
    ^j                        d dl mZ d dlmZmZ d dlZd dlmc mZ	 d dlmZ ddl
mZ ddlmZ ddlmZmZ dd	lmZ dd
lmZmZ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/m0Z0 ddl1m2Z2  G d ded      Z3 G d dejh                        Z5 ed       G d dejh                               Z6 G d d ejh                        Z7e G d! d"ejh                               Z8 G d# d$ejh                        Z9d% Z: ed&      dEd'       Z;d(ejx                  d)e=d*ejx                  fd+Z>	 dFd,ejh                  d-ejx                  d.ejx                  d/ejx                  d0ejx                  dz  d1e?d2e?d3e'e)   fd4Z@ ee;       G d5 d6ejh                               ZA G d7 d8e      ZBe* G d9 d:e%             ZC G d; d<ejh                        ZDe* G d= d>eC             ZE	 	 	 dGd?ejx                  eFejx                     z  dz  d@e=dz  d0ejx                  dz  d*ejx                  e=z  fdAZGe* G dB dCeCe             ZHg dDZIy)H    )Callable)Optional	TypedDictN)nn   )initialization)ACT2FN)CacheDynamicCache)GenerationMixin)use_experts_implementationuse_kernel_forward_from_hubuse_kernel_func_from_hubuse_kernelized_func)create_causal_mask)GradientCheckpointingLayer)MoeCausalLMOutputWithPastMoeModelOutputWithPast)ROPE_INIT_FUNCTIONSdynamic_rope_update)ALL_ATTENTION_FUNCTIONSPreTrainedModel)Unpack)TransformersKwargsauto_docstring)can_return_tuplemaybe_autocastmerge_with_config_defaults)capture_outputs   )GraniteMoeSharedConfigc                       e Zd ZU dZej
                  ed<   ej
                  ed<   eed<   eed<   ej                  ed<   y)GraniteFlashAttentionKwargsaT  
    Keyword arguments for advanced Flash Attention, causal-conv1d, and mamba_ssm kernel usage.
    Use cases include padding-free training and fewer `torch.compile` graph breaks.

    cu_seq_lens_q (`torch.LongTensor`):
        Gets cumulative sequence length for query state.
    cu_seq_lens_k (`torch.LongTensor`):
        Gets cumulative sequence length for key state.
    max_length_q (`int`):
        Maximum sequence length for query state.
    max_length_k (`int`):
        Maximum sequence length for key state.
    seq_idx (`torch.IntTensor):
        Index of each packed sequence.
    cu_seq_lens_qcu_seq_lens_kmax_length_qmax_length_kseq_idxN)	__name__
__module____qualname____doc__torch
LongTensor__annotations__int	IntTensor     /var/www/ramen.bs-engineer-server.com/venv/lib/python3.12/site-packages/transformers/models/granitemoeshared/modeling_granitemoeshared.pyr#   r#   2   s7      ######__r3   r#   F)totalc                   `     e Zd ZdZdef fdZdej                  dej                  fdZ xZ	S )GraniteMoeSharedMLPz~
    MLP layer for shared experts

