
    qi+U                     *   d dl mZ d dlmZ d dlZd dlmZ d dlmc 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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 Z1dejd                  de3dejd                  fdZ4	 d4dejZ                  d ejd                  d!ejd                  d"ejd                  d#ejd                  dz  d$e5d%e5d&e!e#   fd'Z6d5d(Z7 ee7       G d) d*ejZ                               Z8 G d+ d,e      Z9e$ G d- d.e             Z:e$ G d/ d0e:             Z;e$ G d1 d2e:e             Z<g d3Z=y)6    )Callable)OptionalN   )ACT2FN)CacheDynamicCache)GenerationMixin)use_kernelized_func)create_causal_mask)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   )
OlmoConfigc                   d     e Zd ZdZdeddf fdZdej                  dej                  fdZ xZ	S )OlmoLayerNormz/LayerNorm but with no learnable weight or bias.hidden_sizereturnNc                 2    t         |           |f| _        y N)super__init__normalized_shape)selfr   	__class__s     X/opt/pipecat/venv/lib/python3.12/site-packages/transformers/models/olmo/modeling_olmo.pyr#   zOlmoLayerNorm.__init__4   s    !,    hidden_statesc                     |j                   }t        j                  |j                  t        j
                        | j                  d d d      j                  |      S )Ndtypegh㈵>)eps)r,   F
layer_normtotorchfloat32r$   )r%   r)   
orig_dtypes      r'   forwardzOlmoLayerNorm.forward8   sO    "((
||M,,5==,A4CXCXZ^`djnorr
 	
r(   )
__name__
__module____qualname____doc__intr#   r1   Tensorr4   __classcell__r&   s   @r'   r   r   1   s4    9/C /D /
U\\ 
ell 
r(   r   c                   $     e Zd Z fdZd Z xZS )OlmoMLPc                    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#   configr   intermediate_sizennLinear	gate_projup_proj	down_projr   
hidden_actact_fnr%   rC   r&   s     r'   r#   zOlmoMLP.__init__@   s    !--!'!9!94#3#3T5K5KRWXyy!1!143I3IPUV4#9#94;K;KRWXV../r(   c                     | j                  | j                  | j                  |            | j                  |      z        }|S r!   )rI   rK   rG   rH   )r%   xrI   s      r'   r4   zOlmoMLP.forwardJ   s6    NN4;;t~~a/@#ADLLQRO#ST	r(   )r5   r6   r7   r#   r4   r;   r<   s   @r'   r>   r>   ?   s    0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 )OlmoRotaryEmbeddinginv_freqNrC   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defaultrQ   F)
persistentoriginal_inv_freq)r"   r#   max_position_embeddingsmax_seq_len_cachedoriginal_max_seq_lenrC   rope_parametersrS   compute_default_rope_parametersr   attention_scalingregister_bufferclone)r%   rC   devicerope_init_fnrQ   r&   s        r'   r#   zOlmoRotaryEmbedding.__init__R   s    "("@"@$*$B$B!44[A!%!E!E>>Y&.t~~>L+7V+L($(ZeD0(..2BuUr(   r_   ztorch.deviceseq_lenr   ztorch.Tensorc                    | j                   d   }t        | dd      xs | j                  | j                  z  }d}d|t	        j
                  d|dt        j                        j                  |t        j                        |z  z  z  }||fS )	a  
        Computes the inverse frequencies according to the original RoPE implementation
        Args:
            config ([`~transformers.PreTrainedConfig`]):
                The model configuration.
            device (`torch.device`):
                The device to use for initialization of the inverse frequencies.
            seq_len (`int`, *optional*):
                The current sequence length. Unused for this type of RoPE.
        Returns:
            Tuple of (`torch.Tensor`, `float`), containing the inverse frequencies for the RoPE embeddings and the
            post-processing scaling factor applied to the computed cos/sin (unused in this type of RoPE).
        
rope_thetahead_dimNg      ?r      r+   )r_   r,   )	rZ   getattrr   num_attention_headsr1   arangeint64r0   float)rC   r_   ra   basedimattention_factorrQ   s          r'   r[   z3OlmoRotaryEmbedding.compute_default_rope_parametersb   s    & %%l3fj$/c63E3EIcIc3c U\\!S!5;;?BB&X]XcXcBdgjjk
 )))r(   c                    | j                   d d d d f   j                         j                  |j                  d   dd      j	                  |j
                        }|d d d d d f   j                         }t        |j
                  j                  t              r/|j
                  j                  dk7  r|j
                  j                  nd}t        |d      5  |j                         |j                         z  j                  dd      }t        j                  ||fd	      }|j                         | j                  z  }|j                         | j                  z  }	d d d        ||	fS # 1 sw Y   	fS xY w)
Nr   r   mpscpuF)device_typeenabledre   rl   )rQ   rj   expandshaper0   r_   
isinstancetypestrr   	transposer1   catcosr\   sin)
r%   rN   position_idsinv_freq_expandedposition_ids_expandedrr   freqsembr|   r}   s
             r'   r4   zOlmoRotaryEmbedding.forward   s8    !MM$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
 Cx	5
 Cxs   BE''E3r!   )NNN)r5   r6   r7   r1   r:   __annotations__r   r#   staticmethodr   r9   tuplerj   r[   no_gradr   r4   r;   r<   s   @r'   rP   rP   O   s    llVz V  $(+/"*T!*(* t* 
~u$	%	* *: U]]_
  
