Ë
    ÜÓ^j¤T  ã                   óZ  — 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 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 ddlmZmZ ddlm Z m!Z! ddl"m#Z# ddl$m%Z%m&Z& ddl'm(Z(m)Z)m*Z* ddl+m,Z, ddl-m.Z.  G d„ dej                  j^                  «      Z0 G d„ dej^                  «      Z1 ed«      d6d„«       Z2dejf                  de4dejf                  fd„Z5	 d7dej^                  d ejf                  d!ejf                  d"ejf                  d#ejf                  dz  d$e6d%e6d&e#e%   fd'„Z7d(„ Z8 ee2«       G d)„ d*ej^                  «      «       Z9 G d+„ d,ej^                  «      Z: G d-„ d.e«      Z;e& G d/„ d0e!«      «       Z<e& G d1„ d2e<«      «       Z=e& G d3„ d4e<e«      «       Z>g d5¢Z?y)8é    N)ÚCallable)ÚOptionalé   )Úinitialization)ÚACT2FN)ÚCacheÚDynamicCache)ÚGenerationMixin)Úuse_kernel_func_from_hubÚuse_kernelized_func)Úcreate_causal_mask)ÚGradientCheckpointingLayer)ÚBaseModelOutputWithPastÚCausalLMOutputWithPast)ÚROPE_INIT_FUNCTIONSÚdynamic_rope_update)ÚALL_ATTENTION_FUNCTIONSÚPreTrainedModel)ÚUnpack)ÚTransformersKwargsÚauto_docstring)Úcan_return_tupleÚmaybe_autocastÚmerge_with_config_defaults)Úcapture_outputsé   )ÚNanoChatConfigc                   ó8   ‡ — e Zd Zddefˆ fd„Zd„ Zd„ Zd„ Zˆ xZS )ÚNanoChatRMSNormÚepsc                 ó0   •— t         ‰| �  «        || _        y ©N)ÚsuperÚ__init__r    )Úselfr    Ú	__class__s     €úy/var/www/ramen.bs-engineer-server.com/venv/lib/python3.12/site-packages/transformers/models/nanochat/modeling_nanochat.pyr$   zNanoChatRMSNorm.__init__.   s   ø€ Ü‰ÑÔØˆ�ó    c                 óŒ   — |t        j                  |j                  d«      j                  dd¬«      | j                  z   «      z  S )Né   éÿÿÿÿT)Úkeepdim)ÚtorchÚrsqrtÚpowÚmeanr    ©r%   Úxs     r'   Ú_normzNanoChatRMSNorm._norm2   s4   € Ø”5—;‘;˜qŸu™u Q›xŸ}™}¨R¸˜}Ó>ÀÇÁÑIÓJÑJÐJr(   c                 ó^   — | j                  |j                  «       «      j                  |«      S r"   )r3   ÚfloatÚtype_asr1   s     r'   ÚforwardzNanoChatRMSNorm.forward5   s"   € Ø�z‰z˜!Ÿ'™'›)Ó$×,Ñ,¨QÓ/Ð/r(   c                 ó    — d| j                   › �S )Nzeps=©r    )r%   s    r'   Ú
extra_reprzNanoChatRMSNorm.extra_repr8   s   € Ø�d—h‘h�ZÐ Ð r(   )g�íµ ÷Æ°>)	Ú__name__Ú
__module__Ú__qualname__r5   r$   r3   r7   r:   Ú__classcell__©r&   s   @r'   r   r   -   s   ø„ ñ˜Eõ òKò0ö!r(   r   c                   óÄ   ‡ — e Zd ZU ej                  ed<   ddefˆ fd„Ze	 	 	 ddedz  de	d   de
dz  ded	ef   fd
„«       Z ej                  «       ed„ «       «       Zˆ xZS )ÚNanoChatRotaryEmbeddingÚinv_freqNÚconfigc                 ó®  •— t         ‰| �  «        |j                  | _        |j                  | _        || _        | j
                  j                  d   | _        | j                  }| j                  dk7  rt        | j                     } || j
                  |«      \  }| _
        | j                  d|d¬«       | j                  d|j                  «       d¬«       y )NÚ	rope_typeÚdefaultrB   F)Ú
persistentÚoriginal_inv_freq)r#   r$   Úmax_position_embeddingsÚmax_seq_len_cachedÚoriginal_max_seq_lenrC   Úrope_parametersrE   Úcompute_default_rope_parametersr   Úattention_scalingÚregister_bufferÚclone)r%   rC   ÚdeviceÚrope_init_fnrB   r&   s        €r'   r$   z NanoChatRotaryEmbedding.__init__?   s¶   ø€ Ü‰ÑÔØ"(×"@Ñ"@ˆÔØ$*×$BÑ$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�$Ô(à×Ñ˜Z¨¸eÐÔDØ×ÑÐ0°(·.±.Ó2BÈuÐÕUr(   rQ   ztorch.deviceÚseq_lenÚreturnztorch.Tensorc                 ó  — | j                   d   }t        | dd«      xs | j                  | j                  z  }d}d|t	        j
                  d|dt        j                  ¬«      j                  |t        j                  ¬«      |z  z  z  }||fS )	a¨  
        Computes the inverse frequencies according to the original RoPE implementation
        Args:
            config ([`~transformers.PreTrainedConfig`]):
                The model configuration.
