+
    œ°³j5¯  ã                   óL  € R t ^ RIt^ RIHt ^ RIt^ RIHt ^ RIHtHtH	t	 ^RI
Ht ^RIHtHt ^RIHt ^RIHt ^R	IHtHtHtHtHtHt ^R
IHt ^RIHt ^RIH t  ^RI!H"t"H#t#H$t$H%t% ^RI&H't' ^RI(H)t) ^RI*H+t+ ]%PX                  ! ]-4      t. ! R R]P^                  4      t0 ! R R]P^                  4      t1 ! R R]P^                  4      t2 ! R R]P^                  4      t3 ! R R]P^                  4      t4 ! R R]P^                  4      t5 ! R R]P^                  4      t6 ! R R ]P^                  4      t7 ! R! R"]4      t8]# ! R# R$]4      4       t9 ! R% R&]P^                  4      t: ! R' R(]P^                  4      t; ! R) R*]P^                  4      t<]# ! R+ R,]94      4       t= ! R- R.]P^                  4      t>]# ! R/ R0]94      4       t? ! R1 R2]P^                  4      t@]#! R3R47       ! R5 R6]94      4       tA]# ! R7 R8]94      4       tB]# ! R9 R:]94      4       tC]# ! R; R<]94      4       tD. R=OtER# )>zPyTorch ConvBERT model.N)ÚCallable)Únn)ÚBCEWithLogitsLossÚCrossEntropyLossÚMSELoss)Úinitialization)ÚACT2FNÚget_activation)Úcreate_bidirectional_mask)ÚGradientCheckpointingLayer)Ú"BaseModelOutputWithCrossAttentionsÚMaskedLMOutputÚMultipleChoiceModelOutputÚQuestionAnsweringModelOutputÚSequenceClassifierOutputÚTokenClassifierOutput)ÚPreTrainedModel)ÚUnpack)Úapply_chunking_to_forward)ÚTransformersKwargsÚauto_docstringÚcan_return_tupleÚlogging)Úmerge_with_config_defaults)Úcapture_outputs©ÚConvBertConfigc                   óL   a a€ ] tR t^4t oRtV 3R ltRV3R lR lltRtVtV ;t	# )ÚConvBertEmbeddingszGConstruct the embeddings from word, position and token_type embeddings.c                óN  <€ \         SV `  4        \        P                  ! VP                  VP
                  VP                  R 7      V n        \        P                  ! VP                  VP
                  4      V n	        \        P                  ! VP                  VP
                  4      V n        \        P                  ! VP
                  VP                  R7      V n        \        P                  ! VP                  4      V n        \        P"                  ! \$        P&                  ! VP                  4      P)                  R4      RR7      V n        \        P"                  ! \$        P,                  ! V P*                  P/                  4       \$        P0                  R7      RR7      V n        R# ))Úpadding_idx©ÚepsF)Ú
persistent©ÚdtypeN©é   éÿÿÿÿ)ÚsuperÚ__init__r   Ú	EmbeddingÚ
vocab_sizeÚembedding_sizeÚpad_token_idÚword_embeddingsÚmax_position_embeddingsÚposition_embeddingsÚtype_vocab_sizeÚtoken_type_embeddingsÚ	LayerNormÚlayer_norm_epsÚDropoutÚhidden_dropout_probÚdropoutÚBufferÚtorchÚarangeÚexpandÚposition_idsÚzerosÚsizeÚlongÚtoken_type_ids©ÚselfÚconfigÚ	__class__s   &&€Úo/home/luca/screen-checking/.venv/lib/python3.14/site-packages/transformers/models/convbert/modeling_convbert.pyr*   ÚConvBertEmbeddings.__init__7   s  ø€ Ü‰ÑÔÜ!Ÿ|š|¨F×,=Ñ,=¸v×?TÑ?TÐbh×buÑbuÔvˆÔÜ#%§<¢<°×0NÑ0NÐPV×PeÑPeÓ#fˆÔ Ü%'§\¢\°&×2HÑ2HÈ&×J_ÑJ_Ó%`ˆÔ"äŸš f×&;Ñ&;À×AVÑAVÔWˆŒÜ—z’z &×"<Ñ"<Ó=ˆŒäŸIšI¤e§l¢l°6×3QÑ3QÓ&R×&YÑ&YÐZaÓ&bÐotÔuˆÔÜ Ÿiši¬¯ª°D×4EÑ4E×4JÑ4JÓ4LÔTY×T^ÑT^Ô(_ÐlqÔrˆÖó    c          
      óÔ   <€ V ^8„  d   QhRS[ P                  R,          RS[ P                  R,          RS[ P                  R,          RS[ P                  R,          RS[ P                  /# )é   Ú	input_idsNrA   r=   Úinputs_embedsÚreturn)r:   Ú
LongTensorÚFloatTensor)ÚformatÚ__classdict__s   "€rF   Ú__annotate__ÚConvBertEmbeddings.__annotate__C   so   ø€ ÷ $ñ $á×#Ñ# dÕ*ð$ñ ×(Ñ(¨4Õ/ð$ñ ×&Ñ&¨Õ-ð	$ñ
 ×(Ñ(¨4Õ/ð$ñ 
×	Ñ	ñ$rH   c                óv  € Ve   VP                  4       pMVP                  4       R R pV^,          pVf   V P                  RR V13,          pVf�   \        V R4      '       d4   V P                  RR V13,          pVP	                  V^ ,          V4      pTpM;\
        P                  ! V\
        P                  V P                  P                  R7      pVf   V P                  V4      pV P                  V4      p	V P                  V4      p
WI,           V
,           pV P                  V4      pV P                  V4      pV# )NºNNNrA   ©r%   Údevicer(   )r?   r=   ÚhasattrrA   r<   r:   r>   r@   rW   r/   r1   r3   r4   r8   )rC   rK   rA   r=   rL   Úinput_shapeÚ
seq_lengthÚbuffered_token_type_idsÚ buffered_token_type_ids_expandedr1   r3   Ú
embeddingss   &&&&&       rF   ÚforwardÚConvBertEmbeddings.forwardC   s,  € ð Ò Ø#Ÿ.™.Ó*‰Kà'×,Ñ,Ó.¨s°Ð3ˆKà  •^ˆ
àÒØ×,Ñ,¨Q°°°¨^Õ<ˆLð
 Ò!Ü�tÐ-×.Ò.Ø*.×*=Ñ*=¸aÀÀ*À¸nÕ*MÐ'Ø3J×3QÑ3QÐR]Ð^_ÕR`ÐblÓ3mÐ0Ø!A‘ä!&§¢¨[ÄÇ
Á
ÐSW×SdÑSd×SkÑSkÔ!l�àÒ Ø ×0Ñ0°Ó;ˆMØ"×6Ñ6°|ÓDÐØ $× :Ñ :¸>Ó JÐà"Õ8Ð;PÕPˆ
Ø—^‘^ JÓ/ˆ
Ø—\‘\ *Ó-ˆ
ØÐrH   )r4   r8   r1   r=   r3   rA   r/   )NNNN©
Ú__name__Ú
__module__Ú__qualname__Ú__firstlineno__Ú__doc__r*   r^   Ú__static_attributes__Ú__classdictcell__Ú__classcell__©rE   rQ   s   @@rF   r   r   4   s   ù‡ € ÙQõ
s÷$÷ $ò $rH   r   c                   óH   a a€ ] tR t^jt oRtV 3R ltV3R lR ltRtVtV ;t	# )ÚSeparableConv1DzSThis class implements separable convolution, i.e. a depthwise and a pointwise layerc           	     óü  <€ \         SV `  4        \        P                  ! VVVVV^,          RR7      V n        \        P                  ! W#^RR7      V n        \        P                  ! \        P                  ! V^4      4      V n	        V P                  P                  P                  P                  RVP                  R7       V P
                  P                  P                  P                  RVP                  R7       R# )rJ   F)Úkernel_sizeÚgroupsÚpaddingÚbias)rm   rp   ç        ©ÚmeanÚstdN)r)   r*   r   ÚConv1dÚ	depthwiseÚ	pointwiseÚ	Parameterr:   r>   rp   ÚweightÚdataÚnormal_Úinitializer_range)rC   rD   Úinput_filtersÚoutput_filtersrm   ÚkwargsrE   s   &&&&&,€rF   r*   ÚSeparableConv1D.__init__m   s¹   ø€ Ü‰ÑÔÜŸšØØØ#Ø Ø 1Õ$Øô
