ó
    qyüi<   ã                   ót   • S r SSKJr  SSKJr  SSKJr  / SQr/ SQr\" SS	9\ " S
 S\5      5       5       r	S/r
g)zWhisper model configurationé    )Ústricté   )ÚPreTrainedConfig)Úauto_docstring)Xé   é   é   é   é	   é
   é   é   é   é   é   é   é   é:   é;   é<   é=   é>   é?   éZ   é[   é\   é]   ie  in  i¶  i  i­  iÁ  i  i¢  i"  iÄ  ió  iÿ  i  i?  ia  io  ic  iö  iS  ir  i9	  i¸	  i‹  i¨  is  iø  iÈ  i³  iß  i¹  i#  i‰%  iç&  iC)  i"*  iœ,  i½-  iƒ.  ik3  i�5  i„5  iØ9  iÊ;  i@  i—A  iHF  iK  i6L  iôP  i!W  i—Y  ii  i�u  iÔv  i�‚  i…ˆ  i[�  i-ž  iež  i²¸  iú¿  éQÄ  i·Ä  é¸Ä  é¹Ä  )Vr   r   r	   r
   r   r   r   r   r   r   r   r   r   r   r   r   r   r   r   r   r   r   r   ig  i÷  i
  i  ii  i}  i†  i–  iš  i£  iF  i=  i¾  iœ	  iC
  i®  iµ  iÄ  iÐ  i  iy  iW  i;  i¹  i÷  ii  ie#  i§$  i¼(  i±*  i¢.  i/  i+0  i1  iá5  iM7  i+9  i¡;  i=  i©@  iÜ@  iºG  iJ  ikN  i«T  iøW  if  i1f  iCg  iwn  is  i¢{  i.~  iÖ~  i�  io§  iA¹  iÎÂ  iNÄ  iRÄ  r   r    iºÄ  zopenai/whisper-tiny)Ú
checkpointc                   ó°  • \ rS rSr% SrSrS/rSSSSS.rS	r\	\
S
'   Sr\	\
S'   Sr\	\
S'   Sr\	\
S'   Sr\	\
S'   Sr\	\
S'   Sr\	\
S'   Sr\	\
S'   Sr\\	-  \
S'   Sr\\	-  \
S'   Sr\	\
S'   Sr\\
S'   Sr\\
S'   Sr\\
S'   Sr\	\
S'   Sr\\	-  \
S'   Sr\\	-  \
S '   Sr\\	-  \
S!'   S"r\\
S#'   S$r \\
S%'   S&r!\	\
S''   S(r"\	\
S)'   S*r#\	S+-  \
S,'   S*r$\	S+-  \
S-'   S*r%\	\&\	   -  S+-  \
S.'   S+r'\&S+-  \
S/'   S0r(\&\	   \)\	S14   -  S+-  \
S2'   S$r*\\
S3'   S4r+\	\
S5'   S$r,\\
S6'   S7r-\\	-  \
S8'   S9r.\	\
S:'   S;r/\	\
S<'   Sr0\\	-  \
S='   S9r1\	\
S>'   S?r2\	\
S@'   SAr3\	\
SB'   Sr4\\
SC'   SDr5g+)EÚWhisperConfigé0   a¦  
max_source_positions (`int`, *optional*, defaults to 1500):
    The maximum sequence length of log-mel filter-bank features that this model might ever be used with.
max_target_positions (`int`, *optional*, defaults to 448):
    The maximum sequence length that this model might ever be used with. Typically set this to something large
    just in case (e.g., 512 or 1024 or 2048).
suppress_tokens (`list[int]`, *optional*):
    A list containing the non-speech tokens that will be used by the logit processor in the `generate`
    function. NON_SPEECH_TOKENS and NON_SPEECH_TOKENS_MULTI each correspond to the `english-only` and the
    `multilingual` model.
begin_suppress_tokens (`list[int]`, *optional*, defaults to `[220,50256]`):
    A list containing tokens that will be suppressed at the beginning of the sampling process. Initialized as
    the token for `" "` (`blank_token_id`) and the `eos_token_id`
use_weighted_layer_sum (`bool`, *optional*, defaults to `False`):
    Whether to use a weighted average of layer outputs with learned weights. Only relevant when using an
    instance of [`WhisperForAudioClassification`].
classifier_proj_size (`int`, *optional*, defaults to 256):
    Dimensionality of the projection before token mean-pooling for classification. Only relevant when using an
    instance of [`WhisperForAudioClassification`].
apply_spec_augment (`bool`, *optional*, defaults to `False`):
    Whether to apply *SpecAugment* data augmentation to the outputs of the feature encoder. For reference see
    [SpecAugment: A Simple Data Augmentation Method for Automatic Speech
    Recognition](https://huggingface.co/papers/1904.08779).
mask_time_prob (`float`, *optional*, defaults to 0.05):
    Percentage (between 0 and 1) of all feature vectors along the time axis which will be masked. The masking
    procedure generates `mask_time_prob*len(time_axis)/mask_time_length` independent masks over the axis. If
    reasoning from the probability of each feature vector to be chosen as the start of the vector span to be
    masked, *mask_time_prob* should be `prob_vector_start*mask_time_length`. Note that overlap may decrease the
    actual percentage of masked vectors. This is only relevant if `apply_spec_augment == True`.
mask_time_length (`int`, *optional*, defaults to 10):
    Length of vector span along the time axis.
