From 62d1c5041d558190dfaf79bcea0adba7b7a27523 Mon Sep 17 00:00:00 2001 From: Morten Hjorth-Jensen Date: Thu, 21 Nov 2024 15:20:31 +0100 Subject: [PATCH] update week 48 --- doc/pub/week48/html/week48-bs.html | 431 +-- doc/pub/week48/html/week48-reveal.html | 2411 ++++++++------- doc/pub/week48/html/week48-solarized.html | 2294 ++++++++------- doc/pub/week48/html/week48.html | 2374 ++++++++------- doc/pub/week48/ipynb/ipynb-week48-src.tar.gz | Bin 822634 -> 823660 bytes doc/pub/week48/ipynb/week48.ipynb | 2767 ++++++++++++------ doc/src/week48/backup2023.do.txt | 1346 +++++++++ doc/src/week48/week48.do.txt | 1307 +++++---- 8 files changed, 8115 insertions(+), 4815 deletions(-) create mode 100644 doc/src/week48/backup2023.do.txt diff --git a/doc/pub/week48/html/week48-bs.html b/doc/pub/week48/html/week48-bs.html index 16ea128bc..ce7e9f3ae 100644 --- a/doc/pub/week48/html/week48-bs.html +++ b/doc/pub/week48/html/week48-bs.html @@ -1,31 +1,28 @@ - + - Week 48: Support Vector Machines and Summary of course - + - - - @@ -166,8 +250,6 @@ MathJax.Hub.Config({ - -
@@ -178,76 +260,81 @@ MathJax.Hub.Config({ Week 48: Support Vector Machines and Summary of course
-
-
-

 

 

 

- - - -
-

Week 48: Support Vector Machines and Summary of course

- -

- -

-Morten Hjorth-Jensen [1, 2] +

Week 48: Support Vector Machines and Summary of course

+
+ + +
+Morten Hjorth-Jensen +
+ +
+Department of Physics, University of Oslo, Norway
- -

- - -

[1] Department of Physics, University of Oslo
-
[2] Department of Physics and Astronomy and National Superconducting Cyclotron Laboratory, Michigan State University

-

-

Nov 29, 2020

+
+

Nov 21, 2024

+

-

+

Read »

@@ -306,29 +386,22 @@ MathJax.Hub.Config({
  • 9
  • 10
  • ...
    • -
  • 65
    • +
  • 71
  • »
    • -
    - - -
    - © 1999-2020, Morten Hjorth-Jensen. Released under CC Attribution-NonCommercial 4.0 license + © 1999-2024, Morten Hjorth-Jensen. Released under CC Attribution-NonCommercial 4.0 license
    - - - diff --git a/doc/pub/week48/html/week48-reveal.html b/doc/pub/week48/html/week48-reveal.html index 068a9388c..234889bce 100644 --- a/doc/pub/week48/html/week48-reveal.html +++ b/doc/pub/week48/html/week48-reveal.html @@ -1,19 +1,18 @@ + - + + - + - Week 48: Support Vector Machines and Summary of course - - - - - - - + @@ -55,36 +54,81 @@ document.getElementsByTagName( 'head' )[0].appendChild( link ); - - - @@ -186,736 +277,988 @@ MathJax.Hub.Config({ - - - - - -

    Week 48: Support Vector Machines and Summary of course

    - -

    - -

    -Morten Hjorth-Jensen [1, 2] +

    Week 48: Support Vector Machines and Summary of course

    +
    + + +
    +Morten Hjorth-Jensen +
    + +
    +Department of Physics, University of Oslo, Norway
    - -

    - - -

    [1] Department of Physics, University of Oslo
    -
    [2] Department of Physics and Astronomy and National Superconducting Cyclotron Laboratory, Michigan State University

    -

    -

    Nov 29, 2020

    +
    +

    Nov 21, 2024

    +

    -

    +









    +

    Overview of week 48

    -

    Overview of week 48

    +









    +

    Plan for week 47

    +
    +Lab sessions on Tuesday and Wednesday +

      -
    • Thursday: Support Vector Machines: Kernels, Classification and Regression. Video of Lecture.
      • -
    • Friday: Summary of course with perspectives for future studies. Video of Lecture.
      • +
    • Work and Discussion of project 3
      • +
    • Last weekly exercise,
    +
    + -See also Geron's chapter 5. Chapter 12 (sections 12.1-12.3 are the most relevant ones) of Hastie et al contains also a good discussion. - +
    +Plans for the lecture Monday 25 November, with video suggestions etc

    +

      +
    1. Bossting and gradient boosting and ensemble models
      2. +
    2. Summary of course
      3. +
    3. Readings and Videos: +
        +
      1. These lecture notes at https://github.com/CompPhysics/MachineLearning/blob/master/doc/pub/week47/ipynb/week48.ipynb
        2. +
      2. See also lecture notes from week 47 at https://github.com/CompPhysics/MachineLearning/blob/master/doc/pub/week46/ipynb/week47.ipynb. The lecture on Monday starts with a repetition on AdaBoost before we move over to gradient boosting with examples + +
        3. +
      3. Video on Decision trees https://www.youtube.com/watch?v=RmajweUFKvM&ab_channel=Simplilearn
        4. +
      4. Video on boosting methods https://www.youtube.com/watch?v=wPqtzj5VZus&ab_channel=H2O.ai
        5. +
      5. Video on AdaBoost https://www.youtube.com/watch?v=LsK-xG1cLYA
        6. +
      6. Video on Gradient boost, part 1, parts 2-4 follow thereafter https://www.youtube.com/watch?v=3CC4N4z3GJc
        7. +
      7. Decision Trees: Rashcka et al chapter 3 pages 86-98, and chapter 7 on Ensemble methods, Voting and Bagging and Gradient Boosting. See also lecture from STK-IN4300, lecture 7 at https://www.uio.no/studier/emner/matnat/math/STK-IN4300/h20/slides/lecture_7.pdf.
        8. +
      +
    +
    + +









    +

    Boosting, a Bird's Eye View

    -

    Thursday

    +

    The basic idea is to combine weak classifiers in order to create a good +classifier. With a weak classifier we often intend a classifier which +produces results which are only slightly better than we would get by +random guesses. +

    -

    -We finalize our discussion on Support Vector Machines with an emphasis on kernel transformations and applications to regression. The following video attempts at giving an overview on this part. See also the follow-up video. +

    This is done by applying in an iterative way a weak (or a standard +classifier like decision trees) to modify the data. In each iteration +we emphasize those observations which are misclassified by weighting +them with a factor. +

    -











    +

    What is boosting? Additive Modelling/Iterative Fitting

    -

    Friday

    +

    Boosting is a way of fitting an additive expansion in a set of +elementary basis functions like for example some simple polynomials. +Assume for example that we have a function +

    +$$ +f_M(x) = \sum_{i=1}^M \beta_m b(x;\gamma_m), +$$ -

    -Friday's lecture is split in two parts. It starts with a summary of -what we have done this semester and continues with perspectives for future studies and -modern research projects in machine learning. +

    where \( \beta_m \) are the expansion parameters to be determined in a +minimization process and \( b(x;\gamma_m) \) are some simple functions of +the multivariable parameter \( x \) which is characterized by the +parameters \( \gamma_m \). +

    + +

    As an example, consider the Sigmoid function we used in logistic +regression. In that case, we can translate the function +\( b(x;\gamma_m) \) into the Sigmoid function +

    + +$$ +\sigma(t) = \frac{1}{1+\exp{(-t)}}, +$$ + +

    where \( t=\gamma_0+\gamma_1 x \) and the parameters \( \gamma_0 \) and +\( \gamma_1 \) were determined by the Logistic Regression fitting +algorithm. +

    + +

    As another example, consider the cost function we defined for linear regression

    +$$ +C(\boldsymbol{y},\boldsymbol{f}) = \frac{1}{n} \sum_{i=0}^{n-1}(y_i-f(x_i))^2. +$$ + +

    In this case the function \( f(x) \) was replaced by the design matrix +\( \boldsymbol{X} \) and the unknown linear regression parameters \( \boldsymbol{\beta} \), +that is \( \boldsymbol{f}=\boldsymbol{X}\boldsymbol{\beta} \). In linear regression we can +simply invert a matrix and obtain the parameters \( \beta \) by +

    + +$$ +\boldsymbol{\beta}=\left(\boldsymbol{X}^T\boldsymbol{X}\right)^{-1}\boldsymbol{X}^T\boldsymbol{y}. +$$ + +

    In iterative fitting or additive modeling, we minimize the cost function with respect to the parameters \( \beta_m \) and \( \gamma_m \).

    -











    +

    Iterative Fitting, Regression and Squared-error Cost Function

    -

    Support Vector Machines, overarching aims

    - -

    -As discussed last week, -a Support Vector Machine (SVM) is a very powerful and versatile -Machine Learning method, capable of performing linear or nonlinear -classification, regression, and even outlier detection. It is one of -the most popular models in Machine Learning, and anyone interested in -Machine Learning should have it in their toolbox. SVMs are -particularly well suited for classification of complex but small-sized or -medium-sized datasets. - -

    -The case with two well-separated classes only can be understood in an -intuitive way in terms of lines in a two-dimensional space separating -the two classes. - -

    -The basic mathematics behind the SVM is however less familiar to most of us. -It relies on the definition of hyperplanes and the -definition of a margin which separates classes (in case of -classification problems) of variables. It is also used for regression -problems. I recommend you take a look at the lectures from last week on the binary classification problem. - -

    -









    - -

    Kernels and non-linearity

    - -

    -The cases we studied last week were all characterized by two classes -with a close to linear separability. The classifiers we have described -so far find linear boundaries in our input feature space. It is -possible to make our procedure more flexible by exploring the feature -space using other basis expansions such as higher-order polynomials, -wavelets, splines etc. - -

    -If our feature space is not easy to separate, as shown in the figure -generated by the code below, we can achieve a better separation by introducing a more complex -basis functions. The ideal would be, as shown by the code example below, to, -via a specific transformation to obtain a separation between the -classes that is almost linear. See also Figures 12.2 and 12.3 of Hastie et al.. - -

    -We can achieve a better separation by introducing more complex -basis functions. The ideal would be (see Figures 12.2 and 12.3) to, via a specific transformation to -obtain a separation between the classes which is almost linear. - -

    -The change of basis, from \( x\rightarrow z=\phi(x) \) leads to the same -type of equations to be solved, except that we need to introduce, for -example, a polynomial transformation to a two-dimensional training -set. - -

    - - -

    import numpy as np
-import os
-
-np.random.seed(42)
-
-# To plot pretty figures
-import matplotlib
-import matplotlib.pyplot as plt
-plt.rcParams['axes.labelsize'] = 14
-plt.rcParams['xtick.labelsize'] = 12
-plt.rcParams['ytick.labelsize'] = 12
-
-
-from sklearn.svm import SVC
-from sklearn import datasets
-
-
-
-X1D = np.linspace(-4, 4, 9).reshape(-1, 1)
-X2D = np.c_[X1D, X1D**2]
-y = np.array([0, 0, 1, 1, 1, 1, 1, 0, 0])
-
-plt.figure(figsize=(11, 4))
-
-plt.subplot(121)
-plt.grid(True, which='both')
-plt.axhline(y=0, color='k')
-plt.plot(X1D[:, 0][y==0], np.zeros(4), "bs")
-plt.plot(X1D[:, 0][y==1], np.zeros(5), "g^")
-plt.gca().get_yaxis().set_ticks([])
-plt.xlabel(r"$x_1$", fontsize=20)
-plt.axis([-4.5, 4.5, -0.2, 0.2])
-
-plt.subplot(122)
-plt.grid(True, which='both')
-plt.axhline(y=0, color='k')
-plt.axvline(x=0, color='k')
-plt.plot(X2D[:, 0][y==0], X2D[:, 1][y==0], "bs")
-plt.plot(X2D[:, 0][y==1], X2D[:, 1][y==1], "g^")
-plt.xlabel(r"$x_1$", fontsize=20)
-plt.ylabel(r"$x_2$", fontsize=20, rotation=0)
-plt.gca().get_yaxis().set_ticks([0, 4, 8, 12, 16])
-plt.plot([-4.5, 4.5], [6.5, 6.5], "r--", linewidth=3)
-plt.axis([-4.5, 4.5, -1, 17])
-plt.subplots_adjust(right=1)
-plt.show()
-
    -

    -









    - -

    The equations

    - -

    -Suppose we define a polynomial transformation of degree two only. We define a vector \( \boldsymbol{x}_i=[x_i,y_i] \) and have -$$ -\boldsymbol{z}_i^T = \phi(\boldsymbol{x}_i)^T =\left[1, \sqrt{2}x_i, \sqrt{2}y_i, x_i^2, y_i^2, \sqrt{2}x_iy_i\right]. -$$ - -

    -With our new basis, the equations we solved earlier are basically the same, that is we have now (without the slack option for simplicity) -$$ -{\cal L}=\sum_i\lambda_i-\frac{1}{2}\sum_{ij}^n\lambda_i\lambda_jy_iy_j\boldsymbol{z}_i^T\boldsymbol{z}_j, -$$ - -subject to the constraints \( \lambda_i\geq 0 \), \( \sum_i\lambda_iy_i=0 \), and for the support vectors -$$ -y_i(\boldsymbol{z}_i^T\boldsymbol{w}+b)= 1 \hspace{0.1cm}\forall i, -$$ - -from which we also find \( b \). -To compute \( \boldsymbol{z}_i^T\boldsymbol{z}_j \) we define the kernel \( K(\boldsymbol{x}_i,\boldsymbol{x}_j) \) as -$$ -K(\boldsymbol{x}_i,\boldsymbol{x}_j)=\boldsymbol{z}_i^T\boldsymbol{z}_j= \phi(\boldsymbol{x}_i)^T\phi(\boldsymbol{x}_j). -$$ - -For the above example, the kernel reads -$$ -K(\boldsymbol{x}_i,\boldsymbol{x}_j)=[1, \sqrt{2}x_i, \sqrt{2}y_i, x_i^2, y_i^2, \sqrt{2}x_iy_i]\begin{bmatrix} 1\\ \sqrt{2}x_j \\ \sqrt{2}y_j \\ x_j^2\\ y_i^2\\ \sqrt{2}x_jy_j \end{bmatrix}=1+2x_ix_j+2y_iy_j+(x_ix_j)^2+(y_iy_j)^2+2x_ix_jy_iy_j. -$$ - -

    -We note that this dot product can be rewritten as -$$ -K(\boldsymbol{x}_i,\boldsymbol{x}_j)=[1+\boldsymbol{x}_i^T\boldsymbol{x}_j]^d, -$$ - -where \( d=2 \) in our case and \( \boldsymbol{x}_i=[x_i,y_i] \) and \( \boldsymbol{x}_j=[x_j,y_j] \). -To compute the last equation is however inefficient from a computational stand. -Instead of computing the last equation for the kernel, we simply compute -the dot product \( (\boldsymbol{x}_i^T\boldsymbol{x}_j)^2 \). -This leads to the so-called -kernel trick. - -

    -









    - -

    The problem to solve

    - -

    -Using our definition of the kernel, we can rewrite again the Lagrangian -$$ -{\cal L}=\sum_i\lambda_i-\frac{1}{2}\sum_{ij}^n\lambda_i\lambda_jy_iy_j\boldsymbol{z}_i^T\boldsymbol{z}_j, -$$ - -subject to the constraints \( \lambda_i\geq 0 \), \( \sum_i\lambda_iy_i=0 \) in terms of a convex optimization problem -$$ -\frac{1}{2} \boldsymbol{\lambda}^T\begin{bmatrix} y_1y_1K(\boldsymbol{x}_1,\boldsymbol{x}_1) & y_1y_2K(\boldsymbol{x}_1,\boldsymbol{x}_2) & \dots & \dots & y_1y_nK(\boldsymbol{x}_1,\boldsymbol{x}_n) \\ -y_2y_1K(\boldsymbol{x}_2,\boldsymbol{x}_1) & y_2y_2(\boldsymbol{x}_2,\boldsymbol{x}_2) & \dots & \dots & y_1y_nK(\boldsymbol{x}_2,\boldsymbol{x}_n) \\ -\dots & \dots & \dots & \dots & \dots \\ -\dots & \dots & \dots & \dots & \dots \\ -y_ny_1K(\boldsymbol{x}_n,\boldsymbol{x}_1) & y_ny_2K(\boldsymbol{x}_n\boldsymbol{x}_2) & \dots & \dots & y_ny_nK(\boldsymbol{x}_n,\boldsymbol{x}_n) \\ -\end{bmatrix}\boldsymbol{\lambda}-\mathbb{1}\boldsymbol{\lambda}, -$$ - -subject to \( \boldsymbol{y}^T\boldsymbol{\lambda}=0 \). Here we defined the vectors \( \boldsymbol{\lambda} =[\lambda_1,\lambda_2,\dots,\lambda_n] \) and -\( \boldsymbol{y}=[y_1,y_2,\dots,y_n] \). -If we add the slack constants this leads to the additional constraint \( 0\leq \lambda_i \leq C \). - -

    -









    - -

    Tailoring the equations to the usage of CVXOPT

    - -

    -We can rewrite this (see the solutions below) in terms of a convex optimization problem of the type -$$ -\begin{align*} - &\mathrm{min}_{\lambda}\hspace{0.2cm} \frac{1}{2}\boldsymbol{\lambda}^T\boldsymbol{P}\boldsymbol{\lambda}+\boldsymbol{q}^T\boldsymbol{\lambda},\\ \nonumber - &\mathrm{s.t} \hspace{0.2cm} \boldsymbol{G}\boldsymbol{\lambda} \preceq \boldsymbol{h} \hspace{0.2cm} \wedge \boldsymbol{A}\boldsymbol{\lambda}=f. -\end{align*} -$$ - -Below we discuss how to solve these equations. Here we note that the matrix \( \boldsymbol{P} \) has matrix elements \( p_{ij}=y_iy_jK(\boldsymbol{x}_i,\boldsymbol{x}_j) \). -Given a kernel \( K \) and the targets \( y_i \) this matrix is easy to set up. The constraint \( \boldsymbol{y}^T\boldsymbol{\lambda}=0 \) leads to \( f=0 \) and \( \boldsymbol{A}=\boldsymbol{y} \). How to set up the matrix \( \boldsymbol{G} \) is discussed later. Here note that the inequalities \( 0\leq \lambda_i \leq C \) can be split up into -\( 0\leq \lambda_i \) and \( \lambda_i \leq C \). These two inequalities define then the matrix \( \boldsymbol{G} \) and the vector \( \boldsymbol{h} \). - -

    -









    - -

    Different kernels and Mercer's theorem

    - -

    -There are several popular kernels being used. These are +

    The way we proceed is as follows (here we specialize to the squared-error cost function)

      -
    1. Linear: \( K(\boldsymbol{x},\boldsymbol{y})=\boldsymbol{x}^T\boldsymbol{y} \),
      2. -
    2. Polynomial: \( K(\boldsymbol{x},\boldsymbol{y})=(\boldsymbol{x}^T\boldsymbol{y}+\gamma)^d \),
      3. -
    3. Gaussian Radial Basis Function: \( K(\boldsymbol{x},\boldsymbol{y})=\exp{\left(-\gamma\vert\vert\boldsymbol{x}-\boldsymbol{y}\vert\vert^2\right)} \),
      4. -
    4. Tanh: \( K(\boldsymbol{x},\boldsymbol{y})=\tanh{(\boldsymbol{x}^T\boldsymbol{y}+\gamma)} \),
      5. +
    5. Establish a cost function, here \( {\cal C}(\boldsymbol{y},\boldsymbol{f}) = \frac{1}{n} \sum_{i=0}^{n-1}(y_i-f_M(x_i))^2 \) with \( f_M(x) = \sum_{i=1}^M \beta_m b(x;\gamma_m) \).
      6. +
    6. Initialize with a guess \( f_0(x) \). It could be one or even zero or some random numbers.
      7. +
    7. For \( m=1:M \) +
        +
      1. minimize \( \sum_{i=0}^{n-1}(y_i-f_{m-1}(x_i)-\beta b(x;\gamma))^2 \) wrt \( \gamma \) and \( \beta \)
        2. +
      2. This gives the optimal values \( \beta_m \) and \( \gamma_m \)
        3. +
      3. Determine then the new values \( f_m(x)=f_{m-1}(x) +\beta_m b(x;\gamma_m) \)
      +
    +

    We could use any of the algorithms we have discussed till now. If we +use trees, \( \gamma \) parameterizes the split variables and split points +at the internal nodes, and the predictions at the terminal nodes. +

    -and many other ones. - -

    -An important theorem for us is Mercer's -theorem. The -theorem states that if a kernel function \( K \) is symmetric, continuous -and leads to a positive semi-definite matrix \( \boldsymbol{P} \) then there -exists a function \( \phi \) that maps \( \boldsymbol{x}_i \) and \( \boldsymbol{x}_j \) into -another space (possibly with much higher dimensions) such that - -$$ -K(\boldsymbol{x}_i,\boldsymbol{x}_j)=\phi(\boldsymbol{x}_i)^T\phi(\boldsymbol{x}_j). -$$ - -

    -So you can use \( K \) as a kernel since you know \( \phi \) exists, even if -you don’t know what \( \phi \) is. - -

    -Note that some frequently used kernels (such as the Sigmoid kernel) -don’t respect all of Mercer’s conditions, yet they generally work well -in practice. - -











    +

    Squared-Error Example and Iterative Fitting

    + +

    To better understand what happens, let us develop the steps for the iterative fitting using the above squared error function.

    + +

    For simplicity we assume also that our functions \( b(x;\gamma)=1+\gamma x \).

    + +

    This means that for every iteration \( m \), we need to optimize

    + +$$ +(\beta_m,\gamma_m) = \mathrm{argmin}_{\beta,\lambda}\hspace{0.1cm} \sum_{i=0}^{n-1}(y_i-f_{m-1}(x_i)-\beta b(x;\gamma))^2=\sum_{i=0}^{n-1}(y_i-f_{m-1}(x_i)-\beta(1+\gamma x_i))^2. +$$ + +

    We start our iteration by simply setting \( f_0(x)=0 \). +Taking the derivatives with respect to \( \beta \) and \( \gamma \) we obtain +

    +$$ +\frac{\partial {\cal C}}{\partial \beta} = -2\sum_{i}(1+\gamma x_i)(y_i-\beta(1+\gamma x_i))=0, +$$ + +

    and

    +$$ +\frac{\partial {\cal C}}{\partial \gamma} =-2\sum_{i}\beta x_i(y_i-\beta(1+\gamma x_i))=0. +$$ + +

    We can then rewrite these equations as (defining \( \boldsymbol{w}=\boldsymbol{e}+\gamma \boldsymbol{x}) \) with \( \boldsymbol{e} \) being the unit vector)

    +$$ +\gamma \boldsymbol{w}^T(\boldsymbol{y}-\beta\gamma \boldsymbol{w})=0, +$$ + +

    which gives us \( \beta = \boldsymbol{w}^T\boldsymbol{y}/(\boldsymbol{w}^T\boldsymbol{w}) \). Similarly we have

    +$$ +\beta\gamma \boldsymbol{x}^T(\boldsymbol{y}-\beta(1+\gamma \boldsymbol{x}))=0, +$$ + +

    which leads to \( \gamma =(\boldsymbol{x}^T\boldsymbol{y}-\beta\boldsymbol{x}^T\boldsymbol{e})/(\beta\boldsymbol{x}^T\boldsymbol{x}) \). Inserting +for \( \beta \) gives us an equation for \( \gamma \). This is a non-linear equation in the unknown \( \gamma \) and has to be solved numerically. +

    + +

    The solution to these two equations gives us in turn \( \beta_1 \) and \( \gamma_1 \) leading to the new expression for \( f_1(x) \) as +\( f_1(x) = \beta_1(1+\gamma_1x) \). Doing this \( M \) times results in our final estimate for the function \( f \). +

    + +









    +

    Iterative Fitting, Classification and AdaBoost

    + +

    Let us consider a binary classification problem with two outcomes \( y_i \in \{-1,1\} \) and \( i=0,1,2,\dots,n-1 \) as our set of +observations. We define a classification function \( G(x) \) which produces a prediction taking one or the other of the two values +\( \{-1,1\} \). +

    + +

    The error rate of the training sample is then

    + +$$ +\mathrm{\overline{err}}=\frac{1}{n} \sum_{i=0}^{n-1} I(y_i\ne G(x_i)). +$$ + +

    The iterative procedure starts with defining a weak classifier whose +error rate is barely better than random guessing. The iterative +procedure in boosting is to sequentially apply a weak +classification algorithm to repeatedly modified versions of the data +producing a sequence of weak classifiers \( G_m(x) \). +

    + +

    Here we will express our function \( f(x) \) in terms of \( G(x) \). That is

    +$$ +f_M(x) = \sum_{i=1}^M \beta_m b(x;\gamma_m), +$$ + +

    will be a function of

    +$$ +G_M(x) = \mathrm{sign} \sum_{i=1}^M \alpha_m G_m(x). +$$ + + +









    +

    Adaptive Boosting, AdaBoost

    + +

    In our iterative procedure we define thus

    +$$ +f_m(x) = f_{m-1}(x)+\beta_mG_m(x). +$$ + +

    The simplest possible cost function which leads (also simple from a computational point of view) to the AdaBoost algorithm is the +exponential cost/loss function defined as +

    +$$ +C(\boldsymbol{y},\boldsymbol{f}) = \sum_{i=0}^{n-1}\exp{(-y_i(f_{m-1}(x_i)+\beta G(x_i))}. +$$ + +

    We optimize \( \beta \) and \( G \) for each value of \( m=1:M \) as we did in the regression case. +This is normally done in two steps. Let us however first rewrite the cost function as +

    + +$$ +C(\boldsymbol{y},\boldsymbol{f}) = \sum_{i=0}^{n-1}w_i^{m}\exp{(-y_i\beta G(x_i))}, +$$ + +

    where we have defined \( w_i^m= \exp{(-y_if_{m-1}(x_i))} \).

    + +









    +

    Building up AdaBoost

    + +

    First, for any \( \beta > 0 \), we optimize \( G \) by setting

    +$$ +G_m(x) = \mathrm{sign} \sum_{i=0}^{n-1} w_i^m I(y_i \ne G_(x_i)), +$$ + +

    which is the classifier that minimizes the weighted error rate in predicting \( y \).

    + +

    We can do this by rewriting

    +$$ +\exp{-(\beta)}\sum_{y_i=G(x_i)}w_i^m+\exp{(\beta)}\sum_{y_i\ne G(x_i)}w_i^m, +$$ + +

    which can be rewritten as

    +$$ +(\exp{(\beta)}-\exp{-(\beta)})\sum_{i=0}^{n-1}w_i^mI(y_i\ne G(x_i))+\exp{(-\beta)}\sum_{i=0}^{n-1}w_i^m=0, +$$ + +

    which leads to

    +$$ +\beta_m = \frac{1}{2}\log{\frac{1-\mathrm{\overline{err}}}{\mathrm{\overline{err}}}}, +$$ + +

    where we have redefined the error as

    +$$ +\mathrm{\overline{err}}_m=\frac{1}{n}\frac{\sum_{i=0}^{n-1}w_i^mI(y_i\ne G(x_i)}{\sum_{i=0}^{n-1}w_i^m}, +$$ + +

    which leads to an update of

    +$$ +f_m(x) = f_{m-1}(x) +\beta_m G_m(x). +$$ + +

    This leads to the new weights

    +$$ +w_i^{m+1} = w_i^m \exp{(-y_i\beta_m G_m(x_i))} +$$ + + +









    +

    Adaptive boosting: AdaBoost, Basic Algorithm

    + +

    The algorithm here is rather straightforward. Assume that our weak +classifier is a decision tree and we consider a binary set of outputs +with \( y_i \in \{-1,1\} \) and \( i=0,1,2,\dots,n-1 \) as our set of +observations. Our design matrix is given in terms of the +feature/predictor vectors +\( \boldsymbol{X}=[\boldsymbol{x}_0\boldsymbol{x}_1\dots\boldsymbol{x}_{p-1}] \). Finally, we define also a +classifier determined by our data via a function \( G(x) \). This function tells us how well we are able to classify our outputs/targets \( \boldsymbol{y} \). +

    + +

    We have already defined the misclassification error \( \mathrm{err} \) as

    +$$ +\mathrm{err}=\frac{1}{n}\sum_{i=0}^{n-1}I(y_i\ne G(x_i)), +$$ + +

    where the function \( I() \) is one if we misclassify and zero if we classify correctly.

    + +









    +

    Basic Steps of AdaBoost

    + +

    With the above definitions we are now ready to set up the algorithm for AdaBoost. +The basic idea is to set up weights which will be used to scale the correctly classified and the misclassified cases. +

    +
      +
    1. We start by initializing all weights to \( w_i = 1/n \), with \( i=0,1,2,\dots n-1 \). It is easy to see that we must have \( \sum_{i=0}^{n-1}w_i = 1 \).
      2. +
    2. We rewrite the misclassification error as
      3. +
    +$$ +\mathrm{\overline{err}}_m=\frac{\sum_{i=0}^{n-1}w_i^m I(y_i\ne G(x_i))}{\sum_{i=0}^{n-1}w_i}, +$$ + +
      +
    1. Then we start looping over all attempts at classifying, namely we start an iterative process for \( m=1:M \), where \( M \) is the final number of classifications. Our given classifier could for example be a plain decision tree. +
        +
      1. Fit then a given classifier to the training set using the weights \( w_i \).
        2. +
      2. Compute then \( \mathrm{err} \) and figure out which events are classified properly and which are classified wrongly.
        3. +
      3. Define a quantity \( \alpha_{m} = \log{(1-\mathrm{\overline{err}}_m)/\mathrm{\overline{err}}_m} \)
        4. +
      4. Set the new weights to \( w_i = w_i\times \exp{(\alpha_m I(y_i\ne G(x_i)} \).
        5. +
      +
    2. Compute the new classifier \( G(x)= \sum_{i=0}^{n-1}\alpha_m I(y_i\ne G(x_i) \).
      3. +
    +

    For the iterations with \( m \le 2 \) the weights are modified +individually at each steps. The observations which were misclassified +at iteration \( m-1 \) have a weight which is larger than those which were +classified properly. As this proceeds, the observations which were +difficult to classifiy correctly are given a larger influence. Each +new classification step \( m \) is then forced to concentrate on those +observations that are missed in the previous iterations. +

    + +









    +

    AdaBoost Examples

    + +

    Using Scikit-Learn it is easy to apply the adaptive boosting algorithm, as done here.

    -

    The moons example (Adapted from Geron, chapter 5)

    -

    -

    from __future__ import division, print_function, unicode_literals
+
    
+  
    
+    
    
+      
    
+        
    
+  
    from sklearn.ensemble import AdaBoostClassifier
 
+ada_clf = AdaBoostClassifier(
+    DecisionTreeClassifier(max_depth=2), n_estimators=200,
+    algorithm="SAMME.R", learning_rate=0.01, random_state=42)
+ada_clf.fit(X_train, y_train)
+y_pred = ada_clf.predict(X_test)
+skplt.metrics.plot_confusion_matrix(y_test, y_pred, normalize=True)
+plt.show()
+y_probas = ada_clf.predict_proba(X_test)
+skplt.metrics.plot_roc(y_test, y_probas)
+plt.show()
+skplt.metrics.plot_cumulative_gain(y_test, y_probas)
+plt.show()
+
    
+
    
+      
    
+    
    
+  
    
+  
    
+    
    
+      
    
+        
              
+        
    
+      
    
+    
    
+  
    
+
    
+
+
+









    
+
    Making an  ADAboost code yourself 
    
+
+
+
+
    
+  
    
+    
    
+      
    
+        
    
+  
    import numpy as np
+
+class DecisionStump:
+    def fit(self, X, y, weights):
+        m, n = X.shape
+        self.alpha = 0
+        self.threshold = None
+        self.polarity = 1
+
+        min_error = float('inf')
+
+        for feature in range(n):
+            feature_values = np.unique(X[:, feature])
+
+            for threshold in feature_values:
+                for polarity in [1, -1]:
+                    predictions = np.ones(m)
+                    predictions[X[:, feature] < threshold] = -1
+                    predictions *= polarity
+
+                    error = sum(weights[predictions != y])
+
+                    if error < min_error:
+                        min_error = error
+                        self.alpha = 0.5 * np.log((1 - error) / (error + 1e-10))
+                        self.threshold = threshold
+                        self.feature_index = feature
+                        self.polarity = polarity
+
+    def predict(self, X):
+        m = X.shape[0]
+        predictions = np.ones(m)
+        if self.polarity == 1:
+            predictions[X[:, self.feature_index] < self.threshold] = -1
+        else:
+            predictions[X[:, self.feature_index] >= self.threshold] = -1
+        return predictions
+
+class AdaBoost:
+    def fit(self, X, y, n_estimators):
+        m = X.shape[0]
+        self.alphas = []
+        self.models = []
+
+        weights = np.ones(m) / m
+
+        for _ in range(n_estimators):
+            stump = DecisionStump()
+            stump.fit(X, y, weights)
+            predictions = stump.predict(X)
+
+            error = sum(weights[predictions != y])
+            if error == 0:
+                break
+
+            self.models.append(stump)
+            self.alphas.append(stump.alpha)
+
+            weights *= np.exp(-stump.alpha * y * predictions)
+            weights /= np.sum(weights)
+
+    def predict(self, X):
+        final_predictions = np.zeros(X.shape[0])
+        for alpha, model in zip(self.alphas, self.models):
+            final_predictions += alpha * model.predict(X)
+        return np.sign(final_predictions)
+
+# Example dataset (X, y)
+X = np.array([[1], [2], [3], [4], [5], [6], [7], [8], [9], [10]])
+y = np.array([-1, -1, -1, -1, 1, 1, 1, 1, 1, 1])  # Labels must be -1 or 1
+
+# Train AdaBoost
+ada = AdaBoost()
+ada.fit(X, y, n_estimators=10)
+
+# Predictions
+predictions = ada.predict(X)
+print("Predictions:", predictions)
+
    
+
    
+      
    
+    
    
+  
    
+  
    
+    
    
+      
    
+        
              
+        
    
+      
    
+    
    
+  
    
+
    
+
+
+









    
+
    Gradient boosting: Basics with Steepest Descent/Functional Gradient Descent 
    
+
+
    Gradient boosting is again a similar technique to Adaptive boosting,
+it combines so-called weak classifiers or regressors into a strong
+method via a series of iterations.
+
    
+
+
    In order to understand the method, let us illustrate its basics by
+bringing back the essential steps in linear regression, where our cost
+function was the least squares function.
+
    
+
+









    
+
    The Squared-Error again! Steepest Descent 
    
+
+
    We start again with our cost function \( {\cal C}(\boldsymbol{y}m\boldsymbol{f})=\sum_{i=0}^{n-1}{\cal L}(y_i, f(x_i)) \) where we want to minimize
+This means that for every iteration, we need to optimize
+
    
+
+$$
+(\hat{\boldsymbol{f}}) = \mathrm{argmin}_{\boldsymbol{f}}\hspace{0.1cm} \sum_{i=0}^{n-1}(y_i-f(x_i))^2.
+$$
+
+
    We define a real function \( h_m(x) \) that defines our final function \( f_M(x) \) as
    
+$$
+f_M(x) = \sum_{m=0}^M h_m(x).
+$$
+
+
    In the steepest decent approach we approximate \( h_m(x) = -\rho_m g_m(x) \), where \( \rho_m \) is a scalar and \( g_m(x) \) the gradient defined as
    
+$$
+g_m(x_i) = \left[ \frac{\partial {\cal L}(y_i, f(x_i))}{\partial f(x_i)}\right]_{f(x_i)=f_{m-1}(x_i)}.
+$$
+
+
    With the new gradient we can update \( f_m(x) = f_{m-1}(x) -\rho_m g_m(x) \). Using the above squared-error function we see that
+the gradient is \( g_m(x_i) = -2(y_i-f(x_i)) \).
+
    
+
+
    Choosing \( f_0(x)=0 \) we obtain \( g_m(x) = -2y_i \) and inserting this into the minimization problem for the cost function we have
    
+$$
+(\rho_1) = \mathrm{argmin}_{\rho}\hspace{0.1cm} \sum_{i=0}^{n-1}(y_i+2\rho y_i)^2.
+$$
+
+
+









    
+
    Steepest Descent Example 
    
+
+
    Optimizing with respect to \( \rho \) we obtain (taking the derivative) that \( \rho_1 = -1/2 \). We have then that
    
+$$
+f_1(x) = f_{0}(x) -\rho_1 g_1(x)=-y_i.
+$$
+
+
    We can then proceed and compute
    
+$$
+g_2(x_i) = \left[ \frac{\partial {\cal L}(y_i, f(x_i))}{\partial f(x_i)}\right]_{f(x_i)=f_{1}(x_i)=y_i}=-4y_i,
+$$
+
+
    and find a new value for \( \rho_2=-1/2 \) and continue till we have reached \( m=M \). We can modify the steepest descent method, or steepest boosting, by introducing what is called gradient boosting. 
    
+
+









    
+
    Gradient Boosting, algorithm 
    
+
+
    Steepest descent is however not much used, since it only optimizes \( f \) at a fixed set of \( n \) points,
+so we do not learn a function that can generalize. However, we can modify the algorithm by
+fitting a weak learner to approximate the negative gradient signal. 
+
    
+
+
    Suppose we have a cost function \( C(f)=\sum_{i=0}^{n-1}L(y_i, f(x_i)) \) where \( y_i \) is our target and \( f(x_i) \) the function which is meant to model \( y_i \). The above cost function could be our standard  squared-error  function
    
+$$
+C(\boldsymbol{y},\boldsymbol{f})=\sum_{i=0}^{n-1}(y_i-f(x_i))^2.
+$$
+
+
    The way we proceed in an iterative fashion is to
    
+
      
+
    1.  Initialize our estimate \( f_0(x) \).
      2. 
+
    2.  For \( m=1:M \), we
+
        
+ 
      1.  compute the negative gradient vector \( \boldsymbol{u}_m = -\partial C(\boldsymbol{y},\boldsymbol{f})/\partial \boldsymbol{f}(x) \) at \( f(x) = f_{m-1}(x) \);
        2. 
+ 
      2.  fit the so-called base-learner to the negative gradient \( h_m(u_m,x) \);
        3. 
+ 
      3.  update the estimate \( f_m(x) = f_{m-1}(x)+h_m(u_m,x) \);
        4. 
+
      
+
    3.  The final estimate is then \( f_M(x) = \sum_{m=1}^M h_m(u_m,x) \).
      4. 
+
    
+









    
+
    Gradient Boosting, Examples of Regression 
    
+
+
+
    
+  
    
+    
    
+      
    
+        
    
+  
    import matplotlib.pyplot as plt
 import numpy as np
-np.random.seed(42)
+from sklearn.model_selection import train_test_split
+from sklearn.ensemble import GradientBoostingRegressor
+import scikitplot as skplt
+from sklearn.metrics import mean_squared_error
 
-import matplotlib
-import matplotlib.pyplot as plt
-plt.rcParams['axes.labelsize'] = 14
-plt.rcParams['xtick.labelsize'] = 12
-plt.rcParams['ytick.labelsize'] = 12
+n = 100
+maxdegree = 6
+
+# Make data set.
+x = np.linspace(-3, 3, n).reshape(-1, 1)
+y = np.exp(-x**2) + 1.5 * np.exp(-(x-2)**2)+ np.random.normal(0, 0.1, x.shape)
+
+error = np.zeros(maxdegree)
+bias = np.zeros(maxdegree)
+variance = np.zeros(maxdegree)
+polydegree = np.zeros(maxdegree)
+X_train, X_test, y_train, y_test = train_test_split(x, y, test_size=0.2)
+
+for degree in range(1,maxdegree):
+    model = GradientBoostingRegressor(max_depth=degree, n_estimators=100, learning_rate=1.0)  
+    model.fit(X_train,y_train)
+    y_pred = model.predict(X_test)
+    polydegree[degree] = degree
+    error[degree] = np.mean( np.mean((y_test - y_pred)**2) )
+    bias[degree] = np.mean( (y_test - np.mean(y_pred))**2 )
+    variance[degree] = np.mean( np.var(y_pred) )
+    print('Max depth:', degree)
+    print('Error:', error[degree])
+    print('Bias^2:', bias[degree])
+    print('Var:', variance[degree])
+    print('{} >= {} + {} = {}'.format(error[degree], bias[degree], variance[degree], bias[degree]+variance[degree]))
+
+plt.xlim(1,maxdegree-1)
+plt.plot(polydegree, error, label='Error')
+plt.plot(polydegree, bias, label='bias')
+plt.plot(polydegree, variance, label='Variance')
+plt.legend()
+save_fig("gdregression")
+plt.show()
+
    
+
    
+      
    
+    
    
+  
    
+  
    
+    
    
+      
    
+        
              
+        
    
+      
    
+    
    
+  
    
+
    
 
 
-from sklearn.svm import SVC
-from sklearn import datasets
+









    
+
    Gradient Boosting, Classification Example 
    
 
+
+
    
+  
    
+    
    
+      
    
+        
    
+  
    import matplotlib.pyplot as plt
+import numpy as np
+from sklearn.model_selection import  train_test_split 
+from sklearn.datasets import load_breast_cancer
+import scikitplot as skplt
+from sklearn.ensemble import GradientBoostingClassifier
+from sklearn.model_selection import cross_validate
 
+# Load the data
+cancer = load_breast_cancer()
 
-from sklearn.pipeline import Pipeline
+X_train, X_test, y_train, y_test = train_test_split(cancer.data,cancer.target,random_state=0)
+print(X_train.shape)
+print(X_test.shape)
+#now scale the data
 from sklearn.preprocessing import StandardScaler
-from sklearn.svm import LinearSVC
+scaler = StandardScaler()
+scaler.fit(X_train)
+X_train_scaled = scaler.transform(X_train)
+X_test_scaled = scaler.transform(X_test)
 
+gd_clf = GradientBoostingClassifier(max_depth=3, n_estimators=100, learning_rate=1.0)  
+gd_clf.fit(X_train_scaled, y_train)
+#Cross validation
+accuracy = cross_validate(gd_clf,X_test_scaled,y_test,cv=10)['test_score']
+print(accuracy)
+print("Test set accuracy with Gradient boosting and scaled data: {:.2f}".format(gd_clf.score(X_test_scaled,y_test)))
 
-from sklearn.datasets import make_moons
-X, y = make_moons(n_samples=100, noise=0.15, random_state=42)
-
-def plot_dataset(X, y, axes):
-    plt.plot(X[:, 0][y==0], X[:, 1][y==0], "bs")
-    plt.plot(X[:, 0][y==1], X[:, 1][y==1], "g^")
-    plt.axis(axes)
-    plt.grid(True, which='both')
-    plt.xlabel(r"$x_1$", fontsize=20)
-    plt.ylabel(r"$x_2$", fontsize=20, rotation=0)
-
-plot_dataset(X, y, [-1.5, 2.5, -1, 1.5])
+import scikitplot as skplt
+y_pred = gd_clf.predict(X_test_scaled)
+skplt.metrics.plot_confusion_matrix(y_test, y_pred, normalize=True)
+save_fig("gdclassiffierconfusion")
 plt.show()
-
-from sklearn.datasets import make_moons
-from sklearn.pipeline import Pipeline
-from sklearn.preprocessing import PolynomialFeatures
-
-polynomial_svm_clf = Pipeline([
-        ("poly_features", PolynomialFeatures(degree=3)),
-        ("scaler", StandardScaler()),
-        ("svm_clf", LinearSVC(C=10, loss="hinge", random_state=42))
-    ])
-
-polynomial_svm_clf.fit(X, y)
-
-def plot_predictions(clf, axes):
-    x0s = np.linspace(axes[0], axes[1], 100)
-    x1s = np.linspace(axes[2], axes[3], 100)
-    x0, x1 = np.meshgrid(x0s, x1s)
-    X = np.c_[x0.ravel(), x1.ravel()]
-    y_pred = clf.predict(X).reshape(x0.shape)
-    y_decision = clf.decision_function(X).reshape(x0.shape)
-    plt.contourf(x0, x1, y_pred, cmap=plt.cm.brg, alpha=0.2)
-    plt.contourf(x0, x1, y_decision, cmap=plt.cm.brg, alpha=0.1)
-
-plot_predictions(polynomial_svm_clf, [-1.5, 2.5, -1, 1.5])
-plot_dataset(X, y, [-1.5, 2.5, -1, 1.5])
-
+y_probas = gd_clf.predict_proba(X_test_scaled)
+skplt.metrics.plot_roc(y_test, y_probas)
+save_fig("gdclassiffierroc")
 plt.show()
-
-
-from sklearn.svm import SVC
-
-poly_kernel_svm_clf = Pipeline([
-        ("scaler", StandardScaler()),
-        ("svm_clf", SVC(kernel="poly", degree=3, coef0=1, C=5))
-    ])
-poly_kernel_svm_clf.fit(X, y)
-
-poly100_kernel_svm_clf = Pipeline([
-        ("scaler", StandardScaler()),
-        ("svm_clf", SVC(kernel="poly", degree=10, coef0=100, C=5))
-    ])
-poly100_kernel_svm_clf.fit(X, y)
-
-plt.figure(figsize=(11, 4))
-
-plt.subplot(121)
-plot_predictions(poly_kernel_svm_clf, [-1.5, 2.5, -1, 1.5])
-plot_dataset(X, y, [-1.5, 2.5, -1, 1.5])
-plt.title(r"$d=3, r=1, C=5$", fontsize=18)
-
-plt.subplot(122)
-plot_predictions(poly100_kernel_svm_clf, [-1.5, 2.5, -1, 1.5])
-plot_dataset(X, y, [-1.5, 2.5, -1, 1.5])
-plt.title(r"$d=10, r=100, C=5$", fontsize=18)
-
-plt.show()
-
-def gaussian_rbf(x, landmark, gamma):
-    return np.exp(-gamma * np.linalg.norm(x - landmark, axis=1)**2)
-
-gamma = 0.3
-
-x1s = np.linspace(-4.5, 4.5, 200).reshape(-1, 1)
-x2s = gaussian_rbf(x1s, -2, gamma)
-x3s = gaussian_rbf(x1s, 1, gamma)
-
-XK = np.c_[gaussian_rbf(X1D, -2, gamma), gaussian_rbf(X1D, 1, gamma)]
-yk = np.array([0, 0, 1, 1, 1, 1, 1, 0, 0])
-
-plt.figure(figsize=(11, 4))
-
-plt.subplot(121)
-plt.grid(True, which='both')
-plt.axhline(y=0, color='k')
-plt.scatter(x=[-2, 1], y=[0, 0], s=150, alpha=0.5, c="red")
-plt.plot(X1D[:, 0][yk==0], np.zeros(4), "bs")
-plt.plot(X1D[:, 0][yk==1], np.zeros(5), "g^")
-plt.plot(x1s, x2s, "g--")
-plt.plot(x1s, x3s, "b:")
-plt.gca().get_yaxis().set_ticks([0, 0.25, 0.5, 0.75, 1])
-plt.xlabel(r"$x_1$", fontsize=20)
-plt.ylabel(r"Similarity", fontsize=14)
-plt.annotate(r'$\mathbf{x}$',
-             xy=(X1D[3, 0], 0),
-             xytext=(-0.5, 0.20),
-             ha="center",
-             arrowprops=dict(facecolor='black', shrink=0.1),
-             fontsize=18,
-            )
-plt.text(-2, 0.9, "$x_2$", ha="center", fontsize=20)
-plt.text(1, 0.9, "$x_3$", ha="center", fontsize=20)
-plt.axis([-4.5, 4.5, -0.1, 1.1])
-
-plt.subplot(122)
-plt.grid(True, which='both')
-plt.axhline(y=0, color='k')
-plt.axvline(x=0, color='k')
-plt.plot(XK[:, 0][yk==0], XK[:, 1][yk==0], "bs")
-plt.plot(XK[:, 0][yk==1], XK[:, 1][yk==1], "g^")
-plt.xlabel(r"$x_2$", fontsize=20)
-plt.ylabel(r"$x_3$  ", fontsize=20, rotation=0)
-plt.annotate(r'$\phi\left(\mathbf{x}\right)$',
-             xy=(XK[3, 0], XK[3, 1]),
-             xytext=(0.65, 0.50),
-             ha="center",
-             arrowprops=dict(facecolor='black', shrink=0.1),
-             fontsize=18,
-            )
-plt.plot([-0.1, 1.1], [0.57, -0.1], "r--", linewidth=3)
-plt.axis([-0.1, 1.1, -0.1, 1.1])
-    
-plt.subplots_adjust(right=1)
-
+skplt.metrics.plot_cumulative_gain(y_test, y_probas)
+save_fig("gdclassiffiercgain")
 plt.show()
+
    
+
    
+      
    
+    
    
+  
    
+  
    
+    
    
+      
    
+        
              
+        
    
+      
    
+    
    
+  
    
+
    
 
 
-x1_example = X1D[3, 0]
-for landmark in (-2, 1):
-    k = gaussian_rbf(np.array([[x1_example]]), np.array([[landmark]]), gamma)
-    print("Phi({}, {}) = {}".format(x1_example, landmark, k))
-
-rbf_kernel_svm_clf = Pipeline([
-        ("scaler", StandardScaler()),
-        ("svm_clf", SVC(kernel="rbf", gamma=5, C=0.001))
-    ])
-rbf_kernel_svm_clf.fit(X, y)
-
-
-from sklearn.svm import SVC
-
-gamma1, gamma2 = 0.1, 5
-C1, C2 = 0.001, 1000
-hyperparams = (gamma1, C1), (gamma1, C2), (gamma2, C1), (gamma2, C2)
-
-svm_clfs = []
-for gamma, C in hyperparams:
-    rbf_kernel_svm_clf = Pipeline([
-            ("scaler", StandardScaler()),
-            ("svm_clf", SVC(kernel="rbf", gamma=gamma, C=C))
-        ])
-    rbf_kernel_svm_clf.fit(X, y)
-    svm_clfs.append(rbf_kernel_svm_clf)
-
-plt.figure(figsize=(11, 7))
-
-for i, svm_clf in enumerate(svm_clfs):
-    plt.subplot(221 + i)
-    plot_predictions(svm_clf, [-1.5, 2.5, -1, 1.5])
-    plot_dataset(X, y, [-1.5, 2.5, -1, 1.5])
-    gamma, C = hyperparams[i]
-    plt.title(r"$\gamma = {}, C = {}$".format(gamma, C), fontsize=16)
-
-plt.show()
-
    -











    +

    XGBoost: Extreme Gradient Boosting

    -

    Mathematical optimization of convex functions

    +

    XGBoost or Extreme Gradient +Boosting, is an optimized distributed gradient boosting library +designed to be highly efficient, flexible and portable. It implements +machine learning algorithms under the Gradient Boosting +framework. XGBoost provides a parallel tree boosting that solve many +data science problems in a fast and accurate way. See the article by Chen and Guestrin. +

    -

    -A mathematical (quadratic) optimization problem, or just optimization problem, has the form -$$ -\begin{align*} - &\mathrm{min}_{\lambda}\hspace{0.2cm} \frac{1}{2}\boldsymbol{\lambda}^T\boldsymbol{P}\boldsymbol{\lambda}+\boldsymbol{q}^T\boldsymbol{\lambda},\\ \nonumber - &\mathrm{subject\hspace{0.1cm}to} \hspace{0.2cm} \boldsymbol{G}\boldsymbol{\lambda} \preceq \boldsymbol{h} \wedge \boldsymbol{A}\boldsymbol{\lambda}=f. -\end{align*} -$$ +

    The authors design and build a highly scalable end-to-end tree +boosting system. It has a theoretically justified weighted quantile +sketch for efficient proposal calculation. It introduces a novel sparsity-aware algorithm for parallel tree learning and an effective cache-aware block structure for out-of-core tree learning. +

    -subject to some constraints for say a selected set \( i=1,2,\dots, n \). -In our case we are optimizing with respect to the Lagrangian multipliers \( \lambda_i \), and the -vector \( \boldsymbol{\lambda}=[\lambda_1, \lambda_2,\dots, \lambda_n] \) is the optimization variable we are dealing with. +

    It is now the algorithm which wins essentially all ML competitions!!!

    -

    -In our case we are particularly interested in a class of optimization problems called convex optmization problems. -In our discussion on gradient descent methods we discussed at length the definition of a convex function. - -

    -Convex optimization problems play a central role in applied mathematics and we recommend strongly Boyd and Vandenberghe's text on the topics. - -











    +

    Regression Case

    -

    How do we solve these problems?

    - -

    -If we use Python as programming language and wish to venture beyond -scikit-learn, tensorflow and similar software which makes our -lives so much easier, we need to dive into the wonderful world of -quadratic programming. We can, if we wish, solve the minimization -problem using standard gradient methods or conjugate gradient -methods. However, these methods tend to exhibit a rather slow -converge. So, welcome to the promised land of quadratic programming. - -

    -The functions we need are contained in the quadratic programming package CVXOPT and we need to import it together with numpy as - -

    -

    import numpy
-import cvxopt
-
    -

    -This will make our life much easier. You don't need to write your own optimizer. +

    +
    +
    +
    +
    + 
    import matplotlib.pyplot as plt
+import numpy as np
+from sklearn.model_selection import train_test_split
+import xgboost as xgb
+import scikitplot as skplt
+from sklearn.metrics import mean_squared_error
+
+n = 100
+maxdegree = 6
+
+# Make data set.
+x = np.linspace(-3, 3, n).reshape(-1, 1)
+y = np.exp(-x**2) + 1.5 * np.exp(-(x-2)**2)+ np.random.normal(0, 0.1, x.shape)
+
+error = np.zeros(maxdegree)
+bias = np.zeros(maxdegree)
+variance = np.zeros(maxdegree)
+polydegree = np.zeros(maxdegree)
+X_train, X_test, y_train, y_test = train_test_split(x, y, test_size=0.2)
+
+for degree in range(maxdegree):
+    model =  xgb.XGBRegressor(objective ='reg:squarederror', colsaobjective ='reg:squarederror', colsample_bytree = 0.3, learning_rate = 0.1,max_depth = degree, alpha = 10, n_estimators = 200)
+
+    model.fit(X_train,y_train)
+    y_pred = model.predict(X_test)
+    polydegree[degree] = degree
+    error[degree] = np.mean( np.mean((y_test - y_pred)**2) )
+    bias[degree] = np.mean( (y_test - np.mean(y_pred))**2 )
+    variance[degree] = np.mean( np.var(y_pred) )
+    print('Max depth:', degree)
+    print('Error:', error[degree])
+    print('Bias^2:', bias[degree])
+    print('Var:', variance[degree])
+    print('{} >= {} + {} = {}'.format(error[degree], bias[degree], variance[degree], bias[degree]+variance[degree]))
+
+plt.xlim(1,maxdegree-1)
+plt.plot(polydegree, error, label='Error')
+plt.plot(polydegree, bias, label='bias')
+plt.plot(polydegree, variance, label='Variance')
+plt.legend()
+plt.show()
+
    +
    +
    +
    +
    +
    +
    +
    +
    +
    +
    +
    +
    +
    + -











    +

    Xgboost on the Cancer Data

    -

    A simple example

    - -

    -We remind ourselves about the general problem we want to solve -$$ -\begin{align*} - &\mathrm{min}_{x}\hspace{0.2cm} \frac{1}{2}\boldsymbol{x}^T\boldsymbol{P}\boldsymbol{x}+\boldsymbol{q}^T\boldsymbol{x},\\ \nonumber - &\mathrm{s.t.} \hspace{0.2cm} \boldsymbol{G}\boldsymbol{x} \preceq \boldsymbol{h} \wedge \boldsymbol{A}\boldsymbol{x}=f. -\end{align*} -$$ - -Note: we use s.t. for subject to. - -

    -Let us show how to perform the optmization using a simple case. Assume we want to optimize the following problem -$$ -\begin{align*} - &\mathrm{min}_{x}\hspace{0.2cm} \frac{1}{2}x^2+5x+3y \\ \nonumber - &\mathrm{s.t.} \\ \nonumber - &x, y \geq 0 \\ \nonumber - &x+3y \geq 15 \\ \nonumber - &2x+5y \leq 100 \\ \nonumber - &3x+4y \leq 80. \\ \nonumber -\end{align*} -$$ - -The minimization problem can be rewritten in terms of vectors and matrices as (with \( x \) and \( y \) being the unknowns) -$$ -\frac{1}{2}\begin{bmatrix} x\\ y \end{bmatrix}^T \begin{bmatrix} 1 & 0\\ 0 & 0 \end{bmatrix} \begin{bmatrix} x \\ y \end{bmatrix} + \begin{bmatrix}5\\ 3 \end{bmatrix}^T \begin{bmatrix}x \\ y \end{bmatrix}. -$$ - -Similarly, we can now set up the inequalities (we need to change \( \geq \) to \( \leq \) by multiplying with \( -1 \) on bot sides) as the following matrix-vector equation -$$ -\begin{bmatrix} -1 & 0 \\ 0 & -1 \\ -1 & -3 \\ 2 & 5 \\ 3 & 4\end{bmatrix}\begin{bmatrix} x \\ y\end{bmatrix} \preceq \begin{bmatrix}0 \\ 0\\ -15 \\ 100 \\ 80\end{bmatrix}. -$$ - -We have collapsed all the inequalities into a single matrix \( \boldsymbol{G} \). We see also that our matrix -$$ -\boldsymbol{P} =\begin{bmatrix} 1 & 0\\ 0 & 0 \end{bmatrix} -$$ - -is clearly positive semi-definite (all eigenvalues larger or equal zero). -Finally, the vector \( \boldsymbol{h} \) is defined as -$$ -\boldsymbol{h} = \begin{bmatrix}0 \\ 0\\ -15 \\ 100 \\ 80\end{bmatrix}. -$$ - -

    -Since we don't have any equalities the matrix \( \boldsymbol{A} \) is set to zero -The following code solves the equations for us -

    +

    As you will see from the confusion matrix below, XGBoots does an excellent job on the Wisconsin cancer data and outperforms essentially all agorithms we have discussed till now.

    -
    # Import the necessary packages
-import numpy
-from cvxopt import matrix
-from cvxopt import solvers
-P = matrix(numpy.diag([1,0]), tc='d')
-q = matrix(numpy.array([3,4]), tc='d')
-G = matrix(numpy.array([[-1,0],[0,-1],[-1,-3],[2,5],[3,4]]), tc='d')
-h = matrix(numpy.array([0,0,-15,100,80]), tc='d')
-# Construct the QP, invoke solver
-sol = solvers.qp(P,q,G,h)
-# Extract optimal value and solution
-sol['x'] 
-sol['primal objective']
-
    -

    -









    +

    +
    +
    +
    +
    + 
    import matplotlib.pyplot as plt
+import numpy as np
+from sklearn.model_selection import  train_test_split 
+from sklearn.datasets import load_breast_cancer
+from sklearn.preprocessing import LabelEncoder
+from sklearn.model_selection import cross_validate
+import scikitplot as skplt
+import xgboost as xgb
+# Load the data
+cancer = load_breast_cancer()
 
-
    Back to the more realistic cases 
    
+X_train, X_test, y_train, y_test = train_test_split(cancer.data,cancer.target,random_state=0)
+print(X_train.shape)
+print(X_test.shape)
+#now scale the data
+from sklearn.preprocessing import StandardScaler
+scaler = StandardScaler()
+scaler.fit(X_train)
+X_train_scaled = scaler.transform(X_train)
+X_test_scaled = scaler.transform(X_test)
 
-
    
-We are now ready to return to our setup of the optmization problem for a more realistic case. Introducing the slack parameter \( C \) we have
-$$
-\frac{1}{2} \boldsymbol{\lambda}^T\begin{bmatrix} y_1y_1K(\boldsymbol{x}_1,\boldsymbol{x}_1) & y_1y_2K(\boldsymbol{x}_1,\boldsymbol{x}_2) & \dots & \dots & y_1y_nK(\boldsymbol{x}_1,\boldsymbol{x}_n) \\
-y_2y_1K(\boldsymbol{x}_2,\boldsymbol{x}_1) & y_2y_2K(\boldsymbol{x}_2,\boldsymbol{x}_2) & \dots & \dots & y_1y_nK(\boldsymbol{x}_2,\boldsymbol{x}_n) \\
-\dots & \dots & \dots & \dots & \dots \\
-\dots & \dots & \dots & \dots & \dots \\
-y_ny_1K(\boldsymbol{x}_n,\boldsymbol{x}_1) & y_ny_2K(\boldsymbol{x}_n\boldsymbol{x}_2) & \dots & \dots & y_ny_nK(\boldsymbol{x}_n,\boldsymbol{x}_n) \\
-\end{bmatrix}\boldsymbol{\lambda}-\mathbb{I}\boldsymbol{\lambda}, 
-$$
+xg_clf = xgb.XGBClassifier()
+xg_clf.fit(X_train_scaled,y_train)
 
-subject to \( \boldsymbol{y}^T\boldsymbol{\lambda}=0 \). Here we defined the vectors \( \boldsymbol{\lambda} =[\lambda_1,\lambda_2,\dots,\lambda_n] \) and 
-\( \boldsymbol{y}=[y_1,y_2,\dots,y_n] \). 
-With  the slack constants this leads to the additional constraint \( 0\leq \lambda_i \leq C \).
+y_test = xg_clf.predict(X_test_scaled)
 
-
    
-









    
+print("Test set accuracy with Gradient Boosting and scaled data: {:.2f}".format(xg_clf.score(X_test_scaled,y_test)))
 
-
    Setting up the matrices and the problem 
    
-
-
    
-We have the general problem 
-$$
-\begin{align*}
-    &\mathrm{min}_{\lambda}\hspace{0.2cm} \frac{1}{2}\boldsymbol{\lambda}^T\boldsymbol{P}\boldsymbol{\lambda}+\boldsymbol{q}^T\boldsymbol{\lambda},\\ \nonumber
-    &\mathrm{s.t.} \hspace{0.2cm} \boldsymbol{G}\boldsymbol{\lambda} \preceq \boldsymbol{h} \wedge  \boldsymbol{A}\boldsymbol{\lambda}=f.
-\end{align*}
-$$
+import scikitplot as skplt
+y_pred = xg_clf.predict(X_test_scaled)
+skplt.metrics.plot_confusion_matrix(y_test, y_pred, normalize=True)
+save_fig("xdclassiffierconfusion")
+plt.show()
+y_probas = xg_clf.predict_proba(X_test_scaled)
+skplt.metrics.plot_roc(y_test, y_probas)
+save_fig("xdclassiffierroc")
+plt.show()
+skplt.metrics.plot_cumulative_gain(y_test, y_probas)
+save_fig("gdclassiffiercgain")
+plt.show()
 
 
-
      
-
    1.  With a given kernel we can thus define the matrix \( \boldsymbol{P} \).
      2. 
-
    2.  The matrix \( \boldsymbol{P} \) has matrix elements \( p_{ij}=y_iy_jK(\boldsymbol{x}_i,\boldsymbol{x}_j) \). Given a kernel \( K \) and the targets \( y_i \) this matrix is easy to set up.
      3. 
-
    3.  The vector \( \boldsymbol{q} \) has all elements equal to \( -1 \).
      4. 
-
    4.  The constraint \( \boldsymbol{y}^T\boldsymbol{\lambda}=0 \) leads to \( f=0 \) and \( \boldsymbol{A}=\boldsymbol{y} \).
      5. 
-
    5.  To set up the matrix \( \boldsymbol{G} \) we note that the inequalities \( 0\leq \lambda_i \leq C \) can be split up into \( 0\leq \lambda_i \) and \( \lambda_i \leq C \). These two inequalities define then the matrix \( \boldsymbol{G} \) and the vector \( \boldsymbol{h} \).
      6. 
-
    
+xgb.plot_tree(xg_clf,num_trees=0)
+plt.rcParams['figure.figsize'] = [50, 10]
+save_fig("xgtree")
+plt.show()
+
+xgb.plot_importance(xg_clf)
+plt.rcParams['figure.figsize'] = [5, 5]
+save_fig("xgparams")
+plt.show()
+
    +
    +
    +
    +
    +
    +
    +
    +
    +
    +
    +
    +
    +
    +









    +

    Gradient boosting, making our own code for a regression case

    -

    Setting up \( \boldsymbol{G}\boldsymbol{\lambda} \preceq \boldsymbol{h} \)

    -

    -We have two constraints, \( 0\le \lambda_i \) and \( \lambda_i \le C \). To do this we multiply the ones with the constraint -\( \ge \) with \( -1 \) in order to get \( \le \). It means that the problem \( \boldsymbol{G}\boldsymbol{\lambda} \preceq \boldsymbol{h} \) -can be written as -$$ -\begin{bmatrix} -1& 0 & 0 & \dots & 0 \\ -0& -1 & 0 & \dots & 0 \\ -0& 0 & -1 & \dots & 0 \\ -\dots& \dots & \dots & \dots & \dots \\ -0& 0 & 0 & \dots & -1 \\ -1 & 0 & 0 & \dots & 0 \\ -0& 1 & 0 & \dots & 0 \\ -0& 0 & 1 & \dots & 0 \\ -\dots& \dots & \dots & \dots & \dots \\ -0& 0 & 0 & \dots & 1 \\ -\end{bmatrix} -\begin{bmatrix} \lambda_1 \\ -\lambda_2 \\ -\lambda_3 \\ -\dots \\ -\lambda_n \\ -\end{bmatrix} \preceq -\begin{bmatrix} 0 \\ -0 \\ -0 \\ -\dots \\ -0 \\ -C \\ -C \\ -C \\ -\dots \\ -C \\ -\end{bmatrix}. -$$ + +

    +
    +
    +
    +
    + 
    import numpy as np
+class DecisionTreeRegressor:
+    def __init__(self, max_depth=3):
+        self.max_depth = max_depth
+        self.tree = None
+    def fit(self, X, y):
+        self.tree = self._grow_tree(X, y)
+    def _grow_tree(self, X, y, depth=0):
+        n_samples, n_features = X.shape
+        if depth < self.max_depth:
+            best_feature, best_threshold = self._best_split(X, y)
+            if best_feature is not None:
+                left_indices = X[:, best_feature] < best_threshold
+                right_indices = X[:, best_feature] >= best_threshold
+                left_child = self._grow_tree(X[left_indices], y[left_indices], depth + 1)
+                right_child = self._grow_tree(X[right_indices], y[right_indices], depth + 1)
+                return (best_feature, best_threshold, left_child, right_child)
+        return np.mean(y)
+    def _best_split(self, X, y):
+        best_mse = float('inf')
+        best_feature, best_threshold = None, None
+        n_samples, n_features = X.shape
+        
+        for feature in range(n_features):
+            thresholds = np.unique(X[:, feature])
+            for threshold in thresholds:
+                left_indices = X[:, feature] < threshold
+                right_indices = X[:, feature] >= threshold
+                if len(y[left_indices]) > 0 and len(y[right_indices]) > 0:
+                    left_mse = np.mean((y[left_indices] - np.mean(y[left_indices])) ** 2)
+                    right_mse = np.mean((y[right_indices] - np.mean(y[right_indices])) ** 2)
+                    mse = (len(y[left_indices]) * left_mse + len(y[right_indices]) * right_mse) / n_samples
+                    
+                    if mse < best_mse:
+                        best_mse = mse
+                        best_feature = feature
+                        best_threshold = threshold
+        return best_feature, best_threshold
+    def predict(self, X):
+        return np.array([self._predict_sample(sample, self.tree) for sample in X])
+    def _predict_sample(self, sample, node):
+        if isinstance(node, tuple):
+            feature, threshold, left_child, right_child = node
+            if sample[feature] < threshold:
+                return self._predict_sample(sample, left_child)
+            else:
+                return self._predict_sample(sample, right_child)
+        return node
+class GradientBoostingRegressor:
+    def __init__(self, n_estimators=100, learning_rate=0.1, max_depth=3):
+        self.n_estimators = n_estimators
+        self.learning_rate = learning_rate
+        self.max_depth = max_depth
+        self.models = []
+    def fit(self, X, y):
+        y_pred = np.zeros(y.shape)
+        for _ in range(self.n_estimators):
+            residuals = y - y_pred
+            model = DecisionTreeRegressor(max_depth=self.max_depth)
+            model.fit(X, residuals)
+            y_pred += self.learning_rate * model.predict(X)
+            self.models.append(model)
+    def predict(self, X):
+        y_pred = np.zeros(X.shape[0])
+        for model in self.models:
+            y_pred += self.learning_rate * model.predict(X)
+        return y_pred
+# Example usage
+if __name__ == "__main__":
+    # Sample data
+    X = np.array([[1], [2], [3], [4], [5]])
+    y = np.array([1.5, 1.7, 3.5, 3.7, 5.0])
+    model = GradientBoostingRegressor(n_estimators=100, learning_rate=0.1, max_depth=2)
+    model.fit(X, y)
+    predictions = model.predict(X)
+    print("Predictions:", predictions)
+
    +
    +
    +
    +
    +
    +
    +
    +
    +
    +
    +
    +
    +
    -

    -And then we are ready to go. -











    +

    Summary of course

    -

    SVMs and Regression and multiclass classification

    - -

    -More text will be added here. See also Hastie et al. section 12.3. - -











    +

    What? Me worry? No final exam in this course!

    +

    +
    +

    +
    +

    -

    Summary of course

    - -











    +

    Topics we have covered this year

    -

    What? Me worry? No final exam in this course!

    -



    - -

    -









    - -

    Topics we have covered this year

    - -

    -The course has two central parts +

    The course has two central parts

    1. Statistical analysis and optimization of data
    2. Machine learning
    -









    +

    Statistical analysis and optimization of data

    -

    Statistical analysis and optimization of data

    - -

    -The following topics have been discussed: - +

    The following topics have been discussed:

    1. Basic concepts, expectation values, variance, covariance, correlation functions and errors;
    2. Simpler models, binomial distribution, the Poisson distribution, simple and multivariate normal distributions;
      3. @@ -925,57 +1268,42 @@ The following topics have been discussed:
    3. Practical optimization using Singular-value decomposition and least squares for parameterizing data.
    4. Principal Component Analysis to reduce the number of features.
    -









    +

    Machine learning

    -

    Machine learning

    - -

    -The following topics will be covered - +

    The following topics will be covered

    1. Linear methods for regression and classification: -
      1. Ordinary Least Squares
      2. Ridge regression
      3. Lasso regression
      4. Logistic regression
      -
    2. Neural networks and deep learning: -
      1. Feed Forward Neural Networks
      2. Convolutional Neural Networks
      3. Recurrent Neural Networks
      -
    3. Decisions trees and ensemble methods: -
      1. Decision trees
      2. Bagging and voting
      3. Random forests
      4. Boosting and gradient boosting
      -
    4. Support vector machines -
      1. Binary classification and multiclass classification
      2. Kernel methods
      3. Regression
      -
    -









    +

    Learning outcomes and overarching aims of this course

    -

    Learning outcomes and overarching aims of this course

    - -

    -The course introduces a variety of central algorithms and methods +

    The course introduces a variety of central algorithms and methods essential for studies of data analysis and machine learning. The course is project based and through the various projects, normally three, you will be exposed to fundamental research problems @@ -984,6 +1312,7 @@ results. The students will learn to develop and structure large codes for studying these systems, get acquainted with computing facilities and learn to handle large scientific projects. A good scientific and ethical conduct is emphasized throughout the course. +

    -









    - -

    Perspective on Machine Learning

    +

    Perspective on Machine Learning

    1. Rapidly emerging application area
      2. @@ -1006,20 +1333,16 @@ ethical conduct is emphasized throughout the course.
    2. Requires education/retraining for more widespread adoption
    3. A lot of “word-of-mouth” development methods
    - -Huge amounts of data sets require automation, classical analysis tools often inadequate. +

    Huge amounts of data sets require automation, classical analysis tools often inadequate. High energy physics hit this wall in the 90’s. In 2009 single top quark production was determined via Boosted decision trees, Bayesian Neural Networks, etc. +

    -











    +

    Machine Learning Research

    -

    Machine Learning Research

    - -

    -Where to find recent results: - +

    Where to find recent results:

    1. Conference proceedings, arXiv and blog posts!
    2. NIPS: Neural Information Processing Systems
      3. @@ -1028,10 +1351,8 @@ Where to find recent results:
    3. Journal of Machine Learning Research
    4. Follow ML on ArXiv
    -









    - -

    Starting your Machine Learning Project

    +

    Starting your Machine Learning Project

    1. Identify problem type: classification, regression
      2. @@ -1040,111 +1361,82 @@ Where to find recent results:
    2. Consider your data carefully again! Think of data representation more carefully.
    3. Based on your results, feedback loop to earliest possible point
    -









    - -

    Choose a Model and Algorithm

    +

    Choose a Model and Algorithm

    1. Supervised?
    2. Start with the simplest model that fits your problem
    3. Start with minimal processing of data
    -









    - -

    Preparing Your Data

    +

    Preparing Your Data

    1. Shuffle your data
    2. Mean center your data
      3. -
      • Why?
      -
    3. Normalize the variance
      4. -
      • Why?
      -
    4. Whitening
      5. -
      • Decorrelates data
      • Can be hit or miss
      -
    5. When to do train/test split?
    -









    - -

    Which Activation and Weights to Choose in Neural Networks

    +

    Which Activation and Weights to Choose in Neural Networks

    1. RELU? ELU?
    2. Sigmoid or Tanh?
    3. Set all weights to 0?
      4. -
      • Terrible idea
      -
    4. Set all weights to random values?
      5. -
      • Small random values
      -
    -









    - -

    Optimization Methods and Hyperparameters

    - +

    Optimization Methods and Hyperparameters

    1. Stochastic gradient descent -
      1. Stochastic gradient descent + momentum
      -
    2. State-of-the-art approaches:
      3. -
      • RMSProp
      • Adam
      • and more
      -
    - -Which regularization and hyperparameters? \( L_1 \) or \( L_2 \), soft +

    Which regularization and hyperparameters? \( L_1 \) or \( L_2 \), soft classifiers, depths of trees and many other. Need to explore a large set of hyperparameters and regularization methods. +

    -











    +

    Resampling

    -

    Resampling

    - -

    -When do we resample? +

    When do we resample?

    1. Bootstrap
    2. Cross-validation
    3. Jackknife and many other
    -









    +

    Other courses on Data science and Machine Learning at UiO

    -

    Other courses on Data science and Machine Learning at UiO

    - -

    -The link here https://www.mn.uio.no/english/research/about/centre-focus/innovation/data-science/studies/ gives an excellent overview of courses on Machine learning at UiO. +

    The link here https://www.mn.uio.no/english/research/about/centre-focus/innovation/data-science/studies/ gives an excellent overview of courses on Machine learning at UiO.

    1. STK2100 Machine learning and statistical methods for prediction and classification.
      2. @@ -1156,58 +1448,51 @@ The link here IN5400/INF5860 – Machine Learning for Image Analysis. An introduction to deep learning with particular emphasis on applications within Image analysis, but useful for other application areas too.
    2. TEK5040 – Dyp læring for autonome systemer. The course addresses advanced algorithms and architectures for deep learning with neural networks. The course provides an introduction to how deep-learning techniques can be used in the construction of key parts of advanced autonomous systems that exist in physical environments and cyber environments.
    -









    - -

    Additional courses of interest

    +

    Additional courses of interest

    1. STK4051 Computational Statistics
    2. STK4021 Applied Bayesian Analysis and Numerical Methods
    -









    +

    What's the future like?

    -

    What's the future like?

    - -

    -Based on multi-layer nonlinear neural networks, deep learning can +

    Based on multi-layer nonlinear neural networks, deep learning can learn directly from raw data, automatically extract and abstract features from layer to layer, and then achieve the goal of regression, classification, or ranking. Deep learning has made breakthroughs in computer vision, speech processing and natural language, and reached or even surpassed human level. The success of deep learning is mainly due to the three factors: big data, big model, and big computing. +

    -

    -In the past few decades, many different architectures of deep neural +

    In the past few decades, many different architectures of deep neural networks have been proposed, such as - +

    1. Convolutional neural networks, which are mostly used in image and video data processing, and have also been applied to sequential data such as text processing;
    2. Recurrent neural networks, which can process sequential data of variable length and have been widely used in natural language understanding and speech processing;
    3. Encoder-decoder framework, which is mostly used for image or sequence generation, such as machine translation, text summarization, and image captioning.
    -









    +

    Types of Machine Learning, a repetition

    -

    Types of Machine Learning, a repetition

    - -

    -The approaches to machine learning are many, but are often split into two main categories. +

    The approaches to machine learning are many, but are often split into two main categories. In supervised learning we know the answer to a problem, and let the computer deduce the logic behind it. On the other hand, unsupervised learning is a method for finding patterns and relationship in data sets without any prior knowledge of the system. Some authours also operate with a third category, namely reinforcement learning. This is a paradigm of learning inspired by behavioural psychology, where learning is achieved by trial-and-error, solely from rewards and punishment. +

    -

    -Another way to categorize machine learning tasks is to consider the desired output of a system. +

    Another way to categorize machine learning tasks is to consider the desired output of a system. Some of the most common tasks are: +

    • Classification: Outputs are divided into two or more classes. The goal is to produce a model that assigns inputs into one of these classes. An example is to identify digits based on pictures of hand-written ones. Classification is typically supervised learning.
      • @@ -1218,166 +1503,139 @@ Some of the most common tasks are:
    -











    +

    Why Boltzmann machines?

    -

    Why Boltzmann machines?

    - -

    -What is known as restricted Boltzmann Machines (RMB) have received a lot of attention lately. +

    What is known as restricted Boltzmann Machines (RMB) have received a lot of attention lately. One of the major reasons is that they can be stacked layer-wise to build deep neural networks that capture complicated statistics. +

    -

    -The original RBMs had just one visible layer and a hidden layer, but recently so-called Gaussian-binary RBMs have gained quite some popularity in imaging since they are capable of modeling continuous data that are common to natural images. +

    The original RBMs had just one visible layer and a hidden layer, but recently so-called Gaussian-binary RBMs have gained quite some popularity in imaging since they are capable of modeling continuous data that are common to natural images.

    -

    -Furthermore, they have been used to solve complicated quantum mechanical many-particle problems or classical statistical physics problems like the Ising and Potts classes of models. +

    Furthermore, they have been used to solve complicated quantum mechanical many-particle problems or classical statistical physics problems like the Ising and Potts classes of models.

    -











    +

    Boltzmann Machines

    -

    Boltzmann Machines

    - -

    -Why use a generative model rather than the more well known discriminative deep neural networks (DNN)? +

    Why use a generative model rather than the more well known discriminative deep neural networks (DNN)?

    -









    - -

    Some similarities and differences from DNNs

    +

    Some similarities and differences from DNNs

    1. Both use gradient-descent based learning procedures for minimizing cost functions
    2. Energy based models don't use backpropagation and automatic differentiation for computing gradients, instead turning to Markov Chain Monte Carlo methods.
    3. DNNs often have several hidden layers. A restricted Boltzmann machine has only one hidden layer, however several RBMs can be stacked to make up Deep Belief Networks, of which they constitute the building blocks.
    +

    History: The RBM was developed by amongst others Geoffrey Hinton, called by some the "Godfather of Deep Learning", working with the University of Toronto and Google.

    -History: The RBM was developed by amongst others Geoffrey Hinton, called by some the "Godfather of Deep Learning", working with the University of Toronto and Google. - -











    +

    Boltzmann machines (BM)

    -

    Boltzmann machines (BM)

    - -

    -A BM is what we would call an undirected probabilistic graphical model +

    A BM is what we would call an undirected probabilistic graphical model with stochastic continuous or discrete units. +

    -It is interpreted as a stochastic recurrent neural network where the +

    It is interpreted as a stochastic recurrent neural network where the state of each unit(neurons/nodes) depends on the units it is connected to. The weights in the network represent thus the strength of the interaction between various units/nodes. +

    -It turns into a Hopfield network if we choose deterministic rather +

    It turns into a Hopfield network if we choose deterministic rather than stochastic units. In contrast to a Hopfield network, a BM is a so-called generative model. It allows us to generate new samples from the learned distribution. +

    -











    +

    A standard BM setup

    -

    A standard BM setup

    - -

    -A standard BM network is divided into a set of observable and visible units \( \hat{x} \) and a set of unknown hidden units/nodes \( \hat{h} \). +

    A standard BM network is divided into a set of observable and visible units \( \hat{x} \) and a set of unknown hidden units/nodes \( \hat{h} \).

    -

    -Additionally there can be bias nodes for the hidden and visible layers. These biases are normally set to \( 1 \). +

    Additionally there can be bias nodes for the hidden and visible layers. These biases are normally set to \( 1 \).

    -

    -BMs are stackable, meaning they cwe can train a BM which serves as input to another BM. We can construct deep networks for learning complex PDFs. The layers can be trained one after another, a feature which makes them popular in deep learning +

    BMs are stackable, meaning they cwe can train a BM which serves as input to another BM. We can construct deep networks for learning complex PDFs. The layers can be trained one after another, a feature which makes them popular in deep learning

    -

    -However, they are often hard to train. This leads to the introduction of so-called restricted BMs, or RBMS. +

    However, they are often hard to train. This leads to the introduction of so-called restricted BMs, or RBMS. Here we take away all lateral connections between nodes in the visible layer as well as connections between nodes in the hidden layer. The network is illustrated in the figure below. +

    -











    +

    The structure of the RBM network

    -

    The structure of the RBM network

    +

    +
    +

    +
    +

    -

    -



    - -











    +

    The network

    -

    The network

    - -

    The network layers: -

    1. A function \( \mathbf{x} \) that represents the visible layer, a vector of \( M \) elements (nodes). This layer represents both what the RBM might be given as training input, and what we want it to be able to reconstruct. This might for example be given by the pixels of an image or coefficients representing speech, or the coordinates of a quantum mechanical state function.
    2. The function \( \mathbf{h} \) represents the hidden, or latent, layer. A vector of \( N \) elements (nodes). Also called "feature detectors".
    -









    +

    Goals

    -

    Goals

    - -

    -The goal of the hidden layer is to increase the model's expressive +

    The goal of the hidden layer is to increase the model's expressive power. We encode complex interactions between visible variables by introducing additional, hidden variables that interact with visible degrees of freedom in a simple manner, yet still reproduce the complex correlations between visible degrees in the data once marginalized over (integrated out). +

    -

    The network parameters, to be optimized/learned: -

    1. \( \mathbf{a} \) represents the visible bias, a vector of same length as \( \mathbf{x} \).
    2. \( \mathbf{b} \) represents the hidden bias, a vector of same lenght as \( \mathbf{h} \).
    3. \( W \) represents the interaction weights, a matrix of size \( M\times N \).
    -









    +

    Joint distribution

    -

    Joint distribution

    - -

    -The restricted Boltzmann machine is described by a Boltzmann distribution +

    The restricted Boltzmann machine is described by a Boltzmann distribution

    $$ \begin{align} P_{rbm}(\mathbf{x},\mathbf{h}) = \frac{1}{Z} e^{-\frac{1}{T_0}E(\mathbf{x},\mathbf{h})}, @@ -1385,7 +1643,7 @@ $$ \end{align} $$ -where \( Z \) is the normalization constant or partition function, defined as +

    where \( Z \) is the normalization constant or partition function, defined as

    $$ \begin{align} Z = \int \int e^{-\frac{1}{T_0}E(\mathbf{x},\mathbf{h})} d\mathbf{x} d\mathbf{h}. @@ -1393,49 +1651,39 @@ $$ \end{align} $$ -It is common to ignore \( T_0 \) by setting it to one. +

    It is common to ignore \( T_0 \) by setting it to one.

    -











    +

    Network Elements, the energy function

    -

    Network Elements, the energy function

    - -

    -The function \( E(\mathbf{x},\mathbf{h}) \) gives the energy of a +

    The function \( E(\mathbf{x},\mathbf{h}) \) gives the energy of a configuration (pair of vectors) \( (\mathbf{x}, \mathbf{h}) \). The lower the energy of a configuration, the higher the probability of it. This function also depends on the parameters \( \mathbf{a} \), \( \mathbf{b} \) and \( W \). Thus, when we adjust them during the learning procedure, we are adjusting the energy function to best fit our problem. +

    -

    -An expression for the energy function is +

    An expression for the energy function is

    $$ E(\hat{x},\hat{h}) = -\sum_{ia}^{NA}b_i^a \alpha_i^a(x_i)-\sum_{jd}^{MD}c_j^d \beta_j^d(h_j)-\sum_{ijad}^{NAMD}b_i^a \alpha_i^a(x_i)c_j^d \beta_j^d(h_j)w_{ij}^{ad}. $$ -

    -Here \( \beta_j^d(h_j) \) and \( \alpha_i^a(x_j) \) are so-called transfer functions that map a given input value to a desired feature value. The labels \( a \) and \( d \) denote that there can be multiple transfer functions per variable. The first sum depends only on the visible units. The second on the hidden ones. Note that there is no connection between nodes in a layer. +

    Here \( \beta_j^d(h_j) \) and \( \alpha_i^a(x_j) \) are so-called transfer functions that map a given input value to a desired feature value. The labels \( a \) and \( d \) denote that there can be multiple transfer functions per variable. The first sum depends only on the visible units. The second on the hidden ones. Note that there is no connection between nodes in a layer.

    -

    -The quantities \( b \) and \( c \) can be interpreted as the visible and hidden biases, respectively. +

    The quantities \( b \) and \( c \) can be interpreted as the visible and hidden biases, respectively.

    -

    -The connection between the nodes in the two layers is given by the weights \( w_{ij} \). +

    The connection between the nodes in the two layers is given by the weights \( w_{ij} \).

    -











    +

    Defining different types of RBMs

    +

    There are different variants of RBMs, and the differences lie in the types of visible and hidden units we choose as well as in the implementation of the energy function \( E(\mathbf{x},\mathbf{h}) \).

    -

    Defining different types of RBMs

    -There are different variants of RBMs, and the differences lie in the types of visible and hidden units we choose as well as in the implementation of the energy function \( E(\mathbf{x},\mathbf{h}) \). - -

    Binary-Binary RBM:

    -

    -RBMs were first developed using binary units in both the visible and hidden layer. The corresponding energy function is defined as follows: +

    RBMs were first developed using binary units in both the visible and hidden layer. The corresponding energy function is defined as follows:

    $$ \begin{align} E(\mathbf{x}, \mathbf{h}) = - \sum_i^M x_i a_i- \sum_j^N b_j h_j - \sum_{i,j}^{M,N} x_i w_{ij} h_j, @@ -1443,15 +1691,14 @@ $$ \end{align} $$ -where the binary values taken on by the nodes are most commonly 0 and 1. +

    where the binary values taken on by the nodes are most commonly 0 and 1.

    Gaussian-Binary RBM:

    -

    -Another varient is the RBM where the visible units are Gaussian while the hidden units remain binary: +

    Another varient is the RBM where the visible units are Gaussian while the hidden units remain binary:

    $$ \begin{align} E(\mathbf{x}, \mathbf{h}) = \sum_i^M \frac{(x_i - a_i)^2}{2\sigma_i^2} - \sum_j^N b_j h_j - \sum_{i,j}^{M,N} \frac{x_i w_{ij} h_j}{\sigma_i^2}. @@ -1461,45 +1708,36 @@ $$
    -











    - -

    More about RBMs

    - +

    More about RBMs

    1. Useful when we model continuous data (i.e., we wish \( \mathbf{x} \) to be continuous)
    2. Requires a smaller learning rate, since there's no upper bound to the value a component might take in the reconstruction
    - -Other types of units include: - +

    Other types of units include:

    1. Softmax and multinomial units
    2. Gaussian visible and hidden units
    3. Binomial units
    4. Rectified linear units
    +

    To read more, see Lectures on Boltzmann machines in Physics.

    -To read more, see Lectures on Boltzmann machines in Physics. - -











    +

    Autoencoders: Overarching view

    -

    Autoencoders: Overarching view

    - -

    -Autoencoders are artificial neural networks capable of learning +

    Autoencoders are artificial neural networks capable of learning efficient representations of the input data (these representations are called codings) without any supervision (i.e., the training set is unlabeled). These codings typically have a much lower dimensionality than the input data, making -autoencoders useful for dimensionality reduction. +autoencoders useful for dimensionality reduction. +

    -

    -More importantly, autoencoders act as powerful feature detectors, and +

    More importantly, autoencoders act as powerful feature detectors, and they can be used for unsupervised pretraining of deep neural networks. +

    -

    -Lastly, they are capable of randomly generating new data that looks +

    Lastly, they are capable of randomly generating new data that looks very similar to the training data; this is called a generative model. For example, you could train an autoencoder on pictures of faces, and it would then be able to generate new faces. Surprisingly, @@ -1513,57 +1751,49 @@ autoencoder from trivially copying the inputs directly to the outputs, which forces it to learn efficient ways of representing the data. In short, the codings are byproducts of the autoencoder’s attempt to learn the identity function under some constraints. +

    -

    Video on autoencoders -

    -See also A. Geron's textbook, chapter 15. +

    See also A. Geron's textbook, chapter 15.

    -











    +

    Bayesian Machine Learning

    -

    Bayesian Machine Learning

    - -

    -This is an important topic if we aim at extracting a probability +

    This is an important topic if we aim at extracting a probability distribution. This gives us also a confidence interval and error estimates. +

    -

    -Bayesian machine learning allows us to encode our prior beliefs about +

    Bayesian machine learning allows us to encode our prior beliefs about what those models should look like, independent of what the data tells us. This is especially useful when we don’t have a ton of data to confidently learn our model. +

    -

    Video on Bayesian deep learning -

    -See also the slides here. +

    See also the slides here.

    -











    +

    Reinforcement Learning

    -

    Reinforcement Learning

    - -

    -Reinforcement Learning (RL) is one of the most exciting fields of +

    Reinforcement Learning (RL) is one of the most exciting fields of Machine Learning today, and also one of the oldest. It has been around since the 1950s, producing many interesting applications over the years. +

    -

    -It studies +

    It studies how agents take actions based on trial and error, so as to maximize some notion of cumulative reward in a dynamic system or environment. Due to its generality, the problem has also been studied in many other disciplines, such as game theory, control theory, operations research, information theory, multi-agent systems, swarm intelligence, statistics, and genetic algorithms. +

    -

    -In March 2016, AlphaGo, a computer program that plays the board game +

    In March 2016, AlphaGo, a computer program that plays the board game Go, beat Lee Sedol in a five-game match. This was the first time a computer Go program had beaten a 9-dan (highest rank) professional without handicaps. AlphaGo is based on deep convolutional neural @@ -1571,20 +1801,16 @@ networks and reinforcement learning. AlphaGo’s victory was a major milestone in artificial intelligence and it has also made reinforcement learning a hot research area in the field of machine learning. +

    -

    -Lecture on Reinforcement Learning. +

    Lecture on Reinforcement Learning.

    -

    -See also A. Geron's textbook, chapter 16. +

    See also A. Geron's textbook, chapter 16.

    -











    +

    Transfer learning

    -

    Transfer learning

    - -

    -The goal of transfer learning is to transfer the model or knowledge +

    The goal of transfer learning is to transfer the model or knowledge obtained from a source task to the target task, in order to resolve the issues of insufficient training data in the target task. The rationality of doing so lies in that usually the source and target @@ -1592,17 +1818,14 @@ tasks have inter-correlations, and therefore either the features, samples, or models in the source task might provide useful information for us to better solve the target task. Transfer learning is a hot research topic in recent years, with many problems still waiting to be studied. +

    -

    -Lecture on transfer learning. +

    Lecture on transfer learning.

    -











    +

    Adversarial learning

    -

    Adversarial learning

    - -

    -The conventional deep generative model has a potential problem: the +

    The conventional deep generative model has a potential problem: the model tends to generate extreme instances to maximize the probabilistic likelihood, which will hurt its performance. Adversarial learning utilizes the adversarial behaviors (e.g., generating @@ -1611,17 +1834,14 @@ robustness of the model and improve the quality of the generated data. In recent years, one of the most promising unsupervised learning technologies, generative adversarial networks (GAN), has already been successfully applied to image, speech, and text. +

    -

    -Lecture on adversial learning. +

    Lecture on adversial learning.

    -











    +

    Dual learning

    -

    Dual learning

    - -

    -Dual learning is a new learning paradigm, the basic idea of which is +

    Dual learning is a new learning paradigm, the basic idea of which is to use the primal-dual structure between machine learning tasks to obtain effective feedback/regularization, and guide and strengthen the learning process, thus reducing the requirement of large-scale labeled @@ -1630,42 +1850,35 @@ many problems in machine learning, including machine translation, image style conversion, question answering and generation, image classification and generation, text classification and generation, image-to-text, and text-to-image. +

    -











    +

    Distributed machine learning

    -

    Distributed machine learning

    - -

    -Distributed computation will speed up machine learning algorithms, +

    Distributed computation will speed up machine learning algorithms, significantly improve their efficiency, and thus enlarge their application. When distributed meets machine learning, more than just implementing the machine learning algorithms in parallel is required. +

    -











    +

    Meta learning

    -

    Meta learning

    - -

    -Meta learning is an emerging research direction in machine +

    Meta learning is an emerging research direction in machine learning. Roughly speaking, meta learning concerns learning how to learn, and focuses on the understanding and adaptation of the learning itself, instead of just completing a specific learning task. That is, a meta learner needs to be able to evaluate its own learning methods and adjust its own learning methods according to specific learning tasks. +

    -











    +

    The Challenges Facing Machine Learning

    -

    The Challenges Facing Machine Learning

    +

    While there has been much progress in machine learning, there are also challenges.

    -

    -While there has been much progress in machine learning, there are also challenges. - -

    -For example, the mainstream machine learning technologies are +

    For example, the mainstream machine learning technologies are black-box approaches, making us concerned about their potential risks. To tackle this challenge, we may want to make machine learning more explainable and controllable. As another example, the @@ -1680,27 +1893,23 @@ models for data fitting? Although there are many challenges, we are still very optimistic about the future of machine learning. As we look forward to the future, here are what we think the research hotspots in the next ten years will be. +

    -

    -See the article on Discovery of Physics From Data: Universal Laws and Discrepancies +

    See the article on Discovery of Physics From Data: Universal Laws and Discrepancies

    -











    +

    Explainable machine learning

    -

    Explainable machine learning

    - -

    -Machine learning, especially deep learning, evolves rapidly. The +

    Machine learning, especially deep learning, evolves rapidly. The ability gap between machine and human on many complex cognitive tasks becomes narrower and narrower. However, we are still in the very early stage in terms of explaining why those effective models work and how they work. +

    -

    -What is missing: the gap between correlation and causation. Standard Machine Learning is based on what e have called a frequentist approach. +

    What is missing: the gap between correlation and causation. Standard Machine Learning is based on what e have called a frequentist approach.

    -

    -Most +

    Most machine learning techniques, especially the statistical ones, depend highly on correlations in data sets to make predictions and analyses. In contrast, rational humans tend to reply on clear and trustworthy @@ -1708,48 +1917,41 @@ causality relations obtained via logical reasoning on real and clear facts. It is one of the core goals of explainable machine learning to transition from solving problems by data correlation to solving problems by logical reasoning. +

    -

    Bayesian Machine Learning is one of the exciting research directions in this field. -











    +

    Quantum machine learning

    -

    Quantum machine learning

    - -

    -Quantum machine learning is an emerging interdisciplinary research +

    Quantum machine learning is an emerging interdisciplinary research area at the intersection of quantum computing and machine learning. +

    -

    -Quantum computers use effects such as quantum coherence and quantum +

    Quantum computers use effects such as quantum coherence and quantum entanglement to process information, which is fundamentally different from classical computers. Quantum algorithms have surpassed the best classical algorithms in several problems (e.g., searching for an unsorted database, inverting a sparse matrix), which we call quantum acceleration. +

    -

    -When quantum computing meets machine learning, it can be a mutually +

    When quantum computing meets machine learning, it can be a mutually beneficial and reinforcing process, as it allows us to take advantage of quantum computing to improve the performance of classical machine learning algorithms. In addition, we can also use the machine learning algorithms (on classic computers) to analyze and improve quantum computing systems. +

    -

    -Lecture on Quantum ML. +

    Lecture on Quantum ML.

    -

    -Read interview with Maria Schuld on her work on Quantum Machine Learning. See also her recent textbook. +

    Read interview with Maria Schuld on her work on Quantum Machine Learning. See also her recent textbook.

    -











    +

    Quantum machine learning algorithms based on linear algebra

    -

    Quantum machine learning algorithms based on linear algebra

    - -

    -Many quantum machine learning algorithms are based on variants of +

    Many quantum machine learning algorithms are based on variants of quantum algorithms for solving linear equations, which can efficiently solve N-variable linear equations with complexity of O(log2 N) under certain conditions. The quantum matrix inversion algorithm can @@ -1761,28 +1963,24 @@ machine learning algorithms is data input—that is, how to initialize the quantum system with the entire data set. Although efficient data-input algorithms exist for certain situations, how to efficiently input data into a quantum system is as yet unknown for most cases. +

    -











    +

    Quantum reinforcement learning

    -

    Quantum reinforcement learning

    - -

    -In quantum reinforcement learning, a quantum agent interacts with the +

    In quantum reinforcement learning, a quantum agent interacts with the classical environment to obtain rewards from the environment, so as to adjust and improve its behavioral strategies. In some cases, it achieves quantum acceleration by the quantum processing capabilities of the agent or the possibility of exploring the environment through quantum superposition. Such algorithms have been proposed in superconducting circuits and systems of trapped ions. +

    -











    +

    Quantum deep learning

    -

    Quantum deep learning

    - -

    -Dedicated quantum information processors, such as quantum annealers +

    Dedicated quantum information processors, such as quantum annealers and programmable photonic circuits, are well suited for building deep quantum networks. The simplest deep quantum network is the Boltzmann machine. The classical Boltzmann machine consists of bits with tunable @@ -1794,14 +1992,12 @@ correspond to an adjustable Ising model. Then, by initializing the input neurons in the Boltzmann machine to a fixed state and allowing the system to heat up, we can read out the output qubits to get the result. +

    -











    +

    Social machine learning

    -

    Social machine learning

    - -

    -Machine learning aims to imitate how humans +

    Machine learning aims to imitate how humans learn. While we have developed successful machine learning algorithms, until now we have ignored one important fact: humans are social. Each of us is one part of the total society and it is difficult for us to @@ -1809,41 +2005,33 @@ live, learn, and improve ourselves, alone and isolated. Therefore, we should design machines with social properties. Can we let machines evolve by imitating human society so as to achieve more effective, intelligent, interpretable “social machine learning”? +

    -

    -And much more. +

    And much more.

    -











    +

    The last words?

    -

    The last words?

    - -

    -Early computer scientist Alan Kay said, The best way to predict the +

    Early computer scientist Alan Kay said, The best way to predict the future is to create it. Therefore, all machine learning practitioners, whether scholars or engineers, professors or students, need to work together to advance these important research topics. Together, we will not just predict the future, but create it. +

    -











    +

    Best wishes to you all and thanks so much for your heroic efforts this semester

    -

    Best wishes to you all and thanks so much for your heroic efforts this semester

    - -

    -



    - -

    +

    +

    +

    +
    +

    - -
    - © 1999-2020, Morten Hjorth-Jensen. Released under CC Attribution-NonCommercial 4.0 license + © 1999-2024, Morten Hjorth-Jensen. Released under CC Attribution-NonCommercial 4.0 license
    - - - diff --git a/doc/pub/week48/html/week48.html b/doc/pub/week48/html/week48.html index c375cc5f9..6004b819b 100644 --- a/doc/pub/week48/html/week48.html +++ b/doc/pub/week48/html/week48.html @@ -1,36 +1,105 @@ - + - Week 48: Support Vector Machines and Summary of course - - - - @@ -191,736 +354,988 @@ MathJax.Hub.Config({ - - - - - -

    Week 48: Support Vector Machines and Summary of course

    - -

    - -

    -Morten Hjorth-Jensen [1, 2] +

    Week 48: Support Vector Machines and Summary of course

    +
    + + +
    +Morten Hjorth-Jensen +
    + +
    +Department of Physics, University of Oslo, Norway
    - -

    - - -

    [1] Department of Physics, University of Oslo
    -
    [2] Department of Physics and Astronomy and National Superconducting Cyclotron Laboratory, Michigan State University

    -

    -

    Nov 29, 2020

    +
    +

    Nov 21, 2024

    +

    -

    +









    +

    Overview of week 48

    -

    Overview of week 48

    +









    +

    Plan for week 47

    +
    +Lab sessions on Tuesday and Wednesday +

      -
    • Thursday: Support Vector Machines: Kernels, Classification and Regression. Video of Lecture.
      • -
    • Friday: Summary of course with perspectives for future studies. Video of Lecture.
      • +
    • Work and Discussion of project 3
      • +
    • Last weekly exercise,
    +
    + -See also Geron's chapter 5. Chapter 12 (sections 12.1-12.3 are the most relevant ones) of Hastie et al contains also a good discussion. - +
    +Plans for the lecture Monday 25 November, with video suggestions etc

    +

      +
    1. Bossting and gradient boosting and ensemble models
      2. +
    2. Summary of course
      3. +
    3. Readings and Videos: +
        +
      1. These lecture notes at https://github.com/CompPhysics/MachineLearning/blob/master/doc/pub/week47/ipynb/week48.ipynb
        2. +
      2. See also lecture notes from week 47 at https://github.com/CompPhysics/MachineLearning/blob/master/doc/pub/week46/ipynb/week47.ipynb. The lecture on Monday starts with a repetition on AdaBoost before we move over to gradient boosting with examples + +
        3. +
      3. Video on Decision trees https://www.youtube.com/watch?v=RmajweUFKvM&ab_channel=Simplilearn
        4. +
      4. Video on boosting methods https://www.youtube.com/watch?v=wPqtzj5VZus&ab_channel=H2O.ai
        5. +
      5. Video on AdaBoost https://www.youtube.com/watch?v=LsK-xG1cLYA
        6. +
      6. Video on Gradient boost, part 1, parts 2-4 follow thereafter https://www.youtube.com/watch?v=3CC4N4z3GJc
        7. +
      7. Decision Trees: Rashcka et al chapter 3 pages 86-98, and chapter 7 on Ensemble methods, Voting and Bagging and Gradient Boosting. See also lecture from STK-IN4300, lecture 7 at https://www.uio.no/studier/emner/matnat/math/STK-IN4300/h20/slides/lecture_7.pdf.
        8. +
      +
    +
    + +









    +

    Boosting, a Bird's Eye View

    -

    Thursday

    +

    The basic idea is to combine weak classifiers in order to create a good +classifier. With a weak classifier we often intend a classifier which +produces results which are only slightly better than we would get by +random guesses. +

    -

    -We finalize our discussion on Support Vector Machines with an emphasis on kernel transformations and applications to regression. The following video attempts at giving an overview on this part. See also the follow-up video. +

    This is done by applying in an iterative way a weak (or a standard +classifier like decision trees) to modify the data. In each iteration +we emphasize those observations which are misclassified by weighting +them with a factor. +

    -











    +

    What is boosting? Additive Modelling/Iterative Fitting

    -

    Friday

    +

    Boosting is a way of fitting an additive expansion in a set of +elementary basis functions like for example some simple polynomials. +Assume for example that we have a function +

    +$$ +f_M(x) = \sum_{i=1}^M \beta_m b(x;\gamma_m), +$$ -

    -Friday's lecture is split in two parts. It starts with a summary of -what we have done this semester and continues with perspectives for future studies and -modern research projects in machine learning. +

    where \( \beta_m \) are the expansion parameters to be determined in a +minimization process and \( b(x;\gamma_m) \) are some simple functions of +the multivariable parameter \( x \) which is characterized by the +parameters \( \gamma_m \). +

    + +

    As an example, consider the Sigmoid function we used in logistic +regression. In that case, we can translate the function +\( b(x;\gamma_m) \) into the Sigmoid function +

    + +$$ +\sigma(t) = \frac{1}{1+\exp{(-t)}}, +$$ + +

    where \( t=\gamma_0+\gamma_1 x \) and the parameters \( \gamma_0 \) and +\( \gamma_1 \) were determined by the Logistic Regression fitting +algorithm. +

    + +

    As another example, consider the cost function we defined for linear regression

    +$$ +C(\boldsymbol{y},\boldsymbol{f}) = \frac{1}{n} \sum_{i=0}^{n-1}(y_i-f(x_i))^2. +$$ + +

    In this case the function \( f(x) \) was replaced by the design matrix +\( \boldsymbol{X} \) and the unknown linear regression parameters \( \boldsymbol{\beta} \), +that is \( \boldsymbol{f}=\boldsymbol{X}\boldsymbol{\beta} \). In linear regression we can +simply invert a matrix and obtain the parameters \( \beta \) by +

    + +$$ +\boldsymbol{\beta}=\left(\boldsymbol{X}^T\boldsymbol{X}\right)^{-1}\boldsymbol{X}^T\boldsymbol{y}. +$$ + +

    In iterative fitting or additive modeling, we minimize the cost function with respect to the parameters \( \beta_m \) and \( \gamma_m \).

    -











    +

    Iterative Fitting, Regression and Squared-error Cost Function

    -

    Support Vector Machines, overarching aims

    - -

    -As discussed last week, -a Support Vector Machine (SVM) is a very powerful and versatile -Machine Learning method, capable of performing linear or nonlinear -classification, regression, and even outlier detection. It is one of -the most popular models in Machine Learning, and anyone interested in -Machine Learning should have it in their toolbox. SVMs are -particularly well suited for classification of complex but small-sized or -medium-sized datasets. - -

    -The case with two well-separated classes only can be understood in an -intuitive way in terms of lines in a two-dimensional space separating -the two classes. - -

    -The basic mathematics behind the SVM is however less familiar to most of us. -It relies on the definition of hyperplanes and the -definition of a margin which separates classes (in case of -classification problems) of variables. It is also used for regression -problems. I recommend you take a look at the lectures from last week on the binary classification problem. - -

    -









    - -

    Kernels and non-linearity

    - -

    -The cases we studied last week were all characterized by two classes -with a close to linear separability. The classifiers we have described -so far find linear boundaries in our input feature space. It is -possible to make our procedure more flexible by exploring the feature -space using other basis expansions such as higher-order polynomials, -wavelets, splines etc. - -

    -If our feature space is not easy to separate, as shown in the figure -generated by the code below, we can achieve a better separation by introducing a more complex -basis functions. The ideal would be, as shown by the code example below, to, -via a specific transformation to obtain a separation between the -classes that is almost linear. See also Figures 12.2 and 12.3 of Hastie et al.. - -

    -We can achieve a better separation by introducing more complex -basis functions. The ideal would be (see Figures 12.2 and 12.3) to, via a specific transformation to -obtain a separation between the classes which is almost linear. - -

    -The change of basis, from \( x\rightarrow z=\phi(x) \) leads to the same -type of equations to be solved, except that we need to introduce, for -example, a polynomial transformation to a two-dimensional training -set. - -

    - - -

    import numpy as np
-import os
-
-np.random.seed(42)
-
-# To plot pretty figures
-import matplotlib
-import matplotlib.pyplot as plt
-plt.rcParams['axes.labelsize'] = 14
-plt.rcParams['xtick.labelsize'] = 12
-plt.rcParams['ytick.labelsize'] = 12
-
-
-from sklearn.svm import SVC
-from sklearn import datasets
-
-
-
-X1D = np.linspace(-4, 4, 9).reshape(-1, 1)
-X2D = np.c_[X1D, X1D**2]
-y = np.array([0, 0, 1, 1, 1, 1, 1, 0, 0])
-
-plt.figure(figsize=(11, 4))
-
-plt.subplot(121)
-plt.grid(True, which='both')
-plt.axhline(y=0, color='k')
-plt.plot(X1D[:, 0][y==0], np.zeros(4), "bs")
-plt.plot(X1D[:, 0][y==1], np.zeros(5), "g^")
-plt.gca().get_yaxis().set_ticks([])
-plt.xlabel(r"$x_1$", fontsize=20)
-plt.axis([-4.5, 4.5, -0.2, 0.2])
-
-plt.subplot(122)
-plt.grid(True, which='both')
-plt.axhline(y=0, color='k')
-plt.axvline(x=0, color='k')
-plt.plot(X2D[:, 0][y==0], X2D[:, 1][y==0], "bs")
-plt.plot(X2D[:, 0][y==1], X2D[:, 1][y==1], "g^")
-plt.xlabel(r"$x_1$", fontsize=20)
-plt.ylabel(r"$x_2$", fontsize=20, rotation=0)
-plt.gca().get_yaxis().set_ticks([0, 4, 8, 12, 16])
-plt.plot([-4.5, 4.5], [6.5, 6.5], "r--", linewidth=3)
-plt.axis([-4.5, 4.5, -1, 17])
-plt.subplots_adjust(right=1)
-plt.show()
-
    -

    -









    - -

    The equations

    - -

    -Suppose we define a polynomial transformation of degree two only. We define a vector \( \boldsymbol{x}_i=[x_i,y_i] \) and have -$$ -\boldsymbol{z}_i^T = \phi(\boldsymbol{x}_i)^T =\left[1, \sqrt{2}x_i, \sqrt{2}y_i, x_i^2, y_i^2, \sqrt{2}x_iy_i\right]. -$$ - -

    -With our new basis, the equations we solved earlier are basically the same, that is we have now (without the slack option for simplicity) -$$ -{\cal L}=\sum_i\lambda_i-\frac{1}{2}\sum_{ij}^n\lambda_i\lambda_jy_iy_j\boldsymbol{z}_i^T\boldsymbol{z}_j, -$$ - -subject to the constraints \( \lambda_i\geq 0 \), \( \sum_i\lambda_iy_i=0 \), and for the support vectors -$$ -y_i(\boldsymbol{z}_i^T\boldsymbol{w}+b)= 1 \hspace{0.1cm}\forall i, -$$ - -from which we also find \( b \). -To compute \( \boldsymbol{z}_i^T\boldsymbol{z}_j \) we define the kernel \( K(\boldsymbol{x}_i,\boldsymbol{x}_j) \) as -$$ -K(\boldsymbol{x}_i,\boldsymbol{x}_j)=\boldsymbol{z}_i^T\boldsymbol{z}_j= \phi(\boldsymbol{x}_i)^T\phi(\boldsymbol{x}_j). -$$ - -For the above example, the kernel reads -$$ -K(\boldsymbol{x}_i,\boldsymbol{x}_j)=[1, \sqrt{2}x_i, \sqrt{2}y_i, x_i^2, y_i^2, \sqrt{2}x_iy_i]\begin{bmatrix} 1\\ \sqrt{2}x_j \\ \sqrt{2}y_j \\ x_j^2\\ y_i^2\\ \sqrt{2}x_jy_j \end{bmatrix}=1+2x_ix_j+2y_iy_j+(x_ix_j)^2+(y_iy_j)^2+2x_ix_jy_iy_j. -$$ - -

    -We note that this dot product can be rewritten as -$$ -K(\boldsymbol{x}_i,\boldsymbol{x}_j)=[1+\boldsymbol{x}_i^T\boldsymbol{x}_j]^d, -$$ - -where \( d=2 \) in our case and \( \boldsymbol{x}_i=[x_i,y_i] \) and \( \boldsymbol{x}_j=[x_j,y_j] \). -To compute the last equation is however inefficient from a computational stand. -Instead of computing the last equation for the kernel, we simply compute -the dot product \( (\boldsymbol{x}_i^T\boldsymbol{x}_j)^2 \). -This leads to the so-called -kernel trick. - -

    -









    - -

    The problem to solve

    - -

    -Using our definition of the kernel, we can rewrite again the Lagrangian -$$ -{\cal L}=\sum_i\lambda_i-\frac{1}{2}\sum_{ij}^n\lambda_i\lambda_jy_iy_j\boldsymbol{z}_i^T\boldsymbol{z}_j, -$$ - -subject to the constraints \( \lambda_i\geq 0 \), \( \sum_i\lambda_iy_i=0 \) in terms of a convex optimization problem -$$ -\frac{1}{2} \boldsymbol{\lambda}^T\begin{bmatrix} y_1y_1K(\boldsymbol{x}_1,\boldsymbol{x}_1) & y_1y_2K(\boldsymbol{x}_1,\boldsymbol{x}_2) & \dots & \dots & y_1y_nK(\boldsymbol{x}_1,\boldsymbol{x}_n) \\ -y_2y_1K(\boldsymbol{x}_2,\boldsymbol{x}_1) & y_2y_2(\boldsymbol{x}_2,\boldsymbol{x}_2) & \dots & \dots & y_1y_nK(\boldsymbol{x}_2,\boldsymbol{x}_n) \\ -\dots & \dots & \dots & \dots & \dots \\ -\dots & \dots & \dots & \dots & \dots \\ -y_ny_1K(\boldsymbol{x}_n,\boldsymbol{x}_1) & y_ny_2K(\boldsymbol{x}_n\boldsymbol{x}_2) & \dots & \dots & y_ny_nK(\boldsymbol{x}_n,\boldsymbol{x}_n) \\ -\end{bmatrix}\boldsymbol{\lambda}-\mathbb{1}\boldsymbol{\lambda}, -$$ - -subject to \( \boldsymbol{y}^T\boldsymbol{\lambda}=0 \). Here we defined the vectors \( \boldsymbol{\lambda} =[\lambda_1,\lambda_2,\dots,\lambda_n] \) and -\( \boldsymbol{y}=[y_1,y_2,\dots,y_n] \). -If we add the slack constants this leads to the additional constraint \( 0\leq \lambda_i \leq C \). - -

    -









    - -

    Tailoring the equations to the usage of CVXOPT

    - -

    -We can rewrite this (see the solutions below) in terms of a convex optimization problem of the type -$$ -\begin{align*} - &\mathrm{min}_{\lambda}\hspace{0.2cm} \frac{1}{2}\boldsymbol{\lambda}^T\boldsymbol{P}\boldsymbol{\lambda}+\boldsymbol{q}^T\boldsymbol{\lambda},\\ \nonumber - &\mathrm{s.t} \hspace{0.2cm} \boldsymbol{G}\boldsymbol{\lambda} \preceq \boldsymbol{h} \hspace{0.2cm} \wedge \boldsymbol{A}\boldsymbol{\lambda}=f. -\end{align*} -$$ - -Below we discuss how to solve these equations. Here we note that the matrix \( \boldsymbol{P} \) has matrix elements \( p_{ij}=y_iy_jK(\boldsymbol{x}_i,\boldsymbol{x}_j) \). -Given a kernel \( K \) and the targets \( y_i \) this matrix is easy to set up. The constraint \( \boldsymbol{y}^T\boldsymbol{\lambda}=0 \) leads to \( f=0 \) and \( \boldsymbol{A}=\boldsymbol{y} \). How to set up the matrix \( \boldsymbol{G} \) is discussed later. Here note that the inequalities \( 0\leq \lambda_i \leq C \) can be split up into -\( 0\leq \lambda_i \) and \( \lambda_i \leq C \). These two inequalities define then the matrix \( \boldsymbol{G} \) and the vector \( \boldsymbol{h} \). - -

    -









    - -

    Different kernels and Mercer's theorem

    - -

    -There are several popular kernels being used. These are +

    The way we proceed is as follows (here we specialize to the squared-error cost function)

      -
    1. Linear: \( K(\boldsymbol{x},\boldsymbol{y})=\boldsymbol{x}^T\boldsymbol{y} \),
      2. -
    2. Polynomial: \( K(\boldsymbol{x},\boldsymbol{y})=(\boldsymbol{x}^T\boldsymbol{y}+\gamma)^d \),
      3. -
    3. Gaussian Radial Basis Function: \( K(\boldsymbol{x},\boldsymbol{y})=\exp{\left(-\gamma\vert\vert\boldsymbol{x}-\boldsymbol{y}\vert\vert^2\right)} \),
      4. -
    4. Tanh: \( K(\boldsymbol{x},\boldsymbol{y})=\tanh{(\boldsymbol{x}^T\boldsymbol{y}+\gamma)} \),
      5. +
    5. Establish a cost function, here \( {\cal C}(\boldsymbol{y},\boldsymbol{f}) = \frac{1}{n} \sum_{i=0}^{n-1}(y_i-f_M(x_i))^2 \) with \( f_M(x) = \sum_{i=1}^M \beta_m b(x;\gamma_m) \).
      6. +
    6. Initialize with a guess \( f_0(x) \). It could be one or even zero or some random numbers.
      7. +
    7. For \( m=1:M \) +
        +
      1. minimize \( \sum_{i=0}^{n-1}(y_i-f_{m-1}(x_i)-\beta b(x;\gamma))^2 \) wrt \( \gamma \) and \( \beta \)
        2. +
      2. This gives the optimal values \( \beta_m \) and \( \gamma_m \)
        3. +
      3. Determine then the new values \( f_m(x)=f_{m-1}(x) +\beta_m b(x;\gamma_m) \)
      +
    +

    We could use any of the algorithms we have discussed till now. If we +use trees, \( \gamma \) parameterizes the split variables and split points +at the internal nodes, and the predictions at the terminal nodes. +

    -and many other ones. - -

    -An important theorem for us is Mercer's -theorem. The -theorem states that if a kernel function \( K \) is symmetric, continuous -and leads to a positive semi-definite matrix \( \boldsymbol{P} \) then there -exists a function \( \phi \) that maps \( \boldsymbol{x}_i \) and \( \boldsymbol{x}_j \) into -another space (possibly with much higher dimensions) such that - -$$ -K(\boldsymbol{x}_i,\boldsymbol{x}_j)=\phi(\boldsymbol{x}_i)^T\phi(\boldsymbol{x}_j). -$$ - -

    -So you can use \( K \) as a kernel since you know \( \phi \) exists, even if -you don’t know what \( \phi \) is. - -

    -Note that some frequently used kernels (such as the Sigmoid kernel) -don’t respect all of Mercer’s conditions, yet they generally work well -in practice. - -











    +

    Squared-Error Example and Iterative Fitting

    + +

    To better understand what happens, let us develop the steps for the iterative fitting using the above squared error function.

    + +

    For simplicity we assume also that our functions \( b(x;\gamma)=1+\gamma x \).

    + +

    This means that for every iteration \( m \), we need to optimize

    + +$$ +(\beta_m,\gamma_m) = \mathrm{argmin}_{\beta,\lambda}\hspace{0.1cm} \sum_{i=0}^{n-1}(y_i-f_{m-1}(x_i)-\beta b(x;\gamma))^2=\sum_{i=0}^{n-1}(y_i-f_{m-1}(x_i)-\beta(1+\gamma x_i))^2. +$$ + +

    We start our iteration by simply setting \( f_0(x)=0 \). +Taking the derivatives with respect to \( \beta \) and \( \gamma \) we obtain +

    +$$ +\frac{\partial {\cal C}}{\partial \beta} = -2\sum_{i}(1+\gamma x_i)(y_i-\beta(1+\gamma x_i))=0, +$$ + +

    and

    +$$ +\frac{\partial {\cal C}}{\partial \gamma} =-2\sum_{i}\beta x_i(y_i-\beta(1+\gamma x_i))=0. +$$ + +

    We can then rewrite these equations as (defining \( \boldsymbol{w}=\boldsymbol{e}+\gamma \boldsymbol{x}) \) with \( \boldsymbol{e} \) being the unit vector)

    +$$ +\gamma \boldsymbol{w}^T(\boldsymbol{y}-\beta\gamma \boldsymbol{w})=0, +$$ + +

    which gives us \( \beta = \boldsymbol{w}^T\boldsymbol{y}/(\boldsymbol{w}^T\boldsymbol{w}) \). Similarly we have

    +$$ +\beta\gamma \boldsymbol{x}^T(\boldsymbol{y}-\beta(1+\gamma \boldsymbol{x}))=0, +$$ + +

    which leads to \( \gamma =(\boldsymbol{x}^T\boldsymbol{y}-\beta\boldsymbol{x}^T\boldsymbol{e})/(\beta\boldsymbol{x}^T\boldsymbol{x}) \). Inserting +for \( \beta \) gives us an equation for \( \gamma \). This is a non-linear equation in the unknown \( \gamma \) and has to be solved numerically. +

    + +

    The solution to these two equations gives us in turn \( \beta_1 \) and \( \gamma_1 \) leading to the new expression for \( f_1(x) \) as +\( f_1(x) = \beta_1(1+\gamma_1x) \). Doing this \( M \) times results in our final estimate for the function \( f \). +

    + +









    +

    Iterative Fitting, Classification and AdaBoost

    + +

    Let us consider a binary classification problem with two outcomes \( y_i \in \{-1,1\} \) and \( i=0,1,2,\dots,n-1 \) as our set of +observations. We define a classification function \( G(x) \) which produces a prediction taking one or the other of the two values +\( \{-1,1\} \). +

    + +

    The error rate of the training sample is then

    + +$$ +\mathrm{\overline{err}}=\frac{1}{n} \sum_{i=0}^{n-1} I(y_i\ne G(x_i)). +$$ + +

    The iterative procedure starts with defining a weak classifier whose +error rate is barely better than random guessing. The iterative +procedure in boosting is to sequentially apply a weak +classification algorithm to repeatedly modified versions of the data +producing a sequence of weak classifiers \( G_m(x) \). +

    + +

    Here we will express our function \( f(x) \) in terms of \( G(x) \). That is

    +$$ +f_M(x) = \sum_{i=1}^M \beta_m b(x;\gamma_m), +$$ + +

    will be a function of

    +$$ +G_M(x) = \mathrm{sign} \sum_{i=1}^M \alpha_m G_m(x). +$$ + + +









    +

    Adaptive Boosting, AdaBoost

    + +

    In our iterative procedure we define thus

    +$$ +f_m(x) = f_{m-1}(x)+\beta_mG_m(x). +$$ + +

    The simplest possible cost function which leads (also simple from a computational point of view) to the AdaBoost algorithm is the +exponential cost/loss function defined as +

    +$$ +C(\boldsymbol{y},\boldsymbol{f}) = \sum_{i=0}^{n-1}\exp{(-y_i(f_{m-1}(x_i)+\beta G(x_i))}. +$$ + +

    We optimize \( \beta \) and \( G \) for each value of \( m=1:M \) as we did in the regression case. +This is normally done in two steps. Let us however first rewrite the cost function as +

    + +$$ +C(\boldsymbol{y},\boldsymbol{f}) = \sum_{i=0}^{n-1}w_i^{m}\exp{(-y_i\beta G(x_i))}, +$$ + +

    where we have defined \( w_i^m= \exp{(-y_if_{m-1}(x_i))} \).

    + +









    +

    Building up AdaBoost

    + +

    First, for any \( \beta > 0 \), we optimize \( G \) by setting

    +$$ +G_m(x) = \mathrm{sign} \sum_{i=0}^{n-1} w_i^m I(y_i \ne G_(x_i)), +$$ + +

    which is the classifier that minimizes the weighted error rate in predicting \( y \).

    + +

    We can do this by rewriting

    +$$ +\exp{-(\beta)}\sum_{y_i=G(x_i)}w_i^m+\exp{(\beta)}\sum_{y_i\ne G(x_i)}w_i^m, +$$ + +

    which can be rewritten as

    +$$ +(\exp{(\beta)}-\exp{-(\beta)})\sum_{i=0}^{n-1}w_i^mI(y_i\ne G(x_i))+\exp{(-\beta)}\sum_{i=0}^{n-1}w_i^m=0, +$$ + +

    which leads to

    +$$ +\beta_m = \frac{1}{2}\log{\frac{1-\mathrm{\overline{err}}}{\mathrm{\overline{err}}}}, +$$ + +

    where we have redefined the error as

    +$$ +\mathrm{\overline{err}}_m=\frac{1}{n}\frac{\sum_{i=0}^{n-1}w_i^mI(y_i\ne G(x_i)}{\sum_{i=0}^{n-1}w_i^m}, +$$ + +

    which leads to an update of

    +$$ +f_m(x) = f_{m-1}(x) +\beta_m G_m(x). +$$ + +

    This leads to the new weights

    +$$ +w_i^{m+1} = w_i^m \exp{(-y_i\beta_m G_m(x_i))} +$$ + + +









    +

    Adaptive boosting: AdaBoost, Basic Algorithm

    + +

    The algorithm here is rather straightforward. Assume that our weak +classifier is a decision tree and we consider a binary set of outputs +with \( y_i \in \{-1,1\} \) and \( i=0,1,2,\dots,n-1 \) as our set of +observations. Our design matrix is given in terms of the +feature/predictor vectors +\( \boldsymbol{X}=[\boldsymbol{x}_0\boldsymbol{x}_1\dots\boldsymbol{x}_{p-1}] \). Finally, we define also a +classifier determined by our data via a function \( G(x) \). This function tells us how well we are able to classify our outputs/targets \( \boldsymbol{y} \). +

    + +

    We have already defined the misclassification error \( \mathrm{err} \) as

    +$$ +\mathrm{err}=\frac{1}{n}\sum_{i=0}^{n-1}I(y_i\ne G(x_i)), +$$ + +

    where the function \( I() \) is one if we misclassify and zero if we classify correctly.

    + +









    +

    Basic Steps of AdaBoost

    + +

    With the above definitions we are now ready to set up the algorithm for AdaBoost. +The basic idea is to set up weights which will be used to scale the correctly classified and the misclassified cases. +

    +
      +
    1. We start by initializing all weights to \( w_i = 1/n \), with \( i=0,1,2,\dots n-1 \). It is easy to see that we must have \( \sum_{i=0}^{n-1}w_i = 1 \).
      2. +
    2. We rewrite the misclassification error as
      3. +
    +$$ +\mathrm{\overline{err}}_m=\frac{\sum_{i=0}^{n-1}w_i^m I(y_i\ne G(x_i))}{\sum_{i=0}^{n-1}w_i}, +$$ + +
      +
    1. Then we start looping over all attempts at classifying, namely we start an iterative process for \( m=1:M \), where \( M \) is the final number of classifications. Our given classifier could for example be a plain decision tree. +
        +
      1. Fit then a given classifier to the training set using the weights \( w_i \).
        2. +
      2. Compute then \( \mathrm{err} \) and figure out which events are classified properly and which are classified wrongly.
        3. +
      3. Define a quantity \( \alpha_{m} = \log{(1-\mathrm{\overline{err}}_m)/\mathrm{\overline{err}}_m} \)
        4. +
      4. Set the new weights to \( w_i = w_i\times \exp{(\alpha_m I(y_i\ne G(x_i)} \).
        5. +
      +
    2. Compute the new classifier \( G(x)= \sum_{i=0}^{n-1}\alpha_m I(y_i\ne G(x_i) \).
      3. +
    +

    For the iterations with \( m \le 2 \) the weights are modified +individually at each steps. The observations which were misclassified +at iteration \( m-1 \) have a weight which is larger than those which were +classified properly. As this proceeds, the observations which were +difficult to classifiy correctly are given a larger influence. Each +new classification step \( m \) is then forced to concentrate on those +observations that are missed in the previous iterations. +

    + +









    +

    AdaBoost Examples

    + +

    Using Scikit-Learn it is easy to apply the adaptive boosting algorithm, as done here.

    -

    The moons example (Adapted from Geron, chapter 5)

    -

    -

    from __future__ import division, print_function, unicode_literals
+
    
+  
    
+    
    
+      
    
+        
    
+  
    from sklearn.ensemble import AdaBoostClassifier
 
+ada_clf = AdaBoostClassifier(
+    DecisionTreeClassifier(max_depth=2), n_estimators=200,
+    algorithm="SAMME.R", learning_rate=0.01, random_state=42)
+ada_clf.fit(X_train, y_train)
+y_pred = ada_clf.predict(X_test)
+skplt.metrics.plot_confusion_matrix(y_test, y_pred, normalize=True)
+plt.show()
+y_probas = ada_clf.predict_proba(X_test)
+skplt.metrics.plot_roc(y_test, y_probas)
+plt.show()
+skplt.metrics.plot_cumulative_gain(y_test, y_probas)
+plt.show()
+
    
+
    
+      
    
+    
    
+  
    
+  
    
+    
    
+      
    
+        
              
+        
    
+      
    
+    
    
+  
    
+
    
+
+
+









    
+
    Making an  ADAboost code yourself 
    
+
+
+
+
    
+  
    
+    
    
+      
    
+        
    
+  
    import numpy as np
+
+class DecisionStump:
+    def fit(self, X, y, weights):
+        m, n = X.shape
+        self.alpha = 0
+        self.threshold = None
+        self.polarity = 1
+
+        min_error = float('inf')
+
+        for feature in range(n):
+            feature_values = np.unique(X[:, feature])
+
+            for threshold in feature_values:
+                for polarity in [1, -1]:
+                    predictions = np.ones(m)
+                    predictions[X[:, feature] < threshold] = -1
+                    predictions *= polarity
+
+                    error = sum(weights[predictions != y])
+
+                    if error < min_error:
+                        min_error = error
+                        self.alpha = 0.5 * np.log((1 - error) / (error + 1e-10))
+                        self.threshold = threshold
+                        self.feature_index = feature
+                        self.polarity = polarity
+
+    def predict(self, X):
+        m = X.shape[0]
+        predictions = np.ones(m)
+        if self.polarity == 1:
+            predictions[X[:, self.feature_index] < self.threshold] = -1
+        else:
+            predictions[X[:, self.feature_index] >= self.threshold] = -1
+        return predictions
+
+class AdaBoost:
+    def fit(self, X, y, n_estimators):
+        m = X.shape[0]
+        self.alphas = []
+        self.models = []
+
+        weights = np.ones(m) / m
+
+        for _ in range(n_estimators):
+            stump = DecisionStump()
+            stump.fit(X, y, weights)
+            predictions = stump.predict(X)
+
+            error = sum(weights[predictions != y])
+            if error == 0:
+                break
+
+            self.models.append(stump)
+            self.alphas.append(stump.alpha)
+
+            weights *= np.exp(-stump.alpha * y * predictions)
+            weights /= np.sum(weights)
+
+    def predict(self, X):
+        final_predictions = np.zeros(X.shape[0])
+        for alpha, model in zip(self.alphas, self.models):
+            final_predictions += alpha * model.predict(X)
+        return np.sign(final_predictions)
+
+# Example dataset (X, y)
+X = np.array([[1], [2], [3], [4], [5], [6], [7], [8], [9], [10]])
+y = np.array([-1, -1, -1, -1, 1, 1, 1, 1, 1, 1])  # Labels must be -1 or 1
+
+# Train AdaBoost
+ada = AdaBoost()
+ada.fit(X, y, n_estimators=10)
+
+# Predictions
+predictions = ada.predict(X)
+print("Predictions:", predictions)
+
    
+
    
+      
    
+    
    
+  
    
+  
    
+    
    
+      
    
+        
              
+        
    
+      
    
+    
    
+  
    
+
    
+
+
+









    
+
    Gradient boosting: Basics with Steepest Descent/Functional Gradient Descent 
    
+
+
    Gradient boosting is again a similar technique to Adaptive boosting,
+it combines so-called weak classifiers or regressors into a strong
+method via a series of iterations.
+
    
+
+
    In order to understand the method, let us illustrate its basics by
+bringing back the essential steps in linear regression, where our cost
+function was the least squares function.
+
    
+
+









    
+
    The Squared-Error again! Steepest Descent 
    
+
+
    We start again with our cost function \( {\cal C}(\boldsymbol{y}m\boldsymbol{f})=\sum_{i=0}^{n-1}{\cal L}(y_i, f(x_i)) \) where we want to minimize
+This means that for every iteration, we need to optimize
+
    
+
+$$
+(\hat{\boldsymbol{f}}) = \mathrm{argmin}_{\boldsymbol{f}}\hspace{0.1cm} \sum_{i=0}^{n-1}(y_i-f(x_i))^2.
+$$
+
+
    We define a real function \( h_m(x) \) that defines our final function \( f_M(x) \) as
    
+$$
+f_M(x) = \sum_{m=0}^M h_m(x).
+$$
+
+
    In the steepest decent approach we approximate \( h_m(x) = -\rho_m g_m(x) \), where \( \rho_m \) is a scalar and \( g_m(x) \) the gradient defined as
    
+$$
+g_m(x_i) = \left[ \frac{\partial {\cal L}(y_i, f(x_i))}{\partial f(x_i)}\right]_{f(x_i)=f_{m-1}(x_i)}.
+$$
+
+
    With the new gradient we can update \( f_m(x) = f_{m-1}(x) -\rho_m g_m(x) \). Using the above squared-error function we see that
+the gradient is \( g_m(x_i) = -2(y_i-f(x_i)) \).
+
    
+
+
    Choosing \( f_0(x)=0 \) we obtain \( g_m(x) = -2y_i \) and inserting this into the minimization problem for the cost function we have
    
+$$
+(\rho_1) = \mathrm{argmin}_{\rho}\hspace{0.1cm} \sum_{i=0}^{n-1}(y_i+2\rho y_i)^2.
+$$
+
+
+









    
+
    Steepest Descent Example 
    
+
+
    Optimizing with respect to \( \rho \) we obtain (taking the derivative) that \( \rho_1 = -1/2 \). We have then that
    
+$$
+f_1(x) = f_{0}(x) -\rho_1 g_1(x)=-y_i.
+$$
+
+
    We can then proceed and compute
    
+$$
+g_2(x_i) = \left[ \frac{\partial {\cal L}(y_i, f(x_i))}{\partial f(x_i)}\right]_{f(x_i)=f_{1}(x_i)=y_i}=-4y_i,
+$$
+
+
    and find a new value for \( \rho_2=-1/2 \) and continue till we have reached \( m=M \). We can modify the steepest descent method, or steepest boosting, by introducing what is called gradient boosting. 
    
+
+









    
+
    Gradient Boosting, algorithm 
    
+
+
    Steepest descent is however not much used, since it only optimizes \( f \) at a fixed set of \( n \) points,
+so we do not learn a function that can generalize. However, we can modify the algorithm by
+fitting a weak learner to approximate the negative gradient signal. 
+
    
+
+
    Suppose we have a cost function \( C(f)=\sum_{i=0}^{n-1}L(y_i, f(x_i)) \) where \( y_i \) is our target and \( f(x_i) \) the function which is meant to model \( y_i \). The above cost function could be our standard  squared-error  function
    
+$$
+C(\boldsymbol{y},\boldsymbol{f})=\sum_{i=0}^{n-1}(y_i-f(x_i))^2.
+$$
+
+
    The way we proceed in an iterative fashion is to
    
+
      
+
    1.  Initialize our estimate \( f_0(x) \).
      2. 
+
    2.  For \( m=1:M \), we
+
        
+ 
      1.  compute the negative gradient vector \( \boldsymbol{u}_m = -\partial C(\boldsymbol{y},\boldsymbol{f})/\partial \boldsymbol{f}(x) \) at \( f(x) = f_{m-1}(x) \);
        2. 
+ 
      2.  fit the so-called base-learner to the negative gradient \( h_m(u_m,x) \);
        3. 
+ 
      3.  update the estimate \( f_m(x) = f_{m-1}(x)+h_m(u_m,x) \);
        4. 
+
      
+
    3.  The final estimate is then \( f_M(x) = \sum_{m=1}^M h_m(u_m,x) \).
      4. 
+
    
+









    
+
    Gradient Boosting, Examples of Regression 
    
+
+
+
    
+  
    
+    
    
+      
    
+        
    
+  
    import matplotlib.pyplot as plt
 import numpy as np
-np.random.seed(42)
+from sklearn.model_selection import train_test_split
+from sklearn.ensemble import GradientBoostingRegressor
+import scikitplot as skplt
+from sklearn.metrics import mean_squared_error
 
-import matplotlib
-import matplotlib.pyplot as plt
-plt.rcParams['axes.labelsize'] = 14
-plt.rcParams['xtick.labelsize'] = 12
-plt.rcParams['ytick.labelsize'] = 12
+n = 100
+maxdegree = 6
+
+# Make data set.
+x = np.linspace(-3, 3, n).reshape(-1, 1)
+y = np.exp(-x**2) + 1.5 * np.exp(-(x-2)**2)+ np.random.normal(0, 0.1, x.shape)
+
+error = np.zeros(maxdegree)
+bias = np.zeros(maxdegree)
+variance = np.zeros(maxdegree)
+polydegree = np.zeros(maxdegree)
+X_train, X_test, y_train, y_test = train_test_split(x, y, test_size=0.2)
+
+for degree in range(1,maxdegree):
+    model = GradientBoostingRegressor(max_depth=degree, n_estimators=100, learning_rate=1.0)  
+    model.fit(X_train,y_train)
+    y_pred = model.predict(X_test)
+    polydegree[degree] = degree
+    error[degree] = np.mean( np.mean((y_test - y_pred)**2) )
+    bias[degree] = np.mean( (y_test - np.mean(y_pred))**2 )
+    variance[degree] = np.mean( np.var(y_pred) )
+    print('Max depth:', degree)
+    print('Error:', error[degree])
+    print('Bias^2:', bias[degree])
+    print('Var:', variance[degree])
+    print('{} >= {} + {} = {}'.format(error[degree], bias[degree], variance[degree], bias[degree]+variance[degree]))
+
+plt.xlim(1,maxdegree-1)
+plt.plot(polydegree, error, label='Error')
+plt.plot(polydegree, bias, label='bias')
+plt.plot(polydegree, variance, label='Variance')
+plt.legend()
+save_fig("gdregression")
+plt.show()
+
    
+
    
+      
    
+    
    
+  
    
+  
    
+    
    
+      
    
+        
              
+        
    
+      
    
+    
    
+  
    
+
    
 
 
-from sklearn.svm import SVC
-from sklearn import datasets
+









    
+
    Gradient Boosting, Classification Example 
    
 
+
+
    
+  
    
+    
    
+      
    
+        
    
+  
    import matplotlib.pyplot as plt
+import numpy as np
+from sklearn.model_selection import  train_test_split 
+from sklearn.datasets import load_breast_cancer
+import scikitplot as skplt
+from sklearn.ensemble import GradientBoostingClassifier
+from sklearn.model_selection import cross_validate
 
+# Load the data
+cancer = load_breast_cancer()
 
-from sklearn.pipeline import Pipeline
+X_train, X_test, y_train, y_test = train_test_split(cancer.data,cancer.target,random_state=0)
+print(X_train.shape)
+print(X_test.shape)
+#now scale the data
 from sklearn.preprocessing import StandardScaler
-from sklearn.svm import LinearSVC
+scaler = StandardScaler()
+scaler.fit(X_train)
+X_train_scaled = scaler.transform(X_train)
+X_test_scaled = scaler.transform(X_test)
 
+gd_clf = GradientBoostingClassifier(max_depth=3, n_estimators=100, learning_rate=1.0)  
+gd_clf.fit(X_train_scaled, y_train)
+#Cross validation
+accuracy = cross_validate(gd_clf,X_test_scaled,y_test,cv=10)['test_score']
+print(accuracy)
+print("Test set accuracy with Gradient boosting and scaled data: {:.2f}".format(gd_clf.score(X_test_scaled,y_test)))
 
-from sklearn.datasets import make_moons
-X, y = make_moons(n_samples=100, noise=0.15, random_state=42)
-
-def plot_dataset(X, y, axes):
-    plt.plot(X[:, 0][y==0], X[:, 1][y==0], "bs")
-    plt.plot(X[:, 0][y==1], X[:, 1][y==1], "g^")
-    plt.axis(axes)
-    plt.grid(True, which='both')
-    plt.xlabel(r"$x_1$", fontsize=20)
-    plt.ylabel(r"$x_2$", fontsize=20, rotation=0)
-
-plot_dataset(X, y, [-1.5, 2.5, -1, 1.5])
+import scikitplot as skplt
+y_pred = gd_clf.predict(X_test_scaled)
+skplt.metrics.plot_confusion_matrix(y_test, y_pred, normalize=True)
+save_fig("gdclassiffierconfusion")
 plt.show()
-
-from sklearn.datasets import make_moons
-from sklearn.pipeline import Pipeline
-from sklearn.preprocessing import PolynomialFeatures
-
-polynomial_svm_clf = Pipeline([
-        ("poly_features", PolynomialFeatures(degree=3)),
-        ("scaler", StandardScaler()),
-        ("svm_clf", LinearSVC(C=10, loss="hinge", random_state=42))
-    ])
-
-polynomial_svm_clf.fit(X, y)
-
-def plot_predictions(clf, axes):
-    x0s = np.linspace(axes[0], axes[1], 100)
-    x1s = np.linspace(axes[2], axes[3], 100)
-    x0, x1 = np.meshgrid(x0s, x1s)
-    X = np.c_[x0.ravel(), x1.ravel()]
-    y_pred = clf.predict(X).reshape(x0.shape)
-    y_decision = clf.decision_function(X).reshape(x0.shape)
-    plt.contourf(x0, x1, y_pred, cmap=plt.cm.brg, alpha=0.2)
-    plt.contourf(x0, x1, y_decision, cmap=plt.cm.brg, alpha=0.1)
-
-plot_predictions(polynomial_svm_clf, [-1.5, 2.5, -1, 1.5])
-plot_dataset(X, y, [-1.5, 2.5, -1, 1.5])
-
+y_probas = gd_clf.predict_proba(X_test_scaled)
+skplt.metrics.plot_roc(y_test, y_probas)
+save_fig("gdclassiffierroc")
 plt.show()
-
-
-from sklearn.svm import SVC
-
-poly_kernel_svm_clf = Pipeline([
-        ("scaler", StandardScaler()),
-        ("svm_clf", SVC(kernel="poly", degree=3, coef0=1, C=5))
-    ])
-poly_kernel_svm_clf.fit(X, y)
-
-poly100_kernel_svm_clf = Pipeline([
-        ("scaler", StandardScaler()),
-        ("svm_clf", SVC(kernel="poly", degree=10, coef0=100, C=5))
-    ])
-poly100_kernel_svm_clf.fit(X, y)
-
-plt.figure(figsize=(11, 4))
-
-plt.subplot(121)
-plot_predictions(poly_kernel_svm_clf, [-1.5, 2.5, -1, 1.5])
-plot_dataset(X, y, [-1.5, 2.5, -1, 1.5])
-plt.title(r"$d=3, r=1, C=5$", fontsize=18)
-
-plt.subplot(122)
-plot_predictions(poly100_kernel_svm_clf, [-1.5, 2.5, -1, 1.5])
-plot_dataset(X, y, [-1.5, 2.5, -1, 1.5])
-plt.title(r"$d=10, r=100, C=5$", fontsize=18)
-
-plt.show()
-
-def gaussian_rbf(x, landmark, gamma):
-    return np.exp(-gamma * np.linalg.norm(x - landmark, axis=1)**2)
-
-gamma = 0.3
-
-x1s = np.linspace(-4.5, 4.5, 200).reshape(-1, 1)
-x2s = gaussian_rbf(x1s, -2, gamma)
-x3s = gaussian_rbf(x1s, 1, gamma)
-
-XK = np.c_[gaussian_rbf(X1D, -2, gamma), gaussian_rbf(X1D, 1, gamma)]
-yk = np.array([0, 0, 1, 1, 1, 1, 1, 0, 0])
-
-plt.figure(figsize=(11, 4))
-
-plt.subplot(121)
-plt.grid(True, which='both')
-plt.axhline(y=0, color='k')
-plt.scatter(x=[-2, 1], y=[0, 0], s=150, alpha=0.5, c="red")
-plt.plot(X1D[:, 0][yk==0], np.zeros(4), "bs")
-plt.plot(X1D[:, 0][yk==1], np.zeros(5), "g^")
-plt.plot(x1s, x2s, "g--")
-plt.plot(x1s, x3s, "b:")
-plt.gca().get_yaxis().set_ticks([0, 0.25, 0.5, 0.75, 1])
-plt.xlabel(r"$x_1$", fontsize=20)
-plt.ylabel(r"Similarity", fontsize=14)
-plt.annotate(r'$\mathbf{x}$',
-             xy=(X1D[3, 0], 0),
-             xytext=(-0.5, 0.20),
-             ha="center",
-             arrowprops=dict(facecolor='black', shrink=0.1),
-             fontsize=18,
-            )
-plt.text(-2, 0.9, "$x_2$", ha="center", fontsize=20)
-plt.text(1, 0.9, "$x_3$", ha="center", fontsize=20)
-plt.axis([-4.5, 4.5, -0.1, 1.1])
-
-plt.subplot(122)
-plt.grid(True, which='both')
-plt.axhline(y=0, color='k')
-plt.axvline(x=0, color='k')
-plt.plot(XK[:, 0][yk==0], XK[:, 1][yk==0], "bs")
-plt.plot(XK[:, 0][yk==1], XK[:, 1][yk==1], "g^")
-plt.xlabel(r"$x_2$", fontsize=20)
-plt.ylabel(r"$x_3$  ", fontsize=20, rotation=0)
-plt.annotate(r'$\phi\left(\mathbf{x}\right)$',
-             xy=(XK[3, 0], XK[3, 1]),
-             xytext=(0.65, 0.50),
-             ha="center",
-             arrowprops=dict(facecolor='black', shrink=0.1),
-             fontsize=18,
-            )
-plt.plot([-0.1, 1.1], [0.57, -0.1], "r--", linewidth=3)
-plt.axis([-0.1, 1.1, -0.1, 1.1])
-    
-plt.subplots_adjust(right=1)
-
+skplt.metrics.plot_cumulative_gain(y_test, y_probas)
+save_fig("gdclassiffiercgain")
 plt.show()
+
    
+
    
+      
    
+    
    
+  
    
+  
    
+    
    
+      
    
+        
              
+        
    
+      
    
+    
    
+  
    
+
    
 
 
-x1_example = X1D[3, 0]
-for landmark in (-2, 1):
-    k = gaussian_rbf(np.array([[x1_example]]), np.array([[landmark]]), gamma)
-    print("Phi({}, {}) = {}".format(x1_example, landmark, k))
-
-rbf_kernel_svm_clf = Pipeline([
-        ("scaler", StandardScaler()),
-        ("svm_clf", SVC(kernel="rbf", gamma=5, C=0.001))
-    ])
-rbf_kernel_svm_clf.fit(X, y)
-
-
-from sklearn.svm import SVC
-
-gamma1, gamma2 = 0.1, 5
-C1, C2 = 0.001, 1000
-hyperparams = (gamma1, C1), (gamma1, C2), (gamma2, C1), (gamma2, C2)
-
-svm_clfs = []
-for gamma, C in hyperparams:
-    rbf_kernel_svm_clf = Pipeline([
-            ("scaler", StandardScaler()),
-            ("svm_clf", SVC(kernel="rbf", gamma=gamma, C=C))
-        ])
-    rbf_kernel_svm_clf.fit(X, y)
-    svm_clfs.append(rbf_kernel_svm_clf)
-
-plt.figure(figsize=(11, 7))
-
-for i, svm_clf in enumerate(svm_clfs):
-    plt.subplot(221 + i)
-    plot_predictions(svm_clf, [-1.5, 2.5, -1, 1.5])
-    plot_dataset(X, y, [-1.5, 2.5, -1, 1.5])
-    gamma, C = hyperparams[i]
-    plt.title(r"$\gamma = {}, C = {}$".format(gamma, C), fontsize=16)
-
-plt.show()
-
    -











    +

    XGBoost: Extreme Gradient Boosting

    -

    Mathematical optimization of convex functions

    +

    XGBoost or Extreme Gradient +Boosting, is an optimized distributed gradient boosting library +designed to be highly efficient, flexible and portable. It implements +machine learning algorithms under the Gradient Boosting +framework. XGBoost provides a parallel tree boosting that solve many +data science problems in a fast and accurate way. See the article by Chen and Guestrin. +

    -

    -A mathematical (quadratic) optimization problem, or just optimization problem, has the form -$$ -\begin{align*} - &\mathrm{min}_{\lambda}\hspace{0.2cm} \frac{1}{2}\boldsymbol{\lambda}^T\boldsymbol{P}\boldsymbol{\lambda}+\boldsymbol{q}^T\boldsymbol{\lambda},\\ \nonumber - &\mathrm{subject\hspace{0.1cm}to} \hspace{0.2cm} \boldsymbol{G}\boldsymbol{\lambda} \preceq \boldsymbol{h} \wedge \boldsymbol{A}\boldsymbol{\lambda}=f. -\end{align*} -$$ +

    The authors design and build a highly scalable end-to-end tree +boosting system. It has a theoretically justified weighted quantile +sketch for efficient proposal calculation. It introduces a novel sparsity-aware algorithm for parallel tree learning and an effective cache-aware block structure for out-of-core tree learning. +

    -subject to some constraints for say a selected set \( i=1,2,\dots, n \). -In our case we are optimizing with respect to the Lagrangian multipliers \( \lambda_i \), and the -vector \( \boldsymbol{\lambda}=[\lambda_1, \lambda_2,\dots, \lambda_n] \) is the optimization variable we are dealing with. +

    It is now the algorithm which wins essentially all ML competitions!!!

    -

    -In our case we are particularly interested in a class of optimization problems called convex optmization problems. -In our discussion on gradient descent methods we discussed at length the definition of a convex function. - -

    -Convex optimization problems play a central role in applied mathematics and we recommend strongly Boyd and Vandenberghe's text on the topics. - -











    +

    Regression Case

    -

    How do we solve these problems?

    - -

    -If we use Python as programming language and wish to venture beyond -scikit-learn, tensorflow and similar software which makes our -lives so much easier, we need to dive into the wonderful world of -quadratic programming. We can, if we wish, solve the minimization -problem using standard gradient methods or conjugate gradient -methods. However, these methods tend to exhibit a rather slow -converge. So, welcome to the promised land of quadratic programming. - -

    -The functions we need are contained in the quadratic programming package CVXOPT and we need to import it together with numpy as - -

    -

    import numpy
-import cvxopt
-
    -

    -This will make our life much easier. You don't need to write your own optimizer. +

    +
    +
    +
    +
    + 
    import matplotlib.pyplot as plt
+import numpy as np
+from sklearn.model_selection import train_test_split
+import xgboost as xgb
+import scikitplot as skplt
+from sklearn.metrics import mean_squared_error
+
+n = 100
+maxdegree = 6
+
+# Make data set.
+x = np.linspace(-3, 3, n).reshape(-1, 1)
+y = np.exp(-x**2) + 1.5 * np.exp(-(x-2)**2)+ np.random.normal(0, 0.1, x.shape)
+
+error = np.zeros(maxdegree)
+bias = np.zeros(maxdegree)
+variance = np.zeros(maxdegree)
+polydegree = np.zeros(maxdegree)
+X_train, X_test, y_train, y_test = train_test_split(x, y, test_size=0.2)
+
+for degree in range(maxdegree):
+    model =  xgb.XGBRegressor(objective ='reg:squarederror', colsaobjective ='reg:squarederror', colsample_bytree = 0.3, learning_rate = 0.1,max_depth = degree, alpha = 10, n_estimators = 200)
+
+    model.fit(X_train,y_train)
+    y_pred = model.predict(X_test)
+    polydegree[degree] = degree
+    error[degree] = np.mean( np.mean((y_test - y_pred)**2) )
+    bias[degree] = np.mean( (y_test - np.mean(y_pred))**2 )
+    variance[degree] = np.mean( np.var(y_pred) )
+    print('Max depth:', degree)
+    print('Error:', error[degree])
+    print('Bias^2:', bias[degree])
+    print('Var:', variance[degree])
+    print('{} >= {} + {} = {}'.format(error[degree], bias[degree], variance[degree], bias[degree]+variance[degree]))
+
+plt.xlim(1,maxdegree-1)
+plt.plot(polydegree, error, label='Error')
+plt.plot(polydegree, bias, label='bias')
+plt.plot(polydegree, variance, label='Variance')
+plt.legend()
+plt.show()
+
    +
    +
    +
    +
    +
    +
    +
    +
    +
    +
    +
    +
    +
    + -











    +

    Xgboost on the Cancer Data

    -

    A simple example

    - -

    -We remind ourselves about the general problem we want to solve -$$ -\begin{align*} - &\mathrm{min}_{x}\hspace{0.2cm} \frac{1}{2}\boldsymbol{x}^T\boldsymbol{P}\boldsymbol{x}+\boldsymbol{q}^T\boldsymbol{x},\\ \nonumber - &\mathrm{s.t.} \hspace{0.2cm} \boldsymbol{G}\boldsymbol{x} \preceq \boldsymbol{h} \wedge \boldsymbol{A}\boldsymbol{x}=f. -\end{align*} -$$ - -Note: we use s.t. for subject to. - -

    -Let us show how to perform the optmization using a simple case. Assume we want to optimize the following problem -$$ -\begin{align*} - &\mathrm{min}_{x}\hspace{0.2cm} \frac{1}{2}x^2+5x+3y \\ \nonumber - &\mathrm{s.t.} \\ \nonumber - &x, y \geq 0 \\ \nonumber - &x+3y \geq 15 \\ \nonumber - &2x+5y \leq 100 \\ \nonumber - &3x+4y \leq 80. \\ \nonumber -\end{align*} -$$ - -The minimization problem can be rewritten in terms of vectors and matrices as (with \( x \) and \( y \) being the unknowns) -$$ -\frac{1}{2}\begin{bmatrix} x\\ y \end{bmatrix}^T \begin{bmatrix} 1 & 0\\ 0 & 0 \end{bmatrix} \begin{bmatrix} x \\ y \end{bmatrix} + \begin{bmatrix}5\\ 3 \end{bmatrix}^T \begin{bmatrix}x \\ y \end{bmatrix}. -$$ - -Similarly, we can now set up the inequalities (we need to change \( \geq \) to \( \leq \) by multiplying with \( -1 \) on bot sides) as the following matrix-vector equation -$$ -\begin{bmatrix} -1 & 0 \\ 0 & -1 \\ -1 & -3 \\ 2 & 5 \\ 3 & 4\end{bmatrix}\begin{bmatrix} x \\ y\end{bmatrix} \preceq \begin{bmatrix}0 \\ 0\\ -15 \\ 100 \\ 80\end{bmatrix}. -$$ - -We have collapsed all the inequalities into a single matrix \( \boldsymbol{G} \). We see also that our matrix -$$ -\boldsymbol{P} =\begin{bmatrix} 1 & 0\\ 0 & 0 \end{bmatrix} -$$ - -is clearly positive semi-definite (all eigenvalues larger or equal zero). -Finally, the vector \( \boldsymbol{h} \) is defined as -$$ -\boldsymbol{h} = \begin{bmatrix}0 \\ 0\\ -15 \\ 100 \\ 80\end{bmatrix}. -$$ - -

    -Since we don't have any equalities the matrix \( \boldsymbol{A} \) is set to zero -The following code solves the equations for us -

    +

    As you will see from the confusion matrix below, XGBoots does an excellent job on the Wisconsin cancer data and outperforms essentially all agorithms we have discussed till now.

    -
    # Import the necessary packages
-import numpy
-from cvxopt import matrix
-from cvxopt import solvers
-P = matrix(numpy.diag([1,0]), tc='d')
-q = matrix(numpy.array([3,4]), tc='d')
-G = matrix(numpy.array([[-1,0],[0,-1],[-1,-3],[2,5],[3,4]]), tc='d')
-h = matrix(numpy.array([0,0,-15,100,80]), tc='d')
-# Construct the QP, invoke solver
-sol = solvers.qp(P,q,G,h)
-# Extract optimal value and solution
-sol['x'] 
-sol['primal objective']
-
    -

    -









    +

    +
    +
    +
    +
    + 
    import matplotlib.pyplot as plt
+import numpy as np
+from sklearn.model_selection import  train_test_split 
+from sklearn.datasets import load_breast_cancer
+from sklearn.preprocessing import LabelEncoder
+from sklearn.model_selection import cross_validate
+import scikitplot as skplt
+import xgboost as xgb
+# Load the data
+cancer = load_breast_cancer()
 
-
    Back to the more realistic cases 
    
+X_train, X_test, y_train, y_test = train_test_split(cancer.data,cancer.target,random_state=0)
+print(X_train.shape)
+print(X_test.shape)
+#now scale the data
+from sklearn.preprocessing import StandardScaler
+scaler = StandardScaler()
+scaler.fit(X_train)
+X_train_scaled = scaler.transform(X_train)
+X_test_scaled = scaler.transform(X_test)
 
-
    
-We are now ready to return to our setup of the optmization problem for a more realistic case. Introducing the slack parameter \( C \) we have
-$$
-\frac{1}{2} \boldsymbol{\lambda}^T\begin{bmatrix} y_1y_1K(\boldsymbol{x}_1,\boldsymbol{x}_1) & y_1y_2K(\boldsymbol{x}_1,\boldsymbol{x}_2) & \dots & \dots & y_1y_nK(\boldsymbol{x}_1,\boldsymbol{x}_n) \\
-y_2y_1K(\boldsymbol{x}_2,\boldsymbol{x}_1) & y_2y_2K(\boldsymbol{x}_2,\boldsymbol{x}_2) & \dots & \dots & y_1y_nK(\boldsymbol{x}_2,\boldsymbol{x}_n) \\
-\dots & \dots & \dots & \dots & \dots \\
-\dots & \dots & \dots & \dots & \dots \\
-y_ny_1K(\boldsymbol{x}_n,\boldsymbol{x}_1) & y_ny_2K(\boldsymbol{x}_n\boldsymbol{x}_2) & \dots & \dots & y_ny_nK(\boldsymbol{x}_n,\boldsymbol{x}_n) \\
-\end{bmatrix}\boldsymbol{\lambda}-\mathbb{I}\boldsymbol{\lambda}, 
-$$
+xg_clf = xgb.XGBClassifier()
+xg_clf.fit(X_train_scaled,y_train)
 
-subject to \( \boldsymbol{y}^T\boldsymbol{\lambda}=0 \). Here we defined the vectors \( \boldsymbol{\lambda} =[\lambda_1,\lambda_2,\dots,\lambda_n] \) and 
-\( \boldsymbol{y}=[y_1,y_2,\dots,y_n] \). 
-With  the slack constants this leads to the additional constraint \( 0\leq \lambda_i \leq C \).
+y_test = xg_clf.predict(X_test_scaled)
 
-
    
-









    
+print("Test set accuracy with Gradient Boosting and scaled data: {:.2f}".format(xg_clf.score(X_test_scaled,y_test)))
 
-
    Setting up the matrices and the problem 
    
-
-
    
-We have the general problem 
-$$
-\begin{align*}
-    &\mathrm{min}_{\lambda}\hspace{0.2cm} \frac{1}{2}\boldsymbol{\lambda}^T\boldsymbol{P}\boldsymbol{\lambda}+\boldsymbol{q}^T\boldsymbol{\lambda},\\ \nonumber
-    &\mathrm{s.t.} \hspace{0.2cm} \boldsymbol{G}\boldsymbol{\lambda} \preceq \boldsymbol{h} \wedge  \boldsymbol{A}\boldsymbol{\lambda}=f.
-\end{align*}
-$$
+import scikitplot as skplt
+y_pred = xg_clf.predict(X_test_scaled)
+skplt.metrics.plot_confusion_matrix(y_test, y_pred, normalize=True)
+save_fig("xdclassiffierconfusion")
+plt.show()
+y_probas = xg_clf.predict_proba(X_test_scaled)
+skplt.metrics.plot_roc(y_test, y_probas)
+save_fig("xdclassiffierroc")
+plt.show()
+skplt.metrics.plot_cumulative_gain(y_test, y_probas)
+save_fig("gdclassiffiercgain")
+plt.show()
 
 
-
      
-
    1.  With a given kernel we can thus define the matrix \( \boldsymbol{P} \).
      2. 
-
    2.  The matrix \( \boldsymbol{P} \) has matrix elements \( p_{ij}=y_iy_jK(\boldsymbol{x}_i,\boldsymbol{x}_j) \). Given a kernel \( K \) and the targets \( y_i \) this matrix is easy to set up.
      3. 
-
    3.  The vector \( \boldsymbol{q} \) has all elements equal to \( -1 \).
      4. 
-
    4.  The constraint \( \boldsymbol{y}^T\boldsymbol{\lambda}=0 \) leads to \( f=0 \) and \( \boldsymbol{A}=\boldsymbol{y} \).
      5. 
-
    5.  To set up the matrix \( \boldsymbol{G} \) we note that the inequalities \( 0\leq \lambda_i \leq C \) can be split up into \( 0\leq \lambda_i \) and \( \lambda_i \leq C \). These two inequalities define then the matrix \( \boldsymbol{G} \) and the vector \( \boldsymbol{h} \).
      6. 
-
    
+xgb.plot_tree(xg_clf,num_trees=0)
+plt.rcParams['figure.figsize'] = [50, 10]
+save_fig("xgtree")
+plt.show()
+
+xgb.plot_importance(xg_clf)
+plt.rcParams['figure.figsize'] = [5, 5]
+save_fig("xgparams")
+plt.show()
+
    +
    +
    +
    +
    +
    +
    +
    +
    +
    +
    +
    +
    +
    +









    +

    Gradient boosting, making our own code for a regression case

    -

    Setting up \( \boldsymbol{G}\boldsymbol{\lambda} \preceq \boldsymbol{h} \)

    -

    -We have two constraints, \( 0\le \lambda_i \) and \( \lambda_i \le C \). To do this we multiply the ones with the constraint -\( \ge \) with \( -1 \) in order to get \( \le \). It means that the problem \( \boldsymbol{G}\boldsymbol{\lambda} \preceq \boldsymbol{h} \) -can be written as -$$ -\begin{bmatrix} -1& 0 & 0 & \dots & 0 \\ -0& -1 & 0 & \dots & 0 \\ -0& 0 & -1 & \dots & 0 \\ -\dots& \dots & \dots & \dots & \dots \\ -0& 0 & 0 & \dots & -1 \\ -1 & 0 & 0 & \dots & 0 \\ -0& 1 & 0 & \dots & 0 \\ -0& 0 & 1 & \dots & 0 \\ -\dots& \dots & \dots & \dots & \dots \\ -0& 0 & 0 & \dots & 1 \\ -\end{bmatrix} -\begin{bmatrix} \lambda_1 \\ -\lambda_2 \\ -\lambda_3 \\ -\dots \\ -\lambda_n \\ -\end{bmatrix} \preceq -\begin{bmatrix} 0 \\ -0 \\ -0 \\ -\dots \\ -0 \\ -C \\ -C \\ -C \\ -\dots \\ -C \\ -\end{bmatrix}. -$$ + +

    +
    +
    +
    +
    + 
    import numpy as np
+class DecisionTreeRegressor:
+    def __init__(self, max_depth=3):
+        self.max_depth = max_depth
+        self.tree = None
+    def fit(self, X, y):
+        self.tree = self._grow_tree(X, y)
+    def _grow_tree(self, X, y, depth=0):
+        n_samples, n_features = X.shape
+        if depth < self.max_depth:
+            best_feature, best_threshold = self._best_split(X, y)
+            if best_feature is not None:
+                left_indices = X[:, best_feature] < best_threshold
+                right_indices = X[:, best_feature] >= best_threshold
+                left_child = self._grow_tree(X[left_indices], y[left_indices], depth + 1)
+                right_child = self._grow_tree(X[right_indices], y[right_indices], depth + 1)
+                return (best_feature, best_threshold, left_child, right_child)
+        return np.mean(y)
+    def _best_split(self, X, y):
+        best_mse = float('inf')
+        best_feature, best_threshold = None, None
+        n_samples, n_features = X.shape
+        
+        for feature in range(n_features):
+            thresholds = np.unique(X[:, feature])
+            for threshold in thresholds:
+                left_indices = X[:, feature] < threshold
+                right_indices = X[:, feature] >= threshold
+                if len(y[left_indices]) > 0 and len(y[right_indices]) > 0:
+                    left_mse = np.mean((y[left_indices] - np.mean(y[left_indices])) ** 2)
+                    right_mse = np.mean((y[right_indices] - np.mean(y[right_indices])) ** 2)
+                    mse = (len(y[left_indices]) * left_mse + len(y[right_indices]) * right_mse) / n_samples
+                    
+                    if mse < best_mse:
+                        best_mse = mse
+                        best_feature = feature
+                        best_threshold = threshold
+        return best_feature, best_threshold
+    def predict(self, X):
+        return np.array([self._predict_sample(sample, self.tree) for sample in X])
+    def _predict_sample(self, sample, node):
+        if isinstance(node, tuple):
+            feature, threshold, left_child, right_child = node
+            if sample[feature] < threshold:
+                return self._predict_sample(sample, left_child)
+            else:
+                return self._predict_sample(sample, right_child)
+        return node
+class GradientBoostingRegressor:
+    def __init__(self, n_estimators=100, learning_rate=0.1, max_depth=3):
+        self.n_estimators = n_estimators
+        self.learning_rate = learning_rate
+        self.max_depth = max_depth
+        self.models = []
+    def fit(self, X, y):
+        y_pred = np.zeros(y.shape)
+        for _ in range(self.n_estimators):
+            residuals = y - y_pred
+            model = DecisionTreeRegressor(max_depth=self.max_depth)
+            model.fit(X, residuals)
+            y_pred += self.learning_rate * model.predict(X)
+            self.models.append(model)
+    def predict(self, X):
+        y_pred = np.zeros(X.shape[0])
+        for model in self.models:
+            y_pred += self.learning_rate * model.predict(X)
+        return y_pred
+# Example usage
+if __name__ == "__main__":
+    # Sample data
+    X = np.array([[1], [2], [3], [4], [5]])
+    y = np.array([1.5, 1.7, 3.5, 3.7, 5.0])
+    model = GradientBoostingRegressor(n_estimators=100, learning_rate=0.1, max_depth=2)
+    model.fit(X, y)
+    predictions = model.predict(X)
+    print("Predictions:", predictions)
+
    +
    +
    +
    +
    +
    +
    +
    +
    +
    +
    +
    +
    +
    -

    -And then we are ready to go. -











    +

    Summary of course

    -

    SVMs and Regression and multiclass classification

    - -

    -More text will be added here. See also Hastie et al. section 12.3. - -











    +

    What? Me worry? No final exam in this course!

    +

    +
    +

    +
    +

    -

    Summary of course

    - -











    +

    Topics we have covered this year

    -

    What? Me worry? No final exam in this course!

    -



    - -

    -









    - -

    Topics we have covered this year

    - -

    -The course has two central parts +

    The course has two central parts

    1. Statistical analysis and optimization of data
    2. Machine learning
    -









    +

    Statistical analysis and optimization of data

    -

    Statistical analysis and optimization of data

    - -

    -The following topics have been discussed: - +

    The following topics have been discussed:

    1. Basic concepts, expectation values, variance, covariance, correlation functions and errors;
    2. Simpler models, binomial distribution, the Poisson distribution, simple and multivariate normal distributions;
      3. @@ -930,57 +1345,42 @@ The following topics have been discussed:
    3. Practical optimization using Singular-value decomposition and least squares for parameterizing data.
    4. Principal Component Analysis to reduce the number of features.
    -









    +

    Machine learning

    -

    Machine learning

    - -

    -The following topics will be covered - +

    The following topics will be covered

    1. Linear methods for regression and classification: -
      1. Ordinary Least Squares
      2. Ridge regression
      3. Lasso regression
      4. Logistic regression
      -
    2. Neural networks and deep learning: -
      1. Feed Forward Neural Networks
      2. Convolutional Neural Networks
      3. Recurrent Neural Networks
      -
    3. Decisions trees and ensemble methods: -
      1. Decision trees
      2. Bagging and voting
      3. Random forests
      4. Boosting and gradient boosting
      -
    4. Support vector machines -
      1. Binary classification and multiclass classification
      2. Kernel methods
      3. Regression
      -
    -









    +

    Learning outcomes and overarching aims of this course

    -

    Learning outcomes and overarching aims of this course

    - -

    -The course introduces a variety of central algorithms and methods +

    The course introduces a variety of central algorithms and methods essential for studies of data analysis and machine learning. The course is project based and through the various projects, normally three, you will be exposed to fundamental research problems @@ -989,6 +1389,7 @@ results. The students will learn to develop and structure large codes for studying these systems, get acquainted with computing facilities and learn to handle large scientific projects. A good scientific and ethical conduct is emphasized throughout the course. +

    • Understand linear methods for regression and classification;
      • @@ -1000,10 +1401,8 @@ ethical conduct is emphasized throughout the course.
    • Have an understanding of central algorithms used in data analysis and machine learning;
    • Work on numerical projects to illustrate the theory. The projects play a central role and you are expected to know modern programming languages like Python or C++.
    -









    - -

    Perspective on Machine Learning

    +

    Perspective on Machine Learning

    1. Rapidly emerging application area
      2. @@ -1011,20 +1410,16 @@ ethical conduct is emphasized throughout the course.
    2. Requires education/retraining for more widespread adoption
    3. A lot of “word-of-mouth” development methods
    - -Huge amounts of data sets require automation, classical analysis tools often inadequate. +

    Huge amounts of data sets require automation, classical analysis tools often inadequate. High energy physics hit this wall in the 90’s. In 2009 single top quark production was determined via Boosted decision trees, Bayesian Neural Networks, etc. +

    -











    +

    Machine Learning Research

    -

    Machine Learning Research

    - -

    -Where to find recent results: - +

    Where to find recent results:

    1. Conference proceedings, arXiv and blog posts!
    2. NIPS: Neural Information Processing Systems
      3. @@ -1033,10 +1428,8 @@ Where to find recent results:
    3. Journal of Machine Learning Research
    4. Follow ML on ArXiv
    -









    - -

    Starting your Machine Learning Project

    +

    Starting your Machine Learning Project

    1. Identify problem type: classification, regression
      2. @@ -1045,111 +1438,82 @@ Where to find recent results:
    2. Consider your data carefully again! Think of data representation more carefully.
    3. Based on your results, feedback loop to earliest possible point
    -









    - -

    Choose a Model and Algorithm

    +

    Choose a Model and Algorithm

    1. Supervised?
    2. Start with the simplest model that fits your problem
    3. Start with minimal processing of data
    -









    - -

    Preparing Your Data

    +

    Preparing Your Data

    1. Shuffle your data
    2. Mean center your data
      3. -
      • Why?
      -
    3. Normalize the variance
      4. -
      • Why?
      -
    4. Whitening
      5. -
      • Decorrelates data
      • Can be hit or miss
      -
    5. When to do train/test split?
    -









    - -

    Which Activation and Weights to Choose in Neural Networks

    +

    Which Activation and Weights to Choose in Neural Networks

    1. RELU? ELU?
    2. Sigmoid or Tanh?
    3. Set all weights to 0?
      4. -
      • Terrible idea
      -
    4. Set all weights to random values?
      5. -
      • Small random values
      -
    -









    - -

    Optimization Methods and Hyperparameters

    - +

    Optimization Methods and Hyperparameters

    1. Stochastic gradient descent -
      1. Stochastic gradient descent + momentum
      -
    2. State-of-the-art approaches:
      3. -
      • RMSProp
      • Adam
      • and more
      -
    - -Which regularization and hyperparameters? \( L_1 \) or \( L_2 \), soft +

    Which regularization and hyperparameters? \( L_1 \) or \( L_2 \), soft classifiers, depths of trees and many other. Need to explore a large set of hyperparameters and regularization methods. +

    -











    +

    Resampling

    -

    Resampling

    - -

    -When do we resample? +

    When do we resample?

    1. Bootstrap
    2. Cross-validation
    3. Jackknife and many other
    -









    +

    Other courses on Data science and Machine Learning at UiO

    -

    Other courses on Data science and Machine Learning at UiO

    - -

    -The link here https://www.mn.uio.no/english/research/about/centre-focus/innovation/data-science/studies/ gives an excellent overview of courses on Machine learning at UiO. +

    The link here https://www.mn.uio.no/english/research/about/centre-focus/innovation/data-science/studies/ gives an excellent overview of courses on Machine learning at UiO.

    1. STK2100 Machine learning and statistical methods for prediction and classification.
      2. @@ -1161,58 +1525,51 @@ The link here IN5400/INF5860 – Machine Learning for Image Analysis. An introduction to deep learning with particular emphasis on applications within Image analysis, but useful for other application areas too.
    2. TEK5040 – Dyp læring for autonome systemer. The course addresses advanced algorithms and architectures for deep learning with neural networks. The course provides an introduction to how deep-learning techniques can be used in the construction of key parts of advanced autonomous systems that exist in physical environments and cyber environments.
    -









    - -

    Additional courses of interest

    +

    Additional courses of interest

    1. STK4051 Computational Statistics
    2. STK4021 Applied Bayesian Analysis and Numerical Methods
    -









    +

    What's the future like?

    -

    What's the future like?

    - -

    -Based on multi-layer nonlinear neural networks, deep learning can +

    Based on multi-layer nonlinear neural networks, deep learning can learn directly from raw data, automatically extract and abstract features from layer to layer, and then achieve the goal of regression, classification, or ranking. Deep learning has made breakthroughs in computer vision, speech processing and natural language, and reached or even surpassed human level. The success of deep learning is mainly due to the three factors: big data, big model, and big computing. +

    -

    -In the past few decades, many different architectures of deep neural +

    In the past few decades, many different architectures of deep neural networks have been proposed, such as - +

    1. Convolutional neural networks, which are mostly used in image and video data processing, and have also been applied to sequential data such as text processing;
    2. Recurrent neural networks, which can process sequential data of variable length and have been widely used in natural language understanding and speech processing;
    3. Encoder-decoder framework, which is mostly used for image or sequence generation, such as machine translation, text summarization, and image captioning.
    -









    +

    Types of Machine Learning, a repetition

    -

    Types of Machine Learning, a repetition

    - -

    -The approaches to machine learning are many, but are often split into two main categories. +

    The approaches to machine learning are many, but are often split into two main categories. In supervised learning we know the answer to a problem, and let the computer deduce the logic behind it. On the other hand, unsupervised learning is a method for finding patterns and relationship in data sets without any prior knowledge of the system. Some authours also operate with a third category, namely reinforcement learning. This is a paradigm of learning inspired by behavioural psychology, where learning is achieved by trial-and-error, solely from rewards and punishment. +

    -

    -Another way to categorize machine learning tasks is to consider the desired output of a system. +

    Another way to categorize machine learning tasks is to consider the desired output of a system. Some of the most common tasks are: +

    • Classification: Outputs are divided into two or more classes. The goal is to produce a model that assigns inputs into one of these classes. An example is to identify digits based on pictures of hand-written ones. Classification is typically supervised learning.
      • @@ -1223,166 +1580,139 @@ Some of the most common tasks are:
    -











    +

    Why Boltzmann machines?

    -

    Why Boltzmann machines?

    - -

    -What is known as restricted Boltzmann Machines (RMB) have received a lot of attention lately. +

    What is known as restricted Boltzmann Machines (RMB) have received a lot of attention lately. One of the major reasons is that they can be stacked layer-wise to build deep neural networks that capture complicated statistics. +

    -

    -The original RBMs had just one visible layer and a hidden layer, but recently so-called Gaussian-binary RBMs have gained quite some popularity in imaging since they are capable of modeling continuous data that are common to natural images. +

    The original RBMs had just one visible layer and a hidden layer, but recently so-called Gaussian-binary RBMs have gained quite some popularity in imaging since they are capable of modeling continuous data that are common to natural images.

    -

    -Furthermore, they have been used to solve complicated quantum mechanical many-particle problems or classical statistical physics problems like the Ising and Potts classes of models. +

    Furthermore, they have been used to solve complicated quantum mechanical many-particle problems or classical statistical physics problems like the Ising and Potts classes of models.

    -











    +

    Boltzmann Machines

    -

    Boltzmann Machines

    - -

    -Why use a generative model rather than the more well known discriminative deep neural networks (DNN)? +

    Why use a generative model rather than the more well known discriminative deep neural networks (DNN)?

    • Discriminitave methods have several limitations: They are mainly supervised learning methods, thus requiring labeled data. And there are tasks they cannot accomplish, like drawing new examples from an unknown probability distribution.
    • A generative model can learn to represent and sample from a probability distribution. The core idea is to learn a parametric model of the probability distribution from which the training data was drawn. As an example -
      1. A model for images could learn to draw new examples of cats and dogs, given a training dataset of images of cats and dogs.
      2. Generate a sample of an ordered or disordered phase, having been given samples of such phases.
      3. Model the trial function for Monte Carlo calculations.
      -
    -









    - -

    Some similarities and differences from DNNs

    +

    Some similarities and differences from DNNs

    1. Both use gradient-descent based learning procedures for minimizing cost functions
    2. Energy based models don't use backpropagation and automatic differentiation for computing gradients, instead turning to Markov Chain Monte Carlo methods.
    3. DNNs often have several hidden layers. A restricted Boltzmann machine has only one hidden layer, however several RBMs can be stacked to make up Deep Belief Networks, of which they constitute the building blocks.
    +

    History: The RBM was developed by amongst others Geoffrey Hinton, called by some the "Godfather of Deep Learning", working with the University of Toronto and Google.

    -History: The RBM was developed by amongst others Geoffrey Hinton, called by some the "Godfather of Deep Learning", working with the University of Toronto and Google. - -











    +

    Boltzmann machines (BM)

    -

    Boltzmann machines (BM)

    - -

    -A BM is what we would call an undirected probabilistic graphical model +

    A BM is what we would call an undirected probabilistic graphical model with stochastic continuous or discrete units. +

    -It is interpreted as a stochastic recurrent neural network where the +

    It is interpreted as a stochastic recurrent neural network where the state of each unit(neurons/nodes) depends on the units it is connected to. The weights in the network represent thus the strength of the interaction between various units/nodes. +

    -It turns into a Hopfield network if we choose deterministic rather +

    It turns into a Hopfield network if we choose deterministic rather than stochastic units. In contrast to a Hopfield network, a BM is a so-called generative model. It allows us to generate new samples from the learned distribution. +

    -











    +

    A standard BM setup

    -

    A standard BM setup

    - -

    -A standard BM network is divided into a set of observable and visible units \( \hat{x} \) and a set of unknown hidden units/nodes \( \hat{h} \). +

    A standard BM network is divided into a set of observable and visible units \( \hat{x} \) and a set of unknown hidden units/nodes \( \hat{h} \).

    -

    -Additionally there can be bias nodes for the hidden and visible layers. These biases are normally set to \( 1 \). +

    Additionally there can be bias nodes for the hidden and visible layers. These biases are normally set to \( 1 \).

    -

    -BMs are stackable, meaning they cwe can train a BM which serves as input to another BM. We can construct deep networks for learning complex PDFs. The layers can be trained one after another, a feature which makes them popular in deep learning +

    BMs are stackable, meaning they cwe can train a BM which serves as input to another BM. We can construct deep networks for learning complex PDFs. The layers can be trained one after another, a feature which makes them popular in deep learning

    -

    -However, they are often hard to train. This leads to the introduction of so-called restricted BMs, or RBMS. +

    However, they are often hard to train. This leads to the introduction of so-called restricted BMs, or RBMS. Here we take away all lateral connections between nodes in the visible layer as well as connections between nodes in the hidden layer. The network is illustrated in the figure below. +

    -











    +

    The structure of the RBM network

    -

    The structure of the RBM network

    +

    +
    +

    +
    +

    -

    -



    - -











    +

    The network

    -

    The network

    - -

    The network layers: -

    1. A function \( \mathbf{x} \) that represents the visible layer, a vector of \( M \) elements (nodes). This layer represents both what the RBM might be given as training input, and what we want it to be able to reconstruct. This might for example be given by the pixels of an image or coefficients representing speech, or the coordinates of a quantum mechanical state function.
    2. The function \( \mathbf{h} \) represents the hidden, or latent, layer. A vector of \( N \) elements (nodes). Also called "feature detectors".
    -









    +

    Goals

    -

    Goals

    - -

    -The goal of the hidden layer is to increase the model's expressive +

    The goal of the hidden layer is to increase the model's expressive power. We encode complex interactions between visible variables by introducing additional, hidden variables that interact with visible degrees of freedom in a simple manner, yet still reproduce the complex correlations between visible degrees in the data once marginalized over (integrated out). +

    -

    The network parameters, to be optimized/learned: -

    1. \( \mathbf{a} \) represents the visible bias, a vector of same length as \( \mathbf{x} \).
    2. \( \mathbf{b} \) represents the hidden bias, a vector of same lenght as \( \mathbf{h} \).
    3. \( W \) represents the interaction weights, a matrix of size \( M\times N \).
    -









    +

    Joint distribution

    -

    Joint distribution

    - -

    -The restricted Boltzmann machine is described by a Boltzmann distribution +

    The restricted Boltzmann machine is described by a Boltzmann distribution

    $$ \begin{align} P_{rbm}(\mathbf{x},\mathbf{h}) = \frac{1}{Z} e^{-\frac{1}{T_0}E(\mathbf{x},\mathbf{h})}, @@ -1390,7 +1720,7 @@ $$ \end{align} $$ -where \( Z \) is the normalization constant or partition function, defined as +

    where \( Z \) is the normalization constant or partition function, defined as

    $$ \begin{align} Z = \int \int e^{-\frac{1}{T_0}E(\mathbf{x},\mathbf{h})} d\mathbf{x} d\mathbf{h}. @@ -1398,49 +1728,39 @@ $$ \end{align} $$ -It is common to ignore \( T_0 \) by setting it to one. +

    It is common to ignore \( T_0 \) by setting it to one.

    -











    +

    Network Elements, the energy function

    -

    Network Elements, the energy function

    - -

    -The function \( E(\mathbf{x},\mathbf{h}) \) gives the energy of a +

    The function \( E(\mathbf{x},\mathbf{h}) \) gives the energy of a configuration (pair of vectors) \( (\mathbf{x}, \mathbf{h}) \). The lower the energy of a configuration, the higher the probability of it. This function also depends on the parameters \( \mathbf{a} \), \( \mathbf{b} \) and \( W \). Thus, when we adjust them during the learning procedure, we are adjusting the energy function to best fit our problem. +

    -

    -An expression for the energy function is +

    An expression for the energy function is

    $$ E(\hat{x},\hat{h}) = -\sum_{ia}^{NA}b_i^a \alpha_i^a(x_i)-\sum_{jd}^{MD}c_j^d \beta_j^d(h_j)-\sum_{ijad}^{NAMD}b_i^a \alpha_i^a(x_i)c_j^d \beta_j^d(h_j)w_{ij}^{ad}. $$ -

    -Here \( \beta_j^d(h_j) \) and \( \alpha_i^a(x_j) \) are so-called transfer functions that map a given input value to a desired feature value. The labels \( a \) and \( d \) denote that there can be multiple transfer functions per variable. The first sum depends only on the visible units. The second on the hidden ones. Note that there is no connection between nodes in a layer. +

    Here \( \beta_j^d(h_j) \) and \( \alpha_i^a(x_j) \) are so-called transfer functions that map a given input value to a desired feature value. The labels \( a \) and \( d \) denote that there can be multiple transfer functions per variable. The first sum depends only on the visible units. The second on the hidden ones. Note that there is no connection between nodes in a layer.

    -

    -The quantities \( b \) and \( c \) can be interpreted as the visible and hidden biases, respectively. +

    The quantities \( b \) and \( c \) can be interpreted as the visible and hidden biases, respectively.

    -

    -The connection between the nodes in the two layers is given by the weights \( w_{ij} \). +

    The connection between the nodes in the two layers is given by the weights \( w_{ij} \).

    -











    +

    Defining different types of RBMs

    +

    There are different variants of RBMs, and the differences lie in the types of visible and hidden units we choose as well as in the implementation of the energy function \( E(\mathbf{x},\mathbf{h}) \).

    -

    Defining different types of RBMs

    -There are different variants of RBMs, and the differences lie in the types of visible and hidden units we choose as well as in the implementation of the energy function \( E(\mathbf{x},\mathbf{h}) \). - -

    Binary-Binary RBM:

    -

    -RBMs were first developed using binary units in both the visible and hidden layer. The corresponding energy function is defined as follows: +

    RBMs were first developed using binary units in both the visible and hidden layer. The corresponding energy function is defined as follows:

    $$ \begin{align} E(\mathbf{x}, \mathbf{h}) = - \sum_i^M x_i a_i- \sum_j^N b_j h_j - \sum_{i,j}^{M,N} x_i w_{ij} h_j, @@ -1448,15 +1768,14 @@ $$ \end{align} $$ -where the binary values taken on by the nodes are most commonly 0 and 1. +

    where the binary values taken on by the nodes are most commonly 0 and 1.

    Gaussian-Binary RBM:

    -

    -Another varient is the RBM where the visible units are Gaussian while the hidden units remain binary: +

    Another varient is the RBM where the visible units are Gaussian while the hidden units remain binary:

    $$ \begin{align} E(\mathbf{x}, \mathbf{h}) = \sum_i^M \frac{(x_i - a_i)^2}{2\sigma_i^2} - \sum_j^N b_j h_j - \sum_{i,j}^{M,N} \frac{x_i w_{ij} h_j}{\sigma_i^2}. @@ -1466,45 +1785,36 @@ $$
    -











    - -

    More about RBMs

    - +

    More about RBMs

    1. Useful when we model continuous data (i.e., we wish \( \mathbf{x} \) to be continuous)
    2. Requires a smaller learning rate, since there's no upper bound to the value a component might take in the reconstruction
    - -Other types of units include: - +

    Other types of units include:

    1. Softmax and multinomial units
    2. Gaussian visible and hidden units
    3. Binomial units
    4. Rectified linear units
    +

    To read more, see Lectures on Boltzmann machines in Physics.

    -To read more, see Lectures on Boltzmann machines in Physics. - -











    +

    Autoencoders: Overarching view

    -

    Autoencoders: Overarching view

    - -

    -Autoencoders are artificial neural networks capable of learning +

    Autoencoders are artificial neural networks capable of learning efficient representations of the input data (these representations are called codings) without any supervision (i.e., the training set is unlabeled). These codings typically have a much lower dimensionality than the input data, making -autoencoders useful for dimensionality reduction. +autoencoders useful for dimensionality reduction. +

    -

    -More importantly, autoencoders act as powerful feature detectors, and +

    More importantly, autoencoders act as powerful feature detectors, and they can be used for unsupervised pretraining of deep neural networks. +

    -

    -Lastly, they are capable of randomly generating new data that looks +

    Lastly, they are capable of randomly generating new data that looks very similar to the training data; this is called a generative model. For example, you could train an autoencoder on pictures of faces, and it would then be able to generate new faces. Surprisingly, @@ -1518,57 +1828,49 @@ autoencoder from trivially copying the inputs directly to the outputs, which forces it to learn efficient ways of representing the data. In short, the codings are byproducts of the autoencoder’s attempt to learn the identity function under some constraints. +

    -

    Video on autoencoders -

    -See also A. Geron's textbook, chapter 15. +

    See also A. Geron's textbook, chapter 15.

    -











    +

    Bayesian Machine Learning

    -

    Bayesian Machine Learning

    - -

    -This is an important topic if we aim at extracting a probability +

    This is an important topic if we aim at extracting a probability distribution. This gives us also a confidence interval and error estimates. +

    -

    -Bayesian machine learning allows us to encode our prior beliefs about +

    Bayesian machine learning allows us to encode our prior beliefs about what those models should look like, independent of what the data tells us. This is especially useful when we don’t have a ton of data to confidently learn our model. +

    -

    Video on Bayesian deep learning -

    -See also the slides here. +

    See also the slides here.

    -











    +

    Reinforcement Learning

    -

    Reinforcement Learning

    - -

    -Reinforcement Learning (RL) is one of the most exciting fields of +

    Reinforcement Learning (RL) is one of the most exciting fields of Machine Learning today, and also one of the oldest. It has been around since the 1950s, producing many interesting applications over the years. +

    -

    -It studies +

    It studies how agents take actions based on trial and error, so as to maximize some notion of cumulative reward in a dynamic system or environment. Due to its generality, the problem has also been studied in many other disciplines, such as game theory, control theory, operations research, information theory, multi-agent systems, swarm intelligence, statistics, and genetic algorithms. +

    -

    -In March 2016, AlphaGo, a computer program that plays the board game +

    In March 2016, AlphaGo, a computer program that plays the board game Go, beat Lee Sedol in a five-game match. This was the first time a computer Go program had beaten a 9-dan (highest rank) professional without handicaps. AlphaGo is based on deep convolutional neural @@ -1576,20 +1878,16 @@ networks and reinforcement learning. AlphaGo’s victory was a major milestone in artificial intelligence and it has also made reinforcement learning a hot research area in the field of machine learning. +

    -

    -Lecture on Reinforcement Learning. +

    Lecture on Reinforcement Learning.

