
    fThE                        S SK r S SKJrJrJrJr  S SK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  SSKJr  SS	KJr  SS
KJrJr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%J&r&J'r'J(r(  SSK)J*r*  \'" 5       (       a  S SK+J,r,  SSK-J.r.  \(R^                  " \05      r1 " S S\Rd                  5      r3 " S S\Rd                  5      r4 " S S\Rd                  5      r5S\Rl                  S\7S\Rl                  4S jr8 S:S\Rd                  S\Rl                  S\Rl                  S \Rl                  S!\\Rl                     S"\9S#\94S$ jjr:S% r;S;S& jr< " S' S(\Rd                  5      r= " S) S*\5      r>\% " S+ S,\ 5      5       r?\% " S- S.\?5      5       r@ " S/ S0\\$5      rA\% " S1 S2\?\5      5       rB\%" S3S49 " S5 S6\?5      5       rC\% " S7 S8\?5      5       rD/ S9QrEg)<    N)CallableOptionalTupleUnion   )ACT2FN)CacheDynamicCache)GenerationMixin)AttentionMaskConverter)FlashAttentionKwargs)GradientCheckpointingLayer)BaseModelOutputWithPastCausalLMOutputWithPast SequenceClassifierOutputWithPastTokenClassifierOutput)ROPE_INIT_FUNCTIONSdynamic_rope_update)ALL_ATTENTION_FUNCTIONSPreTrainedModel)Unpack)
LossKwargsauto_docstringcan_return_tupleis_torch_flex_attn_availablelogging   )HeliumConfig)	BlockMask)make_flex_block_causal_maskc                   8   ^  \ rS rSrSU 4S jjrS rS rSrU =r$ )HeliumRMSNorm8   c                    > [         TU ]  5         [        R                  " [        R
                  " U5      5      U l        X l        g N)super__init__nn	Parametertorchonesweightvariance_epsilon)selfhidden_sizeeps	__class__s      b/var/www/auris/envauris/lib/python3.13/site-packages/transformers/models/helium/modeling_helium.pyr'   HeliumRMSNorm.__init__9   s-    ll5::k#:; #    c                 V   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                  R                  [        R                  5      U-  R                  U5      $ )N   T)keepdim)	dtypetor*   float32powmeanrsqrtr-   r,   )r.   hidden_statesinput_dtypevariances       r2   forwardHeliumRMSNorm.forward>   s    #))%((7 $$Q',,R,>%H?T?T4T(UUu}}-=AA+NNr4   c                 ^    [        U R                  R                  5       SU R                   3$ )Nz, eps=)tupler,   shaper-   r.   s    r2   
extra_reprHeliumRMSNorm.extra_reprE   s*    ))*+6$2G2G1HIIr4   )r-   r,   )gư>)	__name__
__module____qualname____firstlineno__r'   rB   rH   __static_attributes____classcell__r1   s   @r2   r"   r"   8   s    $
OJ Jr4   r"   c                   l   ^  \ rS rSrSS\4U 4S jjjr\R                  " 5       \S 5       5       r	Sr
U =r$ )HeliumRotaryEmbeddingI   configc                   > [         TU ]  5         [        US5      (       aH  UR                  b;  UR                  R	                  SUR                  R	                  S5      5      U l        OSU l        UR                  U l        UR                  U l        Xl	        [        U R
                     U l        U R                  U R                  U5      u  o0l        U R                  SUSS9  U R                  U l        g )Nrope_scaling	rope_typetypedefaultinv_freqF)
persistent)r&   r'   hasattrrV   getrW   max_position_embeddingsmax_seq_len_cachedoriginal_max_seq_lenrT   r   rope_init_fnattention_scalingregister_bufferrZ   original_inv_freq)r.   rT   devicerZ   r1   s       r2   r'   HeliumRotaryEmbedding.__init__J   s    6>**v/B/B/N#0044[&BUBUBYBYZ`BabDN&DN"("@"@$*$B$B!/?+/+<+<T[[&+Q((ZeD!%r4   c                 b   U R                   S S S 2S 4   R                  5       R                  UR                  S   SS5      R	                  UR
                  5      nUS S 2S S S 24   R                  5       n[        UR
                  R                  [        5      (       a0  UR
                  R                  S:w  a  UR
                  R                  OSn[        R                  " USS9   UR                  5       UR                  5       -  R                  SS5      n[        R                  " Xf4SS	9nUR                  5       U R                  -  nUR                  5       U R                  -  n	S S S 5        WR	                  UR                   S
