Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

fix for pyTorch 1.1 #552

Merged
merged 6 commits into from
Dec 1, 2019
Merged

fix for pyTorch 1.1 #552

merged 6 commits into from
Dec 1, 2019

Conversation

Borda
Copy link
Member

@Borda Borda commented Nov 26, 2019

What does this PR do?

Follow-up to #546 and #549 Fixes #491

PR review

Anyone in the community is free to review the PR once the tests have passed.
If we didn't discuss your PR in Github issues there's a high chance it will not be merged.

@Borda
Copy link
Member Author

Borda commented Nov 26, 2019

now the error is following but it seems not to be related to this module...

>       checkpoint = init_checkpoint_callback(logger)
tests/test_a_restore_models.py:82: 
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 
tests/utils.py:235: in init_checkpoint_callback
    exp = logger.experiment
pytorch_lightning/logging/test_tube_logger.py:37: in experiment
    rank=self.rank,
.tox/py37/lib/python3.7/site-packages/test_tube/log.py:154: in __init__
    super().__init__(log_dir=log_dir, *args, **kwargs)
.tox/py37/lib/python3.7/site-packages/torch/utils/tensorboard/writer.py:199: in __init__
    self._get_file_writer()
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 
self = <test_tube.log.Experiment object at 0x7f533d7c8b70>
[1m    def _get_file_writer(self):
        """Returns the default FileWriter instance. Recreates it if closed."""
        if self.rank > 0:
            return TTDummyFileWriter()
    
        if self.all_writers is None or self.file_writer is None:
>           if self.purge_step is not None:
E           AttributeError: 'Experiment' object has no attribute 'purge_step'
.tox/py37/lib/python3.7/site-packages/test_tube/log.py:504: AttributeError

@williamFalcon
Copy link
Contributor

@Borda that’s a test tube error. maybe submit a quick PR there?

@Borda
Copy link
Member Author

Borda commented Nov 26, 2019

well, I was confused because test-tube does not have PyTorch in requirements

@williamFalcon
Copy link
Contributor

true. but test-tube is a subclass of summarywriter

@Borda
Copy link
Member Author

Borda commented Nov 27, 2019

@jeffling @neggert @Ir1d anyone wants to make a fix on test-tube? :)

Borda referenced this pull request Nov 28, 2019
* min pytorch 1.2

* fix IterableDataset

* upgrade torchvision

* fix msg
@Borda
Copy link
Member Author

Borda commented Nov 30, 2019

@williamFalcon it seems that the fix helps for crashing, could you merge and release a new version 0.7.4? Then I would finalize this PR... :)

https://circleci.com/workflow-run/e15c02ed-128b-4081-b283-ea6e2690fdb4

@williamFalcon
Copy link
Contributor

@Borda 0.7.4 pushed!

@williamFalcon
Copy link
Contributor

williamFalcon commented Dec 1, 2019

@Borda looks like test-tube install has a bug.

never mind, fixed it.

@Borda Borda changed the title fix for pyTorch 1.1 [wip] fix for pyTorch 1.1 Dec 1, 2019
@williamFalcon williamFalcon merged commit 89ececb into Lightning-AI:master Dec 1, 2019
@Borda Borda deleted the pt_1-1 branch December 1, 2019 08:44
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

IterableDataset breaks 1.1 compatibility
2 participants