
    Z j+                        S SK Jr  S SKJr  S SKrS SKJs  Jr  S SKJr  SSK	J
r  SSKJr  SSKJrJr  SS	KJr  SS
KJrJrJr  SSKJrJr  SSKJr  SSKJr  SSKJrJ r   SSK!J"r"J#r#  SSK$J%r%J&r&  SSK'J(r(  SSK)J*r*J+r+  SSK,J-r-J.r.J/r/  SSK0J1r1J2r2  SSK3J4r4  \" S5       " S S\Rj                  5      5       r6 " S S\Rj                  5      r7 " S S\Rj                  5      r8 " S S\Rj                  5      r9\ " S  S!\Rj                  5      5       r: " S" S#\Rj                  5      r;S$ r<SAS% jr=S&\R|                  S'\?S(\R|                  4S) jr@ SBS*\Rj                  S+\R|                  S,\R|                  S-\R|                  S.\R|                  S-  S/\AS0\AS1\(\-   4S2 jjrB\" \=5       " S3 S4\Rj                  5      5       rC " S5 S6\5      rD\* " S7 S8\&5      5       rE\* " S9 S:\E5      5       rF   SCS;\R|                  \G\R|                     -  S-  S<\?S-  S.\R|                  S-  S(\R|                  \?-  4S= jjrH\* " S> S?\E\5      5       rI/ S@QrJg)D    )Callable)OptionalN)nn   )initialization)ACT2FN)CacheDynamicCache)GenerationMixin)use_experts_implementationuse_kernel_forward_from_hubuse_kernelized_func)create_causal_mask!create_sliding_window_causal_mask)FlashAttentionKwargs)GradientCheckpointingLayer)MoeCausalLMOutputWithPastMoeModelOutputWithPast)ROPE_INIT_FUNCTIONSdynamic_rope_update)ALL_ATTENTION_FUNCTIONSPreTrainedModel)Unpack)auto_docstringcan_return_tuple)TransformersKwargsmaybe_autocastmerge_with_config_defaults)OutputRecordercapture_outputs   )LagunaConfigRMSNormc                   x   ^  \ rS rSrS
S\SS4U 4S jjjrS\R                  S\R                  4S jrS r	S	r
U =r$ )LagunaRMSNorm.   epsreturnNc                    > [         TU ]  5         [        R                  " [        R
                  " U5      5      U l        X l        g)z,
LagunaRMSNorm is equivalent to T5LayerNorm
N)super__init__r   	Parametertorchonesweightvariance_epsilon)selfhidden_sizer'   	__class__s      {/root/GenerationalWealth/GenerationalWealth/venv/lib/python3.13/site-packages/transformers/models/laguna/modeling_laguna.pyr+   LagunaRMSNorm.__init__0   s/     	ll5::k#:; #    hidden_statesc                    UR                   nUR                  [        R                  5      nUR	                  S5      R                  SSS9nU[        R                  " X0R                  -   5      -  nU R                  UR                  U5      -  $ )N   T)keepdim)	dtypetor-   float32powmeanrsqrtr0   r/   )r1   r7   input_dtypevariances       r4   forwardLagunaRMSNorm.forward8   sw    #))%((7 $$Q',,R,>%H?T?T4T(UU{{]--k:::r6   c                 ^    [        U R                  R                  5       SU R                   3$ )Nz, eps=)tupler/   shaper0   )r1   s    r4   
extra_reprLagunaRMSNorm.extra_repr?   s*    ))*+6$2G2G1HIIr6   )r0   r/   )gư>)__name__
__module____qualname____firstlineno__floatr+   r-   TensorrD   rI   __static_attributes____classcell__r3   s   @r4   r%   r%   .   sB    $ $$ $ $;U\\ ;ell ;J Jr6   r%   c                      ^  \ rS rSr% \R
                  \S'   S\4U 4S jjr\	    SS\S-  S\
S   S\S-  S	\S-  S
\S\4   4
S jj5       r\R                   " 5       \SS j5       5       rSrU =r$ )LagunaRotaryEmbeddingC   inv_freqconfigc                 f  > [         TU ]  5         UR                  U l        UR                  U l        Xl        [        [        UR                  5      5      U l        0 U l	        U R                   H  nU R
                  R                  U   nUc  M!  US   U R                  U'   U R                  nU R                  U   S:w  a  [        U R                  U      nU" U R
                  US9u  pVU R                  U S3USS9  U R                  U S3UR                  5       SS9  [        X S3U5        M     g )	N	rope_typedefault
layer_type	_inv_freqF)
persistent_original_inv_freq_attention_scaling)r*   r+   max_position_embeddingsmax_seq_len_cachedoriginal_max_seq_lenrX   listsetlayer_typesrZ   rope_parameterscompute_default_rope_parametersr   register_bufferclonesetattr)r1   rX   r]   rope_paramsrope_init_fncurr_inv_freqcurr_attention_scalingr3   s          r4   r+   LagunaRotaryEmbedding.__init__F   s(   "("@"@$*$B$B!F$6$6 78**J++55jAK")4[)ADNN:&%)%I%IL~~j)Y624>>*3MN4@Yc4d1M  J<y!9=UZ [  J</A!BMDWDWDYfk lDL(:;=ST +r6   Ndeviceztorch.deviceseq_lenr]   r(   ztorch.Tensorc           	      v   U R                   U   S   nU R                   U   R                  SS5      n[        U SS5      =(       d    U R                  U R                  -  n[        Xe-  5      nSnSU[        R                  " SUS[        R                  S9R                  U[        R                  S	9U-  -  -  n	X4$ )
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.
    layer_type (`str`, *optional*):
        The current layer type if the model has different RoPE parameters per type.
        Should not be used unless `config.layer_types is not None`
Returns:
    Tuple of (`torch.Tensor`, `float`), containing the inverse frequencies for the RoPE embeddings and the
    post-processing scaling factor applied to the computed cos/sin (unused in this type of RoPE).

