From 86358bc28acf30faa69f0ede6f313292e36a2754 Mon Sep 17 00:00:00 2001 From: Morten Hjorth-Jensen Date: Wed, 20 Oct 2021 09:50:37 +0200 Subject: [PATCH] typo in p2 --- .../2021/Project2/html/._Project2-bs000.html | 4 ++-- .../2021/Project2/html/Project2-bs.html | 4 ++-- doc/Projects/2021/Project2/html/Project2.html | 4 ++-- .../2021/Project2/ipynb/Project2.ipynb | 4 ++-- .../Project2/ipynb/ipynb-Project2-src.tar.gz | Bin 193 -> 193 bytes doc/Projects/2021/Project2/pdf/Project2.p.tex | 4 ++-- doc/Projects/2021/Project2/pdf/Project2.pdf | Bin 239949 -> 239764 bytes doc/Projects/2021/Project2/pdf/Project2.tex | 4 ++-- .../Projects/2021/Project2/Project2.do.txt | 2 +- 9 files changed, 13 insertions(+), 13 deletions(-) diff --git a/doc/Projects/2021/Project2/html/._Project2-bs000.html b/doc/Projects/2021/Project2/html/._Project2-bs000.html index 3658ea708..8996c18a2 100644 --- a/doc/Projects/2021/Project2/html/._Project2-bs000.html +++ b/doc/Projects/2021/Project2/html/._Project2-bs000.html @@ -165,7 +165,7 @@ MathJax.Hub.Config({
Department of Physics, University of Oslo, Norway

-

Oct 12, 2021

+

Oct 20, 2021


@@ -228,7 +228,7 @@ epochs as well as algorithm for scaling the learning rate. You can also compare your own results with those that can be obtained using for example Scikit-Learn's various SGD options. Discuss your results. For Ridge regression you need now to study the results as functions of the hyper-parameter \( \lambda \) and -the learning rate \( \gamma \). Discuss your results. +the learning rate \( \eta \). Discuss your results.

You will need your SGD code for the setup of the Neural Network and Logistic Regression codes. You will find the Python Seaborn package useful when plotting the results as function of the learning rate \( \eta \) and the hyper-parameter \( \lambda \) when you use Ridge regression. diff --git a/doc/Projects/2021/Project2/html/Project2-bs.html b/doc/Projects/2021/Project2/html/Project2-bs.html index 3658ea708..8996c18a2 100644 --- a/doc/Projects/2021/Project2/html/Project2-bs.html +++ b/doc/Projects/2021/Project2/html/Project2-bs.html @@ -165,7 +165,7 @@ MathJax.Hub.Config({

Department of Physics, University of Oslo, Norway

-

Oct 12, 2021

+

Oct 20, 2021


@@ -228,7 +228,7 @@ epochs as well as algorithm for scaling the learning rate. You can also compare your own results with those that can be obtained using for example Scikit-Learn's various SGD options. Discuss your results. For Ridge regression you need now to study the results as functions of the hyper-parameter \( \lambda \) and -the learning rate \( \gamma \). Discuss your results. +the learning rate \( \eta \). Discuss your results.

You will need your SGD code for the setup of the Neural Network and Logistic Regression codes. You will find the Python Seaborn package useful when plotting the results as function of the learning rate \( \eta \) and the hyper-parameter \( \lambda \) when you use Ridge regression. diff --git a/doc/Projects/2021/Project2/html/Project2.html b/doc/Projects/2021/Project2/html/Project2.html index 1ae046634..775efa135 100644 --- a/doc/Projects/2021/Project2/html/Project2.html +++ b/doc/Projects/2021/Project2/html/Project2.html @@ -123,7 +123,7 @@ MathJax.Hub.Config({

Department of Physics, University of Oslo, Norway

-

Oct 12, 2021

+

Oct 20, 2021


Classification and Regression, from linear and logistic regression to neural networks

@@ -184,7 +184,7 @@ epochs as well as algorithm for scaling the learning rate. You can also compare your own results with those that can be obtained using for example Scikit-Learn's various SGD options. Discuss your results. For Ridge regression you need now to study the results as functions of the hyper-parameter \( \lambda \) and -the learning rate \( \gamma \). Discuss your results. +the learning rate \( \eta \). Discuss your results.