ˆŒô Ÿš =ÈaÐV[Ô\ˆŒÜ—L’L¤§¢¨^¸QÓ!?Ó@ˆŒ	à�‰×Ñ×"Ñ"×*Ñ*°¸×9QÑ9QÐ*ÔRØ�‰×Ñ×"Ñ"×*Ñ*°¸×9QÑ9QÐ*ÖRrH   c                óN   <€ V ^8„  d   QhRS[ P                  RS[ P                  /# ©rJ   Úhidden_statesrM   ©r:   ÚTensor)rP   rQ   s   "€rF   rR   ÚSeparableConv1D.__annotate__}   s#   ø€ ÷ ñ ¡U§\¡\ð ±e·l±lñ rH   c                ón   € V P                  V4      pV P                  V4      pW P                  ,          pV# ©N)rv   rw   rp   )rC   rƒ   Úxs   && rF   r^   ÚSeparableConv1D.forward}   s.   € Ø�N‰N˜=Ó)ˆØ�N‰N˜1ÓˆØ	�Y‰Y�ˆØˆrH   )rp   rv   rw   r`   ri   s   @@rF   rk   rk   j   s   ù‡ € Ù]õS÷ ÷ ð rH   rk   c                   óH   a a€ ] tR t^„t oV 3R ltRV3R lR lltRtVtV ;t# )ÚConvBertSelfAttentionc                óÖ  <€ \         SV `  4        VP                  VP                  ,          ^ 8w  d9   \	        VR4      '       g'   \        RVP                   RVP                   R24      hVP                  VP                  ,          pV^8  d   VP                  V n        ^V n        MW n        VP                  V n        VP                  V n        VP                  V P                  ,          ^ 8w  d   \        R4      hVP                  V P                  ,          ^,          V n        V P                  V P                  ,          V n	        \        P                  ! VP                  V P                  4      V n        \        P                  ! VP                  V P                  4      V n        \        P                  ! VP                  V P                  4      V n        \        WP                  V P                  V P                  4      V n        \        P                  ! V P                  V P                  V P                  ,          4      V n        \        P                  ! VP                  V P                  4      V n        \        P&                  ! V P                  ^.\)        V P                  ^,
          ^,          4      ^ .R7      V n        \        P,                  ! VP.                  4      V n        R# )é    r-   zThe hidden size (z6) is not a multiple of the number of attention heads (Ú)z6hidden_size should be divisible by num_attention_heads)rm   ro   N)r)   r*   Úhidden_sizeÚnum_attention_headsrX   Ú
ValueErrorÚ
head_ratioÚconv_kernel_sizeÚattention_head_sizeÚall_head_sizer   ÚLinearÚqueryÚkeyÚvaluerk   Úkey_conv_attn_layerÚconv_kernel_layerÚconv_out_layerÚUnfoldÚintÚunfoldr6   Úattention_probs_dropout_probr8   )rC   rD   Únew_num_attention_headsrE   s   && €rF   r*   ÚConvBertSelfAttention.__init__…   s=  ø€ Ü‰ÑÔØ×Ñ × :Ñ :Õ:¸aÔ?ÌÐPVÐXh×HiÒHiÜØ# F×$6Ñ$6Ð#7ð 8Ø ×4Ñ4Ð5°Qð8óð ð
 #)×"<Ñ"<À×@QÑ@QÕ"QÐØ" QÔ&Ø$×8Ñ8ˆDŒOØ'(ˆDÕ$à'>Ô$Ø$×/Ñ/ˆDŒOà &× 7Ñ 7ˆÔØ×Ñ × 8Ñ 8Õ8¸AÔ=ÜÐUÓVÐVà$*×$6Ñ$6¸$×:RÑ:RÕ$RÐWXÕ#XˆÔ Ø!×5Ñ5¸×8PÑ8PÕPˆÔä—Y’Y˜v×1Ñ1°4×3EÑ3EÓFˆŒ
Ü—9’9˜V×/Ñ/°×1CÑ1CÓDˆŒÜ—Y’Y˜v×1Ñ1°4×3EÑ3EÓFˆŒ
ä#2Ø×&Ñ&¨×(:Ñ(:¸D×<QÑ<Qó$
ˆÔ ô "$§¢¨4×+=Ñ+=¸t×?WÑ?WÐZ^×ZoÑZoÕ?oÓ!pˆÔÜ Ÿiši¨×(:Ñ(:¸D×<NÑ<NÓOˆÔä—i’iØ×.Ñ.°Ð2¼SÀ$×BWÑBWÐZ[ÕB[Ð_`ÕA`Ó=aÐcdÐ<eô
ˆŒô —z’z &×"EÑ"EÓFˆŽrH   c                óÞ   <€ V ^8„  d   QhRS[ P                  RS[ P                  R,          RS[ P                  R,          RS[S[,          RS[S[ P                  S[ P                  3,          /# ©rJ   rƒ   Úattention_maskNÚencoder_hidden_statesr   rM   )r:   r…   rO   r   r   Útuple)rP   rQ   s   "€rF   rR   Ú"ConvBertSelfAttention.__annotate__¬   st   ø€ ÷ O.ñ O.á—|‘|ðO.ñ ×)Ñ)¨DÕ0ðO.ñ  %Ÿ|™|¨dÕ2ð	O.ñ
 Ñ+Õ,ðO.ñ 
‰u�|‰|™UŸ\™\Ð)Õ	*ñO.rH   c                ón  € VP                   R R p. VORNV P                  N5pVe$   V P                  V4      pV P                  V4      pM"V P                  V4      pV P                  V4      pV P	                  VP                  ^^4      4      p	V	P                  ^^4      p	V P                  V4      p
V
P                  V4      P                  ^^4      pVP                  V4      P                  ^^4      pVP                  V4      P                  ^^4      p\        P                  ! Wš4      pV P                  V4      p\        P                  ! VRV P                  ^.4      p\        P                  ! V^R7      pV P                  V4      p\        P                  ! VV^ ,          RV P                  .4      pVP                  ^^4      P!                  4       P#                  R4      p\$        P&                  P)                  VV P                  ^.^V P                  ^,
          ^,          ^ .^R7      pVP                  ^^4      P                  V^ ,          RV P                  V P                  4      p\        P                  ! VRV P                  V P                  .4      p\        P*                  ! VV4      p\        P                  ! VRV P                  .4      p\        P*                  ! W¼P                  RR4      4      pV\,        P.                  ! V P                  4      ,          pVe
   VV,           p\$        P&                  P                  VRR7      pV P1                  V4      p\        P*                  ! VV4      pVP3                  ^ ^^^4      P!                  4       p\        P                  ! VV^ ,          RV P4                  V P                  .4      p\        P6                  ! VV.^4      pVP9                  4       R R V P4                  V P                  ,          ^,          3,           pVP                  ! V!  pVV3# )N©Údim)rm   Údilationro   Ústrider(   éþÿÿÿ)Úshaper•   r™   rš   r›   Ú	transposer˜   Úviewr:   Úmultiplyrœ   Úreshaper”   Úsoftmaxr�   r–   Ú
contiguousÚ	unsqueezer   Ú
functionalr    ÚmatmulÚmathÚsqrtr8   Úpermuter‘   Úcatr?   )rC   rƒ   r¦   r§   r   rY   Úhidden_shapeÚmixed_key_layerÚmixed_value_layerÚmixed_key_conv_attn_layerÚmixed_query_layerÚquery_layerÚ	key_layerÚvalue_layerÚconv_attn_layerrœ   r�   Úattention_scoresÚattention_probsÚcontext_layerÚconv_outÚnew_context_layer_shapes   &&&&,                 rF   r^   ÚConvBertSelfAttention.forward¬   sÈ  € ð $×)Ñ)¨#¨2Ð.ˆØC˜ÐC bÐC¨$×*BÑ*BÑCˆð !Ò,Ø"Ÿh™hÐ'<Ó=ˆOØ $§
¡
Ð+@Ó AÑà"Ÿh™h }Ó5ˆOØ $§
¡
¨=Ó 9Ðà$(×$<Ñ$<¸]×=TÑ=TÐUVÐXYÓ=ZÓ$[Ð!Ø$=×$GÑ$GÈÈ1Ó$MÐ!à ŸJ™J }Ó5ÐØ'×,Ñ,¨\Ó:×DÑDÀQÈÓJˆà#×(Ñ(¨Ó6×@Ñ@ÀÀAÓFˆ	Ø'×,Ñ,¨\Ó:×DÑDÀQÈÓJˆäŸ.š.Ð)BÓVˆà ×2Ñ2°?ÓCÐÜ!ŸMšMÐ*;¸bÀ$×BWÑBWÐYZÐ=[Ó\ÐÜ!ŸMšMÐ*;ÀÔCÐà×,Ñ,¨]Ó;ˆÜŸš ~¸ÀA½ÈÈD×L^ÑL^Ð7_Ó`ˆØ'×1Ñ1°!°QÓ7×BÑBÓD×NÑNÈrÓRˆÜŸ™×-Ñ-ØØ×.Ñ.°Ð2ØØ×+Ñ+¨aÕ/°AÕ5°qÐ9Øð .ó 
ˆð (×1Ñ1°!°QÓ7×?Ñ?Ø˜�N˜B × 2Ñ 2°D×4IÑ4Ió
ˆô Ÿš ~¸¸D×<TÑ<TÐVZ×VkÑVkÐ7lÓmˆÜŸš nÐ6GÓHˆÜŸš ~¸¸D×<NÑ<NÐ7OÓPˆô !Ÿ<š<¨×5HÑ5HÈÈRÓ5PÓQÐØ+¬d¯iªi¸×8PÑ8PÓ.QÕQÐØÒ%à/°.Õ@Ðô Ÿ-™-×/Ñ/Ð0@ÀbÐ/ÓIˆð Ÿ,™, Ó7ˆäŸš _°kÓBˆØ%×-Ñ-¨a°°A°qÓ9×DÑDÓFˆä—=’=Ø˜[¨�^¨R°×1IÑ1IÈ4×KcÑKcÐdó
ˆô Ÿ	š	 =°(Ð";¸QÓ?ˆð #0×"4Ñ"4Ó"6°s¸Ð";Ø×$Ñ$ t×'?Ñ'?Õ?À!ÕCð?
