Llama3.2 knowledge distillation config

[lindawangg]

Context

What is the purpose of this PR? Is it to

• add a new feature
• fix a bug
• update tests and/or documentation
• other (please add here)

Changelog

What are the changes made in this PR?

• Knowledge distillation config to distill LLAMA3 models, specifically Llama 3.1 8B into Llama 3.2 1B

Test plan

Please make sure to do each of the following if applicable to your PR. If you're unsure about any one of these just ask and we will happily help. We also have a contributing page for some guidance on contributing.

• run pre-commit hooks and linters (make sure you've first installed via pre-commit install)
• add unit tests for any new functionality
• update docstrings for any new or updated methods or classes
• run unit tests via pytest tests
• run recipe tests via pytest tests -m integration_test
• manually run any new or modified recipes with sufficient proof of correctness
• include relevant commands and any other artifacts in this summary (pastes of loss curves, eval results, etc.)

tune run knowledge_distillation_single_device --config llama3_2/knowledge_distillation_single_device

Eval results show that using KD yields slightly better results than just standard fine-tuning.

UX

If your function changed a public API, please add a dummy example of what the user experience will look like when calling it.
Here is a docstring example
and a tutorial example

• I did not change any public API
• I have added an example to docs or docstrings

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/torchtune/1690

• 📄 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 f7c5cd0 with merge base a899da2 ():
💚 Looks good so far! There are no failures yet. 💚

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

[ebsmothers]

This looks great! Can you also add numbers to the PR summary on peak memory and tokens/sec you see when running with this config? Modulo that I think this is good to go though

[ebsmothers]

nit

Suggested change

	# Config for single device knowledge distillation in kd_single_device.py
	# Config for single device knowledge distillation in knowledge_distillation_single_device.py

[ebsmothers]

Looks great, thanks for adding this!