allow sequence fill for v2 AA scripted

[pmeier]

#7839 fixed the v2 AA transform family to not be JIT scriptable, since we unconditionally converted the user input for fill into a dictionary. However, the fix made it more strict than it needed to be:

We currently only allow scalar values for fill

vision/torchvision/transforms/v2/_auto_augment.py

Lines 36 to 37 in 96950a5

	if not (params["fill"] is None or isinstance(params["fill"], (int, float))):
	raise ValueError(f"{type(self).__name__}() can only be scripted for a scalar `fill`, but got {self.fill}.")

but the v1 equivalent works with sequences as well, e.g.

vision/test/test_transforms_tensor.py

Lines 583 to 589 in 96950a5

	@pytest.mark.parametrize("fill", [None, 85, (10, -10, 10), 0.7, [0.0, 0.0, 0.0], [1], 1])
	def test_autoaugment(device, policy, fill):
	tensor = torch.randint(0, 256, size=(3, 44, 56), dtype=torch.uint8, device=device)
	batch_tensors = torch.randint(0, 256, size=(4, 3, 44, 56), dtype=torch.uint8, device=device)
	transform = T.AutoAugment(policy=policy, fill=fill)
	s_transform = torch.jit.script(transform)

Thus, we only need to exclude the case of a user supplied fill dictionary.

cc @vfdev-5

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/vision/7919

• 📄 Preview Python docs built from this PR

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

❌ 3 New Failures, 4 Unrelated Failures

As of commit ac6bd0c with merge base 96950a5 ():

NEW FAILURES - The following jobs have failed:

• unittests-macos (3.10, macos-12) / macos-job (gh)
• unittests-windows (3.10, windows.4xlarge, cpu) / windows-job (gh)
• unittests-windows (3.8, windows.g5.4xlarge.nvidia.gpu, cuda, 11.8) / windows-job (gh)

FLAKY - The following jobs failed but were likely due to flakiness present on trunk:

• unittests-windows (3.11, 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)
• windows (windows.g5.4xlarge.nvidia.gpu, cuda, 11.8) / windows-job (gh)

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

[NicolasHug]

Thanks a lot @pmeier , LGTM when green