            device (`torch.device`):
                The device to use for initialization of the inverse frequencies.
            seq_len (`int`, *optional*):
                The current sequence length. Unused for this type of RoPE.
        Returns:
            Tuple of (`torch.Tensor`, `float`), containing the inverse frequencies for the RoPE embeddings and the
            post-processing scaling factor applied to the computed cos/sin (unused in this type of RoPE).
        Ú
rope_thetaÚhead_dimNg      ð?r   r*   ©Údtype)rQ   rY   )	rL   ÚgetattrÚhidden_sizeÚnum_attention_headsr-   ÚarangeÚint64Útor5   )rC   rQ   rS   ÚbaseÚdimÚattention_factorrB   s          r'   rM   z7NanoChatRotaryEmbedding.compute_default_rope_parametersO   s�   € ð& ×%Ñ% lÑ3ˆÜ�f˜j¨$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   r+   r   ÚmpsÚcpuF)Údevice_typeÚenabledr*   ©ra   rX   )rB   r5   ÚexpandÚshaper_   rQ   Ú
isinstanceÚtypeÚstrr   Ú	transposer-   ÚcatÚcosrN   ÚsinrY   )
r%   r2   Úposition_idsÚinv_freq_expandedÚposition_ids_expandedrf   ÚfreqsÚembrp   rq   s
             r'   r7   zNanoChatRotaryEmbedding.forwardm   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$r"   ©NNN)r;   r<   r=   r-   ÚTensorÚ__annotations__r   r$   Ústaticmethodr   ÚintÚtupler5   rM   Úno_gradr   r7   r>   r?   s   @r'   rA   rA   <   s˜   ø… Ø�l‰lÓñV˜~õ Vð  à(,Ø+/Ø"ñ*Ø Ñ%ð*à˜Ñ(ð*ð �t‘ð*ð 
ˆ~˜uÐ$Ñ	%ò	*ó ð*ð: €U‡]�]ƒ_Øñ<ó ó ô<r(   rA   Úrotary_pos_embc                 óž   — |j                  |«      }|j                  |«      }| |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.
    )Ú	unsqueezeÚrotate_half)ÚqÚkrp   rq   Úunsqueeze_dimÚq_embedÚk_embeds          r'   Úapply_rotary_pos_embr‡   }   sY   € ð& �-‰-˜Ó
&€CØ
�-‰-˜Ó
&€CØ�3‰wœ; q›>¨CÑ/Ñ0€GØ�3‰wœ; q›>¨CÑ/Ñ0€GØ�GÐÐr(   Úhidden_statesÚn_reprT   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)rj   ri   Úreshape)rˆ   r‰   ÚbatchÚnum_key_value_headsÚslenrW   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 )Nr*   r   r+   )ra   rY   )ÚpÚtrainingr   )r�   Únum_key_value_groupsr-   Úmatmulrn   ÚnnÚ
functionalÚsoftmaxÚfloat32r_   rY   r–   rš   Ú
contiguous)r�   r‘   r’   r“   r”   r•   r–   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(   c                 óš   — | dd| j                   d   dz  …f   }| d| j                   d   dz  d…f   }t        j                  || fd¬«      S )zJRotates half the hidden dims of the input with flipped signs for NanoChat..Nr+   r*   rh   )rj   r-   ro   )r2   Úx1Úx2s      r'   r�   r�   ¼   sZ   € à	
ˆ3Ð"�!—'‘'˜"‘+ Ñ"Ð"Ð"Ñ	#€BØ	
