From bdd3d80a423df290958c88fe73127ce74d0a57ac Mon Sep 17 00:00:00 2001
From: Riku-Laine <28960190+Riku-Laine@users.noreply.github.com>
Date: Tue, 11 Jun 2019 15:08:27 +0300
Subject: [PATCH] most of the fixes proposed by AH
---
analysis_and_scripts/notes.tex | 94 +++++++++++--------
figures/sl_without_Z_3iter_sigma_sqrt_01.png | Bin 0 -> 50123 bytes
2 files changed, 55 insertions(+), 39 deletions(-)
create mode 100644 figures/sl_without_Z_3iter_sigma_sqrt_01.png
diff --git a/analysis_and_scripts/notes.tex b/analysis_and_scripts/notes.tex
index a29e8ce..d9a1449 100644
--- a/analysis_and_scripts/notes.tex
+++ b/analysis_and_scripts/notes.tex
@@ -1,6 +1,6 @@
-\documentclass[11pt]{amsart}
+\documentclass[11pt,a4paper]{amsart}
\usepackage{geometry} % See geometry.pdf to learn the layout options. There are lots.
-\geometry{a4paper} % ... or letterpaper or a5paper or ...
+%\geometry{a4paper} % ... or letterpaper or a5paper or ...
%\geometry{landscape} % Activate for for rotated page geometry
\usepackage[parfill]{parskip} % Activate to begin paragraphs with an empty line rather than an indent
\usepackage{graphicx}
@@ -14,6 +14,8 @@
\renewcommand{\algorithmicrequire}{\textbf{Input:}}
\renewcommand{\algorithmicensure}{\textbf{Procedure:}}
+\renewcommand{\descriptionlabel}[1]{\hspace{\labelsep}\textnormal{#1}}
+
\usepackage{subcaption}
\graphicspath{ {../figures/} }
@@ -31,15 +33,16 @@ This document presents the implementations of RL in pseudocode level. First, I p
\section*{Terms and abbreviations}
\begin{description}
-\item[R] acceptance rate, leniency of decision maker, $r \in [0, 1]$
-\item[X] personal features, observable to a predictive model
-\item[Z] some features of a subject, unobservable to a predictive model, latent variable
-\item[W] noise added to result variable Y
-\item[T] decision variable, bail/positive decision equal to 1, jail/negative decision equal to 0
-\item[Y] result variable, no crime/positive result equal to 1, crime/negative result equal to 0
-\item[SL] Selective labels
-\item[Labeled data] data that has been censored, i.e. if negative decision is given (T=0), then Y is set to NA.
-\item[Unobservables] unmeasured confounders, latent variables, Z
+\item[R:] acceptance rate, leniency of decision maker, $r \in [0, 1]$
+\item[X:] personal features, observable to a predictive model
+\item[Z:] some features of a subject, unobservable to a predictive model, latent variable
+\item[W:] noise added to result variable Y
+\item[T:] decision variable, bail/positive decision equal to 1, jail/negative decision equal to 0
+\item[Y:] result variable, no crime/positive result equal to 1, crime/negative result equal to 0
+\item[SL:] Selective labels, see \cite{lakkaraju17}
+\item[Labeled data:] data that has been censored, i.e. if negative decision is given (T=0), then Y is set to NA.
+\item[Full data:] data that has all labels available, i.e. \emph{even if} negative decision is given (T=0), Y will still be available.
+\item[Unobservables:] unmeasured confounders, latent variables, Z
\end{description}
Mnemonic rule for the binary coding: zero bad (crime or jail), one good!
@@ -48,11 +51,11 @@ Mnemonic rule for the binary coding: zero bad (crime or jail), one good!
\emph{This chapter is to present my comments and insight regarding the topic.}
-The motivating idea behind the SL paper of Lakkaraju et al. \cite{lakkaraju17} is to evaluate if machines could improve on human performance. In general case, comparing the performance of human and machine evaluations is simple. In the domains addressed by Lakkaraju et al. simple comparisons would be unethical and therefore algorithms are required. (Some other data augmentation algorithms have been proposed by De-Arteaga \cite{dearteaga18}.)
+The motivating idea behind the SL paper of Lakkaraju et al. \cite{lakkaraju17} is to evaluate if machines could improve on human performance. In general case, comparing the performance of human and machine evaluations is simple. In the domains addressed by Lakkaraju et al. simple comparisons would be unethical and therefore algorithms are required. (Other approaches, such as a data augmentation algorithm has been proposed by De-Arteaga \cite{dearteaga18}.)
The general idea of the SL paper is to train some predictive model with selectively labeled data. The question is then "how would this predictive model perform if it was to make independent bail-or-jail decisions?" That quantity can not be calculated from real-life data sets due to the ethical reasons. We can however use more selectively labeled data to estimate it's performance. But, because the data is biased, the performance estimates are too good or "overly optimistic" if they are calculated in the conventional way ("labeled outcomes only"). This is why they are proposing the contraction algorithm.
-One of the concepts to denote when reading the Lakkaraju paper is the difference between the global goal of prediction and the goal in this specific setting. The global goal is to have a low failure rate with high acceptance rate, but at the moment we are not interested in it. The goal in this setting is to estimate the true failure rate of the model with unseen biased data. That is, given selectively labeled data and an arbitrary black-box model $\mathcal{B}$ we are interested in estimating the model's performance in the whole data set with all ground truth labels.
+One of the concepts to denote when reading the Lakkaraju paper is the difference between the global goal of prediction and the goal in this specific setting. The global goal is to have a low failure rate with high acceptance rate, but at the moment we are not interested in it. The goal in this setting is to estimate the true failure rate of the model with unseen biased data. That is, given only selectively labeled data and an arbitrary black-box model $\mathcal{B}$ we are interested in estimating performance of model $\mathcal{B}$ in the whole data set with all ground truth labels.
\section{Data generation}
@@ -77,12 +80,13 @@ In the setting without unobservables Z, we first sample an acceptance rate r for
\STATE If subject belongs to the top $(1-r) \cdot 100 \%$ of observations assigned to a judge, set $T=0$ else set $T=1$.
\STATE Halve the data to training and test sets at random.
\STATE For both halves, set $Y=$ NA if decision is negative ($T=0$).
+\RETURN labeled training data, full training data, labeled test data, full test data
\end{algorithmic}
\end{algorithm}
\subsection{With unobservables (see also algorithm \ref{alg:data_with_Z})}
-In the setting with unobservables Z, we first sample an acceptance rate r for all $M=100$ judges uniformly from a half-open interval $[0.1; 0.9)$. Then we assign 500 unique subjects (50000 in total) for each of the judges randomly and simulate their features X, Z and W as i.i.d standard Gaussian random variables with zero mean and unit (1) variance. Then, probability for negative outcome is calculated as $$P(Y=0|X=x, Z=z, W=w)=\sigma(\beta_Xx+\beta_Zz+\beta_Ww)$$ where $\beta_X=\beta_Z =1$ and $\beta_W=0.2$. Next, value for result Y is set to 0 if $P(Y = 0| X, Z, W) \geq 0.5$ and 1 otherwise. The conditional probability for the negative decision is defined as $$P(T=0|X=x, Z=z)=\sigma(\beta_Xx+\beta_Zz)+\epsilon$$ where $\epsilon \sim N(0, 0.1)$. Next, the data is sorted for each judge by the probabilities $P(T=0|X, Z)$ in descending order. If the subject is in the top $(1-r) \cdot 100 \%$ of observations assigned to a judge, the decision variable T is set to zero and otherwise to one.
+In the setting with unobservables Z, we first sample an acceptance rate r for all $M=100$ judges uniformly from a half-open interval $[0.1; 0.9)$. Then we assign 500 unique subjects (50000 in total) for each of the judges randomly and simulate their features X, Z and W as i.i.d standard Gaussian random variables with zero mean and unit (1) variance. Then, probability for negative outcome is calculated as $$P(Y=0|X=x, Z=z, W=w)=\sigma(\beta_Xx+\beta_Zz+\beta_Ww)$$ where $\beta_X=\beta_Z =1$ and $\beta_W=0.2$. Next, value for result Y is set to 0 if $P(Y = 0| X, Z, W) \geq 0.5$ and 1 otherwise. The conditional probability for the negative decision (T=0) is defined as $$P(T=0|X=x, Z=z)=\sigma(\beta_Xx+\beta_Zz)+\epsilon$$ where $\epsilon \sim N(0, 0.1)$. Next, the data is sorted for each judge by the probabilities $P(T=0|X, Z)$ in descending order. If the subject is in the top $(1-r) \cdot 100 \%$ of observations assigned to a judge, the decision variable T is set to zero and otherwise to one.
\begin{algorithm}[] % enter the algorithm environment
\caption{Create data with unobservables} % give the algorithm a caption
@@ -94,32 +98,34 @@ In the setting with unobservables Z, we first sample an acceptance rate r for al
\STATE Sample features X, Z and W for each $N_{total}$ observations from standard Gaussian independently.
\STATE Calculate $P(Y=0|X, Z, W)$ for each observation
\STATE Set Y to 0 if $P(Y = 0| X, Z, W) \geq 0.5$ and to 1 otherwise.
+\STATE Calculate $P(T=0|X, Z)$ for each observation
\STATE Sort the data by (1) the judges' and (2) by probabilities $P(T=0|X, Z)$ in descending order.
\STATE \hskip3.0em $\rhd$ Now the most dangerous subjects for each of the judges are at the top.
\STATE If subject belongs to the top $(1-r) \cdot 100 \%$ of observations assigned to that judge, set $T=0$ else set $T=1$.
\STATE Halve the data to training and test sets at random.
\STATE For both halves, set $Y=$ NA if decision is negative ($T=0$).
+\RETURN labeled training data, full training data, labeled test data, full test data
\end{algorithmic}
\end{algorithm}
-\section{Plotting / "Performance comparison"}
-
-\subsection{Model fitting}
+\section{Model fitting} \label{sec:model_fitting}
The models that are being fitted are logistic regression models from scikit-learn package. The solver is set to lbfgs (as there is no closed-form solution) and intercept is estimated by default. The resulting LogisticRegression model object provides convenient functions for fitting the model and getting probabilities for class labels. Please see the documentation at \url{https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LogisticRegression.html} or ask me (RL) for more details.
-NB: These models can not be fitted if the data includes missing values. Therefore listwise deletion is done in cases of missing data (whole record is discarded).
+All of the algorithms 4--7 and the contraction algorithm are model agnostic. Lakkaraju says in their paper "We train logistic regression on this training set. We also experimented with other predictive models and observed similar behaviour."
-\subsection{Curves}
+NB: The sklearn's regression model can not be fitted if the data includes missing values. Therefore list-wise deletion is done in cases of missing data (whole record is discarded).
+
+\section{Plotting}
The following quantities are estimated from the data:
\begin{itemize}
\item True evaluation: The true failure rate of the model. Can only be calculated for synthetic data sets. See algorithm \ref{alg:true_eval}.
\item Labeled outcomes: The "traditional"/vanilla estimate of model performance. See algorithm \ref{alg:labeled_outcomes}.
-\item Human evaluation: The failure rate of human decision-makers who have acces to the latent variable Z. Decision-makers with similar values of leniency are binned and trated as one hypothetical decision-maker. See algorithm \ref{alg:human_eval}.
+\item Human evaluation: The failure rate of human decision-makers who have access to the latent variable Z. Decision-makers with similar values of leniency are binned and treated as one hypothetical decision-maker. See algorithm \ref{alg:human_eval}.
\item Contraction: See algorithm 1 of \cite{lakkaraju17}
-\item Causal model: In essence, the empirical performance is calculated over the test set as $$\dfrac{1}{n}\sum_{(x, y)\in D}f(x)\delta(F(x) < r)$$ where $$f(x) = P(Y=0|T=1, X=x)$$ is a predictive model trained on the labeled data and $$ F(x_0) = \int P(x)\delta(f(x) < f(x_0)) ~ dx.$$ All observations, even ones with missing outcome labels, can be used since empirical performance doesn't depend on them. $P(x)$ is Gaussian pdf from scipy.stats package and it is integrated over interval [-15, 15] with 40000 steps using si.simps function from scipy.integrate which uses Simpson's rule in estimating the value of the integral. (docs: \url{https://docs.scipy.org/doc/scipy/reference/generated/scipy.integrate.simps.html}) \label{causal_cdf}
+\item Causal model: In essence, the empirical performance is calculated over the test set as $$\dfrac{1}{n}\sum_{(x, y)\in D}f(x)\delta(F(x) < r)$$ where $$f(x) = P(Y=0|T=1, X=x)$$ is a logistic regression model (see \ref{sec:model_fitting}) trained on the labeled data and $$ F(x_0) = \int_{x\in\mathcal{X}} P(x)\delta(f(x) < f(x_0)) ~ dx.$$ All observations, even ones with missing outcome labels, can be used since empirical performance doesn't depend on them. $P(x)$ is Gaussian pdf from scipy.stats package and it is integrated over interval [-15, 15] with 40000 steps using si.simps function from scipy.integrate which uses Simpson's rule in estimating the value of the integral. (docs: \url{https://docs.scipy.org/doc/scipy/reference/generated/scipy.integrate.simps.html}) \label{causal_cdf}
\end{itemize}
The plotted curves are constructed using pseudo code presented in algorithm \ref{alg:perf_comp}.
@@ -134,18 +140,19 @@ The plotted curves are constructed using pseudo code presented in algorithm \ref
\FORALL{$r$ in $0.1, 0.2, ..., 0.9$}
\FOR{i = 1 \TO $N_{iter}$}
\STATE Create data using either Algorithm \ref{alg:data_without_Z} or \ref{alg:data_with_Z}.
- \STATE Train a logistic regression model using observations in the training set with available outcome labels.
- \STATE Estimate failure rate of true evaluation with leniency $r$ using algorithm \ref{alg:true_eval}.
- \STATE Estimate failure rate of labeled outcomes approach with leniency $r$ using algorithm \ref{alg:labeled_outcomes}.
- \STATE Estimate failure rate of human judges with leniency $r$ using algorithm \ref{alg:human_eval}.
- \STATE Estimate failure rate of contraction algorithm with leniency $r$.
- \STATE Estimate the empirical performance of the causal model with leniency $r$ using algorithm \ref{alg:causal_model}.
+ \STATE Train a logistic regression model using observations in the training set with available outcome labels and assign to $f$.
+ \STATE Using $f$, estimate probabilities $\mathcal{S}$ for Y=0 in both test sets (labeled and full) for all observations and attach them to the respective data sets.
+ \STATE Compute failure rate of true evaluation with leniency $r$ and full test data using algorithm \ref{alg:true_eval}.
+ \STATE Compute failure rate of labeled outcomes approach with leniency $r$ and labeled test data using algorithm \ref{alg:labeled_outcomes}.
+ \STATE Compute failure rate of human judges with leniency $r$ and labeled test data using algorithm \ref{alg:human_eval}.
+ \STATE Compute failure rate of contraction algorithm with leniency $r$ and labeled test data.
+ \STATE Compute the empirical performance of the causal model with leniency $r$, predictive model $f$ and labeled test data using algorithm \ref{alg:causal_model}.
\ENDFOR
- \STATE Calculate mean of the failure rate over the iterations for each algorithm separately.
- \STATE Calculate standard error of the mean over the iterations for each algorithm separately.
+ \STATE Calculate means of the failure rates for each value of leniency and for each algorithm separately.
+ \STATE Calculate standard error of the mean for each value of leniency and for each algorithm separately.
\ENDFOR
\STATE Plot the failure rates with given levels of leniency $r$.
-\STATE Calculate absolute mean errors of each algorithm compared to the true evaluation.
+\STATE Calculate absolute mean errors of each algorithm compared to true evaluation.
\end{algorithmic}
\end{algorithm}
@@ -166,7 +173,7 @@ The plotted curves are constructed using pseudo code presented in algorithm \ref
\caption{Labeled outcomes} % give the algorithm a caption
\label{alg:labeled_outcomes} % and a label for \ref{} commands later in the document
\begin{algorithmic}[1] % enter the algorithmic environment
-\REQUIRE Labeled test data $\mathcal{D}$ with probabilities $\mathcal{S}$ and \emph{missing outcome labels} for observations with ($T=0$), acceptance rate r
+\REQUIRE Labeled test data $\mathcal{D}$ with probabilities $\mathcal{S}$ and \emph{missing outcome labels} for observations with $T=0$, acceptance rate r
\ENSURE
\STATE Assign observations with observed outcomes to $\mathcal{D}_{observed}$.
\STATE Sort $\mathcal{D}_{observed}$ by the probabilities $\mathcal{S}$ to ascending order.
@@ -180,7 +187,7 @@ The plotted curves are constructed using pseudo code presented in algorithm \ref
\caption{Human evaluation} % give the algorithm a caption
\label{alg:human_eval} % and a label for \ref{} commands later in the document
\begin{algorithmic}[1] % enter the algorithmic environment
-\REQUIRE Labeled test data $\mathcal{D}$ with probabilities $\mathcal{S}$ and \emph{missing outcome labels} for observations with ($T=0$), acceptance rate r
+\REQUIRE Labeled test data $\mathcal{D}$ with probabilities $\mathcal{S}$ and \emph{missing outcome labels} for observations with $T=0$, acceptance rate r
\ENSURE
\STATE Assign judges with leniency in $[r-0.05, r+0.05]$ to $\mathcal{J}$
\STATE $\mathcal{D}_{released} = \{(x, j, t, y) \in \mathcal{D}|t=1 \wedge j \in \mathcal{J}\}$
@@ -190,20 +197,20 @@ The plotted curves are constructed using pseudo code presented in algorithm \ref
\end{algorithm}
\begin{algorithm}[] % enter the algorithm environment
-\caption{Causal model, empirical performance} % give the algorithm a caption
+\caption{Causal model, empirical performance (ep)} % give the algorithm a caption
\label{alg:causal_model} % and a label for \ref{} commands later in the document
\begin{algorithmic}[1] % enter the algorithmic environment
-\REQUIRE Labeled test data $\mathcal{D}$ with probabilities $\mathcal{S}$ and \emph{missing outcome labels} for observations with ($T=0$), predictive model f, acceptance rate r
+\REQUIRE Labeled test data $\mathcal{D}$ with probabilities $\mathcal{S}$ and \emph{missing outcome labels} for observations with $T=0$, predictive model f, acceptance rate r
\ENSURE
\STATE Create boolean array $T_{causal} = cdf(\mathcal{D}, f) < r$. See "Causal model" in \ref{causal_cdf}.
-\RETURN $\frac{1}{|\mathcal{D}|}\sum_{i=1}^{\mathcal{D}} \mathcal{S} \cdot T_{causal} = \frac{1}{|\mathcal{D}|}\sum_x f(x)\delta(F(x) < r)$
+\RETURN $\frac{1}{|\mathcal{D}|}\sum_{i=1}^{\mathcal{D}} \mathcal{S} \cdot T_{causal}$ which is equal to $\frac{1}{|\mathcal{D}|}\sum_{x\in\mathcal{D}} f(x)\delta(F(x) < r)$
\end{algorithmic}
\end{algorithm}
-\section{What if...}
+\section{Results}
-\subsection{We assign $\beta_Z=0$?}
+\subsection{If $\beta_Z=0$ when data is generated with unobservables.}
If we assign $\beta_Z=0$, almost all failure rates drop to zero in the interval 0.1, ..., 0.3 but the human evaluation failure rate. Figures are drawn in Figures \ref{fig:betaZ_1_5} and \ref{fig:betaZ_0}.
@@ -221,10 +228,19 @@ If we assign $\beta_Z=0$, almost all failure rates drop to zero in the interval
\caption{$\beta_Z=0$}
\label{fig:betaZ_0}
\end{subfigure}
- \caption{Failure rate with varying levels of leniency and unobservables. Results from algorithm \ref{alg:perf_comp} with $N_{iter}=4$.}\label{fig:betaZ_comp}
+ \caption{Failure rate vs. acceptance rate with unobservables in the data. Logistic regression was trained on labeled training data. Results from algorithm \ref{alg:perf_comp} with $N_{iter}=4$. Data was generated with algorithm \ref{alg:data_with_Z}.}\label{fig:betaZ_comp}
\end{figure}
-%\subsection{}
+\subsection{If noise is added to the decision made when data is generated without unobservables}
+
+Results are presented in Figure \ref{fig:sigma_figure}.
