
    qiU                     X   d dl Z d dlmZ d dlmZ d dlZd dlmZ ddlm	Z	 ddl
mZmZ ddl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% ddl&m'Z'm(Z( ddl)m*Z* ddl+m,Z,  G d dejZ                        Z. G d dejZ                        Z/ G d dejZ                        Z0dejb                  de2dejb                  fdZ3	 d8dejZ                  dejb                  d ejb                  d!ejb                  d"ejb                  dz  d#e4d$e4d%e!e#   fd&Z5d' Z6d9d(Z7 ee7       G d) d*ejZ                               Z8 G d+ d,e      Z9e$ G d- d.e             Z:e$ G d/ d0e:             Z;e$ G d1 d2e:e             Z< G d3 d4ee:      Z= G d5 d6ee:      Z>g d7Z?y):    N)Callable)Optional   )ACT2FN)CacheDynamicCache)GenerationMixin)use_kernelized_func)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)maybe_autocastmerge_with_config_defaults)capture_outputs   )HeliumConfigc                   ,     e Zd Zd fd	Zd Zd Z xZS )HeliumRMSNormc                     t         |           t        j                  t	        j
                  |            | _        || _        y N)super__init__nn	Parametertorchonesweightvariance_epsilon)selfhidden_sizeeps	__class__s      \/opt/pipecat/venv/lib/python3.12/site-packages/transformers/models/helium/modeling_helium.pyr#   zHeliumRMSNorm.__init__1   s/    ll5::k#:; #    c                 \   |j                   }|j                  t        j                        }|j	                  d      j                  dd      }|t        j                  || j                  z         z  }| j                  j                  t        j                        |z  j                  |      S )N   T)keepdim)	dtypetor&   float32powmeanrsqrtr)   r(   )r*   hidden_statesinput_dtypevariances       r.   forwardzHeliumRMSNorm.forward6   s    #))%((7 $$Q',,R,>%Ht?T?T4T(UUu}}-=AA+NNr/   c                 ^    t        | j                  j                         d| j                   S )Nz, eps=)tupler(   shaper)   )r*   s    r.   
extra_reprzHeliumRMSNorm.extra_repr=   s*    ))*+6$2G2G1HIIr/   )gư>)__name__
__module____qualname__r#   r=   rA   __classcell__r-   s   @r.   r   r   0   s    $
OJr/   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 )HeliumRotaryEmbedding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defaultrI   F)
persistentoriginal_inv_freq)r"   r#   max_position_embeddingsmax_seq_len_cachedoriginal_max_seq_lenrJ   rope_parametersrL   compute_default_rope_parametersr   attention_scalingregister_bufferclone)r*   rJ   devicerope_init_fnrI   r-   s        r.   r#   zHeliumRotaryEmbedding.__init__D   s    "("@"@$*$B$B!44[A!%!E!E>>Y&.t~~>L+7V+L($(ZeD0(..2BuUr/   rX   ztorch.deviceseq_lenreturnz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_thetahead_dimNg      ?r   r1   r4   )rX   r4   )	rS   getattrr+   num_attention_headsr&   arangeint64r5   float)rJ   rX   rZ   basedimattention_factorrI   s          r.   rT   z5HeliumRotaryEmbedding.compute_default_rope_parametersT   s    & %%l3fj$/c63E3EIcIc3c U\\!S!5;;?BB&X]XcXcBdgjjk
 )))r/   c                 N   | j                   d d d d f   j                         j                  |j                  d   dd      j	                  |j
                        }|d d d d d f   j                         }t        |j
                  j                  t              r/|j
                  j                  dk7  r|j
                  j                  nd}t        |d      5  |j                         |j                         z  j                  dd      }t        j                  ||fd	      }|j                         | j                  z  }|j                         | j                  z  }	d d d        j	                  |j                   
      	j	                  |j                   
      fS # 1 sw Y   AxY w)Nr   r2   r   mpscpuF)device_typeenabledr1   rf   r_   )rI   rd   expandr@   r5   rX   
isinstancetypestrr   	transposer&   catcosrU   sinr4   )
r*   xposition_idsinv_freq_expandedposition_ids_expandedrk   freqsembrt   ru   s
             r.   r=   zHeliumRotaryEmbedding.forwardr   sR    !MM$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$cff177f&;;;	5 	5s   BFF$r!   )NNN)rB   rC   rD   r&   Tensor__annotations__r   r#   staticmethodr   intr?   rd   rT   no_gradr   r=   rE   rF   s   @r.   rH   rH   A   s    llV| V  &*+/"*t#*(* t* 
~u$	%	* *: U]]_<  <r/   rH   c                   $     e Zd Z fdZd Z xZS )	HeliumMLPc                    t         |           || _        |j                  | _        |j                  | _        t        j                  | j                  | j                  |j                        | _        t        j                  | j                  | j                  |j                        | _	        t        j                  | j                  | j                  |j                        | _
        t        |j                     | _        y )Nbias)r"   r#   rJ   r+   intermediate_sizer$   Linearmlp_bias	gate_projup_proj	down_projr   
hidden_actact_fnr*   rJ   r-   s     r.   r#   zHeliumMLP.__init__   s    !--!'!9!94#3#3T5K5KRXRaRabyy!1!143I3IPVP_P_`4#9#94;K;KRXRaRabV../r/   c                     | j                  | j                  | j                  |            | j                  |      z        }|S r!   )r   r   r   r   )r*   rv   r   s      r.   r=   zHeliumMLP.forward   s6    NN4;;t~~a/@#ADLLQRO#ST	r/   )rB   rC   rD   r#   r=   rE   rF   s   @r.   r   r      s    0r/   r   r:   n_repr[   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)r@   rn   reshape)r:   r   batchnum_key_value_headsslenr^   s         r.   	repeat_kvr      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        || 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 )Nr1   r   r2   )rf   r4   )ptrainingr   )r   num_key_value_groupsr&   matmulrr   r$   
