Training a Simple Neural Network Manually

Created in May 09, 2026

2026   ·   math   neural-networks   machine-learning   pytorch   ·   tutorials

This notebook trains a simple neural network, using the equations for backpropagation derived in my previous posts (1, 2). The equations are verified against the gradients computed by PyTorch’s autograd engine.



Enjoy Reading This Article?

Here are some more articles you might like to read next:

  • Vectorized Forward and Backward Pass in a Simple Neural Network
  • Forward and Backward Pass in a Simple Neural Network
  • Posting my solutions to Mathematics for Machine Learning book
  • Probability Theory
  • How to publish a jupyter notebook on Github pages with Jekyll
  • Gaussian process regression introduction