rope_thetapartial_rotary_factorg      ?head_dimNr   r9   r<   )rr   r<   )rh   getgetattrr2   num_attention_headsintr-   arangeint64r=   rO   )
rX   rr   rs   r]   baserv   rw   dimattention_factorrW   s
             r4   ri   5LagunaRotaryEmbedding.compute_default_rope_parameters[   s    . %%j1,? & 6 6z B F FG^`c d6:t4h8J8JfNhNh8h(23 U\\!S!5;;?BB&X]XcXcBdgjjk
 ))r6   c                 H   [        X S35      n[        X S35      nUS S S 2S 4   R                  5       R                  UR                  S   SS5      R	                  UR
                  5      nUS S 2S S S 24   R                  5       n[        UR
                  R                  [        5      (       a0  UR
                  R                  S:w  a  UR
                  R                  OSn[        USS	9   UR                  5       UR                  5       -  R                  SS
5      n	[        R                  " X4SS9n
U
R                  5       U-  nU
R                  5       U-  nS S S 5        WR	                  UR                  S9WR	                  UR                  S94$ ! , (       d  f       N@= f)Nr^   ra   r   r:   r!   mpscpuF)device_typeenabledr9   r   rx   )rz   rO   expandrH   r=   rr   
isinstancetypestrr   	transposer-   catcossinr<   )r1   xposition_idsr]   rW   attention_scalinginv_freq_expandedposition_ids_expandedr   freqsembr   r   s                r4   rD   LagunaRotaryEmbedding.forward   sd    4<y!9:#DL8J*KL$T1d]399;BB<CUCUVWCXZ\^_`ccdedldlm ,QaZ 8 > > @'1!((--'E'E!((--[`J`ahhmmfkUC&,,.1F1L1L1NNYYZ[]^_E))UN3C'')//C'')//C	 D vvAGGv$cff177f&;;; DCs   +A.F
F!)rX   rg   rc   rd   rZ   )NNNNN)rK   rL   rM   rN   r-   rP   __annotations__r"   r+   staticmethodr   r|   r   rG   rO   ri   no_gradr   rD   rQ   rR   rS   s   @r4   rU   rU   C   s    llU| U* &*+/"!%	"*t#"*("* t"* $J	"*
 