9W	R	                  UR                   S
94$ ! , (       d  f       N@= f)Nr   r7   r   mpscpuF)device_typeenabledr6   dim)r9   )rZ   floatexpandrF   r:   re   
isinstancerX   strr*   autocast	transposecatcosrb   sinr9   )
r.   xposition_idsinv_freq_expandedposition_ids_expandedrj   freqsembru   rv   s
             r2   rB   HeliumRotaryEmbedding.forward[   sR    !MM$4-8>>@GGHZHZ[\H]_acdehhijiqiqr ,QaZ 8 > > @'1!((--'E'E!((--[`J`ahhmmfk^^UC&,,.1F1L1L1NNYYZ[]^_E))UN3C'')d444C'')d444C	 D vvAGGv$cff177f&;;; DCs   $BF  
F.)rb   rT   r_   rd   r`   ra   rW   r%   )rJ   rK   rL   rM   r   r'   r*   no_gradr   rB   rN   rO   rP   s   @r2   rR   rR   I   s6    /| / /" ]]_<  <r4   rR   c                   .   ^  \ rS rSrU 4S jrS rSrU =r$ )	HeliumMLPk   c                   > [         TU ]  5         Xl        UR                  U l        UR                  U l        [
        R                  " U R                  U R                  UR                  S9U l        [
        R                  " U R                  U R                  UR                  S9U l	        [
        R                  " U R                  U R                  UR                  S9U l
        [        UR                     U l        g )Nbias)r&   r'   rT   r/   intermediate_sizer(   Linearmlp_bias	gate_projup_proj	down_projr   
hidden_actact_fnr.   rT   r1   s     r2   r'   HeliumMLP.__init__l   s    !--!'!9!94#3#3T5K5KRXRaRabyy!1!143I3IPVP_P_`4#9#94;K;KRXRaRabV../r4   c                     U R                  U R                  U R                  U5      5      U R                  U5      -  5      nU$ r%   )r   r   r   r   )r.   rw   r   s      r2   rB   HeliumMLP.forwardv   s6    NN4;;t~~a/@#ADLLQRO#ST	r4   )r   rT   r   r   r/   r   r   )rJ   rK   rL   rM   r'   rB   rN   rO   rP   s   @r2   r   r   k   s    0 r4   r   r?   n_repreturnc                     U R                   u  p#pEUS:X  a  U $ 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      $ )z
This is the equivalent of torch.repeat_interleave(x, dim=1, repeats=n_rep). The hidden states go from (batch,
num_key_value_heads, seqlen, head_dim) to (batch, num_attention_heads, seqlen, head_dim)
r   N)rF   ro   reshape)r?   r   batchnum_key_value_headsslenhead_dims         r2   	repeat_kvr   {   s_    
 2?1D1D.Ez!!Qa"23::5W\dlmM  e(CTTTr4   modulequerykeyvalueattention_maskscalingdropoutc                 @   [        X R                  5      n[        X0R                  5      n	[        R                  " XR	                  SS5      5      U-  n
Ub"  US S 2S S 2S S 2S UR
                  S   24   nX-   n
[        R                  R                  U
S[        R                  S9R                  UR                  5      n
[        R                  R                  XU R                  S9n
[        R                  " X5      nUR	                  SS5      R                  5       nX4$ )Nr6   r   r7   )rm   r9   )ptrainingr   )r   num_key_value_groupsr*   matmulrs   rF   r(   
functionalsoftmaxr;   r:   r9   r   r   
contiguous)r   r   r   r   r   r   r   kwargs
key_statesvalue_statesattn_weightscausal_maskattn_outputs                r2   eager_attention_forwardr      s     3 ; ;<JU$?$?@L<<';';Aq'ABWLL!$Q1.D
0@0@0D.D%DE#1==((2U]](SVVW\WbWbcL==((6??([L,,|:K''1-88:K$$r4   c                 x    U SSSS24   nU SSSS24   n[         R                  " U* U4SS9R                  S5      $ )	z*Rotates half the hidden dims of the input..r   Nr6   r   r7   rl   r   )r*   stackflatten)rw   x1x2s      r2   rotate_halfr      sJ    	
319B	
319B;;Ryb)11"55r4   c                 4   UR                  U5      nUR                  U5      nUSSUR                  S   S-  24   R                  SSS9nUSSUR                  S   S-  24   R                  SSS9nX-  [        U 5      U-  -   nX-  [        U5      U-  -   nXg4$ )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.
