§
    9²¸je‚  ã                   óP  — d dl Z d dlmZ d dlZd dlmc mZ d dlmZ ddlm	Z
 ddlmZ ddlmZmZ ddlmZ dd	lmZmZ dd
lmZ ddlmZ ddlmZmZmZ ddlmZmZ ddl m!Z!m"Z" ddl#m$Z$m%Z% ddl&m'Z' ddl(m)Z)m*Z*m+Z+ ddl,m-Z- ddl.m/Z/m0Z0 ddl1m2Z2 ddl3m4Z4  ed¦  «         G d„ dej5        ¦  «        ¦   «         Z6 G d„ dej5        ¦  «        Z7 G d„ dej5        ¦  «        Z8 G d„ dej5        ¦  «        Z9e G d „ d!ej5        ¦  «        ¦   «         Z: G d"„ d#ej5        ¦  «        Z;d$ej<        d%e=d&ej<        fd'„Z>	 dKd)ej5        d*ej<        d+ej<        d,ej<        d-ej<        dz  d.e?d/e?d0e'e)         fd1„Z@dLd2„ZAd3„ ZBd4„ ZC ed5¦  «        dMd6„¦   «         ZDdNd7„ZEd8ej<        d9e?d:e=d&ej<        fd;„ZF G d<„ d=ej5        ¦  «        ZG G d>„ d?e¦  «        ZH G d@„ dAe%¦  «        ZIe* G dB„ dCeI¦  «        ¦   «         ZJe* G dD„ dEeIe¦  «        ¦   «         ZK G dF„ dGeeI¦  «        ZL G dH„ dIeeI¦  «        ZMg dJ¢ZNdS )Oé    N)ÚCallable)Únné   )Úinitialization)ÚACT2FN)ÚCacheÚDynamicCache)ÚGenerationMixin)Úuse_experts_implementationÚuse_kernel_forward_from_hub)Úcreate_causal_mask)ÚFlashAttentionKwargs)Ú GenericForSequenceClassificationÚGenericForTokenClassificationÚ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_outputsé   )ÚMistral4ConfigÚRMSNormc                   óT   ‡ — e Zd Zd	deddfˆ fd„Zdej        dej        fd„Zd„ Zˆ xZ	S )
ÚMistral4RMSNormç�íµ ÷Æ°>ÚepsÚreturnNc                 ó¬   •— t          ¦   «                              ¦   «          t          j        t	          j        |¦  «        ¦  «        | _        || _        dS )z>
        Mistral4RMSNorm is equivalent to T5LayerNorm
        N)ÚsuperÚ__init__r   Ú	ParameterÚtorchÚonesÚweightÚvariance_epsilon)ÚselfÚhidden_sizer&   Ú	__class__s      €ú~/root/private/project/pallet_detection_v2/.venv/lib/python3.11/site-packages/transformers/models/mistral4/modeling_mistral4.pyr*   zMistral4RMSNorm.__init__4   sD   ø€ õ 	‰Œ×ÒÑÔÐÝ”l¥5¤:¨kÑ#:Ô#:Ñ;Ô;ˆŒØ #ˆÔÐÐó    Úhidden_statesc                 ó  — |j         }|                     t          j        ¦  «        }|                     d¦  «                             dd¬¦  «        }|t          j        || j        z   ¦  «        z  }| j        |                     |¦  «        z  S )Né   éÿÿÿÿT)Úkeepdim)	ÚdtypeÚtor,   Úfloat32ÚpowÚmeanÚrsqrtr/   r.   )r0   r5   Úinput_dtypeÚvariances       r3   ÚforwardzMistral4RMSNorm.forward<   s|   € Ø#Ô)ˆØ%×(Ò(­¬Ñ7Ô7ˆØ ×$Ò$ QÑ'Ô'×,Ò,¨R¸Ð,Ñ>Ô>ˆØ%­¬°H¸tÔ?TÑ4TÑ(UÔ(UÑUˆØŒ{˜]×-Ò-¨kÑ:Ô:Ñ:Ð:r4   c                 óH   — t          | j        j        ¦  «        › d| j        › �S )Nz, eps=)Útupler.   Úshaper/   )r0   s    r3   Ú
extra_reprzMistral4RMSNorm.extra_reprC   s&   € Ý˜œÔ)Ñ*Ô*ÐIÐI°$Ô2GÐIÐIÐIr4   )r%   )
Ú__name__Ú
__module__Ú__qualname__Úfloatr*   r,   ÚTensorrB   rF   Ú__classcell__©r2   s   @r3   r$   r$   2   sŒ   ø€ € € € € ð$ð $¨ð $¸$ð $ð $ð $ð $ð $ð $ð; U¤\ð ;°e´lð ;ð ;ð ;ð ;ðJð Jð Jð Jð Jð Jð Jr4   r$   c            	       óî   ‡ — e Zd Z edd¬¦  «        d
defˆ fd„¦   «         Ze edd¬¦  «        d
dedeej	        e
f         fd„¦   «         ¦   «         Z ej        ¦   «         ed	„ ¦   «         ¦   «         Zˆ xZS )ÚMistral4RotaryEmbeddingÚdevicez5.18)ÚversionNÚconfigc                 ó¾  •— t          ¦   «                              ¦   «          |j        | _        |j        | _        || _        | j        j        d         | _        | j        }| j        dk    rt          | j                 } || j        |¦  «        \  }| _
        t          j        |d¬¦  «        | _        t          j        |                     ¦   «         d¬¦  «        | _        d S )NÚ	rope_typeÚdefaultF)Ú
persistent)r)   r*   Úmax_position_embeddingsÚmax_seq_len_cachedÚoriginal_max_seq_lenrR   Úrope_parametersrT   Úcompute_default_rope_parametersr   Úattention_scalingr   ÚBufferÚinv_freqÚcloneÚoriginal_inv_freq)r0   rR   rP   Úrope_init_fnr^   r2   s        €r3   r*   z Mistral4RotaryEmbedding.__init__H   sÀ   ø€ å‰Œ×ÒÑÔÐØ"(Ô"@ˆÔØ$*Ô$BˆÔ!àˆŒàœÔ4°[ÔAˆŒØ!%Ô!EˆØŒ>˜YÒ&Ð&Ý.¨t¬~Ô>ˆLØ+7¨<¸¼ÀVÑ+LÔ+LÑ(ˆ�$Ô(åœ	 (°uÐ=Ñ=Ô=ˆŒÝ!#¤¨8¯>ª>Ñ+;Ô+;ÈÐ!NÑ!NÔ!NˆÔÐÐr4   r'   c                 óä   — | j         d         }t          | dd¦  «        p| j        | j        z  }d}d|t	          j        d|dt          j        ¬¦  «        |z  z  z  }|                     |¦  «        |fS )a¾  
        Computes the inverse frequencies according to the original RoPE implementation
        Args:
            config ([`~transformers.PreTrainedConfig`]):
                The model configuration.
        Returns:
            Tuple of (`torch.Tensor`, `float`), containing the inverse frequencies for the RoPE embeddings and the
            post-processing scaling factor applied to the computed cos/sin (unused in this type of RoPE).
