Ë
    ÜÓ^j#U  ã                   óp  — 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
 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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$m%Z%m&Z& ddl'm(Z(m)Z) ddl*m+Z+ ddl,m-Z-  ed«       G d„ dej\                  «      «       Z/ G d„ dej\                  «      Z0dejb                  de2dejb                  fd„Z3	 d9dej\                  dejb                  d ejb                  d!ejb                  d"ejb                  dz  d#e4d$e4d%e#e   fd&„Z5 ed'«      d:d(„«       Z6d)„ Z7 ee6«       G d*„ d+ej\                  «      «       Z8 G d,„ d-ej\                  «      Z9 G d.„ d/e«      Z:e% G d0„ d1e!«      «       Z;e% G d2„ d3e;«      «       Z<e% G d4„ d5e;e«      «       Z= G d6„ d7ee;«      Z>g d8¢Z?y);é    )ÚCallable)ÚOptionalN)ÚTransformersKwargsé   )ÚACT2FN)ÚCacheÚDynamicCache)ÚGenerationMixin)Úuse_kernel_forward_from_hubÚ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auto_docstringÚcan_return_tuple)Úmaybe_autocastÚmerge_with_config_defaults)Úcapture_outputsé   )ÚOlmo2ConfigÚRMSNormc                   óP   ‡ — e Zd Zddeddfˆ fd„Zdej                  fd„Zd„ Zˆ xZ	S )ÚOlmo2RMSNormÚepsÚreturnNc                 óŠ   •— t         ‰| �  «        t        j                  t	        j
                  |«      «      | _        || _        y)z;
        Olmo2RMSNorm is equivalent to T5LayerNorm
        N)ÚsuperÚ__init__ÚnnÚ	ParameterÚtorchÚonesÚweightÚvariance_epsilon)ÚselfÚhidden_sizer"   Ú	__class__s      €ús/var/www/ramen.bs-engineer-server.com/venv/lib/python3.12/site-packages/transformers/models/olmo2/modeling_olmo2.pyr&   zOlmo2RMSNorm.__init__4   s1   ø€ ô 	‰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                  |z  j                  |«      S )Né   éÿÿÿÿT)Úkeepdim)	ÚdtypeÚtor)   Úfloat32ÚpowÚmeanÚrsqrtr,   r+   )r-   Úhidden_statesÚinput_dtypeÚvariances       r0   ÚforwardzOlmo2RMSNorm.forward<   sy   € Ø#×)Ñ)ˆØ%×(Ñ(¬¯©Ó7ˆØ ×$Ñ$ QÓ'×,Ñ,¨R¸Ð,Ó>ˆØ%¬¯©°H¸t×?TÑ?TÑ4TÓ(UÑUˆØ—‘˜mÑ+×/Ñ/°Ó<Ð<r1   c                 ó^   — t        | j                  j                  «      › d| j                  › �S )Nz, eps=)Útupler+   Úshaper,   )r-   s    r0   Ú
extra_reprzOlmo2RMSNorm.extra_reprC   s*   € Ü˜Ÿ™×)Ñ)Ó*Ð+¨6°$×2GÑ2GÐ1HÐIÐIr1   )g�íµ ÷Æ°>)
Ú__name__Ú
__module__Ú__qualname__Úfloatr&   r)   ÚTensorr?   rC   Ú__classcell__©r/   s   @r0   r!   r!   2   s,   ø„ ñ$¨ð $¸$õ $ð=¨¯©ó =öJr1   r!   c                   óÄ   ‡ — e Zd ZU ej                  ed<   ddefˆ fd„Ze	 	 	 ddedz  de	d   de
dz  ded	ef   fd
„«       Z ej                  «       ed„ «       «       Zˆ xZS )ÚOlmo2RotaryEmbeddingÚinv_freqNÚconfigc                 ó®  •— t         ‰| �  «        |j                  | _        |j                  | _        || _        | j
                  j                  d   | _        | j                  }| j                  dk7  rt        | j                     } || j
                  |«      \  }| _
        | j                  d|d¬«       | j                  d|j                  «       d¬«       y )NÚ	rope_typeÚdefaultrM   F)Ú
persistentÚoriginal_inv_freq)r%   r&   Úmax_position_embeddingsÚmax_seq_len_cachedÚoriginal_max_seq_lenrN   Úrope_parametersrP   Úcompute_default_rope_parametersr   Úattention_scalingÚregister_bufferÚclone)r-   rN   ÚdeviceÚrope_init_fnrM   r/   s        €r0   r&   zOlmo2RotaryEmbedding.__init__J   s¶   ø€ Ü‰ÑÔØ"(×"@Ñ"@ˆÔØ$*×$BÑ$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�$Ô(à×Ñ˜Z¨¸eÐÔDØ×ÑÐ0°(·.±.Ó2BÈuÐÕUr1   r\   ztorch.deviceÚseq_lenr#   ztorch.Tensorc                 ó  — | j                   d   }t        | dd«      xs | j                  | j                  z  }d}d|t	        j
                  d|dt        j                  ¬«      j                  |t        j                  ¬«      |z  z  z  }||fS )	a¨  
        Computes the inverse frequencies according to the original RoPE implementation
        Args:
            config ([`~transformers.PreTrainedConfig`]):
                The model configuration.
