|jnUddlmZddlmZmZddlZddlmZddlmZer2ddlm Z m Z ddl m Z m Z ddlmZdd lmZe d ZGd d e ZgZd ddddddddddddd Zded<GddejZGddejZGd d!ejZdd/Z d=d?d0Z d=d?d1Z d=d?d2Z d=d?d3Z d=d?d4Z d=d?d5Z! d=d?d6Z" d=d?d7Z# d=d?d8Z$ d=d?d9Z% d=d?d:Z& d=d?d;Z'dS)@) annotations) TYPE_CHECKINGCallableN)nn)get_weights_path_from_url)Literal TypedDict) NotRequiredUnpack)Tensor)Size2) resnet18resnet34resnet50 resnet101 resnet152resnext50_32x4dresnext50_64x4dresnext101_32x4dresnext101_64x4dresnext152_32x4dresnext152_64x4dwide_resnet50_2wide_resnet101_2c8eZdZUded<ded<ded<ded<dS)_ResNetOptionszNotRequired[int]width num_classeszNotRequired[bool] with_poolgroupsN)__name__ __module__ __qualname____annotations__k/lsinfo/ai/hellotax_ai/data_center/backend/venv/lib/python3.11/site-packages/paddle/vision/models/resnet.pyrr.sB%%%%$$$$      r&r)z:https://paddle-hapi.bj.bcebos.com/models/resnet18.pdparams cf548f46534aa3560945be4b95cd11c4)z:https://paddle-hapi.bj.bcebos.com/models/resnet34.pdparams 8d2275cf8706028345f78ac0e1d31969)z:https://paddle-hapi.bj.bcebos.com/models/resnet50.pdparams ca6f485ee1ab0492d38f323885b0ad80)z;https://paddle-hapi.bj.bcebos.com/models/resnet101.pdparams 02f35f034ca3858e1e54d4036443c92d)z;https://paddle-hapi.bj.bcebos.com/models/resnet152.pdparams 7ad16a2f1e7333859ff986138630fd7a)zAhttps://paddle-hapi.bj.bcebos.com/models/resnext50_32x4d.pdparams dc47483169be7d6f018fcbb7baf8775d)zAhttps://paddle-hapi.bj.bcebos.com/models/resnext50_64x4d.pdparams 063d4b483e12b06388529450ad7576db)zBhttps://paddle-hapi.bj.bcebos.com/models/resnext101_32x4d.pdparams 967b090039f9de2c8d06fe994fb9095f)zBhttps://paddle-hapi.bj.bcebos.com/models/resnext101_64x4d.pdparams 98e04e7ca616a066699230d769d03008)zBhttps://paddle-hapi.bj.bcebos.com/models/resnext152_32x4d.pdparams 18ff0beee21f2efc99c4b31786107121)zBhttps://paddle-hapi.bj.bcebos.com/models/resnext152_64x4d.pdparams 77c4af00ca42c405fa7f841841959379)zAhttps://paddle-hapi.bj.bcebos.com/models/wide_resnet50_2.pdparams 0282f804d73debdab289bd9fea3fa6dc)zBhttps://paddle-hapi.bj.bcebos.com/models/wide_resnet101_2.pdparams d4360a2d23657f059216f5d5a1a9ac93zdict[str, tuple[str, str]] model_urlscFeZdZUdZded< ddfd ZddZxZS) BasicBlockint expansionN@inplanesplanesstrider downsamplenn.Layer | Noner base_widthdilation norm_layerCallable[..., nn.Layer] | NonereturnNonec t| tj}|dkrt dtj||dd|d|_|||_tj|_ tj||ddd|_ |||_ ||_ ||_ dS)Nr8z(Dilation > 1 not supported in BasicBlockF)paddingr> bias_attr)rIrJ)super__init__r BatchNorm2DNotImplementedErrorConv2Dconv1bn1ReLUreluconv2bn2r?r>) selfr<r=r>r?r rArBrC __class__s r'rLzBasicBlock.__init__rs   /1~J a<<%: Y fa6U    :f%%GII Yvvq!uMMM :f%%$ r&xr cH|}||}||}||}||}||}|j||}||z }||}|SN)rPrQrSrTrUr?rVrXidentityouts r'forwardzBasicBlock.forwardsjjmmhhsmmiinnjjoohhsmm ? &q))H xiinn r&r8Nr8r;r8Nr<r9r=r9r>r r?r@r r9rAr9rBr9rCrDrErFrXr rEr r!r"r#r:r$rLr^ __classcell__rWs@r'r7r7os}I &*59<r&r7cFeZdZUdZded< ddfd ZddZxZS)BottleneckBlockr9r:r8Nr;r<r=r>r r?r@r rArBrCrDrErFc t| tj}t ||dz z|z} tj|| dd|_|| |_tj| | d||||d|_|| |_ tj| ||j zdd|_ |||j z|_ tj |_||_||_dS)NgP@r8F)rJrH)rIr>r rBrJ)rKrLrrMr9rOrPrQrTrUr:conv3bn3rRrSr?r>) rVr<r=r>r?r rArBrCrrWs r'rLzBottleneckBlock.