From c64a22ebf06d6a4b937f23c80d15b86df1831de7 Mon Sep 17 00:00:00 2001 From: Morten Hjorth-Jensen Date: Sun, 14 Nov 2021 16:37:50 +0100 Subject: [PATCH] added text --- .../2021/Project3/html/._Project3-bs000.html | 29 +---- .../2021/Project3/html/Project3-bs.html | 29 +---- doc/Projects/2021/Project3/html/Project3.html | 28 +---- .../2021/Project3/ipynb/Project3.ipynb | 113 +++++++----------- .../Project3/ipynb/ipynb-Project3-src.tar.gz | Bin 194 -> 194 bytes doc/Projects/2021/Project3/pdf/Project3.p.tex | 32 +---- doc/Projects/2021/Project3/pdf/Project3.pdf | Bin 236062 -> 234551 bytes doc/Projects/2021/Project3/pdf/Project3.tex | 32 +---- .../Projects/2021/Project3/Project3.do.txt | 35 +----- 9 files changed, 76 insertions(+), 222 deletions(-) diff --git a/doc/Projects/2021/Project3/html/._Project3-bs000.html b/doc/Projects/2021/Project3/html/._Project3-bs000.html index c79df5535..1144eb817 100644 --- a/doc/Projects/2021/Project3/html/._Project3-bs000.html +++ b/doc/Projects/2021/Project3/html/._Project3-bs000.html @@ -74,10 +74,6 @@ doconce format html Project3.do.txt --html_style=bootstrap --pygments_html_style 2, None, 'introduction-to-numerical-projects'), - ('Introduction to numerical projects', - 2, - None, - 'introduction-to-numerical-projects'), ('Format for electronic delivery of report and programs', 2, None, @@ -136,7 +132,6 @@ MathJax.Hub.Config({
  •       Part e)
  •    Additonal (optional) exercise, adding 30 more points to final score
  •    Introduction to numerical projects
    • -
  •    Introduction to numerical projects
  •    Format for electronic delivery of report and programs
  •    Software and needed installations
    • @@ -344,8 +339,11 @@ libraries like Scikit-Learn, Tensorflow or similar.

    Your task is to perform an analysis of the bias-variance tradeoff -using at least three of the main sets algorithms we have discussed in this -course. We will limit ourselves to a regression problem (fitting). You +using at least three of the main sets of algorithms we have discussed in this +course. We will limit ourselves to a regression problem (fitting). However, feel free to venture into a classification problem. +

    + +

    You are free to choose between bootstrap for resampling (recommended) or cross-validation in order to get the best possible estimates. The methods you could study are Linear Regression (OLS, Ridge and Lasso), @@ -360,23 +358,8 @@ of your model. Comment and discuss the results. Discuss the pros and cons of the various methods. Are there some methods which provide both low variance and low bias?

    -

    Introduction to numerical projects

    -

    Here follows a brief recipe and recommendation on how to write a report for each -project. -

    - - +

    Hint: when you use different methods, pay attention to how you represent (and understand) the complexity of the model. For for example decision trees, this is represented by the depth of the tree.

    Introduction to numerical projects

    Here follows a brief recipe and recommendation on how to write a report for each diff --git a/doc/Projects/2021/Project3/html/Project3-bs.html b/doc/Projects/2021/Project3/html/Project3-bs.html index c79df5535..1144eb817 100644 --- a/doc/Projects/2021/Project3/html/Project3-bs.html +++ b/doc/Projects/2021/Project3/html/Project3-bs.html @@ -74,10 +74,6 @@ doconce format html Project3.do.txt --html_style=bootstrap --pygments_html_style 2, None, 'introduction-to-numerical-projects'), - ('Introduction to numerical projects', - 2, - None, - 'introduction-to-numerical-projects'), ('Format for electronic delivery of report and programs', 2, None, @@ -136,7 +132,6 @@ MathJax.Hub.Config({

  •       Part e)
  •    Additonal (optional) exercise, adding 30 more points to final score
  •    Introduction to numerical projects
    • -
  •    Introduction to numerical projects
  •    Format for electronic delivery of report and programs
  •    Software and needed installations
    • @@ -344,8 +339,11 @@ libraries like Scikit-Learn, Tensorflow or similar.

    Your task is to perform an analysis of the bias-variance tradeoff -using at least three of the main sets algorithms we have discussed in this -course. We will limit ourselves to a regression problem (fitting). You +using at least three of the main sets of algorithms we have discussed in this +course. We will limit ourselves to a regression problem (fitting). However, feel free to venture into a classification problem. +

    + +

    You are free to choose between bootstrap for resampling (recommended) or cross-validation in order to get the best possible estimates. The methods you could study are Linear Regression (OLS, Ridge and Lasso), @@ -360,23 +358,8 @@ of your model. Comment and discuss the results. Discuss the pros and cons of the various methods. Are there some methods which provide both low variance and low bias?

    -

    Introduction to numerical projects

    -

    Here follows a brief recipe and recommendation on how to write a report for each -project. -

    - - +

    Hint: when you use different methods, pay attention to how you represent (and understand) the complexity of the model. For for example decision trees, this is represented by the depth of the tree.

    Introduction to numerical projects

    Here follows a brief recipe and recommendation on how to write a report for each diff --git a/doc/Projects/2021/Project3/html/Project3.html b/doc/Projects/2021/Project3/html/Project3.html index 6e029085f..2e06fadd8 100644 --- a/doc/Projects/2021/Project3/html/Project3.html +++ b/doc/Projects/2021/Project3/html/Project3.html @@ -152,10 +152,6 @@ div.toc p,a { 2, None, 'introduction-to-numerical-projects'), - ('Introduction to numerical projects', - 2, - None, - 'introduction-to-numerical-projects'), ('Format for electronic delivery of report and programs', 2, None, @@ -374,8 +370,11 @@ libraries like Scikit-Learn, Tensorflow or similar.

    Your task is to perform an analysis of the bias-variance tradeoff -using at least three of the main sets algorithms we have discussed in this -course. We will limit ourselves to a regression problem (fitting). You +using at least three of the main sets of algorithms we have discussed in this +course. We will limit ourselves to a regression problem (fitting). However, feel free to venture into a classification problem. +

    + +

    You are free to choose between bootstrap for resampling (recommended) or cross-validation in order to get the best possible estimates. The methods you could study are Linear Regression (OLS, Ridge and Lasso), @@ -390,23 +389,8 @@ of your model. Comment and discuss the results. Discuss the pros and cons of the various methods. Are there some methods which provide both low variance and low bias?

    -

    Introduction to numerical projects

    -

    Here follows a brief recipe and recommendation on how to write a report for each -project. -

    - - +

    Hint: when you use different methods, pay attention to how you represent (and understand) the complexity of the model. For for example decision trees, this is represented by the depth of the tree.

