port vertical flip

[pmeier]

cc @vfdev-5

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/vision/7712

• 📄 Preview Python docs built from this PR

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

❌ 2 New Failures, 3 Unrelated Failures

As of commit 49341c0:

NEW FAILURES - The following jobs have failed:

• pytorch/vision / manywheel-py3_8-cuda12_1 (gh)
• unittests-linux (3.8, linux.g5.4xlarge.nvidia.gpu, cuda, 11.8) / linux-job (gh)

BROKEN TRUNK - The following jobs failed but were present on the merge base 0e49615:

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

• unittests-windows (3.10, windows.4xlarge, cpu) / windows-job (gh)
• unittests-windows (3.8, windows.4xlarge, cpu) / windows-job (gh)
• unittests-windows (3.9, windows.4xlarge, cpu) / windows-job (gh)

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

[pmeier]

Same as #7703 (comment)

[NicolasHug]

Thanks Philip, LGTM but did we consider a base class for both TestHorizontalFlip and TestVerticalFlip? They look like identical copies with only a few names / callable / matrices changes

[pmeier]

They look like identical copies with only a few names / callable / matrices changes

They are. I didn't find a good pattern using base classes:

class Base:
    FOO = None

    @pytest.mark.parametrize("foo", FOO)
    def test_foo(self):
        ...

class Test(Base):
    FOO = ["foo", "bar", "baz"]

The snippet above does not work, since @pytest.mark.parametrize is evaluated at class creation and thus sees FOO=None.

Another approach would be to merge both classes into one TestHorizontalVerticalFlip and parametrize over the kernels, dispatchers, and transforms, e.g.

@pytest.mark.parametrize("kernel", [F.horizontal_flip_image_tensor, F.vertical_flip_image_tensor])
def test_image_kernel(kernel):
    check_kernel(kernel, ...)

If we want that, I'll implement it in a follow-up PR.

[github-actions]

Hey @pmeier!

You merged this PR, but no labels were added. The list of valid labels is available at https://github.com/pytorch/vision/blob/main/.github/process_commit.py