ˆ3�—‘˜‘˜qÑ Ñ"Ð"Ñ	#€BÜ�9‰9�b˜2˜#�Y BÔ'Ð'r(   c                   óþ   ‡ — 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z  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NanoChatAttentionz=Multi-headed attention from 'Attention Is All You Need' paperrC   Ú	layer_idxc                 óÐ  •— 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*                  ¬«      | _        t)        |j*                  ¬«      | _        y )NrW   g      à¿T©Úbiasr9   )r#   r$   rC   r¬   rZ   r[   r\   rW   r�   r›   r•   Úattention_dropoutÚ	is_causalr�   ÚLinearÚattention_biasÚq_projÚk_projÚv_projÚo_projr   Úrms_norm_epsÚq_normÚk_norm©r%   rC   r¬   r&   s      €r'   r$   zNanoChatAttention.__init__Ç   ss  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 &¨&×*=Ñ*=Ô>ˆŒÜ%¨&×*=Ñ*=Ô>ˆ�r(   Nrˆ   Úposition_embeddingsr”   Úpast_key_valuesr—   rT   c                 óN  — |j                   d d }g |¢d‘| j                  ‘­}| j                  |«      j                  |«      j	                  dd«      }| j                  |«      j                  |«      j	                  dd«      }	| j                  |«      j                  |«      j	                  dd«      }
|\  }}t        ||	||«      \  }}	| j                  |«      }| j                  |	«      }	|� |j                  |	|
| j                  «      \  }	}
t        j                  | j                  j                  t         «      } || ||	|
|f| j"                  sdn| j$                  | j&                  dœ|¤Ž\  }} |j(                  g |¢d‘­Ž j+                  «       }| j-                  |«      }||fS )Nr+   r   r*   ç        )r–   r•   )rj   rW   r´   Úviewrn   rµ   r¶   r‡   r¹   rº   Úupdater¬   r   Úget_interfacerC   Ú_attn_implementationr¦   rš   r°   r•   r‹   r¡   r·   )r%   rˆ   r¼   r”   r½   r—   Úinput_shapeÚhidden_shapeÚquery_statesr¢   r£   rp   rq   Úattention_interfacer¥   r¤   s                   r'   r7   zNanoChatAttention.forwardá   s¯  € ð $×)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�jð —{‘{ <Ó0ˆØ—[‘[ Ó,ˆ
à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(   rw   )r;   r<   r=   Ú__doc__r   r{   r$   r-   rx   r|   r   r   r   r7   r>   r?   s   @r'   r«   r«   Ã   sª   ø„ áGð?˜~ð ?¸#õ ?ð: IMØ.2Ø(,ñ*)à—|‘|ð*)ð # 5§<¡<°·±Ð#=Ñ>ÀÑEð*)ð Ÿ™ tÑ+ð	*)ð
  ™ð*)ð Ð+Ñ,ð*)ð 
ˆu�|‰|˜UŸ\™\¨DÑ0Ð0Ñ	1÷*)r(   r«   c                   óV   ‡ — e Zd Zˆ fd„Zdej
                  dej
                  fd„Zˆ xZS )ÚNanoChatMLPc                 ó$  •— t         ‰| �  «        || _        t        |j                     | _        t        j                  |j                  |j                  d¬«      | _
        t        j                  |j                  |j                  d¬«      | _        y ©NFr®   )r#   r$   rC   r   Ú