        Ú
rope_thetaÚhead_dimNç      ð?r   r7   ©r:   )rZ   Úgetattrr1   Únum_attention_headsr,   ÚarangerJ   r;   )rR   rP   ÚkwargsÚbaseÚdimÚattention_factorr^   s          r3   r[   z7Mistral4RotaryEmbedding.compute_default_rope_parametersY   sz   € ð Ô% lÔ3ˆÝ�f˜j¨$Ñ/Ô/Ðc°6Ô3EÈÔIcÑ3cˆàÐà˜$¥5¤<°°3¸ÅÄÐ#MÑ#MÔ#MÐPSÑ#SÑTÑUˆØ�{Š{˜6Ñ"Ô"Ð$4Ð4Ð4r4   c                 óú  — | j         d d d …d f                              |j        d         dd¦  «                             t          j        |j        ¬¦  «        }|d d …d d d …f                              ¦   «         }t          |j        j        t          ¦  «        r|j        j        dk    r|j        j        nd}t          |d¬¦  «        5  ||z                       dd	¦  «        }t	          j        ||fd¬
¦  «        }|                     ¦   «         | j        z  }|                     ¦   «         | j        z  }	d d d ¦  «         n# 1 swxY w Y   |                     |j        ¬¦  «        |	                     |j        ¬¦  «        fS )Nr   r8   r    )r:   rP   ÚmpsÚcpuF)Údevice_typeÚenabledr7   ©rl   rf   )r^   ÚexpandrE   r;   r,   rJ   rP   Ú
isinstanceÚtypeÚstrr   Ú	transposeÚcatÚcosr\   Úsinr:   )
r0   ÚxÚposition_idsÚinv_freq_expandedÚposition_ids_expandedrq   ÚfreqsÚembrz   r{   s
             r3   rB   zMistral4RotaryEmbedding.forwardm   s›  € ð ŒM˜$    4˜-Ô(×/Ò/°Ô0BÀ1Ô0EÀrÈ1ÑMÔM×PÒPÕW\ÔWbÐklÔksÐPÑtÔtð 	ð !-¨Q¨Q¨Q°°a°a°a¨ZÔ 8× >Ò >Ñ @Ô @Ðå'1°!´(´-ÅÑ'EÔ'EÐkÈ!Ì(Ì-Ð[`ÒJ`ÐJ`�a”h”m�mÐfkˆå¨¸UÐCÑCÔCð 	5ð 	5Ø&Ð)>Ñ>×IÒIÈ!ÈQÑOÔOˆEÝ”)˜U E˜N°Ð3Ñ3Ô3ˆCØ—'’'‘)”)˜dÔ4Ñ4ˆCØ—'’'‘)”)˜dÔ4Ñ4ˆCð		5ð 	5ð 	5ñ 	5ô 	5ð 	5ð 	5ð 	5ð 	5ð 	5ð 	5øøøð 	5ð 	5ð 	5ð 	5ð �vŠv˜AœGˆvÑ$Ô$ c§f¢f°1´7 fÑ&;Ô&;Ð;Ð;s   ÃA*D<Ä<E ÅE ©N)rG   rH   rI   r   r!   r*   ÚstaticmethodrD   r,   rK   rJ   r[   Úno_gradr   rB   rL   rM   s   @r3   rO   rO   G   sö   ø€ € € € € Ø€_�X vÐ.Ñ.Ô.ðOð O˜~ð Oð Oð Oð Oð Oñ /Ô.ðOð  Ø€_�X vÐ.Ñ.Ô.ð5ð 5°ð 5ÐZ_Ð`eÔ`lÐnsÐ`sÔZtð 5ð 5ð 5ñ /Ô.ñ „\ð5ð$ €U„]�_„_Øð<ð <ñ Ôñ „_ð<ð <ð <ð <ð <r4   rO   c                   ó&   ‡ — e Zd Zdˆ fd„	Zd„ Zˆ xZS )ÚMistral4MLPNc                 ó   •— t          ¦   «                              ¦   «          || _        |j        | _        |€|j        n|| _        t          j        | j        | j        d¬¦  «        | _        t          j        | j        | j        d¬¦  «        | _        t          j        | j        | j        d¬¦  «        | _	        t          |j                 | _        d S ©NF©Úbias)r)   r*   rR   r1   Úintermediate_sizer   ÚLinearÚ	gate_projÚup_projÚ	down_projr   Ú
hidden_actÚact_fn)r0   rR   r‹   r2   s      €r3   r*   zMistral4MLP.__init__�   s±   ø€ Ý‰Œ×ÒÑÔÐØˆŒØ!Ô-ˆÔØ=NÐ=V Ô!9Ð!9Ð\mˆÔÝœ 4Ô#3°TÔ5KÐRWÐXÑXÔXˆŒÝ”y Ô!1°4Ô3IÐPUÐVÑVÔVˆŒÝœ 4Ô#9¸4Ô;KÐRWÐXÑXÔXˆŒÝ˜VÔ.Ô/ˆŒˆˆr4   c                 ó¨   — |                       |                      |                      |¦  «        ¦  «        |                      |¦  «        z  ¦  «        }|S r‚   )r�   r‘   r�   rŽ   )r0   r|   r�   s      r3   rB   zMistral4MLP.forward‹   sA   € Ø—N’N 4§;¢;¨t¯~ª~¸aÑ/@Ô/@Ñ#AÔ#AÀDÇLÂLÐQRÁOÄOÑ#SÑTÔTˆ	ØÐr4   r‚   ©rG   rH   rI   r*   rB   rL   rM   s   @r3   r†   r†   €   sL   ø€ € € € € ð0ð 0ð 0ð 0ð 0ð 0ðð ð ð ð ð ð r4   r†   c                   ó$   ‡ — e Zd Zˆ fd„Zd„ Zˆ xZS )ÚMistral4TopkRouterc                 ó\  •— t          ¦   «                              ¦   «          |j        | _        |j        | _        |j        | _        t          j	        t          j        | j        | j        ¦  «        ¦  «        | _        |j        | _        |j        | _        |j        | _        |j        | _        d S r‚   )r)   r*   Únum_experts_per_tokÚtop_kÚnum_local_expertsÚnum_expertsr1   Ú
hidden_dimr   r+   r,   Úzerosr.   Úrouted_scaling_factorÚn_groupÚ	num_groupÚ
topk_groupÚnorm_topk_prob©r0   rR   r2   s     €r3   r*   zMistral4TopkRouter.__init__‘   s‰   ø€ Ý‰Œ×ÒÑÔÐØÔ/ˆŒ