You will need your SGD code for the setup of the Neural Network and Logistic Regression codes. You will find the Python Seaborn package useful when plotting the results as function of the learning rate \( \eta \) and the hyper-parameter \( \lambda \) when you use Ridge regression. diff --git a/doc/Projects/2021/Project2/ipynb/Project2.ipynb b/doc/Projects/2021/Project2/ipynb/Project2.ipynb index b6ea3d965..9d93e67c6 100644 --- a/doc/Projects/2021/Project2/ipynb/Project2.ipynb +++ b/doc/Projects/2021/Project2/ipynb/Project2.ipynb @@ -10,7 +10,7 @@ " \n", "**[Data Analysis and Machine Learning FYS-STK3155/FYS4155](http://www.uio.no/studier/emner/matnat/fys/FYS3155/index-eng.html)**, Department of Physics, University of Oslo, Norway\n", "\n", - "Date: **Oct 12, 2021**\n", + "Date: **Oct 20, 2021**\n", "\n", "Copyright 1999-2021, [Data Analysis and Machine Learning FYS-STK3155/FYS4155](http://www.uio.no/studier/emner/matnat/fys/FYS3155/index-eng.html). Released under CC Attribution-NonCommercial 4.0 license\n", "\n", @@ -67,7 +67,7 @@ "also compare your own results with those that can be obtained using\n", "for example **Scikit-Learn**'s various SGD options. Discuss your\n", "results. For Ridge regression you need now to study the results as functions of the hyper-parameter $\\lambda$ and \n", - "the learning rate $\\gamma$. Discuss your results.\n", + "the learning rate $\\eta$. Discuss your results.\n", "\n", "You will need your SGD code for the setup of the Neural Network and Logistic Regression codes. You will find the Python [Seaborn package](https://seaborn.pydata.org/generated/seaborn.heatmap.html) useful when plotting the results as function of the learning rate $\\eta$ and the hyper-parameter $\\lambda$ when you use Ridge regression.\n", "\n", diff --git a/doc/Projects/2021/Project2/ipynb/ipynb-Project2-src.tar.gz b/doc/Projects/2021/Project2/ipynb/ipynb-Project2-src.tar.gz index 91a0fe1348340d54253551558974d9a5930f965e..1084918e1d066f97d1a821a339a7a0d1458c77bb 100644 GIT binary patch literal 193 zcmV;y06za8iwFP|%5Px+1MSaC3c@fD2H>uHia9|^nm*QoUAPd6c!AWWHq~Zol7hXx zeSoeMH${Yeo1bBZVdju+w)-rwck8W&5Rya@OqnL}oFrW88Ko)Eh%%N&@zfxc#f*4B z);sB?Hh3D~00;m8lMY=D literal 193 zcmV;y06za8iwFS0J7r-21MSaC3c@fD2H>uHia9|^($`wB3l~BWFOb^QrrJzRQn0tT z573q3rihSl^E1pa%p9`KcAo|IZoSnILXs$gDbpmLldz?pQJMmcXh>NYr-TBbVa%8Z zWWAGKdSkgBPigB$C?nLnxpAzjKI~aufoJ}SV=WEr^1;@qK%o@{;stVzjW}5r$Zk*t vlqk&91TAj8)B?B~fS0AT5*5GtoyN1~tqJ^JzvDQL<9z7>f~+$N00;m8pM6=p diff --git a/doc/Projects/2021/Project2/pdf/Project2.p.tex b/doc/Projects/2021/Project2/pdf/Project2.p.tex index 90db3aaec..962c42528 100644 --- a/doc/Projects/2021/Project2/pdf/Project2.p.tex +++ b/doc/Projects/2021/Project2/pdf/Project2.p.tex @@ -149,7 +149,7 @@ Project 2 on Machine Learning, deadline November 15 (Midnight) % --- begin date --- \begin{center} -Oct 12, 2021 +Oct 20, 2021 \end{center} % --- end date --- @@ -211,7 +211,7 @@ epochs as well as algorithm for scaling the learning rate. You can also compare your own results with those that can be obtained using for example \textbf{Scikit-Learn}'s various SGD options. Discuss your results. For Ridge regression you need now to study the results as functions of the hyper-parameter $\lambda$ and -the learning rate $\gamma$. Discuss your results. +the learning rate $\eta$. Discuss your results. You will need your SGD code for the setup of the Neural Network and Logistic Regression codes. You will find the Python \href{{https://seaborn.pydata.org/generated/seaborn.heatmap.html}}{Seaborn package} useful when plotting the results as function of the learning rate $\eta$ and the hyper-parameter $\lambda$ when you use Ridge regression. diff --git a/doc/Projects/2021/Project2/pdf/Project2.pdf b/doc/Projects/2021/Project2/pdf/Project2.pdf index 1cbc3c1c5c9e304186b13513806adcb51cde7b31..9e00459a17e4ffc48c6c111e297182b27ce62d8b 100644 GIT binary patch delta 17370 zcmV(`K-0g?(hii-4v;1RF_R${D6vn70e`B9_R$6v<*tUE4kwFaMYE-dmOhszSzzKs zq$*`kc-s`iP=AiJDT}ctM>A2()~CuRyYDRqbS?HV(uDVYcjPW|w4z^)H(}H5>tU=* z7Wa0$a>bJdMv2rr{k%SV+}Ls1+Nm#^xjvHjJ<(s+e;N3gD;r&$(#XIlBj9~*i+|e# zT*t&FJQj7!wW6j|#Id_F5{w7(JgPlV6spumAxXH0m814BPSrD4@7&OCJn)FJ*~suK zgT`babW~jL*`^RsH`xcskbPw%sf58@x)VUByT%?VHcW&~XP1i*@{im$brt5vZI z(=#da!@B2jw19nQUDi(Wv@a$=qWlI410 z*>}URecZ$;ysNIftw^zApi*;WS#m*Yu~cycP^IjS57Wr=zICCo5=rdQ%_SYA0uCWm zj<#sG!@sAZXA^7Wsh+*jcp%!%O`7D+m_$oB_50}vIVYFv{XX%8Sbtu}UybAwYrUSM zUdC*#qKh;27iTKU5(rg3b89HO-tqwmn69DjwX}nX7inbHkm?Z)tKLs-NplBoGEC9l z3=(=qd%yEE)=m8%%eYR0y1h{P0l`F!B6Ezw8NVP^iRFkZKDoSUlZjL+SO%r>R{81p z;AmRckI;<>8KN#rz<)7D`?7(RmLL&kz1*A47Mu_<>++k@mzhtoa~h?{Wm9(9v-H|{ z&BdXhLqBfVSR`?