+
+\begin{figure}[H]
+ \centering
+ \includegraphics[width=0.5\textwidth]{sl_without_Z_3iter_sigma_sqrt_01}
+ \caption{Failure rate with varying levels of leniency without unobservables. }
+ \label{fig:sigma_figure}
+\end{figure}
\begin{thebibliography}{9}
diff --git a/figures/sl_without_Z_3iter_sigma_sqrt_01.png b/figures/sl_without_Z_3iter_sigma_sqrt_01.png
new file mode 100644
index 0000000000000000000000000000000000000000..d1d016c0fd3d478ed5cbfa4c8a06105d77c51f7a
GIT binary patch
literal 50123
zcmb5V1yogC6g7GQk&sY80ZAn<UDB=8l@_I23F(q9X{EdSq9~<wBQ4$C-QD^2#qW<7
z<Gt}5<NL(x^_+e7UTdy7=UOL7QC<og^C>0-0>PG%ex(e7puivyq&@We;FEZQv{CTe
zJv(t36?E{+4gGx}_%nvJw7MMx@&FI<AJV|I#})XH&t5{!{*9H9y`#SE2Z)8fy|uZO
zy}79Y<);s}cBWRAl$<PVESyY~CieE${H(11>kBMaw#KX}4}NArAe0c9S7Ito$y?Ko
zZUhn}4g0g%qLdgJJ_M$iFpOuu4<E`<K2oguLaCjNS%^Ha^cL%#%DrsqwOT0Iy;p1*
zwa>K)sF3lX$jXa;;ljHV)qG*$n7#sY=N`hRf46tGwu%KDt$Vw-k_R-Xd>@LVg9|R`
z%AeqfB7OpYBn_d4-F+&K9^wE0_s42wShor+gext89-g^H8u{=vSoue$G6QYYHVX`z
zAI;#oo7v$G%T0ylqPw<5B;*;Wz_wLj4)u<YFrMwakcswv=;)`+0E0Tbgg!Prq>mZ)
z(AZ4X*bLi|2zG=)kG+*Y7St|Ehb6rK2p`fDXxooh-l)6vu5*n^H_5;4({LRlJxz5#
z`?NG#{7{@1`R;meUgwtvbT1aVI_SZpzY3i?LZ-bKS9Bx%(9&xqo@9skJ#>WO(L8%q
zpbX=K!ebfXJ&f=IDExV}vciP9?)PX+IpccGy8d$JBz{>1=nG}219`lDffYRZ;;-O{
zJY3}%YXO#j`vN{>Gsk0h{T@C<3?CAK<<`P-FIFy|G_a*djcb<8YUwIMUr0e6dZu=T
z;6u!}Weo5R2Dm|tvJUs=20i?S+b}mTdZHlu%R891GK{(B&x;dCD%u6o+ZbhG(=r~g
zP28Kp`OzB&1-*sQPYa{@KUhu{M0c23x^_&hCMs`)EC0xla~FtFreJ`dBwXez!ya?+
zSn+gbpD>Il!GsE;F;!vO+`SfJiJR7v&=+dZ<6`yM$RXaY?cOBztE_1Lg6NNtQD#=L
z*P_tUuDnTP`**P6V!k#zSZ)vusycZH35{Rf|E#X?Nmm!#qBW`akO96BtxN$gi>us%
zo4d?KE2lw@z1R-p@y&mje@wl63m<xww8#sEOG4o{T)MCAVw6$4EXuJyyeNF5C|S75
z^`ZEZoHQr8>U;FXYv}NBj=5Ozl|=EBCJg$)4xJ7@6latq#Q=XgzIFMdGE4$`9H9Ke
zr@IV<&TN<2FD2+(Zn>I~(-IE<lTaBuomqOg4o_OW#fX&>j7$ku3bP~T2xd>&j8W2(
zUoqRXDo$;J<^CWZrt-C>)G%{ye@T8&v558%)i=9%24fA!3E8sYy<S@PiJ$2AbHRIo
z9BI{%r}Swu<J5Chl(^xz-%+&KZuwD!yzoU+?ZJ6_;wz%J8yD;%6L6Q~qV*s<cUw?E
z>;{L5$<OfGaii~@pcEQehm@fdjxmLy(I)lQbBC?I*ZKC%)_PkPsVDx@@?Pt>`A>+N
zS6r7J1Lj%$WDb(1w^Y~WFvJ?Q13At+$0()S>$T`3yZV#{({zS8oILqjmIlLCPiCGL
zRbq7R$z0x|2SjGY*H}f9H69XPv)-vIWfu)E>^$i>%NWXSwnZR#Ua>|(xrK36D-O9P
z<c^}Wh4slbb%_wP?p5q=+iLT&ZG~NzgwZlr{=&f*UWAI|S*vW#Q!|sn)yS=`=Gbg>
z+n3PpQk;&p<K#cC)IU$>s#i919$S;q>sjhz<#8?kKqp13N|<y>ysWd_nd~=T{Va+^
z`P6BO=@MOW*zM-IHNkg-Rzl3J_BA1FuGX`1nS>8tAUs}pD+jNQuH}ue*tsXc30Y`<
zc%*8l7dGx8Pqdp9d2;5cL4FV})kFRXR}}ov=CjWx65Je5L0y3KlapR`=!$L8l4VvV
zlOUiwKqXiQ?l?%gdvBeZjQg0UYblvPGE%8qt8U$gMseb*FJ{z8n{cigYh&|)Id_@>
zk|)w}_i{^vO8Sw7)+KsZbCci=YD0^b2ih5bihHv6*^u4q;_do)iZY843snAuUPok#
zG4n(2N%08z%xL**R%BO;*j@;kd2HLh6*Aw0Go}=NTNkj?b-m9Ep5=}*_>Ahx>*(&R
z95bWWghbrC^@|vC4@~IFsnvO(OpiL2r-^~Zo@RyOGHL$Ph<r52HHy9k(!)Jj2=O2a
zlE70OH`4Uf`+8nOC;2<yyo>^((<&WZfMkJG&c*H?kYGOEOk`C@Zt1L$G&;|8Ti1J`
zUx&uYyQ=H8qTUi;mj65OCx&F0gLVV-+pZNNzE51+`vjgfPY+3&Y4^F6ASuH57~#S%
z0?My^_VYQ2Fa}eq(o5xw)LGE5cmgD}-@vq+&dMt6X&sWwwAlh4UFSr9xnE%5HrGu*
zVk@dM_qlsN2SUR^fBRyuMES?PKJ>srjoaR-6WZ2*PWKL?&m+uH@fM5BVf}``+O<kC
z_Zrqvk+aZNUh!AQHmuFdSZVUjEpv^sLyA(8n@*~(gKWq!jUO7yz=u>}xy37HdMxES
zgNNL`hTV4UOgG6omnH03(Jqao{th$2o3$Y2T)WI`>q<sOy?ux*1|L!-Ooy<Suso#t
z^EW8CKmi)OnYN)2{E0cX>B4H1Cf~u9|JUO->DIwkBiH;MlYEE7ZFtGZtsb@Pwf7+`
z_ZXJjVU%FoK#;WQi6M>DF*amqVVv)PS=5z>8?2KtbczYf-PxGUh#vK#t2loVZW+9;
z6@axkpkZ~tk*yBBGCdme4a_;`O~Av(Khl}hzU>y;-m7f>4Nm%QJ~)!QFa1kKWQb<l
zSIX73IV^s0b$;{i=l-i9mha$^f2^87#&8_ueEUal4mCK{OPpvn{%6Sti%`M`1Z8|W
zNcp9X_gC$S&&W*z_Ps5Fy36b);*s_q1>*(|#wb7L2oiX!%2^lEQK7oY+lsb4<SHSL
z5;HKqi`=SMa#WVmx~BBIHgrd0v)8;My43lGUQ$zge#GqFUQlJN|40?2Rk2HUQbdl4
zxFvqu%nC*_$rViQ+tyVf8DXROjhvINu+W02mgNPTKdL24D(XiRf#1Z2?`|5;!c#Yt
zf<M&;vSamJ+$gm!-Iw_7nr9B;?|>&fS>D{(BDT}ISc<&|E&o_D-vM3nxeIAleuIgR
z>R^eOaQ<;vS;n_qM~uD5bB}|TGY%xUibU`doT9a8;4_hsNLgVwmZ2=wU`5FYe#uCk
zs6%ju<ENAqW%LN4NP3KCwn=oY0r{O*0$cRe9*q^_Q|i!R(na#bE?Kk-MnhyIcyQ8+
zCIfuR?x7i8Yx$o7qEq6r9akj!G?$lEBNtq)0kgkgxyUPWlVQ-LAl!jpM21FHtKF**
zP%xBXaF0n3?8PRXgZ%2*OvY9;H;N=F*Yp?;S~WFvt~an+y_qc?73kZ+E!6l2vTUwy
z$uaESK8!^Wlb_Kpty;S|(I2xdQ+H~Aj#Dk-1mCaYsCT|4>01ftQ%NPFn8@lTnF@Hn
zSfkvH?2_2#s0@oyg1xQmHcN}8YprD)jbZ;Ioi>EMa$Ff~V<q<^xvs8EpV}{=JDUpm
zF!g!<aSYZ;W%x?Gx;iq}J6H^lD^&PiY`hS(9to7d9HZ)-nb+@KnKn21Q25+&kwGfo
zi$$SG7|lgyLti|O-*^Ncf*d-GBT+q@$&Pkut|>TYkF49gy3|2OxiKqK-SB<HYV@d6
zC-_q;jyaoaUc=h?0Few_8Ft(ZVkibgJSlBcMYQtGN>htq+`ZqYGf%|F5&$0zoZ?1X
zBpIvsqH4J1VD*R^W#4&I;otwDn7g=frM&!_t&XPdi%pd4Dk-TIN)alrzfJG3NCRKe
zb|;&QSphx#Bz}|Mkorx1V%t?>SG|L7$bQHHH2?Tf(y@a#;7prDr|Czd8fnOiFtWF>
zQU%Qj<sV3`Z0S&q&BI*WV5_byuM>^UC?b3!thTj%KCk{Jq*k<1>EM}AMk0Je`g{k1
zRtHbC#g|La;0*7wTVtk%vp!cG!@yxtNPa2m8PQokssB>IKGHPBV$+6(o&@daD(y;N
zvrfY+yIbp<?e32CQiV;q;Oybg0LYlSPaM5lR-6SrNeT;7h;(tXd(s_=q)XK0d+*(l
zo$#`dtg<x)yQ1v{II~zmO!5nnZ;@8?kCwFNSDv}8*QO43^t-Z!jIu?3jv|s-6x6Hh
zKDpJOdgFk)_&SKND?Y_LgQzgSlL+4hi>-IvEZB9djyvErd275qHREeGm(x65d6Rrx
z!<vGV_%;{hGxlx|;ufbUrUiz;h&|mVC*9Ph_;v~O`8j1{??-^L|CWH3;>YKfrkd@l
z0V*Fj^<3!xk;!A#;{PM~fn3>AGXK&3f$&%V^GEAnju6M!{h1OD=LgH%Wd6g)o1-?4
zy*mo!fv#c`&FnJOyyT|%ZPeQfPT6v#Q6*m&ZA!Su>pgD)@czo{5Mm2b2LEup+$ycM
zspv~!ASFG0`;0z1$pCGr%tXt+bP_MEDPbFx#X@|xHfdN^X^`?p?^bkFhU5Rc?r24^
z3sZK%!^h{dTfwriu{pgO)6vm+B`G;`xQOYKltl8OBl^$q@bFB=-7^JV_MxQEr|{X`
zS5Q#sh+_>?-bf{%H25{QzV56L2F21aW6`baO5im9u9)$;wh6I=z7J0`&|KEEwX{ej
zz7SNd%0w2i!U@2Szg{b3-f1N#Ho?9iyzoV8sEOr6F*_eyIwG51X<Mc{>9o^~l14v1
zxZD}HzrX*dtt}vu=CyK_z3HZ{xOk(sKu;{Q=H9`<?~(1r#g|S_PNADJFj%JEpU>rF
zZCm5Te4aNo4(BndI`eD&S>@Vk|Fqf3f#_`3dn>Sp`M!*o5MO`)h57kE;m=-LGGvuX
znuSCSrYz@$JbhtsyfN&k(w34a<jP%FR~OBsj^_pTY3v30KUrrPcSJw_XMeFRSErWW
z{qm$`bTn@8o5Ji+RI&ZKvdhU9MuXo2LZ_Sa)&0Z6rrFuF>1s#gOJ1wGb6W7Va+dn*
zlS#Mz7Q(B;UJ)+KDI#56-J053^?HwM=1a@*-+tT{6KZjnC)=Hkeh-kGPNyAn-+T)_
zCM_9Z(ku&p8GzGY=k972rtWheJ1=Mfd?!~a>;CXhRpX`hh=zJ0H7d#AbffOXp`WT@
z^|v<yXsXKp&TwkK?}#3IiYxr(D((k$mQvYIJ8ur4>rK8!_?C>Rx}MiD;2|GAc_OK-
ztn9z6<9Y4ik4+*eDhg3jQYta&|E%kAz8q^iKR<tVcBU4mTBsuzfI~*aZg8*G_4NMS
zW??<8pkT7w`N1!v?x%u6Lg$ye_0Ai)-%LR8ml=94hVT#xxo~Ni@qgNC&yh<qnXMCw
zqLni^So+J;g4DX@V^x364Ia9wv-4|s_`{JxU4>$!?!eGcS~j+DcEeUSr#(HR-qh~E
zM^wMbzWzJV%J~eLXi_;iEav;0Z}PFrW#;3ygYLp$Y0Ir4BukxfyE(S&*tLwhb%Ng%
z(pJD$CG*+&BO>XC3KS0yk4V^!SHN+n`S0H#a4>f3190=%v=WEScfyxjoC^yJ753|R
zrDmfm<Hl*-MOR@WH`0r(AyADHdI*`or}tp7TxKJ|;EB~w^R=tF1;B&7!}^zZOF03w
zN*$KH3Lh1eU!4X{J-Bb}=vXf7XlpBOW}#h{GNNJDohszo9!1w{LAL5gsH4~N<?$;C
ziT;e2XpoiOGzEKmE;Jm{hL#pTjt?DE95H2IzdnFMq2<j4BqW_)NGQD4a}fCJxRvel
z*&f@cmXmg`J5OqoQ&Kp$wf`Z~f=+~MG^tlq6uuwk<At$bhS}NKMIVP5Hp{^7=Ex=(
zXv^2G(pB@@uQd$7U#nNz(y-R-Aw&B6`w3WeNC^oE>n?W$dcYbxJ(fFS7TY7JsHv%~
zk2l`d-QKuDET<~G!KEd>K4~%dRb<eDxhb5QnySVWtGR)Rj!p>zp)-y(DmuCuTpI%m
ztE=DnA2!;s%49B43ZF0GC%!O~w6<mgIP<F@jpGAm^{4H|$;qT2Z+~(awg%guD%46r
zk0TPzhx62<z7n!3XNaSIA!Nl7jbqV1o4vgWNK2ytA^QRx${F+V+wY3rAZrw}B(bYL
zZGR~&WYz0h8qL?_`Jq}UKUrpxt65&)kfy5R><1obhU3~ff#ZYIZtW>LCT1hR7B4SI
zB9E2FDxuuJOwr2ZAbH{-0<w1y9CEd(_WSqL3c}~_LpLflU+ywJc>tc(Q6->)({xhg
z`SZ`|>6DkK)B`mC3&SJ0ZDa>nLfiT$UO`KB&9>v-RtBt)!?!xNU8DXpib56qXAlk)
zh?iGS2ePRDecVnaCU&M(eV;Emyz7y4SmeJr7tekc^uPV6tRc4Y^51^$_eH*f#yLAV
zy{|SmF`+kX3+*nwusS7L5Bl=OV1J>-Q7i1<B|@d~Xlyj8mGJK36PH6KdP{G*%3&y9
zGft!Q<0CvgkQ{`0lO7lAbWeFKyB5PldV+~L^EO35CTtAl`jnUR?ABdcRvjCutTf?@
zWW!<@e0_bbkJkF_Pp<)vVc_7jg5#T;)00o(M-@HYnO3iJt1Q*}h&uu@7K9&~u&{72
zg-Ghs(tCd_VsQYodnYGG1KBbV^XaM>um&Y?crHh4Vs37N5CHoKGy@QZD<UF-L&_^U
zm?M{~l8dLWub+@Q2x_S5Oby=)vyp6<;|+j{OqYbZ|BlYw$teeo0{y-5yT&07v2prM
z6xnuXdhDlMdSFTP@OVBu+BA;~Xqx8@A0z;m!U%k-Rqv4k?%!YIYzGjIFqBN76+}s&
z#^8^)jAd4{@=gBO`3{!ge$yN~J3CC;Rnb0ZxRKG(z97s1I@HnRAfepvD>YNCa@cGI
z#mD*jcvQXC#le4-&8Um0HI%IL;&>A&p5K9iP9b$^?vGDJMTPZp$7296YcsV1n&p;o
z+vN`X?THwG^BmT5qL7YQ=1yFZtA=zjAH+96YCLCQ3H><yqrcp07OQuCq2+5d!`m;Q
zpxpObTU$%!`FwZC?b=YN^$F8Ci*A`Cb@}o<z{Yk^lMzU}+LJ={?3p*_V+OVIQp@Q?
z09p~Oy24Lid~62KI#y{%m&9Y0^DbWl(%jr^y*2hr6AymWp~m!NqRis;&DA-Q$?%W2
zRqWS$O}Kelm3Uq(+Er}kV@0GXgP;(vO_sB=u(ZDZVLjh?bu_4;0<y6qh7qL?Bp?E5
z<f4@g5jBdPo!!aBWnq0i2oQ|X0`1Pd<W!FfQxFb~d-F}E{h6o=BG>kz29NJzv~sH%
z!|C`fuLmg|E^dVl%E5B6qYB@_9yM%y`y7MgR&q(62y%Pg>q#W&#Fnp7`W$3pU!US8
zH+)F$LyWRXzVl{&`5M4=e+MHKqefqh@}UH>+)J?e;gOLBLXzU*;{I(4uB7=6Ic@Dp
zJgg9TdHIpm+)f!8nS|8So9oN*B7GFA+T$m3iCp?a-<8Hi`0mm*LO@wrZ-Clqles+#
zjX`y#Dv1Pjx0o7ZiJYbyl&c*3^#Mp%63=dqh1p2nGjMJI3Gp925=-HCFqtTU7901{
z($e1d0>2-e&U((4=t_Qg%5B~bfUm-4@&1#qeh!->aFRFI7su*V_E5;zC(M~*KKH$B
zSo^`6@_*6L(jo)e6$Eg1XEs{~zp=4V8U{n20~;>5-;57xh|SfR4de-vny8+hURmKA
z$<?D#UC#&+PY-^_oi4z3K<N-I+a&n6$xElR8Tv1YcppCH@u!lSi;$Y*t}Ai8hpP}4
znbO?as+=QBym<g(76Tg_u3h~pnngPiprw4Opy~9^?ykb&N;jXw2ElI4Ar7d*nJT%;
zfI})5>IgN&E9%cT`hi+9Riz8+`~dUf=H_dFrhv|6XjR%0<Ge}YwRtY&ax_=mE1aeF
z%K*?vG%uZ6m;NdTv#aZC1fWb$lR>PkthBVWzN;700v5F@a(j-8AT={HGZj@;13y(+
z?_&`)fU0gXT0jO)ZYcLnK%M)g(dEPkd-=973gflD3|>CI`HhVZ@K?p!-|iBnqBAl|
zzww^arG>SfkMm*rYeK~@=`ZoYJ{o+hHLSsA_&Kk1`T!tLt<`K@(ujnl<iS#u!b)By
z<)Kw`N{itywY{UGnqpWpnat$PqeqYKp`wnnQbw@+Z%n)-$T0F#W!rPA3@O~PS`v-Y
z&qZumgntQC_QE6RTh7MW+moR16SMqcENTB4_+znxnY4c;pxK=jiXnV{@8d9s{}i*v
zy`QSH(bUts@7GfP0r~SwkMou_iHT~$R8R7{D?C_qk)w)Z$`MfacO~R^Y&<%J2YOwd
z^#L`L=Kan%k#HHj|5-_%GX~(kJl(P0qb#wql10+VQyaNFKYM7mtx*sVNAf=<OhldU
zJtCo}A4mO*jrChF)vCB$Wd_{kU7`nIDKs?C8&j43l^=awG@J&;LZVuk6#hSix}p)!
zPeM0R{|=MWz*RWmLrTENhYtnXmz}<aUFSvfXGc%uM*sL&z*-nxrC%VD9gW%9&1)AG
z5YQ&qX}8+*p}`C3(-&Gtu$>W~KAvm=xijZ^dzt3I$HWwTez-cD-}kRx%D1$U)CVo?
z;u5kpCcY@Lo_`rbEyi*}4}~XtLmN07i1iAUKjx2Is3sYFEgjJ}I)B%|o1Rd)?K?6-
zSH6HB`t130WL#X_jnP8f@GOe#e|Y0KK2s~nIuJqCmroK*f$N7kF2GRT-c3&9Iae`q
zft&AuTtr}Ep($#19oC&Gu%~x*VZrUNJCg!xm1f)<*EE1X=vY`w@Hm}J0O!F(?4JQU
zng=-oN*%Aw!ad05NPd>(bk)$0w+MpEqE!JA1tl9~0MK-SUmi1Z*eo>Lht)eD8f^^!
