ELF>@0P@8 @0A0APPPUUPPPHnHnPPP h-888$$PtdQtdRtdPPPGNU46h['04Ɓ˖+I   ! V  rn@a 4[ Mo X ffm  E  ~phd@>  "I  . G v06 SU:1   :R.wX2 L.Z"ZH   G, v  F": kb  __gmon_start___ITM_deregisterTMCloneTable_ITM_registerTMCloneTable__cxa_finalize_Py_NoneStructPyFloat_FromDoublePyDict_NewPyExc_TypeErrorPyErr_FormatPyLong_FromLongPyObject_GetAttrPyUnicode_FromStringPyMethod_NewPyUnicode_InternFromStringPyExc_SystemErrorPyErr_SetStringPyDict_SizePyUnicode_FromFormatPyType_ReadyPyGC_DisablePyGC_EnablePyExc_DeprecationWarningPyErr_WarnFormat_Py_DeallocPyDict_SetItemPyType_ModifiedPyObject_HasAttrPyObject_CallMethodObjArgsPyObject_GC_UnTrackPyObject_ClearWeakRefsPyObject_GC_DelPyTuple_NewPyDict_SetDefaultPyBytes_FromStringAndSizePyBytes_AsStringPyCode_NewWithPosOnlyArgsPyCapsule_NewPyObject_GetAttrStringPyDict_SetItemStringPyExc_AttributeErrorPyErr_ExceptionMatchesPyErr_ClearPyThreadState_GetPyInterpreterState_GetIDPyExc_ImportErrorPyModule_NewObjectPyModule_GetDictPyObject_GC_IsFinalizedPyObject_CallFinalizerFromDeallocPyTuple_GetSlicePyTuple_GetItemPyMem_MallocPyDict_NextPyMem_FreePyErr_NoMemoryPyExc_RuntimeWarningPyErr_WarnExPyImport_ImportModulePyMemoryView_FromMemoryPyObject_CallFunctionObjArgsPyUnicode_DecodeUTF8PyUnicode_InternInPlacePyObject_Hash_PyObject_GC_NewPyObject_GC_TrackPyExc_ValueErrorPyDict_GetItemStringPyModule_GetNamePyCapsule_IsValidPyCapsule_GetNamePyCapsule_GetPointerPyFloat_TypePyLong_TypePyExc_ZeroDivisionErrorPyLong_AsDoublePyNumber_TrueDividePyErr_OccurredPyList_TypePyTuple_TypePyLong_FromSsize_tPyObject_GetItemPyUnicode_NewmemsetPyUnicode_FromOrdinalvsnprintf_Py_FatalErrorFuncPyObject_IsTruePyObject_GenericGetDictPyDict_TypePyDict_UpdatePyObject_VectorcallMethodstrrchrPyImport_AddModulePyDict_GetItemWithErrorPyType_FromModuleAndSpecPyExc_RuntimeErrorPyBaseObject_TypePyErr_GivenExceptionMatchesPyCFunction_TypePy_EnterRecursiveCallPy_LeaveRecursiveCallPyObject_VectorcallDictPyObject_Call_PyThreadState_UncheckedGetPyFrame_NewPyTraceBack_HerePyCode_NewEmptymemmovePyMem_ReallocPyLong_AsLongPyFloat_AsDoublePyLong_AsUnsignedLongPyExc_OverflowError_Py_FalseStructPyList_NewPyImport_ImportModuleLevelObject_Py_TrueStructPyObject_GenericGetAttr_PyObject_GenericGetAttrWithDictPyObject_RichCompareBoolPyExc_NameError_PyDict_GetItem_KnownHashPyNumber_InPlaceAddPyMethod_TypePyNumber_SubtractPyEval_SaveThreadPyEval_RestoreThreadPyUnicode_FormatPyObject_VectorcallpowPyNumber_NegativePyNumber_MultiplyPyGILState_EnsurePyGILState_ReleasePyObject_FormatPyUnicode_TypePyNumber_Remainder_PyType_LookupPyDict_DelItemPyObject_SetAttrStringPy_VersionPyOS_snprintfPyUnicode_FromStringAndSizePyType_TypePyTuple_PackPyObject_SetAttrPyImport_GetModuleDictstrlenPyUnicode_ConcatPyImport_GetModulePyInit__sgd_fastPyModuleDef_InitPyObject_GenericSetDictlibm.so.6libc.so.6GLIBC_2.2.5  ui   ui  P`X ``jjjjjX@ h(@088@`HHPX`hpx` ph0hX(_ ( 8@HXPhx@0Hfli(wiPixi(_  li(НjP \uXP_p_(hx Hf8_P`HXPHf h`0@`(8Hf`a@aPHfb@pb``Hfhc@P(d@@sfHX`jfh`x3gf `ddsfjf  3gf f(@gHp`Eghp|x{RgHfHf Hf`Hf Hf(@fH `3ghjf`sff 3gjf   sf(8@iHP`hipx`iPiPiPi0iXi`i jj$jК .j(КH7jPXpEjxRj aj`pqj  (. 0(10H8I@KHMPPX\`^h_pfxlnrwz|!8!p@p ( 0 8 @HPX`hpx !"#$%&')* +(,0-8/@2H3P4X5`6h7p8x9:;<=>?@ABCDEFGJLNOQR S(T0U8V@WHXPYXZ`[h]p`xabcdeghijkmoqstuvxy{} ~(08@HPX`hpxHH)HtH5%@%h%h%h%h%h%h%h%hp%h`%h P%zh @%rh 0%jh %bh %Zh%Rh%Jh%Bh%:h%2h%*h%"h%h%hp% h`%hP%~h@%~h0%~h %~h%~h%~h%~h %~h!%~h"%~h#%~h$%~h%%~h&%~h'p%~h(`%~h)P%z~h*@%r~h+0%j~h, %b~h-%Z~h.%R~h/%J~h0%B~h1%:~h2%2~h3%*~h4%"~h5%~h6%~h7p% ~h8`%~h9P%}h:@%}h;0%}h< %}h=%}h>%}h?%}h@%}hA%}hB%}hC%}hD%}hE%}hF%}hGp%}hH`%}hIP%z}hJ@%r}hK0%j}hL %b}hM%Z}hN%R}hO%J}hP%B}hQ%:}hR%2}hS%*}hT%"}hU%}hV%}hWp% }hX`%}hYP%|hZ@%|h[0%|h\ %|h]%|h^%|h_%|h`%|ha%|hb%|hc%|hd%|he%|hf%|hgp%|hh`%|hiP%z|hj@%r|hk0%j|hl %b|hm%Z|hn%R|ho%J|hp%B|hq%:|hr%2|hs%*|ht%"|hu%|hv%|hwp% |hx`%|hyPQHGuH3wHH5H81(1Ht$H;W tH wHH5H81]ZATHUSHHtHPuHH[]A\*HPHtwH HqH9~bHDuHPHvH5H81jHu,H t"HHHSH5HPvH819H"HHAHEt []A\AUIATIUHSQHÅuLLH5Ht1H5HE1LL1HHt Hu Z[]A\A]AWIAVIAUIMATDAIULLSH(H4$HT$D$1HH2H4$1L9}HHHTHHHHkH$Hu1Mu L-E1FIE1HhHHHHl$IHtHHHHL$1HLH-MMH D\$LLMHULDAUIDQHAAWHDAWt$@PPt$@PPAPE1H`HMt IuLH tH HH(H[]A\A]A^A_UH1HS1Q?HuJHH5HHHHHyHuH0Ht1 HZ[]AVEAUIATIHUSHt3H;sHu1AtHLLH u'HHsH8t1[]A\A]A^ATUHSHxHtJHwHu Hw;H9t6HAsH5H8H5] HIHu1H+HtHH&I $HuLHtHIHuH uHAH HHHxAH LHHxAH ~LHH}exE1H qLHHmEdH[]A\AWH5AVAUATUSHH|$H=alHHWIHHHLH*IHty/ H= IHuIMuzLp1HL1IHuLIMuLHMuHyI $uTLkJHBqHH5H81)HtHMuH6I $L$HLs 1L-E1HcK<&1D|AL9HD$HvHt H|$HD$HtH`HD$0HHD$8HHD$@HH H[LIHH HHD$0H|H=HD$8H HD$@HőHD$HBHHu HL-IHH HHD$0H=5HNHD$8H]H& HIHLH YH*HHD$0HHH_IHLH HHHD$0gHHHIHL|H ͏HHHD$0!HHTHӏIHLH HXHHD$0HeHHIHLȐH AHHHD$0H'H -IHD HDyD(%ΏfDl-fDՍD=5ffD- fl~fDl=)$%'fD%-Ufm ڎfDmfmD-B %fmDfDl)L$nj ЌfDLfDϋf flfD ۍ)|$p)$=75flŎD)L$P)L$  3f=ۋ)$= flD)D$`)D$0D)$D)$f)D$@cD)$D)$fm5)$gf5fl=$)$(4$)$)$fm)$ =)$%Ȍ5ы)$Ҋf=*)$P(l$f%ōf5f5)$%)$p-)$)$Ɋf-Q)$@(|$ ffm)$0D)$`)$)$D)$)$)$)$ L7H HH#HHVD=͉IH%p=HmfDlD% fmD-_% QfDmU-fDlD %ӉflD(DDۋ)$ yfDl-։fDfD35|fm%fD)L$- hffD؉fm]f flfD i)|$p)$=ʼn5NflSD)L$P)L$  f=i)$=flD)D$`)D$0D)$D)$f)D$@D)$D)$fm5)$f5fl=)$(4$)$w)$fm)$ <5u)$P(l$=q)$%Bf5Z)$p-[f=f%K)$,f-)$%)$-ff)$Of%)$0(t$ fm)$@D)$`)$)$D)$)$)$)$ LH ·HHSHHIHLH HHHD$0HHD$8HHD$@HHD$HGHHt~HH IHHLHD$0HNH=/HD$8HKHD$@H7HD$HH HHQHtHHuH1H uHH8[]AWAAVIAUMATIULSHH=HT$VH1E1MDHx(HT$HLpH@L@ptI$11Lc HK@HKHHHSPHsXHs8Lk`IEHtHE1W1HkhHCxAFH%tOtQHItP=t@H[I=t9HaH5PH8^H u,H"H/ 1HHHS0H1HH[]A\A]A^A_AUIHATIUHSAPHHH@u HaHLH5sH81lLK(HC Mt I9LLIM9s#HWaMHLH5WH81(I9s2Q11MPILH011^_yH uH1ZH[]A\A]AWIAVIAUIH5ATIUSAQHuLHHHHu2LRMH H5!HHaH81HLH:uCH^LIMMHCHHT`VH5AWH81_AX=LHHMIHQHt"HUHuHH uH1-HMuHxHH H^Z[]A\A]A^A_AWAVAUIATIUSHL>.HT$LHtLxLcHHCH=yHHHHH$HHHIHuHt%HIcT$LHHT$LH5IHMtHxIELhHu ^HtwH<$LHIHtSIEM9It^HIuL$IcT$LLMAt H uHLHMuHIuLE1HHIuLHH[]A\A]A^A_USQH5HHtB1HHHu DHuH]H5AH8H uH\HHZ[]AWAVAUATIU1SH(LPH}Ht HHcHcHHHIEL4HD$H9l$IDH'HHuHI$IM9tI?u+HHL$Ht$HL$Ht$IGIIH9tIHu=HcH5hIDHHI$HPH\H81*HHpLHt1H([]A\A]A^A_UHSQH5zHHu1Ht%HHŅxH uHZ[]AWAVAUIATUSHH5~5H-n\HuH5"HHHu$H5~~HI1I9lH5LO*H$HH9H5HHHu1E1E1E1xH5L*IHtH9uH5wLyIHu$rH5^HÅu1E1E1OH54HL IH5Åu L *AL9H1H5LyIHtH5 HuhAH5~LxHHt=H5~HLoxZIH5~aÅu Lsx5Mt.Hu$1LTF11E1E1H$E11E1UHuHYIUH5H81TMt IuL_H<$u *1E1E11H$H $HHD$HHuH,Mt IuLMtI $uLHtHMuHH[]A\A]A^A_AWAVAUATUSHHˀHt)1H9!HYH5 H8U!HH=HvHHH=#HtHHvHH=HtHHvHHvH=%H5oHXH0HH u1LHmvHuh9Hl$0RHHHHA L jRH}PHqP1H 1HrH-1H'HuH1HMHuHH5W1HHuHn01HH5)^Hm}HMuHH=W}tH.H=@}1H5\eH3}Ht =Oru2VHHyH5zH=v~!HHL% HLHtH=Rz赈H|HH~LHp~VHHyH5wfHnH0fHn(1flfl) uu|HuHL%{LpHHHoLHHtHHHB}HH sH=`iH })rsH@H=tHpsHsH2j<H5|H=ta!H=kt| HWtH5vH=|\H=5tHrH=rfHc|Hl|)rH@H=sHrHrH:gH5|H=siH=sUHsH5huH={3H=sHqH=cH{H{)qH@~^QH=GsHqHfHnH%sflHd)qH5?{H=sH=r~HrH5tH="{\HpH=7aHzHz)pH@~PH=rHpHNfHnHbrflHa)p3H5tzH=ErH=1r*HrH5tH=_z HpH=^H zHz)oH@~OH=qHoHfHnHqflHD_)op'H5yH=qL H=vqgHbqH5sH=yGH8oH=[HByHcy)oH@~-OH=qHoHxfHnHpflH\)ndH5xH=pIH=p5HpH5prH=xHUnH=NYHwxHx)6nH@~rNH=[pH,nH%fHnHpflHY)mH5xH=pH=ptvHoH5!rH=xHEHȅxOHEHuHL%LHHCHLH\HfnHu2HMH 1L5aAE1H$HHtHLHHnHtHHt H_LHHmHbH]HM H1LHHmH&H!HHLHJHtmHHHHLHH@mHHH HLHH mHrHmH] HLHHlH6H1H!H_LHZHlHHHH6LHHplHHHH LHHj1mHH H5jH=dHHHȅXHHuH,Ht$E11ɺH=NkLdH|$0mHHu1L5AAH$^L5kHLHHHFH8u0HFLH5BE1L5ZAH81bYHHHtIHtH5lfH|IHtGLHiIHt#HyIHD$tL|$ LI $uLMIMuLM91E1L5AIH=3cLL84 I $uLH uHL dLbH=\H diHefHH H5}hH=cHHHȅ1E1L5A2H uHAMtLE1H AuHHtHMuHMtI $uLH=ktBH=+btEtLDH=5H=kHt51HHku' LBHDH!2IH H5H81n1HHytH1IH H5H817DH1IH LH5WH81@Hy[ff.?HGHHtfD#H=R$@HGXHtHfSHGHHxHtHCXHtH[HA1H[ff.Ht 6fDHHHGHHtHfSHGHH8谷HCHHtH[ÐUHHSHHHGLH@t~$uFHHH[]Au+HLFI$HvHH[]A@H/H5DH8RH1[]fHLF1MtHCH H5HH/H816@HtHL $HT$L $Ht$H`)fDHL $HT$ϰL $Ht$H#HCH H5 HHY/H81跶0fHL $HT$L $Ht$H&HCH H5HH/H81bff.fHwPH1H=ۯff.SHGHHHt+H.HͰH81螶u%H[DH.HH5GH81յH+t1H[H1ְ@USHHHtHHH[]HHHtAHHtHXH@ HHHHH/unHDH .Ht3HGXHHwXHtH(t1ÐHH$1HDH5-ff.@SHHpHtHCpH/AH{ HtHC H/H{@HtHC@H/H{HHtHCHH/H{PHtHCPH/H{XHtHCXH/jH{`HtHC`H/?H{hHtHChH/H{8HC8Ht H/HHtHǃH/HHtHǃH/HHtHǃH/tZHHtHǃH/t-H{xHtHCxH/t1[fk1[[fKf;ofD+>fD fD fDfDfDۭafD˭6fD軭 fD諭fD蛭fDSH至H{(tHhHpH[駭ATIUHSHHpHt HՅH{ Ht LՅH{@Ht LՅH{XHt LՅuH{`Ht LՅumH{8Ht LՅu[HHt LՅuFHHt LՅu1HHt LՅuH{x1Ht[LH]A\D[]A\ff.HGHuHGHtF`0H@HH9P0HGHH@0H@HGHHuff.