Fixed issue with jitted AA transforms in v2 and added tests (#7839)

Author: vfdev-5

test/test_transforms_v2_consistency.py

@@ -927,6 +927,29 @@ def test_randaug(self, inpt, interpolation, mocker):
 
             assert_close(expected_output, output, atol=1, rtol=0.1)
 
+    @pytest.mark.parametrize(
+        "interpolation",
+        [
+            v2_transforms.InterpolationMode.NEAREST,
+            v2_transforms.InterpolationMode.BILINEAR,
+        ],
+    )
+    def test_randaug_jit(self, interpolation):
+        inpt = torch.randint(0, 256, size=(1, 3, 256, 256), dtype=torch.uint8)
+        t_ref = legacy_transforms.RandAugment(interpolation=interpolation, num_ops=1)
+        t = v2_transforms.RandAugment(interpolation=interpolation, num_ops=1)
+
+        tt_ref = torch.jit.script(t_ref)
+        tt = torch.jit.script(t)
+
+        torch.manual_seed(12)
+        expected_output = tt_ref(inpt)
+
+        torch.manual_seed(12)
+        scripted_output = tt(inpt)
+
+        assert_equal(scripted_output, expected_output)
+
     @pytest.mark.parametrize(
         "inpt",
         [
@@ -979,6 +1002,29 @@ def test_trivial_aug(self, inpt, interpolation, mocker):
 
             assert_close(expected_output, output, atol=1, rtol=0.1)
 
+    @pytest.mark.parametrize(
+        "interpolation",
+        [
+            v2_transforms.InterpolationMode.NEAREST,
+            v2_transforms.InterpolationMode.BILINEAR,
+        ],
+    )
+    def test_trivial_aug_jit(self, interpolation):
+        inpt = torch.randint(0, 256, size=(1, 3, 256, 256), dtype=torch.uint8)
+        t_ref = legacy_transforms.TrivialAugmentWide(interpolation=interpolation)
+        t = v2_transforms.TrivialAugmentWide(interpolation=interpolation)
+
+        tt_ref = torch.jit.script(t_ref)
+        tt = torch.jit.script(t)
+
+        torch.manual_seed(12)
+        expected_output = tt_ref(inpt)
+
+        torch.manual_seed(12)
+        scripted_output = tt(inpt)
+
+        assert_equal(scripted_output, expected_output)
+
     @pytest.mark.parametrize(
         "inpt",
         [
@@ -1032,6 +1078,30 @@ def test_augmix(self, inpt, interpolation, mocker):
 
         assert_equal(expected_output, output)
 
+    @pytest.mark.parametrize(
+        "interpolation",
+        [
+            v2_transforms.InterpolationMode.NEAREST,
+            v2_transforms.InterpolationMode.BILINEAR,
+        ],
+    )
+    def test_augmix_jit(self, interpolation):
+        inpt = torch.randint(0, 256, size=(1, 3, 256, 256), dtype=torch.uint8)
+
+        t_ref = legacy_transforms.AugMix(interpolation=interpolation, mixture_width=1, chain_depth=1)
+        t = v2_transforms.AugMix(interpolation=interpolation, mixture_width=1, chain_depth=1)
+
+        tt_ref = torch.jit.script(t_ref)
+        tt = torch.jit.script(t)
+
+        torch.manual_seed(12)
+        expected_output = tt_ref(inpt)
+
+        torch.manual_seed(12)
+        scripted_output = tt(inpt)
+
+        assert_equal(scripted_output, expected_output)
+
     @pytest.mark.parametrize(
         "inpt",
         [
@@ -1061,6 +1131,30 @@ def test_aa(self, inpt, interpolation):
 
         assert_equal(expected_output, output)
 
+    @pytest.mark.parametrize(
+        "interpolation",
+        [
+            v2_transforms.InterpolationMode.NEAREST,
+            v2_transforms.InterpolationMode.BILINEAR,
+        ],
+    )
+    def test_aa_jit(self, interpolation):
+        inpt = torch.randint(0, 256, size=(1, 3, 256, 256), dtype=torch.uint8)
+        aa_policy = legacy_transforms.AutoAugmentPolicy("imagenet")
+        t_ref = legacy_transforms.AutoAugment(aa_policy, interpolation=interpolation)
+        t = v2_transforms.AutoAugment(aa_policy, interpolation=interpolation)
+
+        tt_ref = torch.jit.script(t_ref)
+        tt = torch.jit.script(t)
+
+        torch.manual_seed(12)
+        expected_output = tt_ref(inpt)
+
+        torch.manual_seed(12)
+        scripted_output = tt(inpt)
+
+        assert_equal(scripted_output, expected_output)
+
 
 def import_transforms_from_references(reference):
     HERE = Path(__file__).parent

torchvision/transforms/v2/_auto_augment.py

@@ -28,7 +28,16 @@ def __init__(
     ) -> None:
         super().__init__()
         self.interpolation = _check_interpolation(interpolation)
-        self.fill = _setup_fill_arg(fill)
+        self.fill = fill
+        self._fill = _setup_fill_arg(fill)
+
+    def _extract_params_for_v1_transform(self) -> Dict[str, Any]:
+        params = super()._extract_params_for_v1_transform()
+
+        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}.")
+
+        return params
 
     def _get_random_item(self, dct: Dict[str, Tuple[Callable, bool]]) -> Tuple[str, Tuple[Callable, bool]]:
         keys = tuple(dct.keys())
@@ -335,7 +344,7 @@ def forward(self, *inputs: Any) -> Any:
                 magnitude = 0.0
 
             image_or_video = self._apply_image_or_video_transform(
-                image_or_video, transform_id, magnitude, interpolation=self.interpolation, fill=self.fill
+                image_or_video, transform_id, magnitude, interpolation=self.interpolation, fill=self._fill
             )
 
         return self._unflatten_and_insert_image_or_video(flat_inputs_with_spec, image_or_video)
@@ -419,7 +428,7 @@ def forward(self, *inputs: Any) -> Any:
             else:
                 magnitude = 0.0
             image_or_video = self._apply_image_or_video_transform(
-                image_or_video, transform_id, magnitude, interpolation=self.interpolation, fill=self.fill
+                image_or_video, transform_id, magnitude, interpolation=self.interpolation, fill=self._fill
             )
 
         return self._unflatten_and_insert_image_or_video(flat_inputs_with_spec, image_or_video)
@@ -491,7 +500,7 @@ def forward(self, *inputs: Any) -> Any:
             magnitude = 0.0
 
         image_or_video = self._apply_image_or_video_transform(
-            image_or_video, transform_id, magnitude, interpolation=self.interpolation, fill=self.fill
+            image_or_video, transform_id, magnitude, interpolation=self.interpolation, fill=self._fill
         )
         return self._unflatten_and_insert_image_or_video(flat_inputs_with_spec, image_or_video)
 
@@ -614,7 +623,7 @@ def forward(self, *inputs: Any) -> Any:
                     magnitude = 0.0
 
                 aug = self._apply_image_or_video_transform(
-                    aug, transform_id, magnitude, interpolation=self.interpolation, fill=self.fill
+                    aug, transform_id, magnitude, interpolation=self.interpolation, fill=self._fill
                 )
             mix.add_(combined_weights[:, i].reshape(batch_dims) * aug)
         mix = mix.reshape(orig_dims).to(dtype=image_or_video.dtype)