z3;@Mtd!m$zmexOngy&%3HLH5(t5>i3et&ok0p$cie*hvcPCL%ZmCH|(mw;6a0v?WK
z81m)IbFf`Noh&XcHi2S>j*a~R06K(}-#!qeE>Ie#lV!|+BcleB=gl_-EP^EhhByAp
zP%5;r_AbbZ9ucBZA(%g4BC~ENcB?Zn)KHq`o~Gh}sT!$VF6#-Sb1X$3i)U~ZPd&D9
zG!ygU+|SAzlzX^)Wg{0o!q~W^=m9t%v=JJ^qkp^&jaIh3q#ZoR-u`*H;TC|Z62L?|
ze|Q46@tleZs#U@EnDJe<ac?Rpp8>3OXP;%FXxTxAKIJlL*qN?Yg0b7J$O>N{y%D~>
zvV(YBZu2xYHT_a!0(j%+aeV>O`(yqspqg%bf6&awf1^4&IwCkG5CLs~xULU=!yPTu
zZJVA>`Oq1QP)`uytG$<)ge<bQwj6){{2>)~BeJsE%`3!zPh>V*=f1zUr`I0-48cr>
zs}w+`O5ifZ^Y-=*e){4|f0h*IpPCtfHHfuhlW?=S9GL*&5t+~XADOFmB&n6IzrJ5s
zhnePyqod=fwVf&?*`{Kiho<49?D%ZNm7WJ5e#FwkEF`J132h*Nt4PgRq?1_VIH{Pq
zt3%(CsO3lhh|NL6gsWJk^N0rF(F96!BKgF{5kC2f>Q#e*!AnDDOk7gTEiT(5IUenM
zZ=w5Q5GF6*YVa$ZRktq5Meqr;Mzc2xT9FzPH7%{}Vk@>;Fa%VVu`x9_cf`|C0xoc{
z+&P#oh631Z48z-n>fO2~gI|EpDM^Ns{sml6T3R|C;548B3yX^w7#PpMKdEs(R4y@5
zl27Ip16nD+?fm9)m(O+y9q45ga7iRjPfs8W!~m_=bw7S=F;Suqs#wioj{vv|5H-D1
zm3Ba;FI=4;Zq3%G0WuPvoZJP-T>fgr#VI%}uvjDFpWv56+R!qmaSuXnfCYntuGwpR
zV81zn{k26L99ehbi(*=D&#V15xly*aT3S<bg+cN0@tQhRjW-Az<PLmo1n)c&n#<&c
z#auR5?hmo7baf_5p&kbqs@K;c&<3kFFwCtPZQl=_9kyrIV#&!|eh=t=f9Pb<(%von
zV_2C$pMume!4!vd(~*6ptjFLr!cQK^leVi0OMx(;@N^TUm!{&#qU`~dAZF^~4FY-p
z=!ok{-4%zcQj$Z(`$OzlI-lbwUt3(v&9J-x<#9e(`ZG8fh2UO1Rysf;_iIFjhkJed
zMhC!+&hzGK<XTHx8>;gCA<!2i85ZFG2pIX}gNC4@2z<fDZUGkn#{i5Ao534POG};>
z1dmJPGHnF1)-tFTz^H5jhaboVL|*4A!+D(#sH5l<jCN;g2^drYK{jPYHx3Ts8yg!>
zO;3XrmlU8V!axQke#pxmg;bc1m$3kBOioVDYkOTw4$3tUj#KuQWusMq9*!&r2BHtt
zN%~&@UyZBR+}jhV?iUTa^)J(eE8H)jFCGw7Y;P<qhMbcN-;#ntXs%AggYzy=5ircd
zE>3Sp_A^&1o?WGh2Z9CBKhmQ+Q0yTi6HgRjLYBp2q`XE2tVZ(py_J7F-{akrVrl(I
zh!6$0H`kCn^<sZ;4bp82GBS&88lc>CLQk;K(e*&0l8ShcdVw(axV?64s6bn5@J8v1
zXE#J)dig(Q0h;)?b^!bT1Goh%D=R?Z1`v3I`Q@*I@Bnp*jErmqOzd~?@Zccl^b95V
z3JBBp+aa1odPv^hr?Vuoh|1TPp;d1=6?t$p*lI;>F^;USZ=EDFF#L0GzfG}D@MsIe
z`2aM9g7Dc)9#%}XiAYA9WtinhMJuCtSs@wg0gMb(-VoAD8*yP~6NNrTrbbp}_n$`M
zx;OI%MuYTMR2LZ5Q(uO4fVhzWRb&73l#Z8|$kFi>JoD41FrWwMwX3+0kdTrEojE2i
zpD?})uC!Y{IOt@>r=Y-XK%;OI1&D6*=~E6UL0u^VqlrU9Kt>}(R=&myhIe`GZxn=L
zENg+#&H&$tX4Z@aB?liLKgu+ZS+i`Zg;4iB$T8VOuJxR%mF^@dfQdmNA#j71FMy@d
zNsUU~mH)A^vhE7B^=6E<S&G(RZ!VD!QQDq2rm}%Nb~zEQFHDgiJQ20tyLMz%mW&=b
z9aZ}ie0F}mUQ)k|PfYwbN<lcG_{+`!ZqEcxE&fSF*)^PT=Kp4gzUDVD2qxank@bef
z=#4N{bVu-`;Wiql{ZZ8=#!w2tVSA50l$YRfo=#cuY5Qpa&f}Wvvt17OkVk&0?z`ok
z#hxOq=RCJ})F=?GWhB2fzCz@d%YVv!4{6DoAZnz@xj2H-Vj43wRr0M>jsDcvqx8uC
z%-cY;ZjsxDB7HWCr-C=UTVqJXQ~VJTzW3Cs2wFq^(mef@;d4<6>5f+P@!dF?D5~{z
z{fG2@tgEYf#)wu6q8RK`j@OK@N$|Ul5D}ivxB_r4bG%)ph!>tdtL!T_&d_)f&xywW
ziHp{({Sy%#H~yhHT?veq`xnW%_Tg;n-!TQt9v8a1s^VLrnwb@HFW2;qq0@AMcB@mR
zIjxLw@6!~;51oyD8H$x0=3{zCv<|ze*VIs9XzHI?oSgN>oV88%?-DdS)zldOgW3gN
z&iZ46U9ejX4kDRxC!Cu4UAoyTLejMPgq7QatxQ`Wi>iaQa~;R>c(X-!y&{uJMEfl0
z#^Itky;Fg@Z8j~cO(rwCRXg!W{-5BUay<UA?;~-zLTtG7F8?#X!|y3MGjv=Y{$y^8
z&sXR3rK7<zPZ?C&oryHea<<c2)2-0vBiei%SPQex7rPVO7iIA9v#b`Iw}2nA{QGdV
zcMw=$3~ztD0`$gtr*bt<WGnFvKCLVzH}^?F^W(>lTY&n_?*vL>@px3Xp3z|v6BG0K
zcFFdgo>^c~s~(XKeZt~fKI?ydSu6_@TJJAjRyjQ{5e_LEZKHtu)4H>TlgRAXzv4`)
z^O&#e)}@+kvH7M>*-EF(fK~axa*7u=Q}YBAGMgkiQ;|se<>&b?EGEln087KCliveI
z=Rv;|g$+<QlRsFCUqSi!_@o@*a?w{;SMh-8S+vmH9obGsXIzGzvB~t)FOPkMT<5M&
zpIOadIR*u><7KpwbJBjO@R*J-n(q~vufO$c^s{1(X7)7y<NAY#B+Tl)R+NqTT7|b<
zCl4ZH0rWd<{c8IT43KkMwglg-@K;bmzkq<njSUz`&HPO_H#et?^>6(eoc3#q2+s>}
zPPH;~;=SFl?Ao)1uLxymzcGZ7{{lS0Qepi~CU9H0J`SNn0Mtu{Je2@e(D{4-1;P^p
zRw@o~lfOCc&Sc-j?*SqSs(%JBl+P|MY!{j_B!h|m09v`XzpsCFwvXVXVACK4T9x8S
z+!nyop{1nsY6`$bc?ggyj9ds<Y?%Nz_~i3qnbZ-s&v#&&gTF5?NPA)SB!*Fy7VImK
zee8M-UX%4}cc+N^JBBetC;LMuB^SZT4yx`f*4H!<D{ApD;g48988x4b<az6Q`svjE
zb|)ER#*4%!s&c09WpeWWY>9gQYc;HdaWeEtw`O0K6z<27yuX0sLIGUDawi1>Q-ruq
z>9PMecG@@I))vprPE~((@Bo-O@{W^$o$}eQJrub<!UtYsCa|-(%*P&j0nf*Fsr?b~
zy;=ZxwiIYrI{}}En8%V7IN}7%8U$XOqlIsP?K9)LD+KxRF2CcuQWj)w0R8}c{tfU>
zu>V1is3iN!EHr`hBP}DN4=PhInLsQcKq!)*V>bYohl3KEr&F7_TYsAh?qJ(q!v>lK
zB;XE!LH$t46$d>6YX^t!0}XI*T`UUW<m=1RH{TVXAqW6~a8M~$uTN)JLD>UN%Rq&V
z;YXvR!JdDc`tvK{yvQv(UbaT*+@#gBs9xFnYa`h{P}rzx#V5;JWj`SyIg)o+#d%~c
zehDQN^bD^nz2qqSs}{?w)e5XAgggd*UCqga8IaUK_Sy`tc3u8`8AZ>Rm?GeaC_jr^
zTe5(z0v+@o!8hm?A%GHgL_<-CgcLy+fy<<CE|wp-)XwvMgn6CdvD`q^0-FH{rpKV{
zN-8Qo1R?+qvJF!4VbRP7T=q9GmmvsC5|sa-P;SdcGvLHIAFj|?&C~#QA^8Czi<)6~
z*G#P|28cA^BR~)vy**iu@{ou~TEDoy9M~4+?znLXWz&ozWp;L9M#Kgw(B`zHf69$D
zaGt4QHdZ+ppRPa6pH;p#tQt~DPVRYS>6@8gqjw*L2b;Ry%;R0TZb)G@D{GlYgj^E$
zBO;=9Ktez|4x~_YAqjE?DI0;-`_4+9BLFQ!(63v!d`<QwF13bx@pVc{3gEo3@?<~&
zdwO~x%`Gka6J~|KtY(FQ&G#B8K+t|mAg7~R-;SF2kT^Z|`;D>DrXOOp<Sv9PL~r`@
zwM@I}_O4GAYcH5vBBMh7?tuD3;pzKaB6$4I<tdM)eiaEzn@-MXl}_G;#p7p>U&c?m
zuq4&ZygL3f4DO9+u^erV;?jLXTkT0jTa_hM1Z|9<urTA+VB!%6*VCO3z#<1h9uA^3
z0x<HHuY}`u?gBSAHwh!FcUF^(qbBI%K>$|(_ET$O`eS-UFVL4LaykA80rCKFgYJ?7
z&@K9H(*L}HC{h)O1WW?@a8q+2u986)t!W#WT8O)7RoMpt2N(ebdU}-L4(5|(AwN|M
zft~t(`|q7O?POwVarIqk_{Uy151SFz)m{^`jIueC+c}L&uEEtwKm0Y@2h3#t4^wsa
z<&oLD3LXmGMTvR0?m}rohs2yud(x$<$&a?i@j+M3<L2D>;&k^`ltzX!NEr+z{sOWS
z(bquKm(q{JfKS7Lge<X`d;)AHfE6-eW6ptt30437KIiy&XFAC^&3%MF!6*MS>H~z^
z^l|9B7=XF&hMX^!z()Y@p93pus8BZzRIbJC?GVtW0k-sefQ&rDkDos60A>`z=WY&s
z1Uy03!3xk@<Mp_za5&R)A$MMSBz$vb2u#@bDJq#@hY;>s1eKI>xurI6C%<cyGJvKm
zBu~3~#HBapE`DFmPw|-y-sgAt+z~TBCq^Vps&zw#{%Jom0zHP&_~_|lX3WmGrv5BR
zG?kkxN4jX=rF;IKJjvnb&zzR_=K@qc)8`6N_Y3B`0lLJ4t_#9^2I%a#>J=(!b}cIS
zakL<R@B(0qm3mxEgXp5oH_*jQ6DV8~-UlZCCNLtWPi~%kl`zY9Ffhop?9cSmskB|z
z4q~*p3<wCAjd5w2{=c+APA#glvrUAoQb$l9eB55iIUf6^K}R#KvNXn?e|hqP+h>GX
z^JA@h^{>t8Pe%hv+O8TFa`aWF6Epk=YS|tu3C3LMBriT@-<~aonU5};SRii=j<>Ut
zC%Q)e-V{!WZYluft_L`$maezJ{RKi?97HOxY%@TjI<oWM@h~*U6Q-jbpw%1RDE!$P
zm$ZsBJ+(Ca?Em&>Fdu1^y*JOSFI`&AWrj(9g1a?$p_L064<XOH63svGl(@a!B?TA^
z6Er_*+%Jxt=lsXvQg}4_j-b_2>WF@atf#3XY{ocP<X9LOty_jDz}w}7LW?|VBPh40
zFi;;fYFZI<`WoAmw^~%&E@IY>9vew4VLnwg>5{6W*gqk$D*n*<1c7h>Q|NW;+;&K(
zkYvCl)ZKyCe%grc|9Z4t-dFUjD-ktJNK%WGDIPlw2yHu+g5(I1+6x9JrzIWUM@o<E
z)rb8K*;k421^toh2`IC5&>^5L3Rg%IKHbJ`n<+r6@4Ztm;^<NtFJ#j!yY$NK=wwr|
z=I6DbS46-xT<#<}kE#}6Qb%qJYa+2)(us_YPi}9|r!o~RWIxwB`1N}oRu(+dXm+iG
zaI&sI1N&%u@-*Xhq^JuRH%4qm1y5#Xy0OAo)ni~NCvft~ptj<2@kbjNW~g%#@e@3K
z@p-C8E>xFKzU?vD_~odM%h^%GWeE?Nj{q{F^$e;8MaMm83^mO3UP3)D$+Keqe>AiD
z84YS<quv;swJmT`PeymTjjf<{Hxp{Zzo@W?kzagV^5E4lTe;>6Ln669x3Sd`eep((
z-3Ya~K21(q8XcGk*}%=HqzLo@vk{eBz=E(}>jR>KL8<iP+f>&bKA;#+wL~mmy}HM&
zQPMta2#j<kp%mSK9#djO3<aW%1mQm<#k#)C+7xMNm1#4r$;0(<ARys+T0EK}h;g_w
zokYF95%SYst>0U9q(*!@vXpp(KU&`+;FsgBPL4ua4-ir`G8{5t$}9sKgMbAf+C@O8
zBk=;ZLCB(&{<}Lm>@J=bkfI`yqiFxMM^N*w8tyYr5>>C7hT5w}vsiQ)js7Hba-OaZ
zFT1Tw>0uPGV_6}6nM%}HXL@;Nb>`^|5(d%Z3KPCqt9`H{OeqAbq!$z<RaI5JIN3JS
zUIm`upn`~;tSlCQX29r_LDy|0PaS1y0)_L_mYVp>mq-Xud;9ZU3ux!zknux`jC;v}
zQ>p-3EZ=qOg{8vCySZ|J+#JY}Bhe@`mjUhNcpyUoa%KVca_`<fL~pA>dJP#F86aN{
zsO4I<E|0wsM5WkB4v>_#Xy}`~V_G^o1X=yBw+PyMo&BX671qx{O#t%+ecu!)BBP>^
z_V)JXx_M@s!Gr~9Ewbp;#AmtO)yADl!!}B`@HOwOvyv@phNb!jM!TmAOWE-v6gOv0
zL`IH<k=e~luV&6~$ewna`Tn&15PnFRQ^y~3eR;y^eqmpCeWEE7IdC@;LCPBrrYi16
zA~tGa3BdBu2kjza%iHBR-S2=sR83AzO{tZBe2j`oFb7TyJYyeVE3#ocpuY!NAu=;H
z&XquZLKFn|?@g9lwSlfZkkM?mOVaQ1)FMoU5abdKmz)|jgWAYlgF!3V7tl!#lRip>
zD+b8g`a~%sXp%e!qYi+WvjazwRyH0O>oo@2B4CruQotxgBN&VT4lf=BMK2(r65#8<
z9Z>W@XSy?%*$`MCpdWI8AQqqmgC2MXV3qsF$IAe<S8Gpa3_+t-HdT<XlmEZnTY4YQ
z<~?|eZHR2e)?~pLo8^p-<3E0wE)zACmi{asUsb-(^aGuT;3XBS-~lorq3%+#MOr`T
zHyL%sKf_R_B~N_yx8Ee6n9r6P@x)+)0yMO2fhLkY-3O-*aHqs{&>!^VArHyO<jO6l
zc|UCt!JuH`s95k*mShNMpfm#)3ivA^hA>1C9=Y2Y8<;-Os{Rzx&>#u{*&zvx4ome}
z_-jIL%PI5*kUxCjJD{Cb-s%7xZF6k)sVWD?%hTO&fTTl!Uh@Ic5M^#@=@n3iNNZD-
zoV_BqZn4LAVtmT3{<>9x)xCWzEvZOFA~Pk~d7Y*u{Mp>5>uyuOM)BMp@`cq}Sk$LX
zDFfVCY#smNVon5VfOy9}4*oR0-$xBgIKCRIN-CZK+0mWIn_UXF8iZs!=*!tSIKY9g
z2gu$0!h$#q_N5dwt&)$=d1l?3PPQl2t3NRVPA@DKN;;q7IFkVmIZvbX%lGf!&5wPT
zyOTyQrvCo@TeSt2ju^hVG|lI9duf`Fu$@5}t2K33QBjHAH_2Zg$tR)kyt0Y?u#C-&
z09>g|{HU=BE55XDmPVsm$Z1Uw*4k!QgBYf3chEyJLb-aYi(gA|G-s9P^EEvQ<JCrk
zC8lq{5sHof{;-Wd<}SIT)FA<kd0JuNRANpewbAYA>Nt?HkHDdU@d5ByG+>rP<n{{R
z3y%i22O5@k=La7E<pa}7LJ%;Jk}T|=2n6K*`nNRO?THr#1_uB2K)q~k`={rYmnAa9
z{SfVc$b%<LlpwwVYg$-dZs|!8K%yFdJq(tUBcJkdyZ-{{mP+8|%N>J;bnM*RoMtE?