õ #
Ðð &×*Ò*Ð,CÑDˆà˜oÐ-Ð-rH   )r–   r•   rœ   r”   r�   r8   r“   r™   r›   r‘   r˜   r    rš   ©NN©	ra   rb   rc   rd   r*   r^   rf   rg   rh   ri   s   @@rF   rŒ   rŒ   „   s   ù‡ € õ%G÷NO.÷ O.ò O.rH   rŒ   c                   óD   a a€ ] tR t^þt oV 3R ltV3R lR ltRtVtV ;t# )ÚConvBertSelfOutputc                ó0  <€ \         SV `  4        \        P                  ! VP                  VP                  4      V n        \        P                  ! VP                  VP                  R 7      V n        \        P                  ! VP                  4      V n
        R# ©r!   N)r)   r*   r   r—   r�   Údenser4   r5   r6   r7   r8   rB   s   &&€rF   r*   ÚConvBertSelfOutput.__init__ÿ   s`   ø€ Ü‰ÑÔÜ—Y’Y˜v×1Ñ1°6×3EÑ3EÓFˆŒ
ÜŸš f×&8Ñ&8¸f×>SÑ>SÔTˆŒÜ—z’z &×"<Ñ"<Ó=ˆŽrH   c                óh   <€ V ^8„  d   QhRS[ P                  RS[ P                  RS[ P                  /# ©rJ   rƒ   Úinput_tensorrM   r„   )rP   rQ   s   "€rF   rR   ÚConvBertSelfOutput.__annotate__  ó1   ø€ ÷ ñ ¡U§\¡\ð ÁÇÁð ÑRW×R^ÑR^ñ rH   c                óx   € V P                  V4      pV P                  V4      pV P                  W,           4      pV# rˆ   ©rÓ   r8   r4   ©rC   rƒ   r×   s   &&&rF   r^   ÚConvBertSelfOutput.forward  ó5   € ØŸ
™
 =Ó1ˆØŸ™ ]Ó3ˆØŸ™ }Õ'CÓDˆØÐrH   ©r4   rÓ   r8   rÎ   ri   s   @@rF   rÐ   rÐ   þ   s   ù‡ € õ>÷÷ ð rH   rÐ   c                   óH   a a€ ] tR tRt oV 3R ltRV3R lR lltRtVtV ;t# )ÚConvBertAttentioni  c                ód   <€ \         SV `  4        \        V4      V n        \	        V4      V n        R # rˆ   )r)   r*   rŒ   rC   rÐ   ÚoutputrB   s   &&€rF   r*   ÚConvBertAttention.__init__  s&   ø€ Ü‰ÑÔÜ)¨&Ó1ˆŒ	Ü(¨Ó0ˆŽrH   c          
      ó´   <€ V ^8„  d   QhRS[ P                  RS[ P                  R,          RS[ P                  R,          RS[S[,          RS[ P                  /# r¥   ©r:   r…   rO   r   r   )rP   rQ   s   "€rF   rR   ÚConvBertAttention.__annotate__  s_   ø€ ÷  ñ  á—|‘|ð ñ ×)Ñ)¨DÕ0ð ñ  %Ÿ|™|¨dÕ2ð	 ñ
 Ñ+Õ,ð ñ 
�‰ñ rH   c                óX   € V P                   ! VV3R V/VB w  rVV P                  WQ4      pV# )r§   )rC   rã   )rC   rƒ   r¦   r§   r   rÉ   Ú_Úattention_outputs   &&&&,   rF   r^   ÚConvBertAttention.forward  sF   € ð  Ÿ9š9ØØñ
ð #8ð
ð ñ	
Ñˆð  Ÿ;™; }ÓDÐØÐrH   )rã   rC   rÍ   rÎ   ri   s   @@rF   rá   rá     s   ù‡ € õ1÷
 ÷  ò  rH   rá   c                   óD   a a€ ] tR tRt oV 3R ltV3R lR ltRtVtV ;t# )ÚGroupedLinearLayeri#  c                óÌ  <€ \         SV `  4        Wn        W n        W0n        V P                  V P                  ,          V n        V P                  V P                  ,          V n        \        P                  ! \        P                  ! V P                  V P
                  V P                  4      4      V n        \        P                  ! \        P                  ! V4      4      V n        R # rˆ   )r)   r*   Ú
input_sizeÚoutput_sizeÚ
num_groupsÚgroup_in_dimÚgroup_out_dimr   rx   r:   Úemptyry   rp   )rC   rï   rð   rñ   rE   s   &&&&€rF   r*   ÚGroupedLinearLayer.__init__$  s‘   ø€ Ü‰ÑÔØ$ŒØ&ÔØ$ŒØ ŸO™O¨t¯©Õ>ˆÔØ!×-Ñ-°·±Õ@ˆÔÜ—l’l¤5§;¢;¨t¯©À×@QÑ@QÐSW×SeÑSeÓ#fÓgˆŒÜ—L’L¤§¢¨[Ó!9Ó:ˆŽ	rH   c                óN   <€ V ^8„  d   QhRS[ P                  RS[ P                  /# r‚   r„   )rP   rQ   s   "€rF   rR   ÚGroupedLinearLayer.__annotate__.  s#   ø€ ÷ ñ ¡U§\¡\ð ±e·l±lñ rH   c                ó˜  € \        VP                  4       4      ^ ,          p\        P                  ! VRV P                  V P
                  .4      pVP                  ^^ ^4      p\        P                  ! W0P                  4      pVP                  ^^ ^4      p\        P                  ! W2RV P                  .4      pW0P                  ,           pV# )rŽ   r(   )Úlistr?   r:   r´   rñ   rò   r¼   r¹   ry   rð   rp   )rC   rƒ   Ú
batch_sizer‰   s   &&  rF   r^   ÚGroupedLinearLayer.forward.  sœ   € Ü˜-×,Ñ,Ó.Ó/°Õ2ˆ
Ü�MŠM˜-¨"¨d¯o©o¸t×?PÑ?PÐ)QÓRˆØ�I‰I�a˜˜AÓˆÜ�LŠL˜ŸK™KÓ(ˆØ�I‰I�a˜˜AÓˆÜ�MŠM˜!¨"¨d×.>Ñ.>Ð?Ó@ˆØ—	‘	�MˆØˆrH   )rp   rò   ró   rï   rñ   rð   ry   rÎ   ri   s   @@rF   rí   rí   #  s   ù‡ € õ;÷÷ ð rH   rí   c                   óD   a a€ ] tR tRt oV 3R ltV3R lR ltRtVtV ;t# )ÚConvBertIntermediatei9  c                ó¨  <€ \         SV `  4        VP                  ^8X  d2   \        P                  ! VP
                  VP                  4      V n        M1\        VP
                  VP                  VP                  R7      V n        \        VP                  \        4      '       d   \        VP                  ,          V n        R# VP                  V n        R# )r'   ©rï   rð   rñ   N)r)   r*   rñ   r   r—   r�   Úintermediate_sizerÓ   rí   Ú
isinstanceÚ
hidden_actÚstrr   Úintermediate_act_fnrB   s   &&€rF   r*   ÚConvBertIntermediate.__init__:  s–   ø€ Ü‰ÑÔØ×Ñ Ô!ÜŸš 6×#5Ñ#5°v×7OÑ7OÓPˆD�Jä+Ø!×-Ñ-¸6×;SÑ;SÐ`f×`qÑ`qôˆDŒJô �f×'Ñ'¬×-Ò-Ü'-¨f×.?Ñ.?Õ'@ˆDÖ$à'-×'8Ñ'8ˆDÖ$rH   c                óN   <€ V ^8„  d   QhRS[ P                  RS[ P                  /# r‚   r„   )rP   rQ   s   "€rF   rR   Ú!ConvBertIntermediate.__annotate__G  s#   ø€ ÷ ñ ¡U§\¡\ð ±e·l±lñ rH   c                óJ   € V P                  V4      pV P                  V4      pV# rˆ   ©rÓ   r  ©rC   rƒ   s   &&rF   r^   ÚConvBertIntermediate.forwardG  s&   € ØŸ
™
 =Ó1ˆØ×0Ñ0°Ó?ˆØÐrH   r	  rÎ   ri   s   @@rF   rý   rý   9  s   ù‡ € õ9÷÷ ð rH   rý   c                   óD   a a€ ] tR tRt oV 3R ltV3R lR ltRtVtV ;t# )ÚConvBertOutputiM  c                ó¶  <€ \         SV `  4        VP                  ^8X  d2   \        P                  ! VP
                  VP                  4      V n        M1\        VP
                  VP                  VP                  R7      V n        \        P                  ! VP                  VP                  R7      V n	        \        P                  ! VP                  4      V n        R# )r'   rÿ   r!   N)r)   r*   rñ   r   r—   r   r�   rÓ   rí   r4   r5   r6   r7   r8   rB   s   &&€rF   r*   ÚConvBertOutput.__init__N  s–   ø€ Ü‰ÑÔØ×Ñ Ô!ÜŸš 6×#;Ñ#;¸V×=OÑ=OÓPˆD�Jä+Ø!×3Ñ3À×ASÑASÐ`f×`qÑ`qôˆDŒJô Ÿš f×&8Ñ&8¸f×>SÑ>SÔTˆŒÜ—z’z &×"<Ñ"<Ó=ˆŽrH   c                óh   <€ V ^8„  d   QhRS[ P                  RS[ P                  RS[ P                  /# rÖ   r„   )rP   rQ   s   "€rF   rR   ÚConvBertOutput.__annotate__Y  rÙ   rH   c                óx   € V P                  V4      pV P                  V4      pV P                  W,           4      pV# rˆ   rÛ   rÜ   s   &&&rF   r^   ÚConvBertOutput.forwardY  rÞ   rH   rß   rÎ   ri   s   @@rF   r  r  M  s   ù‡ € õ	>÷÷ ð rH   r  c                   óN   a a€ ] tR tRt oV 3R ltRV3R lR lltR tRtVtV ;t	# )ÚConvBertLayeri`  c                ó|  <€ \         SV `  4        VP                  V n        ^V n        \	        V4      V n        VP                  V n        VP                  V n        V P                  '       d1   V P                  '       g   \        V  R24      h\	        V4      V n	        \        V4      V n        \        V4      V n        R# )r'   z> should be used as a decoder model if cross attention is addedN)r)   r*   Úchunk_size_feed_forwardÚseq_len_dimrá   Ú	attentionÚ
