o #jy@sddlZddlmZddlmZddlmZddlmZgZdd iZ Gd d d ej Z Gd d d ej Z dddZ dddZdS)N)nn)get_weights_path_from_url)ConvNormActivation)_make_divisiblezmobilenetv2_1.0)zChttps://paddle-hapi.bj.bcebos.com/models/mobilenet_v2_x1.0.pdparamsZ 0340af0a901346c8d46f4529882fb63dcs*eZdZejffdd ZddZZS)InvertedResidualc st||_|dvsJtt||}|jdko||k|_g}|dkr3|t||d|tj d| t|||||tj dtj ||ddddd||gtj ||_ dS)N)rrrZ kernel_size norm_layeractivation_layer)stridegroupsr r rF)Z bias_attr)super__init__r introunduse_res_connectappendrrReLU6extendZConv2D Sequentialconv)selfinpZoupr expand_ratior Z hidden_dimZlayers __class__a/var/www/html/Deteccion_Ine/venv/lib/python3.10/site-packages/paddle/vision/models/mobilenetv2.pyr!s<   zInvertedResidual.__init__cCs|jr |||S||S)N)rrrxrrrforwardFs zInvertedResidual.forward)__name__ __module__ __qualname__r BatchNorm2Drr! __classcell__rrrrr s%rcs*eZdZdZd fdd ZddZZS) MobileNetV2a[MobileNetV2 model from `"MobileNetV2: Inverted Residuals and Linear Bottlenecks" `_. Args: scale (float, optional): Scale of channels in each layer. Default: 1.0. 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 MobileNetV2 model. Examples: .. code-block:: python >>> import paddle >>> from paddle.vision.models import MobileNetV2 >>> model = MobileNetV2() >>> x = paddle.rand([1, 3, 224, 224]) >>> out = model(x) >>> print(out.shape) [1, 1000] ?Tc sVt||_||_d}d}t}d}tj}gdgdgdgdgdgd gd g} t|||}t|td |||_ t d |d |tj dg} | D]*\} } } }t| ||}t | D]}|dkre|nd}| ||||| |d|}q]qL| t ||j d|tj dtj| |_|rtd|_|jdkrttdt|j ||_dSdS)N i)rrr)rr)r-r*r)r-@r)r-`r/r)r-r/r)r-i@rrr(r/r)r r r rr)rr r g?)rr num_classes with_poolrrr%rmax last_channelrrrangerrfeaturesZAdaptiveAvgPool2D pool2d_avgZDropoutZLinear classifier)rscaler4r5Z input_channelr7blockZ round_nearestr Zinverted_residual_settingr9tcnsZoutput_channelir rrrrisx          zMobileNetV2.__init__cCs>||}|jr ||}|jdkrt|d}||}|S)Nrr)r9r5r:r4paddleflattenr;rrrrr!s     zMobileNetV2.forward)r(r)T)r"r#r$__doc__rr!r&rrrrr'MsEr'FcKsZtdi|}|r+|tvsJ|dtt|dt|d}t|}|||S)NzJ model do not have a pretrained model now, you should set pretrained=Falserrr)r' model_urlsrrCload load_dict)arch pretrainedkwargsmodelZ weight_pathparamrrr _mobilenets   rNr(cKs"tdt||fd|i|}|S)aMobileNetV2 from `"MobileNetV2: Inverted Residuals and Linear Bottlenecks" `_. Args: pretrained (bool, optional): Whether to load pre-trained weights. If True, returns a model pre-trained on ImageNet. Default: False. scale (float, optional): Scale of channels in each layer. Default: 1.0. **kwargs (optional): Additional keyword arguments. For details, please refer to :ref:`MobileNetV2 `. Returns: :ref:`api_paddle_nn_Layer`. An instance of MobileNetV2 model. Examples: .. code-block:: python >>> import paddle >>> from paddle.vision.models import mobilenet_v2 >>> # Build model >>> model = mobilenet_v2() >>> # Build model and load imagenet pretrained weight >>> # model = mobilenet_v2(pretrained=True) >>> # Build mobilenet v2 with scale=0.5 >>> model = mobilenet_v2(scale=0.5) >>> x = paddle.rand([1, 3, 224, 224]) >>> out = model(x) >>> print(out.shape) [1, 1000] Z mobilenetv2_r<)rNstr)rJr<rKrLrrr mobilenet_v2s! rP)F)Fr()rCrZpaddle.utils.downloadropsrZ_utilsr__all__rFZLayerrr'rNrPrrrrs    - m