    Introduction to numerical projects

    Here follows a brief recipe and recommendation on how to write a report for each diff --git a/doc/Projects/2021/Project3/ipynb/Project3.ipynb b/doc/Projects/2021/Project3/ipynb/Project3.ipynb index 6be6167c4..7d98fd7d5 100644 --- a/doc/Projects/2021/Project3/ipynb/Project3.ipynb +++ b/doc/Projects/2021/Project3/ipynb/Project3.ipynb @@ -2,7 +2,7 @@ "cells": [ { "cell_type": "markdown", - "id": "92499fce", + "id": "26d52fd9", "metadata": { "editable": true }, @@ -14,7 +14,7 @@ }, { "cell_type": "markdown", - "id": "56cf5579", + "id": "236f6cfd", "metadata": { "editable": true }, @@ -29,7 +29,7 @@ }, { "cell_type": "markdown", - "id": "b787f690", + "id": "1d3c512a", "metadata": { "editable": true }, @@ -39,7 +39,7 @@ }, { "cell_type": "markdown", - "id": "7a40125d", + "id": "4f7124c2", "metadata": { "editable": true }, @@ -79,7 +79,7 @@ }, { "cell_type": "markdown", - "id": "8c7b1223", + "id": "ae4474c6", "metadata": { "editable": true }, @@ -91,7 +91,7 @@ }, { "cell_type": "markdown", - "id": "f5f567df", + "id": "657d3638", "metadata": { "editable": true }, @@ -103,7 +103,7 @@ }, { "cell_type": "markdown", - "id": "daee2ad3", + "id": "cd7e6b5c", "metadata": { "editable": true }, @@ -115,7 +115,7 @@ }, { "cell_type": "markdown", - "id": "07b5058f", + "id": "05c0260c", "metadata": { "editable": true }, @@ -127,7 +127,7 @@ }, { "cell_type": "markdown", - "id": "7dc2abb4", + "id": "9108ddbb", "metadata": { "editable": true }, @@ -139,7 +139,7 @@ }, { "cell_type": "markdown", - "id": "af965f2e", + "id": "4e2b05cf", "metadata": { "editable": true }, @@ -151,7 +151,7 @@ }, { "cell_type": "markdown", - "id": "adf60f4b", + "id": "e45fc3b7", "metadata": { "editable": true }, @@ -171,7 +171,7 @@ }, { "cell_type": "markdown", - "id": "4b7d456d", + "id": "f036e12d", "metadata": { "editable": true }, @@ -185,7 +185,7 @@ }, { "cell_type": "markdown", - "id": "c8eabc08", + "id": "52660f41", "metadata": { "editable": true }, @@ -197,7 +197,7 @@ }, { "cell_type": "markdown", - "id": "1c3ff570", + "id": "c1fe35cf", "metadata": { "editable": true }, @@ -207,7 +207,7 @@ }, { "cell_type": "markdown", - "id": "c98d31df", + "id": "a1395ad8", "metadata": { "editable": true }, @@ -219,7 +219,7 @@ }, { "cell_type": "markdown", - "id": "1c27b732", + "id": "aa6df14a", "metadata": { "editable": true }, @@ -229,7 +229,7 @@ }, { "cell_type": "markdown", - "id": "4f88e787", + "id": "09d87b1d", "metadata": { "editable": true }, @@ -241,7 +241,7 @@ }, { "cell_type": "markdown", - "id": "b4fa9a6c", + "id": "1dc28a4b", "metadata": { "editable": true }, @@ -252,7 +252,7 @@ }, { "cell_type": "markdown", - "id": "0787e8d0", + "id": "9e6804ed", "metadata": { "editable": true }, @@ -264,7 +264,7 @@ }, { "cell_type": "markdown", - "id": "f1389cd4", + "id": "004eed40", "metadata": { "editable": true }, @@ -274,7 +274,7 @@ }, { "cell_type": "markdown", - "id": "08aa1c31", + "id": "877b7ed0", "metadata": { "editable": true }, @@ -286,7 +286,7 @@ }, { "cell_type": "markdown", - "id": "7b04313a", + "id": "cb45cb03", "metadata": { "editable": true }, @@ -299,7 +299,7 @@ }, { "cell_type": "markdown", - "id": "aebe353a", + "id": "fe999d48", "metadata": { "editable": true }, @@ -311,7 +311,7 @@ }, { "cell_type": "markdown", - "id": "cd4094b4", + "id": "f6ff7ece", "metadata": { "editable": true }, @@ -321,7 +321,7 @@ }, { "cell_type": "markdown", - "id": "8f948136", + "id": "40735cb1", "metadata": { "editable": true }, @@ -333,7 +333,7 @@ }, { "cell_type": "markdown", - "id": "57d23faf", + "id": "1464be77", "metadata": { "editable": true }, @@ -343,7 +343,7 @@ }, { "cell_type": "markdown", - "id": "a62549c5", + "id": "ed9d7c03", "metadata": { "editable": true }, @@ -355,7 +355,7 @@ }, { "cell_type": "markdown", - "id": "0ee71a0b", + "id": "588fde58", "metadata": { "editable": true }, @@ -366,7 +366,7 @@ }, { "cell_type": "markdown", - "id": "1d1bf187", + "id": "4fe15315", "metadata": { "editable": true }, @@ -382,7 +382,7 @@ }, { "cell_type": "markdown", - "id": "c18bb4b3", + "id": "7fcf0aab", "metadata": { "editable": true }, @@ -399,7 +399,7 @@ }, { "cell_type": "markdown", - "id": "1855e33d", + "id": "3f519e6a", "metadata": { "editable": true }, @@ -415,7 +415,7 @@ }, { "cell_type": "markdown", - "id": "98d1de62", + "id": "f1037588", "metadata": { "editable": true }, @@ -427,7 +427,7 @@ }, { "cell_type": "markdown", - "id": "78f42475", + "id": "3a26e170", "metadata": { "editable": true }, @@ -440,8 +440,10 @@ "libraries like **Scikit-Learn**, **Tensorflow** or similar.\n", "\n", "Your task is to perform an analysis of the bias-variance tradeoff\n", - "using at least three of the main sets algorithms we have discussed in this\n", - "course. We will limit ourselves to a regression problem (fitting). You\n", + "using at least three of the main sets of algorithms we have discussed in this\n", + "course. We will limit ourselves to a regression problem (fitting). However, feel free to venture into a classification problem.\n", + "\n", + "You\n", "are free to choose between bootstrap for resampling (recommended) or\n", "cross-validation in order to get the best possible estimates. The\n", "methods you could study are Linear Regression (OLS, Ridge and Lasso),\n", @@ -453,12 +455,14 @@ "algorithms for a data set of your choice as function of the complexity\n", "of your model. Comment and discuss the results. Discuss the pros and\n", "cons of the various methods. Are there some methods which provide both\n", - "low variance and low bias?" + "low variance and low bias?\n", + "\n", + "**Hint**: when you use different methods, pay attention to how you represent (and understand) the complexity of the model. For for example decision trees, this is represented by the depth of the tree." ] }, { "cell_type": "markdown", - "id": "29edad50", + "id": "4c14d063", "metadata": { "editable": true }, @@ -489,38 +493,7 @@ }, { "cell_type": "markdown", - "id": "27d5016f", - "metadata": { - "editable": true - }, - "source": [ - "## Introduction to numerical projects\n", - "\n", - "Here follows a brief recipe and recommendation on how to write a report for each\n", - "project.\n", - "\n", - " * Give a short description of the nature of the problem and the eventual numerical methods you have used.\n", - "\n", - " * Describe the algorithm you have used and/or developed. Here you may find it convenient to use pseudocoding. In many cases you can describe the algorithm in the program itself.\n", - "\n", - " * Include the source code of your program. Comment your program properly.\n", - "\n", - " * If possible, try to find analytic solutions, or known limits in order to test your program when developing the code.\n", - "\n", - " * Include your results either in figure form or in a table. Remember to label your results. All tables and figures should have relevant captions and labels on the axes.\n", - "\n", - " * Try to evaluate the reliabilty and numerical stability/precision of your results. If possible, include a qualitative and/or quantitative discussion of the numerical stability, eventual loss of precision etc.\n", - "\n", - " * Try to give an interpretation of you results in your answers to the problems.\n", - "\n", - " * Critique: if possible include your comments and reflections about the exercise, whether you felt you learnt something, ideas for improvements and other thoughts you've made when solving the exercise. We wish to keep this course at the interactive level and your comments can help us improve it.\n", - "\n", - " * Try to establish a practice where you log your work at the computerlab. You may find such a logbook very handy at later stages in your work, especially when you don't properly remember what a previous test version of your program did. Here you could also record the time spent on solving the exercise, various algorithms you may have tested or other topics which you feel worthy of mentioning." - ] - }, - { - "cell_type": "markdown", - "id": "441c4d7b", + "id": "f565fe9e", "metadata": { "editable": true }, @@ -542,7 +515,7 @@ }, { "cell_type": "markdown", - "id": "b3744fce", + "id": "a762ce82", "metadata": { "editable": true }, diff --git a/doc/Projects/2021/Project3/ipynb/ipynb-Project3-src.tar.gz b/doc/Projects/2021/Project3/ipynb/ipynb-Project3-src.tar.gz index 45f4f92c4b427cb4448f0988ca32d36e85c05428..857624d163b7acd67e3fa759b84023922d63d2f1 100644 GIT binary patch delta 160 zcmV;R0AK&Y0m1Q3_M0ESVFL zrJiA&0;ZT!LRf-gKsZha3jnQm(o1h0H{&U7-3VocdN((YRn><*%Pa8AKXI(3gqwv@7c%J8Z OUwQyIs$$Xr2mk;f21Nq^ delta 160 zcmV;R0AK&Y0m17}=hoAH#kZiKQzy_*}ys_Mg@DYG1wma$AwrwY0Y}>Y-bZk56*tXrl`PbRCcGaF&@2vNF&hgZ! zC$|}?tsJSIm=>Iaog*zD7n~Ys(bSCFY)AEbsXLWqXA+Gfb!B1oHfmX#Z#Z~+sfbE^Pq&Fbs2JZbFuJR%tXsXi_QbmJ^TL|Exgp%{{MN5Bj*!*OFujk~cyf8cOY~0$}AdLWeMl3h;N}(v` zmZ|?Hd|!uE;(Ts>3H)_b8OXfO2$PwDkNe!!bXwmTZ1r8-yIvd{RPO!W7$g!(fOoA{ zlv1qc(Kj}hLQN{0*VqfKKmHsYy2^IaOdQX6bi=B0*ECYyGf4^)1rK*Js|vPN_dO!;q1_`TwJxzm9aGP*LhAT6dF^#Tptzq>)ki{dfc$D@yDTG zUKAKY&NyYe-TS(x6A70hohCC?cP2A**im8Z#?F4uf%udSI-1PGzbDZ1Cx?gVKcg}1mf4M15lZSib6)Z}js>Ve zwvuBtQAT}5|D>eT#}&H9`Akej(!B^5Lr(cMZ;dlGiVY?1xXH;Y`$Z z-tZ)OZ9oo;hV#f8Xd@hv{Ca99>qNH~*u0CRy3w|b5FW%L$OJ_gWHvqS^SMG(tAZew zfCfDXJ{n0J0GX*|5Y}C)q@_~J3IeUHsYUofoW+$={06obEUJRgADE4gTg}u;z0`l> z3Ika%^kv!Nf743F;)@HKPqmh@56#xMKp}=A=Q9UL{zK|HkCT+aW~Z2Jv(q@47&5|k zS_1ol|DNDO21(4}S`Qt~cE5q#k~VH4`RY63FAoOh(SUFHgWH?-a+43NUONL~aJnIz zkFt_pjznsrZ(iQ0$P0FlXRADK#4c^5f<^7J*NkJtn{I|l&iQjfwgUrV_M|i@d~854 z$B-SMdicD1NaMw^JvY}Bg9S0o1Eu57R6SlIL}!z;q?tK~x!IpMjF>QLYMP-wNED)9 z!0wBQfXqkWoWOBH5k5Hyctvgj=jw|bY)dG~U-X_KW8TdVl%E~kO+)y1OvrsjVqRYn z<98`>D<2QOyCVYcKn-~pq$=1+Oe_(Lr0fk4a|B{VmlA>n$Ii(ip(qLn2AP(piGiQS zYGP3hjDx~9B5&Z}(mkSbk!b12;GUp275SytW)le;v6cxrFJ~5svxgMB(614IBIlpH znmIN0$t{Y-eubc>XpB=uIQ|wG=gR9dOINa(pe!-S`@qCYS8$-A`mq)}0Lj6OwMz~7 z82&(|io7DGf}24zyqQD~o@}a+c4h__@78(E3`iF8K@^{&ZtB~BN)1fFAF2*rAb*@p zFz27v#TYp1P+O(;A{nAEG29_>?enX%2gty4k<1--Wf+cw*dqMoQKJK&rpuAKewk`r z2j78kY8h`A+l=`tQuZ@(x1qBd;e7&go0;IC`eKbl_ib`w)oy5JS=2i2f|JoT@?-7- zzoaAGUv$2iXflxo#saXOB!QkTgESF4PYYVMzJMf*MCH@Zg{8**1YctwhqjVEtlX}^ zhI@1KeNXOL*5J*Z?fn$m*rwl{;W-n=Gz)$sBD8yxe`-UPTmC=^E}d8dLCXN>v&U^P zLH9E3YQ`UqKr|HP#ye4eP}nxJ4&(U>o!yfDI?4>jv8BW-_R=kf+bDUYF9;CsAUtw> zXVk8kN!8ri25ovczJ|eQn7%9_az!icv;?Z`Km9TLSJbP9$n}}l1$yV)k$7VsAMP?Q z1#&vzwFUma#b>uR_q?v9o$p3wUN~sw@+D{C8V+OOi4Xcb% ze)I@cBV`3MEFX&Bqx9EZ|D;i|VIb;HvsQtjuoi|QX<}q#yb-iQx0g+#&N)Ge$QI#< zcE4qxzZ=fvtjWNn1%J$HSNMsl>SvY8LvDOw8lQyhlZ!C;rhzyug#e)p!;pdAEL22S z<{5`HwpDe9jud)N0-pf@If}tn!dCu`{G9m~Ib7w^toQ^nZ0i=Aa2f ziB-R?op*w|oxue|Xu*!Z!M&+GK~2+>lLh1XSIS87?i1MQ@C@Fh!wZ}hSp!Xuolvy` z#P$hWPFP#Jlb2ie-B}#*I$NT|!9uiq^b|+(y;Cd5w*`vkIRgL^`48J)b$G)rOYHF8 zB~4F!IjmqXaz%O8u}LYxzjs55jiW(BCdQ4 zn1ytF;UT%i3|v9wk?Y#REiCVZOg+)MuQnM3&A+iHtiL>(h}ihT&eDnbLKlYm3AVI!7t#^*;UZyH@^AL_Hl{eMsJn| zhahr83jPL?AkZkfl?}YdnUKiX?03A(?g*r*Qq?F-7bmCRh&^r&rcp{+yN`Qs`zYE$ zYyka>oo{!6#pUX^OHi7ho;m2tVe(V=uWV^zX%2_Pi1kv+0vuOEmSHASF!+t)D0T}x z;r{W1HhWt)@`KW8hVkv8zgK#ZA;7McjEh$2TWbKa9I>QJ3AS;N6!`rn1*dQDS1DFd zL!An-R!RelK?oYXpXTFab@87FExn%e>sw>0>d~VT;akHz7E=QYzY!7Mj#OI=Bq&S| zJk;Qfg{AG~enW!Mm7U{@@fO9F{t9~7 z%!a-3Fj8sr--~Z3s|x+-ABIRwP*TUCv^|I84D(Zy7Hx_uqHPMS*`2Lk?H!t=>dN*B z*PaeQr-Ci2icR}S6V*DY=DMw_S6shE+V0EmJIr(UT~6IRD6yUPzc-;~Rno{Z6V-GF z9o^eY(`4~_bgY)<$@ct6U#R=wQ#3i&*nF+A?!ws-DZz^m&AV*8=>E|S9NzTKrTZD} z8Eh@Z#^T9xzLd`y@2YM6L5IEE)##fPjp;V+cRVH1@^-izMq<1~d&asnj%=r<+umur zEZP^-UqA8r7%jj7vH-c~!D-uEQ^la}Qm9Pfe$w+CYr=Ir@l;aFMcZ}D3p{Sy*4=3( z>qR1@Rx(ta5T2E-R6)dNYa^CUk|-X`zsP*wY)<-(FdUAm+Bw4|QkvTT7)S~aP;#@|F?Rpse3Dhsu^fCC@I<^!XqEfT7Q4g2Wnuo-{X9}71`31Q8ctx9 zeb)u;W5Lh{XyKg37*4RR?ERV59sNyBZW+-{al`9mjFzN-nwN5TMR5W6X;4%JC=&qu zeRe6a@IjIoWtLxgdoviKX5wh(TC?p%{MXpi0rK(}Bku(vvZ~GIQqyrooLixff4K=> z4c-Q$UPQK^J40OJuY1_}#-od>|D3d!xNE#}?fI+&?<76PP#qKMIJug7_0IssmE{^6 z^X(8`Z>c?TvNF+XMC@jB0i*Yq(gV1(w8O(U6=o9|0?TBsj&KOKl&N4dA&tZ-M(Dcc z=uL2K^Cr%&5U$(jw*9Fz?$`p~ObXhrL$JkdxU6y;13}cds6hr!`=5rZ&KrqqQ1e7b zaAaD*-%R1d0QpcK`97vM`?