hT'TSSKJrJr SSKrS/rS\R S\R 4Sjr"SS\RR5r "S S \RR5r "S S \RR5r "S S\RR5r g))OptionalTupleN ConformerlengthsreturncURSn[[R"U5R 55n[R "X R URS9RX5URS5:nU$)Nr)devicedtype) shapeinttorchmaxitemaranger r expand unsqueeze)r batch_size max_length padding_masks S/var/www/auris/envauris/lib/python3.13/site-packages/torchaudio/models/conformer.py_lengths_to_padding_maskr smq!JUYYw',,./J<< >>W^^   1 L c^\rSrSrSrSS\S\S\S\S\S\S S 4U4S jjjrS \ RS \ R4S jr Sr U=r $)_ConvolutionModuleaConformer convolution module. Args: input_dim (int): input dimension. num_channels (int): number of depthwise convolution layer input channels. depthwise_kernel_size (int): kernel size of depthwise convolution layer. dropout (float, optional): dropout probability. (Default: 0.0) bias (bool, optional): indicates whether to add bias term to each convolution layer. (Default: ``False``) use_group_norm (bool, optional): use GroupNorm rather than BatchNorm. (Default: ``False``) input_dim num_channelsdepthwise_kernel_sizedropoutbiasuse_group_normrNc>[TU]5 US- S-S:wa [S5e[RR U5Ul[RR[RRUSU-SSSUS9[RRSS9[RRUUUSUS- S-UUS9U(a[RRSUS9O[RRU5[RR5[RRUUSSSUS 9[RRU55Ulg) Nr rz