Ë
    ÛÓ^j°R  ã                   ó6  — d dl Z d dlmZ d dlmZ d dl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 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$m%Z%m&Z& ddl'm(Z( ddl)m*Z*  G d„ dejV                  «      Z, G d„ dejV                  «      Z- G d„ dejV                  «      Z.dej^                  de0dej^                  fd„Z1	 d7dejV                  dej^                  dej^                  d ej^                  d!ej^                  dz  d"e2d#e2d$ee!   fd%„Z3d&„ Z4d8d'„Z5 G d(„ d)ejV                  «      Z6 G d*„ d+e«      Z7e" G d,„ d-e«      «       Z8e" G d.„ d/e8«      «       Z9e" G d0„ d1e8e«      «       Z: G d2„ d3ee8«      Z; G d4„ d5ee8«      Z<g d6¢Z=y)9é    N)ÚCallable)ÚOptionalé   )ÚACT2FN)ÚCacheÚDynamicCache)ÚGenerationMixin)Úcreate_causal_mask)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maybe_autocastÚmerge_with_config_defaults)Úcapture_outputsé   )ÚHeliumConfigc                   ó,   ‡ — e Zd Zdˆ fd„	Zd„ Zd„ Zˆ xZS )ÚHeliumRMSNormc                 óŠ   •— t         ‰| �  «        t        j                  t	        j
                  |«      «      | _        || _        y ©N)ÚsuperÚ__init__ÚnnÚ	ParameterÚtorchÚonesÚweightÚvariance_epsilon)ÚselfÚhidden_sizeÚepsÚ	__class__s      €úu/var/www/ramen.bs-engineer-server.com/venv/lib/python3.12/site-packages/transformers/models/helium/modeling_helium.pyr"   zHeliumRMSNorm.__init__0   s/   ø€ Ü‰ÑÔÜ—l‘l¤5§:¡:¨kÓ#:Ó;ˆŒØ #ˆÕó    c                 ó\  — |j                   }|j                  t        j                  «      }|j	                  d«      j                  dd¬«      }|t        j                  || j                  z   «      z  }| j                  j                  t        j                  «      |z  j                  |«      S )Né   éÿÿÿÿT)Úkeepdim)	ÚdtypeÚtor%   Úfloat32ÚpowÚmeanÚrsqrtr(   r'   )r)   Úhidden_statesÚinput_dtypeÚvariances       r-   ÚforwardzHeliumRMSNorm.forward5   s†   € Ø#×)Ñ)ˆØ%×(Ñ(¬¯©Ó7ˆØ ×$Ñ$ QÓ'×,Ñ,¨R¸Ð,Ó>ˆØ%¬¯©°H¸t×?TÑ?TÑ4TÓ(UÑUˆØ—‘—‘œuŸ}™}Ó-°Ñ=×AÑAÀ+ÓNÐNr.   c                 ó^   — t        | j                  j                  «      › d| j                  › �S )Nz, eps=)Útupler'   Úshaper(   )r)   s    r-   Ú
extra_reprzHeliumRMSNorm.extra_repr<   s*   € Ü˜Ÿ™×)Ñ)Ó*Ð+¨6°$×2GÑ2GÐ1HÐIÐIr.   )g�íµ ÷Æ°>)Ú__name__Ú
__module__Ú__qualname__r"   r<   r@   Ú__classcell__©r,   s   @r-   r   r   /   s   ø„ õ$ò
OöJr.   r   c                   óÄ   ‡ — e Zd ZU ej                  ed<   ddefˆ fd„Ze	 	 	 ddedz  de	d   de
dz  ded	ef   fd