~u$	%"* "*H ]]_<  <r6   rU   c                   2   ^  \ rS rSrSU 4S jjrS rSrU =r$ )	LagunaMLP   c                   > [         TU ]  5         Xl        UR                  U l        Uc  UR                  OUU l        [
        R                  " U R                  U R                  SS9U l        [
        R                  " U R                  U R                  SS9U l        [
        R                  " U R                  U R                  SS9U l	        [        UR                     U l        g NFbias)r*   r+   rX   r2   intermediate_sizer   Linear	gate_projup_proj	down_projr   
hidden_actact_fn)r1   rX   r   r3   s      r4   r+   LagunaMLP.__init__   s    !--=N=V!9!9\m4#3#3T5K5KRWXyy!1!143I3IPUV4#9#94;K;KRWXV../r6   c                     U R                  U R                  U R                  U5      5      U R                  U5      -  5      nU$ r   )r   r   r   r   )r1   r   r   s      r4   rD   LagunaMLP.forward   s6    NN4;;t~~a/@#ADLLQRO#ST	r6   )r   rX   r   r   r2   r   r   r   )rK   rL   rM   rN   r+   rD   rQ   rR   rS   s   @r4   r   r      s    0 r6   r   c                      ^  \ rS rSrU 4S jrS\R                  S\\R                  \R                  \R                  4   4S jrSr	U =r
$ )LagunaTopKRouter   c                   > [         TU ]  5         UR                  U l        UR                  U l        UR
                  U l        [        R                  " [        R                  " U R                  U R                  5      5      U l        [        R                  " [        R                  " UR                  5      SS9U l        UR                  U l        g )NF)requires_grad)r*   r+   num_experts_per_toktop_knum_expertsr2   
hidden_dimr   r,   r-   zerosr/   e_score_correction_biasmoe_router_logit_softcappingrouter_logit_softcappingr1   rX   r3   s     r4   r+   LagunaTopKRouter.__init__   s    //
!-- ,,ll5;;t/?/?#QR')||EKK@R@R4Sch'i$(.(K(K%r6   r7   r(   c                 V   UR                  SU R                  5      n[        R                  " XR                  5      R                  5       nU R                  S:  a/  [        R                  " X R                  -  5      U R                  -  n[        R                  " U5      nX0R                  R                  UR                  5      -   n[        R                  " X@R                  SS9u  pVUR                  SU5      nXwR!                  SSS9-  nUR                  UR                  5      nX'U4$ )Nr:           r   T)r   r;   )reshaper   Flinearr/   rO   r   r-   tanhsigmoidr   r=   r<   topkr   gathersum)r1   r7   router_logitsrouting_scoresscores_for_selection_selected_expertsrouting_weightss           r4   rD   LagunaTopKRouter.forward   s     &--b$//B<BBD((3.!JJ}7T7T'TUX\XuXuuM}5-0L0L0O0OP^PdPd0ee#jj)=zzrR(//4DE),?,?BPT,?,UU),,]-@-@A/???r6   )r   r   r   r   r   r/   )rK   rL   rM   rN   r+   r-   rP   rG   rD   rQ   rR   rS   s   @r4   r   r      sE    L@||@ 
u||U\\5<<7	8@ @r6   r   c                      ^  \ rS rSrSrU 4S jrS\R                  S\R                  S\R                  S\R                  4S jrS	r	U =r
$ )
LagunaExperts   z2Collection of expert weights stored as 3D tensors.c                   > [         TU ]  5         UR                  U l        UR                  U l        UR
                  U l        [        R                  " [        R                  " U R                  SU R                  -  U R                  5      5      U l        [        R                  " [        R                  " U R                  U R                  U R                  5      5      U l        [        UR                     U l        g )Nr9   )r*   r+   r   r2   r   moe_intermediate_sizeintermediate_dimr   r,   r-   emptygate_up_projr   r   r   r   r   s     r4   r+   LagunaExperts.__init__   s    !-- ,, & < <LLT5E5Eq4K`K`G`bfbqbq)rsekk$2B2BDOOUYUjUj&klV../r6   r7   top_k_indextop_k_weightsr(   c                 X   [         R                  " U5      n[         R                  " 5          [         R                  R                  R                  X R                  S9nUR                  SSS5      n[         R                  " UR                  SS9S5      R                  5       nS S S 5        W H  nUS   nXpR                  :X  a  M  [         R                  " WU   5      u  pX   n
[        R                  R                  XR                  U   5      R                  SSS9u  pU R                  U5      U-  n[        R                  R                  XR                   U   5      nXXS 4   -  nUR#                  SXR%                  UR&                  5      5        M     U$ ! , (       d  f       N= f)N)num_classesr9   r!   r   )r:   r   r:   )r-   
zeros_liker   r   
functionalone_hotr   permutegreaterr   nonzerowherer   r   chunkr   r   
index_add_r=   r<   )r1   r7   r   r   final_hidden_statesexpert_mask
expert_hit
expert_idx	top_k_pos	token_idxcurrent_stategateupcurrent_hidden_statess                 r4   rD   LagunaExperts.forward   so    $..}=]]_((--55kO_O_5`K%--aA6K{8'DaHPPRJ 
 %J#AJ---#(;;{:/F#G I)4M}}++M;L;LZ;XY__`agi_jHD$(KK$5$:!$&MM$8$89NP^P^_iPj$k!$9)`dJd<e$e!**1i9Q9QReRkRk9lm % #"# _s   A7F
F))r   r   r   r   r   r   )rK   rL   rM   rN   __doc__r+   r-   rP   rD   rQ   rR   rS   s   @r4   r   r      sK    <0#||# \\# ||	#
 
