o i/@sdZddlmZmZddlZGdddeZdededeeeffd d Z ded edeeeeffd d Z  ddedededej dej f ddZ dej dej fddZ  ddej dej dededeej ej ej ej ff ddZdS)z(Utility functions for Transducer models.)ListTupleNcs2eZdZdZdedededdffdd ZZS) TooShortUttErrorzRaised when the utt is too short for subsampling. Args: message: Error message to display. actual_size: The size that cannot pass the subsampling. limit: The size limit for subsampling. message actual_sizelimitreturnNcst|||_||_dS)z$Construct a TooShortUttError module.N)super__init__rr)selfrrr __class__P/home/ubuntu/.local/lib/python3.10/site-packages/espnet2/asr_transducer/utils.pyr s  zTooShortUttError.__init__)__name__ __module__ __qualname____doc__strintr __classcell__rrr rrs& r sub_factorsizercCs@|dkr |dkr dS|dkr|dkrdS|dkr|dkrdSd S) a Check if the input is too short for subsampling. Args: sub_factor: Subsampling factor for Conv2DSubsampling. size: Input size. Returns: : Whether an error should be sent. : Size limit for specified subsampling factor. )Tr )Tr)Fr)rrrrrcheck_short_utts r input_sizecCsj|dkrdd|dddfS|dkr dd|ddddfS|dkr1dd|ddddfStd)a-Get conv2D second layer parameters for given subsampling factor. Args: sub_factor: Subsampling factor (1/X). input_size: Input size. Returns: : Kernel size for second convolution. : Stride for second convolution. : Conv2DSubsampling output size. rrrrz?subsampling_factor parameter should be set to either 2, 4 or 6.) ValueError)rr!rrrsub_factor_to_params0s r% chunk_sizeleft_chunk_sizedevicecCsrtj|||tjd}t|D]'}|dkrd}n t||||d}t||d||}d||||f<q|S)aCreate chunk mask for the subsequent steps (size, size). Reference: https://github.com/k2-fsa/icefall/blob/master/icefall/utils.py Args: size: Size of the source mask. chunk_size: Number of frames in chunk. left_chunk_size: Size of the left context in chunks (0 means full context). device: Device for the mask tensor. Returns: mask: Chunk mask. (size, size) )r(dtyperr"T)torchzerosboolrangemaxmin)rr&r'r(maskistartendrrrmake_chunk_maskIs r4lengthscCs8|}|d}t||||}||dkS)zCreate source mask for given lengths. Reference: https://github.com/k2-fsa/icefall/blob/master/icefall/utils.py Args: lengths: Sequence lengths. (B,) Returns: : Mask for the sequence lengths. (B, max_len) rr")r.rr*arangeexpandto unsqueeze)r5max_len batch_sizeexpanded_lengthsrrrmake_source_maskks r=rlabelsencoder_out_lens ignore_idblank_idc sd dttjdtfdd}|j}fdd|D}|d|g|fdd|D||}|||tj|}t t t|}t ||} t d d|D|} ||| | fS) aGet Transducer loss I/O. Args: labels: Label ID sequences. (B, L) encoder_out_lens: Encoder output lengths. (B,) ignore_id: Padding symbol ID. blank_id: Blank symbol ID. Returns: decoder_in: Decoder inputs. (B, U) target: Target label ID sequences. (B, U) t_len: Time lengths. (B,) u_len: Label lengths. (B,) rr> padding_valuecSstt|}|dj|tdd|Dg|dddR|}t|D]}||||d||df<q&|S)zCreate padded batch of labels from a list of labels sequences. Args: labels: Labels sequences. [B x (?)] padding_value: Padding value. Returns: labels: Batch of padded labels sequences. (B,) rcss|]}|dVqdSrNr).0xrrr sz;get_transducer_task_io..pad_list..r"N)lennewr.rfill_r-)r>rBr;paddedr1rrrpad_lists   z(get_transducer_task_io..pad_listcsg|]}||kqSrrrEy)r@rr sz*get_transducer_task_io..csg|] }tj|gddqS)r)dim)r*cat)rElabel)blankrrrOscSsg|]}|dqSrrDrMrrrrOsNrT) rr*Tensorrr(rIr8typeint32listmap IntTensor) r>r?r@rArLr( labels_unpad decoder_intargett_lenu_lenr)rSr@rget_transducer_task_ios r`rC)rr)rtypingrrr* Exceptionrrr,r r%r(rUr4r=r`rrrrsB  "