9iSSKrSSKrSSKrSSKrSSKrSSKrSSKrSSKrSSKrSSK J r SSK J r SSK Jr SSKJrJr SSKJrJrJrJrJr SSKrSSKJr SSKJrJr SS KJr SS KJ r SS K!J"r" SS K#J$r$ \(aSS K%J&r& SSK'J(r( SSK)J*r* SSK+J,r, SSK+J-r-J.r.J/r/ SSK0J1r1J2r2 SSK3J4r4J5r5J6r6J7r7J8r8J9r9J:r:J;r;Jr>J?r?J@r@ \.RS\.RS\.RS\.RS\.RS\.RS0rGSqHSSjrIS\RR4SjrLSS jrMSS"jrNS\RRS\O\P\44S#jrQS$rRS%\O\P\4S&\RRS!SSS4S'jrSS&\RR4S(jrTSS)\O\P\P4S*\O\P\U4S+\V\PS,\PS-\PS.\W4 S/jjrXS0\:S\P4S1jrYSS2\\U\R4S3\US4\:S5\\USS4 S6jjr[S7\\\RRSS4S8jr^SSSSSS9.S:\_\S;\\7S<\\@S=\\PS>\\>S?\\8S@\WSS4SAjjjr`SBSCSD\RRS\W4SEjraSBSCSD\RRS\\RR 4SFjrcSBSCSD\RRS\W4SGjrdSBSCSD\RRS\\R4SHjrfSBSCSD\RRS\W4SIjrgSBSCSD\RRS\\R4SJjrhS&\RRSK\\RR/\\RR\W44S\RR4SLjriSM\\\RRS\\\RR4SNjrj\SO5rkS&\RRSS4SPjrlS&\RR4SQjrmSSM\\\RRS\\RR4SRjjrnSM\\\RRS\U4SSjroSM\\\RRS\\\RR4STjrpSU\RRSV\RRSS4SWjrqS&\RRS\RRSS4SXjrsSY\RRS\\RR4SZjrtS\RRS\R4S[jrvS\rwS]rxS&\RRSS4S^jryS&\RRS_SS\RRS`\O\P\4SS4 SajrzSb\OSc\WS\O4Sdjr{S&\RRS\\RRR84Sejr~\S\RR4Sfj5rSgShS\W4SijrSgShS\W4SjjrSgShS\W4SkjrSgShS\W4SljrSSmjrSnrSSojr\GR"SpSq9S\VSh4Srj5rSs\GRGRS\VSh4StjrS\VSh4SujrSSvjr\Sw5rSx\S\RRS\O\P\\R\RR 444SyjrS:\_\RRSS4SzjrS:\_\RRSS4S{jrS|rS}r"S~S\RR5rSrg)N) defaultdict)Iterable)contextmanager)ismethod Parameter)AnyCallableOptional TYPE_CHECKINGUnion)detect_fake_mode) FakeTensorFakeTensorMode)FunctionalTensor)#first_call_function_nn_module_stack)PreDispatchTorchFunctionMode)insert_deferred_runtime_assertsConstantAttrMap) OperatorBase)ExportedProgram)ExportGraphSignature)CustomObjArgument InputKind OutputKind)_deregister_pytree_flatten_specregister_pytree_flatten_spec) _deregister_pytree_node_register_pytree_nodeContext FlattenFuncFromDumpableContextFn GetAttrKeyKeyPathkeystr MappingKey SequenceKeyToDumpableContextFntree_flatten_with_path UnflattenFuncp_b_c_obj_tokenFreturncSSKJn U"5nURVs1sHBnUR[R :XdM#UR (aM6URiMD nnUR5HnupxXv;aM Un URS5GtpU Hn [X5n M U RRU S5 [XU5 URX5 Mp U$s snf)Nrr.)(torch._export.passes.lift_constants_passr input_specskindrBUFFER persistenttargetitemssplitgetattr_bufferspopsetattradd) graph_signature constantsmodrconstant_attrsspecnon_persistent_buffersnamevalue_modatomsattratoms S/var/www/html/land-doc-ocr/venv/lib/python3.13/site-packages/torch/_export/utils.py_collect_and_set_constant_attrsrNBsI$&N$///D 99 (( ( 15  / !(   ) zz# D4&D $%E"5') #s"C+C+C+rCcrSSKJnJn [5nURR HnUR S:XdM[RRRXR5n[U[R5(dMiURU;dM{URU;dM[RRRXR5 U"XpURURS5 UR!UR5 M UR#5 U$)Nr) _assign_attr _AttrKindget_attrF)torch.export.unflattenrPrQsetgraphnodesoptorchfx graph_module _get_attrr9 isinstanceTensor _del_attrr7r@ recompile)rC state_dictrFrPrQtemp_registered_constantsnoder9s rM_register_constants_as_buffersrc`s? #  77j XX**44S++FF&%,,//KKz1KK'==HH))33CE dkk9;K;KUS-11$++> MMO $$rc@URH1nURU;dM[RUlSUlM3 UR HLnUR[R:XdM#URU;dM5[SURS35e U$)Nz Constant z< is mutated in the forward method. Pls register it as buffer) r5r9rCONSTANT_TENSORr6r8 output_specsrBUFFER_MUTATION RuntimeError)sigrarEs rM7_override_graph_signature_for_temp_registered_constantsrkzs ;;3 3!11DI"DO    II33 3 88DKK=(de  ! Jrdnew_sigcZURVs1sHBnUR[R:XdM#UR(aM6UR iMD nnURH<nUR[R:XdM#UR U;dM5SUlM> U$s snf)NF)r5r6rr7r8r9)old_sigrlrErFs rM/_overwrite_signature_for_non_persistent_buffersros '''D 99 (( ( 15  ' ## 99 (( (T[[._getattrsD"- D&A= =q  rd call_moduleTF)recurseremove_duplicater3rR call_functioncustom)rXrY GraphModulestrrUrVr9metarWr\nnModulenamed_parameters named_buffers_opsHigherOrderOperator _input_nodesproxy_COPY_META_FIELDS) rCparams_buffers_to_node_metarxrbr9r submodulerG_argentrys rM_collect_param_buffer_metadatars#%!,,!! yy 77m # -I)UXX__55(99 5 : GDHL t0CD  )66 5 7 GDHL t0CD  77j  -Ii)=)=>>6:F3 77o %j KK77/ / ((66Z'!