cCiESrSSKrSSKJrJrJr SSKrSSKJ r J r J r J r SSK Jr SSKJr SSKJrJrJr \R*"\5r"S S \5rS /rg) z%Feature extractor class for SpeechT5.N)AnyOptionalUnion)mel_filter_bankoptimal_fft_length spectrogramwindow_function)SequenceFeatureExtractor) BatchFeature)PaddingStrategy TensorTypeloggingc@^\rSrSrSrSS/rS%S\S\S\S\S \S \S \S \ S \S\S\S\S\S\4U4Sjjjr \ S&S\ \ RS\ \ RS\S\ \ R4Sjj5rS\ RS\ R4SjrS'S\\\ R\ \\ \ R\ \ \4S\\\ R\ \\ \ R\ \ \4S\\\ \4S\\S\S\\S\\S\\\ \4S\\S\4SjjrS(S \\ R\ \\ \ R\ \ \4S!\S\\\ \4S\\S\S\\S\\S\\\ \4S\4S"jjrS\\ \44U4S#jjrS$rU=r$))SpeechT5FeatureExtractorai Constructs a SpeechT5 feature extractor. This class can pre-process a raw speech signal by (optionally) normalizing to zero-mean unit-variance, for use by the SpeechT5 speech encoder prenet. This class can also extract log-mel filter bank features from raw speech, for use by the SpeechT5 speech decoder prenet. This feature extractor inherits from [`~feature_extraction_sequence_utils.SequenceFeatureExtractor`] which contains most of the main methods. Users should refer to this superclass for more information regarding those methods. Args: feature_size (`int`, *optional*, defaults to 1): The feature dimension of the extracted features. sampling_rate (`int`, *optional*, defaults to 16000): The sampling rate at which the audio files should be digitalized expressed in hertz (Hz). padding_value (`float`, *optional*, defaults to 0.0): The value that is used to fill the padding values. do_normalize (`bool`, *optional*, defaults to `False`): Whether or not to zero-mean unit-variance normalize the input. Normalizing can help to significantly improve the performance for some models. num_mel_bins (`int`, *optional*, defaults to 80): The number of mel-frequency bins in the extracted spectrogram features. hop_length (`int`, *optional*, defaults to 16): Number of ms between windows. Otherwise referred to as "shift" in many papers. win_length (`int`, *optional*, defaults to 64): Number of ms per window. win_function (`str`, *optional*, defaults to `"hann_window"`): Name for the window function used for windowing, must be accessible via `torch.{win_function}` frame_signal_scale (`float`, *optional*, defaults to 1.0): Constant multiplied in creating the frames before applying DFT. This argument is deprecated. fmin (`float`, *optional*, defaults to 80): Minimum mel frequency in Hz. fmax (`float`, *optional*, defaults to 7600): Maximum mel frequency in Hz. mel_floor (`float`, *optional*, defaults to 1e-10): Minimum value of mel frequency banks. reduction_factor (`int`, *optional*, defaults to 2): Spectrogram length reduction factor. This argument is deprecated. return_attention_mask (`bool`, *optional*, defaults to `True`): Whether or not [`~SpeechT5FeatureExtractor.__call__`] should return `attention_mask`. input_valuesattention_mask feature_size sampling_rate padding_value do_normalize num_mel_bins hop_length win_length win_functionframe_signal_scalefminfmax mel_floorreduction_factorreturn_attention_maskc >[TU]"S XUS.UD6 X@lXlXPlX`lXplXlXlXl Xl Xl Xl Xr-S-Ul Xb-S-Ul[UR5UlUR S-S-Ul[%URURSS9Ul[)UR"URURURUR*SSS9UlU S :wa[.R0"S [25 U S :wa[.R0"S [25 gg)N)rrriT) window_lengthnameperiodicslaney)num_frequency_binsnum_mel_filters min_frequency max_frequencyrnorm mel_scale?zeThe argument `frame_signal_scale` is deprecated and will be removed in version 4.30.0 of Transformersg@zcThe argument `reduction_factor` is deprecated and will be removed in version 4.30.0 of Transformers)super__init__rr"rrrrrrrr r! sample_size sample_stridern_fftn_freqsr windowrr mel_filterswarningswarn FutureWarning)selfrrrrrrrrrrrr r!r"kwargs __class__s r/home/james-whalen/.local/lib/python3.13/site-packages/transformers/models/speecht5/feature_extraction_speecht5.pyr3!SpeechT5FeatureExtractor.