
    Uht                        d dl mZmZmZmZ d dlZd dlmZ ddlm	Z	 ddl
mZmZ ddlmZ ddlmZ ddlmZ dd	lmZ dd
lmZ ddlmZmZ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$m%Z%m&Z&m'Z'm(Z(m)Z) ddl*m+Z+  e(       rd dl,m-Z- ddl.m/Z/  e)j`                  e1      Z2 G d dejf                        Z4 G d de      Z5dejl                  de7dejl                  fdZ8	 d<dejf                  dejl                  dejl                  dejl                  d eejl                     d!e9d"e9fd#Z:d$ Z;d=d%Z< G d& d'ejf                        Z= G d( d)ee%      Z> ed*       G d+ d,ejf                               Z? G d- d.ejf                        Z@e& G d/ d0e!             ZAe& G d1 d2eA             ZBe& G d3 d4eAe             ZC e&d56       G d7 d8eA             ZDe& G d9 d:eA             ZEg d;ZFy)>    )CallableOptionalTupleUnionN   )ACT2FN)CacheDynamicCache)GenerationMixin)use_kernel_forward_from_hub)AttentionMaskConverter)FlashAttentionKwargs)GradientCheckpointingLayer)BaseModelOutputWithPastCausalLMOutputWithPast SequenceClassifierOutputWithPastTokenClassifierOutput)ROPE_INIT_FUNCTIONSdynamic_rope_update)ALL_ATTENTION_FUNCTIONSPreTrainedModel)Unpack)
LossKwargsauto_docstringcan_return_tupleis_torch_flex_attn_availablelogging   )
Glm4Config)	BlockMask)make_flex_block_causal_maskc                   V     e Zd Z fdZdej
                  dej
                  fdZ xZS )Glm4MLPc                 *   t         |           || _        t        j                  |j
                  d|j                  z  d      | _        t        j                  |j                  |j
                  d      | _        t        |j                     | _        y )N   Fbias)super__init__confignnLinearhidden_sizeintermediate_sizegate_up_proj	down_projr   
hidden_actactivation_fnselfr*   	__class__s     x/var/www/catia.catastroantioquia-mas.com/valormas/lib/python3.12/site-packages/transformers/models/glm4/modeling_glm4.pyr)   zGlm4MLP.__init__9   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/   chunkr2   r0   )r4   r8   	up_statesgates       r6   forwardzGlm4MLP.forwardA   sL    %%m4	#//!/4i 2 24 88	~~i((r7   )__name__
__module____qualname__r)   torchFloatTensorrA   __classcell__r5   s   @r6   r#   r#   8   s'    7)U%6%6 )5;L;L )r7   r#   c                   p    e Zd Zdedef fdZ	 	 	 	 	 	 	 ddej                  deej                     deej                     dee
   dee   d	ee   d
eej                     deeej                  ej                  f      dee   deej                  eeej                  ej                  f      f   fdZ xZS )Glm4DecoderLayerr*   	layer_idxc                    t         |           |j                  | _        t        ||      | _        t        |      | _        t        |j                  |j                        | _	        t        |j                  |j                        | _
        t        |j                  |j                        | _        t        |j                  |j                        | _        y )N)r*   rK   eps)r(   r)   r-   Glm4Attention	self_attnr#   mlpGlm4RMSNormrms_norm_epsinput_layernormpost_attention_layernormpost_self_attn_layernormpost_mlp_layernormr4   r*   rK   r5   s      r6   r)   zGlm4DecoderLayer.__init__K   s    !--&f	J6?*6+=+=6CVCVW(3F4F4FFL_L_(`%(3F4F4FFL_L_(`%"-f.@.@fFYFY"Zr7   r8   attention_maskposition_idspast_key_valueoutput_attentions	use_cachecache_positionposition_embeddingskwargsr9   c	                    |}
| j                  |      } | j                  d||||||||d|	\  }}| j                  |      }|
|z   }|}
| j                  |      }| j	                  |      }| j                  |      }|
|z   }|f}|r||fz  }|S )N)r8   rY   rZ   r[   r\   r]   r^   r_    )rT   rP   rV   rU   rQ   rW   )r4   r8   rY   rZ   r[   r\   r]   r^   r_   r`   residualself_attn_weightsoutputss                r6   rA   zGlm4DecoderLayer.forwardV   s     !,,]; ,:4>> 
,
')%)/) 3
,
 
