Ë
    ÜÓ^joS  ã                   ó`  — d dl mZ d dlmZ d dlZd dlmZ d dlmc 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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$m%Z%m&Z&m'Z' ddl(m)Z)m*Z* ddl+m,Z, ddl-m.Z.  G d„ dej^                  «      Z0 G d„ dej^                  «      Z1 G d„ dej^                  «      Z2d„ Z3dejh                  de5dejh                  fd„Z6	 d7dej^                  d ejh                  d!ejh                  d"ejh                  d#ejh                  dz  d$e7d%e7d&e#e%   fd'„Z8 ed(«      d8d)„«       Z9 ee9«       G d*„ d+ej^                  «      «       Z: G d,„ d-e«      Z;e& G d.„ d/e!«      «       Z<e& G d0„ d1e<«      «       Z=e& G d2„ d3e<e«      «       Z> G d4„ d5ee<«      Z?g d6¢Z@y)9é    )ÚCallable)ÚOptionalNé   )ÚACT2FN)ÚCacheÚDynamicCache)ÚGenerationMixin)Úuse_kernel_func_from_hubÚuse_kernelized_func)Úcreate_causal_mask)Ú GenericForSequenceClassificationÚ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
OlmoConfigc                   ód   ‡ — e Zd ZdZdeddfˆ fd„Zdej                  dej                  fd„Zˆ xZ	S )ÚOlmoLayerNormz/LayerNorm but with no learnable weight or bias.Úhidden_sizeÚreturnNc                 ó2   •— t         ‰| �  «        |f| _        y ©N)ÚsuperÚ__init__Únormalized_shape)Úselfr    Ú	__class__s     €úq/var/www/ramen.bs-engineer-server.com/venv/lib/python3.12/site-packages/transformers/models/olmo/modeling_olmo.pyr%   zOlmoLayerNorm.__init__4   s   ø€ Ü‰ÑÔØ!, ˆÕó    Úhidden_statesc                 ó¼   — |j                   }t        j                  |j                  t        j
                  ¬«      | j                  d d d¬«      j                  |«      S )N©Údtypegñhãˆµøä>)Úeps)r.   ÚFÚ
layer_normÚtoÚtorchÚfloat32r&   )r'   r+   Ú
orig_dtypes      r)   ÚforwardzOlmoLayerNorm.forward8   sO   € Ø"×(Ñ(ˆ
Ü�|‰|˜M×,Ñ,´5·=±=Ð,ÓAÀ4×CXÑCXÐZ^Ð`dÐjnÔo×rÑrØó
ð 	
r*   )
Ú__name__Ú
__module__Ú__qualname__Ú__doc__Úintr%   r3   ÚTensorr6   Ú__classcell__©r(   s   @r)   r   r   1   s4   ø„ Ù9ð/ Cð /¨Dõ /ð
 U§\¡\ð 
°e·l±l÷ 
r*   r   c                   ó$   ‡ — e Zd Zˆ fd„Zd„ Zˆ xZS )ÚOlmoMLPc                 óÊ  •— t         ‰| �  «        || _        |j                  | _        |j                  | _        t        j                  | j                  | j                  d¬«      | _        t        j                  | j                  | j                  d¬«      | _        t        j                  | j                  | j                  d¬«      | _	        t        |j                     | _        y ©NF©Úbias)r$   r%   Úconfigr    Úintermediate_sizeÚnnÚLinearÚ	gate_projÚup_projÚ	down_projr   Ú