@HHtTH9P0uff.HHt4HH0H9tfH|$薯H|$kHGQH|$H|$CHHGHHuf.HHtH9P0uff.HHtHH0H9tHDAWAVAUATIUHSHHhLG0MtHwHhHL[]A\A]A^A_HVLϭIH1L蹫HHtiHHLSImHtAHhH[]A\A]A^A_HVHvHu[HhH1[]A\A]A^A_ADL0fDIm:H'HSPH5۪H811DLMMtI<LD$HHT$LL$CIH.HT$LL$LD$HtmHBHID$ I9H1HHff.ff.DAoDAHH9uHHt ILI LLD$HT$LL$菫LL$HT$HLD$HH|$PE1HD$XL|$0H|$H|$HM$AH\$8ILIHD$HH|$HT$LD$ LL$(+HD$PHT$XHHHHL#IDIHHT$Ht$LH葬tt HD$PHT$XHT$LD$ LLL$(L|$0H\$8MLL$HLLt$ALL$HL$HH)tK1MM%DHI9tI PH51HEHHEZYHt[]A\HfD[H]A\ܣff.HHFH;!u/Nff.zt4^^HifH;a!HFHu"H+!H5H81HDHtJHHPHtMHf.z ^f(lfDf*N^LVNfHH H*f/vHx^ H鏨FfA*^D$/D$Hff.ATUSHHHGH;Z H;tKHhpHtZH}tS1ԣIHHHUI,$u LHD$HD$ @HGHH[]A\H@hHt7H@Ht.HH1[]A\HGHHH[]A\@1AHHt)HHNHmuHHD$ZHD$x1qfAWH5H ףp= ףAVAUATUSHH(Hl$IHIHHH?HHH)HHHH)HHӍ?HfAEHu1 AIL)LIHyLHcQMI)IH*@ 3IHM~L 蝡HHHEH3LH1H)HHff.fDAoDHH9uHH9HH)HHHvILI)HJ 2HHHӨt~ADLHCH9~iATHCT;H9~VATHCT;H9~CATHCT;H9~0ATHCT;H9~ATHCT;H9~ ATT;H(L[]A\A]A^A_IW0IH@HEDA}H([]A\A]A^A_@1 fSHH$H$H$L$L$t@)$ )$0)$@)$P)$`)$p)$)$H\$ H$D$HD$HH$HL$HD$ 0HD$ԝHH=zşDATIHUSHHF1H@hPHHH葡ŅtSttL1,IHt\HI9D$HLH+tgI,$tHH[]A\DH+uH誝H[]A\H+t:H[]A\fLxH[]A\DH`fDHPfDH=?H1L$$HH\$THHt H(t1:1'fHGPHLVM1ff.@HI9tH9|uf.E1ff.fJTHBt~@tuH9tHXHt,LAM~[1HI9tGH;TufDHDHH9pHuH;^ff.IM9c1ËH?LOIЃtHIALfHtIH>IAHLHH IH H5H81Z1HH?IHHWtHGH8LJHHHAHtHHHfDHHHH H5#H81ڟ1HHLOH?t%IHumHMu4IA1HHt{HH>LBfHIH H5٭H81N1HHytHIH sH5`H81DHIH ,H57H81@Hy[ff.HG@HFH|@tsH9t.HXHt2HJH~I1DHH9t7H;tuf.HH9tHu1H;5&f1DtT@AWH5wAVAUATUSH(Ld$ IHiQHH%kdH)ʍ Hc fA $cE1 H\$ AAK,H)LIHLLD$HcQMI)IHS@ LD$ ^INHM~HL LD$萙LD$HHHCHXLH1H)HHff.ff.foDHH9uHH9HH)H~Hv$II)HK<J<*HHHЃtHB4:HPH9~kILHPBL8H9~ULHPBL8H9~BLHPBL8H9~/LHPBL8H9~LHPBL8H9~ DAH(L[]A\A]A^A_f.IV0INH@HED}H([]A\A]A^A_11AUIIHATI?USHHHF:HM!ItCH9HG8HHHHH1[L]A\A]DH9LhL9LXMIkH~1fL9tHH9|ILH9uHKAcM)La1 uHkH=%uDLHAHӘH腘Hu HH5 H8Zf.1fH9LL9LXMMZM~.1ff.ILH9L9 HI9utHG8HHHL%0HH=0#K1HLHٗHHH[]A\A]fHC0H/HH1L[L]A\A]驗fHff.ff.HH9t4HuH (H9t#HHL9tHuI9fDHKALa1 uHkH=D7_1HAHH衖H8H H5"1H8pHKAuH9tHC0H|fHKA|HfHH9LHuH $H97Hff.ff.fHL9 HuI9pDHLH1[]A\A]鱗HGHuH{.HtF`0H@HH9P0HG.HH@0H@HGHHuff.@HHtTH9P0uff.HHt4HH0H9tfH|$H|$kHGQH|$螓H|$CHHGHHuf.HHtH9P0uff.HHtHH0H9tHDUSHHHtHHH[]HHHtAHHtH@H_ HHHHH/u.HDHAWAVAUATAUHSHHH=4IHDj4DƃHcHD;d1 ~.H9})HcHTA9}߉9|A9A9~{HHHNjGA9ujH/HEH+1HL莏HH2D`(HƑHmKH+9H[]A\A]A^A_ËGA9~f.M}pAoM`HDIEpfHAE`) $賕HHI}`fo$MuhI]pM}pAU`Ht H/Mt I.Ht H+H=3HD52DHcHD;d21~qS9~_)HcHDA9}߉HmHH[]A\A]A^A_HH؎A9A9~}LcIIE;g|:2A9tmD9}4DIc)HPHHHHHHtHHH舒AEgI/D51HE1D;gË1A9Dp@IcHdHHLcD51D51IH1IS@H<$HtH/tOH|$HtH/tFMI/L豍L褍H藍荍膍볿HHJH0H0D`H(HEDHcHL<II?I/HEH/D;gt֋0t"I1AR\1DUSHHHG~HHtfHHHhHt-H>)fHnfHnHfl@H[]fH+u HBfHH=1H[]H HE1L ,H H5}H8R1HÐX1ZDHyxHH=Y[01HmeH蝋XUSHHHG>HHtf輍HHHh袍Ht-H'fHnfHnHfl@H[]fH+u HfHH=%1H[]HIHE1L H LH5=H8R1H@胏X1ZDHyxHH=/1HmeH]XUSHHHGHHtf|HHHhbHt-H&fHnfHnHfl@H[]fH+u H‰fHH=%1H[]H HE1L H H5H8R1HCX1ZDHyxHH=ٝ-1HmeHXUSHHHG辊HHtfHFHH|$HHD$HH9_藇f(f. zH|$H9_ $j $f.zuHEHP覅HHH|$Ht H/H|$Ht H/HHH[]OcfDHHH(L AH H5ۈH8S1(XZH|$HtH/tvH|$HtH/tVH`H=+1HHH[]fGfDHFHHD$HHHD$f˂f軂fD諂HHH[]Ð蛂fD $F $He-D$ $  $D$HgHbH=EYHt$ HT$HHjL &Y^XfUfH$HSfHnHHH)D$~HD$0fl)D$ HLAHMHHH_HGHT$ HH4IHj1PAPLD$0%H SH"H|HHt H|H|$-f.HH>HFHH|$HHD$H&H9_7f(f. {zH|$H9_ $  $f.MzuHEHGHHH|$Ht H/H|$Ht H/HHH[]fOcfDHHHєL 0AH H5{H8S1ȄXZH|$HtH/tvH|$HtH/tVH\H=1HHH[]fGfDHFHHD$HHHD$fkf[fDKHHH[]Ð;fD $ $He-D$ $ $D$HgH^H=XHt$ HT$HHjL iH#Y^XfUfHL!HSfHnHHH)D$~HD$0fl)D$ HLAHMHHH_HGHT$ HH4IHj1PAPLD$0"H SH"H|HHt H|H|$-f.HH>HFHH|$HHD$HH9_׀f(f. zH|$H9_ $誀 $f.zuHEHP~HHH|$Ht H/H|$Ht H/HHH[]OcfDH1HHhL ЏAH *H5H8S1hXZH|$HtH/tvH|$HtH/tVH'EH=ۃ1HFHH|$HHD$HfH9_w}f(f. zH|$H9_ $J} $f.zuHEH{HHH|$Ht H/H|$Ht H/HHH[]fOcfDHHHL pAH ʍH5~H8S1~XZH|$HtH/tvH|$HtH/tVH~2H=1HHH[]fGfDHFHHD$HHHD$fxfxfDxHHH[]Ð{xfD $&{ $He-D$ ${ $D$HgH}CH=XHt$ HT$HHjL Y^XfAUHATIUHSH8LnHD$HD$ HD$(HHHJvHMIH :E1L MHHHH5|H8AU1|XZHT$HtH*uHwH|H=~H8[]A\A]DMIgH}HH|$HH9G"zf.jzH|$AD$HtH/tH81[]A\A]fDsvHUHHT$HSHT$ HM1H jHQ1PLD$8H H|$HH1dM1KG$fDH AL 4cD$lxHT$D$HHhHt$ HT$LIHL QHjY^=AUHATIUHSH8LnHD$HD$ HD$(HHHsHMIH ʉE1L MHUHHH5SzH8AU1yXZHT$HtH*uHtHfzH=|}H8[]A\A]DMIgH}HH|$HH9Gwf.zH|$AD$HtH/tH81[]A\A]fDtHUHHT$HSHT$ HM1H jHQ1PLD$8rH H|$HHP1dM71KG$fDH $AL ĆcD$uHT$D$HHhHt$ HT$LIHL HjY^=AUHATIUHSH8LnHD$HD$ HD$(HHHjqHM IHHH8AUHSL 1AH H5w5wXZH|$Ht H/u>rHwH=zH8[]A\A]fDIxHUHHT$HBH9BtLHTuf.zH|$AD$HtH/t H81[]A\A]qfBHUHHT$HSHT$ HM1H BjHQ1PLD$8 H H|$HHff.SHHGHWHt"HtpHu"GWHH HH u4H[fD3lHH tHujHuPf.HH5zrH8h.GH[HqH5rH8zh|H@`Ht`HHtTHHtJHrH9Cu+@HH+:H߉D$ rgD$ %HHHujHuHH5{H8ghfSHHGHW1Ht-Ht1HHtiHt+iHcH9H[@GH[WGHH HcʉH9tHBH5KqH8Kg@WGHH HHcʉH9uH[ËGfDHuiHtfDH@`HtkHHt_HHtUHH9Cu3ff.HH+H߉D$ eD$ H葴HHu9hH+H7H5[zH8`fff.AUATUSHHHtHHH[]A\A]Hu$HHHPHHHHL%tcHHtH@I$E111HL H=2hH+ItlMIELLHHIImtMMtXHI$HtvHI$ LdH fDHdfDLxdfDfL%$HI$HPI$iInLf.HGHH;u 1cAVAUATUSHHtHHt%HH[]A\A]A^DiHfDbHHH}`L I9HtID$HHW@A$@HXHHJHK1ff.HH9.L;duH}`fLmhLepHEpE`HtH/tOMtImt9MI,$LbL`cuLbbMl$M~Y1 HI9tLI;|u`HHI9MHuL;%-bf1LmhLepE`HEpAE1M9;KtH9H|$(H|$IDAUfH#ATfHnHUSHHH)$~HD$flHD$0HD$8)D$ HLAHMH/HtH HHH$HGHT$ HH4IHnvj1PAPLD$ H H5H<HCHt H<HHt H<H|$L$$4@HfL&H~HFI$L$$HH|$HHD$HGH_HH`HHHb5cHHxLl$H5I9ut6L;-t-ff.fHhuLHD ;H> u HA ?H-fInH=1HHt$ HEfHnfl)D$ cHmHHH-AI9t&I}~LH:H}NHHHHH<$Ht H/ H|$Ht H/H|$H2H/(_HHH[]A\A]aHuHiH5sH8_faHHLl$H5I9ut L;-b|H1"HsH=_iz1DHff.fHHHsL qAH rH5cH8S1cXZH<$HtH/tgH|$HtH/tGH|$HtH/t/HsH=h1HHH[]A\A]]ʐ]f]f_GHH H9H]7H]HHFHHD$HFHHD$C]fD3]fD_HcDH]_GHH fD_fHqH=g1mH-q H=eHH1 H=dgHH#Ht$ HHHjL q Y^)H<$Ht H/ H|$Ht H/H|$H7H/-\#fH@`HHHHHH@HH9~H]Ht+Ht=HtiHtQHt1Hk^HHm#H}[Hc]Hm]Dmt(HL ]EHH H붋]HcH4[-[#[H@`HHHHIHH@H9tzI]Ht#Ht5HtZHt@Ht)L]HImLZIc]A]AEHH A]AEHH HA]HcLhHHPtOHHѾHYH81_uHEH\HtwHm$HYHhH51ZH81^HmHPtHH.HѾHOYH81%_uff.fAUfHcATfHnHUSHHH)$~SHD$flHD$0HD$8)D$ HLAHMH/HtH HHH$HGHT$ HH4IHmj1PAPLD$ .H H5H<HCHt H<HHt H<H|$L$$4@HfL&H~HFI$L$$HH|$HHD$HGH_HH`HHHbuZHHxLl$H5I9ut6L;-/t-ff.fHlLHD ;H> u HA ?H-fInH=1HHt$ HEfHnfl)D$ [HmHHH-I9t&I}~LHzH}NHHHHH<$Ht H/ H|$Ht H/H|$H2H/(PVHHH[]A\A]XHuHH5jH8VfXHHLl$H5WI9ut L;-|H1bHjH=`1DHff.fH!HHjL hAH jH5 [H8S1XZXZH<$HtH/tgH|$HtH/tGH|$HtH/t/HWjH=D`1HHH[]A\A]Uʐ UfTf_GHH H9HT7HTHHFHHD$HFHHD$TfDsTfD_HcDHPT_GHH fD_fH0iH=_1mH-i H=]HH1 H=^HH#Ht$ HHHjL hJY^)H<$Ht H/ H|$Ht H/H|$H7H/-WS#fH@`HHHHHH@HH9~H]Ht+Ht=HtiHtQHt1HUHHm#HRHc]Hm]Dmt(HL ]EHH H붋]HcHtRmRcRH@`HHHHIHH@H9tzI]Ht#Ht5HtZHt@Ht)LTHImLQIc]A]AEHH A]AEHH HA]HcLHHPtOHHѾHQH81VuHEHTHtwHm$Hff.fAVAUATUSHHH=HGHH;u&1ҹHPHH[]A\A]A^ÐHHDHuOHHH}`L I9HID$H5HW@A$@HXHHJH~m1fHH9tWL;duH}`fLmhLepHEpE`Ht H/MtImtyMtI,$u L%ODQHt1HHH5adH81S1fDTLnOV@LNzNdMl$M~Y1 HI9tLI;|uHHI9HuL;%7Af1LmhLepE`HEpE1M9KtH9H|$2H|$Iff.@ATIHUHSHHWH=UPHIHRI$HEHtH[]A\OH[]A\'SHGtHHH[fDH@`HtoHHtcHHtYH@H;.tHHtfHHLH81RtH+t1@H1M@OHuHoH5a1H8MVHRHH5MH81Q@AWAVAUATUSHHHHMH5.L%HFI$QHEHH;a1H"LHHFL5L9OL=L-L9L9tH=#L9L9u L9Mff.H=9HH9GHL=yMIMIHHEHID$HHID$ Mt$(IMHHLxL` IELm(IEHeH+HH[]A\A]A^A_fH=yHGH9L5MI:MIHHEHIEH.HIE I$Me(MHHLpLh I,$LLyJ?@LIHHHLHXIIIHIMQHI,$MLJ@HMXfkMLH ,DImuLIfI.H^H=T~I,$uLI1CHhI8LLItTLk`fL{hLspHCpC`Mt ImMt I/MtI.uLIfDKHGL5IL=L-YLLHH1HE1H ;]L [H5%NH8R1H^kMZY1WfI,$u LqHI/uLbHM DHyxHH=]1@H`MHHGIHI]`L0I9HAIFHHS4@'A@HXHHJH1ff.HH9L;tuI]`fM}hMupIEpAE`Ht H+Mt I/M&I.LGHHIHHEHHLk`L0M9wMIFHIU4A@&A@ IXHtHJHS1ff.@HH93L;tuLFmLFH=QtIM=H9,HI9aHuL;5OHZH=P胴I,$L1~ENfL%IT$LLHIHL=H@HMRGLIMsfH=aIL-IIULGIHKL5H@HMFL3IIHLDLH&E"fHhD3L[D5LND!LAD#M~M1 HI9tvI;\uM~M1 HI9M;luLHI9HuL;5f1L{hLspC`HCpy1I9ItH97HHT$薩"HT$H1I9hItI9LHT$]HT$Hf.AWAVAUATUSHHHHMH5L%HFI$QHEHH;ga1HBHHFL5_L9OL=_L-L9L9tH=L9L9u L9Mff.H=H"H9GHL= MIjDIHHEHID$H]HID$ Mt$(I.DHHLxL` IELm(IEHeH+HH[]A\A]A^A_fH=HGH9NL59MICIHHEHIEHHIE I$Me(pCHHLpLh I,$LL@?@6CIHHHLHX'@IIHIMQHI,$ML@@HCXfC CH ,DImuL)@fI.H!UH=KI,$uL?1CH?8LLh@tTLk`fL{hLspHCpC`Mt ImMt I/MtI.uLv?fD+BHGL5 IL=L-ɼLL8?HHE1H SL =RH5DH8R1HTCZY1WfI,$u L>I/uL>M DHyxHH=,T1A@HCHH=IH]I]`L0I9HAIFHHS4@'A@HXHHJH1ff.HH9L;tuI]`fM}hMupIEpAE`Ht H+Mt I/M&I.Lv=HHIHHQIH2L=H@HM=LIMsfH=IL-IULE>IHL5H@HMK=LIV@HL;LH;"fH:3L:5L:!L:#M~M1 HI9tvI;\uM~M1 HI9M;luLHI9HuL;5 f1L{hLspC`HCpy1I9ItH97HHT$"HT$H1I9hItI9LHT$͟HT$Hf.AWAVAUIATUSHH$H$8$o$H$H1҉$H5փ$1L$D$H$$@$$($$$$$o$D$o$D$ o$D$0o$D$@o$ D$Po$0D$`o$@D$po$P$o$`$o$p$o$$o$$]HHeHH11o$H5$o$D$o$D$ o$D$0o$D$@o$D$Po$D$`o$D$po$ $o$0$o$@$o$P$o$`$iHHH6H=H$HDŽ$HGH;H$H@8HH|L$1HLH+IzHmMDŽ$hH|$HDŽ$pHDŽ$x%8HD$H'HH=H9xGL%MI$ID$H5&LHHHHMI,$H11ҿo$pH5$o$D$o$D$ o$D$0o$D$@o$D$Po$D$`o$D$po$$o$$o$$o$ $o$0$IH$HHHCH;.HDŽ$H4$H$H5QH9*H=ɲH9p*LXM6AMHM~'1DILH98H98HI9uL1IL$1ɺLI7H+IH$4Hm~H4q@L4 HG0H+K8HD$`E1AH$HD$Ƅ$Hl$Pff.Ha9DH=_?zMt1I/I$uHt$F8H$HtHHD$HHtjH\$`HtHH$HHt`Ht$PHtHH$HHt&HL[]A\A]A^A_ÐL2jH2fDH2fDH2fDƄ$E11E1H$E1AHD$`HD$PHD$HD$ 11Ht H+Mt"I+uLHL$0LD$;2HL$0LD$Mt I(Ht H)Ht HmHL$ HUHHD$HH@H13HHL$8LD$0L\$1HL$8LD$0L\$MDLHL$1HL$gfHx1`Hh1`=HL$HHD$HHH81H$1E1E1HD$`A HD$PƄ$H=H2H53IMHD$ E111E1E11HD$ E111E1E1E1E1HD$PA HD$`H$Ƅ$Ht$HHD$HHLt$MtAI,$u:LHL$HLD$@LL$8LT$0L\$ 0HL$HLD$@LL$8LT$0L\$MtI)tDMI*LHL$8LD$0L\$/HL$8LD$0L\$WfDLHL$@LD$8LT$0L\$/HL$@LD$8LT$0L\$fDHHL$HLD$@LL$8LT$0L\$_/L\$LT$0LL$8LD$@HL$H{4H-H=IA@HD$ E111E1E1E1HD$PE1E1A HD$`Ƅ$h@HkLcHEfHnI$$H+S%ID$H5'L)$H$AH9;H@8HH*H1LHIHt Hm:H $HHD$ HHH+MH\$LLT$HH3LT$HHH$HHD$ HH%HHI*H$H@m>H$H@H%HHH$CSHH HH $0$0H<$HHD$HH&HD$hHH@H9t5HXHHqH1 HH9H;TuDŽ$D$@Ã$p HnƄ$HDŽ$H$HD$DŽ$$vHD$`H$HDŽ$(HHSH9Xc1H-:Hy2HEHmH9E1fLHH)$"HD$PHm2H|$P&1A0D$pH$@E=$p-yHDŽ$ L$M5yHDŽ$H$HL|$hD$d$HDŽ$DŽ$4$8$PH$ D$؃$pE$tHE$HH|$?D$$xL$dHD$xL$L$HDŽ$E1!H$$H$H$lH$H$hH$H$xH$H$pH$D$: $$f*$f/rsf(IET$h$d\fH$xLZfH$pX\$ ZPt$pD$0\$ T$h\^XD$p$X$$$CHH9D$HEH$HH$L$L$H$P HcUHH$Hc<uIE$dLH$xH$pPZXD$0D$ AA D$E$$hf$/w $fZd$hA IWZf(HAL$ f(LP Avf/w  Cv]T$HfWYvYAKL$h|$@Y$lZYu~Qf(f(IEL\\$XY\$HL$hY\$8T$ \ffTZPL$hT$ ff.zDIEfLT$hL$ Z$dH$xH$p$xL$ T$h$/ f(Y$XD$0D$0fHH9Hu1H;¤$IGfL$ LZ$h$XXD$x$D$xA A7 $hf/`fIWZZ$|$hHc$dH$p0H fff.ffHZHYXH9u$XL$ L$Xf(s^$X^ff.