functionalsoftmaxr6   r5   r4   r   r   
contiguous)r   r   r   r   r   r   r   r   
key_statesvalue_statesattn_weightsattn_outputs               r.   eager_attention_forwardr      s     3 ; ;<JUF$?$?@L<<z';';Aq'ABWLL!#n4==((2U]](SVVW\WbWbcL==((6??([L,,|\:K''1-88:K$$r/   c                 |    | ddddf   }| ddddf   }t        j                  | |fd      j                  d      S )	z*Rotates half the hidden dims of the input..r   Nr1   r   r2   rm   )r&   stackflatten)rv   x1x2s      r.   rotate_halfr      sJ    	
319B	
319B;;Ryb)11"55r/   c                 F   |j                  |      }|j                  |      }|dd|j                  d   dz  f   j                  dd      }|dd|j                  d   dz  f   j                  dd      }| |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.
    .Nr2   r1   rm   )	unsqueezer@   repeat_interleaver   )qkrt   ru   unsqueeze_dimq_embedk_embeds          r.   apply_rotary_pos_embr      s    $ --
&C
--
&C c'SYYr]a'''
(
:
:1"
:
EC
c'SYYr]a'''
(
:
:1"
:
EC3w;q>C/0G3w;q>C/0GGr/   c                        e Zd ZdZddededz  f fdZ	 	 	 	 ddej                  de	ej                  ej                  f   dz  dej                  dz  d	e
dz  d
ej                  dz  dee   de	ej                  ej                  f   fdZ xZS )HeliumAttentionz=Multi-headed attention from 'Attention Is All You Need' paperNrJ   	layer_idxc                 \   t         |           || _        || _        t	        |d|j
                  |j                  z        | _        |j                  |j                  z  | _	        dt        j                  | j                        z  | _        |j                  | _        d| _        t        j                   |j
                  |j                  | j                  z  |j"                        | _        t        j                   |j
                  |j                  | j                  z  |j"                        | _        t        j                   |j
                  |j                  | j                  z  |j"                        | _        t        j                   |j
                  |j
                  d      | _        y )Nr^   r   Tr   F)r"   r#   rJ   r   r`   r+   ra   r^   r   r   mathsqrtr   attention_dropout	is_causalr$   r   attention_biasq_projk_projv_projo_projr*   rJ   r   r-   s      r.   r#   zHeliumAttention.__init__   sC   "
F4F4F&JdJd4de$*$>$>&B\B\$\!499T]]33!'!9!9ii : :T]] JQWQfQf
 ii : :T]] JQWQfQf
 ii : :T]] JQWQfQf
 ii 2 2F4F4FUSr/   r:   position_embeddingsr   past_key_valuescache_positionr   r[   c                    |j                   d d }g |d| j                  }| j                  |      j                  |      j	                  dd      }	| j                  |      j                  |      j	                  dd      }
| j                  |      j                  |      j	                  dd      }|\  }}t        |	|
||      \  }	}
|'|||d}|j                  |
|| j                  |      \  }
}t        j                  | j                  j                  t              } || |	|
||f| j                  sdn| j                   | j"                  d|\  }} |j$                  g |d j'                         }| j)                  |      }||fS )Nr2   r   r1   )ru   rt   r           )r   r   )r@   r^   r   viewrr   r   r   r   updater   r   get_interfacerJ   _attn_implementationr   r   r   r   r   r   r   )r*   r:   r   r   r   r   r   input_shapehidden_shapequery_statesr   r   rt   ru   cache_kwargsattention_interfacer   r   s                     r.   r=   zHeliumAttention.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'6'='=j,X\XfXfht'u$J(?(M(MKK,,.E)
 %8	%
  $}}C$2H2HLL	%
 	%
!\ *k));;;;FFHkk+.L((r/   r!   )NNNN)rB   rC   rD   __doc__r   r   r#   r&   r|   r?   r   
LongTensorr   r   r=   rE   rF   s   @r.   r   r      s    GT| Td
 T0 IM.2(,26))||)) #5<<#=>E)) t+	))
 )) ((4/)) +,)) 
u||U\\)	*))r/   r   c                   *    e Zd Zddededz  f fdZ	 	 	 	 	 	 ddej                  dej                  dz  dej                  dz  de	dz  d	e
dz  d
ej                  dz  deej                  ej                  f   dz  dee   dej                  fdZ xZS )HeliumDecoderLayerNrJ   r   c                     t         |           |j                  | _        t        ||      | _        t        |      | _        t        |j                  |j                        | _	        t        |j                  |j                        | _
        y )N)rJ   r   r,   )r"   r#   r+   r   	self_attnr   mlpr   rms_norm_epsinput_layernormpost_attention_layernormr   s      r.   r#   zHeliumDecoderLayer.__init__#  sl    !--()LV$,V-?-?VEXEXY(5f6H6HfNaNa(b%r/   r:   r   rw   r   	use_cacher   r   r   r[   c                     |}	| j                  |      } | j                  d|||||||d|\  }}
|	|z   }|}	| j                  |      }| j                  |      }|	|z   }|S )N)r:   r   rw   r   r   r   r    )r   r   r   r   )r*   r:   r   rw   r   r   r   r   r   residual_s              r.   r=   zHeliumDecoderLayer.forward-  s     !,,];)4>> 	
')%+) 3	
 	