# #r6   r   c                   j   ^  \ rS rSrS\4U 4S jjrS\R                  S\R                  4S jrSr	U =r
$ )LagunaSparseMoeBlock   rX   c                    > [         TU ]  5         [        U5      U l        [	        U5      U l        [        XR                  S9U l        UR                  U l
        g )Nr   )r*   r+   r   expertsr   r   r   shared_expert_intermediate_sizeshared_expertsmoe_routed_scaling_factorrouted_scaling_factorr   s     r4   r+   LagunaSparseMoeBlock.__init__   sG    $V,$V,	'BhBhi%+%E%E"r6   r7   r(   c                     UR                   u  p#nUR                  SU5      nU R                  U5      nU R                  U5      u  pgnU R	                  XU5      nXR
                  -  nX-   nUR                  X#U5      nU$ )Nr:   )rH   viewr   r   r   r   r   )	r1   r7   
batch_sizesequence_lengthr   shared_outputr   r   r   s	            r4   rD   LagunaSparseMoeBlock.forward   s    2?2E2E/
Z%**2z:++M:/3yy/G,,]oV%(B(BB%5%--j:Vr6   )r   r   r   r   )rK   rL   rM   rN   r"   r+   r-   rP   rD   rQ   rR   rS   s   @r4   r   r      s1    F| FU\\ ell  r6   r   c                     U SSU R                   S   S-  24   nU SU R                   S   S-  S24   n[        R                  " U* U4SS9$ )z*Rotates half the hidden dims of the input..Nr:   r9   r   )rH   r-   r   )r   x1x2s      r4   rotate_halfr    sZ    	
3"!''"+"""	#B	
3q ""	#B99rc2YB''r6   c                 R   UR                  U5      nUR                  U5      nUR                  S   nU SSU24   U SUS24   pvUSSU24   USUS24   pXb-  [        U5      U-  -   n
X-  [        U5      U-  -   n[        R                  " X/SS9n
[        R                  " X/SS9nX4$ )a{  Applies Rotary Position Embedding to the query and key tensors.