.Nr7   r6   rl   )	unsqueezerF   repeat_interleaver   )qkru   rv   rx   unsqueeze_dimq_embedk_embeds           r2   apply_rotary_pos_embr      s    ( --
&C
--
&C c'SYYr]a'''
(
:
:1"
:
EC
c'SYYr]a'''
(
:
:1"
:
ECw;q>C/0Gw;q>C/0Gr4   c                   P  ^  \ rS rSrSrSS\S\\   4U 4S jjjr  SS\	R                  S\\	R                  \	R                  4   S\\	R                     S	\\   S
\\	R                     S\\   S\\	R                  \\	R                     \\\	R                        4   4S jjrSrU =r$ )HeliumAttention   z=Multi-headed attention from 'Attention Is All You Need' paperrT   	layer_idxc                 J  > [         TU ]  5         Xl        X l        [	        USUR
                  UR                  -  5      U l        UR                  UR                  -  U l	        S[        R                  " U R                  5      -  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
                  SS9U l        g )Nr   r   Tr   F)r&   r'   rT   r   getattrr/   num_attention_headsr   r   r   mathsqrtr   attention_dropout	is_causalr(   r   attention_biasq_projk_projv_projo_projr.   rT   r   r1   s      r2   r'   HeliumAttention.__init__   s?   "
F4F4F&JdJd4de$*$>$>&B\B\$\!499T]]33!'!9!9ii : :T]] JQWQfQf
 ii : :T]] JQWQfQf
 ii : :T]] JQWQfQf
 ii 2 2F4F4FUSr4   r?   position_embeddingsr   past_key_valuecache_positionr   r   c                    UR                   S S n/ UQSPU R                  P7nU R                  U5      R                  U5      R	                  SS5      n	U R                  U5      R                  U5      R	                  SS5      n
U R                  U5      R                  U5      R	                  SS5      nUu  p[        XX5      u  pUb$  XUS.nUR                  XU R                  U5      u  p[        nU R                  R                  S:w  ad  U R                  R                  S:X  a-  UR                  SS5      (       a  [        R                  S	5        O[         U R                  R                     nU" U U	U
UU4U R"                  (       d  S
OU R$                  U R&                  S.UD6u  nnUR(                  " / UQSP76 R+                  5       nU R-                  U5      nUU4$ )Nr7   r   r6   )rv   ru   r   eagersdpaoutput_attentionsFz`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.        )r   r   )rF   r   r   viewrs   r   r   r   updater   r   rT   _attn_implementationr]   loggerwarning_oncer   r   r   r   r   r   r   )r.   r?   r   r   r   r   r   input_shapehidden_shapequery_statesr   r   ru   rv   cache_kwargsattention_interfacer   r   s                     r2   rB   HeliumAttention.forward   s    $))#2.88b8$--8{{=166|DNNqRST[[/44\BLLQPQR
{{=166|DNNqRST&#7RU#[ %#&nUL'5'<'<ZW[WeWegs't$J(?;;++w6{{//69fjjI\^c>d>d##L
 '>dkk>^>^&_#$7	%
  $}}C$2H2HLL	%
 	%
!\ "));;;;FFHkk+.L((r4   )r   rT   r   r   r   r   r   r   r   r   r   r%   )NN)rJ   rK   rL   rM   __doc__r   r   intr'   r*   Tensorr   r	   
LongTensorr   r   rB   rN   rO   rP   s   @r2   r   r      s    GT| T T T4 +/590)||0) #5<<#=>0) !.	0)
 !0) !!1!120) -.0) 
u||Xell3XeELL>Q5RR	S0) 0)r4   r   c                     ^  \ rS rSrSS\S\\   4U 4S jjjr       SS\R                  S\\R                     S\\R                     S\\   S	\\   S
