+
    
åûjb  ã                   ó†  € ^ RI Ht ^ RIHtHtHt ^ RIt^ RIHt ^ RIH	t	 ^RI
Ht ^RIHt ^RIHtHtHt ^R	IHt ^R
IHtHt . ROt ! R R]P0                  4      t ! R R]P4                  4      t ! R R]P4                  4      t ! R R]4      t]! 4       ]! R]P<                  3R7      RRRR/R R ll4       4       tR# )é    )Úpartial)ÚAnyÚCallableÚOptionalN)Únn)ÚConv3dNormActivation)ÚVideoClassification)Ú_log_api_usage_once)Úregister_modelÚWeightsÚWeightsEnum)Ú_KINETICS400_CATEGORIES)Ú_ovewrite_named_paramÚhandle_legacy_interfaceÚS3DÚS3D_Weightsc                   ó>   a a€ ] tR t^t oV3R lV 3R lltRtVtV ;t# )ÚTemporalSeparableConvc                óf   <€ V ^8„  d   QhRS[ RS[ RS[ RS[ RS[ RS[RS[P                  3,          /# )é   Ú	in_planesÚ
out_planesÚkernel_sizeÚstrideÚpaddingÚ
norm_layer.©Úintr   r   ÚModule)ÚformatÚ__classdict__s   "€ÚXC:\Ai_Coding\apps\studio\ComfyUI\.venv\Lib\site-packages\torchvision\models\video\s3d.pyÚ__annotate__Ú"TemporalSeparableConv.__annotate__   sT   ø€ ÷ 
ñ 
áð
ñ ð
ñ ð	
ñ
 ð
ñ ð
ñ ˜S¡"§)¡)˜^Õ,ñ
ó    c                ó†   <€ \         SV `  \        VV^W33^WD3^ WU3RVR7      \        VVV^^3V^^3V^ ^ 3RVR7      4       R# )é   F)r   r   r   Úbiasr   N)ÚsuperÚ__init__r   )Úselfr   r   r   r   r   r   Ú	__class__s   &&&&&&&€r"   r*   ÚTemporalSeparableConv.__init__   sp   ø€ ô 	‰ÑÜ ØØØ Ð9Ø˜6Ð*Ø˜GÐ-ØØ%ôô !ØØØ(¨!¨QÐ/Ø  1�~Ø  ! Q˜ØØ%ôö	
r%   © )Ú__name__Ú
__module__Ú__qualname__Ú__firstlineno__r*   Ú__static_attributes__Ú__classdictcell__Ú__classcell__©r,   r!   s   @@r"   r   r      s   ù‡ € ÷
÷ 
ó 
r%   r   c                   óD   a a€ ] tR t^6t oV3R lV 3R lltR tRtVtV ;t# )ÚSepInceptionBlock3Dc                ór   <€ V ^8„  d   QhRS[ RS[ RS[ RS[ RS[ RS[ RS[ RS[R	S[P                  3,          /# )
r   r   Úb0_outÚb1_midÚb1_outÚb2_midÚb2_outÚb3_outr   .r   )r    r!   s   "€r"   r#   Ú SepInceptionBlock3D.__annotate__7   sh   ø€ ÷ 
ñ 
áð
ñ ð
ñ ð	
ñ
 ð
ñ ð
ñ ð
ñ ð
ñ ˜S¡"§)¡)˜^Õ,ñ
r%   c	                ó¨  <€ \         S	V `  4        \        W^^VR7      V n        \        P
                  ! \        W^^VR7      \        W4^^^VR7      4      V n        \        P
                  ! \        W^^VR7      \        WV^^^VR7      4      V n        \        P
                  ! \        P                  ! R^^R7      \        W^^VR7      4      V n
        R# )r'   ©r   r   r   )r   r   r   r   ©r   r   r   N©é   rE   rE   )r)   r*   r   Úbranch0r   Ú
Sequentialr   Úbranch1Úbranch2Ú	MaxPool3dÚbranch3)
r+   r   r:   r;   r<   r=   r>   r?   r   r,   s
   &&&&&&&&&€r"   r*   ÚSepInceptionBlock3D.__init__7   s¹   ø€ ô 	‰ÑÔä+¨IÈ1ÐUVÐcmÔnˆŒÜ—}’}Ü  ÀÈ!ÐXbÔcÜ! &¸aÈÐSTÐakÔló
ˆŒô —}’}Ü  ÀÈ!ÐXbÔcÜ! &¸aÈÐSTÐakÔló
ˆŒô —}’}Ü�LŠL Y°qÀ!ÔDÜ  ÀÈ!ÐXbÔcó
ˆŽr%   c                óÀ   € V P                  V4      pV P                  V4      pV P                  V4      pV P                  V4      p\        P
                  ! W#WE3^4      pV# )r'   )rF   rH   rI   rK   ÚtorchÚcat)r+   ÚxÚx0Úx1Úx2Úx3Úouts   &&     r"   ÚforwardÚSepInceptionBlock3D.forwardR   sP   € Ø�\‰\˜!‹_ˆØ�\‰\˜!‹_ˆØ�\‰\˜!‹_ˆØ�\‰\˜!‹_ˆÜ�iŠi˜ Ð(¨!Ó,ˆàˆ
r%   )rF   rH   rI   rK   )	r/   r0   r1   r2   r*   rV   r3   r4   r5   r6   s   @@r"   r8   r8   6   s   ù‡ € ÷
ó 
÷6ò r%   r8   c                   óL   a a€ ] tR t^\t oRtRV3R lV 3R llltR tRtVtV ;t	# )r   a;  S3D main class.

