Add missing None type hint to init functions

torchvision/datasets/_optical_flow.py

@@ -70,7 +70,7 @@ def __getitem__(self, index):
         else:
             return img1, img2, flow
 
-    def __len__(self):
+    def __len__(self) -> int:
         return len(self._image_list)
 
     def __rmul__(self, v):
@@ -118,7 +118,7 @@ class Sintel(FlowDataset):
             return a built-in valid mask, such as :class:`~torchvision.datasets.KittiFlow`.
     """
 
-    def __init__(self, root, split="train", pass_name="clean", transforms=None):
+    def __init__(self, root, split="train", pass_name="clean", transforms=None) -> None:
         super().__init__(root=root, transforms=transforms)
 
         verify_str_arg(split, "split", valid_values=("train", "test"))
@@ -180,7 +180,7 @@ class KittiFlow(FlowDataset):
 
     _has_builtin_flow_mask = True
 
-    def __init__(self, root, split="train", transforms=None):
+    def __init__(self, root, split="train", transforms=None) -> None:
         super().__init__(root=root, transforms=transforms)
 
         verify_str_arg(split, "split", valid_values=("train", "test"))
@@ -245,7 +245,7 @@ class FlyingChairs(FlowDataset):
             return a built-in valid mask, such as :class:`~torchvision.datasets.KittiFlow`.
     """
 
-    def __init__(self, root, split="train", transforms=None):
+    def __init__(self, root, split="train", transforms=None) -> None:
         super().__init__(root=root, transforms=transforms)
 
         verify_str_arg(split, "split", valid_values=("train", "val"))
@@ -316,7 +316,7 @@ class FlyingThings3D(FlowDataset):
             return a built-in valid mask, such as :class:`~torchvision.datasets.KittiFlow`.
     """
 
-    def __init__(self, root, split="train", pass_name="clean", camera="left", transforms=None):
+    def __init__(self, root, split="train", pass_name="clean", camera="left", transforms=None) -> None:
         super().__init__(root=root, transforms=transforms)
 
         verify_str_arg(split, "split", valid_values=("train", "test"))
@@ -401,7 +401,7 @@ class HD1K(FlowDataset):
 
     _has_builtin_flow_mask = True
 
-    def __init__(self, root, split="train", transforms=None):
+    def __init__(self, root, split="train", transforms=None) -> None:
         super().__init__(root=root, transforms=transforms)
 
         verify_str_arg(split, "split", valid_values=("train", "test"))

torchvision/datasets/_stereo_matching.py

@@ -20,7 +20,7 @@ class StereoMatchingDataset(ABC, VisionDataset):
 
     _has_built_in_disparity_mask = False
 
-    def __init__(self, root: str, transforms: Optional[Callable] = None):
+    def __init__(self, root: str, transforms: Optional[Callable] = None) -> None:
         """
         Args:
             root(str): Root directory of the dataset.
@@ -152,7 +152,7 @@ class CarlaStereo(StereoMatchingDataset):
         transforms (callable, optional): A function/transform that takes in a sample and returns a transformed version.
     """
 
-    def __init__(self, root: str, transforms: Optional[Callable] = None):
+    def __init__(self, root: str, transforms: Optional[Callable] = None) -> None:
         super().__init__(root, transforms)
 
         root = Path(root) / "carla-highres"
@@ -229,7 +229,7 @@ class Kitti2012Stereo(StereoMatchingDataset):
 
     _has_built_in_disparity_mask = True
 
-    def __init__(self, root: str, split: str = "train", transforms: Optional[Callable] = None):
+    def __init__(self, root: str, split: str = "train", transforms: Optional[Callable] = None) -> None:
         super().__init__(root, transforms)
 
         verify_str_arg(split, "split", valid_values=("train", "test"))
@@ -317,7 +317,7 @@ class Kitti2015Stereo(StereoMatchingDataset):
 