\\   S\\R                     S\\\R                  \R                  4      S\\   S\\R                   \\\R                   \R                   4      4   4S jjrSrU =r$ )HeliumDecoderLayeri  rT   r   c                   > [         TU ]  5         UR                  U l        [        XS9U l        [        U5      U l        [        UR                  UR                  S9U l	        [        UR                  UR                  S9U l
        g )N)rT   r   r0   )r&   r'   r/   r   	self_attnr   mlpr"   rms_norm_epsinput_layernormpost_attention_layernormr   s      r2   r'   HeliumDecoderLayer.__init__  sj    !--(LV$,V-?-?VEXEXY(5f6H6HfNaNa(b%r4   r?   r   rx   r   r   	use_cacher   r   r   r   c	                     Un
U R                  U5      nU R                  " SUU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4nU(       a  X4-  nU$ )N)r?   r   rx   r   r   r   r   r    )r   r   r   r   )r.   r?   r   rx   r   r   r   r   r   r   residualself_attn_weightsoutputss                r2   rB   HeliumDecoderLayer.forward  s     !,,]; ,0>> 
,
')%)/) 3
,
 
,
( !0 !55mD/ 0 "++Gr4   )r/   r   r   r   r   r%   )NNNFFNN)rJ   rK   rL   rM   r   r   r   r'   r*   r   r   r	   boolr   r   r   FloatTensorrB   rN   rO   rP   s   @r2   r   r     s   c| c c c 2637*.,1$)59KO'||' !.' u//0	'
 !' $D>' D>' !!1!12' &eELL%,,,F&GH' -.' 
u  (51B1BEDUDU1U+V"WW	X' 'r4   r   c                   N    \ rS rSr\rSrSrS/rS/r	Sr
SrSrSrSrSrSrS rSrg)	HeliumPreTrainedModeliI  modelTr   past_key_valuesc                    U R                   R                  n[        U[        R                  5      (       aW  UR
                  R                  R                  SUS9  UR                  b%  UR                  R                  R                  5         g g [        U[        R                  5      (       ad  UR
                  R                  R                  SUS9  UR                  b2  UR
                  R                  UR                     R                  5         g g [        U[        5      (       a&  UR
                  R                  R                  S5        g g )Nr   )r=   stdg      ?)rT   initializer_rangerp   r(   r   r,   datanormal_r   zero_	Embeddingpadding_idxr"   fill_)r.   r   r  s      r2   _init_weights#HeliumPreTrainedModel._init_weightsX  s    kk++fbii((MM&&CS&9{{&  &&( '--MM&&CS&9!!-""6#5#56<<> ...MM$$S) /r4   r   N)rJ   rK   rL   rM   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	  rN   r   r4   r2   r   r   I  sS    L&*#-.#4"5!N  $!"&*r4   r   c                     ^  \ rS rSrS\4U 4S jjrS rS r\\	         SS\
\R                     S\
\R                     S\
\R                     S	\
\   S
\
\R                     S\
\   S\
\   S\
\   S\
\R                     S\\   S\4S jj5       5       r SS\\R                  S4   S\R                  S\R                  S	\S\4
S jjr\S\R                  S\S\S\R2                  S\R                  S\4S j5       rSrU =r$ )HeliumModelif  rT   c           	        > [         TU ]  U5        UR                  U l        UR                  U l        [
        R                  " UR                  UR                  U R                  5      U l        [
        R                  " [        UR                  5       Vs/ s H  n[        X5      PM     sn5      U l        [        UR                  UR                  S9U l        [#        U5      U l        SU l        U R)                  5         g s  snf )Nr   F)r&   r'   pad_token_idr  
vocab_sizer(   r  r/   embed_tokens
ModuleListrangenum_hidden_layersr   layersr"   r   normrR   
rotary_embgradient_checkpointing	post_initr   s      r2   r'   HeliumModel.__init__h  s     !.. ++LL):):F<N<NPTP`P`ammDI&JbJbDcdDcy2Dcd
