This repo is a resource for my Deep Learning with PyTorch talk. It contains all of the code that was demonstrated as well as the deck.

This talk is inspired by a PyTorch tutorial available online. It is also inspired by a silly nine squares image problem.

The purpose of this silly problem is to provide the basic intution behind models, cost, accuracy, and optimization principles behind machine learning.

• input
• model function
• cost function
• optimization method

• Tensors
• DataSets
• DataLoaders
• Models
• Loss (or Cost)
• Optimization

• ONNX
• Azure Machine Learning service