+
    œ°³j®[  ã                   óÀ  € ^ RI Ht ^ RIt^ RIHt ^RIHt ^RIH	t	 ^RI
HtHt ^RIHt ^RIHtHt ^RIHtHt ^R	IHt ^R
IHt ^RIHtHt ^RIHtHt ^RIH t H!t! ^RI"H#t# ^RI$H%t%H&t&H't' ^RI(H)t) ^RI*H+t+H,t, ^RI-H.t. ^RI/H0t0  ! R R]Pb                  4      t2 ! R R]Pb                  4      t3R t4]! R4      R-R l4       t5R R lt6R.R R llt7]! ]54       ! R R ]Pb                  4      4       t8 ! R! R"]4      t9 ! R# R$]Pb                  4      t: ! R% R&]Pv                  4      t<]& ! R' R(]!4      4       t=]& ! R) R*]=4      4       t>]& ! R+ R,]=]4      4       t?. R/Ot@R# )0é    )ÚCallableN)Úinitialization)ÚACT2FN)ÚCacheÚDynamicCache)ÚGenerationMixin)Úuse_kernel_forward_from_hubÚuse_kernelized_func)Úcreate_causal_maskÚ!create_sliding_window_causal_mask)ÚFlashAttentionKwargs)ÚGradientCheckpointingLayer)ÚBaseModelOutputWithPastÚCausalLMOutputWithPast)ÚROPE_INIT_FUNCTIONSÚdynamic_rope_update)ÚALL_ATTENTION_FUNCTIONSÚPreTrainedModel)ÚUnpack)ÚTransformersKwargsÚauto_docstringÚcan_return_tuple)Údeprecate_kwarg)Úmaybe_autocastÚmerge_with_config_defaults)Úcapture_outputs©ÚVaultGemmaConfigc                   óT   a a€ ] tR t^.t oRV3R lV 3R llltR tR tR tRtVt	V ;t
# )ÚVaultGemmaRMSNormc                ó&   <€ V ^8„  d   QhRS[ RS[/# )é   ÚdimÚeps©ÚintÚfloat)ÚformatÚ__classdict__s   "€Ús/home/luca/screen-checking/.venv/lib/python3.14/site-packages/transformers/models/vaultgemma/modeling_vaultgemma.pyÚ__annotate__ÚVaultGemmaRMSNorm.__annotate__/   s   ø€ ÷ 5ñ 5™Cð 5¡eñ 5ó    c                óŽ   <€ \         SV `  4        W n        \        P                  ! \
        P                  ! V4      4      V n        R # ©N)ÚsuperÚ__init__r$   ÚnnÚ	ParameterÚtorchÚzerosÚweight)Úselfr#   r$   Ú	__class__s   &&&€r*   r1   ÚVaultGemmaRMSNorm.__init__/   s,   ø€ Ü‰ÑÔØŒÜ—l’l¤5§;¢;¨sÓ#3Ó4ˆŽr-   c                óž   € V\         P                  ! VP                  ^4      P                  RRR7      V P                  ,           4      ,          # )r"   T)Úkeepdiméÿÿÿÿ)r4   ÚrsqrtÚpowÚmeanr$   )r7   Úxs   &&r*   Ú_normÚVaultGemmaRMSNorm._norm4   s4   € Ø”5—;’;˜qŸu™u Q›xŸ}™}¨R¸˜}Ó>ÀÇÁÕIÓJÕJÐJr-   c                ó²   € V P                  VP                  4       4      pVR V P                  P                  4       ,           ,          pVP                  V4      # ©ç      ð?)rA   r'   r6   Útype_as)r7   r@   Úoutputs   && r*   ÚforwardÚVaultGemmaRMSNorm.forward7   sC   € Ø—‘˜AŸG™G›IÓ&ˆð ˜3 §¡×!2Ñ!2Ó!4Õ4Õ5ˆØ�~‰~˜aÓ Ð r-   c                ó^   € \        V P                  P                  4       R V P                   2# )z, eps=)Útupler6   Úshaper$   )r7   s   &r*   Ú
extra_reprÚVaultGemmaRMSNorm.extra_repr>   s'   € Ü˜Ÿ™×)Ñ)Ó*Ð+¨6°$·(±(°Ð<Ð<r-   )r$   r6   )g�íµ ÷Æ°>)Ú__name__Ú
__module__Ú__qualname__Ú__firstlineno__r1   rA   rH   rM   Ú__static_attributes__Ú__classdictcell__Ú__classcell__©r8   r)   s   @@r*   r    r    .   s$   ù‡ € ÷5õ 5ò
Kò!÷=ò =r-   r    c                   ó8   a a€ ] tR t^Bt oV 3R ltR tRtVtV ;t# )ÚVaultGemmaMLPc                óØ  <€ \         SV `  4        Wn        VP                  V n        VP                  V n        \
        P                  ! V P                  V P                  R R7      V n        \
        P                  ! V P                  V P                  R R7      V n        \