    Args:
        config:
            Configuration object with model hyperparameters.
    configc                 `   t         |           |j                  | _        |j                  | _        t
        |j                     | _        t        j                  | j                  | j                  dz  d      | _
        t        j                  | j                  | j                  d      | _        y )N   Fbias)super__init__hidden_size
input_sizeshared_intermediate_sizer	   
hidden_act
activationr   Linearinput_linearoutput_linearselfr8   	__class__s     r4   r>   zGraniteMoeSharedMLP.__init__S   s     ,,!:: !2!23IIdoot7G7G!7KRWXYYt'7'7uUr3   hidden_statesreturnc                     | j                  |      }|j                  dd      }| j                  |d         |d   z  }| j                  |      }|S )Nr:   dimr   r    )rE   chunkrC   rF   )rH   rJ   chunked_hidden_statess      r4   forwardzGraniteMoeSharedMLP.forward\   s^    ))-8 - 3 3A2 3 >(=a(@ADYZ[D\\**=9r3   
r)   r*   r+   r,   r!   r>   r-   TensorrR   __classcell__rI   s   @r4   r7   r7   J   s2    V5 VU\\ ell r3   r7   RMSNormc                   h     e Zd Zddeddf fdZdej                  dej                  fdZd Z xZ	S )	GraniteMoeSharedRMSNormepsrK   Nc                     t         |           t        j                  t	        j
                  |            | _        || _        y)zF
        GraniteMoeSharedRMSNorm is equivalent to T5LayerNorm
        N)r=   r>   r   	Parameterr-   onesweightvariance_epsilon)rH   r?   rZ   rI   s      r4   r>   z GraniteMoeSharedRMSNorm.__init__f   s1     	ll5::k#:; #r3   rJ   c                 "   |j                   }|j                  t        j                        }|j	                  d      j                  dd      }|t        j                  || j                  z         z  }| j                  |j                  |      z  S )Nr:   rM   T)keepdim)	dtypetor-   float32powmeanrsqrtr_   r^   )rH   rJ   input_dtypevariances       r4   rR   zGraniteMoeSharedRMSNorm.forwardn   sy    #))%((7 $$Q',,R,>%Ht?T?T4T(UU{{]--k:::r3   c                 ^    t        | j                  j                         d| j                   S )Nz, eps=)tupler^   shaper_   )rH   s    r4   
extra_reprz"GraniteMoeSharedRMSNorm.extra_repru   s*    ))*+6$2G2G1HIIr3   )gư>)
r)   r*   r+   floatr>   r-   rT   rR   rm   rU   rV   s   @r4   rY   rY   d   s7    $ $$ $;U\\ ;ell ;Jr3   rY   c                        e Zd ZdZdef fdZdej                  deej                  ej                  ej                  f   fdZ	 xZ
S )GraniteMoeSharedTopKRoutera  Top-k gating that returns the routing decisions without grouping tokens by expert.

    Returns ``(top_k_index, top_k_weights, router_logits)``; the grouping/scattering used to live
    here (via ``expert_size.tolist()``, which broke fullgraph compile) and now happens inside the
    experts forward via ``use_experts_implementation`` so the default ``grouped_mm`` / ``batched_mm``
    paths can compile cleanly.
    r8   c                     t         |           |j                  | _        |j                  | _        t        j                  t        j                  | j                  |j                              | _        y N)r=   r>   num_local_expertsnum_expertsnum_experts_per_toktop_kr   r\   r-   emptyr?   r^   rG   s     r4   r>   z#GraniteMoeSharedTopKRouter.__init__   sP    !33//
ll5;;t/?/?ASAS#TUr3   rJ   rK   c                     t        j                  || j                        j                         }|j	                  | j
                  d      \  }}t        j                  |d      j                  |      }|||fS )NrM   rN   )	Flinearr^   rn   topkrv   r-   softmaxtype_as)rH   rJ   router_logitstop_k_logitstop_k_indextop_k_weightss         r4   rR   z"GraniteMoeSharedTopKRouter.forward   se    <BBD$1$6$6tzzr$6$J!kl;CCMRM=88r3   )r)   r*   r+   r,   r!   r>   r-   rT   rk   rR   rU   rV   s   @r4   rp   rp   y   sI    V5 V9U\\ 9eELL%,,X]XdXd<d6e 9r3   rp   c                        e Zd ZdZdef fdZdej                  dej                  dej                  dej                  fdZ xZ	S )	GraniteMoeSharedExpertsz2Collection of expert weights stored as 3D tensors.r8   c                    t         |           |j                  | _        |j                  | _        |j                  | _        t        j                  t        j                  | j                  d| j                  z  | j
                              | _        t        j                  t        j                  | j                  | j
                  | j                              | _        t        |j                     | _        y )Nr:   )r=   r>   rs   rt   r?   
hidden_dimintermediate_sizeintermediate_dimr   r\   r-   rw   gate_up_proj	down_projr	   rB   act_fnrG   s     r4   r>   z GraniteMoeSharedExperts.__init__   s    !33 ,, & 8 8LLT5E5Eq4K`K`G`bfbqbq)rsekk$2B2BDOOUYUjUj&klV../r3   rJ   r   r   rK   c                 f   t        j                  |      }t        j                         5  t         j                  j                  j                  || j                        }|j                  ddd      }t        j                  |j                  d      d      j                         }d d d        D ]  }|d   }|| j                  k(  rt        j                  |         \  }}	||	   }
t        j                  j                  |
| j                  |         j                  dd      \  }}| j                  |      |z  }t        j                  j                  || j                   |         }|||	|d f   z  }|j#                  d|	|j%                  |j&                                |S # 1 sw Y   xY w)N)num_classesr:   r    r   )rM   rN   rM   )r-   
zeros_likeno_gradr   
functionalone_hotrt   permutegreatersumnonzerowhererz   r   rP   r   r   
index_add_rc   rb   )rH   rJ   r   r   final_hidden_statesexpert_mask
expert_hit
expert_idx	top_k_pos	token_idxcurrent_stategateupcurrent_hidden_statess                 r4   rR   zGraniteMoeSharedExperts.forward   s    $..}=]]_ 	S((--55ktO_O_5`K%--aA6K{8'DaHPPRJ	S
 % 
	nJ#AJT---#(;;{:/F#G Iy))4M}}++M4;L;LZ;XY__`agi_jHD"$(KK$5$:!$&MM$8$89NPTP^P^_iPj$k!$9M)U^`dJd<e$e!**1i9N9Q9QReRkRk9lm
	n #"#	S 	Ss   A=F&&F0rS   rV   s   @r4   r   r      sN    <05 0#||# \\# ||	#
 