hidden_actÚactivation_fnr�   r²   r[   Úintermediate_sizeÚfc1Úfc2©r%   rC   r&   s     €r'   r$   zNanoChatMLP.__init__  sj   ø€ Ü‰ÑÔØˆŒÜ# F×$5Ñ$5Ñ6ˆÔÜ—9‘9˜V×/Ñ/°×1IÑ1IÐPUÔVˆŒÜ—9‘9˜V×5Ñ5°v×7IÑ7IÐPUÔVˆ�r(   rˆ   rT   c                 ól   — | j                  |«      }| j                  |«      }| j                  |«      }|S r"   )rÐ   rÎ   rÑ   )r%   rˆ   s     r'   r7   zNanoChatMLP.forward  s4   € ØŸ™ Ó/ˆØ×*Ñ*¨=Ó9ˆØŸ™ Ó/ˆØÐr(   )r;   r<   r=   r$   r-   rx   r7   r>   r?   s   @r'   rÊ   rÊ     s$   ø„ ôWð U§\¡\ð °e·l±l÷ r(   rÊ   c                   ó  ‡ — e Zd Zdedefˆ fd„Z	 	 	 	 	 ddej                  dej                  dz  dej                  dz  de	dz  d	e
dz  d
eej                  ej                  f   dz  dee   dej                  fd„Zˆ xZS )ÚNanoChatDecoderLayerrC   r¬   c                 óô   •— t         ‰| �  «        |j                  | _        t        ||¬«      | _        t        |«      | _        t        |j                  ¬«      | _	        t        |j                  ¬«      | _
        y )N)rC   r¬   r9   )r#   r$   r[   r«   Ú	self_attnrÊ   Úmlpr   r¸   Úinput_layernormÚpost_attention_layernormr»   s      €r'   r$   zNanoChatDecoderLayer.__init__  s\   ø€ Ü‰ÑÔØ!×-Ñ-ˆÔä*°&ÀIÔNˆŒä˜vÓ&ˆŒä.°6×3FÑ3FÔGˆÔÜ(7¸F×<OÑ<OÔ(PˆÕ%r(   Nrˆ   r”   rr   r½   Ú	use_cacher¼   r—   rT   c           
      óÀ   — |}| j                  |«      } | j                  d||||||dœ|¤Ž\  }}	||z   }|}| j                  |«      }| j                  |«      }||z   }|S )N)rˆ   r”   rr   r½   rÛ   r¼   © )rÙ   r×   rÚ   rØ   )
r%   rˆ   r”   rr   r½   rÛ   r¼   r—   ÚresidualÚ_s
             r'   r7   zNanoChatDecoderLayer.forward)  s�   € ð !ˆØ×,Ñ,¨]Ó;ˆà)˜4Ÿ>™>ð 
Ø'Ø)Ø%Ø+ØØ 3ñ
ð ñ
Ñˆ�qð ! =Ñ0ˆð !ˆØ×5Ñ5°mÓDˆØŸ™ Ó/ˆØ  =Ñ0ˆØÐr(   )NNNFN)r;   r<   r=   r   r{   r$   r-   rx   Ú
LongTensorr   Úboolr|   r   r   r7   r>   r?   s   @r'   rÕ   rÕ     s¾   ø„ ð	Q˜~ð 	Q¸#õ 	Qð /3Ø04Ø(,Ø!&ØHLñà—|‘|ðð Ÿ™ tÑ+ðð ×&Ñ&¨Ñ-ð	ð
  ™ðð ˜$‘;ðð # 5§<¡<°·±Ð#=Ñ>ÀÑEðð Ð+Ñ,ðð 
�‰÷r(   rÕ   c                   óz   ‡ — e Zd ZU eed<   dZdZdgZdgZdZ	dZ
dZdZdZeedœZdej$                  dd	fˆ fd
„Zˆ xZS )ÚNanoChatPreTrainedModelrC   ÚmodelTrÕ   r½   )rˆ   Ú
attentionsr�   rT   Nc           	      ó   •— t         ‰| �  |«       t        |t        «      rnt	        j
                  |j                  j                  d| j                  j                  t        j                  d| j                  j                  z  «      z  ¬«       y y )Nr¿   r*   )r0   Ústd)r#   Ú_init_weightsrk   r«   ÚinitÚnormal_r·   ÚweightrC   Úinitializer_rangeÚmathÚsqrtÚnum_hidden_layers)r%   r�   r&   s     €r'   rè   z%NanoChatPreTrainedModel._init_weights[  sf   ø€ Ü‰Ñ˜fÔ%Ü�fÔ/Ô0Ü�L‰LØ—‘×$Ñ$ØØ—K‘K×1Ñ1´D·I±I¸aÀ$Ç+Á+×B_ÑB_Ñ>_Ó4`Ñ`öð 1r(   )r;   r<   r=   r   ry   Úbase_model_prefixÚsupports_gradient_checkpointingÚ_no_split_modulesÚ_skip_keys_device_placementÚ_supports_flash_attnÚ_supports_sdpaÚ_supports_flex_attnÚ_can_compile_fullgraphÚ_supports_attention_backendrÕ   r«   Ú_can_record_outputsr�   ÚModulerè   r>   r?   s   @r'   rã   rã   I  sn   ø… àÓØÐØ&*Ð#Ø/Ð0ÐØ#4Ð"5ÐØÐØ€NØÐà!ÐØ"&Ðà-Ø'ñÐð
