
    qi;_                     `   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 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& ddl'm(Z(m)Z) ddl*m+Z+ ddl,m-Z-  G d dej\                        Z/ G d dej\                        Z0d Z1dejd                  de3dejd                  fdZ4	 d;dej\                  d ejd                  d!ejd                  d"ejd                  d#ejd                  dz  d$e5d%e5d&e"e$   fd'Z6d<d(Z7 G d) d*ej\                        Z8 ed+       G d, d-ej\                               Z9 G d. d/e      Z:e% G d0 d1e              Z;e% G d2 d3e;             Z<e% G d4 d5e;e             Z= G d6 d7ee;      Z> G d8 d9ee;      Z?g d:Z@y)=    )Callable)OptionalN)nn   )ACT2FN)CacheDynamicCache)GenerationMixin)use_kernel_forward_from_hub)create_causal_mask!create_sliding_window_causal_mask)FlashAttentionKwargs) GenericForSequenceClassificationGenericForTokenClassificationGradientCheckpointingLayer)BaseModelOutputWithPastCausalLMOutputWithPast)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   )
Phi3Configc                   V     e Zd Z fdZdej
                  dej
                  fdZ xZS )Phi3MLPc                 *   t         |           || _        t        j                  |j
                  d|j                  z  d      | _        t        j                  |j                  |j
                  d      | _        t        |j                     | _        y )N   Fbias)super__init__configr   Linearhidden_sizeintermediate_sizegate_up_proj	down_projr   
hidden_actactivation_fnselfr)   	__class__s     X/opt/pipecat/venv/lib/python3.12/site-packages/transformers/models/phi3/modeling_phi3.pyr(   zPhi3MLP.__init__2   sp    IIf&8&8!f>V>V:V]bc6#;#;V=O=OV[\#F$5$56    hidden_statesreturnc                     | j                  |      }|j                  dd      \  }}|| j                  |      z  }| j                  |      S )Nr$   dim)r-   chunkr0   r.   )r2   r6   	up_statesgates       r4   forwardzPhi3MLP.forward:   sL    %%m4	#//!/4i 2 24 88	~~i((r5   )__name__
__module____qualname__r(   torchFloatTensorr?   __classcell__r3   s   @r4   r"   r"   1   s'    7)U%6%6 )5;L;L )r5   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 )Phi3RotaryEmbeddinginv_freqNr)   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defaultrI   F)
persistentoriginal_inv_freq)r'   r(   max_position_embeddingsmax_seq_len_cachedoriginal_max_seq_lenr)   rope_parametersrK   compute_default_rope_parametersr   attention_scalingregister_bufferclone)r2   r)   devicerope_init_fnrI   r3   s        r4   r(   zPhi3RotaryEmbedding.__init__F   s    "("@"@$*$B$B!44[A!%!E!E>>Y&.t~~>L+7V+L($(ZeD0(..2BuUr5   rW   ztorch.deviceseq_lenr7   z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   r$   dtype)rW   r_   )rR   getgetattrr+   num_attention_headsintrC   arangeint64tofloat)	r)   rW   rY   baser\   r]   r;   attention_factorrI   s	            r4   rS   z3Phi3RotaryEmbedding.compute_default_rope_parametersV   s    & %%l3 & 6 6 : :;RTW X6:t4h8J8JfNhNh8h(223 U\\!S!5;;?BB&X]XcXcBdgjjk
 )))r5   c                 N   | j                   d d d d f   j                         j                  |j                  d   dd      j	                  |j
                        }|d d d d d f   j                         }t        |j
                  j                  t              r/|j
                  j                  dk7  r|j
                  j                  nd}t        |d      5  |j                         |j                         z  j                  dd      }t        j                  ||fd	      }|j                         | j                  z  }|j                         | j                  z  }	d d d        j	                  |j                   
      	j	                  |j                   
      fS # 1 sw Y   AxY w)Nr   r9   r   mpscpuF)device_typeenabledr$   r:   r^   )rI   rg   expandshaperf   rW   
isinstancetypestrr   	transposerC   catcosrT   sinr_   )
r2   xposition_idsinv_freq_expandedposition_ids_expandedrm   freqsembrv   rw   s
             r4   r?   zPhi3RotaryEmbedding.forwardv   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$N)NNN)r@   rA   rB   rC   Tensor__annotations__r    r(   staticmethodr   rc   tuplerg   rS   no_gradr   r?   rE   rF   s   @r4   rH   rH   C   s    llVz V  $(+/"*T!*(* t* 
~u$	%	* *> U]]_<  <r5   rH   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..Nr9   r$   r:   )rp   rC   ru   )rx   x1x2s      r4   rotate_halfr      sZ    	