#r3   r   c                   `     e Zd ZdZdef fdZdej                  dej                  fdZ xZ	S )GraniteMoeSharedMoEzISparsely-gated mixture-of-experts block: router decides, experts compute.r8   c                     t         |           |j                  | _        t	        |      | _        t        |      | _        y rr   )r=   r>   r?   r@   rp   routerr   expertsrG   s     r4   r>   zGraniteMoeSharedMoE.__init__   s3     ,,08.v6r3   layer_inputrK   c                     |j                         \  }}}|j                  d|      }| j                  |      \  }}}| j                  |||      }	|	j	                  ||| j
                        S )NrM   )sizereshaper   r   viewr@   )
rH   r   bszlengthemb_sizerJ   r   r   _layer_outputs
             r4   rR   zGraniteMoeSharedMoE.forward   si     + 0 0 2VX#++B9(,M(B%]A||M;N  fdoo>>r3   rS   rV   s   @r4   r   r      s.    S75 7?5<< ?ELL ?r3   r   c                     | 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..NrM   r:   rN   )rl   r-   cat)xx1x2s      r4   rotate_halfr      sZ    	
3"!''"+"""	#B	
3q ""	#B99rc2YB''r3   rotary_pos_embc                     |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.
    )	unsqueezer   )qkcossinunsqueeze_dimq_embedk_embeds          r4   apply_rotary_pos_embr      sY    & --
&C
--
&C3w;q>C/0G3w;q>C/0GGr3   rJ   n_reprK   c                     | j                   \  }}}}|dk(  r| S | dddddddddf   j                  |||||      } | j                  |||z  ||      S )z
    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)rl   expandr   )rJ   r   batchnum_key_value_headsslenhead_dims         r4   	repeat_kvr      so    
 2?1D1D.Ehz!!Qa"23::5BUW\^bdlmM  (;e(CT8TTr3   modulequerykeyvalueattention_maskscalingdropoutkwargsc                    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   rM   )rO   rb   )ptrainingr    )r   num_key_value_groupsr-   matmul	transposer   r   r|   rd   rc   rb   r   r   
contiguous)r   r   r   r   r   r   r   r   
key_statesvalue_statesattn_weightsattn_outputs               r4   eager_attention_forwardr      s     3 ; ;<JUF$?$?@L<<z';';Aq'ABWLL!#n4==((2U]](SVVW\WbWbcL==((6??([L,,|\:K''1-88:K$$r3   c                        e Zd ZdZdedef 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 xZS )GraniteMoeSharedAttentionz=Multi-headed attention from 'Attention Is All You Need' paperr8   	layer_idxc                 ^   t         |           || _        || _        t	        |d|j
                  |j                  z        | _        |j                  |j                  z  | _	        |j                  | _        |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 )Nr   Tr;   )r=   r>   r8   r   getattrr?   num_attention_headsr   r   r   attention_multiplierr   attention_dropout	is_causalr   rD   attention_biasq_projk_projv_projo_projrH   r8   r   rI   s      r4   r>   z"GraniteMoeSharedAttention.__init__  sJ   "
F4F4F&JdJd4de$*$>$>&B\B\$\!22!'!9!9ii : :T]] JQWQfQf
 ii : :T]] JQWQfQf
 ii : :T]] JQWQfQf
 ii&&68J8JQWQfQf
r3   NrJ   position_embeddingsr   past_key_valuesr   rK   c                 
   |j                   d d }g |d| j                  }| j                  |      j                  |      j	                  dd      }| j                  |      j                  |      j	                  dd      }	| j                  |      j                  |      j	                  dd      }
|\  }}t        ||	||      \  }}	| |j                  |	|
| j                        \  }	}
t        j                  | j                  j                  t              } || ||	|
|f| j                  sdn| j                   | j"                  d|\  }} |j$                  g |d j'                         }| j)                  |      }||fS )NrM   r    r:           )r   r   )rl   r   r   r   r   r   r   r   updater   r   get_interfacer8   _attn_implementationr   r   r   r   r   r   r   )rH   rJ   r   r   r   r   input_shapehidden_shapequery_statesr   r   r   r   attention_interfacer   r   s                   r4   rR   z!GraniteMoeSharedAttention.forward)  s    $))#2.88b8$--8{{=166|DNNqRST[[/44\BLLQPQR
{{=166|DNNqRST&S#7jRUWZ#[ j&'6'='=j,X\XfXf'g$J(?(M(MKK,,.E)
 %8	%
  $}}C$2H2HLL	%
 	%
!\ *k));;;;FFHkk+.L((r3   NNN)r)   r*   r+   r,   r!   r0   r>   r-   rT   rk   r
   r   r   rR   rU   rV   s   @r4   r   r     s    G
5 
# 
4 IM.2(,&)||&) #5<<#=>E&) t+	&)
 &) +,&) 
u||U\\)	*&)r3   r   c                   P    e Zd Zdedef fdZ	 	 	 	 	 	 ddej                  dej                  dz  dej                  dz  de	dz  d	e
dz  d
e
dz  deej                  ej                  f   dz  dee   deej                  eej                  ej                  f   dz  f   fdZ xZS )GraniteMoeSharedDecoderLayerr8   r   c                    t         |           |j                  | _        t        ||      | _        t        |j                  |j                        | _        t        |j                  |j                        | _        t        |      | _
        |j                  | _        |j                  dk(  rd | _        y t        |      | _        y )N)r8   r   rZ   r   )r=   r>   r?   r   	self_attnrY   rms_norm_epsinput_layernormpost_attention_layernormr   block_sparse_moeresidual_multiplierrA   r7   
shared_mlpr   s      r4   r>   z%GraniteMoeSharedDecoderLayer.__init__S  s    !--2&IV6v7I7IvObObc(?@R@RX^XkXk(l% 3F ;#)#=#= "("A"AQ"F$L_`fLgr3   NrJ   r   position_idsr   output_attentions	use_cacher   r   rK   c                 <   |}	| j                  |      } | j                  d|||||||d|\  }}
|	|| j                  z  z   }|}	| j                  |      }| j	                  |      }| j
                  |}n|| j                  |      z   }|	|| j                  z  z   }|S )N)rJ   r   r  r   r  r  r   r2   )r   r   r  r   r  r  )rH   rJ   r   r  r   r  r  r   r   residualr   moe_hidden_statess               r4   rR   z$GraniteMoeSharedDecoderLayer.forward]  s     !,,]; *4>> 	
')%+/ 3	
 	