 "&"4"4&:M:MN	/7&+# 	 es   Dc                     U R                   $ r%   r  rG   s    r2   get_input_embeddings HeliumModel.get_input_embeddingsx  s       r4   c                     Xl         g r%   r'  r.   r   s     r2   set_input_embeddings HeliumModel.set_input_embeddings{  s    !r4   	input_idsr   rx   r   inputs_embedsr   r   output_hidden_statesr   flash_attn_kwargsr   c
                 J   Ub  UOU R                   R                  nUb  UOU R                   R                  nUb  UOU R                   R                  nUS L US L-  (       a  [	        S5      eU R
                  (       a/  U R                  (       a  U(       a  [        R                  S5        Sn[        U[        S 5      [        45      (       d  [	        S5      eUc  U R                  U5      nU(       a  Uc
  [        5       nU	cD  Ub  UR                  5       OSn[        R                   " XUR"                  S   -   UR$                  S9n	Uc  U	R'                  S5      nU R)                  X%XU5      nUnU R+                  X5      nU(       a  SOS nU(       a  SOS nU R,                  S U R                   R.                    H7  nU(       a  X4-  nU" U4UUUUUU	US	.U
D6nUS   nU(       d  M.  UUS   4-  nM9     U R1                  U5      nU(       a  X4-  n[3        UU(       a  UOS UUS
9$ )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   re   r   )r   rx   r   r   r   r   r   )last_hidden_stater   r?   
attentions)rT   r   r0  r   
ValueErrorr#  r   r   r   rp   rX   r	   r  r
   get_seq_lengthr*   arangerF   re   r   _update_causal_maskr"  r   r  r!  r   )r.   r.  r   rx   r   r/  r   r   r0  r   r1  past_seen_tokensr   r?   r   all_hidden_statesall_self_attnsdecoder_layerlayer_outputss                      r2   rB   HeliumModel.forward~  sI    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"\\ ]5H5H5K"KTaThThN )33A6L..>L]
 & #oomJ #7BD0d![[)H4;;+H+HIM#!%55!)
*)."3#-$7
 $
M *!,M  =#3"55' J* 		-0  !11&+/8Od+%	
 	
r4   r   input_tensorc           	         U R                   R                  S:X  a  Ub  US:H  R                  5       (       a  U$ g U R                   R                  S:X  a,  [        U[        R
                  5      (       a  [        U5      nU$ Ub  UR                  5       OSnUb  UR                  OSnU R                   R                  S:X  a5  U(       d.  U(       d'  [        R                  " UUUU R                  S9(       a  g UR                  nUR                  S   n	U(       a  UR                  5       n
O5[        U[        R
                  5      (       a  UR                  S	   OXi-   S-   n
U R                  UU	U
UUUR                  S   S
9nU R                   R                  S:X  aZ  UbW  UR                   R"                  S;   a=  U(       d6  [        R$                  " U5      R&                  n[        R(                  " X5      nU$ )Nflash_attention_2r   flex_attentionr   Fr   )r/  past_key_values_lengthis_trainingr   r7   )sequence_lengthtarget_lengthr9   r   
batch_size)cudaxpunpu)rT   r   anyrp   r*   r   r    r7  is_compileabler   _ignore_causal_mask_sdpar   r9   rF   get_max_cache_shape5_prepare_4d_causal_attention_mask_with_cache_positionre   rX   finfomin_unmask_unattended)r.   r   r@  r   r   r   r:  using_compilable_cacher9   rF  rG  r   	min_dtypes                r2   r9  HeliumModel._update_causal_mask  s    ;;++/BB)~/D.I.I.K.K%%;;++/??.%,,77!<^!L!!