Removes the interleaving of cos and sin from GLM

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:   .Nr   )	unsqueezerH   r  r-   r   )qkr   r   unsqueeze_dim
rotary_dimq_rotq_passk_rotk_passq_embedk_embeds               r4   apply_rotary_pos_embr  	  s    ( --
&C
--
&C 2Jc;J;&'3
+;)<6c;J;&'3
+;)<6 {{51C78G{{51C78G ii)r2Gii)r2Gr6   r7   n_repr(   c                     U R                   u  p#pEUS:X  a  U $ U SS2SS2SSS2SS24   R                  X#XU5      n U R                  X#U-  XE5      $ )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   r   r   )r7   r  batchnum_key_value_headsslenrw   s         r4   	repeat_kvr  /  s_    
 2?1D1D.Ez!!Qa"23::5W\dlmM  e(CTTTr6   modulequerykeyvalueattention_maskscalingdropoutkwargsc                    [        X R                  5      n[        X0R                  5      n	[        R                  " XR	                  SS5      5      U-  n
Ub  X-   n
[
        R                  R                  U
S[        R                  S9R                  UR                  5      n
[
        R                  R                  XU R                  S9n
[        R                  " X5      nUR	                  SS5      R                  5       nX4$ )Nr9   r   r:   )r   r<   )ptrainingr!   )r  num_key_value_groupsr-   matmulr   r   r   softmaxr>   r=   r<   r  r#  
contiguous)r  r  r  r  r  r  r  r   
key_statesvalue_statesattn_weightsattn_outputs               r4   eager_attention_forwardr,  ;  s     3 ; ;<JU$?$?@L<<';';Aq'ABWLL!#4==((2U]](SVVW\WbWbcL==((6??([L,,|:K''1-88:K$$r6   c                     ^  \ rS rSrSrS\S\S\4U 4S jjr SS\R                  S	\
\R                  \R                  4   S
\R                  S-  S\S-  S\\   S\
\R                  \R                  S-  4   4S jjrSrU =r$ )LagunaAttentioniT  zSAfmoe-style SWA/GQA attention with Laguna-specific gating and per-layer head count.rX   	layer_idx	num_headsc                   > [         TU ]  5         X0l        Xl        X l        [        USUR                  UR                  -  5      U l        U R                  UR                  -  U l
        U R                  S-  U l        UR                  U l        SU l        [        R                  " UR                  U R                  U R                  -  UR                   S9U l        [        R                  " UR                  UR                  U R                  -  UR                   S9U l        [        R                  " UR                  UR                  U R                  -  UR                   S9U l        [        R                  " U R                  U R                  -  UR                  UR                   S9U l        UR*                  U   S:H  U l        U R,                  (       a  UR.                  OS U l        [1        U R                  UR2                  S9U l        [1        U R                  UR2                  S9U l        [        R                  " UR                  U R                  SS9U l        g )Nrw   g      Tr   sliding_attentionr'   F)r*   r+   r0  rX   r/  rz   r2   r{   rw   r  r$  r  attention_dropout	is_causalr   r   attention_biasq_projk_projv_projo_projrg   is_local_attentionsliding_windowr%   rms_norm_epsq_normk_normg_proj)r1   rX   r/  r0  r3   s       r4   r+   LagunaAttention.__init__X  s   ""
F4F4F&JdJd4de$(NNf6P6P$P!}}d*!'!9!9ii 2 2DNNT]]4RY_YnYnoii : :T]] JQWQfQf
 ii : :T]] JQWQfQf
 ii >@R@RY_YnYno #)"4"4Y"?CV"V7;7N7Nf33TX#DMMv7J7JK#DMMv7J7JKii 2 2DNNOr6   Nr7   position_embeddingsr  past_key_valuesr   r(   c                    UR                   S S n/ UQSPU R                  P7nU R                  U5      R                  U5      nU R	                  U5      R                  U5      n	U R                  U5      R                  U5      n
U R                  U5      R                  SS5      nU R                  U	5      R                  SS5      n	U
R                  SS5      n
Uu  p[        XX5      u  pUb  UR                  XU R                  5      u  p[        R                  " U R                  R                  [         5      nU" U UU	U
U4U R"                  (       d  SOU R$                  U R&                  U R(                  S.UD6u  pUR*                  " / UQSP76 R-                  5       n[.        R0                  " U R3                  U5      R5                  5       5      R7                  UR8                  5      nUR                  " / UQSPU R                  P76 UR;                  S5      -  R                  " / UQSP76 nU R=                  U5      nX4$ )Nr:   r!   r9   r   )r  r  r<  )rH   rw   r7  r   r8  r9  r>  r   r?  r  updater/  r   get_interfacerX   _attn_implementationr,  r#  r4  r  r<  r   r'  r   softplusr@  rO   r=   r<   r  r:  )r1   r7   rB  r  rC  r   input_shapehidden_shapequery_statesr(  r)  r   r   attention_interfacer+  r*  r   s                    r4   rD   LagunaAttention.forwardu  s    $))#2.88b8$--8{{=166|D[[/44\B
{{=166|D{{<0::1a@[[,66q!<
#--a3&#7RU#[ &'6'='=jX\XfXf'g$J(?(M(MKK,,.E)
 %8
%
  $}}C$2H2HLL..
%
 
%
! "));;;;FFHzz$++m4::<=@@ARARS"''HHbH$--H4>>Z\K]]ccuepurtukk+.((r6   )r4  rX   r@  rw   r5  r;  r?  r8  r/  r0  r$  r:  r>  r7  r  r<  r9  r   )rK   rL   rM   rN   r   r"   r|   r+   r-   rP   rG   r	   r   r   rD   rQ   rR   rS   s   @r4   r.  r.  T  s    ]P| P P PD )-.)||.) #5<<#=>.) t+	.)
 .) -..) 
u||U\\D00	1.) .)r6   r.  c                     ^  \ rS rSrS\S\4U 4S jjr     SS\R                  S\R                  S-  S\R                  S-  S	\
S-  S
\S-  S\\R                  \R                  4   S-  S\\   S\R                  4S jjrSrU =r$ )LagunaDecoderLayeri  rX   r/  c                   > [         TU ]  5         UR                  U l        [        XUR                  U   5      U l        UR                  U   S:X  a  [        U5      U l        O[        XR                  S9U l        [        UR                  UR                  S9U l        [        UR                  UR                  S9U l        g )Nsparser   r3  )r*   r+   r2   r.  num_attention_heads_per_layer	self_attnmlp_layer_typesr   mlpr   r   r%   r=  input_layernormpost_attention_layernormr1   rX   r/  r3   s      r4   r+   LagunaDecoderLayer.__init__  s    !--(F<`<`aj<kl!!),8+F3DH ;S;STDH,V-?-?VEXEXY(5f6H6HfNaNa(b%r6   Nr7   r  r   rC  	use_cacherB  r   r(   c           
          UnU R                  U5      nU R                  " SUUUUUUS.UD6u  pX-   nUnU R                  U5      nU R                  U5      nX-   nU$ )N)r7   r  r   rC  rZ  rB   )rV  rS  rW  rU  )
r1   r7   r  r   rC  rZ  rB  r   residualr   s
             r4   rD   LagunaDecoderLayer.forward  s     !,,];>> 
')%+ 3
 