     _has_built_in_disparity_mask = True
 
-    def __init__(self, root: str, split: str = "train", transforms: Optional[Callable] = None):
+    def __init__(self, root: str, split: str = "train", transforms: Optional[Callable] = None) -> None:
         super().__init__(root, transforms)
 
         verify_str_arg(split, "split", valid_values=("train", "test"))

torchvision/datasets/folder.py

@@ -305,7 +305,7 @@ def __init__(
         target_transform: Optional[Callable] = None,
         loader: Callable[[str], Any] = default_loader,
         is_valid_file: Optional[Callable[[str], bool]] = None,
-    ):
+    ) -> None:
         super().__init__(
             root,
             loader,

torchvision/datasets/kitti.py

@@ -57,7 +57,7 @@ def __init__(
         target_transform: Optional[Callable] = None,
         transforms: Optional[Callable] = None,
         download: bool = False,
-    ):
+    ) -> None:
         super().__init__(
             root,
             transform=transform,

torchvision/datasets/lfw.py

@@ -38,7 +38,7 @@ def __init__(
         transform: Optional[Callable] = None,
         target_transform: Optional[Callable] = None,
         download: bool = False,
-    ):
+    ) -> None:
         super().__init__(os.path.join(root, self.base_folder), transform=transform, target_transform=target_transform)
 
         self.image_set = verify_str_arg(image_set.lower(), "image_set", self.file_dict.keys())
@@ -119,7 +119,7 @@ def __init__(
         transform: Optional[Callable] = None,
         target_transform: Optional[Callable] = None,
         download: bool = False,
-    ):
+    ) -> None:
         super().__init__(root, split, image_set, "people", transform, target_transform, download)
 
         self.class_to_idx = self._get_classes()
@@ -201,7 +201,7 @@ def __init__(
         transform: Optional[Callable] = None,
         target_transform: Optional[Callable] = None,
         download: bool = False,
-    ):
+    ) -> None:
         super().__init__(root, split, image_set, "pairs", transform, target_transform, download)
 
         self.pair_names, self.data, self.targets = self._get_pairs(self.images_dir)

torchvision/datasets/oxford_iiit_pet.py

@@ -45,7 +45,7 @@ def __init__(
         transform: Optional[Callable] = None,
         target_transform: Optional[Callable] = None,
         download: bool = False,
-    ):
+    ) -> None:
         self._split = verify_str_arg(split, "split", ("trainval", "test"))
         if isinstance(target_types, str):
             target_types = [target_types]

torchvision/datasets/pcam.py

@@ -73,7 +73,7 @@ def __init__(
         transform: Optional[Callable] = None,
         target_transform: Optional[Callable] = None,
         download: bool = False,
-    ):
+    ) -> None:
         try:
             import h5py
 

torchvision/datasets/voc.py

@@ -75,7 +75,7 @@ def __init__(
         transform: Optional[Callable] = None,
         target_transform: Optional[Callable] = None,
         transforms: Optional[Callable] = None,
-    ):
+    ) -> None:
         super().__init__(root, transforms, transform, target_transform)
         if year == "2007-test":
             if image_set == "test":

torchvision/models/_api.py

@@ -48,7 +48,7 @@ class WeightsEnum(StrEnum):
         value (Weights): The data class entry with the weight information.
     """
 
-    def __init__(self, value: Weights):
+    def __init__(self, value: Weights) -> None:
         self._value_ = value
 
     @classmethod

torchvision/models/detection/anchor_utils.py

@@ -159,7 +159,7 @@ def __init__(
         scales: Optional[List[float]] = None,
         steps: Optional[List[int]] = None,
         clip: bool = True,
-    ):
+    ) -> None:
         super().__init__()
         if steps is not None and len(aspect_ratios) != len(steps):
             raise ValueError("aspect_ratios and steps should have the same length")

torchvision/models/detection/faster_rcnn.py

@@ -196,7 +196,7 @@ def __init__(
         box_positive_fraction=0.25,
         bbox_reg_weights=None,
         **kwargs,
-    ):
+    ) -> None:
 
         if not hasattr(backbone, "out_channels"):
             raise ValueError(
@@ -289,7 +289,7 @@ class TwoMLPHead(nn.Module):
         representation_size (int): size of the intermediate representation
     """
 
