ёiD%SSKJr SSKrSSKJr SSKJr SSKrSSKJr SSK J r SSK J r J r JrJrJrJrJr SSKJr SS KJr \(a*SS KJrJr SS KJr SS KJr SS KJr \Sr"SS\5r/rSSSSSS.r S\!S'"SS\RD5r#"SS\RD5r$"SS\RD5r%"SS\RD5r&"S S!\RD5r'"S"S#\RD5r(S*S$jr)S+S,S%jjr*S+S,S&jjr+S+S,S'jjr,S+S,S(jjr-S+S,S)jjr.g)-) annotationsN) TYPE_CHECKING)Unpack)nn) ParamAttr)AdaptiveAvgPool2D AvgPool2D BatchNormConv2DDropoutLinear MaxPool2D)Uniform)get_weights_path_from_url)Literal TypedDict) NotRequired)Tensor)Size2) densenet121 densenet161 densenet169 densenet201 densenet264c>\rSrSr%S\S'S\S'S\S'S\S'S rg ) _DenseNetOptions4zNotRequired[int]bn_sizezNotRequired[float]dropout num_classeszNotRequired[bool] with_poolN)__name__ __module__ __qualname____firstlineno____annotations____static_attributes__r"]/var/www/html/banglarbhumi/venv/lib/python3.13/site-packages/paddle/vision/models/densenet.pyrr4s!!##%%$$r)r)zYhttps://paddle-imagenet-models-name.bj.bcebos.com/dygraph/DenseNet121_pretrained.pdparams db1b239ed80a905290fd8b01d3af08e4)zYhttps://paddle-imagenet-models-name.bj.bcebos.com/dygraph/DenseNet161_pretrained.pdparams 62158869cb315098bd25ddbfd308a853)zYhttps://paddle-imagenet-models-name.bj.bcebos.com/dygraph/DenseNet169_pretrained.pdparams 82cc7c635c3f19098c748850efb2d796)zYhttps://paddle-imagenet-models-name.bj.bcebos.com/dygraph/DenseNet201_pretrained.pdparams 16ca29565a7712329cf9e36e02caaf58)zYhttps://paddle-imagenet-models-name.bj.bcebos.com/dygraph/DenseNet264_pretrained.pdparams 3270ce516b85370bba88cfdd9f60bff4zdict[str, tuple[str, str]] model_urlsc`^\rSrSrSSU4SjjjrSSjrSrU=r$) BNACConvLayerUc x>[TU]5 [XS9Ul[ UUUUUU[ 5SS9Ulg)NactF in_channels out_channels kernel_sizestridepaddinggroups weight_attr bias_attr)super__init__r _batch_normr r_conv self num_channels num_filters filter_sizer;padr=r6 __class__s r*rABNACConvLayer.__init__VsC $\;$$#!    r)cJURU5nURU5nU$NrBrCrEinputys r*forwardBNACConvLayer.forwardns$   U # JJqMr)rNrrUrelurFintrGrXrHrr;rrIrr=rXr6strreturnNonerPrrZrr#r$r%r&rArRr( __classcell__rJs@r*r2r2Uss               0r)r2cV^\rSrSr%S\S'SU4SjjrSSjrSrU=r$) DenseLayertfloatrc>[TU]5 X@l[UX2-SSSS9Ul[X2-USSSS9UlU(a[ USS9Ulgg)NrUrrFrGrHrIr;downscale_in_infer)pmode)r@rArr2 bn_ac_func1 bn_ac_func2r dropout_func)rErF growth_raterrrJs r*rADenseLayer.__init__wso  (%-  ) .#    ''8L MD  r)cURU5nURU5nUR(aURU5n[R "X/SS9nU$)NrU)axis)rjrkrrlpaddleconcat)rErPconvs r*rRDenseLayer.forwardsP&% <<$$T*D}}e]3 r))rjrkrrl) rFrXrmrXrrXrrcrZr[r\ r#r$r%r&r'rArRr(r^r_s@r*raratsE NNN.1N[T U]5 XPl/UlUn[ U5HDnURR UR USUS-3[UUUUS955 Xt-nMF g)N_rU)rFrmrr)r@rArdense_layer_funcrangeappend add_sublayerra) rErF num_layersrrmrname pre_channellayerrJs r*rADenseBlock.__init__s  "" :&E  ! ! ( (!!fAeai[)%0$/ ' '  &3K'r)c@UnURH nU"U5nM U$rM)r{)rErPrsfuncs r*rRDenseBlock.forwards%))D:D* r))r{rrM)rFrXrrXrrXrmrXrrcrz str | NonerZr[r\rur_s@r*rwrwse N 444 4  4  44 448r)rwc6^\rSrSrSU4SjjrSSjrSrU=r$)TransitionLayercf>[TU]5 [UUSSSS9Ul[ SSSS9Ulg)NrUrrer:r;r<)r@rAr2 conv_ac_funcr pool2d_avg)rErFnum_output_featuresrJs r*rATransitionLayer.__init__s< )%+  $!QGr)cJURU5nURU5nU$rMrrrOs r*rRTransitionLayer.forwards%   e $ OOA r)r)rFrXrrXrZr[r\r]r_s@r*rrs Hr)rc`^\rSrSrSSU4SjjjrSSjrSrU=r$) ConvBNLayerc x>[TU]5 [UUUUUU[5SS9Ul[ X'S9Ulg)NFr7r5)r@rAr rrCr rBrDs r*rAConvBNLayer.__init__sE $$#!    %[:r)cJURU5nURU5nU$rM)rCrBrOs r*rRConvBNLayer.forwards% JJu    Q r)rNrTrWr\r]r_s@r*rrss ;;; ;  ;  ;;; ;;0r)rct^\rSrSr%SrS\S'S\S'S S U4SjjjrS SjrS rU=r $) DenseNetaDenseNet model from `"Densely Connected Convolutional Networks" `_. Args: layers (int, optional): Layers of DenseNet. Default: 121. bn_size (int, optional): Expansion of growth rate in the middle layer. Default: 4. dropout (float, optional): Dropout rate. Default: :math:`0.0`. num_classes (int, optional): Output dim of last fc layer. If num_classes <= 0, last fc layer will not be defined. Default: 1000. with_pool (bool, optional): Use pool before the last fc layer or not. Default: True. Returns: :ref:`api_paddle_nn_Layer`. An instance of DenseNet model. Examples: .. code-block:: pycon >>> import paddle >>> from paddle.vision.models import DenseNet >>> # Build model >>> densenet = DenseNet() >>> x = paddle.rand([1, 3, 224, 224]) >>> out = densenet(x) >>> print(out.shape) paddle.Size([1, 1000]) rXr boolr!c>[TU]5 X@lXPl/SQnX;d SUSU35eSS/SQ4SS/S Q4SS/S Q4SS/S Q4SS/S Q4S.nXqupn [ S USSS SS9Ul[ S SSS9UlXl/Ul /Ul Un Un [U 5HupURRURSU S-3[U UUU US[U S-5-S955 XU --n U n U [!U 5S- :wdMnURRURSU S-S3[#U U S-S955 U S-n U S-n M [%U SS9UlUR(a[)S5UlURS:aHS[,R."U S-5- n[1U U[3[5U*U5S9[35S9Ulgg)N)yzsupported layers are z but input layer is @ ) `0)rr$r)rrrr)rrrr)rrrrrfrrV)rFrGrHr;rIr6rUrdb_conv_rs)rFrrrmrrtr_conv_blk)rFrr5rg?) initializer)r>r?)r@rAr r!r conv1_funcr pool2d_max block_configdense_block_func_listtransition_func_list enumerater}r~rwrYlenrr batch_normrrmathsqrtr rrout)rElayersrrr r!supported_layers densenet_specnum_init_featuresrmrpre_num_channels num_featuresirstdvrJs r*rADenseNet.__init__sE &"4) #$4#55I& R )b/*b/*b/*b/*b/*  8E7L4 %)  $!QG(%'"$&!,( &|4MA  & & - -!!q1ug&%5#- '$/ '#c!a%j0   ({*BBL+ C %))))00%%!!a%-')90<0A$01#4 +q0 ;5>$Lf= >>/2DO   a <##566D%'4%2FG#+ DH r)cURU5nURU5n[UR5HMup4URU"U5nU[ UR5S- :wdM8UR U"U5nMO URU5nUR(aURU5nURS:a'[R"WSSS9nURU5nW$)NrUr) start_axis stop_axis)rrrrrrrrr!rr rqflattenr)rErPrsrrrQs r*rRDenseNet.forwardfsu%t$&t'8'89MA--a06DC))*Q..003D9: t$ >>%A   a qQ"=A Ar)) rrrrr rrrrr!)rgiT) rrXrrXrrcr rXr!rrZr[r\) r#r$r%r&__doc__r'rArRr(r^r_s@r*rrs}<O PPP P  P  P PPdr)rc [SSU0UD6nU(a[U[;d US35e[[US[US5n[R"U5nUR U5 U$)NrzJ model do not have a pretrained model now, you should set pretrained=FalserrUr")rr0rrqloadset_dict)archr pretrainedkwargsmodel weight_pathparams r* _densenetr{s  -F -f -Ez! f^ _ !0 t Q D!1!!4  K( u Lr)c [SSU40UD6$)aDenseNet 121-layer model from `"Densely Connected Convolutional Networks" `_. Args: pretrained (bool, optional): Whether to load pre-trained weights. If True, returns a model pre-trained on ImageNet. Default: False. **kwargs (optional): Additional keyword arguments. For details, please refer to :ref:`DenseNet `. Returns: :ref:`api_paddle_nn_Layer`. An instance of DenseNet 121-layer model. Examples: .. code-block:: pycon >>> import paddle >>> from paddle.vision.models import densenet121 >>> # Build model >>> model = densenet121() >>> # Build model and load imagenet pretrained weight >>> # model = densenet121(pretrained=True) >>> x = paddle.rand([1, 3, 224, 224]) >>> out = model(x) >>> print(out.shape) paddle.Size([1, 1000]) rrrrrs r*rr@ ]C >v >>r)c [SSU40UD6$)aDenseNet 161-layer model from `"Densely Connected Convolutional Networks" `_. Args: pretrained (bool, optional): Whether