„«       Z ej                  «       ed„ «       «       Zˆ xZS )ÚHeliumRotaryEmbeddingÚinv_freqNÚconfigc                 ó®  •— t         ‰| �  «        |j                  | _        |j                  | _        || _        | j
                  j                  d   | _        | j                  }| j                  dk7  rt        | j                     } || j
                  |«      \  }| _
        | j                  d|d¬«       | j                  d|j                  «       d¬«       y )NÚ	rope_typeÚdefaultrH   F)Ú
persistentÚoriginal_inv_freq)r!   r"   Úmax_position_embeddingsÚmax_seq_len_cachedÚoriginal_max_seq_lenrI   Úrope_parametersrK   Úcompute_default_rope_parametersr   Úattention_scalingÚregister_bufferÚclone)r)   rI   ÚdeviceÚrope_init_fnrH   r,   s        €r-   r"   zHeliumRotaryEmbedding.__init__C   s¶   ø€ Ü‰ÑÔØ"(×"@Ñ"@ˆÔØ$*×$BÑ$BˆÔ!àˆŒàŸ™×4Ñ4°[ÑAˆŒØ!%×!EÑ!EˆØ�>‰>˜YÒ&Ü.¨t¯~©~Ñ>ˆLÙ+7¸¿¹ÀVÓ+LÑ(ˆ�$Ô(à×Ñ˜Z¨¸eÐÔDØ×ÑÐ0°(·.±.Ó2BÈuÐÕUr.   rW   ztorch.deviceÚseq_lenÚreturnztorch.Tensorc                 ó  — | j                   d   }t        | dd«      xs | j                  | j                  z  }d}d|t	        j
                  d|dt        j                  ¬«      j                  |t        j                  ¬«      |z  z  z  }||fS )	a¨  
        Computes the inverse frequencies according to the original RoPE implementation
        Args:
            config ([`~transformers.PreTrainedConfig`]):
                The model configuration.
            device (`torch.device`):
                The device to use for initialization of the inverse frequencies.
            seq_len (`int`, *optional*):
                The current sequence length. Unused for this type of RoPE.
        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_dimNg      ð?r   r0   ©r3   )rW   r3   )	rR   Úgetattrr*   Únum_attention_headsr%   ÚarangeÚint64r4   Úfloat)rI   rW   rY   ÚbaseÚdimÚattention_factorrH   s          r-   rS   z5HeliumRotaryEmbedding.compute_default_rope_parametersS   s�   € ð& ×%Ñ% lÑ3ˆÜ�f˜j¨$Ó/Òc°6×3EÑ3EÈ×IcÑIcÑ3cˆàÐð Ø”U—\‘\ ! S¨!´5·;±;Ô?×BÑBÈ&ÔX]×XcÑXcÐBÓdÐgjÑjÑkñ
ˆð Ð)Ð)Ð)r.   c                 óN  — | j                   d d d …d f   j                  «       j                  |j                  d   dd«      j	                  |j
                  «      }|d d …d d d …f   j                  «       }t        |j
                  j                  t        «      r/|j
                  j                  dk7  r|j