is_decoderÚadd_cross_attentionÚ	TypeErrorÚcrossattentionrý   Úintermediater  rã   rB   s   &&€rF   r*   ÚConvBertLayer.__init__a  s™   ø€ Ü‰ÑÔØ'-×'EÑ'EˆÔ$ØˆÔÜ*¨6Ó2ˆŒØ ×+Ñ+ˆŒØ#)×#=Ñ#=ˆÔ Ø×#×#Ð#Ø—?—?�?Ü 4 &Ð(fÐ gÓhÐhÜ"3°FÓ";ˆDÔÜ0°Ó8ˆÔÜ$ VÓ,ˆŽrH   c                óÜ   <€ V ^8„  d   QhRS[ P                  RS[ P                  R,          RS[ P                  R,          RS[ P                  R,          RS[S[,          RS[ P                  /# )rJ   rƒ   r¦   Nr§   Úencoder_attention_maskr   rM   ræ   )rP   rQ   s   "€rF   rR   ÚConvBertLayer.__annotate__o  sr   ø€ ÷ ñ á—|‘|ðñ ×)Ñ)¨DÕ0ðñ  %Ÿ|™|¨dÕ2ð	ñ
 !&§¡¨tÕ 3ðñ Ñ+Õ,ðñ 
�‰ñrH   c                ó   € V P                   ! VV3/ VB pV P                  '       d<   Ve8   \        V R4      '       g   \        RV  R24      hV P                  ! VV3RV/VB p\        V P                  V P                  V P                  V4      pV# )Nr  z'If `encoder_hidden_states` are passed, z` has to be instantiated with cross-attention layers by setting `config.add_cross_attention=True`r§   )	r  r  rX   ÚAttributeErrorr  r   Úfeed_forward_chunkr  r  )rC   rƒ   r¦   r§   r!  r   rê   Úlayer_outputs   &&&&&,  rF   r^   ÚConvBertLayer.forwardo  sÀ   € ð  Ÿ>š>ØØñ
ð ñ
Ðð �?�?ˆ?Ð4Ò@Ü˜4Ð!1×2Ò2Ü$Ø=¸d¸Vð DDð Dóð ð  $×2Ò2Ø Ø&ñ ð '<ð ð ñ	 Ðô 1Ø×#Ñ# T×%AÑ%AÀ4×CSÑCSÐUeó
ˆð ÐrH   c                óJ   € V P                  V4      pV P                  W!4      pV# rˆ   )r  rã   )rC   rê   Úintermediate_outputr&  s   &&  rF   r%  Ú ConvBertLayer.feed_forward_chunk�  s)   € Ø"×/Ñ/Ð0@ÓAÐØ—{‘{Ð#6ÓIˆØÐrH   )r  r  r  r  r  r  rã   r  ©NNN)
ra   rb   rc   rd   r*   r^   r%  rf   rg   rh   ri   s   @@rF   r  r  `  s   ù‡ € õ-÷ò ÷@ò rH   r  c                   óz   a a€ ] tR tRt oRtRtR]R]/t]	P                  ! 4       V 3R l4       tV3R ltRtVtV ;t# )	ÚConvBertPreTrainedModeli•  ÚconvbertTrƒ   Ú
attentionsc                ó¢  <€ \         SV `  V4       \        V\        4      '       d#   \        P
                  ! VP                  4       R# \        V\        4      '       dZ   \        P                  ! VP                  RV P                  P                  R7       \        P
                  ! VP                  4       R# \        V\        4      '       d‚   \        P                  ! VP                  \        P                   ! VP                  P"                  R,          4      P%                  R4      4       \        P
                  ! VP&                  4       R# R# )zInitialize the weightsrq   rr   Nr(   r&   )r)   Ú_init_weightsr  rk   ÚinitÚzeros_rp   rí   r{   ry   rD   r|   r   Úcopy_r=   r:   r;   r°   r<   rA   )rC   ÚmodulerE   s   &&€rF   r1  Ú%ConvBertPreTrainedModel._init_weightsŸ  sË   ø€ ô 	‰Ñ˜fÔ%Ü�fœo×.Ò.Ü�KŠK˜Ÿ™Ö$Ü˜Ô 2×3Ò3Ü�LŠL˜Ÿ™¨S°d·k±k×6SÑ6SÕTÜ�KŠK˜Ÿ™Ö$Ü˜Ô 2×3Ò3Ü�JŠJ�v×*Ñ*¬E¯LªL¸×9LÑ9L×9RÑ9RÐSUÕ9VÓ,W×,^Ñ,^Ð_fÓ,gÔhÜ�KŠK˜×-Ñ-Ö.ñ 4rH   c                ó&   <€ V ^8„  d   Qh/ S[ ;R&   # ©rJ   rD   r   )rP   rQ   s   "€rF   rR   Ú$ConvBertPreTrainedModel.__annotate__•  s   ø‡ ‚ áÑò rH   © )ra   rb   rc   rd   Úbase_model_prefixÚsupports_gradient_checkpointingr  rŒ   Ú_can_record_outputsr:   Úno_gradr1  Ú__annotate_func__rf   rg   rh   ri   s   @@rF   r-  r-  •  sG   ù‡ € ð #ÐØ&*Ð#à˜ØÐ+ðÐð
 ‡]‚]ƒ_ô
/ó ð
/÷ … rH   r-  c                   óH   a a€ ] tR tRt oV 3R ltRV3R lR lltRtVtV ;t# )ÚConvBertEncoderi­  c                óÖ   <€ \         SV `  4        Wn        \        P                  ! \        VP                  4       Uu. uF  p\        V4      NK  	  up4      V n        R V n	        R# u upi )FN)
r)   r*   rD   r   Ú
ModuleListÚrangeÚnum_hidden_layersr  ÚlayerÚgradient_checkpointing)rC   rD   ré   rE   s   && €rF   r*   ÚConvBertEncoder.__init__®  sR   ø€ Ü‰ÑÔØŒÜ—]’]Ä5È×IaÑIaÔCbÓ#cÑCb¸a¤M°&Ö$9ÑCbÑ#cÓdˆŒ
Ø&+ˆÖ#ùò $ds   ½A&c          
      ó²   <€ V ^8„  d   QhRS[ P                  RS[ P                  R,          RS[ P                  R,          RS[ P                  R,          RS[/# )rJ   rƒ   r¦   Nr§   r!  rM   )r:   r…   rO   r   )rP   rQ   s   "€rF   rR   ÚConvBertEncoder.__annotate__´  s^   ø€ ÷ 
ñ 
á—|‘|ð
ñ ×)Ñ)¨DÕ0ð
ñ  %Ÿ|™|¨dÕ2ð	
ñ
 !&§¡¨tÕ 3ð
ñ 
,ñ
rH   c                ó\   € V P                    F  pV! VV3R VRV/VB pK  	  \        VR7      # )r§   r!  )Úlast_hidden_state)rF  r   )rC   rƒ   r¦   r§   r!  r   Úlayer_modules   &&&&&, rF   r^   ÚConvBertEncoder.forward´  sR   € ð !ŸJœJˆLÙ(ØØñð '<ðð (>ð	ð
 ñŠMñ 'ô 2Ø+ô
ð 	
rH   )rD   rG  rF  r+  rÎ   ri   s   @@rF   rA  rA  ­  s   ù‡ € õ,÷
÷ 
ò 
rH   rA  c                   óD   a a€ ] tR tRt oV 3R ltV3R lR ltRtVtV ;t# )ÚConvBertPredictionHeadTransformiÊ  c                ó‚  <€ \         SV `  4        \        P                  ! VP                  VP                  4      V n        \        VP                  \        4      '       d   \        VP                  ,          V n
        MVP                  V n
        \        P                  ! VP                  VP                  R 7      V n        R# rÒ   )r)   r*   r   r—   r�   rÓ   r  r  r  r   Útransform_act_fnr4   r5   rB   s   &&€rF   r*   Ú(ConvBertPredictionHeadTransform.__init__Ë  s~   ø€ Ü‰ÑÔÜ—Y’Y˜v×1Ñ1°6×3EÑ3EÓFˆŒ
Ü�f×'Ñ'¬×-Ò-Ü$*¨6×+<Ñ+<Õ$=ˆDÕ!à$*×$5Ñ$5ˆDÔ!ÜŸš f×&8Ñ&8¸f×>SÑ>SÔTˆŽrH   c                óN   <€ V ^8„  d   QhRS[ P                  RS[ P                  /# r‚   r„   )rP   rQ   s   "€rF   rR   Ú,ConvBertPredictionHeadTransform.__annotate__Ô  s#   ø€ ÷ ñ ¡U§\¡\ð ±e·l±lñ rH   c                ól   € V P                  V4      pV P                  V4      pV P                  V4      pV# rˆ   )rÓ   rR  r4   r
  s   &&rF   r^   Ú'ConvBertPredictionHeadTransform.forwardÔ  s4   € ØŸ
™
 =Ó1ˆØ×-Ñ-¨mÓ<ˆØŸ™ }Ó5ˆØÐrH   )r4   rÓ   rR  rÎ   ri   s   @@rF   rP  rP  Ê  s   ù‡ € õU÷÷ ð rH   rP  c                   óX   a a€ ] tR tRt oRtV3R lV 3R lltRV3R lR lltRtVtV ;t	# )	ÚConvBertSequenceSummaryiÜ  a„  
Compute a single vector summary of a sequence hidden states.

Args:
    config ([`ConvBertConfig`]):
        The config used by the model. Relevant arguments in the config class of the model are (refer to the actual
        config class of your model for the default values it uses):

        - **summary_type** (`str`) -- The method to use to make this summary. Accepted values are:

            - `"last"` -- Take the last token hidden state (like XLNet)
            - `"first"` -- Take the first token hidden state (like Bert)
            - `"mean"` -- Take the mean of all tokens hidden states
            - `"cls_index"` -- Supply a Tensor of classification token position (GPT/GPT-2)
            - `"attn"` -- Not implemented now, use multi-head attention

        - **summary_use_proj** (`bool`) -- Add a projection after the vector extraction.