&!A!A H,$ =MQ[7 CEJ "B)7 H '&rdcT[RR5(dg[RR 5nUVs/sHn[ U[ 5(dMUPM nn[U5S::dS[U535e[U5S:XagUSnU$s snf)Nz6Expected only one PreDispatchTorchFunctionMode, found r)rX_C_is_torch_function_mode_enabled overrides _get_current_function_mode_stackr\rlen)torch_function_mode_stackmodepre_dispatch_tf_modess rM+_maybe_find_pre_dispatch_tf_mode_for_exportrs 88 3 3 5 5 % P P R.-D d8 9 - $ % * @EZA[@\] * !Q&  #D Ks B%$B%rgmcdUR5H'nURSS5 URSS5 M) URRHnURS:XdMUR UR ;aGUR UR nXP;a)XR5HupgXtRU'M UR UR;dMURUR nX;dMXR5HupgXtRU'M M g)ze Given that we collected param'buffer metadata before, we put them back in newly traced graph module nn_module_stackN stack_trace placeholder) valuesr>rUrVrWr9inputs_to_parametersr:rinputs_to_buffers) rrrlmetadatarb param_namekv buffer_names rM)_populate_param_buffer_metadata_to_new_gmrs0668 &- ]D)9 77m #{{g:::$99$++F < ; G M M O'( ! !P{{g777%77 D = ; H N N P'( ! !QrdcfURRVs/sH1nURRSS5cM"URSPM3 nn[ U5nUb UR $UH:n[ U[R5(dM$URR s $ gs snf)Nval) rUrVrget_detect_fake_mode_from_gm shape_envr\rXSymIntrb)rrbvals fake_moders rM_get_shape_env_from_gmr sHHNN "D 99== %  %"   *"-I"""  a & &66## # s !B.B.name_mapfind_available used_names orig_namerGis_placeholderc[R"SU5nUnU(a*U(d#URS5URS5pUnUn X;a USXS-3n [R"SU 5nU(a?URS5URS5p[U 5X:a [U 5X'XU'UR U 5 X$)aI Renames nodes to avoid name collisions, with suffixing. name_map: map from original name to new name find_available: map prefix to available suffix used_names: cache of used names orig_name: mapping key name: candidate name (potentially suffixed, e.g. mul_2) is_placeholder: if the node is a placeholder, avoid detecting suffix (.*)_(\d+)rr)rematchgroupintr@) rrrrrGrrkeyrtnnew_names rM_rename_without_collisionsrs" HH]D )E C ^KKNEKKNHU!N/!345 HH]H -E KKNEKKN q6N* *%(VN ""YNN8  rdkey_pathcUSn[U[5(deURS:XaS[USS53$USn[U[[ 45(de[ U5SSnU[USS53$)z}For a given index into the flat_args, return a human readable string describing how to access it, e.g. "*args["foo"][0].bar" rz*argsrNr)r\r'idxr%r#r&r)rargs_kwargs_key_path kwarg_keyrGs rM get_keystrrAs$A; *K 8 88 81$vhqrl+,--QK )j*%=>>>>9~a#x|,-..rdsymintrkeypathic ~SSKJn [U[R5(a%UR R R(a[X5(agSSKnSSK J n SSK J n [U[R5(Ga2[UR R R5S:XGa[[!UR R R55n X;aVUR R R#U5n X:wa+[%U5n Ub U SUS3- n ['SU S U S U35eO[UR R UR(5(a[+U5X:'OU "UR-UR R U5U 5n U cC[%U5n Ub U SUS3- n ['S U S US UR R SU S3 5e[+U S5X:'UR R U;aU"X R R 5upUS:a0X:a+[%U5n Ub U SUS3- n ['SU SUS U35eU[.R0:a1X:a+[%U5n Ub U SUS3- n ['SU SUS U35eggg[U[R5(a&UR R R(dgU[+U5:wa,[%U5n Ub U SUS3- n ['SU S US US35eg)Nr) _IntWrapper)_convert_range_to_int) try_solverz.shape[]Expected input at  to be equal to , but got zExpected input z = z to be of the form z, where z is an integerrz to be >= z to be <= zt. If you meant for this dimension to be dynamic, please re-export and specify dynamic_shapes (e.g. with Dim.DYNAMIC))torch.export.dynamic_shapesrr\rXrrbexpr is_numbersympy@torch._export.passes.add_runtime_assertions_for_constraints_passrtorch.utils._sympy.solverr free_symbolsnextitersubsrriSymbolrEqmathinf)rrrange_constraintsunification_maprrrrrrsymbol existing_dimpathsolutionmin_valmax_vals rM _check_symintrSs<8 3 %% '' c ' ' 3&%,,''C 0@0@0M0M,NRS,Sd6;;++889:  $!;;++00AL"!'