q !=43K3K#KK 55mD 11-@??"-M-0NNM =43K3K#KKr3   )NNNFFN)r)   r*   r+   r!   r0   r>   r-   rT   r.   r
   boolrk   r   r#   FloatTensorrR   rU   rV   s   @r4   r   r   R  s    h5 h# h /304(,).!&HL%||% t+% &&-	%
 %  $;% $;% #5<<#=>E% 45% 
u  %(9(95;L;L(L"MPT"TT	U%r3   r   c                        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Z ej$                          fd       Z xZS )GraniteMoeSharedPreTrainedModelr8   modelTr   r   )rJ   
attentionsc                    t         |   |       t        |t              rmt	        j
                  |j                  d| j                  j                         t	        j
                  |j                  d| j                  j                         y t        |t              r7t	        j
                  |j                  d| j                  j                         y y )Nr   )rf   std)r=   _init_weights
isinstancer   initnormal_r   r8   initializer_ranger   rp   r^   )rH   r   rI   s     r4   r  z-GraniteMoeSharedPreTrainedModel._init_weights  s    f%f56LL,,3DKK<Y<YZLL))9V9VW :;LLSdkk6S6ST <r3   )r)   r*   r+   r!   r/   base_model_prefixsupports_gradient_checkpointing_no_split_modules_skip_keys_device_placement_supports_flash_attn_supports_sdpa_supports_flex_attn_can_compile_fullgraph_supports_attention_backendr   r   _can_record_outputsr-   r   r  rU   rV   s   @r4   r  r    sp    ""&*#78#4"5N!"&5/
 U]]_U Ur3   r  c                        e Zd ZU ej                  ed<   ddef 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 xZS )GraniteMoeSharedRotaryEmbeddinginv_freqNr8   c                    t         |           |j                  | _        |j                  | _        || _        | j
                  j                  d   | _        | j                  }| j                  dk7  rt        | j                     } || j
                  |      \  }| _
        | j                  d|d       | j                  d|j                         d       y )N	rope_typedefaultr#  F)
persistentoriginal_inv_freq)r=   r>   max_position_embeddingsmax_seq_len_cachedoriginal_max_seq_lenr8   rope_parametersr%  compute_default_rope_parametersr   attention_scalingregister_bufferclone)rH   r8   devicerope_init_fnr#  rI   s        r4   r>   z(GraniteMoeSharedRotaryEmbedding.__init__  s    "("@"@$*$B$B!44[A!%!E!E>>Y&.t~~>L+7V+L($(ZeD0(..2BuUr3   r1  ztorch.deviceseq_lenrK   ztorch.Tensorc                    | 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).
        