__init__Ns5$ wl_lwpvw(%:"($$("4  " 0%5='74?'(8(89  a1, %D4D4D4K\K\gkl *#|| --)))),,   $ MMw  s " MMu  #returncUb[R"U[R5n/n[XR S55HlupEXDSUR 5- [R "USUR5S-5- nXVRS:aX&US&URU5 Mn U$UVs/sH=owUR 5- [R "UR5S-5- PM? nnU$s snf)zK Every array in the list is normalized to have zero mean and unit variance NgHz>r) nparrayint32zipsummeansqrtvarshapeappend)rrrnormed_input_valuesvectorlength normed_slicexs r@zero_mean_unit_var_norm0SpeechT5FeatureExtractor.zero_mean_unit_var_norms  %XXnbhh?N"$ "%l4F4Fr4J"K &)=)=)? ?2776RYSY?K^K^K`cgKgChh ..q11,9)#**<8 #L#"Vb"bUaPQLBGGAEEGdN4K#KUa "b""#cs;AD one_waveformc [UURURURURUR UR SS9nUR$)zJ Extracts log-mel filterbank features for one waveform array (unbatched). log10)r8 frame_lengthr fft_lengthr9r log_mel)r r8r4r5r6r9r T)r=rW log_mel_specs r@_extract_mel_features.SpeechT5FeatureExtractor._extract_mel_featuressP# ;;))))zz((nn  ~~rBaudio audio_targetpadding max_length truncationpad_to_multiple_ofreturn_tensorsc UcUc [S5eU b<XR:wa,[SUSURSURSU S3 5eO-[RSURR S 35 UbUR "US UUUUUU40U D6n OSn Ub?UR "US UUUUUU40U D6n U cU $U S U S 'U RS5n U bXS'U $)a Main method to featurize and prepare for the model one or several sequence(s). Pass in a value for `audio` to extract waveform features. Pass in a value for `audio_target` to extract log-mel spectrogram features. Args: audio (`np.ndarray`, `list[float]`, `list[np.ndarray]`, `list[list[float]]`, *optional*): The sequence or batch of sequences to be processed. Each sequence can be a numpy array, a list of float values, a list of numpy arrays or a list of list of float values. This outputs waveform features. Must be mono channel audio, not stereo, i.e. single float per timestep. audio_target (`np.ndarray`, `list[float]`, `list[np.ndarray]`, `list[list[float]]`, *optional*): The sequence or batch of sequences to be processed as targets. Each sequence can be a numpy array, a list of float values, a list of numpy arrays or a list of list of float values. This outputs log-mel spectrogram features. padding (`bool`, `str` or [`~utils.PaddingStrategy`], *optional*, defaults to `False`): Select a strategy to pad the returned sequences (according to the model's padding side and padding index) among: - `True` or `'longest'`: Pad to the longest sequence in the batch (or no padding if only a single sequence if provided). - `'max_length'`: Pad to a maximum length specified with the argument `max_length` or to the maximum acceptable input length for the model if that argument is not provided. - `False` or `'do_not_pad'` (default): No padding (i.e., can output a batch with sequences of different lengths). max_length (`int`, *optional*): Maximum length of the returned list and optionally padding length (see above). truncation (`bool`): Activates truncation to cut input sequences longer than *max_length* to *max_length*. pad_to_multiple_of (`int`, *optional*): If set will pad the sequence to a multiple of the provided value. This is especially useful to enable the use of Tensor Cores on NVIDIA hardware with compute capability `>= 7.5` (Volta), or on TPUs which