        P                  ! V P                  V P                  R R7      V n	        \        VP                  ,          V n        R# ©F©ÚbiasN)r0   r1   ÚconfigÚhidden_sizeÚintermediate_sizer2   ÚLinearÚ	gate_projÚup_projÚ	down_projr   Úhidden_activationÚact_fn©r7   r]   r8   s   &&€r*   r1   ÚVaultGemmaMLP.__init__C   s¤   ø€ Ü‰ÑÔØŒØ!×-Ñ-ˆÔØ!'×!9Ñ!9ˆÔÜŸš 4×#3Ñ#3°T×5KÑ5KÐRWÔXˆŒÜ—y’y ×!1Ñ!1°4×3IÑ3IÐPUÔVˆŒÜŸš 4×#9Ñ#9¸4×;KÑ;KÐRWÔXˆŒÜ˜V×5Ñ5Õ6ˆŽr-   c                ó�   € V P                  V P                  V P                  V4      4      V P                  V4      ,          4      pV# r/   )rc   re   ra   rb   )r7   r@   rc   s   && r*   rH   ÚVaultGemmaMLP.forwardM   s6   € Ø—N‘N 4§;¡;¨t¯~©~¸aÓ/@Ó#AÀDÇLÁLÐQRÃOÕ#SÓTˆ	ØÐr-   )re   r]   rc   ra   r^   r_   rb   ©	rO   rP   rQ   rR   r1   rH   rS   rT   rU   rV   s   @@r*   rX   rX   B   s   ù‡ € õ7÷ò r-   rX   c                óÌ   € V RRV P                   R,          ^,          13,          pV RV P                   R,          ^,          R13,          p\        P                  ! V) V3RR7      # )z*Rotates half the hidden dims of the input..N©r#   r<   )rL   r4   Úcat)r@   Úx1Úx2s   &  r*   Úrotate_halfrp   R   sZ   € à	
ˆ3Ð"�!—'‘'˜"•+ Õ"Ð"Ð"Õ	#€BØ	
ˆ3�—‘˜•˜qÕ Ñ"Ð"Õ	#€BÜ�9Š9�r�c˜2�Y BÔ'Ð'r-   Úrotary_pos_embc                óÈ   € VP                  V4      pVP                  V4      pW,          \        V 4      V,          ,           pW,          \        V4      V,          ,           pWV3# )aI  Applies Rotary Position Embedding to the query and key tensors.

Args:
    q (`torch.Tensor`): The query tensor.
    k (`torch.Tensor`): The key tensor.
    cos (`torch.Tensor`): The cosine part of the rotary embedding.
    sin (`torch.Tensor`): The sine part of the rotary embedding.
    unsqueeze_dim (`int`, *optional*, defaults to 1):
        The 'unsqueeze_dim' argument specifies the dimension along which to unsqueeze cos[position_ids] and
        sin[position_ids] so that they can be properly broadcasted to the dimensions of q and k. For example, note
        that cos[position_ids] and sin[position_ids] have the shape [batch_size, seq_len, head_dim]. Then, if q and
        k have the shape [batch_size, heads, seq_len, head_dim], then setting unsqueeze_dim=1 makes
        cos[position_ids] and sin[position_ids] broadcastable to the shapes of q and k. Similarly, if q and k have
        the shape [batch_size, seq_len, heads, head_dim], then set unsqueeze_dim=2.
Returns:
    `tuple(torch.Tensor)` comprising of the query and key tensors rotated using the Rotary Position Embedding.
)Ú	unsqueezerp   )ÚqÚkÚcosÚsinÚunsqueeze_dimÚq_embedÚk_embeds   &&&&&  r*   Úapply_rotary_pos_embr{   Y   sS   € ð& �-‰-˜Ó
&€CØ
�-‰-˜Ó
&€CØ�wœ; q›>¨CÕ/Õ0€GØ�wœ; q›>¨CÕ/Õ0€GØÐÐr-   c                ód   € V ^8„  d   QhR\         P                  R\        R\         P                  /# )r"   Úhidden_statesÚn_repÚreturn)r4   ÚTensorr&   )r(   s   "r*   r+   r+   s   s.   € ÷ 	Uñ 	UœUŸ\™\ð 	U´#ð 	U¼%¿,¹,ñ 	Ur-   c                ó˜   € V P                   w  r#rEV^8X  d   V # V R,          P                  W#WV4      p V P                  W#V,          WE4      # )zÈ
This is the equivalent of torch.repeat_interleave(x, dim=1, repeats=n_rep). The hidden states go from (batch,
num_key_value_heads, seqlen, head_dim) to (batch, num_attention_heads, seqlen, head_dim)
)ºNNNr‚   Nr‚   r‚   )rL   ÚexpandÚreshape)r}   r~   ÚbatchÚnum_key_value_headsÚslenÚhead_dims   &&    r*   Ú	repeat_kvr‰   s   sU   € ð
 2?×1DÑ1DÑ.€E Ø�„zØÐØ!Ð"2Õ3×:Ñ:¸5ÐW\ÐdlÓm€MØ× Ñ  ¸eÕ(CÀTÓTÐTr-   c                ór  € V ^8„  d   QhR\         P                  R\        P                  R\        P                  R\        P                  R\        P                  R,          R\        \
        ,          R\        R,          R	\        R,          R
\        \        P                  \        P                  3,          /	# )r"   ÚmoduleÚqueryÚkeyÚvalueÚattention_maskNÚdropoutÚscalingÚsoftcapr   )r2   ÚModuler4   r€   r'   r&   rK   )r(   s   "r*   r+   r+      s™   € ÷ %ñ %Ü�I‰Ið%ä�<‰<ð%ô 
�‰ð%ô �<‰<ð	%ô
 —L‘L 4Õ'ð%ô ”S�[ð%ô �T�\ð%ô �T�\ð%ô Œ5�<‰<œŸ™Ð%Õ&ñ%r-   c                 ó¤  € Vf   V P                   R,          p\        W P                  4      p	\        W0P                  4      p