Args:
    num_class (int): number of classes for the classification task.
    dropout (float): dropout probability.
    norm_layer (Optional[Callable]): Module specifying the normalization layer to use.

Inputs:
    x (Tensor): batch of videos with dimensions (batch, channel, time, height, width)
c          	      ó|   <€ V ^8„  d   QhRS[ RS[RS[S[RS[P
                  P                  3,          ,          RR/# )r   Únum_classesÚdropoutr   .ÚreturnN)r   Úfloatr   r   rN   r   r   )r    r!   s   "€r"   r#   ÚS3D.__annotate__h   sJ   ø€ ÷ (
ñ (
áð(
ñ ð(
ñ ™X c©5¯8©8¯?©?Ð&:Õ;Õ<ð	(
ð
 
ñ(
r%   c                óÈ  <€ \         SV `  4        \        V 4       Vf   \        \        P
                  RRR7      p\        P                  ! \        ^^@^^^V4      \        P                  ! RRRR7      \        ^@^@^^VR7      \        ^@^À^^^V4      \        P                  ! RRRR7      \        ^À^@^`^€^^ ^ V4      \        R^€^€^À^ ^`^@V4      \        P                  ! RRRR7      \        R^À^`^Ð^^0^@V4      \        R^ ^p^à^^@^@V4      \        R^€^€R^^@^@V4      \        R^p^�R^ ^@^@V4      \        R	R^ R
^ ^€^€V4      \        P                  ! RRRR7      \        RR^ R
^ ^€^€V4      \        RR^ÀR^0^€^€V4      4      V n        \        P                  ! R^R7      V n        \        P                  ! \        P                  ! VR7      \        P                  ! RV^^RR7      4      V n        R # )Ngü©ñÒMbP?)ÚepsÚmomentumrC   rB   é   ià  i   i   i  i@  i@  i€  )r   r   )Úpi   T)r   r   r(   )r'   rE   rE   )r'   r   r   )r   r'   r'   rD   )r   r   r   )r'   r'   r'   )r   r   r   )r   é   rd   )r)   r*   r
   r   r   ÚBatchNorm3drG   r   rJ   r   r8   ÚfeaturesÚ	AvgPool3dÚavgpoolÚDropoutÚConv3dÚ
classifier)r+   rZ   r[   r   r,   s   &&&&€r"   r*   ÚS3D.__init__h   sÇ  ø€ ô 	‰ÑÔÜ˜DÔ!àÒÜ ¤§¡°UÀUÔKˆJäŸšÜ! ! R¨¨A¨q°*Ó=Ü�LŠL Y°yÈ)ÔTÜ ØØØØØ%ôô " " c¨1¨a°°JÓ?Ü�LŠL Y°yÈ)ÔTÜ  R¨¨S°"°b¸"¸jÓIÜ  S¨#¨s°B¸¸BÀ
ÓKÜ�LŠL Y°yÈ)ÔTÜ  S¨"¨c°2°r¸2¸zÓJÜ  S¨#¨s°B¸¸BÀ
ÓKÜ  S¨#¨s°B¸¸BÀ
ÓKÜ  S¨#¨s°B¸¸BÀ
ÓKÜ  S¨#¨s°B¸¸SÀ*ÓMÜ�LŠL Y°yÈ)ÔTÜ  S¨#¨s°B¸¸SÀ*ÓMÜ  S¨#¨s°B¸¸SÀ*ÓMó-
ˆŒô0 —|’|°	À!ÔDˆŒÜŸ-š-Ü�JŠJ˜Ô!Ü�IŠI�d˜K°Q¸qÀtÔLó
ˆŽr%   c                óœ   € V P                  V4      pV P                  V4      pV P                  V4      p\        P                  ! VRR7      pV# )r   )Údim)r   rE   é   )rf   rh   rk   rN   Úmean)r+   rP   s   &&r"   rV   ÚS3D.forward’   s@   € Ø�M‰M˜!ÓˆØ�L‰L˜‹OˆØ�O‰O˜AÓˆÜ�JŠJ�q˜iÔ(ˆØˆr%   )rh   rk   rf   )i�  gš™™™™™É?N)
r/   r0   r1   r2   Ú__doc__r*   rV   r3   r4   r5   r6   s   @@r"   r   r   \   s   ù‡ € ñ	÷(
õ (
÷Tò r%   c                   ór   € ] tR t^št]! R]! ]RRR7      RRR^R]RRRR	R
RRRRRRR//RRRR/	R7      t]t	Rt
R# )r   z4https://download.pytorch.org/models/s3d-d76dad2f.pth)Ú	crop_sizeÚresize_sizeÚmin_sizeÚmin_temporal_sizeÚ
categoriesÚrecipezOhttps://github.com/pytorch/vision/tree/main/references/video_classification#s3dÚ_docsz®The weights aim to approximate the accuracy of the paper. The accuracies are estimated on clip-level with parameters `frame_rate=15`, `clips_per_video=1`, and `clip_len=128`.Ú
num_paramsi0ô~ Ú_metricszKinetics-400zacc@1gd;ßO�Q@zacc@5g33333V@Ú_opsg´Èv¾Ÿú1@Ú
_file_sizegF¶óýÔø?@)ÚurlÚ
transformsÚmetar.   N)éà   r‚   )rb   rb   )r/   r0   r1   r2   r   r   r	   r   ÚKINETICS400_V1ÚDEFAULTr3   r.   r%   r"   r   r   š   sz   † ÙØBÙØØ Ø"ô
ð ˜
Ø ØÐ1ØÐgØð\ð ˜'ØØØ˜VØ˜Vð!ðð �FØ˜&ð#
ô€Nð6 „Gr%   Ú
pretrained)Úweightsr†   ÚprogressTc                ó^   € V ^8„  d   QhR\         \        ,          R\        R\        R\        /# )r   r†   r‡   Úkwargsr\   )r   r   Úboolr   r   )r    s   "r"   r#   r#   »   s/   € ÷  ñ  ”Hœ[Õ)ð  ¼Dð  ÔSVð  Ô[^ñ  r%   c                óè   € \         P                  V 4      p V e(   \        VR\        V P                  R,          4      4       \        R/ VB pV e#   VP                  V P                  VRR7      4       V# )a_  Construct Separable 3D CNN model.