*=gaSN*D"(.>|nJWZV[\ #&++**ELL99 +.c('$UXXfkk.>.>%DfM#%g.D}'!A.&)$s3%8''-{{'7'7&8X /2(1+.>O+ ;;  0 04!++"2"23  G{=%g.D}'!A.&,TF*WIZPSuU!=%g.D}'!A.&,TF*WIZPSuU !" 1* FELL ) )&++2B2B2L2L F '" = gaSN "D &6vhjN& &  rdinput_placeholdersc SSKn[U5[U5:wa$[S[U5S[U5S35e0n[X5GHuupVnURR S5n[ U[5(a[ U[R5(d#[S[U5S[U535e[UR5[UR5:wa2[S[U5S URSURS35e[[URUR55Hun up[XX$XY5 M GM[ U[[ ["45(a=[U5[U5:wdXh:wa[S[U5S US U35eGM{[ U[R$5(dGM[XX$US5 GM g) Nrz&Unexpected number of inputs (expected z, got )rrz to be a tensor, but got z,Unexpected number of dimensions in input at z.shape (expected rr)rrriziprrr\rrXr]rtypeshape enumeraterrfloatrr) rflat_args_with_pathrrrrrrbnode_valjarg_dimnode_dims rM"_check_input_constraints_for_graphrs 3'9#:: /01>WX\]`XaWbc8>>"c#))n4"B:hCWBXY!!) 0syykD +4C 8>>4R*S&&G'88+T 3s"3 4 4CyDN*co"(H)=(>>NxjXbcfbgh/>%,, / / 08T 3"Nrd)serialized_type_nameto_dumpable_contextfrom_dumpable_contextreturn_none_fieldscls flatten_fn unflatten_fnrrrrc ^^^[R"T5(d ST35eS[S[[[[ 44U4SjjnS[ [S[ S[4U4SjjnS[S[[[[ 44U4Sjjn TbTOUmUbUOUnUSLUSL- (a[S TS 35e[TTUUU UUS 9 g) Nz7Only dataclasses can be registered with this function: objr1c>/n/n/n[R"U5HbnUR[XR5peUcT(a$UR U5 UR U5 MQUR U5 Md XU/4$N) dataclassesfieldsrGr<append)r flattened flat_names none_namesfrGrrs rMdefault_flatten_fn=register_dataclass_as_pytree_node..default_flatten_fns}   ##C(AVV 4#"4  %!!$'!!$' )z222rdrcontextc j>Uup#T"S0[[X 55D[RU5D6$)N)dictrfromkeys)rrr r rs rMdefault_unflatten_fn?register_dataclass_as_pytree_node..default_unflatten_fns1!( PT#j12PdmmJ6OPPrdc>T"U5unup#[X!5VVs/sHupE[U5U4PM snnU4$s snnfrrr&)rrr  _none_namesrrrs rMdefault_flatten_fn_with_keysGregister_dataclass_as_pytree_node..default_flatten_fn_with_keyssC/9#, ,J/2:/IJ/ItqA"/IJJVVJs<z7Both to_dumpable_context and from_dumpable_context for z must be None or registered.rflatten_with_keys_fnrr) r is_dataclassrtuplelistr r ValueErrorr) rrrrrrrr rrs `` ` rM!register_dataclass_as_pytree_noder s  # #C ( ( A#G ( 3 3d3i.@(A 3QXc]QWQQW#W%S 78J2KW *5;MJ#/#;TU$rr)rb split_maps rM"sequential_split..ys Yt_rdT)keep_original_orderkeep_original_node_name)torch.fx.passes.split_moduler4rUrV_codegenr_)rr2r4split_idrbnew_gmr6s @rMsequential_splitr?ds:IH $   MH" $   $ $ FHH--FLL  MrdrVcNUVs/sHo!"U5(dMUPM sn$s snf)z:Returns the nodes that match the node_call_back as a list.rrVr2rbs rM nodes_filterrBs!" ;UTnT&:DU ;; ;s""c#8# [nSqSv Uqg!Uqf=f7fNT)_DISABLE_ATEN_TO_ASSERTION_PASS)orig_vals rM$_disable_aten_to_metadata_assertionsrGs&/H&*#3 *2'('s c SSKJnJn [(ag[R R RR[R R RR[R R RR/nURRGHnURU;dMURR[R R RR :Xa,UR"SURR"S:XaMUR"SR$R'S5=ncMURR)U5 U"U[*R,"UUR$R'S5UR$R'S5S.S95 URR/[R R RR UR"S4URURUR0S.S 9 SSS5 SSS5 GM g!,(df  N=f!,(df  GM=f) Nr_node_metadata_hook_set_node_metadata_hookrrr)rrr)dtypedevicelayoutargskwargs)(torch._export.passes._node_metadata_hookrJrKrErXopsatentorNrM dtype_layoutrUrVr9prev_assert_tensor_metadatadefaultrQrrinserting_before functoolspartialr}rO)rrJrKaten_to_variantsrb tensor_vals rM$_insert_aten_to_metadata_assert_passr`s '&      &&  ;;* *   EIINN$J$J$R$RRIIaLDIINN1$55"iil//33E:: GHH--d3+!))