[CHAPTER] New chapter on supervised fine tuning based on smol course

[burtenshaw]

This is a draft PR for discussion.

It would be cool to reuse the smol course chapter on SFT in the HF NLP course.

Here I've just copied across the content, but here the next step I would propose:

• - update the structure to course norms. i.e. mdx and no ipynb
• - change the style
• - link back to relevant course chapters

Let me know if others thing this is useful, and I'm open to feedback.

This PR now depends on these notebooks huggingface/notebooks#553

[Vaibhavs10]

Very nice start! 🔥

[burtenshaw]

@Vaibhavs10 I've done this:

• refactored the SFT content into an NLP course structure.
• added quizzes
• added tips to do exercises

Next, I'd like to add the colab links to some pages

[Vaibhavs10]

sounds good to me, I can do a review as soon as you add the colabs? so that I look at all in one-go? does that work?

[burtenshaw]

Makes total sense. When it's ready I'll change status and ping you.

[HuggingFaceDocBuilderDev]

The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update.

[Vaibhavs10]

Very solid content, I definitely learnt a thing whilst reviewing, most of these are minor nits! 🔥

[burtenshaw]

@Vaibhavs10 Thanks for the review! I responded to all the comments and worked a bit more on the SFT page.

[Vaibhavs10]

Requested @lewtun to give it a look as well, otherwise this is good to ship imo!

[lewtun]

Really nice chapter with a great flow and clear, easy to understand text - well done @burtenshaw ! I left a few minor comments, but agree with @Vaibhavs10 that this can go live whenever you want