\        P                  ! WP                  ^^4      4      V,          pVe'   W·,          p\        P                  ! V4      pW·,          pVe	   W´,           p\        P                  P                  VR\        P                  R7      P                  VP                  4      p\        P                  P                  WµV P                  R7      p\        P                  ! Wº4      pVP                  ^^4      P                  4       pWË3# )N)r#   Údtype)ÚpÚtrainingç      à¿r<   )rˆ   r‰   Únum_key_value_groupsr4   ÚmatmulÚ	transposeÚtanhr2   Ú
functionalÚsoftmaxÚfloat32Útor•   r�   r—   Ú
contiguous)r‹   rŒ   r�   rŽ   r�   r�   r‘   r’   ÚkwargsÚ
key_statesÚvalue_statesÚattn_weightsÚattn_outputs   &&&&&&&&,    r*   Úeager_attention_forwardr§      s  € ð ‚Ø—/‘/ 4Õ'ˆä˜3× ;Ñ ;Ó<€JÜ˜U×$?Ñ$?Ó@€Lä—<’< ×';Ñ';¸A¸qÓ'AÓBÀWÕL€LàÒØ#Õ-ˆÜ—z’z ,Ó/ˆØ#Õ-ˆØÒ!Ø#Õ4ˆô —=‘=×(Ñ(¨¸2ÄUÇ]Á]Ð(ÓS×VÑVÐW\×WbÑWbÓc€LÜ—=‘=×(Ñ(¨È6Ï?É?Ð(Ó[€LÜ—,’,˜|Ó:€KØ×'Ñ'¨¨1Ó-×8Ñ8Ó:€KØÐ$Ð$r-   c                   óX   a a€ ] tR t^¡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VaultGemmaAttentionz=Multi-headed attention from 'Attention Is All You Need' paperc                ó&   <€ V ^8„  d   QhRS[ RS[/# ©r"   r]   Ú	layer_idx©r   r&   )r(   r)   s   "€r*   r+   Ú VaultGemmaAttention.__annotate__¥   s    ø€ ÷ hñ hÑ/ð h¹Cñ hr-   c                óš  <€ \         SV `  4        \        VR 4      '       d   VP                  V,          MRV n        Wn        W n        \        VRVP                  VP                  ,          4      V n
        VP                  VP                  ,          V n        VP                  R,          V n        V P
                  P                  V n        RV n        \"        P$                  ! VP                  VP                  V P                  ,          VP&                  R7      V n        \"        P$                  ! VP                  VP                  V P                  ,          VP&                  R7      V n        \"        P$                  ! VP                  VP                  V P                  ,          VP&                  R7      V n        \"        P$                  ! VP                  V P                  ,          VP                  VP&                  R7      V n        V P
                  P0                  V n        V P                  R8X  d   VP2                  V n        R# RV n        R# )Úlayer_typesNrˆ   Tr[   Úsliding_attentionr˜   )r0   r1   Úhasattrr°   Ú
layer_typer]   r¬   Úgetattrr^   Únum_attention_headsrˆ   r†   r™   Úquery_pre_attn_scalarr‘   Úattention_dropoutÚ	is_causalr2   r`   Úattention_biasÚq_projÚk_projÚv_projÚo_projÚattn_logit_softcappingÚsliding_window©r7   r]   r¬   r8   s   &&&€r*   r1   ÚVaultGemmaAttention.__init__¥   s¦  ø€ Ü‰ÑÔÜ;BÀ6È=×;YÒ;Y˜&×,Ñ,¨YÖ7Ð_cˆŒØŒØ"ŒÜ ¨
°F×4FÑ4FÈ&×JdÑJdÕ4dÓeˆŒØ$*×$>Ñ$>À&×B\ÑB\Õ$\ˆÔ!Ø×3Ñ3°TÕ9ˆŒØ!%§¡×!>Ñ!>ˆÔØˆŒä—i’iØ×Ñ × :Ñ :¸T¿]¹]Õ JÐQW×QfÑQfô
ˆŒô —i’iØ×Ñ × :Ñ :¸T¿]¹]Õ JÐQW×QfÑQfô
ˆŒô —i’iØ×Ñ × :Ñ :¸T¿]¹]Õ JÐQW×QfÑQfô
ˆŒô —i’iØ×&Ñ&¨¯©Õ6¸×8JÑ8JÐQW×QfÑQfô
ˆŒð '+§k¡k×&HÑ&HˆÔ#Ø7;·±ÐJ]Ô7]˜f×3Ñ3ˆÖÐcgˆÖr-   c                ó`  <€ V ^8„  d   QhRS[ P                  RS[S[ P                  S[ P                  3,          R,          RS[ P                  R,          RS[R,          RS[S[,          RS[S[ P                  S[ P                  R,          S[S[ P                  ,          R,          3,          /# )r"   r}   Úposition_embeddingsNr�   Úpast_key_valuesr¢   r   )r4   r€   rK   r   r   r   )r(   r)   s   "€r*   r+   r®   ¿   sœ   ø€ ÷ ()ñ ()á—|‘|ð()ñ #¡5§<¡<±·±Ð#=Õ>ÀÕEð()ñ Ÿ™ tÕ+ð	()ñ
  �ð()ñ Ñ-Õ.ð()ñ 
‰u�|‰|™UŸ\™\¨DÕ0±%¹¿¹Õ2EÈÕ2LÐLÕ	Mñ()r-   c                ó@  € VP                   R R p. VORNV P                  N5pV P                  V4      P                  V4      P	                  ^^4      pV P                  V4      P                  V4      P	                  ^^4      p	V P                  V4      P                  V4      P	                  ^^4      p