zDQOpQsS$%jNENorUyfaX6{!~04frW=u6?R)L+{|y@j3gWC+@cQXFt<98NEZpS>0dE
zDrZU-UhO!;IC&lUDP=+eFhr(HXX5Tc?$yy#51NBjX`p9YfbmBo)8QSEBlVX&S{ZR5
zz%%CnIStYs6k<W(I-~>7ti*gA5A*^N8has#4g&Po2ha{iOim!$>7#UXbc2<4#+g}J
zKz@Ui#(cumUZmf&2964Wi8hO^j?5=Ouz=pZu7ROpuG+8Xb}L;-dOgW}ASr!7C7K;_
z9V#|f$jr>#J3f{$llx`RqE_p2Ea#9a;D|#Qr>ScT0&dv{R~Td`E!cuN(9HxGO0QZ#
z3I-BL`D}eBO3fC*rULVNXMztCQLmPwTaY=w(QwZD5Xno#&C9H2b!}9~$uC~FZyMD{
zu*yJXYU;P$g~#@h;8C!}+VLbSk1Kz$DPYnH3tZf=d~8P*W@fQVw1fG7>w|tzt$uOT
z^Kbf3%0bYCRw1=3D@-_XIs0*$prBx2P*7$5BHZFOFfh;@TaMr~7ECq%HxVNRMjn^n
zRbL$w$|j@h>(b&|*yKE$B%09U>1E#aeiLNic<sbzbFd-9pTE_N`^_W0XPlo@-S!@+
zU#(#j3Se4e+hg-{R;_BZ@@46Whk8sDIh<>!wN~c2JIZ~mqW=gP-@#@f+b1R_t?x4O
z1A@ID{huUy-v;aKHCnp{v}HDZb+TwO%hrO;2L;lNReQ4b3STmTIe|ExYys+^NNbbw
z$)r2>Wm8?^zdad3;di}4(!LrizKg8OzINc5JH!s5W%>MEV(Y1)y%yIs&yjJz!q@NC
zdC05p(iXK+?SMxb0O-+pE#bNt<^LGm5mm;^(=t7x+@5~W^2Jn==PGsd)2D0`F*-F)
zY+LIGVwUIgN4WM{;}IgaXHimA<jYI*QG%yacCu;0)BJzqD(rBt@*HtHgw&oJXHO$Q
zJVJZ4uxaTpu>MM{a7PQo#x#|#)Q7^9CSTG|R1%Z&8Jto>p)^6KWwla1>ezvyfwa9x
zg0F6gyFnKQWN$6CjN>@Ubk@?!N-L1oQ044EdlBJRB_mpOZp1u1Jgx!H5dedBayrXq
ziU-t}RweTB#-CLRy1orj`kpioOq$n?7sq`ym9;NvHZ%xQR6iMwW@r;kDhGd!AViGk
zfSMk44;lxCJe9$?6)zYxC1lm9S#1mb`ZWr7;}f_sEe{(aSaA~&_HB2-%K#46)63nx
z^Hnx0a%yTKGKrS~jc<CYK?^-FnCr&lhL3>BNx`7EYd|-4z}3PE?-R{j)o@I+TToY;
zNE$$t*T$kOF6sC*AX*qyzU##qO1*le`12ieHL2zui=5i<yyc`dQ7gWGu;gtyt+%K5
z{a_87=mlnLh|d6Ae~`B5yjI%vSF<KjkqzXqyr<1Bif+Gc5iywp9wrM&#S@%e?^98_
zf3v^~&urHolFP_6)Qdz!C1aV5KE)Wo6%n0p+?%~^^g>e1yjNsKShAsS&w0FKLBOQg
zVEHY4qm5JK;NT!GXoA4+b;X08^AtGYI(gK8p0l8}<&@X?VMxDSTd3IiK`U<S$r{qm
zj;J{k)&;BEKC0jaz4<uupFbx!NlVf1TenD?zAtB4tFL|z9T1#6PzPg`tyE!x2D^LY
z4F7r1foPJMpH{9<Lu8(^=-d+#Q7gj&a|2q+BIlN~mhWZh<QvQi|HQM;t(%TkAjwOl
zWgdR%*-|S*Sw%NKs9rTbX&krT8dGwK*b)Wr^&tYnX7+1{-y#ppjwOT5Rq~HrFEKvX
zz5AbEv;_uvm(OslAQ!4VYHSK2rX+1pm8#2J-AdKEz2HywU&&9zfAWO3R{9gjF>oCq
z6$(>?9^`8_jr(R2I97I@27FP@*GLqf^~4o&6;~ms*}w00mce(wx6T01fT+j=a@e?%
z(fC*=RYz8x%P&sXLmjM2_J6MGVqeN+uyu{;N9-*%>M(GV7%!}lzM}eL`xLP`UD+6p
zY5Uwic`(*GQ|#(&v+*@K(Gjypv~z?x;R$yLXq`<6a3S%I3qbCk@tI{?rbKHWKM#+g
zWOZi2?Ua_@q0x68vK%=dAQfK*kUp1z$;+%X1>$0?J!;5RkXNG2fq)@JqvIdLk<@$-
z*Z&){JB^7ZF#NL1ZW#Q^^xT&I<6zJ$`;rKv?nG1tL47cNQJ6ZtQ0l;c!PTqaoa(N-
z<5S#M^TEzYJ`IWyiPG`^F-1|l&1_en1%qDRF5G!rzC{9GQ76h?L}u&M2Fp`TQbQ?C
z<NCjH0cA>ogNDm58+kN5U-<pN_VGz=HE!4@ITK*h4rU`VkxS{e`BE~_Wh_~wRHPn=
zYmBm4=}7$qfo5JQ4rgaMaw}4Imv1hLA~rp$Ie6=0M1yYiUCmwTZTtnKPrioO`|SNB
zX_Wh8T_%p}cvJ>b{Y*Aw1m7_k-+9&7YcI+@j(YgAe0^Gt&xj`3kCr)+&2z;K7EQ0e
z+xLpr)^06D#9GVj8qG`jBed^n5iLx^Y|=Y>$BR3y{C@1m{6+65Sni^cyU3Eke>WAg
zn8jhOfyl@d*-!}{23G=UXw>TuCdgv#V=4e?5wM0FtdyG7<a!7mQ+1<ptseMyF4^_A
z1-Ednjfr^>u6u+hz1)SNd;_nb#cx_>MsI+3K#|-qYf_`!@91%}k5wxc;KRUSQ{Tyx
z8Bh=*j^aGpKpp-ewms1-bP)KGS_aDR$3s{rQ+sh-QM1E=on|UM^oyRqc1k8P`_gwn
z!&rOc6T`b7WPK`T(}P+&$Y*Bxw^3r}p+@JHuv};ZuEwU{N|}1Y{ruxET$e}pT~MOw
zl`>C&xUgH+vXlbyjGUbO1n1>g_`gX~MBTjWxJFIICFh+Tbi(qq)cyG#3fMjl6*@Ul
z&y$k8KIgNeFSex#N=;2(n~ZxPm$blX@(|F<D?!sp?L}b<*VI36(CjQ1`LqrzIrc)l
z1qlz5wGDsMNx2q+R)9(U$OP+8+e*Hp9W2EWqe`qjRodk^<TrM|3rAe|*~l7tdlk_5
z+F2P<faO0Yy^n5+FZF-2dj4vw7K&e<>`gneSyQ^5sOxJbr*Y&|knle2N%5Jf{d?Fi
zC8%D(rhrVq)DE2XtqB(0yz9?Am-R9S;1y)F(Ul~LS;3>z=hKC|F;k>2<%CXJ(6?gH
z7&Jp$|I#wlXyvd;CT_IQgx*nPvS?*z@CKJxo3~IU(a$?Dv3@8OCN2tj^{SyOfz3m(
z;uTqdahhM8&gX@e5DIzB2R@B{+hi<|{$KrPj{C-QhbT&;1o%esdb6?axbBhmBAun=
z;Ejt&<&Tg(<~gjc5e*6@SS{Mtphh9;eR{aK*Q+^(zV4}0(6?m_4x)(ky5U&78^{8o
za%e<FXzSpYuQ|WYZ^L$he|BaFCIjovkab+KwMbW2MeA{H6IrIK9(wYpGI+lotFmW=
zzwX#RKCRYr_cP7sfshR3k$gD33Xwgzet5=qn)z>lJ5=1vuw(a+Plz+|NP*u$y_Nsd
z>-jUAwLT<Jt)$3*Ieve)kFj5Ju0X=wJ>mOCm`+lTyk+esYnv>#;x#GVW5T-(#fAIc
zTbrFFG8(OV!F0C7Dj0z6TUhVk5fefZBQCy0wjx)yGXtB9n=}aetyraFTE;RCi?MyX
z1upW{_K^G^=0D?b*w192J_TEA`wEjWjXdm|?qjuboEc9F9EwlhsiExb5j`jZ_yd&p
zIwc;s3$Hh=Yk#AWae>JEH^E%=74%Er<p*PIJ*mAEAX6>n#LEK5^EWEPK*BY^DT?Cs
zR{HA+bA|l^9_lsbZ~giE8*g)scZ*)JMEth|o8Elw2gjQ)uR;jQc&@`kvwk+jFcul0
zx*c|J)5tt}M4HGpU(l8qA1{5Ye~s~-qyc^yXL2mh*mG;G6Pb;c{NH@sRoXYv4e0-v
z1*omY8T_F!%X>4;iS;bRS2_#*{xdkalBl9MDpV|+M8=f*zJJd%cy{(n<#4?F5)xJL
zY5}AB@f!6aPgbuKgOFB2yf5vvb!2OHXL^&}6)jUsQ)w&>v5vMMr0^3Te5h*w{8K`v
z{5@-lDay^woCI1FviWpbi~^%)NAC5vz*{4?oZtlU7Uai5+OZYh{4mhwd%p0b+*0g#
z19;zx=jUnX%O;WowGpIcmVQVmH4Na5wt&F`JAs4c-Zn5R+<xoe5qj&Q^=zFuGc*#g
z?ZAqu1kSzjZxQBWTdx(qJ(ztJ(wMNvJd}qaJag=L3WWN%E5c>It5JP7+w(pvJ&_3l
zE1yN)R2I1@X)zg)*RQlb4oWr|(^%Oaf6*r&rdMTsd^C>ME3&&ja6=2G%DhsMj@G9G
zzMm<*w~SAq5sKZ@_%SM|o#^EC`QQD!^7M^!j~#54rL^bDRVN}{ZC$~Xr9#{>D=`1F
zV5JOOleM_EVv8&x7*|Z;PBsh9Rb2ZIcQ@?7tY2XJn`av<7FAEq{TP^aDC>(8LWE+Y
zTu<!_AIN~Y@3YIxC{|D56S7IE0S%e+^?<aqvW0R?nO<IpXvET~B~e*5<U)Qt>cTw_
zoKPQO`#8+l+<8HPxH-b+)(H8yHC~El`zJt7`29xQ8<`Zg(r>TvS=0Y*U32i@H@*|o
zH<&SUe;UILuq-R$_5;4)v6}HKsGjhE`y&Rlkx?HMosQ`&sDv~H+EA(Kcvgx>UWNTT
zxr6u5pX)72du*OjEnCSizpKe+N+6T@%fTn8Jzfun()_?155LVuS=zDrek|+Bhw#-z
zTwp82(dtN&(ST;!BWJhG(U;!HvL7pBGRtElso&U+ScNG3<**Ub9-qA`|Ef}?Cw7uL
z_CQzkP`G=1!uF?l<W$suw_an>^$#TrA{9tI36f+Blq9%pU~?^d`Gp{rWoMV$YDN}}
z5{`o*Z(*y8@T=k_*&726cyY{$z3D;=!N1MB2=)fqXrWo49lG?jOW=Hc?I)K7rVpCH
z*|~F7!>!}xi3MUfA2A3LMUA84hX>j)zqVctP{S7251Kon^g-YNrH{5IP(w(XsvJn&
zit*BDZ&H2e7C2qFxXjnq7j7&8haM`>P6CeFBN0!x=OSO;n<Qj#F^*xi*$8T<HX~!+
zWx#>npFe%I0{8BV|9+omSseXk@$~{P4x_4g#Pd{^J_hA%^f*@C6<}h*b?ZHv6`oN1
z5Fc?Y=E!QqvW5R$UdN+(qNIoi_A|nj^ZrV2z!T=VC7JK4VOF{r+0JdR-uL2sC44!O
zm&$T*I}IEo6f~T^YDcTc`lw{aUEPeomvTgV^ANl<nmv1GyiSvi6hvBGP0ly%{iRdP
zK+cvHJDZq2@3113o1n>5fu1em?qbM*seFXvg*ark#=W2O0akpvPe#?d0tcuFBA0%B
z&!R@8_U-P81?W7w9DnAu0n&rzt%W3<b)m=zyTlBw*43x5@Y{pZ9brP~%+LJ_?G^5p
zS^A#r-%;QVzx;2VQEIaP5;PA0X<^;x>&pyf&}Mz#TWariK~a6MbGe)0;-Eli>gVUz
z5BgMy$)N>;AQ?5rXBoue&NioCuZC9X+(?;|Bk8iz5z9Hq6R^9lQSML7IR}rM`<9me
zUM7~skKyfH<timT`iyhipq~Z;Fn}juK3kVk;x}NsGR;ZQp%b_(d>zVJ-9t|;6Wn+E
z6_I+WZy!BkbK_lVM><>)E4KuyC+C&SeVh`_XdMq?Zp)+4Z~}62c`#23=G<1>|57#o
zd=j<^FW}6)LLV8?czqH*E~K3)C^4jjI5+1f7Gsc*r+-HB_xh#S4DP+}Y~7yhD=HIu
zQ)VC$&xL_Udu0Es$Zl00(c2%%*OWPb2V;3sCmobbLv;9;h=ZncY5iacnthkn?ZrzS
zk4eI3JV*cu886l!=V*@&0h2c-N+gk04nFF|f)tTpGuDQG5`wm^;cfqQ^Z9d}h~?)A
zM=jJI7YB5H)HcNb@Ff+sb<7mYVzP0$v+p-t?xjp5r0kbU8Ix8GI(RqGGV<H@Ol=ad
zi+-67(nc;Tv<_#gLeuI6sa}UI`MqLwsuFbW_?IPu+n;v*TSMkYb+?VOl(HxtcW&J;
z11QoZv85xBNAi7YoQ2dF%f5nUIVIv{CM#rJH{sJ~I^IrQ%tCRPg>U!FZ4SEn1_Tf>
z{_U+aI*|X)2V+HCmdSGBj^dH*4<G7POI6yAXJDt|$bjxHTroo&(d=+4Y*{qAPExte
z<abTnQj*3iSC&G6{*tL77LyiwcT!UjzT!8WqfIpEo10ylj10x<vgu5ftoy;lDEo`9
z1xxMi(#R+%>%do)*Fs|u|9wksc=D|ndTuMz`<S+u2^9fj5#RoT3F<sz&_m+0n<$Q$
zaX+<U<#Qg`k(#J_d~q7UZP8#OsKQ$0dFuh*XK4Xurr2j5JnQ0SxkR@<U*VREa9$k)
z#O1jDyWF4cNianRM(Tux*{svI>Cexvw?6=+thD{~b|zEeB#&AF@vhTYkv<Tr5#M>0
z6v+o3%eVRd#&=>JbWLTHz4z}zV^+r>&o&sx%6lz&F|mA))^Cr0L*Jx(xJ$?3fVrhn
z-BZzgM^ZtjaDImk#EcK#?y%D<H&U@nyzPC~Ed}KId*E72IM{X%!BYn1_oVu<*83W?
zPzsX9vXmCr-ad*1BaSc8stMOq9OrsPga{*npNDwC5cEe4u2+uth-(J69&@|Zde^mF
zGEfJND&B2Aby-kw300kd!)ML%bTic?U%V@U_fr-6#X}1%pyh`}aZ@<#WNCM{rw=+M
zO~B{d@7zKAklscoWO8fidgUgh65Z}je}}%!v?M&@S&~H`9uTnNP)Rj17ET?)!MJ|{
zN0B3IFp<9Ms<VN~OBEFY#MqzQB)6}GpBQobrKt{WbV`F+u8y6j*z^AMJ8-}w(W*2M
zkj|27dcX<@HE?83<%G>f#r-fJK1S_~n;dpdMT`n>jO2IzA8dVPSe0M&rG%iQlypi;
zmo$igbR!_$Atl}2(jg!Xf{1j-r5h;;>ArM#$DHf$Kl98p&wSz&_rCYM=j^lh+H0*1
zW1S3bL@b&6pwWx?nLSNg*@f<Y%N5TR{THyh(g+b$3uXxy@mXi30l_4ey5D<=!-a{<
z5}vi$BzfL+V57(5kgbOe(5}GY=`QKhK4QQvRZ6ip?TO=dSHHxkJ_-Bpj%xawDdf^T
zKU;HjF;~l-NH^lGPlg3oqBY|!1?u-#iSaa!#?7vP0h$ha2bG}H?^Stoq|?3mte8Y<
zfoFHs?WB!1PV&DWxN7t6UA82qCyUBz7a$o-6q%1mHpVb&?8!8JD6-bvSt9pZXsR3{
z=nQ&}V=-F!l&90Kqxa}InVfS6s+=bqKiu3MCP?{jXJOe@TEyw<x$Y56{4&K+{H2EJ
zHL%$1Vgb7Fpb(iTgeJ*6ZmjhuOq6IO;|*vmo482n(eN$#Tr+32J~T8+VKe?07Tf~A
zOWM5EY9d&4L-4tN3TQ+*cG@i@uQfvI?e`n1f>9b;CLn~ySAcVEIorfQS@$j*SFKPr
z*f2jd?xZ2{fcZT&(stWNusg97LA@jFy<9V8Jx0jZ7X9EOCB=ahg62P#PpL{Pz&*b}
zxqY@hp7@Z*1xOH7yzXp;+689EvE6cIyt<UqMneAuJVDbTB7tJVLp^%^McOPhLb99M
za<<#rW9UlNh{11Msl-0?^F_l*!`W&}F59t`*fdHdY$7h(w~sM530yuZmv~gG@BX{e
z4_pq?NiT(*Uz_y9%O*`vDsiQY*o{bP)nV|vh8!%A^Fk?mG63q|c)nYC)|(#-$H6tw
zw$_-ykedJV_R9-<gye@>2a<w3DDN#vFW(lgLa3)#>sKhRCrR(j3h(epG}NSV*E22}
zY7IFZq0*RS03m(#WnA3ReqF4Pz+kFr!tlYuD2B9v%iH#Va?1qL`-ShJw;)u4+t2Li
zcTrin(~L?wHU_g~ok`iFUNE)$>xSa*YrN5{Y}i*Xc%NH4k(kG^n~mtQ&Y`UF^aSxo
zj}vTRs<oUpr}d=VlH*dqzK=oTOYG*(hv{18o5Pvv=R`cUtuGmoUP;cj;mV`pcO>h-
z8_RPU1{q$A@})7q?}MN-(S%`t2Td%7p6vyPDzV&p$(db4*)uF2$Mq#qTto~aV+>;B
z4!>`2XCV3pkspmCpEsXwdr&l^?!w-&eQS|Mjx^333cveUZGd`dqzib*9+$PAaeAIZ
z7&f2HrC{=P)=jr>G$JhRn_}dRat9r@j|(W$ClikAWcf~Q;Sgbw=Ym_VYP!n>5aF0Z
z-eFn__${l*k1tMXVCi^$7(3v;R@<WEv%VYu;ZmYAwuLEB?>?T%Dd5eoR)VtRjr<(D
zzI+3s)8rP%kG@;(wrJ#rrDu1{&q*{yVV;_P!0@)Klf*dEie-=Ys~>hxiifLk83ry_
z4;Q4@st8WrzoRE*&b8$tmwlre%KOp(?^|}`!5rVFGPplaC;)}y+H5#|J3edO9uXNi
zp)Uh!zLs+)*OZVMaXHF)tl+`2NTWEA1W4PJAUt~dJc5pJ-y0|T{yN)rk7$y;#~Z5v
zefRaa1N&nmKs#~|Z1zl0DqA%Cr7jc=8MXmgDDyq?B^?(v?h;kYM2ULx2&Z>+8FV)j
zKGV;q%#Ga6lNRuu7OW%tAIqm2D!)5n*L${@dMuxb=&$iA*k|Lo22(#V9^4jaC37L6
zd*+67rLDx#6Zy0Uer+Uc370{I+RQhA?gcJx5YOLN9<+-xBi*-K-`wbg@L3a>2!a0Q
z{*v=KIuIlqF29vZdGS7mR=U|^?b5s2ufqQOLX6Ws5tHT<(*d)ml|7o*mq46O^NFW;
zXqEh>m{#8uEjedPs%Y(IRm0sPyF{x-lcKdZZRz~?YOKBuUWJ#v_N(w8{95J997#6i
z4bYT&2pk_u%jjH5=oT|#RN)WdT`~P(_t-H@1pb`7xcAMpv%~F$XlKY#7^UTgoZF-K
z(P`k4Z)l|{`9o$3o9}~HWm9}`_bEP3otw`4LE`q1VPc~XQrO|rET}R)DbMyC8rR$O
z+pLWTZF}wENqQL^H%$@z#M{z+BjJ4Urq_6fT@27&V&eZ*iEdfOy?c2;*KUn(xvvE>
zI+X&Sya92E8Os%YFT+r7>-lGZsqteCqTApQT$z;&Ypl*6LE0fgi|fntfn4s8=fcs%
zm=@T_94tbgCkd)>HE&|doSS!{tC#Lehkh<7u-$w4<x7CiJq1OZ&te(5*L{u6Q=!~t
zgP(c}ELBC?Q)D%`Jho_4o|vdFS+d_R5Y5*+1YUJRIe#6}f7&Pshn5a0QRxll|GUE0
z3UTnB;<LtIEsJjbX?IBgfmH4r*{(xx^o$v(XD&E~(x4|f=-`d!zwabs%4{Z3=?wK2
zHu;oLn5S@Mq01Im@a^9ZmvRaUYFq1nGep8`pe-DM-$|DxctJB&5_qys$%F>7Z)_T)
z^2J-tE9kple@ovVts7v!l<n(k$0<xKir=+k0`<sBA^2&1f$75N8T_I>kr&I;CjHBQ
zH5%ex_2LBQJ;zxN#N1ndxEnvG1e5|RFO_v1;oCFLfDD7*XRMv!-E_F|;s;{?{VKg$
zo4{0VN=Gw+{!%HdIry7XMS-6y>{GoG;%|%8U079CIxZY$s(LNfA1goizZ)}`9f_Zj
zmC|Bj5J9@{>zG89&ZKCt0#6lTxecq5`g|oM+7(6Gm)pxdYBG$b>MIs_&&FgGy{@gM
zvSsizYX|il{!oNLJ4B*jh1YBJU*%34DpJq0FmKY2<$Dt7OD{<t7*<REyX@bI8UDt7
zhE}~vo-b=&F#Wd-v)}l~uEVk)=`0X0f{YZMsH@NP(CT6ATCMCuMl3329(_^(-_r0(
zpoZ}0eR+vYs(*LS$sWJw%lFYi%HNmAX!_Fd0faB}mCS1J^1fK>DUfFg3tvDxf{pja
z9<fPC<noceYQ8Zh6vX5XnKT&GFb*T9YuUAZeS)h3D=Fj5M}{+MJ-KX%f<oui6^qo5
z7CqVd)8+^~X;mU##=Z(AjZdbLjnsY6-(XIYJWu`mT!>uemr<c6n;G1r`?g3aB&SbR
zu9F0obI%w=E|8p!O2#tc52imq-<ipGL4DbFd)R9sg@Oqhkg0EghG*q+Nf$ppATaRy
z=Z>Tr{RjU)VM;rgWg~^_1>i+y`u(f^Pn;mms6P~HNb`g|DQvX17Kn-71mQC1I1I=p
zXB$a8L#Z?$1C}H$Ei4UqB22gbMBG*{JcJXgyT?wTSLFR^3}sffv?pl)isrb2@#|BU
z!_mID&nxh2{Rp5J^ObR~(<tk0oQ)WG#|8^REp?uCy8W&@i5Zy6C9^mSJ1qfX0Swx$
zOhx9UJ<=P0tW7vWxX1!xzCRVCV*f7(%w+Mrm_8UyWcYkrD}-D7fzcB~MUJ^-AR%<Q
zNaOK*9dKEHeb`q!uCZDvsWaMedE~O1Vfr-nYiP)kM_QjrXn>X(dVW?0*)XFTp}+qc
z2p*)%r>E`rSEXOLO%IK_pEMwE?#`1lvX!bWA`e6wr}mN2$}{q;XAGe4r-Jf`DS7Dc
zBn*EgmUfiP)maEMGWTkbX-Hw)!~M&1s4p75iGL{8k>+IxP|6c+jUoV&S|Y=ON=J$I
zg<Kwk>fW=XaGaHfO``{4)ntWypM{%we$R-(ZjLIi69+x%OPRqXDe))&eb$>sw^5{A
zxo_DDq_;KTzoyx@=4WguxA#8I6{m7Mg@@mxXimY9o)RlPwf*Z__#o}vVH-T^Zf%ZH
zCRiLxy=t@rOXa2ZDqL9st1W?UPF<l;#pO}C!iYsz%at`Jrv;b(Niu8ad;=)z2A7Oi
z<HNarO@)<L88+$MnW}rW1~6Iby1hsUyNmfR*76t|3=t6s5+pJCyB&iP(_#i2<Olof
z^DQ=Xv!wiqVQ1SbbxV71n^T;wdrW1x>AOQd5+BL0%3j(pY-^QS4K9j{ll=z{HmPc+
z=nlMbW3yhuC{zJ;{sr#TMB##-f?K_f$y3C(aGY3rWkh10n>bqe!A0}4)zNB;>40)h
