ML Vision Projects - Hands-On Meetup (please sign up so we know to let you in)

## Modifying a Neural Network & Seeing What Happens

We are taking the CIFAR-10 image classifier PyTorch tutorial and actively experimenting with it.
Instead of treating the tutorial as a finished recipe, we are:

• Extending and modifying classes
• Observing class imbalance effects
• Freezing vs retraining layers
• Tweaking parts of the network architecture
• Comparing training behavior and results

The focus is on understanding fundamentals:

• Look directly at the model
• Change one thing
• Observe how it affects training, convergence, and accuracy

This is a read-and-modify-the-code session.
We will be working directly in PyTorch, reading real model and training code to understand what it is doing—not prompting a model to give us answers or “vibe coding.”
Bring a laptop, or feel free to look over someone’s shoulder. We will help you get unstuck and explain what’s happening as we go.