
    inX                        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mZmZ dd
lmZmZ ddlmZ ddlmZmZmZmZ ddlmZmZ ddlmZm Z  ddl!m"Z"m#Z# ddl$m%Z% ddl&m'Z'm(Z(m)Z) ddl*m+Z+m,Z, ddl-m.Z. ddl/m0Z0  G d dejb                        Z2d Z3 ed      d>d       Z4dejj                  de6dejj                  fdZ7	 d?dejb                  d ejj                  d!ejj                  d"ejj                  d#ejj                  dz  d$e8d%e8d&e%e'   fd'Z9 ee4       G d( d)ejb                               Z: ed*       G d+ d,ejb                               Z; G d- d.e      Z<e( G d/ d0e#             Z= G d1 d2ejb                        Z>e( G d3 d4e=             Z?e( G d5 d6e=e             Z@ G d7 d8ee=      ZA G d9 d:ee=      ZB G d; d<ee=      ZCg d=ZDy)@    )Callable)OptionalN)nn   )ACT2FN)CacheDynamicCache)GenerationMixin)use_kernel_forward_from_hubuse_kernel_func_from_hubuse_kernelized_func)create_causal_mask!create_sliding_window_causal_mask)FlashAttentionKwargs)GenericForQuestionAnswering 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   )MinistralConfigc                   $     e Zd Z fdZd Z xZS )MinistralMLPc                    t         |           || _        |j                  | _        |j                  | _        t        j                  | j                  | j                  d      | _        t        j                  | j                  | j                  d      | _        t        j                  | j                  | j                  d      | _	        t        |j                     | _        y NFbias)super__init__confighidden_sizeintermediate_sizer   Linear	gate_projup_proj	down_projr   
hidden_actact_fnselfr,   	__class__s     l/mnt/e/genesis-system/.venv/lib/python3.12/site-packages/transformers/models/ministral/modeling_ministral.pyr+   zMinistralMLP.__init__$   s    !--!'!9!94#3#3T5K5KRWXyy!1!143I3IPUV4#9#94;K;KRWXV../    c                     | j                  | j                  | j                  |            | j                  |      z        }|S N)r2   r4   r0   r1   )r6   xr2   s      r8   forwardzMinistralMLP.forward.   s6    NN4;;t~~a/@#ADLLQRO#ST	r9   )__name__
__module____qualname__r+   r=   __classcell__r7   s   @r8   r%   r%   #   s    0r9   r%   c                     | dd| j                   d   dz  f   }| d| j                   d   dz  df   }t        j                  | |fd      S )z*Rotates half the hidden dims of the input..N   dim)shapetorchcat)r<   x1x2s      r8   rotate_halfrM   3   sZ    	
3"!''"+"""	#B	
3q ""	#B99rc2YB''r9   rotary_pos_embc                     |j                  |      }|j                  |      }| |z  t        |       |z  z   }||z  t        |      |z  z   }||fS )a  Applies Rotary Position Embedding to the query and key tensors.

