hZNSSKrSSKJrJrJrJr SSKrSSKJrJr SSK J r S/r S"S\ S\ S\S \S \RR 4 S jjrS#S \ S \ S\ S\ S\\\\ \\ 4S\ S\S \S \RR$4SjjrS\S \4Sjr"SS\R*5r"SS\R*5r"SS\R*5r"SS\R*5r"SS\R*5r"SS \R*5r"S!S\R*5rg)$N)ListOptionalTupleUnion)nnTensor) functional Tacotron2in_dimout_dimbias w_init_gainreturnc[RRXUS9n[RRR UR [RRR U5S9 U$)aLinear layer with xavier uniform initialization. Args: in_dim (int): Size of each input sample. out_dim (int): Size of each output sample. bias (bool, optional): If set to ``False``, the layer will not learn an additive bias. (Default: ``True``) w_init_gain (str, optional): Parameter passed to ``torch.nn.init.calculate_gain`` for setting the gain parameter of ``xavier_uniform_``. (Default: ``linear``) Returns: (torch.nn.Linear): The corresponding linear layer. r gain)torchrLinearinitxavier_uniform_weightcalculate_gain)r r r rlinears S/var/www/auris/envauris/lib/python3.13/site-packages/torchaudio/models/tacotron2.py_get_linear_layerr)sRXX__V4_ 8F HHMM!!&--ehhmm6R6RS^6_!` M in_channels out_channels kernel_sizestridepaddingdilationc RUc'US-S:wa [S5e[XRS- -S- 5n[RR UUUUUUUS9n[RR R UR[RR RU5S9 U$)a81D convolution with xavier uniform initialization. Args: in_channels (int): Number of channels in the input image. out_channels (int): Number of channels produced by the convolution. kernel_size (int, optional): Number of channels in the input image. (Default: ``1``) stride (int, optional): Number of channels in the input image. (Default: ``1``) padding (str, int or tuple, optional): Padding added to both sides of the input. (Default: dilation * (kernel_size - 1) / 2) dilation (int, optional): Number of channels in the input image. (Default: ``1``) w_init_gain (str, optional): Parameter passed to ``torch.nn.init.calculate_gain`` for setting the gain parameter of ``xavier_uniform_``. (Default: ``linear``) Returns: (torch.nn.Conv1d): The corresponding Conv1D layer. zkernel_size must be odd)r r!r"r#r r) ValueErrorintrrConv1drrrr) rrr r!r"r#r rconv1ds r_get_conv1d_layerr+;s4 ?a 67 7h/2Q67 XX__ F HHMM!!&--ehhmm6R6RS^6_!` Mrlengthsc[R"U5R5n[R"SXRUR S9nX R S5:R5n[R"US5nU$)aTReturns a binary mask based on ``lengths``. The ``i``-th row and ``j``-th column of the mask is ``1`` if ``j`` is smaller than ``i``-th element of ``lengths. Args: lengths (Tensor): The length of each element in the batch, with shape (n_batch, ). Returns: mask (Tensor): The binary mask, with shape (n_batch, max of ``lengths``). r)devicedtyper&) rmaxitemaranger.r/ unsqueezebytele)r,max_lenidsmasks r_get_mask_from_lengthsr9isfii %%'G ,,q'.. NC ##A& & , , .D 88D! D KrcN^\rSrSrSrS\S\S\4U4SjjrS\S\4S jrS r U=r $) _LocationLayerzaLocation layer used in the Attention model. Args: attention_n_filter (int): Number of filters for attention model. attention_kernel_size (int): Kernel size for attention model. attention_hidden_dim (int): Dimension of attention hidden representation. attention_n_filterattention_kernel_sizeattention_hidden_dimc >[TU]5 [US- S- 5n[SUUUSSSS9Ul[ XSSS9Ulg)Nr&r%F)r r"r r!r#tanhr r)super__init__r(r+ location_convrlocation_dense)selfr=r>r?r" __class__s rrD_LocationLayer.