r(   rP   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   re   rt   )rv   r1   r{   )rN   x1x2s      r'   rotate_halfr      sZ    	
3"!''"+"""	#B	
3q ""	#B99rc2YB''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)rv   ru   reshape)r)   r   batchnum_key_value_headsslenrd   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 )Nre   r   ro   )rl   r,   )ptrainingr   )r   num_key_value_groupsr1   matmulrz   rE   
functionalsoftmaxr2   r0   r,   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                 
   | j                   |j                   }}|j                  |      }|j                  |      }| |z  t        |       |z  z   }||z  t        |      |z  z   }|j                  |      |j                  |      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.
    )r,   	unsqueezer   r0   )	qkr|   r}   unsqueeze_dimq_typek_typeq_embedk_embeds	            r'   apply_rotary_pos_embr      s|    $ WWaggFF
--
&C
--
&C3w;q>C/0G3w;q>C/0G::fwzz&111r(   c                   
    e Zd ZdZdedef fdZ	 	 ddej                  de	ej                  ej                  f   dej                  dz  d	e
dz  d
ej                  dz  de	ej                  ej                  dz  f   fdZ xZS )OlmoAttentionz=Multi-headed attention from 'Attention Is All You Need' paperrC   	layer_idxc                 d   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                        | _        y )Nrd   g      TrA   )r"   r#   rC   r   rf   r   rg   rd   r   r   r   attention_dropout	is_causalrE   rF   attention_biasq_projk_projv_projo_projr%   rC   r   r&   s      r'   r#   zOlmoAttention.__init__   sM   "
F4F4F&JdJd4de$*$>$>&B\B\$\!}}d*!'!9!9ii : :T]] JQWQfQf
 ii : :T]] JQWQfQf
 ii : :T]] JQWQfQf
 ii&&68J8JQWQfQf
r(   Nr)   position_embeddingsr   past_key_valuescache_positionr   c                    |j                   d d }g |d| j                  }| j                  |      }	| j                  |      }
| j	                  |      }| j
                  j                  |	j                  | j
                  j                   | j
                  j                         |
j                  | j
                  j                   | j
                  j                         |j                  | j
                  j                   | j
                  j                         |	j                  |      j                  dd      }	|
j                  |      j                  dd      }
|j                  |      j                  dd      }|\  }}t        |	|
||      \  }	}
|'|||d}|j                  |
|| j                  |      \  }
}t        j                  | j
                  j                  t               } || |	|
||f| j"                  sdn| j$                  | j&                  d|\  }} |j(                  g |d j+                         }| j-                  |      }||fS )Nro   )minmaxr   re   )r}   r|   r           )r   r   )rv   rd   r   r   r   rC   clip_qkvclamp_viewrz   r   updater   r   get_interface_attn_implementationr   r   r   r   r   r   r   )r%   r)   r   r   r   r   r   input_shapehidden_shapequery_statesr   r   r|   r}   cache_kwargsattention_interfacer   r   s                     r'   r4   zOlmoAttention.forward   s8    $))#2.88b8$--8{{=1[[/
{{=1;;+T[[%9%9$9t{{?S?ST4;;#7#7"7T[[=Q=QRT[[%9%9$9t{{?S?ST#((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$2H2HLL	%
 	%
!\ *k));;;;FFHkk+.L((r(   )NN)r5   r6   r7   r8   r   r9   r#   r1   r:   r   r   
LongTensorr4   r;   r<   s   @r'   r   r      s    G
z 
c 
8 )-262)||2) #5<<#=>2) t+	2)
 2) ((4/2) 
u||U\\D00	12)r(   r   c                   "    e Zd Zdedef fdZ	 	 	 	 	 	 ddej                  dej                  dz  dej                  dz  de	dz  d	e
dz  d
ej                  dz  deej                  ej                  f   dz  dee   dej                  fdZ xZS )OlmoDecoderLayerrC   r   c                     t         |           |j                  | _        t        ||      | _        t        |      | _        t        |j                        | _        t        |j                        | _	        y )N)rC   r   )
r"   r#   r   r   	self_attnr>   mlpr   input_layernormpost_attention_layernormr   s      r'   r#   zOlmoDecoderLayer.__init__&  s[    !--&f	J6?,V-?-?@(5f6H6H(I%r(   Nr)   r   r~   r   	use_cacher   r   r   r   c                     |}	| j                  |      } | j                  d|||||||d|\  }}
|	|z   }|}	| j                  |      }| j                  |      }|	|z   }|S )N)r)   r   r~   r   r   r   r    )r   r   r   r   )r%   r)   r   r~   r   r   r   r   r   residual_s              r'   r4   zOlmoDecoderLayer.forward/  s     !,,];)4>> 	
')%+) 3	
 	
