+
    ƒ°ñja)  ã                   óx   € ^ RI t ^ RIHt ^ RIHtHt ^RIHt ^ RIt	^RI
HtHt  ! R R]P                  4      tR# )é    N)ÚTimestepEmbedderÚ
PatchEmbed)ÚAttentionPool)ÚHunYuanDiTBlockÚ	calc_ropec                   óz   a a€ ] tR t^t oRt^€^^R^(^RRR^MRRRRRR	]R
R
R
R
3V3R lV 3R llltRR ltRtVt	V ;t
# )ÚHunYuanControlNeta   
HunYuanDiT: Diffusion model with a Transformer backbone.

Inherit ModelMixin and ConfigMixin to be compatible with the sampler StableDiffusionPipeline of diffusers.

Inherit PeftAdapterMixin to be compatible with the PEFT training pipeline.

Parameters
----------
args: argparse.Namespace
    The arguments parsed by argparse.
input_size: tuple
    The size of the input image.
patch_size: int
    The size of the patch.
in_channels: int
    The number of input channels.
hidden_size: int
    The hidden size of the transformer backbone.
depth: int
    The number of transformer blocks.
num_heads: int
    The number of attention heads.
mlp_ratio: float
    The ratio of the hidden size of the MLP in the transformer block.
log_fn: callable
    The logging function.
i€  g†8ÖÅmt@é   i   é   TFÚlayerNc                óJ   <€ V ^8„  d   QhRS[ RS[RS[RS[RS[RS[RS[RS[/# )	é   Ú
input_sizeÚ
patch_sizeÚin_channelsÚhidden_sizeÚdepthÚ	num_headsÚ	mlp_ratioÚlog_fn)ÚtupleÚintÚfloatÚcallable)ÚformatÚ__classdict__s   "€Ú>C:\Ai_Coding\apps\studio\ComfyUI\comfy\ldm\hydit\controlnet.pyÚ__annotate__ÚHunYuanControlNet.__annotate__/   se   ø€ ÷ Z
ñ Z
áðZ
ñ ðZ
ñ ð	Z
ñ
 ðZ
ñ ðZ
ñ ðZ
ñ ðZ
ñ$ ñ%Z
ó    c                ó¸  <€ \         SV `  4        VV n        WPn        Wðn        W0n        V'       d
   V^,          MTV n        W n        W`n        W@n	        W€n
        W�n        W n        W°n        WÐn        Wàn        VV n        VV n        \$        P&                  P(                  V n        \,        P.                  ! \,        P0                  ! V P                  V P                  ^,          RVVR7      \,        P2                  ! 4       \,        P0                  ! V P                  ^,          V P                  RVVR7      4      V n        \,        P6                  ! \8        P:                  ! V P                  V P                  ,           V P                  VVR7      4      V n        Rp\?        V P                  V P                  ^VVVVR7      V n         VV n!        V P                  '       d   V ;PB                  R
,          un!        V P                  '       d9   \,        PD                  ! ^VVVR7      V n#        V ;PB                  V,          un!        \I        VVVVVVVR7      V n%        \M        VVVVR7      V n'        \,        P.                  ! VP1                  V PB                  V^,          VVR7      \,        P2                  ! 4       VP1                  V^,          VRVVR7      4      V n(        \,        PR                  ! \U        ^4       Uu. uF.  p\W        VVVVV P                  VV P                   RVVVVR7      NK0  	  up4      V n,        VP1                  V P                  V P                  VVR7      V n-        \,        PR                  ! \U        \]        V PX                  4      4       Uu. uF,  pVP1                  V P                  V P                  VVR7      NK.  	  up4      V n/        R	# u upi u upi )r   T)ÚbiasÚdtypeÚdevice)r#   r$   r