Vw  r¼\        W‰W¼4      w  r‰Ve   VP                  WšV P                  4      w  rš\        P                  ! V P                  P                  \        4      pT! V VV	V
V3RV P                  '       d   V P                   MRRV P"                  RV P$                  RV P&                  /VB w  rïVP(                  ! . VORN5!  P+                  4       pV P-                  V4      pWï3# )Nr�   ç        r‘   r¿   r’   r<   )rL   rˆ   rº   Úviewr›   r»   r¼   r{   Úupdater¬   r   Úget_interfacer]   Ú_attn_implementationr§   r—   r·   r‘   r¿   r¾   r„   r¡   r½   )r7   r}   rÃ   r�   rÄ   r¢   Úinput_shapeÚhidden_shapeÚquery_statesr£   r¤   rv   rw   Úattention_interfacer¦   r¥   s   &&&&&,          r*   rH   ÚVaultGemmaAttention.forward¿   s§  € ð $×)Ñ)¨#¨2Ð.ˆØ8˜Ð8 bÐ8¨$¯-©-Ñ8ˆà—{‘{ =Ó1×6Ñ6°|ÓD×NÑNÈqÐRSÓTˆØ—[‘[ Ó/×4Ñ4°\ÓB×LÑLÈQÐPQÓRˆ
Ø—{‘{ =Ó1×6Ñ6°|ÓD×NÑNÈqÐRSÓTˆà&‰ˆÜ#7¸ÐRUÓ#[Ñ ˆàÒ&Ø'6×'=Ñ'=¸jÐX\×XfÑXfÓ'gÑ$ˆJä(?×(MÒ(MØ�K‰K×,Ñ,Ô.Eó)
Ðñ %8ØØØØØñ%
ð /3¯m¯m¨m�D×*Ò*Àð%
ð —L‘Lð%
ð  ×.Ñ.ð%
ð ×/Ñ/ð%
ð ñ%
Ñ!ˆð "×)Ò)Ð;¨;Ð;¸Ó;×FÑFÓHˆØ—k‘k +Ó.ˆØÐ(Ð(r-   )r·   r¾   r]   rˆ   r¸   r»   r¬   r³   r™   r½   rº   r‘   r¿   r¼   ©NNN©
rO   rP   rQ   rR   Ú__doc__r1   rH   rS   rT   rU   rV   s   @@r*   r©   r©   ¡   s#   ù‡ € áG÷hó h÷4()÷ ()ò ()r-   r©   c                   óT   a a€ ] tR t^êt oV3R lV 3R lltRV3R lR lltRtVtV ;t# )ÚVaultGemmaDecoderLayerc                ó&   <€ V ^8„  d   QhRS[ RS[/# r«   r­   )r(   r)   s   "€r*   r+   Ú#VaultGemmaDecoderLayer.__annotate__ë   s    ø€ ÷ hñ hÑ/ð h¹Cñ hr-   c                ó,  <€ \         SV `  4        VP                  V n        Wn        \	        WR 7      V n        \        V4      V n        \        VP                  VP                  R7      V n
        \        VP                  VP                  R7      V n        R# ))r]   r¬   ©r$   N)r0   r1   r^   r]   r©   Ú	self_attnrX   Úmlpr    Úrms_norm_epsÚinput_layernormÚpre_feedforward_layernormrÀ   s   &&&€r*   r1   ÚVaultGemmaDecoderLayer.__init__ë   sp   ø€ Ü‰ÑÔØ!×-Ñ-ˆÔØŒÜ,°FÔPˆŒÜ  Ó(ˆŒÜ0°×1CÑ1CÈ×I\ÑI\Ô]ˆÔä):¸6×;MÑ;MÐSY×SfÑSfÔ)gˆÖ&r-   c                óX  <€ V ^8„  d   QhRS[ P                  RS[S[ P                  S[ P                  3,          RS[ P                  R,          RS[ P                  R,          RS[R,          RS[S[ P
                  S[S[ P
                  S[ P
                  3,          R,          3,          /# )r"   r}   rÃ   r�   NÚposition_idsrÄ   r   )r4   r€   rK   Ú
LongTensorr   ÚFloatTensor)r(   r)   s   "€r*   r+   rÖ   õ   s¡   ø€ ÷ ñ á—|‘|ðñ #¡5§<¡<±·±Ð#=Õ>ðñ Ÿ™ tÕ+ð	ñ
 ×&Ñ&¨Õ-ðñ  �ðñ 
‰u× Ñ ¡%©×(9Ñ(9¹5×;LÑ;LÐ(LÕ"MÐPTÕ"TÐTÕ	Uñr-   c                óÐ   € TpV P                  V4      pV P                  ! RR VRVRVRVRV/VB w  rWq,           pTpV P                  V4      pV P                  V4      pWq,           pV# )r}   rÃ   r�   rà   rÄ   © )rÜ   rÙ   rÝ   rÚ   )	r7   r}   rÃ   r�   rà   rÄ   r¢   ÚresidualÚ_s	   &&&&&&,  r*   rH   ÚVaultGemmaDecoderLayer.forwardõ   s�   € ð !ˆØ×,Ñ,¨]Ó;ˆàŸ>š>ñ 
Ø'ð
à 3ð
ð *ð
ð &ð	
ð
 ,ð
ð ñ
Ñˆð !Õ0ˆà ˆØ×6Ñ6°}ÓEˆØŸ™ Ó/ˆØ Õ0ˆàÐr-   )r]   r^   rÜ   rÚ   rÝ   rÙ   rÐ   rj   rV   s   @@r*   rÔ   rÔ   ê   s    ù‡ € ÷hó h÷÷ ò r-   rÔ   c                   óÎ   a a€ ] tR tRt o]! RRR7      RV3R lV 3R lll4       t]]! RRR7      RV3R lR ll4       4       t]P                  ! 4       ]
R	 4       4       tR
tVtV ;t# )ÚVaultGemmaRotaryEmbeddingi  Údevicez5.18)Úversionc                ó    <€ V ^8„  d   QhRS[ /# ©r"   r]   r   )r(   r)   s   "€r*   r+   Ú&VaultGemmaRotaryEmbedding.__annotate__  s   ø€ ÷ Oñ OÑ/ñ Or-   c                óä  <€ \         SV `  4        VP                  V n        VP                  V n        Wn        V P