&6XnT*)F7}M?YF&Lf~VwjJPyqF3Z{ocS(D{k+G6B;=lhc-4yUjFCINzyw0sNb0uO`1gJO3dZWWQJ`8VmyFD`KbSD5X zySD(Dj|azTcVO7Wu4i7oyKB0;ojbuc1y2J}+d4v-F!wwb!)W_m)s;IuSm^GD-Mh!4 znOxw#$MI>Avw!5K>yP4CT`=-*VX6TuH&9&VN@SUTuB<7B%O*aDBY#x&u$}E4UFK|+ zVeR>SRZfG8Z8={kYDk;M!iA?o91{=n$x+R5EdA9u3ersjcjBJ7bKs=!G9dW=$Ho?e zpXc0%hWLcXsTA$*=tqq8sYZ{O?ka4l1aFtro1(`PkAKm{^8*v^C(^94M93Le03RYZ z#^0NNY2D7SRPmDwGY5*HahOv<;U5)_9(t5}pkw=t-3@+)gPq}*eHbFb;S=yx_P&@J zzuI8O;j4!u+&5W14dG|W{3Od*-_J@$FFDOJyy8qGX$+W)h4fjQM1R)~OeVfz$meu`^OO3HM`qzqo|JkIBF>42Bg>v?fo@upOn{V~Y3kUW95uOA zU1=eWCzp4#6XIcx=P$JT!mVej`->;8JiyAcBSTO3HlxOxENKlM%pJw-gTylY<S4XWgZ_)E#anJ=W|kO#;E;hkv4WFBNB&o?9R0mz}H6PmR<+H6UE_ zT6&CkT%L0;8jee*GnY)6OF~YL=>jqG?oMOb-BBY?Tt}EcfXs8D=C@YaO^pw+JC56x z7s3N&Mvw?kR2~wgSPxlpSpZO+>p-Tt(2+ONI~du0Yf;jp3@SH-Zhg^ynWIlFe9*NV zD}PezJOw!1yC*F*=5xxE=~ejPoWc=e?+WN9-mF#Zm-7=>W|?(PQlVp?fSvK2RhCZZ zp38g(KnL++jBBuDyaBiHba2Z5H^!yG``(!AUoR-%T}+R$QL-GPQ(X7 z_O(@}h9cZxe5MoLw~P_(&6vG)H(`(K>G85_O@tC>oMg=?|;GP z2;>E-OZo05PlFWMG?Hu9l!I-@L53JjJPdqRvHze33VNQB36mXbNO`>_CWR3n^*3$B&Z#j z-P6;PnCecn-Oha)0Y9_H7XRQjZCo+#?w<#mi#0Gi-qmeo|76mnQ(+Rh-qE&u!HXZ8 zS1#`<^YOA2`Bd`Zx^*cm}CtW}K@EV^?lK`i^ z=iT@Hz#IH0cRIJSd%7dE&3~Rqg7N*(pAgAch+nGgW;kr(@ZH;2n_ua|$E~j6_6I&= zQ#rxX0TPfEd76AnYjTZIMvG|LzB_nC(By#Nxpkczo4#YfV>2Fk)HcYFhrqM3 zmwzGx24UBr^nk49x765OJT&^j35Q_5gDad|{b759tmVmlf5GebJb!az#5~Q`sOwCb zk}=1}zsxIRlONpR1l=``=WgWx?}v~0^b-jYfotN@TG>qG@@88Z5ORyx7*Gt-CKxg6O|t8| zaB0VgDWKH^o&YJb*MD_;Idl{VQ|kvXGO;#>&yrG4v<9`M-sDxOY%Q$+=eE*Ge*ry9 zt7L~FwP~54kn$rYF`=f`a2fuiIwjjJLs!d+zDz};`XH}?36-FFf>3@M)>46bFFkjY#1+#6X z%1m$a!YJT4zxkIQ74DHVP#y<2W8u{XUbUG9Bv6MhJfV&KrQP$iZ64?(&rp5vKa7|C zXCC&zkAa7KH#%AOW1vSs_ecv7$Xd9(<~;P#fc)fL>oo)OOcf?$V&pY+(D(-Vsx0~C z$c!TACT;=TYbX;03TLjpdX=0eP`Df*n4p45L?JgyoBB(93!JRH_5D4sO-!S}be5AE z4Ih6?5T64caN5a($78cY2_1QQbb$wKCf;|lA0*cJJx>||ieYOCIBOsL5*fNwr5154 z4gV%D3kaa(M@pNVoG1>w&5&l~fA2y=gq2338vsAE+A);86aaVq>C6&?iHkm-#;*PR zUq2m-Lx*nIS=*#&jkHOlWV zx2N{D>GuAUIv>whPI#5cK4UJngk{MMLyW8J^!=xuZ`(y6T?XG6sN0@te&FflED@dT zhAyG3Fswp$(}i<0P}o#o35Jj*|IL3L4y{+aDSey*Al)_NeU*1a3X4+JR`L%2pNnfL z;Uy$E86Ny(N`^?JY&*v!#l|GXK9i(iLfH>~NjFh(F5JdDt<<`pQ}0niPc!vYvW)CaII*+cBjD~YD-dk23Z6!Ykg0JeYY$q#dm z4)d6!KWhxusx%fo@GwdK-%upL(bOav{}J}g&QwPN0RA{M5p3jrSnp8;j~zqbk9;Sa znG58qcD(j>aQkLfMmYMN54#1xfK$uU_iw+0zl?VfaWxPJ*N{AJ-M;DwP?L z&(B*3J}it1*o!xR1A!!~L2`dJxLr1fIt3SGB2loh$74hkOy0A)03Gl{MD6%YqzILV zAwtm%Kmy>Zy?@Pf^p8Ujbi(M*Le*ETJS|iPzJ`eU9urENLgGmA+uDp19snIMY9oMz z2;n$OFik4zNuzKPGzu14o^+DF(Rpxzt=oB?G#&3Cwjcm5?e=ciPtSihokKr~n_7sv zRs^nd>Klm6S7j^?nJJz&)EYtCxIk8$P zH$bv@eXMOUM|0!Qe3Fn!A#RFGefLs+Vk)h4DWY)gp$8>8p19Z*W0}$g98C5B|DfEw zF=CG)o5Ev-MN7sB3a)=UIl-3HPf094bM%tj>1Cd%#c#ODrtq)lGj6|sk)kqHCadPM z^~Qk0ocfdGYnKy47<(8%Hi`^_T(NDYY7^mu{0nx=Gat9P#cqip7Z@5YQ*H?}C(w64 zzURzcE?q|jux{{y$`poi{`Nb(^VFA?WB|20x(;udmYu+WC<2nfakRqJ(tF}6@p5JnG18i_^fb#$i` zHN@DtY6_TTM!AUQXlctpZf6rSD-$y;4h)=^ZZ|G0B3-*&*raSO!eC7FocoIk2d)$} zGk@-KK;GqIXKH_OiZ}_H2fxyb8F+#3&yrKqH8&sJ$O`18DiOLMC`&F!(UJGWErEBi zM-)$p02iLkKaAGO*40ykN7h8fh8@@f6O_-LnfAJy<0LaF(b$a?cUFuYJ4=X6kml7~ zo0zK-1gm-S*j?-qv7n>tU0S!zHYQ6|ZbdfYpD4=uejmjYsT;X-qEO{Cbm*;H;WTmU~e+Y~MIGJ*8$pQe~ zPt@9V-LLCvZep-Y<2EhCajsn^2U|veUK6`fWyZMo4KP~ftz-u^As`SXd)gtIY)O6wKQY^`(4x?Wlg^3aX`3#eG5Ar@3m zaLK-ci#4SxAq-xfqcB{P=(nigPn|&yDg=>V6{UYD&9zlh{k`f9tUS9ev?{fpx|BI7 zgf3D^@^z7kY9MUB9%mFual@4PFr4fU;aracB=`}_ID}D24WMc5r|nHfJ(J7e3{I;1axB&**tmD>tgvN3kTci$5AN`!;ffbS5p{o&mmje;?FC(^21hD!UNksU zsVdXUOi5gVgxd6p_m6qH9-$Hd~I3xKze)FzId z_uu6Lpte>f&z~UlH6LJe$P?-nA&|^=&zpZ5IGU+@vgn4|S1V!&c%r;1 zXmpW1sTIpQ`4DTxw8UDmNT%wyL|<;Q*b&4W1KXq>Fa6}%vTy+{3zzU>vMk&%m4q*p zeiM?W20kFSmPktCFo?1>r#Xs1O7NOzDC8H+HLGfPqKF#Yh_*H%>St^sLhUMDKPG<% z0B2@SRA*A6cmWq&H5FX1r3oAsK@E;SO0_-5LJ69S5~;>;nkQ3qq**_ZZ=%8zfx_Y! zs+p`TRVqCWa&7W$qDTLXGRW=Hh2pe1oiU{QSn#l%^O#M;;De(SQjVXgL3?(G_$?Ak zn?+&>)A^$=ztGJWK9^%=MGnz1e)E5vKHFk1Dt>?ce`=F5)+Q#g+3)_UlR;{h+3UuZ zH8D0>bZU}LV|s;i`eufD$kS!YWA&o9g_EiEVMRmfWc<*Bxn?6@dZKr-tN_4k*FEre zmzQL2GnG}szJ=#4a;Zrzb_oZr%K#`JL|}w%gSdt3jGj_tm@+t@tJ&o_{5V@Q_bdP) zy$2YvDU`AI0FVqnUgVTkV%_;%Zb4M&T4~8>LAPQN8NXZrAEaL3f~wj1W$lOMXBg*- z4KwydcUHu2!1b19b7N{%6~M4Pyrr+U{Q2(H{{Uaof5Nl(71IcRQl>9!EBYJ-NHWQp zbpY-y&;Z21Pq95+P{b6rfY378!X+Tie*lbgZYTzvvxPlzKi6|wg`=+6nvbAM98XG; z&_(U}^61LClybrtc?4WJ_yZiIe^z9UxyT%2IlJphSJbUfi_t+@5UXKv?vIeahwQNE z_T|toq3f&Txhn7TC3s!eDU*XH^S4O+*6IN0WBd=W6HAza8mt zz<)zK_?&e4?pRT{<3)6VYY!p4!gK1RayU3$0j3cdR&jSOGsH3=3+=fBgGidIQ0n=L zFP@C?m~=)W0X$%}bmy?)bNwm+KWfDC%K?{BXDcjfh8UGG{wnuRea&36z*VVYgz=BPgUW{iI_g@$|cVcW7U6NR6D(VLGMmy@vPSYUkJ5E@iX>8kd~GtMsw z*8%1~Q-MUCABn4GTOW7)vXDpjlm|28H`9lSJgTekCK77s;1jLPk^ z*AX|%ow5n^rgM71QThjJq~*1U37Mn?A&r;(8Y}w(><=daM-nY(%WPcqR97CABEb}Y zsvV6|4I&Ni&e#kOWr$6#X#of+JqtGWCw*{~A^?p4tDnx2Mdi2qugf%KOx4w{KTxF<77s8~UvtzA* zl^MWyz;`m^OX$fQz?%ODtGOWilky@d0Wp&y7AdnuBgqJVg!iju`9BrMMJ9?wF1LP1 zNgrc;eOH&5xv?5(y2tecH+wz`6YwKC}e1afH~x77^E*gN`z|5a+-th zBR@kSlfZ&lcOGb<|Gw)l7TprUPFi*#SDva0O$q^j5)~rOvRj^B$ZmMG$7F>z_@-p? zYDS!Clfy__Bb4T+%J=m{c;-z>duf4Xr#}xgq3(ck?+~6wxvEHu@P<>wEo5cP{Y*X^ zryAo9z>#QD|7k`Z^Gh<_HI8L-5x1m);xyxTPLEMDw*oEh-Sk#EO#d|T(qE z2W+|==$vOb4fQ^RK(vM%QX%=V$1LXjdi&--md@p-v#2eA`VcTU3NK7$ZfA68G9WWF zG&h$)I{_2~G&MIllOYx;f3;P2JXQbyw~|q0Bov*GJ-Xa$uVh?