&?7O4C7XwqJnr^H76n0JI&t^JEv-ML5^f8lE!df&PR+ z0hcAZQx%4=ZqKe9d2^LR8!SLZQm_%a+vNmmpnWScwuY>ZZ7kMp5t%e4edM=u{l}U% zl4f)h8F0nx2PNe78;I$XK7=u`YJ#{!C>;ZKTSV)X;HKU98yCKI%?udLA*jsBLGJip z*w0)D&&hIP@vc&{KnKC!mEyiuZA^|Ad)DT5Kj{plFi1rx91|j+W|MyvNy0Hmv?SxB z>O&1_D@=`PmL!a_?^3f6O&uEB`$Ah4g8b1e<$txvcGRl1lSO)@!STJeqgDFaC8z|( z$Z__DTgQsP%{Cmot+k1374fym?|x&sR6=BBvKfOGG3u?FjztmBJyO==fQwcyAas!> z1iRhHH~eCX@@*d6WzeNwP-vD7Y?Y41N2`mDTS@4B?FvL@=M)4w4mkRM|IR>A5Th<$ zJ$@bF2K60wB+2xV2s~Y!CDIHGVbldsy{Y2~7MA)BMaZ~Hy*WFk=kTRU*b)f1>|P0q zHDuQP9Qm?}1Wk(_P+dMPaG<=tzK6h1Vu~{&{4O+3LHkE9gcsVuPF*`;ois$mc_*7# z>=C>xmX6F2KFSi&_kwmca3KR1oPlXlHz&n8DIIu^n8)dV^`>l7g>e^gI{?5q@9D0b zNEHiCpHG@YPBH1NdDqkid4*0MlrtFL+gc+(SmuvZ*P5TiB{kgzDG(7F0QuUB{L3`3 zFd#J~qkCc_17<=<#uXynT#1*pAsOt&~TEgr;F4g68Q!J;lh{8Op}2*HO3BHH0Qs=DyuSV zgR2AfbB@)|;gnSx!0zc?w&M=weKqRj%ar-peK()1f!H?OP$_=xd+ZU+#x0~~vtBnh z52zT&3p1UKtL8h^CES94I-O6PGUXy;xVy5Y1@f9hlqj9(XkdG5fdF6OU&W=?kFS!q z41ExU!7J59m=Ew`Sjmg?AdfI6<#ze2Y$kY8>TmDuUwhelwBjIHzYE`-Z2o!CW$aHU zq(meY>cn31IwY_7`w7TZkwWe0$Kr`Djn!yu2C$4pf?ZThfEe?u zS2Zy0up!zO4iWui8;ADt_P~4%#>EaSl z1#B|ENCjr$wJ@2n;AuxRB#9%YbaAv6Ga=mKCtY1_Y z0a7x6maNeLy%6LI>(XL`P!f8VxXE(bk^Z`BqJ+0<0q7(@hD2V|(tIwA_P8?Ukv}L* zmT5hEs|6Q6eKi{NRJ-{jlNp>P?!*`Gsz6I=lfq>1xcjbEZbT;X2g!(#j{I3Mck$Ph zGn|z*4Y?2Fv zqo_nFF9719$!8AjD@?%duS}zOS$pOq%yR z2H-Ief3(0&G*&6BqNL;UuD)I43>aS-K1KZO*q%h#yH-g%;Vq;oY4qOfB|&^k5h3#a zyQcybMmO|4xiG~mUC%n2AJ$0K<5yHSE9)B)Lz^^AS(MWCfUk(He-}i%EQE5V+zGi) zMpDcphrXm$d5EWTUXnU~Nb5QPI>dv@2b7fX@7)=oO$D^ZRh~H{Z~ShI-(WkDUHWIR zLAKB~RbPHha=TRQX*Ht#b<_thx$(y@e}}CY<*2q7-{Vo0N!28hKDpCj_?C6BA>(3g zxE1j;(%%T7{zqELDy9D8+vjH|0G{l1*ZKdD=>HF*+|0yG#OVKnqXB?u{0ayNS=_Z4 zNyC69nfeAaZ#8bwKet%J5t9nLzbEuZ)G8gj@U_xv4h>`d-AeuxNQY{|b{OLKC!_LQ z)JgLh+w-@u48|6lSslFi2;t{#7bC9j&4q12XO`_N9{cmHyqO!zh#i(wp{04}NH%)p zc<8Z6Tb?_KRPC-aDD)VLG2wC!ZS=pQ{%r$O8L#guYK@KU+(W5VIvC@0e4Sw7{sQ4a zq@GS}RK&YrzS;9(BklFLj|pg24#pjgS7p*}jQK9qvxJ@GPXOryP>dKId zku-!KCOzka4R_3Lu{t0G2aqCoDiSgCH`!a)Z(S1$$q0w@ctfS*sKbF8P=lgel0$$D zL(UP#B=I|?Kh48oj%~u=Yf330$F5q66P$^B%c+{=!Kf*e<6rFqL?_b1Has7axllrm zjen@qSg~x+pf9+Ywg-RbiTzkE2Y1KqWSs9Ki>oV3BZWl3s}e~iiQcyyZJOg=_}(u3 zZ05xW!yk-jkx}K_i}23GYQSB^E@uGKJ~|q5_2>s9(qS99E&<5Khx6xnHy8e4O75mx znxUi|$jR*R@R44RE|?5&O#6%pjogX{jBlitOKk<`gHke=vxEBefU>f7ZeMINeQjgiGG{Q?gTJQy5@2v zFAN!}9|)u9a~&wlDGZ4{i+Vu)eOe1|HYdC?wl91mqk-4}77a4SQQpBEtnjO?w5G;! z)&<5-nZn$ZLJHbEwqpexN;reUeT;PQL-NWE}AL6k1S)2R=eG!;vjo5NM)2zMEi;M;zacV}~l7ikS1O z3Yp~NTRBk-R;d*8(+)L^=$8EV;(n#yp?Ir;CLD!oyoX?0$Mt$lZfHRAN8_>i@_gDo zB)Nh;-dt4ajB!bwY+v?MNqw(0UtlA7eZCHK5PYdev~RADM`$8-J3XM88M5Gh*{NHH z46R=*u16VT=S~KKR_iKu2US|)XRcPeoNkj)KOZ!r7+_psFIB)SM6T8Bxo0^$mL5)mXE~y@#hb&M_r=SU*WrB-m{MT_Xo)vV%J&zk&UG&AgcA{ynp)f&CY+Q;phz= z-*Hz`L0Un!6sjb75%+Lb4vC3r!cpF5i}}cMY`P!V5p@+f1(X1-MVEpx96H0VO5CGz zh-J)d@rMt~+==n~1u+}BgK!muuDP227fn6Xk-Eh)PI5+h{V=hLC)y+sMle;_CA1E1 z0yKTm3J(_Ec^YKlC`)h(r=6BN9BUFg|6(N}35unYZ0)B){^5g=`yqVW~E^rxu0Ir=_PvTL&*1nM=hR!J^nbs+8iA>zJPNa-{OSSLe0bT{sakFeg%1% zm3Bbxrtj|lMsDmDtT}rTI)%*rD~dD!=9_pVDATwjD`*Rt&h|oyL1|KTiqCl2Xa50- zgIZcM9mk|c-dhf)MH6N1g_GSB7;^gI#SIyzL;c!P7yg(X5Yp!MYT8xS&=lm@O#eF)vv6^7v7{B?f>Qy_e@s=fHx}4& z+1eq)grMF3=h$bDY;EP_g9dhXTHo$o_w#Ew{4u$C$h~YDeR)BWSe(*QWjFs@*+Q5v zB}hZP2$#8v3p#8gEY}?Ef}jI}qe-ESCyGJ_UK3x_0_z_U8xQ~)o2;@5YH?~}2#?0> z0_Fs|nllA@KO$!z03d--LP5!KDfx4uK)85lvw>p&0XGSd3#Gkw6OsT4+0z&5U!IRx z9-F(f#erdI0s*Dj5ec@v+9aT#n#%DzF(@th%sF#!Mj;TIm|LA#{yRSbW&pzi`fGlC z0rb#DcPAPX1Wp^J55~gI!WOQc9kLvR9i(!iPCh_ym%& z=zt(Wae$Cf!Tayo{t_3a9BAd>u&*T8^J|kI+OLt9Yel+#{LZwZi?5*@d)QG^hL4Xp zyQ$3rxHjgi0_iVv0uo8T?|fPNKN~VYb3hP4^7%nd&1~epv(D?J-fYc2Z2hbQ+RAcj z2>O=h)=^Jw4dHyegC1PiTtLCH@v{TAbH8<7^&v*bkeC{n+`RDa+k(Je)ZL`wD;WGA z>>siczqQ|IpOrscf6?s7l}9j)trHl4v4s=?X-%$PcuEF9ToW=?($8 z%+BcoEp0_|=;cg|!4P76-FW0Q@;-PES_OR^08pU)^$wBuA?BjKuo_rFPKUpUDM7l2 zh#p6XKcTys$3BGZ5gLIxyr4}0brjH+3lThJC|e*JLdaMD3eo_1r2U`WI?-dP@59t1z^by?lpSfK;I_U_D@UZH*HM$UHGliNjZ3j zX6YX8{D%B~@u2zetNj36ZCO6J?Y}*Dzd}-S@NH`R&oL?Odu=`j#h2&C`v+vR_O51( zeVTfGKX>FizD0%JPddINo8#kavZxp)H!~*S1>|5)v}j{sBfY-+s@Zt~))t z2;X6Vpa+OhKHeO*o#pm&I`0@u33Ks~B%6KXkI3Yt>tg(*|0DuhLMIyEMgnB*!wXE* zc)^|CHO3(N9W{x--zz6gg5?T9Y%qiGrri`3M*HhxiUu2VlFQJ@QYsE?+A(tM#G-EV zbP_LMJVoq1DIobgUgT42+pa{@Y+sRH88XGn>m8ZkM_NpitHIX>F~_ap zfg1sBTx+1mk6u6=A)e_fUAq{&aXeliH!Y;?%3?^Ncurzfp}syhU;sv-`Zp-fyhsLA zM8Wl;uA_fUBe*=C+(63V6icL-%;_CdfaW%brkzXlI97VatK{eGCbX6ZkNWf)-Oi=Q zv$27oOK}g+cBWHw85Jb=EVE=8sv*eoS8=64u+AIy$ppau;p~S_ZAVk7YD8XV&| z#iba0a&xVl$HuKEmkCpZ<`h%sRuVNoETR&`f{O*;^>v$VvuFcl3xc&#=ryY%UO`=e zuD~y$~5YL=`Cg_5Qdoc zv_e&KXtu~L?mb-Dz=+70UDY<|DKpg8(JwOLrFi)4x?(EejIt!Dlz;p;v`MOrQ9>j~ zL#cBIwl$*ksN=`%Za2$b-B!L*-plhPSlGZrlNzvz@7!zF53Ygi)`l!~U88i44({hy zn7b@8pxfB-ta=d#MKF3nZ#PzoJ-G?#(}k)Wlm)u=u3bE8^5_sp1AmQN+A-oMXV zy{OO)LzSKyZlEiLOp9vv0)2pJ0!5;|yuEa?q{QtvRF2b7bJ+1@J+VlO_{l8&bS{vG zr0BP7RS3LRvL>y*W|;b29341GD)f6>B2wPUO)gutm1-V`3O(R(BVKY>L-(5$Q82h} z{_)CwZAXxY!5Wx+xiwjFC>f$MkhJU_dAER#0`h%?Ks|n!^H1ssCc$xDaq~ml-`NPD zldKFi2;~su5S{CG+*4o)d6PS1RRXo_&dzBeQk)`9is9}`hT=hA50jF-s+Tjo2t_2s2Bot>Oe+P-IfPh zSXWi%29K2KxE!d>3PR2jU$pIDa=ane}ME@cdd77#Uk%Pm$ z5*{Ey#;~s9Pvt94PO9;>f`LTH!QZ>LoIhTV7HQ8=p()Xbj&#=(Z|gH!^ExBDtj)pC zSntR|BqZ+#w99{?@tF$z>!$V<#zOd+G4sH+TUSIs`SZfriu>PB90(8V(A1Jr-E}is zbG@)%(7rKYpRJ--Z}gBRPLwFN{{n1o^HRnF)p#?(o|Ix`(T&98K867%afopr6_|Xh zs;FL|Mjh=H?WmB7anZMe^x}Z$F&0Pn^s_RNK}z;ln)9WoW6^(ZNeQ_hE_=`(7^o~P zvRt_A=m~+(*w1FT!Ar(rpT!ceKk)a3!IC>T_XJP^9pDy!w5Tu2T zqjRGxs0h^{N^im#p&5a{5ZwqSJx6gRU6Ls|=#o)K35lcy7JX`&y3a-EL71qFZT-XP zui=xqEROj*`88@tu-iCMUfvPw2ynM=0TPP{kL1Wb{s=g1YN0ip`Yx%8W$r6m zx_7Fn>{;z;KLjUC_92^KPt;@(>GebR>v?dE3k{a-TtrR7Fmr%Ocxd7P8>g=E7~sR8 zFZqYoa$C~Yvg{FsPTcM?g*4Vw2xa5=aFfVDXr(f;E>96+w8$?Xu12T!n1@4p$Zu7L zlkY9lvebyHG<;1fl%b&_QmjwdoS;Um7*Qp z+Z%$V0;ERpsf=C1kPN&ZsA=bX|<-kP3xyFGa^|>XFMJXXGe^3MV;1us?8G zaV^#T9BNSVsatO;b&?;EC(SJOuzeb$^2gqm`YCTNJ{~p?~n12r= zqX))iojw0@jk2Fq^e8^979Zy!c4b2pnRLp98@&RsHT1&U{eL?&>{=Z+pB_OWPhuHR ziJW$o8L(!*f2~p#p{7JrH8G4 zw`!h(!P9^{m~&MU6w2x_ke41a1;2-gVQHbCraEmq0DhxB40}vL3Kv8b65KhoP4m-w z1}7U}ieWz|l7=#%9PB?1(OSE$0nwfq#1RN-aH!(c@Z zEGj4os-s>Y0H)LKxf;n6n3yp9@ld(Ms|`&KfHRu{Lz3 zR?J@g+ikjQz+<>|&4LL{ds97H!W9{KHmQ+2p`CSl@bYQ!uohC1T7LPF)uU@ykesD3 z3+`0QAq!KS9asxVxg;;#9v$$$h5rQT(eN^;`z`ArXX;GfRq))?AFR0OVvk`3)`tZ9 z*eMpmC|Y2uq098tJ3ZwbGRXTy^uQD(vld!L1znG{<)>DR#q6{?#kjXS<%`A8jpcgJ z;Tx}s?pl4;okg8J%w8%MgD{OHvJjDY27cANM36;|?E4X46rla;^}h$zWyu9R9(^Q! z*wkjU){5r}uF#6-t*AI*n0~yYmeMX$uTQBX&t061H$}|V*yl8z1`w4;WBaqY&K|<4 z)G%`u_C!_SDK81{aa&ianQThfZAqq=lT9Y>jcFnnT^TU%Ni2xHg#KB&-q$|3Y+j6+ z03J(%uv=LYwym`{L%;T|6v#jcs72Cr`l_{KDXAu&zz$5A4Ym#`jk_y!}dOo9MCMK!lRFotv-+|R~j33pa zror-|)kQb)g>G!8UOogsY6^?TdS3KGz#^>>@t|C&oA6K@CSOH#teYQ0me*J&!#b9t zjz1#9h2#sijSiaLf7Oqg{)KdT%!W>!1W(V?xp+}Fbhhff3?eR=McCd8QLkQ2>=qcY za)F4Hw;H99xe-Yjcwf+uvK)e+y|=v+bD?`RG~*4Fz5V5$Q#K0h^|)-9YB`fhKu&9{ zF_pBbtN3!4Zpyfq8IL5O#@%lh6%>NSvXZ^g@RUF0vL%{IKTj4Tgx>k8twg#6E4veO zWxLZY&@a2&ruC#}dl|Acf`lf|74Y4Q^U?~c9rfq5oHyXoYL3Gj1JZG&2w{QL19@F7z^ABzh(~s zp7z~F_v6*3qrpZNVL2a#G*c2?QSBYoysJn9gK}CAE`lI`(OMHe18myLKeHAMja;0g z9qh^)iWd!HpvD_$(8lTaRF8vCs2e@i)cVYhzr_2RCkg;gE*6DIJZH-IQw+4~CU`Lp zyte(4xduDlRib*_$zdcxa}|=%t|)d`;&m8R$<%~lQg)GwJWul0LqcpAo!m7SeKXY3 ztO;ZUOc=tk$&%xN(Jx06us+$q4MEc2@$;e9`i( zXsm>ixPPi(By*KVJ>{lZbHR)Y9B7J!%jfw`liZnUe)jUmJ;8efD1=Rd&PU}FNq3vO z&_L*vKLvi%WcFVMg&LFAw=JC|baZb@|DXT*>|%jB`F3c*U~SEurv$<)g9G@}4}4=7 zZSuQuukAZHG8WQwkQ;O`^^DZQ(@~w95YyhYZla+*V=&Ry3@8Y#a2iL3+X}$FXJ444wk0`5mz%guhrC%Bg%x35w~CKpXj^@A?o$Bk{D{Q)6JZ)Q)ZCrQDHn9YQj z(E2od2y)-ARh|pYp?16OY$9m8qlD9Sxj-Nki<=4!+k7Rdd<%o~mfb=eCsV>Z2frs@ zSy&!c2w_!SYVj2V(x2HP2MzKK${IJ?mon)`;}wdxO(w~uf0F+|%$WYk#d+r%kE8YB z$T|9xeRUMiv?H!(*yyWN{oi-($ut89npq(m__wFEjzzh*!JZgs{`qSQ)3VnEGHu{q zQ$XVO&OR3vm*c}-!ZO4EMlC1wc&?$s*Xmwmu`Lbi^y2cfsz?oi1cft{!BSL9RcFXZ zGJ-KTK5*gr!Q{#>0u8KnJMicZ8m#de!@uvJhcE@fO=9_-#R?QA=a0PuA;uU4oDY6- z5@@Hk1kpW51rd56$`UO)Z7vnrgauGY3G{LRFdyk-SVF=C65#`xtvT(J>kmUX(9k(B zHct7$`hHY%n(XZVZE+RzfjygldvnHXWGw=pgonIy%`S79NB8GoOMqP2H> z)Lpfj<1zlq*1}~ zez87!q_HBL{q-E{>Ha3DGOQ@`04ZP>1^76!5&_U z7vcsJTN#XVb1}^PsJ;%rHcB}iHI4s3m_zB%rE)CXIx zvN$Z=4;(e`8oqd&OK0+P(g3=H?u|=$$LZJy)6`ojOwP;;^F-)%p(EG|d{dQGbGBZC zDZCN$rV#plG3W@0L=)?3CB<;tr5{D@yNB{xOH9^X36>Gl`2)e}*A;-;2*AyA_B z-S%&_>l!I5x--P40DOtk`_OSSh}qY3E{;*-r!>Qi=r-{$&HL+bE$YAfn!B0qV557Q zW#R&T@rs0pozLfsZ4Xn6WEAj3w5HRh>D;T@WfHH(l4QL2-y!^)2gDD|crL9lOJW66 zKi9XvrzvW?;AAuiqyOeTAUpaMenywSUA_& zR5#-Nd@@F8Pm&Se<6^|;tKp1~{)&(7R+0k#dOgkU>4*6c zn-PHChzr@6my~Av1N%+LT=zTMrDdq=GS=rWAf7!a8e{IbTwLK0XY@PAc?rX@$1y)T zBrH54wM|io(%N(&_86jy|4_V0Sz5lNTME6_x_*+QX>-ZcX_ptLXc-Moc5An9TMgQD z-0gHQ&@jc||~!M-`6#toq|b2aHw*04dL zK9069Eg|gnliiP+DZPFeZuMk5?>rR;_=^k0Z4xfoXgsBANIgk}&;*oJ`p%3Q%q6G` zj8cXfdm1(ve>NsOjq_$_!a%ZJ$^4_iT;VSJ+6I7F^7Di^C)}t};)%y#EM{fK>iU-Yfa07q;HDRYSyS zHK#EC>8mS%Uk)t0Bc}VXaa-*HV`hXU*ZnANLqiq3qZ&4Yzq%ftFZ$DBl;1$~8bi*x zd$zMm{Sh3j7PMrk$7Z1 zs#ubCNe{CYavy5lG!mMCFa#)9Xk8D(P8gC7HCmA^ifHXehTzH60{wiDf(ShR*1D@! zt7*Iw9k^L4Vwp{FW=x6WPr)(yqj8AFkMucHoqp5Pg7HphnF_^6|0#j{d7U`Fy-o2e zqzT?Re$>#%>Wz7KN-idw9QE5*ucRzz?0FeSj`-={iIsh>u`0tr%NjyWWuUCj^@%`b zR6V<%IGzu7N9xx5PaV!+0ez@0sz^oiO!Nrzm3W7KYxCF;?&zbO%Q$?bD>K~_6|d^_ zGv+fQ;j_eWrTlaEqK8A7Edl+SlU4gWQKt#m?t)V8HupI17dAn-WQeOP>lQV|3$|s- z-(FFQ3%a>W&^U&`L_75FLQ}HE|C|=;HxB64ccZJXOuSg>&=Ys(DQ0dpayMIjxAV%? zO{6_ZS%eRxVQ#&@9?&5)I8_dL%k8+~v8B5T5%cz?Lii4(#WNB=h5l_4AWW z^g8_%CD-5LTS9@ssNjpV{~Gyantck5Yy)_am|xv_6Oqa%1=)VC;6^dLjJrQ)sRvbq_i6PnjU*UBQm&{UtD+)P9&xj3f*YkVKsc5w7j&3X}WqWrJW{AE! zvh7UI0eczxfQ9>R^LSCs8cY`X|e zs~$x7b74)tVXOPLjBrC$w5lOzGqi_mhkYzVJ03g;U?aSSZ$jDku5z6)e{e2vF>^`T zh4edhH#&lEA0My!wM~k0s6}hCq3xf`0RM!%OgWV#+p->ToayoV`axuNAKpw4V;%H~ zHUWKfxE?=A5I+r(+t91P??~1f!K)hl?1qR3+rO<(up9{~4?m;B)nF0H>!9#9$XHtPhI*>DeqrBEmzV((!3~ zY^_=_3*=ECIBgQmI@&KRZ_mb9LP;(=_oq}O*t=hoada6OK3UmJs%*0^iA)v2kw;FY zp6+=T8`SeW2?>+?FW#^C6)2+HhT2{v-5)~!rv8qz!uBHNz^t_|_KM_)4E5ZnQYje% z5N9nE!BC%jmB!yc@wz~HXS@c>=uey)yG%LJ*hMIZGitn_JUXL23hs^3z`{?cY#VszT4`VY3AM)2 zE8OguqG1ZMPU2vsVgk2lGD;a;7CYk3KiSp(ZK96I+KaO#vkd!PBn+$A8t#gEv%e=* zXrKvn9{vOSdyC+eV0s4YH>Dz&5*;)0hg<*MjV{&rfe75TdTu;M zsW+g{T|P2_3ssn3j7G|0xq1#KK_ud!-@<5bm}-uj4y9Ds$OwslFBi+o z79g(0tMnJz$GPXhhC_1J=Awl@%6A<$ANY6Qu`Jr0Eh>Lt55XXuv=VV7E{TEkU!nG6 z^}Z4|^5qHzag$}P#D4~2N4pE?Z8U%91m@clP*_BkqLe#z_8Tk{oRD>ep=ZJk1 z^g`p#F5=KPp7V#?{u2LH@cqsSWv=%r94Wm(A?&XYXU2%t`5Otn)`5689lH4!7;^qD z{TgmD))YKe*qG`c#)c799`umBhE@xOCU)~nqGb?NN{mUrnFh#29DbR2ZVZ1w3;(q` z6tgjrBS!jo>2&4O{wW)fMm(XK00lnT9_?{kU<(k%l=i|2niU>Rb1+DNd$o&n z8oWHifDwKNtAMY)pT3y3%)J$)YME+%Sq7we>C$tfl{`rDkvH5wF`M{_zH#c03f1>( zPh>vxk*FXG*r7dO>Pa9M9L$Qd8)Igj$nuBxH4JTM+3U&nxPDbC+_F*X?B)4%f`GhE>};QvOkpi_+u_74dz`d9G;R9QB9*Aj;Hgeu zu+sd6fc8_6haE|Ht*?I~Qqjeo>b5Ve%qcapVF0*f4cB8XG3*tr6s5u$cv>R-<42ZS z^&123#W~U2WpVsL4;W)EaLQJ=t(>DhnTw9nMHrUUVY7~F>tnVO@eP&9)_lNaCIx<$Y7F>4$Gss*Z;-F+9WL7jA_vtc-z*@UMBX8&nRnC8xUSyYNAtZ0(APkxwL1(Qt zdxt~2!jgpwdSMSs@E$7k-HBnp$p++NM;HBa6{-pD;V=sx70=40v}D$%abvKgWIDmX zzQE`tFfuqNHo%tv#%rzy%fv57FjYCskE`n6VX4_ATB>^M3~?SG^Qn2M3mIXkvAo7@ z-T96JrOAJ>l7^G=?gI$qGUAA(7xHf3HIAa_)60KF(m4S?9tM5$D=3q5x!k=O_GlN1N7G+$y_OrJo z+<$*=e15qWBsrYy{1zCg!HnbP^4V$9bVB7+kBPP0Cz-^0VE0ZBQh#4zSt>(-e>!nY z*gWcW#Ls6#<&9J}k3FQc-<7E%h@yYUWe=i_9}4S*gqU8VX{okhBr0MTR)(6TJsfC6 zkBdkU0@K?d>Y>f(9`DkIZTYxN+jF3!gzFHHjN}b!Jo`G5MFCNl7&Ub~%E<%3| z`Pnysj{=LP1~%_<%2TU`U%rW9l7(F!;GKMD?YL~k4S@#@Q^;`$)lyI$N6`2FwClR? zdsOd(=}4b6rbSsjK6jMejB^*-on7D01-?|uUWOgs zEQn7_8@fnK#C zuS&WxaL({_16Va|x-LLWF@EGsxW&ceWv=uXdEY?6rnO7VHQr#F6T07h9_WOCjYFuY zC;mLw`iMJ`d=;h61mhNcm2J$zwO+FLsZiq`$Z`@4dYi5qV>kU{?U0U$e7Ao#?n4iT zZgrl2*I0||sctNL3z4FR#(YY?gvjN98Q8>h{fAZfQ#Q$nI$E>1ou%^XP$-R&E0ajT zK3^*H>eqhMs0yzE#5t;g^ga*TLnXWszE-ZWwysbBcp;wWM?_M2%PE zSZ9uqI+$(akTB5GxMo(i>M(6M3Inq_6E|!v)$AONLaP z%b)Xnte?7ExapzeqDpN~Susep)#>*>U5pZ2ZL> z3)moFmV3f5&KGmD0EkU3kP@I^<;GXTKj^<0qq?JuT->UzzMOQ*aPuRA`5xu&Xy~q` z1*>eLDQRadX;}~>UN80K7dou@CY{E6T&|O>t%(Xn5nn`8_}hPp+vENUD4btwWThh9 zR*naV!TNe#HwEn{67u4#RQhT z()=t?5PPeJY)Vo&A*uA@%i>C^7JZKO)uycu6XsmIw`F%D>X+yqIuwx#wV!uvWw<|$ zYI;?uED|f+&AoU_HBMH+e;a@JHj1`$(1pO~hi7l00EN>0xt#-51hYr$a>XsWY#ZVd zAEkfs?r>^pSOcDq&IJOjlmFut{=C#QW?1b5_l6$Z#JAuihupY;hd$nV6ZJiwS_i-? z^05EvsDbj%gF{LgR{b=CZ*E}rIF60>68EV<1^=y1Xu`ifVX8d}%0m5#Aq6vJ(TYFi zoh`e>DWqU}I)*~A&T09MqK?wWW@CYBnF4>TcpI?