This is a presentation and demonstration about PyTorch.
Template.ipynb
is a skeleton Jupyter Notebook file intended to be worked on alongside the presentation. Complete.ipynb
is a completed Jupyter Notebook with some additional bonus features, such as class weighting to prevent bias, more advanced logging, and training performance plotting. These files demonstrate loading data and designing, initializing, training, and evaluating a basic feed-forward neural network using PyTorch
The slides associated with the presentation and demo are available here.
If the slides become unavailable, Presentation.pdf
is provided as a file backup. However, because the presentation contains animations, some visuals will not work in the pdf.
This demo was originally created for and presented at the Upsilon Pi Epsilon's WPI Chapter's Lunch & Learn event on February 9, 2022. The last major revision to this demo was made for WPI Data Science Club's general body meeting on November 16, 2022.