    -

    -See also A. Geron's textbook, chapter 16. +

    See also A. Geron's textbook, chapter 16.

    -











    +

    Transfer learning

    -

    Transfer learning

    - -

    -The goal of transfer learning is to transfer the model or knowledge +

    The goal of transfer learning is to transfer the model or knowledge obtained from a source task to the target task, in order to resolve the issues of insufficient training data in the target task. The rationality of doing so lies in that usually the source and target @@ -1597,17 +1895,14 @@ tasks have inter-correlations, and therefore either the features, samples, or models in the source task might provide useful information for us to better solve the target task. Transfer learning is a hot research topic in recent years, with many problems still waiting to be studied. +

    -

    -Lecture on transfer learning. +

    Lecture on transfer learning.

    -











    +

    Adversarial learning

    -

    Adversarial learning

    - -

    -The conventional deep generative model has a potential problem: the +

    The conventional deep generative model has a potential problem: the model tends to generate extreme instances to maximize the probabilistic likelihood, which will hurt its performance. Adversarial learning utilizes the adversarial behaviors (e.g., generating @@ -1616,17 +1911,14 @@ robustness of the model and improve the quality of the generated data. In recent years, one of the most promising unsupervised learning technologies, generative adversarial networks (GAN), has already been successfully applied to image, speech, and text. +

    -

    -Lecture on adversial learning. +

    Lecture on adversial learning.

    -











    +

    Dual learning

    -

    Dual learning

    - -

    -Dual learning is a new learning paradigm, the basic idea of which is +

    Dual learning is a new learning paradigm, the basic idea of which is to use the primal-dual structure between machine learning tasks to obtain effective feedback/regularization, and guide and strengthen the learning process, thus reducing the requirement of large-scale labeled @@ -1635,42 +1927,35 @@ many problems in machine learning, including machine translation, image style conversion, question answering and generation, image classification and generation, text classification and generation, image-to-text, and text-to-image. +

    -











    +

    Distributed machine learning

    -

    Distributed machine learning

    - -

    -Distributed computation will speed up machine learning algorithms, +

    Distributed computation will speed up machine learning algorithms, significantly improve their efficiency, and thus enlarge their application. When distributed meets machine learning, more than just implementing the machine learning algorithms in parallel is required. +

    -











    +

    Meta learning

    -

    Meta learning

    - -

    -Meta learning is an emerging research direction in machine +

    Meta learning is an emerging research direction in machine learning. Roughly speaking, meta learning concerns learning how to learn, and focuses on the understanding and adaptation of the learning itself, instead of just completing a specific learning task. That is, a meta learner needs to be able to evaluate its own learning methods and adjust its own learning methods according to specific learning tasks. +

    -











    +

    The Challenges Facing Machine Learning

    -

    The Challenges Facing Machine Learning

    +

    While there has been much progress in machine learning, there are also challenges.

    -

    -While there has been much progress in machine learning, there are also challenges. - -

    -For example, the mainstream machine learning technologies are +

    For example, the mainstream machine learning technologies are black-box approaches, making us concerned about their potential risks. To tackle this challenge, we may want to make machine learning more explainable and controllable. As another example, the @@ -1685,27 +1970,23 @@ models for data fitting? Although there are many challenges, we are still very optimistic about the future of machine learning. As we look forward to the future, here are what we think the research hotspots in the next ten years will be. +

    -

    -See the article on Discovery of Physics From Data: Universal Laws and Discrepancies +

    See the article on Discovery of Physics From Data: Universal Laws and Discrepancies

    -











    +

    Explainable machine learning

    -

    Explainable machine learning

    - -

    -Machine learning, especially deep learning, evolves rapidly. The +

    Machine learning, especially deep learning, evolves rapidly. The ability gap between machine and human on many complex cognitive tasks becomes narrower and narrower. However, we are still in the very early stage in terms of explaining why those effective models work and how they work. +

    -

    -What is missing: the gap between correlation and causation. Standard Machine Learning is based on what e have called a frequentist approach. +

    What is missing: the gap between correlation and causation. Standard Machine Learning is based on what e have called a frequentist approach.

    -

    -Most +

    Most machine learning techniques, especially the statistical ones, depend highly on correlations in data sets to make predictions and analyses. In contrast, rational humans tend to reply on clear and trustworthy @@ -1713,48 +1994,41 @@ causality relations obtained via logical reasoning on real and clear facts. It is one of the core goals of explainable machine learning to transition from solving problems by data correlation to solving problems by logical reasoning. +

    -

    Bayesian Machine Learning is one of the exciting research directions in this field. -











    +

    Quantum machine learning

    -

    Quantum machine learning

    - -

    -Quantum machine learning is an emerging interdisciplinary research +

    Quantum machine learning is an emerging interdisciplinary research area at the intersection of quantum computing and machine learning. +

    -

    -Quantum computers use effects such as quantum coherence and quantum +

    Quantum computers use effects such as quantum coherence and quantum entanglement to process information, which is fundamentally different from classical computers. Quantum algorithms have surpassed the best classical algorithms in several problems (e.g., searching for an unsorted database, inverting a sparse matrix), which we call quantum acceleration. +

    -

    -When quantum computing meets machine learning, it can be a mutually +

    When quantum computing meets machine learning, it can be a mutually beneficial and reinforcing process, as it allows us to take advantage of quantum computing to improve the performance of classical machine learning algorithms. In addition, we can also use the machine learning algorithms (on classic computers) to analyze and improve quantum computing systems. +

    -

    -Lecture on Quantum ML. +

    Lecture on Quantum ML.

    -

    -Read interview with Maria Schuld on her work on Quantum Machine Learning. See also her recent textbook. +

    Read interview with Maria Schuld on her work on Quantum Machine Learning. See also her recent textbook.

    -











    +

    Quantum machine learning algorithms based on linear algebra

    -

    Quantum machine learning algorithms based on linear algebra

    - -

    -Many quantum machine learning algorithms are based on variants of +

    Many quantum machine learning algorithms are based on variants of quantum algorithms for solving linear equations, which can efficiently solve N-variable linear equations with complexity of O(log2 N) under certain conditions. The quantum matrix inversion algorithm can @@ -1766,28 +2040,24 @@ machine learning algorithms is data input—that is, how to initialize the quantum system with the entire data set. Although efficient data-input algorithms exist for certain situations, how to efficiently input data into a quantum system is as yet unknown for most cases. +

    -











    +

    Quantum reinforcement learning

    -

    Quantum reinforcement learning

    - -

    -In quantum reinforcement learning, a quantum agent interacts with the +

    In quantum reinforcement learning, a quantum agent interacts with the classical environment to obtain rewards from the environment, so as to adjust and improve its behavioral strategies. In some cases, it achieves quantum acceleration by the quantum processing capabilities of the agent or the possibility of exploring the environment through quantum superposition. Such algorithms have been proposed in superconducting circuits and systems of trapped ions. +

    -











    +

    Quantum deep learning

    -

    Quantum deep learning

    - -

    -Dedicated quantum information processors, such as quantum annealers +

    Dedicated quantum information processors, such as quantum annealers and programmable photonic circuits, are well suited for building deep quantum networks. The simplest deep quantum network is the Boltzmann machine. The classical Boltzmann machine consists of bits with tunable @@ -1799,14 +2069,12 @@ correspond to an adjustable Ising model. Then, by initializing the input neurons in the Boltzmann machine to a fixed state and allowing the system to heat up, we can read out the output qubits to get the result. +

    -











    +

    Social machine learning

    -

    Social machine learning

    - -

    -Machine learning aims to imitate how humans +

    Machine learning aims to imitate how humans learn. While we have developed successful machine learning algorithms, until now we have ignored one important fact: humans are social. Each of us is one part of the total society and it is difficult for us to @@ -1814,41 +2082,33 @@ live, learn, and improve ourselves, alone and isolated. Therefore, we should design machines with social properties. Can we let machines evolve by imitating human society so as to achieve more effective, intelligent, interpretable “social machine learning”? +

    -

    -And much more. +

    And much more.

    -











    +

    The last words?

    -

    The last words?

    - -

    -Early computer scientist Alan Kay said, The best way to predict the +

    Early computer scientist Alan Kay said, The best way to predict the future is to create it. Therefore, all machine learning practitioners, whether scholars or engineers, professors or students, need to work together to advance these important research topics. Together, we will not just predict the future, but create it. +