>5_#BAA2Gj{z+=b=U60pa&JAmz zg*olZHzoI}Sp8vmg;$0!a92AuogINV+IkB|Qe|5OQ)o8Jn_VP`c>{f^)TnvUwU~E9 zZkVDSY{~tNITv^QdiruJj7%C`5_S?FWAm9E{5SzwzH1*U+EU(M$zd(*hhY<(PDXzV z)1)*_cvht{kNdLrYqXwS;-)I43VTPWSET7E?zOd(-pIpTp3`G>HxGpMjdf5D0RvvL7&@t(M>%B08HRk~Dv1Oha2dt&958oZ?KB(~RNR9C?k? z+T(tT?fcAERA(zDYIX^eyo=Hw7$JZ9H74(+r#WKtM%_4QR_!aWiT;&bmyqrKTGmuz z#bvK_fmm3$Y&AHS)~hnlf`y(4GS1h+WdRQ981Zxxq|d#{Yu*c zoDLPt=%>$9xSua`8Gn{Ihd?o)dk7*C+ja|>j^TudW6`>CRA+kZj-A7?(W8I5;C*}_ z-v|XN+O*LKhQ+@LfZ75gkse5zRCNL+%g!p4OujNdQtl&v#Cl}ePp5+SJ826V5$R#S zkR<)7ZhDSDq5#HzzZc=pnGew7*Ap5hr+1fV0 z&JDsT1>kcFqUaHk5dwYDg3^CSZ)yI)O1a$#ulldC_VM5|{QYwB1{a%5eOwqmm*a@< zH;rX2vaBfysujF4Ntd#Oo)g?Y3`GycK0?>MOVsuH5*YzGQJhKFCNHJIBK0~he8s7W z2a)(16Le-b^ddWI_gIZSGGAk}>+Cfc4Dnh{Aw*rx_o;&f!mFqR#AAQ@XqosO@?$#7 zlbgj;!q_y0qN!Cp{0Skv7SXPg{=rYd*!%BJCU7F?T5);ttcf=b($#16^*T8Gvf?zL_~c7?%?8es(F3OjW$>r8xxl!XiWOr$_F!oUnMF zdh}?gb0buFJaHnjGG>3#n901a9`eZx)BXn42B(*)`IXj1|J|&o$BpwyNmNhWqyMTm zkCWj~xKUa1ylJZ}0{Y7plx zx8{aQzXuF_>$}{WDR%3##A*;G=>K2 z1g~aR?hQu?f-(ixBHs+_AGB`OI!3a`kGx6Gp7a4t%ulhmdrRNh9?#IfG= zBR5p3kxnsT?K?S{WQSZ+lYMZaZ6Am>^R)8Bp}@dG1U-LrXiAwuj+Kqio?!svXSepW zt|A3525)Ha8tyz(9!w|&C({4p-(drVc7_an?|zYsScNpl;b5yg>FiBRBGCSTa48v& zk+v2>9eujDOV5oLxypxnw;!`{4;*L@7yL}jfWuxE;n!4U0=}I50X>H(J{CjOEZYz> z#FkCdw2FTQJ|^t&-I*0c7+W~|dlO)>PdfpxT3wOu4}Oerx3WGBC{g~PF};&CT%FY@ z){eVKx0YsXxAKaiGTU;QTNa-A{TFY{zAcB?Nu%|tCsC#DCiJtzE^4u?iHzd8VQBPA zI&dS)4i8&L5hP-M4{I*QW)B^ucdB*Svs~N zov@KehHP0DbJ|#dI+FH&5^{~G!_0OzwS1RelV>I7j?rQ$#+KYsn+(2689C;`n^a@J zP=9~OZ>!i67uT3RAOvR?(_(LN-ToGtglgLJliVWZn*Z4(bRsXZS{Re50DPPA zf6K^CbRNSMb8QIvc;e(V=z$|w+i?C0^^kud13^=GW;7`c9YmZ>7O&0&7s!Ci<>vy3 zzV~&hctI0?bg=Ex_RZ&kiscxZg+@r5&T~51G=Roe5o_F7-3x}ZmP2|5u^mEQekBJFyIF#Lby zKTI1eyf`Vre#|SHc9X>CGkv!Sfm4zR}Lo7qkR>2kZMR9{bOC*ZM1(k?t5R|+vw7)jqDv@>ba6Ov2*384u2GEXPRAe^ z(KdziIX{XuJe9J!WD^}F*MK6<6F7)=?&-NTKJ8Joww-PS&5cQjg%UEH7gB#VFK8#w zhI}(G_catMhXu+I-ndr3Pco7OwLQb`aBdp;%~N%U7YYJG7R@)7Ikb+6#q^BNm+A=MDzPWP;# z&1>DA;fO=xg^m(Nc!zw@I^lm>EpY`Fo}{TOxgE9I`Au+;M<34>;?Ilr?}HY9@ja{m zu;ncLQh!SN;O4(vf53h;DYmn6HaMV42Rh8ha<~6yP7okuWcjC`6SR(@`Ts3C_{K7O%g2akKJ0|x&wV}E(mm724}JO&ZbYij(plG^Xr8liulaR1Wq2T%{Az->{l zHjDT&;FbS%(w10TrgD%*F3?^4o1UX>Q<(&gsq_ps9IIB@JC=uN?_22r(?&X>rz2n;J z5v3c5rOadPeR_YF&E?)Nqr&UEDF`3rBGVj$$8xf|xzWAVP9=8GOND2Dnu;vrfT)@F)mino z;y>!Dq)6!>R`9!_d$?BSI*#)0S-;E|Psbn`gEiT{+0%btZGEN~5E`jP75`o|5UXRC z*4B@VmlRbl;O$!F%~IY&k^n7-1&`Pf82`mhcLFTLI!VP64Sxuk0MAzoG7McJ+{}}4 zW)w8|E;CMltw8)-9zoM8?Y$j<xs{k)^S=i2FAqVAU1 zH!gDdOhtcE+^@n5^GbCLJ)r{_uW71!EXLAG=grN0KXpbFYfaw%_~4vd`hwk+%wAKr zV?@-rz{bkezeRrBfkKa(ek~>t_#M8WH8R=7o&bH=$1gENQm@rkU)ExTU3I$7|Hz^{ zs0F&XkDeNLCDnO;iGyEg@G0O)ho4%Jww5&zhO~dZOT!Gd;ri&&`@lOjbfbr|N7gY{ zl%m=Cm_l@da-tYfPgaNrVUID>5uQOoPVMaOSb%^qm7ao^z@f z8>Xi9o=m?5jvhnz9XK^d&`E4(VlZV+_w;c)-!91F$0p+{%G=IYg2T4t7`v>eW94D6 zVVz&bzTR5P${|3*evAw|4){@JcNB3@ym@~Fc8Him?~m?+s8JG3Ovi1$>T{eRrHOoR z{Ow&Fo_n2Rkv&RtRkRpDP+dQsAx}Mu4Hx=w$L~pVOC(l6L2{CCj#ywGJ4Tl73+DJf zDno{i`ttS+W$K3Y*0IEtmKNmx+G!a*p?2mVUMg!H(<`6Cl%X#T$9W~)2J|oG{yBfd ztA`%@no2bHUjZv3g59_Y(k+I%ZDt`GK%)Ap>se5wnjuF!DU!x$BtPrjN%y0+GCwJE z?+-p2pAj6b(UPeIgP}YSbDf=U1r#LdFOs&G=>99~02 z{W!G^XcG*k_$h_`z%K#i5e}<{q{kczcNhY`ElZ~I;Zs(|E@sojYfDY_mt}usUXf_g za;l;RKNDL-2)baV3+&yX$#5mGnU3{Q3233EeL+KwuZgDjE0rN>(6V_9M9nEah<**l z%6mj)C+>A%iy`7XwcYg8Os|dt9~+t9p2WpzoAdG+FZ5xIB$etkxV-1r?qvN%f}5y5 zoLt|wDS(pJa$s%+pN8i5YRG@{{IcEkK-pFwYUvLL9o8vk?46BUMJ1)iEF#L6vkyT+ zk6%H~*I&)TgQJZ-pj}q*q12on+>U-aq8ootz#2t9J5u#ok~{kF7}sU;f$Ppc509g* zrNZT&vL#5QF+_KT=Gn!1E%q}Up8&Ag92pDgQ;n(^0 zG)DFz4^Q?F8Lb$aw#CoqHf=9(QW27Q2`R^vZ78!4rM1O^%{rMA*^8w3=L5R7eo+2J zChftrMCV5qh$$t0A@zUZTllpemLm7?=8?(B4enF{(_zX`<9oa5E#C#NWsy!ZEOO6R zXh~;=P^3mywo_*o(BA0+8|-vuO1R|Il=S-i2?r`YoU(5}_VVl_=ka!@2=K2?@P{t; zT%3}$ufJSeIuPaqkK?!Aqh;~NG6c!8Z#mC>z*QS!W6z1_UAKSC>|CQVeqSR?J!*?W z^CFE55U!$LEWBQTF6j?nuoN5rJ|6l0(Ac%5Mbkrz%5r{0=qiA6m1*cEB|;%OnR%Z2 z7*;WN5!kAy1XGPER*+x@@X#E-vKg$~)?;Z(q|yj^i%Hp>phiOPU`2QixC z=kVmv+)Ar~Q)jGos;+mxb$?f?@r`L@nn%8GxKDn+z7=4HqP>n#o_X(0(P zoQo@7yzD+aRaqnyjV`hOUmN$ZB#NRaNVII*wr$(CZQHhO+qP}nwad2sPIN@i?qB4Z zq(}21TFm7u_HU>@CR+2}&81~kcuZ8jtRuV-u%0*0xDm8kB?q0?#?KM`WW!ZmK2_z`G^h z8Bp_;Hia`fBIeY66!IhHQq>l8P5alfqjH>xX~loZP-p}ABC>8c8#Ip~P`JdRF=vN+ z(ofV@?W6Fjc8mMV&GOr@T3|pVL10f>(b@yY@7j-Ow)Wd%Pg>?6tgX0B=l@m`{9gKgYj49I)ROUUPMlIJY|LjRA5om$w^D1Pp zdn@8PNGzCtOc`2st4>SA&(2keq#@Vx2IM;SbatUc=kQMT91&(^u;V-5OJ<9rJR%5i z6gQX==%gLfK7V>CD@Lvm?d+BB&BeiXVN_(E?KspW?adzp#t@A!pv6ke_(C^QqZ^hzjrpC{^l+#$LTzmgc7cl-kL*)#Y?-Mv8T| zi8FKIeK%2RlU#%FqMwLepmtsSJHye`sMubof@l+hyEzmM)$Ne7kfY6h2js%C zCEbd@9>O+E+_dJDDyS({`vu=r%QJr=X{0$AkG2ECdm-XvMMUpsmheI;_2;c;BdLF| z9Lri4x91CGQ46Y86u4-%qW%zvQngq7+>*yUzd^^Z>r)Jnh|q`a=R;sMmKw~6w>N6Y z>gHeu<@*qAoCi8M2T1_} zk>zlpZ(Lw6!|)`r*muRTpSkq7^9ZFOh$0ayLwSF=%=K`b#&?@( zQ}LR#G0DkB>*2bhLH{b-))eHM>o~O7FFN@=7dIQuMOhUIkjKR>kJDf4VEWIHvg0Sb z)v@V)-s+?a#jZExfk_`-zB0O9z2-8%%F0q1@YRr8d?=@Y9esP%OC*2SyC^!NpWi5Z z*^vE$<-q9>q6XWGbGnbtbPIqHbzIPfvdH_i_c8t#Uogi6I~qTY`D(#Uc$%224N1A9 zKAK3`YG*G@vGyPH0|LgmJNLX&w3dAg+Yk7Gqtse{v!~n*7(n~BtN@e(xhl`TXhgM_ zr4ZeziQtqd$EKUgBw~M~o%W}t5L3sktd@H#+}x&BE$(54Xoi`gzd`8#=4=n|ni|N8?^+*3TN>0)W?_SW$E4)x#T9(uuU;ehs+cZ; zY08{it${Cn7&&E&3Uj8A7|XYWCyRJ~1wuF<-wH8qtdGxz^OXGxWH^8a)%Q-&YQdT9J z%{G4$0uaAqd}}v?C7+V!K{@mVKbgxoCZ4PSVlf_3337`xD@|L_I#UjJ=k|5EPYN~o z=&!erp4%V!mJ3D*!s=on4c&t+UNR-a3)7jIdiH?zB)OZA6=}}FL-KwGg=vzd<0RKi z07!dyMEKkXY1yBy0xi5fsHu z?+FXS7|wR!-8AAo0dC$1u8m1^IJvtl|MrW}TEZd1i}v?PFCYc;3~(J;^t2nQSQ3Bb zHLaLbCAD9Yq7!NdH(YH-H6<)a&dU;!6N`PWVBS@PD{F^Y3FUGEEdc_gJ~{f0vl&d7 zdeqG{oyJG#EkCXkm!_U={U#}g@Q(F)fMN$CrD{*m!j)e1tWygC<*CRP_to;L?7E#h z{+k_oGScliC3aohXK~zMNo`CALPviEr*)3Zl;l>-gw)h*b4*`hayjYI1cv9!*>?yz5zmjx<~eCau>0o zRd#4i=Ff`=zmF$o{ zV56mcxTt~P2(U?0qHxq=-58JXabuUk5t6G#U`zJ2PwX$;y~Z?^tUiozY{{ypDUX*5 zHV`?29F2N-`g>|tyzCBUoYH^Q)IA9iH9IzaW^6rdQ&W<*mB9Zfe%!h{EK$nK8NVD% z0wJ>K3V^nHZ?V!uJfpJs-PB3u1{X}6ag}QQtbzaBLKZvGZ$>f6!`vjOp?{bO&k{QSw7c5;KPp3JjzW35rb5zC%b5#|(n17K9B3_@}~ zGG+mFws^A@g7a{)CJlcA1CTBb|EwS+aRQoupqsI)=uM846prF$MOcdF%CQWD);#@4 zAl(0Ivzpzyd5(P%qaUT%->zlb1RiZoXYXDcqEq(54870ayB1D9)6O|ew44B*LSe-t zV3BMXCIpwBces$#c@9pY;%G;Li|b24AF0QabN68zewuQ(K1hG3U=5<|UMeA&p?zc; zdK;2j?dQ0jN4cfH7vX+$&pcLNO(1b(bI}*%X6L#Yj~bZE#JG)%mF~S`>$!w?j{&*i z*F*6fMzQ_3P;0%LxHz?3Ufz)-??va>LB-JZpdCx1=3rFT_=8ssH% zrpgV!Q{~imO7fv1d@Yv*t-e$S3O?&`sw?!{;CbB9Q~#~^Oaw@ozD+$m=!mgmDlP-q z3|&#I-*Mawiqs&e7GLNqUF|+o_brc!hKen0hH`|Tii4cAfqCiJ%{eE(zLDl1v=woZ zNi-k-gxr6*VE_$;7V6D8du)R;elFyQeYN*}A`(q^vjyzctcYiP-TOWZ!sJtEYW7W!;A`u7CwDr~#aBrNO*nd&k!Ts2U~scC)G1)>UZYX4v=!*MqxVCa#;J5YCK z43Xxj&wHRhTd>Frlh2=^W70NV{#c!}p~`z1$^5i6(aJ*`$_hrLprZX!&(OI$5#x)s zR9k;OBY`?~2qmoQ0>z#Vo)a#6f+s8+j+$8E+5f!oi$QmuB{>vBYpg}pC;oXY7>A~8 z&9JV^Vss^jV~`Uoei~_phZq^D^uXxHJ@E$Wc{vp{ChgU$LMx*Z#{-C@MhYBE)d*hb zInD6&bL6#*RTIQXS;l)Az_1_LKtB^v_(OlEC)5H~N)C6MVOyg=tlWA#6oye`mX!hx zK{x%c-kL*m7ju67C!HI!rbu&TO>KVeny2OlCDPV_Wfe+ABm!rd#8rP1 zR=QSvYW0@F{~n=UwWPzOgXp;Pe$q@pG(8o%29|)YlgeDb%n!RYK42$%<7|gMES{vxb68Nf~P~m24&wr z&ZeL6Vr(x@rG6N#93z+jvWjgC0LC z?2i%9aZ%}@kHp!UTBe^_>{Q)_X`O7}9(oDqIu}XzY@DRY@LUN4HWwUDpB;YV^-hb;c?ilGQ+C3Oo*TdNmZ)p2Lv3LVNrql1JSIn(20iKkK#05 zXhg$gY2wn+319H&5Th`U$=`ol{(5Ptp2Un0rhYy$8$TB-=s5lkwjjO5;?|A=(R-hf zDX#lTjipw+^%po;UAdRz-l%c|)2dL>lQwQZR^Z1=D%D9eid~kIKG#ZH9+L2m;gmkA zK<~`Mr5(I4@^n2=!u@?+jps+$2$W$F^bMu6jm%EO%~qM&AK-`fXxV>a^=SS646`#6 z!cCtwDnJ}8e7h1fcB}%f&kpD`&aRr$S2gW4>sN)~N+kFp1FHnml^87BlJ#Y*eCyK( zv6adlk7vx!_Q`{Rmy(KO2m@hD1nfoIj&dkJu@3^kMI<);CHCW2Don?Q6OjDD;u_F#nFf!9F<0Q;PNPxX}wGmzlCUG@5t!j|tPDy#gD2 zc|!f>;Uuoh%mHp?pu8jm?!Dx*{S-AR6xa+Bm*=mlzG3u;Uq>jlha!}0<-STR z9Sn&NP?GKmD5_G{q-kQRHSJ9`$Q7&PJ=zyDXCBmDDH?b@2p|NKe_wO^t!n{ z*`A&96(%fuz;w#^6E+cIzCps?>7LL9dK5dY3*8UgD%$Th#NSfiX{CgW$T!LE+P}{# zP?>^%B=7^W=p$&^QqXg4vi!BLC1dQcxJT#0To4gF6=k z-b+E55ta*gQ4di)(rV%8ve0nDcJ&Zqn^4qV|Fp_^n(*wBaj?ElhF9Rfueu0twnTe&6j1y2G0#6 zz*HFkD7*(^pSR>_C5`x!mEn$tRT;xlb5O#ttr&kiMWSu?zBZ!#J#)C?t-L$$;q9QW ze>~s-=e}=lFwkUrmXL_7jM3V4Dm)_de%c{aKn0U;d4Q)`2BeWV6v?TZOCeO#n}_^P zM=yX-iGo$%r_S8$*iQzG#nz(Q(31j-wMQcoQ8%LidV&2OTiyg9-MX1A$i=1Qg|OLe z*Cc;G-m?D)Egm=I+89Bnsj2F;R-qIH-?ll8SY?6_`25(y&+sK%CDPD zQP9=W5ICYo0)Nrc0y*#rgh?Zhh5j}a=5v&0&Qor~V;)|r16kqoiGoY%qa&b0N|}FN z5=Yxohk!PSauqXxvAVc-*SgK!k!{01);&jAo}Zf8c^vfjahj_r zE;pkd%q?kWi8E2$8yA;Nr{ejsKr=QYE2+OV$g3i$Np-LLb1G@I-t*vwZ+2%OQ;;EM ze!g+@0b{ER`x|tKGL#G*2I6pDBRM9^p~W-rQSl;EF6v};Ww z6ZNrZDDEJ$_lC&zN2j-fJ3k|Pm2qr$ffYbERqXspQ4gVT`(DPSTTxF5*c^Wkg}8Q@ zP}eYh+tH&pg3f%=10nfS?a1DR{jvwRpFTG|G0d-Xj{LGe=WN-Is>JwuKKnK%Ead7| zUt|IW4Uq4#mRfG;r{gG8{>0{=nAdXS%t~ruT{QZ9D*@L3=obM6@);s!9VGEdIrrv% zN=tQQ3)VzHJdovUdX78O@JD~^5gkO8=o`*Sw9zx?{XfrHw_@lTG>3Xt{^qAWsh(OpwhtC zY6koCd`c=i_ z%Qc4S1+sv-$KegmpxXUW1N^@wAn_-uc@gVLR5lIHa@QA<8F)yPz5l)a_^?gvSY(>x z6p-KnX$B;OML7n?KZI3Q?Vi}<#imn1^2*z(=SzV=YueyMC$L=DJ_x0NkJQxqklqM? zYY`H;Nc(4Vc)P)*PUL^!@C+c)o(hX)0m?c#q~_ zJ_dzE3C`cNJ(t~VPlJ^r)NZV2xObY=F>EGY28q+TP_+ELcQ9DrLI(aObYDk)n_RnG zEWP-iCt3nH_qpo0EzZbyhf=(1(bnlRb3zUBfOdgE0XTT8&G)8pH z16=$$0#h423|d6(=Gp-iinn%7N}rLoVh}b99nD>xHHg0Lo2*7H6MjFuirbC3>9GQL zor8jaTpdc!70BbY=(hbSKxPkVPuywLe4hj*aOr=jPz&AFQjWM8kpZXg<| z_0K*8UV7}nSTa&!m$l*jjW_g;ImkXRCIFe#!;B@ESs8zRhbVR5Hiv~1_5n+N6kFbP#0glhKE9o*fm3<}N@bk&q*&|d3gyXlFFD&a>w-^eS6WksJfZNrnmL^+;8%UmZ1-$Df^ZlL!E1E2{tEp1qQQio{v zbl!d5dMkBO1N4jq_lI5t6@b#7TZ!vBk=x-|#Yhmbkt|!*w}yf93RN^WU>G0<_+;oc zEDzH@#SHxKg92#GQDW-89x`mZq>@uvT;6{@fp2N}+5ev$E$CFpAC6Z5JrI|Mqpkt^ zlaJMy4K&D0z)$xhj;)*~Rz>FI*=P09k*f>PCvp%ANQ|_6x8AfEVW=OikCICz4t601 z=zNgM=7NFSl{XprUvf_%%s@IKd8hv#Y(h}V~T@;3J! z{qhxUV3Nn@G+DmkdGY9#K(~PVfi5mY*|w@K8(CQOhP;d1(yVt1__^g1+trW9Z6_OY$wak_x#HZ&01rSf|aJ zps+JEcbv$W@Sl}{mM8-3GOK@+$mzY`_AyIoG9%$VkpKzuJTY$9Y}wK66eH&?jKva) zN;lRX>8BL>@$WFv?1eu9QABL=?yIEkxY?RBlXae&$Mht-+%2qCq*@>G8Si4a@=p1- zZ{SIpf=Nn=yGtF!_?+@?BRaB^`%qXQ3k16`QsPpN7;cqNRD9m6b}@f6$i?3EC?|iy z%)+GB2H0v>4$QB(5Zj9{$efjK_LBtat7U_Gae9DdUudqsBv}tJ7Pb)BE@9V5Fgc(= z>Zd|2VMN`hQ}K(XMaNat`Z(74&~7;NnH393+rSU3kcW%M#j;w8&a3qL?rc3U&8QJtINh4_JLo z`?jO>EP60X7`%+IG{nu7)t{&IX9biU-?@_A{cAC#fiPmCH%EU7P_zaC)7ydB9t$!A zo!}zuBqz}VSjtaeH=|0hU{fu#NX7zcJbYlbFhsccnwsBAP#~FWDBVKIfG+YY^Y6aG z3XQZoX$Wfgy{-(K=|F0iY-W+HoW|kiVZFDQ-OV3XF$Zq~FhZ7h zUg>=Xxe%}4K*)a-3N!7F!cN{gQWPGL7R}nGcl3Gr?vms4S4DmT8b@8CdO`l$>y8$)wA zmr*h`9yszmg7}t=OI+c##kgD&Pt$-UwZb_ zxlr0ZT2l2AypNYmh=RD|zv*4wZrKG}o#clr1Z7wltLdIsHJu!1(rd9FskD)w*!{UiH04so;wAA}!a-y&%T+a(07KqtI$ z+vyevHW+_0o^&qhrDG-OMcn>#sgr`P`;~`SG_WQY0l;CBP*Lj6AZGmXDV_7o&=WDj zIadD-4xrygwLDxmj8thCOq(=q#wVs5pth28LFht-w1^MEW&3_ZnO$ub>Qqi>F0Ej} zLE~m3A}U>`2(>XB3KsEg>ex-35;QPL5qm;a8z6tkPMRM8cK67dQh&~4pa|qtgS2oC z`beo7YAL(pP^+GvK?p|&_J&jb8Z@<4Gt-~V*a2fB#!B-2l_#Z_R*}q9Et(&|2YWMI zyb=5m%v{dVUQ$%xD?_?f?3BN+tW{NHlB$RBB=!2fJ=ZV^!NXubrSH>UtnOY%bRe9RRWOYIZWml(ZT>Ut@1YV+SvxwHph6C-k zq3x(374=C0%rQY6ZTZEAKIw+V;jRGbN;VB_3ta2j0m&WQyS4fjT5AVoA1O3QG@zdN z3M^ZAp7Is$IEMrbwNWDYN}{X9GOpm*56J*w-bH~nLdiZ@%-hL#3H*kHZhQTKRp4pU zDXA2P@G&KS%_-Jx$ALCR$i3wo#$g{I{0|q%Pl+MN{7q^MUEMK&$&&(g_5C*$^vXJ1 zQCb}Q(7`Y`22RYTp{;!La4ZQANrd(Qm}BUdyf-~g1UzrMifs}izusp`l`iIxScm%_ zV|^OSweh3Hxa6i;0}#!vZ*q0K3jUWT{8u7GQ*YzQh>6Be6?N<`j2pJ5V}RJ|aCHv^ z87jk+{&O=;F$sS1X|6a8KL+;i0K-JZ-`C__^;@oiCyN}TT3NwdUXBaZinp+7)skJT z%rLh*u0V>YEcj)dw z$?V5d-LS)cU%QW~)*$N;0U)CI#?|-VvQcqAi2czCT>Oe2?&T2k`g;uSw@(|p*3E7I z46sVmdFK67Te}Ty1{CMXXMXv)p}P0XC&^zf;;Y^kFOw<2=9SW41v}bmI;3o%&bb7Q z6nqTB@Y?C0SFHWT)4-HX(p{TQeN zeTC7~zJKI7n*-ZT%c3}_KX_VAgI$1a2f$Vn8spjz5@~;E+sRxSnUh_b`+=G3_Q3Z@ zW$@6sJ8B{UVye(`drNX;0HQc(BYmVdKO#s1YS#Ni&aOh1ICOBcHTv(N$=`Py;22Z_ zcVuvt;=AHi(V*H*M!4)-XJ|#nB@D#5g^wd?AToIYKUKRxFomJVwTG|-H=Z9g+4*}&-h&Ef)pncqaPvYf zr#*xJuNw8r4f{+DUkf}M)93)^5^3Flo=8C+KashuV^k!v*gI5ZBv<*bdX)<1T`eHv zrD7#>fYO8~R}uIouE$R(j@{{-s)016Egk#qHPd46L>C0}<~+t)y^}i=Hb_AT(F8x6 z@mN+L3BD#xTP-jTy#qx_isNi&IZSuO)P)z^5TM z6DdVhoWpkUef{9xxckA`M*{)&Saq~SGp-&1#KSyj}Pv|BlC8q+6sx^aMu;sgI%{pgH3X$-?+Os@6V{(k0ky15-te_1A9Xft0nP016a1cEiNP;K@l3| ziY8{AGk@I=+0S~+wb^eR7~U+xo$Su8MYKKUDY>6S|3cdwo{zH;a#G$<&V7~)$4$X{ zkpd5^3-6QghwB|2I=EJ{!U>ghGI6NJIsOrd$cMGjGr+>S&=wwHLB`9Gqe#cXW-lPfL@Y>K^CdKy}$MK%eBrT)$8DgzD52%?76u z-{HtBi%saIv&^@GsQBBHXeoF$@8M(B~BP$|YO>ciZ9p5L3Kn#uv-mbUyE8 zg0Ob!+lP+9p_LC=>_8Eztd9Uryr*Fgnpar z;o?6O!Pqg~>s$`Pd-M=PkqIh$`6+P^8kzcCCcB-MZ8SPlm%r9)J-d4{Bxd_1e9T5a z8`W1S<~vtuc7Q7NgcExOx$M&XfR&92z`kv~4N<70v!qVGC6-MQK?I0*s78fHStzm? zq9O=ml`G3rO-u__Dpcamz-l^}@Fd{Zt7)f~q!-m#Nc+=Ih#CS;Ri(sFBWXdPi!