Reference: `Rethinking Spatiotemporal Feature Learning <https://arxiv.org/abs/1712.04851>`__.

.. betastatus:: video module

Args:
    weights (:class:`~torchvision.models.video.S3D_Weights`, optional): The
        pretrained weights to use. See
        :class:`~torchvision.models.video.S3D_Weights`
        below for more details, and possible values. By default, no
        pre-trained weights are used.
    progress (bool): If True, displays a progress bar of the download to stderr. Default is True.
    **kwargs: parameters passed to the ``torchvision.models.video.S3D`` base class.
        Please refer to the `source code
        <https://github.com/pytorch/vision/blob/main/torchvision/models/video/s3d.py>`_
        for more details about this class.

.. autoclass:: torchvision.models.video.S3D_Weights
    :members:
rZ   rx   T)r‡   Ú
check_hashr.   )r   Úverifyr   Úlenr�   r   Úload_state_dictÚget_state_dict)r†   r‡   r‰   Úmodels   $$, r"   Ús3dr’   ¹   si   € ô0 × Ñ  Ó)€GàÒÜ˜f m´S¸¿¹ÀlÕ9SÓ5TÔUä‰M�&‰M€EàÒØ×Ñ˜g×4Ñ4¸hÐSWÐ4ÓXÔYà€Lr%   )r   r   r’   ) Ú	functoolsr   Útypingr   r   r   rN   r   Útorchvision.ops.miscr   Útransforms._presetsr	   Úutilsr
   Ú_apir   r   r   Ú_metar   Ú_utilsr   r   Ú__all__rG   r   r   r8   r   r   rƒ   r’   r.   r%   r"   Ú<module>rœ      s­   ðÝ ß *Ñ *ã Ý Ý 5å 6Ý (ß 7Ñ 7Ý +ß Cò€ô
˜BŸM™Mô 
ô@#˜"Ÿ)™)ô #ôL;ˆ"�)‰)ô ;ô|�+ô ñ> ÓÙ ,°×0JÑ0JÐ!KÔLð ¨Dð  À4ö  ó Mó ò r%   