+
j1 ^ RI t^RIHt ^RIHt ^RIHt ^RIH t H
t
H t ^RI H
t
]
! 4 ' d ^ RIt] P ! ]4 tR R lt]
! R R 7 ! R
R ]4 4 tR .tR# )
N) hertz_to_mel)SequenceFeatureExtractor) BatchFeature)
TensorTypeis_torch_availablelogging)requiresc t V ^8 d QhR\ R\ R\ R\ R\ R\ P /# ) num_mel_binsnum_spectrogram_bins sample_ratelower_edge_hertzupper_edge_hertzreturn)intfloatnpndarray)formats "q/home/luca/screen-checking/.venv/lib/python3.14/site-packages/transformers/models/lasr/feature_extraction_lasr.py __annotate__r sP "R "R"R "R "R "R
"R ZZ"R c \ P p^pVR, p\ P ! RWVR7 VR p \ V RR7 R\ P 3, p
\ P ! \ VRR7 \ VRR7 V ^, VR7 p V RR \ P R3, V ^R
\ P R3, V R, \ P R3, rp W,
W,
, pW,
W,
, p\ P
! R\ P ! VV4 4 p\ P ! VV^ .^ ^ ..4 P V4 # ) z.NumPy-port of the JAX mel weight matrix logic.g @ )dtypeNkaldi) mel_scaleNNN:r NN) r float64linspacer newaxismaximumminimumpadastype)r r
r r r r internal_dtype
bands_to_zero
nyquist_hertzlinear_frequenciesspectrogram_bins_meledgeslower_edge_mel
center_melupper_edge_mel lower_slopes upper_slopesmel_weights_matrixs &&&&&& r linear_to_mel_weight_matrixr5 s: ZZN M #%MS-Ucderest'(:gNqRTR\R\}] KK%9%9q
E cr
2::q=!
a BJJM"
b "**a- !/N )9j>YZL"9n>YZLCL,)OP
66$
q'9Aq6&B C J J5 QQr )backendsc b a a ] tR t^Et oRtRR.tR V 3R lltR
R ltR V3R lR lltRt Vt
V ;t # ) LasrFeatureExtractora
Constructs a LASR feature extractor.
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.
This class extracts mel-filter bank features from raw speech using a custom numpy implementation of the `Short Time
Fourier Transform` which should match pytorch's `torch.stft` equivalent.
Args:
feature_size (`int`, *optional*, defaults to 128):
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).
hop_length (`int`, *optional*, defaults to 160):
Length of the overlapping windows for the STFT used to obtain the Mel Frequency coefficients.
n_fft (`int`, *optional*, defaults to 512):
Size of the Fourier transform.
win_length (`int`, *optional*, defaults to 400):
The window length for the STFT computation.
padding_value (`float`, *optional*, defaults to 0.0):
Padding value used to pad the audio. Should correspond to silences.
input_featuresattention_maskc < \ SV ` ! RR VRVRV/VB W0n W@n WPn \
P ! \ VV^, ^, VRR\ P R7 4 V n
R# ) feature_size
sampling_rate
padding_valueg @_@g L@)r r
r r r r N ) super__init__
hop_lengthn_fft
win_lengthtorch
from_numpyr5 r r" mel_filters) selfr< r= rB rC rD r> kwargs __class__s &&&&&&&,r rA LasrFeatureExtractor.__init__a sn wlw-w_lwpvw$
$ ++ ' )%*aZ!^ )!&!'jj
r c 0 \ P ! V P R V\ P R7 pVP \ P 4 pVP RV P V P 4 p\ P P W4, V P R7 p\ P ! V4 ^, pV P P V4 p\ P ! Wg, RR7 p\ P ! V4 pV# )F)periodicdevicer )ngh㈵>)minr! )rE hann_windowrD r" tounfoldrB fftrfftrC absrG clamplog) rH waveformrN windowframesstft
power_specrG mel_specs &&& r _torch_extract_fbank_features2LasrFeatureExtractor._torch_extract_fbank_features{ s ""4??U6Y^YfYfg;;u}}-
T__dooFyy~~fo~<YYt_)
&&))&1 ;;z 7TB99X&r c < V ^8 d QhRS[ P S[S[, , S[S[ P , , S[S[S[, , , RS[RS[R, RS[S[, R, RS[R, RS[R, RS[R, R S[R, R
S[R, R S[R, R S[R, R
S[/ # )r
raw_speech
truncationpad_to_multiple_ofNreturn_tensorsreturn_attention_maskpadding
max_lengthr= do_normalizerN return_token_timestampsr ) r r listr boolr strr r )r
__classdict__s "r r !LasrFeatureExtractor.__annotate__ s EC ECJJe,tBJJ/??$tE{BSSEC EC $J EC
j(4/ EC $d{
EC tEC $JEC TzEC TkEC d
EC "&EC
ECr c Z VeR WP 8w dA \ RV P P RV P RV P RV R2 4 hM-\ P RV P P R24 \