        - **summary_proj_to_labels** (`bool`) -- If `True`, the projection outputs to `config.num_labels` classes
          (otherwise to `config.hidden_size`).
        - **summary_activation** (`Optional[str]`) -- Set to `"tanh"` to add a tanh activation to the output,
          another string or `None` will add no activation.
        - **summary_first_dropout** (`float`) -- Optional dropout probability before the projection and activation.
        - **summary_last_dropout** (`float`)-- Optional dropout probability after the projection and activation.
c                ó    <€ V ^8„  d   QhRS[ /# r8  r   )rP   rQ   s   "€rF   rR   Ú$ConvBertSequenceSummary.__annotate__ö  s   ø€ ÷ Hñ H™~ñ HrH   c                óÞ  <€ \         SV `  4        \        VR R4      V n        V P                  R8X  d   \        h\
        P                  ! 4       V n        \        VR4      '       d‡   VP                  '       du   \        VR4      '       d1   VP                  '       d   VP                  ^ 8”  d   VP                  pMVP                  p\
        P                  ! VP                  V4      V n        \        VRR4      pV'       d   \        V4      M\
        P                  ! 4       V n        \
        P                  ! 4       V n        \        VR4      '       d7   VP"                  ^ 8”  d&   \
        P$                  ! VP"                  4      V n        \
        P                  ! 4       V n        \        VR4      '       d;   VP(                  ^ 8”  d(   \
        P$                  ! VP(                  4      V n        R# R# R# )	Úsummary_typeÚlastÚattnÚsummary_use_projÚsummary_proj_to_labelsÚsummary_activationNÚsummary_first_dropoutÚsummary_last_dropout)r)   r*   Úgetattrr]  ÚNotImplementedErrorr   ÚIdentityÚsummaryrX   r`  ra  Ú
num_labelsr�   r—   r	   Ú
activationÚfirst_dropoutrc  r6   Úlast_dropoutrd  )rC   rD   Únum_classesÚactivation_stringrE   s   &&  €rF   r*   Ú ConvBertSequenceSummary.__init__ö  sg  ø€ Ü‰ÑÔä# F¨N¸FÓCˆÔØ×Ñ Ô&ô &Ð%ä—{’{“}ˆŒÜ�6Ð-×.Ò.°6×3J×3JÐ3JÜ�vÐ7×8Ò8¸V×=Z×=ZÐ=ZÐ_e×_pÑ_pÐstÔ_tØ$×/Ñ/‘à$×0Ñ0�ÜŸ9š9 V×%7Ñ%7¸ÓEˆDŒLä# FÐ,@À$ÓGÐßIZ¤NÐ3DÔ$EÔ`b×`kÒ`kÓ`mˆŒäŸ[š[›]ˆÔÜ�6Ð2×3Ò3¸×8TÑ8TÐWXÔ8XÜ!#§¢¨F×,HÑ,HÓ!IˆDÔäŸKšK›MˆÔÜ�6Ð1×2Ò2°v×7RÑ7RÐUVÔ7VÜ "§
¢
¨6×+FÑ+FÓ GˆDÖñ 8WÑ2rH   c                óv   <€ V ^8„  d   QhRS[ P                  RS[ P                  R,          RS[ P                  /# )rJ   rƒ   Ú	cls_indexNrM   )r:   rO   rN   )rP   rQ   s   "€rF   rR   r[    s<   ø€ ÷ )ñ )Ù"×.Ñ.ð)Ù;@×;KÑ;KÈdÕ;Rð)á	×	Ñ	ñ)rH   c                ó8  € V P                   R8X  d   VR	,          pEM9V P                   R8X  d   VR
,          pEMV P                   R8X  d   VP                  ^R7      pMùV P                   R8X  dÒ   VfH   \        P                  ! VR,          VP                  R,          ^,
          \        P
                  R7      pMdVP                  R4      P                  R4      pVP                  RVP                  4       ^,
          ,          VP                  R4      3,           4      pVP                  RV4      P                  R4      pMV P                   R8X  d   \        hV P                  X4      pV P                  V4      pV P                  V4      pV P!                  V4      pV# )a#  
Compute a single vector summary of a sequence hidden states.

Args:
    hidden_states (`torch.FloatTensor` of shape `[batch_size, seq_len, hidden_size]`):
        The hidden states of the last layer.
    cls_index (`torch.LongTensor` of shape `[batch_size]` or `[batch_size, ...]` where ... are optional leading dimensions of `hidden_states`, *optional*):
        Used if `summary_type == "cls_index"` and takes the last token of the sequence as classification token.

Returns:
    `torch.FloatTensor`: The summary of the sequence hidden states.
r^  Úfirstrs   r«   rq  r$   r_  r(   )rU   r(   )rU   rŽ   ).:Nr'   NrU   r¯   )r(   )r]  rs   r:   Ú	full_liker°   r@   r·   r<   r¬   r?   ÚgatherÚsqueezerf  rk  rh  rj  rl  )rC   rƒ   rq  rã   s   &&& rF   r^   ÚConvBertSequenceSummary.forward  s\  € ð ×Ñ Ô&Ø" 5Õ)ŠFØ×Ñ 'Ô)Ø" 4Õ(ŠFØ×Ñ &Ô(Ø"×'Ñ'¨AÐ'Ó.‰FØ×Ñ +Ô-ØÒ Ü!ŸOšOØ! *Õ-Ø!×'Ñ'¨Õ+¨aÕ/ÜŸ*™*ô‘	ð &×/Ñ/°Ó3×=Ñ=¸bÓA�	Ø%×,Ñ,¨U°i·m±m³oÈÕ6IÕ-JÈm×N`ÑN`ÐacÓNdÐMfÕ-fÓg�	à"×)Ñ)¨"¨iÓ8×@Ñ@ÀÓD‰FØ×Ñ &Ô(Ü%Ð%à×#Ñ# FÓ+ˆØ—‘˜fÓ%ˆØ—‘ Ó(ˆØ×"Ñ" 6Ó*ˆàˆrH   )rj  rk  rl  rh  r]  rˆ   r`   ri   s   @@rF   rY  rY  Ü  s%   ù‡ € ñ÷2Hó H÷:)÷ )ò )rH   rY  c                   ór   a a€ ] tR tRt oV 3R ltR tR t]]]	RV3R lR ll4       4       4       t
RtVtV ;t# )	ÚConvBertModeli?  c                ó(  <€ \         SV `  V4       \        V4      V n        VP                  VP
                  8w  d1   \        P                  ! VP                  VP
                  4      V n        \        V4      V n
        Wn        V P                  4        R # rˆ   )r)   r*   r   r]   r-   r�   r   r—   Úembeddings_projectrA  ÚencoderrD   Ú	post_initrB   s   &&€rF   r*   ÚConvBertModel.__init__A  sj   ø€ Ü‰Ñ˜Ô Ü,¨VÓ4ˆŒà× Ñ  F×$6Ñ$6Ô6Ü&(§i¢i°×0EÑ0EÀv×GYÑGYÓ&ZˆDÔ#ä& vÓ.ˆŒØŒà�‰ÖrH   c                ó.   € V P                   P                  # rˆ   ©r]   r/   ©rC   s   &rF   Úget_input_embeddingsÚ"ConvBertModel.get_input_embeddingsM  s   € Ø�‰×.Ñ.Ð.rH   c                ó&   € WP                   n        R # rˆ   r€  )rC   rš   s   &&rF   Úset_input_embeddingsÚ"ConvBertModel.set_input_embeddingsP  s   € Ø*/�‰Ö'rH   c                óþ   <€ V ^8„  d   QhRS[ P                  R,          RS[ P                  R,          RS[ P                  R,          RS[ P                  R,          RS[ P                  R,          RS[S[,          RS[/# )	rJ   rK   Nr¦   rA   r=   rL   r   rM   )r:   rN   rO   r   r   r   )rP   rQ   s   "€rF   rR   ÚConvBertModel.__annotate__V  sŽ   ø€ ÷ 3ñ 3á×#Ñ# dÕ*ð3ñ ×)Ñ)¨DÕ0ð3ñ ×(Ñ(¨4Õ/ð	3ñ
 ×&Ñ&¨Õ-ð3ñ ×(Ñ(¨4Õ/ð3ñ Ñ+Õ,ð3ñ 
,ñ3rH   c                óò  € Ve   Ve   \        R4      hVe#   V P                  W4       VP                  4       pM#Ve   VP                  4       R R pM\        R4      hVw  r‰Ve   VP                  MVP                  p
Vf   \        P
                  ! WzR7      pVfy   \        V P                  R4      '       d6   V P                  P                  RR V	13,          pVP                  W‰4      pTpM'\        P                  ! V\        P                  V
R7      pV P                  WW5R7      p\        V R4      '       d   V P                  V4      p\        V P                  VVR	7      pV P                  ! V3R
V/VB pV# )NzDYou cannot specify both input_ids and inputs_embeds at the same timez5You have to specify either input_ids or inputs_embeds)rW   rA   rU   rV   )rK   r=   rA   rL   r{  )rD   rL   r¦   r¦   r(   )r’   Ú%warn_if_padding_and_no_attention_maskr?   rW   r:   ÚonesrX   r]   rA   r<   r>   r@   r{  r
   rD   r|  )rC   rK   r¦   rA   r=   rL   r   rY   rú   rZ   rW   r[   r\   rƒ   Úencoder_outputss   &&&&&&,        rF   r^   ÚConvBertModel.forwardS  sv  € ð Ò  ]Ò%>ÜÐcÓdÐdØÒ"Ø×6Ñ6°yÔQØ#Ÿ.™.Ó*‰KØÒ&Ø'×,Ñ,Ó.¨s°Ð3‰KäÐTÓUÐUà!,Ñˆ
Ø%.Ò%:�×!Ò!À×@TÑ@TˆàÒ!Ü"ŸZšZ¨ÔCˆNØÒ!Ü�t—‘Ð(8×9Ò9Ø*.¯/©/×*HÑ*HÈÈKÈZÈKÈÕ*XÐ'Ø3J×3QÑ3QÐR\Ó3iÐ0Ø!A‘ä!&§¢¨[ÄÇ
Á
ÐSYÔ!Z�àŸ™ØÈ>ð (ó 
ˆô �4Ð-×.Ò.Ø ×3Ñ3°MÓBˆMä2Ø—;‘;Ø'Ø)ô
ˆð ?C¿lºlØñ?