__init__s   JFj4/011F:YxUCCC :e$$Y        :e$$Y 6DN*A    :ft~566GII $ r&rXr c|}||}||}||}||}||}||}||}||}|j||}||z }||}|SrZ)rPrQrSrTrUrirjr?r[s r'r^zBottleneckBlock.forwardsjjmmhhsmmiinnjjoohhsmmiinnjjoohhsmm ? &q))H xiinn r&r_r`rarbrds@r'rfrfs~I &*59%%%%%%%Nr&rfceZdZUdZded<ded<ded<ded<ded<ded < d d!fd Z d"d#dZd$dZxZS)%ResNetayResNet model from `"Deep Residual Learning for Image Recognition" `_. Args: Block (BasicBlock|BottleneckBlock): Block module of model. depth (int, optional): Layers of ResNet, Default: 50. width (int, optional): Base width per convolution group for each convolution block, Default: 64. 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. groups (int, optional): Number of groups for each convolution block, Default: 1. Returns: :ref:`api_paddle_nn_Layer`. An instance of ResNet model. Examples: .. code-block:: pycon >>> import paddle >>> from paddle.vision.models import ResNet >>> from paddle.vision.models.resnet import ( ... BottleneckBlock, ... BasicBlock, ... ) >>> # build ResNet with 18 layers >>> resnet18 = ResNet(BasicBlock, 18) >>> # build ResNet with 50 layers >>> resnet50 = ResNet(BottleneckBlock, 50) >>> # build Wide ResNet model >>> wide_resnet50_2 = ResNet(BottleneckBlock, 50, width=64 * 2) >>> # build ResNeXt model >>> resnext50_32x4d = ResNet(BottleneckBlock, 50, width=4, groups=32) >>> x = paddle.rand([1, 3, 224, 224]) >>> out = resnet18(x) >>> print(out.shape) paddle.Size([1, 1000]) r9r rArboolrr<rB2r;Tr8block"type[BasicBlock | BottleneckBlock]depthrrErFctgdgdgdgdgdd}||}||_||_||_||_t j|_d|_ d|_ t j d|j d d dd |_ ||j |_ t j|_t jdd d |_||d|d|_||d|dd |_||d|d d |_||d|dd |_|rt jd|_|dkr$t jd|jz||_dSdS)N)rururu)rHrgrH)rHrgrH)rH$rH)"roer;r8rHruF) kernel_sizer>rIrJ)rr>rIr)r>i)r8r8)rKrLr rArrrrM _norm_layerr<rBrOrPrQrRrS MaxPool2Dmaxpool _make_layerlayer1layer2layer3layer4AdaptiveAvgPool2DavgpoolLinearr:fc) rVrqrsrrrr layer_cfglayersrWs r'rLzResNet.__init__s       5! &">  Y M     ##DM22GII |!QGGG &&ub&)<< &&uc6!9Q&GG &&uc6!9Q&GG &&uc6!9Q&GG  8/77DL ??ieo 5{CCDGGG ?r&Fr=blocksr>dilate nn.Sequentialc @|j}d}|j}|r|xj|zc_d}|dks|j||jzkrJt jt j|j||jzd|d|||jz}g} | ||j||||j|j ||||jz|_td|D]4} | ||j||j|j |5t j| S)Nr8F)r>rJ)r rArC) rrBr<r:r SequentialrOappendr rArange) rVrqr=rr>rrCr?previous_dilationr_s r'rzResNet._make_layerHsk%   M   MMV #MMF Q;;$-6EO+CCC MU_,!#  6EO344  J E  !   0 q&!!  A MMM;#)     }f%%r&rXr c||}||}||}||}||}||}||}||}|jr| |}|j dkr*tj |d}| |}|S)Nrr8)rPrQrSrrrrrrrrpaddleflattenr)rVrXs r'r^zResNet.forward}s JJqMM HHQKK IIaLL LLOO KKNN KKNN KKNN KKNN >  QA  a  q!