o i@sXdZddlmZmZmZddlZddlmZddlm Z ddl m Z Gddde Z dS) z-RNN decoder definition for Transducer models.)ListOptionalTupleN)check_argument_types) Hypothesis) AbsDecodercseZdZdZ       d/deded ed ed ed ed ededdffdd Z d0dej de e ej e ej fdej fddZ dej de ej e ej fde ej e ej e ej fffddZ dej deede ej e ej fde ej e ej e ej fffddZd eede ej e ej e ej fffd!d"Zd#ejddfd$d%Zd&ede ej e ejffd'd(Zde ej e ej fd)ede ej e ej ffd*d+Zd,ee ej e ej fde ej e ej ffd-d.ZZS)1 RNNDecoderaRNN decoder module. Args: vocab_size: Vocabulary size. embed_size: Embedding size. hidden_size: Hidden size.. rnn_type: Decoder layers type. num_layers: Number of decoder layers. dropout_rate: Dropout rate for decoder layers. embed_dropout_rate: Dropout rate for embedding layer. embed_pad: Embedding padding symbol ID. lstmr vocab_size embed_size hidden_sizernn_type num_layers dropout_rateembed_dropout_rate embed_padreturnNc stts J|dvrtd|tjj|||d|_tjj|d|_ |dkr/tjj ntjj } tj | ||dddg|_ td|D]} |j | ||dddg7_ qFtj fd d t|D|_||_||_||_||_t|j|_i|_d S) zConstruct a RNNDecoder object.)r gruzNot supported: rnn_type=) padding_idxpr r T) batch_firstcsg|] }tjjdqS)r)torchnnDropout).0_r^/home/ubuntu/.local/lib/python3.10/site-packages/espnet2/asr_transducer/decoder/rnn_decoder.py ;sz'RNNDecoder.__init__..N)super__init__r ValueErrorrr Embeddingembedr dropout_embedLSTMGRU ModuleListrnnrange dropout_rnndlayersdtype output_sizer next parametersdevice score_cache) selfr rrrrrrr rnn_classr __class__r r"r%s*  zRNNDecoder.__init__labelsstatescCs<|dur ||d}|||}|||\}}|S)a%Encode source label sequences. Args: labels: Label ID sequences. (B, L) states: Decoder hidden states. ((N, B, D_dec), (N, B, D_dec) or None) or None Returns: dec_out: Decoder output sequences. (B, U, D_dec) Nr) init_statesizer)r( rnn_forward)r7r;r< dec_embeddec_outr!r!r"forwardGs zRNNDecoder.forwardxstatecCs|\}}||d\}}t|jD]M}|jdkrB|j|||||d|||dfd\}\|||d<|||d<n|j|||||dd\}|||d<|j||}q|||ffS)atEncode source label sequences. Args: x: RNN input sequences. (B, D_emb) state: Decoder hidden states. ((N, B, D_dec), (N, B, D_dec) or None) Returns: x: RNN output sequences. (B, D_dec) (h_next, c_next): Decoder hidden states. (N, B, D_dec), (N, B, D_dec) or None) rr r )hx)r=r>r.r0r1r-r/)r7rCrDh_prevc_prevh_nextc_nextlayerr!r!r"r?_s  ( zRNNDecoder.rnn_forwardlabellabel_sequence dec_statecCs^dtt|}||jvr|j|\}}n||}|||\}}||f|j|<|d|fS)aOne-step forward hypothesis. Args: label: Previous label. (1, 1) label_sequence: Current label sequence. dec_state: Previous decoder hidden states. ((N, 1, D_dec), (N, 1, D_dec) or None) Returns: dec_out: Decoder output sequence. (1, D_dec) dec_state: Decoder hidden states. ((N, 1, D_dec), (N, 1, D_dec) or None) rr)joinmapstrr6r(r?)r7rKrLrM str_labelsrAr@r!r!r"scores   zRNNDecoder.scorehypscCsVtjdd|D|jd}||}|dd|D}|||\}}|d|fS)zOne-step forward hypotheses. Args: hyps: Hypotheses. Returns: dec_out: Decoder output sequences. (B, D_dec) states: Decoder hidden states. ((N, B, D_dec), (N, B, D_dec) or None) cSsg|]}|jdgqS))yseqrhr!r!r"r#sz*RNNDecoder.batch_score..r5cSsg|]}|jqSr!)rMrVr!r!r"r#sr )r LongTensorr5r(create_batch_statesr?squeeze)r7rSr;r@r<rAr!r!r" batch_scores  zRNNDecoder.batch_scorer5cCs ||_dS)zNSet GPU device to use. Args: device: Device ID. NrX)r7r5r!r!r" set_devices zRNNDecoder.set_device batch_sizecCsJtj|j||j|jd}|jdkr!tj|j||j|jd}||fS|dfS)zInitialize decoder states. Args: batch_size: Batch size. Returns: : Initial decoder hidden states. ((N, B, D_dec), (N, B, D_dec) or None) rXr N)rzerosr0r2r5r1)r7r^h_nc_nr!r!r"r=s  zRNNDecoder.init_stateidxcCsT|ddd||dddf|jdkr'|ddd||dddffSdfS)a-Get specified ID state from decoder hidden states. Args: states: Decoder hidden states. ((N, B, D_dec), (N, B, D_dec) or None) idx: State ID to extract. Returns: : Decoder hidden state for given ID. ((N, 1, D_dec), (N, 1, D_dec) or None) rNr r )r1)r7r<rbr!r!r" select_states *zRNNDecoder.select_state new_statescCs@tjdd|Ddd|jdkrtjdd|DddfSdfS)zCreate decoder hidden states. Args: new_states: Decoder hidden states. [N x ((1, D_dec), (1, D_dec) or None)] Returns: states: Decoder hidden states. ((N, B, D_dec), (N, B, D_dec) or None) cSg|]}|dqS)rr!rsr!r!r"r#z2RNNDecoder.create_batch_states..r )dimr cSre)r r!rfr!r!r"r#rhN)rcatr1)r7rdr!r!r"rZs  zRNNDecoder.create_batch_states)r r r r r r r)N)__name__ __module__ __qualname____doc__intrPfloatr%rTensorrrrBr?rrRrr\r5r]tensorr=rcrZ __classcell__r!r!r9r"r s  /  "      r) rntypingrrrr typeguardr-espnet2.asr_transducer.beam_search_transducerr*espnet2.asr_transducer.decoder.abs_decoderrrr!r!r!r"s