 @O?Z?99;`aCRC^!?!?di ;;++v5>T]n%>>*'7 MM	 ""&,,Q/!+??AM nell;; $$R(%7!;  PP+')#))!, Q 
 KK,,6*%%**.DD%
 E*..I0CCK[Kr4   rF  rG  r9   rH  c                    U b  U R                  5       S:X  a  U nU$ [        R                  " U5      R                  n[        R                  " X4XUR
                  S9nUS:w  a  [        R                  " USS9nU[        R                  " X$R
                  S9UR                  SS5      :  -  nUSSSS2SS24   R                  USSS5      nU b  UR                  5       nU R                  S   n	USS2SS2SS2SU	24   U SS2SSSS24   R                  UR
                  5      -   n
U
S:H  n
USS2SS2SS2SU	24   R                  X5      USS2SS2SS2SU	24'   U$ )	a  
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_valuer9   re   r   )diagonalr3  r7   r   )rm   r*   rQ  rR  fullre   triur8  r   ro   clonerF   r:   masked_fill)r   rF  rG  r9   r   rH  r   r   rU  mask_lengthpadding_masks              r2   rP  AHeliumModel._prepare_4d_causal_attention_mask_with_cache_position"  s}   < %.*<*<*>!*C(K* ' E*..I** 0Y\j\q\qK !##jjqA5<<>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 r4   )r  r#  r   r!  r  r"  r  	NNNNNNNNN)F)rJ   rK   rL   rM   r   r'   r(  r,  r   r   r   r*   r   r   r	   r   r   r   r   r   rB   r   r9  staticmethodr   r9   rP  rN   rO   rP   s   @r2   r  r  f  s   |  !"  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r4   r  c                       \ rS rSrSrg)KwargsForCausalLMiZ  r   N)rJ   rK   rL   rM   rN   r   r4   r2   re  re  Z  s    3r4   re  c                     ^  \ rS rSrS/rSS0rSS/S/40rS\4U 4S jjrS	 r	S
 r
S rS rS rS r\\           SS\\R&                     S\\R(                     S\\R&                     S\\   S\\R,                     S\\R&                     S\\   S\\   S\\   S\\R&                     S\\\R(                  4   S\\   S\4S jj5       5       rSrU =r$ )HeliumForCausalLMi]  zlm_head.weightlm_headcolwise_repr?   logitsrT   c                    > [         TU ]  U5        [        U5      U l        UR                  U l        [
        R                  " UR                  UR                  SS9U l        U R                  5         g NFr   )
r&   r'   r  r   r  r(   r   r/   rh  r$  r   s     r2   r'   HeliumForCausalLM.__init__c  sU      (
 ++yy!3!3V5F5FUS 	r4   c                 .    U R                   R                  $ r%   r   r  rG   s    r2   r(  &HeliumForCausalLM.get_input_embeddingsl      zz&&&r4   c                 $    XR                   l        g r%   ro  r+  s     r2   r,  &HeliumForCausalLM.set_input_embeddingso      "'

r4   c                     U R                   $ r%   rh  rG   s    r2   get_output_embeddings'HeliumForCausalLM.get_output_embeddingsr  s    ||r4   c                     Xl         g r%   rv  )r.   new_embeddingss     r2   set_output_embeddings'HeliumForCausalLM.set_output_embeddingsu  s    %r4   c                     Xl         g r%   r   )r.   decoders     r2   set_decoderHeliumForCausalLM.set_decoderx  s    
r4   c                     U R                   $ r%   r~  rG   s    r2   get_decoderHeliumForCausalLM.get_decoder{  s    zzr4   r.  r   rx   r   r/  labelsr   r   r0  r   logits_to_keepr   r   c                    Ub  UOU R                   R                  nU	b  U	OU R                   R                  n	U R                  " SUU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                  " SUX`R                   R                  S.UD6n[        UU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]`.

Example:

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

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

>>> prompt = "What is your favorite condiment?"