%W@IE>?zvp9d${-F zBBPSbA|jMzG*B|5kCC!U8BvivLeelQnZI-O8P((a`SbU}vW2w(wHFm@sSAPSJeY#Y`PnS$`Z zkYP7S^uiNCX*fk6@(&^5y}T*38j1)6tr9I;PYy7}c=|%YWM4dhAz}ejISV5pwf${2 z33xK#-_hVy90~He{9cZ=Zz-=!4Sz8fZrsL z0B#c#NW{W)QoO+*1!%hinBzS`A{nG58A3mtelV4A5{y&+84MvY)^B+P`!xOhL3tdYSLjzko zggLA|BEkZK*&)hNsFdw!T0LE~xdx!3t^}aqho(~lB32*r^Mj`*v(e~7<6)jC5Gh3d zKLqYego24-|K-KuiC7$M7qEf;^7cf0Kp<#>{woKDf7t%Tyg&+o1i%0Qq>t*@QW>sj0pk(3Mmi_5Bu5itHp*=1+aKe3M_i~;IM61 zHzDF6K>Y^{Px80xFF8ml$ie4H8a}jGh)4(lupo|2-V&m~;*k3PIdi}DG7KaTEHQqd z)c+>-fA?sNAD$5MKjMFja{y^`OIboBKMdhFAD(Q8r-E2(JjK)d2gyJ1AEk7OUIY+8 zpp@j0iYf{}JodEHMu24vUrl)0*-4b$z(tYx$OdBW`3I<-VWOAfc!Cg zb8FjUvi~8b?O+3F}+P2gb0C}5#EF~pC-V+u-<{v}>1w-)N`wxIZssZwZpZqiv zDFlTCgN5XnV!8rW+``3wI8i@aU)PBm=(}SpC2t@?|p`Z-Yh*VMoRF#q8zhrv; zxW%?l3cPuL+q5eL06{9~$u{r`^3;s>$*7LG9bu5)U9>}7U2eXFPutNng=rwGyKX-} zI(I?@JW_C~CMsK^05Uh$I1zE_6!C_me|Yp>!sbYA+MTXt>`|RbOoT`!M~^hXsnh)BOPy_V;6Q|U$agOmrtfPKfDzIt8b_aueREw06M5tpI|L)Ui%7xtm< z^O6YM)c!q!_51s~>**<5-+3=2Vjk%Z$n?7B#qf2QGW$0*@=Vbu79LqokebRyf6g)_ zdb?cMdZfi6TN;4KGMi;5sy+EqU^0w&ZlcA;Lr4Ev=$_m}gLUFk(?^!eIDsgK1pV|i zRLBy25Eq@OkOb#b&l$_1&UTimdbVo2qnDQCV-I9*ss)c~J~4E7Gf@u&g}!?e?kF=V z4w>4tL|zq1&_2*Il)5o;K18T&e|nXFthQWF%{xR!_6a+6;5&AQSes+_=30P8)^42g zuGfhdo*3LXoVc@n>jZB6^?>L7W08`aH9;@7go75=1^XEcr+9i*Z}Q`>&CIhf8LvM# zbM7}#IM?snCfTHOqrB;>n7X>phxUw@hOH}NTq36r#%^X`SCVBieV4B_f0bVR^7S|0 z19iQq+(f@V)tl${8fZ+G72~j7^%rX|%Mm!8??_OY^ql8yH294(H4qF=(^e@K%Gy0r za}Ja_=-kLgPhGyVu6E&TSG~z9oNYFRx0H2JOX%)KilSt7$Hy=?567 zPSLHfCpMBzo6v_W0w)Wzf7KyImx$z;~u(guJ zP=;HI+%?|U2MA6Onf#+CK4`0WRMW*x+TY!*NH;QWSy5?LeW+J$Hdz`n5?EqMvhRGj z$l;ep_aeJ}piX{aEa^N0r(*paOMbj-^79vJF$7%Wr~R6k#bf5rfBV_HjU926TD_D( zl6RMrH*vs8a&6_gf|!G>cLaUNL7kU^qU_hWKlS@WCg)}(E&{74(@IpvCh??|NOPmP z#kZ2SPcyA$HkZxDwubld7T0#3F>|VMTDgc~5LTLK_}Z3qEMvl@>NP^fEh=?yEdRW_ z=;5g}k@G!=+lz5$e+v~wb?X?40c;W7+Jx_hq=XieA&=x*5hAv_$B4RjnLjRSetB2# z9bxIwd?S4Elhl#;(s&L@)nt$Xou zANeCW9@p?KS&A=n@_@$q)SNJv^5K2+`52!G2D>0FhB%+LfBbao)i0lV==ZB?iaReO z-oCA<>^J|svuU9074pjRjo>2lEBv^HGo{q$^G*f$ECsGV^7(T=*Kj3D92cS}EmTzE zYBMxOPraT`ckg5p;MdqZy2p7l_neHl4O){!Q7A#p;o8hhZQj*I-;*6f$D*X=oEe$! z-J8jCWB#~re{`5^Ssq3bb|E*Jxf~wxXu7+j%N;MV>fB&9e!9|q_U6$7z2II3NFZ{?o}dW8XdNttisKAn^QCXtXaf&&fL*FmZ|2@e~(gC zmyWY*vM&d!wpbZ31#%-c*^DYNO-oscZ<^*rFOLgbe=h@u!|>dpw|CiN8frJYK5gBppAH&G zdIgwi-?3(RY>a<%eRoz{E#XvXuT536I+gT6nB&=CEKh7agXTHZQhWVKOc|H>9v_xHYOrQK* z;~^w`Rq*+jU_ccE@DCHS`@qG+Pn*P_Ii}EU-G1Nu3;|(XO~Q~vx*y7FjtdDpf8T8g z80?vh|1Qzuc8N+8r{o= zOZt|kj<6ek=RO?Yrf*i#Nr|pqL9_FZB|Vu_9puvs$3Q)0YL-`RW@EU*kM?vN(bdf_h=JoN}esaL=gNSEExCos&tIhpKo2m#F8-f0i_5+dNTn zL;Wh^apn9Uh$|xTu{&n%o@YKCtrZbLdl?q(_`*VVu*tb^d+~u0K0&9Ly+u*Yrnv&W zr?qiu+`sTt8U5GClC}&YC*NA=S&e{)Gm5@J%avj+vh+9lIwZM|&lgVC=1s*29g{I_ zxX2OYe!C!tF_mQPnq7P7e@ZufU+R^2tmNt>zS8%9RDH=!P{8)AK{uBfuk)|!TykDb zN@X|PIR7cK>Nd&e@|Qy*bTb6BrgDG+2HmVDZGjhvyVT~of5rtNX8)ky%7T=hke^f= z{IaAgDK_zlf^trmb2rx!o?BInUdJQ9i8Nqu6|>E_y?doF@-RA5f513ExL9S~v3&G# zq2}cM2Y0|sbpl8CvdqoSAc_kggF;4F4Kn%W(p_uPA%)eCi9{|w>Z4JY>XdcMIJP&V z&+4LU!d`f*c$uD61|A?nYd_lkbp?<{0#b7HqD_v@lv?GY|7$NYwK z=@HE$f!OPY!VwQwbdb?_t{Y>~TFuJu%QdEZf;IJg(ir=>mo)?gd<>2UupYQ&(j4#^ z5zGehR6QyAx;_*j}f zv%3ZzPwqZ^QZ^EFKF2&Hxe~d|22pjNF6HcO?A?sD4D$ZqZg&x-q#5~~_%MXqo^HQl zdb+|}RwLn7e_=*@}6wx{8&{K-fO8Q|HI`gn3@W^hTxlGxXQ`M~qh*U?s$l6cG|t8)?_dqS zRLr^@e<5_LUyZM7wdvljllMaj&7jw@Pkk1H^u8RPpBcJN4>+X$*)OU*q~(0bUUGqt ze`W?3>|`@;KsYa$NEk9`dm~jXlvf5Ou~)|g-8zalskc2_Eru7D6aw?KX4S+L-!?G| z1x}aeRk>+ZZn%1UcO1wv%0B41Ja;^Qv!T5|e`6#{<|25j;?uo@xogeCQtoXxl{yAG zE3G8d?b{n4-t$b})fwK!=3rQrgr16iak;d_iDz7N?g`UoT)>BA|A{S!GHyg=*`7To zgl-P<8mmf9R}ObqH2d)&V|Pc~O^Z{W?$t&*?^+rxJcYv%1^p{naULhy9V=&~w5dnDj2dY;txT28CycoJfv_3fo{zaueQM@(Ouo|)4xRb&$D#?C|~ z<0FK#4>Fg(+qw7TC2!*8>n$XtLE_rJfBBEIi-1BOzWEQU`ioEKq}q_O!C%{hyXtGr z=`3SG{a$9ZwqAJS^vy0rq< znyoaaq;h3RuQ8n~Y>i>dl1Z`>Ns=IZp1)_VIwTc(D&m|IVDkzglWN!$NRb5o1rJh9 z{>GP47XlN98an}p8ao2F8ao51QvxwGlOYx;m+X!M7Jp=WV{Z(3?e!r>Fe1kiV~WG& zkhYYk&sWt$awtnSMLPQj!4FV9J=4|QRn=8rRX2xHg)B^ARACF{QsD~YT%lFrk}iA^ zjVzQZlvJ)z8dnzzg^!Z5&<0m)4iYZWqhFznjRjhrGDQR(hS9i9Mi)9_P-hCA(9dH^ z^miyQv48dv3=m0@2L@FcLSgXH5!}SWpi-EGYtV(U_$XgkT%E)fLwMp|fmG_k;v+Q* zG-;gynacY+n#SDY7hXQS{t7dPW8)X$-PI z?IqPPhfY{3g%wan(|BZ+#%X_`G#O+F^b9>rN;Cp`SnDtoY(`lR^(m~Ta?)}?)C`Sm z$cnL$;D}ligoLpuuW53a32H$d{Pa^HUVjwgQS)n4hzG?x17A8W#L;N}zWB>u(0m|% zy`0vCcwU{>2jYG+nb(tfixG(MKpfX?vzQI)wn+KGvwAqH?ll+1JEHK4)>pqjK>ry! z!c#KY-N~fEFz?{AG@;`XDxeFK6z>l<=I0Iv;$AZw)-xVkz88e-`5N=g;M0)RdD(wDZMqRL?|xK0$e2%_r4de7J1I(VG_r$`+%^u)a8~C#U7v z{CxZ)1lUamEnNvWHd2-30XD?a!hddz(q(Ywb|SQp;dC@VTbz`G<{T?{K7D?6*^UOS zcvcP0Mw9w!UCkz=$*DLQHzyFQokO}|GZ4>b%|F3UD;|ET#*2yue>AIxM7n-*)XdJy z^WpBo!s^Qq7(7c45CQ0w){ykF3@QGPNoydTCm&Y61eQBjvh#=1q#YKPjDPcEb$U9k zS=_~7bUqd*vl>zz4ywtZo*jN3wS#8TjwXl0YF_miUwixOH*X7G1ppnOxdePn7US{z zn}Q_XWaGu74vw=;fHeVx7kX8)P1X7d2K8Soq+m5Wm(J^Q$0r1tg8X z3X;a|hLVBV$b#s6qr1B-k;%fCjG!iqW6WMiYIQ6lMQy7~Ls^Koa7;vZXG#R4+gmG4 zbUMx6Sawc2YQFGG^2O=mM0~F6k5&mn(p;B&qN)nEj0pT*0Zy?tHh+YY5+}GLBE=5! zR5Nh?sV+$#xL#Qg7Xr+ZpS3lBPMUT;8vIG8J=ZGtIvLmEycsUWbt_J5guZHC z4@LVi6D_4J$4xc-M`gyIe`M_0cfGp6fTXfS#tFj@WeSvyBBD$6{nslqe{hfInR!zFrmUr(k@PcV!QdqRujy8j0POPygq^6pm%>*8aLl+W9Ls`V_6UqJ`92Zso7n`{?Fwf(~@- z@`ldWmJUv5aBA7Xe-=*f994kNZO*=ARXD>}qs~ZR~Vhp{ysT<57FYaif}z+GcWi0#K;y$>F#;9Z_a| zIIB-*b=zXxLjc8LhK|D}cFY(q=SMPJcuMQag~&2~H(h2;-%D_R|7i8#Lrb5k%H! zQac8VpsxmtAa@HEKJygZerX^;ML`Ez(VKzYBGp`|^kB01fA?