,
(( 55mD =0 !55mD///> =0 ")++Gr7   )NNNFFNN)rB   rC   rD   r   intr)   rE   Tensorr   
LongTensorr	   boolr   r   r   rF   rA   rG   rH   s   @r6   rJ   rJ   J   s   	[z 	[c 	[ 2637*.,1$)59KO+||+ !.+ u//0	+
 !+ $D>+ D>+ !!1!12+ &eELL%,,,F&GH+ -.+ 
u  (51B1BEDUDU1U+V"WW	X+r7   rJ   r8   n_repr9   c                     | j                   \  }}}}|dk(  r| S | dddddddddf   j                  |||||      } | j                  |||z  ||      S )z
    This is the equivalent of torch.repeat_interleave(x, dim=1, repeats=n_rep). The hidden states go from (batch,
    num_key_value_heads, seqlen, head_dim) to (batch, num_attention_heads, seqlen, head_dim)
    r   N)shapeexpandreshape)r8   rj   batchnum_key_value_headsslenhead_dims         r6   	repeat_kvrs      so    
 2?1D1D.Ehz!!Qa"23::5BUW\^bdlmM  (;e(CT8TTr7   modulequerykeyvaluerY   scalingdropoutc                 T   t        || j                        }t        || j                        }	t        j                  ||j	                  dd            |z  }
|#|d d d d d d d |j
                  d   f   }|
|z   }
t        j                  j                  |
dt        j                        j                  |j                        }
t        j                  j                  |
|| j                        }
t        j                  |
|	      }|j	                  dd      j                         }||
fS )Nr%   r   r;   )r=   dtype)ptrainingr   )rs   num_key_value_groupsrE   matmul	transposerl   r+   
functionalsoftmaxfloat32tor|   ry   r~   
contiguous)rt   ru   rv   rw   rY   rx   ry   r`   
key_statesvalue_statesattn_weightscausal_maskattn_outputs                r6   eager_attention_forwardr      s    3 ; ;<JUF$?$?@L<<z';';Aq'ABWLL!$Q1.D
0@0@0D.D%DE#k1==((2U]](SVVW\WbWbcL==((6??([L,,|\:K''1-88:K$$r7   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   Nr%   r   r;   r<   r{   )rE   stackflatten)xx1x2s      r6   rotate_halfr      sJ    	
319B	
319B;;Ryb)11"55r7   c                    |j                  |      }|j                  |      }|dd|j                  d   dz  f   j                  dd      }|dd|j                  d   dz  f   j                  dd      }|j                  d   }| dd|f   | d|df   }}|dd|f   |d|df   }
}	||z  t        |      |z  z   }|	|z  t        |	      |z  z   }t	        j
                  ||gd      }t	        j
                  ||
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.
        position_ids (`torch.Tensor`, *optional*):
            Deprecated and unused.
        unsqueeze_dim (`int`, *optional*, defaults to 1):
            The 'unsqueeze_dim' argument specifies the dimension along which to unsqueeze cos[position_ids] and
            sin[position_ids] so that they can be properly broadcasted to the dimensions of q and k. For example, note
            that cos[position_ids] and sin[position_ids] have the shape [batch_size, seq_len, head_dim]. Then, if q and
            k have the shape [batch_size, heads, seq_len, head_dim], then setting unsqueeze_dim=1 makes
            cos[position_ids] and sin[position_ids] broadcastable to the shapes of q and k. Similarly, if q and k have
            the shape [batch_size, seq_len, heads, head_dim], then set unsqueeze_dim=2.
    Returns:
        `tuple(torch.Tensor)` comprising of the query and key tensors rotated using the Rotary Position Embedding.
    .Nr;   r%   r<   )	unsqueezerl   repeat_interleaver   rE   cat)qkcossinrZ   unsqueeze_dim
rotary_dimq_rotq_passk_rotk_passq_embedk_embeds                r6   apply_rotary_pos_embr      sD   ( --
&C
--
&C c'SYYr]a'''
(
:
:1"
:
EC
c'SYYr]a'''
(
:
:1"
:
EC 2Jc;J;&'3
+;)<6Ec;J;&'3
+;)<6E s{{51C78Gs{{51C78G ii&)r2Gii&)r2GGr7   c                   >    e Zd ZdZddedee   f fdZ	 	 ddej                  de
ej                  ej                  f   deej                     dee   d	eej                     d
ee   de
ej                  eej                     ee
ej                        f   fdZ xZS )rO   z=Multi-headed attention from 'Attention Is All You Need' paperr*   rK   c                 P   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
                  d      | _        y )Nrr   g      Tr&   F)r(   r)   r*   rK   getattrr-   num_attention_headsrr   rp   r   rx   attention_dropout	is_causalr+   r,   attention_biasq_projk_projv_projo_projrX   s      r6   r)   zGlm4Attention.__init__   sD   "
F4F4F&JdJd4de$*$>$>&B\B\$\!}}d*!'!9!9ii : :T]] JQWQfQf
 ii : :T]] JQWQfQf
 ii : :T]] JQWQfQf
 ii : :T]] JFL^L^ejkr7   r8   r_   rY   r[   r^   r`   r9   c                    |j                   d d }g |d| j                  }| j                  |      j                  |      j	                  dd      }	| j                  |      j                  |      j	                  dd      }
| j                  |      j                  |      j	                  dd      }|\  }}t        |	|
||      \  }	}
|'|||d}|j                  |
|| j                  |      \  }
}t        }| j                  j                  dk7  r^| j                  j                  dk(  r(|j                  dd      rt        j                  d	       nt         | j                  j                     } || |	|
||f| j"                  sd
n| j$                  | j&                  d|\  }} |j(                  g |d j+                         }| j-                  |      }||fS )Nr;   r   r%   )r   r   r^   eagersdpar\   Fz`torch.nn.functional.scaled_dot_product_attention` does not support `output_attentions=True`. Falling back to eager attention. This warning can be removed using the argument `attn_implementation="eager"` when loading the model.        )ry   rx   )rl   rr   r   viewr   r   r   r   updaterK   r   r*   _attn_implementationgetloggerwarning_oncer   r~   r   rx   rn   r   r   )r4   r8   r_   rY   r[   r^   r`   input_shapehidden_shapequery_statesr   r   r   r   cache_kwargsattention_interfacer   r   s                     r6   rA   zGlm4Attention.forward   s    $))#2.88b8$--8{{=166|DNNqRST[[/44\BLLQPQR
{{=166|DNNqRST&S#7jRUWZ#[ j%#&snUL'5'<'<ZW[WeWegs't$J(?;;++w6{{//69fjjI\^c>d##L
 '>dkk>^>^&_#$7	%
  $}}C$2H2HLL	%
 	%
!\ *k));;;;FFHkk+.L((r7   N)NN)rB   rC   rD   __doc__r   r   rf   r)   rE   rg   r   r	   rh   r   r   rA   rG   rH   s   @r6   rO   rO      s    Glz lhsm l4 +/590)||0) #5<<#=>0) !.	0)
 !0) !!1!120) -.0) 
u||Xell3XeELL>Q5RR	S0)r7   rO   c                       e Zd Zy)KwargsForCausalLMN)rB   rC   rD   rb   r7   r6   r   r   &  s    r7   r   RMSNormc                   ,     e Zd Zd fd	Zd Zd Z xZS )rR   c                     t         |           t        j                  t	        j
                  |            | _        || _        y)z:
        Glm4RMSNorm is equivalent to T5LayerNorm
