Add part 2 of end-to-end tutorial: fine-tuning

[andrewor14]

This commit adds the QAT tutorial and a general structure for the fine-tuning tutorial, which all also include QLoRA and float8 quantized fine-tuning. It also connects the 3 tutorial parts (pre-training, fine-tuning, and serving) into one cohesive end-to-end flow with some visuals and text.

Preview it yourself here: https://docs-preview.pytorch.org/pytorch/ao/2394/finetuning.html

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/ao/2394

• 📄 Preview Python docs built from this PR

Note: Links to docs will display an error until the docs builds have been completed.

✅ No Failures

As of commit 2c50f0a with merge base 8b12ddf ():
💚 Looks good so far! There are no failures yet. 💚

This comment was automatically generated by Dr. CI and updates every 15 minutes.

[jerryzh168]

thanks, the same image should probably be added to other tutorials as well

[jerryzh168]

when do we want people use this?

[andrewor14]

I think if people have their own fine-tuning frameworks they want to use, e.g. the Llama-3.2 1B/3B quantized release did not use torchtune or axolotl

[jerryzh168]

nit: I think we can order these from most popular to least popular

[andrewor14]

Ok, I will put the integrations first since we want OSS users to use those first

[andrewor14]

thanks, the same image should probably be added to other tutorials as well

Yeah I added the corresponding images in part 1 and part 3