o 7?e*@srdZddlmmZddlmZddlmZddlmZddlm Z ddl m Z ddl m Z Gd d d e ZdS) z7Keras abstract base layer for separable nD convolution.N) activations) constraints) initializers) regularizers) InputSpec)ConvcsdeZdZdZ                   dfdd Zd d Zd d ZfddZZS) SeparableConvaqAbstract base layer for separable nD convolution. This layer performs a depthwise convolution that acts separately on channels, followed by a pointwise convolution that mixes channels. If `use_bias` is True and a bias initializer is provided, it adds a bias vector to the output. It then optionally applies an activation function to produce the final output. Args: rank: An integer, the rank of the convolution, e.g. "2" for 2D convolution. filters: Integer, the dimensionality of the output space (i.e. the number of filters in the convolution). kernel_size: A tuple or list of integers specifying the spatial dimensions of the filters. Can be a single integer to specify the same value for all spatial dimensions. strides: A tuple or list of integers specifying the strides of the convolution. Can be a single integer to specify the same value for all spatial dimensions. Specifying any `stride` value != 1 is incompatible with specifying any `dilation_rate` value != 1. padding: One of `"valid"` or `"same"` (case-insensitive). `"valid"` means no padding. `"same"` results in padding with zeros evenly to the left/right or up/down of the input such that output has the same height/width dimension as the input. data_format: A string, one of `channels_last` (default) or `channels_first`. The ordering of the dimensions in the inputs. `channels_last` corresponds to inputs with shape `(batch_size, ..., channels)` while `channels_first` corresponds to inputs with shape `(batch_size, channels, ...)`. dilation_rate: An integer or tuple/list of 2 integers, specifying the dilation rate to use for dilated convolution. Can be a single integer to specify the same value for all spatial dimensions. Currently, specifying any `dilation_rate` value != 1 is incompatible with specifying any stride value != 1. depth_multiplier: The number of depthwise convolution output channels for each input channel. The total number of depthwise convolution output channels will be equal to `num_filters_in * depth_multiplier`. activation: Activation function to use. If you don't specify anything, no activation is applied (see `keras.activations`). use_bias: Boolean, whether the layer uses a bias. depthwise_initializer: An initializer for the depthwise convolution kernel (see `keras.initializers`). If None, then the default initializer ('glorot_uniform') will be used. pointwise_initializer: An initializer for the pointwise convolution kernel (see `keras.initializers`). If None, then the default initializer ('glorot_uniform') will be used. bias_initializer: An initializer for the bias vector. If None, the default initializer ('zeros') will be used (see `keras.initializers`). depthwise_regularizer: Optional regularizer for the depthwise convolution kernel. pointwise_regularizer: Optional regularizer for the pointwise convolution kernel. bias_regularizer: Optional regularizer for the bias vector. activity_regularizer: Optional regularizer function for the output. depthwise_constraint: Optional projection function to be applied to the depthwise kernel after being updated by an `Optimizer` (e.g. used for norm constraints or value constraints for layer weights). The function must take as input the unprojected variable and must return the projected variable (which must have the same shape). Constraints are not safe to use when doing asynchronous distributed training. pointwise_constraint: Optional projection function to be applied to the pointwise kernel after being updated by an `Optimizer`. bias_constraint: Optional projection function to be applied to the bias after being updated by an `Optimizer`. trainable: Boolean, if `True` the weights of this layer will be marked as trainable (and listed in `layer.trainable_weights`). validNTglorot_uniformzerosc stjd|||||||t| | t| t|t||||d|||_t| |_t| |_t||_ t||_ t ||_ t ||_ dS)N)rankfilters kernel_sizestridespadding data_format dilation_rate activationuse_biasbias_initializerbias_regularizeractivity_regularizerbias_constraint trainablename)super__init__rgetrrdepth_multiplierdepthwise_initializerpointwise_initializerdepthwise_regularizerpointwise_regularizerrdepthwise_constraintpointwise_constraint)selfr rrrrrrr rrr!r"rr#r$rrr%r&rrrkwargs __class__rs/home/www/facesmatcher.com/pyenv/lib/python3.10/site-packages/keras/src/layers/convolutional/base_separable_conv.pyres4     zSeparableConv.__init__c Cst|}|}|j|jdurtd|d|dt||}t|jd||id|_ |j ||j f}d|j|j ||j f}|j d||j|j|jd|jd |_|j d ||j|j|jd|jd |_|jr||j d |j f|j|j|jd|jd |_nd|_d|_dS) NzSThe channel dimension of the inputs should be defined. The input_shape received is z , where axis z> (0-based) is the channel dimension, which found to be `None`.)ndimZaxes)r depthwise_kernelT)rshapeZ initializerZ regularizer constraintrdtypepointwise_kernelbias)tfZ TensorShapeZ_get_channel_axisdimsvalue ValueErrorintrr Z input_specrr rZ add_weightr!r#r%r1r.r"r$r&r2rrrrr3Zbuilt)r'Z input_shapeZ channel_axisZ input_dimZdepthwise_kernel_shapeZpointwise_kernel_shaperrr+buildsh      zSeparableConv.buildcCst)N)NotImplementedError)r'Zinputsrrr+callszSeparableConv.callcsid|jd|jd|jd|jd|jd|jd|jdt|j d |j d t |j d t |j d t |jd t|jdt|jdt|jdt|jdt|jt|jt|jd}t}tt|t|S)Nrrrrrr rrrr!r"rr#r$rrr%)r&r)rrrrrr rr serializerrrr!r"rrr#r$rrrr%r&rr get_configdictlistitems)r'configZ base_configr)rr+r=sl        #zSeparableConv.get_config)r r Nr r NTr r r NNNNNNNTN) __name__ __module__ __qualname____doc__rr9r;r= __classcell__rrr)r+rs2M47r)rEZtensorflow.compat.v2compatv2r4Z keras.srcrrrrZkeras.src.engine.input_specrZ(keras.src.layers.convolutional.base_convrrrrrr+s