q !=0 !55mD/ =0r/   r!   )NNNFNN)rB   rC   rD   r   r   r#   r&   r|   r   r   boolr?   r   r   r=   rE   rF   s   @r.   r   r   "  s    c| cd
 c /304(,!&26HL|| t+ &&-	
  $; ((4/ #5<<#=>E +, 
r/   r   c                   J    e Zd ZU eed<   dZdZdgZdgZdZ	dZ
dZdZdZeedZy)HeliumPreTrainedModelrJ   modelTr   r   )r:   
attentionsN)rB   rC   rD   r   r}   base_model_prefixsupports_gradient_checkpointing_no_split_modules_skip_keys_device_placement_supports_flash_attn_supports_sdpa_supports_flex_attn_can_compile_fullgraph_supports_attention_backendr   r   _can_record_outputsr   r/   r.   r   r   O  sQ    &*#-.#4"5N!"&+%r/   r   c                       e Zd Zdef fdZeee	 	 	 	 	 	 	 ddej                  dz  dej                  dz  dej                  dz  dedz  dej                  dz  d	ej                  dz  d
edz  dee   defd                     Z xZS )HeliumModelrJ   c           	         t         |   |       |j                  | _        |j                  | _        t        j                  |j                  |j                  | j                        | _        t        j                  t        |j                        D cg c]  }t        ||       c}      | _        t        |j                  |j                        | _        t#        |      | _        d| _        | j)                          y c c}w )Nr   rJ   F)r"   r#   pad_token_idpadding_idx
vocab_sizer$   	Embeddingr+   embed_tokens
ModuleListrangenum_hidden_layersr   layersr   r   normrH   
rotary_embgradient_checkpointing	post_initr   s      r.   r#   zHeliumModel.__init__d  s     !.. ++LL):):F<N<NPTP`P`ammDI&JbJbDcdy	2d
 "&"4"4&:M:MN	/v>&+# 	 es   DN	input_idsr   rw   r   inputs_embedsr   r   r   r[   c                 D   |d u |d uz  rt        d      || j                  |      }|r|t        | j                        }|E||j	                         nd}	t        j                  |j                  d   |j                        |	z   }||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   )rX   )rJ   r	  r   r   r   rw   )rw   )r   r   rw   r   r   r   )last_hidden_stater   )
ValueErrorr   r   rJ   get_seq_lengthr&   rb   r@   rX   r   r   r  r  r  r  r   )r*   r  r   rw   r   r	  r   r   r   past_seen_tokenscausal_maskr:   r   decoder_layers                 r.   r=   zHeliumModel.forwardt  s]    -t";<YZZ *.*;*;I*FM0*$++>O!CRC^==?de]003M<P<PQTdd  )33A6L(;;'))+%
 &"oom,oW![[)H4;;+H+HI 
	M)	*$7) /#-	 	M
	 		-0&++
 	