        N)r(   r)   r+   	ParameterrE   onesweightvariance_epsilon)r4   r-   rN   r5   s      r6   r)   zGlm4RMSNorm.__init__+  s1     	ll5::k#:; #r7   c                 "   |j                   }|j                  t        j                        }|j	                  d      j                  dd      }|t        j                  || j                  z         z  }| j                  |j                  |      z  S )Nr%   r;   T)keepdim)	r|   r   rE   r   powmeanrsqrtr   r   )r4   r8   input_dtypevariances       r6   rA   zGlm4RMSNorm.forward3  sy    #))%((7 $$Q',,R,>%Ht?T?T4T(UU{{]--k:::r7   c                 ^    t        | j                  j                         d| j                   S )Nz, eps=)tupler   rl   r   r4   s    r6   
extra_reprzGlm4RMSNorm.extra_repr:  s*    ))*+6$2G2G1HIIr7   )gư>)rB   rC   rD   r)   rA   r   rG   rH   s   @r6   rR   rR   )  s    $;Jr7   rR   c                   ^     e Zd Zddef fdZ ej                         ed               Z xZ	S )Glm4RotaryEmbeddingr*   c                    t         |           t        |d      rG|j                  ;|j                  j	                  d|j                  j	                  d            | _        nd| _        |j                  | _        |j                  | _        || _	        t        | j
                     | _        | j                  | j                  |      \  }| _        | j                  d|d       | j                  | _        y )Nrope_scaling	rope_typetypedefaultinv_freqF)
persistent)r(   r)   hasattrr   r   r   max_position_embeddingsmax_seq_len_cachedoriginal_max_seq_lenr*   r   rope_init_fnattention_scalingregister_bufferr   original_inv_freq)r4   r*   devicer   r5   s       r6   r)   zGlm4RotaryEmbedding.__init__?  s    6>*v/B/B/N#0044[&BUBUBYBYZ`BabDN&DN"("@"@$*$B$B!/?+/+<+<T[[&+Q($(ZeD!%r7   c                 b   | j                   d d d d f   j                         j                  |j                  d   dd      j	                  |j
                        }|d d d d d f   j                         }t        |j
                  j                  t              r/|j
                  j                  dk7  r|j
                  j                  nd}t        j                  |d      5  |j                         |j                         z  j                  dd      }t        j                  ||fd	      }|j                         | j                  z  }|j                         | j                  z  }	d d d        j	                  |j                   
      	j	                  |j                   
      fS # 1 sw Y   AxY w)Nr   r;   r   mpscpuF)device_typeenabledr%   r<   )r|   )r   floatrm   rl   r   r   
isinstancer   strrE   autocastr   r   r   r   r   r|   )
r4   r   rZ   inv_freq_expandedposition_ids_expandedr   freqsembr   r   s
             r6   rA   zGlm4RotaryEmbedding.forwardP  sV    !MM$4-8>>@GGHZHZ[\H]_acdehhijiqiqr ,QaZ 8 > > @'1!((--'E!((--[`J`ahhmmfk^^UC 	5&,,.1F1L1L1NNYYZ[]^_E))UEN3C'')d444C'')d444C		5 vvAGGv$cff177f&;;;	5 	5s    BF%%F.r   )
rB   rC   rD   r   r)   rE   no_gradr   rA   rG   rH   s   @r6   r   r   >  s3    /z /" U]]_<  <r7   r   c                   F    e Zd ZeZdZdZdgZdgZdZ	dZ
dZdZdZdZdZd Zy)Glm4PreTrainedModelmodelTrJ   past_key_valuesc                    | j                   j                  }t        |t        j                        rY|j
                  j                  j                  d|       |j                  %|j                  j                  j                          y y t        |t        j                        rf|j
                  j                  j                  d|       |j                  2|j
                  j                  |j                     j                          y y t        |t              r&|j
                  j                  j                  d       y y )Nr   )r   stdg      ?)r*   initializer_ranger   r+   r,   r   datanormal_r'   zero_	Embeddingpadding_idxrR   fill_)r4   rt   r   s      r6   _init_weightsz!Glm4PreTrainedModel._init_weightso  s    kk++fbii(MM&&CS&9{{&  &&( '-MM&&CS&9!!-""6#5#56<<> .,MM$$S) -r7   N)rB   rC   rD   r   config_classbase_model_prefixsupports_gradient_checkpointing_no_split_modules_skip_keys_device_placement_supports_flash_attn_2_supports_sdpa_supports_flex_attn_supports_cache_class_supports_quantized_cache_supports_static_cache_supports_attention_backendr  rb   r7   r6   r   r   `  sS    L&*#+,#4"5!N  $!"&*r7   r   c                       e Zd Zdef fdZd Zd Zee	 	 	 	 	 	 	 	 	 dde	e
j                     de	e
j                     de	e
j                     de	e   d	e	e
j                     d
e	e   de	e   de	e   de	e
j                     dee   defd              Z	 ddee
j                  df   de
j                  de
j                  dedef
dZede
j                  dedede
j0                  de
j                  defd       Z xZS )	Glm4Modelr*   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 )NrM   )r*   F)r(   r)   pad_token_idr   
vocab_sizer+   r   r-   embed_tokens
ModuleListrangenum_hidden_layersrJ   layersrR   rS   normr   
rotary_embgradient_checkpointing	post_initrX   s      r6   r)   zGlm4Model.__init__  s     !.. ++LL):):F<N<NPTP`P`ammBGH`H`BabYfi0b
   2 28K8KL	-V<&+# 	 cs   Dc                     | j                   S r   r  r   s    r6   get_input_embeddingszGlm4Model.get_input_embeddings  s       r7   c                     || _         y r   r  r4   rw   s     r6   set_input_embeddingszGlm4Model.set_input_embeddings  s
    !r7   	input_idsrY   rZ   r   inputs_embedsr]   r\   output_hidden_statesr^   flash_attn_kwargsr9   c
                    ||n| j                   j                  }||n| j                   j                  }||n| j                   j                  }|d u |d uz  rt	        d      | j
                  r%| j                  r|rt        j                  d       d}t        |t        d       t        f      st	        d      || j                  |      }|r|
t               }|	F||j                         nd}t        j                   |||j"                  d   z   |j$                        }	||	j'                  d      }| j)                  |||	||      }|}| j+                  ||      }|rdnd }|rdnd }| j,                  d | j                   j.                   D ],  }|r||fz  } ||f||||||	|d	|
}|d   }|s$||d   fz  }. | j1                  |      }|r||fz  }t3        ||r|nd ||
      S )Nz:You must specify exactly one of input_ids or inputs_embedszX`use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`.FzBThe `past_key_values` should be either a `Cache` object or `None`.r   r   r   rb   )rY   rZ   r[   r\   r]   r^   r_   )last_hidden_stater   r8   
attentions)r*   r\   r%  r]   
ValueErrorr  r~   r   r   r   r   r	   r  r
   get_seq_lengthrE   arangerl   r   r   _update_causal_maskr  r  r  r  r   )r4   r#  rY   rZ   r   r$  r]   r\   r%  r^   r&  past_seen_tokensr   r8   r_   all_hidden_statesall_self_attnsdecoder_layerlayer_outputss                      r6   rA   zGlm4Model.forward  sT    2C1N-TXT_T_TqTq$8$D $++JjJj 	 "+!6IDKK<Q<Q	-t";<YZZ&&4==Yj I /DJ+>?abb  --i8M0*nO!CRC^==?de"\\ "2]5H5H5K"KTaThThN )33A6L..M>?L]
 & #oom\J #7BD0d![[)H4;;+H+HI 	6M#!m%55!)
*)."3#-$7
 $
M *!,M =#3"55'	6* 		-0  -!11&+/8Od+%	
 	