                  j                  nd}t        |d¬«      5  |j                  «       |j                  «       z  j                  dd«      }t        j                  ||fd¬	«      }|j                  «       | j                  z  }|j                  «       | j                  z  }	d d d «       j	                  |j                   ¬
«      	j	                  |j                   ¬
«      fS # 1 sw Y   ŒAxY w)Nr   r1   r   ÚmpsÚcpuF)Údevice_typeÚenabledr0   ©re   r^   )rH   rc   Úexpandr?   r4   rW   Ú
isinstanceÚtypeÚstrr   Ú	transposer%   ÚcatÚcosrT   Úsinr3   )
r)   ÚxÚposition_idsÚinv_freq_expandedÚposition_ids_expandedrj   ÚfreqsÚembrs   rt   s
             r-   r<   zHeliumRotaryEmbedding.forwardq   sR  € ð !ŸM™M¨$²°4¨-Ñ8×>Ñ>Ó@×GÑGÈ×HZÑHZÐ[\ÑH]Ð_aÐcdÓe×hÑhÐij×iqÑiqÓrÐØ ,ªQ°²a¨ZÑ 8× >Ñ >Ó @Ðä'1°!·(±(·-±-ÄÔ'EÈ!Ï(É(Ï-É-Ð[`ÒJ`�a—h‘h—m’mÐfkˆÜ¨¸UÔCñ 	5Ø&×,Ñ,Ó.Ð1F×1LÑ1LÓ1NÑN×YÑYÐZ[Ð]^Ó_ˆEÜ—)‘)˜U E˜N°Ô3ˆCØ—'‘'“)˜d×4Ñ4Ñ4ˆCØ—'‘'“)˜d×4Ñ4Ñ4ˆC÷		5ð �v‰v˜AŸG™GˆvÓ$ c§f¡f°1·7±7 fÓ&;Ð;Ð;÷	5ð 	5ús   ÃBFÆF$r    ©NNN)rA   rB   rC   r%   ÚTensorÚ__annotations__r   r"   Ústaticmethodr   Úintr>   rc   rS   Úno_gradr   r<   rD   rE   s   @r-   rG   rG   @   s˜   ø… Ø�l‰lÓñV˜|õ Vð  à&*Ø+/Ø"ñ*Ø˜tÑ#ð*à˜Ñ(ð*ð �t‘ð*ð 
ˆ~˜uÐ$Ñ	%ò	*ó ð*ð: €U‡]�]ƒ_Øñ<ó ó ô<r.   rG   c                   ó$   ‡ — e Zd Zˆ fd„Zd„ Zˆ xZS )Ú	HeliumMLPc                 ó  •— t         ‰| �  «        || _        |j                  | _        |j                  | _        t        j                  | j                  | j                  |j                  ¬«      | _        t        j                  | j                  | j                  |j                  ¬«      | _	        t        j                  | j                  | j                  |j                  ¬«      | _
        t        |j                     | _        y )N©Úbias)r!   r"   rI   r*   Úintermediate_sizer#   ÚLinearÚmlp_biasÚ	gate_projÚup_projÚ	down_projr   Ú
hidden_actÚact_fn©r)   rI   r,   s     €r-   r"   zHeliumMLP.__init__‚   s¸   ø€ Ü‰ÑÔØˆŒØ!×-Ñ-ˆÔØ!'×!9Ñ!9ˆÔÜŸ™ 4×#3Ñ#3°T×5KÑ5KÐRX×RaÑRaÔbˆŒÜ—y‘y ×!1Ñ!1°4×3IÑ3IÐPV×P_ÑP_Ô`ˆŒÜŸ™ 4×#9Ñ#9¸4×;KÑ;KÐRX×RaÑRaÔbˆŒÜ˜V×.Ñ.Ñ/ˆ�r.   c                 óˆ   — | j                  | j                  | j                  |«      «      | j                  |«      z  «      }|S r    )r‹   r�   r‰   rŠ   )r)   ru   r‹   s      r-   r<   zHeliumMLP.forwardŒ   s6   € Ø—N‘N 4§;¡;¨t¯~©~¸aÓ/@Ó#AÀDÇLÁLÐQRÃOÑ#SÓTˆ	ØÐr.   )rA   rB   rC   r"   r<   rD   rE   s   @r-   r‚   r‚   �   s   ø„ ô0ör.   r‚   r9   Ún_reprZ   c                 óª   — | j                   \  }}}}|dk(  r| S | dd…dd…ddd…dd…f   j                  |||||«      } | j                  |||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)r?   rm   Úreshape)r9   r�   ÚbatchÚnum_key_value_headsÚslenr]   s         r-   Ú	repeat_kvr–   ‘   so   € ð