benefit from having sequence lengths be a multiple of 128. return_attention_mask (`bool`, *optional*): Whether to return the attention mask. If left to the default, will return the attention mask according to the specific feature_extractor's default. [What are attention masks?](../glossary#attention-mask) return_tensors (`str` or [`~utils.TensorType`], *optional*): If set, will return tensors instead of list of python integers. Acceptable values are: - `'tf'`: Return TensorFlow `tf.constant` objects. - `'pt'`: Return PyTorch `torch.Tensor` objects. - `'np'`: Return Numpy `np.ndarray` objects. sampling_rate (`int`, *optional*): The sampling rate at which the `audio` or `audio_target` input was sampled. It is strongly recommended to pass `sampling_rate` at the forward call to prevent silent errors. Nz9You must provide either `audio` or `audio_target` values.z3The model corresponding to this feature extractor: z& was trained using a sampling rate of zB. Please make sure that the provided audio input was sampled with z and not .zDIt is strongly recommended to pass the `sampling_rate` argument to `zN()`. Failing to do so can result in silent errors that might be hard to debug.FTrlabelsrdecoder_attention_mask) ValueErrorrloggerwarningr?__name___process_audioget)r=rarbrcrdrerfr"rgrr>inputs inputs_targetrks r@__call__!SpeechT5FeatureExtractor.__call__sb~ =\1XY Y  $ 2 22 I$P**+,**+9]O1F3 NNVW[WeWeWnWnVop\\   (("%  FF  # //"%  M~$$#0#@x )6):):;K)L&)57M34 rBspeech is_targetc [U[R5=(a [UR5S:n U (a'[UR5S:a[ SU35eU =(dE [U[ [45=(a( [US[R[[ 45n U (a5UVs/sH&n[R"U[RS9PM( snnOU (dC[U[R5(d$[R"U[RS9nOo[U[R5(aPUR[R"[R5LaUR[R5nU (dU/nURn U(a?UV s/sHoRU 5PM nn [SU05nUR Ul O [SU05nUR""U4UUUUUS.U D6nXl USn[US[R5(d9UVs/sH&n[R"U[RS9PM( snUS'GO[U[R5(d[US[R5(ahUSR[R"[R5La4UVs/sH"nUR[R5PM$ snUS'Or[U[R5(aSUR[R"[R5La"UR[R5US'UR%S5nUb7UVs/sH&n[R"U[R&S9PM( snUS'U(dVUR((aEUR+X4S 9[,R.LaUOSnUR1USUUR2S 9US'UbUR5U5nU$s snfs sn fs snfs snfs snf) Nr%r$z2Only mono-channel audio is supported for input to r)dtyper)rcrdrerfr"r)rd)rr) isinstancerFndarraylenrNrllisttupleasarrayfloat32ryfloat64astyperr_r rpadrqrHr_get_padding_strategiesr DO_NOT_PADrUrconvert_to_tensors)r=rvrwrcrdrerfr"rgr>is_batched_numpy is_batchedfeature_size_hackwaveformfeaturesencoded_inputs padded_inputsrrGrs r@rp'SpeechT5FeatureExtractor._process_audio)s&fbjj9Sc&,,>ORS>S FLL 1A 5QRVQWXY Y% ve} - d:fQi"**V[]aIb3c  IOPvbjjrzz:PFJvrzz$B$BZZbjj9F  + + @T0T]]2::.FXF!-- MSTV2288*DEN $ 1 1D )>6*BCN  !!1"7   .%^4 ,q/2::66^j,k^jUZRZZRZZ-P^j,kM. )<44<?BJJ77Q%%"**)==S_,`S_%U\\"**-ES_,`M. )  bjj 1 1l6H6HBHHUWU_U_L`6`,8,?,? ,KM. )'**+;<  %^l.m^lUZrzz%rxx/P^l.mM* +T..///OWfWqWqq  -1,H,Hn-n\`\n\n-I-M. )  %)<<^LMCQU*-l -a/ns-Q, Q1-Q6 )Q;-RcP>[TU]5n/SQnUH nX1;dM X M U$)N)r8r9r4r5r6r7)r2to_dict)r=outputnamesr'r?s r@r SpeechT5FeatureExtractor.to_dicts2"^D~L rB)rrrrrrr9r r6r7rr!r"r4r5rrr8)r%i>FP@ hann_windowr0rig|=r$T)r) NNFNFNNNN)FFNFNNN)ro __module__ __qualname____firstlineno____doc__model_input_namesintfloatboolstrr3 staticmethodr}rFr{rUr_rrr rr rtrpdictrr__static_attributes__ __classcell__)r?s@r@rrsN*X()9:""")$' !&*::: :  :  ::::"::::: $::xbe#2::&#8NPTUYZ_U`Pa%abcstS/12 s SM s  s%SMs (~s!sJ!78s }s sp 5:$( ,004;?Ubjj$u+tBJJ/?d5kARRSUUtS/12 U SM U  U%SMU (~U!sJ!78U Un c3h  rBr)rr:typingrrrnumpyrF audio_utilsrrr r !feature_extraction_sequence_utilsr feature_extraction_utilsr utilsr rr get_loggerrormr__all__r1rBr@rsT,''\\I499   H %j7jZ & &rB