//3yy}}]/K3799==AR3S& HH** >>FF"iil_%/%5%5&0&7&7&0&7&7 + 43  43s&AI5A8I$I5$ I2 .I55 J c SSKJnJn SSKJn [ R RR(dhSnU"U[R"USU0S95 [U5nU(a"[UUS[UR53SS 9 SSS5 [U5 UR!5 U"U5UlX4$!,(df  N9=f) NrrI)_graph_output_nameszUFile "torch/fx/passes/runtime_assert.py", line 24, in insert_deferred_runtime_assertsrrLzexported program: T)export)rSrJrK4torch._functorch._aot_autograd.input_output_analysisrbrX_dynamoconfigdo_not_emit_runtime_assertsr\r]rrrrUr`r_ user_outputs)rrArJrKrbrrs rMapply_runtime_assertion_passrisY ==   ; ; 1 %    #}k.J  /r2I/()LRXX)V(WX  -R0LLN#6r#:O +  s 5C CcV[X(aUOS5n[U5S:aUS$g)z Returns the first node that matches the node_call_back. If no node matches, returns None. When node_call_back is None, returns the first node in the node list. cgrDrrbs rMr7nodes_first..sQUrdrN)rBr)rVr2rets rM nodes_firstros, unDU VC 3x!|1v rdc*[[X55$)z:Returns the number of nodes that match the node_call_back.)rrB)rVr2s rM nodes_countrqs |E2 33rdc(UH nU"U5 M U$)z Sequentially visit the nodes list and invoke node_call_back on each element. Returns the nodes list after the node_call_back is invoked on each element. rrAs rM nodes_maprss t Lrdold_nodenew_nodecURU5 URR5 URR U5 g)z- Replace all uses of old_node with new_node. N)replace_all_uses_withusersclearrU erase_node)rtrus rM node_replace_r{s4 ""8, NN NNh'rdc[U[RR5(aN[ US5(a<SUR ;a+UR R SUR S05 gggg)Nrr~)r\rXrYrhasattrrupdate)rrCs rM_update_gm_meta_if_possibler s_3,,-- C   #((8"456 ! ! .rd call_mod_nodec b^ URS:XdeURRnUce[UR[ 5(de[ XR5nSURR5nSURR5nURRVs/sHoURS:XdMUPM nn[X0R5H;upx[U[RR5(de[Xx5 M= URRU5 UHnURRU5n URS:Xa~U Rn [!X5(a3Sn SU 3n [!X5(aU S- n SU 3n [!X5(aMXl[#XR[ X%R55 [XY5 M [%U5S :GaC[%U5S:Xa[%US R5S:XdeUS RS m [T [RR5(a'T R&R)5 [UT 5 O[T [*[,45(aT H!nUR&R/US 5 M# [1[+UR&R355S 5n [5U U 4S j5 URR7U5 O3[9S [;T 5S 35eURR7U5 SSS5 UR=5 UR?5 U$s snf!,(df  N5=f)z Inline the submodule of the given node into the parent module. Note: we only support the case where submodule takes tensors inputs. rzNc3H# UHoRS:XdMUv M g7f)rNrW.0rbs rM node_inline_..!s K.D'']2J44." "c3H# UHoRS;dMUv M g7f))routputNrrs rMrr"s  +ww>W/W+rrrRrsubmod_rchURS:H=(a UR[R:H$)Nr})rWr9operatorgetitemrls rMr7node_inline_..Ls)O!;"8 x'7'77"8rdc<>[UTURS5$Nr)r{rQ) get_item_node new_outputs rMr7rRs -%"=#5#5a#89+rdzUnsupported output type z2. Expect it to be a Node or a list/tuple of Nodes.) rWrU owning_moduler\r9rr<rVrrQrXrYNoder{r[ node_copyr}r?rrxryrrr>rBkeysrsrzNotImplementedErrorrdelete_all_unused_submodulesr_)rrsub_gmphsbodyrbrphrrunew_target_namerget_item_usersrs @rM node_inline_rs   } ,, ,    * *B >> m**C 0 00 0 R-- .F KFLL.. KC ++ D &||11 I1tWW5Hd1F Is../#uxx}}----b0  " "= 1Dxx))$/Hww*$"*//2//A(/smO!"66Q,3A3-""66#2OOWV[[-IJ $ ) v;?v;!#F1INN(;q(@ @@*J*ehhmm44  &&(mZ8Ju 66&DJJNN6!9-'".,,11348" "##..}=).tJ/?.