__init__s] ,q0A56. - 0 5f rattention_weights_catrcnURU5nURSS5nURU5nU$)aILocation layer used in the Attention model. Args: attention_weights_cat (Tensor): Cumulative and previous attention weights with shape (n_batch, 2, max of ``text_lengths``). Returns: processed_attention (Tensor): Cumulative and previous attention weights with shape (n_batch, ``attention_hidden_dim``). r&r%)rE transposerF)rGrJprocessed_attentions rforward_LocationLayer.forwardsA#001FG1;;AqA"112EF""r)rErF __name__ __module__ __qualname____firstlineno____doc__r(rDrrN__static_attributes__ __classcell__rHs@rr;r;zsA   # "  *#V###rr;c^\rSrSrSrS\S\S\S\S\SS 4 U4S jjrS \S \S \S\4SjrS\S\S \S \S\S\ \\44 Sjr Sr U=r $) _AttentionaLocally sensitive attention model. Args: attention_rnn_dim (int): Number of hidden units for RNN. encoder_embedding_dim (int): Number of embedding dimensions in the Encoder. attention_hidden_dim (int): Dimension of attention hidden representation. attention_location_n_filter (int): Number of filters for Attention model. attention_location_kernel_size (int): Kernel size for Attention model. attention_rnn_dimencoder_embedding_dimr?attention_location_n_filterattention_location_kernel_sizerNc>[TU]5 [XSSS9Ul[X#SSS9Ul[USSS9Ul[ UUU5Ul[S5*Ul g)NFrArBr&rinf) rCrDr query_layer memory_layervr;location_layerfloatscore_mask_value)rGr\r]r?r^r_rHs rrD_Attention.__init__sr ,->[`ntu- !eQW ##7G, ' *   "'u rqueryprocessed_memoryrJcURURS55nURU5nUR[R "XE-U-55nUR S5nU$)aGet the alignment vector. Args: query (Tensor): Decoder output with shape (n_batch, n_mels * n_frames_per_step). processed_memory (Tensor): Processed Encoder outputs with shape (n_batch, max of ``text_lengths``, attention_hidden_dim). attention_weights_cat (Tensor): Cumulative and previous attention weights with shape (n_batch, 2, max of ``text_lengths``). Returns: alignment (Tensor): attention weights, it is a tensor with shape (batch, max of ``text_lengths``). r&r%)rbr3rerdrrAsqueeze)rGrirjrJprocessed_queryprocessed_attention_weightsenergies alignments r_get_alignment_energies"_Attention._get_alignment_energiesse**5??1+=>&*&9&9:O&P#66%**_%RUe%efg$$Q' rattention_hidden_statememoryr8cURXU5nURXPR5n[R"USS9n[ R "URS5U5nURS5nX4$)aYPass the input through the Attention model. Args: attention_hidden_state (Tensor): Attention rnn last output with shape (n_batch, ``attention_rnn_dim``). memory (Tensor): Encoder outputs with shape (n_batch, max of ``text_lengths``, ``encoder_embedding_dim``). processed_memory (Tensor): Processed Encoder outputs with shape (n_batch, max of ``text_lengths``, ``attention_hidden_dim``). attention_weights_cat (Tensor): Previous and cumulative attention weights with shape (n_batch, current_num_frames * 2, max of ``text_lengths``). mask (Tensor): Binary mask for padded data with shape (n_batch, current_num_frames). Returns: attention_context (Tensor): Context vector with shape (n_batch, ``encoder_embedding_dim``). attention_weights (Tensor): Attention weights with shape (n_batch, max of ``text_lengths``). r&dim) rq masked_fillrgFsoftmaxrbmmr3rl) rGrsrtrjrJr8rpattention_weightsattention_contexts rrN_Attention.forwardsv.001G[pq ))$0E0EF IIiQ7!II&7&A&A!