Search⌘ K
AI Features

Stacking Fire

Explore how to create a utility function that stacks several fire modules in a SqueezeNet model, helping you efficiently build complex CNN architectures with repeated building blocks for image classification.

Chapter Goals:

  • Write a utility function to stack multiple fire modules

A. Utility function

While the SqueezeNet model uses very few parameters, it still has many layers. The original architecture, which was built for the larger ImageNet dataset, uses 8 fire modules. Our model is a condensed version, and only uses 4. However, that is still several fire modules, ...