 B§I¡Ið °$÷ ñ r(   rã   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NanoChatModelrC   c           	      óö  •— 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 )Nr9   ©rC   F)r#   r$   Úpad_token_idÚpadding_idxÚ
vocab_sizer�   Ú	Embeddingr[   Úembed_tokensÚ
ModuleListÚrangerï   rÕ   Úlayersr   r¸   ÚnormrA   Ú
rotary_embÚgradient_checkpointingÚ	post_initr»   s      €r'   r$   zNanoChatModel.__init__g  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‘mÜFKÈF×LdÑLdÓFeÖf¸Ô! &¨)Õ4Òfó
ˆŒô $¨×(;Ñ(;Ô<ˆŒ	Ü1¸Ô@ˆŒØ&+ˆÔ#ð 	�‰Õùò gs   ÂC6NÚ	input_idsr”   rr   r½   Úinputs_embedsrÛ   r—   rT   c           	      ó\  — |d u |d uz  rt        d«      ‚|€| j                  |«      }|r|€t        | j                  ¬«      }|€V|�|j	                  «       nd}t        j                  |j                  d   |j                  ¬«      |z   }|j                  d«      }t        | j                  ||||¬«      }	|}
| j                  |
|¬«      }| j                  |
«      }
| j                  d | j                  j                   D ]  } ||
f|	|||dœ|¤Ž}
Œ | j                  |
«      }
t        |
|¬	«      S )
Nz:You must specify exactly one of input_ids or inputs_embedsrþ   r   r   )rQ   )rC   r  r”   r½   rr   )rr   )r”   r¼   rr   r½   )Úlast_hidden_stater½   )Ú
ValueErrorr  r	   rC   Úget_seq_lengthr-   r]   rj   rQ   r€   r   r  r  r  rï   r   )r%   r  r”   rr   r½   r  rÛ   r—   Úpast_seen_tokensÚcausal_maskrˆ   r¼   Údecoder_layers                r'   r7   zNanoChatModel.forwardx  sW  € ð ˜Ð -°tÐ";Ò<ÜÐYÓZÐZàÐ Ø*.×*;Ñ*;¸IÓ*FˆMá˜Ð0Ü*°$·+±+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	™	 -Ó0ˆØ!Ÿ[™[Ð)H¨4¯;©;×+HÑ+HÐIò 	ˆMÙ)Øðà*Ø$7Ø)Ø /ñð ñ‰Mð	ð Ÿ	™	 -Ó0ˆÜ&Ø+Ø+ô
ð 	
r(   )NNNNNN)r;   r<   r=   r   r$   r   r   r   r-   rà   rx   r   ÚFloatTensorrá   r   r   r   r7   r>   r?   s   @r'   rü   rü   e  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 ˜$‘;ð2
ð Ð+Ñ,ð2
ð 
!ò2
ó ó ó  ô2
r(   rü   c                   óB  ‡ — e Zd ZddiZddiZddgdgfiZˆ fd„Zee	 	 	 	 	 	 	 	 dd	e	j                  dz  d
e	j                  dz  de	j                  dz  dedz  de	j                  dz  de	j                  dz  dedz  dee	j                  z  dee   defd„«       «       Zˆ xZS )ÚNanoChatForCausalLMzlm_head.weightzmodel.embed_tokens.weightÚlm_headÚcolwise_gather_outputrˆ   Úlogitsc                 óè   •— t         ‰| �  |«       t        |«      | _        |j                  | _        t        j                  |j                  |j                  d¬«      | _        | j                  «        y rÌ   )
r#   r$   rü   rä   r  r�   r²   r[   r  r
  rÒ   s     €r'   r$   zNanoChatForCausalLM.__init__¶  sU   ø€ Ü‰Ñ˜Ô Ü" 6Ó*ˆŒ
Ø ×+Ñ+ˆŒÜ—y‘y ×!3Ñ!3°V×5FÑ5FÈUÔSˆŒð 	�‰Õr(   Nr  r”   rr   r½   r  ÚlabelsrÛ   Úlogits_to_keepr—   rT   c	           
      ó  —  | j                   d||||||dœ|	¤Ž}
|