hidden_actÚact_fn©r'   rE   r(   s     €r)   r%   zOlmoMLP.__init__@   s¦   ø€ Ü‰ÑÔØˆŒØ!×-Ñ-ˆÔØ!'×!9Ñ!9ˆÔÜŸ™ 4×#3Ñ#3°T×5KÑ5KÐRWÔXˆŒÜ—y‘y ×!1Ñ!1°4×3IÑ3IÐPUÔVˆŒÜŸ™ 4×#9Ñ#9¸4×;KÑ;KÐRWÔXˆŒÜ˜V×.Ñ.Ñ/ˆ�r*   c                 óˆ   — | j                  | j                  | j                  |«      «      | j                  |«      z  «      }|S r#   )rK   rM   rI   rJ   )r'   ÚxrK   s      r)   r6   zOlmoMLP.forwardJ   s6   € Ø—N‘N 4§;¡;¨t¯~©~¸aÓ/@Ó#AÀDÇLÁLÐQRÃOÑ#SÓTˆ	ØÐr*   )r7   r8   r9   r%   r6   r=   r>   s   @r)   r@   r@   ?   s   ø„ ô0ö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OlmoRotaryEmbeddingÚinv_freqNrE   c                 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defaultrS   F)Ú
persistentÚoriginal_inv_freq)r$   r%   Úmax_position_embeddingsÚmax_seq_len_cachedÚoriginal_max_seq_lenrE   Úrope_parametersrU   Úcompute_default_rope_parametersr   Úattention_scalingÚregister_bufferÚclone)r'   rE   ÚdeviceÚrope_init_fnrS   r(   s        €r)   r%   zOlmoRotaryEmbedding.__init__R   s¶   ø€ Ü‰ÑÔØ"(×"@Ñ"@ˆÔØ$*×$BÑ$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�$Ô(à×Ñ˜Z¨¸eÐÔDØ×ÑÐ0°(·.±.Ó2BÈuÐÕUr*   ra   ztorch.deviceÚseq_lenr!   ztorch.Tensorc                 ó  — | j                   d   }t        | dd«      xs | j                  | j                  z  }d}d|t	        j
                  d|dt        j                  ¬«      j                  |t        j                  ¬«      |z  z  z  }||fS )	a¨  
        Computes the inverse frequencies according to the original RoPE implementation
        Args:
            config ([`~transformers.PreTrainedConfig`]):
                The model configuration.
            device (`torch.device`):
                The device to use for initialization of the inverse frequencies.
            seq_len (`int`, *optional*):
                The current sequence length. Unused for this type of RoPE.
        Returns:
            Tuple of (`torch.Tensor`, `float`), containing the inverse frequencies for the RoPE embeddings and the
            post-processing scaling factor applied to the computed cos/sin (unused in this type of RoPE).
        Ú
rope_thetaÚhead_dimNg      ð?r   é   r-   )ra   r.   )	r\   Úgetattrr    Únum_attention_headsr3   ÚarangeÚint64r2   Úfloat)rE   ra   rc   ÚbaseÚdimÚattention_factorrS   s          r)   r]   z3OlmoRotaryEmbedding.compute_default_rope_parametersb   s�   € ð& ×%Ñ% lÑ3ˆÜ�f˜j¨$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                 óì  — | 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 «       ||	fS # 1 sw Y   	fS xY w)
Nr   éÿÿÿÿr   ÚmpsÚcpuF)Údevice_typeÚenabledrg   ©rn   )rS   rl   ÚexpandÚshaper2   ra   Ú
isinstanceÚtypeÚstrr   Ú	transposer3   ÚcatÚcosr^   Úsin)
r'   rP   Úposition_idsÚinv_freq_expandedÚposition_ids_expandedrt   ÚfreqsÚembr~   r   s
