
    iA`                     P   d Z ddlmZ ddlmZ ddlZddlmZ ddlmZ ddl	m
Z
mZ dd	lmZ dd
lmZ ddlmZmZmZ ddlmZmZ ddlmZmZ ddlmZmZ ddlmZ ddlm Z m!Z!m"Z"m#Z# ddl$m%Z%m&Z& ddl'm(Z( ddl)m*Z*  e#jV                  e,      Z- G d dej\                        Z/d Z0d9dZ1 G d dej\                        Z2 G d dej\                        Z3dejh                  de5dejh                  fd Z6	 d:d!ej\                  d"ejh                  d#ejh                  d$ejh                  d%ejh                  dz  d&e7d'e7d(ee    fd)Z8 G d* d+ej\                        Z9 G d, d-e      Z:e! G d. d/e             Z;e! G d0 d1e;             Z< G d2 d3e;e      Z= G d4 d5ee;      Z> G d6 d7ee;      Z?g d8Z@y);zPyTorch StableLM model.    )Callable)OptionalN)nn   )ACT2FN)CacheDynamicCache)GenerationMixin)create_causal_mask) GenericForSequenceClassificationGenericForTokenClassificationGradientCheckpointingLayer)BaseModelOutputWithPastCausalLMOutputWithPast)ROPE_INIT_FUNCTIONSdynamic_rope_update)ALL_ATTENTION_FUNCTIONSPreTrainedModel)Unpack)TransformersKwargsauto_docstringcan_return_tuplelogging)maybe_autocastmerge_with_config_defaults)capture_outputs   )StableLmConfigc                        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 )StableLmRotaryEmbeddinginv_freqNconfigc                    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)super__init__max_position_embeddingsmax_seq_len_cachedoriginal_max_seq_lenr"   rope_parametersr$   compute_default_rope_parametersr   attention_scalingregister_bufferclone)selfr"   devicerope_init_fnr!   	__class__s        /var/www/vps2.regionflexible.com/Desarrollo/venv/lib/python3.12/site-packages/transformers/models/stablelm/modeling_stablelm.pyr)   z StableLmRotaryEmbedding.__init__;   s    "("@"@$*$B$B!44[A!%!E!E>>Y&.t~~>L+7V+L($(ZeD0(..2BuU    r3   ztorch.deviceseq_lenreturnztorch.Tensorc                 n   | j                   d   }| j                   j                  dd      }t        | dd      xs | j                  | j                  z  }t        ||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_thetapartial_rotary_factorg      ?head_dimNr      dtype)r3   r@   )r-   getgetattrhidden_sizenum_attention_headsinttorcharangeint64tofloat)	r"   r3   r8   baser<   r=   dimattention_factorr!   s	            r6   r.   z7StableLmRotaryEmbedding.compute_default_rope_parametersK   s    ( %%l3 & 6 6 : :;RTW X6:t4h8J8JfNhNh8h(223 U\\!S!5;;?BB&X]XcXcBdgjjk
 )))r7   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   r   mpscpuF)device_typeenabledr>   rL   r?   )r!   rJ   expandshaperI   r3   
isinstancetypestrr   	transposerF   catcosr/   sinr@   )
r2   xposition_idsinv_freq_expandedposition_ids_expandedrR   freqsembr\   r]   s
             r6   forwardzStableLmRotaryEmbedding.forwardl   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$N)NNN)__name__
__module____qualname__rF   Tensor__annotations__r   r)   staticmethodr   rE   tuplerJ   r.   no_gradr   rd   __classcell__r5   s   @r6   r    r    8   s    llV~ V   )-+/"*%*(* t* 
~u$	%	* *> U]]_<  <r7   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..NrO   r>   rT   )rV   rF   r[   )r^   x1x2s      r6   rotate_halfrs   }   sZ    	
3"!''"+"""	#B	
3q ""	#B99rc2YB''r7   c                     |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.
    )	unsqueezers   )qkr\   r]   unsqueeze_dimq_embedk_embeds          r6   apply_rotary_pos_embr{      sY    $ --
&C
--
&C3w;q>C/0G3w;q>C/0GGr7   c                   $     e Zd Z fdZd Z xZS )StableLmMLPc                    t         |           || _        |j                  | _        |j                  | _        t        j                  | j                  | j                  d      | _        t        j                  | j                  | j                  d      | _        t        j                  | j                  | j                  d      | _	        t        |j                     | _        y NFbias)r(   r)   r"   rC   intermediate_sizer   Linear	gate_projup_proj	down_projr   
hidden_actact_fnr2   r"   r5   s     r6   r)   zStableLmMLP.__init__   s    !--!'!9!94#3#3T5K5KRWXyy!1!143I3IPUV4#9#94;K;KRWXV../r7   c                     | j                  | j                  | j                  |            | j                  |      z        }|S re   )r   r   r   r   )r2   r^   r   s      r6   rd   zStableLmMLP.forward   s6    NN4;;t~~a/@#ADLLQRO#ST	r7   )rf   rg   rh   r)   rd   rn   ro   s   @r6   r}   r}      s    0r7   r}   c                   @     e Zd Zd fd	Zdej
                  fdZ xZS )StableLmLayerNormPerHeadc           
          t         |           || _        || _        t	        j
                  t        | j                        D cg c]  }t	        j                  |||       c}      | _        y c c}w )N)epsr   )	r(   r)   rL   	num_headsr   
ModuleListrange	LayerNormnorms)r2   rL   r   r   r   _r5   s         r6   r)   z!StableLmLayerNormPerHead.__init__   sT    "]]SXY]YgYgSh#iaBLL#D$I#ij
#is   A0hidden_statesc           	          t        j                  |dd      }t        j                  t        | j                  |      D cg c]  \  }} ||       c}}d      S c c}}w )Nr   rT   )rF   splitr[   zipr   )r2   r   states_per_headsnorms       r6   rd   z StableLmLayerNormPerHead.forward   sM     !;;}aQ?yyTZZYiIjk2E$$}-kqrssks   A
)gh㈵>F)rf   rg   rh   r)   rF   ri   rd   rn   ro   s   @r6   r   r      s    ktU\\ tr7   r   r   n_repr9   c                     | j                   \  }}}}|dk(  r| S | dddddddddf   j                  |||||      } | j                  |||z  ||      S )z