$fZ$h\D$ ff(fWsfTfUfVmD$XYD$HYD$8Hc$d$IH$xAX$H4=f.f/f(f(\HYXZH9XHcff(ffHH<LZ/YZf/vXf(^\f(fTfZZZyf(\f(^Xf(fTfUffVZZZ0fDqD$$(X$\YD$8^T$HEd$$hf/DfZ$l$hBHD$xHDŽ$D$0 qfTqf.rRH$H$H q1ff.@HH9sTq.sH$@MB"H$ %IHVAH=LLD$U(LD$HII(;H;H$HL$L\$HDŽ$H8L\$HI+X;Ht111HH+r<1E1E1ADHH@H9(HqHD$H8)HH|$H5HGHH(IH\$HH$HM=(HHMH$#HD$H($IH)HD$HлIAHH9XF)HHD$H*HH|$LL$ H5QHGHH)LL$ H$H\$HHD$ HH<$)HHHSE1HI9D$'*fL$LT$HL$LL$)$&#LL$LT$HHHD$ H)HTHLLL$HHLT$HAHH$H$HUHA HH$H#LT$LL$HHMt I*$I)OH4$HHD$HH&HT$ HH$HHI,$"Hg)H$ 1DŽ$ HHHrH$H9&H$SRLf=L$ZYc+fo$ fH~)$fo$0HD$)$fo$@)$fo$P)$ fo$`)$0fo$p)$@fo$)$Pfo$)$`fo$)$pfo$)$fo$)$fo$)$fo$)$H(H+"H$(H$H9\$$|$@H$%kDŽ$d$XH1H3!IYl$Hf(f(XY\$8\ff.$$D$$^l$HEIGfL$ LZ$h$XXD$x$D$xDL$@E jIEL\T$XY,kT$hP(IELYZYD$hD$hP0$ZYD$XXD$hYD$8XD$xD$xt t$8f(XYD$0XD$xD$x$hf/fZ$d$hf.HL$tL$THLT$ [LT$ 0DŽ$_H|$LT$ 1LT$ I*`LSD fDHLT$ LT$ fDŽ$0$!D$E$0ft$xH*^t$h$f/=i"$4d$h1$8\f/G‰$4|$h]$8$8$9$4|7A=l$Hf/-,i=DŽ$4^-il$HH$ HBH9$H9H$ fD$hf/fZ$t$hHc$dH$pHffHZHYXH9uff.zIGL$ LZ$X$Xf(^]$;HqH@H9/$LMA%II@L$LH5»HH$L$IIHM9$IHHH5H9p$H~HD$PH%HH|$PH5PL$HGHH %L$IHt$PHH$HM$HHD$8LD$ L$L$LD$ HHD$P/%H I9@0&HD$PLLL\$ HL$HDŽ$H$迁L$L\$ HHT$PHH$HHI(H%H=лHL$tL$HHD$`HHH&HHHeI9CA&HD$`LLL\$ HHDŽ$H$L\$ H$Ht$`HHD$ HHEI+.H<$Y&H<$7HHuH-H9hTHf(f. 6ez.H+H|$heHGf(T$ PT$ _HHH<$HHD$`H'H+x D$8IH H|$`HHD$ .L\$ HH I+!H=,HrIH H+"I9kLL\$ L\$ $(f(f.%-dz3 I+LSfDCH߉D$ 'HHH=H|IHH+H=H HDŽ$L$LWL9LM9IXH LNM~.1ff.HTH9 L9 HI9uApIB8HH_H$1HHI,$H>3H+|$ @fIWZZ$|$hH<$HÉ$0H vHl4HZH53H8cnH%DŽ$0FDffIGZ$Zl$h @IGfL$ LZ$hDD$@$XXD$x$D$xE$5$hf/UfZ$d$heH$@$0hDHHC0DHD$XfH$$hf/fZ$t$hHD$ E111E1E11E1E1DL$IEP(ZIH+H5HHD$ LL$ HH$,I)5H"H={H9x!-L bM-IIALL$hLH52HH-LL$hHD$ H|$ !I)mH11ҿH$H5\H$o$$o$D$o$D$ o$D$0o$D$@o$ D$Po$0D$`o$@D$po$P$o$`$o$p$o$$o$$HIH H\$ HD$hHH$H$HHHDŽ$H$!zLL$hHH$,HHD$hHHI)H$HlnIHX HHD$ HHIALL$ LH5hHH-LL$ H$H$ I) H$HmIHHHD$ HHV!LLL$ HkLL$ HI.HHI,$LL$ H$!H$pI)!D$0IH~-H5XHHD$hsLL$hHHD$ i*I)3($:vIHG*H$$0f$H*^|$hL$HH.H5H L$HI<.H+)l$x^l$hf(L$l$hL$HH[0H5HLL$x LL$xHH$0H+>+~ :fInH$fIn~`LflP ~%1~ flfHn~(~-$)$\$ $)$fl)$)$)$)$@u tEH$R @u tEH|$ O @u tEAL$ @u tEH$K @)AAA,A.H$H$LL$xH$HPHF HDHT$ HBIQHTIT$HWH$ALL$xHH-H$HHD$xHH(H$HHD$xHH(Ht$ HHD$xHH(I)'I,$'H$HHD$ HH'H$H=HHDŽ$H$tH+I:'M,I,$=)HHWH9XV-L%>Mu1I$HI9D$C3H$fLH)$`tHI,$(H-Ht$PHIH2H+c)H=H$H9W3ID$t H;ш<3L+IM/I,$+)H$H=EHL$LL$ HDŽ$sLL$ II)(M2I,$)$ fDHKQ &E1 HiuLcH=QD u/H4$LHD$ LT$ M H ,E1fZ$t$h*LL$ L$LL$ HH$ I)$IH H5HHD$xLL$xHHD$ 'I)L!HD$hfIn~%fHn~ԞL]flP ~-~ufl~ fHn~$~5|)$$$)$D$ fl)$ )$)$0)$@)$P@u tEH$Q @u tEAT$ @u tEH$I @u tEH$AI @u#AtADEHL$ I @oAAA_%A%HT$hH$H$H|$ HRH@HDHSID$HDHFHWH$H$HFHTAIH$HT$hHHD$xHHH$HHD$hHHeI,$CH+"H$HHD$hHHH$HHD$hHHHT$ HHD$hHHH$H=HL$LL$ HDŽ$fLL$ HI)3H{#H+H H5rH9p"%HYH$HHzH9C?$H$fHH?)$^fIH+M"Ht$PLH$H"I,$"H=H$H9Y"H$H@t H;z7"H$HH"H$HHD$ HH!H$H= HHDŽ$H$}eH+IV!M!I,$ $$f/57J$41X$Pf/$G‰$4$_$P$PHKQKH-LL$L$HE1AH$HD$PHD$HD$`Ƅ$LLL$ELL$H3(HLD$ L$LD$ L$HH$1ɺHH=HДH5є̬HH1E1E1A,HD$P0LeH]I$HHmuHH$HL$HDŽ${cI,$HD$PtHJL8HH=HVH5WHD$H|$Hw1E1E1H$AHD$`HD$PƄ$HD$LH=4HHD$ 11E1E11E1E1HD$PL51wAHD$`H$Ƅ$INl$ MA3 H$@OHH+!E1A3LT$ fH$ {H=gHD$I,$H{v1Ƅ$1Ht$`AH4$Ht$ Ht$PHD$@HD$ 11E1E11E1DŽ$4HL$ ]L$ LJH=H_H5`L $'L $HD$H|$I,$HuE11E1H$1E11HD$`AHD$PHD$ Ƅ$HD$LL$ H$aHD$ 11E1Ƅ$E11E1HD$PL5#uAHD$`H=L $ϦL $HD$#HWBL9HG0I,$HtL$11H$E11AHD$`HD$PƄ$HD$MT$I\$IHI,$Iܽ1Ac$(H{H+tAHD$`H$HD$PƄ$HD$s$f(L\$ L\$ H1HD$PE1A*H$E1H$E1HD$`AHD$PƄ$HD$MH=zHˏH5̏觧IMH$1E1A&HD$`HD$PfRLt$1LD$1H$1E1E1HD$ E1E1A&HD$PHD$`YL$I\L=HH=HH5L$զL$HD$PH|$P\H$1E1A&HD$`+H=XIHHD$ ~L\$ HHH HD$P11A&HD$ H$HD$`nL$IHqf)$ LHL$ HL$ ?HHL$ HL$ H=L$FL$HD$PH$1A&HD$`&p.5H=31tTL|HoHff.ff.HH9AHuHqH9,Hff.ff.fHH9HuH9LIcLH$E1A&HD$`HD$P2M`IhI$fInHEl$PI()l$ tBL$foH$H)$s[I,$L$HtIL5fot$ LL$!L$Lt$1H\$1H$E1E11HD$ E1E1A&HD$PnIkI[HEfHnHl$`I+),$t=)$ff.fH$HZHmH$tILpfo4$)$HY1E1E1A&HD$`HD$P龾L*YH|$ LL$hLL$h6L-HKQ1HH$LL$ LL$ $H$M1E11E1E1A$HL$@LD$8LL$0LT$QH$@LT$LL$0E1LD$8HL$@ HD$ H$MH$L2H%|LLL$011LT$AL$LT$1HmLL$0E1H$HD$`HD$PƄ$HD$鼿LLL$ LT$LL$ LT$H$E1E1A Ƅ$H$HD$`HD$PHD$HD$PE1E1A(ϼH@`H HH H<$HH HlH9Cu&HlH+$0HH=HHuH$LL$ LL$ HD$ M11L\$hE1E1AHDŽ$$HL$@LL$8LT$0L\$0H$@sL\$Lt$L$LT$0LL$8HL$@L\$E1鹽DLUL LL$ *HLHIGLP H$11HH5K8H$o$$o$D$o$D$ o$D$0o$D$@o$ D$Po$0D$`o$@D$po$P$o$`$o$p$o$$o$$HIHH@L\$LH5HHL\$IMI+2D$0LD$QLD$HIH$H$HD$ H9 LD$0H11L$H56o$$o$D$o$D$ o$D$0o$D$@o$D$Po$D$`o$D$po$ $o$0$o$@$o$P$o$`$HL\$LD$0HHH@H5HHHL\$LD$0HD$ H|$ HmD$pLD$0L\$L\$LD$0HH$pLD$0L\$xL\$LD$0HILD$0L\$L\$LD$0HH,L@LX HD$ H]0HE(Le8ܶLLD$RLD$H>HLD$0L\$-LD$0L\$1E1A隷IL\$IX1A{I(HD$ AηHL$ HHD$HHE1AHD$ M11M飹Ht$ HHD$HHAI(HD$ 1E1AHT$ HHD$HH^AI(hI11AHLD$0L\$ALD$0L\$HD$ I,$l$ MA3H$@H$M1E1A$L\$[H$@L$L\$1HD$ HD$hML\$h11HD$ E1E1AHDŽ$L$l$ MHI,$LML\$h1A:ML\$h111E1E1AHD$ .LD$01A&L\$H$L\$1HD$PLD$0HD$ HD$`HD$ M1HHfL;L.LHD$L\$8$MAH$@Hl$P3fDL@HHIfInHHd$hL$HD$ HH)$t[HϺH$HL$ L$)$OHL$ L$HI(LHL$ THL$ wHL$ @H|$ fo$H$L$)$0OL$HL$ HHLHHD$ MH$LGlH>E1MIAL$11H=9H:H5;fIMH$lLL$hHD$ H1L$7HH$H$ME11E1E1AMH=bIwH$HMHHL\$L\$LD$0M1E1AHDŽ$HwmpL`FLS.LL\$A;L\$1E1HD$ бL1I鉵DŽ$4 HLD$0L\$L\$LD$0]HLD$0L\$L\$LD$0MI1AHDŽ$HLL$xLL$xXHLL$xuLL$x'HLL$x^LL$x у`HH81MH$E1H$AD р1ML\$hH$H$L$AH$D S1ML\$hH$H$L$AH$HCHKHHH+HD$ tHH$HL$hH|$ H$HDŽ$AHL$hIH)t&H\$ HHL$hHL$hH$HH=zHH1M1E1H$AH=zH0rH51rLH1IM1H|$ 11E1E1AID$IL$HHI,$HD$ tHH$HL$xH|$ H$HDŽ$@HL$xHH)t&Ld$ LHL$xHL$xH$HLIE1M11LL$ E11E1E1AHZHE1A3DLt$ 7AH$@@{LsL(HAH$@I,$t!E11E1E1Ld$ 1ZLոwHLHD$zHL$A3H$@vwHIaHL$H$@KHL$H)tl1E1E11HD$ A1靣H=5xxHHHL$1AH$@1HL$E1HT$ E1RH1A1E11HD$ E1)H=wH*oH5+ovHyA1H$@t鯡ff.@AW1IʹAVAUATIUSH~bOH$HHH-ufHnH-flfHnH)$fHn~&OH-fHnHflfHnH)$fHn~NH-fHnH-flfHnH)$fHn~NHflfHnH-@)$ ~NflfHnH)$0fHn~NHflfHnH-)$@~xNflfHn)$P~gNfl)$`~[Nfl)$p~ONHDŽ$fl)$~7Nfl)$~+Nfl)$~Nfl)$~Nfl)$Mt*MBM~!H)HhJcHI_<IuIXH AHhPHHzL H5_AE1AmH8AT1E1H;XZH$H$Hu ff.@HH9tH;HtH/utHH9uEuH $A8MtL;%OtAD$8MtL;-OtAE8DHE1H=@HL[]A\A]A^A_IHHH$HHH$HHH$HH$HH$HHH$HHH$HHH$HHH$HHH$HHH$HHH$HHH$HFxHH$HFpHH$HFhHH$xHF`HH$pHFXHH$hHFPHH$`HFHHH$XHF@HH$PHF8HH$HHF0HH$@HF(HH$8HF HH$0HFHH$(HFHH$ HFHH$HHH$L$pL5M1DŽ$ LHL9Q H$R<AULkHL$tAYAZfo$pfDo$fDo$fDo$fDo$fH~fDo$)$@fo$fo$D)$Pfo$fo$D)$`fo$ fo$0D)$pfo$)$D)$D)$)$)$)$)$)$)$)$D)$D)$D)$D)$D)$ )$0)$@)$P)$`)$p)$)$f$HH$L=UKL9 fD$f(f.zH$ 1LHDŽ$ L9Z H$R<AULgFL$q_AXfo$pfDo$fDo$fDo$fDo$fI~fDo$)$@fo$fo$D)$Pfo$fo$D)$`fo$ fo$0D)$pfo$)$D)$D)$)$)$)$)$)$)$)$D)$D)$D)$D)$D)$)$)$)$ )$0)$@)$P)$`MH$(L9$oD$f(f.%zH$0H$8HD$(HGFHGHHHHHUHcЉD$\H9|$\H$@L9D$f(f.-zIH$HL9D$ f(f.5zH$PH$XLDŽ$ HD$01HL9H$RAULD<L$oY^fo$pfI~)$@fo$)$Pfo$)$`fo$)$pfo$)$fo$)$fo$)$fo$)$fo$)$fo$)$fo$)$fo$ )$fo$0)$MH$`H;=^GH;=GL9$$H$hH$pHD$xHGUHGHnHkHHZHHcЉD$lH9|$lH$xHGHGH/QHHdGWHH HH jD$`|$`qH$L9D$8f(f.%zH$HGXHGHHoHHTHGWHH HcЉD$pH9FH|EH5H8D$pHL94$A}A%HHH$HHH$HHH$HHH$HHH$HHH$HHH$HHH$HHH$HHH$HHH$HHH$HFxHH$HFpHH$HFhHH$xHF`HH$pHFXHH$hHFPHH$`HFHHH$XHF@HH$PHF8HH$HHF0HH$@HF(HH$8HF HH$0HFHH$(HFHH$ HFHH$HHH$IBH$2HH:M1jJ4HLPAPI@jH xbI~DIIJ<uHBHL rAH HH8ATH51A[A\H$AE1E1Amx|$f(@|$f(@‰$Nfol$Wwt$ xgd$8H$D$`L$pL$pGHD$`H HHAH5H8D$`HyL94$A{AfDL$pCGD$\ZGD$`1GD$lGD$p|$pH$HGHGHH1 HH H_ GWHH HcЉD$hH9tFH@H5H8D$hHtL94$A~A|D$hH$H;=@H;=_@L9F$H$HGGHGHH0 HGWHH HH D$d|$dH$L9 D$@f(f.-z H$L9uD$Hf(f.5z H$HGhHGHHHCHh H GWHH HcЉD$tH9H?H5H8%D$tHtXL94$AAD‰${OL$@f(_\$Hf( D$tH$L9Yf(f.=z H$L9$%$D$Pf(f.Zz H$H;=C>H;==L9$$$9H$HL9Q$$f(f. z`H$HL9$$6$$f.l z H$HcHGHGHEH HH HGWHH HcAH9H<H5$$$H8谿$$$$A$$$$$HtrL94$AA@H$‰$H@H$  HmH$ Hff.E1HD$(H5YH9pt L9D HD$0H5.ZH9pt L9 HAWH$D$ $Pt$x$P$P$P$P$P$PHpfo$@$fo$P$fo$`$fo$p$fo$$fo$$fo$$fo$$fo$$ fo$$0fo$$@fo$$Pfo$$`fo$p$fo$$fo$$fo$$fo$$fo$$ fo$$0fo$$@fo$$Pfo$$`fo$$pfo$ $fo$0$fo$$fo$D$fo$D$ fo$D$0fo$D$@fo$D$Pfo$D$`fo$D$<$L$HD$p$Pfo$ $$,$fo$0$H$H$$fo$@$$$$fo$P$fo$`$$HIff.HH9tH;HtH/utHH9uH$L9tHA8.M9tAD$8M98AE8$ ImLWT$Pf(0OG D$dd謽HD$dH CHvH"7H5H8+6D$dHL94$AAfGD$dGD$h|$h:GD$t|$tHH AGWHH HcЉD$\H9sHn6H5wH8w肻D$\HGL94$AsAE1GWHH HHcЉD$\H9fDE1E1AAnHoL94$AoAE1E1L94$ApAE1GWHH HcЉD$lH9H5H5H8虸褺D$lHL94$AzA9GWHH HHcЉD$lH9fDI,$[LNH $HHD$HH HSHIm3L5&HcЉD$pH9kH&̹H)fDùHcЉD$hH9H蔹HfD苹HcЉD$tH9H!