sg'- d Z ddlmZmZmZ ddlZddlmZ ddl m
Z
ddl m Z m
Z
mZ ej e Z G d d e Zy)
z&
Feature extractor class for Wav2Vec2
)ListOptionalUnionN )SequenceFeatureExtractor) BatchFeature)PaddingStrategy
TensorTypeloggingc h e Zd ZdZddgZ d fd Ze ddeej deej de
deej fd Z dde ej ee
eej eee
f d e e
eef d
ee d e
d ee d
ee
dee eef dee defdZ xZS )Wav2Vec2FeatureExtractora
Constructs a Wav2Vec2 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.
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 `True`):
Whether or not to zero-mean unit-variance normalize the input. Normalizing can help to significantly
improve the performance for some models, *e.g.*,
[wav2vec2-lv60](https://huggingface.co/models?search=lv60).
return_attention_mask (`bool`, *optional*, defaults to `False`):
Whether or not [`~Wav2Vec2FeatureExtractor.__call__`] should return `attention_mask`.
Wav2Vec2 models that have set `config.feat_extract_norm == "group"`, such as
[wav2vec2-base](https://huggingface.co/facebook/wav2vec2-base-960h), have **not** been trained using
`attention_mask`. For such models, `input_values` should simply be padded with 0 and no `attention_mask`
should be passed.
For Wav2Vec2 models that have set `config.feat_extract_norm == "layer"`, such as
[wav2vec2-lv60](https://huggingface.co/facebook/wav2vec2-large-960h-lv60-self), `attention_mask` should be
passed for batched inference.
input_valuesattention_maskc H t | d|||d| || _ || _ y )N) feature_size
sampling_rate
padding_value )super__init__return_attention_mask do_normalize)selfr r r r r kwargs __class__s k/var/www/html/venv/lib/python3.12/site-packages/transformers/models/wav2vec2/feature_extraction_wav2vec2.pyr z!Wav2Vec2FeatureExtractor.__init__C s0 wl-_lwpvw%:" ( r returnc |t j |t j }g }t | |j d D ]m \ }}||d| j z
t j |d| j dz z }||j d k r|||d |j | o |S | D cg c]< }||j z
t j |j dz z > }}|S c c}w )z[
Every array in the list is normalized to have zero mean and unit variance
NgHz>r )
nparrayint32zipsummeansqrtvarshapeappend)r r r normed_input_valuesvectorlength normed_slicexs r zero_mean_unit_var_normz0Wav2Vec2FeatureExtractor.zero_mean_unit_var_normP s % XXnbhh ?N"$ "%lN4F4Fr4J"K
9 & )=)=)? ?2776RYSY?K^K^K`cgKgCh h L . .q 11,9L)#**<8
9 #" Vb"bPQALBGGAEEGdN4K#K"b "b"" #cs :AC?
raw_speechpadding
max_length
truncationpad_to_multiple_ofr return_tensorsr c ~ |;|| j k7 rAt d| d| j d| j d| d t j d t |t
j xr t |j dkD }
|
r&t |j d kD rt d
| |
xs@ t |t t f xr( t |d t
j t t f } | s|g}t d |i } | j | |||||
}
|
d }t |d t
j s8|D cg c]' }t j |t
j ) c}|
d < nt |t
j st |d t
j rc|d j t j t
j u r1|D cg c]! }|j# t
j # c}|
d < nkt |t
j rQ|j t j t
j u r"|j# t
j |
d < |
j% d }|6|D cg c]' }t j |t
j& ) c}|
d< | j( rK| j+ || t, j. ur|nd}| j1 |
d || j2 |
d < ||
j5 | }
|
S c c}w c c}w c c}w )a
Main method to featurize and prepare for the model one or several sequence(s).
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.
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)
Wav2Vec2 models that have set `config.feat_extract_norm == "group"`, such as
[wav2vec2-base](https://huggingface.co/facebook/wav2vec2-base-960h), have **not** been trained using
`attention_mask`. For such models, `input_values` should simply be padded with 0 and no
`attention_mask` should be passed.
For Wav2Vec2 models that have set `config.feat_extract_norm == "layer"`, such as
[wav2vec2-lv60](https://huggingface.co/facebook/wav2vec2-large-960h-lv60-self), `attention_mask` should
be passed for batched inference.
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.
padding_value (`float`, *optional*, defaults to 0.0):
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 .zIt is strongly recommended to pass the ``sampling_rate`` argument to this function. Failing to do so can result in silent errors that might be hard to debug. z2Only mono-channel audio is supported for input to r r )r2 r3 r4 r5 r )dtyper )r3 )r r )r
ValueErrorloggerwarning
isinstancer! ndarraylenr) listtupler padasarrayfloat32r; float64astypegetr# r _get_padding_strategiesr
DO_NOT_PADr0 r convert_to_tensors)r r1 r2 r3 r4 r5 r r6 r r is_batched_numpy
is_batchedencoded_inputs
padded_inputsr r" r s r __call__z!Wav2Vec2FeatureExtractor.__call__f s L $ 2 22 I$ P **+ , **+9]O1F
NN\
&j"**=[#jFVFVBWZ[B[ J$4$4 5 9 QRVQW XY Y%
zD%= 1 lz*Q-RTR\R\^ceiQj7k
$ J &~z&BC !! 1"7
!
%^4 ,q/2::6^j,kUZRZZRZZ-P,kM. )<4<?BJJ7 Q%%"**)==S_,`%U\\"**-E,`M. )
bjj
1l6H6HBHHUWU_U_L`6`,8,?,?
,KM. ) '**+;< %^l.mUZrzz%rxx/P.mM* + //J/OWfWqWqq
-1,H,H n-n\`\n\n -I -M. ) % ) < <^ LM ; -l -a /ns ,L0#&L5,L:)r9 i> FT)rR )FNFNNNN)__name__
__module__ __qualname____doc__model_input_namesr staticmethodr r! r@ floatr0 r boolstrr r intr
r rQ
__classcell__)r s @r r
r
sO B ()9: #
) ad#2::rd sD ) ( I 4 9 9
H %Q 7 Q r