    -











    +

    Best wishes to you all and thanks so much for your heroic efforts this semester

    -

    Best wishes to you all and thanks so much for your heroic efforts this semester

    - -

    -



    - -

    +

    +

    +

    +
    +

    - -
    - © 1999-2020, Morten Hjorth-Jensen. Released under CC Attribution-NonCommercial 4.0 license + © 1999-2024, Morten Hjorth-Jensen. Released under CC Attribution-NonCommercial 4.0 license
    - - - diff --git a/doc/pub/week48/ipynb/ipynb-week48-src.tar.gz b/doc/pub/week48/ipynb/ipynb-week48-src.tar.gz index 40e3c4820fa0e38f99784df238f7b7e8175a423c..fde56c91771eef7fc9111824b4038d08c8f5f79b 100644 GIT binary patch literal 823660 zcmV(vKrJLZrxM2?ydJun_F7AI+;0t zVEb(5VPp$pvvM%A0Q`%=&d$!m#YOcG%FDz3pC~)Wf4=`u;N;+;0&((k^K$TVgFxI= z>>zF)5C}lU{;yW#zXexk7bB-vRcyalnb|p;+5KPjd+la!{_n{7&r_lLFZ6#FAQB)0 z-~j-6BNHk`RVpp2f8rlJz?;{?@mdOA%Rlbt{>Ma$sHv&^^XEUFNPR6I|4kRT|H{WV zvA1P2a&WLQV{>#hax$`Wv9vP-IDTMf2ZJCy?Ci$oU`|em^A{s0OFIiw2p^9qJBPT4 z7%Nyzl#7*1LX4MHR0IrSWf$e-5EB&xaf(ZT{{?3GUj^p>^nWEI4=FPvQ!}Rz|N4F4 zzw7^;od3rEc{#Z_0aPCUDh2*M{4ehRAPx>HIcZT5l}}RA8scmoMlLQ+Z2v8f6qFYe z;gVN#=8zW?wbZe)fXYK%phl7so<^G7b~<9t+zQsx?()_eqAH)*xm3iZ+2zz!Bvr&r zT$HWFx#U&Z-IcB6JypbI#O2i#Bvh;{Q2BVNR9^>`)Bg7c{lBpZp#Fbt+W#r=zx)6H zbJ735-@6Z~&>-3-EUt@CkqjL_k0Q zBED9Lh=@qYXeh|9zC06aDv5F74qAAtNde?S1j z!Tlem0YpGVf`dm!dGp#Vi3Na%1HvQxLv?sW_*ebm;DG=HY{Yj+I3N*ZTxBC_NB=k! zJdRv^8c~&+?q39)o97p_#%~BeIT6vB#Mg4Eng-;(3L<$O0toki8s}Bg>qxI{bl9(a zczAf=YYl|^XB#%~9S#TqS45fG2#>=N@$V7;}XL@0o*yU){ih*~gB-n6^4pg34&0#Eh$tz~c0YY{PBR zD)0ItF;@vg3cthxf~~Cic+-$jCRI6w8;Q|6GaEsgDwf9sh}`*wP{QQjK32rfB&zGa9v&k!Y^ZWum-sb?FBeoHR)+| z^M%p?ep=-w8nvEj4;P})`a%7d0x3Wu(aKtkS&1d^jY@cs(n|nRKCcFbg3Xfbr6|sfd4||R?%7qXg;viYPw!) z(yEpvA7EmTHb$_3OgV|Zp`Gg2A`5K53}~U+D?-o!jaa_}kSZ;|8*q*)5$<=Vgf@g0 zESLq~<7cTzj4_xE(SQZZrPSlX%!jV=elAo#!1N_`hKG=iou%uZ=;?hLTuvI9HGz1k ziDWz56!qzH0xRgMLg~xkz^;-`y8p2ITE&0KLyGz4@ z$65>}t;jt?xvG@|AUj(Prc5EZe8;M&vdIZ?0a>b3Np@y;4=M_Lx73B`9|=}5o~;S6 zIghYFFMk0yg(_5jTglakc1DMfe*sJ7@xFxc(#8cYd4a8I%$ReZ^THWQrCuvkhCILj z5CK={Bx=+Me8XgaO!G2vyIK;+@q052pKCTfzM{N~ViIj~tRYA*(-^!Up-X%uXl>P2 z9aQ+d%@368pATVMoVXRt3{7(F-)&4VMIe#QS zU!`^(<);&cp0CuQ^vdGkyZ!yHi9fdmY%D38uc-jAnf!oEPu^azTDK%rbD5^(t|Y(r z(Pnc?(~`zPnEp|k-11~>90izy9@Rx|{gq#p3LJ_y^@vu{tOJo3gt2NR4+!+mrHl7c zb0g{QvlYa1Hruqi!}ZwQrHivL!s8W%pPA1LF90RQftZB09WP||83`5!ff zv@Mh15kGdRo8sH^&a}5O!UKJG`0uoerOZs7Kv_5M;3N&qy77hFUB%bPcn5Y|+<7{f5A z6G;fC@X0)5(5FNw)epmO=HZj1NZ1jxG<2?%WB5noT2dXn!(i@-#=wY{l51DZ(u)KSv5|1HM`RtiR-?V1*1*X*AYgyAu`*N@(U^ zSiasmWR{WmDQx=zE$P0%LXUO>v5e%bJOU*{Njqax&{tb;KdDt{2ha1%+ zNqxn^=adZTFIAKbrskwNf1?k-u(cbqNo;9U>ankg~g^~uHDUQ22Li{Ek>*MMw zi}cQC+9zJPb!`gYa0rbho*k8r%e(Y?LLU6K@sgHhV#pi=WHCBc4{bT+Ek6YQ0D$5h z6=~^CzS9H8g>`{2^GyQ*PwbNnQ&n{1O_!22^{@gPL*q247yn4C6PujDa$ZWTOIF1SyBfrj6dRF~Xao^)B0DWJ z9Eba8$-9-AV`rcI1<1(C*J;HjtLWBn!LI4@2(jN2E4(2;5{nfMVU3rx*HM!mL4w0u z8WCXuwWyGi?qv)&Be>-;N{GiIeVxob5Er(F(y;!Vx!#~j^eNB}BR_)j4gpT0O6`>W zBgQ{!UE@;V#Jaw0HIR%2!gMTgd= zhSc!7gwc*JP8I~d<{lp&r|}VBWxKv@J%d~{9zj%#Bmrwb!kw zjg4-FKF*+A(EyU)DJm=c>KY*BJJfZ&)Ee2SLBnG~S~%7w0&=6m6$kOS2z1@zhY=ceLiPUKRB%M9 zJwHL($l0s`-xR z(tWW4B!~dKP^UUw_Rpcfb?AVziLbuuThbSUH=3zAdp9NNLncu}G~PQDH#FQcj%2wA zYM3}H`C=4|m>`J1=64g>s35Z>6JePv3csw@Uqzp#n&?c?ki53ao2op~n}2wsL?MG7 zq7IW%u=Vu~265g|i}DUWKh8Xu{87KCtZr<)>Zv>+V<*LKuwuy-^)5gW1T3Xl5>_-? z(GfA%ATMQ!l+Z>x@Zf)DasPdh2LVg*DPcN6re&8k#P))O2bEW50w$^+-H_L6C3$!z zbFWJ;2QTxs)xTF2MRP=s9<(U#&6`=+ZjdBM_TQqD zn#T<`gb8fgsT_d|X9_2-gu8nh^yD!*h3KLR4ob0->Nlm#Z*$}5KhnM|W%!6I&a~Bb#v-CNBuy4NBw z<7Px1qw&F=3LaZrarj^%xjkr97RS~=ro6uHaj+&$-A^7&HL~=EZ&a=Ct5jZU&rw?BVWCi zA@ntl(JG7My*jV(|1c35)!68G=(jPC3+YmjY!~-zHKSD#ER{MPW5ElMMGrmv47xRY zFzD+cw3Yc%xNEZmo!;$?>SFV2gv_8nt?rGQ$cJqk6c{}GV^an zvqu*|r{Ye}qHfj`-Y@xS2Tv+|FvpX2=ZO1R3%0=|wYxI|#gQg^)j(AuduIptC<)?5 zr21B0I=uQ+T8(+8I->-mV`)9=Rt6ecckVsi9|uva2(>FAnzY76+tOl^qbcZzTGRQ$ zYR2r2jhu?%UEd<3opp@VIv$x(o%-V?KQhxcC2Dsn9WZuCn*?`Lv@>g35-aw>A%kx! z+wA3j*H#B>XFRk3TwuGV7$u4Z^{Zy?HZIgn-dJylj8^9KzH9RvxG9`h^-~%$@oy4s zo11z&CDT_j7R<{Ii69udRO;1r!e|e^akgnvIm7q2h+N}y8V(Ciiz$`o4B4$R z72DDXKCC3sSLK$oXv$-ejljnRNa~nm`%KSj*(|DUW*16{bIEMRi;OYlxQYcMWiB{X zx}AzkIqa1eAF~s)+F=FJvQKP8=igSYvHY&EN%F03kKSSr3Slo?T_BI1)%6!hM#l?0 zQ;GY`kG;nx34>;M*G}5m+@IdPJB)4YOQUwBJ(>D$MUvK!%?QMi=rb`<%J@*lmRqK1 zLz-IT8G6v}#@#4Td(-Gn`dG&q;2b{W#E-hD?CIHxm`G@F(iZ|Ez@xrt&-SBB#39Yj-mENsh+39f}^Cox<>x+c`z|O#rHEYPH{Y?*V)0n5{DUIMjikO zUEJa)5Fk3W=(4s+*a-s(GFu$ftX{Q6$h$I3;e&WXe!S~(&pCc*?%+dF!sf5f<`QZK{7UeEkaQzFYQ`I=vU0SNZD#6a_0zgrJZ{HK%Wa&L# z@7(>;*OsGd*T4Z0ftM2F#Yf(APW>Rn)z)1<949L&>k5EHb(14HI|)P-X>-U~oM~S^ zd4&|Ud3t*B&5C$&bb01+8DqIl@W0HW8=B96ty7IjgA+?W%>!bxtqKfM!QbRm!>P%IrO6h)29p06U0FJn;qm!;nk*e|x>)DiCzEKEt zKIr(I>@bht?$LDPU0m(AKmT7u_C9M7~z<%zrzSFw4_6gc8`ra_{ z9s{NlZ{?Uf%W^t@r}TvAB|ISnb)>vxg~UF0y)n1wlgELH8(E>_OS@2I=V2Y0gh zA9mzdQ;lfClT(S{NtroO_WWmm{D`mRKznKqvEN|JDN$DEHY`=T*!K1-N%@ZApY{vA)Y*B_b`-a-q zv$1zrKfm7-A&xh>&`qcw0Ik4vK<8Ue3ZgcmJqk21a)OcH`JXcHrEFP<)@)Xh+(g*Y zvM4nSC(-DC(-JE9THfH%v_(kfaJD(!TS>F|Le)CGH?xH`VX`}lRq5n}nNvFDa?2h| zwmB0wvrpxHAWh@F8M||;wuAeb&qJ|4hS-$?Yo@-PWb5%ynu*}}wFf8FE->--?a%8o zmCj&a!sflnepY6&j@kXr+8K*W^WcpDLStznz&;JRqddu1BsX=kJyASYK7h}4hNedJ zu1QNw>R*7AEB(`aj1W-=yY*;ajRPUtdP9cWZ-QH`Xj)H9sR_3Ml#eas*71(}h#n!E zEIJl=iv6hO9aN=FsdWPi3Pe1MUE>mQ#Zv?-i`|h=SGre>`qPQ-v?qFA=QUfllVN0t z`&QU-rIwXcElv{Tp+iE(OvCqqGgbAzNZ@kyX||pE_WBx6!|(k0@az&!kTmH$mk#Cu z1^izjdcY2A^CS-ZEOsxn`4C^^4(f|0?E9iy{=WchgZ0I931_k(2Zs^PYKjKJ!48^> z*^n)s?N4=E5u3$ykJQ4bmzPNh(OE@-YlTU?#^tAj#1&+e5eX|;Zrs~{0RwmB7xrk@ zWX!g`k0-O=GGMeG{aZ0jI}!qh7qL?}iB;>F~GrDG%u|{`8X?Ue6uqbIm!Xj8UN9VtpG~559(}cu5sB z#>NqM8KeGjUHkEQ!{_wzPtSL5d6iq;@8-=8lg#h{wBD+b=E3bwt)?v$FtS!tFY8_= z*I>${RwC`!H#CQO`+m(${nB7;4-Z2xJ*=`4T zQFwuIg4uLPr7M{nboiOOjpHJ{?<2`Jsd)3ey&db+2Hki?0sbI68pf)`HtZ|KuPTY$ zLTxJm9w9qz7v=K^=Hpc9qu1+R@z2wi$k;u7$Xhti5>dBgQR0OK9!N8k)2Mz~$iD71lNui~HA6CmU zkoy;mGr?z~%3Run#X$|r0TQRqMpxU|_T%#~JFfzVkwH8Eqt6Y<{UV077Ax}Pcqu*@ zqDxG2Gm|V8{pN_L@CYfwAxHL5^E49sy}u}$}vxQ5p{q2vbN$d`^yiRZ&wH#NOTgZ`5DRU zM};RS?w}m<7ceG9!u}Q*&mUgtI2S`uXwSOBRwIoC$%>Rz*89^V_$l1B*$sT38KK8a zhCS;ysP_!j{FJ4u?XNe}yB1VP$)t~BFVRkv1)6NE#pn$rf&x_~?Z^fgq!HFs0 z>`g!8ef-l0-y<{6akpT-S6}?3`Nf86*N+zM(Y*@qBKdotySKU!+9wMSv-q99un`o< zOT}Nny39@IW@k_86W3<#9gb&O!%@t5=j}E4@JI-H+FPr6+pu%^s#>{z$T7nEwHKEwa$#A}(oa+Nd~un3)1VgGaiJ6Yfy#pKFFS-yhy z3Xa|}L*;brWF#N{N6l;`9ZWg8FC#{$7kBdfW)LdB&IDL6Pj*5rAtgJt;)s85Hy1hK z#6xr}_d&a0q>2q0s&gbE3EBIeYwno!jbwlFee-qL)*DlTAQLc^Vcb#_+>j~THpE`% zsB!mKAn0x8VOA!yxJ!-y`95(4ZThk-FQK(1bwZk?VJ^zbT_J5-02pI4q@={@SD7;b zt&J3ZVPDy~FDnLUsww4ljjQPmna4KsDNE5!o#y(mY#-{ZunYi)x@@Pu>v9D>5Cu5T z<@LI9l6Pwj(GZtBOO6Wk$I8@3YiJNF@v9a#8Tx;(Wh5<<7}++%Z^EBl%1Cg0iyr(P zwIyJVLI3hC)qvo*b%9mg0ZFXua8xvgI|B5%bah#b2EEpQd9Lg+x)jY^IG)j13RPjN z-GI?)IwYYuzZs%iJNr97dNKAR=Ln|Wsn6_3P~?7*d z_zwRl1_eAviPijm*`Of`7C|VfB2>ST_3U2x5S@?G1On1S++ZZ)M$txEPEtVH7}d1~ z6qM4b<-;caz#9A*cGGyb>xv+QoBT~vT04}bLW#;xWbvQ{l~^nbJwFV~mWAXA2_F3R zWVs{VN-NeU@zN=FiRY3E`|O;GS@pbdh8!l|u=<0Buc%k@-#@(#)d$=$3I8D@iLW5r zw;^*Se@Ucd14l?~eL#Lv-ZhP9hVk#GEt%8~<s5en9({9>Vq)5Sco} z66Bq8Kj3aN?{lHp^FnZe(WQGlCqws=p4zpq9KCar)m|Pdw9y;1bj{%?QNqf9iklJh z7huQamf2Q^wM9uFpn&^gAe$+v^5b>&;^5&Xyr4rRFFJF@dwq}{c>L`oS%WSj>;kbt z%V+w30X@ENz(T#&4rz3u@rMaUsdCAKa$-Eh+1T=oKa%F0~oK8U?3H%Ms zDl7dmaToCXBsR-|KJf9B_sHRkCvy^`{z(Yiv9<;t=bzI50f!FnagWH}gt?aCgC=cPRYLIcSe%GOoK9E#RX7E&3K zLkbwnXxqLzHp^qa5mE4e0pb^v8Z(v2YJsXrcD4nB*FoyEU)5%2LqcBgtui>c+r9dO zueAB1N}Nn3UZzlEF_29nHEW+#&pU$!B&Tua^%~)3oO+#!?&r^%V2Hp&d@_vFc@BJoKB8`{5pT4(vSG~Gb3<6=+xB<`T2zqPK4L-Gd>W>}+ro`o z9XK=h&_m&Xm&IH|-})O$N4Yle28__J*@_iYA}Sdr_-*Hr^E||M4Mr}k+Co`!OEI(9 zG!_NaoD;ygNVcSH{Al}akI%>`+G7|2vxGXQ!K9yL{A4~fcuCo2zKi=OC90GN)0+p> zyW&V6z-f0Q|a!C}_N6)_y7R(V7N@1ygOpl0R>u4o1N3-oKuQTV*ojK~1#Avh_TdBNlId%;cK z;K#%^(I@f&!XJ$=drP`D5HCG*GT@30k31CUtJ8A49yo#5HdB}_J6{Dejtl!DtS6Z@ zTJt>a;)^JNwyF5mp5l*O^3my|^6*1?1`^;#i4cekESY#U?&K;^As7?u{ZvDgf-Mer znp9CMLaAZ|LgS!q*n*~R^Z96Z(2UC;uXVi2D)Tm1B||ehCtuAAxGu_&v1f5YeYb+E~rqm&2lknShy9(*Etae+cloU0=?hd_42SCU~DTMXO z@e9-!bJ+nYsaVFmA-KngEYgRbY&fuSIqIlp^Fz3|ZLn z*WjV`szT`WKqoxdzV9l(BfsQ`#9<1yw1(gCCRlX@$l-Dx~VV37%ZY78%Jb!V8ghA}elH z`8>(hOS;_%R46A~T3f(cKftqJC3<&fFa_!6N0yi{E+4wN+$_GaJOb=*%;WK`Y>rVs z=DiS-k++zTkU$BiR!peB{>vL`tSPw>(UWjXNEL{>A=$vapa22hvW0ZYHseu6h5UB- zVX8uOKweVXt!!{#eUvQ`5^owU6aje5%1ie#eSRkJE~AN$G&0sv{==T)yV_4= z?kBnzTt!GKZ-9cBFyTo~F1ZJ=p!Lv$is8vtI>;;O>0mTg^3%ffz1`wU*;&*uzD8%} z(_cW$W(0(r(YB<(t}NcK*wA^dZPaeHjv~1!ZUK+4`=;c$)v!zobu=qvt`7ec3uHJm zA+JjKo>vYDLb4tcKg!iBVC0^Hg;uT7ygRh$+#n%%JvYzF!VCW5P6MXsQXO4Rr}Otu zEq96s&_K}cD$Ob{%<{*6wY>B1B>+z@wa3IHujoxH4EC%j7o^X&&YZB$ZPpCCER3AB zOhDsK)_z21ezk^@$P#OOS&i1m5TS9x>>TN^zR&ODpGk{r)C~z(rs}H!^`f2JEVWzT zzngabhEY%pQy1mr9HvFi z!+bfv3=#4n)3;0Iz9fj%q=-gDr&e4IXmt>4>jkR#EephDoG9IiOQW?6mD}SlrGEIF zo%_CwB!)%XZg#^v!`rFKB>chHxma&gCYX5QhJ$uY*xx_uLXFWrYfjxJ15|F8ULtXO z@GU?KIvv86)O>5;JfTU1ELiGUUBk#47%ukPKL&;y6vv^kuvDCUExlN9c`CimHQo|T zh68uNXn{m6y`pL_7&*v%o={yWt25M>gGeosW#@NSWN&vyT`vVUDHE5T>emO1>SwhL z#4-kSuccPW8M<>2t)n8VK+TgcSHQx3z@O1cD%L%GFEAcsJ1-5&wEsIRUa#3s%LOND0lO~ZJ+INbq zqu#F$K|)@`@woei?Ui)y5o}ft& z6tN?$kUj|w1&#?xYXe@fg$5d5T~k}_r@OEuVvkLU5^7RhY1su9@b@vD*SHAFnCq8FtlxE0Lbc%>0-+V18Q4Uz3a!SbB&dI@!I)`jGKTK)zVlVa5PV=%jmP0sQ z0B8x<@FkYpltEVp{U*r7hXc1x_{J*zsR@2a3&$cqfxj`{S~wsVqJFXy8G=|ewbNIx z0R5|#XWI{@{7?WB5RC05Kj!9@o8h~MIbq^S{c6E-E+wlmb?15HV&_X3z9$@0p-^dz z)7K-)$mS~HO>*A*6fCRcTY`-Yu41x5asXgfw>O6Ho-?N5xSc+v0gkJ&IxMq$(ciH3 zt5nWGIAVbPYE~hin1^oMsJ~!u%3zI0>x|^12+mBt!;s|x@1`+h*lQ{NOX2jVV*dDtFyBqv4u?N&R(4;8j zsciOZQfKQs_7G@HWA%)yCL@7u0;J~bY^{*%EA^J|zFsM+7e#|bI5^)=q6`2)Eo%Jk z7WK;|)ny39koGliJM~TTiMzRjrEd};)0E=pkCin8rn|ib%LOdVYnM2ymjnl&pP9nr z<&-3e=ya%A3X(G z+#2$-`q(4()q0*9t~x)y<(likGkx?E4Qn?CLN>R_&xJCFBzTziZ-;V;aHI$Ji*+tu zTARFg;`#DcFYpTv*tq87zJ@$oPFvkE>FAz(lvQUPDJGY++n@V!xugpqqvuU|?y|px z?KVx@8?p*kC{hT?qIW-PDpFz;9j6OCl?PmqY9E#FvQ7_-aFCJrJ3q??H|R#JQ;(d! zt#tOo@a!}Szvjg||Crz{cAn#xvGYMDYI6c!pJ)H-t;Z$26P|z5iSBl0&vjw%2tyXJ z59z7vxVWnVb1VC@*vKL3sV*{PhrFZVSP&xHIL(^ZSO@?13o^BE4kGyMSS^ZLw;Y)j zp&TTzRPgaN3)>oSq!o4_nMSIlZym8{nVQ3oHZaB1vDuu87Rbe;Dq0jIv zK|!IAhRZe{8M@}3VGZfk@D38f8VBse{d6lKn59$Ua}qw&X{@1v2T<)-S!Fz2%u8kP znrAp$Pg15Ki)74mY@w{j`Z1Rr5mKKn3~ZdTOP57A6%y~g!NI}wwl%QlebDqpzle5r z$;}biC+n!s$`+W5l!;f9!kor8o_FC>!_yiF1~UU$J$i+m5Ey07HWwi zhQC9$n42F}f|P-v>!tEGX^+D%mkf3)AobKm;^pP6QQXne(RV|}e{h#7`0Xz9a6JD4 ze&qdn`w?T2kJhzRmR)^T1WwXt|Ad_3_vrePK4uTXdjSWfVK}{Gn06Ug{b~NOt z6I0gre(Ub2774v%sRI%Yf5R5=xFAj9^jASZojR#)ngewH*SSfBfh6m1jE0I#Yood9 zDBxG+k-{p4{foaUD5ZmMMe&WL>yLxqI{)nw_=)F3%jVp7i za5ob8SA;$3VATOuj;Y-Ras0iv7DX6d^V4e#%zpvsRu!Y1SIho+HMG`~~J2)?V{lGH&mpbcMOUtV?Sih1-F~9obb9U>w_p}*(vGcIksh?3o z8{XwP5bOrtR%Nac9A>=fn=H5bNbBfjIgK*9T$VA)9M7|Y9qC;r)Z4+(w$X{J;zQio zl02ligO*t+aWBbEYC;czD!JGX^5T3f-}2@m_>tRO8?XXL!rKKKTb%5;nd?M)deaRx zNGWs_jKP-v*jl2)ys`-2ESGQ|7rB9jsdJ^dB^h|*gLy%+MQ)(A!yoYC^ifm_wJ-Yl zmV3q*eFeYjJo=lr){mCRL*+$clPKbr0R=d!lbLW#3r9+V^LaGC$J-WbQ_|}P*o!d2 zTpQ;A*@wOsTV#X!jvK7Tbz?;D!=LrhJQvG-Ud^`lp8Ma&rs^Zt^iUOE51THez`1Dl z%?{pknwW~dM}OSDo-gJ>sghK~$be^P^bDm8x1hqPTodPK{azcuC|3p2+AL9qNiXN& zWNCkec0-;J^9ZHca;NW7+t>TLIn%AzGdyJ!fwXaoawzhy6JL|}KJrp?3YtJKVrDa<>Vti%n>N8?So7I>Id)lc$1F5j1UmRr_~8LMravk{IWzfyuJrNHB4 z)2FzME!}p2#P>Um(*Q?T!U50VTAS^C%2uYrUqHlfFokDyx>-(Na_xZDcB)^0cBRxF zjcplne><>O9#?#NccYReMRP`yp?LuQ^bHFNV28ZqW2!j)callq8@ytUB&za9vd693 zoeT!2@Ls0gN5w+V>2vY{Xff6~;%!{nC@ri=fx6L+Paw!MX8wJvWVp`pd3H$J9aew5 z4J(B=#XWiZ0ntTg*a~*H9BwE$U~a@o1e;)#w*$ac41b^0Enq`JK{j7ewW}l-D>nGu z{SN4x0`q>@;OzBw9CTl*3d_)N8I=IE7w1d6S~pVd@gndW4r z`54q*SnfaiYRsXb!7C`+&VQ#^|9mca-tEbiB9GAkTj@hlkrj=sgj?@G z5j>9eklg4b^JqZaD=%J`MEoJ$-oFgO0S$w@Zjg#4vPIZD?!LiJPfm*FE0T4C-m{Zk zRxgvc_sgvwx$)c{x^w>qEhjCcz-duD+*T!J_~tocB45Zo30BH-xi^}!ow}=Vpsa&w z5Ht$}yry@iJKwf(I+D*sFx8uV*V_r<;}f*PRMc8@OhH{+)>~P!c5~|f(xu1GxBICB zZh2vfjBhZ5rb=JhBH#KpC4OKqQ#l%XHiI7I{dg|f`yzP3F{oa<`K&oo)41#P?R$E% zL^o3_+RjD{OvRKe{y?H4ugC3qJzi5 zaJOVKFE2q@!Da2lZ-;Rn2fY2+X!NgKo3#tF2}k!Th$ z`Buv!xq=UU0CK;AKw#QJX8Mc^*J`@=9k;=jsoih5rTDK`+7@umQO6Q-p1$%tjl610 zXPcs4x0T9U0&&Y$%6iGD--1k*oe*8dMI9=o!9<68A^n&1(*-+_(mmjcz1}(vCBr=0 zjfMeE%)K8D0fo#2t`Rne)+ji7ck{j8eA1T=nE?JfoRWps8d_ zTOn>Y2T;~0g?Pm=lUO_@H~e4hH9@s66eoN5^4(`z@6?P#e*)my5pJk!)fjuR6rC*1 z=)O3(2}0)br229UpN=h)68%zX@5oEB_pUbfPZZ{sJz<-P6b?kVD=L@;I(+DOh7|jF z2RVAC+ce-pnz)cd;)3|u!XF|>0=%}6coP8bd4>vMBm5F=lt@pcDRj2*^4B+dT9=PW z`Aen3dPu~>?~Ar+4+-W4XH00xY5?wBJv>nkZ%rR(N42NqCi+p{MhBEOJ-Rer8oHV! z-QrIs3)T-Z7TKl0nFo@l&8z;TZ{8d*{^sXd7n5mKa);Yb04HHsfN?W(dVQ@xz|)0c z5kovEU#)Qv{bG;*Wgk6gmDgoy$2h zT9YlvT*Y8QDN?zGwaiQ@LfTBc8!WH{n3|yFNke0^7w^_qH?1IlmDXFwc!QsT+o8a$ zxkqEtpkyKQklkRZ-BkdiwTn0STb zz>aB>cF^(W9#U@A88Y#{C2KsVeKr?<#xR(vqLK$XW&4#eN?m2TZ(`^en1P z>$eZBt-%O?0e_@m_p-ePh?{)gYvh~hk?4=-@6irI}EG(h5p& zV6qea^e$EoJC-MAUQ8!`X~E*esHk)V>_6eZS5&)RPi2JVUVS8pMY#@PLsbkmLBn7o z8ryN{^V7im^V75+c40>NNqIi0`4h-13G(UQD;qxaTb=N2Qf=|_tj5j6!5+m3%v5}& zfyd$NeLmbK}p%VF2*8qit-^&KySEA%>kyEcLM-%KtnN8qrME;g6 zg?T6j-dhA2C$o|NI=MdBv$G^%*^CAZ4-5Len(Wj4!ujQ2fYb%acK8V_Yim-~p5GE% zk%=8nG&@!A*@{#60<)Wx0$2)%bC=jjH)C-=J~DpqoxiLbc5`(x#sVn&UHA z%!vEMZ=;PB4}t9l0x~~^wki?l?p;HAeZQPr`~@_t4Wtw=TDnD87VC=J!UOQJ9grA@ z0hxD2;~lz(a~fobTjc#Qq~}*w1WHmWUp%D@IWW|A(6-mxT{|dW)rPEnn6mNk?fH@) zyk{D{C+pY$lRQj8lSc6F%wTAn67!xjqUUwRWC)s-IzwhXI`hFRZ(-J$6;^HBX0AT>tenAnU;x^Ne>Kse zik?VM8>Ej!4ysX8plhHl0g%!oV*mn)sD29`GBOXR73IV37w-*PzMN#q>xyQ5;M!(o zVyfwS1VA(y&J_n<=+5&dt|UX`gKvo!l7bz&^NeEE01|!K&WE2xn{~h4I|o%1ye2a5 z*A<%*11|dBph>v%(a+5Ks-6e-$}4KUfWCG=y4Wu& z2wKOoZX94<>`7KCK-BZ5@#!*dN52o5$vQrt>}m8@H}$TXpRXQy zGZ(K+D4Vc#`!nSLc&Kaucz-lDK4Q(*M|`DM{%Om`V(W+1)7_$kk=l*iAkvg9bnm|I zqr%1*pvFYhPY)VeRIhuno($@`Sga%>+7=Ky8-qKrpck@w)a)fg@6lst6Y~3_@x!Bq zD1Bt_{v3sPSTL-lX7fV`{7sebCacgD>$}9^;9;A6LFmq2I?q+i_U$2M2>2sDn|VB; z*`R3ir|jh z6r#TCZ-$J50t6yJF&k`qSMYn0vzv1j=M) zfe)MO2!`R5EyH+@n+oN=*-T-4nu<;8{v3jolvtKK zSyJTO*jG7I!|<%#Q-a2>V4Zq#$E}ins1jwdCTqPe-aTG znPIrU)RyIO+?QCmf{TUlh-${7GO}s9r{DB3iF=&n3SY;9mfB=2Cm1rHJseb7!4U`a zG~0DHS{y!%Zon?>yVG&Gbsjrrd}eJ_cowp1Sk+yTBsK=ZHHcWbMUJ61YjY-~4N^d& zR;_pI?a^0OwL|Ocn_|mKSRv@O(3s_^%0}frVO~|14XA}xK+VjiUxgxS9eM|%9|Th-069R$zcl$^sYH$Vm*=>w+$GPE4L`@say#F>+GeV~1m2H!#BPID2>pesv2Tvm+Tu7Q4qd8S#1Uhv z3K{6xpni%X(zPW{56FVn>J)mp6C2lM&oy@8ZIzgeqKZ};NIri7m=o!kVT;uYe>BD@ zYtt$w+xvYLjMLK@Fr<#Gk?sSGkC34v--1S%l4xBk#4%$FvcrbgxxM0>ehrKErt*j! ztNXm4HcxXn4}Pb>4neIQqz82q;qmwNMFvK9-0;X^^Wr_f{gh}`{gF6d5>5IrP9JCQ zOMi1jDrE@Kb%~3s%KygMR|UlZY*7vp2yQ_KX9!Mk7zPc&-JRg>gM>kX1ouHR4DK$$ z1Hs)A+}$-eB#><0)>iG$*4DoL>D$#GU3IIvPoI1H^zC;}ao&s}{Llm|{UNs3DgIxz zTK@v%?c!8%KZt<4Pdc@{)`)-YZ(_!$q4a=UOb6yU!cpc7J-#`;3ms(|P0D-39^-Ox z<6=b!Me-@ozOwSa!z^yK~>m}a6&9RBsLBJyw)Jj z;p{qh$vw?#k&QN(P&h&%~|4XO-|8DftrT%%6WDB>o5kYNCP479j~>t8!n(u*V|E=fHQ@U&sM>4`f45zO;du`xLvrhknCw_|y=(rwxta znr70;4CiVdve2nvVYxb8TPb1RC>zv(jC&wAbtT^?cfb{1$SVzj=a}b=T6u!>rNT1H zN0A9bjJfZ?!`J|9PxWePdwrbO$p9&j_`d+hG{0q56aX7|y~8azz4A+9YpP;FtS$XP zjgz*&E7mGNh~h#wE31tf$@}|=p$hNV zR-n>E_e|)Fhs5Y0NI}=uXE6&LWrGSiNlw}!T#1>r7-W~FOis(A{Be-mFif<;X6``&d+RL8ZAaBf%C+h zUw+LS`b7$f&SxHQKm=yn@fPT(xlfS$Ul_CR2HvR$)vH;u1kCPvYl4<_R>LW#mK)(m zZ=ErxI^}?Qk)K;Il5%|T_H!S-$;GSFd7h4?Z3%CaZuSl(yv_tc^4O!=7kDj=X!VbFYDB)o$jk6msalXNp}lFnJ`6ijB0cFBU+c|)y>xcDUkv0xAptl$w$kTnx=7R6TfH51QouAAQ$Y&$A(~47o<-r zJeZc4Eltey95XD7(^wy=i%2C!z2Jo*bv0cLHlkIYgJY}oIM6ujgvT>w1y5vt@CsP4 zvb_x4P1|I;@S$W#Ek>ssE>((6kauFRH^uw>n$kwQIi)K9nt?iU8@z};l*t_A?fv0O zPWqv&rZZOFe-a$Pm^dX4O743vubrWUMO}b zLA>ruYth@kDJo4#-O31>>)Pd}QiFOy{{n*MqN{8y#-~x`VPRj`7Q1?PGQ{y1#A30{ z0nqiLu1JCUJ)|{}(Uc_*c~Z)WOp~%c7mF}^PBwdXztDn~ba)h!>bT)(bqRyVXE8*D z+Y97PUwBXkz}2x~MzFMqe`g^wcv%Dz5yt~}&Azi{2rzf+R5})WeIpm+vaqw7(m#8ibXeP-g<5vnQ96fn8mtH)xuGn zn&Equ6~BTVN(q)W)>hMkDGCS8^cD#;&0I|=m5-JpIcG|jR*P5a>?mnoY2!H6Vg2KnelA>&{J7nc+F8-CI(6_O3EoPFCx1y({zo$iPejyS!zT#0swLy)mK6SAV{>5H>+(dJjA>h~wyqb)u8970qsBO9 zwFI@ncAwZPQ0*Q)Q|hhw%NoOhW$(1K3|>0OV$~(q?YSG$ zR`6$wrHN5)PY{i}B*$KaRjvDDH52po-s2Dt@NB)7Z(20<$!szO9k_B&%Hjkwot0ZR zv07-W*fv2uK7kWB`k@KHBREkL`f42c*RloE9bk=}ap77kJJrg|YE@GE!?zVWi|ulW`p zBQv=wbzX6G3uBYwxZV19RB_d*bY-E1i#ZXgB=1ef@RHu5Qjb1IO{v#cdzUvFt1-m% zo0he@dr)9<3XK4*1e(J8RI7$;D0!vh65M9h<{b2olYbDd~6E(?n zt)4nVdt9>um)X!@^Y-5ffbXNh;T^4Gq4s>t`u$x2NkUuI69Ay2 z_uy=fi@hkHQ;7%(j&6QMWc;%5Pfu=|OM8L_@0szS&qKSeeeYnmHL2!Nj7pJ;g6+n) z-gHY~v3*Z{u+N2|iI4J9dVaCPy~nGZIcDwIWLtmXf0%~-;mPmU6t{eY zw$x5M_6u^D%7AetN^Y&u7x&w}T~rWa-eeP13Ax|t|2oTj?9!-z;#k({Nm!lUb`;;u zOjBp~!C>zw!0*(r8NA2aa%HV&@iVE^O2r#bVP!D%s2ELLVE`87grimXNE9qeSm+ow z-zu*<$H?Yqfj$Z&e#vwevp=@^=6(TbJ+}py`bkVGx0`*fE_>tow#vHaql|tI&re54 zXoLNNi|giI63VE2{vx7q^#MA#?A@Ok$`J+lW1zAdczI!9^Rw#7$aC!UeEABbgAzi@ z;viy!9|?A*voM+8|MM51)G350BCuENjl`5xo0IG_kw;4*v%(f)jvkR{ihp=iA8frE zdK-GoI$B1bbF=;DK}@rcvj5YTdTFO__^d~3a>5!P{>lXcO8VRZe{4gSD?-c3n)k(# z1ga@&v*@P_@5M*;o(2+pT!*(UjC=54CY6oA-@XmIDG{0;0BvDM;C-djRB7#4TKTy* zlmi!Ye%Tw}-E*XwAZ$N5k=fIKtaq(3OFOzV`WKLDxz80Wy!tz)#a--j!|d}CuH%5A z^vPKK2|~Q7(FE*9n`O^znEa{x;4_BW!AGT|Z&W>;&6*|@kGdKK=_^@4E4MU<2*`+o zbrb=)cWjw=Yk@aW|Br@(umz7$$3{eCMi8_93q9)eDbom9pFrO}ap5RCt2u2 zI=NJyDTXC9Y27k!)5=mQf<|Y~HdwehSk1`k;l!*DOx{>2*86yUKJ`c?r5K1;lQUk# z4}MUUIyTBk4@N2&xMCgYqitc5Vh}Y~1EMkeCbBPYDsl-Q{07stMdj^Y93wXeCUQ$W zG&6ceN+cO|9sm3i2y$cA$(`=Ld0qdpc5U$}{iocSdm<~h=q52czt)e3^l7<--;v(T z*gotP(!UXA%YD(Rxu41L9Pt3h1NxHtnNKS9m4VY$0W#DME;f~hL*Xf@V&>;Cv_GoK zC(R=j?y6TEvRV+m02pCW;;OD9mq^UNMYNwN*A>3)9L5)nL7*9Srbqkg4~uwJdJE%D zjy10`q2_3JD845I)u}cyX2;OehO<`kgF76bBZ7g#CfU#+ArHkYb}NwX>aI9lLpu{m z`}MB6)xT%Oh4D&NZSuQV6#*;#9u&EAp;!GTToCdB9ljz^W3uam2GM;KsnO!yo|MfE zNwee_BU?Pte-55q-m1?qXB=PAfbq}hli!u+$^r$W@5i6 zasH-P7@TT>_SLn`0{?&+`@gR5@*f!(mB=6%nhO&PrASMWDEQ<|wG<4Zbq=UhZJx2# z0&a`YHkOb>2&zFH&n?vr{&Ow@8|XLSyDD0n6QgUQZ%*dQI<5!{aOTYfyU%07Z1QN% zSE@zxC&RK;K_JIHwyNDY`VK|WP`{@TX|r@Un#b&neS`2rr%;~t{d&mj#R2=(rG3m! zt}q`8VD^cc>Gg8+{tI7V%=da0*VHt=qk48xa|Fs8DJi*U`=K{V@yjl0@&8WJu_VBr zmhgJQdl)$I*d=BVL)CnfSaHC5S5HnUF8)%J-Dj2%!z1~~pckMpsqQ-W&EMQj4o|Jm zGyW?nUUvO*lfA`gL}DubN!Mekj76A%HTvwa;bD3u)dt>?(I&4W8LtU@99yy!<}r`| zg4jX__10kOh>a`fBJwYQHOJ)H)2ZR5qC{HNcc#}bDQPCOF2}^V`0Eeit@_Sfr}-cK z5C+}D8j$$lBi*m@Z=$%On@zcQLyMpQ;f({cQe}f-hB<{tJ-kMC6e5}M!nN{+out;aCftsyZh8^1r{u>VZx0+0GT|-<|_jf zyPFq&fOy{0+lUR1@?bviw_QEJ&fnbY^SA%-S|j^&wZB)Kt{l}vG?bv3#3y`TNkHJJ z{HW}oY2n?BncC}-x>5d{V#)FEKK7Misvox>jXD#{(ndaylX7t8f z>>bxXa|v`nwiKrc&YAq93^qY{8^ql0s0kdu8n!o#os*|yn`I@Y=?6mMH|}wjB&c98 zxI0bvC2Ed1s!4o&gFV`IjPF}yVl!+PfY)teC26F@Z7v0xzv3YInPNO{`T1ks5Arbx z98=}4=tWoNUfm|(-Wg%$HuKUeoogIaM?jmcEX25&;-pTju}^Mv(m?EKKpk3REIs$qP{4b#`xucnlJ@$`ZFMaem+zK=neWcJN`N?eXO}& zHj?>sYEDr|n6eRjkgJ$^(ruwqm3+6aU}qlcBk%LA==Kprq6wck*ZA#3gJ`y zg9d4|{X61QurwmGzW^L%L+;@(0l~UcY?2a9KW=kfV8HSC_r65=bU0)bf?cBFV`Ans zA)LasVrfYR)t_460``7sXw*at(D{^3|C1l&0(&a-?4G>cXaKak()sAo8*}?X@O~6A zAHX8zzKXNVr1UdY%UwF9-~pbB*{U*nii$OF5WK3}CC+Zcac;LO=MEBG6P?3@S`{Y+ zfLZ2h1q~Ic6xxs0@<5X0)jLr#vQkMc4k@ey?rSjJJ{)LMrS+MpO0vGSbcsR(m?O0O zTgcBo+Tk|O@UTl+8iw3~5^moY8;@R~8a&oD$SEA>x6p|i3F~1}hf7V$p#5n&Fcsat znm_lAS}j!m3lLNsF^3|br<5}5zRjFgTPKdpGLhga6^F=NZHTK%qz2Zk+kWb|j)MX` zlKLor&uh+PTVO2H0(_aoeG+F(bpFOhgzBAXU*H$qL6JX$l6 z9Op&<+qV?a+*RNxl-|6Ad$&tS83lxy|JIr0qH}79kVQin%%;SVom^EZT}HnPayH)) zVS(*vuOGEUP@c=60Rzz6uz6=Fz`y4g?u}&UD`{5mZT6*qHt-G)`g@iCN~~XeSHDXj zD#Vh9`Ek9Om;t@AoXM(n<>1Uj%~VFQb*`|>e<;R_g6|6<)<)e*S25p$H^j-X3sSny z)|wmz_p)WqpaDR|}+hk0Gd0tqkoR9aB_H$iSNrHYFtfqx-ZvDvIR(lX1QaMXVozbxWNbel~7jOkE(-WaxI#DnB zZMt>Nh>~3#Z-vG_5P^RqKG9B!N3SuKblR7SxHhYUOJvhD3Rk{T{#mZ?obw32(CH!+ zJkwYC&d@^jIW=Xp)pqzqTV$%~G#$>TA#f>YVD}>8FMyqLlPyt@55r@eFGzRCm2$qx z;pW74_6?O*p#wOve7S~9G_SqA*@1=oGD!;GcW6E%p3Z58_tDHeaVj$NoIGp0+5B&+(f;hSVvZ@kAW!78PrfgkZ0t-nmka z=C){MR-Gi8ELZ1qp48L!+)8~1aL{Glsr=WoR<#JD6DhdGqAf~Zk=n2>w+Vu1#&pv zLfLGGF`Mxf4^>5CxWKJJ%2-X6J}n~{JmVDJ7m2q;+wb;Ps!Mw%jWEuYaS49So9?_D z6MX?p2U)iIPnB>LEJeE-t*Lbj_0jZ9+RxQ`-8UvM^AE=+$oX`md!b=e-*Vi2b@b!x z==~5a4iK4IZ@|U`!+8(+XMgMC0sZR>M`E6wFG!$4>28>1*yKj zTHq6tE^G_9uCr9c-1jhS=x{RF=lN{KzdP)gIDk9=OB@u-&sQJ#Hsz3^KaMO|BzIb0 zne*E04c+4|pf|Tab!PRvSS@QHHk7dg-JW#8^svC3s_+TMA%+)enVU(_g=#YolI%6K zd@YswK?oTW=AI&@?nlGDdrRRP#@4ji2gN2D z&QD}T+o{hBGC<(E5F%1*RLUBc?^&xZG2 zMDB0nsCl_^`qv1~Y=6BrjT^2Q_{IRo*re9w>(iw5hkcz&9N*>4j~X#Kmq-Qu zXpDKo#?1KP<|haD5FGrQO71BtZ2zl)!{{|rVQ^` z#TUXLKhSd-HMkH@us${nSv~r0sO((wl1GC~gpTX_J};SHt`+~V=vQ*r zXas_%$8VPR5DSp3+0u@Yz0CI=9 z9;%>BF3BQ#O&iAaI*c>mspHtd+axX~F_1vT<(=c}A1}Uxi8+LYS45u65@lwzHASNu ze=!JhcK7;r-k|XUbYenTTV6g z;S__q?~NyN*;~OF$P_kQb>3gO7SfFH!OgT}3j9_u+JPZ)zciyPy*GGK$0a?a*szaN zFWre)2}-YLHE}&6*+gg$^P=)vc6Kjo$pEWarQ`-2+~>jh(Ie?f@<(?f88l6ZMTV` z0Fd-5A^VW-^A0hm3uC3QtW{f8;}0f6Ytx7wNMf=4RDnZnK*pJU$+_eVWBooI*X5#5 zm2Cb|n2so7ne`hJhSm_srJi!gVmVZrnE;*;hiL;A%-oOf+4Z&*4VGpsIJ?2)n}5DI zJ?rQ1dg`UfvAP8XR{GNrDAxPgG&LlEZKuGvLs{Z51Qs2y+VQc}N$cwd>%!s@y<~nawXcEuPBk_4FyTa5aAf7K)gx z@?(^KiS#SKA8}BHV4~(*a*sYy)j?!4fio^q3m;&d2b%ldNbg!^rsgcaxiwlvBaR+d zjyKNtE-%T?N&`o<21YD}xhFG5ZBFhDxIsS;(gphG9r+fZ_*pkZN$&CUqqYw6hMkJe zFBoxsy6%DWiY@XE_ahT>9!^>59xHRS4ykXndGz9LN{q!6 zDIxnso8bmqNHQ#I^@t3z<&*~Qk|gXm&3Ms3)N{6-6qJB**P)7&&|`Zs??Tfl9Ckv~ zZ;y|<%|Bk7w`Ms*QL;=?xhT>U7j_ zU@qi!V0?!iX~VM3vfxs$9CcB6^~k{fa{nRXH}07hzd|^gk<2RtG-gFkg~v0cFn{r~ z@JSWN$@UmHqVk)<@~%B6B1EVJ~%X3oUNwR{9W*pQJb1(1e3e?r54$ftTz_3)!U-n zdEt#c)taUw6H4X}+ZsGlleaEW9~w_e<1-mt+pL!BX*%EqD@EpfbWy*iyN0qbm@W3| zS8D^@)|38mS)px3EbF&CttXoP`FKi{F8i8~zPxahJ;=ea6QAMKWKCUNTwEMo?1^z~ z3g`juC%w0niYf^+S{SyoY)Prga~`5#In%{e)M<5-fYN;-_@(7O;lQ;3N~Fp=d<2*7 zLzYUkAy`i{b=||y<=_V&edJ^Mc&j9qONhi`dy+=47n?`S8+_(&Je=1N!Tl_94;Fl-2f}iHR|V3)h!1zm zs--V-)eL$pAM4tB>U~Wd+EvC%Au($lbv9dNVInN`IQ&}&d)COiuYYgTRqj|C%rfhy zxPBDmQqkwt-vdap%yn5G_UDpxgE=12waG^P4yvUH$%gvNb&?`kDDfRjjycE1?oR3-~SPiF*7KCS)vP@-<<&4ghe zLbq{rRURGbdE7#zI%OwN$MCFKK0M{xR&CkZip)qJvs-YxR9%@!;Go z==_9}NT^s*6uF)i@KWw%4nDh!0@$&Y%X?EMr6>6UPlS9Mz-*l1h1lIY@; z=$LE9ps1TJZNQzfxA^x@!p&x_M)RjybTJHG5=%;M{pP9OSTd|)bVdRY+ zv?eUPpE#vPfIAmP>H@BR0pk`7f0o1fiDkHYxE86+XQF6!W&L#6Mkp~o685Xv#e4$x z<@D0&8QFzs8&i&bRbpmVupXrWG+$ckNsA{Yn(AIQ@PEO7=2zH|IWVADq#E)Vv-tm{i_mEliZ`x)f z82RH3XwL>cl#94JF7{rC)%#1*P^=+0pf4@h*6kHWuy>r0*Svin2Xf8Zu1f_1ENTc@sTiTl-GfSH>7 zL#1p|oi{FZa3MPF!>hsd>4`QyEWU#;&TQHa?kpU8kQbCFQavKopwn{6aVdKHLXG+q zxV*_cQ?ixV60SH_SsODeAPp4V)2%o7^0p*pm4H7LOuJP1aHH51VX8*RJB;SlUDaGq z?XyK06SiE@d)V_@nwXa!u!hK^P4(OiHbb3CtE z0W`xGiU!75Bk|l?)KFSKtUV^66>!GS>vx-SXFKJN2HNn~@(oIw;}$U9dpmttOgp}6 zr-!OVdxETuh*frvIGer5M=fg!)vt;nLY6+57n)#?*Ft|N z`B$#P(zGzu<_gv#-J=FL2 zd~RA>@=Fn|Qg?NjL3Vb2Jg>w$&1^a4BJ*hZ33ba|c6geneTGAarxKia({LtIJc5#lcaF2We`50sAE-v8jH?hN(A9KpoSwe&?+h zjSa)%;{WB4dCG${u2HRtF<}*Wm1M2i`OqxS7TQB`N~PFtdvFRN0jxJ{%8A>y;!w{@2+f zOcwmhA#QZPhs3<1*4jglTf-8OY{GJRmA;$#iZRE`wfknM#-V63j{8iI zT=PjVmKc^(xTnT>=IsxD+7U)z2kYFR;H;jxc zLWouoj@1uQU&Hh|T0v8cA+6}R1g@z_QQHFKOifAh@K^Jcl$9oV zKn~@s3(jjrAoq}1u=&}AGIv^&QTx`(5T4-NxTCSRc|Y1+Sw@<`u9IDT!tGix@u0y@ z&?8f^(MF8dSIdG^xfuRnKK0)^vIoRR)2nGSRc4A5yI(`%NLFAafOEC)*@|X)g?swUq_RjlNmx zkB~6VA1$ogfAuY6sQu1N#tZVxU(-6Pks%3lgW!Uk43x*nUQMyni56MJnW$~ zPhpVPg$7k3>5@tRy30~iI-hOIjFNvH1C-08cE@xy`DBrHkPlzPFT4CW2pnFNY< ziR{W1Q#BXj39M&Xg;l%108&A`k~Dvm@Tz>nS5JnR^!GF#$tF1mQKi$=0(dXxz9bJH zt0~ilwaCP@$o4N#>85qwH6O;i^Hp>jB!8bd8x1PYcvUTu8{8MoGE+k%MLir@NqrID zZdsOcv>fwZ+4bkJf+V4MW%VIa+Nsj`6LG3LI8EIghHcZo%sbb3z%Wh{HeS_f4BR=1 zIof{IpTYemc{#e>&>B7ZBB|^oPCFdlmnrW#EvIvkw>leC=;2S6FQA_!*l`?03PI}> z=xoc;Xn?$wf2;4Q{>`Zi-!KJn`Jtrdy^6{s`=mHD@lsQmpN!c&xsWEpUoUB`K)&&W z;YJ{S;`)jEh{7x{?wrGIqTKmrZfQx)?{h{cqb288l7xflbn?3^25;(o`Fi3+?q>wPeSRg&s5HC+kQ#I>>_b!`96dz{$n#IsxNlF6O04XL zE98!s7GW`d(jH&uFC^J0;0&&WE=ZdSj1w(BUi;aw?2!{hm8xF=i{1gTg1QV+Yk~MU zR6pIGm?7?3$WO8RC?-2PqBFtXc-uCB+A=n3-!hg~p6^x=jb>sqYdpj%m0=@A(yZ}r zN@KR9O)rf`3W!T!^jWUj|I13pCA^ekgXOL&cBxuNj39yrO?!*Isj&ZUQ{il!BK~f8 zfWgC?WDKZkz8LF_@+Ig7$LJU?KiaEFmqx~iCGM#8yZ)JMgW zoaau*N=Y(f14EuP?n-s7*ry-%ymy6nAN4$+R{^ zbIxzGji#_=P>2k^91J}Aq&E)!JPGYgAe!fBBPt-tZhMNVyVN-n-3OG{6ZXNeM{^hPkrQl3b^X zCHyGFSOLL(D_IT)M0Qv?#ZWx8)&ik#8l%dZk zaG;j#sN5Z_v|4T>-XO=$Py>>~x7r&N&H3c>7?ZlSlFahTVOx6wlp^P;{%1>fB!IRP zj2{2ZY%Zo-BhMM-&`EG4qxml&G={Z|z95D*`M=d2VP*~{+I?V}IhccU1Vkp=0Iaiz z(Q$V`G5&+FZ=XvPU0iJZ(FUlc^#p*p00L*^PN&8>OPX{nQJVCm0kP6SfR5L-tU%QX zZdEs~B8yC}3e`!93OPp(Pkje(?l>=NM->G?XwmCtfIo=7$-NuHT(J-34|`pmF@Nnf zA;z@x^oN&~k0R9XJPqgNBhqQXh5%(dPa4n0+$ za(2^Eqn(jqovkIrG4`ae0=9xR?@2lQhTt+BoCA7T@@(#EI(QKf7sT%V?t-z`rG;CI z;hbcsj5FTsTc*WOPjSF{Mf=1xy#Z5gBZ+c}^FWs~OzgIGXw<&++P-thf|S9#?_iFL<)#^^(r5z<^(__Fq-`(2XP7 zg}x|#ZBJ4Pr(C7pfy7r7`vnR>2S*s}7Y{JecdGThoA|yfwX**5%od0ck2`-z!cddWR+GOc z(QdzrjE!3N!Fn}}w8{zQX6KV8g+pjUYs-RJTOwWKUuh)MYW3m5WeaQHnDc!mHr5`H zIPZQ^YC<7)EhKH;)U9 zVOqwqKPCv+BMwXmW}*6H9F>i5w%5G_JO?i^Rbuv=cVJ$MoNAZqSVosr%(Jd!TYM;V zb1kqP6ZxWJqZdzi3mT(LxauO5m=~s%Dsx0kx+aY_b=3&?Jm;z4m||7UCJ7lPCt&oT z)B9&Y?3)(hH~2f_t96so4%8STn2w0L9~U5A6dxl26Lx_3uU*~ci(cbc#bp7F z(yu|Y4oJ#M8NvRP{Aj6XbC(qlvl_|QHcX*EPnH3khydP^N0Y}$Os2fQfWQh86J;zW zgd07Q8Y;1Jc4QhRw>8-QeTh3}c>QiNa&JMt`u~#XPhC<%TE8oV(#^ltUU#naI-p`fE>D=}FR@g1q z(a*&avqy-GPiX&jdg$Z}&mt@Wb5Xr-?6bPfspksGVlzug0rGp_ZL^!hF;A2W8d#t4 z;-1~vrSt8rT$Tyt`n~DcZ8##ih~Hhe{mGwoSuc^b$t{5|4>$mYKkui3nOgq>iXn}& zXF21atLFuQm`qkz)$=2p%;&BxOdM1p$=Aabcln&~jwrXK43M7`E?xSXz;0TGQWKdD5hL>?76 zyD?Tun|#y&KRY$4{sPRB{Y(OgXbY!{FT_Ne6~Y{jC^GBEzQIW9R+if?1U8zxOlQW0 zsW~JR>f4gF-Gg$Krn$(jBxIFKQlP=22dxm;g6mu>W{hQ9>QH)!!br3nKzveG_6Py7 zM;uFkbcZX2jq~jIaJ467pD2zAL=X8K1W|Mk@HNH_H#pnK&=5Fv$ zlT}vVI&q_rQ7YWgS13iTz(D;C??Z&LS2rbS^WarZ?VylV;)-hM_Wruo(FD$mJ(<;n zlJf5!gY_SWgnzx7ZcHNcj!Kdsg7deO4cMd`cM2e(Gd1#CiI_NT=Z$A_Tby+!6OOut zk>z`?&e`qG1Tq?k40O|#CH1l+-{?w1CUZri5e;=M9*(Ss_hlXuX|*S04xr1$sr@{Y zU1GloIe+9_-_OktwaVLPN(&mHvh@f(sTR;y7%`THcp}Uyr?$y+dscqK;5c2QYCKLR zDAXhV*?)Rjd2MKaf-^ZhT6Y9I0@Q#c zXWNs4<5k(d7VVmNy9z^aFlt?-*tpc{S+ghqd|qqliYa`gfQn|kPC)`oQ<`qSW2Hmu7o9 z`A`8P^RG6q9q={m{(r%IGlZ;5G@rgEF{j_vTS() z6UYBEF>&{>GV$8)1tHHfXEy~^(RcQlqk~%ReKSyP_jMe9b2INo6XIeu#o_x2y<1-2 z(w8oU(eKO&FTTNCG(~N87hs>F8T53_$w8aAr>c5Z(~~_#jzraQOHV2fI0k7YHr+;u`hd0)|>TSdM%bN|saRF7c*D>p8AI4s4 z3`kRQ`b<4knvfrdjmBMX?($LRuGdYr+Z_EktLt(mTt~Z0RYlJ>CKEbNE!6hHBkt}||6PkQbWw?hcAk@c5H&faMXCSSyBA9D08>q3B%eV<3_WNp zh*VG>t{TL8`XKepnDKMfPos6h(amqZD}T~ZK5jjP?pd2Ocsb+e+|C+y_o0Fl-z$w3 zG*~DA&I27EuJ;(ld^UewV2>4>fix$W1wXbyr5wt{gkE~cxV81kj_ zyIZ~)7#H|ZTK0u_)v`sM3FSoXZf{bb2(lA?=OzgXS2(on z`#VkWzUa5(U=i|MV)97ngN^$6JCzq5&T_>!HuSKjCF77%3`!k34V$rexYibY$UIiycjT2>X%*Hl9PWR-Mnodh|Hqq>fa^_i#2DG;R5ySy${ zaxcT=-B|TnZs(Z_I0mVv_d0c@kf!c3HLzmBN(8{Ah8Z zr;h4nD*G@w&E&-BsW`qDaTJO#E<7M+bE~F0{Ow`MuhQ&tZf@2+Gck<`&LM8}{h z(1AM+`nfH*TW^qe@$AGDn$Q%>Tf4^GFE(IM0-^w;hs-QAL22Byn!>a=@5*qjQYGBQ z@dv+tsQ(KfCo@?4V)t>(BWt)MIr9m7Xp3tl|8~IOgN$}(5O~JYiFseStS#iQ%^hF3 zBB-#ubDSf0EFAA={N2!7w!Z)r6Xm}E+9MRZbUpvFwpc5)ZHhtR-?J+Vw*2aa?(pvH zOQiMfgqvb7>Ym@@3MJ^am;>jm_&*85YL!=-2E+SIl=QBw{IXgd%9oJv4sm6cZAiUq zi|4Di>ME$7fz1`Ae*w`1Sl#HCTxAmB2_FXp92Qw?#0j>A`;@bl)_ceF$hFj-@&=`` z)m08i7j}}(7#Pn>1Abc~1eRlZyeYZ;xKNc zbU*HDV7hD~kQyz1__ij$_t#ttp}ihr0Nxt9hnDATjdl`WUZ^O_N(nlJZY^)s^R+G@dcoBL7vi6^m%hEIe#?3nv zzWv{qP*=3|P|j%fK4OaA#MmyS8rCMu^??35G_)B(7R!>g<0QDq5mIo&WyQ_q;{tZ?iI<;p%isp8AB7S?~MWZ z38L}Q6??Pqah5~+{!k4#1Y*-{}hbVf?70>RuZZ!#Oa-$y?a_9NdG*g~vCcZJMnFRkf z5#<%HK0rhc+kjtuo&AKEy%05#8PAB4V~&oRBxtMr3n2WdmR)gJUgo}DPBYi@e&0~J zaU|rM6u_!R_{^l&80)MiXI;l8SyD@=-|DVngUZJB7tF4Qn}5FeN8#5wX5m)P92sgP zDu(|@#F-wB(|}I~Iib-XGcPl3GGTx1SgPUE%8@pU94l^w5RmRVzP_6^xXuDa;6S#uW32(iXDpAs0#l zq+`}L(`~#JqeL{9TUBp}mHqgYx=;@?h2;j_Wj$A(ToN6dGipk*omEbIy=aZXo~693 z&Y>!|Sx2>0>pq!VT?SK4g4_92O2?T5Ot@BclaAt)h2HTaDUgfBdCC=IMY1tjY)0NT zbZ6hY$_!l7F-Og{G2}VcsVwyD_2zY5^2o@c<5;OW$PdZ}rWl-?N~^X`=qYm2=p!y; zk`&DhqTk2PuY=?|?%Yx3>oa7cilREIdPk$%%3G2HXDCM7sh<>@2kKu#zwZ|b&~k)M zD{FH`*kxboWM|OFpTGyo@T9ZG!0|9ROgl>OMHBr>zcqW1kZZ6H(@D(HPzeAzK*qmV zz7%Ig@nw9GK?08-{^spEa z?AziJ-XE{C3-*}N!dCTc(w0I@xaE~%m7~uD4)Lxlh6x4<%x)%~Mhy3o#U5W#S5jM1 z9P%3O!KP;@vaML%J@JV7>$^d;o7x<#=ZM$IbI~6EQ2eaa=R8i4&sFrGysE{{$)V_S zh>>uZVkaP>iKO}vX1!NxpOlniNO9`Ra#wD=^cTR5_JwpxN|Y)#dAAjp+E0_VbiNwQ zSFvglem2W1^*<>4s-QTSVBN*t-3br~?(PsgxVyW%yTjtH!QI(qad!>wZo%CGA@@I3 z=RV!1J2h2P^VU7pJ>S>UJu{STKzhJ8N(Y`Q+IIcCeMGa_ynK&J;^QII5r_6Elct-8 zid-zn;&kg|0@r0=vsPa3B2CP;ZP^m|UEjvcOUgM`S;CEBV#KfScAHYt0hO~PDbmkl z_d-_p9N*R0KoLDZRml9SzMzFgL4rp6i)s$Hzsip=`Zab9A^^~_&+*JEqBpM|--dRn zm3>yCaIERia-;I9Y23A<$fvpa7b6=352z^D-asjLo5qsP9xzyQIOZ^)JTeenxaPE` zCm){Q%EW?x5*irVG|Ks#;Z@&^zWdUog%!8tyR8M&ut^81i=Q^e9rVJkT6Y&Hff|F7 zpH49C=SxGzh~H6PPK@lQ12XhzG@$WELGw-r1XKJU;Kb^Pm(9GTFkF~XsHLTH%yp9K zgQbJO^5{s#nC7r*cDv?0liAcJbADJ%yhC-R{uYTQ)EE7*iVRM|uCplYLr?BM6r!!? zXECly(MDSP_bgou-}3PJ~0Rf{^r?IxitJnSTJEmol{8{DqAe$1xVf3^~4&(W}+wt$-TkFs8%!VMC(+2TqT;IBEV6)2PiG zRn08OJZ66z00yDbRkx1O!rV+o{WBwXaA#v)3ReSDtBl+|ErZ(hLSIE%cC+H3q@?{c zi5yY?wb9r7>q(=#DO$}DjkuE1-V;J|Bg*Sj|3+|hcd+MB8$lLYO`rMgKkpsK7^a!xrKey%$9()xq@Ay~!6Z8d`nNmHa8DzkfO{&OM1!Q5t82LL zu5*b&J@)Dzo>~(K02*U4N!< zW9?x?W__~X(Uz>{)^3hqXB8S8uch)3!FuBfaJ|Sx@37p;kScy{i>?{@x*wI35*ntiEo`O!JVGiud`Q*ODOwBFsJ zVeRR+`vW!Cd|5-Oud}K3*Q@Peuje~ft``1&6@lC%f$0Az6=R@|ZKYvC%SQr7N`y!y z^ftC0n5u7vLKO>g{fk2)Z z!(%=YVA{onF?YAzzWxr%p`&Ef&9U4yQ$C}>Q5YMdsDrlght^$Csb|@3C8`Wp%7*^g z`D~2~jbjYm7PJ&X)#L7PbmU9!+!2mZ6drRp7f^!it_?oeE^S`O`*u}lRa-th<|=dl z=>EXq>I)qpp?2DjU|;X^WvJ}HydLr9D10^%Dp9;mjE-nxv&Z!x0KQJ_t}p8GB8#S? zl~hUmB6{E-0GC4S6*Ag~!KZJSCAZ?a#(eVyCPycktsm3R4@=8Y9S9O;4IjfZ+!-U! z7IWnZ!Y;!WIp~IaT?FTDj?EYXmA<4B3RO^BL31NeP=md9ab6q6+CQlExcq=>T=P;+ z@zWfnU(}O@5xL0nLb`SLf_Qr~$D*I?{c{eD*{YmEP}3e8f)A7rtV7#nbONc?S-^5s z$w-Au)f)U;PU7SI@7e|b0C2fR#Z|6G9r%T_4Sx)7>q`F^ro&OU1u_<_t?Ri!fFd&F zXZOmscxSw>3eBFZh34&Rz`~y2BDuGj5X<2JVavL+OU0fgd#WQKFhrK`T7WpC=0B59 zak(_0Hvb1e9zZ4j5qlz$f~_eDkljivc-g7@pIp95=>KMNfEFGdnaaX)RS%=0NOKPL|SATIO4|GhO$U?MF} zcdErq`)5^U{^|*Zu}v2dS$a4hDp`CJ@5h}?>`he7AA^)|r=bk(fE%#352LO>3Z%6& zxy*Mmys5#U$Kz_qOCWN)_h%q$0gVJoXmcjCB&j!Le$X;Z!>tE#$4Q{P+9YQxE~y_W zaw?=ZgwldxP`r8eQ1q>?`wANu2+xH6cBb=Y0ns-P=Ryf2mZB)2L#Q5m*LK*S{SvJx z8)kq+>X`r3KwYWqLUX^uHn8p+JG;4RxHrQy$wAwWLQOfDsxyDfk4 z4`AHMHsH+3m|VB#(7{)gpfKb_32x8Tue7gq3sp}JO6wgnZ69x_9)egI=_n4X4#sB= zCbn-v>UABUD4zoa0Cr~S&EHD8&?*qoyHsPF1rqDeiGJdT8g7L}XWEdT1?k7iS7 zA#W+%#JD6-5S8MN4dOs7_;Ur3FLR~9Fq9DGH0y66%@9U2&q!(GI6Kv!CF%yJ|G>)- zAAt1`KDz>!rEpo6!?KN&gBPAA0@_iw^c$JRQdUXAw?3b=dO=`NTPh-Ms?L4$r7$BS zPj}lFmLiJWhkcsr-*BQ)9y%epwAAff3Ehq-k!hBy3S#qH)08prOl!MfX?FafE^3Ej zpuCy-va+9htW%gYE)!M=I{cA!@yma}`jg3Z|C(F&-Kh;Lgs7pkW#b`Ibpx$52`X@nJmMm|B%J@CjUkowhy$(LViEb%q;(nvXw{v z>fuH+_7Mp*!(V_^?nQw}@>12V=Dl=XoU@kdD^VC?LcCz<;nQh3T5nsOuCE`-A%Yrr z>iM#HCU48zygtipWXs}vJPOY*HfA*1(eN|qPe5yZ3#icgV}+tz!24oNy(>^-+aR_8 zD9tqpCyL^nf75I0xZ$wmS!*9l)x}wNMI*MKY`%{>_@V<^JWe)^XCJ3)=EiWWGw6RF z+(O`}5hU)SE6!pLwoK$FrBe^4!YbdaNG7+WthxS3CDtZPH{_!5T(^>HCz+RtHSm zL5WY)nt^=%6ltgiob;?WnY1vY6W4zm*8z$Gz_(Nd>GHaN)9&d=<@-+niHTfy}fpHHX&3DmOgoe3Bn~Vd^xQi z7w#8KdE$mC-OB8sUpjiOaVlnN zUKgAx)Yw9Q*Sv|VcWi`>DQekdOHNrz>$}om;pxiR*^&MAT{|P7@md|2_8$@8{?WYg zKwlcjYV3hRzKg#{UQ+y8?8I{7TLR0&Lse*WHY)DyZ;!-KlqDRxk*_z0+tifG?bahU zYfon2ETcKe+^(**bc<`0Iy5{BWl&2={NB{GCWNbWf&rbq$Ub< zs$I$(gRJ``;CGpWs9)fL<6NGjb48G?f_{Y|)TkSRG1yO(d<%4P)R|~>uPYYJT&e6% zY|il(sw8_3&*8VSk~ZQ%a}3s%Jj}@BM-Y4zK2sE!Jy%p>w>4pja+~H5d%IBrt5fap zyEM&~LC5imp^qUr{sSy@9WACFEAsE?-ztW&4}Ksji6NM|1wA?3o}aAqJt*79n=_K6 zYgm%V+@bgRC4zwI%(R88r2PSk&CY=_)j`O8I_+~m7FTAN5l)gb1CBl7gRdu?;yR(H2Spt0dGHz47+ah2Xy6#zm~(2t3rxv&{jf6s_R){6Zrz;;xjvOpS6HK|!&fvwcGHmG5M&*D2G~%Uy z=-EPVZU`WyBgyb?^Z3UT8HxN-di<5h#$MXYoivhgXRgEM$tjvmfd64q%>35&+e7yv z6!%aAHeCPK)uQRzQ)3w^Z`nwx4ZfdZ?}Q_Eg|cL5uSiF^1J?}C0J5&m=4RA+a}mNd z{cU-91HtR)?fFY$fFLii5SmD7WqSsdYG#TUzvc_gyLN(*;2lcR2M56)tB4n#5^HjY zA34V^?1j-TjAsvx9LaKkr!~@dXH^`R>zJ34Gf+J3Pl~!>MA4^;nO+^&kv_b+#<{@Z z%$G{l4k{U2^VdXaX%}Jk$3}@Ds&1ZTzNDcF^Pmcpg6IiIr+;!b zQ83(cq0|_H1USIM-3#|pPL=!)hXB8mp5oVrJyCqmPdbX*jM?JAQX{F@`Y-(&lu+SK zg1|!1*lqQzU*9k8(Su(87WO&k6kD4^O7`9S({@Vp04I0FJP&2)0HLni)`88uTETCu zK2ayT4{D__A48M!w0=j>@N10}uNk$j(ZF7sR;eKr=jx~rNQzBDh&HmD&eLY##hELg z7zd#r)-&$luEL_DC8z$`?Z6jdt*@}n{B**8i&o2hFg{i<5#5{%AVq1d)N|aEw7i&^cKjsJUHUi{$aBrK3zJ?&gseuxPxKfog2pLxE*hdXxK`6 z3n_w1=3ol~6C=xffmY2AmNDK*Fvb4-RTh$}0Vj+;Se<{~7!5W~fO0Ko(9CpU&=zW z6|f+;m_)PO7&bsr($zRk#|{cqRQE}W|EfDw%+5KkQN6a8%riDy z>vWPUx7cGFZ%LkA6*`bHj`|J_kG{53PP=f;0vbu(IIu+KD1r(T7O*Z?=lix9Jj4T0 zGvZoc6N(!)l?6bQ?Rs5COKJT4zhqd?S^~(zAmZ=De8F?-)E z`IRm-@Dl3!*RW$Hf!rEncX3#=1MY7GL}W?`dJtE*kzB;Ij3fb=f=G?VFTAW&41x>= zJvHqIkW{!W`4%C0+0pfiP9thc2*Wr@6`j-OyT#Da?D0?T|(NmPyE`> zifIw?H+0?1q+tha$mJN(M#W<_Gt;5XTSKuWVWuwsck?x@UIDKgO-BYY>65=#r4fYL zsSVrwu1kT^^m4;#Rjo^~&7&AB(rm81ipF|UdN*k@ec6imt#Cc7S+}M6?xc@|A@j}m9L!`fbpW4KTarE}j9g5${;^>a$wJ2-~ED6~1BR1+R=LatV7G9eXMq^?K? z2q$zfUz?Nev`njZP+(0>DMp71VuW@L?^N9uOi?=Gt3GOr+T%a{Ztem;4?uCR#fgp* zFxU_1T9<*nn!?x$fkJ9|Hii8C*i-FePP<=}EcZ3LAl}XXzR*1UZgQrp&?UzU7Mlvp zdU#sd>{wZ(4Sr}{8c&Tl`h=5~<|aJUq6FDHFMpzp ^6=DgfXhZmgnxHgwkMjPnBa+)`Tu#Y6F&t>X~~A1z)%)uP?RfL z<$5mnEWi~IM#6qR;lNij@F$EXQX(OUL{%qxu^u#S%#lR-{`{&>cxgibg&EfWji3B8 zL+7Wq&JH&dGNvqb@lI$ET{>TYw7|^@IA_J1Fms6+?fAAMqROod+us~_QDgCIX#bB% zGMBk63+Ke;Jo~VFBK{cF(B`MAZ3RP*$ay&mtbwO7alL)rd3U^S<;-@iuVM^x6aD?q z8>Nj7j?M|)UpDBOkFjemQ83*&FKzjs7_0Yj{155TxtZz zDawrK;ldayfUAJk&2ZcUpKpd^{}b6&<7 zd?-z7FVoMHj!Keqx#|mub}eq(r2)M7FJQ7j8ra~ez(BE*L_06_yP3-SNlogQx7X1G zVyc9+4tkMS!za2Z;qlSjWbz{G@j?kaN@PN|2Lk13^NgTe=~AU=wogeG(U|$oQ~lTd z91B%JiAfC2$Co;(LAB765#N)Pzx-SN9#@8_3(47a$?2IABv7gB86rmo@6VM1%wOwo zVmg^Hiz8$Nu;)=9On}z>k5$=@+$37+WToS{z{HN-N)4jYCc-sgCh+2ipn!2R^$ zDm~*2GmSSpmLYd_5qGbQ8pPb5XLL|YhRuS%xFVRd!sB7h8YadE*St_tAsoY*@(buB zQ*U7n!#4+!U*V92{Sa1QtyualNonzyZ|%=72S2h4F%Bft=}2$7YX15;`F={ep^l3< zRKS`8rguMi_ySnU$l;!>4b@*8{X_os4kPk@pxpLZ)yQrY>#f)z8B30mEJj!yPxFJt zXL+QT(zWMONYaF=HFFISRqP(&1ok?xRs8H%@h|ux)TgJq$X%MWhho|Otv~@S8z|rk z;%;@_3GkOmBy|z>LaP)}(EuRfPh#}16(R7%rWu=zS)je$H}*;SNSb`oqnGC*00)6- zVOXmaJo~jcT|F}UwSU^qf3xnqpb|NLx(@Pm0cbal%4YJy3@|QI9YceT^_pKmuSYSA zV2HFALfijSAvDCBC6JJAwpYZ46&dhBcf6B}W~2u2va;H;u3SjT+$V;QoUV1LL$2AY zLJ0Z?khgTFVywF!?pOn&#cXuq{*bKVdo7bp(pzAbw>tgeN=&*}2uAa+2H8L{Xtyy4 zKT*NoW3ErY6cbgnB`Ls-yrfD!bqCaWQI&)0<(%uM=IGJkp;v{LJfj={!*IV5=TtDzO;Y*&am`JH+PyAW*D)j&F~1NwOP?5dBB16{aFUw zX@ZwzGWQ})n+V6NqL;H}HO3L?zO=Z1F3(tIfF=d%XNxN~H?l9U%~K`hGijbTI#%)J zh`O+8K2`zA<=!x4FIDyh64qweUMjqVTpJf!?bbT$P_QP==SsXa>mCc11fS%~o?Hvm zQ5Qz+mR>|KBmp)17!-NJ*mF{g*U$ys?=8Wq{d2F6HwJ3kQ+kjB7K!i>rN4|=HRT{X zYqvTm0C~<+jn6WKJZkCV`=Ty&8>QUwCKagYvxPJ9v%SobuZP9q$o2?X=C{FP1A)Dd zc}y$~YRFaiFR#j}`g`noh-0Xr+9arXk-Nu9Au2vsZC2ZiJ%oKBSI=H`bF@Sb$!+aw z6>rYiD0yE^bygt6d!USRsVq}UE2ATJ0)c3RruyDX(0|GuJ)n`}Y!)wMyjtIoj1s62 z6fbnoqDsQ2o>)m(RS0E6wa9#m1*#%2TVu}mtWDNW+P|MZHZPRCmG_fI7Y*8G*g<6p ztXg%%tLfa?76*9T$3ct3k^+NQiW=xTKBhJd#=ju1#a8g43w!-ps>D3wlz0%7Vyc~q zy&flGO(_&#q&9Cq%XO|g+wf=^adxedpDZFgc7e%6aC2b?`l&N{;_n<-U5{L=%^rjK zikl$LJlZk&kpm>q>`RGD?-8DYFs^|}pN8Ej_e}N6dDO0(d{x_4Byv{?SFr?Y!-e`| zL_(qN`G@BDDN3(=XDb`7A4jM+C+_5?o%!Yt#C0KH$UN7!exdJ)h2>Wym~`jQ$FB45 z@NV-AGk4!$Q!&K)OmhL;cBr zsk<)#Wt9zg){k@zL(piP=Vmd)fo5e~PVAXWIT87nsn8rQ=sU|If5>`U%czw}h|G-S zkV7-*0)9G0SYVI{KC$>O3hHH;vaPuFOf3Fdw*r0sJDF&9 z&w3HzSl7VOF(Ganf@8p&B2}VImBr1`1Lan zZHe95!c(RGl07Z%6mr3(^m|Y?9(p+6+_9$5FkYYQ(Xpp3r`~0IG(9r~+z+AMenH3v zCPpu_-5gzdAEk<8Y&em{?I8x{|AUYcTT$ zql<+ty0}p|*^F?Df`YxYwNAvPi(zd4+d(P%xpSFiiL%>%As2G}o#!rmQ9DE>=VOZT zdm;y6fQ$&(zTIOnI&UsAaEJvN(!hD9DXcS4y^Tkp5G6ERsR`5I)hH8!T^^d!DY|NxGSs0 ziQa%~1owq5i^rq;ZQY8_)=(gh&g{JfpdW44|K^iEeaayV(2YSvw+{=Ic6MJUSv@_n zoJ-3ft#iuU=jU=m`cSJ#{{S6UQ-sOEQ(Nb2>0C)Te~njvN^>Z>B+MWa&DF*KS~0Qk z6^_n;?5~u#K0&k!hmxL`CY(lScwvagPQ3v=Y3jwZ8lza9rP@5UZ&2U=ZCDG0nB{ahK zaUa;isk95=wY9G_66)2_$xq`4;N++EqI>l>IPMqL)=KVSkm=|PRZOT-nBRt3h$;C! z7M&Y7zqYSBdFD9!&_(a%P7*Mm%Iko_D`eGynv4BEhyV5w*kS!PP_G=RIA@8zzrqq~ zNk$qtp*aOSH}V9L`{czxO4h93TElJJ!CegtwMe8dCTK$Bo<_YJ#ey)qjhefg z7Xn0yg|rsd*5T~?i#RdmX!8H~Yg2yM-&Ujy2GOpIm#mH1Woe-ZeUv3HJg_mj`RyT} z8{xTVMV5RLH}L%j;Y~=x2tQD*B^211ph)tqVGGmojq!rMUhccG6xWd$a^m0ZFl1nxm%YHI#&3s zCLF$sVq5*FNn7Ud7W5VYPyPd7*Seb?X0}$!*1OtQOsDA6l-=LMN+xt+D=VGUe^PGx z3>_4PSFTI#$AX(dh-UVDw}{81M-a*UU;fsE*Sz!XFHy|kZK^#Er9M*K*Ajlp>UZl4 zErrTB67NiF%o22KHx~fj>CJ#P%v~_7OK+p?@QT|yN?v3pqU9JHlo*XUHPM{kSzip& zb^QmyL2W$M!2y|wOhltvzXKcs+?~vwuD|0jfdB=J&Af3oX{j>31F3`jePPZMZe!R# zK-`39FD-cglWU=KF^(&G=qQ19x8UIO#z`%gQn4j$ z_Psb0yciAEuCQHj%m>|F8qlZnLbr2Izh2KM2+lKW%QkEi@h0dFVA=a8L^teH?+6ZA2)QewzweJ%v`SM>xae@2 z?BN|dI4kyH<~pSe^hN$lF@u@LCQ8j9`aHIG7e&Y9PR>rK+w1{8#=O^h0Zq}+8xaKFRly)fkOWC?fygc zg22joPZ;>cuq#gPRBG~0b-pD!@Rd3ib3q1(0 zDRJL|k+-;FCy_py$-6?7L_qk8f_K41BH_V-{$<$>C6)^M7Bt;Or*252@bE7Fi%!rV zMrRuhEx+zWQac?}2zB>qBIzo5*#$p(F7iK7gjn|}Fqzvkjc8jHhU};lNz-oKYb~P? zx6HYRkLT+x?v#wk<8zorf#77zb4+;^&zn&3^B|@V^G&_;3Njcm0-Fof9#>j6+^G=q9!3O&ZBFTJC#-K| zH=dkksl9>32zS90A8QtRuQ}6&5E>B{h{-d`xVlAFJJaBDoLI>Pzv8>EOB0i{Fbpjd z6p>=gSKkNp!8!D+-v+r&nXNI1Idc$fX2Wb%jh`9I_y}r)4$z4fpX|k`6R)>vvG$<5 zC(#kfMs4MD&Y=KMJ8=UJz3llO!nNFpJ}f@uw(ARGE!Z5cS3j&<%%QM5tWE^~+Kst- z62-K9r#@>ZFJPcOk!DQQO~)3a^nTM;%5kE!XhZfTQ(&!^-l07CVorzvUFS>_seIpW zEUEfXGC{{M3?(dEUEt5G&7$q#>z@iIbIyY}PDpxMg=gE{Wb;YcQDLr_=^|f@r||Ww z`p$~Z&s-!zfPi%mt8Mj;K7=XxExf|c_c1=o3Z64W-~dF-8Nnon(7T**n5U=bXkM!~ ze%V?rCb$?HX!btxjWqA*Wrp&P9%^z(_$t?ip$KzW7^2{}-*&8|3yXs2Uzf~=td7Nn zYiKXj?#n3tgo%?uo$1e&F}7PxLM)#uyai;~g^HqXR%)LluWVU*r)Ef<-&FRVxK{5I3PQPUNAUB<=I`JRCM^u20p8U+qzCxDy|PyS*$hUH=18Wj z)~g(eMq!4TqbY%#)5svbARa&L&*sWy9Dc@?UNR}n{XD!FF2sv`-l+?%-*q>ZC-`v| zTVXmMd?d>VVc_h({VF0Wbw{TXZpzOR!N2b6*!C@F&kk3dmHadf(m@g}?Z2<Ef`W<84=OmvI|%M-9a%f%*K zgUwy_8g1XGGB*zA+Eh_3gJ=xj68)T|^ZYEArm<;W6ZFYAsTg)y`^TX-YD7@fg6RD#S}Yl<6XueL-j*3&u&8AJ0B1@bHfCVhG30O( zi)7yac#IzC%uhU(8pjPPzeu?4LW%j+)PzdbAT#^ErE2_X_8)*@ilxT#gJ)losS}u@ zr{6k0jXKbV>cT_5Hll3LGvAY6>poRAeCp!|GAAi${<5V-`q!;3$pI)_#v_4$Mivgh61+jnAuJmp$dc~d3#vE1BBy*fk@Oh&X zJvHrj3==CpTMP6BRjMzEAU$2+L+<>BqT-4_eWsMrW;k7zdt4?_p-WS zg7KzXhHR@o#YUwZi**btnCGQRrM(c$c2By3sx8U7Mx6aU%qD+^93s#OE(%*m zT{$-VNc&C)@<1apigV!|AGlD6_%OQvrp?2-r9b43%8MPZ$+V+`z{gMR2kgkT*Vnl* z>%2c^GSac=cYuM2uW8IY5)d$8kN#bpGEXx?S*a)?XT0`qdX?qrea!ss_K-z2aVXov6u7wnE^UG>u6VAPyiF@y%Cl0!3P#ts z>RU&fL5wc2M6cnV+}Z*$xpG2b(^r|+=Iu|^V1WgcbpH;Leeo9FCkn&bAXKNNqs~Zh zUe)2!)z$IGknK^Sr5KVhC-ARrC6;87xI1hcm7J1J0JIC|AGbg|j(-6CZrvkz zN;ArYDA&U5icT(w)lE*Q2OyP=zscVj?DwiD`|!q^p`f$D^@_R0H5~c@f`Qp|%?p0*|12h3S>hm#;~H9>;8NYH|X#O>}{yYiC=_Zas7q&KA-g zP`BbH9E^7Tq^0Qq1jAbHDg8OwYwAQf5SqjWC2Gs2d{ymIvjpHRE$;1n|s!;aLMk%E4-qgi-t7`%ulvXmH}AVti+c^^jSVIcOo9o)vASo?l@866yPub2U9m z0Sk*$^4A%^xxpkDk57Wz%!>%Hwft^eR<}CC zoEIYajP3u#negqmdVa=$GHV5GjJ_$ft@O{;kkZ=@UKnoEzp_8oVe`NA0eYHRQBs|b< z3|I;jgb6j{0Egwc>yViLf{#KRpYqX9#_jEmvsGoN8$VTC2O%28iof1s6=>iLK@&|)jD1jAvDLgW1?B!|5ga0=#%V24$Hieydn0bK-~JHfxZ(ky<^waKf3&$X5h?Vp=a$g9bxa@nKdiD%gdvN=8e zexgO$@Rw7OTOU`qyOmJ$XbRIW-$Gj-PPYtVMb9Ng=v8knyGAa1&$&v*U?hm*)rQkO zDhFd~N@)st|9qVFIf^r9r8_qNT;2+Lt16YP!}Z*9~J^NrUc4&F>uR? z(#4RzxvwS^x?#5HKb~a^v@4Ft{GkC!)l}*~tmjDkp)Gm4mv*L2Sob}K*km}RNwx1r z@pSP3EzBBtS_;*-QPPMTwdgpM(H@g_4!^W^=GxkS(Mz;FH1VmzIqqNA^JW!w=JS3`suq)$)S-AG{ixG}( zl#W)Nt_r9%AJb1GZ90Kg6yZ5I5a@+D7>j4@8Z7)n`l}{c0Pr~#4-Dxl;mstuKXJ$S zI~TDW7|V^sD}G$IN-~)Xro%KgJYjg2oox-v2y{N{ zav;8|YX%P*@HcFaA?xvD=}|#VBAuhVKlS5T6MwAU6Ulnv@Up4N3>6OT)TY?*%s4VN zZTq?<4C1It%-%CbI&&w7uU*Q3%?o)I%(c&S#yN=68{YFNDtPLzxXpf>crgfaJ~{i9 zk{O*V6vLX9(*MbR z=^go5ZWY=@v}q=?DLu`<^6cy2nn`auYur_5b{a;3nX=-|q1fx9F}{oq-DrV(z^#O- z_v&p}OQ;mCklFT$l{Cf9Tv|=DL*!{C@$s{`9LRDb|499UJoBVGQydiFbf@X~95U!} zsTVjC{v&~OimSWB(mV`4ux7NLv{x^79R}l9JLS!y-w`p+ma;XMn^AV@{e;XWo$HmB z@PB+k#XD?ivOA^^hM$_bpwr)~H1VR9N!$0d4U*nR1@$uq+l{2@UDG}eEROYGbZHW& z1alde74Dv0q!Jx<5dqD5l3S=bEBU_LlJW;89s}S!(UxRVVuD>r- zw^Jyi&3)K4CW+mWUTFG;^s3-NVW95+7v8@XY|&#WhAvBq)q7;3b^rUum2<)|V7|&U z=6UV#Sl;$oeLZBgt!07;;Sj;RZ?-9Wxqkb*or>6ey#Ix7B^JHPx!bhd_^cAAA~J94@lwD=d>@;!p1ONW}g7k2a=#rb) z`^wq`oy}xN7nvSkrh>}6L5M;ZRA_jc9o4B$hmVEf_YsfHBT^vSZ?6%f_C)*mOWtuegRFrgakHx>3^1f<)=R_YY8S&`tE~%l*R1If5-w zl>262O#n@vy|5!o)XB+jNh*+C1ZveTIhdKHl_mSa#&N-B+gGE_jc3%W7JLJn)$~(k z(Ne_3$mxe?|Jg#vkGmaxqvlSPOPy-i&w`8A!^18Jfr#9uqq2O~xG`P{_`q+25WMYe7GF5lWcF@wmo61?1|KMDLogHOktR?vK=XbSc@afM2o zMlfTOxiLy#N@TpEt7=GqFKRlg_ao`S*4(&i7+IpFVZXEb0V@nIlP{q{T zfHLZf4v%l@@JX!1(<2o&w}5NF+FyGdoihjy4sMiKd>TF9sE9Fx+QB_i1hdPPkZU7D z696rWy$ubDu$_&8X!u82=*-gLner=pt_b_W_kwYZ-mrn z-0>r`TcKE0!x!z8=<64>6yBp!bg%(}KY#cm|6sbW^{rcIcA)0lg7!X>-#l~8rZ%y# zjZf}B%pzJNe~&v~geIUTISx%8n@X30aqyU%JJs9Ni;uU;ZB}Xh?v;icO)R9NLzKzveM*HO@G^KL zqDhFIdyEZr{BZwLS@C?p-{xmF_860EhLI+!k_c9~j#uU$6CVmADiV7WtfMmK#*0p8 z>bGAq$Pp@|J50x3=k9AC@9N*cj^YJ{J9}ac<+>`+=xdZmIKBRw@oMwy_0$mi7|R?L zE7{hW#o}=oJUsk3>NY?UbBlP;?ptM?zod+F0FAx?uF>W#iWv61yr9-W!kHqRB_z}a`Bu(LKvLyJ8gT}8;yJ0`e@gZiee zCW{cD?y=ZGY`Z`$S)+Cv%liWD8UtLcC;BLE3ByRWSS`BORE;y`RztU`#Hufrk%S`N8uGT zJGH^Hsb9#LVMW<2i0!8>dwXhV-*v+{Jux~Ym}W3|(qDaI8n^hPV~Hz4La?fgz1y17 zy52lfz@F0&`Oll)5gp}Z8@i07!uB}?sVqo>hnHh^lh?!8!Mwd8*mT&{c$53!2eDa) z$DHIs7G$ZRMAaF$*Py$L{VL5MAlvJFh`zFjPwgIMoa=Y}ctNQOguySd(jib_RRP@c z@tnLkhF!LH5`x}A;{3jfcPnpk+#Yt~q5olpjdA`zM=Y25{9Y@ao_*Pgi^1xJITM># zJ?w+QSeY8-Aw1qFxK2Y;MEQkd!Xi5SwM4Qfp%K%BVI1qbuDQUjkGhqYGE+U(6L#`` zWW7mkeRGa4sa#2o)%A7Rqdc3kZSmMqR_-68{l|~gJ@mqn7n@&On(mu{A)G5kUk+pz z4K|gDe67s9KT`^K!HQe1u$bR1WF}DB={A}=#upR{BI1USB+a!WP~Qmqt5+Ll2<)mSuXYb5m)5U-|zw)CW!P&8x1-F z21lPUYoe#I^9e@br4gJmWDehJ1sTmyFX%m(Z$qO9&x3UPtFPG<7&GG*?zc-iBgbRH zqz&>zynp+zqTR6?e&X3P$5($0pyKIVhZMZOY&5!TDe-?fXmxatIqvjXO$XqKZJGf7 z0r1v*c%_fy=+!8*`(5CN@TR(HuXbJgWpm{HB!abf>%0j z$hAP-1*}>GdI3DNJ2IFL75s^8UX22;q2)YT!)iT4S#jRXBn@XB<3tw)=Mz<~{Jn7p z3sO3TEEd2pH>JJcxN1A%sA6A7GSGBO2|GMz2g%bvfD!ojt?QQ6MZDm8#ZS-U4(yrZ znp5Pw)5XSmW>>iFLFUurYwD0Tls$8z3q5gY8BCWQv*y>yCz4Cv$I7NHC7k;^JC6E3 zbsV^{*M@|asXxy6sW(%?DIXWI!{m~ghIhR#f{P0C+z|`n-z+UD-N-9NeqN|G8_3~J z&p{{iYp@AIH4_oOFZeCF8^B_ z4RimG5vfuC$jc^FY4vuN54Lx>9jECwJ1i1Clj|%<4soO_#8(Qi5a~Bx#A6<4F1>5W zY^s6N{RPDNJG9cwMDIyO8AEjhCsHYB{aj0c_PuROy;71d*770k@$@}XZ?;QGWT$t^ zgPQDbRlSP2dD^wT-El*#0IsT(*#9@i{5N3G4UvK3B4XZ}7dVvPDv_L_-C98B!zocP zuoOSk8EGT3hph{m!(_xCsu4RrhN4lms|+807IWsgrRpil)lImoPZoXsZp+gEHUj#O6k7mvz;K^g4NC87<}j?Kq5PViA^fnC zCOuqMJnB8`w;ZJ}OL*U;vI5#JTPfzDso<3$rV~nF~~kqpDufVmZLjch^hD7 z7TJ-@9yMuuZlg*12e{xiN|QxkGL%&>G|w;Ge69^pw%f}DG5l>#`&49GI$6S|48vzj z`xU~*zKKVys`xXiq2ZX(CiOq93ROe`&&pR3&NF}Uc)aIDlURgCQb?+4M;()WbdYmM zOQW;dQR(;S;ihPPYB)ul^1sK)g7T{Lw@D`W+UHnRdIlcs_d zPUQr-0zBF#+x=A&X7{&4*(vS97o@it0dC8Soj#V-{{T2%ff8~7ZO1t(K)8Ovd<*gd z#@^dgtnRip09in$zjh+?1VYP0xweqwFN#rHkL*?Vexy4^USW+K-AiRIEeO4X zIV4>+#4;NzVwC6agTdg6dFk1g`6W2DuS`m~X@#bjoN;_g3;I10`;XhIl?t)n(^`@x zssNbiBAP^hXxbWbuY0WVE0U!u>!;ZjGP3BtZM&})P1M{S;4M4pB--q0^}{p-cb)Yx zH+;&j3vA?I#SS+B#=fonEVSzr`N^e%Qouof>i#DzbjpA<4I<5P<``|XcJ9wLC&!|>rBGex9E5gSC_(>5>NRY{~%3;1qgf|F17sd+H_ zCHGWkmHgmZ4oNv0W7RQ9Im6$4L;B;#l6MDVN3ItM$NgGJgzo?v3+2+EWPB6x z));VW(+7H1FpD0_(by6>c)Jr+eXDvF0j~hS7Bkw)8kU;}d*8Dpi}_1^U;$%Q4pnng<4L# zjCPE061<0jP?&aMOQ#X$RiM&KwO`qqbcLCjxcUiJiADHiO4@YkPR-qPfdzLo;rKRb zR$9!a+m^oZdwp5Xs{tXx)H3uW4kQ&+g(|6Qj6nTki7IXt)())wi{3zGtDI0B1;P#v z#(wTBz$|dvsb$kRYFuS%Bx~g~r*P3ao_TL`kbIa55YW4;;_9-gQ#%9yavmrJ?PJ;M zEH7u2og2X0eNbGEr=dAGRfGYxpS^5Y!9F^2FxwzoQNSqA` zIkK$AT^jahaxU#M=YC1vd=S2ZC{VXEMCx>#qJMaW2m{8qf>4_pqOm7@tf+nUXK-oq zOFX2fqVVPl&3dHwDL=BqO*XdB%o87cNX6E$ZcOwmg7%k5#4U0Z;H@bEOs)wR3Iw6EY1hgj!bPk$V5%#Bjt z@op%aptjdIv3|54x&X7af0b$%Gq(;f>>q$7&DdQ*EA+-itiz=@Y*n?PA&epnAk*#a zx^8=BY;wld8jiY^N_}%rPEXt1J+QU03Ff4UV#f$+Z06J~>1vG@m}E#>EuekNCZsJa zH;94hO|W~|E#6h%d}wl7gNsK5>ZJ~EUotQpOq(G2E_^BFR#1-Tf1)GS79R_^m!omd zoj%nZtZI6ntViHXu@0L@Yt_6!A-m$PL(43@Cep5XWfZsCU(U%8^O)cbIMI$r%oS-% z|5n!c2N)ar2iU21i16bezBVEJj3i#9sBp#SRRpA)2NwC%3C8iR=H(!J%@kN z+z$Pje{wi@1LZCsOjjUmD-Y_XGqsGw+dGpxu?)Ep&t~zMz+$xv$KB05{b^0*9P|U) zQIS>+#2JMp3FVuCU#AaKbdos3apJA1gTB4w_j=(r5^6(`( z)@CMjGj!8V^U_q;`~b$&BKTuqztCs*W60#Ws_OU&sHSYZBRjkP_m}`e=jsbXGP(r| zz?${88?2Ba5{?aDL|~%V>HTBbmc}&$aULgeix@jR_VGM!I#Tg`#?m@ah^0`!dX^uLkmgT>j?!33 z3GX(yk8KPIz-n#0u|W(;cFP-k1qrSA@>-S0U5^GtOW}hZ%}j-k1=gD zMa~%h0mAu8T9CMd_5yZ_u9kG@nTxG;CW$C zKM_S|zdv5(o+CT-I}#Kub{%+V=7*B@C~c#+N9JNA$JmaqW1K$%FBIC6x3-!hWsBiK zyGQ2RgQo9CzCWZ4UL0TWWTXO_-6A~s1EhnDCG&nhYAJ*%6JW}58bZv}{Zd~u*YHa$ z#Crg}QVr~;mlB#66C`YOTX+p>e+bRC=24jkMu#q|nArgiRia{!AtzVzy}3169iJcz)O+=8t%sk3G+2S&1k z&^8byf_F~vJZ*1GeQNSRX&qvyBzT-i~ zulR~%Z92Ka0BAM*LPjw?Ik=_(3{bjsgQEXk3GCe%x2koSKP`uEqGotHHTLp>wR%>a z7_Gu?tA#@{_g*HsnZiJFC5HCjSCo1eo92voDJ1>2glgVTUE1CZoRN?A$wIJ}p#sF| zsnzv352HRj0`Tr{zqGfcVcwDx6jG3|0VXfDMxR1)80|x=S$(;6s!K7zpn$~9j1z0L zfJYc0e-_}9+#Ulh7Q1ZI+W!Da#kV8L^Gl(+vE)Xks=YG9myTh%d$>T)jE`AY7eSEk3&CZ+fYhiKpRdgRze|9lE|ZwMo+}H&~Z1cn%Q(> zGR*my%rB0o^$&2WHK33!w ziKhA6S6vDF9_pj;4$c~^;kJrQWa-2y%UT8#s6_Cku;A6ZOQTTQ1d+LoSMyt_Z^cbb zt~rM5%GJA_j;Wx-v|XO%$45`o6SO@@r{8V|3BKMmDqEBgr}jcY(qqFLYEYt^q(gUUi}WMzj;!RU(Y%xU36n4wK|$)-BQd});K7I=0P z)6g8^m2T0bp?7>)b*m9_gSHSc6>|=^vnFKruf%9uRkYtK4;7~Aasz5RBD55FO;ayA zErTW%XhnyJw<1s9ES!vDO+GV%fv6IRE)c)c)kl0h93vGy@;ZL86XWvaS-sLb>m|!1 zl$36kkNDr29SfFX@6GS7sE%KY+}=Wg>Tk;2^$Bkz{??Wgqds%Zx`=wF!bf@y1}|J%>}PWjT$YRXo#@RtWyg`Gc=-Rq?7y@_yDQD0 z5w_ZSCu*2yWo`IbE9b9}#Xo?yt3~ZLyu;oMu~+flyCWoL^gvaSMn>WgbdonOU2<^) zC|Yi7JG74d+I}Y_}LC!HFq>Dd#CB$TqEOb(C7S;&`=} z^j>rD^>nmx7er;4}~r#cHxXAi94$`Lc>~+ zL>03Fw2~C$+wMnfG3p^OrHrBzT9CalJGFzg_q;z7j}gj%!gIA!=CW+oB*qm$_Lch$ zcag0(!O{&HLb9_ZcZbjmC;DJypG`9N^H|QR)ijODRRK_Ls~!yq&Gzxz9~inIH0o z{No-&p#*M_q#tlouv;)PewiAC(u;Pa4+q+OwQ5QbK~DwOyV>C3 za}-N6-_0~l9B+e7C;)NvX(;tc6zON35DQbQbrN^6b2-p2sNVYqITAYeEjy1~H+r4~ zYgOv-#Enum&4dAhPj^MqAytO&uSO4zWZI8Rlqn93(d~j^u}*Wrr+M>Xy~!w}Hwfai z7%yJqpSv07rucE&ebRl>G6plHYyG88Ba=?a_~F5ke^mg7D-miR`QSoyQFVl~*~0qu z(x0IiTf{$aiXSskL06I~afwOQ*U5mx-qSey=RslTHil%$geEi?Y_cai?dDSxAO$h5y=P4~+|5dz6n*%P`w*C0 z=wuNkA1H7AR`p*cb_lDNh%Bo9_WmBK5CjxWP_^kBvyM>*% zB~L%_M(wR0>!s&Lx~oXsE(Lk$WJlK`%$6}TjM|7>Q1{~g`qwBo9Md^@aU+=i^R)hD zbNG6vZ>Bn6(Q?ep0O((?TDPuTEjmWzt<38Vdn8-5x)F7R;MSq|{Yn+H=}u==xQ|bz z3y*sxD&P$eWz(f*1@}Z;Z^6`F^eix4_h%B6zGBeqM%702V%o2QVv{?ZfIlz+)1_2z zQw7dKNOqDAlOD**9YGLI zZusp8dCRvdS0gLm$alHcZWw6j5n8+GwrwvtHPog9QTDI2h0rDG{vNBF?`Th&y$Poh zP4O&Sh32@iwQBwp!JSVwFI%KjRUI*6mb^WUg+9xrFE*AV9n&{V{I3O)mO!|PnjMSO ziLu*l_0dyrVJsTgBc3QrPwRc=%1FM$Ju6$6URs<*^uz z2%adP?+zHt-x~!DG9=Sa$@y{ zUP9k;$BOQC#%n~iJX7^YY;JG!kf$J=IhO}I7vlgiG2Bw;z;yPMtPzmI((NCE&~WsQ z6m5?nKaqk@d#{*-=^iJ2!Ur*iICzC8)r369l;wfpiC$;U-*GOb=bMHO{s9d2-b4Ci zZz^B1=f1BVQH9*)C6|`kwYTo?GsM-QyUA0O%aROLrVLj71CVAx3{e}{bx=k(ZSsl@ z9mc;}x3voN`%q%})b|4Krp+Mag9jz}PwzUhSpEUh%e}=U{SZ9a#My}?$FM4}2f_k7 z=0}7I`=hnNP&mR1%}j-7s?{q%IESC2m*NhsTVd_uYT20+n_&5`=1mm=tioBegTJ*q z-Y_lnjbfX$*yxdNiEqZthe)HJaf=y(A~&SZJfykEAUo>6uJsSDu0OID39%9D&-WIi zlpkZTe=B=gIq%ePG}bpiBdSllfRWpMqnIy)dQI`#F%E%JxeBiK{tsbpcBlHqQ+ZSK z!g+KvH-Ek=`cK${a3(-8BTIJs7AFHajjuo5nX#vqy1@R3&3yqE*G$2+PchmP{=}A^es`s&zF%9@(m{8ANLqD&9&_Irel# zX!I?vAfttYb@JOuvtNBu+0_?OOAX=hkhdCUUh_TqL0Air{q8tCpKlM*f4!18E8M38 z%PM^ev*d6vJ^l=`RR1_}XKIUo|DM`TL~@QR$e$E(VQ|S?oE^2E!9T^ik9Id{r*X)ZZn9wTEvl7A0X{>a0zc#5c@-M^g^X9fo zr^ogewdT(D9T{dOb{Q;&nT~s+<;EBH22Ba_1dS{MVOLS;?$~SiRbTrDR-Oeug;;Gv zkXPl0K>BX5*~#Jw&X@ygbCfyLnnSo?f{H>af9Tso!F+?pkrn#8-9 zDD+myTv@+Np}~?_=lk=YL1FyyM2^ew7NPo+W<^&v?HRrtJ(YCjByz5}@KvS~aXX!# z-`wSiV|xGU<&NI<4Vu=l{$%~~cr{!%f{unXcsLc`FB961R{w zqrUVaAJ5Er&QuteX$;H)<`zncpY9f*5&E%U=DQHi_nOTnSq?~Z8h+}ibVQWX?=ZP^|rDPcquqI z)#`MXgmD*NZe4 z`D9S>3)63569n&>`|BO!H+W!WN#6g^tNjy#z@O%x9pS~;PUzp)yU}Ggm!WNLX-eOm z%GTyT%t|=98g$Ek)o`R)mQ`Z!aB#r3Lpl7%V33SxvOsQye-3|O!^+=UMa<(4QR7P; z*fj3?I$Nh@kK8&y{I&wh&g3UefP2^qD@zv(i=fm#9=2PFS5%0&H*7@$^K>r4P$R>j zw||h$WW>|OimZs&g|j18(h-N-=Zoh`W=h1z4ac}Z_*H1hbpK5S75o1(;1QAh`fER_ zY4;U)8R{-GAJ3;0@xyr^TmmaaCmAjc;WUcHf(Xe)42>NlEi^gTNKES2eU8k{opkz2 zES$%^gjB>qwgK?>(IkP zePJD?V+r!^z4VD|wTLwWL;d<;ghxoxU_g+o_uSa`4C_9!-%G!Pkhx^ULchGmt5iaH z9jvrpl9*GU|J)%8XKWIUC6e;=mU9Z$W&|%l5EoWl6$x_PHjm zAR%#5b*2fEq~TJqWXQ2uEY??A3|vXqFHCKj)q8s>a7RhPZpiyWdFL|ld%W3jMv%96 z!1Ny^f{dua^R)3-zlP8RnXGsc{a0YhOLNb+Z*PN~J>2evIC}xFywZ;rPxWs8OT>)q zO(Bvx#!mF1PyeaIZ+fhyXU4jtv$b6aOx-t^`g>{E0$D;voZ@t$ULCJnk}`oiMnwPF zCT^js(EK0Z%?s1)*n~0&`vK}#VaCn`@nz?^rR39d28YCw)=fb}P9*$e@Kf*g2BnJE z(2`Ucee&|5X@!yd@AgMh!31h{BA|Z${&MVZN%jR(lF34bnYHJLvgg)RM9ZGx*H7Oi z>56Cvh*8U!G<+$g{{f)ChB;b5nh3FMH;8qeS}suo+_!g9f2X@A{gNS81O{1U{`K+5 zDE_1AhqMI|?vG%!@~L@J@;ev}ej1l|zjsg1BSLV@g(Km0p3}mNiy1s+Q#SSaQy8Hq zA;M967IJp>Tq_Qk4l}n|s1iJGShXM}#vxRmdXQErz5ODho`ucW5ZHhP3@DybT6>eR zt*(7odQCCfygy9+77Nx$^Xibu2iBelpg7<-K7`M4qS-+4cvS z_JVlSq~gNOjOy?7Ng6n9lm1dMHcrg4ntSx*mB!MM6fn zbUydDB6irv%O}Pi82pqQ1va%Bd}MXb;7N!?us&C~r|hx&_hHs>es^>^s)pKoS#Z@0 z(hY+K&6=bctR(u02!WsYgxrYq@nuXjQ1oUUDf?_O{~+_Sf*jqR5t5wg>ja?iVYP+# zGnNX6(ZUce<%S4Ky@TKRoO)%9p4b~zDP`{vPLv1kDSsnpp|IT}qjnu7EE>yIUx14c{NQ-41yfT`4b6m@6iNJ0hN*Mc#2el?y zVvss!U~cZ88dKExo*HXqbL9kE(5B{op;t7l>J;C~%}PsTe*kshl)Bh{?u~{}J&ckePl;6n8$|5%Ii_*naJ@7CX#ng$LqO#!&l6Lesq>9TurNZ2W z>M#p!wJ=V7aiw6^A=WRLIAQCiB3@CWMO`@^bpl!eu*UhU^47Zs8({5F(UPpmU2nCl zStORlE{M`_=pDipLS^os7W{~1@mpxNvY0q3-V;0`e5e(e{*eb$DpyYUst#F`;o{KH z7PFhGk?dCWi_k2T57?6a7KZ0YIsM(GXxJx^ifiIUS2s-P`)rG#fnVuykG4 zrAKPT(aVZ8dc%sYZIP+7&3x4nodFUX2Bvrn2g}#EG&*Crs<&{f2;j$&NcvXFUVctc z?@>+~X2eXTZJ~~JeGdH16Q{!{kY-zu_X&P}DsWIwk1_E2K)@AIRL|VtsddqsuO${{m#@QL7bd3>3}@1i zl)0Urf=3^|hyMd?ud2T=+mv@-jdFyivO^N~5|(O)U5m{6xZRQjllm*z+U{#~xBaVy zX^rj{HufE_Jxyrv+_h~infl)aRo#A>z~<)fenPI!Aov*U7)0glT1y=20C+05q$#)6 z3vX7Mbl;K#{$GD4`(KdJJ2JXD7W=x-g<7;@D=~imK$AT5odfD9KsIo?#fyPBP zUd&Phwe@AM=4TWf?dosKXZ%v=pu9lbP9{n8t^9*`fCx+ZC|0;PiZwssSmAreW>Mz?)rF^|ej33WS+l0W5F24- zN>x*xEfSX!+8@Q^{3vw-v9l*>@Nwd6pk1=iY%NI4_p4|R?g?0~*_6CkK8o^&Y(s}_ z4riPM@F|^b-#{|3WUkieay>KIC*>zh+g&7Uqi?3fK}dk%$*^I}DLG|s_f1FF7sQvo zUlRJAoj`~Yu4#z6`=(GG$kIzE&1oF=HEwi5MOgu zH*}-JtQc9DG{eQ;XVow&)deNOi*O`l1J<~MTkf8`k4quf&%Wlk0C%;}bp2SYj#s~9 zGBg1_TxkBonmQD)B1&ue`bs1xUG}TZW{I8EynXgwNfEiy1Nin_hssrW(WT~F(iZ&r zc-^-{U9K&jB6rFDEW}yRT-Ue}nJl^QfKKEI{^c_Fq2*rGm3bH4fVJ7xQJBfx>Bkkp zdUpkc#OVwYJ1Kw@N%Oh;9d%D?&y#S7R90V9jKv*dQ}Z*hX-Jg0g7(-|K7q@7YCsSz z318;=^f9iOLt`t>x1#OT+@(y)*lUPM!6EJmN8v=h*@+x3?7i#t%k|p|Pxrw41@`B9 zWI|z6*JqXN6}^$_vCpZwF`&DOr<#_mk%Cpv%Pd9VwQ|Y4rn9l35Q&>n67Du?QsnHt zvAhmOs{m&OL)@ACYB$S=ALoVyC0XTM{HKn-#?rLChr zZj?-koRY}gYU5;!QS|!ybe82WeZp7VkIIx=D3^822s66`7{rj_`&$XsE_F)(trl&VC_A3rif&6kqr{M|T&Bk{9o9bV=S+da+?+yKE zF%8QU4vWH1XKtJ1kZ!6&+ut#{ZVW(83UXw$hsA!ABt0=sKG9@(k(S*4`1}TOv8Q|V zq@=t}O45mz?A3m9T}3`$`qi4iZCgj@H>Rg(t;Bw|Z!^XxDLn7#46#co=q2t18gdRs@&2 z*{%rcbam&?!W!+2P`K}Dc1qoadg31a)8`StUTX=+Rjr1nL;m|xXI9p0bTw>0>IiN9 zl%9j?*PF2JM%M8(Vv}GX98LxQ@Y~TlkyU%^^Ou$u-U}ft7+&;dkV1 zW;#bjYYjg85vX9ypR`qLUaMwZavEVolDR#8uLcAB)9cU^$J4qEP*n<^49jRYX7;W_ z!wIR6q3{6zV7@9d3k*f>bF78uL2gf^`xkik{^c1Vqv9V4y0NnaV{c&@Uy#_j)SGJl7eBqpm zOQFJ>#Vy%0pc0-TIJRsAJOmQV3;LGw1|xxXiz_WHO+=*QST<*zkTs<*sc5Ofdt1Z{ zxfbgOfcm73F#WHqX#Cu5MpIt80iv@7L*Pe4>FTuEX0eub5^GE>v8i0D~l z%`$>$IkVpFNJzElTn6(JBCsh`4op$cdxGY0op|lN>(cdCXH(O0zwAhzx1H$5&miE( z-prL8Xqce8RfNjdD045AKF$G-+f!5ikxR6Ly_-Q6^5SZ+!%?@hcRCfHGkZ&FJ1rC$ z-#jK99+Awll=q`3dF&lW33@}gCULI(ehNBmc%|ugaJ^@9rkOJ22d19SM3A1?KCecf zuJI!6F%G`6T%N_U%qRUEO9hHZ4AA2|uPAfx@%&ZaXU? zx4pJlBS%cMwhC%WgCgM%`$Je&V8b?KVV8%!r;;_E$6pnsK}a2ivrd?NukJ=_I;eVq zD=Q#Yv-jpi<6N!Fb+2A;5}q*MBAImgYrCcvvd7<|DwFy&52ayg!%h2ZYIg|q@-l{U@n3|{xSvfi!mGQ!XQdkZ2DgLm z9Wf^8#uNurWKVN{?uhzncSQ~2=so@nI{IW8Hv7%Yxu96q(5<0jpwJ^ZYF{6}KBu9I zs^q&`B*3iwzEXskDS>yYnlTB*{pN`NE)l9=^34E8VP8xm}xb&^H{6-7nqxxabR+%9CSI3 z=en-wi!_r`Zlv|YKRCzk{qL_U{B*olaxBejC=|T0|H^VA##qfcyB3bHuQvFxDUI~x z>25*2&$87nPiy`yq-#n-UFF^idBuk=j?&K=1dQf3pB_x>e+GzzUhIDTPlBBpy5kug zM!--SkADEQ_~x`FDT!@5n(fs-KQFw1`#!lqN6Iz)4#*#agQ805?zI$!IDjN49>dLj zjing%+N4OenXqiOF2l9jnU?DC$6R&SVEsq)&bOV-@ZxIFSa?()<&sB}(kCkD3OEck_B7rG>xlvhYi*uKz zqqXr^jp|rfg87tZb;-#!)n-Pw=T$*=z@!HduCqa>+1U3G>)kc+j z^;XIN{FtK~L{$!-bzv-L!T*VpDu7-L{O<*2#Ld4F1tN@@fp-<AoaQ)=k-bRh@$QS`qD=^v6LA%D942cALa*@Hlbri&;ddU!)77PBBSrP`NU6L~WE2 zPtMxpuIxXF{}N+;IIaH=;E+8?%3-rlb69?&{z$1ELlJx|1u^qh6m6R`Nah)_xRHG< z&@LPQ2bg6i)X>{m5G3TMnD#Pz(L|X-v05q~YRXa|6=_iVSzHXFFa3r`d`)Ld@Mrfa zWYWy7slDcViieXUtOW~&$M_8If;6VN_zhK@fU}W13O1^+hmz^> znogAa(u%Fan7WDtErRbrnmsn{jh=)REt~@q8ed>SbVkAlDg_#43cr=)^~R30tA%@Q zWYf-%6SdA}d!T^tkicTM@S6n>z#%t*7u+NJDW<~7jIIJjy_?B29k3=gU5|9~KoK7K8+*`dDz)S0;#GaD5qnwA&z6CZ@=jA?0u-HYfrV~?;g6qa z!B*;diDsEgLA<_0d~c6q4?PVKESVI_;M06w3~obO)xd zWf+Ms?c^5evhWO~-{U0|U~3)D$v&2L!*<*}6h^T3y3w`-;3Yi2noiadXFn(MIGhrb`cZC^xFn&Lp|0~%q2eQ^ zr0TLP?B6PiHzKk&l~!w7JfC{i2E#y0q{zRTM%$guZl{}Vg4b!d3cW*}uV}GutKv&b zs1ls@L6Y>^X`Z2k^Jh0()>+g1QpI@>t0HqZ@k*LbKtW9LzalkMxymV&Wpd5PQ+YH7 z0a{NX?QiW04*;SdlF-{}o3@k*f>pdzeWR5x8klZ0gv}nDyj;0W=cLgp`YAtC!u3}S zZ%ACyFULtTcsnD&$I>q0_!D@b_&&^*TY-KAv+JE8u&a0E`<_16Y`XKWg32<^Vw!;J zpC645+0Z=HxU%p6&yQUnD?A*DEku{9N765Hm%?)|l{ox3Q0#23%kXXgU~BV!gs#|L zcKrjas;9VLRM6~u=^jv+CXnrlfc8`6sgjKMRfWabI5D*5UE@@ii-lGdBzK_(=0AFQ zohRq3<8Zz6EN;oeQMs+dz<1?HnH5#%nOn=h&SLeAiQKA(=h`}31O)-CDez_SVSz(g zw>e!Q;lSuQ$^*)32K2Q5FjgksOfAbB;scg~VK52G*k*s7_y<^~^p?c$E;mFGRNnJ2 z7w;WqNOS=&|M=ZWmx?rd|1u)@*%tLR@a7h=!G_N0271ZX!`k<{sK9K*G1s1}ZRSy? zmQ;cqsHw4dv|izH*)oM#SnVc~Gc`pO%L zU>z|jy%b*0@t26|7QI3fWz$G@5W3%xK)0h{j`MDL_MAh)nkpE8-*}`t9xB6Yz%BID zXgCvmNokx(u{?^NS47KU;)o?9z$Wl33@^K72En}`(U{8qnME0VQ*zqOYl@minNkJ) zxWQS>t60TJQb~A?b(}A|dagPsvgd>jpC(Os@Ane_B=-Fy!pyNWfOh`;p~(f{IP_2< z={m9p_=YDpLO#`W`u1_&5to->A6uO6u__*3rvZ}!VP+EOJ|(Qbau!_E+hrP<60AHO zhxOdjmoJ;S5QCkAAhc_|{G&{fuv6W}jV>>uSzdcLrM=U<@#_=Qrg^(I?^d>P7$Fy{ z=Xk6BOK4NOq88&$?Y2+U?PF=r_Mh;=byaM8<}KSGn!WI=TF^t*L+CCqUX$y6xQCk^ zC7&C%Zo*y?eh6NBe)@{i^;kPSwm1+{Pd2~xS+_2I(>Ovb;{861BAEBHZHSNVA@Tdf zl$4@IVn+1H$b_X+^B*K4)6kobJMVhG{V26!-7a5&P{suQ#NPs@k|Bb6s7=X-t3YQ>maKg^Hm$jgAQ9GkG`0vP zJ5CcmPB>|t(a}L4Q-^Bc(_AJuueo5LV=79ra<5(rP)(;oZ+QsQB{#p~!-v=%C1uaA zOzf6Va7cFc%!iTXF_3wVY&ja-kmL3GnRYfnx>tnD*{^hYXoLZ8vPJpu8Ts&n;HvUm zB=0YG!N$&jG+@`1mC8H%kIjg)p)rf>|0UUn_+PHSYi06zUkT5NVnv&v7`mN;l_@FyW zqhTDtfjYcZS3lLbjeFkZ7sI7T(w2{4=415 z=7gVuK730Z2I>~)g=NqG9$}b9e~&d{tuCp|6 zM1D)Tk~8dn3Lg+U4w^qx+b2E`;fSktQk9N(CCfnNd$f)SC~NFvmn!5JJLa;Fg8lH0 zAatog&a^RDcc2QiRZm<<&~wg%OZFRJtPBK2BE{(m-Ttg4b}UVSNW#&$`;(WOj-wB9 zvYco^2cKhIy(St^@_oP^Yi-wb4xEHHd0Dp=F?lQMWfNtrNaEYli$P{BJuYi)#k`1o zb*yXjkNJ))-UfEecNM&p+xYLk)V{4rZq)L`1p!`7>%xw7-Lo&ZAxbMw*|~prR##RU zrQ*mq@%B>cYSll;_Yz&mBtGlw%*{f)539tdlL$l~Z)x%QwhvKPE{t+XE#!DbY(a5l zm8krl5^h2()%W+s<5X&BTzq}-1I6=m)BOx>#jmlY8fR{=O{xD=Udm&C%DK9@0${QQFTWhi4&hWi1&-}wqazLtgRUL}*!sLh1HI8`+)C0$O2 z!3#P2CIG|E9k!t|c6Q)DJ&5^3n@_`Xo zZs4H!0De$uV-CJKH;Ov~@3n;`7=I|Sj7&9w_*LTx%l8)h z2bk0lmV3KDoNfzB`VRRA7(O>&k*8hF84t3dX=OM4nNUn)pU{ui%+x?O%XF9pakn=q z`#3xP^Y8Jz-D@;kxgqjC!{g!AjRN11T-$i;2@UK71+kcL=G4%`;M*#?7|cxA8U_#U zDT8!k?UGSu;}p#?heu#pC)VRBt>+;05E&F_TiJ0NbzAp*gJ++IGPU#%aH1rUpD9kS z*7_EG!T1elwHe4nN|BlsKvQ(sUUJwFug40M@WG>(^MPGWYN_q68;N_EE4N;`W5HrD zn1+3hWj?#qX}xdTZx4$oZ@-%MqgB}`mniwg^O}RXoc25P3Uik^vSw@H^h3iG8f!lPRc`NJ!aB-7RGWBBLtPVk zBwEIplFvt!JneUwXrQtyWp%*J6^TFdx8hh5f~9?N=EaScyP-Xo1-Y65hwM=&gL|1< zR`Qr-Rr1Cw!lfpKizD;0|o*XPsKT@x0tXPEl zv5<+Rvhw62eADGqk^n4tjq}=%UV@EOG9!Q+qn94)d!guTpFD!rnA-r=bPT1U72SBB zYiq4;pR_CeSHiz)`qKuh)`w~J#*WrprA4e*GSoS&ZR{eabMfVZch5gF)Ssd z36*_rWUyrRqS2^ND2R189U0LpEPXCZCKSN>@$&6NQ=NG)5EvA$-{Bt0tRf@#9!R>7 zD^J4x?UUOfZKu&7nYi_*Bq-JN0gz;yCt7mOVVv@t0WV~!MfHC}GU7YBqTp}BGcc`H z%zWJ~*LHZVmhKA$xKa+3+L+y9ySkY}vJivab5|m5?=%koo`AoZlo*=iP$yw*Z3UDO zT3?Hj(CHrwDVh202v}KT5P+pQIA=q%`y72fmYc7J>M-U;kYsUV*8B;mHTq!L-@!)c zq+>cPj4FqjJ&@-Q#5?G^?4z;&Gg zGpcMJEjL3ax)2p};#EnGM%#v-(AVOUpB^z@9yvmK8PTba{515>UJLy(8tL zol8dxtqV!Du_rSn8k0{I9Gy*kALxIs-w|3cv%cCn7zGocM5{U6e)N-RwrqrJOJxWD zOnOZ^He3Ok+Env!4iejg@PSz=Y9dm8@%xvjz2@VC2oDyuB)y`B{FJr$+fpTyy`xUS zk1FtBFTYvy574ZC$}9W4@iL12s*prU#iwrdzREq5DRwq6+PfMxF&O_u!);aS`|@x6 z$IL>HSvlZg4>si=quJIpw72ctQDM@pCA%>^|M?{Or9X~PH8CL$e98mD@%+Edh#`|{ z7Z3W84z~-7NBK+w>!PRWf-0HM6{hRXcUKAqk5w02{{Y4sA=jGXyV}wJ01C+ppQmHM z+ByL>+Wj5rwZ-pn*$8y$IerQHSS<2b3g_?BwwO{K6Uq0<$upOOmhxS`?BvnSw&g^h z-Ka0k6m%#Q6XzK*g!Ql4|2?XS_kk9x5w&>>N_rrXE6N$TOuz!TJEo2+IBianUg}wY zOH(o8pn;o|Pk+$(qR(DHi6-ylK zV=`>Vo}s$>)%r6wPiqd5 zxKwf&PTs>{Kz+Z@ASSnuJlV%3GRJ2<5~o)fZ)eMgw&*uYl!nzIL%#WVXYPlJmB+Yv zu5nlJdRg7rBN>{9kRL1W$d4L;9?}CoRn@a_lEBZ_UVjXRFM{M`K%>%^dR-h4h{r&O z*B7CG08oIefPM%U1d1T~X6xzgP1p=k>47HXr5YJ;%>3`3n23vFlfff?8any*(y9Vd zc(G7aX{pMbu(1wL_7p@JK95ykr}5hzM)3^ot!v4NPsu_b{f4P}iiY6k)n2$tS0WP#&n=G;?A}`Kj~6ey79YvBR!2Xo z#hk(teh-8h5L#nb%=+GLR0KJf94)@NpF?z4<5+_^V_~&I`%G@Rm|Ruo1ZGk zv&UI5Rm6ugID#;#|3CKLIyjCeN)v1`Ggzz^vn*z2YRO_|ZZV^U7BjOf<`y&9VrC|b zS(e3MS+M@i&fG=J-BR4_U2N=q5wBimR#sBg_uk8YvYs_E_7$U>FUb2RUh(zsL*4+4<6ZMFd|fqJ>pN?X_fmPZT$UcRP_) z^Ig$-A@$B1i0SVPL^|ud$}#BrnoMBns++*TN4VYa&vcN0*ytSaD&vi88f^Rsh4-~S zdl@m^q`OOhUg+lRIB4e)O|{CTk$R>s&b$4s26RC zymCXonr5T0{)y}SqiG%ev>(Q&Fif0JxdP!DwIZ1lsu4@YYDLF*jjqx2KHSYaigT+% zr)$#J_vucae7Yrw4dS_b2jI??tH1~|ed^6G{#?Zw(c(wyC)>K@zLDf&?i{w47AOx$ z%6tOb<09v5^zG$uJXM@>1-_ff{3gWVP`iUfraG^tu;m~wfCwfi3*ER5&N%M`Y z0sNtl*awoPfa;K&jB#N|72FNlW&OB&df!8*5jleFHXRm`g&kiSv`S<%f7`$}{pzAa zje4*TL)$Tf_O!ISjcGS$Z%2BZZ5GdCPe?!l zX#XYmdL|hwWO8*L&T8Gb^z>CF{9Zi^k~Qkep2!|wLav7%n*VJP8VN{}wA_(__q4J5 zh^nXitr#Lj(-B1zjb~y|3=khtGKs&ZDfX`J)vH6B806WMHRC!_xh=5<7CSZco;Q#y z#1EaY@ohRf)1zW2kriJUe2w^yNk(WVuzuQ_{aT+dFVr68ked>mkTtSZyJ^#zL+u(DLFMqprR_p8x&%Hk#a64zplPI%E3Eqt|1tr4we|3 zNcQpKM{xXc-WGaGE)c_|Lthk4v6Gqf)k?-V5t4@p3x}|I4>|l6&+lWJDEF=9dk$_1 zkMmZ3OW9Z9wW6%7kP1tq9e1t z+E&t>Aa$l9BaJ8pLZpW(4L6LYyhv(FH>d*g7hQ0ttgV>Is9@NRjzk0Rs%#by>`H^C zCVM!~4g8mTW}>)hSDFTiezg%aN1sW@>6ozcY`Z;FJ0?{d+dQVv>lXVi6=$2b72+0w}4E*jA|E-&*8FfMH z>hfJV^*g=}Vx9RCw#3{h2Oq3)HdmN7sV5z)K4eXu8`4uuR&iod!E1Xb9a@^AN>wRGMdF&hg42Vq}+y0 z%l@>D4d1{)m(y!d=~P)itUwC(K*7gxdFuNU#-1Vkv*t+CpZOZ)|h_yF#EvxUG}BiKfu<%ThfZRQLe7e|AupVvLr_#2EE{{_I1FzxjhG zShYoV@xF&&7Y#D~6@SOxV)JLf-|)*k--zAbYURl?dmY57mxeq7{-$wo=aw9>N_Am~ zK39y5VpcN321kGvDg5TMUXjhQX=FQZn}J9jKGvuHx)YoGsw|S`J0EK{4j(sAV{U&B z!?l0#!*U|pRGdwouumAFVm+CIae6W#NCaMu2R8OVEom6`oU(yp*acL(sK|dr5?zo^ z)p7Ah+oCJp)eAz?FOQKb$<|A-8X~X+-cV|Ox#dcN|KzL_*Z25PO?r~GgS_D58aBRP ziD$Uj-N%qb>7Ixx6u2oIc)8p#?)mM5A^Av;Ajc|05205M5!=tDXF?BFfSP!4Y5 zuTcRUCHKz~NX)PfDbPF^ys9f)_eZT!u@A=pPds}+b=Jx=-NCZ}Ku~U)Fb2d2;~J;m z^Er@MJ`g{T*~)MuNtWXDoBXuh%B-O*vv^iVih0s{vG+to4PSy}`uq`YQsksVkUo$E zb@}D97XSVs8C>TqH=6w1-oi3{(5oyzcTXFF(op^x!3rG=;TT!Vddn8!x4Fxbg*r$+ z7j@<+sn>deu|yX|X*>dQ2lJpkpmxlQ;4`nI?MJu`Nq3kA%Mo^D-rvaib3&YLi|mYR zat-%KrV01e$ME{juY(EP(@jpe4XuO*cYXX97h0or%#LdpcWct~WwKm4!`tA}IQISd z81vmrG#!Ei@8WdJw_FR1Je4>C*>RI-c1&U*0l7<$lC*UO^EE)`Y5FfO99L+{*{YJF z!;>kw5l!a@Ib{W7I%rTx5a5*M2=9BRbZL7!h|P1Be^xr2zInqo@pKk zP4v9Y;1`p90yUSm>Egn@DfT4QSQDIytAoc8UwzQG`Fk`qMLYd|YtS>2t6&E>@y!3uYgkpd)+s?aF>moC?O12^Swa-LeH+ z2s$#{fc1g1v@aeM0Oz&6E?+c`4A{NrM8tY`?!wtuhGo+bBzKG|F3Oq6XlD9#x9S}b zi_SzeQ{3oStypjy-nkyG;FFy^UKQZ09BF3iPR1N*zz*quyu|CQO8sVlt{#lR+o~@yDw&fkGg=e` z=H=fZitm7Hv^ft#Zqg6h;q&s$g8_<41&QR$E)TVzpRD$0FbkI5w-D#(Ql31fMs9r) zU5oI52m3$=b+t;D6)asB@_4s1A)`lZTeG{UIW`>`1Zmjd++dK9*&-i~*78Mdx4^7Z zvz<=Sp|^#4y0hCuN!^W0ns?e(q8tCq{>#8pcFf ze&o`C>H2ywW8DdQ-R41U46`SZFi)A202c2zPUv}yujMyPelgfwf5SJ;p+I~i;w|ss z-38hdVxCg-k9oLY4jn)M)GZws8Sl=NVtA{4nl zz2Hklku!**z02_G-T0N%oZj_5xs>%#p%yX<~h>16HceGrJ1NS{e1lFYU`Iir61CT%env&l=fkeCnlb81G<;|ArOeF-6`f<$A>51x86_Ct$f?6rRdk399IpFNIzwlYhgV8BLY;D-W` z_EZ+jZo2GP6&HK++U=ojHQr9qWyE;eR}Bv^g~<4t_h`QZF!D*kM)c>&Cs=maIyyyM z$u|wdVxt?dMdC(Iz~ghAIca^SR02%Mb`_1j`Ta*@K3JGGZD;n?yuT1 za*oy+xxZ0Fj+%>YZ5kcKE=ZQx*+~SPF>R77 zX$J)&yb)h3m+uK@V2@e4nV0G_eb3iM9*hk`_~sJ&RqyfYMBvp{;cuY+MF^8BOgDj! zX_>G-2qxugi`~q1dg8%Xtv)=oNQm24MOt+X8DnVa}aUTutxn>dj3)yrC8`z172$E7oWH`XE>SV zIFzsewcvPxWLwjt`X^ltbAO`V7BSbK%J7G+qAdc!6bF2f4T8{&3-b%4$+1cm#FI}5 z{^}RhF*zl1D^r>i4jV9{r}Euilz@CWl*{Ejo65r|11lcQ^SF5NikXG-P}0?^>zBz# zY=s(pW|9!E4d22(4`Pb6VPz4VQ!OJl#h6G(p|oUO%_snB2HA6xUpO)+wXHtMZ1{`F0*(Dy`l`{?8iLpTrO*XvdjR_)OWM=%8^08=X3m-b{s(cc4Sq zPcy*pE@aF>PjxZ^xkhD>Ls_K_Jd^Cm5_hrK&T+%!wAYxEMI^V11Y@2SLC;6}C!s~J zV0nIVgrnJ2;vG;XYsUT)u(90z4k)>XZ?wJvwe*GFg_!{8ny`6KTNDjvD%U4qJ42BM z>JQ^|cF~`P+4Fkr2!-?wXcA;>{lzkj(JHNapJGff${23!c<7KC6WLrppC{S)9(d`w zA<sq)deS?G9)Su=z0zw^55567u7+;hye%}=KL-qlmQcx!<`^NQ+IGp>t&5>CLqI03S zsp4$w&)6}>Yl-=YQ#p<-YL+^Ut`@k>X}AUPOO6UNtyt4TG5lFh?Hr(skJi{O32A2EY}0kd`hd`2LA3=gR}|_- zfXoh51H?*UqBi%Te|ce2^^mdEEt$d4TCS-bg-M}ap@&*Be50Up?CuZZ;S*^*=jkV+ ztpq>n!k;DDq>DI{B6;);OmWB>BMOPTy2S}`_K;r?o#7jX_)k0t zqM~WieRzjK#vIG?NTW~@YQE-praKT@2fpZe7PU3=tOHddzSCqzMy5rZ4}iF7z|s93 zq*{Yf%-LDTqT)hd%OmfflT;Z>I*oFe%XHn(Ws~9+xfta|N|KUof8Y{GYkaCG)IbX& zfZ4zeTZ&7YfbVE(%M0q|kg5^;ooG9Ri#Fds1hk=QY}5IAPdQkqf5aZDw{g0RKaa%n z)4|MTE}2G6q`)dqRc!gWf8|qHL3dm99`iwUfr*GN?ux>ky957N{epvu{Vy#b#_V3n zsx^U?fiF}KO5SaVtfW}nQcKSwEmMq5EtRn5EZZNVxcc1y^5zb^n9sEHywl1QM&F9T znZzWaJr4E!FbF*`J`e}i!khBua832Z+TfsPtQP!MEzgdZmJ&?sx(uS)0B`6F7}&hN zqM(A{(0Y+BkD_SAJe3G{!P^<1lJ)Vm8TK%FQO;;DYjbm3#E|uI#9}IvN&-NFGyL5- zEVekEu=uIIL+*}Ma0!v;SJ-d%_&g2ruVdJie}c28roWmxzdYh=E$}`HDUAG*<}V87 z34sqCwV`4+zr%wISRG=kpXIWX*VA`kU8bY+#G%^)WSlvw5c@%TL_7P_|9)83Z63?JI;K){qVyA!L$ z()?f$NR!vg6=ElUPa0nLJ*cB&gwgFpHyuX)-sk zsu3Dq0D!oj|LL>eLxIBb`tlZL`W3YT0zG1(Y`M5XTo7Ko@fTLVqgG)Xc4^#C-lRK| zfMeg8bR~uAj!WfmL;$PXszzB-p^@QODwM`AMQf@UP~IJi7|3N9u%*9@cGO#?&`_M7+jhjR&5MqU#nO7 z1zbEVhc%Qdm|>6gf_Vqv*|u~X4%3dp5PafP!~bBN<4*0{BH(fT!v9=JhT5PhbSnrD z$Upk3dj(0+PgNgBa9f7BPLPBqrD=Bn9rR=wc}vy6U_|aXMR)Wkxg953<)vx$njk+DMK2^okl^Vn6NccznPIbS9?IM$9E@)Y zjG2>2(h(T7*rUUwX#&u5=CMazB`U_wTymFKIhOy-=T)AryF(^;;!wXl0)qmEtFvbS zC!O6&WrJ(%#6FVzf3i=s*+jFZCX#aZ@f!Tvly!EIgYn_2OY1 zCSjqr{!O>)tK|mVwd{4F7lje0tuEG#{j!(}kex$i4SiUztsEf&DM;SVh~R`M-~SI5 z@xB${AmCnl#UJj7WF2%6l&r!41LKqqroO&|wEpqfuY7aUPCgFLMpIb#APLjIyS#`l z^KO7!I$Y|;R~sU$#CiuT(kk~_Q&y~thHUzkZPu&@)H*~8Vjeq{u`6>YAa;u@rc}NI zaMFM6HtNnMi~oVrI!itkEZ@@K6hplkL5?^$nBG8LJRpi??{i$Pfy#PD&%^kmjj`;l z66Jnu(cPVaa9>P<^ML2E<6uUOucvW?L(Pz^->|7A)a?oyx;}^r&k~?I`{;}$oq=61 zSl_$xOPb9ulW<0$N%6x?lrW-T=$~MH+YD!E2bKvsfvj&FezE4Fd&%G9FAZ9Gvm_`bicA_NY;u2?^G|ObEM?RISTpwNX$IWgF)r1aktUqyLKuKmMJ<7U3>IgF{(7p zj6RouAH)a*u*uHi2rYWteh?AC9zmrAEOTCREQ%#eDW&U+%YQ zIR1?|+=xl9xx+JAi);OIVSD5#sfK-_{25c`=}j|{lyeQoTIa_>1GJ`Zf(tzYzcvji z4L-7i2B^qm9&3(#%=L_f?B^~vlWe;codJh0zu&Y3+yVRNnnq;FANDq$UiKQ@0Oi7? zcrHs07f#E>S0%(M@m$1-GEdJdB(vQ7#E~t$S<#a;%+UAJq|*Uua}lijsP)&i9k%`} z%WkxYY{^K!t$kD5(&C~y4f99?9>^+XP~LEqfaiZ%1eq zCcZ9Q{CQjI?qP5A)1h1}%3vaWk|oBYcKV}{o>N_PXyGh(=C)kn8$0{LSx6GIG{p7B zKy6ua%+f0H9K^_~Czbr{exa@tE1v2c&^$k*p&gaRUQc~_U&YlqZm=fP>tnD_Fr#D- zlG1Sj#{2B*x{CE~*Qx5d-hge(Kaibm1VRwH%-x*LLR}`yod-<+ELx7>>Dzk#w9dKp zbme8P@iI*b@^c^K)(kb&_iJQL?8Z*HGUo@r10Zssx~n^mtiUfqlh_H>*=vr?<+>NtB)xH0h@&&Ji!1|O(7LJdJDyx^(C1@i-4mWsWO>dC znbUtgCp71`wfhUzO?_K82iMDT#p87apJdH#^LB(wql~y=#}3+x$c-o!$@3mLDF;nr zIONtrFFe)g5uKa0qOi>TW3Ef-rtf0?zht!OTrSqDkMd}=u!tM`LQ38N#?NZ&T%N~K z%sHd7xpEz>qRRWF+;(p98cE^kE@xob2+agc-CH(P(g0MyXdi``h3~*OH(rKKjF^Up zQhNSdbeDyKqDnh~e|Dha(zvRW|HEgqQ#rSxa_pV$A}Lu@J4DxK>P9K*yB2yL4a3p& zV;dUl{$4HsVPsE?%Z~z*HrC6FEQ~x5(grX!;&YbTZ*C%Uama_vejTi{x7}Pbd3cbD z++%oHy14OXCJSPE4H<9?(jTT7sa0Pp9k)c*vV3)RcG;>sYe|+?zEVL!H22YrM=}i& zBlTTM5KCKM)sD7gDb2ZIvnLDN`pCtv1%A8=7YqhLxN<|B$8nQZ_%4!(QtUVW#KIr~wmb&Qs+Se!xNgj)nmzETstKT*WLi zq`=Wu&Y%XsWcg zB>(3?TW+@2i82x7hJ)^Zi?i|hJNWouzp3=SV0oC3dc_&X5PX|x8WSB*IMf2|hM|r8bUktpw!68B z%tn3kVStJ9z++~iltdVE5;thB(x!<`9{K!tS?@i%(^34cV*6Cm#aI^niSkL9D-Or$ zi2_Qibf8!8IfvO76P_1)hC4HeTsOx{A~&3UqZ{8G=T*nCs9|cwYMOaC`p3;;Wd=~& z|EyT?t#IF!{%{elk}5DL43U5$JPEwy%3MxINphpEGG!X?5G09TVuD`Pv)xZCtH+!}CL2CtZ0VyOi^+kv7 zb5(-%5Ra-0TBjnh%r{$XY)|6H?PCpPhSnu6{(ZI$f!K7D$&Exm)Fq;KfYD-6N3vFA z$Oskn0nrSrDB}uhi&5k9QCrBq@)C>b2mIh`E=m6o8H`cX$FlVrZK4rO@=<9gOPldh zHhEyucZ0Yoac_n?n-}f=n9?nJqF9?DtjxnJ9+*H=_4L+V$mT z!p^euUEi29{i59zv%7|P9y^VSSyNoRrE+5^sCV$F_)b^wICsnHn7tZLdErW1llU{f z2B<+%k#j27UfliAL%i4jW~G%z=@Gar4ul45kfoj-E4MCuv~Hd^w=&1b3=xl5#R-l+ zJvg{I<8SC6Z(`ZgS2AZmz;p)uWGMZyh`!m6-SG>|^6@AI)SCSU1-HefkunPvdS7Cl zEMNBba}qn^I}xtVE0WO_SF{aGe7MIC3!+Va>lbPMUqxsnPTo_ z8Bwl2%VB^p@2P5o)J}+GzW>9Mv_rM+Beu-Sqqo#3Q+^gHLn-8jl36_h{TSJ$m67IW zqs<$=RrqGzxzJ#|;gE21R&mTbKn)pZ7V7OTCrg4!PoRU~2*%}7GCfb3giwwuV*$JX zKDnn&08*gJuYT1_iI#4N1&J~QjQg4_SKSWxrZ3Ze>!3-Kh$ROr zHV+>Ds>CSvos**m{6bJw*>^M}VHI&t)G2C%=QgPOR9rWmrp^$PSpfsV zYiOF({q3r~sy2(KjlL=>5Qi~C5inqWTN?W%_*A{w)--zWN3r7|S}Pu@P?>~ypTSRt zB6jD2$ss7A_87NYY)`T@!`gL|q#!6NVX;&BFC?o+y?1w|$WfGpuZEQ$LS*>VB)!V! zS}o*8H!nAEm%SLvfr$bRif?KX$mtXblKRp%6IiHjw^cfN@yTa&vG6iLk{u}3OJ)`3 ze1|Tj8uoBN@J(e0OYx``EIX~JSAUCSf!o$6eOl1jlSanvu?m6Y z3O9XR32VZI(6LZP`(nK-=d>VbV25yQh|UMOzo0w7jNE>~GDTUv2#n7oK!VR`7QAPq z7GdT^7uZ8L*llTAZp+*_StLV@DRck1YghQA9;*?q#P&w+(r17C-Ju@d4Ao6TH*d=+ig}Ve^Ws9+MJZUw0uSUY=Ph5+pB&+C zF5HRDYS_fboXTEpr_=fbEK`Gb=P3#J8MRx~uFV69_tWC3cH!R2MTy@5*o!R3a0$JD zH5u7I+QqE*Om=Hu=Jn(*$D!2N;Cnqq7N^`5?F?lEo9lJ7`TFZ#pVhN!cR=z*(!j78 z6;q9FS6`OxrY>?lfLHZJI*7FWYAB^Zf=Z2p)jVwP;ncS z(99QC*d{3jso-*;J(&0{G3i{b$SnSTV<(B1hm%z?%|1Lvbym76c zfoSOxDKe$$KY0wU!=1G6UdWq!@N1ce2n!(l6gY233CMK0K;&6^PCH?yBnHEQBxHGB zw4H;@!yJ8>@jEGE?f#ezcrfj!+8NDDit-t;KdopL>5p-6far+G6YT(~1%J)t^+(-~ zdTb-z5^dK87T~l7OgrI4T%}CU*Bzk|#GSkdrAxg{qo(m8oqKASSd*ph+O$Qp!~Kl3 za$0#@lVf{_HWh_IY&aSAbMYo{tcvea%`PIC4z>x5+cKWe#0}7T&Bdgr+=;kWR zzy!q*9)eA@>-lKj4Jrf}Ck6WPzos@ax6M=AbB-5a z@;u_+kaAqab$jxLEfyY6KM9%(`k$IW>~(ITjZLaJ5M~6Isv}s^ZSs<_NUCs?1+b!H z+mG9%rWf=Tr<8l8y}5?E*y_^7V69C!ToM#IHc$*fz+@C0*q_0fj*{$vw9N-Wt@LE6 zX_oF1&BF=OM_^!&@`>A!*^WS6`rxktj~N$@EP3}>arUW12H z3UfhB`NQEIkFq#5`nb_XwYtgG-Z4S5WFMiRY6fIs482hw)Zk*+3Q}oJ2lY^#0=2np zo$(6iGcjQXPG90woP49=9(qTUTz>^I1cy%lGnzDg2FtK4bi_>u`LuM&F?`(klE<<# zdA;3&n5;-bwH}P|bSblr)Xw^!0@*Q`8rYxS0nC=lNF#a!BjeCFu|#Y5YN3;8MDbtZ zzuoO2?kBoNN%Goja%IVxHwH%!&TcfCUVrs%~BgmUfase9}Mal}_TL zXrZeQWv5Nw8Ih2npj&weN~D{?NX{Ce4O6^6C|DGUp>aW%dO1y*JNYY#5Z<#MknJIvSfk8_Q!?+A*3Wdd^tp*? zBm&7quXNIs7;)X(mRNmoo-lnl@e*>8Dpa)>&zve>w%OZ zd8SD&@ji5mTn(ejnyPTtj>BICXX6I4v2(L=^8kQs{}aaYe}RXa zyQ%Bnp*YxpEgjt~9sf)9{#LWH`meO}pQi%+KlJ}9fWt=sKm!0EQ*)q-I#37rulT17 zfb>^5{z}PT`6s{p-`AwLhKAa|CjWUN^RI~fyDCusXE~m^lLM=%v$MS=tBZ%JtEr>A zt)nHt$p1&FZ^!^VV z!+#O~oAaNYg9E4}FCnfbDJQQb&FXFH?(WL^f5%8+kd!zVNW+Z-Bqd?12e$SF`MUd> z%F6hdYI8g4Nx5+=+sS)@?6f4*B-yysq~+O^G}L6(q|Dt_?WDOt>TF)BV33cRw1PB9 zLs>=*Y>mv%3snExsgmx0t<(PpMgaN$>!tl)0{_PU|9Rg3zrlZYb`EwP_J7BJE*>t9 z|KI=rzk#Sv%Ccz4ME{TlOFLjuZ+Dcs*Y z&|FRp1OWKb0sui@0D$MeML{QjSMu-sIx_(P1pYeq56-!rszQHnz?&(`O1=LJkaQk5 z|CS&*%jvrP)pGwLGGliC#0CiN@*ru1Jy|keT<=l+OP6`~%*u=1LN7G6YD$*kaarrI zvaLK2Gc2N`qeB+>^8SAOj`Dsh`kr#OS%R5xt;Ty_X0ON`k$kSOtUs*Ms zF|g}g$@m(ejs1PhWAYT2_f*44ZLtNk7Ph$fM~{56VXNFtS$f-JfZ)EcDmueEexx(% zPbb5(4*zvUBZTM0i$C3s^tZlHSe2T-TE}d3yVJY#_st9MO_}e4vqkRjQZ1@=%0B82 zA=#dluhp4VZkt+kEmk<@%r7$U`Q~_bpB?sCwU!#zON!3D{I$#PfUm-5GS97Z+7M&= z`@WtxmbbP_o{z?wLQW4CGiQ#aTBqX%rKxwbSBonT31ENQvywW{X0>0&`g`?s3v1Q% zFL6V|yLwXNXL$(nvsUF;`2+1UncWmVl?Lxw8CmaFat9xWr0imOX7*WQjD6(+ujN^z zkF&TMzOC?i_+a`6C&QkAme0s5KJ(3w?zY9mnukv(=qsn&HGQL#EDqW^OH(^po!z`| z%W$G*#?ek5Y7y!}=i71iXB9?L0g_IOl=(k@k!;X?3$XtQwJG(k{YTd~e1JwZ3~ zxWli*(B`AhrlTj5tHQ-2VmLBIUq8-lNnlt!|Eck3sf6jbwpS+6&=O$>TPGq%uRNQv zN|Tx{wX+{?ijHVVOnUQv3v6rHf@SjoCL7yCL^$|uRy)HuEc5m}+y#1gN~$sx3m2p_ zyh7RG^#QObZSP-X7SptRrYn~jwcQF)(zM@(l*Iyx8z1R}e19!}p1QE}j(Th2BHJUG zFkd9o4*O-P{)5mr)dqy#37TSDSvY&Mc3Io6DvSslD;Mr4Yo5s|5%&)?$=n}0a{ovW z*a~isRTyn%f^Nv1=W)2z+V>sxYmxY+_dVET37imc`M&O=snGdtdLyxw@(Z{|rXaOM z=2Ggl^{uu%>Q65CT3Q5KeJdK5EG4m8AX3~>&{*@T$^2N`*GyAtp6K(PSl6$at=5YJ zbBWo|4Hgc1PN7tVYSA&CdnV!bKd3G=g{1-OE#xfVZ)8Q=6|ZaWkoy6z9}3F6Hq@R- ziB%|aGvsIP8X2vJ$>?uA-@Mi{+p`3!nHGPadkxcfJStU*Q5gE~B9Va7$dc9YQR%{~y00t8_d*r%(b+8f8MQ-sk?Xt!w+7~OQ|P$&Ezeiq>Si<*tEF}MYfdf! z?P}ZPpGqk|`Rzq@^vo|d**1TzzWTC{vhbQk*LmW2q-3%u?B6du`;BVn)b|aYhe06i zpu?K)ZPh<5vPTeN5KOdrITE3XY+^0e3P2Y9meo=w+0KNLDR!*Pf@?a~Bu(9NVD};U zz#lQpRr?)PUWKFn?$zd7%X){6&yqc$2;U4Z_A~s(nn~t7Jd-qciZYzp>&eexE+>ZQ z#p(;>mxr_SDCr^;uy8RH58HRW7~Qf;ky9&aNy9PzpNoG`&Ufa9+d&yMpWIU(JNnMW z!wxxcCr_f{GRIiYhp{@P3w$Yc=+PzNjGEh``No_2)rEO24Do}DWZchzBe*SFxTQyV z_OcC(h?(d_DEtf_iWhYj@Mjev+N!^(;uTl&CALI~bO%(zCa^kPwE+I~H(sX&v^Lmg z?A1TcP@rj_@wkT$ldG43UOM7i>%BUcGQ^>ZM;{3y1l(Ip;!Ijy5ZPcO=!^|@7K3#^ z-EtGPAn4aO_%&lO+W6CVnii|d6MixI6kR57HIQq^`q)oZq}H7tu1C5uIAIB93+4r3 z$@E4@>)v4!#?TqdKm)dp^>`WmKBrU37*ob$Jw+WjK@T;nnI#K40d~hq$sv0=$zbX+>6pSk z@{E(<0>&Db;`NTo@kUaj%%x4HIix#GK4!+e4N;)0i_#)>%6Po1Sjw&zIgQ&Rqr= zrp91W+aJzU;;|xmVBJ386IviNkCkv)L4KZX zmDaqBXa9>$o{gTRTZheCxA$N+ItOxy)IBrpREb`tv^!LvP6q{e$vQ} z2gp+q!?Jm^W{a-)`zuwFXG!mMO}Q)bZ4$bSCn0ho&|l;B0;ZJk$rllt>CqdBb_|P4 zrRUj-IljbL-A}gba(4oR_^V#Ck?*zQk=ziu`{3VMy0Db^aZj7COBfu3q7K-y(2+|$ z3ALdedJvD3O24diO7X7gcBpZ3zY?jo*G^ty7k{NI(c17N>m$`(aBO)OJAjv+;pSD# zX%BvqQ9`g^|Dv9h)!mgJ^L|+{BMn*a=UHkwvYcC9mF>i4=*Z}X$~(|hw8xuUFJMBU zMcse&O(cU>u&O6Xf~$xV>#WC&LNdTU8nbimkU1`i=ZfmWwIYQ`h;3eT4Z3%5k4hZ# zoW(|f%rOReP`hV!_XnHL_oJ6ARIGZzx}hoyuHoJ!EmE*0VEeeF#v9%f?@|s5t@k_s z--w_jfx6%Z?5Cu*br4zk)JKe;;B;5!-Yx=i~>N-5(FL43Nan zc?U^-9MC+!_ghOQ!xoR~m35Kv22T+;5JAT2%N}Z_Q??!U2|1_|P(M%AFcjgU-V`xN zEo@*qo?08+5b|aw4jddW(Yi;be8REll`mEX7MoA%rX*b> zQj!3uco=u^M?IQvZHk|FT8O1(iIje@oAde+Ai93Hy2fT&Ykcw77?=``f0k!yFtpsj zt8WY?U82`>l|`l1P|P<#)pE&U(fVmbDq@H9lWoO*^HE@99-#x;<}9jOom6222-hZf zl9Eny(iD9E;g}qI=0kyvJEAQ3vramE+G3nka1d5uk(J_ZkVu(TA;DNQ2Ddju;?g>l zPr*GLV*H>#p->`{WUz8?Im%;g^^}Gu7k#e zpNkPB?b$h$u##H(Ud2D5y~%ZM&uZ}PLO{s1OWeYeB8xUnOsHJgAutY)`bX88G&gj5DRb6Hu$%G7V@18}chn0csHyEyy$Irx=G%#==aVh5ymoWZ z`{wI9vg~2_7B)7EaAlp#YLQ1BbZD4Z`5^^F;d>1!=lHNlHfHqg{gFsEKlVPNdpaf> zHb_v)2LF0H%hiR^yC$#&ad}ac^;20nE)vS^*0J`p#&hsU1MiogVS7Us8=Z>vm%oz7 zy^9B~L?f~_!j?iCeneR{53ie?D$2b-5-M4Z2;y;P*S1-esfiISAY5$K|7{sOi)lkZ0QY0 z)i{>>ZP*m1>Y)eT{UQ1o_3$q}+1hwEoA@Ny*rusR;IQA~^Hd-}Qb7uBZ96pxLqmy| zvXRaU3EtA*woertuLF9rU)D{zHUKY_KVB2-bXulG7gL*t( zn1b+p&$eMdEYnF&xLpBP`A~!82HcAtA~8*J7g>Mu3b3;Zjq2&JZF%B>pW*Bi2~=sa zNUG<4&@NEH#e6pW!mF-ibJ zYIRe(sEQ5#KGG(Jrb_6kP)?i*Tv^@K%Gjg|qAGQY=ZbwPa@X@GPc-Qd{$NVju=6Z~ z&Ki~WAMnw5_^{eow80T4lwU;lGPCIQFIEPb>alS5Z@K7i?84l|EE8z#;wC5amsrnD z@R`oFWcS7peNi08`=_#^1LAgl+qZ?ZwEnzelGDmVW9#& z?>SGVRgIO}(^Vuo>vZB+?vlPUBT8%}%3qj)h5Vcvlb1G8Nfo_QiP)vRXl2&;!9r>T zlwI*<21YuoNt^kb$RC2Dp=-V<7yYiH*guwIEl_0QLMzf0BMs2Lh5TlUog*X)&B>VK z{NSxv%&m~~VHB%f#oHU&ssx@NisfV#4BDwYVK*(UPz!qrk4gW#SJcKi8f^tH)#gjRqrQ0=ta^Sj31;u zN*zB7mdK)r-u=FxwwPH?>VK8xhVODdiP)kmLXu3jEN&&Ws|F~cM29V};$w~Hu6#-D zMy&2?oHwFysvr&nQ%{lZablT`SlR*6o%tiRCB6dbf3(_N&>kQ^)>|4z#>X?$G%c7* z?UFK(>TLF+jjK}g zs!ieF5rSJG@#!QqR>W++h+L?99~yWURUuc5TFqmlCCpAu`*}Ru%jx?`RYGJA47PeV4hU)@6GOb1?A_?dFRLEG;2NJIbyAfXFY5iu6avl)-g=g4)sC8?ZvOc2g_)CZ+-&nGhO=6>vwUDu&l77E*P-}c zCT;wq-Q}m8VP8@+m}#}lXiDRbP7+(Lx)%v%<~oVd=S8hzDzNptz%ql}EURi}hKW&yWlg^1|;?GEA1` z=RuU9Xb=DM3_Fg7qDej?UGNU`^Qlhra^nJY?`qV~R2XUO|HQmEcvsRqe+RoUHqOS} z*xGDt+cq|~ZDV8O#pY#5x$T47JHrWn4_yqJNKLidyVOBjz3@&MD|M#}ycUB{Ec`Sj-I@4v{~cPmgOquJ1Wf8Wo`Ht;>Pv z>Wj}44Q48dXVns^tpk}`CrmkfiSbkSWWQ&z6FQGoPDU4iF!DUAdI*_cpXfnlnO)ia zifXX+|7m08pby18V91E!CUmP{L0n4?9q13GgS-a-aog!8SO-K&a(xV2iu z8W^;Kx4AYCiUQ~ zNGm9-$ZkO%$v9&e6(9A1+$d(P(r@r9RmN2e55lH2O^t9EKSLBrvZMbFWR?aZ99Esx zSxa!;@kgU>BB|sX(@58<2cxwKmi0-|>S_Nqq1b~dsqKYI9g{oOYGR#brC4`eXJIjw>8ynaDTt0iQ$QK^o1xM#@4B~BfM?3| z6_JTKN-f*{tcIXUSzI+?c1FK__LM(c96-u|OKhi4PzFfPl$VHxNaIh$`^scSp%h4# zTx#T4&GsW39~BFlqLBI=mRP$y?kCPsbf#ata50cHs5gd(%%t-C;*<#rjYD*yFG#MaU9aneyAg2y+(h^aeFA zIMn!yu>DTWY(2mC519np;C>YD7=L#&R6se81|h> z6Pfkdu>}%z{|K8A=QpMaFfRaGI0ES#_bfiZ8;Qc@xIyo+kc zG)cs2eIusGMyH^=nr+_{{@YLBd{7W-ly&_qY@@!N>>l8HF{a!ypt;f?*-Sn)Q5Lt< zgiph;(L&W3Q|JQn8tmekwqSnwCtR?$)euX5DG+mtLJR8&zo* zWuM47g4ybHk?OitvG8OYcv`-RDzR}&P9eZsv-ap)kUNzUvzOzIh{~@^yBBu(tA4wKJs90i?uRDLd06?9$`@10 z9VaX3!XIg^GO{JYeI3|Fxc7LSXS>G0ZcoAM`UAb6VHi>q*)M3BERo4nIPg?t4!HfV z^QNm!ei{&Bsm@>=efrmMM#2a{SW>M+Ui%M!#s~Dj#MUaR0Z( z&sVs*;4)I_Ir+P>!hi{pd2-vu;vSlJj!+E27IP8RsqJ-aq`WQK(zF zqr4DZ=EFO{H2#kNVS&|YQ_k#PHsWl#9C!I4ctBd7ohCD%k|p#Ttpu9-L0Y$|C~I85Y;#05#N84}5MM(fg(dz9!@q7`YM5XrQOD;H(@H`>$g~4(3_g=v@8T5l zOw90I87O}@8j;7F$Xs4Y&#$th(ti_73Lw`>a%^SojR*Pxb_`wD%+;joN zY^q;4Kz4OEv+!?E0=yFioy~uz8WiH9wzz12X9n;%qtu1$DzTd5ahu5>hDoNw;!r~{ z!O@}OtvMyQ4F*ej&PKc;dEaR2l*uP*qEBC(%CP*TD|eMg6G_}@p^(f-RhZp_sb0^G z_OmNl8UGkrDEW9@##H&Hj!wNYpA{j9Z`>9@$`>Do5zg+;(Ng-_1Fke6HVjTmKm-OW z4Wn0sP>*us1*>gTg8|cq%E-korW$AlfKZ~PYoBSeUl3}Jt^vS`to~>$Zs_8s)JI&3Qg9T!@9lJ9 z??Mc4u2@!lqWC@9w~os1aEp%Hk4I}XyktmoT#vpNCsRB~o1Tg)^M{0jmz*y>*O1Gl zdVU!3il+=2#mR!3hk(785gTrdw)kAR=vF%>LVcQ`^fWv>Z&F7tPBIAnF&pVtIl6n$r{X*b(VP;)nfqR7*j+mQT6(lR` zXVc^KN1+992up0>5-n9uJwCXAy&q+P4S{nL_lSIJ4q+mr85_Zgdg*B?aI8-0Y+kY| zI#mRzl>!GBO9?Tm!|e0Wq+Zsnu=tJpaXC3yH^F#m0McejL7_SMhv2<46JX&*NKjZe zE~if+kSDEQ$vV+Pct!kxyoFvHUD-t-T*KZSYko?uZt~lAF+efPNU}=BU!^N?dBl5| zR2()H<(oSDaAVlo2Bm3b5L|)*Wf;$axsbFfJwA6IxHjQTkWC0N{+s<_usF__4k8_4 z%QXv%WAWKE5gqObv0;`PR5%aTkcC6PrrHUu$U5BfNhnXXrui6PCwUMva9@0#$^$WC zJ)WU}68!M%-euD@CiHjB=J%h|3+mX#E~Al{Y@Ew)I-Qm)S5_-kHl*yyFd|vP5aAP} zGq7uesKqK2ssZJ{5=1ahRY~a6pZ2slWGfhFW>i(3TU!X6f8>ShZjUm#*y1JRxhRoF zE?AG?;s?81`di6N()`Sfa7+)EvuXu;RA`v~vbTI?VIiE3@9z3^e3{g$sF7_+% z0!P&SwMF}Nz0R^*rRrzdg4-4?KyRwse6$wE{U zcZk3#r4J>8Bv_rMQsfoM2A!u<*dwAFA^%`P3#hSdrLG&>CcXYh6Q4mHPsUW(H5ozw zHQ5!?G_OMQCYYc7(DvdE8O?Db)f5oz_yc7{wtge%5{g>2P37Cq3L)sB6L(YC= zYK<@_n5pq534Q!hT+TkpQcC^{hA!Cqct%Nd(8bT6K|GD>@<&VA?`sd-JoksIW#W3a(bQYvc7K*tj7t0G#}WVlp^1&OuW7 z<6Ef{527XOU-(LaOoVUI%`Wny-cF3E7HiE)#nAvSr}t45_umT~?mq6F{84%d2}~2z znJEH>7zp}PYWrDjy=WUde<^o`*=-RCOUXgJQ`ajt6 zzW~_)=`;yQVX?`CZ0B5#BGB$Dg-jlolPB#gL!CswT_B;<>=kX*hAFvnr<^7VVHO4s zaObbEfG!OI83<27t}(4NLYA^z(4DIRs{7xl|458|FmN)JXSS3JX zNBGKc;kYb&+A0k=)iBB%E~SmLN@-e2VI2!0abl}#Px3H(1&3PCVd79Oq9D7eAMz_O z!B_GL_k0S31XRzHRemL%VgQ%Puod#ekt$qbjUx-a4*rsy2C9O_)vp!r#T!UUW5U&lD5tZ04T*lZ zGLk}23a#R%lSsY<2gWym`VYl+c+9F zXJ;o!8)9LPr9WaJi>~cD6u$-Kpd_n{xm2Tx=TzdtRCvpavL2meQgMDhA#&K-DB`fo z!R(W`N%2~Zm^jL194|*r_>@>BF+ci(5eZP9pBIWUKnT%f@8c0!Q7-sqyvmc+iA&wc zRVA-nfgHA>fM9Lyr*9uxZ2TRXJLK_RA>0|9JAtECzUZjO1d!=!Q)|0rS}Qf-d7*xs z7|-edyLo-XwASF3C&`6YLutuc3pp2-QHl6iVud0%&X}HF&D?c{tD()Sc62 zidd{gb-98~0IfGk(*@iyZQMmnww>tuF>bn$wi{hC2pNwv{AhNmmC$l{LR{KR0bX}o z_|*Q8w|dw3EVL;Diriyizb$tyj>1gY5_$bzRTgN@DpiJ3T)aA_LgyUw{L1>>FX-5u zJ-8Ui++Q0#&aSeavN3#QP*Oq7cL|31qYG&(TpH+XM8CDQdrwy3fvN5L$ygYDn%=={ z^>PANlkZOr-C2t8SZXcC@exAoHt1vW)+Rp+NF0?gMO~2Qtq|Y17bctP69Q2=*|Qm2 z&QVP`2uBLpU*)w)q*u^?5ked#9qB$R=S8n9IKL2RG@MZVB0G_}7csf$m_Sgl^tF@l z^tXx6mYs+CDJrqqn8#H2YhqG`gweJVKwmf!lv;noqJd-EPb3QAwlaMrqdz&a^C9$G zWq7{m7^Xw1IhqXh!FVfTt|Lf2JR;@Hc-6c@^NTGhJuTOgFo_V$TZ}sXS(|wen_*J3 zZ+)8up#f^}8JAQRm*gD`8#UnyeTvKTh8OY{ai!E zXNP(9!rZRybG#US>@mToDBc4wu)UZUJXmLym@NczU72C!V;M&u>%dX&uV{5GQ(gYD z3@SN>$o#W*K{l*pa%Qn$xW}(dtaV1xK*LgIoQjzAo_Cq4ZD25G2=iJ)99QtOV+hNf z)L%qp?CbCJPOwcYnfO}D7*~S74fq6j_(Jt~ReCETzgwVfldP5SD?F8@>M2MU1^$XM zj{JbK9g4kRKPiqN%b@Aahlk-Gfr@iXk~jkor=PK)f){1?tH=faNg<@+(s~I5T!n>i zwi>5U&Pj-xrIe6fAL7p&VIrHVw_m0W)QKoJ(j}x~p`2sye=RpF{ENOiCq`Kf7ZL}M zyY?H~EK{Gn2oO-dE&{6v2UT#%Zr;UQuPVuV&S=Y9E^7ptr{eSUGE;_&J^wW!9kE8l zF;m9L0B=8U7n_ZH4vG3h*lu=jw?u)}E^Mq0G}w5e41{x)*>PRf;Il`>LM3Fwt%bwu zi>LhZ*YkN96DhXmnQ5d@yD&zk9WlHA>+F~uZ|m;ETmT(H<^{A$EQR+y!SDj7gyqXs zL*5+jmMmY!udeHZwpVaFgO@KYI{>_P3eJ$DSX!>pTF*a5a!ae#j{rQ0S_HyM?4nr5 zFJZ+<&IJ!o56U#l!$VTYPF93CtHr+%fuz8)u`_?yi;GuTIgFux4%u@=&vk&vIp9&asUP7qiH!hO_c%J&eijCNK zOTOW}`k5hgX6EO{!p&z2ZJ5day3#x2U$r!ulj2pW8@a`^`A$``ye(jVbTeP~67k=S zJT2|9eQ%aW!XB$kf%*c9q8n8-a>pAkDFg<~mMb2IeP#TiGYW>t&f^Adf@Tv(Ab*47 za6s|0%bF)!D5Ig@8WW_KwJo;o)n0ft)YIzZv`E_#tatVlRSP`Gtr_pKH@)vqwHY~I z))I~lbQ8Tm-mK$3&5gm4`C5-sjHg^=bY@9H2gEY1cx@VV!;^zM$47H)`h$N@0>=)) z&ixNd?U^~v*M1B{`c+r8{>kHON1wh#782_p-%3+5nWJzd#6;J(YlPq)DO~XsJ-MnY zmt*8T`Jet^V#vKLX)3Qrur5SgCI%%tS31tc8l{w5YzqT3id$q~H+x$(@ZYFx2&Wy} zNx1cx&x+pEoYEHUutPU+NoO$RnE%PzgkNI}YD`;?G`rLu_!`TtRi3j=*V>L^wyep&Xpns|dzt z$yeH+W(vZcoub z`-FWA8V(j%5frhkqG?kcD_v9T_YCh#opxP9-W_2;J-g|Ikv>|R`P@KC;Z1aaUR)WPT?ETV9-?slK!n4bM(D zX~4-$tc{d$74<34#Rb=Pm5}}GJ|k{xr#4HQp^3*mU!UgwSFm<2w5Kn`urgYC6;BwM zjNA1qB^>o9y~WmadzV~Kt?TGT+a4z-ARn!2lO6qQ#t%v*-P?t+$tkB6A?I-9=Y_` zbIA*zBpK-4T7(XjavLT@sCLfzYH zI&F&BnK@f@Xnoz?+&o{#4jkKgZ_3769w#>eYPviNpewKImptSwS)TQ&`EA##%qWqU z>|ci)rkmzCguRN0#v2v)RZm!IKApd$A^UqV5h44zy`I2gv#*Ii`yH||NjW@~uT)+z zb6`x>^#(XZhdX_6uJf_T|06SD>IezFnD+qS1_a`?sqY76=2Ou=?BOkRN|6HRz;p?6 z)y5KJs$SHRR5nXUP>u}wCyDK8t`31=B+g`6BUak)TpCnpSF2y+2@OhrN$wrk%bPN1;@bL{z2Rtp*>dZ9M&L>ppsJ!#;TYXW(?s{BHJ@42Mw*Sh zl6bqG&zIYc#lWj8PGCqyS@a*Q=*cGZbt2px6J11j(pu+3#0I&c@J0)?rGgX?>hxcZ zW+G2SxDs@OBdQ9~=>{ANFh*YC zAJrum$`&`{dapK*^Nr}U&)cs~Miiy+dDe8X<2w#smqf`vg0p#eEeueZ+uZ-4mXi4lM4mdvEjKC<4Mkc4z5?MVq#TI}8;ie4WRPGnB%vA5Sw%G+ zck0|y4Z0p*3r%Yy;j*-`PmjQ^uzcF(k4YE$X&g+>F!sjfKa)eTa|%v!z8T@wWA&4w z4F!!CV*^FTY4+kj^*_#O1RoBg$U&*yaf>S#vA&Vz13pQ>w}39{O?;^>J7Q-fHTL$T zWk4kl&v=4R=Ox7nI*0Kc;1>@I9jKMDBGFg)->5D;-?|Efy8MPOyMyDc-Q0Z|F~yPV#&~&!CAqza*N))&z@JtG!hFuEP$p&B{#kLfn!U$${)P z14~nH11gMh+<*oH?LuP@+n!B=soV5lii2aOxgcmxS+i+h6MjVxT}C0Fv33b98|j3? zUYY+wD1hb`9p&jdAW|zYpgAwmdxFWqbiZGnNhhgZbKauj<+r6pJqmJpRIG$x4!@GA&w!qF{ZJp#LN zb(xbJ*OnVb(FG&x09_g?!AK znkEuqy(TLJp>6z_?ZolgGUzy6K|IvP_-eUBqPRYHk85NcYZs-f-qV^fNIkn2B)Gms zBFN~(U7c~EeVobzk}4DH^Vr6=o>5=EY4^3Bweh)mXY&KJ&Be$>;u-iX-)LPj&B44R z&~rUR-WlsGlIavRuBIgwK%$;Sv~&+4sA1%UPJtT}VQOf3@v`V@*>&`Z&zb!$+Tj`? z=*xBqa_@i`u8cb|0!w2Ol~SI&;`HOD$m0cz`|839L*y=bs~7L|T`q!wQUrNxxpNoO z+vpK)soL818lFU$f9u}ESFNaI@x)v)M|EcH6(4h{8H&>E@^`wg~ z&pyJ&o#((+4&*Q`7tWVB50Xp0oJRa#ZTi=`$^RBKdVmJIsq|h^Q-=8*(|^Yx(}7?= zJfN}m`LH1tsUEE0WHIKxK(`4Pv3S}W%gODjdUGIk8P+%24M8#{JL~QzF1I4|gcNoc z^3A)^2%w_$0Tm@3-HSlL|i7-`b^|NZULia1$USZk6hH zcPhWYe=RvPy9j0p(j~rE=<7<$7y|K-nTOy{B*_;S#LBT}bq$himFRzbGF#>rK3+q= zIF}E~a?H$M)t2~xxO$(uT`sR9SOzhreSu)zmt7NkV)}2vI1kN|>z_X9Xgp4Ud$24z zyoGyE!K>qQfc;!nkJ-|Q!p~~8taeF=zl$t=(y5t zwSQEP0#CH#2$nXJf6dMtN&N8_cc*ARY=rx=&S5#ee)#i%GOy;7gWV*Ky?g0xXn8-Z zS~JlgDIG5q&cm ziNS-4s~wLkDG1$L=Uo|Joh`?0Di{wQosn)WM&H|)IM>k_i`V3=6)c_xQHWUv&3!&8 zbY7d-GtN3Y9huj*j$ugN%T_;PazlHWS2vk{8o#K0lJ`|y>OFYB#5^bIr^#2Ukw~ut zP-1;hiMz9jQo!MP{`^(eqtXN6B{Kri$t2$_fU#g8cAK!frrc)JC;oeq$queqkoNU4 z7uQ<~LBSnZT*?5NuG-Y943m&Xm>O7G31zYQ6-{UmH*g~Td)J*9!&VqkyBMZyCdy+w zldB8+VIgDVI&9ebJaBNGuaT+Q?D=gWy0tz+>Emi}{bn|V6^WPy%F_WNoBdFBG3ejR z|4P!#SmeaNpBy7$n+$r5X)|%TTO6Selcq!=w4U@%LDmy6HU!_Q`Tzh+Xe?cpynhHg zqtPn)&#%)l2a*|{V5Ys3*BScie4c%2n=#)RK}F;+E1yq~l!8EkyjKhtz#L89YGClw zr0g{V|M0QCob2C7bkvZXE~{9b*l&$Cu^Z*Ntsa3kPm_`KYc0MN^v1WCqw^pU;%;Dx z)=yAp)_EB*>B{;QB8>YD$(dl<(EV1P&lGI391^TXx!Gf4P|!Q4IvmNzITLX$MdI%^ zkKsrY-=o~PcbO|1_A?)Oq}Pg)5TM_%Q^(@h_?gi&tg$VRF0NmP#&<9n4%*KBmRo~zS>Tpnlp`nrcabhn{$ zxalxt@1wtkX!M#p?Ovb-3Ig}?u*7}0QrA?NLN-%h!Q$H<&kx>yRd+Mi4aa!=dUn(| zKAi`2gOk!s>Skk~pRjX-kMNYenvf+ko)9V)TU zeh+y6_3@Dy?ZlV;73SW``_*LTM_vSUZEv>mjQxwKw{;G^Nh+-CPJR07k<3*i@1zVn zGWRIArpOh)qJ8FOKZd0zlyLCYO z5ayh<@9Q_uOknY3baVyA!T0QGHTqQFiz&goPHt8XFkMJ>D)PjaXmvSywXGW+ZdQKI zl%(JYSs-(2`HRcqnQ!y+4INWT2{JsdpzZsoXOyYG&r36H>HDQ zf84OJw#5YB0O+1Iqh7g{ADsihO@-VZJC2sRveVusyfm)Su|+J&9g<95d*e9G$1GQF`a^$yxdNPCoZQwCFGyVLb}|IWMALIJPUHSb@>kDsiD!O8xnX+ndK7P(7xRHw}EsIen>9Sw(TB=d;fJW)M2W@FMB5HK((jyFS6@PfsDH@agW{CO?N-HUU` zG?TAzmFg9}yY+(4sq)Naj8`hz?5wYSx0YAx8qB2L?~VIe=sbrBeIU}7|M zYt~D_Cbr>_ehWD z(o1X;488WBbCl39uMU9oISaM#CQq{7hr$Mh(RFENiCZ3lk93zG>y&2mLX#|`SJAF` z-F5L93O0t?Hi;ijl&kI62MnEVucUy?j4P0>ZU=~Qv@)bw!!!|1W9C_u6BCbt`E%B- zS2bNaaTS54{XJhPj+S*IU$;etW#Y3;6CTs3gb(WyW$_J&?>X*dGixx`vhT=w8tD8^ z`29dI*n62$NUz|w(mPHEP!W_=ALU>=v)!ODS#$0#`kgmP-9wXgj#8;xAq0qi)D53r z9_=7AqhI}B7rEH0P4u~H_5u4KAU(_;kg0{l^*eC~HyDG5F8z1n^#-LUM|K6%=oPyj zO|(N{HpnJ%hbyCWh=mOqP_ZRz;Kni{qJzsa@*L^XOV;OV;az0W5F%6O!8X#}(CzCl z@ou}e^}#s8VR~orr~AMRa`stjW?PR_#I{M|9mpk8O#>Z0dKlTxZv1DgV>nDA1lTcP zJ2@Fx`|W6n=p#tfbg`vDPL}qx3Bs zRdfj3LPkq9{X%fJYDIs+5Df7hi;|rp@*8aMnY~t)(fk0UU*4c?m69AyywU(s{e2bxDB{An__Nr?+R-iWXu-u`>42-RTd9k_?z<$;-Smey& zT=g>eiHk!Mx8$S0s#3Q4{qvs8>SMU1sTm6qkrS^qf8@Yjy*O}{!{ z(At~w>n@L6tU)gy{yMKw<8KvJm8!A6hD(D#4t;f&RB6unYAwsroImedpArC;<%KUf z%Z&?3wEy^X-!r_q2c{<^q_y2rw`u1qopR zWw+IfEHiENrKO=IQ8W}OT1rEi{s4SZbw=ul2y9&H{N>tjTsVvnqVfs?162y=h8&Cf zeE|^)wNi#X5r|)D)ZKO8o{m~gGe$;lbQrbns3;5Q)@5|}zK~o_R~YMs{cL^%82O26 z2HTBsd6YM1KDDq>zNeIObPDAIkM^IieRgA54d7HK^gSmtndwz+k845f8({GL3bgZ+ z`Fv!4a;~a=%jNFOtL(P)PmG>K10Yn8YCk|^TS*2623Uco_zfSGNN3=xd>SpVCO-|bNj1fUXi$s>+ zf-ae0h}$_y20iSh z;w?~8B8g_o97lngM4`>ZI!o(GH@;ltHfh~LX}YLUU78NpzMaB(E5I9`wY>^&nHSJ} z%=wV3_LdBN%o1au6=zGhEc|jBbMb zniZc%a&r?^)^@$he%cqu219aW^i2>`p{&_vgVGxW)lR4gi;`d6r-$h4?QpZDv|hA+ zPdne)Y&MfdYSxv`xzqL7dGn|Xw>H1n3gelYz`A4~(<(RCweVVWL5QKnz%D_##28@| z9s5i`xAth>Z4YwUUUyNwOyOc3U`pu9YH>B>JixTQSZ%l5vEkBo^~gJMFB+1&dpQUV zHzVNlLBD;_INS)jUQ8uuFYAdD4_21HpV~1FJ_RDfoxscPR`h9VX;&kV;$N2 zebuxC?=|zgV^L3<>#;*0rA1itcC5TjlquNkCiWO~sEzr1tLq&ZQpXlo59*=HXlxdQxjM#CV}=@0!LR<;?N=zEuEg|9oydo7QPJ4As#a(n6w7O<#TdiJu(?^|@^s`yeI2it*eXlYMBu(?^wR~&v^L;y8 zs2J%h?Aqh+1JLSg<5h&JbtSIWqcm0ApFa+go(!ZPP`2W3wtcu1+>KA{c|N^SbDhh3 zS|V-P&XpcB989eCY_IPe?Y{F>CFgo{x&2dGKa^*Zy^{+V&g*Mp9WaTWG=NPTB#ly? zwi~_gnbB^%fTax=8i=``s*#Cf^g>+ZoR-d7lDw6W#?aNY`ibC4ZOWrmb0m+?~5p8zTy)m$}BYU)60 zFY5;%97?V{G>=v_l7)(;vbPO-+@()Nwk4j zzY#Mzk^XlD8VGCL-_b0 zb0_=_ZJp$}1;izsA1YF1MX!!KQt!zARf<{c2;m`oGp!sXSYnsfR%>_O^}4Tt51lLR zY711aeyh|N$yu;d6AyBsDw|4$!!ZY3A(wUt;RtPLPDi^*{nPymd?Hae+@msuiXa+- zp_X%@f9b%~FvLc3B{3hTQn6IoS!@AU`7v!2Oq(|KXdUEW2H1)noZ>cc$nUlY=z1-n zGBS&$V|7_gZ0sgTSL{|Zo*Ll^-2STe(gsz#EG2$6R{5s&Y=rA0u)~zfk43*cyV!c} z)Uo4OyIBaCf~3DTHd1asWpv)mtuj}03}*o8^C+r}{XK9bCBFN)@zK)u{DJs0%Q>)A z5a{PvzL597Ea`M#Qh!%mT`w|dnp!aT$9nJ{Zg)+S^XR@inSLy=sj|!!#H1!T1EI9T z4sm_?zNs&dJbo^B3uQ`|^wafmEQ`&DZ!|m7XgXjvya5b1m^Mn(gzxrf%r@gTJ=grPI#&CEtUqljwW2%f*IOZrV?92b<@7xi)#g*=Cj2Me1!wW@i3?2}sv9j;g zM>TxXJc;&;?Q}o9Urr{5Lu3A~i_tlyX)?H=ueC*FTF@$T+4~hTvFTIGC?_?;96~;C zFHkMiOB60lr^t>T8GfL6QWpuP&rY;O+eyQAsUGYkGv0sN1+i8b6&jBy0;GH*HN?JP zS8(VNHq%EW={>2Q=o*N?HGRN3%fpK=JI0%(WYR$KK&Bz70IK&5r`39ys6#8TS7^Yo z)^x@9I8oyhc`G|qHmQ>&Rm7ZwyYYvAbP$6SjC`No?WFBe2vjT-OKE18l6>i#06ZUj zw=`%NZE8A}S6b46dwS({beW)RHQ=nj6ttIBr=F+p=Y}d)DLYZkE0a&679>Yui{fs< zNT9ZT*@9ruSq=RMXd0wksQ7o8P# zhp2VjG(!S{<4Xg47D$%${#-5b)4Xg6s4L)R@F8i*`p&`*;`>eN6b)qGoR?Hbj;SkIxB&1Hv5rNg)!6o<)1d_hlmsp*+ zg-Rh^H;UtbN4!%yuwYEPX;B)@NaS-$>q`CiJ`;az^1WxEk=D{>IYi1(Qb`-YGJ1x~ z(Keu3M)RQBm%lU#%M!qUV$9)%pD*(sCij6hRVr?|3!86{k-s$CvBVc2TSjn<5$i-{ zp&~nKeF*QV1_I#h$S~cAabuUukQ+j#GI}`PY`Rd|e|hfnVF)3@ZA1mjxD?@iLR1p9 zIV0Pl<^hf^b4`|T&Sh5V^vL^)|LaNdkQ$imR99j!HE=UMP2$sBout~eE+bpyh;R( zZ|~j4!^N14<@^S2?O=IwaNlQr7HnL}W_>tqZZLD^b|sdK;Guo(9?!k#eseJMsQy30 zCXZeQ&gQE1`du(WsrSl%)@=2{M{4s>eN|Lcl*MO^<+&ZA)Oq-}z*F=Ws#>|J((8V?Jmcx3-KJwoimM&2OL% z4|;qmbD%eDKbbJW-(Iu83Jl)T^ObAZ=Q31R7{~OCL(JU~J8GJ=zP=zjBc=(huu9v} z>~<#ev;o*mZs1fKrqod(-Yk*a2HGrcV#^SG786cW{SEN~2RMjhHzU-y-WgEY&B;Ez zJjV_LW!<$ysmdcCuWNQoI};whnyt6RD$Y8SrOJ|K`t!wXnNy1yS61#H_@tP&oe6LI z*V~wS+&Tr600yn9nhWD~!~ow_b5YV}nDio-xiVcqhr2PH`P4 zSRhDB`Ue4KKUOGcdF+H#L<iJFq>Ni)Is51)Jq;%T@g=a36J|n=K+oVp(NrLvN~v4u`W< z7&&T;<@!ayF3SWufJ^^HSR^-!v#0J%bF%GJ$A+@B%GCnCO1#onsmzaFoGj*51duXY zo$>VE_R-5_BYRI44U~ z9yE9xDUXybrD54gYYsHr|E@pW&~mcA=2{BB)F7ZAyRCR>kW{_L=`#@AlE4Z`FM8o^ z|I-HTscyMk`BZcfk69x+I67|my||6=OLw49g1YLwU_&iT2tc{bCA}^ER(R2ju^i=J|Idbk@;d#Yot(KvVP& zMHpPRQjuY|$5tDyxdfY_lgt=+9C8KFLJNH$YEL(DH zxcVkaXLdvOaI*Cug+}=i3@miFNMCKIr#2~|Y_RVhv*=u3H!3N(MOSgoJN@QhKI z0$@n^{V`8_P}m~@)pPMCm?sNEc zI(;L>6QNdQDxj@ov$mH)?diap3-ZNF>#cxLWN9!3Dl^iapsh8&)njsH^08Cb`g59~ zrT$a{MuKH9-o;~2^L3ntA?h@s>s9=5mrsS@X0};0es8h360wApm3z9vi(`tB&%)<@ z?m|vCr@gz`*B7XR8p{^+7&2fP$%4sYnv;b>NT9M+J~{hK>Yiuu1G?ASzQ<%D8M+>v zOpvu6>Vfu>?{OnD?S+6Q4|W(;|4gqb^df76!FBNoM4555RWJXznJrSw2oKTFs&-pL z=(4*;;vRWOr)l7S@+NdEdM0&sFkc%d_Af+jN8Bz9sCgQ-Tg2M#byZH=@q?Q0i;SjP zYiH-07F2$zv^-l!P6jMEjNe8H{U!n?%DUO^H&z72eE3HN`tJ09_C|8Zyfpm z=2X1F)#q?iwv6f_`SLxlsg2`UJ@wo`M8+`S^+Dm*@c-~=bECMcW&FZds15`EqUZK! zVY-+2oAuk`w@uudqhXJ|Oc2OU1Qdgl3^Z8-UiejoflSCNR`~1H>lwJu94Z9Bp#~=} zwi05=d}H#G%rjFPH$o(|o(AL~OS_eiU7MN_V`pXsZ|A_?5sJ1ywQE03 z?+Ya%yA1h65n|iY4dEC*LhX0ad(VEWp5ZNLuVU)QXfwphXmVgKO2(&zn4GPAp~-(B1y(^{T^P|Au)Fia}-Er`_mLR)Lp?rboMntw&7QB z2V=1e-QchH2c-il>ag(75G#3AY&Fq$TB*`zJd0NkwF(7SfxQ$Srq;WEg{uT19ln_o z^CMrwtF9?KAGyX#mS*@bS zy5t#iHc@o#h1WCLm*ZA>!K3)u<}|p~9f$kgT&&I6x_%fmM0QIlJZQF|G=BR(cHROg z?(SOgGwQRc!SPW{S7Q?|Agl*8 zBUHFle?vSOGugdj>F(Ejrb_r5D4Cl+_e*pJElMg1Z3;Lzix{O-VPFe^v%K5UO_^1; z-WzjzCTq%9aK#EbI6hxaWryltb1@bMRu?{5&NVS3DL-D+cV<~?N`*?dw$3X&Yku!#-bP6p!S9m7-V;O|^gdJ)ZR@@BCuA8@)LWPhr zis0t__eL}VH!J_R&qOo%%tWv`62uD<@ zDEk>I;5ZS2=r?rF2Th9B?t0>?!SAzw&~zZ3Zl!i%ICFY!li<_&^aWGA`)GQJSFw?1 zZO486_8H`M?)#e=$=ZR&U^C7mp{jOv1I`g!)4L2;_2o!NXp_LboZpNWvcXHRQKeD9pfaCua4DN0iURQm=3 zn5H^WC?_)%uRywQ3dDENUtxLA)^6Ml1_n`F{E=9Ol+-6HW#Kb?a183-j5L~?r{iyS z${k;3Va-U7EVHZChlG{woRp3E4>oP!|Kx6a7Iu4314 z=ZJxS`!ytxxu^~2<7x)K9l}nsM8-<;dd(Z7nz+nFaD*9Uz<#2F3t~wTrIz^BVuX)? z|J}{y1tRD<_Z-9Q6124X=yM|>6(h};#J0F1n4|c|KZ2dX3zV^*W2Mwp$DzFJ1jU5d z#_w|P2QoPJ(Fn__tyCHyp1u=d3$Y~*q$C&4MI-Mx)}VlZ~9R>0zBUGd~N?TG}fP?;iR8% zLrORMh%zV7WixX3>TPnegGyV;^qv}vU0C)U&f#!N6xz3YQB&H*J;x>Eg%U*R?1Vv> z3Y(_6RX(AD$GTpyyP5z2Z6SDvg_<}x8Oar;Z&>olNbCLS9cn}@15rTWN$B^k*(q!W z_~;t|)cV%%tf{H#hunvlx3feZl%Iqst%q)0a%fFjB#U@Wsk&Kyz9~FfH)Ump5l9>S z!vmIP5yq;96Yb9msn#hk$qi(B)v6nVv_u0NnD+N-8>oaRZ;h^WpdiVj6EQY@9e5OU zx>7PQcM~qnpm5SP_64YNMr1t7yo%9KH&3fsO2C`y%=V%~-4eK+g;n3elOu|E|18+JA zE^r}eDjWkZuUz=VXn1X5&coPxtvyzNLQPbJa0F8+#dM@w`2m9H$4ZU~zN1f_fwZ)7ho?d~s zJP;%A=`6Qn&HD<}cFIzwG|mquK(h6D;=%FM1xvT!g^m<&Gda9Y#hFhQQX~xGq37Lw zZKxbLX%j$7zzsEPQUwwsA$CYL1*IHzCqe1z*Eg}&&vwrbG8p`4abL*OS$IZOxxGJe zICKzQ9C^VExh^UxT{8dlZPQc|B|kW@@*5&{)s3h1hAw>KCh{;j^NFy1?)2MAFpzDc ziDy~_Wk38@3Dcv;FLjB*-}3I0O|>RMOhlE&FU!pHP#u)>sbS6i4_=Zv;wV})CtqAB zzrAX3!AQ49aHX>V;-j%rpxBB&c!xF@uv4M(pH){(xT)j8*I}wp!~d{I-^720TdUjJ zor8mqb%=`_@a;5mZ=pyt>_FEwn&lf9xb1p=y0DY}O%uV}LlAOQ^aM617%2qAf`yU% zNUqg3C|P{L5va;+V2U+_>R})L>zx_S-IUh2ZihQnl%R(_k>EZwpyf~^ZRcLaojg@} zVyloEm79&zj}mZOI3L&p_Pw@@-smd~$IE)cj{al8voKRzlFbi%P}pFYz-vD;?_NlY z<)q=EpIb!P$TNaG{e*?A&|{EHa=5FCK)bJ|+p8xzRs;IM-(LDUpo_J#1vjX7@PQCt z5Z_Jgjah;g5U~;4*!_WM-Q5IIev&Y6<28UK(hEz?-b18>6@mlBD zWH9!qNF}p$!QE}CuZ4%Rp#!+f5=!C3@gN3z-z@yam>&S=55stbHU?4FocJur{Yt$x zx~ls!@57N{f5}X}_ck0_hi)u4@lIwM7uqdorrpVaWzNqY*(G#-{nc9vUwcb);O-*v z4dJFCn&`UX(O1=?<^~uGy>6o-kBlot7L!<*7R48Qr)y)`(5O#x)|%a&n5l;0F&ih3 z<{ig(TD1XAouc`yk9a4nA&9zCgJf!0Is82aJTT+)O$>YE)qq%lQ`HSSI(MMUv+hiK z5RekPB!YgkEa6}=&K^j;2c+yi`XGi~hkOKMmrRhgyH(OhZajQtBQx1V+ab9x2tT-&kDz@2%Gw==iqYSRAA(xMz;g%!+>2r;zsfp$n5Mp|OHe6PIohcpO|KgQCr;mfy5!L3_>u$8K@IG2Tw0wTzJ|}(_anRXc{d$z_oPUv&h4!x zQMon>*`SBuog?kp+$*7AD6LsC332$!^Y^sB&QG zkVZ4;GH)}rKp$MHjiO~oO*5qgaac)$87HhM55z4*M`zk3d`>p8)f@M*tp^FbBJ|N{ z``6T^`j*;Y)93KpQne*tN+&ocRhSrFuu=ojUzgG*ULy!tYlsB)5OtJXED{T*?h3y! zr0nfn&pWC2uCB80(eyN=R$oXmFouX2-?OFrh2LUUJVza|(C98(O&&ah*+cxg8JqrB zm*ZLsS{=1}K=C^{%82>k_)z>-1f-=Q(@2rS^f%-Z>d;vsWK>LEia~j`U7L zONe835sZ0yGq{r#%~DzrLx}v9TI`&J zSo@jV%{EEzk_-eRZ=decCpiiY_`qGyMkmu7?bz^@*qWZ*V%1lMIW7=~T|OM$d&aJB z_&Uxt!ByJ+$IhDtv)0a%(C~Zccx)1!oK$UOb&3L#@uir#RK= z;N8bmfamw@BWwrJ?i!t&SsU%=QF=_DOV%|~gSd-EF_lV2o>j8gJS)mHgT|j>SNLgE zzB|nFefPot9dE)eF*^9|$+L{H+X<-qlZYAH$ETS#K}yikTcI}xf6m5|+xEL1Osc`5 zITg0y<9?gY6O#{x{xJ1TH2a*~l-|^RMU+0N)~PCWs3rdK_G^@G?rQR!ZOna&}&t2mbg>@#p~Inyk49kWvd z|7>HoaGu58GRlO;SHN+<(+Mh8Cfk;rO^;VtUMOpxOg|lL`xtpSLw#GK(G+#di(Xb| z>8UuT*Tc8@n(10&KegIl-KJ^!WoDXW$t?#S{(6TlFqS9P_AmT{l2RBeXqUb@|edUdEaumkvg@+ zxk66m<-G%m8Jn_4DG5z)#PsL7IqfT}=)orv0A|<-hdOCgEFo47I20}M9~lou4-vfX zpt)eY$vVhwcX#!VeR^ZbQ`NxvTM=fK^{3OyG?Rc!clPXe<<8`X(vvJS`wjniaX?*V zk@s8_)=RoRDE=9@{^$*Woli3rj>GjJot8@$JK>sz$?Bfi??XkQxoFKl^fqp5PN#2B z=d5>s9bc2MJ-7G>VO{qY*HsR90^)5AMC?gvzJ3wpG-OaV0_WD^@cRZjArUKquQ|J4XqIz-=nThs;1ke7bLi7L9xUh0WdA=v>N~!p~SnygTkuSnkx$C$!V^Nyap%q z)T^N04(vNfXE&X>t{!4Wdxe9mPc(kzX@imfmdb*Lr^Z?hai$^btHK%1ALa*2sm4S) zcL+Nf;2|I{56nYL{Mn9bM+UYbtGhio>v(k6L~X{>z66*Hml-xdM6hx@kml>%2Y#>j zzenT-Nd?+Mc?`h6&KF0RagADoED|U6cNip(R<(nbpr8yh4@Evd2PbDIlh{t$bYJ7) zSAyq`7~c3g<267RU?Ir8WwlB+SquED!O&1@IJ(J4$Ft?kZ*Z?bAoJ{OcCKhvho^e) zlPU!@^#-m#_<+56KG9FmS3k5DjiBE>d_V5L&PvcuXf7QV8-)fO)RngZ4klLTLatCb z6wlka+fkQ!PJeBcwPssvtK`xof6fV^o+(??X*Y2=rRhm!wDuxz!jW6f-P>1X&#A0a z5cg;Z1(Ah-vLs+y*CS7XO>!vDzuk#?&?y=%aor?A|D!(4rJ z@!Lm7`YP6-36b_we^{vtk#N>qJn$9QY^>SEn$`5)Hz?x~OaTKEzf->*AB3&{ZPpOa z%2LQN$t&PpbX)3Sij}UCT@0JOr5|LK#@M>RAsCdU2cwse$fYUac^A+)BFOk&dn!LH zlxqayN~s;P{Vh&ExhhQC`MAs+wSD52mjFF%c{s88#f>$#ZeT@s{#Azghv$^26M=8S znbyLjb>YoW$e{9{(pNj8G`Z)%(4;S(Zk?caV`fg;Dh;*E<>o(wzi9H+abBS>88?O- zk{0-wR)cgN_Q&sD^_v77q~7S-r3E1|L~DDE4L)+$VP zbK|Lvr&WhdN}XSBh6FoqzFm4khxI;hf8I7lfv{G3i?{f{s^e<~wBCl$E?mdC@3Ut} zBUe_oDX9WVr|WTYJ07a9&9XJDmZyNs`QPHHY(lYcoZ4YCk{c*`6^BrclwlXqkV+HuHMt|`e+pb0y{!EZp#+0t@ z^OMbgW)_RPNt6Jeet|+^`YiHyH<6)eovW#!HhRu5UU4C^7T0ib&a=(ut^E=!{8DAE z=YmaQsqJqQdqVX-6om(hY*&^1dK+6I8m`kVEfwbMD?| z{YuJjA!qw97~g)?FGX5fX!`l9IJ~Qz{Ng=%nbkTVwT&< z1LJV_gt741~hnlvE|UIQ{Kxqh{UbniPc*Z zq!!BZftYN8XBB-_8veC;uV`~R_kj93!MTP5aSP(%zm68MVRTv%hMGz*_iLa%kw_7r z+xO+`&;nMNc~2q)uvn44OZq$;M2gr=?dD3-N97ipg8LoODxa8PuPZH9HWB&$*C3@l zcF2%b+5~(3q@;w9v8;Pi*TI_(7u>tRufiBDgVBMS{k%UrxNo>*3~i>Q@UCOLZARI2df>?X0%?-jZj!r$B`0kljr zW#dQ-6#>TJ(x?tF7BY#G#W5a*2VhQ%j>YCl4~3idIX7oHp>`0(I^Z%0nd59iP@Agq z%3wi{6x^ykv)(lFLn5{Z$|hU*^G%t?O8xcz0%;{ zPVi$i;Pjr^NN4$tM!ODAr5$Lui)_i;yktx$qrzKx9&b+TAjB=M@2wTs#bad8d@jn_ zd$69Hqub`Im1KpEjX-;nK|6ch&A~jetd=;&ieVJY!*;tq8z+WS26o4}?bzN+F{YBy@ujPVoN^v32WvWA}aVRElJqc-tn zs-btwnV@a@JYmbBhdqii->aI^8^v?(hwSY-l!bgZWa9#8xGm|0-Lk+2#7r)Kuf)H3 zFSg3Qq6drQ7e~7Vf>MGH8$|%c&yZC-KZ%?MStV)VkIfphM4OU>b{|<*U(tOH;c~F2 zsDjKJJf$+PFGa0b7VfSlQTgd_Z#hLYhH|gEzC`Tw6}A|Py5;CDvX{f{idER8QZf4CF`v0nED2mN^I#xHy<_Sz3Tf&TqA z=j>T}5S*+QyvR^FA}%(Z%Bm6S63)uo2H&aWhWP1W7Hh5so^`EYa@z&s*-&;ul*2!n z&nhQO33{+qEs@R0~@0ss;-e z`g|}C)4!Au*N6gPd=)eLHv;6ao9sfbtN%6x0dWDfpE(||^7ENGz@JG8Pf0&$;Bb1+ zyK<@4Phb|y9m~#8@`^FSh`aq`RXQQ|;1${7)E@$e)5jMLDD0e|Jv+!L*L0sv@*mS{ z=IWQ2yD@ggWVk8|A7rA?|1QD@E=5G-FLq+t_ahO))qqPKRu1w8=YIt#saCkyX8+Jm^~`(iSnzh<<^wnxbVa`x0ovg5qB9YrQ$?