Upgrading torchvision 0.12 to 0.13
Unknown0.12.0 → 0.13.1
Fetching 2 releases from GitHub…0 / 2
UnknownUpgrade verdict
DepDelta is still reading the notes for this range, so breaking changes can't be ruled out yet.
In plain terms
- DepDelta is still reading the release notes in this range, so it can't rule out breaking changes yet.
- This page updates as the notes come in.
Must act Still reading release notes
DepDelta is still reading the release notes in this range. This updates as they come in.
Moves torch 1.11.0 → 1.12.1. Your build may pick other versions.
API changes
Generated by DepDelta from published Python source. Not mixed with prose notes.
- changed:
constructor HMDB51(root: str, annotation_path: str, frames_per_clip: int, step_between_clips: int=…, frame_rate: Optional[int]=…, fold: int=…, train: bool=…, transform: Optional[Callable]=…, _precomputed_metadata: Optional[Dict[str, Any]]=…, num_workers: int=…, _video_width: int=…, _video_height: int=…, _video_min_dimension: int=…, _audio_samples: int=…) -> None → constructor HMDB51(root: str, annotation_path: str, frames_per_clip: int, step_between_clips: int=…, frame_rate: Optional[int]=…, fold: int=…, train: bool=…, transform: Optional[Callable]=…, _precomputed_metadata: Optional[Dict[str, Any]]=…, num_workers: int=…, _video_width: int=…, _video_height: int=…, _video_min_dimension: int=…, _audio_samples: int=…, output_format: str=…) -> None
- changed:
constructor UCF101(root: str, annotation_path: str, frames_per_clip: int, step_between_clips: int=…, frame_rate: Optional[int]=…, fold: int=…, train: bool=…, transform: Optional[Callable]=…, _precomputed_metadata: Optional[Dict[str, Any]]=…, num_workers: int=…, _video_width: int=…, _video_height: int=…, _video_min_dimension: int=…, _audio_samples: int=…) -> None → constructor UCF101(root: str, annotation_path: str, frames_per_clip: int, step_between_clips: int=…, frame_rate: Optional[int]=…, fold: int=…, train: bool=…, transform: Optional[Callable]=…, _precomputed_metadata: Optional[Dict[str, Any]]=…, num_workers: int=…, _video_width: int=…, _video_height: int=…, _video_min_dimension: int=…, _audio_samples: int=…, output_format: str=…) -> None
- changed:
constructor VideoClips(video_paths: List[str], clip_length_in_frames: int=…, frames_between_clips: int=…, frame_rate: Optional[int]=…, _precomputed_metadata: Optional[Dict[str, Any]]=…, num_workers: int=…, _video_width: int=…, _video_height: int=…, _video_min_dimension: int=…, _video_max_dimension: int=…, _audio_samples: int=…, _audio_channels: int=…) -> None → constructor VideoClips(video_paths: List[str], clip_length_in_frames: int=…, frames_between_clips: int=…, frame_rate: Optional[int]=…, _precomputed_metadata: Optional[Dict[str, Any]]=…, num_workers: int=…, _video_width: int=…, _video_height: int=…, _video_min_dimension: int=…, _video_max_dimension: int=…, _audio_samples: int=…, _audio_channels: int=…, output_format: str=…) -> None
- removed:
export _HAS_VIDEO_DECODER - removed:
variable _HAS_VIDEO_DECODER - added:
export VideoReader - added:
export _HAS_GPU_VIDEO_DECODER - added:
variable _HAS_GPU_VIDEO_DECODER
- changed:
function read_video(filename: str, start_pts: Union[float, Fraction]=…, end_pts: Optional[Union[float, Fraction]]=…, pts_unit: str=…) -> Tuple[torch.Tensor, torch.Tensor, Dict[str, Any]] → function read_video(filename: str, start_pts: Union[float, Fraction]=…, end_pts: Optional[Union[float, Fraction]]=…, pts_unit: str=…, output_format: str=…) -> Tuple[torch.Tensor, torch.Tensor, Dict[str, Any]]
- added:
class VideoReader - added:
constructor VideoReader(path: str, stream: str=…, num_threads: int=…, device: str=…) -> None - added:
method VideoReader.get_metadata() -> Dict[str, Any] - added:
method VideoReader.seek(time_s: float, keyframes_only: bool=…) -> VideoReader - added:
method VideoReader.set_current_stream(stream: str) -> bool
- removed:
export feature_extraction - added:
export get_weight
- changed:
function alexnet(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> AlexNet → function alexnet(*, weights: Optional[AlexNet_Weights]=…, progress: bool=…, **kwargs: Any) -> AlexNet - added:
class AlexNet_Weights(WeightsEnum) - added:
field AlexNet_Weights.DEFAULT - added:
field AlexNet_Weights.IMAGENET1K_V1
- changed:
function convnext_base(*, pretrained: bool=…, progress: bool=…, **kwargs: Any) -> ConvNeXt → function convnext_base(*, weights: Optional[ConvNeXt_Base_Weights]=…, progress: bool=…, **kwargs: Any) -> ConvNeXt - changed:
function convnext_large(*, pretrained: bool=…, progress: bool=…, **kwargs: Any) -> ConvNeXt → function convnext_large(*, weights: Optional[ConvNeXt_Large_Weights]=…, progress: bool=…, **kwargs: Any) -> ConvNeXt - changed:
function convnext_small(*, pretrained: bool=…, progress: bool=…, **kwargs: Any) -> ConvNeXt → function convnext_small(*, weights: Optional[ConvNeXt_Small_Weights]=…, progress: bool=…, **kwargs: Any) -> ConvNeXt - changed:
function convnext_tiny(*, pretrained: bool=…, progress: bool=…, **kwargs: Any) -> ConvNeXt → function convnext_tiny(*, weights: Optional[ConvNeXt_Tiny_Weights]=…, progress: bool=…, **kwargs: Any) -> ConvNeXt - added:
class ConvNeXt_Base_Weights(WeightsEnum) - added:
class ConvNeXt_Large_Weights(WeightsEnum) - added:
class ConvNeXt_Small_Weights(WeightsEnum) - added:
class ConvNeXt_Tiny_Weights(WeightsEnum) - added:
field ConvNeXt_Base_Weights.DEFAULT - added:
field ConvNeXt_Base_Weights.IMAGENET1K_V1 - added:
field ConvNeXt_Large_Weights.DEFAULT - added:
field ConvNeXt_Large_Weights.IMAGENET1K_V1 - added:
field ConvNeXt_Small_Weights.DEFAULT - added:
field ConvNeXt_Small_Weights.IMAGENET1K_V1 - added:
field ConvNeXt_Tiny_Weights.DEFAULT - added:
field ConvNeXt_Tiny_Weights.IMAGENET1K_V1
- changed:
function densenet121(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> DenseNet → function densenet121(*, weights: Optional[DenseNet121_Weights]=…, progress: bool=…, **kwargs: Any) -> DenseNet - changed:
function densenet161(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> DenseNet → function densenet161(*, weights: Optional[DenseNet161_Weights]=…, progress: bool=…, **kwargs: Any) -> DenseNet - changed:
function densenet169(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> DenseNet → function densenet169(*, weights: Optional[DenseNet169_Weights]=…, progress: bool=…, **kwargs: Any) -> DenseNet - changed:
function densenet201(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> DenseNet → function densenet201(*, weights: Optional[DenseNet201_Weights]=…, progress: bool=…, **kwargs: Any) -> DenseNet - added:
class DenseNet121_Weights(WeightsEnum) - added:
class DenseNet161_Weights(WeightsEnum) - added:
class DenseNet169_Weights(WeightsEnum) - added:
class DenseNet201_Weights(WeightsEnum) - added:
field DenseNet121_Weights.DEFAULT - added:
field DenseNet121_Weights.IMAGENET1K_V1 - added:
field DenseNet161_Weights.DEFAULT - added:
field DenseNet161_Weights.IMAGENET1K_V1 - added:
field DenseNet169_Weights.DEFAULT - added:
field DenseNet169_Weights.IMAGENET1K_V1 - added:
field DenseNet201_Weights.DEFAULT - added:
field DenseNet201_Weights.IMAGENET1K_V1
- changed:
constructor BackboneWithFPN(backbone: nn.Module, return_layers: Dict[str, str], in_channels_list: List[int], out_channels: int, extra_blocks: Optional[ExtraFPNBlock]=…) -> None → constructor BackboneWithFPN(backbone: nn.Module, return_layers: Dict[str, str], in_channels_list: List[int], out_channels: int, extra_blocks: Optional[ExtraFPNBlock]=…, norm_layer: Optional[Callable[..., nn.Module]]=…) -> None - changed:
function mobilenet_backbone(backbone_name: str, pretrained: bool, fpn: bool, norm_layer: Callable[..., nn.Module]=…, trainable_layers: int=…, returned_layers: Optional[List[int]]=…, extra_blocks: Optional[ExtraFPNBlock]=…) -> nn.Module → function mobilenet_backbone(*, backbone_name: str, weights: Optional[WeightsEnum], fpn: bool, norm_layer: Callable[..., nn.Module]=…, trainable_layers: int=…, returned_layers: Optional[List[int]]=…, extra_blocks: Optional[ExtraFPNBlock]=…) -> nn.Module - changed:
function resnet_fpn_backbone(backbone_name: str, pretrained: bool, norm_layer: Callable[..., nn.Module]=…, trainable_layers: int=…, returned_layers: Optional[List[int]]=…, extra_blocks: Optional[ExtraFPNBlock]=…) -> BackboneWithFPN → function resnet_fpn_backbone(*, backbone_name: str, weights: Optional[WeightsEnum], norm_layer: Callable[..., nn.Module]=…, trainable_layers: int=…, returned_layers: Optional[List[int]]=…, extra_blocks: Optional[ExtraFPNBlock]=…) -> BackboneWithFPN
- changed:
function fasterrcnn_mobilenet_v3_large_320_fpn(pretrained=…, progress=…, num_classes=…, pretrained_backbone=…, trainable_backbone_layers=…, **kwargs) → function fasterrcnn_mobilenet_v3_large_320_fpn(*, weights: Optional[FasterRCNN_MobileNet_V3_Large_320_FPN_Weights]=…, progress: bool=…, num_classes: Optional[int]=…, weights_backbone: Optional[MobileNet_V3_Large_Weights]=…, trainable_backbone_layers: Optional[int]=…, **kwargs: Any) -> FasterRCNN - changed:
function fasterrcnn_mobilenet_v3_large_fpn(pretrained=…, progress=…, num_classes=…, pretrained_backbone=…, trainable_backbone_layers=…, **kwargs) → function fasterrcnn_mobilenet_v3_large_fpn(*, weights: Optional[FasterRCNN_MobileNet_V3_Large_FPN_Weights]=…, progress: bool=…, num_classes: Optional[int]=…, weights_backbone: Optional[MobileNet_V3_Large_Weights]=…, trainable_backbone_layers: Optional[int]=…, **kwargs: Any) -> FasterRCNN - changed:
function fasterrcnn_resnet50_fpn(pretrained=…, progress=…, num_classes=…, pretrained_backbone=…, trainable_backbone_layers=…, **kwargs) → function fasterrcnn_resnet50_fpn(*, weights: Optional[FasterRCNN_ResNet50_FPN_Weights]=…, progress: bool=…, num_classes: Optional[int]=…, weights_backbone: Optional[ResNet50_Weights]=…, trainable_backbone_layers: Optional[int]=…, **kwargs: Any) -> FasterRCNN - added:
class FasterRCNN_MobileNet_V3_Large_320_FPN_Weights(WeightsEnum) - added:
class FasterRCNN_MobileNet_V3_Large_FPN_Weights(WeightsEnum) - added:
class FasterRCNN_ResNet50_FPN_V2_Weights(WeightsEnum) - added:
class FasterRCNN_ResNet50_FPN_Weights(WeightsEnum) - added:
field FasterRCNN_MobileNet_V3_Large_320_FPN_Weights.COCO_V1 - added:
field FasterRCNN_MobileNet_V3_Large_320_FPN_Weights.DEFAULT - added:
field FasterRCNN_MobileNet_V3_Large_FPN_Weights.COCO_V1 - added:
field FasterRCNN_MobileNet_V3_Large_FPN_Weights.DEFAULT - added:
field FasterRCNN_ResNet50_FPN_V2_Weights.COCO_V1 - added:
field FasterRCNN_ResNet50_FPN_V2_Weights.DEFAULT - added:
field FasterRCNN_ResNet50_FPN_Weights.COCO_V1 - added:
field FasterRCNN_ResNet50_FPN_Weights.DEFAULT - added:
function fasterrcnn_resnet50_fpn_v2(*, weights: Optional[FasterRCNN_ResNet50_FPN_V2_Weights]=…, progress: bool=…, num_classes: Optional[int]=…, weights_backbone: Optional[ResNet50_Weights]=…, trainable_backbone_layers: Optional[int]=…, **kwargs: Any) -> FasterRCNN
- changed:
constructor FCOS(backbone: nn.Module, num_classes: int, min_size: int=…, max_size: int=…, image_mean: Optional[List[float]]=…, image_std: Optional[List[float]]=…, anchor_generator: Optional[AnchorGenerator]=…, head: Optional[nn.Module]=…, center_sampling_radius: float=…, score_thresh: float=…, nms_thresh: float=…, detections_per_img: int=…, topk_candidates: int=…) → constructor FCOS(backbone: nn.Module, num_classes: int, min_size: int=…, max_size: int=…, image_mean: Optional[List[float]]=…, image_std: Optional[List[float]]=…, anchor_generator: Optional[AnchorGenerator]=…, head: Optional[nn.Module]=…, center_sampling_radius: float=…, score_thresh: float=…, nms_thresh: float=…, detections_per_img: int=…, topk_candidates: int=…, **kwargs) - changed:
function fcos_resnet50_fpn(pretrained: bool=…, progress: bool=…, num_classes: int=…, pretrained_backbone: bool=…, trainable_backbone_layers: Optional[int]=…, **kwargs) → function fcos_resnet50_fpn(*, weights: Optional[FCOS_ResNet50_FPN_Weights]=…, progress: bool=…, num_classes: Optional[int]=…, weights_backbone: Optional[ResNet50_Weights]=…, trainable_backbone_layers: Optional[int]=…, **kwargs: Any) -> FCOS - added:
class FCOS_ResNet50_FPN_Weights(WeightsEnum) - added:
field FCOS_ResNet50_FPN_Weights.COCO_V1 - added:
field FCOS_ResNet50_FPN_Weights.DEFAULT
- changed:
function keypointrcnn_resnet50_fpn(pretrained=…, progress=…, num_classes=…, num_keypoints=…, pretrained_backbone=…, trainable_backbone_layers=…, **kwargs) → function keypointrcnn_resnet50_fpn(*, weights: Optional[KeypointRCNN_ResNet50_FPN_Weights]=…, progress: bool=…, num_classes: Optional[int]=…, num_keypoints: Optional[int]=…, weights_backbone: Optional[ResNet50_Weights]=…, trainable_backbone_layers: Optional[int]=…, **kwargs: Any) -> KeypointRCNN - added:
class KeypointRCNN_ResNet50_FPN_Weights(WeightsEnum) - added:
field KeypointRCNN_ResNet50_FPN_Weights.COCO_LEGACY - added:
field KeypointRCNN_ResNet50_FPN_Weights.COCO_V1 - added:
field KeypointRCNN_ResNet50_FPN_Weights.DEFAULT
- changed:
function maskrcnn_resnet50_fpn(pretrained=…, progress=…, num_classes=…, pretrained_backbone=…, trainable_backbone_layers=…, **kwargs) → function maskrcnn_resnet50_fpn(*, weights: Optional[MaskRCNN_ResNet50_FPN_Weights]=…, progress: bool=…, num_classes: Optional[int]=…, weights_backbone: Optional[ResNet50_Weights]=…, trainable_backbone_layers: Optional[int]=…, **kwargs: Any) -> MaskRCNN - added:
class MaskRCNN_ResNet50_FPN_V2_Weights(WeightsEnum) - added:
class MaskRCNN_ResNet50_FPN_Weights(WeightsEnum) - added:
field MaskRCNN_ResNet50_FPN_V2_Weights.COCO_V1 - added:
field MaskRCNN_ResNet50_FPN_V2_Weights.DEFAULT - added:
field MaskRCNN_ResNet50_FPN_Weights.COCO_V1 - added:
field MaskRCNN_ResNet50_FPN_Weights.DEFAULT - added:
function maskrcnn_resnet50_fpn_v2(*, weights: Optional[MaskRCNN_ResNet50_FPN_V2_Weights]=…, progress: bool=…, num_classes: Optional[int]=…, weights_backbone: Optional[ResNet50_Weights]=…, trainable_backbone_layers: Optional[int]=…, **kwargs: Any) -> MaskRCNN
- changed:
constructor RetinaNet(backbone, num_classes, min_size=…, max_size=…, image_mean=…, image_std=…, anchor_generator=…, head=…, proposal_matcher=…, score_thresh=…, nms_thresh=…, detections_per_img=…, fg_iou_thresh=…, bg_iou_thresh=…, topk_candidates=…) → constructor RetinaNet(backbone, num_classes, min_size=…, max_size=…, image_mean=…, image_std=…, anchor_generator=…, head=…, proposal_matcher=…, score_thresh=…, nms_thresh=…, detections_per_img=…, fg_iou_thresh=…, bg_iou_thresh=…, topk_candidates=…, **kwargs) - changed:
function retinanet_resnet50_fpn(pretrained=…, progress=…, num_classes=…, pretrained_backbone=…, trainable_backbone_layers=…, **kwargs) → function retinanet_resnet50_fpn(*, weights: Optional[RetinaNet_ResNet50_FPN_Weights]=…, progress: bool=…, num_classes: Optional[int]=…, weights_backbone: Optional[ResNet50_Weights]=…, trainable_backbone_layers: Optional[int]=…, **kwargs: Any) -> RetinaNet - added:
class RetinaNet_ResNet50_FPN_V2_Weights(WeightsEnum) - added:
class RetinaNet_ResNet50_FPN_Weights(WeightsEnum) - added:
field RetinaNet_ResNet50_FPN_V2_Weights.COCO_V1 - added:
field RetinaNet_ResNet50_FPN_V2_Weights.DEFAULT - added:
field RetinaNet_ResNet50_FPN_Weights.COCO_V1 - added:
field RetinaNet_ResNet50_FPN_Weights.DEFAULT - added:
function retinanet_resnet50_fpn_v2(*, weights: Optional[RetinaNet_ResNet50_FPN_V2_Weights]=…, progress: bool=…, num_classes: Optional[int]=…, weights_backbone: Optional[ResNet50_Weights]=…, trainable_backbone_layers: Optional[int]=…, **kwargs: Any) -> RetinaNet
- changed:
constructor RPNHead(in_channels: int, num_anchors: int) -> None → constructor RPNHead(in_channels: int, num_anchors: int, conv_depth=…) -> None
- removed:
class SSD(nn.Module) - removed:
constructor SSD(backbone: nn.Module, anchor_generator: DefaultBoxGenerator, size: Tuple[int, int], num_classes: int, image_mean: Optional[List[float]]=…, image_std: Optional[List[float]]=…, head: Optional[nn.Module]=…, score_thresh: float=…, nms_thresh: float=…, detections_per_img: int=…, iou_thresh: float=…, topk_candidates: int=…, positive_fraction: float=…) - removed:
method SSD.compute_loss(targets: List[Dict[str, Tensor]], head_outputs: Dict[str, Tensor], anchors: List[Tensor], matched_idxs: List[Tensor]) -> Dict[str, Tensor] - removed:
method SSD.eager_outputs(losses: Dict[str, Tensor], detections: List[Dict[str, Tensor]]) -> Tuple[Dict[str, Tensor], List[Dict[str, Tensor]]] - removed:
method SSD.forward(images: List[Tensor], targets: Optional[List[Dict[str, Tensor]]]=…) -> Tuple[Dict[str, Tensor], List[Dict[str, Tensor]]] - removed:
method SSD.postprocess_detections(head_outputs: Dict[str, Tensor], image_anchors: List[Tensor], image_shapes: List[Tuple[int, int]]) -> List[Dict[str, Tensor]] - changed:
function ssd300_vgg16(pretrained: bool=…, progress: bool=…, num_classes: int=…, pretrained_backbone: bool=…, trainable_backbone_layers: Optional[int]=…, **kwargs: Any) → function ssd300_vgg16(*, weights: Optional[SSD300_VGG16_Weights]=…, progress: bool=…, num_classes: Optional[int]=…, weights_backbone: Optional[VGG16_Weights]=…, trainable_backbone_layers: Optional[int]=…, **kwargs: Any) -> SSD - added:
class SSD300_VGG16_Weights(WeightsEnum) - added:
field SSD300_VGG16_Weights.COCO_V1 - added:
field SSD300_VGG16_Weights.DEFAULT
- changed:
function ssdlite320_mobilenet_v3_large(pretrained: bool=…, progress: bool=…, num_classes: int=…, pretrained_backbone: bool=…, trainable_backbone_layers: Optional[int]=…, norm_layer: Optional[Callable[..., nn.Module]]=…, **kwargs: Any) → function ssdlite320_mobilenet_v3_large(*, weights: Optional[SSDLite320_MobileNet_V3_Large_Weights]=…, progress: bool=…, num_classes: Optional[int]=…, weights_backbone: Optional[MobileNet_V3_Large_Weights]=…, trainable_backbone_layers: Optional[int]=…, norm_layer: Optional[Callable[..., nn.Module]]=…, **kwargs: Any) -> SSD - added:
class SSDLite320_MobileNet_V3_Large_Weights(WeightsEnum) - added:
field SSDLite320_MobileNet_V3_Large_Weights.COCO_V1 - added:
field SSDLite320_MobileNet_V3_Large_Weights.DEFAULT
- changed:
constructor GeneralizedRCNNTransform(min_size: int, max_size: int, image_mean: List[float], image_std: List[float], size_divisible: int=…, fixed_size: Optional[Tuple[int, int]]=…) → constructor GeneralizedRCNNTransform(min_size: int, max_size: int, image_mean: List[float], image_std: List[float], size_divisible: int=…, fixed_size: Optional[Tuple[int, int]]=…, **kwargs: Any)
- changed:
constructor EfficientNet(inverted_residual_setting: List[MBConvConfig], dropout: float, stochastic_depth_prob: float=…, num_classes: int=…, block: Optional[Callable[..., nn.Module]]=…, norm_layer: Optional[Callable[..., nn.Module]]=…, **kwargs: Any) -> None → constructor EfficientNet(inverted_residual_setting: Sequence[Union[MBConvConfig, FusedMBConvConfig]], dropout: float, stochastic_depth_prob: float=…, num_classes: int=…, norm_layer: Optional[Callable[..., nn.Module]]=…, last_channel: Optional[int]=…, **kwargs: Any) -> None - changed:
function efficientnet_b0(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> EfficientNet → function efficientnet_b0(*, weights: Optional[EfficientNet_B0_Weights]=…, progress: bool=…, **kwargs: Any) -> EfficientNet - changed:
function efficientnet_b1(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> EfficientNet → function efficientnet_b1(*, weights: Optional[EfficientNet_B1_Weights]=…, progress: bool=…, **kwargs: Any) -> EfficientNet - changed:
function efficientnet_b2(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> EfficientNet → function efficientnet_b2(*, weights: Optional[EfficientNet_B2_Weights]=…, progress: bool=…, **kwargs: Any) -> EfficientNet - changed:
function efficientnet_b3(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> EfficientNet → function efficientnet_b3(*, weights: Optional[EfficientNet_B3_Weights]=…, progress: bool=…, **kwargs: Any) -> EfficientNet - changed:
function efficientnet_b4(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> EfficientNet → function efficientnet_b4(*, weights: Optional[EfficientNet_B4_Weights]=…, progress: bool=…, **kwargs: Any) -> EfficientNet - changed:
function efficientnet_b5(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> EfficientNet → function efficientnet_b5(*, weights: Optional[EfficientNet_B5_Weights]=…, progress: bool=…, **kwargs: Any) -> EfficientNet - changed:
function efficientnet_b6(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> EfficientNet → function efficientnet_b6(*, weights: Optional[EfficientNet_B6_Weights]=…, progress: bool=…, **kwargs: Any) -> EfficientNet - changed:
function efficientnet_b7(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> EfficientNet → function efficientnet_b7(*, weights: Optional[EfficientNet_B7_Weights]=…, progress: bool=…, **kwargs: Any) -> EfficientNet - added:
class EfficientNet_B0_Weights(WeightsEnum) - added:
class EfficientNet_B1_Weights(WeightsEnum) - added:
class EfficientNet_B2_Weights(WeightsEnum) - added:
class EfficientNet_B3_Weights(WeightsEnum) - added:
class EfficientNet_B4_Weights(WeightsEnum) - added:
class EfficientNet_B5_Weights(WeightsEnum) - added:
class EfficientNet_B6_Weights(WeightsEnum) - added:
class EfficientNet_B7_Weights(WeightsEnum) - added:
class EfficientNet_V2_L_Weights(WeightsEnum) - added:
class EfficientNet_V2_M_Weights(WeightsEnum) - added:
class EfficientNet_V2_S_Weights(WeightsEnum) - added:
field EfficientNet_B0_Weights.DEFAULT - added:
field EfficientNet_B0_Weights.IMAGENET1K_V1 - added:
field EfficientNet_B1_Weights.DEFAULT - added:
field EfficientNet_B1_Weights.IMAGENET1K_V1 - added:
field EfficientNet_B1_Weights.IMAGENET1K_V2 - added:
field EfficientNet_B2_Weights.DEFAULT - added:
field EfficientNet_B2_Weights.IMAGENET1K_V1 - added:
field EfficientNet_B3_Weights.DEFAULT - added:
field EfficientNet_B3_Weights.IMAGENET1K_V1 - added:
field EfficientNet_B4_Weights.DEFAULT - added:
field EfficientNet_B4_Weights.IMAGENET1K_V1 - added:
field EfficientNet_B5_Weights.DEFAULT - added:
field EfficientNet_B5_Weights.IMAGENET1K_V1 - added:
field EfficientNet_B6_Weights.DEFAULT - added:
field EfficientNet_B6_Weights.IMAGENET1K_V1 - added:
field EfficientNet_B7_Weights.DEFAULT - added:
field EfficientNet_B7_Weights.IMAGENET1K_V1 - added:
field EfficientNet_V2_L_Weights.DEFAULT - added:
field EfficientNet_V2_L_Weights.IMAGENET1K_V1 - added:
field EfficientNet_V2_M_Weights.DEFAULT - added:
field EfficientNet_V2_M_Weights.IMAGENET1K_V1 - added:
field EfficientNet_V2_S_Weights.DEFAULT - added:
field EfficientNet_V2_S_Weights.IMAGENET1K_V1 - added:
function efficientnet_v2_l(*, weights: Optional[EfficientNet_V2_L_Weights]=…, progress: bool=…, **kwargs: Any) -> EfficientNet - added:
function efficientnet_v2_m(*, weights: Optional[EfficientNet_V2_M_Weights]=…, progress: bool=…, **kwargs: Any) -> EfficientNet - added:
function efficientnet_v2_s(*, weights: Optional[EfficientNet_V2_S_Weights]=…, progress: bool=…, **kwargs: Any) -> EfficientNet
- changed:
function googlenet(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> GoogLeNet → function googlenet(*, weights: Optional[GoogLeNet_Weights]=…, progress: bool=…, **kwargs: Any) -> GoogLeNet - added:
class GoogLeNet_Weights(WeightsEnum) - added:
field GoogLeNet_Weights.DEFAULT - added:
field GoogLeNet_Weights.IMAGENET1K_V1
- changed:
function inception_v3(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> Inception3 → function inception_v3(*, weights: Optional[Inception_V3_Weights]=…, progress: bool=…, **kwargs: Any) -> Inception3 - added:
class Inception_V3_Weights(WeightsEnum) - added:
field Inception_V3_Weights.DEFAULT - added:
field Inception_V3_Weights.IMAGENET1K_V1
- changed:
function mnasnet0_5(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> MNASNet → function mnasnet0_5(*, weights: Optional[MNASNet0_5_Weights]=…, progress: bool=…, **kwargs: Any) -> MNASNet - changed:
function mnasnet0_75(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> MNASNet → function mnasnet0_75(*, weights: Optional[MNASNet0_75_Weights]=…, progress: bool=…, **kwargs: Any) -> MNASNet - changed:
function mnasnet1_0(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> MNASNet → function mnasnet1_0(*, weights: Optional[MNASNet1_0_Weights]=…, progress: bool=…, **kwargs: Any) -> MNASNet - changed:
function mnasnet1_3(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> MNASNet → function mnasnet1_3(*, weights: Optional[MNASNet1_3_Weights]=…, progress: bool=…, **kwargs: Any) -> MNASNet - added:
class MNASNet0_5_Weights(WeightsEnum) - added:
class MNASNet0_75_Weights(WeightsEnum) - added:
class MNASNet1_0_Weights(WeightsEnum) - added:
class MNASNet1_3_Weights(WeightsEnum) - added:
field MNASNet0_5_Weights.DEFAULT - added:
field MNASNet0_5_Weights.IMAGENET1K_V1 - added:
field MNASNet0_75_Weights.DEFAULT - added:
field MNASNet0_75_Weights.IMAGENET1K_V1 - added:
field MNASNet1_0_Weights.DEFAULT - added:
field MNASNet1_0_Weights.IMAGENET1K_V1 - added:
field MNASNet1_3_Weights.DEFAULT - added:
field MNASNet1_3_Weights.IMAGENET1K_V1
- changed:
function mobilenet_v2(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> MobileNetV2 → function mobilenet_v2(*, weights: Optional[MobileNet_V2_Weights]=…, progress: bool=…, **kwargs: Any) -> MobileNetV2 - added:
class MobileNet_V2_Weights(WeightsEnum) - added:
field MobileNet_V2_Weights.DEFAULT - added:
field MobileNet_V2_Weights.IMAGENET1K_V1 - added:
field MobileNet_V2_Weights.IMAGENET1K_V2
- changed:
function mobilenet_v3_large(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> MobileNetV3 → function mobilenet_v3_large(*, weights: Optional[MobileNet_V3_Large_Weights]=…, progress: bool=…, **kwargs: Any) -> MobileNetV3 - changed:
function mobilenet_v3_small(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> MobileNetV3 → function mobilenet_v3_small(*, weights: Optional[MobileNet_V3_Small_Weights]=…, progress: bool=…, **kwargs: Any) -> MobileNetV3 - added:
class MobileNet_V3_Large_Weights(WeightsEnum) - added:
class MobileNet_V3_Small_Weights(WeightsEnum) - added:
field MobileNet_V3_Large_Weights.DEFAULT - added:
field MobileNet_V3_Large_Weights.IMAGENET1K_V1 - added:
field MobileNet_V3_Large_Weights.IMAGENET1K_V2 - added:
field MobileNet_V3_Small_Weights.DEFAULT - added:
field MobileNet_V3_Small_Weights.IMAGENET1K_V1
- removed:
export RAFT - removed:
export raft_large - removed:
export raft_small
- changed:
function raft_large(*, pretrained=…, progress=…, **kwargs) → function raft_large(*, weights: Optional[Raft_Large_Weights]=…, progress=…, **kwargs) -> RAFT - changed:
function raft_small(*, pretrained=…, progress=…, **kwargs) → function raft_small(*, weights: Optional[Raft_Small_Weights]=…, progress=…, **kwargs) -> RAFT - added:
class Raft_Large_Weights(WeightsEnum) - added:
class Raft_Small_Weights(WeightsEnum) - added:
field Raft_Large_Weights.C_T_SKHT_K_V1 - added:
field Raft_Large_Weights.C_T_SKHT_K_V2 - added:
field Raft_Large_Weights.C_T_SKHT_V1 - added:
field Raft_Large_Weights.C_T_SKHT_V2 - added:
field Raft_Large_Weights.C_T_V1 - added:
field Raft_Large_Weights.C_T_V2 - added:
field Raft_Large_Weights.DEFAULT - added:
field Raft_Small_Weights.C_T_V1 - added:
field Raft_Small_Weights.C_T_V2 - added:
field Raft_Small_Weights.DEFAULT
- changed:
function googlenet(pretrained: bool=…, progress: bool=…, quantize: bool=…, **kwargs: Any) -> QuantizableGoogLeNet → function googlenet(*, weights: Optional[Union[GoogLeNet_QuantizedWeights, GoogLeNet_Weights]]=…, progress: bool=…, quantize: bool=…, **kwargs: Any) -> QuantizableGoogLeNet - added:
class GoogLeNet_QuantizedWeights(WeightsEnum) - added:
field GoogLeNet_QuantizedWeights.DEFAULT - added:
field GoogLeNet_QuantizedWeights.IMAGENET1K_FBGEMM_V1
- changed:
function inception_v3(pretrained: bool=…, progress: bool=…, quantize: bool=…, **kwargs: Any) -> QuantizableInception3 → function inception_v3(*, weights: Optional[Union[Inception_V3_QuantizedWeights, Inception_V3_Weights]]=…, progress: bool=…, quantize: bool=…, **kwargs: Any) -> QuantizableInception3 - added:
class Inception_V3_QuantizedWeights(WeightsEnum) - added:
field Inception_V3_QuantizedWeights.DEFAULT - added:
field Inception_V3_QuantizedWeights.IMAGENET1K_FBGEMM_V1
- changed:
function mobilenet_v2(pretrained: bool=…, progress: bool=…, quantize: bool=…, **kwargs: Any) -> QuantizableMobileNetV2 → function mobilenet_v2(*, weights: Optional[Union[MobileNet_V2_QuantizedWeights, MobileNet_V2_Weights]]=…, progress: bool=…, quantize: bool=…, **kwargs: Any) -> QuantizableMobileNetV2 - added:
class MobileNet_V2_QuantizedWeights(WeightsEnum) - added:
field MobileNet_V2_QuantizedWeights.DEFAULT - added:
field MobileNet_V2_QuantizedWeights.IMAGENET1K_QNNPACK_V1
- changed:
function mobilenet_v3_large(pretrained: bool=…, progress: bool=…, quantize: bool=…, **kwargs: Any) -> QuantizableMobileNetV3 → function mobilenet_v3_large(*, weights: Optional[Union[MobileNet_V3_Large_QuantizedWeights, MobileNet_V3_Large_Weights]]=…, progress: bool=…, quantize: bool=…, **kwargs: Any) -> QuantizableMobileNetV3 - added:
class MobileNet_V3_Large_QuantizedWeights(WeightsEnum) - added:
field MobileNet_V3_Large_QuantizedWeights.DEFAULT - added:
field MobileNet_V3_Large_QuantizedWeights.IMAGENET1K_QNNPACK_V1
- changed:
function resnet18(pretrained: bool=…, progress: bool=…, quantize: bool=…, **kwargs: Any) -> QuantizableResNet → function resnet18(*, weights: Optional[Union[ResNet18_QuantizedWeights, ResNet18_Weights]]=…, progress: bool=…, quantize: bool=…, **kwargs: Any) -> QuantizableResNet - changed:
function resnet50(pretrained: bool=…, progress: bool=…, quantize: bool=…, **kwargs: Any) -> QuantizableResNet → function resnet50(*, weights: Optional[Union[ResNet50_QuantizedWeights, ResNet50_Weights]]=…, progress: bool=…, quantize: bool=…, **kwargs: Any) -> QuantizableResNet - changed:
function resnext101_32x8d(pretrained: bool=…, progress: bool=…, quantize: bool=…, **kwargs: Any) -> QuantizableResNet → function resnext101_32x8d(*, weights: Optional[Union[ResNeXt101_32X8D_QuantizedWeights, ResNeXt101_32X8D_Weights]]=…, progress: bool=…, quantize: bool=…, **kwargs: Any) -> QuantizableResNet - added:
class ResNeXt101_32X8D_QuantizedWeights(WeightsEnum) - added:
class ResNeXt101_64X4D_QuantizedWeights(WeightsEnum) - added:
class ResNet18_QuantizedWeights(WeightsEnum) - added:
class ResNet50_QuantizedWeights(WeightsEnum) - added:
field ResNeXt101_32X8D_QuantizedWeights.DEFAULT - added:
field ResNeXt101_32X8D_QuantizedWeights.IMAGENET1K_FBGEMM_V1 - added:
field ResNeXt101_32X8D_QuantizedWeights.IMAGENET1K_FBGEMM_V2 - added:
field ResNeXt101_64X4D_QuantizedWeights.DEFAULT - added:
field ResNeXt101_64X4D_QuantizedWeights.IMAGENET1K_FBGEMM_V1 - added:
field ResNet18_QuantizedWeights.DEFAULT - added:
field ResNet18_QuantizedWeights.IMAGENET1K_FBGEMM_V1 - added:
field ResNet50_QuantizedWeights.DEFAULT - added:
field ResNet50_QuantizedWeights.IMAGENET1K_FBGEMM_V1 - added:
field ResNet50_QuantizedWeights.IMAGENET1K_FBGEMM_V2 - added:
function resnext101_64x4d(*, weights: Optional[Union[ResNeXt101_64X4D_QuantizedWeights, ResNeXt101_64X4D_Weights]]=…, progress: bool=…, quantize: bool=…, **kwargs: Any) -> QuantizableResNet
- changed:
function shufflenet_v2_x0_5(pretrained: bool=…, progress: bool=…, quantize: bool=…, **kwargs: Any) -> QuantizableShuffleNetV2 → function shufflenet_v2_x0_5(*, weights: Optional[Union[ShuffleNet_V2_X0_5_QuantizedWeights, ShuffleNet_V2_X0_5_Weights]]=…, progress: bool=…, quantize: bool=…, **kwargs: Any) -> QuantizableShuffleNetV2 - changed:
function shufflenet_v2_x1_0(pretrained: bool=…, progress: bool=…, quantize: bool=…, **kwargs: Any) -> QuantizableShuffleNetV2 → function shufflenet_v2_x1_0(*, weights: Optional[Union[ShuffleNet_V2_X1_0_QuantizedWeights, ShuffleNet_V2_X1_0_Weights]]=…, progress: bool=…, quantize: bool=…, **kwargs: Any) -> QuantizableShuffleNetV2 - added:
class ShuffleNet_V2_X0_5_QuantizedWeights(WeightsEnum) - added:
class ShuffleNet_V2_X1_0_QuantizedWeights(WeightsEnum) - added:
class ShuffleNet_V2_X1_5_QuantizedWeights(WeightsEnum) - added:
class ShuffleNet_V2_X2_0_QuantizedWeights(WeightsEnum) - added:
field ShuffleNet_V2_X0_5_QuantizedWeights.DEFAULT - added:
field ShuffleNet_V2_X0_5_QuantizedWeights.IMAGENET1K_FBGEMM_V1 - added:
field ShuffleNet_V2_X1_0_QuantizedWeights.DEFAULT - added:
field ShuffleNet_V2_X1_0_QuantizedWeights.IMAGENET1K_FBGEMM_V1 - added:
field ShuffleNet_V2_X1_5_QuantizedWeights.DEFAULT - added:
field ShuffleNet_V2_X1_5_QuantizedWeights.IMAGENET1K_FBGEMM_V1 - added:
field ShuffleNet_V2_X2_0_QuantizedWeights.DEFAULT - added:
field ShuffleNet_V2_X2_0_QuantizedWeights.IMAGENET1K_FBGEMM_V1 - added:
function shufflenet_v2_x1_5(*, weights: Optional[Union[ShuffleNet_V2_X1_5_QuantizedWeights, ShuffleNet_V2_X1_5_Weights]]=…, progress: bool=…, quantize: bool=…, **kwargs: Any) -> QuantizableShuffleNetV2 - added:
function shufflenet_v2_x2_0(*, weights: Optional[Union[ShuffleNet_V2_X2_0_QuantizedWeights, ShuffleNet_V2_X2_0_Weights]]=…, progress: bool=…, quantize: bool=…, **kwargs: Any) -> QuantizableShuffleNetV2
- changed:
function regnet_x_16gf(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> RegNet → function regnet_x_16gf(*, weights: Optional[RegNet_X_16GF_Weights]=…, progress: bool=…, **kwargs: Any) -> RegNet - changed:
function regnet_x_1_6gf(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> RegNet → function regnet_x_1_6gf(*, weights: Optional[RegNet_X_1_6GF_Weights]=…, progress: bool=…, **kwargs: Any) -> RegNet - changed:
function regnet_x_32gf(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> RegNet → function regnet_x_32gf(*, weights: Optional[RegNet_X_32GF_Weights]=…, progress: bool=…, **kwargs: Any) -> RegNet - changed:
function regnet_x_3_2gf(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> RegNet → function regnet_x_3_2gf(*, weights: Optional[RegNet_X_3_2GF_Weights]=…, progress: bool=…, **kwargs: Any) -> RegNet - changed:
function regnet_x_400mf(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> RegNet → function regnet_x_400mf(*, weights: Optional[RegNet_X_400MF_Weights]=…, progress: bool=…, **kwargs: Any) -> RegNet - changed:
function regnet_x_800mf(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> RegNet → function regnet_x_800mf(*, weights: Optional[RegNet_X_800MF_Weights]=…, progress: bool=…, **kwargs: Any) -> RegNet - changed:
function regnet_x_8gf(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> RegNet → function regnet_x_8gf(*, weights: Optional[RegNet_X_8GF_Weights]=…, progress: bool=…, **kwargs: Any) -> RegNet - changed:
function regnet_y_128gf(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> RegNet → function regnet_y_128gf(*, weights: Optional[RegNet_Y_128GF_Weights]=…, progress: bool=…, **kwargs: Any) -> RegNet - changed:
function regnet_y_16gf(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> RegNet → function regnet_y_16gf(*, weights: Optional[RegNet_Y_16GF_Weights]=…, progress: bool=…, **kwargs: Any) -> RegNet - changed:
function regnet_y_1_6gf(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> RegNet → function regnet_y_1_6gf(*, weights: Optional[RegNet_Y_1_6GF_Weights]=…, progress: bool=…, **kwargs: Any) -> RegNet - changed:
function regnet_y_32gf(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> RegNet → function regnet_y_32gf(*, weights: Optional[RegNet_Y_32GF_Weights]=…, progress: bool=…, **kwargs: Any) -> RegNet - changed:
function regnet_y_3_2gf(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> RegNet → function regnet_y_3_2gf(*, weights: Optional[RegNet_Y_3_2GF_Weights]=…, progress: bool=…, **kwargs: Any) -> RegNet - changed:
function regnet_y_400mf(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> RegNet → function regnet_y_400mf(*, weights: Optional[RegNet_Y_400MF_Weights]=…, progress: bool=…, **kwargs: Any) -> RegNet - changed:
function regnet_y_800mf(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> RegNet → function regnet_y_800mf(*, weights: Optional[RegNet_Y_800MF_Weights]=…, progress: bool=…, **kwargs: Any) -> RegNet - changed:
function regnet_y_8gf(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> RegNet → function regnet_y_8gf(*, weights: Optional[RegNet_Y_8GF_Weights]=…, progress: bool=…, **kwargs: Any) -> RegNet - added:
class RegNet_X_16GF_Weights(WeightsEnum) - added:
class RegNet_X_1_6GF_Weights(WeightsEnum) - added:
class RegNet_X_32GF_Weights(WeightsEnum) - added:
class RegNet_X_3_2GF_Weights(WeightsEnum) - added:
class RegNet_X_400MF_Weights(WeightsEnum) - added:
class RegNet_X_800MF_Weights(WeightsEnum) - added:
class RegNet_X_8GF_Weights(WeightsEnum) - added:
class RegNet_Y_128GF_Weights(WeightsEnum) - added:
class RegNet_Y_16GF_Weights(WeightsEnum) - added:
class RegNet_Y_1_6GF_Weights(WeightsEnum) - added:
class RegNet_Y_32GF_Weights(WeightsEnum) - added:
class RegNet_Y_3_2GF_Weights(WeightsEnum) - added:
class RegNet_Y_400MF_Weights(WeightsEnum) - added:
class RegNet_Y_800MF_Weights(WeightsEnum) - added:
class RegNet_Y_8GF_Weights(WeightsEnum) - added:
field RegNet_X_16GF_Weights.DEFAULT - added:
field RegNet_X_16GF_Weights.IMAGENET1K_V1 - added:
field RegNet_X_16GF_Weights.IMAGENET1K_V2 - added:
field RegNet_X_1_6GF_Weights.DEFAULT - added:
field RegNet_X_1_6GF_Weights.IMAGENET1K_V1 - added:
field RegNet_X_1_6GF_Weights.IMAGENET1K_V2 - added:
field RegNet_X_32GF_Weights.DEFAULT - added:
field RegNet_X_32GF_Weights.IMAGENET1K_V1 - added:
field RegNet_X_32GF_Weights.IMAGENET1K_V2 - added:
field RegNet_X_3_2GF_Weights.DEFAULT - added:
field RegNet_X_3_2GF_Weights.IMAGENET1K_V1 - added:
field RegNet_X_3_2GF_Weights.IMAGENET1K_V2 - added:
field RegNet_X_400MF_Weights.DEFAULT - added:
field RegNet_X_400MF_Weights.IMAGENET1K_V1 - added:
field RegNet_X_400MF_Weights.IMAGENET1K_V2 - added:
field RegNet_X_800MF_Weights.DEFAULT - added:
field RegNet_X_800MF_Weights.IMAGENET1K_V1 - added:
field RegNet_X_800MF_Weights.IMAGENET1K_V2 - added:
field RegNet_X_8GF_Weights.DEFAULT - added:
field RegNet_X_8GF_Weights.IMAGENET1K_V1 - added:
field RegNet_X_8GF_Weights.IMAGENET1K_V2 - added:
field RegNet_Y_128GF_Weights.DEFAULT - added:
field RegNet_Y_128GF_Weights.IMAGENET1K_SWAG_E2E_V1 - added:
field RegNet_Y_128GF_Weights.IMAGENET1K_SWAG_LINEAR_V1 - added:
field RegNet_Y_16GF_Weights.DEFAULT - added:
field RegNet_Y_16GF_Weights.IMAGENET1K_SWAG_E2E_V1 - added:
field RegNet_Y_16GF_Weights.IMAGENET1K_SWAG_LINEAR_V1 - added:
field RegNet_Y_16GF_Weights.IMAGENET1K_V1 - added:
field RegNet_Y_16GF_Weights.IMAGENET1K_V2 - added:
field RegNet_Y_1_6GF_Weights.DEFAULT - added:
field RegNet_Y_1_6GF_Weights.IMAGENET1K_V1 - added:
field RegNet_Y_1_6GF_Weights.IMAGENET1K_V2 - added:
field RegNet_Y_32GF_Weights.DEFAULT - added:
field RegNet_Y_32GF_Weights.IMAGENET1K_SWAG_E2E_V1 - added:
field RegNet_Y_32GF_Weights.IMAGENET1K_SWAG_LINEAR_V1 - added:
field RegNet_Y_32GF_Weights.IMAGENET1K_V1 - added:
field RegNet_Y_32GF_Weights.IMAGENET1K_V2 - added:
field RegNet_Y_3_2GF_Weights.DEFAULT - added:
field RegNet_Y_3_2GF_Weights.IMAGENET1K_V1 - added:
field RegNet_Y_3_2GF_Weights.IMAGENET1K_V2 - added:
field RegNet_Y_400MF_Weights.DEFAULT - added:
field RegNet_Y_400MF_Weights.IMAGENET1K_V1 - added:
field RegNet_Y_400MF_Weights.IMAGENET1K_V2 - added:
field RegNet_Y_800MF_Weights.DEFAULT - added:
field RegNet_Y_800MF_Weights.IMAGENET1K_V1 - added:
field RegNet_Y_800MF_Weights.IMAGENET1K_V2 - added:
field RegNet_Y_8GF_Weights.DEFAULT - added:
field RegNet_Y_8GF_Weights.IMAGENET1K_V1 - added:
field RegNet_Y_8GF_Weights.IMAGENET1K_V2
- changed:
function resnet101(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> ResNet → function resnet101(*, weights: Optional[ResNet101_Weights]=…, progress: bool=…, **kwargs: Any) -> ResNet - changed:
function resnet152(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> ResNet → function resnet152(*, weights: Optional[ResNet152_Weights]=…, progress: bool=…, **kwargs: Any) -> ResNet - changed:
function resnet18(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> ResNet → function resnet18(*, weights: Optional[ResNet18_Weights]=…, progress: bool=…, **kwargs: Any) -> ResNet - changed:
function resnet34(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> ResNet → function resnet34(*, weights: Optional[ResNet34_Weights]=…, progress: bool=…, **kwargs: Any) -> ResNet - changed:
function resnet50(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> ResNet → function resnet50(*, weights: Optional[ResNet50_Weights]=…, progress: bool=…, **kwargs: Any) -> ResNet - changed:
function resnext101_32x8d(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> ResNet → function resnext101_32x8d(*, weights: Optional[ResNeXt101_32X8D_Weights]=…, progress: bool=…, **kwargs: Any) -> ResNet - changed:
function resnext50_32x4d(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> ResNet → function resnext50_32x4d(*, weights: Optional[ResNeXt50_32X4D_Weights]=…, progress: bool=…, **kwargs: Any) -> ResNet - changed:
function wide_resnet101_2(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> ResNet → function wide_resnet101_2(*, weights: Optional[Wide_ResNet101_2_Weights]=…, progress: bool=…, **kwargs: Any) -> ResNet - changed:
function wide_resnet50_2(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> ResNet → function wide_resnet50_2(*, weights: Optional[Wide_ResNet50_2_Weights]=…, progress: bool=…, **kwargs: Any) -> ResNet - added:
class ResNeXt101_32X8D_Weights(WeightsEnum) - added:
class ResNeXt101_64X4D_Weights(WeightsEnum) - added:
class ResNeXt50_32X4D_Weights(WeightsEnum) - added:
class ResNet101_Weights(WeightsEnum) - added:
class ResNet152_Weights(WeightsEnum) - added:
class ResNet18_Weights(WeightsEnum) - added:
class ResNet34_Weights(WeightsEnum) - added:
class ResNet50_Weights(WeightsEnum) - added:
class Wide_ResNet101_2_Weights(WeightsEnum) - added:
class Wide_ResNet50_2_Weights(WeightsEnum) - added:
field ResNeXt101_32X8D_Weights.DEFAULT - added:
field ResNeXt101_32X8D_Weights.IMAGENET1K_V1 - added:
field ResNeXt101_32X8D_Weights.IMAGENET1K_V2 - added:
field ResNeXt101_64X4D_Weights.DEFAULT - added:
field ResNeXt101_64X4D_Weights.IMAGENET1K_V1 - added:
field ResNeXt50_32X4D_Weights.DEFAULT - added:
field ResNeXt50_32X4D_Weights.IMAGENET1K_V1 - added:
field ResNeXt50_32X4D_Weights.IMAGENET1K_V2 - added:
field ResNet101_Weights.DEFAULT - added:
field ResNet101_Weights.IMAGENET1K_V1 - added:
field ResNet101_Weights.IMAGENET1K_V2 - added:
field ResNet152_Weights.DEFAULT - added:
field ResNet152_Weights.IMAGENET1K_V1 - added:
field ResNet152_Weights.IMAGENET1K_V2 - added:
field ResNet18_Weights.DEFAULT - added:
field ResNet18_Weights.IMAGENET1K_V1 - added:
field ResNet34_Weights.DEFAULT - added:
field ResNet34_Weights.IMAGENET1K_V1 - added:
field ResNet50_Weights.DEFAULT - added:
field ResNet50_Weights.IMAGENET1K_V1 - added:
field ResNet50_Weights.IMAGENET1K_V2 - added:
field Wide_ResNet101_2_Weights.DEFAULT - added:
field Wide_ResNet101_2_Weights.IMAGENET1K_V1 - added:
field Wide_ResNet101_2_Weights.IMAGENET1K_V2 - added:
field Wide_ResNet50_2_Weights.DEFAULT - added:
field Wide_ResNet50_2_Weights.IMAGENET1K_V1 - added:
field Wide_ResNet50_2_Weights.IMAGENET1K_V2 - added:
function resnext101_64x4d(*, weights: Optional[ResNeXt101_64X4D_Weights]=…, progress: bool=…, **kwargs: Any) -> ResNet
- changed:
function deeplabv3_mobilenet_v3_large(pretrained: bool=…, progress: bool=…, num_classes: int=…, aux_loss: Optional[bool]=…, pretrained_backbone: bool=…) -> DeepLabV3 → function deeplabv3_mobilenet_v3_large(*, weights: Optional[DeepLabV3_MobileNet_V3_Large_Weights]=…, progress: bool=…, num_classes: Optional[int]=…, aux_loss: Optional[bool]=…, weights_backbone: Optional[MobileNet_V3_Large_Weights]=…, **kwargs: Any) -> DeepLabV3 - changed:
function deeplabv3_resnet101(pretrained: bool=…, progress: bool=…, num_classes: int=…, aux_loss: Optional[bool]=…, pretrained_backbone: bool=…) -> DeepLabV3 → function deeplabv3_resnet101(*, weights: Optional[DeepLabV3_ResNet101_Weights]=…, progress: bool=…, num_classes: Optional[int]=…, aux_loss: Optional[bool]=…, weights_backbone: Optional[ResNet101_Weights]=…, **kwargs: Any) -> DeepLabV3 - changed:
function deeplabv3_resnet50(pretrained: bool=…, progress: bool=…, num_classes: int=…, aux_loss: Optional[bool]=…, pretrained_backbone: bool=…) -> DeepLabV3 → function deeplabv3_resnet50(*, weights: Optional[DeepLabV3_ResNet50_Weights]=…, progress: bool=…, num_classes: Optional[int]=…, aux_loss: Optional[bool]=…, weights_backbone: Optional[ResNet50_Weights]=…, **kwargs: Any) -> DeepLabV3 - added:
class DeepLabV3_MobileNet_V3_Large_Weights(WeightsEnum) - added:
class DeepLabV3_ResNet101_Weights(WeightsEnum) - added:
class DeepLabV3_ResNet50_Weights(WeightsEnum) - added:
field DeepLabV3_MobileNet_V3_Large_Weights.COCO_WITH_VOC_LABELS_V1 - added:
field DeepLabV3_MobileNet_V3_Large_Weights.DEFAULT - added:
field DeepLabV3_ResNet101_Weights.COCO_WITH_VOC_LABELS_V1 - added:
field DeepLabV3_ResNet101_Weights.DEFAULT - added:
field DeepLabV3_ResNet50_Weights.COCO_WITH_VOC_LABELS_V1 - added:
field DeepLabV3_ResNet50_Weights.DEFAULT
- changed:
function fcn_resnet101(pretrained: bool=…, progress: bool=…, num_classes: int=…, aux_loss: Optional[bool]=…, pretrained_backbone: bool=…) -> FCN → function fcn_resnet101(*, weights: Optional[FCN_ResNet101_Weights]=…, progress: bool=…, num_classes: Optional[int]=…, aux_loss: Optional[bool]=…, weights_backbone: Optional[ResNet101_Weights]=…, **kwargs: Any) -> FCN - changed:
function fcn_resnet50(pretrained: bool=…, progress: bool=…, num_classes: int=…, aux_loss: Optional[bool]=…, pretrained_backbone: bool=…) -> FCN → function fcn_resnet50(*, weights: Optional[FCN_ResNet50_Weights]=…, progress: bool=…, num_classes: Optional[int]=…, aux_loss: Optional[bool]=…, weights_backbone: Optional[ResNet50_Weights]=…, **kwargs: Any) -> FCN - added:
class FCN_ResNet101_Weights(WeightsEnum) - added:
class FCN_ResNet50_Weights(WeightsEnum) - added:
field FCN_ResNet101_Weights.COCO_WITH_VOC_LABELS_V1 - added:
field FCN_ResNet101_Weights.DEFAULT - added:
field FCN_ResNet50_Weights.COCO_WITH_VOC_LABELS_V1 - added:
field FCN_ResNet50_Weights.DEFAULT
- changed:
function lraspp_mobilenet_v3_large(pretrained: bool=…, progress: bool=…, num_classes: int=…, pretrained_backbone: bool=…, **kwargs: Any) -> LRASPP → function lraspp_mobilenet_v3_large(*, weights: Optional[LRASPP_MobileNet_V3_Large_Weights]=…, progress: bool=…, num_classes: Optional[int]=…, weights_backbone: Optional[MobileNet_V3_Large_Weights]=…, **kwargs: Any) -> LRASPP - added:
class LRASPP_MobileNet_V3_Large_Weights(WeightsEnum) - added:
field LRASPP_MobileNet_V3_Large_Weights.COCO_WITH_VOC_LABELS_V1 - added:
field LRASPP_MobileNet_V3_Large_Weights.DEFAULT
- changed:
function shufflenet_v2_x0_5(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> ShuffleNetV2 → function shufflenet_v2_x0_5(*, weights: Optional[ShuffleNet_V2_X0_5_Weights]=…, progress: bool=…, **kwargs: Any) -> ShuffleNetV2 - changed:
function shufflenet_v2_x1_0(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> ShuffleNetV2 → function shufflenet_v2_x1_0(*, weights: Optional[ShuffleNet_V2_X1_0_Weights]=…, progress: bool=…, **kwargs: Any) -> ShuffleNetV2 - changed:
function shufflenet_v2_x1_5(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> ShuffleNetV2 → function shufflenet_v2_x1_5(*, weights: Optional[ShuffleNet_V2_X1_5_Weights]=…, progress: bool=…, **kwargs: Any) -> ShuffleNetV2 - changed:
function shufflenet_v2_x2_0(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> ShuffleNetV2 → function shufflenet_v2_x2_0(*, weights: Optional[ShuffleNet_V2_X2_0_Weights]=…, progress: bool=…, **kwargs: Any) -> ShuffleNetV2 - added:
class ShuffleNet_V2_X0_5_Weights(WeightsEnum) - added:
class ShuffleNet_V2_X1_0_Weights(WeightsEnum) - added:
class ShuffleNet_V2_X1_5_Weights(WeightsEnum) - added:
class ShuffleNet_V2_X2_0_Weights(WeightsEnum) - added:
field ShuffleNet_V2_X0_5_Weights.DEFAULT - added:
field ShuffleNet_V2_X0_5_Weights.IMAGENET1K_V1 - added:
field ShuffleNet_V2_X1_0_Weights.DEFAULT - added:
field ShuffleNet_V2_X1_0_Weights.IMAGENET1K_V1 - added:
field ShuffleNet_V2_X1_5_Weights.DEFAULT - added:
field ShuffleNet_V2_X1_5_Weights.IMAGENET1K_V1 - added:
field ShuffleNet_V2_X2_0_Weights.DEFAULT - added:
field ShuffleNet_V2_X2_0_Weights.IMAGENET1K_V1
- changed:
function squeezenet1_0(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> SqueezeNet → function squeezenet1_0(*, weights: Optional[SqueezeNet1_0_Weights]=…, progress: bool=…, **kwargs: Any) -> SqueezeNet - changed:
function squeezenet1_1(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> SqueezeNet → function squeezenet1_1(*, weights: Optional[SqueezeNet1_1_Weights]=…, progress: bool=…, **kwargs: Any) -> SqueezeNet - added:
class SqueezeNet1_0_Weights(WeightsEnum) - added:
class SqueezeNet1_1_Weights(WeightsEnum) - added:
field SqueezeNet1_0_Weights.DEFAULT - added:
field SqueezeNet1_0_Weights.IMAGENET1K_V1 - added:
field SqueezeNet1_1_Weights.DEFAULT - added:
field SqueezeNet1_1_Weights.IMAGENET1K_V1
- added:
class SwinTransformer(nn.Module) - added:
class Swin_B_Weights(WeightsEnum) - added:
class Swin_S_Weights(WeightsEnum) - added:
class Swin_T_Weights(WeightsEnum) - added:
constructor SwinTransformer(patch_size: List[int], embed_dim: int, depths: List[int], num_heads: List[int], window_size: List[int], mlp_ratio: float=…, dropout: float=…, attention_dropout: float=…, stochastic_depth_prob: float=…, num_classes: int=…, norm_layer: Optional[Callable[..., nn.Module]]=…, block: Optional[Callable[..., nn.Module]]=…) - added:
field Swin_B_Weights.DEFAULT - added:
field Swin_B_Weights.IMAGENET1K_V1 - added:
field Swin_S_Weights.DEFAULT - added:
field Swin_S_Weights.IMAGENET1K_V1 - added:
field Swin_T_Weights.DEFAULT - added:
field Swin_T_Weights.IMAGENET1K_V1 - added:
function swin_b(*, weights: Optional[Swin_B_Weights]=…, progress: bool=…, **kwargs: Any) -> SwinTransformer - added:
function swin_s(*, weights: Optional[Swin_S_Weights]=…, progress: bool=…, **kwargs: Any) -> SwinTransformer - added:
function swin_t(*, weights: Optional[Swin_T_Weights]=…, progress: bool=…, **kwargs: Any) -> SwinTransformer - added:
method SwinTransformer.forward(x)
- changed:
function vgg11(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> VGG → function vgg11(*, weights: Optional[VGG11_Weights]=…, progress: bool=…, **kwargs: Any) -> VGG - changed:
function vgg11_bn(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> VGG → function vgg11_bn(*, weights: Optional[VGG11_BN_Weights]=…, progress: bool=…, **kwargs: Any) -> VGG - changed:
function vgg13(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> VGG → function vgg13(*, weights: Optional[VGG13_Weights]=…, progress: bool=…, **kwargs: Any) -> VGG - changed:
function vgg13_bn(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> VGG → function vgg13_bn(*, weights: Optional[VGG13_BN_Weights]=…, progress: bool=…, **kwargs: Any) -> VGG - changed:
function vgg16(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> VGG → function vgg16(*, weights: Optional[VGG16_Weights]=…, progress: bool=…, **kwargs: Any) -> VGG - changed:
function vgg16_bn(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> VGG → function vgg16_bn(*, weights: Optional[VGG16_BN_Weights]=…, progress: bool=…, **kwargs: Any) -> VGG - changed:
function vgg19(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> VGG → function vgg19(*, weights: Optional[VGG19_Weights]=…, progress: bool=…, **kwargs: Any) -> VGG - changed:
function vgg19_bn(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> VGG → function vgg19_bn(*, weights: Optional[VGG19_BN_Weights]=…, progress: bool=…, **kwargs: Any) -> VGG - added:
class VGG11_BN_Weights(WeightsEnum) - added:
class VGG11_Weights(WeightsEnum) - added:
class VGG13_BN_Weights(WeightsEnum) - added:
class VGG13_Weights(WeightsEnum) - added:
class VGG16_BN_Weights(WeightsEnum) - added:
class VGG16_Weights(WeightsEnum) - added:
class VGG19_BN_Weights(WeightsEnum) - added:
class VGG19_Weights(WeightsEnum) - added:
field VGG11_BN_Weights.DEFAULT - added:
field VGG11_BN_Weights.IMAGENET1K_V1 - added:
field VGG11_Weights.DEFAULT - added:
field VGG11_Weights.IMAGENET1K_V1 - added:
field VGG13_BN_Weights.DEFAULT - added:
field VGG13_BN_Weights.IMAGENET1K_V1 - added:
field VGG13_Weights.DEFAULT - added:
field VGG13_Weights.IMAGENET1K_V1 - added:
field VGG16_BN_Weights.DEFAULT - added:
field VGG16_BN_Weights.IMAGENET1K_V1 - added:
field VGG16_Weights.DEFAULT - added:
field VGG16_Weights.IMAGENET1K_FEATURES - added:
field VGG16_Weights.IMAGENET1K_V1 - added:
field VGG19_BN_Weights.DEFAULT - added:
field VGG19_BN_Weights.IMAGENET1K_V1 - added:
field VGG19_Weights.DEFAULT - added:
field VGG19_Weights.IMAGENET1K_V1
- changed:
function mc3_18(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> VideoResNet → function mc3_18(*, weights: Optional[MC3_18_Weights]=…, progress: bool=…, **kwargs: Any) -> VideoResNet - changed:
function r2plus1d_18(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> VideoResNet → function r2plus1d_18(*, weights: Optional[R2Plus1D_18_Weights]=…, progress: bool=…, **kwargs: Any) -> VideoResNet - changed:
function r3d_18(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> VideoResNet → function r3d_18(*, weights: Optional[R3D_18_Weights]=…, progress: bool=…, **kwargs: Any) -> VideoResNet - added:
class MC3_18_Weights(WeightsEnum) - added:
class R2Plus1D_18_Weights(WeightsEnum) - added:
class R3D_18_Weights(WeightsEnum) - added:
class VideoResNet(nn.Module) - added:
constructor VideoResNet(block: Type[Union[BasicBlock, Bottleneck]], conv_makers: Sequence[Type[Union[Conv3DSimple, Conv3DNoTemporal, Conv2Plus1D]]], layers: List[int], stem: Callable[..., nn.Module], num_classes: int=…, zero_init_residual: bool=…) -> None - added:
field MC3_18_Weights.DEFAULT - added:
field MC3_18_Weights.KINETICS400_V1 - added:
field R2Plus1D_18_Weights.DEFAULT - added:
field R2Plus1D_18_Weights.KINETICS400_V1 - added:
field R3D_18_Weights.DEFAULT - added:
field R3D_18_Weights.KINETICS400_V1 - added:
method VideoResNet.forward(x: Tensor) -> Tensor
- changed:
function vit_b_16(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> VisionTransformer → function vit_b_16(*, weights: Optional[ViT_B_16_Weights]=…, progress: bool=…, **kwargs: Any) -> VisionTransformer - changed:
function vit_b_32(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> VisionTransformer → function vit_b_32(*, weights: Optional[ViT_B_32_Weights]=…, progress: bool=…, **kwargs: Any) -> VisionTransformer - changed:
function vit_l_16(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> VisionTransformer → function vit_l_16(*, weights: Optional[ViT_L_16_Weights]=…, progress: bool=…, **kwargs: Any) -> VisionTransformer - changed:
function vit_l_32(pretrained: bool=…, progress: bool=…, **kwargs: Any) -> VisionTransformer → function vit_l_32(*, weights: Optional[ViT_L_32_Weights]=…, progress: bool=…, **kwargs: Any) -> VisionTransformer - added:
class ViT_B_16_Weights(WeightsEnum) - added:
class ViT_B_32_Weights(WeightsEnum) - added:
class ViT_H_14_Weights(WeightsEnum) - added:
class ViT_L_16_Weights(WeightsEnum) - added:
class ViT_L_32_Weights(WeightsEnum) - added:
field ViT_B_16_Weights.DEFAULT - added:
field ViT_B_16_Weights.IMAGENET1K_SWAG_E2E_V1 - added:
field ViT_B_16_Weights.IMAGENET1K_SWAG_LINEAR_V1 - added:
field ViT_B_16_Weights.IMAGENET1K_V1 - added:
field ViT_B_32_Weights.DEFAULT - added:
field ViT_B_32_Weights.IMAGENET1K_V1 - added:
field ViT_H_14_Weights.DEFAULT - added:
field ViT_H_14_Weights.IMAGENET1K_SWAG_E2E_V1 - added:
field ViT_H_14_Weights.IMAGENET1K_SWAG_LINEAR_V1 - added:
field ViT_L_16_Weights.DEFAULT - added:
field ViT_L_16_Weights.IMAGENET1K_SWAG_E2E_V1 - added:
field ViT_L_16_Weights.IMAGENET1K_SWAG_LINEAR_V1 - added:
field ViT_L_16_Weights.IMAGENET1K_V1 - added:
field ViT_L_32_Weights.DEFAULT - added:
field ViT_L_32_Weights.IMAGENET1K_V1 - added:
function vit_h_14(*, weights: Optional[ViT_H_14_Weights]=…, progress: bool=…, **kwargs: Any) -> VisionTransformer
- added:
export Conv2dNormActivation - added:
export Conv3dNormActivation - added:
export DropBlock2d - added:
export DropBlock3d - added:
export MLP - added:
export Permute - added:
export complete_box_iou - added:
export complete_box_iou_loss - added:
export distance_box_iou - added:
export distance_box_iou_loss - added:
export drop_block2d - added:
export drop_block3d
- added:
function complete_box_iou(boxes1: Tensor, boxes2: Tensor, eps: float=…) -> Tensor - added:
function distance_box_iou(boxes1: Tensor, boxes2: Tensor, eps: float=…) -> Tensor
- added:
function complete_box_iou_loss(boxes1: torch.Tensor, boxes2: torch.Tensor, reduction: str=…, eps: float=…) -> torch.Tensor
- added:
function distance_box_iou_loss(boxes1: torch.Tensor, boxes2: torch.Tensor, reduction: str=…, eps: float=…) -> torch.Tensor
- added:
class DropBlock2d(nn.Module) - added:
class DropBlock3d(DropBlock2d) - added:
constructor DropBlock2d(p: float, block_size: int, inplace: bool=…, eps: float=…) -> None - added:
constructor DropBlock3d(p: float, block_size: int, inplace: bool=…, eps: float=…) -> None - added:
function drop_block2d(input: Tensor, p: float, block_size: int, inplace: bool=…, eps: float=…, training: bool=…) -> Tensor - added:
function drop_block3d(input: Tensor, p: float, block_size: int, inplace: bool=…, eps: float=…, training: bool=…) -> Tensor - added:
method DropBlock2d.forward(input: Tensor) -> Tensor - added:
method DropBlock3d.forward(input: Tensor) -> Tensor
- changed:
constructor FeaturePyramidNetwork(in_channels_list: List[int], out_channels: int, extra_blocks: Optional[ExtraFPNBlock]=…) → constructor FeaturePyramidNetwork(in_channels_list: List[int], out_channels: int, extra_blocks: Optional[ExtraFPNBlock]=…, norm_layer: Optional[Callable[..., nn.Module]]=…)
- changed:
function sigmoid_focal_loss(inputs: torch.Tensor, targets: torch.Tensor, alpha: float=…, gamma: float=…, reduction: str=…) → function sigmoid_focal_loss(inputs: torch.Tensor, targets: torch.Tensor, alpha: float=…, gamma: float=…, reduction: str=…) -> torch.Tensor
- changed:
constructor ConvNormActivation(in_channels: int, out_channels: int, kernel_size: int=…, stride: int=…, padding: Optional[int]=…, groups: int=…, norm_layer: Optional[Callable[..., torch.nn.Module]]=…, activation_layer: Optional[Callable[..., torch.nn.Module]]=…, dilation: int=…, inplace: Optional[bool]=…, bias: Optional[bool]=…) -> None → constructor ConvNormActivation(in_channels: int, out_channels: int, kernel_size: int=…, stride: int=…, padding: Optional[int]=…, groups: int=…, norm_layer: Optional[Callable[..., torch.nn.Module]]=…, activation_layer: Optional[Callable[..., torch.nn.Module]]=…, dilation: int=…, inplace: Optional[bool]=…, bias: Optional[bool]=…, conv_layer: Callable[..., torch.nn.Module]=…) -> None - added:
class Conv2dNormActivation(ConvNormActivation) - added:
class Conv3dNormActivation(ConvNormActivation) - added:
class MLP(torch.nn.Sequential) - added:
class Permute(torch.nn.Module) - added:
constructor Conv2dNormActivation(in_channels: int, out_channels: int, kernel_size: int=…, stride: int=…, padding: Optional[int]=…, groups: int=…, norm_layer: Optional[Callable[..., torch.nn.Module]]=…, activation_layer: Optional[Callable[..., torch.nn.Module]]=…, dilation: int=…, inplace: Optional[bool]=…, bias: Optional[bool]=…) -> None - added:
constructor Conv3dNormActivation(in_channels: int, out_channels: int, kernel_size: int=…, stride: int=…, padding: Optional[int]=…, groups: int=…, norm_layer: Optional[Callable[..., torch.nn.Module]]=…, activation_layer: Optional[Callable[..., torch.nn.Module]]=…, dilation: int=…, inplace: Optional[bool]=…, bias: Optional[bool]=…) -> None - added:
constructor MLP(in_channels: int, hidden_channels: List[int], norm_layer: Optional[Callable[..., torch.nn.Module]]=…, activation_layer: Optional[Callable[..., torch.nn.Module]]=…, inplace: Optional[bool]=…, bias: bool=…, dropout: float=…) - added:
constructor Permute(dims: List[int]) - added:
method Permute.forward(x: Tensor) -> Tensor
- added:
class AugMix(torch.nn.Module) - added:
constructor AugMix(severity: int=…, mixture_width: int=…, chain_depth: int=…, alpha: float=…, all_ops: bool=…, interpolation: InterpolationMode=…, fill: Optional[List[float]]=…) -> None - added:
method AugMix.forward(orig_img: Tensor) -> Tensor
- changed:
function to_tensor(pic) → function to_tensor(pic) -> Tensor - added:
function get_dimensions(img: Tensor) -> List[int]
- added:
function get_dimensions(img: Any) -> List[int]
- added:
function erase(img: Tensor, i: int, j: int, h: int, w: int, v: Tensor, inplace: bool=…) -> Tensor - added:
function get_dimensions(img: Tensor) -> List[int] - added:
function normalize(tensor: Tensor, mean: List[float], std: List[float], inplace: bool=…) -> Tensor
- changed:
function draw_bounding_boxes(image: torch.Tensor, boxes: torch.Tensor, labels: Optional[List[str]]=…, colors: Optional[Union[List[Union[str, Tuple[int, int, int]]], str, Tuple[int, int, int]]]=…, fill: Optional[bool]=…, width: int=…, font: Optional[str]=…, font_size: int=…) -> torch.Tensor → function draw_bounding_boxes(image: torch.Tensor, boxes: torch.Tensor, labels: Optional[List[str]]=…, colors: Optional[Union[List[Union[str, Tuple[int, int, int]]], str, Tuple[int, int, int]]]=…, fill: Optional[bool]=…, width: int=…, font: Optional[str]=…, font_size: Optional[int]=…) -> torch.Tensor