   )r   Ú
output_dimr#   r$   Ú
operations)r#   r$   r&   F)r   Ú
c_emb_sizer   r   Útext_states_dimÚqk_normÚ	norm_typeÚskipÚattn_precisionr#   r$   r&   Ni   )0ÚsuperÚ__init__r   r   Úlearn_sigmar   Úout_channelsr   r   r   r(   Útext_states_dim_t5Útext_lenÚtext_len_t5Ú	size_condÚuse_style_condÚnormr#   ÚcomfyÚlatent_formatsÚSDXLÚlatent_formatÚnnÚ
SequentialÚLinearÚSiLUÚmlp_t5Ú	ParameterÚtorchÚrandnÚtext_embedding_paddingr   ÚpoolerÚextra_in_dimÚ	EmbeddingÚstyle_embedderr   Ú
x_embedderr   Ú
t_embedderÚextra_embedderÚ
ModuleListÚranger   ÚblocksÚbefore_projÚlenÚafter_proj_list)Úselfr   r   r   r   r   r   r   r(   r1   r2   r3   r)   r4   r5   r/   r6   r   r,   r#   r$   r&   ÚkwargsÚpooler_out_dimÚ_Ú	__class__s   &&&&&&&&&&&&&&&&&&&&&&,  €r   r.   ÚHunYuanControlNet.__init__/   sx  ø€ ô2 	‰ÑÔØˆŒØŒ
Ø&ÔØ&Ôß/:˜K¨!žOÀˆÔØ$ŒØ"ŒØ&ÔØ.ÔØ"4ÔØ ŒØ&ÔØ"ŒØ,ÔØˆŒ	ØˆŒ
Ü"×1Ñ1×6Ñ6ˆÔä—m’mÜ�IŠIØ×'Ñ'Ø×'Ñ'¨!Õ+ØØØôô �GŠG‹IÜ�IŠIØ×'Ñ'¨!Õ+Ø×$Ñ$ØØØôó
ˆŒô$ ')§l¢lÜ�KŠKØ—‘ × 0Ñ 0Õ0Ø×$Ñ$ØØô	ó'
ˆÔ#ð ˆÜ#Ø×ÑØ×#Ñ#ØØ%ØØØ!ô
ˆŒð +ˆÔà�>�>ˆ>à×Ò Õ(Õà××Ðä"$§,¢,Ø�; e°Fô#ˆDÔð ×Ò Õ,Õô %ØØØØØØØ!ô
ˆŒô +Ø˜u¨VÀ
ô
ˆŒô !ŸmšmØ×ÑØ×!Ñ! ;°¥?¸%Èð ó ô �GŠG‹IØ×ÑØ˜a• °4¸uÈVð ó ó
ˆÔô —m’mô  ˜rœóñ #�Aô  Ø +Ø*Ø'Ø'Ø$(×$8Ñ$8Ø#Ø"Ÿi™iØØ#1ØØ!Ø)÷ñ #ñó
ˆŒð* &×,Ñ,¨T×-=Ñ-=¸t×?OÑ?OÐW\ÐekÐ,ÓlˆÔô  "Ÿ}š}ô œs 4§;¡;Ó/Ô0óñ
 1�Að ×%Ñ%Ø×(Ñ(¨$×*:Ñ*:À%ÐPVð &ö ñ 1ñó 
ˆÖùò1ùò2s   Ë44OÎ2Oc                ó  € TpVP                   ^ ,          ^8X  d*   \        P                  ! WÁP                   ^ ,          ^ R7      pTpTpVP                  4       pVP                  4       pVP                   w  pppV P	                  VP                  RV4      4      P                  VVR4      p\        P                  P                  V P                  V4      p\        P                  ! VRV P                  ) R13,          P                  ^4      VRV P                  ) R13,          VRV P                   4      VRV P                  ) R13&   \        P                  ! VRV P                  ) R13,          P                  ^4      VRV P                  ) R13,          VV P                  R 4      VRV P                  ) R13&   \        P                  ! WÞ.^R7      p\        WP                   V P"                  V P$                  ,          4      pV P'                  W0P(                  R7      pV P+                  V4      pV P-                  V4      pV	e,   V P/                  V	4      p\        P                  ! VV.^R7      pVV P1                  V4      ,           pV P+                  V4      p. pWP3                  V4      ,           p\5        V P6                  4       F9  w  ppV! VVVV4      pVP9                  V P:                  V,          ! V4      4       K;  	  RV/# )aK  
Forward pass of the encoder.

Parameters
----------
x: torch.Tensor
    (B, D, H, W)
t: torch.Tensor
    (B)
encoder_hidden_states: torch.Tensor
    CLIP text embedding, (B, L_clip, D)
text_embedding_mask: torch.Tensor
    CLIP text embedding mask, (B, L_clip)
encoder_hidden_states_t5: torch.Tensor
    T5 text embedding, (B, L_t5, D)
text_embedding_mask_t5: torch.Tensor
    T5 text embedding mask, (B, L_t5)
image_meta_size: torch.Tensor
    (B, 6)
style: torch.Tensor
    (B)
cos_cis_img: torch.Tensor
sin_cis_img: torch.Tensor
return_dict: bool
    Whether to return a dictionary.