^2#pnc zY$JSVi$B~r^|THR6w1>1$wfJkvLkAA0O`5EIoWcb3g9vM1-lxI=YHUFA^QF4Q#my1m{7~i~1>Nat$jibr!uDx{y+5Ej6kq&8G~%?R!EexS=8XuV z^}7{a35=RQZgqlCjSA0}U5utxP0B~kKS_CUI9CVL9ni>XG(7r`$t5KVXDj4DW{zAr z%d?lH60MKf5gS&47PIOV*l>Pek$DkC(TLLVrC)PL>Cs@8a$_oXHfy)0MoO*VFIPP^ zyf_wRuLUp9k-Djq9OH6~At?OV7q%jP=^u3l^Cl=oR`6iL>lfvE;YkxEIqaNFnl7(u zT~!3j(jy8tv~hDek3MK2=H8FpeK;$0PYEBt#gWG`esnWmRT2s|pkA(>ckYYsY55pq zHS0h4lI^b_9kgz;V4LRN;;+QuKSHVvZ|g^N$yvACB-v88udZmrE>%7aLw>24;X=r6 zyVdxy0?S5Ak|AtSWMJH@j`GkAK}L+AT2L2*DLzdPo2~w}k%?1UHSVzHRGF%fMAtOk=Z^L0+K zQ;>f9#D!rAKoF%19FxQjDS%$joS&T)GYi$RggMMkoD;NIP}y*zQ!bZv|4C!mvZ>b8 zXcRL?UDI~dm&S*RroN^{B0nAEBmo-@SBC(f?xdC)$W(s~sNMti*0yGJw;UOE_18*q zkQw?_D!_bYwyIyYKKsucXfSam zws;j^gnx-vT4Z;o*HMY|kT<$Qr->lSnB0^Wl#4B%+ntQozNgWqX(ypRA*RW0R!t}B(5u~0cML>EG1 zOIAB%dn(zPJ|k73Rq;&hl3GG%qMUKJ_Qur(aWLiWD_h`7Q|1C38m^JiMJdDWW6nS8 zHB$qfEF*1eE%n_-53+TQ=C|w6T?64Wl38b3gsd@7)8vZ3b5=nJq)l$BFa5W0w(hp; z`Iv2aE?O}`Bbh5z!w8+Vny59!+z<1tagc(-(&uKHDR~=(mN+}a_jt7i%kNiU>WwDr ztBCead9j&P`BFa%U4XZV;iu`0GupCij$?~@bN&K`wUDgplS0*0IsNvUUO$R_`QkB% za(w_CBH=HbOetlrP`{bU>Kfakf6}R@EBLQmUr%1*9xZ4v>Dw0if_<(}rt(viVpnv| z3eJY1ww06q(r=TFB@@ae_K&-U3BqGFwnfnYstev(rl8i0{&aGf&COVYlek0c%+wIi z{I)f3#Conkp$5XT?jf%YZq@uO%;difw{cixE^|}`2C^*=|D1YZf8IEANi~yD@fi;n zMg@(35+=e;SYNaVWUqD47R+tgs9bLy3AJ3Jwj`_3Xrh8k=`Ww`sj(U9 zIy@gSpMUju+L!vi*dJfH%29%PMD7lmOv(yIa~udKJdHks|wv#R5$<5 z)Njv@H@CZqZc3+gs!SIyPl-dQjISxS7O~%BY9!cu4yQK_pP?Sm&z&pWGuvEW*x*gy zIft7I-cG^v>gf9f|9s`HlChNiJ%N!P3 z!6A&I>sl04(^8v*FKXEB%q$_Fl_QK|>vz9mk&+5YxeE7JhY9%?^YJXt14vNoTFKzs z`m2$iQs9=@U^%Aj6;YC8D+8>ih-T|~It$U_Up6w|H&G@zeqN9?hP2jB;8o35Az=x^ z6g$`wLX^ogaFv?Dp#-#N|FK@^e6<&Sr!qwJpaOc@`_Tp4xxi8vAEyR(;IY0bV~fnA zkT*?J`An}pqr!>8tE9m#sINh~Vo74O6xD?&?=1!ZL$k1b-`vA2w-DR!0{CR$gJ>OF zZSjS>=mp~6cb{jQcu+1|gHg62k5SUAJr%`jq^@2ZiT9FzS>BhCw}IxN7$H05WwQE( z>TTxXRmoH9lHhi0`TCkLt%Mtq@VE1XcU;}YJC@}Px{te7-&!H@?6c~#8AunEOS0Q>73`V zXwWP5kbCsTjI^>!R$m_+fropcyeGaKKI-Y;CqQK{q1iCx?0*x#wP<8Ul6pn&eRQ?^<7dHtVn*YJ9VS=K=j%1c#Q?*6JYAhG>c;a?0p?PA8=+zQk-1 zL{W*AC%Pf8jb`%vl9ple3#UbVI%fA)r%d7pc>`l8q}3vPMc;qtEJ-*^mIS1ySS= z1j3{-Oi`$78t>S&ne<#Wr6+mw3!bv9RfQeX(|&j}bKAmm*SzM7eh2-2L>pfPj?pP1 z<)4r@Z~A^9+x~2LP1kJ4sLY^mq;4EN^gN|H(g$qHZz?!MH?SAaoXD_QfX7hMY7e|b z#HDKu=>77bjY%TRC9ydfoWE+L1j5W~hS+I-UAni+e)YQ;JhVPP_9_c9wW;Wh%zqE? zJ0-0D{3HP>=7PCVE?^Z`TwJ5$9Ak+aL_~@;<<=D2K&fsR&BOGnaaIab$=6R@>jcVQ z0q`$s{v0vIZ+y{_!ht}o-+qGaPoJO$wi`$msy``?Fyn0J*LA{aAw32r4o zi6rL*`+UE*a(nX3U-SvvbDl@j+SLjO+d(lEh9#n_dLEQ^x>0u-zj`? ziVMSJeTLryXz`URI`_j&fNY#{3HBc$fD5h-TwLXvfFQfVOzm_XKFG*MlPGpK%7d0s z`IdT?A)GAa|J$z>*4N+VH;o?;3HacvU4l+k)$}H1xrAzOVy7 z6k~&#uGz~nF2fnq8MNqanMhLYu_08d$`z`JH{TkVblAf)BFo8;#{dogAj_M)mUm!*Rb^e;8md6fp zdqipIHRf|2rN8gi^ek1d%Wsgr>j(O9Wx*I^M4^)s=KC&aiCbNuqe&AekEC!V1qj~+ z^gPHrl=kU|RI7FDoM#Jiv}XjVoaa&dD^pFR@3E|&U%5|9fePT{3p?&RGjt3w1;_JX ze!VUw!p*_UKFwJo+1RHC1h{d`Ujr)EwZ!=@gsKkV6`{i0({=@Oeca%#!a$*;eP9r) zn>0*bVTv-2HfhmcXCfonkt$2^S{ZaNm^Ug^`if1ezWq?_YCv8Es&OfR)9)w6-?mbN z$-p|8ne36U4u*N(lH^IjEnfp%o*;%VaJ}Wqny3oflcHy`cE_5Zv&_&;iS4KTnctsE z@0ZLG0`WN9+}PHPUZ00;67&z#1hG~%z=r1b)IYDbeks3T2&pdw&%|T7<;g1@^3AR4 z%{b3s32&<+UWxG^Hu%6K6Z;6k7)*tm{mC^7gbghhPecCM2fM#k;%BEVf$Wj4c*|Nc ze%0_mdW<(@*Q;$ug=e5Bt8i1N>!Th=8COxi&YJQ}t-Dl%|Nh>`i30)XR?EX?_XVTC z`??2Hpwvw??YqoeGF2a5#1=W0o@VMFv>Dx>@4c6KPo=mksDJ8GU6a?k0LM*}eAL*N z$f6YZH{j(b9j?L51gX5Z%kLLAxPpufn%39<&5~z(BcEyP`NeSz`nc>yq$T=OV81s5OHA;54{t{5JF3s<3 z3;!%0%Mg=c_6Ycshihp7YvXlJV@sc`mWP2m=5(XcI6#7#C@%yN<~op$HZ1M?l@xPf zP=*AMdoax4?Hvq!`jfjN2lLnnc;r#h9gxL5L{rK}%OV&g`HN`g(mwF#uQ75bhJEDv zUFdSLxZ>>_yT2ptZLN)(dj6r2C<*W`^?aDHDkK7!m)l7VW!@pmB z@YZ`@XWU@|d?&=(CG0wnBAx;@Hjz@NBgSoeVGCS49iW_?$O&rk!#p)_O2&4O^wM&# ze0^tR(SM6e&IWEZj8=Z`8~a8mN2rR?L;LnmaI*QRCN>VE!{3$)#p)qf`3+X;?Cszv zPt8)ANefBvlcmpr}j>VT&p5kAMFB3JO^ z^c<{qMTgY&u<#!4^v$^!sS)K?;tSk0^uJy7*Dblb48>9L-Dag zo|Zn=B={cA2Q7G;vzi>WhZnAJ<+i#QxuJOe+LvOb>EJ?1NZ8>1@!eiMXyefwQto%^m%;9_Wgav?jm5QWVDatkE9g^iUq zLPVN2uP|HF=@BaRnk7Z&e%ml!$^sXQ+|cQVf@mt_PmH*IqlN?DC?-a@h`ed>jH&XAP3>R(Dpt4Y~XC9N|qdzQmM* zN1%Z}sfo5oH%-H|yU780A>;3N;z@$t=UE4B)i&12zwKZEPO718av z3VR*`vQHI7@$O#DJ(mWvFO%kMct%c?ICLw5X>Jt+Qk&aqo^RWmEWMd$@~xN!mJF{6 zC*Qza%p%Yy>tv0K)N)q_1dR)}otj~IVek9N->P4t6*_18o=}H^N_RXJ`qoe*@rMq zg+7}l=et&FHy9{R<#liUb28ir0^;qAxX@OMjj*5I=@(!iL|O+3z$mabNDgvQyjBsj9NGSx6K*f4s__S9p6m zl62L5ltH8&T$rg3PCLKkfDYr>6vIzW4~q-8419s~7MzEz0QN7c!ek~BzHk5h5d)h^ zU!W+JHi9T227|FCp5EJ$$I=mJ6@TkRF!_UDJ;f3y3#-Ji`!byuQ>xH|rD&&UcSXkb zFpd)VrfzdSz!|=mh_Txa*ztt>#Q*6C#n-^*>!W>x6}2VbQB0U=)=tsd0gpv?Rt9pN zx{{4=w8*FNdg;|`)uLlLxPS?1?@{$g)TXxhGJgK+&-GFyZK|yiPxYF&>5~}RCu`|_ z&s(y?G7%%rTCbGT*PdnsRB3ad&KlGYxUg=sEuZ>0Ui0s}XS#Rv(c^G}_Wt`1LJ zp?<@k7w%8=nEl1jesakN$tw)XmR|Fa&#Jo1eQlOf`cq(}mx%AluNR$kENOg8rp=Mf z-Y2fzLu9&^&`J^<=B!V5UEWBGmmcqH8VnD0a%DZ>mCyBDuUnf%zt82RIRkg;LmUHn zJXR{dx|Oh(mfBa^SB^Mm?~f)v)V6rI{tM_w_$;;Ldue|FCi4Cm7)GZ3p@zADxc~+5 z!JrRipQ?R0mBhaM+qSRneYud$zWQVE!nOF)zH$;%ln}0mi2PVcCuJeHg<#hJ5p{S* z=z@W}^7096haQT7C}Unwup!SKWIO9_RV%Yk$8IS9Hn$3xt^1Wq__?;eznuQl+wT#i zkqpLrA8sZ5LwYKlyDlbDHc0U&Q4806Zx%-jslX@6Dx$B6&}JU12gL3^UQmf3t+Sk6 zB`3I#*bTrDw*RDT<}C5i~f@|?a~wIck_b>uj% zB%?niQF>H`A(+y?DIF7hMJ4_Y+C0+&X{n zn^*Af*Utn9G%HV7IETxh;-ocgALp&68Y^?PuayZu7FJm!dbh)U7OcduKebqLc};^v z64ZW2OaNQLShlPJJr}CZ-D93W8I?&%C-?k4L%IG711Ww?JYkU^aok%A^Jq&9PF0q# zeT=cZUn&VQEV52Rlq)2lO;SXS@GJVVeqg8K_d8iw(;bWDf+=kcH1D4-?)`p`dwm!Q ziTkO%ZE5mQM`-y8j@xK7;wbv?-%E!Ym}ls}uKO(reG;eF>c0L^&?GK{P=uN76;uFSp(930k}Pf95@_FXH(n)6>1(;(C4 zZWXa)XVOV8W$_{(SISfa$x@1qA&MyMffv-$x= zF)5{@;cQQ*b`DcB!@$xjefyAbbgd`$1=;-qRd%?XlJL!Qb+lrVw*?f8!NU3=Fv?^P zG1+x>(Twz=?mzWyP-!$CpK^p+xP-8}Qwbu>Hlf}7Tblwq0{_rO@BaP#XnZu)`4v9d zZ;qVwksdt@*WHQPf@yPodt+NM4ddr}1!lR%a$SpeLYLA~}nrU0=l5(A`J zTyek(5n2LJzp@D&n}D0WeDAIOFDwf*?OR=NS@G10ysEC=#m{Y!MBN%ux(V{`exk3n*ipr2rEKabY-%(mx13W~h^MnaRP)x+aB3LbKM4&d)zQg}QI+EN-w?H#Vo&L)p3 zrkann)VE>5(p91JYSu#cNE9bg{hJgvobVr$!uqx=I1lPxumaLmJgg6yu9+9<$4(2u zZxPZfMCYLb_is!T6+k*sXH^;(|*lwq9!|L=B3PC84 z*89aPGNFf+@Ofyo(|pv#6-dzE?YMf?6>t~7Bp!k$a=Co@lX+e2o~@(LU()t;P-{di zQfJFyD_^nV5MBhMOcW7Cc-h|e-zQL^J$n){G+~!);Rm{Xd023-=5IpS$SM zjur_$F6NH(7(~{e+s0@&vZ`l_tBpGdke{lEo^DOQ5ai^`0kJB_&W^06dynJgdi2Wt%?=Y_FhMeQ)x*06-RM3 zHUn2=3mi8_81l_(d!LlaD=wM8dd0~IoE}LU{8T0uQ%OEkVcLWfoykbl@)_|-gvE~& zZ9(5>!A;KZsLzYAnLJEaCWB1`AB#>LMq-=uG24Zh*gsbLnGTM#xWn&Ul%C#pEqsd$ z&OF`FxGO5ZO@h8`X0FrvQz@i$_c6yQoehq$Ca7ip@pAA%cOQ*1M_N420YYVkQ_Fnj zs|NC8Vy)1q8ry9h#*b<3vu_z~OXv(~~Lr%g9QSPOJixm>-Y14L}jg=X#bT1W-a^E|I;kmi)O=WD{= z;fFAa7{C=K_2ln}PH`8tPuLbKtK6*rN2&Yj_I-drf*afJ;0s zc+Ez1Y^q`$S$}Tk<6aLFv?z8l;4Ls#21y%=DN6nJvU{Eqs4)L#X)4iJr{ zB}lg<6o6YdKx#(w_=GIQ?Y-2QABk1oSMG@cXkgFLa70!@Ryxzwa9PA?UU5Gj^goBn zX|a8tp;#LpfHz*j7R$MQSnVKZDrFquOVELa+T3J)@8goTqr8A_dA1U*Jy`@%+db9SPfnTY;M;Hd(` z*~>15Cn+)38l|Wu_|pEbG+v0JC@qB;L8Ut{AFHFBT!O9MHJcDOV`hn0;w45!kXLpm zl0_p|<`t101>gPf*6&0;l$awNl19Y`HS!MAz9@Dw5aayERGGpwR${x$r)Q-4(v4pB zoi{-km{MBv+Zr#uxvUgy{_eI0@>wc{Rerjr^6~O>Vqz3hb?`|(5W;_fP3Dg^M`{vi9xD|}ObhPW7#O6pR|9&ekQ?2EzTWtD#7~!KuNPB_v;8e? zXc#!y8zOY$_^qzazW$$rg&}k{jtJtq7#6+0Hi@06#s`rdXMajTm`&k z2$ui=0Mss$S}vv_7c+hnCo|X^fSr|%mx+~=iH%2{jf0ykX$D|A1?lIXS~-2!O#cu`zQ2MS?ZuVZYNo{O$pA;1~b1SrY%YVeCw7-0Ezs z{9Ii8T(FN^u=Tf}H{JcSVgIF%y_34Vz0KcQ$o$)aDa6${rUVM(VIAoI>7Zcd0)p){ z=wA;Lx#ds#@295l=UD!AA2-y%9#|iqzkL)G_$ADotlZ2@-#FRZ0$g3KO!>vwxVU-P zIVAZ+**GP6+1Pk_C3$(pxh2Ip#l^TK**F&V2`6DFxc?mrdlOe%GdmX){xt@> z(iC182L6A;AYP{d# zX9tjznVpH}Be*8_*^hq+_%FBu$U_-qVWwu~1*1uO7-DB;f_!5b5Z1o~QhWe$2C%KU z$Y#KP0?0@zN)(G31p)qd`yZ*9TUod|nK`pC8!DO^8=HDMu`@f^S^VdY!OF_Y&B;me z@Xo`{_2-?H{qggk#=*`_!N$SE#lz0S#m2@(!OF(X#>xqxVExaX#(%rPWDm#*HWgb- zYco4%GrRvD91P6d{9oXCTnfd1dHtzg-~a#xkO_s78if|c-}dkw02TIPhrM!O zuZPc&|9dD!)zwx1{`=75BkcA1KY%>>x9{;x>}{DJ0B1m$zZ9Ywv!g2rCVyP4?92d; zEUc`&Yy#Y@tj6ZN92^49mLQl2voICl;}&COmlTy?;*}8NWa4}y!NVjb$_vxBVjS!e zViIf|l5cqb6Sn1lmG%$+D}y|w%|NDRPAvcVa^U}x|2a7R1OM}Ib8rJFJpOZr|JUt* zasFpxXQz;t5ffDvmzL3xWcC2LxHvKYw>VNrK|+*MLEV{MK|;(*$J)Y6!OO)9B=yD< zq{(HcBjL=Y2$peI0BeY;inDU6O3JXxtE)<>N|?B)fF(H<)L7kBtQ9;}C1oWQ)D_>T zT3ev;@ldG229?+TSA+gP5CUlbe{b#o6zw1W|9?;W|7-qd=Vs&NVP*dh{LcmZ>VND1 z|1VmkvZ53ajp%_{fHKk&Fuje8006+>pdi7%kzPZ?1pqLfTZxM+%ZQ6p{NsoMq$86x zk+szN@Y8iwQBaw%pS;TC5%o%5j1F)qeVPEs?8HWS;byARR`vBo6rMh!iUU&sBmIl0 zP|4EsuAr!LX8h8xM>MEkcbboyj(wfChCewE-E7T&N_`Gkx7CIdMfwg9p%IoMT-on> z7p-KJa0g%)LH`DjGa59uKo{)j5C0CopmVC7u>r(6L?rUg zK;|~(_VI^;G}*-v5)WtP^f7$ zg3tg?s}}=$ZfNhQXalAbhF10 zI=4JYBAW=8WbhCrVrS1QDw;rp0C)hdgVyUd><*nQ3>B}+ExKxRRO;ptGT_Dr!kd`Y zmn&%zqZgLU!c?55oI_FScFTl|LxX2ABo#87mxy8*nnYfl!d&zk{m*QFZgho+F(Nx^ zx_Ox%eA^GcQQse7BhI7~8_*_Sb?*~vK`8*{ef}hrLSWuzF0-i`I!#5(MKAI@y34U* zO{|;Oq4ssNDkfq0&segJu(rt!JhtDuFMCnQt9q)PIc}(OT)Q#Nw%)wIH0TRNm)Ajt zst_^I>GhKn~|seZ|dob$xM!e3nYhMM#co{<9KRRH>c6qgBE zS;EuTK5v424rnFx;)M3LsGn~H>3zFrFHU|vC|YjHsks`DnJ$(V`D4wl7M%<(gRE40 z?S1X|M{+ek8;vZM(QrbM6ROYej*Y}2&r4CC>@WQa0u6`GMAjYjI9=e!9wdz@cu3SL z{-T^mVq1=U+5c_Q#vt$jc~hPh@1)hzFMEc%?W3;=cP*uMuC9IFoI2$>xn(mZbT7R-ubqrlX>9rN6#1-8SjU*5 zluAp$ji@1j%LEt@`K6tx3=6FZ}wjS-S? z4s@#Z`)fy@r1 z^P%8uf<^>*BNPlO6frSN6jBxO%BS$17-errp7L}Opo+eT^LY7GM)V>AMLU2|TsBVC z6ulh1;EkW;d#e0+bZH(ZV$X;ZO8Yn^x-eB0GxcX&613UxztIxCdy~mAip%+bu=m$d zQFZShFiea%2na(r10$i*-Kl`mA`*ghr<8Py3Il_JfHb0nNH@}`#7IeZhlF%9@SeH9 zh2FRN`{!NjdDePZivu&W_u1FJ_7$Hi4&Zet;Pv!diRe9}mtN`VU1Hu59SN?xagisW z%>-ZhB3}zFk0U$2Qo{=##&YoW&+Ub`-;>)0tsh+*2;Mbic*vyM@~ToqKNZu1zw>j+ z#?6CkK6nov56;>jXq_-h2KX~~gQ-D$L41r+L<)+GUzkZD?O<=*7O>^?#5ZPq$g>+b zJx`Y~_ASJDB7{>b!zv?m6zmw?Nf<~j-n#OFx(z;e`x2~3QREHNSBtNhU$1^8yTkPj zTQ~N-N;%8%bCOmcXUr;iO{&iepT{R~arO+TDRh>dgr)iUE7lTu3xsm?7xPySUKxim zeR?````PO^8nJiUua?D>#CXS46c-m)7poTcnt2v0nIX*y+fw3u*^ZyTyCLw_OVd*Q zW?z4wZy))x{qlY42n)sSH0Ew+t6b~;QohoQrG(C|B6wBOB9UVzRhO!~s;)a1%$d&- z%t_2!wk;y#+8CbKvriU4GmNsQ@*3{wyireg(x1+3N@FtUxy_@@qxtaVjNd1`u&$Si zw?l8&-KKpj`*!@cWej~UPc{8`VrM2-#(H|{+x=@KFC${GU-rf*#SmV5bd4~ZEobQ6 z$LzVByV*(V-;9m$<>aBT;G)on@ehMHt%DXoUASYo*ID~iD|x4Dmv3khYUSG$*(5K! z^{d$^3mBHUmbuvw*zET|bkp3mUUuoP$W7H+Pt6wZF=z5N+5Ff{vTl_|ok)PN(%{l0PZW<6zoNj9 zS+~Q+M0YFqfK!!m@U#m|D?>WNCxbe}`!m1kqgnRZ+q1STH%JtsU8C(;3^KEGlC9ra zKmKOU*Y{-tZdu;(-kaF?Memi^g$r;HnkQ0y>u&k)5#@9GF zAI}xUy2nz*db6+VUDdsyOQs{BTU#VpW^rFz$FL-Q_+Xg5+`81$-qr$HHtyh!jC2s{ zcc?PR9jEZV=$-DH=bIW_C-=Jf-jl@0_Q>hTT-q)%NK#D_xtM^MuWR|E&WDn#Cvo9b zO4UfexJ6BPrc>7HtF0le!I$&Z$g-A*Z?Q||y?HGLEy?o_zMZzpbp_8B4bHyGdS;oh zh}ot6AohW5f$4+w+T>aauZjaB$y3QslF4w4<$`!F1QJ9>N2e?V|Bdf2t^hzkXQ zF22XBq{^mhz5M<%E!Y{XN7+ZNOc04bL43mYk^cteSAMc&YJN9VpLaS=`bl>vHxW5y zx#X%}rH0471ION{y-itK%HGt*WJS=|5-;dQ#GL)-)MK>!60$$^eC~O@G(dS6PsP*E zvu^5ZqGakl!_vY5|3F$m-xcnt3}ya!WmSbT>ARTk?9K$0J71W^SiChqpX0(~-Q^Nq zi{y9?sYHQeMJ7fb!H51cmiH)gUkJpL#+M3M2ds5WeDrJEoU)tRd1z4ecFeLZzwK>X zm!`hvK-SLN>a5ted)qah_dN=FvA(*$sZ}FXnv9ct0ZG(|yVa@uMZ8mcCTn|i!K$Mz zv6q9JgcR4SIH^R?f&d)+*t_KU>Z{=hk@k)3UPfMW)D^GzmDu~wkBuF^jKMs^JWpyO zJu|oOb$Z9H$DJX4d?da7R*gLpaXun(SEyk8#8YWGc7^N#ncIDV^8B$QKN`!?muvK+wvVj7 zvkh-rvpO#38JvDtzD{r*=}z)!@w;C&tA*MzS7iO&b*;(!#rsIf)4G&*Pv4c>JXu~w zQ6CTXr3t)FsKe&F>`J#IGNfbs;6rYthPE@9mYB|K({`pn=<-Dd2Zkz7XU z>|0DP;=X0YgapZ>1-;SI^3IY*%dXUIu}7KXC8$2n!R-!;Bre+eC6A)%7pS*Kmp??B z){mn6-uKp(oEA6Ff9A}}&-{LMoVB3RP{g-?ak1cJw(fknX4#y-9>?7v`j)H_)ZpLrRV?AgD(284rt}UmhU^rT zxtJXek)|=1q@RK(M&DDiKtb3 z487tv`;tXO(^URh#gqo^rhyFehNM(>FKd$d&=Di9lcKH*1_n7Z`Wy4sohus{7?`Qn zcXeEKRFp;k3hvS69!?K?r@sOZaSu`8S9^0;W2lF{or8<0hXm7KBSeAU(Qk7xLH`=! zYAeB{qoNL#b96R`3UKmsax+O@f(Qe}oR9c8 z9i1(?ctk`*xVU+_czHR15gaa_4z9)?91bq9f8FFi_qk#2V(M(|gf7TI(XWk79Nk&wc%CA#)Gw|E=WU^6zZ{8{|TN!o|bM&Gp}Vqi>IXS5)2F!`x2ihP6EqV*+bP z@(76tigW$rNB=lg_uN!&UheZ#b^kV1oa?78{fl2*|7MuvC2_9*GGFo%g!%d=a2~YQ zH`MN8{4383d|{%ic5MIp34CLR-3fTDOpbvejq%q%BE(!B$1OE5_dMGaio|CmxuvKm zi_nxqk0soH`lN(wZCDy9Kb{bmcs!v zf@Hz3fQgDXO53F-vn2QHUD{L}9KK8JkGa&X>xmsF`R@7cAN%b|>Q0|{OpdR-jd-g= zsxFOr;h!$l_tG+?u=khybB$G2uIge`p;{;MGc}(m^i<;#^fj5$w zji5uiF9tIHX(({%%LxcHN`63@_!j*_u|c{}zB@(!pLa>R#0e1n{l$I_dKq6ZC3v94 zt&r?jbj0COg59a|$k1d+$O3m(&}X5^`2`)W`oMr$cY5^a{6PRsF#Vgy|9D^EVkgre z?Sv=n_Xa&U5Ex9NZ}uVkp9gnv01Ox#oSZN#N4~+Ejw{p!u5khc+#{I63VEk$l=7kaRCB?KJ zD`WXlQNPy>u@^+QQ0brLk1tb!9S$WoHGe&^S}P3}D4+3|YTOG`+!H~*)VlxVve3w* zR)DpWp9LTM;?&gBdM=0(Tk-3~_bd$#ow_AUE7VDqX7<#{;RG2xLg<$KGn3JmQI78C z(iuHX@HP3;piDAg!He(zqEv8AyCL%Wug+%&6c`Djjv%+#30X0tSM$pRW@x+LW$G1sW-=LlK(4U<{*_b6+ zR_f@pNVEEDUhu?}yV5VG@qqr#7fS7~4_u%8#g~%5_%i-*-&6HBM|kz$d~v^1m59Yf zf=*P-laZ-v;tOyJRaZjqRr5Y6c+RreeM1tRqca|eZj9dBdv++ zdZpW{lshAbvod+@?qA|2o#FcgA0)P0Ud;PzQvbRvD*@V7m;WSB<9DL~j%WfhFZQyx z{r9&c&E^6O!BI7R`L8U+e>6mu46sFaRj>d0?EbZG9S9$^vf?#F{fCW+Y5^P36vrRt(=EyX;?l5psZR@f z@KuS+EnBMAYj?T(A^n?Qh$Jw&9K0d6x9)#-Y%g+Uj~U-*H`lPXqs0^bVK)_EG?UY< z=zb~FDvZ+5gBGdNhl$7S4YAsFTQ5g_k}je@NoE90FICt;`ma{*CdC2;(Iu~SUX@Fn zAuo!))b|!J<^~VQY2N;l*ng?RBH(=7BQwf=GX{N8j{~T8^VGGA`R(lhQegn8B9Q~H z|FEtB;D`t2&5JaCIcoRX#$Rl=KF4;uf`GuOy-Qv-zftd#Z)*M{3#Y_5)L;j`M1X(N ztZ4s9{HFhx^zc6kfPoTcMMsq`P~QUN^Lc)z?=R(Z$r+G1PS!mT!1gxunc~~H{hlY6 z`l6)`E(Gm@Oz;0enyo;abh2IIoHQr~ z>Ya&I4=f`61dm6g3y0SORF~@wr&NPVe}EB7g#l*lAo(8>CX&KITh|KpeiarSO~hOKX%MRn}`<=nujSl5&H<5+%y zKL=&qH?dA1-J)pJ%ljYGx4do1L?3a=Prv#0vY)8?P%Qs1WKnM{4)puwt zyZ+~rx^=M1P9M4Ebq(rh=g!vh5Uo}wE8GK}U}raUL(6p@Ajy{?l0}4 z`&TF`+c+!7wa#(t@&$lfqBU%SxBmCu_X{?t^(DR3g6~|?5!6;t9oUk*RD|J_e0|fN z+BrcyqS&ZcT0Wy5mOb|#$-h~BM6#YjP6c*HcsD2ab2&8|+z~s6XQUA%z+A5}=90$$ z5$y(5U-4HMJo0gq)~8Sb;9xkidz)sOqH@tt&ilNgsG%43IwC7Yo#Pme5T ziQc#5evPo-3LlC%NYVS9MlE5X-K_Lln-(>AZ!Xtz^mo=PFb zzcV-J=#{`jqasTQCulTdvFNwTj{>m6z@)&p0W{OD*RyK>&hm&F00zGImbd$3oFIwg zqhYV#+Osmi{@6Ze@pps{%?SmJ|K`qtp(KE9sZ^@=rQ;JO508~umD+*P# zNEaq@1!Vo6W?f)(w2Zfx8$dJa(um*hXg1OW4n;*G;WRnWsCNvxAe@h)(w>nz7-9*;lfAG$U004Jbo)-H9SEz3T;*rjomHdZz?cb8&ll2E@ zDuUl#xN3+BaH&e1#2J}d^nS3GE5H%m7jh>2yPQ73 z2&9*xd)ytV5diXLFYw#VH9dCpKOSxJ13iWx!TLTNE+o!M7*7=%H@6QJnIbtqtP-b| zFdTC?MOEhrg$Eklt~`ehjF9Ms_WmImZj~i1Kr}~`dGsZKTWZ!9#~Tz1@<~OlTAIcf ztxwb%fvPp?RivxJZ`==Qufxc-p{-qwRc^C!Pj-zGS#=eArO1^ zecvamCVb*oHhtWMbQB*A&fWK*Bp@)YBIs`d;~FN769k`weS}rcx<#(eUI7KwSk6t> z?oZq&=hW9Zoi{vvE51MGvU$=VcUk03lFwG#Jb|N!9GrMA9-qoZCEg196n<`FfqQ_X zD(`Bm{kx~Tjzx7c?{enM4}Ahz&=|5r9qpRWx4)eK{$ZUHkDQ%N>U0x<2Ll`WY<_ay zXuKFlq9$HM`)KT8A}B~#XjfR4n++v*o=t#$l_$2-{SDt5*vniT=l(QD zhN?$lx`cp6Fvc7i`%=VZA;$3RXwdxhS4 zjxoR$>Czt*`WPL)FU~#qCke+#fmfZq)XO_kBtE>qUy(odXwsL5ILau>SI2$CItzTs z@xxjWHczRuRA1}Q-QI5eLu#H9B*p2rG-dBFZ0##<%+ZpV|c zLj1XEHe^E{iL*uxDqj#a{R{_#EIX5^7@@`!Q;orY33mn<7tenqm?5SoLm|$0f4m|I zpu~K2sAyAnMLLS;xh1%6L7>3eGIvbn+^P%Y0Id4H;<)}ZTbEohGtcH;1$2pTE9=&H zJk%|-(NZS3Ev&dPSnq$vYx41F@!9EdoN;sL%Xws63;%1tz*tl9nkj|^StdP{hm)J_ zKp#VbNQ0zh6=DYgT`%g8`r*`mqj4DNbQy;O1%`(A^7VBsIuf+VIrK;ia4{ZDcrH3O zU|zftHTdDtI^}fBPO36PwOQMXtu{5k;??SveDkI5^!e42E>02_Wj1Ua{HH`wUROau zhoY$}L!Ap%=PkVo;75l_<0gKtD)CoKN4_x)>tMHosUxk0az{OrVH?ig`NI>{>J-d_ zfS_H4U%o=5Uu1F<6a**M|KM&fG8ZMsH0F(3F=Xfs;5BDEE|QX&`TAA~d6U9`G~IZz z&KD`Nn54;~nnVf1*&Z8CwNofFtm-H`zif?ugY?Eci^I?OF--wndb!ogHPoes#co!_ z5B>$Yu?m<41JBdL#o1wBvDHHC51)9%7ZH$0Ah(*+S%$MAV(Q}L*^lO5An}^QXo8G zLNHwW8Q0cht>J=j$G|vO3w+mXVTz_Z59pDLq_Xt`#>PH{n}mDJPfh2bz5ZKAZLFzq zD8llL=PTly6z&FPt&mlB8hDM8s#cD=B|JPr!m9 zUsY13a^AZTl;u#dECfuMkp z2xXG?D$nG{Q_n?3SQlY)rqgcakDmwjW~p(3f(yLISK6OKVC{#Csr7ByO*ZkZw&6d- z5l`mL`{zSdRPTwoNj7#&_v$ zWo2b0f}~_?(&e6fw#AZ>U!8BV^gY^ncI52UEzauZq#9I%e_gj%GLzQab;{t;I9V5y z_)fPR67UEQIBPbVz~t60p@CGO1yRv?h9B6L39v1Z-|ugr%NGxz%q&dpwkp|eZiVia zgaCvWC1$H)g95$h74j5_qK;yAO9_T2Cp6F`5}#JU<}R(6$Hr7_k6(yy-z#H=X`_|y z$ZcoIrki^mlH6oVy=qbh4uI))o`0zwi`t)*%A~J7ml&@)CE>y;JRAz!33Fj9}c%8#}d#l^|(b>k&lqI)z3fnmR5ac!Y{1lI`2 zP(VZyITh6Ig{c?%pPfR8h(AL1%%>&r-0r+fXIQjg)GabLT#641iD}~sJ&5S zzqX-4C?NwqjG+Exo~va^Zn?vBI8~J62d+s1#O;Yx#r=!|+*qS9b<2za;xaaPiOfi$ z#?*}lk)7T(oY3kUrX5(FMe%RV|?R@sx&k@%zGWa+R z1OC>uy3-GJbTdGaGVZap4DLyMXv`7rnsX^*#)+`hx=Wu?IjaJ}*UOT#L$u#o1SS0! zURk!osrcISDki8InY+kp=3Xc<(VE%Z=|uXKW{p(XCEEUwNT>-2oLVp#d-t=@WodPo7*S6*XeBcnBTy_SBTz65-CRk4D)UhiN!~4=?2*$ z1sxnhj3T#Y^-1@bvwcoBTaXb#6O=34oni&6<)aSsy9NHfN(QS+
    rKUD70>;VBR zw?Db@Hzz^fT9`M|Y6BGp%mc6EM9o&>65{IIvp<~X6kvee(vJKJEPm=azd&t>hLeR53B~`M7V`8GOLgj`o9Rd&w z-p5PX8X4`MC=6*ap&8`EPh?xI{!e((;}r(uy;z@)>fIO9WTUt(;LqT#kxP9;slMCW z_Q|fY^y!BGe-v8`<2KElau<-Q!If$qIrcA!XSVg-_ZEl!u`gb> z!^Iq~v+8-x>$31=YonPk4uJF@t)!QY<|d;U<=X3xzu%Wf&>P_3QxbJyn$dJ&k9jU8 zZ}W3tviQDsn&7w)%W=_t3-9P?CY(#&=r)^T+O5JvwX+>Ho`lVl?Jh3w?^TwsxQ%5) zLMz2&3!}M$`u`Vl^S=lDPV)mjv7ZtS-C9^7+TsidqC#pA@f}FE7RBCM^#ASYvc=-Q zl!3#iZ&r^#uS3#TkhqruL{ng#zrx=LIx#_c9NrF=-hi$5(pb{6 zb&|eEH|Rowbc^$bJM9S9a?R7}4!_78iTIzM6a!H@7=$Z=5T1F8HU%*`npyHr?UL+2 zK(e7~z+eqTjLH41ZSHt?<-}zLW%I{krx`J|S94C^?{!O%MSVKvf;lT6{M(P2eVh7G76+Z`3g*P#Sa`C?o@nG}IA8POp zy{qfAUsY=Ow5crCeoW^+#MmU&^kEF)>yb|n!=&JPpr7KkoI?pY0AS!J*MyMA@RQj? zuMELRoFrcJ+u{^X731!<*4Vp642!LT7Ky@Ai|t+2d`8G}vq{~>av(@z&%B&0Xs;)N z+UZ4(0y@eIKmqRX$fGAepdc*(1?ArO{xjBS$ixEOJN#0+%t{4bD7OK(*hZQ%F2Oe1 zugX!OVjF2QP+)8qe&`AiblzlZ%@&dUSY|u;x%AbC^*V3kux=jH&oQ^6Szb=mdbhZb z#CFXhrK1G3Mxz30>@AK5 z6Cf_;*pVRc81hh@BFYlvjL*o33DtZ?@3oY^Dc-O-+Y*PiU9HlxpTsY&nVtlmJ&>QuTJV%cxx|e{D zH18+rOv?d*`<(BN4v0P=$P*4ti`PE3Pim&86hU`gRbPJ;%IBDkLi zM9E6@Yq%waLqa@N_om%axDeK;l(lgKD^|&2clnE7JAoKF?lB%kEIJGK_>_>EKf7k0 z!|^Ii9?k}sz}4>--scNWULk;8LR2aX-ua1}DIjpZ#{dItNQW|S@E+U2he;Htd<|sl z;g{w9Tx5;s_9f#fF2Bw2Vu$hSVxz_&$~APHflgBJx-NBZl~>eqly8dS8B;yh2n{h$ zN^}#Ln)L|@fs316VS=p{&()O*AUe91jU%qxzwxc<%(;Qirkt+SpXR>TDPl^=@@Tw3 z@zF{~R_6Z{STwL;fm}kILE2QGiy|5jQwR-z_|7Og&d{-g9hlO#!35PCX z@se4sKoKs*W>GCXpeHf-gtHx~XGVsXju#g_*;n-A^IcqIW>Ry%?U13~bW1i#lgkSi z(-G=B)*UW5Px|h^1Wd^Xedji2bS4n5MAZ>ab}q{OPrY{S#gDEk`U0hjLR-T%xx(f0 zq4cebG6P4uD|j;Gy=KG~=4qcS`)&=zWGrMyKez0!&WS_~u}UFD58zH=)s$88MrG(U z!BZ)Dl*GwKGddADzc=cLcpeZEATmEc)fk+iqKv8OZ>rm$H}@R zrjh8y^x9H{;GC%YNT^gcOs;&uH}9TYeR1uu|B((-mS#z8`WlRNb6RvGVMR90tC>J3 zCO5zOU6S9?w@v)U`h)QLr?!m(iwEz61xB|LG+9whI-g1gd~@GsES`%Uj9k#=L(PXj z*>gslwmE*qQiE>TQf-jxE$% z`cgr6;d-W5rrKX#N1hyR-1IAfnXXGCKrnT%oj{i zB2{{Iw5=SUxrD38RP3y^n)mJ-d(r`2!+|crJxi4{7jH4OGJ_BEp1QMeE*Mi4XANrz zGto&8!jDw*iE6TF89uFW;Rd_RNP8KwZb888{@ycD9la-?M*FP@&t<36F5*&x_tL`6 zHGe=+ffhv2p>#8&mn*AcjB0LeC@PiD?pTai%viEbOp&ZlZ4JO!!%w6USH-EyTI$$! zrQ9HJ4Z=X;QMA(D%hMy%)aKMFOk>xtjWQ`0m(CLAto(!0e z*$qZxGGoTI_S1P0myX5$ragO(p}PoVA>7{cL1(k=(U8Ffv5^n~do#uwq~JpdTtmmI z`B+Nk*0Pyxw9nmRRPUJ8(d0Fo7Z{9O8y?gPO-`CTm;aD{1OQq0eX@;{pQ6-6F!!!y zs2oU(s&!LTj{S0L+V#PZ+pCG`YHo(*B$4J;A8losE;VYTz_3f1m0~~gt%A=|{7(<# zsbdDk@uI#-i%^V~Y^+K()oZFr2sBA$;xK=CJ+X?tnC=(5ikGTsMS*JN4z~v-#jGmq z@d>(bAXW`ms^6`(ZM$ud(uu+w5KT{|M+Qco+J0*Na=czI6*P2_-Xrb(gWBpbkJ*Ta z_<@`)MRyBq1zDt!b@QT9=W(xe2D%z_L{{$qqgWD>3<2h?G`^b!yNvc~EGD%W(bnC; ze(9(+TD8KM<3j&k-%!)l7Au)fCCXdEH=(WkFX6Xg2QawFUFfLCJ?NToA46A_Y0Tuu zXC!!CnBoiZrcAA{wDu#uIoRrJJ5IRCr2dvm=ktP5QO;|2Eev^iTT%G!#eSV}3sq_y z1Nvf}D)EW-?NKpCF@qsRE{mNUgc3XIRi{TQ$eB;!Fe0-zMv>p&K@wMDbWO*cPj;luJIH;P*Ltb zDz(O?Q}~YFG2|o4prdUr&SRF7ABe%H(zR^Eh}bsnXj<&w5ZKS_r>4Nt3u;0q1QM69 zyX@z5gV3cf80E?O+37miOd*n%YLFBbIekz=UuILXsBzwD0()9HTpkkqaZk@K@aovs z=D#ot_$1f_m4#uY<2Qj%hn|Q!PYT!$e6-?|kKYup2f_zdc@%^n0wsoTj!9v?#l*%b z1(_X=iOORyWb=AX^JEV?P zQ>CV0fX6rD}CvIeo|hN2V*qA33`9EPtMzlqCj_} zy=!ZGsGGpQd(sXsgWsB_%i@he%*ja;9v6HO7hQic*N+3@o6mmw^<#*YTem7!?+=hM z{D9F_n2ex9S13jteoXR8+)SgMBbvAy7r7HC9bt)?uU}RCB~_~aONxY8ij5&!{~_Zat?WCqJ8U;v!2#SoYHQ*Ud;k^5B)%CU;zJUoNsnr{lQGQX?wSEbLgj- zJk}Hs6DmP=Wd|p}&aA3Qx0(Rj)c`#*MTsa)85HqSv zMH0cVCVs4Z{4sE}M^{2rWSZ^a*8F43!_6e=lv^!)O9$}o=RKD28L_3b;C8?(<8`IO z?v4QNYSyRkX7&x?Pfew1rO&o4y|aUvo3F`dZzeo;?;+naH@^FzTxx&X<5#LA+^9AxMZzKPM3_WXU|cMYIAE zu$i(5;`Loj5^*u<*h%w|_h>1>z(t{Bo_GuA_L!@iw4zN7Ni)$oH-;aLSGz$}A?jU^G+KN?UpJ_GyXA2WqJ7=f?&Eae+wWhgN(r;{0XhLK%2 z>7$384dfe)786ZCu!v6zi8zb=aqN^aZxaav(5bjf^7tcvv@+zjuskYH$pHy{whywc z$$D^U)XKCxx(iH%o6f6)oO;&9TlvP`Zc%u^Gv7`wg>SGguD~UAFF#8q@!mpbG7Tex zFEeOuvfkhL3LU0-nJ=Zq=;(ca%kNL{24mbILl(w&S=UVNdvQIe5tjEs$J9K<+W@qU zToLs9s(2U}fgo8~zAF97tQ$!`%elv_b6Pmtp*-ecGX# zt1Yro&3^1YvBpr?;nB`Fl16gPz+$dOii|HheD4qtM*u%v^*!E;fJQBL?yd}SW#Z=H zI1fuc)qCP=!H|ASo}}NSJ~{T$2!k z@E91>+)J|@olcQt1)(+({`%hm2wn_00xRq;1+yzx(%Cfai7zkgBw44(P+xSa6DPDc zZ)-Z;33l(f9Bbq6T)&t9%~rK1!G1<3(`#N_F;qHG-bQLd<5Vk_jsZY6r?yW!= zJ;h*^V+*2f;>F@vo)$2MGR{0#pr>lQ7b+hj0mr$5G3L@GVQMy({B)dh^3okUl6(dGVEGhy&cmjHZkA3lwfi?<{^i>36&~-zm1vjxN~_5xpP) zu;nTcGR!AVo>&~VsrlVm=t!iw) z4~?*u9g-KmB}Vz_7RoMH>u{(=l6=2010ku`DZc#$UCh3*6n=dzgF&abn^T^~`E1Jc zyf=LXLO+LU^V4&vX`xu4TckmhvXt@xMH|(kFW}tIdVJxou#blg7bUxS@7V1u^*pwQ z<^-pONk4Id=WyYoC@-|w%I4tb4k0j}azD*s)LlIEP)R2U(O)ekNgBSzXa^2(iEO_TK!v-26-7 z7a4rGGiAno_s5aeS;_Zs2&hTd((*trcjCO$ywnEXMWiU4Qw%R6Zld>3a*b$y%*Nr6 zwoW~bdMaxaW2}wiYR78FL^^jE04G2BP=EXZjDvge)g|YdkI#5&S<@L9?+Or^xzgb6 z@`y$>i4>b*=n&+9OV~h8G58ZRDiNm*p-$JZ`A1rg#~u(2g5{`~6~PAp%KFGAlLAPvSj zSI+g$rcI!qty>ElqW@6>WI&X!ZZh6@O+*Q~Qbg2b?)1^Si954-E*{BMU-MqSiZ{hq zLHK5&+t|aFPhesBT5~YjONWzSQ@+A>qcBmvG!v=FoVYvo+M$^`b8HQ(*fLi~3r+O8 z(J=3E4iD}g7%UPSz%TbPT2=IF5VyPz@eX7GGBGV@q}brqC9(gp?=>3lhEs4|cF`BU z*qcktwpDX8g{PHGOrBtiYTCFQ5fCA&>%^S)wj-t{b2AgiAAtIzSdHI$vnXU6bxSN> z;&E0&zrVsJ6zBe4M(g>Eq$}|X>G__KZacIgaR0RS<71WTFib3tOMyl>@*#>LZX$QW z-EfSk_#t=m^vd3ytXfK4SBm50P?JKr8ZWD7Hv6o5Ptw4?%EKukuzKOpjqt*2!6`Ou zpH}pXN2$n*JNM=-4Qe~1>mh@E)UvKKt37O%wX-)SaQ$TdbsmBrq`J)<^G5YjnawfG4I&tnz`;=X?zYqkvuhSjtYDDAv+J8{r zMWt#c>^jp&7zS*I20&aQ(;*}?!=CUrSFX%?pp)uGA=%1hVRBwZ@^lD^J)Mw%iZN#; zxbVjWQ_4U^6r8+fqX`e@WcCKSX4)qChQa<#p8mSHcAmIM7rHggbTIzKfBqIgBM&C8 z=F`ix!rb!weseH}XOz55b~C~-F)b! zF|n~Rj#i>Ab;h9IcH-nma8*`{*E{88Q30$%zAo_+dDFx*VBS8hvjg;B9D%89WFM9e2o|pez`0*dDzJf#h()nszW1? zgEEKl>Za0H@b9KaYE80DMeIr8+lzhdfo5Yghd8gu$<6__p+EKT(HaA}uW8Z5LCfHQc4ic2DN~$Ycxp20Q zPbJz*o*6q&e>mum8U$DaIM@ZSoSMjdOl6$aeViZ>0o%NhZ=UD7sT|OM$e<1@#qh^q zj0qEqmOM!2cBoPni;-THvw4Uv$RtEd+1T@Vts)Mc6h@=aOkcBhby=C9cjB=*0#~W- zd`gr$YZTl=62S+@b)7FMr5whsv34hjJ#tVcY)9KVxFsvE$~8?-;t>O{Z*S)Tc=2p; ze!ArB^ti^QT6tC#T3bpeEjtpl zzSp0fHj%y-vDvUSJk}FeiqZ(vk{d}1gck5EbwE;W$^F2DZlM;!f`ytObi ziUqGxrL0ixdtB-~8|NO3A4d8GVxbMgz&b8U6sm2lF#eiB;xQ`{%l+T*#Q|yCU zF+@NS&wZ8CZ&NC6wD4VIRZF4GmLukF zCIH~W)6c><%N(aq>_$K@UdgQZF?_aCwKUo2xqwn0-XPz4g#Pwen^x|W5k%;BTLj5q%brvfCC&Q|rlsCj#ghu<{1K<@^*;!J1l zf)f<-T-Px)T-P#hKt$_mO6lV!0TxBc^_uM&`m}mS-RCchhvob9@&V0Q(#-x{YoUDY5K|07*c$ztjkUW0~3@ z0hGs+`NixmZ6i?&k25EpTg|~tIE3k5bK6PH^}*7|l0BS9ZFjRkw&*r6_!rX{`gHoh ziHlMPjq%GalfL^AEXEb?Ya?Jq&+#Fs(HYjA zjQYb4M<$Oe{3rmn(2$2bww82R_(J&{IO_w%wCt$I9<*y}mMWsu96*YW!>Jn)rhH?} z_+P56Oz#g?Uf1s>X~t*PB_ZWSUE9Z%VMQ6enL@|cGJMB#wKrSaR_w$y(-wfBebP#* zK=TrKTrJ$EfP$5&x+YcgxiR%3tF96BWNJ>^I5s-8v=h!5iPnZAgPML*ypf6&ZU zRpqiM&*rh1{OEJGY*3*_hWtyR+HWd6Sdkd;CPEB2@vH}VCNvKx$euq#NzlblmSsfR zkkYdPh#%vx#i&O?fJ?wdX6XFhmZowN|Iz?E6 z#F}K}bvbnkq08oTj2Rp`(a`5kF_0kZ+iABAKtX*Sd;4|!wVNl6=qx)Ml2I-+l^mITwbXK_7GjOMdUu6G%`< z1SlI$dxhv^Y*CVe5jIUZx)~3cXWcp<&k%SiLDqJcR6RHE^XGy37c&JrR-$B>VBd<( zvSTw=oV_^)h8OJQR4=x>=Af}#)%WU^nNw&8YP23T8gKH_JJ_$tkzU(gETv!L$Ndv#3mrVe;u= zpImAV>k1Q*nHkjWtsKYt%YgqiA*NwoIH)k^2tNxwma@lg)SGZi`KgQ9C>QGx8}n>h zH9I@;iDc;BIlS-u(KQ@ifF9**&JapR$Qx>jU0O!-Zp z^A5!#s}IxN3_|gB*yC|yZX5tWK1l``w-8Kh zEyYNWNtrPNt_D|8Qr+tQR2mP*H|vN=@5exdiKOf~sa|T?N_qQeeS(8iA2AkJK6}&Z zo?CM)YUlxyLBFuAe9l)|YDj+C1@2I@`Pmr&fH$P#gLa!4PMts*%FhzIT;7b+DopkU zu6z{yqFW&dk3RTcI=w*}a{4lCnMx-!tf4^H&-;SipJ+<>!65G(qk4utTt0(v zf70Jy8P!&T8pt=G$)LmC^q631Bz?@FOqustmb-N`B)E^qV`GZfYiB98n=4obu-@E? zMvA^ki0UOFr37NX=>{v|@)AO0=FyE#(;-CHc8LwNRWhbDLP+2sSh8DwJqCjn z|BUi_^Rn??>B91f6qTQZf(FD3Z{G+@3pHX$B0wq4|dR6oTa}Zrgb?pfy`KvISv^%cAR%8Vd^2{hDZ6JXhAr zaQ9uH(Ra-ifwe3NPX;Hyat$l7j5BWUnP;psx)FvaGZf8p66>C7R6-8=BIq(9#+=@i ztVB(xcbMpqyu|xl`G>@3=bFbXFi-QekWKZo?2>YSd!8_@6U5t{bNWQgm_Uja*xbSU z>cR-4aHqAw`ct>6Tv11qlEd z*j;m*ei%ZWHV%uEwp+>f3d>4YX8KgSRAC%vI>h*h__!%3-5Vcv4B67A(;r5+yHh z^i?e?W{;E2mN;AzmJn7c)gu3dd4v{K_CU4&u+V>>LXd_5@?K=q&w9#XHGHtX7qL+m zVWRc(qwhv|SX5hPPor%wPfHo=@qEe(Q zMWy$a1PE1WL6jn05G>TtTc|?lAV`PML+_B#f8?2YzvrCu%=cGj9A}v1zV@}(+Iy{a z-$xVO9X$r&0ak3<)sikV^}eyQ1fBC!4e>)Bp4BV%Ro(I|KAP$kFZ0)t{;-^BfG!}Y zApjSvXn3llrC$9@@+2{v&B>Pa>+O?4m!oXVaPMcQgA^P?Ux!16m=>br?k~a6P>4oy zmsoxOk)!rX(&*h{%@NEehlySlxi=msu6mUVN`cHyk1m;?KbrM9{A1K{zT0$-`lHkp z>bA{<9^E6D|Fkh>B3Bo@nX#CH!`1iwWH<8v$qqcLh%WVh+U{c9F7ur@jl1jmJpM*I zOGtyytW(V=PC=Kb2VT(O#a%4N`kBC+K#?48NCqV8*Yim%rY-cMDaueXH#N1=dGTNUQ9Wv{FUc zZ^f|5h;K<8^*`|#PU)~e_YXrJf@11bund7%_#}=*OA>^MOmiZ+O8*5Tx&7 zSe2cjFCF_WBdOm#AS2GH-mXzzq*&t2=HywQdR8)Di0YB`v#F!&aFgw$B9@zvQh__K zhX%4$O)vDiVfi=;-SRSRqZW}!h zTYJ#Ly|m18GG`g1L&Zq;v;WEwtx`$%FfS${RyU|;F|CaGl!qLDzX6>A_`#GgP}(td z;@Mov=<$NLWTI2eV%+TU0+{RC;f|^#_)D1rm>By5c+p*b8N(sr}vn=cL*k>u0GWv14?}sWz^*OO<*jJA7_I5i$xLQTeH(b(Ve9n(&b#Y z$!N{r;kdPpEIj;lk{Nh$?0!XunKJmath5JkgK@A2869W|mi<5-^!mrqTz+Hc!b@1u zf0;`fVT<55z4>2%pb|UzBK7TM|;w98{r@p$^BE@v{| zII7&kerI=3&#atGP>;dIHZuxeisH$G*55aA(8 z61h(w{+Oxz^FxPuV~ZiXi9YIs4SnX9PqG?FrtAS%$9>8^n#$%GyY z-CP=nf-ZI~Y4Zj(+Z`2qOYLHfT+kclmFD@-=+g&jPv0MU%&Pv2Cc#H@t ze79<9S!ozAvrk$b%I^R+$3Yf(#44~WgE>_{<(PMI18`Cb0z>Q~4E34A@Jk|rG8&+* zict$|G)CD>)yxkzx4adz`8M*#-_#n9RSCHtF9YuXUG>V_6gMo$0Q2p$3L33h+^rD- zpOx)ofG)@5N`i1~`l9F-dj{C-9X2b|>(2Fgu8u|fkeZmyH2(Dh=eUrIGtJ7zS|Ljh zMMdlwr2szK>1;NGuRnd8%uA5G+$hV7Q!A*wp~Evh6iyhlhI&qUfKCP?xY=wzS}<W@s@%QY|a!=;CnGAf9sSgeA@ z6%)6es&*5b7J&W=SM<(xmB+QUQiIK$QOC>GQhngdGjSUVev;Uqu>z zU9tPsq-^kU* z@VgVlvV^W%U|jm8{_=b~F6$l9U5XA5%VivAE4Xr(1t*;i+Wwx}UlX`L?sN4$Pw;60 z$5`;CAHm9fJe;q1D?i)0MIlz<3RLj`3GVe*A31|ZzdgyBukqIL>e?Od#mQ-2bBDlu zjIlWR$qupdd)JybOcjdDHp}CW%FIgL)cAs@|1&CFthSJ5_ni&4Zc)@(8jo-Bf*g(8 zcelhm?Yb%kupq(C76F3!PEl|B^#{E8VRInfToi0Y@lND#TD5OM!B*4_u9BwzZMeGt zAT@q26A=VD^H|ir9>zj6uuu61NW$FEQVO7TBK2Z1+-a3Oq|>m0)-OXoQXUwWl~b*W zL(u07SzRC}Cwry|t?bQEu``;+w|{;Bh5ZsuuP3Z31^&Rsg`BGf7XlY?VI`5L54T-r zL`*9MRhn5{>cvR1DCZP+-|4scM4ZN)#zBuQcA-D?Oj0pCB+5rzzPT{Q*zRpg%(d91 zGeMszH_d2`I{bXpf3%W7WKk~!f##XkQ@~dk~%uj3K{!SuCq;heLko%Hjf^$r5G2;D{Y^> z?asuTlRI!k@;%!W$pb>euEM|qLdyd8r2^@nhQ_i;+=1*Ku)rsY){Fk;@w{inX`Gbw zOXF{DrW_(38)N*g;x$q>@n=@)e5zDO3(T;RJ7xygDVs%@EQG;> z+%;Gf-oz79Px(}usd&G`Te07%oD0`~v{+lX3DLtVbCu6__-HJOzFn)qW(aKicF28U z59QcMYQD?J0&8!+$Zj$OhyxeunxJ}!R%qLK_v>Q9j?*mg{E#q67gA_{PBGzY_16H{u$$)vE)pj⁡Yvxle ztLS?aCSN!t}SS@`^?oGLyCLqhS@hCT3?ueZYo=Sh=E&t6y_Lidy^bTo z$>(5!zsO>$lYptffn@HzD+m|O8EJoH?%v5K#vR8gOflr;j=`-C@k{-}!TrVA=DlVX zhI%|*gS#y%{QVi z7SBuEm%8w$yP22AnXS=p%vj#}wb5!aY*AluEN-hq0)E)q` z-=_5-S?QtG>YN_ObMplwSiSt<3Y zex_oB{CU$wIY!dxknpc>u6%9V_*MGUN#8x2X66fV1Fmw?a1#&ez|$PPtfIzt;ua zaH(C*OWd9aBEB|XP01-EMa;Q*r>$RwS=zSiBQtx4&nM3Ki>q&fw3*u`>&wJo)_fk~ zzE^oo{N|Tu2|K=`i)|PqK&2yLuKI$c+4GM4=IhSW;B~ZLw6U>rgXC~zKuDK z`Ha5P52s19u>o;UkO19{{XD6Ldn9`~v9dpGfAmoT_Q&0Zn@fkafxkY_#>~F=*g2oA zhuDuLoyu!phZ|W!y`p3J%CT~qksUybSe|eBRH)G_I^JIMJrf7%IF1#68?z1U7|MLh zsinYhp-DUu|CGXBZp|+p#=P)olI2kk#whb-IDdW{c!3Rn6{k>{=K1@PP~RU;Pt((m z3aqTVv&H2Z797`@T2!}HD-OZ$;Oq8>pX$eyBul#K?r*m_DI}f#T^&13*aAaI<^P_+ z7RWdo@dh@F)r2OpO+RIiS3$JLokth&lvw3r$oSJ9Q-RNR_RjYEJM86sqU`slullxa zTO_U6(sbR-VUh8^nP5ihFOMR#Y&UVwb!-NTIIGXs$=8ooC>UDv6K$w8@c$gcQ9{q%g0w;=j+9KVTzyml|+{yTT zlIFOOk`hg16UvE`Yo^D^SaW78bco-M9YlIy`(lU?+hja;;}Q1QLAyTi+u?8c3a+mwFQ?-^i#Gr{(QY*$LHpGTf)&kj=p)55UX-w#^2IZg*q%0;*pWd(U+cY z-#0qWeIcx9;WY5*fb9k;xatwlVnDa) z44pNaKoyCvIoX0Ilz^R{If9P)PB|BBI-;NBuqiiL#Ca?8Sx>FhFWs^^6ko}C4Rrhm zqI|Ht*h`|V7Zu_jO4~~-FSif0z@oWga}*qN8d=^y=Ew4|S@#|zZ-&ZA(?RS~jh`lc zbpz~WWNX$h!xn0v0zJpt&s9oL&eOGf6~#~?l`Qor##Z7kE3Zsq-9K21~G(x_%W4z-y35kZ)A!d+vlGrd_z1K*di1d#v?4;0cs70fsTjfsf0bu< zhj-cJr`fd&)NI;DM@vAFNV^|6muvGSEAfNVLD1jD^wU>JXe|9C<8Roqh!iiD1R5;c zX%+c;qUsMzbU5li$M0r<(vIh(14H{kAIBrYHvKq`c!GEwXK%Pw?yD+*_vq=6CDfCO zxSIY70FbaO1r?An0?(yyfNd_pM_UIbWj|!E;m@z_@mU&t`eAIr^;SoVhFkp6P;{3( zwK(u@$V3ou0-bC}H$W39E!sb((;F?M%X9OA1aUsmAg}i)+xwtBu0WL;%x~<-oNfrz zMQXZA#K9;%Ag^K6;?%@p+q5FXXOl-@wX}mHRq3{K=`Y5^EecC~3&TC++w#qC1z*>T z8&Njh8Y)}RD_|hV4nWZ&+IcBfcRN7EAyi$OihJ=tq-3}yZRE&44^37_PqP@VC2)y4 z>%5<>cI%=!s|q}=dPfk5ter~Uez`mwLY+Zgc>`1%idR}jyb(~}v$mN!;W5R`Q1k^j zJ3K${t~7`cw135m$<~cEKwdg}0(a^Vaz)2P@JGt1ubf23p@nj`S>w)3Z(uEC{LdN? zdp5iLNR_7e;AjJUjLg=o{#l3Dj5-rTm!}i=4X}Jj+V}E^i#dW%Z^XvBEG6w+P^Zv{4(yb zJwY}*6gL zfyBgiiwT=jq?HlBhC5) z|40Uw2k<#mtA-KrL9gUw0Z`c#`ZgkxGj#??>3(sXd0BI!%%&HsrPd5h2}W_@S90PQ zh1(q|o-XG4i+?STcuulo!v&z_0DjtquLKu!Hq^e+4;PawBR*$y{3Xm4yn?Y^Pz#DH zx%WF9EH^gcr-kNBAFe1Ap2wPEPDVnvhn(s*E9tMA+`Th8>#N(LU2lloM-0idxR|FOk^9aO} zgz5>#izviI2}iof-Yz?Nk9}XWJ)juIb{KtcyLZ|;eHbdze^TW=Fm0Uax17?l38xVL zc6^Uwf@?@7tkC|&!40NYh6U6koGj_chnc?f?@s1gGcQ&%M=HO1!lNWbDP&SFtXI-Q zEaFpfQ0IR<%hv!&-mAyytaVVg{r7*p9c&V#1g~58>!Os?1>OUnfOi1z1Bm9ecQ}IY zlnCOZd+4PTSHlsvukX;ZoN#xTxLO36X$%q4G}I!Tu9IHcUhfLxzLp))gCFZqG#n59 zsgV-2##5UpXku`l7bxx&ha*h5=o|6BMAicA#1B63C2!m{F(hELwE^mM$A|an2c-rN zL|!`wDA;e)GPGxj;IQdF7BQBRbDH3j!IQ1*tkb~r(ZJ_-a&$fu8p5Ga)$u-79}Ov2 z_)@l8b^~huQ-Cv0jrOA^Wy|#L6_&4^;nK3Ev?TW5=8mwK8N5lXWHgqr?f3q~T#^IO zPNw{sC*W44-fr|I`inZSOR3DM$j@UXJ7Q!ip12#^W8!q25)cNycgBPIJZnB{Ana1J=eWvjoSXL89N%sq}ws?8um^e(-O#ioKCTs9NALPwo zC3-k3$>CPnKkC>jM@cEj*ww#2`bJO#NRj@op6Hldvn`90zQW-%(hXXxNMS&TpbW8n zZUzzf<;`IAaI40B1iGXmo3i1m3zt@^FbFc?H6!0(cWx@0PIZo`GI!mUJwF&)Xa5ph zZ*v*%5+E;2z-NEFw;OO7U48VSnzdZ)5L}N52m34_E4WDpaJ2|NQQEVJ{4DuYP5~Ma2@Y9=;6;+r zM7Uc!x*rMy^(t70f>EgW!NrajE zVT7o~Wj^qd%h?cQ4s@@z=6h^N@l6V__>GlN8nthMxH^o`Jj-|3n03Ub%)q^b%(Tx7 zo~629ogh$fT#v5gxXt|4MO7r@eAQ=2Jm}uU6_lfP{0>*UM*pbZfgP1)rhg7x<_ff6 zr&4W*;R_wXCx;)796oH20I#kX*LuE*uWXNtE|cu!ti01?$l7mjP=1qR$J0eY6b7Sx zIdZbv;Hc|8d3CBNp86IQA?P#rVb!qWA{f}=U)4LpbXC1YBDr;$|9t2&AxiKOk`Aso z6*d_;z;Dv9Kf5vV%1cR>up0OnyjSZ!v5T@(8>6_628f$dnKLVnP3c)@pQ;8B@>6>& z)dgj;nC248wMs*bOr>!l`3=+@k3gOyT!7bW37}Wq*;ik)_DeyVA!MNI4dc^4-lYmfj;jHf?6K*ZDdV%mw zH}2@kuIBE6yn23G=P?ET&dB^)oE-;vw-T~gb{)O5ggNVvBc{PEb3@VC*0X~?x$Oc< zGcb{o?T4;ptu7 zUlxsRkRLJ$zSh7-FfX{7aoq;yH66nv_^dB@fT5kdoLIOMhs0~aaaO!dw;f#kSZmD}7+G_;eB zWxD18dBRB1&@2zFNXOCd90vVzP#!Wfo=rzh{5yTf7bJcqBvh%0FH#TZc;keS8lMe1= z_#@M1VepCT-!iNu=6eHz=gL1R^9P9x=|6t)SUxl>8f2m#sRqbY^=4OTs`Rq?5!Y5? z&dI*QCoLoz(<67Xk|Xf5eSbx#iiZ5a$mn{s3mLc2%5j(oB>O~#>h~WBn=BZO4}6-* z%vG6|4EfULNXVkTT8qLo8Ft)lSyXrt?J$nCJ%p+ddzvT$YlW(&kXo9r$YdfI>gV4J zyzPPSq94Cf1OAB68ZN%}E`(WaqFQsAtS=!<>7m9eO^dDi{l2G*${TIMJCZYwnfG}j zoors@&Pvy0z75k|G_y>)TI?bXlQzaa+vvFrlBNa$1+r?d%{XQSk_&4I-y~LU} zy9qtt(en`*@mK{*HdGE%rf#^^!NU5)E~ZC!F?ZX4#&bybVJ7Wbq$oz}L0|!?=d<$X zOk%#H=03M-wkFdKG;+sd$6i=%K7s2W914bJ$m5+aC-3j{D#`=K zyKG!xSxVtt{g4HhvR=ub(H!FCw0NIQB@XhmQ@ac&B3g?`K@=vaC019R0$aqd8?H#%C{V=Yk%WA}+6 zm1OryYWMLHhjl$106ep4UUY&y1ne!vB1vF-_rWS!B&O-@$Ox|0rb9 zpR_a(#!x-i7gk0HGui`n*>4NXjJzPYZpjN)O?Wp+2RhWc#C+#C@DdlupK*oS-vw;d z!XwZzTJyfr#h+qW z13*rOuRF2m9f)Hs)t48&o1z#C|JO{=;)t6Px|adJk`xqMAFY;QJFb zxT#Y#`=~zDtLG7Bq<$&Uv}?O}fv_Ne(afgiOvGmw$q>}ZRa(CLEz~5CA^KydJ~jRQ zrrV1fDXqqABiC!OHTx&`2b+zfJFRs^`w7OC(OUb$eN~cBE0mb*)AQe%VrB2>R9veS ze``)haQJ^in}Ahz(w|Rq^ruP{qYXnHM5F4hn2@e{{@!)}gjO0>=3E1zv$ZP7rGM6U zy$cNun?`*n8yU;hRs2nmIT~>?-VZahVKR6uYVp`uBQ8k7fOGQN+XbbSaqi*7FtoTb z#B_n{X8BE)s+S*a`JYm|E!S)cleQlQZ~O+&Q_@Ci4J9d(R*-(P?W)OZbi13u!*aN7 zYSV}6<)1C&1fpqpM~@O|WrB0FA~q?Y`8$J}ph8Mbmggv=aJ47s#HDKjn3>;k8%n$c zf3=Bn;2!CDgDSdO^ZJ;GeG^4|8c)v@JMU`toeQ+k3=p9VqBC zZkzO@)p>+2!h7)sCS_&J>Gr>vlGnQ{rnFAioA~FSm5imK$zo*o)=(m2IGxFG+Hfqs zcTI2Lh2&>WQ8-a)SP6MEFM8v}j-}Gf$frIs27aLPy1uC+cgY>ilt-w9CH9nwpCX*G z8+7>m5Xm*}9`ti+E`BYcC{pXYT24{A&2LyVY-DNZ+jG(tA)tY(4#JXU;_c%AclrlL zqrqn(#Nt%nXY1@DdyFih6_MoUnQrYYw_fMR)gDyF+;yy+c>D9DD7%ol(-3{@V?>S{ zd0w}>N)VzDBp&~R3jg7d5LtnYgkZH@XS{P!>2?qjoEDfR0r@@pmlXN<>QLr|r&HbQ z^+s%FAjAi(a1gBeHSsFHN$ffPtdl%}6CcTCJ-Z#I-11`c{#gpPUBQvoX z0Y%I48|87ZgPe*_BAE-|Wu!NiB8R~{`R(H}zqg7!dx(wYWZZI=P#~?h7T+jjrmjcK zmA*^^oV9J7Onrk3Ue{?fh`m=<=j8gSnNU{&FHjA#+!u^YyiK{3n;)-OKHg z`A5k1J5LXaE%zj!Woy1KHO+UNBoBx}b!H63GgG^Nb!K%Xgb~8qSpI2v zKdI+aKpcBzF}-LdsU1+6*df@KMG4-erE0%O!eT>FDL~RjrTp4H7;7zR%9=Y&h~>1o zrGlXB2kfbC#q*u&Hdd0I53B^dL?+Q##Pb^qYd@Ad=MuhU6Z*0MNM1`8ImG(?=(o0r zuAn6fGM}5EcV8vm(TlyA*!p!nK_5cvRe3S2XZU9^^7R1jdwzzjIvKi|+Cl z)jWOpR^KAU#SI<|DMOGZjQc>icIC@Gk|qydE=2@j%W0S3fQP&!>?_69?UJk%3?5hV z$P1(J^J}W%Uz}8J1t!RM3ae&*P;#=w62jaVh`Tjn@fsy)mc*a?p#HjoKmdNA+9_O- zF=kKju$R^&NQCRp2iu6LWy~3Qa66X95euPBEwdtt=5_;9W5 zS?8LzDHd&Mo|zGUjo)|7Uxrus6e*x%Flil%5YN4gle1;$vU!oIaRJ&st{!Rjlrin}> z=BEU5C^5bL~O1}55 zaZq1OiX-rB_Z=Z+41IThdnG5X=&;?eb~$=C-bHsOgxKcuhS7O)1pGLqKDDxT&*ozX@sO|N~wOIO}?OcZ8=g< zr15O8C5D~wo{D$2;&(c_%!OD z2x5eXv*~_y1~)!6tBlA6sX}Mw%~)^5uk6|Erfe0;*`kCnf6N{n@sT@3(1{dI`vkc{ z14ljr1a;DUW4z1sF)v|O(w(+6QK>`+Obl6MU8}LkiobVs=9p-0};VMHEDBz~bw4Ry# z(fli^z3Nii6;AVsQtQPZ7H$O%Mq^s%L3S2t{Alt5T{Xg=+?(` z0mjwa6E`2(UgM@3-w9`RzS|QY*?4hSEbwhRy?_T}{)&2WKGD?KB-aW}Yws&Zb9H0L zdVemuS$k2d83d9Mx1YzJ{pGfuE^ERxCL+05jO>7DK4l&MN_bOGeAh-HJNOBDP9O3( zT*&=o!lAejMBZ}+)fSUQY(7h@ttVhiZy=r#Cd9q$V4_)VAfK1`_48^;kzTTK_<^eQ zXP>LW_o=9|Kw38V*3;qVU*b;Z;{DsTq)eN+gZ9D1s1YZ%uR1oPo>{B`E$XIF*?CPCiD?CD??GE% z5c|p0zs^k$!&N)7e{WXx&w1b2YZ{uxThUCEdRfxMR?07eu?&kd(lvK{MM?yj;}#Qi zd0q=AXE-7Vxd5LO?wG1s+Wd&Br#t%tqHE4B^g;I8Bm(;;RYY0TW+?iGo@av<=pV;! z!O@;ME^Bsd4(_pRUZkC*HMC49o)uQ}NJU}?DoR}R#MWw^_$-w~LxiA2l`{WEE+YIECrI}u=5q?PkY_|lEOlpcscrD{YKi1=SdN3Jk^^+!X2tv{+Cj3CLUNuhkCR!!w$#e)V;x zJlX!AyM+SqnZ}IiZe{d8n^g=+Fu=m8ivO|Tjv$fT)CAQWuk?Dui7L0&0O$>F?Pswe zr_qjOUTjd~Y{_hCf$ZqS_jO0p$57o^l1;Q?+b2hEWfv9^l8#iT%JE5HG@z0mp+6uS z8j2&cs?$4PWp2TvZ)Hg?(=b(Pt8a1LpG$aYMrP(<&kA1ryh&rrK=5Ne)T;%4cF@Lr z1Yi#}eKb9lLho*|pOA6kH(dhj7Rix3+HB&di1#OmIUm9}Ciq>w+#Z_eC(4PQblsME zV&oa?tB?Soa!L>tOJ=3_*Bn$lbw)RxYV5$D>Tm$D37RzbucYx{rDXvpmBpQSpHv-6ZdnD9hTgYx<^&TF8HBiL86CPJ~IyPQG(kkC=0v)~gCbbM(L(&Dmgbzo4Mc7S{=SA>Rj4!dmC*p&Bd8vuN zQQLg7691wrnxT_RjC?#RO}qjqSY2l^Prr5UpW$;Nb{#h)wD|iL*}R~Z3xs46bQ=m9 z#D|AwoHXm66DAozsgwjXy(tqn(+d1C85gr-PUh}lfV434-5pOLlA?%5^dMw$>}8Y? zIlpN_3tb-YQD~MdnvyqMM|hhCez^;}e9|_51s{ys(kbt0R(h$4cLCHyr~KCozb5qv zoO7w7u9}OHD`?9zk4cBk^OcX}wzwRPXw_43{qX(Hy4{+VBwLp-8npxN{doNh0id$! z2=d`x&)KK1xSdw&(F#Y&3rFFJz7N}YuArJ=!9}TM;>O(^BHj#`Lc*h!aRkk?OnKXC zjhd>MJM08WzQ6qiH%Tvb8%v80EAnt|g6kUNJHp)T51UW_Xs?Mn`zr}h50eA^OM2lh z8Y$$b9_#by1@`kld?D=OcHRb(q!!xlm+#m~+eDMoS4!>NaxKjtYgh^5dM}VWb>152)Mo-QT z6U#w9f9ZgHV~^T@&l=YWv&K=r&q6={c}xtsIf{p^ z_B8MiJJk9Ka~5X!3<<#}PHu9|en!FfGU8dk#5zL1=ZN@^zo5NG!jyZRq`7)k%iC28 z+`eQ!!=o_l24Wur2zw%+VVR7JTy|zzPZ|flBe=hK)&}KR;OHk~XyK2#D@u4I@m@*yh7za8=fFJy zJ`WzOd|bO@r00C*8Pm9ZgY;@{LRY&@*P=R*JMEYR`Q8$?9e@!)>a$<8i45~}Pt7HX zzus59hKj(;o$a*#q=xn7cOI~DNgeRJuWDPocTzo1o^5A!5wx^;*riiG{cXd>-Bv6d z@_Qz^{POB{GhI+K_-q7JR?Ah4NfI8B*{8Utl~`7RplYVBl{>eh&nBuC&A2Pv<#7QA{D3*QPHye z{=#EmT-qo<;Q2)>$OH7o(H#CT01aqZ`|r$vYMBZs@65>zp+wSg`W(sg z^?T@FXYXXYj@rM7ODD}E{WD~LuoG>z7qO^Dva|SuLKP4p#Ay9Z*qrWhelgGaYwV14 zr*{*k?4XcJhYfr>-cRmx(rKgOpUdfV5R(-mHt8FO=h1JvAViqMiQVo6kneJ%i>3)y zH*Psd2u*y8b4RSW{{i0+tu+8sK&-!fg8);^4(shTQj~Ty6n%mtVX+5S3WY5Uq z13gi3-&MWn*|zysdU=e^v}}txs!UT@$qKozN<&=XyF34a2ZnD2F+Qs9ypuXyb%y9e zd9jRFqRR8`E)xXvPgwGS?Y9IB8NRQaCM}CtUL6S`ep)ZcANdD&JoxsfYE?#5sTH2F zcoMQ3;-?{dZS?wr?6%+E|&v#HTQaplrDFGtpd&W4gg`()P(UX=B0&Z=2Z zRU6;1UOnHbv2O)gGDn~~5BrP@8nu-6dsqe^@ne>FOrh1L+VKfHN=-47L(;dTO-#`v zhJ&Eg&-=*xh4h}1_GN}PH@_q~{PODBk;U5<_SJr6uftRQbkeiDf;Y9JmrRZLUoS*# zM2G$mi?~NPXhN7`2Y>mE1$wS|{A({IN`Hkm(#YOU=lx4XprVxZTq8WIP^pNhG^=2? zjQe%eGdkX?U%4Api*f#)X|(l*nxeO5)MFj90N$Oh=^J8;2`40TXi{EII;9fXao3ZP zd=;tfaN?K!^t|4Mlp(F2**PeB^aZr=;i*9N%^OgvrMFJ?GdDv{GN=#Tbi-@gw?xC! zo`Q+b%D)zSlVDD~UD)2ct&GaC7&`{DDYtN%)7+pkCHKerl_WIS1{nJOT#yT1Y%ylc zI$Gmaqmk{TnerUc{TW9}!T3>;@Gdpmr(|0c%iSNowElQmUIZG6^g(gmjnb_Y-Mc}%$(mKS zS!sGqN}@doMpBUd+QvUX(ysE?`L6!{`YE12>M=;%)V6x7dLdRF!qtR#@igj)7n?9h z(x36erdQ+{8^wYz@LfA)w~G8@qyOAat}LQ8Mk}6k3OV)rt$s}!dGSV%AOlhd-YJs- z+n}fTy3cqrEj+ydmfH-62T8|8!u5rPJMo`ZG)c7{?2k|rpH(-pP5NM)jeqoj!p?@00yNKKBaOcyXFrWx z-X~LR8=0A0W>9vx`MX##?Qmfr!%X~uwK0Zc(Ds|A<-g@%5`>vaO4i7~rmT|p0RYqZ zKzK=ObXDf%8TJ+f(_Kb3rYyb3-ts`b@SsW}F?eT1nJSI3X@44N3TZ$6j+aQa(3tRw zAU+KSNcv;S{+(B}0Bo;ptqE^ay%v3=Fk*XVb0Qhdn8hIi->Lr3?CisWYiP^3= zp{mU)3@VaYje2j66@zbYTU@k>msxuq(0Wsffw7WSz5I%*&~Y0@&s~*|dOWWdL)$is zeVSOXj@l04yl}#=o*r{1Xr{0TjZh`wnQ$z=$NKS-8-{=Z*#R=(`qQmC2VwuDZQL$B z?kwsul{%8s#EEMNVZId~j(k?AaMpJwjh_v1sC!|PUGY{Lx8fjd?6A@VqQ#6?GR)b= z>=!EJN^|Q5xEqv5gxR}_Q};p-ZTbgK_qs(yYR)ht`R$>(!}-T5{ca81fsmUe7Aw?@ zkz&5aSb3v^dm#qlIWkUyj@Km4KSv;iVl*#J0^7aN)d!?ZH_U&FnAz9~n^qQ#T<;%W z9Z;n%^m94a8mXexUkK6mG{Shjci3h}w_AO;1T%^WFaFs6ymv1`%s7Om!*w_lhG%j- zyDYsPmOPx~vjenPZS4Q|u1tr3QS;DV|5~ZUTqhQf2#H*4l zkGtz;+r=DNxQBeV4(59QbA)kbNgC5Oe|q7b5tfRs`PCS;S>UnIwU zYsVvu^Uns@#yO=qNdA}%`oK>%;&;JUO!iP$vW-*jtQCBw8c&J}uwy&XTx)tBPG#1h zdY?Na-v2}ao96Kzoxt~TuZ2Qn*cSnYgtEuOJj-nbf6v$N_X!wWD-C{ZHK|G^!^?eA zA55#1yek2_-Izlc7}MC1pK~AVe^u07nSuMM7=}j?R;r817&IG8Biiypy~qXl83R$h ze$T)-^8^81bL<_zg1_qxg)*x_M2q>rdnQF(9LMLSIYJJd%JO?YeCr{o0+wZc*kh|+ zWkoLj;|f1p2ebV_+njjXB*@g!^Kj1xW1mmjwl5duR7mBppTtOePkac61!lp`$EEGK*zj|d$- zJoBVSzk-q1cvr2Gb8m5d)CxoIvSMi5k`8=JNkJDHEInp|cZNqfY;v#Y)}=0L*EPktmxeAe{ib3qDR# z1|K9DfL8%|m0uhJ;@HXX87z6GvcwZkGd>zbp=+T&mJ>ZKhLa63*t=YUT-6a8JyA9Cy45^J%*HK# zjZ@X2YY?ef{Lpnj4LtGwabn(y77fydl^ZZUDLGixxbnWdn&6zaME^-NZ#?-^oEY=P z{@!GuB#>k|&#xt^hgNb$rcU8XXaN`4?_WOF}l2WhAFdKW*?ibW&TAaU*5+k`W zr0U6BAI{K9mnic$99$a+Qxw^2^K2s&zZO-?xICT^f4m;I`KD7v%U&`5%?u?}8nz|4 z@Fz{lbMPJb_wf~PPC+b1BXr^i9jFl4f;a{+#FcY}W_jchvO-L%Q!TP3*lazc=Iy4c zm7nvLv`Vb4@cOmolOd@oYv%wve;%Vb_+1G-n>+GzeX*rnXCKFew&i&y+3sMJx1;<0 zgJ?dM@|Q_(rWc-e(mY@UH13u#sW4DXK97Rb>`)W14cOIjL}vtbAC1#v~43~0Te z<8*k_n;X7fE9&_vAJM;yRbsPh-8hNlBNDJAGg}c}w#Q`QtRbAWK4>Cat?QhR(0ufj zgkaWqccA1X9ru(?A8%8!h`W^?q%?{1@(L0@*j>qq8}Wd4#eT6Aea7@eHIW+__e@!y z(~5@O6LC}8Wu|@&^ysbtIawszjCBNi7d5_n+&Uzs|>!4e^=^aFZ@ zgfpafny&q{gcdtHYvh%AA4o9ueEK9{G$OA)@@Cy!3oIM|hJ&-K$8yq|d-rUoWopMv z!6Eo)gid@V*7teb22Bp|$pZLfCFf^)WqTK_{RcQ~v!FXt9@*(-Y_HP(J>+#rs>2b| zJGo=c*S;{H|JXI0FW_$bJqdR zx}FnI-1Pcoh||l?lGvumo6#WsTkVooG8bBM0gdgyf2omY?r$`=n*f_S&zpx~uqz>t7Ev@U5Hvfrp_WI*j^^%et zYCvZu#nJy~Fkj+4+>3PJa683^%2=T2X>_m0jQ~zz1X1N1HSQ*AG9r=1zanps#V;tf zMF=t3S|jPj=dZX#T8n{IX8K%VI;R|Wlh5Q>`Z+D_0Vk{9oSf|p4}ZCsvl@{Kvo+!E zBNH>!@%3-jW`S!C*#F1ad&g7#|NsA-jvOP7P1bQvQkjuGj!lS48QD3a&XG;BkDVQb z%oddq$H>SrEl$d-a{kwxs-v)wR*1p8mEY$g|b*D1LL`j>w;6n>3LE7Fz3F13O zm_&Gfrn{cQ%;=5e)7IO3)yK?N8B{h38F+WYJuz_`^*5P2f8^~itE{g6TijHKo_EY1m*4 zF-GckLdY5hT2$vLizk~Bd#pEv z1=tjzAI~@TO-BYzn=V+Od)NOw+r_>_`k&grC@jr2f5dbTV&+_+$|V)mII>5^7AA>r zKG2b{lE^8dI-BY7c6B7Qx3|-N5z2VQv(KzLtet4-XC8>jPrM9bXvtAkR4n6yUf2@S zbrFWI6PYbm31kns-%I54QT648442~Ao9pB~dnzMDFi9Yr=Fg5=IFE`T?tAajF9Ivc z%Ul_-hRc(X3bIR|HP>4R&}BMh`guriqlW+5wuq=3eZqe8R~vb~=&gDec9tZI@$0tZ zLM|xJYUeO0$eYCB>@XS&=0+H=jYas`;p^}5fy=!@BixjC7L13Fwj#Gy=vpl>28!9*@EpxC%?Ogb#^`qF40(B&A&?R2UV?S1D z#QWdKm6idzB_JoeKKAk1zx_8v&C*e=J;lyevX@GZN;yfN=GnrL#~4i>GzF+ZCCf5> zrE>@PX-_a%OgH`(Gm+z1OSjruns$LseN}&RBm~5Y(&AfYSr%fUZoF2YtLW(P#ZTER z6g~!;>=|PAqp51Pu8ePx$q*O-Q%LN_S|O}Ue|WDY7DCnBub_RJoq0y}tz=Bw*h6r1 zz+ie2j01|MrW1G1W(o(vIK(U;oaZc2Kv*|0Re8%Ky#ANa(EG*2^0We(oaL(NkMxU7{J}8GI-l-{9@K_Be~01@ z8%@h8s)-ux_`PobQMX09d0RY&VpMJ6?K*-<8l@BOEL%jg;InWQ5E}$dK88%sMDmOA zKj=Lmfn~AFZ5yE4#Mk&s3S*Cw0~rX1*Jpcb=GuUsm4~PqASjh*%d(mFIoyyg%_T7JL7eLT~3K2!e0E`hzGq5$+ z2=j+KVKvdYRTRk_h#T$lT_31N2DBD?cB?%BDepYGXJlb-?`yluMg-OXRi$R(+Kd~3CyK{Z2V zo&+SLrB}x9W1@kn6&!!Ep5D&&$y;~yQ>+XTQoOW<5Ne@n(#XwOpBA}GNdF*uKX_u$ z(Zeeecz5k@wL;~Ry-%32WtS!Sx?Lin<@(ysyhYwq8D9_&H_|cr#*+Pn@&> zkGBQ8TT~aTf5u%(eb!6(g?BInU^ZL1X#kbEjn1v;Gc`f{mo~$g}C?BA{i~Y13XBZ=i zjStK3f8<9e@Zp~$Yki_L=4=(G7|R2e1CQ2Cfs(()NUW*@#0zjz7{n0&Id+Kns_5zA z(%II`_WTZpB3Yply+nys7^j@^y)lpNhr)TK_+?^nL*d zXL>{4^k-=9p`liZX`gY4q85dz#DJ9JHtL-*Rp*VLHWxKTt94zVF|UNrt&o z`F=_WWl;EnyATGMRWrH#fu=viKRQTg#L!Xj)LQmr(HWE$xykjU<&yJT=?32@7%Isb zW0WM;6QzO-rN0$Mf4`6Y`H1=)RIEVCr=}s#Lb4NPQj!Tq1$-U=KhvkFbmtWxXLKck=l&L>l8cmf<(RVpm z2v=EF4_f`^)~|K1ZDdy-#h*UsSd$jBt9tQW2Htm^8Acj0$-!St4_a%%k`7>$7POrV z{_{yX&*<@|vLG_4}@&f}4O@SO$y)f{DgeiZrd+>+W11g_M-$f`rS@HC# z$0%?+86$dhnK8gn{%^|lGj-hYYW?YIi!634Zd+PKLkCC;|LqvnPc%Fhnj`EGenf?d z{>-$O^7(^FSjfC)39_F_RXHN%goFzzlxMzKlzS zZ5Z}$ElmkU^ERfC)(;6f`~l_uR5M2W$Rl4Y{PR|=t;<>o^ulDeBCFi5Vpl!onu0l{!VAr;_A!}Cs~-P(G~@wrwS6bw z?SCE(8vyV|51kB1p&`;fVTf5oztbz8NET?gA=OfW;%p2|R|FL40vhj7T2s1(YQWV_YlDZP|JWw%wgVtI5BRSO&-=1+_nNS&+deNrJl zCojZmqfVT;US+CX9({73edz;}z_l29H3P>;h;sU3w%_w47L-ejy7|{s>zB(d^HH26 zGJ^YeQo=S3r4Oku%cd2M1fMw4QgU7wh?#EDradaXDKckN8R4Ka()^B>tU>4?!eZ(W zxb$&UZzcZ1Xci8tnZ6IY;-L(rS6mj4T=bmviTXmmuEaB~`)h#Hn>Rr4mXwO9sgatL zGsXbwnJ{n)(-W)r)Ikpu2siu(Ih8|07CpCO35|^&FQ2INU1D;&R7=1f<&nt7m^TLQ zJmHZaOZvzw((YbF|9yt}@abZ}id^zvC6T#qjbyP)GN_i`EmX~`&ZWY1&V_k69-pIfDgcqN>5(C$~8jCZ}x-=`i@=fFUX8d}V zF%Fuq|DP8_zNDyrI?VU*zXFDUklU2_EofQ+=)qNrEY&<)H+0USN@^FaS|H^^_o1W{ zffqMGB1Gi4j+QMe&6N`IO<~(tN{Qw$iP~pF-@dN#tJ%7PsTmH?pE0kPWAC`W`8i|u zGv~H|ix@{0s7biBquA9V3l7?jWDPz_1oK=hg9%kFHw!rRa1f0pflRWb?O3HSHQM{p zMgmk|O`N@&@2N+-U}{Q`h(%@L{Oh{*s~N23u4EPS)wK1q#JiXQMEsWgl=i0rmAFQ; zW!5uS#!xXA9SUm++UF&mQXcHj7hsZYc9$d~nG_f5P!fpDA}%I*iPY$D9?dF5XCL{L zV?l1#@sHuH522O596^6|@N`@b{_TzUWBTgKXXWvmY+N>=ZzFuwJojInL0+PypNSxL zgcY{C6YW}^W9Pr=2vO6!Ut)>$(%qWS3?rM|JoCG6#^W`3Px`qd62fnEQ$@?hd;sTj z=lQFn-Bm8seIjhfeGU0wTOSvXsyop zKcZ|k2&I!+KP#o3184AlASVy|gFS<3Q`HgnAULT%VMT;UP;|rG0N$ti>F^M=)B)xZ z2kZfkBn=Q3YB(&1>M6={O9R353@=}JpC`$O6c)-DDlaOV<CNP@>7mPu!PvFB!@U@5UaF#ekCA4(360(r20}gd`l%@MOm z$eK^Q+xp6%gl5;yd$XcFrh9QOG;cIsRI5B($QU*|?$?>GVN@wr_Q`SV`C4EzETI=Y z`PImqHcOc?J{!sJ+q1%*eA8lTCX+uv&Z0osXbSJUZQFWrbN2)umP_DTkvgNcgmo@W z2PY8@$Y+CC`KMV2^|P^3JzsVrwJp50Wlx@dJn_`Wzjfno{jub5XEwhT7l<`(H?`W~ zMdB*QpIe;8AW|_Eq4Ww(v4LBH0lo`auz)uF=AM%C^kU6_gN7D4XduV=%>M}*mqUQj zR1lUhgK2kS4`llf7bT?Rx|w*IaIH8O8(`;9J0-Zbp%_w3Lbsrc(b<45z-K_?{OAqC zKI^yd6;@#AB>|^bJeQf#ZG3)fgNl2+fSy*h%HYDlnxDLs`^N4O9>q*`v}@?6lSAiE zf>^v^KDlUxquFkYaa_AW>@it3ts60bV}c{5LVoU#A%&X_F236*k*2BPW>~wJdt7FW zDnwi7OYPzTrr`dt!fE|u@xW{k|A7~^=YOAwfgPo-GE|209%oI!^m)ka_<*x}+L?vL z|92&8WF=QtL4ZRoNBH_@NB75FY{qYo85%1S28^QkU$%&UbU-a-$K(0DP5j0Ceq3?p zE?|^&_`4PPe}{^y7G@DuG+SSau!VQHzB`s|8uYsyTW0xN%^FImH> zbrC9%lv4@UzOVgc!JzdoeJ$&OudIJw@6vQ(chyF=?;?oHaZN9~)-b0e?7I4}EN9ke z4=L8E+Yc!wuPf=nL765mRM2DQV4v65oaE-Tn(_XWF|8JnF@W;eZWl^b0>V(`#^qV< z{zKV!c9&h9*G}sbsW}c_tvke}s}z$Pvy~~obhIaCUHEmMwD0bAHi6;f`X~PC$>q)$ zJQ94B?Pn#^AMd38XWkZ0p0{;oE5G~C;)q!{uY}^0tpi{yo5qaxm^M}c?1RoYv>yB( zbh1kO0(JqYS)nyn%?|de(q5}#SDHYTbzk+K3OMMHV4^2`8^wS=bcl!)SHBLqQr3P<a4Vs ztpHdp9*Ul-I#43u+`47vA#|A)Ne&-31-U+B6&JeN%9(88yL}o3#Uz?PXXlUQ+Hb-_ zhKm?{4vCT)XkD=#)tHfkxiG5ckyL{x;2?#;tE{j)SY)-oVE$n#x@~ zTU=P?aabMdJ5Fp3C$>N)UmjvgtGn^2&Q{vdu_R|~SxDh*p_;~>UAFq6c=4}oIUg>- z?e$Nmvq+NLWUA9PG9bsbiDu89N;AH41xlyEyexI9Jrz4tRdLYe=e(@RL16V7a?^zO zwpT`UuTulcUSQ^wnQ=>sA#@w;;R_3Y1_`GJQAwQSBjo>!4k8GZo}r)3B0 zK92O2qgc?*`5Q-=qRO?S_xJF8>c*|GBLSe)54+cghHycJ`OL+!rQUZZ139PVIJNZT zC{0P@rLRRKf|h^wD)Nrjm7uM zW&=ZaPQ)AS2`o!OZPMn&Y{Nib`*o#D!M7krS0f&%id63l_;o(vtP>}J}PQ>7-EDv!7yS|0%MvV zo@UC;(>XBc|5iHli5hwkzet>28xIaF^CSE!cZs6SlTkv-%+Zn~uemhoiS2Q+ zh`FOD?acmP@u2GCC;3_4-&B^K2lPkeEcq>U`o|P+y(c=aRQX(@c07>(@MGVzoa!@y zro;X3j>Oo@)gqyIUI&HFz&x1ygVMlDe+Qf4$6NkpyX(3EC)H1cdySRoHX3`SJYqzb zTi7u(i5`-(BdlBt$|7C7VWjz`NbbW*q5EKb<=-KGMTK28EHkgwqWNwYmp10*a)_%Y zE{Q7(cA*DpybBN1%#%{yO?TWV5^9g#4eDesaV@KC46Qp3niaxuFEbR(jO;$8=Yks) zDMy*?*|t*KMvLs975)D%M-2rj@2dS_rL?6qQbVad9yP^W0me{mQi(fLuZVyzGaKO< ztJ{MiI-AB!OIj_Ei^;rUa$tiuTagCfVC^*=k!(Es$`$;F-FE0Zvhh(Pl5+Kpu<2WC z5T%9YRU~X~X-5&(u(6g5*-O;i`u$d{ABFPh*_tW$PCAODbM|zAE8~7yS2r)nQq{Vk2Y$*;~C&rRCp)*2QG8e{q|Yu6CnNN%~Hyb^cgF@ob;yB#JbAWF6%@B%|{;9 zpY*An=?d*x=>)d&jy8L&EOl0#@@#}2lmxCMedWx!xx#UiDrg+rBw;?C?O6At?)vs` zM(Ja@=F03-YO}M$))xXf)~)UAacLXAQK{H=lj)>W zsFd6j<=I6%-)c9;p>yRdIaWJOJ`wG?d)7vBSbgv^**=yO5i&b=raz}8@$I81d>}rj zCqq{J+ivy061aHsIWUiu6J!4W{(dhmm}1UA50KJaDX{O~d)szo-wEBoK#6L!3o#7Q zM`Ta`2+pS{&;3U6tt|K2_T9b#DYXGINc6^Cv1Gtan6b9s2@#J!FB@$_o@|p-3@rnU zi3-6>gQ4A}D;aNfcGpw^m2~Rr<+ZvR1k~+=N<7dc^m05H{fX{b0n04TFYx-vMwV*O zaa_=gR7aZDg4OC!Bhc&QwuNl6!zi0-(AF=GP7W|N?imoqKaB*@YuSu#aUrDO<3VR9 zoi0?9?Iyb)r~V8#pYF&29}DlJjo1Fl$97vF*5YI*&BAi!uX&)(E8GInV278bunak5 zYUOcQ94ltfj063 z=3)ii^`|{`7PoUqu+FLxVNN8{DpG(KT!n*&iDf2QufQG%>3HZpDe{_pBs%xj8o$tW zr}A=JJVU=mQ1gLrPcg0h93k@H;nNGJ?Ph4f31W8hne1vDz2j+uNZ56h!;nP|&sY;p zd~4@h2D7G#GwrE9EzCPv-P!f?5_Rk&L*;<8T*hX0@rG6XKUyXxGUGwJprYrWsVWrs zE3XOoO=#gH@NX7J*MZ{C#rR(68tcLG_ zz%Nvbdo$^=yk5Ja5K@Qm_d$Fvh%dxAY>uqlq(#tWKbunp(58_W>TDuEtga~o3mi%k z3$iH+gORCTp%mQt4XL~{oeLX>2?}e=)d$xnzkYnqKp^VS=BR0YO8h(7$XM~i;S1_& z`Zt^-%kQR^jz-GI5RSewRv5=@^R8>Pkxqb+0qMo>*@#-Dv5 z9@2^bysP=aJ8-axk(j^ZN%zO`e#CzI?W5-d|0-gY$%Aq}YpdFS!vTFr`b7X$2-St8 z06;dn_y2HOV*wdPoA>jgMAvu#Ie=A=$|!xa$VGD)Lf|4$-C@snK|zZj@HO7%|<>%?e{yBT%B zR(r)NMrXeJ*S&`Hqs2F4l#G;R4ki_^Kyr@A!XE^ff6hjC>9c++J54yx@9(y?HM<&@ zw<*V?4dZn$aTV(_687ws<_FRX$y18aMloHeW7;LU*r8~6TJypZnFU`3)h2Vn4U$Sx zltJzz!o^iV_wuEIk)VA+ zhB+xxl&8^2j!|-w_$k)VzO0q=2W;;U>vb_(*V&m!nF0B2=NzYgact~GO}LoUb&aF0 zPXRyvvu))&uMG_uzZd@p>Jiaka3KW!dW_Z#I*JOo`mt5UUGh6nhNbU;P`W(&4Mq#> z)`KNP%4y<8Zrrg7v5tg%vn_YTgaJP=DWHE##!SALi~B})z$lNpfXH2IMljNT=)fk* zFPTY?UG0-H57X7yTb@%$7v3F|{gB^TqY?Zaa($L1DV8&;JBt_95M;R%@?I1;%+||- z@-g90N_y=GVkRUGg04AoIUR2#>F|AZx!u7EOJJleSNy5NY2(^R<+bo}o>K>~e|$wD ze@f+ykCNE7klx$Kh?gjT4hbLW_TT>)(kWJ3s4L24OeR7($33+$Kse8rCdfbU+2#-L zB6bp_)4mi6JMR{V9o6_GCa{*j-DJYHw|bT?9Pj_A&oV20{431m&Lh}swJsmr?=zAZ zjvqPw(9}x0!b+&F&{2oi>w#TZqLekU4Q{t?4%#Ow{7kW(uhY0U_2c1*YSXc5;v>JJ zkMjcgg;PO$ra=#p-;tkKV@C)5WMUW^vn)q4V%lHshzJ6Wdwe$xS~pQVWp$gvY1#*| z(yO-q`1J0GIgg_3%1v|;zf?)I-)PYS_%&ESBxcly%W>QO?UE_2!Cddk+O4<+tBQX= z)_3Hk`kjWFm;aW2<4kE!A$RYNgtT}-CjFE38y>mZxDNO^s@)Coa~y>M(kxu+yaEVD z!B!1cuBH@Y=3p)^7%{rR7-&GHz8C7sn#pcvqI368WlklRfto=_6}W-@Prh+fo|X6L z=O~T63XX<6ZV?_5ts19qfVCoUT3Di|vRPH-D1NTHc%>|)2GlqbL9gc2=;lRC&8nBk6VAb2Dlm6_UR(DXX*uE4>J$zr zi4@{)y3qpS86DYA)`bBh4qt&S%bo6xqOBsi(3LF@5{hQ{r|uf{Q!C1iSlyiRDC#A~ z`heK52V|toXO3!tl(P84Jlo2;ax8?B$qI4l_OtED&ezhc*{(ZvS}m2asFwnBhH^7u z&{C!}|KCZa`E?!==o}>Hiw$~H>euZ&CEkyElR_kcOkYr%m4vECoj*n%$3)kkuGWh| zJ2LGeNBrbmX7u?3lAem@%XSsj(*~q0*~o^%xs3Qe2W~jdjlP&W*zK&DSE$6*NUC55 zxBi?-?u~YzNmdeje6YrWsBz>}${*(+6E`$Gt|InUNcb~;Vbsv)fB(xTcMjD8#XjL(mqBMAytPSiF z+l0GJ{-nJ{r-Tgp#b!FV@7WH$5;PTXAnNWmY~_=MiGx%+uc4H$Mcqggl~@Pg(h}hb zB+jr|YJoxibR*yB=(yOe@X7#zNMJctgE3f}#uIjlC;K{z-902u-LgcP-`(3uL&XpC z9@!fiMp>Wq>)GnvD~vX{x9@x5Xa0vd5I@i9P)w1(Vtg+I{s$(meyd+6M~bGw@z~?A zi(x@pu>^*pd9s$!C#zFLy{R<@%Mh!ce+@lF06RyHcHn0_G+byr`#60Yc7B^iMgYFb zS-T25UEqbkf@@Nrl>N0i{$=yPYA+{{w7i|<(D{~)OJrw+{XJ%S$%y^TYKqFYeGyS~ zIe-13&3S%~K5r4qq6Ahyx>!Mr7?pSXE_0k!lu zma^Ra{8cSFk!IBzw#Qe(xy$5_=FJa89+ng?|K6$uw;OS7|bwEd5v}S^oZ|e+1)J67UZb z8e0**rlyL!Teq0)_bTf3b=@C{P4hG(umbW0)LWxOp=lNQZ9$dq{F;>p8lQAif50f} zy9!3qx|Pp*&yIULRB9LAA!*?);3m2}^6CF*m-KPB^VEfhizUqr87nO{jU2PoTOlLy z6B_ZG{0F4{P)cNIk@{$l4`KJt4{Cn@5=<787)W=@{)mihU8SlWpyH~;($(LRB}(P| zFeIG{sIx^!4}KJmMbLKq+|i)D@lL?(|6>6VxoR~+sLdHZ@j#gvs(tpmDHn*()V()p)V(a_%wdoeBBjm97Qm zda0|wh$%BwD~%&%u|@&I^JYF|@Z8((Ea(0Xq2mUsxBlLFBQ1vg4tIL{J?^af)cuFo zp+V+#G=I1FUunZ8R>w9kq@g!p!-GwZJ&aZ2t(xEqG1cXCio9tIq+}X<-kX3W(c`+@ zqYUWY>wqo>HK%k9bSBnDa8vw6PHui~`ScGJs$yr^YuCi|@AW-!wRw_6Q8rycVvtO{ zL(}?9kzPfIago83q2H_)`C^CK+n_IJVj}Xctxs>F6sp&TcuBE@$>LJz>R4QAxsLZH zyWs zy}Q=z=I8(XhRRaKwGrj8Y2oib?{gWF-NwyY*wB0%3~8`61U*KcOILB{aKt{L$Z)W% z_(ejP-uNzuO}pZ0b8312kF_6UKlHr|5o8c5W_H*2?BpOdf!L2mDOuPV#z#qDTUzls ziS8szn_rpJE;wDYchr^iZ|=nnR)%9l^tt@VwrEJJlp$T0)f`fKQ0w!!6sxN+u)Gd7 zwjI1Z-@tG&aC(NCQ}RMmutcc}3^zk>T2rl^)}&>##(S)6oZ#<4y|4dPNMA=ZfrO`qZ z)j1m<*(qYtX~d)Yvm+=kSZ?dmzP{C-a)JnUj_)>lxAtUzJZ7_O=!%Uh0OpTNeN%b< z3nNrnw+zeF+3A6>vEJJhRRwrIwbhW1$*}ZxZ)NP50LL7>@=f_0hx1nfc4wn(s$65i z04DLFWEeX3D7x?f29T*_`^-=I$Gub3xKhJm!EY6?pL9rCHJq;!OZe1a+`|>5&wrvckYiv35Fc{?303EI6@+N)iX2l`MAu^+54&Gs?^i*{~Vj;C8)BX2igiP{tf^1 zU~+W_Hcd30Hd8fMwch+)mKcQYMUTV-RjDMX1}pTzx42Bvc=RB4FG3{cdR}}bIul)> zt27!OrP>nvAqep)Rstxj6}|1SmQ zJndQ%Ee*sk;z6il21XtVx>tH0-q{62R-=t*yvw+bu8n;)%7bAYxU~?B7)q|C@kdct zS}q*H=*f_pv1jaCVLY1OAf`moYLro46LsPZG{NQPR%;8nSw?)O7p98-wTNi-_EiLw zJg}wlLx2?X1 zcKdbm@jnL)6^jyl3g-VFQ);)SOtf9Ul2#sAXVHW|nt`ZY)) zD4x!YR(IhjfUII~y3{XZPkCKvbM56$w`@)a5uC=6`}PphPy}HB5{6Qjp=P2IudeBV zE#cUytGGD&<9Q*txdmDQ<^(os*YhYUhEX~rVCuALSIpwgSmjKI5B|zLXsD-13k^!7 zC^2GB6@YGxjQ!{t(Uw^*_>O(FXdf@17p0_<%b-|xN0+mA&`}78Ueu+r5W*}TU=Qk` zgm_-(o9+C>sM%VJ#Y_*wLt17=Hz%at=$GU4-o)GNHbeP^4G%xYW|EaC&pPjF#%^so zI3GSe|LARdZbUK8D{^18;D~l`CgY_YOyB*$sk9_l$ANz?=VjypF&_;cr{7%*-Hpmp z=Pb>^j)=ANrv&$hSA;V7$HG;%T37@_vzVR*wm=4w<3Unz_QZBwp;NEZ!$l*5c+3)t zI|kU_OA_M}Mz`B_m3$C(AFnEW>d(v+Kfj|vqtbJ+M5`FK_&TZjypLoJaazM`tq<}( zIo$HH_kPh+x!EP<0p>%^rXQ^I_j~kiU!ahwBAd8lE1}YDh*#g*iap}MmMr|yR*`2h zcGMcb&P>vaF!?qwmp_xHqJEMAohxUsAjSn(Pxyr7nxBI6Xyt!q^(1n}-;4A*_}?hs zkg0i|@l|a>>d&+G7$O=b!+m-6TEc@Z5xqw$l>Yq$$3mDOWq~HY>!2a9*{gVF1 zOmEGL>2oLZR;09UM{GnNUEbF4p8@46D&vDOxk7Q>KgNdgpO4ui$abxr{3rEcRPOjQ z=|F`6X5PH%KUpD&?Iwq zM--m!DYs#u^;$Fl3mBvcrhTaZ{N@l5ewX!_6{KAWc>+{v|D9udQ+@^JI=I*KqYbF# zut(#)DjROOo+u6Y7LQALsNwkOwd4GHS+B|)Ui(@-!MWI<)t-jaTq>o`J=t7-8#8NM zM~%0c{P$Jdl~|T~qz;}E9tF4)Bd4X^=dec`!jfhelYRVe@6(NsK$+cU6chcSVG+o6 zrl_ft4Ig1CM)|M@^ZY(j0Y2^8r=!_*c-p2XsbYp6X zx_yQ2sNIclp1is9$>2Zyya~r7~8ly8J zLTiZ;zPI_|(3`zOvb{?(Ei#-jl+j}I4$<|x-!tO9+$3;>?RCxV}~98BN^C>B!fR0S2w=?!-V7o(12+TZ)`+{ z=%}5M!6dR(K(PQi1CuwXT07hOwTa>A#SLEc`*{I>%VeSEPT4wOD3vA?a8iKJfv$a$ zDObn_!eyR2wZsOa(M1YXy@Q4tEKCoX+E`H{LrlWLx*(dlkTZ$C9B+^H*vyb3+n-WL zP2)@!^VN`wH32zH7CKEgH-neW%QAVru&(Svx<>q(sx#GrI2n%_I03te;;u%GVxUyA z6P^Q5!2X2K6!c<%<8!mb>PQ}UT#L*&%6N)~^@p>bs2+V@8_UzaC>qnkj$A$Cxch^$Uu@6ea$%kRYkZRm1 zsyl(tcRr~DkRr$)q>3AVxp&@Q7E}P1V?OO%pbUJ9g_RTI@)r+MW4;Qiyn~a{Q1r3OO^N7Gp;pS_>^)vd&A&JVYluJU|mi8lGAZnUBj?kp>S!P{0!Gg zr2Vwi0`<`+v!LTdN_o+Ofy<6{nhf6*4x5y#~Ko$rHOrkhBW=euJ@J!%zr{6qB7b8GlpsWFKddg2&ER`8m zrjX94q$hg`TMI`D@ALOM8ezQQSAA3@-9cGI$ir+a_c&>+Y zPq9Vzh2Mg!i|B4MXZql2Wc#a%Y+ly}N(cQj;lskFo~V2FbrI(5#Gz*5Ni$tz0%kOcUttEgr}dHaBYOBS7eD3%MUc2=SC|1tKg02&1x zr26~@QJi=k*5vJh?nKbw&iMU&PCuF6J7H7`il}$GQ5x)t-L%4 z13!*-mk;D8E-ap827>jBFZk+2G?dyZS^+1Sq)BwlFzUpFet|H^{kz&sXuY7(Yw;Po zDC-4Hof9fc4kV4hHO}v`>s*gjH+lI`EVZ<%WpN7oPgn@MIhWUFxn`0KddapUs9rn@ zh7_kdM7OxM*uA0`LR|bEuIxafLBniff61)Gtsn58c+W5A=U9ihX@EOMx7^6Pvn`tZCLi6ErbAOC$CGPddxuygZM*-5wv#_E;NilfDck8Vq@ zy~+oJshmq~sq2C`ICfq3;__pwFUj{%{I7>NnXmegoRGx`&bCxOT-YyE%xUpjGHGYE z@L^7oygj74yY2G(1Q~pyxE@Z#822GV(TZ(d(h%=im-Y+3W@l;sD#}NPVxX<2Y|+o< z(m-#PHD1{7_XJ5+#;Sa8?V9;f>sSR+f6-Qzp!_sJ2A(!B?hwE5q31Xt-Udmh$hE$I z7hCN~+@`zV=i-d{TxVgUHHFIzudaFeZyC%Nb)Lxla{C|s{oe$~&I|y22Q*&K9*SF~ z;<1_w{?SnNp$;N3Yoy9CTJ8+77?hu737KV<9 z6jb*Jy7HUToiIr4uq8e*_GY?kf(|96zp_?D>^1|G}9oN7dfmLv+8@U`0xPSVh zu~-nQRXjfpIS40N<(AwIrn~a;@$W&t!Zd51`gDy?CW{gb^0mZ-+iw3X?@wj zlH+$w&M~nKZjA{b7AXdHCiuHelL@u42(AzNa@Z!g!*#zlo#k%RuA!{c>-dr4t4JLE zbFQ}R&p&q#4}r^+nI z*iTZ)f$7m4$vo9F?6b63vs-h-rOzA_tw%#WOBLX;G;u%T5i9D%dCq757N@z;n4lk? z&w<}_bo{s*L){pWH9eQkbJA963G#T(`|m`=n>-HRcxhnq_AG|Spo(Sv>-bFh-)bVvrulDCD~IPGmwqw zg~)__TNAKfIQWuiNP@iQnYlIqHiu=%(={BVpI$qWm+Ljy! zj&ysKz8OhfM~5EP0M`#GKWK$S=SqK&r9_21lRG;h1bz2L;4oqDr)A_zD@rCwta&V?I~pE=;nPy+NfhiO&Gpc`nk_#`D+L4Qk3ExpIWdWZ3HMIXp+&NZ zFO$NtVV&FYLdH#AB?p6Ld8h!jESkU3fh4h0g}H{=>=LAKsp z<<3;s8x`^%4R6DF?7CRJr6z%(jeaw)_%6IB_Z!0R~&-2zTe)wKD5qLy4Jirf>p!cpWypu*g<6BnbIFam~@MX-yhdKrgf2chyWjgv2n&W%) zA^hZ`{EExlfG!tm%$w6BuCr(Au0D^-+x}UN1(G>Efm?+B|81*A=9s{Vtk>9`9x~tZ zHU-{jqjcC43j=TK;$ta_Q4Qzc~S?(NYh|J4?{qXG-gP zJF&scw!AEMg0@Mg%@HLL%J+Z$v5JBd(4mfIFmne>s8_b%Mm6oRP*s8@?$?HZV+9*X z=N)7K@gy(HR_@s060H;LB~v{K*~^!GvXO2ea^=!Elx8^w5;%1n4&f%t90uidsWn;eOJU<;x5trDH!$72PYtNUA`X)^WIimtLl$4JXUUB; z$G1Vf-pJn(4uhP9l*PE1Bjf$F*bhb3kbeT*X2Ln}?-)xX%nWCQ=QkhvNLu2ziXp(| zrXLQJmN;(ED4-o3g#^jgOwIr3OLp_A{k(f<3w>pLb(mw6<=XVtADC+B9% zNbNr6HXE)x*$H{TABlJEr@9s8bO^iS>9x}HodW0G+|Vuu2u#|yMDhwm&0`iVRs(Ye zZl=Y@R8Zq~D-Oo#whw-B(2O*aDkhhnJnyjd{FCmx)*-X$NIilvKlUp0TiczD5yz)8 zRz9D2G%pn-V`n^%C1QLQ`v(6gCpduWpsNxh zxz}(@ded(n)1YWoX!q$B=we}yfiE4B1pWZU6O^`bD}=gOZk=2K51<_HdbCqN!|kZT zEkjk#q_8FV&4vs4Wu?I2cDF9uU3EO@8n+Qsp$>(DyIB2%AdP|(m(>Y+wMOI z%Vyo~zQFEjxkj~*?w~36(;fqI_e#o8Z8@~J$3NCkNzA>K(WsZp9Z&G>^}A8vR?EK$&q3vGiO&N)O=|SUyLq=l~X9%^yQ%Db4$s? zS}3NA#@jFSdzZK9W&-Pc{1joHA1@^`I`#5Rc;u$q#-kX#2s{`Gx`BMupA%H2{vmAb zGT|ZDLEOf^%BPB{`1z0iN;Z!qumP~JT*ts|vQ3lc4}w6{x(3-|d*zUf!CH^QU#fr4 z&HtgGqSXj9{4?qK+`+zuWS51&u>XW{>V6~rX(R2z^-+z1gv>#%W{0dUDkH9D{Dfet z<^7uegn*a*g3^lwHwJT~W%{^6XDkv)%`Bg+XU3V-#)ahI6VvdZIEcL;R{lL{A2T5TnDm?Y5)n1e)bU5?Wq=gs)< z^j;Jbc4J$Tq+}!FT?aPl>x%i?w3p%hxaXhCd|$8Di5Oo&#MA#RIDTnR1x%c~{XPhO=&iHf5asKeFCCoa+DoA4f-yeU8kmGmJ!LHpd=CRCag64#z=OW>G3Kj-70AtfV0;yJH=DopT)gI-l$MUf288@89P?{_r^W$L)51 z+(x&_Y(F2q>T?Fxg9_W!lJEjp87!Tgo3?rIH85t&9 zd)wSpCzw*d@6{o2Gu!Hl@9B@M3D>V~F6t1jN{mYp@)gjLfdWn;`?6yt*B0+BonOe* zILSK@@OIybzZ5-7ONYm4PTCb8p!N@J1B~-DjQOSR<#&k%cMZ5{cU?N7)Ap^7I+(t+ z$KZXCc!_#ViPL6rQ*36zvURL|)0yBcRftY+bN70m-Qcb`TQ*jzwd;O7_wz3~#v|At zf;3<2IsEQ_zyFHWMeliY8yZeSHUR@*Bg{4Kct4`xO=BNx=PaR|+qeTP9gy4BQN7~a zc5xeBwR<<{rm@qw%LY)i!)=q6h{)mx3$D#QkADFuJLbrnts=q~fGV1fT+F7Tdfy)D z$pAk;uPJ(0RK0Y?!kU4o^6u z3iS2;Tp8pedt?d2+vCO<=LmS~RYsU!dIVmjEMc|QM&FldGc^f|4mKj-p87z{ z5>`bsq_l8eggQ96+X!g*DA+bkRs1=qH(4t>@O-EybgXsbaAQokk!7rh40@N zcX4Qz%NwNKFa2U<=ZSFAipU}(hhs~XJw!%l_2qo-hWny78^^4WdPle@?f#ig&;V-q zUp(Kx*ifm_6zyi_4#?NT|D6U0B7oqb0h}0J!OG|? z48^zhizuj1TtU<~={A#D)!@Ln2PXw!31a_20U6z{GBsBz%Ak8TAC4p^6xFjqkdse& zq*qlt(dtR#hN&J;g@Wf3B=58Js;rhxr@p|q?|#;-#4~8n*8Bdou8Q{_HvP_o2?;9* z=J>tlJ2h0>u|6gZ2Da|zf3+OFcnUWO;6H*#>1(`+qjysLiH)8Pr+h1vRfP9WZ?M~z z`0j^{?JyyC>Y--7yM!U5)vAmZ`U0;{*2FVmW*p4!`CT(%*q8Y(;ihv9$Y}c{&c)-; zQ)IcB%$(o&>mdopQEd}UsML<_Oh2VeWuNuBpBAD5&;3hUS~0JTuIHv`oIJcVlH58T z7hv2VTnCk2_I(TCi`{kzvbFLKu0@^4TsEelQ9sFLRl&qc~j+EHIi z<&j>_+}!)bn=>HqDwsv?GPofs_#4r4;b7{)Ib!5rj76HpSZ+6lRQ=~Z5oKVOm6Vqz zWt;ijptDk8Q{>2jMXZn;7ckN{jkZLfV&C+xTgh8HJ5}8*;u?weBkvZT+9RQ9#%Hq@ z^p2iqXgi*UtKU}lU5H{F2+0bq;X!86d`*1HS9K;9kVuZ-VY1>8|e0o)i; z7?zRJ73Au3-yD>TwDC`@u4-~QVHh%1Gpx{6F6Rj_T!#GoUUNyaWAs-@$HQorBZ>oJ zI9^*!PVvdmb7i*~Qq*?RP-&7`Y`a?Xz=3&YQn5N2|2xacF?c(H=9s{+xd|e7EdTu1 zVGnBIDeuXd?lI^uJ5p_oy`!ckCsA7p1x?WM<({8dZbJOttzPd1&ZR%-H|OkxG^3@~ zo$|f6BBxDZ-gBkl%Oj8Dzl>Y4G>2c;`;h-uqEeY60gk;L0raL>km5sz7QDLN2`#`W z!!|}}+wmZyT?1`0TBO_R-aSAQB;S`l!$1$gsw)eTr8F0n3d!bRGrSg{_)xJ9dleJ% z%s-y0)M}v}Ac7x?a5gLg6Jf=sXa^v0mGNT!ZLho6;BPJjgWY?k}!%&yU9 z3=`;ipb=8Z0qajB{~^6?MA`IMxxVv1_^vFd%8bY7lSM;Q!;!F@)KHx5F`@{fm6lLI zQTu>Jxo&Ijv@GcmdMmME{eKXEofx3gy(^a8Kf1&{n_p@7UC;6GWdJELj)))i;}h zqI-K>-_Y{@0N}~6-ECq;%C8SUUUTw99CRfGt-WvHU2%29Wmu}sMeq2x-r+#SY)aXh zr8X-ak^)B)o2wlJ@Gnv$yy5}!xDX>uN z*mqR>wx5w=>1QD?8nCBw2s)FK9!M1QN)^OBa)U3j?U;|Ao2;rn^t#Y4haKj-U14vF zoB^8%DI^Q$AP1h@G3XYV&upOuM6>w8Z!e2QynoMn#RTyKH zD$&)L$9vy^3AugS(ymVt-fTVEz;S=og9t-AQU)1Q+0O0iK88i^l}4RlF8tPs*FAQt z>sy4TV>Ux(jxtemtWGxQFcPd>No`2Y9fv;-e%(lAC zuEHWh5PqvPR*oLxaU7PX9U-JhV$BPscb>$)^0=faaa-`{jE?@7i+;0`6jobRtD!1#3pu&Nt1EfRnJr&7ZQmWnFo*R>mhL5D9zpsdff3dbx+p;|xN|bB02{#eW zd0$z7@I3VJS3{WPNWtCV@7`X%!#yHL$-@=1prqw5e`8rV4a?Sh4_Ex}_k3er$PMs? zTU;`J!!;DZxw=5{_7z#ug@_!w+WJ)x3jNCxK}IqTG1dzo8_VM472SNG0oW%t7s3Q)WvCp4bM(l}@7o5n@wM&h2?wxY;;Pr^(rG&`1P4=%vC z(AWhT#zMv`8J~p2lHU)__!9dvWf?hE??~xxqML^?hkF7ZvDtboosKhr{W@1AdJi$7 z$86&r7ONyCa%C&!KG-W>I9leWi9~6004ipR1jpgWOhu1Zq-yRR`6P+tLN}i`TPng1|}a%-m)l!%2!poQ`>i9X8MKgyHSSI-Sl6&&*b` zjN>kYM<%+s#|Pc9H50gN6F5G*on-Elw~*97SJoNP%S$+gBI;^_Pg)xx&f>bNBzne!#jjt zE%>$0LtNh}$DFPd%Ld%)eMlY8;vk+oQJJtLrjKHk=geeIUaP|dwuefWU9~?oyfpc1 z4|1L!co{kt_TRkxEhnH@M2>E07IVPs7b;CV77zxAQpe| z=xl0TC9utNg z>4ke`TY`(?IjhS_VqYx+;AvR7Eo)cEWwN{gDxrefqov)YD*DGi%4D z+O?CO&p7$_-Um2YH`8q}z-tk9Z8mx9N2UXh##Y}IVWPN~DIxdLNdAmq?-&5 zz`G;B&DZXC437`bNQ6L74-Y>VbTRE2(d~HmdfIWeI?s7uvXkzagHz@F(--#s>ey-F ziRMTQ~GpUN{kT-AfeYLRubRnF{r=1CF*b2 z@bHr2jUN7+t2EV4XRHwA>i~)`L|408zv>#@_|&=XSUz*CJ^H(>fFJssykrz7DH&mM zN8>y%{c?VLLjbXK=DD)xMZ3@5j@3%d-IXMEyVU{Jy(KGVuXLzdMap)%!)psWM6nYN zsVKR#Tp{vEpL22H{}c*=i+OAa1b^GS4p2NtRkGlT3%IZ9(Z@GQI)NR*0443$u`;rY z>@MoPnHuw)0R+Y~yGD)zaiTEopA`;vhHzX+btVZsMLv`Jy8DzwmkT+Z`&2kP{1B(I z&bNiJOl7uebso~Vn|^6bw8HCwLvLWM^;GgJ#P1;xdR)2>mgS%%>6-0H7;fM}ys^;q z{lby(a3jkpENCcyc!#I;_n7@jQ}F)ks&^hJe-@YpB+T~_avO^Lm~$sfvjegA_q_JYoS0NdBu40xW99Mc~&i{WJcf!QY3>k z7>;Np?&)=Kt)%rA!5k%J>_t<$91!yd7(FNi&yeNZu)X)qY0JsS?+*s@!qc~vidD%L zcPe<|8QX>>8aEowG;rqdVH|xw78Rd$dmwOBUi10l%d-jCqA*~o@K6X>oc6X-8aiLe zr12udK($V%Vf*uA=LHsWrv(D;FB}*TLXm%G4H!e&$K&Lx1oc5HYw_mY{HLa~|DNS{ z+7{?gkg!VX;>5el;H;&X+74{-9oB9@Wu!ne^e?lt z(mHO+l5W|m#q&p}|KU{0f~uLa1ZpHFs;DRZ?cG-N0nMi(T=?|O;6xFp1NdLU*>kor zPUL>;ccPu{3VZ?1PK(?O29eCo7pWY~tbqbyUbCim0%~rL4+I;vwvQl+Ic+gN4Bvd% zJ=6M@JukVeJ1aS-%#rPs-|SBAIPXsHTAU-IOw~H%*|2Qbn*z?~4J}ESiIGTAN7VMG zy4%FUi4R04)DfCh_6>RMoIW;h@W;nzNoPbiaj)Gq8y@N?F&qpc#d~{vI8GKEiKSI- zFWGb7T83=?nfJG$vE(+n@-c6u{H{mHw>TZ@+{6(|FLGwm8>Vxl7K(pm`Il_S*jeKL zN>MT@{;2ihI!H3v&cbYiV}!2k6VCD*whC zmL%&+>f^Ulfg+1SH+3O<^^*^CB~Gzgw1*Wby_;ef$Xhnu=%n@rS4b*TICVkPQz-J5 z*fYq-nMMO+1a+%Rx%Go|Fq>HvlXPDFtID$1ez!p4ERYx2;g3FjFELG_j>w!ai@M<0|6+ z{l*WByd-R$V`8Ctxm|}4{1l>bPiay=)*nPx z+E+E<%ja>rT$MG>2{s!|JK0JNy(-c&^fO;)ho?;^Vb5>S%TKg&avAH-?NF=SChL8P z!7EiULG>tP*lD~={0#Pi+wjpPD`aD67vFvAgS|jJACgpGTyq>+ySc@<)AwbDCpVrxeid>O`?K{#+}Sa7JHaCC`6F~9EE!g zXgJqyQVZDw$TK!Ee(IwYy1c7&xIs-F)PB)lOl$nE`j`U z^mnA=-&dMN*i}txJ_gI;byN(B#CN$AdQ@<$pg>fdH&lOYoFAtZO0e7T%3{+y(aY); z4gUxf2MD{!wGAw_C@gnp;>D3%$AD?@2woygEaaI^Kk3_a$ok5MVQl^FfHA-QoISE9Fn=Y-IESP$0(SWS*JPlZA!J&$xLK%e{YsD5xd!^^N@ z4m~j&*{uYk-n==xVZ@+L$EasT)9VR%40V(XZ8G|Tj{4^kbE5alaWl)3y@b1d5VOsQ z>=Q~4NAK-{;(rOU2*w%S`3a~^&>5_44^yrWak2VAm8YvlOXQsm!K4Q(AqT4#tz8J-@Pf71P6b4_c`4Oz^yzP`tE@z+7 zzu7fQOL{&g%}qOwu~*>Y3?|^cC;OG*JR8JL5z{520hJN#d0ZpllBrXjt}juD)CqO= z)^X)5{L$O?WAk0waVjy1PiIFw0lNeRtOrq_TCOCvE2!?~oIg-roW0aAbyH(k-M6l7 zqM^KQ`$i3cIPdv~R`%2mKM%=rh|{G4TYWG?WXNRiS>xKuaVFGTI{M{J{+B-y(B|h4 zNEdP^y@k$sp6WBC^Ef>oZfpf8slTpX?mQgNlk@w#M4*QzFg7m!Pf_$(A{cb<)?-d? z1gj#8*T6l{l%I)65bsB8EW9994^2iF;HqoEjhVmDZ@}Gt9@3`gUs~SIgtX~Oz%1)i zAZ>%^qv^Q%8{G?n?t-nH|6S%v}kTNW^yS7UWA>EHeIl!wQh zFGAcp)gO96I^sN+UXeB3?dFzO(I4@@K59!-or!|mtR3!vI64)EsP(yiiFlDk(H*wP z$jb|=RmOWl!OJg3O%Ia;@-*RiDab(7Mr=mP32nlbiGiHUo^e=5#|KKc~WI%tfeBX64MP)7PQ58RjnB$Wc{db=AG zIhunDgTp{=OVTIPBY(vO>B8k{izbEh$NyXA+wO>ofD}uM7c%N;OG4!E0*Z#$t>kZR zfoj!uQfCdW_UxX4M@7@nC8Q-V+k#yT+zF=t z-9W7M1-xR&|!IO&HcjLKJ>6V#k#i zq^-|g91oUY9R3q`%Cbhz;D-7c{)0s zkgP0&aM#)!B)YaF(KoxcMTMdtk}8F9OlK7Z$YR1|tzUQa;`X+QsvUy@O)X>0dTU(9 zqp+iXs$sB=FJ6oP-!|Z>wPTRmbJ9Y%S5M`b+FF_@O}pnN-OZodp?LBf&a*DM{cY?S zdtI^R^!n5N{Nu!yf-tqo=q; zwEeRU_Kc{UsL?vg{hdf}>V(jH?eei&u5?;+SoYxd{29s@Pis;D^G5hzFWv04m+tdc z2ZsN7=~e~;B`z|{ekk+ho8rs!b`_ z(TcVuj?Noo?!5qeEL3v=g1Kcl3T4@QXwkY5;%`#RrFqdV1cehLSq!27FbAM-GorTl$T{?q`ZJ}Sf&lli`(>@ zQLLIQsDTZ%f?W*{N_gQSXhUQ^sS>gTXam;*T>=)*EkewNlm<>PB5u;_a*eZS*lkdd z7shZj)4rRlC0V-WYG$FA;6AaRK$ZRwhqJDF|D3ZIm7zS2^c#nd~f96h|BZ4%+&M`c(Ykj~Ssnm3@%sU4vH@c{`I(bUBR z9gOYcGG>8UId{v=4u*NHa2Padp3ju)14X}=4mlToC83Z|qZyLS%&QVy6JyS2EAUOE z`_yWXomX`Rwd)sKG-DdWsL1?CENbw1G_y`|Dg0q^f?9E5k|NkEJK82ELVBEJ?(cLS%RXE>E``Hejiu`} z?HpclnD*H0pQ$0LKG;RSIsGd`gq~L7P4}woKLs=tH0L=7K zL<$?YJ&-8b$@D@(cO`8fqANQlXDXYhPe=@7n3*1U3^wG^D;}VUW|0^KzJwO&JUy~aGi%PO-@MRa)nR+Q{Em~WLv!6AEvRA4%|rmF5O~O#;ntKl z6YRP^1f~T-pf_+^G_^M@3vby*0^1og1yi)^X%ZTsyX`;J;q7!CQU~XK88M#D5rqO% zkHxQJqUgjBl-LtSI3fx-n~uNQjcH0X;%kle^0SwiIsFocxYhmpXX`=T&V#kP)x0V( zDaxJ?9**POo=>@2A1vum8Kv$PI-99Phzbe6y-?y9s`LbU zSq=MM5=hxNkg+Z5*E!ivnOMV4pe%Ub9GiqV-P&S2TLL~#HE4~V)xWuG%f~kLp65tZ z1~ebZ9eplim$yG9U0>d~$tZ~O<*ih7+ zNF7K+X)IgdN#`)x`Q&r*p{?sDo%isLZjl!A)iv`qx<#F~j zp;54$&(rTGIbt40r4T7nQ!B| z-2eZw05atuN_&!VPLPA|f>GLgMl85CFFu5k%2&g)7k30!&|3PSoVdfD7@iPgCjm7K zAS2n-f*8J#tePo0a=Kf1`ezJD?e;%G@Acl1z`F73v7!$a!X1$=;dttO+* zGshcW<8)w=gNH>?9Pb9J%w(HNkg0lMTkfn2rLrp5b!NnUQ$!G@55HRe$c-XHYw9_3 zp3%Y%U)P)>eLFWOYn5_NdgvI^EPSzJgpUmyr>-(4((VLmL<*YNsZ}S0$ou}35Z~@rw1Lk@3f&m0tLm;m5!|FShL6*;7 zt*ic#FVHgVuDFP%%12i}ddW$K%6}L9`CD5Mk^)E=lGO8RY*(Vc^LSUb!{xD3L4fSz zecu$0p4aE>dg~_|U3?X-YWHO$y^h_0)4^-+je^_3SM)|3TKC`IFB?jbEU;OO3#7o5 zLH8&hSl@gD6XG>({m99>t~jW^VMOSpZln6ugWp+B6-8P%L5`GQ0uwR^uJ|b?VnNze zoB|{a^ys_5w-NhqEicVAV5)y7?#0!ZLv|%OjR_U9HY2~zkJWlurCtVBAc0?_WO`Fq znF0kK$5+$RQdTbcrXqU2W_~8DeVzGe-;pa6gV_Ctgd;aEcUca0D80P6nWh{!R$NKW zjZ=SKIgGoQ+D7yBvzi2$6s#CvR91f<_ZN(Us69W=7^H3 z%|;1N$_K{_?oG#QhX~KiHxO#>bj8A=gi@Xw>feX&mJ0|JFCOVf9be}ylTk#x;Q#h9 zG(f;0B;+l{fK$5`*qahcbX7ikyyC^-n)LGM$u;-<^sF@UVgW^a`=O(4cn&8@%YB2O z9Co07Jhd<*zu&O7-78wz94aG#p4if}jX2;6$vEzr&i5iKZ1gOD{c9;gd`Q#v{E9z( z^1okYt^>G0JbT@!eYU$*-(Zh^Ic>H@YK z4Zi~H7^$Uazhbl&`d_DAc|@%4y)8ePz3R~8YguJ)=gOQ155Qh!Sj5G1sLe*O;QKF| zUl)D^0i~GhTfppF{}38$ko9P66LV18RwqKp`FvG~`g}R+W(I9*x7+du%qN{QD44B^Lh=|&d^uJ`cj8%t!d8T-aGdCJ7KwWpyDX1@XH z_i_KsF)IN$XY{zYJfoA7V1}1*y<=no~bG>y{UN@ zY*TIGb-RCXr*KeD_0M1EF(#%}w;bkR2&HX7Qq){0r@&wSMqjq$2ISg49-KFV=I+0D`igTYsGXFD^=H}y?*!o$aUDGfSOi&ar4sP7`k zw)Ccll{ohuk5+-HGS|~@f1S9&bF>WKb6eV8|4YVLc_JTSpJ5nsR=L>tWO$-Y0L;%> z90E{|}rm1H=OkxL_*WGR%Z>NNp7pdxFy^u*V($OMlfcnxC?Z;_6QK+p+SA68# zNL_AF7JG3@D|*dugefP2)aR`nb4wAS*q55V2lf5+!QW__R1VyF9g1%B%XyuS=oB`7 zwJyL&l^Q?`j{$LK117K8` zvpG~v|7QJbnd^oMFawnr!HX45_fsyvV8!u|ybkR#0)|NPr{uZnjD5`rSG$_3Q|{1Q zk2k74%GK+5 z13Fl{FaXpGuBqI%?{Tr?YaYqCW;J*%H;0&~;{8K-#aNE=?FYp)(*NpKzt;C}M?c$6 z{K;>Es^NvO#LWj&jn#;nw$Sd>uEn^L`sw$c*h7sp<^-_zh`hAaE(1Y1=OmTyPku~8SfF%3#htWQTG3% z`h~OwP~r07*nb9{E&zaKlx4iZc-etjsBBl}-ctzk9Bk~JW07)6%v)yMyMgLLd=?0e zrm}oB)SYRZNQ>gu&sHOV@60~#E+{I88}LMq|6#$eaGwB9j4i^)7wBZBkS&`MCCow@q&XRl{9Z zi!rhVyB7H>=n<)DDWAQh2%pk8jZkO(%wba<6}|orQH9v+Zi0p5JU0?;)K00p(zD6f zksi&QNCtwRZX*ZVGM=_SFnz(II`qXuUsqq+~CZMV=ioB=j_T=-oeMs}1zy z&nqErgMmrl!8@AuSNw7Y@(ZYu(>>?;8U=@iNIW>xkY2q%V84eK1FW;|`>wS4Aj%t_ zjzy6#UGaQ6OolOv#=<)%M34w7Fp`av0vRU@c4{Gy@5Bk;K5@PD#z6M`&w?bZv7%$} z$}1y@e+vmkiv{En%_9c_Hedp?l!;*djVNF(e5&(hJn!yBN+Ygk?D-|KfDk!CjsZ<)lDZ`?>S8WHoaFg~#H9_e2(bcPnH=@A-@vJa7mmDNo2U0H&Rfz6j29Ik z81_~HCva)s$?@h#2BsZdd0al$-m+W&c(bkl@L1qN-L@?K?vb7-Ql%)W0vQmed!J0c zKzYuj4rC&Aa0!rT1ETLBGTs1oZ!>Zu>(pzo2jSViL!F&$X<(SW2Z@;Jpj56^O z#SOj+Bleam$u~RFEfKb2nKIvL63!T9{p;ZYY#ah-I99PrEakJs+DYzqdxDiZg$^i; zeK34z-6>~n^+%a!IgwBPk-2PPUBUNV?bhWt^GMa_#mcpzHX#H9MB(@8;JHAIP~FMk zX7o9W3pxhC)hL$d^Na>GE-@6pT*T!O;9A|1a>uMbqM^;!C-3|L^L~$6tYT7M zis;v@2RF0)71UEsrv8R^7BqO5fA7%dKU2>cz@^K+0(qk)V9)d2oVjm7cKPl5h#FlF zAe&*la~V-f&!^*Yq}_v5C_^Rg9bUPh7`h)AUEo~is0lhe1*j=L^Xk3SiJtZw5%KU6 z65|$jj)X*TcDCDuO=39+y5U(tBiDqTUdlZ-d;Mw#Zb<3l-XkAOdduN@)oA&-sn1Ni zNYXOgiWy_Wc_eX}6+I{4=au?=AHKphcAH5Op_2Q>nr-r`yC(BptXJ@*q?bf#vEzvj z3o7=Kv%G9aWL}a1lQs2@?Ai~0Tc^b*ce6krsJpMS9N!fukVPlG$EseS#E%$5Wj}6p z?l|lk#D^1<3f~?NdFSoGclGS;YDUww!8z_N+nQlsGMjr175K~EN7?>~_@64c^3XLb z=TTa8^)kM}c@^)N^JNRk7qHMPMd(Bj*m{ZoR6m#)`k61>1NWku=)o=6o?sD9v ze4siwy9emE!@TBLy4KIEyc_cbMOF=E2_6XG=+}j}hsx<~QgfA`KiUbaK-fOr)A_sR zTb7o_?*P1;_#dO9rLnHO)g!)H?m_OcTY&eB4hAoQ6%r8@j+z~J^=uaDV2s;)xii$) z=xK)W31}6U*tMRdqJ|=so1Zor`U(0X{NFY|G2;oU(|}*c2hM05Mr4nJyqMo!X+}4? zsXfxM*({LdWnx1K?F4*hj2-&%u~zSLwh6kIfXHfpg6@LyE_0jr-RgNq|kiyWCucUxYiP6G0=!;XBK)3n|XXI9A!lBFb8>v>0b7 zPw0cTh{ltCzT^;dF(2Z08>D$ zzg>oeQ^zst9MKAB7II2o>O3CcM=@1LEE*18@Ek3dQuq0|*?RO;NNXWU>Ea|%!E}Zs zLpJEej-c{!XV;JmbK^MG_EUJB7(LCvb%);xY+o%`^^Tc4^{l?M~Gizi2$ax*28 z5f%i4GT zzb`(!@e%DF40#MW%)l~(Js{SS3r#609OUheFDi>DD{7EI(EHi)rwY$oyk-kRpPKRq z`kV40pKb%*8t5}Gvb;IErqVI_c5X~X42{6c=-GtYPY4M1Ej*N^<&(io4#4?<PvFjVR3pOpbm|=zPL`^w&DS#z(TZggG1un5T;hSg&q0P6V`t0f^ zCLfI41#-MziBu3xFHnYj)uYH;}Ic-VOy-cyQNP`a$JM5RObSa>x zNpjk&WDP@Q86>0+RuY1NA>G4XWJ?~Cl=EER{Ow_ZS){V2ZpoEd0gJs?xrg%Z)`sUr zRhZVsMCKa*?d^w^+Y<*QJjVNM?wHmS_+ru>1Thw+ zD!MMDH{b{~-JKOpK~V0k`x9gnlZK&1N04CT+9<9M~Jqi6+f>5dz@TBZ5T z!GHHV-vbcNMM?mc&K-;VCs3X-I%3uX{f^=J!mPMUS4c;|CcuPI-2qBp5Jp*noe4jh zoZ*#%UAVOT4w#oNN!qk6vBdS;JG^GYo%4{H;a-FTrol{vx7mX9sv)5V`lG!|spX3kX77V+V4Y-b z7pzcQf{8@4jpYP%`Ms+c^t6FP!mKL4d?vpz>V<-~<+N3A3U~cg(I*|}6joWKTKPM} z$P8%2$%f6sfn*4W%$@$;c~+{3@r3&?H?{{o8vIVpABUg6h->kC9r}`cGyLyS_lq{_ znwghHUjJtRVsz0IJd@rvz3cQAS-SO*rECDm+4vbL^fLX%CA<1t^8en=2|z;J0s6lf zix{05?HNxP@6>sK9q<3lqW%GBaySc#w4f)P=N49?gdisJlVgizs+M9EGyUmv zCMF$^OD4YS&f_N@*E-`$oclh@c5^gC1@}@P;9I>U1p7%FNa*Ed(GtF4m>}Vo<4?6cl2HtVG&?u$T5?ebAJQl zahcEe>UGgkD(7~B`KubXOOJ&Tt}@wt`$odwvq_Da(qrjNPo~MGpPg1G5-r*edro?4 zZjU8Alp?%%6E*Wh`1O!oC?^M@^b4Hiwiywr4z~t}$J?H~pFSU*bGTxeStpk+J8t7O z+&bL^ujb~r5w+Fn5vi=6{eA3}*ox}QLwgqtO@?)=zPGJMI%!Dn;Z&zysS5H>I8+!`W1uNY>CE4Qw$Y{0N2lyfK|R`M1O>+ma@$wA>3J{^wyG z#K7#sa+}qW1^c0S=W`9@&8+fcqc7u|_n94ab_bO*5^$lUbvZrX$bSTGp}(8zBAFPB zt_!M1OhH+%K(Z5^^&=G?rhW-v3z(#KthmFv2V*N%vkm zM|N|CkuQ2+yEXFeX9K)GAk#q_Y zg{fdhJIa8BlI$tWPCFDFYHD6Wmo_q1Kz0BM74|6}(i&t5&N4AVn=1*|jYV*xI&LZH z78)MG2*N$Ddhtncs*kd}fGtp`?S=a>ByBk{JEB2ZK^c^`r*!w}@YPj9l+COOp+iK9 z%CfQ7|EU1;gw}eTC|^(5dsTBFza;I|{h7)om9f7Rs5hRUOitA{)Zw$gGpAVLpP_85 z>Lxq3Jdjn+{AXxWCZvTnJseR9$IDOUOQY%{l?e#Sy3NThO zWJt>(THMr9Vw|8b$soB4XT3A$i~dP4mHjxobkArpPDbH@!E*tn$a|sAneA=EK#gdw zPLK78IwO{M)?DSIPWjR*NSy`2#P`_SRo5kfvhfub#$0TmH&r|~?oPvamfiGZ_lK*4 z85_v+;it!kl@Ujj{k2kNmk%D5?DlSJ%L52hEUFb!#K~8GxG@(eDvRrU;^!A+X&qn? z4O78r?cl{lPqyOBWcANDc68>EROX|nXi?^-+5r(U*_FtSW({joV)~ptMl4Fg9g>Fz|a9=&iAL*gYrRht+~P@$#SYKDxRO6W+6%J0Rbd zp2_jLZ*IK>3>bOjzOe@@l$AX(;vf?xi@*J=REdl!c+?}X?`T}R-<9PS^7Z#KD+?*% z^#`7Dbsi6LX3Lez=`p~5JJ%QQthM*W!1DpDd8s)&&g@Rsb|@87xcy>=N>JTr zWGZpS-+j^{68Y3VwobVx4wk;z9ey7xOjUY2(8y^!BDFqSSvBw(Qyul9`NLmxa!ed} z8QN&(>;?UQIZlcU45|$G8B{9(;pfH_REN>p4E__q1D1_J?k7S_F?-iQ7XmkvC3T-U zrv_X_n@(JyEEX)iBMs}MW+`i=)mjL~J zy}OYc}MVm9Dsb+VRXVp5}71GgD=vDk1z-;N+PO(Ub z$G8IQl|7oT)k(k}BlLq;!GK9D*;!qGo3d6}>D+XHUh(7FGBJcg0r9B_f~8Jjva3V# zVbm2O+HSO_`GjG23!hg9PyEbq3_PpWNf!$o9F#u9OIY6E48#oy7*`pn;RBcjoGY)+ z%)YPGC|oh+bFC3JUyhkF;)1Y)xE8xC>;l}R`Gof--66yHWSQVjJ3U$an8BjPH0bly zo~T4K(co(JUwD&?Q1)MazB2CuC>MLo+h;z`^#+b*41qN!cBNE|c_GWQ`mc%Fi}B=~ z%Hf`-y{~48&os%4>8x-t>x{XN%x&hOv%$zM6=IUdi~V|-Q! zFxvy2B-kvJA6V0BB?=c*q|cD`OlxHJQ19gBp8Gx1|c#WsZNP zq!BD1i13T!iuE6C_a&X)bKB`_l|23h57m7A)E{W>z16DM`<$K z3VQ)e^9lB1x?a`qd%nWzG4SMm?OwP2*we9S(S-H?s4F~d8Ab=W&egLH&(GC?2Pc0& z8NRx#fB~Uj$NxD|x&Z)bfGj{LfP`H94>2R~6csT0UTaoQZi76hJmWOOAJE*5?E=h)S#2(9zHhWqy9? z!V(x_ys`3adTE&xdGAS4ZnwhPhgf`~53g8xkQxwdMXHSE3cUCW{*B|Vk2&&NP!MbXC z;JdF(WaMwhHQZ0WQumhi!@N6xEB9#sD`2Bt~iBoW<8@?G%SauyTF=#iqlz~*B#V=dZJmS|HWYb=Gz`RsR0L1g_Pf6t4w_=cOO6_gigCBs-29My;;c6hVUE zA&M$LyS?r~5gohP>@wlv=k_x@YCssKe_hM&etLy+BlXMk-laa-wS4Z~y@{Yhh03o? zzmgg+TeublEK2cS=nmm(V1o9s&AFo(MToB*oz_A63ZK>>;SlUL2D3VSF(Q%x2{*3F z0auKXp9|ROzKw(ltgd{aX}IJz3s?Vwo&-0p&zg+eP`kG2c-KzktQ+iyhzhx={y5wY zs%pBjT>`I8p`wn~jU`JI{?ATA0yr)Bq}4eyS+lr zSZ_dwUN|7%Y5S4+_}3UOQn3JjrH7dB#t9cHiq8TOd>haG0_vHS7RwpHE|iaJP(hAl z=t49%B~T-_t_ID!yriQc%ikgScA`J#XDt1!2J-pryV8z#yQcJ#7dSinT&8y{)JH}h zl=Z9qEaZQPvcG_NSxs4xID}E9oq*rco^2TJ0FYH>Kk-aMK(PQIorrypg2b|cmFFyX0z$y1N+P}R=_PRWBo z1#f1bUnpUe!l7Xt6WO_ANm0iOWxL{shqQ)+SMqY;HDo)QTw|toj-9-Tzfb9vN^{Xv zj^YM@#C)BmqOgJM=$r)QXKg$A#br**DemurC}!?wj8gIy2be@wio8#=emi39^6+3W z1%nYMuE%C8NhjbMV8|qpr-QJaqAcP$`8dbI;OKTY zeCY4gn;nm{YiA}^s!K_^->jf%+W1F@tJ}=!_S91`zgkgqZB@+u$*6wEtQ=xqzZ@MN-XY=MMT;O`YGT!NT3EjjSEnN>A zBuwpFU1b|TkF8N!-8J`887++xSAx*i}dbM`Tb&j#jd1<=%S8i{DvRg$Eo^=8O#QOKAVxU z_Iz8OEN9=HE{1@obN~Xnd8W;`LrNcDMP#2BfUPC#QW;Wp>CgZu*aDkZr^qo4?hCnE z+f%h=ZGbsqo&*zPSFO@y=O=xQx5x}T)nLRHpol!=CE|6KPo!d%+tHM}S;1@jeEp3c zU8l+Hl3rfhrwvELjde~;+5M>%_dAcHwCx}BztS$Pagwy5D*<*$#_ zXPH-ig%;4WYeEAEgpOpJ^yiiD>X-kOgvlWm#&L_4QLl|<_mMvgc2%xlWR7PO&my;w z=2V}RCD#xbB-6$T&C{ZyBe@T+!ds2c=Glk3{D*)O-q?S0HOY=r%hq#akA{k+(2=z4-yY zMg+`+xrck$FV8Exy+EB(sK62^cAVYJ=ore*+jI4#@@XfXBGW&h+h;}W6LUcJL61DvyQ}*K$$}_9MU#Rrq*eE3KGJrjV(;L$xJ7-%A z4id7Dfgkd2{VyYczGJ3rEHkdjvutBz<6l~&Bw6Y2Yiejs<>2KJ>CDha!zMr6r}RchN;P z0h9>noxFabAMbB|LD$7v;Ool+AS6W1@H=)%nFvshzDWd^Rl=wyU$C5-DkjgoahL-C z1dNlr#A$Re1yf#VByBsKU3zrBS(4hxFPkL)st*`>IVny1OZGzpPFinFzVyxGQ}tA8 z<)-eJF@?{|zHPQ;CXke@%kD8TSG_HFVY&CM$enBs|H;a|dZ$?B40x%eR%&gUIO!z|$4$yLwdu!#=_ zrZo$fp;$uIJx?2KYFPzU6FrG!!9IH5808sCN~Oo5IF!6xoc9Hr^-g9x_<&N`{K{z4 z1ae_~Zc>?NHmg}c9`WaGxA4tVZ3TF-K}yt&W5U5LeLV<&sg*N^%E>=9GXux==VE~$ zF3=pP*n@ zNtAvQ!x~;~#a9=AIt2YTw%ilojR;du|avCu%;I>$!JMu6yO88EXIeW@T zaHwB7uegg_zkIY7l+YX(V@^7fhxxXN36+KQwIAeze*oGT zeJmGzj~!C7&a+aaaMn?AgjUIbpU%;J&DmLL^YztwbTzNx3Hw^t&G{?C>9PjzI!^y! zR35d%5mYXJ(oN>+{x}jS0$vMge0mEy_YP}f4@*nDYxE4abVTS;Am_8X8K#X$3G&u_ zcQD5Ew-lO|6DyQ-)f6}iY`woJwE6a>AHvVg(}$_3rR+xac8Nb+DVvs+bwqBi>0LSa zp99WMKl<$|OJQC=m!)4-C;#SZfiAgPpxmRR|F8oYNihOS*pscMrob-W1a*ebyP#uX zCB6*u%t`V@K%p^lxbAaOL3dnQZG2f~%C7<}|I~4;@uAoH z{+1uHTml>wRKz1qDDfAcZ*zDD$p*N-Y=BTt@5H)mQ}`9|lCOX}pszM=KWJa4j5o^I zH)wKQ8@|aD5l|j~m%BS8o$N@sf3?@*8~Ep6JojkqdEo4R&xgZf^qkAKBdB}#xsvgV zBll^~rT9FHg6?PU?GX(vt&IWN#Vko0hfJfuc^)H+rP{T^x;t0y7ju0KS8FxIzeLtr z7mo2*IYBs13zK27C%@Z6G|Q^Mc0C#(NrW7!hc~KmyqW+^8$$IJTge5do;~;$*Ff!g z8`sB)X^XYW9!TiJYqQ`e^bdqLdTB5uQ0dn{fP}i)wcF3UTG$}c66OU8`2@c3A_o$^n(Bf{koRl)HO)G0gxJr@al>0t8^(%ST{$c-T>eUv zTyot{@#C*V;bV5XO5r^#xQm4#qPa&NJw0wk?GxvU$Gzn^mVNOa9v1vzgWXKrMr>-1 zdh7lABgenk$0ZD(AIkWEjD$0g1hHd zdK68$Q)}Bgot)uwIp9OHT66vCS4&L><{Ho@*Zcej*EpFG#c|CVV!n$@ zKWdLAT*g=EmOfh>|Hr20BN$B*|M|TZy}^L)wy~a;qW-MTz(7iG@=m~OCT*S++Bos6HmC4= zW#n{?$FKW0ry4s}*xYWF4BQ6qh%>x-WtK}Pu8bpmd^_0U`v_fZVNh-i_1@hgOgbwTOW1|vFwN9nkfuSVfy^6!0 zn0jW6 zb}W<)0Bj&9?e9GgOGj1M#istO*lP1`1;QH&Bz(=&z$=q0`9%`;P!ZY-YSAvP<6YYY zGdmlFU3-^=&WLPF{srEXo$J{5GdfQlVlQ)uSImt^TLMv-RMcQd>^lp7*-r^Qzw${n zj>jkFDKN?D#Ud=^w5!fx6l?;kbNMuNg801Y){Nf?EGa1za*%^LO{my%@6wo{RjQ5H ztn~A8l&L3+*$}_jhLPh@D`jL`^w) z?G;sE9aXdu#jvWc*1%3|;3o1Q)uk^F0cToxNRq-8TdML2{0R`G3qRZEcFMos2+gg3 z+5Pb*izlwClHLKRfare$I9q}0WpK0JRfN_7@0TNqyDt)l<0>O7s8|vuvAK4js+VC>zVFxvFF z=C=mMPHJ&tK+)<;W2vV_G|gI~P%LBowU37q!4xN8k9Vu2lkbVw<$lFSM zU=}$1}pEjzl4;Q0Ch%T7_1C43j8F zPX-ijT$MID$K23->`~kxdALsy8auQdGd{n8wkcBg!dz`@3%dS%yoD+thFZlj-?*mq zJRNIeb$f+EOO=UOitS;rpyaO%fG7}rPHAhpb;qJnh7N7wu{$vqg~LXtG1R`+y~QDt z!=EJEc8~fLj>{B4>ceFpN-hr(xNVxVa^9f)d@9L%-$YN(f+^uPKQ^w6SvQV%by?M& z#%8T6%BHv=lW3ynB5AiH{3fJ~J=@J@?d81&#u~r$IdGI#c#gBIYuV-eoSuw{3NMw~ z7OfrVC*aq*uWFX&qt@_#vpmKiIxM(7fq0cK@K>J)2i zU+yF%82etYFMzZzEr(j&qsWN45*6JdIXrP0v*WEbED5U0d1bp%oB~zcM)b4k`i<_1 zyB0cqL6sP%XI*Lde=x*$mBrtEN$&q=U%Ezmv6JyR2^^m_54c;@a5#SLDpeq5diT|x z7$3|_B_%a83~!=;O5Nu~|IgrzBVkh6(gX32=1n9YVSyjX-jboo;>pa9*Ak=2lpY7z zXv9-`#eE>jB^i~b|8DkfPu|x6dYqwdP66DWaq7I7t!*zN6fHR!VLV}{m$68JwPY(~ zYo{iBpFRs_(0*y&>Qtfzk{LO~+A-^~>xYhI+i^j8EgXjo?a=awwb? zjqmxyu&&|%od3bQ9Mr7iYFTK3QL;$~tt^BhAHLS{wMomb+RC0na7+0Xq@WFZ`xn;1 zi88l^-<#y|ZBjb~I@xLc`ryQL@aKT6qooW#-vb_9zeNAKnp>lVP3z((qy4^zWxqTx zH6XR%fgTla#NUx%qAAuMe3NPu4;AcSVeS~BQ1S;Vy)DkEY3>tOn8j@FE8a}?TFv|L z2-`S4^gO~L8RgYlCK}yZL?Ee4sq$fD!-;49@!00m@FUcu(4BC1eha1Mt5jVD8<&SA zP-`Lv*pp)xR3(>-Vrtm0ybc}>ykhqLtjl2?)4{a>*j0z`Wr`2}S8ylOLjhAO_+1^P zCtr&n3VR9c-@9RATv7(THh?%OJ|fh`VrA#Ud9LY7zqD)m@S0j9Yz?a4KZ7u%E|n_Y zxy-$M4R=E3W-I{%Y*@ zwTh=Of{_FfIE|ho?3B_)FWI^;D5hIM>qi|g0m5bhPKWo;?M;#B^yCyz_U_S5PnHX6 zVa%QXF#{_;;zl=Rm8qGZR@qeg)cO3I02pvd#1HU7js5eZ&-B}%3o>;|5#G^DTZE@1 z+U6W3U`K&&HAiTZT$aN*Klljn1@QARpsJpwM;|)M*O6Di(+a2nv;s!SP08Ppladbu z`lAJk@7#uV@P+e|+^Ob$_3>wl3~T!dfXFxyCG6t>a0~YT`R3dv3DlZ1rJo^5R*;wQ zi2vtZU|&1AJu%&Ia zj{`I5g82eS?lUv00;mf zw2O;NLQ>6lw?i54B8d@WL|;W;FRRT)ncY}vm$y%cgO4hJ`wo3>_^S?=%&URZ@}e2* z!w}QaPE{+&Ie3oJ3pp3mY^`Z|7VQp!RL4z&{7#+XhTOoP{2;_$Ui<4VBPTufh)8>G zf0uFnf9$Jxf^=wn$j!bxqh9~v(vr)L}~N}(9xr0F`T|F^vIHg#fM%? zA+ldbn+#j)y7paxxC#~*1_@4lEk zK=cNqMzE%q*S>UCdx$ z;GnyWouqYa1!vB!wb$5v3{hor{aH|AcA7jsWy)drwI$juSXibTR)^H2Wio12oj#kT zfag|B^gXD_GR3Ryaw~(s1ZxT&9pt4qE=ZBKjd@FV@gMBt(EZ!ul3R11d}nhWir^Ca zB`dM_73{^J3-mL;_mZ<5M&)`&dVDs<7#wg>;0~6`*=yo5>t!GAZY}2bi;u1tm?4MM z=TMDAp`-dlEob_xNNYjY^4@yKP?`W!?Ov_D@C+h%;Q3;lFQpkQWg}if81;?W)+s0l zcBwAwy1zCm+RmYGo3wMYi4(IPp61jgF4NMOmVp#khq0EKbU4R`-(lf6UaUHOR4-{d zqA%v&W@CSyke!5Ez13gucGobky->1nkIuC4byoV$_LQt0rx;z&{k_Rle&9{i0V8eO zhTeWVpIoutD2s+Am~^-n5xYDyBWp)wWFAa=Xv^`?{2K;XRRuPk#q+7A_H(eTKTM0&6s-&Txb(mUyc9`XHWk?hV7YB&W zUtfs`R*~kZW5_Jzc}iS09||Cucz#hbv7`H+9hZ@*_dPpi6rwK% z+RJq=@W#XHEGq5B9=lLGMa>?tN~5O5!+xurG7=WV0+f`}Cm7VT#XuYMN{lu+eKe%?GS0{JM`F!-MW#Y(PX)7O6l_%K zNlFnu-3vf&OMOAffhEP!*7Q=GeVRC>IkaO05Ei>BW6Szve*j%@%oIMW?Q<2jr}I~K zTKn#0%h^TH3K|qnmRR$q_sj2T6RlV>D|%`fN~gv%HyAWHO02|yL6%wzPM`?g{M}r5 zQGvczn2ATQs9f0Zl5Y`Vm2Qt~D3}mEW41}nQ!G3{ff5>)K+?%m;g^` zr+-3Omxuo21rX-I5ElNdE#F~e!OStisIXf@4r9LlN!E7{J##t`HY(F%*ds3+7yin4 zcnXJtH?9_Q4hH!1aanR{`=onta5n*!+@h`JE!$vz?KQmW^RHmr8ij<^c=p1v8lXOR zq}#DB#1wNclY=yjwN2`Xz(P8~)>gO|1b2RK-*~eu)y)Ld)>~O!W6e<<*U|Ir zPm1689U1RlKi^V&`R5mt30IKmMrI10UzYz@xkwU#RM&}8hk=+BcoaQUXZ25v+WAJ<*;b{ zU{|_E{wQ72*I@`5mjbkqE1_?M*j6Y>+Vs&&p$~~0icRg-M<1r7R`Z}Ue|UDmo}6wB z++iDc;*VBG&u$DpO83r(LP!`2{K9HJ#P2L9z+*6Du`EseVV^b3QD#d4U8#TJ+W}#yqUps46pe=Fgy+ z@o}FeDdyI&Eu=Y}&Z*?pK;UUOxvI8r%hR_zVTy60>~wK`TjwJ#)DPX+w9huR$1$rP(^jiE&y%=?UnzhSOMb3yq1+Sm@rf?w5{!ID_SL4R_wb zNVo0jDtMIkemp&%K58`ZT=l{;%tRYamb;tuAS?TQA;405-|??GaF{y%`-yo=iqAC& z6H!>wO?dq->tjPTkN0M6;A#Aq8fCqEDD1dRS@l|C)ASz7UX5d}*{r%+ic1Cm$t&`l z`M|k%%$xiw?SV03Ghpi>4jSWOXtrRdMp*mIwa_az0j zmq*!u4F-;4U^JX5hZK?_HB`P*`|n}#5b%P2GG*C)<{z*V73mM(-J|e6p9M8(NJ4OE zUEo~+>#u>WOb9KdV}~6edy1z+W?{zI2$)Cz3g><744xR(@Lf^i96{H(Q29LuS%l`Q zzUP^ff6*=B3u)o-6{I0rkC2MpMK!sg(wqjjJWg@mHL%7**P?dBS#7P4-;A1hDog#o zuXbtdUWY<^H?+w(^Mj|dHbjqpkA7}!O6k`7t?<$Hz7qXNO%+FF4yIEvXpe%G+(ftw zH>c87N{3kP&y+7G!Z@W>Kel;SCV5C9$V3jckScbg{^S~kiR&F?^ruFsG zvej|hqPu1b)N{XeB<<(4J|l)UZGE=a>i7D7j9_n)(FMq$;>%mmX1gEkm9N&#e{2uj z9rLhwp!q?i7~+3yqJbNbBE8@DTJWu<&MiixI@1V9+k@MsCm*22vsv~2>~iZX2{gF= z)Myc0x@6?>e6Cs8+IV$3BS)|92+!~Ef+@Q;ocleZKp*mchTSc{#rXMj+x5r~^#gqz zfdTb;lS3?Z!HU%_PmBrN>Hz7zv1G^WM(>^HX4mX`VmhrJ+lI{~^Somk-e>dc6|f)D znOF$w*YSzWLYCJ_SNm*ZDCATfTs(-oiy=);yD>CmA~Oqd8m|~1MWo}yDuA^3!bVQ# zN6`wa20f!MDKe?45~kwp-z+dXxP- ztp#0A&pOEsWfq`ym%exf5+d|0VZ+fzUnY4^RwGP37z`m=AAH*$-oVi6)*Umgl!?%+3rfXeo z7c<(uxGN)g@S!pRxrtI8KAsN%J+!=aUJHxlEXH~Vl+B;5KX#*Z=7bD_013w)C6f$Ga3)mpp z43onYXWR{ALlhH{PDao+04%z}W< zLUBW(8Fmcrb3I(z*a4Os@|GVpUr8mG{OtS0=jO;)`Ks#8pYOG8EcP4YzuF9eABP<+ zqtha>vUMLvYZv0Ru4$EM^+Y%$cQp;AqwJkCZrhoJWSCg&Mrz)+SuASY;0g|?_Q?P0 zyLezlp9B0gueyL>;<{$HPC5`YWhK4$TI)8a<2eN*j!$QWXC@I)&ukQO zI4|Y6E!|iY_7{e5SdAd#p>4O*+h}>3qeqhyxv16}tG9;`5Tf?FC|yt3!xYIsV};NR z^3_rd9ntc1w?zT*PIvPK$ebEA_JvM4MRcb(iID7x;As2#jbo?@+A=!V4s~IbYWJ;SPXv(a};^89snP&?3P0UyMwbVpozWnafeLhOLc_#jp`xkppuxCWxNMvbu} zA@gTXF>H+Cl|pehllA|R1FD<%g9id<(=inV|A8F50QxY^J67!TeJ5xwi0ahdxB~W; zCDo&lCr?LQN6?-fFxA82%sj}CuxQ@zYJX_fizIeQ6QJ zd>*<$^irF(0bSENbt<(>kfK_4v1D!$+TBoEGyIQW4lA;EfHN~S=S;g5PaZIpBiw15!!ls9D^()lXSem;B#45^O|%)R_My z&G2!d|1!g)nvrJoz0|Jq1>ogj1@9XM1}TSWV+6SIwj*B7jEIqvYn;o4BCM}iAM@C; z9$nrTHs;n!G6B3E->*RBDtbwkxiMkjMbWR=w>X}gHHEr|;?UVymvFuN+CxQ75qH1n zv&%qDV|!F=5chRTyZN-Qi_**4|7J5A>BG<+nCBE*Ond&m0||3Lc1&8e&+Rnep_ni~ z%P(AyH0IdZce_)6U2QCOqx&Xs{H~Xk$QZ!P1*@8+s(X(WsQH>+!n$PD!E2@rhG7P5 zYpz`&hkvw(imr`(Dc)KAQS{7j_4sihF@rPkFDml~s9R=9Go9sMNAAn+#oC8eL^vR; zxEKb{(iDxuQqI>uCGf*imh?vr^BGg)HQ0(=EG@XC^GQ7yifgzj7QPw&WN36og}>#B zjBs7);6VwIn!qB`>R0~T)af3X_tk(KfFtTc`-=bA9fSQpD@#K1=pO*pu(txQ2i#z! zCoVT!$y<>nqmv22?xo+|YOrP^4PV~V6mmQJ2<9HjA|F$E%Y9Sj(^nJV;HYYNO{j5( z17$J@)55yH2JYDI@4LEC)lR}ajh=G5W^JHxjISz*rIfByLE=C=F+Eyr324uPv=KJ< z)kDtRVQ@FOW8b@ zZ;Y*e6=uXG_li?wIG+@c@@?N6=GaNsvnlj+;20lO1Hx@pkR7S_o6OwLxBM=Qe5a6m zy}Xa89OS@0^xs-50vdGk!6m@a>V@Jhf6}_+4ux4|rvyHW$=rE9jou`lyc~g7TqT z_RXRxo!tl88|v@j!!L*ua^atL!$%QSVECBmRl@d2wfR+wouM{_An|kTF`j)gmH5hz zc4NPn)^AJ#R$?&9l(U~u?=od;KlZ5p=!Hw!Bko3;TsJUdiA8d_KNP~$jayDik zHKGgma85>RB#HC6l@IctrV@FZ);Gm<%+dp5T3&z+M(^=2hD>3Dq4VP2!wIt_f4haa z_QJIpte)K7eUNikR{N4Zp~8?H<|wH{?uMDFAY`UHMMV*;N2;MWZ@?Z^e%yPx0&}8lFX0V*Zo%5kjJTjVpi;uEc`$wu&7>Xrf^OrWEYw|0k8yquJJ!r2XxETDFOL3E#+G2n^eKU3WLgVL7LuV z%#S)a>b9Gy1kEbuqIgj3Ps1E}6XQ0H@fnOMIC+V!ilr3S0H`L?RpHZbPa&OJ8<-94 zcE9(5GYzqzc+DxQEGQR|lOm%o-*xK$93OFvr!p(M@sg|Mc@x{37PR3B*^R-FVsn<% z%Ggo9X10W#wRTMUiRCTHRqd;Tq028h@bV1I)2gh#uWS-y8ySu*m)yB19|J_{palYM zD>6|`9L)2S8Tm}}J+KL6v3HMB5+s*Cpy11{I(ucz-8FP|Y4^u!TNM$1eg=K&dP!3X zGor)MqexMPv&#s@-JqoVueRAgV1?@_D8(Fk%o%tWit5dyikH?0a^#Q5>1#HQgceDJ zEV|s~)@HRAM0{n6{K)j(0*Nu&xa)|sn;GZKBaFIr)Rbjd<1^O8#_;dsd4lS8HAQLj zpSTaV5$w;7*{bfp<>g=eXzPe(pN>+(-#vk?g=Zq^Z8I2(n$Ep?A)c9ODrTo*M+Yk| z5m4HZG|Ez#E!N0P!{DUxj&$UB=1x6OPk6UyL%ZI_1pJm z65zHYg$0MmJ{u`-OB-Xmn!*O;B1_8dL%RhSg9#FTPQWo@uoIJ2m7P9__hER{%=`B$ zcp9e}o6>&M8mHFl2ESRqo$>FaS56^U`+G$tXQ|*$_l0ZI07yW$zpEpzYh#HkN=f`( zjKp38kI0gyq00=hy_C-xhQ&kwEwp?EIOHI^3*3~kn{fVTo9E`CXS`MuaZlYaBwF<| z=NRI_>le0rp8$X%=K?71N;hE@#t?RC1-9%cAD(L|>MCdYyPyzX=TQJ7GakB_#I za=5JE`U=V~++3rONl>|d9202H_aZkcTUC==MCCBgAs@%dwvAwVdmPa}j-^oRP;XIcnH$gfiJL)20q9pKX)$(v^l&*N1Rg*8?@M;-t5| z&BsH4C}NQRq!YIwS67$=vI(xLA!jF>y!3K%373w81-%yIM&ON*zZ2Jvmza!#pHxy> z7Vmo7m*m~88|L-FglFI~W2JWnkFBED86J-gQ&7)phTr=20@NCl1aVWPYW`8Z+KtnS z;HGtYMj&UeW{?|o-#(C=A?g%J1RIfoO-_8N~j2Ap0ffE-J zy7`He2GR_SL<*Mk2^oBkE)LAm>Mr-0+mKkV+1Lz~5v*EhoV+D`6IDygohpg^$g`Q( z&k{BgIkGLH`zIu&I+^3yM2F4y`F^)%CnxsValuR<2#r8zBYRw~i*&ySZ5vmf`l235lXi zLBuMrgVp8%k9>;LCjbW6?rlpjZYIk1nRYJk{I((m(e2A5(k}rs9d+pMn{sH+Q#GNo>%TnLjWkEm z{+kkCq|{$sc->9J-XrpU?3A+vDo+-IU9g3aX0aVU_lz;pr z*rV0KbEB#8upmQ1JwuSZ^KU}9`fkX@cF+gO zV{KzfoB1_`o@grJ)$kw9g{;!M$shXt^C#M{u^d)Weua5Ox(Cd@w?M51CYCH0i(kr& zZ+&B{nsO=*t1#?hM@Zg)Eo*2K@eRJ%j_QR>x5B-wP9vz_1SJ>Q+{nun%fIXA7^C{J~uPXX>{i`zZiS4M__%)?dL|b zUpqk39iEMh)iDguf<>InN-SIBkEXqg?7DbM3@hF|?fsFgl%1~2@T2w4^Qwi1k+$2; zKb3{QR@+27M`tvK-?AR&*lVi~%Q^U&2*b9W2Jv_gm$q`4oo>4P*M8&p`4Y4SFs^Xf z{Oc11kgP)h%NQ>S$Z4PS1T|CUC5`~$RyVlm9z}GhFmgB1qwPF6=l{_10&WGq6GL~o z`HqqYEvPv{pcRx65v(|-T>~X^?v}QN@xoBP$~LR#;hmFFcYPMpceq#R-v=upL`)GH zznPTL=i|kk7v%^Zw@Iy;)!Ux%LKfUp@cZO5;XU#FV7m%u+a$BTM0S% zWKLduyf@G)Fv3{%_K6A9<-MYkF*y4&=wk7Qj6E$q6Hpo(eHoRcs@3c-FG*Y3%hqTu z%pck58W>SBZXY&q2(;i<4d&d4mf8z7;dQY-9Y3~;fz)CKge9uS9%)4X;-4Ynz{a_B zP-#XjhBbVtS7gbWSw4DuepxhoScb6sNBDMNLCJ}Uv}m!)fP!V{VcQOtN%zV6SfyLd zTlqHSh53wTkR1M!dRR5=0a(92T}Ovwq%B z%!XF};UU39JVfjmD%ZVVU(jdNsQZ~eGZvze6j}m(NMA`AyxZlWQ2Dg+dkZKGd3}!# zepl`%8n&J+||!18oH2_oOJ~HaW*y`i#Uap>9$$O~;K(d2nNl3yjw( zMR(%$#4VOqrKS}*+)D@K!*rS1OhXED>wyV_bNeIrGwHSd{$Gx9)ta*8p4Yh*>X_Cw2WSOC!ha771F{W( z!r34(Hd54kRm+QI4fuBxB{vATTm`$z5^!Vw`lAAH{cj6DSO^@4&>jeRk6>ZTNq2Ds z{{mY^K-&VByAgxoY?{vjc{>{B5d|eh$CMUQoBEtSp+a0$DiqbYc+75@*IQDVaZqAK0}%*h4At9Jji5}9nZ>)2QH>IM>jL*bWd7_ zXeG>4#=&KKzq+I@jB^kDnb)D$K(%b9-9-U4aVweuKx_`3k{@er2{my%#yQ|aUa>(R zfs4e?dj!wuxgsc3`BBitI^Sp&9~V+d+6~L@u@m|wWa;5{w{5itGk*Z8n86}6gLZG! z@F2#HKlEp?p)K`sC9KsZ^JAJR(z?68YH00b=E64yGk(-e1$}?Fl0Rs`Vu_=ku=wbU z9fKYHA(q~LG@ferr>*5@54IxLFt~_6Wz%oSW^=nOf(Ok2SP0j6U+Uog@g&w98(Mv{ z%#GG(O^f4T$!B-=lWn1ZP|Tnq@D~O-Suru9)GuQ&?#x(zAAw{br>cAoFFfX<6|pPlcvT12nr{_Xv#b18Oe4dt-TkP+O;_P)F?`2J!<^JTDS2;| zzipMVtNg zfvm_SkQI@x{P0f=_y9=(H=HT&hiR9q(6fc)t1Cn1V8_c@+{-{pS~*Q+yUQ(WyZF0!HE@gVK3F#gy{}J}^z}UUF{+>+)>iE|0P9wKo zUDT#alFf0Cyw`%+slmXrvzN36c9!RlX{JIAZW>8^Z z4@^|C@NNmEkvJm0yesBFNjn|&p`tssVxz^lN=vVPbYJ>mU4r@k-~MF4>BF9t6yXyL z^>}UbP-T;8i(@-9(eRrpv#6gH%Dj}WKZ5QE=r?!)2kQP&&r z9p2;c>kO!B>zp{kGRo&*-*dvg>S-gxJM-0U_U+lSG^_2zp$MA#f)#+qIp%R=mrdXu zj^X_0`obkeTqWDt2|9P+So+Y*w7v)wj4A&IlJN>_F~guibWE}cZ_c9eumP<1?r~48 zuvxohyu8?3++$tg5=srZ;^;6%K6k7cZCR^StF7eTbj1SHqQ})KKw@V9%TM_|B`#YY zjo%V}gazOAH#uB>i$%Vae+rE#2sAufQ+ux-A~aJEn#z=!nFW;#ICc!j7m|}a-{o@P zx_yqO3~YQkTzW($MtazA288wLjPw1R7g5+=j`F%EqShG+mP=0M+xvC7rxLzy(GTGO zOH|iAQ^a0X;A8_~KFS@N8LdO41Ks$sAdX3@msCHK!@9mL{MP7Y%71uzePkj1@Ht{> z+%Uhy_5T>Q1cWP61yWf4_U`$2vIdIVA9+n6qFz}NLeDS)L|8FrG78kXt__dp1!}62 z*(B1>fvr*`*BPisRjdR|n|8Yd)k?WdD5>1`z{+s}5OM|!NR#qB56u`IQTpNSQ5v>H zw>A!LF;$Ke;4Cw4mg1sZW0vR6#_t_gaVuU9g6b{Ba;9s$hWCQ`z1zOi?h5;q{D6n8 z(H7OM&j9Hc>2V#NqSxI6Lx}I&Cp+alXVC`ukxT7mT|hVw9x%_rP(=TwmzVpRMd)mB zIO?IyeWClXlpn{X36Y%5ah_JDymV=c&^Z1kl!kc{wQ|C_LTo(Pvg-5s`fu*+Bq}$W zE*Djwwz!heQ}yFI8`{P6G<7VMT+Dhi&zMJ#M6l*gaiDd!Ai^%))$zSVlLPD@+q&ghT~?RM z#%%@Lh{|hn+e06HxSat?(tB)@k(H}!dz*Dz!cO5L{5ru(M_9xO7~RZlUATi{T`BV? z$z6tuzAsaS%yVAEj3p{B2(qgRCbf)a35Y-Az6Bz-Xn5>`8u1@lyB*Hes9(0HihSJP z>M?6|4g3End(WsQ*KTWA5g~{|0w^FQAwiH$7wHfL0>MoaQ32^iQ2}Y22oi!6A=H5M zE;dk*UJaelo7B)lN9m!45a5;PJkR*fd(QXg{hND?jQhvD*0tum*P7EE&%&eeNwMfj zS7H-MC4x1@UdpqFi|NNUq)C)yr+Ydl7F}eQq=uG79ds+60K=p1N~GDl-F}qCDK!W4 zJ+xR-t0|vl#;@cpdkVv=3NrY-s6&;~tL8^!v_}?>CM(~weCB&VE&Af!?G&tVY@YU% zk+x2WV~-Q4T+~4uAq5J?$9_pCQM;99Gs#@_e1Ti>oaleJD(7^O2ybQALq=G=^oz{V z{2Mdr(eg>&;i`NM|GL(W!iauKi$~1xNY=U_J64HjC6G30Dgec zhV;L_KDoVtym_Nh^BG7FibJtL(qr`6H4%lR0gY`XU@XS8gvTBxYkJg^v+0Fcg`hkEu#p{lj<;uzCn_ z_Zu#;6FNtI@S}Wy+j1GQGf($T02Ao^Q@|5%rTe4$SSPI8=qP{_&O94Qd<-k5N$GrZ&78;KOaTb6+fM`XV_t&A zNu$Y0?k3RsOT7#MxKiu_1Z2u_y?afZoPV*uCFIp(|YSdNY^R zn$j$u3{w1xH88*rGve|KTs+jE?`^-?*w=Z3Dowkziuo&0{gNS_%262v?>Kn4tw?poOZKYj|PNVri+>si1qWc zP&rNLs@piao^Z8mVYuRyoV#r07Q&&Yb3s1#dLy89)oWwkhwG|5BR8t44HsBBLTK#v z^A(!+EPWKS*z}eLQLn}u!22Dw)fJQknCaK~a@YY(DnU3QT0o9)?I< zZC^k0k5)w_wxx~*yt#+?1g#nxWMIl=nKM>@6tWvBiGQt~t2Fv(&T$&I{z<}GqfEg2 zDnd|=uiffIprTegp3M%JB5<9V#mJ?LbFf%`Xns2qtb<|&YAW2VtgeyFYS%6L}vdT%Aw`YeCwg;o3T!Zm@n*gs4^ z96ABopw35BAf$c%t{M=I4Vx-=B92TBt7S)RX@8RD{LT~%#t#B6(83Luv|3X!`RBtY zm+G6NmF<=0BV{@Z0nb;SaSlu>96s+|O%nGr3$2kSVOfGzx57V(qfp z?AvoYXWKsQZHZ(cSLD=nM2#Jf5Gvn?Af(K)H@1_kdvH9R!S}K#lMkU@p~@e}w@-QR z+62fBs(R{}ve)oY`w%l9B|nmMajjjDsvZ1~qgcr|%nG+|L^y@hbR>i;s)_$2 zYW~cuf|vd=(1JG||DV)pEsb+NR|hUV4@X}zd;3ZEaMk+svGRP(_&9_X6M_=^8d`&R z`WeCpG6jeuI0b_I=ohvZyPV`X$QsICLwnZ`xwA&?;=~Vt!5c7!c~7b*-|sNJ;a09W z+`a&5=-~Cs{*lILDhoyw@S{wg;RK{VDf}L&oX*6Ib>HJ_zmNYrCKYqaoRY_iS+r3@ zWWkLog)h-sQlitV*1p7aV@cGI?&UT)%gvG^fZ*o4Q#VdzaQVuo3`3|3@i z;(Od-8>gMTdpIc=i1%1qf2~og2G6;Lpjpkg{sQa z>}>#U{=ygh>=kPTY+$@*OKWN}TB?o^sK#Rb1@Z|b4-g18vc9=R8RTJ%{W&@$!{E1? zr1omk9T_$Fx~uMp8tT6llyNZrOIg}6)W3E8C(D=CoC4lJ(KN|=n#W>-0OmVM8c6> zG99DWtYSsnk>5-D|2bZxiu=>b`|{6kWdD2G00g>o8_EcJr=v#CF>v)Tm|kCfNMJ}+ ziKF_tNu;F=3MQ}?6Efx%Sj8+5UW2T%@)~~Vm37no?Bm6TqB0N|wgIat)3w#LmH))8 z!EfZ0!|cx|QUDtQ2)Dj*_R0d=zk6^T<6(unE4SVn{C$4|chs*Pa8aQtACc$}743TLY(JTy$ae zrdZ+07JWTv{HmCuh}GZQslWi6N~ag7QFoy>&@a1Cns#QIFY5O*8EyneQ*cH#0YLXo zv6U4!I~=k#U(0KUEXH3MGxWZV9Nps5^V(~3teYr)%9&X=ByG|I_Ue{K?K$zT2p>d| zcu`ZPJ=Qw`Z`_)ggyR*WCk^2{;-l3JdELyYgsq}BPsSnXE1UY}7jUF<6~1clye zWW2H4V&cTo-^t0nJHN$)BGi(OH(uQk16kU#_{vdc%Ztge*v3p>x1f5j%TK8f`!;d_ zfxRsC0=eHUpSL<(CYWHdJbM*&pN3M{^+q4{QEWW_u-Ak>{sbnOgP=-T?Tiej8*7bS z1xWjAPJT_@oLu>pZk(bhY|BOsNoz*-*(o~qHK6~4N4r|y?(kn-In>9+-DZ>r1v+H z|HVgFzaPDl_W$UoQN9+8wyP&MH!LggM)~)o%9iWit8>^W?g&Y4>`)R&lBnGsP>L|7 z_gU2&Vdpxu(_4D`x~QCI2o}uLDCL=4IcF#TnKplLmR6K?fY~Y4g-KnZOOYHbi(US| zrpHV#(=Y8*UaF^VCbVFCla8LLZ*?W_{zs=d<~Z*k9NAJ86N6Rve+Hr5k15tMRd$IP z<>y2F$_|w%P9toJ3Y-8#oTEET36T5SXY^_N+IbyF(Zu4}QY~ACrHdwfg?%g!G8WsQ zqk96mHLO3Y<{!p6zGI%%emA%T=6_oNakz{*IX5Wu$+h5foo9ifa$x`D>NHQzkq>W% z%qeBl{TU633H`O?8O&*~(JSg4eCGyXt9kZ1>2Ou`-DAuBTGryukRQ{{q8CF}a)y!w_rmSYq;eAAC?IugDx_mf|U>o0~Y3isnr5WJxLu9h9x8X%2(JI z>_63lr@MKE_*Vd>F{Zt4nZEekkBONSA zx}Q&M1u%N{ztkB8i2&G82sZUH*fe|R2^r@)4*rVK|D>6ltu+Ia#bE|yr@Y$f{AX6;AZ z>xvg=SnEe)bhkl?c0vYgplL!p-!Ts?eG!eC;D#NwG7Ebn&^^6c#L=IbYSf>VhfPTf zkup4MVM)o!tSF{VMrmo*V}?wwaV-kluy=CFV*wr z^lBV9y1QrLZK)H>{5ws=ozd_s?lCqmL*J`lUFM}tW*m})C*TL`y@E5HjukEO4Nf^( zK2N73xYx!Oz}?sl!V8?55F0zwq2Ok>vFVXUAlfeHzc9`^Sk{rA`bD4U^55qp;ehUC z*Q>|7B+(sp--C%VNYxQn#oPbLk@SCH4zY{%(hy4m; zBdLS+PHRyv+d+odVGetu_C)+Cn7ne*d*YVn+q03f7I-HUW^~xgi37?qd^jJ_=oFme z+w0ttDKn?U1vR&QdpoLrqmd=LojP;wLJNgs&t~q&56f%U&If|>TSPCq^KV<$5R?Py z`~T*#3$j4IkBHTj$XzaeZ6|KE6h5Vc2@>^`J-kD+N#j$5`43CKTJzA@_ML>1Vh&Iz zK--$9BzvjFo7M~OE`TJFGHN}FnChG>Y6)bEgG{WJ$~c^+%?0|II%D+388k={3LOh% znr)Hw>8e=0bhIVIe_)e`S^Dy0qd37AeYc`RR@L$4ycGtk`l`<_>{Xo*FpT)}?}`-@)kproQ%eUizYYU6tg5W2OM z>M}o*0}aJG&`Q&c_=_D~Y@7Y6grC}d_!Zj60^jMfN47keS^!D2JcS$21(m+u9vUq( z0q>)YZ#`1HV81G+SJq=XEXxn4*F20{UBVr}Nt7#`g)P0K`M=%D49=3}Z$Q|6#s+ZG zy?YTPaxw2zi~mxEtuIgAD__PFX0 z%cb;+m2ZxmDgy2(;5(+@7|A64V{DH1DATHEA1yAZXfG`bo+G3g4TAOP;9UpYwzA0%jb&Z~rrK{rK zQNKk=pe-bTjYB=0w;Rj}xYM#%eX%%Hu>7M|e@H`U@ZOOjpfLsq(4o)7!~~6xUAZ#^ zar36<_W(ZsdWB8qqEGG?^ZAZGV2!}Lbp!SLt=4;o-8qphvR?5+JftqL-8_mq=ex4V z53cRy>iT7Hq8<&yfzq+YgpvbiXxG>sbSuk;X0Y&UPeL=$iJ zy?iD~yVpO}8544aDo_jI!nZ8bVcOW|+mowUeS&m#&Q_3(mYBe$zi7TRa+_#<428cZ zbcVlmtv;goUhNfN&075Kk@q%mdMng2ocxzF zFcy%5@&&n$f`}Q5g4S#~golh(HziC51{e<<2j zj;Z@JSBiDzj;OioN4$z@d~6j-j$BZ+}KR$?? zw_bEz9n5$&ZRR;y;J+HKV!RNFPO4vuf|r{ZMyI7sD+9JX+M@Lo!oMo!NRpR>B!On4 z#~J;fL8Djc7e>mSsBC_5nb_m`@9z5y0f)5jO_#Y-Q9YiNoP@VvlHSA+x_7q%AE&@r_fKJYWiCk5fJATu@L6qFb+c+-$(iqhdZi_ z9V72r?I2Z`N845Dh3pc6K?+*YP%BGKIc4^@z&(I*U_^xg@sNj2enGZ`r#M^%`2&i! z?RMmvP`7;*%O{FjhdewKW>{Ik;<(MTI2i%}G60t|+_x!rHR7k9A_VR9f>kQYH*Df`*)FEI zsn45fgTS`)bFyTEi_yBr2)~HULU62Rd6#6fWDn2&@VdJ;^U|e=#kSh8Yq_~hHG^8_a&Xj8M(tE?3a%Qq>sviy`m)Ul+8e25oEhzGj<0lzMWHV@{E zU5%X4+>egHY%|`s+PT2}TUC0%i6B81g);H9Yk2{axjF9-tpdOGig4E}IY)1PHs~d{ z_#))fihH8xt_Kcz5pE5nHUupMGkxIreYZ@#4Ek+qkYDi|-{h@F^m)DjHwW?UYd%rA zTAkneB(fO;R*i65TkW#z6mIk0aQA=AxabMyASX)alPYDX7^JbKmqkwb{h9#`kz4xD z?*YDbmB;UXLLC>fN&O{*zt27{ITSC4FD(b(VB?H0VX zf}{bCsXNJk@4)M}uE?H}g61h?KI8TPCCSAcoy&oq1xOpu%xOm@ViYk$~vK`Ck1gY;yND;6uA8hT_ zvQ`m*l6BGLCf4XMYd5c|EZGN_VLCjan#CwK(c|76$!s9~x^{2?-^y*8Xg4^Ef`IxAtJ# z+_R_Z$89=%;GHEx!H|AN_f*a1|1}5Zd+A@bJ?X^viJ{g)-g*UuGECW>`VRpGp8sUm zuD1dE`#5_R$3X;Prj>ynSWWsGv8(hqWzRQCFv;(xpVyyv5!B95U`jHv75S3~a&83OLR z%#}K=h0<%~e{q=H$Sd?Oyt>);WaEge>H_@-5^4-A5qMI*ob+1fQQ;J927)NQ~?h@<-ZQbKFoWRAXR1867Hd%#qHmEu~~vS=*?(@$=b|nt!F_; zj>@G6>D}j*pD~3l1xZgXT@qApnb+wN zU8tGzSb{Waz3dPO6cji&Td)Z*4aVU8nD~~2(+bEpcAPe7G5)kDNHQ|G_p-(n>pctv9!Y*UQ|hz7XKWFpJ*uWM;@wXJu?#(-x8HwAK_3IvvAlN8)*nJj%f;RIj=s)0 zFLPM*kh-4`qj0KqjU@YomJF92-`@~^(Y5OlCsF;r7e}HAm*C+igE(UN304N>T5Rhw z%;|4T95tTF%@u8102ZfmrS82|L%R!(Y$#F^s+;PRs8DDZpYkz5-+4G8Em+x}6*FG# zIJGQqMBFfQ55-B36+fNarky#JNt3616(vo=V%m5;^?H~OnqC~!{IBFJeeZ?@(qN*^O6eWjr0eLkJK>rDnd zLGMj-G6EKoEIc}*6)e4mAMW`kIp6f?S^aZTdakrgmeh@L>)m|&b}EiG^oRdu1*^M5 zSXTOArae4|^USO#Pq+Tu8aKdbvjBbErH8Kbtx?0)r-OjCzy7dP{9|iX&z%1&zAqx6 z0e$jl*5tIdgscD111t`}aL3jCd- zC{Y(C5Wv*#;w?iS9ETu=QhTd@?5)bdM9>Pm?I!NPQHibabTF%R-Cl64VPwr_8kq`>ZKo?eC9$JIFiv z1nFcUlk^tm4SEQa-d%mof@qN8ZqV*Z7FzQW4g_PV3&)r3#+d8KCGafbC3%?TFFWpmEIy6fYCE6L zc`XWM)yi4z1lK!nNBNz2>!58)T-=F# z-&=5~IWv>zgQ$|S-fPK3W;9I~7;rtaZ;LlAvdu?bEKx6$6FyKprqD~=k*hhKT!XL#HJvIoj5uPuc0 z&frEB9M`2;FrwS8DBuF%E<%N?SXgeH+;YTGwV3H-g)ig>&mL{yH%C6Yhxwyt3a3hf z(Q+jo_;r-|46D78yhr%VssP_;c*{u(}%!AgP0)N@&zI<@u%1gYXKg zywx8^!HTPmg_ym(64wGNG`Z!*Ky@F6`oy@mlgIA+TrJ;6`p0s6u1P}!kU|Vi^{?ruLz0C8nDx0>fo1|@H0*t{ACUz%0pYS^V?~?i# zEgQIh-mj0RRw>thzx_%IoN1aFee|CN5CQraZ>(|)SV)IN%bq<>{)O1}0ozS{+=%hR z)=jw9`!RmS6`>$$v1@>bc=%c?`MW>8{w9P9NX@(J{ntLK-5HF(vn7U2Kp2oTq}de?Wv%uzf=PpEuQH*NHZxozp);Xpg} z6Of}czx^doMkBh};~`_igc&UbW5eDt?kBxI-TZ-YyDVRT=N|0~j}d~;=d;yzd%6)J zPTY5|ARJz62_TKzE884U4`qf@Sa9-1i)GNY`fmXaH_ZH@gLm7~}8U+<4O>fVtd z!A5|uyA7S_-2n_>zg?rAt8_8tm!8he*G=i5W&N4%7eOS(LBDhvX%L2{xI2zWN%szz zTuTt-BDjibsl7AUtA$CeZ(W~>bFWS!JACh=>W&7VG6|>O!+l&}T;s6sOKouVYl^sX z1Kz;d+ZgV$`nFkh?>jdQ1ojg&kx`vgwAoBkF-X~k?k>0?)U4*WLSudYh?%V_&?gPs4JOV=E6=c4CpR?4GH z6>c+mb5=`2S;oZG8WY3xNf|k^!$i{S?vHD7HcfEnvcH?|rr4!GgUfG#T}V4w1^qMKFOdHve%6ENJ;XE9;1zHdh&s-XY^ zd6gr+UGx7?-oY^70GDkV+nm&9#&P32wa4k8D1frE{#ucG^0fcyv6dbb7@G&sQr?RR z0-zD33#=;laEFNeOpi_xtz78Wy*{CVez2l|k>#tNfS+0=P``B+yS@!NoFbwMrU1rJ zW}(|*B8O4XPv4~lgrnpMKwHjAKnlVmp&=Hi|^@xoEs-={473t#~O|)WBsBW;q-9Le<|vX zLX+ZLm^bh}KyHFDJ5p-)plzYp6bK-2lEo{r*ARSkJ_reoP%1&RDYKd_(DL$LTV1+M zm%XoQczg_21Ucm7q6875VRX1pux8hb*OWgw^- zwRH=ae*^XEM8`WG3} zo_)cO0{&PWWgc>E8@KiOJLz_LL?3mLoKE1gHgHQe+>(=2W$AZn0yz`p;MfT1CdNd< z75we*OmnvRi0@d<-W4UDIG;}B{I-?f38;-`1tG_}C6seApzrsMGaGub=TryUNTHT| zX+$;uMI5 zE~bO^IB%Xm4fP3Rt;$ABGT0S9-K<{Dd?DPnmYA62HdtTR^S1l0AA6yV4KlaJ<5!Ss z@sf-hfvN6LZ^r5G?fYQ6Y*+WEUB3IMN`f$^k%6B!kLxppekQhllc-<(+Uy&8B(hXb zuBWc#B~+rM4MS=U7O$_e#2T2Nuv_e{N=Z0MI~s$)(jeAq$t;_O5XGKK4rq-+L!U*l zxtv`sEuOu1yz#hOrY~RCcC6%ybpNupO%gCBXE1SLw`~0f=D!1e5Gs{b|8Te8zfdn? zw`s^#-Obj`S@hozj{tN6lve*3h}?4GJZufLH8}jtY($k*NgqUlTY1ZQ)e5 zbwC~M|9p~fdL;&+UO*nlbOSVqygK&;MR{~A@jA4SxQvT{d+TfWG<2X_v5oC|pOm(F zI&zOkpHgQtp7V-K zj)yr|Zd)v5f5hB1K|$-mJ4d@@G~{Q8Fs?&}F=pB0G0Q@K7S^ARyEsNcuzn%P0)(V1 zk;8g>_DjfpU<$BM4ON8ZtfS&S(maIh+KuThgDG>WX&Yn_pSW~eU0Z7!A^-ZbptlmPLY{i|)DwGiDF^?(beO zK@IJ^UJE*Ma}TL@f1G9e#g7_Lse&4@nk^-+go#Tl;4TjnC+%v?1}AkV_Y#_fr0dRABt5Tezrsbw1ErCjnru6jaiweY@_ zV{^6yX?T%cC9V*?9E85dX1LCA5ZUibzK+ZJYi&+#{I9he3{qZGytk7G?_6m~ICNl? z94#aTCP$FC(Uj|BQ(IyTWI!dGi*`G%FC^}@QCiehEF%2TGLPyGBa`Arl(`oAo42|4 z#~ZD9xY(B*L;J>GP5b05rSYBq9^|xGi@XMQU)D@IQ{B5~pX)A-|372?*0F7@nuNxF z)TvH~)4`DDhJaq!l{2rlbuOPXi)@=a3j#ns?A;a#KHS%dvggLJr|oTnhHA&BfWVc| z8V76jXMa&~IR1pNWdF_DgYY}7B9Z?f=wLcbZxIYl%-I*r{+60iODp~iuF&L|cv9jA z6an^*&E9#YS;fWxszE=XM3~;V4B6vz&g+vXvGk8F>n}V2JSa?vOyDbwEf7Ju=Q`?9 zHQ(1C?@ws`U~`B}Ifvj76aqTfL`%s@_!a*K+|4=}KxQrD-Xte$N^uIi3q`Zn$F8^g z9uV%HBujeqY#7+ZnF8Hnx7yMBDj_Bq-=SP3yH)ZJ{&k(*pRDhu&?RY7K92HHm<8WU z1D00g0O0H;aQV;UMd##Yqr`48!JmuibeRm^` zr(m_OS2>VV8T~}q(_Hn#M~ez$LqB}=LM^p*`QAcN%@fW<(Bqkf@kDX+zt>7j!?2AE z*}+=}3vq^t*uQmDS8_jgvL>?)X7Z2xbpk?!g^jaGQM862zZ!ez*`riHGyt4=@7rVTrl}kCe>cDSQca2G@6>-p1t5<*Xa1x1sc7V4s`CO8BxAHYrbY( zg86NTQDifBH}pGhV?6j?X(qha(}sPp7M?xzM5fubJH=!pqMFnM6Eb5yWHeURwiVP~ z;v&%!zpsvchF_y^nJ`_5C2r|3X}9t^Q+iB7ov7QXS*;o5*4Bdb@qjdQoy#v(X`hx9@+U;yXprZ*Qp zl#u1zX+9TGvp=1F7hdFe0&1YF$Lt{7@Lp>tO55!)r;TLMlr)S}-v+q8`AbxrI9o6! zfdc%xXoWcE8Cggh&<@yJ&hodaZ(#&`{-j-H-K*(CyIcFWb`-c*`!8p_N)!*QFK!UV ztp>wX`~>b?&cC`<`Sfh>VNuk4Np|+;&Xk0R*UV>JcXqd?&s@1zr)phzhT_df^h&EC zDn-3p8Ncjmwm__?J*T>G&Ocm3qvGhN=`g$GF}bTo82cN?S1pLl&!)Ls@;KW6>~%U` zQM(~`SWuOvM15H1CBA7H5tCdqUn>2}*Fuu(cbn7kCtUiN$>d<-6(_j~4}z?_TqU7r zbPUE(?WjWE@rvYEIpVPs;lI@DirGo)?DFhWt!!;%RERug1CJJO>fV^U^XC50d}ygO z1LLy$;kSe8Hf<~1#Kww~LWl5MzWJF!c0&~-cqe1Ea9>cRQg}tH?f3R11e*yZ#TfC# zF1;o}241*Ur*Q6-Pr=Sj)q2;TZLCZ`gN5%Z(fVdfRMaKh;No=W3j>UWx7q# zVY{JH+9f-~_IKBF!&#|_;F8tyW|ph-H&P)F%7f1HtbE0B$VyZxD$jg`v?lCsM&>?5 z`WBW8Z~XbkrfrHg1$pk0G_QJq8t8H-1+#4^C55`NT}f;PDKD1bz~p856kKJNGmc(%7{*2_nAy zbKUTLkcs6Is7D`ecV4J=(OZhEtCd^E{^Ym*OL*aP#+@aC1MoWlfKI}^0|9N|9mnH{0LBRep;8R=F$IAtW+7GjW%i1&RpV^+!OGzu6N%)~T=!E6g&Adfwg%BT?De z-FH%dJlyaS>$~zB&8rrkBf`F| zFu0pXk3wX|&z#H)U(f@%$ygk{Q{a2P^djL)Y4+Rlo`a6(>fdR$3QV`o9k{miIU3P8 z{Ex}HceAMaV`Ho2@}aa%MTMsuTdF>ejs>fz{m0<$a(9wd^F2QM+nDOuxjNZSG}-6w zWGRYL^JA-tS@#YHBi>^D+BL?5uw^L_5BdJ6fq&y~N3|ofmY!uuZ2#*dNulI}Z6Y