 2?×1DÑ1DÑ.€EÐ  hØ�‚zØÐØ!¢!¢Q¨ªa²Ð"2Ñ3×:Ñ:¸5ÐBUÐW\Ð^bÐdlÓm€MØ× Ñ  Ð(;¸eÑ(CÀTÈ8ÓTÐTr.   ÚmoduleÚqueryÚkeyÚvalueÚattention_maskÚscalingÚdropoutÚkwargsc                 ó  — t        || j                  «      }t        || j                  «      }	t        j                  ||j	                  dd«      «      |z  }
|�|
|z   }
t
        j                  j                  |
dt        j                  ¬«      j                  |j                  «      }
t
        j                  j                  |
|| j                  ¬«      }
t        j                  |
|	«      }|j	                  dd«      j                  «       }||
fS )Nr0   r   r1   )re   r3   )ÚpÚtrainingr   )r–   Únum_key_value_groupsr%   Úmatmulrq   r#   Ú
functionalÚsoftmaxr5   r4   r3   r�   r¡   Ú
contiguous)r—   r˜   r™   rš   r›   rœ   r�   rž   Ú
key_statesÚvalue_statesÚattn_weightsÚattn_outputs               r-   Úeager_attention_forwardr«   �   sà   € ô ˜3 × ;Ñ ;Ó<€JÜ˜U F×$?Ñ$?Ó@€Lä—<‘<  z×';Ñ';¸A¸qÓ'AÓBÀWÑL€LØÐ!Ø# nÑ4ˆä—=‘=×(Ñ(¨¸2ÄUÇ]Á]Ð(ÓS×VÑVÐW\×WbÑWbÓc€LÜ—=‘=×(Ñ(¨¸È6Ï?É?Ð(Ó[€LÜ—,‘,˜|¨\Ó:€KØ×'Ñ'¨¨1Ó-×8Ñ8Ó:€Kà˜Ð$Ð$r.   c                 ó|   — | dddd…f   }| dddd…f   }t        j                  | |fd¬«      j                  d«      S )	z*Rotates half the hidden dims of the input..r   Nr0   r   r1   rl   éþÿÿÿ)r%   ÚstackÚflatten)ru   Úx1Úx2s      r-   Úrotate_halfr²   ¶   sJ   € à	
ˆ3���1�ˆ9‰€BØ	
ˆ3���1�ˆ9‰€BÜ�;‰;˜˜˜R�y bÔ)×1Ñ1°"Ó5Ð5r.   c                 óF  — |j                  |«      }|j                  |«      }|dd|j                  d   dz  …f   j                  dd¬«      }|dd|j                  d   dz  …f   j                  dd¬«      }| |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.
    .Nr1   r0   rl   )Ú	unsqueezer?   Úrepeat_interleaver²   )ÚqÚkrs   rt   Úunsqueeze_dimÚq_embedÚk_embeds          r-   Úapply_rotary_pos_embr»   ½   s¿   € ð$ �-‰-˜Ó
&€CØ
�-‰-˜Ó
&€Cð ˆcÐ'�S—Y‘Y˜r‘] aÑ'Ð'Ð'Ñ
(×
:Ñ
:¸1À"Ð
:Ó
E€CØ
ˆcÐ'�S—Y‘Y˜r‘] aÑ'Ð'Ð'Ñ
(×
:Ñ
:¸1À"Ð
:Ó
E€Cà�3‰wœ; q›>¨CÑ/Ñ0€GØ�3‰wœ; q›>¨CÑ/Ñ0€Gà�GÐÐr.   c                   ó   ‡ — e Zd ZdZddededz  fˆ fd„Z	 	 	 ddej                  de	ej                  ej                  f   dz  dej                  dz  d	e
dz  d