3"!''"+"""	#B	
3q ""	#B99rc2YB''r5   r6   n_repr7   c                     | j                   \  }}}}|dk(  r| S | dddddddddf   j                  |||||      } | j                  |||z  ||      S )z
    This is the equivalent of torch.repeat_interleave(x, dim=1, repeats=n_rep). The hidden states go from (batch,
    num_key_value_heads, seqlen, head_dim) to (batch, num_attention_heads, seqlen, head_dim)
    r   N)rp   ro   reshape)r6   r   batchnum_key_value_headsslenr]   s         r4   	repeat_kvr      so    
 2?1D1D.Ehz!!Qa"23::5BUW\^bdlmM  (;e(CT8TTr5   modulequerykeyvalueattention_maskscalingdropoutkwargsc                    t        || j                        }t        || j                        }	t        j                  ||j	                  dd            |z  }
||
|z   }
t
        j                  j                  |
dt        j                        j                  |j                        }
t
        j                  j                  |
|| j                        }
t        j                  |
|	      }|j	                  dd      j                         }||
fS )Nr$   r   r9   )r;   r_   )ptrainingr   )r   num_key_value_groupsrC   matmulrt   r   
functionalsoftmaxfloat32rf   r_   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$$r5   c                 `   |j                  |      }|j                  |      }|j                  d   }| dd|f   | d|df   }}|dd|f   |d|df   }	}t        j                  ||z  t	        |      |z  z   |gd      }
t        j                  ||z  t	        |      |z  z   |	gd      }|
|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.
    r9   .Nr:   )	unsqueezerp   rC   ru   r   )qkrv   rw   unsqueeze_dim
rotary_dimq_rotq_passk_rotk_passq_embedk_embeds               r4   apply_rotary_pos_embr      s    $ --
&C
--
&C2Jc;J;&'3
+;)<6Ec;J;&'3
+;)<6Eii%#++e*<s*BCVLRTUGii%#++e*<s*BCVLRTUGGr5   c                   >    e Zd ZdZddededz  f fdZ	 	 ddej                  de	ej                  ej                  f   dej                  dz  d	e
dz  d
ej                  dz  dee   de	ej                  ej                  dz  e	ej                     dz  f   fdZ xZS )Phi3Attentionz=Multi-headed attention from 'Attention Is All You Need' paperNr)   	layer_idxc                 |   t         |           || _        || _        t	        |d|j
                  |j                  z        | _        |j                  |j                  z  | _	        |j                  | _        | j                  dz  | _
        |j                  | _        d| _        |j                  | j                  z  d|j                  | j                  z  z  z   }t        j                  |j                  | j                  z  |j
                  d      | _        t        j                  |j
                  |d      | _        y )Nr]   g      Tr$   Fr%   )r'   r(   r)   r   ra   r+   rb   r]   r   r   r   attention_dropout	is_causalr   r*   o_projqkv_proj)r2   r)   r   op_sizer3   s       r4   r(   zPhi3Attention.__init__   s    "
F4F4F&JdJd4de$*$>$>&B\B\$\!#)#=#= }}d*!'!9!9,,t}}<qFD^D^aeananDn?ooii : :T]] JFL^L^ejk		&"4"4gEJr5   r6   position_embeddingsr   past_key_valuescache_positionr   r7   c           
         |j                   d d }g |d| j                  }| j                  |      }	| j                  j                  | j                  z  }
|	dd |
f   }|	d|
|
| j
                  | j                  z  z   f   }|	d|
| j
                  | j                  z  z   d f   }|j                  |      j                  dd      }|j                  |      j                  dd      }|j                  |      j                  dd      }|\  }}t        ||||      \  }}|'|||d}|j                  ||| j                  |      \  }}t        j                  | j                  j                  t              } || ||||f| j                  sdn| j                   | j"                  t%        | j                  dd       d|\  }} |j&                  g |d j)                         }| j+                  |      }||fS )	Nr9   .r   r$   )rw   rv   r           sliding_window)r   r   r   )rp   r]   r   r)   rb   r   viewrt   r   updater   r   get_interface_attn_implementationr   r   r   r   ra   r   r   r   )r2   r6   r   r   r   r   r   input_shapehidden_shapeqkv	query_posquery_statesr   r   rv   rw   cache_kwargsattention_interfacer   r   s                       r4   r?   zPhi3Attention.forward   s    $))#2.88b8$--8mmM*KK33dmmC	3