    Args:
        q (`torch.Tensor`): The query tensor.
        k (`torch.Tensor`): The key tensor.
        cos (`torch.Tensor`): The cosine part of the rotary embedding.
        sin (`torch.Tensor`): The sine part of the rotary embedding.
        unsqueeze_dim (`int`, *optional*, defaults to 1):
            The 'unsqueeze_dim' argument specifies the dimension along which to unsqueeze cos[position_ids] and
            sin[position_ids] so that they can be properly broadcasted to the dimensions of q and k. For example, note
            that cos[position_ids] and sin[position_ids] have the shape [batch_size, seq_len, head_dim]. Then, if q and
            k have the shape [batch_size, heads, seq_len, head_dim], then setting unsqueeze_dim=1 makes
            cos[position_ids] and sin[position_ids] broadcastable to the shapes of q and k. Similarly, if q and k have
            the shape [batch_size, seq_len, heads, head_dim], then set unsqueeze_dim=2.
    Returns:
        `tuple(torch.Tensor)` comprising of the query and key tensors rotated using the Rotary Position Embedding.
    )	unsqueezerM   )qkcossinunsqueeze_dimq_embedk_embeds          r8   apply_rotary_pos_embrX   :   sY    & --
&C
--
&C3w;q>C/0G3w;q>C/0GGr9   hidden_states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)rH   expandreshape)rY   rZ   batchnum_key_value_headsslenhead_dims         r8   	repeat_kvrc   T   so    
 2?1D1D.Ehz!!Qa"23::5BUW\^bdlmM  (;e(CT8TTr9   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   rD   )rG   dtype)ptrainingr"   )rc   num_key_value_groupsrI   matmul	transposer   
functionalsoftmaxfloat32torm   rj   ro   
contiguous)rd   re   rf   rg   rh   ri   rj   rk   
key_statesvalue_statesattn_weightsattn_outputs               r8   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$$r9   c                       e Zd ZdZ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   deej                  ej                  dz  f   fdZ xZS )MinistralAttentionz=Multi-headed attention from 'Attention Is All You Need' paper	layer_idxc                    t         |           t        |d      r|j                  |   nd | _        || _        || _        t        |d|j                  |j                  z        | _
        |j                  |j                  z  | _        | j                  dz  | _        |j                  | _        d| _        t!        j"                  |j                  |j                  | j                  z  d      | _        t!        j"                  |j                  |j                  | j                  z  d      | _        t!        j"                  |j                  |j                  | j                  z  d      | _        t!        j"                  |j                  | j                  z  |j                  d      | _        | j                  dk(  r|j,                  | _        y d | _        y )Nlayer_typesrb   g      TFr(   sliding_attention)r*   r+   hasattrr   
layer_typer,   r   getattrr-   num_attention_headsrb   r`   rp   ri   attention_dropout	is_causalr   r/   q_projk_projv_projo_projsliding_windowr6   r,   r   r7   s      r8   r+   zMinistralAttention.__init__}   sl   ;B6=;Y&,,Y7_c"
F4F4F&JdJd4de$*$>$>&B\B\$\!}}d*!'!9!9ii 2 2F4N4NQUQ^Q^4^ejkii 2 2F4N4NQUQ^Q^4^ejkii 2 2F4N4NQUQ^Q^4^ejkii : :T]] JFL^L^ejk7;J]7]f33cgr9   NrY   position_embeddingsrh   past_key_valuescache_positionrk   r[   c                 .   |j                   d d }g |d| j                  }| j                  |      j                  |      j	                  dd      }	| j                  |      j                  |      j	                  dd      }
| j                  |      j                  |      j	                  dd      }|\  }}t        |	|
||      \  }	}
|'|||d}|j                  |
|| j                  |      \  }
}t        j                  | j                  j                  t              } || |	|
||f| j                  sdn| j                   | j"                  | j$                  d|\  }} |j&                  g |d j)                         }| j+                  |      }||fS )NrD   r"   rE   )rT   rS   r           )rj   ri   r   )rH   rb   r   viewrr   r   r   rX   updater   r   get_interfacer,   _attn_implementationr|   ro   r   ri   r   r^   rw   r   )r6   rY   r   rh   r   r   rk   input_shapehidden_shapequery_statesrx   ry   rS   rT   cache_kwargsattention_interfacer{   rz   s                     r8   r=   zMinistralAttention.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$2H2HLL..
%
 