Ø!Ô3ˆÔØ Ô,ˆŒÝ”l¥5¤;¨tÔ/?ÀÄÑ#QÔ#QÑRÔRˆŒØ%+Ô%AˆÔ"ØœˆŒØ Ô+ˆŒØ$Ô3ˆÔÐÐr4   c                 óº  — |                      d| j        ¦  «        }t          j        || j        ¦  «        }|                     d¦  «        }|                      d| j        | j        | j        z  ¦  «                             dd¬¦  «        d          	                    d¬¦  «        }t          j        || j        dd¬¦  «        d         }t          j        |¦  «        }|                     d|d¦  «         |                     d¦  «                             d| j        | j        | j        z  ¦  «                             d| j        ¦  «        }|                     |                     ¦   «          d¦  «        }t          j        || j        dd¬¦  «        d         }	|                     d|	¦  «        }
| j        r|
 	                    dd	¬
¦  «        dz   }|
|z  }
|
| j        z  }
||
|	fS )Nr8   r7   rs   r   F)Úkrl   Úsortedr    ç        T)rl   r9   g#B’¡œÇ;)Úviewr›   ÚFÚlinearr.   ÚsoftmaxrŸ   rš   ÚtopkÚsumr,   r    Ú
zeros_likeÚscatter_Ú	unsqueezert   ÚreshapeÚmasked_fillÚboolr˜   Úgatherr¡   r�   )r0   r5   Úrouter_logitsÚscoresÚgroup_scoresÚ	group_idxÚ
group_maskÚ
score_maskÚscores_for_choiceÚtopk_indicesÚtopk_weightsÚdenominators               r3   rB   zMistral4TopkRouter.forwardœ   sÅ  € Ø%×*Ò*¨2¨t¬Ñ?Ô?ˆÝœ °´Ñ<Ô<ˆØ×&Ò& rÑ*Ô*ˆà�KŠK˜˜DœN¨DÔ,<ÀÄÑ,NÑOÔO×TÒTÐUVÐ\^ÐTÑ_Ô_Ð`aÔb×fÒfÐkmÐfÑnÔnð 	õ ”J˜|¨t¬ÀBÈuÐUÑUÔUÐVWÔXˆ	ÝÔ% lÑ3Ô3ˆ
Ø×Ò˜A˜y¨!Ñ,Ô,Ð,à× Ò  Ñ$Ô$ßŠV�B˜œ¨Ô(8¸D¼NÑ(JÑKÔKßŠW�R˜Ô)Ñ*Ô*ð 	ð
 #×.Ò.°
·²Ñ0AÔ0AÐ/AÀ3ÑGÔGÐÝ”zÐ"3°t´zÀrÐRWÐXÑXÔXÐYZÔ[ˆØ—}’} Q¨Ñ5Ô5ˆØÔð 	(Ø&×*Ò*¨r¸4Ð*Ñ@Ô@À5ÑHˆKØ˜KÑ'ˆLØ# dÔ&@Ñ@ˆØ˜l¨LÐ8Ð8r4   r“   rM   s   @r3   r•   r•   �   sG   ø€ € € € € ð	4ð 	4ð 	4ð 	4ð 	4ð9ð 9ð 9ð 9ð 9ð 9ð 9r4   r•   c                   ób   ‡ — e Zd ZdZˆ fd„Zdej        dej        dej        dej        fd„Zˆ xZS )ÚMistral4Expertsz2Collection of expert weights stored as 3D tensors.c                 ó´  •— t          ¦   «                              ¦   «          |j        | _        |j        | _        |j        | _        t          j	        t          j        | j        d| j        z  | j        ¦  «        ¦  «        | _        t          j	        t          j        | j        | j        | j        ¦  «        ¦  «        | _        t          |j                 | _        d S )Nr7   )r)   r*   r™   rš   r1   r›   Úmoe_intermediate_sizeÚintermediate_dimr   r+   r,   ÚemptyÚgate_up_projr�   r   r�   r‘   r¢   s     €r3   r*   zMistral4Experts.__init__¹   s£   ø€ Ý‰Œ×ÒÑÔÐØ!Ô3ˆÔØ Ô,ˆŒØ &Ô <ˆÔÝœL­¬°TÔ5EÀqÈ4ÔK`ÑG`ÐbfÔbqÑ)rÔ)rÑsÔsˆÔÝœ¥e¤k°$Ô2BÀDÄOÐUYÔUjÑ&kÔ&kÑlÔlˆŒÝ˜VÔ.Ô/ˆŒˆˆr4   r5   Útop_k_indexÚtop_k_weightsr'   c                 ó€  — t          j        |¦  «        }t          j        ¦   «         5  t           j        j                             || j        ¬¦  «        }|                     ddd¦  «        }t          j        | 	                    d¬¦  «        d¦  «         
                    ¦   «         }d d d ¦  «         n# 1 swxY w Y   |D ]þ}|d         }|| j        k    rŒt          j        ||         ¦  «        \  }}	||	         }
t          j                             |
| j        |         ¦  «                             dd¬¦  «        \  }}|                      |¦  «        |z  }t          j                             || j        |         ¦  «        }|||	|d f         z  }|                     d|	|                     |j        ¦  «        ¦  «         Œÿ|S )N)Únum_classesr7   r    r   )r8   éþÿÿÿrs   r8   )r,   r­   r„   r   Ú
functionalÚone_hotrš   ÚpermuteÚgreaterr¬   ÚnonzeroÚwherer©   rÄ   Úchunkr‘   r�   Ú
index_add_r;   r:   )r0   r5   rÅ   rÆ   Úfinal_hidden_statesÚexpert_maskÚ
expert_hitÚ
expert_idxÚ	top_k_posÚ	token_idxÚcurrent_stateÚgateÚupÚcurrent_hidden_statess                 r3   rB   zMistral4Experts.forwardÂ   sø  € õ $Ô.¨}Ñ=Ô=ÐÝŒ]‰_Œ_ð 	Sð 	SÝœ(Ô-×5Ò5°kÈtÔO_Ð5Ñ`Ô`ˆKØ%×-Ò-¨a°°AÑ6Ô6ˆKÝœ {§¢¸8 Ñ'DÔ'DÀaÑHÔH×PÒPÑRÔRˆJð	Sð 	Sð 	Sñ 	Sô 	Sð 	Sð 	Sð 	Sð 	Sð 	Sð 	Søøøð 	Sð 	Sð 	Sð 	Sð
 %ð 
	nð 
	nˆJØ# AœˆJØ˜TÔ-Ò-Ð-ØÝ#(¤;¨{¸:Ô/FÑ#GÔ#GÑ ˆI�yØ)¨)Ô4ˆMÝ”}×+Ò+¨M¸4Ô;LÈZÔ;XÑYÔY×_Ò_Ð`aÐgiÐ_ÑjÔj‰HˆD�"Ø$(§K¢K°Ñ$5Ô$5¸Ñ$:Ð!Ý$&¤M×$8Ò$8Ð9NÐPTÔP^Ð_iÔPjÑ$kÔ$kÐ!Ø$9¸MÈ)ÐU^Ð`dÐJdÔ<eÑ$eÐ!Ø×*Ò*¨1¨iÐ9N×9QÒ9QÐReÔRkÑ9lÔ9lÑmÔmÐmÐmà"Ð"s   ¨A>B2Â2B6Â9B6)	rG   rH   rI   Ú__doc__r*   r,   rK   rB   rL   rM   s   @r3   r¿   r¿   µ   s€   ø€ € € € € à<Ð<ð0ð 0ð 0ð 0ð 0ð#à”|ð#ð ”\ð#ð ”|ð	#ð
 
Œð#ð #ð #ð #ð #ð #ð #ð #r4   r¿   c                   óL   ‡ — e Zd ZdZdefˆ fd„Zdej        dej        fd„Zˆ xZ	S )ÚMistral4MoEz:
    A mixed expert module containing shared experts.