j                  }t        |t        «      rt	        | d«      n|}| j                  |dd…|dd…f   «      }| j                  j                  �G|| j                  j                  z  }t        j                  |«      }|| j                  j                  z  }d}|� | j                  ||| j                  fi |	¤Ž}t        |||
j                  |
j                  |
j                  ¬«      S )ak  
        Example:

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

        >>> model = AutoModelForCausalLM.from_pretrained("karpathy/nanochat-d32")

        >>> tokenizer = AutoTokenizer.from_pretrained("karpathy/nanochat-d32")

        >>> conversation = [
                {"role": "user", "content": "What is the capital of France?"},
            ]

        >>> inputs = tokenizer.apply_chat_template(
                conversation, add_generation_prompt=True, tokenize=True, return_dict=True, return_tensors="pt"
            ).to(device)

        >>> with torch.no_grad():
        >>>     outputs = model.generate(**inputs, max_new_tokens=64, do_sample=False)

        >>> generated_tokens = outputs[0, inputs["input_ids"].shape[1] :]
        >>> output = tokenizer.decode(generated_tokens, skip_special_tokens=True)
        ```)r  r”   rr   r½   r  rÛ   N)Úlossr  r½   rˆ   rå   rÝ   )rä   r  rk   r{   Úslicer  rC   Úfinal_logit_softcappingr-   ÚtanhÚloss_functionr  r   r½   rˆ   rå   )r%   r  r”   rr   r½   r  r  rÛ   r  r—   Úoutputsrˆ   Úslice_indicesr  r  s                  r'   r7   zNanoChatForCausalLM.forward¿  s  € ðN ,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Ø�;‰;×.Ñ.Ð:Ø˜dŸk™k×AÑAÑAˆFÜ—Z‘Z Ó'ˆFØ˜dŸk™k×AÑAÑAˆFàˆØÐØ%�4×%Ñ% f¨f°d·o±oÑPÈÑPˆDä%ØØØ#×3Ñ3Ø!×/Ñ/Ø×)Ñ)ô
ð 	
r(   )NNNNNNNr   )r;   r<   r=   Ú_tied_weights_keysÚ_tp_planÚ_pp_planr$   r   r   r-   rà   rx   r   r  rá   r{   r   r   r   r7   r>   r?   s   @r'   r  r  °  s#  ø„ à*Ð,GÐHÐØÐ2Ð3€HØ˜_Ð-°¨zÐ:Ð;€Hôð Øð .2Ø.2Ø04Ø(,Ø26Ø*.Ø!%Ø-.ñB
à×#Ñ# dÑ*ðB
ð Ÿ™ tÑ+ðB
ð ×&Ñ&¨Ñ-ð	B
ð
  ™ðB
ð ×(Ñ(¨4Ñ/ðB
ð × Ñ  4Ñ'ðB
ð ˜$‘;ðB
ð ˜eŸl™lÑ*ðB
ð Ð+Ñ,ðB
ð 
 òB
ó ó ôB
r(   r  )rã   rü   r  )r   )r¿   )@rí   Úcollections.abcr   Útypingr   r-   Útorch.nnr�   Ú r   ré   Úactivationsr   Úcache_utilsr   r	   Ú
generationr
   Úintegrationsr   r   Úmasking_utilsr   Úmodeling_layersr   Úmodeling_outputsr   r   Úmodeling_rope_utilsr   r   Úmodeling_utilsr   r   Úprocessing_utilsr   Úutilsr   r   Úutils.genericr   r   r   Úutils.output_capturingr   Úconfiguration_nanochatr   rú   r   rA   r‡   rx   r{   r�   r5   r¦   r�   r«   rÊ   rÕ   rã   rü   r  Ú__all__rÝ   r(   r'   ú<module>r;     sÑ  ðó* Ý $Ý ã Ý å &Ý !ß .Ý )ß IÝ /Ý 9ß Oß Kß FÝ &ß 7ß YÑ YÝ 5Ý 2ô!�e—h‘h—o‘oô !ô><˜bŸi™iô ><ñB Ð*Ó+òó ,ðð2	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(ñ Ð)Ó*ôG)˜Ÿ	™	ó G)ó +ðG)ôT�"—)‘)ô ô)Ð5ô )ðX ô˜oó ó ðð6 ôG
Ð+ó G
ó ðG
ðT ôR
Ð1°?ó R
ó ðR
òj N�r(   