|1* z)##=zycV&Jv3cot$7ur4D97!Q{jq^MN(f8el)_E%bXh`Iar3QeR8WzI^Hv?!E7<&< zP3K?PjtpDH`G);vylLY*^{N2LI&;D8^h_d2Vg(jB-|GVCu6tQ~zGyT^(E4-uDFjH2kvpqD8 z#bIzaydJGzVKu+k7a{+oP57kM)fAftr7Pj1GRYlb$~cn@f=z_&8RTwRVTsItu&8GW zU_uOu-($M*{;{IVs?7)A z(Tw}+q`Q?L%jos><95^Au*el+4l9&ld?QD)UVG5)eiLNH7zZ<5Yf`aU24}VrO?6<( z@(+m+R%VOaY{7%g;^{Ij`Eu_qBH&sp?4I&Fe&PC@!RsBNAXe%ah|%lXG0kfsh4$n7 zOb_k)N-AcEonfQMN1A-22K5g)d=^f}y?w%kO;Y_|y8@4ar=+`o3qAcOat35=cM{5@ zhGUO$*o&aE0)dQPN% zv=jzPjeLHjy$_q&1Y0iL#gO!v+#QmeUJXO(DDp2EH=*_*!S(h02zqnn5q;WuSG>?| z#TWF%Q+ieF33pQ{%*?+bGA{A|w7N&A&;SR!^~)y@kPayQ;s7eLe%&SnM8+}tSaCRENgy4 zQ0P_2-0QqwR>31YjDK&t9`nTR{hakrFRsB)($#w2tN+M9La_0+!1R6rm~0(RG~n5o z;2a%r{NQFN5bT^C4zl14kr3?M9mEUZC^8W2yd6yV5YZlAt{w3ZSjJzj(^nArOpt6G zoE^{%klV-*9NZm8Jdi6|nYILwv?%QVClm<;vSx}hK+=J(_4VDic$x#(8upKz9K!)O z165UZa#t42uGIK`)TtX9Bo^6Ct;v${5>$5|ulJsiDYRR1HVi#}BueSfq9TxIAM@Nn zt&G)TxOzAhNnBX%^Q&0(jf>dqILy!FIN-$1^k$mD6;V>SXYqr`Wzsz{;bnmh2+$Dl z7*u!hqGuqjbVjUhPR=bcf;z1#y2HNS+Z|DBt5ovBsteV9@p0VZ02|JTU3oz%ggDp{s`SD2{n0pCG-)YQJW=6pG z#we-`h%+o5h!z}1+Q>l0x=GvrJC#{FIfqA?J(82k~$8PVA2BreJWda7>Ztqr1? zs@*^g)u7D*gMSatZyh=LIQQoBVYMgqvQA1mSG&u|9<649c#CJS0^)zn{{gI!i^Y;=#35FLty0EraNc@e9sBhE_#y?Dm+`l`1&W(qqkEqTH^I-3zDBr*E5DmbqDrk+S6ja6LX{9kE%|QL;*|QrzwJlh zK8{*Em%SMLsGHvpib%ZJ0UO4}(-r|x`w8T7W}P?=_M28~@G&=?KRh}$wg{l&MbUs} z4oi%FZ?Ev*g+8ki9l$nvOaF-5Jgfs2*B05VZmB`C9C_Hv^n|daqZaY&=XYeO=P6i` zjvU?g9bu0gsq))e9V)zkd!-2YV0YSj`GmVWU#fKDw4KzY$hf{RQ)7_`;1u`abg&-Q z`q6I`V-3n1q|sr^He4kx`DGU?hiTBbb(vNN`z zkw#=@E_D-eY9E@dqhxya2n30f%F8;<smyjIX*m1av8W!B(T$E2F4+QN316AGSD{c$*{? zic9~~U9j{weajR%cdgq_lK8YI!Ugpjy6JLGgq1QdOY4Uou(XBX6++wVy%w#1wQws} z=_~?O$}q#SIj&sa+o4A`Jodj~#BBRSn?`0*BoX<)114NK%Pr`~x+ci&Pun3xZTWE(ZCT3HhqVZB4sr(~ zgbaB)CkGKonJRyn-~xs{Q#2+&JHmoC_4c-WzmAwwMkk@A*QO0v?$(UUWXbsR;A@*; zx{ICzBn(^Dq9DVXbG^OQkZ*Oh_H`-D{wQ+6J4_tzf7x4@DC<8_y7}kbjmRk$du`6k zO4=^nE7mi!=XYYYBLSy7b%a>=tgt6bC=nL}!{Yl#xe4`U^0jTD>~LCgk&zqU`gp3W zGnXdZ7&9IZkDy9oWEouSNL;DOrG_`g57hiaF#}oA$Uym5mdCMEYS53Qmc;a@j?Rj^ zR5N{Wl_Zw03)Pe99vmByTI9xwfzDXBHV+@HN|`Z2yUmfXIU;G?ihvO#YP5gIm~b0V z7;|K@q>HwEl&EM!rue?AZ07#DC;F6PyG4S}uxEJ9OP0KO$kH=cW*}~i+v74Rh+o!3C*g@vZ;lI|G5rsg=8yB+& ziwc)DlyP0?J?oW!ce%zLGiJaKEd#R1y9knCGEj}@!x@6XD6Gj-u@O?H$M-GIx~7Y3 zJulfcSQb1-jb59&V)@GT*j15fxu>Jx!^;_t%{6B+-%~&$%_BTNvlw$jh;> z)Zh}NI%N`Ux^02RpY5M_?T-?v`bZCGF*Q><-nQ~kn(~QS$*v3PvIOw=-?9_xWij^V zNkMTg2>AG7me$k-obhuE-(}46AS{i@RNe3?W|kaXe~+B=QUk|T!@%m`SvE*ug=E9nwjXB_Qv@P{cUb&YIiqUZvnVC{p4%o=UMu2sU-{Wc zlIrH_b?MvdC?Dzlw#7DG^&?foCznn(AP8}XAHOev{TUlrB~9*$E+n`h>NvsFltT67 zwBlHlESs#B{Y%br$WVKG3*8!$;um+s>XC8&#<^E5mOR}?0v}F_Yw+Ca2Kj3gB@Mep?&FOV{D7%2k z8JsdMl3eSx1S!fl#rrL&%GzZspPtn?CKLVS9SQ}gvJ;Avtebq!CnVvlBHZ|} zq;68!NUnlS1qY3V4;bY&$vqxVVYiAc+RFls#mE;td^zn!t1EZKCxOcJOY}R7T4CbV zZ#KH9Sg;^vARW3&D=Krd0Vu~SgO2kyk&}oSd*Y1f@$gvbo>j3R+xq9Di0Z6Vb8@Sk zQKr1^LD`Pueq{>3$+7PrT!j`$Y~{Qk+8s)fT>avygG$%+^#2ITT0Iq(%9VjIW1@t3 z+*|9{iV)>|nJRG77mD?Vu?Lp+-w9qco*Se>x0?QadIQ%Hl3+J4gh8aptAvZ^Gd|F3 z>Jtp&_*gWPn@NUcz7Xe0>@%ryq!jwA?LeEgOPsmM9037u5kHZFkpPr{B>b1^8IMi* zr=}EXb)d>s!-{VKu)e3nlP=P5uHdXQExnxhhKsl02cmd=<%~rRanc#G()1T|PDp)5 z#ZvA;IajywWdCqP9f-;Df*X6=Rq|T{SNsy{Jb+@v_>=R4E}4 zYW3S(w?}E2%n!_#I)1=gs_whGf zxV#c%MWGEbA*lYByYFbFw*#fBJ0I0Ko_teE9yGBood-X)d58%2vp>9vyWns9Op$xh zyC=em((rN*dTsUdkUba24jalkR@be_2gW~T$5}7@J1BTF=vkVojw&*B4}79dvOPmb zT&-1m>lSsuA&06?7P7R9uG(s^{3)4kO8)xH(ou_DE>Nt8?bH2l1VQCMpS$;&yFDeN zJTC3w>{=2*+X`Gj77ELkhHIMgv+WRt@8cV~T8dc8lWf)bb?W#Ox}auI3vpEG!F!~y z8}F>Jt>}dxB?a|s6R{Fa>3GYOt)T{_>*}g@!};TNuy-mf8J20nor`FPI|w~-6w{S5 zVO0H20rU`emjxCsy(J#sA-wR@Gp?@~9^F~DgSO#y=g8~ueEbX3{3`!bh`dd4y}KH+F2m49 z8B|maKQ<*PSf)Kv@#rQ?+m?FpU|lK>1AJhRaCIjkW z*zPSEoLb^;D9o!5M>IyC?Pko2t{r}IIoQV4c!`1>((_E+bP~VTB9@+c=c|%L!Kejw zq_NQLIzY^GZDwM{V_zQn`$`WePNr#|XOQnO$i4f2^YHy+74P|uuxeE&Gq!jj5+2%= z@OCup`VF}se8+QuigKKAouEkatQF)%33xRMKG;xhPzpSMof)|jN=Mvqyx2!k>r&Eb zd`nLlDX3p5*-8Inv~chonK|CyKYMU{^?LCmo4Fr{qDU#Rzs>LMPX{X`~+^yJcD*aj_^mD$=r0sqG7^e^`&jxWCBWL5_ z2UF9Y^)oXB_HZ4Udb#`eq#iMp-;nY|5*CUH9kbg9F23mX70Q#NkhWw_h|lL}YvNVl^w^PGq5l## zUos-*=uvMS3XG;j_NI0)oMYz-$5kqiwZUUt>T13#2d@aUj$P<;_X$n@#gX&h7!D^# zW=|qy7!3zE2M+@)8x1Qf4IKiDqO*mBsfQ(iR)UY6m7VRY0ol0zkvk_g4>%ypfm%JWP`tl$-*~VLAWLj zs{r6p;6U6m!!pRy)4XYDN&xC;F{xpjICls*b|hLT(@jx$R^;+G@F@jL$Ur(+95V1_ z9tr5=wA{rgh659^k+z7lApnUTf)14Hu!0CxIO$k0J0)EDU?eK|$EB+&x{y-khGE_; z#CGECt)tUY71R74)ZCGWE+FD`sw@t7g2Eyc6Qqyzv}c*Nuj^EbQJW9uvv1S%v*osp z_DhOAf2;MK!2kW(jWOnpi;e;?91E8g-OF-m>?7buo4oIx3>FJ_2jZG@$1=PC9=M!EkFLJnz0SN(%i>X;q!yGg{V`E zC*OR2+(-3^q=X1XBnVcOLIp;t;i%_Nu?z2VJA;eG#i;Y)MPYos@2RqZ10pU3F>~+O zr~q>tZ2jQz)Wlh6Vigk;NX&ZJ!*^$85&#Y@t-Uc?|eNh|IfZ3nR^2d03Z^Siu&Ib8F zuLC?~(nffXX!^>(X1xHCy+O|kUr?NiA!nzqW@znXkcQyd1tn8HH><7BH7 z>N*^p8|@?Pk(3~yc-5Q+OUTGA{(`l zP;ZjVx*%mM;%Z;90pAneFtEqEy5vFO`|QC2Rpks=m(0Hw(Cqg20$)1D4xosIMlIhy zv7`VlNVgVaMB30T;iKrF*=QQb?j%B0!2D#HxY{|m=4(Bf{K-^O`IGEGu`Po3x-Gu- zxwEFHr9LcnM(a!~LYLJVxL*O(TuiH=_)|=KdLwl8Fm$p8EE4-NaUt^z$~D1w&K-)y zI)J!+M?7GdtC=5ET^`El4f12798{Et<%Me_)+VjpaHS+Si6KcR-BNv9KTYWTb=D>f@4kSz>EB4_I z3NqwEduj*V^m|Hvci1&^`j)k$`(EByYMDJ!Wj|ONu83OW5>;;i^NIS{H=_GLaSoVS zoC!$=;^Z_$U{SaAwFIznaw4#31N7Md900bjNi}C@kFQB~04qQZfkoEI%K2-=@xLPp zfW82as1%17E4!$u6dRWqj}#{xC#$3=kc~q^oJX9EO$;am`2Sr{z83o5Fv-Eo3izKe zNwqk|rSFP0e%CuE8Y?G8(8^3!45vw6rPbzSYqMZyc(I7{hy?>9RokX`p;(K8G+k@C z%*=!);{i=dhgn&kT|;foqJkJzwggu_rlMV^OKF2dI++P!=O^l0n;^nYcrF%iLSvgamq5EV>=&`vM?)=f@jAcb&XZ#D9xXsb6~`4FD|g z6H@;ChH<0s;wU7H(2jeq@q=$Lh1wnB#|w4M3X=UQ{!VXlH0Fec_C}S=vTErE%^Lp> zmj46^4>4X15W)6Y790ycLlv5W_YG+4gi+~4V5cAU?Y$|UJQP-Hu4v3I<=A~JE_K_H zUD}F2-lkXH5{x(t{F~o*nOjmuu1^$nw9D%3H1^j|8#*t!*)5B|H zwKIK~%n1s?gJvTzl|{ZTT7KJOrE!DB*%N?4uynF;_i(c`b^MRzY-Wex?qTZY;q7K= Tg}~17bzPhYR8*46QV9PA#Z(uZ delta 44299 zcmZUZLy#q0w54y_woz%@wr#W0W~FZ0wkvJhw#`c0w)*exc%yelYsVg(6R`(t#kWsu z8Dee|VjVFJ7#k-?8j%+m6|h^&FmAIA1vp!JTL^z?O&*y%R7!fKUqcOwK1Q#EE>F~i zYS&0B%Ro)O?y!S1qmr)`l?JX_cb6$@ixO$ILMAJ%!mb>*Nh*ITF|CmEDS)3`Nl3uh?KLARcI_^;8;wl74pR67kP>+Xz*C{-XdZ@dYiec*xj?INDyfZ z8K{ZFU%i8>y#ucg{2G#8T87VS87B1$Hrh3V-j?Teg_pl_2cdWZx`Bwhb+!*{{l}en zrD;}>GW!T|xaTekNhf>3kgmN?z z{t*jg_qL@7$`Fn~Z9rwjpB`%4{!>6V7n9_k{Sh1Lm)VJho0wyCJB|oo=3+>iwLtP5 zZj_p=S}W_4iIN*StZ3-mEffiR)OdaB;D>1yE{H&=2mCLi$9{04wGF1ad6Hvcg5Q<`fPeQT?mMCqVEAN^6mm z>Ip0A)iB9VsO~9A+GA0O&4N!7cX8B>JSSoEod`+Mq5Nm}R3VLc7ywL&DE>|u_P27R zyZVeZAY8d$kjIlMrNj`@bI6IRZ$TiJ)Q-Qd)0k87m2vuPC)e}unrt|c?1IA{fUViA zOpqV}$P_)ByBdrRJIW&Kcz}9VUw~G^XG9$h*SS1a0H_{J%ecucqA(2^ZYC=$D2UTF zB@L7LeNuNbRAjj<Io*J)~vx2jF)sxZlvimV95WECT{v#of(^i1j)TSm7^hLd?7CQF>ilKVuFtmRDgAG5;W}4EssCj|4 zj5a0xbwU8>9s*@ocUd$9(lD_g9A(qBwS2g9tv4$_$eX0G8utxIeWdWn!_ zcV+N`4y8JwG`Q|Ei}Jx_#ZDu;fEmocDwPg0W6|W#7#?31D3@~jQZ8DJ-_{#F>1;6- z&xq!(lF+@>L{y?Z!gucMG93LJGh8%HO~G3FRYrdu9n3wmnsm2)ER1- zUrZ#}Uzxe}25f=7759#xElP|MtKd*)K&l$}+Tk*Z-QJ z`LXMBn?!@#HMpQv90K*DkBoIgmyeFQ@yJv!2_krA1yQ1sjuRtV{iypsdd>sMa_;){x=Wx9&N(hX`n&pCZ)7jhT!g_|afeXQI-PVyDZ?8e0TD1_f)ZX?g-9m2 zXMc@BsGv3Jga;FqzchFL>55AvQ^Q@8Z$ELqcAXmq^qY26+4fID?Q*_*WqFqT3B4au zP-4ubjkwSDH-^W9)-b%4;c_uRXtgJLVB07c5O~ zQYTaRr;sh#{QZCktDAF?Qh?2XIlp!YRS9=7kY=Cr3VbtYE`VwljdmdE=G!RFs9_E0Rj=f+~TD3Lkv2QT`mnA24*PDg#I2S{dDny2`CsB5K6wd31} zx&g=0nFk0f^JS2nyrKsan+gwKK%qJ_WLEXvJ#`4JQ}}6~?Nd9FN%`Hm_07J!Wa~Uy zouGP;lJNn|UJSu`r__e_)*c#Zc*k@69kMYu9sVLx1~MQ2;AF-r8udWpVhZ8hIhaK_ zXn(JmM|S%fM1nQ87@Y#OM{e)MWvbe-`JJTv5ZPPMZahWJrHJ(}+xIa_;Ea!jtag-X z-R&QcAJ=7)9xMov6JB-n`a~S*AH`(C+G$jnfVLNpcNj9~<3%2n1YK)AE34uHBOfJZrExA?QO~yvjw8r2;R}au5$UQ)U5`T=4T#P` zt||*;l;-LV#4XZ?c+CYopqfhzRl#rko{R{vz=#+5rgDEB$iSD9(N~^Z+-km742Q~a zN649RpN50|a3)azGE~UPAzfYcU=|G(&Pkvz3!xS5Bu_PJZb?O{v?*Q{h>RGp5W1so z5qzP|Mr&loSWO0kbmd@3ZN>_FZF{bvH9M9`)NN=#4$N1S-%IuzNEh>vt=}g(d4XHg zhYnld4y!RC)UXJhzVvKgeQ~7I$BbR??>W;VlF?EwQ0Vc`7B@2#l`xXgIod+^VSKi_ z+R5dkqRx!kEO^c-q2Y;B#B3f#HL5}cKXkP*;G$M>zF^r}{x-rkF_U=E zrvo$qTq+7Q8zi>Bu?HEHrTie9>}@YFm>Y+2HNi%j0GnfEIf=i~*ME3aj?f4LQk0#s zCvfg@xr34MAfJ_gdi)&=Q3@6{pw2CTsXgp(1{E;?N6@#;z5tdL5tod`^nr(j8miTi z7Nk`E0muIXcw}drita=rb=&LdLbzf%C3YHjSe%Fa6#owm?&agLJih}((uZhf6 z$XsNC1?bn@aIS7zU!hxvVB0xuFVaBnvVHU~XOYd<%fuU# zznoe?K3sXU@)-ufL8GJ0Ew2q#D!PrFz>yr)%LwiOGUjn;a&%+yHY7V`Y%jf3x0Cm` zGIkmJf@#DzOG&`3D>?Maw(JiL9&nWOAb{EEz9@oC+BD9ab1 z*X3SpZmD8BF08r*rQ5rbTTT(3aPnK?67My5BwnRc+^kZ~y_9PDZFSH^IM`YklJ@uaRWN4Sg;`p?7gCfNX%U+a3DO>? zmI_O%FZ;7GIb<>pHs(zEl5QdeIE*}Slm<%Rd)rY^N~EFf?fY%57BtC$aRt+L$g@_`)OtcR<1DW5U?Bc-~8GE2BH?6ZG|U{;{s;|mjQ@` zX-LVzopE_y7y>`F%EkFqFQXoTLS){-jUbV@_K4NY)gYH11SI$iV^994=&s!1E$Srj z@ajX^Wwz91$fY>1hatrw)v@>pWNkJ_z_AP-@`{{Tmq zTe2FO?p*_{=OQXfRA@)0>K^-264j;dou{ll`_N-T+V2g;>OtcjX9JB_^ZNztn(Q79 zFo=!amezA*T~&!Sx^QrC=(#hmjTr=VaOgW%A50R*?yPj1OE<^7n-$d)m5L{Kl_Dtf z*m^1C-%B=9`NzEqgdT(OR1h+3f=svpkF3TGhSFV&+m#CB72R=Gz?@PNyd&2zU5ZQf=3D^jQ4I7wWuG%>yOIpJdHvJ- zbH=(%s!{Oj<@X^sNv4rc`2*|s5j;V3a<=+$U;&H4HM=GFgTI%rw@h!}A`M*t{i}`<>J_`pEH}U^Fu#6H`wyx&R#EfZ-USQb3W^LWL%@GX0 zms+L2G9;M8?YF6|L)R44uxn+jLXmAqdd?aqsiY8y%@f}rxp>qnw44$FCLjqeTxqN) z_c;!~3zzLNs{Ops^!b5Zz{YM>9^E73+*32RpEeZ^L>Ii;L@tEp2+fz}^VZ!cjrsST z7vNt=`x{pQB6JDd!+9BCwOV$FG!Guo|L)~N4(e;t%hVMYo7_4m3R^`3H+8u5x1xJj_I1R^>p}%vN`=6Y;Zrh|&97Z9eLMDp!xa>x4Y z+CA04XoSc<6M?%p2dZ|cS+_Xz$FEk`;JU#oIX25r_9J3|uk$&SIe&g1;Uq3k0kf@v zArrFOWNeDQCxPFkg@vu6=If%5Kh~G>z>6>$au*u||DQ3!0@T}x;mn6fCehIp5#Yxq?t~qP@W}3Zwoz*@Tsc?N6?B~-8CmgFv|Gc7_^<4o2K-=%=QZQtWFZI7 z+{2nix&972i2@Q~vIgs{F3nif*YPt0J`Ku<`W54^ELhKPgg9-?~i#YOtwuoU;ay^+>hS8a!)AyDD7L|0RIcb`A~`3xc|&uP-WZt z6JJHpu=m+K!G+3iftw3ds2gtb-y6CqRr`Dhb^{Z$u66ytj#F&foJtt^$u`TfiwAY3pvQV^3|)c^Etrx7q($i8 zPH{Ci<1^(V2;!TTOBxY_!Xgt&4Tzt`&!73b$D=PE|0Ha!@Pi$*i=c%gjevx;QJs=kLbOAEN|0P)!yfEzJ#TsJNZW5$vqlf)L&22^ zNh*&3?-evYrCR!YH{=2bjk-AGCDG#6Zmr>zQAe~{>;m|eS*rlAHnb!DwS;q8&PdtK|9t)fa2i{SnnOMKmOgA-%q9#rl3brPGub|D8WxIqc`H zBP{{}8x{2~&EghG$N8Y(I;UMkp@lzwo&0}i_sXvaS~e4ZapKaMsy1#gh1Q4Yb%H50 z?s%;Vf`#2}daZSwt6`ol&>vpA?6m?-MA(n3A`X}!OZ#Y$X+xj5Mfrt-)eTVky-)IZpgJ5spJ zf3KQV{Oc$ZWBdd-^p&NHU(d(>1uJLLOS;oDC9j4H5S6OqQY7fa;*B)rki)=X52xY? zKG~l?ON(s{u{9*55A`cP-9DTd!^HuC=tz^?KE7pH1wXhpb%2$T$Ry}IgIKKS>9UbC zy57@^pTNDp#5#ARYfNrRk`iie=%2S z`Pu626(dp5`gqmQawcA#E;E1(%QFZVoukDV+b<2I&zsqjkHj8Z0jf1eQgI=HVBy9M z-R1gjVFka(Gyv<)VocV`#gfH_M@YLZ?E9Iro~Jk_HbR9B-0BBIxH9g8jB-(rA|41u z%V~3Ohy4B_X(ZKC>bGEX=2|l(PGyV?Pm719XN!kDwyrxdrbyT0ha62A$NUK%i}Rhg zBEh&yZ@!$MGkuI*vZFC2ey4Cifr}$EtTiQ3T;Y!amU8V!ma=#gz#;sNnQE5B^oi zy@VN$YUW5kSVdh5I7xs@tGgvet$TGIMRZGj%0L-z6gA8|mTf}xu1GVv_*=t+(1{KI1S~b0RfpkxHkNNSeKgn5E$f`C=0m}I3&m10A;4Er{-^3B3p+_FjXC8PU-n+<4zyiH1#&1!I}v|Ddx!4i!HQzc3{$ zGq(-~DtoL*Pj;*U88Q!WLK=?ka_L=sxEZ8C)tgURlGs5h`3@COK$i_0+*1&@8=|Dk zt+~BSpL044G%9&OU%jKi=70b&;)LyfMqPmJ?|Ujrm4{az48 zrT#T46ez_!U8hM(ET1{UMa_fDps!QCQAyI53^x^_rACW?@VUjDq8@KQ zh}i*C;j;H&EPV%a$aR#SS@&SoiM{Iq|4|k(+8Ez|)1Hfv`}Am=jEWZMRE0U*5HRDp zukQ%G0;|Ia0u^%jpcVZ*=zNk~cst=i__pRF0wIO~fh8bVmu|QU4ZJc%h%mj)xcq&v zn8wjQxLzg|lJU4DqNS=v4x(gxw;0@8LoVmu7?O{-|2;h`9pWo7v9*fe@*t$O2;)G1 zbI(P8VSToE`^98{jR@D}?c#g8?0dfKl0459H0TK`1SdtX+1RD? zM>yeccrc+DbeVD)w%`EpYXbd-7a!4|y~WzJ#Y*7cUE)p{DC(^s7)$gxSDV;c(0sJ( zDIv71-c+9mC2kxZS5?8Pm2_@S_&eMwb$nrvBul^3tv5w$0GV(qgQRe^tS#rLCO$0pPZ&7iHU?CF<6NdDajMe%@U^;nVe$k=>|jIaLxELW)e%m>M$;7i zl$ssnjK#rL=SLwIfVK;iqMf!tuDP}UqE#MW+?J$#1x-Ns91V58*NQkl4J<}sfD3Y_ z>VavpNTf7M0N4q9AY$N{srcU!l$+~+M^I)q7PkMtWSrKNjoaiv?m5vs)xgzaE!yU3 z5lR$LA(m<_ErP0%+Z>umBvDy^-SG%zy%+7IS0N6|6+;X==oSV z_4?wlOO)U+*P(dX6q8LENR)9cXq3}1&q)?1U#yb4MjEHQb5}1l!PdN}Vi@*btzD3j zGqmFQ1LS@0T;|c#;cKs>EZ4O{;fv|5iquP=U-$n|Aj5BY+~OjVP;y67brhY|<#`G3 zNd3L}X8o~@A4?!Rm`-`izqxkj_v&^RC@|qn&su*g&{1+D`FoK|)}RBjM_VVm7Lda#jhZ8)uRg8=Cb6{c zADi0B^LHggWk4|tJ zR+7MYGWa4_bG=Y-#)RE);60gXYZ5}q@GrvYnR^;2SbT6gciZz3R6)Dmdf$sG;BT;A zU1g32Dwl@-%|aacKIk$yr!Y$NDKu=&av-)HFFs-mlq+5$xLynm80;gyuoKFD#g^W^UJ|BEiIuEYK-2oW~BJBd4$$z;@R!_ zB2K8-ub&}3u{^Wf=uR_2zo8tJa_dmdrG~1=qA`HM^vriX=hd9U6bpQVoZ-s;6!6cx zN{jy`jxU(|{E}o%bhY^mH+t|L%~eq{=^nIWpuMN{;fstuozjDZEUmi;J(>6(=dxE3T-Mqa& z7GT9Uqg~VFGW{A|4-UdaVRjm|1PF<%hC@lxP=QQHn9Fq_P)X+?4J#%P81%kHFg>^p zj*H`QONaLgJ{4i_U9heWowy5_bAL8s?AIUK{3TkvAe$^Rsk3U4=q@@X>jWeZ41P}o zV>j)wRA$>Lf_O81)Z4nfk}x1e{)|Vu8yZd`N-L#z+yO^7oxYG*A@^ST3-qZA!>T~X z#ECJBySvL?j(TqjwN^WcDwFADuUH@#8#9t0x%Lqi_=mJ;KUL+1&Jr=qUfXfbOR95z1NDgr%4l0B=O z`7ka(0g;Yk)6>RGv63Ms1w6;s3V$uLaL)M?ex)$OVb+eiIlnP^#)wMH2;wcyL^)QJ z&5ht|!i}wK(Z=Dy>}YpGuRiUTIVnM(%cJtKz->N&w7E8`97872t934c6?oKe{eH51 zm%RoAD$01S(W9|H3TtlBoV=(_LW-j9`6@KOLozUY7@Sf!aD#ta0=Ej`E@m?84Ve`F zebMHz=eQiy-T+ShAR;hw&4n@n`cVO&U%O& z3(gl2{%OSIVp1V57ta~JKi!&jDKM808Dratg?~-|uBZIHK)wClH~L_It+%me2pWln z*V|NX?!