@@rs    * *= 9e 2h##%LLN IyJ 2 1s 7NNA>N F,N  N.c[R"UR5nURSnS[R 0n/nUR 5HjupV[URU5Vs/sHowRPM nnUH0n U S:XaM UR[R"X55 M2 Ml [R"US9$s snf)z Get source code and parse argument names using AST. The function returns a signature of the forward() function. # TODO: Directly provide inspect.signature compatible TS-d module. rrQself) parameters) astparsecoderrPOSITIONAL_OR_KEYWORDr:r<rQrrinspect Signature) rCast_mod ast_func_def arg_type_map param_listarg_type param_typea arg_name_listarg_names rM&_get_torch_jit_trace_forward_signaturerdsii!G$+LLOLI;;>"c)ns;7G&GG  J G  , Ec *44 D DDrdcURU5 [R"SU5nU(aAURS5URS5pT[ U5X:a[ U5X'ggg)Nrrr)r@rrrr)rGrrrrtrs rM _build_cachers]NN4 HH]D )E KKNEKKN q6N* *%(VN " + rdc /nURRGHvnURS:XdM[UR[ R R5(dMKURRS:Xa]URup4pVUR[XR5U45 UR[XR5U45 MURRS:XaFURSURSSpUR[XR5U45 GM"URRS:XdGM?URupn UR[X R5X-45 GMy UHup|0n [[5n[5n[URR5HunnU[!U 5:aNU UR"XR"'U UR"=UlUl[%UR"X5 Mc['XXR"UR"5UlM [)U5 UR+5 M g)a Propagate placeholder names from the top-level graph into HigherOrderOp subgraphs, and handle collisions with non-placeholders by count suffixing. Different HOO subgraph types have different input schemas, so we first enumerate them and gather the top-level named placeholder nodes. r}condwrap_with_set_grad_enabledrrNmap_impl)rUrVrWr\r9rXrr_name_argsrr<rrrTrrrGrr_name_hoo_subgraph_placeholdersr_)rsubgraph_ph_tuplesrbr true_graph false_graph cond_argssubgraphr body_grapharrayrQhoo_phsrrrrs rMrrsRT 77o %* KK77+ + {{  F*8< 5{"))727H7H+I9*UV"))727I7I+JI*VW""&BB $ 1 tzz!"~#"))72+G*MN""j0*.**' 4"))R!2!23U\B&0#%)4S)9"u  !5!56GAt3w<&-ajoo#*1!*//9 DKTYYC6j))TYY 7 (1!0rdexport_graph_signaturerBc  ^0n[U[RR5(aJURR H0nSUR ;dMUR SXxR'M2 Sn Sm0n [[5n [5n [X#U5n [U 5upURVs/sH9nUR[R :XdM#UR"RPM; nn[%UU5HQuunnnU(dM['U U U U[([R SR+U4SjU55-SS9 MS URHnUR[R :XaM#UR[R,:XaSnO U "UR.5R15n[2R4"S SU5n['U U U UR"R[(URU-SS9 UU;dMUUXzUR"R'UU M URR H6nUR6S :XaM['XXRUR5 M8 URR GHnUR6S :XaURU ;deXR=UlUlURU;aYUR R9S5cXxRUR S'O!UR SXxR:Xde[UR S [:5(a URUR S lMMURU ;dGMXRUlGM [=U5 UR?5 URHnUR"RU ;deU UR"RUR"lUR[R@:XdMgUR.U ;dMyU UR.S S UlM URBHnUR"RU ;a(U UR"RUR"lUR[DRF:XdMeUR.U ;dMwU UR.UlM [IURK55HnUUnUU ;dM[U[RL5(aM1U UnUU:wdM>[2RN"SU5(dM\U[([R@U-:wdM|UUU'UU M g s snf)a This pass is run at the end of _export_non_strict() to assign better placeholder node names: - User inputs: These follow the signature of mod.forward(), e.g. forward(x, y) produces nodes x, y. For nested inputs from dictionaries, lists, tuples, or dataclasses, the names are a concatenation of the path to the tensor. e.g. x = { 'a': torch.randn(), 'b': [torch.randn(), torch.randn()] } produces nodes x_a, x_b_0, x_b_1. - Parameters/buffers/constants/custom objects: These follow the FQN of the object, prefixed by "p", "b", "c", "obj" respectively. e.g. self.bar.l0.weight produces "p_bar_l0_weight". - Effect tokens: These are named token, token_1, ... r~cURS5(aU[S5SnO$URS5(aU[S5Sn[R"SSU5nU$)N L__self___self_ [^a-zA-Z0-9]r) startswithrrsubxs rM _strip_name,placeholder_naming_pass.._strip_namesY << % %#l#%&A \\' " "#g,.!A FF?C +rdcP[U[5(a-[R"SS[ UR 55nU$[U[ 5(a[ UR5$[U[5(a UR$[S[U5SU35e)NrrzPytree key of type z not handled for ) r\r&rrrrr'rr#rGrirrs rM_extract_pytree_key4placeholder_naming_pass.._extract_pytree_keys} a $ $SZ8AH ; ' 'quu:  : & &66M!4T!WI=NqcRS Srdrc3P># UHnT"U5R5v M g7fr)lower)rrrs rMr*placeholder_naming_pass..s#L8a.q177998s#&T)rr+rrNrz arg(\d+)_1)(r\rXrYrrUrVrrGrrrTrr)r5r6r USER_INPUTrrrplaceholder_prefixesjoinTOKENr9rrrrWrrrr_ CUSTOM_OBJrgrUSER_INPUT_MUTATIONrrr]r)rrrCrrfake_params_buffersrB custom_metarbrrrr combined_argsrrrEuser_input_namesarg_path_arguser_input_name base_namerGconstantrrs @rMplaceholder_naming_passrs6#%K#uxx++,,IIOOD499$)-8)< II&$ T "H%0%5N5J.ckJM3MB+666D 99 ,, ,  6.11DFV-W)4/ ? &$Y%9%9:((L8LLM# .X'22 99 ,, ,  99  'I#DKK0668IFF?C; "    HHMM  +i 7    #4?y3IK/ 0I&-38 77m # " j))TYY  77m #99( ((&.yy&9 9DI yyK'99==*2*5ii*@DIIh'99X.+ii2HHHH$))E*,=>>(,  % %? YY( " +DI"$B'LLN'22xx}}((( / II-- -$++2I"4;;/3DK 3'33 88==H $$TXX]]3DHHM 99 66 64;;(;R"4;;/DK 4Y^^%&T? 8 J ell% %  ~HD HH]D11 4Y5I5I JT QQ&. (#dO's ?"W%Wr`in_placec0U(a<UR5H&up#[US5(dM[XS5 M( U$0nUR5H<up#[US5(a"UR5R 5XB'M8X4U'M> U$)z If `in_place` is false, return a new copy of `state_dict` with "proxy" removed from `v.__dict__`. `v` is the values in the dictionary. If `in_place` is true, modify `state_dict` in place. r)r:r}delattrdetachclone)r`rrrnew_state_dicts rMremove_proxy_from_state_dictrss $$&DAq'"" w/'$$&DAq'""$%HHJ$4$4$6!$%q! ' rdc/n/nURRGHnURS:XaXSUR;aHURSnUb4[ U[ R 5(aURU5 MhMjMl[U5S:XdM}SUR;dSUR;dMSnSUR;aURSnOSUR;aURSnUcM[ U[ R 5(dGMURU5 GM [X-5$)a For a given graph module, we look at the "val" of placeholder nodes to find the fake inputs. Additionally, if gm doesn't have placeholders, we further look at the "example_value" or "val" of other nodes. If no fake mode is found, we return None for fake_mode. rrNr example_value) rUrVrWrr\rXr]rrr )r fake_inps fake_valsrbfake_vals rMrrs%'I$&I 77m #(:yy'H# 8U\\(J(J  *)K# ^q tyy (ETYY,>H$))+99_5$))#99U+# 8U\\(J(J  * I1 22rdc# # [UR5n[UR5nURR5 URR5 Sv XlX lg!XlX lf=f7fr)r_state_dict_hooks_state_dict_pre_hooksry)rCstate_dict_hooksstate_dict_pre_hookss rM_disable_load_state_dict_hooksrsn,01F1F,G04S5N5N0O!##%9 0$8!!1$8!sAB"A3& B3BBrWrc[RRUR5[RRR 5=(d1 [RRR UR ;$r)rXr%_dispatch_has_kernel_for_dispatch_keyrG DispatchKeyCompositeImplicitAutograd py_kernelsrs rM _is_cia_opr sX 66 GGIuxx++EE  K 88   9 9R]] J rdc<[U5=(a [U5$r)_check_valid_to_preserver rs rM_is_preservable_cia_opr s #B ' :JrN:rdcLUR5RS5SS:H$)N::rrU)rGr;rs rM _is_aten_oprs! 779??4  #v --rdc"[U5(+$r)rrs rM _is_custom_oprs2 rdc [RRS5nUHn[UR S55up#UR S5n[ U5S:Xd[ U5S:XdeUSnSn[ U5S:XaUSn[ [ [ [RU5U5U5nM g) zu Utility function to query C++ dispatcher to get the all possible CIA ops and populate them into torch.ops namespace rrr3rrrrZN)rXr,_dispatch_get_registrations_for_dispatch_keyrr;rr<rT)cia_opsrW namespaceop_name split_listop_overload_namers rM_materialize_cpp_cia_opsrs hhCC#G "288D>2 ]]3' :!#s:!';;;Q-$ z?a )!}  GGEIIy97CEU Vrdc[$)zU This is an special marker that tells our infra that we shouldn't decompose this op. )NotImplementedrPs rM_special_op_to_preserve_ciars  rdcSSKJn U"U5(agU[R;ag[ US5(dg[ UR RVs/sHo"RcMUPM sn5nUS:g=(d UR RnU(ag[RRUR55(dggs snf)Nr)#_should_decompose_because_unsafe_opF_schemaT) torch._decomprr metadata_fnsr}rr r alias_info is_mutablerXr_dispatch_has_kernelrG) op_overloadrrr#is_mutating_or_aliasings rMr r sA*;77&333 ; * *''11N1q\\1NJ)AoO1D1D1O1O 88 ( ()9)9); < <  Os C-Cr)maxsizecH[[RR5$r)(_collect_all_valid_cia_ops_for_namespacerXrTrUrrdrM-_collect_all_valid_cia_ops_for_aten_namespacer+s 3EIINN CCrd op_namespacec[5 [5nUHSn[X5nUR5H1n[X45n[ U5(dM UR U5 M3 MU U$r)rrTr< overloadsr r@)r,rrW op_packetoverloadr&s rMr*r* sbeGL- !