             r)   r6   zOlmoRotaryEmbedding.forward€   s8  € ð !ŸM™M¨$²°4¨-Ñ8×>Ñ>Ó@×GÑGÈ×HZÑHZÐ[\ÑH]Ð_aÐcdÓe×hÑhÐij×iqÑiqÓrÐØ ,ªQ°²a¨ZÑ 8× >Ñ >Ó @Ðä'1°!·(±(·-±-ÄÔ'EÈ!Ï(É(Ï-É-Ð[`ÒJ`�a—h‘h—m’mÐfkˆÜ¨¸UÔCñ 	5Ø&×,Ñ,Ó.Ð1F×1LÑ1LÓ1NÑN×YÑYÐZ[Ð]^Ó_ˆEÜ—)‘)˜U E˜N°Ô3ˆCØ—'‘'“)˜d×4Ñ4Ñ4ˆCØ—'‘'“)˜d×4Ñ4Ñ4ˆC÷		5ð
 �Cˆxˆ÷	5ð
 �Cˆxˆús   ÃBE'Å'E3r#   )NNN)r7   r8   r9   r3   r<   Ú__annotations__r   r%   Ústaticmethodr   r;   Útuplerl   r]   Úno_gradr   r6   r=   r>   s   @r)   rR   rR   O   s˜   ø… Ø�l‰lÓñV˜zõ Vð  à$(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*   rR   c                 óš   — | 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..Nrq   rg   rv   )rx   r3   r}   )rP   Úx1Úx2s      r)   Úrotate_halfrŒ   �   sZ   € à	
ˆ3Ð"�!—'‘'˜"‘+ Ñ"Ð"Ð"Ñ	#€BØ	
ˆ3�—‘˜‘˜qÑ Ñ"Ð"Ñ	#€BÜ�9‰9�r�c˜2�Y BÔ'Ð'r*   r+   Ún_repr!   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)rx   rw   Úreshape)r+   r�   ÚbatchÚnum_key_value_headsÚslenrf   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 )Nrg   r   rq   )rn   r.   )ÚpÚtrainingr   )r“   Únum_key_value_groupsr3   Úmatmulr|   rG   Ú
functionalÚsoftmaxr4   r2   r.   rš   rž   Ú
contiguous)r”   r•   r–   r—   r˜   r™   rš   r›   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*   Úrotary_pos_embc                 ó
  — | j                   |j                   }}|j                  |«      }|j                  |«      }| |z  t        | «      |z  z   }||z  t        |«      |z  z   }|j                  |«      |j                  |«      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.   Ú	unsqueezerŒ   r2   )	ÚqÚkr~   r   Úunsqueeze_dimÚq_typeÚk_typeÚq_embedÚk_embeds	            r)   Úapply_rotary_pos_embr³   »   s|   € ð& —W‘W˜aŸg™gˆF€FØ
�-‰-˜Ó
&€CØ
�-‰-˜Ó
&€CØ�3‰wœ; q›>¨CÑ/Ñ0€GØ�3‰wœ; q›>¨CÑ/Ñ0€GØ�:‰:�fÓ˜wŸz™z¨&Ó1Ð1Ð1r*   c                   óê   ‡ — e Zd ZdZdedefˆ fd„Z	 ddej                  de	ej                  ej                  f   dej                  dz  d	e
dz  d
e	ej                  ej                  dz  f   f
d„Zˆ xZS )ÚOlmoAttentionz=Multi-headed attention from 'Attention Is All You Need' paperrE   Ú	layer_idxc                 ód  •— t         ‰| �  «        || _        || _        t	        |d|j
                  |j                  z  «      | _        |j                  |j                  z  | _	        | j                  d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                  z  |j
                  |j                  ¬«      | _        y )Nrf   g      à¿TrC   )r$   r%   rE   r¶   rh   r    ri   rf   r‘   rŸ   r™   Úattention_dropoutÚ	is_causalrG   rH   Úattention_biasÚq_projÚk_projÚv_projÚo_proj©r'   rE   r¶   r(   s      €r)   r%   zOlmoAttention.__init__Ú   sM  ø€ Ü‰ÑÔØˆŒØ"ˆŒÜ ¨
°F×4FÑ4FÈ&×JdÑJdÑ4dÓeˆŒØ$*×$>Ñ$>À&×B\ÑB\Ñ$\ˆÔ!Ø—}‘} dÑ*ˆŒØ!'×!9Ñ!9ˆÔØˆŒä—i‘iØ×Ñ × :Ñ :¸T¿]¹]Ñ JÐQW×QfÑQfô