rope_thetar   Ng      ?r   r:   rb   )r1  rb   )	r,  r   r?   r   r-   arangeint64rc   rn   )r8   r1  r3  baserO   attention_factorr#  s          r4   r-  z?GraniteMoeSharedRotaryEmbedding.compute_default_rope_parameters  s    & %%l3fj$/c63E3EIcIc3c U\\!S!5;;?BB&X]XcXcBdgjjk
 )))r3   c                 N   | j                   d d d d f   j                         j                  |j                  d   dd      j	                  |j
                        }|d d d d d f   j                         }t        |j
                  j                  t              r/|j
                  j                  dk7  r|j
                  j                  nd}t        |d      5  |j                         |j                         z  j                  dd      }t        j                  ||fd	      }|j                         | j                  z  }|j                         | j                  z  }	d d d        j	                  |j                   
      	j	                  |j                   
      fS # 1 sw Y   AxY w)Nr   rM   r    mpscpuF)device_typeenabledr:   rN   r6  )r#  rn   r   rl   rc   r1  r  typestrr   r   r-   r   r   r.  r   rb   )
rH   r   r  inv_freq_expandedposition_ids_expandedr>  freqsembr   r   s
             r4   rR   z'GraniteMoeSharedRotaryEmbedding.forward  sR    !MM$4-8>>@GGHZHZ[\H]_acdehhijiqiqr ,QaZ 8 > > @'1!((--'E!((--[`J`ahhmmfkUC 	5&,,.1F1L1L1NNYYZ[]^_E))UEN3C'')d444C'')d444C		5 vvAGGv$cff177f&;;;	5 	5s   BFF$rr   r   )r)   r*   r+   r-   rT   r/   r!   r>   staticmethodr   r0   rk   rn   r-  r   r   rR   rU   rV   s   @r4   r"  r"    s    llV5 V  04+/"*&-*(* t* 
~u$	%	* *: U]]_<  <r3   r"  c                        e Zd Zdef 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 xZS )GraniteMoeSharedModelr8   c           	      .   t         |   |       |j                  | _        |j                  | _        t        j                  |j                  |j                  | j                        | _        t        j                  t        |j                        D cg c]  }t        ||       c}      | _        t        |j                  |j                        | _        t#        |      | _        d| _        |j(                  | _        | j+                          y c c}w )Nr   r8   F)r=   r>   pad_token_idpadding_idx
vocab_sizer   	Embeddingr?   embed_tokens
ModuleListrangenum_hidden_layersr   layersrY   r   normr"  
rotary_embgradient_checkpointingembedding_multiplier	post_initr   s      r4   r>   zGraniteMoeSharedModel.__init__  s     !.. ++LL):):F<N<NPTP`P`ammNSTZTlTlNmn)&)<n
 ,F,>,>FDWDWX	9H&+#$*$?$?! 	 os   DN	input_idsr   r  r   inputs_embedsr  r   rK   c           
      X   |d u |d uz  rt        d      |r|t        | j                        }|| j                  |      }|V||j	                         nd}t        j                  |j                  d   |j                        |z   }|j                  d      }t        | j                  ||||      }	|| j                  z  }|}
| j                  |
|      }| j                  d | j                  j                   D ]  } ||
f||	|||d|}
 | j                  |
      }
t!        |
|      S )	Nz:You must specify exactly one of input_ids or inputs_embedsrJ  r   r    )r1  )r8   rZ  r   r   r  )r   r   r  r   r  )last_hidden_stater   )
ValueErrorr   r8   rO  get_seq_lengthr-   r7  rl   r1  r   r   rW  rU  rS  rR  rT  r   )rH   rY  r   r  r   rZ  r  r   past_seen_tokenscausal_maskrJ   r   decoder_layers                r4   rR   zGraniteMoeSharedModel.forward  s\    -t";<YZZ0*$++>O  --i8MCRC^==?de <<(;(;A(>}G[G[\_ooL'11!4L(;;')+%
 &(A(AA% #oom\J![[)H4;;+H+HI 		M)$7*) /# M		 		-0%++
 	