 !0 !55mD/ 0r6   )r2   rV  rU  rW  rS  )NNNFN)rK   rL   rM   rN   r"   r|   r+   r-   rP   
LongTensorr	   boolrG   r   r   rD   rQ   rR   rS   s   @r4   rO  rO    s    	c| 	c 	c /304(,!&HL|| t+ &&-	
  $; #5<<#=>E +, 
 r6   rO  c                      ^  \ rS rSr% \\S'   SrSrS/rS/r	Sr
SrSrSrSr\" \SS9\\S	.r\R*                  " 5       U 4S
 j5       rSrU =r$ )LagunaPreTrainedModeli  rX   modelTrO  rC  r   )index)r   r7   
attentionsc                 f  > [         TU ]  U5        U R                  R                  n[	        U[
        5      (       aA  [        R                  " UR                  SUS9  [        R                  " UR                  SUS9  O5[	        U[        5      (       a   [        R                  " UR                  SUS9  [	        U[        5      (       a4  [        R                  R                  R                  UR                  5        g [	        U[         5      (       a  UR"                   H  nUR$                  nUR&                  U   S:w  a  [(        UR&                  U      nU" UR                  US9u  pV[        R*                  " [-        X S35      U5        [        R*                  " [-        X S35      U5        M     g g )Nr   )r@   stdr[   r\   r^   r`   )r*   _init_weightsrX   initializer_ranger   r   initnormal_r   r   r   r/   r-   r   zeros_r   rU   rg   ri   rZ   r   copy_rz   )r1   r  rg  r]   rn   ro   r   r3   s          r4   rh  #LagunaPreTrainedModel._init_weights  sB   f%kk++fm,,LL,,3C@LL))= 011LLSc:f.//HHMM  !?!?@ 566$00
%EE##J/9<#6v7G7G
7S#TL#/*#U 