r7   r    input_tensorc           	         | j                   j                  dk(  r||dk(  j                         r|S y | j                   j                  dk(  r't        |t        j
                        rt        |      }|S ||j                         nd}||j                  nd}| j                   j                  dk(  r(|s&|s$t        j                  |||| j                        ry |j                  }|j                  d   }	|r|j                         }
n1t        |t        j
                        r|j                  d	   n||	z   dz   }
| j                  ||	|
|||j                  d   
      }| j                   j                  dk(  rQ|O|j                   j"                  dv r7|s5t	        j$                  |      j&                  }t        j(                  ||      }|S )Nflash_attention_2r   flex_attentionr   Fr   )r$  past_key_values_lengthis_trainingr   r;   )sequence_lengthtarget_lengthr|   r^   
batch_size)cudaxpunpu)r*   r   anyr   rE   rg   r!   r,  is_compileabler   _ignore_causal_mask_sdpar~   r|   rl   get_max_cache_shape5_prepare_4d_causal_attention_mask_with_cache_positionr   r   finfomin_unmask_unattended)r4   rY   r4  r^   r   r\   r/  using_compilable_cacher|   r:  r;  r   	min_dtypes                r6   r.  zGlm4Model._update_causal_mask  s    ;;++/BB)~/D.I.I.K%%;;++/??.%,,7!<^!L!!
 @O?Z?99;`aCRC^!?!?di ;;++v5>T]n%>>*'7 MM	 ""&,,Q/!+??AM nell; $$R(%7!;  PP+')#))!, Q 
 KK,,6*%%**.DD%
 E*..I0CCKQZ[Kr7   r:  r;  r|   r<  c                    | | j                         dk(  r| }|S t        j                  |      j                  }t        j                  ||f|||j
                        }|dk7  rt        j                  |d      }|t        j                  ||j
                        |j                  dd      kD  z  }|ddddddf   j                  |ddd      }| |j                         }| j                  d   }	|ddddddd|	f   | ddddddf   j                  |j
                        z   }
|
dk(  }
|ddddddd|	f   j                  |
|      |ddddddd|	f<   |S )	aM  
        Creates a causal 4D mask of shape `(batch_size, 1, query_length, key_value_length)` from a 2D mask of shape
        `(batch_size, key_value_length)`, or if the input `attention_mask` is already 4D, do nothing.