^2V(sJ!R;y98^4WzXiTlRoB?siyv>|oon znpDV!ht-~edu^Bse@ww3EpmAscsB~;+|4kd+OuYHdNya~rZenS26M8+W_tM@cvZ%i zlxxU@U7@EhFtIB%x6X%d+FE@F1`$P!%HFPR{%%I89g_-}c-zcAjGNEQEMd>HA$&ti<*&?}-bK zZIyxNsFL16e^D@a2m-7DnZY$6zX&BA%uwAaRjR$pZcHvippk=L*E3D`#1iKL5iS*h z&UcaXxx!PT2BXu|GZ=cxbO!G1=nW(Ej_Bcmlk}2KZzLWy_wQ!FZWw($!qakwOvzQ7b^WZb3wo*@f5`w4B~36_ea-QBS`n=z&WvFS zv+V7$k8%0fslW!YTWA1c5pf6GFto(E9+W&SVdPXtHh43k%o;uv~y=*HM~C&z)rrYLd}=YEASFKu6P!%p-L5 zsHj)Mf9|WBrJI*aCF!NE-UDybN{GHdoe40_hN^gA@-w|rH;8FrgQinfuHB>_POFU_Sf1QNMU zZQK6o-Wa)Du8{?Zkz+HoJ92LqJgqgYCGeZff7v2Epf3o(VIX*+_al*o^D|IdwhCAj zy22h^Dr;Dcu&z1m!pZq1y?RM5 zuE;MUAe~>%1V`Qy0E9@KNA?SA61lP?5RB}nK1w`EL6RRyhKekcr3^`OKw-M!rxz>tR901UVUYAV-tcFLVb7yFpaKh4?W%Rm7w~od# za4Xf(AyG(?gEn}TlE#RP3k>Gg5^!0pAH9KZ*iNI^gloZSoTY#gNaM|OtOc|+i z8QD^Exn326pc;Emb8W-I|HIZFz9=(le|AsEDi8n-138{@z{_jhr3G0BPA)a0IECQe zu>KgbljZBl0nlPDiBiO0(e(re8E!8!GEoZPQJqvLD>{kqXJ z9hb;~X@&cbJP+PiLs!p>$ut{E!oU`z#I8mNOM8uVYo}g%MrR=1eO$4%rp)b@-m{k! z=?%So(`(5BxUy3+9rDL|q~-oJj9YcUa~8AXY&wZwTYiK5?T z7%g(0+P3}cdUKh_>z9sEyL-a%EUL(f9C=^_$AtRX_b9bbyiZAltJyEv(39lY#imaw z8ZJZ<83-sOsIDG8L1I=>Bnk$GJh4*jE*%$K21)3#mbt5Ctjy83d$QDfe@@K1IvvbF z^NQDaIc=wp9BpOj?(%9hbZ^>$a>Vs^O3909eL&#&JKFjz!LnZMoim zp6mR|lHzE=_qIo|ZZoS|ZrP{oM&zjm$ON^Qq-ZC9ctd!SXHJ3hy71wU9jcgA# zt=s6(kOw7r=4;+4TEm$)8=+olAwNxqZEd|O+;2sj+pWufNGvz%#OggmD z7l)dD9{WBfp{^m z&tK8cPY%TEt`mY0#yG2HbolXuxGV092jWOP5>Lcm#Z&Q2{34!WLVMk76v&#Y8lsnbcw` zrgSVeBM!v0o{gHJn2A=jR?W{IUeF2ls3E&Ou1^=^YIeO!wI$1? zuRN`zd_%7_YGq%cvZqeu+Bc}v_qAzMUz_a$I-|40d3`b8kTZWLwXK}5UOs*P;#DVS zX5w2y8Vw=yfA7RNFYSBh6{|5{vzkM4%1%&uqo7Yjc?ByNSLY|g3Z^i>vcCD0 zb6dHe{`Tnkn@zdH*Tq(y)Qa-E?;>^3H%dJclTG35ZG`*n^}~B_9{0fZtAbf0$!f`H z`4^$OhXXem?G6@uml5EBc*s~whv+}A@b+&j#C-|Oe|_Dd?#W_2uIJZ)FOM(k-xt+b z)E5Jo#1?qGC5Uv`JRL7u@w@naQMY_T|NEku*K3$*>rWW89bH@j@HO~ch)eM=@vnN; zY%b)j(ERO_`$un|FBkF~25O_Bb)=sFA)~-L+CH9d#?820+zdA_XVyL`5;AX!cq^d& z+wtpHfA@};BF1ls=zJ%lO>C-ee-ZsxEH-BbODoM5dR;~^Xr`ChIF?b2tndpH%;q;4 zL1`-+%;v`CH9Q$#lM`f<>I~|tIhk95{Vy->A3uG))OGrXuIPz zhXOC5-DO5Vnmk*nt%0_3g#R+On_FT3-B(A?f8M_A!DD9jCXMA#O80cN^)s z*828Q<9-(}NGZ=ZJZ~KcK72K0?A&eot*ravss^t?>W7m-LqCOuRHF|cY6K1XLj0W$ zf6M4IE{s`*q@_<6CiTO|bx{XW=ZKS6ypy71I5SJGzl7V4Zw43^odXevq^uE^nbm}S z{3?hQ-hC?y7f~+*F*dU~78faWieFpeW;mT9Qua-eq3Zoss*7?nmEEqF3X}~NT6O6l zEG?@JHYuL45Om@uM)V&G#F+lgg_zJ*4ob}FUkr##p61^->nod=Q5OOemwcfEf-5pG zFgY|iGdU|TGAA%FCn*XqO>bmGVRU66C`39kFfuVQFfuSOIW##lIhQG-1HuC}IUtiE z7ATh%qXQQaXp%yMpdk;ZzHq@ImWa*4qL)6S1Cf98_hkZMH=E|OG@A$EX?lO3tdeY+ ztl@7Sh07SlBDuHazg0Sau)uZrU$V->JWE3REyuox$v>NT@g;co?&Yk6$?)rDu}CbI z&xF%_`MpS6iSFf@4Z`u}t1YB6k|1oy>%_f*u>WDy>rMXb4@UW8oRR2FJ73I#BJ4C- zPgj5G(&j55#Esnu?dfW5I}h;+s0zYa zdobDrW@=!%5lkEd?E0qLzCH6QHcIB3MZBuB(U*!wJvMp_w#(7Yq&?cP>1H%Ws%VTi zl`T3-**=XsqyEhf%^W~uRW#m+M%XdX-cNsegS#tlM9*!!NT<)BiJwP{je|EI%~YSZ%Ix(THY_f%fsN)4p|cZ4lds?!Pw4)@gotJ=7+hjyy&u_J~mh(U{j0 zPOkbp{4vElqVJ{So?Azh&sNKq=H1HW<>q%L%C3rf+wkC7rNQF<17Ed84=dl5_gnt$@XGANn)R9Pwt*gx*F$mEiAX6d>p@o z^{LvXt6NwnRnJ-|Ht*Srsmootw`fK4{C=67rQQ2>PnJV#Y|@r*n>4Q+yS$54?b8;= z*lS->I$aCC9J%gsKFehHv@$)@DTn%GpI-8QY9&=EWxZ;~a@k!UOcu>v8>b%J{rr?X zYaDO-^;z%3oPT}Pt3Om$X{xVwZBtT7?R8OUv%@||+hiZFN;-V5FqXCU(1YJ*my-fj zRz5j)$Rldw!yR5>8z0^{mBx0n>d`KDjpzG*-z&U%r#$fIlH2OH?S3w_{1uaQC;Gzq zT9>QeTK#&S$Jz0pYWGtPpeeNbD6OJzWlbn z`!<&{B&WSt#vB*qW@hPPVqj+BZ0YFgY-s6ZXy|NiU~1&(WMb@SXzpTer(i=+Ni1w3 zp@OlQ0XQGlt6nnqV-6HJ_HkxxkiEgNueMLP&vNGD?PI2~7 zosmD=TahWX<`S1b`|6|j!ZC#L@VQ1NRmu_Ad^*y-x3r zy})|+0%xs2Yn?=c=ZvPp7onYTw{JR1G$l$j21+zboH@|qrIV(KZ=%Bhc8O^-77F@UMwkd z_CbcIc+-Yo{Q3Xi{=K48f5+GS7guUtN^wb1YGN*km7kQwRa}x-R3cPSl$yq6WNc|@ NV!@@V>gw;t1pxlafZzZC delta 17529 zcmV)6K*+z8(GJbh4v;1RGLw)vD6vn70e?zG`)CuYNV%(Fr^CtOSkY`LqNUHJNfwwm z5vfYq6W%t(Fw~zTZOUS-$Q86mjgXj0EF>JdbJ*6oo4FQAiT*Vdbbjj8pZ@)jKz|8xK69Y&J6d z%AheB2ptvId$uVA)J^sQGGt%bNGf4)m+l16>8`OyiVYKC)7j-Bg#079mu~Nxd{B^?-FOd#W3M)}k4@(+U#TH&u4Nk=$IIL1&qWmA{d*qkn6iex)@; zFf&yWEPCk>)V;V-jA68c_^Gc)%QC(~$QDl_y?hL^a*}eSxzQ)(uo$nj@Ko1l{lfJ*1}^nZj7X5V<`8hx_-a+G zLiw7skFWsUn>!PU=}hokH-ESrhp{gnJTNX1W3in|q)3!rZ-;a5*rFFuvz(ZxfMmIz zSoYm8Y#%pq3h$~bZ!1!)7^u`7S(aRoS}au@0aPiw?d8%h`G#-d{bCV{yGbYgzPW^s5Le9zMdcRLRA%B+F@mC}H#9FWC zsFyKYtLWlP{l%GzvIIhv&)gcyuD5&u0;X%IdoArC;zb&nHKck(!>adFThiQtn+#L5 zH-m(p(cbSojdfH1$1<*ypl&aeen2o0qsSbiaKyj6ZW zJ~*1z^&@m6LWZcz5`S=v(Y|b8r6ov&Sugiyvjrza%)0!h^kwE#?3_j^a@mw!_AI?N zUUP9M=+KWFHWo=->O^_6ATH4EMRu_!?TbOFV$_*4(xe{6CpDSyM;j)R(;m99VJ#1%tN0&KU zWmtQDUzO9~Vq4A^iW<`9v2fw35XZ!Wd~#HC97}&Sj)HX4z@4}!?i@Jjy9@}v|FN;f z;O9B_p&>ruaVkZ-JNgl0eX7wTrn?GTD#6<&^`_|Y#D8OS@%+Gq`-wEGED>_X6~Kqc zjq&&9Us|^_ELHsE!pwnUXdLELQ20lMqlX^l9_ZLUV|RmJ;b3R@Wgmu!aQFm#mAx;f z#;-Qmaro-t2=`5vPeb@wGC#>O*7viL(MwLV46is7Ng4y@Vj+E&vP%-l1eRc89Ql5U zS8`Ey41dwJ1Cxnw81gwC;QXY%8Cr3># zRaaU_i*)1D-W>p?8wm5z0IhxCQDj_2XjX;`yjCle|dBm z%U4TkGa1Bt%lpG=*I9RIE_H{SNsl#qOOrrw_C7cl=8}+;W4b_$yt~s_c6Zds6W0;u4i&8E?QXJRO|!