)Údim:NNNN)r#   Úoutputéÿÿÿÿ)ÚshaperA   Úrepeat_interleaveÚboolr?   Úviewr7   ÚopsÚcast_to_inputrC   Úwherer2   Ú	unsqueezer3   Úcatr   r   r   r   rI   r#   rH   rD   rG   rJ   rN   Ú	enumeraterM   ÚappendrP   )rQ   ÚxÚhintÚ	timestepsÚcontextÚtext_embedding_maskÚencoder_hidden_states_t5Útext_embedding_mask_t5Úimage_meta_sizeÚstyleÚreturn_dictÚkwargÚ	conditionÚtext_statesÚtext_states_t5Útext_states_maskÚtext_states_t5_maskÚb_t5Úl_t5Úc_t5ÚpaddingÚfreqs_cis_imgÚtÚ	extra_vecÚstyle_embeddingÚcÚcontrolsr   Úblocks   &&&&&&&&&&&,                 r   ÚforwardÚHunYuanControlNet.forwardË   s¹  € ðP ˆ	Ø�?‰?˜1Õ Ô"Ü×/Ò/°	¿7¹7À1½:È1ÔMˆIàˆØ1ˆØ.×3Ñ3Ó5ÐØ4×9Ñ9Ó;ÐØ)×/Ñ/Ñˆˆd�DØŸ™ ^×%8Ñ%8¸¸TÓ%BÓC×HÑHÈÈtÐUWÓXˆä—)‘)×)Ñ)¨$×*EÑ*EÀ{ÓSˆä+0¯;ª;Ø˜Q §¡ Ñ 0Ð0Õ1×;Ñ;¸AÓ>Ø˜˜DŸM™M˜>Ñ+Ð+Õ,Ø�O�d—m‘mÐ$ó,
ˆ�A˜Ÿ™�~Ñ'Ð'Ñ(ô
 27·²Ø  D×$4Ñ$4Ð#4Ñ#6Ð 6Õ7×AÑAÀ!ÓDØ˜1˜t×/Ñ/Ð/Ñ1Ð1Õ2Ø�D—M‘M�OÐ$ó2
ˆ�q˜4×+Ñ+Ð+Ñ-Ð-Ñ.ô —i’i Ð =À1ÔEˆô "Ø�‰ × 0Ñ 0°D·N±NÕ Bó
ˆð
 �O‰O˜I¯Z©ZˆOÓ8ˆØ�O‰O˜AÓˆð —K‘KÐ 8Ó9ˆ	ð ÒØ"×1Ñ1°%Ó8ˆOÜŸ	š	 9¨oÐ">ÀAÔFˆIð �×#Ñ# IÓ.Õ.ˆð —O‘O IÓ.ˆ	ð ˆØ× Ñ  Ó+Õ+ˆÜ% d§k¡kÖ2‰LˆE�5Ù�a˜˜K¨Ó7ˆAØ�O‰O˜D×0Ñ0°Ö7¸Ó:Ö;ñ 3ð ˜(Ð#Ð#r    )rP   rN   rM   r   r#   rJ   rE   r   r   r:   r/   r   r?   r6   r   r0   r   rD   r4   rG   rI   rC   r2   r3   r(   r1   r5   rH   )NNNNNF)Ú__name__Ú
__module__Ú__qualname__Ú__firstlineno__Ú__doc__Úprintr.   r�   Ú__static_attributes__Ú__classdictcell__Ú__classcell__)rU   r   s   @@r   r	   r	      sd   ù‡ € ñð>  ØØØØØØ!ØØØØØØØØØØ ØØØØ÷-Z
õ Z
÷xl$ô l$r    r	   )rA   Útorch.nnr;   Ú(comfy.ldm.modules.diffusionmodules.mmditr   r   Úpoolersr   Úcomfy.latent_formatsr7   Úmodelsr   r   ÚModuler	   © r    r   Ú<module>r“      s-   ðã Ý ÷õ #ã ß .ôf$˜Ÿ	™	ö f$r    