        Args:
            attention_mask (`torch.Tensor`):
                A 2D attention mask of shape `(batch_size, key_value_length)` or a 4D attention mask of shape
                `(batch_size, 1, query_length, key_value_length)`.
            sequence_length (`int`):
                The sequence length being processed.
            target_length (`int`):
                The target length: when generating with static cache, the mask should be as long as the static cache,
                to account for the 0 padding, the part of the cache that is not filled yet.
            dtype (`torch.dtype`):
                The dtype to use for the 4D attention mask.
            cache_position (`torch.Tensor`):
                Indices depicting the position of the input sequence tokens in the sequence.
            batch_size (`torch.Tensor`):
                Batch size.
        N   )
fill_valuer|   r   r   )diagonalr(  r;   r   )r=   rE   rE  rF  fullr   triur-  rn   rm   clonerl   r   masked_fill)rY   r:  r;  r|   r^   r<  r`   r   rI  mask_lengthpadding_masks              r6   rD  z?Glm4Model._prepare_4d_causal_attention_mask_with_cache_position9  s   < %.*<*<*>!*C(K* ' E*..I** -0Ye\j\q\qK !##jjqA5<<n>S>STWeWmWmnprsWtttK%dD!Q&67>>z1bRTUK))//1,2226*1aL[L+@ANSTVZ\`bcScDdDgDg&&E    ,q05@Aq,;,AV5W5c5c )6Aq!\k\12 r7   	NNNNNNNNN)F)rB   rC   rD   r   r)   r  r"  r   r   r   rE   rh   rg   r	   rF   ri   r   r   r   rA   r   r.  staticmethodrf   r|   rD  rG   rH   s   @r6   r  r  }  s   z  !"  151537+/59$(,0/359\
E,,-\
 !.\
 u//0	\

