torchtune icon indicating copy to clipboard operation
torchtune copied to clipboard

[EZ] Fix config bug where interpolation happens too early

Open EugenHotaj opened this issue 10 months ago • 3 comments
trafficstars

There is currently a bug in config interpolation. E.g. given the following config:

name: REPLACE_ME
output_dir: ${name}/logs

When overriding name=some-override via the CLI, the expected config should be:

name: some-override
output_dir: some-override/logs

However, instead we get:

name: some-override
output_dir: REPLACE_ME/logs

This happens because OmegaConf resolves values under the hood when you try to access them.

Context

What is the purpose of this PR? Is it to

  • [ ] add a new feature
  • [x] 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?

  • See above

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.

  • [x] 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. Here is a docstring example and a tutorial example

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

EugenHotaj avatar Jan 07 '25 23:01 EugenHotaj

:link: Helpful Links

:test_tube: See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/torchtune/2236

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

:white_check_mark: No Failures

As of commit 2bab3eca24afef8ef4aafd49a7f0f59e1d624673 with merge base e420bc0c733eafca563d7efa88fb8c0d1663137d (image): :green_heart: Looks good so far! There are no failures yet. :green_heart:

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

pytorch-bot[bot] avatar Jan 07 '25 23:01 pytorch-bot[bot]

Codecov Report

All modified and coverable lines are covered by tests :white_check_mark:

Project coverage is 64.31%. Comparing base (213f386) to head (2bab3ec). Report is 10 commits behind head on main.

Additional details and impacted files
@@            Coverage Diff             @@
##             main    #2236      +/-   ##
==========================================
- Coverage   65.41%   64.31%   -1.11%     
==========================================
  Files         344      352       +8     
  Lines       20658    20569      -89     
==========================================
- Hits        13514    13228     -286     
- Misses       7144     7341     +197     

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.

codecov-commenter avatar Jan 08 '25 00:01 codecov-commenter

This is a great catch. You may need to update the unit tests to make sure the mock value returned by OmegaConf.load is an OmegaConf object. Would you also be able to modify test_parse_known_args in tests/torchtune/config/test_parse.py to test for this edge case?

RdoubleA avatar Jan 08 '25 01:01 RdoubleA

@RdoubleA added ptal when you get a chance.

EugenHotaj avatar Jan 09 '25 18:01 EugenHotaj

@EugenHotaj looks like one more test needs to be fixed, you can also run tests locally with pytest tests

RdoubleA avatar Jan 09 '25 19:01 RdoubleA

Ah missed one of the tests, updated

EugenHotaj avatar Jan 11 '25 20:01 EugenHotaj