\H$fDG؉D$\$$$*$$$HcAH9 AH|$8HT$@D$lH/HT$8>HT$8HT$8HHuvHT$8ӧHT$8H^Hv"H5H8蟥HT$8>H@`HHHtyHHT$HHT$HHHtbHD$@H9Gu;HT$HH|$@k=H|$@HT$HD$pH/:HT$@yHT$@&HT$@HT$@HHuHT$@ HT$@HH!H5иH8դHT$@֦H|AWAVAUATIUSHH$H$$o$H$H1҉$0H5F$1L$D$H$$P$$8$$$$$o$D$o$D$ o$ D$0o$0D$@o$@D$Po$PD$`o$`D$po$p$o$$o$$o$$o$$FHHHH11o$H5$o$D$o$D$ o$D$0o$D$@o$D$Po$ D$`o$0D$po$@$o$P$o$`$o$p$o$$EH$HHQHD$H=m=HDŽ$H$H$ HGH;DH@8HHL$1HLH+IHL$HHD$HHMH|$ HDŽ$HDŽ$HDŽ$kHD$HmH<H /:H9HH-:HuHEHEH5mBHHHBHHHmH11ҿo$H5 $o$D$o$D$ o$D$0o$D$@o$D$Po$D$`o$D$po$$o$ $o$0$o$@$o$P$CH$HHXHCH;uHL$H5BHDŽ$H$H9&+H=H9+LXMjBMHM~-1ff.ILH9r9H9i9HI9uL1IOL$1ɺL艡H+IpHdcHPiH@HG0H+ 9Ƅ$1AHD$pHD$0HD$L|$HD$fHDH=ϩ Mt1I/I$uHt$F84H\$HtHHD$HHtqH\$pHtHHD$HHtfHt$0HtHHD$HHt+HL[]A\A]A^A_fDL bHfDHfDHfDƄ$E11E1HD$E1AHD$pHD$0HD$HD$@11Ht H+Mt,I*u&LHD$HHL$(LL$ mHD$HHL$(LL$ Mt I)Ht H(H\$@HtHHD$ HHH8H).H!HHD$`HL$HLL$(LT$ HD$`HL$HLL$(LT$ <LHD$(HL$ 趛HD$(HL$ XHHL$ 蓛HL$ IfHHL$ sHL$ Hf>HL$HHD$ HHH8Ƅ$1E1E1HD$AHD$pHD$0Of.H=;Hz4H5{4LHHZHD$@111E1E11HD$@111E1E1E1E1HD$0AHD$pHD$Ƅ$H|$H7Ht$ HH7HT$HtKHmuDHLD$hHD$PHL$`LL$HL\$(LT$ LD$hHD$PHL$`LL$HL\$(LT$ MtI(tRMKI+ALHD$`HL$HLL$(LT$ ÙHD$`HL$HLL$(LT$ f.LHD$PHL$`LL$HL\$(LT$ HD$PHL$`LL$HL\$(LT$ oLD$xHD$hHL$PLL$`L\$HLT$(HT$ 8HT$ LT$(L\$HLL$`HL$PHD$hLD$xKHH=9tHH@HD$@111E1E1E1HD$0E11AHD$pƄ$:fDHkLkHEfHnIED$H+p%IEH5;L)$H$AH9H@8HHH1LHIHt HmH|$HHD$0HH`H+6MH\$LL\$PHhL\$PHHD$HHD$0HH%HH:I+NHD$H@R?HD$H@H&bHHHL$AQHH HH $@$@HL$HHD$HHHD$@H3H@H9t?HXHHqHV1f.HH9;H;TuDŽ$D$`ǃ$ H.Ƅ$HDŽ$(H$ HD$DŽ$$fHD$pHD$HDŽ$8H1H b/H9H1L I/M2IH-I9A1fLLLL$PH)$LL$PHD$0I)!H|$0w1$H$P>$M$HDŽ$0-BLl$@HDŽ$51H$XD$d$PDŽ$DŽ$D$H$`H$0؃$$$tID$$LH|$ $$H$HDŽ$H$H$(HDŽ$1!H$$H$H$H$H$H$H$|H$H$H$$ $$f*$f/rmf(IFT$x$|\H$f(LH$XL$@PD$(L$@$T$x\^X$$X$$$CHH9D$ ID$H$LH$L$L$H$P IcT$I$H$Hc<uIF$|LH$H$PXD$(D$@AgA D$ED$E$f$$fTfUfVL$xA IUHAL$@LP f/w  ]D$PfW&YAH\$x|$`Y$Y~Mf(f(IFL\L$hYL$P\$xYL$HT$@\ffTP\$xT$@ff.zYIFT$xLf(\$@$|H$H$$\$@T$x$f(Y$XD$(D$(fHH94Hu1H;$"IEL$@L$$XX$$$A AoP$ff/}$IUt$x$|H$T1,3HfHHfHfYXfXH9u@tHc YXL$@L$h $h^$X^@$$H\L$@ff(fWfTfUfVt@D$hYD$PYD$HHc$|$IH$AX$H4-ff/f(wvf(\HYXH9zHcf(fHH<H'Yf/vXf(^\f(fTfDf(\f(^Xf(fTfUfVXf.$8X$$\YD$H^T$PD$$ff/}$l$xHDŽ$HDŽ$D$( fTf.rDH$H$(H1HH9sfTf.sH$PM`H$0#IHBH=+LLL$ sLL$ HII)<HY H$HL$LT$ HDŽ$H8LT$ HI*J<Ht111H 1H+<1E1E1A_H'H@H9%(H%HD$H)HH|$H5 /HGHH\)HHt$HHD$HH(HHH$词HD$H)7IH)HD$H ?%I@H$'H9H)H%HD$H*HH|$LD$0H5x+HGHHA*LD$0HD$H\$HHD$0HH|$,*HH_HpE1IH9E*fL$L\$0L$LD$)$ DLD$L\$0HH<*Hw*H *HLD$PHHL$HAHHD$H$ Hw+HA HH$(L׍L\$0HL$LD$PHMt I+~%I(Ht$HHD$HHH)hHmH,*H$@1DŽ$ HHHH$ H9\'H$SRL=L$&.Y^6,fo$@fH~)$fo$PHD$)$ fo$`)$0fo$p)$@fo$)$Pfo$)$`fo$)$pfo$)$fo$)$fo$)$fo$)$fo$)$fo$)$HK)H+q#H$HH$ H9\$$|$`H$(-%)DŽ$d$hH1H[I)YOl$Pf(f(XYL$H\ff.$$A$$^t$PIEL$@L$D$`$XX$$$3 #IFL\T$hYVT$xP(IFLYYD$xD$xP0YD$h$XD$xYD$HX$$t&|$Hf(XYD$(X$$$ff/$|$xnHL\$裆L\$fL\$0膆L\$0@DŽ$H|$L\$0QL\$0I+L:DH(#HL\$0L\$0fDŽ$@$$ $@f$H*^t$x$f/5#$Dd$x1$H\f/G‰$D$H|$x]$H$9$D|7A=l$Pf/-G=DŽ$D^-6l$PH$0HFH9$X;H$0$f$$fTfUfVL$x$|H$R&HfHHfHfYXfXH9utHc YXff.zIEL$@L$h$h^]$X$t$x@HAH@H96$H!HD$0H%HH|$0H5%HGHH%IH\$0HHD$HM$HH4HHH9X%HHD$0Hy&HH|$0LT$H5%HGHHz%LT$IH\$0HHD$HM%HH'D$HLL$PLT$葄LT$LL$PHHD$0%HI9A&HD$0LLLT$PHLL$HDŽ$H$LL$LT$PHHL$0HHD$HHI)H&H=%HLT$BLT$HHD$pHHH&HHH"I9B&HD$pLLLT$0HHDŽ$H$LT$0HD$H\$pHHD$0HHNI*7H|$'H|$HH}H-H9hH謄f(f. z6H+;H|$@HGf(T$0PT$0_HH H|$HȆHD$pH(H+ D$H腂IH(!H|$pHHD$0LT$0HH!I*!H=#HIH H+!I9j@ LLT$0调LT$0$8f(f.=z I*L H߉D$@߁HH%H=H4HH H+H=W#H 8#HDŽ$H$LWL9LM9IXH LNM~&1@HTH9 L9 HI9uAIB8HHH$1HHHm3Hc4H+\|$@}@$IU|$x~H|$趂H$@H zH4H)H5H82=H%DŽ$@J@$IEd$xIEL$@L$D$`$XX$$$D$E$f$$fTfUfVL$xmHD$@$@d@L}HC0<HD$hf.HH}GD$E+$f$$fTfUfVL$x@HD$@111E1E11E11:}L$IFP(}~IH,H5JHHD$xm{LD$xHHD$@+I(HH H9Hb+LiM.II@LD$xLH5HH-LD$xH$H$ "I( H11ҿH$H5zH$o$$o$D$o$D$ o$ D$0o$0D$@o$@D$Po$PD$`o$`D$po$p$o$$o$$o$$o$$@HIH!H$HD$xHH$H$HHHDŽ$H$LD$xHH$,HHD$xHH I(H$HIH HHD$xHH= I@LD$xLH5HHI.LD$xH$H$ I( H$HIH HHD$xHH(!LLD$xHF,LD$xHH.HH}xHmLD$xH$'H$ I(j'D$(1{IH#.H5HHD$x!xLD$xHH$-I(($IH0H$$@f$H*^d$xzL$HH /H5HwL$HH.H+*$^l$xf(L$l$xBzL$HH3H57HL$'wL$HH$3H+x+HD$@fHn~ fIn~%L ~fl~fHn~P $)$ ~-$fl)$$)$fl)$)$)$0@u tEH$S @u tEH$K @u tEM @u tEH$K @*AAAa-A1HD$@H$L$H$H$HPHCH$HDIPHCHTHUHV AL$HH"1HL$@HH$HH(H$HHD$@HH(H$HHD$@HHp(I(Y(HmA(H$HHD$@HH(H$H="HHDŽ$H$H+H'H/Hm:)HH@H9X0H-'H5HEHJH9EY4H$fHH)$HHm(H`0Ht$0HVzHH 2H+)H=H$ H91HEt H;l1HyIM3Hm)H$H=HL$LD$@HDŽ$8LD$@HI(/)H2Hm*$r2HKQo'1 LiuHkH=xru1Ht$HAHD$0vL\$0MIvH+E1nLLT$ksLT$2HLL$PLT$OsLL$PLT$L8sHLT$0&sLT$0HsLT$!H)D$0rfoD$0xL$0uL$0HHD$pE1E1AHD$0 LfHH9t.HuHhH9tLHL9tHuI96HWB'Hr1Ht$x uH_H=wxqHHD$xHH+uH Hm#Hr1AaRDHD$@111E1E11 @HLD$0qLD$0HqHHL$qHL$gfLHL$qHL$-fHhqFHH5{H8qD;rH11҉$H$H5HH$H$o$$o$D$o$D$ o$ D$0o$0D$@o$@D$Po$PD$`o$`D$po$p$o$$o$$o$$o$$oHHHAH@H5aHHHHD$xH|$xH+D$(qHHH$HH9X( HHD$xH$HHDŽ$H$H$H$HHt$xHHD$@HHHmH)HZHH9CzHr$f(f.-ͼz H+r$+ IFLP(pHH H5z HmHD$xH-"H+H HH9X#HH"HHCH5+HHHv"HHH+H11ҿo$H5ȸ$o$D$o$D$ o$ D$0o$0D$@o$@D$Po$PD$`o$`D$po$p$o$$o$$o$$o$$HHHH$HH$HHH$HDŽ$;H$H"Hm\H+EH$HHHHEHD$@HHE(HCH5HHH="H$H$@H+H$H!HHHEHD$@HHEH5 HHt$@HHHt$@HkHmH$H$H+TD$(mHH"H5 HjHH H+$}HH $@f$H*^|$@ImIH'H5F HH$6jL$HH$W'I(<$^D$@lIH&H5 HH$iL$HHD$@4"I(h $lIH"H5 HH$iL$HH$'I("HD$xfHn~vfHn~%1LflP ~-:~ fl~ 6fHn~A$~5)$0$L$@)$ $fl)$@)$)$P)$`)$p@u tEH$V @u tEU @u tEH$I @u tEHL$@AI @u#AtADEH$I @ AAA&At&Ht$xH$H$HVH@HDHSHEHDHGH|$@HGH$HWH$HTAIH%Ht$xHH$HH9 H$HHD$xHH HmH+H$HHD$xHHH\$@HHD$xHH]H$HHD$@HH)H$H=" HL$LD$@HDŽ$zLD$@HI(H+%H+HHWH9X}%H>HD$@H.%HHD$@H8H9Xo%H|$@H$fH?)$HH\$@HHD$xHHH~#Ht$0H4lHD$@HH#Hm !H=|H$ H9g#HD$@H@t H;BH#Ht$@kHH#Ht$@HHD$xHH!"H$H= HHDŽ$H$H+H!HL"Hm $d$f/=$D1X$`f/$G‰$D$_$`$`fDHKQpH0eHLT$eLT$HE1AeƄ$HD$HD$0HD$pLLD$0dLD$0HL$fHdHLL$PLT$dLL$PLT$HdH$1ɺ}gH]H=HH5KIMKHD$01E1A{1E1E1AhIiIYHEHI)uLcH$HH$HDŽ$HmHD$0tIHcHdH=UHH5HD$H|$*H1E1E1HD$AHD$pHD$0Ƅ$HD$H=`IHD$@11E1E11E1E1HD$0HAHD$pHD$Ƅ$_hH|$@MaH$PaHH+uHb1A`H$@H=,HD$Hm"HE1E11LT$@AL\$pL\$0L\$Ƅ$HD$HD$@11E1E11E1DŽ$DxHL$0aL$0HaH=vH'LD$H5#LD$HD$H|$HmH5&E11E1HD$E111HD$pAHD$0HD$@Ƅ$Ht$3ffLD$0HD$HD$@11E1Ƅ$E11E1HD$0HAHD$pgHWB]L9{HG0H=kLD$!LD$HD$HmLH50LT$11HD$E1AHD$pHD$0HD$@Ƅ$Ht$@L]H]IHHmWHA1Aj$8f(H_(HƄ$AHD$pHD$0HD$H\$f{$f(LT$0ObLT$0Hc1HD$0E1AH$ E1HD$E1HD$pAHD$0Ƅ$HD$|H=HH5 HD$0H|$0U1E1E1AHD$HD$p,fL^EHHT$11HD$@E111HD$pE1E1AHD$Ht$HD$0cIH_^qHHD$0M^LT$0H=HLT$H5$LT$HD$0H|$0HD$1E1AHD$pAH=` HD$0f1HH HD$11AHD$0HD$pHD$@bLT$I~Hr])f)$@HHL$@O]HL$@HHL$@8]HL$@HmMDt$@ D[H$P&\H\$A:f1HD$1AHD$p!p(H=ea1覽H=oLT$% LT$HD$0|L\H\HfHH9=HuH$H9(Hff.ff.fHH9HuH9LIMiIiIEfInHEd$0I))d$PtHLT$foff.H$H)$ImLT$HtI9L[fo|$PLLT$o[LT$HD$E1AHD$pHD$0锾L7[9HT$1H\$1HD$@E111HD$0E1E1AHD$IjIZHEfHnH|$pI*fo)|$uLZfod$H$H)$HmHD$tIHZLvZH$LD$x_ZLD$x1E1E1AHD$pHD$0酽H(ZHKQC1H$LD$xYLD$xHY9MH$D$1E11E11ABDLD$PHD$`HL$HLL$(L\$ mXH$PXL\$ LL$(E1HL$HHD$`LD$P"HgYMH$1D$zL@Y=HLD$H1AHD$(L\$ YLT$E11H L\$ HD$HD$pLD$HHL$HL$(HD$0HD$@Ƅ$ξHLD$0L\$XLD$0L\$H$LD$xXLD$xH\$E1Ƅ$E1HD$AHD$pHD$0HD$霻HD$0E1E1A邻DH@`H HH H|$HH HH9Cu&HSH+$@HWHmHHu.HWHWHD$@M11LT$xE1E1A1HDŽ$$LD$PHD$`HL$HL\$(LT$ VH$PWVLT$ HT$L$L\$(HL$HHD$`LT$E1LD$P阼@HVIGLP H$11HH5tH$o$$o$D$o$D$ o$ D$0o$0D$@o$@D$Po$PD$`o$`D$po$p$o$$o$$o$$o$$5HIHH@LT$ LH5HHLT$ IMI*bD$(LL$ jWLL$ HIH$H$ H9?LL$(H11L$o$H5$o$D$o$D$ o$D$0o$D$@o$D$Po$ D$`o$0D$po$@$o$P$o$`$o$p$o$$H$HHHLT$ LL$(H@H5HHLT$ LL$(HHH\$@HHD$ HH$HL$@LL$(LT$ ULT$ LL$(HHL$@H!$HL$@LL$(xLT$ ULT$ LL$(HHL$@HHL$@LL$(LT$ HD$HULT$ LL$(HHL$@HT$HHLHLP HH(HX0HP8酵LLL$ ;SLL$ HHHL$HLL$(LT$ SHL$HLL$(LT$ 1E1AfB1Af5XLT$ II)1Ag閶H)1AiHD$@M11I餸H){AjI)11E1AhյH)xAfH|$@LL$(LT$ yWLL$(LT$ HI)LL$@1AhHD$@xLQHQLD$x^HD$@ME1A-$LT$ PH$PPLT$ 11L$H$QMDt$@;THHmLH_Q?VHD$xLHD$@M11LT$xE11A1HDŽ$L$LT$xMA-HD$@(H=HH5IMM1D$HLL$(ALT$ PLT$ 11HD$LL$(HD$0HD$pHD$@LgPALZPLHD$ HPLT$ FHD$@M11LT$x1E1E1A-HD$@M1D$+HOmLO$MA-NH$PNHD$L|$齱HOH$OCL@HHHIfInHHt$x)$HD$@HHtcHH$L$HL$@H$)$IHL$@L$HI(LHL$@OHL$@HHL$@L$NH|$@fo$H$H$)$ӷL$HL$@HHNHNL$HwNdSLD$xH$IQHH1IMA-HD$@1H$]HD$@ME1A1HDŽ$7HLL$(LT$ MLT$ LL$(7HMMH$D$H=x3ILLT$ AhMLT$ 11E1HM(HrMHeMLXMHLD$@FMLD$@yHLD$@/MLD$@EHL$ML$ HLL$(LT$ HD$@LLT$ LL$(HT$@ZHHT$@LLT$ LL$(HT$@gDŽ$D6HLL$s уHD$@IMA1HDŽ$L1ILHL$(AhLT$ 0LLL$@LT$ 1HD$@HL$(釯H L8HKLK%QHQLD$xH$H=:HH0E1E1LT$xMLD$@11E1L$E11A1H=5HH5bHHMH$D$E11E11ACtMH$1D$HKnHD$@M11LT$xL$E1HDŽ$A2nLJHLD$@JLD$@dM1D$1E11E1ABMDt$@i1E1AaɭLlJOH$MHHH5^H8JILT$xM1HD$@1E1A1HDŽ$LHTeME1E1ACH$H$D$MH$E11H$E1D$ACHLD$ 1AxI111Ht$pLD$ E1E1Ht$0E11H5HL$1H|$@Ƅ$Ht$; рH$PM3HHIHHJH$11LT$xMHt$@11E1H$L$1A2IE1LT$xM1L$L$A3H$HzH@LmH HLD$@[HLD$@HLD$xDHLD$xHLD$x-HLD$x[HLD$xHLD$x'HLD$xGLD$xHLD$xGLD$xHL$GL$D уLGHH5LH8;H$MpHH$0A HH HH5FH9pT L-MIfLLT$ H?H$)$LT$ HHNI*7Ht$0HvLIHHmH5hLLT$ ELT$ HHI*HfHn~fHn~ ?LH$E flfl)$)$@B) HSH$HEHT'AIHH+HmrH=uE1H$HL$L$ˮI,$H&HH+DRDH$PDxHhEE1M1D$Ll$@1E11E1E1A@xH.EhML$E1H$H$D$AD:ML$E1H$H$D$A@ 0ME1E1ADH$H$D$H=HH5oHHR1MAF1Hl$@1E11D$E1E11wE1LT$xM1Ld$@L$L$H$A3HCHCE1LT$xM1Ll$@L$A2L$H$HM1E1LT$xLt$@11A2H$L$$MA4;BH$P~BHH5HH8CH$ME11A5U$MA5AH$P!B=Ht$@EHHrEIw1LT$xMA3H$L$H$H$<1ME1A5H$I1M1H\$@11E1D$E1AF1LT$xMA/H$L$H$H$D w$MA/@H$PA.D р=H= NHD$@H|$@1ME11H$A5H=HH5HD$@HXHH@HHD$xHHH$HHuHKAH|$xH$H$HDŽ$@H+HtHD$xHD$@DHAE1M1D$Ld$@1E11E1E1AEE1M1D$L\$@1E11E1AFHEHMHHHmHD$@tNH$H$H|$@H$HDŽ$xH$HH)t,Hl$@fHH$-@H$H$H@H=BH|$`$$f$$fTfUfVL$xnH?DAZ^>H$P>H\$Lo?Hb?HU?iDAZ>H$PU>Hmt 1E11E1HD$@1H?