 "%\
   1 12\
 D>\
 $D>\
 'tn\
 !!1!12\
 $$89\
 
!\
  \
H #(BellK78B llB 	B
 B  BH 444 4 {{	4
 4 4 4r7   r  c                       e Zd ZdgZddiZddgdgfiZ fdZd Zd Zd	 Z	d
 Z
d Zd Zee	 	 	 	 	 	 	 	 	 	 	 ddeej"                     deej$                     deej"                     dee   deej(                     deej"                     dee   dee   dee   deej"                     deeej$                  f   dee   deeef   fd              Z xZS )Glm4ForCausalLMzlm_head.weightlm_headcolwise_repr8   logitsc                     t         |   |       t        |      | _        |j                  | _        t        j                  |j                  |j                  d      | _        | j                          y NFr&   )
r(   r)   r  r   r  r+   r,   r-   rX  r  r3   s     r6   r)   zGlm4ForCausalLM.__init__w  sU     v&
 ++yy!3!3V5F5FUS 	r7   c                 .    | j                   j                  S r   r   r  r   s    r6   r  z$Glm4ForCausalLM.get_input_embeddings      zz&&&r7   c                 &    || j                   _        y r   r^  r!  s     r6   r"  z$Glm4ForCausalLM.set_input_embeddings      "'

r7   c                     | j                   S r   rX  r   s    r6   get_output_embeddingsz%Glm4ForCausalLM.get_output_embeddings  s    ||r7   c                     || _         y r   rc  )r4   new_embeddingss     r6   set_output_embeddingsz%Glm4ForCausalLM.set_output_embeddings  s	    %r7   c                     || _         y r   r   )r4   decoders     r6   set_decoderzGlm4ForCausalLM.set_decoder  s	    
r7   c                     | j                   S r   ri  r   s    r6   get_decoderzGlm4ForCausalLM.get_decoder  s    zzr7   r#  rY   rZ   r   r$  labelsr]   r\   r%  r^   logits_to_keepr`   r9   c                    ||n| j                   j                  }|	|	n| j                   j                  }	 | 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 )ar  
        labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
            Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
            config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored
            (masked), the loss is only computed for the tokens with labels in `[0, ..., config.vocab_size]`.