                  P                  R ,          V n        V P                  pV P                  R8w  d   \        V P                  ,          pV! V P
                  V4      w  q@n
        \        P                  ! VRR7      V n        \        P                  ! VP                  4       RR7      V n        R# )Ú	rope_typeÚdefaultF©Ú
persistentN)r0   r1   Úmax_position_embeddingsÚmax_seq_len_cachedÚoriginal_max_seq_lenr]   Úrope_parametersrð   Úcompute_default_rope_parametersr   Úattention_scalingr2   ÚBufferÚinv_freqÚcloneÚoriginal_inv_freq)r7   r]   rê   Úrope_init_fnrû   r8   s   &&&  €r*   r1   Ú"VaultGemmaRotaryEmbedding.__init__  s¬   ø€ ä‰ÑÔØ"(×"@Ñ"@ˆÔØ$*×$BÑ$BˆÔ!àŒàŸ™×4Ñ4°[ÕAˆŒØ!%×!EÑ!EˆØ�>‰>˜YÔ&Ü.¨t¯~©~Õ>ˆLÙ+7¸¿¹ÀVÓ+LÑ(ˆÔ(äŸ	š	 (°uÔ=ˆŒÜ!#§¢¨8¯>©>Ó+;ÈÔ!NˆÖr-   c                óP   <€ V ^8„  d   QhRS[ RS[S[P                  S[3,          /# )r"   r]   r   )r   rK   r4   r€   r'   )r(   r)   s   "€r*   r+   rî   '  s/   ø€ ÷ 5ñ 5Ñ0@ð 5Ñ\aÑbg×bnÑbnÑpuÐbuÕ\vñ 5r-   c           	     ó0  € V P                   R,          p\        V RR4      ;'       g    V P                  V P                  ,          pRpRV\        P
                  ! ^ V^\        P                  R7      V,          ,          ,          pVP                  V4      V3# )a~  
Computes the inverse frequencies according to the original RoPE implementation
Args:
    config ([`~transformers.PreTrainedConfig`]):
        The model configuration.
Returns:
    Tuple of (`torch.Tensor`, `float`), containing the inverse frequencies for the RoPE embeddings and the
    post-processing scaling factor applied to the computed cos/sin (unused in this type of RoPE).
Ú
rope_thetarˆ   NrE   ©r•   )r÷   r´   r^   rµ   r4   Úaranger'   r    )r]   rê   r¢   Úbaser#   Úattention_factorrû   s   &&,    r*   rø   Ú9VaultGemmaRotaryEmbedding.compute_default_rope_parameters%  s~   € ð ×%Ñ% lÕ3ˆÜ�f˜j¨$Ó/×cÐc°6×3EÑ3EÈ×IcÑIcÕ3cˆàÐà˜$¤5§<¢<°°3¸ÄÇÁÔ#MÐPSÕ#SÕTÕUˆØ�{‰{˜6Ó"Ð$4Ð4Ð4r-   c                ó”  € V P                   R,          P                  4       P                  VP                  ^ ,          R^4      P	                  VP
                  4      pVR	,          P                  4       p\        VP
                  P                  \        4      '       d2   VP
                  P                  R8w  d   VP
                  P                  MRp\        VRR7      ;_uu_ 4        VP                  4       VP                  4       ,          P                  ^^4      p\        P                  ! Wf3RR7      pVP                  4       V P                  ,          pVP                  4       V P                  ,          p	R R R 4       XP	                  VP                   R7      X	P	                  VP                   R7      3#   + '       g   i     LH; i)
NÚmpsÚcpuF)Údevice_typeÚenabledrl   r  )Nr‚   Nr<   )r‚   Nr‚   )rû   r'   rƒ   rL   r    rê   Ú
isinstanceÚtypeÚstrr   r›   r4   rm   rv   rù   rw   r•   )
r7   r@   rà   Úinv_freq_expandedÚposition_ids_expandedr  ÚfreqsÚembrv   rw   s
   &&&       r*   rH   Ú!VaultGemmaRotaryEmbedding.forward9  sE  € ð !ŸM™M¨-Õ8×>Ñ>Ó@×GÑGÈ×HZÑHZÐ[\ÕH]Ð_aÐcdÓe×hÑhÐij×iqÑiqÓrÐØ ,¨ZÕ 8× >Ñ >Ó @Ðä'1°!·(±(·-±-Ä×'EÒ'EÈ!Ï(É(Ï-É-Ð[`ÔJ`�a—h‘h—m’mÐfkˆÜ¨¸U×CÖCØ&×,Ñ,Ó.Ð1F×1LÑ1LÓ1NÕN×YÑYÐZ[Ð]^Ó_ˆEÜ—)’)˜U˜N°Ô3ˆCØ—'‘'“)˜d×4Ñ4Õ4ˆCØ—'‘'“)˜d×4Ñ4Õ4ˆC÷	 Dð �v‰v˜AŸG™GˆvÓ$ c§f¡f°1·7±7 fÓ&;Ð;Ð;÷ D×Cús   Ã&BF7Æ7G	)rù   r]   rû   rõ   rý   rö   rð   r/   )rO   rP   rQ   rR   r   r1   Ústaticmethodrø   r4   Úno_gradr   rH   rS   rT   rU   rV   s   @@r*   ré   ré     su   ù‡ € Ù�X vÔ.÷Oô Oó /ðOð  Ù�X vÔ.÷5ñ 5ó /ó ð5ð$ ‡]‚]ƒ_Øñ<ó ó ÷<ð <r-   ré   c                   ó^   a a€ ] tR tRt oRtRV3R lV 3R llltV3R lV 3R lltRtVtV ;t	# )	Ú!VaultGemmaTextScaledWordEmbeddingiI  zT
This module overrides nn.Embeddings' forward by multiplying with embeddings scale.