-    def __init__(self, in_channels, representation_size):
+    def __init__(self, in_channels, representation_size) -> None:
         super().__init__()
 
         self.fc6 = nn.Linear(in_channels, representation_size)
@@ -311,7 +311,7 @@ def __init__(
         conv_layers: List[int],
         fc_layers: List[int],
         norm_layer: Optional[Callable[..., nn.Module]] = None,
-    ):
+    ) -> None:
         """
         Args:
             input_size (Tuple[int, int, int]): the input size in CHW format.
@@ -351,7 +351,7 @@ class FastRCNNPredictor(nn.Module):
         num_classes (int): number of output classes (including background)
     """
 
-    def __init__(self, in_channels, num_classes):
+    def __init__(self, in_channels, num_classes) -> None:
         super().__init__()
         self.cls_score = nn.Linear(in_channels, num_classes)
         self.bbox_pred = nn.Linear(in_channels, num_classes * 4)

torchvision/models/detection/fcos.py

@@ -218,7 +218,7 @@ def __init__(
         num_anchors: int,
         num_convs: int = 4,
         norm_layer: Optional[Callable[..., nn.Module]] = None,
-    ):
+    ) -> None:
         super().__init__()
 
         if norm_layer is None:
@@ -376,7 +376,7 @@ def __init__(
         detections_per_img: int = 100,
         topk_candidates: int = 1000,
         **kwargs,
-    ):
+    ) -> None:
         super().__init__()
         _log_api_usage_once(self)
 

torchvision/models/detection/keypoint_rcnn.py

@@ -199,7 +199,7 @@ def __init__(
         keypoint_predictor=None,
         num_keypoints=None,
         **kwargs,
-    ):
+    ) -> None:
 
         if not isinstance(keypoint_roi_pool, (MultiScaleRoIAlign, type(None))):
             raise TypeError(
@@ -269,7 +269,7 @@ def __init__(
 
 
 class KeypointRCNNHeads(nn.Sequential):
-    def __init__(self, in_channels, layers):
+    def __init__(self, in_channels, layers) -> None:
         d = []
         next_feature = in_channels
         for out_channels in layers:
@@ -284,7 +284,7 @@ def __init__(self, in_channels, layers):
 
 
 class KeypointRCNNPredictor(nn.Module):
-    def __init__(self, in_channels, num_keypoints):
+    def __init__(self, in_channels, num_keypoints) -> None:
         super().__init__()
         input_features = in_channels
         deconv_kernel = 4

torchvision/models/detection/mask_rcnn.py

@@ -198,7 +198,7 @@ def __init__(
         mask_head=None,
         mask_predictor=None,
         **kwargs,
-    ):
+    ) -> None:
 
         if not isinstance(mask_roi_pool, (MultiScaleRoIAlign, type(None))):
             raise TypeError(
@@ -268,7 +268,7 @@ def __init__(
 class MaskRCNNHeads(nn.Sequential):
     _version = 2
 
-    def __init__(self, in_channels, layers, dilation, norm_layer: Optional[Callable[..., nn.Module]] = None):
+    def __init__(self, in_channels, layers, dilation, norm_layer: Optional[Callable[..., nn.Module]] = None) -> None:
         """
         Args:
             in_channels (int): number of input channels
@@ -332,7 +332,7 @@ def _load_from_state_dict(
 
 
 class MaskRCNNPredictor(nn.Sequential):
-    def __init__(self, in_channels, dim_reduced, num_classes):
+    def __init__(self, in_channels, dim_reduced, num_classes) -> None:
         super().__init__(
             OrderedDict(
                 [

torchvision/models/detection/retinanet.py

@@ -65,7 +65,13 @@ class RetinaNetHead(nn.Module):
         norm_layer (callable, optional): Module specifying the normalization layer to use. Default: None
     """
 