            device (`torch.device`):
                The device to use for initialization of the inverse frequencies.
            seq_len (`int`, *optional*):
                The current sequence length. Unused for this type of RoPE.
        Returns:
            Tuple of (`torch.Tensor`, `float`), containing the inverse frequencies for the RoPE embeddings and the
            post-processing scaling factor applied to the computed cos/sin (unused in this type of RoPE).
        Ú
rope_thetaÚhead_dimNg      ð?r   r3   )r6   )r\   r6   )	rW   Úgetattrr.   Únum_attention_headsr)   ÚarangeÚint64r7   rG   )rN   r\   r^   ÚbaseÚdimÚattention_factorrM   s          r0   rX   z4Olmo2RotaryEmbedding.compute_default_rope_parametersZ   s�   € ð& ×%Ñ% lÑ3ˆÜ�f˜j¨$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)r1   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   r4   r   ÚmpsÚcpuF)Údevice_typeÚenabledr3   ©rg   )rM   rG   ÚexpandrB   r7   r\   Ú
isinstanceÚtypeÚstrr   Ú	transposer)   ÚcatÚcosrY   Úsin)
r-   ÚxÚposition_idsÚinv_freq_expandedÚposition_ids_expandedrl   ÚfreqsÚembru   rv   s
             r0   r?   zOlmo2RotaryEmbedding.forwardx   s8  € ð !ŸM™M¨$²°4¨-Ñ8×>Ñ>Ó@×GÑGÈ×HZÑHZÐ[\ÑH]Ð_aÐcdÓe×hÑhÐij×iqÑiqÓrÐØ ,ªQ°²a¨ZÑ 8× >Ñ >Ó @Ðä'1°!·(±(·-±-ÄÔ'EÈ!Ï(É(Ï-É-Ð[`ÒJ`�a—h‘h—m’mÐfkˆÜ¨¸UÔCñ 	5Ø&×,Ñ,Ó.Ð1F×1LÑ1LÓ1NÑN×YÑYÐZ[Ð]^Ó_ˆEÜ—)‘)˜U E˜N°Ô3ˆCØ—'‘'“)˜d×4Ñ4Ñ4ˆCØ—'‘'“)˜d×4Ñ4Ñ4ˆC÷		5ð
 �Cˆxˆ÷	5ð
 �Cˆxˆús   ÃBE'Å'E3©N)NNN)rD   rE   rF   r)   rH   Ú__annotations__r   r&   Ústaticmethodr   ÚintrA   rG   rX   Úno_gradr   r?   rI   rJ   s   @r0   rL   rL   G   s˜   ø… Ø�l‰lÓñV˜{õ Vð  à%)Ø+/Ø"ñ*Ø˜dÑ"ð*à˜Ñ(ð*ð �t‘ð*ð 
ˆ~˜uÐ$Ñ	%ò	*ó ð*ð: €U‡]�]ƒ_Øñ
ó ó ô
r1   rL   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)rB   ro   Úreshape)r<   r‚   ÚbatchÚnum_key_value_headsÚslenra   s         r0   Ú	repeat_kvrˆ   ‡   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Tr1   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 )Nr3   r   r4   )rg   r6   )ÚpÚtrainingr   )rˆ   Únum_key_value_groupsr)   Úmatmulrs   r'   Ú
functionalÚsoftmaxr8   r7   r6   r�   r“   Ú
contiguous)r‰   rŠ   r‹   rŒ   r�   rŽ   r�   r�   Ú
key_statesÚvalue_statesÚattn_weightsÚattn_outputs               r0   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$Ð$r1   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.