zLlf4_3_MkhSa$iIl;-d-!NmU_EqNf>aF75{SclUqbiW_A1cHL<>KJ^UNnQs2XbCZi
zuyw%<jUvX2cd1<cMtj1WWmyF#f$`QERd`F{EoX#Fp%~8mUDM|hYoit^uub6>ih!Ig
zjnfwk2aFLMHh{3tV2b6Z_S8Nl!)p}KA>vsqXS|Q6uGpl$KMqBgVQ--MwAGQ5=gCc~
zZEmw(lUrtQ{m33njfJWDuQU_vCAeD0hcww98-d&K)8D;^@9BI(>SLfV%SYXtHYC<^
z+<e#Iu#D5`+wFC;V=2S+y1kSS5|%)Zfk20r=e$43?umot&h#%1t@^C2)ar7A#hWqe
zb$2~L)AxSL+O!-OAL)<BX#IA!9!t+>CR1BGALZ(%=d2m+e)F<G^;AJT^9!|kMLJVf
zOL}$Tc+Xyx_*%t(<-#rK$3uJC%MD$ebfQUIL{!dqVk!K4e``S|H-Lo=h4Fg=kwuoe
z2xx!^18K8YqKf`kqr>#(E@^_$wqyV5;&F}*J8CLL@qd?r&fy<Qz|N$V^NW4qc`!O%
zp;T!ySZ3~txH~n6G4|`j(P{*nL*4=g9;@Eq(g1<jWy8Yu=r3ety6=7WqP9u$`aP(Y
zg5#n%amcFwu(v6;9x@ItH#R0sgE`I*MIBT6Vh8hN%r?8~#IZ?C6{@|+m?7_@bfzX|
zeGNtOQ<GMYFhh@JmnE04-lk4j9;Bv)OGPUE3%MzHScV1UXk^wrw|P>A1=B!PW+?<g
zHovqzx(p@vDNoq-xCFcv3gui$h=S4!#HTh0Uj&bw<CNpPICqJWs>%KM4Be`>(wH(@
z67dDAF=fSb761KYqn;Y8i>B{}oy%ADeM#so_mj34(kXww=acdpjEJVPcZN{;`p!+u
zq)s?r&fA+IXH{PX;hZWgC%474Ly^bxTyWA;(ivH(SpI|46gT>?h@)F>F#-Z}q_7Cs
z;e{Wjzu(3@i6#ps6EZk4fgJwS6|k6lnd!Uee2Er;_I)VwP2`ms;zkgqew-ImW=l#n
z*};$SzVH9V;oj3;au@f7CigclJ`@_wak-y*G;%0pR1WO)L{;T;TOpD1zfI*TZ9dGR
zd&tu%a!CL98L8FH2kqMSS=26<I=(dZ%m1#dt%$JlKtd$+%2&{Mnkwd0uvEGL(eAxI
z{Zm(_dyG#A?jF$sg;?Qct3yxDJKCWF?oEj8QagK}KCcLB0V_-k=(Hat(PjGYUS3-%
z)vx>B9Ewr1JEQ0EJOndTtA@-QI{;Ou1q}gt_QZA}AeDPx4Z`$M$M-P%m8jY$jqnIm
z*453;-q;E;3mrvW>bPlze=p$R@X8+YXfe-AoV4hP36HQcuGl^`HU8KBzLDAd$e`Z7
zv-IKgGIVc7*K>&~;mhak-!u;7zlW+b$Cv4G>2>_vRX5`n|L!wl*`xle!>{YmiLGx^
z1f~G%e-ngEfB4Lj-RRgC0<5{<MasX{Cu?*K<yriE-yFrR^Q@(9G3&C@s+(o?rkm!o
z6Gas9&j{)hu5cG62NJp*cG|A5-5{cbAxpgP$d!0QQKF439Y6hfqL|QRaBF+#38dKe
z%wlW2_dxQsE)hIDg}(6Ryl-bAIv2@QtPsvc+jIIOJ-7{p4B7Xr*GrP(M#Ex(@lf7J
zxE2I#I6~6iesnIb9<TMCRgemFp$7ySu8I?IAYzb*41U}1b>4Zq+Lc*~;aAQEk@#h<
zC@@;8$u4klo8da|`1ndA`p+jH#+Q*gPqf>4a#$@z6bG_6AJy=s*BJ0JI?>_%vD;`s
zS>m$nt?`lu)TOq-2Ld33gdA^AH-L!RrEsA=@K0{AMo?bZv$T~;^U$6sn-J}0BJ~mn
zMX5%IQ<RdVTB^C^8O=W?aKdxm7GuTF@fCGUWS|Tv(o~8S7R=3@J?U4yt<R`*RinWQ
z)z5DXft|WH1CSxS*C2BQ=uBqmo^Nh0&}#~$thB*Fs%<Ce>6PK<<K#Q2Itgm*aWi_I
zY2B3Bbb>m%Q8+(c(d&7if0)9jpd!Q1$aY#CJ`sTOD(GbM9VJ^0z|t-PJiMnO4;Bbx
zvo*EaK<>FuZ92TSz4OpZ0NB2bF2M^BN^=EIYt?TY2hXK?2sVH#lsc!Dz1M~PG$vuC
z$?{zI(#dfzc~Kzs30k;<(kDc;r(<w*{&;>Dbl|e>LZYN(bFW-*bOOocZ-6yUl%Pr{
zcr-RXvLpH2^k;fkn?~q<;&G%^`-LbxIz#^gVKPfcaUhua+Oa2#cb62qOVQ6)-xo%O
zZgM-GR}Bhti`qiW-|I<XJ$wj&@*cvsHvFIl#g@BKf#Tft3;@->GAfm4>kg>52zDKp
zj*F_7g$!q`#E9&{RV2`f{rz>oczX1;(@As2pO156`Lqgi-tQCluhB9)5x}|L&7>0V
z+7;i_7wz<K8of$|-Fv{K2fTv8Z!Yt$02X~56*#On-=BS@SvTww8xTO@?Loka^l1u7
zMC5Y*@nmS=>?R*yY!kiv&SV*JjX5?Mt5fo5QroVf;2GAj)+M@3G_(of4q0lQSD^^f
z@)hyuPI(?|BqIrl8E30N<hcvEjvyCt9gy*I7ZnuUCd;Q;ztyTm8_V-I>;!mf82OUW
zWh<$DnjMdi>?JwG(Wak!8l!ML@5f`_gIbv$DY<LOISy9YMRv4`&fbi)y&q&VP!IWa
zV<kOUr0|eGN+}>wLNHtso=e(ZTKW-f(l4mH&<ZZxP%Z#dkBHdG+KeG)dJYR+^7OE|
zj_Ay?+Kt3zMK42!lX>IpE&i6`0ofRB1%9L$KGvcAQnpFtN7Q0x67zWWjTB-l*oH#p
zgVhApr<JU_l~q**ilJZd7V2#GvGx{F0d0?2MLOZ#TR$5Y8tK&*A|H`sfs*7pv-_R%
zE?qo)mbZn9GVH*M4sb0uTHc9MPlgSX<tz2tBRwfYc(nRLY%Kiw;|H)@Cw2m^vj?Nq
zmM_RKOb2G~C4*7i7am>#X=IM4YYIOeHnpEZMz*t+ROY-lH%X&c7>0<e!NF`pmo^^e
zg=o^=M}m|47cxmq+yA}<Y}-_%L3rWzXq2+&I$^`zq`eH`q|67XUgnY8w=sux@H=o=
zuB9J0TNfC6?LWb195z)Ku$wtJ9RMusM2#KEVRK{Ql_Ul4bg62QKDkzHP_q7olk=t1
zDk6Z^acHOeo%Mau_vgfZsT~xp!1l;pK3%6TA%<1Ec{AoQcF*^bWnnIxSzHm;y{7Om
zSMWR@iv*mD_USKOnR~Tw%X#>sp{S9F!d;oI)uZ<55<MwGuFckru}cIvJG=_U*&<3a
z_UX8OBC+IbZ5slI%_kNjnIO?+GDR!aGUYknn-8~3=lxTF$z}0`LB*}H(Uae?CyCcA
zlDSp+)zt>^=gpw~IjlOu2778aQ8*oa-wZb^sQvAY4S=9!%k-M1GWct5zhK2_GV_Zv
z+zpqRlk{E_@S^9%^xUGrBS*1^!jbfTd!G380Y5R4-Ri*O-XnW9BgR#*>ev6~0?0mL
zjABndY;vVh!cM`8x&;_Qm#x+0v|qI?sY&0HnM!*hAFp`lt>x|C2ivO>0x5=F&z_-H
z0qEW}MDE+XMj(3#pXbIO!w<;Od%W-B|4Oq-qF*l#^#Z5K{$$oH*uxbB7)QDvp^|Lk
zENf68T-@bRzjZYp%nfZSqf#0pP(2*p{OeyA$u56*dV|&W(O6Eg*M9cNaTL4Ip;sB|
z=WzngBl1JbCtOdY0QbKI7<&otZ92mo*8nqnZJT3_1%OP_|A9;jW26Fog7#GXbv96!
z--#r*C$htdEK%XaQ;b=%pFfkBjVJ&!j_n5Lf#lWh5#_24_OX1(_9VZH-dz7rvF}&A
zYx>kEGU7**2EYJH8W^B4s23ANM_kysLpia8@a@)YM)H8KV_FpZHTgv&vP<FtQ`Uf4
z<TX}360Ua;(ZRW0zUL2)Ui)8$$uoF9iI+siJ@CeGCz~U7UvL>to~<{i1EuGpS`5I$
z7=`c;znibZ`Q8}{NENBKgNDPOoGz<Bej2O2n_mT3kAG}BKtqIXEtcQbd*HqPP^r4F
z=ehrM`Dxu>`=b+4{^FfBOaUK?<D{-GU7Y04I10e(Z2M0R0@&0}+xN}e-})ZM9gl@8
z&c6w=(Gwa|frEp9DBM${XarGu{7MF~KC}aP6^0W$3bC>6()l9Qqnq6y;|T!it~S51
z2KrUclTSMn7j4t;qxQmhS#gOz(ZyfxTP4qZWiT$7UGIl3^0Nf%<@T})5mXo=+Fx{2
zUZ<|<@Hqwp<~k>E+Nw0^UtJ6kUn|7T5Rz$K`uLIu_1X0yeiXYl*-1;(+c&-Tl}`%%
zUcdbd(O>2rPHQ9B!)`ux^|s>sGiRUvi1Fh7*%m>RmV~%?(^3-i8;biY=wnvX&0me>
zRg69;s9XUVNUwPhOM%=QQOE<p+Xp33^)BZ3)&!y>`W_DVYtDK)msAC{S*z<)`ewt4
ze7KD2gUwIF`&7Y6*%op2L9gGnLULQ?gTZ*c_Nhi<F|2Iu1XbY6pBRj12r2x7@dS25
zV~CA^6Bx4yM2{m7*FJ%N5SSJPoQuiXmW5Jf?eaD<+6f2`U<`!gcs5pgqF{#hC+x#j
z0D9@|Vdvfq07?!@Ite<97pk_d93`VOg=K!&e0%o5m-2@?2PR_%5&|qb1*Zq8S5(;E
znD=#G*f*3fXzov&W&QIXue7Eh!YmJdmkLj41rM5GYtd7x>k>6P-w?}HC{H$&HW$;Z
zWI_!jZJZ0BQ$Go-+l{*CeM|StdVTVTe!sEit?+lc9gj0{$5zl005~W5bKp`5yeAKs
zrNlD{Z+7v))uS;3nL^6@SVf^swKdkA*^(-F-MO?2r>y*Dz<C$R_ull9j9Cwj+;{K6
zuljM!2e$L>nop(Og=q-Y&6d0TaRBBUlp_`zw2`I4GyganqC#-+F{`|T6SXW1eH
zz0;Et?YVm?y*>isFrYgm6@c|<3|?zsczMlh4|M;DKrBW#nE0XG)r^(Q^QNV}G+WnX
zFeMK7vVwB|X>!!<ORR=8SoJI=$LBSzzz1&1)t;J_pElZ3Uw@B(T}WUrcgPnaae=;f
zyEX-v%r7-Kef`Tx2gcQxoew1M%k(|5fTkiXbjjrMZh^!BW3<`R73S~<&zgZmPo-E8
zibT_EpNXJ_Kp4e-4$@R4lUB!C7_)~&07=NRBTxDTUG#$QM2k?KG!(=^t9`9G(VKIP
zz7(vUm{1~a>+1tt)7+E>D_T<DgTCXU-hNLS)&gJNp$*lvLY)Q<paG);&Bp^~8S$Vk
z=R{igG=J9vFCXj_1Xu^NuWG;Rxj=*Mof{;4v#4y315#}XxG&x%`gRDQ2CW^8dzE9W
z8-_n}UsEj9nDmBmu6+B9wldV5ZtAy=^-Os%AhvT_-7qcsK6+o`(@;4`It7f$_N4rd
zhI+vnL^f<&*q4We2@WV^g$R2a>}f)LN_@Pxrm@BfE}cv-Yn&OK_CLkg%6$D@Wj=QR
zsgJoRnyvc0^lt1XFMq2vw>YCN33Tx|fi?@YJVVPMQxD-`cVbQfDk)Ij=jG;fpb;h8
zW|OQAlJ3<g&>r@=H#oBh3-jZ5UQV;4P^DGWaM6dj&Q;m?UadUF*nXNE#8`2xL#4#u
zYN-|Y&J#uJuLv*PaQA$-MZcGo$0x^f{0DG6idxg-$5Il{6dHXl^U%gY0q9!eey;vn
zrA#uQbSMHQV)wUm&Yuhom8yPZ_{GvhXxr-NJ=1AU4Yireeau8a-ZMGb9=D-CxvWd*
z7qWL6A391yyLY5-_T87aaHG%nmaX#L@$s3zeFaoC=D_r->EUMgfK^U>4UUWeg!PI!
z)I{V(B$Di4_KjTPPr%6VxL7vXVEHY;ZEP2VSX--!y(&#l!2&v;{UmMT0g1HhL}><T
z(`I=8nlkN|ZGnQ*p^>G}t2fWT&@dt~8XS*x18<qHPf=eg0dGOrjxIZE134OdG?a<0
zXAL2#zK|9i#*s#cZrfEt4n#mwY}zZ;rIf^gEntDdjHxrUXQnC!SU@?_2y(lzZdtr2
z(Wo{X_{O{@of&#L+}b<8=m@M^zXJmcFJP0<6+^zn*&QAxF6NZ92wMCFe(Yx72a3RV
z9Oqe0S@RMJ5|rwP%6#2R<!04;>GB{5=zB{(sBv*hWn@k}YYh$9zPeWtu+8B1Xf#Hr
zp!llwJ=0EIm`y<)%$ozqDrFm;wgpww!}?x@DJZ>0;c_C`BzO^^>ti0t-k_uP4ICfH
zd_7!e$c(8ApvYdIyrEb>S$v$(ZLuC8eKS~!Qmowhrg^^FW=DTGG`&DuiKoYwOtz!`
zxn*i8197jN$=S^)XenjEEPjRs4h8StQBmnp@rxEF4iteV&T-JPfk3ul+Y<<Rp&a4X
zB>JM)I0yw(=esWmEYkFNY^Oi?KRm<=LCkj=2z?UT&i&7Hr;`61DSc{DOVDPtWq0^3
zrIIhxeRT{i{801~*tBLcD2cDJOVH61CTjqzyj6vq9JWu1r03o%@IZo!paLrE(!Ca{
zS2hh`@7V%>a+hBV#CC2!GU;);CrDMds~d6X%ChnX2Tc3y>5;#;63GLX10WyGl}%=$
zKJQ8q0wvvxjO=Pe;v^CZ;n=6s$Zm%LHXu}bHn@+IVb2v{rVAjGdz*~+%-Mn4<n~Ks
zbs$kRnf$v@j5p&?5#P3WWfobU$NjDTI2ux7UKjt^ea9uc82Ma?w%lTg3TS#IfVaSm
z(BD;TtZ+(9ba1n!yZRpRD(GeB47>*Uf4F6xU8n2POvUE{#ZF}MHUOsnj4fdF%nn4(
zQD?rj$`y}e(m&|*eSkK2T<A=TLB8&5Iv70v7H?*x)FOH(cw;`xYN-gM=9-lz&w&Z^
zg^oWJ6+1mYX?`nvkF`BXujdpkc<FZ<ngl@Q1mb3Y0=NL#Y*K~V|3-i+iopdW+Vk};
zuifYCDepE#`UO=+8kQZekgqvy2Y9i2<>lH0;B-FP2V?B>zYy{90v4B`cpnNU<|&oj
z!NIZ*{uidT7S0PXWAW>kE5E=fSZZ#(UXRT}?&s6&ZkwdUDPQmqSViW~rrCe0B7*`m
zp2p_Y-T>=9o3JY?I;QCndEoY?wur67e6Dm0t%<<5C_Qcu`-6EtU;rQwMA$Hcj(Pp`
zu&@Y;g71k}r7^F{J(GMs%kzQCMv;Q5!zB_Pe)-5yAl>lHf?l~}7a^XV%@7{6+3wx;
zG(BnK>3&B?elhtS&ETZB@J*FRLHXhDiJG)lFq|$Dn-}!+qR*|(OIAHOI94(rN~)8*
z+$zaJhyV&WlkV1Wacqu1NC!WIYW%R-jos_6VhWkMeD252RA{~|jJTD;VSXXUT!c}U
zyZ^h^S^yFS&-&89EoWtroIPu#z)Jz}5rEzZ1!(5}*4tsZ?=0}p^F>#sAV(tc^d#C#
zOMss+3J?WA+8SN-yqrZa+8hCmPGgC;97sPDL(PVkCz(CntP^OL!_g{l?`&sFtp^m5
zeW@HP2fZG*FVl5=;&;f^=zuH|xZs=saWX)?sV?#9s8-=Hcj(cbwDG_f#8AEdjg$){
zMpG<q_s#;{r8;~gd6B)J9|XB<(M~sJnAJd~Uu%6(2DLu=*jT##`>F&gO~S)|A8|>y
zdtF@p*dzqZ)pG%(5RHf{ELqQ^{~m-ew?E#2@$6<L0;sT+qBk}er2;7E{@gxq{Ws1W
zm3&Q{*tQTLu3etx_6)P$om}4CT_k$f@s8lEm*VDeGON6gEfvrD{rbe$?a2yaM)hLQ
zI@C7>&f{KJE1?I>a^hXG^Ovo!qoIQ>r@IQh_AYm)<d<LRsi;N-je(0&UlzIWEgI;|
z?WoPk|I*=`49Xiw-baL>-~qIGy7K94jia5^7x3O!C&QaKR`Kov?_lJxNeykxf#v3a
z!{f~?Uk;PL7r<1xG#2DX)yT}vR`#1yiu20gxb(FWATX-Gohk*kyC(fcy<QqcNcLpc
z^lxLd8YdJgAfMZZ)4^?hcb9v)m`(8vWflq7J(ozGOiMcLtIf2%Ei$M%A_&*b%3as>
zWN?|<!NG*-)rCASs6YHtQ*<H+1z3Av(qcx#=`Q=_*H^$U7+tDJW6tmF_OSMz!<Gbq
zd-qN&G~hB&FEO}tbAzIT&fcl~<z)Hp9Sdw1R0a<8#WSH+y&k68U$3tI`6r-6yG{kT
z?*QkMp0ooNuznFjP9*^Iu@i`{0#&d<uYCteZ#cLQ8U?=Kfd1j1*&3z1XMmms{1tvG
zyR5l3dsMJc-`{br4O*rdkn*?vkUv0|Hn&aBdC0w4Bg;rTr<>y3`Yu{bUlnjIw0R1+
zKaR(%5`Y)c&)RyI3l{MNbM)hc5d0N9?-S}LJ8+%49}_{26X<h`{spg!a=Q(VKzROP
z_9s?X)b92UnY7e1E}KtPdUsbBlDp%`tKE*{_nq&y$R;w9Lt6*e_7g0R)lpskE=lzL
zdg0I-mj-TWn6dT6!GbKzP|m2G-WdzaHDVIp&UjF?*WN3NJ+4i>*WSm~UKhqF)ofi<
zsL-x}2T)MkTZ`!z*wjZ(rZMZFzl72*7Mg7;yY(0KA8-Q;4r8^_P9MjEIor*Do@vo|
z`2I=T7oB5BECWIgGMFa`8UdT%U?%^lx;S+bxK~k}N$6P0hwRfI+6XCfV8CEp!Qv13
zHx$geC=99{ZT=Sql&<+-2wZkwI&H%TB8hGcx;znE={SHqo&*k9R#l=yh7VUW&JNiw
zPaCLPQd5w=5|x8V1r8g7{UD(OuKs{(1TKv3+^VzEn~W*$r|<mG-ah>YqAFIDXaoB~
zdZjj(gqw@tSCV^%?<%;cZRVf3pUxIfW%~=Q_Ulqw_BN0vrf3DkkIP^ak^N4Wanw|E
z!5Y6&c%_&r2;<NKYZrhTy#yVmILXgT(EtzzX?E{ns)sA;)<I?^=n5BBInp`pnkj0_
zC!u*<1Oc#K8XM5hr#m_TSq<_?Z=<EzLLz+ACv++74dv2RcQstC4X2LCifqI7pBPci
zl7`a7HCwrq>NU#%?J=yz0mCj17-YoP7*4YDT~Ym-o<u|M3?h~B0r8V;AyepLIeVt+
zUv`-08cze#I7O3QjGxM#E-{}bgDA!3PhUL^eag992ZwfNOaUxLsyzzuG8T8cYR~7~
z3YG=-<{J`nB_cS0M*&#J?VA8_5vgFJ$hgL|_3!Mw1HXu&%8ncnu`QA?mW|SB#{?7i
z%lbl8*fStWdDE+s=NF8cBS$#3#4{$&VSCEkSiIfQ5bLPs)6CHvao)3;ig{i%wF%lh
z6skYufRAA^sHv^uGs60CJd~6;F0VDg7d&8)Rqu|1{&!WyTutB`o4vLWP1{4gIicBd
zIKZ95dZqJdOlaK4&5w`FO17px!Mr$h#De;ZiNgFO9k?RJ0Ppr?C{zqMm02t_az8iP
zu6~P!Wp|pLop?LkUWpOATq<Tp+D2e(beQV^QU`a*=xJ|B(3lN0Y<-vw$(twyB??YN
zBq;oBRGAFlrEsGPKhB?D-k(pGIcY6~>u$PYP0_K4cs(FwnLb9&RGF(5+Ry;gh<<#1
zFHvCU1yb(7T~ur=_mFuWfKPZWd<3zuOQknvt}}pE?uov>=h&y)3zx=r)#X2wKXr+H
z?*Fn>5OX5wxwY3^&U<-qW=uJm&9s~5$tNaRGgM-UPzisuHO-NVqq@7<Eo-{oNCP%w
zfq=Q$P)h(JkXkiHIvy6*b#emTetn%iw9P(S`4!L<jpc)v&)z`*T%Tuag8@N(j9Q?n
z{j0X`L*o&?ZeSw^09sY3r9>ZB?=4lzpHpgx!F=MbwKk!^AIKcIF#vmQWPSbnj2~fP
zMgOX~?|hCunciGxXyOv03yna~a)ir#&+jw;pjo6oW4{FL6}verg)GG}O2P^tRTnwR
z9Q}89sJEx*bNV<Ae29;BcVKb^H3vsLu$W&3W~P;z!t6@o)Nf#o!MI0_yl3vEdcihx
z5BI?Ibg#b7t~L1RGlZ-}m98&+X5X~&3wy2OL}mU$GM7iS(A&umx_O*D!+%bbGb86X
zF!agwwVD^cTm;WLZg<gW@z@LjK|NUbY~Vx&lz5<tW!3ltB-hvYuo~XIJOUpi*aEBj
zlgvN{3M%(xy%VrpJA;J8UwrPcfDVdQA*h&zp9HG1MRRjM8G}Xq{skGHLNm|q==J?@
z`mE=dkiWznfJX%!DkOm^t|i!Oz+;LMO!MIejix>!SW~yJS+@l(5DD~jUtq(3D0(+l
zdi0wyvy|yQq4B@QXHL!w0zlIaj8Pn1+U?mVMl8iN`9mI9Z-0yQ*Wp8FYq<zA4_V%g
z`QMLKX6XI#G~a<H0XCszLb1K;EmqvL46N_=&=Ves=<AzC5D>UeV(se-EcO+XUqV@Q
z{D7>W!Et53Zod+;H)H%D3^=bR`#pBsx7zwFA$oZ_(GFLwn<M(IcM{!3cr&#&#J~gC
z_)oSd%>Ak>9H-X{3)}%xur^^AzuOB8QmW%`oG)HH!FxUJm&jRAz~~tM6gAsWtlps;
z<IrlnN1-l4dL?c}ZkpfUPamEa#53~dGpvlIgTd0T5c)?wcMNi2xr@D-Zy>RRxi0}D
zg{T>LaN;W(3=G>J9seIb=%aV9eJyV^&I+$Wx33rLn0>m9>%+&j-~1BdnSd0_=G#KL
zpTta@E+{RYD_-tgWK$&=Ou79VPXO>2`2_mCPR#`V3et+Hn5a>!y2{b?_K|RDjfD;N
zb5zyO8=;W1%5}Rko`7^I?)8V)FxM8~QUSbSYTp5LB-#F>WWPO$z5(?S<!9{aE#6@&
z<H1abo(sv(r_~L;G`2*>C-uLC!Rea?mB51Q1zO&C!T#Q^-xJNo#6M{~;c1)OcT<Oz
zrY?Jz4UV-RCkt_|uDj>Yqu80|>tctC1gdU3Y)_hELk}aoM=qYuUFUY=14^VS7-BP9
zXEy@uF=C*<#lC^139on9))igOX5V1*gYt6vSGGL{fVRh0#-T(*IR+LMoae^q@2_m1
zaGCV13kZM$SA<n}wpPvOfzWpG%qgG!pa<PzVRi%b>q8|XDD+!WAT8`6A3`-oEuHD!