r3   )NNNNNN)r)   r*   r+   r!   r>   r   r   r   r-   r.   rT   r
   r  r
  r   r   r   rR   rU   rV   s   @r4   rH  rH    s    5 "   .2.204(,26!%5
##d*5
 t+5
 &&-	5

 5
 ((4/5
 $;5
 +,5
 
 5
    5
r3   rH  gate_logitsrt   c                    | t        | t              syt        | t              rC| d   j                  }t        j                  | D cg c]  }|j                  |       c}d      }t        j                  j                  j                  d      }t        j                  ||d      \  }}	t        j                  j                  j                  |	|      }
|>t        j                  |
j                         d      }t        j                  |d      }n|j                  \  }}|j                  d   ||z  z  }|dddddddf   j                  |||||f      j                  d||      j                        }t        j                   |
j                         |z  d      t        j                   |d      z  }|ddddddf   j                  ||||f      j                  d|      j                  |      }t        j                   ||z  d      t        j                   |d      z  }t        j                   ||j#                  d      z        }||z  S c c}w )a  
    Computes auxiliary load balancing loss as in Switch Transformer - implemented in Pytorch.

    See Switch Transformer (https://huggingface.co/papers/2101.03961) for more details. This function implements the loss
    function presented in equations (4) - (6) of the paper. It aims at penalizing cases where the routing between
    experts is too unbalanced.

    Args:
        gate_logits:
            Logits from the `gate`, should be a tuple of model.config.num_hidden_layers tensors of
            shape [batch_size X sequence_length, num_experts].
        num_experts:
            Number of experts
        top_k:
            The number of experts to route per-token, can be also interpreted as the `top-k` routing
            parameter.
        attention_mask (`torch.Tensor`, *optional*):
            The attention_mask used in forward function
            shape [batch_size X sequence_length] if not None.

    Returns:
        The auxiliary loss.
    Nr   rN   rM   )r  rk   r1  r-   r   rc   r   r   r|   r{   r   rf   rn   rl   r   r   r   r   )rb  rt   rv   r   compute_device
layer_gateconcatenated_gate_logitsrouting_weightsr   selected_expertsr   tokens_per_expertrouter_prob_per_expert
batch_sizesequence_lengthrR  expert_attention_mask router_per_expert_attention_maskoverall_losss                      r4   load_balancing_loss_funcrp  /  s9   : *[%"@+u%$Q..#(99^i-jPZjmmN.K-jpq#r hh))112JPR1SO**_eDA((%%--.>LK!JJ{'8'8':B "'O!C&4&:&:#
O4::1=*B^_ 4AtT12V&
OUKXYWR,R	 	 "IIk&7&7&9<Q&QWXY\a\e\e!q]
 