++-H!)6K%k22 K(. Nrdcn[5n[RRHnUS:wav[ [RU5(de[ [RU5n[ U[RR5(aU[U5-nM}MU[5-nM U$)a This is an util function that gets the all CIA functional ops. The algorithm is in 2 steps: 1. We first query C++ dispatcher to get the list of CIA ops and then we call getattr on torch.ops.aten to lazily populate them. 2. Sometimes, handful of ops have CIA registered in python dispatcher but not on the C++ side, these can't be caught at the first step. So we walk again to get the final list. Note that the output of this function should never be modified rU) rTrXrT_dirr}r<r\r _OpNamespacer*r+)rop_namespace_namer,s rM_collect_all_valid_cia_opsr5seG"YY^^  &599&788 88"599.?@L, (?(?@@CLQQA DF FG, Nrdc [RRRnXR;aE[ URU[RR5(dURU$Sn[ R"X S9$)NcNUSnUS [RRRn[RR UR 5[RRR5(aUR "U/UQ70UD6$[SUS35e)Nkernel Expected z) to have CompositeImplicitAutograd kernel)rXrrrrrG_op_dkAssertionError)rQrRr8dks rM_special_op_to_decompose_cia9_get_decomp_for_cia.._special_op_to_decompose_ciaCs! 8 XX ! ! ; ; 88 9 9 KKM588//II  ==5d5f5 5 F8#LM rd)r8)rXrrrr r\r\r])rWr<r=s rM_get_decomp_for_ciar?8sj    7 7B ]]:bmmB.?AUAU#V#V}}R      9 EErdc## [RRn[RRnS[RlS[RlSv U[RlU[Rlg!U[RlU[Rlf=f7frD)rXcompiler_is_compiling_flag_is_exporting_flag)old_compiling_flagold_exporting_flags rM_compiling_state_contextrFUsx::::?,0),0) ,>),>)-?),>)s5C.B&+C,B==Crc"0[URSS95E[URSS95En0n0nUR5HDupV[ U5U;aU[ U5nOUR USS9nXt[ U5'XsU'MF U$)NF)r|T) static_shapes)rrrr:id from_tensor)rrCparams_buffersfaked_params_buffersmemorrH fake_tensors rM_fakify_params_buffersrObs s##U#; < s  % 8 9N "$D$**,  e9 r%y/K#//T/JK)EO$/S! - rdc .^^ ^ [T[RR5(deSSKn"U U4SjSUR 5m S[ S[[[ [44U 4SjjnS[[ S [S[ 4U 4S jjnS[ S[[[ [44U 4S jjnUm UnTRS -TR-nS nU 4Sjn[TT UUUUUS9 S[[ 4U 4Sjjn [TU 5 g)a Registers a module as a valid input type for :func:`torch.export.export`. Args: mod: the module instance serialized_type_name: The serialized name for the module. This is required if you want to serialize the pytree TreeSpec containing this module. Example:: import torch class Module(torch.nn.Module): def __init__(self): super().__init__() self.linear = torch.nn.Linear(3, 3) def forward(self, x): return self.linear(x) torch._export.utils.register_module_as_pytree_node(InputDataClass) class Mod(torch.nn.Module): def forward(self, x, m): return m(x) + x ep = torch.export.export(Mod(), (torch.randn(3), Module())) print(ep) rNc>>^\rSrSrUU4SjrSrU4SjrSrU=r$)=register_module_as_pytree_input_node..PrototypeModuleic>[TU]"U/UQ70UD6 [U[RR 5(de[ US5(aeTUlg)N _proto_cls)super__init__r\rXrrr}rT)rmrQrR __class__rs rMrVFregister_module_as_pytree_input_node..PrototypeModule.__init__sO G Q 0 0 0a11 11t\22 22!DOrdc4URUR:H$rrT)rothers rM__eq__Dregister_module_as_pytree_input_node..PrototypeModule.__eq__s??e&6&66 6rdc>T"U"55$rr)rrMPrototypeModules rM __deepcopy__Jregister_module_as_pytree_input_node..PrototypeModule.