zWG?nGj<mirXnE^0yd>ScaSl5#pxUaonu-R{OYeM?gm{;!4q+l_t{vRmQ6EjdY7%LI
z@}~Jc%PwO#6P0^OM|Cn-5AwK1GEOAW>{Khv))n}vzvO&c)-0tID|E6w&3<d_dWZo$
zC+E+7LE|Mu-IWT|_k;;%d5)9Adzr0i>Zn@xR$X-<M3P0b_i*F8q|3a0ZY<X<kfni)
zLS&q}0If;zBadQtJiM8W=t&s|)ZnM;2;GtFp%2F`NN*ARxGcxk()b*I$oeGp0x(Q2
zS3F!$H)LL7q)TAC0Z?A47a9`q@tLXqvbHYckPZKxZ{XAYfrgKQ*D-G{BA1!O93B~I
zvU#*(%fh>{(74`NZ~yTd53A7jSnya*r{1C))FV>&(cnmwmgHL-dGaI0$8?h>-+S-q
z65gH*n>p!3dMuDheFt7t)F3N(e^9W*xG4GWyZ?WT)4tDBo7?{5quy2Q8)q5%m;%u_
zh)vH)ZhEc$Z#>-uuQwaaplK>31H4#u4;J>lS#E#~zK;n3ol#2_i-~7M+*C7GWpR6k
zhfB5lg>JO1kG7aGLZ90*!zT(x90f0R76}?I9%OH~8-FlXxr|U1>Z|-pdA;|oDpX<)
zzrm^Na{Ndq^UeSC7rG|RSAS>!57TrRSLx|bq6{AX=me5y+&%Mv3QKdu@@L8P>zA-D
zmo%Qa&s^S(LfqWGJ_I6U;64s|E~J|=H{WO)%7sSqjF6*k=PzJw4eH!p)j(n3C&&uB
zIFISw^Qk@Y_f0qAzJW0?Vy{}jA?EmxY6>uF_yIV5g8+83Iw&t(8?EsmRj)nDlT-h-
zwt{G=6HiYQMiWY81Ls;hGDOmH#M_pJ>ckoaJ@^}^4UA?3`fk5I*dQ}JifP~qI=`_C
zN#(YA1%N<eAh~lt+08^1mORw=<n%bT7?khSE?6{r9J^<Nlt<PrdPKCf`-5u&iJUw+
zvmO8=wi{yz=27hb2cljK(~Z9zxxB{^^bU0eJ2XhhbI%|I0O2Z-DZN0@9QAOPy7<%W
zteA218DzpB!Q-+t*jAzcu+v}WxxRsi^dKBd>5H*4xvt%0e}#@wBG`4<-U{38eVlFo
zxTr=_ngoMs<`7A%so757*Xlt%Q8`MKZtow<TO14ascaTLIj;W6Im)ZT3Fr(Uyx3$z
zx)b&*qAQC(FDp*JT1SLXxmy1FI2i4xD~*&7-nbW7U4gO`F!Q}{XWgd?RXrQ|V}K?_
zcDK=`=JfT(P@uzl=3Z(Yl`Xs3IDtpm?$EXOrZnUg^VAzXs8iE8kSV?W_kP%6Jf}rk
z4s1a1S@>=`U=4<^DB4_Zbz-g%E}>TWw(e7v=wpK0;M74i-{WpqLR^B)ea@uusWWtK
zGJ$$75$&R5$b@Hmq7VmoPro|nB#{Ac{lmXcT^%I6IuHQ%;R@+~)s{K3_yXT3?ptde
zg&hIS;OaA=<qnsm2$TL&bHDo#3Q)%xi21nb2YifDfj&p#%s`x2Hr;oPXgjO%SW8q?
z*q*Tb#7#vFxA>#+92F}5JMo}>B6LC{h?I{{X6YYB;GLpDT&jzj<z=Mt>WBQT+Z0Dn
z`C1z`Kg(`kDIVnLb{v!CT@MMNB~MPqd#{MFXE064Dq%9qOri64$vIW!EMGFvVGTiF
zSx*ML9gv~kOl#FtGGfrl`2o62Yee-%AJHtcZSRlZw4;SBR33^$ZTD@q(m>dU^)11k
zu0Dd@#dzqaGVraRZ$Uq7v7*CPD}Dx4SdH(>98RbwijW7A#g11m?&c#2$q%)N6>PH4
z=O>HPnla3)5b9Ps<H^T>7%A%NB>>f6L?HVbVjhE8F5+01J-@DS0He<d{r);I{@0z<
z7)UuR7UwhLfXCnk+CpzXhrq~w-ugsRab_)%iyx-2+Ap>6Xj1-9u%Ig#3u`_USufdk
z!aCa|Xh3)tXjG&g3=5(e8o=m!qgNtgx%tvyRorbmU$S+$eE`{#zm#Esjycr(MK3Gh
z4Nd(phC~w`GXoSR7pxn00)UbE907%(b&;i@NTs`09Ob2#8D>aSG@s)h;@PqR0oOCF
z>L>b3l8~NoD>@6_JW4<a?IhvuQY6Wu-B0~f`a$5|50(7)aX1%;bql%D3Cky@eVL^2
z2>bpb{Rv=Cm$__X1JO~HnW#U!OHX%EneZ(4u#SLObIbXa<(OhhRY=eXU7q~mn`+xx
zE!c)$tAM=-!P2A80%M*N^qS2?t6gAdg3aYmDJ<uhT*FZ+i$85<1A5Ki61x>KcFJgW
zNaA1EE#JE}O<@el+fEp>6sOagu*-q%xf5|%UPSiy80unn4xmSQ?r$73BI{Fy2!I5*
zy<-#%3=z1_q!TDyZ3Siyl+75^`F+1MX6#lP^`OX=Q%cn0I+2fMl>PF~|F?r*!9Feb
z701|vrO%m#W@T5j+3l6_oY3o!C!;*$f}QHc<zkP+>HfnR;H-q%o^JLXzna+9jNuk3
zZl8aX_dDau;zM)c?MK~z3rCO--opvS=J1TFLT_~NcS#rxb{cDEsxE^%ymr~ZVs6rG
zNJBr^{d6xgZ)>{QoF}T)a<EL~0Yy1>r$p`S@=&oPzD7r@3TzMv?8zSb34jgvL-@nv
zG~{sWXT-%l^icX09DL3z(!5t=`OiaW;i>3J@lbP*er`=g8`;s{F^#-ZZbM4(!xdHA
zBT_>|#LbgZWq=dc@<Wu$SRzkIc>#}TSe3SS&QC{|ws7C&nUP?^YPOMyDg4<p4rqw*
z^@D`HPlWF&A2q9=**5l@4&5FE7Gw?2T{p@mv8)_O9&Qwr`3AhDpUNS@A<JqzPu49X
z1e}~>dXV17$pK6GYKoDDqyNnXuw<dInFiKE3-+M`(*=4$b{ZNKqyk8oB4}WyN6n{p
z*QNIS<je;mD`wanF-+lOidL3eymvX)mz7OdeG&815AUCXPw&SE!E1G?ICy<Pmp7Gf
zyt&G5yvLak3|s2xXwo(Gud)8N8{K06-Rkka168Sbx-cs%HT5?bX5_ZF-8<dn!x<#D
zEu8QE;J-<@XvXUgH;YV5MRhF+Lie4j5mi@Lz2DjC$N}kHxuNaF&ySpT8k^j4aWBZ1
z5<e6@++Rw`1khzUwf~gC1G2}hvGs`&5n#awo76PYd_*EpcNkQu<e(&-Np{awSJM&|
z`<B@m7c1_r0PKYa(|s-cXaI%+vWmTN*p~pSGN4Up($)399ALG5X*onRHFJ=ktZ?6Y
z4n{)(r_AFqMpk=+CAgpP-Jy7!qgXKxlDh)*L^@o)5umf${~kOtN^2f<3TN-rJaQjD
zDhY$OC%|@Tw)Ss(`KG80ir{0D)tFK#2>^R{Gc(iv$4&aQCyz%;PAZ81g^M%6C1yjG
z=iGv9I2B4U=t+2krB!jee-MyMMLcXm0OFPXSvJFfv=e&m$vNP>yF1zNn9ryIyz!4(
zAHlQ|IuSe2^@R<f1Jhs<Qj<{dMWsStyw~SWMP~Jmt(4#fOS>unHgDc%X?f>y!m`!_
zg*WJEPRz}cQ4Y*42=L&A<-{dzBuz$<?j}tQeM7jv$+&9*j>*RzsH`xB416F$#Sk;x
zK0LvwJEE~bbiPCQ?`!=gP}~uMKq9qL?nu{kO)K|p)mEc$l^<o^0afkS+Wrno#!%Wy
z5AjMb`N~ew!ml1M=8SB_`WWm8))&OYvYKuL?HK{mwsC{!Hf5XPeRMxR{7RbI0jtV~
z1NJUM{n)s!Yz@fQV}u>_T@MdZIRT$X^ajthCBLAvEBt6V(6(sG^RQvv<=b47imK{2
zm?#VkdIX-!KTI0paUqoi;ZZmZjs1TG7q%^lzS^xV;DgjA9@zu@c{;M#Y`X+pACG#;
zOFi4XFWXD%8dc@HUjaXgG`kQ`agzznuj#HwesKK8+(HK|E5Uh??k2(uC%_f;%5Y;T
zbLAUx@(5A?s_kCm0~$P4Ibe4!;t)!YU%8z)S;W+oyH%tKrE((2C*&AQx@{2FZxHb`
zQ<51nH=Vz{TY7Bx_MK<R$-h6P_~lYr%gt+;n)#x;rtq`u(}!PY`1?h5n=`{B4j-ci
zn(nbd7i2gU+)r1vVxjXlrGV;bPVtp#8zY#M51Z)CnTvDq;&28KlzZJUfW98>$uEtO
zPsc)*W7J?0sKsIqUw0{UPg1<Lg@>r>;=$b;p--1>KhGP(!f=&f$2iUBVJ7%{3W1bd
zP!jWkTF-L3QgKRM5#w2L@RmjWse&uo-^T}JFrLhMw%UKCQ#a!9?97Q}*O|~W^<ZVw
zFRM`F2U!*WP_BNb4<dmxJ~(Ot^59r_^^Sf&dbqW9UpQHR@p6OzX(Y*ZK>S)nBCc4g
z53oT4-T;R!sK^*h-+g_jRv`e^(Q)4Y(8c`)n$~6aQ@xVNPXFS~6i-8D_LF;1=Eil1
z`i_cs-TK@Pahm}Gz$$bhzVYz^zqFN%imFLBV@XQde*pq?C7svzu5QN?vYjDWj=LqL
z-L7C%41xXSvxnkKI>*OU*{83MTFD-b#ZR$tpO2BTV_`AqfWy%xFg9jzsQ1NYZX-bX
zGuzqWp;-*SQ{N5`Gg)#H5~o=kjv{OTjHV`sqA*{~mIZiXP#_APorzihY6<W6m7}E+
z4%mbUj$|ayOIN&L4g|~4Aq|${M8l!dYM9NI#ClQgtyCDqdt9C_Q}Z=ha^^wOl+Xo1
zu|8%L5h*pV+mqgy$2J5^YUNfR?;b^xk=YX~cl-N@p#5jY)}a;ia|e1<@3_b5sHlQr
zJc+fK?~=NlA;(Zk7K~J$&=7P!Y+UiKKuiw^f2$EQU=zj&qqN=u_u%4O$IWms=T~0l
zH#<ChhdZ4aHe;v8NgrzrK+wJWEqa^?#IB7{sz8t+hpj`^DwOgy3UD{L*mpE1Z@#A0
z@oCuQAtgWDN4fvav(z?SQEnoQz5aVB$7Mepba6l_&`}SJ^TR(*d`XN*KFJb!5L5Yt
z&BFuvCMh5+^EduOvo6?bQZ2A#ZDhI%S<k&YI1hwh4OYP2oUOZbG4<K*CD(m>b1}Cc
z6skxicqe~xaGw?y#@_@)RukJQp<6?vB+bh%&*bubn<)heuc@7XPaL~!gJzT-?9^>e
z)!M)UB)5!AfIq@YPH48Js-4D0GN}D~qUATo6<cYkO{jXj#EqzQ+}FRNK}-qe2CU6b
zSQBR0>=3YWu!5|~#Fu+ox&#-O=JF!|>+~yF;B<0es*TE8{~o;uj7m9bLr|(Ktz$fu
z-JJxca>rh0qiXQ*xqk8hQj6BMb@!X});e9u#B4a2+U3pkNPqKT_Q*ENmGF<n6Cns5
zjj-+jB_Q{J^<A64yW<pm6dcSz&B^(EyZ`yHS?PnPR+Hht%lh5Yzf3v=Pn(cQ@=XV7
z=Q$iWq|f8stcJ8BZ@oym#lAQiClEK3tlMyv{G1U6Cn6p8d3zsGftG1R2pLGefN$Ze
zn6|A^rBa>6Rz+*3A|vDQGtHa_<AEd-+h%8$B_9DSO*thc0x)2C{kI|Ls5FInPR@&U
zTp*^ZS#%ILFB(`tbAs@nU%JeI)fJY7?cZUot20&n*q5v>n`BYM$lu?KhGQ?54zreM
zRs<@4rv;XtxXX4jNyS6^-+PrCta!(H6wUkM>z<*bm>x+463!S=C#$s}e(&@G3os1R
z+YB=?4>6YjV``9i{`wFI3P=!f%h|vs?^;|lnDvN`Da;!;l!pIgTdl~2=F$D=lgpH*
z{6wmKmoKs6edW1({w*V#6-T6sVSbZtE!bNAu=KbfDIIOR078qN#v^urb94Y<bD~E?
z_5{blaqAD|qMPw+Xs=|2@8kPyG<4!cvDVw2Q|~q>vRMz#_S?YisdDmH+L3w2=h3M`
zppSl?3|m^0@bJ>+%b@2k@R6Cd>Iw5(yeTQ#=Kfx103UpSU*fofnIcjO3W!>c_Er5{
zt>+Y302vvdyHw%KrG+*|LVuX?D7R?3+a)tmQK>8g98@al$}NR=loS_YN3P>S2yOzW
z-3=`t*%oO9g$>|>3OG}_noU=p6u1O3_ZtWd&gYU$@{lMns{Ri|&+dIKfpi@z1j}Gp
z2n?2ZXTA4YDFPj2dX8i;A^@YGhX31Td8_?;k{DM+JvbI5ByFor6VE*)1M8mT5lyQC
zk4kl{)>?DSraKOCiFV*+QG^fFhWm4#5RO(HeEeR(0fKdxeecil{rzkFo~v2Shks8N
z>JO>ZP-?I+C9`|r?M%ZgkO_ai$nsHKJWuIF6Uc%Q;c@(J-;$Ki6I8WnL4@bVYY}+h
zFW2j9xsG`MV3)W3j<mw`_@#%{T5SV{4<4FDfdQ3mPu#XBEq*yVpCr^If$7*u;W%_B
zU`R`RW~PFLh2pYHhn4!K((UNpPj`_8D_3)qi0iXoa}n83n4iIDA=HWw`OS#aMkV(M
zZ3LV+TL^bJiN!VAoMf^i_BO7|vR3>4k-e>mrBBS<%Jr}wkLT=4Ma7(|wN+d@(P0O<
zAQ~KGFfNkZ3^deXIo_OID;Akv2sX__Od#^3nASJ(t*a-Z^6Ezhrz?>b?BBHVKe<Cp
zKxevjr9WGZa5`>(mpA*IPe=P&@Gs43t9ohBW>zSsu-_l0dMHNyrQXlnb?s|Q;7lZ!