-{e2s(Wn3uwe7zfNBdOXwUJZJQ+g`kbAQT@kWrz|s9IH`68Mpy#m7zkJNbQG z^>Wkd)qKfA-g&N;vsjt6-p}3R%y@8osuf38ICQ zk+Hr!x!@X66)XnlDCA&Yk>lh^G_XVIR<#U_~R-%T&m8`ca3fyi5Lnhtj`l40nEk_>`nlsAVT8unu_NSbp&Xy0|#LwB_(g&AA%G4 zyVmxuw#>TBmP9VN#T$G89e^`0sd4$H?q^@(Z3O-{Y$L0@J*j)k5CpK4%D{lwB`0@% zeI1QiGs6)5tbiGf*aEy$ZESLeY}@1ot^>BWbNpNSdSMS>S-fdWqsVU^pn6-ui7m`w z!MVYJc+zV>-IgyJ6P$&d?UTTWZj6DWR>jTUWsn)6-JjrkHuvvI4FzdwL?a7Z2xpMM zXyFovNQM!Hni)W0ubG2l9?tR(cmP0Ild^P&`S8bK{H)jbJFX-AOZT@|%XSZN=`B9iAZC5oQRf7@xEZNHX>)gg77u?HeozGd{AvVr0MbE(Bmr}SONnP5 zWxYH{t?5Utuhrl^+lbGI02fnTi@e$t?%OMb?alQy1ZYPmXNY&NAD2BK3LhL+-_*p? zQT6x^B*Zu4%>uEDH(+*WpOfOJ_3`gj>9>uP)i%Y_6rtG#3=>%9AS;l1t+R_A?CD34 z^|@9HwWg&7cEnxi{D=Ouw>kQ%%9j5}^Tv=Svu{VO{D){nV+K`&P>7YB7E5)qxAMJJ zZ)5ni`Yc2kv!$<<;7zvZ+dI~F2lx4EuKwNkHwd=Y8uw50^GZ|+2XS5_tN()o>#i=s z_wY;gZSOj0IvJZ<5j(Ky^m~~BNF5O*cQsulgbC0aSS z!=K)1CIQm1g`XK5C;(8JScd-X-DsaTF^7i$CYT@0uNW>6IZ_}pmcR5P(mo_p#FsE1 zfa30ll;M~&7C4usxGs1)UHXsc@mTd6$?Ku~)^Ol$YL4%*(&_!M@?Ecp??bP$l=34MvivN1sCaz*ckAdEl9_Mg%O~9TjS%%$kMuVWRmKlcZVgzo)>~H*mDG)(v^a5jbulcc zb{M0dUB%~`v7^A9%K5f=@FQ^cqX>u;gnAUL5*NGV@4{Kz=|~Pp2?OgJcx#EFqqDW` z`7yP5f3bU6h52wt|0MeMT?Y6IA>~njP84)g8-}}i$!dlC^_+JlXM;dG&}MJ(Pn9EG zl34+|Qi9a1jq0o9895p03UQd%cW0y5_83|dm#IaOf=@oXHS!n#J`c-inBz1R{38fz z=CpU`gCOygXZv>KKVh$uX`1mVa>eqzemNWQHv;7RfUM;iMz9u3 zs55Psvc z>T$d)gGn>AC5{Zzg6RlD-lhh7tY1h4#EeRq#_nWDdOhKq zW*SSTh5IpMp`1}uyzXM&1%N-TmCC$Ha(=8Tb01nWTxP^>p#C-fk~1^J*@C$!wCY9NdR22 z&%Fdmf*VDgih;NP&;rLLVzn3*45kO7sZ{!|6dxo2l0sO0^0*yJ@t&{>g1# ztG%E&rUD;A)}9JPB_%IIX(Hg}6)1?RcX)L%;~j0$xXZsc(vY{vk)sE5VTghF4gAF@ z8`owX9&5v2edPpxB|z_T5+%r>g$s9Ub9kDqX4CW1rL|2p(u12Te1?SSR!0Q(Vpl1` zXV}U=zgbaj={)e~8p$)`6Sc74a0WmaoT^DJh7w5?iol@S~}f zj>M=j=BsMY95*kcWjjE;CT3=V^J((dq6Wt69@J|0kmRnUWs{K8mT$eAd7jx7d|U3) zypT@+2eRzn4F!b}y>r27>G4+xXA;g|hynLdKIH2-uhPd?LjW-8H$)mpnm$dL1M%xo zxDd;WfJxa?hM#9R>WNZt_tU7PNz))3kANmr>{_-Sm8TdEfgWH5^g1TCg7vt)PBxXv zmpd`nXKK39YG34;FcuF-BVR>CWLZDQQg!Vu+Vmr1|4sZEvQx1&X@PsyLz&>m!%8fO zm!4(lu6}Q(qFpWXaQsnejcUtXUF5w*N2+^eKe$7(j!deKVp7t$PVf{td|w6nq8H6s zp$X{;hBaKIn=Wt$Rq1&d4$7k6B6<|9MSGUlQp^fZ zT`+pRJhE~hc*Zd2*RAdOeqC|vmP}W(T^gtoN}S}w>QFdr1)Q+`At{XyrQkXcR_$Ji zvy7)}3WsU z18+%ZDhB?=Wp)+yeOrx_rxD?1wug*#?q0RN_q+aOWSdTUv!Z@EE*SFEHX@kP>S?>D zC5!fZffpzDQ^|~Nj zUwS6Hd7hk4RKGp_ldU+*O9w5}7r&dZX9fA0IT84w!k6%kDDNkX%Aac}5t`kBqiOEP z?yn3gA=j{hNMsWU{g>AFe5_7evk(`yfS2&<8@Cz({ZO)K%7rjQVF&WD1hzX67LP+c z#HK;dJ!R4Cerd<>K{KT_-77hjZ`}L~V2?FknvJ5@4Lfb;!qP7@U9!0!Faq(bM+qH9 z2okuC@v4$>eaU1h8&!qN^WXo1*;kpqo8QGtumY*=YC}Tp!82y zlLObkyVI&LPP+HVugi9|0Vxj}0SzpqOFf9aA-VRu;N{9Xu-w-;*G8$na08>Bue_g~ z$RIW4APrv?D6jLyW^@5l{Uwqxwb6(=py9wW?fbVvtdCRf|Hfn9z=0?E9`d?p_!$Q8 z_H{#e+6zhj^$OqYpKN!sjNJ69r0dp)EQwEYlNkMAZad433=}zEjIB<*A~#dY?zIjA z%{I^Q$}fs7PIoeEz3S08kp_b(U5wnzoG}i(FNpQOAcxmw=g2~j(Vk%8ITcA!VFVJQ z{<#}5y%l{iF6T0F#3?nqW~cH3&%vHf5NI_9!xcn_B*VgpMx%k4wW(KtwYKCeyjS~V zi%PDvbYyxn{jAKYfU0FN06Z~zMZlvc5c>qrYZG1_xtE3uctrQLFK;TUq~RoWLS%AO zfI1GT_zRTq3u3az@r7$D_@#=0I0J)x*}sZQ@~MOsBmC&2vJxYp*>(O9>E!YswTVLa zDA%L7<*r*0JB**G66oGmH}5s7D+(_YnSva_@wC4T?c(oQM18^cszE${zV^bCf^z}` zc^*Al&r2hXlBDs;1Jax3Zj#PYK1+nIQfgXJe%%EBTjqiPHS5BVS!ZG9OAb7SBpIHJ zt<|5iSHkA0&Gxk>=28Y~+y@tEdVG6VTi5eF24!f0Vz>(R8pS~MkDdgF{v7GKXoz*B9-|MB)& ziCAgYYzZN*d#Dq^_<30#LJ&QN*aBRUg@AH-vQL>kjqzUCezq0939=kr2TIV4q%%JT zCabo2>@CChf1Futxhp_sY2s4&vE8gp@k982P{iT_bci_SfO_lTzkzz^C3)8_m^~%Z zvHaUGjvQ2G{j7a#4IqKVka+|Y6|Mq5VNkb!l!Y80^q?BY^x%`yjzyaZzQ8VepYtV)WOCKmOH;-DeiTTNHY zGBQf`7ha%uc1NHDBZ^X$Ejy1!brzL!Os4ppi4Ljt=fje^=zEApij^-xO-E=0Uax*z71Gdn`{}xR>fbRF_uz_3t2AVedvW=j$>DcCct0|-&9Wl4V(pgej`Tf z-D+-Kinp`GD6;N z&r|LcLfKIMk4dHUgWE&Vj}I^;*HSEoUI{R47>>ucxp=gyXo_C&2`&5{$_W9T{3=>Y z!8!KH3{69tM4;nlsrDvh@~~`<%#^qthKEyXoLHUP9YsCnH>CeJz~S3&$Ut;6(429= zNWF;oFhf$qBk!HE5UnGuZ3M=gU+p_;#mhX^f$A?3N+~Y&!(kg-swh{-+VpFDGqbO8 zt(!MEhi3=DW1;H|0?f4`Zv9?`)lDPMmMlT{?dv+mg-yIQO-vwd6s_hk#F*Ril%*Em+c3vpd zOAehp@?-DAE8b7ndEJ*bvtXa`m=0@(l5^{jf&02TaMt61Ngk~cHaVBCol-m~$8+U) z>Z8$-4!9|DIEr(_ZBgHGJ3}pfp;~gU?jc33(uc5N+**G;jR;pN2t%ZnxFv3e*p~VN zA6VL-7-ml}+;=AXSgA6%4MhdjXgPDWTio^-VM%DW$QG&^qU)AUiEQ05p%tHdOB1aa z_PNw+U3X|;U)P_2w)9q4g3HCe=M*HNk%*;o4Ag7-Lo4Ug^|lnU)I_l=4pw6Y*WKsfxq5{jvi+^q^Cf9frM3cmJC%%|&AxkTUober0no&RCN+5`f=(n!n~Gu%zXw}v*b_tmz~xWMOyW-37J za}l$FB7>Z{5dKxEc9H5&D*;2qd{bJN!8p|WxpAM929bi{JJanRxQ_4Ed9j|%| z4Gq*hY2nvg7ss)P3>ws+(b%g-{)3}hRsruoCYBIA_NQ6ZkR#DvWrjnY&Vw%^qVO=c z?3$#uy`6~Wt8PLO^a2r3g;RiB3T#s%Y9l6ipKBp}k}BGWTP5VK*g7??R%ehr0JxZW zqKHp=IAcJ%hKf_wkDZ4zMN&FpvMaWb?Py(|I_nPzr6}{a27$GleNC}7Zg3kK`>_Q1 zUoA`$xiK}oe}olXQpK$Os`sYj<_lz}Io|BUH1TV2J_5ylh{MMOhZOsO1`fWev2*8U z%*C}h3#FxL{c1BA7qr95ai}XO6p4V@{&lZ$MiqsI@y<{U+2wa#5m>!+gru!-1j1&p zOy`NtcmC~3E#GdWa3)vH-*!$BRpNJx|C}}-RG+DAXja>G2e%krl&oTp>aBo8SY<0b zHJB3K8WWlDQw#z=@@xOK61Z4x(P}1}iyn~iGQvvX36&K1lea!+?1RJA;)rgY#Zrdn zS{XQm1f&0Cg%0czvYT_7W^I=?>_(1b!FNKaA~5PV?ZT=Nam{VWgl%&~Q#x zGF$OM%v-0y_qKrdn3k6lKn|~L&5#VenV%J9bFB@ZT64l=>6l_o0S>YBW{8}WcZrt` z6VLqP*}mCN4w8mr{RPf_s6{DDq=X4RQ{|$7yH8YM$Ce;NO>eS8e)o?@ash%7B8Eu4 zi+YaLvDN!{ejtpMm~d?Y+r?m)1yZ8#-r%+ID(yL2Net)4NRM`S`fcRFe)u4uniA}8 zTT4A{Xadc<;tvRBIB=YCTUp0F5Q!qA-M^M&Yb2HF4hIZ3lC<$EQ$aBxYa)hMFeJ;EI?x>hC2LDcE`ncB1Z+9fM z`9&k5YFq9BFf$ZNnz(w?d1)n;E~CyruvKwmHAHH-N;#SKuRuq_N75)BQGad?=)mUF zCfkIDZ-6N&i8Vw0G7!|Ce%*Gam(ud`G*X%!&0=G{&b&429i=a zHfrep>l29l{QhBO*>sJKqxk}s-BXSgvF!M8HoXh^N_@_-E10@h=7_qs6bjGd-fVGg zbE?%KdncQT1khaxGd;o{ORM1`N=aOt+&&_4_L{tqlSUWG=l=VEBi9fe-RYlB&`1ar z1b!4|g$So0NrpoT4>3<=l3A}!;t-481>I3*dJA2vBZHTF#>h*Ewp3}@H^x4$>98Yf_%mcDl30O$puGxmibe>}b1~1+Tz^aVEeVf!b%F z1hoXV`PVo!F{^V-Neg$rclkRxpL~w|pYPKlxBNesy z58O!iK523H-E1YpIN>vx9BYwO_&Rj3F}vvcl7I2O=*{0Pt^(gkzA1Y zRnIg2DakE*6E+yid!QSlpOb7^4H(w47>6w}?@Q&PzD2dJ-*56SDUJOblRo;_IpbB{ zDj*e^j2&$vg3do}7MlUIq~&cc$*fW50F&AkC1hYy$##lLx+^Odj2D5x{t zaUL^*Qva0e9V2xXJG;@s*)Bf=vMSIom_0TSU%t;p?oRdYcC(lxe$`0}Ci%R{ARen! z-0fpkfK0><63L~T*0Wch4JW^^{SJJ~X9n(27YR1ufp$NgKk=I_2|INfoMUr^H9{9ck6}B&$%9qP9?!MF}jz z@y9xPx<@^184>*`N9}QPVHRD8KYa4z2M=QlrHy{+Kkz-4C|0S-U6aF2Qc(3BZ}OC- za2IKNtJGW?A*G2#i0>)Lf!6n|H6*HP2HV&KFEr$}JK)5Bl$O<$NlP{CSzsO-$w@p7 z^L1#}m9Rq&`@WT<#t*Q8t+Dc)69k^%{hfpN$m|1*fVrH3YLlws7;pr@;HOo1jl%}0 zTz&sp&zNeXBTT_*y%H$k#t7A2-Vc@t&*Qxhx~C++4`jI;Xczi|FGs$7J$pf`QWL*r z?>Y;8F;EM|e&YgFn^PTjmV$5Z$iPDwoABT@#g1-~XQbw&wXMvh;^a4j_yoFwYhYDx z$i$UByFyU%E2rs7R&aq!t(}VNk|*DD{MI{VaP0Uc@aUBy*VM83nD8#~k?m4h&Hr2S zd(jE^YLCfvT0xH>W^N^W)Q?ZwAn9YdLl8(B2YymThPE=XB{vA|TV2UqNhh zIPUwHd)`*|A_ z(}d1*3Sp!9o^gm%4K*Nlrwf+jxXzWK%^BfuE-aiIEyG_J#H16mq5Yf2+>yM>lr!6O zUVS~=T41*>5YI+FxP>BNUE!)~Aw#`ROSKud9~_~pX47srY3I4)-6mSMy^X;${-o34 zhr^Pt5Z#IMF?8#ydj&#T_geH@JcS@3s+~1B<>~ha=_yZdkC#0<3C`N3@;#|t{^AU> z3^i3srn1RtNiy2Wa>E$%j@Pfy4d!I_W>N|AZdKKv=T@WdVbTGQO)J>7KGwW^*cliJ zs_dD4X7))iVY+WF3yb8cOK42~Qs&3C^<=&UWBRgnUzo|6yWWXUg;Hm&MhF|unu~zi zl5>=sg0#g*a$;vcB{9$BVk18l^)oOU9|H#}R1K5if$|h+*!gFvt=*RhnzH_+g<3=c zf4T(z`s&1SdRA*>scz#w)OPaRr5pGNd!RI(DRPd(kstyGSig^UzTFPTVtBY$5UDRI z;^~`>a{WSJf^iM@FXziwzvU6z>o_d!mpw-mil)^tZ~hMeB|zH08bfiWNG_OtXG))C z4WGG>zUUD!qe0yVvLF(m|1qA!vyavrQgoz$CJN=_D}l{w26c&LJkgfKG)IdwKSwj`%2I-UY|9Em|)9!hJQ+<=c6zhS23&jw*}ez z!V7Bexw2T5-1?Ae@33|4{jn)a^x5t3txH|zmhFv&|HT*d@V(uSlSQa2Hp)W% z8|66?#V;H0@R3|?^n7>@^u2ON3=`-jH{{LA`X!WMCV#6JAx0?k9CZBPJ#Dy?isMIx z4T58xUOv>LsjVb|{papa>adJaypH~U3{kFJVX>j^^vxH><|Ox;NZmTyB5+V?1zXCT z-N>+EJs{!wtAv#$GC_-iYcNAA2CchMA1}rHZG3M(!nxPCx{>wxZa!mwysuhRZ0#t) zAzDw5rGI&diT(9@)$_%L(0X(FDhNV5hTS;DLLLf=YSQ?Mr(J6JAhKFW7 zCt&NtLSv~aTcta6@GHc=$hRv)6>>Bd(AD!9u}Vp`3=cxPL;M`2qP;94xOLQah_I1VO~{j0|=$*40h$ zs(4;g=ll-7dS1BeNVQpsQ1Y{wkR(TJseu)q(S!U^?I#EuKgHhjt{7286=zgfEGP7* zI&uLYje<7x#GZJP=emec`*BV*TKV|ia^;bXCEJT~xgKWnvlLM7l!w-+ufBP;BY!=1 zwTQ_$i%PJn9ogk_V)(2&=m*s~wnTUMkV{hj!lqzST3L?aBb=f(ebu>-?pkT8k=hl( z$7Dk~l0L(A@0G_|Z)N1ia5Mg$wOkq@zvAl>?dTk4nRVqEVcg6ht(=|KrKLdDx6e4v zZ3on-lIE+Q;o0Q}_gV5Gq4Dw!+JC`^qw<;pPSw+CK0hl3I=--dC4F)S|5eDoS_x}# zQNN<1@&k&$s=el7IVsQ7&j?M0Zo=Md zti^jvqWZfBEa zwd{ZHO}_mlKIMR&cd2m#U4LVE5wOHbF7G-CyhaojBudx#1e@(ji=h&2MmgaG1Mb@D zcZE?vBNH(;MjqJD%f7Y?NND$|yhSHkco}X!rs8NTTN$1ZT+y`S{w!2tX#o!9V zBAK=kb0({Zg9==u3*hj*5jFAW4F~a2=B_2C0&!zKg$%aZ3WAFriGQiBVz{ohvfv;a zUQ&3U+4RLKqlFgdJ;6mbZ4yBeTh9l99z~0FJxYH zJi{}avS6U~%ZHnNPLYh)bH1EUj4w2teNDo2=6K$}#w*QXj*=S8B9T!v=}<}-`|{1~YxM9E z@6&zqs35dkMSmpz=t6MRzoi?{ojCQo3uW(%sMh{tOXilg!mIFOAw(pj%lQ29O4c85Bl7ccTn?G48xP_I-YEvHuKar)C7$g~j&KV%3^ayU^cMrPF`t}K(t zn5>*$CV$8;`~@CAG`r94QZ!Q_#16T?n?=V;^HZEB+%Q_)!+_4xdQu17a754qe|{Y; zXJoD{gXl)-WYGGid0a^(a5U!mV|uFjaZrI?)u*>kPO2~S*emLprobJETMS>y)|3&H5i4T7N?qTR+GNjjJQjHI8oX#heA~2cYKd zi;nE?G_ksmGljbN)U;g4OD|A~XnxKjk3EYX zoKL%`+>vI1DCTdQs&s-KCo*uVn&l=RUVExg&>>*;A=-Lfks4H9jI*qs0 zI=)yyql}8Pto;aZv?n=yX?c15A%9GMJl*{zI9iJh&%-U%W!ij7?c9Ksqt`Eu+;(XH z!2rteP-#^rTS#y=c|z1O=8=zk5kvGt7wJn0(hpHSgs5|C)Aqgdi&X zNe6#0b}xyVUhxauV61|Es?XJ-wn)B z^bp#dh)|)i$Aj}$Pvu|1iSQ&Z=ff3E)V5E(WOTu*)DguoF$%dt7!pS!`k^3aH|X2z z?N~_5xI|T<=NXQNS}+r$l0?7_wn+OJvU?|cbm7{GRvG$^^i=Sjg3^$EK`j>_$MQ&^ z;B$RHIByEvpmB-OL8bZ!34c)I(0_s}x}NdY`6Q^=$~?`^r3mm%d$4t0wc&##fPpRL zK89{DsZSstcz53Oqx5S`|C9OHfGoCERTCj!jKlMnd6Q=;bIh`42<2tt1GKAm3F|Bi zSGwIjUoVAnYUQsZj_;SDnq8&v*d(jmp$YhpRE227Ip+INB$WjQ8Gp?^zCcC5idtK_ zn(4&CSau0fF22I7+fmh}+!?uMdwT$Enz!ATAZJ*Jc#`h%2|lpZdX2qnrsmYyC*z%L zHZKT2?7a+jLBz!)Q8ti#S@`^nKb3Y9qsa>E5qpzw%Fer4vHZSN3k+mE4G!~-u^wwb zi>P5lPfW2_7yqdbOMkz8ZZ>TxSvFn!52AD%w&Omudmi!4M(GR_bub%3OSkbkUo=XJsqj< zxmDK2K|Xx21DSMK-KF9sKT*@n^~xRM_1?OE4Gmj5LWnjGm4^CFehz1>u)-N2NOq23 zOkp9MjwT?p5mG zRS2?sdREcF|7~0@%?3K~lPj3efgV|o`^USZpM(L_{eMs0jJX5N{=&3+y@qAlnKE+q z8BMdQX=%<|4^@+|mN>v>A&bH@=1GCLyCpz;Mv06NH3uJ|2I0}*)dcMWWAyTFef{;c zOSVS`m3*`Tue;joD5ShL z$$V;NuARRUzo7p>i^cpjK%dSGxZU=SaGpKx4eh9>0{Je7HxtEXQ1owYgWRqe8!H}rC=uiwHmUWKGK z&V?)!MkU5+HH77}Fw(wM3zZGz=bFovRDXCpL$eoFxS@G(iB^aBP^}^vA8DzSyPk8{ zkCi3y{TXan52=xuO196Nncoxllj4H=^%_WssodrIkzlDkZ558#eEK!`o>~irbGw=N z?gU|?E1{XA7wO9IWBP**_8x%M@msOC)&VeJpq#p&?%;y5ZjpPd*n`#FK7#>LVSj9# zsWnW|Jfx*?zU9Q$A7lmDp&<{}jX70h3c@m&Bv&O>)U5`bo9fL!J55>g?%!8EjA>qD zdg)O|D>bC<*~{_0AJ_J&QClWcdRTb%m1&u-L-;cJ^kp1<@2Ce+Ab`NpN(maZEwz&y zO$@tF=X%W}w(1+?H6d!t!|}|@sDBoM0HYfuc()+Y4dJ59EOtc06W^8r=hTHKBy(g#iDx9WSX8*$A{7F1J-4(ubp&G$^z3`-e{lnC7 z5|W4Z9ZLpw#OgEQjBmdD3Xh1A`S}DY^(K!s7TRa;8wVS2Bm;w7%O% zhFml>U5Hv7USYu(;>XCAgB8o3&a|+-A+;&gDO0!Pu<8J4JQ3MWtxpX3xS{{oE7} zq(8@-s-~a7vpeybWN^fPA3u;PP}XQKA#QOEo3@WS7#t=3F|O#fuYWCS`%cp|WM1PV zu$Ac-l|BjQyN$e=a1DcI#`F?KCbRDqGM~VUjKpE^;I7a0 z+Ml?D=VU^2BOrR0*na{ID%-Zx3WX!Q3xeJSB2yemo7Hs#rK>J#RLnlIJ<}Yb5aB$t zA7;@a1e|q*jEVJeT}e}X*EGLGBv%6GdfX3Jxt;9l9A@_HtoP8b!KXoel=_6OQnxe8 z;!546AJ&*XPde$- z4B09}y^Hft1h*(cCJs~KtO>_id30c2boU6 z{ctEShgs1F?Pu)X=U#ofS$v3fUO#z|I9T%NE#%(SkAL_TMd*Hk?tnMQ)&9ufW(YR# z?RDomRuR+J_#C+I%kN_R9e!M%>R^<;5KF)RFpWhzzh1Y|jis7$F$5QG;j)=Q`=HL8 z{-vTCw}_AFGaZO$pHF+sGCL7r6&{HtQjTA~Z4@@ZOBts^f-q^*vAprv1JVImDFNwr z*}G$TVt-oxo8&hlb%GWFktXzkNOCvFVDDhuO&?Nz2U#lZ&L%2=;nI&jhjV0}ON5I2 z9uzH4q?QZ~ZFJ^l+a#Yh4!$jCDdH+6jxzsj+Y#-_@|a0o@9`dGe5yr@+F+TqB#jI( zbMOEr&5tH~{b<-E#A#1A(5RmFNsLyvC&4{I(0`>qEomf!xu4CQhu;le2?Ps-tLlv` zl`NOfs^^5__^^10;4GFBjB6n3Eb}&x?XvJ5qfQx>^_$~q$%sqCjUy=Y5v;NsHeZU9 z>ZNpOh*Woww{mZ`?i7Zrv@j|sZG5Gol`|0N5>+m`zw)0v@i~ zPrZYwX+(w{NVkf~I9Xc}wDF&h4q1hXVmHOmV25#AkHDeMNa0j6W<0K{sDRcwGlG>j^sMyFiCKGTn`C!K@OztStWWY7>HgnkAFKI z-BM2U&}LFXVLH2|i|3K2)Yx~oTR~_Oyu0V`-AdgU)1rhdMcbbw$D*EDp(MelVlW$6 zsEr-DHZpd>t(`w3pK+2)hwWOI^VnH|x>63llkkpd!p?WMwttn~P~;%vi_>8)$CcjG zm=3*3A3NnInAYIB(tIlJsM(d2)PI^iB7tC&(BW!#|Mn$11~-4v5)?4HLvKB&rbF|y8gxd(DB?O^Du~+vqdgv+eR>iY7#A5>jx3= zYWZXLOo4UHrFH_`Ju&0#`WoA?PT#vXXOjm#VDXZ05mJQ~jsiLSKz&tCq0VNR*q}4X z4B7oJX0q-Nh@(FPM;Sszmw%_FxK2NaXWpd<_{~0S1Gzn_Vu{L04Rw)+I5><^S&G;= zK~d>bQ1-yc{flbyUZigbQyp7`3|TlBqxIDmjI9Rc*OY4B9XEzhiTb&eSs{pDH%qE@ zY@XVZ67&S;hq3mn2;iG()l;oLqBB_S)U+2R>Z#U5=^phr+pG2^6My;-Rn?{-&0h)o zbw%D{c)C99ua8HwEidZk;M0WIR3vK%G4kLT?`V>c%W*G_D6#A%gdW`y7Zul~<&}o_ zdd?pprB3qR<$hPQc9j1(ZZ^SSJ^2KSG&3sHXRVory#imSFA{pCLHnL$RZt8kA?mv^ z+WoDqvf?CMaTqu8QGZF&YS4SikDJMB1K2IF2@TCLRNTS|64MNmr1!Np`53^*oA7S( zWSPShL0I&z-IR+4ccTv;=bse>;zM3oay;x+0$+$aDaN(E1k+>9ysS7*+j&2Q$vyrO zB}v7&cA+~OaDt@kGQ~mJF>}i9;-FIO_4nV2pb0$fs_Qj$34he)8V76KGRcgySy1F( zs35-n?!Y&BeCT3WxZmJ?8P*McOSGNz0VyrySLFsfnG`H!!0kz(ieh+p?RAxhhg35} zeLJuWtDipcZ5oN{Vn!IX$esn>Xs6ZhFuF_<4SPy z^C&mh*9Ki~(gdDZg0c5JAI!QVjeSO&?0cA;JYTHw@hu1O2-@0yi6hwiRUUpO99x0c z)J=wXfuFEbPCYv|Y`F5Ju*2XHoUAf{dU?i>u&}U_dVg8BectCAFwtqpn$_$@CJqJ9 z%h=>@ZaUr4o_N0Vy;k+`^z6y?TL02)C4go^@?jKa=e(QiAcy9T7)|OcCS^tqn+_CM zRy8!Y!ibfMNr)PrwmlijG?Sk5d-@8(wT!H6xiPEcRWCt@HJ5jAQ|$Z3B(BxALEN4` z7~;f_qkp>uIYxJ~R0r7!QM)4^q>y63I^!TK^_Eu3*+!