à)ð?
ð ñ?
ˆð ÐrH   )rD   r]   r{  r|  )NNNNN)ra   rb   rc   rd   r*   r‚  r…  r   r   r   r^   rf   rg   rh   ri   s   @@rF   ry  ry  ?  sB   ù‡ € õ
ò/ò0ð  ØØ÷3ñ 3ó ó ó  ÷3ð 3rH   ry  c                   óH   a a€ ] tR tRt oRtV 3R ltV3R lR ltRtVtV ;t	# )ÚConvBertGeneratorPredictionsiŒ  zAPrediction module for the generator, made up of two dense layers.c                ó  <€ \         SV `  4        \        R 4      V n        \        P
                  ! VP                  VP                  R7      V n        \        P                  ! VP                  VP                  4      V n
        R# )Úgelur!   N)r)   r*   r	   rj  r   r4   r-   r5   r—   r�   rÓ   rB   s   &&€rF   r*   Ú%ConvBertGeneratorPredictions.__init__�  sV   ø€ Ü‰ÑÔä(¨Ó0ˆŒÜŸš f×&;Ñ&;À×AVÑAVÔWˆŒÜ—Y’Y˜v×1Ñ1°6×3HÑ3HÓIˆŽ
rH   c                óN   <€ V ^8„  d   QhRS[ P                  RS[ P                  /# )rJ   Úgenerator_hidden_statesrM   )r:   rO   )rP   rQ   s   "€rF   rR   Ú)ConvBertGeneratorPredictions.__annotate__–  s'   ø€ ÷ ñ ©u×/@Ñ/@ð ÁU×EVÑEVñ rH   c                ól   € V P                  V4      pV P                  V4      pV P                  V4      pV# rˆ   )rÓ   rj  r4   )rC   r”  rƒ   s   && rF   r^   Ú$ConvBertGeneratorPredictions.forward–  s3   € ØŸ
™
Ð#:Ó;ˆØŸ™¨Ó6ˆØŸ™ }Ó5ˆàÐrH   )r4   rj  rÓ   r`   ri   s   @@rF   r�  r�  Œ  s   ù‡ € ÙKõJ÷÷ ð rH   r�  c                   óp   a a€ ] tR tRt oRR/tV 3R ltR tR t]]	R
V3R lR ll4       4       t
R	tVtV ;t# )ÚConvBertForMaskedLMiž  zgenerator_lm_head.weightz*convbert.embeddings.word_embeddings.weightc                óæ   <€ \         SV `  V4       \        V4      V n        \	        V4      V n        \        P                  ! VP                  VP                  4      V n
        V P                  4        R # rˆ   )r)   r*   ry  r.  r�  Úgenerator_predictionsr   r—   r-   r,   Úgenerator_lm_headr}  rB   s   &&€rF   r*   ÚConvBertForMaskedLM.__init__¢  sR   ø€ Ü‰Ñ˜Ô ä% fÓ-ˆŒÜ%AÀ&Ó%IˆÔ"ä!#§¢¨6×+@Ñ+@À&×BSÑBSÓ!TˆÔà�‰ÖrH   c                ó   € V P                   # rˆ   ©rœ  r�  s   &rF   Úget_output_embeddingsÚ)ConvBertForMaskedLM.get_output_embeddings¬  s   € Ø×%Ñ%Ð%rH   c                ó   € Wn         R # rˆ   rŸ  )rC   r/   s   &&rF   Úset_output_embeddingsÚ)ConvBertForMaskedLM.set_output_embeddings¯  s   € Ø!0ÖrH   c                ó6  <€ V ^8„  d   QhRS[ P                  R,          RS[ P                  R,          RS[ P                  R,          RS[ P                  R,          RS[ P                  R,          RS[ P                  R,          RS[S[,          R	S[S[,          /# ©
rJ   rK   Nr¦   rA   r=   rL   Úlabelsr   rM   )r:   rN   rO   r   r   r¨   r   )rP   rQ   s   "€rF   rR   Ú ConvBertForMaskedLM.__annotate__´  s¨   ø€ ÷ (
ñ (
á×#Ñ# dÕ*ð(
ñ ×)Ñ)¨DÕ0ð(
ñ ×(Ñ(¨4Õ/ð	(
ñ
 ×&Ñ&¨Õ-ð(
ñ ×(Ñ(¨4Õ/ð(
ñ × Ñ  4Õ'ð(
ñ Ñ+Õ,ð(
ñ 
‘Õ	ñ(
rH   c                ó‚  € V P                   ! V3RVRVRVRV/VB pV^ ,          p	V P                  V	4      p
V P                  V
4      p
RpVeR   \        P                  ! 4       pV! V
P                  RV P                  P                  4      VP                  R4      4      p\        VV
VP                  VP                  R7      # )az  
labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
    Labels for computing the masked language modeling loss. Indices should be in `[-100, 0, ...,
    config.vocab_size]` (see `input_ids` docstring) Tokens with indices set to `-100` are ignored (masked), the
    loss is only computed for the tokens with labels in `[0, ..., config.vocab_size]`
r¦   rA   r=   rL   N©ÚlossÚlogitsrƒ   r/  r(   )r.  r›  rœ  r   r   r²   rD   r,   r   rƒ   r/  )rC   rK   r¦   rA   r=   rL   r§  r   r”  Úgenerator_sequence_outputÚprediction_scoresr«  Úloss_fcts   &&&&&&&,     rF   r^   ÚConvBertForMaskedLM.forward²  sî   € ð$ GKÇmÂmØñG
à)ðG
ð *ðG
ð &ð	G
ð
 (ðG
ð ñG
Ðð %<¸AÕ$>Ð!à ×6Ñ6Ð7PÓQÐØ ×2Ñ2Ð3DÓEÐàˆàÒÜ×*Ò*Ó,ˆHÙÐ-×2Ñ2°2°t·{±{×7MÑ7MÓNÐPV×P[ÑP[Ð\^ÓP_Ó`ˆDäØØ$Ø1×?Ñ?Ø.×9Ñ9ô	
ð 	
rH   )r.  rœ  r›  ©NNNNNN)ra   rb   rc   rd   Ú_tied_weights_keysr*   r   r£  r   r   r^   rf   rg   rh   ri   s   @@rF   r™  r™  ž  sF   ù‡ € à4Ð6bÐcÐõò&ò1ð Ø÷(
ñ (
ó ó ÷(
ð (
rH   r™  c                   óH   a a€ ] tR tRt oRtV 3R ltV3R lR ltRtVtV ;t	# )ÚConvBertClassificationHeadiß  z-Head for sentence-level classification tasks.c                ór  <€ \         SV `  4        \        P                  ! VP                  VP                  4      V n        VP                  e   VP                  MVP                  p\        P                  ! V4      V n	        \        P                  ! VP                  VP                  4      V n        Wn        R # rˆ   )r)   r*   r   r—   r�   rÓ   Úclassifier_dropoutr7   r6   r8   ri  Úout_projrD   ©rC   rD   r¶  rE   s   && €rF   r*   Ú#ConvBertClassificationHead.__init__â  s„   ø€ Ü‰ÑÔÜ—Y’Y˜v×1Ñ1°6×3EÑ3EÓFˆŒ
à)/×)BÑ)BÒ)NˆF×%Ò%ÐTZ×TnÑTnð 	ô —z’zÐ"4Ó5ˆŒÜŸ	š	 &×"4Ñ"4°f×6GÑ6GÓHˆŒàŽrH   c                óN   <€ V ^8„  d   QhRS[ P                  RS[ P                  /# r‚   r„   )rP   rQ   s   "€rF   rR   Ú'ConvBertClassificationHead.__annotate__í  s#   ø€ ÷ ñ ¡U§\¡\ð ÁÇÁñ rH   c                óî   € VR,          pV P                  V4      pV P                  V4      p\        V P                  P                  ,          ! V4      pV P                  V4      pV P                  V4      pV# )rU   )rU   rŽ   rU   )r8   rÓ   r   rD   r  r·  )rC   rƒ   r   r‰   s   &&, rF   r^   Ú"ConvBertClassificationHead.forwardí  s_   € Ø˜'Õ"ˆØ�L‰L˜‹OˆØ�J‰J�q‹MˆÜ�4—;‘;×)Ñ)Ö*¨1Ó-ˆØ�L‰L˜‹OˆØ�M‰M˜!ÓˆØˆrH   )rD   rÓ   r8   r·  r`   ri   s   @@rF   r´  r´  ß  s   ù‡ € Ù7õ	÷÷ ð rH   r´  z 
    ConvBERT Model transformer with a sequence classification/regression head on top (a linear layer on top of the
    pooled output) e.g. for GLUE tasks.