        Example:

        ```python
        >>> from transformers import AutoTokenizer, Glm4ForCausalLM

        >>> model = Glm4ForCausalLM.from_pretrained("THUDM/GLM-4-9B-Chat-0414")
        >>> tokenizer = AutoTokenizer.from_pretrained("THUDM/GLM-4-9B-Chat-0414")

        >>> 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."
        ```N)	r#  rY   rZ   r   r$  r]   r\   r%  r^   )rZ  rn  r  lossrZ  r   r8   r*  rb   )r*   r\   r%  r   r)  r   rf   slicerX  loss_functionr  r   r   r8   r*  )r4   r#  rY   rZ   r   r$  rn  r]   r\   r%  r^   ro  r`   re   r8   slice_indicesrZ  rr  s                     r6   rA   zGlm4ForCausalLM.forward  s   N 2C1N-TXT_T_TqTq$8$D $++JjJj 	
 ,64:: ,
)%+'/!5),
 ,
  118B>SV8W~ot4]kmA}a,?@A%4%%pVFt{{OeOepiopD%#33!//))
 	
r7   )NNNNNNNNNNr   )rB   rC   rD   _tied_weights_keys_tp_plan_pp_planr)   r  r"  rd  rg  rk  rm  r   r   r   rE   rh   rg   r	   rF   ri   r   rf   r   r   r   r   rA   rG   rH   s   @r6   rW  rW  q  s   *+=)H_-z:;H'(&  151537+/59-1$(,0/35934G
E,,-G
 !.G
 u//0	G

 "%G
   1 12G
 ))*G
 D>G
 $D>G
 'tnG
 !!1!12G
 c5<</0G
 *+G
 
u,,	-G
  G
r7   rW  a  
    The Glm4 Model transformer with a sequence classification head on top (linear layer).

    [`Glm4ForSequenceClassification`] uses the last token in order to do the classification, as other causal models
    (e.g. GPT-2) do.

    Since it does classification on the last token, it requires to know the position of the last token. If a
    `pad_token_id` is defined in the configuration, it finds the last token that is not a padding token in each row. If
    no `pad_token_id` is defined, it simply takes the last value in each row of the batch. Since it cannot guess the
    padding tokens when `inputs_embeds` are passed instead of `input_ids`, it does the same (take the last value in
    each row of the batch).
    )custom_introc                       e Zd Z fdZd Zd Zee	 	 	 	 	 	 	 	 	 ddee	j                     dee	j                     dee	j                     dee   dee	j                     d	ee	j                     d
ee   dee   dee   defd              Z xZS )Glm4ForSequenceClassificationc                     t         |   |       |j                  | _        t        |      | _        t        j                  |j                  | j                  d      | _        | j                          y r\  )
r(   r)   
num_labelsr  r   r+   r,   r-   scorer  r3   s     r6   r)   z&Glm4ForSequenceClassification.__init__  sS      ++v&
YYv114??O
 	r7   c                 .    | j                   j                  S r   r^  r   s    r6   r  z2Glm4ForSequenceClassification.get_input_embeddings  r_  r7   c                 &    || j                   _        y r   r^  r!  s     r6   r"  z2Glm4ForSequenceClassification.set_input_embeddings  ra  r7   r#  rY   rZ   r   r$  rn  r]   r\   r%  r9   c
           
         | j                  ||||||||	      }
|
j                  }| j                  |      }||j                  d   }n|j                  d   }| j                  j
                  |dk7  rt        d      | j                  j
                  d}n||| j                  j
                  k7  j                  |j                  t        j                        }t        j                  |j                  d   |j                  t        j                        }||z  j                  d      }n.d}t        j                  | j                  j                    d       |t        j                  ||j                  	      |f   }d}|| j#                  |||| j                  
      }t%        |||
j&                  |
j(                  |
j*                        S )  
        labels (`torch.LongTensor` of shape `(batch_size,)`, *optional*):
            Labels for computing the sequence classification/regression loss. Indices should be in `[0, ...,
            config.num_labels - 1]`. If `config.num_labels == 1` a regression loss is computed (Mean-Square loss), If
            `config.num_labels > 1` a classification loss is computed (Cross-Entropy).
        rY   rZ   r   r$  r]   r\   r%  Nr   r   z=Cannot handle batch sizes > 1 if no padding token is defined.r;   )r   r|   z will not detect padding tokens in `inputs_embeds`. Results may be unexpected if using padding tokens in conjunction with `inputs_embeds.`r(  )rZ  rn  pooled_logitsr*   rq  )r   r)  r~  rl   r*   r  r+  r   r   rE   int32r-  argmaxr   r   r5   rB   rt  r   r   r8   r*  )r4   r#  rY   rZ   r   r$  rn  r]   r\   r%  transformer_outputsr8   rZ  r<  last_non_pad_tokennon_pad_masktoken_indicesr  rr  s                      r6   rA   z%Glm4ForSequenceClassification.forward  s   * 8<zz)%+'/!5 8B 	8
 ,==M* "+J&,,Q/J;;##+
a\]];;##+!#"%)A)AAEEfmmUZU`U`aL!LL)<V]]Z_ZeZefM"/,">!F!Fr!J!#>>**+ ,Z Z
 u||Jv}}MOaab%%VFR_hlhshs%tD/ /??-;;*55
 	