V\ P 4 ' d \ P ! V4 pM\
V\ \ 34 ' dj \
V^ , \ \ P 34 ' d' V U
u. uF p
\ P ! V
4 NK pp
M\ P ! V4 p\
V\ P 4 ;' d \ VP 4 ^8 pV' dY \ VP 4 ^8 d? \ P R V P P R
24 VP! R4 p\
V\ \ 34 pV' dd V F] p
\ V
P 4 ^8 g K \ P R V P P R
24 V
P! R4 p
K_ V' g V' d6 V U
u. uF( qR, P# \ P$ 4 NK* pp
M'VR, P# \ P$ 4 .p\' R V/4 pV P) VVVVVVR
R7 pVP* P- R4 pV P/ VV
4 pR VP# \ P$ 4 /pV' dV VP0 R V P2 ^,
RV P4 13, pVP# \ P6 4 VR& \' VVR7 # u up
i u up
i )a
Main method to featurize and prepare for the model one or several sequence(s). Implementation uses PyTorch for
the STFT computation if available, otherwise a slower NumPy based one.
Args:
raw_speech (`np.ndarray`, `list[float]`, `list[np.ndarray]`, `list[list[float]]`):
The sequence or batch of sequences to be padded. 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. Must be mono channel audio, not
stereo, i.e. single float per timestep.
truncation (`bool`, *optional*, default to `True`):
Activates truncation to cut input sequences longer than *max_length* to *max_length*.
pad_to_multiple_of (`int`, *optional*, defaults to None):
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)
For Parakeet models, `attention_mask` should always be passed for batched inference, to avoid subtle
bugs.
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 `raw_speech` input was sampled. It is strongly recommended to pass
`sampling_rate` at the forward call to prevent silent errors and allow automatic speech recognition
pipeline.
padding_value (`float`, *optional*, defaults to 0.0):
The value that is used to fill the padding values / vectors.
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 of the model.
device (`str`, *optional*, defaults to `'cpu'`):
Specifies the device for computation of the log-mel spectrogram of audio signals in the
`_torch_extract_fbank_features` method. (e.g., "cpu", "cuda")
return_token_timestamps (`bool`, *optional*, defaults to `None`):
Deprecated. Use `return_attention_mask` instead from which the number of frames can be inferred.
Whether or not to return the number of frames of the input raw_speech.
These num_frames can be used by the model to compute word level timestamps.
Nz3The model corresponding to this feature extractor: z& was trained using a sampling rate of zI. Please make sure that the provided `raw_speech` 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.z2Only mono-channel audio is supported for input to z;. We will take the mean of the channels to convert to mono.r r9 pt)rg rh rc rd rf re r: )data tensor_typer! )r N) r=
ValueErrorrJ __name__loggerwarning
isinstancer r rE tensorrk tupleTensorlenshapemeanrR float32r r' r9 squeezer_ r: rD rB rl )rH rb rc rd re rf rg rh r= ri rN rj rI speechis_batched_torchis_batched_sequencebatched_speech
padded_inputsr9 rs r: s &&&&&&&&&&&&, r __call__ LasrFeatureExtractor.__call__ s' H $ 2 22 I$..JaJaIb c))-););(< =))-););(
! * 9 9!T__q=P=cTXTcTc=c:c dN%3%6%6uzz%BD! ">BB_ M. Vs N#..N()rB rG rC rD ) i> i i r )cpu)
FNNNlongestNNNr N) rv
__module__ __qualname____firstlineno____doc__model_input_namesrA r_ r __static_attributes____classdictcell__
__classcell__)rJ rn s @@r r8 r8 E s7 0 *+;<
4&EC EC ECr r8 )rE )numpyr audio_utilsr !feature_extraction_sequence_utilsr feature_extraction_utilsr utilsr r r utils.import_utilsr rE
get_loggerrv rw r5 r8 __all__r? r r r su ' I 4 < < *
H %"RJ
: MC3 MC MC` "
"r