
    rh0S                        d dl mZmZ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 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"m#Z# ddl$m%Z% ddl&m'Z'  ed       G d dejP                               Z) G d dejP                        Z*d Z+d5dZ,dejZ                  de.dejZ                  fdZ/	 d6dejP                  d ejZ                  d!ejZ                  d"ejZ                  d#eejZ                     d$e0d%e0d&ee!   fd'Z1 G d( d)ejP                        Z2 G d* d+e      Z3 G d, d-ejP                        Z4e" G d. d/e             Z5e" G d0 d1e5             Z6e" G d2 d3e5e             Z7g d4Z8y)7    )CallableOptionalUnionN)nn   )ACT2FN)CacheDynamicCache)GenerationMixin)use_kernel_forward_from_hub)create_causal_mask)FlashAttentionKwargs)GradientCheckpointingLayer)BaseModelOutputWithPastCausalLMOutputWithPast)ROPE_INIT_FUNCTIONSdynamic_rope_update)ALL_ATTENTION_FUNCTIONSPreTrainedModel)Unpack)TransformersKwargsauto_docstringcan_return_tuple)check_model_inputs   )BitNetConfigRMSNormc                   ,     e Zd Zd fd	Zd Zd Z xZS )BitNetRMSNormc                     t         |           t        j                  t	        j
                  |            | _        || _        y)z<
        BitNetRMSNorm is equivalent to T5LayerNorm
        N)super__init__r   	Parametertorchonesweightvariance_epsilon)selfhidden_sizeeps	__class__s      }/var/www/html/ai-insurance-compliance-backend/venv/lib/python3.12/site-packages/transformers/models/bitnet/modeling_bitnet.pyr"   zBitNetRMSNorm.__init__,   s1     	ll5::k#:; #    c                 "   |j                   }|j                  t        j                        }|j	                  d      j                  dd      }|t        j                  || j                  z         z  }| j                  |j                  |      z  S )N   T)keepdim)	dtypetor$   float32powmeanrsqrtr'   r&   )r(   hidden_statesinput_dtypevariances       r,   forwardzBitNetRMSNorm.forward4   sy    #))%((7 $$Q',,R,>%Ht?T?T4T(UU{{]--k:::r-   c                 ^    t        | j                  j                         d| j                   S )Nz, eps=)tupler&   shaper'   r(   s    r,   
extra_reprzBitNetRMSNorm.extra_repr;   s*    ))*+6$2G2G1HIIr-   )gư>)__name__
__module____qualname__r"   r;   r@   __classcell__r+   s   @r,   r   r   *   s    $;Jr-   r   c                   *     e Zd Zdef fdZd Z xZS )	BitNetMLPconfigc                    t         |           || _        |j                  | _        |j                  | _        t        j                  | j                  | j                  d      | _        t        j                  | j                  | j                  d      | _        t        j                  | j                  | j                  d      | _	        t        |j                     | _        t        |j                  |j                        | _        y )NFbiasr*   )r!   r"   rH   r)   intermediate_sizer   Linear	gate_projup_proj	down_projr   
hidden_actact_fnr   rms_norm_epsffn_sub_normr(   rH   r+   s     r,   r"   zBitNetMLP.__init__@   s    !--!'!9!94#3#3T5K5KRWXyy!1!143I3IPUV4#9#94;K;KRWXV../)&*B*BH[H[\r-   c           	          | j                  | j                  | j                  | j                  |            | j	                  |      z              }|S N)rQ   rU   rS   rO   rP   )r(   xrQ   s      r,   r;   zBitNetMLP.forwardK   sF    NN4#4#4T[[PQAR5SVZVbVbcdVe5e#fg	r-   )rA   rB   rC   r   r"   r;   rD   rE   s   @r,   rG   rG   ?   s    	]| 	]r-   rG   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..Nr0   r/   dim)r>   r$   cat)rY   x1x2s      r,   rotate_halfr`   P   sZ    	
3"!''"+"""	#B	
3q ""	#B99rc2YB''r-   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.
        position_ids (`torch.Tensor`, *optional*):
            Deprecated and unused.
        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position_idsunsqueeze_dimq_embedk_embeds           r,   apply_rotary_pos_embrk   W   sY    ( --
&C
--
&C3w;q>C/0G3w;q>C/0GGr-   r8   n_repreturnc                     | 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)r>   expandreshape)r8   rl   batchnum_key_value_headsslenhead_dims         r,   	repeat_kvru   r   so    
 2?1D1D.Ehz!!Qa"23::5BUW\^bdlmM  (;e(CT8TTr-   modulequerykeyvalueattention_maskscalingdropoutkwargsc                 T   t        || j                        }t        || j                        }	t        j                  ||j	                  dd            |z  }