    42`}lzJeH9GlYaK{(Q6Rehj*J!T$L*uHV`(+xWI>U@b0z<`cAqBG?3-YSFv zGQMu}li{xa@I#ZU>MugyS*o||1?Qp=G7jG?!})!lE}svKjd|n!PhWuGo{knVHjz34 z&cb5(!R^AKt7tt{FHlGeT9n(*h zq!)tDzgSbEAR{B^Mb|-<(DH$5Ex)W__W>}t>I(;ipL1hAFU^^hP6ENF1CVA z?5UUC!$I6=)FGFI4xgZBj zTfwJiiy`B_w+kwkk|$-rD*Js3s!fBk)voDVJrb(W=A10dW53sv&-jIN4VJK66t?ru zOjm=mRo?xq!R++NJ4KNx2JM>#%!V z=2G`@IvVkI&0hF)0#oM5HNW4XRb81U>yafwN^6gUzLLZo(%K!-EIxKLYEqt67c<;t zv?#lGrGwMQ#_2ny#X()HZ25bK?$J=)Orpyd5OZTfQ_1kpy@zR8*BL$}n65l`p2%5vKj- z-@4kR`CwMRSB{|f(gH_>osx(M(;Wm)@zbtUf;Bv4lbmu&ldA8LUo}elUz76B==T}U zY;3L!;rsqGWX5uyp5Jc7wvwS~+MWh{!+#oT5qT$D!j8Am`}FvLGazEYy!^$(*cu(f zPakBq*=(`&kIMF})CBldA73`wh)qE-60$R{pu&N4u|&kV1=t+d-F$xa^M}~(veHY% z&kwF{gBqhDdw5s=sMc$E9Y2k1K0VP#`MDM7c~<|}jduOL-9|V_+Z}v&bE_jpdAL zctz;Vobt6VT&tcoE)fb>y}FGaP~=?F-3t*hQP=Ld!i!7jI(-~?{)1GgrH4M4GnH**$irVt_?oN)R(8W49v z`i%VfNvD%)*JwsZ96hR!E3Pp3&n3t!hta@e_^1ImBBn612*61+fFCZfU92&bX7K6} z)$+A>%mOb3R8XNp)G1AP>M}i>CmjPy>oJUXo1|_d#d$3B@$e=8q22l7Ca+^{;BC3p zpu>;;j-x=j?=Iox2)&)RjKx9=VX#vyKfQivMgYQsj>(r6lHe6$SzHv2P4PvAv)xts z_VA@5;lc*LwOHa>Q)o}Vrt@~f={)U#nzwCHATqZmD+bV5YijeFUzAKg6l0ldses49 zsQXHrQ#i@$0^C9X)u)fqez@=?D#c}fDDk~mU1_r`H`RkNvWK~fvKG6%#Be&yX^3fR zk@Y*I0WaVsd-yzc_)f3^sJZ3*x=HTkXRS@dn%=SUDe*8C@pU(c5++moC2~v^sPHjp=_J#D)wYnD_-_s*xnH2f3R1W zaoA(HZyB1?huQCDE-tP}(LAx5Nt6Di>t{dnu{^NKwU+{or4*6=w_(%ufEL92i;XP^ z{LzblC!gkk?i4^T#F7CP^=b~B0iL5U0Xf==*q7O!SJt*cj6JHI7miM~D`}(Jy8p z+sC|`{R#3K*Betvz{^bh{$qFSpfZ}+alh=XtIt;;k08&ZLD0@-A%~i)L1&xyB8ku@yqH3XW*y-m7sF^qs64MkWJra%;cZ z#yjK>EIk`cUs8L_s@+$<=3G1R==3h@qDftw|3~tMeOTYT=gL~IS;3Amd^O9kPQj)! zGb7exX}e331lj?VwG!m(7)`iJW>wczu8pEABe+9a_uFAOj;9@(M1tkzZa=bcQ>V!& zl`hKSvK94>a(YA@Ekc3GRlmp9iE zj!6oKJUfUg7KmOqxYq~cooCiaWDTZ=D-A6()1?A5-XnT_^|9{zxev6=k6|ifx27Yf zxD{WZ!(Ss{ftvgQ&nnx`vKgUK8AuVpLyCaQUL&~TaJ`((dGM*V>W!lyjC#AOZYiZ-Q4QKXtTxC~!@&li4N%;2@(gH*%6RmG)=T|1H7A_J-{pD^YlK*t|yE3HG^tW-+9mvh^AcVTc4@E^x z0SS&#WB`A4er-KOP>L5GaK3QL*g5T6*@8y_mC zFb*VKH8K0bV8xJ6(Xb&iCFQG+ZSM<5y`s=c*?!*G({NhI+wy$f|G$<#?5w2^LmdqN zr=^F!5k!o?RnTJl3)9A`h;>GO0{iXuYZZkIX=#r$y-h~Aym}q6_(rt8O3f~%h*-TW)+~>Q^(kRoXV(j$J6bVQoYYWccPbfhRoTXG5gNn(@ zI>2pi`SW_#=mmBX(`V<~gYX;9ntt=`NyJu)>_LFu9jo$8*@nFWHW6dne3y@_$R;Gk zp`B2s5qGzs&&}Wttk=#Q-X|bikvwseIcaO-)$pq$e(ND}sQk$}0o;1UHA?2*M{9W6 zg`;dIf~`vcTTu~FfDm(R|qb<$j9Idw){`W1W*G;?5Ae zdTi?Wmh#)p6~)qs?kCN!jyt~v<{@RSv*TAszKcrRMrOIcz@=x1(H>C_yl%E;-U>@t zt#-A5FZ<5j&xn+DBnm~K?2qCO>VM%CgX9C#u6PQy0{#zU?;V$f|9y{_Qeu{ZmRk{U zRA!3iR$Mqr%gjp6jhL32nz?6AaD-+#a?pC2l?(U8g?lv&(aeFRr8saRu3x@?@9%g# zzOO$%|MCy~aXHU>&b{{>FY?`3t=9H%!hqDnJfByN_^fBy_aj`JB~6pZhme}XY`-7UxH>C+1vZ46pGU9{e zR=A<-+I$jzqVzYSSe%3Yg4m~VsdP5J)}$y<+hs~8t$pUK>q}x9y_tNzku#*cX*iBY z3zyCMB14~YF5&ym`o?q8sR#Kv)`Xmo5W&AL?T;Jc_KK;gc z#wm&3_#giAGhc3p&qRgpCSKjTv8q02ci@;iplQ*6rwUnojP}pu1>e%&R@$VtxaRpb zU9B#8#C+}ihmZHQI)AEC^U=HW>e6$f4x^A=@$yAbC`x+^Ha|Zp?9}d$ui4Bu@+(N0 z)E(K|AA7@%WKJ1Y7!aCNM*m2jT`=U<0E$}<>@CF2U4F!Fz zig_?!hgW;V{i=ui3*s%BuD&K#cNHihadxa%A^QT*U9zBS3#w4VjYkdZz;4|ZZ%*zJ zyn9@atTuXO_IXzxSPxvkq%I!v`xR_IC74cN2)cBpMkL47{U-Q^ME@tgv~IuoB1+dUe#PjCW`B(bnVT5h zuR6!QCV7n$NsF`onHSlkU^(fA@Q&bogsPS@%^1$Mma4`SNhuAuF^sMCSNL$1GaRD4 zJLUeBXGSOpbbhy?C*K_9*y9f68BR*;C$&mAo$q#Xp3_;6tM)TD5|+7N_feb|K_80v zNQ{oOZT}~cHbazijGO%bAbPXbum+)tff2k6Ob0_KU=tGQt~+&U0?V_nF!eLtyo3=P zTUNC!=4#{O1N+2#Hl_8A!CLoq!JgiyDIW5lGB-G85+D(^vo|Iun$PZJPq{7awI0bK z8&TCLCHA&$z((qpyY+s1-dWdivf=CIz+zs<{A~fvw8J(2`KSdff?VJ0=`qkk} z{8EBDxyo(%Pu0an9&7GSDP0^(?CS~Y7WTV6?xqt?%+qiQomX3TqqCHx7r)e0gJ135 zrSf#-D!WDuJojxwi4eI({PKzK-|^Jinmz$Wm%DUWil5TlOgG;p-k#292(=5Ah7AU5 zb(XGX4+Y=6L?i@BS5S6W2h;`{ZoaUwF@c5RfM?m=nxzqd5ijhy6+0{HCUm38gbZRakt0Z<;7 z#rD2$QRn>rO~)xX?0mldA8u=%3Y72!rva}^+MA#O;*5nq?+7Kvi#=MpSopiNMQjl2 zHC=SMEj>@y%#!1>U=H-Q%cM5I1!x>A2^lvA!6XbQ_55v!fJ)2UDnAKL@H3=G67U!(-AQZQknr(HK@5?l(rj*^Z9=Kp>dPFV8a(|l6whdvlycz~SD)M0*Io+^CEeBPe6Ke5 zf~7tl<4W=#%FsASf?Yek$B-e_-6ah)*KVl4j;nhr<~{)Wfc--OLa%$`%zfidZ~eZ) zh^rNd1e{_-X$e%It#W@8QWFZd-1U6rq*Lqzo?9DL_uZ9xP#*>L48?8CdfFYW0%*Pd~AHT74UfyawIE~nFZhK zPZ`>MXrBxQq}&Qc=wDdvI?;Ll`0%g5&J%t}{7m{)-}I-*DvR+UmZI@CJ!BV`GVU@t zR|)z4P;Mjln?K>y0$JHb@$R0D0J^pDgGUh%PY=8PpexG&Yq!xNnwb%8&HaH_eh#LerT_rN=AJUJrrddt4G$%yytr^f9RDSg)B z>Zo$i{W-r)fk;soZUK1*I~SFJnQEiG!9O!!Oy%X8>6ZVa?xYn4D*r6xoLK1J9ke>A zO^b@5>Y7NL2BA{J%8!G7n;?bE%EK!SAH>&k_h?8vvmd{4qj?E>2HZH)#^08v&9R!U z?dG)(KKORa&Y%c(bGx@>Xh{?FV%PXD>j({Bk@8Q?6ne)(ouF~*v4KBM9Eu`dRpy$@ zZZ3Q_z2>C1?#nCG8*MPXnf0&c@@VVZT(sMt2U}$V((2EVIOP3F7vqM~_v;G)&AIkU zgHx$~Ci7ud-fm~s5`v8TlOW`}ZQVln)&M;1{A7ppXTj-ZiwN)^5OXZs*2DOAs@YkX zA|g?a#+!C6CtEKUt%RvV@YP_z_4$IIo0C=BPB^E!zf9RmV2p;GStFkYgws-WfFb(g zh(Y0MeoQ%72K(64#&K7fC+ud|D|1+GlWs0w%l7ZGYr}Z=8BIr~r~KoAjS)x3J&*zL zIpCyzd^pF0@{DS3+F1yvaTxeuo;Z!gQ)(GDHJx1?*&*}1jCGGCSTgc37Q(wcJmuCw zu-3Z}Pn-0}!;77!gZ-U2xoKRtJGT~$nk81IJg0zgfH^0IqTTi+u7e{n>MumB-QnWQ zXF)?Rsv+mlSHq|=beR%)+v=j(STVmDlgpuVQ>me>zK|Z9X6^_w2epnz$!*e4e)B8R zKitaLLgp_#*ZSsNjbL7D;pdpmD5VMKD_oKpoP930cJbos4`oKna?w?P>aDIFX}9n=LIERHzE z?jgl+vTwmtyW_Jq?E5hhy9NJX1NXCdnCjp8`Y(4bZ*fnc$KYwM&a>hNr~f{C(|C{m z&vkS8wg}7;=gj_8RQDoY7&9WAuKnTPsOi^3@$ug4AL0KiNZlO(isvRv&`zhByi5g4 zaCQ2&pHdyxOaw(?O)*gP4a%oOSA=MZ*Xd%CuyTHA7B9?VCrr8o>W8~rqQ)QTCBB3PbNR^>kU&g+6;j6#8Lwj{VCxbkOO^`t9+pAI=fQ3871MZr`;>lH;f{ z6pRk|2)FvrR9@uzbiSIjF!(EYHdfagz8orRC~jMT?=}44zc>&%HdqAVPqM!ztFhosiFHg z_wtR?T%FwlNJgF@S|^Gaw302@mJ!P8jyNKq;?&6v?ivk~s%GAo-@ItJm&F$#3ZA9* z<$LVgk4#vf+4{n<^IewsT)80Mz&2rsu6|==j@lQ##kC2By~;!2PQ>_gYB&bi93M)o5zN&A-n>7~!Rc^HrLEz{Y{+lobq z$3((A=jYEGTp_`HfkQhS0H>v7gQI}ZYcy(QONa9p?=KrXRehJn!fu$H+ggfJ` z%55=f5)Ti<&SBf2=O+4qlE4#QW0mhG77QiQ)!$0&q?VcVJp-jerRg@{m?yIb{~5W-b6|t)4Q9tHf!CV#DPv}%j`eK=)IGAt3aH0VhrjK zWR*_@vANy7du63hICLgO_551AuD+o9N(;NdV*Nr%6@Rk1rhneXKi-oY@PHyywk(wi z4`a-A<8Y414({pfm%5E^Bg(M!T4`z6Iu0m~FtIk(>2%-yTc(a0T4NM7n8I%Z?F4`F z5HYI4AEXBq64y3M#t?>fmYWY$4vM&hOv(vNS9`G500-m;{KXy3u+KG!?%s;oemo;` zPIK~H8ggskXpUk9hJSw2T04So0(IH%tu}(J)k29tk8es=W0uPUv`^|V^gqp_%WGh^)OnFnszjMToFqtm3WhW5H?UF1u$tTkZHyG4vxK_#k zao1PSF>$lAK<+@vh=B8^4a{JRAM5N+^Y|0An}z1Bt$faW+&;zYhrWa_e=;=R<>)_s zMXiRTEPYZZ>+W7p?hi$+*Tcff4xy>l^dz^9(*d{$?&Y0ip3V<>+VB1io7WCE0oERe zY>obJRtO{t&4I#!9%wcTkPY^kQ5$>%{Lfgopbro3EW7S;i|%gf!`l#F+{z)1t?3@{D3%K8hVSd;&`Zuk8A0ha{0MuT*iXAC=}>S{IgHUD;xvD?hLyK%yn9P9u*IhX5qtw|=?=}6|qwS*(i-Q)3C z0lsWMKQ(n7bgyulq3JUTN|rGqrZ%(3_y~G=^SgJcwZbx463<${rpb76b}K@|Lx%WI z^J#=M3Aak4<7D!wM!S5~KxfKhL4z+IvUfXbHsE||Sp{NHoKuU8$KV}bxlwOQuJ1jW zjFdk3V@^3rRI;}3M{SG8Ab2O-bLEWmxlWUtK7u=kzRQcFiGp+BdX)SqDh%Bx9GlX0 znpc;OU_3ToYDH=8JTZ>sy*ZE;9Jbmc!^MSu6Jk>urn68~afK;C10Uh6r!V*|NI0#2 zMLLI#^}bjQvWNst6U(^S-c`}XEf$_xUK@@7hxncALBnG8eC8=3OYyGX#Ejjo{pjBlD3+m8OrXVc-K8z5A!QtBVEen8)4IiJ$-}Jzg5H)eIbb7DZE*$Pox>*5!JJ`-r2J4Obo>)S4;e2AH;mk0(XECu z##z@+CK^Xl^;;5@>3g3!ch6MEq!4%dNuIhLg3g8b{}@KO_mOUPSyR70(V_E+;3hg3 zNa%~?qVR!Gg>ZodPd;G7zYK8{AizoqHArC~;lH18aX$u?i5gJP0nGps7E^EV$PdW? zn7~uet0#@96R^(7h@C}VE>IpYGWL(!$C#@eX^-u<(xy#pH>i3SJ}7M(+K;5ZPDFBw zli-`-Zl+>!d&65o&Nw@nC%?^c1C`Ejzy(>{f`;QTP4F}e z!}sdG-IBU2`4%j?E4KH81SV*Nx9uhcA7DSrbJSPjVrOh58qPp34Iox+jfB8sxrrS3 zzBArR2FLTd7WV#_M}FwPgz)P)V|7!LVR`{Y>bqlE30*I`Q}qGpAAZLa2MJeeY&JcK z3kPT(Z3#iGuif^F6{u)(uTq| z><^Y~GsVI8%mBRjn>Ve*km@;*@C!EOZz;JK&bGyE%o9KEKI`d`9>)cu?6c!y&$vX_ zbiw+bK3uUi>wYx6MW=gCa~N!FEaFdu)o`@*x@BVlgCh>1#k;3@eEhua>?*vEdBYOx z;OD`LoN3%azxzZn%b%MUXkA&%Z03NCL%@JOR6T9PDzp7y_$VLpAq`h}` zaGSnEEhRR;UnqKewVc`y{=Ru@PdL=4N_;ivh?U>dO`)~75SHK;frYjbaZxZI{L>Yd zD9a}H>e_BlX$ey$a_o<$6!UF3Ygit7J=}Ll+;?X>0aD^=V^cEQtzD~msLJZsFz$Vf zL^z+qT2QTHge(=la1>Rh;oKe59~XwtemXct{YGeN%b&HFR`_=9fL=EwANs;yMYD(t zVcK%nV3PjZ&lr87cBAfJChuZ;f16)1z2fixt>>^eAl*;URnQ}#`J@=|3I^g#uNXpp z3=G^ZdSWKD15$iMmJv{0pB6t5=;H1sLIg)Jb@sYHY zU5qA1^5JbgEtp=~D-&zL+oAb{-IF%-hI&s46(w%chh<-HI1=8!@ake7{P=uM zDOq4ES*k-*D6QUuQgp8E*;TyL`hL3yC)v7s7L~JZznxjUo25GPPlX2P^62KY7Lx z55yFy0h8fdz1{XiwBbhH##=Fc2hp_@A4B?6=-K?OF^^K-MCLnj0`3fy)j@#!%8B&U zLkof%3r-^USCR-U+1My_UdB*i&@Ab47AANobFy({gf)BLLWK(i^)t`zTUjX?$TrYx1wDO~=k# z+BL|>4{9zqdNn6ej`$7H%0@7krYh9^>>2j`iRO`?oR+65b|@uT(Zyy~YrlP` zJ8S%@>EzP(<$U$)BajYjQ$WCivEvo=oxT65d~5<^14nWb(%zeByG^F$mLH2?Bso5l zNa#13@6o>;?}WX*^%sJTV+l&HAC><$)jnpC^oZxZiKu}-O!~~mdR*)?XdHbVs6&EE zx}Ja{U7-?2+en*A3#OOd6Q%%6opSS+&j%JFB$_>g|dv?Fy(zW=eP z(KDIsIa~PzQ(%)oDXQJDwg}jwy5NNgh(y=$A#Ne9L7Kc_?tJK-Ic6FsO@S}m`Pt|f{ ztUK%WkTS;A#5}x%BNo9hgv^5k(z3?>;UTG|HqNWj*ll->L9#9uygeU5kC-8UNmT|A zB?+shSAT!-&D&)3!}x)j*ipZ8vlaZbsNkOQEXTjBskGFG=8Jm3gTap-F zC}&pZ9LsM=+iG#!Fl!o4)SIM89ZePdt$lCvShP@bk>N&q_B&D)x9rxs_XDb;k=I%m z+pKZ*Q$`jSEItB$v*qp{YvUJWr}MXHi!`T&dU$Bc_MoWICto?k3w1p3nWUG$ zeUr+{y|Zoi54MLQOO%=`cfxc~Dn-l6G5&bekaVadHt{J2@ z=Z7B8{gv9Q!)Y|Z99gPI4sYaY9s zf*zy7d~Yas6_f)OqMA_wJ0h|}j8__q%C7~l>l_y0OPXfr;-St#*4^)uE0CB4eM zoe<>zV+y=*7UKHo~ZASS3Ch*KLRILPi_BUq>%a9IGb@iEd5&H9A8OGuJ=70li*BG=npVq zB9MPIqsmS!y$+k{Ldls`?j)iJqcP3#CHwVEQ9PVdPAc*%?b)k(YR=8a-+xjPyd z=N^+U=6rAUJR=5YZ#2dh&c&$`gfqYL>O zSen0I#?0qMfqOY0kM)tD4$u+T7@SgeJ-S!V2Y$UmW|P9xmJy+KI~8eNt*Isur?QoV?dZ$lzh;shM5EB!$lMd2e;063mA0OL3}~J}EL2o9#iGBk)0-!9G(jhz42Ha9O$mn&M6Yqa|N|6k85RR%{|=XBXL;4i@e zKm#(trB6Z}M8>E?37HpcZs;eN9nF5+8-Ge95sSK+up*M4o8ICsD>8i;;l&yQt%LC5 z>HKYwp5MT8NxvnZ{5BRNE4~)L7yQ{!;Q867{d)R>%fm|7mT8I)Y)*I1 zS|t1@aDeR(-u;^cKtDSM1s45AgLrNoB^OwKGpNuTWtD(Csrq0l+{2GR?%Z z+QIqwDlbxO1i()gTsV}`*Zcnn`~QCe61~GFksLl=;Nz80F8xo z-6=fB=49kS?;}Z|yUbEbk8tHRYFsw5D7Y<8`|&9^H4q}DOE2|$kr-N%F$4jF#KUVK zAoTWY(pITW4JC`18#4cXW7NU<>I>N&yf!(UBX<~H4RYlaFMqYY6qTd&S*pWYboMU* z{)`wzpWw#HyGUAVuYWQ)qcp1mvl)}zcPh<`DU=e|evn~((oHRUhi)*>@tw+RFxIey z9Ov69^Se6UUsZN**YlTCOw;;Ik>l2xx$#o2H<60h+&0iUi~X5RcDrTHL9WLO~OnrGj9 zg-Rn~WVx7myGtur;N7G{IPnb$nbHIKvtqcb-XMP9I({e8mgAMZUE&YSl~n`n%?-C7 zAzSbK&0dbwFw046zWoZm@8obaWu(YCIh4jdIh|U-+g8gdufFK{c?kW80N#t^JLf!MY1ddF#4w{ z48<*3xK#yJ*)Dc(m&YL3p1V(pVZva#Aa>4+;yh<#czQhTfKDnv2^1gTLL1*TWW@r# z0a3B6G|3z(>HSQFO1K`aSt#@yK5UiB?Co!x0h0pE&Kg=TT)yZ7rlL}j? zykhok(!H+*u-_f$ZrCToXq*@SlEL>#%wFw&3}n~o`x{p30{xdo6hI4Y1>mGrZG9q2 zJ?W5)wijD05}KKkhY`9BTln8m(yaXtNIOfidi0GEM|r4)cD5VMA%n*Zd(Tr7)0KD& zCLBG zHwg4+j&Xvr=a1EaaXg+JQzid^=`oIibSOt~SmG5nwDFjXdqSa`GHaL%CI2uzFaDTG z5L}62hOvRjz4Xp0SMsijJd-`;~=HX~&zF$$= zWS)A%q|sH4AOKG`V%Nflc9JqIph@-N-K_8A(n-^W>$^Yg3(!}a(Qp^|>iZehG`YxJ zrzYzpmxChzh1Q6vn!J`S?pN8M-!;Jj)e=L`Ch{X! z!qm;5mA9FTxTkii9_W`EFu_FXm$}$PX~F4kwyH) z{?7cOhi~*ofKTF$oh^`w_PMYQ@cO72KnKlG5LP zHB1_i*6>bqHGJ0+7xq)*y+%KjyK%;Lwim<{KH%|r629|H8KoxB;Da7n9y@dv9sT5U z5TJ5+vgH~3{lAV5EhB7?5q0(23E=qdLEy=W8bdYRSR-Kiz1T`5 zgUo08d-d8S$#ZGPwz0cwLPA${QBUB)Vq?LP5H>|2BukPGwa1<7zQO^GIVE%w#TTm; zecD%eHC<2-@XQd){EI2Q5Sx&BXj|HEsCT&A(9VgpqHc3^R8cXg6gp8_OG|{g*}I8= z=Y8XV*N^4zVl-_|mxPMK^is2M>2GqTi@VofJ$!fZm)&yeA<$EfzQ_`2eT*RSu#(vK58yE zR`{Ieu(J2R#pzuUL!EDkF$af}!Qf%LzhL7s=^*HQ#XkCZKnYlkQ?L9?N3{0ICrj;5 z$}r2rnQTgnWP2Y4$fhgpU9hTSrxQN*<31G~qfKQPgi2MfHC-H>CR7qUX(23s^Jevi zITJ$NT*O$5Wgak^B$Cwb)}c1c$dlu+Ey?!V>iDqyObHXNdBfAHBeS`mmhfou6ZLgXUq*+($Q?c%7D9T_wX$zaz%e4G zQQ&4Hs9h;wqgZKgVTkm$xJkF)ww;p63Y(767~;lvx!O7|3CYlt6I$du8{a((4^Na& zXu8^On*iuKbVu3Oaasm-U5OBZr;c>|OTw)1<8z1U%BOXgF8;5~StJ`&&mfs4{!}^k zg9jGX)}AC!vX+!#v*3r6;{~Gt>w^;d0x%Hg$5XEfUa=UI-cQ(}eDTvL76t>a#~SZJ z0>^y~@AM|>7HCu&ck^zhregynH6J37aCuG+6bF*{&U?Q*Kz|C>1FlB$0hn`2tB`Y$ zr}+wPIzMH=$NvOxe6x~&Ayn>5mCTaxb49_t4Iaj?UmBcZF}fg`Ont3y`MP?#Qb1wb z%x%M#Fc`g~kylY`ZSI9&U*Mc5g(KSGLbnE=B|4@|<^d)by8UCz#6*%r1;t<^H)^Q*^^!}l*;pp1%pK~%QJ3VA#T8)<3}wFyf3@}ZOr`7v1+ z&5|zjt@WI>+B}uIIZ`OMi(h;sNkoVKP_8YrzBNUAukETmQ6%SA%ZmB#)yToIL47RnL+)2Zwi(I@#gMVfETcVlmLMK}={39{Zz8@G$Wh(pFCRfhuSrK`g*}@1VuuB~Qol`#R9q|H*pVl*eP;?i-R)ydd9%?lqxo+%T$; zPo>cmv~7N*@nC;vAuXai+{S-Gj34nN2>g0joh+eCHMj}WZPMWs5{6~#zP$Jau>o4} zTp9A0BrT7vm|G-VZ;@7{JPunApXp37OkLyjeFBp5xu3L)`U$(k8HF|V`eESH8>IWt z!Kd|@l$OMJTA?imZ)P~D@J-}M9=Jxxyqu^d?*F%^k%MhAXYleCRiew0`~WsEq`#@o zG;Evx0-eV!saLz)S?DdHw+vT=4QT|7>wwy zj`&)m#Ot6U=&V;HXIou_-7?_OuB6@4^e5M3-6o??I2KhM+7ivLDSHsSDTx8tFqMva zgUH=-Qrim`vk9A^?3?F0-Ipcmg&LZ^PUse0x{KG$6H`3l8WQ{y zjh6L$V74MP*!EQ6P0IdVIn#{lW|>MUb1ua4vj){1gKP_0h&^H8%|H&XHC(1 zaCkxOJ9i}{erFI8UONQgfD99jZYY{lUcvT@Z!L5HvozFPR*^)lwhJ=nd5#vzRxcWOp6$F>5+K>xt9$VEUw9uk>@O;FvyBY= zx6Da4K#6r}uGNc}*A={Mi0JG*;V=&Fst$3MThj60#YU1YhCUN(Cn zedvvas`rGQek=*oXE^>@zM1K4Cq8r8Y{TSMb|c`1&~q{tBR?8Lze;&$cGd-+TPVwS zCdJ;@S{1d@BLW9U%mP zJni8KwQl>dJL-oC2YX^{apztqBa%3fn2Xd)9AkBE2jDjfXr85|Y(6$#JrD&YyycwA z>seLhon1q{(WSh(Iy$GSy5+k*v{~Z4JXUx)eAtb3%dr_}b;d3xa=t~0 zuA#<*p&O-}NV@>(8=w|%~@K)kWPhWaSWZHtUJIqov))|Ne4b12W!jQj7GMo>N? zvSdcf!hV{g%=~2XJBS+(uHhm&dyTG$XUp~t2|v4)lXYnlXX<iRjN?{xS8io8NI^59eK-d80(A@(p-&@Bis5jGSH^ z?n_*)Y$*NjC&!X>fem$P*b7@EiRifG7+?rMHG1^Jt&fRkC6l2rgCnq#v|>~$;*Tg> z29JsHTiIsNyRq13r*OF6k5ckLRltW{sT06E!zXR336%+znQ zL(nBM*7~758#WZ_VVqUYk9EZQASA>qpI==-JO@ByYnd4~ALRj&lmw448w0?xh08^n zz+~F(_zx^kdd*~nlz8tE18S1t$VRJtg5a%n*Q+4m{IJ!Y-kXxVjBla6MbC@07lCFh zXVvOtK5mwim6(v1VjlK13c0b}lQE}79qCa$NC=r78=IQ{YR6Hu-v)WE&j4(X!d59treryuJ+xd$L>_bA^Ht^;dFenvP47W3V=v}ShhV%*25tFWp%d?da^(Dar|$A zQf{Jb{;G*Cxk-(;kJdG9@rLr_KeyXo9&?CWqhA(5;|R2y_7R|?$#K_P@|7=FN}{C( z4@o0$0uvrzEQ2hFbh+e!SekkhnrJ37^A+NW9j8HVgI^6RBRl*@y-5v&8LZ8dg2B1x z!EG}qT-RD@^$R7kMbiU1ewbX=Gi`T!K1#8wATKQ09@@M6OFr^k4y7z2+4#7`P68w% zXnK0jqXc}{`9vTRxx3Q7;b6)3x-CI;Q(3LprlO!ki0hGt@p3lH)p$z=IBBk!|dQcUNg6hrB#x|+Y+j{NROvUy>%{~-e6v`W% zonTJCYoU;290u^)=0#l@8r+l5hkNWsI&hI_7uv({z1N}NKQ7;4@$+@QTP_;#71`K5 zUdZm|5`8R8QGD?WlBCuy+aW4losu^|Hkik1)5I9wePxrOk&>R@V%`a+i+z-@9;i~2 zxweM4f2{^)9&XF*_rJ2<{677o5uiqEXoL8==OrBr9jf=$O#==)WN2Y8 zC&E?!ySlCh00sETVT~m<)AXge&?0*0U4X60$&z6bTm~oV@j%05m@haTYk-k#Q$$@r z-?SJ}PVnHCgnzMMuS0+jH`9us=kvto@4rJfpWds@4=pit(oN-40#%5A z?yfe&JzH-V*91#%fnB~!=mwDl>lZ%lUrH>LPwB^Osq-yb1?Ad;P` zipSn^#~*zhduoUMB(G^1pKM7E*P46(s3p@l*To>zPl7u+#(G|?*puPNf6n65j2N!6 z9M*9@j4Gk=x1Xo-3HZ2t)niVV_M4e1MWS13;lGn-V$Bl~9dq>xyXm3Z zB|m;vF`B~m%iY3-;W^w?vg+1Dg7%0dEyklRe`vd;BqUaWsh;E<#60QfdzSRWyt<&u zja0{b(!uLFH;xJt-IzyD)Cr7Cqx^{C$&-P@IV<*Ktp8UR>|3rN@%WHpBNi>X> zQ*@)ly59{jpZvMjshxv4`ldfm=Q>zm6SmxUsel6#RyO}0GLDh5)d9m&Obg!V9JwJM zUcp(~pH@L*VTy)R{#8QAOFo>t2~DX?{&()Sp(}x!v}-hsh1GF_$^l{|_()nD&;u=? z&+A*r6YJK!hp{b-DMYy!#D+<-)h1*aNyDyJ8B0GXtzFj5hs|!M5usPsvb9kjSIE(J zkV%8Y`E+aDs8nYuQ;<6Sb12X}Vm=e!nI6#@TZlL`l?sDj>OT2+h)(weuXc|9ZA!Zv zp}hl{S-7)`Pc%2ijraL5Hbbd0Of#U&YAeff%)%^&pKJzPHw-Jum0x8(@EyriPyQ9>hHc+e$TV! zM}Te0K*-nTm_p7_^pF0PE)GSAB~>L-eBh78D}}Gk5l+WtqMcTxE9)cwzN?vP2E!6B zd%wV>SK|uVxpp`;x%UORz#Y4-u~7z2@G-&C{i!*H+Y#U>aIF+SUCrNank;kB)xNwWX1*R|*8Qn4ToEwC8#$iTJ)$ z(4a)90UifgYK2NIuKoxP9kIiEBO^zjYG>ld%vtcxVT=CIU9JapD&jM>-b`j*bTfMi zO}8jfhjEOLW16Ducf#Av?sHINWe!iW1?mA%5{4tcJp#`u6Wu`V8& zh4L3Mfo)%`i)0hp+?p+EV=NGX^((0kK{<^S%rvtd*~`b-An1Yi1doxDVhxvM%;!XZ z-?O)iZcP^R1g**L|L)UTk?n9JCC|xjvBv)eI|mQXe)4b+zW+C_mpLY}LNFOSP&opS zJ<9fKZ?#X}>h5Vau~%4Y84Y^Xh=ZZgJo{0h&(nRpEkMhk=XpOrmM~O--3d2Vfys`p z+3R#OKu>_?DxlZ%V67X94;H0OJ?l&I*}?sp8w(?uKL5!^zYhNMex^Ua2!tf^(Z+Uw zoZ?^YmR6U3FLq{2J^l)ElCM6k+-T_Ab-0zog zSgf1Nmul%gfyR5`Ife(vz>qN;85X}WSPWCZd!L-A_7xd@j0lqq14yvND9?Rz>tfA>j0(TCm8=fMSD z05q{h#jYcR(KaNAQ=k3na6FUmn%b)|kp#r5SbUOS%R>^uMXw>`0Yhphk7=Oxnr(qb>LK~K-kOFYC2nt-735_{q1{IOf2!JdU7C)#+` zlTeT3Mt;o1>~O;XifQ(-Ff^}kimsvcPV!Au8*CN)VeHpu_ij`2`v9OhLA&4-6!-RO zWSCHTho?jwcrHx0v3C~hUAN?$+Wn}*8oZXs5X@nLO~Q=Ug@^0|f81(2nR_>(Ul$$A zkl5J%{1;RetP<;P^uP1q9wBqUY#ZYm1!4e7uTNU!>Kgl84kxQy#7khaWKLsUV8k4;@*YMyNydC6~4UzUz;nc z?ui)+N|{(UpdKww7wK>~8;0($WGp|M6&6n@Lu>uSG9g_Dw|D34(3b`*cE&zTzSIq_tkYS+yltQ$M!SDtV(Mf)dYC9WxCtR+{B$fl!TvNdmGsi0^)+a69`J4!ipot}1ElyvBLx~bT{aXIKr z1BKAk9s&{wTd^_YO4G|9^Wf-(QdaB+ZU-(p`4!9vIY&|63{<_#1x%7 z;sRagb3kX`IMSJ7a}~zl%AwM^VioO3Ns3kRohrHb?PD?|Ec=4Ahy4G?*n0*w*>+vS zf>DAZBy^;NBvk1jO-d-CC<-DfAU&a|bddlOdW3`)0xI2zdWi*)CW0U}G-;xAkrJ9h zC;@>G+9&Va&-=|h@Ac#Um6`KTW|DpGW37GcwN{XDUC;4kx1s;CWILKfW%2s4F|-%8 ztM#?!_g-qK@)XqhjFy}1uGIRI@)qOehg|SuX%C@X;0zy_Iq4YyoXvs9yu470zk<_( z@yjLwP0}CiL0R)hVg@p;p}Ros2aM?i0aEP`zgj+_vs7N$i|*Ecu3Q$EfJKt(%f)#m zq{@e4M*i$S4PTLUYYL|%x5vD^0Tk}AiM;dT;Alem@=ag4d|g9ILx3Ujh-Fc5BfM6L zBS=5uf~lxcqM%_c4}g-rBgIQ#*=)tDk#f>v#|6A;(6Q00ws|}Q6dZ#`4GY{+JV~N+ z3VlQyk{T&b;~?1Y{uIKI4e*?_p%ntDOzou^{HiEa+-pX1&XcrrK;b{7Y090=!P2B=7}ya}~5z6O78s zJvhc*OnE8x#Ff#H-$Y(Ks9ven%Q5dnY*CHZN*C_thF-HR5{&zt*l0X8pb>FrDTsQ* zA>yF$`9#z?TJML_1)%9+@#j#S@2T@WudywAsv#KUR1Dd0-%mnspi&%+ zBY8&hg>cscIm0)WD(oytO|mw_yRy?O#LFYIFD7|np-p29O8#e$@2{?G+UN6F>6W2# z@hcy#eZZZY|3sCoe6-42pB$sYYv-@*b;fV%{I(w2tE_VWy`EC-wRi9};R?OTh0S58 zXVBKgDeKJw7wex^0Sx1WSx|;*o5t|JLWKWk#fCpd33j)kX*)UkVn6M`hfdH7 zPEK0wJrYJ3-O3-0xvsxCVDph(YV7{ss_o9q;EM8mPgkZm)6?gIuo<%#^6CL8a4R(`#&TUV_YPjv4^H^Cq&vHHX2IOY z*}bR(B8g|VmX9-I3-l?}e=?nK*}ZwERK-_yl&u-mvhwX83%TzCCHC*uq(SS0*rTYd z8!pL-XA-o^Ptufn4;ZFf3?O+ zc%&?Hly?JQ9C&{wMRTay-GJHiWN>NQKG`vHXO7RMEt3^}KtH~cT`|7;Zzjt+!ZG}S z5RGQvhX4O&%)F~c>s=2cbUxeb9J&?V?>^U&PEwgwaqqY`3%jTB61L7=qs0#ZZ{Y0! zGWWO%iRaB8aN9Z>Ym^Jb9Q&Pm)#M3`NvZ?(r0tbJ>mwGFU}7q%FSmc0RApRjIk}$U z0$?4zN;yqs;eMom{u;6f4FSK!-4^WqS&=2 zzIQ{XU}QnKW$UC}2UrcdxA}OL=|zkhO7zhrp7;}LrE>7qm}^jn;{HH#CHY2e@@RGI zx!t4dMuHeVauhX`a!`ao30Yqtb*fc z0KYAuK>BGYs(;8(c@{DG^P^Wz)Yd338Yd8Vv8c@DOQnWV{Zc{%HBIl^CSINXFpJ_v zygE4a)Ng_ovbfZ)Vh>{^_m$;;s>xgp(r$68D{Zo!(GUwt3*WlMSXYn8-7bP`;@hc` z4JtIyZqFOU)xb%n02fV`+OHnC2y&R;&O)cYnUoo6PG+>GGGOW1kh0alvB)UcEO>y% z4!E%0L%1+RGMi=ojEk5zI$2E53WwUpHy%dJ?w3IlT1rL5XPSR7?`^ zglcGV43{|yBWfOk{C#wXDFiK7PlhI_Y;e`QEvfPmw*gnEUcAO=f{eoT;Y-`2UgQTf z;c%be>?i=`21_b9v6Dmk+WP`nwp+`y_@Ypb z5fOL%&j@8cFpsPn6v!RQKFCiQ4aGkj`f(&6fgNv}`%Bp5$fE>tFq!>s=^sx0{ls0` znXOZ;W4Hoi2<{B4)zGU8wB{ri)sNUA&6t+Gj_XoIzpTPBy3XKw>_;|jyr5XaMVfnp zmT`x2<|0^bSuX%ZG>6ZOh%G3aZh7?3I}NjD*jR=fg$v8F&0;g`#E;dO2At;tn~cWQ zr0;(Xi7*kVEbRE^-c^sbnh|c8YJvz>_f@T;&X$xWJfyFgiR1+|v;QcK*c$bc+r8nG z#f_mgQ3tSio7hr8agJuZ7N4OSoz1~cMth@Yfj&KEtAKq-GUnLbSGL?>507XZuHIRy zt?qI@qqDg`0XkOfyRndyz7zK%aI}K`?rQz6U$KCSo7g6cz(`_29{bNN_NrxeF`+Y7 zOOuc;20YrGuraEqGE2P6F#3h=a_kh}60|ge7ubvLrX!xID3$Z4kVe{UXU-Fu8bk-S zX+hJ-q}GH3N<$<=ex4>=8i==!c-QqbFbh4ruX0*MH~xpSo9-5ClzbCcFy-cu=e$sf zPX(L9!ha=Wa|Iim?EN6V1#10Um7L}bGe|@w@HUP?G840>3ahHk-e+DULr;)C;8)#k za3aH;P3F}G$KW!g@eWhAGC+SP1Rzc2Qy~B_{+^&;dEm@N@qfU%<~N7IbeeENUoR8s zyl570g14;#;whjlzK9LJro`moiLA6rsNRQ#h9Y*%=t%hLX93ZuJ^GE|5z_mko(w!a zuZ)ThTwqyq)>wzM2#R*pe#f^DYmCeOzo z&UZ1Ta_GiTT2tGe4EB-RiI5~irDl5RD79-lixaqNWMIQ2z=qRb+61rk;a9MfyjG$d3pObc7Hy~AVgY3>7o_Z z={epdZ?vM>a@#hh;<#@5tUm-reY+_A7<7AO$NiuuWzwjuQz=2OEidbPcTHd}n|&wh z3zN?((u=Th5YRY+6*x{kp@F(tm+cZR5|zL2$=ttZ<>O$p7oX2aH7Ih%g8Ha~+uiD9 z#^`@LulPMr6^DVE@B=~OE#ML{bb_M^tSi*$9weZV#Y6(G1CF`GOiojgVjIOkpl@%- zRo-NyzjIl3G_HCTntZlDM%iE%{#x9VlnJ!Lq_uR|0RD;O9UirK!e(R(S%F7nJdATS zefmY&ot(mm2Wfcji-)A-`ZMDa%6I^z!Hn2vfhp}I`k$~*K=CTl5LltRSn$5+&WcZD zkIudE@kE87T)S^Kv4j(4zbdp$9(B!wUjW#b9{!5GE^P-ErJSIOm8Bww+g@n%EyZ+D z5D%|JY!Hap!H<{{f@#%-l4v*#O{&3`Sr27upsthW}?B@Q*Lx`Jnj<% zU*4lC1?I~95+jeGGT!4rtbnq%F^|k+cda_kQs1%K?&|<~0wm>2JucJQs+i1J!Y47} z7ypCXY!3RJrGG}T9$e9+9Z~_i_pk>zM{X-`nQbqck3{;q)#Y+fzQq%s81iDh2M)8L z;8nN}85Z3tQq5#NN=fScZw!;7S@VL-v5f9Zd_te=Psn8{EPmaZ+Ips?=NCz%~In~j&ZoFwmb#K)!e2pB!1e{mh)s@<|i zul=-l*x%(qqpwq{S~eZKHs7x8AXbFlhX%}(Cq-DR)RMVFaPnwqG`&Tng_4&9hKWN4 zKO%IJ)em;!IH@_^)(xjo<`CtFW>{sT@t?XW0ssCklHJ@oNN4EcPx1>CAKS&nHJPE&!h>pXZM#m9u1!w1b^= zHvi40Tlvb7{!MD^r|o~Jq2?(ToQx4j1^>&$8J42g#tBVfnC|g>t<&Km;PFVnh51{c zWSwhx)sAnrE)gHhFCq0~zrPsPUF3NqUjgb|2K((vcCV?ttPAYi6F(&=e2-rxNuIOp z4P-NIvFSwrspGLH${9xyt!nq|=Px0F4Csyjv#V>92kq_AYrdcY>ch*`Xb5%9i;gNx z7*Q$=IY4OS@#iP#O&{Kse0od8i&D>)b&u+eVmcyJvW+PBtZ3`jh#|86pLgPugW%zk zBn9t2d+WZ4gFAA<1cYiTo}21&l9!Hr!dqrm$Abml6ne2;K(Md+Y}5T|T~y;Qgs!>s z{S~(+KG-MhrF2Zk1Gv^&Od1H1p>g{we^f&+X6P6y8`YZD90)GiH&iEf17tdSJTT^1ZiZ~SpVO+n;jWE z(4hR_QgcHb{F=2?TKf{TlyDm8ksXM?32+5A*TlwjkCS5Ske0A3q(Y#(gu)~+2zd&r zGztD~?wl=be&acL33_ADI0t@bC$8G;6RczNIM;T>lg}}ECR5xBpi<3AY*0R_T~*0l zpeZ#%A}n*qM0{31U6)K6P2~+Ls`gwFfp+wuspnbaL&f?1M@kEZfY9S_6sw8|6n-~c z8ps-F5dB)3D*jazsS~mM_eU^!?cmT6x=E?7|LK+C0;RV4qPW2j7M#nmh)vH zFTyz9sOQyu>$PL_`Fp$QpnLUyi^k*l^p`fPjMo0g3m~rJh1m2B`L#}nwajQ7%?g{4 zOJC2{3+k+oFw+UoKA3i2<&##GTJEGs?>aMM2w@ufDs9sC_mMI)AFR&JKRM=p_(ga0 z`zW^>l|l-OJ>es|M_}BJs+>Hx0)E-eUY5o+6sw{=ZlyPwLVC{}5Wp5g7<1qrEPwla zw6EY!UUNLka#TEZQNR!B6WNj*VE*z7LHVxVKp$gVI#g<_g6p=oUj&&~a%VZG$hbjt z3xI-%^y%-?nn%!Kn`q0F&T2w!+ZMeYVq4b>L>Z;@fl zZ!#qB)05lcW%nBH%pQ9*TpZLC6G|QGU7zccyP2%VxJzy(cay!}@L=<7C>f#i=>Ntb zm0=16T?3ab7*xTmKg)3#5?*p;f4V#P7WRSEVBVsl&G0Z}k2&;MDs(o#R1cONDt54) zRZMq^@96HnOe!$HFJyoo#XxHTx=cvvgWq90{C>LwBv=oAFQ|9^N9G z=cmfzCw=htRMCEe%SS=Va~3%HdLGN<;Tbh{E~KOp8;eukQXi8PIA@)UlX}iu%Zyn5 zFnkFqZ3HR>f^i%(m8L*;@P#}Nfp>+UIb>HqC#^NPpx18OPze`sn0+4F6pNBR_p_-x zk$AT`J6`^Ax%4F$-JGbb&igzGsweYB_gp6@I1#%=5#L2hTcn#GU6>!XYh{tAJU0t} zyU?+>(GyaIe}%hZ^mLVaYG{7q!MBrAi%D|~SIa-LC2kbN@<@eek+ps``E&j=`S=fm z_L-MDwHM|U782YO^)6z5HzaGu|pSw`PI@CRnP6Pd>YW?F3Zhq zWr2>bKKX3uVB(;EJNNTjKWwCx-f)G|E2qUj|E4Ft1OE$eE5WM=|FeXv0>mXF4K9E( zIFSn9`0pgYM5>bSf4gY%fO`s4Dc%zDGWnu8rHn%mSA7|{s*)=Nyi3eMLu!_vnzjE^ zq6SG2!a?uOSix5E`b0Dgt26hv7()8ww6In zVp50TUak2W0FeXGl}g)8klEGh{jDDsmpQ&+B?E59&2H<(veHc*xnqGcOzHC=Z4tm` zah$T{Ywq0)?;HbT+EuSEapVIE+&VV&A}x*HZpAj;We1dMd+WPa#UIh|mwt5lM(|jP zEUt~7-5N0`LLaiQ;LOpEI0ek$uqHiy>UlBh{TJPe6H)aY)c@kHpmLZ#t> ze*XDh>h-9_-wt}jV9x}`Z+*A#Gal?6RNHZesInQA!8pB7J9hBXp)tGow#g`qbI&qg2M_9;I=cKWw|a zGgq?fGS#zoBmDklrSaC$uC2n$+W!$=`irod>(Bb1Dj!v61sGmE$zja_Rycm5K&HjZ ztSieWFX_yzE{s>G8-Y|ZdP?lLxebSjKPSD#hf17FhZv?dnm&OUh>pSC*tzjm5}McS z=DGV>_<<4b8Zi6Az3qC9EHFs#6>(YdG#d-Oz`Y)8>dfN|tFQJ{)?KTMTn?Ac7}C74B)*PvPx*HdJ;4Oyj_Vf0o3xfS+1 zLnOWEEK`s~B%OsH{J_=CvE;W^r(@`L;nL7#cWIkDjN9+eMO;}bp(M=Tsr=AU8pA{b z1i9bjZT3`YYxqr0cZKZ6aLA5cKTsB@rm29fbzydGA))-W?EE2HuiU6%+y|APQ{`q3oNEZatjuGDG(H%s-MQUEh*EL$8KXb8ld;@sf&}+){oq8G zE7`S|-tT)6u&)@JTv-!eK+jToE)iAimAw=|Vt*%y5ZP4d{o`+0EQDcLz2;?GAy_Ys z2+~;Q_3WCTh-Zbc=Xj%CyRz_aDFJ0uwSwkC?fKqgFWsLq-ZWCOME@(Lj_^YRytxGV zc-R|pQ*eg4tPg$c;OBgFW8qEwaUAe@|XBcbEB(+4g2n0PtlzCwjKVOO{YkI0Bald)Yu_fzpOj4s2pKLWu zFdFUa)!QEDj_H0^RsI9$O+iPk9h+1zKA97$x>W?nEQc^PIqiN}>~MH*GZt!>eC!2K zKlQILt|h!5J4^h;j4P;U*?Zj=m9RCR&?w#2U)MtRSC)Dp)P9uVHY5E@NR6zzV3Db) zKyFo285Fr?wFck%-E#H=v8w}}+2E6P9FsgcN@Ad_Xb*-<^qSFcr#%ilQg;0$tnNn?(#ih z(!dv(Y0onl$}ofYNE{51<>yavURuj8`aUWA|CaAa_EC_RTEFj){}8W37xCy><6AHW zDF%obH8r=Tz_@vSyPD{rKA^cU>Yj2PIF|kbTlX2OX1XZ-PY8!-tbl zE`Q?r{}l>?IXenA#bSbCn#|WIg2LI8N}%JnmE_{&V6^ zb|JrvOirxIIRb=(K$z?$udhrtFS&2bnun(@6zk;$OV{p(ZPh1yDEY9md#7Y+JiJAw z7CK7$5LYU_GNeTe`t#a+vUMSbO-yFG6Jk07HOG~qQ;sdE6b49!Hcvco8~!JdU_}wE zZqhNz7%wj=7u_0yX^2Jb(JM6Y4J{8|fykSko`ERMf^=mG{Q!P}U{_FbrSvnQ!$04x z-tX;A#c?g_tH|j{PCxIsW7lN=^!EEyZ3*Wgr+p9G9QDDh^~}%(`Kl5^ZQWFE9}|Th z>+s&3J2$SX)SUplg47L+%g{xiwsx41THT9@ub=9&WQ$1`{R@D-2_FH7YA``>A|p*^ zc+|7^i-KSdI>U`kZnD+6=ue9Vy0mLGX^WykWEp&4=Ru5nkCn*badhj9Ec;$qF+Np@ zkg}e1j3@TZmxng|#e_E|{WG*lH2?>6#i+=d!wE$2A1m>PxuUK244-DdXJQ&ni-4Kb zYRvLjqMq=k!YcMAgT*THpXx%ZgmC5G|F*1Na5RZ9CSZvFx!YKKjEwGx1|kJvJ82Q- zUesf-EScD}$`_rr-D%P2^fgz*{#Bz(=H|`hW=j~Wu%og21Kp zW?HTJRmyo)KK6I|_YgP@Fi9|0;-fe2ecNeMF&_NOP0=@(B=EnY&Es7%TTZi3szok| zNw&A&{8T6w^Y9rGFMd{hWsWd1CMq>aflOO|#p6Hx<(Xr8W`H!cwlwnkbq5p0_?IE% zXU5l&-%erJETzs&|H}<3fJzociekL#fmGSKTXj9-}Di=s%4Tk^SoFbX-quSTAD}WvPGJwF-AonjjoI%7R7iod@z0n;Y&5>* zQStWj7hRuNZALz|aiJje&7B4E`z++6dEtglHkYjGHmRiov+vxHxuPxPWqOdi2r9dd zbleAeAxx7TIs4q3c`v?UsclAD-VLBhzi;XI$6rrUe3w=qSc!(zj$D=xN_o%R%2rrB zw>UaIPs-D8xql?);hOGg@a7|#W?7G79`aM3#&MT!V z{0Wi$o1s```B7gpZK!&4LEuf!TBl7(TV53oF^u1O(i#WXW)UA#IW@3Hr|gN`3gPq5 z8lSb^D^(j1vB<7&u4?YpQAs7rMLkrhJ|$7G*nh{9Go0&H`lixShIiX}=U#m|dP_*z z;(uAF$Jp^84tjzX@)<0Q9b)bas1k+!#TenFWZ?7;Kay&63GX*6>i{kXdg5fLL+Z^y z7=M3~yD5cR8AMru$LO6S-Hs0Z@SURP?sux3H~Wiiu4gMqdpe7cZO*8~thXpi0Cc`R4KVwP zH!$J2I%1Tj)|a+@Z|k9e^K0GnWFTN=RYYv=PG(vlXd5R{BX!tgqZ7Tg{KFQ-(w{1G z#iuMz<>6~rjt!gvN>f-z^vGI;{oG1RoSlIo<37AoqzutyEcwI^-SN9uS{a#a+2U; zfDBhbPdi2iy=1B4XikTV(9|#kq{l^BXrI`dv{QNomXUPZ*8|*e8vYt6wZ?VkbuYwWBx8&>Tezk9e=}$I9P-~G1+=56Tg_!*E@2>-5?$& zXN>$27W2%sh}#-e*WoeyDxAXm@}pC9#~mGgMiSB_ei4$rdKNNhcy8&qc7ykD$znU& zTyOp;zxM5}&(eB>nSq}~UmMx}iHLv~Yn&j@mD!s1AnxqhN8#&}<7@VQJC=f15yY4G zm8fK0H62MUKsd*QYQijLN`dv88fcCULVQrI%jhm zOYeMtWp5}86LLv5lY#k!*-903*&ze=H-x2dr}nuo(ay&T)AQcrxOc1u=$tySjelv` z;@N_0;cPmLx5XG`_${Hs`CNa+!{x~G)U3F%auz=dxS42578S6YTa!Y60@b!&^(4v%aPB+E#{KPlv zxzA1^mUJh3Y_-O;y%o8^bMQU(pKU6yS_Bu%Z%ksjZ2~25Jj6k8cu z>0~RjrL>_fH!~;CH!Q51Ppr8JW1272z@JiAa)R0h3%&o11Zq9{i=sni`d7;SXRvk+ z2s)n@W3HbN>xdM&P)j^-cHu`&%_T_92e3){0@py*<`*TTS-AsK@=Tbu^E%v* zmCWzfUZ6E=FTQ?}2#_wKK#;}x%0SN$V)lWjhss)1WJ!qPO7;-a+{zQV5w$0+#ciQY4*BTXSsd&NFdGqmbY5J^LGYQsJ{ z!nvUv;0;uXf0%h%y4<}>2+XINLZzLz^?CPt^6r5w@sYSsBXec(z;K-cM>2;_ezLFg zSnxb4X5xgG*cW@^enynfRAw3e%6yz?eVZh1v~}L~?`f@9wFa%PDLL?u+{m>jCtF>H zrPC!>FNhpkN9()3)1%|pP(II-^8ai#WCo|Jb=E8YHVU1bD-^ zfZW2=+L2eKE;JJS&$xiFMlM3X0&W%=^_QiIUoCVAE9y|${9|@r>VDJH2BmriYWbA( zw-f#iP5=|5XFr1xJDpaGPxC6;gcG=^)6ojMM?UE)yj)w2)=pYqbgCI~++m9$%3C{n z)-GhsT4EB`)p)*e$Lh}`2TCCIedH|1WO9@KWQxVrC&E~j!sNN9Eur^r3F|y+O|O;- z8M)1|O|h74t4hj<2uLX0G!PAgOAibZ5BTW8E{4cXFwAqH2sEQ@~3zMq!uj4D?a~JSfyYym_6})zJLF=(J88mB*nL5+A+lR z2`raX4)lnu#(a@2uj3#_s^{^mi; z8BN}u#m^`t$8(l%AwEkZRR`GHAoJ9@WJj}GM1uFMX1S9*X$B`u&#<9@o~Ml)BqzPu zZtFabVl(W@GuYxm&nBs)t68pd3`A;A9B7e%BzkN zJJNhILc649KIg&s%Rt~Eb`yAY1mstV&dlmkint}*nxZTNp~O02A33zG`RQ-cj2*k{=I7JIM!8#G_**C4B!_wgBd1`S)=W58#4ebyZL(O+gq~N-^FC z1&!WaStB!en|_YeZwUFVDO+tCuJN4qR;EqQ2bNRhF*R1aAG^N8G0V{&Ge?oU`z>w3 z8{Zt!N)E|0^la*qrt(gs#$<}uLcZvy^^9U+O+R_vDYRc`x00@#&%v6LbBC?=)1TE$ zWzKQcLmKB}>4Q?F!%X6Pn2RHi8d2~k0yytqM>j1tq&RJulTq@D%)p+UdQ?wNyO5*R@FA=h+a#qs@1DCTA941{gMV$;*N@uu>$H`?|7q9v zfa05J5l}qonYq5ZES1lYa2d%zHWz1rOdbncQ1Kaqz8t)P9V&#+x&e{fHk0ZzvuB?> z1uD!|td>+L%~r0q1>97k)O|_}B;^8wU;3C^%_^%tSr!uh{r>6Cv*vaZD&+%~aNm?B zmwI?AyS%wCr6&KZt#gPW*|px{5;j7+Gf&j z-Ha3T6(}b5=Jn_Fx~dl!d%1aLYs({cx6WE10N@?@VRmE88*Z*{_Q>1@+_C)%TR-lr zjFo!tdA`Z17fN2s%XVqVGcj+sF+mI}^e#h!6^DiPjCxuxMO0vtq;;&HMGEN^T8|#-jI5a5C zC)TM!{;=ByMxTLRa{gs8IkezPi7dNs<$}7RE42^<`Gd|g(qG|EpzyUBE4hWe`_a8uIj^ERly#6h_Gid|D~v~Ps=;ogQEk?=q)>25EW zn&P`qURGqqOpp~GX_i90nKWoF(N;2Dx6-TDa_sU~)^S`~HvyD;M$s2z8k(y9-^~gU z!M_ljE7^MZ|Gp9P$YX&I*NkOgmf%dFzB?4gO-fxeO`C|^T{ijkDL&vslEmzJx&ga2 z9%OPA&Ov%93k`KTy)&1?+({ki_CiWPO%G}GpMqn*S_Ekjf$hQXg`$RlQ!iy>d8Xe) zC3rKperW;J%X2_+yu&vQ^?9IajUDcQ3+n>Vg9`F~5zbnt6!;Q=CGHZPz-Hx zLPxFKm^tw32Z3A|NzmkoDmg2C#4fGku`%_sglQWOptUzufqDY$J171N+QEJd$EV}h z%~xZjC1p@um>&jZ`{VWTjXpcqJ!^-sxJO=@fdnv4u9?jmCu2C0Ubr;;o}3>w|Fq`! ze$aaG=HlRELzoX20W5pfQ_eH9d&JV(Z7im`Z`6-RC=X%?}sPvoK4!fE#gm2?;! z2>o+D_kEe$di6@%K`1R3k*R-?rg9!B{Mp)Gd;j0pBM25)%ZM5c=|O+;sI!e9#i}s zch=~Mi@BAF6F~mj6Z2*>irP0}+2YSFil-ulTB_H?&Bf8+R^ZxSyzx$~EPlw(T1Kt_*o`V*3;0>TyjN{qwf#_Fw4U@wYMDIE2*IFGTonHTAm)YG5RtFT_K{oAp_ZyW`dcJdJW9EMpuRIsz3? zx#gE*gXA+hSwJxVbvBA!9w4mUbCTF}pV)&uu`$}KSaZ2?8UbyDbAE>OgizB(ODTL> zzuR?HUp*&x2M)U$T^D}|a$|Hx5@pW2P)`fjoj&svAQu5T$tReN@EH`K4L*FoV}Z=& zU=_tC%Ua5oeitCIiT$YZl|!^P!Nmx%7=s;7i(ru>NeFu)&ZSI}lTfR&M~@cF9hN1) zqj=Hl#`4P(<=)A+jY}&lE>D2Zw<7f-zJO<%2T0*ZWKM_B)OnFwSwAiL9vD1yHY;`5 zNV_jY`>j>5TeDdtSV}E5|ILuN_c=N-3X=oA$!@yRq+j+Gg4m&@@i6E5N}OshxzBw( zfKlyjM`fN2InrLU)Tg%I|n}59bvfbFk3P2^LUH-8<{*IVV7ovtoLk%m}{ttz|n+NZ> z5TS1yqDEr8>VLUR;;e#l4sl*V?aU^{FgKBAVH)LFB&zGAwx^Gdr5A%AUUy9R~)kg*NYA+PoY0q9LLp)sF<3lVDpl$An=wneCEx~0ZfgOYqb8%w<=$@aY zBsMdduv4|n{NF>Bdq&SrG|llYZ+-aB&AK(%z){=iHJp_XAhF&TLME`_BX!QFeK$AP zIRj)jN6!8e`}GoeOgtU@PcOa|lz3xja&tr5Yx55nCcr|#4nSfD^(fvQE>*Rr6vL~g zuT}hf4CHFe~?8g?n0%sdUJ$uC(WJ8uw z=~iu0TOB4wNxIZprr7(7(WC}Sa~gvgESM4S+^T*tP zq}uS*?fTzwS)K=>-I6pnLswXI(pA~2;T1VSL=xj$NvtCzDi zSgU-vc7!IfJEs(7@4GquDcEEJL>Q2s7&((|cl-FqqvbtR_}~rO8=M^o9fjf=47JI? zfjBoy6b+-H3nNAC-=WnY@FzVSHA>_;vwPW7v!K&XDVe7iwHUOgb{#xRZy8O;CoqwVU7zGv8S)d%cy^tXrTXW^b+IvFo{{^^;f50Z zqSkeb$xd4$??Vzj#Y^l>`$DeYr#m+THrmB8?{aqAyb?BJOF3;cLcIzFe`V<5G=jTr zh`-OrKk!+|nB^&(&BJPS|>2Y?Vv)tR%7R#%Z^3#9K>+4E**_o$`!^| z@Pkf|!Z2&YslUF6u|8Mn?wVEUrdaANi=YA2aUzfVNpaqc0)uE*&nj=hm4&Lnalj@yQaRk*+W}={xj9 zJB}ZALkFpRO$7ks?hly}7b`q%G`OD(F*I^%bK`;nSQ35C>?Bc48)2771vB}az`?XB zp#MvydrlE2+?G^)F#Y1a?AS4Q-YAMAm@(UbzrzuEx7Qp?J>|kz)!f`PpY|Gyp&Y)| z-D_!AOo}tWWR3>%j)NH}$#`==6@w>T@W7B%19?q~L+@?)Jd$t6;#eBFM7n;SB^@$q z`BFtQJp-jQ1RdfDnfdX-_Aiw~P#NqrHB1=D2!27rjog zsmkgSivu2Cy?I##|8$H z>QJgm1RoOoY#jqzeGWi_j7uZ1`(e7TRi-a#-Rtw4oqZYEdILKEowr9&G{Q97DC!9! zPrhS$Piu9s1Q|O7J&ZzreB7c~u6xeLC2=d@W?yIk$a}|lbg_`(P416pJ6PTN*D8TB^h zR?9^dt1V_(V1#bH->drKANa6Az0uEb#cn}-a6B$cR%HSFGYnN z8E38IgTnulE4p@wGw6_W=S}Ymp}dL(as}d_Nos}Kg8R?7)p_>hz=TNE-`)RKc@!lY zw8#|OCab{iieKc5@F-+`Zr^Vi)p1kD0#paQG5m*29~CnMu_uGE4HP1ca6KJC_EwpumDtNz++vqNDUB=x@R?iZb%Tx*oEjuFN>m?%^|Okq!u|v<^;XC`#ik+emU^5R0ou zlrEL9sR^wuB)mCW^y5A7w0{D*Z1!zlX?U6-_2=WGZE_0$G{D4NoR;N6JC8`dB2&Fu zzk65zfy=WxscVA9{r&*~H%Hnbs3!-~aWYN(%~d;vDHmuVf@VK}<|QVT>&~Xbdd#mS zQbDP{iz9Pw5LmKyyVNS{^*+62JfiI% z%7_4Dq}LfnG}bK&9#ZudJS_5EEVYPQherK!73H<6?znT}Y1_9UFAbq~-E?xB_(&5Z z9$?6L+ouuu#0J%)U$+Zo!Ds+h)Om{le%7R93J=q%{!0z^V>qF-u~T6&BD_!=my&f`0aK`fl%s%8+$l; z=XOLo5!J?}uQhebALA}2B#o(U;}0IVauE?gAKh1UK_0`Lmd&<3edmYo{~pxL9mOkj z$&M|1{|CuDhlAEn2|cX50g479B$Q$WegSupw4J@3Q{8FeA}Jj@B*5@ShQnYFz*RuP zBE(EtY|6SCkP0p4PGm;uFh8*0chc%sf|b(PwVSP%6ElV&lW}3Ud`Imq@3RVSCH`n$ z0L)=-OUiX8m)VPy)v`n%^xJsyEnrF2o86|IY^!u7F{*d?jH6>A$K2wDa!UjQZ>Xm$ zR|nmlDW*h!r?fzDJ~z@y*4SbLVDcoKgoOI*YB3i%g~X6$RKji@m*w>+3NR=la{e6h-L5euGN@mV*ZVkb(!| z*d+x7hZgc0a6nY7I!COIb!F}&VN*vLgP4NdnhJZd5Q`v7b(5j!+L?uHyqqoOI=K*X zeH^8z78YTJ5ykF4KiuCL=rCCAc8pRn9clNQj^PMo*Cuy7kS<{b;T}lcAPg{au6UXJ zt%>N~l)&&b*YDLO1?2|^d#je|+UI)ouGUvE0@t+D;r!$Gr!Du5n+xR zQ`~1r{X-IBC`tC1`~YyLC}S#X^i!eKGoF3+P;!4imtKu|P;T|W>*gzJq*g`}(JsFb z-h%d>_+B>uqF;2Ndu*_JbNXRQC-7yv{%NiM^y94!0Y27Ss}=D5@2qcU#;U&p`Y3`i zJqe0q!nnW=z}M{XPqMH0IR3m76%SV-ecm1_o4+;g2lKp6rim_KdJSKL?`I@7fY43) z4jBJ!jf6F_i*OMys7;lmG?k2C_~EZm4U&0uZ4I6c35F(=HjvGvLPHdRmCbjzA@OJ# zsQ+F2=4aECp9W_}bgBa5Qk4Xe(STEleKg!zuI<9GL#i2v*Pn+ejfS;mLjJLMe@!qs z0Q+G?8iNsLIkqd`Im2=XpCG|DleY7I!Lu_{+mJLoC>cK$$yAHizoq}1S7B4{B_MdC z){4s^!q#}DweARq8uZxn%+#+-->dFu#I}_-M8!?(|F~@lEqa^!Nz{hpg!k1A6YY{B z<&9BW32%y^NrHmQOo6daMS&P0DtJg5XTOlSy*}b+#q*C=QQ|lwRZgsw9df+qmtNkF zq*vU}6z8!JXkFJ;yBh~zn~YL_jYT=e>`@KQz3>>#JXFIYsvnRSRgYu#+6P_3TewfnT-F`NAA z*6R~%Q*5-f&gyTK`Nfm!)ceG|wFoI|OH-45k1TYNJkCr?Q1}K_{=o*3d z6QYpn{+B_oggF-BaAduIUi#H1;`+ckzR<3Hz(a=q7K_*H|y7B@mI8pqq(V&2~tG1JfT;gw5VTboM+ z4>3rT|J(`9gi&_ek1?Ve%_~O?Ko}eSEB~PZfFUN*&0U3FDN<+0#8NutC{`% zAhjQ*8^#>~cdV>DyXboIF{3kttY7L>C4RnLw})$qI_s$zyLkWAR3aAG6*9VRjz(4( z>pI3}1kRmOY_olL2Ao{(81%G3+Wi%tSm}W}E~*zw*L_^hTyM>l2OCkRYrLorl#Dmz zbf=xpKo{_%vTiC|HtjlIYedybBIaqh;lbA|@k%Pipt(EpakHm}?z5jp1dbI2waiYj z>gJ$Ip?3QX#6$JXtjvQxUy}3ArppzADqx;T2(fY-DI++5#1?bsi&c?G!mxV3NK@{2 z;)8uaGU=sNsu#w!5Rm=WUPfrLuX1}PQe+xpo7c#E85M?c;(v#l2rB~`WH|D=@yHDD z4YYCl?>V0R$P0=+3v1te6#ppS?qVX8141VLbcppPH8mK-#5R>LZ$TVy7#4^+E-tVzF%)v z4~o5?|KJ{_Ral+2_aRE`8^!-H$9Q#U(3BZY(2(J`)0GU5V|G^DsuMW4cG-5oEJ!98 zW$BH~ke}Xr`~9oMSq5JJm>H*R+^qhsCDs2K`K`c^^Q0Nf zX|#K@{XxUY@*BVPX29PEKgsU^e^AHDeO&rxhaGX+_or?)`;{!p1T`H@ELU{H?CJ_H z97?RtCRc^jsGs!wRJmv3x03{cscA?wv7^@01;CxNlHG_rb{~gaVuK57719L#5nN{! z)Vxb|dIeeU)cfamKZK3rBw5K)WHC)1b|y3?BfDS&QXiveq2&zN^~`KGU0t2t^Bj6Wo$CG66>Sd1nc9Ct>2YGd0Bs|6_k)ctTACb%0+!vF;MECUz(HMF66Xb*t#rK&f}YjvpXq+ZhMY?N@de@iP*7Ynwgh@|A-gH(W(+e(*x3 z*TUSXD#@$s|LoIlXV%|C@3aY*xlRL2DNk238PS9?p>YcNZqvJq<9R<9TR7^~hn2Y7 zfy*K&B@eNTi{E)nae;lw?_H`(l9bd78$HAPM%NRsj5S=qmADVz-}qf6MI5Bbux+|V z6^K00phC~F`A4(t#Wf;>6Y9b_1tp91aS01*9fPZYV9KIe%(HCq zt=Xb-*@yixN8;iR3#Dx5hXTewrKVYu+@@PNq4gJQI+G^Vc;ZRzZRp)A&g7|FRoMbu z>|RW&1}cT^mp`#T;=tv6$b@R(q=e3m>EML+iUsPH)p^_qryWe*#*@Uh63ER@aQvGX zm54)z{)!;Jvw5Vrsn7sKY<8t783YiU>jrHQ9Wh3_4#(S4Hf^oI#E~K$U z=8s}7dz^ntELY_mJ3CB@@X8d%LU_$P^={PEL1i0l5LN*D|{W-yp9jhQ^3$ zV*;8q=F=^FXQHYGH82t0sn1BzWi5vuv_rHA5@0tIBXS=ltvmdAUfX?@>r!Jia|&DL?O1rM-W7T!r&Tk%TCwfrt;KPD#`&=y7g_G9mC zKR^6=SKb%)X&!NM1$9D zvL@!l4jTuB2*Fk-F#fE1yHijba(O*sfYi+Rl)1`W%el30_@`@i=%*_uTM(g3=(CQ% z5b)-2kQk2_H7{9b4J`T9Jhl5ig3O~CV36?QV0WJLghTJ&QBO1WZJE^oyP?>nr| zr+C}O1NJEnmT8^+D95kzE`#-48XpkjTTf&zOb*EDIMTZ0%*#I^>hpHr%OmT$R!xxK zzqvdtJB&pzKRPUOYf}rua3UdbO)*6&?(1iR`q@&OmG_%fHgwG>pV^J;!HUT_Y2#A6 zSzJ|{oids)2AyR-)+47eHGa2wR5^;m-EjMo=UO3j!KO-m`xV>+rXrW(U1&LXX?ct~ zrj$VH^&=MbxDdFt{jj zHu?$jvBVySWOKKTXSHv#!T`Rc+1wrY0R#I~NYm%p!_r~Tt}cy#&>?MFN&F^jqap1H zsqqPowF=2EK;;=6C)k4>@HAj2Rj*{@gd+B(GyVXOo1`a6``hfvlK+l+NgQX%hQA>A z>L@ICq5hZ<8;2;5mL&e5+Y;I&e}J?WVGaH2e_<^}*x{m#t&DTa3H=|o32f^(MFV90 z5;rix1C34Enr~6bmmK01EOmBfBFgqWY>%HLMA;eIN`L_!o7YI7GlOw3J5UK_QpS%; z?T*?N7~cI%FAxoBuA1bnGk}+yN!p*Fg`!c>%?2-@18{U|Vht?3t=vmY+B4&)pY;3s zI9>id%+)0Oj@s*@^z}3=j~yZ5eD1>A#<#gW=JfhS9GxB(wm)!WlO){Kag!=rGY>q; zMdAa;z%9iZIsWzk?w|a}BdC$}$zD{T#D`Y%L41+)mG!oFfp%vgd`oZar)UI=rg_6$qv^v!2lZ?NmD? zt^jc8bE)xaoJ%wo`#Mr$RskK8mBrf|i})f$W)(i%bfd7#lpCO~a|xS1EW~%ZMJS>$ z4;)UKsz(a~ht(PPHcwcVtQ+b8+w(uq@z}u4ucfcUYq&Ij&3-Q)on&s9*{C*5d=K7t zT`rtPFbd5O0kd!31S&MAS1)`?`xdBL_i*LLDYro|lo^8i8<09^nW=>aKSiZZcAvE# z9F!uxvJMX_Hqe;uk+AlGq ztqYHAHibzRiLfl}l}WTGkQvWa3r;+Fdc$2m?<;RN&y&ct8`1mIuIx55GN8!!x&X;J ziv6eRFgO+FE&TUkO4RnGc=bKPWn5>R{+es9&uV`cy*>0Kf+@Y|F6qq??B?NgOP$>+px8aOFIXtCTJ^LepmNpf?7Ch zz{-P_)=OSLhXH!#5=nvHr*41ecI-S;|Hhb=r+}mPh-<&R_&|f*y2TIOdd23)LNrd)Ssb;dw_E_`i7r>)(~@(uw>&LQ(&GXij}Cwq z_tCvWZ)In!BL74OcpE^i41VxBu2!4MF)kZX+p|oz>YP^-iw`gwHXMFruk*!}3B}aH zgy3iQ`GdR_6vHC~teY}K8MntN{Vg*@TPfI+6XN^dJZ}uNfz&ZSHCX+o-UQ8D$?NUG zmA|GJwW`!hP-jen0+qNbMaiu~?P}n{rXcAZDl|fvr5|$7k4jNox`fB1i3z*1Wp*<7 zxkK&x(TeL$5*<`Yxinsq#zykQDTq36LpzJleUzO<)UKlpL4|}$h=pBUrnj?Mskd>5 zzOPUy_Bpo_k6~&?Wi+@IO}~bRJ2>QyyyASOCco}6+y>v@&GV$RAfT6~Tr!@~+X~L@ zvhW+nl#ngV(?)!5juT^jQslwM@Db+}{|~TpHlx2meZsL<<-ac;4GoWhqFJRegP$i9 zF{FQRJ+OCR&O`?;(#&%X`^W^tP*8k!i^<1h`gaf}y@0lbR`e(5>n7&oKYip?{oB4C zB$kK)o9;KFKl!*};0v5}xV!yaUhQgAZHzp$6)~e(KMFaQG0SSZjO5`4$GlhdO8~=GOFdIwQV+i67 z_Q@IlwUei)v}VIq`wa{N2a;@?$t?ccjf?HDac)0uI@EeFpa~mf6%L|>S#cPAjB9%E zj%~2q1zJ~Ir)oG(4#O*ovTD|!BAB~rg?3#%&bhOz|4(sJtgeAm4~{_U3S{=m+yEN$ zq|u*RexD`_d*-RSb;*y|&lj#-7|LE#R*$|?*PV)#Tg0t#CgoYKNxdN0XtCd1e=<-2 zkBH82LJAg-I-1TF_|rPASfRTYEw?xvH4O6BSF|)Wi-Tw&O;!#Q@34dmQFaT7DMK`4 zVLC8m7FRF79h?2PrF~{9R%gsrY)Gt2wj!$qAF~&!lq@z#7=6}@QzV4}zIf7#5@Yr>S+Ge(sX)Wn}*}LmCk1DPu!F z$%$-Maq+q8s-oAcLYNe<;qaN0?n=Em_X_^QS*>a}5EE|tNQmza7L&&s$UlYk5 zshHitJ{bO&I&k49psqATfdir3AeZeKyf-W*NBJ2_g}sj)(f0tS0Y}RtYtKZ_rg!Q_ zr>dX2Ac-0R>Gh-pp02;Avv^@8n`=^68+P%PSahs_rJ65URl*&(4s=QwC$NO15drOD z)|fv=5;_O;X0=MUZ~rcKB(@3V%)NY485~nhr6l zWN>fu+3*_UJ3N7yi59=*UeIZE-cHK2g1iVYih5ChD$R6=FtE46E;onWb;tu8tatd~ zmflo`^H1ll7CW>}K!WX@p#&MtcNkf*YuA|)7bL~I&4eFe^WoOd{P(v8T+d8h4_7G3 zb=9i;d~{*3KJ0QFtD&nN!g8a}=Tk5^I)L3^dkNQUtnQE`kS~JlsXH6A(h&I)R!AeN z?N^w6QdwoU2Y#5~$N%8fTw)=#tSAZ8Ur-gWK<^U<{i62V?%VQnBa1p2^UB!hTPVeX za?gCA`S`A+_}^zT-Y=(BMp3y?Qx#=)*b?YS6+uQDw)L~I+!9|W7qqg6G9tfjMM;?I zLb~DplM!!}r?};Jv&2BZDjuDQ(x}z(CFve6E)4ie5gQ_}ct?45v^-k+yp4K^-*&q} z$l~>w%87#~GJMLK%m@j*tXIo_1y}mbs_S9hSK80`@SYcAdwRxasGw5@k1BF=*mmc%#q4Nk8Vt>gZ8-7z{*OIdf{H))xr^a&1PL2IL zG|S5Dh>)EAag_Kkb0CrP>;N`E5xbQ>=W#w-!1kaEzhW$w>skc;Op8c8=aYr2uXIj+ zh{hUzi+pW(9Rh6eHf0I_ z$TF3NH-S$Ja(3l;;!p<1b(m`n_qe-Ef9M_OO{0fAkBUapR6?;{TMNHy_S`Ijsm4^! z-PX(>t=jIk2AAjaGM73`Kf|o-N{_2G*+OA?x6IMT&D;A8X7@Gs0dxnB0aLm1PwU|r z2{IDRal*>dUeSE1`QpsO_ZZR`xWXPo z4&kFx=7|JMC8>)afg;SkEr(%wEkdR6ZSFpe>g4mnA7ceHFrvmIDeO!%W}dV9GhF69ZeKlx4_s_8(%RZTVt%|+R%UJ&wU0bQ z;Hl;wUk;)2S6FA-^shIeK{>wvmWn?gp$ttmUbqMTtAl<8J7$l?v_TA(PwQI+ro3&M z5Dl0~EDrj*q+1*5Z)i8GTL9Ki%)xXdWU>FPVf;f7?Zjq`CG09F{N*dZONP&S3b)D>6;OdfEH--$Nmyx>C7Hw2@Q|?Fg*B&!zcV>%am&EgYu%^76qz z&B49Z5CibxX130i?o1OWEjDoGKJ_jgcq5^L6nT!d(80Fs!b+5F{H^z5!V|o&H2lF| zEtKK|PYF{J7J{ymAo?#;4eh)UH_lbE%ob)u&c#w)b>fbMk-Trg+c5`|{ zQ-`82E&Fyys*J@ z@!O2^p5=OUDQfE8P>#6$x#Xe8xrLlaBAfic#Km~#&lM|kBv<9G7L;DGRFF#>6cmqR zkqH{!GSYfJ4kD9e^xr5d1@~<=oe$3&_NdIKhG$nSj(E3T;R`M)w9z5mu2(zEzqc0l zR@E(9d#b{vGn1#cjpew!&V+&>Ne`#}MyKCUF(gf^`Ndp?K#r$&m|_~XQEXsjK8%pD z`$cqwa+-HTX$drJC+{~RQ{tlCpA8;Y7Spj`fTJ1Z2aKg9oDwshBK0Jm7cl6A+Ab8o z@227xr!2_a8hmiMbJu-&BB|4@d#nKFG9B>gk5`kZW$>AgapT&lxl3wfm0-!K%L5{d z7bcLYF-60A?m|Iy9qJo6DSb*NvQ(VNh4;m zA9G=7YuKh?#nCGXw_%R(A({UR^GhZKuc9&p3FUUSOkd z`AlUw2REKKc0Mz(pkF~>^usm8ismaB251wI{=*L=gNU{y3pDATnN9Z-jo}`j{H`-F zquYOP;K-}vS1_LA+Sp`Hjn_lENB+jj_4>D>9jtojw`$1D2Q{Q$AnCPI=?6V_@+(Xp zWnb`hjg4V}{$~f}WHIBl_9}-gS89?>huGe?0*aI0SX*I`u=2)r?HWF&MEu+FBNh-# zTA9&>ZK#3fPH!LHcMO%(gPMpPKOgXJJSlGwDOfZv08`su#C7%qef}Kj7wG&0x<2%| zS(T9Bc+gzCnmYQ*+pvwIAxJW=e9@@rHOD_Po73Qk*md0x4H>r(s590qpUtRtbEw&N zNT|A(-P#S)W%A$2+xZVR(0_nT>?4kaoGpK`)n}Ok7h^F#XOI961?GUt4|#7#hZ_=y zLiEp2N<9&1<2>-wWuhJm+z!N>SZdk=@?o5k^7U$~Ui#_a!o)sMuwawecI0z|j;B!y z(&-Z8Tb<5quiELLU~Q!iIb~M26+)dA(u7B7Gc5dIw=?)*(1z$>t8qAi5cdH=@;Hr8 znC}4X(7#ATz+1O|bn)9b$cI{sLO=8KiBl1o(_$jHp01pM!$U_M6fK0ecNNWB%f8G(lQ##wLH<> z(WY}ieu#ci3i4{?GN?SwVF^tR83xdDzRa|ym!%43n;x`%Y$Z=)L-;&Zv34(?)smz5Po zxK^^`4#NKS*V35`cZJT;lp5+k(u#YpEzGQG^_=o|@!=~iYr+%SucjQews+64O%u2J zZSF==&lq7s@4)E7P_#bBH-*YjSBdDbWAK57udh`(Dhd}1Mn)#&xbW`vW5zGa}( z1>nnP@1axYX|X187e1OhUZrC6Ga+WcFEAa+E8PMtuC~*zVa)s7Kr<`dMV?~`?J$Ak zY#-C&VuFrzB3E4?>_Y@6zn*rraoZhW4e5j<1T&bUjo(b6L4?3%lR}_lx3RFu6PgtA zsRJ-e;$Gc9PGFEe$F)BdSIbu`{6$uzwy$d1&9|jsyl4%+0$S2%UuSsgtnrY&HOW%ipx1&B}6LhK|r!=#_bazto_Q|<*P24pnV`eK; zW~MGB)%*`F^*VIzi5quX93Nt~ME?h*ySbDK z)B`-IYO2H0uB-!Z?PVwxzDpO`qRnKZYoOTO%^vd<`ixg~jng;(xH$L@EvC?=`uk2kO zJ8$dcgMWj*k65g34|{&6=Ob!@?r)ZI{rTHhsQS?%k^QHqZKlyX!*(v8{nt9RgToV0 zg&o54vndF<5q9iaFE`8Wsx64T9lH%%-GgA8QjaU+TBrHAyqgNl4lt`u0OfYm>BX>i za&jj;S^Bs`SpsLM9P{S67Ts&8%fp(8CmI7_dEkqDF5|#;u9<>R_IrQ_7&CO+ znHhJ?9nv$T5GW4rCxHPDErJ<|-!XT?g+hDfI6^~>Q+a$7N``>;7!0c+%(vN9vOMF; zqQpk-0&OPE<2&+8)fl< z>qcJMTQjMY!i<{5)Dz?+{zY0^4RJ4>=6z*+76WA$Koy4dYdx0~-^3H>t$w(bh$K-b zVuP$EYflGdNMoD(c!oyfd0*`gg`M=lbd+OSCs#?ireBjiab#7Hjk`nMb zo3@otH0Y?!sBzoLF*}}0Ir8}F!+G)T8?ie=YL<1Sn}KT>&;Izc!m6*W^3P%i-}qHL z*s#+-X4-|j+9G8&C}sRll4GHw`Am$4y$9WZ-D+W6i$?|tXr1UZci-;{J!scE&}isn zwvL1mM88lg=u@$r;!7z==A+7k*(;Ze8Z4_uTny&4#b@Gtdmgl(DnEwsjix-4QL?en z4odnX`_Q}BGZCJ`eEog*#MybK2W`hQc+Y0gJhcRpP~$L{NzXtIJyH42{J=6?)q2PN zk|t&2Z`7w0E-QXv7A1dwSY6=;Tw-_~0KbBSheXDZkNVn&+z|WF@DYn{OgS@1|X29l&Z;*--p+I`G{fsQ_ z^zGb&&>yD>3qsyQDzG0*vh25E-vUj1om@_T?DduOJq-lzx~j3Ip-MSV(qdVSrXt-2 zB6Ftgav6+&psxh=@}Um?ZoD5m%ASZ%$w58GiPgwFiJZ8=eruP$ETiS?(Gvaxa6i7u zTrQ;)ciIWs8Sb+1&5|rqDl2#8wS)~QGcoqo>&F^3BaXTwVVZ?OyA_RC~FBjXC)N%`$K#f+|9)7 zm-v8#=aBUICn-xBmGUdzR^o)ZRJTNk+jQ1PvlYXtXWILPJs-~6^v0cCC)GI*@QugW zUq<@h{ZMYt1)QsDk}naq8Hb(s<$#J=aBLj;FR@+(o5rN(xS*zMH6q?S6ttp=8|Fdc zNAq>bqoBtI&n3yL6haoGaw3%TI+efmT(D%Vi`Pv>Xc~t#Bkk|h5qR!6c8_GfJh4DT z-Tt8*tQ?%@vN(e~j!`V%Y0^0$xJhnH@^9Mb&0G|T6ZFDK%zL%=Utdb!d)7_1zw8B0 z<_SA8#d@Iw!7jTMKP04OECO3F_oyi~B&8kEsa&&lwlQXW;anxHan$iuzA{-+hhl!t zC)J#!8H@o1QKhj@9OrFNvz28JAU}bg!Dm1_RWi4*)Mvbw$5XU;bm;Ujj^H-?c3y6< zl=y4&SPM(;(m}+o#cVw|n@PPuu-eJtd!^3}`9JDRu{{&dO=FXXWiN6O+6fe35{S z=&y_j+|HSc#Jw;erPY1B@-m#YNUnV#EPf9KT>3Y%nW^Z}3j77@@WH>*v{clwdcG%u zU3DI(W-b>ge9(J5b3WNy9m zz|XG!4~XgV%vXU0<@3!dl~}BdWn8v)h>J78njHo^gEI&hCa@$I7>ano)=SWutEj^L4&{K(5XobnUB$lqArWPV!N z@Vp4e##27`m?TyILWFR@36~9QYVuWw_ZiUXBxS8l&>R3kR(rHmL1vW)WXF zv~nK@-hMJoj=Ln?FmgT?+rS^SI`qra|46w)ipyH(NFMX)iCDV9CiwnhtM9&0(m1s% zC5X?R{qCBKGfzI}@L-vf&kb4pFpgFKYAy;}Lga-0J$tpcadOOXCxN z6!EMo`v8mC)RAZOFs|MARy|nxlBp?fa5_geHMz%ud>*;c`GC`tmZ(gP2-^~oL;eZ) z^v&<=12)WO@1YlNTE6fU3nVTU2u!BC7=OxmQK%THx*K+O`l7kBd?i=HyV1_tw1TGc&78$;YUwE5~cii zl(PA|ONRe_J81g9Q5_DwLv#NX(s9T1&p(Mq8R|oKMtlq-A()MWuWuUma`k7x+KI$P zNDi87_jC1y7>oagOa-n{;Qtvshp^)eu|U9&*b_LDMe&pWd2}hYi(*_fQm}`Xq#R z)G8D*!f_SoNa7BWiwx%|)8M^I#`T-(I9$klCjE>qHpyabe(Ytg;^2gGr3m2?vglq7 z*M50{=Q8Q88!7bGw5GVX`QgylX(1N9u58zJv#|?qw=DP0`JUVQ>Gvr|Q*Ncj#&HPv zInH2(=ftEeg8eLvD8}0Vq3Br4A~y7>^gC^v+w~+Z zhT@PJZ9yudbb9)><+F#bXP7iX1wMJY^YE)@*WS@)tD(P64R|d%KBCMWXP)YWyUMk3 zf!6wfw{0`ud>WWY_t(hWIYynqVjz^TrvZkM|lwk#ov!EJ>)KbYL5MP81{cqf+q@fB|>4%u1<3G8xi~u zFX;0KqR4_S=MrP@n(1M${xEO=DOhYq_nz?*ebp-XbqRV$-F-7my8m{uO$g+HDG-dp}yG#s7k>cR^QS@kx8&(sJ+%VQfK9!||C=&bPw zJFFZ5+|2!RzspfkM!ngh_Am3J(?(iUK1OXjz%EOS{Nkv$TDQ%ke>t4-^CtTePM>o4 z^3ZJ_U*#O@x^}HCqcMFoX0PRto_~e*dd=QNN=v%~fK{omv|=md)19 zG;u8|Dd+sx^t6!|$eq4BSUy6Wt|J$#1@fzTDu_%8xa2Vui1;ed$hPHxr=iVOgomw| zzsZ6|9q~tQ08E)5v3!c=HC!8}h&kIB;z-h^q1f`yp`vyb)k4`&tuqdnn4`Rwl~rwC z2RBZI$WPa*NvSrc=8A&Wlmnz@OUkd! zCN$5uyM8>qKWfA@F$LLA5bwynpZ_hW%|I3NS%7X<`%h#i8OXVe&6Xpiy(Dul-{)Go zh@f`QrBn=EaM1fxVrsK8D8=sSkfyV)aO%w8nrs8ci=tM5ckDl$)k}EH#_N?bO@3Az z|NV%Vrs3&smyMbZt2c9R>=6Rvuq>v~N*N@|u12I6nXc1!`>@yv(0o-D5vDn}VVJk1 z_CM0~Wnb0@0NX{N+upyFj*~@>H@5;(O)ghDvWK#Uj@T4C-{g=W0elnkxpops&8?f0 zB1#tU@x)0JOV)3$@x3gms~WgfkgG+_#WRNz>xs@Q8k@Z9{MKnffV^f0Clz793aG9} z&dCGmE#bV(kgx^761y0sCloZ@=&@Mmlsy6j$N9tJl)d>4{bR)q1s7Do^Mg~$uipKB zS>+DkYp)cfRFc##I73YPOB3*~ni**#e zu9T;!SI$2mduu~0PWgR2>2=Yasui10wi_<{lXO$;tqFT6XY2(%m@k}zizw4UZuLC? zf@cf&3e_xoZ(~^u>Vb+;39cd^qx@diLsMt-u84O-9SM5D@gqqFT>3bFo*Il?lD!;l zaIyjlHxIO`kTF|n;oQe9)8L}6@yu4gtt3ykz5Uf-MCHrn=aM-*F^*Xkr_Y?52C|mA z58tn^L9Mg#KPxr*Ow>4L;nf$ZqhN3B)~Ow=^rNrS19oT7kotz^4n@f1)=BX>rMRC0`sO9 zT4qD-zWq^3D@(0IpXU2NEM*3MToWu69f(Fx@A>>}`Pr!6L@Xz#uju$@ z%-2iqG%bJm-rnb78(h#i4drrjS)4|$OW6$6L_I71RSfWcjc{BIO; z;LJNs_Lm2WIXTWpM=0Lu<#2{Vaz_Uc43UAYx9y9c94~nhfR$nCA?N`Mr zH>0hhu24}j3H9lzwI!rBm#x3Ogzu$kxFSUXY^ZYX+8y><<7x)4l%w$*28sj2ZsgDh9V zlQ($9*n%{o4d&Cr#0(J@jxelCjPrg=(JC!N=Me4f6A47>Sn-PX0e$+X#cPS=3@>3- zw;UpPOwe?k8UfL)3cPp7rT<-}=fzux8?-GG3tBN3uQ{(gAx#i@t$5zGL9X)LL0V~a zwiUzUfB{bwKETv^a)O4fJ0$xHUA~J*jxfTw>p^>cevx?J)))3NSsIR(UZ)AK-N$Sm zqwlB@f7v@$@6M(5nTb7JRggxJ)HSXcK`Z4?F}Jh_vN!%wMfG(s?|Mf@Or^U{CA&WJ zZiZO3q)cSTt3FXX$Xrs`wibVq(kU2NfFxUB`Gptj*KnwQ7wU63f@}J#XS`GT4-155 z&`c)kx~9tj_ZIG({tx3DLxK~73C|HozqGO9-`{E1P9DVZTFN!dRr2$aMQ1$-Pt}2&W>_5 z%h8d&qK)S|i2J3hD*u1mJ@F&kJ?;na(*H)aNU{Q0G4ikz$2k?5%#F^J=Cwf=e4i!} zPMU}XK zc9)1A^@EQyey23}xdRGUjE_Bwqlar5fgxYFvaV=y=t3dofL8L0h0iH5b{!qsp&2iz zf6FU;`5^}o_k6xoQTr7Rjk=A^M_3VzCzag>H-}umze$)J4RO;OX_8@OmJ+w}dA^KH zXC0i|-@UPEGHFA8X&8Rqy{E(1Fdf~I2up!Tesn!&w2@wYfRi3<=(Xw5{pl}uY*ueJ7W!znh!F2Nwi`RQ}`xGT;KUKxi- zsP^&O@q{<-T)vy!Y9(daXsTbovk-UClg>i}IFC>H1;5F5Hh^tXkvXQBl!CTR%8&3rqCC8`2hv;4v$DishGGeDQI!hm4f1) ztK2D=+OhG+Ts#8#KsO&8+S{z_q1=nAeBXzMDGy&fnasO#Ip2M3qbX=3>SoXm^OTh< zC7)|}`Jd-V(d_|SpZ1s*r=qF9;U#ag=_%UHADF%&+-gj(H%uR#3tHB=pJ=8pu(Vi8{w&T1TH0}MF zXWIMHk>s`)^FhpZ^uX(0FSFY&P^p8sy;KkU)gG6vAFY31JM=fjdp$edb9S9if(y4S zJlzZmw*H5CRip0Taivk0u=oFI)`cANKE}e+eNu3y%L}RHAIwuU9QX!MYGVEgc)US_ z^$iqYO>`CSZk*;!Y+<(@%7nb6cA{co6Nqx~LSc*~dLMvtHg>kC!8t~fS z0q+uMqiEvQa|7U8 zaZ9-=>YvBxL|0IKZ>#q59r3R?EI&jQ(mcAOVc4-&S8&A;;g*Cz?3D^|T70U?$Qk-2 zLucH0_V?db5_f*4%9TcH);q(aj23A4ARe_uY`tOyh!Js+!vP!Jc=*c`iLvCPCUY`#3Hq28x+^l08xPzUpU|L~<8OUbA z@)gW6cQ(nuEzh& zT9e_tzvRkDonAy~GqOErVl3s0r)v?B;|ZC7e2}%}I@)IryGGeFV5_{_er=obC>!4I zvtigF)Vg9oi!^7|bA5T7-wU)NXrQf;SN$^#J*RH7x!)p#f9*LWV2WZs#n=*irV{{<}-v{{fwADr}|jgBZav8N{P7Y<0g(UhtFdb(PpB{-TiOJ=k69X z7EGv)x-dj{l8D@MRS8J4lZ>r{3!`O_0?kcp_(`8GUTR5FB=?kmT%i+J*=!ufgTHk~ zzNXFp5Jt4lkNAiMo9S$5Xa6t0zB8)H?%Oi~Bq}8;D!r)?6&2}(5~5T^K}AJ6zJQ?g z7HS}XiV6e}kroKmV4?R8q5%X#Z=sVALg<7V2#nmhcmD6phgtjEleLm3YwdmZ{*`mi zGx#H!&EIL!|ryOprws+45{O5pmR+iTD7ViG*Y3e&g z#wkH9jm!ykt-5krgoNnvBF0Akjp#W~gjK^Y&@X6F$nlWJhPQq=f~7#Gf0{JB^G38@ z48^+7N(&aNz7uvj-%budu${)75MS=%B-igP-f0-|w#A*RQFH0`wVAY;%TVKK7e*wj z@W9CB7ABN)=D*n{ZazHIzDul%L5RE8MLZ^$X?Bc@?oWa2!eP?{*fYrne0ddQIdD3z zYv$_HLDrRBa>ZuYaR_K5?qed(Rpx%I(Ym-_@&YK}MM^W@AVl>;rwrnNs;luW;O11y z!dg7nfm;*f8`s+M-(G!v^o04o=PJJrOt2+MG#{7fn$Lf+YKi%pG8nK@XIvxm-ed5J zCr8kSO`VNU>2QzW( zY%(%bFQZ%%SJt*b_!?&@bchBd?y<=)2=9~Su*272PI2vpl0jzQPkde<$8^|yC@Z0; z?w^iO&=yvgav`e_HH^C4-R^$bRa%%|M?>8=o}Yhg zC2~3$8ly-vB(0=9qpXg{2H{$|aZv0>#4fBp!I3}M$r#T+*`D|^fnMpd&MopGpeWXr zk+>zpB>#xon&;5@utG!+^_I3lwPx*nki~!71;#}Yx*a$5Bcv6QkqA^ z8ZUjT^1n+kI1-TA5;ApS8t#z9|LZhMiZEpP@05Lb;!cFn{k9v&Cu61Fjh$LUjEcXk zlLjZH2eVy2!wdaz{MN+V+@eu~z2i_?rJ=2C{y}U7=WAw<2bj5W5;=C z&n%0b7XN_KJNtnz29P-&bVB>G7`ymueH6Ts@9OandM4**KWXfQ1^=F7I?yq$!IR!E zWam0M?ojUeJZT76SswUqT(E2zurfrfn{)us5NhLY&0MT}ARUdRmxh8rP*Y^dh^?Mp zt{`H}2f3-pIZie9Z!fK2%ul-P;(qo?_AwExw--GFsnjOE=_`uHfE7G@Yd-ggvgI*? z8bIR#*3&pRGC+1A!e(=0Ca85N}$F6WAr>Qa`H3n#y)+^8RX54DC!hetUiB>7#EpKvI%5h6X&=-{=(`+Uw= zSJf{Q*Mq+PB~*xc=$k^{NpZE+T^Or$Kk0#SulMU|R8g)q8Ku4NY$L`bh_t$-4M~@7 z4sfeiuUVlEa6S1i3>20=>bj-vlYe#zeE~V5OBBXEid8BGe@K;or2lY_eKpDl!CL$z zl)abscWh8>bva0`SiFYpPVI6wO`ydVz-G*_n@wX68XTj8UKQ}_1?mhF^i&2!?KZ}h zJ*m^3z;;BD#A0RF>sKIG!V881!cquoPwIGzHTQb^7M5SPMVyY2M?0k&4prB?1*#fb zb*IUEnt9%_s{aL@^M`%p;a27S?%2(U6PQftnv9eMO7WFQw~vna?@x6qoYwl zs@&>cQS2yo8hWA_z8h$kpX6eOuCgSVB5hy^q^$Z{D8VzUynOUZut9}imz@tH(B6A) zOI5MHg<}n4!tO-V(RfrYbcvxyd*C+C**&EryrhvD&}e9nRj%l$94m3U6{iC4lO@W@I^VckA#?{9N!+8BruL+7(AvG5uSK4l>S<%(YGdiBWmAvj!#TVOd~wq{lWQ4Lw35+ zsO`_uqM&y>0%1a-WziseMU$odK9KB3lZGY#k1B>Z;`yu-h4HxSOJwtVhB&za5K_Td zn>&CTtuI~YwV^NCY5;Tm%_lmcoNORDRW=G&vfVV%CcQ{FS10IbhreW@a-+LpE3%+p zs84hkbCa=Lf@auZ@?2v{SA&SoE+l6dDT8@imT#zT!%&c=o?Oh{8o3fCvXS1SLFWsm zcN9&HRTc>&5=JEUH*gxrA1@8=ug`=O=qY;Mg9*ROoApB)cU6~2=GU(ODM}fb3!guA zJ#|VO{76RCiQ7?s)Jaybbs}A&Byj74=2-Pz`?|4=6{ioaS6(^|J1hwZ5|j{DDWQx- zRF^B`buqN#Ym`Pxx<;E|-KU8)%2_p!mN^yaYHH?lrPB7;HksCP(r>h&Ah#+#3mTfB zNWD9XPZy@GD-)yQ5F$*fE;u~BN~-(V$gk)?@pG4D>hQo2IE-n-(LCNJeZRYbG$CUU)64Arty+xs69(mvu~Sv?@& zeRtcjwJ>b%q86*E3QvXRK4ss8;#D1Y?eFizweL6-Fz|u6^c&T_{fDUzbG6j=@8g8+ zWEc8~3y$7fl}dXmnHn80st=aKrlLu8Mk9aB1fKmXl=-=?g-^9PxDHtSpJ&5i>;Q38 z2As#*I6Xu-^l%k?c=JH|#FS%;=z|TB?aG!DW&ObncL$(Esa?hK^z%0tgsi+!&#^Bx zIj(_pM*Zwt;w*TD+)>VltT|&4^^ifPu2_h&uyQY;264gt08qX8zJx1=bCEr5zCaN` z#P}Zz451-}fv@8tvbrG@!pP#;BG%W7jq==ls0tfD_;UuE&H!PNvzwdAk{q&z2_UXv zO%a^P13Fd|E~~z0B~M_-@*XQ)XqQMZ_n@xgo9I&@4d2oei(eA>eeuSJC~U~BHaP57 z5CO2x!WREkDJ+MfWjSj*m?pzg$nt0Y^FI0Ua)}=XKm!g!bCf4H;?Vta=O3ZfRh8$v zm8Uch0sfw$Lc{^E?6)kgSki-0T5p8=wA;-J7Vn;v%-$bV~-SknL&KbChzdq`%CXbVbq4l_h5nBy$z{ef4&-WD{=?7 zkWbHTy@jWUD!yTjK*}91J&@ClvSCV^A!8u|uKuOV9E7rr$mNNx@(QjLzouinoT?5M zi~WX6Ux&uJ5lg!ow+vJX3JX`l6BXmc141j@KVEH4upt!fz6!q_ZjWf)zf`j2Qol=? z=F3r19C4`&-OVyE&@281A>QvaaJ$qS0sYnVr%CU8DJ63E_lwv&>O&xp5;^D%WPX4C z8U(iX-e##1zhl0+R51y|1P&5Of3v5DTQ#IYk9{cJF~91?hBp7Mg^}{rHubI=JW7bg zqFw4UWFS0DV8aZvr-#Mo`q|#&puFGg4|Ka{@6aJZy5gPYG7x3#?|Q9UB5X4g?I+n) z^cWruI-qqpj*LmljjZb8%&&Uor;%*CTZsIxeY$66RX&uLXH&e9+SswRT<3YAxGa@N z&R!x**G4bb#$oZHPQH4iJ70(~y!V~cC!KG?ymTsPUCdWMUU}kR1Bm$|BBbV27ePPX zyEgS|`&*S8T0TGM>u~`F@n5BBSZZP*wHj{>LRNnrr{OuANfNF~x5m$XX8(5>gtLFv zT=jDBUNiqs<)PVGIQ{=F>@2r;0uEmsXAMMjtE-T9xh0OIsi zg*gATvicBUmfoRk(j{Y3EtSQABSQa>=ceLViVNx)v?iWO%*QTYyxd!ZS)i1*^E`1_ ziyb*uKE_9i(ovT!3Rpah?IaEnnne#*i{7@mF3;o?5aqGGVZ2Wl0k&4FvuEuunrj4p z_~cYEju`D$ud94fs9#r#f*d&Am5%mp3n@wJ>n=kze@@NhJ*3~Y+QtD7yK?%>_fvJm`xKWGG=5 zl@WRJ!ECdXd#5!RQi(R}k+I$n14bG7&R-cjAM&qv8hP6bCOco0^IOvzE%MxGLt)|E zOYb)8{Hcz*SaJ$(?Y1;tN`#hvS{02`%TO6@>49)m&*WC_-pVq+tHCf=fBC$|(z&ZP zE1n=5-;g>475VM3Y30x%RbXM~u!GPQZ8Tk-{B@8e`@%FfEZ`w z3}A+HdC#8FWB z+clF+oq$Fs`JHy~Soiq{+^$Q&H*wz&+!S*wxOOgA)SOgrd5oUvD%>A@Myzvi=f?E z)>6v7fWJd+0Cjb`Ioxvdn=#kO{6nn*EV;F(tt~w`wEhyP3bk~5mSjuUbOAu?Vo&qE zuaH?i@eCWYxvQbJ7z!k(h~f<`8a28x^e#dp!qG!=c>KhVZIj2f1JnJ}tMR=Nu#Z-I zcw=n8%mF6rPphsg<377}{rEo=_Nwx|$CKqowe!mU%(LX4C!X!YJvr{ZFMyY$ZrJPX z;#4KTBW3p~lxX_kdp~QDq-d|}sdkf5k@NOS-sLajk__6Me#}>`DIh49QZtK;?uquw znAXHcJ7IJdLw8%xFfR>e4QNr+G9`fu4s6=zk9d9-JEDysU9aMb|T=2jsM?k`pb*51FG76<#ay8S9_ zCtgTUL67hizwASM7{C;>cDogt?Cs*+#bQ&hIb9L_dE@0MM17u^94nwvBOc!{bxfMC z{Jt=^{*VOroE>O`HLvyHYeD3)a$4(>%5;PH<`4&DL;HnF)(T!9i)ohzTlp~pA869` z{#IBz2hvct)5y=O{e0lxV&o@H3!p5sjeC^8hgHOIck&KhR9nsefuLGMS{1h~Of&{I z2^RRKYhHBb>{0%l***v@5MHS8Z-av!YN=vR3{!}5b7Yh7wWn5MCAH++;k@;p6Jitl z>3J@GwDg3NfBLlL$?tKpNW5r+ALsY)X|moYX;&s)f?+F_kH^W)uBux0=;LKy$N&6r zcJ|Te@y$?hZI!;h*l%Fha(DX?WOv5TUadQVaY6a1ZvQ0DS1YrCJ1dcgmcIEa`mFyB z`9h&kN0Dn>{}YQP>@57R&il?Zi2~S;&;F`f!wZ567d&4zK0p^a16Ukba{&mTJ?aKL z$m_yQN!@b*$x~Mz+({um1ydsuE?Y5$;El$+4sACs>AwVg^xS7d?IDzl-EXkLuQr1V z?=MT`SN;R|gRwxIaC1=K8|e~hRD>8Kxloj|y^s)Y4^!YKTkUKdmxcc$80&SaO2z2w zuo%{lBtYVlQ3ly>Ey7BT!RD)@ThZv+!f-YQL(Ht}W!P#*=N*P(#*s7@^3uB=$;N)P zgbl&uL&CT`=HvlyXRw}Sz*3oR%j+;5jSxzO6Fj&EbWCLqo})`pQN=1VTOIZni9P2b zC<}FmK!XRZyST`{MswxBomGG~(z1UiF>HNx?)c5~_|vX|Q3~p;yI1=(WRO`BIQQee z_qQpwx0vg*J$c=}S6w4@n_x`0(06#r)x>*@Pa>uu{Idu33EhEX_boi$&G(wiu0-AO zNCMk_l&0w_0#a=Zgb7Y)31-;@<40C~*6B_<{)HKVA?&bNy9@vP7m=leu0d2`$2^Ti zl9DyzzlMLdr2g73Pw9k$ApD(e8`&a$YvMQLUDJII-J@c8|pmG}ZqdwF! zv+PoL0V;Z}n_a}Ab)XV&Xw~zKb>}_pdC~73Iz5!B8-BD2%rTuSk_nebvY9>fu)EM(cq>B>6|4{=gu;nG=Ex+(#;o?xT;&h8y|Q@c9K{hEnRJ{ z@E+Y{;Yzz85X*c?NJt2bil+N29y;Pw;u8i9*6+moon~*KR0b}6vfr=pyP9I>gKnty zq&+jq@b1CLYYf~GUGGUO7@zEx*$a`ZiVQ(%)O|p|aWZhG8UrmD)WrNZgm|+Se2Nem zlI2l4l0A42sb#Q*{tFMVMPCEohiu(KsM!=!njnYB0`z^S}m6@aT%Bry7+UuEb@^zV%tM~T`S3-6d)S^nfvYG3x_Yj^e4(}6W9_3s3 z5kEsd>bm{7UgY0=e4V1R&8T(n__rPl8v=>81 z5m;xllpIW+)AKoa*BSifDo&NzqwCGl_QRs+RPbykSo)NMnV=AEP`_FNr&1*q|ps0ZvYN4|C0QW{LE2`~F1C$ z%O&9&M7XjxCTqL6uS#jP?~VV{_=I#>;h`uN@IIRLEFeWl3wEwe z+;?(s!UvFl(7^4$LWYBD0a|AP@+dCrsI%#yDbO-#SqK*Bi@FK@Q2IXlUBzPyF>cs8 ziXSdgN@Se^YO}mimM%Ni|HPQ>3`>CW8lcw!@Dq8PXSJh>J)8YTCg{wrykt0abn;-U zDB}L)rr;E3wq#hYbL_Q7fshU;TiJCgxSN}+F6Sci3?LpLx)Bl$zX_N=g6Gp`r?WX$ z*WTZ&?iS@1HMlmFQy$Vw&M=$Xp>jZ@culzt(C6Tlsjs~jqWo1IUFtRmGcN9n&=XrC zQ(q%G0_iD}LWd-FOpck#U9ixp4VSc-b^~svI0#9DSoDPjQEgN*!BaK$vCU9_=RPGj z5mDJbY%Yv{)yPor4JBK)g<*AEjhuedxZ{q;E8aW=6sW8R=sZRW!!oGfvkv$8X<%XM z_77HB58nzi{Frz_YI;Pi;iMEi>#EZ}at560P7zzz zNHYu|_Th;oI>)Nv7iYew;)%}&?5jPl878)G^;R_5(zB4rPcv=jW1IGsG&XYTX*J~= zRqz`d2ONeYi>I=Ay!(yc8YIp>k#tbC>U}dSz;sFV8(AK;^d$_lS6Lu3SQj z8I8Cb0ZxHi_Uy#=zR{%=1f@FYMc%t1ZPy zO5r3yJgUj{7OER9jAl(igxaf?LnuvjFWlHOet+WyT)<3*vLm7U&5NK>0uQQuv@Q;dq=FxgZ>}tCF6*CIeXQ!$W-8& zK;_=c#^&Lmx0|cwSR#x*dJ+YEocR611C~VVaDf&c)Uhc>@=)yyKo%TL;rDO35W_ya z5dsdhR(`LXjJk%A-eJ7~=tjMVodCh(KpTL?E3B1wfnv02uCo`zg)QMP4K8!D!bR;L zH=gv5jU@K*PRa3i&rEwJVqP317^VQu@2^dj&TukGf4g1QRKa09i{hwXLJ&aiLcQ0p z%Qdyql%S3n6J2^_tf3Y8*p4zU^yX|F?}OtH8m8Y0;^#Dh>5Zada&srbg}UV|_*rE~ zLO?MUnjf#>Sy0C-B^?bo)?M-jIXI&QuO8}Ie6(5rlpnl=u?%7dWN z9%B)v!4=czU#=HC9x)dYoz<=$<&Ix^930>>Uy{zPp4%*8zMF8O@GnXXuerILd7Q-p_0zIN4Y#Z1wKFd) zDZ>Z{LML558J|%a_|j3#Z;nCV(K0qQnd{woQa&ysuwbj&;Z2gkZM;Btd7}fH7MZ0Xo1u{Fn&vDe7CJ1wiBh%hXxd zbZj|*18`yv7=kjj!)X7}N@qL8Di)$-qJ_S0Ee<&OAVUO(GWYIY5ajP?tyhLX+!Ezl zxo;WB5g;43#|qtl)-8J%2sVP$)tK{!qN-<5_`!tZ=BI(b8DO?6P6bT_{j4k==RZSe zn_5PJ)GqY$(jQjkQGVcScH4@DSkL(URyZ%?o8iX$^M>52{s7b!O~+cATv98mm~6dQ zwrj;?(4L-fT^c%_5M14_UZQcji|6Qjt*XGm%tu|a$O z6qH^V9p;00j#nldp#^LDB>VpfH0LUmc52yDZvEcj-lY6tFjuDZ;Ml@eD0&m;D`(uD&Tqc>g9BPPRT zEx{$PT3q0Wx6P~j^e$2B>Y$fq$iuI%?8|mrDVO7_&rh}WtT&4uxUH^Emarr?WhJmV$-=BRqpLhw+H!`<2*ov?@6LOm(pB@YuthX*#`}j? z^bcKyq1c%AQ>PS0!v##_DW`rAx@G3v?exKSrcchoqMMmWKPn)R&=dea8^7QShgL4!cKk}(nv*-WTsTcbjJ&vcP zlIs2I}Lt2c9Y`BB+7Zc0Kwu_^;NF0^pLgdzca`@mC~ zshWQRXSmXz^Jb~FAl^Fk%`L2g80o)x)Wv=iV&P{3dw>rCF|Pi7Dd^?M60>8iUIR~z zqYl`ra~JpwnOFK-mgq}?HtC0tP9MLoOXQ_Cp?T17)i2aW18sxrjk$*fjJ4=sH9VS= zVK~|7(a2RN_&50@RfksgCJzr-ALBd$il@eePPJu8Zy1OY4JJbjg|dmPU!*{_j=C1E zS#hNIhJiEP{7~xC20*Rt^u`@tO7!lpk#8^+&xa|f09k&XMhNpAb@e0ou^jgW`>!yM zOAf}!M&7E;aF< zrIf4Xgp4Bh;Mbzlk>TDF1+p#2W;2UQ2m_j3=L_8}L#H-u%(Oj2>0NWmt~E>Ln!J|2 z)0gczkg`1yQcl2bA_QJXZhHO;K|JS>@HIhfSY>*ZU2OaI9V~v3A~>=7_Bo_OMy<1f zToQ-reD1jRoMg^ZTRN2-j*e=}mltibu+rZ&e)j@{(bk}Z=z-@Oh1{JDC-IFtfg4$( z!qQZ}fT34I>ZZ>Dij*)o1H!aXval15ybF{Vxk`VL+6104MAWD=-9hUb>Rn4ytkoE44h=1vdLw#Nz zYfr~H&XMg7*;>}iTJo!~9vZenZC}kxKaVw-7%s=w`d-}p_;_2>x6&RF{1&ns8B+91 zdK}U4)N->e48cvWV%(m;SAqWb!cX8n(vnAH2;UR*S?%-KDWG)bgP;DDMDTU&uCs3? z`1MU)^_>(|bPhZWjHGJXRA5q1&pE*)b~AbF8~`VTuA_inJGmE+wX%r0RLy>N=S-Z% zuwV%EGwLYd^%#~Nkn;31maA?85~2@``B!g~H-N1yf!b5)f$J2#OM*8w4*}(W6Y#R} z7jAf@IjwDh%fVV6GNdqM?s@jvTJTd$WP(-D?n8g>s~#-xfEUQuJCN~1elNWqd6bRz zxTuMtx1goU%aIYs&*n&>O{PkO0|Cz5MR4)zFGfgwoXu};Lu5(cZ4G4_^+E8b={;O9 z1E&cg_Jd`CAs-Eq&O_X=4V@dijhAh*gAv^^OG#aUl*>4Kk{A(rMPUnS4c&^2uF8~d z^z~kCsi?AxY#W+uWXagzf2VB;Pm+=3HR(~tcgdvD`KGRaLdSJ>-ShlGaxnIPC2u{! z53;f479CMegic3&RclH{I7RkE;1MEOMi)D0iUqj}c{AM|u9;Sh@XQNEW+}W(6){>)TXQ7lyA4MgB56h>iIZF1gxg!OsW$r1Q)ZL2BIaV& zRvB*0;Fjv~wr2DcJr28XYhKgp(h&ghX>}zDZ7au@B^zmABsIAJnY-sdh zA_*^GEh>1cls$elgab-hC(Abc`g7MO+2@azN|cpX&a5uPnJZTp(i!Yc@o#%<8j#wx zF_1!kQH|Dj72-*!uBhnL*m93ALeBh6F6n*6gI-EKAiAGWQxaU?5b_LN!x-hb2U<+o&BN%(E><|FVff-#bZSjejoQK=nyvbNW7xTl-y?m(%c z@LdMpoFW6C0`7#cG>!sPl1n7JSq_wu0HH>1Zt-1=iEXpyo{+0A&=FbPPdK&i+ss_mG8>?-Yriz-VlfaA2jzYMzQx0ES%euF_>4xHp?EV@kCx?_ftERTlJ60^`4K-2HZAe&mZerj58sR zY(Ig>3KX*3;CkReG`*_c!3Y40HNi%yw}=2YK*+zDly$`(@19|!E+J}XWJGZ*OI{+X zcC_;p(o(U*Q~z~=*7N+13o-Yn*a$Lt#~k~8b!-gE-Wpnl{(fE{P#C0TV`R-k=FjR~ z)OG9|F1m<+QkXx&rg$x&V=xiy6k^ePBr4kA;?$jo|5+-DUwpvy z8t)y-|H^oZwHI!~Hvhmv55p{;TRO2h&tAcVJm^$m$QH5SZ(t%iVwRD7hL ztCfc_cBVpf?%_sDYt*s39W6T@bHuTNn-u}IiWLRVdkCAtD_Xo!r8b*c*YmDb?~Xll zjrNfhl8!QV)>6d|8QA4G{PIl{{AZWG|NaF<5S!XdaoZd&y6s0vw;57+iRgCUEODnO z)j;g5TIb)$A$~Ex2PoM+@>-6=_^4m1kp1eIRsVkZJ!Wt6eqbGdVkajZUZwjhi#Kn~ z%C8A6>^a#sxciaaM7x|{+Pk#5d|-&73H~#pyD{_sa9k~Yf2SG|L)Gt2M>9_35Gwo& zlM}*tP=l4tD_;&0SUm;KK3)lUb+Yjt0CPmrvcgvny-_CE4;}pBcyqai5RBaAGw}3U zKCv^_<$Y?$B-2@4VptPD+T0bS_1`cdh0K+6{)s+^x@LQJpqXbUv-8OD%6jVH=#)mp zB{z14)m2yh(fhhOQzygEHY)GZ5)|XUZ>dUOZ}>!SMjt2W8K`-o_-h^jhET%rQ{$K8 zRD7ok=_?wr%xJX~LHT7~bqmnd}OfptGl=3uIU)hEJU2EZz$(c?bhts&i}I z-=EXappl)0LNtZkP?w6g8q_hRWEsQKJ1ua}sdG0^#$;>kB&vE{0eop~7e9_s+G2zr zU^-^sIG*#%{Su7PXP03RYE#3OzNxICZxph{!rs>7yZI;)v zTU@6IP1`xr$d|((w`=Frb?3snL?+#uxx$oE7RLhDWSjru+%VwVhOM#!Vfw;F(cM-z zA&;Bx?5`G-i<%9W;xD*ICM0sk z`=Ij=x{8&T)hs<}mFrLicHb1%9{X?|JO#HWG3S*Vm=@O)PuQRJULIECp&2scDn@r> zCOYk8@0jsEDi2qWn^gazzNZGR^j-d~gI!uDx< z+U4DKL(@@tp3y_f^Tx$r7b=C{EN_;k9L!JqJ|DDGQ4?Wa{kIuTSy|v2cTQ3~-4M!i_o&-l%YbG23o!Hi3iOP$`DKKY8$G z%?Ol2K@Tga-}rAIzmN_OBItPXv}WYXPi6ZP;%D_dIL#X)vf$8z^$2zdTt~$EHdDx( zc9M4fkLPmYHNY9`R@KQ}!Zas|jqYb)kQ;HM+5sgDGYWK^275p$ht?0_{%zN}#yqZNmeIS0|F<iopRP5T8l_i-X&nExgmnBwYJtmwJasgyx!qIe}+VGH|?{jhtAxS!20b$)uuHi5!;^;`H`%`?|tv4$PL8)!8ULC*%6hW zCbsN9lb1UDC2tK&yh!<2bQ8kHaCPc{swjm=Ryhy^=FTmvaFr5#Ou=|}oPzDWJfuOe z;}~%>6zE?hEDLuC9D$TN5snDzBG1~?Yp~~m%*`Rub(RQ!O0zwTo|NTy7%;gvQ7B~? zv>w+IxckSrUiQ(_2vRn_{ks3)% z0sG4rN9Va;vEUp%UbKU`JdR*eIuHkIgtljO#CSpr%@ylPA6x#uzMlO%XQ-H*nw{t4gvh~n0Z&?@)cvh`t+Q%5DaSzD52(YX%_#wp?T>>zdu{0IiJ7*onfZEp z>kH#j|7GEMtvy)J@&sCLqV6J2)NE120Low6|0bMQZ1q>ljrmz$hFC?6sb>@)p7m>ExV z++RaDnFF6OBvmMw4w+ZmQCmx7pcXK@wf~tS1>cnjqQ-)_h-CfG8&F(l1hjxeihpVI z#wZI&vgb~uN5!wj>vpYo@U~DYZQP-3hEWHGZzB{g2xQ@J1dT4(->Bf|PzcQ{M1=IM z{o_hmQ=kkMD=?2_)Zr6|*Y~HW^ea8m!Mp9vO*m&nmv;zr=kI44LZOKLX&Z>*1l&y; zS{QM3W#C(3yE)|5m5jd-s&dd>XS&nnn|=MuA#R6thWwSJ&PfJQ|6FRE{c?q8M!gP1 z8;1ocR6aH7!@DnkuNRDS0@%OpG7E73&8KnXnmD}wox27#=#H2uTCf^oqu;q#Q8O@y zP~SJa^f$>n^vxJ`qv>e}Cy9AW(A4OxxWR@$v%;U?5Qb0YLN2t;(rMwMx--t}Nm^B{ z@KveQ`K;8IT1WGj@}pntt0tR_`*mdll0K-%c~IZO!aH21=NnQBs%hVfbgB&>4{`7g zo)M%55vuoyL-mS6;YTloXQ$kWEwhq z$2q!5k*Q+!|69D%!&8Mn80ZiBC#qb3r7o4oav4wrH=5EroZPM+s&fe>Ecpa)`?T2i z*Irixd8&zVgHgGD$87;8m0g9Mx(s(|dcbqH#S%yWu&j2(jR&m5z@1TUz{y*q$2PSA zZxw;JgwTj@8&*a~Ob%6JiWR&z=Gi3-c$ezFNk3BB-Oe(y&y1`}Ic*pHU9bC#RNq9x zDH{(Cdj!G$pRQJ9y=l%f9LR7tFhudq{{SQPtuSd;Oqy{EM73-;(q7F**&u|{C4V$u#P^oHlsnx=QW^?!341;9J zgBqyM6`jcBa>LiJz4kh{m-u=WMvQSWYI52rN;<#R?9FwtF{> zJ}irko^vFsl)FS#m6_iRrxF5X(~j)0FiO~(UlWJrO~i-qVeHt@1y}kv*+6aoasAJ5mq)o( zXSK7`pF`5tORs#i_+}Nlb=9Uld5#eVsUc7NF`!*cu?U{rrzo{>m>n??8J>z*E7t^U zJx=OMz3T#Lr_t{dnZ6%L)rLC`ol^aIgWbK^?6m`O>iG1>Co4Q4ks{0)<(p z{_w_R-5Vtr>1&Gu`)JA(A+H~Yqqhz>ETG{+4^-X1d5Gd>?}&D}j{Ht*y0P31RK2bi|SjgSF;1o+SxK5`e}m{E;{r zLqzfW`9JP9B3wZ<`2*ihyPXf~6tM!gA2ze2o8L-$>m1X|obD$YmS}cYxhr->eI52) zuKWFjZmVD&WfCx$vX(UTMkbk604N;LUcb8nU+94D{fzx${$LU_b;=f33uyfEGFmf+ zPJp~wJ2&EP2aC2(OSh*lCa$f6OggWt#?}cvR}u>+GnaiDdd>^xIM)(XS89t?J0oPs zHjcyD%_+ii6%+mol_Hpi5h%$Y$Y=OsNMeJ0j?Alz-w4Zpw)LZ7o%`-9r^s7$l)*8E z%%InF@{|ghKJGBzr~Yo?y1)F6ma7^GV#qK2!D0JM`GP`uoBy>m;-r4!F4h4l+iNH; zlCwTXBJ`b3y<+I|XYKI~m8bQ>_TwEbG!D-c&+$Vk=3&rWwP)((jVPfuV}U#`N>9;a zEvq!B_|qX|^iBG^;3Z?FAkQGXbqZ-G;VF*(XqP^$oxn+b#hYsLb)TXzuKamUWquf$ z5Fk~)d?asYTojy3lUAwZ4cKkB3TmA@=o(QzHr|(EA3XoNEIu!r*E{dlc9`XuNAQq( zwTi=7-$<2IS+iLUz1@VW#T2(%FrPr=Vtc6-F9&!3&89RRap)o)?iO|bCD&@Qig;|v z`RtV4(7iA33gCpYg*Z*fvdizNNkU={A2XIF50<7+alWRf7vzKd(plVqHvzSU zqq^&_W zhg?ppYLHolQuWu!PW1Q1822N%6+XrYNl@Gb<^I!h(qxvGQa*dcqA;F{x{BRnM4aEb zuK`$u0q^R<`H>Q=Q^`VzvP)JfnY=xM6AT;|asX6g#p1A9?)fn;8XXyO!1@M25eh(w zs%%xM(L*PgKD^dUkj73EfYvn$Ap6py{cW`htdz#wniC;WMQe2}>dP`-?IRs0a?yV2 zqQpnhS*Y5~q8=Z5)B*X}h4~aUXNYEXno;%XvXC^PdDZHTiqxyp1yJ+U__SK~4j&t! zwLJJmc)E>^b8Tb@HSkT%Q`W)Q;b)&desh1-d0DQ$cd%;HNw~OfFn>X715XsN|2}f3 zM?s55C*(aB+n$cfBItt`F^5~tnpGC9i^+x}QZ|BUCzQt!t_8MLWV0Q4$SL0t`07)g z`;tu@Y621M$?yxgt%_(H^Xj|beCN(n#jl4p1NI2L8ioXQMM#&j5r_VhylFgf>&}7{ z^Em#to(38!-|;5eSx5j~$LTp@mmplX0=KE3nr2uNvP0D~{I36(DI>((m0cD?k%jIl|Ds zr+*%1RfB5CANBDD4hFZjJ)OQwt9M7ha@z?fd`$x^f34pBZ5Yhxd#*7a{zbR^v{iXu zFqcY8UtMmDNH*m@^CLmr;%;BFc|H37D~}#!xRZv{lKtrSJULxfksa3A4u0!Uwe(yY z@D@f35PoFfV7)xW7tmoN;z^wb2k-RwaaFFKlFmGKekx=6$Zqurdsw~Es}D+lrJg>0 z@66H|V30);a0!*%ENGqJdAys}`NBZjj(isQtI@UwiaLL|<$9+!92fVZKLYXbm+JEy}sbTDS&E1-Eqd{|`$VGC@@yL4#FB-Q57HdmQZu2rS zEL3oENLJZCNL4pq0tzd7uo|wJkMTZ`d)pji(3UO|gvrEQg^L6(S_6l7iIx zhuCw`;{feE~I6>xcAJIdF`XZvTN^>WVqfTtk?$8=$ah9wp5&D^%r}- z<+3#%`Df{a>YI5j(S&EXplMF)6j1zxz_Z_=J{pKeXk%dulB zH%0KjhXtiv%}2Oacw%4T z{uf*C8P?>wd=DptDkUfi(h?9gD!q3IBBG*zvK0{sxT#8&KnM^ZfQm|0M3fc~6%|P6 zy^0b*dhf;136O*Uq2`7CKj(MOxvux~^L)vZduGj=S+nkgUaKu}y6n}YgByz_N=o0Y zt7fxz9a>^nJ6uuUQ+P@&UX@Z?@V*Ffl{ur#y@h7deVmQUr?~-{d$Qn+1}S^KQ)-dAiw&1ZzF`pZ0je zz)?acQ*CRkMKqRGrzBnU8}~9wa4Y!I4jEonc>G8zRl*bB0Y6?Z0ARhwuhVAd+ilc% zkyB@J+V(q-Y%TUKcEavp)-G0kUvm%hW93M)4-oAd21>Bp(+t7#hB51(>KV)od6rE4 zIM0!B@8@7j*CC%|>S8%@p(G6Lk)qP{yg&+^JJ!HUJApeslTZ&{0T=@=Vr}?fUK+^M z%;TC1bg)9Wytg3~01JDxP#VTpc#KB6`zq2w=PilU?&;x;b#-AcK8WkhoPKXX90EKd z2hQB@w+xoDj5X1roY9sLwI_w{iv$G^;6p^s0{LE)jmCSf;Jce%x5tV&WUL_7svNQK zCiZeq^%!a|6gSb!5xm``UaNQz@thRTlN#hmskrNvh%|p=(1oKLLwvUXvg0$D{d1It zrZ1FA-?@g1EeVU1s4I4uFF8IgF*5CfN9H;x;k_TL+IM^?8~9w|EXT%BH(s#4;_pI2 zKmKezhU$SYXQ_ygq|6+VpJF~diJB-VaISkRYCXuWmLB7+*GoE=A1ZbgcRPG{ z?k*A-{+Z%c8?ZNz8}D#`-y9M!{gf0h%krkk3UyH6pK6X5a9#Oyo?~7zXqN5xW-JEE zWDtqmXYo=V(kr^|vP=kl_C7o`w+nM8L`IJ$EgfU$;j7pC4i;ltc`FM#`midk0t15} zD@YFggMwAbwblv!L~SdVns4(5f4>XfMbC9BCVw#7_J4QY&=(ncY1Fh^Eq(5n*hkq; z%?H76=9+Q4Cs$D4J@-~~y}A>d-0<1Kvn>#hfIB-$Z!9NmgvQsWdi{F$)3F`%V=J#} z=aSu;vR@AqS~ZrtDkq~{l*)xv)NLYWA0^59?)wp1)PgF0@2iN)Ok#YXD+pn)trsY8 zW!$GhJ+8I)TQ|c#`GEG{lH@spMH7t5mR6mbQNV z0QaMb_VkN8M8i51_8XkFBS%WqMZlf~EWxG#oNDcUo50~gdu_tQ3Q+R?Gc3sns@n?6 zl0AI5_0HN;cyIU9XB4m=Yx{_KmxID~0;J07b8Lha?rNgVHoIf0@5hZKT$1_EN z8{x||lch%%;F)U+2gqO0sYLmpYL1P_) zSvnH713H5yGL+BzUYgSoi5}=Y$|AL{61(=i&YeT0?&^}cp-$weuY2e)-QUh1T?8m8 z0e>3ghPV=V^97!JltjhFIP{VwIBY_X2JnODqsizoq_2G zyiktsZRgXAi{MVB&-h3XCD=FjAZSJPgNM6Ds_?vk?QG$|YhU0gNY(DCbuy`b$yk*^Lq zT`kJ&XXYW1qT_v*8ymNBCs*273wvuY!Jjzb+7Gl!l;@A2rHM}@@Q(_;!VKrjcA_N< z95!)yY`I#S5q8B6d(G?=V?ugurt-TE%YU}x?RXvIoKspLCx>8?DKI~eDN5Ow?!Zh5 z1&s$NL9=CR?>1B+5dC%SGbq*ou8u(GtCEMN+f6PY{q1t0Y=>_o^K4?Ca1ZxX^Yx!4GdX--A z?&u_hX}g)WhME6r*rnPy)VoCu-+u10@^~zIcaPt;uh>J1q)> zqjICy1p~U<9}%gt;}`AH12hTk3ub4>U(Ubm8|7}##ntpp610N)TM#?hvTENd?{vRU zUUz@zupZ|SJ07%XVy8v!DN5hSeJCAzu#St956&+`oQbLpDNk}X{Jc6(t}b28)!ggK zaf{K~I>+%X$bEg%^RmZ~(~59|%`&8_vj$fb(C{Nz%lhJg+Qb=A-TB2hiOzz)eft1L zLcp{CUfEPOIqGXT#|*wf{rf}$T%r~4* z4_XEA{Fw$$M1&sKpgfbuJk;3Mh>vv_dxd3TiqB7vuPWdQSk=7wfNa%IqRnFASWvuR zb9^t1AL}mmVx&Qo0w`uv@)q2_c_N>Y6tbA~Jd3j&7!7cY7uTf-mc!?;r{Dy@+phBg z0LhM)uevL549QHkjdPHhTLc=5&VrR*T6bB-o+C^8GptF5`Bd~l^PrGVi~2iX9)%Mc z&cUq$*GxlquE=9yrdRjp-LqhU3b6nQ?MKlc(o`1cQ6b$RQZW9ms2G;MrX12F;%y28 zi1{pnYA=Ij19s<%GSR>GV?WJ5bJ{&}%2PbN$!B1G%=AXe@82mF$oQK}uja0-2jO#MTyCi#19ZAQ z=Dl#nRcYU)uW3)b{8HHCn%etvyCNyS-Ai7_w7BEt(#)0ZG`Y(Emk#f3-VOcp@aC}b zA(2F`){X{?L{3e&M;$-nTg1?BJ7X@mrmZ{d)>?HLUusc9HXEy<%w2HG^XD~pS0UXt zWpdRorV$#u(Q`*)wV~7KdNom zkwbj=b8-K36t5l~#U`Hq6aTiPnaXgjZ!C#?45mYdbj>;BG`8}=Dm3p^eN4z<<9)J8>tOX73Gu z;{{@iN4@HTF|E;tJYb!p(Ln6Zj9E6ihb!F?=+63>TF!#~w--QHn1oju5WG>6IEHpL z4jM@Y?O@OBe-RCwIsG$IE{0oI*n-q!Q*RtxKusLf=uRw9O}~R=%kQ}T0*1gsNsG#v zzQ2IiudaNN&d{}Ld!70Utui+wNL9>kE7-Xac_zJRY{upvF=VJ}y_%zRp*XEja-e2o z2^sHx+pSW_(1+id*k^$Ywg(RZb%w&i)0!E>i9ZI0TRINByPVt(mg^4O$b!WdT^q6D ztq1v5ELLw{H^`yYsLw?`PX`=;27mM5&pAE}8onU^f>C`ACdKh-Im-=Fv2~(DO>{u) z!+iAe4ed5fTRY~?U|c8P9KQ-jtWc#~GHi4wH+RSNsYhjNT;d-Gh>CpL{QYn8OKRY@ z=?*vP?bTfU#S4g$9}d~Ok*-%Ato2)+hf80r(dp>wwPKnaS5LQG&WiRn!mn`fKzR4E z&L`Tq@dpd{bfL3Ds(c10ijjO^H)BuBdV9&TE^v&aDpYlAEa?Mr?(6R54ULT<1JnnB zIOVwZUv?wGmW6NMMA!_5dvpZ!NZ~ z40-L0h&1RDK()(_I|FdPobe}$qy?!Mn z_1#FXy!}c0)>Sj4LCwtddyD@lRs3Vv3k?T++$_>d+SGh=qiqX%-|LpurWNUvpRNxS zV(4khR@EOJ2CL(ziZb<}t+h|)D!!Co%@m^E+l=+Eb|CANNt@eNuLy3PVabM8HHM`5 zPXshp%&+%45<5+J5j@`rc=er!>cwt+5CnE%q%i4q*JOl9<;17Dj*USLl@2`Y=+F^O zjKso2X|PZwx+%dmx<+9Tw=rOoGd3jD8ucZR7Bv@Te{jyym-4PBVSc4o*-)E(Ixqo2l z>AI}HG~N0^ek%mag_U;e3I}dkFkT&04ey_^ZhbP(&0n{!U8lSd{KU{49~)G#s#szt z5jnK}Y}dDRVqa2w#i;g>>bsB-$Dy%sc?$;fvi!rb-H9#8)(c&`cy9Tvl%jDnp`d;z z8vehHxDGMruq>QV4YbSquVuI3=1qHfbUhzDk>frf6*7!@w@y0bh2EP$lYd10A6fL- zz%y_iI1rW`8M>h}1-chxPlTFJ#XPJw!a0EGhpA8?)##a?@Z+*bpN{i@aK>i@v;q%%u1>!a{GTz+%O9)I{h=Cd(qLAqw zQggQ`MT_;pKcvpT;_v_t!6vTZ$@Q${bl=zaInBddL#AGt6ky4J(U z6kyx8Fg-tw-e7MSFvFAsKNGZfWjwIVAVW zBZ7Oz+pGQf!o&xQEZc7&KKHbQC@Tl!57gE+p8Rit_|M?W zId1Yz$iL4f&j1{}u6ZAO3_>;vz=j`4uOe}~V*ue;_tc2cO_xx$iG>pDAB~k9zV*_^ zQ%7QTIJ6CrqJQ5SD8=%STdTnI84TNa2>QlnN(dyB` zt-SiPsBPfcJ+EW-4WnIfZ5<7kmMAX(8$1cL9;v@K4M+|aF|BA`2*FuqrP zMf)~oeNGvBmDQb9B}D`q--;EUaf_TXh0ngc3DcVNP$*rr8wY%bi)p=-e_$}e@J!m; zR2QkdrxyW_XqbzPjY$b`2stfUS!!)8tl5`kj0AYg&)$6Jk0x>6e2!+BE-`uT`}>LO z-v47J`n<&lxyHm{u8ToZ50=U2Qy?|sbpsU=oe!L)ZH>FC)C%ebip9^mhY1;1`_^w3 z7W5lQu@v+t1ukO6UEDO*H2|Mcp)kY2&q>uY=SWQF&PhydtOd16)JseaYKG!8W6CoQ z3mir~^ANIa5{A6?&&PG%blW-}hyGbU7w<1GDO#OW>A$75ShF0z9=`mMDj2PAoY5IO z&+khx$eug(Kr*Cfsnts^*|2r4;Oi}wA2wc&VkF|2O5`)VU(=1Zag#bI@vKwD0{hM7 zq)45a+}uE7LXO_sjy}l>R?f0va{6m>Os|wq!HO<={czAV~>1m z)Gcdc0%lg4$n8>9TJ*ss2 z#r0XoxsmTx|JlXwI@-nO@VH_b`ftEq-4x-x25{8mWA!FKtXkN^9W3ISvNd(fk#jGO z=D=w4EiP5?A4nh28O!2z3S@rN%>hP3CNz98g}r!Q43!5C;$d;u9Rqm-IVYUw;3tAa z@4pi4%Fw0mEEAw!B1e7BeZVWtH(fwZN-s=XAKRhlcXH}U$K3k$dN!;Q{}0o8uh895 z2O>0|70uZv0hUPCaotxxS&L#nACVRYKcz!Ap|?o+X`U(h2HnZ)+{s=6)YEHIxC970 zQfVMz)<;fv=GV)6vNLY-E6NqJN;G?XY7bXQFXj4CtGL_keQK2i%N%IqN zHD%mG`)5YU!_!An2zFhqt+CbR&~^1tBze=hs4592t@kUT8-|dU`7gm2xsLsvp*Pd7 z_QC%(cz&x+8HBI{0h~Yf%z60%zT22D%;P|w3qp3gBZINyp?cE`V~s%o13<2=APw+H z3>l12=?9Gi1oae`cpW;_JG4!&)?bAaI6@>i^su%*Q8Tp3v!T0Xq1vrTIFO13FO6;0 zZi$;}V5CDA453T3@BAZ&o^#^=W+%@lYoxyp?jbS#m3Du|jNkl|(LT9CmcE5}K5Nn4a>aHVhShT7Y z+e;4JOd8+BeOU`$3nG?0Ps5YJ2UB478X*IlWio;aExChxb-rqx1bD2?ZYyL18)|cx zv=_y$?l>iBWKSb5t9i-$b-vYV7p7^cbO;$gx$!wG+KCtgK2pPSBxU;8tq$|xXn7ngP7yv!R2`IY=X{h@rJQe^g0bC z(?8jzHwt{JEf0e7$_C|J*KTt9Rwm^S9;l{&9#N}nham?GnqxmC5kfX{BJ-T$ssY4l z)JDH$oY2lt|7GL}yY(Is?@*6Hd^G}SUXbo${NR=s!=CYe77_+y01dl(toKFIv>KqaX)f&Xd2_qZ z;c6n#V3t3w%I2yfv_3E5v!yIA{Dr{~3D-&`e}GZzucb@!KVjRCO+x5Mh*o7nb~^xF zr~!VfG)Zn7a+Du{m@TU`yVz+M=M><1eew*_H&TdjKn*_jx?W*-yXVr zSc2YYykoF48yRjE30ny{i_JFh7 zhv_)7z5$T2;F;i_gmjg98!r*k6Gsw*oRG=L$uNSVv)HC;CCqn2x!dvT91a%Z z)hq_$mTI7nG2#c8FG5OQ%e2myrgeI}Hrl=^AJ@^6X*c(q=NPHr$I*!9z!fXr?0kPY z&o0k8JMZCqE8rL0J_GO9j=DnH=^l9(3E>UmI{B{fm;2??G~P22Iua@|M|B zI;YQO6iJP3aJ?T_ZQxcw4HtvPv*UohOIJG|Ug4W6uYG|}>46iTc)+Qocujb5s%n67&PVHjK2O zmI!TXa_785wRYJ%l<4=W(ZHb3gA*V4qZ<)F5IgFfkxh1wV?T(_Jw)~Bdh;0|`^4t@ z9EkYxKd%fU!B7bDc#V%6C~huWB{IWr;oM6X{oP@QRcnEP7@?UX&JDGg&EWjh0%PkbI!c7Na`TiUSS(sA@)qnzh*Xt)>zJUbn?$BOCU7be6B_ z`UsoobYCOsU&8;>&<*azZd`CQb57}b;LOdsT)feNDtt)>Rk8HRI_g(n$;2apUa0pg zPQBD07QwG$d}3n{Aw7Ipa!|Y>*4y;UkwHYGUeyHJ;1^&1ygqZgk!u#2h(AX@#c^i1 zdFrzOt|aKE6hsmA0$4VD#`FpbQN3EZ%pSjL5gf#|=q~TKw(7Ugi*_V@wiS3|%Z#Zq zEB3kKJ+=caMscdmF;q&IP-*4}y@N%I^wOEcwZK2MO{wMEb1C5My31SAwuecwO?vR^ zUO#eQ-KCmNa*|wE44J!3`;fI%krZDVVLRSRxjI=gSS|ae|B05#0d2}>z@a=VB6qgo&IgyU+S*3bzBNgdr7RsEq0BsX%^d=0tqca!dSq2Xi` zgwNE_L}|r9#I8bNfm135(N}bS40vR5>(%Pm9W=J;0f5s#PWs8J9={lwuG;gL=9T%+9+4sG#`vyY`*B4 z_dkYa-p)s6ot?SV2R;ApFo^+p0brdy4j)Z+LGQT$g)Mt9gANZlT4d+&2Jh$%BD;-* zVk!Y#7_7-cm_KA_nOb~7~(VXL+MY3o7uwHlfA_o2c{|u_x=Hy z8ukZvRe6tOo$KaGKD58U6|VSooX_cGcg`8UuO`pIE!xsim}XmKV@&y5D)F1!{Iw3g z3?+pRD?|6(xAq?JXKnUkG7V$1W(CU!d_>4p<6BR;5!@57yB)J^Nfl|8k1}` z`tNI~CG{h7$8=OW6G`%?r~+fsTB{)@AH*G+KwvEkyj*STMk2m{dZ92y2)d%$vGgi! z*#&5}K}}hgPWmIg;AVj9&x0JW6`pq3c?HvY&GWJHj_>SiYh1HbmzOW?8RMIwP-{Eb zK-cfeEE{owA`zh|!L-}^WfxsL40MZioped)9LBtC#9Mky-%UYlG-Y-A1wQO3s^_+2 z{7{4z5$A9OxBMtFWCAf$6cq=WRAv<=*Oj)oc%F7M-B7Qd=9N-?M{9XS2vNiDT*BdX z&0HgGT0B-4j4MA%tz^q<^~^gwTPW>a5Y6te{7+q-ZzuAe;G%7t zU74Qr<~J2nYgd?lP5ZYb`pkm|D*xqQW)6b93p)u(2Z3gQ#{tS8nE=-G9q>8q8#Car zaD&wxP&{}lf`f9fPFy`!e+qC*uZw%A$GQ%(qg(D~$#aE;2VjF0O!b=hy7fUJgU!6H zQ5SGInD!ca!J8DL9y=b5^7)#xryfF5zd8y4sVr^tyf%|~@}TnKAC*NTM^9uP zFMyZ)xPuqc9bo z_uiUcdfCAi=wo^iPCY(X-MT522Od$NgH+CN-DOzO&}rRtanf<}i-X>{T7{v_QVpSrXP>zk3t4P5+Mf=^m0! zGTUAIt=<@nI{rxEX>ES$zkDx<{whxMLQhT=v)6f^PrdYl6N<%QDYp&|SxM z-`uJpzbvKWYMR5$Obvu9ieY(^iPOh%bGHIt4D6hP+Hx&4Ul-gfpU>NjIt>SCRhLlME@VN+(QW|J`>kx(b4n>vl!O>Rze34X8&q zvtRzlRDJU$NLab~EITQ0f_K`088X_h5=7D zyA!i~?B515f{<3h!Jk0zXMtEwuWO)dA+!V>Pd=NOQdi@8`d0OAD3qx`qij;ovk$-k z5R{`^qK!4hyH(*l`-DNhQRCH|-nGK@GnZ$%?hcYE{FSjt)ObxV2sL z{4;QDJpGE?pj_}-D2x~vLeILrn*2$AXRdPkz57z7T70A5)hgQJWvkceBA`KYl^KE# z?AXDs_`)y7HMYUfK<~U9{SM8yn9HfkKWZ8g;^8CRUi*V`6Ld)}b%Mr?u_Vk8({xEs$(gf4f<1y&JL!qJOKz4bLhYJc4;-_FW=9MOP)BdOud=wi z`=R9V$s%VXKPGkWBFu8)%cZ-It*-Z9k-hctJu4flRJ7~aE9U%HJ}O3sPl4;dGNj*X zoOl6@`RngoC82`zh|hgWmJ>Mr*($sJ4PDPpp|>C8k#ick_>s<#^qMM5!qNRutDPe? z(2eiC!P|YeAle!mvdAkd&tmlq?Uvw%X^cCj@#4JnW|>f+$VO`wZvDMU0WZPTNIC71 z<%+>1Sxk4bL^v(JMRl@Yy1Qn1Ndk)Kn^o8AmU;iplur4A2<#H1J@((8-S&GK^49Xp zm3hX%yVdF2L7IJ1EpaZ9%ftCdwOseD1?6DIDx~Jm@Up>n9dn1ke6V5)Wn(Q#_yn2# zSS05Oog3IoP?M;thfH*tPA(;m8ArxUbvLL#l5f~VZ@n`2I#2*TK*GP+x?8tL+paUH z?v7G_T8{7*3`TuN=oE-D6~EWfMQ#Uk-r6TdM7g)z{wS?8ZCus;TC2I*tOZGzF&hlA zs9IZW-|DV+wO3IPIn7)BRzx8^G1^4j7U{`1ZKI_%;kWODgGrz(tnHVeywU7eL0Mr* zY3w&4H!ra8Vxz-Y-~!xqA~Y1E%M0cKG>!QU+SVcS!xHOp=cNSu;T*mrk_-K5g5*|o z&tU`+?CgQwKc$~yy_Z-nu*%*4%8@!89ERUbXKbiw+Xw z9zT4SE1lwhYZx^j@K{)aR0~T4hSY-4*ek7|^D_pM%p|}|tZh9foOK((f8G*ZB6S>n zB@RJgy#l-nSUR$ikiqgZ=xJ*hm#~%t!VB`zRYF1w+!C2Weu4cb9-v=y*!kX^FX;xH zkgt-{W{G2$t6-<)3Eeqf_tUGSYX zAtq#DB|^G=FJSfy@XrB|$7lFXD+O;B~Ey%Hp)(ZF?)2n-rLv={6`J&63u4$<-HG+5Of#sac zJB9Qhd5+4@O9^0Gj+zT*TfWYwe;-WML3PoNW1f6}%F2Y_M|KGXb5|GW0-YRo8q{Eq zUBm*S2j@=JYWHQzt+P)KXJtIAFz|`Dt;&yZyaPS3MZQO7zz6^}(TAZOd_ZHAC+xEi z{jprOUgcjCOqg`z#IXMOvCV0-5Wn6y&EwiqDa(=8`PLF_yqll&8d7ORNZP{h5jFAO z#`BEsrZ=@#m&y}v6+W8Xjj0W=*69X~Hs4!F<}2YAe%ly1Ir3B3i$?mandQS`Hd3iJ zB{b6r(@hYN-KK=e&(8Xn%hla8WCTkd3;EgohAk=WKdI$yPYXj$B``r4i&Tu#X^?@O!bkx%tdB+28OkdC*ZVN~& zIWCIqlC0c$_0%r-Igs$w zBXb9ceYbXgsIM{cGdX**Bw|vYK3+5OW9xcW*uNr7!<1L$YEDtXcO>StiFi*pGQhMo zfF&IE`P}okf4zbpepLTw{Ick^uBX6yPVWt<2^PTZnz00f3VL4xVs>nr~(@B)9MF0kmbj$y5qHm+1@g#iJQQzG8n8tWLYYd5dLC9o@Fn}Le< z0)h3iQ&^DqwEB+k3w))Sm;IMzEbeF~#$C;z#)VcdRc(&Ag|g+cp!R0%4rW`;I$e8S zFFTG$JnAGa+lc;?IfQb;vbbA?3?uD)03k7waPl2G!~dJ2bEw=4b2z_idmXKx1$KuN zvj*0o)E{%^#GV#lpb1+le1H_)E63$KV$>t8%|WTx7I&YbJMV zekIj5HnH7ke>9|+P9^C7GTBc>44)TX4n^mB%pf@#Z%jhHJH>8}=5WjNRf$r4^VwkI zS^C?3R&Z{QKxE&*++%u{K4~yBcNdMkV8>iH58FGjv-%<8E<&?fD@Pu};H!)ZU5VqH z=-w2U4S7nS+z9q@1!D>G$zDUdE+k z2A?x7oU5ALT{8|{`~4TH<39n&$E%{JxmLHf?3q55ytcc(^CVpFyaYO2#Nod1zwwR0 zF4Gidd^Ft_be(>Qs&9ktcSG}@?e9)()v}obpfcq^s}uFd>LtrLyC`-c>7Pn7INkw! zWtwI=&Ni><@iz;DX0Gcq*GIO2 z_JBYDzsvzp0Cg!a=0XVe+v;m|Z;ty~igz z>qUnxD8x1)iCnLa<*I}uuu>V%a^Y9$D`!~G_84CmRtfPR^-y~B52D)~<1Nmw<{y#0 z269jIi}9LWPRp|HiVSrRYn{K2`5UWW*P?%eCaljY7_Er7)oypJUf3|&ibaM)G=DV_ zwWnS8xNQmv>}lcx@^5#)Dw|Ia^6Z}86>ky@*gg>TEPu$B*_Ck22k@HuFcnQ{;INDb zWbmfWWydHs88`6LPXCUU_LW4#Ia+(U%#+8KRNlo84_$>dT_q6$_wPXuVnF!rsK)z} z+X&s5B$lwII*V-6JlnTR#!&Fcl+U@!cuIU;m%i}XKmoXbyT(}f?Lk}r@xc18yTcFa zEHmvej&329YwBs5+9J4K(vcqJ)L(aT>fBu)#BYV0gx?OoO#Y2O2j(AccWf&qwMsjz zKt6e?Z5&CXO0o+?#T%4R*h%JIC(~vnPrUsb>v*v^WcXt3(D#hQ&5iS*^^24JFl>A6 zP;4QKG|Txp;~nfWy8Q4H+zTEVTApYzx>`z7Ek`TEilpvHJX?7TuKHdargpuKMLUEP zHD0vrenNhA8HskO>S#Tde92<6TX4yKE&Oqb(dH$_SIBI81Z#Y<^1vi(&q716kI-L6 zgMOtS!QVl_T)~xhvz2Wt<17P-Ih)?R;yVeQYqaSo#2&tSje!*JtS#cbKF*dr5_TWw zf}F#Wf92e2pJ%O0oxNN46jq#whSugx4zJBzCaRXoeBCl*(#>-C^g0^( zFlXze;sQaMjojG%vu1(Y>1&M;Do$f7ShS%aMyDDK>=2!u?2N#`GdL#g%jnSv zao$6-D#1k_S*_XooBOlSrS;3(^E>~lHlx`Qk9kjLi{xloMpj@Am&R%mQ0dsI>u#^7 zUHaTJ8ORy@`KQ*$gH;uF`%Y*?duMCvV?8!!d-_Q>ag9WeJC$uHxyBNY{y_cX>?-3sZ0EFHDd6*(WIj_^vG++_khV zp~F4H%&}@U(`()!6%WdY%QrPpq)nq^kV$TjZXx z$I4x>a~E%2G3cqEfNw5Pf@svbR*h)riBLD4Y1p=YzmZHkDNah39R5?f{Rv|Lc#%!LveVg5R9mOg(*HRdIVdJ-2H%OxZhcShpl-jFpb z!AJ~H7mCw+7^ZWfjjt<2ij1q zg6z8#VtLa>6@Rg_va)dqJ3=++XHwQg0{qkRPPCpeA~4a; z%A@6AsIqY^hF5#W;+;GFDvZ`81g)>_WgGv%L{I(x;IT{4&{7+`V^(WuMIU}^FKsLC zxh`Mjnns<4)b_kCTpCbYolXepk1`%4rKD4H%nEHiqRdqI)b^`Z;iC`vi0O;nejZa{ zs>SUwCC{dkP;_l;3w968^z|PRl{V6LE<1EvL_nWx#d-d`_H+!bGER5NI39+N{*; z%(6Mcck%XL#Z&^gm*B^v{Urb5{-0Rw)%&Hg*KmaMp_LDg zjRtp(|NjP`3I}{^;k}dGRJ;}e3Zpn2mZJH$pMU-rSe6&60#X6*0Dl7@A=WQIyjX|u zaKRp|0Q2+N3=VTvcMSl`d8i;I#2UTzQyJTT25SGbG-GJZ(+m!}7IwVVsu)xVP=$+d z`w|-NK>68pGx!xu^vtlJU)V~l1RZdjLDNzM^iQ)h(033a0aT|lDSj3P<;V~;%9y~< z!pbTJ6s2(e6%P<>Z6yV^s0RKJ0+^?DpKxcr`VH#ME(ZFWB}OMRCbWvT`ItR0`F+Og zBpKgGGrx(D-Ie)m;QR6~2uzFe|KXsRt$FQxt#q%VPXHBDrH|Dgs1MV z$o<)kHs}D@9waemvJx zPr)R1W=MIAUAUt9Bxe0m4#a_H2jcsPeBAzQj^bMzRCbn;ehE%z$-*?hN#oqL$2=Sv zz26FbO`Axj^rYCudh`p}UcAtn{@yfru>3lEpr^VXV%Qvoe7og4RjCt_LvXx#FG*{W ze54vJ1ce!d@c==Uf7h|_kh^15SDH49h0=sc|2UYkZBd##SXqbL8wACa?nyU(DM_39 zxE*KtEo&{0ZDE>#YJzFsxmR`FcRK9CO8eCrCky&jJ)}6Wx~y)0MY>rEPSzUf)Y~No z!lj38+DEfLdUJ&tVG9Jf7f$)e93_U}Q&ZQqi=?z9z%@C9DP0lYmpGP} z3=3F5_b(|NX&MC1>xM%5VoIxhB@2uv6BB2an`4VliNwTT<50QSZ5oveHK=K zPd9iCuAJ`=8DQ~E;#n;Ss48K5Pt6MD8lY+->`dwqmNy_yi&z=*xlz~1OfAOWVF8t- zpmRQ>8^p-nNy~jRg$WTj^ruVZL+Ao$J#G}S)_Jdg>yRp8x=%*Kyp*yh?{=?}p5cj| za?2R;e^s8{^-kUJb-;YE(sMsD0rzuFfM4fjo{~KC@unC0D2)o zLOWr((rrqy{+*J9NvvuHJj_U&9dHc$1CWZfR0;F0zf{j%&tV791iZQaMyX>-_Vq=* zRO$^&T8~f&Ru2H#p`BHRRG&|*#tN7PLEW(#mbol3fY|Wd^uy}=3r8V~E7mIgi<}0x z%om5L3>`r6c}x(36!8Cw@Tt#T_1$c+#3)WS}T5dl|cY06oA^=zZ$%rdLP3Zs`SfnR{1ln?MKSAQ6lkP|$^hoBDBN zqhSQaEliUHaEza{hNYjd2M5TG(~K@da30lypE$%ZqXwZ4v)r7a#&f1qB)?vR?$teM zK|AhAWKD5HS{!*Ass_)9RGlS#rs?KKcscaqnAlo_moJ`)(HCea>`8W15q(ae#5zQ~YGd8=skTWy?#n zV-Ld8v(p0~vn@xr_>cPy_N1UrKILyqF!nac@2|dY-TQfS;5=GzMk*l^!*vh<8tS0z_*V*2Ha6HA3`T2@(8hF)kUZbxo9f$+hHaEL&1g~<*s zbfg+8v~}o}KGbu>JB56zdySJlhT%H=I<#H=AC0mM*R3QB`sB}_Rd8NmYOV*Fe5c9p zFo5FtO1I-GSq0mx`nuijGpE1@_ATcCR|#~J?2&}_Uh(ms2sK*W@1^aX!MfU!E>$Gl z!%*IRG)Jc2#4fQnd*Sz}oDOS7=bMBTX)~7{g}@UZhS+oqj?;sUi8}9B7KrG~P~Q@H z(F`Q`S$Be=UUj#4isn`nGi}bF9z@p_tLLMes5u&^D)FuKq-tCH+x{T9#iFydxB7F$ zG+W)0E>gNFIA(Mra@ zfOlPjI$T335m1{_$#%4>Na`pek<~Ja> zGq+ciM}0Uv#i(fNZo)Ap62wV;ODN_5%88o}c##c9&g zY$rCwx&4Lgiw*JM)mr0ZO?bH9{Ao(zMn!M2ErLXdr8y{9LXb09N z4f-F9cukd5vs?loNmy~q5YY%mP}nQiQ!~?{>%=Az`2WY&n+8JN#@*wy7=yuxvYR1F zlCA9PNRh-Sl}eUElr`&&b%si14MhoqN~P>e))_6fEMqCV8T;7B*k%~~Z})RQ&+pCu z?f1p^)pf4Vxz0J)x$5mL*dIn`NRIGZbXJqB;mCkA*@Z1wyeeQ;o5^u_@#uKuA9<8K z$~le4QHcgDZb}10NDe zY|f;iLVGI8D(^rS8%3>}XXGFp2cAkS50`jtw^@~m!PGr}CjGg;^ZHfef@Trl zfN;jqG7pgDVq1dE<~yr3In7g|`Buh)yNxowN`O4?@;(*m7{|u*nqPh9_4 zLH-M0s|t0egwCm}tl>l*$SaMyzvmzdJMr*C5jSWc;5xR(284`1zN2VZle&F4%{-biLrqUTsZjwLhN)1Gc~)2EPp zZwP03FAd|q98>7$elg%NC8D-SGE-sJw@V{}D?98+=&>C$L&gh7Zmv?WD>*ea(6Y9- z#8$C4hoELaQ2b;|bb=1a+j+)F*C&ndu0-xX2n|=LIVi{`M&%;5okSKLu;m-ZbxAj+ zrj(@JzvifvR2&%f9x=zprpJFwTIhX3tJ+8`LxCno&4=gk`fO;6!adgeX-bku~d5?k- zlLoll)=d8p3Rn*F=6{Zi4k!7h{FNU$z&i6>?SnfSMP~gocW2g^E)ByPdztVVXM9%_ zvI!B9hMumKT8aY)ti`m^|L_C38YD?@HfIgYMKrkYhsdx-)DzsZ{Wqnx5W-h}J-4_c z4Df2aY|si!tu&Cn6KR~3|*3F^-0Sl2n+@3JaAk)3B2PYZ2ecCpc7Z! zH?9J@w%8PQk#0@B?8~o`4E@Tnn(9ArUP1df?YAb!F8vYOtI&|Lai~^iZg6C!PuTjN zQ50eBUH?a(fw>Hb`?`@{X?uOkiI%%D>6;I%J~%E^G5bV1QrZ*O@0Ty~`lENZUxd2! znl4oeS@*?M!1Yw2KP6ayObkOzkSjy232Oe=%`;%ut1>~Hj59;#k-0WNZJ=z;D|DoB zB79Z1$Mbifmo0g0|5|~N{pMNCH63HbA48@Sj5*dAwOP)=KcS4#pBBd>+b zjdbbSj$6U_7gaBwEDtAt^^1Cw;JJSA!ZM4LzRiS10K1aIhD=`Xj`V7*(?+6sL7zS2 z5_;gZLge@G-KTMFgD*ZS(HE=wscUASzGFboCC`{MdM+jbIQX_~K}`j0JJD#5St=4n zZPYb3n%*jx-=D`DDvaj1iG0?GCRHrB4>!8srrNCAGE%oA)<<@A68{@??$mep3DsxA zG@fki$Q^it$#!lSiT@a^;uWy)obUobkuZGF<__Tne8uSquaS<$E@s`ne$Qy!+5y0O zf@1*q6eNcGjDn)T)@CSr^=y`%vHF%B9jp#Opln&E_JEFiFf&6Z1rleZ@G=Ps3&8=B zAauwG@B_dE$3x43t1^ZMeXxMADIiz+iSb1Jw&Y(v=y0$^-Na#QX=2E9FsIc=i6maR z5?MeCNG>lj>-{Xti1_`}2}~tI5+xd*eo0U~Zv$LlCqZ#6A7}1JZ^_9rmWOjZDM8KJ z`l70}PCrDQ%&92>i?d{xmox=8I36|Hi}z}RPRg+2z!-*jgN>c_leg~Sa~6)buzjwQ zr@guVCf!6?qlAhR_;=Sy?G^nFwN2iQZ+<8dUH#il9Vk2DWm zO4+<4i>WHh$4Rq*%)#rEcZgf+r6&C?%Ubqn188UB&ibd_x`;9sN6k=C(B-iaoc-%z za_wgKasHrFOzO+BHmvGJswKTSGV6tjT^ZI`!Oz2Iiypi+8eF8={WnPCg(FJb%U=dB zq`6qd1|=dc3AW$OXyZ1``5iI%WIv(o$19~WgQc%TuDgS*=2v8aq&wL`59dTc0b>{^ z?-K*2nnlrDb~tPLdCFNkK$D`vv|6xS5%z7>MOwl3yNIcMp24>iP}EFV(Dxo0hdI8) zpL(1TP(tY|4HYUmhJocNTF08uc$G{LyN8XF&G~P6uS+Y8&o3%otsOJo@uE zVITP?W)Sob-A{-OhC`^`#zdxZ%rfNnx`Cc@g$9Bnda#V}E+|)in>xH)x7PW`duG`4 zN&E}5of!-A3D@@o@%8`h0{HDo+6dpj#))n#Q-tcxq9NaDeac2lUn~5pEO;>DPt z!!6J{YXqI=oAv++1OTdXI<|r0aW`T&i=HBC&%nTEK?AiP8< z2k%H3%v?H8>TW>b?nZyns-=^u@6T`1K5Q|wKd$Yno1*BY_K6>VRI-04${CE@)mJ?p zisAGDWrD<-@S)4qZmo|SGVcT_b6*wJ>NeS*FDP=Ty~oi>k9uW)p-IU3gYDWl_eUQ# zA$HPgP$cuhP>Y7yBt&e`V!vJHYY>3qj=*UaAG_@*|-#Or2{f||SN;cM+B=IaA6b;q;rSb~a3VN=ITZP9iG0_@p z@O`-Z*iokol|Ph7&TgMWJt=nl13Ir)ecU951yya0sw39(kVxg2jM!}9KGqJ7c_a~)U~cGsF%FTj=yP<(})mSzXUWIR~zhr)_G8LclG{YM$p7c`~IM11odMA9VL8vXuM#bXD>M(QD2$5^clui`nw>m zv{0u` zx7hd-;`7b0J^jb<0UIRwd$`l_ED#NFN0^BLq<_>R+l-Rl0YuPuz7VMCrG}s;FLXh45p-2Gu;APiA)?n%Gs+vUy`o0m052d`jF=ar9eFLiI zc}VlvgOBE}D7Xbc(D(+_J3%rmk|et^-s~+Ja$Lj^Z??{XNofYI=g0?G1e^f&U1XZs zNF7faGTG_DuPotvW-OC0c44^#lv!15gfb_PuwE&uz)&(V4k=fI{I$)kRwL3LCjK!Q1PzO`ZE zu`(orW)u1V-YB&()L0Zgm%cnwR?xz3^+qo()T-T7c%*d!Tkg|*G zWRwZ!dViy^eA_NuOApI)h(OwhZI&^uFcY?u1lVc;wF z3(hWt?2S2e!9Sizf*BMmY_JNw)41!zE?l)>b%0zMrMA^lOYM9wZ}u zQuQRMVChc`LX4@$8WfFK7l{dh%t9UvX6$%Y*A%2B-jNMER-a^RSWdeUnhvcsHA2vO zUokx3MNX~@+d>UrAk6ycubBodq^eT+4ML8-Sow=ReD=D)K1QEr84T#5mpxq1rAEnc3Ayp9V8YbY%xQH+31-iWhENH8{ zIgee~3sgPIqDMKh$WRhM6!(WDSqFS&foYT<1W4m95T&~PEDcaXghQxCAttl9fP*%^ ze@lOt{~Lg(Cnkg@N`^{;%5yA)sY+PfH5AwB6R;DPv-J51>*Jg1wf&Yk@;I)=wE%3$ z2ipL=DD$c^3Tz*eTjcquol$b6#;+@wF==_lYB+D9hK*M#bb~Y6q7(H0gvR@Y9 z!<6@_dPb9rReLMceP-m|i%8$Dg&pQy!YQImbwBfa7hs1K;mm6UU7Ss1ophUgLV7-W zqlpJ#Bz>C^k4at5K*)t1G90oVbxR~&SCyME11C z@lpw|V6`IN_Ze3vlXY9ZTGg02?hFRHNowtvcyvnsjJArZiV{GedN_j=OD$M~X*%cq zIcm4re>bDxQvw4*r-W|Yl?Zxz%F`yJ*nWe%&b@4Bxz=Y8-|+MLfOd#y>~edW zRkmUM{FyERPq)aK8}8d(7l2l^M^^9d9i^Ge^Ec*1D-GOjXv}KZ8C_8hPs+$FkqL21 zTr6!+`1mxkbY2l-4c(v;De`xA<^_nq6<{3SR4?%Pb*X1rM;v?4f&WX+;8TK;HK zdZP51@Na)3`|m@y=;XR%01j42(9&iqS&&TrLJ32>nhoqeKUlxWA;6@ z*w1y$q?p|YHtQSQ8+lz>i?=pZ9|ft~=I@+w>-n#pa z0dqB{A;&g?*H)RUHLgn?!4n-a#79pKdr>m{cMJ~_Tjl*`|z+8KaU3-h{yLAk}388sm zMrZQ&V)ELfdihQ~k_1BE0s~L|19PSAKhOAm`BWi)=iKudC0BVsBK83G7k@gM&DcDvi$n=;K8kfR8_B`BWqZ$K)Wlli3H zaa^9o(_jODpp>3|Z)Pf4`xZmw7F{NOyvI35oYQd@A}+|l`lmcRfmDVC1QDY87+K-` z6n*+Su<_62H5}?*Lx-pwZK9FiNqxT^o;?;N(G;_r;AITh2<^yX%!ku*VuUfC?8`-@ z+=sDoF?QCR%nq}O%>6X>OfM@W@S~0I&zk}ew97gn1`xXnj%*UDo9GfXcL=~w&Yvml zuGfVqKrA`B$fe*bbRfWc5ts{0<>C^Trx(R1D2geTzpi`_Z}oBJH+*B~9=y-`b)7bI z1;X*iGG^V`EEn~pWXsYgq<5ktUCCG4+O>heS9b&u>7eByY5U~si{@n=U5(BOd%Ho_ zRn_h~3!O$Iyz+j9TY}%tQdN=tm8g?5`kn&|yn{TUXB-*C2Y>LMK^&PR^@MeR&I#yY&TzRZx~e84N`hTERit} zk8%L{MQyofEN~_1!sH$}?-n#L7R$N{{i4uO*)vzdVhv?fPZD(`?LS%U^u*$19R~Oe z%j*sER|>MTSUi4hZd%GxFH$bCB^qcMbs%OCvNSyC$6iH##KfZSSJn~i;7zaPoxQ4^ zWRIER`7+rA_#Na>M#oUa$#e@N#%B|U4TeOax#ta15l$L;fw*|uAicL>0l7Dhd^JW( zRl<;?Ge4W@XxTObSMu9kBY`JO@qEiC?+E6zcm@IW+lAJopC;}X{N+mR_xu+_u8slbxw&? z_pK98$5B`3Lm&K~N@WcO@Wk4}Fe&#?rmbfw;f)HgTl?V2BL;xuXqX)U907PQ%n%;t zPjQb~5N;IiM2V8la~n{h);3O;>hZNWahB4%m)w1Tnvax;k>GWuzxBcw!tKra1M3KI zZ2=Mo3Q!NwWK>@8ZE|zsxh1?=2s}Cx45*=VbLkGC6wn``SsW*m3fKS{cAiP%Jjd;~ z!w#>fxqZs(_V`IuBWr*sK(xFYa2@y_bT>6emQ4v=skiW1?#^Doo$&+f(@b}$zctY2 zFs`BdQE;z<(PP+)bH*8yK&SJZ>gC{3cL@f7I3=4;VXO>mzX`2BS!CaHIgIeEA@qOX zo5;y9*^Z>DME;P!z&?_#@admdx@YwOPm9<4g!*j>OzK6cG!apUh zOa!z1@J9%>4^5?LTut1C15&a@r-QI-!$H?pu?~%=w4Jo1kVEMwomQ*6!o9mWr*`&F z;)RRo0q~Zt;I*8npiT#fyX2t}-~)+RreEFgp&6}e$P$MXB#SKID|q;@uYmUXACX^H z$4NocfS$dg1ZCPUerde?hHk2@yp*UO1QJ&SGIrQ;YtOYnRp}-wcIW4@9*20qwtYqf z){mYUT*a0E8=Hq^cCD%D;^?hv#ZHL-(FC`0DbXG7CN-N8UFI0#$+uKKWn zUY=KH9TcBNJgF-_<1vIHSYUxBf$lc7kZw}G4$xsyS?K9nc68M78K5V^ zzA+xuOoR1k{Dm3j-d~odH&(VhH`>o`HXSTBG?Ooi++DD>9_mzJE5ve2t#3O9eefx; za6G);NZicTGvEIywC(ArS8R#h2*y^FI3==jDKlq5t}%8_qZ*9vPtZ>-a^F4}>i#88 zcSrU$j!kP79d^2szxSmp*_dh@v1gV+#^VHQ*4rIgP|ZW;bgoFy`u4NRt56Br%bcic zrl$OqS^HMCVdQB-%O6dG_d*x&J?yI+RBaM|zRUeR= zn=r?8S~WF`!wkXN3%wY+R;hh6w9{=_$W}88vaU z-92{AGv@REaxE*GsjUAT>z)tReLnNArDIWmXOBGodpdmMtcE7b4BT9R=jK8bHRM&L;5Hwrf|R~1({ZM`d&_X8S}#Hb?HzoYsmBykVazI>T7_fo$d z%x4dXFY^F^@wza-SPTlcgLy#-p5>1XBJd?OO(4GAZ(LHahhZs`K-@ch00a$ui-S<` z*w(~s6d&LM6?eWa%*AZW2^@$ELusRfVKX^=KOaiflK>%kBGTsnVDk&s@}#u67B~!fYM(@4=~3Im7^=BsG2Dcbo^t)+%%8@-E3N1ae$wC(~gPwjU4| zv;{IPhg5^kkb}A^P^~B%FRO^{)O^iU5yu2iQm_K~*9zu_*8gm0xf)51-774JTf`GI zeY6o-&3Q_AaeN=HMWO`5a{$W*gvPBkEc+!R?6qH5D)hu+6IdAh-ar(u>3F;-I&pT3 z<%I3>0$vCNhC{@dxAgp&x|0d^1_1XxNu@^pNdX7;akvA=VG1?dO4Q<4aH-WDciXw{q-oc;( z4gfHP6i_RT9|1}1Rds5|`>d~q-2S8{H~P6>!z=Ae;dnKUfWzUW^>R!;F?Jn9B3HZ9 zr`*=S+kW+GSFm2=Z)zY<4w`d|MsI`8BLtJv#sIi*xT3d^sAej6wq4F)~- zjOi#lFga;0)zh|L&`_cBatJY4G6j z41r5W+|Ts)R~~bf>JG?U^pvMG=a)!eTNa%nrw$dL$#-*R(9CA4*1PiCCbKFT0uaO2 z_Lv7++wV<6x@XHPbVBqokBc|`jq2UudRGk=34C)(HJ$vrNkmG3Z1w`EZD@XRuq;+2 z%>7Kz_&E3WWh=4MY2~f&4#ZT9MpAYy`PRnx|35?a{jT8z^m*ukA38evPRz;qP%TvG zR15wT+=xvFz2qofyjslMaQVV$0~!Myl?Vop7eWL6+|P+?I3Gw-l!grm;PQAw0AJmK zfK=%96Hq`g9m)iZf=XLBJn4FbcOL+{fMSq+Zerxdg`l2RkrPp4F0I%e*XrH1|A=UNFLYzF%#pQJ4lH%L``Hdjy%KLk z5|&yq=C8Z&Sei3;;=X^t+`H7h+#^Xn|6`Zge&A#x4Zo@3DO6kL<%q^^193nL_z&y6 zVM3xNC&p5S_zTHcV!vEr^X@x;0sIBiS;uf*&XNepuq=-z-SQK2F3C)_HgomDQWQw?5JQLz*cQ zjMCl?V>Sv)lk(Nl_GgR?*hk5I%1Zh>`)A=Kz)bf{(Lr4vq_V*A<~}?|9y){(p;{M7 z-D|=)C+?rhv)z9FS8z|%aNuz{GL0tO<&tai%&;yh$&j{vQ2U~MQu;zf-~$V{${V8% z7V}&4Yi7vsfVp_S`^c>&eIZ`K6T&Tr;)>TUWem8^${@FfGv$4MKY{d-)jEfCKI~2( z%hQS1iK&KWY@%Mql#}mKKsSfll&|iZ8d+Dx#FKmP6>`Uo+lb-$@M4v+s%K51yCrrZ z$A}b6bM|I_aHA4s^@(O}Hzh0$ojJoBb#eag!VF>tJ7j8YCFP!Svw9}QSt`_)8Nw#MF*Q_g_WuMz8TFLZZ2C4x=hsB`Hhfk>Kz;OuSU5`B$#eg z^>!BP20SAtKA}3JTHz*3VHaxeuU+>uS~Jgot!43*L3bJMXwCX$R*-t)J+q+qt3FdA zm;NHBpFLp=<-IegnH866oxIdP>L1@>8vE8|hB746qEd{m6;I^kcp2Ph^tLs@7Oq6z z7!9V>v!$@*Rhk!lf~B7ZQ;+y+i_7KN8wh>C-s7D zxplVJf#>mu_n^OIfQJC7!i%`0v93h{U|pUIFZ&h*4zC4l_Y?OvAm@M*iw-tL=PSvt zy}qL3nM+kM=ofCQHWdyaXC8zwZnMN&MCh^!;cdaeiHXyW ztYCs3b?k?PJKZdbH%!DSJyCO+E+3Y@`gGr0L>llq8|WvI4``0~UG$b9=z9y*w5(if zLzkhSP@M_PCfyI%H51O_@WS5Sr)PUk{(2yP8B~ zy~2WlO(dSEhhPV5sLX z!}r3Il?Q{uzi)Z=S67lFUbepS_%TGOOSFH3>?t`7$_TI$p3ePBA;-!g!`+FXC{N5224LSb(Z#=|>p0VSiKO)=x~HJBInnk5k2p_qGi)e;@HM=D&%L+8@O?!1 zwS8GL|Jg-h^jzcXy%ELkf`$2%w1mFtec8e6)t^=mG=C8%jzzpYw)NkbE#9&bxw5v_ zA*B3{J#~;2R{naRB;ViA>-#SdMdkU6d;IO!=U-khr+(Z^;bGwk;SE$;Ga}UUpJ|rN zkJ)HDit1?OtV>r1AHqSi!Z^>~lmx2-ZvuWLhs{H3Zd@>8O{EklI{|&0#OP5_RLU=X zCmEbji}Pi^05ArKIkkva4(hr2&W;4Gr}J~2c)=ZxRqC8hp+wcMjKVIIB%P3EcKAg2 zD>47@rGuM3N#JClgU<9Efy`UI4NcNe73WGaS@#3MKGo zw>Z-6*h82-ilnB}H&r*rYP$q!5z zbHiQfu}swju`asP7w}?e;7FsUa<1%n>kk!@Cp?;JMnF-!251d6tWZM5!#@}|Ndfgk zKXpbx7B0)iBIPiy&%FkC&Qj4TiL{oF2x__0Gqqc|3&qd%oAeAF2~i0Ks%H!RhYLho z9}9CMa1z{W83QABVXXP=j+T15ZM;$!b!N5ojgYqeVhZHKwRt%wy?Vx@;Wt3&y1 zZNB8te9#9Ay|=X2d$bnJTfXjmou_qC;fuF2-Uw;udTD)ByDV+2C$ za@KIS8KcZnPLv&W7xIb@o-{H@evvSc7U~(B^-^bGJ2rG`Bg3owtWxnk9@o#H2$>0!4dM?XcH8`oKq z9_J&jGIv)ZBj=)sFPk5BX!M_Qd@gLi7C%D>u*_#}wOut{E-0*d(7lo4#lke6*PoYP zJR-&Z06{>$ztpP{_^peRewLF#wk-YpBe-9#^duTR_`HS#vN|ing-K$%T{+TM0i3b?g56QY^`u{L|zVw$8DgSYydGTWv zRRDGM6FU5XQn^wtZKC+Gc5OYH+ytRVFUz$7p%!AZ!gEpu@@wnoi4JCRqLE69^fDb6E zZSW>oUzi4YJ_wA4u6GmcLJIgQ6b*JLO3$ol;*nRpa>B{Mne{JPw}xC@vktDrQ(bd_ z5mCsQTIQ2v06sJj^KCD@4tU{0cN{f9oJ0Q zp3YGv(JPZ8@m_7>y&Sg8YZ2M#>DsuICS_M3J4yv(Dzy89@Aj+(WC^!DpN8BAJSdNI zZd-fdBEVFe8Ml#s5h-#>^+OV05nDC#i?Z|chd&D?U1{Anjyk!gy?A(S)IlpgVaH*T z7P%SMyFQCQ;%B(K-s^Gcs;Sd2?JYb!j9NF+dQDHN|Mg<>_FxoceL-vVevC@^oW<<$ zM3aQiZRb}zSFk(v%6k*w`YV)aqg#oamsN%j3v42znlAeb6oqO;JWy=3U6N6i`-eFc zy1$MsmRNfHq;g$1NWa8vIJG~occ6zF+CH=ECADm8nL{kQ97BRsna{%ux# ztRiPLC+hap-#s?@pj!NxP=p7VGAFhxI4IXY8r(Ni*ef=AV$zqQVSHb6{$)nM{jA`7 z8AS-sM!UV|38hEM%MSUGVLBXi@}ar_DKFQ){0>@yRQ6&<=O2@MAurqU9g1^r;}`7O z24xhO^Trx=+EY1OW(cec3l(9hdQCcQmn^jNXGL6sUMO{O^+ZvqVEDJ?Dib~C*Km5C z%AmQz$2qNC%=-f9B*({p&D=CJczfm7yV(r++q{_yXAX#=INv1{y@wX~G2y>t;u3_z^{~|Wo}O3ZJ?F#4+iM<2w3ih3;49Y5)5+=Dm`%|opDoN z(Rtjv+tXMX_fgW2Lt3f4UfCQG2v9z?2ZXpYRRm$7gmWo@(mcMv;+~&Kzm^AbNniv@)L@PhvNJqY;iprbv&B``Cb^JKM+c`mvdT@Iv8CkS$Rylcb=-lbOi%kQCt|Cpy$^HXd z_a)k)NJ)e@CEst_Q6vc2J#2q#dH|#+=kKgDwC-hF4|unlpj;0c5p8Oc3wvWUt|-1N z;WKw#qBh~lLX$=1KYItlwf(Q5=P!vPaSNX3hRs3OXJin6)joKfDIbJqYyr}6Tw~-a zS!)4uWcKS&ZWnd_$(P=52U8e}z{tYi&0_~IvbAfNBmDUSC%b(@&Z5PL#{#4(6FT278tpyKRP=u(cMYq$STa7`PJ;GCPw zL%wiqDYUdyr*8apLibjtcSYvr#fb8$Zp3zK5DIOVmncjmWUimzwO1Vs+?sqnwr;zh zW5&srE1{=ebl2)iM=z4Jq^Ugou{W?x$n{Z|W>qXSOE6=nZ1N;0vb?_qW zUjzSeyvbx#wW7hNHQ%+8U1W~#BH=o6RNi*`p%gk`>P@sHrgMKs$*G`dq00S+{3>P7 zCt0c=-KX`hg=@$+)vH`?y7%ya4Ep+Mr~(^PhBbc-{uh^I3%ir#bc!3#4xH`(-k(xK zlE=$BJvLXv9r}nV#;0UiRi3CqRelUMax%si5jl0eBHF7N*oGz1yJMwzkCOed* zeVUMH1|{Mqfv9I=lo6fYm~jUg(A-XIQP8Qs#5OjfQN1ZC$Q8-P!b|;`n<62`xog`1dhn{M~2}S0~W=r|J~zk8#vdoP{%eA zz*amw%-)$N`=_B zYV&|c%(FV2Z^Nqq7R+l`jcbSoS8d zOT}qTe&rP}Y>$R;HSx;@-LXyD{Hq*b?$Z>ubk{SDtg5VN$2e)ajQY%+I2d{&cU!SGM(x;sOVw)+a6JRob-6xQ!kC4cFis ztin3Y;8$G++b^_6U&D$`;C8LO8|&r{yHZld$x~`XXUf$1Y<_s9j%;Xe;YM&>DUsF_ zyR&YW0Xz44OF_`)3S+isBa>>q;at492OC*KUm z_`&x(=M}4u!wq|?*b%=id;$(<7n%o50>MCIO2=xQ(m8s@*WCA#p_h?jay7Cdj_&*x z{aa(9D?Msr?Z$)rqwcaBx^%fuMh4qo`B@vtfI0hc_RG^l32{@ZUGG)=?|8+a^)kNN zuV3Dbt~C4ImgW?=&~Th~d?XzD6X43W&4y@+ zli(!0t@${g7kZ56p293`&Xw>b|1t1RRukkZ0CJJUF%7N;s4yXY>Z@BkHl_7TUOsmH zn=lOZbr|Zipgk}T=SSk)P?*#cqj4f}0sy!*yv5IZ*z1BD++dxY+z|;%1}9NBvS|7{ zi)9N69{53$f^HId3mLyxc0o*kxX~Cn#o*9UB@2yWIG%qi4@9;(2Z~TaZd7dai-VX(+$=6yu^5PEq{C_GYyWOKbrDTXg}y=cRtvj z!F9BC)>xqSRh{tVn}&`EA1(pyD=K3gZ}=;v`9j#lwTJJULYBru11IxL^gRAqme_%( z)L5sF`d8|qazF?wNkN0B;%BH|1(F=Y2c)%+ckfR209w$Si& zt>QCV*RXUT>$#_XlhWhYRrgMyHW@KrcO1do(^toEY@meiIS3Wu|ts;(g|Orn5xbxI7{(t))YAAa5T# ztx`W|BDNH$-kNXaA5<>s3CzYc^+q^6wGDY^r0Zt38h&prGoa0_#k5xM#B$0pTuOhP zqW@g@$T1DLE?k#@0#lSml)flESIPn^6#|QJJh)q~1~++IQ2j^sQH{BycZW1!K|obh zJXH*EZInUS`05kx0d#dpJG{~c{LaB)R}AK+I~_`prz+4*8%RgXyCI*G?MHWG;sM#A zQ$9IGU(Q*X+heYD8q1?jL49S#UP4s?S8-9KD7P%^y;a}|AQ^awem)U(X}T4x%}Ix7 zCn?l(-#S_39f-9Ty9uaY9XE;xNaGyzc2a?V-F-;>HJB@jnbFJLj~C{Fg}@^aBrtLP zX@VT-Bs`1uTnWP`9zkn6+77Jin?YS6PIryry4EMfckkP%OVP|-{JVV0ry&9rH~cBx zxjt~jFOwjAmItl&iMIlwb$6(ay==ZZ{>4hc@MQ`-a$V)q4xT}4vp}~Y@ zL=DYB_u*IldFR=Ar~G66?>ighi5>}4wvqkGkwr$X|4ek{$DB;A2wbFF|MvOtZCnv!keRe+7<@eVO`6Sw}rGkyZp z@3*}4TEkYo4@)ybYu|g3g8<8=;)&<%9OI#IN6|ELBeWqNjM2K+-FnbHzXmkkI~M_z zue%WC*aV|i`7t5xJ3e;KygI{=oAwzd2|*r{(CmZm3R{9nvB^z>0rib!l*jyxf5U3L z&ljoKxlD&d#bae<(UC4lmP#1e_@5ufe{qQQI#m>=AADWekrrARRe4ydoaF<(u!Fg z+HHg}+F5faIop1KuO(q5_?}S{c6ur3n4%86O6nS8@`aSd>|j=dkWWMJ*J=)3rmNoK zE&6~@fBfuXbUh(n-s@43;BfM~Qb-_;=>Fc0(rRqxY0H|C3vkrS$-A^xzg@@}^;fbp zM{MSK{TZ#l=Cot2@`m5se)S2I)#{OT*bCblt>uQ5oO|BxVy&!q*{Fe)m7M9~NlLi9 zwfYI3A+)o9dF{-StkVD(I&WONc}WXnoJSdI^r<^3TP?q|c1-IvM5g{pujJ*>$Z?qC z4S#HiwESM{$ND*wo0LJ7(A65(wE-wH%y*D#C@reuPK9>3d<>rm_c5AEPL_9+R-t$z zzGf5hjd;HFzOHN*34ddg;oqCDdHCrTJ%^O3>6i5S(l#gZW6?~^AM|zID=H3>Q>v!v z5^HI5Ez#n%9X-HyvewN^3zs~2zobi2aL@mh%9w(QpYhyWd!G$w`u|_kOgZN%eY}kD zitw;-F-Zg#jLQ8*w7mklj*mkjoDyARJaaz3J#TQU?FQ{uKpNmV&Kz%ymm@rdl&lDk z3p0d??uv>z@4buK8SnJza=wR2zWw&WO5(vI51X;YJji*dKENMmfzv9lAL46)P)m!e zNNHvUL4wIhcR(scm<(j4%EZ=yU(n6QcBAaEupndfG=LLus)na8DW-50$z0HIW^$YZ z2;)m}y!zzJfHMYfhqLY}1G(wa+5M8(BQ_gh%pXVxC*{$k%NZfL652;U_VaT{A433t z0C)+!H z&5Qb8NUChn>9(GXcl`P}MVW*Ad*BK-fAJ5IvF;bNwjB1}8{One&AnNYix)C6FQ?>F7=J7iT_{_Y>b>tHY!P84w)*#b1#9@XELNA_ zgiZxp?%LZVt_0}aSNHv-LCAKA*}k~jvgJ9za-WFYkPnaJv)Be^t56%a=)|xS#C)P= z!?{@Gi~jU?D+8bM;UVt83+>4H{CQ&&tSg5r)4PAw+qM!1qcqKiq?;m~O4*q59jk1O zPJ-p~f%Cg`o$+0uyP2Y+t*vbYzmV4au#>-( zuay3t*%qnM#B+B>1)XhJ5Bd_ZViR!NX3WPd;&iQ_qh7CFh_y;c(>>=8qvwN$Gtlnx zduyY^`>#r1`0`{a^C4R&oxPUvS{!hv`vV9Lk`L14K3XhRw7zK+%}_V9hU<|l=ac(2 zf{rJ~)DDObJ7s%j#*2j#cSctVSF@!s&VYu2ZZm?~tj{TI1SG%4>!_21MZ6opOG9a)~|$A+#- z{H*d{ic`@nuaXTQ_iM6F%(l5HW|kREf6n)aXSsZEEd6M+%T5o;@KES_@<%r5xJvy_ z_C-C8jr0Ddx*Osbi!$V8Z?#_3YK=qKfBxq*_A6Ugu7AsoGHHKvX-P)k*g|h5#Go*Y zR#xWXoKJg`=WgWX7c^||Zu|c+b?)&@zyJT=3=1(SDZ?mIIi(zO*i4ELI+7gbR0vHu z#>ONhXDUigsZ`G7%r-g8VRJs8W6m4Hu-Wh3?fd(@Kc9cEKVE;luEXu|d|c1#{#X&fQ0nG10 z!im!z!NZw>GR)-GDU$yDZkY&sB}?!j*V8St)vFUD3Zg5A%_yWOyiLtp=l`NJPVl>X zrFitv*abocm;Pt8)`kE6%7aRj_Sqh`tEIQquRoco0R5&qA>%7E7o;)OCo~+uE;lxj* zzf5kKf!+@*Z#V^wOD`!Wf$^VCqjVGC*R!I-YTOuH$KJm2%8(e(=EDIcPZgQjGRa_M zTrCk^!ibZ72t;`2NgywX&K>UnS%1@G%>kbHm%UNr@CA6_4*8~3qWw4oPrkaoJ|ber z6D_q_&%ML}S-fBK_obvU`rx(<)MhACmqg*Vdr+#_y|^^L92?Os3jE zPn^DZQ-Hmc$6qEwG#Yk7Hee->;7;6+Kw%B64pq+F(|R2Ev+YL9Myb{gX}tHUh`4r5 z&f!2MSFr0On5?gN(wVSWpY_ChJLFdR0_h!PeK%L)XRCoul0GO|544iJs#$u?ceQnU zM|)9ZV-8FQA+Df=@)8&vL1w6a2wCmS*#O!8P>NTzMkP(v};|TzPC56 zhNsoPFk>nr$kR1_dO-lX8)|>#t-*zm^`~{kq1u@A4NRi=7>5W(yzWob+@_U8&}wmu zgq2Faf#SnhIoE}i`jc#W)y}<<*n^dh>B-h{fqKU079q{V(q)0w~o1#qaCVCtkcC znYiCJP-|Yh7c)B|oe^X4@Tag=OKw%jv-#tlM_TRO$@4CyCktOZw1DFt9!OqzVj=t> z5#xRs8-YJfj+v_P^_=wa=>@`@Pa_r;SOvdEBQUWeFZkPYS*>&5&;tWy!SzR-XRAw} z?Mo{KFuYMVL2fy>R|{aXDE1qRq8>jM5c{yBlys%#WUzmiL6xEoDl@MHp1o| z`O;3Y++byg?c5hx;6v;HW{ep}xM_}E66|9SwIIspO`CVursEFUbQL4z@w0b6CDCot zyVi0Wup5i)y`$Chkc{09=Xc}n5sLfB3N;G>HU?Z{XT{NMYutgpY~=qgW6g*$wAr&I z5?C<#801a+L|9Z8+9xegznZy#2lwI#gLP{@3;RVu9ud;x%SrQ%duiju6;Z6_sfd6@ zdsN)Bfm2Wfwb2&->KP(UGvq2&!Cu;5Zg^m9Z z2ClhD#*1bZQ+YB&{c10h5XzPo&@|450g4 z4EQjo1fx(R%9+oG&w{fg+w8mI?QtUy=z@y7ssNRIs%ZJBpX zhwGeUaSV+4JIXIJMH9Zu7jKBzA(;!r2~66wfNV)mTMcY|vLSg=M z{G|JY>84)8HH2eQcqMrIT(xlLZQlS>fW3wIy`?XFwdcSp9O+qHRDeG421!02bb9H!>Z*TW3B5;ewZ_`LiA!qe-2EGeh$m+8TAN@vcZUb;85SY^x$BY z|9Ur*TYRm zEq4)nN3D%yDKmqNPK!NL#ah})Z(V&tTgOfSHR1-dwis9FLyFP>>U{cWgIG!wQNj#A zmGPLdlPil`XPiw)637VJTaTk8>< z8{ocu;_g6Q4;jMqd%l)dXG6KuN}91^Me zb46Kx;6E$?`_y*1S&1M{M4vrE8X)gVxML+ag000aXiZuIJ7m-S9k_(KuJIa`IBqcV z6*scFB|f%YS04sxNrSC&WPPU1*sk(Tgusoeme^kGY&P@D$dIL#p??pevv}wdOAUs! zN<<-c-aMT|Z@rtlr}LZv_W2#9R=1a5Soz5P-h`hDSr#J7NV1S{%n#}^t1MuD1w#&# za@A(EYMaM%QJtm6?x6x)NsOZ&t?B*W71>8t%XXW+HZN*x3<-%R>LI|kZ41>e7wDAg zzF|M_F^gD2$)%JDYVag>M*${iv+={Y=6B)8z&Wk4`Sb(uXs#P!yR1+_=a+<0|LI`u z>4tSa6Y^_QwJ>wvi1KAj>hnx5FPTmIgYGfQ)Kb@)%0(Me1}jjQfX^y)^IA1}hHt8b0IEV0kr+HpmzF2-POtj2pz9E-f(=hD>2MJGip7N3*ENdqJ5Ho ztY`{PYlezYx$J&2$}^uN#Kc}JFXd9TYwWEB${~+o;3$q)0MpubVYJx{Sa3}V&9WPERd2$o{oh?TA^XZwUgy9Fi?c34;3RU3} z>2I6v7^XRauJnxlO<~)tIltONj6!CW;O<~9W4dEMg~e9)<)`DWOx@j>vgqle^cdIZ zo?k#b7jqXhDS!D@8p)9Ya(RBmrj9ZCyX@i<*wXOMYC-pN73%HB$c!o2kb2|l1tKw~uoY;uW)Z+Z0M8W`yp~M<)5Mc^uI3~ZrMJoS>JH{^) z1L?qt_UcCwcHUgPTX%v4%xu?X;}@j(XXP)9R2p#O^KhBf)Wx^S&5HGao!wr)x@DW+ zh5ns6v>3?cAxu_v$WS8f7aafgNt<)@Z!i@ufFX&;4_EsAR;b9Mg}@H}vOFcO)XN@F z*EKbLJEovIEondDtjBI`l}JV3JBblY2~uV-z6vK~A`tgdV){ ziW7)2d?7wB=u0lC~0|~LxNr(QH+nsS*uz|SA8_J`ptCq+tb1O zQj%9h+CRwhq(48)2Uq*>WLRvsb4z-6_!Rb24$Hdt2_t!MIzWYTTzv5H`Ur(mr6yhR zOkiuJWq8CGPHCU5xf`g}>$-I=NuD;O`sc@>--bnO0p(+%9qa|Ay<)TnzG}C2HO_ll zwXQIrKX*XvQ_wPj`bOQLTY>?l54YA2?Cc7Ibz~Bv(xEkrA}Ig2t!G!CuRn0!Asd>z zfxPZ%Uh9TRQcVaO#wp_+wc}#q+Zbxqichh(EOt`o!LgdFP6WE}2a_ofQ_Of4E&u*BA6y zE99Viu)e9aKtx9>KCkx_SB-uCw{VE&4*LGrXSgl0>GAJbA@UiB6P`Naf9OVuoi6|; z|4hL%1b%#oco;VYouGybzoJT)URx)JRfOecUn_6Q#{fWlo?xOd$v%s3tc2)0IWuxa z48nB_XSd`Dun2j!22u`j>%WIpXz!1Dx6dIzS3Gnad0OrP9+cl_tWT6hic+`^gGwHx z@YMnatVOQ#*zw#1oSKHkErA5(;*Q_Ua$A$*1JE^#(LtN3#yfx0cRBVrSpsVDVB$_f z9L5O2}3iLFoVZ8 z{?P^Hcs^~8782KIz(w*EJ!sLri|ujgb2V-P7+H0E;vI>d8L}nfxpc|C?bMZtL<`qq zB|wcG$V%Yy!`T&%!X?j4IgJ+Rz9&gxm-YqE!m@NFLU=Tf%@xh!;Jl}@Mc9#2~-=~?U=a>8u zV*gU7g|2d+lN^c_#X@Z&4y16>Q_X{m&c5#!<^DM@QF9CEx!!g1!Y1RTZ!c(@d=eeVf=citxsL-fTXZRvH*v4w_R z4GNybj7>|1nwo#uH_5NXA%&ln97CVhSxi+aRzAdBET`&@G^V@^2g4ZZK1P8VrHa)kJfD03f%IH zN#}gB2v*RQg6y5xEQ`?Kg7Xe#MWw5Qid3~Obzb6(;JW?BUeDc>ghV2z43N6W2e>Na z3*X6#-M?h6^LhS?aoYWO`5<;CsfC1*=C75<`PNcj0?8_FYm=^BpRYLbQis{7u(4 z3<}+XmV-G$I8t3c{!L!tc;8MyLwyp#-3Gh&TF^%RG8M{_mfOCcF*`pdt3Jc~a$^!F zA6(Pg7QMr3JO>j12O1*HqHO+A z{^o`g$IA*zmGtCCg4L;aZ^P#?Z_eUWsnN+Jl`Q-2)o_#7042B zL!k?J7he-0A-crvpY>hhA}ARsRZyuRxoGNQR11ZwsVA9g#jDTr>4+qYtSiK zop_;}F0yF>=(?=|;a8wzqzK8l@iSfR9MW%gbymh#U&~sQwa~0~yWtFd)O9%q?(A#* zn<|&$OhX@l{BXDXI7?SoC&pzT2Gd4S`w^3pbbevhC0b z+3X7ZVgoNH>-$15c!P(&m~1`5b(^G~CNsFTJSY;+qW7995koVmkW|gr6<^wLux&}0 z=gR|J>QdSF^~*Iguu6z`qsRp=CGY4ntjv}HwZV_hC}yeE!cdG!tD~0*3aGKvWS|0y z-qxfNL>BoWNYmW)A#jl?+i77`7;|o}--3t#Vsi%GrXjmd&FOT@mPyOrmDJ8HuCsmm zaBVGp3#*OM-h9rBo@-eV5)W_}q2G)peB2pEkb*yhG@X4iF^vRZTSTk2#3V6?RS&*- z*QqVfZRgNFK%h#^c`2uTua`ugM0zMIk1UN!`vNNFIiFfaA5V=&u?v3}KQ|CpGJTcb zd}`0%%cNxfG;Fu^!M&LP_P#I!NC#zt=pQ(RJ=DP3cq&^)?cZ>k<-HEf<=@%Rm) zrz%~J5ZU;U!!)yql=n(AqZ?ZL1fM=*!5B`Tw@t?gdbG?Gm%D#~(Gr6IECd_t8d%(f z?di|8NRA^PsVUBwQ?Pp7GH(?i(mVWq(ooU(RByh>OX=G~B)$uC^))XOKNplUEBB@F z=k48Pk{pA2d36mx?k%z(%tUCy|dO&>R1d}r$QqQpZ(~L zVDM;^^=F)Ve$@}7$};N6l>aw!*+!?@v=tLNojS@Gn`k{1&$~Ty= z=H!OQsoCu3<-N2sM3>vk?LBg=U0JWERwS!xjMk(dke+roDN)rjzE^&! zYe&i6HA8H%)BmX{)R#KSIBtv$zYF{8eGGh*QX|Vvw8n(1@ru%p5=vqJRu?W)I-)5r zr~H5Yt2Lh70SW1tAND`AB`IGur8~3)fD}l~hdjpdOTKi&@w5@muuFINC^RD9Kc{6N z8?g2<;vu9`q8&~hA#uk8cyS;izc>QI0_auX{B|!T`5dh9lGCATYWEpT-^3a8=!Ch(Qt&sbqbFQ1z&Q%N5K5o`j#B4R@&Udhz%iU6QbsasFR1tE)FICsW?TE#|@b5|WU~ zTX`u`;HftR^0_pw_Dr4&&PpBuN7HKe{;0b@-2{r=Pc=MA#IVkZ%h=*Ku}*2(7ppXgX#tZ!HK{-~Ea=3i`JT>f7bcC?jwok|4G2Om1Y2qw1oozoe zdmo`m2fuRV;No1k*89QEBmNi940XKrImkhK9o4y&+_|V<4-4$cH4C53ejA8s9Gd>^ zlA_KUh$y~t9Gd{;-_e(mi8;ELsM2SeF+%>LWJ0}^VIqG-e_d^m_HtWGVR%DD)8;pL zY1Ec&6xfo}@7JYoM#njOxYu2v{-xT-c;x!WT-)~QzNYTpY1YVJ{`2{{C`jqs#-#4M zpU^8t9aX{esFn8P1%XF&uaGMLEHu9?7E~J1Y<4V>j@$g_lZ%|cTDC zB|lo8r0IoDZa6Ac52I?fGlX9>S+iPDSWycmHEl$~Lpz}%{Lwaf5U=uX@ecgy9rpg^ zdOzBdrp;PmR$RH;UfK6Tx>$C;vU`XBZkPYF{AB)dLXC(oA-C@7i(qIP;SP4?PulFi z^@Q~?{u@`QpYA`K`;VCL8UzplxIR7kn+E4-AQIrp9u{l$p0f{lf&(NSl6RSUlW&SD zC_c1Uf1f58T@HQ(8IKiiA%f5n%dpHpomM~d#wa1$MMxWuGSC}o~ydX9M4!(@(!x8;%h`6i((DzyT9`%;k_> z-*E`hI#YZK*qVulV?mFlyhXo=Pvi+G)%NMBd*q7K5hPbzxm7kZ!o1>3~^i` z%+V^EcpB082opapwwu0F>x_L}^!c;dC^^mFF6g38GT;l0`8_+h4dg?KjBVuif+0wy zkB!}_GjRy)W5J720tMtu!G4%;cqK=sN-Nv~1CbqUOnmB3g4DqIf1*>%p zhvyB$t%fZOYJ%Llg9?1soc0S`7l!b&?vu8oOG+;wj&ERhMj(m;b z!E|@k-^<{(nB!@~(gM3NB0-JP+k$S4)}XLVx1~t+mARm^Rpj!tx!lRUsNgNrjdY2j z-ny3MLPAhzx|xXIWR{JOR=K)9gCBM<3u9N#WCrRpjRMZ?#l*3cC1$-G z742&K+85gu^gj-xhwa3fQTBda1iAfx8X25>a1l0;T!|~fkncDEBzuB)hj`u)bd&4m z)`b-hX9(k9fxl?U^Rm2Pi=HQ>XeoeH0BMnk6`qQpd&>#aB%0ZYTrh!qc!ZcDbs&@G zHd8!0#4~LdDnOJUpu7^FD5QXQtMevN9I`#Pg$t^c>BU)Y?*TExwd3;2VBGVKHrc^Y zEs_A%3bN<2Z(_c&-12lo>)l7G!q1ZFI9uc+Vk!G1NDMS*Sqzqpm(2Sdw|6bh-T%v_ z5@TPWgY*voiE+4fzaA1)jz8jww;KWGN|Va+?BFVg6Gg{a{jDIz{5b zVLIRnP!`*Mn0N>9A*IWK!{C?gG{9@iiM6-2MYHrF`W*2`Nc|Yf4hQE z$LP&bb8kEd3?_jIS9K<6e=1B~+|H0W+m?zZD+Gg=P$P+70QU^xkv!Mn+lD~_JvwO| zkMg^K4sp32iQ{Yw`h)%^! zRw@r1n>E7xJjpg^MaDVgJzWmtRG~@@JeI?s-2;qpZ0E0BnXL!ItnDfpFP&GJ1kI=2 z3%X72%H6+@CC1Kt*SLBL=Ar~#d>A$pw~M^O?Llvi_M3e$b|koI%+`*Sk;0AKK6isa zV=trF0W!}Ug|T8D%C;XbwYwGE>0w#}?AZu43mrDT$gk5?s%%s8UgR!rUFe0aSfclg!TykO@JgP?7ypkCS_CFB z_**V&p)W!3v&bMR88cTJu^RL$A7k{cpzKmb4-Iv&k^IFkQe#KVsVLD%3VN{1Gde}K zez;!IeLK}lCfxW#F-)iALy}1wF}UM5GYv8oWh1jsUH^D5%)Ih(LyJ3Op3ofrJe)nP z6vU|BcGp#BJnK?sC7E$d)apIqDVm#c#7x#X)>c-O6hlAw=U8~KzAlWYLv}`OS=&pG zl4oAJb^R&a?Rxx^)i$xczrCai)ATiQB?PS;+?CzKp6W4nV*WC=aHobqFm_v!U9)jI zcyF1L-6HN51=f~cH34)YO6K|E+l=Cz`=`#eK%)KBqWrc=v4rWScdv!!n^J(sUK zqO2R-0x~nyw6)eNN(MV^BLV}&bB&c~apCx0P>o-H-M(Y>kyLFqNlB|GhAroEBQ=V8 zgYA#Jbg&e-{ohD|3O75U7wf)|6TAO;b}Q}%I0STAnikfDWC7m4u2>DZHo+)+%@;qH zHL3T5GwkIt953nY?(FnHXnurOQ)nr8CU&rX8;qnE2D9AY3=XjLD_q&$PW)QP(ENQE z<$AOWSOs(*IGNW$N_N};q~UICp@QK~N!Nmf-p{@B{sH&`p9ARuA0pKtj3uio@O{o- znf;-J@;RC82tQ17upLMoS9e7}bfa1g;yx>X`X5UjIRtlgeJy}dJyIv%YHZ&+P zln1!V(K69o@N2Mw-A>z;RuP7moCyL}ES=OF**t<+ zVBhQcV#1Mu?vD-rS_m)9{lz@V|1p$X8p0?Lk%{J-?O-T>l^JThtX$9VwkC&nA9pNM zR@`>%ynLq*^owIMq{Jb$dor@Zq54IvMwj*}MZ?c0vj@wQk4ZlPY&!g;$)0n0n^fui z(NHuIw7xU$H~G43Bt;z8(s!LJ}BWrnLzjiA5r)Ca|L@VI*~jst;qt zdkU4x*mrcAZyW9bK5VVCKPKjwiF_E9ZpC#SWW3dwDQX(=cd->2>?ob*)F#(edIX7R zEslKjv;9(4^W2=!6}d>W=y?${Zo!;}1XmVciSMH36 zzO2)75Xi$j=E!EzK}RX0BF92$oGN0tCiH@t(BXaC{x?>~JLc==-o5QT6UF_cFq=hD z__Z|uxMn>6$;5SO_`>E}mIRg{?5L+jAn%N_vIFHaUJO<|A51CRh%<9VH)SxUf=197 zE03j#_!#DK$Z#Kg+*qmZr^_e%!-1Ajikm1_b_{qgu;!C(tt>jwWBX!_(lLV3KMElh z@*2wM%{bWQ+|Q1ZEslZ2qqI+ppoF_=IY9xLFlgQQ6J0(X7pG;HKUUjDgF8XiHq6|X z9ZI}y&lk6#(Vh(WsO!->Ooys=R+!SlmV!Cd%XUOirzg~m6a`+B+j6~hK+5So_dg9x zObbwwAhS?$-sQi#VE_jfajg>8s4_!jV}dT@o-PQ;TII8&g#PMpSQQ)5<(vK z3{xaZR>N-wwsQG$tm6c5_w0@lMP7Yo<9IO7G^On!wx%-V>z_}4Gc-d#LW5+M?#=(jc5|%ZF4Ho*N3Q+WfLdwfzz&EljEf_4GICQW z@eM=k6HtI^(+{AbpB(2u=iJr8rB7*Q9C1vwa{6>>aAH0EQ-1M#^vI^!mT-w?guz9V zll)_@{D7ty#{+@8YUPY96NhND(W6bT;+v=HPt0e_MipmQzEgq*Vji5DJi+~oBPXu> zbHwM=T_bH_bfb~s>A|#1&4BHBStYIOPhLSs0^SpcEDSKm9C5W>#@v?mvFY9OCuf_h zWAdCX=ESE(s}=X%??WjA#IdiFUf+xLmNZY%qALl^j%%01DLH*LyB6FGe?Y|24BW|u zLzytm{)*ku6^A;_5wanVu1A*^f~5SMnlPSr8fHV{!Ty*HYBD}%E^GMldDc~H;qJc9 zir4+#>64cYk};hc^pdb3kHupO#lvIy!SuSh7vY-)y%kRKFfYQD9eQg9!-2)XeD$3_ zTTU&BeW^5FnpO)0>-JJ+7`}QQbTPpy)x~&(YBu`+Dz_WEe+AK^zQ_&Ro86CHd<~oa{WWgh#G5p^vtaG( zx9T*U;bpzQ@T zd(Jw!@%^)#PG0v?gt>RQky2 z;&pfD(+z}1qaLoECVSg_v@b$`1n#>mtLr?oWwtIXUO!U7hB%`z3;21E(HJEzH~t}#VkA)@?Y!FS9Qw1-<+q9YPTFBT-4W45^)>y3wIv( zvN67bTvz7q^CAFei2ODuNNQ-1w`*unzL^Vt#QGNZ)oa__iL_CmreTpPfXIV#GGz$(083X9q#R;F8*yG;5{tFmSk5qfV2wUxKr@eFG zY+OC+MSs{UYsGOYWXS*;&u{m$ho53Hi9A&D;0wT-b8afmL}!m2SI`gk=N9PY69nnV z^yB(5AO%VJ3t?U4%u4_n4GN&p{KT24xr_ieK*+ypq5>lHotIF!5QzR8Pj5v~z7+;9 z0B)}s8Mt#1Ilg7Cf2i9_%w%$}3nk2PSQqmLK_TRcWlI8uV}X*y?%nCnwXg^bil6)* zRk#vnTTaT@1}ShIDnD zxV2gH>I6$37j*h2aFBLm4O6MC2cdthL;RnEz$j5ZwO9{}0ps1}PwF%mDL+9Um4!q! zhz179HH!roeCtjo>q2b+c?=nDsKNX>DFQtgngukuyco=`UJ)W8Rjbb3d>P$EL-_28_6X;Rt&T`D|T zPz9XsmvyH$1wy%TWSP>Yu~GdbTU{yNdel~%7OAIn<+=+wznbncItqnvdv=}qx@{HA zGB=yi3)DdC`L27|z(|A$5uO*z=E-mZubgz2Z(!zY0H);oL0`B5&-H znDyyQCIX%~UYod{s+5+b{e#4Zpd03FI~fo(QY6TI;Ju~YI;Eft?G|%k@3Q>6eyA4J zg{^Ef6s9>m*;ZbVVNSBYgY6AXnXY*IwE0KMBc|aiGN0xeYB(7`?pK_wq)!DJI<1a- zo$OB3UqdtaP|A<6bs;y@yh+mPwv_ky#pJpjOw!_K>%eiBbxqM3(a9biwoKgp`W>>` z!I&0FJV>~7*_v1AeT;w5 z-93Gb;TNl9z5yoGgr}RpSu5)aq&tNO8;iJD3MYKw(*oBCT0^B{aDZ?ID?@+BpK15M z-a(F=o%x~kf@BS?e@zCf9wa2E;_h>cHa!(w9|&I54&6=E7E&a}{ap;p_Q&u4kwQ7} zTn9@64gvDT0#n|RzWuYvbSzh8+@$q!h9O#jGpsdOA~0vd4txcO2e?R711}0+47d)~ zrW4mHHpv6}x&*h7yCe5VTo+v|bX$%Kf?hs7e9Y5UN)9_c$O|z%mtm{vjx4+?CVBm$ zz=^BefHUI2BgZgh3A*GKOFqk?=n0T)>u)v0Q<-yZK}4}!6`<={(X*>S+#zs)XCpSmwy+~y8cVWODr@JSRs%8bxvw&QXhc{vwJ ziDa6~C-P+$C=g}$DWgvptmAZkGB|bY_)hqz(jgohv=>n*y0K<1kLrieCZg4CC+ z=Ca08e9kzbfvYl~d<+W)*tIOCq=R~Ygm%3l6FXIiZ1-kDU5dkD>bdGOT>n zcr~g)=0+p`tPtsmm+i+3!6$ydO%!N)jLrVum?2y$w1+4?gnKKWXs6uV=i&bPXmcE% z>ZZ-n%rZkd@{rboOW$c9Yq3dA0ey~2o7Zvdh`KbBH1u%Hc>a}7qDPFy0;K_+y4+&r z&?UB$1;M&(g`y&ZB3m3{?@bD^oCLG6rz&^(1kKJwrXh|*npM+h!m3VsD1FH^4yL?5 ziF};F*BFsvckL)iv|D!d!@+uAy3U-%UM7qSwtFo2bVbT@tX0^0QkJd2({8EyYuqY* z&W^r|V&cQrrPl~4c}mr~sM-M~tq=RmV3js&JVoY+vUtqjylFQd-{KMGj6`H`icUwd zP#AONFSusLh+S6evJLZAJCH9@`u6`DYdgT)}h*#{2Z~Z&4_0Nz%+m%{}i}Gm9Z3 z*38@IwpQC6FzUBhT}-=v*mxkVJPq4|)VBh8g5x@fxd$7#J9 ze^Q*|&SL`ugJmzGzkcs7a}d3-x#<1YEZZyhTJ5uS=RZY^X+%uy@J9qiMdF{e#T!9DhC?10%(4^Nx~W2P1`iB238cqVbjZj7{NU;?U0FhuN}ws;J?8o3`#4 z+3MNeK&!jx|2hSgum0X1Ui-P!s`CFg|H9W0+>27*Z{j45BH#s%vxY8)JpgFPNeOqE zi?|5fd1Q*3#0xkbKd3el?(k# zoEOPbABfuJ$$d_ca@i=iJp)+EICP;Pn@gJO>vYo+S1#_nJMfU7fakCIrP4Ca4E*Ac zd(q}gfHi3u;2WGO5*oks8+EX;AuM@*^s5<`b_?_AEk8{UhP%Lu4`b#RkOk~MV1Ihu z=-vjZYy+II9IH0HA_!nDLpLLqa)0KijW>f6dE$_!$U!s6P2#oG>_R?r?CO_YP4eu z?$+$&eYNfrVCKmE3qj|b%@0z|X88(*RGX_SA0dhjFAhc7ES+?wWR3ctdt_-(sYOUX zYczDcjGUA`Xw=}*RB1hjXS3FWQMJ(CIUkdZy;7MteACszjzS?Bgw)xI_v(wLHyQ}{ zr@tIx8B;BRXU;5#rpq*v)LO3p_+)YW_H85BNMxKtZgHHtvL3Vz=dvsI^SR}x2-9A) zCLw8``3*t+J{5u~n6{&q2ub-YySj2bwCyuE>#68;1cx?SSXiRA! z&|D>a(L~6WiNxN;GrP&Vw{@7e1LfGXIXq5&WnfH?%_!Z?Qw*LRoSW`#PQ9vnF!cNP zZ8{OPrEB4DcLz%|@exh!blhGV+OS$Jy@VZNL@Yabc62QjYp}kVp@bLOKIQvjxXhTJ z@%aPP4?#~baR!mVem+Gx4sod8S%HVyD8_OEnMdy7tP!s z4|iVZe$c3baFw6wiy}(l+<2TgcJAOArQ!VC-@1@lz7?4qX1+r{Q>X^kbo1;w576;^>Wm~G8!S!=)1;WkmoeizUNfEKc!<7NVH0IY;D!#xcL(*#ZRQCWjP=6MPS6_7%_ z3pi!N(U|w(Gs+kAy7=u?1rj(O*m(MA8c0$4H>H<56C{zyMIfrSsiHrCBBi$gS4V)l z0Qrb#KO`*D+mmB%y9!Mwg1Q9HXJS!=8Gog?i98mM@PkW+B@cvA- zv{=>Fy^8{^ZQZsQS`*jR2jZ@8ISJ+rm9(?;6f;usWFWLh`90H2VoJ)Bv?N{KKqzNq zqI0!zP08h-l{IK?g>do9^KSI-FwAVj7EVttfIiK>T=FHxde76^mAQH}%O;iOGWAIG zZ29PB;K4&BU(sB(sS7$u@k_{nc(Qea3C}aIvDAwguIO7tbR!7-A;nJMGEn`RCTj6S7m z^B!Z~PZ94fhoN&1x7i=v?E4}&YxTlFTTSQH5z69kG1(^%77u-^`1EqDz!lw*qcrO}n$F-JU#C;tJBP@NrlJRCz0d+X)U?Sp(F z#KH3g>nu|(rLn17B_kgf?#!CMdp`W?BV{R>8KLn0aWo|vmHh|bW_*2Kb4O^d1zJOX zt`-e0pS`s7&Dh+#uK|X-fRNGatXL^1D*Yrdo^S`g6{SW05wmRW*Aufnm)J1}_FjlO zL6_&TqJ6j82+-g&L97Pd#3V{8@L+$+c z=xFO(1$+Nen+d;%PMpVHyc@k>z`Pr{HT-UZd`k?~cf~er<@IEKke`_`+990waC1c( zab2|3V+ZZs?7W!9zDKcz7q_cAV*jhx6+dyGk@7#1uX*u5Y2Hv_9(f2fBp^4Ptw^`;{xAn)oB&3v~aFX1_^ZqC>F9>fXQXv~8c+-J_?j?sa0l`SXEaX=JH%M$B^9Y~QJM@Q3<7ndmCoirOC3bVgrF zXPl)@UiP>_yXF`>dD4$LoKkeMu*pKDud{pitA^|Y7bNeIl!~-= zqu>@to{R(_pP@WZ_Ez$eFR*f6D=aAZ`F5ziY zwX`2*Pou0IGA>_J5(;guNqIWtkn%D%h|=E<-S)D2(9K66HIZeH#R-*zgRuDO4Ga4xa+=TgT~sR|(wFgV&CV<} zMT%{QpU&xo7NxCT_E8Lz)?)FiqxPJ=H2EcXRr&|jD0YXyX8kTC8JvDFFPSU6oBK9t zu6+8ps+V8xCgc6Sx}nBqrBl2n6Zxzi;~q^C-eZodmXJ!7XRDI~se|pB0m_8wJK{Xv zs|-(!(M#i%DiKhjb%8`p>h+Pm7g5{gxIGsPhTLbY1>(8T&ZGAOpJ7b96MJVf!yRqF?ph zIJzi8jWGie@*D{SMy>9uKG=ULsj+EQJM|W#wNTL$#Hwu4alkoA>Im}v^G}p}SbD5Q zxsk4ur)RV#y}EzLBE*d(9F?u8tIhAXW4f)JR@o0#_lgthW;vC0QurGA_GTe#&pTb2 zt^Q9x2KQ)*Vw%zw`?FeR%=hMqYA`hkOQ0uq=I-7 za$v4|pT%_;@QA|{LUQ|J=U+ z?ICH{?4>=T+~n0sRSavZFxY15Jn9n)xj3;So@8$pb-1}?b1QV<|0?nfcixQ0>y?w8 zA^?vdH!-T8V^qfXmhKosgu;6wOnlQY74cJH{!R{blZvIJJ@Sbx<_n`&+Gt0 z84ui<=dZ1NWc?V;%#u-?z81&=?Di{3ot4U7^TThZ-5XTCAc6(p>(oYO1|Q20=9ie$ zBf*<8!<72RdaHeMQWVVt8LdVsHeW&I!>t$ug~+F3FuEj#*;hDz-9-%T^_j7H$(z3P z;cfd&xbQh|hoh(4l)k@0@hB9NtlbNQLOX|z%58MI79sD216iY#m?y1`um{6u-|hYU zv=78pFRy&V2ka&|E<`Pm3n7GM19xu;{_=_!oxI56v05jI?7#!lLi37IN*sg|l~2Fc zO>3NRDBHc(V|x4B)qJz=4-L|^jx?tEr|9mPC^3q3oN!RkLg89xPr{r4qVDtT#~3Va zSqAEb&QPqF>0;+LD4k>^R42wUd*G=DG8@s(ER}6~ce1o>r*mwKVZf7u4XB4#t@NQ` zN*gvy1l{UVo4IbQTaq|I89dPk6Q@L>HrLHCGe?(gb;W-Qd$`Z&d5m>SLp11Zlfzko z8hi4lS$gKPaq|A;oc#nLBJB#8(Mu5iF$Z_wsZ8ws%L>~_-P7lrbbpvc={FbaNv;~@ za7U< z#Y2?6z4^1ca_Cy28a@aAsm@EJP8$8R5&aA*f|)FRQ7K=3g6+UTP~C3wpe3WBcf7yC z*|)_uz35^DEAId@_R-vb9t-yV)$F}n7o83MmGQ=+RFc0|Q|OsY&dNXa$EbiQfEYtt zP_i%7V-P2g#DiSg&K z10JcX5B>@4oDlebguPc-(@(Sonh=WAARQ?I2}Y$C=>!DnO+`he7XhVrgaARM_adN_ zfGDU)uhJ0+AT2a$(h&j#NFYE6?drY%bI#+<<2>Y>Z{{~^uUUJ~+L;X>=}V>o+DM2E zhvx#%4x$q#Asj7C-6y8#cf9xK4-9!<;EE(e1(s_n9ul%c|8zcj5^|1s1a3e1?K3QP zasuU6!X-L2rifDJTQ9)G@{izDUlMj>jBZ$Q2g$^5(stSBe+Tb^^>qSr;b8>Mz=&ilDJxMb>&>SHu+8 zg=tUqUuB8(xoZ{p!2NQ*`G%YMK)-(*1=1O?DbP67&*U&H)8x+wW>(~DN_@a1af!uX zpD{^I{yzTaJ4JvUe>4(d-C%bkcD>k3)r{u|E>%F&D=(C4zyd2OMN|o_Jv*S^%9z@Z z2@N}AWy{a~#_gT9kNJbY`&WI={6=;?_icTX&;E*d)!4j1TOu{+^tI8ku4<*cd(qOQl~JhpZD_Y?X&QpIC*Pc!1~@0ZF?Nt`lWsu!Fl z--uuC9bMR&$(ezZLgmr(iECOaCjKkPklu1Gk+pm)XfoTnT56wNOR$a-Tjz#RK~EZ4 zWi}82=j36dPFa_7X0w0Knss1&TtJ zAQ`}QND~__8~XMYpqZ-26zD012U2A71r0;x5BgkwPgTRN-Ye1o+!k*%@crCe={UN_RS9SRIR3>8c_zgERV<5v+&3#HWaxZK|(jS|NE8EBv zJLq9*m){Og1z7r#3gNCpg-5^4J;bg6zl4`MeAGOgxcL8qex_SO(z4a|Uf7A1e3>IjeBMa>r(0i0UB8EyME}2?=D+oLL-gZN;KhLUxP!)5B zzUgac*wn9_8?@K@9P^%8`@HgB6At~+6B2a)Rkw;Qd+RYXA0p%T-qi%oKw>#Re{w*ggK*!d+pc3kau?!j{EEpcv_za`}g6~ihHvb>+-%oi}xUh1qLevir$#& zAkrwO;<0$%*)Cl7Ma1*h?MQOdFP(#su#5!OZsVr>;B4G2+m9%dbs>$J#QJeoxMj zQC?+f>`9x8vo9>kOG0QH+S}e+0oJ{|!wIJALr+Om7ggRPSwU;>oHEKtNuvtWW1j*} z_ki}Daj$mh6Ev%;GyRFLi*W(>BUK0Ed<`9yejf1!Bo3Z(^mGI5v-sWCsE*puktfeZ z{EmbDjf$PQctW#zivBMPfAxh8%PY>2uR>%0asQlb>Y!v5m#CSS*czG-H~8Vg;fNjd z#Y75ojWoR^Xpw-VH%zGjyr6WTxT?BJAs#Nx?-|^CV1wKCL`Kut*qDmY9`GRx!KVOQ zfG(07`~%6%({4Ll+(<|MX$9t^F-|U_@dp@D8TnxtF@T&D9z10ZNagvKo3D+8&7D_2 zzusDmc9|ETDSJ!+T3m&3k7$*{5A$Jp^LhpHAzL7^Md0_`t+WC%l|GPWdcsK=Xrt~? zLd04GtLK?CQ?LUA<^`NUdUt4V=CA*8d$wo`nmZl7?FS{{Z9pocRH}f769#bKabV&L zkb_%<05Vk#fFZF|yh#FLzUfl_din_)CP48#oUE>L0n-byfh~aU5_qaWZFtysy=wye z^T6_UPPx0BDD~T+4uxt00=W;S%tK^z>qlU@OS88^=@ex6k^6_fh7fo;52mRh$!