    rR   c                 óì   •— t          ¦   «                              ¦   «          || _        t          |¦  «        | _        t          |¦  «        | _        t          ||j        |j	        z  ¬¦  «        | _
        d S )N)rR   r‹   )r)   r*   rR   r¿   Úexpertsr•   rÙ   r†   rÁ   Ún_shared_expertsÚshared_expertsr¢   s     €r3   r*   zMistral4MoE.__init__â   sk   ø€ Ý‰Œ×ÒÑÔÐØˆŒÝ& vÑ.Ô.ˆŒÝ& vÑ.Ô.ˆŒ	Ý)Ø¨VÔ-IÈFÔLcÑ-cð
ñ 
ô 
ˆÔÐÐr4   r5   r'   c                 óú   — |}|j         }|                      |¦  «        \  }}}|                     d|j         d         ¦  «        } |                      |||¦  «        j        |Ž }||                      |¦  «        z   }|S )Nr8   )rE   rÙ   r§   rà   râ   )r0   r5   Ú	residualsÚ
orig_shapeÚ_r¼   r»   s          r3   rB   zMistral4MoE.forwardë   s‚   € Ø!ˆ	Ø"Ô(ˆ
Ø(,¯	ª	°-Ñ(@Ô(@Ñ%ˆˆ<˜Ø%×*Ò*¨2¨}Ô/BÀ2Ô/FÑGÔGˆØT˜Ÿš ]°LÀ,ÑOÔOÔTÐV`ÐaˆØ%¨×(;Ò(;¸IÑ(FÔ(FÑFˆØÐr4   )
rG   rH   rI   rÜ   r!   r*   r,   rK   rB   rL   rM   s   @r3   rÞ   rÞ   Ý   st   ø€ € € € € ðð ð
˜~ð 
ð 
ð 
ð 
ð 
ð 
ð U¤\ð °e´lð ð ð ð ð ð ð ð r4   rÞ   r5   Ún_repr'   c                 ó¸   — | j         \  }}}}|dk    r| S | dd…dd…ddd…dd…f                              |||||¦  «        } |                      |||z  ||¦  «        S )zÔ
    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)
    r    N)rE   rt   r°   )r5   rç   ÚbatchÚnum_key_value_headsÚslenrd   s         r3   Ú	repeat_kvrì   õ   s„   € ð
 2?Ô1DÑ.€EÐ  hØ�‚z€zØÐØ! ! ! ! Q Q Q¨¨a¨a¨a°°°Ð"2Ô3×:Ò:¸5ÐBUÐW\Ð^bÐdlÑmÔm€MØ× Ò  Ð(;¸eÑ(CÀTÈ8ÑTÔTÐTr4   r¦   ÚmoduleÚqueryÚkeyÚvalueÚattention_maskÚscalingÚdropoutrj   c                 ó  — t          || j        ¦  «        }t          || j        ¦  «        }	t          j        ||                     dd¦  «        ¦  «        |z  }
|�|
|z   }
t
          j                             |
dt          j        ¬¦  «         	                    |j
        ¦  «        }
t
          j                             |
|| j        ¬¦  «        }
t          j        |
|	¦  «        }|                     dd¦  «                             ¦   «         }||
fS )Nr7   r   r8   )rl   r:   )ÚpÚtrainingr    )rì   Únum_key_value_groupsr,   Úmatmulrx   r   rÊ   rª   r<   r;   r:   ró   rö   Ú
contiguous)rí   rî   rï   rð   rñ   rò   ró   rj   Ú
key_statesÚvalue_statesÚattn_weightsÚattn_outputs               r3   Úeager_attention_forwardrþ     sé   € õ ˜3 Ô ;Ñ<Ô<€JÝ˜U FÔ$?Ñ@Ô@€Lå”<  z×';Ò';¸A¸qÑ'AÔ'AÑBÔBÀWÑL€LØÐ!Ø# nÑ4ˆå”=×(Ò(¨¸2ÅUÄ]Ð(ÑSÔS×VÒVÐW\ÔWbÑcÔc€LÝ”=×(Ò(¨¸È6Ì?Ð(Ñ[Ô[€LÝ”,˜|¨\Ñ:Ô:€KØ×'Ò'¨¨1Ñ-Ô-×8Ò8Ñ:Ô:€Kà˜Ð$Ð$r4   c                 óL   — | dk    rdS d|z  t          j        | ¦  «        z  dz   S )Nr    re   gš™™™™™¹?)ÚmathÚlog)ÚscaleÚmscales     r3   Úyarn_get_mscaler    s,   € Ø�‚z€zØˆsØ�‰<�$œ( 5™/œ/Ñ)¨CÑ/Ð/r4   c                 óª   — |                       dd¦  «        dk    r8|                       dd¦  «        }| d         }|rt          ||¦  «        }||z  |z  }|S )NrT   rU   Úmscale_all_dimr   Úfactor)Úgetr  )rZ   rò   r  Úscaling_factorr  s        r3   Úyarn_apply_mscaler
     sh   € Ø×Ò˜;¨	Ñ2Ô2°iÒ?Ð?Ø(×,Ò,Ð-=¸qÑAÔAˆØ(¨Ô2ˆØð 	0Ý$ ^°^ÑDÔDˆFØ Ñ&¨Ñ/ˆGØ€Nr4   c                 óœ   — | dd| j         d         dz  …f         }| d| j         d         dz  d…f         }t          j        | |fd¬¦  «        S )z*Rotates half the hidden dims of the input..Nr8   r7   rs   )rE   r,   ry   )r|   Úx1Úx2s      r3   Úrotate_halfr  *  s]   € à	
ˆ3Ð"�!”'˜"”+ Ñ"Ð"Ð"Ô	#€BØ	
ˆ3�”˜”˜qÑ Ð"Ð"Ð"Ô	#€BÝŒ9�r�c˜2�Y BÐ'Ñ'Ô'Ð'r4   Úrotary_pos_embc                 ó¾   — |                      |¦  «        }|                      |¦  «        }| |z  t          | ¦  «        |z  z   }||z  t          |¦  «        |z  z   }||fS )a…  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.
    )r¯   r  )Úqr¤   rz   r{   Úunsqueeze_dimÚq_embedÚk_embeds          r3   Úapply_rotary_pos_embr  1  sc   € ð& �-Š-˜Ñ
&Ô
&€CØ
�-Š-˜Ñ
&Ô
&€CØ�3‰w�; q™>œ>¨CÑ/Ñ0€GØ�3‰w�; q™>œ>¨CÑ/Ñ0€GØ�GÐÐr4   c                 óÎ  — |dd|j         d         dz  …f                              |¦  «        }|dd|j         d         dz  …f                              |¦  «        }| dddd…f         | dddd…f         }}|dddd…f         |dddd…f         }	}t          j        ||z  ||z  z
  ||z  ||z  z   gd¬¦  «        }
t          j        ||z  |	|z  z
  |	|z  ||z  z   gd¬¦  «        }|
|fS )aû  
    Applies interleaved Rotary Position Embedding to the query and key tensors.

    DeepSeek lays the rotary dimensions out in interleaved pairs `(x0, x1), (x2, x3), ...`, each rotated by a
    single frequency. We compute that rotation directly on the even/odd slices instead of de-interleaving with a
    `view`/`transpose`/`reshape`; the output is bit-identical to the de-interleaved `rotate_half` formulation while
    avoiding the extra contiguous copy.

    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.
        position_ids (`torch.Tensor`):
            The position indices of the tokens corresponding to the query and key tensors. For example, this can be
            used to pass offsetted position ids when working with a KV-cache.
        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.