%
!\ *k));;;;FFHkk+.L((r9   )NN)r>   r?   r@   __doc__intr+   rI   Tensortupler   
LongTensorr   r   r=   rA   rB   s   @r8   r~   r~   y   s    Gh# h, )-26*)||*) #5<<#=>*) t+	*)
 *) ((4/*) -.*) 
u||U\\D00	1*)r9   r~   RMSNormc                   h     e Zd Zddeddf fdZdej                  dej                  fdZd Z xZ	S )	MinistralRMSNormepsr[   Nc                     t         |           t        j                  t	        j
                  |            | _        || _        y)z?
        MinistralRMSNorm is equivalent to T5LayerNorm
        N)r*   r+   r   	ParameterrI   onesweightvariance_epsilon)r6   r-   r   r7   s      r8   r+   zMinistralRMSNorm.__init__   s1     	ll5::k#:; #r9   rY   c                 "   |j                   }|j                  t        j                        }|j	                  d      j                  dd      }|t        j                  || j                  z         z  }| j                  |j                  |      z  S )NrE   rD   T)keepdim)	rm   rv   rI   ru   powmeanrsqrtr   r   )r6   rY   input_dtypevariances       r8   r=   zMinistralRMSNorm.forward   sy    #))%((7 $$Q',,R,>%Ht?T?T4T(UU{{]--k:::r9   c                 ^    t        | j                  j                         d| j                   S )Nz, eps=)r   r   rH   r   )r6   s    r8   
extra_reprzMinistralRMSNorm.extra_repr   s*    ))*+6$2G2G1HIIr9   )gư>)
r>   r?   r@   floatr+   rI   r   r=   r   rA   rB   s   @r8   r   r      s7    $ $$ $;U\\ ;ell ;Jr9   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 )MinistralDecoderLayerr,   r   c                 H   t         |           |j                  | _        t        ||      | _        t        |      | _        t        |j                  |j                        | _	        t        |j                  |j                        | _
        |j                  |   | _        y )N)r,   r   r   )r*   r+   r-   r~   	self_attnr%   mlpr   rms_norm_epsinput_layernormpost_attention_layernormr   attention_typer   s      r8   r+   zMinistralDecoderLayer.__init__   s    !--+6YO'/0B0BH[H[\(89K9KQWQdQd(e%$00;r9   NrY   rh   position_idsr   	use_cacher   r   rk   r[   c                     |}	| j                  |      } | j                  d|||||||d|\  }}
|	|z   }|}	| j                  |      }| j                  |      }|	|z   }|S )N)rY   rh   r   r   r   r   r    )r   r   r   r   )r6   rY   rh   r   r   r   r   r   rk   residual_s              r8   r=   zMinistralDecoderLayer.forward   s     !,,];)4>> 	
')%+) 3	
 	
q !=0 !55mD/ =0r9   )NNNFNN)r>   r?   r@   r#   r   r+   rI   r   r   r   boolr   r   r   r=   rA   rB   s   @r8   r   r      s    	< 	<3 	< /304(,!&26HL|| t+ &&-	
  $; ((4/ #5<<#=>E +, 
r9   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)MinistralPreTrainedModelr,   modelTr   r   )rY   
attentionsN)r>   r?   r@   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   r9   r8   r   r      sQ    &*#01#4"5N!"&.(r9   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 )MinistralRotaryEmbedding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defaultr   F)
persistentoriginal_inv_freq)r*   r+   max_position_embeddingsmax_seq_len_cachedoriginal_max_seq_lenr,   rope_parametersr   compute_default_rope_parametersr   attention_scalingregister_bufferclone)r6   r,   devicerope_init_fnr   r7   s        r8   r+   z!MinistralRotaryEmbedding.__init__  s    "("@"@$*$B$B!44[A!%!E!E>>Y&.t~~>L+7V+L($(ZeD0(..2BuUr9   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_thetarb   Ng      ?r   rE   rm   )r   rm   )	r   r   r-   r   rI   arangeint64rv   r   )r,   r   r   baserG   attention_factorr   s          r8   r   z8MinistralRotaryEmbedding.compute_default_rope_parameters$  s    & %%l3fj$/c63E3EIcIc3c U\\!S!5;;?BB&X]XcXcBdgjjk
 )))r9   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   rD   r"   mpscpuF)device_typeenabledrE   rF   r   )r   r   r]   rH   rv   r   
isinstancetypestrr   rr   rI   rJ   rS   r   rT   rm   )
r6   r<   r   inv_freq_expandedposition_ids_expandedr   freqsembrS   rT   s
             r8   r=   z MinistralRotaryEmbedding.forwardB  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)r>   r?   r@   rI   r   r   r#   r+   staticmethodr   r   r   r   r   no_gradr   r=   rA   rB   s   @r8   r   r     s    llV V  )-+/"*$&*(* t* 
~u$	%	* *: U]]_<  <r9   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 )MinistralModelr,   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   normr   
rotary_embgradient_checkpointing	post_initr   s      r8   r+   zMinistralModel.__init__T  s     !.. ++LL):):F<N<NPTP`P`ammGLVMeMeGfg)"695g
 %V%7%7V=P=PQ	2&A&+# 	 hs   DN	input_idsrh   r   r   inputs_embedsr   r   rk   r[   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      }t        |x}
t              s*| j                  |||||d}t        d
i |t        d
i |d}
|}| j                  ||      }| j                  d | j                  j                    D ]  } ||f|
|j"                     |||||d|}! | j%                  |      }t'        ||r|	      S d 	      S )Nz:You must specify exactly one of input_ids or inputs_embedsr   r   r"   )r   )r,   r  rh   r   r   r   )full_attentionr   )rh   r   r   r   r   r   )last_hidden_stater   r   )
ValueErrorr  r	   r,   get_seq_lengthrI   r   rH   r   rP   r   dictr   r   r  r	  r  r   r
  r   )r6   r  rh   r   r   r  r   r   rk   past_seen_tokenscausal_mask_mappingmask_kwargsrY   r   decoder_layers                  r8   r=   zMinistralModel.forwardd  s    -t";<YZZ  --i8M0*$++>O!CRC^==?de"\\ "2]5H5H5K"KTaThThN )33A6L ?-F ++!."0"0#2 ,K #5"C{"C%F%U%U#
 &"oom\J![[)H4;;+H+HI 
	M)	2=3O3OP) /#-$7	 	M
	 		-0&+/8O
 	