-M z6~3e-*L=e+zVbQ0K#~i(m`j7{EGDeX;3{@3P<1}_CDwAm%27b*xC`#3!{FjBS{6Hv zEr@HqlX~bY5t=$;RUlC!&}|rVv!=GmJfaNmU~kqM3F#504BxPVdS7_``T3Up1Z!Z) z!(Fp3K~#^z&hUz)j%~(dTaIZ5hrYh@E%MlJ489`a7~QkR3}1t{JATSP z;a7QgGzWQCjVrxU8*9mu1diQZn+dF!p3hQWP)&sjDX+X9Udom*pln>ghe1Z{Ok_!H zs9W{=k_Ef-464>t?U^q-UY=jq(-yZclrq&wHYCe=5WHHQPBE;ITS;6chs1a$OF4dm z(DXAEysB@!@Z@w6**OkOSPhb(OV+VmpVGG`n_(u~_7P8W<>v!eBEzcbpHxd&jo;&B z)!bC45&%Zg` z;vg8hPYS=VWOb(`(eQtQzx$O2J|w`0Ul#O; z!Xlgrz~sRKYC>kf=Pq?sR#8;gq-j-> zdC)3BQ34>|PFZeE9?i@!d=!rXwzCJ)`%+wbt{DlC@ZNs;WAX_-e72Y3u6o))JH+E! zwFEyDQ0@V!rc51<2et#|Nq|6oGO%@8mj_i)eI%!-43dFHP8>XfGe4|il%&juO5BY@ zwCi3p(W&7+@!P!KK3MYksLtHZjS&sA2_uqN&Q8K^5$UH2O{Ed%`zY(}9MBsIHDT7$ zUkvdn6_h{m5D^1nr~EVkP|oKG!Dd57QHLADl|H8&yEqSqDSTfB-5DqX%OSCuI=jAv z1=&2l4KmsrWo-kvk6u;;R;gPyQt$2G092Dn_rYf~Gis;!PL~|Lbc%O?^*}Q`nv zHfA>Z(F^P==F-$8rsx1yizVi)jfB7T@{I_PnxBb?^wv#8fO2RyLVn+)a zf;IRZg72Wobnoj!xz6$6LcGEl-F$!xf+g_r!KeXtoX32=OV9_1EOZxnkeZ@nzTB&_ z?z2c*kw^{P#MJtyx}>_{dIAO#5E@~Da&7Oy!k_`wHQUgH#N!cFj!>c-6IvdAyr~IR z^<4>#PQs=oCHJqHCgWTS%S+74#TUFGXtm?m@0lieBl0zxwjZDt;wXb6@4ktgcR8|% zvY`>p&A-7B=DhcixY6z5)AZICVYPsMQ^#_38ECm^9^7Yq?oudHvjtP@K!x(G3cT%JJjZavo;7u4>&_nTRzC&1ZZv(F3up zZt4;X9&^*Gnxwz?h|{GIVvjkyg<9_+6f?FIx_A_p43)o@IW6!wdbnB<^pqvl2s|5S zzFN2y+j`R#HjUU0V|~R^h!*^gePr(KvQodCGrD`QLUArAe^$Y_P!?!qW3Wp0W~~S+mO8)YQ*k$9zo|zN$rq)Bj@>)?^C05CZ)DU0h7De!BNRZcMm2nk?am zwQ`Hn>p4F5Lv}->*VOM(vfo|?(6Ko5KDUWZwFPi{bA8C?0sz^^A6+)&l3bTt&sn*c zY${)Pe8}g}=*2XjSBokp=a|-Z{%ks|e_^(Oy*T((a-ZcoZ91hsg>ZOkdpuNf>^V_e zzVY5s)52NVNx${^=qK~YYUwyRGjyd#HQoGd6uEx@CM*DX&5oZw?;%baakCOBzv zCo!Pe{s^n1))+j$pEs$x|8Cgz^xqYKcWo$lR&_5W3HQr7>}=;YIOuLbdGK7xgcj@^ z6B1|s`$Hd7d420aDUVU7RPl)_@$dY@+*Poq9YR5U+2^ONCv&^=A%r+P z_&aY8>A=ynyzy0?wSUdqntD`syykKgeP$G8^L&OBOHI5__A=Xk8yOn7HJ?xVOhQLP{w|XHX2a@k`Hrl{7ZjfR7+Tn^t>@IFD$AF} zjO(<|Acxz`B4hH9ahf)!5W9%1l%Sz~zjNV?g2`sIL>nQ&!}p(dGJCpT01o*)V(ZSL z+12rP^FrrxcM5G((+odaf+U-gl4m;mZFkbn$ zX)eCcTKAsk-HF;=uJHd?>1N)xra*O3cG49`+RIs2DL!tLTP`KtiTUY#wqy%fnKhvHQ0`$0HAYHK?PmzREE7T>Y{&mVH46Gc=wI`gj z?&|rm_~CnzA=RVbJO7>}XM^NsU8}E290|olEQ>bjjbys43j87)MQ$!maG{49aEGxX zWmsdSjgOA7%~96e25Z}9+9aP&#|Jq^xn*4ht$wN6Y zj^v)65go`JZqpue@^~f5(0c)&y>MDIS6><6)9`)$U;Df?POCgZyi?6_x&@w>n><+8 zx}Ur23(N7;a6+iyEB}6h#dAU-``HVwD_luNsf2&4DW^6{J5Ty_@*};ARY{gZxkx4i zuR{*$H=iGET>%f81pZ)c^a$B9FAbZVj(j;AaHeH#7rkqD={(`GT674P%%9~riJfaO z*ldgB-2BP>k$crd2GiiNNLhp3mX0-c419c&u>SAO{M8+D`C23*rqa*y&?$N*TZ9=5?+K6QR=-Re8v-J)DWO{FmBrPPgb5=ih|6q z>QSdW+jK$g&O{~=m9yc5zC03J3;vi*3Tdg8xhj%T(m#A08SLhjhnQ2VPWtCkE4G6- z3wfcF$h1)MANZ>4*~sZHYM*^14=a~V_u3L3{4X#+yB_||xRB#LlA@gh&`H4w04ltl z>brWFWuN#w&U`UrYx57OR|h_dj6C{#=CgCBbY!*?KZ!Rl^7MWa6A*0-B#)uy)tf*}3i$_mYUPMOB0R|{Z+_jT;Tum4 zQQm2=2Z?Sx3PFB_3V1%)yWDU|5~>pf0}DvpRaOezGGCZc*R0cpY`V1`uMPSm-=fZm z#ZCXzr|@*-1Aw#*{&auXgSbSGY~K4RBCo6wS1DY6&@!UUb4xN(Y&1zDaWw+cn0eQe z(bDybvtCB!?R-}Nzs>N!axQ9%G!Nf0lZa+S0DK!rTxFCzI7>AU-{3#2h`MEC`f550 z7+F(5PvetETw;uN@W-R{f|`AOGD7u3i#CA^8D=wlS3wlTV}w!;Zv>_AuC@x&G=3q z1XprMAjTde;>t;~t(38pVCIG9A~pK~lv-QuA02*K$qX8cE|xyd%9WoYxeaY2;1C5T zejg8k$hiKsYm2Q4?(Tifs~ueS1$sO&%}_7lOW2>MQq!nK0!kXZ`8=G*bi|}Kg_FED>$Bhp}l?7gEfuM zC5+gp$TEf$KeUKdb6C&Q_MTJCrGa45gCSM=Qbn(rw()|3(Zzpha!nt@McJ4sT_f*| zNlS)O^&c6;=&e~>%^S<&=NtXEP_ByJQZ)Jld#0@#VB%0H^y&1q5N8)&rAeK)e?cT4 zdiO7G3~19n(D0OdQ5$&kKMYCHYyaTY*iA8E$8r6#WS58f%D-^?ub)k@dGyoxyny+a zah!E%Zt=T7YM_;Bsaq7Q4Fib+SA%)Wn_@`8{7gli!Gv*?;3S}z%B_z zIqTA14E;e2F?R695y!peC{(@Mo4hbotCDryN}jaIAU4WnK&9{n;CbN79_9HFB!@JG zqg(Z$93!baK%K%yw2=^_V^0yI;X-;mW8zB_Eg;{X<%1n*gC!*-X(N<+JHmT ze^FBUYe)}40ox!(NFZ!M)=*s#f2s&_q%2RRqZG0NhQ5@ava%a9=k0^?%qeaIlnA?2D@8d9kHDicr5*hU_|=m8SPYbG8tL> z(!6o-nCYG$YlRz7jHLjSozmb@hv`#A5_Q`%lC4VRgP8ME#}g*0D2W+i%C2=87Duvl zp2-2Lj4;oDh(Fc7lvl*Uo?xrXDaP3qZpSN?*tTh^vZ|KKi!ihHzorL`*4&J-Pt{C~ zN|j07W_dj`vz?|0{Mre8EUwA$}F1kxE^x4YjREcXQ(~cF&}ub<`)B1F)(y*+OBhJzIPt; zP12dej zxDh?pg~-#?Asd~lG;NDu^vjVrtD3-2M~T26^i7radAW+%w6m&^!|DLFV!Wxh)b35! zY)B>5emQ)KOYuGDL;yq(XfXeG;^s|d{7vbLG8sX>cTlczYVXSDz?EKy;(C$(M{fe< zjU5uTo0n6DF(of&EHbhRJcRpTZ3GP~2YtstIg|Zova`42j5O z2O#3hNH6V-T_fX7RiMQtM{r#C=@3VJ%fut|S}&+~m#~s1BxJ@f7UEkX!v7uYKa|ao z%FJeskC=?__uwONhd6V7e^dAA7#e&O_%EN)bW|`|Y1PKk;~G0O<^k^u&+C2S`k@m9 zR;eIFnfNMwD98M)CvIewMr`Ypk0klHMvJrW*|;$6C43vVO**}+VpDz^vG5oi!?>(` zmbK;G1~-591o8I2JKvsIxc(*a1+v7B!T(1?7K8K!=t&#CpsvcheByH9)GUH|-kFk> zNGdwO&vDQrhY^V9kL3joE@%8uKx;M71$BmVnpH$*9&QGu^53N(0BHiqVbMTO1nvA2Ah0Bz-4$8?QTo--pM=<)jf zQ04nMlWNCB4CU(2w~w#dzs@0Y9U{?ZiYU8RO(GEGh{Hue^ZZb@!l-&W@;`0vxw(&v-=q2ACmIgClIaN$R&{6Z~*fa-nC z#9pP_0Sl=VlNv$XY4t}0Bk#E0%1nMc^|_U+9L`^pAWlRrkdLyB<1=3hu521yoe)>k zV<1{r zgXi&V`_Hj^xnb`z57O$v`E#mNZJv;y@)LsANAG-hCCv&mRnS3gklhg7`$H|+jc8jl z2lgg!CTm~O4xRed*Cr|>uY)ORZB9vvVg74NQdB0c-3?6M3b=~S_jz)2zGK}UZ}Y*v zjb&WQdx;DdD0e{-!XxioM6>a)GrE@qTKK06+uJhdk+dDTf5Ueru2*-qMo3;udD2m1 z-ShNfZFd9=CzLEM1hjs}Wcx7H-QgekWYQ}73$KOZm%ZCNmV?gdQaKdQz{($GkbSNf z2zw{NPvz>oM!wABEwJM>LFv=qs?j!QH~!-g7fOFw1-{K#SnKY8Jmb_l6{$7v8y9i4 zor2k>^7CU-Fx}J+MSdn#A=UC_9o~^thW1kA^Y0nm#fPkCTP?`IJR}V%{s2iYDQd&b zJoSSEPvM%{3$S=Xq|hL`sQ^+MkvwP06u(G$TD9aSy-E<~t8TqiCPN@`-d30OryM_r zxcgsCD>!G!Qk3aYQ-oS%LN)7QKNAW7OGwLhb{md-&M7)#$7?U53vP+RB8=`h>1#6k z)2qetaoRzJYWYI>?x>K%=K$ekwo6OVqR;?>hSOGAllyFo1~q_5nKadBU`-ySu1e`Y z{L3e%sF5C4U~1VyCxAR3@auFWA3MN3WO`1+%t6Q4nQP09PI5O)5V!#q7+Ly#E!%a} z!2}G*=!1uUNrVc9h${v4n_wUbwsUTPS-XM^b>VAMLkPgmvlOJMol(@oK z`Q{}?#xBZhRkrE{QsZNsJht|gZws130{GSYTLEsX`BNRBiZPR!O|Sy2aufgJ5#qQl z1NlIYvd%!;;O0!~r>)A9@i+*TjrV*5xAMWJkG}!3rP{^6M;lQTeOWPQMdHTL@W(8n zsk|uKsp5|m7Y_Xdp7nU)zU^QZ89-PcwO1D}OYmFW5ty}pJd5&F9iJ+C1QiIIE?|PQ z>eST6>3R>TTmw2xvfj-A$8I4*2s)~?)Mqizu=HRud(W~zpK8`-9blUaXJnsofR^|F{p_NTDfzG@S zV;C!Cg!iq7PCkOA9!Xtxzc8j6T*xXF&DEM4qL&z z`~Kb#y74=Bs&d$4t~_5ejO>@BErTJCZrA`#d@KchfQ818x%pV@%;d81esl9C(%#@&8B6|5UQgt2(m$08sk zeXWTNOi}Uu-&^u-%9z`S9n`;Pt1LZmHf;{W;OV@E5BS$zvUMHqZ2qj)Mb>BCho`@+ z>1Jbxm3oH0{-0pM#`5{i;`NUf_WvUwe|w9J_-z>LkEDouKz~O)2^I}BWw`}Vh%;7h zRjGW@Ls2oPNN*Hp5H}zy4pdd=Z~6(4;3!41G}Qy#$&AdOT0%wjRr$cHK)dIxp1qIJ z>P`UDIE%RFaaJqT?Uyb#m2!?^>H#8~kFlEuZBImbU}u0(QOve+Cw>0&2zbXE@aYI}N#O{W6>_aoWM@1Iz8S z5P@u?8%Gi*wmx^Z$4tK{7+V;-ye#p8%Qo9FOye$N3SDM_)oA_l`7wB>VXfDEZL zm&4i6&^387Y(wQfSm7KSS@aRaz6{|FlF|i-CZGWA3(%dG3+5>9@+|j!ka?j;Nn-24 zBjE?f5pUj2LA5Jf;NsjXOEg}CwOuqZ3SM)ku*-iQzDd2ircn<6#XoGd@7P+Mmb~C1 zVf>=R>P)8}P z@g3gw@ZTS)q={v-AA9{0-I6Lbh}vv0`(Q2maza?yCYU^q_Sjp3VP(uJT%zV&)>yo& zvxhZ>T-$H=pr1{yzM-je=JlT7KO<<#mGG1M;tY>JI>^p%fLqEc6=fB30*ssDU0kE! z*qE@dP6&IhG~&q=bmfm7i*Be;&wmFJ;a z_l;jA>}*VC&E$t$u$}T~-k$WOR~?JF5WFsUy#q?E7kLHNNZW?i1ij?RbR%2_BP*4g z?`Ri97}rjky%jb8afZH^RynOrrq=k45inq+X5$P`R{|#l;vQz*`K(@bp&ep38q+xiX z`}8q|87{|rh~hE{#0+B6>($)@fS|g7n;gj;nd4z>mLH(1uhA=c^qDRiT)0$(>P}-go(At2AQ(F_tD4 z2B*N?zqDawHO%3IRv3We)POCOetw+Bxa!%;m)D%+``|&ORgE6MW2T)@!$l!PxNbd7 z^#=JiL=nUk9_ji9HBI%wx=9bZe2j8g{UJfHuV?1bh?~2~{aQ?!%^sN27m<@IU-m(4 zmO-tcpNmd;qPXzqo)FN#oq9@(sV=qCZ-ISsnucfK`-c9_)*-p$9K|27g=mv^p!iO! zmGxydl)_HK%KOsyX(SPK)jCn?d?OyOcEq*F!fb0@?dT``$Yf`!1^g}L@Z(Q%%SnM* zmN+Y+zH&a6{mrQa=sO#C$;gI$lYO4Bhi16`9Gh0Cgl~6GS3Z=A@X5MrL@re1Mwi5a z&=;Z0RJw48$95drCZ?gskcR8gs^cc%b%XcUCPt>F^q;p#2QUsW7xA=erF~Q8Q3$f31@`=+gD#q*R33{@Acc(Od zt9MAD?*(~2?BdK^jT|ecTd{nbLYlS%`eC4N?e=Mjra112REDMR+O0wQo_qe!(Fh?EllMwm^PcSZTlv@%A_0&=~hjjZap<26buRSki*SS}W4ER>lg~ z-9T1{T&RP;$J|0dwa%vQxkj|5aP!UNZk?k;k{deUBTM{!ir%T|4hE-hA=R9Xd|Rx@ zRa{Q6u-+p*%kmvI8tl)hKn;&mkQ??<4$8eiNKW4HQmi@4tCSX#s8=4|Oc$wnu07(q zgkL+Zdww4m^-cr_zdu4NnC`JtwdXhuY+)Od{gH?eJ0m>jjUa14xGrF%c3#kJaF9B=>-%+flic-)4nQ-4K zFjV>$p~ru=oBTNn{2w$pN-+>?+7F<(Sdp?RmN)A3>lCsD6IVH}FP}2=-cHjJW4@7} zxImVx+ua1z*Z5HSQm}wNBb|}b%DU(vJhiSp>@O8hb=WlbH5dR809v8D^$Vd!?hf{@ zAx>UpMjZ1gJp4I6@s0U3e45h{)aqBbXSs#$`H_~Qe}70S;m_5(}*XGy;QLlGCio_*eAe&Xdz|WUBz`c=hZ%c_ww2Jm2P1`;tpriVx2q zdA;d7y^yR+IUjWD`qzk{J2m@Ey&pVG_Qr15q=f)jrX$Z2o-v%~D{$JRl~&wn5*(q+ z;oiKg>cDkI9d8>pG@^=$imw)$2y8bqSx)22H)dhcjh8$Bftt#(jBuv03o6Ya^g@!* z_py&(wR@}x3#Pa@bQ+I9W;TWcLO-n{PVR*{wop#pvzxekH1sE$dFri_t9wzz;3jAC zt}^ojZz@usV*(HRzVc+ZN=>%Q;5uSfTuSfKX?9b5%A5Sv zD^FfVQ=yQn_c8~%(Pke+P($xzy1qjx*9YE>*;O@Ql@wFr>^DP@=N6NhngvI^io^aP zPCTi6{^r%f$nVEF&4UJOcTmf1mEFMy&{yM9;tJH zB{B!-w|w#!nn@@DYCjSyeng2#E#-n(JHuoUN2%7>&(J|EAgmq=}@2x2}D_- zXQ$t$4r${`iVE&a?>-jg58#E#eg(4zISarKPVzTs|e!>TC;p7kymm zS=Jil{9PcnsVnN@MEFSutlJSjxZcIsx_U{~zbR(JpU|yy3_Uqex+S}33wPpet4-K0 z3HbBYtZqso52Uj_v$VdPV2L~(*Q7!?Me+GRHUuGa!j4hynF5*<~XT0TI$Vsax*^*Yfx(N;4xm{j}v^6=WF{iqXOw?Xh}PqU3qN$7wS z*~!5FR^#wh)7CDvsYsKR3dp|D_tXLt?;rKhtYD3hD`}^F>G0l8+o2-mqm-Q$mGr9x zJ^mvltqa8BFzIC>bbFyZQ{ZBBUan7a_Nxr6RLFOl_ROgQYw1+Um<8t~__+bt!RdE% zin{84X&3MG=44=?&tNgx;VK8I4eh(Lq!8q7-IdWRPk6~_EW@paXP zp&nB0$;D_jr!a-uTvkl8 zTcYWzWTuohJ~Tn;M?W#7xG6X7kXF}A=}Peu$^gtnveTOgfSG$J#{BXxC+m3+Qn2>E z2&X-zz$P9nvSS85C=XpRtP1?){NVL)pH00(J&oxOz;DV|#}~Rt8gHmX6HUvjv9vvaF9J9i zPV1oIj@0jiBMgQ%40zfS7@I$u)=~eq2O(YT0T>HV%@z&wpD;b(5b{ZCkU!Ojxyr*C zm->COS#PH5N_g(3!Bs{ZnufQ%>DNJrl@SE^ZD)pdS}TlGJ@vavgO#dh%EBQ@d@0$D zbWg}kUAM>X4l8;z_svf>H>TszL9^OC$vPU$?eQF~MXKsFTzW9Xs|(=eK7chzHe+}y zov=@*T?N%_EaYhxzVRU>)~5{gs6-US63w>)UkF)i+IJO27VLJqlu5YR7P^isDTf^? zS%GP@Sdm#)5~7iDf@_ps0{-I0z?tr=eati$tNnoB(08kvX!VlHF4zGnnk2yu<=+#j zG|cg4#BLUVy!g0=j=E?r7T6R-zW>na1}R4k@US55B1tjub7LzkIyQ#6u)QT^JZa~9hiU{z!Xukcm zL36{-(M_D%!=xQH@4HxGABxsF6NZbZMeY9`rnt-rs$j-HVfuL?htp!$;%aAHG=FVwy>?4!qfFjk zyAxz*MnwB}g}6<#SP9eh(&zdnSRcMX_;1z+OVkOBthV-_y<4s8%D)3tb};bosHXK$ zP|x-+ON&#~gRJSCYsl%{Cv96fm)8Dzq30nN$1i;+%4=FY%oAL>Rn}CeIM)0uamfYk zi5Cu0hALr-J{N}_UGX1G>9||(YsBvwse)Oxy|q#>{X3xyN;x=<9n&6#?h7{w2%!Q4 ztu8R{C^Jv2nlbE8o_iP`baCtvkjyWq6AX^N%R9jAI4@XCp!6H*CYzep%WM$Y;nYkr zai#Uew>FymBijyC<%p(x>q^hjMh5;xR6fnLcRAG9SZt(4&KVY`6~|Fa!+W~hv88|M z1335Lc|B#X&sQcP4=USiNC0?^%Zw-PAw7nT zRxm%yzx~l7JDv&&r7bLV(4zy~%#Y9#mF!s#yrfIwzpH5(=V4Fol1QeuwlZ3K?tjYF zvvno>7RS^~iWI(Afn)r!h#c_kFRGiei~E&vR+02K z47DU&e!0xYhm>sRw#2-?NTJdFXoBRKqj64Sc0d6X50@SWsXs!hz>%@3feuVqyqXly z#$TSY_eY_n8X03;HFL|PznD*1XVzGGdwhtS)y7%g#C5?>$zj)F9Y8#0sL&n0v8r*1 zv^`tt+wu*Ur=~r<>B}L0xr4xSC?n~chT5vk zN3f)ep}N=GKc*C?b5gYx_UB~)F7*I>p4EgJgm8Bwu{c{3m!+&P~rLG zg+l)8$tqG3`l|+p`S(Ugn;mizgDtMe~8uN5f0-0)l#-QvzLTpPG^R) zdKuq_G_tMSOyQfGDHs6DP1(FgD`$U{(4l+E9%9szq11okv~1;I#6PlrH^ads(P2}W zuj}yOwvKB|f(a3v9KE$BeWvu=wA$$0Ce&w&TYRSpC7*GAf2tfl{jH&^uH^qG0kC2i z8&UW%wi{~Gt-(4)fv*rCQp)1d;;M`E>XgZp%|LYuUEmdtZyagJN&s+ZL*Aq%O1qB% zK$EOvxgcgyOQrpUL=mHMMz=Z=)%{gV^eNy2)_}#J4apSuBIHeEFDQv5ZS|L3*q&k+ zF6Iay>xvu2B87{HU{PQ?@El-`_Ou!)ej~snN)9t|F(q8;Y@nej>0%PK99;wOo^l`GaZ0Z87q zT&H1YOyssxG3ZJ0R4Ik*TgfELnTG(;1M`tQCce|uVO(VozY(}yP;a>TJ+mb`_Tm|D z%~DkB*S}1_Mu({z7Ob1|)2Zy(p{G-6NX*Ul;xw|H#1_y+t>?8LALv>fv>c;AFPkZ9 zGq(0QUd-RQQ>RO?kKT*Qvo!4dXE8xmPOdIXhnB7NDlM(-+Xs51sIZtYyXY)f>y#dJ zM5r!W%9(?{#csM`Etb^MRS<-|+e2c}a^^6=+a=y^svHasyj6U~tT{E=%dKi5H+}W) z+T%&WwbtjwZCG#;T8`hkr}!M$OJdzhR9wsY^l&2Vn-!L|n2ji+xSM9tcaK7AAd{J> zPLtQ=<%Ta3KN0efo+vii^D^?Xyaz*lU(4Agdd$lNq;RSRCBH!=tN?|j57jI-FsGez zxED@ozEts?YN}ja8GSkN-Cj*sgp)B&Og1`tLshU9-5N~6)L`Vn`n;j_&_*`X;OkV^ zwMUf;&8~Ug4VkvZ{A)`*YrX3SO$!53u0ltGB(11M@-0+~*69eZcH~e2Xl@PSi|w(9 zTl?L=_|QD6(YdZ><%$Vv^-oTa>wP}Bx*7kDpiI-dfm(-L`1BjN_l2X1hp)}p z>JK$cbbufJGb2C#8^m5AUJsT!Q1Y!x-(WK(^ZPK@X6p0R{MPR;s|V#=>>FUZ1)6Qg zvQ3^z9sbiS?^!@t#!?4$f7hQa23Dqq;y5!Y_thBA#yDP6*owp;EHB~Rn|!m#DhTcE zv&Ov8v!TaBV`*;$)G9QSTc+N9gX zs1mFRT}K1qL*2W?&f#MUx32_XqTc+cr|OKfkPAt^J?E|IL^#hwc8ItPhd@trl`@$Ut&#jy(KHN6h{n>}xOf29OL>?*mCP=b||1kyv6CG~+ulYobk#_;_*(0++6KQiH| z^p@pQf{TaUrJAJ405p-S8^1$c^eYbJ;iq(MVs( zMTqOEDV;JM3H|_nM6~#6e@X1>o1y>+fS$;r330rtGR)&^+9XP!3d#Uo0d5>`zPyOM z(l$M)v>xsLuN5swC;DO5!(gXScme+lfTW`+E~*CHw?dR=!il2+mZRbh`}Doq*@e8_ zK)RYRXTzO|)cLECK@<6aVA32MEy>9<{Hf!z8ZTlq6g*W>st*B!&_SmS&XwMDDriF5ye#XGb zlb-!UX-zJlPRYK@O`k^>4URc&HjC*Ht^zlTL@^eN70X-a^w8sG>OaMc`aMu_tqBqf zg&vc19bLS;nWC=_i8|*6^0o`lij#c`%0=0RrS~&uQn%_`DU&~F8A6j;5H|X)bl0!HM{q5iDCK6Gx+O?ciY*f*A0w<;YMEfPoSF}y?OlAwtL(C z4t3e@oo&x)UY`B-NW^~YVZDJE*(zl-e|sUyh-rs+FWd2V2x?lndq|lquYaxu|A@Kw ze)iT+=tP75APUwR_)V-g2BLA1blQCqyG=|$I4_LZP~O1MpREOX4_K&o4AdP9HXe@0 ze0I(iw>xTmLNa!p)MK11jHIo2p`iQ?GQWx~v3Z-@!RT{EsaJ)rq?ws9WmQ$@ zkR@8tVnV(4EA*PitJKOUui@0w@dA(Arv`K`r5U!Qmxl`2%BCg94#WiRPqkm`rEn{Q ze#g1vw)hSHhL?O2#qi z={r$q8{CNH1SByErn~!+j-BpwGI9A)IPR^VQnba>0~0O7qO=76fhib;V7Uv?$^W5o8aUIxCe;x%(kPycw19( zI+l3G{8_uKEuXmB_{s?ob%JN!c@RR$n!~u)yD5oSUcj4dKoW!IAw1=}QNc;>%O69! zcZ}ux7@Sp0Hpk!20ZSo`EGu$WG5Q(x6$QWZP=HlxDXZULq1YzI zVHev-azRxkJ|o{PwtysYW#I%oRmixQd}+$B2jFUR?HtqEXjn9wU7l|J)o4NcvU0X-^C=_x<2ZNHKxo}qd=1UjAfsd?54 zO#$vr6vkVRpI4m_Rmlq(c0%y|;TWG>a5&pt9`eCPHet0m!Wx!};x($RQN_dR`kcq( z;)S2FyJ~?oa{j?~QEU0E#T=TIzk0Zn!b!d2TVANh0u?<_GIvKom|yr z-c%>nxsQu?&g+1th(A5P7cL%=#2^nNXT3hn^_<^v*)DaEj1hy&ktMQ-7)|nM z?%9y4DN_=h;`n|mg%Mt-ycrGNzXCMFqS}MNg>N{}D?<(Ba7NPLq?l(;zf%V{JmQA4 zg^T`JWkor@uamyD?_OU}>&J9~6)gv+^iS?v0RhicQtc9jmTaKf`Oxu`fAqjOl>G~VbIb|x~2;F*!k z>W0<$kAT8sgAxdpNtwakX8vy)_!-j1gfw$#6UU`j*L?~#114SJtZjIHmi?gT9%pl~ zx8XTs5svZQO>i?45{@4*jWECuPXtEgn(Jjr1Rjw8w{|~u15a4xsUAY8?$oil( zHGvioV}!9IFBb#wi)DFW8%b46jMcS()FinZh_7$ zo645;xKfk8Wyh%-&h?f-%sqI{j;-jgfE=)^nR@En`;;R=p;Y8cOV=JLp;i}FpCQF9uo($twC7yhuO zG>bag;D-9>6Q3RD5z2ZUOdC&U4Ejt=rpG^LFhL8rdQUYC1`B&R-^lQ*X`JVP0XD_N+G-DPM1$Ut)cB zJQcd3d02dAwi=aaeAzWpu?)0HO`6HzEedF1RZrEdju)lZpe_q|=CO6ApLSSBs)^q< zXJmhiLZ0{M_gU*aAa|C}rTxqPG*K>^sP$93+CwVq3)TA}-ZrDjo1Zs`4ud1Ln<&f) zUQSU`YhW-{qT41t?1~=B59)IwEYLM(9HR4f?BKuEJ~q$)ZTV}SdGFc%2Lb=f;uU5fr9GFM({K4+bl5JvwUk9(8X&RJ>I=*b2By&WO0YC2^ z)5&3ygRfog$V2ks5?*E4afCi|^keOP;J1FgjKnhmXi)cq#D0_mU1uTFeR<^8Uid`1 zQ7B$=T578hgqk738v)nL0O9~umqaXpzS879kaLS+td&S1-mV+k~20ZN|j=ZNBbEjtbldh;*wFzgTr-Z6}B<1oU@YYK?LbxZZ z9jV=WJjXi!ol+ez_&48+%(oqBpejqgKiNEibsLR;6eqG|9M)1*#r|^(V1(;>k9d@k z%D*z6$Z-va?K7znH;jO)K1`sOxg}|r>efU1P;+b~SifNPVCZTh^f>z@fy48s>`YeV zL_syBXQN-dEdtP-H?XXJk*)1)J_hO?`RhT7`gK76MT<^`fLu~uBz$GOx#JjK#GTyg zF7m4^Z5*V&hsTtpk&^SoAc+ydbZ>nS++y@y+I=CBkNVYpzvNQcXmBsltK&Xm=%BW33kb$Rj>! zb>GHaa$|AI{gJ(f=&Hv#{W;1P;wM-q`=JK>LH`uk_KOP;9c?i|jdDX74Ib?J3w7)f zTWlr*_!~vKtmxXl?Nda@{m$8GmN55ZR-CoE+cKrXBH&&8=e6mt_y5@UHFvmf4imyBDMZ+OH=%fgJKb=Cz;vA>OsKGyk$$(YRDc5QdA)mb+ym{54^mf!@=kXj^*7%b)wxG?ovTzMDm|O{ z{=H89r*p2;r6wV3@XVG(O@Bjg9P5?U^3I>PLasyvzn580=_pHCv8;P=u+cFPwvl@H z)|x+kZ+ifB>;IbrJo|SJaCQB1#ee1iIrNf%T52tzzl1#-2n97NyFFk!&YA|^12{>N z;rApj0?j#}(Q?`VH;l`G1ps^+Kx)-6A0P!V0?0Z>GhDle>Ex_J^2pOAsu!=11TE4< zU=WCQ;B)EY%|42xaWkAnRO%5NCrTs2S*`zDOMT)!!zR_ zMyGJ%pj+R}FV@)@X2*ce0DML|Mx0f(wNotD*Up%VyF6Adp1gLiI6w6J9^!1~GsL-7&oQb2j=_?B6lBZyW^50oXHycmP|j1~dj&yb&^<{YW>teEY{3C-=F2 z20@w^YH_`~M_^qoB5Up82s=H-w7wg%*C%~xaUPC|Y_ z@#)omkaw^Vriv9v_U-vOer3w^eg=MVWP$E18+XCFi=pZ;Z(Xpbi-&5@VOHxKk0eLimakilWhhzShtm|19kiOU~AXQJq(!pU# zuDXkl=WHt%p2cFK$e~o zAZR(h>=8R{iOMDd!R5O_@tY(^2$z+p*>D0RK#36|Vdd-QY z0n!tg2I;)LJRB<%UJ7PbXnMI$dFCmk8Qg9E-MiO-Qsxj);4yf$W0S%kPIeGhJYZTX zSI4^)l@v^ub6a^D&44FfB!Zn^`hUGTASRkaE|ePoC|>+6lIUU} zSy=n1zAIQ%r!Pjc+M`~?$Z#_&l6>8Q+eDjP1_Eux3kCcM#ol+>Y&g5Yc!XDD8rYa1 zYL%^Zp1e3cyf`NZe*7(ctK<2)v`(qTdX}ZwVsADS??C8)hsY zeE<4$c|zm_Dv&o))MQxy&Qi$UaG#?3^|Hdote5WJXHN>_J&5yy*8Vun!1+EKn(w9? zWIu{$uR^oW@AA_7rX7hsqwU_o?_}Y{gy?g;+A#~iWDc4Za%|Rs?=Sk_k~t8L{NwRvZq#m8 z)4692!=DtES?xin)u4%Ms#=c*%x-oXof?ObHkxC++{>{PUR!g|QrQ0L?5Kv)+%|xI zCfq~+|C0*e`d2Fay3XVMf26{<+3^W=Y;}N7%9OX%jb=R&T)Qy4#_$^COWzN~tXgZb-@ue%0G#zS2GK`QzF>DK3D14Ir<~y$#>jSA(k0pm zlI+}kW%Vf+AhEs;j@)`%UNnjM$MLCk5zr+CJkIcar6by>EzTMd$B2VkLfe+q0GA3m zztY5b1*`QG@SS0Sz!{3dPjTxkHhO8`07hjHl;C=6OXVwjhNG%8U9)KVw)r4MAJKM~ z+3DC|`7iusi!^~mtDzR1prf6y*7Go4U>)#EH1*UPHv^$DocrFV+iKN0?lOp&U^!xN z4D}MkXA*!3l(3(7-4u9rBo0+Yj{tGqAcC|F9yif4M`3!EV`$ZTrB8YP+;w_GaNq+D znoF|#-Qu|hI{7>4>yKa&^o?UM6pS(X* ziR-gAN!;)_!%+8cXZ=y#JC@9i@agE&ZZkSV2h=|iyRj|{E?DANZbPD}-B8B`mt@55 z(?9M+tRiYsdC)}L&}qv-HV)b(&A9OGrH6T^CgEfCA0yrXo+7w$ugH$6lWPa7lIPy~ z*5#mLEASC+jHr)Yc(;4*MdAo$DXM5-<*X+stibE7IA>OTuxE_j;|k4qS(Gc^{;M}` z_v2MG_L_?*!Q* zCzB# z1~%gqOHG!Rk6j*@$VVA02AYI^%)Tp}g_>YeT~ymtoG`?(Kl!F#`NQMCbAn_%JN|mzo4Pl( zOq40%VJ9}ntJO}Qo-v59fiMh~bsb;m*vFaXf#x>zwkSG13_S*l#{ZKoHd|2aQUPBY!?tcXO zKkgao?b;Be{JN<|C`khqy0A`fuY6KOx*e~ido6I~*}WF^wPT;}SQEpP&Ow5|cE21_ zU1BOtG~TnsisAm4}S-IxP0E&^o9PTyV8UE+P7{x_gpvf_N zTaDILDk}=>IqH;e(vIo89^Ut@`5__g{;)Ban#NaWMQ+o&GyHUuswX!&QB9wJIQ5{J zDtj2mxFd@-5q;Eg2fd128=hAqP-KRCA>*euSeJk^_#-!Sy(ZaS2YxVi_tA#pLH3K& zmpvk<<1H5I8HI;7XS7n=K%7#O!XX1TS(3`HsdIohS{n)VwJ81&0d68q$B?b&j*B8k zw;!7!#1u*Wtz+kA_85hUz(N;W}0&EQq(#8X10`(tE4BQKHCnPsFLgDM>#qX>RSQ z`DjpTYiF9j^nRQ5igm=?kJ_Jd@t+SyW-svXg28GEJ;PFJ*W<@kDg-9~&;7t>NY3QD zqqY}S%axazaCH6jrW1ZHT90qKkKMxXvxFhjo@!hN;4WBpDPY*2@*I>}iOQIrM?+ zSX+Rj44|K5q!Q@u{K@_1aTNjJMV{w6PHMq-U0f{9aVRcCQC7Vg z3Z0TIPV+lVh>21)&>t+lFfM+`;x_AP0_XPS zEop&q)(59r&^d5Y!BD@@5(wQS(AWtO)cji(tE?6a*!ih<9GnCZVquPweg>Y4bI8-` z*|R=IDN>nc;$>OA1tzc62uE#UQ?*^_xtAO5L}!m4vzN?J z{tiAB)n4g>EZoc`YtmBrNs)qzS+O|Lig-Cu8Mx&DODLSfi*J%V#P{-a`*%e51v#s_ z+qvsRA-nw2H06Z^F^MsCq_*qC?TttS2Z)ai5q3A6m{8e&DH0YsAMFJxR$Ii~wnH0l z8A&%l1J1NJQMzmOAUha)w&zm_oYLRGey6+KM^*2WAMu0Ff#w=>OZNy@#1DJhDlH|% zAo8)I_)wdVPyz7+L4ql@@F{gbZ zH@eDSi8Soo-bE041!)7yMv9H;^E&L_w1C3%9l1vw&6jQcVEx|5_F=zs3cMx;;(gRh ztu$Rq3tv?p8AE4?60No0##&yFCGrlT4ek4q9MHHz>FIcG6I)gNg7^?o#DwDCEM}8* zdg(Z8WKW8OWL%qWsaEdWs4>YyY$b;*s`TF1iXk<}|K7S`D*8CLL6BJs<)N9fJeGSp zF|Zrkp|+l*q{LkKW;iWxU=9Fnn2u7VCywR2K5zC zRJO)=_@K&iT`a`X>G;8nL&_u z^=>iY80CMaJB0`%?>KohM+whmA98-@oORL6&LX&d4;J%s;vIS9(Zf7LmqDg$HKu*h z0bF0EyFwH}GYW!rfkE(_z!nNAC=8>6ysQR%Jo18G+7iSR(g)9#pL(FV`dvtseCf(E zpI~Rj_m@rNlH~jkfYvql*FH^a#39H!Aw#`1qIsN$d>lOGHMW-j;bkOkW&Ns$k(<1c zU@R5`4;HR*t9bxLIwCq@aQ+|8t)9VA?{Dy4q~dMS`SJG~-zBow@Ez%TZ1HzNc#g7EaoyRXh`m;KIQQ z5oVmjw*M%+G2Q;e6_mI_+Ti2Ka~t45{Oi1{$d&wit*bayhxbG!_q4;>GYJ)tdi*0O zZRUHq^_>U>&kd{K*|23rhc`WG0}>Ro!DLRU$2Wq~k>DlyfPyf`jp+t6hrpukX#+e= zt(`pB(x-ctkJ!zKkT{uoA#$NNI3-&8K$I-YXgi>DW2y0Ef2wNO+57qN^Y@C38b`1v z>9^@LG{_-my6C-x^zJ#8%fqTbsP6c?0wGITP+PW*DO z$&H+}8XY!`YNML{sfy)36DP^1X@^*nq?+t=>zJLJO_bp$(7t;eU-rpIXrcepTv~@eelFE-*`CM%X1wi4RuGh3^zJ?-hHUgW;eaLi;XU9^J$G zQISW32j*()bsJw;WIPjfBA2xisCu;bL;us{OPc(KJ=12<-)m#|K#l;l`_8b24Zb!jwRk(2i>ZanvXvIB0LQ_s z+GRyu6Ram&eV@ofT7^5e27BQ7Pe4o-sw=OaOEOZYR(c4Rs`i`>Ch_ej2Xn1^3T}fh zpMSO~wwQk;vEpfP?Wl{Fhbmv`6ffuiI0hb%C(pHUGBc^;W#4n6nq+?Hp-^u2iYw+mTIe@tU{7tM*fd$MnVpT!b zL7gFOH*8;hLbLh7a&$vts%tI>Y*>@mb%@7#yw!tdDMKhPnu9n~y$o=FFs6c^n!ws!yaBw2JWWypHK zr<;Ofb{IT+HKMQH$rVmCbwy1I-F-I`if!JWYUZ^wsQH8&4nx*7bW({eupe0C%(cf-F&y(bF+p}8w1*D9(nxGBnQOnVkOfD@SI5*K2Z14qYm--!Zt7xt96?s=hksbHm@9u5dG( z#V#u(rFoNx9W(|C%bKvaI`T+uD4>5 zA|DW6cpSo>uy;`g%ThjUt$5V_-V&hY?pvu-2A~ zwxji(sbD@Dh=O++#!aZ9L|b zKhVQ;{i5OqVk|~xBXeRdO>HA|oqp_p-}EFw|Aspc)5k{1{|@)xI&^IpSV9unl8spO zy6A`@W3^j>GfwOPdO@I5LzxY^34&qHC{61kh~A?CL;*UI*Q9KKj6gfoyWIU%dYfvS z+hZl8!gn-UQ<;1H`L8EI9XWipguA$D0D{f+zbTS9upx_-Mlywh9bb^Kjn;DygwEsX z7P6k60#(~I+gMdab*0Lpz1gFXYRB&unK!Ges~PKo$huFp>}N%;2lu4Uz+b}_A7;+P zs-$X$@;($gTH#de^xiUle(q#A#F&)f;VGl2Lm;uDL>*cu1XLpIxHLeWqhr$Y-mNG5 z>-n)VmIeH-;&G~LacOI1!0*QWDi#L-i!%;mCkz_k43s*D>_5|Waa(h1f5KoE6+nv> zw4_&N5(R#z+r&tPrg~M|06sv$zy8K|Kshsw8HG;J>8NEkC8;y1n^1@=Ow z{~^0@u|#!88k0P)j_37FLw~Y?zuuWUdjNIx%TpFxboH0}IEP)(Z2wb~S$}(hZru16 z@sG*J+A%g=2k>O#ydVpDioxx)BUi`YBJ0bvBv3bwDEdWKLpTqrWky3LOTgfgog{1a$Q zAU*P?>mA{rS!E$_6zM)mOCzuUdjgi$Uh0w~ns=sIUU`_wH(gHDIQls_t-aYjtYrHa ziPB*KxeWv6D4H9~68yL{(aZSdBG3Ag%6fl0RWWy0sNTY!PocuRxByyrvaV!! zb2}1MBa!Sm5vSNa5Mvgqn;n428F%i+4wi#wr_b=fMg%EC=U%``6PF_4b1^P=dcs>k zxHFl%x1P9&8!4974j_9K{=kYwk3WZzNRH{K5e-=fF^8Ju6TQH-wD$opbEU~zpXL7Y zz;FAMY^uc&q)`EazDCPslmC>LJB|g*{C^Gnv;PA9*LKeTxfb|73&Ze$Gm-PT&h^d^ zT*_12W}6^c3=wr7$FyeRCPRp{^^^e!WHCR1H1O&&2)h3%YDRUhnXV2}Q9`HrfUW z#&Ky38&JMV^982o_O=H#zII0DBQU-g?eI}(gowcjU4HNLRN#dPHs2@aigT#lpJ4o@ zgffkzVi<;C$LAU6g_o+zVi*TCslR}BK5M(}kjcG8uW9?hA(&R}eAYzt*>CU0xB@r= zOTa!VTW)SbO@O6pSlxVjOE%cPm({`Ll@Gc>px~b#g|pmiHSNh`&2OuU*u#zjU%9h9 zLu_|)OqtU=et%r?p`JU%yoDyod_R}o$+p_4{%iS!|GBB#=?*<}i{Aht^^g>j_yYak zC^rsFE;xSKI=-}ubMBH`T@BUfuP4ddwHI_)TcY{9 z@XZyDg#>EtPZn78j;RYq>vub~ehH=&f3WlPoIcB!Ge{T~AURrBgBS5p0m;hJ?LFT8 zQJ%z>Sl!&sRsL{t@%m{8M82^`jp)G#_qeneHXbue5qUH}{jQ>Na>=aTIqjD%wJkeB zgr=};#8>-hKN`9-65sG$1Jbyy7rAu3zFTI(bc3U?1DfC0ai(1-afF7)Y-u@7)p+j%cF!IK!eX{3tiV*O{}uEY{e&MzKLGqo+GPx73?)wO%mBP$ z9gbpK-5cn#eJ){Z>dC0n_hO4>%8I)0#k>_IqK+-{Ekl(o-Is^a1%y=a6)jlXEeB+d zXffs=XC!Oy!bMHo|O^0 z%Xy4%=z@Q-gu*i!xas3~<|+yS>P58u&QFMFYvczZ&Cq+3>@Q@zPX5^Xhy|9Jk(6Eu z;TXf-DXEj$djpqVxG84X@;7dcw$20s5J41^c)|$K5`=TNPF6h?XO9Rlr8He8pR+OV z3C2ABwy{PjwZOF21HznFK)qkn#*1fOui`CmiDxSvk;kpb;NE@WJI}_IvH^!L*nn01 zipoivwmcvK7>R75ax?%K*E?0JAZi}f(-ilpIaZm66NIqkH z|68q1qe?M{e(21UG}J-sAgo3h&CeNOX7C|cRQ3?QsngqQ5k}bc1H;a{e~_Buo`8n}2(}T=6RH=6IB66q9GtkA)`)zP6jE-3NPJt?RSdFMp+@rX)^K z#?3ESM8VxXJ+_`T*~(q&<F%Qji( z&SV0kgsoL|oZWpl-8Ra^DMn>wxUMkDk-7LnI$_$j^K{tNGjO%we#$qP4k(f%!pSx< zs+*YRso_4~&QCB&t9;uNJSxk`n;A%oaeu%?bHUm-T+_!(Y{x^k$x$Z@*^(65I&0@x zu^U@6doMWE-V94@6hAV=&K2K2J0`05*Cd;+iHSEyuvU5miVZGmGHc!7*_X4r=jNWW^j(*9clak8KEEb>89am^dX#1~QE?9SPOQH_@o!GOx!&@s$a zq!=&lhw1+&G+ZwqkiON74!?Gz9DQHBjZbv25Q0sSyY zW@)4u_3@X&ZWtxKrvI6_d{#M~sb$&sBEl|D6G%BjDkHiX7$3D0IC1_-PFq7(5dnxYAc4)ZP4m;=rsU5 zk);AK=W>)WmXyv}lavm~1muvMmYIp_iackQt;R-Cd}{^|Mj*kYgw%4)`u;l; zUdOr-h?S7FbZ98?$=PI~6eFPB&TR}6Ec*iH!Y7YMW4Cn(uVoTJ*huK3V92;Pig0dC zUR-%|e(u;n*f2#aU3Rg+=}Cj>89`hul)=yC!R+<2&=|$zf|!1V$&%YUU#Uw9O&(j) zvN169t`#$Hmr4z?W3{4c7UMuMiZi`Ju-sygtWom}_Fe=slJUJI#nS)V8DxGh%KRayJ*Jx`u@A+RhOSET};< zK0)yKH8%BA>Z0w=c*~PmfkIkTJ4Nt$^5z;ndh$ESPK|gC62Z-!iWq=Z0=GYIM~#bO z*W3+abXv)^wj4t=Ux5}RdTyfHE|}wldsX!~%I;J+ED@pT1ab5Tt`3`eauy(YwNGH2 zF@GT1GefE|-ATvAkuLxo{=?6ryZhs>oQGZIQjDG-P_qFWkP+m7&f?^ceC=-Y+_c=A*Kkf~w36a4VoaB}N7?~AE zf3Mg+OjAs1gzT4wCL|Q*l@DvBUk$GLej%dR)?ZicbUKE+;Dh0-k^zm^+lOVM#qnv= z%eaX==N&$h z)_;Gb{%fl-2GRHQeP8N82JefB58vwVkR3ECGT;#@d=qRb3x%l4w-yMWTNNFzrfe5xPxplvQU{*4}%(2wux=JAMLo!s{+cy`$%P~Ga#zpf;! z8DOa=Po9i!+F>K37Z3es!NhJz@^Q<#_;D?v&)eCQs8koe@(egy{D@K z*Hl@#=POy)s5x0)e6@4f`r(EBs^-RttGo}!CU0`J=z|;vA$H|QUxC+Wz|VGu6y=Fg zifhwnTg;+&Uj*P8gr6-N(A$p}$?D7>I!qK=Xb{UvVqN!SRjH0|55@I#_+?aleBQ1`f8Jc;)}4BVv{F7x&?rozgP0-lL3PFQUXZj(u-GMGUK|rscwjy z6#{`wk1w0Ne>TQww|~`E#&}rk_-1bX3PQtas+)i8hY=T1qDogR^j&jHBvD6QvSC)N zl{Knvu@$wA*$6B9qN8=a`G=UTGmE>Wx#s8qg!tX&j_vyIj;eo6raab^ej4aRI|pV- z5pYxbQ51`rN#aKPPaj0jDvI*TV5d|rN}KDw`twg1Y&fcnTc*!E<0Soznr;oF7$7tF zYaSdNZ@Ori4&{ax+Y$PAhl@5Rt$ozHxO#pka#`0{b7|g1vN?<%{Oe>+!1FBA{@C!+ z#8kQ+hU7FlQ1c*|&B5!^I4j~j9x zT&V;W6AW!pQQcz&$ALPcf3@+(fdKs3YT23`;g0!zubB%6g2GJ2_)P}QjAy`3c`o^e zjhupx-`^m@?prhVf;#o zdYt;vS}n`{8;1;!>9UWpeeH~^Oni^N#j4QiO6`})1~No&;S-G5yx0(ucO7WC{l#_` zAl24qP1wm{z}<~zkUZ`+K-BE4iT~L%(5w_&KSC+|u2hh*xA!{dUYpzdPf~D1IEg^p ze!|z=RMUqGXk3Uy;+${X;>0lZgKYqPy$im!Gjyh4_EpLvp%*rLRC1} z3DN{0fYC+r%;%faK~^vrLEp|ZmUlW^*0{{UCW|W1K*Oe;5iKMZDRS?(Gi=A^#n$mY zBLOk-{thmJ?m2A;LGuw16t)M*nV?Vmk}mz;ZeeBmz$$SL_~CTu`!bQ9KIGN%VlXKnr5g}*t^{uY{|V^t zVctJJUelgOx<*){pVmQTxZRIhW&$d0U5O>K9V7D2uPGZoFULClg@^ z`+52yhQ;@@Vq7#KA62bcl(0sCXUE*J%gL?opyMwg>Y2o7H@f=Mjti}++ow?$0?qOe z=Elobz3I}!eg`QcclL=wUMzTS7V_mVUJ#jnrbab|kZD`v{;|xB8=cjgZ^ncyl3BxnsG$nTOAnvnugUIn9qAv-JY)XkwuqP6&EbUis1fcXr&La zS;6m-jXG2T22&Pds_fmtbHq2Ss@Jw3OuV>Vek@T`{1JmY4sjvHxF~?a2Chu+j@70V*9RZlvu63wAelF z8=`$>DBKtB>|K9-qoDRz+kXYVvOUBUxfOb-_7VDDK_Ac)eG)rT&1{j)on2Jghc+rY zsrJZGx~h1ymVZEG?yebGQAoaBUh90Gb`|q0kDYN+Cg0*yLjL{y{sQ;LxF{eeCXKq1 z&+Iv$-K_Zg_b|k=-Gw88RAlsC@tL1Rle_=W-l~o6dgPXaz1P>OQzH0!HWUK$}*=*W23$k!U4{0NT1>Y zb%$vT@BD==^O@nez~!;5t$9(aMzn&Lt4I#~589s}A3elwPgGdm=&Qn}ZEI-w!H@=o zjlMrx@@wR8G*|o)CfsDr?kgK;tu~kCxpa8(PSMK-<+PH9)6Ln-B!yIdxeO76GZr`#2D z%bmI9n#;^(Gu!#}{d~{&{Br(Wv#;OG9U0ixM%#Pr$2teG&(A!6W=jZXmqVVB|_ zSC9HoH2506PI_>p4PnWNXv*;&Dw49fc@mD5O_^Cr~RhyZ6E$ zo_`*RF9X6^o;`b)3sJMOaBa}A=FaM+r?hkRL(w}8jKQM9$|xI0nWC(Z=Z*2LcV;E< zWTfPGHDc*b`jgoxwL(Cil1rK2@53+s|8bn=HXx4BDX*Tda*IZ4vp0H73tf{@6;H6W zdy($Uifsv!`-U)a!_*EMHz0>j>wA$29ez#i?&$w&+Fkf2F`;g7ckg9mPYUOfVGvpnrLc>iJjX`*og0Rl=cHoe5PKx&{Xdy_9(ds*Qs2*AR5He z=-xp8Edvh3VkHAtP)obH$xjIEY~#ZYWieB@l^&z}G4fpoxMH2^ZoME-fR` z2Gk__{Z2iK0A7m~kt|l@=8ytgeEC>f1NwM5d>OHTtfBUEV+0S*pXNODrhWz4X~&`_ zivV~Zf!qDXnYW0Hfldb~kP_7mcy`4!s5Y8%KH#4DDt}`fbmXgSwSMN@A2ISJ%Pgs4 zc(JpLge6d+J>gKm%7k~;kqJO1cpMI9ODqd&;1){hc_|OCg7SNHJpn_342SX!Ty%| z1q9%$<^1}fMP>VA!(gKdknq%=I$roH@L6Avbrn14UE^|IvhSh9eB)+McyMRZfH|@6=&oC<2x|lhtDcdm_OC=J_%-ZQ6rE#j#i(u% zCv4fp&4sj$`b?dH^xW5`8?{s4{gjH4 z%r~`lntwBHUexf_qlidrS=QOk)~{d2G%_7ONIeyYlRLr#1(WP95fdwyj~9JEcWl<_ z0{(rim(KowZlF&o6~SBVz)&^~!Y;BxW~{UUeN_R)-`tr|Z*^rjA=-OvMy_}hgz z==_JOvlrrSkk(53b8W&$R(a+jx^?aEK6UjGMX1)*v0jBp!t$n*b6@$`R1dDO}(VK=lsQcdJsm z=(4_1DfsOjaI>U4rTaS!UFY@ffOqT~OaqMgU>B;~IZ8OS?AlitZNTmVGVhv)@ZuE3 zU+ZuFtiRc|DzYg5k;732>|t7VR|gH$Y61h0#Xp0ug7Vv66mltYO~z@Hb*))BJ!hoo zz&l=T*UkL189eUf0|D&M@xMMMw)h9~QeAJp(Nwzm508UU^ zo!bCm%Y{WKg^yYU;YQr(Ks4ZjqA*#!c**k#gNxK_S4pxPnY*jzH7KP`yPdI^GmEII>H+SUAEm(ll0u6<3MoE+f^zdLgb2y6?0Pl@aFAe}VM-*#+dk)`cpKCJ*?C#$_ z?H?E#$5-kaBDG5Y^-90U>RRyduL6m$6~>u;`EDv_LT#QhGUvu547m&d1 ziWRYu0O3vMyL08kweqK0Jyv4oM9u6XQ@fgJb`Q@<`IAFU_ZG!QJ{nqY7Cmj(4Wd2h`kF^v z6W=AvZ>NXKdzVW3lO}uTQR4TkLkTW7)mWtuc796{bNJju*Av;}3#cj=8HJi39P@f= z!Maavf8X78S+TsmA7Eqqj5P%lD^i`#2qN3ac~m?vV)c9taMVXto&C&ZVm7VkiNHevY4yl6LE+$0(80G*QXSc;lUPsi;&{O(cH=~Lp+vL2XBfuG3Cgzw0^CX)oqj>1|K($mJz0f~BN#5}MEwnzwwI;tL4_b! zyNeTdeL3(U4NUkrRW<&k%fHOJ(av}Jx%YZJkvZ3Q!=8Vl*%Wn1AtM=-QVtzw{=G&> zkkxRLfmq+bcET%*qDNqy(P3a(Vv!=Y>~Q|$kqUkPMZCR ztcf!hgIZH;6-|nPm^0*9^u3r)*RjtLwg7t{TLe|`n{}00>uMOj#yxOjN!tm+F(FhA z{IF*3KtFpj?NU+Ao{;y0TZ(O4+>P0dt&&brW*|Q372dh_o?vdoX^%TI4&zmCCrXYw zfle~+N;IFucr+cVo4!@QdZPHd$23?f^Y%<~?#}4b9}4PMV5#p!J(hrI--QgH!t(&w zW`YFW#Jb0|^K(~Hpt#`-r&JQ>oN37A|GNtyc#$jBb`+b$7?-u|V4q_&DO+z0tl!*jB( z8m90?ZoHSVWvQ^r1$i;G;ln`|Emta8cE8FB?JLfMZI)e}1Zsw#JvPc170DHDmUBw; z%|*AK$8X2+O2$R{?cu(kEbofB&-eqCP^j>VIYjOVN|XW9re8nq&NGX8f$Q9Ji!_t1 zmhn_N!5ajQqP>m8IbPL|taZ5=+@$Yao!7Yh%9-lh>Iowb6hwid3m?o*2CUmWndXJq zY?uoqD(p?eg1;``%3Cm|3blWLj~=Bth1otjq3K2yMFpGiQ&(bg-3*8^$iSM>XE4Tp zl0-SeGkvf2&SKw$$-(3{(K3&|DJ3bb}YJ|Gn zqs~Y9dfeokOd)hlH4j2~1)yiI7a5G?_RkXld?R#PoSL0+5by>`-2Syg`~wCC4S0;~ zAc;^RMyZQdI(jUC`4bsWp(#=WvKt!=46+T}Y}tOQt;m0o>*WMk%J?nt`lxCIkw%q3 zaAKY{7Xk!jo_4x##{c8C$#I8ZYLp2)P~Dz;Oh%J_CkIBdM2M3P7vCRzew2Ui(hhc4 zy`i{(3dt_>!^meTk$D8pa+9Ia++3iWOG2@?$pRl?iktKs_RjfGLXj6{MAZp&hlguf zEw)V^6Pwv<$@sYrfEKKg_3blE=R3dz&z`*2K&eh^3gJm0o<&~}L4yRg!!l<+RbK0s z`LozB(y;OT=qaGY8*Ww?+;r}zf<~r_#6@}jjV4+=Ux^P2UTm;iFdPrLPguvU0CS+* z!>SY7%K;i74;z zh$}O*r8PO*cci2fHadMLBZV12Uy)%!0jkm~YGMEReGg=pF*Y(s??;1H8iYVCLIoPPavvVnCQ>bV|ID_^)^W*oeH`+m+c?yhq zu?d~BEGen4*y~q`8^4Zyut{s6h})Lq_}Il~mF03@!;AKsU|i$WbIefGH3qhC!R1B|%hu1<4c@pJ;n4!F z+w34O7xM33qdia8QrJ8WYI0=i2h;SOER}A&P|a2l|C;`QUzn3>$>bNmyAT-T$v#L8 z>OZcr)^J(Ya4xK|6;Wvrvjnrru?vi5LlaE{sMj0~OlZ-p63N;8qv5~JIn&@2x!le# z@pRGuisot3`6u?E(6;vRPyYX!VOjo^l(&?W_O}<)4yBH#d;tp?JvGXndpsq+g^B@V z&QCx2Io5_cIQ7+D@V&f{Y&^~n7j1KE=pWz${9`kYe>eILzyL~*T*42 zeO;c2TIkc*qbSaY#RkkWpQ*{b3Xt81Y!=w!$}ikaNzx6S>d=EuD>of})-3nl@nS?# zC17?GN&_zx)a01={k2{h%n`rNm&T zk^a*I%PzQcM0)6j;|-L3tgElugNu*uNZ1FvdBseY1V3#?(VZCbSqR?~9kN+%Y~I)s z(@HaKCV|gl`Mm%U{CjtVf8Rk47ry@6P)pZky-&3eaFdSLf#wGdr!as!t8e~*TcDqY zk?5J{hgC%L62RTpa4&kMMF=${bD-eU9(}IsXL9fd@i7xKAy0e5f9;Xfz80nrA_5P? z!#LQ9owkFoMF0T@x?l!b1{Qq!*fu2$CLctZe(-@ArJXvT-vRE~Q`b2H)M*lHI@fe6 z2!@WMOVweMSC&gZsnIW*V{dY-)9(dJo3hxuLF>D}e06shPmILI2dWN{kpz91t99)A z^Ic`n&c80&UNwkjR#9Fv9c3=cH9Ck_cNoGtQqSTwer@xPb)8)V&hY*?w!eswd^gq# zmox3KB+9#oD9*+`)(zJBPT@vjrcd>&cnXA#T1Xgmq00zb8-7x8IpTpW8fN+Mm$wH&r6} z3_&0sjR!l4#4k3<8`ztu5%t3LmEUGein7LHG9}+r$wYR$=&6LO{t0s$W8kB`&}5d7 z$LyQ>0b9RM8oxT2XZf3z{nzNRaK$xBs^1kY2iL+ksR@m@q2?%gXC;pd8|39G=gcRtFt!2ilX(qS>lyw@xh2RM1;Q?_1l}IT$ zOOJSG0A!{@EVl|BbTdRW^cw&{P!hmWt`|qh1Druqb-=aYT*AXO5ts8NSR$YOCE5Pk z9IqdAHt4W#^L;P|d|{(aJ4;Bs#Q^{X#JPbN15t;rf6&Uz&p9fL^@=-5k4+4yns_?q zgdls9vduE%*gbu!V@Q3seOyvy!5L8YM{qk2Q}73v#B-!y9dI^FZ1{mwIEDImsrn+8 zv-q|k1-t{e91@F3n`{u4D5lPRk zGm$XNXptX|aR@LV2p)qp9}%F9l&!s+DyiIPSN#{4{lN55MgHW)ngEh(kR65nXln_; zd~Hn%J69POjsGME*O?lLqD&4ovbN_}=nL3C77x0H%FMjCcCkv97oap}bZa!pOs*hO zd{u8hFs5dH8Z7O`BkbQQ;1yr`;6TaUuUE%U^@IFQ7cuKOy!g@~z}b&4Z9`=6%2+=j zFm|Qo$n4?$1=(KqYFSTV7QcW6;*wO5PVFdJHviAw?aKA7N1$BJJPCp3lXb;!F8I@> zp3r_E1uQTdVQ?mer=3T5`e^2<7whkrpG2+Sgp1b5&kuE)PPU4ptg>3mvPG1%PR1P< zYtLN(vPQ1R@8QzVyN}e_A9ry(I} z-XZ33mxX;`lTeQst^Mo0L=Dk^q|ItF@0i6U{Q>XCWw3cjSpD!flgKoeyfNRLvtl$p z^lVfBwI+!zD_-|R@?aBa?EAWndeIyy)$62nqx%EjF3oH-*+~Q^h4&JI#ic1ZJ`B8x z9=&FYqR-8|uy@xnz{mIGP4<2;VF4O92kRRykyOj|k*K1ThR0T8qq(}H$+YJ`Y#d%^ zYWH4}>a?8J+rj-BmWyF5#n~fl1lnbk8HkrZ?rVGcciZRto;v+HJV)G$99+Teu|eoj z-02rmIxO%7%}Dz*;H-4K4a>zNwUPtCYrRV9jNQsLUmzr7BW1i|y2XUr?-fnWNtbA< zULh`p2cj9hF-h{rhK6Oea{{PRxP}ZZ+?;ep9-1l^qm|dE+fxM!Wsc_lKa5KM#c14k zX63&aT}c6lXd;b}Uy*x+X!32leSzMr#5mv-Zm9XxLFEJZ^_3rV)*Lr*J6tROOP(WJKPm>A@&Gq~ljlZNLsQ*2h$=wjtA!v<$R0rEQTohj zM;i*h1bh_u@Q2h-><;iav<{f^ceJq-s8JO8w9k_Eu}j;lL@57sNGvH=kbMd|i7Z>| z3A*@5TOBhseAb%p^6$2&Ke$*{uZ=3O{eCS-1!t_@Aog+%4Wb_tCtFLd4!_dUk(%G* z2?{yjsVkP;Id%0zBs-`-x|s#ZghUJ%8Eb|v1V#j)6n;44|G3&&v#xElY&|Q**fwAQ zA@IAd%b2QU{^6pAGHbba11~a*g0}O!IEx<}OewfGNvd}GY^n!x24@HRl0Mi7y%bzL za%gVM>cTxEv8u&D6o-v27OnvAU8(!0O8?}3EHcIarwXd&hrx#+8Wh4 zY8b`Wy;Z&JZHpeXQ>V5Lu^B~dtB`S&@flLAX(b@*u1YbZYT9XB$?-N)-9gZN8e?dZI_XH4WRG6Yt)4*WXYdlDPOK0M}d=K@hgH zZG^~b?nrEL25)d02KQ!}f7uF3PIad@jV#oZLv7;CbfX3ylf{pyI#{6nI0KS!uS~ zvSZrlmH43P0>ZidyJefE;x6~}nH?=Zr8XyqUzzUsemWOYZ!6Zvc|aAldfFrFyK^~T zA=#xT_<7}}>weoy$+>)g9-{Y~r(BI)dd+oC5$oQ2`kvwXYTSJIO-H{6V* zH06##;EVR(Rx%@EBffACs{JGDw(r+*zu{KRYH_@#2zPMjTZK3=+Z#+FG1RkRN(U z0HF0wWvx9T%K~%bdI7r~vj)J-Xbl4J?D{7OtEM6loraJAj%Z&8v;vIWS*a%T80Y~3kr z-1x@3^7bhZ&44bp6~O{2;C!jTzNE|Mo2iY>&Tibc4~u#bAnOTpj)K+jhOrovJhw75 zKYwvRP5XyMvD!y#{-|VDzjX-jQ!Wcm`_c+*zi=)^At7&95HQI;NuRRRk<7Om)<*X{ zU2MRFcs3pWQw&6bFN$p}%Rqa$L@E&|@!zgdP1}u* zPGhCbzPOdSRlQ?1P4WXVDF1i18Vg?sHyE#kX;dJ+)U zO3Kk{@~*d7qx25bpi-LEqp&;V2U=dZLq(`MMfV(o+SM1U(Dsgyj38<}@lMb4Ouh=Y z{B!%?f386e+VwQ&I)819RhzwQ4Q7QDd?HSaXjVPwVOKo(t{mzt;&5JU1EV?GqUJRt zGC8uiRe{_l%g7ypUR_cN3g|VcAJvs*^#T8=S0?ua|13F-M+C6nzeI^oBcl?xTD9Zc%Ewz6 z#ZsBZx!H7P$mK%rRx%iqWT2w`Mjzy!LIC$Mchl6Kgve8E$1ztB5`<1!3INM81Y0nM z9byV1DmyBuqqnYPJ9}|1I(>7`KCmLaHCs|=-OU?8?tjY%@KX^EgvLPJuA4`|Sey=k z+jh3m*ixxViICC9D<{maikWmSAVerM?9G5&pa1|IXM>4KC!~5oKknAA<2^bK4%UO& z)Z-Xoor?(KQ!>vNpcD_ofbq_FcL^LTc9Fv#Oa(x;|Kz|A7SGseFN1bKXP`;|0>Mrr zj}7wnIsym0!fFFmduJrLr1vNqgd=>6Sw4?#t0kkNrTTvqa?n;(cQ4XnSvlOLr1W$^ zpEG@9@2b>_3<;FxaXiiEz$PkI`Sz#BuJ~@r%4HTNN#uU7W3$xL31Xy z6-ykzlFUn!UyGqv@e~n zSL4W@FudNXAh#HG%S-U*a*l4OOY}&h^r@&KT&IN&=iqT}tmV(&Lbs@~IT8Q3JF4RG-GQMXm=tY@o(1u>329FX3D{Q6}YGmGCrrC}Iiy={ix zsNi!O?JGnyvP3<53#NjpvcaXDIn_#zCmM|rgBCr%wUotPjHLEd&D7RF9oxsY05y=l z;fwwQGBJaNt?Mt~L3>mlojP*`)>WLE)HJW{sGaLg5iw|Gy@k$NpY}h$t7S~DtSL~Z z4abLOIG)09-T7Viea^uXju&jG-mzM;NzC{Te`24{$Nxm2%Jz6&UaZ16DRFb!hUWghXSOT*?H5B_*4WL>UbF4hPw057y3!*HF2e0irU z3F4puj7Lp=14P_De@`2l%=&Okm1qw(IU{S&)NI_+%$M>Y^U~f8$)*tGMiWG?elH^5 zU>jU{0sOkDLjrMi%T9;YROFxtd_<1)oobDI721u)isu{%M-2m!B~*rf(dEX$0x6Y) z7CUAOHI?yFwY<-H9!DQcPF}nt$zS)BKR1o;>Tq$wfk-mcNSVr8pv6b?3IkLQUI^Cu z@LlZW4`%NM2h$(y{Ky<`F#Y3y{%Sq$yR6noIOR3sc@}{bRXKbz4e(1i(m{E3@Ux6p zR@(aAp4Hp7LE+k(JHb8$@**DvVsG1~fnG5_NIJaP0XE^{>=auqo^yTW9PWJWj)LJc zZ_s%EEH(?V<6pigCpX|I;tX(8l+M&2|soZz6tfc4a9#`3zwdr0JL?N!)05*p1{3eG`fe ze4qTh?_5wnGUf7bN3nC6>hvI`ycidaaV%shX5nUAXv~~~Kq?BbH|POXe;dW;*ct3< z7_8mZC+bGFo6SDDMtfDa#4#ZP*|oIN(1X(~gBW_d3Ew?Bh8=$-%4(87)@Gb7hWtw7&&RFbX&$ycF-(x< zPViu7A5=+pik9|+4&MG{UbFMhzc+B+<2i!$uC^ggP3%w6h3OnrziYYw2eJaN$#D%u zr^N3pK3Cmtb9UX1cYpJ|hd*CPVIq0zgWcj9(`re(Xg>1rYD__#8h+j`(vSu--OONi zJV(u(9vRjG^LF+*C}b8NuGE{_YjvE~Ys1~haO~4o+6oMK5ODsNA+S}RSgH3y$uapL z(_G!C@n3KIh_d>~9u(@VFRa_~Z?&giLh<~=Zx3efm5%1hZU7#Eu5VbL7m^Wr>?8S!xp-R7h50>%f}BfOQT8n3>?2^p`3M$4rqjBN2j_S218D|7ub>yOtvOdpzdM|Y^F&ME{Q z>^HLa=Lr-$_O@LS z$^NaqGV5Vi*Kx;7`hA?Nl$& zATZkLspxR*(Fa3&_jQ&1x1WEBvTi-2@KXi*?z`_Av4Cx=!grC5Eu*K<7OWW9>K zxF>Jq4B9sPurVXrWGK>ZGmRCHsyDUSmGV)NK6p+vEWshiu?W6b)MMLyPi^kR>67Tx zQZ;&%8*So)lz9D+timu?Lo9E=tC%^Nmx6zmuPA30w=pPh9V06*6>p9>*DP_o?daoS z(VDL5%qp>Fs~Yxso?~QL*L+t~zXu;H(`SSv$7#_|1-`Od`F(-8sElt>pz)OFv;S6e z19S7euI&51Z0o!_x-*mH{2mmF3b^)*8Xv~od%3<8|0JFT=iCTqrM@_((4LOYBhJ+? zz%%0ln2xV$dVaGUmUN@@#Odci|Emk_N7zeGEA>nY^W%-$2e;IM@7_7`ai=i}6x#XF zeA3|m)nd8-s%Po1T5kUuU{4Y-{-}_2{Pp-G+t$=vY~ z(H*R<9&{rh5R=*w&)x*gq-3RSYXYBkJ-@W*iwv7Gy=5#GqB&#!c_0yF2?kM16nR6BqkmBDCv+31A z2lBseu;6=XYtIW10xT*&^J-bRAM#q2Ez9*quc!vHOQ=INCdg;&+crQ`A~_hJGk8Vh z&-xbt1S#N3DV&axzaHnN-Xli1xVy#B`q$a$_pZLW@J&dH-2ur2$g`avmf4wiF6#_s z=;2izLja*nCE8dPFN4z`SL1-p9LsfQe^bcrNK zn<*_8ceF*j*KM{sIe?x9a->}m51B``7cv@oH1+S777Hr@`zBp9X z>f9;B&Q4V$zCfzq*_%N_pik{RdSJcX1OxieHNAPT8+LR~l!~HQ#wt5kFjdyaje09; zBDzB?BQO^6jP(5IJM8^8;F|feX=0SgZLCdUV*A(y>e=1`s*53MxJx1vdf-!hbhut% z&NI>Ro|C%9o|Au*npjz+whQgYY08Pc-(@9O+g`aIxt-A*Pul5v!2#iv**xfTNPWW5 zvnoNBC09KLRz|i9Jn$<#?cCwq;k=J5^{(1QX*zyM@3~!i+Fz#Kh&UJF>=oliR`7Z- zrOhDn&7|_xuZiZ=yl`IjS<9r#{G)T4biMmcduF`)w!Us&m2#^#Ws^@5H8!uY6P!H@ zc~RVNW6?Dgd2SK-tT%2mA;$H~Lmz5}4^2{e_o7f#t>M-?tV@K_5)(_mbaC_Dh@vc5 z=RYq@us{E!jB_(+%U&A?wfe)0M|0IU-^uz?QZFBL`l^22>ed|cgfYq$E=Xy%IUIQZ zu@Xw8w%LCdWcp4)|z&=-Q$6UA$tq`S#ofEY!b&@c3>(q2rdn5Ndz5%=O2k_1v~+42}K&BH7pBZ-LgssM@sX#Xmq+Q{*(dY#eQ5l8cNk zOl4l(DeQ9aTy6N-{HvK8aZ|}s>9EL=NAQ;%m(_=n2gn6vb!W;EpgsUc%8(@nJ~4K| zb97uj_@??v>3p)jwabMY*zVtN9pi8(zH}q+5I(s?7mhTl^5@DSKN&|#&enTUADvZq zA0x^NJy9vyxl(|LW~V@S`h~JeTv_?t$L2k?kP}NG{vFkf3Il}LA_W-_smfz-|HzyIEpPG;=WYwbdkiU+uB})5*P8w#Cxrb|YBj zhYaQ!I}LCb%UTlyk89uNUHh39_<&#^x^4cvP!y*{_RGre_m4aS%A&B2ky0P$t{2V1t2vl zhT7+*ao@qyCjLcXzpgmx);*!+8}NNITLcC${kLpnwGEns6rNcR-j*o4h4{3|*dnr- zrP?Ul2^!IZvQ8K9_LHYKCZ8WlSy~YY#Q4A0vU|rDI4OG| zP7Y$j;!Y_Gg7#p?4S|FIcehTa_-qRXC2jaCd3xki?O@AW=A{5PS4 zl4&?_l{sEnh`iA$7OdrY zMb_?8UZS`0#1d?*wAT!Ov@aa?sr!rdWBqDnLMh-YjJLg^dVMgcACU zWK6z-l&1=$m8YKdvmTNqK~AR|E}QdnW&d;meEitIjl?6Zdi}$P55lFNk4r0ED*~uH z>ID<7!ccn@&|QEEvIY5*m8PFH0vXE{+b{d3~iLhJVyswPneJhi<+~NTc^8>a*TN7>+9X+{LyU_mxAV)!q(Lsc& zo_V0b9p~DcJBKVj$tys!IDPZ9{byQ09*?av+D!M-CG6p`D^M_|q@l)HCcBCaZ0O8^ zlRb*~7?3;V-@tf`J1c^pdIPop0e^7#cCIHz27e{bQwt<^CSQBL!b64^4l!zWIOsQb z2j5L->B-~VKA~PJuz=Kn{K9yAV$YU^zBnM$P4UaOTq`n7H>nR?9AT{i za?RvF^mk8(BN5Mz9$UR68JH{*K}K4d9YUDL*uUO^`^~-4gIUqv@s1f@>0C}mBs+65 zAH}d8asq+yR#UF~qX8Uu*^A5`HnV45f21A}I(ReF9Af-MS}39T(qWhzD?Zw^Wfl|~ z$R$c^fh?@$tf&aVpWQFt(DQXY+bY-4ZHu{%xJ?OP-BeefR@Tne?Rd|+D66TJpl`dN zs(}%<-4RGXe0K9DSMc~(VkTMl&l%^sCS|$6fg;-`%_czyNqo)rFYDw$D19;UDee0n z#uT#D9{Xca@j%EZSIA{ho@7UaY_*T?^>#gM<@Sy%K)q{w1^3V5n;$rCZ#>m|+^*>g zBiudh@p}!Ltkkj`UKZXqB9W1!@A9U-a85(C?^eKqud|U}y#LPj>6P@lUc)Jy_$kj+ z=Rk+kmSY13rgLV*$5wU_$HX$du=BUsb3@@jt2KUb55LGNOnxIFYp2qgqur*e)%3b{ zt;TgzvR=k3tN-)eHdW$pY4{e5)+&Avw!0>Jwd~i&U31 z>#6DY)^1_Q!EG(^E>-$ra~8E;Z>Dm(7@hhuyeeH`J&R?*^|^cXyCJKbjeEF*l9n{a z`}7s+VlC6_S0?_pi4AlwVf$>W@K0r#z}B|WcJv0@<(uJ9q)ta@aBsK>K)OaZ7)m3{ zz$MFk*3@Hr&YOIS_t-ky{WR`V^0fh75nPe1s66t=|D8Lkb5~acHVqXlhHv~Eh`C9) zC{0~pg@Ww0kim1T>6}526C^dr|qOQe-gqYZ5X*3c)M|;mxctQDktI#jt059MR#j? zc{vmLlLkzYBj}2>_czx$-YW`>zuyJ`%(`_8P_EEEZ0l;O@BUn{d zepL=Ioy3{NVLR)9(#at4s(KegiviCjKXdGWN4{j^WLFVAlK3<5Z2DLk?zMZ@gB zlMZ`bbBq*LC7!e1)CpofudH~`5>oXMzhi2fA2i_noQT?$L+hFN0`rIrD+s*357edwUXiGCkO+m)^N zURFUnR~1IW1LBx#z25fP)^X#`VWDSMucJt6?7L=(>3j+-v$oQkp}4d%=KxPUvy^QCgXTx8DwBeGiE0$efxf8{cAwwjMzt;5_z2rVWSq z81LDI)5-gwv{tx|oR*ma{IeTSWC1I&B}?6BUkYGuM9u||Q%yr%Gd%;ZQFG%_h8+Xv zgTMQ{QEQalv}B?zxpOVwG_!fK5}cjJ4=8B?`z03}V~V0XzU0~&ua?g9<snRJaoM8;xcYF=uM5JKS5|0um}CZ!lU4 zBgIF*88VO?A8j;}8u|=Wx8A|hJ#2!7h*Q%D(^Q2&t&1<*! zZAMaDpVvBY`BUDntuRle;04X8YEVrmsTJ4cKzzF0*Mvg6CfRBep8u3o;qnsleMIv=q~TN+F}cUf&4)W{NqOhL2xHHRnjw0Fqgy>3 zUNV^>0j**9zRZ#rYt#s@iIU1X+;6>$Jw>BgL!4v2pX^2__R07RPp$t|qE~)~aC6yu zM4RCMCH6>95~L$H*T7@HeR68$ZGA*+08$Ql9uOcT1(3otKf7i#xwnl~p@6s=il!i) z&)VjG&T(>a9%4-HHdG-4mN^nq$J5F?JX~)sS}2X^zFt`_F|XX0>;F3{(SBugy(3&6 zFa>xVgJLYc6#%cs9nU&+^*+@o6G@iQ$O+uOIJx3T?;ex#peff@w1HJgNx1IKw|t$v zR~>RN+ypOsZbj}#z{`+J98wR!=t1bqfpE|5;>U=zC^ztpJLqvxB-ZC6=hwV*^H6g8 zbj)ZwXq2 zqcmtb&oKZJD&#+AjWA6kiL>-5H*0+XDIlhW%0nULz^gLjnm46i@U4}8#ymeJns(Y!8fgeB7f4GRC~fe_|%(UB!eKYC|OTS4#tQP%@RAAtJkNN-o6>pzgt zQ(tgNQ?dng71DY5aeIewd)BEm^qJCfy4ZWmu(31KiT=Z%>fO+l?Rk7d?+>)y2C&rlysQY4qxbUb zQr}CJI1n)^j9TjvpJ324AUAK9pi@?JExPwj@Z${kH`Q!qY4fL7x<^~lfdb<>yuQfx zNLBCY`oug|+D)n1Dy<%m5S;~E4gam3FW=l}wcqwW%r9E2Zpgdv*3PX+im~>dqbUnZ z6c)(!@RhC@5|*N9;#(cO0-^<)z2NHu`9TXdw_;|?GTTjqmvl}`4yZ-Q%f8MWN|#%( zt--4I*mi#IzregZ7O0?o*VT711^s5gSqs6bk$y-kjd~+IbRx1M@+HQJkogf95%9eMPuD#(nhvAU-?ucM$P)*{$_|qqrOimr_qJ z_QScPsb?*75-tu;9J3#KP#=DSK#mM38z`Mt2F4Kp`a)TPrxuj0{&DS2P)NO#$Gu$s zcr76ZZpQKR|xQ%@ez#QuZRn(O-5o zbQ?ihamo|db`}z2yx=aeg)5ptz{!oX1Rgk{n+lTtl5KAqErYm#FjfS_gb4XwyF%cq z44+zlli;H*Rm;!3|Dw}Z*D9;qx*y}zaL@S#C9HvS3+U5dUfLf#s7h&dPA#=hIamy31!Q}Ay>%nwHM-YEKAYXdWkhc(&@xbcn%27yHPcSP zw1-S08pejD$)n3wq%*+|X3@z3!)L9A%L!Sg5ME)>r6_K}eF6LGa&2piqdD?G9hPn| z#B%sfP(DVBnCAhoWnB#l(ZdVZ30>Cd`n%PfiP+yBKc02+qIB$cQ>tl?*47V=WIM$+ zFkIj9BxH z&ea5I&ULNFW%P8-$X0J|IbJo7ys@|wc?NSCz8G#AM6Fml#DF}L&PP=Auycfh*J{@r z-g|nrd4ArH5?RbGqYYf-qR0mOQ_{aclMZCF2Or@0gA1;0)j2@L#XX}k9|$Q(E8XSv zZBdok`C^BYfJe&gP6n*^IpLH34`J^a)dbgdixLQg4uVpoDndk%(4=>iP=r_>L8Sx* z1O=q`7K$PQp-T@{P${7!Ed-FRBs6IXL0Uo$5L#%*=l#w*?iqKCbMILH_rJO4Tx+j6 zcj`khs_3DJ;MUTOydd#|(KnU#Xte)Z3zzaJVZ~Y|&uc-7muw~Hq&9}LHb2i)Q?F5+ zvUBPqLjn9Vj#|5!dk>On+`VfZ)UFa%GTL@8*b}H7FPyc0->_Yq>Rk!@eLT~CIwKx- zW+<%o_Jfmq9Q%gI5^2Euk~Xze26BA5qfP6tvR`r!0$zMCkd)EFp#wcAB(N79H=?yFI=N82Rf8S=d z1!l;3+~CU(-OI|q)vq-5#SHs?T#JU+t#e>&_TPWrSAm?L2VW$AlHryLSiE)Q)wXz6 zT-yPVuSxR5UL^Ouc+MsY=m7))Vh}nhF1_}~4FjWhVq6YtO1L*S!h33mch_KdG~RI> z^C0H@Uv*a~rvNe#2KgY-^o^+$fX%$6tWP~bSs}# zwif12>o~r3S@ceJaI{)#n#kjb7eO@OpoaHmA~b4Rzn8sSv_x1Zm%L=GzE(}stPWog zP4K0pbKHkK-Q&2vrDn^JFP8U6%c;TGa4y_+XGD%Bmk_Sp9R80JNVvc#>>9>E2Y-Xh zA#L~k+2M*x9>eRigUkW^&A?Z5*9OJD z%BXai1G(6hL^Z7K*GP7*1K?q3gNq$(fbU>qO8VHr1mWBaG3-Lp@1@m;EQ*OdISRfJ z?;7~_>qf)OhuoLYnUjl`>3;*wfx45$j?94DOIEcJj}iIRSAu?8D8&lnM4iTpE z&7BJofWvTeT~VCy^zBHEVB~&<7A;nrg4d7b2!5_>nWU1vOO{9IQiAA(s; zkB=h*;%WbJ(P*!_tg@VU%J<-*zmqzot9lmbgmBo2E9+V?@gNaXQ^U>^5zO%ci zgYp&T@Rt~?mVYC&cFw7y_YBkHcg)-7=gpoyt6P5(}ep# zv2KEd9SyEE2UfgjuZp*{(JDw87?xTvG-@XsSyIwMnWckJJ&i=&+U>a0KNYrOS>#xm z1tvAPQE8vCUZdLh6;*?lu~`rKJJ}lckMyKag|IfK=jWB_h1NmikFXC=)FrK__==Eh zuZi)`x$6_)!H|$|VcyJ?9Ts!r9p&}(iog$wzB9F68qY_~uhpt*q@IuFol!2&kd)xz zTgSz4C%hJQI7+G4B73#$hz-7A0$g2Y*-n4mC${0B(@r)=WWyJGHAf3npW>gK4!`Mf zE~woB*UueHd4vpWg(TWYie880rbDzN!Ie`jLkUxdato>HE##Dq+FNJenwtM)HXjDy zN*3>&^$2_OjgECUj_NLvY)=W!eR1cd=hKvta7V+9&n7|M;}iErLeBetEU)dZ9-jQy z8s$y^s{HNHxovk8<^BWwp{t;*+YGGv!?y>_ab`~4TPsv~TqBSwjG?~uf&Zq;*$dLX#CxDkANgSBwWcdSx3kR<;|Jds){!fHR z>$RFI;1whZAS5rz@cl3~0p}EO{9)FgPQyL(?h>q$^wC*7#Z3+1q=shN`cta48rJHu z$%lwXpY<}+(5gIFcj2;q*GL0$=)ZFO@aNmG!*ZeColp9q`5^Z(3BvVHps@$w8PGfb z(dKbM?CTx2dyho2uYpA+ZM(G+S#0%|S7}yhF~As(#>T2sj*Da@k|s2~iw1rKD4U*7 zr-5Mk0NSEgW$`|`2+GjP3>o+}XX~AkRKQpPj_57&j)gEGBRro#08~XoMv<_Lrf3-T zjWB>N#E-2)3$ij<@;bZuhND|bv&ET@N9bF+_BV7cnIgprBvWcGX4o}=CUZTCZPau@ zkux`>X4gAO#E`e;1^R3gU{;^PVNjT+?1Esc_1I4jVIFa!%U{(xJ|}>*4@>4wf6rOt z0INe=*Q3q71;i*nYTvBThjxZ>42l^(7y&H}JK(lRdDw8Bq4{OlwWao*G<3OE?YVh0 zld)u}5W*e6fV)hazPVj7y%aoq#XYJhEo-D z8u~-}UFNfb(OV&f`aRqL{h9-pSC>Ak#dGvAFUfb8Uq)^=dsFeC}t4+DR$6 z#t}FktF`g9bYnUlIx{e3f%ZhE#WO6qR{Fu;4JLPhcq?4cgz{GZes&TJpzApa-nrefKgDoaybkUBCmU7H}izde!L0-GfjIZrhN z+ZPM_OcFU%XX3br=u@^wC!$;}%@zXauN}e^_ zj??>k0%}tw={zn@7L8}0<|gxnS0YAH+40;d{~OV^D(=aG83=-rXVCQJ|A{1cM`wD? z)Hcg1t4d|IR6;$%XEqC1k{O(N9qZ^^Ciql$dt+VtYjEt0pYZJ`N|#BWM0`$vmP^%8Ab6YN98 zjZ@AOHkU6TlB7`bm)`2YabADIbbyHtL&HDq+wnu7N|KunkTWa357$6LM92`LqAZje zpfcFF)kh#vK3nz-nqUIt32h&TvW`$8r} zL7&hbzWIki!O?fbYNGy|1+YR-seXB8CTm!glyIRBn&5b`18A#eYptGl{A=X{+_v{5 z@%()!Z^*N)h)AbJj&bpO-O6N^ZEhT8-0}9&H3+oYPm$5=x&S#Lv~|(#-wAzMBQM*S z--8%-LvMr5mXix=IVm0h-O$64{V5HcZG?;W9n5MqTF}~9L~v$b^lR|qqFU3dK@|e` zcmGrLITc;Tre$@HZSLRDhDRWq6f_mFvs!3E=@@ic96TSDY~6p(1V^)GuBJTt=r$V> zk9t(EYaY=t@HzH`g40}~E%45eX@)|8O!)Pbc&BZyI9xQduQMGyWcukj5xuQT7>bpq z>hxMQLfW@ugs1Sv>Bjya&GtyO@f$VVb_;F-Vq<7G|JPD(K_};~l5N+u)l#R@#l$8y zieMr)irBr^*B;o+CI>wTIv9)g^-)z;#{au1>c4H}p5Q``KD!21dXxxrzb7ivuBn|E ziF=vUih;wnvv!fe}-wjyA#by<_d^0eeMv_*=;j zfNd)K zHnY|qM8kRPg7x^T6K>yd^O5A>tnRft=x2Pa7fK@^Bzhe`<5)SyPv{f|5HTqFoG`4yBSxc?}0afH!l^Ay1`q#|&EcUNrt#fgWkN`4l}0!$bH0l&;?YJf|?@ zQ#Z??@3%;l=EE>;`=q#oi4}hpvaw*kf=it1uu!Nn{0nP#NhD=VQ|a2{JL&Y@ll7PW z=`riz+?yJi@P6>y`zNMhQd6L+T0DTBX6kAd4GIMUjvR6Ot_j*oaWuD{beRD?Cq>(_ z=}n~~tM_aO_c(jN!r=>A7d~6#B#)!C&X_8QbSJwH5s$GIut&AxLvF7=k<8nAlJbkB z54c#PKk;UR{-RpB%#N&CL412UC@=mBRkwZic=kXBZWc!TVEes8iRU_$Vfw;#Te@H^ z&40Ap^Wdeul<`{T(+fGn77m){m*K!{Vk)dRSs{0O?k~^2bZAy{r6xdgZJNB zkbprQ${}l6)n@HuF$@!FX}qvSKe&MZy_V^u#z6BT3vvp_r<@S^^;t}oJqq=YS?1|- z;KBMJy7i0@QiC<|q78o@&bQME-q^LEzQAoBDlI5yHzZHSs!?JqvL8Oa^6KRDuvZ&{ zVk0Vv=!pDDI?JbnEyd$1p|uNArv3GZeBRhGh^V0!wP^*q_b0u^=6${w3NjgeDf!UB zglAx|%!bl~TUCnav@UL(~r2Ea~AwK=Rj#Zk?K2=A2I0Pm6bjO-9J^4upQpeX-=Ww!c zy!d~&X{pc&I$d>57SFtRE8HcS&QP{ro=H9lKseGycDbf;lIio8HXKu4r>Le};g;_& z?!ET?n}++Rm<^0jfGH|91^2BFXp}EXV4tML%RTY55*Sbd#J$)sUd{4u7Mz4l-q$lA zheYrvBTwX<r7^{8v7R0?6PMdvy@%P{#xS(wbL1qNvVC!3MJdGQa!1TKe>BSoY& zfqa%8o?UFp<-XxQG%p=U5S_=C=HcX!Lj*MV+BCcJIb+=@@?5DiQb)?iiQ8>VN)-u3 zezdPKTXs|--w@2VP{NGa6v87ekv?qCXf_YvX!Ap+KbTg7l~|lWhMRP3CkQp7YkuiM zh^LvO>yC42K0UniwN=n|H}8ZOP{-y0#F7^Iv^v@bC$L_6?z_v{395p0v`4bpUwY>y z4Z&NSQYXBO7v6ai@3)Iy)fVlexyTbd;#P(T$E;Fb!Zj}5#)AcdeuAVXCl#}KUC5G# z9Y%m_r#mCSu($2R3w)gqr6%hj8R6z<`a1WthxBx0^Yn!mR|FHtJ0m=tbq{)CMa{o4 zeWB)j38uaiX&9K#T?o!rTk3-Dr1Qf1tVce`pTQC|wuezx_Aon!W?mJqXIR8z>}4?|*f3WO&5l36QkZZ!-u`4nCHK}pa`&Lw=J5@7&< zE5BXZFjB5JjPbI<9?4GRc=G9w?#ju8;8mBLk+;6vrePy3-T52PQzzZBC_nF9l+O>J zl~vc^l7m%G&6Y?Xs%zxX`+vEw3w{lo4@9;%?#GXseugPKzlMdJ* zwdrxe#06S9?Lzr}_=krIO!YXQc$wxcc}bJBG1JBa@TfVt$8MR6nr%2r7E+YpC z=`on`snwIWA)RBfwkGFC;4CG0a<^D1?GZRI-qB?rslNGm;j@RSr4lCgHT9&T&1*-| z_t5@&8wz3jAm{YI#^{Ux8lxZNpnU&FcJt38V$8U`Qc5MBNN7pi+swQRcthibjgO#} zxHJ0#7)&!_ETJ;6kPBPxK8mf#Km;O!n*Mm*cx`=lgEXw;1r&gbQt0ZP7d!7xUvn zj!y<#bB5sp5AR?119;S4kVI-V?oRE7tbl!|Q5?>@{zbLbqUMfuiJ&1CV`4qf&5NA`nM!0ToSl@#kL`UA^~B9{&o*P?~9<1M)Db?uH1<=<>gB3@2I z)#r-QP7e%hl-?=%4nmnO=SQOJR-x83{vFZ)Sl+E3Uft^flXWny=B?WN)tbQQyzS?- zYO8tsFqI-tZ*!D-Ew5N6?^J)TC_#%_5^1fI_&d-0 zs#vfgcwN26B)A83Wx*L1v%2Z`C~ceAC&9NzuUXwMe8sGj^%HI0XgJWG61LTInecd8)?X%DOocS9YLo>j>G5C&bkZp|g<$)OA7Pd0$@haMMxWtQ7kduMMkNS_ z4K8LO$OEHO6ht$7@3{v21Vl`Hj2$;lB6-NP$FvflC}BQ3q;%odlbsLZ_-`xs zvg|F=nUH}wGwNG58ySZmXs$?Dm{ONO@)>=Akh@$Wfy4qI5!>gTPIT+p`BIkZG^$H? zzl#`tLJ2G9-t)qNL1*d)>Vz|&!5SOb zTpD&FaF4;316cD^M_`pkxM0q)?NGfV%BgCADY0C8r{Iz;Q!4P@)2U?bKL>pf)qH?9 zxx2@&DIwaQ)+tL-@aqG1ypVvV9c)Tb{d{T1>j|vn13>~7lwxOh2P7Q_ep?o{lhs~{rrhF?VaSmeHLMQ zRUQLuKU9P(y_-X&P)zfU$Pez-APC$rjcOwK51_fw^_IJyHZ>`)vY#jl2{q+p$HwRV z>%wEsSY^R5A^wD~$l)B1)O+E~rm?f%u^x#QVJ zih;HL9Z31LUdNh8?WQ(re%pV#KQ@9av1x*7+jHSaQr!;4Dntoax44+W&PbDZ>-R9s z4meE==hY%sVd}E#(kJTdf@z4O@Njw@-rW2$hgC0cVD~Igik>#kh!#LRh_Ke}cS)J$ zE2E^N3aV|Rw7zhgZq6)73Kt25#aV0?Ozqj{eE(gWwtujcUI+QyayxXu$JAzHrmhSX zVm(~tW@0^j4Z?4Tv~DCt@5kfpZI?dT=3`N5g&+ZrJ{W)KH~-DBJ5yQdOyBv*u|k2u zvN8f?+ikG74SH3g7+-NpV%KbGg}?;4Xl_+XbTTZK`j4vcPW|z1n}KgEf!M|XOuos9 zxYlSu!1UOC)rq0Nosm-2Qr;>uBAD{qApKg%7&v>4D^mq5aoPNa$CQebWW$DoVrKZ7p!fa(8{*XT|FcXAQ8%7ig=t7#Kc!Q>KfaByQH zG|Ua&-k%cFXawd2GTM)@z5|s^(g@H9a61{bn0`rploylPpJilzU^Rg;O0>&XN(Hoj zb#xv9RmpV(62d`0eNrHV%tkDS5{+nLk>^epU)J^P98CcEm>RswC59TxkEJAEfazR(#2I?sWA5IzBfUWE&Yzs}A5t zc-X!>M)(9Es;68<87#|R?=^`KpLci7?|)FQx_Ga*VNc)k-%y$C+5Lvb3fwcmdqip2 zzGGO!^j^%(aQ$rM({RaYkYLclWlqMZfZc{b5Dn!cB;@50kEjWN>GU>lNKw#9c3`-R zj)+gt!x=f*BnWWl5cZJ5BIv~HG!E{k`Mz46Y!G@{b@jIJ0U6j%8Rmgotu*3abSbp@ zSX(By{`(|72vD+k9z@Z~E`?LxZ=}*JdZ$^}uqg=w|5{XrN}>7 z?rzd~)X!%|oO`Yxb=VxAT(C@o32n|N`2sKDE6rR9AJr&uVSwNS{mVm^{1BYBe=Ib@ z0hRlU`!TKBF-~`aUb0UQ!09Hd2Ral{g2z)PnIB%T|P#%kcI_E z$i3MYC!#HsA1i9(vBJ$RwNF>vM_o_#5tU<2bS3Sk{hO+qj=XGHdy86(Hs^T>%cV7>#=v^O z2jv6@1!5f9o13Gq!fi2{=j00a!BWK>sn;gYNz_U3 zZQsen1;5zF`)>ri6=d#n5~!e}X1izZ{Cado)SkV_dApSQZ`k|a^I{nLJ3%7*xN)J;UcElcwN^H-8ssYXg!AP zAZqvHhTZ5cxI1F|n{DH6bal^l2K^aoLB(i2-R#{I+UO+O(iBk5pq%k+pIG$#Q<>qI zP0=u$VFL^I*PaHR$)3e2%2oa7V|%L)R@@hN^F~uqhRR76EkVS~L;LaKUvVDJj*KDS z07XJWs!K#3=)?vCzCle8LB(d`bWIPOBE^=P*D3|59S7m5^J5z}RI~GSwVH=e?fLIU zMM6l&A-y5x4n5@a7S`eTB1bPTuXWov1EOQ&0wi=hcKf%knKdPqw3JMxzLOZXY->N; zLw7Y*=JLXWJ&5%vdR2o+xr7wjb0g-&XZ-C+bRA6c(FZ|VGcgNOt4(ODCz>@z%89z3 zX3IhJV!~iQdv=W~`VOKHyQR;p|3duyuTl$@r60fN^~cVYdu#9bSTz8(8p zuOWBY(a6eh=kKjOkMdNzLly?49_5AXPwDzK4?DBmWY?Ek5J8bFzpg1CE<5j)HPqR~ z!GkvHAF2b}-jAwGFLmEXA}MH(Zfz7V8be*h>f(DJ54h97s`@e}k5-;MAG(fsCnBvU zr{$aBwH#UMx^PvNk7F{#p#y{(_L~z_L@)h=$!cG>_N>H#+ONP0jin9jy^vBI5_&a4vTPkTK8_>|Lu24D8uj)=6F*;u_vtU4h@>1+i^!@gPSqJ z;7}rvA9@m-K3~`x)WhlOmpuj=WtR`^xzEKZCC)pYLJ$!mGoq*)Zei-w;2GD(csUSv z2;serdt_{oq1+NLTo!8$?y9orX`7wk^oXPWF>lOp~&vM$* z0pD~OOXHNGs*!i$QeFO@5dSzF0M(USV(_8*D zmwn|Z45nw7q9EbME(2fHM@T+3yz*K5kX3XV_IauwW4jKaO{08XUDv+X0l z*DEmNydETn4UPL_#6Ydsq&MoQshxBT<>Q5@D}#%ON*Z42Kts!jAm8uvAyuS&%hyc~ zeQ#-y^0prvf#+k}t57d?_Qx%%eVieBFM#gMj3XSCLG&I~t750RmHiTJ1Z``*xuk$4 zq5wd=BP2FEbvIXC_Ix`bj5S&-u7u~wjFqeaHDi5;ZoZd&;p@xn=}{sewy@L50*P}($`u4{gW62tS$>5# zQR7;HOf{=1JLFD*o<~izTqy3G?d=H1jQu^!7N^Mb)bu#*P)@@-J1(1uQLaP(+*zp9 z9_c&zsSNE zCn>C(UWy8Q6*qIxetvo%R8D*Y#&gOH&8M(6d>F3%m_#L9Os(rO-(k}w+dK4u)DH7p z3iZ{P=Lo9Q5g*xYn%4fyIrB^(Si`Ew8S?&jgfiD(!rH}`j5m)tlvJjfwPkZ0!O8qp zX$c(lr8!!L(^pE}ySh*U%i$=%Vh?*$aNL%n$)RZN8@(>66=5KST~%X^B2broYKGG*HUD*zrD@_=;;X+ zf6HWdZNapK{36u{CAZqVvnh$=lWYQQjq-C7yaGM`aW_`@(AMzwg2SEwH$ce0x+8Cq z#6mKlIA`Lit2Dbu%TAq?Tvh!br#)es^n6^5FGrPiUkr-O>BPOaTI#xla*kvjn*JT5 z4yIXnzx(D2PISh~m+1bLiFH*wGNl^a1)V$aS}#1+Fk==}WKLxA{NU07In^~77;yjT zPImX5>>y}2iNkaL43Q$eFP_qUP)c?+yQCH&)^L%1`ku|7iv?26FB}D@E`8g5su?Kh z2!WQ-)A&BEo6e@otgN{#AZ>OT>3LcS%5oSZpnI zNV?Vh_8~M(C9AHhgIMoLvTM0M{J!gccFsMLDB0MXm)glDgsmAgAkh4Oa zrSxU;#0BJBR7t-mn&6-6;vSW}vdrxJV}5P#mmX(ukcTIRGGbCB{N!ReA#Ur@OR_8K zYTs=)r`}3ofT(8yDpzCi*O2Crxxf2r@)cbb_NVG-m56@1#{Dg7M^~Tsv5UX?DMvx);Doc3!ZpEHY)bs9|sxP8dEYF=I@7DK}0_Z z{)VDQTx|}%i1-ezvocnC09i~~)bNht3!qBXPngmAo(!3b|9phg+x@VhpCA_mtSuKN z+}tPVM5Ck5r{8=I30HS@M%&T^7G$p-c1(`0`CEz*B>i2=r4b50BbceArj?3|J>wzl z!8kV8Kj2+^4RM-8;pp03GnK^^qGtU?I0 z!&tWY%-6~_9v?`It@Gn;%e)c4{lhUGVonL#Y2pb(!%AAyERy5ONPj)qdz{8KbU#&I znzYnl44u^oU{6sU8BwRwYyR$$ysg6=g6ALoJ5(?KM4odAF)*~13G<0hh@V>Ux|%qZ zYOm|KCf)Msuz&@eWU=-PEGGcF&|axU zU$0@_)d<*rq1>-=FVpkB7EDI_Jb*$Dgary$QVnN`uT_G~_&DZ3<50VN?b)hP867IU z;Evlbicni_&>OiB0xGw2xn=g=aG_;L_R9yh|E>1xvUCy0cSuu1#n@tOv|e~z(r=|9 z@+A73&aw^iEx_7aOiRwb!G487orjSX=&VGGA_hnC&BVV@CN{kfFV@tRPNq^l{~7aIpIgzln0wD!)+=bZgI3Hp<5Mqb^DL{x`SEV>1A^5( z8o2Rdv7>sgBk~FEmv2|VhlAkdlW$$%57tulgBlUFLXa7w9+Gj%pO6C-~^fr)xCc-a<$b?@+s7@hW3mXMH%HO{Ql~`6qsW@9+h`5ZDQ9DE_Wg zB%*}&F;!$M-%GI~9yx$NpDt<Lc|+44c_l`^oR{E{XB|V7Jm2e z+b?sPp9;5WO!Ip;J#p7qCWrySvGDJHHx=bxuM%s07b5`Vqd&auDg%h6N2BDUPF{>G ziq2bEBh5r6P7}r5dcmC{-8D#Lu?S8`=;M8Q8zPIp@FPuM-=D;b;a$I2$En!cekN5b zJlb>VE~%~O*)I)S?X}RTtd$bX?MO?vHXjX>?SwZV3b5%mBsdbh{UtYN5m$D-)+6wd zH1&^1%P29f2%3PDD~1;SFLjh%307=2+qov+c%$kP%cTdSzQ z!&x{x@#-4KKgAjt&G`@4;RzG?r7A?c1Hlz@6Ox1G^DR^ql(0^Jzf-!-{C_w za>@tXz+33L!SUS$Ym?HHJT~&D(GXNc@#rrPK|AKu^600(4LKS^G_j+a! z@!aXI)=6Q7w zxC{xuO;ZyzAV)#0C}O{U4%Gp?&+ak{>M>uNS&VTpw3k&~*FKsjEPo@|_LE?t<4d9d z6Nv?+piKWNP3y^5sFEriyQp*JnDSOS$!kf)K&4m3vF+?`0P+$$eCQJLMRU|>icF%4 zH$4aO@*%UHJJRFC$pW+8BLLf2jRZsQht$f?nQ`x_x>5#_7n{vQ6Bpiv&O1e=tP#q3 z`FlQ>xd|@j7*gk36a)u1*#G#yjRtPH8!EpPiKoAd*UR>OG}^}V0B=py9xcX(-bdsZ zYnE=`(CW@cZJ!<7^?7ZkBb+LSu6lTnGd|~<0_GMJXXwp%#2a@-z8yI z@@>4w3=UO|r`;>9PfGGAXf?u<90^e(CU|e$B0&`8!NuAowp@YY@awre^PG5x|vTbg!1q}y@)}A25 zL?{=}3EGE$cAehC$>9@0S?YXt9glT+>-_S5F@He$%?8+=3=bhS6hVVtKO@2^5d@jz zj)S+`_ZNyxRX#=XVh!xnAfdA?H#^2l5S+O*z8W09G3ADv(3FU=h#!Zyb*h3Ec`DZe z$sd)m2!%k&NOL671SVNJ_rTkeW4$swmf`S$^%7PBw3@{cUWBX}A8+{8=6>-Xu~=HC z#cV|TWq4bQ(^ z_}@A$#)pKY`Sw75^&y8I2>241 zzSOA2g%2knAkZcI?6GC8(;JQG%XIMt+;_e`O^>NOUYNskNJlW6epWx3p9+WzKPCzSej+syb^bLzZn4D;w(1zS{%D^eL8> zG}EiBM~EFCqG~T{>=a}Es?*MuexBYN{PGF+hJtmMTY0|$B3qri8q*0(5chuW73~1( zXx^FPB7@ClQn%Qk^B{>$4aBM|vkThD&ie`$UC=(M6K4UEN{_gH>+mMuZ|)o7@it8O z3}Ogjl|w7IBVqQ(D1P)0 z=R#~H(Zy__fL$vq3(ma5&ut|8U6AwjayIH#Ps&*Wc&}H1Mw{QFRRVE4f#vWIyv0`o z1l(NLXmox^gH!1K#r4`|jnQvO14q~xPoqAUraaOP(s~W)aMq>8DFEnW&B&<;`amey z%l(ud6kNB7YOvB*cHe`GP@rKGUf!PE?8usxJfQI79q5&U(v=&bdX7-%tb#n>`UG@b zw)|#TqCKkPLwNRnaLZ%$o`&L^l2hRc^t|>XC!jpM5LYnvzLFMALR`PQZvy)Rq?ylL zy^o&X(t`TA_)<>fhRxmyIM$x^`=Od(s#KTuW8$me_yp1RfDdPkzUgF*tObu#x*%|Q zWr~l|&4A-Yu(qnVr2l66_($Ac0_rVC2G2`u7yENHzL36axim6X8&=o628tqGs#!G6 z;NcsYI>>F&wkQc6Uc9MqYvbDF1*<^%+??!`FK&RG=Ip`$A8`Lw;@>k4tGJ5jX4Msf z`9$5E3un=!XNo_dR#E`odUCLpbwrTA0kBVN4v)gzT zt2>C@W)>I{G(g!yVg8ms2Ki~B69mblym=DeiHFzG@LQmp1OKZy#gKrF`UwNLikA{k zua+G$GO7q9Xq1~NLC>yE?>Uk@0ub(3eNq%@^Aw}s*CFk+5b*jS{g^PLtBd^c z3>c$gM!G0V{}dav`$IW8c%0WYFqY;QiE(27B}Q)u7}%#iMW>H2oUt9@SXQRc@=y5f zihHq!i!s><4!AcAMw&e|5SXqtM4ZfpouAUNPYx7}u6(dg^u8EETVf5*>GOVAB_Oat zjP*#w1B1`4=Uizm%qLXTVX)4Q$2l^h*AKyme`el!r+4#5MDchLK;CpPW4CT7RR`QT z_4NoMwem?b*lt2mXhXHS`uhepb(sD&u?!gBk8$j`dw7JPG-x zOdDcD$H=P+ZnHrCVUi{1<;#C1^XPQap&t3B^4HG1ZgGuk()t>`vi%-I2Xb9`Snvlf489_54W{t1hFwv*4X zI42{8Hoxt*V}!XmyIrn4~t;~J6A#m>RwEe1d1E9?wEk9?LB zkLyjRe@K1KilqZ#2+eQdeM%ZwML+JSE@fL%ZJ=lCN1>`8w{jz4Hsb{r^wuONuUZ}q zR?~mLpRMBefQ7ZsI5B+%C}fCnhMsGb5r0%Hv8gTRA99>Zs}Y~{`Z5gaF>H_`i+}uo zj?0tgOyoW0(5YAIYTNeqs7d#GY;DQ1NWKO^7QqSDrH$?jorWH{+vRr|F44W?epCJ2 zM_T{wJiSbYsy&}orCg`6?bn{`1Nrkg)qm3Tzi9$=W$Tm&NekIf4OV-_@E#&ZZ>@0Y)fr*}|W^7j0M z3KGgq(dUpvfS#zj95`!@U5k0$1@?^(0??k*LFmZXVdM$c#c`=k8r#3QcfY_`&FTd>7L`n|1ngt>){RdX@cHX%`yf*&x zlG8`NqM#~^)KxLCN6aWoh{`+Tfe|=H=$%Z?sNnjvGxQ{+VtUm%ObbM}7Q^b+4E-co zN~HM(WQFE)wcINZ`D-nv@lDOdSS=@`yvxv2?^$+TJ21_&f;U))aLd4{=ZpZjO0-yp zds@z*dkJ)O%Q@#a_d9(P^dxKhUqL#c!ZiETd)~a({~xY3)hCC~2XLsIP*D0q{SizG)1RHoX(XkF1zG=naQiQ!@>4-G)!Kdt{My zbdTC}KES_+OK@Ts@1GA4ejj>Plypq?gdH7rdFhAQ`&@G`L$5dgzE7Yz!#+9RGZEQ? z?ub-ql!^qCJGQ?c+?9xpXh5CYe`km7T(4yzeh`BqSs$a0muY(j9t(I;o}mfRPODk_ z$#N?B5TO`c(bnsJPzmM<{<3No1KsV^?K|5fk0KJ}O))q))2N_-ux{^!w( ztxvp5B<~3s!D0>6x=gq!&wbhju>RTaBOi$)nYexEm^&;}QkNK<_MPL)LMO55s;-30 zjfN}@|MpJ?+17nf3Ymc$ZXHJ&cmOwb85_b<*nJ*7n2Hlvk*Boz6Nu3 z`O4M^+^QctXAjF)b8XL6SsT9OoEw7NIl`3_Fb9P`K~-q+=sfnFeNNCFj)e~%AFX3N z`b2lmw|hlIq`Z6Mg|w&f@YA>YA?zaE{GD-X2G9DLEqe+%TD%yUUae-Q%?RF7XPaF8 z@fRV_ifB|nf>f^bU0S&lbFy?|bk>}1Co@${Hzr;r?r=^1?($;gOAZ&;Pc1*M^fw0j zDtcj02q0}f>bn#c48dPh==58_x1(>FE2!`Hw}?IuBYb_`Rn94;+M1LP`$YYrtLWpX z7H1x?(*(0ybD}<@%SOnW<=o{VYqQ&Gk3wRFuvNC5KJ4yG^uj9pDn6|{&W;b@=MwokemW}!|+<6M^xg>KjT_4obsFKKrb6-(4>Bp?dMSnS> z7xqhdm|Z?{i7*wua6{3Gd?asr;CMX3;=iQO)5r`q~~ zR4+Y#&kjux7COB$L)F_PG0XlisdKQv@3*kjS6bWn^gof#h}suiwq;=+xm9W8ZK(EC z#B((vJLm)eebt&PpzKYTccx1ho+5Haj*hhHnS9L`9+?gg4UkUc{MAY>?MAketA;;;yH!tmfc`QOhW;=-Lf8d`9__Gf!ocZak2fA8BdQrc28Myposoe--^s)$abwfaV*wL(yv(9#yMYsD_w zx0YJ9M`&YJ?7gB^>>!HN_|-9&j=f6DH^_=s%@B7^699cD-8Q#dsfsH%l z%I6;zhj#J&vZRpcJ_e*#cS3xDO}=d|C)Oh)F+NWCixkBq~0#n~1w}4RK$oz9^yDZ;P#2z)VjH$L%HhF`Q}L z6_KmhF716`qRMy-b9dpwZ~Wyj#iC8>h;L)`H^L6E@h=r7j~l=o3L|}rI*4u2jGmWk9_YpGYy-Qd3V43nF=ZlbG&Of-t6P>q!Kg3kl6T6$?Q(Zhz~{srsRNvnuo1lJ=@H#sio_~dFAtCSk*ZXh*FF5^5j4WAGkgX&M^l z9Vts8nWc2?;wP|$7_nW66*5V~KdI2M>I`FFBtT;D$_ydK-$%%n5v8{?ye=OtT`yK9 zU7}?mBuSSwdV!yTts#!2OHO}5c{|vA%BcNmzMP=*)$d0uIcMeP6A%luPx6ESvNO&I z4p__#!NCv;*8f% zEVwO9NqNxO#!7|2ge|FmA-ecE*Oq4eUW1d!@k{-gVHWSxhpLE#WfmhDl8l_l;EA@O z!<2;?kU_3B85S08{aEr=QOZi=qe#$vpqB)V!PdsVmt$z3_;0d86YddCC_|&aRF?gM zsjI~usz{=7GpeYlgm^(6MGt@GCnZ1b@iOuoNvmg3N& z^rv;iqn7pu;7;BX*2_BKJall-$g*d6Hg?3D!)h#US++DnPv<$0TTE1WBC%$4A$G<+ z&iNYJ>F3Ey44}#h+L)*)bcs^svOHp8CrJG%*F+b-^QJzQBh71nRga8XTTenn{RyW5 zQTntFravLxhOuj;&E<1w#iWukM?#au(KGE1;U38EF#MxHB`|X5oyCdD3PaCv?MkX* z$!F2YCmT#=w7ldd{9}<;Fx$nCF4N>jn#gjL;8wR5l~h14(J0Ysb&*+)3@dYfCSx5J zGSFMD7!?gv-~XkvPrR&!yARiL(z#(2di~_~Df#_QKaV#Lc*hk1V4lwXvU7dNkrU{! zam2#jv;bM(l*+F5i`b&$5N6p3{Ng|{v|u^m7bfFWzO8MFp*`{v=HX;?;q_bP(#PI6-iq880PS><0GhZQSulE#Jk1wtTOyM+I=FjT_x*sg<{M4k z1d!tJpA!j@PcSk-cK}+SaDOaKtuFG5EEeGju$~BiO-B|k2NBsAY7%kb=nj2CjNaQ zE%+tg9Jbu)Kvv{x(qIcRyaIq)5(?xSv&#N#WT;iG3N>C|0a~3bLcJWvtqT9`i%~d1 z00}Bm$Z;v%*{(}R#u(6;RFnQzVfUlridDvnRcnY7>An5nqpdGa6RH8C2>u|j7lZ_^ z1-~IN60F%R;Jg~wqe`;4O9y;F|6=8^^4<|CYuF9j)n|${S|a1jFm_%lTAwe7@HoII z6w(InB*6%LgzK9|$SdUBsc-Ld=}&fhJwj065RytgJA4i4{mD$Cn<_VC;e4$fMnvn! zC8g~#3t?9WXyrg&*co!c5Tp)V{49eDJjT_Ey5DEjXyW@*v9or49dYASg=a^`cW%$R z|5c2qcfEuNbE)pds8UAF#i(YH22d$oIE6k-7v3wi@%tW(7*i?iij%}VRFX@O*4!dK zr3BsF&S+un!MV#-kO%flu5o??mHTZpceZ0dp z`5Vei`z)5Q?Pli10jFh8RmA}(AIlO${30^}T#Ya`2Z`JWo^b)Zs+Hc$hNgaozCl^a zn{v^Y_;zTUDssRB#WKp)#ydtL)qiMmq=nW#@2PIOzF%!-G46jyr**=8@IKhNy-s7< z$QELy+b?In$6W^hp|hCBV0O3==XWQQ7%vDNnC`qOg%Vh5I(-i$K02!|1A#izT<{EK z4el11D2+f=-c`Ys=72+$=GSpMx%xOoiGppJm|YW6(v`Kgbdv<1IjJiQg(b0bEAKr@ z|1ATiQEAP|s}88fzHXzM|D6Rujj>(6zj$eDZB6>|`39;2?W|`}xnUxzykbhC?>wI_T}RM4d{4lMk?;dV9~jmKhyi&3 z?a9}fu0(wKE~zIR>=QRYFINzCul=q=JnG+0K5L-N3eXhK-XQ`@NO^BN&~empo8f`g z*Vbo2p8Q_G`VLap8=sc%mh_nphML`y^2W~XYAos-%M+eQ@j3Ct4T;K@Fox z5G7oxf9t*gjgW`kp~Znv%|GZ$&#;PXMwDOMAqNe4mEf%DVmlHLR#HCy9`0Fc6AKM46K_9sE)< z?R(#HO28QzhuU;J~-vpsf=HVkN`LM^${0RV!XtZ9A#ElvXH-MiZ+&Xk%&XkNL4I_R91 zUl5$wQo(bOf!;V`2W?CF24&}8$CDz9Dh?0tyfScpb)TM%l6<~PMmuEA?M_D~k4EtU zg|=7FhF-W-=X56yuo=@A7eXme#Q3)6DzEs{>NS>g@>HcXL}`ni=<2wY!1|_Q5ir4k z+)A|>^|-|ip09fWmW`3y)0_|A5O(r^_jBm*!<@*`p4IX6v)FAd&LB;5D|j%~LZOYf z^-@bB%Sr3l^2pjM+blkZ?jbof81OTni_%VS3wSQqJ+Q5cUZf8p`g2RJ@NsS8z*)yQ zo6wX8;zxUSO3%7Goe`jFTJ9j78!TIw>%kpt%Oce1mCp0gV+wxyt&=5o0}sFQ3xB1L z5#$qnnZ2~d);7eme9?~o3AA^=de?d| zG6oj&N$~?uCsPA!+z^|uWC6Xkqfaud>`>=qcnhk5*E-%1@LRq%^(;_zL+)< zyVgm?lrG*RW^WZ*tZE~P|2BJJCdUx4dI9;z-}Nlms@U4x<#4wY$mQkw!VV4_GQaan zR&_j_C2?#fXqAegn>18tfM2gdvqT_1C#cd4H|lZxnnT9Yz)UZscJK;WO82&*eBELq zE`oZ!edO9@lx3s1Y z5$@HI849Z0VRe)DXPbNkM$$X}#-bkj@G=MfC1#LnduEf=WiU3J^2ZD=wtMer@!F|x z&V$$4MB>uV`FOE13Y~-tCf)I(%qH@-v{`X&DtyQtPJKD+j{rXbQ&;sc5@i(yNu;Q7 zW-@x^cSVxEE_wkHRq%V2we0ePxV;NO?6)x$pdN3KL|I5DXk=^P(fo#_fjYY+j)w5j z*L7-Xv004-EBjPImz}*NGmKEEc0+gk&B6$MwBNRCw|OM?6U;C$JHZ?4!HZ=de`0*^E#}d_wqw z)r$c?ro%BLO?QlMX)UZHK~<^DjOayW1tOXLl*#+fF;Ba2*oMpd&igG{Ys4juYX#^A zQX#zAE`Wp_P|Ey+f@lM(Exs1F-9JJWe!^t|_(rJ63vno-xUm`bsr@BCXtAlgPWorKr z$iW+NzOf^wrDC$fff7QiS1W0pldr`~DYCHzZBr_a3U{u^D234doa_kgCK_onE&x!^ zpkm;~Y|p{c7#Nu@+Z{^^T9{8s!JFicS*Ax%8DTug64ddNA6@d6^M7gyU997HPIkhA zBAD7z~hG0N4c5f$=-<;Q;9sg z1v0h{c6nlaQ9LtzA>Los>E2CVG|@ksr7Lh#xbCxNtd?odHz6)PIUI+#&cweez0|KP zUw<$g2ncaxJqy%!mBlIh3O~-6^N&^WSHV+rZP>oWNI<;JzA+n|Feb_kIoev}W7Tr=j*=aX0{&}v{AxE)y#GN`^Bx?CV( zmJQO>-M^wGBl+fF!KmZlPWPRGhMkbI-975LL+-y1ob_iI5&rJy**ScO%BB;a0A5LY z>7O<2prQbDJTPtJlQ}I}39YN=gJCo|{bINjOioW$u;03~Bck`vo^6noeM#U|!nH`= z_IFP|m9DQL!kz=*$Y&SZMw);ZBICJEv5%V(t#_X$qY7d`-1?MSrSk?;vU@tB}3|}pN7Lv%TK~_tq=`F+6cQ}_w$lK{0oUtN%f%*t$H75 znoZ$Z#>Gh-1d)TJrz<+yH*M$iBw2&(wQScTorN=?KW&KW| z0K)#IppzJJEjEUWklY~D3U$3qSo6awcnOlUDofDD8DLz>XbU9zU+GWD;nKw!3AfOW zyCzG%vmGvetyO!wG(aP(Jv zRFvWp^TCXPH*K7IQY+xWpu|Cc<6?_z5GZB-d+2gH0$(bmr|V!p3+o7E>aXE0%zHhf zj#ScF?&s;>vA*|SsIYQg+)|SGKxBYjEWqN47&Pje?UKn^V6`Kfp}_~cu70Xe>c?xw zhk%NJ|Jf0khRPF5gK3&u{SJ-tLa@6hT< zFuh8=L5e_ zd8913Qy;>0k?RoGbG|ZO%8EvvZ@Ba~=>pCxI)`MTxlXQXmX3xmn|LRjiuKIH-*@p& zKBmOmz*^RH0)cTQW^wg!%+9Clw~OSacXX(Om4}n1PpMowJ*yh~{XMyHTIXLg=OiDENA*hpN({*{xZ|ET%->Yww83)Rmi&P4@%1KK1vLtXsSNEoMg z5+#P+sc>KQpuVUypfH992A5!x*AJ(4y`l!-MkwjWdbWOignwblezk-X=RotyjNQLh zp>`CH{yo3U?}v0iDu^Gi(ENHee|KfN@_rJk{RR?)=@k>P)MDbwP#B$Rq}y+Obq%@m9OQMvSmeRS#{eQ7fb!Ir|vYWKH4wplyFbo3*Nqy6kz!#A>)Pp@{}`;t(qK%wT5qCoH^&M7B6zP`gWVX- zfziM${ZGDFI?or|b;GF6@Q}K$6@S6wqhmq-3#+o_XI}C=jF=RyGw)>WNB?6L7zU-C{)n+5x-`bm7ZI-q#;b zosb%$sR$X00i9?q^_LK~+QyvVlU^|Ek^%X%&lB9Ib9OxNZDxA8>Ck}O=f@|#r;Vns zAN@y#G90<{Bal*4{(z&IRMr@lt^4`ZvP?z_RxND=JHC(?|gDs2LbgL@-sf0(d4`7S^AR_QC?a*Uoo z6^D0jnCve5#{$R^N`6NxI%h}VkUvv%LdUmqctw$S@Pa~q-?b{xeX*H)GvZ`2b7to? zLlhl~&BN44eAJ7+`P)v-Op$MO1)w*bf#H0wU!HMA1ToUkrP-_0^nf9&lZHS40v~7R zwI`JPZ^`=fe1e75r90iWU}wOVg|+98+zUcwf<6O)-qz%aFiI85SH#9Bn^+2hj!{BL zc~SVbhuyhYPdDGGh9mf9AIb8Q%%&}9=lU7ia{vp$>vwcjC!AFi|vV= zh;yoF-7Aj9Rqf!Kp}sa+FXNT|gXOUB3cA~QQ^zL(0lg5Ib1H*fYKcTVm$ZC{qF&s`+Y)m=?H7*AmXfv%f2r? zwZezu)4_-xh($)CV4Oityk9UN9r0sO8}=bq8>nl@zL27a|GETpFqwF?&|*KvU+UY& zISo`ED^#_er@?$jPS}DLuvS3+5{dcmQoYWX&0L<9Wcx4=)FRfQ)sv(=NfC|Gs^`~n zV|Xw%$((~8ZqQlZ*$k+~S!BV>Ag3!RgFb{Tr{&9GdPZgZ8Wy_pb>OXlL;Umw+k z*Hssn?Wl&|_Bnt7twG^Qwf^sB9M(|rJl0nk9kCBrWd84*$NpgQl_`zcb^%nAj9N^+X zv4y{Xo{etS7h_^{NbPjtRM+firl#3T9s>)M!rV9ImvVI-s14<}`3cRF{7(H~mCP)4EE zl4akK37az6Hltm*Hh(FA{$yMaKUu0snYG=xg67q;{LPx}*peH{r=!-9t32GUz8D8( zW6lt+g`F;5wzE11(H4x;yLvZB7y%qFr;&u}+{3u9-58pQdB=a<(mi z2B;ujCHdNz#ZAoI=9p52bHtnaalIO{b1H$&u_aqQ)H8FJ@iN#RG}^pK!7~P5Y~*P1 z4kJnhNBeg7kGR62402Zc&zO9AR~S-eA}T}t)cV=H2P^xuHP-^+%yf@Cf6RW5M(#D{ z9~97IW#MkNZ<^`Kyx+enqDVE}E@t^BP;T<*1w5vd6jtQ?k9F*87v~bekedH81zNs2KJ#i74AH-TujiXdI(u7(*0{c-u~Y zirohZ`7*OImvV}}F7JweERaIwmgRj&9Q}5xl+3#t^SI<{HGNKu8)Z;#0RZ}>)U7yf zji980-Fc}%^A+<@rCqSlAA*6KBw)q-pKu)VmPc)t+@rC3YRx;e)|83r(uhlak{qEE zC+Chq4My^3iY@Gbtjst5X>Rzc14xO&n`JmQ5O|m~lruZKOWVe%i%J~H#*?Ju%#`qnyH>Sdpilvbl+mIvGD6bZa6nBQz7 z;r}~H)62ahGAwQexe9?htx$Ut%Aht&En{!jl3H(8DFX&9sFB(I!YHBrAmeDEenbEV z2xWkZPjo>JJ$noHaoarHo^_9#oJr3IG?Lsf{(c2dX!Qub3kw-_d5W$C#76kMq(K24 z+v)m)s#gGwqi1P0O5Fc2^CI(Ch&->zlefHKX0%#;Tx_lVPl|eCTZpBK2S3UDX2-uK zp3X8`ct^c767D^N>XQXdGP7YhKP0m`4haNZy&E{%zMEB- zJKOfUe^O4l-l*@ymw#8r3z`M&|65+cX52cbOi^z+>Lp(3Gpolb^Yo6Wx*z@pyUKGu znkPHe_8{9mXh#FSiXlp=&YL5r{S(C4sc-pqQD^LHy72A|sruF_npydIR6rkyHg4d_ z;qaD?%2R-W=Rc+J;BqN!1vm}W>4aqGqew5H;qSS*D3cG5Bgz(MAJU!bBhknP5lE4CDDp7Kwz`A@Pc_nAxw7~Qk;3ccn0ww?$6M^Ulq?%&ux zxLIw5IFIzZ_gX(Z@p5cUi#f=-Mao#T$YF4Jk!1LJjrsVi9-!{Zgts-o)EHh{WCXl7%H#Tuvi9d_+1bzTtgPin8}J z{q(cd^8!?XHv=fF(y2?^7hL<8rHl0R71~V#h!7Nwqrcz|f~ z%MI4NoGJ$xE0%csU{H{eyYZixg$_FU1JZNwnr84Xl8&W!lZ7qLE=*60ATn01pEG#W z9I=GeA0jYO@6oiYYCb$KMLuCAwlrO@sDKkYvt6>JlT3WJjtXNC-$c{JG8cK`k{D^- z1Hyw;V7WYx7mvAL{>@QRb^nH22-{Y$>;K^_Rv zUXE>mf1>VgKv>6x=$ zJ+cnapgmSw=V)$fvK)i)3H58_Hh(RZKJ+XF5WO|^csY|U+?SRg<^3YW(C&uAacn`> zEagU&Hh;HA)BfoAk;CVH&XzRg1yKjY*4lnJK9aBxEx&kcj#wZ}{dexL-%rSj+bgXg z{4LK0<~!mvqR<8zTP~PH{AFBL!SLiR?w5vUG?bHEM+|Qk=t)1#qVsEOY`yoM%oQT4 z-+a~+sWetB3|(P-*Dt8UPrn!J8&iJMg*G%tx>eu%?RNEi+>}R4Fgd}daObqX*IR&h zYhmW_&zxfDnKboP*~W|gCgU>YB2f~^sKI{QWHBOWVjeLdd19o^a7Y$EWfCYQ16%9| z+NrKEP(!pJ!d9PsctFlpg8tAduyn!ffFu{t3wpi9l~$>FFPQCNL27|@u+m4H3C!)P z?Fa1SFyRdIS`G z5cag5=>cFdpEc#K3g)$QK9rO|J727{9T)&HLgu;qJ)L`bk?3fiHr(d|HF3yn6ehBJU#qW_fz{thB(?Eq(yAyjX5pD*ees}3rhc1 zb&zQ6?D*tP5y{~_3Z8<{lbV0A_qzN6*{397(VPWS8z(+9?8@8`;bo-$?LW@n@6p@3 z)v?=k_{{aE(gtkr{wrOp*b0dNU*IJIHT7UVZRJ$5A&zVfW{! z&-F(wREEyZ;Q&7)*_kD7i1V^-ky$5ZU1>rrdNCG`B1{;HlUVCtPkUKIRRYGonCMrR15xSQ&sJZ zw!KwA(ta#X)e`0#JdY9F$xJ@FMdEY ziezK}H$ce0{mJ$d0Cwrh`?e~!3xL&n0UHcE2BaA;byc6H>-0bV?tP1arzLn zEI4B>HwluveS0F{IL5nIlXmVK3XVMvTlNXtBq;MR_4vIfGw}(_>l{ajci^UHr+#W&f<`vx{dzP^rUFdKJsx+iq{tFwlgKF&=Zr$tM@@B&L@ zeNU601;vjx_F{Rrx_=opI{lpCk>R>nRnUzxR&Wr!sIrCH8vA&9=Fm8x zb4W|-7hC0C8ZPW;9Z(CBuZsk#~UFThzz@oUIUMw zkbW`qxV6P%F10ppH~+yOI#=d?pee4)`*EGRP`yKG`fQ3E&m7QKxGR3(jn);_dTP6x zQcU-f^l6()%g%f5?XnCDgY}PxEB9SRLD2UO)1vBOJo$ApNaK}1a4Eh6=v^~&b3V7{ z?WoUTxj&wf`{f-9gaRZvs(C25>)(}+q0OIuST^sEK7j>;72F-k)zx?xuE!y~Xri)r z^8B&G2j{q}gs|Oi-b1f!bYUE5Z>^qiwc}I18|La^ZAZKz4`25Es$nL}EhC_bblsO< zj(+O}zge$&Ph0v*I_vn6Z}mB=gvsi}XgDR$r1NvaoJOn6N6-B5XEFBdHw~YH8~&@X zkf$4z#s0QNumO+LA0@HwzAi>FC5RC?*ahGkf7x{sU~ES_y->PS9)dK}f=z4*)bA~J zCFy@78EYi&!EI{&e6B>P2KTJ?>6xyy^=vg#E=sU#_3+&mMz(>y6b2!5VV6OWXf^H` z`zY3xE7RapMxNlB$XPUxjj1}Rl3~SpK_J|lr9%O_${x__-*A7!dEot--5GT(s@cWI zk9zCqzXij?t^X=8RBopPn4`adPJNQQH0xA5c3eAp9Ag}klTxPp0gjb6=(X|JCCy=w zV!7gznU`}~&H9LtTwq6~2DaZZ-qkC^rRSO$j;5a0wxW(!Yp7;}_~R-Yca)0Yf;Ar@ zR0*y#z9Z(hpwR(6Y4tDFUimV^Pl@$^>jG-?M*F}0LrKWJd;;USQPAg2cOuT52el@t zZcbiUjAdNOdO2HLob7)ZIt)Yk4BWmCI9+8wIDH_M{T74#EI3BGQ&u-Dq0#qphh#bT zV+e-h)qB)X9?oc2JJhiM^h-O%ER>}0RJEG%`nvbBavxMV{3Y=P=VmX*&Imp~iIGfl zW1qgAQz&4Q<+2SdA9qT4(Wn4GTl2ICR>yI#*dp?KOPlS z9A4gjBJ)c}!`0m#vl5PYeN>$Fb!wZ5S2&n5pjaj$F;C9n?>XZJ3XzPu~Lf;~t zypf+<4U2qxj|AVw{k?ir&^1fh=j!P!QKdKot-XsH0FH?!=}ss2*M>+_CSo*ug~g7> zV2@*@?TrLCw|+e_x7UD)>oAInVBD?v*HUrjmy!F2?Xzs-94Ftofxoez0d+#p=8VIJcnjJLK zjpTDvhl6VeUvAh}7|I89h=ha6(W)+Wfu>hpasvxe! zBKjV76NOU97D=XXCd4sTQi8XG_vLx zQ$)9;ot%{Dmu*Gf>j%U(-C^>@mV7e9-^UOLh{R zI*bzj`QPXP{G#*H%%wi@DF~vItGVbwVygMmBva~Z(N~qO?A+9}bPwbd<53cJ})l>&YRI|X&R^MP9Kwgu+cDnKGVG>dA>>~>_Kx)*R z2$c3j@AvX3ZSR5*Fu+)IDtgZ%w_Qtt+n9e!J z@LJusacq&b0Abu{blPV$=(VTr1w6&ax={@6@*j$S~x&fZtIlg2amWg;RXtI?q zkNVXmKqHo3Fdz2>sRP9VUnlS32K_7Y_(BSdP8Qr)ojA z0V|0G7GAgFslvG|i!3I)Uz#3#Z$n+=J##KoyPEuY-b!g+CnRO`SW6E3gCo0b2>D%; zTPwqY+1Yp!jv?1ChVl`b`_{fNFA}z9C>KE^QN@dI&PWLA!zntdU(Rtn#0Mm{dzTg9 zEdlceWw4A-joc1RM;NU|*}-VNG|Y&E+DGDmt$0zzv1#Ea4XOyhKdBxYvZk+ zc<1Qa)brcY)lyy`iBiMefF#i~M%R7VAKSQhTh1_87kP|(=El_Xk2=fsbVFMUEI}?S z`}nrh<$_5AeZTgh@qxV3E~@)yve@)r`u)FM3*#3UpXnmihclcc82^UP6-K3NAk*(- zkx53iE`R%jbvYH@Gzsi~Kxj3G#bl2oqdD=Q@*mE`%hEMS# z4+urPT8PohmyKi+jlj{{86uHN3d0Kw<#&X(%vF@bKg%ptB`M$L6`gc~ezJ?qSwAh{ zsBbkst6u-33>5c9v-x3fq{xeqM&LlvsvFOZnJoUHIqG8`ECPle{DptTS0gM#dFM&_ z*O zDTJ14eS#40UJaN^)kV9zHvO&`v&gN3hdgxVqvYg z=orY-Tw7BOz60?dX*k~PiAu5~|4_~jiVuZhEW8ZWW+KKkw;PoyNi^X~%d}Kx9a_HK zk9PJbxe}NVWM#k7rp>pK1StL-CXyu*=AqqwulJMmVXOZa4&V4;3?%B?JiWO8avIZK zXT|#W#=b_JmiG&xfuLtA;6o0?Qq?E<$w@c5g2qS|xioxFUGs!qq3l9%uj&|Zy6L8w zZfm~mLZ?(2N5$#r7a5rA55HHYxVMIO z<=Ef^dQi=f_PNvFHMj;mflwYNuNSPRuf&X@0Z!>djzz$WklxYOy){+2uIsiUoTm|@ zd}lM_!JJ?7_sCzi$vB&-!pYI152j;|W%XRv6>svxM5cFC8VlDn2lKoi1sr^vau;wq zCc%PtLQOr+#CGj;qmP-3sK2GsjkNnOV!Q&Pt>bxoum3R>M<_%1?0qP%XX4Ctpk z`@jdid}7`dC?70*$CX*viU!@V1xw<7co6mq@ z;-|4R?NPJ^LY9+HA0o{a^Fi~8wBK{*DHk23(RpSuKo*%XUS(fZB@);iYzJN&PKBru zdp{M%0Usb8nXX#TC!?2ILc*q0Req7yHGv@Q%k6Q>IYTD7?5;IrgpSZftI9`lKzcy+D10lA4m#znZb(Td zNYFA-V}@=8*N4Fe*kMgTj{Q}(`e;|r$1LX5<_j|eCHfFL$RJ5jVMKHVI5+Om5%j%t{bNEzUIZxHvmbeGM>nF4n%4_G4HL{-+!xD;thIy zcs@-iT`~;uI}ys!RgDl15pB0dpS8Q8Yvi$)~_6cz-IsOrd=9L=a zg_C)HKp7-#+Zd-8j}2<=b&H( zgY=XqGfSCNnN6SnvZ~mHKN`=YO&G}&_?<62oD#tYtqljfG-^$u&pB0w`#zr(TY2OM z8a>}(nHG^A^%}9Kst~U8mkCk}E(G_p2jpd*c{kUU%orT%UZf4~H~DJy$4@eY`7l3g zNZqrppk1t4!&*B6xG1(qUZnvYBW{c1_ z^Ni!3_}=ixFem$!hP^&$9gsj@QbR+E%I*5vVFi0DT#Ok2$5h@*$4<)tJlltP`QssV z2FAl<^>6q)+|1W&@JwgC%6;~IE=nfaP4#JVcT0VLg*LwbTpQW3Q}f}=jCkW#ER9F= z(en5bA}WV?IrzGmK4Ct8d-4JRb@|(Jw0tvogSiY?$Tq=sDAert$mT>f>hdk8l61{< z7@E&CrNpjy-Ap4H7_YjSR!n^A9hUlgSav-kSG#FVGi1YdK{uPYm{PiE@b!zMb-$se z<3Kn~!_}^sx`c1IJOs8t*p>`b@b#4~kcL>K&MJ5eWv^4hrZHaMBqn}Cze?)nB!(XI z#L7f|)$dnl-MP|qeYN1n)v2D!*Fv`f{ldy-vI~i8safRZG^La9XB4`uD514|S4DCA zdu#Xz5Run@TX|+Pvw|yyE-tUatH2!Y=6Hda@_Uc4Z_Ya~A`$N0Mz= zA{iGV?zJb-dEn__>=Agm)V1fCJG2sS6YdLtkr^>#Dh~16$9!|dC$1D+gtK!o;V~c{?3p6jboKpdB2w9HTgcdT^dw4=}S;>8fOLesYoI3 zN`=K09!@4;J1J}s5xd32805!zMx`>jbWiwfI4)h^ecDXtW2&pPv^^}xbJW?`X1}U1 zTg{PHXV(mdY<`?&T57%U#?xK>5`XanWeXq#K-0b?uWnxQ)(7&(X|kTid%fE5`?@=?S?Ft1sq3$ z_XIZwF02B;8L){FCw73gO*@}S2qXnJ2mlW;QaS$KBl6xU8qT+VZ*u8z?nx-9a|3Ea zkQD_wmIF@$$!3Vrj)93K3F`itYaIT9tI?B+v(ok&HO8tS`ncw(7LVIgde4g!juy>A zc9Zl%HKdz$%~tkB4=3p?G)Z5|l=usHn#@+HhSw#UvJ@u`9_-y)7(LAU^5exJ;|0cP z*4q(I7Q$Ha8;E!|Dg8>w&6#e>er^e{%A^ zs?1}`xW9D|x6a*EKX}|n$V-Qt(KezH{fgP%n3tU#TIjR>QOg8P6wMxOpD*pr#yjtK zm`+6I+4cqW9m3<_Sg;~~c2B*|8Mt8KS6fu7aXDn%Eu7OvJXb7i)u+l2Z*4@^^1+ma zy(^v(z5z#F3JxUYMMZAqBFL2$hagPv+n>?eHI2TU!4+C5vFvNNNd<8`buNck+v`sj z!y1Wp$TpVV);HQ~`&kxfW3)_k!Rb+7D7*jF1w5(w&z9!vi1J$k=v%&N8oqZcRI3dL zvT`gsW$P-iq<(@mJXj+=qz&M#8~R%FSoTm?xPo9F)w8@C6CS75+7U|qpgmLkCDWtX zpX3_Q7JO__#%mm1OX3O;mYc+*oZB!X>b^yu@8=l+3pZEtE=@+TNqFyD@8)a`g&u!# zb8wfb);s%^fXDf%`MPNH-^$SWza}Zfe?bPc?%wZn^U{1ln2|PpM7uO}aV9L~E^Vr8m zU49M>N@=)JrgpTVr=!)r`$%(GZXi(TKv3C`PYh^a70Nqx?j%>Z%^MDiQWAY9`tv{% zT;5p!7X2C?lZ&L!nQsSoTfQ#)6P{{L^7`#rpUd-@1^W24tR#n+U$hq(e5!2ro!^|eTA5fIQf#7eDAP~5>WM#k!rGH3UZ^YKD}6uLvpYaRqav6U|2CNZ%+0%WF>4)Aa{u{}a2!Qmjuiux94 zL`lu^!(dzN<3ll*H;0VOjfcMc#TzN`{3M7wH~|t;Mc?OLMPrSr=>0*Z zd1yUg+-(xSM~vf4f(G>dvqys@*KgaWD?5*2Ujw~%UyPkgP8+VWdjeqAHh*{4?Dl~m z^ydVhz84Dn^wu~YaF>8_E|Y6?SZA%*zHumHY%7ZnyJIk<;%9m*=iw#qezhQoALGYP zEq-TVP_Wq4r0-0y=E$&ei-OLFA(H$dtCAWAQIdO@L0E>OmhF63=Qve>e(WHl)uWUy z5;fuDV3*0rFaX>H|BfxTXv;7geA|5h4Vwy>y+W`zCTggcQdf0?jf*pk+1gG{4_eOB z#&o^jq0*#1pv0SYUTp=NHZNyXDe5yZ+y8p}WZ)(`+TI!aC3yTX)VoTFyKrEWE_|{* zY7!@XIIyMD7q;Ar{4s^MRcJ+ZV1$nyEbc?!>h#kDQ-(q{dpDiGkCCeje2Ru1h|IFrFIk*BQaX5{aPhzONuH= zX}?;dsFk8E6+4LCQYE&cHr;CPnz6NrpcSLUD6xYeVm z?1#vo-ZI&fvHJ~cEoR@y?Dpi+BDbr(>6D0?6f5`TZWPm7(pZw9B!<3qU5%MHd%=YYF=yU8n@3@Y7m&#p8-6(IY(o+_Xgfi3Q=pg4(N-A(;%g0vh?;!3dJw{WrzH@O#Nl52Dh1nA`>ms zcD2nTjMzo{N*&&?gp5uzL7FILH7x}lRon#0m zITSHV$FE;gh0JkyCMcg{o2Deo`aUkZKVAk#-4F3l^#8IW`ywug z^sn!+N+FZ!G^y zyccnQX`M$|T4wPThgaFKP0Ov+=^rYMjBAI>`JzgVm1-txu675vRIAp5-w}r+a%35z z8DbaHZAXP2hCPur-ub|hbFLSU)0UCDA(zKTB@^0cKgwZsEn3VZ*{ds4$Z)fSPoyrz zK~~B)neuqpENTIxXJ4LcxL2;Rt9At`|Az|o%PuQgUmI84!hX*SL)`sWKCruZwGUzV z%F}LPUD;sXJs>xpHIua!HPh$f4#5h8)D3d0MfW#;&1uNq+T!qmQ+*7v!&inLqHs_7fBDivx{Bh4xNzlvnh+4ZZM z8ei+inPO;j1a*`fz_mtSA#RAt8E-f|w2fsJg=}=vhmx_X7`;au{q1z{lv>|EIy#o8KdH^XbUH+?!F zCmZ#%A{#-HtwBA;!*_?xLbO)=B1kwOnN6SG_4pBF{ihcW`dLvQy4+)~9yf&Wv7_hn zK}6j5-R5w&#dF9FYX{vN-R!R3G=Lj?TA=#8$?wt4p7i3lxe+eY$RA4O@)&O}j6btm zBYjA!tl@gtltfy51KRLj-HD3Oat-OP&^x-fb#< zl>9QZ=wg<35QQ`hSPES2lok;ycasc0ZQ$~ZP_NSx+;`XIXH^*C2YF_Fu%>$|V$@b* zqhz-CW(}jC-~OHZm66x8jC|Ubx=(u$yH@ZRr_9g58rbaqG5Cll{Y- z_8$5%kR=Yz+4~N%UXf4}0do${a|28BR(j081;w8-KJmEm6-UrBUbe_zEkBD7z6lfu znIMZ8C)=}~nP!mwn1$;hXSGo(|57Y0{xy3rngUJa?#8#fInhNr%RcE<{~9iA+Vd;y zsAL3%Ho0Ry=g?~6use`&SW{tWCL|ARqiau>u(b{-h0OXnlh)&GsQ#ou2n z&+W|#@D=!_+lzdaWa+JslNhChcPZ+D)JI0?A=-;VvQpPFuZqk`4Q?7ENyM>@=#ZHVhF{vU>1LUEb&c9kA_C)K z!Ar&YIS+{q(fqCzjr3ptiPB;yS#Pv9w59NZpg%71T zFQ2Z(?e8T%yunY;r!l66Jg+>D1!t|!6xZ?r4W zCZ`9AV%g;z9GLWnCz~zp&`&s4?(7;aL~=zY-o*1fJL{ck*?(ejfgJ)D{jF|)!ynx* zllD>3dm`N1{5*h<-th)~a#J|VM~)ez_?wsvx1v|D=}}$+sdyEAl3VL z>^lzl4Ubq2UB!MJ7~ee9?xX~l@oxUu`vO~w1CDZ5MLxsH^$9Vai4&k2E{Ykqw>`PZ zsbxau>^En2YOcP6x7KxER=WM3x`--GfDzc7U(JwIZ<0>ab~Kd+mCE9ehyIgVx~@vWgNK~Z;4vHQ!`$FM2RbV zKs|tXPRMC>1(3%|U?S?I=(?FrXug%e+f>Sg(|g573#1YS@f#)8iraz?n}E`JDb{}W=Db=O>StXj+d2~`GRn1K z6>LKPe36523V&W;(atH2eZ>?d>FdNYS8!LrQf62Yh zHa7P@j11|apV?eU-*@|WnE2p?4CbnX^U&orismv%QCUQ^`3>2kZSL`PJ&>ggglq_t z={l_)BtMN+{{n^2$I;x@InT8Lym;bLc>ODky-Taqkflhm7DS?iq{QQYd~*^8Vp)tN zj5K5UX*UujESm?V`*qqR>C-8*S@QU)^&Fa{M@v>$nc+)oy0qsh*ZWQJVf$RRvCz8s z8Jh|R&-Itg*+Z5DbWNf0`~#(bz-%gCwGLjhwYb0u(qA`WC?5EX`*s~$tJV1~@%=cD z`(~HeLR9n%a)qrQc@d-L=^DkemjZX5z$9WqN8we5(Ss$gjuKMRHa~fq{+uZkwv+n4 zWad+F`ycsKi`i09lpl=1$)L zjD6VeWMu8}2SFMYdyPMihEWjtvPvTE*;&opB*7^xKNege_j5a6E6zYgH_-U*j?FAA z^_8^Tr3Y72-Fn#b>&Fy=uvIz{X4vv7!h1|>h(|+x{p(A+j{SO3H;3nMVs_>4 zwiHpN4wdTPO@sALy&)>egLT-BjvT-|h^e0&yyg>zHVXA*r$=ym4(b38`Fn@w8@YfF54#81K^g;>XhmyI1E3n$0BID2BU>OT{HyNs|x;c^JyS3ZSb z`Ms7WllP_7GaZ@MX1($r&3n%cdUUky9$^ioTn`IzB$e{&E2vxLq|?8F`V@Vw9_g$3 zsp1-*)e&A|U?HjxYFrT6pLssBWP7>n)dxHOUxG znF&!*&l-E^sJF-Y-t5alslI{=-X_!c`NI(aneb@;=OxFQqT9{;Jz&rHVB&pQrV(jp z7JH4)qCv92YtLZ(>|6^hi9}%!uZsE`d%uaUh4`%7yG!{P@%4tvao{Fei#=tD7fl@s zXoaFaB>Q+vD6hx=FL|!-bS7?Z!HXM z_s6K9&l?MgA9qM^JV+b(i7lpD|GE!i{+~krfe?9hD~J`O;l28uaLLPZNFH12;GxAB zw*&Gg00McOzjL1IHKV?{%2uH^%`?oS4)HQow*q+5F4M_}d$-=h*Q+qthg@){{ff+L zHyM_2kB!pigGNZXSaPHmwyf zL%pb`-gE+IV&!T$C+a?qzh_TmJ&%bZniOpDYWS~o?(R={iuQk?-RF`67;mx%#dB)N zY-+M0mt7qw5ZLyBHrCvA9OX&Rh|9%4bkfxx&y{>XTu=sU2Xv}6NVi`=3#zgMslTh8 zQ^F&E|GD)tKHyt%%8QxmVBfJU(*`rumH4-DOgz~#^fpS}S)7ITw_*`r2TV02%p`It z(vD9ZleHNvM;`pU6b*hRVDyL;N&_;}Sh)No!z6qpLw~>tAOJ;%yA;-3{elO|lMu3M z#b8rp;$+2Rnquq}n|n9d>Ejo;-^jv!_$wQKj*stiDMMK zz&v0{!REZ9-ZmH*^i6kvQ%>sCDu22lnCm+fcJW~K6Bp>$`{hNU0W|&T9{}(6!wvji z3e~t(9_8va!+=?Yotb!BIL#InAfRgGF@8A9CJ@G!`_2<|EcXxAJOZjk`6TLb=(kQ6 zW~%Ag^@#CS&(K+YtVcUk4ewdgB~CMI3jyJp!~RZq6H2UOs>R)$_D{av? z_3St_$zzglE#@qrDB zP1XE3&ALyhrYi9Yt&NuXn`@m;37oDxr1gKmEX?^{05i?FNU9SS*Bt(#XEdU|$pd=_ z%&xFr^}OY_)Dlx&&GyI)l>Hx&+99f^OWU&`MX_QU+J-8LbFbLqA@l4m<_y-gGpg2! zJS*#mWN)4&2^xD7Q(vNC$UZLgJZJ0W6A}2(S?gH!sZ&_Tb2-aSaIbPqM~SwPnMIK0)Nd@U zH2dJi^g##b2}^qehw^E?wy5|gTSBvY*nKSXFSnkiJiMx6#o1Kv>&j}9KZ*a@I^#51 z5KXf<=CG&fD>S#bpk<$96FD8h9`SV2>XA5^@@6H&=ggueCakScnhOTHI}pPD7!@Jq4-6wnz(fR2XA2c0>yK7zM<{DpSVOG-Q);;oEB43_KGXa^;ut@?V$6BW>`J(( zMildz1~BBH`yXQ4ZZJq^?GE9WY>h9otZr}wyhixvi`&gPFwbg}my3gT1x?w|kQ6rJ zs3n50Y8WMpxCccENu(KmZpcR8qEElU>1U+gqW{b}M)Qq}0yH=lj z%plaH4p?TmLnWQ9*NI(rj8b5mq)1<@$~mQTHc^Oso(l*(Bz5mDPY^= zusMI_@<;&M*Pj>{s$D{KT@cTtK;-#0$ajRt8La4qtU?23Yg?#^(kw zzH@Lbf=9n(Y(dY+*+OMWjs7IkYPmZib%36x3n$&zFn40GquA}dN+2*5z_e4ul9L>a z32Z^Bt$0N;=?l$)a9?gkZ#hS@tYiyn&hsOnPgRUx&L(4 z;B07uaW~>^0wQfOqcgHNRN}S7QuAD=>6-b`dJZCi>9z1KNd5$Ok5wvGMXfNFvx{Gg zPER*zHZ&K}Xw_h!3lR%LegsO7F(;jpbqOkT+TL^37KZGS?Y-GE$pM111BnLb>n z9bL1T011CsVXJU8&8H<~V zro_T)ssjV2y?Ut}A`giDdE^k;s>1v;gro7LtlJT;0Jv+D61-%{YiPpQqv{9t2JhEx z{FL^(OcR@Qb_aR&uK`0jsyMe(;NDl}x}v=8Zt#HyMY6F>Ayh{6JX|hk{==fCAXaa| z!?rJRe{oQaSlR`Lk83Y{ldirS<+1!U*?FN`Bs)01_D1WVY z)!zS5_v1&aD?rSm+d%LEH4z# z+SvO6Rpv47E+SHS8*RPQZqXQH&=Wss^g5Ss10$pTFV|3sEbzaTV+#BNz>cbK#;tZo zn0M1-8J3$aT|OU=e8ui2`;3)tov(D);Ua3S@?HPEZ-Z6PNuXOMA}Hju*Tp`$|qgc$EU@# z%}U;-)4$_y-R~8QQ+wodH3ObC3p*F`l!nX`z+ICKqKUcSd{of8W%t^Bq7Y++a8EC_ zI>0>xDw^rpb5Kc)d`il=-H!81@TyyxzP2K+jtZB0S# zEFY?Vy*Kpn^ugBAhM)#9L{dN4w}|m}Kb!C+c3Ar3i07d;R#jHTuqK2!Mo~q|H}mwR zf-K#ks2^pJx~oZ`kJDX`YE%=6_Mu&!s_}JcwLCe{Ougbghq<@2YTmJ^3$EeqiSDy|^IKqt3n@Xi$w8AxF223>xYg$WB|XQ>Qy!V)r!aw3BDB z|DhPuj+=f7+$jjzk6JGP!>O$x$2k+w^wUl|r5mAhb981?))w*K|4{D#)`u9zr#fUx z{9ci-7jeuxd_mStwxNfGm8;2AeUoP((!^BtgRR57&!tFkj94mEo7pq+H^U8Jpia)y zn32F^wX+X&^>u&~0}J+zgKj&Z45xkA#*=ax z7WljjDc}Bry;-C4N>I*^U6Trlk2{Cph~(?t@}^e-0YB33bZX+rW@3!?I(ggedf3da z`0IVze*R6rSZat0cetlfn&#wl0TsjL&yUcYy-0!dI{2zVhKgHaT~CqELGD6woiJ9( zl$891y8vrHj}>6Cl?+?=jcK~hE$862{6MVk+kAU%2YprPFd#@zl_aEHCD5x<0bYO3 zB$MkfORCo3xv=6w>sP7x;hz2cRp;hYBbj^_#yF2uM3va(!@Ej1b}UkQFJ^U`t{SAg z1<;MVx87jO^|gDv9Fm`SF5~ttC{|xT7EH}lKnuOHV~h%wAytVHKD@@ynHM8_z&J9a z=K!4NR~{CSm+(i7!Q6n#`R_`G5Sg%)yFs@>{ndz4=*hia=%uWL?gcLcG-!04_pHry|;}Lzq{7;m_u7Q8nidpYcU+qT7UK`Jv3aI?jpeMntt@qUtOkv`;AXNKdi1K5eZ4Fi?%GrU(*O5)9ny z0h$#Ma=!&OIkz)I&U@J4jOS;mukSHy$x2RW~m7+b&=XTcLurAHUZf$re^B&n(^#j=Q=f z@BG=`do|2xCnhxjEm4203%>84Z9gRq#6LcEZOw-N<2wS?(u;D%nEOpz&X+=Ie17Sn ziK0`YeoIwn&8|PzrbCvEYd@KyUg#a#c1`V+0?$wr^EW3~t znH{H`+-)i|oQkagEJii2bY^~D2*WmJXtR5s^aP`#eHth>E#KvO-k7e8My8)N&O|zd zFolV7d900|lZZ1|%uFEMSS(dndOUiushgk8WrJW-6ZFuCYvNIBTKU4dir}L#G&a0h zlY$HAkk+~ea(k@zdVh7H1V(UVCxV8B-C1(*Cgao)>Y4+z`{NAIB|EtKfjx*qj`Rdr z1~HlKb(^{;t9EZk4n}0t)uDroh0y(x4Tf-{>ne!(SK`YvHih$t=xsyj}IYg_aT=H z=Ey#1fHcG-_MuM9Vz`*Wb*YME_arh=O;bWj?I35JAx@p(SAMj+*PUrQY#H3v#H_jj zu!8s-$@R;c9z$x_5}4fPVnmeL&JRmImWsG!Ade{DikS_YKSP{IW(Y>HsOD*a$Svm$JhTID4>m3p;|6I{quGU+p zN2Ne1-#l3bS^4(qIar;HuKEvJ^RU~pK0NzgUAKufnoUV?zbuSMTwmNbiM;#gSt)4O z#o5?+iWz=3EJNu%Z%G+Ty)fyqf1fACV-<6uZSL~+=(ITNJ5iM=&h+!Ci%yJkXg8ci zS(O@D$0s~qlk`#dX8*gW(?F@iTks!+89=4g>737bb^U5qSg&09YF3A=-NB3@Sepws zW^+XQl$Ba`qJRjEzfq-YPpammKaCN#TNK4Rf0bu_TG+n4Dq$|?PzeALTD)8p6E-_A zThezoZYQI;ef~%jTokQT?&Mtc+pA|u_MK8h#%yQP56kYH5UT9@tz9kK2Wih4!QuQV z_!->+{@(|W*W`HE^HL@-?Ad^TkcW>S4`A=+fM;pj*@N2ka!gfbu#(5(Ob@qIG-E*d zvP{Ny)bIr{X?Jw%W7OxhMi|*fTMLha4!07(_x1dQ6YOeQVfilogpaICn+eomN}N}EeyUV@AxOV_E=R30~akd zPm(mffzrFyJ|ghpW@ZKS9cIz7xByJaY(D}w7SLiZ?_cia<^N}=0h;2DehE&`CvOTL z#MhnhwC47ksTJIf-EVB!?mX}e-Obh;?|;sjfY|Am&WCu&<`6GVL}@mFm@xI-GCtZT zKYJnzZ~-X>-g7xlW7U*2DgmCh458D$JwbvTMod~x+8sCt~xo_cauw%|oc=Ibzk zD($wxL|9PnH0lWKj&@?F`mc7Tn$G6}bapyH*_y+BWrZ1MKcr_w0~eFGr)*#JE&|bY z9aio1CRNfIVoQAG)vOqsg>WepiCnDCz@a?x4C`l(_bJ=gtTisk$=4 z7$zSm#gE+X`54%B(NtQzd!F=T{C~Mf#(%g-!buVO4;S$YhYoVTo}NpZGO`+ZAKvp! z16WNgD0tQR>)Nul!+m58?#7;cFBT+w`3JKZop@4Z_WghL$ro3|+IRb9jFkG>J&5c zGI;`C1MY*!OtxGwDv{+Ln=TWPxW>-urqd)5*Ol056;UcODRDCAiH^qN@5HbCb0)-X>-Cg&LuU#XTf$OCSgr;t;S(Nx^wRZJ!Ep+7q4YMSAn7 zddYmsT$b>Z9Ls~TaP|fY7mUVeSrl`UMo9m-^E0+lXQbOS6t|(9(NJB`E)|6lX&@jX z5_mLdNnH_)C;prwn)24ErHV0zY1Wq?WlKg$91w=wzt>*kk4q>?Wn8+bWa5&l?yY&G zIcV)W)a;VCB$^!9`VyK|It<;Al*UqW21aX_G-)?aS4`L=mxBCO1X^B`(z`NH_hos$ zN=N)-Ez@5T#;!sfRs8ymtMaPB*0BDvto$~(`7R^ySmNFUvww6sAiDnC1T!q|P%zY; zO@P*|P}k{HP;X7R>RZLT9=km0_PwVtnQKY4Y3VFsXQ9R~F~pqtEcAFmG!X+lbSv#wQ{3#ChYh5 z;N9x406OfmzT+27pU;|`XSkF9Gk4sy*XBg6z z0sZMPd_O2rB!=^%LzXM?apf0PrWHi3 z>c2@;3?gaSr{z-WB-NaBLx{6VG@em)yha}Y{Vuhl^ zoMP}3(?zqQ*s}I2`sG5mt;DMd}=dqqf6&?|W%D=t}Cf*b{Y9j3k;$<<#f6Xx0Sw~E%TyS)H=#luzTQUq;Fer8Q==7%dF{>O&_l>`!{UN zhqtfIK2m!(87|VNw13}&92e^qh>P+07J)^Dr0G);c}z)X!q250o;xp-F27uDtq#uH z;>YzxGb9Q|KRvjx7o_scJ#4t3^g3T#ilLrp8S-1AK0KG#Aie^fG|8_p zm4!-V66}p6uM*$eD-*OFW18QlcfEPbDS*VN&eadaF`hjGX4dyPePw~!d6hS&s8EMrZ0hOd=Hz4SJ3nP^yPTQGnn!lbYV_MdaneC#;>(XC{9aQ%;gTfsmEGUV$U(# zb0_!;Xd}mZv}qg$4NpDkUvatm=oq;-#~XlVsZT-YCLl(SVA=X|{#dEY&36dr-YE*f z<#!XkeA?^xLl+8zfi-jNIn9rpu%jJu5-rb_ve~84o!?V7$P6)P@G8JQ9Qi&t#h=dT_veo*hoamJVHn>c$$@NW&UE(z^tx16)l zq7zk$g2A$ndah1{+iUG`c!0QuA7jUwol^Ve=H@bIwE}QO45Aa)XmjLtMlOz4aKJ%i zoEd&TYmFLffWoLEMtz(OW^6U1FrJtf)fy-*Aq0oo#r~1Z!M~n_hMO^T87Y-qvGa@Dru)5>yZ> zMQaDMW0$sm)%7V@>|}k*Iyau~cF=T$e~`P7gC7hXVvk&`2k2;h91GAIf#<0=v&LB{ z@=p;l8aoM=?n|(Q%>~9rdj0`6JM#2<4yC8~^Bl*HaX(Z+4k@Blre_djho0i6L96~E zMdfod`?Yl4(UszqfvH>~J~;HEO=)XuJV1L-N;R*d7_>myB{(N?3iUnrUzf5AuU2u> z6>lEygb9Z2M<(`dg2TKEte$rn9P}sBNW7Lp(pTMu>cmK;Hya2n@3oITBw&-ilV=Gp zqG&zg<*y#13y&Nsxv7u8Y`FLwN;+=;A>w@nQw|p)}vKWE-t(pu0 zAO=Q1^;w9oX``OlojU>xH427nA_k#lO|$QVcCGDSeT~$@4*d;;O^y#Q5qrVtnen#W z_>!L8Nqf4vf)_n7NXv&%DZ8>aq`AfMVEtA49p39~bi?;&=ge)E4`divbz+HTNiG*n zh!-QCC-TV5Bven0$={{k8fJaTJC9I9|Wq)YfL8SM@pF9RSvt#c-?_9`Z(9>CPg&$2Q`ls;qv*wlmb((mKwTPXi z++I(7%cbd6vEh=DR+ZG_g3tB6U6hzNMMo4=OHtvmPf=Wdqxz`b3&&8r&*PLg6(Q#+ zn~bPz_hA+67wSjEf>g}rziWv(@&eao1jg;dP6R-)qO@?$%G`~Q(Q9l1Ne3$oUPjGKSRQnT^4{S@Ssz@}PPdFsy@y>%w6P`Qc z#|F>(-*(`L`Q%Sy=A_JS1r5)c`!Ub(q6Z|4cYzi*_(@+=kZ6L8HW)T`(9Lp6WpYQu3X4XBI@s#*D&T{hW z@<)OES1Pnl&3*1)DeL8Jr!=y#jV&FQDcZ=Hbo>2jz-!l(85-l(-xml4PQtnCyxcOB zp9l!OoMV$I#!rk0u8|zXs1tfJz!OIv|CHhbt-_G1WF;x@DP{2LIHXef2g5l&!7 zE{G?&S@fmxY~{9{YVE6&?IS)Ai|UE&}g7@zKmUI0g#`s)ZIsZnGdezlo961|9U z-~#5w4L?%k=?YJvN*S>;-L#u|yXurEv*Z^(POYr@!5Wk}K@Q6% zWqJmi&*7}ggGPVd(BmR;0I429MNC%5GSF z`xwvzaDPu)u?SAjp+$Z${O(Njgk<1+@b*`6--^UNr~@U}#=cHBc$ z!z?NYR7$*0EJ_$@*f$A|@k(uwd3>{HI4<;zk>y!rCmZZBGdyI-`(Cc!Lx|cZh}e|x zFj#0EI;&32tGyoTgF+fZ;e}V_JtIp0pZ!TjCQRbrY6ZA0pRv`1b1y%dY`T16dd2eDNKylPaA^bWrf*r4woR zVSfa#dVOY@XPozH`|PsDO5VP&VUE5$!ABzw*1W|2;5<{R*1V_yBC(}1++`ad&F&l* z<7**I-DZ@lt5hI~XH!XJlP{aq<>XGN(&bau)PRRm$?C%Hkd{j^h?nv!uKx=~-~U2U zP;cnnRyAainppT;gEw-yLlH!WOjB=wghmw589My{3rBVq%k_6CE^8Tc`0Q*TFI;E& zr@INVvLK;NqWg810Av<6R9#`#Y=0Q|UhZhH&TzA_H`d;C))cVaMcF4gSoP=$y;@K#D5H|moe+gF)M@(o`a$=zZ zPEyq-4VwSJ!NhYVMdE{sZ@UzP_N;YpfI;tT-PA^O=;}I4gYN97 zb1!CrtzGz~lF9XOe{_07w7Af`ut4=$lH3ORiYA@CALSi&BisPh=<>bg?j!p4B2a!p z`EQKOJ|?Vw7N}KIVNG2$v>+Cng(#cS6!T_C)ks&J^Mh3H?{f9FRu0wF_L-^^o)eV> zQl}bqx=nEhBzDLsYKuEeB=-mYXD+0xWqAvX zT4T)Li)vroP=O30UT~@=guLf5y|o+LI`=X(geZxk^rDHAXgY-xAue{E)eik5)GeTp zr|(yl*^eJ?>w^JWsMRPSF)L^_)w+Sn{$y1k00%*R7WFP8@!K+IN$wI$7-0ldxI!7; zT+?qaBBl&`y=lMax@b>$>w+|jTu*mH72EB@{v4@11d1O<+`XCa&TL(QbiV6RKk*Xm z+3ZE6>qRlvlZE(~&pt*i$;i1&sO!bN@b}EXV#0XceYtJqC^pcmiEV7 zIvc{&rN1L+CkbRtavEB43+d?7jiu49e5}4qmG$}Ukb2=;Zgq00nw@I-<|TiEi?*5RgANJ zFC>7XR`~O-wQiP-Tcs3kJ}znC2hJ&qe1ZQ@Pg(l)-=1<8@Ru&yi|FeW4joee9q#iy zJ;#zB==$J})m6vd$|IQNr3SodJ60WZgGFjGw|_zoIl@fMaT<@rH8v=aN8U%Jx(5`v zmH~nwGzZZ6r@dbMzAQ=<@jKyQbn}mZ8cFWIUU8jl;?JQQu~M-T)D9!jNdA*vMzzRM zD7$Z@5TyfP62ih#d*AF1tE*0aUF7zw>yQ~_c(vY+$yo|d9;@wfbnhSa{{q)vaGSw6 zY55EkNpI#WL;Cqpa~!C)$6>O!ZC3CCWE2)Fo}EWfKwX>=cn0E^^m8Xv6n1o6Isz=a<^{SC zH1d_A6=(!~k8OsN6s45Z(BSVcPL%C!1Ah4}$N_x9UDwY||IlX+mQ9XPXgMj8(}Yd6 zxVx(S6Kb%ap#ZdvXGD}1FC#t{2M_)5DH~R-TEQ^Wh*!5#&4Mfo`4MUI62*Cg71JLy zL-tK}W-hgKt^RFPI@AkvBvSu6P#hf-M292keo^2od%p*Hg3?5PXTb(~LC_x&(5d#< zWazN7e0_L0to58-DqW1Y2KDB8T97;I$!}+);Zf@Ynu$oue)jE}2rO8!Y)_brzNTGv4PwW*BmUyV3sC|Y6EH-?vWFfr;fd2wjhir1 zOfXFuUJUrPSB&v$msya56}9!K!v1i*2}?`ac|^>yXz#B&F@lXfna#ceuYV51A40Z? z7)^hTI*NfGJ2rEES4n=1vUE9f+~gP8HrCW|*v^IomnfWcDl|wvb+{KV^M+i9yEmfg zCzmwe1i0iJgLAuWww;eATg$4ckE-v#JgrRb5s+20^4$%*6Ss$!-{&})>CqEE><#;R z-5J`JbaZ^zk z^{HY{C67>U&zaz@md^c1E~%ME#qstUhB)mH-;oEeD}&seVj%9Cus%Jas+PyIia7?` zg-+l+YfJM8gB0x0T?VXc)|b_p^DQB;X|!PuQ|W9rCJCi3T(ux~ImZ;0X$T__|Cdaq z4*n%mRlS~$ZEDCTYVBG4bSxI$4(#$i4uD&*Hf&+W01?eT1kj1}Gh&Q;;;-t11aJ{G zj+Yj_zww3qrUaECL+#e=#fbF)Ltx#RU;!MP(5&m_e~yWc*>-^@oQw@)NP0+%$@-E0>=xhC8~vEIYfBBGwZpfe#F z_n@Oep$ynbl_lcW`dKsbc%T>Y3Gupju&r(1wcOjd+7AH}7RU)XSLP50=^E##^*&S2 ziOU`n7h`XFGF4Us`)xFO>MCNpx*A6!tbI|F7ut@!Q%8|sGO4pQpo9ll#HZfUN{qAf&$$n{aU-Rq~%NBr`ot+6jLqhW_wqo=N=c;~Ln@~k+8)85-m`BdD4^Sq zeyN0tt5Rg^vwL=4fwPt>YtIx+ETu;Zh(Y#5 zSPl>VYqE6p`P*c0S%(UsK(CI>%z#?h10`NCz#9&lg_sEKJLd-Yn@mm9@-9DeaaTUH}Ca2?YzScs9-wjh*mA&|M z>fZ_KRFy{_q&zQ9Kc>wEoaB^iKo`U}IR(qu4g1t9y!_BIF{WD3vO*|YDAF*SDxUX* zWpVOAJZq44l;&qJJ8ZR4drn!%4lI#Z+9u+Hr^#`PuluB+beB=Hn`^sr zs=BR9GW|nfzGX24>m(y+^{P-hmfFeFlCp80>Y639y)TO4fT@l_+;)45cVuFeu7xRu zv&xgOo}S3${D+oQBYntBcvoV4Z*INo8ME6A9y_$}T zeA~1{TGgHGEB=`z4OJ}M|6;zK0)G+6jy0N`W4|j60AJ1~J3YRj(Ov|e+~EgLqD%B6 zGuRxvxfnpTm?a8ZYP}rFoRhYL z(vNMDA_1KmtoAGJ&teX{#itaj@3k99=fSYdheq8ww9BG4EC0lxG5t$FRdU+pk<<}3 ze#D}i-?-pG>#B$2?+KD~TwUeoWIh)QH-*j#%&q`2(Y9M;N+~pCO7*{40Ef@b5BPZ* zr&fjt-nyW;- zYElI=&VoBU)HaT?>JL|}TY{YRRW;9eFl%U+8`ml7gQ!myhtox~YZvzOK`8kZRAzML zpL76Kyho0rSsr<|UVI#)R@Uv#t+WW@kI3C-x268%Nu-pY?u|Gt+2yu{rlkT&nvw+Z zJ}_56!*sr5%b#3ZKD@{m!r6QIjKeAQ?Tkp482?kZ?7nj|HZsDl|7QPw_P)F!g`40W zCXc$yxXRf`sc5@j&`g18&s7%R_kGHyfhhVw>pG%KhN!)W8A(`LZA6s)@AH1DSqA2^ z+!#~p69rgE0G{5I&W`-T>$_Ha^7^&8ub=J=ZZYPQKuh;X8#Q6qok|zh7OC}J;Y%Wb z$xHG!voB^NU>4<-VOFZ7N(Jb{@Hi*SL-+2gLfPq~trkxOQw>?#ayG0Z&){S1=f9PE%O#NbZUzlc zN;G{Prcih0Lp!SMrSHwc_$D`%mGqhN+bJy1YRxuVVWBskXM&3a5hO!mozh*Uz+e3Z zo{{aIKHLmyGTpgG_U;&p5uI0)BlAz=Y2N4@>RL% zjI>51_}*4~t(>dr^RTa%G-Qes42{UOp%!h<;PO`&?;rVOlxX(#90ugcs9!Ia*_Q9wiHT;IJpBIw=$ARUS`_#8rnKy#pf0-=uV`oeurBL>+_d zh7YV1N2maq{E}m=vlpozO8=O|o6e!=g9edDD{xe2c|O;)M9E(*<3R*a}f4Y5Ab7_tfD ztLAhZH}T;LWEb}IqtAN+pX40)gLF^EZmtxa{LDgDe3F>jG9CH=^V6{!^|`1mf(h5H ziOlr}OiOKk1Z5BrEtPtH+7EOC3!N>NGb84KV$FilU&B zlEAz?0S^F=9TY4$Nb6a8QZe5?`=-PVJ zPmccMIi`b2OeD_%9UqAt*^m{ODM0Q+&>@m)wrObmq=o7k@khKRVPY z54kaZKCXIBck98Egz4Q&JQq`>UAUT$G~Eq&UQVeS%lKji`1>ukYqSNxLuJV?^UOcuR(iI@=8FH4V~ zB5*svL-DV4T7_evT+Ial6g=%aEU#vy1r^F1&I69Oq;TtRtx9b&qKkbqB{ckj41A`* zcWa!QcEmf^tn3cqvhHInsQWjb_-m`g|7hvhrN6W^&7hXo28xh?9_Km>oj%s}ic_&p z0fSk}3Ah?Yu8}fK#$ZsqX<|a|-W%L0$GUQSR^RiU5;+iApcLy#7uUIE2gOPqErqpx z&L!6JTC7`$Fv5P3Y$8j6K~q_rmikT2GWjy4jehP0J!B<)Iv!@me*GXSHH>GCx?_hw zZL^#z%hBm?%s1aAbuh&#bgNV*;DTY<8M_ml1OS5LMqKa7&BH6qFk z%br#mtRr4EeQzvR5l3GPp8Y*zbDUm%Ecnb5beD#<6T3xlP=@`#B=)xA{4w?F#KQ_{ z&+`xkXSw~>-Eh9Ns4CwdEGd&HN%%4OyD(CE1N}oe*!BBA*z1 z%m*X!cw0zoTwpg|X;5V*?2~Rx`&m>i0Ok|PeKV$pUn$I8WAX~(FJ6sn)gNvcVGCMwmZ`RaoWfG3eqY$cTL^mnI>OY>XZ++qG8gBaP6 zuJWLtR<%Nh_RWgjpYMj@_RG- zS=HV}y0|y(^%|aIs9IP4d5yrQjxDd4&hiAgm=M*`DwkDqgf%Vv^FMJ8dn2V^GYXru zFbj1-xGwFd-*|#IBpW~)jM1dZ`yu297K0igkw(#kaP8Ma?o&TbM<$#Wj~3p4V);m4 z6&n8*bpSu4UI_SSDRxj+%t@HOC7m1pu47YX8zPsDF`|h6aIIo`am>czAv&nU9 zs`^B)*x2*i&XDe6H*d9Q?ZMU%WCS|e!Z>!tM18UUtOl8v~puI&#G0X0aW4dMYJ@0D||1a9){ugZ$ z)(PR??+71mW4F-m`d9rP+11B^TQfy1r3>uI1%n0-)JQyQP|80UcFayD96@=3cZu#{ z2WMNe2K2$%e8}B6(&ey~&`<8Yr)%_N@{ddTY{;TS%8*ORk;Ulnl2$S}Ec`SjS&SBY zHZT3@zg(i8`Gm)9P)3A=YSq%$$lgbLEl(#13O1=%L_^-oyNghN8|}CYCzC$7VE~p0 za@`yKgTM_oBW;9<_j`|}x0dgDuaVCh?>!Qa%OuCnIEAm`2&Ha}25v8fs;l;-2?t` zd$kT8K0le)l*~5`DA{5e1TxjLsEuwZabU7x?Wi`AJt%hqezKWv=aFhHY*>6~$MUxrMo%*k4BwNc~bz=s4j=yv$gr*AW@Cv%TY;qRN$^mX9SO{h?Q z%JvE(XqY{u+BomiAM>XS_`bbhslB%-ZPOn!;F|RKd*q8*-#o#5gGIA?3*0{MLS+=R zW!0*I$!QAd&^ZWO4!v`iE&0b!Ff@+a*R5vXp9L}T_v&wfiw%+^BP02J$_zQbxN><(lV$UxHnc}`x~f0+>we6!AlbnE z+cV!tqBQXz`j-6W-d0=XRQyq7bdrbGhSDjoCY{?o1Ff)4Qda*Z9_D6^ho8t!)rjs) ztC7xErVxi@nI@$JjeE@hVPXY9gbvpqJBJ2Ra-zf>z^_gihqZ||t*PR)rqpp1T0>zDjzI5Y-=zE;K*Mv_f0hP!@|G~( z^s%mj8t;us*?Rv_`JK4wD&YkrLOWq*wO5ibJV`?wp+SR)v9-B-p%eMA5+=E~uve#y z!p>L)t!j+MVe&aZt-A>1byriSTL|^rnb)<&mNxZ2?_g^w zoSQ8F+Qd@tjM9VV?8{e49^u`#55Zgsjuiqt%|*hF^ze5tb}=${4AdpRO1X!SV&&Zr zrZ;gKYn+gm0=AfcJe9`o{WxRI%-p$rM5zh;CuH2b!WG%Bh3VbIJEF|^BPPyzABXL? zY7P}iNe|(wdsL!%3az)FQF3`*+m=m)v;&h7pB!Q z5LlMQSZMPH)N*2!Rk_;Xpi*4MBBytvE%o2{NKUtqC7+4*Z~%rKUE4_iZW;;SKCM}p zqxcmbhiyng$ca5(6tELIqShM zO<`7g1+&Lu?qS(qRA+m%NbK;r(L?ro9VPXP@Hdjf{MmD7QQEC_>2FTh;f=1(t?2>A zI0ReB${7{jE5g?~AUoV`hsDu)hG#f8D8k8zc|kg}m>u-wey%flLDE z{1h&e2<%1eKOTDR%`*k&c^W}X^GD0ASO?-j9$^TR?)yM;fxMmA+kBjC$k+&ayVlrOjCHYcBJ#-Ko3&CoV!j$uJE^E|+mDZJO z!k$aUN`D=@;zavCDdJfW)oK|sCmaboq(pk@$p?Fdv#TiD)Pv;x9nOqGIQ)mS@dRtX zu?y9I@wQL^;C19U$JR7*`Sn_`GVOg|P#$XNG}F<({_+wzF~_%Grv78*p%jQi9JHb)_G!UrQWmvF8877UWpB%K{~ zN+y{Qo@8z;RJR}HWcQK8KGw-;Ni9Zg?IG23bLX|~nhe|Ws-(Jk3W@Rt0_1 zDiTd2WIw!6{6wfIV>G1oME2sz`48?M<*l{Pxx3Zs<6CH_jXI|rmz#I1{1SkT{(`8+ z!EW$HI1B6beR?q%TZcxnvbZkZ`y3V%>=fpfjW^;3kk%%zD` z{yc#W1+Dw~y+QrTjr z=2WuBB@fDcayEFB2k{4FnRZJ+OaP>(kAowJ)N8ly((D>MbIaQupS-QdT*c9*Z7P0u zDY+ZO?D~o~FP`IH#~8Ns`8nzCP5E#l$1Ct2&Lh}$DiRmGaeZnyW@+V1#ean;ufK!E z93b1L-@6o7wbb4qe|<0@kMggGkSbrLfLjdyq@wnDLO?AD z^p*FqqIYXoxC48p6tOi|D}v)9IlP)FEW=m0jz~vZ1=Mh#=io6B&^KW4y(7!iav14i zen~dw43SX|^*=?K+^11TG8q7CAvUxu{D%)I4kfdat@NW1QyG!oANnguPO>y>Y zkg0PO(Jn5FP035~7E%$Kq9+~(Nk5%IKTIU1hG;ucSUt?=g93E#Hx$C0Nq;}28XsJ>x@qgGB< z{_M&%r}^;R34)bcmS7K0?NVlQgvRd-hH$C<{1d}a>IhJv`nh1*=c}b1XW?nnK`S4M zd1k0Li8pjfGSW^B{+9Ylv!tRSv@*J!g9 zezQ0o*DelD1SwV1WwEC143}09;INfg7+KfZ2`-~6@UZpztW)i^U9S0$)Zm{5P}I9# zr`s8&(nM^NEm6%+xU>3-43Ab2sSxUNI%*k$-4&Qc4g-+w>WkP5R~oqiW?e(oVUZd$ z(5p#sOslA#sLkTz#`1GrtVgIjphV9ICj9YtXJ%qLkdeOv~teu zO$W^8gFveQbKI(>Zs$FkZSaBaR=2c^S$vZDb)|==+}4%!c0wuC{xhIcp0`>ZSr>5& zk*Hv>J;uF{_M>IiThKAic*g8mk{0UV5mNE%5UkNma!dx%c(6Mu+Gz)yVAh;knQuWc zFmrZlS8*@wn>vg-J3eAuhEFmU45zNicY|J!zjf6Nyg0LB(_Q_(3nXny{YqL=wax%h zBl_VpRR8``83;Iy{1X$_lP;~mEVgt_FI`^0dLYKRv;U~`IFa@``z4lMlr;shLi=Pkt>$GrkuHn$QPmfVP5W(+fSl zF{5F;9Et9<2{Ci_avN__E*J`ML zt)MAyGd~@wQ;7Pl&)0>jTj^2?xgap?ri5%+r>zvl1mb2>oF&%w#uWnL}RhHNx7?e2lQCI*kC>>FIghKvMnd~T3=b}nuvW& zliTY#|dC8&t{Ydj}JQiFFR=5Yf!Q^Z%@~Sz3tmumd;C@ZS|*U*pS|vZY+y8NM&Yk>g)E8qI8W_ z2=}E!EG*_m<1%P*yT zlg&JMxa8k3UB(^4k977JjBprTJFd&`g_}HCz77Y_ zZ75QJ!&$LR?p`}D2df&~a%$A6AeY2p4ZCF^lpeK5l~y4D+Ah@sD+-!}$rg(wUhBP` zU-4iRcfA*QFzs-nta8#{N)O6B zKC=phJQ|tTAcvPfEcF^Zu5yYquV4g(>`DLa8gK7m0G=3P)Ai5vd_Ii~NEXD_oZW~A zxEiiKhT{Daa6TM&V_ImVX($Rz5&K7mG7?u_J5sq-LXO>%z~y45SsX@h>$q~kKrhuH zdC4AY1hM>=yRUfY_hR3i?C(2h?5eE+h|=1$H>16HY+LlMs@uKeUny}g=1-vFmT@$c z@v(1j%JKg~1q=U!3RoBHL==J9T0rkw&3DjC@i~td`!mcI9;!Ggyb2@ud>OydDg~1qUM1&-J5u3xy>w9vA zSA^_#gmra{D)akJiy5^XDy+;mEKxaUPkW!wcm?%ID3;yhvT{$g3yTcBh4zdP1kao; zSt@GInfLRXwzAILOn?N3k_1)4Q#@D%(+0hSaJ>1PX$9WRgUgIV7ki#V%jIe!Ra}Ki z*w(B&32Oy}(yG{=fWGYbJR#|R3#TvD8XK=A`^8FBYWX$gRJijH(ISN3pGgM`XDeNUtArPX z9uU|MY1)uC&CUJnp7mV+N=##lbPhGg#NbyN-0{I7whx|=4iI_|q5W_$Vz>(H$%jN2ly{rTs0|adhd!}@*aZ03?MP(f_ihhy>u^T3`l~{aon%h_Z%lGuewSEXjD?$OVPM`;`gYd9 zr0a9twdZ30g7lZ-Ij88cO9`LKf$6?mf48dmc>pUw)V~VIj?!}vT1~wf`ZKsrPzBoM z#OsjHdoZ|z}fNU<#&hQ-HG#^yK*#{+jLUC{Ow%R#P61f zje}*&&{H_rtR>95O#E9F;+xnh&LjEbx^&Sm&F@`KTn~A^=lvy_-B;Xay(j6(vTF4G z`-ri8kKt$1IcJP|6tGf8-b26=()I;YTfu8``ACNi97Yk1WaA7<<51N)-9n;Em=B-&^p2!ItTzs45`IGOKI9A) zRGgRa1-l`V9;aj^0PFy9>CiC>dSZvPKKPHu{1o?>6<;}(6xIK?%G5&A1FFd94ELJ! zzZ{KkJOBJqxp|X+3DYd`*+({fRzS7`5YaVyI-iZ-&KuG2D*OK452d<42?M(1Gr2U! zH4t$jT#hkw~(&}QL8YEm%(M>eU>5Ui0}ZVk&I{p5*rQ0Kt_Z_ zc&MRMsRYxGKk?{ruO&(E+7X9K?BgEoaHQ5r#95lI)>Kw$+tA$)O|w2aM3Pt|;}}f; zHBFduteN{?j50}t%k=yy>(MHW4e$0<2I*dlkE&w}*(lm>e*1hJr!@qy>$mun<^rJb zKw5K65{=6Sm10^S4+)xh)L#yjN&0s>?R}Y;e4U}C#7enyN|a|#2<=^9e5B_h(+yI*3Q#D>baih zcwb(oLQ&FqD+4t4_-I@H^_Uy{!o6{euN4MDnqU>`6-OyES`Z?Ly10MAT zP0147IL>quTi}zv?c~{oJmj!KHQw%(-ODeF7l=QKVXii`sVc49NHT9p@Y=Jv*U-p* zuzT_2dSlqKV_xFCDZyL2^PDZnf3)m6ID`Y$=?=j-n6^z;Pp8(fQU}?*ZEgV z-kl1Fx)@Q4@3j6N;ej-s2P6sfCO?O}`4@8`Li=qVco6XX2V~G}H%A0FPuy$D_ZVXJ z+!sR;L~H-pBT&sT7ArcHBfyU|VuQae!SaDwFxcekMbrus+p3Db>$EIs;Z{XeBp4)( z*Moh)7|!gEpz0tWSZ={vHQix(yPPF0jKpW^>%4}ax~E3GLWY34d#ixMYgA-cfVV>K z2na)i+WB0Q70>*25{Xnry_hA{6~k&%090E4fwwQ*@!W(B{({>nQ^cV4!LU`_x5djl zBRf3cyHh!N+dXo+cA*P;I@Gy@4*3=@>cw2JPhY9i*C$50%!{LOZY2IqO1S7@av*Xy zd+qh7`Qj(b_5L{R#V6udVDq2}MP9AA3yRt0!VFG%Ouo)I+iM)-3p&VsN|8e7m5+n~ zMx6+jI(Zn`Rj0}*OuHMw01`LYD>ptxSyNA0t~iSH%x>&A-lfaFV|5#$)WAR62WP;Z z`pF4k1wuzgoNP~6)TD18X0EAbI6kWbUO_dJR#V)hH>qzy0OO2TwE>saLh|B4$UxjOR7fHH%G3E zXM(L#oy&&#TY4%2`y-?>NU0t!G0M~=TRQpqor^SS$Drl67j`>mvp!Je<2nbAig%&} zneJrkmIdzYMH!^(o$*(le6tM!c6R^2F!aBXj4R@CQ9Xad&~NX2CzI<^PlgglIfN3w zN_G1HmAV>a%JGqr#PJ@)u=M2rSR!K`t+JwhtVermN_fI5tO(CuogV~i2pi)d}^Px__=L&>()f1^4O=XA5 z#32Qy?1c@*A-z0V$@4;}HDRn%p3KIuIiE07ro8K+@n_TXTA?ded8hmil0<{#`Q+*@ zC&?C$S$dDF^{6%Ejl5zwr;60BC}5-Gf`PHZ+CAeZ8mT7b;8QYD&>4<@>H6}&d5vNv zAlDU<*R(2n!D5csoe|n%HP<5HVUcXL3?+M6H&l^I0GLY$S!PN zk}XNIH>vo!{}eHZ%(8Gdh@O(SF6qe)e1ZEon*BKO79}R=q+gan!{}XmD#L@BKf8Em zTl#l?PVheOLYA^IV|!FadEdt64tI}5q zHP)q%N57v=%detDJa%h*Qs(UKR#Y#lLY_a{6CdNG^&={&4NZ)XD_7uC-v6vLIiwCx zSDAcx_K$v>4-;kdp z%{bBw<3kbJErur5MZE)>-I)*Y|6>+$qmJu4Z}C-gnT4UFq&xI=6yo>W^7n~Bm;z;o zI*{Yu*zPg$?;zv|dl$Jr$eV& z0$qk|5sg=I>xyn)CVVGZ!td{_1b&wK)d?*R(>=NY?EIoB(23PVMRk=LjA1~x zI)tR8jF~r`f97xW)-6{)l@Tt9eBb;e92WG$4_ZSa)OocX&$%2|TALqkwvuX(55=xn zJ_@n_eJ{*V+0p9yY)cDXy1rFZU!}Q1`~HnatGv=@uo;&2#M-+Ak+qB`Z3>OMyD*!T z;JA4EOLsckGJ=PP&*WN`*pe#Fu?eod(V|9ust=JZQ<~~jV9`GdR)nY>RK%1vZhfPk zP%VaJ~2s4K)oa3cA^l;^ka zxhg%9D-gR{F(cn%U3e71>Skh%d$evnEva-&-hfPQRuLeqEAmIUj@ewvYPLOoan>k3 z%AK{~_&x9!ydh|w&aHwWkS^q0wtkv~t2G3n>hnq{Zg2k!ecS(QF2;2|c5X&^e?Sag z87PkxmY;YN8F~JR&Gag&ctqy8w90kQoByaSfWDjOn|c2r8*2n!=otZIy!yr4;BN`R zzEAyQBi7YZlTb{`j!E26pqk0>z?E+i5Vz?!QUGm?w_Z{+^sToK$RQ;xttXeH;Wl&T z4O~wAkbh8Sc;Sm{;h8!ZQa>nFMA(R3jNr?DMp!wO+Yfv7YGzN$VO5ktBb%^2t)cw7 za)Ig1t~a+^`B_3T$8ChBG~^<<4ue~jgS6FmbS>fd*JMGb$)t+T`rnD2+^B@NY8TIH z@TqBbX$Z^AKjaXwOjUh)y39S|&#jSkH^YB`A=>%7sNTS|HziL;e|`I;e!e8HBiuHd zvj?j+_%bA&lK$^hrFg|w#1coWtkwstRx2FrbnR$a_26r4_(AfzrzobNYfbdGa&lYs zuv^dLrAVbg+ymP%7fXEF@LhIQ_NxbFCl4Ezgil-+J8DzzsmSW{t5i@SM}R{A;NRr@*q@)rwpzy>cM0@oS7_;Unmf zyr!f7gR;4shsif8^mDt<)k#@@+oJa1g`>a@P+_tA@xHfl z%%*B?-{lOqmcyYw;*)l1Y|uv0tDN#NsqTU1c1SZcOX~UnkPtpyu5X%e6O8#tta%C) zMZI#W)hhX8%$5jnuzQq7{Z@MR_Wr*`oOUinP#9OU#E6DJtDzHG8XwuW`2$_iD*6ew zP9C}}fm~_fzH&6*&9m>?RNo`0b?sZ^?Z5}$AO=RGotz*NB$x0y&?SE8a>78*n&xc7BA=EndDbGLb!$uJVPXqF-FWN)-Xzn*1oB|3&uZ^SY+i}%gB3YP>d)e* zTOyAqUR7~6;OtB*8zWAqiu%z6y_n&;xnvi_2q;5cr_)ZNOlI$g-;X?z{%`bxW}Ic# zz2k3MFzYSD4Q4xUSq>$rf*q~U0Po*3Ve9@QZv8qv!OA3$kubkCi@kM2b`mdu3$Xy5wUg2 zwngp+fSES{pq+!iQ=29qu_q$u`R;k9QFWg&d>(9*zq+8u@8eEsj^^#3$00N}26iM` zCeB7Ib^hQHIta=Qe>AWJJ<{>P|Lh_xYPUvxahAq*oVgpWkb4S(thgcQDPD@FLsL>* z7#BMSbX2}>_uL|i)rKt}6%FbAIVb*A&j)ES_@J0iIp4V0+!w*~A=5`J;t>>mJ;*=h z++IAV7@bp@EC};9^$C2iTKivV{18zF<|58!D8-8clw|=Ajp%O#NclcN>lf+oc)BL5 zNFS$P%2}uMoanMwyE~X(ZJzy_Rlk{AZ_bA3FEO`kyuI}WjrHJiH4`yAkNA;NLp7T+ z1Di#gUB-_B&Bm*^VuTnKd`Vhz0&;d7cHAF-Tw1Q^<|)CAyds^h1Gyb!(lfHN=zN8f z{o=rh)xW(u`x(diT#JZY`gJDwN2xs-&mJGFHXL0*PJunmpPdp;wW+;Y>Q4rne~XWO zBY@k!p@}36)}Xk1vL`S7^Yrmk#j@!1i7d0%lyP&9%M1wDtte%7lYk|Dg+mqvelmx|%P6qYH$fyZHWrdE&rg-h+4S;ryGl;~;th z;M+ceqhcaF3g{Pa>$P~HoqyExU+FyVjr~egL!$M@_!RhvJ8NH??4(F&4fS2Ig<$8C zp0W(Jk^3%A=R^gpo|i7Nfnh`Y7U(M%C`s0-JSsj!rm5@h zs(uVjTl%6IL%ev(di)(X{9Snr$j%jWt+ThT=qaO-8l?hd1brf21?>L%yMD?)*2`+K zmxG>~v(<63`FW{G8Kn7N&fNINY72q}+s;RZ-O3fP+21pJ(GGu$e%!_CS+gu- zn@H`Uit`DM728KSZa3?XV^KJ_j#iTB{wLuHtDKZ?4VPo*;D0nad{j5JWBsyKfAam zZVnzxjP_H?5j4S zAA5B88uKm`u>{R`6cW#~@;4B-U#oZV^Vl@j#hVllznVbQgL1>rOxcWr1gU7JhA;tJ zBVBIAA4NrL>3EUf{Lx`C9Y{jJE)Ud?(y+^c0O`;Nc_SJrhAOA@#rfkR+?~SK*G0Y8 zP;N??DBPhbx^~-==ge`MzM}inv1yt?sIY$sv6#>w+qgWKkO-huRj)>DlyI z%4@;OFv~m3Wk(Axx?|6U8tLL8Mqlx<2}6lqMU|U@#NcR&GQEe=T(egts6DSV_QyE* zbv+ONPH#VLy7sy5TwMx2s7rh1g6An)cZY{QQI_W`Mc2TN&)2tpLh_z{d0^D7tc{14 zg)-e{2fsN+HVA0ITcMVhMWN$hIM31kHQ(d}L;2U0%ZqOAc$M{T&EIb&#u5Un`)tvj z{W?T;EoW9F-%kR8KrN>L^Mp2&uIT`fy2!XO9P#*iMd|%xYnrTm+y-s-8rAe?cg}^H z>j7T2wg2}2A{aItJ%84Vj1e3#NdjM@O^*fMv7yf!^}-wQ{o8S+%e`VDW<;5zBlnOU zK}TdHze9vZv2X8_|L1J}*XiH<|8@F5LJZ0*U*z)n$NXROY;I%3tei3jQjXt0#it#6 zXfmDBY)7;}FZtxGXntE)774qsywUHa&iM29A$`92DIaiv7V_tmajon@XJnVu?S1v1_BH#y0ZjN>>Wr${`LMb@09$}<^`c;+yg<|R|lW#xEOKG+4Z7g9?;q+v1{t}_|!tt z!?naPAv^pD-pH7mN>DZaWUwqXTAtQhDTTTPT~O#mV;i6htYaBXQd#th%47a7{u4mvZ2Egw)#L}$0s zYLN5Z+p!^c=t-fhd0tjfTY3Ub*T!Ac^KB$mKOLv|z1)!F3-vycfuJ~>I_y@Y3+fJz zp26#JO7;v?CQz*4s}+;bJ_;)kb; z5Z7Uxe(RF(No#@4F-Km3W}3A>wPPaO(4HnH&3vSRR|A9=np_|#QHnr9QBhHP zmllvvqzX#tp%)QD523w!@9+QdUGsI;nzH8Ev-dgs>}Rp`N#{NUcQ{ujXTeFp}ZULKO`(ftBd)-wIMtD>G*Dts2!j=l#qz__PQ9I@WAn5175p! zmBo|X3tRB0*#B6gJnEDcGWa}+n=N}2xXvJ39=|c`EWa{!qp}u+DvwHsK9fB&8@~jZ zdj6KqW43V(8?oKOJ8g0Q|Akji{|m2P3O$7GCbFc*v1AVnAw1RQ5zIK~7uJY1(fP^) zz62dMeu2Ka1$HBx$?@WxNIdOz3f4v_Jfg{~yye5s_?q7`<-YYx30LvFkQBxOItv|H zWV|>Jf@bk>NU-aUzj!YqMQ|(;S+dg<6r#k4^z4L{V9YS|JJ7nUitD;A_^PokT*~1T8{OYc_TB_Ys^I`L z6d=RcEGPP?u0PC+)s?jxu)pA7CE*>(;f?d${Y&MN2$A3;lqGYRMlK0-_nmw^r7+Z{B9jn4#4n%?PmS(H z^l|jf(lk?!YrkN|sIc)*`ny$@h)9rG=7XE-gZ4zeUAB^?#_LE=fX<9Dk=17A@nDWD zReDg9@y#Cc)df7fVQp634m;bFf69Q!z>IvEx`N@Qs=-(Qf6$Q4%lWoeyn|Ng zb%P{;R+1zVv`*+ccKqzRzxi=PmNFrFQXud~gJq;6G};UpgksJ#k4W z*Wdpp>SbWRFL9-?KHx)W)#R^d2R|U_eAITi1p0b2KzZ4l$@DjCD@8_68pLi{@<3Qc z|70sm9=j1gzR0F$xvTg5UOBzcEa-Lbo)FIsS1>5COVhim4 z;Git57L?#G#iQqUD{}S*+sV`}p&QU#h)@L{MqK-u7KR8grN>I05uo*n^dI7u{Ko^s zH`E<&zqTFXx7Ho5c=_PDYgcf62k^Z4yjG^~XWK?6Bh3W|(C{Xhu1d&rx)oW5B;6d* z)~w0{VMha)klV(BnepP2XKLqJ22{g}EPVoZU$qQAt?vTnLp+#s_yYDGqnn~5YflyS zn*3d$=FD159#K|_cFbiMmMB*>t)uNBajWhW2RKzMvV?G5JL+CyYyd&F)^DM^;gO{(zVC&d^2nqFl3DQ-eeuo&4btQP=G$bD5BU7`YOb@i!!YA;=yAtK7_>xMJ)^pqQDcfIWF z@*}9D;@?z`TzR7w0}=XDrrXV95533yaaS(}(($=Nz=p?|GGvlja=e78j% zWTu?3kDUlAbVlb|KQcmhpCL1u(pWpMD2Zz(5+x9kss*~B_2b7MuFlWa}6!E$|QgHMqP(a1A)Nbm+ zFkE|p+>qJ7IF{v}{F#dfs<&@Vj)Gq|9SWUmJ@HEp4|?wDTDapX&L3XZz9AgGaV}*h zu7(T~Sq)D67cQB?JqO5Mk)PB@mB*UyQGp1nQQ?pjl+NJ{xa3>g_|$H7lzg3<`}e0u z(Sx-I-Ek(K<^8R#XM4$%Me7Z4Q&A7OP1$OLohvmoG*ruLqE$!P%rqJ?X8Av2QOo&{ zR-`|C$)B9a!XL@v8o2u0i`hk~WpF7Y)ch+eNzTwafSo(j*!QJL`QBeOAlGl5o$^EZ zw$j6hKcUl_2r%R9OC_c)XoBvB5V~`PI~;O>yU|}QYx!PrQU)bc&}LpA&qA1Kelg$1^N{Kc0;!k5s*Qx6z3%xNacPutOC7y z-ypi4c4bAWtbD*d)kXE=0ag^Dg7C!#Abb(=bVnt2FoeulpHN%T)Y?(qnz#}k2dN)U znU-?u_*nfH`}{>4*$w@43fBvP!WCdUOZmppw%m@@QpfF^xtYh^YdHmxA6ilz8ggHY zl4qb3-wh%YlQ)Bsnw;xJL_A_z8Y9&Ya^-7F!7HD07O}<59&3@i9&nqMVyrWhuEGsF zl8T_mZkIoc@EF?s2CZjYbIe9}xX-`+%6Otdr3o&I`e z;jDt*4{uF6ot5h&{iq3LuB2MYc<}zm3-1|ap+Vg@f&bXU#Kynj)=CX?-o8b(&>d$^ zQ@uuKc3=-0x?MWZ<Q_sijtf{?zlJQY)j+Q{s-L=qpbL78#Ax8VB1VBJ?-H zS{r&qrTZIDIqVvS&l?`#(S}B5K2&@tatsWQTrOdQ_T%$7x9>}dpdaq zhaHyG8iMv`R}}S09^F`H$kINvg{foP?@=cMqfqKz6CO8ANf8J z9rI4T4vgEK=hNDeEa&N^b~#=1<93~U$H}?*#`fe+sj%e2;Cjfr=_ zR)1@WlWmxxis>sB!uC{VFsx$;ll!8<$S*eZ(i2 zF6&ojQenHQN$K9PR3VRLn&)?o_bk6Uf(wN?Wl;dfz$};I0xm(Fe+d+E7o#h@zYwov3+NPE9=-` z?1<92S=IaH+H8FCoy^A?-Gqt&N~^?VrByc})Koye)CC40xF&Em)tui?!Hwnp%K3v? zqV0KDX}foe;1iZxPf#MkPSl%hvN5d{$pGlnuKZZrG9JfQ+&yS&bj*XxtL|wdb?cvl zxTlBSkr|c0E&k7nveo&oW+8&|oM}+=;ZSesLK@9dGTD6w)W0yZ2Z_J-p=%*rA5eeh z=Ni5=tWtT4I9=>2^l^x_FEDWDt(o`8O2_;RmFlooefi?|Lc{;4WZ&zc{_sZOY7@#*8;j8cx9>Sbz{;&#KI0^4`$PwSPW+!b19SZRpw!I3kq!C87+`j*e3u9G8H~ z)9iE2PRN{`hg>>gMl-){VIv!Nh|;p$wtU>rf2;ovBsFg56)4r)Ox8^_2K3j>hlfbq z$`{V+{OK$|><_k$I!`)2_Ky`_J~?g(=4$#RV&z~}8@kygg`(*z^k9Xppa51hCG^Wr zz?T)wMImZIcFx4W|7UswYF1>Q6WE|7M0H2Uaj2wTk7}Ne%1(7w5>rpT-G9|j zLK4+CWU}|NzhA9>e$@5hRqfH)D%a29RZN!enz0pQHUdNPlX#cl9hrE|h&V9YQ#!VN zBmp<*!-z@@Vs2vYg5D#bXA@>%E#vvi6@;Bnw}Fj>9fvX|p9NS;RzyoW&#gUZ9`aRi zC^n4eeJnU8-%aSXsB8leQwW@J)Q1X+P3M zt>W<+G~=aY02Oj>qk%7Zo>dm=maH9o8!m*IkaCn;C@O2G0UWs}ep$A*^|g6UP{Q-! zoIAF2pB)c>v9q1CMcMe6`b!iAboj?kRP6hl?36@=g}oRPtBy@98h`X`iRXy-YvDN5 zi@WAJFNlQ31!=X}cwbb%|6K6__V9pDN62U}+F(@sFnM=%@;@l8$ilKwsl9uY`-}#asV{**Hh^Hyu_J3X$eo6 zaW;RYnk$>3O{YfdH_37%@Fl^$?qsdre2JrXfBtD%Z9x7`%9HUBBxUJ~9-qosbP;*tAtDr2#=yzu`a+s-28=0-jvN4*m zY#~&|S2l4}uU~xt7B1+rZZzRphBPP_Lu8{u7|XQw11zk<5ibgxA?aJJ-ig8aSV9yrUL=- zQFyFqc=g~m1g)^ap)-s(c{3pQOzIgn9UwmAbtl#fZ;4kzw*|f{i#&Dwie?Z>pK&F% zu0KseC2Q7FVMP_25|m%HYCc82qUvVVnj*2Tfs^KD#XV)!D~}W=9OZ>qazg@!)L}1V z`K-z^7Kg&)ZFpWLpOJ7eV!ge5#fDT}*t|WKCBv)|lnMF{y7^e1g9dufacM9#zxzgP-Hwsx{E3Etc4Le#poQ0~`FZ0o#-pk`R z=93>Rq`wE6CS-Sy-5_)k`f=N7$Irb(Fs0fll7W->(A|)!`o-xB1-~YZ-p#PcrzKP96;Dx2A0{hXh9ZECW=~JhW&V(9A+L!(jNg^$L#mHWR4uaa!o=>|@XtPi` zflVU&H$=9jW(qVmvr{z}XgIHh)Sy9gP3Mg)%ytjc-22$UG2y54Ha3-_@GiOFjmJ-% z=X{{gdk2U3NCOy(`l=Um&h+-$G7R)% zpjkOpWGw0I*tgYomZi`pH|c0d$}p3(&D%*sJXbpMtRbrN(*C=WE@`!7z3pX~iI3FQ zc8gThj|z?`;zPMEi!!q{-f>sbhwB&WjjLr0QUj5xC*@LbCy?jVy5(pV%{Ls6wc=XqIZqfZd@yZXnd>#Tv<${Pgd7)%x z`*RV_+Z_F*cKF`FYVaFxm|4go>AWU4xAtW+gBjTTvgUp#Nb!N%F& zZ4DtRBe~=zMlEZ@8AC)7dV!uti*icR`)@Av6n9W8#_94k<)*j5(IBO?2f2DVHrj8^ z3fm#z{~raDJzRmnm2@#nWmWWMs_M~KEt@qE!`Xqv?+SIBkKZ?nTTSD(_C?@0r6s(1 zuQimm;A6vc`327{sI7`%@ZOhcZNxfl@Ld-iI4)t7+112B(B%M5&0iAQUVwh^7PNmbOi)al;XsZ_D2ZNl z-vr_gaCBU}B>Gi$hqALnw66$4f3Osn$Rf3%TYX-81%3Bm`)gUqtHI3f<(o^?_juDm zoxSC)Pbu9Vr%H3rM}8z2j{sB~ZWkhLbf1n&!6aDkQOdBrYX!O)k*_t7uL8%brEOZ^ zS_%{qHT81dRgRlwn#{;GysnI`)0b9&#qnl(bJBr|IN$uEXn=E7qDyTH1isx?9E_yH z1rza5{-c!h8np8wK|McN!C;cVeUQ$V%OkhF-zN4(S+iw_8l=i^%5`?5ZXgEtui*i(IcjqiU4GT(ltcVsbCw1(X>21F_k9$t1P6+Z zrzjcHgdV8)l_j797rI~Zx-GhHUfFr%>CS0+o}%5%uN2%STuBTG~4-N0nf^j%n!%zgTI2SNU(i#AE7VSR?e zeqtnjxD&KF6@Dbhf-F$Xc=$YzUVuY}Jcwk83o0~;&$4%qhVQe(!-5?<1y%|GdY8#I zM|oYayKZRuOnf})}R!A*?GS7Vc*#Pg54yc(XCsaa##}f9zO(72ihY*1l7| zGNBz{w7xgt@#)nEaoE!8MPSFqdyRrwj)>0ccDZVk#V?hj0*B;C&f%wX*IkyeT(5m9 znG%rYOTclhWT7V|@GMrko|3k>ql%JXL?RLz>G8;6X?y3dYLC#8@~9WHrwxX?Ko5BiNwWOOf8gV3Ly6J+A}VO9!a6+@6|r7q1f|Fa+7_{xDvZir zp3%;7>ys@i@IJg6aYW5xIFz}QO)`6Se`S>bpTf@3zN(>-ZTfJWYne3TI#~0%*Wq+N zrH#Ld7OST^mL-1h$NKWeZ-4ksTHLpva~z%$-|ygeJ~VgsD_45f92-FmNqgC=W;Fi8 z<$r^TTK@{Cgu=o5wQ}6HkGOrB`cq$9`KI~5Bq`=WldfDWOYCPI`#Nvp9Wvw-px)de zsjx@COWvQ~_D{>n63CcfM;t^T@quV|tE?Lt5Xcj_T76pTAHf?dF!h^;f*Pu3ftPf_ET%4mPcJoc=We|99nn z%2xRSJ~!j#=Tr`KMQ5l9vY-M;ZFMJBy0CM@_$;l{tyl7}=`A?Jw0V=O=-1n?X89i! zs1h{0(T-vg?NS};kEIIQ{w%zfKXJz7&dStQbEfgFbhwT68e z97Ql8RBljXa-@BXpN^U@$SC4!QRwe_4z^I&y89C$O$~8e~8>1%eHyrCtYy<759sh{m~n^P98Of=LPC@ zg}C`_IraNa1LUoV{FtR`zw#R>Qa%d?N^ju}AFD%o8!_i~)yHyaO#54iLO2=u@=Z)f z-$i5sHkbw$Y;AVJrp_2Mz^X`MtS;jqt){`*1M|q93X&&1p*1u*KJ16dd$5Jm+9o0} zp%8F5^H=6c2afWO(S?lo^jHm*w!$r{n5f;E$R(TLgIixlO{q)SH1kaH`w~meVWm=2 z8ZQS(s+(z|q0H4m<-UnOtt|}NR^mybCd719%@(e!^N$_d$U>zGn2_ld4NfZmZcCvRu?VB&+z%u{6h zx;OS=QL=f{<4W^At>u1#DcdV`v*+vw@|t3SC}%f51$h#nWU1`eR4#!JjmoOr~+$oyu`CvnBOD9p?7t4sE0M{o+}d#lbo$P63SHjH2ij(!=Xy}7^u>$~Z zH{iA~MvlaL>n}tnDRO_J(bm#;lv{M(4Yxyk64j&7Qg|+Vfu`r`)8;O09 zmi=#>o-NzIIK6KQ-HGJ(C{PPi-}+;`ijXx)nE#x;uFN;AkP$o`j^PU^17hw8biU?gxt*R20HKm~NPz`$`wwuY}jZU zX8PhMH03@><-|&LkZczcSx!cTL`Rf_UuaBm%4;b9!u<-6)xK%@aVf*A9u>dRBziGa zw6+~lYRGKFZ0*5(WMOX;2*_eJ$DI3CN4=zZ3VHw{lc15jEt@Ad*MOmCJUtTp(I3Gs zK_oJ-52YhwKRrS)3$>P7Gw_HdA)NXjXcx|#HgM)>y{nt@^WaZ&#NYMk2#JrBRdahZ z*JdP$>|QVPf4i3BzB5|=N)8T8l|`Nk-Y}Ms?|bzk>?HKsgVQ(fcWSKe1JgF0hi$3c`ZCAaWMB3jp=DxHcRbJ_sdH3>W_gMc@_i?KhT%?vu@ z4f~mEX%(|E9(@C0{-N*5u|i4gRE*oZvLbS zt}j0yh%`97^3EwrYw$71WxlK8ocD65R?V&G^Z+qBX-!Q-Z3aBKO(ukQ<+~zn*i;Ue z$-ajc@yPFIy>YEiA7s?zlABheryrf~Y|U2v`Z7v(P{^%1x?)~;X_ zODL%6IE0^Xva)?NPS^LTN=OjiGX4NhG-Kdp&vf@k9B-kTRkF$pJn|XzQ*sB7@iDHb zdOdzj_-I8>p}}BJ-}Xt^lury?J0{UL1}0J_O-|+7#=O{1Ss!#%O6DJ%4Qr@+w%SlX z7R-_pu7YHohxwdWCbP*pw6a9Zd=%MNlnXFy4Bm>Wr%4$pVT67FetillBzk1koEqF5 zA|3}c$d0B80wd%jmtIb1y!AJMK{g&&jhb!d($sg{@X=l+W$8sN&D*y8_VCl38l zRid>^o%Ij?X&_1UmO*XL?{VT;Tfeg_#=UOpvEi>qf$?W>hb&+%1)RyxmH~otk6}>G zP_aydkli;#7y^#4%)T8p=WfF@YctLGnZX*lw?#ZTyavKjyMt7tLGlVFcUa4X)Q=HZ z+AQV3TedR8Czsz8E7`Xpnw|B$r7+xB%p}{S&-f#&9_`y#1mj7Gj`R21kc`zIh4qLi zgk86>3WnLB#cZG}%!j(*hAK!%pDbV&VirVPoOyVaS)TrFhXp>eo<1P=V@e&FuQ%yH zj`w-$O{^c;xG=Wo`EzHsC3r?xLs#kzPPL7Q{88-~avh~%=DketkK#uKQZ6&4O)Zud zO=*W*cqUnU5uIP`w*$yz^#z2E@B29W@Xjhs4qe4sM` zE!Ml+7wPR#jz`u4JlD1!Ytnz2m&SY^j<05LyEH@Ci~HxJ zSKspEsNedefyDL4g<+#x!rbxW{A?1xSzpe)`OG*IHo)H)uwL_9zUG6! zn`SJVQj=7S5)bFeheT&{q~3@K9zu_yviWnlD5Dl-ilW zb~wd)f3Q>G#u(<%{6Bm9$o5$CL>KMOv&hV)3FQI%u-j|(fYVL#=O}hNw&SIXT>Q0l z7RC0ptcfR{7+DJe2#DJm`p)fBmh-Fs=o(2Sn$t}Asnt{$dzLQrD6fh znlM_ngqLj3+&ASW>Cah1)b9A?NGEbHC?Br&d+r7l*=O$l_$)(T1rR;(Od!9tC=a>o zY{{Q}U5TA3u3ucsROy|?uJvV1>1#ihmg#&t`xg&cH5{TWB533Yzqo1gd1Nk0J9Axa z%D-QmWg01bT1q8x9UEoJSTNB5^+9!KO5g$azve4@WSc#?RMANedVbx9_EpML5?0z$ zC>5XO$>-MA+@(^>uQ(;M$Y{YQ^6i(AK2{{^GJaCGBeNrwONWq#Rz+wPqtvZxQL|Gacz zjygt?<#EP3wu=h)uPN*~#SjzP8-+!gnhBzaH#>W(#QA2Pbz+)iY<&NTw@oH+4 zA+Fy#(>T+^_t{MP&{{U6n3(+sE5~3N`kLhLwK-!p3s5kAVsbgml&r1xTIdD>Az8W; z+r*eZ%rA4_quvYNjvkcA(^$>jiE)5*Y*KrPZxiEeU%gy_!pkeNN0 zuk2*4x?M-_diBpOY;sL5Fd|Tf72A)z>5xY8poeU=S9FZ)c{DF2iJrHSe+B~XP^*~F z_<9R_e~k*bD4g}{`}NFbYOqOjXFlKzlTH0SU3|l)omSzB5t1eNtVCe1s>UUb_ucDw z8x0Ozuz>e%pA)jq8jWW%a)AujM8=C~xZgH< zE@Vp|^|)5Je{RZv#NPy&?pZROfton%%Sd8mzFW!A;!aP4Wa>fuwH%i!^!}CR*~8oV zwU0ASHH6z$@K0fUxw&7H~2lf^YAfg@QrR50v!Veq-} zi~^44>EqoR@RZdjpJ_4f{pbR&v$T2pafwI1kx6;paRtRms$*8y&g(-Z@1J6}6yBv! z#(ZX*+hq?^VdY?{!$&16N!84&Xz#V%558;5#~4#pnE=KmHXX{WQiCjyh7=x&`ZZdp zbJ&j0EL{DU6sO;R*sMlZvLpL5evMu=)h?3x*fhiCE#M5j+1izn`vTLqncI`m>!)@u zXEy#pi{NqBH9AuP<|HY}jA7z3ARaXnjN_*T*=K$d&NiC#_iSi8`XS+F#p?*u)b#j~;=!RqaoE9A!CL4W*w@RoYaAkWb-r6P2=2?bkJnIscJC1AvZJ;#FwZ(~qCj#9u#7f?#BJ^Yp6G z(nBnc*TEif4?7qhkI^E5#^f(c_ov-bCDq=hC~N=tj6|Z`e=oW_?<`U3!iR^BT@176 zp2oWuGq%dMRz-dkhyLFjpU;bbQ+)E@-Q@27Yd#(G1J&}>KTcj2d~V_;bHi>tC;|lK z5FPz+K|y+;=(ET7;cF@Q$JiF@?3}4O_#SDbZ~kipf<0;fWnzDB#^ZEaW%a4bQ=eMAOIZ^sFOUx< z*%xdbq`t_Y#X+T~Kb9x2B{xydRc9tY9BBch^p;!eos3kwGKapwos!jb=C4|csrgJ% zptZykTP#)XiYjKEu9o0=hKzD>4r&C^-LkX;EC?)c}r`WWAMXE&tY zRgrExDDCGJHNYkW0tQ4Gzxxt!*G}U@lFaF=sgFl1Y={8qDp}f1`{Adb%|Z-*Oo0_F%;dwjZacgD#2zw} z)_3{o_D9|S+*Q2uVOv?#T2q&+OZkZ#Y+lUYnLjs`x8PP%WLr>LI~Lq0W)Wz)TR5q; zVPswyG2g82z}tudZv=O-Ym!;&*o7PG!&MvrQ<@$`y%Mb-$6XCzuce;P-sn7eZsSl_ ztzASF_{}I7Cj`p}cNvjk4g`1PSwn6aW0?No_mZ_S^hyCMp}ayup^HUeUY{8>Rnb!$ zk2ta@sUQ2Ewf-2~zXn0Z)*KI;7johNAt4V3hg(I%1zpP@`l6<367~VlZ0naGY4hI} zBN+hSvj;yFn#oT7{hoP#i}(LQ~6*SGga!4Wz*)X+cm*ebyvPbM=aAYG~!DFFCCrA zNARC4=-#dTXud8YaIocY?`e~Y#huEZ7CC)Oj-Jx6bNHVf+Xl5g43`@|B&?I8cxqNne0 z99rGt1VWSpPIq!#%cQEm&TfI_Fw~7+uc#Rv{8vd**MuCz->exhZtMZxQ8^PxNeP3yIDxL+ix1;SaZ)7a_qW%K7IE#wNlX}9j5D;1*Y4xD4Gn9&F| z_Lzd^SSFv!bj$lfwSvI!iK~M!jr6IDQjSg*xdZ+gQ|xHaa0s->rsV}D#H$EJu&Yr= zH3np)oL@1^G_J7?4R<#y`3~BY=?of&K7TrWWn#rJ$|TH?(FJ5mH(>&MC~SY$RvQC$ z+>Tg_a*)z4OkpM*a_I%rSQlQ|%;!f3+;&L+eM1UR8*w-X2ZuWFCUBMQ|0f0Ox}w?~ z>-jL9!lq|gYqSQmo!CRbB2YfUQ){3GRu7o6XG?O>(&Uf}G3oJT8;{cWZz6&&!v`X+ zS=*MM`oNhyw%j(G`UV~_lkT_q==}4DH4`K$!uwO1KP_Y5Q;W*ZKL>5<;-k<(a#L`x zM0Vs55OrL3{fQRo5Az1JoO&9rVjiaU0y5_%v!8^euW4p)gZRst2^!@{+fe-ktyyrk zlfur7G)&f*{>PBWB`Lld`N0hHO?Afw5?ou~I9h2oc0+h?Dd3xOQCYcKv?bJG@QEU; zUBFO8?Tag0Y=`&_kjXK&g>LN-Rltce?n|oc|FHT^tgNQBv*5Z}V6vg(6`rYZnrSBH zC*LIbqasJzhLq;E5Mh1DvfSKyfy{?Yj`G_sGAcZm=o4tnT{)O?QslIS@y(4%hi1w= zlaR)Mm0Tuvh9cC)Y7689)i1pTj@_6kFZmWEs2V1qoW(*B^oteoyJ68KAVPxc3?!y( zGZQKBT{70e?}OP9UOH3go(s2uMN11zt77yAb2jqjiz}$34@K}qyZKvn?({izHQ@N{ z?{?o&758^AuYAz8nH z3T{;mT@ZmM9+$S!Vw9)4Rh&pf(-tn<-@^OeXzz{m-YAgG8*SU<(VFI?--=KMQ?zl0 z>(Fg)&RP7%fPCijBqllo4}YM31=s;mpe__mv#rmzIx%J-<>j?X&2|}N!a~mRy7br)=UJ-Xc zmlu#916}Oo{#Pw#|5q&rDJCU8`)9zLQ+w{e)M?%KVNutwUcD31gG>pi1zcLQ-Tv8e zFtyjD_;AYXb+oc(Zl4KX%THjfIv#O0h`C18E!F8x_+QqToFE+E=p^=D8!a?GYUnNI z=_meSPRA638?cSo%U^4cL&=0l+weeT8leJV-p ztzo9!D)uhIhWdA5YGj@YF#?u^t#(~&zlJ8ZI@^a=xYw3c{v~0w6Cz7M+UHAy2UzCy zGJBlg-b)Td*xr|EVG{qYK2H}L&|A4yUQJS0dfXuwcVhd~`1^cWI6*{;sScZrzz@57oFC5IJGwv!nwU z;x(pst_BfB499}(!$ug_YZ%yC=Em*;kGx1z8-XeXDP385i8nHjTCPg*I7MyKdK#*{ zUhR2)=tS9wp__snWeGg~)qLzuzk^=zl&Otse^6 z=s4JMxub(~XvLm`@JM_pO}xS*#8GcODIEY(V)b@L%i`sC)3WR1U6lxJ9CCg8lTP>J z#3%k40bPZ>R4B|+44KxJvs%Qyih-*JGPZQ53xbivPvBA*DDex%%3FBd-0}P1(f&jQ zt_PjYRXV`$L-ng$apeLJ5cQ!gc$en!*jeEg@GE2A-mb;kW(n|dSn z@A}{6Hx7#Fed0U)D7oy);`R@#(Bu3poriiax#sNM%pNRr2XCkI&JR*-?p7&G!X7g8 zK~siqG94wE>4Syf`aWn&qQ`w!gg5oDh(Z3c6!w?HTqZjI zaSDqngeygZspqoVO!%oTZ4h{e7_UgdkfjSU)vQHIlSUv~qjlvvna4c075aCLb0Y|@iyHw$Rv{{@DMte}p#z*l;JtoJmPA{{^RmR`T>TYXfOqR%Zp!phB*Gl#j7K8k=0-OFPo(9!uW;Q$% zu%CDZ+{}Z=|LcRpyO_8gtKV`+#9IlXvRHZt_cG7oeeiOcOGok8g!E4YmC-95emF1O zW7K804kp#En&|IR`X&~}2F7>Q=Yu}P*>kH1r23hD@bJr@;5J2#FwZfYAR4+^PZ|F- zzGaND{iT0EyJr@}6&@T=dvrm**E5^{8aDcP&?2nC-0RMzNXJyT9by%s%zO@E?{h|` z4FoyKtYLUzg!)rg_hdXRv;N^@g@2^BgRAgpML?