|#|d d d d d d d |j
                  d   f   }|
|z   }
t        j                  j                  |
dt        j                        j                  |j                        }
t        j                  j                  |
|| j                        }
t        j                  |
|	      }|j	                  dd      j                         }||
fS )Nr/   r   r0   )r\   r2   )ptrainingr   )ru   num_key_value_groupsr$   matmul	transposer>   r   
functionalsoftmaxr4   r3   r2   r|   r   
contiguous)rv   rw   rx   ry   rz   r{   r|   r}   
key_statesvalue_statesattn_weightscausal_maskattn_outputs                r,   eager_attention_forwardr   ~   s    3 ; ;<JUF$?$?@L<<z';';Aq'ABWLL!$Q1.D
0@0@0D.D%DE#k1==((2U]](SVVW\WbWbcL==((6??([L,,|\:K''1-88:K$$r-   c                   6    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e
ej                     de
e   d	e
ej                     d
ee   de	ej                  e
ej                     e
e	ej                        f   fdZ xZS )BitNetAttentionz=Multi-headed attention from 'Attention Is All You Need' paperrH   	layer_idxc                    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*                        | _        y )Nrt   g      TrJ   rL   )r!   r"   rH   r   getattrr)   num_attention_headsrt   rr   r   r{   attention_dropout	is_causalr   rN   attention_biasq_projk_projv_projo_projr   rT   attn_sub_normr(   rH   r   r+   s      r,   r"   zBitNetAttention.__init__   sj   "
F4F4F&JdJd4de$*$>$>&B\B\$\!}}d*!'!9!9ii : :T]] JQWQfQf
 ii : :T]] JQWQfQf
 ii : :T]] JQWQfQf
 ii&&68J8JQWQfQf
 +6+=+=6CVCVWr-   r8   position_embeddingsrz   past_key_valuecache_positionr}   rm   c                 V   |j                   d d }g |d| j                  }| j                  |      j                  |      j	                  dd      }	| j                  |      j                  |      j	                  dd      }
| j                  |      j                  |      j	                  dd      }|\  }}t        |	|
||      \  }	}
|'|||d}|j                  |
|| j                  |      \  }
}t        }| j                  j                  dk7  rt        | j                  j                     } || |	|
||f| j                  sdn| j                  | j                   d|\  }} |j"                  g |d j%                         }| j'                  |      }| j)                  |      }||fS )Nr0   r   r/   )rf   re   r   eager        )r|   r{   )r>   rt   r   viewr   r   r   rk   updater   r   rH   _attn_implementationr   r   r   r{   rp   r   r   r   )r(   r8   r   rz   r   r   r}   input_shapehidden_shapequery_statesr   r   re   rf   cache_kwargsattention_interfacer   r   s                     r,   r;   zBitNetAttention.forward   s    $))#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%#&snUL'5'<'<ZW[WeWegs't$J(?;;++w6"9$++:Z:Z"[$7	%
  $}}C$2H2HLL	%
 	%
!\ *k));;;;FFH((5kk+.L((r-   )NN)rA   rB   rC   __doc__r   intr"   r$   Tensorr=   r   r	   
LongTensorr   r   r;   rD   rE   s   @r,   r   r      s    GX| X X: +/59+)||+) #5<<#=>+) !.	+)
 !+) !!1!12+) -.+) 
u||Xell3XeELL>Q5RR	S+)r-   r   c                   (    e Zd Zdedef fdZ	 	 	 	 	 	 ddej                  deej                     deej                     dee
   dee   d	eej                     d
eeej                  ej                  f      dee   deej                     fdZ xZS )BitNetDecoderLayerrH   r   c                     t         |           |j                  | _        t        ||      | _        t        |      | _        t        |j                  |j                        | _	        t        |j                  |j                        | _
        y )N)rH   r   rL   )r!   r"   r)   r   	self_attnrG   mlpr   rT   input_layernormpost_attention_layernormr   s      r,   r"   zBitNetDecoderLayer.__init__   sl    !--()LV$,V-?-?VEXEXY(5f6H6HfNaNa(b%r-   r8   rz   rg   r   	use_cacher   r   r}   rm   c                     |}	| j                  |      } | j                  d|||||||d|\  }}
|	|z   }|}	| j                  |      }| j                  |      }|	|z   }|S )N)r8   rz   rg   r   r   r   r    )r   r   r   r   )r(   r8   rz   rg   r   r   r   r   r}   residual_s              r,   r;   zBitNetDecoderLayer.forward   s     !,,];)4>> 	
')%)) 3	
 	