&r zGOp7f1n+KM^K1{eEPi6%PqAXPQo~t~yt*KZbg*)cG2XU+(t)Hvg0FjM!jZsUg*2;r zPzPg$@h0YqZaL<)VK7TiCULnr|xBYCJIZQQ?l~kuORu-U_3tBJnYnjWzJR2eq2qY5XR+}0?B>YL&%UfI$6C^$G025 zMsa*)jS)SPIm*bvi2d5+OUz50Eq5Gukr`KOZ|EnV1cxtq+xVx!J7IwIP6wDkfM4>S z?tg&53Mr$jG{xzXWcG*${A_gEIvVe1IdSUa)h7F|Xw-ESDf9P?hOOHJ8$qQjkn{*t zj*nhhMs~>AsPtY1MOWKp-uNUWTB=@4q!V`#6d2-^Mw=X2g6sU@Vu>ATbFv#73WTwhMZf@A8qMiu(daB;Z@qbhSwfMT(Zy!z|a77f2H5)nl7o>dG)~VJj zojr4tVTaijCie7LD#o>mL=_1TK*ZY&VgVN>6vueJ5kGG6djgbY?i$Fc=W}#Q)~7a_ z%gR_)$F#unmLu75Qm;f4Fr1(8dn&3wgmI)jGN>%~pfth~{T$9rR_Yqu_MUFL%zv<& znf=baAQbT7CO4NzI7m3^h?PQjLcf(x1(MM!>gPMMU}Ul@~6Jbbs z#@gty2cn*h`IsrFTM>IlsgC$ExH=3mDl})6WT*0gJgL3)BzIkgs|UvHS*z60k+fow6QhoZ0cHFq|h-LI#F)_J%)2!idkAHy=?xxwiv}lh&K6=>Cwk1#(nX3f%RMNWk z*5g0~W2xHK!+8p&N`K*vlhH}G?~yCMHEfbxrZhOWw9lHvhtdJ@o9+^lo|z_?>|dd4 zUz~#-aw7nQhUa+0X^5}1^j$#~GTXeL=B)IK?{A~5BBrAi=hI3-+}N3B@V|Qe{CX2! ziz}*TLKNB&2-EN*5k7@g3e_TqZ3n|{ZedRrK_oV5=YQfCjen3x=W+W^#^g3S-RsF2 zvKW7?Ofx`Aet++s6y%mI|2>yIItG4F(h?W_TT^WXAmJy{h7!Sh14MwsnE++_U_Ta3 zz5e$Gi5+VMX4&Ap=X035uMMZW4as*1ij7{R?}QHy50G5mf4|(_L4!Gs-A!>(eTmtK zbMIxKvkWVk-hYD0)C%E!kjd)Lg&_xWf8NCDY@WV|yg$RYkDss5OM>@7CmDp_5R7!=SwqFvHIYYJ28S~+6ZUyh`a;D;2Jqu-jfhMsM5vY_HhMn zq=#!sUGFp+gmPx|>^nrNl$s;LEJ0)JrEC$S{NQsmLwa5%jpS9Yb95&snzGbciWMz# z+99*ARz`C(s=5t}$OPhne?G}96p?lPG~VnxBW)OSXM-EKV?* zvW{)Z#*7AM>TqB^vz8Y+5wOAmU`vd+tymK}qQTE#z}255;gul4=~>7G9%HXy+~QO* z5i>;TWgA@)Jth7au~=$bNI;N3rIA&dS`MdTHh)Kg-=+7`!EVr3QF8SU&dkx^bKP6r(^q= zURsPXP}|gvOgy&=heze%cafZrR`&|e!GGh)3Qt3ElhtGa7nh31_h-x95DS0*VR6Um zGCxR5+>!y=O}>i@h_kEJ(~l6HnIBIKQ;rsRt~c6QL|#e=s(8-CWw`A%K@A=oK}Akf zz`_jc9vi|G0bamnGLMzoKbkLmX5|>I5aC-9ksw>a?pHEiGPD=1xL#GiP znk%K`84WLSU>Ua@>IZ%`_#|tutq3iLf#7GbQda|3v|)3UTJ9pjj-n>;bAP<8^fU0e z(S}m7r!j$}s7>E3TWxlMT8o%0S}vpX30(GEdq6u%jE(?Pg3_fXA3LKBSa`5jhz~>W zk_G)qhGkF>gK^&~%KyS}X2;|sD0vf_px}???_TmS|45pB>?N6s)JzbgSj1L*#x!Cu z5+Fjy_k!XEYc|Ehve3?T>VNMI{*{f{RtfBi%L&E?Hd8M&QN~SrV!Oa?=i+=Z6~@Fe zQ=>KI@7CJ3iA+Fw{IyU-zslMnmhlsCHIhCVBs!KKYYEfw{IQkKHHYK|rYIg=ea5>sB-E!OY zIEdpA;Ro7~%Xn=Bnt#+&i>>d)PS~3Wuc9AWMH5~Xk~e?&1lNCODQcp09R`|uMN>iT zeUvT)@%((4XpEH6x<4ORk&+Wso80FKJhxT;)FaQ)GzFIZuJ}DwRsv#0yz^%6)Kg(r zDen}eHTp*iD7E*-`HvlobkGVY^w@DCu3`*O#qRd#3)?*U{C^-K7Cj^TM@@Pjdi3*Z zd?uCP(cXo9O?_DwzEdBoFex z&6-LN^$@43S0t^GmbraM&yyv4V>)82uYck6Pm>83V=m?HwrC<>o?xPwE~*?&TtN`?|^)Ju4JZ$({4sjeYBgb`XiYdDe272EfKF~1OfEYfSdZ2YGwwz0(WN&QyYB0Xf7 zHkEHuwjZ5v=8=$%{JlQ#izurONliJz#t^ZPA)}H;k*uZcB_mssTvv2f6z>q?9vizl z{ML0gT7UfwHYtaeo~biM&7GFp9Goi`7kS1mko!<-hyE2VX|`uQQis;Tk>b`(_!P>Y zWt2ou{ynDq++Z@5s~9|Do$I@}+}u8llbO@2)>~HgVxrZmZ{dxQR(u;D>1Gj^qMH@^ z)_b!*eptjgAWBYO@=_?`foNuRYEF0s%Woe5w0}l4<)+QxlatLV3}qVzxe>pAUcmdB zQIF0k*tYCbMqh?%Af_wF6Or8uoKDf1uXgbhYj*b2bWNB2D?+>UHV%F;CJWZ>j zY@JE$?2wW2&aJuqS^&1)tZ~P29HkadWC{)bjL)Yk;%E3z}$>~(*s4whaPc<7(vL^q2_g1 z63nG@hLLSj>&pkqr`Rb<1ui7y6Ez!dFQuuHi-71Sbw22f6e+*NR*t? zEQ4Mu8!4NL!5TbO>C8Ub6}(rV*8;}NQ$Rz&aEb1nZe5xP2B}dKukso-<}_^$Lw}1) z0e()VNmVLVmj`B(`Yzd9D6lvTFUV~mIfD`(amN{dS$At&@$zO-bgq7oTiM0@4Hm?S zIxU13ZTh#Chw*v2Z(}TFL~nhdrixAuu(0J2m~E=wYIBD8&p+D&+sQVJ_4X#=K>nCI zwx3ivL?bUn=w6+aczXdasSkgd`+s5-GPq07Sj}xl$xxsT6J8G_%l<}0hIGkgk%?P_ zcY5r)9edemhwZb52~pq@?S0-QOYFkrx7-mxk;G#ZK!Td%fZRoZbqqMK-gzX5?Wh0y z>)a8(u=>-vef{QOSqzi7kTg_GpYKZ}gAIc}HG5c+%rM|9{Z?={(tg zcQ2KcZQH6{C7Tw;0#FNE!_443AXN5~1D2)9_&!8Ooq7Ulw;h=1%FWhr1`vv{PgNga zxzJ{ndw;wUeHmr&^kzLp6cm&!lteveAy^)6m8v<%PHfM1Rro`H^5=B@jUWC{8r3_P zchMB;x##*m!Ek@iJ-X%h(0|lU+^-)UDE3U3h)N1sZpv!kdmMXpA6Tv&VFJgOLA0aqfDlY&WYs7H!B?X_}hZ^PJcmTb5=bjM)A^s z4YnPW^(0&sTe7hgR5(j0j_$g(MCYc`#=rzLl#Jf=VIG)<`H53EbDLZD7Q*U9=kJ(@ zFE=dk`Mv!x=fw>Zb#HLMz$q3vDFQI}`@!QV+Z$a}BR^9z>Kbkp(v5xB z0$=2cL`05eF9gOnMpd#)R8MsXs5)=117ita`9gAjWrQz*`vQ|4W&CcO!+Na%=|SpWsS0RLc6Z+ic@`F zGUD+eLzmoUoqs88ECH$X=i!1MHHhtWVo|4|Xm5k6-ZYlW#}XVr!48X@Y0V)7PF5xx z`ioGIDsp$Zzr}mXbI1ej5Nn72ND_2d5tTM0w5i`XY~bJKuG9dNQdQ8njC@`fD`nS| z+CCrhU9ARERqFyc$G|e~Wa_0QUdK-#UGT!0wI>=`SATfcNne5dS{P}1Zr|GAfr2Zc z>0EIzx}apagT^9KmIwfaWm0j4V5uZ-$fL~ZJP zzw)BW4DZNXnltl0@;LUPtBB!E4VuRr<|nX3O0@a7$+9@ExjJF>g&Mtu8HzvX;woFw zC1y)iVSiv0&x``+8_S8>OyOg**uE3TVhVl~7G?|-c8Iyh6TpiI%s{`_@agU4#lvkbv| zWdR9r7?^b8Aw}f7fo1Dzsr-{YEb`yQr=jc2_r3LYCgI}*!w>VAccfRWcqT$2gl)n3GoF(q2sI zYKYl?-022Ik&XGWYI$3%J#Nny5-ER0EQoWf0R)pqbYc6bCY_-H=qiR#1PGjlQ~4He zwryrCmu&z>NrG5WjI1_-z8OyW3!G*A;D6Yx99Kf6_u~h$iyfg;ZFS}4^eS64TD4#X z>bD4lV!CRtGfGKvk_N6rF5xmCMN`{>sX-DHJ^vn@P|noz%H=S_wQE zP>ZHoD35q+VUFi}ZKW+sya$FhH%aT_4y@oUit*^N>Y&2+rl=rYXMub;+Vz`n*?(tG zvvB`c`*XwJV!03M<_I*>#whaV+|}x_Y-!X#vV$uYTnR!5%)0&i1*d|R5@NXm5SfCz zTVzx5eJqF3K?XzaXBaxQ-S}96rr;e5xQ(97a{(!1stPU~C*86!jqm^yOF!Q#9=XxN zPDkO;*3_?U?y9LPs{~u}MWO2*Lw|H+EE+R-Ru|*DWaDE;ubB=#)m4t=j}s#)r({FJ zCd{m)UUL0a{FP^XB*CJ}ty4im(4g*6jbqh1=QxDtYDY9u$LK_3Wq6LpUpE5D+Q_Bh z7We}KaEo#MmQ7U`&e1PW*Q}VTAvb0q(cf{S}7a{V8)P%DX_XNY+yZ906D)-v_YhAdqs2yU0IGSxk16wMIBJ) zxcy~69edbjp-5FzhMC_=aH8eLICX!nt|#|iEArZ2lCLa|&=Q!hq~7VjqHooKg!jXQ z&=-Ngcmq?~5O(kOuSIhgU4O1nad9IPGdNgoV1-0Q@Bkbo>AntMx`S6O4yYxcS>g@- zb8fso{n+V+Hpr#*uxxK&^i{RLC@oEc5!P%e1z%P(60gFpeC1rT!QnW42m0tKq?Qmf z>yE<4-Qo~q5~o*KyJ0y08W;~Ou*mQ`-bD|AcI+N5K52r4Y$BjZjepq!;q&{&rNt^( z6Q%z&H_SKOM75?=va^1Ly-z&5r6x-gmZ=AGFbDiY(qEQrgb5ZJct7KRhS9;KGvS!k zr*sN}0@64V8mhU%Qin`9`(bzNM;7MpTY$z|WHn-e6)WDp;G?afM_YoKT~5fXtS3R1 zqx7-~L!t?jF{5&Nk$;G&UdcO*8nTMbr-bW=7Wiik)$c-#&%QZbvbDGD0@AM-q;)+1 zq7Yg7c7_|@Z{VYd4UIUv9Q1nk9EnCVq=|AUHK<>t@^6p!rF=TVk_`%6*`>VEG_1f8 zN$Q*0(bitQ!I3GxT&W{71JuQ+e@aHbWNq(?8-Yx1_xy)^HGd*oUrT^31O#`aNu}db z$<>v~{$2yLfFr#9#j7prLri_s0FQ#LQEYgJ-s|60WjF!z78LTM>S7=j< z>CWJ~&Geh3FMq~~f$gJv_wK#wFhSoxtoXAF1jKQ*(mjp~46MCU)tFNUMe)D<2q|Zk z7pH&6L+RwCetyqav+@1j`5p+1XCEKUJ_K@tVHRkswhKcBr!^1Q347!%Yp|ghPYBYV z-4wBydSCsCVes@?5_y)sbiyO$>2)0tsT9TNxXIXu^ndp?EX=>0^8ce_hLuwJ1N=&r z+-*#TL42$oH~Pmrea-PbNL1M))_kHfEG`a=(M<~lEt(EWwdX{cTN>V}9# z3megExPRZs*2sVu7s^#~*!<2a2zN~@r5TQ$7!3_vT!+CVy}y2Bewi<%i0bJS@drIY znE6$I--oLGkP!R}boHt=*tC_S)lCH2$kvo%j_Bm|zrBtLo=_t3A%5>E+Q>^Bv z+3~NbYU)rdCE`*5HcIB=VTQP0o9t*X*f}X4LVx42CmfDmf=nMf)H)b^==QR3IJ)F~ ze|obnLzhBjg*7S&!0f=rl>I0qn07MmfK)SFO2f0G4>R*VDiV~nUSTFl4J1Or(yNHS} zCLZ$u7eRgZ+;v;I7GoxWPAi8F*nfMulccKiJWv#`2?F@Hn*5|+@Zf3LC74-| z1N+lX^{X(>sV&MSMcHzc+IV8AtSla3$YY$9mjDNzU z0=AYV5gl?jz)fn(3UP8JT;>GB+;0wSY^?T4tG|6ax(@N_HfOl_(wN-8x=Fq$kljhH zE#0k0BreDgv1<3PXo0aZK8Z3zP56dScYV5OoTGD$cnfB5lx+hZZuh7N1x>@qz?NzR z48@42Yi_z_QJ(J+g={5`nS{lbpnu*^;WPy_&E$GaFTci}2D2al7SIPb-Lg6+PASd^ zswRTiHZ(bfW#pA2N7<{u)RI_m|Zh8D^*NxQwid7V`6Ky-}VxFJA-AbEsZ3Uzk}e1B@#DbGI8 z?&quZ>x~1GWQE^dFNkCPukPBg_d8K{*2%f_CRDCrZ0tx6DJzI<+f-wbzFxL`-0xT{ z@-`dJ#?bfu?+(df zgtH!;dug(IQ>$C@WKv$baIuYbXB}`~oroJ#P)aq4n*%DSm9v zf$P3Ql9+Yw)A3D2qkmit5Tc@30jfVE4mk#a%DT?Zxj7Zv6Frf07)(Ek@}0Hb6^(Di z$Zakpd)##(R9CSxRfN?|3|jazt)`W)bp1V#r*~ctrydh0)I9~VHz~BsF8_@m2hmfL z9pn48-ER;B%wqyK zaD6j^gOwi~s-N^big21!aL&Dfr=>J-uZTN%DSsm2U)F55dTo)&3zgMAwi=Z%u%Axm z`SexfOptZd|8zHkjEk@4ZB3FVq`}Q!veRzngUttUl6QKwb3tFZ9yhLkZy&f`7j84v#C#bUn3usum@2&9$gEDWCn}k@uULR(7gX<^3(s8gcyjJ5rK7 z7K}?(-t5#ai#CTT7bGS=gNT}GQ14%^SVMSw6Slh{0+e=ztA7y6ja&wD1#jVK!`p%{SgYj!(Qk3iCHG(sKEeJ=eJT2^x8NO=f(i*v7E7Hw zMPc)5sJVSf0@ldFAb8OZR!Nl*?`K>HdsX<7<#tGyd`Swt#bo!N6vtpx!+lGv70+d& zl3~TFQOfT=6&w-7MR+fKLRamXEz38?r6AB%et&51n92n1NcLYUqAonc%UH57{2#am zW=?tDTRrBToE`Cs6`QDGr!-L7d6qOUDH~tjl5pbr403hYmqjZ8+zLd!tAX36I^oct z*MIzC{J%PhUp=qP|-4G3=fUs$_X6InVnc_T^q z{DEoZqCj3P7m@44G@AC>?Ll8oFaE(>n3Z}|xv@fzgxth03yb6D0JIQHqE%`=NeN~g z%N@RG9gKypgusLBe-OLz6@Q>=jcUBju77=V8yxw*5u07H>esY+sb~i|pcT=12d?&CSyMjx5wa&e)kdW@_ zQ<2SA+SuG2lfrrx#8;+0sG*&FKE72u_DE`0&O3=oc<^mU%Oc3cr`zd^8zo&QHGjh& z63`XgTyB)dIJ@y)k=JzQdW`W$T$5}1ovG^4nZ@e`!S?iq(u`(^~g~UGfL0} zB~Xrvyh(f`mKmCKde%z;I9fkSH*i%N zyCGwv9b(jL*OlE-RtM>V8p#g=;#;c2OTJH8!HX1wyZ_iTRMPKkI!O#}PbaJY2E8tXByM<6$W;{|3PkM7C%g!VF2Nodj|UoHi? zWzhRY)dP)(`ceRuzu=f=HK!38cHAnY%wJQ62uaTw<#-fvSi~H#8&e0emS+AMttJJz z>%i4YE)nB;EGkqg(0ACS0IW+_A$=o~Soa(ggsXSv#je>l=+4mpdKNJIM+9668kc@W zO}0g!>r^kXhB1`6{r;va=Rg`LTsN#KW2gdSP%8P_8x2xN>C8YPI1JCkA`=bWpU-83 z(kB2>nV@QkcH_Y<@7A*7jjQ@(+J<2M6B+ZvOAlvF^Ap~MK?2_Y}z`rlF zjWOQIf85PP)2uK0j%>Iu5jnW{7V>XEy5SOm<5Jsfhzq_@-ZtDNXC}<;f?}NMl&AZK z`$lr&QYo?$>Q~4@^wFTXgglHEKSA2f2w-z^I4I9O@Z6 zA0tGPSzM*;9#S;0x^-*f*h)do@lGR7jhOpk%dYJe9mO>WE%EY$j0zz-1#ID42wMVk zdJvf(@RtgRuu{H<8=fy|zjagHUEX3m<&g*|@Ye@`^~60Y5epX@p=BaJsoP(nL02n# zHT#-JTqFR#kDSw;2I2Wf7Fa~CrE+<;bhq|r3gBH)Pj&Gf&$ zi}6MvrAGoE@+RLWc?y!SnhU6M13OdRH>X+2y*&dy-|f0Z_lL$>h4gBK2{{Prxc+=s zde?OF2>uOOTveVB8-p)xe?!v?{j7zfvq{E?B7V4~)MlUP9XOB;-vaNz;Egf_|M$4| zJ}Oq^b`%qMYwhe!a<$}`@D5$+MnuF#@!o>pysnfO-s6Th`i%z>MjvAWR$(XF81cLr zU8f5m{{`=R#8`TVwXM>C6#Z#rJhJ91@Ny|wk_}7qmB25o%Rm);;5xyYpvt9l;vi{i_-$XlrYm{Yn|Py z_RSG-@lm?d=D;FkM6oF4ca9XhPm z7NQ)?e`M>z26c93AVo=aBxiV9H*i#*d`3==W)aK1O08YwOR6*(%tR+*(aDPj8SNk7 zY6X5TNfsC~`mn4n1 zyX#p1Z0ueq(sQ?gWL}x9@e`=qqV@_v+1*Xv`hDSN7E5uWg21ewLV+hjwlJwMDmSI@)X{wuUej?E!D(GT`~W|6~gube4*^}FISBu zE3q=i(t97mrQxp(U}K|1YMT1j?S!e1i9?*MMniC+9Cb#0jX`oj`YcP3s2P;2XGG8^ z%S>CRw6kjA zU6!a740O#Bp_y!2`+``I`71#$cthPdIqPIr;&c^E3bGcNIKuYg6&OS`&(%kcb(u(# z9H~nS!{{WRQ}j9cy>7*>)?^^`tLQUUd7j$yS*HL#(!3TT~xI03WTmY zbf~8Y5PRbHH0tI>=^+qfrYSPg{hKXAY-1mZrPRsUypxN%<$utkfu{%;hWRdMSdhQ7 zDi;K260MI;Yqc94mo!Pn7Zbk-TmSWs-)OeaVSu@_q0$^7+V9n+JA9$LIA zzH)DT88O?fBn9gWADikaW7;?nY%R!dc9AjY(GV=5iwO@=!szIMcN+7ijszi-;*Q=N#q{Gh{9 zPs>E5a4{{vMGsd_)J!slx`zP-+A@16OpL>GHyGb;WyKN8O8Gd#VZ}KyC7Wl z#I8kUJzgBQK*WpZv-_)r=1$&0G~s`Y?s6i9F0OK_hSKOA2ct9VXh;&93E|U1Rzq5Do}c zZ#DUO5LO2pAPl$BNeG9}t}WH^x+v06=8xm%xyKj^#cuMXmBvD)JobmNkm+JZzbVoZ zRk!)0~7Id(Mf&n^~?3`Pi zcX2|y_ad#jaf3}K?+P#y3conj0meLQUeP1pn!BUCu=k{<>Z7&X%Mvi0Y?cs13gE{0IdwoMd< zLvQt3lP-^k;B}6v;u3QiEN z04x5PJUc+{oC>UuL8rMe`C=d{kNeWE4XN))+s^UK?xx2>{=XN((VARA8VCK5DiH$W zsQnqG;?bk1!p2TAQFuGnaIY#zH&^O`Zn(*xeC6Ivl;lYKMj6HH>?d}}{v2GzpIpb& zkk53!!~n~8Rich=$LDGrTq>2&goJ5+0MrFb@*}K1Q2Z@I5N~W?=L*ms`0H^*9#jOw z39g?h@Fda@(?@-~3R)+{ajZL57fPc>vU`i>T+m7GSHVpTwK+5IxH=4g-0&-V$lK2| zs9(1S<~lM8|G-M<2a`(PXHjQV?9M|P4j8TdLEYw1o=nnUKT?h_{6{VXS-_Yo1h@~= z^B3yIL%X8w4E|Fp6B=o`SOiBpF8YknDGH@n4SeN%%K*YIqcUrsT)aE_OX`6C5!h0M zqHAf#_hoNbPtA<{^m@ut>U_h|GuN8rVFOmh`&SeW|Pxa zoSq58$ELE@Kd*_nrfL=ukiuK?fR^MYq&C$A1EK%a3d7eJa=)W}NL zxziJR0)}HvvgK1f!aLKa)(@q$?GRT}C_H>XkfJMjvwx+nT!(tI{> zaERnAc8Tf-f>0K!CY?rWA3E?J8k5)ikZXCMYZ{R`fO4Heg*kphSvk-mY*I~BG*b~p zUtPXF>n})>%7)8SAB#h31mHYkt3wDfwIhJ~Gkm@2-p#?u?c9`0#^km7uV${R*@gV( zrN|MkNpCFuUdW#z3GN&5c1oVa{+2UM)?kR1Oewzh4n=Hnf46*k1AE7#zEY}`#AnwT zn;giwYwHCWTK)v1&SufhwCdR#mQZUgC4YRs?xqGLtrbtb_q67N3~+<`!SF2@p>k+@ zL>7c-QV*mwfvw{oC4`b!?TJDd42T$$SD7Gv!>i}6Q=A=_aUZm91TM0D`92CN$tBM* zRTyI=OKEr|E*Ye;o2>iSoGs767MGL4Nx+7@t=S&#dR+544HRe|p%aPHE@cGL-na)< za5TPt1q-Xy9^2Bf3=oNRDn?iV`kQM-c6+QyPbUY?HAIkxJ6q=EY4!bslE%I;^73Pp z?~Dqb@Xt_mY6_vNq8)I>B9T;OOIY9VV4@TqEN>`L0i>*qg_rL~25@b0Ig+^U zjm*{WTFys{VXU&HRPE;TyDi&d8+Cjs9I7d|{DE})HM*r@x$c{g~+ zmhwoOh#Viso81pR_`G~=PI-1rog-3_U|mW&3vwihb631FpzY$P_2L+6)k74kATbKt zfx^+<^=>?70!B59Dj_(Jx?)dpjD#>Iz?%LME>hvTF)_DSOmhYK_<^o2N;o$8v-hb4 zik!0w`_-9bgky%*U=kd(s|^;`OLGdzvqgC_T^*IKK>c>)tx6_B9sU`o@q&^Mj`G0A z;$i-msg@FKs~rYZY6nL%D}|mN*O*qy9d<0dsLCMg7J&;^s0* z^0}JUf*n`cKN052Ov&-rK5cnriW=Ya5>c(rxf6l%#f#L9n1yc1EsxfiCw@=*Q|twg zJKH%}v!xIFGJe!A@ zb!&Fw_+O+_H5fBRTH)D8l5UlWwgC$BYkouWnxmir(aaZ1LWN+1*Up)%%3}2itY|t6 zoe`#tOv261^4%W=raeWW90j=SsMWy}N(qIr5s>F@wJsO1Of#n7r7owo^GCItM#1p5SWiE;zZ?%Xv^cr9`=aP06$%VJ-P^ z#kGXJuA!a}5qvhYIz;29pH9YRD+mM_OlX$fZa&yZ{xba+lxih$&ntJ^Q(=)cB6p!W z4Jdft;jx6e*SJC*?IxftdoLK$Yb#~VnSG?0?xaQ+0Y1RLPT3z-TSB5|aDd^}1~wDP zP?&Xf+S&YG2W#4%v6hJvI`D=n=Fau^lJ{VOZQ5SRpF(Jn`kS&c@{mbWWvsebPY;f+ zvx@oxM#)fTR$`v2?NPZ0ws5CO5Y8+X1Gw{9=ch;x)75q=Rvttv>fo6XE^bdzoz1VV z5~VBmq&j;7U8??fFNr@Mm2^lXy_=L$liXb!szbOS9T{jL<}Gg{MKSY3w&?(S@4QWV zJ9=$2FQUMT0TB;fBfAimxOx#(=FHWj1oE+}S)2MOiXnbakp$UpZTq8HnLza~0-)=* z5?HYuW+XzQfgLS+?Tuwrj>Og=zDovAv8+%@(FtpD-V^-@vp%>Jc-o$hMIm4P=`KsM zs7yO+_j&x(p@&>f@LurQ>8;A-5Oz;XeU@e25(ilzr^x&~xiQ&Sv;1tl3Zn#8=kXDL zd#bg3=IIKa{|)S6`7a5^#Mao!Idyds7y+2=zx2e$gVsNmz#RWsqqWMA00*N2bG7~$ zh6o_6O#k(0l@tZe3I$>PFKrR6^=%UPO#+1Vf1+EV(Lh|C|2MoeDGWppnCm}5=&co} zAYt^WXC9zbh|FC7$6#b*X65=1T9*!h+luacTDvyt%A>nwmn~~%%`Kfh7EVhZqa000 zWl7&q(gxSwRQ>hld#W$RKt8VQ$Qo)LonDSp%d34lgTS>y8+gmNjc9+b9O~a0+IN&TAAgjt_M%+k71jky~C>@v( z!L^SnZ&q{EuQvv>TRB|)B|1lW4io8-0M36RHI9aWBuRG=Rt&`hpVC(a3cFEAD3e(R zk)uBNhCmmz2OH}UuQ<>K#S;r~kO~uP4Kz4;ZRd2=ptC8sLi6 zLxR*&hoxX{c0pp%&x-7Tf*|qB4RZi~1q(xK9_jIO3ICk`xpgWGR*7>+@~m(OtjGnW zD0)@dUoh06(jsZ@)<-T?2bB-RoYD_rloJFJ=YD)eI}QqCfFcZVL4MsV3JyYJ zUGqW|h#@3>r2s+${6STB4(&uvoC;#7i1f!Bi^?7k&p5C`=nqt-pRi;M4H5(R0*T~h zUj&Jqat6(usBR+G3tN7HluA>PaZ`*u=x|0U9Cc29e-m6e{v+gU#wO@8ld2b$AIb3| z@*Hge6GUd2Bye=(V-zwV#2J}9N$po<&%Y8Rv*6d7uXcu^hl8!Y0C`vkK8qa;z(3dy z$b-{N50x;r!AeV_mj6laPl}l##!