    .Nr8   r7   r   r    rs   )rE   r¯   r,   ry   )r  r¤   rz   r{   r}   r  Úq1Úq2Úk1Úk2r  r  s               r3   Úapply_rotary_pos_emb_interleaver  K  s  € ð8 ˆcÐ'�S”Y˜r”] aÑ'Ð'Ð'Ô
(×
2Ò
2°=Ñ
AÔ
A€CØ
ˆcÐ'�S”Y˜r”] aÑ'Ð'Ð'Ô
(×
2Ò
2°=Ñ
AÔ
A€Càˆs�A�D�q�DˆyŒ\˜1˜S ! $ Q $˜Yœ<ˆ€BØˆs�A�D�q�DˆyŒ\˜1˜S ! $ Q $˜Yœ<ˆ€BåŒi˜˜c™ B¨¡HÑ,¨b°3©h¸¸c¹Ñ.AÐBÈÐKÑKÔK€GÝŒi˜˜c™ B¨¡HÑ,¨b°3©h¸¸c¹Ñ.AÐBÈÐKÑKÔK€GØ�GÐÐr4   Úpositions_idsÚbetarW   c           	      ó†   — d|t          j        dt          j        | |z  ¦  «        z   ¦  «        z  z   }|d d …d d d …d f         S ©Nr    )r,   r  Úfloor)r  r  rW   rò   s       r3   Úget_llama_4_attn_scaler!  r  sL   € Ø�$�œ 1¥u¤{°=ÐCZÑ3ZÑ'[Ô'[Ñ#[Ñ\Ô\Ñ\Ñ\€GØ�1�1�1�d˜A˜A˜A˜tÐ#Ô$Ð$r4   c                   óB  ‡ — e Zd ZdZdedefˆ fd„Zdej        dej        de	ej        ej        f         fd„Z
	 dd
ej        de	ej        ej        f         dej        d	z  dej        ded	z  dee         de	ej        ej        d	z  e	ej                 d	z  f         fd„Zˆ xZS )ÚMistral4Attentionz4Multi-headed Latent Attention (MLA) from Deepseek V2rR   Ú	layer_idxc                 ó¸  •— t          ¦   «                              ¦   «          || _        || _        |j        | _        |j        | _        |j        | _        |j        | _        |j	        | _	        |j
        | _
        |j        | _        |j        | _        |j        |j	        z   | _        |j        |j        z  | _        d| _        | j        €)t#          j        | j        | j        | j        z  d¬¦  «        nd | _        | j        �&t#          j        | j        |j        |j        ¬¦  «        nd | _        | j        �t-          |j        ¦  «        nd | _        | j        �)t#          j        |j        | j        | j        z  d¬¦  «        nd | _        t#          j        | j        |j
        |j	        z   |j        ¬¦  «        | _        t-          |j
        ¦  «        | _        t#          j        |j
        | j        | j        | j        z   z  d¬¦  «        | _        t#          j        | j        | j        z  | j        |j        ¬¦  «        | _        t;          |j        | j        dz  ¦  «        | _        d S )NTFr‰   g      à¿) r)   r*   rR   r$  Úattention_dropoutr1   rh   Ú	num_headsÚq_lora_rankÚqk_rope_head_dimÚkv_lora_rankÚ
v_head_dimÚqk_nope_head_dimÚqk_head_dimrê   r÷   Ú	is_causalr   rŒ   Úq_projÚattention_biasÚq_a_projr$   Úq_a_layernormÚq_b_projÚkv_a_proj_with_mqaÚkv_a_layernormÚ	kv_b_projÚo_projr
  rZ   rò   ©r0   rR   r$  r2   s      €r3   r*   zMistral4Attention.__init__z  sR  ø€ Ý‰Œ×ÒÑÔÐØˆŒØ"ˆŒØ!'Ô!9ˆÔØ!Ô-ˆÔØÔ3ˆŒà!Ô-ˆÔØ &Ô 7ˆÔØ"Ô/ˆÔØ Ô+ˆŒØ &Ô 7ˆÔØ!Ô2°VÔ5LÑLˆÔØ$*Ô$>À&ÔB\Ñ$\ˆÔ!àˆŒð ÔÐ'õ ŒI�dÔ&¨¬¸Ô9IÑ(IÐPUÐVÑVÔVÐVàð 	Œð ÔÐ+õ ŒI�dÔ&¨Ô(:ÀÔAVÐWÑWÔWÐWàð 	Œð
 EIÔDTÐD`�_¨VÔ-?Ñ@Ô@Ð@ÐfjˆÔð ÔÐ+õ ŒI�fÔ(¨$¬.¸4Ô;KÑ*KÐRWÐXÑXÔXÐXàð 	Œõ #%¤)ØÔØÔ &Ô"9Ñ9ØÔ&ð#
ñ #
ô #
ˆÔõ
 .¨fÔ.AÑBÔBˆÔÝœØÔØŒN˜dÔ3°d´oÑEÑFØð
ñ 
ô 
ˆŒõ ”iØŒN˜Tœ_Ñ,ØÔØÔ&ð
ñ 
ô 
ˆŒõ )¨Ô)?ÀÔAQÐVZÑA[Ñ\Ô\ˆŒˆˆr4   Úkv_nopeÚk_rotr'   c                 ó4  — |j         \  }}}}||d| j        | j        z   f}|                      |¦  «                             |¦  «                             dd¦  «        }t          j        || j        | j        gd¬¦  «        \  }}|                     d|j         d         dd¦  «        } |j	        g |j         dd…         ¢| j        | j
        z   ‘R Ž }	|	dd| j        …f                              |¦  «         |	d| j        d…f                              |¦  «         |	|fS )a“  Expands the compressed latents into key and value states. Args:
            - kv_nope: key + value without positional encoding, shape [batch_size, 1, seqlen, self.kv_lora_rank]
            - k_rot: shared key with positional encoding, shape [batch_size, 1, seqlen, self.qk_rope_head_dim]
        Returns the key and value states, two tensors of shape [batch, num_heads, seq, (k or v)_head_dim].
        r8   r    r7   rs   N.)rE   r,  r+  r6  r§   rx   r,   Úsplitrt   Ú	new_emptyr)  Úcopy_)
r0   r9  r:  Ú
batch_sizeræ   Ú
seq_lengthÚ	key_shapeÚk_noperû   rú   s
             r3   Ú	expand_kvzMistral4Attention.expand_kv±  s-  € ð (/¤}Ñ$ˆ
�A�z 1Ø ¨R°Ô1FÈÌÑ1XÐYˆ	à—.’. Ñ)Ô)×.Ò.¨yÑ9Ô9×CÒCÀAÀqÑIÔIˆÝ$œ{¨7°TÔ5JÈDÌOÐ4\ÐbdÐeÑeÔeÑˆ�Ø—’˜R ¤¨a¤°"°bÑ9Ô9ˆð '�WÔ&Ðj¨¬°c°r°cÔ(:Ðj¸DÔ<QÐTXÔTiÑ<iÐjÐjÐjˆ
Ø�3Ð/˜$Ô/Ð/Ð/Ô0×6Ò6°vÑ>Ô>Ð>Ø�3˜Ô-Ð/Ð/Ð/Ô0×6Ò6°uÑ=Ô=Ð=Ø˜<Ð'Ð'r4   Nr5   Úposition_embeddingsrñ   r}   Úpast_key_valuesrj   c                 óš  — |j         d d…         \  }}||d| j        f}	| j        €|                      |¦  «        }
n;|                      |                      |                      |¦  «        ¦  «        ¦  «        }
|
                     |	¦  «                             dd¦  «        }
t          j
        |
| j        | j        gd¬¦  «        \  }}|                      |¦  «        }t          j
