[WIP] Sample packing deep dive

[RdoubleA]

Context

Deep dive in live docs on sample packing with in-sample masking and position encoding and how it impacts training throughput and model quality

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

Test plan

Please make sure to do each of the following if applicable to your PR. (If you're not sure about any one of these just ask and we will happily help.)

• 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.)

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/torchtune/1083

• 📄 Preview Python docs built from this PR

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

❌ 1 New Failure

As of commit 9286a0f with merge base 0e0aff0 ():

NEW FAILURE - The following job has failed:

• Build Docs / build_docs (3.11) (gh)

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

[ebsmothers]

doc build failed