&DfM-55a8 33r)rercrbrgrd) rQrRrSrTrUr(rDrrqrrNrVrWrXs@rrZrZs.. #." . &) . ), . .*Vvflqw*4 &44! 4 & 4  4 vv~ 44rrZcT^\rSrSrSrS\S\\SS4U4SjjrS\S\4S jr S r U=r $) _PrenetizPrenet Module. It is consists of ``len(output_size)`` linear layers. Args: in_dim (int): The size of each input sample. output_sizes (list): The output dimension of each linear layers. r out_sizesrNc >[TU]5 U/USS-n[R"[ X25VVs/sHupE[ XESS9PM snn5Ulgs snnf)NFr)rCrDr ModuleListziprlayers)rGr rin_sizesin_sizeout_sizerHs rrD_Prenet.__init__ sX 8in,mmY\]eYq rYqBU7 wu =Yq r  rsA xcURH3n[R"[R"U"U55SSS9nM5 U$)zPass the input through Prenet. Args: x (Tensor): The input sequence to Prenet with shape (n_batch, in_dim). Return: x (Tensor): Tensor with shape (n_batch, sizes[-1]) ?T)ptraining)rrydropoutrelu)rGrrs rrN_Prenet.forwards6kkF !&&+sTBA"r)r) rQrRrSrTrUr(rrDrrNrVrWrXs@rrrs> s tCy T   F  rrcR^\rSrSrSrS\S\S\S\4U4SjjrS\S \4S jrS r U=r $) _Postneti zPostnet Module. Args: n_mels (int): Number of mel bins. postnet_embedding_dim (int): Postnet embedding dimension. postnet_kernel_size (int): Postnet kernel size. postnet_n_convolution (int): Number of postnet convolutions. n_melspostnet_embedding_dimpostnet_kernel_sizepostnet_n_convolutionc>[T U]5 [R"5Ul[ U5HnUS:XaUOUnXTS- :XaUOUnXTS- :XaSOSnXTS- :XaUOUn URR [R"[UUUS[US- S- 5SUS9[R"U 555 M [UR5Ul g)Nrr&rrAr%r r!r"r#r) rCrDrr convolutionsrangeappend Sequentialr+r( BatchNorm1dlenn_convs) rGrrrrirr init_gain num_featuresrHs rrD_Postnet.__init__*s MMO,-A$%F&0EK%&1+D%E6K`L$%!*C$D&I%&1+D%E6K`L    $ $ %#$$7 #%81%<$A B!"$-NN<0   .(4,,- rrrc*[UR5Hyup#X RS- :a<[R"[ R "U"U55SURS9nMS[R"U"U5SURS9nM{ U$)zPass the input through Postnet. Args: x (Tensor): The input sequence with shape (n_batch, ``n_mels``, max of ``mel_specgram_lengths``). Return: x (Tensor): Tensor with shape (n_batch, ``n_mels``, max of ``mel_specgram_lengths``). r&r)r) enumeraterrryrrrAr)rGrrconvs rrN_Postnet.forwardJsn!!2!23GA<>> encoder = _Encoder(3, 512, 5) >>> input = torch.rand(10, 20, 30) >>> output = encoder(input) # shape: (10, 30, 512) r]encoder_n_convolutionencoder_kernel_sizerNc >[TU]5 [R"5Ul[ U5Hen[R "[UUUS[US- S- 5SSS9[R"U55nURRU5 Mg [R"U[US- 5SSSS9Ul URR5 g)Nr&r%rrT) batch_first bidirectional)rCrDrrrrrr+r(rrLSTMlstmflatten_parameters)rGr]rr_ conv_layerrHs rrD_Encoder.__init__ks MMO,-A!)) 3!4q!8A => &45 J    $ $Z 0.GG ! %) *     $$&rr input_lengthscURH?n[R"[R"U"U55SUR5nMA UR SS5nUR 5n[RRRXSS9nURU5upE[RRRUSS9upEU$)a/Pass the input through the Encoder. Args: x (Tensor): The input sequences with shape (n_batch, encoder_embedding_dim, n_seq). input_lengths (Tensor): The length of each input sequence with shape (n_batch, ). Return: x (Tensor): A tensor with shape (n_batch, n_seq, encoder_embedding_dim). rr&r%T)r) rryrrrrLcpurutilsrnnpack_padded_sequencerpad_packed_sequence)rGrrroutputsrs rrN_Encoder.forwards%%D !&&a/3 >A& KK1 %))+ HHLL - -aD - QYYq\ XX\\55g45P r)rrrPrXs@rrr]sS '"' #'! '  'B6rrc!^\rSrSrSrS\S\S\S\S\S\S \S \S \S \S \S\S\S\SS4U4SjjrS\ S\ 4Sjr S\ S\ \ \ \ \ \ \ \ \ 44Sjr S\ S\ 4Sjr S\ S\ S\ S\ \ \ \ 44SjrS\ S\ S\ S \ S!