ˆŒô —i‘iØ×Ñ × :Ñ :¸T¿]¹]Ñ JÐQW×QfÑQfô
ˆŒô —i‘iØ×Ñ × :Ñ :¸T¿]¹]Ñ JÐQW×QfÑQfô
ˆŒô —i‘iØ×&Ñ&¨¯©Ñ6¸×8JÑ8JÐQW×QfÑQfô
ˆ�r*   Nr+   Úposition_embeddingsr˜   Úpast_key_valuesr!   c                 óª  — |j                   d d }g |¢d‘| j                  ‘­}| j                  |«      }| j                  |«      }	| j	                  |«      }
| j
                  j                  �´|j                  | j
                  j                   | j
                  j                  ¬«       |	j                  | j
                  j                   | j
                  j                  ¬«       |
j                  | j
                  j                   | j
                  j                  ¬«       |j                  |«      j                  dd«      }|	j                  |«      j                  dd«      }	|
j                  |«      j                  dd«      }
|\  }}t        ||	||«      \  }}	|� |j                  |	|
| j                  «      \  }	}
t        j                  | j
                  j                  t         «      } || ||	|
|f| j"                  sdn| j$                  | j&                  dœ|¤Ž\  }} |j(                  g |¢d‘­Ž j+                  «       }| j-                  |«      }||fS )Nrq   )ÚminÚmaxr   rg   ç        )rš   r™   )rx   rf   r»   r¼   r½   rE   Úclip_qkvÚclamp_Úviewr|   r³   Úupdater¶   r   Úget_interfaceÚ_attn_implementationr¨   rž   r¸   r™   r�   r£   r¾   )r'   r+   rÀ   r˜   rÁ   r›   Úinput_shapeÚhidden_shapeÚquery_statesr¤   r¥   r~   r   Úattention_interfacer§   r¦   s                   r)   r6   zOlmoAttention.forwardñ   s)  € ð $×)Ñ)¨#¨2Ð.ˆØ8˜Ð8 bÐ8¨$¯-©-Ñ8ˆà—{‘{ =Ó1ˆØ—[‘[ Ó/ˆ
Ø—{‘{ =Ó1ˆà�;‰;×ÑÐ+Ø×Ñ T§[¡[×%9Ñ%9Ð$9¸t¿{¹{×?SÑ?SÐÔTØ×Ñ 4§;¡;×#7Ñ#7Ð"7¸T¿[¹[×=QÑ=QÐÔRØ×Ñ T§[¡[×%9Ñ%9Ð$9¸t¿{¹{×?SÑ?SÐÔTà#×(Ñ(¨Ó6×@Ñ@ÀÀAÓFˆØ—_‘_ \Ó2×<Ñ<¸QÀÓBˆ
Ø#×(Ñ(¨Ó6×@Ñ@ÀÀAÓFˆ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#   )r7   r8   r9   r:   r   r;   r%   r3   r<   r‡   r   r6   r=   r>   s   @r)   rµ   rµ   Ö   s�   ø„ áGð
˜zð 
°cõ 
ð8 )-ñ/)à—|‘|ð/)ð # 5§<¡<°·±Ð#=Ñ>ð/)ð Ÿ™ tÑ+ð	/)ð
  ™ð/)ð 
ˆu�|‰|˜UŸ\™\¨DÑ0Ð0Ñ	1÷/)r*   rµ   c                   ó  ‡ — 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OlmoDecoderLayerrE   r¶   c                 óð   •— t         ‰| �  «        |j                  | _        t        ||¬«      | _        t        |«      | _        t        |j                  «      | _        t        |j                  «      | _	        y )N)rE   r¶   )
r$   r%   r    rµ   Ú	self_attnr@   Úmlpr   Úinput_layernormÚpost_attention_layernormr¿   s      €r)   r%   zOlmoDecoderLayer.__init__$  s[   ø€ Ü‰ÑÔØ!×-Ñ-ˆÔÜ&¨fÀ	ÔJˆŒä˜6“?ˆŒÜ,¨V×-?Ñ-?Ó@ˆÔÜ(5°f×6HÑ6HÓ(IˆÕ%r*   Nr+   r˜   r€   rÁ   Ú	use_cacherÀ   r›   r!   c           