$PoAgl;ZN6ciH;4@li^C!D4Fn>8%cL!G(w%yi@v%fGty0y#VA{z%OoeH>GwEb-xJGfi z`VHVt`P&FbuIBl6t1Z3G&VOyA^M6*y@5Xr30Z+Bk@|AyHvl$uWZfZB!l|Wi$NQAnX z>~6t_)3+^^P}ld;n-3Sxvp~aEiLxC=-v6lW%paOs4R{yLXApFJQ3UpXyU-751YX+h1)kSCBqRY-kz0 zeEEOiP?k}i&6s|l0dOsD^!x(D$@Qgr>rkyb5PDy-HnPO~w{0{~RqZh`{2WU|9?WM; zMFyiC-Qrjy>()JgHDdn8^n8dZ+TV6;I?>Euy~$8`=EA)W#RH}EKmG1H@IfJ=oyn{} z($e=z92X88fA1|6w;jC?E%NvutcdwiiRGebah6c=#mOF@>G!|uYyO8qd_Xm#M`BOi z;@3qhjn__@+w>F;(@N22hCT5*hck*Biu;PZvqF8+eM$vQpNV0hUBW7{Unah7jwUp}cYB zsXST+CY>R)UblBORF?1m9$xj|2}^BYd6Wkrr?2lq;a)&UwW3FchR(rgf1DjeL4(3< zuA%73%w9{^-Tu1z=?kW%&Jzs|WEorLL)5={eYK?nbj7RLfghKn2 z%_*MeP8AfRpl^MTN!LrkHM;4Hk?NBtkN&U@4HOceHvY0J5T>kM5CAL>?|AUW=>H`o zsaO*`BRxMQ^l38O|9GvDKtf%s;jIWV9rrpSeCH6ZLkXv z1lY(vALtsdR@IgASdMCnjDg6&;`E5k4`litMSYZb|MIp-P zsJrRX!@*vgOAh&?vr{COHV-#lVr){d-oTP3X^3C~VRtDU&N=GwC=5YIig4s~({6G< zT~JZ6VvkALc%-@UAopB~tQ$k;?(z9%sO?-*Iz7e&~C|^8^^k%VkD(A5vGDd!? zby&{LBQZ{U!aweHG?HC-wrhxPcPy2-9Sp~2_RNE!h=Bo4Wa1CnT9eU1DUprmn?*TsaL_ZiLk7 z63HDkfbV!vNXnqP)>uclpuI6n{k|A_j|wkEr^X-XhqBq%7w5Tepl zz(_IlCJIsnD@YAW^(H{*p+itnX(CGRh^X}5ArT3L79za}p?3@rAcXeb?!BMmJcAWAl)0}rI~muwU~wuw51{2(6VP`)vea`k^jrf*T|gqR zSGYouvOf+g-E0e4`_{EUVOEZ6#XErZRTt_BvFc$3uzKBH(VZREkgd++3#PJ#uBt2R zs)(4iIg7Xu)te{w?N#9-wV1$)m)vu5<8f>KDsR>7+DB3K%cEf?{@B-qCoOhwt$?g< zl&sSy^$~7#r@#(zrD6WGO=ALgDIQ(KJ>0npbVcJVRAo*O1j5isDRw#8KK@M!XPhnR zmRrhe82iE3t?%svja!SJ`2K|{2O-Lud%`FmFZ<$JymL&RY&=u^xs$}_w`}90 zSzWgNl~=x1{uTY5aqI>QYvDjwZ^K^24RNNj`Jeb8>^zGm3j})|ap5itWG%ZIdjU1}Y5g8BOxovtk8R{A4A0ZR7!_t# zE3KGDebj8hNgm`iX{rab8I|E_*n_TN5o>DWh1_oJ zX6OeB7H>2j?&@1DGVTL=C8fb>A6bdKP__O*X8w#jjwf z)O`H1X%MwvM9WD%GJu+U;*@(ZnT6%8-y!l>;BX;lZC*p3(l1iqA=aCs9nVI`I%RL6 z#}zMjM2c4%KqBF28{PmF?m+scg4m)o2jAhV+=(D>QTcF?@Ifd)1`!JnF+7i(|;YAM6Lln2b7{1DUeWi={g-8&8FB@SuZLmUG3;i!4s1~=aI%^$JhdR`?+$)r1taiWEvmQr~W%i!YMwlQ~6qd#KGJdL$}~xYV-;=#%C3J z9ymkZbqqKj=N78aY^kmDuh?p~+l`FsGl(2Myrd4a&f~Y!lodzgA+y(Kv#xC&B(3t=aw5SC&yacskg2VPK%JQdgz$i1q%qWaT|BAD0ReSG5_wI#$HCtBx8G~Z`#R>&d8Hd7&Y(`O`(U$huwBy%Ya9;j%i zBdm9zEwMjJcB4&)w>Ne67W`KO|4V=x|5JRjch-^)2TOs4y26AW#$3vOzulGHE%=1+ zY~Ie|>+GBctxwyQXP%2Ub8IO-Ln$I=Ecvlhs^|DmW)!#}AkAlZ!@YCeReoCDtYY?; zoff)?vzQXsc9)mAw$&h{ygugMjL^E``2j>2k5uXG(-y{_;AOWu^dSiG>|8B$al5nc z3k=f`i=-y9QFm;gJ6xW-qG^Hy#9$n54Ha9(3&taEI6h^4yRNITK@>ac6j%M1=@eE} zB3=HP-$o?n-FK7aZ^hy<^P{)wiMwSGmaC}i4Od2X8yD3|97wg*(qA|&IZ+j7StM)> z)&$2d_WcBpev^LM=30I6ia<_xnJAhCc7)eDbMK69GMg5<=@N(OA_}U~ReZ|9OaSm` zcc{W9)h5a2WtP}@m(h1gCJ`nMjYuF%979ukm+0O*c6ZdktE5s9EeFOF_j(L^Gjf8; zKQ>-{WW3vA=Jj!HTNg0kyn6Qwn>Mk`M#>LJN|obDI%K!s(H>s9(o_`V&&WMs6MX>z8|nvr6+oBqla_N@X-)6t;A zwIjS~Xc~Oc<*fg8O6Rb9Z$zagQ61CIL88ZDa7G*hrwFHpX4r)1OUjh!Dd|zMo4qm_ zS+ExJD|XsfSP{LedLLkmDf2|QypWj9V_@>sROfnk25&YuFW+;!&|F$GLiq+&oeQ*C zoMr+GEPcdX`SR&i+rxqUQZiuMck3t(l#=e@~4CW33321dP`M{dxYwJ zJMus-W8+(df;uP4y-ltQdk#8WTs9$AE^)1>5l40o+ld(e}pUjOad%!?B9Xx zC-YiL7nX9Na=5sUV}H*?9d~5-#E=Ic)D4i8%qM*17q^jHYoRjLwXRW{+R(#a35J5) z_V%SwB8rn+qPc%C?J5L%zb&ZKi46j$^y)0f8{g)mN=y*b7$4N>x0WiDKCMf@)E&Fj zH_0@Ek^!R8D(E_`7b5_~96&(jKZCcj?Yw>omRnqgbW$I(AjMe;PpztXdJ#4_wtFs@ zw>d3$rd@kcm-^E7D3lUp5);HtzZe3+>>g@r1pjD#`thT|ijOFqAzDL~+uOMJk59q` zj(5!tBvfjD)Max;VS%6{o(Wz4>Rx(x$?O(*^BioN6@jtKQxFdd>cFJNN1BC(wP%=N zhMv9wQVQCpXc~v?+4=M^gTR#3Hb>o{yMvLi&hLwdwU1Ty9D}~AU)W^k`ZWQg&#(N6 zRroQjv$u=>yBzm-{1$K;_FARDr5A$nS{g15PeT5SEmy#3aV3(F4lhe@PPvfU{D*QM zv&n!zO|~Cqqrto{+)}!YzA?r;SMUp8!}|faCWZ|DE1Xxnw=TU@+y&cKobEYvuh9C0 zF&vwf#_3WUF2;{@2Y*q&r;j_cB=a_yM%HZyw7qN_M~Jhtkj34_)fDs@ARqa99e#8B z290Uwau@8_oKWD&-ZP$=M`jG!I4zClL{$jR%kB?ZW~#)xU2RITWmG0(s&KNt<*#EL zI~!^PKH{srQWk)E^ESZ8KvVzytuy!Ew}aiZJa$d$pd&00#03tcvh!;&0o17mk)5W{r4t8-q6vzE`W!Fz{DQAmXQe`>jK9Jr3SISHgO`0)Y|HB^KKP$Yq5XW&C!`W zT!yBR9`4pIdyZxtQhMPU8GSnF}VB{s?bC6FPJ<8i~8DKt^4jmvh-i}TM15{ zhzoL4?z`!Bur1h8?XyyIKdWcDTHr36D5rG2)+iBc?1S%yrDRT1<3|ANY{qO}RyFQ- z&s2T}D|!hVGWYXXkr#z%gNcR3pNl-b%!P7KddY;>6o{fU$ElAwZMU*Qo7zG@)Q`~{ z%i1)3IxU_b3c2yjblJwLA+WV91L`|(@(=Y8*dQ$tl|&>*i2T-XDwiuC_q1yNdGMy< ziQW;<;jS&W#9h%hG1qP z7=wps=4TQl8-|}MeRxaYIelXX_@!A4-3qVkHwy5^lGt0hGRA64jP1*1NoS8ja}6~H zGw%zpPpJG8_?;i7axIfIU3fy6$al0T6%*l8P~}xaHeUTR%En*mmyoR+Kzd-$XV?)J zu!NHYX;B}OEC~+URO&0EL@LYCioTO+Lb1sod0l$W1EkHYB;8}nMoRik3NYF9sQ@ZF zt$Pe*bM}Io)8Fcd*G-zaudaf`onFJ6s+BreKDF^iX4*|v9J9jcI8CVV!mY_4w8P0* znDwt=*dYa|t)bdF8p_{7Z?iRWu%tdrv3EX)(!d}jX1Nq}E6WA{7INxLp@dSkhKZV2 z1@S6+2Rw&jZc#JYa9rQB-H4-dTzkkb(xS%6{)OW!Fvf{(7*+%#Pn zLLZ2gxu3`_C=|3cMSacvPJC9iE8wtk^lYZx>?v-BoGZvIQ04ShW!A9%?yL>J(D*ktv_(44HbWQ1n2ps?ADamxT_|I3tg$IHY~Jow zX~$+AQ#1(_iK#1Msh&_5{9WtxT;PV2yXvXy^#zbdLe(n6Sx6{)zS(-m&Fi#=_wM@Q z>i(e)b*SmTGANSpH|T+^<3+Cc>&2wArrqtms2gkB$@wI~HsjHX`TH)wY}22hKS@86 zPMcDuP>pl#>nuw>$v)Zc?~(m zBcHnO&OfrH=|ZTW@RIJ&fWRF?6I`WAzv#%^PrDhLacRzwcv%jONP{y<$#r%#Q74c0^j9cnjl!?ytnvl`E{;W5#2;)11NB zAnf;o&c6_;@rHHnlvQs=Un|W*$wG`SdQprb>NGRqz@khdl&W;H-5L?_IctgM(?yVt zP(Jz|$oJ-@kKXG~t>xn00f1p=eYzJv0%S&X5W|oMYqhdk0f+=OV7T-)J0d{3k7P2o z#sm*CSGAb}*Z>s!+N@u!R+wOKp;YJ_oEbY(f^S?j*0K?sS0F8Grb6cJ@#xUQ4mE`U zc4vb>b9PZ+lLR-d#E!3QC@sHw9QWCmqI+X~O^nTN+WuPH6Q+q$T`T&u{EsDbDX4xK zdzKSEIe;{=84<{-e7j~z`S)XFgRCN%@>QsZkwDH#msP}0}M+4*pvu}WVb@yVMfa1K27CjsBiK& z6WIPZ`VaadZzcL*I%>i*o%f>S56V3_U9becU}=G@-#KFAt>BfCvR30rHcDwL#+ElJ zOFpBKF<2rzYZpbksxpkU%|_N{amRGxr}}d(A6h>YZzVnmpc2{Bq6%`=qxhSN9~c|X z&J(z3(+#DPZf`RS7YZbHFuX#3mOnGx_halz#(;ZBF}GK?9&FO*as(fHRanF#$PXV< z8+X;i1*K}SROtS|6ykBmmm86p*{nG8z}7I>D)){>V_!|KkkH;h>W=Rj&v&p{X6r*% zSWNk`{o;ZpA4_I1V|g=5C-JtpTCOzVo5a4cN+U%f>+M8@T+mNA=nbD5YR1qc5S$#8 zeZmJKmY!_d0VxsI^{CTJrKH3Umq`8~3hB2xS8>T4bfY&9?HNN*s^f^r#`kKJ*-cNy z;h^JrFrWFLE==VM)Gs1Jy=QMJ&G^yx{N{Q>IM<^;M2XDC=8H?#jkkbUcv<-JtXD3 zX{X6_BtStv8}?_qAZANRSFUM*?>HmUj@xx(+c7$QV!LoL-r0N8=Q;@9xVyIH03SnL z@r)Ms$MHS2^Pw#$k0N00o|x?bF|RlF^jmDlf^I=GEdeo0*RoAL(kja7OO~)r$At$} zoMABaM_m-(?)=x_UtxsL&fd$BBGz@W`2-?HIe`XWWeuKtnDSyRFyu5r`G2?F;{SOY zB|jE9)D~vY!Q5*DzAW)Un;7P`CY=AkW^9i7H^~P5l`XOa#d}J{*;8LBwwxjKPj3-1 z$GP_^y;Cti&BOaM!<6)MN>uax=;tqP$Kx_%62^wEJH)rIosCqU+F;|gqrsRi_3a;K zd{>sXk>YqI0y-kH-ih?Kx9)YrZVZdA_;a7Jl({6!9}sJC3vo%GsFe4U)oko+aNKKH zn%&V4tB_y89$yWoY-W(0LRSy3X)Mb~oP6Ff{SSDyWU{WYTRCLzzE#UwWqHCPz@kxp z@tuuQfM`{#KpGBGM6=taW?q+j++cqG#{X%j%9`A6Ut`y zt&5|`^xNQ4(V$^hQoDDPV(B82Up6=K$=>2>2607c_;ulTZT97JCt*_Yx?y)f;d9nj zdqkD}oqvzCx83G1smW-(vX;p@-q&jD5csI>7$7>B>l=8UIUtyo?!}X~g4@&aX9g#E z8CxZ{U5AIqse3Q59RSWQJu9k`zF%eW)g;d|7-K zbridM7}cxb3M8nz5Sw5)dgX zzO?h|Jn3nq-%?yAX67wkJJpN&-fELwux?lwK9V`Ai<69u#JsRwQjMW{B2&6HOCKw# zLY&=%7)W)H?AmJ-ifVR5pB+)m22D`CTqf_bcj%p@>sr(7&kkBl9=pZGbggk=)Dt_0 zWimk;k>O+0Y#N<8XKNW{n2#$6tDfeaCBb z(dPbbA)MLcHMkM^XHpWqC(Xbw6G=X)p7^s|t+gChtiLiad9(1FNm0{3gOeoYSkIJM z(u0MUp4M!3LY2Qgx}-%k#g^9czXd{_cu}5kL0i>!Kd>_Tn9L-6$QYc=7MGZC&@Vkz zq}0SprL}*-y&e%KFHGokk0+o*?H}Rebm8XK`<?24_(TOs`wn z@F3*v(!}70hD@3_Qk#B}m~-wmD@>1?9{~^}T0Sq+_o~<|9MYIXFf8JXQ-tj3qK$jb zVH!T#XmkkwXP&H;{MwNR%yK8~9g2eLf%qp_Qg-;J(|JYDuULuXCaL@50%8Z9I;Wmu zrJa7bIu0*3ulnmzp7LaoXzg-Mc0qlQ>`852PeVrKEYhp)EQV~SPDN!_&T7+V2#U@L z>Zb~Sxhn$3mkTv*BO<&;Q}}A7%tTg1OX=6enC~;Lqu=pSi(HNqZ@xz8rQXEGt_yWZbo7~?srk(9M=ASB$Oau@$Wx6s6&8T1jo*E_bhiK5uxx!I z%g-o7+G+RBrJnqwnuh*rij~e@giO&1bp0`s3FAjH^hrv~p|LD8{G_Ok39_*iDJ~{l z6=9J1(8XrOR-O9zzNtp_dG%@`aB|@1xjZ8n?d}+{wW$!+h`vj_oe2&bgd|4$LJenQ zKjq}fD(*zU(rBNAa`d3T^E5Dj@*RIm0E4KG>!WbR#!W*x4GfD#1E5&x%gK;k`A|mj zGCX_uf{c!isY`Ba>mjRa;|Ivdl4fGn40XaY3EVFkP~05a-~QCOn^CzSv=QL_Wd>q!eThUwCyI_XwEjM@dZ5J|QLb(6Ar*`5Gr)@>gP)kx1&JKB-*f zc#p!A>JXRGwtYueBy`t%i0@z?Hx+hSFifkZJcyBJ;XC2p9?GeZp0b_<2Na1ZQeVM< z4~MS#l}a8VZ=SE^{)@OzcVTxraPzkJ}y}Cs`e3@P%LA>@^^rGp=#xul+V><%1@6Qpx5XNN^6V> z%xtvn>>Q}h;Ch`MN?$RLvScc24b4wl_mD0~9B_9Xyc#|K)nm&gSp8_(m|GQL4h9^B zL<&8<(Ko`5@~L;ZOlJM!U&rAvIX3>Wr~lfCnW<9A0bN4-{nuek&S z0vfVXvXbHd1m3jSzQPfPSf_uC4g)sRVf+mmRo*2b-&Q+UOzT8TMbQp=7Y}SJ;PwY2 zlBqt0VP5vJgOGm3^0x8764wpeZhglXKsQWn zWF^91IWD6%UvfEE=#)Tz&rqiS9%C`TFGSDE1^iT0OWuE?goXO{FjKSMXgF%wY-7w+ zVPpS1OA+)`&b^SGL(nSt-cn*^eho{=O3+(}9*#Gs4msC29n5{BBNs3~vHcEW70N3JG%-4vvsJ{9MyG zFU`s=xbbI=<#PEd=EIcs*5^ z`^V}%bI=`rwSRj0ELJM2oVl9?NoO(*{s$#2f|1F7!dlyYrmjs|^ z2L%FW$o}Q7w6Oj#X7%M5bwZiJ&?k7VYHpF3EI!p+JRojq;x|75tgb&K0c(v+JmiXy z9%b?n)mR4}6B}e~r`#d8bpE9g(JdIA^=5L0uenRD}uMSYKa{E~%V|F9I}Y zu82N5jyf5=+hXp#$q^fVPjTIXcuV}z)?A1l1_w#$GgcTer1og}d_d7eKsS1r1(nYD zym1U3klfMxLJ4<)8^Ah-nI(=7Dq(dU$Uxgf;P&-mbph)%^Gs+%1SYUKA8a>8Vm~Y5rVjq}IH%E%eCv`U=rVZEUFv zQDvP|_fVtKVDr_`SOEkYbNE)_&`Gyl|G)5)e~TryBaHIR#%|h@=jp$rm;g}B7IJ}v zrCkS0=prN@sI;&BT|x6Rj<$B(u@%`i-4z?jWa;JG;qhZLBJ9VWoinM&@V}v18iVuRE_kDO6UFKN0jd z3Je@P&9c6#PY*;$gTczwe5AKrns*Sx;G$JVxcyfX<3GRHJn!-M$f-$@1DwMq6PM(z zq_WzPSC&E5p-NAOu#~S=mqJ}#7=r4-yh%31!;4ga^t4x~gW9QLCHZUb#T7XcZuW=^ zNcPB6?qDNhXS!h2eRb@1AlOUa2@ozsVang6SwpNXB&qzOhF``u3rm7!tt-2D0*Mxo zMklN(IwATDX6F}^%545NV3o}yRU7DttqOKf3|wP1i<4Fk{3iN-m5n2C=G*>2Ooo}s zN*~V@CU&?JkBieFI$S4Tu7F14CC%C-g=jdfi{?R}<|IMY10?Ab2=guvlu5g9^eMU{ zyz$h)3Phbwe*DcEQLRo{eqq1|+hKj&W4y&@_Vd`M@s}AMN6YV;H-q{V(KNXs zA^l=UODktq74;4+*$nmfMRhFgs$ZRvvaz|kXnBOcZX2zgZfAMZa=M={t&?$e*7tgD zQ}Dm%PL%MCt_q`-+~zWRm4liE^syq{qsM{Hd8?dy{^R1#;UKSC0`##TB4EKH!i&$4 z6g*pm&lvVf2#&v*Leq9bgg554n~RDZ(Js|ItlZ`v;XKJa69g85*HW-$*0_%}=~Q0g z1r`YS2K^!Igx;>JCLDC$!K&JI)IFjgP$I$=EbLAE#>=E{qPC+hv)baryS4`v_w2Cj#G5J+Cr#<_ z?Emsjg}h#~XN3c^?pS$hAutNSIoBvljT@W8 zbc`61+IO?&1sq}_tGg0Cx-ch{R`=V|Ss~xc&?hNUV%1Wu$pAKy`&NUs_jJXMT5GKO zp*~uVyy9Q%2^Z4{d|`adU0vMr-#=Eum}OOyI!E+>X8spR?(k3wJ#0Lq;_zj&9Ui9! zJr@S{iM$X_4f~lQCnV?C6%QUyiU&>cl@MbC@7?5$234a&UF?*)hOK;mkEC%*jeh13 z+OLBceFd0;#Z{Fkh+>e4$LA}5>6y?oq0d!DWstfRN~k!QawcAt{lCdX zBae1FQk%geLgzSEmRd77%PR_4R9KJ{j%#7WjB;BP9I|3Yaq;kZ-1m*vg*^7?%L7<6 zz|W3qAaxv9$&u=$(?0yc$V+qpM?kp0Jvk&t`Ql^;AEkw{QUE-me8R;!q*_@8#XRZt znM$%*ioM?Gw|g#7kuU8@h`&BgsB^(>h|_yWfAy7cl4DSzv_EC+(AoKTP_SRw3%)Ai zTM`QphlL^Mx&UFmVX_S8^=#I{rCD1JrY(uSi=g|b3g_XQ&uEF!5vD9aYx%olgZLLw zRsS5B*93m>`kh3xOb_mK@^w>H#eVtV>`na9bWFgnS{s3vLixGHB7y?Hv_^2phE}s0 z!mo0C-kJi81+`6_JW&tX^Yph{e07Z(?L&%W4;}-f-WIZn235^CYgHG7bF5cHb#OL) z8asLPBmVZr!@c6o^tIn^ybV56L3>cOGq)2%D>FUMp&Vv$tSa?F?ESh&`k0@({$mwc zyZS-h1xifuQMjTL!vb|QBv(E8%Ix_GdeOjwf9iC0#mT*tXF1NPl=lw)H9-ppPQ_{C zONxR|VVlKPNdFjPM$tg5RQ&8|$7V41m=#c@K3pRE+Jy7p4xs zyQQ;FjjV;YiUr|+=Bl0ccT8zzGl{sr#%I-~Fbz)%A$iUFs0uvv%3(7q|D}*o<(T@R zMfn(iydlOCpsx=j~BC8^Ue6?qBS&L8JPq>OrF5+Z82RS4yDvY zh`agx6lhF_AqJnFhG>OAqkiP3tM1Huev^aqTp(**7i<0uXelcXm)gm&qHeXTrrN<% zD4^d2oUyCT2i6I~hg@ZScnFD8r?N{KI#^^b}K*N3|O z3o7Kuj0L7>Mt()nxS46kgLi!NjyK)(6T5XKmX=?Zt!FcAveg)=iFANS`fW);a)^9^ zqr}du3O_d1H7D^{+NJc{-1YE=wQ?_~d>@`aX_3Z4I1YJtFR{c#aZl(+n&y$BN1BGw zw*#T4dEJ{IE0jr?JC@7b?;cV0OYIL!(@e=+EPpVo>M5*Ly>t`IZR4I~u+sMu zPD8)hw0k5@>D{cpIwMrj6$jMi8JMPz*^H>-B8{miYsw%vqRb`qM@y$z*xxZ@L4hnJ z>l|7})bUb29TpQTH?RsUx7Q;1FOA$x^+a@Ikg(CY|HLEVA#U{*k5CEGN=a+-0oN$3 z;sKm|kcAteP*%AMr#_$J`{wuufXDryp*d@$Tl_-glEE^qEqaFo9CsFqRMa1 zQTgv`MQAfM?k8U-vG_cGo-4^+FC49VjJvdF`Cz>{E1Y3NTYE@YJ?(6jDado9h1S)G zsJT+oyojDyiWf$;(+W#6!0f!Lb!80wjUu(xudP1y()V{O^VQZ*$a$G-=)zU-Tu>N` zu6(OZf7YaVEZ^$XO-Ni&)+)h|8OMkhI~tIjZd=#POq^Ah;PpydX+gx1O=b!GXRZV@_8+FWPAmQUi*Vx2$aM$s+QK&D6G9Y z#uDfFU6wA1?2)65Y=4uySGe^V302_{u){d3L*S!y=jtf<+IwCSaep&5!c!z4z7jn=eAB<6-}vE5u5|)c;z@~AkH?#Z zPOpMyPr)dC7c1S!dYdQT<7w%h^%^yGF)fVm@jrGbzm>EXDPmo9&`GbP!kxHyWAwFf z{-$x+kwLcuU!>p!{eDPe*?9HB4QQWm;SeBc%mT9MbX<6o6H3!ebxv+JD;SbHV9{Cq zXNW`%g=b%C#_Yot)MCRx7^5;(b%9>i>V91V=&vn31-?!N^yz#5mdcUCc<)! zwBCR$Ei>INkJG*s48HIIIz`}rfAm;ybmD(h9&Fn7J8$!@kS#^z?VNA*q0EWaBjo!gIc&AFeX#u!mSkLbcb?Q6u+-yhAq z^r1&-mDDASTF9ca<^`RE1Cw1QcYR5xx3*8S_h)C7+0biXfiI3F8d0|@@gh&Y#||%C zHScymS^}6;?!&_)H9R=LlVYBn8RbqEU`!YZhnCh7?y?EhzOo4xp2S>~CPD!i{%{YY7KOL1bYxT!}m@e4Z zWcB{81})p=3%Baq3L-OsKYy?(I_+L=CgJS;noB<-gDH4!ZaRVwYl^sncq8;2)4{RM zNKhQ{z^DJ$A^CScF$E02R#aJ*cn=6ToGKd=s^?7O;44)?%%IL&w1QO%8xA%abPu|$ zcgxR8IX0jeY?8aaVLXV3#OBW9z1UX8RJzOaf{ILPnO}3(?$BXT&7?9KWT@E3dnVlA zVmrV3+T~g?_|vR@ixU=E79Z62W9!SCjY%QD&XIm;*?Z$ncBAxz2g~+0%|o_|p~x{) z;SPiR@Bd<=Sg6FBSy|_RLM+xsG>0PN%gT$o3_+1s6vnhxo5GZ&6aq z&V!FAN{t6&_cbSsBerg`Ewj<_A5%b$xeC?;tB+d9`-&kK@=@1kV5fJ822K08y}!|C zEWORagodmt{-LN#y1PPC#2!-T_9f^!QMZVfbDt*<$;^lkD8Q4=FIKYbQ1a}gB0cJk z1%tO?+p_Kpd~k5;Xp<)}LnY7UUv@J@YI3*pE*ub8Cl~daUz{3C1?ANzZ}?ea%oP@F zjFzs3lfCUruSE|c5FOZwG?k*9a$`rbwlU1A(T*_kt;=1Dsyct>|m z4ygK5g#}c_&%fI7F=Z6$D&4tRv^rzYrjR%H0NzxPf+6 zKVeEZYX8fjRUkz7%TVdQxXNAze+KU%aq8rFPuq>An~RjZtSkF7(a{T@BbZo-XNGap zIHdKJ#cUMP1ePfZUhEsMceqVYol`8wGp8-T(7X@xb_!{-Z z@5}#L&?bP=A7J&4X%C=krXRvHtv~&OQK{Ssb@2nqKw1_*lxH-)z;71lmytaj`F7#4 zT5$9y+~+PrYhrP?Y8%N?fxSfIf3)K6Us`dFGdSDoFRk$7p@g~*cpw@UYu)D@7KcnF zf?v;$IhQ~o3aR{VwqxBROz0ULr%8;M?n4OheZ{Fv=eLhePbwWPx-OJwx{A z6JbZlAcDKs&L@04>K|NKxNWE3dCO8xYJzwzc3H@(a{ubJn8M7%TOz>WFCTeHO0oQ< z_%_?pLA@0_TXOmTY+w@>Ho=j%4>cB-HQ%e*Izxvz-*NJdJgx{D8$tZbBGf=OO6M?t zGg26!`^ypsX2#4+8q3|soSNaBe5nL*#WV{G3m~9@@=;rtIdd&B^Zi!8IhzTtADqdg ziRao)3A$-?gMg2aV){axKcf-_1M*{Ciq%XzS`~dvQ&^d-BpZwz^hu{>nSoImSJ~ng z91}D-R3^6cN_1bPlcl;&8bWN9d ztz;0Exvd+u7t^S+s2~ z7!xk!|E{kbDs*p7Ei0~CO@!xKr%OG7{a%@t(J@m|TGnPil-PX(-P-mA)B5I9ik`uQ zLh*mtSQ?KAiUc|b0$=FD94#=4 zKjqOg|KR0(j2K}mtrxX-hYjqp{Yx)|tw1);@b^Vr+zWbrwas2wp{01`VPdLly^#j_ zGq?YTY#;fq(Ocy)Lr!kJp0B8yb%r)kf0wI>qzC$7J7;{rDyG`&net5W;T`HR0W)gQNrY5TqjLA-_GVpvrDR~W z3d8?cP6q_IxL!x8LJT?9I=ZKq$x6R`vsm1YoZ&1 zGvL&%XQP%#)$#^Pr1AUn>DZ~XoSXIJQO{SWr~AIZ>ZW@F$-J0lUUgz?MN#W@MeH5p z%FU_h{$bT`#=#3(mnAXI{gdVhorj6nhY=TID)SU5A8GKL)xQouWb(wW{2mAOJu6?0 zB+t1BQx*F4Ys7BoXSmSMYsM2=Vkb86+A1kREMsYO4VIuw zNg{u^?m$>#Bn1c~X-;mlg5W?=2qg63p+IE&r@uTW0Z`3j0~!_gX}g&!9@?f}=8I~d z+&~-M=ya&>sbT_JNKr+zejJ}l5ZL}#==lJ6M{a1PAa?bUZXo&6WPONtYH%r_edO+u zkoc*zC5iPv%~#G@Zu{K4uoE}ZHvBIcYSjsUa^|*5FaQ@QeI<)h6cWb2`;?-8KQlU7 z!$%N`>Gk0j)jckq|5dH}OVvqNLy>WA;{Jz{#k+dx>cHu5q-&(poKH2>!aptCgx)np z%dUpwvKCuHABQQBP{C|!My9yx*SKvdZsS#xAQwu%#_C-SG$47pgWE~f(&>HC**z9J z0tW*70y+RwbBde3kG-1B7CmGiRH#tT#W(fl2$x!h;;1Doi?nlNZ0_16O8tg73z?I} zC3sg;U{}ukk&r9$i+FC64Ru4k&ARCSkx>4CrKTJwIZhoc3HHWSI~>`ig+B?ipid*d z;J(*s2ONfSBDXUNbCNCd#w2y;nM3+YxUC)QYK??9->`e)GVNMYwQEPUtH&&zP4iCz z1@dFS-WW}d!JBSWKz~|Y=JG|;N-P997v|&t#0t~*5~!TR?dK@5Uym+bB-EL7lJUEc z2=Vpe>3B;L?N0vYg6;_g;wBuOyeaS7NxdB|Q@lFW6uhmt83G&(~%vYy^f}R<_%Eem($Syd(*MPX!;W=T={DE{rHw3NKx; zwp$+?v0A-;bSL&^uSA7$)xxG7bXqNV8o^GoR<-XJ`fz2QM~K?vDZn-D7%=KZ{Sb{P zCF_aPR2w*aZ{WRuTV4?P=G;R)h;kf{8cghAB6;3PM@`ZU2x|#`&jR|MSt9P~`7kgh zSNAJ3m=~HIe;oWvF`k{^se$f_%YxZ0Z~?Mtt8Gl*9X@dA<}Lp22yN-=d+C{?gTuyN z5a-~I_uYNq1f@46Ychw#$;jK*X9(fSwm4({i7sk(M4^9mKbs3h(tvVzPBebhGXEqk zC(qlWq0CC5`M^a`^+tF4guvEC2+4X>`EtN2_VA$mJ?VOUx<{l)00~I#E-j(*WH}>$ ze=s$G9EhPe3mw&*0e1lxN8r^9USJG>RZWK&WW@1BHz&sSHZ&6AOPP( zSn1L4XP3=KT)d<&4aB3l^|9yxWQDL(%Uw3SAu_5#eq8=1+ydmcNQ+X$^ zWM`HdnGw$H>qYi9EB)m^AD-35xP~i$TFPg`VSo&pVE0&|3kL9nEXFZB>_>sb@|B;N z=!YRO1FyMD%_FEwys;bPS?h$pjZ!4#>6x^9&tpd*VT0SfDsDo7W&GGMsswU&lz%j_ zb>Zj9w46!kglRlszf*fFo=BZ3`{|kk?ZGUciA*DlFT*7&2pMTX)r(b*YfSlZmH9e82b?3dfMw?OZTmYCs3!lM<|P^MlVi9+7844mBM3V z2u{QNP)3$mdQ`+5Phf59l)yHt`WT3DE&Nk=l;`h5MbTknpEIm2`zE`GXs+u|%ggM5 zunWtdR}VwBT@7isL+oYWs+SV)`fz=jn^X-Ii%))eERD3O=A*EsHm*3Q$#^cCA6R0U zP34^)@o$j#1wttC7c^R(?xJ((O@#HWzNBP%g^Zf`1H;oIKtQe9f<_ zn$=}lfwYcdtT}1~&@}s6BG7MIzu<~OXmz@oRl3H7P~<-TwN$fS@Um2iN;dcuyJsCTt=KVV)MyGdPhWXrxXFUbH@^JZR4;&wM=bk?FgL3KQ;-D0Pr zQMbJZR?eCDrTCLoE`PU@;PBI(DpI*_rutY8^r(0-82?be$GcI$Hir5$vF6KR_jOo% z7SEjWq=typ{jNOHL_EH`OEWzy9K*t6|<3G1u~1(C~l% zUey^M3T5BEB`-@RABoQ0jmX4G`|qZIDFrQ4(G zK#QrM^|^2dk5{^sd%uJ;SWc~6;7@y!;m5O->=vAtL8KLYj_oc%D{3#Qtu8y_M`+K6 z&XYW;OAqz>%yNbqKUiJq&>121Kun=-1eC>w<HYWk5;tY6j?0{AozALi3rb*QWmW>Ce`8iJ z+3nsCRe)c(JR#~?W{(>Xi>mjpr6>%g)%RwkntBu1rZCj$yqZr=@%YGNn+S;}xyh={ zPj)PPN(G+L{1~y)9kPxDEth*J+7A~iGe#{E4z9Iwe&nY5v^bZQl##b$SAN;TEb}nP zcIT62*VvBfA9HL$^#k`gZu5ngF}!+&o>2kY=_P;5wd1Ay^@}4*X^(CcL`ux7X1D?> zKA@r`Y_B#JLo`YR*Ljq}v~=`?@Jz98Ga=_gsNpanAa-#hFb)jZhD1fW+FJD+CpR-R zoWRmcH;_xVu!9lY9p4LkDZ`BpOT<~>z1q$Bq|6)n&P%1ScOt#-?yVm&H1ZdQvtcKa zK|884Ly^Y%vR~k1$m6XHqQc9{z0Z z{~x!Vs*0AjXhQT@Rf-z1g`=f*tHU0xeTcpHXr(=7joLyLb(E;RCDjHs5;I~(DKQ%) zMy$L2-oN|)3$DlGx*m_$^?trTuh;AIQTQOh-aVL^FWp6LX zl@JIpIb7#(jr?y6)zflXcx7y(|@lxw| zNv7ic_{P>{_GQH0U3iYAKa}>BbLkJ@IUWDIMZYbjyv+y$!Kg9Z~!@p*N zb!i1XV9$J+6s6TibDMpu5#iRCFL~3Cy1AqNVNpL7u)270`ttVH7AF|gj7G{DP`@>gR8lg zL~BxOzr(Q{BPtsMjAcTA>X7F=B<+f?y5G&3=pjvYU5@CxXXGVg|2pk@bi70HSLhO} zbWnR+Zpx$9vjSi42ho{kxMfS^x+C1uBMuQT-Bz4U4c<3Z?_Eo}WCGvj@EeM)3@|KI zBuVmy=>j@ES+W&%7Y_NVbAk3Os73A6Ko?s~8ElrEQ2GH0)$hC{#y80j6^{T0d@J>x z{vo5hNRp|{;Go2z3cJ5$e-Tz$9}vWdTYJYv%%3UED>K?+Y!C(HhZU(Z3>Y6YnM1p- zg%rn;Fh~u=_le~}f>yuDMD6gciq<$?sCbFwK3pH-xzYE}fJdZcBcMv|XrmM;MU`|` z{ab8^_2}I=2nkaeu!bx(1}D3yAV))#vqWXT2y`xLXRCfQuRHz!x&VTB!3`yP=fwFD zZT~5ZV5b5Q?3m;X%i2`p&-3tKnpQpqkG}^*AEJ^pDNcu!=SkdNq6I!L!&k1%x&>E5 zX0chM&(cmOBU}LiueH3ZOm%@u=teP2mcZ>|0dT{QEU*7ebt&tP1yMDa%=eGk1>649 zN%c5w)X$>BAJL1-`N8Q7KrZGk z4(V|km;5f^D{+gtrj=ylU!0YV^sH4$P{XeoVL? z@Ur>`l44BOR6W~o?&b{Jxl|Mu_4bTX>l*b=q3Z|a&Prj8!ueUSJaRB9Ln>Ww7n{1- zb)*I9eET1&Th%Wf6j_r|s_Dd_2i3OS`!L`6glG{+T146cAam(mAsHjLUqQOKyBd2c zSlZ8McTLo6wif#{Y&5vkO~cC38!^1bo|nU&tM?=pcgOZt5;S&S$c5O|#%Fg7GnA5{2-e5eviS6&hlumm>d*A2s7=)wL{Q|9n>I9^ z?2UEFD&%u)yjgeMaX@If{QUbz`Jqa#%cdp=bHNL>xe=274$^~#UFn}go_<-)OHuu( zh>YSyo448hn3XUjH{mWc+W6kSpu-`9lE(0qldIOC}a3PQZe28ojki zyUusN@ePvHaWEg=i<|WAf>C0*F6r5+>E-R2{~%@ianCfgaG{JDD(aP1Fm)qwN{Cq+ zls(M7^J6&AmQc*l&MVIF{^5O?d&iNK!|sOtdDs*mMvn1e)3biy#jbjHnXr*W zU0FRXguo@>@yb>>;BuU_Cs<9)hY1z0rOnGYaXM!`GhK8%y9%^Y1*7%y!$hY%cm7by zm7JY=Q~Je*EZm%c=fye$Ptjw@rG2}`U!{V?hEV=oG}c zP)pS|HCtPgi`9dYbeu zURp+kUH8j#TYb0p{usvWXeMzrG@+z@opw!15>qyRVzysM!wc^dPG+< zBy2?{=@oW2Mt~x&>@%>v z7u#-5=n!O)v0I#7;TKgMW39EV>OuPY^!YSAOGknGgH`nU2y;T%-rv!Kr5k2)4_8Asi5~ED zecZFQJ5`UNVyc;DVIPcY7IO{#)x`Y_#DV{e1$*^wKv_%8&dGm?6DEcNHe{auQ`H^j z*w@vX6Znm{kca<-K1t^+);KyEKpfx{OJFs9>bY5h}F<8TW=md zz=Zd435U#PyUGishH0N$^s0wd_{yCBtBw)zqc!;6%6)$S`_HiP&X=s_;s>D03{45B zj@|EfQNUaS`r;TkjghZE*6RoJ$vQr%ds%C;%!wn#cgT24+$`=Uf=gq&)KH@k0P+gK zTaHSX+_V^C)wHR~{s*F@cQ$TLR4fub*m*MZ_Iat6aK~iKSC@2@M&F-}?dR^=cSETv z!06h{>5d0|USY$>R(Sf(r)HN=!v4~|l63q;7`Tss{6DR@^+zkXUv)nGpH|qg`%Ec^ z7ThrrcTMHN`esbDW~(@VhmIB;wr{s>6sY~`@b738*V>BKB8wlE+iP=#u&{jai5Ij>aBqK!X z`b22{0)fE9?0IYEw$Rc!%fy6i|#m@j?)RGOkbDctlk2IY3BfxcycPCPL;T~N=U_NZ6UiR-F z*b+y>;-i>hL*_{#J}ik}Goa}#J&|5xJx08|Apn-0VsFCY7)oCyRY%tOpisPAQhGe}@-UerM%PIJl&im%g_W}V z$23%LZ)8L|1UD{PV-exG9W))D6SJmdlL_dJpy)5m37_c~9u91SKn2*R2|XIBWg)>D zgjT)MHu&i*?=U|KTIq|)G2B|Leq#O~DxAO#Ucq)YmU?#0dSnb4w)CRskk}F9B{Hw& z4P?=%D|{XEFl=~)8K-`*VwUe|J9?)q+-I<~wRo@E+uv&|2SOkf=h-A35udOAgt^6J~Bcaoc(fX^C~ z*fdqt{Z($Kp15n0dLJ*5WX1PW>aq=v?gDu-frtpk6;WpUJ(a0=knswSx@6oj$aumO55( zW&r1o;I_1=aO&-TTGzWgg;EY#$#wO|M@%nNKvu3K?@jy4EZMRnj+AcO~m-fSk%jNsPZ=4?|qeTzfO37MAKbxj3y~$r)=^B4idpTa>YFSg&{Oc9IN( zM?uVIhGyi?pRV}UUiFo`d2H&GpK+TtYqgD5wK&2ZAJs!wUhvlq-cz?*x617!s(P|x zR6{mj(LUK>QQ?P94RNBc5x66*hqhhUrfEDUF42_IW%CktI7!?iD!b=F07>+iv;m{4 zP#X4~&x5TVW+VINXQPg{919vRN#}HtZW?NWVaZ0Mn0yohdOG+uI7hiv0u zLbI|ql}e=wvyRe|Ps4%AHjqrTgvY`UV#=GM^>)~N3xCup2$dAI>u+poJ62Y+CLcA% zNB|NWaWF%-q}3Wl=d(=ivufwPQC^RtcH#xvN|P&e&M2w~M#}T{cau&o4Xzmj$;oD8 z`>Czjo;Lsp&rhm8>H5yQT1p}G4>FBwhPunz*ZVo`sF{m@XLnI;B$jYDGE_r2xZ08c zhH;V|+Y{yhV8}>?fU#!`O~U$M@-wrxPttQoFr{Nj;1}OMzgI+>d>i$qT0_j1;4F~k z$W-9MI?qrDYoeX<$KCWePhj#nhAT>TnwU!w2@VAA2(V2X?kawYJoz!w*f<+2)Qb7( zAk5dF1Qww+YFO+QpEMB;Gy3M#TYHtd49Ll*RPw5z$&!K7rHUXp>47|ar0iD#ZOq=j z)+2s~55B+N#Ux_CKUQ)XmVRp9^o%Bs zeSnTBW#Q0_z}1a-=l9`q$as?kO1wDhj0PV`Jl}`U-rE0)=F)fU_eM^ZcG&a;flEEa zS^s<`;>qS2`i^4IEcmF@%LMaOY2$S`+~Jj>?IB)d%`qieH2h@4(RhRzuouA?6{EUm z8jKmbp8w0qe?D|IIkLY*2Qf^X+0t>^07tZ-p@>DuJ#a}6%?5QIVZU%*&QgDMbNx%j zS#cY7V)Af3Yo=+t5q^w=PnW@w0FQH8^WF2xg>KzXRsXO3azo4K9k$Kxf4P(?eW8kB z>nx@o=o2uf{Utw=XYnj<7~m>fA$M~H?RU`-%-foLxgxWkU0Uz4MlYT(-wR|!Ou#1? z$)Z<=%8xB^VUsGRQM(zbqi^?G%kIopsqcxVo#7SxH3SThX*yKX=JpA`oJ%>-pH6UoZNiU@4~AdNaI3=``4r<*R>vqCOJ&}*_l6q#5p?-O&+!3 z^6SIfhw8LMZEM{V0C1)sU9~%phnc)QOPW7+21;JIV`0nCs|<+j`tlrw(ScY2A)4w; zcbKFD1UtJfQ1`6PykD~*&(B2W`O{pj<`E5+2W7L|9zJSJeB{yS2&A8r5 zJzgI$wU&QIeN}^p(zRPd<)ZVmnsl!&QK~KtLBG)XM zCn3khsmvm47BUZ<^Bay0e#%524O|$ua&&s%5G4>6;P#WmoH}WT;ExM-d$O8*>K_a{ zt)d#$4Fl69+eq2T*AA~qQDLU6v!@HA=}68%w^T@-OFi}Y2XD9X=FMX#b#)+M>Yi*} zsT1*p#4)$09!tClR2!*(43VW^<|RgAOizb8rE+AH=agH}3QhOC$gym!o};PZU$ zZ0L#2Ts5W}ZpX7XA5V{GySBwudQ05syIi>4c+jEP1$F%A@%r%|@;m~}cF~~6-Q*zH zeopOU@xbIwOG|t2&;e_#)#G2XJz>X#uA1^i$x(CyZbG$K4tliMZ=>p?fNgulE<5fxnExjZMX&h@sAN!F#S#O#b&>=t~n_T#%Iuf_qyWPI7W(eq5FzTA@a>TdeN9t*ov(?5I^v) zO4v1`J=F#0LH}VMHQ0@MkgN#9K3LTDUn2QcOF^E~8tzbH0TCsUlhv=GTxK2LP`&|k zRs~4Tl?M;sR(z&1uM8)GIO`#3Jc){lbwMS;alSVo_Ox5crk z{t(r3#N2Biojw3}tp8V~3kfp=f!%VsL`Ov8E5;8N%X3XTerCKxCe;aq{N(|S7i`w0 z^z=UPqI)4*sl2?uw@M{HXpOzTN7oT&((gW?2w4;X_np>R=>%3;os9guuL& zs-obDf>wD|;URE>S^MNt~ieZ|YSKhmpWbB_}JpA|oWBapIkBOSJ7kJ7H)aFx}czsDU-^y`BXl>@OReH8M zi_XW4y2^?9I6kRNkkq&pyBze_?gKzle@YH=#inppeMDQoFlt4NZ@q!z<6#~`px&3o zkYW(sCTbx)wONbT9WJ*_GfFdz`7Sn&wem?8Xc*ZpyRiQ`rcO)qaMO=bPg@(KsV(pE zs^$iBGdcyj1eS{Xm8MgS!4@T7>$`sIPK#SdjFsPIC-l;vO7$xw-@Zrhk3--++ijOp zJnHm|$in87q-%VGE5~o{GQK}@y>9Ke22k!64KgV#}@dqok&ud3&|_!g5*mKU7;^$REozJ0hm+2(s$ zyW&WCGBeOi<2}OmHH1>9w2qE5gvhX6c;T)zuL`gZO6`rasp(JO#A~OlPjNJ1<$=<; z+X}Tj-1LbJ>(0jhgwHy0=W))Ph|f%&Os$7s><>@WcPn70Ya_Tez3!L6gt&iK5-&}G zpnv5}eRE;}giYOX97)AG#p^~cvI?v4m2yIA>b2eQ_9>=;ZX&{nE6e@T?=!p&N9s>od3Ur=6Xh z{?e_>={p5f_0p={WTkm!>V&e)Rn^OJ)_=kLox5m(obN^#U)Rw}V8riqkBn-ruy_~v zx=6JzaD3J*9cf=sZ0qR9yZ~$&+e~G_B|qI}pGj6T$Wy0~a@9W^O2IhL-5S zYSB5X`nLWF(jIY;o2434Zg)*|xelEe1^IwcV6Ex)k-%MX;o2ebkV$~$&izjkw5~yy zcIvBzGZ|0}q@ze8FU8^2I3zz-3#F?3`<(WwkXL=?x{1tNIBn3^5+A|3_F2 zFm{9TI>}aq7A+ZuN4!O8PS$$(-O1%TbvZ^2&*GM70O2_UyFW%A>FKyK1xU4880M_b z_(*M1xU)nL3yODD*wyR!!|OQNHQ-{U`4k;nklhWMMt=8#gKx zq%GoKBPinYf?v`m0`*CcgkuAR4nO98|JL_Mjm{023OG#$xuOqycCy$i3hNiO(wMQK zV|qXCJpl6(RH{pN`;c_L$3-*ASt_wu!?@djF4xjK>w%ofLOTm8cst4WmppO`9)N(r zv=4)o?5l1T(SnNHwc4bRv1@xnjGE8hdjurJ59O|Awxfz_D%EZgKTI!ArHzKluJ|q% z(Tv&ynGr+(iX7G&htw8j!WE_&JdcMDu;`oKW1kPF_SHbYxY7xqmt}DknoeHZqHuW2 ztP(HlGinx@P2(Xhf&Ez?BwsZa*pvc1Ulxb)=ac{3;;N_gis>nncw82=H_B_Ee3~%L zG|nX6cx0*Bz&?M~>3R|Z0A^37lo+D{aG;CNSUQ(Deeyy)|JHq#+IinWF3J0=( z86OYq=)+Wg-3U5kdrw2*O=l@Azf#^?|EcZQFh(5n?w!Oua$dm!J z9_0^nlbXV+7f`G0PyhMU|EHbaQF)4bV1fIK5l&B%ro*d8-8b+h?KDBu;hRZM&Zwnp ze-q;1lPR8^YF~S^jZElH%}HxVgxKOBKi|vd2IX!^Y#uxUlY+;K9x(giquW}r78atw z13}dQHwt5RxM*noUF6|c@dQ(VuDYsvAKL})(r>1GOTCRdH6zZ+RsJ#$H(6ij28G~1 zeUMy4SA2H*n1otSH|$!SjEZ3vY};ASRWZQIxS(QfNZwe*n|IAf^eURhkO0 zu2n@if`#5zJ9!P})1KSDzc#d1vvH@B9)r4vSQWNeA8c-&d?y{B2&*NHVRc)wsY>%V z#!}23H3&V-HNNEW1aHl%0%Hu1&1B-4!!9o=$bk)F6TzVBA@G0>=dnW9JzFeZz>*Vz zvRPEKgvajE=${;=S7fHqyWfPOC{v6=ky^ZYpLjzKwS3}q&WCcN4xgnWCUpP_D*|4% zv{y6qzw%G~=;gK&0LOvj2&A10N0JV&Ra|N6hQSi1h>+E`&$QAXrsdji@kT1n+Pb!l zV3I=PX#Tv$5#p(2Il`!F5VUz+HFKpdjQ&2docUA0g8F;&H-dX>tt)vESw-FJnMHvs z35AJWr*i{xE0Z#RQ{dGes=jMw*+3+28oJ;!dPnZN7in(RJ^t9psJ|pM5#f~5%q`%H zLMb#)AUzG79MAeu_YziQZS~&;5NyQkPib#~UsN_K6%8uLsUA3(|N5(&LAjci1Kg_B z*Y=zR+O}Em$#`3olATvwz#&CP_d$@?nx93v6yGMQ4~%PQj=x2)paStH6j%c!v+w$) zjC)okj{mTS5%C3I2vs(z^>uw}B7nFD$ss$N_xQ5h#>2)W4k5f)psc@WaPl zNi(A00kF(CHn|-6`SJJp3$G5=t5Aa4H@$P&c){*iYum$TN8~gH+qZf42h`f3R_L&T zZH^Lo8vYj<1yxy9TU7yoHUs3;NyGXYt{MU#dconeI=aKuEpnn@2)sDn!?vV)C|~GZ z)8Whdu9_+??|djID=S5YvChzIDD_HzbDdUxv9AAf;Lux!Hy+x`VJ^%w1yj_EH|p#I z`gJEd$gab13i_lMcIG?G%%$#=axOe&%2fyNa+4cgF}-XnJ3zP`p!d*-G#86(UdQNX zW8ln{kV0;o2KKGw2yK3le?y#%>@o=+&6~FK1#rAhImS!J7^IHCZP&PV0z#r0uf4`6FgHug&{>Y*3t>kjkJ znKZHuN7h?oAkd{h)o>QQz^m50XR8%5Y;$AQ9ybcBH7NMRXR_!F+N0I#h`QQIq6nj0 zOJgj*lp|UOeO1%3|Mqb?m-VCDQ`=P!)awnMJ}s0KR$m_9+l$`Duq^C-Y9+NFA>$9` z!+%k>{y~hLYtGHEMs&uI(lD5u;=6h1F$Ap7E0$fxvX&z*wJR#_u}0L+&`KYLFE!ec zVv3|@Phc>^{3wA-5|(SfZfF2?1|9uS5DTXoE^FsYpo+|q*qV^yjDXAbn}tEKkITXo zD%5=N@kQ*^L9pZq&aOVANFZEIWz@Yx3AVG?^-_=DrUNaCx%u{u)1da!4|~gSKaPC{ zA2iZovRiuSW|$%M3XsZIk(@Nxqe!rkk1E>!1Hi?g1?s5jRR3>uwa$H@&`zZ z(pyUz>3C;O{c6@dSGzK|8gLfiWMqCiFc}c$XooRotL8cpeS8cdoUlb$ zsjcW0hUzBzu89;fWt&vGZk)iBHj?3T6nD=@qVht zg-(U$L+*$qSv0Vn4|Zf@@b{TJGCKe55=aNF0^$>g&|89d8AR0=BJ@(c zcWa+fmSQq&4Z5wF-TpFkoKhe-SXmd*uo>O7N1&rdvGi`ukO{q}Akm-}7QsAn$BqH@ z$9;)hUiR7Dk5i{N)hriU$5T?X-mwhpvVUU6)ca2I;{GUFZFc@&8v|2i(k+%ELzqdD(65-S|~sk%Pys++eU%M_Ic%g;P)s=V%? zr$m(4`R7vD&m#Hnt=q+oH%zP4J4GF$tfN;nPRpa~!ijmZtMqa`dj4^j)S_WQ`|*z~ z!Lp1=kbud#V2T3EENdx#DfhW;9MK~vK!|1P9}A0K zzYCh#L;>_`{^l)fku)*)UxrMwtiM=~h_@QlKB4oY$c%cFhi`NnFve%={RsuhZvc^F zH+`mFY&liQSe8kCH*2~Kj!cs1VRi9ON|{bCV<->9`2-GuZX}6wHRXO9#|SUI@?-_q zXSj`)*F=#4_6&layD!99#n*(?=JMRTvalk;K4WiIE}@)0WE$=%hKV-ZYY(T02GZ7| z6gJNWeQcDv!u)2%8uZEUsIMRj>Wz#<>WB4BaM zeK~dI@b`4O9x@rirq;5nPJ0zDQ-#hd3J~Dj{uxpYz9)E;k<_Of?)S|5)42Lj?I)hd zwJVXuyxW*tLPm)3GM;Ihd`G%NrwaHR&qB|+WwT)4u>aG3nV1ECP}FV(G62UX_riRX zKXN%B!?-v;0L83bcaNhpxX)ldJUpBk*GRBaTY1e{Tkaf*S>8+I39~Xivi0QbLN}C? zVbNIiJV(E@3DCggh9S@?j%#B`(T;Sp_wYHPp`;=Rn%Y6R*raIxU9)zvW+-*EfR8Xs z_LdSX3+SgOu>~}}8>vZkGIGQK6PifH3pOsM_*+ZNFzmqegv43~v?+LIG9o5bY*$3w z|NT8=n4c2Y${;wG`GiyzY6h|Qd;)z0>*yn+<5dGn3XvXNoi9J7B(2q;SWiu@T1JFf zdll`_f0r)P15DlC;s4Fvo>+7ohz~mL(+8=1^r&fkTejR5whO?W5XxC}&Z+JS+h=@| zWn9<|kjC_w{^0~pcDAZ{<;6v94ZgoYq^GD_Zz-5%Vxh z9VQTjK4#jlp@~28dAt#4v?wH_-glVQFI;$)r#`y$7;C|S#fR3VsVDAyt+>l8*M@0` z`cj$+0%U9MOw3lMx>82obAet+&06QhYh+^$b|!KR;2N$rs0I-qsQ#aGLgV_*kpIOh zZ?gOFDc|jSafL#IIe12SbKyuL+=63XzZ@r=|k_+XPVYIwON8p3l)g zNAvs$mx{hpk;uH9Y)-UZLI4GBfF7C%cR(HB|Dn^+LyR}B0Zz0XPr2Oeo>sK7cq%MI z|63z!JB_Kdy@y5C`99!cwf9tIO@=?AewG+XP4S(DyNp#rZFPc%(#0j@;1_6FPfZu9=OWzFwMvT31rV? zPrs_{F7N8qSpwHtOLn7|ZT)c{;EgdiK(6+{AR)KMmxk~2+A2VEh_D8tdIj@5sT{Z2 z$PMojqn1~5{@2A+PClDI=u(Cc!1F`WU0fzrY6py^8ktYx^1N8yz`E#B0fopD~x_btI}~+Klb1g8?r{&OI3J zhY7At0|EAz^N{hn8*|TJ6=&<*RUH=mutX&)P96Ro9hKchQXB8h(CQY->M@N>$~W}E zxZCm;=-T^vbFvT8RYRi>f?)ZJ{1VoQOuJFL#>LmAeK0RMOpEfJzNF+bpF3W)JDdW+ zBPG_(-_^yEp)h~E()^D0JCaf`nSo2{upeU~A($jClY1(sgssfpDHKGPm767jtUkK&%dmu zyky9Q?P`~9^0cu)a>uSXLo+`akV54i)*=ajbgML07+~tpcZfX!RcO6abuDFt(uzSaR1X!r?>MGb*77;S1WL0l%}zQh@OjwCj_awlKbEtDSu*(-akSmvHlTk{(Ci+(@kUzx z{Xoc1ZHeFbLgE(>w@YiZulBjaM)@kP>3dk4Z69@HkxykB+Iiz8mqoew>$4B!9sVJ+>||Ku^i=1c|5!lf74~A`t6t8HFhzZJsu;q7r8gpoy6a(4+ zrFY1}C!pNYJ64Z8-MlFrpfJow#mp&&k;e2u>T{_k8#sPZ#y!`lH5QGMfl0401$oD! zp;j8PVC-=qHFnsw*EYQJW8>SJ!8XO7GgAU_Fr}=>12GDa?7bvl=sWP0O9)-yXfz-cKewxNWJZ1C5PeyF* zwSFW0OSI55nkiqZCtkM}U%~a8xvNo7nO{{()&Gk(el=d z4<)TK&3+Ez2EAzk`2}REwqxA21^W9tD%J?Hknr(e9{+lc-1I}&+rxL5xCLdp4u6cS zP^L_S=~g)09aXpJnVI>ZaXkv{8vpzLlMO&BP2c`VfvUfM{$ue+VjAqm+GSPZ#U-c6 z{!Bqpdrj7EEDBi6B}jLlQ|tH;jR=wnNpIKLamv)R@YtI+g0AB=HWC7V2wn>4^Sl(1 zwceL>LUE6>as7q9<&2=ba+PDWhQfSGcxs|$r~EwbgOsuFFc}T~Zu;{h zWvD2h`#pRO$2`AC<%wDA4NCKTC#Aedr{BDHtHXk0GDbySiQe#QC7YknmMo~e4F8vn zR&<91xCE}@z~qG0a(fS4EafY!NkYoKJk!;d6w*J1X~+UF zSo4^e^A6ZeddoU`CF+Qrx;KYK<2L;zol||N7Bn>Dk!q9B+-`=Q; zo;^zg*Ygd|oo;A?-f`<)w2}3{jCto5*5;kWEl~HQ+qeEtLPIun?VSUk?ebK%rxck@ zuf5eo+;#54@TEj%X|3;{oAm>QE&x~aw{NTL`VA}Pka}{9Vwr1RYJ!^GKHp{FxYVCR zW)kQ74!?+P670x2vt?Qy#-?$$<|57HytXAL#zYuZI|&+~TeaCWm}*Yr5pSI#uej0Fx2T?OFPxjfKXp{_3XXhtbY&Y$$1hxmAZR#792Z7P4e-+Le2( z^u`vWZcb@82@0%%r#1ya|Cv$-MNeP$G`@r`t?}_I@H&u6HDMW9R$&Y# zQI$a9XEuHF$;pwztFewx1RgHltalXM!VXHbEcSxH+_vM}U9Iwf^=e+#pN0y!dpMQc zF5BNj+ESGt!cZo(-!S0}_8TuCCt)dccEfW5P>NXcKBfVAt#O{5gBl%S>W+unFK3u` z;M4;&_LTPA1^d`jMI(QvpUq|5%>7a^-0@?fGd|FM6kqN1D|AithDTleqlm3vlLGla zcy0SIC9}@iwoo88y-=nW-53N;67%pFWy~}3-a7ge?I%a@;gw3L%H)*gJ#QgKi*%h*Xa*`w|mSNCX0KzoH`grJNtNl%SfJd-yw z&9E_SB~S3I4fU{pUabq6UY@+~ajkIWb*Ci%SOf;4c_6CZw^nernA^K%Z?Ev(O(+{% z<(MP3PVB zNtU8&rQ<{(+Z;&T+O??UHD7(NSY(Cx&PA+cfNczkhrdPVx0>DOwgAMrjyTCY%l;FmGh!-W8Ggl(qJ%re z0lh7{4my7cXkabPPtIl|gL$)P>FXQrK1%tN#=)#S@Lc%; zDtvAOt1BSLZew!2qJr~chYECY=3VmKJGV+yQ`WFLyTSn#AxJzK<5KOd{P}Q(GF^bb zmVXv-tvTSo4|cGyM|ErIW7iPU&?=^f0h#i@T<51BDw#Y)>sBWxCZ`I$G=8XXEXSb% z{V{695#lr4zyk87e_!)Hy}i-?J-U;JL+Xar1Z>jcGXAbS^ z&|3K!)sA64gfVzyjng}%$D#+gwgVT)q_9?~dYGh5E7CPK>SPQ&sIzl8J?Ojzg?LFb z6ZJAVSQcSYS2u6NVb!Y3Jhgt&Mq(*{YiF=gY{&VMH4�rh7x?c-P=zHiW$*Y9*%Q z+uTvILWa__dgQ>SE)kUx%c4}fGi@;)nzCTK^lH^1kQJbQ zqJ@wYNYXSv{Q6jx!*q}L6u<9}Jf1}p*raU3|J?XU(19FJ^_pzd9%Tq98`V~B3R!~_ z_A|syfkpYjLmMl`Q1L|s%7L~dPqO&CryxsDzMvft|0%9gKvSYeO_ox;pH*TiU>*2$ z+fV`#N~)L79|W>=-Xq2qBP~Q0oZbXS*?!}=|BXM~a8?LN;vlMY`nef!4HAIQ(axot zAe86i3`^>Y-LqnIv%_}!3$elkK#LDrhn_69^KZ|ywiVaWZ)QK4KEZPoGI-K@u@$o>9PMmjrA zyLlTgC&KO7+X!sxKBMh-6y|r|3Jh!_!6G4h&m9ccXZKi7%A=PYAn1anR~&mN3*G35 zE9Vpqa{8g60zbgsNyg;8sX%BN*W5Fg;ok72I(>8JL$S3~apL2y%Mg+mxP_!2#BjdL z;CIoa7QJD2svTy&Kx^|Ob$*YfX;C_F6SHcvm?Z(jxy(!~+6g4xsNH^XJ+0+p6UvWj zYwdrf7_RtSFcZPNJ{z?uw+VNSyX;+Ob&q(pp4=eR__0B&o!ZrF9iNDVuSnXgK&PwMzK~*GA<^b&!xZY1XC`_q*z4xsSq@mgU&qtgzpkY< zKeuI_tD)vZjq*gUxpT0Xt4=9jw_PlmeLwRTs}Y4|f>c?bQ_KF(mFGJ;tBuVuwfv$K zM8*(^gC*oWhskuGDCLXH>8twVB9C^vg@)5yrI1@=SLB*ne;SzE{5<>wIZLmYeUTL* zLR&8IE{UecRTdP$t%#reo^D5%l2P*sVucj{2x`pvo|ru*v@Cw>U#}2ip95gC{9E8sg1zh)oFaHGn`%AwY9N;S^`@3)Q5W9#O~jaaP-|fw ztgHT>Ju#M3tk&~&dHZxh;_%D(A-AY_V50R@v844!&O@>R!j^dDG+1a^Yj&j(IeUK< z06WAy=YZxLkZ-s4g3dT>pwE7|?rLBa^*$@YcMMk$gRbioW)w*Tp{6?C?>2%8y)f$| zTfvVjuMQm=0Jz0uL`~m7s>AJ68Dt469~^#(zvDFYSu8g7FN#X9`n<@IGPvD1aJG17 z@qXo*2BF-QHe%B~{ZSHRy<6b5vu3t6*MqBvQbxIc+3dZ7&&S>uq?7oO`qwF1_L}3m zH!VmnQ^S^SWiS4)gbvR_A`WBB1uXezY#cn>2BfL-QOeZW7N1+ORkjUzrKI=pmPxzN z!@vpNOD$=gP5qCL-D36zL6M&2VeA9iDXqii>r^6m)m6LNaEkl5?N9!-gG_R-Cnkkn ztH#n$TvJHJ*V*3W=YZ&di3s|i3UIQUx-RU-%GKz$-^4`C{88z;wUoO$UExh-+;ThH zus>Hlc^SwHT7-ZvbA zw4b-iy$0!0@nv{K%MY2{w;EO=P07JsTH|*6S!E$T*|7mB-9CJk)x1Lg2BKxfKqWl% z8wf8_M%7usU#DZ(n|IHLQ~ymRahSqKZq*CRV$Agb)S+_;+^-#dQbtyjj}U+mU?-q1 z5)wN^;P3r^XgU{vCfxu1=eRVcQVwm5gya~Svypd6PL*opv>c+$SsG?eAvs0}Ii*s` z`7GvqK2wa&p^ACQ^$ah8w@r5td+6E4vpqv+W;>kZbvdAt@I$y{xL+r8Wp>tuvd3 z3a`Hpo+E!S3}{+C6@J`&IBMZ0SUMrS2ML@jgsTP@oZL4dUBI?vxJtG}ne);NOh{!? zlioK5-f9xHz{n9s>M?=97tq@;Xk8*xNo>6l@rjRSDGIZuhr+NB+yN!^Igf=|Zv0U_2G^i1eK`tV=doASNc8|F66!&xjb z9({WB*^a(x>q~Pi4lZ+-L6`Y<&Qx0~-HYYJz#3W{=Q)0zn*1>FLQi&c=-P;LZsU}|IZO*ytNk{3c+jZRlkovfAryuWGJMpxn{ofUS-zoR! zXX}O;zbg3(5#)WRVAQ*<_?D2NX8i59lmkRME6hN~6CWC?R}4iHf`hZ#d~3gOsSQ)8 zd_syvn4v4_0CERkXA=+xD68TGfu+!kEb& zzxrYo?fyk038IFez22eV>1$buI1-{Zzc*YUYt=w!0M;>H0LvT@yXapGtZWw&^rkEo z!kU*aSd$bqm-tYtI{?syK7r-4HmW6T$JwnygzoUhQ$?z=)45?AEnzn9e*oC?7R|aY zrRNZsAm_aTYkXbTMNoT6Oi_JrReX~2`~f$v>uHu8r*nSv*3NmRz@_XajjnzEh&ZPO zoJY)ah)VM`BkCD~M#{AuHCC9>`L2eLci8tV{cdDe`_-;JQm_3n!qWuNGN9<&N#BV? z<{Al_#Rx!@!s7Pqt2ARNgK7&urRzeULmVnzZe&K2}= z$M9@E)aI`(T#gNSU7LJ-xCQm=Ifh&r)~^JL*Fl~fT;tGmiV81`;SSuLjMS(|<*B7; z`@h@;E>ArcP>oN@;;Z!N|4L)CnesWy?pC>Tv=tag;{lHz2zy=2ocrQ?F(aWQX`#j} zRH)bp5K>6{e*7s)_R_!!eXPOw&~C}S#mx&Icrk-<5#Gc1=9N8s;{{47{!5?dnz~Cp zRnq6&{ybj>lZl2HuVR0%l;{2ahPjx0er7jR>xkWTEun=QGo6fU;|dU(75Wp{tsLPf zz0GbBs3IAKU_8>-tPRZvS)rScsiw)?Z(^yR@;#3lzh@n`!$4Z``rb9ydPXfPicGfM z#q5>}@%(11?|(Hj)u~FlK@eS!EZpx>wI5g)f(&=p%>+NRjbwiDg`i2fv77NrTiE?y zBZn-@U(;P{Ndf5&)5}q_p`h1|8mp19a}xL*yr%iDuTkA3Db1w;atU8NX#Jv6bb^|0ZA8xG<}y*Y)n`7aWlC8H}n3^{L&-`2DC_MD}xS&S>*+m9xm!C6O9vv+(LhSzjLaqQsxaHDP|oKXBf6VrR=-2Sxp! zwiuTdFC7wDp&XfQ!UBi0R|~zDD<@sSKrO#Ku7l7mA#3;T#HWnK{34thZ!~XA7X5c# zICSsgsWigP2y2D)lFF=~|P-NqsSDr_gr&aFp@1DO7$ zXWPSt(wd6nzl_nex8^RmIpWkchHhI_Q+HmIpNlpJbkPzN-1qYHH#CeL)^?j0g1ETV zZjMslXeA9BdEAc7>?)HT`SN$rsx=Hl;Vl2Y8)KXn1jI-I0kU@nHfe7P3+ff~#^52VH*c8u0fac z@@YagTdN&DeIm3G@#;##w#l3$ruw^-b7)}Z?yf=*pG9jZSzN+paI^R}#&Kl&WI|^) zBDtut*A8?wngx2IczOo&hUGu(cMqDjXOpsnyDUkn>HaXtmEwItHjwb6|53N&K=~us zrH^#e246@(^pMJGV>q0DPQbWHH$S?8-Y5xdzvFzp@|<xsrSm-tNjVnBX=zTM;0 zVT3;_+DgTH!UD&>v-{D=6t3yalN1}UqCP8-UgX`MHc!gS60#dK-Lw1YhTTcejmBEP z{PRdxukQ=Z8RTFBnz&1RKNu9ZRAVhTu}F(C4p>6}FjROPFlf}@{d=-bjuN&G3Fxwp z<{01^Nr{A7O(VAnn{0}#3!(rpDeQpM&wG{L@fa7Oh8=s8^C`@Z7b$~OqF(Ni! zY7f+ZA1BVIakbCWU2Zfvz(xES@4!wJ<3_`+6TGjS;ssGsKG)G$=GIy6;&*+|6hXf& z1%IqA%p&5?fu+M{F>MP1HZF-51%8OeCE!Y7LzOZ+Q~++Vv)#sbDTxwjiWjrk_GI@v zdi$Oq(D!P_ccs|eC%;d_bpg8f0);j!8*%Zm-afH7fRu1zx1ZA5CD82_bb$i$Lw2Et z6)7i`?L7PJ%mLG)hMGH-FeLy#uxKiMU@p|URIw0SF%p-xkoOZ%0Z5M<2}e2lhKWvG zd$pmJHn!FsywDwdViCR7Xh;2XFE;z9u`c7Us<}e47)*da7MCw|I?LJp zV0Yo^TAbE{F{JAa_0c%AD4LHp(A2OUik3nLNLjag^=8bKe-MT*48 z%Rq=>wdN8Uh(dTknOVj2r>2kNZO#^C5hmNuA5}_Txf+qqCSgWSZbCua8;dz?&pCSJ z%1stG*-~!jX@)wK0NE#pQ#3nK-?V4U+|Lm$?p3=b-nCNA{KNs`jpb#hS0K-rxGdx! zpT|tKNN9~>mB#6cRiF`?6aRMlbCq56CJVLse&_@v+g z=39ZYMhy4F)gt411qAw?gdgdB+XsKdcJaIA7aoNfT(W9ma>slLMf@~P_jxdOQwEHn zwuFCl1Y2OXo|*Sf8s-9dGum_18D(1kHjbVJppTc{?$KCKz~xV0D!?rCVcZgdF{_=@ zX;%2iqEv`RH`4qkJy3fB93^XU7OWdcWd%Y9?cNnkjR`qv|FVxB)=@BlJ6BoHT0pW{ z><8S)T;7w4{__^J8v_quX)>!l)>30*;@nvV%13dwbW&a>aXQjLF+-@IOf}(-DRRxU z(HIi~O_Ul{awVfi-fkt8hXJbbI!o(DHt@tYHx>J_O%^Dvu{iR0KjnBCR6IEH5m`6{ zHG=)DSA}A6mMhMv!d&R=C!hw03Hgk5g5S&T=sEMpMV#>Hjt>4w6C|E+S*nX4Q|g^=e@*heEqEhUad|DRts%AO-o-x$gYsCkPUApmTF?86>!_KJ zBQW@i+WzVg=~BZbU2}7bCHEp}fh#tp(sN+nMhSE`SuZOF|^3gVhQycyA$Tb z$47od&4PbEq0hmOUY>0&oE|yEvqv+1tq_L)`t?ga^T6!L0j_K~rfzvr>OD2((_e_> z-1K=1U7Z?GJ1^~hmvZYDx)7RHU=AtHnP$?$79%#wMeGJ}*3V7~35{|6{5@sYpFw}f zKKTtWHCI?LN!l^NuxHV#8jf6gB}Y+C^A??v+zRX;TebgkOjUWw*P%>q5m9V7{5Jyg zx!!+Gff8pH`H`#*b-v+#;6^{Cw?Vt^Lb#o5YowrbLRzR{g{!>jiGK9)o3udnd~meS zK)bw<^`~3@`T{de)8l_gxge9yr^|=EHXDoQ)ye^RtXPg-)wO)AtLfxX;m-IUw}pEL zepY|g_AiP407wBIgMLC({%Y?%{9J|+gopMFA75F0A)e!|Or~LZ2y0?-P}6+Ze{RFd z-`DbD@(JDZy&SqnK;+4B7Sk5vf9V3q$O6s@fuMrNo>sKV9fz ze0q`XHzC6e7H<%;%zJ7Glx)NkCErvh42=!(uC(>)wheVLBl6FBbv=9Q2}B~~$+t%N zgg~Bm+W}A`8uA^|4j=L(3ryMjMv>-Sh^X`|ZFyBZ$SSeXUtTS)>@qk#&U+7yy8oS1 z!-3b_{W`na-PMW}#m39Jb{z&Azatzm-%d{GX}Cr($SFhuV&J~dsl2_-7zjM}8bp=E zm=O1U>WsY8KM*UOrlhrgH@(iHr0t$TrJv$$hs}46oZJ)9#STx_&7Ao&D}G9&Yx;Kf zt#U3Cn@bctwX6LDwmnBzz315(GyDO{%ju5t`vKsM+GZ`3Y_rMUc>wJ=OOK>683`99L|m-c-Rf!Ix4pqq0F5U%vN0NuO#x++ru%XF}!Nmo=ZGO z2xL6XJ1^)6aM@J5ij<;enQWg}ZR<{~jV-I8>0sSMR91c090HSD&!Uj71-gv7Y(VAb zLcQGk-pnS_48axLV&X7+L97#-p^pS{Aa=P zf|!H1Ej@46S)kMy?%iK!3u*5-i9czH4GW06d&2D6e6wxh^}j=Hy2;7`8}~LRXOH*D z&tb)ao$vUdxte&Ddn|mAg@4~hw?N1L8yjtc;|BwAlht&}TxOJMXL_912LmBnfDI`p zxS#yp)>b7X2}QX1*juM&Z}c06+5E2e5ab>hYi`&(OOCW|pO3IVq6yQvttBU4W^9Ev zOc$=~(%AvyzwzSRpZKh`CoPt-a1Rqbhx%0oFbg{f5bmGiV;BkE1{7Ril1Yp^I!#?8XdP~FU;?YknbCK-s z_{X#o9qcsX7b;RFio^bT*f=3peUqPZ#!RfZnac%L$Qt;v{a!@>c}&v4aYJoY?L5m# zTb9|r#^HKd%UDZN5nkF^XeZ?QIF7#`(JQt@Wil=u&;1WE_}{Zt!is8UlR(z?ea2}> zX5jIG2(aYzY}2cQzzAEImaacS<9ZQA;@}Abd$5$!QQPA;$X{BogihO9_r)8OjgAOm z_s;Y>VhV30RzI@9y|-ByAQ%l0wdu%A>Zhyz*#Jx_jji9ir;bDUxKG7B7;f3PV3V}) zsPtYqVo`<^YOF%wpdlN-P@dsK3yS>yYNNRyJHYaL{vqALWby0Ft~>`9y*9bh;~S>F z5YiaDtL^?L^qPfYPw}bYGHw0y z*-k6?f!``Y=V)iBg$>7a4>I#86m~ssGz-_+79Z(cxJfYO{}COwLHqm=Bhj&h2akp7=+9$Ci!98kr)FSO?m}SFJ?eDY~P*Bok_ucJ*?A76AZ3e(oz^h7SO=x?|h1@cGCM0Sp z;ntOz=D)D-eU`28B=Ac;n@p|C-2-}yZTF|sZj|~qRy1MN4y92V&!Ud|)PvqWFy@tM zACM*YFmqzNyYS$;iJ^<>H2cMs+D=aH-_OUuHL&H*ct*JqH+g6v`X)48c zq@cgQ<$h|%V6&+7Xk&ND6p*u=i;bH*+eg1kj(g1bEyR^@eTXe$;+yW_=H}*AL`Phs z4!DjI7(%J|Xk_7@yX-PjHXT@4WNcI2B58GhMOARnB%{(( zh58fs*)U|%=Z7HuMb-53Rn3-^DVi%LTv;=I-W+4<*m7Z-YTz$WRPNEJm*=uM?B+*m z{3!Dwk65UHVec8Ir%DhSjY)B4jmuZ3_b+n_O2I~AXIhvzcZy>yOWqx9{&O7=a;;Z@ zC(b77Xh~&&@BQ|V-z|TEX60B@zrYt+ULh(Iq~Fi+sm@=-&ll`|G;7LXnK;V2Mv*-z z#~+bfxZ@Nmo%@y@R6S2Qm$eqrR)X43UU(CUC&D|Dw!nxGYsOGAs$5K>x@P}Rbj(YN zXoiTK)(^SN@<*;HxZKF|N!F4~)O40|Ws-v;dJEuBXmVbks;fiQ_EEo8PyiQO zA-|1(@9xQilhh2*h&g)R9Q~zXj#IWSy617dwv*wG<{{F+n#}X7_}pFyj?gSYw-hUv zOsRMTN~ub31&~X5uYC>jr+_ACXEbKbt=-p`zK@QI+#DnlkQ#o(FEKur^9l=L$wo-4%ZcBMP+xxst)4njXh}jfE%~{GvtVyHaERi!! z#O?O6Peq`{62zNps$RST;4Dtc<&29g&5?0puXYMjM96(=7eNzBufW{gZS7)8jR+`5 zCUtNMWl7%6a@f+EZnPU9J#)7j8QM5Jnh3i8*kmQzChe)@2uv$JK>r}^uM)QYGoXe0 ze=hw`7yUt}Pus&~?3kWYRyK@+zax8dva_{c%*$vii2%i~H0eb!!IJ8auG}e+q{g^m z?CJzo$+rH@o4$u?L4|=c_d%zBbS6wCd|~P|x?8xK5TV89$cD5Ey3+5)q#0)1 zn^(Gr5Jc*8-d*ogR4X-7@imr3zliQG4+YNrC=h5anJ-*pyK7pjOOas%|il9=|E6PgJ+r^tVyM@|Ib#V2KsdD-an60*0%RDn6llcX@#($m zxP4R}UT1IauHMMS>b1m=5aOkd+fophQhfGH;qj-6SSN1wnoB@AattS_;d&j3b95{D zL=eaw^_1?Gn*wSUv(C@c>mK!HsKg5mO6@Jk2V3LMK3@pImFfB2w9yz*9>AQNrf{;O zVQt$R>s{fT0R>#L1N3C0tIiOY#%s?-*|Ii5`oxK_kK@ezWcpvMyH4~Vbc_zNg!|w^ zp4#CQ9&Z_(-diSw%`U$>h(#$1;Z+u9U8GZ?z2>$`-joK~SGkYdR>aijF?whj(uhhf z@_@x!659cE?@MKzdt$Gz2|q2N)%sz(|Gauy1FGMDG^=OQ@xMg6zn?Xa z@{IS0k*+mb>yY9l;JBH%PO_#WInB+>(T}tG2FDU$%Y5|JZaw|YGI;HPjq1A3%VN_E z-*~M`SIL`YYF9zsKO-j z{^Rs8)}*}|?%+K@Lk^P`W3=iX2yvBRzw%T7dIHqc)=pQSSDAeNtJiD887N!#>oLw6 z0s0vEuUJm4E*mX?n5nw|R26;`Ur(|xansp2dsBE}fAYxt7EgIlc9+&yk`%`WRvc1; zMhF!plzIf%`YvqY4mS^r8c%&x?HkxRntE0r=noOr+dhA%>x(_H=2S^-i!~dH0Av>1 zZ_ifby%}=rKXz*Vx zZnwywG>OFZNu=K+UzAQ^xLry|7PgmHd&b*|vKMrt=2zWzs_la>-w3D3ydf+?6D>}e zAy2QzIO>QX_Y*~kK~-Vz+klos!h==u$SU58GRO+Y6y{e^p-1PjorKy;uhb9BdB7&f z@%0i(`;vL1hv2664ghT%Dn{-+CG^Gq&E_kqFWhtUPfBYta0!vOZe-2_Ro@jTZzbi< z-sCyL5k#DA43klx5)*F~FY6seI4SF3?&z@jF|JRJE5Gr89me<5L;c zLQ{}6k*ns#z*Z$Gtjyb_+jzLG-6N?fe}h=ecxsCIUee?!I>DdESWGmDjWOB@>-#+8 z9g$~rb8jvv?u5!5#-;0Z8s7X>uAVxq19)N&v)d8mLQX`_A5u;m9$$f!UpKKgMcoU; zbOZjgse4<5bdq|`b288Lfn46zxx9osy}(EP(|-H+GdyZBpT@n>*&9$Q>}7X{93p|) zmTdnDpm%wNnE1+OZRG<7R5D5YH3I7O-;L|9JM15RA!qo0{#j>}m1`!xmmTW;1sw$X z;M^)jm3Y$CrioQObzxGa@oOOLL07ENhp|>^bZ&j$qw7H|H7HX}ut(bpwZ9&g;D=7- z>AI^L5V13Jp1VY17la*8Ix#UXodZdn0H>KkYsNM94h90TG+q@rotSHfdFft`N9~Ma z(7#RT`f+$x%^A5i#CG$q9=Qh0_-7lQvkhbcoY3bLT(MHhju;jf=aN{oNa&v!FsaR;M{<7X`E{kw(GeKfu!lJ_otNKdPFNY+U(5Z8rA7+54j534I74~y zMfT{*+w#`9m_V194=?-pjM{1tc;7l2X{g3&iKuSZkcSFweR9D?*y8uT( zxWBgfON=X#x4;<8 zqi6(djC|{4@pRsK6?a)*Sfw+ujKl_32yx7 zX65O z(AvUZ`N3LF=_%k^Vb7)!w@r56&Gjme2YI-K-t~uZ1KhDY>9f%#f693FUL>|H$Hs_a z6HJOpwxYL}0jS~Ju(+z{9q{+4&6SAqIf(;br4p&*thlGn%MyFnnwYQ{<~qt9cri6` zUT=?hZ7vM5(XOCe;sFS9dc#eN(mIagzCI8ie9+7nW^GUT6LD}y^8_GAo(py(J`fjB zJh&aTEs#Y!hg&s3zKAke-Cv6kTEgUbEO}p1DWphfDHT?3?yo8C!(Z&v-ahjST4AVZ z*~7ixO_i5DbuLsid9$H8a%20l>0YulrAdSBt12@b&*l%?uZVo%YMKK2C2r@b2p!6| zf3wxDSg0BwKzs5!b3f#QrZV(!n=6xU>4SH$GlpeJhK3f!(%8ThMo7$z`BrS&&q|68 zAFdVbnb`T@QyfXsUqZKpO|?XQTf3;JWV{s3Wpeaiu8}8f%~?C3>%SV^G5gJYUCcg@ zU-zQT34Lr~#k`A4$bJjSJ8&SXHYW_T$T3+Id;HZ4#6;7F52R!a>(f7T)ui$@j&Nt% z%2M7{H!s)ZoBj5?Q9r_dgn^n7_4gNAkP{Y+;8)LaMW}ilNpv8L50zm<(vFAJUtO4u z{s3+RY_ZNBob@sNBtcr#A@69?heS50QbVR{Qz<7$U?%75JJ$mgzEl7y zoBx!h*5IS%^q2+m?U2w9Owt>FBu9*EZb0??=n1W(Lnl?& z@QA{S?K*Fbd99|=2Ow0a6_e6!PWL{1_uHuQSX>^3RqU!!|X zrZ4GzyzAJ-FN68O_qbJikE2|1KPQ-byQ~+(mp0P%uhDR>W)>#^`yNAk@^@Bu7TR_5 z^A{b{yB=u5U9Cp|^^C754ULqp$AR7pf=5WZzRC9Rvi}}qq^v&i(IS-#X@BZ{6l#a^ zgo?z_`em5gP1}0u#P9XB$VV&YFA7I0_pH*DhY{$X&zBmp5HoJp{e|u)mH)zvCCl18 zMf!ghrS6>5KKj=XFZeYVH0oKp26p1^pzt@R^Rn{i9tMfsIeU6-B}kk6FVYZ$>}Ix3 z7vDePhfb9hOcotIse}W_HiNInwxziT?4P6ty{ltO>YkF~p`M{Oo-3DnWfOnxo44Fm zo}ee62W5htqXeTkc1w>5e5WeSI zezG>VhO{Nh+Wbnkfr4!2gp)^`9vHRG3o2~C&iqi9@q@fTC?P$o6;<@_xzJ4u7A68Hd2 zu*pj_08yLt5i;>E@)it3`F*pOx&PS*dtFYr4BZ|xlpL+x^EK!eSYTd4mAd<-&B87* z<$6=AE)X{ZX*=|`^rwro1X#yAC=cm#59&9NQu+c^h-=R7ehc#9xT?X20FVZWw_d`c zS`uJkK@V02Y1-2c)tCyXu;b>U*HznTu}}cBZc z%9{U$qDA8fO1SHwpm&aq0QgfL>Ai18Z#qveiT>^-K$}i9sHDD=y+gAr%mKh|k#fyM zg+lBsw=VhKsA9U{3f$N1Su#b@Z9Q(I3{0gp*uw?| zwT8^)q^++?VjELjI}}TK*7RwMzWufr11e06L97QCYK6sj19{bZ$!wh1Z9WjIVA}Q& z=1m}5a0BjPyFYIEU?nSXcIG_zFpP;VeC*X2IzZK#jS%v4B1YimA9~0@*-g+?g^L1Z z5?XEbu(l+OTGm?Q150BQW9f9Ng3WWtbxzF-S<7)lA90c+e3?X5!{I8WSwyAv><6j- zFz&@I`54s3MKTX2Le8{-)+QIrpDH~Y`QI4+RQ=_S!p%96ml++oiaVCgj*<(yBn1nK zn1JR>+DKZ=>^#tas9)=?uL+~q=l9S{J6NIAxV!Dh`sg@AoN=J|vim@(a-mHN*VprX zeJDy9RP(d1aV*(op`^&8?G{GbOo%Q;}^t7O7$EyX`h(-tVS@-?w z#&G3pYQ-zg_DF08W_Hb{Om-h=ACbcNkW*4CiFhsyIoh?cLvF-LR7%1G$K%#?TirZC z-H*t8iQt%7VH_4!Yu_~X)|YZd$caTc<8D>g%?CC@Y(3D8s+R^swai4Ll z5%eK|uaOjBZ=umoxEhKWE&XS-#A*Uq*~EAX^dAjgb=pqKB4`I~=Ib=AYpACG+&vc^ zB>JX!NQ_1iPcvcz-Z^5s=#~^xKkcd>KVE2)HITs)vPg>^V3u@X{)yd7fOI`ZN(ViZ z(0+MwDup%L`zF~~jx`Ed%XbfW$0|n8hN_&LU2?sG4IqRxXRYR&lHz+aEP)_{O|oBH zzknr0YfW~Uk@D4Cnv{)gZk(Q|2+r>l)hY9jVvm5~G)jD3C}Gz7X#4w4O_!_p%3|Fw zm;oDi%dYOM_eiP$%XeSRqUTuUTx3q|C3Ep@WpDrVjD#nvRDDFso_kaZ`4-SRF?XA{omk=V)O1#d zLolZZ29nk&!qHcUw6H-7OQ7l(D}y0*26Qk0>2nEL z$405sSq%$8nlOT6&ek?I42j}W1_+ZVHTyvC+oN#Y)%Qqbm7->?_PH~zFu6Zig^4Tk=l9EkpjJ+%&s*h*27vK_t$JmCV zB3vfUDn?NM0Kx2l>4|S;Opm78pyPKgmQnN+XUXg_#33Pw7griy89SH0LOWY91dX#i zVBYr7w#N}EDGMJ;3$Q}M?-sqR zjn9H=BzCikAwEdJT9M06ke|w}Etus-vreFU3w|dofgF?>;kA$A3Q(B7AwgNn6a#1C zd~U!af%6gh{Nd6IH#GHev}=4TL;kg>86NH-U$tBCwvYCJyv3V{#=L#&dsI&m!S0y~ z77Gnqts6FJb`XfTgi_1gzr5p)Y{sG;Cen7cOIu@!?7u zn6S^iW(*ei<8-N_I?u-C5t@Q$1`j?bF>ax(^|5#Oi4Kot>tpJEvW2U_!ghw3X9Ua4 z9PlIxFsX&xn;fIssA8WfPj1BS?DXcgZ0kTvbGVmZz;_Kg1Z#_}0Q5ZwEsfA1I5Q$r z``_lI`wvsMpNs2hTTV0oLE%gCgHf15P|JaqR=jl&$kDPOW<0wFKaRTbVf}%^Gq1?X z_D8p8@fNgg=WcxUxFtlc{b`l;r^X^V%p6cumg20`-Xl>a;Io44VsJWQKNuB5E01%Q zW-Z=Kx7n}X3K6X--_*9F#jrb494q1E!OKnWwg|6iP?{>?KXnZvN2Ynl+X_N=gPCZ_ zp{)*i@O4MG0j(85Q+?&EEGPT6V=$nFOF8<0*)4j&@gOiC8MRZsAEKVQz;=lsnmiXp zffOCFI?I#!GjV2>n=ARh=R{9BQ!ZuwJouEKQo)D6l;x^w*3UN#wexH0wvHZ;>VNKR z^hR#<6MDYqw$1%CMM|vI|9*`scWz1U?AN)h+B^gqh`cyMkGQDKCr*jYBRY@-%}0huwS_pXVqnqs(t^I^l34-Ci|CV7-IBCEvHFE>1n?I ze!*O89fw%Bqe>Hy2-igY_@bC&!pI{gIezB<{3K7u3Kd7^Hw2B^8fWf4hEr>st^Av6 zmD2@3X4|;CD&psFHbOzXT+FCosrs#K6IIJ-DftdXCQVD1G-7BT?DgFIS$$-I6+uj2He+I_F)p}M>lyegm1$ta0j3q73hfQo|ClTf5B~&&dR%p46zc-C@Fd4v*hm8 z*|at3ZPd|d*T5T1X62jt3!L15%H80TL?ZiSvB%B*1DEtxZ#aCTn{QahZ}D!?INx+q zqoe7vjGBH?#S%-I2&|w;3%RQMYD9ly`C+{led3<2iOYZfC8E<~$!!9reGd0F}kxklwKc02~=x`m}B9U5@e?B{drLwmBxy)N$dPz2)~3 z%U^L-g~LC&bCCzR4fKiLCGz4R~X9qC8c^|Eu<5oaOmsoX9@mCRr^9|EWQUc&*TlM49 z4O4`U_sc3 zh^J^^WbY#0P0*)~p^P7c&UEJ+NGxA3e|NwOYtbUgX;ROFetou?=f_0~jRe*jXC;9X z<4{wDbHio%3Ja-upZH={drREi&FnPi&p?+LF!f5Ib0g~Cy1%w+p*y!;3;@)>&AZw7 zZ&U<}eypE5?dA-|IA+n`@!+#f&1pi(`xPyY9tM<4mxitdOIldW?{+0hM8I|IBu|in zw$|BF>w^8Xs8?jxh>hGhOkHcF4T$y|JO&h})71tTU%6}Mqdc{`J~ro(xZUbjc3L?> z#s?6!paI)QUS}pY>-T5o;FOF^g0khHoAgmz=rX)?&C#MQF|#kCP14EJZ8HRV@FnW_ zA~yuW?v`AGXP|dItKYzmTAy>n8RTq18T2Phh`1JyPs%!} z%u!`d>_|T1gWHHeT?@eLb^czvOrZk%_<7vUK_7|Q^x;oTIw_9s*@g@iHL%LJDUt%r z|D{cSiP5!xFK?s(#va-Fe_qR`R~!9g<^#&OnR1v(*D-Jnr%YkK7Sff~tRQK5R+&Kh zNY-s(+v`{3&tK^S!#T9G{%7JqJ`QY*9eS;3GcKZQBwle8<@3Bh6SALdWVBt`8P96L zhGg@|OdR1f@Fy>y>MQe*kzM4yMC~Y317gx&$$WZ39%-XGL6=vhkmf!xn;nZl; z+r|pEv>^UMulQ^Hc~w|YsgX$(S3q6YPq@m2Ii&nLs;sBE>7qtmv82e^z1zPOeC%}C z9Zz1}N@w-{ ztWuJuEg;!_m(K{zfdYXcbaU2v>af8=_MO9jEkDLp`FX$cTBD0>;4Ahe-{Fu%9`j4k zjTQB|a~hrRjptKty>*TAZ( zqdshhZs0EqRuGN&@Eqzp%h%|Fp5$HnVU0xXmOu>cQbUpfeOOY%`d+TJQvJ>L4XL18wOUA{#>u-cisLtu7uZrS=*>IyPG&U0Zq40EP|AW-kF z?D#2iV2;ma;^0hFk-Q^Wxf>OZ*C>7Wn7N@Osj0*=Bfr#Tsyrgyuln1Vte`VilxN>% zo$?&x^4){@zzzY>Vl<{i-=A@{$jU|E>gW}vhq(GI7Kg(nY%uh}?Y7gZe^2(?l+X{S z3eHb7HQT`!Rjso5&~F+xNiqlNroB1TXye3CExeK@jE^p9?SS0jb|;DnTsPGBS9n*! z1~hj6Y_XZt#Fjyg7reUDEBB$OG?5_GCWR#^W+r!@+i=pCrLbh+L;ZYg=KZNLj*Ihr zPk+w-o=fR-u-V&F^rl?_Rr#(1A9%*My;=4KgwWRH1-4dLbQBg3iRa*Lh@hgWB$F~# zqkj7k-eVvqx|>n0pkgrbreSV9`@CU<$o$KO#vIKWAEX_I(GsO7#z-Gwpn7Q;Bp{LxLv-(g;j?hI11Et3pVn?rmAIslqIP1FMMd${$qPX8V z7X!6d@KTVyIkx)pbhiik9PZSdudowvL`*}t6xJ$VKx^|%o(Wg}5I4^Mg*5ul20O5J)L!D4J< zI|6p@0|=p!gR(;Hx^*ki18a@qv8pO!|Fa=;x>M}_`qGShb`hpX&5lq}6ZDZNVz%sJ z*W6O(0U5vMJR2Wfrb){?oZz2S>0VqHssAAj_LN~uSsyF+BpCj??Hp*$sBm3s><+f; zV&W+jz7-@kz*8#R4R_nW6&Y5i-S~N zHspO8&FWWCfny>=ZK&g2)vLAv=I_kXXMxSnv?REhvOINC@+(6$$7%vTu+2VlmwJ;Y zK_9&6&Ni9JmfD2!Y7WW1a%T(yRXf8wU3%(My36Y)pT|F*MFMebd(`{IMn!Dj>+z0> zQL-ZoPSG$O4UNX)REy%Ds4~{x^~SHET---8{e)b>^b#kf%0?>!ryP7Cr~X!<2l*5R zQ;|A%wtwExz3-XyYng|ZF)P?*W!$IPr&%S`{hwS|yoz(vVxE7@f(slh27Au7!W%{7 z;H%S5riHn|N+_picr8E|5-YT5OtR3q4Qu(q)fI%ab?3dBzNh60Ik&i!FqUq0+(-m=yeMI5p-e1H=PI78V)=<)sk|TI{bnD zES(;o1*4w*KbqdctqJ#i|CW|kkaECON+m|WuS=4UIzV&Rqf@39=eeG+(ZtMau)w#>#h&*+x5C3gHu19d8^wZM>Y@;D;$01K`Mt7XJd^ESMs4OxSF>2Hy-0XYhdV&{aCWApK6rd% zhIB-!fN1s;*r^WyPf_%%8#A1tu=?SnTQ7+60C+~BuvUtK7dZR8;W`6s z@s13P@3^IVK5D*?Hyk#?Ix}1hPjc{Mg^o!Fbxj-1&dStdS6!FzH_!H-e=qv1BW+8i zj})dh{e8pV9$j?ABT|SChyh$pH##MA6`WiRPvl;A!mNI-4mlIy1T!tOKnW8}i0G_P zrKeroZ8$1Qsi?a6}47Y2ow>`&s4m2QK)B zmOxXuYv7TN_s-L=Z!lXOSwp?}70WXtV-6NT4_?Ju+^d$pbvF{Zu5?T=nz5@@I~{>` z8Lj{pdJ43Lp&I1gx2Ho(kx5yXCeVkgQp-VL*2O30hvc!-+{M+d(K2Nt3}OH}aowF0 z8EvcjlzcARS8>7+wjMs}5d;p8NATp#TWeK*t=AxmbV#cSfZ zw_86Fr1hU8%kiiK zm6Oi_XInxlU*qyr=dX*ObX-fEvTgB84eU=R8pg$}zX)1px=PC2 zj*1VYeoEHz=hY&;Q9ApmzUhrH(1JCVXEM0{6RSaz+h1S*-?QF`EBh&vkT0Df@0WOs zi0cMh_5QUk&*ua#6cVqd0e?C#bWvMY?6(XY|t1a-9jk5f4vDB#eudF&< ztqG`2MlhgdZR|cGLspQ9#P0)s4u64TWST6+0z$dcWa?zJWb{bSCdS5v!NFpfAUCXO zjaQivil3F#&5*;f?7B=yutms=TJ7E|{ZaMcpnK2l zr8AQIRXftK3vVIGef5oKoO|JzT~KUM zki$4m26+)P>VtRN;0&Q;7vW_AEQrI|9y+a{cQW{^r{0s2Az#G&!n#(tMGW5BC69IP z&Ulxyue4mb)SK?men%O^-HMiUI*~3r#Ve_4d|Ch_o72pp z*)MyqndieaAD7zo(`Khp0Wb9WFEGfQc4>Fk{r(;Pv>GeCP4Vh{sw(8FMC=XSWr#EW zwX94Vx`$94kjLM271(5JO|Mohf?Nux4ml~Vxl^N)A#;e>DIxE2Z>$bH}s4~pkp+cr2!dz`4fTN}gSD#QDjRP?l< z;&~Tpl)Hb`>g$LwB0JE11cJuQqo+w3TO*qJYxuk}BRDk0Zr0Fy7~E5)%4h9l1yR+A z0RlJWM`>t8pXoQ{BL^P_^|T`kAMH+D2^1j@y_%7n0658w3N|~T(|U2@7~g8$=_sKk zW_f>l#=4A81YYYJw9FYoj)5pvsI==Ow#H!O?Ov8utHgO-V+VQd>JAT@%)6etYxVG+ z3SUfMI{7K7M$!h}G-H7r^$^Lg~ zLB*Msdg-hC3gDvKIp3+Wcitt3pq?uN_naNCATCcwI2eZt&E~LlP_=^gD|V;h7D0G1%Sn;Ac*n=>j-gzrZ`ZNg#p#Q00mOtXc~YSk4U%Tx z&APDFOI`h7gBO`02po~;+}>2ymT{iHCvy(t-*K_9L(sl9lf#ipKL1=zyn$0~|F?RQ zU@wisQ9$f4mkhYuUoOxLH;FGcTFnHOlBOBMFan z$g7@ppVN!9H=6@LlD+U=2-M-H4J*MLaTDDCNbB0EM`U{b@If?0)NT&CytjO$s@-tw zMmGOej|qgX&g9v>??s-Gl`{V|V^_YZvtPi_zap(z?biN>Rz9ybOOrJ?Y$CYKoV<~? zY>ae&*Cm?FE=6TW8Ntdr0=8xxSvV9)^-8lZo)Y7{h!P_JuXZ09Nx2Tnk6JPzCX#>J zmZyg3>*fYOg>zU&%!n(NlLzl>EyY)1gV|FSocBZvDU4@}w)!BEG6B*6i+Q3%;(d0> zJC^@hWS0!kH}pvqmW-4pY{{ez##X({=(~iYz3dSss)=r-ZnA6jm^D43Y;7=YrZQzW z;X?sM2-*K;7^(h)u>~*=DmoFqOkY7NbLBbJ0cu6u1EpssL!4e{f05D$CDw2|8TtPz z*ASNB!4^ZZz590Via@v~*xQUcH(Nv7^t`BwT751loH3xVO`|e7J7qtkIfNTZa)Gb1 zhL=X&9Ra24A|ao8O=ofylFMw>Nt?>^#PYA4VPjSJFUh0N8yX|~i{`qfnnHqJVFGbD zf*G7=A?$FWXB$7{MUnb?>-xEQOX11d86_J?jnN$-wFw42>-M~8YlQpM5 zEWi_b-Q3L~u&c%TmA%;D$@Ud?#v0NKJJsXB3-%g&_dz3gY0+KYt)+Qa-SKjPLNrq+ zmB*X<6V@cV6ZGecshzp%{Bk1LmoJXi-IcYl+UrUv8I#MKaAe<0SJO({DF>hoesIYV z%{*n)E^+m{!CyN&3+QzFY3TCi&j@Lr9$~_@3bBNdr+d6dFXMnu{J%?8rUy_;8N~94 z{%LYV_p`mx{;|7r@v;?G_*YpA@80U$f!lgFj3e#0| zA!_teev`&#e^Y+*Z4(FB5xy~y#m{1K#K)2n>W4>$4ep?W#~4o|7#%b6)sCuVzDBTb zZ*GvB9KZST?^O8dN3El`>S`QU!h3q?>P;N-`i~5dkFB(7djycierptePbuxr+`?rlm?M*=E$?gdBILp3ZlMVX0@;K-Z(!|TxejS_l$rV+PlE<8q zQ^SShJpB!lX!Cv;&YW>%)IuG>G7H0Til6oVbk36#^>?_5SiGO7`k^p!Z_I1T=Oo5r zR(RchXx^Fw`__I49=DyR`)~$$E>WxCG2=_S9S|9JlqgH=NqK+#BRN$m^#ve-d#fkH zDWzX(uMu8I*boQl=0u-DmeTiGPe%K;CYFpg<2*ngv%&}AUcmu|3SQ87UnLIgqKF$sdV5M zy;3LK5PGmy?f}WUJ;Do>%iDSM($OdSb+XzWfBQH&D*m%C`At?wFR$a>HhQn~gaO=C zFCs6X|3}=(L^KHT{QyR%?CwJSzB}g~n|j9OIo5gtyo#qA2!g>84t~GeiN8r>q6`|O z(|?_XZ)N|wzXfb&WgU}QJcHnnevrU2%Un#HD(=MRUu91ANqa|n$6E0_yK*Hsi>Bhq zU{q(nH0>KC9A!CGO01Y9*6pAJ>qA2mersgatrgan6#nZ0@!O`Otti;0LI@F_OtpIeg8UEZA9^G^ zB`rLFBrjLgYJ~oQ0?_|O2zu}v5+?=(`2~M3yJqBiT#S8ovM437d6tmTWtX;-te(h|*fz#@kaQ1Ky$avhdsXUK?0rkO4|kUuym z4hkM3P3TOyeAX4ulfz#;YpnbxeROeN_9^_NtFK#c!gKjkGO9neGg>#!jo%B5JYM1o zdc%T`bncmTc$!ELxX=GUU0uB_sVYs0y82%c6Z_Sq3Jpy&(0~`EB9IydImrS477rWJ z>rhmTNPJV>->sWy=qt&a$P38h52Or5f{rHmRF#^F^ux)MP;@uCpXn;m4*(<9MbsIrRt}kmlSsf+bWKU5B!DyM z0hf1tr?;4$#v1`ki!?wzD4H~SU0xWQXB_2VP5lfK;tqqgm9u9tzzS@C+yWR;BH^1! zZG{U)g`a5UOPh(VLnBw!|NOP|C(Q#*m=m+s5gzePPmd9-YwKS ztPve_N&74D$dmu+cUQO}7<$Xe3GI`Kd+H2+2X7#+ajhIZYJw}>5}onSpz<9I+3kwK zt-dn3BJXz`4WGPG-V%}X)>B^0(;eFcN8s*u_7!dFo6_)~dma$F=&FYC9*Cy#z%^Ym zPz#&{PYI6iV=fJFtMegV$O19+=yj@IJmcd0hhb8H?(!b_Xs}FEy|b|N0;dx)aYe|@ z=Rxmsj=dxE6ecc)lQg_Ov`$J2be5%Pkb0p=vLER=6pBx1UiE|v$p($aNtye_sOq*V zf`W=!Oz~I=TFH1jKDRJ(^9&)?@`FKv{p#U|WcA&%W$thl>Kh2`<@P1%E;|j0sS16v z$Raj69F{V@d2M!Z;@Nl0^|?FK+#Cq3m+59~<^7E$R*0I+c5&8}#EQ6oym^5vRvGikQU^)}T>ZUZTBsqT2KdSu zk<1}SJ-L8PvqD_+d+ql_uV0Ln!xxL0PinS*I6O)S9npvZ`#8jXvd6Qi5WQmp62mu?@ZA(OmMS6xuWf(7-<@~LK?jsxUw?GF_SsY;<+8}Z zUDA}+HI3dY-hV1}e7>r@6&6L1(6tA5nCE^=8;5KH5w*EdC3H(zca6VYPe3nwY%{;Vs+0?>TDbEwB(ztxM)abn58XH7|AnZ|RPF1O zSaauL8Q&E%8oM4|n}%6|7??PL_{EM5B#a}MGj@5THlL$&bqkomh7-daYg1>^Ice_{B^~8S1_LAJ-r#Sj>zz49GtOipL>4MpW zG>!!-f9tI?uN%0K146;I^Sx)!-8b7#v;=(R7jn}MMbs$VGHAxT`Mm78#EGn9gtJGw z-2tNnjIG}Dd#l+`PR)I$&A>kcgAMv)n-x5!O8LA_}*)rH-xh8@S@Sdl4YYP!!Yod$dg} z5?{;>2fX(e6!Gpc@@^zQ|KjWWRNvO^LbPVf?JRtZocynMe-^uyPND@g&AH6wC-gDs zkC<%{`8%S>NY}kFJ`{fq{Gu?1s33RtzUM)>Q>}?pJ)N(ZfY;?7x&oS^4LSo`&Y{VV zDBL9H%2{)zN!XO;`7f^Ii%Pn`rYzQn4*3}h6&P$t2R7Ea_w>$9_y^ZhuXwN4n|Szc zj);aS=q&ZbBSQjWI(>6w(bJOeeFL1!-a_5EU{L$rf!R6+RFhtGpZsm!i**`|(ngk_vvNKz-IB$Di z=Xkxt81oF|m=M_TdPmxtCGCW9X@Gesp1bDqc==$G@adAdR@~AhzX_|hFGXkYoQ_wz zg@}lNll15dX(HDyviuM+)R^@--r5Ua2WKK8%nWV#-jN%r!e$DLGzr%I=yY$=RZE_t(w zm~R9U&l3J4D7ctGril;f?Ne=os7Z;87RaH72drOP60Wgytfj33ani?rG3>(V%t7md zU1B+JA=zYChQKuUv@q-{gU!07mW!7qOY09#)XmeY5g#D5vN_sh$XGl6kLvEP#lpcIAsa-0;zZ2c4d(oR>mbrG;N78qq^ixQ8rwPHxG?{rqYU3vrovm!o z2_YvU19RR*7FvAK6a|lS$5$qAJQ4w-?@GPOD~vx*ngKZ#>vQN9wSVOJ(4$ErveU?D zf$+^Y3SCX%j>17X(!IgNxSi9(B?v59^jOwb^!D>z28a-B;vqW*6cfo$1=rXJ54{4!l2@NxDWC$$DUpc&_Gc@ei zFJRJao82UKi>%_piWjRI z8d)XWYAx=N*@b2Mf3gyd8J0m>CTnI8BkA%4Ixn3o2-dR(DfgZG7z{$!II$etF8e(33hWrDuj4=CR5NA>&h{Vg`hoYz2^ z1AAe1R{uf&=^>zsLr*~8&&<+0jw#Mz&WXI zg`6m7!qsX79pP1p=?Ci8tO`EDOD0jCefbHAs3~&&Pk)K|usGpmqO!NX^7m*4E%FZK zVL_j`X6YU+L*KBYS5CtFD1n+fK>N1Yy{6svBfo2XlWwN&LCAfe=Jza(EMio^rtMl4 zid3WO%#e_b`)>wQzM;jkup3~f>(3~ScmU~mtOO3(t zMUK&KzXWW^k=**+Mcs?-4u1aa@R*+lzI8CT2-o{}t$Rr3|wYSKeJo z1dR=FMyiW+ltG$>pHkIp_MJAoo6)76v?phg!b#jxesO|n=j+_<-@ZXqkPjcK7QJ0U z9&o3u)9nu@T`zfQ#eagTL&SwT;gvfjZ$&36o;(Sk^JwOz!IRXeWlD&8+DS=~nVuMu z2P$NLTla)5K{DOzSwE$I`!E%~tApT0^7b*3*qi}&mqR-r2I`CTH(ouJ)! zQs@i5Ffw%#z87^}?DoZB!l=x25KKCeN zXwlD#X7l`pM8{!pt|)Hx((3fuQmMJ8%uT{5G@{yuEE}?UTck^x_NQV2SDsbS%L^Aw z{X~QRkqgGRaxizUzgiejt`Vzkz=;Fa)rW?sO>OykjasKuw|Yrj%0GBRm;`*-j$@R= zb)j?f8(Dz?0eaSb1L*1VBEk~!TlKT{E6!rMOC@vm7zolfFK z+eytmqX50C-=}Yy_TSjK51qBYS$@FSb~Dmw04J08p)2F+1gXoYtLP^wGmL8(Rs&~@uBdOx9&Lh`7V1YiI$SPn$-wY_dUh(faPN2a5dSgMXQx7 zeyG<#tH>h($n~_lF`(|j5A!6;08n(aslSAu)6eHZXTQr#_F{JZAB1|m zk7Sa<3~wgwr-B5@gQzPf-#$|>$`3wVlsSfO*_r}-Ul5zXF&V3sF!B}E0*k3YM;3|@ z_bI-Tfj7LvEuz^`kq{X_uzC^d{r4Wx6&UH&Z57JMh3&c&I&rt5ON5K+aGEI9f5Y9Q zkzpzErGyi?0EuIV?ZQ8Gx=A&a&NS(CkL|2U=i)_{Xsz2D*K zqHhzy>~@()YMG`#`1~%|3Q5sO`HLm`91i&feRlP?PWJbe)kwoy7jq_D+qNWLwREwb zcWHAY#w2Fb;sr5UZ!}B%^*C5QAcsj&$-ko^K@XaERAG~|gF}xxTou9F=kbZQ* z3rbni}&-bI#tL8wNa(WcSSR5cMY z)qe2b!YhowS?1Gfn_u z^X56y`+GWm&CX}Cj&ezqp|6`|=QqaDE7QfrC}*GyXLl0VV{^Z{W}|Hf_A(-_#UlA= zyieOc2x5y?OPTc1Uy4oXd-aHqAFUPIk7%OpKK%=pdEXwCEe@>K<`7}B!d_6uk=PG# zfY2zjs3j%JZ|N@CM^`AU6sB=}?PH1s$0ZA^n)_LO2Ju{rC&F(d9ywlrlqt_DK6A6h z?NLV~hYR|9%rg=bn%)Ney$B`Fo|ud;=Cnnyg-E!7rI|4FF94y$X#EM)TogRs+NbW;3GSAno9H{mzBl?&=6yWW4!2I<-a{;j&;Pf5;yw2bQk3Nm( z!X$%;0ir8?G($oyDNMian)DF61`^dD#$HN)9_8C+c65<7BPLv5@fSbjw399IVbDE& zinF_SpL^5#3kN6f8UZy1VCt;l_OgRCy0^%cQWBxN&2u+iS!Ju2p3VlFp%BST&{G_U z#nZ)&)+w_Vo4nx>q1h|()yybf)nLflM&Hm^+i*ws3AAQNw7^-uQ7#|%;hB&8n+BfEp4k`h6_}x`ZV`Q-N{vFPx;5U#@%0kryGy zczMlpHpl96%%S?23 zl*th-{u}Pnn$%{$-c^Ku#=>Ua7=A;8ZQei5D}<}M*4$SNLeKmTlfJA|>_}X6V5$pmj|9M({1#kb=4|Xps{Cf!4!bRVvahGB9RD zW5JMf+mr6m>g}aHGb<&$JQ^yT6vgZaiGR~-)du|F_kjdaDX%!ugA7YI?A+Y7BFPAisDeMKgZ0b_0&fYNT)TMWLVodrtfqKbizAJ;I*lo zmx~OBFP^Oe?Z!^kr0kmXwxc}E0)@l3? z3E`B4W6eb7wkf^+(Xc$P|IOvTzrOt7SkpbexlFMfC90-8Ub{z2W;y}<@J?5csEvS8 zafhR*g6FCbwyQr`scwSbWQ9SM0{{-kt*&Qz5Cu~aH$wIERJpa}DUZo3Il{dvSy>U4 zoQp}1H3(uer3;Aif~-qpVY|;m6y(A#uOpiK&PmvqvUH6uHkZTQC1Qy=rJmJeNPy$n zF|AES*0$Lb+%YmjJc&r|l4#dCPsBH+75}6qx9! zbVNjf2u6^g)z+YQ6_tfYE7>A$x`C@pK@z$fTe+MyvH|I*ZE>*)O{-ts{seot*QXpd zF;2zeJH8b?SV-?m@3KjY7&4Z3&{4vU^A`C1`dD$1nen1z0Q-$a$NE7zw>#L}o6pcu z_z0v3r$2LP0g=)IwW&2))J#I=b#%)Qu$NYx{d2%A)i=NCtuaFonK%#(UhdmtJVcK)n}Yi|^A5KhRkY zABAyGSK9&o)4U$==of(nHbp#UqMLchZr3xp$T7^BU-<&ew6_n0G6s;kXCn=JZs;PIp|R1eOP8y+bV+|O*I zZf+e~?h+`nFoJy1uvdBk;{2HKwGS7yWbOxle~RvtTXw0~c6VwQ?5qc^9VXi(IwF*` zj;Q>XyOzbM!KW^^1wCFILcF$Y*BL-*_oJ&@WG8T!7|q6jdCJra?C)uj+(oap=`&Iz z8U7;J)nsx7v)1|4=qB&x`#%^DT0f`0RD9I%c(AXPE$aYbM61p2(?3d|y{vU7y9oVe zGAkSA%*&;biHU&%YIb?u3y%U6xlL(9EG@Bd1#TelIv~kpPb*wJ1?Ge(6_8D`w9^q%cGF={>37o- zL`WT)QQC{DDV_c?l&r_6N%~B~>i6kg3;)61>AUYg{+z&)438frU4uGp zfUEGZjdgSI^#K=MQrBmdx_h965|8FS{}ox}CHSxYuSom1iG6A-py@(a2rYxDXatPQ z!1c5jfL3*9ib3RgCozQ58JQ*3uQst!!=01#);$HBcO~n<19EQEzk177O8QJd z&RvwkS#_=?o2sWI@*?g!;2NMX zKpgDh89ZWi|6a8gzH6pFEbo3+=%!5G=y-k3>-J`e8ZhII$EnQk4~!oew@VpwaW~M6 z75;YG&lIk+*_x0Re<=E}hcI$W+JrG(L2X|)$mum9(oa6y)?^J*H+tf6ukwc56ZY-) z;G4KBU94=IFG%+>^l}Hwt=fjK!#gj2V0trK`z}gu10VU=?p}`jU8|o^b;I~oJCa*1 z^bwrl;WbaJw*5_fy(OJ!JHJn$mG{k?2gbfkRl@-i);u>)j!r|-?P6!o zJw4BzBDQ4sqopq%G%4l7479*IS18o;OZ?v@99Cf(tPjc0uI>DJLm#5Df^DgPTiS8K zn*QSSa?_C6HBJ^3TPfBEjM_@iwjfUGUCu8f?N z`nxxud(ETpLe^lDjw{-tNFFxNXF&h-*4{Bp$wZ9u)DX~`{k~hxOoC~flkQ3SL!_Ut z;RpKw?j`W#1b$3!I%qWur{hIe|59yVqadxd2{m38esDb=9S3?mNcdb)c$H#wB9on- zakiSaw-zaV7dTAJm)N0 zU#v|wz(_JV2W|v#68y@4jJv&%CWgB`Q>zRZ`tHH8zc&Fh)?fWrAhtM^ZgUq!=Bq{x zW*X3hULw-6(lt{O_cyu?B5XHv!)kmIxwoQZYy2=*GsgL*k1_KsBZB@0Y;VF39+GpAi){+=F2A~2Ix#H8zgj+1cvAbG`bV?J`QVhMz00Jp&YP{g77cfM zMvBGee6?7Ee+xtnvDe zK`jfq&Bs-fnc#m}ZxRFnC^C4?uf&H;JU>v3M!}3kseB_}Uo$6$kS(@b zGQGZO=Dq9wYh(`l*=9Bu4jv<;X(R`m)Z8_Gp7Khf;}wHsusNl<)iK3Y-SIw|=XkgN zHR)wESJ-K9KG-&$H^0lk6 zNTJHYL?9ou>(st!)D=0EguA*&ht6=js4?Y#j+ zsIx%fo|8QpO{??{E-$d?mW)!d*mg9Xm{NsE3b>p7kQ8{q7lti5*NJu?fCDv2E!C%6 zeiL!rBn@4$rD4y$ySt(rZZ%pfpZNUVTR*f5j>s}N9f|iozG12$C8X}ib4vcxcSE63 zW2D7wY@i<~LR#y8=u7$pTpSz1rr0ulW7A6Buxj>9My=ESWj$8a!PSE=#!HIWQk~k> zr2qk+()qHfk!e$U6#Ytz6p_h{F$dI~Dh(fRPI;FhP)XkGyX_7CX<7YlDVJTN{_#Lz5>rTTq6LtX z^ez?09|l{*-RGWlhd#G02N$-#g6UmDWdh;8Hlcv6_C1t%WPdB-{ZpeGGLEbXjE?kX_lk=RE|XXn>zE=l z@u?y;tojGnZ^~*9Bj*FFOzz3H(`qy5{lYdPjEc)UZWDiMoxq{;GC_RCb-H=EQBMFK z^qLJjO=IjNrbxF7yj)Dxw~t}O+s^}HD2CHfc6T1$wz0y0ls*;br=?dYGfMDoOH-mgz~n9nXb>K++|&A^+$o`J zUxv|33#gi5rY-A67uzuw_YxIs69~E@MmMm#fb-f2_=?GPp4b%exO=> zO_xC%mct9aO@IpV_cQ2IqCTXv8b-&s0BUKWRy@dt;?xB^(3oAARLo=zHU+uqq`HTE zo{+M7Pz@$FyN$b?+`koquce|7p2pZohz&_wO)?9Uj6NoL-6@%gQb7)B94(_&ty69b z5}PiWdZycgWnccVbB-hWcOv^GX9roWeV+~fn)?r>MaHZi3GTjFTX`5^3Q#_YR2XO8 zEU*dc7j;$l?l@0kj9UBeRhdPQz0}_M3lVNpa@@DFPd?5ZC6m-uBV1i~Y#DS8epR;W ze^0!d?63K+CAn%vc34s5_2wtJxsw%Y&@lrD_LSP5^Q&{Vh4L)HMdZFSBmJ|U9T}B%Lm5ys{Q_wyPp#? zs_U=aXFj=QD~hBpUza+Qw(4p|xs{o*8>jdW>0=p|Vv%;2n@IB`I*4r5tB~2-CxBq; zkaXST#%!az7_&U2UaX-sUPc7-z4&^Lsw=nK_J=tSpB%rY9`2KWe!NCy4g@JB*fCko%Qly))Ok(^UmD9R@V zfoNw$S#;`@>B$XO32w|-=eh)9PGwnaFc$M@kd2G5V3fIX`R}R&IJ6hyQH9Tt8|-OK zHx@>ll&{`JANUm$~eacBB5w7 zi?e(4tw6v<$>2ou`7GCxTcX(Uv)>zSp+pz}E7Q47*iA%SQR_vzJ=;XTJG;{2Vi`H3 zQYwbLqy>J%wQ*AB?$=Z%kTFrA2@vQ+8&&`jBy^2ZZK`05@Mfln#sSW$tWz){iE09I zX5%`l0|LCZGj6GFO37XFt2N)F6AE(ppI*U~7bnAaSH&j8^*@OPFfDE4&X0LMK%}@q z{iR|ZIY3-(a544t9~(}wh8c#~%mCg{Cd7*lK9umPQkAbC?&53)mOx-u5FvXold@i;mC$(?dDE=ALYcSu8 z3Ht&Y;SzopME+gav=s+WX7{}Tz~B@XNk-d;US8l=Qms#Lx7F?Af?N|}&KxTO)})~! z9Ek&=SPInF_=EApX=ss*7AY~$6}_wldGULgEg>Yk3;fm${&A@Hyu0yzrM%dRK4r+- zgKg@=Lmo)OKqU%BuN-nUd|SHnXw0ksyzbq6>+;d|c|+G#prO6tE`r z-S@AxPsjOaSWObA{;%wxn5^gc1gWc)vqj~C42Wo-s1(}>dGm&1dql0f(>*`~Y-0M9j{muqjnuA!F!#71Co>geE&ii`y7WUJka15T2 zU6wy=;-S{VpnZZq8HSFl{waw}=+n^WwOD1S*xSAfB^#|eyjxIGBmnV;dk!7OJ<#CI zo|XK*urs>9s0%;(b$HI*7WbLB?XLc<*TTl|^J$%|7)lr+S}=$IZKVSL0TJR-*Lba9 ziaVQ5XoDoDdmJEp>G92!^G3_Lm68~l5bdHdZ9iDk?x0vIhfNJotH6}1=zp+hzAlH# z=UC^wN7IqEfEAd!oSo-wzyvvm-ir=SzV;QEp6D$4Rd0E^{e35;HI}o%&BavVoV-t~ z%-o7c4t`2lvMlnxO_OA&Y!=DBTB^rm^zXO>IlcQ6k*{i8uDS|l?dJ<8*_|jK5;8jv zlDyo6Kd~cYjgzH!Z=A!6APNUFs&nQZJOFk?$^&B28Kt79C?425!JibC%m`V^kdsND z%u~48gjYXuBRtcnE{2BQX&tCQQn31Bo{g<<z6{)@AHk7;==p|97b=n>83K zj6n+0v1j9yB${la=&D?WHk)iUkvr=>XI~Bc-@v)gif*}I8!UX?JI3SAIJ&}| z_Bs{hTlFG!VjxjyeC_!<-+`p^YW*@Sb=|vxn0n^M%qy}SY2?Fg;+A7{x*j@?js)0& zSE$eZ*v?pm%W!+T&pV=ME+#wSLS2DMSNo2<21iC?+c~mbrMasZ!3S_p~hsx`uIut z8tI!`;qzk9umVIKb_D4Z-AQ@bksKn!IP zA;Rd=YB=ZD)eY3f%**NM?7yiNi6UdM=H1KMkPg zL0Kv_28G=c@FMkc!GFJV{{vdlBF!-5q~9$(dIefpNzBLPqqg32<_PqrP1#N)B7Gey zA!p=N{SygSKIuv6`GgJPu8t?6$RnK{iLPJ2z%mp6Z761;<@wNKvW$prtFQB@nD@%- zLaFq}Ar2t&-3#9!0($>+clUpV?3uz@nXMdWOR(A~dPsB*-^I6}^S$UyIxG3-ieAsKNc-1~;x4-j zTj$e;uW+C6+?Xgsy6Q155auT9Z z!Ox;3YcJ#2UTOApYFpbSej<|0T<|LoYGmG)`vu8%O-ke!T2AaUC4L;ri`br|>u>{I z+u$a?-4bUd{n$wap;WdrB#;tbWL@rAi!Q@yusRcopIg|bUW$*yP$MTY^hLiT}<`3E*9=_37;xTj$@`ovG7|do&Ao}dZ6w>s$S`tR`&eF zm&5#*ejzOv^|cY%Hx|Z|{d4=&jpHM$fNK0e*vVBf`$~Q-ssQ^7{yM~w+hYth!ph3` zC(!dR0MaiiJo_JZI{$YtAvHeCJq?J69Q!$ZYe}J5$Se5EVf|n^f7#W5>f# zz4?ayg`tW0?Sty7LEC*?$P_K2kj?F*)qx4T_3Nd1qquSGTd8S(do6-_M`S8<%@N4& zL!Y)hf;^CwGo8+m4Rc4?=4&J?#4<`$c>L>sIT)2Y9k?oY&UA*H8M;lAzaE72>5_yMH@ zJ+2^)&uM_uk_Qszz7x&qCTh3w&fJYY9C`ysmVcq?K7TdUH;*_rN(%{CMA5gtIECLr zVeddTPl@I}sU5$IwoX@TvlGsD`{Ks}?b!Y{G5MeNTAsoIOXzhgm{vSa$>lBrJHY@5$Xw z2BFFC$6vDI#xjwMt(<}Y)^#S;>nEF8&?1Sc-})PoiFD)*SUZ9ea`*6PyP$fTpP-$a=#9hAfOQR8&U{Agm5?8yVa#(;PYXjWH$4egl)#ISJ<3fXLs=JE-W z`<1+472HPY&e4=7{kbTQO*M0af?p)4xEt$2eukd>6RY5OWwYFQBQaQgv2n_Kc{is1 zm&E*2Yi%=N5wqkBU#(0GH*1>@Me8xYSv1!WVdi0Kl@d4Q`qSs!0qy#CVfWDek9Mmy zLeyukkG3>;Nz*3AyF22Z8#ma=*@&6z%#~lxM_+Q)s1-8mg9Ohzf zCEWBsS<|si8!*_9-`FS~l^~b=e-^+In^9?k_GEbQJC27NT61GmV(di_fST3;lN+-3 z;k0DswJ#%sU$vq#f%S^PBCd5Kv9K_+bc`Vy5g%3;TZ5%a$<;bVV_7|5VGJ>L*vz0N zr$`7V-OK+&WmD4mcCpzf3Z`ZF9HZQF?(1j{$Y}naX>C!rvs54+{VK#m23JF9QTTnU zC^Znxys^A!1ql|(>uCK2oS4mtu+)?@QbCW5$T_#^|K|abv~k9LRgP1R3wdr9cUesC z8&}4qQDENXM>GK**&AX_{6A-(#ojC-OU9q^AS7+4alQHY;3|opAPQ9)JTZxB`<(?{@ zB41ji5i$Y@#QkAZ!!r|GSoE-;{8))~Zgp4XtjC!>g_Zbr`alF^mRG=rZvD3J&4TPP$*k~r9IpeJ(#jLEa>($!G! zUh|iTPSM#({SZecnxb6_v1d($zbk#FHx77;B!6bKi{LK)~Ddw??MbEaI zyE0fl**TYXUtyB7v0uBtW%$CW4awqJRSV{(W%^SeCi!Y`b!HO`{go(t1s+n`_7$hA z{-87$+MxNmuutW1q&wtsL01!f-bxo5i6KTwGUJ*p;eVoJLNQN;% z_oyU!wS8zd*J2xKVNzoeHXq*{WaP{w-ef4blkTq~{z}kUK_LkSDH9Ye(}6FnUdX|p zK=<+r(;^mRtM^*Na|I4rkLHWp zeRGL{^@Du(Eggxr>ve6TMD;f~t)-uKLK0W0C-aZDP%rpE`b1#o>ywyksG!E zBo8fE-*>M&!3sH;k~JRKk?Nb1H9k1cna7jsJALOE-XUx*dFp~ni(Tx*$Tf;h2$u5{;#xA5doFfRijlBV~}RA7V;YXsas?C z^#ZV2^{;M{j2ICtvwOFe@HKG=M^1GHopa}!KyBoJ3V9Wx1)7X zY5D}~JSOu*ltm3ze37N)Ty`TWLZppgC4qN4IT!Ix9J)7;WAeNasF`zXRPK2SW5#jw z8ktqX7Ld@Q!~ePqx2J<6EkyTmgSazHjRM@$_&s zaqW+C{EtmfjSZIh*?QX2<)4{}6_(R2v<6qRX>CyYaX)@F9!BPQLO$KRR*AV4mPfa& zy(@ZtMnK}*{}9@O<9d+Otdpn8spms)O~jV@uPZYFOYvGk04)>S&0F@ZT3H^qMmLRs#;}21;vIPs=%>y^ZVIq1zd-MHapFC1d)TUfTrJk^f-c3Yv z^xS$ls)Hq>w;ymE^>xC0SQoKB-2gQLd{PqnGxc#IX}{a>^G^IP{TG?aV>BgS?LNId zlCW^=*27{rS3)&IdWdNV0Wrz%Oyt$Nb1X3r`@w|%7z(>F_Tnn0xI2KEBz&kYot88P zM$s*k^u?zTh(n#DjT(C9JA>)AKJ!hmyu2{n-}FKtz1nF5eT7XKkz}?5v?RHnu z-+0wRq&^wHW7bJjEP-V6Gmyh<%aMjDhu=rVcpLXk%dq99Ot&cFO|tdKyuJHKpCG$c zpU<(S`Bt1ej)`lwpX{>k9S;3# zfDXFx89;eFYcR3td%50bEx7Z3sfp`G%u*@}*SUjgWjao`d#IeiG zv!-sCmXOFdPP-m24w&D6Nh2Q@=16W{rJT!Q-t149gVXBHjW#;HI5=D_kL#$vv4+15 z-e?QVyp-~FFnry>5j9~XTo!A&iJ%|#VdnSvdFt^s;sDJEY%6{;B zAe^K)Nq!gh6h?izKWJZVBaffZ<@r#uyOFi^Lq zblBj2&ppWrj#W1pSYoSO+7-txePjC;iB;0`QND`w>qDu?_#;EZsbh0MYR#~!acsKv z!DN4nzfJAHr9wN*Kjlik2h8)Mf9Ac7*|UBSWAul_o;c zM>UBHFxQo#Z#?93iJ$R>G6S!UoGEyf>^AV@G$dyVqiBaxx}6Cy{}h$vYN6ptgzpXB zG)wzYuAUp0u3mz}Qu*#&Y*s4hu>GgKL1F!CA{IIO&lfC3$~Zne`Pl8Pdp1ROqSph< zGG8MeRWajUW?uQ|4PE93(_{Y#Sv6idO)3>(BjTVzOv{_A8NfVhmQoj}8Ad`DWr`hS ze)=r0CS%9iHR|Kile3+ed9AZP15_SaK4xmG&t@dqz?+4@yFKM#gCQ)X=@M@m(HbCv z=+l@9K>*PikA7Q@uxdT~N7VJ7sp|(dD9mx$BM3JKo*y#rHDs+07)D6Pkj2%BaQ0q& zf@C{vN}0^OM4NdH)g=Fw8YQ^);}kd}D=g>Y@Tuk>E1Y0U(yn&nt$8R)$O=X97ACe? z1>~gWhj5kPWR;9b9NcJJ<3yg&FpUp9-+4G>%4%R3K+)IKs6D@mS%L}P(G}k>+g-e0e8)4df)qHS+p=-qk8?a; zE@F1f>sp8ug?^)6e4MeJ(oP?3dBYcZ+VTAU#>REeOaM3_VD0-UmmCnwqW*kfDefR{ z(XVdEtp@sWSfMCGy%TCpw;Uv!;5?jb+~w*XS!p(o0V63K#eDLu5>X}_V`yZ;=yzd_ z%_3()I{Mg$RD;$*(P%_jYIpNeLF&iF+ux^6wAZtyHj!Ax>;)y`Tq6pPK}L%yI?1vi zj#v9_t(7u+BumJkG*H8DtQnTJ$-wTA-lhs*7YRQc08!w>lt0=95N_|m)-n;}vFMK% z`4BlNjSA)SDC3(0dM@E2^30_wt(hXNQuR-BfpRl9-5S*AZY`e<+k_XVx_YZ0owi=| z@2mH$e8CtfE>Wj}7STcAfYEO^0vpWU%wRcD#?wd-u{&e%kE#HZ0oVohzi(z<7v(OH? zD~vRkJB#%uwC&ULHilyGDcT-6sw~qGZ9{6SEa6Pyunsh9fR7y5V5i}erscuPlPhK^ zK(vk@&NSUs7}1t6))9#!@CWP+POvh#pYD@6c76tSn#}UC`sfR4?9#&GqF*V#>(>e@ znOO;8*!2ukl=kO2Hr#W=@Nl5$GQCO{`rN4N-J>ko#fW#<^+@Br={a^%8}W_M6!M}R zs_naXx(N(nJNiH(Wca^4`B}+m9GLXS%_HQvOR{fQ?T=V@<8n)fb2kY%N!XZcekQ}_ z6DIL;m$*%4IVKCdaavR#)i8qdTB!d*Z9T(}paEuG(g(CHkOj9o8@(gMnV2zgRI^jF+p3e!~JJfR%6 zeof^Qgl{f=mD*8|RW#7^Wo-4!V{&&(%hGZmr`Fr{vFQ5vd<gM$V>X6QbbkJ_wRUZjKp+ep5w0`si#AI{3Uzsn!#zOk2>j>1XI-@8 z?NX7quWRtD@fVcyh54%%kGgRE}Kuk8!CY+G{#P z%j`#aoFw#t>#iv$(fk1nyHjrkx_OM^HRH;xG@^l`gdMk~f z6aEzMIL*EKbG1Bi60v}>pkrqW?bdm!~c%ynH4H`JK|EO zC#;A!a}GXlu~QURqfRZ6i{6|u1)UOG=l)X0>C$sAG zN`V@S?ad4QOLkx@&?4i}`1#x{ll?RPa|DYb#4`^PuRH4-JtfOKr?6(OQ-d>7bLz&& zE1zK1gOSZz&96vH{{UN==QVcYLisfgx&h50_lRO2$PB|MoCGe3m{s@rW_W&#(Yumg z4$Ax1pj(){V6XOCD=e3r&0$Xt@{mI;=Ec0t%jL=Lq-LUM6kjp;yz>m;4uTUlwdUte z_t+XQ5;>p5v}DYH8=cB0%Wy&_xIp^*qJ(tG+QmBlq{W~OG2*Es9&F>xV2HTK66X-H zK8jcmV+-U|54*TuEDdR2>zw~5BQ^cuD!wfdP6SKm`m@CUMfLe7lKzzx4HV?UF`O4Zyo9N4A(MA|#T*4&7Z&GNp74gz-98BzD1jP>`VPiky3_=nTYb&_ zQ9vZO6!$^@euBJqIq>89H8B%h&rslk(cE>=`aSsLuYbfD)QexPe5IGUGsO_P1BlIrVbzO81Yt=?=$l4Mpt<%7v&ZY zC;#Zf^LK?jv@gBceDAm&n@B9)ul~fxH>tAG;OWm-0i$OGKNb;kte3PawTWu>c6!~K zV_&NH>4d6_pN}P*-;MDu{zI(zFtfGzL+s*|@sZWEAyzGJnW46RQF7~^>pZGzqsHq7 z@PMsYJfMf>9F;7F@lcf}$#=@!J7i7 zY6Wa8(1XQ|o6NWDIF2~1&&U(R7+B%wAZ+OW7rjyDo@B3tDad zGWy47W@bn`TYturE|bm9c*%P74jNFUZqvVn6*TBHX7cA6go~^!3(Nibz{-SD-R^mK z#eGf?G1d2jMAchmZ}CK$>vXL7A}4tBYR-`%ok})AS>dKCQI;ZkH|m}8iYF~0`+su( zAU?|zbUNz+5iQv*q-4JOdzH&1=QOl8ba)Y8^{*bQ%^tYDPfuT?y^-I&4WH@k>t3Cc zR697()!g^mLD)JXd5_!+taY&TQa*oPKi;enLeq(}@|Fo$+L&U*6GdvCN76tdMJ3-G zia8$1l4+jH{jGf8@J)=?(hh#7Lu)n637g@N8Qz?6I0-qz*zAn2O}*@bfZ};cDt26p z>Tv~!|Ey_5nSAKIxvw1On9d`>2*zBya@zA;&Wzo4!n1j+{rG!vX(9bw>T%h0XsZPG z(Pjla^^{}D8H%cu{p^fbNgRYF{)~Twr@}KbQ5dDl1s#vqA-pdUms2q=YRWy$kWdn6 zRzLMD?lGJRJga=GgP)PhA$D0)>F!}o37uS4_1Gw1VoEdXN^H72r!LFpCs%2W;AZE8 zU!RXzDEo+>9u$y%E8(VmKYm%_MTh!RzCftJB^2qH*Wk&C%v*uXh(BM)6S=&#MXPAF zVwdfDl}?N(tIz*_`ynKnyg!SB3h$xnshD%g86~%+`Vy1|%IaSo|C`CzcH`GRxgA6N8A8 z{NB{GgqceKDFIHL~)P6@n!Vp z?G)V~R0D%7y+Z;D^OZkx2#iTE7yznB7?S|i0P~d_2ZP@%R66HUx8@TeeOl6fl`5+l z>XT9-TXVoBh4eiFAnm=Jp$hx;KEvB_;fedpI<1Q~2!P0)wS(O%23=32S+5qAxB>&4 z-9(HQmjEUh38MxPzIG<*)|gU{>Ae|c8P>zdyE=#fVS*TF7@WXq3Qr4I2@HV8^i6-d zNk0Wb?dmkTP|BfjRUi3esNyC$wEShoKo zHcj<6tX_%PxLECXqrfI1isk9*rDu#jk+_UZ%#^0SfrLB>F_x&Q`{!`L;`G6-x4hu# z>ZbALOxYBL##-RYAh*f;4bKY=SekR!B$Rfx={w$ZEk&d6;ll}i{CLEHL%jGf?S6%Q zZMT@LjqyK=$2t#R(07mvxPVQko&E1{3QhNAV%2_xdU(uzL)S_nwOo)IKIhX;>h^{hN{jp`j0RIf@(p=2Zqo%SS2aXp-4>Yha8!3m> zYP?@=KOcoLlGa#=;y*xZO;MUd+V9U5?H%s_O^bxdC~1l@K6?QkUCVlAiPl~P>hH13|=pKYTCG7NGy2&A?Ezuk@b6W zF6S$rapDQH{iz1gJ_Z?eZRyxeD@DwM(;@kb4DFgd0bx@u3+19d0Bb7+Ia#BhBzDJL z*NgjbM~)Ko6)dx1fY*KeGf}J~M|IF`YS9_gtFrFUR$?&j^E~3h)Fac#yL!v~feh{9 z{1}!@eukGUiq7NA-x-pBk-3+70s~b;2jhweLoE@(PuwW#@-#O^C}u5tsmJ&dA7U8x zon_tGyyxdQY7%mXoFwzBB5a`v?lYGV7UMgW`Q*ty&A%uM)}dIn~02`lf=*emX$ZO)RP;- z*&A(en4FsxY?Sh5tSO(qL#=|@bj1bKC8CQk1J@^pvnfHZ6kDsJYtKK>3%J=f@!vCN zwrl5(ad^QW7Xe7y=l_~3GrMd>a}GY+;w?1KzE#y9bJwJH7j@m@P^x{-6RU7VRUcS^ z<#LQG<``}hz;CB<-5`jwp6+w0wF1K8uh{3kJg+bDVKr)(y&!U*?r2L%eqbXxCKHK2dzdSljxjh~xZ7V`u$l>aU7Ve|7(Y?uSL z`?%$R&4ZRCjkp)Gb^OhDgI^vhuZfz}GDJT}(3Pz_4j@6J?k3K%licYC83|U-jTZ|5 zta+aFD71Q8FK}O>VVUnh*=gCVEK#IZNem59Os>R-#`y0S9~b2d)ZRCIlt#_sd92=B zS$s7sha&jHLy(8Gq$j&CjhjFNa~-2`A{Ao zW#6r0IA$XlkZ^{8VEuz>3)w=n69|fY?$%}Rh%A`so-r-F-}AB&!i~DlfBw^JG4Ic; zLdn^+V=v=l#-HLhg56|~Y#lWG;c$<|7}v9}fz7=mOKHm#Ease9kxu{8RN}X1F7k8XU!txjGkEj+V$kliz^!rjy z={ZnJ0Yams-awf-3F85^J!U5OE8k^&Voa?9qX;GeXPpR%^FPF6F-k5uu4&DZT569r z$P)GgCsHP1*g;2SBi)0;olY>8YNx!KA^wq+LPVJ1-_OdovrmG|ptp#Xu)*mQ1y8Zn z3H*uvn)F}eEcwH)Txo6n-DmOiZlfag0lzOhWhbWa$LF2JfmlCIJ@+-C7Dbw|V^UBF z;4dv|;Vn~dnIt<`DHw7ASI>st9^#wfCCzOdNjgvt&80gQs%s67T%`)I(ZEf5b~a3xB|!C4RUc zVmyt=GnrZJ-(H`UHviAau<&pjs~fzppjfZW9a0zU?S2|6zKY~B8QF4v!MNvsz9JPGM5u|0HvCcZXYXP0%6ihQAk^ZCRi7sE5%i7k+?a)C@erh*JO5PZ zXfYk+wtZ55=i7}#9nn+cV-nk!J2yR%^>U975LpWXdpkODm;}jhl;`L%=v(I4y2_sT zAH(+;-Xk^ZU&UVVk|7g2=L!qE`)~fzUf0o$?VH-Q^0=~=w6_ZnhcPyByg`JGqsgwM?7!TjvJa*u zvZeS0^0G3TF$olo0~lez^0ml=mwI{oL!F4SGVy3ngG`|UhL(I`{vsh(QdWPkNOm4i ztSyhCla!~`9lJa8*kmEc?@e5eZLbX`CR{H+2JdRN4aV&XweC%KEid`y`}PGcf$bNh zx6epUqSj7I=ebTr{`pWE)811_Lo34{jJ^D;wL@=XPmL$C#)MNtSAEayo~{EIQ5BDf z?t^&;<@QE!D2>LKi(b3NHv2K3;NH(x0}Z_`}{mq?tDLm+0z-xLmHvWlrvuk^u zRi18Y=%>9e8@>dBFzEq3>`%c)zyzC~_)GO8g=QM=yGt~h}w3A5SnMr3r1eXV@ z5SYeuKVCEL<*B#xW=HHBL2^I`z@2!{{-T7XBC{_iG=jP$o(Px$!ix9+IY7q0Cjd{l zE{jf+@KS9rC zY4xgMwhNo3Lt*?9Y?hB5I*U@8?m&8k4Xs15OViuyqjw0rm54HcZE}1V>{Zjk2>OfA3%(7yna^s2(p1(M4H1U z`31n(W@U1G3iMX=0qnB`>e%*WR}9E|#4*!yUCP4XFw4#DOv~=B;n6Lw{g0m6mq3zB z$%2lV8~s-fcRkP)ev^hyVovb1&h~}Bv~$%<uC-9^{i)ma(l9#jDP^$b%QJj+s_YOtkko|+;)s1#)wwG!{I!wx|>+KeJ* zM0Kx{wN;Qt?$OIDUh}=0__u-cjrx>YDDCq3gz@hUz;LFFVfD*751!+3SfP2g>>$XYDPKSEEPztKh5~C8QbuGp=qRjfLdgfsUhB6JKKDbDNP3lO{NESk;EjYs z^txB-uql@0_GTW)XLm^Es$a@}luDfX=VkgYZ`(q!Jbc5$0-`1yNdmfaU{ zW})fArQr0Vq$)X+j0A%pSHVrPTkmYZ5M${nWdFnN z%7Pau$m_FLA7?FBxzhX3)Ew=f1rtH-3)tG|5AQyFE}~~=w~d_X(h(jI3ee5pAY}%A zpE}sPw^4oL^)?b<-1%hw;t0L8a3xl;I&R#cFv=%H$aIg$ZFB-RR#U@~3h)MrFk{wm z98j#T279k52D=joA zzZ$TYT!Wrdq2w zLV#u%?r|bWoFdt7r~+u!IqSj+eST`2(ZWFXA`qkjn$<6mzp9>!RrLP2#`SGHV zkFXP7``o-~1)(k_KPY)Ov3LQdK{-1PLYe0h8sZ{UF0{o7QhU$KFLz0eLqLCXyWCpmbzUqH?qhgwfRyMk`N9 z8RY7?#4<|D?~kFs`-cDXke$!-m7n?^nQW~7r1cGW-Dzp93LxwK1gdGhM5W9+k~y5w zDO$>fQN#rkU!64S?M3bxnXI1lm}C4%=@S4E?J7YQQKec@ZH_xxP8=oQ9+kctw_L}&#-%6hFE@C4G%7^J- z7_?%}eB8Ync3+LTg!pHb3PqpR!9}xS`vI1}`rNQ_eDQ_```0DiXUmPWf7y`U(oYB)zuo-! zImU)-K8lu!D*UFGuZv|&zv1lM`CVBf+_Apso}*8hb*j^r!n!;bk#zoP4EtLP!RLWn z>O8ZC+J~_$8#LucSU-%Ahk88uG5cT`^U>oWTpyV(RDSULdu31HS|9Q`*r$cTY)D*P zv-R`p49PIcRWunJUEnPq-C631A=LcWc_xn~Ps9rA=0JZH!Zg;8*btXI=TudjMvAfS z=++Av{IApE?5;m;Yk|lN{=LcmN;jgcpZuH?;RRT#5|*;A^S~@5clU>s9^V&bB;$)J z%eQqY4qhs$&&W(+#Wa42=!;n3ZG4O2G(|qkOu#Q(^N5W8rRpIa@~B|6-9i}aFv>E)7>LGq|JQ_u|(Kd|oS?RDos0_b;0cSw@(iA%TOQV&3Qp{!1tT^Ikj3zbXkC)^}&v9h0byZGg`&PM! zvA!>8A$sn2kZTMnsc=mQ*+)C(L4O?-+H`n|_cb@Fedv;&iU;@0x_B_=HfWcej@6yz z&5xVC5}9ZRUaZ6z{@2S1my5*vwjs;l>TwNT{1;{(?BxiQ`n{(YAy9-!Y_PSyygK}aJCQ?;xdZnS$rB6L{Pt&vc2kZfos&du!@iYCKDqb#q+uBhKKUlqYJ6Edz} zTxM3>Pw*gAy|SuJm014ovD07z z0^}ZclaLUlu5{^}lG?nQ7AN>{xi8}m*~FB>X>?i#JxDUqH1XZ=>?)7__2IT{CeOIn z1#g}s3GG1#3oFx4sG00xTl(Nt?V&{aqw+F!c{PVDAy>=t`Q#7pL>>9nUHlg9EOCB_ zRNfW!W~2K_4j;o(MQRycXUa^MdF73>T$9D{bTb>6Xd&zCNFpSv$inV_kHGWAKhiGKH7i2_HyIzjFf8IJU$Z&>_z#OS&r z5nmVHxC=C$J`FIX?D2synn`%*BU^?+I{I)II1Hr|MQAgbrf&uaOZCN%9Ccf9f1IoUHUBv#30G-86&1u3g2QnEpah8m z7NR}RvD%~4G)?0DP`D>J0OuB@63Us@R@A}F_FPX#-&9@jgc1Jg`DAkc)UUB<_Pb6W z)o=^P$5g2wFmPDYt_=0e1b^7fMc>u71xL!sJ0JT;x=yvQDirHQoeeURf-k2nWx0Pf zh&;LJ=NPjmWkck4wcD2Sb)V_?hW(bUVgz6R=pespFx?pjMm2?h_*PvWMOek)(=_@& zvnM+I!j@&zX&G_l0*Q;8gg{b4`I@EV2(=YfJ?We(Hp+N$dBBP7wY(9yip!#R>EyqG z1oIuN)y+yAnjDa#03VimO;7`T0nGnhxa6#?HXf&PfIb)f{%iQ@5AMNH1MZOWMDpU0 zB$6Wq|9O_15{HLz{Su|s%PnF!T2e5+ z_oV?Bc-iSEtZF)XXL9Bb60s4u+=7Fthpw~WJMQb@6qIyyTxBE0cd3q!I@$|WI zYDc?syG`J{GDN~=ZEn-XkuhGZwkhdco-aQ`$kGzhFj6*Rrf{=CuIT=nVITcD4_8>_ z8F*bLmKkC2!}^TJ-Kx5Oya(PWO%VB6oAK7#oS(EZhkmR*xa5gC=I+qpE>-Q*kC7=0 z*{oPBwO})?_&)#~Z3|Dy zb^1NO9BYeTPMf@S<>kfOU9Vj?`Drsv?-HSekD?v(u|Dh1zm;&BAX08>UV5)oyv%F5 zP98s>v)pyQsO!f|ta7GuyTl9i2T`A%Mbr6@kljaS&o*rh3B3KYA7Gij>s?8{%X03U zr45fRYxf0^aMlkz|J=82Yq3|pc4YKTx06h(l(oXUUml27a6^Pc^SEco=N-+{hzt^ku5Y_UCPiSw1+j-h8gxjZ0+z z;fg7naraYEsedPf59Z;`xu^z;XUz|ZZ zUtE3LDk$c?s7D%^pSNA0 zou1!$64gaYY&S|v`d3Meo3;^vMc7UA0Al+kcg(4n%P(Qn_5p%qE{@5tyH`I}4E>%4 zRSe&-ew}mL)aP)<*~IB%*!77Dy-hOeqh<}+(VN4O86RC6{i55k&Urm_nvJTywvmM# zbxfSe>d5~zyf5(rdHzG`fgWBH#qc%=J?tkE2Q+XeP0Wc*NrNaKB{RE=hL0%HU{~%D@(Uv!eA&SL)8INmb|?r{JxFd}#z#P-%)ogAw1_bHhTID4n9t5axo=aj zQf0Ntn1o!~;5}W+Av&#vW;C+mV@&Al{JPbuf0%_e{u9wIokggYyET7*Bz#1En7WA& zwq0H7)+di#bwe(1IL}IB%{{!o=7@!maJRaquTZMF}t?S9I2e$Ed+KnRZHRc&@TEj3TM{?&rwJWF#Skc`) z={)Id98x?SU-OX&i7c-)U5y$MsS6A{+llxh0%tR0eVqnIO=yoiM~NilC)&-mnR^Nm zPmM_hC7rw6_OtiAV@E~xp41F05_5&j+nx`@*l*omDTrW4q~<`o4vnAH38D>56y}us zb^KV)xa4$Dx%9MOZ;kN?@i>}c!3QtUXxH^We3_l$t!uEIRa(=&{+Y?WxAtMvx9Fq@ zm7=J+vvqaec{VvoK9V7JhY4J`Wi0&@#jH!(nTiUW-}VyJrDvqA&Btu5DKpos5R2o{ zn5#ZI&X~RbvJ@lSiuzU=vxh6r;ViHGT)(<74p_vX?~>GI)XcqN#MY3`s9;1-RO9f< zCelfCh0hhK;CK?O49S0lm^w;V>Mu$LFz;qvzT9nEZ=RnuZ?=f&)-x(TkpDCswiu!v zRv4|f*kxz);1>Id>!5JWrXb$n>w>B&1?o~xJpjjG<`>K@!O*Xxq zMBO4wqKs97V(nm|_|Lyy{^xp}*y~tqn|n5TEl+p!YxtIn=q=IKc2T+$PYRcm`HS)r z#}gALG5qyvwHJ{*gQ7w>*Ld)a7FJU8$-j_}XBpBmH9)EJzISCs7!XBNL6(V2KVREhrpl%cxW8`=9IevyH&PD% zk|X9KZ2sZ`@dVsqf!MOvI}TD^-S|W^grn|4Byf~Mt$$?eg+eX@u)1sFjP=X#9hPWo z(^`t$ayIfyJjZl5q~Q8thRM(<<@jw1Y$`Q`MM%Vf*ToZB`VIt7#8}<*EOB<-9!gt` zn*Xw$YIFEhM01t--o*6#-b~zIMWYaz7_Ym@f+X`ab|T*IjIXw(-q9ZytDi)tb{LpZ z{ySY_g;XkcTV<)c;lRRyWnHs72>Lk~XHwd42Ugc4*F}gsGJcH>l0j<==%BoC5!ewQ zuZOdFWxVxf8R!G~&ZmR&FclhG&DjH=;CnR{xwxA?oR!-4Zy8(2uhC;+>KDD2zRLz+ zXBSBUbwj610T%6blI<=^y25JrZx?uRvsGl+BIis=6J3k-CZ_r(q`3Qv;_eT=b{^^d zns7_5UdWD?X6sdFoufI(ln435%u@{_ep`Z9jeKHO8(f*&DB&1dcV&nFA`y@)?Z5mC zmHaTp*HG6n*=9h>&E|mqBT=~PT3(jiQa^f_IMlalBq1`^vREy51A6F}eP`D8;mVVv z=qb97B}2eNnn5N<8;GU{!2sPp*ESYrT zvhI6!f`4$5R^S*U_lfktlS0sX_D730pXN4uDS#QwHkfzLX9AuR-=FM5K z98CdwyKBTr2G<2A@Otwh*FNS@i4Y9hycivyf(#mvrbBGJlaA^Jl^u19R;OXj10Yk) zu1A)?B+0Kl0QQWmbjq>kt}~L;UKchBA|}rWPv18k^S}Zm3WA10GGwJS{*R`!jBE0J z!#<6G3gbM*6r@W8q(MPM>5v*dLTVr}a?}_-dW?F0 z{-5XhTraOT*PHt|<35k?S(o4gk8NjL#7_6deKFZPLnSi|@X0z}P9;LKBN_3JzzRLh zShl*) zCMJ@(=kD_Ya*Jumw2%qztnLpeZ1Hal*yx8HhRg5yfs1<_YXzrwhal9%Eb0U96IWvQ zl=n&VCj3X4l2uAfLd?m)FTuh(*W^r z|A^z?UaWV%iA++~p8xvlpY$8CPoz*~-z0MWy|fEDW>L)weDdHHo@+yn*?-#bU1^-f z-)bsd((NIReI_~~5gO$qPFOXk%LBI*iY_(S zaX+V_l=T)5S^Uah0a0&ypJ129jT#F+-Nev*F=Gy*_C1pDM@`Khda}1_acJK&)>7heNom!xTZ4m$$3wSRvT~VFl%ZLZNT=qr!%J1wwcUo!`U5lpOdmR^bf}w`Zlu*| z2+B`u8hmn+B;f$<0WQ%{s$wX&cH_@xhgQy zkd<$Nn|plF($k!62c4=JmUxp=*f-@?lBQlRpcr8iP3XXMZMNPoQq9 zR8`F96r0R5$jc|Lq#be&5=S2b8-H}r!`YelmH$h51#{zJxT<`JOh|jyt*1v0d`*d; zfOL517Ox+Wov2x%iP>ml;4|rhG$v-v*fv;b`4ao{E*ZxceSNYk5@9E89SQLS#6c5H zrFJevyJ=%EQFdsZToKPm#|T;Z)t@`eDN7MQk*!+c=YuS~J!5Xk%GZ!RQLHv5KFQ-dChqtbg+Am{dx zFvYonl5RZ(DNX6$-Zgdhk|SXS-h|#{k@nAhlB9f2iItb#pAdO*0*6shX60++tACV| zFV}-bPq*5On`(Z%?{e=bOsm9NXVHvv7YNof|5qaEUHou6YN2hvs^e=3=ak8fGbaw- z_h~5b8Xt(EE)R86g8F=0!I5 zg=)#FhVJk4`|2qHhg%Ng6?q8q@9oS!G(i8A;(m+$P06bXA2jkA1o41-@1icT*SwAY z$cJdrk72}E#Z4w{%a46)Tr}KoHrVc?@f~pF_5JnZ>lk;w#%N{fkCmbyl3eF4Myp$o zuk(PJV*nL~^?N$9?uDk$g#prCySWE1;Sd`U#6$RohBg0hGWg$GE9bJunPvZ~*n zFR854T5@{tEc3$vsYJ-Bm+KY2AQ8Q?JHw~>tYm^|StN()4yN$Bw4|y(P~>)d_PNOH zaH&~;&jP!Q-%7E>9!GOG_j=A(R}xmw@HpQ7LGk%danl8N9>n4$x%f0WxPoshV>UG_ zR0$U&bxuX;QCIL!DvG&P0X63i|GO7J##ha@9_8hajErE%Dnm*D-ftlZ%`P8L6LTpf ztlXDX+4LJCP{tCUDUED$M)ShO)x?DMCTj?KaXru`KXGCRw_%{k#i?O+v zyvmhG`wdz~&RL~VmKl+~_yEHF(lOr{;La>LD5T_6)D3=*la+d*hru}Q?$Gns-Cax} z1{~aZ&$a(!dsvpYjk;EHX5OmaRs{ODC)S^=2?&{I^ye`H)K~D-44>xP)(AC#huL)A*X4krx03|t3+xRjA z)oZN(3q>x(9u>$dzgB~?;iMxW($bPGI&|@)P}FX42@O0I^$|ufXo2nYwh13D`}i zxIe_xXni|1zKRP`b^F_|Bl~KL^`@M({v1_yDyH83J2ktt&1VjDcM!VwR-&#)U2Fu3 zVy|PrS!g`f2LNM|bJgWeX!(>7)NJG`ZU?KYI_`-)O_fa-=SR(g-ea;mkO$wqpoIb` zy;>^6cp<++e)9=z%;u8txoAfvr~C+@fATZkGfGspuj=|AiDpInkGzNAO{vr=XCp<={Z#~D;8X}Tcfq{1@rF>kn{ta(>9lw3*9|0QXd zeUp(#j2Hv<9s+^3=Odd-)&|cs<>amJk#hJm8z51noRKCeR|ua~(9Sx-WI_sMY`-2* zo|-cBjG?!10MmxAc$TxxL| z8kp!GEbCtTR9_bd!nZ=axB4aQ1I`Y{!=s~DmNuOM@}u|dri&k;iE+w?$3vlxWy6Nf zwnno)!r^Nl>`DxA{tjyK{fS9-ly9&%Rn)GIx&vr~_^Unz8;9qlH#dZH>KGbY- zMbdvjx`@o^i3-2PYur7lmz5wAQmcmSWM+rxN^gy%A6 zfL|b{L!%J77ooIw z?N0ZE1EUz_+xW7GZJcLITKYt!#y1kg&Rq2CZzm0NL6%@$!WeS?qGWUSC`K#`2 z_%|NS)f;E9KJD34kT3p(XFMhwqX1ng`);JKP>zOmNeIO^ODJ?cA#7lT?(Gz0x_W*U z@QIMdYHlIFbu)G}d-`*whhfa1ivjbw;e7~7qm_ zz#%6gpBC=k1H~nK$=3Y7sx({4gqECkzg_ZFA9Ptfb*=tO$gf(me4<&I8Zv4oLUlDS z2H|i#yR_hZhkN5YtL15a^5vcfDtC_SIRfk-vhXNgobFlWkJdAvMUMB3^Ts`bQn=37 zR9a2OqeOzLZ+N?S_`#iR@p8iV>p&x()&FbkOt9OVrgcSt)LRu8Ba(!M22@`dj=drb zak{&V!OAmWs?3aNdTQ#x>!BgN352SmLDePVGF2O+qc_YxLox~l@-GxK%Zygz>Jn2n zZ6yRw$z092blcQn=n^mNLx!s=y{18$Jjko?wwUR6iQ5KTC_N!_&xRT22bq7>vaff( zRV>^?$T0I}`4G~u46(rx8J4ObpG)>g-9mm|(&4bS1gr6nbU-(!{*u+3=T}B(s#+ce z&dY8ly{r&<`b{NPInBiAn_Af|XQe;Z1>$vBNL}y^krX@?*kR2(oJ%GzO?hV-?BrD+ zmyfC!(_kjU=B1!FGz-x+(V;DqEn<%Q)NMF;h9DESzN;m zoUsMgQk)i{;WJFvTGkO_dE zu)c4I5v7Sl&pi*t>C>d2imw7t&08wci1f>=5n{s-rKBc<35=qrjy@tq?8nrpbM)l{ z#u{5UT2OtiNj-y!!thsrzYk@%AJ28@1`9%dOi<5%)?y-#ydFm*PcT4^a? zB_ZNse3FdK$Qf+g9%7uFaHtAr1Rc#n!YNC%Y>M~R)KX*#;`n(XT9^*tIMFXXH7Qmy=lMwhAwH4l)PC}(sVe)?`76$w~jE|foOdJ3vB0l zh+wPXp{X4A>chHFhAN)?XlhEG51sf~6mqifnWpd_F#?VElW5&ETi(LXG`L8gX)#hn zeDUW0rQLxWaa!)*1G+KhhKu!&O6{gJ5|2A3Y?bFZ_QV|4J~DN3qp%M%-kx_04BuC( zWI7aY?J5{eQmU*z`^vTsr`++!e=k(_MszFe=l=OGIvJN|qx0tWx^UIUAieVMK7FQHfxL}-0dR9p`G3l&1g)7Lv=c|PFLBD^O|;ko`2K1 z>81E*pQ`4)72pf$3D*Lj`*Cx0Z*?nIz{h3@%h?+bOT~dB$0mnJXVoR;i&X=Q*&yUPz(R-d<|685EzCB;E zm#isEYTWG=SyaNMXY)VZ4Hy_55jNvC=niX+5KlLnY}fU2k5<`4Otcm$HxQi2Z=amL z{UiJ-$`}WIjW55m)w0>s(-Z%zt~>XhVov|*%`HJjPYiHK9Yd<1R=%caO}IP4m;Q*& zPiwZMcV|8lrMS+KHIsf~jyxDGFQ0sKck^^1%GbV1e5jN&P?c**;xpc4wF{wk_K&0+ zVyBqMb3U~{Qn=5t`_=1gNI`LLs-yi>cXzoqq5S-NtX$Y8md!X`$fP3dfw;56$G#QI zY-fduER2}wWkrkDB3hBnlzy5#PYnIHt+Tz zujMPQhp>uSk;RTXDRRcWqw_kh%yKDGT|iKpU@x|>ZV47_B^{VVOjCojRNMoK?gtI} ztbBD@GL@?ZnGDolRw?AtCAp-EMb^3gusIO>zFE_$2(XL7K~G>8H1}g`4T3FP{Dt&x znY%7EUz}~T`P^P~H|gHzfM#?RW}|%ZS3_NvS{o}I%$W)PJ3%TWI&~oRPG`?Dc|kg3 zNr%9m=CUi+?Pf%8=5UE}C&oQDhEA^xpES=HXF5`BNB$I)D}Z$Um>e>n`Bfa5Aa-25 zV*cEaN+SSxW3VRh>M!!)94-Eq);yPqv5@Icd_pe`N&WNF6x8l_E5`wp`TvU zb<;qM0G)IR$zI!VqgUEDQ&`xKn%cx!qJ;#wu(wW`tgm&6F>eeC0jSUH^{aIf9ORZdm}3tKRke;M4F zr0r-??%*H)0hQC#A5JMKb)eJ#uWV~%`m}%b^f-JC$Jl5M-XxV%v+zUNY=XcuMjaY{ z?hOmtd{Nw4t>3t6@=8Kchj)|W=nC!jZC09!&Bn2K0@&fn=(dm99Q+EDVTgXF;sOyX zkdazokS^`eJ@Ijjty&Jc3ZVc`SYsur1}qv<8VfbIXGYvSa~!q!c}$UA3!$jOX6?%J znWO3ZuoC=wyR&a;b3ChwtoXVKYJo|qSL`u&K8WIcv6Q3_NOve_`MJTdJz0*dkd1{= z0}4C&6<>fImlOT&Su?6f^PwL?g^e3;uZpY%uCJwuF3%S6Wi1t422>?i-Jn%`;(T8j z&mm!*7mM2D$0dER^XQ(tnIWkf3%#%!1tIkQUc^K}0b=A-+30n>Sw6MaDxQMbeJ3{LIf&TaD$yyBjfy9!yAO^>YCyMD2qFbLW&^R|&eoS%(tjS`}9xu=D+`-S&v| za<8A$-@#V<6a*B)b~)_9%qE})e9auf7hSIs@YX%zp?KU80|lw zs6WS`K8<8bdABi!`6N1O&#&QLHpz+5YL3mdgcR27>zaE<=#lNXCqGpfdbGIHvy@`w z47KFWm+Al)A$_|{#caUY9^I`bp#%v4h!) z>-Iu2Xy|6CyVtEUiS?fN5$^Amiojp-xwl7TQKBzG3YxquH5$vj zY@n)N>?*VzaKHO6pQ)W9&T$Sq>&ap@!@!@GhVb8MkmPi>vr-w4V|?pFS@Zc%(_Xqj z_Zxcg{Xz@$iiYEju}*w(lrma~yX|?!e3yQeOA%PX;&BNnH?Dap>K?4)4nX)vaaJ8` zoO42&r!y~PzAD#V=f(;muyIrc2aP7?!{X4Cx0)_&JduA8>^y>M|KZN|05O|N1R6d&;Xbh4#wiVoDY%&I!7) zuDmn0y1FXnBeLig;CRBxe9%2yAI$n(`6e!Hz12(KiMSlFCTj_`DppfnDum;{e*?P@ z(PtcRo<%k-myX3Xjt6uLo7||~{}bDC=jx|uA<>-R_K)YU0L<*>x6dr?co)7~A zo2*sCP@65IzHI4sZh-l6#(IbfUo`$p_)atfGNR---SYKf=vRI?%QjT4ClsZC=+hW! zQgg81Z?5afoa+j|8%9;S6RLR2!LWYfe~4X^=4{_BUIakMomDUV31Sq^{u}hKjgq5H z!xo$QevttZ!(oK)xH=lU*wd4gTr@cUsfB1ZAed_LvKiXCaLx5E%%dd3sLJ^7b6aT| zQ*T&diG}kT%njI$=Q6#$;w-PV_KDlu*{xm^$tD(}o+0OHv) zmgrIS4+<@!6wm)b4;~EbhXuWT6F(v*tN*=Jj4`9p+U9xv6M{PupMiO|8g8QC?D2Q5AVxXbCTf*BjxdjXV2tl$@wn4 zq@pgMpH7Uq%-5u+mnhWqR!t+=rA94__+B+r6rX|+Z7({*0cDth_%IH|vMMfr^6T4_ zvr4}sd(Q7gCw~p}k+!w;YtSdSyMCdXwv~w4i0479#>@lVEd{)3M7FGXYr9|(S4{0W z6}Wx^(s-_^-g*}`PiLY6M6guR9Ryp!60?i(EQ3%3<}&=DxS91xiT(CP15p7*VDJx; z&&)Bt5($W642B_J22L8jTJy>7&ofgxqceAH%XdAzEIj?(p+2@PJz)4-*V^E=#EH^z zS=9cYv9-CSZ*x=EC13Hl;l7~N&QFl!E5>V@8`nWB(Yy92APUMx{numSuvZ7 zZUr5X7vtjUG0&?Y=}SNr;-MZh+w|6GLVdmL*09Lg8i&1RsZU#*IPcD_`KzLJp&Em; zS|mINKfChIXPs=ntvbn)knJ)|Pfd%KBSF+yA?kTd>yqZcDkBJHZB<_vsoZC0t!C(- zWxNs+@3#**dBWI(4;fnIlBm=f9aT>2S6tj{q)dJR{P=+|8o8|h8wpPJX@)}$X>BltLy1}Iv z^}#%j_onk9xKPH-wRUm z)Hdmng4G=pE5@}UkHf@SET-gdLjKFM*}C59xBeE=1nMfDBV282Tk8Wij_m0sU*dyo zxi(>;>>gnrbJ7YaOHTJEov?X-!0g-vdVf)^n;-OVr;;LOa|S5Aq>v;be_*qDFy%Fe z^yh@%4az{EoA;fi&#fW&tFqNZI+go;<`~kQb ztxNb}myPu7R#RZK?$?5bBZvt%REFoJ;QEmqeXAH=Dd^{=<{aI|>*z8$4BHh0*yRsi z=Tye2N4yH%G0Og{AApoHYU|6C9(|M)DQE7tZH5PIxN(kwzy5FMoZohntn!Gc5 zPbtsGn4d~zTvgaY5v3$9#r<2e5RUB%*`3((B>tmONP3RnuEomiUg^XY-{=PGuL9>C zh6_+v`jl;|Y`=Pg!7q#gpqpcHEyEgp>TKk3w5LA{bmO(LjO^Ter;@R~ev12t3-|?P z+$!k{niz(hM7&I4ErXy0({rWy0^JKT1e7YdNQk?^JzT>nh>v@c*v^{n%>Lbgo>$pl z-QZmn&nSyglO1X-4VL-jgwCc@L8Kytd?ERji5oX$)?504Xf-cUP<3X8C{mr8pTa^+$n)=rr?AI%# zgR<%Gt2$9WDz}1oD!ioFLGx-#2P1WX7$!Ap2zf0brT9ElfLdH0SB=-l7m24^rn6_V zSYt)-hH=PwPX5W@*7ND?tSrD8R;>!6Flcd184arq)pT8n$wb%(ft8Hdgvqt67s!OxbzEj)>{lWJAGH|^ocH$e4}f! z-}7Zt+}@YQ`HzHNlJKfk?+`E2`;)Ctbns@J=jf3+z|*5q!Uyt!OMs+w6dRpfsRbRc z^xQf~;hW0iG})caOM1Vb87?8v((+$|s5pr!_3Bk@=Tfmx{2OyVi!JqYe*x5E;R41Y zjLM~v)nB)6^gK*~QfSB(%6X-$UJt%1If4AO) zuZCiPn=z8!iVZ&XGvkr?K3|X0U*(rU#sTg^PG#K+IE@t7vZ_(T7^iAUNIo0LitY>Y z>SmKVkY~{12}(X0lraWnm^5nVA_D`kwg@T?F@pCQj^@>AnkMo1pTUw_Ta@T6oWCI? zF#oW2gFkC}ktwERCtTTmflbdUP2D8iwRhf9UXy~W-JDf{zfAEmG=9Av^ z`|zx#nG|l!fXGyvefX1>H*=FuyjdBuh;*XzPvgW6Rp^kMZ%oytkK_4Y&yNz49^+`H z!5SgH;QjwZ-sA6C8HLpUTS>dxWfN($J??$69bM_UhI~3x_v^%BaG7G}Iefwscd~C@ zcX*)dS)|B4YcpPQ!lbNq#$)bX|0ASyBZ=+a_>r>0!d{u$;QZJtB$DRx@~GDVx_*8# zCI9gG_)ghbkvZ(H*XE4r*xBWk_O+K674X%J45HQmA(6$WBHy}5?~bw8uh?R#w9N_! zC~g#VGu_S%iL?KuZmG!<)~k1kz;y+9aNFA)8|&&Jl@t@%tDu$!FjI z*Jo}!OZBga8K7lgZ+-;^GF64hPmm}aE-k$b#zWp zMBg0Ya0VD~OU<<+chwmT#$woJ`GYxGOmPI6d!6m4c?7Y&@8lLtPBEN*C{_oXFJL?* z=1&Cv*;7WffYwkd2BX}xP;iyKUcoLr=50_bd;RWT={zFN1vtbZiSzPU` z?|b5uJk1yU{s&0i2%}sie1ah@DZegNd?9vat~0hv-1W1yQ8mToF% zk-Y6~{o_}EV`cxiTMoCfJQXb~RAu$aMOja0inW?QP^DGRbN{Q-vB%@I>7imB@Ebm8 z3yA@Z7GLaCpv3Qd)4?R!p`ko&x)T2K=?VKk2JviQiufea6iixw=)7Yi;7@9OBW*9l zu`YlJZi<)oM}7H(AgX4wm#z!0C78$%8PEQEbMm?@lmREh>%SQxE@eNx=k{}B`XcMo z1U0TjSDZboN6lw(+PnPR)QHI+AC}V^ED1{byaN)f9<|l7_}{_n30Z6{G`b*Qux~Un z17OAx1-Im%(X^WoN!ZJbSFvBAa&tg!?TnWshihLTHiJ?|_Lpj;w%Zr^S7O>bpH=2k zr96m~?f^ePz`v;@m@@xmCc>JwuT5-T{N*HH03 zh{Nx3?9v}}+oIhYBK76MM?J5u-^#hFlOJ?>==QoG%>#(m&WAALSnCL7|4PXIe5NSz zb=e0)68*k2{;GgXf{=1IXDJlBs@nW)PWYS1H0E+mkohjwOAOi~h!0j%QG(|3I4#AB zOQj4wVIZnC%RB!NoGbpc8OF9{uMyWVn}pmy5P!WRsO|mbI5%q6;{g4_#D?^D7P(DK zKPzWB!qPGDP86b_x+cnHgO&vxFi&d7t^oAU3_Q0@{&N)&Q(#j zFqCaV8$&vMh*c83l(W}V+qFr&Ro6zQnWmfkzVy)2^2I9m_`$D4)K>xSQLkjN?>D)X@$e~oJCQjlkzvT1ffdY$gQJpH-nDJ9bv7Y?n!7%|Y> z_X44ezkrUfbAkr-*|lelTNb2^8h3QU_U(0SIY>qu`3i>w9=*jZD0+9C2ZiglfTSoPOH31ZmMYl?Gef6KV=h}&YCG+G2cal`U#2l zS4YT^eWq-+a(}p+rcSB(01FD`*MD0+@}D0}(Z^3r+>`z8lqK?ts%xIx8}#4?<97eD z#c>{B`;@CHq&W6i{2$=Z@%G9BzdZFLmb2Zl}&eI4bhB!wbnmguBonO=T9PNd0az zkg6>~c}Ko3q7Wwpo851x*yow8$OCV^>bG0}+-R-uOOT zfB%jkm+xQzM8y3yW2?$V#NbAnZiTuB4(x5+23=uQ^-bm{TUxEH$;#HKeY8*PjIh7b z?$V!;yW*)#&Du8#B{Py*L3$Ns_t%=ET% z{k)_fhmLDWdyvd;stw3YlNYM;vPGE_i=!Zl&M%q_B^W=U6HN|Q#eH6SHW*ffG#~Ec zvofBz2QyHA^kSE{KKPWt#oMFe`>xcS)bwY}mklc4vS#M{vJ2@wWOy?Gs%7Uc3!!{d zpjf(D$FUg({lGuIQk+Fa8$+~^M(D-Q_lkdFD$5(ZaJlkGuE<^zeucr%vwKV|kj2O| z-KIL4;E;ztFk$!TYr7Gc>>gNX;o5Qt_#EH4^8gbI_#bo@IiD1DICN$mewKc}z*R6Iz zFZj4=$q51YRz`tJ&Kfmw(ma=ARX$@tm7A=u>1{<}bCqM)DkwzJRYA+H(-QzF>9_mHlClFm@6j3A>j=sv<^ zV#Q_g=~Hovu1MU4vBplwD>AkrdxCyS#t?=pVdMXT7lxc&60aHj9K8wr5G~hl2V}ps z56g9U(-A~{)^*3bB7TP$>0}5?yrD=DN{cy`-Y1dV;r#mW|o~CaY6qF)wM{pL*}R7Xjs}nUc*;0F%GW z4nE}J5F6{?yZMFWw6SMUOfWyxU|8CeY5s7R#ll)w7{`pt=KviwX#)_da<6wFek+j> zDWS!G15#F!JusW;%BOF2FoQmP!K9hD!(L*F3lY<2%v65AnnjW(nD}#eB<#nyddB>N zJe3nZhc`HXQ%gY%Sz=4d4mzIS z?rjT-Mh9&W!0zqqk+ZnL?^$vh4;#wOGcl%aO}bL)e9-LZVvNDw)qvSXTT_qat`6zf z(JPqc_v0lote5upBC2KMB6&zc<~OX|N+hQ>VkMgTYVny_S+VkjXH}%t&dc`FcDbE| zh`I=WvM{;qPFlte^pD+EEm^ZLz==v&sl}{7H+e zLFO*ke24C29qlchGReT3d_6jTEr_}%oqd&s$4>u!BZ%M-b;W+xU*OLG?4uKMJP1qal0DBsVMqa_8Y_S0sP8b83by- zsEuo9Q&A*7k~ZHj7nusum7jJf2m6Q!DtP9w)(xT)(WQO9^b`B-o;|~r#WmiQTJIOH zil=`fABtRMH{08(T4D4pVDU8^FeDV`uqetzq~p>;EQwWGx?pQ!)C1Rf{Q}LJ`KJi0 zKwpcqler6YbwR#(R=0s=Hl+wnxX#)9sxMjs8LG3}KBR-e-z{9Ji(?I3AN-K?&xzxF z*X*(wljuJXiX)S=Jwb;T^b^^o!QYeI-a#$umIyP=aNVYp&v?EJlG3}WnWVNn<3pM^ z$fBJLyB?mhxBSH*oFV91cv5ATs{5162!JFE^lIQ6RZt~^a^YtVaaaQVGmzdkqAv$K z`c}`up{C^*Kwu#bM8}-bk+iulI0^gfT-25D7o&e~@^Sg?uDWIKCp|&5cNghdi8oZX z&1@R+DL=6fm3x!twO!!?^iyEc*Iu*gq!y|;LG1z(lYNVz?4T4k_!IW$rjMTUF0g7r z0-aCKiskaR@fK6xHYJ!T?$5sDcgFV%)4x~p3kOj4vd~pmx$z+-^n7RVkYHB_66A`l zA_pmdgkRy{Hu+kgmgCBPpkg z=7DNdzirheqr($=B+(0MN#9&?*gHC|LJZBEr#@S{VornpmvoF z#WXWn$rG}wVpiYPdTOvdxT!i^3!(kIVqg_4>xNZz4YQ8 ze)&3GAV$DdGvC$CNs1<^@z2tu0cyTiUq2l2^0Z<1)eOHGeu*2bPNcIL1AzS+ql92?D>C zDuy8v)Mw}B!k%tAmWzp~s({Qt_2A^O?333tIA%T%iNZObQ=*wwiI0}CRq>&X4d(8- z_*`ofE7f9yggd!#t6##x2rp~PLwh_&U|+0px%c`K?)0S|@jwh#0r#ld&SYh9Nq2$Q z9n+$D*j^7!V8UER2~l+pwZ~Kh0k?dSy3~{CIr0^Qg&xBv3p6MhMZy(W-&=>f3*)Ch zGRZ9k;p7&W(;_WkvyWpfEnr6ou$-U7X(YWl`8^*cB}mKkYjwa@6bXBrSMvL@36D}Q zK1L%w*tyMSR!DkpAZAmfWMRefi8n6P=kK%?3c|I#7N0KV0x(WLN`!kZD)}CBJMTGl z4`g5fe@9*c!^yap;n$GZ2u~@RloH(4M!@P5g3dV)IE*cPOcRWBA9?~`iZhpTeL2ox z-s~SJA;~3wdTQ%=NtV|UA$3FWLU7oZe(tK0wO`HgL!w%`2W@&VN|y$37`A=N?0}$I z4AM&;sbw`vkDZK{_<~cc|K&z91L=f>4+;?;Dj2U;ZWt?|lI5-AG(5FGFKr?H>+V+e zd^w9}@rIsH0C-moniss(8F_!#h6B{ zR)KpQn|v-yO^{v12taWs5~U?P3Guqe%Mz`{t~43y@jbbboP-GO+j*Q~p^f7=c~$VM zak%q;e#Hrj+4TQY$tO%pe_b4{0Ti@Be(a~cjype<8UOe>d~!>Lxn_vP^4X;<778%_ z^6Hf{&X5BLhesy;R0@wxTFy$cb5mFFFtcHKlhFgzZ(X2zZ$}-&(M`Sb$QN>FW#8>V zYbYByI5(KdV`qe>)ORvhO{0(=8heuwIOx2F_eE87Z^Ls@_Dp#JOi)YW&LILZu(q!_s%N6=(?v0!6&KggvVF!d+A zcFPB@peUJp5)_Nwe9v2z z^Xy8WIbx{XSmsu({Tl;f4RT#%0sRDs^?*2asLv+eg< zN7`1{POas##vI-KCgDEMq4aIbuM34uQz|ah-Caf7ZI7n34sEr2r=P&j^3Y#P1FcM{ z)XbzlDYE>e$LT$&JcggG8P;N0J#4=O*TJ`VVm88mm4_o<`fV;eGh=Uo_4!u7lI zB^Rhz7+)&$$(Jy>A}=!!9sq`CgpNbRo~DiCWeJti{y7oT882itSGB#X>D~$q=xK#m~*rQ z;4;GIsEnldi!4U?jAI0F?*=u_cj5vr&ywsC7})HlXi`s)nlF&ioWOli(V!p(_i`1M zYJdzG?u`*!28MxtIJbmtV1ogLb30C^U)ij*PJ`~u|LP!ovRhfw)9j7O%<35(p9RW! zO+5rs1OiTXZkYJq;_~bSbcZ9N6Hck!`IDmO?8&%66tsjOq>{cIK#jeVHN+rG>7JO- zes#(m<+7oXU6o_2SoJ_l7LZt`zbh8`*|*!E8%|I^g#>K8KPv$lPQy__ljwq7eVu0Og5&8xHn)Mxx?;wUZ#5$g-RD8NbQxCe{L2WbDs;l z9&ob}uSJ%OO>ZWRVm^98xrZLSvSeU1^_)6wQ)!9Q(Lfq6Kh-@CaY?`2w|J~n;%~nv zKH2)*PS3lmT@jMUC=CI-6~SH&aJ4oZht2j0F3_N3k7uS=sMvVEaY=QE zn6(E1%4sJiTjHF23Gi!>kI()JsVU0BbF-bCwjcp~a6?A{DjMy!LZ5w14HM+Ni;d5U zwkJg?_avqHbQ#Y%v%%^>cyc&j*xv(G_7$Xc6XjRfcs#q{>pY}~+Q$q&V-(*>|J{u# z@kODo#UWy1@o)UdOZrJEx>}5J6HF&JWUKZvJ6=J#SE-z`R!uLzrH%WEy*Vp$6e5=O zcT9ggKl^7-|7572LpL*WVGz+lP|U&>177!}>W~WixROio*A(7`D-Z7CY(*^Jqi_1+ zBH$0MJ3*~OI5NrB`lqj8$1(^eR)yx0YH4m%}KJD;1sX6O#nqgp?AKje|KA2+&ZUZTSF>VI3>e|SpP{9k&i zk5G%x@~!_QbO6w$@qFOwb8(+S<*sxCp>aLqpfFdx#KSd{=>^2QK^H=x%WE?^uZ+7v zw2ZMKwv}{^z~iPB(j5af9cMhPZ1~HvXT2^4j4>JN&Uk#NNq2J9N>1_M{L+j3q{>TK zb9uExM97_Pnl0!JdKMrHJqHNcIB{(uRAFX2>WZG+m-)rtE78o#Rh*=ppvFWY&=xjP zyqo!A4Rholqp$c@A}F`fw->e&%o@b}rNv?7xpR^D>nCf0w8Dic2u1DLm(B9kwE133 z7~qh2(6i@Rncr0u#@u|$2%h#|Hm|4apxjvnn%cxzp3{lxZ88JtZ>{~Ik@VfR+o0>P z(2y1AVbQDUC{Q#}tf4;o=l1_VAH@Q+qJOYos%!!zDy;q&Kj{Q$R8>JR^1;RZMk_9K zPWdJD6b(`p;1}O5E#J+K3|2qGT#)V_NASyB3da2i)P2xdGEX1Jy2geuyr(Cy7N2$j zR~0Ab>3i?%N0^E!Do{tl-Hlpq2$B@n4bBH?>oX>WOsJc3uAy25ln)$8gmXSNut)!v z={L&NFZB7|>>F!~OOfFl)-`)^ZEm%*mGluxDcDdH^r7QY_NLt&=J#0#;MQiyOa9Gd zdXqW|se3j-tw>poZELju@W8D{q25v(MKjhGN|ilZB^SrAdDn|Iu3YEhjgxzOSc|Gt z84)5^<`vQ8T?$f?4!|$|I$ZP9-`;o-0NoP!(vm!ZmTcF!@*0b{p{7E^)g)Dy`%7Ki zwJl!)bHG#=8pr;6{w5zTIMI{BjQ+a+1c15Tm`mF?cedr#j_Na>Ts^{;=;!Gw*;FX& zm(0#5rH0^UC7z_%eie`lvYu=6RXz6a=jQ3Fo$3J#J+LgFIKMIQa7nvIV1jT+W`pHau$(45IAM1RY|Z#&&j< zJc)sv7;cLcZ`IE>v)>y`?}|Qhvs5FnM4)55D#4QD_rrPMF)IWXx3Zl-APbHT?-d-( zp|1wp20!=^>oQxw1H$uFT3Z#DBM-x)1Dc`=Fm1%3{7}^IfW7p&==L9IjFc0(xe8;# z=dspK8%tRbhW|WKaReB-r+)AsDgmGzl)014dY&ryr>eCW#DO}vEuTm-UqnWeI#d`; zkbJMLJDs|?X8C1qK^U#b0#NN%C8vR0kzLvG`w|HTWuNkoFC3a@-v}J3-)x5W(TMX3 zN!?SK+NjccTOtYBS#Q;Ruk|w;eeS^kA{gwwTx6U3aC&0MWNqnK8T2H}pSAouE!6Tl zqg$T|0Jvyyl9ZG*&owdqP+I;Lj08DR30m5w5?0_S{b~sRQ553jL26U^O542uP|rF9 zu#0Djmkt=MyzooeJXzfzgAtP)_R^-P`>yOnqtO+9=+|y~ z(L?odd`{~H(=9i~J~BPg|75c^Jer(JVd2@)>Zy>}BFmv4?jP{+?{I(yo@ZyY*1Bjm zTjWJ;_0WJ^3mB{NY5k!OQ|#4g%boY$7;r9rdNX z$_^_YI`31c@XfyaO;FeuVI}?Ll*5VzivNC(M zAzS9cbT_AYb|;X0CT)%fYc-DuH?lSO_<#3(wFHZiL6yMlcHT?f!RNkf6&1X zYD*1=b7KSts;Ou%Z&oJ^{mP_@7p0>IS=N75(c-&v){v**+w#2Dci-^2`1?hrFJCRx z!@VZ!G~gbelKmJp2IPI)1t+e2H~Icbqktwq*rhcT|Iz$YyJVsF4G!OLkaV+PxJe1_|o|m3er>OwAjnY)|_x za6^IvXtj%yq-@)&iEK2UzRnl~ZjiUF6h_9gOlS4bXF{zJ7+9S)fcOm3;B@-s;KXGv zDKj+I6{@Ybuc~RyEwLstr$4KWcehFDwQ=r|ikz@`D zt=<)np)8#C&1S6gg&Qpm4M$;<*%X=@g$1}blN}D8_Mmtx+3q4HJp@uG0Ihm8p+WDB zzwMH~VO-1Jl_($ZETaJg!$TKyUZAH=#b4ivX5DACGgaQHjiOxE0{28U8*c@>*)kOC z@-^hYBuFeS2a<7%<;gniLgDSvBoPkF|5weQb3W^q4@7u8F z!m@7XdV+WrT?ssoa0!fg33CNZzW)yl-b<^&RdN1sv#LalaNZ8|{m$#Ke>4<$P0;=b z@3U7TD4)&5&AF0^!uuRYVfq~Z9=N_9kV?;julKtqmV#DljeI01N$;uM!EPPvmSw2m z7KfgOz-riUOyPX+X;hD0^o}j3(CMV$zMUJ=i3@!{TxCbtYhFn#Y&wZz%07f}jlvh~ z^&YvQ8;DN{QmdIIUbR`1N({e1lQxdFABDbfdEvR4`HqdE5YL-s7flI$bmG+N>EA!` zkOJmzZ?X03Cv^Q@w9mwc z6g(u}PWlP=8)f(h@Yme;+>8PK*ebA`W1}G~>R*dw*>>$4i!EW+HNLO$o_-fJX5<+~ z9{5fvEMip&MjFa4E0=$LmHI)fhNF;feIbY>7_$JDoMSh~bd+)%lOh8{iM)VnXM_o6 z&+mHgnRG5q(et z6J?YFeUZ%mQhkg|TG%G*u17@i>aMLu@X6T;_G4hXS^P-T~LE8e|w0lcQ%m8ReJLAlQo7)Xz1ptnwu&hT_9>~%F>!)i=q=7c)m#xclh}{GOE}}f5yRoCP{iGgd(LFFE}SZUgz4=9i_w{z zxQp=4fKiklIPVMd#aO|Se2MegKWO{KO8J^&St9nmAg`V&U8LVESDSgA$PVB;rzWFx zd6Cz1=kE3J0T#jfLI@_Y)xWM%xmm60!Ich=$8aICkg2${&cP~{AB>d*Q z!zRxRR=nh@9(!H9aa7im@72-=HnWG)5BKIkbz5&RtPU6x#jj9It>g8-n^}>;FEP$6 zd1V}$zbRIZ65@BN*Z?V1zwB`zF4eh<8keX8!*8jNAaWPl8knRDsVwpU{#0un6x4Q3 zPZ{>h}p3d=%>%=S6RluRaq8|WC)dKLjmKXBOUayI+=hw(=+9Z`}0CLe1 zkXQ0GW+Cm_WUCTXMTp=1H`$hmF`Y!SEeBBu!MZ<_^kxb>z(^?V;|Xcd|M{C&C@X;G z>gFeGIdjg!<$+7xJ6~Z-%^ks<&*dpZcx2v1 z`;H0e380Vt@Om&v|4v)qPEhy>bBSl@){>Ch1a)D+IHe1Hn=2V|JVXEF8|hX0d^$O( zz#B&YnuETR4ZfDYc#@&*B9=??O;^@0Uq(3HdNnERvTS?l@ze^EdozYbs7RHAYx>|5 zT7?xKD~B)n(UP#(%%(<2085c49Y$?;9wywH)$kiVSKp`5IC|GqfcxZTjDG;SfC1h5 z2psqYLSU~uU^MeLW8@lDEM00S*3nnZyDWJl={LhQ0sOPoJiWrcSO6e@Vh># z{x03Arkaz1t*4D6ySnCqsWstg3k->|9%nIr4)W0$v$3|;DXjMDjzS4;e37_>A@LFN z&GwdQG^r?v&{o!|EC-`ZPpzmlt@dk@qH+g1N1Yh0NCq>%-i7>~jNnng_LTAOXLCQG zNC>~YH|WOCW!fivG_>J&(MB{}){R&6GUT3}5F=ww?}BG$|VAVgQ*M zSnrP@60DRi%neL}Xj7qmO2p8pOIL6$`e$}qy%v2AWP)Zt!Jdd41q}0wVkD(qkTczi zs5tG1i=)IqMGQRzy zNR^enz6(dnp`zq}*iJjbg4iq5LRy;HXrph;K7B&Gjye#g&l2JL7g65?557-(_n#Ru z;L9G&&<@whmeH8?Xt}BcNe*d}Hae8mmmc8|LpV~c3o_`5tn9X*fV1Na`9A%h2OEP% z|G`wmWib2yg{c{=s>+3sbW^715vH#{Z{E0(qxxCx&eE4SiR6877)*4k2om@1i&|LW zV^gq*V>^y?379lxmUC82c1?ZEeJEY;OZ98~g#Wr`z(=jXHPk-_or=M3Oc?`>2|B^5 zSt5^9d_|&^{r}FKSE;g}`M{kdY_h_21&8ADD3Yk|RXp`ND>5Z*J@}hi5kvehYaOKb zpVFZ*EUwDcNZYYpM<#|zFfvZeZ-DSHP-iXyB1J`TNUw~&2j60qrsBLV@9AgY%+KIT zJ_No}oAQg9tPTO+XBlyqwt>@Cu^;LZj5H1Se{5n;!p0&(2sazGtRDz$$~yFz_Xk)a z7V1B#}F zdU6K9UpyTVlRJX|=ep``+r47uIpVTmea3uWm^(;euIhXgeQs*ZK7<$A9r$(WvajP@ zMH*Y1YxCpIKd!NsOgDv2kEWybS6v7K?4&CWZN80SU`h;o0q1sm>GB%SaRomXT~oXI zs+GTUdx}^g8{>l?DQ|VOBu9Ea(8sd$&Sa)BQ%41Ej|ylXjJdXT>-n!3M$#wVqWq`P zGwYz&qyH$_z)C@5fsuii;L$DS>7c_`b)zwIYen{r+}3es(Ao5fEkVco#duw_$nsn0 zW{S1WR0y^NoR3MS;}$FjzRGLXHHG7vH7tW|aLh2#mVbaEh8l*VyRB(anubdxuK+jN zEU&D9mF6p66WoI|TSuE8L65&Pza8eXY!QyY>AR#vMx~#1fM!&u+YLey+ZW9O)ODKw zK1kYEeJ7VcdUy3&^l!P^*s2NDNVkIc(fjuMKp}K2^&Rh(eg@Ja#crFGS5fVX2Ift? zGLZ{uwVuiW@w%08f?vd+r3%T2WHmmF7eiW~kr3RSSAr~kPz=dUBA0%OKaXZ{xezsf zPo=m81WUP3Jy&;YPVfkQZEsJzNXo-S(#kc_H^|zvTtoAE^MGU~^q6(k?o%<_(s~T< z$vCt4%KUuQ7F&mp}0uRex!OvLSV!{bv7mJobUFn7vD@$DNHn5 zfqi??Gq`_6FJ z{_SySa-_*yr@2`xFjg<Dif2C8j`^e7O?P9D-&(O&3x@rV?4qJDcg2%xq=R5;CX#FS2XLjOLWVPtk1g; zj5R3f>Yk|~9RrG0YYT$D@HpQ=_f6ae^z`V3#6+*NxN-%TX~3gZ{f~57UnBOM=2o^a z01wUtK{-sm31MO;61jW&thYG_LdOcoS9?-BPjt4I?Qg9^+Vj9$qHDkMBBD^~Raja{Upi zpLFyxH3J+ummEJn>x8HM8Am5R0PW z@y6#Mc#sQB&(FE2l>f|sx_S^$bd0U1&=Y!kl_qJPUK?F(^0uioo{+ps5{_H`yWWt+ z!BMb3Jr>g1QJXqsywg-VAe~tP(&a+-vH-PXb4*bGHGGBB1IR_+zAtN49d@e)=vodU<4v~zGQo2c#iSAUaq^vinmR6236A+7fPd??+e+(wh4)A+}O0z=8AX5T^M@KMopQ-Dgbz{>B)B>2O(638jaa6T$c0|Tjutsv%Zz*+~NYhu#ZiFRqc%wZ|x#r`}b$?gc+X#192DCtL>af+RQ*^J)bFGXkDKd{t;&(5T(cw>{ zCJS7Nemv^4p0bYZp2~^G0lX%3_T#I~|JjmyTr_y7s>Y79siJQ6W9Q`PiH%E|J|arm zWVuw{FQX{KXjPPB$u_#YR$YKNI@96a7o@nNnj~;w<9G+!DUy$2aVSBoB{wQ~$X~T> zl8~X!HdFZIQIZw|THkYd{EK>wN6XeTtb>#2q70kMnLxZk|!<5GiJJ|E} z=wPA1^w=x7q;~x~hk$=z%%G#PSTm9b|0IA|6qIfeu2^*mx(9lbXGV7Wo1jOGe_=#B za?Gumm-2;0w06d`?`wCFhoWN&d>y0W+PQUUqi3Rahp(+&E~s)nf0z+Pt+;ASZ2)vh zfXQ0TYykq zW|a{oy2G4w%Eoh4A&C5&PG^|)JRR!WhYq+aQKbaL&HnNM*|^~eKPd3u>|*MUz*)4# zhDXzU#*8PD>~&>lxR_yfIz6K=Bf2@_cNBLqn*L5Z&0Vu%-}Uz82XrC#54!Y5`Y&p{Uw+i|G|3%<*tJ?mrEH#J^H}9tt_z(Y? zjfZPL-l~2QfuxI7jW|nyswZ^U$^I|htKP=5{|z&A9LQ_Vy<0J&4be<5B-a{DLL4 z_#3Mg2<-g`v_vZ1xBxr#zdL(76^ z5&xhwp9`*<9>#kFHoD?3I-m?pWvf==t}#UXyrpRI!6)yr!HE9Ej;YVSGg#qG(T%W`JZI5Y$4MdbtqTRYaA;@OKxRd464esC;nbzDV~Shy$o z=7)zRg)=u5#&}+Lf}&{y86)uXt5h~v50~<2S`?d$K*Oey?G5{xCk!vH5)W*zVhyV}jb8*cGb_@UTiVbDyBmMUL2O zirThtcNLlz*v;3Kt-XX44>->?2Y+Yctvpu%5l|q52kw+a#X^OlTUPVpz_vIDs=dVI zQBLH!OGO{stlys=+vJ$Cx>S3O=){{-2E%@EpbXz^}ajr3`CF&O8BR73{uDYge zd!pl~WKZK8Ev_tY*#>CWf&ryA^68G!7uHCN|1?tZmX?$%dh&6--`Sg;l=ds_pLiXZ`JL#-Mdd%EUTtmy)rg3 zm@_vH|I*c-; z@-}`!>LT`VOdNk7#b`=kj&wmwpkrc}HXDkYWp_CB`r=gX%F2PP!_BPTGAwbCK`8v) z@>-8@IsS5Hl?D=Qk}{$~59ZB>a$?$_ha8ChxOG}S&TuIVKKp{ZmO5r64(PRfEu>;$ z!`C6oDU^auf%hE1HG^(PU2&N-L#h#4IwtI#CbBSXlrJ5TO>$HFcyDpx<_RN|#iHHW z%A@MeF`;KaZiaoUYLhI}^=cmO^7C`LbcC_kE2&*>()4;zNlSXVh#9c|L zSa_nCB$VU*ICqtn|DI^8N?S%h)!z3hupCD6VMy!+u((hvRv}`eD3m>NDdm`UnWXfm zprGYyg(nklzv~Ac;Y+aU6b|c7$QC|RY$2!BQop3Bo} zRfj;|lAWh0{gGpdk$2-e3s=a7U}p~RqnjQD%MCY@p3fUprRfK10E@Rjn;`_1X#J5S z$%r;QNDXzX-%Vyc6-<@M9)81cDTv&@4D-~li8tMzzbf9Ct!LmkSg`A&l)Q}*_)bYb z)k$YC-5OyiRew{#@$=5A^_{FS0VmBoDg}c-o35Khds^m1-$A=h=YTi>@7MLFKDm)2 zM<=DL0~j4yDq*qrLg_d28fCmORquSl%L5sRN`A$Chs0?OMG1+dqTvMOT}GzLD$jX{ z^IZ1%O%XE$3r9cw0xjxUky*T0Ui4SND8HE9#W$bHq)Medke&-4y(c~iiP0*c$i9us znM!?x)B5uxu-$VEf?-)1xUm4<+kZ_kb2oaw{X~mw@&5;}ASRv+@v9q&6+_n9s-7!j z0tIhuNgHR?TZhBe$iLfk3@&`WSfgEapD2EV&@(u!_LCfU^$fX?FpmFuH-H$>r5Y(> z8!PfJ;%=CcQJ6_&;bgB$H9rSTT3C}D$&nxxU5YHm>tuhn{>DBH(I68(7RC#KNyM2x z+FAAo+0oPVj>HvFl@jBvj`d+4xJGU z&kO_WGb6KYOx72^YUnaW4uA(5m?RKQ$%cuZxY6wXE~-ggeedDYs~YaBiWf_9PZU5x zSu=6cdGWj;1JTJioi^sGn76Xo-EHKx_&U6cpn&g*NQvV^vY`Vh3-u4Lyf@{wyw*52 zKEd;F$N|O2z!juRHsIVWalgNF+VomZSLVs|;j&rlgU*ABS##Og;)I$StPR`+^adKN zWUj}KPp@5i-k`-;@E8CY9O`THv8AI($NhJ zTD&3q)1S0Qa6Q|ga&GxZri>tm(|DK#hlroiy5zlnk3YDkW9xd#d+ntI>qC>aH0zaR zAI0JU0}fnku9M5S2BX0R;*m-z)SO*wc%i-Z(ligbfMPj%rR&|ZIE5%b?{nhv)LLq zJz2Cjd4`=MoN5#>n8IzKsOa8c?xJ*(T~m)GpBF3e(14I+uWZf<@Jy;vw%$RPWY?@2 zjjDz|eBlc$9NInBI7z^MUcjM&KDyGBTA(tk=fmS|HNn_4i&o#t{Vun(Rq+>89d)nR zQyi9OT!9bqt_k<}-jk$Flk}T$wgXx2i3h4nSf+el>D~JxQ{fO9yZy_>bROV^?H>~O zE0}f90^-p^M5F!Ln5^d?d$UPR`Kdc z-$XrH$|dLM`R>zNWgdWDS|*X3>_*57LC+p-t*BZk_KjtQkk08Z5)J;wsPV%CQyQm` zLzeAVucJ0LO1-y)EEiM^%FG}ugI8}u6@ljNO-Y4R=RG^X?!LMVhjmr(_ zL9~$^Zszs9@~Z>&3%StI)Z$iyA!8tdC+lV8WDd$ikaqizy6ZC^cL_KruQRt| z%j~7!#K>Sq$vsUJSQ}DR(2U|=%xXW?^870hnpyQcg#OZ_#RlNW54k*9AM`;~BN&&Y zl>$Va<>{NKlH#Z^ZJYE_tYYbxehzMwc6gJ5B@;GAwdOJ>bvs0B-9R)#$tDbYyJ%>k zUYo>)G-hbHj#&(6GE53w>+q7kri*fr<{9{+rqD=jQ4r>LbQMSo!bwEFqK zPrYGG!JiEEGvEl%2^|g^66v3#|Qdr(Y+LFrK_tdhU$cqrzLi zxex=mhlXsPj(+KC&LYY`KBg^*@3)!;EDr-qyC4HrR9HK(zu3o2W`UfA)19WR~j8@Q`tw=|m);>mv( zk!032%yEn0jp6+2L@P|UE^2c9(USY|HnC(BE9^Sz3JRYxhjA(K zsgLRGISlGG8TR*@L#8SlEAW^67gfn`Y@@>D!&I7Zv8SAt8 za;MI>;HGEWi+8D07|@l`rb~tEF;(lO+O`OMf}nl`L0RhyJu*^D$iCHoS1TW57&&np zJJ(1lnf#^Hr})J=P&_nec}iD`nlZ6J^C^~_aZL&o_YBDtDW8_{!_e|>0?)A!5wzIvG`%gd?kcw*b*6yXG$0l+{8H2A@P&)p;!?6C7d+RB)t}UM6UQ zmT`&b@hGrw^)hDR*P4Cz1LHeI2~7LaO3>JaL<@R<`7FNUMu0bDU$$}GwR@AD(vf03 z5tz2%dD{bX{_;q!xyXn4DVn$WVD7YU!eXq(eiJSGhim_!(*&r1Zv|(W^J73y!}`Wh zO~j49RPdPSB<;ZtF@?sO(>KhFJX;Xwcu>r{sqji|P9&@FNG~gV{d~zG(QXftQ)9mU zaq7+F^ud0C>!3V2Mv%gmM}NXspof&*mw}=Mvx(~ay`8zW^}gCl&^tY|)4_+qLb-g! zvk!J=2deNgl&A1ZYCpy}uJP_mpT>ZP&~A*Z`Y^F*gu0ei@4+cQMN;e?xKA*A!7>pD zW7`r$vi8*<@&v~HQs9r)pWx-`nmln$Wd9|!{oH$|S|3ik7Q5DTb;{XAn<>cp+S z&QnK1t#lXOY~8e<{l{@|^Q8f@!K4q!*o;e~u#6SKxOuTJr^lAGuFO{M&XsWM?zM?I z8$V1mMbuWkk~uB>9m~y8(E7K$#41S#iX{OMm#3T66q;t?o9MPYlQyI^nREsoq9aXH z7JFfa>JC5Hdd1T7-f_*)x#Y__$$UrH9?mf-V-M$KOjnXme~pF@+YdRNbR;ex5$HET#ViNg{`@A3c=i{YRcvQOB?)!1xX#sig zK>lWTD>lr;@G5T5xH`yNUgGLi8mK*c_hyE~X6{&cX0QjN9#%b< zZaLR5`at~7wj9TG^rzk4|M^M_(F@ zIw*raqujl8`6#T0EytpM+udRi2;@Ssj+Wl2u^o61TT(8KrfMB;H@i5+$PPRQoRF@D z8bVDi52Dr0qaDbOzOGDbL_5hDMNO$FCI%M5hTCAO67t)oAJyXFK2R~ zUTmnTlDFd1Tp3oveg%tR|K%z&mN|V7Rx%&$HhJx0nZ7&cIL2FkwBP3svl4gwPHzf< z-z^XZ=C03Rw1=psgIsi!vbK$E<3&&7g;i7{a58K7eB>&v!go~3Y-j9dI4);)1pyaR zs>@#F*yNEX3>-|RBRSm*nFM9`bF^!xD?=e0y|)C-);~Z2pqUTFls7)}t-|g@%a zXs#jzYl(fX3hpY0%RggQ>GC1(md7BFT*4#_l7r1P8O!o$f*f|LxR=Hni^eH{n!FVv zO)^4+wTN7bUOw-9!~Z0RFihJBMkl03Cf`CPpqS?I$5jmNzs|6!jU))j7NfB&!0hcq z=d|j{PNE3?bp?R;Zx?xt2f}U{%&-+GzRfq7&q?zlDONQ+2@&sZe}VQPc0THV4+hMx zJGx|RVa`*BA+}#(@&O-rMF%B>vsJ>Rfqt(y=jzZb>XqGan7!N5wA&Y~bwl|*QELF! zNtjGhvh`s+G!yPoid-_p{}PsmV6EQU^|R|OXQ|WtpQL*PxBij+#Zk9U6Dr10f3)re z_jdD@%5;f)gsoN%zIB-M-hc0LDA9ad0W%!stg@9ak!BdRfV!%s-8mc$adKeRiy&Ns zWv&hi9w-$Oc?i{EF~H2w^#@NifBNc|bflZyn3mozT0m8Wx(-g4oBL7T_x-KcS|YD+_0l0* z7;*aNGEainrpESY{hdN@ifV^ntll(I0#NPPz!EaT+tc`IQmBe*h9C`U1j?ul2D*KJ z{&s%aAn44v6Mjo*e9Nq2kKQcV0E%tB+5viX(D6Txiv~%*Q=^#uN&o5C;foxpO2f&I zr#LUMVlS+-5*0dcZp7Ru`5tgi&0a)pN?>CqP9Rq#z^zWik~W~jp<1XtT?pS*%c&Ew zPw!gSqA{PX>Th|Yiudy;%2@@Hak6I~gB;XGC@)NVF3I|J%sM-zo8NdyJ`V<6-rWf- zXS)UwRa;yiR=YgmM`eqLXcgAo1=_Ua3~q=41!U#2mOPsVJg;H}!#r64C7!=Mgw-b3 zC9}C{BCsCo+L8X@JKkfRX^#%Rx}exbL`q1BB`!RIVVs!J`D&={c{@*LiK4F=G5Q(> zmLwgG-tG)`YH3a}stnekdCX}9TekMx_+VRj2o)SVk8JtkxtFSq98QtRdk`gN8gms> zQIhR_)kYk3IKW~MK^@=bl%-xbs8Cb{fe)57hPP4x>h1vM+vL%Ej>GOvmRxO1jiOaj zMOQ2YxXELlZ)eNQvij5R>0S$K2&B$bl|)r=RjC`riR_jBagF6Up+#f6+eOFH%YPGA z-@mSZ&KHVPIu!FoH#J~|6d)Q-4czW700oRYl+Q)t>ec6}6erjN*q4I-*QfiH5447Y zrb!DbpNyX#*tA->yME@+%)FRa^U>RGEkh2^63N**?9t(s1mskUoUtq+n8B7RlFVF; znLj{Jgm!^QGOFb^G*eOy1g6s}hiHmoni@Qjr9)!jiRtuBK>B7iWkXnuJo!!XAZUAO z@FfIMGjJ*LaHH1#q|*Xvk)=}9v8c3YDL1>&*LTF*gl94ZM$z!68X`P_KbqSS-49C8 zS+1+Lm_Ffk6D#erkg+*oUz(pgfc_+-ZZj^u*(~L*LlF6N4AVq;Zu;BNz6?CPPhCK~ z=TiLoK7uEB-n$k*hDd?*VLYTF?)NWQj+QroA}Lwr~d}i;RQZI5Z(ua)h zMGi+|$om3fnErd$+u*TWa@5Pnp5vYF!0j29Fl?pz%fD1MkKA9=9!aZl7yj-#cjGz=LCRTc}Y}n)Aa`lRnd!jy zwgnc*PbKuV3QY+v=yIWs*x??Hkgj*3n05JAa=cfkk<>I6+K)anmk&F01voQ9VxXt>*|G}HWn}jPspnY; z#aimeYLD7>nym(}N%RVFknA`3&^ehG6N7ZBu&sxiJd;7h+KwEaU(JDkim?6%5yf(I zZ`9$BSJPrWf2Ei-1OJr2@15T9+MW5P)%0NlN0@h8J*Y6{=xdr>H|G_2(9{(YEf=}d zqy2r6mp3-(r?DkAMhwHw*nIkFV=2<(z^ld~Jfx(b02XDioq7$$+b?b7<7tk6E^EJ7 zZ4@2_zm%a{yzY$eP2QYPfvbx(nJYI{Z_DRl3L(YKZj8pC=!D~S12T1g>J z725Qj3MLir`B0*H+C>iQ1&N9cxxx8e@1N@H z*MOo0URYMUi4@mc83pl<9xDmpu$ON})7V93lrzR;yq&R*msUx^k4>Ab>M^lYxKnHI9asmrN; z-C5@H9))FUX{}^(O;2RdjC`wj_Fw*>SIBAE^$o2eKAf}D$WK2AFo($QJWt8J03)L{ z@;*;NM7&3{EbBh-*w5A=|M~F(6g2z8O~Oxw^)L~){0jR=g&7c=HlUBoyr&Z*`Qeht zxfhpWOI;tnBKA4j4S1otx}-eWDweJ@--~@KK4%?sopy-<9AS^yNawrlJ%6$1DM8js zZ#q)*^7=E7AAf`471j&^BPs4IU~E;NBrNWqWFw|py60N^fCY16e|v6bm;>)1i=! z5Qpga3Ra^gw+*%5#8HR^zEk2WX8Cxd@}y#Qh3BHk*};m>gLnXt?;5JNgqnGQRB)c= z%I`_a_LfZXz`6z;^|P$MbrebZZ9R&H+5|nZJ=m07f?sG``U+IT7HyDo;8tgHE;IxF zz3X4fdVhRY)5j`MzYSYM$uA_>o)AQf*ZoZvx6|#7G&fm+Sq@y6AgoT$2}lEt`_2jOQd59?jkmjE)f`4Yik{pz1_BDY7|exZ_mVtc<@G=Ur5gW& z1hq+ak$af)W71Mz1naV(CX%ODFXl3X0Kh>>&`_oZSL5btt^hJjV@3siau;9j>tH5d)N8jZQolSvI8?enx(= z^Et)M^&N{p@AYQw?b*b4|U#HGZpf;l%nWBEB|gP6MgM~^nfab_cLg2s26 zGK<}0wwF;*vfvc*Z2#tW-1czcz;2UT)SJn@gu0)4N%)k{{t#9zmuRHw>hcCXt!lwy zbS>)e<*M@~?apJw&cl*vOS^2?G8_U6SC^kRn8xQtIa+8;k1drIs|ogtZgqSE7tYTD zZ4ehi@ss`D4P{bM?a8ixb5^Ts((S%zb;c{5iU5Pec?4JtEK3A81)-OFa7J2td0cjL z@tE(}z=JEh^molqw1df>#kZ#;Oa=C{1RwoNB_8?rtDp2kl_zDWv+dZ=FI+RHUy2su zOQFgbZX~1MRxOPiGEIJ8oTj+?ZBeiw?v{0vVdIOV<;p7smk5nl6~FS@AmTLsjuLhB z{7aSzv*SK%T;PAoP=AX=*OG=Bf&=?|?u)?t+8?nAstbZu$D{F8(c$QL3L(dJT~}eQ zKFt7`n==NQYDQmM-Dl3A{9NSdVMqcGAJ-hMj(_le@KUdUm(-(q!6J>z>bgSk?}41K z7@ig=kYd}kHho)*Fy|368uj@&&5JfLzGQH#>|YwJG3|s-{9KQ?3A(h=z#aL*gfhT^ zn#xa)Paj@Dmm(emZUOCQul^t>umh-jwYSfoJgfOIB5)+z_8tf>douIJU`S&19;_SA zx|2Ata-j#Hc>Gu2QKI1@=B#Gc*|RIGNz8`24xCBi4{+>QI5;#1_qx)~?)A3p0j$C|1y zf@Qy|s7b!22b<}x+7?p1vfmbNPYo&wBUudiYUJ{;FPmB9x=7!Zr=*^cZU*75O#3C& z`|3BztfDh%_fuyS-y&1cYfQ_yl9Sgh4|F2ktG35b7Bl7ix%VX(KqcX5?$ojPx-Nu= zktaT*5Wbi@0MhV#n~R!`JbV$Y{us7aldc!wJ+UXs2VNCAKogxLUtU(m)j@@t`5B!C zJVB1eBRp0ghE~YuC35d;dhZPA7d_|^n)VEz;-S3>!i$_N1%F!!?SbYPAoXZi=#CRTRamq%ME5VSmXk|*BS(A_fL*_h2JRd$)+1I$B70{%iJbI-_ z;iFN|!f?qj8<(m8b!M%5&gv|t>WE_i=5cE^TXB_`VB!KIBjxE70VQ zjR|dPe#B)FmHrZ@w6kkL(u?*x7)&dEf-YKT)iWuT2)yZLJVC3)pQ#t0KTHwvo_5S(oVCS`Pii0w1mDoqw4bjM3?}sIn!SRsgU@((ZY&3)bIQTqSQzqGd)L^4 zNAR+e{JfGtmPW+AoXIOmOKRozh1=GuuP&Ns^Fzmc({m)osvJwXyA(W)OP!`}LAsOD z)xP#{5Tfp#iC%LlfCXGE(Dr<4Bo=N%un4ly)G{kZSXM<84~;18sScfO)0QEgKiD*h z)D+~+Q^CJB{aW3&jrt)2RhoXLCl=Wt+=+9VP$Gt^#r+u7YI}SUHhOs_`t;;!Xu!>; zY-`46Gy@7dl(ZCR&fWNYmZ%bm=UgQQ`4_0hi@rRT*+QkszJ#<6H6POxSg_&peWe`W z#+MZnDEJVj(f_V|;D=i*T))n*1@AS>_&0h4?MgWmIA)~fTt6@n*h|)(No?%6P`~cx zu{5}#rJmhx!{S*n?k;)!h(kU~lu2MSbAtZ1eaOD>d0pSweU){7qo;&oRoU;3{6X## zdB5kX)Eo9DRAb_47e&D`$&bhxu%^M+(AefMit0DeD1KwS)nJiWX~ic-J_yJWdRTmO zDpsj*J`^VV>(tk>2l2oXQhhM)gKh0{)@{PFeijUavu3dgZ2c#32z9^>7Tg+T$LoO6 z6Y5EeyV}02;teXS?V@m2$p0Ym);o^eg={*^H*_=gbT*oC_b-4b4MN-7d3Wy=#2M;V zM8tJ4!~e`&Xt-QjLnt12RNX%%GqClcCN0KxO-l9ctNYT=BoLPxmSTpoS+9z18J$vM z>R9aW`K5&D*2Q23t0LljxeEAR51qZqwyqG9oOafm?A|TKA-hQ?#ryZ$&Npk?yM`q>)R3j+R zd`-boqnlDJ0!qvL9Wr!qwKEsOYfd;anYjn4cVR~WEhP>4=_qze{SeIth2JkTpC*;q zWjPgik)1*tDq*KNCU@{AX36ZLgRBnp`F8D1`$!CGBvA9ZD~QE5>jE8}qLjbQ-w!|g z)=c|oeSMZFDmr&Vd-iipgx^=*p6{PuQw=}J8NZ4}ZgZd^A=I*>$@_+4B9i8PoWJsW zF1M9I!>d>@1E!SoqRCwzJoJ{cXL5#wpnY1S1|KXnBHegt=ze-ob;rmCE<2U{6&mH` z5qq!vr2yQCkg6{vTtRA*II#OAb7iaE#bCAg$%x1FT+uY|;~5JK4-ogBIlL0UDA^0h z>%$FKGJa&SF>?trE=DHGgu7NuBg;If8hQM5Hbc)oSMACM$?8mFrf+SN$pP7$DNfYJ_@!l^L&Wb< zmYm`vkqpG0k;=^BJOUYv?oRkPw+84!ws7m_sym+NK&l_t!XmHJ*M~5Pw+ut6o>y6U z>i~kJr!`g=8VxtqN?dADX)E(8huGYBI32=Um@(`KPNb34^kU#fd;25uZ+`=StMg;uGp zTi3RN)~$oP%s`^su|Pa@7(TW3L>rUx1ZfK3gWr)I{x;F>psH+OW3^!$sh^>>Ruy&^ zb~LvjhLQSv2|^+T?$`E_cXhxv2!AucAL=i-U1v^F#fC!=<&pgXpaRWLFTPQn$pl<; z#i+uGw-<9|eu@u&B3`E+Xuj8U{C`;hyb&$4{>ns}dr~1QG_nEkX~nmWnnxGDTu{Lb_M;V_!$CwjJTJH@LUK9MfPF zn8$d4bcss3A>#CZ*n0DSDE#niI9ozYk##W1R#}GB*ayiLrBWf;mr+L67|W1#vTqf_ zjHRe#-?w2bgTf4wu`eTr(2RAA?fHK0=e}Rh^TYi=ykF;iu5+F19NgKDQ6-njfaBDd z>4^ve!g3C|XJo0wn~*wg`2LV5=zFW_iBs|u%TqXu*HJ}iw;FE=NsW(7M8`c?g7*|L zhQ1fRGi;Q>9>70D#|a8A%{xg0zd|Oexi$E=`%G3Z+Am6MO?7`_@!(lhRCdv5DDS8b z$v#KuR}3$fB-^YPC9c%Ymfh6&gS#wIJdCF}tj(tK1l_ z^A4ZUy1%tI<`<`e1fXnX0!q?A=4Y=7cxL#qF@$xJt>$Du+Qs;RslYDl_+3w`>;n{= zIAs`^%=7tOi3;EsfrJ)@-fvuyw#G&+ zode^)AU;*UYAO(*+nz49A&m{Y!CwxsuIp0$NXOa~gqV~qrJYE{V0YKV52OP8J)Dgm zp8hu8Q_w8}2Zfv*tgK{f1GyOfm9qX5I&G$RCqTjeKXr&HN!GwN>b6RRz_7yiF8)ml z5%uZg#+qfwjlv50I`psr6xxAbY#+36zQbz*@*dn9PO`5hs@C`Q)K4uHvs7!=4hNE3 z5Aoj@YgX2@2X3n_6@_mZJFGs{q4`dnP|pW~g{0H$aI}@JQ{R@6V|0Thq_xj2m|CV{brWT-`^oc@STW3Wa*VGg9R&{}> zMe&^ZUU}lp6w=M7y~Xb!blsGJNg*47YWl1CCHQ%ovcEwfYN6<32K8mvY!;(e=R{ex zLw*ql9Hd8qkBhc5+mBp-{DA!(twGQ?1KTyHs88^p4Q@jL7|g){tl0A~h`7v(6!{r? z4W4N3JKC;pXS3CXd)_14Dg42|*sJwWCRT$|$OOc`m&fk*=@1o~X}O9cU&9<-!fJ_s z*f>wbVzCngc61mtn{eYkZT{dJ7emhihdB|Q`Io)ot&F9s#V9Q1+&>Bx0>MPP#v7Vw z=fKq>piuTn7sEN+?}0k%ib~R`0N%l9^RJzQ@5F5YPk_161l!D0`q}-3?*K+w*z4TU zYazb(2z8@Rfx+IM-=FOK$@luaH{?GsOKQB`yH-eWBh}O5>tP<@<>2d+8TvrivJ!Iw zJ>>@@v@AThopE^NeIehk*~(M;hkqZ>VE9f)G08mOL*7UB14!6(r@B4{w2q6@db5tYUHc!=3s=c9oOXD zj1AeUK`jbD*64Ons;_dE z&$ej2<75}luG|hm7qyHe%bW~bJhyym-C_h@J{Jrz%4(t5ncX)+q*{=kd)&X;OV(V} z(^*npdApC4JP+9-!XKJlR@6*vyv^gl+r~rEfeqd=(j;&mp8&wciVTaeh59@ z9fxTL2~X2X9QPY`v zfAy{CpLb4I|{u+!6QyE5gzukA;Fz~%>udTe2n46{932(zHX z|L}J3ZP&S}bN?@QHxgmOx%SlQT2@|H;~n7WxE?u+=C9?7t+Oz9u=4(rq)XW5{ji2u zhCX9#KPOt|S`L!wV%0Cu8pzGB4x+ zA}w;X(FG3|o3_@W&nTQi(r=mFB3|hPtqwvWzIzFryRAKHc(JwFQ1(d5D$>g9Of$ea zF5(OcmSr5eWAe&G8r$~uYU7@T9O88<+bQB(hu}S{W&tzP2l9pAz)aOaFK8rl%=@dE z^J{z{DvnQhr0TTf4j7yE=*IBr)$06ww1Ztc&qn_ScIkE@XGomd0qu%}L_4d|)|D4k z2iI`5m%k~9V6&ZZq}%PybBDA0#pV6iu0^IvspTM!XubcAPNnu9rJu9Zc=#_Ap4fRu zQ)qzzqc50ilSWW3lZCBP!{+n2a|#@9m+qt+?z_aZ6T~J8qK9NsSQav3{Z)j;tQV<<>+X| zoOIo1QXQB8bVl}2Xh36tJ@1cX_|sH29ZLSczC9R>Y@QV?FgpZjvC4ODV63Q6G)}$8 zPN;DB-7gao&Hfv~3hNLZ3;U&^WRZ#;$L&}s<)dvSa<|smEB{3!tj-n`gr9sfDUTah7g(sVdzPQRC=S5@97ZQ&P94b9G?FjH5K9v zYa=?2%8MSYXRt|<2@{U9OZo$XJGbqR*iSq6fi0m*VQru77oupffGsjp`;N2L!r0NL zq#h90WF-&GllHgI&fy-E7Au(I;F_#TnEX|D247O3S4==|`dPdRD9A559WBHb!5H7? zXH){x@~YIcNat^YvqI?4oFCX`?Q|xqGw%)UtbQy@!-{(&L55@&(1mKL zP_S{#p}8fx8}&{W^be6(ZM?yn+usi-M1yu}eh9|8RCNDbT^T*po?AghotgRU!SL>w zPRk?yd8S2XpWw$_T^3wvaw~-=d5YpiaROpT^6idcFs@}A$m^G21kS?1R(fi^CIpXS zz(rHgRrkg>JuGs|%AsS9e;E$roI%v|TuK^Knfp*Y84yYwo4ploIX89@ljNe<79r@<1X>`Ap}t{dn8HaWGpk}g)?@(qNl8`47_{5PLwT~ww@-t3RmTFi|vYf`}E-)Bm}pzy)uj_`XM($ zia|H!oD^vBT2`%0n$;ukb5ZC8I2o@cds4aDjd@DH#%^;daIcibF6)T{6X*!NxD@2v zy{)vqj$Wa=0oOI#7%%*)&R};6z7zre^7ymq)2>W+tNX912QSJypSyRIVRAnnD%_)u z&m#=^O5K<|Usqt9a^h7OWc=#Y_B}NGXibo3?d?UlM(0P{#lNyT2mJ%6g+T#QUOjq& zAL#==5W}~=zNBuozm(vl)Z80|cwpD(B^0it4pNp0T7B$OkvreCg3dv9MRLdI>g$P*q<7jl3wd_0orzCs!