    )r6   Ú	unsqueezeÚrotate_halfr7   )	ÚqÚkru   rv   Úunsqueeze_dimÚq_typeÚk_typeÚq_embedÚk_embeds	            r0   Úapply_rotary_pos_embr©   ¬   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1r1   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..Nr4   r3   rn   )rB   r)   rt   )rw   Úx1Úx2s      r0   r¡   r¡   Ç   sZ   € à	
ˆ3Ð"�!—'‘'˜"‘+ Ñ"Ð"Ð"Ñ	#€BØ	
ˆ3�—‘˜‘˜qÑ Ñ"Ð"Ñ	#€BÜ�9‰9�r�c˜2�Y BÔ'Ð'r1   c                   óü   ‡ — 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ej                  dz  d	e
dz  d
ee   de	ej                  ej                  dz  f   fd„Zˆ xZS )ÚOlmo2Attentionz=Multi-headed attention from 'Attention Is All You Need' paperNrN   Ú	layer_idxc                 ó,  •— 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                  ¬«      | _        t)        |j                  | j                  z  |j*                  «      | _        t)        |j                  | j                  z  |j*                  «      | _        y )Nra   g      à¿T©Úbias)r%   r&   rN   r¯   rb   r.   rc   ra   r†   r”   rŽ   Úattention_dropoutÚ	is_causalr'   ÚLinearÚattention_biasÚq_projÚk_projÚv_projÚo_projr!   Úrms_norm_epsÚq_normÚk_norm©r-   rN   r¯   r/   s      €r0   r&   zOlmo2Attention.__init__Ò   s•  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Œô # 6×#=Ñ#=ÀÇÁÑ#MÈv×ObÑObÓcˆŒÜ" 6×#=Ñ#=ÀÇÁÑ#MÈv×ObÑObÓcˆ�r1   r<   Úposition_embeddingsr�   Úpast_key_valuesr�   r#   c                 óR  — |j                   d d }g |¢d‘| 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$                  | j&                  dœ|¤Ž\  }} |j(                  g |¢d‘­Ž j+                  «       }| j-                  |«      }||fS )Nr4   r   r3   ç        )r�   rŽ   )rB   ra   r¼   r·   r½   r¸   r¹   Úviewrs   r©   Úupdater¯   r   Úget_interfacerN   Ú_attn_implementationr�   r“   r³   rŽ   r„   r˜   rº   )r-   r<   r¿   r�   rÀ   r�   Úinput_shapeÚhidden_shapeÚquery_statesr™   rš   ru   rv   Úattention_interfacerœ   r›   s                   r0   r?   zOlmo2Attention.forwardë   s­  € ð $×)Ñ)¨#¨2Ð.ˆØ8˜Ð8 bÐ8¨$¯-©-Ñ8ˆà—{‘{ 4§;¡;¨}Ó#=Ó>ˆØ—[‘[ §¡¨]Ó!;Ó<ˆ
Ø—{‘{ =Ó1ˆà#×(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(r1   r}   )rD   rE   rF   Ú__doc__r   r€   r&   r)   rH   rA   r   r   r   r?   rI   rJ   s   @r0   r®   r®   Î   s¤   ø„ áGñd˜{ð d°s¸T±zõ dð< )-ñ*)à—|‘|ð*)ð # 5§<¡<°·±Ð#=Ñ>ð*)ð Ÿ™ tÑ+ð	*)ð
  ™ð*)ð Ð+Ñ,ð*)ð 
ˆu�|‰|˜UŸ\™\¨DÑ0Ð0Ñ	1÷*)r1   r®   c                   ó$   ‡ — e Zd Zˆ fd„Zd„ Zˆ xZS )ÚOlmo2MLPc                 óÊ  •— t         ‰| �  «        || _        |j                  | _        |j                  | _        t        j                  | j                  | j                  d¬«      | _        t        j                  | j                  | j                  d¬«      | _        t        j                  | j                  | j                  d¬«      | _	        t        |j                     | _        y ©NFr±   )r%   r&   rN   r.   Úintermediate_sizer'   rµ   Ú	gate_projÚup_projÚ	down_projr   Ú