r7   rT  )rB   rC   rD   r)   r  r"  r   r   r   rE   rh   rg   r	   rF   ri   r   rA   rG   rH   s   @r6   r{  r{    s    '(  151537+/59-1$(,0/3A
E,,-A
 !.A
 u//0	A

 "%A
   1 12A
 ))*A
 D>A
 $D>A
 'tnA
 
*A
  A
r7   r{  c                       e Zd Z fdZd Zd Zee	 	 	 	 	 	 	 	 	 ddee	j                     dee	j                     dee	j                     dee   dee	j                     d	ee	j                     d
ee   dee   dee   defd              Z xZS )Glm4ForTokenClassificationc                    t         |   |       |j                  | _        t        |      | _        t        |dd       |j                  }nt        |dd       |j                  }nd}t        j                  |      | _
        t        j                  |j                  |j                        | _        | j                          y )Nclassifier_dropouthidden_dropoutg?)r(   r)   r}  r  r   r   r  r  r+   Dropoutry   r,   r-   r~  r  )r4   r*   r  r5   s      r6   r)   z#Glm4ForTokenClassification.__init__D  s      ++v&
6/6B!'!:!:V-t4@!'!6!6!$zz"45YYv1163D3DE
 	r7   c                 .    | j                   j                  S r   r^  r   s    r6   r  z/Glm4ForTokenClassification.get_input_embeddingsT  r_  r7   c                 &    || j                   _        y r   r^  r!  s     r6   r"  z/Glm4ForTokenClassification.set_input_embeddingsW  ra  r7   r#  rY   rZ   r   r$  rn  r]   r\   r%  r9   c
           
         | j                  ||||||||	      }
|
j                  }| j                  |      }| j                  |      }d}|| j	                  ||| j
                        }t        |||
j                  |
j                        S )r  r  N)rr  rZ  r8   r*  )	r   r)  ry   r~  rt  r*   r   r8   r*  )r4   r#  rY   rZ   r   r$  rn  r]   r\   r%  re   sequence_outputrZ  rr  s                 r6   rA   z"Glm4ForTokenClassification.forwardZ  s    * ,0::)%+'/!5 ,6 	,
 "33,,7O,%%ffdkkBD$!//))	
 	
r7   rT  )rB   rC   rD   r)   r  r"  r   r   r   rE   rh   rg   r	   rF   ri   r   rA   rG   rH   s   @r6   r  r  B  s     '(  151537+/59-1$(,0/3*
E,,-*
 !.*
 u//0	*

 "%*
   1 12*
 ))**
 D>*
 $D>*
 'tn*
 
*
  *
r7   r  )r   r  rW  r{  r  )r   )Nr   )Gtypingr   r   r   r   rE   torch.nnr+   activationsr   cache_utilsr	   r
   
generationr   integrationsr   modeling_attn_mask_utilsr   modeling_flash_attention_utilsr   modeling_layersr   modeling_outputsr   r   r   r   modeling_rope_utilsr   r   modeling_utilsr   r   processing_utilsr   utilsr   r   r   r   r   configuration_glm4r   !torch.nn.attention.flex_attentionr    integrations.flex_attentionr!   
get_loggerrB   r   Moduler#   rJ   rg   rf   rs   r   r   r   r   rO   r   rR   r   r   r  rW  r{  r  __all__rb   r7   r6   <module>r     s#  , 4 3   ! . ) 7 > B 9  L F & h h *  !;J 
		H	%)bii )$71 7t	UU\\ 	U# 	U%,, 	U& %II%<<% 
% <<	%
 U\\*% % %46'TH)BII H)V ?,j > Y'J")) J (J(<")) <D */ * *8 p# p pf i
)? i
 i
X S
$7 S
S
l C
!4 C
 C
Lr7   