$$A Ar&)ror;rpTr8)rqrrrsr9rr9rr9rrnr r9rErF)r8F) rqrrr=r9rr9r>r9rrnrErra) r!r"r#__doc__r$rLrr^rcrds@r'rmrms**XKKKOOOOOOMMMMMM .D.D.D.D.D.D.Dj 3&3&3&3&3&jr&rmarch _ResNetArchBlockrrrsr9 pretrainedrnkwargsUnpack[_ResNetOptions]rEc t||fi|}|rq|tvs J|dtt|dt|d}tj|}|||S)NzJ model do not have a pretrained model now, you should set pretrained=Falserr8)rmr5rrloadset_dict)rrrsrrmodel weight_pathparams r'_resnetrs 5% * *6 * *E z!!! _ _ _"!!0 t Q D!1!!4   K(( u Lr&Fc ,tdtd|fi|S)aResNet 18-layer model from `"Deep Residual Learning for Image Recognition" `_. 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:`ResNet `. Returns: :ref:`api_paddle_nn_Layer`. An instance of ResNet 18-layer model. Examples: .. code-block:: pycon >>> import paddle >>> from paddle.vision.models import resnet18 >>> # build model >>> model = resnet18() >>> # build model and load imagenet pretrained weight >>> # model = resnet18(pretrained=True) >>> x = paddle.rand([1, 3, 224, 224]) >>> out = model(x) >>> print(out.shape) paddle.Size([1, 1000]) rrzrr7rrs r'rrs < :z2z D DV D DDr&c ,tdtd|fi|S)aResNet 34-layer model from `"Deep Residual Learning for Image Recognition" `_. 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:`ResNet `. Returns: :ref:`api_paddle_nn_Layer`. An instance of ResNet 34-layer model. Examples: .. code-block:: pycon >>> import paddle >>> from paddle.vision.models import resnet34 >>> # build model >>> model = resnet34() >>> # build model and load imagenet pretrained weight >>> # model = resnet34(pretrained=True) >>> x = paddle.rand([1, 3, 224, 224]) >>> out = model(x) >>> print(out.shape) paddle.Size([1, 1000]) rr{rrs r'rrs!@ :z2z D DV D DDr&c ,tdtd|fi|S)aResNet 50-layer model from `"Deep Residual Learning for Image Recognition" `_. 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:`ResNet `. Returns: :ref:`api_paddle_nn_Layer`. An instance of ResNet 50-layer model. Examples: .. code-block:: pycon >>> import paddle >>> from paddle.vision.models import resnet50 >>> # build model >>> model = resnet50() >>> # build model and load imagenet pretrained weight >>> # model = resnet50(pretrained=True) >>> x = paddle.rand([1, 3, 224, 224]) >>> out = model(x) >>> print(out.shape) paddle.Size([1, 1000]) rrorrfrs r'rrs!@ :J I I& I IIr&c ,tdtd|fi|S)aResNet 101-layer model from `"Deep Residual Learning for Image Recognition" `_. 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:`ResNet `. Returns: :ref:`api_paddle_nn_Layer`. An instance of ResNet 101-layer. Examples: .. code-block:: pycon >>> import paddle >>> from paddle.vision.models import resnet101 >>> # build model >>> model = resnet101() >>> # build model and load imagenet pretrained weight >>> # model = resnet101(pretrained=True) >>> x = paddle.rand([1, 3, 224, 224]) >>> out = model(x) >>> print(out.shape) paddle.Size([1, 1000]) rr|rrs r'rr!@ ;j K KF K KKr&c ,tdtd|fi|S)aResNet 152-layer model from `"Deep Residual Learning for Image Recognition" `_. 