        || j        | j        gd¬¦  «        \  }}|                      |¦  «                             |d|| j        ¦  «        }|                     |d|| j        ¦  «        }|\  }}| j        j        rt%          ||||¦  «        \  }}nt'          ||||¦  «        \  }}|�|                     ||| j        ¦  «        \  }}t          j        ||fd¬¦  «        }|t/          || j        j                             d¦  «        | j        j                             d¦  «        ¦  «                             |j        ¦  «        z  }|                      ||¦  «        \  }}t;          j        | j        j        t@          ¦  «        } || ||||f| j!        sdn| j"        | j#        dœ|¤Ž\  }}| $                    ||d¦  «         %                    ¦   «         }|  &                    |¦  «        }||fS )	Nr8   r    r7   rs   Úllama_4_scaling_betaÚ original_max_position_embeddingsr¦   )ró   rò   )'rE   r-  r(  r/  r3  r2  r1  r§   rx   r,   r<  r,  r)  r4  r*  r5  rR   Úrope_interleaver  r  Úupdater$  ry   r!  rZ   r  r;   r:   rC  r   Úget_interfaceÚ_attn_implementationrþ   rö   r&  rò   r°   rù   r7  )r0   r5   rD  rñ   r}   rE  rj   r?  r@  Úquery_shapeÚq_statesÚq_passÚq_rotÚcompressed_kvÚkv_passr:  Úk_passrz   r{   Úquery_statesrú   rû   Úattention_interfacerý   rü   s                            r3   rB   zMistral4Attention.forwardÄ  sá  € ð "/Ô!4°S°b°SÔ!9Ñˆ
�JØ! :¨r°4Ô3CÐDˆàÔÐ#Ø—{’{ =Ñ1Ô1ˆHˆHà—}’} T×%7Ò%7¸¿ºÀmÑ8TÔ8TÑ%UÔ%UÑVÔVˆHØ—=’= Ñ-Ô-×7Ò7¸¸1Ñ=Ô=ˆÝœ H¨tÔ/DÀdÔF[Ð.\ÐbdÐeÑeÔe‰ˆ�à×/Ò/°Ñ>Ô>ˆÝœ ]°TÔ5FÈÔH]Ð4^ÐdfÐgÑgÔg‰ˆ�à×$Ò$ WÑ-Ô-×2Ò2°:¸qÀ*ÈdÔN_Ñ`Ô`ˆØ—
’
˜: q¨*°dÔ6KÑLÔLˆà&‰ˆˆSØŒ;Ô&ð 	HÝ:¸5À%ÈÈcÑRÔR‰LˆE�5�5å/°°u¸cÀ3ÑGÔG‰LˆE�5ð Ð&Ø+×2Ò2°6¸5À$Ä.ÑQÔQ‰MˆF�Eå”y &¨% °bÐ9Ñ9Ô9ˆà#Õ&<ØØŒKÔ'×+Ò+Ð,BÑCÔCØŒKÔ'×+Ò+Ð,NÑOÔOñ'
ô '
÷ Š"ˆ\ÔÑ
 Ô
 ñ	!ˆð $(§>¢>°&¸%Ñ#@Ô#@Ñ ˆ
�Lå(?Ô(MØŒKÔ,Õ.Eñ)
ô )
Ðð %8Ð$7ØØØØØð	%
ð  $œ}ÐH�C�C°$Ô2HØ”Lð	%
ð 	%
ð ð	%
ð 	%
Ñ!ˆ�\ð "×)Ò)¨*°jÀ"ÑEÔE×PÒPÑRÔRˆØ—k’k +Ñ.Ô.ˆØ˜LÐ(Ð(r4   r‚   )rG   rH   rI   rÜ   r!   Úintr*   r,   rK   rD   rC  r   r   r   rB   rL   rM   s   @r3   r#  r#  w  s=  ø€ € € € € Ø>Ð>ð5]˜~ð 5]¸#ð 5]ð 5]ð 5]ð 5]ð 5]ð 5]ðn( ¤ð (°e´lð (ÀuÈUÌ\Ð[`Ô[gÐMgÔGhð (ð (ð (ð (ð2 )-ð>)ð >)à”|ð>)ð # 5¤<°´Ð#=Ô>ð>)ð œ tÑ+ð	>)ð
 ”lð>)ð  ™ð>)ð Ð-Ô.ð>)ð 
ˆuŒ|˜Uœ\¨DÑ0°%¸¼Ô2EÈÑ2LÐLÔ	Mð>)ð >)ð >)ð >)ð >)ð >)ð >)ð >)r4   r#  c                   óÒ   ‡ — e Zd Zdedefˆ fd„Z	 	 	 	 	 ddej        dej        dz  dej        dz  d	e	dz  d
e
dz  deej        ej        f         dz  dee         dej        fd„Zˆ xZS )ÚMistral4DecoderLayerrR   r$  c                 ót  •— t          ¦   «                              ¦   «          |j        | _        t          ||¬¦  «        | _        ||j        k    rt          |¦  «        | _        nt          |¦  «        | _        t          |j        |j
        ¬¦  «        | _        t          |j        |j
        ¬¦  «        | _        d S )N)rR   r$  ©r&   )r)   r*   r1   r#  Ú	self_attnÚfirst_k_dense_replacerÞ   Úmlpr†   r$   Úrms_norm_epsÚinput_layernormÚpost_attention_layernormr8  s      €r3   r*   zMistral4DecoderLayer.__init__  s£   ø€ Ý‰Œ×ÒÑÔÐØ!Ô-ˆÔå*°&ÀIÐNÑNÔNˆŒà˜Ô4Ò4Ð4Ý" 6Ñ*Ô*ˆDŒHˆHå" 6Ñ*Ô*ˆDŒHå.¨vÔ/AÀvÔGZÐ[Ñ[Ô[ˆÔÝ(7¸Ô8JÐPVÔPcÐ(dÑ(dÔ(dˆÔ%Ð%Ð%r4   NFr5   rñ   r}   rE  Ú	use_cacherD  rj   r'   c           
      óÎ   — |}|                       |¦  «        } | j        d||||||dœ|¤Ž\  }}	||z   }|}|                      |¦  «        }|                      |¦  «        }||z   }|S )N)r5   rñ   r}   rE  ra  rD  © )r_  r[  r`  r]  )
r0   r5   rñ   r}   rE  ra  rD  rj   Úresidualræ   s
             r3   rB   zMistral4DecoderLayer.forward  s¡   € ð !ˆØ×,Ò,¨]Ñ;Ô;ˆà)˜4œ>ð 
Ø'Ø)Ø%Ø+ØØ 3ð
ð 
ð ð
ð 
Ñˆ�qð ! =Ñ0ˆð !ˆØ×5Ò5°mÑDÔDˆØŸš Ñ/Ô/ˆØ  =Ñ0ˆØÐr4   )NNNFN)rG   rH   rI   r!   rV  r*   r,   rK   Ú
LongTensorr   r²   rD   r   r   rB   rL   rM   s   @r3   rX  rX    sÿ   ø€ € € € € ðe˜~ð e¸#ð eð eð eð eð eð eð" /3Ø04Ø(,Ø!&ØHLðð à”|ðð œ tÑ+ðð Ô&¨Ñ-ð	ð
  ™ðð ˜$‘;ðð # 5¤<°´Ð#=Ô>ÀÑEðð Ð+Ô,ðð 
Œðð ð ð ð ð ð ð r4   rX  c                   óŽ   ‡ — e Zd ZU eed<   dZdZdgZdgZdZ	dZ
dZdZdZeedœZg Zg Z ej        ¦   «         ˆ fd„¦   «         Zˆ xZS )ÚMistral4PreTrainedModelrR   ÚmodelTrX  rE  )r5   Ú
attentionsc                 óŠ  •— t          ¦   «                              |¦  «         t          |t          ¦  «        r(t	          j        |j        d| j        j        ¬¦  «         d S t          |t          ¦  «        rNt	          j        |j
