Add finetuning exercise to transfer learning episode

[qualiaMachine]

Finetuning is a critical component of transfer learning, and I think this episode really needs to explore its impact. This added exercise has learners unfreeze one layer of the pretrained model, and compare this to our original frozen model's performance. We see a noticeable improvement with finetuning, as expected. The solution discusses the pros/cons of unfreezing layers, and how to balance overfitting and training time considerations.

[github-actions]

Thank you!

Thank you for your pull request 😃

🤖 This automated message can help you check the rendered files in your submission for clarity. If you have any questions, please feel free to open an issue in {sandpaper}.

If you have files that automatically render output (e.g. R Markdown), then you should check for the following:

• 🎯 correct output
• 🖼️ correct figures
• ❓ new warnings
• ‼️ new errors

Rendered Changes

🔍 Inspect the changes: https://github.com/carpentries-lab/deep-learning-intro/compare/md-outputs..md-outputs-PR-584

The following changes were observed in the rendered markdown documents:

 5-transfer-learning.md               | 114 +++++++++++++++++++++++++++++++++++
 fig/05-frozen_vs_finetuned.png (new) | Bin 0 -> 112734 bytes
 md5sum.txt                           |   2 +-
 3 files changed, 115 insertions(+), 1 deletion(-)

What does this mean?

If you have source files that require output and figures to be generated (e.g. R Markdown), then it is important to make sure the generated figures and output are reproducible.

This output provides a way for you to inspect the output in a diff-friendly manner so that it's easy to see the changes that occur due to new software versions or randomisation.

⏱️ Updated at 2025-05-15 20:12:38 +0000

[carschno]

This looks like a very valuable addition indeed!
I have added a comment about the added figure.

[carschno]

The image does not look very convincing, both curves end at a similar value for validation accuracy (~0.6). This seems to require some explanation.

[carschno]

It would be good to have alt image descriptions everywhere, e.g.

Suggested change

	
	{alt="A comparison of the accuracy on the validation set for both the frozen and the fine-tuned setup."}