Make training args fully immutable #25435
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
muellerzr
commented
Aug 10, 2023
src/transformers/training_args.py
Outdated
|
|
||
| def __setattr__(self, name, value): | ||
| # Once fully through the `__post_init__`, `TrainingArguments` are immutable | ||
| if getattr(self, "_frozen", False): |
There was a problem hiding this comment.
Since adding _frozen to the dataclass args earlier would make it show up as an available option to pass through, we don't set it until the very end of __post_init__, hence the getattr instead of setting it earlier.
5 tasks
|
The documentation is not available anymore as the PR was closed or merged. |
sgugger
approved these changes
Aug 10, 2023
muellerzr
commented
Aug 10, 2023
sgugger
reviewed
Aug 11, 2023
| @@ -205,7 +205,7 @@ def compute_metrics(p: EvalPrediction) -> Dict: | |||
| logger.error(p.metrics) | |||
| exit(1) | |||
|
|
|||
| trainer.args.eval_accumulation_steps = 2 | |||
| trainer.args._set_value("eval_accumulation_steps", 2) | |||
There was a problem hiding this comment.
It's better to create a new set of training args here. People sometimes look for inspiration in our tests and we definitely don't want to advertise that method.
There was a problem hiding this comment.
Done, there's also a method in dataclasses that lets us change frozen params the proper way with a new set of args overriding it
muellerzr
force-pushed
the
muellerzr-immutable
branch
from
3e3e1ee
to
02c2b62
Compare
|
@sgugger can you give it one final look please 😄 |
blbadger
pushed a commit
to blbadger/transformers
that referenced
this pull request
Nov 8, 2023
* Make training args fully immutable * Working tests, PyTorch * In test_trainer * during testing * Use proper dataclass way * Fix test * Another one * Fix tf * Lingering slow * Exception * Clean
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
What does this PR do?
This PR ensures that the
TrainingArgumentsare a fully immutable dataclass after the__post_init__has been ran. We'll find that the tests suddenly fail now 😉 Should be merged after #25390Before submitting
Pull Request section?
to it if that's the case.
documentation guidelines, and
here are tips on formatting docstrings.
Who can review?
Anyone in the community is free to review the PR once the tests have passed. Feel free to tag
members/contributors who may be interested in your PR.
@amyeroberts @sgugger