q !=0 !55mD/ =0r-   )NNNFNN)rA   rB   rC   r   r   r"   r$   r   r   r   r	   boolr=   r   r   r;   rD   rE   s   @r,   r   r      s    c| c c 2637*.$)59KO|| !. u//0	
 ! D> !!1!12 &eELL%,,,F&GH +, 
u||	r-   r   c                   ^     e Zd Zddef fdZ ej                         ed               Z xZ	S )BitNetRotaryEmbeddingrH   c                    t         |           t        |d      rUt        |j                  t
              r;|j                  j                  d|j                  j                  d            | _        nd| _        |j                  | _	        |j                  | _
        || _        t        | j                     | _        | j                  | j                  |      \  }| _        | j                  d|d       | j                   | _        y )Nrope_scaling	rope_typetypedefaultinv_freqF)
persistent)r!   r"   hasattr
isinstancer   dictgetr   max_position_embeddingsmax_seq_len_cachedoriginal_max_seq_lenrH   r   rope_init_fnattention_scalingregister_bufferr   original_inv_freq)r(   rH   devicer   r+   s       r,   r"   zBitNetRotaryEmbedding.__init__  s    6>*z&:M:Mt/T#0044[&BUBUBYBYZ`BabDN&DN"("@"@$*$B$B!/?+/+<+<T[[&+Q($(ZeD!%r-   c                 b   | 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        j                  |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   r0   r   mpscpuF)device_typeenabledr/   r[   )r2   )r   floatro   r>   r3   r   r   r   strr$   autocastr   r]   re   r   rf   r2   )
r(   rY   rg   inv_freq_expandedposition_ids_expandedr   freqsembre   rf   s
             r,   r;   zBitNetRotaryEmbedding.forward   sV    !MM$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$cff177f&;;;	5 	5s    BF%%F.rX   )
rA   rB   rC   r   r"   r$   no_gradr   r;   rD   rE   s   @r,   r   r     s3    /| /" U]]_<  <r-   r   c                   J    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y)BitNetPreTrainedModelrH   modelTr   past_key_values)r8   
attentionsN)rA   rB   rC   r   __annotations__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,   r   r   0  sQ    &*#-.#4"5N!"&+%r-   r   c                       e Zd Zdef fdZee	 	 	 	 	 	 	 ddeej                     deej                     deej                     dee   deej                     deej                     d	ee   d
ee   defd              Z xZS )BitNetModelrH   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)                          y c c}w )NrL   )rH   F)r!   r"   pad_token_idpadding_idx
vocab_sizer   	Embeddingr)   embed_tokens
ModuleListrangenum_hidden_layersr   layersr   rT   normr   
rotary_embgradient_checkpointing	post_initr   s      r,   r"   zBitNetModel.__init__E  s     !.. ++LL):):F<N<NPTP`P`ammDI&JbJbDcdy	2d
 "&"4"4&:M:MN	/v>&+# 	 es   D	input_idsrz   rg   r   inputs_embedsr   r   r}   rm   c           
      *   |d u |d uz  rt        d      || j                  |      }|r|
t               }|F||j                         nd}	t	        j
                  |	|	|j                  d   z   |j                        }||j                  d      }t        | j                  |||||      }
|}| j                  ||      }| j                  d | j                  j                   D ]  } ||f|
||||d|} | j                  |      }t        ||      S )Nz:You must specify exactly one of input_ids or inputs_embedsr   r   )r   )rH   input_embedsrz   r   r   rg   )rz   rg   r   r   r   )last_hidden_stater   )
ValueErrorr   r
   get_seq_lengthr$   aranger>   r   rb   r   rH   r   r   r   r   r   )r(   r   rz   rg   r   r   r   r   r}   past_seen_tokensr   r8   r   decoder_layers                 r,   r;   zBitNetModel.forwardU  sT    -t";<YZZ *.*;*;I*FM0*nO!CRC^==?de+0<< "2]5H5H5K"KTaThTh,N )33A6L(;;&))+%
 &"oom\J![[)H4;;+H+HI 		M)*).-$7 M		 		-0&++
 	
