Ei()4SSKrSSKJrJrJr SSKJrJr SSKJ r SSK J r J r Jr SSKJrJr SSKJrJr SS KJrJr "S S \R05r\"S S 4S9\ R4SSSS.S\S\\S\S\R04S\S\\\S\\ S\4Sjj5rS%S\R>S\S\\\S\\ S\\S\R04S\4 Sjjr S\!S\\S\S\S\4 S jr"\"S S!4S9\ R4SSSS.S\S\\S"\!S\S\R04S\S\\\S\\ S\R04S#jj5r#S%S\\RH\RJ4S"\!S\S\\\S\\ S\\S\R04S\R04S$jjr&g)&N)CallableOptionalUnion)nnTensor)misc) ExtraFPNBlockFeaturePyramidNetworkLastLevelMaxPool) mobilenetresnet)_get_enum_from_fn WeightsEnum)handle_legacy_interfaceIntermediateLayerGetterc^\rSrSrSrSS\R S\\\4S\ \ S\ S\ \ S \ \ S \R 4S S4U4S jjjrS \S \\\44SjrSrU=r$)BackboneWithFPN a Adds a FPN on top of a model. Internally, it uses torchvision.models._utils.IntermediateLayerGetter to extract a submodel that returns the feature maps specified in return_layers. The same limitations of IntermediateLayerGetter apply here. Args: backbone (nn.Module) return_layers (Dict[name, new_name]): a dict containing the names of the modules for which the activations will be returned as the key of the dict, and the value of the dict is the name of the returned activation (which the user can specify). in_channels_list (List[int]): number of channels for each feature map that is returned, in the order they are present in the OrderedDict out_channels (int): number of channels in the FPN. norm_layer (callable, optional): Module specifying the normalization layer to use. Default: None Attributes: out_channels (int): the number of channels in the FPN Nbackbone return_layersin_channels_list out_channels extra_blocks norm_layer.returnc>[TU]5 Uc [5n[XS9Ul[ UUUUS9UlX@lg)N)r)rrrr)super__init__r rbodyr fpnr)selfrrrrrr __class__s h/home/mande/repo/quber/.venv/lib/python3.13/site-packages/torchvision/models/detection/backbone_utils.pyrBackboneWithFPN.__init__!sJ   +-L+HR (-%%!   )xcJURU5nURU5nU$)N)r r!)r"r's r$forwardBackboneWithFPN.forward8s! IIaL HHQKr&)r r!r)NN)__name__ __module__ __qualname____firstlineno____doc__rModuledictstrlistintrr rrrr)__static_attributes__ __classcell__)r#s@r$rr s2159=))))CH~)s) )  ) }- )Xc299n56) )).Df$5r&r pretrainedcF[[RUS5S$N backbone_name IMAGENET1K_V1)rr__dict__kwargss r$r?As(9P)QRSbcr&)weights)rtrainable_layersreturned_layersrr:r@r.rBrCrrcH[RU"XS9n[XcXE5$)a Constructs a specified ResNet backbone with FPN on top. Freezes the specified number of layers in the backbone. Examples:: >>> import torch >>> from torchvision.models import ResNet50_Weights >>> from torchvision.models.detection.backbone_utils import resnet_fpn_backbone >>> backbone = resnet_fpn_backbone(backbone_name='resnet50', weights=ResNet50_Weights.DEFAULT, trainable_layers=3) >>> # get some dummy image >>> x = torch.rand(1,3,64,64) >>> # compute the output >>> output = backbone(x) >>> print([(k, v.shape) for k, v in output.items()]) >>> # returns >>> [('0', torch.Size([1, 256, 16, 16])), >>> ('1', torch.Size([1, 256, 8, 8])), >>> ('2', torch.Size([1, 256, 4, 4])), >>> ('3', torch.Size([1, 256, 2, 2])), >>> ('pool', torch.Size([1, 256, 1, 1]))] Args: backbone_name (string): resnet architecture. Possible values are 'resnet18', 'resnet34', 'resnet50', 'resnet101', 'resnet152', 'resnext50_32x4d', 'resnext101_32x8d', 'wide_resnet50_2', 'wide_resnet101_2' weights (WeightsEnum, optional): The pretrained weights for the model norm_layer (callable): it is recommended to use the default value. For details visit: (https://github.com/facebookresearch/maskrcnn-benchmark/issues/267) trainable_layers (int): number of trainable (not frozen) layers starting from final block. Valid values are between 0 and 5, with 5 meaning all backbone layers are trainable. returned_layers (list of int): The layers of the network to return. Each entry must be in ``[1, 4]``. By default, all layers are returned. extra_blocks (ExtraFPNBlock or None): if provided, extra operations will be performed. It is expected to take the fpn features, the original features and the names of the original features as input, and returns a new list of feature maps and their corresponding names. By default, a ``LastLevelMaxPool`` is used. r@r)rr<_resnet_fpn_extractor)r:r@rrBrCrrs r$resnet_fpn_backbonerG>s%h}-gUH _ [[r&rc US:dUS:a[SU35e/SQSUnUS:XaURS5 UR5HLupg[UVs/sHoR U5(+PM sn5(dM;UR S5 MN Uc [ 5nUc/SQn[U5S::d[U5S:a[SU35e[U5V V s0sHupS U 3[U 5_M n n n URS -n UV s/sH oS U S - --PM nn S n[X XX4S9$s snfs sn n fs sn f)Nrz3Trainable layers should be in the range [0,5], got )layer4layer3layer2layer1conv1bn1F)r rAz6Each returned layer should be in the range [1,4]. Got layerr rPrr) ValueErrorappendnamed_parametersall startswithrequires_grad_r minmax enumerater2inplanesr)rrBrCrrlayers_to_trainname parameterrRvkrin_channels_stage2irrs r$rFrFvsf!/!3NO_N`abbGHYIYZO1u%#446 HuOOE**H I I  $ $U +7') & ?q C$8A$=QRaQbcdd5>5OP5OTQuQC[#a&(5OMP!**a/CRS?aQ1q5\9?SL !1l IQTsD; 'EE is_trainedtrainable_backbone_layers max_value default_valuecU(dUb[R"SUS35 UnUcUnUS:dX:a[SUSUS35eU$)NzChanging trainable_backbone_layers has no effect if neither pretrained nor pretrained_backbone have been set to True, falling back to trainable_backbone_layers=z! so that all layers are trainablerz4Trainable backbone layers should be in the range [0,], got  )warningswarnrV)rgrhrirjs r$_validate_trainable_layersrpsz  $ 0 MM==FKGhj  %.!!($1! 1$(A(MB9+WUnToop q   %$r&cF[[RUS5S$r9)rr r<r=s r$r?r?s (););F?O  ! #s?';z'IMj[\nM]]detduuvwx xCL_C]^C]41M,-A6C] ^M\]_]%56CC_] %5|ku   MM  IIhrl//q A  &E]*_]sF:F:-GG)NNN)'rntypingrrrtorchrrtorchvision.opsr misc_nn_ops'torchvision.ops.feature_pyramid_networkr r r r r_apirr_utilsrrr0rFrozenBatchNorm2dr2r4r3rGResNetrFboolrprt MobileNetV2 MobileNetV3rsr&r$rs,,/jj 1E.bii.bc ,7+H+H+/,0/\/\k "/\bii( /\  /\ d3i( /\=)/\/\  /\j,0,059 mm  d3i( =)  #ryy.12   F%%'}%% %  %4f ,7+H+H+/,0 ` `k " `  ` bii( `  `d3i( `=) `YY `  `$,0,059 -I))9+@+@@A- --d3i( - =) - #ryy.12 -YY-r&