ָL>SHHD$ >LT$ DA[=H$P=wL>DLT$ b=H$P=LT$ I*tnE1E1E11LD$@1A[鿡H=IMDLT$ A[=H$PE=E1LT$ 1LL$@1E1sLA[=1E1E1HT$@11JH=HH5IxDA[H@`HHHtyHHT$HHT$HHHtbHD$@H9Gu;HT$HH|$@˥H|$@HT$HD$pH/:HT$@ HT$@&HT$@u[HT$@HHuHT$@iHT$@HH H50!H85 HT$@6H|HHShared Cython type %.200s is not a type objectShared Cython type %.200s has the wrong size, try recompilingbase class '%.200s' is not a heap typeextension type '%.200s' has no __dict__ slot, but base type '%.200s' has: either add 'cdef dict __dict__' to the extension type or add '__slots__ = [...]' to the base type__int__ returned non-int (type %.200s). The ability to return an instance of a strict subclass of int is deprecated, and may be removed in a future version of Python.__int__ returned non-int (type %.200s)Interpreter change detected - this module can only be loaded into one interpreter per process.unbound method %.200S() needs an argument__annotations__ must be set to a dict object__kwdefaults__ must be set to a dict objectchanges to cyfunction.__kwdefaults__ will not currently affect the values used in function calls__defaults__ must be set to a tuple objectchanges to cyfunction.__defaults__ will not currently affect the values used in function calls__qualname__ must be set to a string object__name__ must be set to a string objectFailed to import '%.20s.decompress' - cannot initialise module strings. String compression was configured with the C macro 'CYTHON_COMPRESS_STRINGS=%d'.%.200s.%.200s is not a type object%.200s.%.200s size changed, may indicate binary incompatibility. Expected %zd from C header, got %zd from PyObject%.200s does not export expected C %.8s %.200sC %.8s %.200s.%.200s has wrong signature (expected %.500s, got %.500s)Incompatible checksums (0x%x vs (0x%x, 0x%x, 0x%x) = (%s))Acquisition count is %d (line %d)invalid vtable found for imported typemultiple bases have vtable conflict: '%.200s' and '%.200s' while calling a Python objectNULL result without error in PyObject_Call%.200s() takes %.8s %zd positional argument%.1s (%zd given)sklearn/linear_model/_sgd_fast.pyxsklearn.linear_model._sgd_fast.SquaredEpsilonInsensitive.__reduce__sklearn.linear_model._sgd_fast.EpsilonInsensitive.__reduce__sklearn.linear_model._sgd_fast.SquaredHinge.__reduce__sklearn.linear_model._sgd_fast.Hinge.__reduce__sklearn.linear_model._sgd_fast.ModifiedHuber.__reduce__sklearn.linear_model._sgd_fast.Classification.py_dlosssklearn.linear_model._sgd_fast.Classification.py_losssklearn.linear_model._sgd_fast.Regression.py_dlosssklearn.linear_model._sgd_fast.Regression.py_losssklearn.linear_model._sgd_fast.SquaredHinge.__init__sklearn.linear_model._sgd_fast.Hinge.__init__sklearn.linear_model._sgd_fast.SquaredEpsilonInsensitive.__init__sklearn.linear_model._sgd_fast.EpsilonInsensitive.__init__cannot pass None into a C function argument that is declared 'not None'sklearn.linear_model._sgd_fast.__pyx_unpickle_Classification__set_statesklearn.linear_model._sgd_fast.Classification.__setstate_cython__sklearn.linear_model._sgd_fast.__pyx_unpickle_Regression__set_statesklearn.linear_model._sgd_fast.Regression.__setstate_cython__can't convert negative value to unsigned intvalue too large to convert to unsigned intvalue too large to convert to intsklearn.linear_model._sgd_fast.__pyx_unpickle_Classificationsklearn.linear_model._sgd_fast.__pyx_unpickle_Regressionsklearn.linear_model._sgd_fast.Regression.__reduce_cython__sklearn.linear_model._sgd_fast.Classification.__reduce_cython__sklearn.linear_model._sgd_fast._plain_sgd32sklearn.linear_model._sgd_fast._plain_sgd64Unable to initialize pickling for %.200sModule '_sgd_fast' has already been imported. Re-initialisation is not supported.sklearn.linear_model._sgd_fastcompile time Python version %d.%d of module '%.100s' %s runtime version %d.%dinit sklearn.linear_model._sgd_fast_cython_3_2_2.cython_function_or_method_cython_3_2_2._common_types_metatypesklearn.linear_model._sgd_fast.SquaredEpsilonInsensitiveEpsilon-Insensitive loss. loss = max(0, |y - p| - epsilon)^2 sklearn.linear_model._sgd_fast.EpsilonInsensitiveEpsilon-Insensitive loss (used by SVR). loss = max(0, |y - p| - epsilon) sklearn.linear_model._sgd_fast.SquaredHingeSquared Hinge loss for binary classification tasks with y in {-1,1} Parameters ---------- threshold : float > 0.0 Margin threshold. When threshold=1.0, one gets the loss used by (quadratically penalized) SVM. sklearn.linear_model._sgd_fast.HingeHinge loss for binary classification tasks with y in {-1,1} Parameters ---------- threshold : float > 0.0 Margin threshold. When threshold=1.0, one gets the loss used by SVM. When threshold=0.0, one gets the loss used by the Perceptron. sklearn.linear_model._sgd_fast.ModifiedHuberModified Huber loss for binary classification with y in {-1, 1} This is equivalent to quadratically smoothed SVM with gamma = 2. See T. Zhang 'Solving Large Scale Linear Prediction Problems Using Stochastic Gradient Descent', ICML'04. sklearn.linear_model._sgd_fast.ClassificationBase class for loss functions for classificationsklearn.linear_model._sgd_fast.RegressionBase class for loss functions for regressionneeds an argument%.200s() %stakes no keyword argumentstakes exactly one argument%.200s() %s (%zd given)_cython_3_2_2takes no argumentsBad call flags for CyFunction__loader__loader__file__origin__package__parent__path__submodule_search_locationskeywords must be stringsdecompresszlib__pyx_capi__functionpicklePickleErrorfloat division by zeroexactly__reduce__an integer is requiredpy_dlosspy_lossat mostat least__init____setstate_cython__tupleExpected %s, got %.200s__pyx_unpickle_Classification__pyx_state__pyx_unpickle_Regressionname '%U' is not defined__reduce_cython___plain_sgd32_plain_sgd64builtinscython_runtime__builtins__does not matchsklearn._loss._lossCyLossFunctionCyHalfSquaredErrorCyAbsoluteErrorCyPinballLossCyHuberLossCyHalfPoissonLossCyHalfGammaLossCyHalfTweedieLossCyHalfTweedieLossIdentityCyHalfBinomialLossCyExponentialLossCyHalfMultinomialLosssklearn.utils._weight_vectorWeightVector64WeightVector32sklearn.utils._seq_datasetSequentialDataset64ArrayDataset64CSRDataset64SequentialDataset32ArrayDataset32CSRDataset32sklearn._cyutilityarraymemoryviewcannot import name %S__module____dictoffset____vectorcalloffset____weaklistoffset__func_doc__doc__func_name__name____qualname__func_dict__dict__func_globals__globals__func_closure__closure__func_code__code__func_defaults__defaults____kwdefaults____annotations___is_coroutine_sgd_fastconst floatconst uint8_tconst doublexKwV 4:h(2)!:Iv{dQ#Kte2K-RK-eZ'}JW[:gf9 _}JaRQ6I74'd]6MjWPI0jK}T֥YnL3Lj*篙Pͦd*Jö*L¬ZVY&I3TRSc* UUăӖh״:֩n椧O垚v UUIe햴f2Y-k bZ 0P4TnxJl4$G1m (|j\)5)Tv$fRUV; ʶq l_R3U_#NU%aݺ0;fVe]v)\[SCDiu dޡa2ۗjC+ ]2P#y=7OLPaL5!߶[0 ЪMAEv?e{ = Bj!í <)Yeif udHJA!KMeÈ$/㹨(ĘRkB˲ Aiq+ɦXD$BTt^QYۢz8̴,4Ӛ )'V/T&Q fTX?*r{ڰTdQ'њY FJ`5t *D$ʺ 7đkNِk< $H,Tp_3M25q50&64:˂6AKG / 7- J1 K25NOBET9 hLc_K৆ai 4տ9dع f ZO<3e Pi+>ɌIP:Tr}'VdnSu DL4BթۂГl[mM]Lid!оa9l^iD)A(Ц0,Ӊ]껸 |JripIhkGFŔ`8@q9b$5Zh5Ƽ)+/Aq&ȡab5 dv`ڡƁ!i-?P)uZۆ% g/ԛ}v(T K18_ßAX$[}Q)1^^ޛ77Nٶvij~D;<ŗ@w\3~+AFH&r+C\uNIgXnF(ȽfwJҴevVg"4wNIiEuL\UX6xe8&qXM+ -zBm?`e`LB1m.eQ9rqEuS+ ,,F=#*%&~QMYU`*liVl" Te{@$+Zq!ً\x!|2|s⾑.xa3mFU8C)Ê'Q.֢=tꃉ8I, C8H Cxe֨unyuQUW =st@hK]<śq 1.7Ugb)B {E¯9  D5 l1oW~_-*_A91& ZH_I["F%>,Ǣ# "R!] t E;v+,w5FV|iMr^ګUo)lWv 1R q;PGa/9i,%䱩 P7nB7wufsd"oɮg`<ĠVtP7p$+Hy9z13NgF~=S;{i.Eٳ^)ʞɌ8M']9\e/ENeypӴ7e%Wx@[e?\Px1z WL{G#L6->=gNˮ]R46uwu Ap>xV:Kؔ[re\m/GcLt䤫c\[yD-۝N-x+^_KSn{,;RGF {`!(ǶT79T?.~+A%7 .ɝrlp/ 2uR 1m鬷O8`M f{xHiw!ƻ'? o)sK)Nyyuoƛ`iE^AEZD[{rR1,Ŀ>u ҪMn˝ę/<dg.؀8v%h] n+7GWpu_ Ά ҝ.CL0|<{;xQPp6XuB|,S#' fgA^6;2=wNG Rg-uH /Kh#C h@r0ZĂ}MK=7Sh5_>U٭@o5 cfw[ q?NI, a;iEQ˸Yx(cc} 4N"$v6hzǶ~Bez7oe}x@p$؞ 2ݑ?84fC㌗@Cx;4Ј.VK>~/+c`&o}|{ˁ[N@{]k\90$g.0WϻW0Doӽ\z}>sdbA"eLkxU B}v) NFۖ>ywcQwxʣr > GwoS>Äy)l/>  x "  " "  $                      E455 YY  97xhXH8(p]J7$3$ԨĨtdTD4!