r-   )NNNNNNN)rA   rB   rC   r   r"   r   r   r   r$   r   r   r	   FloatTensorr   r   r   r   r;   rD   rE   s   @r,   r   r   C  s    |    151537+/5959$(8
E,,-8
 !.8
 u//0	8

 "%8
   1 128
 !!1!128
 D>8
 +,8
 
!8
  8
r-   r   c                   `    e Zd ZdgZdZdZ fdZd Zd Ze	e
	 	 	 	 	 	 	 	 	 ddeej                     deej                     deej                     d	ee   d
eej                      deej                     dee   deej                     deeej                  f   dee   defd              Z xZS )BitNetForCausalLMzlm_head.weightNc                     t         |   |       t        |      | _        |j                  | _        t        j                  |j                  |j                  d      | _        | j                          y )NFrJ   )
r!   r"   r   r   r   r   rN   r)   lm_headr   rV   s     r,   r"   zBitNetForCausalLM.__init__  sU      (
 ++yy!3!3V5F5FUS 	r-   c                     || _         y rX   r   )r(   decoders     r,   set_decoderzBitNetForCausalLM.set_decoder  s	    
r-   c                     | j                   S rX   r  r?   s    r,   get_decoderzBitNetForCausalLM.get_decoder  s    zzr-   r   rz   rg   r   r   labelsr   r   logits_to_keepr}   rm   c
                 z    | j                   d|||||||d|
}|j                  }t        |	t              rt	        |	 d      n|	}| j                  |dd|ddf         }d}|* | j                  d||| j                  j                  d|
}t        |||j                  |j                  |j                        S )a$  
        labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
            Labels for computing the masked language modeling loss. Indices should either be in `[0, transformers.,
            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, transformers., config.vocab_size]`.

        Example:

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

        >>> model = BitNetForCausalLM.from_pretrained("microsoft/bitnet-b1.58-2B-4T")
        >>> tokenizer = AutoTokenizer.from_pretrained("microsoft/bitnet-b1.58-2B-4T")

        >>> prompt = f'<|begin_of_text|>User: Hey, are you conscious? Can you talk to me?<|eot_id|>Assistant: '
        >>> inputs = tokenizer(prompt, return_tensors="pt")

        >>> # Generate
        >>> generate_ids = model.generate(inputs.input_ids, max_length=100)
        >>> tokenizer.batch_decode(generate_ids, skip_special_tokens=True, clean_up_tokenization_spaces=False)[0]
        "User: Hey, are you conscious? Can you talk to me?Assistant: No, I'm not conscious. I'm an artificial intelligence designed to assist with information and tasks. How can I help you today?"
        ```)r   rz   rg   r   r   r   r   N)logitsr  r   )lossr  r   r8   r   r   )r   r  r   r   slicer  loss_functionrH   r   r   r   r8   r   )r(   r   rz   rg   r   r   r  r   r   r  r}   outputsr8   slice_indicesr  r  s                   r,   r;   zBitNetForCausalLM.forward  s    J ,64:: 	,
)%+')	,
 	,
  118B>SV8W~ot4]kmA}a,?@A%4%%pVFt{{OeOepiopD%#33!//))
 	
r-   )	NNNNNNNNr   )rA   rB   rC   _tied_weights_keys_tp_plan_pp_planr"   r  r  r   r   r   r$   r   r   r	   r  r   r   r   r   r   r   r;   rD   rE   s   @r,   r
  r
    s&   *+HH  151537+/59-1$(5934=
E,,-=
 !.=
 u//0	=

 "%=
   1 12=
 ))*=
 D>=
 !!1!12=
 c5<</0=
 +,=
 
 =
  =
r-   r
  )r
  r   r   )Nr   )r   )9typingr   r   r   r$   r   activationsr   cache_utilsr	   r
   
generationr   integrationsr   masking_utilsr   modeling_flash_attention_utilsr   modeling_layersr   modeling_outputsr   r   modeling_rope_utilsr   r   modeling_utilsr   r   processing_utilsr   utilsr   r   r   utils.genericr   configuration_bitnetr   Moduler   rG   r`   rk   r   r   ru   r   r   r   r   r   r   r   r
  __all__r   r-   r,   <module>r0     s  * - ,   ! . ) 7 / B 9 O K F & I I / . Y'JBII J (J(		 "(6	UU\\ 	U# 	U%,, 	U& %II%<<% 
% <<	%
 U\\*% % % '(%4F)bii F)R*3 *Z<BII <D O  $ K
' K
 K
\ S
- S
 S
l Hr-   