        d| j        j        ¬¦  «         t	          j        |j        d| j        j        ¬¦  «         d S d S )Nr¦   )r>   Ústd)r)   Ú_init_weightsru   r•   ÚinitÚnormal_r.   rR   Úinitializer_ranger¿   rÄ   r�   )r0   rí   r2   s     €r3   rl  z%Mistral4PreTrainedModel._init_weightsG  s¶   ø€ å‰Œ×Ò˜fÑ%Ô%Ð%Ý�fÕ0Ñ1Ô1ð 	XÝŒL˜œ¨S°d´kÔ6SÐTÑTÔTÐTÐTÐTÝ˜¥Ñ0Ô0ð 	XÝŒL˜Ô,°3¸D¼KÔ<YÐZÑZÔZÐZÝŒL˜Ô)°¸¼Ô9VÐWÑWÔWÐWÐWÐWð	Xð 	Xr4   )rG   rH   rI   r!   Ú__annotations__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_backendrX  r#  Ú_can_record_outputsÚ_keep_in_fp32_modules_strictÚ"_keys_to_ignore_on_load_unexpectedr,   r„   rl  rL   rM   s   @r3   rg  rg  4  s½   ø€ € € € € € ØÐÐÑØÐØ&*Ð#Ø/Ð0ÐØ#4Ð"5ÐØÐØ€NØÐà!ÐØ"&Ðà-Ø'ðð Ðð $&Ð Ø)+Ð&à€U„]�_„_ðXð Xð Xð Xñ „_ðXð Xð Xð Xð Xr4   rg  c                   óâ   ‡ — e Zd Zdefˆ fd„Zeee	 	 	 	 	 	 ddej	        dz  dej
        dz  dej	        dz  dedz  dej        dz  d	edz  d
ee         defd„¦   «         ¦   «         ¦   «         Zˆ xZS )ÚMistral4ModelrR   c                 óÞ  •‡— t          ¦   «                              ‰¦  «         ‰j        | _        ‰j        | _        t          j        ‰j        ‰j        | j        ¦  «        | _        t          j	        ˆfd„t          ‰j        ¦  «        D ¦   «         ¦  «        | _        t          ‰j        ‰j        ¬¦  «        | _        t!          ‰¬¦  «        | _        d| _        |                      ¦   «          d S )Nc                 ó0   •— g | ]}t          ‰|¦  «        ‘ŒS rc  )rX  )Ú.0r$  rR   s     €r3   ú
<listcomp>z*Mistral4Model.__init__.<locals>.<listcomp>Z  s$   ø€ ÐfÐfÐf¸Õ! &¨)Ñ4Ô4ÐfÐfÐfr4   rZ  ©rR   F)r)   r*   Úpad_token_idÚpadding_idxÚ
vocab_sizer   Ú	Embeddingr1   Úembed_tokensÚ
ModuleListÚrangeÚnum_hidden_layersÚlayersr$   r^  ÚnormrO   Ú
rotary_embÚgradient_checkpointingÚ	post_initr¢   s    `€r3   r*   zMistral4Model.__init__S  sÔ   øø€ Ý‰Œ×Ò˜Ñ Ô Ð Ø!Ô.ˆÔØ Ô+ˆŒåœL¨Ô):¸FÔ<NÐPTÔP`ÑaÔaˆÔÝ”mØfÐfÐfÐfÅeÈFÔLdÑFeÔFeÐfÑfÔfñ
ô 
ˆŒõ $ FÔ$6¸FÔ<OÐPÑPÔPˆŒ	Ý1¸Ð@Ñ@Ô@ˆŒØ&+ˆÔ#ð 	�ŠÑÔÐÐÐr4   NÚ	input_idsrñ   r}   rE  Úinputs_embedsra  rj   r'   c           
      óH  — |d u |d uz  rt          d¦  «        ‚|€|                      |¦  «        }|r|€t          | j        ¬¦  «        }|€V|�|                     ¦   «         nd}t          j        |j        d         |j        ¬¦  «        |z   }| 	                    d¦  «        }t          | j        ||||¬¦  «        }	|}
|                      |
|¬¦  «        }| j        d | j        j        …         D ]} ||
f|	||||dœ|¤Ž}
Œ|                      |
¦  «        }
t          |
|¬	¦  «        S )
Nz:You must specify exactly one of input_ids or inputs_embedsrƒ  r   r    )rP   )rR   r’  rñ   rE  r}   )r}   )rñ   rD  r}   rE  ra  )Úlast_hidden_staterE  )Ú
ValueErrorrˆ  r	   rR   Úget_seq_lengthr,   ri   rE   rP   r¯   r   rŽ  rŒ  r‹  r�  r   )r0   r‘  rñ   r}   rE  r’  ra  rj   Úpast_seen_tokensÚcausal_maskr5   rD  Údecoder_layers                r3   rB   zMistral4Model.forwardc  sŠ  € ð ˜Ð -°tÐ";Ñ<ð 	[ÝÐYÑZÔZÐZàÐ Ø*.×*;Ò*;¸IÑ*FÔ*FˆMàð 	?˜Ð0Ý*°$´+Ð>Ñ>Ô>ˆOàÐØCRÐC^˜×=Ò=Ñ?Ô?Ð?ÐdeÐÝ œ<¨Ô(;¸AÔ(>À}ÔG[Ð\Ñ\Ô\Ð_oÑoˆLØ'×1Ò1°!Ñ4Ô4ˆLå(Ø”;Ø'Ø)Ø+Ø%ð
ñ 
ô 
ˆð &ˆØ"Ÿošo¨mÈ,˜oÑWÔWÐà!œ[Ð)H¨4¬;Ô+HÐ)HÔIð 		ð 		ˆMØ)˜MØðà*Ø$7Ø)Ø /Ø#ðð ð ðð ˆMˆMð Ÿ	š	 -Ñ0Ô0ˆÝ&Ø+Ø+ð
ñ 
ô 
ð 	
r4   )NNNNNN)rG   rH   rI   r!   r*   r   r   r   r,   re  rK   r   ÚFloatTensorr²   r   r   r   rB   rL   rM   s   @r3   r~  r~  Q  s  ø€ € € € € ð˜~ð ð ð ð ð ð ð   ØØð .2Ø.2Ø04Ø(,Ø26Ø!%ð2
ð 2
àÔ# dÑ*ð2
ð œ tÑ+ð2
ð Ô&¨Ñ-ð	2
ð
  ™ð2
ð Ô(¨4Ñ/ð2
ð ˜$‘;ð2
ð Ð+Ô,ð2
ð 
!ð2
ð 2
ð 2
ñ „^ñ „_ñ  Ôð2
ð 2
ð 2
ð 2
ð 2
r4   r~  c                   ó   ‡ — e Zd ZddiZddiZddgdgfiZddiZˆ fd„Zee		 	 	 	 	 	 	 	 dde
j        d	z  de
j        d	z  de
j        d	z  ded	z  de
j        d	z  de
j        d	z  ded	z  dee
j        z  dee         defd„¦   «         ¦   «         Zˆ xZS )ÚMistral4ForCausalLMzlm_head.weightzmodel.embed_tokens.weightÚlm_headÚcolwise_gather_outputr5   ÚlogitsÚkeep_full_weightc                 óþ   •— t          ¦   «                              |¦  «         t          |¦  «        | _        |j        | _        t          j        |j        |j        d¬¦  «        | _        |  	                    ¦   «          d S rˆ   )
r)   r*   r~  rh  r†  r   rŒ   r1   r�  r�  r¢   s     €r3   r*   zMistral4ForCausalLM.__init__¢  sj   ø€ Ý‰Œ×Ò˜Ñ Ô Ð Ý" 6Ñ*Ô*ˆŒ
Ø Ô+ˆŒÝ”y Ô!3°VÔ5FÈUÐSÑSÔSˆŒð 	�ŠÑÔÐÐÐr4   Nr   r‘  rñ   r}   rE  r’  Úlabelsra  Úlogits_to_keeprj   r'   c	           
      óP  —  | j         d||||||dœ|	¤Ž}
|
j        }t          |t          ¦  «        rt	          | d¦  «        n|}|                      |dd…|dd…f         ¦  «        }d}|� | j        d||| j        j        dœ|	¤Ž}t          |||
j
        |
j        |
j        ¬¦  «        S )aá  
        Example:

        ```python
        >>> from transformers import AutoTokenizer, Mistral4ForCausalLM

        >>> model = Mistral4ForCausalLM.from_pretrained("meta-mistral4/Mistral4-2-7b-hf")
        >>> tokenizer = AutoTokenizer.from_pretrained("meta-mistral4/Mistral4-2-7b-hf")

        >>> prompt = "Hey, are you conscious? Can you talk to me?"