      óÀ   — |}| j                  |«      } | j                  d||||||dœ|¤Ž\  }}	||z   }|}| j                  |«      }| j                  |«      }||z   }|S )N)r+   r˜   r€   rÁ   r×   rÀ   © )rÕ   rÓ   rÖ   rÔ   )
r'   r+   r˜   r€   rÁ   r×   rÀ   r›   ÚresidualÚ_s
             r)   r6   zOlmoDecoderLayer.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*   )NNNFN)r7   r8   r9   r   r;   r%   r3   r<   Ú
LongTensorr   Úboolr‡   r   r   r6   r=   r>   s   @r)   rÑ   rÑ   #  s¾   ø„ ðJ˜zð J°cõ Jð /3Ø04Ø(,Ø!&ØHLñà—|‘|ðð Ÿ™ tÑ+ðð ×&Ñ&¨Ñ-ð	ð
  ™ðð ˜$‘;ðð # 5§<¡<°·±Ð#=Ñ>ÀÑEðð Ð+Ñ,ðð 
�‰÷r*   rÑ   c                   óJ   — e Zd ZU eed<   dZdZdgZdgZdZ	dZ
dZdZdZeedœZy)ÚOlmoPreTrainedModelrE   ÚmodelTrÑ   rÁ   )r+   Ú
attentionsN)r7   r8   r9   r   r…   Úbase_model_prefixÚsupports_gradient_checkpointingÚ_no_split_modulesÚ_skip_keys_device_placementÚ_supports_flash_attnÚ_supports_sdpaÚ_supports_flex_attnÚ_can_compile_fullgraphÚ_supports_attention_backendrÑ   rµ   Ú_can_record_outputsrÙ   r*   r)   rß   rß   M  sQ   … àÓØÐØ&*Ð#Ø+Ð,ÐØ#4Ð"5ÐØÐØ€NØÐà!ÐØ"&Ðà)Ø#ñÑr*   rß   c                   óò   ‡ — e Zd Zdefˆ fd„Zeee	 	 	 	 	 	 ddej                  dz  dej                  dz  dej                  dz  dedz  dej                  dz  d	edz  d
ee   defd„«       «       «       Zˆ xZS )Ú	OlmoModelrE   c           	      óô  •— t         ‰| �  |«       |j                  | _        |j                  | _        t        j                  |j                  |j                  | j                  «      | _        t        j                  t        |j                  «      D �cg c]  }t        ||«      ‘Œ c}«      | _        t        |j                  «      | _        t!        |¬«      | _        d| _        | j'                  «        y c c}w )N©rE   F)r$   r%   Úpad_token_idÚpadding_idxÚ
vocab_sizerG   Ú	Embeddingr    Úembed_tokensÚ
ModuleListÚrangeÚnum_hidden_layersrÑ   Úlayersr   ÚnormrR   Ú
rotary_embÚgradient_checkpointingÚ	post_initr¿   s      €r)   r%   zOlmoModel.__init__b  s¾   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‘mÜBGÈ×H`ÑH`ÓBaÖb°YÔ˜f iÕ0Òbó
ˆŒô " &×"4Ñ"4Ó5ˆŒ	Ü-°VÔ<ˆŒØ&+ˆÔ#ð 	�‰Õùò cs   ÂC5NÚ	input_idsr˜   r€   rÁ   Úinputs_embedsr×   r›   r!   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_embedsrï   r   r   )ra   )rE   rþ   r˜   rÁ   r€   )r€   )r˜   rÀ   r€   rÁ   r×   )Úlast_hidden_staterÁ   )Ú
ValueErrorrô   r   rE   Úget_seq_lengthr3   rj   rx   ra   r«   r   rú   rø   r÷   rù   r   )r'   rý   r˜   r€   rÁ   rþ   r×   r›   Úpast_seen_tokensÚcausal_maskr+   rÀ   Údecoder_layers                r)   r6   zOlmoModel.forwardr  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$7Ø)Ø /Ø#ñð ñ‰Mð		ð Ÿ	™	 -Ó0ˆÜ&Ø+Ø+ô
ð 	