    )Úcustom_introc                   ó\   a a€ ] tR tRt oV 3R lt]]RV3R lR ll4       4       tRtVt	V ;t
# )Ú!ConvBertForSequenceClassificationi÷  c                ó´   <€ \         SV `  V4       VP                  V n        Wn        \	        V4      V n        \        V4      V n        V P                  4        R # rˆ   )	r)   r*   ri  rD   ry  r.  r´  Ú
classifierr}  rB   s   &&€rF   r*   Ú*ConvBertForSequenceClassification.__init__þ  sF   ø€ Ü‰Ñ˜Ô Ø ×+Ñ+ˆŒØŒÜ% fÓ-ˆŒÜ4°VÓ<ˆŒð 	�‰ÖrH   c                ó6  <€ V ^8„  d   QhRS[ P                  R,          RS[ P                  R,          RS[ P                  R,          RS[ P                  R,          RS[ P                  R,          RS[ P                  R,          RS[S[,          R	S[S[,          /# r¦  )r:   rN   rO   r   r   r¨   r   )rP   rQ   s   "€rF   rR   Ú.ConvBertForSequenceClassification.__annotate__
  s©   ø€ ÷ 8
ñ 8
á×#Ñ# dÕ*ð8
ñ ×)Ñ)¨DÕ0ð8
ñ ×(Ñ(¨4Õ/ð	8
ñ
 ×&Ñ&¨Õ-ð8
ñ ×(Ñ(¨4Õ/ð8
ñ × Ñ  4Õ'ð8
ñ Ñ+Õ,ð8
ñ 
Ñ)Õ	)ñ8
rH   c                óî  € V P                   ! V3RVRVRVRV/VB pV^ ,          p	V P                  V	4      p
RpVEe˜   V P                  P                  f–   V P                  ^8X  d   RV P                  n        MrV P                  ^8”  dQ   VP
                  \        P                  8X  g    VP
                  \        P                  8X  d   RV P                  n        MRV P                  n        V P                  P                  R8X  dK   \        4       pV P                  ^8X  d'   V! V
P                  4       VP                  4       4      pMŽV! W¦4      pM…V P                  P                  R8X  d>   \        4       pV! V
P                  R
V P                  4      VP                  R
4      4      pM-V P                  P                  R8X  d   \        4       pV! W¦4      p\        VV
VP                  VP                   R	7      # )ae  
labels (`torch.LongTensor` of shape `(batch_size,)`, *optional*):
    Labels for computing the sequence classification/regression loss. Indices should be in `[0, ...,
    config.num_labels - 1]`. If `config.num_labels == 1` a regression loss is computed (Mean-Square loss), If
    `config.num_labels > 1` a classification loss is computed (Cross-Entropy).
r¦   rA   r=   rL   NÚ
regressionÚsingle_label_classificationÚmulti_label_classificationrª  r(   )r.  rÂ  rD   Úproblem_typeri  r%   r:   r@   rŸ   r   rv  r   r²   r   r   rƒ   r/  ©rC   rK   r¦   rA   r=   rL   r§  r   ÚoutputsÚsequence_outputr¬  r«  r¯  s   &&&&&&&,     rF   r^   Ú)ConvBertForSequenceClassification.forward  s§  € ð$ 7;·m²mØñ7
à)ð7
ð *ð7
ð &ð	7
ð
 (ð7
ð ñ7
ˆð " !�*ˆØ—‘ Ó1ˆàˆØÓØ�{‰{×'Ñ'Ò/Ø—?‘? aÔ'Ø/;�D—K‘KÕ,Ø—_‘_ qÔ(¨f¯l©l¼e¿j¹jÔ.HÈFÏLÉLÔ\a×\eÑ\eÔLeØ/L�D—K‘KÕ,à/K�D—K‘KÔ,à�{‰{×'Ñ'¨<Ô7Ü"›9�Ø—?‘? aÔ'Ù# F§N¡NÓ$4°f·n±nÓ6FÓG‘Dá# FÓ3‘DØ—‘×)Ñ)Ð-JÔJÜ+Ó-�Ù §¡¨B°·±Ó @À&Ç+Á+ÈbÃ/ÓR‘Ø—‘×)Ñ)Ð-IÔIÜ,Ó.�Ù Ó/�ä'ØØØ!×/Ñ/Ø×)Ñ)ô	
ð 	
rH   )rÂ  rD   r.  ri  r±  ©ra   rb   rc   rd   r*   r   r   r^   rf   rg   rh   ri   s   @@rF   rÀ  rÀ  ÷  s0   ù‡ € õð Ø÷8
ñ 8
ó ó ÷8
ð 8
rH   rÀ  c                   ó\   a a€ ] tR tRt oV 3R lt]]RV3R lR ll4       4       tRtVt	V ;t
# )ÚConvBertForMultipleChoiceiE  c                óÒ   <€ \         SV `  V4       \        V4      V n        \	        V4      V n        \        P                  ! VP                  ^4      V n	        V P                  4        R# )r'   N)r)   r*   ry  r.  rY  Úsequence_summaryr   r—   r�   rÂ  r}  rB   s   &&€rF   r*   Ú"ConvBertForMultipleChoice.__init__G  sM   ø€ Ü‰Ñ˜Ô ä% fÓ-ˆŒÜ 7¸Ó ?ˆÔÜŸ)š) F×$6Ñ$6¸Ó:ˆŒð 	�‰ÖrH   c                ó6  <€ V ^8„  d   QhRS[ P                  R,          RS[ P                  R,          RS[ P                  R,          RS[ P                  R,          RS[ P                  R,          RS[ P                  R,          RS[S[,          R	S[S[,          /# r¦  )r:   rN   rO   r   r   r¨   r   )rP   rQ   s   "€rF   rR   Ú&ConvBertForMultipleChoice.__annotate__S  s³   ø€ ÷ N
ñ N
á×#Ñ# dÕ*ðN
ñ ×)Ñ)¨DÕ0ðN
ñ ×(Ñ(¨4Õ/ð	N
ñ
 ×&Ñ&¨Õ-ðN
ñ ×(Ñ(¨4Õ/ðN
ñ × Ñ  4Õ'ðN
ñ Ñ+Õ,ðN
ñ 
Ñ*Õ	*ñN
rH   c                ó"  € Ve   VP                   ^,          MVP                   ^,          pVe"   VP                  RVP                  R4      4      MRpVe"   VP                  RVP                  R4      4      MRpVe"   VP                  RVP                  R4      4      MRpVe"   VP                  RVP                  R4      4      MRpVe2   VP                  RVP                  R4      VP                  R4      4      MRpV P                  ! V3RVRVRVRV/VB p	V	^ ,          p
V P	                  V
4      pV P                  V4      pVP                  RV4      pRpVe   \        4       pV! WÖ4      p\        VVV	P                  V	P                  R7      # )	aœ  
input_ids (`torch.LongTensor` of shape `(batch_size, num_choices, sequence_length)`):
    Indices of input sequence tokens in the vocabulary.

    Indices can be obtained using [`AutoTokenizer`]. See [`PreTrainedTokenizer.encode`] and
    [`PreTrainedTokenizer.__call__`] for details.

    [What are input IDs?](../glossary#input-ids)
token_type_ids (`torch.LongTensor` of shape `(batch_size, num_choices, sequence_length)`, *optional*):
    Segment token indices to indicate first and second portions of the inputs. Indices are selected in `[0,
    1]`:


    - 0 corresponds to a *sentence A* token,
    - 1 corresponds to a *sentence B* token.

    [What are token type IDs?](../glossary#token-type-ids)
position_ids (`torch.LongTensor` of shape `(batch_size, num_choices, sequence_length)`, *optional*):
    Indices of positions of each input sequence tokens in the position embeddings. Selected in the range `[0,
    config.max_position_embeddings - 1]`.

    [What are position IDs?](../glossary#position-ids)
inputs_embeds (`torch.FloatTensor` of shape `(batch_size, num_choices, sequence_length, hidden_size)`, *optional*):
    Optionally, instead of passing `input_ids` you can choose to directly pass an embedded representation. This
    is useful if you want more control over how to convert *input_ids* indices into associated vectors than the
    model's internal embedding lookup matrix.