hidden_actÚact_fn©r-   rN   r/   s     €r0   r&   zOlmo2MLP.__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�r1   c                 óˆ   — | j                  | j                  | j                  |«      «      | j                  |«      z  «      }|S r}   )rÓ   rÕ   rÑ   rÒ   )r-   rw   rÓ   s      r0   r?   zOlmo2MLP.forward#  s6   € Ø—N‘N 4§;¡;¨t¯~©~¸aÓ/@Ó#AÀDÇLÁLÐQRÃOÑ#SÓTˆ	ØÐr1   )rD   rE   rF   r&   r?   rI   rJ   s   @r0   rÍ   rÍ     s   ø„ ô0ör1   rÍ   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Olmo2DecoderLayerrN   r¯   c                 ó   •— t         ‰| �  «        |j                  | _        t        ||¬«      | _        t        |«      | _        t        |j                  |j                  ¬«      | _	        t        |j                  |j                  ¬«      | _
        y )N)rN   r¯   ©r"   )r%   r&   r.   r®   Ú	self_attnrÍ   Úmlpr!   r»   Úpost_attention_layernormÚpost_feedforward_layernormr¾   s      €r0   r&   zOlmo2DecoderLayer.__init__)  sl   ø€ Ü‰ÑÔØ!×-Ñ-ˆÔÜ'¨vÀÔKˆŒä˜FÓ#ˆŒÜ(4°V×5GÑ5GÈV×M`ÑM`Ô(aˆÔ%Ü*6°v×7IÑ7IÈv×ObÑObÔ*cˆÕ'r1   Nr<   r�   rx   rÀ   Ú	use_cacher¿   r�   r#   c           
      óÀ   — |} | j                   d||||||dœ|¤Ž\  }}	| j                  |«      }||z   }|}| j                  |«      }| j                  |«      }||z   }|S )N)r<   r�   rx   rÀ   rà   r¿   © )rÜ   rÞ   rÝ   rß   )
r-   r<   r�   rx   rÀ   rà   r¿   r�   ÚresidualÚ_s
             r0   r?   zOlmo2DecoderLayer.forward2  s�   € ð !ˆØ)˜4Ÿ>™>ð 
Ø'Ø)Ø%Ø+ØØ 3ñ
ð ñ
Ñˆ�qð ×5Ñ5°mÓDˆØ  =Ñ0ˆð !ˆØŸ™ Ó/ˆØ×7Ñ7¸ÓFˆØ  =Ñ0ˆØÐr1   )NNNFN)rD   rE   rF   r   r€   r&   r)   rH   Ú
LongTensorr   ÚboolrA   r   r   r?   rI   rJ   s   @r0   rÙ   rÙ   (  s¾   ø„ ðd˜{ð d°sõ d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 
�‰÷r1   rÙ   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Olmo2PreTrainedModelrN   ÚmodelTrÙ   rÀ   )r<   Ú
attentionsN)rD   rE   rF   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_backendrÙ   r®   Ú_can_record_outputsrâ   r1   r0   rè   rè   Q  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r1   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
Olmo2ModelrN   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 )NrÛ   ©rN   F)r%   r&   Úpad_token_idÚpadding_idxÚ
vocab_sizer'   Ú	Embeddingr.   Úembed_tokensÚ
ModuleListÚrangeÚnum_hidden_layersrÙ   Úlayersr!   r»   ÚnormrL   Ú
rotary_embÚgradient_checkpointingÚ	post_initr¾   s      €r0   r&   zOlmo2Model.__init__f  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‘mÜCHÈ×IaÑIaÓCbÖc°iÔ˜v yÕ1Òcó
ˆŒô ! ×!3Ñ!3¸×9LÑ9LÔMˆŒ	Ü.°fÔ=ˆŒØ&+ˆÔ#ð 	�‰Õùò ds   ÂDNÚ	input_idsr�   rx   rÀ   Úinputs_embedsrà   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_embedsrø   r   r   )r\   )rN   r  r�   rÀ   rx   )rx   )r�   r¿   rx   rÀ   rà   )Úlast_hidden_staterÀ   )Ú