c                ó2   <€ V ^8„  d   QhRS[ RS[ RS[ RS[/# )r"   Únum_embeddingsÚembedding_dimÚpadding_idxÚembed_scaler%   )r(   r)   s   "€r*   r+   Ú.VaultGemmaTextScaledWordEmbedding.__annotate__N  s0   ø€ ÷ Rñ R¡sð R¹3ð RÉSð RÑ_dñ Rr-   c                ó–   <€ \         SV `  WV4       W@n        \        P                  ! \
        P                  ! V4      R R7      V n        R# )Frò   N)r0   r1   Úscalar_embed_scaler2   rú   r4   Útensorr  )r7   r  r  r  r  r8   s   &&&&&€r*   r1   Ú*VaultGemmaTextScaledWordEmbedding.__init__N  s4   ø€ Ü‰Ñ˜¸ÔDØ"-ÔÜŸ9š9¤U§\¢\°+Ó%>È5ÔQˆÖr-   c                ó4   <€ V ^8„  d   QhRS[ P                  /# )r"   Ú	input_ids)r4   r€   )r(   r)   s   "€r*   r+   r  S  s   ø€ ÷ Sñ S¡§¡ñ Sr-   c                óŠ   <€ \         SV `  V4      V P                  P                  V P                  P
                  4      ,          # r/   )r0   rH   r  r    r6   r•   )r7   r$  r8   s   &&€r*   rH   Ú)VaultGemmaTextScaledWordEmbedding.forwardS  s2   ø€ Ü‰w‰˜yÓ)¨D×,<Ñ,<×,?Ñ,?ÀÇÁ×@QÑ@QÓ,RÕRÐRr-   )r  r   rD   rÑ   rV   s   @@r*   r  r  I  s(   ù‡ € ñ÷Rõ R÷
S÷ Só Sr-   r  c                   óš   a a€ ] tR tRt oRtRtR.tR.tRtRt	Rt
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VaultGemmaPreTrainedModeliW  ÚmodelTrÔ   rÄ   r}   Ú
attentionsc                ó&  <€ \         SV `  V4       R VP                  P                  9   d#   \        P
                  ! VP                  4       R# \        V\        4      '       d.   \        P                  ! VP                  VP                  4       R# R# )ÚRMSNormN)r0   Ú_init_weightsr8   rO   ÚinitÚzeros_r6   r  r  Ú	constant_r  r   )r7   r‹   r8   s   &&€r*   r-  Ú'VaultGemmaPreTrainedModel._init_weightsi  sb   ø€ ä‰Ñ˜fÔ%à˜×(Ñ(×1Ñ1Ô1Ü�KŠK˜Ÿ™Ö&Ü˜Ô A×BÒBÜ�NŠN˜6×-Ñ-¨v×/HÑ/HÖIñ Cr-   c                ó&   <€ V ^8„  d   Qh/ S[ ;R&   # rí   r   )r(   r)   s   "€r*   r+   Ú&VaultGemmaPreTrainedModel.__annotate__W  s   ø‡ ‚ áÑò r-   rä   )rO   rP   rQ   rR   Úbase_model_prefixÚsupports_gradient_checkpointingÚ_no_split_modulesÚ_skip_keys_device_placementÚ_supports_flash_attnÚ_supports_sdpaÚ_supports_flex_attnÚ_can_compile_fullgraphÚ_supports_attention_backendrÔ   r©   Ú_can_record_outputsr4   r  r-  Ú__annotate_func__rS   rT   rU   rV   s   @@r*   r(  r(  W  sy   ù‡ € ð  ÐØ&*Ð#Ø1Ð2ÐØ#4Ð"5ÐØÐØ€NØÐà!ÐØ"&ÐàÐ/ØÐ)ðÐð
 ‡]‚]ƒ_ôJó ðJ÷' … r-   r(  c                   ór   a a€ ] tR tRt oV3R lV 3R llt]]]RV3R lR ll4       4       4       tRt	Vt
V ;t# )ÚVaultGemmaModelis  c                ó    <€ V ^8„  d   QhRS[ /# rí   r   )r(   r)   s   "€r*   r+   ÚVaultGemmaModel.__annotate__u  s   ø€ ÷ ñ Ñ/ñ r-   c           	     ó6  <€ \         SV `  V4       VP                  V n        VP                  V n        \        VP                  VP                  V P                  V P                  P                  R ,          R7      V n        \        P                  ! \        VP                  4       Uu. uF  p\        W4      NK  	  up4      V n        \        VP                  VP                   R7      V n        \%        V4      V n        RV n        V P+                  4        R# u upi )g      à?)r  rØ   FN)r0   r1   Úpad_token_idr  Ú