z&Kn2I%7A%S27d{?Ydi7Yf;U3qEdtV_EQIVh@)LJ_%lRe_Ho@wCJUAYA(R6_-wgRpD
z*G-cIj*Oof-vSHrFff9`(LOUNHg;L{Jd8Xhye1Klpms(_6ZI+rCXBknAP0cm>WI#O
zg;j1EaeIXAOPD+{p#~df4MTImNF15ZpV^{V4tg$H-<P{P>eJT*Q{CHb{<eLt^C6%0
z?-%vJ<|F7L`DwGdFJVH!d+4es7j26)_=B>O;>JPW4fmnSTleX-^UJCC8})DAY$g=f
zB>49~0Y2c>adB~cS*zR?6%|)1LaU+1u$yNKBmDkY@~09ob9-emLH{X{UKHJKDmkY-
zW&jwO!gd&JXdCdmg$?eq3R&p99Js0aWU5VHk)hR@xw#_nnzMS_D(I=_<CCQhVcMr(
zd%7<z<h&BVAKVG{V-qDUj0yIh2}Bq8K>r-~rVGbqK4ZBJ<&$It<6+-$s$eB8qQP&_
zr0Ehp^uI-rq$QpOM8id2oCD0_8YZpkFtF7)>=p&b#>clOn~9S#8p0Jt_J@(j`Hj9D
zu8hr@9gn5lRR1c~w*vvNrN>ybUn}iXWCH@e2fTk>NzHB~yCNK8aSduDdx@X>TZZg9
z_6=z(n^#u*aK?k<_2o~G)Xa>LvG{^PNg%wyZ3(yofZb83YS{PU8ODW+v(E~c+W5$2
zt_e)0VMESmwaZNHu2sC)A)J<3x)KHzZl#3Hi9hLxlyfHD{tFyb1<Rukm6Nl<Z5<?+
zLP`4<4)c|b^QGRzZgUCe;kYQ_;b3TpT*?yx*LM*w*<4uw#|v)L=SgAu4DbpG1O72!
zkAwN<f0`s5Lqzz}v&JylY;0XMc~X|kq4QpLBvA4ax&C;YCf!N^dZ+titiE9Qi9zNh
zyQAm+^x`nIPYSr4ZA}i<+`WvE0h4{e#AI+{ut5GEM<uZ}T8At#`*LM`VfWhIM_bby
zfs!)maXzi<yYUy5Mki&vWPpWGx#wk5tNhuUL11k~Q&#@G_Dsi}1{Qc`-9g|f_q?=)
zxh#X=GgD)Q|JO>*yzdd+P<zN`aH%oUFQww>)H4VXk!?qzGle|_1+V%{8HAo}pe19V
z2f6w&bfH+;fdc@pm_ms1scw+UcOzN@u1nsRijE=R6tL`i5ZXxAm?|GpneBm&!D$xV
zRwyy0P}oW*dW;}wMP(jqgBw3A24pA4&S119v`5sbdd|@a8U>3KklH#%#<OS=z7EgB
zW=UW$V>ML@o2zoxR7!`;!rzt~B!Nhs^kcY(N9u0%Q#1+cTH`KSj0i>zh!>!zip8e^
z66}^w-6&J5o{NH0lb&Z5^_yfr<E<WX_+lo@^dG@fzLJdoQ!j1ov9y%7o}XMrVF~wD
zWoL}$z)F2E+60>O1PLREF4GSQy-BwL_J&wtpV*}ZKtorlJusM795Iq~_npd2<zi{6
zlB~S^I+$T@vf35yb-P~&n`3SnBP(tBtpn3iq&%s{spvDF5RXu&Bo6WMxt`eJ;?aca
zpX$Q8B2ab3or0{r3ttm}gRAgk@ia;~0RotoYf^v&1<KA(5Z6O9V4n@F_P>E5RwZ8x
zuW@NM<NWXBsr(L4A|~_5J-*PpV`tanp6N=j3RECoD#!|XvvUG=fQkjU``lE)xy;;8
zn!dkbR4Y4ZoLO-Du5C+4W>~Tg+TaG@iD7MU@XEBDttEzbKUH#BGuU^H7Ru1+S=>BI
z>9})`M57q(s#`2A+TcY*o?S4)Y#tPWD)Ttx-(edfvYA<ZQzw^(6o@QVt|jd$^X^@T
z$Pgiz0U_2O7GHfhmAUU+S0(tFFaMpaz4^F6L;g%j@pY|m3}!$S@d9B>Su4f@WH;<J
z_*h?%SH@>y;{c_83LBiu0Vs@C0x`1{$)`pY1dTws4bA|&^Nt#<k+%@BGj?%NTg=oP
zFX10yp`FjHSnU<?zZ(0>u&BB=++hGgy1PV>E>Y<QDGBKgY3T;(5)=e!lvKJwVrUR?
z=<Wt->25e{-uL^ibH0CP{h13G_TFpl6;IsHecxd}I2gAw#oAhba>3cqbvm*+DvJb?
zlpTm5=xTTf8jz0z7$wtt|IM+7fRoJLV_3*3Movk|;%6`_;!FZcHG~8NlJNsd{6iOz
z)Q=AIO>+D&17A#7P17g%8rOO+*CQ7cc4Hh+RQD=MsECz&Xto?UiqcJ>G`%DK?g|r>
z-R=r0$OSGmgJNV5>6%R>*7G^rF-5$^ug^Zz)$y}xmB|%v>sOZb2quy0Dp!5%PZ@I|
zkfOxR^^$XLxu@ZzgaF%eyy&OCBO-O1vltDQv%iufs<bdrcQXY@qHBBLc1eUReG6au
z6WMI2`^~)60<h~NhR14vN=*%<X!7t){o1DtEP2Bvad=k0Kbo63dZDNJHjL*z2p(;m
za<v448l})h4j9l_UhqdK!u`Pn*4IG(>9*+{Kg@xRuFbvNX9=mma2E7H;lqay(z3F=
z&VQu+Z%?(`6+=f0v>(@w&9pZVY+LJ;J-W9)G%j;=5dn66;H=Ndl-U6QfLRwkBs2k~
zdU|8MTra^G`F;$J>$<G)t+zb{`aM3Kev^hygCeT+KfrQN0Fn%HV5#r7Typ{&A@5!D
z-?92ioK!Y@OTVdiSA08zT_lg|t7mdkqC`w6`PzYFc)j14U6|oNgRSmW=V^@CPcELI
zGR6k<>giN>Q4?IRCF9{Q=dtcJjVft&c_)HUSSM{jAIL0U{w*#?$HB=v%dO$T#+vQ1
zUuhp9J$e&KT3@;KDS=IEFLN7ga`_p-h3fOe8fI`1sE{aBTkt__tADN$qm)zwlsP#O
zP$6R0wn(madsuxThCn&UF8@k(Z0XLCk4H!pSg<AKo(0a>6}E&aEvIg(x8fqKUByyP
z6cB&d7V>Pb8f;H>Y|#Adxd^OppqArvh$s@33qq^|?Lr_jTdV&DW{EvkunZQi8)4}C
zca05_zx8?nzcfNr1tI;;$QbsO<E>heBFmZPysUMhYR20;B80)H=9NoWc6;sU8SSW%
zK=}jzG0<YK#!y29HKM_5_`<;Wv)X2F{v5d1zut5EHEf%cQtg27QDR4jG(vR&oMGcd
znnWB%39N=i6*U7|wMo+oF@hpwz_k$h3pA#eI*g6ca2w!=7-na5J;`E>$b;v$@3#K!
zLS7bw3+uy&+!h~w&YJ;i`P(%IWA{Y;UdD}nz$%WAy7S&h_e4n55ObBtZ6W0fsBS=H
z6{#&=Rq+7hFPH-VxpKo<7d~k;C?b~%2mFSU)!&SoeFMnBONdivc=pK=qqxbiYZ`1t
z-2j&6Sujvs-Y}akSg#B*Y!H8m_FTXdXe#?2bu$Uy-yE0dR0d})(Q#V4BR#|>xG3&1
zb}JkI63g*&B;TJ_5-O$Ka5u#RYXYQUQ!w1C6K5q#H;qKZ#t*<p^j`qKUb6A~Nn_6{
z`DUL+5*z#3ci_l-EFzK)7=7>w34ehYYeY;86CHct)8?Lv);oWK9!4ka(Z;%(-=$-0
zsn-drARIAt(8uf0#o0PnbJemM=j+b{dIIeRG&jm7+O;u#<;hac`g1;~>BwLbi`|C;
zUaVc%BgWsQ^Y4J!J~suHA;yFrki?O3b9)AqM9e{9QYZ;_?yKXc9g`7E&6a|vZ^z8c
zBV$h-r!p&+Dhygx-WJWWo{gzj*i(W^5ne0F&gjb<0I8C`kq|uKV?xVJGv4oCn=x5}
zW<6CUQd44;K;`xzYWgTMkbJs5-%y%DhK?0V=@^Pe6Nt%m;$aci4m3xU0hcrn50Bkc
zMNnIoc+YbdS(9t==t)s!lX~$zTQAzv_-)>sp5^=-(XFk$&r_YW;r*%eT9tqdRG+Uj
z6vSt!=rQ;FQ+&;Wyp(DEJ&+oMs;MR^vY@ID3oxqTVc!(UX}`dFAzx<eA1?!*bx9U4
z>R{r84l(JMBhN%>0JGj&?>xEO@rKm!cc*{$kH55jGzjTz08&A`3WNd@^K1pXgkP{+
zX|jSznE0NZ7gGo)h4gk$DGwaPKvTERe~^gY;L-RIO_aI&#E9Mul}2cY0|&JG_iOM>
zPaSCo*JjkrQ8V7BQTekjOUBVYztL1BxHsXDc=*tP;!kmTmIlkD5UAW>R)>RA+8W5P
zm;%inAR>H#Ky)M}y_Eih9$b$8Bv%3>C=ps;x@?d2_F2T)_)Mk~vz<9ZM<<_!W^KS_
z6$!w0q2>^?WjPqaIvm;{&nb!U@x>DBaW`M}F5#@X*pqSYXu$g+Ygl4mXzR;;#XKu3
zSUreirQdCeEeT*KsMXz`rRwX!d<A~TP=|C#!)Ui^XX%9Jz(eT(rf%P}X)Ca8SDE~T
zM$?KZS04RF-@%j9`}HjNEsC>iNGkc5H6C@bCj&Pfx5}dDFg=+NiT}mV^o{EQA*XqF
zaPnvXGlDF4<?JHu?K;!(%DzDvHa*+wGf0EWn}cpFDMly7_0^KC-Ebn&+1h%R2y$MN
z)3<!sufL<n_9o?S<|*On9ZYmx1%t>X!Z8ApCZNg?aL3c+$bF!nvLvApS52pT@SLn8
zEAqi`z2!zcPN)cwRG-5d0qkbE3T1~tt4XS>+wdpw(tI<2_wS`~-t2Vr6XdSfO+eP1
zALU~>J*?*nVN;wCr8`7NWn`hcyQ8!m^jNk&eB)`H`eKhgCkWZ}z$9*Eg{A1Tn^s8;
zLJTt{g$$us17zJKzDb}wLl4Zr8=(s=rZnu&dKt2*S8Q9hL7lGP@<WL3Y)~w6I1f94
z4UMZA<*}?}N`BrwBAzn-dY}o#fz*jT+4gZ4hR|uZAGv|d&0E^fT<WTE!P^I>DSU%r
z*#Q4k{j6@2-s?;k#4WW-OE%!}AFMBGA!X60(Le&{I5XNEXn8WxaR+8KNDSdha$<Z_
zSrJ9dl)aipLW%^~LA@CoyW!W<N*6{3MhQ@aFetiR8%fNS(#Z~v3DeK*TZ70wHxXdA
z_W6k?25KRu>&lD;>FFlh=VL(dijid2X|gOy8V$%Jnn%lCBh&cey~lEvkiSc8y}wHZ
z;6Xxa>V9yVK+p?`q0MwT^61T}XO<?}K_?}n@!x;u-zx>|q+m);hYSy-?^O<|0=5@$
zoMKyux0%xT^S6(3sesKy^UO~?e;Wlx3Nz<yq%&yiPDA?pYOi-m;k%!B7)bD5w2Wy0
zk<7^i(*<FE7Zg?vCv0-Sc{Ncw&UIz1ASY*n@NQ&fkAXT?99TLBs6U6s!?H6<-#slD
zrYkG)Vf-!SlDSt&0(7{-zA-A!ZHyPsR|E+qu|6Fs4YeDE*IW+F`BeJ~bZne)xq>~_
zJ2pBMbq2!@xmIdwnY->iAsD23T=Y57tUd>)(k}3>1gdenVt{q*sG#Dx-(dD;ZXvtb
z=bO3w<l}c;c;^H9r9*d#tE=Xs_cyNKC`=Ix5bcbiWh&6Xv#}pVe@oD@w##Xynk{is
zxEM|#qmbry8~gZRRE24PWOsIhhRukIv%VvJHkCXIq{KH!0o*-Imem1n#m0$l-|OMS
zR)IVzD3rR?W-=9lDVW}ME3UDN949;^q;yC<(NG8$k-=oc?(BTo-9ob=|KDNck3@$a
ziKuGt-U$Lph$n)AU%ymYx@;~zf__B(?EhL2i4<+<^2Y#{gK$dA>9rbbiAL?V$*=c>
zQo}Pf4h0a3XHLIHYfCX@uqh=?76gNUog_}cxnqtiKJD0vWFt^kh^~RA2fYoXAUD=V
zd)dDHIQ)Z7F5|q*9#KNab*-KAvNasbbMLh(9tcU>_6+663<wFLwccH^fm7gBOvXT2
zonZ=INU1H42sX8lr@37GBZJW;iXIXUOdv#dU}HD&v?0^qC4H+ceRwtx8-(~pja?o}
zr>!i0o6?E+WQh8<>JhyfKQ(Ht>mPR2v%y9IXJb`}!$iKI6N~O6=403AOKu{eCo%3W
zG~kVS0z&Bh*8TS3qqV)3XzJy%#w=qokfLimB{l)1+!jEAv`OgJ@WpLy_w8^!UfqKk
zciYa->r$JG6=D{WY!x0HeHmY$>sQ;vU(+BVIdBbsex$3o`FCPmiyLhJ%iQ@H*fSq(
zKYupFQza(#<=p6)&Z4Mu4XI{O_6-GTluTqalRX32J`n1H`4@Y;0ypIWb}*~vKiOBH
zz=}|T5*l4F{r!V@Z=gtA#d^Awak8d9I@;pF5Bqsk_FKpUk)@-p;q1M_wcf@xSHAgG
zQJD2Nt$ASJ4?@udyh)g^cS8iqSR}PJZF)XbNj!I=UVXOx7!};yOY03iyUtSk@J6PT
z0tP^A#E9+XvHhJm^$0`Ndh>>!*8k#>qmp0bRJobFilXPdcY}C#bpNLBPmzR~#dxJ&
zh2x%Lr*kb&LnrQMwhtb^H18+TEH9g;2k|wtpw@0bd4u(hP^w9y$7~nXE1=(h>y%yo
z5-^2<Tc-tdcOOh!Iq1#@qdLrdcd>w#IDm!1X}I5WpO}z91Qf6_7rvg6WU^m`uC8>e
zLaiwVP{+%@SuO?eH=Mqi5TopAh~gC2EN;;KKAdgH3~CkAEdXhBm;r?Y?ogp+NG(#Q
zs{vsUrbtZ?XIhhN!81WFxtsUk#<Vsz4g&L0<RG8*Gek8H3yX5SeK<F{;S-n31PCSp
z$m>HypTXH5>&1-!5k+G9WoQ`A<AWt90<9$TbA<?8*YNrU_hd|&QvJp#R(kpk-T#=5
zI#L)+wF3h;Y0d5)3emPb%TN`wB?V<HiE(GN{Er9-&PKZ)Y}f*hHJM_1?f|RIbu=c8
zZg6+qZdx@)=&w1pTt89I|Bct`$%TR<CC9{_O;^aYgZ0hD-g<O2?0c;}NOj+Gk>ayF
zi!(z}&iyYIz=?XrMV}yfT2M%&^Le6E&<24C9g}Ai+xlUV3<fqTqkzx$q*cxF-K{73
z!=MZiL;i*%1>ix=5|ilIPC8iK-1{V7(kg8n+;P-Vm(K^<KI9xpU+0k8nyKWspQX@)
zX}JPuE|ep-WXQE>VtN{NsiA!w!nv{<?X2nCPjsWpUh>gUCMP=m^Rc*u{E_AC3r}>r
zS6DY!#riKN>v@H5xD8zU>uC#C!be8uD_<M?*Jpz521qdvGte5x9a`%{+tJj^EriOp
zrUNDBi$^QI4V#wH)1!7jQWl}CCZ0-Ox%^qDoB3B9SJ{X}o4jU#q+1q4!R2khJ$98P
zNWansxSiM7Am86|VUa<I+1|lgpq7iGfXpvlyV}c{9#r%#O;wUD@zN1X77?iB3&m>J
zY_y~{jMkX4MSV3wxOkXP#@iH>so(zrGe$w{z~*N+^$Mpr^@_A_1eh=&>+tX4+v=f2
z!P$VG;e+!47hv+50Y2h$927|DQm;0+vDhu7L48r5P^6%pF^3W}RyzxnHIyO4&RfzI
zi;WcmhmC;*L}o1X2FI@JVk+g(8#Un-$3u=P5x%sNC07<Eu1{>?g76PSbpK5PI8$;g
zR8K_r9(v)MQ0y&go*{>T*!U-&=Yfk(t<jJqJzH6j2xtX>3E!g+n6jC^P<eSyJdum6
zSx>iNnT4{0G?|x@zP{>O-UUL#d-J+M*4+d;mF>5VsES)7Vgv>|O2Hp}#ovyYGDBrJ
zIosF#0HScoT5x<v+9Ysd44}`(k(6<yC<Da;$Lt1+QO3}Ee(^-(zu;kKLzl^-V^E9m
zVh6<l(4Y@>>6^}NE@pjD+ubWW6-pec>WPULaI6PLkOvmRXN@(rh53JveOOo=;LJ}D
zboY;Cvm{aLc3+T2>%k^;W%EBLpa2cTKX$`syL*+N@vIl9%IYn#1MY5Fi~TeDgKoo@
zxl#DA@)XNC=ltg9-O4A^m&7@qk>Yo27uMxE*Q0*lxek*IY^WodL3a9CU<8-_5VavK
zspd;y>i6p?+Xep+sxrWA2(_!6xC1rTUml(y{XokVZ6hRyZ9vM&?RyT=PTXLV;Mstg
zmujhy<aYfD*b>U=s<1#p-R&zKb6lC$n?D0%sDPQXX6>%^9uTd;j=KMBTponfvl}+h
zVxdgn<jfCw%=xX%pV*9qe{RFuZ=nOhC#Q%KyaW*}d!)E3k#b?q8v`XRvZ9z&>XA$-
z{MHRb2EubpHxFbW3Ny?~m7c??pgrcoTBBsgm`-G5wxzcVAvGjM6YWKsV4LjwQelP`
z8YZ#70P<l5&ZtC(=SJUk2?*zZ)bIS<EN`7NGc_G5BA)YuVak^3OF4Ck!j)(`FJP7x
zw!54VXwieAo4oa?7Vl>vGg$P`n8zZ^<)m{yD|DvwB+Tv%8Bb@Bq8LPAiZ=^5dla~d
zPO#CVMpsU}Tc@Ptg={*MHZ!hlZwD!s)OcF@3^YV-bXfS_3OPJsMPha+op+3~LFz-J
z(6z<gxNw}qn8N@}0AVD+pFHWZz2pPX1!kkBy2-n1Q)lM{KzZ8}aE$%C){#ks`*R3~
z{*Qb`(leN&%Gb<vcenF-C3!`~m4JKw<3Ya(Ad^hCJ(+y5u@WQpF+OAO$@5|j0fk@h
z!|yKvmMAuhuD8;xGv&75PW1YS*Q0zM+s34aq6k)|yua7HHC@5<EeJ!_du=z9Sl|o1
zgYR;?4uuAj!A_P;21c5@8U20l`0RV=?=bw7wCB6=1R3qaZ)FAaPB$NB#%K#hV+-Ez
zyMML3G1QIU-T!z2xD|8{Em~eYNf&h2vnjRr&fD;IdC7v~d7_)g+E%76%rrnoh?0PU
zqm#T42Pvsx<ArJ%n#i0*mj1sf78)<|du2O2_Ddtc(usgnoLT?IfOQrCT*-wg^y)PG
z)%1`1SCf~i5$IE`_dpma_P&pZL;nF9PF<&Bf)lT_)cZP|P4ITVJe3LhE-nMYa$bHD
z+ppeP5wzYOP`>Dv<R5GK{({difY9@-IsSltdBOJsYLd3U)@n6K1r@5sM521Q0A~J|
z$_A+fkTDdBrKDa9&ctHepB5o3HVSHSahK|@C>@=hqzgPOZ5XG63P_r&ogMKKHUg@h
zWa@<H+m77a&pKU6Abb$Tz`HiS2|e2=mT=5hHt-ZZacYA_@@|`Ze?V}yD|EFd#HTd!