>>> inputs = tokenizer(prompt, return_tensors="pt")

>>> # Generate
>>> generate_ids = model.generate(inputs.input_ids, max_length=30)
>>> tokenizer.batch_decode(generate_ids, skip_special_tokens=True, clean_up_tokenization_spaces=False)[0]
"What is your favorite condiment?"
```N)	r.  r   rx   r   r/  r   r   r0  r   )rj  r  r  lossrj  r   r?   r5  r   )rT   r   r0  r   r4  rp   r   slicerh  loss_functionr  r   r   r?   r5  )r.   r.  r   rx   r   r/  r  r   r   r0  r   r  r   r   r?   slice_indicesrj  r  s                     r2   rB   HeliumForCausalLM.forward~  s   N 2C1N-TXT_T_TqTq$8$D $++JjJj 	
 ,0:: ,
)%+'/!5),
 ,
  118B>SV8W8W~ot4]kmA}a,?@A%%pVF{{OeOepiopD%#33!//))
 	
r4   )rh  r   r  )NNNNNNNNNNr   ) rJ   rK   rL   rM   _tied_weights_keys_tp_plan_pp_planr   r'   r(  r,  rw  r{  r  r  r   r   r   r*   r   r   r	   r   r   r   r   r   re  r   rB   rN   rO   rP   s   @r2   rg  rg  ]  s   *+=)H_-z:;H| '(&  151537+/59-1$(,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
 
 G
  G
r4   rg  a  
    The Helium Model transformer with a sequence classification head on top (linear layer).

    [`HeliumForSequenceClassification`] 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                   2  ^  \ rS rSrS\4U 4S jjrS rS r\\	         SS\
\R                     S\
\R                     S\
\R                     S	\
\   S
\
\R                     S\
\R                     S\
\   S\
\   S\
\   S\4S jj5       5       rSrU =r$ )HeliumForSequenceClassificationi  rT   c                    > [         TU ]  U5        UR                  U l        [        U5      U l        [
        R                  " UR                  U R                  SS9U l        U R                  5         g rl  )
r&   r'   
num_labelsr  r   r(   r   r/   scorer$  r   s     r2   r'   (HeliumForSequenceClassification.__init__  sS      ++ (
YYv114??O
 	r4   c                 .    U R                   R                  $ r%   ro  rG   s    r2   r(  4HeliumForSequenceClassification.get_input_embeddings  rq  r4   c                 $    XR                   l        g r%   ro  r+  s     r2   r,  4HeliumForSequenceClassification.set_input_embeddings  rt  r4   r.  r   rx   r   r/  r  r   r   r0  r   c
                    U R                  UUUUUUUU	S9n
U
R                  nU R                  U5      nUb  UR                  S   nOUR                  S   nU R                  R
                  c  US:w  a  [        S5      eU R                  R
                  c  SnOUb  XR                  R
                  :g  R                  UR                  [        R                  5      n[        R                  " UR                  S   UR                  [        R                  S9nUU-  R                  S5      nO.Sn[        R                  U R                  R                    S35        U[        R                  " XR                  S	9U4   nSnUb  U R#                  XUU R                  S
9n[%        UUU
R&                  U
R(                  U
R*                  S9$ )e  
labels (`torch.LongTensor` of shape `(batch_size,)`, *optional*):
    Labels for computing the sequence classification/regression loss. Indices should be in `[0, ...,
    config.num_labels - 1]`. If `config.num_labels == 1` a regression loss is computed (Mean-Square loss), If
    `config.num_labels > 1` a classification loss is computed (Cross-Entropy).