    This is the equivalent of torch.repeat_interleave(x, dim=1, repeats=n_rep). The hidden states go from (batch,
    num_key_value_heads, seqlen, head_dim) to (batch, num_attention_heads, seqlen, head_dim)
    r   N)rV   rU   reshape)r   r   batchnum_key_value_headsslenr=   s         r6   	repeat_kvr      so    
 2?1D1D.Ehz!!Qa"23::5BUW\^bdlmM  (;e(CT8TTr7   modulequerykeyvalueattention_maskscalingdropoutkwargsc                    t        || j                        }t        || j                        }	t        j                  ||j	                  dd            |z  }
||
|z   }
t
        j                  j                  |
dt        j                        j                  |j                        }
t
        j                  j                  |
|| j                        }
t        j                  |
|	      }|j	                  dd      j                         }||
fS )Nr>   r   rO   )rL   r@   )ptrainingr   )r   num_key_value_groupsrF   matmulrZ   r   
functionalsoftmaxfloat32rI   r@   r   r   
contiguous)r   r   r   r   r   r   r   r   
key_statesvalue_statesattn_weightsattn_outputs               r6   eager_attention_forwardr      s     3 ; ;<JUF$?$?@L<<z';';Aq'ABWLL!#n4==((2U]](SVVW\WbWbcL==((6??([L,,|\:K''1-88:K$$r7   c                   J    e Zd ZdZddededz  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edeej                  ej                  f   dz  deej                  ej                  dz  eej                     dz  f   fdZ xZS )StableLmAttentionz=Multi-headed attention from 'Attention Is All You Need' paperNr"   	layer_idxc                    t         |           || _        || _        |-t        j                  d| j                  j                   d       |j                  | _        |j                  | _
        | j                  | j                  z  | _        |j                  | _        | j                  | j                  z  | _        t        | j                  |j                  d   z        | _        d| _        | j                  dz  | _        | j                  | j                  z  | j                  k7  r&t'        d| j                   d| j                   d      t)        j*                  | j                  | j                  | j                  z  |j,                  	      | _        t)        j*                  | j                  | j                  | j                  z  |j,                  	      | _        t)        j*                  | j                  | j                  | j                  z  |j,                  	      | _        t)        j*                  | j                  | j                  d
	      | _        |j6                  | _        | j6                  rbt9        | j                  | j                  |j:                        | _        t9        | j                  | j                  |j:                        | _        |j@                  | _         y )NzInstantiating z without passing a `layer_idx` is not recommended and will lead to errors during the forward call if caching is used. Please make sure to provide a `layer_idx` when creating this class.r<   Tg      z?hidden_size must be divisible by num_heads (got `hidden_size`: z and `num_heads`: z).r   Fr   )!r(   r)   r"   r   loggerwarning_oncer5   rf   rC   rD   r   r=   r   r   rE   r-   rotary_ndims	is_causalr   
ValueErrorr   r   use_qkv_biasq_projk_projv_projo_projqk_layernormr   layer_norm_epsq_layernormk_layernormattention_dropoutr2   r"   r   r5   s      r6   r)   zStableLmAttention.__init__   sB   " !8!8 9 :, , "--33((DNN:#)#=#= $(NNd6N6N$N!0F0FG^0_ _`}}d*MMDNN*t/?/??QRVRbRbQc$T^^$4B8  ii 0 0$..4==2PW]WjWjkii 0 0$2J2JT]]2Zagatatuii 0 0$2J2JT]]2Zagatatuii 0 0$2B2BO"//7t~~[a[p[pqD7t77V=R=R D "(!9!9r7   r   r   r_   past_key_valuesoutput_attentions	use_cacheposition_embeddingsr9   c                    |j                         \  }	}
}| j                  |      }| j                  |      }| j                  |      }|j	                  |	|
| j
                  | j                        j                  dd      }|j	                  |	|
| j                  | j                        j                  dd      }|j	                  |	|
| j                  | j                        j                  dd      }| j                  r"| j                  |      }| j                  |      }|\  }}|dd | j                  f   |d| j                  d f   }}|dd | j                  f   |d| j                  d f   }}t        ||||      \  }}t        j                  ||fd      }t        j                  ||fd      }| |j!                  ||| j"                        \  }}t%        j&                  | j(                  j*                  t,              } || ||||f| j.                  sdn| j0                  | j2                  |d|\  }}|j5                  |	|
d      }| j7                  |      }||fS )Nr   r>   .rO   rT           )r   r   r_   )sizer   r   r   viewr   r=   rZ   r   r   r   r   r   r{   rF   r[   updater   r   get_interfacer"   _attn_implementationr   r   r   r   r   r   )r2   r   r   r_   r   r   r   r   r   bszq_lenr   query_statesr   r   r\   r]   	query_rot
query_passkey_rotkey_passattention_interfacer   r   s                           r6   rd   zStableLmAttention.forward  sp    &**,UA{{=1[[/
{{=1#((eT^^T]]S]]^_abc__S%1I1I4==Yccdeghi
#((eT5M5Mt}}]gghiklm++L9L))*5J&S1 1 1112d//112 	
 s/d////0sD--//0 
 2)Wc3O	7 yy)Z!8bAYY2;
&'6'='=j,X\XfXf'g$J(?(M(MKK,,.E)
 %8
%
  $}}C$2H2HLL%
%
 