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:`ResNet `. Returns: :ref:`api_paddle_nn_Layer`. An instance of ResNet 152-layer model. Examples: .. code-block:: pycon >>> import paddle >>> from paddle.vision.models import resnet152 >>> # build model >>> model = resnet152() >>> # build model and load imagenet pretrained weight >>> # model = resnet152(pretrained=True) >>> x = paddle.rand([1, 3, 224, 224]) >>> out = model(x) >>> print(out.shape) paddle.Size([1, 1000]) rr}rrs r'rr1rr&c @d|d<d|d<tdtd|fi|S)aResNeXt-50 32x4d model from `"Aggregated Residual Transformations for Deep Neural 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:`ResNet `. Returns: :ref:`api_paddle_nn_Layer`. An instance of ResNeXt-50 32x4d model. Examples: .. code-block:: pycon >>> import paddle >>> from paddle.vision.models import resnext50_32x4d >>> # build model >>> model = resnext50_32x4d() >>> # build model and load imagenet pretrained weight >>> # model = resnext50_32x4d(pretrained=True) >>> x = paddle.rand([1, 3, 224, 224]) >>> out = model(x) >>> print(out.shape) paddle.Size([1, 1000]) r rgrrrorrs r'rrT5@F8F7O $or: P P P PPr&c @d|d<d|d<tdtd|fi|S)aResNeXt-50 64x4d model from `"Aggregated Residual Transformations for Deep Neural 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:`ResNet `. Returns: :ref:`api_paddle_nn_Layer`. An instance of ResNeXt-50 64x4d model. Examples: .. code-block:: pycon >>> import paddle >>> from paddle.vision.models import resnext50_64x4d >>> # build model >>> model = resnext50_64x4d() >>> # build model and load imagenet pretrained weight >>> # model = resnext50_64x4d(pretrained=True) >>> x = paddle.rand([1, 3, 224, 224]) >>> out = model(x) >>> print(out.shape) paddle.Size([1, 1000]) r;r rgrrrorrs r'rryrr&c @d|d<d|d<tdtd|fi|S)aResNeXt-101 32x4d model from `"Aggregated Residual Transformations for Deep Neural 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:`ResNet `. Returns: :ref:`api_paddle_nn_Layer`. An instance of ResNeXt-101 32x4d model. Examples: .. code-block:: pycon >>> import paddle >>> from paddle.vision.models import resnext101_32x4d >>> # build model >>> model = resnext101_32x4d() >>> # build model and load imagenet pretrained weight >>> # model = resnext101_32x4d(pretrained=True) >>> x = paddle.rand([1, 3, 224, 224]) >>> out = model(x) >>> print(out.shape) paddle.Size([1, 1000]) rr rgrrr|rrs r'rr@@F8F7O OS*  @F  r&c @d|d<d|d<tdtd|fi|S)aResNeXt-101 64x4d model from `"Aggregated Residual Transformations for Deep Neural 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:`ResNet `. Returns: :ref:`api_paddle_nn_Layer`. An instance of ResNeXt-101 64x4d model. Examples: .. code-block:: pycon >>> import paddle >>> from paddle.vision.models import resnext101_64x4d >>> # build model >>> model = resnext101_64x4d() >>> # build model and load imagenet pretrained weight >>> # model = resnext101_64x4d(pretrained=True) >>> x = paddle.rand([1, 3, 224, 224]) >>> out = model(x) >>> print(out.shape) paddle.Size([1, 1000]) r;r rgrrr|rrs r'rrrr&c @d|d<d|d<tdtd|fi|S)aResNeXt-152 32x4d model from `"Aggregated Residual