|+! zTg1bX^ly!@SVF;+N-3Qg%X;gdf3YO4xWt~gG+@Q$IM|+6!uxiGh}|19cGd9lRGO}t zaB$-PBn5LXf@1(A$`-PI~$kICrh=5j@MZSy9;)8WT+mO^AnWA`!0jmIE|csJ%4HGO$?n9(O5V=LvzvN0L!bmvd`%E(dUeTS>pK+yUfGjSI@yp==~6jesaB(U zIgS01m|ZFrWu^epqOVH+g`~NGtzor_XJmmi>^6>I5ElKYJ!ZVMe8nVQVbNwgVb|*N z2;W9aB5uJ7Cxd@cU-RDbr-7R03|4#l;ZWp5+sD$DH(yIYv2;^jQE&nF;tHr)aO{bg zCI!I&cApUN7JOG}P=q#v{wP-*YQ^I{{8O^fNGQ>h%ecQN$XYr$MtT#hl&{z~E zpY$X^#$G8F)>{} zb&HS^lkY5kbGKXd@I1Hd>MX4I*@&Pib(`x17_k68CE-6ru0{huTgL=E=QSybHLk=e ztLgo<7&Y%AfKFaC99Bg-Rim3x_e4zs=;I-*;IE5OQ5Vwlrg}FSms6LmQ?^mma<%E~cqz)N zuL;YpyW??sILdvSSd#9hyUp=WT!^0w*(>099t)c7IhY>{i|y%)cW?uFF*=RLv)8(2 z6Z-u()p-#-t#Gmev-pPwwzEMa9%38qR|x>}SHc)I@#Dl}dtr!M^&>n6D>%?ie`DcH z3#Hg`g>%4H^sVf);tDwu=bKO-#znKn!}>}bHX5a9@K_KzjoX3}ZiUV;YZ8kV_It^} zmR5pyRyt)iGy?;U-B%-5sI)9=J?E810`Ph;=@H_{bs(?m8+H$HjLg3!9Rmcw+V*a=^U_9v zF|F?jni+W=57j$^(;Pt#eqrO&)iqt_!cx5 z@C7qJ4y(vbJd{sZO=^@7K&T%w5gLn{d3XY&oxOKWn}yjNxeoaT;^IabC-(u?Toq=i zzvPj60?LW-6H(67mh~C>oJFt~NLjJ=_-fSD4zs$aFxjsPi8sqmIrRe7$TT*nfzqXI z_TrO5TsI(hjY+FL1ub%9&%OVwMAFt@Bt_k z(uJ3kNyAeu>grz}%WR@dPA8PF)-0P1Y`H)Za)r>Xl1!|2+71orHVuGp3Wzi?U$vx= zw_u7B4T`i2p;-LS<X~8;eAC?JlJ8sxLAJKg~LNj z^l2l|plb@<9S51QX4f@STC83sZ!2nJg7RHoN|LI63dsbp1y9cCA(@CmYjw zUj?YgO0~E3x^(Zf6#we-d4_XhxFug6Q(_@6N3_qE;ha>n3u`!?kMlHE%B5>rFk^R0 z6KYH_dA?cFDT>#KcT7K>Fd5mFytxi*^hbWslDWKBHgolqS84<#MO>FP#jleNm@UKs( z%D;J%l}}?FaS&|I*z!e>J!hV}Vat<0nR>LYMA7SO#dnI-x$|js;%0lAy|lTQDLlEx zxYBMhxRfx3NBxw*Oyh6B9ZB<>)vU(0xjpqX+m0IUj*#%4JfT932%6$O8+{}oA_X`} zu)0Yp9c3=t+er6GIo||&g$1aZsU+;N&}>YDn#|de@jTj8z=F=!amK1O-02!nMA#R! z8eL^j5t=K)9v28K2Us+C4)0ufOA03?)-uRNNq!r5%B%od!t`ewEb~`y%X}GH%hXkA zLz#2*auX1bv73#HJWfd+EYFqb;}AKjpq+!$&hx;-1AfCfFEd&osG3qrnOcp5YF!a- zATJzf_JiJ=_A3bnso#&Jm3Ou5L?+WgI|Z_}h}L4R`ZX9%K8tK;vHRgn6c9k9<^0p;d`M8Xv)3Kog zDrWrL%&U-8>L3x$$KIxqzQK`loEol@mm6U5fh1*oSUM%t{B%amVj4#ByN-1XmUft8 z!Li?DZ^)h2No(>eo_dA8Jew_daD&Gu6X*#2_R@15BR0!!kp{HNV(*l2ZM$ z?I?Tw*(9kfL~z4PLJG-)`Po?Vb*0#-zfnr7#%TXT;+li<63`?dyAqyN|gWB1P!YBYrF!dFs zulZ@rOdEt=bb8Y+JqAN{;<6<}hHx;%s#T2CDp+n)pvWwol6-Wk_n&JYPE21_glC|L zZLuBoMoJegpY5;A_v*2(^PhS>d_ER`;1$5wOe!b{;jM2TFeqwvRv~%H7p! zF%G1ohF?tuYD^7Db|TB$Fj#*_da8!Er!96yTRhc_H-lF*oGcqE=|u#qM;tr1;YEtG z9g3-hC%+O2L}Ut@(#25j=^X=Nh1?h&UW`6I-g9*cKkNF{0SJdeSQ1orUkbn!&P*|` zkVc+Yf%USv{jc0}KK*AshQ_yfG&}pdLBQEzfXBhVFK|Xrz<0F<)&eW^Nx4JOucfbM z&1G)P&4~SvXYai}?|;#me7LPgG;FDEEnJ#25HMeitl097yQUpiR+a%fT4Ncr#P2~# z97dNWkeY=Zb2&mPtWj!_mXdm-a+Svkik1|E=f_}mEi|0D)@-)q0@sEIgaPAjrL#MH zpDvJzcOH^|BEt%%&MJve9(6?}eKIydMna+79iJ!fYioICE*b|kNXPYr`DDI1WzJ*h z$7`G&X*0KsL=hTzHW`r8iF=mD~xwqy%C(%w;k$m61SBhHl@{O zm+k6ur7DA}T*awJxLQZYVo&duTJK ziOU7>C{+NrBu3gDYmhmPja2k#^vl17o(g@_6H65f_+p-lNw4=;XOG5-+eM7}HmecF z`1tV7d{~!Ff&!v^b~=NMLT>o>(Bt<%k^OL}+AZ-7YDbQM_&_T!Ge^iaR|Hu!N zq@ew(pv>%CsS%N&L6j`)tjx4b%#=(_l+-Yca(2d|2F@l#RH8gAOf1Yl4;wSbe_T@8 zqd;SRKJg=4;Nh@xz%YtiI665Ku`)6LpMt^6&dvTG1%nFEoo45h$%Qns`>uRu;>6i5 zhJwPhcrN?ZVv$AByh+BH-9yx3?KHfHvqkaw_yT1gK83d83XarzA=n=nBj%*Ka&$m_yIm( zrpdD37Z?iw5_yvyNb!#z2pWhDxONyOv))yfg{HtiXF^oF>BRT11S_+|21aG=zU3IX|^=royZg)(h_0FT6(cW|AI@pU40J8Vth3#weH{zw; zGUp0%JC=MO8;nNDPG#hM+&3`A$}W-ewQlI=%Qyy{cPX#k4p8Ud^|tf%w0|{vIUt`T zehoe{M6=#py-c|yhqjpE#ju{F{6aun^m?|?1G5&K^V`>EomI}|b&i9VO<+~{VSU4x z*~zly9Lc^n{NkwbX_*CWXX=ri8f>U z@VEo;)(K)xUz6RB9C1QDG1cW5p_rWLv`C_3$zV9(%NR!VB7d|9QpDTfu3GhYvZPgs z+0pG0;BBZVl6qt1);6}#)1WS?Srrr(XjmM4r;3JpI# zmqvm1TNb*QtuQ@Z;ymso={H8`dXS^$9OoOLhbG7x)*%6I!5}FL7&BUYjvYHK$_IrY zTAVGy9L-22v7K;eKdFadVGHhd4lU`$X;P{YY0!|h;^qXMZ`!cPI8MHlSdPFHHIa^4 zW{A)pr&fcBRN+-rs%AhmH+fhV&0(LotwRvFbyOD$`=kGHvXDd)WQd;u5{O&3OaYn(+om zUI4dqua14`7L+WXEjU{921ofVK~gi|4dK_mx^2U?;m^QzF2s6&&Ke-J>f}F#c^g+O z)8$dP_8pm_fs7Qb1nNYorspM!gDY;*ca>C|3G$X5T0_h&2&=#iOkPwUKDL-3SKY%p zx@j%QnKw*RQ9c2utKE&4T1kPrJn*PXfrD%Wt~gJi*fgye+2Rj$ty(G2)qVUmxm4f0LSCza7cPXDnsva_Rt4>s^8q?}**)DC zJ+)>{-cJ017)^7|_|cb?Y@?e1@Qt@t7f|@ok!f>b_~~uR=ZzP~d3eQTaBCHwn~xpo zQ8kR*iu`5_LIHhW?CXz@B7l8^=w_Vix~D^XKhOLT!`N4+7)x24kFnkDOH5nPg~k=83=B64=r@X>>9sJ>(jSU@OJ#v{+H;%?U2<30lX z(Ie2Qex_Z|8YC+QaMG_AgCxfEUaduF&|7qOH|)T(c6-WocvJ5PsK+#Wo%k0y_MaS_9GGldiZ1!R1NCpMc34Hgz8dtmlZDYzAW1-3=#byvus`dT zvEv=p29XqVMpY)crxtusD(rd!);f)#Ysn(4#skt=OFuN@Qd7`wSsPez{}dffGBI97 zm2Gy@@=FfW^YaY&(4LRZd~;rdZe>sk7@C+lSxxj$D2#>LY=~C4q3UkOaG@G z>D{L;#a|q(kjz`Ad|b|%tbJA5z4IYb?Q(#$L`}w<2}`IP#Yw|KS`=NeQx#ENvHjxH z1ArOs#DJ6LAK)ah*IILF)%@*hkyX6C<^bPkfJe33!0#|+4W65%;g@pF)bz&hCtBI} zM-qITx3u(0F#?J3>ZMfHL_D;8EWM5|%#LIjl>bR#?A)nM@u0*24kjHKMr8|66C!2~ zCKyIFA}wYjRwCw~mr8ba&P2@YoJ34SN-&I4wx)JJ7T5n;qC{GJBHSEI?3~QP+^ifN zEL@zz?A+qw9NfYjT&%*JEF5ec0!06R7xQgGi9G4a zz#)N$ghL`kt=WK4etDy)O9vZ6w4Krw1A#ISED&XTwD*@94i7RD}pa84$kmRKQIUkbbFcy^CU00Z5AUN z&VfzKw#-`kVVnhUB&%0cn};Ht88`UI5;|8pQ*I;?oCO18h59D%f9LXp{jf_`+l8_G ze`6W-`3{lE_Ong((M|QU$d*kBtd%^}0Y%mgF}QBglJV}Us2?r>XzJt}2(06-gg6ag zTMQsw4G^FAfe=(u?5^7-Xzcupjh5eVWhg6t-KWT_V1B*3mAqj$JJ-}unVySmZ$}o( zxa<<9)UZJz24KQ;zTcrPe4C`Atu4YUaPn>6B~4<2(%&}?DZ?tz^EKZs4JgCvRrh~d hwe21zegJPk@Z-azA|`?2!?Lrnu)vU$iz$f1{69tW_BsFn diff --git a/doc/Projects/2021/Project3/pdf/Project3.tex b/doc/Projects/2021/Project3/pdf/Project3.tex index 76f4a7b73..b9caeb691 100644 --- a/doc/Projects/2021/Project3/pdf/Project3.tex +++ b/doc/Projects/2021/Project3/pdf/Project3.tex @@ -286,8 +286,10 @@ your codes from projects 1 and 2 as well as the codes here or simply use libraries like \textbf{Scikit-Learn}, \textbf{Tensorflow} or similar. Your task is to perform an analysis of the bias-variance tradeoff -using at least three of the main sets algorithms we have discussed in this -course. We will limit ourselves to a regression problem (fitting). You +using at least three of the main sets of algorithms we have discussed in this +course. We will limit ourselves to a regression problem (fitting). However, feel free to venture into a classification problem. + +You are free to choose between bootstrap for resampling (recommended) or cross-validation in order to get the best possible estimates. The methods you could study are Linear Regression (OLS, Ridge and Lasso), @@ -301,32 +303,8 @@ of your model. Comment and discuss the results. Discuss the pros and cons of the various methods. Are there some methods which provide both low variance and low bias? -\subsection*{Introduction to numerical projects} +\textbf{Hint}: when you use different methods, pay attention to how you represent (and understand) the complexity of the model. For for example decision trees, this is represented by the depth of the tree. -Here follows a brief recipe and recommendation on how to write a report for each -project. - -\begin{itemize} - \item Give a short description of the nature of the problem and the eventual numerical methods you have used. - - \item Describe the algorithm you have used and/or developed. Here you may find it convenient to use pseudocoding. In many cases you can describe the algorithm in the program itself. - - \item Include the source code of your program. Comment your program properly. - - \item If possible, try to find analytic solutions, or known limits in order to test your program when developing the code. - - \item Include your results either in figure form or in a table. Remember to label your results. All tables and figures should have relevant captions and labels on the axes. - - \item Try to evaluate the reliabilty and numerical stability/precision of your results. If possible, include a qualitative and/or quantitative discussion of the numerical stability, eventual loss of precision etc. - - \item Try to give an interpretation of you results in your answers to the problems. - - \item Critique: if possible include your comments and reflections about the exercise, whether you felt you learnt something, ideas for improvements and other thoughts you've made when solving the exercise. We wish to keep this course at the interactive level and your comments can help us improve it. - - \item Try to establish a practice where you log your work at the computerlab. You may find such a logbook very handy at later stages in your work, especially when you don't properly remember what a previous test version of your program did. Here you could also record the time spent on solving the exercise, various algorithms you may have tested or other topics which you feel worthy of mentioning. -\end{itemize} - -\noindent \subsection*{Introduction to numerical projects} Here follows a brief recipe and recommendation on how to write a report for each diff --git a/doc/src/Projects/2021/Project3/Project3.do.txt b/doc/src/Projects/2021/Project3/Project3.do.txt index c3c1b8811..86512bdd8 100644 --- a/doc/src/Projects/2021/Project3/Project3.do.txt +++ b/doc/src/Projects/2021/Project3/Project3.do.txt @@ -179,8 +179,10 @@ your codes from projects 1 and 2 as well as the codes here or simply use libraries like _Scikit-Learn_, _Tensorflow_ or similar. Your task is to perform an analysis of the bias-variance tradeoff -using at least three of the main sets algorithms we have discussed in this -course. We will limit ourselves to a regression problem (fitting). You +using at least three of the main sets of algorithms we have discussed in this +course. We will limit ourselves to a regression problem (fitting). However, feel free to venture into a classification problem. + +You are free to choose between bootstrap for resampling (recommended) or cross-validation in order to get the best possible estimates. The methods you could study are Linear Regression (OLS, Ridge and Lasso), @@ -194,34 +196,7 @@ of your model. Comment and discuss the results. Discuss the pros and cons of the various methods. Are there some methods which provide both low variance and low bias? - - - -===== Introduction to numerical projects ===== - -Here follows a brief recipe and recommendation on how to write a report for each -project. - - * Give a short description of the nature of the problem and the eventual numerical methods you have used. - - * Describe the algorithm you have used and/or developed. Here you may find it convenient to use pseudocoding. In many cases you can describe the algorithm in the program itself. - - * Include the source code of your program. Comment your program properly. - - * If possible, try to find analytic solutions, or known limits in order to test your program when developing the code. - - * Include your results either in figure form or in a table. Remember to label your results. All tables and figures should have relevant captions and labels on the axes. - - * Try to evaluate the reliabilty and numerical stability/precision of your results. If possible, include a qualitative and/or quantitative discussion of the numerical stability, eventual loss of precision etc. - - * Try to give an interpretation of you results in your answers to the problems. - - * Critique: if possible include your comments and reflections about the exercise, whether you felt you learnt something, ideas for improvements and other thoughts you've made when solving the exercise. We wish to keep this course at the interactive level and your comments can help us improve it. - - * Try to establish a practice where you log your work at the computerlab. You may find such a logbook very handy at later stages in your work, especially when you don't properly remember what a previous test version of your program did. Here you could also record the time spent on solving the exercise, various algorithms you may have tested or other topics which you feel worthy of mentioning. - - - +_Hint_: when you use different methods, pay attention to how you represent (and understand) the complexity of the model. For for example decision trees, this is represented by the depth of the tree.