%
!\ "))#ub9kk+.L((r7   re   )NNNFFN)rf   rg   rh   __doc__r   rE   r)   rF   ri   
LongTensorr   boolrl   rd   rn   ro   s   @r6   r   r      s    G&:~ &:#* &:V /304(,"'HL?)||?) t+?) &&-	?)
 ?)  ?) ?) #5<<#=>E?) 
u||U\\D0%2E2LL	M?)r7   r   c                        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ej                  ej                  f   dz  dej                  fdZ xZS )StableLmDecoderLayerr"   r   c                    t         |           |j                  | _        |j                  | _        t	        ||      | _        t        |      | _        t        j                  |j                  |j                        | _        d | _        | j                  s0t        j                  |j                  |j                        | _        t        j                  |j                        | _        y )N)r   r   )r(   r)   use_parallel_residualrC   r   	self_attnr}   mlpr   r   r   input_layernormpost_attention_layernormDropouthidden_dropoutr   r   s      r6   r)   zStableLmDecoderLayer.__init__S  s    %+%A%A"!--*6YGv&!||F,>,>FDYDYZ(,%)),.LL9K9KQWQfQf,gD)zz&"7"78r7   Nr   r   r_   r   r   r   r9   c                 F   |}| j                  |      }| j                  ||||||      \  }	}
| j                  r,| j                  |      }| j	                  |      }||	z   |z   }|S ||	z   }| j                  | j                  |            }| j	                  |      }||z   }|S )N)r   r   r_   r   r   r   )r   r   r   r   r   r   )r2   r   r   r_   r   r   r   r   residualself_attn_outputr   
mlp_outputs               r6   rd   zStableLmDecoderLayer.forward_  s     !,,]; #nn')%+ 3 - 
! %% -0Jj1J$'77*DM   "22H$"?"?"IJJj1J$z1Mr7   )NNNFN)rf   rg   rh   r   rE   r)   rF   ri   r   r   r   rl   rd   rn   ro   s   @r6   r   r   R  s    
9~ 
9# 
9 /304(,!&HL(||( t+( &&-	(
 ( $;( #5<<#=>E( 
(r7   r   c                   @    e Zd ZU eed<   dZdZdgZdZdZ	dZ
dZeedZy)StableLmPreTrainedModelr"   modelTr   r   )r   
attentionsN)rf   rg   rh   r   rj   base_model_prefixsupports_gradient_checkpointing_no_split_modules_skip_keys_device_placement_supports_flash_attn_supports_sdpa_can_compile_fullgraphr   r   _can_record_outputs r7   r6   r   r     sB    &*#/0"3N!-'r7   r   c                        e Zd 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 )StableLmModelz
    Transformer decoder consisting of *config.num_hidden_layers* layers. Each layer is a [`StableLmDecoderLayer`]