\ S"\ S#\ S$\ S\ S%\ S&\ S\ \ \ \ \ \ \ \ \ \ 4 4S'jrS\ S(\ S)\ S\ \ \ \ 44S*jrS\ S\ 4S+jr\R&R(S\ S)\ S\ \ \ \ \ 44S,j5rS-rU=r$)._DecoderiaDecoder with Attention model. Args: n_mels (int): number of mel bins n_frames_per_step (int): number of frames processed per step, only 1 is supported encoder_embedding_dim (int): the number of embedding dimensions in the encoder. decoder_rnn_dim (int): number of units in decoder LSTM decoder_max_step (int): maximum number of output mel spectrograms decoder_dropout (float): dropout probability for decoder LSTM decoder_early_stopping (bool): stop decoding when all samples are finished attention_rnn_dim (int): number of units in attention LSTM attention_hidden_dim (int): dimension of attention hidden representation attention_location_n_filter (int): number of filters for attention model attention_location_kernel_size (int): kernel size for attention model attention_dropout (float): dropout probability for attention LSTM prenet_dim (int): number of ReLU units in prenet layers gate_threshold (float): probability threshold for stop token rn_frames_per_stepr]decoder_rnn_dimdecoder_max_stepdecoder_dropoutdecoder_early_stoppingr\r?r^r_attention_dropout prenet_dimgate_thresholdrNc>[TU]5 XlX lX0lXlX@lXlXPlXl Xl X`l Xpl [X-X/5Ul[R "X-U5Ul[%UUU U U 5Ul[R "X-US5Ul[+XC-X-5Ul[+XC-SSSS9Ulg)NTr&sigmoidrB)rCrDrrr]r\rrrrrrrrprenetrLSTMCell attention_rnnrZattention_layer decoder_rnnrlinear_projection gate_layer)rGrrr]rrrrr\r?r^r_rrrrHs rrD_Decoder.__init__s$  !2%:"!2.$ 0,!2.&<#f8::RS [[)KM^_)  ! ' *  ;;'8'PRacgh!2?3Z\b\v!w+  3QTy rrtcURS5nURnURn[R"X R UR -X4S9nU$)a=Gets all zeros frames to use as the first decoder input. Args: memory (Tensor): Encoder outputs with shape (n_batch, max of ``text_lengths``, ``encoder_embedding_dim``). Returns: decoder_input (Tensor): all zeros frames with shape (n_batch, max of ``text_lengths``, ``n_mels * n_frames_per_step``). rr/r.sizer/r.rzerosrrrGrtn_batchr/r. decoder_inputs r_get_initial_frame_Decoder._get_initial_frameI++a.  G[[4;Q;Q-QY^n rcJURS5nURS5nURnURn[R"X R XES9n[R"X R XES9n[R"X R XES9n[R"X R XES9n [R"X#XES9n [R"X#XES9n [R"X RXES9n URRU5n UUUU U U U U 4$)aInitializes attention rnn states, decoder rnn states, attention weights, attention cumulative weights, attention context, stores memory and stores processed memory. Args: memory (Tensor): Encoder outputs with shape (n_batch, max of ``text_lengths``, ``encoder_embedding_dim``). Returns: attention_hidden (Tensor): Hidden state of the attention LSTM with shape (n_batch, ``attention_rnn_dim``). attention_cell (Tensor): Hidden state of the attention LSTM with shape (n_batch, ``attention_rnn_dim``). decoder_hidden (Tensor): Hidden state of the decoder LSTM with shape (n_batch, ``decoder_rnn_dim``). decoder_cell (Tensor): Hidden state of the decoder LSTM with shape (n_batch, ``decoder_rnn_dim``). attention_weights (Tensor): Attention weights with shape (n_batch, max of ``text_lengths``). attention_weights_cum (Tensor): Cumulated attention weights with shape (n_batch, max of ``text_lengths``). attention_context (Tensor): Context vector with shape (n_batch, ``encoder_embedding_dim``). processed_memory (Tensor): Processed encoder outputs with shape (n_batch, max of ``text_lengths``, ``attention_hidden_dim``). rr&r) rr/r.rrr\rr]rrc)rGrtrmax_timer/r.attention_hiddenattention_celldecoder_hidden decoder_cellr|attention_weights_cumr}rjs r_initialize_decoder_states#_Decoder._initialize_decoder_statess*++a.;;q>  ;;w0F0FecW.D.DEaW.B.B%_{{7,@,@] !KKV % GU Z!KK1K1KSXh//<>@ #--a3>>@ ##A&DKK8#((%0 #--a3 :55rrrrrrr|rr}rjr8c [R"X4S5n URXU45up#[R"X R UR 5n[R"URS5URS54SS9n URX)XU 5upXv- n[R"X(4S5nURXU45upE[R"X@RUR 5n[R"XH4SS9nURU5nURU5nUUUUUUUUU4 $)afDecoder step using stored states, attention and memory Args: decoder_input (Tensor): Output of the Prenet with shape (n_batch, ``prenet_dim``). attention_hidden (Tensor): Hidden state of the attention LSTM with shape (n_batch, ``attention_rnn_dim``). attention_cell (Tensor): Hidden state of the attention LSTM with shape (n_batch, ``attention_rnn_dim``). decoder_hidden (Tensor): Hidden state of the decoder LSTM with shape (n_batch, ``decoder_rnn_dim``). decoder_cell (Tensor): Hidden state of the decoder LSTM with shape (n_batch, ``decoder_rnn_dim``). attention_weights (Tensor): Attention weights with shape (n_batch, max of ``text_lengths``). attention_weights_cum (Tensor): Cumulated attention weights with shape (n_batch, max of ``text_lengths``). attention_context (Tensor): Context vector with shape (n_batch, ``encoder_embedding_dim``). memory (Tensor): Encoder output with shape (n_batch, max of ``text_lengths``, ``encoder_embedding_dim``). processed_memory (Tensor): Processed Encoder outputs with shape (n_batch, max of ``text_lengths``, ``attention_hidden_dim``). mask (Tensor): Binary mask for padded data with shape (n_batch, current_num_frames). Returns: decoder_output: Predicted mel spectrogram for the current frame with shape (n_batch, ``n_mels``). gate_prediction (Tensor): Prediction of the stop token with shape (n_batch, ``1``). attention_hidden (Tensor): Hidden state of the attention LSTM with shape (n_batch, ``attention_rnn_dim``). attention_cell (Tensor): Hidden state of the attention LSTM with shape (n_batch, ``attention_rnn_dim``). decoder_hidden (Tensor): Hidden state of the decoder LSTM with shape (n_batch, ``decoder_rnn_dim``). decoder_cell (Tensor): Hidden state of the decoder LSTM with shape (n_batch, ``decoder_rnn_dim``). attention_weights (Tensor): Attention weights with shape (n_batch, max of ``text_lengths``). attention_weights_cum (Tensor): Cumulated attention weights with shape (n_batch, max of ``text_lengths``). attention_context (Tensor): Context vector with shape (n_batch, ``encoder_embedding_dim``). rr&rv) rcatrryrrrr3rrrrr)rGrrrrrr|rr}rtrjr8 cell_inputrJ decoder_hidden_attention_contextdecoder_outputgate_predictions rdecode_Decoder.decodecsHRYY A2F +/+=+=j]kJl+m(99%57M7Mt}}] % +<+F+Fq+IK`KjKjklKm*ntu v/3/C/C &6t0 , 2 #3"GL '+'7'7 XdGe'f$>3G3GW+099n5X^_+`(//0PQ//*JK        !   