|BZ2J@V+*=9+5=Ftv`tACUe$puH{H(t1q!(SLjJ zIRbfs>QcVD$y=~S3Ru6MNUUhv|_ z=9SBP3i;F#8-I)m3-nxl`HaBM+JCk@OQ+B7O zQZqTg5?W%KA{Byi+xhyu4HU(0*_m7<;9-0F#&oq?rK`6uQu+Vm-HY2dx?a_)E_AWF zyI-ZMu4-6ISJ^7dD^-;Fs>n;FO}@H2tbR>`AJ;d=lu3V&|3%Uq+|CDv4gP;P_;m)G zy0xy7vFUca@M)bFi5r^!6!_G=cjIaBGUL0-WIP@EA*l6@wRLj0_rWvV4W2h{o3J^B z;J534-u;D;UE*qUr7K3YVI%+-z^M6z6W@M;i*$)=@DHnd&w$o%5Fws){piC>d@@Y} zoc5k~-}eJ=@Sok`*vjtdj?jNLTOtX@_d|a`BwrwYsj{nKw~E8}Z(po_r3)X|x`x{y z_>4{E1WN}$v72~k^n())!F&T(IJx@c`UY9cll%UJ*KdD$=EjJ5nk`Y+ znKC6~j*owxSH>nkx($kO-*ohpP2PDobfRJwb&v4qLo@cn7Y4Fo*+`vD{@$S?fB=Wk z0GcC?7kj^VZ3BYt8pm@t^8XLRCw%&egowZ;acQk=CUSYRt_%pd#cK>G25A$FnDr*v z^j$c$W5g8DY64Gy6xo04x~&}A3xuil0~ncD8^h0%Qctu7wWVI=RjF((tpCTl(n)^; zJxi-(gCezQnV{tHMI3`kdLA_$@Z}&%h5H?f9$dj``enulh* zA$js~BK(rm1V{U$ho=X>dVMw{yez6!Ugj#dnki>hs7Mt?99)0sfm!H*5j`+p*Mk|e zZKcXgukyku;5fhe=N=XAku*>q2RCEk)dpU*nFb_Khfh4Ajs2n{941gW?I4(-f=NUnH%goOQ+x}Yti1KZJ+Dnnqrh~O zlNk*ke^U^j10HbN$%Dszvq1^%dAfIj2W%$Zcd{QO*7z+?8Uc!7YYI3^AAF4rU8+)x zIOc|bla~bqQ1TT$QwNDa`NI_~;m5B&^>koLOMrW+cntTcV6K1-s z=9mdu2$i%3)LyZn-o^b3Csm`9nqM596b22)ASZ8&pQ3A+mk4cub5}^Qb*8FZgmDpV zf3?-KXp6Ha(NKnljvgE(If8DP2YBQyeQvjs9&Ujzs9g)JKRJE17!~gr_ zT1t2c2~LIwKbevt5-HoxF-fs8NwLo)DVR|9gP)R+;LlfttHICaxvGWbzhI_;r}7%~ zv2A%2MmLNRNCpoh^s9LVO11-Z1Lv>+h7k1;EKK$gd-F=7>GIyep9sY~x+8#Xe|qx6 zoTI}$=IGBFgS9G+MGrholK&SJ32-zuNydMKJ+m{_kpO@{4ow6bc^}q$6v1Q1&<`Wu z$!6vPxvCAXy&c@PIV&R^eeJ_$1~A~%^7O;o*YKC|4kE4w;@}#R$L$-ltnBh7ky)iO zBl7uq4Z(+nQ2~4L=I0 zAR$6H&Js+Mih9y0Tm+4Rg_b9sq;GT{Twv=qo+nMmJBTd^fJ?ir8@AK)e@*AmPvWK) zqV9M#CYstRtwksSY*@Q${ z3*`n#7O#)BE#_!$9GcG(GAYDOajEaF!ToS@*kf0GkzN&S?>@-urc$(^3(iCX-In`{dIdOqW}+p83nsWMr0 zE?aL5D9oWhNWOMDFody(0c4}dAjlQlW~w$3KFGgdH$U@fom=dd2y%g;;WXuzFmnQZ z|KkVF+~v|`Q~>J+AE-=W80S~7@y=H}ut#>_F?h+skBY%}v0`q~f1$+Q2+XNB5=+@{ z_Bdb{mjTq{fa5luie4hlasF$h%)~9^!eF5dmk|(*1*+D|0Ap;SpdgGMj5HF9)XV5j zDQbwZbI}wq^Ney9&C$~4f!v-=%q&dIusASqp1RFAvxs!*a$%FQxd?+X(dXP>R5);@ zpqcsSE(hdYE_SA7f2W9(pn32sy_kU)`2H+8G+lG^(T%J?UaAtI3xcxbaugkTPuvoC z2YW>Ega~lq+5E$3t!!OAHF#uAWNg@gEigg(xiiyVc5|F$CM6oXk>bvZv14ZmkqOeg zI@c!Vq6EQeo;-FJdqgbg=z8bYZL^ihQk7egjrb=F7<=|6e>9}x)@J7hnC0@WPGn2} zk+rv@0?(f7MHn&*U=aI<&xh%VmdTCEArf5Rb_(Ltv2YL6gNzH55vj&5YE*&K!P8!h(j0^)J1w;`)PfXQOD#oIC=0VPbD}7w0?MC zAc+MJyz6%jCA?QE3Z19$FARA<(|BF#S&Bv}PKGoWOmnFt%`l3MSsY7KK_@?d0TIET?%68A28n~IMcQWgR+T=tZf8)NAy*yVbVgPueyeVjOkv*vt%R2cO zE5)?LO0h_$+P6ecZnD@9#0&$Qq#ZB)xgNmBzKkXv&k zC2bf)*_y*ShCo8_l4mI7E9RP2H9S#74Q@nRn-KLgwh*Cmm7X7y1ArqlC#o|kQM`c5 zf32Fzt(VdQ4zrL3#~-EEK1V_cnX>|^#$cK!Q*5MJKag*tvJ-*A;w!aGR+cK24hOk5 z`7+U?e?}SPcJ4xP+8mA;(0wd-Sk8IOrD5>VQ35H)kJO$$x?TJdiG|H9ErjX$lPwK z3g`624E2zw^MuFhS!W9;Q{%&mhSJISnFlk?M!xez?_^m4fY+{j;IA&%WNtH+Rl>fR z=Ph!nNGHhD4#=MgiV9Eg{zF75@eV%I6v32^KLrQ&Ndl=X1FQQK4(4C7%V|iX~+HZUKCdYJtnCX5;6T9~PfsJlAWOug`k3B7Os| zw=}C8Q>&@~hHc?3eX-?FcQ5`6L|1<*vjP^=2!9e1gBP_GeU1VondHnm0QVMX0Ak>$ z*q$yZVhUS8Xqjx`5)kJ<0LD2t6a&uL!k)OF>p88$QCDouN6;mXCnZVfqV{}wbY)#i zIbn=E0BDiWR9_%-F2la>ei>l=%6f!)v!4CN66npc35=#a_AS*y(AYp z&3{9&glEWABJxycI715y?Al0SusDOWss)}VB7y0nNxSNEwe#oSj&wNSzo8v`PP%+| ztSH>^BD%n}hmc<3IdxJw9GtEI(})bKxI32_Vi}Ny_S}I%B+XSQ^?bz_PsVsmIwO$) z9v1cG4nC?Ik&Srz^4aZHMhMViib4pq1pFMy#bPNu8k z3n!&;2QypKBCvd_dan_8Aq}EZJ+$v#X8(CKQar1HbFh&)@HLZja4b3eyG&%9E=Z36 z7?}U?x+N&cd~=gekGT+@?5k1;@{OgCxti|Wm5*K7b6KsQWgrxMdIvBK1}3MU4=K1P(uf|Fn*oL!|Q~JebHf5ZlAr5xLNL$O`tcO z(+iH$KTsnruSHDABrOPOyyVwd*%x4cI0-nCXgOPErr*jQidh0OYqS1>o8FvQtQ#9j}9fQQK7l@g%sl|Lp{0>CjFcpYXz*#0KNmhlNnz^ zPv!vD{6D%nApDd4A}IkflaM$lvri++2!A9Y^q1xTR2&zXC=$8c`W+>GjPdndU1H|O zYM|*J*AI|;%eGo!QB1vcJN8I+D=Xv8F+99r(a=3JP?DKQ86&h|1AQthD8f9eJpPC* z9RA=f8O-1dNEjgFK;y6jb5!7JmmX%NE&v^6ct*Vu@5p%DV+kfULczW68+sx8ZD4!)24424Vr3u4`Q zpn?ATuD@7xO9(q@*@0Ynswy-o1b;|Wh&an`d3qtc;ng0K724pNlF6$Xai&cUBWaCL znx87)*AL;DHzn<*1(u!uJkW%?1IoQacpBxZA}zujP7$||l`;1-`D~nOj5`2FqDlRy z8F|bv$#mB^md!=nk_L*?jNdsuM$Oy`w77TETj?(x8zQ4V`nyI z(>4szja;ILP!d9fE{G_UqFj<&#flUnl0+iH?@XUd_4t1N{5|&A<#m6a=ll7354PLh z5pIMb`GMvnA{CBAsOkbXwzg<04n;*1@cu|uxC0mvNgt0I5mE0ghk@6%6(R$pDn1!vzB@Kq5$n;uyd$0Ro6G(*c2Pgw}PEedN2TEj=+nPBUIl0@LAypX903d-dkW9fd#{Pj# z4nhX}3k@p8l1agv0{{gal^UX}q7oh+jtB^)P!S|@03w92iO&g#r+)z9Bytb{eUm`~ zxJgVX5d+al#eqK?Fir=s!TWy@VK_Z&y4<%C3)KCiGyXk}8AV&TtKoBqqC6gI5Z2u0C{~7Z)bYl|q zWS)emXmt3WCx#}5Qhy?U;pX?p_9qc3cnX#Bvm*##@dS{8p29dYJaN-!Yh;HqH*<7? z+d$eQ!fib6ERrEDPTfFRGf+U!=WI`^sgGo#r3Z*0Dq(cssMNx0BQa>6~^dI zRwH?pAl!qwhSKSPO5Ho+EhmM1R9Z+#` z^gx+gDE)_+Hj~YW{v-^Z7yzhgXaZ<586Cl;3MoWQLw^H^LPEmFfV52s0V)V0i3*hf zA)(Z0081isF^*YlFQ9^AxHc_qH9*DTA4^>wQ1OSvkN(G_hJ;+u-TR9}6;KI4|C)nT z)do}uzhp2>bg&=hmlQ@7@fTAQP@(>7sv!ZDh)wf%UhEmy#^x5N{!7$f_vxnXNF|ek zKvz5l`hPe1B?TgnjHew^g(QrG{Lt&~UtYfrfc?9Dzf>C=lW0-!y=nklO#@<9Rb3m< z(ol{5tdftytt z8~tKWgP4eE{-`u)SaiJZSdLr~$;MLGE9U%h;(t{)Sgasn{Y`zw_3p2jgZszOF}5*c zJZ46nc`gX&6FIhn$I6zM0llgw3Q!~$~|FO!Uf+>Q?aNxPwEZ5-o`T(wRI&o8JXh+kOO4j<8LH+OwC+6aV2 zPQ8kD+cN?qSvj;G&6PfFAl^EZzV_x!gnvZk#EQt<`YK~>T*Mxwc5WK|J7$aQ6StoA z)x)|Mw_`Q74J4m!H@mtod28E-7xvu%-T#iuQTZKpVSO8tVPAfT4zie!