    Args:
        config: StableLmConfig
    r"   c           	      V   t         |   |       |j                  | _        |j                  | _        t        j                  |j                  |j                  | j                        | _        t        j                  t        |j                        D cg c]  }t        ||       c}      | _        t        j                  |j                  |j                        | _        |j"                  | _        d| _        t'        | j(                        | _        | j-                          y c c}w )Nr   Fr"   )r(   r)   pad_token_idpadding_idx
vocab_sizer   	EmbeddingrC   embed_tokensr   r   num_hidden_layersr   layersr   r   r   r   gradient_checkpointingr    r"   
rotary_emb	post_initr   s      r6   r)   zStableLmModel.__init__  s     !.. ++LL):):F<N<NPTP`P`ammFKFLdLdFef!&)4f
 LL!3!39N9NO	$*$?$?!&+#1E 	 gs   D&N	input_idsr   r_   r   inputs_embedsr   r   r9   c           
         |d u |d uz  rt        d      |r|t        | j                        }|| j                  |      }|V||j	                         nd}t        j                  |j                  d   |j                        |z   }|j                  d      }t        | j                  ||||      }	|}
| j                  |
|      }| j                  D ]  } ||
f|	||||d|}
 | j                  |
      }
t        |
|	      S )
Nz:You must specify exactly one of input_ids or inputs_embedsr  r   r   )r3   )r"   r  r   r   r_   )r_   )r   r_   r   r   r   )last_hidden_stater   )r   r	   r"   r  get_seq_lengthrF   rG   rV   r3   ru   r   r
  r  r   r   )r2   r  r   r_   r   r  r   r   past_seen_tokenscausal_maskr   r   decoder_layers                r6   rd   zStableLmModel.forward  s:    -t";<YZZ0*$++>O  --i8MCRC^==?de <<(;(;A(>}G[G[\_ooL'11!4L(;;')+%
 &"oom,oW![[ 		M)*) /#$7 M		 		-0&++
 	