vocab_sizer  r^   r]   Úembed_tokensr2   Ú
ModuleListÚrangeÚnum_hidden_layersrÔ   Úlayersr    rÛ   Únormré   Ú
rotary_embÚgradient_checkpointingÚ	post_initrÀ   s   && €r*   r1   ÚVaultGemmaModel.__init__u  sß   ø€ Ü‰Ñ˜Ô Ø!×.Ñ.ˆÔØ ×+Ñ+ˆŒä=Ø×Ñ˜v×1Ñ1°4×3CÑ3CÐQU×Q\ÑQ\×QhÑQhÐjmÕQmô
ˆÔô —m’mÜHMÈf×NfÑNfÔHgÓhÑHg¸9Ô# FÖ6ÑHgÑhó
ˆŒô & f×&8Ñ&8¸f×>QÑ>QÔRˆŒ	Ü3°FÓ;ˆŒØ&+ˆÔ#ð 	�‰Öùò is   Â'Dc                óþ   <€ V ^8„  d   QhRS[ P                  R,          RS[ P                  R,          RS[ P                  R,          RS[R,          RS[ P                  R,          RS[R,          RS[S[,          R	S[/# )
r"   r$  Nr�   rà   rÄ   Úinputs_embedsÚ	use_cacher¢   r   )	r4   rá   r€   r   râ   Úboolr   r   r   )r(   r)   s   "€r*   r+   rB  Š  s“   ø€ ÷ ;
ñ ;
á×#Ñ# dÕ*ð;
ñ Ÿ™ tÕ+ð;
ñ ×&Ñ&¨Õ-ð	;
ñ
  �ð;
ñ ×(Ñ(¨4Õ/ð;
ñ ˜$•;ð;
ñ Ñ+Õ,ð;
ñ 
!ñ;
r-   c                ó2  € VR J VR J,          '       d   \        R4      hVf   V P                  V4      pV'       d   Vf   \        V P                  R7      pVfb   Ve   VP	                  4       M^ p\
        P                  ! VP                  ^,          VP                  R7      V,           pVP                  ^ 4      p\        T;p	\        4      '       g/   RV P                  RVRVRVRV/p
R	\        R/ V
B R
\        R/ V
B /p	TpV P                  W³4      p\        V P                   R V P                  P"                   4       F7  w  rÞV! V3RW�P                  P$                  V,          ,          RVRVRV/VB pK9  	  V P'                  V4      p\)        VVR7      # )Nz:You must specify exactly one of input_ids or inputs_embeds)r]   )rê   r]   rQ  r�   rÄ   rà   Úfull_attentionr±   rÃ   )Úlast_hidden_staterÄ   rä   )Ú
ValueErrorrF  r   r]   Úget_seq_lengthr4   r  rL   rê   rs   r  Údictr   r   rL  Ú	enumeraterJ  rI  r°   rK  r   )r7   r$  r�   rà   rÄ   rQ  rR  r¢   Úpast_seen_tokensÚcausal_mask_mappingÚmask_kwargsr}   rÃ   ÚiÚdecoder_layers   &&&&&&&,       r*   rH   ÚVaultGemmaModel.forward‡  s·  € ð ˜Ð -°tÐ";×<Ô<ÜÐYÓZÐZàÒ Ø*.×*;Ñ*;¸IÓ*FˆMç˜Ò0Ü*°$·+±+Ô>ˆOàÒØCRÒC^˜×=Ñ=Ô?ÐdeÐÜ Ÿ<š<¨×(;Ñ(;¸AÕ(>À}×G[ÑG[Ô\Ð_oÕoˆLØ'×1Ñ1°!Ó4ˆLô °Ð?Ð-Ä×FÒFð ˜$Ÿ+™+Ø Ø  .Ø! ?Ø ðˆKð !Ô"4Ñ"C°{Ñ"CØ#Ô%FÑ%UÈÑ%Uð#Ðð &ˆØ"Ÿo™o¨mÓJÐä )¨$¯+©+Ð6U¸¿¹×8UÑ8UÐ*VÖ WÑˆAÙ)Øñà2·;±;×3JÑ3JÈ1Õ3MÕNðð %8ðð *ð	ð
 !0ðð ñŠMñ !Xð Ÿ	™	 -Ó0ˆä&Ø+Ø+ô
ð 	
r-   )rF  rM  rJ  rK  r  rL  rE  )NNNNNN)rO   rP   rQ   rR   r1   r   r   r   rH   rS   rT   rU   rV   s   @@r*   r@  r@  s  s=   ù‡ € ÷ó ð$  ØØ÷;
ñ ;
ó ó ó  ÷;
ð ;
r-   r@  c                   ó„   a a€ ] tR tRt oRR/tRR/tRR.R.3/tRR/tV 3R	 lt]	]
RV3R
 lR ll4       4       tRtVtV ;t# )ÚVaultGemmaForCausalLMiÈ  zlm_head.weightzmodel.embed_tokens.weightÚlm_headÚcolwise_gather_outputr}   ÚlogitsÚkeep_full_weightc                óì   <€ \         SV `  V4       \        V4      V n        VP                  V n        \
        P                  ! VP                  VP                  R R7      V n        V P                  4        R# rZ   )
r0   r1   r@  r)  rE  r2   r`   r^   rc  rN  rf   s   &&€r*   r1   ÚVaultGemmaForCausalLM.__init__Ï  sU   ø€ Ü‰Ñ˜Ô Ü$ VÓ,ˆŒ