է§vcP__pyx_fatalerror00010203040506070809101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899SGD for generic loss functions and penalties with optional averaging Parameters ---------- weights : ndarray[float, ndim=1] The allocated vector of weights. intercept : double The initial intercept. average_weights : ndarray[float, ndim=1] The average weights as computed for ASGD. Should be None if average is 0. average_intercept : double The average intercept for ASGD. Should be 0 if average is 0. loss : CyLossFunction A concrete ``CyLossFunction`` object. penalty_type : int The penalty 2 for L2, 1 for L1, and 3 for Elastic-Net. alpha : float The regularization parameter. l1_ratio : float The Elastic Net mixing parameter, with 0 <= l1_ratio <= 1. l1_ratio=0 corresponds to L2 penalty, l1_ratio=1 to L1. dataset : SequentialDataset A concrete ``SequentialDataset`` object. validation_mask : ndarray[uint8_t, ndim=1] Equal to True on the validation set. early_stopping : boolean Whether to use a stopping criterion based on the validation set. validation_score_cb : callable A callable to compute a validation score given the current coefficients and intercept values. Used only if early_stopping is True. n_iter_no_change : int Number of iteration with no improvement to wait before stopping. max_iter : int The maximum number of iterations (epochs). tol: double The tolerance for the stopping criterion. fit_intercept : int Whether or not to fit the intercept (1 or 0). verbose : int Print verbose output; 0 for quite. shuffle : boolean Whether to shuffle the training data before each epoch. weight_pos : float The weight of the positive class. weight_neg : float The weight of the negative class. seed : uint32_t Seed of the pseudorandom number generator used to shuffle the data. learning_rate : int The learning rate: (1) constant, eta = eta0 (2) optimal, eta = 1.0/(alpha * t). (3) inverse scaling, eta = eta0 / pow(t, power_t) (4) adaptive decrease (5) Passive Aggressive-I, eta = min(eta0, loss/norm(x)**2), see [1] (6) Passive Aggressive-II, eta = 1.0 / (norm(x)**2 + 0.5/eta0), see [1] eta0 : double The initial learning rate. For PA-1 (`learning_rate=PA1`) and PA-II (`PA2`), it specifies the aggressiveness parameter for the passive-agressive algorithm, see [1] where it is called C: - For PA-I it is the maximum step size. - For PA-II it regularizes the step size (the smaller `eta0` the more it regularizes). As a general rule-of-thumb for PA, `eta0` should be small when the data is noisy. power_t : double The exponent for inverse scaling learning rate. one_class : boolean Whether to solve the One-Class SVM optimization problem. t : double Initial state of the learning rate. This value is equal to the iteration count except when the learning rate is set to `optimal`. Default: 1.0. average : int The number of iterations before averaging starts. average=1 is equivalent to averaging for all iterations. Returns ------- weights : array, shape=[n_features] The fitted weight vector. intercept : float The fitted intercept term. average_weights : array shape=[n_features] The averaged weights across iterations. Values are valid only if average > 0. average_intercept : float The averaged intercept across iterations. Values are valid only if average > 0. n_iter_ : int The actual number of iter (epochs). References ---------- .. [1] Online Passive-Aggressive Algorithms K. Crammer, O. Dekel, J. Keshat, S. Shalev-Shwartz, Y. Singer - JMLR (2006) SGD for generic loss functions and penalties with optional averaging Parameters ---------- weights : ndarray[double, ndim=1] The allocated vector of weights. intercept : double The initial intercept. average_weights : ndarray[double, ndim=1] The average weights as computed for ASGD. Should be None if average is 0. average_intercept : double The average intercept for ASGD. Should be 0 if average is 0. loss : CyLossFunction A concrete ``CyLossFunction`` object. penalty_type : int The penalty 2 for L2, 1 for L1, and 3 for Elastic-Net. alpha : float The regularization parameter. l1_ratio : float The Elastic Net mixing parameter, with 0 <= l1_ratio <= 1. l1_ratio=0 corresponds to L2 penalty, l1_ratio=1 to L1. dataset : SequentialDataset A concrete ``SequentialDataset`` object. validation_mask : ndarray[uint8_t, ndim=1] Equal to True on the validation set. early_stopping : boolean Whether to use a stopping criterion based on the validation set. validation_score_cb : callable A callable to compute a validation score given the current coefficients and intercept values. Used only if early_stopping is True. n_iter_no_change : int Number of iteration with no improvement to wait before stopping. max_iter : int The maximum number of iterations (epochs). tol: double The tolerance for the stopping criterion. fit_intercept : int Whether or not to fit the intercept (1 or 0). verbose : int Print verbose output; 0 for quite. shuffle : boolean Whether to shuffle the training data before each epoch. weight_pos : float The weight of the positive class. weight_neg : float The weight of the negative class. seed : uint32_t Seed of the pseudorandom number generator used to shuffle the data. learning_rate : int The learning rate: (1) constant, eta = eta0 (2) optimal, eta = 1.0/(alpha * t). (3) inverse scaling, eta = eta0 / pow(t, power_t) (4) adaptive decrease (5) Passive Aggressive-I, eta = min(eta0, loss/norm(x)**2), see [1] (6) Passive Aggressive-II, eta = 1.0 / (norm(x)**2 + 0.5/eta0), see [1] eta0 : double The initial learning rate. For PA-1 (`learning_rate=PA1`) and PA-II (`PA2`), it specifies the aggressiveness parameter for the passive-agressive algorithm, see [1] where it is called C: - For PA-I it is the maximum step size. - For PA-II it regularizes the step size (the smaller `eta0` the more it regularizes). As a general rule-of-thumb for PA, `eta0` should be small when the data is noisy. power_t : double The exponent for inverse scaling learning rate. one_class : boolean Whether to solve the One-Class SVM optimization problem. t : double Initial state of the learning rate. This value is equal to the iteration count except when the learning rate is set to `optimal`. Default: 1.0. average : int The number of iterations before averaging starts. average=1 is equivalent to averaging for all iterations. Returns ------- weights : array, shape=[n_features] The fitted weight vector. intercept : float The fitted intercept term. average_weights : array shape=[n_features] The averaged weights across iterations. Values are valid only if average > 0. average_intercept : float The averaged intercept across iterations. Values are valid only if average > 0. n_iter_ : int The actual number of iter (epochs). References ---------- .. [1] Online Passive-Aggressive Algorithms K. Crammer, O. Dekel, J. Keshat, S. Shalev-Shwartz, Y. Singer - JMLR (2006) Python version of `dloss` for testing only.Python version of `loss` for testing only.Python version of `dloss` for testing only. Pytest needs a python function and can't use cdef functions. Parameters ---------- p : double The prediction, `p = w^T x`. y : double The true value (aka target). Returns ------- double The derivative of the loss function with regards to `p`. Python version of `loss` for testing only. Pytest needs a python function and can't use cdef functions. Parameters ---------- p : double The prediction, `p = w^T x + intercept`. y : double The true value (aka target). Returns ------- double The loss evaluated at `p` and `y`. SGD implementation?@C@m?mBư>@;r ̷$P ( 3t O J$plL,|@Th |< <$l<Tl,<L,@Tt<l,L4XL < ,T  \  D < p   T L| \ 4 \P p,ld  0$ &L' (d)*L,/ 3l6`9<@@ C F4 IlJKlMLO\X8acc4dPlnw\ slzRx $PFJ w?;*3$"D=XMl'%4-P\9%M    4 H\&Dat&Da&Da(&Da@&DaX&Dap&Da  ,@\ cDXAV \Dt H H H lht pEQc D L/Q]@ AGG0g GAJ d DAG Z CAJ P<4d BDA _ ABE AB |Ay F g I (lyAAD T DAA 4xBED D(A0_(C ABB$AdN <jA C H H ` )Act|oBEE I(K0J8D`hFpHxPLGAADAABL`m8D0A(B BBB$jAIC ZAA8ɲyBEE G(A0[(C BBB(XRBAD GAB4BDD  GBG AAB([d vf `H [ A UxBBB B(D0D8Gc 8G0A(B BBBF K 8D0A(B BBBH Q 8F0A(B BBBH dX}DE G (ADD O AAF (@ADD O AAF hDo E ohDo E oHOBNB B(A0A8G'8A0A(B BBB(dI AAG= AAHKBEE E(D0D8DP 8D0A(B BBB@BHD D(B08H@U8A0T(D ABBT KDBEE L(D0A8B@HIPKHB@m8A0A(B BBB@xBKA n(_0[(A F ABH QDB$puD n N r F  L PBAA G0q  AABH V  FABI O  AABE `8 'BSB B(A0A8G` 8D0A(B BBBD a 8A0A(B BBBE  PAG\ NBGA D0n  CABF R  CABH J  HABC L  CABF H BBB E(D0A8DPy8D0A(B BBB$d `AAA ZAAH ?BBB B(D0C8D`!8A0A(B BBB 4  0 }Th t}Th 0 Dr J H H T `h d PBIB B(A0A8D` 8D0A(B BBBK ` 8A0A(B BBBF p P BKF A(G0E (I DBBG  (D ABBC Q (I DBBN (I ABB(@ [d vf `H [ A U(l yAAD T DAA d T;BBB B(D0D8GP 8A0A(B BBBA  8A0A(B BBBL @ ,8AAD n AAJ n AAH K(\0O(C @D (8AAD n AAJ n AAH K(\0O(C @ $8AAD n AAJ n AAH K(\0O(C @ 8AAD n AAJ n AAH K(\0O(C ,DC I K\ OAF C 4@AAG Q DAA L DAG lx4WAOL`ohPpCxBO` DAI [hfpHhA`~ DAC Y DAB hEpNhA`l$WAOL`ohPpCxBO` DAJ [hfpHhA`~ DAC Y DAB hEpNhA`lX WAOL`ohPpCxBO` DAI [hfpHhA`~ DAC Y DAB hEpNhA`l#WAOL`ohPpCxBO` DAJ [hfpHhA`~ DAC Y DAB hEpNhA`t8%mBID D(D`phSpHhA`t (F ABBF a (C ABBG nhNpDxCO`hLpGhA`t'mBID D(D`phSpHhA`t (F ABBF a (C ABBG nhNpDxCO`hLpGhA`|()*BID D(D`VhEpjhA`q (F ABBG W (C ABBA ~hNpDxCO`m hEpE xhLpGhA`|+*BID D(D`VhEpjhA`q (F ABBG W (C ABBA ~hNpDxCO`m hEpE xhLpGhA`p(D-BKA G@NHDPjHA@s  DABD   DABK fHPPCXB`O@[ HEPE [HEPNHA@p/BKA G@NHDPjHA@s  DABD   DABC fHPPCXB`O@[ HEPE [HEPNHA@,\2^AD K AG O AH 4@3AD M AE G AH V AA 8x4fBBA A(D0T (D ABBI @6+bBB A(A0D@S 0D(A BBBF $QADA HAA| 7BMJ A(GpxPCBOp& (D ABBA xfHxApM (D ABBH dxENxAp|@BMJ A(GpxPCBOp& (D ABBA xfHxApM (D ABBH dxENxAp@ \HqBBB A(A0G@w 0A(A BBBB 4dJYBGD p ABH IABJAX G XK BBB B(A0A8DP 8D0A(B BBBJ [X\`OXAPXT BBB B(A0A8DP 8D0A(B BBBJ [X\`OXAPp]QdBBB E(A0A8G nP ?  