r7   )NNNNNN)rf   rg   rh   r   r   r)   r   r   r   rF   r   ri   r   FloatTensorr   r   r   r   rd   rn   ro   s   @r6   r   r     s    ~ $   .2.204(,26!%3
##d*3
 t+3
 &&-	3

 3
 ((4/3
 $;3
 +,3
 
!3
    3
r7   r   c                   *    e Zd ZddiZ 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 xZS )StableLmForCausalLMzlm_head.weightzmodel.embed_tokens.weightc                     t         |   |       t        |      | _        |j                  | _        t        j                  |j                  |j                  d      | _        | j                          y r   )
r(   r)   r   r   r  r   r   rC   lm_headr  r   s     r6   r)   zStableLmForCausalLM.__init__  sU     "6*
 ++yy!3!3V5F5FUS 	r7   Nr  r   r_   r   r  labelsr   logits_to_keepr   r9   c	           
      x    | j                   d||||||d|	}
|
j                  }t        |t              rt	        | d      n|}| j                  |dd|ddf         }d}|* | j                  ||fd| j                  j                  i|	}t        |||
j                  |
j                  |
j                        S )ui  
        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, StableLmForCausalLM

        >>> model = StableLmForCausalLM.from_pretrained("adept/persimmon-8b-base")
        >>> tokenizer = AutoTokenizer.from_pretrained("adept/persimmon-8b-base")

        >>> prompt = "human: Hey, what should I eat for dinner?"
        >>> 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]
        'human: Hey, what should I eat for dinner?\n\ncat: 🐱\n\nhuman: 😐\n\n'
        ```)r  r   r_   r   r  r   Nr  )losslogitsr   r   r   r   )r   r  rW   rE   slicer  loss_functionr"   r  r   r   r   r   )r2   r  r   r_   r   r  r  r   r  r   outputsr   slice_indicesr  r  s                  r6   rd   zStableLmForCausalLM.forward  s    J ,64:: ,
)%+',
 ,
  118B>SV8W~ot4]kmA}a,?@A%4%%ffbAWAWb[abD%#33!//))
 	
r7   )NNNNNNNr   )rf   rg   rh   _tied_weights_keysr)   r   r   rF   r   ri   r   r  r   rE   r   r   r   rd   rn   ro   s   @r6   r  r    s    *,GH  .2.204(,26*.!%-.:
##d*:
 t+:
 &&-	:

 :
 ((4/:
   4':
 $;:
 ell*:
 +,:
 
 :
  :
r7   r  c                       e Zd Zy)!StableLmForSequenceClassificationNrf   rg   rh   r   r7   r6   r$  r$  <      r7   r$  c                       e Zd Zy)StableLmForTokenClassificationNr%  r   r7   r6   r(  r(  ?  r&  r7   r(  )r  r   r   r$  r(  )r   )r   )Ar   collections.abcr   typingr   rF   r   activationsr   cache_utilsr   r	   
generationr
   masking_utilsr   modeling_layersr   r   r   modeling_outputsr   r   modeling_rope_utilsr   r   modeling_utilsr   r   processing_utilsr   utilsr   r   r   r   utils.genericr   r   utils.output_capturingr   configuration_stablelmr   
get_loggerrf   r   Moduler    rs   r{   r}   r   ri   rE   r   rJ   r   r   r   r   r   r  r$  r(  __all__r   r7   r6   <module>r;     s  &  $    ! . ) / 
 G & R R G 5 2 
		H	%A<bii A<J(4"))  tryy t 	UU\\ 	U# 	U%,, 	U( %II%<<% 
% <<	%
 LL4'% % % '(%2j)		 j)Z55 5p o   P
+ P
 P
hJ
1? J
Z h(HJa g b%BD[ ar7   