Adding vignette for how to use applications in models

[davidrsch]

This PR introduces a new vignette demonstrating how to perform transfer learning with Keras Applications. It provides a clear, end-to-end example of using a pre-trained model (like ResNet50) as a feature extractor within a tidymodels workflow.

This documentation-driven approach was chosen over adding a new helper function to the API, as the existing framework is already sufficiently flexible. The goal is to empower users by clearly documenting this powerful and common pattern.

This addresses the documentation request in issue #28 .

Key Changes:

• Added vignettes/applications.Rmd: A new vignette that walks through using the Keras Functional API to add a new classification head to a frozen, pre-trained base model.
• Updated _pkgdown.yml: Added the new "Transfer Learning" vignette to the "Guides" section of the documentation website's navigation bar.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!