__deepcopy__s"46* *rdr[) __name__ __module__ __qualname____firstlineno__rVr]ra__static_attributes__ __classcell__)rXr`rs@rMr`rRs "  7 + +rdr`rr1c>[UR55n[UR55n0UEUEn[UR 55[UR 55T"U5/4$r)rrrrrr)rrrrKr`s rMr @register_module_as_pytree_input_node..default_flatten_fnss 4 4 67S..01 >,> >N))+, $$& ' C /   rdrrc >^Uup#UbU"5c [S5eU"5nT ceT "U5upVS[RR4U4Sjjm[ S[ X555(aY[RR RRU[[ X 55SSS9 T"U5nSSS5 U$UnU$!,(df  W$=f)Nz!Module has been garbage collectedrCcP>[R"U5nURR5VVs0sH0up#[R"U5[R"U5_M2 snnUlUR5HupE[ XT"U55 M U$s snnfr)copy__dict__r:named_childrenr?)rCrnrrrGchild copy_modules rMrqWregister_module_as_pytree_input_node..default_unflatten_fn..copy_modules{))C.CCF<.default_unflatten_fn..s=&.default_unflatten_fns! ;#%-BC Ce%%%!#  UXX__  =c&&<= = =))??T#j12T@"#&  C   s ; C C!c>T"U5untp#[X!5VVs/sHupE[U5U4PM snnU/UQ4$s snnfrr)rrr rQrrrs rMrJregister_module_as_pytree_input_node..default_flatten_fn_with_keyssW)3C& &J/2:/IJ/ItqA"/IJ M M   Js>r3cTUtp[R"U/S/[U5-Q5$r)jsondumpsr)rrrs rMrAregister_module_as_pytree_input_node..to_dumpable_contexts*zz44D6CF?455rdc>[R"U5nT"[RR 55US'U$r)rloadsrXrr)dumpablesr`s rMrCregister_module_as_pytree_input_node..from_dumpable_contexts/ JJx uxx01!rdrc>>T"U5up#X1R:XdeU$r)r)rrEflatsrrs rMdefault_flatten_fn_specEregister_module_as_pytree_input_node..default_flatten_fn_specs##C,,&&& rd) issubclassrXrrweakrefr|rrrr rrdrerr) rrr rrrrrrrr`rs ` @@rM$register_module_as_pytree_input_noderwsH c588?? + ++ + + +'++ +  d3i.@(A Xc]W8 # %S 78J2K $J'L>>C/#2B2BB6  19/3d3i ! rdc0[U5 [U5 gr)rr)rs rM&deregister_module_as_pytree_input_nodersC #C(rdc[U[RR5(d SUS35e[U[RR5(d SUS35eURUlUR Ulg)Nr9z to be a nn.Module)r\rXrr _parametersr=)srcdsts rM _sync_staters     + 3%)*+      + 3%)*+ ooCO<[TU]5 [URU5 Xlgr)rUrVr__self__r)rmethodrXs rMrV_WrappedMethod.__init__s! FOOT* rd)r)rcrdrerfrVrgrh)rXs@rMrrs rdrcN[U5(d SUS35e[U5$)z Wrap a method as a module so that it can be exported. The wrapped module's forward points to the method, and the method's original module state is shared. r9z to be a method)rr)rs rM wrap_methodr%s<   + 6(/*+  & !!rd)r1r)rjr)rnrrlr)Fr)NN)r1N)r&r)rWr)rrmrr\rrrrr collectionsrcollections.abcr contextlibrrrtypingrr r r r rX torch._guardsr torch._subclasses.fake_tensorrr#torch._subclasses.functional_tensorrtorch.fx._utilsr"torch.fx.experimental.proxy_tensorrtorch.fx.passes.runtime_assertrr4r torch._opsr torch.exportrtorch.export.graph_signaturerrrrtorch.fx._pytreerrtorch.utils._pytreerrr r!r"r#r$r%r&r'r(r)r*r PARAMETERr7rfrrrrErNrYrrcrkrorrrrrrrrTboolrrrrrrrrr r$rr'r)r]r+r.r1r?rBrGr`rirorqrsr{rrrrrrrrrr _subclassesrNrrr r rrrrr  lru_cacher+rr3r*r5r?rFrOrrrrrrrrdrMrsr   #$%'@@ *D@?KJH',AQQ$"  d t & OOW #(<%   %4 (  # .D  6'(<(<6'c3h6'r.)!%c3h) )$) )8 $uxx33 $*! %38n%cN%C% %  %  %P//S/0 [ #u||# $[ [  [ } [  [ |)UXX]]+) )\)-,04 +/9==A$4 c4%4=)4 #3- 4 ""56 4$$9:44 4nE'Euxx}}EE    ((-- ehh  !  B(B B$B  ((--ell&R R ((--R R  ((--ell$ ehhmm_eEHHMM44G.HHI XX><UXX]]+<UXX]]@S< 33-UXX-A-A-d-`!UXX%9%9!J04     ehhmm 4tEHHMM*4s4 T%((--(T%((--=P(EHHMM(UXX]](t(7EHH$8$87uxx7SW7L L(588;O;O2PL^44GDUDU42E&,-(<(<--`g$ g$2g$ g$CH~g$ g$TTTd*3 3 e++::;3< 9 9 9>d;~;$;.N.t.n W.4 QDs>7JD D**))"C$78F: ? ?    #uU\\588#5#556 67 *Ad588??.CAAH)UXX__0E)$)  "$UXX__ "rd