8D0A(B BBBB HK[A@        81BLB B(D0A8G 6 i N A  8D0A(B BBBA . L \ B  L [ B  L [ A  N K B M t ` P B + B D S D H H H H H H G . K R B $_fBBB B(D0A8G nP P   8D0A(B BBBG UK[A       W81BLB B(D0A8G 6 i N A  8D0A(B BBBA . L \ B  L [ B  L [ A  N K B M t ` P B + B D S D H H H H H H G . K R B HbBBB E(A0A8DP8C0A(B BBB\$!BBB B(A0A8GiHHKz 8A0A(B BBBl ` `jRjRjRjUjRX@h@8`H` ph0hX(  P HEPXo` p 5`` o ooo6PFPVPfPvPPPPPPPPPQQ&Q6QFQVQfQvQQQQQQQQQRR&R6RFRVRfRvRRRRRRRRRSS&S6SFSVSfSvSSSSSSSSSTT&T6TFTVTfTvTTTTTTTTTUU&U6UFUVUfUvUUUUUUUUUVV&V6VFVVVfVvVVVVVVVVVWW&W6WFWVWfWvWWWWWW_M 4 B@2GP3@HI@60Hfli wi@i0i((_I liНjP \uP_ _ Hf_ `PHf h` @`Hf`a @aPHfb@b`Hfc@(d@sfjf`3gf `ddsfjf 3gf fgpEgp|{RgHfHf Hf`HfHff 3gjf`sff 3gjf sfi`i`iPiPiPiiii jj$jК.jК7jEjRj aj`pqjGCC: (GNU) 14.2.1 20250110 (Red Hat 14.2.1-7)GA$3a1GA$3a1PPGA$3a1HEPEGA$3a1iGA$3a1HEHEGA$3a1HEHEGA$3a1PPGA$3a1PEUE  =P M  0' `%Q -  9+ %z PM   % К? &z 0 &pf P&x &,l & &T & @  P  `@5 WXL j   Н  E  0 P/9 U @k 8X~ `| y 'Yx `A j  )!  Y2  /[jB  [y\  \Ro    P  @[    p}*  J  h  Ph  h  d]O  _I  z  z  z  kK  A  @}v  }  Gm  nD  0  c#  uL  `n  p'  {   z  pN  co  p`6  ^q?I  a  д   P  к [M ym ; 8P 8B `8 8  , `WshX W W, Wn m Pm *B *  - ^J ^ f + rQ H pb 0qw Y   x#p9hT   QdGbx82hG(jPx@&p<hWm`X p|81$hBh _fPH*@@8[0q(h 5Kf| 813hR re u!  )bX:`} @ @P x     ! !,!xO!`t!X!!`!0!  ," q" " # W# #` #@ F$  $ $ % i% % %` 7&@ &  &h& z& z& z& '@*'C'@\' }' P' P'  '0'@)(@m(@(@(`@!)@a) +) ,)+* m* * {* &+hQ+ps+++++,!,L,a,,,,, , , ,@,X- `#-P,B-DP- HEV-`c-`---8-- P--..!.6.B.P. l........//$///A/U/k//////000.0=0U0e0w000000001!121F1_1o11111111 22$252W2o22q222222233 -393Q3c3w3333333334)4A4R4o444444445505>5M5]5o5z555555556$606E6W6r6666666677#707E7U7q77 77777788,8"G8Y8g8t888888_sgd_fast.c__pyx_f_7sklearn_12linear_model_9_sgd_fast_13ModifiedHuber_cy_loss__pyx_f_7sklearn_12linear_model_9_sgd_fast_13ModifiedHuber_cy_gradient__pyx_f_7sklearn_12linear_model_9_sgd_fast_5Hinge_cy_loss__pyx_f_7sklearn_12linear_model_9_sgd_fast_5Hinge_cy_gradient__pyx_f_7sklearn_12linear_model_9_sgd_fast_12SquaredHinge_cy_loss__pyx_f_7sklearn_12linear_model_9_sgd_fast_12SquaredHinge_cy_gradient__pyx_f_7sklearn_12linear_model_9_sgd_fast_18EpsilonInsensitive_cy_loss__pyx_f_7sklearn_12linear_model_9_sgd_fast_18EpsilonInsensitive_cy_gradient__pyx_f_7sklearn_12linear_model_9_sgd_fast_25SquaredEpsilonInsensitive_cy_loss__pyx_f_7sklearn_12linear_model_9_sgd_fast_25SquaredEpsilonInsensitive_cy_gradient__Pyx_CyFunction_get_qualname__Pyx_CyFunction_get_globals__Pyx_CyFunction_get_closure__Pyx_CyFunction_get_code__pyx_tp_new_7sklearn_12linear_model_9_sgd_fast_Regression__pyx_mstate_global_static__pyx_vtabptr_7sklearn_12linear_model_9_sgd_fast_Regression__pyx_tp_new_7sklearn_12linear_model_9_sgd_fast_SquaredEpsilonInsensitive__pyx_vtabptr_7sklearn_12linear_model_9_sgd_fast_SquaredEpsilonInsensitive__pyx_tp_new_7sklearn_12linear_model_9_sgd_fast_EpsilonInsensitive__pyx_vtabptr_7sklearn_12linear_model_9_sgd_fast_EpsilonInsensitive__pyx_tp_new_7sklearn_12linear_model_9_sgd_fast_Classification__pyx_vtabptr_7sklearn_12linear_model_9_sgd_fast_Classification__pyx_tp_new_7sklearn_12linear_model_9_sgd_fast_SquaredHinge__pyx_vtabptr_7sklearn_12linear_model_9_sgd_fast_SquaredHinge__pyx_tp_new_7sklearn_12linear_model_9_sgd_fast_Hinge__pyx_vtabptr_7sklearn_12linear_model_9_sgd_fast_Hinge__pyx_tp_new_7sklearn_12linear_model_9_sgd_fast_ModifiedHuber__pyx_vtabptr_7sklearn_12linear_model_9_sgd_fast_ModifiedHuber__pyx_memview_get_float__const____pyx_memview_get_double__const____Pyx_CyFunction_get_annotations__Pyx_VerifyCachedType__Pyx_CyFunction_Vectorcall_O__pyx_memview_get_nn___pyx_t_7sklearn_5utils_9_typedefs_uint8_t__const____Pyx_PyObject_GetAttrStr__pyx_CommonTypesMetaclass_get_module__Pyx_CyFunction_get_doc__Pyx_PyMethod_New__Pyx_CyFunction_get_name__Pyx_CyFunction_CallMethod__Pyx_CyFunction_repr__Pyx_PyType_Ready__Pyx_PyNumber_LongWrongResultType__Pyx_CyFunction_get_defaults__Pyx__SetItemOnTypeDict__Pyx_CyFunction_set_doc__Pyx_CyFunction_clear__Pyx_CyFunction_dealloc__Pyx_PyCode_New__Pyx_SetVtable__Pyx_copy_spec_to_module__pyx_pymod_createmain_interpreter_id.17__pyx_m__Pyx_CyFunction_traverse__pyx_tp_dealloc_7sklearn_12linear_model_9_sgd_fast_Classification__Pyx_CyFunction_CallAsMethod__Pyx_CyFunction_set_annotations__Pyx_CyFunction_set_kwdefaults__Pyx_CyFunction_set_defaults__Pyx_CyFunction_set_qualname__Pyx_CyFunction_set_name__Pyx_InitConstants.constprop.0__Pyx_CreateCodeObjects.constprop.0descr.6descr.7descr.16__Pyx_CyFunction_New.constprop.0__Pyx_CyFunction_Vectorcall_NOARGS__Pyx_CyFunction_Vectorcall_FASTCALL_KEYWORDS_METHOD__Pyx_CyFunction_Vectorcall_FASTCALL_KEYWORDS__Pyx_ImportType_3_2_2.constprop.0__Pyx_ImportFromPxd_3_2_2.constprop.0__Pyx_RaiseUnpickleChecksumError.constprop.0__pyx_k___Pyx_PyFloat_TrueDivideCObj.constprop.0__Pyx_GetItemInt_Fast.constprop.0__Pyx____Pyx_PyUnicode_From_long.constprop.0DIGIT_PAIRS_10__pyx_fatalerror.constprop.0__func__.48__Pyx__UpdateUnpickledDict.constprop.0__Pyx_FetchCommonTypeFromSpec.isra.0__Pyx_GetVtable.isra.0__Pyx_MergeVtables__Pyx_CyFunction_reduce__Pyx_PyErr_GivenExceptionMatchesTuple__Pyx_PyErr_GivenExceptionMatches.part.0__Pyx____Pyx_PyUnicode_From_unsigned_int.constprop.0__Pyx_PyObject_FastCallDict.constprop.0__pyx_tp_dealloc_7sklearn_12linear_model_9_sgd_fast_Regression__Pyx_CyFunction_get_kwdefaults__Pyx_AddTraceback.constprop.0__pyx_pw_7sklearn_12linear_model_9_sgd_fast_25SquaredEpsilonInsensitive_3__reduce____Pyx_RejectKeywords__pyx_pw_7sklearn_12linear_model_9_sgd_fast_18EpsilonInsensitive_3__reduce____pyx_pw_7sklearn_12linear_model_9_sgd_fast_12SquaredHinge_3__reduce____pyx_pw_7sklearn_12linear_model_9_sgd_fast_5Hinge_3__reduce____pyx_pw_7sklearn_12linear_model_9_sgd_fast_13ModifiedHuber_1__reduce____Pyx_PyLong_As_long.part.0__pyx_pw_7sklearn_12linear_model_9_sgd_fast_14Classification_3py_dloss__Pyx_ParseKeywordsTuple__Pyx_ParseKeywordDict__pyx_pw_7sklearn_12linear_model_9_sgd_fast_14Classification_1py_loss__pyx_pw_7sklearn_12linear_model_9_sgd_fast_10Regression_3py_dloss__pyx_pw_7sklearn_12linear_model_9_sgd_fast_10Regression_1py_loss__pyx_pw_7sklearn_12linear_model_9_sgd_fast_12SquaredHinge_1__init____pyx_pw_7sklearn_12linear_model_9_sgd_fast_5Hinge_1__init____pyx_pw_7sklearn_12linear_model_9_sgd_fast_25SquaredEpsilonInsensitive_1__init____pyx_pw_7sklearn_12linear_model_9_sgd_fast_18EpsilonInsensitive_1__init____pyx_pw_7sklearn_12linear_model_9_sgd_fast_14Classification_7__setstate_cython____pyx_pw_7sklearn_12linear_model_9_sgd_fast_10Regression_7__setstate_cython____Pyx_PyLong_As_unsigned_int__Pyx_PyLong_As_int__Pyx_CyFunction_get_is_coroutine__Pyx_PyObject_GetAttrStrNoError__Pyx_setup_reduce_is_named__pyx_pw_7sklearn_12linear_model_9_sgd_fast_7__pyx_unpickle_Classification__Pyx__ArgTypeTest__pyx_pw_7sklearn_12linear_model_9_sgd_fast_5__pyx_unpickle_Regression__Pyx_GetBuiltinName__Pyx__GetModuleGlobalName__Pyx_PyNumber_Long__pyx_pw_7sklearn_12linear_model_9_sgd_fast_10Regression_5__reduce_cython____pyx_dict_version.21__pyx_dict_cached_value.20__pyx_dict_version.19__pyx_dict_cached_value.18__pyx_pw_7sklearn_12linear_model_9_sgd_fast_14Classification_5__reduce_cython____pyx_dict_version.25__pyx_dict_cached_value.24__pyx_dict_version.23__pyx_dict_cached_value.22__pyx_pf_7sklearn_12linear_model_9_sgd_fast_2_plain_sgd32.isra.0__pyx_memoryview_fromslice__pyx_dict_version.70__pyx_dict_cached_value.69__pyx_dict_version.60__pyx_dict_cached_value.59__Pyx_Raise__pyx_dict_version.68__pyx_dict_cached_value.67__pyx_dict_version.66__pyx_dict_cached_value.65__Pyx_TypeInfo_float__Pyx_ValidateAndInit_memviewslice__pyx_dict_version.64__pyx_dict_cached_value.63__pyx_dict_version.62__pyx_dict_cached_value.61__pyx_builtin_print__pyx_dict_version.54__pyx_dict_cached_value.53__Pyx_PyUnicode_Join__pyx_dict_version.52__pyx_dict_cached_value.51__pyx_dict_version.58__pyx