        >>> 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]
        "Hey, are you conscious? Can you talk to me?\nI'm not conscious, but I can talk to you."
        ```)r‘  rñ   r}   rE  r’  ra  N)rŸ  r¢  r†  )ÚlossrŸ  rE  r5   ri  rc  )rh  r”  ru   rV  Úslicer�  Úloss_functionrR   r†  r   rE  r5   ri  )r0   r‘  rñ   r}   rE  r’  r¢  ra  r£  rj   Úoutputsr5   Úslice_indicesrŸ  r¥  s                  r3   rB   zMistral4ForCausalLM.forward«  sô   € ð> ,6¨4¬:ð ,
ØØ)Ø%Ø+Ø'Øð,
ð ,
ð ð,
ð ,
ˆð  Ô1ˆå8BÀ>ÕSVÑ8WÔ8WÐk�˜~˜o¨tÑ4Ô4Ð4Ð]kˆØ—’˜m¨A¨A¨A¨}¸a¸a¸aÐ,?Ô@ÑAÔAˆàˆØÐØ%�4Ô%Ðp¨V¸FÈtÌ{ÔOeÐpÐpÐioÐpÐpˆDå%ØØØ#Ô3Ø!Ô/ØÔ)ð
ñ 
ô 
ð 	
r4   )NNNNNNNr   )rG   rH   rI   Ú_tied_weights_keysÚ_tp_planÚ_pp_planÚ
_fsdp_planr*   r   r   r,   re  rK   r   rš  r²   rV  r   r   r   rB   rL   rM   s   @r3   rœ  rœ  ›  sV  ø€ € € € € à*Ð,GÐHÐØÐ2Ð3€HØ˜_Ð-°¨zÐ:Ð;€HØÐ/Ð0€Jðð ð ð ð ð Øð .2Ø.2Ø04Ø(,Ø26Ø*.Ø!%Ø-.ð6
ð 6
àÔ# dÑ*ð6
ð œ tÑ+ð6
ð Ô&¨Ñ-ð	6
ð
  ™ð6
ð Ô(¨4Ñ/ð6
ð Ô  4Ñ'ð6
ð ˜$‘;ð6
ð ˜eœlÑ*ð6
ð Ð+Ô,ð6
ð 
 ð6
ð 6
ð 6
ñ „^ñ Ôð6
ð 6
ð 6
ð 6
ð 6
r4   rœ  c                   ó   — e Zd ZdS )Ú!Mistral4ForSequenceClassificationN©rG   rH   rI   rc  r4   r3   r¯  r¯  æ  ó   € € € € € Ø€Dr4   r¯  c                   ó   — e Zd ZdS )ÚMistral4ForTokenClassificationNr°  rc  r4   r3   r³  r³  ê  r±  r4   r³  )rg  r~  rœ  r¯  r³  )r¦   )r    r    )r    r  )Or   Úcollections.abcr   r,   Útorch.nn.functionalr   rÊ   r¨   Ú r   rm  Úactivationsr   Úcache_utilsr   r	   Ú
generationr
   Úintegrationsr   r   Úmasking_utilsr   Úmodeling_flash_attention_utilsr   Úmodeling_layersr   r   r   Úmodeling_outputsr   r   Úmodeling_rope_utilsr   r   Úmodeling_utilsr   r   Úprocessing_utilsr   Úutilsr   r   r   Úutils.deprecationr   Úutils.genericr   r   Úutils.output_capturingr   Úconfiguration_mistral4r!   ÚModuler$   rO   r†   r•   r¿   rÞ   rK   rV  rì   rJ   rþ   r  r
  r  r  r  r!  r#  rX  rg  r~  rœ  r¯  r³  Ú__all__rc  r4   r3   ú<module>rÉ     sù  ðð( €€€Ø $Ð $Ð $Ð $Ð $Ð $à €€€Ø Ð Ð Ð Ð Ð Ð Ð Ð Ø Ð Ð Ð Ð Ð à &Ð &Ð &Ð &Ð &Ð &Ø !Ð !Ð !Ð !Ð !Ð !Ø .Ð .Ð .Ð .Ð .Ð .Ð .Ð .Ø )Ð )Ð )Ð )Ð )Ð )Ø SÐ SÐ SÐ SÐ SÐ SÐ SÐ SØ /Ð /Ð /Ð /Ð /Ð /Ø BÐ BÐ BÐ BÐ BÐ Bðð ð ð ð ð ð ð ð ð ð
 PÐ OÐ OÐ OÐ OÐ OÐ OÐ OØ KÐ KÐ KÐ KÐ KÐ KÐ KÐ KØ FÐ FÐ FÐ FÐ FÐ FÐ FÐ FØ &Ð &Ð &Ð &Ð &Ð &Ø IÐ IÐ IÐ IÐ IÐ IÐ IÐ IÐ IÐ IØ 0Ð 0Ð 0Ð 0Ð 0Ð 0Ø GÐ GÐ GÐ GÐ GÐ GÐ GÐ GØ 5Ð 5Ð 5Ð 5Ð 5Ð 5Ø 2Ð 2Ð 2Ð 2Ð 2Ð 2ð Ð˜YÑ'Ô'ðJð Jð Jð Jð J�b”iñ Jô Jñ (Ô'ðJð(6<ð 6<ð 6<ð 6<ð 6<˜bœiñ 6<ô 6<ð 6<ðrð ð ð ð �"”)ñ ô ð ð "9ð "9ð "9ð "9ð "9˜œñ "9ô "9ð "9ðJ ð$#ð $#ð $#ð $#ð $#�b”iñ $#ô $#ñ Ôð$#ðNð ð ð ð �"”)ñ ô ð ð0	U˜Uœ\ð 	U°#ð 	U¸%¼,ð 	Uð 	Uð 	Uð 	Uð& ð%ð %ØŒIð%àŒ<ð%ð 
Œð%ð Œ<ð	%ð
 ”L 4Ñ'ð%ð ð%ð ð%ð Ð'Ô(ð%ð %ð %ð %ð20ð 0ð 0ð 0ðð ð ð(ð (ð (ð ÐÐ-Ñ.Ô.ðð ð ñ /Ô.ðð2$ð $ð $ð $ðN%¨%¬,ð %¸eð %Ð^að %ÐfkÔfrð %ð %ð %ð %ð
K)ð K)ð K)ð K)ð K)˜œ	ñ K)ô K)ð K)ð\,ð ,ð ,ð ,ð ,Ð5ñ ,ô ,ð ,ð^Xð Xð Xð Xð X˜oñ Xô Xð Xð: ðF
ð F
ð F
ð F
ð F
Ð+ñ F
ô F
ñ „ðF
ðR ðG
ð G
ð G
ð G
ð G
Ð1°?ñ G
ô G
ñ „ðG
ðT	ð 	ð 	ð 	ð 	Ð(HÐJañ 	ô 	ð 	ð	ð 	ð 	ð 	ð 	Ð%BÐD[ñ 	ô 	ð 	ðð ð €€€r4   