q !=0 !55mD/ =0r(   )NNNFNN)r5   r6   r7   r   r9   r#   r1   r:   r   r   boolr   r   r   r4   r;   r<   s   @r'   r   r   %  s    Jz Jc J /304(,!&26HL|| t+ &&-	
  $; ((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)OlmoPreTrainedModelrC   modelTr   r   )r)   
attentionsN)r5   r6   r7   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   Q  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 )	OlmoModelrC   c           	         t         |   |       |j                  | _        |j                  | _        t        j                  |j                  |j                  | j                        | _        t        j                  t        |j                        D cg c]  }t        ||       c}      | _        t        |j                        | _        t!        |      | _        d| _        | j'                          y c c}w )NrC   F)r"   r#   pad_token_idpadding_idx
vocab_sizerE   	Embeddingr   embed_tokens
ModuleListrangenum_hidden_layersr   layersr   normrP   
rotary_embgradient_checkpointing	post_initr   s      r'   r#   zOlmoModel.__init__f  s     !.. ++LL):):F<N<NPTP`P`ammBGH`H`BabYfi0b
 "&"4"45	-V<&+# 	 cs   C5N	input_idsr   r~   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   )r_   )rC   r   r   r   r   r~   )r~   )r   r   r~   r   r   r   )last_hidden_stater   )
ValueErrorr   r   rC   get_seq_lengthr1   rh   rv   r_   r   r   r   r   r   r   r   )r%   r   r   r~   r   r   r   r   r   past_seen_tokenscausal_maskr)   r   decoder_layers                 r'   r4   zOlmoModel.forwardv  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)r5   r6   r7   r   r#   r   r   r   r1   r   r:   r   FloatTensorr   r   r   r   r4   r;   r<   s   @r'   r   r   d  s    z     .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 )OlmoForCausalLMzlm_head.weightzmodel.embed_tokens.weightlm_headcolwise_gather_outputr)   logitsc                     t         |   |       t        |      | _        |j                  | _        t        j                  |j                  |j                  d      | _        | j                          y r@   )
r"   r#   r   r   r   rE   rF   r   r  r   rL   s     r'   r#   zOlmoForCausalLM.__init__  sU     v&
 ++yy!3!3V5F5FUS 	r(   Nr   r   r~   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, OlmoForCausalLM

        >>> model = OlmoForCausalLM.from_pretrained("meta-olmo/Olmo-2-7b-hf")
        >>> tokenizer = AutoTokenizer.from_pretrained("meta-olmo/Olmo-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   r~   r   r   r   r   N)r
  r  r   )lossr
  r   r)   r   r   )r   r   rw   r9   slicer  loss_functionrC   r   r   r   r)   r   )r%   r   r   r~   r   r   r  r   r   r  r   outputsr)   slice_indicesr
  r  s                   r'   r4   zOlmoForCausalLM.forward  s    @ ,64:: 	,
)%+')	,
 	,
  118B>SV8W~ot4]kmA}a,?@A%4%%pVFt{{OeOepiopD%#33!//))
 	
r(   )	NNNNNNNNr   )r5   r6   r7   _tied_weights_keys_tp_plan_pp_planr#   r   r   r1   r   r:   r   r  r   r9   r   r   r   r4   r;   r<   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
 $;8
 ((4/8
 ell*8
 +,8
 
 8
  8
r(   r  )r  r   r   )r   )r   )>collections.abcr   typingr   r1   torch.nnrE   torch.nn.functionalr   r.   activationsr   cache_utilsr   r   
generationr	   integrationsr
   masking_utilsr   modeling_layersr   modeling_outputsr   r   modeling_rope_utilsr   r   modeling_utilsr   r   processing_utilsr   utilsr   r   r   utils.genericr   r   utils.output_capturingr   configuration_olmor   Moduler   r>   rP   r   r:   r9   r   rj   r   r   r   r   r   r   r  __all__r   r(   r'   <module>r+     s  4 %      ! . ) / / 9 O K F & I I G 5 *
BII 
bii  =")) =@(	UU\\ 	U# 	U%,, 	U& %II%<<% 
% <<	%
 LL4'% % % '(%224 )*L)BII L) +L)^)1 )X /  $ M
# M
 M
` H
)? H
 H
V Br(   