_dict_cached_value.57__pyx_dict_version.56__pyx_dict_cached_value.55__pyx_dict_version.50__pyx_dict_cached_value.49__pyx_pw_7sklearn_12linear_model_9_sgd_fast_3_plain_sgd32__Pyx_TypeInfo_float__const____Pyx_TypeInfo_nn___pyx_t_7sklearn_5utils_9_typedefs_uint8_t__const____pyx_pf_7sklearn_12linear_model_9_sgd_fast__plain_sgd64.isra.0__pyx_dict_version.47__pyx_dict_cached_value.46__pyx_dict_version.37__pyx_dict_cached_value.36__pyx_dict_version.45__pyx_dict_cached_value.44__pyx_dict_version.43__pyx_dict_cached_value.42__Pyx_TypeInfo_double__pyx_dict_version.41__pyx_dict_cached_value.40__pyx_dict_version.39__pyx_dict_cached_value.38__pyx_dict_version.31__pyx_dict_cached_value.30__pyx_dict_version.29__pyx_dict_cached_value.28__pyx_dict_version.35__pyx_dict_cached_value.34__pyx_dict_version.33__pyx_dict_cached_value.32__pyx_dict_version.27__pyx_dict_cached_value.26__pyx_pw_7sklearn_12linear_model_9_sgd_fast_1_plain_sgd64__Pyx_TypeInfo_double__const____Pyx_setup_reduce__pyx_pymod_exec__sgd_fast__pyx_CommonTypesMetaclass_spec__pyx_CyFunctionType_spec__pyx_vtabptr_7sklearn_5_loss_5_loss_CyLossFunction__pyx_vtable_7sklearn_12linear_model_9_sgd_fast_Regression__pyx_type_7sklearn_12linear_model_9_sgd_fast_Regression__pyx_vtable_7sklearn_12linear_model_9_sgd_fast_Classification__pyx_type_7sklearn_12linear_model_9_sgd_fast_Classification__pyx_vtable_7sklearn_12linear_model_9_sgd_fast_ModifiedHuber__pyx_type_7sklearn_12linear_model_9_sgd_fast_ModifiedHuber__pyx_vtable_7sklearn_12linear_model_9_sgd_fast_Hinge__pyx_type_7sklearn_12linear_model_9_sgd_fast_Hinge__pyx_vtable_7sklearn_12linear_model_9_sgd_fast_SquaredHinge__pyx_type_7sklearn_12linear_model_9_sgd_fast_SquaredHinge__pyx_vtable_7sklearn_12linear_model_9_sgd_fast_EpsilonInsensitive__pyx_type_7sklearn_12linear_model_9_sgd_fast_EpsilonInsensitive__pyx_vtable_7sklearn_12linear_model_9_sgd_fast_SquaredEpsilonInsensitive__pyx_type_7sklearn_12linear_model_9_sgd_fast_SquaredEpsilonInsensitive__pyx_memoryview_copy_object_from_slice__pyx_memoryview_slice_get_size__pyx_get_best_slice_order__pyx_memoryview_err__pyx_memslice_transpose__pyx_memoryview_copy_contents__pyx_array_allocate_buffer__pyx_array_new__pyx_memoryview_broadcast_leading__pyx_memoryview_slice_assign_scalar__pyx_memoryview__slice_assign_scalar__pyx_memoryview_copy_data_to_temp__Pyx_ParseKeywordDictToDict__Pyx__Import__pyx_mdef_7sklearn_12linear_model_9_sgd_fast_10Regression_1py_loss__pyx_mdef_7sklearn_12linear_model_9_sgd_fast_10Regression_3py_dloss__pyx_mdef_7sklearn_12linear_model_9_sgd_fast_10Regression_5__reduce_cython____pyx_mdef_7sklearn_12linear_model_9_sgd_fast_10Regression_7__setstate_cython____pyx_mdef_7sklearn_12linear_model_9_sgd_fast_14Classification_1py_loss__pyx_mdef_7sklearn_12linear_model_9_sgd_fast_14Classification_3py_dloss__pyx_mdef_7sklearn_12linear_model_9_sgd_fast_14Classification_5__reduce_cython____pyx_mdef_7sklearn_12linear_model_9_sgd_fast_14Classification_7__setstate_cython____pyx_mdef_7sklearn_12linear_model_9_sgd_fast_13ModifiedHuber_1__reduce____pyx_mdef_7sklearn_12linear_model_9_sgd_fast_5Hinge_3__reduce____pyx_mdef_7sklearn_12linear_model_9_sgd_fast_12SquaredHinge_3__reduce____pyx_mdef_7sklearn_12linear_model_9_sgd_fast_18EpsilonInsensitive_3__reduce____pyx_mdef_7sklearn_12linear_model_9_sgd_fast_25SquaredEpsilonInsensitive_3__reduce____pyx_mdef_7sklearn_12linear_model_9_sgd_fast_1_plain_sgd64__pyx_mdef_7sklearn_12linear_model_9_sgd_fast_3_plain_sgd32__pyx_mdef_7sklearn_12linear_model_9_sgd_fast_5__pyx_unpickle_Regression__pyx_mdef_7sklearn_12linear_model_9_sgd_fast_7__pyx_unpickle_Classification__pyx_moduledefdescr.15descr.3descr.2__pyx_CyFunctionType_slots__pyx_CyFunction_methods__pyx_CyFunction_members__pyx_CyFunction_getsets__pyx_CommonTypesMetaclass_slots__pyx_CommonTypesMetaclass_getset__pyx_k_SGD_implementation__pyx_methods__pyx_moduledef_slots__pyx_methods_7sklearn_12linear_model_9_sgd_fast_SquaredEpsilonInsensitive__pyx_methods_7sklearn_12linear_model_9_sgd_fast_EpsilonInsensitive__pyx_methods_7sklearn_12linear_model_9_sgd_fast_SquaredHinge__pyx_methods_7sklearn_12linear_model_9_sgd_fast_Hinge__pyx_methods_7sklearn_12linear_model_9_sgd_fast_ModifiedHuber__pyx_methods_7sklearn_12linear_model_9_sgd_fast_Classification__pyx_doc_7sklearn_12linear_model_9_sgd_fast_14Classification_py_loss__pyx_doc_7sklearn_12linear_model_9_sgd_fast_14Classification_2py_dloss__pyx_methods_7sklearn_12linear_model_9_sgd_fast_Regression__pyx_doc_7sklearn_12linear_model_9_sgd_fast_10Regression_py_loss__pyx_doc_7sklearn_12linear_model_9_sgd_fast_10Regression_2py_dloss__pyx_doc_7sklearn_12linear_model_9_sgd_fast_2_plain_sgd32__pyx_doc_7sklearn_12linear_model_9_sgd_fast__plain_sgd64__pyx_memoryview_refcount_objects_in_slice__pyx_memoryview_refcount_copying__pyx_memoryview_err_no_memory__pyx_memoryview_err_dim__pyx_memoryview_err_extents__pyx_fill_contig_strides_array__pyx_memoryview_copy_object__pyx_memoryview_slice_copy__pyx_memoryview_get_slice_from_memoryview__pyx_pybuffer_index__pyx_memoryview_slice_memviewslice__pyx_memview_slice__pyx_memoryview_newcrtstuff.cderegister_tm_clones__do_global_dtors_auxcompleted.0__do_global_dtors_aux_fini_array_entryframe_dummy__frame_dummy_init_array_entry__FRAME_END___fini__dso_handle__pyx_module_is_main_sklearn__linear_model___sgd_fast_DYNAMIC__GNU_EH_FRAME_HDR__TMC_END___GLOBAL_OFFSET_TABLE__initPyUnicode_FromFormatPyNumber_NegativePyList_NewPyExc_SystemErrorPyDict_SetItemStringPyDict_SizePyMethod_Type_ITM_deregisterTMCloneTablePyGILState_ReleasePyFloat_TypePyTuple_TypePyObject_FormatPyObject_ClearWeakRefs_PyThreadState_UncheckedGetPyModuleDef_InitPyEval_RestoreThreadPy_EnterRecursiveCallPyFrame_NewPyMem_FreePyCapsule_GetNamePyNumber_InPlaceAddvsnprintf@GLIBC_2.2.5PyObject_GetAttrStringPyObject_CallMethodObjArgsPyImport_AddModulePyBytes_FromStringAndSize_PyObject_GenericGetAttrWithDictPyObject_SetAttrStringPyErr_WarnEx_Py_DeallocPyModule_NewObjectPyErr_NoMemoryPyObject_GenericGetDictPyObject_GC_DelPyNumber_MultiplyPyGC_Disablestrlen@GLIBC_2.2.5PyImport_GetModuleDictPyObject_GC_TrackPyExc_RuntimeErrorPyErr_GivenExceptionMatchesPyErr_SetString_PyObject_GC_NewPyObject_GetItemPyExc_ValueErrorstrrchr@GLIBC_2.2.5PyExc_DeprecationWarningPyExc_TypeErrorPyGILState_EnsurePyInterpreterState_GetIDPyTuple_GetItemmemset@GLIBC_2.2.5PyMem_ReallocPyErr_ExceptionMatchespow@GLIBC_2.2.5PyInit__sgd_fastPyOS_snprintf_Py_FatalErrorFuncPyTraceBack_HerePyObject_CallFinalizerFromDeallocPyUnicode_InternInPlacePyLong_FromSsize_tPyFloat_FromDoublePyLong_FromLongPyObject_RichCompareBoolPyModule_GetNamePyErr_ClearPyCapsule_IsValidPyDict_UpdatePyImport_GetModule_Py_FalseStruct__gmon_start__PyTuple_NewPyObject_GenericGetAttrPyThreadState_GetPyExc_OverflowErrorPyDict_DelItemPyNumber_RemainderPyType_TypePyType_ModifiedPyObject_SetAttrPyErr_OccurredPyLong_AsLongPyImport_ImportModule_PyDict_GetItem_KnownHashPy_LeaveRecursiveCallPyObject_VectorcallDictPyTuple_GetSlicePyObject_CallFunctionObjArgsPyDict_GetItemStringPy_VersionPyUnicode_DecodeUTF8_Py_NoneStructPyExc_ZeroDivisionErrorPyObject_VectorcallPyFloat_AsDoublePyObject_IsTrue_PyType_LookupPyImport_ImportModuleLevelObjectPyObject_Hash_Py_TrueStructPyLong_AsDoublePyDict_SetDefaultPyDict_NewPyObject_GC_IsFinalizedPyLong_AsUnsignedLongPyDict_TypePyDict_NextPyBaseObject_Typememmove@GLIBC_2.2.5PyObject_GenericSetDictPyObject_VectorcallMethodPyLong_TypePyGC_EnablePyUnicode_FromStringPyEval_SaveThreadPyUnicode_InternFromStringPyExc_ImportErrorPyDict_SetItemPyObject_HasAttrPyExc_AttributeErrorPyBytes_AsStringPyExc_RuntimeWarningPyType_FromModuleAndSpecPyObject_CallPyUnicode_TypePyCapsule_NewPyErr_FormatPyCapsule_GetPointerPyExc_NameErrorPyUnicode_FromStringAndSizePyModule_GetDict_ITM_registerTMCloneTablePyUnicode_FromOrdinalPyUnicode_ConcatPyObject_GetAttrPyCFunction_TypePyMemoryView_FromMemoryPyUnicode_FormatPyMem_MallocPyErr_WarnFormat__cxa_finalize@GLIBC_2.2.5PyNumber_SubtractPyUnicode_NewPyTuple_PackPyCode_NewWithPosOnlyArgsPyCode_NewEmptyPyNumber_TrueDividePyObject_GC_UnTrackPyDict_GetItemWithErrorPyList_Type.symtab.strtab.shstrtab.note.gnu.build-id.gnu.hash.dynsym.dynstr.gnu.version.gnu.version_r.rela.dyn.rela.plt.init.text.fini.rodata.eh_frame_hdr.eh_frame.init_array.fini_array.data.rel.ro.dynamic.got.got.plt.data.bss.comment.gnu.build.attributes88$.o``$8 p@ Ho4Uo  @d```nB55p xPPs P P~WWh HEHE PPP PPXX`` X @8x 08.h *. 08O