 4At+,V&
O[QRWR%R	 	) "'?=]+]cd!ehmhqhq,!i
 "
 99.1G1Q1QRS1TTUL+%%[ .ks   Ic                   D    e Zd ZddiZddiZddgdgfiZdef fdZee		 	 	 	 	 	 	 	 dd
e
j                  d	z  de
j                  d	z  de
j                  d	z  ded	z  de
j                  d	z  de
j                  d	z  ded	z  dee
j                  z  deez  fd              Z xZS )GraniteMoeSharedForCausalLMzlm_head.weightzmodel.embed_tokens.weightlm_headcolwise_gather_outputrJ   logitsr8   c                 p   t         |   |       t        |      | _        |j                  | _        t        j                  |j                  |j                  d      | _        |j                  | _	        |j                  | _        |j                  | _        |j                  | _        | j                          y )NFr;   )r=   r>   rH  r  rM  r   rD   r?   rs  router_aux_loss_coefrs   rt   ru   logits_scalingrX  rG   s     r4   r>   z$GraniteMoeSharedForCausalLM.__init__  s     *62
 ++yy!3!3V5F5FUS$*$?$?!!33#)#=#= $33 	r3   NrY  r   r  r   rZ  labelsoutput_router_logitslogits_to_keeprK   c	           	         ||n| j                   j                  } | j                  d|||||d|	}
|
j                  }t	        |t
              rt        | d      n|}| j                  |dd|ddf         }|| j                   j                  z  }d}|* | j                  ||fd| j                   j                  i|	}d}|rYt        |
j                  | j                  | j                  |      }|+|| j                  |j!                  |j"                        z  z  }t%        ||||
j&                  |
j(                  |
j*                  |
j                        S )ax  
        labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
            Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
            config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored
            (masked), the loss is only computed for the tokens with labels in `[0, ..., config.vocab_size]`.

        Example:

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

        >>> model = GraniteMoeSharedForCausalLM.from_pretrained("ibm/PowerMoE-3b")
        >>> tokenizer = AutoTokenizer.from_pretrained("ibm/PowerMoE-3b")

        >>> 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."
        ```N)rY  r   r  r   rZ  rM  )lossaux_lossru  r   rJ   r  r~   r2   )r8   rz  r  r\  r  r0   slicers  rx  loss_functionrM  rp  r~   rt   ru   rw  rc   r1  r   r   rJ   r  )rH   rY  r   r  r   rZ  ry  rz  r{  r   outputsrJ   slice_indicesru  r}  r~  s                   r4   rR   z#GraniteMoeSharedForCausalLM.forward  s|   J %9$D $++JjJj 	 $** 
)%+'
 
  118B>SV8W~ot4]kmA}a,?@A$++444%4%%  ;;11 	D /%%  ((	H !11HKK4LLL(#33!//))!//
 	
r3   )NNNNNNNr   )r)   r*   r+   _tied_weights_keys_tp_plan_pp_planr!   r>   r   r   r-   r.   rT   r
   r  r
  r0   rk   r   rR   rU   rV   s   @r4   rr  rr    s    *,GH23H_-z:;H5   .2.204(,26*.,0-.Q
##d*Q
 t+Q
 &&-	Q

 Q
 ((4/Q
   4'Q
 #TkQ
 ell*Q
 
*	*Q
  Q
r3   rr  )rr  rH  r  )r    )r   )Nr:   N)Jcollections.abcr   typingr   r   r-   torch.nn.functionalr   r   ry    r   r  activationsr	   cache_utilsr
   r   
generationr   integrationsr   r   r   r   masking_utilsr   modeling_layersr   modeling_outputsr   r   modeling_rope_utilsr   r   modeling_utilsr   r   processing_utilsr   utilsr   r   utils.genericr   r   r   utils.output_capturingr   configuration_granitemoesharedr!   r#   Moduler7   rY   rp   r   r   r   r   rT   r0   r   rn   r   r   r   r  r"  rH  rk   rp  rr  __all__r2   r3   r4   <module>r     s  * % &     & ! . )  0 9 Q K F & 7 Y Y 5 B)5 0")) 4 Y'Jbii J (J(9 9, $#bii $# $#N?")) ?"( *+ ,2	UU\\ 	U# 	U%,, 	U& %II%<<% 
% <<	%
 LL4'% % % '(%2 )*@)		 @) +@)F0#= 0f Uo U U4><bii ><B J
; J
 J
^ #
*.	O&ell 33d:O&tO& LL4'	O&
 \\CO&d e
"A? e
 e
P fr3   