^Y?3&h~V=k zXE@j_e{@-U44SFM4+cGvZ{B~ks`;C&j!xjGwyYQCkC$b6rB6sDtmhP{E3sKk6@TiF zXO_PhSPBwv=tt%!2S3*;IU{JMJ62ha#dJ5GtItLdc6eNuqp=x#%sA+ZSYFVDv$#(j zNVC;2=vA0@rB0LL->mgE6nYpL6c2bCkJVrubEpD0Ip_5yZmgy4m9Oobjq>qx>#jeo zkhBNb)o(hr*QdBQ;V72@;kjxVIDb~0qeEi#jERl+@$lBN3|;q8Lb*f|ljB4W4&FTw zxR7vnV>yMT61#A8+q>t_0nt$+TS)BaCj(8tTBgJ?=Ns$QnHH9<%bG1(4~%QA$8JWv z39YarJ3o6c&l8-%)R)snZ&3O2HsuVZ%WGOpVlN2I&UuP%3~$+{}xVwfYO_P-F7;m+gxJQ#R1H9sqP9#}zI-9~1u z6OSB_ZnynBig%cncXQ>OL9 zQUrUdcSGL`VahBe3EQ-ksvvY>OkDR&`sFCks$t=oLUiCLi&L0BOMhbElfq2Ufjo z4oOhezf|AG)@SLC93Pm;^nJ#*OGJ15;7*VA{P;aE2U9(sy%MLjUGpX<>#yX_2Oa4g zayh1m@L*-Tb!+m94}bft@W?R5t}2Qw=}Bp}_T2ZzuldH7ZeP6Iibs?6yA!v4Ka?CS zG7j%&A?>>I<YFPw%nMd?YclE!6*$OroIdih-16CV4GHhm^A|{4({7Je=WrCP z-&FKWigvP;U%Q;5KD2m|W^yFMOX5gJRkk@tcb$-q!JAC;D!NLFfMF-|$0CREl{7xj zCzGu|*oEu^_N zxmLNs!1ww>58RU6OXj~s$bMSEHcSRT?s!c0gYj@*y?^sqo+7}4<}r^~et(1eR8#$W z_vel4jT2$?l$U_D!F79w)N%A~>oxq=v#G)6ETu5pa zA<>>?zy;oYn4sOKu|J{jNR}1o5W-V0^G{IE#!UIh_+MOn;8D5HLtLRCdg#oGq24_o z#S51rd4F_UJ*gV|;R^E~Y!MUt^$&iiIADS*vbsK^8WTIGSe@6qb;w7r0>G-w)zS%KVO3%b&c>F9T;*+=38ha_LjFit#;2;A*!@pgLKM1CFi zaab={dk?b@)jNaVo%=Ejb_U6pa9Z+nan?IO4S&UDiVY7ZNuoHwJIN&1R;q>7I*RGs zfrdySQ(DDwOUejmfw?X@*a@VmpF6CX-o0Q<26J=U_eS(&MjenNN zq%Y&SW#Y(Oou7ncu4va{IG}|FM23ku&%yb99nG+(ZfQ&#*FN??g_AJes3_GTqfeD} zhkqm_J#I7|e%(8k^j)si=RB`w_(FfJm!l@qd+y5R=iNTUroAPLUF>&*&Uf=^4=PHz zpvq3^IV3JRf9g?5DX*?|aPujxDb(~v*gfM+nvuUyQh#d-HW0g*ky9*MdubLI%%T3g zJy$a2HAiDBpHfUg?ff*iDSLq7$TE3>e1AQ07&tO?a=x!}Yq^OD>ODHxQ29qs^~%W3 zwJn`v?VhK4zfzo@h6#p8c!eUQSM*!90Lrdv`K*gUk|lb2_{r^emB?N-@}jGB2aUEL z`%r?x2w70NY!ysP5{uI!7THUqY>c)gs5KeBKerY;y&}2I7p=ysY+D-Re8l#cKz|{z z4t@7b8cT-Ry*~erEf;k9q?*QOdGyS;+j>*>P=}_LE@XMuv7?H|m=T2dlX~}xWj_Lxjc@yb_Kc1N@*DSKwJ82ya z?3)}Ua68a1KV8ncsRp{56dNx)r+$>%oB1goAn*z9P(wVspI?tLsTn`W!#}R22cX})9n0DdCelU{;IO{<}5R+KblW77- zO}^#Q=KK1&ef#qUZNtX>MMi@W-r-n&=8uPc0}hQIi)~+w8&m8l60H0{9Dh~{d@yz} z+xygI-uOw`LUW(`4;(6V7xTQ#g>{-L_sR27q2u8psg@~2H8HFO75Lg9bJN^_$tHWv z%Es=JG0Sl#oT-2YO4;$U#n5&7k}{jD!pORhcM0&AEs>_XKzNiOK7P@GjX2uPQ2Rn3E^yX zxN%S2DMGUV|NW#)Z-KqR+pb_?vzviwR%tsHh&fTG43oT;;#RD zW{r*kO||9OtSXj*P|4gQU++IG?b;R5mKl#N)YZ;CGyTC;uZjuPr6(g`aRVc7tfTJ| zwK&T4$>(PCqbg&Ymzl%gkESzg?*4c`G>z2L9zHt9HCIA?rlit^DP|e%oNBGefTyNO>p)w1 zT)UD!?Negt@Hd@`EkXmBDSF2xl$m|eCB->*d8txeVfxZ8$A84$Tj#?ozwU^eMSPnG zbk2Q$@)UDOHcBh9-0S{vuEz@5Q6X6orAd4gDWNc`>5gG~+=PX;RbsqB1NDuMaDGoy zeeCi17;U}cD-Q4QkIo!AMD#zs_VRYzVAy=N&%`Y2+=0uY6rIO&-5M>zj%S#`PrtBcc!}CZ?Fe-=ouxi ze0qY@Blqy;h;z!dwpH7d-K{1bs7<*Qf6_=acuTtfB7euFLDAzu^mF>JFcLc@N?A*- z*~2W^T~{5}TGuZL`Q){wAT+Swcgvp2zS1X*9lhUl;q}In_za9Cft_#8VgA+FYnFUs zqM$+I=&azgX*-|YPh6$dJ{l{pm4g;O#;Fs9Sixfa}9Ru{N*R~T&>p?575_1R}z#>Tdq)P~}y z-C_;7F>0JyR4?D=NDUYrAMqg$y_*iWW~y{YE~0N<+T7Acbt7}eCz7NRj=vf z?SBOvL(B#IujKZ7n7zh-EaiDp-Z5VhP?LHn>99HKeffT6O{U782X~e$^6}Ar<7llk-Q$$k^%aOH)={Ql06I6nKzFsJ!6#gzZ{% zX<+kJahm=5Tt3{H_1>ybQSM>i&m~`^BQzQfw+FPzHN-60VwZ6^I5xQc?W0~-_qibl zhs6bVNv5**Rb45q{4bun$iILk{2xStSdoWgIsu1cIs&(2Is>~?0y8z2AdUk*f8_PX z-Wc+(*M}Ivh#XUlDGrxI+E$)EUsVsup)A=H>Fgf_KS1^LOjmbTRabpg-5g34vM_~F zg)Njzg)5A6g;s@2y6{CbvQVl}Qn^BDTwN#>K1#|$8(gh9O1MOieuXkN7HD{{K z1PUw203s$t1uYI7DH;v3f0B`48bcM(J7_>jOaS$ul2D2E2nK|q5^RpGOG>JdeR)L& zN0z0?pcFlaLaBrS(yWT?PkS~@%%KxRLJN$~C`?6G2@{q_N|Qm7jEam9RzV}6hqVw| z!Df^V;0OI+v9#tqqcaA48nR;MMN%tSFpNcc4SRw!pk|~x`r(H{e>^Y5qvn^U5D$ts z2EKG!h?DW+ZSj}Cp!rDrayhFD@vIuvN8)}nUDVS>ixG(MNSxMfvz!m=wn+KG(|S0r z?lqr^H$>qTt*?H2g#L4Mgr{V(yVGfdVcx)JX+p;(R6rLdDc&A!&CeZ<#Jy%dtmizo zd@FtwkHvl5K$^G2f6HJ|yiqom(veNOQp0?()Fr9Q#ZanfLHB%lw!oO;$$0u+-2LGP zopN07N@ng9hH|4Yu?P56`*K<){Oi^A`i)pnG?=D+$ z^6I%!&WZ7KSbsXMr=#+GaWVNG0_-P)mac>wV}VNY02^Xye_?w@=`uKTI}uvQFd8q; zmuKamxxflu%$}WJw&Ou7o>qhN@w9$YSM%w3Iud7-<_u!B3rII?2IASg`6u{k#lsKP zWLeSRkLJ~oNHT7U>CKry!_0;Sk+n z6ydY&>Ij<3b#>CN@9KyXfYe-fSclZr5nr@i)qR05f0XNp4|+8wuwCad!3LSf;CYP| z1QtFxDa0@Jr^Qu}x&o5MUj<2H_e06RY-B-nzR}%XmdIpbOh!GswN6P?YnHK0xUEpo!-O5^TFVri)Of-)U6oR2z}L}9*XvTCR#~bPMT`?e~*kk|H#;L=z4X50ZCh9HkBko(YJ=8@!UXK`y|dv9t0X4s8AqK$71O ze~^qq1O*@=B`+gw8<7EFSEJ$7Uhm_)b#>OjC1;mp((JxHf9)F2>uPaP%?`~_6%rS0 z+7x1f#J)m1RVM7PuwyQhvwwJOP*>%l?85Dy-neU$79$~=-HBEsGf&Fjl(qCTl78mV z&tNv0%AVPVbyu?uYxdleW&UhVCagXve~U0lmKuUugs5Z~OC&h330R`_t+!*-dO5Eq z(|Yl-nZIwV>2O%rvq@(khr)}8QPai)yDv5XHj>K!m+H~*i=J zgS&Lq!ClzT!Cl@MxbE;J1DekgyTguOv{sx{TLgjtMv|7L!`93qX^Uo$H0J`cEic%! ztMM_mv9n2qvYw76Cn@1(J_G=?&m@LvUc8ouUW z;~J2xrQIAiNkBvj`{jG>Y!1!;1q zwUhZY{=gxnI%|ajld!U=%{HLAxHF(`lXeF@DwQVMooKCzr9^nyg_f3!zM|QQ)*>XE z-FPqx%OB!Sv^vNVaeg;iK$x}mpkT4a{woP!?>K8Kh-}WJ_6!z5Ukw&P?iVb4<|(+t z(m;TUf)2EzL#4C^8BJNXwEOvNcP_twiUn;w|1gHAV!RMYM+SUiE0BSFs9Nz2yjLQF zB@CzYujE)1Y11FdI!rXEoxOUSbmb+;$I)5|_B7OJuNrF9dmfTz zUPC?~4cYfXI_%q|)!KW+aU_=-NNY8=>;ZAy?XmXQ!M0^Jt&j~5s{;e~#xNCsn1VrC z?H3$ny!cKHo>RmPW;YsiFMp@;3zuF%{%AG&F4 z^$i$A6fr7$ySDke8Kw42Dq!MmGk-T}J~ETrjvROY`XyfTO^HMH6IIxaRwMhCb|+eF zbxW6aqczxbskj>tl&vMFcB2JKTX0jYz5DSZY^X*Xa>SAi*Ju#D=PK#_MypQOg_ z)0f`T%N8)1+!QdeXum)rX z*MR&Ylyopdb*EIR4l27bxeS3u4t`V5G~E+RoCid>R0KNTMb766Pl+0gPE*fd=qb|~ zxU-`-jL5w30Y8hAGUdx5qxlP@O0MVDL&`i(F7qt?