[Draft]Support Optimizer-in-the-backward

[mori360]

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)

Please link to any issues this PR addresses.

Changelog

What are the changes made in this PR?
In FullFinetuneDistributed, switch self._optimizer to self._optimizer_in_bwd that runs in backwards
Running in the backwards could save the peak memory cost during loss.backward()
By the local testing with llama2/7B_full, peak memory is optimized from 31.5GB to 21.2GB, saves 32.6%

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. 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.)

UX

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

torchtune/torchtune/modules/vision_transformer.py

Line 285 in 6a7951f

Examples:

Example in our docs: https://pytorch.org/torchtune/main/tutorials/qat_finetune.html#applying-qat-to-llama3-models

• 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/1530

• 📄 Preview Python docs built from this PR

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

✅ You can merge normally! (2 Unrelated Failures)

As of commit 0a3762d with merge base 0a3762d ():

BROKEN TRUNK - The following jobs failed but were present on the merge base:

👉 Rebase onto the `viable/strict` branch to avoid these failures

• Regression Tests / regression_test (3.11, nightly) (gh) (trunk failure)
  tests/regression_tests/test_llama2_7b.py::TestLoRA7BDistributedFinetuneEval::test_finetune_and_eval
• Regression Tests / regression_test (3.11, stable) (gh) (trunk failure)
  tests/regression_tests/test_llama2_7b.py::TestLoRA7BDistributedFinetuneEval::test_finetune_and_eval

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

[codecov-commenter]

Codecov Report

Attention: Patch coverage is 0% with 32 lines in your changes missing coverage. Please review.

Project coverage is 72.13%. Comparing base (dd348ce) to head (5ec6f68).
Report is 5 commits behind head on main.

Files with missing lines	Patch %	Lines
recipes/full_finetune_distributed.py	0.00%	32 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #1530      +/-   ##
==========================================
- Coverage   72.26%   72.13%   -0.13%     
==========================================
  Files         290      290              
  Lines       14554    14576      +22     
==========================================
- Hits        10517    10515       -2     
- Misses       4037     4061      +24     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[mori360]

"lr" here are all the same