76\+CDmT

76\9K+LM}] 1 7r6   r\  )rK   rL   rM   rN   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   rO  r.  _can_record_outputsr-   r   rh  rQ   rR   rS   s   @r4   rb  rb    sy    &*#-.#4"5N!"&'(8B+% ]]_^ ^r6   rb  c                     ^  \ rS rSrS\4U 4S jjr\\\      SS\	R                  S-  S\	R                  S-  S\	R                  S-  S\S-  S	\	R                  S-  S
\S-  S\\   S\4S jj5       5       5       rSrU =r$ )LagunaModeli  rX   c           	        > [         TU ]  U5        UR                  U l        UR                  U l        [
        R                  " UR                  UR                  U R                  5      U l        [
        R                  " [        UR                  5       Vs/ s H  n[        X5      PM     sn5      U l        [        UR                  UR                  S9U l        [#        US9U l        SU l        U R)                  5         g s  snf )Nr3  rX   F)r*   r+   pad_token_idpadding_idx
vocab_sizer   	Embeddingr2   embed_tokens
ModuleListrangenum_hidden_layersrO  layersr%   r=  normrU   
rotary_embgradient_checkpointing	post_initrX  s      r4   r+   LagunaModel.__init__  s     !.. ++LL):):F<N<NPTP`P`ammDI&JbJbDcdDcy2Dcd
 "&"4"4&:M:MN	/v>&+# 	 es   C?N	input_idsr  r   rC  inputs_embedsrZ  r   r(   c           	        ^ US L US L-  (       a  [        S5      eUc  U R                  U5      nU(       a  Uc  [        U R                  S9nUcU  Ub  UR	                  5       OSn[
        R                  " UR                  S   UR                  S9U-   nUR                  S5      n[        U=n	[        5      (       dR  U R                  UUUUS.mU4S jU4S jS	.n
0 n	[        U R                  R                  5       H  nX   " 5       X'   M     Un0 n[        U R                  R                  5       H  nU R                  XU5      X'   M     [        U R                   S U R                  R"                   5       HE  u  pU" U4XR                  R                  U      XR                  R                  U      UUS
.UD6nMG     U R%                  U5      n['        UU(       a  US9$ S S9$ )Nz:You must specify exactly one of input_ids or inputs_embedsr|  r   r!   )rr   )rX   r  r  rC  r   c                     > [        S0 T D6$ Nr\  )r   mask_kwargss   r4   <lambda>%LagunaModel.forward.<locals>.<lambda>0  s    *<*K{*Kr6   c                     > [        S0 T D6$ r  )r   r  s   r4   r  r  1  s    -N-]Q\-]r6   )full_attentionr2  )r  rB  r   rC  )last_hidden_staterC  )
ValueErrorr  r
   rX   get_seq_lengthr-   r}   rH   rr   r  r   dictrf   rg   r  	enumerater  r  r  r   )r1   r  r  r   rC  r  rZ  r   past_seen_tokenscausal_mask_mappingmask_creation_functionsr]   r7   rB  idecoder_layerr  s                   @r4   rD   LagunaModel.forward  s    -t";<YZZ  --i8M0*$++>OCRC^==?de <<(;(;A(>}G[G[\_ooL'11!4L?-FF++!."0#2 ,K #L%]'# #%!$++"9"9:
2I2U2W#/ ; & dkk556J.2oom[e.f+ 7 !*$++6U8U8U*V WA)2;;3J3J13MN$78O8OPQ8R$S) / M !X 		-0%+/8O
 	
>B
 	
r6   )r  r  r  r  r~  r  r  )NNNNNN)rK   rL   rM   rN   r"   r+   r   r    r   r-   r_  rP   r	   FloatTensorr`  r   r   r   rD   rQ   rR   rS   s   @r4   rz  rz    s    |     .2.204(,26!%<
##d*<
 t+<
 &&-	<

 <
 ((4/<
 $;<
 +,<
 
 <
    <
r6   rz  gate_logitsr   c                    U b  [        U [        5      (       d  g[        U [        5      (       aC  U S   R                  n[        R                  " U  Vs/ s H  oUR                  U5      PM     snSS9n[        R                  R                  R                  WSS9n[        R                  " XrSS9u  p[        R                  R                  R                  X5      n
Uc:  [        R                  " U
R                  5       SS9n[        R                  " USS9nGOUR                  u  pUR                  S   X-  -  nUSSS2SS2SS4   R                  XXU45      R                  SX!5      R                  W5      n[        R                   " U
R                  5       U-  SS9[        R                   " USS9-  nUSSS2SS2S4   R                  XX45      R                  SU5      R                  U5      n[        R                   " UU-  SS9[        R                   " USS9-  n[        R                   " XR#                  S5      -  5      nUU-  $ s  snf )ax  
Computes auxiliary load balancing loss as in Switch Transformer - implemented in Pytorch.

See Switch Transformer (https://huggingface.co/papers/2101.03961) for more details. This function implements the loss
function presented in equations (4) - (6) of the paper. It aims at penalizing cases where the routing between
experts is too unbalanced.

Args:
    gate_logits:
        Logits from the `gate`, should be a tuple of model.config.num_hidden_layers tensors of
        shape [batch_size X sequence_length, num_experts].
    num_experts:
        Number of experts
    top_k:
        The number of experts to route per-token, can be also interpreted as the `top-k` routing
        parameter.
    attention_mask (`torch.Tensor`, *optional*):
        The attention_mask used in forward function
        shape [batch_size X sequence_length] if not None.