r/   )NNNNNNN)rB   rC   rD   r   r#   r   r   r   r&   r   r|   r   FloatTensorr   r   r   r   r=   rE   rF   s   @r.   r   r   b  s    |     .2.204(,2626!%9
##d*9
 t+9
 &&-	9

 9
 ((4/9
 ((4/9
 $;9
 +,9
 
!9
    9
r/   r   c                   b    e Zd ZddiZddiZddgdgfiZ fdZee	 	 	 	 	 	 	 	 	 dd	e	j                  dz  d
e	j                  dz  de	j                  dz  dedz  de	j                  dz  de	j                  dz  dedz  de	j                  dz  dee	j                  z  dee   defd              Z xZS )HeliumForCausalLMzlm_head.weightzmodel.embed_tokens.weightlm_headcolwise_gather_outputr:   logitsc                     t         |   |       t        |      | _        |j                  | _        t        j                  |j                  |j                  d      | _        | j                          y )NFr   )
r"   r#   r   r   r   r$   r   r+   r  r  r   s     r.   r#   zHeliumForCausalLM.__init__  sU      (
 ++yy!3!3V5F5FUS 	r/   Nr  r   rw   r   r	  labelsr   r   logits_to_keepr   r[   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, HeliumForCausalLM

        >>> model = HeliumForCausalLM.from_pretrained("google/helium-7b")
        >>> tokenizer = AutoTokenizer.from_pretrained("google/helium-7b")

        >>> prompt = "What is your favorite condiment?"
        >>> 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]
        "What is your favorite condiment?"
        ```)r  r   rw   r   r	  r   r   N)r  r  r   )lossr  r   r:   r   r   )r   r  ro   r   slicer  loss_functionrJ   r   r   r   r:   r   )r*   r  r   rw   r   r	  r  r   r   r  r   outputsr:   slice_indicesr  r  s                   r.   r=   zHeliumForCausalLM.forward  s    @ ,64:: 	,
)%+')	,
 	,
  118B>SV8W~ot4]kmA}a,?@A%4%%pVFt{{OeOepiopD%#33!//))
 	
r/   )	NNNNNNNNr   )rB   rC   rD   _tied_weights_keys_tp_plan_pp_planr#   r   r   r&   r   r|   r   r  r   r   r   r   r   r=   rE   rF   s   @r.   r  r    s/   *,GH23H_-z:;H  .2.204(,26*.!%26-.8
##d*8
 t+8
 &&-	8

 8
 ((4/8
   4'8
 $;8
 ((4/8
 ell*8
 +,8
 
 8
  8
r/   r  c                       e Zd Zy)HeliumForSequenceClassificationNrB   rC   rD   r   r/   r.   r$  r$        r/   r$  c                       e Zd Zy)HeliumForTokenClassificationNr%  r   r/   r.   r(  r(    r&  r/   r(  )r   r   r  r$  r(  )r   )r   )@r   collections.abcr   typingr   r&   torch.nnr$   activationsr   cache_utilsr   r   
generationr	   integrationsr
   masking_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_heliumr   Moduler   rH   r   r|   r   r   rd   r   r   r   r   r   r   r   r  r$  r(  __all__r   r/   r.   <module>r<     s  *  $    ! . ) / / 
 P K F & I I G 5 .JBII J"><BII ><B		  	UU\\ 	U# 	U%,, 	U& %II%<<% 
% <<	%
 LL4'% % % '(%26> )*A)bii A) +A)H*3 *Z O  $ M
' M
 M
` H
- H
 H
V	&FH] 		#@BW 	r/   