Transformations for Deep Neural 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:`ResNet `. Returns: :ref:`api_paddle_nn_Layer`. An instance of ResNeXt-152 32x4d model. Examples: .. code-block:: pycon >>> import paddle >>> from paddle.vision.models import resnext152_32x4d >>> # build model >>> model = resnext152_32x4d() >>> # build model and load imagenet pretrained weight >>> # model = resnext152_32x4d(pretrained=True) >>> x = paddle.rand([1, 3, 224, 224]) >>> out = model(x) >>> print(out.shape) paddle.Size([1, 1000]) rr rgrrr}rrs r'rrrr&c @d|d<d|d<tdtd|fi|S)aResNeXt-152 64x4d model from `"Aggregated Residual Transformations for Deep Neural 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:`ResNet `. Returns: :ref:`api_paddle_nn_Layer`. An instance of ResNeXt-152 64x4d model. Examples: .. code-block:: pycon >>> import paddle >>> from paddle.vision.models import resnext152_64x4d >>> # build model >>> model = resnext152_64x4d() >>> # build model and load imagenet pretrained weight >>> # model = resnext152_64x4d(pretrained=True) >>> x = paddle.rand([1, 3, 224, 224]) >>> out = model(x) >>> print(out.shape) paddle.Size([1, 1000]) r;r rgrrr}rrs r'rrrr&c 6d|d<tdtd|fi|S)aWide ResNet-50-2 model from `"Wide Residual 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:`ResNet `. Returns: :ref:`api_paddle_nn_Layer`. An instance of Wide ResNet-50-2 model. Examples: .. code-block:: pycon >>> import paddle >>> from paddle.vision.models import wide_resnet50_2 >>> # build model >>> model = wide_resnet50_2() >>> # build model and load imagenet pretrained weight >>> # model = wide_resnet50_2(pretrained=True) >>> x = paddle.rand([1, 3, 224, 224]) >>> out = model(x) >>> print(out.shape) paddle.Size([1, 1000]) rrrrorrs r'rr:s+@F7O $or: P P P PPr&c 6d|d<tdtd|fi|S)aWide ResNet-101-2 model from `"Wide Residual Networks" `_. Args: pretrained (bool, optional): Whether to load pre-trained weights. If True, returns a model pre-trained on ImageNet. Default: False. **kwargs (optional): Additional keyword arguments. For details, please refer to :ref:`ResNet `. Returns: :ref:`api_paddle_nn_Layer`. An instance of Wide ResNet-101-2 model. Examples: .. code-block:: pycon >>> import paddle >>> from paddle.vision.models import wide_resnet101_2 >>> # build model >>> model = wide_resnet101_2() >>> # build model and load imagenet pretrained weight >>> # model = wide_resnet101_2(pretrained=True) >>> x = paddle.rand([1, 3, 224, 224]) >>> out = model(x) >>> print(out.shape) paddle.Size([1, 1000]) rrrr|rrs r'rr^s6@F7O OS*  @F  r&) rrrrrrsr9rrnrrrErm)F)rrrErm)rrnrrrErm)( __future__rtypingrrrrpaddle.utils.downloadrrr typing_extensionsr r r paddle._typingr rr__all__r5r$Layerr7rfrmrrrrrrrrrrrrrrr%r&r'rs#""""""******** ;;;;;;!))))))))55555555$$$$$$ K !!!!!!!! c5*5* 5555p11111111h>>>>>bh>>>BjjjjjRXjjjZ,EEEEED E E E E EH J J J J JH L L L L LH L L L L LH"Q"Q"Q"Q"QL"Q"Q"Q"Q"QL$$$$$P$$$$$P$$$$$P$$$$$P!Q!Q!Q!Q!QJ#######r&