-
Notifications
You must be signed in to change notification settings - Fork 7.1k
Ignore deprecation warning from traverse function #6641
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
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@ejguan there is only one test that currently uses only_datapipe=False
:
vision/test/test_prototype_datasets_builtin.py
Lines 104 to 106 in 7046e56
@pytest.mark.parametrize("only_datapipe", [False, True]) | |
@parametrize_dataset_mocks(DATASET_MOCKS) | |
def test_traversable(self, dataset_mock, config, only_datapipe): |
This test together with
def test_serializable(self, dataset_mock, config): |
and
def test_data_loader(self, dataset_mock, config, num_workers): |
are only there to make sure our datasets work seamlessly with all torchdata
utilities and don't error or throw warnings. If only_datapipe=False
is going to be deprecated, we can simply remove the parametrization here instead of ignoring the warning. cc @NicolasHug
Since you are already looking at this: do these three test still cover all utilities that might interact with our datasets from your side or should we add or remove something? 😇
I agree, ignoring the warning will make our tests break when the parameter is removed for good. We should remove its usage now instead. |
For sure, I will do a quick update by removing |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@ejguan I took the liberty to make the changes to fix our CI.
|
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thank you both, LGTM
Summary: * Ignore deprecation warning from traverse function * remove only_datapipe * reinstate warning ignore for DL test * ignore warning class wide due to priority Reviewed By: YosuaMichael Differential Revision: D39885430 fbshipit-source-id: 2c3dbc36cbba0c24ae18b4dfa18e5f7256420b56 Co-authored-by: Philip Meier <[email protected]> Co-authored-by: Nicolas Hug <[email protected]>
Since pytorch/pytorch#84079 is landed eventually, the CI for prototype tests will turn green tmrw when PyTorch nightly is updated. However, this PR introduces a
FutureWarning
that will deprecateonly_datapipe
and the default behavior oftraverse
will be changed asonly_datapipe=True
. And, this PR is suppressing this Warning turning to Error in the test.