rmel_specgram_truthmemory_lengthsc URU5RS5nURU5n[R"XE4SS9nUR U5n[ U5nURU5unnn n n n n n///nnn[U5URS5S- :aU[U5nURUUUU U U U U UUU5 u nnnnn n n n n UURS5/- nUURS5/- nUU /- n[U5URS5S- :aMUR[R"U5[R"U5[R"U55unnnUUU4$)avDecoder forward pass for training. Args: memory (Tensor): Encoder outputs with shape (n_batch, max of ``text_lengths``, ``encoder_embedding_dim``). mel_specgram_truth (Tensor): Decoder ground-truth mel-specs for teacher forcing with shape (n_batch, ``n_mels``, max of ``mel_specgram_lengths``). memory_lengths (Tensor): Encoder output lengths for attention masking (the same as ``text_lengths``) with shape (n_batch, ). Returns: mel_specgram (Tensor): Predicted mel spectrogram with shape (n_batch, ``n_mels``, max of ``mel_specgram_lengths``). gate_outputs (Tensor): Predicted stop token for each timestep with shape (n_batch, max of ``mel_specgram_lengths``). alignments (Tensor): Sequence of attention weights from the decoder with shape (n_batch, max of ``mel_specgram_lengths``, max of ``text_lengths``). rrvr&)rr3rrrrr9rrrrrlrstack)rGrtrrrrr8rrrrr|rr}rj mel_outputsrr mel_output gate_outputrs rrN_Decoder.forwards,//7AA!D 334FGM#BJ^4%n5  + +F 3      !  13B:\ +!4!4Q!7!!;;*3{+;`_. See Also: * :class:`torchaudio.pipelines.Tacotron2TTSBundle`: TTS pipeline with pretrained model. Args: mask_padding (bool, optional): Use mask padding (Default: ``False``). n_mels (int, optional): Number of mel bins (Default: ``80``). n_symbol (int, optional): Number of symbols for the input text (Default: ``148``). n_frames_per_step (int, optional): Number of frames processed per step, only 1 is supported (Default: ``1``). symbol_embedding_dim (int, optional): Input embedding dimension (Default: ``512``). encoder_n_convolution (int, optional): Number of encoder convolutions (Default: ``3``). encoder_kernel_size (int, optional): Encoder kernel size (Default: ``5``). encoder_embedding_dim (int, optional): Encoder embedding dimension (Default: ``512``). decoder_rnn_dim (int, optional): Number of units in decoder LSTM (Default: ``1024``). decoder_max_step (int, optional): Maximum number of output mel spectrograms (Default: ``2000``). decoder_dropout (float, optional): Dropout probability for decoder LSTM (Default: ``0.1``). decoder_early_stopping (bool, optional): Continue decoding after all samples are finished (Default: ``True``). attention_rnn_dim (int, optional): Number of units in attention LSTM (Default: ``1024``). attention_hidden_dim (int, optional): Dimension of attention hidden representation (Default: ``128``). attention_location_n_filter (int, optional): Number of filters for attention model (Default: ``32``). attention_location_kernel_size (int, optional): Kernel size for attention model (Default: ``31``). attention_dropout (float, optional): Dropout probability for attention LSTM (Default: ``0.1``). prenet_dim (int, optional): Number of ReLU units in prenet layers (Default: ``256``). postnet_n_convolution (int, optional): Number of postnet convolutions (Default: ``5``). postnet_kernel_size (int, optional): Postnet kernel size (Default: ``5``). postnet_embedding_dim (int, optional): Postnet embedding dimension (Default: ``512``). gate_threshold (float, optional): Probability threshold for stop token (Default: ``0.5``). mask_paddingrn_symbolrsymbol_embedding_dimr]rrrrrrr\r?r^r_rrrrrrrNcx>[TU]5 XlX lX@l[ R "X55Ul[R RRURR5 [XgU5Ul [UUUU U U U U UUUUUU5Ul[!UUUU5UlgN)rCrDrrrr Embedding embeddingrrrrrencoderrdecoderrpostnet)rGrrrrrr]rrrrrrr\r?r^r_rrrrrrrHs rrDTacotron2.