ee   de	ej                  ej                  f   fd„Zˆ xZS )ÚHeliumAttentionz=Multi-headed attention from 'Attention Is All You Need' paperNrI   Ú	layer_idxc                 ó\  •— t         ‰| �  «        || _        || _        t	        |d|j
                  |j                  z  «      | _        |j                  |j                  z  | _	        dt        j                  | j                  «      z  | _        |j                  | _        d| _        t        j                   |j
                  |j                  | j                  z  |j"                  ¬«      | _        t        j                   |j
                  |j                  | j                  z  |j"                  ¬«      | _        t        j                   |j
                  |j                  | j                  z  |j"                  ¬«      | _        t        j                   |j
                  |j
                  d¬«      | _        y )Nr]   r   Tr„   F)r!   r"   rI   r¾   r_   r*   r`   r]   r”   r¢   ÚmathÚsqrtrœ   Úattention_dropoutÚ	is_causalr#   r‡   Úattention_biasÚq_projÚk_projÚv_projÚo_proj©r)   rI   r¾   r,   s      €r-   r"   zHeliumAttention.__init__ß   sC  ø€ Ü‰ÑÔØˆŒØ"ˆŒÜ ¨
°F×4FÑ4FÈ&×JdÑJdÑ4dÓeˆŒØ$*×$>Ñ$>À&×B\ÑB\Ñ$\ˆÔ!Øœ4Ÿ9™9 T§]¡]Ó3Ñ3ˆŒØ!'×!9Ñ!9ˆÔØˆŒä—i‘iØ×Ñ × :Ñ :¸T¿]¹]Ñ JÐQW×QfÑQfô
ˆŒô —i‘iØ×Ñ × :Ñ :¸T¿]¹]Ñ JÐQW×QfÑQfô
ˆŒô —i‘iØ×Ñ × :Ñ :¸T¿]¹]Ñ JÐQW×QfÑQfô
ˆŒô —i‘i × 2Ñ 2°F×4FÑ4FÈUÔSˆ�r.   r9   Úposition_embeddingsr›   Úpast_key_valuesrž   rZ   c                 ó
  — |j                   d d }g |¢d‘| j                  ‘­}| j                  |«      j                  |«      j	                  dd«      }| j                  |«      j                  |«      j	                  dd«      }	| j                  |«      j                  |«      j	                  dd«      }
|\  }}t        ||	||«      \  }}	|� |j                  |	|
| j                  «      \  }	}
t        j                  | j                  j                  t        «      } || ||	|
|f| j                  sdn| j                   | j"                  dœ|¤Ž\  }} |j$                  g |¢d‘­Ž j'                  «       }| j)                  |«      }||fS )Nr1   r   r0   ç        )r�   rœ   )r?   r]   rÅ   Úviewrq   rÆ   rÇ   r»   Úupdater¾   r   Úget_interfacerI   Ú_attn_implementationr«   r¡   rÂ   rœ   r’   r¦   rÈ   )r)   r9   rÊ   r›   rË   rž   Úinput_shapeÚhidden_shapeÚquery_statesr§   r¨   rs   rt   Úattention_interfacerª   r©   s                   r-   r<   zHeliumAttention.forwardô   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ˆSÜ#7¸ÀjÐRUÐWZÓ#[Ñ ˆ�jàÐ&Ø'6×'=Ñ'=¸jÈ,ÐX\×XfÑXfÓ'gÑ$ˆJ˜ä(?×(MÑ(MØ�K‰K×,Ñ,Ô.Eó)