?+i)d6N6NQUQ^Q^6^*^^^_
3	D,D,Dt}},T T VVW#((6@@AF__\2<<QB
#((6@@AF&S#7jRUWZ#[ j&#&snUL'6'='=j,X\XfXfht'u$J(?(M(MKK,,.E)
 %8
%
  $}}C$2H2HLL"4;;0@$G
%
 
%
!\ *k));;;;FFHkk+.L((r5   r~   )NN)r@   rA   rB   __doc__r    rc   r(   rC   r   r   r   
LongTensorr   r   r?   rE   rF   s   @r4   r   r      s    GKz KcDj K( )-260)||0) #5<<#=>0) t+	0)
 0) ((4/0) -.0) 
u||U\\D0%2E2LL	M0)r5   r   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 )	Phi3RMSNormepsr7   Nc                     t         |           t        j                  t	        j
                  |            | _        || _        y)z:
        Phi3RMSNorm is equivalent to T5LayerNorm
        N)r'   r(   r   	ParameterrC   onesweightvariance_epsilon)r2   r+   r   r3   s      r4   r(   zPhi3RMSNorm.__init__  s1     	ll5::k#:; #r5   r6   c                 "   |j                   }|j                  t        j                        }|j	                  d      j                  dd      }|t        j                  || j                  z         z  }| j                  |j                  |      z  S )Nr$   r9   T)keepdim)	r_   rf   rC   r   powmeanrsqrtr   r   )r2   r6   input_dtypevariances       r4   r?   zPhi3RMSNorm.forward  sy    #))%((7 $$Q',,R,>%Ht?T?T4T(UU{{]--k:::r5   c                 ^    t        | j                  j                         d| j                   S )Nz, eps=)r   r   rp   r   )r2   s    r4   
extra_reprzPhi3RMSNorm.extra_repr&  s*    ))*+6$2G2G1HIIr5   )gư>)
r@   rA   rB   rg   r(   rC   r   r?   r   rE   rF   s   @r4   r   r     s7    $ $$ $;U\\ ;ell ;Jr5   r   c                   d    e Zd Zdedef fdZ	 	 	 	 	 	 ddej                  dej                  dz  dej                  dz  de	dz  d	e
dz  d
ej                  dz  deej                  ej                  f   dz  dee   deej                  eej                  ej                  f   dz  f   fdZ xZS )Phi3DecoderLayerr)   r   c                    t         |           |j                  | _        t        ||      | _        t        |      | _        t        |j                  |j                        | _	        t        |j                  |j                        | _
        || _        t        j                  |j                        | _        t        j                  |j                        | _        y )N)r)   r   r   )r'   r(   r+   r   	self_attnr"   mlpr   rms_norm_epsinput_layernormpost_attention_layernormr)   r   Dropoutresid_pdropresid_attn_dropoutresid_mlp_dropoutr2   r)   r   r3   s      r4   r(   zPhi3DecoderLayer.__init__+  s    !--&f	J6?*6+=+=6CVCVW(3F4F4FFL_L_(`%"$**V-?-?"@!#F,>,>!?r5   Nr6   r   ry   r   	use_cacher   r   r   r7   c                     |}	| j                  |      } | j                  d|||||||d|\  }}