__init__s2 ( !2hE  %%dnn&;&;< 5Nab    !    "  ' *      (=?RTij rtokens token_lengthsrrcURU5RSS5nURXR5nURXcUS9up7nUR U5n X9-n UR (a[ U5n U RURU RS5U RS55n U RSSS5n URU S5 U RU S5 URU SS2SSS24S5 X9Xx4$)a Pass the input through the Tacotron2 model. This is in teacher forcing mode, which is generally used for training. The input ``tokens`` should be padded with zeros to length max of ``token_lengths``. The input ``mel_specgram`` should be padded with zeros to length max of ``mel_specgram_lengths``. Args: tokens (Tensor): The input tokens to Tacotron2 with shape `(n_batch, max of token_lengths)`. token_lengths (Tensor): The valid length of each sample in ``tokens`` with shape `(n_batch, )`. mel_specgram (Tensor): The target mel spectrogram with shape `(n_batch, n_mels, max of mel_specgram_lengths)`. mel_specgram_lengths (Tensor): The length of each mel spectrogram with shape `(n_batch, )`. Returns: [Tensor, Tensor, Tensor, Tensor]: Tensor Mel spectrogram before Postnet with shape `(n_batch, n_mels, max of mel_specgram_lengths)`. Tensor Mel spectrogram after Postnet with shape `(n_batch, n_mels, max of mel_specgram_lengths)`. Tensor The output for stop token at each time step with shape `(n_batch, max of mel_specgram_lengths)`. Tensor Sequence of attention weights from the decoder with shape `(n_batch, max of mel_specgram_lengths, max of token_lengths)`. r&r%)rrgNg@@) rrLrrr rr9expandrrpermute masked_fill_) rGr"r#rrembedded_inputsencoder_outputsrrmel_specgram_postnetr8s rrNTacotron2.forwardsB..0::1a@,,F15 -2>2 . J $||L9+B   )*>?D;;t{{DIIaL$))A,GD<<1a(D  % %dC 0 - -dC 8  % %d1a7mS 9<KKrr,cURup4UcJ[R"U/5RU5R UR UR 5nUceURU5RSS5nURXR5nURRXb5upxpURU5n X{-n U RSX35RSS5n XU 4$)aUsing Tacotron2 for inference. The input is a batch of encoded sentences (``tokens``) and its corresponding lengths (``lengths``). The output is the generated mel spectrograms, its corresponding lengths, and the attention weights from the decoder. The input `tokens` should be padded with zeros to length max of ``lengths``. Args: tokens (Tensor): The input tokens to Tacotron2 with shape `(n_batch, max of lengths)`. lengths (Tensor or None, optional): The valid length of each sample in ``tokens`` with shape `(n_batch, )`. If ``None``, it is assumed that the all the tokens are valid. Default: ``None`` Returns: (Tensor, Tensor, Tensor): Tensor The predicted mel spectrogram with shape `(n_batch, n_mels, max of mel_specgram_lengths)`. Tensor The length of the predicted mel spectrogram with shape `(n_batch, )`. Tensor Sequence of attention weights from the decoder with shape `(n_batch, max of mel_specgram_lengths, max of lengths)`. r&r%r)rrtensorr%tor.r/rrLrrrr unfold) rGr"r,r max_lengthr(r)rrrrmel_outputs_postnets rrTacotron2.infers2%ll ?llJ<077@CCFMMSYS_S_`G"""..0::1a@,,@<@LLrDs|8// $  cCtQTdidldldsds*59++++  + eCeCj01 2 +  + ++ XX__+\Fv".#RYY.#bT4T4nbii<:ryy:zEryyEP}Mryy}M@qE qEr