Ðñ %8ØØØØØð	%
ð  $Ÿ}š}‘C°$×2HÑ2HØ—L‘Lñ	%
ð ñ	%
Ñ!ˆ�\ð *�k×)Ñ)Ð;¨;Ð;¸Ò;×FÑFÓHˆØ—k‘k +Ó.ˆØ˜LÐ(Ð(r.   r    r{   )rA   rB   rC   Ú__doc__r   r   r"   r%   r|   r>   r   r   r   r<   rD   rE   s   @r-   r½   r½   Ü   s¬   ø„ ÙGñT˜|ð T¸¸d¹
õ Tð0 IMØ.2Ø(,ñ&)à—|‘|ð&)ð # 5§<¡<°·±Ð#=Ñ>ÀÑEð&)ð Ÿ™ tÑ+ð	&)ð
  ™ð&)ð Ð+Ñ,ð&)ð 
ˆu�|‰|˜UŸ\™\Ð)Ñ	*÷&)r.   r½   c                   ó
  ‡ — e Zd Zddededz  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HeliumDecoderLayerNrI   r¾   c                 ó   •— t         ‰| �  «        |j                  | _        t        ||¬«      | _        t        |«      | _        t        |j                  |j                  ¬«      | _	        t        |j                  |j                  ¬«      | _
        y )N)rI   r¾   ©r+   )r!   r"   r*   r½   Ú	self_attnr‚   Úmlpr   Úrms_norm_epsÚinput_layernormÚpost_attention_layernormrÉ   s      €r-   r"   zHeliumDecoderLayer.__init__  sl   ø€ Ü‰ÑÔØ!×-Ñ-ˆÔä(°À)ÔLˆŒä˜VÓ$ˆŒÜ,¨V×-?Ñ-?ÀV×EXÑEXÔYˆÔÜ(5°f×6HÑ6HÈf×NaÑNaÔ(bˆÕ%r.   r9   r›   rv   rË   Ú	use_cacherÊ   rž   rZ   c           
      óÀ   — |}| j                  |«      } | j                  d||||||dœ|¤Ž\  }}	||z   }|}| j                  |«      }| j                  |«      }||z   }|S )N)r9   r›   rv   rË   rà   rÊ   © )rÞ   rÛ   rß   rÜ   )
r)   r9   r›   rv   rË   rà   rÊ   rž   ÚresidualÚ_s
             r-   r<   zHeliumDecoderLayer.forward(  s�   € ð !ˆØ×,Ñ,¨]Ó;ˆà)˜4Ÿ>™>ð 
Ø'Ø)Ø%Ø+ØØ 3ñ
ð ñ
Ñˆ�qð ! =Ñ0ˆð !ˆØ×5Ñ5°mÓDˆØŸ™ Ó/ˆØ  =Ñ0ˆØÐr.   r    )NNNFN)rA   rB   rC   r   r   r"   r%   r|   Ú
LongTensorr   Úboolr>   r   r   r<   rD   rE   s   @r-   rØ   rØ     sÂ   ø„ ñc˜|ð c¸¸d¹
õ cð /3Ø04Ø(,Ø!&ØHLñà—|‘|ðð Ÿ™ tÑ+ðð ×&Ñ&¨Ñ-ð	ð
  ™ðð ˜$‘;ðð # 5§<¡<°·±Ð#=Ñ>ÀÑEðð Ð+Ñ,ðð 
�‰÷r.   rØ   c                   óJ   — 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y)ÚHeliumPreTrainedModelrI   ÚmodelTrØ   rË   )r9   Ú
attentionsN)rA   rB   rC   r   r}   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_backendrØ   r½   Ú_can_record_outputsrâ   r.   r-   rè   rè   H  sQ   … àÓØÐØ&*Ð#Ø-Ð.ÐØ#4Ð"5ÐØÐØ€NØÐà!ÐØ"&Ðà+Ø%ñÑr.   rè   c                   óò   ‡ — 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HeliumModelrI   c           	      ó  •— t         ‰| �  |«       |j                  | _        |j                  | _        t        j                  |j                  |j                  | j                  «      | _        t        j                  t        |j                  «      D �cg c]  }t        ||«      ‘Œ c}«      | _        t        |j                  |j                  ¬«      | _        t#        |¬«      | _        d| _        | j)                  «        y c c}w )NrÚ   ©rI   F)r!   r"   Úpad_token_idÚpadding_idxÚ
