pytorch-notebooks

A collection of notebooks implementing basic NN models in pytorch.

Motivation

I started my ML journey in Keras/Tensorflow two years back. While I loved the simple interface of Keras but had a love/hate relationship with Tensorflow due to multiple layers of complexity involved. As my work involved deploying ML models on embedded devices, there was not much incentive for me to try out other frameworks.

I always fancied the dynamic nature of Pytorch and it's Pythonic nature but never got around to doing anything substantial with it. So to change this, I'm gonna create simple models in it. There is some learning curve involved but since I've some experience with Numpy, the syntax seems mostly familiar.

Hope this repo proves useful to other individuals starting out on their journey in Pytorch. :)

Contents (so far)

  1. MNIST with simple neural nets
  2. CIFAR10 with a simple Convolutional Neural Network(CNN)
  3. CIFAR10 with a RESidual Network from scratch(ResNet18)

About

A collection of notebooks implementing basic NN models in pytorch

License:Apache License 2.0


Languages

Language:Jupyter Notebook 100.0%