|	| j                  |      z   }|}	| j                  |      }| j	                  |      }|	| j                  |      z   }|S )N)r6   r   ry   r   r   r   r    )r   r   r   r   r   r   )r2   r6   r   ry   r   r   r   r   r   residualself_attn_weightss              r4   r?   zPhi3DecoderLayer.forward6  s     !,,];+94>> 	,
')%+) 3	,
 	,
(( !4#:#:=#II 55mD/ 4#9#9-#HHr5   )NNNFNN)r@   rA   rB   r    rc   r(   rC   r   r   r   boolr   r   r   rD   r?   rE   rF   s   @r4   r   r   *  s    	@z 	@c 	@ /304(,!&26HL|| t+ &&-	
  $; ((4/ #5<<#=>E -. 
u  %(9(95;L;L(L"MPT"TT	Ur5   r   c                   N    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dZy)	Phi3PreTrainedModelr)   modelTr   r   )r6   
attentionsz0.0.5N)r@   rA   rB   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_outputs_versionr   r5   r4   r   r   W  sX    &*#+,#4"5N!"&)# Hr5   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j                  dz  dee   defd                     Z xZS )	Phi3Modelr)   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 )Nr   r)   F)r'   r(   pad_token_idpadding_idx
vocab_sizer   	Embeddingr+   embed_tokens
ModuleListrangenum_hidden_layersr   layersr   r   normrH   
rotary_embgradient_checkpointing	post_initr   s      r4   r(   zPhi3Model.__init__m  s     !.. ++LL):):F<N<NPTP`P`ammBGH`H`BabYfi0b
   2 28K8KL	-V<&+# 	 cs   DN	input_idsr   ry   r   inputs_embedsr   r   r   r7   c                    |d u |d uz  rt        d      || j                  |      }|r|t        | j                        }|F||j	                         nd}	t        j                  |	|	|j                  d   z   |j                        }||j                  d      }| j                  j                  t        nt        }
 |
| j                  |||||      }|}| j                  ||      }| j                  d | j                  j                   D ]  } ||f||||||d|} | j!                  |      }t#        ||r|	      S d 	      S )
Nz:You must specify exactly one of input_ids or inputs_embedsr  r   r   )rW   )r)   r  r   r   r   ry   )ry   )r   ry   r   r   r   r   )last_hidden_stater   )
ValueErrorr  r	   r)   get_seq_lengthrC   rd   rp   rW   r   r   r   r   r  r  r  r  r   )r2   r  r   ry   r   r  r   r   r   past_seen_tokensmask_functioncausal_maskr6   r   decoder_layers                  r4   r?   zPhi3Model.forward}  s    -t";<YZZ  --i8M0*$++>O!CRC^==?de"\\ "2]5H5H5K"KTaThThN )33A6L.2kk.H.H.P*Vw#;;'))+%
 &"oom,oW![[)H4;;+H+HI 
	M)	*) /#-$7	 	M
	 		-0&+/8O
 	