z9Tj9WK&4wE>q6O3F#(|=?0CHRk!TCbd}HIyR6(o|6HJNnd^jseNQ$o!_;d@`Cn84a
zN36xIMoM~;6rqyiwVL$+lu#egb(_GM=S~T1aSyDiz7M9>$Q2J^Net@zoDdloNwV)R
zme`7ao0)3vu+o1x@nEF?^N0phf!rs04<g!W|G{@wuW^9vn5d#%uMy&fnc@&N9c<et
zz`t@7-KGvt(V)X7vndr>D%8VF3q{eXXba086bm2IDA2~G37Av*Fjl&~i_y#~GH(I5
z88sOE7>a_94cJrSBR)LL)nwxfabv~w*J-<uU(r3AeyVnsD~_jOtrc_^a>eU$p~L!e
zGs_S^R{cjCjY}r=%-2J&yuWpju^>X^*D_<4`qE7U*&EEdwxm!@*ST|@+SyL$PkgRM
ze55Jc-7^?zISdy&yudj63H6jyNn|an0^Bno(!U5hUik4lEPHFzbXyA@Hyk*nQ`}xV
zJRtnVZ`iQf;%Plx9_h7rx^b}=zt0a!9LSi*D~Pn`{fT21u9!)+8n^?Dqaq^872)o&
zy#hkj_>=CO>aPvZ--hT*#gV`NYcBkR{!g|rMZzjoZmxt~38pMylJpf$U_s`aTWWuO
zea2KyddwS+^Rq;k+ntr&Xy?ZpPz%(@>vTH)dAf=ohIM&ND|XzMV|DT7j+DVc0VT_;
z|IbNdBW=vW9kbF~zV1$K5c$kLi@n|%qOE&Q`-OD{I#gpG_U&CAb<f>xzJfFc06rR<
zaon85W6A>i{GB=}*WSV@F8Lod@rnRijju@$jQzLThi@*OZurN_aR<e4Ck;KX>~$$U
z9aq2bWiGGaiazmI63&sZ4pVJd97&gbDIyuZPJOf=29ErX8U_d5<nLcVuEEZXURI+<
zICpsPC8&VY!8kTT?$P(GRkn#zSA?Ge<WqpvWYlcH|970t_H9Q@so{3gMDsb#&H{F>
z;xABTVPjaVRX<3PJL}f6cp<X=r6XvStH9Vyj6uq&+eGkWzwH(_KbRP+{oOTMrh}^^
z-iK@knq%%`ggR%ZDyE!9T@;P;h8Vx~Th^+QKCD5?#Ow{=Lxn|^DUy;2P4v^rNP9I(
zme0ZlZ}-0l5(6wua4=9Whh0@@m(LCM%)6T7#*2OJL!%97Yhc`#n@;-k0ykVdvW%l{
zJE@WE-MA_g`E=4q8spnD6UutD?;>EM58bBMrrrGY0j)Brr}LeRv~NPNn*VH7r+v{E
z9^@o;=kpqS%=_E&ZDEi(^4vZ~0+xC=6DO75(FfbVlO^O$M;}v7*o^NwHu7`x54X57
zpIS^1`Cpu&;gpCh=?y`1Z4>iu56<T1?kzu}k>rC++vPn75e{hnYbFEs4U|bOAuF;O
zX}EkSxx$L1D%X#8RlhlJ>AU7Qed*O|qJ*Pg-=q$HqJewf)@&Qk*YOCSQvvaM5DNxF
zQAMIv=Jt;ZDXbBO0U~w)TI#$)nAz&SsP%pQ<>7F9%+$@O<;b0j(8mx(l+he1!mHE0
z&IqK(1v-z46Fu$h>_l7W@nc<0g0kC;40ywC@m%Oo_UPEicKgn5AXy}w5ciD^tyzoX
zN(cN{Nok30DQH~W+#^u{9%~OPYxeB_%552yN@6qCHOqJv3MhF6XHauMcvG#R#c#eP
zyO+yTaTgM&9{=U`)@~vFnzNULW2FoJ*g?dPPQy+$Ryjn-dAm)2YpUTflK;X(RcT*U
ziZYgV(I4$9>l<@>vzSz3hx$ah%&$7bGC9-uh1?#sxZ)LqSqWyy!$j6$(?J?OSB*+~
z0K;n^W=y`g)s=<F#xyv%tLL|C8`s|}?autnk#{&XU7sj%7n(Q(<-Zp@RfAKtE|h0k
zLb=@4zRdPzcE6TdLh1|ml@j>pw<>`Xqdzd6(c~-I^#QHak`yQ5SO1L$lyl(BeSM>I
zS@r0VLK#P4h>nc((TCu;a0XdvRr3k<h$xij*ll_j%V?z0fV&nz!StSkgN)BdBOixK
zNwh~Ov*lORCGW*rxH11#=W97M!Z!sI)f7M%Qy}|&)OX!qmJ9v@^lB+NNND_xORmC`
z`>S_FNsMf(t0eDS76y7>4Jn$eluE`$$cd_{9)%h<>JU&Wnw#gq*bW=S?g8#KU&+;Q
zFp@ar?iLJtG&M&rAJ@v(CUeW(k*!DR0_xtJc6)L7sJFG%`%_Bz5={t8oW73`fLjfC
z)LPMUGNkv~shkZm9==NX5T;pK9KkSxyl5aT?E>ghHvmZ!w^eki2sp-IGZn5@AZ{Gd
zbORYSR{qm6xl!RUr5CpnhlI{fA~&z)zBQlCcTavq`&P4bZ4K~|h%WmMoT0e9si;p;
zN6V=CvI~;3k^n<E5Lx|fa>((oMhYKpG<8UW`|=knc1_BF{+s>my7R>SX}&kK49G~T
z4GuOlF>a-e<$v80zj8+qdE|fH6)-&L-nJV=O%u3~&dRj~84D}m1Sta)Wmxw4f-Ch7
zJ$=?<)-6gxddJb}u<%{a%aJpS>OD@vSMg;^byq#b<j<Xm)_N1fetDAOXXZ`Q0K8-e
zQNn7;?jkO)RYM?%cA=o;xH|X#PO#_xQYs|6+`NT0Pg!76;PqOD-Z9F2bt%@(ZUH~L
zEP%d+jrxMJo$r04xNF{$Yu<E5pGkc3bVgiVRUjfKhX`xPD`5Cu6=KQ)&My|0Yi4z4
zQ47vAc#w#q&Z#;p#5D1j^=LVE$Q9Nbugv98Qp|IgHTt6%X%n=HTOwMaKyzz}NW${o
zwB2s37oP<Dq(YvG7vv+(s|YI==57{UOkj|=zjkxOxIX84%zyIOZi-6oGUPTO=pnC$
zy-{Lsqv61=ltaGFeBJ`5eur2tvJbz&jU9-T1w`9IozmxukV*sn+t5Cze3h%*e8c47
zGHMqCP6cw{D|heB<56QfzM5&@*Ka{OW=Y`+%+6tu@LF@yJKOF!)ixfi(Kb+&bBcf|
zlCazm@E_Z$0$=|l^0>3aW?a<Zv=}S$Pk}OTq@+$xrXPkmvPSPgTvY<3-yl?CI6S<#
z$A1(vE1mn_EMb};Ly(pw#vBYIjYQr#cV-C{Xv&v1$_x}gTjElt4z67}L$)9d#1tG4
zoM?k6<49%*jsmgZ4qRsD%gPhC-hh1eKdt2)=G#|^QzyU02~bd8|7mw|+gylTMt(oK
zF(UeQp2LNcfZq@O<I3uERvbt^)TU7dvwmA#0ufD`9I+~Y%gMz?U?Z^feQil;4k!yd
znTMndv5wg-ytj3XSRDwu|K=O2=NY*wPf?LNiuMgeksdPGz9^t(3ePI415iDH+sw|=
z=*rE@dwO16-z4Bl5h-TaS=s8j<l&dhz~4mNFNpJ}lE-gzdhcn?9!Yn<>D*oBjuQTl
zfE^Sy;A=~Okknwe#3uA)RvYxJb4$hjQT1WWr&nnLu_`2-tAoTs74qG6pD;?9)tYEy
zML11w;zexc#7i5059Ye8PiI|FO`;%aItAq;*AG5O><XU&q4bVyq6CZ}v(9Lq+w83-
zzI_q@yE`ia&nFFNue>XyKy2fSk)NQ`;wA;aEq~=FXmySl4bMui%XOVjgrjCqh|7YZ
zCU#cb05ds9Rb7=eu%zau`if~G<??~C-zBlo<W5JP4xlZd>!5!B=9OoIhiykcAL>Xx
zZfKZy<J;I{cH*hug@af#(fTiv4e=b{`4mp%Ob@dDlE)Aa{`-I%EX>4sXe+o+Ee41z
zxl;PeBLf&O#h*g*r1^jMoUCjvnC9fW=b7%lnR7xm@Zm6U3QCS7xrDKMr(!D%IvbK3
zY}Vs#c63)s5D(dX25Vv!?8z|-Dk>Yrv@LIrO1gc|`#pE&>n597u!0W0Y8Ghtaii5P
zv@eI8qGJc8I)+m^1X}i2S1bF@@t6d_JGrfi7cm@GR_WXrI^7?c_djrda^5WQ`)CCf
zTsuNc73}^(9(yJu*YPR$-M%mGs$<7cCKcEUT<}icJ<ET47`vnA$+102vgXBY;GK&p
z3poDdB9bFX@8dt(`ZSBpb_NOi`t5W$sz<#!KvyM!-IQ_;CB}!Sde1#722+UsQe2`t
zGA=&;%l&WPg)hfAif@vgtY;sci}NshIty?%mk@sHnd{|>)H|HcfB(q_IJr${6u=s=
z=<$O2voNqbaH$xDD6j8I<zHq<)a9XDJQaNl;rQW~zFTQAn>^eQV9Wfj#W)4@rK>eR
zD`9ZLQD=~Hzx8Lk3hB{Ck(j)SLO4iPkV{5x>vJCzfh3M43hlthrCWLKeL6s>dp?i$
z>pkG(TkuEyGaBuB#w(1*e@T_KL(k7MH@22j;W{$8To--f?ZkeT0?sP1jQ~<TTp^cL
zbY8Kq-6yztRY%u}RJYd!K0pkqrTVF1qq~LqeXItCL%+{H`2r0lC0_%lySDLe(uDgK
zG-xJp&-iI_O?p4ppue)-ghw#9GS&-H=f7F;m+mYl`jUuTYup0VyKQ7m(R;a{w9etg
zeaE16G<}o+%hdhXrjrs{S@V_Nd!rW#fn!dh7ZCWugGFED3BXL{bbrfj0!bM9mWZ19
z`#B?jDtX3jns1klM^GA>T@iA>z%7=FjRAtw96wh0wdzq|ZSzdK5jQdRbDyys?HHrE
ziSri}hC}GK*g#h&Tedw-;PYn)s5M%N`E@$kIQu92pV6_>o6w3Q{Pl`jS?rhk17?Yg
z<QjzVVpnT+)+!R0baPv*l&qK6)UQy~#sK&t@L>RroXPA9;KN9S%#AcCa%n2=`zF5B
zg4I?%6qQvu7U4~<v3URGWtul_OwvD>x?pcsIt~mg7ZLaM6+S=1G`nI4bp(Lq015^A
zWL|p$o3o@J-y&aM#^daudLRc__zw7AV{SS%g$P+O(ZCou`9Oy)i=avZ$L#;PBoQly
z!p4`HiU6^W-^96ZWI$!o18kEy{<=NE0)uG~Ddp~THf8d;=;pxM{n<~nMuen+0;Map
zrU;sRkkJwp^zVbAgNvl|IM~b&XG=s;tm(aj*%juxGgZ{_dMQ7@j=$XwsW+ev6!JX@
zj*7{fb}$aaeEB7_s{iDM?D{c4U~b$a>eVCUz9B{mo1(4KEc|lu9@IoA+$kX6t@l-l
zH9GUk2!zEEdr&~q1q5ykAPK)XVgcbDz|!J@Nr#T<s<?;o$!yjBY{PNR=v$gXEV0K-
ztj|R~gah#WgO&+wNTkY2JjzClh!Ypad{gJ%nWLxIuFT2X$Uzu&73$>H@_WnqP=`D@
zFwvxQAg@(f?bMt<VGlgJJ|_J<TcS;-`)Gqci;^ik^P{+K^^FqZYZx4;xxw_Yqg(;K
zP0h)3`o_sm*fz%)&ktOY@S>Ai5wAaDQC;=uNu*HMGp^pn=v5a#JO`UuRPpCa1TMH@
z#DVlQzSM}lfIciLIF2btJhoZJM8L}EE-9|}^f$<dDh0B2`Q59K8x-3ix`8iHPLv@0
zji_45&gbaXq8mdPyQ}$HUuN!z@1H*(QlK)bcu~K&`w{yt_7oV_sB|5@?*BZ8T!xq1
zauUxvhFRZBvpy2DKB;<PqYkoTfA~so6wmV?A}^k-+<k(l>}8Z3H#O3D7}rpiLG_R{
zffym;(k$Y_B!Hhcs7*QxZ3vI-O_WW_ip{T}kcpX@w%#^O^2rw{50F1O3AufL+LDRr
zcSJ-dh!kxRm_*i`;o90i<tE&WUlH&NEJTDK4WoF#-M&oBFxc*9wk1XFgQ^hF{y1k~
zDI^9_Pq#$()^Sa4=-8?UN!1xSc-LzohFiMhvG8=?Bbb?@D^1wDqK(xxy(k_S6b98O
zPz{=znyO+Sb_G#4-y)jflnd5Rk^A7|$^ZReDKPufokbLGYxcPiqhgzA@B=R#1&J{T
z*}I9vcGukGsd#S^hrPM^>SfVw+HW}gkK$<ryhGc@z&8ej(f;p;iUc*G3TLzaKPcX3
z@X)2NyzE_vX{c@3<gjbxMIBpv_-zPp!=G`KzYsC+3ax+MOl*Lc-OK+#aWAOQQ5I?^
z%6Q?zs3jE>oR9@*9;I3Bl4aZjD+3U&;l<R!&4I%7ZAXPQi~ku)IS(2jr=-`2r<1bM
z()nXv6bFm{cg-Ml@ZAqkM*&*woY%>}0&}2v=|6(O(hO(_h7`nKsQ}XIpXTz9bOZQR
z_rLlt(*M6-GAljxz}^VsvE-`gpLI7^Cz;yVrC&v%AY9CGJ{0-i4-hUKFRfV{ILVPD
zOaMewmSHuuxv=fe?Em|<|NMmN{;K*!`Jbozp9dDU`T@iI9RDBW^#6Vkb@2?-{I_q<
z{?%}$j|&MN5qJtWFpZkFj+%ZG<;|SjDMVrB{{o&vVRq${{W2gvoTt8%tZISmxR__%
z*6fwL<Jgwt5jW@KPdv6!JjGEwVq|6%@3NHxT=ICj&zkPwWFX4tL>-CoRrn@t4qP<#
z(Dg?TnQJ89-xEIXOI;o~R^P?!4-f*4UwnR*DtuGYl(xHJ9?$b+fzloP9Hsc&tPciH
zF@CxbetPmU-8(nebEA!n5Zw;ifCO>}CJ{9rezhB85CzCN5@tz_>rt!PeQYPs77BaS
z1kH1J6qLTSCZtKNBC~6yE}>$AW}wcG_<+`Smw2IIl6KYN)zb`89A9WC#8bysd)2e_
zjy^_=i4V$ATI1=JD<_uF>0F512Nl$}#icOo(YL)nQ~2pmSKiwCL*!mL>r2hq`c3i3
z{pz8W#nS_@N>8H#@C>a>OP4>rJC@{C*jWFcwR`(;Vy1Lr#yH5$k!Umg=;@P1J=+*d
z-+H2gxWVUF*-cTs%XY>j$S39e8SZeN?t^Cs^iINJk;m%Z)P&u}-sH*nZPK>85|A*4
z7@^tB#7AvWzJ-0QYZeCV4p5{lh6*;?Z6RB5j^sV9csx7lOn(01JZ(1x3r+OYS_bM(
zMQYZ!3r|^ecI=bgP`AJP8)XKzpa&W?6aHzLVV9LncXxl2E6pvUdGg-gC9F>#J>_%{
zYJjKwhNtAh$@<^>Xl2|_Sh1IvKlF{qb-!!`k4xC*4o~4XdE{humvFYv%$!ac-|Kg!
z?X=%>oI+++hO7r&5z&hkr|JByCp4_*^XPz%dtycN5;hwBf#(Uti@c6+ku|RACe0UR
ztYX3l%NO<4sh4J4vhzdeUr21-bR^#=9(<az-pus=>++v%RJQhC(YEoKxK^H`cunf*
zqi#NYHF)8(67O4E0j0<3x-0kAJUC>L7W<sf!vl9ir9mqDG4W&#balrrK+6*92c9L(
z=2d%8WT1{&pH>!jg*Y~~@ZTsgriq!`X@4svB)wJeWp=LC)<1!`r8O)QGh#uX>XsNX
zoL>Z}KghQ5ykrFTrh;YS+U^q25izH7jo)H|@wWUY!aYhx7<_GNu(bB|W}dv0!AxM7
z!9Z?A@aQHs*}3~J?>lLQ0PXg%Y(6!i?d~I6dtF=mPf?$06`jR+->xqScwY+C)6FJp
zBekxX`Q5Q{1#E91r$eH6ESwb&@46;gZxu5Z0?5t&BIO1$UK(V!_Vm`-p*29jVp|Fx
zS@11AVh6*IY-`WzgmxgX7c<UoyDMncr%rCz$PKUV(FqrTqyVLWjpmP-txZu69SRMd
zO*~%Cs)hmb!j-zY*kLVCVkcaukmR0IOm8Z}Zdq@t_+fG}ref#o3x^Zsn!Cp_D@Otq
zi)2+a7bJCydX0-h#H9XKhu5~~+@nRmU0cBFn!}9hx}bi6TsNHDDZ7h6NrC1v6ooKs
zfoXyRu#5jR0R-~^qOJa`Rtn+~A)wd}(f<9Ci*gi?Hl<k!PoBT%9pC;7I8lH3{X(m0
z)Q4yA>SGUZ+8=?*1{kI4DWErjLI&KebJ$)YAfp(peC))4xTu;qKxY8E(K3n!ryA{j
z{A2J(8g6(A54_|BJgT>@yq0IZIO-bgYX&LK_qXtrHL$~c;!&}j)BZ(P^^=VE`O2LD
zJmot)WmueXNgwP$<b!_mWL5Ml+!IVy;}zp%$L8<=Tc!1qWS+b_p7r$m#f9cD9nIg)
zRQGp83hk9?oKL<g4TW{k=cG=hn~F16#F+$hE6_7UvG06?ekh=MkMWM%uOi|{hn)$=
zj}Od+Hc<+$4H8^I)9NAL^2ViZ-c4;UneuhwWIG4@XC_yj1(esE>G9%*-|KFdsJ}6b
zbKp1>mCw$-OM2RRl_r*Ol9RLBb{ppqcT$*RXyH)Ba0)qx;wI46w&;y)#nmQx`#-Tc
zJiK)T=l`KwwMsj94|m;aCaJrj*<1Dy2y{taT0&!=aW}qt=uBB&v8S2Mjy7q`@NX-I
zu)UA!$TO3VIigF#OD^JVpL1*jMcPo?=4)jc>`}C4Xg*g{)GeHOZJp!KUNb16E+o5E
z?pC@u&Yelhvjv3*61SPP`OlIf$$e$ndbadBYx3-oJT7u^P$yH$IZJK1eKl%XwY@xB
z;PYDAgcImN@otcsBjG{GZVY38_SNYXX3XI}Ek6Ez9D@Q71!cbq+EBPWb!jHtL^Z1R
zBg<moF>Irn4PiVJ(kSK^r>!VH;Z0|Co<Y8`_Q~lmM$$q;CX8BnA%CvgT;Bg+9r=^=
zR{zlzs*@ti`A9X?@|skhFUjL^97GxNQoql~Ro$DjLU&|mWIde3{$0>zu45p1T!S60
z-D@#=dO_w70ImG$V~%<dWgu!C)+<zY8|?9(&)Z6R?x?~RK7OV584s@tj2~}N6(xS}
zRcJpZ;4Urfp<O#wiEi4jp;+{(z1EMan*Nq|blk>U8_r;eR$@PFiBZvY=4gKcx=5Y0
z&kPu*%sG&aYW-?=X=bfnm*yFuVj+7zSI&zcv%iK^Xpf8#2vyDnz{b*z{<t!vzs}%M
z8Tqc}Kt+M%gty7MMy~Op*I&LWkkLNMkF2iT`E~8!E`&r;5Z0j_=jKo8myUI#(bJpy
zr{64X^GkB|u&1=Nn;b!IL_|u7zexqjEXJIJ4<$E>FXxtG5ixK!0yX2{fhvK#_PEHY
zfv@F_ksO^EeqP|{82$A`-d6BMK8Vi2wcF2|sXfQ8z_+#=7y8N+_$96-sXF!VwXztp
zC$v0qV;3?baOBHTAu}S|=B0kav9u01#nG63z{L?R>#OKKB?Wv^+3vepP7v#gh-TB@
zU#M-)`XvQ6i2H9PU_cOU%v=Vcc}q|K{RtbG28i}tKH;C=(7%oQ@0S=}BZM&{A`0Zc
j1K28Puj8fv^Ka{1I+iJarQfVtK|qt2QI;-~GztDcxl7J_
literal 0
HcmV?d00001
--
GitLab