2VhZ1dZddlmZddlmZddlmZddlmZddlmZddlmZddl m Z dd l m Z dd l mZdd lmZdd lmZGd de Zy)z4Keras abstract base layer for separable convolution.) activations) constraints) initializers)ops) regularizers)standardize_data_format) InputSpec)Layer)compute_conv_output_shape)standardize_padding)standardize_tuplecdeZdZdZ dfd ZdZdZdZfdZxZ S)BaseSeparableConva)Abstract base layer for separable 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. Args: rank: int, the rank of the convolution, e.g. 2 for 2D convolution. 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 `input_channel * depth_multiplier`. filters: int, the dimensionality of the output space (i.e. the number of filters in the pointwise convolution). kernel_size: int or tuple/list of `rank` integers, specifying the size of the depthwise convolution window. strides: int or tuple/list of `rank` integers, specifying the stride length of the depthwise convolution. If only one int is specified, the same stride size will be used for all dimensions. `stride value != 1` is incompatible with `dilation_rate != 1`. padding: string, either `"valid"` or `"same"` (case-insensitive). `"valid"` means no padding. `"same"` results in padding evenly to the left/right or up/down of the input. When `padding="same"` and `strides=1`, the output has the same size as the input. data_format: string, either `"channels_last"` or `"channels_first"`. The ordering of the dimensions in the inputs. `"channels_last"` corresponds to inputs with shape `(batch, steps, features)` while `"channels_first"` corresponds to inputs with shape `(batch, features, steps)`. It defaults to the `image_data_format` value found in your Keras config file at `~/.keras/keras.json`. If you never set it, then it will be `"channels_last"`. dilation_rate: int or tuple/list of `rank` integers, specifying the dilation rate to use for dilated convolution. If only one int is specified, the same dilation rate will be used for all dimensions. activation: Activation function. If `None`, no activation is applied. use_bias: bool, if `True`, bias will be added to the output. depthwise_initializer: An initializer for the depthwise convolution kernel. If None, then the default initializer (`"glorot_uniform"`) will be used. pointwise_initializer: An initializer for the pointwise convolution kernel. 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. 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). 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`. c t|d||tj|d|||_||_||_t||d|_t||d|_ t||d|_ t||_ t||_tj| |_| |_t%j| |_t%j| |_t%j| |_tj||_tj||_tj||_t3j||_t3j||_t3j||_|j|_t;|jdz|_|j (|j dkrt?d|j d |j (|j dkrt?d |j d tA|jst?d |jd tA|jst?d |jtC|jd kDr>tC|jd kDr%t?d|jd|jyy)N) trainablenameactivity_regularizer kernel_sizestrides dilation_rate)min_ndimrzmInvalid value for argument `depth_multiplier`. Expected a strictly positive value. Received depth_multiplier=.z[Invalid value for argument `filters`. Expected a strictly positive value. Received filters=zCThe argument `kernel_size` cannot contain 0. Received: kernel_size=z?The argument `strides` cannot contains 0(s). Received: strides=zW`strides > 1` not supported in conjunction with `dilation_rate > 1`. Received: strides=z and dilation_rate=)"super__init__rgetrankdepth_multiplierfiltersr rrrr paddingr data_formatr activationuse_biasrdepthwise_initializerpointwise_initializerbias_initializerdepthwise_regularizerpointwise_regularizerbias_regularizerrdepthwise_constraintpointwise_constraintbias_constraintr input_spec ValueErrorallmax)selfrr r!rrr"r#rr$r%r&r'r(r)r*r+rr,r-r.rrkwargs __class__s b/home/dcms/DCMS/lib/python3.12/site-packages/keras/src/layers/convolutional/base_separable_conv.pyrzBaseSeparableConv.__init__Ns4  !-!1!12F!G    0 ,[$ N($ B . 4 +73 2;?%//*5  %1%5%56K%L"%1%5%56K%L" , 0 01A B%1%5%56K%L"%1%5%56K%L" , 0 01A B$/OO4H$I!$/OO4H$I!*?++#TYY];  ,1F1F!1K$$($9$9#:!=  << # (9448LL>D  4##$#//03  4<< <<.*  t|| q S););%,,H!!%!3!3 46 &A c |jdk(rd}|d}nd}|d}t|jdz||i|_|j||j fz}d|jz|j |z|j fz}|jd||j|j|jd|j |_ |jd ||j|j|jd|j |_|j"rQ|jd |j f|j$|j&|j(d|j |_yd|_y) N channels_lastrr)raxesrdepthwise_kernelT)rshape initializer regularizer constraintrdtypepointwise_kernelbias)r#r rr/rr r! add_weightr&r)r,rBr=r'r*r-rCr%r(r+r.rD)r3 input_shape channel_axis input_channeldepthwise_kernel_shapepointwise_kernel_shapes r6buildzBaseSeparableConv.builds|    .L'OML'NM#YY], )F "&!1!1   ! !5 " "& !1  ! !M 1 LL5 "  !%#(222200**!0! !%#(222200**!0!  ==||o 11 11//jj(DIDIr7c tj||j|j|j|j |j |j}|jr|jdk(r!d|jdzz|jfz}nd|jfd|jzz}tj|j|}tj||}|j|j|S|S)N)rr"rr#r9r<r)rseparable_convr=rCrr"rr#r%rr!reshaperDaddr$)r3inputsoutputs bias_shaperDs r6callzBaseSeparableConv.calls$$   ! !  ! !LLLL,,((  ==?2!TYY]3t||oE . 1AA ;;tyy*5Dgggt,G ?? &??7+ +r7c t||j|j|j|j|j |j S)N)rr"r#r)r r!rrr"r#r)r3rFs r6compute_output_shapez&BaseSeparableConv.compute_output_shapesB(  LL   LLLL((,,  r7ct|}|jid|jd|jd|j d|j d|jd|jd|jdtj|jd |jd tj|jd tj|j d tj|j"d t%j|j&dt%j|j(dt%j|j*dt%j|j,dt/j|j0t/j|j2t/j|j4d|S)Nr r!rrr"r#rr$r%r&r'r(r)r*r+rr,)r-r.)r get_configupdater r!rrr"r#rr serializer$r%rr&r'r(rr)r*r+rrr,r-r.)r3configr5s r6rWzBaseSeparableConv.get_configs#% & "D$9$9& 4<<& t//& 4<< &  4<< &  t// &  !3!3& k33DOOD& DMM& ()?)?..*& ()?)?..*&  #L$:$:))%!& &()?)?..*'& ,()?)?..*-& 2#L$:$:))%3& 8' (>(>--)9& >' (=(=--)?& D)4(=(=--)$/#8#89M9M#NK& ( R r7)rvalidNrNTglorot_uniformr\zerosNNNNNNNTN) __name__ __module__ __qualname____doc__rrKrSrUrW __classcell__)r5s@r6rrsb;F.. ""!!! /Wr0d.  ++r7rN)ra keras.srcrrrrrkeras.src.backendrkeras.src.layers.input_specr keras.src.layers.layerr keras.src.ops.operation_utilsr #keras.src.utils.argument_validationr r rrr7r6ris7:!!""51(CCAVVr7