Ø ×+Ñ+ˆŒÜ—y’y ×!3Ñ!3°V×5FÑ5FÈUÔSˆŒð 	�‰Ör-   c                óP  <€ V ^8„  d   QhRS[ P                  R,          RS[ P                  R,          RS[ P                  R,          RS[R,          RS[ P                  R,          RS[ P                  R,          RS[R,          R	S[S[ P                  ,          R
S[S[,          RS[	/
# )r"   r$  Nr�   rà   rÄ   rQ  ÚlabelsrR  Úlogits_to_keepr¢   r   )
r4   rá   r€   r   râ   rS  r&   r   r   r   )r(   r)   s   "€r*   r+   Ú"VaultGemmaForCausalLM.__annotate__Ú  s»   ø€ ÷ ;
ñ ;
á×#Ñ# dÕ*ð;
ñ Ÿ™ tÕ+ð;
ñ ×&Ñ&¨Õ-ð	;
ñ
  �ð;
ñ ×(Ñ(¨4Õ/ð;
ñ × Ñ  4Õ'ð;
ñ ˜$•;ð;
ñ ™eŸl™lÕ*ð;
ñ Ñ+Õ,ð;
ñ 
 ñ;
r-   c	                óF  € V P                   ! R
RVRVRVRVRVRV/V	B p
V
P                  p\        V\        4      '       d   \	        V) R4      MTpV P                  VRVR3,          4      pV P                  P                  eO   WÐP                  P                  ,          p\        P                  ! V4      pWÐP                  P                  ,          pRpVe   V P                  ! WÖV P                  3/ V	B p\        VVV
P                  V
P                  V
P                  R	7      # )a*  
Example:

```python
>>> from transformers import AutoTokenizer, VaultGemmaForCausalLM

>>> model = VaultGemmaForCausalLM.from_pretrained("google/gemma-2-9b")
>>> tokenizer = AutoTokenizer.from_pretrained("google/gemma-2-9b")

>>> prompt = "What is your favorite condiment?"
>>> inputs = tokenizer(prompt, return_tensors="pt")

>>> # Generate
>>> generate_ids = model.generate(inputs.input_ids, max_length=30)
>>> tokenizer.batch_decode(generate_ids, skip_special_tokens=True, clean_up_tokenization_spaces=False)[0]
"What is your favorite condiment?"
```r$  r�   rà   rÄ   rQ  rR  Nr‚   )Úlossre  rÄ   r}   r*  rä   )r)  rV  r  r&   Úslicerc  r]   Úfinal_logit_softcappingr4   rœ   Úloss_functionrE  r   rÄ   r}   r*  )r7   r$  r�   rà   rÄ   rQ  rj  rR  rk  r¢   Úoutputsr}   Úslice_indicesre  rn  s   &&&&&&&&&,     r*   rH   ÚVaultGemmaForCausalLM.forwardØ  s0  € ð@ ,0¯:ª:ñ ,
Øð,
à)ð,
ð &ð,
ð ,ð	,
ð
 (ð,
ð  ð,
ð ñ,
ˆð  ×1Ñ1ˆä8BÀ>ÔSV×8WÒ8Wœ˜~˜o¨tÔ4Ð]kˆØ—‘˜m¨A¨}¸aÐ,?Õ@ÓAˆØ�;‰;×.Ñ.Ò:ØŸk™k×AÑAÕAˆFÜ—Z’Z Ó'ˆFØŸk™k×AÑAÕAˆFàˆØÒØ×%Ò% f°d·o±oÑPÈÑPˆDä%ØØØ#×3Ñ3Ø!×/Ñ/Ø×)Ñ)ô
ð 	
r-   )rc  r)  rE  )NNNNNNNr   )rO   rP   rQ   rR   Ú_tied_weights_keysÚ_tp_planÚ_pp_planÚ
_fsdp_planr1   r   r   rH   rS   rT   rU   rV   s   @@r*   rb  rb  È  sf   ù‡ € à*Ð,GÐHÐØÐ2Ð3€HØ˜_Ð-°¨zÐ:Ð;€HØÐ/Ð0€Jõð Ø÷;
ñ ;
ó ó ÷;
ð ;
r-   rb  )é   )rÆ   NN)rb  r@  r(  )AÚcollections.abcr   r4   Útorch.nnr2   Ú r   r.  Úactivationsr   Úcache_utilsr   r   Ú
generationr   Úintegrationsr	   r
   Úmasking_utilsr   r   Úmodeling_flash_attention_utilsr   Úmodeling_layersr   Úmodeling_outputsr   r   Úmodeling_rope_utilsr   r   Úmodeling_utilsr   r   Úprocessing_utilsr   Úutilsr   r   r   Úutils.deprecationr   Úutils.genericr   r   Úutils.output_capturingr   Úconfiguration_vaultgemmar   r“   r    rX   rp   r{   r‰   r§   r©   rÔ   ré   Ú	Embeddingr  r(  r@  rb  Ú__all__rä   r-   r*   Ú<module>r�     sS  ðõ, %ã Ý å &Ý !ß .Ý )ß Lß RÝ BÝ 9ß Oß Kß FÝ &ß IÑ IÝ 0ß GÝ 5Ý 6ô=˜Ÿ	™	ô =ô(�B—I‘Iô ò (ñ Ð-Ó.óó /ðõ2	U÷%ñD Ð)Ó*ôE)˜"Ÿ)™)ó E)ó +ðE)ôP&Ð7ô &ôR3< §	¡	ô 3<ôlS¨¯©ô Sð ôJ ó Jó ðJð6 ôQ
Ð/ó Q
ó ðQ
ðh ôL
Ð5°ó L
ó ðL
ò^ T‚r-   