Deep learning with PyTorch

Tuesday, July 12  &  Thursday, July 14
9:30am–12:30pm Pacific Time

This course, suitable for people with no knowledge of machine learning, will walk you through the core concepts of deep neural networks. We will use PyTorch, a framework extremely popular in academic research.

Instructor: Marie-Hélène Burle (SFU)

Prerequisites: Working knowledge of Python or attendance at the Basics of Python course.

Zoom   Day 1 – 9:30am–9:40am Pacific

Opening session

On your own

What are neural networks & how do they learn?

Zoom   Day 1 – 11:00am–12:30pm Pacific

Overarching concept of deep learning
Which framework to choose?
PyTorch API & libraries
(Optional) Local installation
Higher-level frameworks built on top of PyTorch
Resources
Jupyter
PyTorch tensors
Automatic differentiation

Zoom   Day 2 – 9:30am–12:30pm Pacific

Creating a DataLoader from a classic audio dataset
Creating a DataLoader from a classic vision dataset
Data pre-processing
Building a neural network
Training a model
Saving/loading models and checkpointing
Machine learning on production clusters
A few notes on PyTorch distributed