(jjq(T|IyX+Q&MRJp`i3 zwm~l}U=zS{0iIn)X(By$mNlqJM(?RL(Ki4cO*b-+(AA@&UJ1K@uWpuZUM`iSm%4fn zyiIE%`T}((z%*N`;(^J}^j6&WbPyWq(NXpMpEWabuWL#B3tV29qDRAqh!O3T&)OQY8g9KoZb zRu31Ukv+F6cF@Q`1i119y`Eu#-ZQ*f^d-2B{8x1l)+20c4!dxc@DlItt-mk6lm zmo~wXw*(9!66cZq#hOGe?g$(s`>&4@Pg0QNUy>mt%Va5kbCMirm@a!%B)$7c6j`&0 zSAsOcuci_lG5GYYB9|MjSjB+>edsmXM1X5}L^_X#rU^eRy>UjbfpqI=ECaey9UT&g z6gilKS1DFwTy9py zAlSy<(_Gtsvi$$BSA@^XjHcZavI+!7!vK$`90>DTcj-dbfs-rEC{8W7H>|%%yhc1u ze#U*f(P^K!!udQtTOlp4UQh2R{idNef_Ams(aF9F_?hKJ?$_!2KHsD06qujtjXJue zaNE?G!X(NQOycqKg7jLm0J)w^o*BlVIg2uu%sJP8vyu)e^EkN-DRGS5l$5EM)sw(s zE4wqj<%K0f#>Z_QZS*oVL0+aaL9s(A?yd8CVQ@~p8YlPb&*(VIK)-JEOvfd1V_M<< zBhQ2P)zH=RVlvH^k}$BvD6y*%!qOgO)7q(*p3&JzcOTbmZ76fQrT6b8WdMfW%IURa z0dk#xON#BRB5yeT|)cW)D0IRi3|i3 z5>!`@o**%+C=vw&L!MYEc9)I|E`ucW*vQ;})iPG*=-WM6={+axU7ZeWpn1h>yqdPt zM~=2KboY5Jx_)a~YPY=Av5re#nN3^P64mfWb^t76Jma{ZOUEK-$F^MWKhJf3l}T~5 z$oWN&V$){UwcK(@*^S6$xB6>)>$V~xN9_z90Dw&E*?kRf#=D({($2_n!g@jlUY^=KGDX!jU3FT7v__SJ_Km0@|+vA4p_Ou=?C)NCVm1;|tOJ9XrNBM@{ zXw=HSLS;{#$hB`!r|)aisJ=GaC3Hq-ii`TwVoT2aUDd8~zIgun@yVlqPR`84*Mu}0 zLgwF(abDW@&MQ`9zGOAW2YsXLV=>(p&fZ42-(Ehv_v*(Ui~9FvH4*iv z0Zd{CRNfIvI)t80maX_*{JyMPKEMBc*(~Y}*tGQr4BC!AU4iip1bq^h;$Pxl^}N|$ z$Xj9h>&N#`UO!tcWZo9>R&e{b z)0Z#qovuWTUlGyyPDGp7RNdhs`Y%{)&JR{rnl1FYykOAGF0*m0Vi{TCXC|1L#Hn}D<$R^bp)Kzmbw}SkipWi=y@^Yo?^c7v(l{ryFgRbMjx=uHP{eun# zeuZ{dc>!tie66;B2HMJT{__ZLZiW4KUz|LB{c8^%GqX3DTQg4CXSuE&Ayg@Nr$#fz&zT!nQ22MZM_IV`m5 z@(=F{PcASkQkW5SKjPzX7sLG`5$ppaYc%F)}bY zGdVOhmrbDq8wfEnFgY_hG&PrSp##DLHZdTRkT@upU7`aQ5y(TKLC}(fQ(w5SNh}eY z1EZIPq63kC)Ayx?u$#`Zd6M1-;c0SrXID0z*){ylqi`9cSlGL~{I^Q(A9CP2{Likk zFiUNif6K7%q5WqQFTMot-o2caFd2T`EEYD$@|kd!Ex#9OE783y&4X~f`I;Bf8A%Yf zlx^Ug!O0;+9%jr zy^4+O{bmub>TL9-qEU~H9)s;_bTes>c5J#CjgcxE<4t9Yj#9RdhEx95;{op&KATf}HuW=X!}#8=gRqlrkOX@+ zbQht#xQ-RkACfuVYKkQYYw-|8-$4~jpbM8T2a=PfS#bhFiKJInE0`J=$A;p|%T-D*fbf8IsdF0bar#9CI2uJ5{H!)BPN*sx<< zRczQfuB)I=7<}JmS}ckrmE2euNh>xSSIH_i99!AUAX`sd_Xy?Oci%fMoBJ**pLF|V z+b2IXN1scKdkgxxkEq(d(24Qf%kG16A^uphWjRPWjmo!!Y;MEPYkRfH7fI???bXtz z4DXBPV1K&(54Kzi2$!&+0~D7orUPA4LPR$?FhVj$I50UkML07=IYKcuMl&@;G&wgi zH8eOuJ|H|oL^nAwLNZ1;FgZ6xI5R{!LNPW*Gc`msIX5yjG&n&%T?#KuWo~D5Xdp8+ zI3SadI4FN5(nU^0K@i6As(#oEHaLU3yZZoxySuaJ0^Fbp+yNHc07y6lw;(JC2{!=# z#pai&%62I&x1kxbi{1j&DR_IXZt=6nJRDWouP)-!Z0 z_zY5rlnIpg5$0@xIcLKx$c7xqg+fSG3BA}0@*sa73ZMvzp#&-*s*)eF&81KVhG zPX&96eOnb&Lk-kHJ5;JvCs>twXn;m&f@WxeT4;qf=!9PAP-%U&0^QI9UC<8$FbG31 z1=BDIeJTT2c6Jy>U=+q62IDXRvk+IA=(iIyd>en#+i(spU>+`E0nT9&me_19una4( z3TtMt4jZruTd=JXFR}7FDo@{*c2&#%uiR5@h*;WJ9gbQ$P@M}~I#hjow{)bg;L_5u zx|=UcC+a?(rBe->VwTP{cz7mxdc)Hjp5E{@gQpoh&ERPjAB7?f?yvva`|(450Wb+R U^p|U>10M=AGc^h&B}Gq03cX9@82|tP diff --git a/doc/Projects/2021/Project2/pdf/Project2.tex b/doc/Projects/2021/Project2/pdf/Project2.tex index 13ae9c648..efcaf844a 100644 --- a/doc/Projects/2021/Project2/pdf/Project2.tex +++ b/doc/Projects/2021/Project2/pdf/Project2.tex @@ -123,7 +123,7 @@ Project 2 on Machine Learning, deadline November 15 (Midnight) % --- begin date --- \begin{center} -Oct 12, 2021 +Oct 20, 2021 \end{center} % --- end date --- @@ -185,7 +185,7 @@ epochs as well as algorithm for scaling the learning rate. You can also compare your own results with those that can be obtained using for example \textbf{Scikit-Learn}'s various SGD options. Discuss your results. For Ridge regression you need now to study the results as functions of the hyper-parameter $\lambda$ and -the learning rate $\gamma$. Discuss your results. +the learning rate $\eta$. Discuss your results. You will need your SGD code for the setup of the Neural Network and Logistic Regression codes. You will find the Python \href{{https://seaborn.pydata.org/generated/seaborn.heatmap.html}}{Seaborn package} useful when plotting the results as function of the learning rate $\eta$ and the hyper-parameter $\lambda$ when you use Ridge regression. diff --git a/doc/src/Projects/2021/Project2/Project2.do.txt b/doc/src/Projects/2021/Project2/Project2.do.txt index c09edff79..a7d732300 100644 --- a/doc/src/Projects/2021/Project2/Project2.do.txt +++ b/doc/src/Projects/2021/Project2/Project2.do.txt @@ -49,7 +49,7 @@ epochs as well as algorithm for scaling the learning rate. You can also compare your own results with those that can be obtained using for example _Scikit-Learn_'s various SGD options. Discuss your results. For Ridge regression you need now to study the results as functions of the hyper-parameter $\lambda$ and -the learning rate $\gamma$. Discuss your results. +the learning rate $\eta$. Discuss your results. You will need your SGD code for the setup of the Neural Network and Logistic Regression codes. You will find the Python "Seaborn package":"https://seaborn.pydata.org/generated/seaborn.heatmap.html" useful when plotting the results as function of the learning rate $\eta$ and the hyper-parameter $\lambda$ when you use Ridge regression.