vocab_sizer#   Ú	Embeddingr*   Úembed_tokensÚ
ModuleListÚrangeÚnum_hidden_layersrØ   Úlayersr   rÝ   ÚnormrG   Ú
rotary_embÚgradient_checkpointingÚ	post_initrÉ   s      €r-   r"   zHeliumModel.__init__]  sÆ   ø€ Ü‰Ñ˜Ô Ø!×.Ñ.ˆÔØ ×+Ñ+ˆŒäŸL™L¨×):Ñ):¸F×<NÑ<NÐPT×P`ÑP`ÓaˆÔÜ—m‘mÜDIÈ&×JbÑJbÓDcÖd°yÔ ¨	Õ2Òdó
ˆŒô " &×"4Ñ"4¸&×:MÑ:MÔNˆŒ	Ü/°vÔ>ˆŒØ&+ˆÔ#ð 	�‰Õùò es   ÂDNÚ	input_idsr›   rv   rË   Úinputs_embedsrà   rž   rZ   c           
      ó<  — |d u |d uz  rt        d«      ‚|€| j                  |«      }|r|€t        | j                  ¬«      }|€V|�|j	                  «       nd}t        j                  |j                  d   |j                  ¬«      |z   }|j                  d«      }t        | j                  ||||¬«      }	|}
| j                  |
|¬«      }| j                  d | j                  j                   D ]  } ||
f|	||||dœ|¤Ž}
Œ | j                  |
«      }
t        |
|¬	«      S )
Nz:You must specify exactly one of input_ids or inputs_embedsrø   r   r   )rW   )rI   r  r›   rË   rv   )rv   )r›   rÊ   rv   rË   rà   )Úlast_hidden_staterË   )Ú
ValueErrorrý   r   rI   Úget_seq_lengthr%   ra   r?   rW   r´   r
   r  r  r   r  r   )r)   r  r›   rv   rË   r  rà   rž   Úpast_seen_tokensÚcausal_maskr9   rÊ   Údecoder_layers                r-   r<   zHeliumModel.forwardm  sL  € ð ˜Ð -°tÐ";Ò<ÜÐYÓZÐZàÐ Ø*.×*;Ñ*;¸IÓ*FˆMá˜Ð0Ü*°$·+±+Ô>ˆOàÐØCRÐC^˜×=Ñ=Ô?ÐdeÐÜ Ÿ<™<¨×(;Ñ(;¸AÑ(>À}×G[ÑG[Ô\Ð_oÑoˆLØ'×1Ñ1°!Ó4ˆLä(Ø—;‘;Ø'Ø)Ø+Ø%ô
ˆð &ˆØ"Ÿo™o¨mÈ,˜oÓWÐà!Ÿ[™[Ð)H¨4¯;©;×+HÑ+HÐIò 		ˆMÙ)Øðà*Ø$7Ø)Ø /Ø#ñð ñ‰Mð		ð Ÿ	™	 -Ó0ˆÜ&Ø+Ø+ô
ð 	
r.   )NNNNNN)rA   rB   rC   r   r"   r   r   r   r%   rå   r|   r   ÚFloatTensorræ   r   r   r   r<   rD   rE   s   @r-   rö   rö   [  sÍ   ø„ ð˜|õ ð   ØØð .2Ø.2Ø04Ø(,Ø26Ø!%ñ2
à×#Ñ# dÑ*ð2
ð Ÿ™ tÑ+ð2
ð ×&Ñ&¨Ñ-ð	2
ð
  ™ð2
ð ×(Ñ(¨4Ñ/ð2
ð ˜$‘;ð2
ð Ð+Ñ,ð2
ð 
!ò2
ó ó ó  ô2
r.   rö   c                   óB  ‡ — e Zd ZddiZddiZddgdgf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HeliumForCausalLMzlm_head.weightzmodel.embed_tokens.weightÚlm_headÚcolwise_gather_outputr9   Úlogitsc                 óè   •— t         ‰| �  |«       t        |«      | _        |j                  | _        t        j                  |j                  |j                  d¬«      | _        | j                  «        y )NFr„   )