mask_time_min_masks (`int`, *optional*, defaults to 2),:
    The minimum number of masks of length `mask_feature_length` generated along the time axis, each time step,
    irrespectively of `mask_feature_prob`. Only relevant if ''mask_time_prob*len(time_axis)/mask_time_length <
    mask_time_min_masks''
mask_feature_prob (`float`, *optional*, defaults to 0.0):
    Percentage (between 0 and 1) of all feature vectors along the feature axis which will be masked. The
    masking procedure generates `mask_feature_prob*len(feature_axis)/mask_time_length` independent masks over
    the axis. If reasoning from the probability of each feature vector to be chosen as the start of the vector
    span to be masked, *mask_feature_prob* should be `prob_vector_start*mask_feature_length`. Note that overlap
    may decrease the actual percentage of masked vectors. This is only relevant if `apply_spec_augment is
    True`.
mask_feature_length (`int`, *optional*, defaults to 10):
    Length of vector span along the feature axis.
mask_feature_min_masks (`int`, *optional*, defaults to 0):
    The minimum number of masks of length `mask_feature_length` generated along the feature axis, each time
    step, irrespectively of `mask_feature_prob`. Only relevant if
    `mask_feature_prob*len(feature_axis)/mask_feature_length < mask_feature_min_masks`.
median_filter_width (`int`, *optional*, defaults to 7):
    Width of the median filter used to smoothen to cross-attention outputs when computing token timestamps.
    Should be an odd number.

Example:

```python
>>> from transformers import WhisperConfig, WhisperModel

>>> # Initializing a Whisper tiny style configuration
>>> configuration = WhisperConfig()

>>> # Initializing a model (with random weights) from the tiny style configuration
>>> model = WhisperModel(configuration)

>>> # Accessing the model configuration
>>> configuration = model.config
```ÚwhisperÚpast_key_valuesÚencoder_attention_headsÚd_modelÚencoder_layers)Únum_key_value_headsÚnum_attention_headsÚhidden_sizeÚnum_hidden_layersi™Ê  Ú
vocab_sizeéP   Únum_mel_binsé   é   Údecoder_layersÚdecoder_attention_headsi   Údecoder_ffn_dimÚencoder_ffn_dimg        Úencoder_layerdropÚdecoder_layerdropr   Údecoder_start_token_idTÚ	use_cacheÚis_encoder_decoderÚgeluÚactivation_functioni€  ÚdropoutÚattention_dropoutÚactivation_dropoutg{®Gáz”?Úinit_stdFÚscale_embeddingiÜ  Úmax_source_positionsiÀ  Úmax_target_positionséPÄ  NÚpad_token_idÚbos_token_idÚeos_token_idÚsuppress_tokens)éÜ   rE   .Úbegin_suppress_tokensÚuse_weighted_layer_sumé   Úclassifier_proj_sizeÚapply_spec_augmentgš™™™™™©?Úmask_time_probr   Úmask_time_lengthr   Úmask_time_min_masksÚmask_feature_probÚmask_feature_lengthr   Úmask_feature_min_masksr	   Úmedian_filter_widthÚtie_word_embeddings© )6Ú__name__Ú
__module__Ú__qualname__Ú__firstlineno__Ú__doc__Ú
model_typeÚkeys_to_ignore_at_inferenceÚattribute_mapr.   ÚintÚ__annotations__r0   r)   r'   r3   r4   r5   r6   r7   Úfloatr8   r9   r:   Úboolr;   r=   Ústrr(   r>   r?   r@   rA   rB   rC   rD   rF   rG   rH   ÚlistrI   rK   ÚtuplerL   rN   rO   rP   rQ   rR   rS   rT   rU   rV   rW   Ú__static_attributes__rX   ó    Ún/home/mande/repo/quber/.venv/lib/python3.13/site-packages/transformers/models/whisper/configuration_whisper.pyr#   r#   0   s  ‡ ñAðF €JØ#4Ð"5Ðà8Ø8Ø Ø-ñ	€Mð €J�ÓØ€L�#ÓØ€N�CÓØ#$Ð˜SÓ$Ø€N�CÓØ#$Ð˜SÓ$Ø€O�SÓØ€O�SÓØ%(Ð�u˜s‘{Ó(Ø%(Ð�u˜s‘{Ó(Ø"'Ð˜CÓ'Ø€IˆtÓØ#Ð˜Ó#Ø%Ð˜Ó%Ø€GˆSÓØ€GˆU�S‰[ÓØ%(Ð�u˜s‘{Ó(Ø&)Ð˜ ™Ó)Ø€HˆeÓØ!€O�TÓ!Ø $Ð˜#Ó$Ø #Ð˜#Ó#Ø$€L�#˜‘*Ó$Ø$€L�#˜‘*Ó$Ø+0€L�#˜˜S™	‘/ DÑ(Ó0Ø#'€O�T˜D‘[Ó'Ø@LÐ˜4 ™9 u¨S°#¨X¡Ñ6¸Ñ=ÓLØ#(Ð˜DÓ(Ø #Ð˜#Ó#Ø$Ð˜Ó$Ø"&€N�E˜C‘KÓ&ØÐ�cÓØ Ð˜Ó Ø%(Ð�u˜s‘{Ó(Ø!Ð˜Ó!Ø"#Ð˜CÓ#Ø Ð˜Ó Ø $Ð˜Ö$ri   r#   N)r]   Úhuggingface_hub.dataclassesr   Úconfiguration_utilsr   Úutilsr   ÚNON_SPEECH_TOKENSÚNON_SPEECH_TOKENS_MULTIr#   Ú__all__rX   ri   rj   Ú<module>rq      s[   ðñ "å .å 3Ý #ò
Ð ò
Ð ñ Ð0Ñ1Øôr%Ð$ó r%ó ó 2ðr%ðj Ð
�ri   