-    def __init__(self, in_channels, num_anchors, num_classes, norm_layer: Optional[Callable[..., nn.Module]] = None):
+    def __init__(
+        self,
+        in_channels,
+        num_anchors,
+        num_classes,
+        norm_layer: Optional[Callable[..., nn.Module]] = None,
+    ) -> None:
         super().__init__()
         self.classification_head = RetinaNetClassificationHead(
             in_channels, num_anchors, num_classes, norm_layer=norm_layer
@@ -104,7 +110,7 @@ def __init__(
         num_classes,
         prior_probability=0.01,
         norm_layer: Optional[Callable[..., nn.Module]] = None,
-    ):
+    ) -> None:
         super().__init__()
 
         conv = []
@@ -223,7 +229,7 @@ class RetinaNetRegressionHead(nn.Module):
         "box_coder": det_utils.BoxCoder,
     }
 
-    def __init__(self, in_channels, num_anchors, norm_layer: Optional[Callable[..., nn.Module]] = None):
+    def __init__(self, in_channels, num_anchors, norm_layer: Optional[Callable[..., nn.Module]] = None) -> None:
         super().__init__()
 
         conv = []
@@ -430,7 +436,7 @@ def __init__(
         bg_iou_thresh=0.4,
         topk_candidates=1000,
         **kwargs,
-    ):
+    ) -> None:
         super().__init__()
         _log_api_usage_once(self)
 

torchvision/models/detection/roi_heads.py

@@ -518,7 +518,7 @@ def __init__(
         keypoint_roi_pool=None,
         keypoint_head=None,
         keypoint_predictor=None,
-    ):
+    ) -> None:
         super().__init__()
 
         self.box_similarity = box_ops.box_iou

torchvision/models/detection/ssd.py

@@ -54,7 +54,7 @@ def _xavier_init(conv: nn.Module):
 
 
 class SSDHead(nn.Module):
-    def __init__(self, in_channels: List[int], num_anchors: List[int], num_classes: int):
+    def __init__(self, in_channels: List[int], num_anchors: List[int], num_classes: int) -> None:
         super().__init__()
         self.classification_head = SSDClassificationHead(in_channels, num_anchors, num_classes)
         self.regression_head = SSDRegressionHead(in_channels, num_anchors)
@@ -67,7 +67,7 @@ def forward(self, x: List[Tensor]) -> Dict[str, Tensor]:
 
 
 class SSDScoringHead(nn.Module):
-    def __init__(self, module_list: nn.ModuleList, num_columns: int):
+    def __init__(self, module_list: nn.ModuleList, num_columns: int) -> None:
         super().__init__()
         self.module_list = module_list
         self.num_columns = num_columns
@@ -104,7 +104,7 @@ def forward(self, x: List[Tensor]) -> Tensor:
 
 
 class SSDClassificationHead(SSDScoringHead):
-    def __init__(self, in_channels: List[int], num_anchors: List[int], num_classes: int):
+    def __init__(self, in_channels: List[int], num_anchors: List[int], num_classes: int) -> None:
         cls_logits = nn.ModuleList()
         for channels, anchors in zip(in_channels, num_anchors):
             cls_logits.append(nn.Conv2d(channels, num_classes * anchors, kernel_size=3, padding=1))
@@ -113,7 +113,7 @@ def __init__(self, in_channels: List[int], num_anchors: List[int], num_classes:
 
 
 class SSDRegressionHead(SSDScoringHead):
-    def __init__(self, in_channels: List[int], num_anchors: List[int]):
+    def __init__(self, in_channels: List[int], num_anchors: List[int]) -> None:
         bbox_reg = nn.ModuleList()
         for channels, anchors in zip(in_channels, num_anchors):
             bbox_reg.append(nn.Conv2d(channels, 4 * anchors, kernel_size=3, padding=1))
@@ -197,7 +197,7 @@ def __init__(
         topk_candidates: int = 400,
         positive_fraction: float = 0.25,
         **kwargs: Any,
-    ):
+    ) -> None:
         super().__init__()
         _log_api_usage_once(self)
 
@@ -462,7 +462,7 @@ def postprocess_detections(
 
 
 class SSDFeatureExtractorVGG(nn.Module):
-    def __init__(self, backbone: nn.Module, highres: bool):
+    def __init__(self, backbone: nn.Module, highres: bool) -> None:
         super().__init__()
 
         _, _, maxpool3_pos, maxpool4_pos, _ = (i for i, layer in enumerate(backbone) if isinstance(layer, nn.MaxPool2d))