to load pre-trained weights. If True, returns a model pre-trained on ImageNet. Default: False. **kwargs (optional): Additional keyword arguments. For details, please refer to :ref:`DenseNet `. Returns: :ref:`api_paddle_nn_Layer`. An instance of DenseNet 161-layer model. Examples: .. code-block:: pycon >>> import paddle >>> from paddle.vision.models import densenet161 >>> # Build model >>> model = densenet161() >>> # Build model and load imagenet pretrained weight >>> # model = densenet161(pretrained=True) >>> x = paddle.rand([1, 3, 224, 224]) >>> out = model(x) >>> print(out.shape) paddle.Size([1, 1000]) rrrrs r*rrrr)c [SSU40UD6$)aDenseNet 169-layer model from `"Densely Connected Convolutional Networks" `_. Args: pretrained (bool, optional): Whether to load pre-trained weights. If True, returns a model pre-trained on ImageNet. Default: False. **kwargs (optional): Additional keyword arguments. For details, please refer to :ref:`DenseNet `. Returns: :ref:`api_paddle_nn_Layer`. An instance of DenseNet 169-layer model. Examples: .. code-block:: pycon >>> import paddle >>> from paddle.vision.models import densenet169 >>> # Build model >>> model = densenet169() >>> # Build model and load imagenet pretrained weight >>> # model = densenet169(pretrained=True) >>> x = paddle.rand([1, 3, 224, 224]) >>> out = model(x) >>> print(out.shape) paddle.Size([1, 1000]) rrrrs r*rrrr)c [SSU40UD6$)aDenseNet 201-layer model from `"Densely Connected Convolutional Networks" `_. Args: pretrained (bool, optional): Whether to load pre-trained weights. If True, returns a model pre-trained on ImageNet. Default: False. **kwargs (optional): Additional keyword arguments. For details, please refer to :ref:`DenseNet `. Returns: :ref:`api_paddle_nn_Layer`. An instance of DenseNet 201-layer model. Examples: .. code-block:: pycon >>> import paddle >>> from paddle.vision.models import densenet201 >>> # Build model >>> model = densenet201() >>> # Build model and load imagenet pretrained weight >>> # model = densenet201(pretrained=True) >>> x = paddle.rand([1, 3, 224, 224]) >>> out = model(x) >>> print(out.shape) paddle.Size([1, 1000]) rrrrs r*rrs> ]C >v >>r)c [SSU40UD6$)aDenseNet 264-layer model from `"Densely Connected Convolutional Networks" `_. Args: pretrained (bool, optional): Whether to load pre-trained weights. If True, returns a model pre-trained on ImageNet. Default: False. **kwargs (optional): Additional keyword arguments. For details, please refer to :ref:`DenseNet `. Returns: :ref:`api_paddle_nn_Layer`. An instance of DenseNet 264-layer model. Examples: .. code-block:: pycon >>> import paddle >>> from paddle.vision.models import densenet264 >>> # Build model >>> model = densenet264() >>> # Build model and load imagenet pretrained weight >>> # model = densenet264(pretrained=True) >>> x = paddle.rand([1, 3, 224, 224]) >>> out = model(x) >>> print(out.shape) paddle.Size([1, 1000]) rrrrs r*rrrr)) r _DenseNetArchrrXrrrUnpack[_DenseNetOptions]rZr)F)rrrrrZr)/ __future__rrtypingrtyping_extensionsrrqrpaddle.base.param_attrr paddle.nnrr r r r r rpaddle.nn.initializerrpaddle.utils.downloadrrrrrpaddle._typingrrr__all__r0r'Layerr2rarwrrrrrrrrrr"r)r*rs# $ ,*;)-$ M%9% #* &0BHH>""J##Lbhh("((>FrxxFR  '   , ? ?(@ ?  ?H ? ?(@ ?  ?H ? ?(@ ?  ?H??(@? ?F ? ?(@ ?  ?r)