r!   r"   rö   ré   rû   r#   r‡   r*   r  r  rŽ   s     €r-   r"   zHeliumForCausalLM.__init__«  sU   ø€ Ü‰Ñ˜Ô Ü  Ó(ˆŒ
Ø ×+Ñ+ˆŒÜ—y‘y ×!3Ñ!3°V×5FÑ5FÈUÔSˆŒð 	�‰Õr.   Nr  r›   rv   rË   r  Úlabelsrà   Úlogits_to_keeprž   rZ   c	           
      óx  —  | j                   d||||||dœ|	¤Ž}
|
j                  }t        |t        «      rt	        | d«      n|}| j                  |dd…|dd…f   «      }d}|�* | j                  d||| j                  j                  dœ|	¤Ž}t        |||
j                  |
j                  |
j                  ¬«      S )a€  
        Example:

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

        >>> model = HeliumForCausalLM.from_pretrained("google/helium-7b")
        >>> tokenizer = AutoTokenizer.from_pretrained("google/helium-7b")

        >>> 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›   rv   rË   r  rà   N)r  r  rû   )Úlossr  rË   r9   rê   râ   )ré   r	  rn   r   Úslicer  Úloss_functionrI   rû   r   rË   r9   rê   )r)   r  r›   rv   rË   r  r  rà   r  rž   Úoutputsr9   Úslice_indicesr  r  s                  r-   r<   zHeliumForCausalLM.forward´  sÙ   € ð> ,6¨4¯:©:ð ,
ØØ)Ø%Ø+Ø'Øñ,
ð ñ,
ˆð  ×1Ñ1ˆä8BÀ>ÔSVÔ8Wœ˜~˜o¨tÔ4Ð]kˆØ—‘˜mªA¨}ºaÐ,?Ñ@ÓAˆàˆØÐØ%�4×%Ñ%Ðp¨V¸FÈtÏ{É{×OeÑOeÑpÐioÑpˆDä%ØØØ#×3Ñ3Ø!×/Ñ/Ø×)Ñ)ô
ð 	
r.   )NNNNNNNr   )rA   rB   rC   Ú_tied_weights_keysÚ_tp_planÚ_pp_planr"   r   r   r%   rå   r|   r   r  ræ   r   r   r   r   r<   rD   rE   s   @r-   r  r  ¥  s  ø„ à*Ð,GÐHÐØÐ2Ð3€HØ˜_Ð-°¨zÐ:Ð;€Hôð Øð .2Ø.2Ø04Ø(,Ø26Ø*.Ø!%Ø-.ñ6
à×#Ñ# dÑ*ð6
ð Ÿ™ tÑ+ð6
ð ×&Ñ&¨Ñ-ð	6
ð
  ™ð6
ð ×(Ñ(¨4Ñ/ð6
ð × Ñ  4Ñ'ð6
ð ˜$‘;ð6
ð ˜eŸl™lÑ*ð6
ð Ð+Ñ,ð6
ð 
 ò6
ó ó ô6
r.   r  c                   ó   — e Zd Zy)ÚHeliumForSequenceClassificationN©rA   rB   rC   râ   r.   r-   r"  r"  ï  ó   „ Ør.   r"  c                   ó   — e Zd Zy)ÚHeliumForTokenClassificationNr#  râ   r.   r-   r&  r&  ó  r$  r.   r&  )rè   rö   r  r"  r&  )rÍ   )r   )>rÀ   Úcollections.abcr   Útypingr   r%   Útorch.nnr#   Úactivationsr   Úcache_utilsr   r   Ú
generationr	   Úmasking_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.genericr   r   Úutils.output_capturingr   Úconfiguration_heliumr   ÚModuler   rG   r‚   r|   r   r–   rc   r«   r²   r»   r½   rØ   rè   rö   r  r"  r&  Ú__all__râ   r.   r-   ú<module>r9     sÅ  ðó* Ý $Ý ã Ý å !ß .Ý )Ý /÷ñ ÷
 Pß Kß FÝ &ß IÑ Iß GÝ 5Ý .ôJ�B—I‘Iô Jô"><˜BŸI™Iô ><ôB�—	‘	ô ð 	U˜UŸ\™\ð 	U°#ð 	U¸%¿,¹,ó 	Uð& ñ%Ø�I‰Ið%à�<‰<ð%ð 
�‰ð%ð �<‰<ð	%ð
 —L‘L 4Ñ'ð%ð ð%ð ð%ð Ð'Ñ(ó%ò26óô>>)�b—i‘iô >)ôB(Ð3ô (ðV ô˜Oó ó ðð$ ôF
Ð'ó F
ó ðF
ðR ôF
Ð-¨ó F
ó ðF
ôR	Ð&FÐH]ô 	ô	Ð#@ÐBWô 	ò�r.   