r   rx   r   r/  r   r   r0  Nr   r   z=Cannot handle batch sizes > 1 if no padding token is defined.r7   )re   r9   z will not detect padding tokens in `inputs_embeds`. Results may be unexpected if using padding tokens in conjunction with `inputs_embeds.`r3  )rj  r  pooled_logitsrT   r  )r   r4  r  rF   rT   r  r6  r:   re   r*   int32r8  argmaxr   r   r1   rJ   r  r   r   r?   r5  )r.   r.  r   rx   r   r/  r  r   r   r0  transformer_outputsr?   rj  rH  last_non_pad_tokennon_pad_masktoken_indicesr  r  s                      r2   rB   'HeliumForSequenceClassification.forward  s   * 8<zz)%+'/!5 8B 	8
 ,==M* "+J&,,Q/J;;##+
a\]];;##+!#"%)A)AAEEfmmUZU`U`aL!LL)<V]]Z_ZeZefM"/,">!F!Fr!J!#>>**+ ,Z Z
 u||J}}MOaab%%VR_hlhshs%tD/ /??-;;*55
 	
r4   )r   r  r  rb  )rJ   rK   rL   rM   r   r'   r(  r,  r   r   r   r*   r   r   r	   r   r   r   rB   rN   rO   rP   s   @r2   r  r    s   | '(  151537+/59-1$(,0/3A
E,,-A
 !.A
 u//0	A

 "%A
   1 12A
 ))*A
 D>A
 $D>A
 'tnA
 
*A
  A
r4   r  c                   2  ^  \ rS rSrS\4U 4S jjrS rS r\\	         SS\
\R                     S\
\R                     S\
\R                     S	\
\   S
\
\R                     S\
\R                     S\
\   S\
\   S\
\   S\4S jj5       5       rSrU =r$ )HeliumForTokenClassificationi.  rT   c                   > [         TU ]  U5        UR                  U l        [        U5      U l        [        USS 5      b  UR                  nO[        USS 5      b  UR                  nOSn[        R                  " U5      U l
        [        R                  " UR                  UR                  5      U l        U R                  5         g )Nclassifier_dropouthidden_dropoutg?)r&   r'   r  r  r   r   r  r  r(   Dropoutr   r   r/   r  r$  )r.   rT   r  r1   s      r2   r'   %HeliumForTokenClassification.__init__0  s      ++ (
6/6B!'!:!:V-t4@!'!6!6!$zz"45YYv1163D3DE
 	r4   c                 .    U R                   R                  $ r%   ro  rG   s    r2   r(  1HeliumForTokenClassification.get_input_embeddings@  rq  r4   c                 $    XR                   l        g r%   ro  r+  s     r2   r,  1HeliumForTokenClassification.set_input_embeddingsC  rt  r4   r.  r   rx   r   r/  r  r   r   r0  r   c
                    U R                  UUUUUUUU	S9n
U
R                  nU R                  U5      nU R                  U5      nSnUb  U R	                  XU R
                  5      n[        UUU
R                  U
R                  S9$ )r  r  N)r  rj  r?   r5  )	r   r4  r   r  r  rT   r   r?   r5  )r.   r.  r   rx   r   r/  r  r   r   r0  r   sequence_outputrj  r  s                 r2   rB   $HeliumForTokenClassification.forwardF  s    * ,0::)%+'/!5 ,6 	,
 "33,,7O,%%fdkkBD$!//))	
 	
r4   )r   r   r  r  rb  )rJ   rK   rL   rM   r   r'   r(  r,  r   r   r   r*   r   r   r	   r   r   r   rB   rN   rO   rP   s   @r2   r  r  .  s    |  '(  151537+/59-1$(,0/3*
E,,-*
 !.*
 u//0	*

 "%*
   1 12*
 ))**
 D>*
 $D>*
 'tn*
 
*
  *
r4   r  )r   r  rg  r  r  )r   )Nr   )Fr   typingr   r   r   r   r*   torch.nnr(   activationsr   cache_utilsr	   r
   
generationr   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_heliumr   !torch.nn.attention.flex_attentionr   integrations.flex_attentionr    
get_loggerrJ   r   Moduler"   rR   r   r   r   r   rn   r   r   r   r   r   r   r  re  rg  r  r  __all__r   r4   r2   <module>r     s  ,  3 3   ! . ) > B 9  L F & h h .  !!;J 
		H	%JBII J"<BII <D		  	UU\\ 	U# 	U%,, 	U& %II%<<% 
% <<	%
 U\\*% % %46BH)bii H)V23 2j *O * *8 p' p pf ?,j > i
- i
 i
X S
&; S
S
l C
#8 C
 C
Lr4   