r*   )NNNNNN)r7   r8   r9   r   r%   r   r   r   r3   rÜ   r<   r   ÚFloatTensorrÝ   r   r   r   r6   r=   r>   s   @r)   rí   rí   `  sÍ   ø„ ð˜zõ ð   ØØð .2Ø.2Ø04Ø(,Ø26Ø!%ñ2
à×#Ñ# dÑ*ð2
ð Ÿ™ tÑ+ð2
ð ×&Ñ&¨Ñ-ð	2
ð
  ™ð2
ð ×(Ñ(¨4Ñ/ð2
ð ˜$‘;ð2
ð Ð+Ñ,ð2
ð 
!ò2
ó ó ó  ô2
r*   rí   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OlmoForCausalLMzlm_head.weightzmodel.embed_tokens.weightÚlm_headÚcolwise_gather_outputr+   Úlogitsc                 óè   •— t         ‰| �  |«       t        |«      | _        |j                  | _        t        j                  |j                  |j                  d¬«      | _        | j                  «        y rB   )
r$   r%   rí   rà   rò   rG   rH   r    r	  rü   rN   s     €r)   r%   zOlmoForCausalLM.__init__°  sU   ø€ Ü‰Ñ˜Ô Ü˜vÓ&ˆŒ
Ø ×+Ñ+ˆŒÜ—y‘y ×!3Ñ!3°V×5FÑ5FÈUÔSˆŒð 	�‰Õr*   Nrý   r˜   r€   rÁ   rþ   Úlabelsr×   Úlogits_to_keepr›   r!   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 )aÉ  
        Example:

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

        >>> model = OlmoForCausalLM.from_pretrained("meta-olmo/Olmo-2-7b-hf")
        >>> tokenizer = AutoTokenizer.from_pretrained("meta-olmo/Olmo-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."
        ```)rý   r˜   r€   rÁ   rþ   r×   N)r  r  rò   )Úlossr  rÁ   r+   rá   rÙ   )rà   r   ry   r;   Úslicer	  Úloss_functionrE   rò   r   rÁ   r+   rá   )r'   rý   r˜   r€   rÁ   rþ   r  r×   r  r›   Úoutputsr+   Úslice_indicesr  r  s                  r)   r6   zOlmoForCausalLM.forward¹  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   )r7   r8   r9   Ú_tied_weights_keysÚ_tp_planÚ_pp_planr%   r   r   r3   rÜ   r<   r   r  rÝ   r;   r   r   r   r6   r=   r>   s   @r)   r  r  ª  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 ˜$‘;ð6
ð ˜eŸl™lÑ*ð6
ð Ð+Ñ,ð6
ð 
 ò6
ó ó ô6
r*   r  c                   ó   — e Zd Zy)ÚOlmoForSequenceClassificationN)r7   r8   r9   rÙ   r*   r)   r  r  ô  s   „ Ør*   r  )r  r  rí   rß   )rÅ   )r   )AÚcollections.abcr   Útypingr   r3   Útorch.nnrG   Útorch.nn.functionalr¡   r0   Úactivationsr   Úcache_utilsr   r   Ú
generationr	   Úintegrationsr
   r   Úmasking_utilsr   Úmodeling_layersr   r   Úmodeling_outputsr   r   Úmodeling_rope_utilsr   r   Úmodeling_utilsr   r   Úprocessing_utilsr   Úutilsr   r   r   Úutils.genericr   r   Úutils.output_capturingr   Úconfiguration_olmor   ÚModuler   r@   rR   rŒ   r<   r;   r“   rl   r¨   r³   rµ   rÑ   rß   rí   r  r  Ú__all__rÙ   r*   r)   ú<module>r.     sÝ  ðõ4 %Ý ã Ý ß Ð å !ß .Ý )ß IÝ /ß [ß Oß Kß FÝ &ß IÑ Iß GÝ 5Ý *ô
�B—I‘Iô 
ôˆb�i‰iô ô =˜"Ÿ)™)ô =ò@(ð	U˜UŸ\™\ð 	U°#ð 	U¸%¿,¹,ó 	Uð& ñ%Ø�I‰Ið%à�<‰<ð%ð 
�‰ð%ð �<‰<ð	%ð
 —L‘L 4Ñ'ð%ð ð%ð ð%ð Ð'Ñ(ó%ñ2 Ð*Ó+ò2ó ,ð2ñ4 Ð)Ó*ôI)�B—I‘Ió I)ó +ðI)ôX'Ð1ô 'ðT ô˜/ó ó ðð$ ôF
Ð#ó F
ó ðF
ðR ôF
Ð)¨?ó F
ó ðF
ôR	Ð$DÐFYô 	ò c�r*   