>B
 	
r5   )NNNNNNN)r@   rA   rB   r    r(   r   r   r   rC   r   r   r   rD   r   r   r   r   r?   rE   rF   s   @r4   r  r  k  s    z     .2.204(,26!%269
##d*9
 t+9
 &&-	9

 9
 ((4/9
 $;9
 ((4/9
 +,9
 
!9
    9
r5   r  c                   |    e Zd ZddiZddiZddgdgf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	j                  dz  dee	j                  z  dee   defd              Z	 	 	 	 	 	 	 d fd	Z xZS )Phi3ForCausalLMzlm_head.weightzmodel.embed_tokens.weightlm_headcolwise_gather_outputr6   logitsc                     t         |   |       t        |      | _        |j                  | _        t        j                  |j                  |j                  d      | _        | j                          y )NFr%   )
r'   r(   r  r   r	  r   r*   r+   r   r  r1   s     r4   r(   zPhi3ForCausalLM.__init__  sU     v&
 ++yy!3!3V5F5FUS 	r5   Nr  r   ry   r   r  labelsr   r   logits_to_keepr   r7   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  
        Example:

        ```python
        >>> from transformers import AutoTokenizer, Phi3ForCausalLM

        >>> model = Phi3ForCausalLM.from_pretrained("meta-phi3/Phi3-2-7b-hf")
        >>> tokenizer = AutoTokenizer.from_pretrained("meta-phi3/Phi3-2-7b-hf")

        >>> prompt = "Hey, are you conscious? Can you talk to me?"
        >>> inputs = tokenizer(prompt, return_tensors="pt")

        >>> # Generate
        >>> generate_ids = model.generate(inputs.input_ids, max_length=30)
        >>> tokenizer.batch_decode(generate_ids, skip_special_tokens=True, clean_up_tokenization_spaces=False)[0]
        "Hey, are you conscious? Can you talk to me?\nI'm not conscious, but I can talk to you."
        ```)r  r   ry   r   r  r   r   N)r"  r$  r	  )lossr"  r   r6   r   r   )r   r  rq   rc   slicer   loss_functionr)   r	  r   r   r6   r   )r2   r  r   ry   r   r  r$  r   r   r%  r   outputsr6   slice_indicesr"  r'  s                   r4   r?   zPhi3ForCausalLM.forward  s    @ ,64:: 	,
)%+')	,
 	,
  118B>SV8W~ot4]kmA}a,?@A%4%%pVFt{{OeOepiopD%#33!//))
 	
r5   c	                     |r_t        | j                  d      rI|j                  d   | j                  j                  dz   k\  r |d   }
|
| j                  j                  k  rd }t	        |   d||||||||d|	}|S )N original_max_position_embeddingsr   r   )r  r   r   r  r   ry   r   r%  r   )hasattrr)   rp   r-  r'   prepare_inputs_for_generation)r2   r  r   r   r  r   ry   r   r%  r   past_lengthmodel_inputsr3   s               r4   r/  z-Phi3ForCausalLM.prepare_inputs_for_generation  s    $ %GH"dkk&R&RUV&VV(+KdkkJJJ"&w< 

+)')%)

 

 r5   )	NNNNNNNNr   )NNNNNTN)r@   rA   rB   _tied_weights_keys_tp_plan_pp_planr(   r   r   rC   r   r   r   rD   r   rc   r   r   r   r?   r/  rE   rF   s   @r4   r  r    sQ   *,GH23H_-z:;H  .2.204(,26*.!%26-.8
##d*8
 t+8
 &&-	8

 8
 ((4/8
   4'8
 $;8
 ((4/8
 ell*8
 +,8
 
 8
  8
z % %r5   r  c                       e Zd Zy)Phi3ForSequenceClassificationNr@   rA   rB   r   r5   r4   r6  r6  /      r5   r6  c                       e Zd Zy)Phi3ForTokenClassificationNr7  r   r5   r4   r:  r:  3  r8  r5   r:  )r   r  r  r6  r:  )r   )r   )Acollections.abcr   typingr   rC   r   activationsr   cache_utilsr   r	   
generationr
   integrationsr   masking_utilsr   r   modeling_flash_attention_utilsr   modeling_layersr   r   r   modeling_outputsr   r   modeling_rope_utilsr   r   modeling_utilsr   r   processing_utilsr   utilsr   r   r   utils.genericr   r   utils.output_capturingr   configuration_phi3r    Moduler"   rH   r   r   rc   r   rg   r   r   r   r   r   r   r  r  r6  r:  __all__r   r5   r4   <module>rN     s  , %    ! . ) 7 R B 
 P K F & I I G 5 *)bii )$@<")) @<F(	UU\\ 	U# 	U%,, 	U& %II%<<% 
% <<	%
 LL4'% % % '(%2<B)BII B)J Y'J")) J (J(*1 *Z /  & M
# M
 M
` o)? o od	$DFY 		!>@S 	r5   