labels (`torch.LongTensor` of shape `(batch_size,)`, *optional*):
    Labels for computing the multiple choice classification loss. Indices should be in `[0, ...,
    num_choices-1]` where `num_choices` is the size of the second dimension of the input tensors. (See
    `input_ids` above)
Nr¦   rA   r=   rL   rª  r(   r¯   )
r°   r²   r?   r.  rÓ  rÂ  r   r   rƒ   r/  )rC   rK   r¦   rA   r=   rL   r§  r   Únum_choicesrÌ  rÍ  Úpooled_outputr¬  Úreshaped_logitsr«  r¯  s   &&&&&&&,        rF   r^   Ú!ConvBertForMultipleChoice.forwardQ  s®  € ðV -6Ò,A�i—o‘o aÖ(À}×GZÑGZÐ[\ÕG]ˆà>GÒ>S�I—N‘N 2 y§~¡~°bÓ'9Ô:ÐY]ˆ	ØM[ÒMg˜×,Ñ,¨R°×1DÑ1DÀRÓ1HÔIÐmqˆØM[ÒMg˜×,Ñ,¨R°×1DÑ1DÀRÓ1HÔIÐmqˆØGSÒG_�|×(Ñ(¨¨\×->Ñ->¸rÓ-BÔCÐeiˆð Ò(ð ×Ñ˜r =×#5Ñ#5°bÓ#9¸=×;MÑ;MÈbÓ;QÔRàð 	ð 7;·m²mØñ7
à)ð7
ð *ð7
ð &ð	7
ð
 (ð7
ð ñ7
ˆð " !�*ˆà×-Ñ-¨oÓ>ˆØ—‘ Ó/ˆØ Ÿ+™+ b¨+Ó6ˆàˆØÒÜ'Ó)ˆHÙ˜OÓ4ˆDä(ØØ"Ø!×/Ñ/Ø×)Ñ)ô	
ð 	
rH   )rÂ  r.  rÓ  r±  rÏ  ri   s   @@rF   rÑ  rÑ  E  s4   ù‡ € õð Ø÷N
ñ N
ó ó ÷N
ð N
rH   rÑ  c                   ó\   a a€ ] tR tRt oV 3R lt]]RV3R lR ll4       4       tRtVt	V ;t
# )ÚConvBertForTokenClassificationi¤  c                ój  <€ \         SV `  V4       VP                  V n        \        V4      V n        VP
                  e   VP
                  MVP                  p\        P                  ! V4      V n	        \        P                  ! VP                  VP                  4      V n        V P                  4        R # rˆ   )r)   r*   ri  ry  r.  r¶  r7   r   r6   r8   r—   r�   rÂ  r}  r¸  s   && €rF   r*   Ú'ConvBertForTokenClassification.__init__¦  sˆ   ø€ Ü‰Ñ˜Ô Ø ×+Ñ+ˆŒä% fÓ-ˆŒà)/×)BÑ)BÒ)NˆF×%Ò%ÐTZ×TnÑTnð 	ô —z’zÐ"4Ó5ˆŒÜŸ)š) F×$6Ñ$6¸×8IÑ8IÓJˆŒð 	�‰ÖrH   c                ó6  <€ V ^8„  d   QhRS[ P                  R,          RS[ P                  R,          RS[ P                  R,          RS[ P                  R,          RS[ P                  R,          RS[ P                  R,          RS[S[,          R	S[S[,          /# r¦  )r:   rN   rO   r   r   r¨   r   )rP   rQ   s   "€rF   rR   Ú+ConvBertForTokenClassification.__annotate__¶  s©   ø€ ÷ &
ñ &
á×#Ñ# dÕ*ð&
ñ ×)Ñ)¨DÕ0ð&
ñ ×(Ñ(¨4Õ/ð	&
ñ
 ×&Ñ&¨Õ-ð&
ñ ×(Ñ(¨4Õ/ð&
ñ × Ñ  4Õ'ð&
ñ Ñ+Õ,ð&
ñ 
Ñ&Õ	&ñ&
rH   c                óX  € V P                   ! V3RVRVRVRV/VB pV^ ,          p	V P                  V	4      p	V P                  V	4      p
RpVe=   \        4       pV! V
P	                  RV P
                  4      VP	                  R4      4      p\        VV
VP                  VP                  R7      # )zÃ
labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
    Labels for computing the token classification loss. Indices should be in `[0, ..., config.num_labels - 1]`.
r¦   rA   r=   rL   Nrª  r(   )	r.  r8   rÂ  r   r²   ri  r   rƒ   r/  rË  s   &&&&&&&,     rF   r^   Ú&ConvBertForTokenClassification.forward´  sÇ   € ð  7;·m²mØñ7
à)ð7
ð *ð7
ð &ð	7
ð
 (ð7
ð ñ7
ˆð " !�*ˆàŸ,™, Ó7ˆØ—‘ Ó1ˆàˆØÒÜ'Ó)ˆHÙ˜FŸK™K¨¨D¯O©OÓ<¸f¿k¹kÈ"»oÓNˆDä$ØØØ!×/Ñ/Ø×)Ñ)ô	
ð 	
rH   )rÂ  r.  r8   ri  r±  rÏ  ri   s   @@rF   rÝ  rÝ  ¤  s0   ù‡ € õð Ø÷&
ñ &
ó ó ÷&
ð &
rH   rÝ  c                   ó\   a a€ ] tR tRt oV 3R lt]]RV3R lR ll4       4       tRtVt	V ;t
# )ÚConvBertForQuestionAnsweringiß  c                óè   <€ \         SV `  V4       VP                  V n        \        V4      V n        \
        P                  ! VP                  VP                  4      V n        V P                  4        R # rˆ   )
r)   r*   ri  ry  r.  r   r—   r�   Ú
qa_outputsr}  rB   s   &&€rF   r*   Ú%ConvBertForQuestionAnswering.__init__á  sS   ø€ Ü‰Ñ˜Ô à ×+Ñ+ˆŒÜ% fÓ-ˆŒÜŸ)š) F×$6Ñ$6¸×8IÑ8IÓJˆŒð 	�‰ÖrH   c                óN  <€ V ^8„  d   QhRS[ P                  R,          RS[ P                  R,          RS[ P                  R,          RS[ P                  R,          RS[ P                  R,          RS[ P                  R,          RS[ P                  R,          R	S[S[,          R
S[/	# )rJ   rK   Nr¦   rA   r=   rL   Ústart_positionsÚend_positionsr   rM   )r:   rN   rO   r   r   r   )rP   rQ   s   "€rF   rR   Ú)ConvBertForQuestionAnswering.__annotate__í  s¸   ø€ ÷ 2
ñ 2
á×#Ñ# dÕ*ð2
ñ ×)Ñ)¨DÕ0ð2
ñ ×(Ñ(¨4Õ/ð	2
ñ
 ×&Ñ&¨Õ-ð2
ñ ×(Ñ(¨4Õ/ð2
ñ ×)Ñ)¨DÕ0ð2
ñ ×'Ñ'¨$Õ.ð2
ñ Ñ+Õ,ð2
ñ 
&ñ2
rH   c                óì  € V P                   ! V3R VRVRVRV/VB p	V	^ ,          p
V P                  V
4      pVP                  ^RR7      w  rÍVP                  R4      P	                  4       pVP                  R4      P	                  4       pRpVeÄ   VeÀ   \        VP                  4       4      ^8”  d   VP                  R4      p\        VP                  4       4      ^8”  d   VP                  R4      pVP                  ^4      pVP                  ^ V4      pVP                  ^ V4      p\        VR7      pV! WÆ4      pV! W×4      pVV,           ^,          p\        VVVV	P                  V	P                  R7      # )	r¦   rA   r=   rL   r«   N)Úignore_index)r«  Ústart_logitsÚ
end_logitsrƒ   r/  r(   )r.  rç  Úsplitrv  r¶   Úlenr?   Úclampr   r   rƒ   r/  )rC   rK   r¦   rA   r=   rL   rê  rë  r   rÌ  rÍ  r¬  rï  rð  Ú
total_lossÚignored_indexr¯  Ú
start_lossÚend_losss   &&&&&&&&,          rF   r^   Ú$ConvBertForQuestionAnswering.forwardë  sŠ  € ð 7;·m²mØñ7
à)ð7
ð *ð7
ð &ð	7
ð
 (ð7
ð ñ7
ˆð " !�*ˆà—‘ Ó1ˆØ#)§<¡<°°r <Ó#:Ñ ˆØ#×+Ñ+¨BÓ/×:Ñ:Ó<ˆØ×'Ñ'¨Ó+×6Ñ6Ó8ˆ
àˆ
ØÒ&¨=Ò+Dä�?×'Ñ'Ó)Ó*¨QÔ.Ø"1×"9Ñ"9¸"Ó"=�Ü�=×%Ñ%Ó'Ó(¨1Ô,Ø -× 5Ñ 5°bÓ 9�à(×-Ñ-¨aÓ0ˆMØ-×3Ñ3°A°}ÓEˆOØ)×/Ñ/°°=ÓAˆMä'°]ÔCˆHÙ! ,Ó@ˆJÙ 
Ó:ˆHØ$ xÕ/°1Õ4ˆJä+ØØ%Ø!Ø!×/Ñ/Ø×)Ñ)ô
ð 	
rH   )r.  ri  rç  )NNNNNNNrÏ  ri   s   @@rF   rå  rå  ß  s0   ù‡ € õð Ø÷2
ñ 2
ó ó ÷2
ð 2
rH   rå  )r™  rÑ  rå  rÀ  rÝ  r  ry  r-  )Fre   rº   Úcollections.abcr   r:   r   Útorch.nnr   r   r   Ú r   r2  Úactivationsr   r	   Úmasking_utilsr
   Úmodeling_layersr   Úmodeling_outputsr   r   r   r   r   r   Úmodeling_utilsr   Úprocessing_utilsr   Úpytorch_utilsr   Úutilsr   r   r   r   Úutils.genericr   Úutils.output_capturingr   Úconfiguration_convbertr   Ú
get_loggerra   ÚloggerÚModuler   rk   rŒ   rÐ   rá   rí   rý   r  r  r-  rA  rP  rY  ry  r�  r™  r´  rÀ  rÑ  rÝ  rå  Ú__all__r:  rH   rF   Ú<module>r     s,  ðñ ã Ý $ã Ý ß AÑ Aå &ß 1Ý 6Ý 9÷÷ õ .Ý &Ý 6÷ó õ 8Ý 5Ý 2ð 
×	Ò	˜HÓ	%€ô3˜Ÿ™ô 3ôl�b—i‘iô ô4w.˜BŸI™Iô w.ôt˜Ÿ™ô ô ˜Ÿ	™	ô  ô.˜Ÿ™ô ô,˜2Ÿ9™9ô ô(�R—Y‘Yô ô&2Ð.ô 2ðj ô/˜oó /ó ð/ô.
�b—i‘iô 
ô: b§i¡iô ô$`˜bŸi™iô `ðF ôIÐ+ó Ió ðIôX 2§9¡9ô ð$ ô=
Ð1ó =
ó ð=
ô@ §¡ô ñ0 ðôôE
Ð(?ó E
óðE
ðP ô[
Ð 7ó [
ó ð[
ð| ô7
Ð%<ó 7
ó ð7
ðt ô?
Ð#:ó ?
ó ð?
òD	‚rH   