>B
 	
r9   )NNNNNNN)r>   r?   r@   r#   r+   r    r!   r   rI   r   r   r   FloatTensorr   r   r   r   r=   rA   rB   s   @r8   r   r   R  s         .2.204(,26!%26A
##d*A
 t+A
 &&-	A

 A
 ((4/A
 $;A
 ((4/A
 +,A
 
!A
    A
r9   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 )MinistralForCausalLMzlm_head.weightzmodel.embed_tokens.weightlm_headcolwise_gather_outputrY   logitsc                     t         |   |       t        |      | _        |j                  | _        t        j                  |j                  |j                  d      | _        | j                          y r'   )
r*   r+   r   r   r  r   r/   r-   r  r  r5   s     r8   r+   zMinistralForCausalLM.__init__  sU     #F+
 ++yy!3!3V5F5FUS 	r9   Nr  rh   r   r   r  labelsr   r   logits_to_keeprk   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, MinistralForCausalLM

        >>> model = MinistralForCausalLM.from_pretrained("meta-ministral/Ministral-2-7b-hf")
        >>> tokenizer = AutoTokenizer.from_pretrained("meta-ministral/Ministral-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  rh   r   r   r  r   r   N)r  r!  r  )lossr  r   rY   r   r   )r   r  r   r   slicer  loss_functionr,   r  r   r   rY   r   )r6   r  rh   r   r   r  r!  r   r   r"  rk   outputsrY   slice_indicesr  r$  s                   r8   r=   zMinistralForCausalLM.forward  s    @ ,64:: 	,
)%+')	,
 	,
  118B>SV8W~ot4]kmA}a,?@A%4%%pVFt{{OeOepiopD%#33!//))
 	
r9   )	NNNNNNNNr   )r>   r?   r@   _tied_weights_keys_tp_plan_pp_planr+   r   r   rI   r   r   r   r  r   r   r   r   r   r=   rA   rB   s   @r8   r  r    s/   *,GH23H_-z:;H  .2.204(,26*.!%26-.8
##d*8
 t+8
 &&-	8

 8
 ((4/8
   4'8
 $;8
 ((4/8
 ell*8
 +,8
 
 8
  8
r9   r  c                       e Zd Zy)"MinistralForSequenceClassificationNr>   r?   r@   r   r9   r8   r-  r-        r9   r-  c                       e Zd Zy)MinistralForTokenClassificationNr.  r   r9   r8   r1  r1    r/  r9   r1  c                       e Zd ZdZy)MinistralForQuestionAnsweringtransformerN)r>   r?   r@   r   r   r9   r8   r3  r3    s    %r9   r3  )r   r   r  r-  r1  r3  )r"   )r   )Ecollections.abcr   typingr   rI   r   activationsr   cache_utilsr   r	   
generationr
   integrationsr   r   r   masking_utilsr   r   modeling_flash_attention_utilsr   modeling_layersr   r   r   r   modeling_outputsr   r   modeling_rope_utilsr   r   modeling_utilsr   r   processing_utilsr   utilsr   r   r   utils.genericr   r    utils.output_capturingr!   configuration_ministralr#   Moduler%   rM   rX   r   r   rc   r   r|   r~   r   r   r   r   r   r  r-  r1  r3  __all__r   r9   r8   <module>rH     s   %    ! . ) f f R B  P K F & I I G 5 4299  ( *+ ,2	UU\\ 	U# 	U%,, 	U& %II%<<% 
% <<	%
 LL4'% % % '(%2 )*>) >) +>)B Y'Jryy J (J(+6 +\   $><ryy ><B U
- U
 U
p H
3_ H
 H
V	)IKc 		&CE] 	&$?AY &r9   