ValueErrorrý   r	   rN   Úget_seq_lengthr)   rd   rB   r\   r    r   r  r  r   r  r   )r-   r  r�   rx   rÀ   r  rà   r�   Úpast_seen_tokensÚcausal_maskr<   r¿   Údecoder_layers                r0   r?   zOlmo2Model.forwardv  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$7Ø)Ø /Ø#ñð ñ‰Mð		ð Ÿ	™	 -Ó0ˆÜ&Ø+Ø+ô
ð 	
r1   )NNNNNN)rD   rE   rF   r   r&   r   r   r   r)   rå   rH   r   ÚFloatTensorræ   r   r   r   r?   rI   rJ   s   @r0   rö   rö   d  sÍ   ø„ ð˜{õ ð   ØØð .2Ø.2Ø04Ø(,Ø26Ø!%ñ2
à×#Ñ# dÑ*ð2
ð Ÿ™ tÑ+ð2
ð ×&Ñ&¨Ñ-ð	2
ð
  ™ð2
ð ×(Ñ(¨4Ñ/ð2
ð ˜$‘;ð2
ð Ð+Ñ,ð2
ð 
!ò2
ó ó ó  ô2
r1   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Olmo2ForCausalLMz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 rÏ   )
r%   r&   rö   ré   rû   r'   rµ   r.   r  r  rÖ   s     €r0   r&   zOlmo2ForCausalLM.__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r1   Nr  r�   rx   rÀ   r  Úlabelsrà   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 )aÏ  
        Example:

        ```python
        >>> from transformers import AutoTokenizer, Olmo2ForCausalLM

        >>> model = Olmo2ForCausalLM.from_pretrained("meta-olmo2/Olmo2-2-7b-hf")
        >>> tokenizer = AutoTokenizer.from_pretrained("meta-olmo2/Olmo2-2-7b-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  r�   rx   rÀ   r  rà   N)r  r  rû   )Úlossr  rÀ   r<   rê   râ   )ré   r	  rp   r€   Úslicer  Úloss_functionrN   rû   r   rÀ   r<   rê   )r-   r  r�   rx   rÀ   r  r  rà   r  r�   Úoutputsr<   Úslice_indicesr  r  s                  r0   r?   zOlmo2ForCausalLM.forward½  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 	
r1   )NNNNNNNr   )rD   rE   rF   Ú_tied_weights_keysÚ_tp_planÚ_pp_planr&   r   r   r)   rå   rH   r   r  ræ   r€   r   r   r   r?   rI   rJ   s   @r0   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 ˜$‘;ð6
ð ˜eŸl™lÑ*ð6
ð Ð+Ñ,ð6
ð 
 ò6
ó ó ô6
r1   r  c                   ó   — e Zd Zy)ÚOlmo2ForSequenceClassificationN)rD   rE   rF   râ   r1   r0   r"  r"  ø  s   „ Ør1   r"  )r  r"  rö   rè   )rÂ   )r   )@Úcollections.abcr   Útypingr   r)   Útorch.nnr'   Útransformers.utils.genericr   Úactivationsr   Úcache_utilsr   r	   Ú
generationr
   Úintegrationsr   r   r   Úmasking_utilsr   Úmodeling_layersr   r   Úmodeling_outputsr   r   Úmodeling_rope_utilsr   r   Úmodeling_utilsr   r   Úprocessing_utilsr   Úutilsr   r   Úutils.genericr   r   Úutils.output_capturingr   Úconfiguration_olmo2r   ÚModuler!   rL   rH   r€   rˆ   rG   r�   r©   r¡   r®   rÍ   rÙ   rè   rö   r  r"  Ú__all__râ   r1   r0   ú<module>r7     sñ  ðõ4 %Ý ã Ý å 9å !ß .Ý )ß fÑ fÝ /ß [ß Oß Kß FÝ &ß 5ß GÝ 5Ý ,ñ ˜YÓ'ôJ�2—9‘9ó Jó (ðJô(=˜2Ÿ9™9ô =ð@	U˜UŸ\™\ð 	U°#ð 	U¸%¿,¹,ó 	Uð& ñ%Ø�I‰Ið%à�<‰<ð%ð 
�‰ð%ð �<‰<ð	%ð
 —L‘L 4Ñ'ð%ð ð%ð ð%ð Ð'Ñ(ó%ñ2 Ð*Ó+ò2ó ,ð2ò4(ñ Ð)Ó*ôF)�R—Y‘Yó F)ó +ðF)ôRˆr�y‰yô ô &Ð2ô &ðR ô˜?ó ó ðð$ ôF
Ð%ó F
ó ðF
ðR ôF
Ð+¨_ó F
ó ðF
ôR	Ð%EÐG[ô 	ò g�r1   