Zj]UtSSKJr SSKJr SSKrSSKJr SSKJrJr SSK J r J r SSK J r SS KJrJrJr SS KJr SS KJrJr SS KJrJr SS KJrJr SSKJrJr SSK J!r! SSK"J#r#J$r$J%r% SSK&J'r'J(r( SSK)J*r* SSK+J,r, "SS\RZ5r.\"S5"SS\RZ55r/"SS\RZ5r0Sr1\"S5S9Sj5r2S\RfS \4S!\Rf4S"jr5S:S#\RZS$\RfS%\RfS&\RfS'\RfS-S(\6S)\6S*\!\#4S+jjr7\"\25"S,S-\RZ55r8"S.S/\5r9\$"S0S1\55r:\$"S2S3\:55r;\$"S4S5\:\ 55r<"S6S7\\:5r=/S8Qr>g);)Callable)OptionalN)nn)ACT2CLSACT2FN)Cache DynamicCache)GenerationMixin)use_kernel_forward_from_hubuse_kernel_func_from_hubuse_kernelized_func)create_causal_mask)GenericForTokenClassificationGradientCheckpointingLayer)BaseModelOutputWithPastCausalLMOutputWithPast)ROPE_INIT_FUNCTIONSdynamic_rope_update)ALL_ATTENTION_FUNCTIONSPreTrainedModel)Unpack)TransformersKwargsauto_docstringcan_return_tuple)maybe_autocastmerge_with_config_defaults)capture_outputs) ApertusConfigc.^\rSrSrU4SjrSrSrU=r$) ApertusMLP+c>[TU]5 XlURUlURUl[ R "URURSS9Ul[ R "URURSS9Ul[URUl URS:Xa[S"URS9Ul gg)NFbiasxieludtype)super__init__config hidden_sizeintermediate_sizerLinearup_proj down_projr hidden_actact_fnrr)selfr, __class__s }/root/GenerationalWealth/GenerationalWealth/venv/lib/python3.13/site-packages/transformers/models/apertus/modeling_apertus.pyr+ApertusMLP.__init__,s  !--!'!9!9yy!1!143I3IPUV 4#9#94;K;KRWXV../    '!'*>DK (c`URURURU555$N)r1r3r0)r5xs r7forwardApertusMLP.forward7s"~~dkk$,,q/:;;r9)r3r,r1r-r.r0)__name__ __module__ __qualname____firstlineno__r+r=__static_attributes__ __classcell__r6s@r7r"r"+s ?<[TU]5 [R"[R "U55UlX lg)z- ApertusRMSNorm is equivalent to T5LayerNorm N)r*r+r Parametertorchonesweightvariance_epsilon)r5r-rJr6s r7r+ApertusRMSNorm.__init__=s/ ll5::k#:; #r9 hidden_statescURnUR[R5nUR S5R SSS9nU[R "X0R-5-nURURU5-$)NT)keepdim) r)torNfloat32powmeanrsqrtrQrP)r5rS input_dtypevariances r7r=ApertusRMSNorm.forwardEsw#)) %((7  $$Q',,R,>% H?T?T4T(UU {{]--k:::r9c^[URR5SUR3$)Nz, eps=)tuplerPshaperQ)r5s r7 extra_reprApertusRMSNorm.extra_reprLs* ))*+6$2G2G1HIIr9)rQrP)gư>) r?r@rArBfloatr+rNTensorr=rcrCrDrEs@r7rHrH;sB$$$$$;U\\;ell;JJr9rHc ^\rSrSr%\R \S'SS\4U4Sjjjr\ SS\S-S\ SS\ S-S \ S \ 44S jj5r\R"5\S 55rS rU=r$)ApertusRotaryEmbeddingPinv_freqNr,c>[TU]5 URUlURUlXlUR R SUlURnURS:wa[URnU"UR U5uo@l URSUSS9 URSUR5SS9 g)N rope_typedefaultrjF) persistentoriginal_inv_freq) r*r+max_position_embeddingsmax_seq_len_cachedoriginal_max_seq_lenr,rope_parametersrlcompute_default_rope_parametersrattention_scalingregister_bufferclone)r5r,device rope_init_fnrjr6s r7r+ApertusRotaryEmbedding.__init__Ss "("@"@$*$B$B! 44[A!%!E!E >>Y &.t~~>L+7 V+L(( ZeD 0(..2BuUr9rxz torch.deviceseq_lenrKz torch.Tensorc URSn[USS5=(d URUR-nSnSU[R "SUS[R S9RU[RS9U- -- nXe4$) aH Computes the inverse frequencies according to the original RoPE implementation Args: config ([`~transformers.PreTrainedConfig`]): The model configuration. device (`torch.device`): The device to use for initialization of the inverse frequencies. seq_len (`int`, *optional*): The current sequence length. Unused for this type of RoPE. Returns: Tuple of (`torch.Tensor`, `float`), containing the inverse frequencies for the RoPE embeddings and the post-processing scaling factor applied to the computed cos/sin (unused in this type of RoPE). rope_thetahead_dimNg?rrUr()rxr)) rsgetattrr-num_attention_headsrNarangeint64rXre)r,rxr{basedimattention_factorrjs r7rt6ApertusRotaryEmbedding.compute_default_rope_parameterscs&%%l3fj$/c63E3EIcIc3c U\\!S!5;;?BB&X]XcXcBdgjj k ))r9cLURSSS2S4R5RURSSS5R UR 5nUSS2SSS24R5n[ UR R[5(a0UR RS:waUR ROSn[USS9 UR5UR5-RSS5n[R"Xf4SS 9nUR5UR-nUR5UR-n SSS5 WR UR S 9W R UR S 94$!,(df  N@=f) NrrVrmpscpuF) device_typeenabledrUrr()rjreexpandrbrXrx isinstancetypestrr transposerNcatcosrusinr)) r5r< position_idsinv_freq_expandedposition_ids_expandedrfreqsembrrs r7r=ApertusRotaryEmbedding.forwardsN!MM$4-8>>@GG HZHZ[\H]_acdehhijiqiqr ,QaZ 8 > > @'1!((--'E'E!((--[`J`ahhmmfk  U C&,,.1F1L1L1NNYYZ[]^_E))UN3C'')d444C'')d444C D vvAGGv$cff177f&;;; D Cs BF F#)rur,rqrrrlr;)NNN)r?r@rArBrNrf__annotations__r r+ staticmethodrintrarertno_gradrr=rCrDrEs@r7rhrhPsllV}VV '++/"*$*(*t* ~u$ % **: ]]_ <