A Numpy Neural Network library and much more

Introduction

A libray which focuses on intuitions and visualizations of neural networks implemented in raw numpy code.

Why it matters

1. A minimalist and intuitive API similar to pytorch and keras
2. Doesn’t confuse the beginner by providing a long list of hyperparameter choices
3. Compare perfomace of different optimizers easily
4. Plot the hidden layers as weights effortlessly to visualize how image reocognition models work
5. Option to display the math to get a idea regarding the theory applied

Contents

Neural Networks

• Neural networks aka deep learning
  • Intution
  • Before Deep learning
  • Neural Networksg

Initialization

• Initialization
  • What you need to know

Forward propogation

• About Forward prop

Loss Functions

• Loss Functions
  • Contents