Returns:
    The auxiliary loss.
Nr   r   r:   )r   rG   rr   r-   r   r=   r   r   r&  r   r   r@   rO   rH   r   r   r   r  )r  r   r   r  compute_device
layer_gateconcatenated_gate_logitsr   r   r   r   tokens_per_expertrouter_prob_per_expertr   r   r  expert_attention_mask router_per_expert_attention_maskoverall_losss                      r4   load_balancing_loss_funcr  N  s+   : *[%"@"@+u%%$Q..#(99^i-j^iPZmmN.K^i-jpq#r hh))112JPR1SO**_DA((%%--.>LK!JJ{'8'8':B "'O!C&4&:&:#
4::1=*B^_ 4AtT12V&OKXYWR,R	 	 "IIk&7&7&9<Q&QWXY\a\e\e!q]
 
 4At+,V&OQRWR%R	 	) "'?=]+]cd!ehmhqhq,!i
 "
 99.1Q1QRS1TTUL+%%[ .ks   Ic                   \  ^  \ rS rSrSS0rSS0rSS/S/40rU 4S jr\\	         SS
\
R                  S	-  S\
R                  S	-  S\
R                  S	-  S\S	-  S\
R                  S	-  S\
R                  S	-  S\S	-  S\S	-  S\\
R                  -  S\\   S\4S jj5       5       rSrU =r$ )LagunaForCausalLMi  zlm_head.weightzmodel.embed_tokens.weightlm_headcolwise_gather_outputr7   logitsc                 J  > [         TU ]  U5        [        U5      U l        UR                  U l        [
        R                  " UR                  UR                  SS9U l        UR                  U l	        UR                  U l
        UR                  U l        U R                  5         g r   )r*   r+   rz  rc  r  r   r   r2   r  router_aux_loss_coefr   r   r  r   s     r4   r+   LagunaForCausalLM.__init__  s      (
 ++yy!3!3V5F5FUS$*$?$?!!--#)#=#=  	r6   Nr  r  r   rC  r  labelsrZ  output_router_logitslogits_to_keepr   r(   c
                 z   Ub  UOU R                   R                  nU R                  " SUUUUUUUS.U
D6nUR                  n[	        U	[
        5      (       a  [        U	* S5      OU	nU R                  USS2USS24   5      nSnUb  U R                  " XU R                  40 U
D6nSnU(       aY  [        UR                  U R                  U R                  U5      nUb*  XR                  UR                  UR                   5      -  -  n[#        UUUUR$                  UR&                  UR(                  UR                  S9$ )a  
labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
    Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
    config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored
    (masked), the loss is only computed for the tokens with labels in `[0, ..., config.vocab_size]`.
N)r  r  r   rC  r  rZ  r  )lossaux_lossr  rC  r7   re  r   r\  )rX   r  rc  r  r   r|   slicer  loss_functionr  r  r   r   r   r  r=   rr   r   rC  r7   re  )r1   r  r  r   rC  r  r  rZ  r  r  r   outputsr7   slice_indicesr  r  r  s                    r4   rD   LagunaForCausalLM.forward  sO   . %9$D $++JjJj 	
 +/** 	+
)%+'!5	+
 	+
  118B>SV8W8W~ot4]kmA}a,?@A%%fdooPPD/%%  ((	H !11HKK4LLL(#33!//))!//
 	
r6   )r  rc  r   r   r  r  )	NNNNNNNNr   )rK   rL   rM   rN   _tied_weights_keys_tp_plan_pp_planr+   r   r   r-   r_  rP   r	   r  r`  r|   r   r   r   rD   rQ   rR   rS   s   @r4   r  r    s5   *,GH23H_-z:;H
  .2.204(,26*.!%,0-.@
##d*@
 t+@
 &&-	@

 @
 ((4/@
   4'@
 $;@
 #Tk@
 ell*@
 +,@
 
#@
  @
r6   r  )r  rz  rb  )r!   )r   )Nr9   N)Kcollections.abcr   typingr   r-   torch.nn.functionalr   r   r    r   rj  activationsr   cache_utilsr	   r
   
generationr   integrationsr   r   r   masking_utilsr   r   modeling_flash_attention_utilsr   modeling_layersr   modeling_outputsr   r   modeling_rope_utilsr   r   modeling_utilsr   r   processing_utilsr   utilsr   r   utils.genericr   r   r   utils.output_capturingr   r    configuration_lagunar"   Moduler%   rU   r   r   r   r   r  r  rP   r|   r  rO   r,  r.  rO  rb  rz  rG   r  r  __all__r\  r6   r4   <module>r     s  * %      & ! . ) h h R B 9 Q K F & 5 [ [ E . Y'JBII J (J(M<BII M<`		  @ryy @> $#BII $# $#N299 .(#L	UU\\ 	U# 	U%,, 	U& %II%<<% 
% <<	%
 LL4'% % % '(%2 )*N)bii N) +N)b)3 )X $^O $^ $^N P
' P
 P
j #
*.	O&ell 33d:O&tO& LL4'	O&
 \\CO&d S
- S
 S
l Hr6   