#~}6s}9baBMxt|ZOx}8E?pSK zvDyN^{^qA{=o%e6sqY6&oO>SqpMrkwJ<{uc>F03PayDjZQE4w`6Wlx9Mp)zz=9l9< zdTo%%zs%PoFkRXPm6n}P&j=_)l!Z(RM4LFUzT!QPL(f~xQ_ycNXh$^t75KdSnFq@j zKg_G!B5OQeqOrMPt^2J!`oW0g1^x?*wp5c7W`~`-v+@@fGtCgb#_6nw$6!{jH1{&5 zVpdCAt#i8lVTXGg{iq~`9R-H*O{plT>2=oH$%P5#Z_z)dJY6y`{7TWyN@Nts`kIQ& zn?R0(<*dYxGw#{o8l-xbGXg5Rr|xobBRUL;g*SJDvzR?=U0Sj>Za1dp^c%&dq8rxP$tUh4wEyul=Oo(H2ki*WvN)p7Ui74d}# z^Zn3_M^~g*JJ{t(5Szk~UB5Zq|y9Uf9+D~VbAsW(uDR-SNReNw5u?tgUu!UeXjgf+-kz0ui?iT0U zC#yjmnM>npspqaw));J&%v@6@zi{!vLaT3ad~Nps4zSwIxY3*Tsfs-(p1*b8665kh zA`+?ISHxqzV{(6=isz9o-o)K0ar2HQVLypa8FnocDRWh=uI}P-Z*taZCWu{`E~DSGq5K zSXHyyxTtMq(sEJrnfDz++H<^vfp*U}qa0gv8(hM0)JKQZ@2!esiIEYs!4x4Qyu#Gh zv>MF(=vsE-U+;X6(9%+S_K-)lWhALv$p=?S&qTk4Y0+i!*y13p!+6aTx(4iZog{mR zG|N=41hNe7>-KS*o&rN$Y*?qEA)Bvo|wv z06Wp;3jy;}p^T^Poq|LbU-O&-KCgRC_ z=32@tfTLB+^hAoE^a(?B6S^-C3;^haejn7WroXt=N-N#T2Wi|C=y&5$0T*~Y#|bDV zDen;XpWD!P&BDj4T>LWAk^pXj%kw+zVDe9-Ukt&2zNSb>PprP^KJDUMq|8xXodqF# zc0x*<0>&?&RZAucyufj{5>yTZm6>!1M!~o@fCbn@ zjF9r=Q7m+g9gOfEGNdhvpS;Z6NyHz5s-i_fF^)V@r9tR{W##=DKFP|c|C*~8qb+?hLc2|s zm&KcWg}S&sw*d_A-#XtNex98Ge6+h>YwhaH1z=nnO(jfByLFjvrh%nZCWeiS&f|lIUtTP6^HO1BDNCw>-d}}L#XImTD9II+Z?KbQQ zbu>|||5SdmG`4y0nzwdWHb4;@*?z>&4eVwZ%n_=JHqdI(v-2_Kf@#Fg46RYE`Ei|{ z6=2MtImQzZcKwbvLj6~Svd#jQ`<9Rz{M<hLt*v*TX`90X43ofjPp|-n?t0G zUf#U5u56XJKo&o||Ofl!58NE`Z z`!rSdOixO_dA`C6T)QqwTT9>)SMl)U-p+e^-x)M1Fj{6}SSH zGW|0wl@ypdOOr&aQ-N~~>ht#WYhGAutJ?fT)f`b?wuo;rA>+Kaj+R1pPUaR?pZHSxCe7&}UQo=Mp8r?+?5K)fvUyVVO4>X8?ZJIy+s zaM_B#GFtEclB(oAhJQq=Q$9!kEQIF$6+BH>fS7~};CKJY!lYi(-@AMUSrr3=jRcXN z6GGYFB%7k&pZ+xvcQ#zSii>T72}JFf2ZWkBI5O*mK4YyqFf**3P_!f93CF~2RfUPG znhwBq#_|3fCb%Rd2dN~%XXp6hxf+ENKu@DA5Rj10fJ*^Hl4Bqx1g|LV4d;{D#}#}2DDm_3=gvuav>=2p6@@=H*Z09O+LA3Ceiy_D5Nf}>PXd5k zaa~h*g{NPCx>fi8*Yo7ZM2vNSdL%c^$Y`A$H4yl+^_x+2yCep4(IRCI5}s}8c`iLBd=#@%IiBJdn$C6X z-C_Ru(8cEOD9k5bs7uQ&P3f8UEm=lC3`?wIhLOs)4Uj>2nc+rC&~ABG_V-MMh4y;q zf0Zk;-Dek>_V8}2Q%G{;v9R1uzrZCemyE%m@EX|UC_C`cctKkf_kxiI2^T44Q|Eg!jyL-wyp&j~!D~)J#5j{@g{;kqiFE&Oj#~y4{a-s|fY> z`+PX@a9bwKGQ9Uyk{GJ0%Whanaq}tlO>%J3#hnQW&7oM2IZ+O-Fx%;&3K4?c+X$l^ z^F`Li`C)UTpRP-Gc8M7Yqyd$orzal+bLCtLqK=f0Xt4DseLP0h9;SPD37f@$4LaLV zcDv`dZ1SHXbPx@X5Fr0dqjLpGh07!>H5`~(QW+aw;UPvCyGJU7ega*!`Yp$jHpAJ7 zLc6wnMXmE!>p$W4($Sxs$zy6>Xu98KnAfrh=0kFRc3EXvxZsj7*oT)*qoI zM?S{UDCXqmG48)!EcLylys!<*OyFGJQRkeHM9l$@LXbw2=gvsW0kM66x^6Sa`2GBF zI8$`u=aAjVt^2eJv5!g;nS)x&i6M~vC4G7zg|B#t_`7jCL90g-e!S5zhd!<{_bCgo z3oSOUQV*?*k(@qGF{TyMZ-%^)%<1g?zoCALC^NIA-Y5ObD_1W_CKBR<)0@-Pc4z#o zF7w?nWxnLeIu!X>uj&o>bFg#JAL9&f}qR$K4XYfP5N6)d*trW6)d-WE9pWHDhAGh zUFzzDHi@R&^+$BycY2wYoxBjOD%5G&xnB`}>@gYJ6HqTv@Ie$eR@l_c3ud@-(VZ<- zy=HSE*AEOv{eq*!68nc{=iQtmkEl_az<@!!y1IVp7im(sMd znXU_-H-z+E8chB@D_j=NiD|fJfxMeeVn4R}Ucy&@WwnGQptR$H*M#~B@fSF6^{1YI zm2k^aM@3bU_;D4ViRd#*SCsr(QNLZHkBOSLP&?!t6e~ zdi~4mygKSD?-MR@G5W)uqHx%ri6P`0z)H(~SWp{b@tgTgk{dZ%9zE6`##sND z7rfjD0v`G=13y4c6r}Jikl`y+cYmtv|IgdwfqYXq9=i`T+9-MmqVZ|3Og0)?sDEA% zZ;uNM!_j+nfO;9)se_S^=PukltIbuy@L3k58C$M>9gWRb{oe1=IzN@B75d(1+O6?@?22@J^4>4v2kFUjih0Xx{V3-Be*~ujb5S0w9#Q~s> zUdgcbqe4=bC<>O4jrRb4SXSnGyA#3Fj)K^Xg4#MKZ^_FgDI*v3?03%gqH&@+*rs`O zFl9R}EIxY%cMd}7;Ex?NVn_y7lbHm@n^I}jY|wahv1KP_gorA9uCA&tfN~Z2nM-X{ zy^V-cu~A9^ny;5NKGODIO>7H(x_(FXfIVO{KRh5F<)Y65fJ3Hk>pL3J1nrmgb_BRdwnAv{^P@U(jl;_{9$EX+Mc58#p#|JMdT(3rP9KABXmiC{(4?N zWLs#er_@jD1R(=w6=1q2a@*I^xSnJZ40BS6m*zL-p)tg~qczdvAd>|#O z+;giDns}ftJ|GvGtv-+oU~gQQE4Ixl>`Gp8L(>56KjKKh@H;U22apn$RG$V;ehj@?*{HTD^rRo0(s#uN4~agadW$>pK3b7{~f ztlaju6d9vuky=6xF2oP!ndA|%gC+Q)Li4B{#yd(-tlByVeYjTRaigDOrc2hQvjt(89y3tcc?_imB>=maviKfq4-BPWV zhE>^!GFwZ&!Y7hcg-ym*_*$$xy0DuqKAbM3 zb+DA-=hzwh#ibe}lEz9*(6!@bRw2zazxMcDvno}!HTr>9=Ex+A#5L7jb2|6ND1nUq z+*RCnC!l)qr&IY-OgZg}W94N>+?*v`w>|AfD)bk0$uWzKv}fJ_a^%oMethb7D={!Ca6c-MEl81B7oxN-DP z#cD0#MEV|FpZRVIythF~4zfTwGlzFbxCOWLH9s40_bFi_md)>mOUSTO(WJ4~gkWBV zSc;3QV5K3~5dD);R&Z>&4`&VM!5C~bJ}fD4S2`D-cz^iTW*1Wc|Nf;}vf2!}_Cv`o?}(AyCL0-lt^50x<0SJPR}qlFLr zRLw5Inm_T8S?8z^=1~OyWs&x^mh$|IVy&I$wb{}0y@H)x^uRC9e0*n*xLXHR@DV-w zH-s#(27|r1+~Gg5bY*3($;q(>=+0o~x+`c0QNOw~K zE}zh%Qas18%RQ#ky6IWTT-1$ObK4TCWc6jup(=%`HFZ>w3Gt zI;I69C$#M;_-soA^XzOroRFT;(6UbmNn(S!?zWBX%QGf8bV@qObG?M*-azz_IzbAivc3U|;aHkgz2 z)lNuDJjHLlvtQ*#*_LW%BZ{E(2!w>87F7~!eZEXh7b9cL(5fh*L;ChTXi#EM|Fq&P zVmTXDtc9-*aG$&o9SFAhmU8yv;>#nd!P@)wzf%p8j*^?Iy;P&)Uu{_;DWNviojJ#G za!cG0Za~-B;Z%<*70An>KOqC9<*U)~@?l>J*Xc4+Z7X~YQQVPqo-=P)XT4gzdH6n3 zsIfx(BV6x95yG2UsiPdXhp3dGxZ2++^(ueuW}2jPRAp4ZxL))8UAQIc)gTLF+$$ha zV0t&stCe4FRsXz9%8mO;)`+6P`~6WyO{aH9 zHC(bNyt~1eX7qGVDm?Q2q97kpz;to0)kS6GmRN>mYd1FW)=khfe z&U`N-U-$8a)hYN|zN7=cF=tqvkAhN{DXvJ@x;^qmM9vB48teDSXE0nL^ncLah;7K7 zgZB{fHkvQ5Zl3Lw_l(uKSN&@jesNm0aj*kwr7|MnUNhekFr3-H_(+|^#VVH>peNfZ z_99GTvO%%a4oSipojDoO{AU*Au(E%kpuy?Z4;C)9`Vw80q(bO(lVSxE%>4`Jz@ZCV z+u~@2ZWG=e_CPCKpH7G@=)6MW)KSLHVjhyBZ?EKP`robQIzctZK_(6HpSz{gcrVEd z_->u0!avwq{(cP;3X?+f6x7L()!Pmz;S#AR6|KCEY9M?vVZ3T9FS{DpyCvVz6#xN;Ey`Sf|ldn}zBvzqbG zdDLA~(Pphz{e()P4&Fru^JPOPmZJnTqH{14x5P>O&5Z5y0EUJF&14e~vWZ}8aRjY@ ziu@2+4nYQ(gDBSC#w){1D&P2cw(<0utv9hnk3egcvOAYB2S@H{V>Ppo`iY(}`fUhB z`0#RO%vtA@Fa_Vrv=;uYs%v|AMY@Dq_^PS>M@VFH_15}I-tet}>y&7p{l59I;3_r8 zT(vG@oq>pzNVvd?{HKBohmvO^eiOp>YMLm{$x6K*9ypET7rlYcvX)Q|EEWL^ zxvywv03=G(>{Iw?xR6G}Ioa60Fo#&x<{p+{%FBvy!p;V^c~g_bJ*(i|8ADLTP+w1G zu@CCOoK=WNo8g+zI($>3QohgI&y^xq2^a0*7w?CaC!CN<=#ip~`4&pXMm1MX*yJ;k zKhcM))Bcn(z7kp-*0}9KhY)O6ZqNP{u<9CHY(E%4^o5aPm38Mrxu#-D4e;=>R%5}V zb5k!w$t=qgd4vf)XL-%Q-?G|YquhF|ZkpeO48DNpA=k`p2(>Va^+7s>hq_2Lvpvz% zgt)%iT$!7Q5i-(}v%AdIBsW>{nUFUQJ7g?KIDD^RZ+0==ksHbYAzt7}f~ag+x{5|I z-$^E5k#9iQ>D6A2XJHg%YB=_I_Fm=z_|eylGDc!@i2LK|uF`;b2ESj*;X3h%99}Ep z$75i-okbrp4aaA5-`h(vs!h6BEh6b{&80)>xsCl*$IW~E0v6R{7<)uOK$3S+fmZo{ zRf6c=K9z!(+K8S1x*o&jVB6DKCQo4-o#S;O{=Mdp=NFC5%g2=MfTakZx!lcc7Oyo5 zUkc|5+=*u6bQ27ZQ>Dh~Y&f2g172U|Z(2Mc8Sji_DJQD}^B6@r%Y4q5wk*P+wq2O< zQy;lzJJvg7B_}_u{$^*&mieM2^hce>&QeWoms<{`E&7dl?P?rm6nxU32)=>p`X;OP zUAgT5$c=)n_QDts`b_Ub77aKKFlsT=)CQ9mqw#r{d z@SvwhOEsAu2hSPCzugyaO-1l){$*2~Tf1o_%taZlntpe$^FHl)NMlgd*iT#kXw^L& z0CDZN+T8|U-=?LHFQST59v>_z1PCdu8oJL9yY}iL4fqKBEvXD1D2u7Qclz^L z4#P+vv{?`V_)pnMxX`!~J8!_fxnh^R8n(8n@xLxs{~H4kS?A70U-Ci5{;N$vNIPj0n)X$w+rWVe1|O6p-pI|HSRs{-u{hJ~a7$=8;p75J!?j!>4$& zNiHk4OemKN+ZrUA_zQE7(2G}PV#B+9c!OBh#mMuX;b`wRbx87xDF0*{UYY7(b(Oc) z=2zMSQ#RfFRMF0D*qfT#}p!uSnc=ZJUk+T`usE zj8<(Nim{m(3qfaS!~0U^yx-0-9;f`XKza55YbmLvmWY4P9rkt8Vfep#P8D2#BPP}> zL^kbvQ|=^BHKM&dPl*!UjS$2vET9+HD@pkP>{HCTy)FAk#8UeKScm<&-cOa?^jjZy zS>~40Agl`B)vuj1w1NLKW#b#R91PFCP6WM9o3v_> z`Lsi)=}R#jr#A2G>mGC`uV$Yvf!dB7C4Ra|tRGy#3Reh7Pc(;xnhM#zp11_c9m?A+ zowuMhg&@&@4?2|(j6(K|0FI^&wnj!``=oo#Z`u90+g~JiRAh1{$~o`vZA|&eXM-K+ zv|bP)DR8?iDB^@pV??E3MOgg*sAv)~x4kGir}BHZ{w?yyPNp*1`qKmU)@!m*JaTq! z3pMhKiv~6~)&Rf9>!MW$L|p=2iE38#_$gnSb~GDbEZKrix zv^?TFb^mUS_+%N-H(qShSi=Ju zv%*R~a-|0TTk!JbfT6I&hqbLFXNlj;udE6a2ui;MQijLbh54)P_ivX>*i;uAWtADU zZH^78!#ym4yQ?{;KX~TvjUVaEb!Wy8ugcpd^z)#)1}?(aME*h?2F#C`W_nRj(&Ue* zlkt_ct7}g{6^L(nC;Vi6TCUhlcE!(w^Dxg53o$O?~Is^ZvJr^Oe*h> zpM6$1r442G!WUtwWgX)%BScGyc3WD*;ihn4XlkZ+tNS$pYWk*+>)G}AjUV`J1$LAcB2gjkBPNhI!=_O6dfGe1zBlFvAF5dO zkFJJRK*85{gmu<)?oY&=)J3%renYzfN_Z~29vGuG&Rw)i#hT{2-f07(-XbM$22eHid>}b*AG#OaXg z*^!QVg`f1j<<$7HIDY-&JL3G!+I@E%?NL$gFzkn`?+G=@eIu^j|4H}pyfun z`HzU#FIbLi*FJOi69_ehOWhRE^juoix2@*TVq1!IVXDY~(wUrWpnrq&m(K22^TBBS zGiCx2wP7@eh-V={Wr3G$!wjjeq$;6O?d;mBh+=cKyRW3qMZWG&JXNu}eR=i$1sv~F z=@swava@1PO{vw$7pMNu(=Y4xmc}S=brP*@W5h+%V|i7rOnBh#!|d4(EY3)r*~oRK zng{>vbW*NIyn8$hn-hQch4lPAJ>6g~-7nz_1jxp~E-(Z7-!weSbG&l#O&d2aS16qb z8rZ#NonA})`=5Qo)5ysz+?~|;x}L%nvei!QDosn2WQAcr`4&_mP}2I4%Yy$w)0nmJ3GS} z^44lnf|LH;XIo!y=6rA;wrs7ENVtOM3(@a2FG-KN?@&5#pjlnif0oik(glTMUq?Z0 zyONJn#a>~r6@mIcdr>OY9$}jm$hes^4Mg3UHKqMfK9-yF37qgl-w=u|>G|1uESrMJvEqY~u zQIg2$G~DSe&#U{BP&@3!<~6z4RpZ?$DLj#Gpde&3KnBr;NrfgQ-mWm}2&yL?e>s?f zYG{r>nP09+0wNJWdkp+mTS#&&Cc2LP<#dS$xI>gJWo8fdHB6DG`@S;kM73W@z$A9J zW9f1|#UBYg;1r>8yfjjTul(Sm+SQ|L_Jh5(5`AZBJd#30zdh=D4Phw2{@{n@GIIf& z0aS2q890vJ{b>)>Tt}5nIS^BVsA>?H!nx+(nFf~|w?`dW1r1I&cY|hI2|7wqxAOAS z%G`O#>qElHaGB#$Lc_vA+x~3MHdwpM;J5<0;jujsZoNd7JMiIp+$PVWvn#fvC(w*f zsUoT%)u&fL3emVh)uOR|vk;gX)zzO84vx>`adsRM?Ng)<7=|X;ifu!hYHX0r*z%ZH zeOW_EdES2iekGI2vP=_ir(k1#f`rHri;1s3j`G*=k;>p@-V|yL{Ek>cuc~39p(SM* zwPU`IC@^P-Ry?8_Kw&xJ*aMs!-L<|0R;@sOFGkgdi_ER-t*;J%^P>O(KVBUziHg#h z=alNQ{@TcRC%p2j4Pg8dw7+d=(0=#CHLS5XoN|fh;Hpz~1sq&8`dVl#`TjR9dqU~o zfn>T8!-Di#U+bFThYV@~JOb^Bf15$NZgZ(&WZcy>v@+#xeI)3{NJ}se4HUMWhpu_O z)ntx$@%Fq=&OD{}>I_xMmb;ZwDuGrMLg;nH53=-wJCKBc0sU@P@&j=nEhY>sz8*>A@L0d+b)beH_seFQ7VVXIA!hO3%pkCGww*V=EWE?Q4Z8`Z5kC!^jsMu@L2 zR>z=s09?UWem?7K6Tpo1a^-bpzajxKtK-4+7tMT!X=9@ID||U%y-zr2siiIiXpZK9 z;FeRf6#MgTD85dudz#q>7LeqH8Iv{L@}D_Ul#v36EQ{Amx4WtOLZN!0ow0-bC!fzu zLfE=+ruFMqvu}7#7ZU*+R1f+56md~)%f8;&+LfqQ(U&Udychorly9gleexE>kTc4v zRmoa2VPxQ37HK#_{o_ABPcxeA;^n_e>r8lT+RGEip9{KB$o%;GZtTuC=l}`TMIW0* z{FihC{x9jKjW|L0-`|0Qq&I+-^-cOyy-bJuZ)A-5zjK_GhOjlTy;%DKM);ZQb=B2e z{w4GF%_Hi#YO+*Lx@bgY$1f+;(n{?)QCbX!@ikfX+;HZSaXsq*bAVpunO|FfO^qr3cA112k>CQe~g!-3+ zR}SUe8kDJQ?xVt`F^^Vgq(Oj&PG#xX)`@2)gk%KL*LT-}(6(EMs+nSTV>RIxHzd zGc5PEadq3KtXkTV3W|~m=oqfB^rRUCdIn*LK@_qZU=T({0h{D5u3|HNtGb4S6~xqt)9-$XVb}emZqD^inuQZasAlB1?@p zDyuJT>%4o)plzM8Yl|5atfwyuYi)x7e=!mw%n&cTJEm+59kC-^l1$cUa|oE-ZqVPVxW{CS^3U>SuvJghc&JV$jwqy zH}Q1;$tu8bxx{ib%>iEKrpnWM!5CVTjfCsL}v=NuQ;4Qml2p zZLSasr%E_)*Z7D$ADAZT`Yp~`U&!PR>VjICL~^N7HOFgyXd zLu0CfH1cxi-L0Y+LC9T0q>xwW)_>BNx@zCbFP~JsftUBwWwY9MOY-Oz-zl!9^#FGI zeB3}N&5`s~EzexQvM3wuJ=6HQ%zf(C300wlm)vP1?*$h!uSAlQRF|!8oPI7j`W7}H znR$2Xuj*+&HrVBj>0Z_3In9poV1iX|#oj8X_)$8>{uV5YJirZ)n#nrM8KpyRWK{rI zWEmsuO$+Wd*@FiGAZ$BhtYYYSOwxjw`J%qx`+L9}1+p@!A(M%p^k;i2r- zAVZFQe7s-Gy32?lv|Gdp;8#=w^95~xQSJ)wT2rFt^Fi&WN~t3DaAO_nS0SB+6*SEI zqTCnr3V&DmepT9LO=eTG5&NQte{Dp?mCOKwA0!`KPYC$cmAS$*?yqO=zv93SnOpsl z+kswk+Cn(z*9rK2Vl1crz8G)d(CapkRC!BawNs5T9|c$za`}f>3I8K47ht4A*=ENB z#CLNPDY5r{whZfoJi*35cJCc9Z*AtGo5%0_)EBLV1*tPkf{ch#X0H9;RS?j+4lu&} z0tpy7b?N^DdjnS1z(_5Ql#Y8&S1$9tZIZpfkKh_&?T&{;TEXH_kQu=TFXgUYqH|uA zzRIkT;0AdnADSTRl6o5@^DV(ABuw#cT$ zitwyLl(}BpseyDvRlmzN-@k@eGasANU5s_9-^?NU?CDRe(g60Ae)Sz@&|UbF30kTU z(N!9KW)JImm1D~~oKd;z!7`3N-}a@VEt=%?Et)gjI~s&OUgysHZX$a^m5Hcfihv#r z>u0X7^DAzed|X(xxw-mY9`V%Za64@}H|d@=SIEhVxaO<0Nkl$N{l78?jtiosBo+eo z5kaNjTFZ*$0Hg=z$9=dnf14k@Ze*fXBa!XTQ5 zq>&C>H^liNGA8*Q)QHe@=bn2J&zuEaq30LK#DAi+MA&U3j|U2>2u8(TEtd zf2ro4-v<1gh+nAbRvL?^PL6A6rDVX554A?B2RoTVwq$$TBb-)?o8%I0Bx=y~w52kP z?(Zj&Q)q6rnt@q_C7Pa?&e4C2TTzxu=wUa3WV#MRTSDA)&DW8=h-F2D*v3D+W8E` zj76wpDZF7$qJ9G6t0VX}wt$?si$bSzy;>_k#|ajoenOeu_-&|$M=lV=iM&~#o;N!f zcf{K2dGxKK?9~3((RZFhRM`QWg6Bg2GE(sauiOG|o3Wj{xk}WopRPeP&8bXW()u_z z_@3dkl63L7g^*0=f($9n$g+nxuHYqw>D&-Uw81D0woAKjSl_E!>Em~|J_*Cm?z?5b z)KVTZK)sJD{XVFj6etv6i`>8Nzx+iUiOXnr^w{YE5K*K7*fx{B4o1(e=$`DO{4}On)4nPo-j>rA~ zhpzX2XM+v@{##0_v081=AZ@L-c2#VtT2)jx8pJAEv1&9WF-z>())u2T9cr)IGo+L# zEwP0XJ0XY&p8NhB&*%B+`#s)&!gXB7`+c6*b-qrr59R}41;N|i^M)%7=Zte2{h$yT zijGKh+FP^1EuPf41%^~y(7gF15tYEZbl-id~f6x7eA&e2e9^#xmRvR|OQ}g^$9uV_IW`MI3UFQqwe@-`s>k9@yL_Im#MenfCP>MGN-b{IIH8tClP$`V#^%9C-yfd{wL z_evgIWezI@p&JokH$kVj9L#Y>_a*2gsu=7$60VX<+#*%<8O~n$FTERlS_}MtJ^y;X zQxZ~P!)&T2)7wl#Wa?#;&L@2{Ncx%+^7vK0{Ijt6u!(WYxC?m!j*m{P+&9Xzc=$Kh zu54*M;HfBZIRyp`VGd1SiRLgNWW8;Vw94Wg*N~_|Nn-T@5KyWFBIWAIZwq>XdmIH} zPu7WB_<|7lec8~nmky*ir$+);J{L>1GHNr(-`kp6XuKvTD?`SQZ+282gE=*;C4A4Z zg*I@Q&LC-#xn{aRtuq-KGYqD(wc-M)+n%Pj8#a?7FaTH8e+Sn#+H{C6(dNrYCU z6WP8TNd&K0X4jT~g?Jj?y-o>5(i6>_;3z7|{^TUMcEQ;Q#Aj_1vy)hg3+zl>uuS1F zO22cr`MH7;Ex2cvi{0N4%iAq1DpZcaG;RHmsSfz>&xT(eQ?lX!-;SpL%7##I;aJGiGo< z1g!^q|1v**YUsU2!a4l7-wa9H$dC9MkuS(ngJ;_*tR=*^@qnE2!2xzlj3G{5r;>8v z?rr9Xq0P+oCu!f{RK=o^M|Oh2D;oda7tCiPb5IIZFUa$RjML{qovO=wy>XEm3*Nij z7#0|@@_kqKIF6Te!9QJVJU^pk$eeSh(<37kQuNLf#8qd#BCT-8T)xWx|WeXO@$MTVNfF8yqn>;c?L!)PHVdF<%kAz zwfMRY50lR#p2ty*THf~}>hTH}7uk3JwzR~l4>HbZZJD1D``sg6qSaGaS5CkieGSwS zTk_G8Vb$Zps?wcgGh=&^;zI>ryr0j-$P*thTrh3-TNBE?bD9T!STMq+o4`P3R~oGZ zhHl&&cPlf?-U733GHXBC5vQ?w)c_XO{LB32CeeB%_8F8$gZ%w&hNG`vS}_ z3kWUc+$Mgzd0T4XVld+rC&ksQEgiUEWaz?CiFE0GF_=X;?OBujGceA$x6~nB2a4t*Y zpuexX0W2NFUl+xFf3>B={TsW^KG=9MEJR0+<`r<hM>t-;ud?*Ku4>FgAOX{ z&DKfAFzmoHyX|b5Z$_~DcDIQN>l@8uS~-FE>*+uwOuDg#dUjBDJ?`-(pqd=+=zGaa@ZueEj|ar}v09O%0OeC(snoPd~D!B@P! z&{iN_j;{b#-^)ks>gikhEAQLpJa!|K(&_WN{rUcgZD=alr>Ds~8n$l_y`AXG0jMrq zp0N-za`4?XNZitD3Evo5lIe{!^a1tv%6+)%OlK>laT+yL%AYV%N}|O;6fUDYcolVO z&I&HWhYWey+zx5N0~8@?-_7SmJ%$4-nR3y<9cd3Iy^XJKCEum81RwO9`yTzblWE); zoX7<2QDU_zVC|+~vO( zjuQ+!{zhBZ!Vq~h#;quS(vHC#0+{58pi=|Pw^{n_VV})p{d!FyAY(~K(6Gl$i?-jj zH)r$UkFR?LFXg;;ekiNd3Bf_}GCGIc^AYAba7jnmE`)*&I9vB!{ zEQ%dWvDqOw)96Mp=FrIHjImRip@|EeN3KJ43z<#O)5`!pogP73+3dMko+Qy3lrtFl zO~32-mqQh@P{O}P4t;0AeR{u0671pvm63S0(($wY5_lY$^fQU}A2xTQpmL$t`LptO zB#5YBO%*qOt9leK7o8WO7cHh;1jkg{-;!cJfPg?}CG(I%QYorh8uY#o68^49SZ+Mi zTh?xfLx&_fJ?oNv@PQ)(sd3X|sG~{`ricej@1OA{9?W!B3xbca<8gFwpVx0;MsW=+>Hv3@c&z!pL z&_7^0IuciCkhVay(3%DC_%G>jk6{lk-1w9zb&>hV!dn!=#Dn2R*w*L8tmIwBU1ziv zziLpkX|mksLJtENWw_am?tZ$`3ih5u92#579yxxn8%}eph-2}(51C6-BH@Q`#HQso zX5%pjR^{_Tu%H%Yv3)j2qI0V^!80ihvn(*k;*@2m=%mYnEJ(uB?yjPdi4np6%nG$6J`0q&tFmdFG;UOE_dz`cpsG0=4P^YW|p^JO9Hvfm4U@t%S*)+L!y2 z-HFS^g=3wPrA%Ev?u%rHtYkc7yLK+J+wu5|IU2KW z+|Be~CRGRw1&-Un;sA24D~SkSXMqLBLFUS{?5iOAI&`GB>m)df3=V4kJ|3&O5~bd+ zN6v4_F}L-(I^p1k|1|cK)lo~|yaZFP@7@*7Yur)TC{Co$XLMqA@7o;x zV$kIfqqYy{V5CcdPe<##x@>Ge&FZ-~Q8~&HON6zAQo7Ika|nj*n_@&0(X*QN*&OO* zIyd8j-i}@~ELpDF9WNk=T^J=_!G@J8MFlH;P7gV*vLCdVK%O?LbuA2OU4+%>@+C40 zWaeDHS=v3D~Fv#b}0G} zaH29fOIY(GOQX;}`z{a0Kahs!qqYNKMwqP&^ndtvOA+2;2VQN)fhe^&%~}doL{#DO zEUEq+;NrhdluOF07-_X`Qvj=$(SJUsagFEZBnlZp7*DV^8My53IQ-fD{wJtmY0X7; zqEZWe^|VWQxdE#DLFIJQ8z0-%PM6)Zftn}1i(~&EIC?vADuSP%!S&}%{i*nKvOvZL z@brI8n-#9H-xa-|b7wad*|ZEWvC8MbCtKRA_+`ok>d#b)G7Zcj6rhC3ngm7%sLwGR z=IlN)To+<d_FuL6C0wl}=7!J~+T|05X7f5XGmU?hGt*(Y} z#nr%sU#9It)PA(iVd`Lkct6tAD*b)$EU89Elo1;4xjmPEYfKvrWCiav&irgBe@Bgw zHvJD$W}Yh%gc`b65HFbJ=oBfpKc_qGds>gcp*82CWNZTjfuORVtbDgw-|D6_W7$C} z67x>mau?O^oUqgT1D&TN@XB6o6VNHw>l;kC8#1Dc_+kL}vPVAR&P@R|5X;Pj zXR0g8!%9zGG^5M<8ei%)xCbK?h-#cW!^NvbH7KTj1V`6#00~OmRb1&WUA(99Vy0Dc zYMpW4oC{Gpkm!_nFHgGyY2zz`Hl3UvYrG)Wa#8dIcett+sRft)JkSZ6K>YcE_BQt; z&>R_AN;Bl$vYd2fjoF1Hb|qmZoMhTxL8R7H-t>N)4gYrcm%UjU5;y~O7)wwNbQldP zH4|;n?HocCiPv|g1kL5i2UcAz6(hMvRe4s_=I2m;5*5|JyV)4F-oB_;`I(CJj5b=qxI+y^% z*R(wICZic)ZTAN3z%@SKBiP#1V1k5EO%E+_=HsH-Gw1aSvWmfLXDvE_KxFy8TM|Cx zq!l~o1e<-haH8MYs0+xk-9mcdB%>>*?83V*>!rEQlT!mpl~vc3$zAHL+tT;+ot}1P za<4^Ws%&Dt#;>D(Lt2o%94W5P~XqyGY5qG-l2m;`>kaQ#F45nPE8nC zI5#V%-kYa`VU!NW=+KLKAx^4|_Ja;TG&h@+RVA4jw(bG3T~b#ku6(7Pjr8gPHg(@@ zSr-eEEr}AP&M7{1>1mNOo(;p&IwGhqPQjQ~N(S=Gvmr`<>G|ECqjA`^ zsO;%87o4-DEP|u2Yi+7uH*Nu*C_EUW+lspYrY~6G@VJgK=*CaecMP zEB5H^-o-5Ic{qLD$Iq^69JlJGsh$`fKa`4n)Fs*=0GTWV1Z+_w>SY#Vv@6P5RjuFZ#!uZb8iZEOMvN<)^a~E_`#zSb5s&)y*LV zF(Th1bY6IC>1)0q)TpY}#m0D`(_?d3%pl8h?PW=56;Iz(uL?uQ^zS+uClX%5>)2{} z_1+gfUfbL$5ePjT2GN)H3N94k?Avr#;P8qOl-05OOUj#b{_AC9MSvsNHKt=vAm#tYzv`C z6oJnQqV&1&ReV~Sv$;X3@&haNh=fly+{A%}~ z=2v^$R(lzB`u+NNWZVCWgFh)RbL}ovGQchQo$;=O|2dR^^bE%T1KpTDlr&;8JEC$; zx#F7D2V#S?`r6g}KUlXFHR84noKQkYdpjawbmN8h<6lu0sp`>{AD>`SC46fx3j|CD zlwWeu<3x;`MEMR%Y>{G=Z@fYtWS05$ocN#dz?)QL?B&W@8 zB1NqFYt`J@u8=;QPc=&JpIAsaA6Q@xajG5D;{`uMUArs|n3fy>bu^i|UM~4%Ue5UZ zP~b+>xR>5=`6=f-t}p`)we*i|-3ubqip$ZjM`tj5Q|fQ(+E!l#3mBNU|5BM6Dy$nZrP@@r z-!2W@+xTgGYK?ZgbsDLBvgXcQBfIv_a@=~p{kfN65`!SDxrDqjEl~Jtc6|btKBX%-2%m^LWw-jEfmW6)Q6zj}~PHi8{S9$Lv`T z<3n3~bXJvNz7&Lw+K?u~v6%kDzf10@RDR+iE|+K-tMhwE5Kh&e{w+{k9jrLLTVqF& z=J@SLtYP}DRQWUg%n=*csLvIob_a6XTx4Q25*WYIHiveFgEtBb8EZiRk@J=Xi2!Vb z%5HIUAh|Wj{%lZrcbSIHYN8JLqz!IAYZZCsp3@@0U2)!6yD4cZGOydqGqOT(>{Fq_ zjtJ!{s$GEz`XO5%`1KCh{i``A`b($Ie$kOgxvt(&Fk=8XpV+cG=mPcVpI03}q9q1r z(~{i(KmsM}P#)R+Jq(qZTSmBPvJRNWKFzGX?Ao92qGV>{pJ_w2Iqbx?T77I+Xh3oo zfx!DI8V)U%sU=5g-x#4qmUHp?O6DRk;jE83;No{q%ab*4w$B{RhEC1m(2BIh2BqJY zZw0dgK#5JdlAx`cv%>=X8VD zByGuVwrXT^Xf-t7;#C#L-hM^)4^(e0O}>^k1398%han_&O$TwP-ro}-*6-)5LCkmG zrbHc;G=e^enC)Iq3+-%9=1VrVg&@$kzW;>QI&3Q2IKTedVxJ=F97_sgmY)+4T{ABE z2e>zu!0M^Lk`N^wq|{Eb{J<0YfO$UHMmh3V^k7SCu-478L=g@?y)s=Quo#h;khrLmi{Bt&LEDtKPcV&C%dquQ^D zgAcI(9QFUD+D)}(U;-(4)_J}yd@(L4U@;jFZ~4@mohkB1hT2`Wumyd+e*<^Z!qfwu z+7oaBt}?o3WvXbKFJu1KZS69`2(am8Br`#OBLil`@{d7J@#RjN1U;WbYfqt0613b2? z_SAPXSpi72?2$x#3LIBs0l^UJ3^aTtt5eKU4fm!}P2J{Y3BGr{rqa?2|M?c?Nhoij z&dNsDq>FpMk4l~D;*3wRW!3S!I_iq1YaOgrRW);ov8V#ma!}MWVnIQ-gYgYhQ`;Qc zw_t(S0&kiE3yJ!jk(tF-lH)5WdJhu?n_fuk&ct&y7*~ZMJY!G~`JX~-O)nCE4LG$e zzUNU%J&$T5_}$35MZ^-WPfIWtJm%j8$)yahA8fY5R+9u3C~w?4-ezf0Yodw~j04ic z^@*;qy-qCOlJ?;XpVzW)Gty@PhD=pm!lEoRy*er1I3tMQS@2Q8-&MKocEYMxw<8qj zc%z_DCwq2z{pJpF+o_eHIr+QlnyGv>?G#wRZug8bs@(X-P);B+tz{@*8kOgn+@?S7 z1RRul!ca<1;0zoYi2O~moW1n5b0HN8s;#NI)!{`(lsTdf!0h@)NvClI z+b+3pobQn-UjKx80M9sW3ER5P=EyR`WMGwqJ_QlQUt)tXwJU~z;N=nFBjnqumoDuH zFyP`!Csfpbkfe>a7*Y819q_2yp65~Q4su$X@re(%Mah~zkDGsBW1u-Y62Wm0f~~hd zXQm1@oaf^)IMx|fo~D1^qzpfgy~f;9{?1f;O#LFNbpbc@i27JJc)imPy@E~Xbr8r2 zRwG^*u5NduFLRqSz5KiOVl)F*$~|%282?s}1lQs`Z;fd!s`ED?NgXLq`tQzSve-DP znrO%wE6i{nC519q_N^X3Js+qx7FtyQ_T$-HVdj<&gGEUcr9b7uzTF~xps1dqF5uKw zFxU^ux|rB?lTNH;I6jELv<KMbJMSU1ZZ$|(xXNkU0V4I7v9<)+S)cVwrPK*;iuH! z6LiNac7FkUD-$Abmq&u7)^nvOz_M_RVGM zgtbs<@AUqlj5fx+i!)6D^i1*j>UeZA>uHd^0RBn?+63dccAS7CLjU06MeJ0!wEQXd zVo=F@Z7yW-|Y#BC&Gh3+9JNc^JO#KwL$}!cKbeH&R!}enhriK2^)puJ;K!3O)6|k`D1f>{r741q*rQ9YI>WHK~ z=BkPgH%>HmyZN;+xJlDQEN92P!#=sTc;%VC{K*gK6amkbJDHN5l$7DOrZca5iUf2r zQQho1cf)oHJ-6E1wW|qaE*M<|WD8%&KZogV?G_QRHLfzQG0yua!jlRU$~bNLw!F7Q zG&6VOHx8G0BVX0~_q|shC!ax5UI1I9tfgueDute(+jQp_unl1RscTVD-ImQBdSq?@ zuMP`!&!1!uE&nhd9WoKHIqYb25;I4d2D@fl&Hqc}6Mc5S9T&I|CVFhAw@?*AE;b&@ zg9Dl9q=d0+Ym`wE+l$)J;M98!k%`4vQq16nI`P2_IeHg`$F`8J*jS?9Uij`_UkR%$ z{M**ZM7}eno5-`~dOENH3({#n&(wb{3qJINv4-^O!g_p9vhV&Fv|r_NJp`py>psGC zYp>@Fxg2k(Tl++5ue6Q5riGMMs#Llzt$D%XGW$mnn3Xhi%Y1UnoJVJ*c9kEt?&Xjk zjC)LtM7owMGGsJL{*caYq;Ye96x%}dW~|+UFE{#ZC4CR1`qv}c6|^YHgK%`=YxKNv zpgURf87(iNeBZ4lt)IJX7(3}-0jn$mi*iz1fI*ln| zws;()0=L3`$donR=<%t=Xoy{*{t*daS&A70rrUQ#V|SiU zP-TlSpfwrX-PcaMuSY3ZdAlpNF2J?e<0_dWGb8d>J<|ti>+Mq$aO6wXR~H-jW9Q;( zTd4etEvk`KGJanK6MM4bZ7Ah(9GFpJmtJLz;~JhF;gkflKqUOQ*K#nmIpBRq)ik(8I=lW-QSpkzow=%#&kOr(1??|*~;>j&(~?y9<}-3V0&E$cjt>L@Gty`u^t@8Ozs(^z1#c2|y0v*ZTFr;62WDzj6Ioa&xgE_YK(@>Qv_fcDN z+(EQ%Bu@<%R=M~n_%Wmo!p|a7;8by)#ZYL=k{ut_L7D0F2x&g~MWHEExjd2nF+N~R zFTL`~s`4wv?Xkvf^Y^GXPD!1P=Ox6}FXlGk~Jw1pzH@rEAn@v+64(mlbiq$|@D zs8-EugteDVFiJ1@|0m9$VoOoLo~k)D^x7D=@GzP2MB;Geu9GD8oH5~Rbfy!6%_Vso zY3eK%C%q?GYW9)%j=MnG_?=>Ai1ACwFHECH4xbX<67?Tdx?O(C_AMCwS=#@4IMm8$dFH|U*Wq<2-~6D~feaPxt*YGS-7aAnE7YeBKN`*N&9iFKS>K8HGy1$-MWNcO&)5}?&L&(*}El6nmKZf%j+{Gtq2T?mH<6zZX0vhy*%z)j~(P%hgGlkwd` zE4=WsVQYU{aK`F0~ei?F&o{ z--Ds&F2Gp5qubVLQxe7tC(gg=ID=e=KyeSX)*)CGXyOSfAxsC7++bo0x zQ|3YsXWt_#JgdhZ97gGuS9%q;D0RDdtJp;KA@8(~#{$UT1{eIN*b-|&;j<;*BF=k< zZ_cJ!ucb`SN_-$D8eDWt=YV?mEMBu36Uq{2=iYBR{`s<`^27@?zv@se<3;{2Jpf++ zIJlSmF^JyRQh8zLf>8Y?vw~w&v0FhzD?MZh>KT`YM$DAr;#uv^&htGR$+(p75w2># z+QXAwW{0k*2otN@%%L}b7eksfd-Exs=^orz!5kq3?aFQ?yUQU_dVCF^wX+a#+ODs$R7+v~Qq@mQ z8o~KgbLUOFTSePu$iWeoF zX;_PD!91n-LpHan_4ef1l!QgT#&yS`XgM}fWm_C?ipRR@dN1czaCqi#u{V{Q-0Ul? z*2S{C&W?^j;7)E{@?-KbmsT5b7b;^v*jos;31C?zD; zdb(-*&xl8yHr~D>O5VTAG?PH+T``uc zk@MJ#|28n)V=h=2VKkh9f^rNg<$jv&M&QuyO7i=k3yyeLWdP-ec9XDAe8zNb7i{7* zOry={=x%My?oc;Go4j{JGPndYGjs;fbd?$A81%%(ect8iG3 zUpt<6L{6|bZk=}nAQ8C~yRujJzGtg9S^n)+s5H1aKZNRx98-lnu>kvZak%VhWg8kv zix1FL=DEcFe%TjmRebtpmJ6|{OmA;lwuWuJ(by(wgiQ8zYO&V;V*%VKUZqB`Ps@bj7U$eG+L6JcUUW+Bwa;8;4;#QoEdQV{Z7 z2WKmkm*%v^C!WgB^fsHJe-z8D@gsf?Ksoav2@9O9Oz8gYKK+>ZWInUQi6*^am&%Fk z=V`n*@?Sjq>C_jN%#|Ei?kpoWkcG0MTv9GEtL?p*$SiUL2D_iGjkSuANT4kS2UetSgW)=_*CbCNQb#T~N= zQas#LfX3`eCG3j_*_GrR4ZRf*yYPzHzh{jtlYX@PgAC9BWz!=Pn647iN6Tr5Spm5w zC%zWVUmRKl=c)^GM<>93WMEnOa+`}K!;}|ujX~tKspa5fR$Yy+^%tcT^Nibvi>Y6z zq`K2O!p?wATv+mNy3Zc|o1iG_=eOESy?OYie9ScW2=csRZx@UxE6S5Eml_|!O1Hrp z_H_%F%QktH3`7I`$}K#ZdXCN><4Ov9cI-hJ8$J_7`qCu;%hwQYm8Gh&6VIXr>h3&L{T56YJh zlBGw(L4mzL{s+&q*`}ZG_G%SqSYDfK7p@C_HP&^`llqxAgV6Ek5P4-9nVlUM0ilI9 zEjYrvZ+pIkQ}bokY33ln_xCLO8r++CwqRl>m>Fa6@J2^uKi*Vmzf>Mx9U2HFE5Ys} z7VNWH8he~)6jTbhHCKTH4!?n=59XmR<-hsFyV*)ma-Z#UK|&yg!tF9_uq9B zj)gtNsDh{i548QTZ?`R_L+?t{7!g+oS3(A-~x6(>@S2}MNv z#!{Chox01etPX7bZLusiRtmSniGy%`&v(O9&Uxb04UNj0%cvvD-7!o+5aCN~cNyR5G1+PGWyLC~ta zdG@Re0MS$ zP*j0d)D4}Nqn6yBn1$6+Ro)~(F8(qfC4aLxzVG=@60t`ULpH|TT)0E+iFKn^=D&zj zb-Apmu~Nx{spM&*YikcKYhos~V(B;~eT7LIAJ!LjWf!?-Fbri-#l%)n#oDFLE?*i&{pcwa^ME znyj(!yX5x`I>oV8Zw{(1mt)&%FV_Qlcm4Ecuk-VVa4dd9Yq5Df>x;44`{U*|)faSR zKGZA=gEwZyhrBIuGP%--<2E@hvvRG}czWc0_1zI!%6EmichtL-u`hZBFSa|ceB-c8 z6(oE*GUd+3Y~E_^x*x1#k~ySwxl1`fDp0D_j3{?h+g4B-uu>>vg)i&=H$h9ugC0^} z%>Q)i&w4h1($CULL169><;L%z+^-vdku#l}vQIu~e$3(hwn>(ct8aM}tiO1g+LBt~_h*i0!-- zG6hBh^1>MHd{VXL zbSDk5-nm?}8@{I3n>ZXC+}ocOnBP;I?}K9&E(zvjAlrJ$VXG@}5#&R=8}4V@9n>)y z3t1&e?Qdf?sKD2BS$X(m#Vxmjl03Huy@sj9H{Abx!a{1v+`v6L03ZOZgqIfzq}uNE zu9%6{&vCe?Hzc^E!%M1eIeNpvz%y$<@p9UKAL+o+x{%HEG#i%<{Yku8is!+`r!3_V@)`G!)W1 zLzs(yXbxC~*a+yZ4?VERIs=~Frl03sx#sm&^sp+}|DISwL`@6jqWMs_rm3kZfzsl} zGSCM0#25PNVwwNQq|U{xjxSFL=is*;&TH5dLRhn%p`NT^M;M7xynfzq$cq9NWn86v zFKhM6b7^rL>ND@R%GQU^t@@fTIq4Zq+-1>46Pc8TK|=ZT{0utdxC4%RPu}tBtce45iA|QRZYfm_n@h$2r3%j1@uEZ^;6jI_;mtN-C@>i3|>ct!; z3$)^e=vr6vAz(55*&tI+cN=No0TD7$aT%pS7SocuxPMJeE9%iok=w2#vfS8N;5N5O z9a-n%LPVDI_ijk4oT4YG?CtPtk&jZ3UjDy>j$BJ0jfzslFE z-to-lFJu&;$f10|mVHxZZyoI-FSFsZBt)yLH84gQvkeI74PuCc!>?s`d8%J%RD4Q{ z^$b3k@DaDj1-Loh$Qk0&ZWEYhebDI=@+EyA^F;58{`nh+>WFRGi7iI?dCEbgaRH>E zap(6>0x`mM2f`};eLbX_ij&9WWFBD4+DXN%p{=_WN!HU_HEOG&%Ft^tr66!$m)ky( z8EL0RiTw#7`~2P42(tBib@u4@YaxIr9jNL5Tozg+6lixnEv6-atQ;6xp~ubdM&SzT zzNMzJTi$v`w9rM?kJks?xWsRQDMPrY_Nf3SmY;b~zV;<~@6f;T`A+S9FA&o}-`1W7 z2-}1k9M-!uB{*ljcy+=@3;@$N61zWry;b=`GGG9&;qRK8GBpQ0M)M{#CL?}y#%q&H zIy(zo+otOy%ymMf9ZzO0@|_7-GWjxsm-TEN zVrM!_7vsHDBZrp=LgBR7<2B{n?9nj!H4)dRkWJ>Rk-hPa^uclicDcjNgPYGbzHhW) znZ7j$JCQ+@St%njevI8EK;?*n8=%km7}#J%ZX=Lv7{Gq#86F?w^&Lm~w>mo2#Rzxm zy4=Ivev7d)2jp1%Gryr*S@DFkBq;8dImW*W@HY;B5ra#xeh^pPa@qW5-L9LmtI$^& zRGtyvc8tJUUYHT)%o1$PcxTlj;qUXJ;%~CNx}o}XPmEQcrrPkv&*rpeFp9A zB5I<;*}#5#gBx?kfi4aA0q!%R6k(zFhr5jBXBxioVkV}}okqbU%U)cVoyypnjomkT zvqGmTErQw-%(!zTL)J*`Pf554{?u3Ijm9tKemvuUCy{yLrJL^5m47%iQs4Mir$uF2 zW~N)MdFfl+ARvfy>{8dH?n0jtZ>p!x-c^%pyFPx`^>>wBVJ?T@x?~R{aJmuWq(Jp@fQRAYWIZ<{d{)u)Q4Qn295$w zn2C>PJhX>n@gUriS@luS8oA9pkPafy)a;_5N;DG%=rfoLF`*&vwZa4|XSsG*gc&P2YZeST3g zZmKgrUO#coNreyz5VMcrRbP>FYi`_Ivo;=eP%Wv+w{76Is z(tp7_#o9Lvu~Y5GV^+V?>4%DzS$NLAYzaz)KuI8g>z`oH>aC5A@b>A39t<}?9qRCg70nJX|o=>Nx75C zLBcqS35MygNjBfLOCHf~`O<8O6|~Ib(hGPrD3up)gpPSZ{)$@vdD8Ps^@*R0v;p&F zqbBmO7nMI!P2y?9S+KyS>_~;TW2!f&!Z^{uGPaOhM@x@GN;T2q5bj;8eW3N_iwLiC z^RoT12LpI}zNGTkEV8-j*<$+yX`yCn2XMP&ARNHAD8!{oui4>@Q0=6~ync0+-ZL#i z(m@!26V_}K+T`t{q;Jl0`mk06%z`+jcdT82o~?(wB__d?DlU!TvM?B?Dt4`ZU+R27 zI4pxOU*(1zR!L7Xr%f$+y_T~fHWqw?Cd7={i&UtaY_1?Kc*3d+@?xlJT0?%vka|c&1lsH+qSjyvY>o>pq7YGJJ${9rC`> zSjyXY$=va?W#os=jf$Vc6`%M3T2)|LZdrwadMqu>tWpIFkfQN`wPxXKdc$fSYwz{u zMK{P|;QkFVViK<@7Kf#La(h&Q+j?wN5}~f%`}fW!p4~DAC>B;lYUdm2Nko~vrgU`T zHFPG^dKlkr{HPWKcZ}A;n^6;7?v4sp%WB~9cr?f2e_xHIi*z~#G3a>#8_YMQTcV?s z9Br^nTn{<)`^i>1^>p067x<4a8xXyHEM_MucsTd7ol}JIBztv`^g!#?v;xY&NZVc1 zY1wCQ#OG5n<&ns)SAy~+ z?GZaxwZ|366Wk0#MXchj%@*S58ZS+|<-@@6g`6sH|7CtiT<199{bUaoD;N*|TfNMp z{WEnHd77okHW4}7I#0_umbzqHbbdj8c()={87K^ry7_K$S>hW zTI$!`CQ7@*19etCWx#B>wc?%cF;D(UBbCQc=Q{a z+ThE}4#h8$J_NYB6?!yDDO*)9`B2n*?K?ble-@Psg-Bx_6GuMWBXkCj<)I49Yh#Oo z#vVtwYxJ7Q@9jiG;b&hjKdiV`%p3Rnfd!nyj9&A9XnG5`CLHd4SR^JAQ&GYZQbQ#M zqN6vFQYjTBq;n`kq!Bh!n6!WjN{moYQDAhAF6k0RHzR}%28^-6@A-V+>-`_jb)R!z z=ML{`DY=dD{EQ%OuL4EbIS{PZjL-%PfrcG%G2UAYRRx!Iz8-~u+91Kxd>H$@VC9mK zIrq|1Z`D$Ztv>P68}e-hJHj#ZQ?J(|cxzgs%}2;M%f;%dYx7RXr-=>oz*(ON z500)9XG@AxhUfLVi--0u1}^{1MNP_cH7N{HzuAf)30%@=q;YXexv z=J0whtPCO?xii{=##>|P^7}Km+15E2In@1>s{Ho`l_V^;}@(_u3K?;;bZ;*&nQYbA(dFKM6{StZa@AN6oF|N8XROWU*5NVVV z3W|%L6ZV?#+yziu#t^<0VUH`%`qz|sn&KC??qdqkn%0jxJ3-++=W8h*kU2f*3DvsZ&qK{tn2F{L z@!_;Qc!heu7tNJGpO}6$+fqerEiN9++}V}!>5Z6i+@btxP)zs6Zc`*XhyxSxQ4#%w zb(5tqZST7$tp$XEHwXT2x~o2c&Wdky1FEOF2cu?xYtNUizD@^7M)*f}d|^2^N%F$d zrP+$uDaq4&f8LnyjHnhf3eXJo6L@qn>BG}>SLIZvH)1t!tu&$^rybj{Zx;D>m{7<1 z?_jyU^4;JX9UF&PYyyOC?~k29U!1he!(+FWF_%>KLgja#bTm?A0Kt;2^oi9_9TQ)pj(fRZ z|FTS)=Y>Khv9QWSNqnhmUf% z<8@Nb@R&SMvs!S$88*@$&pF-I;Cada{B_iif(d&fGC3X|TjQkL=`@~F!dphtx2Y`< zSI=?F>yx;OrG52B>>Iq4xk^37EWFo9dzU+OHhqAQWcVQVJeWa6_Iit$Qlbz%A}PXf z!nY#quIGSmSfEo>^n-y+z4SORF*v3-Jfcc`EUc;k(nyMgpl*Yk65^YfJh)R7O0Pm< z4!36P4Mf$fZ+|Lzz~9Wn3$-Cg;}44?&hxgFzah7(&S-(53AP8hB5=`ndLU+> z`)mKDJG$8ZU#Weu#rI~&(ZYSH$&IpTDeGMh#sV2BvoS@d?3JQM)1!{QS8i-p;Zit> zewX#A7fx{#O{??$rxhbMb^G5n>SYYb%%F26dpz~qRd%e7>MCt14Dto8eS%C4IS%4g zSZG_S-AeG)u}t@p9|zWDA1=3q)JHK*FE-zykhWFTB-Kv+E1Tz4g)~4MGY)| z>u#()Z*594MwEPm+mF9-)X~kq)W*3a zs5+Z_jRRYwHk+o+%tmV!w-S^kCIKpoG+tfoBz&ogY4C+`H)9wR8Ka4LXKiV1%)#r; zXI0Z9Ftx}g&@qE@@Y0M}!j9Qob^>(h`-QQ|_%XxLNuTG2xU7$gj1T_E5ud7K1nM)% z9mbjK-m>z}Jm+op8~;@OZFq(4j`LR^3EMfZMp8NT$%lLWDB!AJ_-y@RE%@IblZhXZ zZ#;r_?_jVkAEyR?9Jqq2)}s?Eqt9AW;?#E8Xbx+^ax?ypDw!i3_+2hl->)18Yxv;7 zr=$CCG$<>t<7FJ)mF4+$_#Vs*us;9V4w)St0cK_F7gulFi~Gr}CA*bKJDv$@k+#+C z)a5+=7?O0n20uMQ8cUJyiNN$i{C3}-yw##cnx!qo!ZKnNmY3kojNY z4!A<;DPpBh`fww)ni9%gyZD`oh+CXc?K8evyi?Q5poxW4wb%*rbSO@MT z?XvKBcIba+OX+B0l>-5m-XW5HrF9u7Eb|0w_;cqjNH#b$0U9tCa)JR?<|a(sbotY5 zEgs>A{DYv!hB)n-bkg=M;5<)!MEpSP(tv1?FGS$y+OJoLLxM?U@JxZ$Zc<@O)kLf+o)AikOFvtd!Vzvk_b-lU@(x2?NEm9@Kk*r_*K^Pc|+mP7yzmJ^j>tL8(0vjdP* zn>Mzgd?l>&;qn>hXC*&`@58t;r0HVs`zNuZDx=#EY}6#CbZAu<+%A-dc|v(LB)?pG z9Z^`oAe zLgYaXYszU5!-1Uoalu?7@dk|k5FTnJ2gNW*oT zhl$FF#9^aHQVr!e5XiO`*37{3{Lqc&RwlPJb58pppL@I$0m+Zipv3;7kz!WuiDAw3 z(2d-03GGVI!i4A&oy+K@?+MLR*U8Jev5Dpe1w4oLPM(0oW#h(;n*qV&e<^Ede8hD>!Z0D{Q_!7dJBFcu`1QBCwNIS$hTq)VuQk{WlZqb zP;b_`?2$SkbzXzwbJvz0a%j-6qlOWx`vQ2|e2sa^OlYAE;j|(kM5TA1DkuP1)D{wfYPKo%WI7$Cv_{lO!A&Xr6_6D7r z9&e7zh;NnF7_zNj*lT_#v>bsmq-VK|1D^N$IHvTUg=G-u|7(U_HWWJflSQ}sv{MoV zpY7K1ScF^S938);iG3n-zzXHmg<&XVJ`kCc+5K|;B=OLs##P#VL<+Uei+9bd+7)PgJ?x@kR=ijLrth>Vp`rCdZ-A5{y5#Yp5K@7# z9-6Rn1!)==bw$MPbbBc87?s7?Tk~Q`KakG{9{jdtYz`(&@P2^eF$DwKz%1CKCCr{`n?5u9rsqET#N(mL28}+%ZF}Mz1VA~O7y=^QRk+w zPo+t&PP7fCGSAUW2KYN|VbuAE8U-=9ed6KvOuE-)2#I@B%H6TViRVi7jdj-*fZvt< za$hQ^z7E=}N%WWcnU+t~Uu13Jd#@O^Mu3j1I9pI)WO4O1o1Qts6-M+MYS*iMK43pt z2us!$p=mX`QQOqgn;UT-WpVm}JuW{7z$9sG`TQ&TWpBY*o6f4~CqvNFh{)U!EWuWJ z2{TPU4_kBtg$vhT<9S065gCHp9A&==S6b7cA{^uAe#GVd5I9(6U8vYzHGtL|0z$(;IOit)W>!mR2wg^fe_P^11ES5igB#YC&nyvGf`GALKT1+s$Xy4*6S?XBswKcfK$S%@H_+ z=F4)qgi3_WRY$_y0+bFfHBv9B$0toLaTFuL4Y@x=0;+dtrotjjs=0L}IinV(hPqDO zF<(;9p;cZ8bg4F1P9s{8BI7~^`NoeSt?=zYswq_aa22%m$-1k3P)8p`e$Z%1xKRD@ zKMO;l_8#xx&G$mmF?aneW3M4eeD8&>XyP@Sn-*-8`Okg8kFVCdbxzxE27kdq+@!%p z$-c2lm&A;(q~C8E4FC3ae7R4e9}$tmM3CUh+7JlP$J><{i4mI{E~y{VvhbY!l#M!? zt10K-qJ;;K3u<9Y_8Kq8HPG(PTJvJ(oD?@oP3DCDymYkQ$vQ@w3Y|!FmlI`-R8WXA za00!lt%V1XtgoZ?z!{zP{{CCT+ziR@MW4%k4K3X#&sj+IC7T3Kj>Pjux9RE3p<)II zqb=#}QYnvl^iiMb&#&KZcE9)AgX1H@QL`t!>;YJy{Vb2iKZAp|7~W(`UV{_JJvN0V z3WT-s@EGuDc|~0b!SD#Lhiy4jv!~+A6&~m@bIX^!cFG&dmHs^13B0-pLSEUIVNq_A z=lDe(t!n4ahjefPYY{$X*DU^434vSwicvTihJCkONc7;8|#{{Z|-%e!fpI z(19<2TakaTjON3|3}yWHx}M&5%V}K|WDG@;HBZ6E;2!FTFq)O?kgSaXo4?+BtEOTf z?y-M#ev2TgX75;%$HyUpG#7EjdAJYBv+MJLUnTg6&X!3AzWq^)A4ySi9t<;dJQq9^ z7qVMsoT7uE%(ws5sBnMMRcsaUJCOQNZq2oBZjGM8Rf53J7Yug+m>J1sPmr3$!4rU9 z85?31mnO%3RPRuS+?)O1vI1Sz!xB|#*~aK<+Z+TwD(cutFP+XSc%)pPj>;<(xm{t+ zH#yj**Y0v@lDA?)$~FP+mq#4AgUTsrn0~vp)xDfCH=I|y!aMoa|F-?AB4Zcm=QzUA z$cvv4g9AZ~8-$1fs${w{J5QjAUA_lp{R7ujeVJzmWLWz70+)Vlf>i4vLFr^RwnnwE zoNfVM(b)~lyF-mVp9z|@v3T%58*q375_N6X7AZNj7L~GgG2UBbJ$2o@n@bhildFdjB~w=n-b%hkhb&di!$luz=Cz;%82RJi$uK1X+Q*{;8Qo zXg3<(!3VUPk1*cj3iuuYQI@7eYQLLk)|4cFXz}%(Mdr@Mzg^F&6V>SO@yn_}{Sp_-V?T{O^JO*H38M-uz2b zzIwHHI}aTa2k0j1skW^qoFzZN~`EG8Hgyz4{0`Bw7Qpp70 zZ{Fi5Ju};_cQ)lYjjq1Mg3bPiqmRc$$3-zpEQCg!ajz$TdfQ7+72ydZ^n%VDd068B zAfE0;Zwipc*A@W-cA1vWf|l~=S0VTyX=|g+@5LRjma+t&3f|@$>~EEkza!O>;=mkl zwQS!*|3N<5o~h&UhQ{)#ibW)c5P`+qo}V)oZwritNzPLAfXfP<3MZ4jA40O?-|z?3 zC_(?k^G$;G#>nyo0q^OwfTegPu<0Vq^5P*O)*C2GmAqKGtD?evItGVPG^^s-ixyIm z@=aws8+EHN6^4Pb(0nt<)W?Wzf@G^R(_!c?G~+D-+aTI{HZBJKv)-#YCKsU@>v}J2 zZeqgNZCe`;yg2!XRDub~;#m_=ReT-@NFTlvT=lU~Gn*U@Z~9^+bY)@$Q`Vx3;uRm& z*F%a~|45XpyP!AZ$E!7QujZ1fE6G`5Ox?VG1Q19h*)QD+nfepe^h~V8{hW>Dm#R=x z&296?S1%;G`UeGOyi?MYfGXsf^(qO@?OQ}|t zP1hP*7k@Py-<$k8f*7aYX3LO0f0a`*5<_Yq|C6r$=URI3{L9xx!dHpgscek??YxMA zkEVzJ19y?!3ge(-yy$y|^YQV!>Ydr(B!kiyejUPBpv0@k-hFS!{8QqPr6&d(yX?M% z#ppw|G;M8V0tnncjp}-I20fGeEbUf`S)FK^P!lX@fs zLwGqT@!ZoA8(GtmWiY?G8I*n-!t)$>ySd#Zll3WwE%D2IGy_})-kO!erTrB8HMFZO z`(_vcg<`!GUWE^WXjwU+VNK4x8Gpk3XXy z8-i>yKDRz6#u=0=a%!tB0$%gc6dqpynwaOXTY3%sR@7MF(dAW|kI*Y^u@3SfM)eaf z6gzcoQvtIieKsG$JDukqFP^_|mYNIav()0@Dz_Tvc84w6X%)4aYM@3$L60c^@CN}D z3$^kYx3e6Ml`Y%u=W1VRJ_j)f@4|naT?%MURPCP;Z*-$kEa0$KwkT3(6{~#kcjI(V zic#?4;D}KWLxmLA|J+VsXM@(=>-1aL<_dL-j~w=oz7ycxmof#NT{pEQE{F@gOwXW3 zaWG2QJ@vDgEz|oa%l7|>3eV@@SrseoL~TTsIJ@jGrpNKVUO!`G9|Q53;bU&|=kTP( zBPc`OK89$8@$k%cHGHUT>3T|nT}&4BPbIaKe=sVWRo0y-S>LYmMi||)GG=C*SsPUl zf0yybQqT}*X-SV(S_Gtj1ID81Pee{cuEKdNyUwoxfWp08nqG^|E~-Mg9^O4s?@q|9 zdZn%a=60|+!EOUzOUq;SwdTQ6^WyiREEO1L_yPwv$G83fN8n^~Hy@c(8L9p*Z28f9??FKG0Nd~O>r1(UnVgFY!8ziQTQocI z#j}EKc7bgXZZvywVpbgak64W-`0Autmv{$pWPi-4(1g)nRM5Kre7|_}b3@8U^a4nD z`Lzp#a#k!)2yS2nKNAfUw3RZ>0U6V<5adMO;XS0Cndqbn1 z#FfSp>i>X9S)lexpu`)+%WIj(=*GUM2F}i_v>>N6l<)vzfdS^YnG^I%R^UWD=>-D-y@}Px~ zYPcuvegFD<_n0Y)J?FA<@V!if)c~Suxo-k&HqUmB)~&bXt!689mgMUMB=NPzXyEGx zwkJSBI19WM1bMjhu$I@q%^n&dOhY}G-K-0ARVzEDFD|)y970(VbnK=ssPp11Z5qQ? z%@si36Wsk%Hu&l#`CuTHt7kHV*gTo6hNXt&MG2kxJI8D3Z&_{r#~`?}P6!Rr`KN0v zkO+GZggL(qIkG@32Ris^R6kr2$7UT1WqTQ*28vrK{Js20?RIOuCUlax6l-uvxmH^# zb6l=A(;5gUM~(aae=qGIr3D_Ti8zTwu4vJs^zt9A_-PY8Yu|~N&>L?k)*vjQR>y9x z;)jrkmgQo1l|jArk_h-G7`N>FS(^GUpjJNv9RAn zUU~xs7YAs!I`QE*T0YRZ)!#pU@+yZ1ym2S`PPce!`w^67AsWHawtRc2;_kl}jvTj& z)>Dw+4x$S3UQ*xLkL+54~*Yllxn~hFq|35 z%5OPe5@46}g`Tze7lP7I=>`*+f_&qJhw=cZNVS=kC|Lx87`syBoKj*p4z27u1Mt&3 zB@R`1vT39Z76cEy2CFetN%nDe#jn5VDO;qG_Tu^}%s@b$=j{hiMS#?NE$TwE??D)N zilSY|s^G^7I2>yiQF}ZnDt>Ylc2-UOxY3Tf`!u9-z^1(2_#huXH}s5TV7z688U%ov z^YAfDJ{*wo7(*h}z;r)v?aK~LBDUoO1XuWx#VTb!0(!pD!sM}W!xi@e{nIkX z3?)t4E&r`5|NH%&;OSN^`cJLX|71I(OWob4jrIjxR%0B}zT}+joxPVSC;>EwO@OFY zWN!?O@UYb4`!^F%^g^D7AXU!R9*#2I`kr}AoloCN$&A&bY-a^he+*{sxHk7jX~6-h zzZMeoIEr1dSy@Jt!V)-7_nhhsnaPq8&+7v`$3*gfRw6H)P;B&KcABp4ZgT8Sj1tmT za3OCahG#Jpdsu$%C*CuvsJ>#$T^p}wdOF8b%qw+NuVa5W92We1E3A7Qe^3$4l=;3j z;%30IlDue{Vu;tA)-5XT&3UPbqygO2KNq$7W-F|G6m};DZ{IlixEehyAS8)xZcIBm zIo>hOeAnI~SkDilMtMF!u9_Y0%wcJvqEi`r^lj3>_6_pcGQ59r;7t0Jq9c*CU6ab6P4A~2 z!pittJ>uXUEC*wYuqk_!jbV4PsCwWd=2ea5R+oy4z;eay%cpVDi;S`XnA(yoXaVAY z$CwZw2?hV~0L;c0P;yHg^=g}V&@T(+EDbSRrMX$q_ItP}Ve%Jsx`DlIhs}|8q}4_W zKnTsdIUypmulIL(%(O8;(ew|34!PswZZmP=sgl|+*a`M)d=i|KVhEG$MHg|?QGWRt zl6SmqiiwR=a1&z12AQCt4L}2?BpL(QJ#gCG*AH6WG2X?Jljb%QpBJetCUUn7+>y*8 z7s_P@T-U6B?~yY;04dXfc71ANxQ+p>mn{45v6W9;0(~<|$9{fMb?W(ut@LOC&v3LYyUK}5=a5autaqrax{}SOnP7o*` z7V)P2KLInLNZEN?6GMHJ3YnDIe{|M5d2AtMs^y&Z9&bSHoUR}v3a%RCh7_~R@xb;l zh;g%-W6!$2X0JRS&)nN7`TSj|%eV-2lHfQ#e z+&N7L2LTp@(4F?HSbJTlIyC`+=rk>2W1_aEWW_wDY?Qer_~h!r)@Ogl`Xu+|3jFS$ z%x#Abt)=>aWD)V!JuMT(YfSU4$F5*>fW25Zn38{<@RWc8d-~Qmahuf+iT9=1NT~)wcSzZcBnOXY57%zmo52xE#?hxqlVMq;TvxsFl0KPk_|^yrGioyk29;F7F3b z((r6}gwCx%zX9FP?SK?C|C>#WRY`L(vydto`jOu~DU4xGdi{OCXu%D~EbF)Mn}q|+ z%Q6_ng)SN3g%QuqCY-nYW{jI$WF`#eTPmxT7l>e3LAw@R12zqwVNk(!`>H8)`8<;p z^q1W1%(&FC7?6KS0300oCpaamBvw z_YsG^f6y3kLn_+;#1`(tSCZOZM~*$wrOHD7p_*5UMb- z5@weR%UYMpqFh1mU#{L5_xE}~V3|Y1JvK1K?&rv=X+mr&3jFcNy!=IX+4?;m)cZy% zCSkGm;CwLoU)?cnMP&Bv*e8*e)QqFs>Nkh9I}u)G&IeLO?}1#Uq?GenA4uY<;Zc3E z>Mt=$e({j}ky64LM{*5rpxLrrDzhbOuWm8fP!4BcrDH7fWIz;_ zy&i8zaeqG1%ZHe_n%)#bFBiRUEFZM;(F_c6trR^cu5-Nj!Sg=-f?k|S<0)eYJ{ea= z7}s)e5VPE8%Ta#E20sjmn+UB6p<5MVq>)*RVIl}%|MFqu0a4$QY>M_P3g!rcP8@-! z`1rc#PfeQBjYsmddZVf~g}$Mr{{y%s{AiX1f%~ zvx5Z~^t}v3?)QiC_I0pO%M>O)D3w^#m!^CV*Ap}XQgk%xOB8*HZ1}D7VBW~}7PgAW zC-fz^MhP>-#5!Lh(X>q=NLru4el>zVZ*-7r{>ht)rj{?uhF)2|kLCt#(qzk?YT15kJzq%+6eeg~Ja zO{l4br6dZMy`lXU7OKT~@V`S3tzx!Y?UbCF%_8e#Dr2C_-Y}4(^7D>IDrb*(POo)J)l$u1vY=L_`b8`%0YB0>&|h**I%@&!(rSbkH!8{NF!U`<%nIjL zS^4b%4skX?J)>J5KWoId>IN}ti@qBJ``3GX>Fs3$$Y$fJKPm3nIcvXfDa>>{+bvTp zpB|rmoYj(e`PL9N;`gQ*TGdSNWJM+zhfrZQSSGltO=Py{jz;Ei2`W#tg|p&E9hFcz z)*k0}6}?Co^?)irr4uj44lSe-lbK{0i!&Mbr^_Gr?UX-YQ@z}Hbe<5G1Z8PQEnNut zD-2$}i5f(x&I5d+rE2%zzvB-}C4J>^UtS<)CSXG1;-LaZ-_r7amnoeLJsnVf5VFxN zbT!ufde4pJS!(v$8a5=@;VaNFoZQhGOUg!6Y9%~W?UhV+ZY?l z(ywPBTUueypmE$;^Lg<>ACbvcFgIflWHTt)ix1d6p9F6XFNztGA5HWdx@QZ2{j6&w zpqcI9&AZfp^ni1NOjFl&j|`L^RxTQdoCld8`qtOt$$D!0!#@TpOaMX$UmFIHz378@ zU^#))^1C%^tkZAyLiNr6`!w9!{WN%kC@7r}>NgepRV$RQx_j!b;r7{+tG}#G;7IPw zIzJv<-LK!~PSONQd7;uXiK1jn57V@2i_jGLhX)2$63=-*&fmB(Je$+>=*g@35Atr1 z2cLXE=K5ZS*Cal$*^PtZ(0&7FhrTRl7VvnLMJ^Y3t7lhzN)5w}2nO?Cpi^qk260`?N;roK+EQ=mv72mX>xxPLG#|@Sv~&)Rr_^JrAhH2$Ex z@NLpU2-kC@`_Oqg=$Jvv!aV@de4>mRnpo1xaU$zS8OcIwqiZX!XA9AdJC|*zojFRV z!>RZW8>gE}jYoacU2jt>mm}FfX9d&NYtmf9G7<>U=DnZVY{*uEn1;x}*bB(NVrP!C z^>@nz>c4v!45n6nBXxuKb$IJ@d-A&$zqGptSeS>29-Csh?Yc122&t`Bjip${O#T#V z<_UVu@pZ}iV!?RVN*82$BAF-&qpDUaW8SGQ&sV_F5dowqA=Zrmow-KMO$ z1>9A=3QFmsI&x}@2vOIk?x>fs2b4ahPF(OxG$Qqwk_sE#wzLFuLD>DCN@PD-?Z8NB zk{>(d>l*QC;K~WBT5Y>h!*|T9Z)xNHh+r0fAajYV$MvBcbN+3g{Pzl3_QgPcQ}x#@ z%;{xxVt+u|U7^xW#h@7AL=#qqJFT^4WM?s_9nQbWGuywViZ6ehQP*L%KEA4yAsdK` z@Fj-|p3f+~f+erSE2$@kLFTQ4(3Z$2o&Ks}X6o6U9aJvR;3wO`q{?+GG(dIZkS)T` zy`#HtRCA6=pCiBr>G^rb@=Tk!?aIjOA)p||Ij=v8tzq$Sf7}*ZUt)Gxud~g>nieZW zD(&vtMjDmqs2&_ly^3Q;HW`1ju+kFg@cA8&N=BEOff8LC_xjbmkr7Na^Tj(c6Wmj& zHi0i3R=Ux#zPymJZ~o*dG3@E%G`X&Gpi}>NcY&U(AV#lr!v~KTz-w-AG(xJs^Ak~7 zn*O2T&E^*);-Hv9eZ0z6TwHwoYX=a$hhN{+v^=(P(iC#7Z3|chIAexE^yo&z=JI)3Qphjabh3Z`3(Ri#E1#vlIG zW^iSgnO{7dv0-9cbf+K3b13NcSP%IF>ky%8EiSDl8j+g-BgM4&uXRsldC9Z`J?BH%>@Zmq$m7pU<$rI9s%!&mMv*9NB;4eUvBs znHDk$m0OA0HAd)ciwVQ;%x3W@5qV}h;N?Q=R}AOK&pZu1h2&@>sh&^!0phST0J4dk zXvhM9iitO0&jO6A%ZSe*L{Mj5!CKiwsCvd*ky18o_k*?Aps$d7nbw~CX1t-V z7fDmcU@!~#Cogm1ya@6A4{Q!me zeRMiDYe@tD7p<;IeYP4y!Gz0;J>)52ZdS(Wl) zQ~l;m%h#gSDK2__Gmp`&ETtK!Kb_NXUdaiXI@&!~!s?iGc4|>N^Ot#HXXnJ*vSitx zX4cT}K3TS(;K&q_l641wb(uX>6ykimJ#JiU+l3TSVku$?bJrq0w5!+9zABjb@%xzo zWL^3r5iobyr2YGA#n&L3$?xvJm0%683TsX=sd&)A7E`hjbm%d8R9Y#f5}vo*nSLuP zyOdwzGi}4HQ8fk4^El$0R`|Cf7XH}{i-52?iJ7K#)dq7;`I6xB>$KezETVjma+vrzxm<7v2us)r@)hK`DIPkyG-M@S| z=A25uxA;7nUzP`wE{tleKqkk#^f+98?rMJKmguJsxo){VB5e@#iwI!3jNnt0ztb_< z(E-79)*5ffP0wx>+0V%g+_oT!F^eR8!dH;L7P5cngT!7;UW5jB6^Q&xcS?yqed)2+ z(Pj?9-?VuL0_SpL4M$1h>vEM#>>*Ismgcib!G=ByPT?<~5Wh@F9OM;KZK1rsro7VKI)LrK)?9SQ2zs0J= z$qZ*cN^_zNrFqa7_c+LB7*n|roqtSTQ)?`HBqL0!p`2+Xal;XMyBm6fZFSS_8tKQ| ztym0{*5rE~&Kx%(#n@B8W&Xx-9d!A3Z<6w_)VFcb&Ie%^3I`ReDQKJec`W6r zrH5{EE6AC$exdOQUmat7NC@+MCIjInD=GA_sW&WPrzr!K#A_k4`0N+cOX3b(L7^H} z3q6T;-7r+X7U+Vo9674*RaH9vzEvfwBYMLU?U&_|1z_L0Nw|G=ykStFF)?IUdP~|e zNxXS9ADXcuE$_59>DB|JmXWz#xE(*aQ94$a#V4b~keBm);@KF=gAR2xa!A9(yEMes z9mtbW#2!UosbVBGDBzI5@M(70kg}-oZCF&b{B~H>8Y(>0s#rr+R!5QqMo4S;z?CQ7p|N3D*wl5SFoA6Kg%+LIlb(2_@l@D;r;9b_8~>jW>IX8rp`UZFv}vh8r5=7!MrdE);C1L5^N9e(qVIa(fGO89 z`!y0;$su2{fdtrZ7`xdKcwI{kouZRtIky6Hy0|tp8iHdU*ihLOP*nz{hjyqtX7DF+ z3zY#5nvIr;%Pz)*vodxvr{TX9?pvz24%nMa9{H#TyvslE5`-eAb$W`$oAj;!kr8rl zxo4ajx3tlWn_Ks+%}W5#uZb}`K7R?mb60#+kK+=8Hk0BLBYh(WftOBKkz* zngq1WI&H|&XA@vd2tN2@L|Uv3q#asY9-;;S;Jqnq=e=zHdwj3`n#L_2Ao^nhKW>b^ zV%TI(>R^P1L}!KzHn$>A(BoA>8)IP0cdP$#iu3=+DejVahpw9oNk0kOnebD}ff-!Q zAWJK!jQg_xVtMQZX1_%R5heAx&Uf?chS!vaGzKcIpe!sOz?d^4MR+QC`D{?+_Rr;6 zagknWY7*8P53419CCf#N^-2~NJ?P@>-N0;~;`}|-DzA)*bfQ00U|XiHc*HAMrMXHt zdB*_M-|h)i#av6i_V5w6jnB0U_wQE^sX&vXUb#q)dH=R} zme)azf1IcLtgq&3f57b@wL4c2WWTU}EDa@qBMS{}yK^7DzQK39Kx-x7UeyQdQN6>e z&8Q}eb0tF1r=i#ezsB~s$}OXs+Dmn>E^ew#na$+`r(Ei`D}MMPQ)nd@D(uRdU$i!9 zq=UmB-KByry*Bi=369>ht% z8nXIZ2HTwn3{d7Qvkyh2a|&lenH!#ROc4V-KuG%C2iVCB89b~ z=FF!6@@;uFlEf{cVG%}){;7|V9`#Vj{Wei1$V)_bD~xeN9GIDITyD<;eM&TdZytO30Nv;HQP&Yyfzw^j$6AJ*_N2q-)Azgv>oe3;8(nJSj|RhD4q zCeHEc36zNkGP_St!nLnenSZq=8(aGXuETSBE*WlhAD z(xTT&DAu2~m56!FJ&7=yD@_Ljix=lVipvkK80)xV3@g)L64~)7>NSrtxwQ!trmevT zhk4!WGcz&_1iS9)st<`@bSrZOA6nD63$%R z$g%HLJtx5@!kIUM!vRD9;qC!Rk$6&X)X9Evh#|a#mvO49YA@o`Uo}5eO^HgzwEd?C zCBWazAON%l-~`~*WotgY_?wj3tswbUb3{SM=zrCVN(%A%7&MX}FaKW93HSLwrZU?Y z%%ElH(Iu|qyB2@HG;_R*rH3-JRQ4NzMwY)xpDtan>4?V{;6MqfAma>Kul{97T+#_B<@F(OR}K)6qAR4uZpg2hqwFlA^JHfOKr;y&N% z!CM?VMyNJ)O7C8k!|nq({3^iIABhDWYtQk)Mj_imBNm%8FR<%VX*UIqFJalV94%zd zAQV{9-)*<<`pcKTF&dc6zES30-NI(=ozEX7JcL=V>PR=;D8+82%NKXzxeheL_& zCbUP1eS1Kkda%!jbv_8~(+GB8UGI2XPaHM2?}TCPB7UD9EEFHL9d2I>rlPkbVhNpQ#g?G8>ty6l+Sl? z4fgg5;rH+MUp_?*$6M!lJ8S?~T;Ey82=v}TA%Q%kE8nQuj_eRp%F&D3YKs=Q&3YO2 z%=WCh_e%PQD@yY~EKP!ghCH4P+R4Pdf@QoK1Oo)D%c`htCeEvdux{qduNQj=7galA zg6Q9sTNg@{{u0hx-V&}*;y#k9+=5L}gp504Cv^T6crsZY$^YeW`hb!uD^y(+Rvs4D z$XWTVQcL`&x)d?n{qYnEFLQ&Bb4u4}%XzB=_hD{5UuAK;`w-&8%)v-c?2DGag&mGo zKnNd5pK-cfHLNsk*U#HBh1ioLP%o`)J*qz=jls3|brfkm#4%f}3*>t=?Ej6Mo6uty z-ZIm0yevX{%dRdlNB(H+Ua!pzaW3t70$oxJoSf;+s_4vvb!!$Q)7+~|$G4bXIZ>cD zvh0lYNVz_;(!Dq;qkCn*|axeZ{?V!Ac@?g(`P%%?QbyGDh&8 z-?li%!6MVf3O}29R>v4@X>R^B(r_GnX0Fa(>JP(N-FnoNr7`?}vmPAP&9fMOo#{Px z?UMYPv%k3H!#-a7&W?Lva%%q~*^2o#+MHUmcJuX360gtHuR-1HSiXIhv~Yy+{mlWf zyVsRm>*XBq#IhnYNOYazU9tu`ybQda16$%)u617(->`ng9)28aHqBTuH^-g=*KMmzTY|GlQ^b(tY zv79vVbk@rF3m!w+OlEr-*y)HHleDEg_sa`Hya9_KA-g0F@Fn$iA=}w`T6tuxaWZKqELij!*FI@t^wKGkfisEe=~887y$Twc81;7d*c!>i#;3%NmZJ1PKHj z%!l?Jg$XMC|5*Swj!pJLpgrg%3p(&A1xatcH_X@TUcKJheC3Uud)&w&$fI5X%^I@s zC?b;>tDz#H__u47>&`)}lHc;nk)+e*wemNVAw5J8K6J~cJM}7n?p3*RD$+8+3_@aF z&A$bBk-wolmckib%c7vbqHdK)Mob7XkxP~0(uq9tid)NG3j=AS?~VfK@d$jU@tMg- z`!4JuEJ1$FJG+TKulWxiVz>c<+cn0CC0-@5yp1o^OCKmVf7of|v+1>1y<4dhe3AvP zgK=Bhb3*nde9#>Lq?tk2^_(d*MZ$UI%~CQ$6w?dt>hLG)loaUKI`?cP_7g`@3-9#> zG={?bjx2B=4}3TjKO38Idt;BTsM-~iOaQV=gaPW1N~WsvZNSn4_(lm2)cRfu=bv@? zOw-vrGeZ@$jdjZ)M(U{I-;Ya4?s@JRI7wKHF(N(Su^#Al{ImK$eS9TP^bAC!7RDP` z+4ZZh7CJyr5g!sS&5;aeFZ9CaFBTJv>O1yMcVDW%Z>#~xE&C?Mb$r8ov5J2|=KuAf z2c>+}&g3%y6pj*Z&*-?0B7C5BQCF4-7t|#g&<@ff%u1qqzO&-1V91!~9O6)Id6kcf z^i1b831$jG-KHP$C#hhcwnS|z4DjkG5AC}XYL&O|6a-5tGo8zqE z?sMOd`={^s@%#&3*W)@ouhZ+IA$}|ROaCJd8Fb}m=L_g&;04_+u>#}(BdFg!yKYnT z2929KcO@Y4bokg?!tX@w@KK2u!8K}o3RuiZdfyv^NRKI3II#rqsOk!wdYuSRek*$* zLw|Kdiqo<)X(n6`83s%lp8BE+?a=D!y+i!)oJXZ55)5rMi4EAux)J1(?Eg7lL~1|@ z7i!=x+~K-b+x|>x(?L%f+rKnlbdF=G0q9A_QNw2=Rl8&zsM`AfJsuM>K)i4~q)OVy zt3|tnxlfeJ9e!1qwg0O!>A<;^WpzLFlDzK~hvm|Cs}Cb@8qBgfZ~pLeBjS+;?@|CZ z26gaFhgGyfaqG`KBcQl?m<$%_uWzXLB|Dw-5ti%z`&A?pYji_9;o^f zN3_Meq7}bL_Y-4{j3TAQZa>qyN^>Qc82utOdi#lbyzSyr6kfHFbakDNke5NVLI(Ge^=h#;CjC!KNAF(UT#E~*N3uO1t`DbB#Q8@q$W5F zj1{-o8>wUmO%r(xA6~5MiEdm8wtQ^kSG4-7Ux+yZ1i5W|ze%zEv+~`|vvH$reaR|s zJZ~f`)DxSTk{Q0=zfv5+)VCs)Ci^kI@vWA*oXjAK@y}#5J&TK{zx!U8W?AXzHSh6G zcr>gOwl}u8CiHy;q(rm`dWsv3CT6Fr#3jld!EDdPY4lc<+AtXVAFBY}SD@AR9I009rrbJe+khyz$@I`zDhg zhet5)iy4%wBj-UtM(o)JNU>M4-412#+E`phhmsUhnp|RvEhr~>rFn9+Eu5KJR<0i& zONoJObJf$lIQRq-!o6#=ZIGWmIwRu6FVXd=Dke2|*12nhenPY-wfQoO5L_pvvTe zz8eWoK+32z+Rc<6w-8#MeLeG+18F}`W$-%WrWxc!cL#KN#Cghk041AX$s01LfM1R& zfqZriBWXx>vqbWDt^GVWIdmYw|9EF5DL@P>%e?;cF>tMWMe6Tv&1+3Ss`~p?;$v#h zh@tpV((|oKZRUHWt+uxY9)lTr>Q1XdOj-$l(3()cpsp)u2=Y47a<4doX#{dOIQ7p< zW}&!0(QgzBInWCWRw_<2V-!NSCrwZ1lU_FYG5dmc(+pc4Iv?-wda^2<8aj)i1CO2) z`IwCJS&7;%F+bif-hOShoRH}HwRya5&QAaEYm6Ys*?ei>?78%#;k4l)DesMUX%2^p zlK=Elk#0X;X{@&-Be&mn(uy`$kaRuhXTh8uc!^-;sHy*0e@fNVIte>>7u_rlr2j`+P@&tsJI z|9H-}mO8yl_C*%iNNyP(%xCzQ6v0y>{!UCkR!**|weMC?RA>})D3F>*XdNjO(#^t~ za^W`I=GZW~^~8_i+&LRbirYL9o3Gr%#HJUNw<^!#nW}Dlq`(f4#2+E4BwpxNZ(lx+3D(>jnnJ&*d#M&U%^E!R+O;** zs4|&dIA+Gfs4tHwi4b%dYBG#|*?|&sD31G9mx^CFsZl}CG?1EeazS0;9^e5^cO#(` zfe6pv*;;6YqIySuLH}t2-{|tyQ1o`+hOKu?q~u%A4z-<&%lc?)SpokaHJVFqM!evG z7ap{T#vq%dm8LD^ZO}}M`(gCu1${ce)(x0dE`viDlu)BeXL^H|V^rq$1MnkiHMH^~ z&8lTlN9K79bLU0y<(YTjeY&8&+g>8VRVwahI9VmXmqL6Iglt=EB3L%x4bpE`j;Q(> zG-(A5n*)}mhm1BbG>%D1FuV7`n(=oKlMmW5IA!&su}9|?duVZs%+a!=N9N;YnjXy` zClVrcIW_|RJLzu|l3~8{@yPd9O6UH!iF@_XWk@K^9*p{&DDd4YdHLD2Hs`amQR_yu z`=@`8JD%gz@o2rM&}dbWi$}hP3aVSOR`__TGA)_8)m&v<*n&$!uP0oQbD})F4})ZH zRL#1}$hVA)3g7SHye@%PMnJ@Qq-pWsvAeqsK@mSjiPayV4r4(}Z(WDAc)K@d`wa0m z({t&pX!ZS>`{f_`y}bS`G0>y)_jJ!if1Wb+i0{q_Vr9E)iapugYgQJ47%jHz4~>zwv>p$Z~`bdhko6THa%8)E8J- ztUzNc-;p_e<7@twPjKebvXT!eE3?+NVs&7Zea}T`({M(SU86_i{e?>-HU*1)^x}|R z>#)n!QSqy9X{%G?WtP|R)lcxF)6i)NHVFS;q8C+&sKPucPjL)UUfMC7{HRd#;HudI z2kDbSSi?D4Vp0d**K~iU559GlNr>ri#PUN#S39}OgCJLwml74iaeaKd57wl6-#au<%@_f_pos%r1WC z;WH$+8U-zR__xWNIu?xiU*CNTNjP&{0U00v{KuS;$jXWVF__D`27g(7-r@dx@G=N5 z^7q3ts|1;Md-*goK8Kzzmm#|_@0Ts`PvC7K$Lk75Y8Xp)z6i#ElI?-6USvjy&pyRe z`)z*jNAq3f;R7RVf9Z}PJfL1NP zMJVRv04WmXLD4_xFC5Y5)aEnLILYqEok6%p@Xi4Xv`bA0UQ>aUwM%@*KF-@}~&I9dkFKDqCynh0CVXXIM z$xUIZvm9Z8lfkKRjv$K-@rY__`c!xESjyXHKVBXlI08j)6><>0VFU7$_lhA>C0Xx8iZht zowYf3Xbfzs>w}an<+6QC41f7N_@ACYN74@0E{1l99zNerDOi9V^p;!P_ugKPF$6UV zULSTUf;<8eAQaIbAinYs*8VZL8R@gI9;Xj#)IP)PO2ej9FYQr;N$t0(-?G9v2bwk2 zZKAZktoDeC1#U_~pLGy`2jHJ-^hiP1^lkJaCr>58XZpZ)tyWk5yt8(>tv$ zpi(DL=;Oov#ZP#`si39rVmb$`jmFJiIEDL&vMG6IZ`vQKQ`hD|Y!b?)Z(qvPSvH_e z@D}mpgk})K6YvbJ^0&|&(o1Cg7d|dlte`0dL^X}P@UD58-)>4SZ0c%*c>k``0A?L1 z9JT4(GqWepDo&yI;{4+c-@5;AzM#KuNKE7#UE4Wz$QLEUeL90V0ziKNU8cz|ui|;t zSUFIV{gZp9;rJn(#`(V+(238v-s+|Yj?C)hD^g>=*T$R;apDE$R?wHQBG2$~R{!Jn zHgo|-JYC>BbTupADQX{mCVxwECF46p2|A}zmrzV=D^9iDOXV0xYyJN5z4O!dD&vFb zI{CaKp*9T&V;M@hhW(sB!}mb#Wo$2ILAk2k-(%aHX3Qsmr+q^{>RqQ%9>t4pRcIXI z{27FH1|r7!#+b9R8yZKRm_3i2pA^$)z89{d-MI^nm{avXW1x5Hnl_fC{hTZ`p(F0E zggi`-(iLrHoZhER__Q|SjA4{h`sz}McB!r&!75DBUl#xK>fN010hKEdLW;mwPm8DC zy$`eM)P5D&`9dafzBn$>AndTAm|t~gO+RPm>NA%dwXKBnqff)H4ja65XAKGK`J`0> z9%RkPw8}J+HpoGl%QklCa^ODe6+o)DV1CiNn@bcIb@0aC>xl{CA7(UV$|ZRmT#Y%UYM$ee4Uo7`= zq|H}`M2olPOW!*s_xV4^+rGV6=*b06eUvfW&?d)%OLgT&rcy?gb+NLTjZHyMS6Qh< z*uGrVX#d7_wfx_*1r0(>_6MzJe=A#X{Rer1_|j{dZpn|m#Z?+$?0_yjd+ZTZ2JLzu zu|r@3qJ(eUS~<=d)xb6UWdhmBOV#?%F<))>_Rl9_i&F zXLEZpk-$E-}6`BzRns*8x8Gr>kJ$oao_(oeXl-I z;manJYNK!aJn%IM*D!CBk1X~1J-B!p0jlujkUcVqwm*tmzqup#B@3FU)*b9MI&8Q@ z8k{o53S@|ENCthGkdNd_?K1+L7jYNyh43s>GT3-qf=dPR7M*h_HRD% z7CtrVgAqTHK+-dJ65#vDQ(Myzwi{b@Yh$1PCa7WX_!~x$*{_W~gWqH49Il+FV z|99l40@Wh@fF;gPhDjUJ6Q(aYbTrfY;<(;Qy>tpD<|LsSMqn0R!?Pf0IgyV=zlggI zNfBg`IsR6*I%{FV#y@!YU(>G`KjTsaxw7NQ!YP5^KCaUN9$LM#5UDN~|2TnJuw)p- z0g)K)o3Xie8p@O{>brN-keNn`s?a41(ll**=w`{`>WY!U20%s=4zG+e`aXbWFp*EJ~7WfbR4~ zh(F~GU|R3gJP1K$<3NiS37+&CXk|<3Ge+qjtQifoHZ}kacLod|RD9tU4W#f)cOM;? zK+nR7oX;7ynDaz7#|+hBr=dAtYQdy!KzdhV1gk-zPRC`vf_MnhmAJ7Z96k6_NmTEhMps!`(3j`*6;F&E_ttO5qSQ4BE$dtS%0F zz-<};09CYHn23gjy;|RWl6)JJqKwn`b5}wk3f9aVJ?uxWZMa&72N|a{h3a>bx8qdN zKa9<_clz&B4r72&oy8L&A7z=3wQ1IHNr34B7cTK2+Cf5jC}tCB9VFJasV}j(DN(l` z2ZJ~ysWtdrIa<04x*4(ahQIc`n#;9^hqAUi9aHToYIs^(t|#SeK{qs|(@L{`)2A~S z;wgc1JnBTHGBlc|`meP`0<9M!?9*4Xv(`c((9Qx>*Qdq>(L{j<6}%zCwe9kWo`6V3 zF!|T9*kZu@-dq>uSA^wOh9Ji_d^brE`fBsk_RH9$JKG z92Aq?SFX)=vvXzM8j3?|@*%TdiqZD+dXrI7;pyu|I!mQ=8P!IEs zo8#}fikZ^il0*zn&Mp6Y3F!)yt6D^B;~PeywM4Jt#9My$h^k!~v10RvB6srrPn!X; z3ar8J9XIjd?Qh{?nC7O*j7;-VC4qv!rFrwA2G{&Wlp~%FeP2nlx1G-1ulK~4xb3dL z@(y?uR+~L?f6xB@#`t@AXw7Pb;6B1#wm&l|6P;x<0atUf*q0Og1~0DIETw5bbA_q7 z73EDA*;P%fxs0t%w7ry4Sz;K6PzeLiwtJi*u+LK1T9VgxoAXI~yd=?itnTtgq88N} z%!K@+xt5&6su)dHlB?>jB}iyCtqxtK%#A;OsfWVEHw~CgFcvtFbz8Pm{5$pwjcpz5 z!|&^T=ZleBzVla`p?lO+J_lF%7zm}dKv)x&t$?FJMd1$7k2n=-WHyt;7^R)Z9e&`HZHHlR`(`@ZzPJ?EnK~Z;l@M zL`!IpD@T%LR#lO{Ur5`zf)#St_Wc)|45tD}AnU(8jTraK*q^T}we-?JSfrVFIPw)v zVat5AzpMjTzf6uopoTbdy061f&%%Qo9Rt2tz$hZq2D-^pG-Mx=i@k`8L}a6f*kLs(*ff_(ia_EfRa3V4`Q^C5^dr!tgkV? z-j_&I17 z&;@jh#}EX*!q{6;$7rUKgaH|BtunQ2TK=ooemNXv_=qeTLYs?ek68n$_UsWli!j*gynt%4aD zJCgnLqYBW;##g-eHFYyGbJYc8A*)WsL$eb}bjPEuEPsRS%??vyP6Y0R9q>(?xs~Xg zNzHk|@+j3$8nt3&qor%Mb}lCtq^6wsM%}{ij~Z(=8*j37^@|-t&YH|R4(8`C-S#_v zA40&iOBJPkr``D{Q=U`sY=y2zgYwf0QV#8pm)@Es+E6L8mq0etGVvx_-p+3^&HFz# z`!qf8Hnf<>mwwoFaB~dmu-Qvib8&oMMl8UYl{s}wEv3OASpzA$Qj)ughI)T0X9gZ2 z60al+5-{nKD!c8G74iM7{RJ+YOT`_B*RjqL+2Rb}LG{%b60{91%t!5Vw)Nz@5 z;Sb*m5T<4?AG9?bk>r5-e^#r6!E@NjitEr(<8zPJI$8uID>7oeUfH^No@wfy6&PPi zkG1NKvZ1~R%jRo?JF1ItytpDowSkSN14mso?k8%wbVi+qT})6|=J#4BwE{{ckWc^4 zC8$|V(M=uvHsfHKz5YH5wPZUxy$_)cZv9IYUT(}*ES*nn{UG3S`;gCowzw;hJvA%5 z*BfamqV?)DeY(G%(BS(t*)65K$mT)IWrtbr;{`s0WuT#+jI5*V)CJL|*VF>R?K=vx z@&g!hRR8uVDLWDDIbqkatRN?TVMx49_Gj5fHJo?Ym6}QG4XH%eiU%&`8ckQn1@cnk z?RC~(fS9Db%o-So{eQ^f{9hF={o+%X?L8U9`z$Um&jD#2CC0@W;nX*r)*@%3of0zO zu)_dugXlYK-(u4_RJ^Kc(`3}QUe{jF-HQgY-8PPw?r?97)l7IEb>SSlVW8m;ZGs2U9i!J zv0{_4zvQkJa81g%qR{q4qNrY*LN8+OsqxIE2ckxNVs85t9D&_l)&j{Mb)w&W_U!%E zr|=p_Gd6yfqoq5W;adlSYGIEqx6<*vFrW9qe8$Z7jK0{NXl+~j<@+cDwTP_mO*l?? z!sK0Mzv;~CwkD5=DiQ;_5)63*{ zshg1Es>&zh!s{a5%nUn3i1P!xX_ZsEPO}~o1<8FHP;y(I<3L%TOoB$hZqe_@(CMH1 zsgqa7=K4&te2eomEFW)CWBNzlgoOzA&i!5xn=XFl@{28%?unn!DW5Il&qv41{|#>1 z&ZCqGJgvx9NDOOgY14Fk2}p3l7(q~8teFY;MwLz{xXCF#`MQ=A4N2%{!**yK7Q%z} zRo)&v@SbQPe>(p=F=7{SQY)nLvjX2kNS8`J&D;bmy8kk|^_v zIX~Q7e?6jncTWbXQnDPg`TriNa3f`Lnv1<5r*UX)b)(`y=pAFY;G5pc9mUo&Qyvxe z%nC&g^X)58uk-Vv5^)kU)SwN7GP zy`ZGt%zhUS91bNMMggRif5AgwjBA%F|@5$3OGHvn(AYxc~))B2ak z(-bvqGfMNgB`4|mLn$jkjwDHliGx9C$0nq<%O(H6{o3Zk_4XTKVGRI~ zL3!}@2i(RhAig+)n&Plzl2dO2+u;TH&1NMPcF4Dh2<&b}tW@5~`9)?={B2VUrKB9# zHw!xApF2rV&8N$QMx-Pu!Rr)mX224Hp}rP%+zJr?Pz^@?Z30oHf24~`;a42N0N2)gw?1*0aG#jR*%xQ`lQ_z z^b2hrnyEt49Ua>8R6vC|AK>Wf+I-8PsVwqiZ*aua*S2wTicV)3$oqh^EdfnQB4PjlJxGZY0)g?O=5t>cF zBMknE!}i`f=3ILNV~I*Wb+KCXreLl-a7mh^aZR0fn^+sg-W2TU_-_eoH=^;Ymy6{Y zKRE~r>(B?mYGzj=NLLO(6CQ<%30Y*yc0UX@*1n)ymsBl?mo9C!2UFEVdb(0dSi*GuH(59@5orIN?S8mJAt`0NXCv-)bF@ub^G3{V>^Pv{ss=qBGU{TjD38} zZs2s-zU4ztG{LluThePL7TkR4kN$ItKW4z=%g^LKxdrHSf^eH;GpYYRdDS->sM|}Q z^M+<4X{X?6lWI6xN{0$+d@!!%e{j2w&_9JQc{{;?YO_bPv1J<>otmj^F)5tQmo}5! z=j31qL`5OmaP5&AF_mA4-{Lb2uO~bt|nZK=y7;=yY z=g&sQCFVc6tME9dHS{OQSEc>?Gf_wj#>zg9V`CEsO`}bMDWdZiQB5wP`BsGKOVRvR=Y{It?X4+0s zlF~m$4I%RkcmI;7-x6f2|5o9H;DDBeT(=(mReh8`y}$C`X!T!v`j?U|kA{?di_-Fk zWs_s0&c!%)8M(S&dtRosEz$+bPsSFXyZqhOATsaTyi4bMsuzx9LXcz4=Fnc zJG^IlKTzdrcMrJLPcoFzUfTz#=eUr)H9=e8lhDjIM`JM8b@NGnsqQgR&DwDE--+oX zq7#yq9f3s5U|B*_GwkjlL)TqN`v-4bW?cMFdwHLsiK9p&jv|ffmAI=8c-!&`+OAgQ zQavEp*6y&#J*InB6-oLSx;{6RJyctO{AA~#zS<`Qs04eZ zVaTEy&FHZJF+k40-y9Ym1#0RV`Q7mbVYJU20ZOTC^xZ6fZ-#MoEkF2>w6}C)tsNx; zZcJTfhXep3!ay0sgri13=(?C&jr`{Q#H!EQnZa+f7U z5}pr_>=S@6gI#vdb(*00bW#MqQf?>RO9&K!NC0QInZs7f6}N8A@i(++HJ*jfPgih^ z5g+JjI}9L97|u3w-U;O`sOAY@=gmqL?+}i zNv6|FZ5fFkrGR#yG)jTyUTJ<^rXWSAm*#FE?m}yA-bA0sUrtKW>D@KWs}m7`P-pr$ zydioA4|gT6Ro=nOHJ(mb!A#y_%<%pDphjAy7Y#(8#wqz&uTW@EfahIap?PHPV~Uk(fxkJ#}SYa-9e z!XEcFE+V+$Yh%X`15Oa6oHi6@23{n+aNnPu_m7{V35dFR>Oz;1VJ?zo2WPy!7Q@PAd_rCde>>$_FE3i@IqeS-ziGlxi>DzC$(GWSaN^w<^tU z54`ybE{PvhjQAkH^l1~G8gztwF;E^(V@g47md^>gZi!0?9d7HEGk-cBX#zhpZu;jR zuP@&2Z^dh@Pj#bQ;vC;`c+IcHHIWN-+sOKom|-I1(T}hj!N@NBleKc|n)6Z`fWgXs zq`7h@lrmH8tK;Ng4%<;&9wVn$Zrb`J!(1_dam}|B&58DJ1^j_a{ifYsy~~x51``fo zIyv$OA&NzO?#=VmlfTj*=Jad&K}t(U%*l@)Ln-~s-?qy9EEip0FZUU`6t{0C!1~Ho z#6npmzPAbj(o!_lD%t!q>P-!)x%ucIpH^*93d>;H%)viKquVAG0Ni=|$Vb*Oe9)a$ zewwZkF-v>x)s2p-Up0dE!H)%Vns_9s{G4AM^!dtvvt-pk_tXlbV#2#X>+d;xa$DJd-T*h|Xpca%Z8@M4#IKC|beq6$Rls<+i$TYQLRh{+&rh8X&wp zu2?T13A`7XzqhgdlToCg`93CCGu!cz;R&vYqlxU9m;pJZyt)eUutFK4J65o3so#Ao z?PvcZOc^Gww}0*=x!m!l?!)m`4D>+4#k!<5$QA#t5#$d9yEZKr%@iha!tn3VxYGSa9CduJFQb z)nyK~>g({UCLkM@rx~{w*DnmtgJv(ys}_Yvn5W8RcVHE;vyYvqQuRw$UwPvRHyopv zAqGqG4!dws`uir${dB`0@4q`}$^p*Fg}r`S49!Q|*np$&VJ%CUd;rY{uV>PFR~j1N z+|wf?wVZZUyD~ECatEvh7P4iy)%PQ!ofmpug`iuF>s zav7MlOZJ4Zu)vBG&j5N{{sodku$^zH>U&WoDM|@I8a%!D89Om3O?1u;yJ|q9HwF zEJQL+Jh>*6#RhvqILUn=>~|}GlcQoU!}OoD_yV{7>Z*S!Xp;qsFXi%2%w!NwiB7ZBpSXXzKF&PpxhuL!*169Yy7&ym=zkX);H`x@~GDg&V~=x zhbwv6gMIYKpBq#{ovit{?asO5s>V1})$+=zIz;Mhe>N zxr6Zks4Gz5b2a7>CU;k*|7vM{;0r+TdB0_1i9a%u2O|1#V#fD)6dnrJOH=pO$=JcbrksgsCaS!kw3 zoa4&FdK+sR;YL%$ggOaC9WMhwv-o&xpS*0|cka2S;gi8v4f&8!u1 z7z4+jOkh3d3t<7s!dpufZqNdB&jF->!XEJy$xy{9=$tE?pA9{b4T0Ev1cz9up|qY+ zY9K`Aa}Yy0$VU03>H8oi%~V)ldirb3h)7kRd^O4OkeOIov$OhD(6P&*());?d+qZ) zB3&{^pPb;^1xdQ+jo7k9f-N5mii?*yHbeg|hW7noUV-k|;TKw3>#oEh*v!21={U%| zS7K%3yp#7X!Ibv)dwbc|-R%9tOZMG4C8ZR0tQ%lw=GfsOY%s<5&$^Z8O>Zhqffz&Z zVSWMxT_jjO2B{ZExQuZK?sUDu$9~Ao18uI6qogHY-DIsqn`kw-UOPD8Mg3s^s_8e}4QLo6oODg0d_I)4-J;&xo3!w>L0y4iwZFoAhR=~b(1yNJ zlaRp2$p!FvP7s$@ zKU;9jTHN7V(AjU*%%H6AcfX5lWCw>YJkhoPD znLBP}U8qGYDOyc(>k5+qu)9#eh&rAEN;nCIY7~VJ+lb09NYh``XM*Q8ckLf_8w| zz2|04UU290_e+gZnG;o8DdmMRYTh}U5V2?HhfKp*w(aLkSW3sPtw@U@E^yx>YUrsp z>#e{QmLkyZa@Xq30L?Z+E69PNgB71&<8^^Gn!+Hl$d;M!bWTy7_fD!f-Xbus6=Bp0G@cZw~toNsSyF#>gQL3TxP4s9I!zxqsTMip7_&k z#3djE-4l-FZq4B*ia%TsVmg`7GPl{7vduajkK)t}(nM8umrTG!?sGdM6N&2lvSDc^ zp#5cQ&_4ZIc*T-6NP{nAPSF5o9fFvq-A!4)u~K9r7&zP7)FGqZFD2pJI_ac3?&{JB z)mP3KkcqK?7*qnN#Mab4JEj>{!gH_;QclbWx2HeVIx7GlR?-e183BXMAykL3?t3*k z7mwa=A_pJHielTr&0otH>Xw?4iQI2-t8W9FurPmuyOqs+APa55W!z=3NVduj#< z9Dl8X8TYJ?(=~PGD%;%~-7O>)>Yfx77}+4*Vun!aqkY(Ottt6CA64q<3rP-s zo}NWhte4Oq&rK2vuBCXTYn+4JcIqv@3?aQ4zIi_^)y0xtm>iIJ@NoYt`qvwKIcR*p zCit)T!=RgHJL`Q{T)N?dTmH^+uy>tiI9o=KM`7`A_q~CrLf>D{evp1Dpa?{kNwc6d zrGgNQ;hti=7p{`vE_73V(y8{rG04Qd=^G(=yMA%Rx+XfgYR;wPoTH!^vM1C!2fhd= zr#CPK0XF#V`ME}*OW0GDtpO9DCAd4EG*6WXc?)3WUIaUbm8v)-8gG5C$hWCk$%wJ~ zm+T&h2kVJqX;};6F7UhkVnq>}+nJ*csMAB)d<+B*yEzAe(wI5}+6#ZKq;d5LjeFd5 zjR7l95&#a8m+4Pf+W28Dz|5UcH^HBXk{>2lAl#nYalH-_R8n%VGl?$`YqrRav_tr2aI44 zD78oCYg|@ykOE1MO5d$-#bDpaJ=ZIMx8Dq1xwQkAo|^s5E~lGw*4Nd8fzt>jiqo+a*V6+>2YrXZR868UIF zxc?xRXq8Ri3)`*E9LXGhrw=g=p2zSxyv`X5c9GIvSZa8l6;96%SHRIRhug7^n0lq0 z|9wyp*D_-CztY~hkl!B|uXuP2RD8G`*>4yZ&lDRCN89<#v|KHed7$3SH^pfqa?Nnk z)x^Z;{7gD0d$fk}TdrCJ{M6Uv6UHZGcpo4ac)oJkvtRs!8WNEd2Q!3k^SESk{8 zInbwQUE|l5*Zm$Ep>FZR8L5*d)`M0Ns7!Pa5W+0h!bX&z;^gb)>d(I&6CbMSfP^la zO*1fXUKQTMdHVAo#r@Gv>m$PCuV#*#zh8ILOl%6qzcxF6h?t)JqxJW0e{h3*MraSB ztYot~FeR5+dj0HSdWzO zM#+KmcXU`#l)^usT2fRto^c19^xd}nFo&m{m+|0Qg&Cik-j(Dfi)XDS{X%Oa-SWTf z0<%jQF)M@bN~9(mgVYWi)Rr?OLBTtozTKVwSn-~WQfH;FG7j)Zq#XU4Nq^}Gxn%d5 z-YSxSOa}i4m-Kq1B1W1krie_CieJDDzGjs$X6%r0I5X^wNHXnDXV*)ZmurGaP)~6y zyQ8ZYTmPFwIKP*2I*Ej-SqWkOCUQn>6arNEo^0ue-37qVIepxy8Nx!N5;zSu00`S8 zI%fhei{f~zmZE_33ZN{HHpO$SM%`p7#Mk_z0ea=lCBhgMT#EYR?HV|ICAc*z(Wl&Y zM(g3mD&gI!spVgfGj&^x-9btLSHYAwO%+%0@7=9I5U8iZHwB7%-xP^nf6ZZfq;GMK zH=?J@V*`=lqX3i~EE8)r<#lk~`yk086!io~=-SP-#c^^;#dA-W*zm5_ z{#kQFtOeu&@@S?5WQeM~Gck(<MbdQR7uQU5G z%i*od%XP8hQV5nboZ>rwB27{s9|x9PcJzzI@bgtX?aWWSm;EvV&l?$+Bz>77`v(=L z1rmW@_YW6~n0YN(2;}hUqDr%?dDPEE<+bJ}T{n+dO68hD^_tEUe*YWrRI84@lpiAK zzTE!l{ErH6>afA4m*o0^3`AD@h-M7)DK@L1Iv#ttTdz>lycCD5ntiIG;POC`eQX{6NRng5rjWq$jb;_sVFtDt;(VbY#pJWUo1=GlVH&#Os@khAx0$7sNx)$FfT z_W#mj%msIMi$0;w#7D@b6hNu^GbsvQ2YNb$tIo$zYKNF_^=Tf_?%+cm*AK!$Klhd| z*tgsPmo*)ZPE>G7oOW^E7TSbB^1kpoCk$`8+9xRm}KM{F4+gnQ9iXCR&P0 z&-}S}n)b5P7h#dk@rhM|0&&ktg@o;n*tpk=!4kLE=Ur0JH)0#`wa-heV?Gda<^7Lp|1b02BAlFu5F^g0 zXusB!6RoaO#FP`pr)KkKuHBaAIU`1Xd79^B_DKuQ3(@L)I;~Z)SttH0mn%0iG5X$Q zG)&Aa4&*ZAV)Tq&m@%?WeE|*G;Em*b%r1=8EM>pC&v&2Gcy*4enO&Pt*El!n{Qi#0 z=hM>Z_Nh;0f4Vbx=HFlEHn#mT0|-*N%=y4)D60eV4XWYwSd*vF^; z-e~i+AoT9gIoj_OORLt;E%y+LH{t{gkX<};$R_I9l#kW@OE#b2yVNpihkC@L`cRZ8 zqm{=CPq}ETaY_AAna8SdSV6UdUBUygaf2fN{oxGC*9MVW>i12dOcVHhsZG?DuEa72 z7}akzJZ5F@mZ<>BA>A&`KYD$jTyF}tU*FAeXjCW+21;TU^7Y(Akt8_p;5(e$yUsWC z8fRU;TG)^vS*E!~P+6LyrR=0U6yibY%J80K!@u$8D=sMj z4wpb8;qUv%qYBO5vL<)?<+;Us6$MWt8>)@=j%x>`LCtYZOotsh5noo|(U98jkt6_ypE&iu9{TaKr(~Z1F>}?e0ZTx1dCU(68;~ZNpp*FY z+2#E$-2YbAidC1cHP1hxve{#eDs;H*kCiZn>z(Ka&K&bX2{VMpP7QS$kbCs2V{4R8 znK5Uf9q2D=VriMB)fr^9sU%AAX0R(y=(Cilp#bVA*CkQ$u3fIV5;l3cP@7ixIgW#IBG|mWgi^36W{=?A{zZW$Q zy#lhE*V?O=wd+1snak>G`-BYNIveo$d-$%*69C74$O`y4Kq2{~{!wKAl}}wAQiuE- z7;w1G-`Qr$o7zWP#k5H9xGQbQamQp|@HdSe=qC;%YK7iaRm2D&=^V9?d!zuc+Ej~r z+)Qg63#r(!l4@y(oP5)zaVuOZ}Bh!@c zSTnzkT_AvDl4V)!{{W}Ep7}*-RwdK}DdxC=d(s`ojOLv-h7;8v(NBN_`plUKPampq z*EeP4|3lN8M?>L%|KrJ)WQhzKO!mq$L}Q<^MfOUCWEaZFlB_c$#E_jNSw<;ZWM8sm zH;gcr!VqQ^Kl=~bMNEst{L2z`zf29zhXG&Db&^_ zUD$WN9)4Fex9@}XqgIPQp>b$!I=Qu;K!yjzoxGI%%dpG|o{YY9VxDsxRib&L#eG0j z1KS)Fa^HQw60+aO7=qBHJ#wMq2~-@UCFMD_7pF<g(Xep+eJ3F)1 z%l^Y83|(N)J@rz>;;kBEn?((qM_|1^xy=i9UKZ}`RT|BL%S3iB23km)Ox<_C+Opsf z6_7<;d$K840mb#2bVHdyoiqMHqv$|GiM9)_mL1a~FB*`j7`OX91u(e?<{!84<)#64YF~j?QJw{NShcDy8Pem;b*cGRQ?p=CXJ<+2~ z?#t^~H04HynA%nL)I_>-B7QAtQY+BWbfH?Okr>2a=m7Y12~`d0V=X{$GE90jk2#nw%ESL z_;w@_ZS>Kc$g}s))}D2xmHxe_tS@lxy;6dZWrnG%to`{gQk50=?%B_5H%A0so{jk{ zokX;Z#LSFUgshfv3$cicvZwU-EtyLgeEO#CDe#cHp6y4NT+X_Vt_4lHCr`oIm0k)v zdOX@)0Uh~Wsts9X>-8*^-#^k#Z2cr;JBO2PzU zf9mnx|A`@`$GyC+6>VV`e7cPIEv)yhe5rut>b+;tv1{`FMmj4!`6F}ybg+A+*GS|_ zSMxPHhVO`B+h(ZG%?bkFF?rna4j>YP*v35!9`>-=s}X13Xv0rET{+<|;tnx5f$o@n z>Tex@e%pGN>g0yes}#+39`P`x`3{txF!?}&;t^S5bW*jxOBG*4m^OO zjH>su@2zQiahA?N_!4`P=}8v0^PUvzU%HLoz;#9}I|FgK%rj3yp{x|%yz8&{Pg)H44V1{qh}7B3=+mgCau@vD%{6+RA>@>wKMcZoE| zT+*oV-zIc-Wm&0fLxhRk;zwD6Jo|l;-1t@wQYHGf zkM_|9DC{ZkjWalRSmJ?X5Q5!eXKRDwg6_wjv0wj@ZmFLIyD` z$JZ8AQk>dH->36zAYNO|1f(v9*|&Y;ixL^TvQ9(G5&EK|wt?R&gI$1rq&LoWvC-Q7 zlLF#&QEl~P*0-5BWNVM3-(qxR)oUIlE(tP@uD>5Fro6)Er&`GKc(+fk3fS_F@6QDI z@Rq#8UDw^jKs7$GOajK(@npB(f_r&%Hn%n*H*?I6cWvz4eTRDULVGUp^`__zGQaETh1 zgwKxS)E65_OBbh;$4FimJmusb{^Yotd80gJC$s{F8pXB>+Q)ZK1c#6#0HIuhc zwJ;-32a%i*^7VK4T6tf%a@hJCb#fuEh1qv|zZK1bwELa$RwI1e3NrwOEIxvj_nh&} zaxF1i6oL6p- zO4OHcicYU*4~d@&Tg%mBJbsLF79*XC^{;8YGdKO--8@9m*`OECKL3YRc43o4fc^)h}f@5nLVKN9O>rmC)&I;bErrzL4DKIobA5A(=AL^YOzUxS7v;2CO*HE<86Q(o!ycT~ zY?p=W@Y5=zp<;$r!W$#9;OY0^1#uXelDL_HNgj`1x+F2m5lPq|TUP~7?&7WwzEBe2?mIP4XjwZ2ux4cAUb@t}ny1n5_uZ~$ zm_=1Xht>UGdGx$k3V$~Kj=}}hHsmhH=EQa|LSFRQ4NIOI_Ca|WF*MpP7e{i|x$D=u z`L$40-C659hnQ!hBLjx{^yVl2{D|kt%^ovLEKsnu$xRMr#PGq(fPEbjBSAI@Ztp-& zj6Bv&AuN5H1p#Dit{J(bE8V7{Yo>($iLxI@g#Q~$>iVB;-Btj*-}(%c&2n8Xisc)| zv*2l`ukR)mgeLS3lhg!IOP6R0GUfOZ<^+D z`?Q8grl`MPnK1dR%Fm-B|6rW_q2iZP(3fkNwdtRWp;6yKa{*Z+!D%h!LDlU}UwIq* z%B;T}yWMIG89B4X_+Wz}72C8_0e6QDIk#wQUt15UMG{PMF>1Ig>$v!nv+iY6Y0>O^ zbKVX*`;;-gxktg2Ky8SH(goc|J~d;f=NXggC3b$Rd`c}o&7bhnr=**F)~6JQB*=Nz zN)L!RdX43f7w*ALTM|ca{^FJYstu8E7GO(yvhjdd{bUKJE0Ox8A7TUv!pe`zZF?eV4)v?SEDoWDU*yCU^{BT-x^^12QafF!ZpSd5DafA# z(FIwU=kYBNW{E55F>q*!6*tiuKVR!`z@eZOE&Az}%N9XOJDp zJB`7<&rEdkWGm6x?4b)-F@0t3r^jMTG~%TJu>_5|Po^Mb@M@P=CInlGU%bs(4w0n! zNvnWdcUWB@Wqs1q!ElKf30apdx0|?6AZ_#&=WO57a){sMcHxr3b!;1h0JS8r9$w#ii-A9_ zw9%Z8D|btcACM61Ec#_KTo z#Us;`@YT6rV2Opa#FE#AR140qPhd^$0asHj_fZF@)qWqBg2?9zm!P6F0D=~q+WNP* zqVyo;{5Ii2C&;mg+ha^5h|Xpg^0elbp`c#yDn4E>@}my3C{F(9@i*m0Bo_U8lVafv z%$_$x-=ut!zSi;jdr^l(?F3k27(wk3UVbadi8va#G*t}|Xd$-Gz60y_ad;5k?z9m% zTMsuI9mpG(T2^zEdm>9cB$h1*=N)$0xf@Y*(WY&EwnLKq!RcU%`Tej=z7s||5hVgF z!TJDG`-{^|uZ5{ih*V}wy?sWJw7y%8@_&=SPL<6PmtYhNifq3RQ?2+M%lVH6K!x4M zNPOrSawET&+b3v(9zT=7PgUPWwM$;Zm}`XDmVhuLy9B-3nPR(r3J|aIJWWqsc}ga} zCITA`oh%O|)nE*MzlmB7g%4cK(X%W9zokltm3sUy9bz=hKmvs%R{6=#AQP2D|K~Dc zuZOLk26vt1*?)wocw!!MEfW+!qqaHde|++Z^A_D_hXp8Tp&8)1gQ$O2QX&@?~G7GmIR5_aa zoblga*KRsx_2}&l@xiFwf@2s;2jR@z=w$8YKjH3h_*C>=v$D>)n(!Stc;C~_m^x|) zx-%YMBwRmWcQMDC5LDZ?ZZ3?Fa5Vj)k=+oKOlh}$RR{WUUb>A)1q zQLBcoG9%>U3X{P*mkP=Fx9({)17+qk`-$zn1K=>5bBAb4U9r7%l?dcKX&7h&dx1zhQ za+Kk=cEjn@2SP7n&fA%Zo1F^ErzNHmRUXs{Up|}6=6+fy>`VfC+G**a?<|iu*!&p} zV%=F}b=lCTuTCB~pSme$Aoq}cermzAARF9rR$y0&|L(Y~f6>sKv`36HFq)mjU*WzR6Z%NL7+wm{Vkd`=Jvfa3s`W&^9C^@+;`LTv z=r$S)v(vzZLVcS z)#~r4u)Db5P8`z4Qr09IXH2I<6OqMT6AKO?-WIW;Ng3(pKn2n-mb&8>gPziZvVy%? z#bt$&TlK)>QIj|;?;SS8@HL48rK8$YfB3{3=WWFikC)rqOSsGWd}1d^u_u1fg;2$Q zKi5eLX<}TXbDIQU&_8$_4Rf`+owb?stJe$&u8iYF%!MtaJO(qLz|64dt2XkRJNuBE zM5V{x&c%YjtE*G?9Zisv#xR1bN0Q2`9p^>I88bE6;R&YWtsrBCNKL z$jLQESiER6_O0VadMad*sHPZ7sLCf$?N6a+eq!%(YifXgk0O3!2G2uQy=8EhOj(uT z`G~^$kfOzuo4OtOg(CqkKHIdW4Hn~giH`&eH-84+K$FK(B0aGhsLAgdrYjRj|6`() z(@}bg<-EX7$yQ;xx5tt(lGKAQ7i|zPDTqch}|4&;-S8vxa*iGo_j(H)RB*N6vwJ z#s3|%ts=Uu^(Vmy12G)2 zl|LFmZ>=#?qqt7`d*XOI2olrAM#$aCG|kdIhKxMZ6L^GRZ|mUrIio*Emr7!@*^5&? zw8h>Z9^upruN?F$JJ9;sshW2;6$U*8UII>j)Sy>~qPpB-PDHzf+p{I8r7jG!3i<>u z$aDBU5M}Ie<#|T%z@Ou01wk_l7r*E z42?Q3k?-a${=y}%jP{w&qsK%ZSghmG)LO$6Q3eWLT`I9sk&t5e6Xdzd84V1R|P5MyT0ym)SY7Ker>SE5h zY2lMiBOOOnRDTPSjj|v8hR^&KD9sVnz7{Z@Hv{j_GvVF*-%g|O$|09LFgaGqM*+y~ zoOgTM zhJgeGUgGQiz49sZ#bKy6bSK4KoKHp6h8JHA;=6crmGWn?3aY{JQ?K;rdTX6bluU@* z9Ac2+dsh6-J5C}CEM!0end2HcYYd4nG+jEREKkT%I-=;0Js3+cnn4a)b<2J^v*1Kn z@UaCzJEJFV$=ClkDI0#k8FAY*Mi`EBd-#=mA#C9rb6pP9f!V&d#6peMvT;-=DvFMU zjSQU(EYuJLPrB>Z!!<>uEwnws^DKM4t&^W3*DaI(*~w8#!7vRo0P2+ltw(v+0yRgH z8ZR+=#E77Da4vh-zbe;iN46Zc}|?y(~#}i8m|k(VxXQu?SDJ85z|KRrR(**ms}X_e3-HVbPxiI(K1$)m)m)Rsi${h zj>#MnOK1I;ixcZb60K^Gdp#4rruXYl`ESU!WGn~Md3@P}NTyEc?|TDf2DM!ZI?>#O zZd>dnWI~|0`8+}Mw|wr=Rqi7PaQjmYW@49zuXu<7fXfeAlwxN;CJD_ZX*vBUZKOD;6B2li6&N3b2__tJ(6Iou8+8m5}|XOHLQnr9zeN zzRoz$3y_BZ*zU#H8qAx+!r2~&$uwBWL^rlOQ=X6nthjlA1p3JRbY5wL7M=j$AVB1F zNc3lY=gKdwt!|1%wF#qXaWyxh+wYzT91M@kOHkg-{*kSDNRbb-7IwSaj*hCO@$uC@ z+VFe(SpL!@?#0gqkpA(=sUjOlMDH`)&j4~EwC)P=M$OCmk1!F3&*?&`!I!_&{DWDCo+VwS#sUQ_8MBw($7jNGM)tJYW%XwHWrC z06_H#ZDP}dAV&8wby;)^j~RPN3H|2tMbA3dJ;J0E<|T&&+^f=*+wy4swj(X2iyohN zh&ZjSchl|Rfpe6P?>7)LXMO@%T@}gaOS#7@9&?o0R97It2JT^DEW z)$$%(948OtfE*vFaXatQ9*<$;p)N}`3ZUC_^eU0-pvgW-;B+TC96O(v_UI5>5qDMZ ziML=f_#ZbZK1$?6nml5Z^?*zeLvtoYsQa~2!k0ApN+#{e!z`^TofSMDc6m=As)|e0 z%Q}_a3WcspJAW39gz?9mLa2!8ev#gaCPuG@#I6}SOfFVF(=F8{&jeAHB6KZF?=W61 z0`f&BI2%VPlCQUvc)J;IJ%|en(=1LqzhaF^)6h#ON}*hsILn5emjvNJ)@m5DFml?Y1_${f=N>WtDq!M_M7ZSZ}84 zFsddLtdlAB2h{8GY02ur>;w0IJV2lHWCK@+>g4VoV#XusX;*Ikc{$~-o)}qRcRlEn zL>#?4s`dWmdmFjsk9QhMuYFt}=ug(nnbQiKH#WY1y?|yhr~iVl;J0*E!@&U>0@#G# zL>C_dvRYQGt&e&>dVIr~2Hgwr&(cUX`0H6nz8 zGaldQJ`0taJjKkUGza_NsZZBECHsM^0j0o)G&_;cz4Bxq4*|a8v;KQYY5r|1vGko# z)cd*Z!0NAH!2RKFAl5UMMT>fuSX2;WiYI=eI2~I{PytHY_tKuuUv6Gln^ODngd6N} z)n(VpR=Y;Ya)s0wLEFnc!F})hRXRzC5sjp1Af5B82yy5dhDw}98Di%gZ&-Rx+pr)Q zCOeSR&7L>f>Xy0TX1L`!TJqzRHplZh=`=jGW!8Nta4QHInrRR)>h3L3#idt zu_e7UjSk3X_A~F6cY6C3X8qEX=B%>!hcAM$2CWk&@spWOR5!=K|Ej@$fojXp@3}8-S5g6D!SR;_wOV5psqNl zgFd%KZOCY!^5*X?8nOkrKzl~(i)&yoZBl(ud(>W#jcA{TSXD-M`TQx-J|j*#PJ{8q za-ooF<9x+i#jDc+2~n*l$jnpT0iV+MV869t;#*B`QLy-Ed9zTtahsJ|%xeO5Z8CO9Xd)3@iOk4wv zkSU&G6;P}J`pD`MttvLG=wGJ=FV0Dw!m%Hu=j#J$DafJqQ@%*67RfaGSZL$VTf`{U z{>KNbjdf~!zf37{Y8WBG0`Y@WQ|8c(cYNy3I%}M?h7cJ}PeBZBpe?!A9WxmZYMeYT zL9XF`clM}Ww)uHf_NU0ha8uX+Kun5nZEE&ckW2ql8>hh0N;F@=e-#;p+9XqpFzTU<6RQxh*6=H}^k*L^kR zx|w`F_4!A^@UDLy_75|~a{kun%?d1Rj&Qz#a{>@Gkktkl`CE4imI8@hNg}E#FVqw$0srlj6s6l!%z6J!+_Prs-hJ`-;OS)ZUpdZ|ov|=$3{CwX>SgLg?%K z;H#AbyniYR4@aKy6~0<~cDC^Yr!V{W%@wuvQp|nfA!eUm^=edb5Je5YoAM`Q>Oq)( zh1tjBZ`QlD6P$>?rNc4WRbi(i6^A3`UrR#GX<-3r2qj2%{3_@($jb8*L zG_Fw7*wi}2N4=iu-L8r3JX2QC)b=F){3%z;myb7x^cYt~Sf@f%cWeCg;a&w5zy8iEx(g( zu{V!`go2(9ty?|aTeSFDH}ir*v!C3-WvX%nN(!KYeBf_E;gT{vWQja%j4 z1i}!Bv~~J~`D{BkX`#&gunaB87e7F4xZMEM$MQ9AU)jB&i#umW>+x(#5M&Ngi$O;| zt0*K=@3|Z}Z-mufcT#l7$DUj)e9|$-!FH~;O63=f_UO3_a`#W;=T%3Gd4QO%G;74- z^U5|9%?D8|?6luVJ@!t_?QEKtFvQ#N~-ITDbs zpqFM3Xe3H#2A6Vhv-rv)w z?_I5p$h%`L^)wD`b=srIA1S)CtC05+l9Mb{>2GxFtA-$MY95omau_z2$Xh-V zB(g9j1f?9#Y@sxI#k5}La5zV-d?KXIE&}|!AHT&BTzwG&%RSlSdPKM~ek&PFlWAbu zZo08PXm-{oEze(VbN7yei3lg#l~WA2xzRi{vexfm$75on?XmXw~>V9s}g0z30OhJjVwtz|OQ* z&*(rTE6@J_{$UbD^~g5XxjRUbs=q&r+sd-L z4NmS$N}bftS)d3qYWQ4AOhcYHS(hup)C#L@;w%+FA{&qXk7xw=CpFl>9sR-|ilZN{ zwEQgO>jfodHO$Gt1a5Mb8FxulE?NjE8HCsbU9PJ3kqS@OeV1=$A|oLnY8KmVmH9aH zQz5&m_j9Xye~X+ooa|* z-q73by3r}mC9QU>@;brnSP^1wE}OY(mR?tEQypO)nj~GAcbE3CTA?wE*okv9_+*^H zEG+$o<9mS4DD&b*)EUL*uggfjEXEYwnJ~7~;dWZhV8saZs>sx^Mrqdhl%>4PT9-Y5 z2#sDpsFuvtljai_lK@jdtiO}?(bnzlWi!xgU1>Fbpoi0#BZe}siq(!lnrPJA<6ud2 zX9?U+;Cn$N`5Ru2Al5eZIBBB3QsY6`qCELCP<(rVRLC|XGv=AclneNulb3u*@hZOm6itRQ?@UbqAytMmL>@sa&8bp~IAXNY}r7szTu5&KHWUD#=|Yv~5cd)`v{ZDodP zUrbJ7EtLzUGE$G?6W09;OQv)9X9XwiPem5zuG6}N`I2<83foQoW4|@+)}z6&qGonB zsyN4xPX|X-Rn|Jg!XMgOJH3C6!N-&Utg2EA3VBDI%rQ5di1~O%h2zZ-%s9 z<4uW4F@$n!$sIE;>#jVOPUf@s>B4MGzT6sNO(lQuUylbmz&~f+Q>^S@)A8P=Tg9!z zv7AmdKTliiNy_U{_p)pjm-*m^P<-F1XPP5}Qk8E#NGak(gZmyCDOL#hY~}kspD(@M z$()|$lLIEb_y-@h@;vE|nmdDrTIta4a%=!{@AINI!&6)ZV}`WO~8Sk z-Xs6R-t!+2P)RiOI=%s#)074^Q_pWNDyS@jgXg_zS2!kH)Sill&V=^0{k3)VAPsAr zrdhltlA?lyV8ipFa4Wf!oU=>ooxGcem(s;kSbiZ_I5_R!Ul`C8>! z!MPz`)Tbg&Q;009!|GSBxbyr0%|I*z}GK(^La>G>hc9t!txvcT8cgL&0jbeo99BV3G>B=Dr)y@ zAz5$oesxiBi*`P?0pWM0Tjh9H(l(R)51gPYH&{>}@L}t@k&4=Tz4rPgi@Q;JN=?%~ zFB(IHbo*k^hNz9d;QMx=usP7_Z?J?0h&8jNl>QjiUx)=JtwT7weUqkJxk}Pj?ba7B zTbGX##KuBLerp{4{#=y-ekF_zhKcqM{r^ilhKP=L=C7XEnu=9L$5t@kA4iw2%}@64 z=ldS_dXEU&UWH`ztxO;(3&mJX)oPcsFN(vIZ5|s9+stR)K9|i+u%~UEb&bza?QDSW zIvw>TU2;!_T5v>6fPcg6?EHMw;;TkUJvRfLn3SD&&X&b!>K)y2n}0x|z>CW0^HKjU z1lJjSs`ynrzaXnjb@=-SFHU;VW(b>bPaD0C{Acb1G@a&GUuqwblL3;h;P2Y-N{u%Z zO2sGLzycSNBNpRIsomUM!5;F#Dfc!LGdw>@HwCE3Rm1f31T}tMHUWS$9^)(i@PRx& z3rmsLFQIEN4#L2c;>tTT(Ft0|uLg0#OYub>+8y*3Ikxay>6RA!Of^V6X#snyF~@&L zloK~tdNBVXda#-Yb>|E^t4a`_S!(FvbUpAlZI`kt-I0z#t!msGyQ4 z56V<8E*gI^>VEj%nt1g8u*?&UG@^vpklEL4gC;~TtkvN&`Lmv9`6QVL+-%DL6VlJ+ z)>=HY6p#s%qt>)%v{&9(3go8IE%&$Bnj{an;SRDha1&=oH?4x za-{x6CJN?`4kBr?x65JQf0xuWImvS~sHqa+f0Ej?|L_SX=Dr*hE!@oGrxDc5F8;+` zegFE$gCf0DuF{Wpbb?9I-U#aWtT)|a{$ZA_WT~$=0*8K66H^lF#4Bx9D=I7b`cjtw zj7zRjY^rh2yEw-0{)sTM>0M zws4#nF&}5*o`qeRv4tHFiH$c-O)z#?#72_*8Vbzo0w;s6K+UdG$APDI;k<$_q^xUxR4Zz}`^Tn1wRtzBA!tUtA?~y)J zmVx($UAS4?*)a_PFY3)&QuGDGp`N3EPp4{1m1lsw7OO_0-yHkTw>{MrZMgr}$x(N{ zbagZu9+tqfC}o7Q^4BG?MIxbaw-@pbuK19`MA3lZ`0wOj=)YZKhHfzYO*@g|_urcU z_iBF)t#g5{IS#Z>G&pwK2I4-HQePuzswTt@{Uq3j8&LJX-i{OS&3~-i{H;9tp>I3S za7=aO&cbv3zr_kWkRPd@LoIsSCaD85#Sm;zc!#jkYEeN&bA6bxNqrorOHWch#H!~W zV;oG0=5FpXY*BUYoOL`on@bfmO4gcsfcOG-OmOG(Mi6_26&Xc0tXs__TKjrQes0!L zBOFoMU%@WbkIQ+F+;AB3U_B%OvF9g*Otv~Wos9cj#BST-m!I)(&Eb}>RM)P))9+E9EE4sI_Lwp#Z2@!=w$#5X z?S~@~`4-=EV$KVaPxIk%$u8B+Az@4BypIP%rjqUjYu+Ol!|{Lq!0=bd8C&)9jjZtn z2c@W0U?j@xdsusDk1aU;ts}*;>G3blB(AA3&}Ps;9`z(|@`1}z#e96&Dc|Sxml0Tr z`W)L4ySDM64@!<1xMxQ!Ddyd%8x~F{ z#c$3r@$@f+R-8u$4~6r(oLqU12POLuciZ*b@}Q5u-H9Iqv8EY2U6`xS{g_NS2X z6XJ+DLee>=U73b*pZGNYg=c@)mX%Yyk9)-m{+ZGwja0{;?-$<3tP07GbA0 z_(A1&tt#*N?;<{y?iOf~_NCQ$je^=`pnFoFqt0ZQQ7B9hmOrLxS=@Q z9NAMhLnkR}n~OnwyxwjaSuYg1E4X(7gfp=YPJX<;dI@a^)57GlC@o-wW4mpu>qZ`V zIH62Tv`(MI79U$cYkv-#^g?DY&beBKw3>xXlB0rUa{h|Ww&Ltmg>y>DdG0Du>;;cv zZ@3xYhe%eI0dInwSby3-_NfVZ*Pc=uYtV3)4gM?sK2{CC!d<$ywsyq@91Kh&&%=4K zCkROWNbVd7{3gMJW{j5b{Om%JHbpAV%q+03E@8;?k|BgLnFe8>;g>ApB;q^dz#rYQ zr+ULCHK>epL$4r_g!7aC;RhSr9#4lw<#rs|ew&ax^&hQpL}f|C(2qJg+1ib7y00i1 z;F98QX2}btF4(`kE5jk+W@eXIX>2qqO;vHrS-+{NBw`k3aZTp_s=2YuMcLHNY+~(? z1!q}8fyLe5IU;7KoSM^~a@|j~RDY^tI{hpdVbXvf^2z2WRVhuC1%HT^NX{qSxQ}ak znNbMn6Z(o%gn5o>4{o6cgau7W0v)j+Ps;(ye9|vC^76)WFe7vM~;P`u3!}6n&RI`oqwPAxOGzb)vy)Z&ggbI38 zf0Ku%fmI5)mLZ; zBg;ZBNp$@KtcKfpv6{G$YoCQl2&DuJi?$XL$ojyj$>Hc0pOGvm*-@sb!e2XK>liO^ z`08X{111z(Ym}qAzED&V2;$*>z(gM$a1< z!{V55A2IH73;lXqjf!_{pH#I!>ccXS)4lk(UD;jx2)7JOqb$a*%gGz&DIe4wTY#6> zsw&w8fj(|8A>xD=CzmS?(U@>+iKDF!d;XfLm0?p#>=j4H%9m-3W8q7O_LIL;pVDsp zcV}R$tK>N{-^v7vpv*d8HOu5p0tZj{IbXNlB;O)E)FukL#oN0f(b zTU^QqiGuCh5g$@QQC+Q!3@aN*RfpMq?igI<_Rzi(qciO(+>{eZXi(QRvHJ>^?7BL1fJ_jvDBPJecfi1|P{?Rh;oafMH=a|oZE62e z51^5B$|$L4Nf_{EMG8LxLRuK`egY@td3s2P#HB@6@`utS=PG{-E4^NeoUdw0|K9D< zdLJI&Hi!({wmINBQ61`&r>bmb+<^UQAoX(yF&6Qw3IU~g-sLb0-K>Q>LkHiU1->|< z8Gyb2CZ~V{Ep=6&+ArqBZYzaDBd6%Xc>0+++f~_XU-EYd+rJ%#S5tzW+v@CCck%b*U?Zcz z!Er`~@SDl*Nq=b*=+~$uarXp7D&KFuqqA}L#3<>Dhdo|_7nQ;}669uQ)L2NviP=vEFek;S2dt^~-MWCr5<6ttm zrt1fsIgjV{h{sdXz0)23O0nuCA2M$t?)H{w!Fpri<^SY@i}J+Gm>W6zwWVOffzEWM zOLT$i=1AWs>Jr~Ln(~N~@yiLpZ3XJEalBT;dXzH5h!<|9MDuCU--*Q|TB~>*!}B7^ z6c2mCj|=;5H{^F$E~ifOd+Hz`IB8c3LZ?K)esY~j=YH}phKG1Xy`4|1&^lMJvH-Zmn!|wILBmzjZImbM#Cfh zh&gPz0cPEnyfOl?1DuW5=_eUoA7_BI0KAM=<}I_YY~#kA^ND+ML;;Kdmmq&)0p!(7 zGecPeb!vKxS;F6)l3ct11S9)IRyGsH?zAssE?{l;%D7xA@q4PV5AuFNsa`3o#B@v) zBLA?=*{szU2hfsnp-I|sy=6;n34X?7TPySYmG8;e`PNzEF@QflxGW&7@t8c~31e8W=jIxYs`~H;`B&LO|KiPJby0o&lS7;Vf z4^Uir1)7q}B)u*m6F#;-zoCb#kpZ~~YGp}g-KF7r{S`Ux(%FB~F^SQ8i!!l^(pod< zh9)^?*#0^2Zo$ajxz;Hi8A9P9;8cs{Wh+H7t4q|1R^N!>d!(^f%fV8m?7Ug(+;4)b z=|yR8FQ?nh{6Vi6hqMN{#8q4_XqfPH4c$)9RU)0XsE%}N|JWDvGEwBDQvYf7xr^=Q3x85;+BmLOZO>=atMv1nH$}th$ zK!2ZF@6L&0l+e3#x(&5|l-ew85Th0YF=Z&ZbVUG%Yma{qe{fft1jT^Us}kV&uoMJ; z^~bL)p=VKX{C0Ot-X4Ib+jh>?S69vJmE>ZyWTgGNVV)%;>u^5ihyOZlK%)WYac4_H z&y~w%!`1CQ=`INttoDXD2p;|oqZT#eJ!^fscvBf?EJ-)Llg=^11Aj}MNv5DvVws*7 zz7Y%+7aZ4@9%!ZZT%0WKE$w#X*7HmUQVjeQ5Ha6LJrjPYx=Q!h3&&yYab+|Cg_NlI zD=@>c_Z0t@l-v|0z&*dS_MfFd=g8$eii)RqL?P{xZNMY9f3k{uvV1OHru57k`GeN8U*&ZMgswTN)V+N~O^NE8snysl zGLxU<@0hkpnfEr($=eY`&Z#}GtJ1nWH@4rcjv$hYlC6+c!O>FpSTcf@ai=?xD5v$*C%|$h}X?4qwxbf(^CG(+k~OB>j~OF>G!VLZSA1O z@ieEovX??p0w=o)`#n5Ktw9P^FNP?%a>!DT)m+h^v~IiO?{i6wh_n;}znA-W)Cb;& zyEZ?n5_v8mJ2-~i;P3kElB3=GS7YfrHI!uqIPuosEyhmjiLe1Q?3M63d*t52@@AxD zdkJWIWej^SJbH5NWPCbcykT-eW2ReO>w4ari_c{&T=hm-eeTkL#&6Y$8)Hd)*i(Uj zIRxjnEzAD-E_Ka+3Vc_wvVc{cLtUFqKrh;nZ80Ukw2n{nZI4-T$$M)Fy2K}T>Gs5b z78azPk2}Z6KPEZ8nQv|REOM6BBc1-JYyuku>XVixTthj+`t&ppq~2F;e2{oelqn6H`Z(RN3C z9NQr~h&`pDnMr4Q|?I{>FA*dCqE~HrOUHhUDz8xQwWsxP0 zu=;Uz?-jVX-f4SBv~ifGSkoJVtswz&Cxcx>n%?GyzORlg#wQ9!Q3 zvS%qHi1%Ey>@ZXBKMj8>_AZ#|1&hywIwhK5{(NsQ+M7vA3vn{lfl_wP@lwA&eNMl_ z4x-VMmQ}`w-k*dGpsMuPWE^Gu$ z>{?((hW8K3RHO4#SUgn$bc7nDf~EoNMX=}{Tbpc9q=ez*_o>iF0!iq}x8r2luxIP z-nHiVX4PrKOQ_i(Ih}m6QCQuhaiaSi2+IDe;<8yDF@KR%n5dlH4w z=Ob}nQrOiO*iQ`d9zrtRo<_5ct5`urK7*kj)*i_=azu?=54n;|fx6w&Apu^tRWry1 z0o_R!|LXo{!|M0j$Mot)!6d_w<4dVjTNJPGw*9SIO)u5Kyghf|H1ONOM}xVNEDux3G$c)CfobX#ydrbWo5MqzKYQ zR6vj}LFv6ELAn@?UGNC`zGAW}k<(31$5!~p-^=ljpi`JTC-xtV9y zTJOATy|unQHj7`d2CNl8MK+oSTHLZc+R*Ns(ciujY~!F_jfY{Yl2-%pyQ`4aZ=uiSi|7VRKqQGb zxc0VF*Z#J-qw^)shI{z#dMC|c%Q)uP?CEHFPGay}tZZ0=Q#8WW8 zA0PNM1UDtfI`rhiXVSXvEZJ*y4o`j7*>f{!Z_b4ppLp-@%`;ikmwWj0&7h51PwGXf zWiwBxSc<=r{H)Ym?nA~i{MJxfhVI7in}<^xA@j#L!9R@tkpA2Iu5sI+*X{o^$jhs2 z4j;y;nT$?!{k%i$U!g|l9#7oF2$_*lkSIe=UUrFa4?(2@0ng-UrElH6TBQM=!<^>> zGJY_J+N7UVHw{MrHeO98n8+6dbscU8$xFeBP=mj7{gdzbd@4MZs%#4c%htBU)UiFng7s;Zn{%3UP7euM|D3#HZ^&Jv! ztW1TC3tf#rlLoJO3AK;tTzfN*#(5~K z)X=JKts+^f@c)_;)@$aGo6PIj`N_;(w4qcQM)(fQR6c`b5~{0xQU9j_%T(;R`L^pn z5^p-so5{}(s>EEg%DlNJpXQRlJLzSg_hK#wwo{Hvt7jMg;Q;X2vbhhR_h~@%sCZ;@ zES>u^{FLL!ck=t1|4(tBaiNY^`~O`iyHs~uC#XhF80?7BL%+qAQvaLCdwTlx zLjM~&zJb=yXObK-Rq9oL%IBQE&!dG9;FALS z+^>OY599u3hzn?AlsuW{LO5OL7HGoMznbGzn)iI!t3;0PZ}u3ipmY^o5Xb0O4>nH4 z1*HilI0=_bdWCsmOML7(U`x+9J@+=h$Rg4s=+twFHswyXL_<21B zx~6?GsefK(o7n}Vte8djN$?URhLVd_N)@!zq!xoIsj|l}LElv`nsNV2A+0CR=%;bM zxoLxR0CQsJK-8$0pYW#H$K78~Y7sWrggW+wgMPfpAlonc?1+g{R43ho&?tO8DZYsM zpvJ?ZcxQ8IYu#BGC`}kL=|2f|?7@T4xF+*UQjc^G*5tS3fOqNQlC~RuGx75p`oqM9 z(tz3Gm2|DvSD$z9qY-K_)px8p2Fq!s?(KRrR%T1MvNQ4)u-l$;vGJhRIMw ziX!u$=FR~2VTmZ8DD?5<4*W}UF0qR|U}363N0Roy+3fl??{HxzZVG%a0S!vASH z2l3S%tB6`3T)Nx#zMPN;11j$GhDw>b9GuF;Dx66_ zS<7F5^YlRP5KWJBO`MVF;*XZfL0YRGbrzVUD)viAjoP~56{i&cjr82PUT}G|HKc{U zqyoGBP^723SI4QbO6m)zUh}QKRG>bNpgp*>Y;@w z)sJ`A<*wZOd~0-V`Gb^5fno`&lbQox|7r-pu4`aCa>*!5cvGFIktAj*{`k--+m)0F z?4Tsb9pbD2NJCan81{=zbWETfk0}Yn{k=}XCBT1jtvnm@1`D!r$jI@Hy__KqY!6s& zH%)UJ(6D`@p|JgWoM$W)Ea`pPxu&L7`L4li~?_kI#j ze#(>x4GxpRE`Oi%Obp6+9^=`CRhECGp*Hj$IVN}z2~_!d&2>&haLpga_|{Rkw`Oq6 z*ke0-CmD}M|CUz~9JU)x7^|e$Xy;y6P@x1|D?k%JL5 zwLDLE?ZwY9u1N{E4-?lEXQ?k0nW^;1C&srkQZON>_&npq5>>r8j^iBN60F;1IEEtv z4U?)S!>h6<&0J6GG7`tGM8`1(G`s)>7joUJcA4)nM}NnH5Mfy{#qBDA)vFp02fR7v zJ}gwC}6GD)=T;XQ!uFT{^Vm`OlGuHGXy271V))V7k3Z0Sf)5l=+6qkVYNa0b^O)$ zzpq6M2MD4JzJyH3$;M5Ouq@(M^N){ zYxc7O?FH0dxu79%HVJR^56Q6JC)A_Y%g%`?H<#mb%Yx5~-A(5cmTGIG2hX6~VOBcy zG@N?-t?SlJ>n6%*k$P00{+>HKes4%#udwgSwL@z@#g;eyQzOX~Ro2!MD=v5Iz^fB) zIher-olH{03&ql)DQ_cplqWU4rl6bQKY4-|{uI#e(7Rd8o~n7O$U4vZBD?@#ot9jr zH}?txL30PBj8^X#tP)O3mp5m5{1cQTv>5CorQZ9a`SwKA1bg;hMen=+;`wj;Xw)1L zELM=5$C^l~N#|sKVgg%)W`X-wm0Fq_}1G2NJ@g2tb(L$jTH z5yQ8|Xc5~^H;b5h?Uy#H!WX7p?c)FXL!wz};<(kq^G zGRcJZ_;TY&wVq>;bi++vl{kX`Zy&wl=aE3~e$#!Woo7|Gp2l_0$Ih$evw#5^L*t%e zF_}O8*#vdE*><*gv}i<+-2YqMtApgnJXu~B4DQInsu>gM|YXN|7?|Wsnx+#<5~dn z2>`7A@{xZEcz(}k7F((`PLz}MtCadJ)ws;#+^Fcye6L)DBRZ??hcylKA+M&rdb{ps z|8rkn`S0zI@rt}CQtAgFT(B!YFY66(>2znuqdCshg5pI!iZvBa_l50~-*xXdg>KA! zjBEaE5P{KKX@+MVSYST++aCDax?%Bd1Hm!ueI=&!7@p;W!Kk9Kg7P&aF4{$Mq1g&FxKylY;-zysX3(L2Vx*bm&vj}3 ziR}FtKH1oSc=eZo2=%*Nb#_li|J7^9*Z-mm$t!(nfh|7|7N}WP_xbZ>H z{E#IuNSjo=*JGwTyHTH$Y1YckWS)ufKvj{92csE-AsO~$WPz)ZAYj170E2TNon)S1 z!S3gFl08?=NB8ff==DWEh7F`)YfBuOE^6s$PA^xel&4GPe=g^kqbqzGUKK0f^Y2Io zxM_~p+edE>Z4hn`q9voJW}%G?o#I~~i$%@of)LH9l_AI=bbGy|JFf_)S?nk#$VUGt z-f}HB_$fES&qJ!^h1NrO?NxV2Y z&~lqn>H6pEO=RK!7B~d=#s4}165M*9l&(&en}tzr621G(QJx-V>VayrCWbO{k6+(% z`jm6pQpfrR=#GGP`ajL`a#wcX+kS2Z$h#=P^nA{)bTNp4Q2IMg9fW`~u0o@&SlgLw zh&M{;-%2FdiIK`xMD8$kdEURYzDWByE@k~)j*FOyg%0!TdXiwi!Yj~k z*oiQ8J>-l>aH%*_y|@*Cw;De$gWf+^TiMn=@~j|mIJESjj>6w3jVZ%l)%g8oZr0f9 zE52wklWa0d28o<8{gFto`8n7t<7ZmHcf+FxtZv6%VRs|f&?#FFvCw)z z484R`)v@jxVM@y@?4-%F5dK zuo+X={aJ@}(ZC?L88@jo+Vh$xJUF`ai&Iy6D!VtgQ}{q1Q4K;iUDA3)aOJOo|GsIM zBvI%w9waXGhta^Ubyb3?C%*6OC@eMi6#qD>`ET%nlZRHzu+uQEX1Kbm=Z8N0#F+oB z_=8>~-dVLnROOq9*^g!?JQQ}HXl=u!U90~srj`-k@FS}0?+;Pi0MvIy5;wxsP3$Nt zBI25`fJY{zlYrj+4Ly<;PAFO!)qvQ55vGXf35?>_@&$1Czu3I>3>)YDhc(Uhrs7M$ zAKRISyN|^FeB1nNGfTY~r{il3IN+DxY-WNw_uL#s+GD{-&xAZxIx;*aNt?b-1R>*o zX|;Qwrg!PD=oRsVTUK`%X3MXyu}~ox<&@I0H=f3B9?dr1 zzS|q!=#cDPLWw{{hqp)ATxyB1!SK5morHt8SSIM}1Tmf|phwq*`=1l}JK8$(C&NwE z{ES*kAz#5iwAY%aS3sueajqtbX*x7k*U^3rItxpz__RMj6e}c~yk`!e!|9D{{w69^ zp&e|_$Drrgzsf>SZ1N5YuVTcvQtCKI0&6K%x8Rd{Njqa2!?%1I2cq`+Pp-Cz0}$Z{ zs2o=Bd}=Gtn_AcXFs1C! zk4?#9))II-gcPpjWhJ;o7BmeHwT@iPIO^{D`Gm%a|9OS0vvE!9sG`!P_q}FDO-+sb zd(Zm9^WSLf0ef>fy>?7~asJCHmtWeccLSMi8~1v761i+?8Gd zyWqwu_seh~j{?Y*yE#325058O4OZCV;HLwwx@duG*^2%~2&_W%B)l&Qw^ieU^!T$$ zALw)~Fljk1^c86Z;D44)54S4$V4Uxd1ZyCyDTL2VN0Z|>Q7m4D?IF%3Fx#NuMJN4+ zE4w~>m;d!3y(43TcfVJ^%iGTH{x1uS7aI)vq-It?A1SCwVU2gbH8=89yHh56IY%-7 zQ^C0yA0kL?Khr3^)Wpj35^iRY)N2rvn`hy;i1@pm4~hXNs3)N41$mC$)8l8}_M6ws zI$f1Uf3R=|X~{vshbT$T8t-ar4Vc9T@+}p##P+w-XY>bJH)B?*o%^wWUCJJKR!X)x z>(En!@D``aw-pWF))gnFJlM2z(-0<`=9bloNj>WAxK45YF zr8mp@CVw;diUr^;7A|N*)E86AXU9OFP+Q9^d|$%1WQ=c%*iaAtyduxIM~PB; z{EB6{VBNwHX|*1Q72v!x7dMvfThT!{6r4LuH7BnxT@h>mJPY%~ogNZP_O*DAyA_qj zoY&O;iCjEo@?9Uf4|W8fko!iS=H>GzBHCU*)1A&P-ZD23bcv+h2j}l=2x>%7WV$fj z2|0M?^z*d<$AzB=?7Y+W{ds5`VPP_8#RKJc>cg&cVM-~&zwR}F?qmej>E#;N)Lr+@ zxqxK`9LliBc}kQ{%g0msyp3{NTLhDj_?GHyxum$~n$OoNg2_%-ai8cJSlI%zk%kSk zX2^0Ff&_DZ?77w&n?)xnP#Y@aI$SBW zeZi)L4ynWUmR>L0ZQ^7%uBQ(QNWx6w?a~xIks@9m>;XZoeO$Jyr71Si^iMVEY)pX5 zgd8S`ccLk84I5pd|F>jQ6tY7Vb$l`8nSKyHHQoyK?!lM2t1kyET#~Fal#}`bbvF(S znEM{)CFHc6T=O@&K3~|yMeJB&$H{kK+V&0gUK7d8ciwT?AX6~dOCxsgw=YqVr#Bn3 zm!Cb04^XX3$KZ+0#^d#18f!{oc;bZOrbdC8wDzl1zn5pQ=dpwS7tB_4{wuhT}LW78cM zOO-_XdBo&3Pml4=icVUIhlzE8sF9v4!i$@E>wi$ z^a?5L5@n+}ICMBncuHvt6fJML)hu(w@8f>Sk1;SIR?_hyQ+ZDT-%;NSbBq0u=Y zZ|tfYl}9uB36i6OVTX_pnpN>PnF2`A8aR#((OL7QJ(B((x_N!Z$*q7Wo&AM~e>pFu zYrKJY1pOae>HR;rGP;a!XyauTt+%;aLSSJ%rZ1ZZEPx;(RK_(s)Wu6j!br~K;~ySG z(B*o=vg#BS&PvSP7q`qKSzQ;<;kaB}E4z(>8L{Fe=nASeZB^dfU)`J~=MPo|yA?MN z|ME|8C^Kmikl?I4ZLSWU3_Wm3lq(InJJl1@dC-Bco^%xi;mhuIj?c`#k3-=sIAvlt zx<~qrPRTo^pfUB~uRoY)DapK{JJRn82QhDPNfcX}SR1V?e!S+PG1tQpsQD<{E99q7 zhux*R+?VCHrs_RYM%*k=KDjP8h+i{b)%>-$C4_2_@mBaqg4D}B=G#6#1Zral{E&Fw z|ME!lvh}CGnrAzQAE7R;SOfO17XtDjHKk~^TCQX6_Fr%qtEP9H{vk{?7gz5o z&^cZ#*RCmaL}>6`sUkzUS{VYcE2R0RO()v*f(2SsQ08+%P4E38aC(wrf22=J-CLIatq8LxV3BJBE z#85oa$GBrf98rz)2{DsO$<8$%HbF2OwZVb+$zOcDNxG`nTsLzU8_h8Blj);_r}rte zjChf9y(Ee6&%&6xQeT&Fif1Z^0Ox773ay#e^HeQI@xklHh}E_MAglkv_CSy(aHO-a zA5zS);%Lbk7`%%;$#GTbT%WbFkvt8>3+GPoe4}>h;biJN_A7rHn?*Q-al`n8x?BhMe=zGBu(@BF zZ&>%?>gc64nxkfK-}_38$P8Z)vd*_m4)avHmHtoNsL|E-K9n_8IybI*?xe)O8%n0a zkl5!s)gC7V!_QycbHI$iDW!er=drJ$)-ZngWwPiSW+T01m!JpM>ezP4{QQ+asUipD6@{`d7z0Ss+*R5|pE>%}=@+WdoaQ zxfQt-EGRZ=M+RYJFfoFN`H7^$UyGXO9{x>rk1jw?4CtLcg!s3@LX4Vk@N8T8nLAQpuGAjHLF6c`#7=wsIK9Q!WxKr>A0h#KV2B+u=UrjSiOU{ zxgNF&OXc;Q1JWy3*&pi=F8y&S_FMEf zZelcO(kXV(lppmR#eP{ryWZyG^G~{5|9*O5{_zK^_V|CwXVF>w{~`~@YZbwNSK+ex zH~XXp==a!}s4m0Bi#rYd-)<;B{-=oaU~0c$iREK6DyIV!w+NpJkme8IEL>Yue);n6 z(iMT2t5)KmtVO-t6Z0^&NmM_bk==Kz+WM~5o?MGSbWpMR;)72)EN2tV9QlG~9LS4& zz5HF4w+*eF@4;TkEegn9z|{P;PI$PH{ zD@+>A>eId{x0h7%@_;@9F%PHvQa_#MTO>Vg;JoyM+Lsh4px-E$3i(dB)2u~Z&uEH7 z&{6^j#{Y^{Qevz3f()lWXaM{g<$`N``vcT%l*_J)R*V33&FhjaD#GQw=&!<2KlJHx zY()>JZR=HVfnHyU7v|_EML{gz>y3s;@!HUmn-#%-#FbSI3m)oeQ2tb^Vf>>a?(%t+Hsi9x#1O9{!4cMZV zb^tv2oGyo-h>cAyK>+j9xFRTo>&TA{GN1OU!r%C8zN0H#{ z_`D9PaFAfbV?qyrSDj{nz9)7Myh5owuB}*AC7#rq>rQ0DK>q~b4tA!YGnb2IK2A87 zBG>lp8lB-LViB}CUlR>ig)VdQYlF{jo8`M$*;t&J@*k>T{BUVj+?>;2Pju`KQV6`o zN9ZkFiXQg?4^gg#XlbX`B?iig&VgD{LV)@@-8g5(m2%(o%Fp+f3+yK2?xW8t&Zij4{D}EZa~VyPZf{KYp$fpgg!3oQd%1sa z9+}wZ4dUj`bOt8$>+ShJc(7%0MkJY64-OVQrYLEI_ADi$HR=Bt8;E>xP<*Shh5Jxk z*WVMFHfbx==3BwL;0^|t;1)fV;NsrDxS3aVK^Wl#N=k#K|Fbinrr4!42QXL?6~(Xn zF&j%T0zg(WqazbV4<)uL<3E_GV9fxlMbYRnTw*cyl$(me1hOXXjBsRWTJC#70(_s1 z8>M+_U8(rDQpfP5$W{aO)dEW#>kSsm&%C^Mf0@f|DYO^xkTq~e`?co$s&Ury5J4%( zuffEM7xrxIU&5qBYyZaQ1DVcTSS+6H z`pGk(9q@UTn^7DI@!Q;5<|a2qn*%T_tMOv7=A&97xFLgneVyJyTPMBprLo~h)DXhU zpJ|)6r+x(me_bSeDstSo@Jo}$NEAo!KG3^~40`Zk)eL49uc=&?qzvzc=;mY|Q7$%U zchsvw-=vze}3^VhfMZA9NLo&OMaf0 z)EsjYGOE>B#}BTKu`Aau$aSL%sEGe+(>x8=*zF*7jsgDav+i!cJ-o84#SdQgO`xm#dH(c=IU-*nemwK1XlWK0OyTv@ zIwn)I)cPjj+TTJ(2QqS0Z;li`fVf*H1(GTGFm)*pd!64!o@|F6J*J+Tm^;}BfA^yy zEs0QlD$n+OLt13CPQSUAv{-tk|5M4Ot2Jgz`q9H9U!-UTvH#eD!;1!L_O!C7E7=gH z@0ZhIkYBBBd#dPO5@6!Y!8}F`r)ngUFeG)Yk zSwY27X+|aH*PyMwSCuA0EpcaOv&LImD^8^HCyJHFD=|Nr^l$|XO-x}6VotI1(#tkW z01tj?`k#$D_-%!$NXm3n^JW#0excPD3Za=+s`m7|&~_W%98S#vZqPpIdcdMMxIU6x$d0ImM4a`|#A zdg7MIxAYh04e3v{W~@GeB&a#CG7p-fT@IPy1WeCMF*Sbe;nR{El=@0^!=GLMgmR<6gb`1sb z8{1A92d1BU0KQiJ15LMl54ta4k#^eFNC_8K#rr_OzGG!s!^-`N_7#S4cCqRc^;@=$ z!*@g9OU{8_0h_Eq{~|`&b}Q$(hlD}U89+U%XGQxkZu3V$tiP!yOdQM)96Ia^{!*DZHL;rLiu5QN;yhF}eO10=C4zqG}z3f+5;$t76MvtsiX z<)y?3-UpfX$SMD~%AJ(K=`|*gzl&~Tg@V*7hj-mme?bXN?}M-?7J)QefM=bIZ#=|v z(l@oDuah&hU<$Q%O9=-0y)sV!Y>wJmt6FC)XMis#m7%A_`WSK&^ecZ0!kIC^>yE)C z<&aS+#v}C?c+%pax|s9rYma3xyyy1!jN9^EHhyoFWd|bLghVE)h)#$!`9k>p!s)KC zZN17OO@*-y*!?w9InocY{U&2vl_(bOC9IZU&0z21w8i4wX6ka@b~T=qKQkl0f;y^P zq)&XWV+zJEvw?|Bxg%mS^@Pf!SNfRqyX%QHh=Hg9`>Xvxo$H%%z8jUKGN^issn;t} z_@DhSxM;(x1 zhZ1*FM8fG^Ez5EzkF4cs)0PiK9uR0sBJb4rRSr!Ro6J8{$A#A?q`jsq@Wp_^&wE_Q zm)X3*@s73c`aID$lcSd4QOd*)>Y!LdJeA`pN9yBA36`6Y>1JI<3<1c9c=6yR;8SSLzvF2tAC3Y*0jBXfbJ!@Rajd!`sG z`h}%_Q*}Q}pYZeQIXyjYd;D#^BW3y6UK^6T8`2hLr+h;82<+2mBT?Un#y?*JzMr5R zv_tupctBGA^bV|i7V?L;FV6&kp2=jssN-*fi2+%O3*!_IW9o1O9GNw!wLkG_gxMZf zJK+jvz7So6Oj}DpdQw?w3(-<%Oqq>i(o4ToQ{(Iuk3mVs&IkKpGr^P1?mLuF-u7pl zLQRRCDL%|-Mw&!M#FkYx077@4O6HE)7gqU94S;86VQ2h!Muf||Ft)PYP5!+GbFYHF z!253@RD8Pg489oe$sH=i3#km;Xj=Wd6|Dg*9ysSE+r~1NnzXSQ&>{)LxB&=U^MpGT z{zL@>I89GIBKsHG?FAi#)~e{ve-iK-bNZ6G7t%yBAXdVjRY|~7ck7yyf_&wbZ%V0z zJOe!c&KgcFq5CRd35>}xQ+}Fz-Q(!*rrUAskx`b2zt^3@dt`pN-ni13My#l5Ml1>#LUKdw*^t>=}*g@!L|8nbCYeSVkUCjViHCk zB=btQtpVusx8F8Un;)61Mu)_`lmn$PO3rQO|`l%hfS%ar%{tO9%=`< zWNSBj=4#h_E$xZsd^3RB1oSFloNyM+WpJQ4H)p%^t5*6;4YiabCS%EWuhMqQmgC6A z;O23NQtFiDvAQPmJFyi%RHtmr(R6ioda=e5S2Vic%Wpf2TMJul2s%T1xZ1#SbeTOS zsnX&?b58nf`7W&CNoVal$e|*wr{j&%wt~~%CGGa!43VlPMJyrjYz=C`-?u}+@=>oj zzS_X<$QR%_vX3X!SZz{g;9$#`cfyqqjQ!YV2K5RFX@$&UDsNHe*4h7A3o8HlX%304 z_U*E#tdlR0n$6il1&}D@xWMI`ckECpuh&o<<{joKB)3zKd+rdu@2nJqtOP*$IZrKp zlrEaTv=+DG^Ivc@QFCghToP>*0Z%qvA}SdNQDTtJcjOFd_SOQ|Ih)MW)7vrf`v+)} z4bjJ=+w(QM7UxIM&~Z>ndy{X@Wa{ZTfA9S3j(K&reT_WoZuCN>a~4Uo_j&BP$VYp* zdp&HX8y3}HBiv9oqyB*4<2d!c=6+sygnk%YJiT6vy21x& z%kPf00sm37F;~QGZEKiUm|s#{4w4~Bj>@qFh~RtRDr=LX6>uF$kg?7nB2~uC&&;pL z@TYuC7d$SNjeGE7e=z=Nrpyvzir<(P#DY&(uRrSTIkP8&On(OTuDkwS=DJIPZ+9|+ z;%?TW?QZD`)$@?F#pugQby}MYpM*jMxy5Gb`OEc6OU6=~tXo8Y1@Tzv(9&1(L}f!fMpBi0fA$sYULn~x z+L=v`U-S3yI`vahg3K)jzi$UP@cwqZ-g_|4I#vbW+gW9ED^Lla>{`AVuT#!B1Z@J3^f8j8cv$*7df^LX7~nN$u5{0#3>G5p&a zDJe$)Op@*jYnrUV?O;=CNY)P6B%6@z?A|f`@L5@Sk-ixlV$xGZQ%iEWyegJ_4P2`m zLqJMB?yJBl1b!G9B24wT(QLO}(5=yi$I&Zhz?+@bb52r2*)=9AQYOMdcqPdfnE=sd z|LZ1cc;C&|5F2rU-Qg^VVt)bb4$nP@4%J^y&+5)T?lF|QFZJ*9=$WpT9=Nx>1Z>&o zRl9kdkfJ6K6!aS^7TR~>fO)QUAaZ};wC1@a*n8o>k1MeaGSxbYJI_J}BUb;(<=S(0 z{HYLrcNsOt$S*tYeHIujpDU68!0yx~+>~S?C2DDNrCBmQ{~*(-Hi|753MyGi2J6SJ zJS*}LF|WdvX9h6?&a*`R!sWrCg^{$mVrA0(Aw!?g;mc%ie$L410Yy#C`@QVHfKC}n zzE9G?g~Lx41ZE|oYk6Y`j)5nGH1E-6mLfsr(KbG9AU13?RfR^lrJs=polZQ&Ezb(l z-&-3ZlU0Z8#UjYvf%ByS{Pw#0xQeRWGs{7%$q-<#|Bhc!0>4tzBeB-JkZ0EL->2@w zo)u@!-1jao5CptT&|m?RlRtZ@+xeSIAqg{mB#0~mGBA!dL+h;^iHa2 z^M%n@(lB)KHeaVv`DqXLpAAQR2GDIrRnYwMV>?03T2%Hwc-y&`fk5HRs8CSkOS2w~ zp6@$*9XF;Zg81~57<@QCfPQ(YDM#_!x>qr7=|{8nc0-?M%hBp56>^m37$a3Y2|ym9 zYtszH7}2!&NHDi9P598<;89m63~}z9?D@lnO{VtWx|--mJi@Dt=LESkF@eS29bX-a z2GkE9>ZXkCI~=y;6}r)`VWqR47ahr~w$*eV%#UXJ>pTN^W?S=7ok#fZd+`j>6OHz-tn_+1s} za5H6+V_)=O8KbHS>^=kJgJ0m)E~9ELX(75Z0juBJP8H6~NKukRF;71>jr&jZ_zVa! z@Ar!kdY@JE)qg1^7)d2@i{%(Q$1dlnm?&V#*oHWI@1O^^L@Lj|hs&A){)>DB6TsgK z+*?Q64h9+Q{{PUS_kXOI4PGtw(PvJB*esGd!e(FN%C%%EKDW4ZD0K?=(R;q^rO9~n zsJ|z}^rg3?5>gk@+gFC>%uYu9IQP<#VZbpi5RWS(X+gkOcAP%7iUysv{3XXk-48#{ zj4hbZz$yKZ!=gORKbbtQM!7^-?tuJIP91Wy^S2b`2#`21gWGm{`Gl9=0@Qf>bI6jk zUH?d#I*+32w7Dk|j3vK0H!)UrzE!Tr-=khO{q@VT@29NSHUyxU4TIAqdjL`6Qr&SU*+b^EGo`Fd66D;`*{&Nl{C73AXi?!817}x(cw2t``Xg zwOR-uSgr;3e5r>Ua1}I2rdz-TZZ~XcG2%oNI&Wq?{9+QD!~h74Sr;wBEq`y^a_==i(-!wPZxkrJ&n+k&1mmXBsO;VVdkIb8WMg;x-+UW?{s-)41bg(PN!(-1v@7R+U9+W@^j+mMG_ zBo&Lu;KeDsinRvX@iHK8ip`A$uePOcJ$-fbt_$2?eJOxof_Z8ga?2n!g>aiwA^OuS zS-5@qn+Eo&ywsqe66giFqNJ|gmUwT_vN~+LxF$}AznWi*%W%vjSN&3lJCEy)@_=6 zJyM+eGHlVgz1Z2LuGTOf)*l}9XFgmYH(;la-E_bxv=>LhWO#R{?HWXP4Bym@{$Lro z=fP$PF}4X9&_=3h=JZ>`xh0m)y{(CyFTZK4>98lu)KS<#u|52=ywe&t{_pAA)QE$* zM8&x4U+-loupC?h>HvJiOFF@Ocg{UzG&*tz;V1c%7^j`wlYtN-M?2FBqV~*o0;wo5 z|A1t?WUIa^9qDxb>tL_)gRPAMei|3MHH?r8nDgGxhYQtDr{k%zSX-!CUeWmj&fkcg zmoV~2<3ax+JOQ65BwD1eku}W^=T0SlaLlZxqTuw7kW&vKHGmO?{=3+6rFMz5t_>vd zbo{l0?nISRV*_N@+_a^`UK;PiuOm>{j->WFpJj*pc;Gx$Zq_@rD-!mCkCK>lP<(Hr zW{@W$lVd9=>=vdBd2DWtIDF;rmqC-c;Ig+%j;-5L9RweMHN=RUx1j&tU(L_ye$rVO zDo5T%mv@4M7}J!~s&mT}@A*(~dTk1?>NCCN#yZG9958N8qrY>bYA^tp^LPTPb5Hi* zg)CWP@UjYSSb}itm_{>A>FnnZJBt7vsu;1k3L8t1qaU~KzyY*VLB44xcGb8S4Q9-|` z`7GPDU%iR(jlsQIU=(ncCf{)dHWm^n-Cz{&NQqMK?S~m{eO9~Dbuky=ZgtkOV?dK} z&?Y^6{4+@P-t0o3$ks7#a>lY!rHe%?RCrU-cgw9`_?Q^qw*Eaz$yX$4qkyF5Gh8%$ zLq#nyYfwdRxyL8;jGB|n8e4IGrOxn(Rz}L8UL?x+Hz%IXA+ezKZ5hfdgQ(SnY7F{f zY^#|)JdrCL!Z#n{#VKX7TQIP0R|LVoOkpVkqYwKP1w%_sIw`LDh5y3unON%oz6W_M zVIrA>|9<|pZC)2~07{uOMVy@ut+Lj+~52HXcMWSB2FB>`&zf#D3@pv;$m&sod4XbRQZ zj)qQG=G!A&>UyLR!e$y?DH*SN(Dyvk!`!F!wa5us!W4h%zXG6s%FH(I`f_H?8hw%- zG0qZHnWfyoWJ{Uu7yA4u5&vnGL*B}Rh-2c*kFp8t*Q*t?c=orpy@A2jf{v?K*m4It#L9)3~fcgW+ z&(D0PZOgtk^C}S@$7yJ?%k6B$Asr>mKby=E-T z%X@LtIn?6Dx|J!^EWPdYdEkZxkKD7JdltVywR~dH(K*%+)xA~f!939an*|WyZlSsK z#w2ZvVI-J|f{JONe4DP$8Jd+nPV|trs|N@MokNY8JX=uJ$w(Fq6bPo#KiUcMayFZn zCI9#(bes-xv3cbmsGgp7>~&VK1Ks1Pwl$e@Pf&ah_P#v*hF~d_Yz4*J^1 zgMbb6CmlZ-;+RV_sLf7S?Ru&!8T|9_R;rtzDqKV&jaQ)MqrKYP`=scEnwBzsD~9l# zA`u0!WvwXhbWi5P*NI!hzoJXhwP4F8QOxTKe(mH@j@9hg;$szVJwob&XQm?vNLB#9 z!(sSpWt!P1nHkujY#X6%an{ub{$~p;BHoOm-bD{pN7u8r_q8!ex3fP1mk8OA9R>ru z-4BEeNbdg}l%=(--JTfu4$phR&pV)h^!`JdEB@Em&+J$-!Lj+{f!o0je5k(c`D#mc zXf!&jrsSCJ7MCuZEF$N7deTtzY}u(NXR~s{O^8xpCfDJmHM>u?xTs56`l~LnZ}|T~ z{rFqhS$V(8xoW@ipHADjzK;lc!1j!t158619vSq`&uV4h{aGi{RV1F};wmV`cx!Cc z$dJY^XCo9Zeo&FmNy8v@VRNUFAnW+|SvT4Bn$LMp?LLNib<>3_X||!YxqkkMS$>{K z*p3HZ>pckuo|ukNmE}G(<1p|h`3)A=rE6)v%w7=WZuDT=o;X+0*rU*OanqxpBS2|q zdA2mGNs>7EK2r(t$i>g{c*p53?d|nLr&>d_Rk*oL{mO1ADH2vXa^F{kdptZtK>E<+;wkXt_QxFwm&cN|NEQx}toEm6Tm-4W{R) zOsT|7B&d0y5%*s&)XFLLt9(ugnmqc?FS!09BKP>)jmYiMMaXn1eygXNxL&1h$4}+=1R5F&qkfL5BWN0u*NJ6W9rzj)aU@*3% zA$uxW$5LsvN4BvKDvT{-9lNp5FqoMJ-~0Z)>-T-v`TxGI=lSC~_c`}{oo7Nru(6!} zo6OX3>T=56nai71`&VqU^5~PIE?Sg?w91x)fH`qzS!9|LHdwS2dqZL6MJ~s9 z;u7ekTDx9gQr}orG?m~rdnw1OsAHn}$D7KWj$`!u+O#IiotemuM+)YrfzCV;t4Rny z0^WPkX>2+julwX=u`vCgLpdP3LlI1V(dhoKKtX2mEoFPcNQ z7k&TmlRMjxWyzo;RN}bhT5B;1@RC;Y`0qeXsIW@9w~HQo>z0I#cKZe0)@b&Sg@1XI zb_zk@+U};yV`z-9TNDaYD2ovC*#Rkc4N8sON{L`K)W9zl-zjxy%M)l-saSCix!WHUmdFCRyv#cmye= z{L^IF=wQjGEVB9>fBoqih6$Vk{ma66p`$%UNJ|0moetr>{@B6aT3^9$)7!jg@(0@Emx}&Fm&aa3#MJV6}Q7h`T*xgYNd2sV1X*>r24{g4jo7VkICb9#% zZ|y8AlFA%~d zgf6Gcrk_P7Td-xakxZlNVdZ*_jUoD|`x|`bEt*iMhF{@bBo1EY<1&F!F_~Y5B_9UN z<#|E!MxUh=bd>b2bs;ex#&y$p4zKdsbk6TBD0Wq-1-@HAaUK_V?&B zJDuIHx(GS9-2o{lP6i>xAW|-!GxhFkD3%8%$He3t&m|`^+}G^hUAl4#N(Sl5g+BT6YE_&f!(srUZAo ze{OUy)W_uI4jTvw>}){~Tz}Vff~oeB=+pR}&R*(ywwt`qO8%y^Y#{36P|apCqZhgH zs8o4G!Hg6>Qv4TlFQe%^#+GHEV3Wg2t<;ESmC$1aL#gLCpg3aV+wKukXSPZlcuVvy z3Jr&-QO%u*`OH})Vm1cjYLM!*wFIf)lnNpHKEQWFf8Ee>1DLrz(?&B|K(L#HL^H&z zL&YcCTyr|(Ge?_B$`jt6BXoVAZAj7xJe`mMIjxvn``jnuS_;ZsirWsP%^>3i z`JPT7DYb9(I&AHiZvU;mHs#s3nrVkuXLsRcQPmb;xbMfZLYi}fn=U> z9CGO+I{AESJyp`C>N#V4_@eGu7G%f8G$hrje`$ey0&ZT_f0z=%C|Wk9UVg&5oVl;9 zDapKu>r(S@Q{1-MpP4Ms{Ok0GW2B3IDs^QVxyMw3ZU;7enWHsD^;gNn9lF1LwuiUy zjB*h6aC>@i`k+H(G|5|Hg(lSpSsLvW-|1lhiZ}6jBlP6(f;_)Mk9OWiyD}&OSVrY3L!ggV-ZrrowU+v zXpnm*nU@)HK3=NMFK7MkE{hr_)jG7IW7R zJ8o5U#HesS^7?$~Sx#zLyG}tJ^U2cf4q?A4npfHFdG7jjbL$6Yy5Rg}a(9DyPMOS~ z=h8N(y+CV$qE{r0QaySimx!1I{#5ep1ucbKG0C(%I6o_qmyG3cIA-*ye>zks1yJ!#uYx?Bclk6-sD2 zoGls7bKaT=L&OISRRByfi1W%kt|yuu)Fs9Q*btq4oAhw z|AQY8fi-~^uiJBSTOY+!sjpu)1mKIrj_l|-AN2+J>lyEGL8iI#0cBiL(&RJA_(7g) z!(7L?9wj)-9fy31dIiaNCM=;D?rX> zaP5L$A9_n{v@OAUAAno+QYp<M5r#OJ~H#_Do`{Mz>0Z!zA z``%x3W4RP7E;9PGcW6)PAk<++q1j2Y&P}bM65ow8;cs#e@S)Pg`16vW1Rrf0FNahr z?_$ngqYLe4V}WNTsbE+ zIq*I{hTg19CE2Nip`3A68KjY%=4)rxt>VK5+jd(8k}Qt80yR1NcpK~IN`0$7+azKvUYJ>uj_2WT2*;KbuT0;&;6RMt;X```h&$tW_?(&0NkZ*E%^hS zP-;AvOPL;zE+tAKR?zv%y*BR63G{*zg2b^9YF24lG@fo89w+ABRxdh>lC`y?EEpu* zCLY&fM74}G9qiPHULq!bPoO_PbwIk`=gLh;G;wnv%p4L+4i;|ho+xd2FfG5KNT7#= z_wz%i!DiKG#{Jqw*R^wGMKr4likVX& z4J|_Q$ghxzc7%_ca2}Y3oq}N<@$94c}X+k4a|*0>uxS$PL95mccOYA zJzfMcfD3b9_fTRfA9NL8Y;1O2b48Ig6%Px#tFcblGs@YTr>kAmBC^-Vu2+dpiZ)5G z`oeYZgMild4}7pC@0rrdP-V#14pGrIBVGKm%*n%%1j1yoa)mg)>{XH@<$I+VzkyBH zT}mLKHJBQ;zyBi2ul>VDnw?^iLU^IFMypC=rU?%I-TN%lL}%_rLp{ewfEE@iVoJ?& zctQ8_I+&i0oRLR7)Ixg9syn0Px505PSn%R(uq#fx1xFgfC-Jhn4AN&MbX7L{dkx^6 zTlUcpX4L7{7ug-k5h~$*b~gS1{2-qLi-*Fjg>82G4re-6brw~&siT*VDbJ>feuyl; zkw?PUt+Z$plQi<@AE?}$Q0tQopgS~dE1Vk}z+=|EwX9o%=a#CbS8sQkS5W(86aE?b8My5t+pi(PQ-KoI>)-#>d^aO46AW>H=yynB%l;(}KlBCs*^d#jL?mq%lTRe3TRGY-B$%ntePAy72 zTHOu|O9Lr0p5g!7hu-C_Ngd^K`nR}G?QIx)UJE5tO}oMf{$EDHREU$IPc1B)ve}0u z03aVAl>1j(IZ?Mc+Q6nST5MW2KUqrp7+~}gpycK=0K~bU=g0|_OxpY@XVgkLmnzIR&N`lBHj40j;^ zBud}%>Ys3xb?P?{m(GpExSrk|v;FwaE~5E4^Ks>}sS|r(=z=bf^YR7V`ziQ3qQ}su z)T2u5cgFi)W0eO*?A}_Jan+9g)X~dTVkK%^BM-ehW>z(<22p(W2ghrf-H$G?KzKFG z44&L{f0U@MavkpDP`Lh$A8v%-qy<`ho&Cj=2qR^ioT-CDP_E3sy_g^g!;fG|aeYwjRcVJb2tA<#2imnHm+I7t~e1E#=DcFq< zJU?UVIop;N-oJM#dXs^Y5R*r4AJwH0k;Kb7g&tqC6(1~_c&)eeemUz?l9Xbzg)+zPi3WjjJ_#iIf3u_l z(g9iE=5N4SIf?{(i;qsY$s=OVDZa5qUoF-tBe^$N4Fi5O3y3MT!lIXM|QD3C>3 zN}dwB)p+Z&WSRQG^3x6K)5CtTan^F+I#aNge$|-qRL-u|Ok`y2N8lLVRhKRCLZJ~w&JYD5c?>#(geS#>w|R^%ObS&Qn;MkpeD&ChmAa-_0R^j1CR!OjK7 z(8)dQ#a{7JDOVv%yNk-T<4;ZYpP>V+bSIB&WAyZ)e!T|Ti^fuDVKo%7n-_kfUB1@n zxO@zpm8+RX$^9ABw8^ymQ_ab?cUPzU-6@UK**={p<0K|@Z1SU5uS`()Au+l3q_(`% zuW;8N6Z2Uh_XP*~74Bi6x?TG$)dU-d-zFK+!D%FiHn#HqBf7u zxmHY!aI!e)WIpK6PxEI$qw^KN=)OPI=4J%f=Xp2-{G;=_wb#tmMT+@r-|Dbp%;eEs z3dHFm>vYU${F4a-m>F`G>tTeRj@)Quyv8l&le#3Wu5>ygX1H$I4< zr{&3mh}45bA3}-L@%5X=Seyh4F*f==f@aw?-_EZ!*rZ)V;I;_$>|={J4uIHF4`)D( zZX+VszVVlc9J-745BWcIyPf|}C93~JB}S#$E5OIn&SGzUHv-OD0!~YFiJqoJYc;oc z0C-7^XBn{wypm%bxSh+Uzw-LA9Bz3wTCxlHL|f^|t??68(mc|((iNWI_K!YicsT$E zh~-W5n;&f7RoeKgK%9YrdI&DPS6NuwD4oO}10lXW&OM0p2WcVp@D-x7jwQs zw9`jo%*%AoedwYTElxgN^782SL%Rp9%|21)6?Qqm-)0|%zNaq9uMhcE$-Gyl?C{>4 z12qT!Nfw~bSB7@KE?o@k-B56MgL-2ZWt+7Sa!3e(%S(uF>-@^g*BhhlPtQw{`rkNZ zH;|fQ4l8TB^Sfp3;4Ri6*X$fR0)*Cl@)-MS- z1a^Dgc~tT4q9nD=Cek&!?QBBdTwi$lZ7iL?Lnp%6YUPFS{Gt_eCV+?)(kHBF*~-yz zDvmc3b9eONK`nRBY3)pADoU;fa)Q5(;O9LKqDAJOvrgioH);Co~ zZ4l$eCc4*XR7z6C#%VTtbF3mVyd%c1C_Ej-NK_6=@QARF#}VfAWrFuV-u2T#=7tBu zd|vdG%yrIa;YC&@T~Cj;JC1ZXHPm_JTv?L!+=d3CIB|hf|I8MZ2v;%;%&+jmAbiy4 zjhw`{q3HZZjC+o_DQ0dD(_(;dvx25E;uaUW!nVr%@;6Jd3km^6Yinz$tp#a z$!xnZ>(t9gh8KPI67iY93GHs1v2Utgt0w}q%iLUA-)bCar|KaaUXIpkUVFHo;DNQY zF926n8dncOoVE;oiiu(83_<(d97NjcG?uyj&@hrd4_?>+IAl<5IY)}@~f}b!E3||HvCI#%V8k{1WEttuY zRH;MOM;uAp`rSo>vl+VKy?h^3!Lw@_^0D74C7I@OGp+lnY?gcZBd?YurJ+t}sz5Ny8pw%rpnjClFXp69CGEVh zmih2wQ!4lG+O3o)4oi_^!6|+0wJepTX|0m-#-~r#hm1RC>gv8HWP}q!1jeWLPI0(`eCpJZw%A*(j#TCOp*aL#XxW6Zf*)set_ zTx8rn_X4ZAXa}Zj zgoDCjWt-pc!K4RXO6V@9#N|sb)g<2OVvJx!R9#^)Dz|U7@~Oy3&v?JF@nPA~zh%n5 z6F|sjqqU#rfLjGrFtqv}HU>p$#Jbcxx2XH_c^uVGWix-EN1ET_9}NhWRWhGP=1-pe{UN490}H@H7+5(G>3EOk zo_D_+=Kvv7=c~BHz#BlO(s{PIyq>&Rls7bj%UXAbJ16AMuWQ#~Q~70@pdW?2KOPxh zOurDiJ*p#pN>je3MydN4?LO^jD)%4s2PtPW4ZRDdcS^IGeQ2&>y=1 zlsTDrFP}uZO!n+rv*kvJ?MQgt>?+3HD-?dT`lR>E_C9j_8$FDA=rvIrN^5;tHad8?NVbdSi#k-p=Z+PMsYlsIk(cgV_p>yoarxcFtk_egijY z7d=qOD|j?9_2n!+KsD$OU*x>@LJIi)PyxuJVGySKr|Cy?@H}*HIjJet!F;aI+m;5S z?8x`g%72y9WFFRcHtmeJibd`?e7$9*tK2|qebq)3WY4CFUf3P|s%MJzn{J8`;Iis-8z$XvRfh9KLYWa>BIDjd@2;AwA7dYRo0%~pb#(YlDP*J6=F84BI* zbLP+ljfgv$;5c6PfMeqCLne0}TKZzY4wTds(e-bQxv?`Z-^xgdRWheN7>U&p^yz4! zK!%dYmJ;)7(Wb(~rbHO^mqldt4Gt+{LAp+TX?Z zFf|X0-Jo%^iPfH%^yJ@4dAqxk4%5au?E6;zx|5nr-pCioCSNnVikC&?YmkxoOLeUl zWB{B#`8@w%Ijyf!KzzdneRV`eo^(O#H1fgM+4~1A%4732E%AHIB@pNCMMN?u1b&G- zyx^CzB1$Z5;SSpiTWHD$g2~`wAJm#ixp+n3G!XaO!+cy?9f{Es@dtK3jp~05wmBTm z6FYCxAKk_Q<=mel*kj?CfioBXzKfYPOekHOCkW_Yo_AuMZ&EVv&we{&COmiTpkJ^F zogx^L4vMVjcpT_2F)A0u>fU`>yK1%8iIbSTo6_`nGwX-%eHWM;W&gV%>tQ8v{{AG@QTd((FAI9G+H+q7uAVcik#CQ6+@()tOtYzyEFc8+x~^3>rUuqG57-6*H95a6vx52pvlmKCU z8_f5apQ z=MmFV(Ct}7lc3sw!j+*xC%3iBptr$o3V-sP?~yZd7{@$FX|?CaUw(RNF?*82cJ*q2 zSakW?PREt0?9=4mS~Oug@%a1ip@g_#^r~?a+hBb3c6>3hC@e*Fqp;@x4b=DkG*6vq zsClxR3nfLlm-6tK>qmz*#~1bW%1P<*l#RLO#Xni^O#OPuQQva2(WFMV$jlGbjS8JXBYEEdt$SB0CSMKAy6onh7idA zEe{J-=psuz^Usx=JY0AAYvttSKQSzAGiPvIVl+^eUUO=bbI z?8}S06(?W5CZHBEh`W(;InlS^s^t}N4^>_!(}hxbQPK8>hZ|S6-1a_GM%C9@-_)Go zGVNbXzVyKvLjfsf^s&~jM?tmpq~{22-nG5)oVH*(6JgBL<3IyE9cl2bLxr#*;> zi0&v2r1?WaE+td6?#b~Fx@_MjjXM(_rrD5r%bUj~iC+kczqiwNWVm()1dz*981;+f zXHMcv^M)BnW)A&+IVM3P)LG--)*SuV73=4x5>oCvNc@Q3ni*%h2G7 z+!G(k%FVR^A((BWLbz*o5PGS&fcdQ4_3i-Ek~)dG-P9q*FSvCoZ(gC?s8`j~ zf{W1ft=HIiu=QOB>(mVvf9^^#)&S|3Jf%Z5%!t_GrV!uwbyyyo8hUNabh`4ESH5PAQmr3-Auj`km@sW%f)uPxX&@mtc3{VpAW9hnzjV?UP@DWv&T z2ZK)a%U&5uLsnP~zL6l*bJ+R|aOKv)_NT{D`;x3<56N}1(9Zc4jqg1NEB7d@&vnEq z>%T;(&u@$jAv8+=VvUL{vIRUFO&)+gjbJ?o$M?DBU|bJC_$~E)x*48TuIVnJjhd-^ zQO(@TZTh>`!skbl5N(3fnAo~I7WP$-cuM2EUEk!aBJ5>Sx;Ek07UZG&C9Fs0+ri}- zf~S96Yuee&>T`deV0?iZ|Ba~++w>gxRErm|svY$}aEKZN6+ejy14~5p%=2LIv$)Aw z`A;_AO8y&AQIgdHKUP z*K~IJh(@a-oiQKTZv0m=f6pILD36Zl*_#1g83bNr2MS1~wXwMC{4ZgrXY5tJ155oi zb*@lj93|*6R#)yzybAaR?2ooz(P{uF^w)r<=KVOUF-~*1NSlOcepEw&VepDAtRUcu zrMV@o0E{~zi&v`s4_;Fp(Rk}k}j0t`SOT^%gSF>CT{Zs4>j zA4%2vx+tal`#Hv4dB46mmEQsT!>DFu5Qy*3KwzRA$VdM;YxvEe$a#wHzbr`p`2`mK z=QoNL>iXD4vb;{dqR=`> z{dR);3}A2pzPnhz?o@oUyn+#;Aw`CN39i@NgNRqg^=MN$hITX9s?YilN(Fw-}N-)17B|oA=Zb zAVH|LwEkJ)=?}2<>ZpqNl%w3YzQLp?T5~im(@c#^YK@JWd!C)}o<5N}^yVuZ*yuw4 zUK+m6kK?-6=nb&8)x7-anvnEJ6W)PqfIAekm;MrBF$(x0$Df^M=7ixRH~@Lf{s8Q; zk1Rw0RVMlH=&Pac+yEnIWW@!uKd?RAEEer9^Ih9{P|+rZaj^=#8Ps)yZ}$N|`@6|P zq|h`km6tsH{XE|fJ9jIF;h|t4y)&aL)~Q3iLbLu&>2!CX(20@gCkWr%6hMa@bnF7B z)8!@K(44_c!hCD-(ZXFu2Z7PvNBd>OY|oA__)DKtL(o5OzCd^udRKH>Fk$sCVctFk zTl^87ozzyIj`n!G958aE#)Q|&FaMiv#Y{|NSF6ZP$0RQ(+WjF+l{ymbL19(JABE|q zG?GPi!nIa9pDO8WD=|=xW$}oT!T0hXg+~;t#AS)LBa1jXE{EZ!X6Ph0xy+>$t6V9+ zU(nMF$u@h;yaby*WL$+Ye#&c~b)LwYrr+PoBH1*Y5F+1e+DOYlQh6!I>`FVr19N>o z&zsKw+8Yek-kbcFP~MF*lSbwX)%GN)TU>ip#ZHud^FL$X>XvQYqDt%o7*`^gUqab| zeSve@QjbS!Bk1d6k7kNdiF|+Qv7Vn+0}gsL7$&f-*Re+;i1JL_DvL1Dp~1|-X%3wP zYX2xG$uZ3jLGgXVdugAX`km>f-7k{0eA`$|{lUZM)5pEJ$MT1ZL1|q+U%Tj#52G*4=%#XeqjI{_cQ0jf z=B+N!7N$iwMo?q=a|nw#-qWP?IR`R%KI&++Iusca7T1v7<7u*Eg=NJYl zCFPm)$=PBEqXDurw6w?4w0j7y3!*g=o!Fa>>sVvc@scZllSMAmDt{FhbL~2U;rA=f z!B(~}WQ>)+5AY_}^syY)D;ts1T;=Ts{H|r%8^z)a?qEv1$6~f?DRaA6cmL5W=Frr7 zE!F2|-;|?JDWg~!d4J-zMxLA-n2EDL#-fi0GhDoDZ{<|kxEYWF4Jj)L32p81LK;I*P zuht4BQq!~0bzNC1;sFpm`BVAQ$=}g(zn^|k*=UjVI`!}KS)dx4$}i|<-bX+iD#L9S zfImzf3uUdJ!l-ur8jj0R;lg#-jZtlpoXlj5`vZ8Dl4f#-Hl%cQmPA9iwP@T z87}8)??@tCjIH}6er7h(qVrK^2fdsT#+Jng778Mbjl1<~h|(d`{y}ph%&tB9w?-q(gpf+0erb%KgD5`bTWdJ;rGpHPjY3}gp<#Y6nBi-9+FoZw}~wO#9v=- z#vlbZ-)G;2IT-$RjrUr)&sd$N3~x>$eFRYd18oA|?(^xnRo}aq;{rUPH@14ilTpqc z!Fb^)Arc@IAQUxr*HJI6c$-V#nb>;V_u@DV;+#lkC5?(2%FAp`t+!?;*$ zA%5M#g5(a_^dnIdi~P>~&1@TF&V#NiNsdM8sX8a?M(r$Q zIOX}&*}UUEujI>J3V1Z#KWoeD4H)HV7K{1-ofbCRg7G!KccSe06qI$`&reo_Use1Z zT1QDQV9fNWrGpb5%bk(Ch>lIMt9fEZGZN90V|ll7B}4ccKP^5AsdLd)otwFY;d(ZT zgZ=Q}29XBV7p$wQSX8b2IAnwwTJdeasZR;^@Gk{_8PGGuSFUozx6mMRUackHF8#RP z=yU1m$5(IGpqdFQB(;46*^-j?V=f@-Vk7JL={Y>CUq!roB^0Tz!CokGAAkY+4ygGa zXRd;J^)Sc{Bzx`H(4Y!5>$o*mPSzIE!{rYxi^f!(#|kQ=Kj};{hB01gBRV@#5ki`~ zvweai{R<$u5crMdc>E~O!vME7a{nK)3jJ(Nt8w_!-L;ZP)$(t#n7Yfg;eXd_mdJB< z*x&BIcnacSd6MpyPFikvb*f&WGX$7X2>8nIoH1$^IXNhA!-~hB4&5E~D86WxtYBtK z>#3na%&J(%5ozmHZStW+?r~4AGuQh#@0r9BYJ!^FrfZ5H(TPmNz^lE_zeR5y)h2y9 zpcXT`%)`yCSOygn^J2OTdlzV|eaRqi{kL?zJ+ETql|L`1NVH4=1;*RS?{jWHZ^Ude zOC%%J<;;j*5dbFWt!i_y^RhxJpib&(Zx6m`?s&Rs^eB^a;IH zC;afgOg*ePl&Fq?pJL*zjUax8_9sj`Y<3!Y))ADgO2f|bWf^nZS2R*0+qDs=)DBHh z=k#ZL66Fu~r}hXModk^YnwZ7xNBK|P9|-tqb$B~(=ov79`SpkNhoABgML}8c$F^kG zOpNQ`uOF(&$-7p9%I1XyNJ5fK=zE2r&L+*2;GLa?Z=mutN17`qVbE1LV%ZkyCbaht z#J8Q795uCk%S8dvk&Ubp`P(j7XPEGfO48wso(P<@fySnk5MD@8Y&yBFu-*zFRoTgT zbD8BbXD`MbWez4cDg}*_WRxWSG#s&Nb z&cA6~a6dB#{+yL9(@>+BZ$k@emtqYc{)pp9^}2d!_a|tlDh;%)F^Y8euYEflJVyw- z^`dkccbp+_LDE7p4CHDBBNh^fmvoD3Tw{>kS4TWQ8iUHtDK0N@y6|SSn?~X3vWNd5 z#Vav|z`b1Ib>`fRMd56#04$BTYb;*3= zV%r3*dAS|dGj{UIN~gz2NkzJs=R!Ju9vfM2B2dOVz6Fk*6t2EaE)N?W2^^8=a6>K} ztk`V{k8$!vn>8e6;~JLqEdrS{EEa2maVgSh9LjN(!fvc9fg@kE&Ehaha^xc_@^IWA zxu!u-9n(Y(tLxd13`K5R(c>k`${N6+?nhGgf%uD1k=YChMSNy(t1TmvoKJ-g zH}%|9k&sKPTu6k6j#}x2H}foN`exc2O!*$u=G9uZa2hN44-52CAkeIn9VUrgjcy`?)wYiwkN$a8wL#E2EXx z+@5NMBQe4|-q7SQ8Z1i;=S>P+g47LEHY;5+=e;w^4b zy;IH}{lkL}&GCkT&Og(AUqqm`tk0{R^zS>P?JF=uo&cQY@(9$HS@y`|UT5<@+~GpA zSf>Fw+8EEl6X&Z=ti)fkcnZ{^0^jJVD>-u^Ky`VfRcw4lz*qg#J6JkE+q2W0 z?mX*$O23r)H^ees(_1mLqQ+3AAr<5_AG!QK<3nPzl3fLFogB}av8D9XJDelczz$xl z04-y-2LwG9O1L=G8PGhIDL|)aR?s?qM)_ggW|93psI)!OY$;)edyxP5Fd;IksP~wY6BTwcx%NLxWV)=G%xW zj?10%NmE^({)Tcw$Njg{&_5}|1>=~qf`6~(Ov$R;VRw-Q;(g=@BQ~^q z@$kv|jcJ@isiU`fY3bgVn57$+{}QHF#WVW@!7#=vvoAuDOX$K}f91T1ySlkskE&S7 z5%x8iXAfv*=)zBuPN$UH!@N;|E@962Q||1?vX*o^s4!uQabkfyQRyoajNQ(r zV5g*4~bCOE8V2Rm{&hL)uxGh{FHsB1Inp6G6 zfH@1w3$NLl;!Y)Wn!aS6iaE_AEpCF4Fy{^dQQOq%2_-07)=k|hp2`d7BKl!?kuTKW zSCeATq8SFAE3?OO+7}&X)pms8I)(m^sE^emw|9EXNFA`$Va5kNNDVm^I zNOtc?&=6j>?S_%oq(veN+^R3{cCX3nZGT!B(3w(X$AdA&8t{z6{?Luur4LQr-=|)B zz{+iNNVu5e@OF}Hg1wB($%OEsOKu%OxL0+OLcK%AeUZjuec&%q(04+7bwI!JUHz&k*HZPwrq8EqNb?$3Ox zypOE(&8q1>-61Q~r)BOB7nJ_Da=BvVaCaK9|r=g+Kh z*iQf5$c!*AX*% z486g=zchyM`1Rf4%2m6Oj;OIX9I>i~t3wzw&=xXFkWhTnz3u-^!dq(1Z!EU%mXnvZj7Nq=#{c;uff0@&+0NBp;c60*?Qm5&-P92oV!NU4 z6AjZQ$y0tRc>d$6ZS#Anr)VA+(ii;pxS+q;$<3~3CEBrg;^aS_UpY~cH2o)_-1BPy z*)Bs`#>81CQoIBz;)Wn6esFPWwmt}hQ=z>l-xI3usqYy^-OH>T6h7W^(5Q0(Hum*p zr({_gVg5WbEs;Q}4e;1H)|amj!^%jWO*x>W%mr~wr*`&&T?CtP$^Z34oYm#l3lU?I zj?rdRJM*W)ZFZ=ug2-;lH1}QwZ~e5s7IICiDS$E~?yv_xo4Eu0l4^il6j}S`6ftA< zfjVN{FE8$HU}5L~u}_)W^RU7T)edKtPXS>hrBTVX_u&14VjUd>Z9S`xtuIUY08?!_ z>zl;s+xuQaH!~*kUU3+~^e>b|Z9m4wYgsi{-a0iVUZjlDpo{ zW@C>jqs!T#tmOF|$rRoBdyHWfta<(K>;pgaFOhKGA(4BO zv+LQ4ddF@7q~COLDe1Kbf}Bq{=^eiXcx4mEeMaZelanWI-4-DM#g(FLeuG|de#aRa zSX|~ZiOM(Q3v`U!-#Lm_diKeOPMQaLTMPqz{`<~!>Cj*x!|+b?5m7ZU)#$Cuxv{eO z0!{}$xh4Wxd35J7s5h~)4S@U=U|4TdO)*f3+eY)M(5;$}d_!8)U+H^t zLhyfK%lTqJLCz+)xh_SK^ zdFm+R5@1*Nkk$9;*MA2PN(64MeJ$Zd{42BWQy57lv37_Dy67mo&=?trGm+fw5_Ld^uMxE4~YDS}eFf$lqiH_F)frgpP<1PKSRT z`JM#1@PDojrXNX=(`|&x$6EMwS+3#z4w?Mrb29MT3X=eB#_x=rm9V87bj*qEsq} zR+dOUA57OpC+e5u@M&La7)s~q4oLgs*5O^m=Qx{or;xamel980n$3_bNIY+`SR*!3 z5Z(Q`@yB(iOMuDIHGi{)GdKc@-R;IigTtW@A*G6_G5Z?ovkwh(H}_n#*hKNrt=FxP zS>=f3(-!PM$>vJdFHXW&Y@N5Fc%;aw>s{9#?kp4a!LHAdxjRKVu&wDN*A|>NHG4nY z7z%cGRilqTPo{4~vZ*PO3xaz)vB)V_Ean<~_m%u`js6P__!MPHl@z0Km;_}ZKP4-L&@__0=+Ho32;jg zpkIx^YO6tdY0f#KOY^P3+-6>e#l`T{Zi9Oe+{_mWXHi3~uiI~BPWdoi*WZ%V0<%Rg z<$1AbnHRfl6kRofd~+kEj5aLs->R04s|)x= zqkCOMw@m{4n8b-#?4{QNh8Uwfy7DRebfINF%R;T}m%gtebhL@eQ~|k{BF-d#b5Xg; zF8o%l2BosL#>7Sapw*nlwGT5E+y!JwEFrSSsQgne>vwoTn!kcawO@P1$Op}KuASYh zl&qZ{6pxQU4fXTp_JcWVi5Km8Wy6^>IJJzy9+!Dt*60V3+-+XAI}{g2OjdacS`{IR za?Bw6&8sV-cJm}~_@fvKgV z^;p?_eeCf)L0~i2h30S7fY*lR9#Q>){DE@+ho?9Hhl2gW$7ig`GL^Coh7`$`ZR|57 zB+KMcAL8{RiAX z+^_qb>s;qL*D!dce8?&%D;oA;c~4jV`o#zwC9VQ<0A(KqWrsTsJ_8sOO_}B$iwhAftu1$F_yLWJ=uDdreaXN{5+A>xFqm)d6Ww`~$UTW&CnXHGen1_+DsIst_LNv|uGQs@B={<-o9x zi0-+^^od=PJZcs;3wDAHPI4e#G*w+5O@#EWpGH>l2nB4yy>on&kp=80Ed=QQpJNj+y~Mu7iVr8)Q-d3 zT0F%a_zr>#5_4_1L^nx0h=dQLKPA2p%%RtYW1c6?@J%gM35dwTC=HyK4&vo!IQ-mC zg^%;MPmLIuL_UXz*-1XWsBlus8-GBvL-M`3?;pqFSxz; zAvmfxYpmVASjovFpUQ*kbrhzBiiFCCj^3TBu|!msng`dG8U2STlfBZarmcUO>3RB_pwE5zsUV!BTDdf4tiN2qz1V*{qBW^;ebFw0 z@ih*MIiE$H*S8qKkW#cB9>?1bi4x6xm4!+^E?QL>BPGwsLR7`j8Np_ zXZJ4*iDRmf))TH~%1uX1wEh#o@$g|Cr{cxn?hx8i^%T%=e>7e51(c($w;V5$#^sz% z`Wgd`SR;Q#9<9^3_jsE#h8Yr4aVy%CW0wZ1YTd{;8&xrzQ60j=sd zeF0=gZh$&pKAge}mn?CI@2Sni5JkMNsiR};79?nOOg0I1=fB=W@@8jFW)Hn+?wvC`pB`_MQhe5NFaj z-IP>MDUC(+UG+P(CX#iO=B{qE?bn4*u5HdK|3B5!|0tyPNLWY+B6B129$-<%pIvA~ zpo9C0{}lyLak{@y6}vwybNZt|)ttO7l+9_ zPdF*!^S3yz0O_(CP@v)|V69=9l?vdX?QlLX4ES2f6<8>=MloJE#%1&xPOo)p@;;2d ze1_LBZxa}g*UH&>){dVK+xYZBYjMhyhoN|@81azTQHZ3t>)=7l)Q}Y;R~K!ZSTBKZ z+TkQ-3scZV>38*{ULPB*C2$Adr9U55;l)BhK&LSEN2)QqLWwEROd(x1cO`wTfllr- zAPUT+s_4EKcC~}Ex|U$=w4jQEwqe(du^U;|q)APEE&s(dN_C?7b0UJpT> z)=LkVQklEu|0ZQaC2EwrZDS$Aj5SR7L;MDa7%--hZ${BfqSn%@g!ItqJCNi4GD4a+ zYq%?+$rBYb{JBP3#%1Qt^vI>w6ycz{g6F4nULtqQf@YF`Ds55!v$n*Y?Ad~hDHZ~n zZ+l#xieuk8d3t(InE3?IN8}jFQ7c~t9tBElFN;Ak;>Hf|)IjIn${bh4|z?f?~x4F8#|YrqtC9--rA!SGvgv>dtSV@8pE8~PI<)mfw4uvgzUcg z7unQ2xz#=|b&Wt-u=mi?uPca;rm6VV_n+ry#7dIG{OtM>Fp^baaBz`Kp zOer|-$%5@3MUOIIz(~5@kV21BUe4wQEd0-=9dEyTY0~kew+x@PtMJqO{ekwL z;9W`vXS+35p=YXrE-N(#N9xT)=v@U z{X58yujsiJD#IFECIdU|xi;}V-JLB5j9OoISYR#+22}RmZwdYxcNH7=o0c@a+_+5@ zCK3#E-J=gYijY3v#s~&@ujJ-b2-IIzt8K@YH2vB6KNMIP$)8|MXqfR7PonN-=zRAl zt71AZT9rrFQ(JlT5)jKT?5~(+>y^sH_m?i`us#(0o5QrfTDWJI5SP$%N-_sd6LnpJdz{x_4`WL=OpNH ze+h_1=?>9nu&OyDRHv2#RVO{Ek`1pAy)L`si5CxZC%!&f(x~}vdV$Z)_^tEW55gtP@%WNk4dml#@o*#%}v}73f%ab!3$oc;NP@9F2>EU-+EX=IfES@(m$}JQ)5jY*>zi)bJa0*`Uk) zBD!udzl8*lxyc_c+37ESjKt#q@d~U8x0?)6DhNJt`#$Z9`#*>sM*L5Eg!9}$O8=CA$|2EWyN6OjB$+CzqJp@K=rHiJ`5+ox3}8n) z0*Oa|Cw^e-6be8_>`&@YJ4zEd+9s5KVwfY(V&}(9rya#)uK+$ILL#^~Ev07c7J>}7 z2Z|9IK2o`KOxUn#73a;oEcEYzo5!3(T$MiQyO(*pS&2W^cfnGH^9G@bD{g4aSQcz; zl#tdasu*3T?yzGN=!yiUHXz(&Qo~Zj8$e{qUu`v4y6BhLy8@z zF+0&1#_RwguX8u~jj!p>(@VoP&me#dq5RovI*N1S(h+NRV5X+P+?ArJ!+KStC--om zypG)4Vo|JTE-8-tYq83czuM<^^)jDK&)y3U`t})M0ew0tZ)$(1dZ)_Z7(JJFL*B!d znBC=P-BL7Km@J)4=z{zH?NYZfHt(b#I*>jjC+l7+R zJr7aOiHT`YFS=Iqugh3Qvohz&a(+0WMBCH&cDQQEiHtB*Dyn7r zR19fmOM@==ujZOu#TkS;TUQr7vbmM)W`TMhIg{+09L3F?0NnixIVZa`;XaS|akG22hG@k?{lf{5lnO{(G-R^K6`G zz-uT*NNcRN+5VN^7)tQDU&q<%6#A!6{8xi7fx6v6rR2*Ml&EGy1qdZu4RN~s^ zmYY15W-7gKW^cmkh&~&Bps7rIV0_yjSEV~X(MlaNz){gdCFY2rmv$&y@bgr4 zQ8#U434Ef~?;qbEc)NNpu1!a)`Z*0kmJ4RM{4jA}oixNR!g9ECjlX+c73kzzj>Ag5 zP58JnzcRI)j@jxwQ!^Nos%!Vg9jbew-ns94*v|;e{SiWY6KRR>>6BA6qItu8sI(Wu z&8H~PeQ!6%>6xbp@4d4sN86$PEi*}Jl$U<)4XujvNjZ3h-XX>>;K}*&0y~E_^LDzM zTy4K7e&YQPE{hy+9;)A#{%U%-an?eyPerf9Uz|bsG4nzMzu##+YhCyzX(4pAz+6y5 zyIcPzK2FY6GOUj~Dw=y}Ux3}S=J{WEBcwsV*jboC1+Qo43kEyHTp894i`;#U?Bpiz zIBXd?;;&~~TkpU9RLeQL^PV3m6&I=-Do*}&B6*LH7P+%tgb2<$q0GKe9V#opde1+I z(`eu06(n%vn@aAM%B6LGXHo~xGnbOq7lor>o`M0()0(UdLFMQa3O{U=>tQUi_0a&* z03CF9AZ;fT@}(?|xgd>k{w%8kbN$z z2jv(Tv7;x(InTtNQ2BSOOHBd}>awH6@i&F?UTq50m-cDVjESu1ZAx0}7*URWXyn3g zMoFM+ZiPU{oTly*ob%`;Xs-OdIdJw%l?u1$C^?h-PT^oD$4T~`Y~aYIo6Q?V)=OXd z%fy)8Kq$F9$Zc%wQl8&j{ff-oO#OJ_{O39mpXBUkyMDSPKh}kahF-Z`f#B=55}AML z-llXL^$FBo#l|k!FaGwl_*^V#;=BviBCL8>wLyLI(yHYgNxkoK@BQQ&j5vPxKiLQoP?(6m7vsQGZs+{j{qeuLLJtoq90W;pL`&^tW02<4u5jg%o$ zwsw6&2^`Lzg{Eb~$%6ZNlFrkO>Y^bGKmJj@KQ+z9#qfi7T4tyRK1z5=v&|h@VQ?t# z8j7lW@x2{0--PowOdNVpBts$-gCrQpJB&5svpETnZi`8_-!W4nG7*DIg&Vz)J)kNx?l(GEQr8ZaK8Pw@MELC^JSa1Y#>DhiY3PSRR$ zk?pG^G8l~d0+Bdm!?#A*SP_Ex+^Bz3tgy%U*d+4gaQdlF!GPacv356MRC1a&4G|e} z1wT>KJ+mlWeIe@kbH(ZbMvsSaL?Oz0^D4T|9Gk^*;js1i~-UO=uC93+_za z&A?oS=b;9?p*B5N{BNFBNsegl=@wqGx>lC`NLkb+hO|<9)D1gu^LKXjm6ux!Vkf8BCDdMq zGryTp!ASa>lhHF~U$jIc65t-D?e^j!@#1$rM@<-br3= zCcRGDOW=;LhuogpoFP^!=xwPpD~8`cEH6#6*(sz`gnHy43BcNl`j? zMq+)2iNo=(e6`&nm>0$)1Qzi}@~5g2lgK@<-Sm$c6_g?#^)--W9DDxkUj_~DRHnbh z$FZCg^D;Y~c-mRkzUo;eATZ$xKfD3&W@*oU8zdQRmja%aSRYlYxRE6g+zE-OwIfM6 z3_SB|;)VXoBS0d5sKeIg#Ai*VoY>`dz`C8hOoEUDBOw@VR?fWh&qwS1d$&cOKP$E$^Smz2O(jA8iU0icaaU#sahf64h+lz-*v)5yuYjHo!1i=qB(P zzz`7zjJD#6p36c;hO$)+)^xBv13ClV9FOgZ+Si{seLm+B@FqY>5^c%$a2*&EeeQ|( z_1Hoai&n>apth3fEO~t7VwQ;r;IDSC`*fMC19}i>Tc}QPnH?qT>Gl0JdpfZ5KYK;x zO-2P#FJx=TJoUivnkaf+v)CVE&(*La;C26eJss^o*E0%E#MI7mKg-ov!&&rzOvT=5lM4U-?g_cY#-(Tq{0#n(SNaX^FpkuUY&|- z6{2C!0f=Q(GSSO*!F2TX2eC_a;aCkEjI$Lh=|>y#Wqf2(9VH@PRk^LgIeogN0rPQj5TsK;&guf7RXVgabPqMJd(dVsJy3UC) zf`)UR_P=FUIRTD3zrl#l9)T{mnTV-o4JS*$y2_w7$-GIzSS6;A9a>)`51XX2Cyeh8 z8C_(jFNP2ugNr&rU{igZ5KbtO~l(v?7V^eBMRU(ySMYOPzX&y3Zx&k_E z#KLOXx*4w}bMf>vx&agO3l~)6w=23Aa9g#g=_S6&-R zyu^vzO|4n=0h0Px2Hnt2V=|&X)r&k4o~T>N-{iodhBtBO1Wz6FcCU#y`U#TXM?pC| zqX>t=#-Lw?!|>nwm%4*#HFNjA52{A?-v3S3V2$zSP^-wEl>@)s3vi^e5WIX;ZMliI z|E*RJTj~X+-;)!#=c=Ju0ucxz2DR#5tDP*&8_#+fMK!Mt+r#O(^xYzUA!(z+Pex327C@{H*rDt z^ja>LhqFP&ClS{A@nL1|ByLP5OL`i>Krr@1=q@ts6x zB-1-HOW;Q>X(9Tgi8zz_$03V3`qxAw8Ny(ylA4#?^gpEAQ{GzzZ+E3nq##H3d!6n1 zRggePv%|>K#m~s_BHnEFGUKH7Pvkr;3#LR4YO*)=%Zt{_0$U{v@| zwGLzikIw`N81k;Hce*e9?9+-W6~Ar(N1;ZzQlG62!TUz)45_cc_+F9|OJz zF4yC+khT?>l|!9%25+}`xl9uA z4gVYc+Hn~DuV1QYPZfJ$pGncoXp@sqypJ<{K?L&hN1tEh>u)^s$mrZD_WApp9EFN4 zfNEm0RYk|ybnYKOrmO%U)kMh;>fq5V&Z`4RjRcxRlsLyiU0_>k%ROs+5wT_aF3VJk_#@c$- zOJ>L1n2yaWzee>b1u0F(IA~a^DSCuiBwMvc0EfcnzU74rDH%9(qbFpX^Aa%3!V0Le zQbSmG&pTk6&#B|zeo=W$t}dWWe2RIbyV2e7BZp)tPYLMhQTqq1LbbK5e%{Spm1Dq} zydjMxce#k{O>!0;n9jhN=rGH?gMh9Lq*97kIY@(Yv3g#_xbAuyE5T6+vwGr5i?B)@xI4aWREx{c;&K zJxY&ey!P0$cu=VA9XPX>*CoZ`I4(-V=l-2#2r%cc^1`Ql?y^R4L6HJ z3ptWSc`dzmZPmn#3AIgLrKdn%;ZL|h^o>vcq_-;a1nv_#VaR)}h(A#w7pQXg*T2RP z{Aywg6~{;0O@wt$E0GHMdjxT7-7#D)U*bKJ3YWEAF0?#Q3F>)4$R3trzUrnRq6nH! zqjY7LVwv&;@NQ+cbS3QtypYm~(KuL}h*RplK{I$4+o+P-IMNy;ohDD0<`xxS=7&&M z3Nc&1BcZ<>eQr~g@a}GjIEv#?S>t90-u==XD^Zeu<(TH?_A*spT-v%*$g36{B@_sw zFUC0_BVSpJ#d6adwX_))tYuLpY(_yP#Tk2Up|vIj>Ovha5zEt<#0tP>ge?& zYId3~*6zpjyk_gG0_3RHX4gr78SI$#S}t(sv(u^*o1;^7YP-dElaNnh-sATQwk`8Z+oHOG zd-bzZ#=|R~DtP+xPlhPoRVgwvmeDyP9fm<$NMS`{OWa`JQ zc4a?`^TFvCxUsq0^i&8Kzmv*c$DsHRJvzev{YRyZHeN4hkimls0<3o39wCR|Ak53$ z*&L}8X-?M2DYc4-jvY9?6I6Cgt>VwAhbu9q<;3~JXX-UqC?f4x;*6>_y*YAi;OTRf z+B%tn7Hc^I5dO4)4`TySd6HBY7wNPL!uHUw<0{>(NFvZNf@g&MPDTkRD1X1yZvm{n zPIuB8MI}OVDz_gCbMP;y24>eRqOarkak0=znlWdk0N8Skp8w$Blf6K`M zfGc=k;tRFNx8hR6}`^zS|9c~6)%JEo>^UN@UT z@(?0$iCtUsQJ`s>M1KVm91f55C!;4w8{mXALI8Pa@m?_)G)cP5k! zTZdssfDT!?4!xGS|4Oj@uyXI9{u}Y@;p6HOB;klwz&FD-6SchDXk~-(Pv%}nkJwvx zwWb-o;5b6s#3S0{OXI9)+%oshhI!PpTHZxgo(XwpJ zWkC<9>VP#6Oy7SDxw${8#DQFi3AAXDp6ZyJ8*vG{15 zH#S}ovoEo|X%4a?{MhjHQaV^)0KYhSxE&dcR`?1_4VTZ;o0S$wmg(x?IC-Da1&EC` zQtgkCQH{QSjjNQMOM{0;Rv;e(jFA8HR|TT;(B5+*rM1;8K->AvovP2%|AehG?C!RQ zJ^nNqRIKRiJ#n&l=_CIA4>x)#-b50+Ap^ZZ<6n(s5Tty1jY;XM9ZzGv%HKI-H5k}WDVKxGivkc5+{@#y9fI3cwtFq z+MQ0TZk!R<;g_-^T~n_kPPBZ(yHk$$Euh8l2%81dIOTd?KauZLQ`tasU9~DaSA z!Mls`wd5i=Nbyw;V#;n2!knsH%x2hq30jknBvnd@|M^PmvZke}pmJ5kqJqqcsrzOJ zQd(`4OnP6{@k(I>-SvSh+c6-#Xbq^yI=p=9a8gbVZfNI<03b%4zI)X*!^-*9#*}^z zLKF5Qz(d$RkqUt?|H$aXi>7?1)WLBNJtKX^J=s|b7AoXYJEs4qg;ctgU*sV6l=D-M zr|-7vj1K9!e^TFnJq(BZ*Upo4;inLp4KSs(!4c<1bN7+x@56%xAG1B2KLoFJEW-P} z;@YuDc#1#1)(stJJ7hzUZy$p~MfX1GeHu(zT?rtWk!t;`bd(gmEmOHk<<$ASX#z5! zsmcd*OXQ=^K@&GP+jBJL?vFjg%FqD=ET*YjcdpfA(`RTS!J6baEoS%S_^zVW0;M>n zkeut^8lE!wnbe~AVMG7;wFAR=A|g!7Y&=ayPiL-vT*$f0V?`0f^b&-+p+HLEl%ZD3 za<7i3)sW;a?ji7$#pY@DtKD1Z1n!nT?k~CBqa?qroIgJ@k6eCf_IE z$jLHHY+#_GVxL^MJnKAehhI^?TdF)J^k=Dz28D}%2AIZ@mShodn#D3m^~MRK-5ne4 zUV=_U$NSBT@iqNR`I3vy%<0D0Ms@}o5vl;lW|2Td8Kg!2ecqfjZ$ZtcUCf|ezJlBu zu6E{B^V;>$<4eItAvB5>Y&;HDkuR>~GX{|;=k*C%VC0tkx?;9bw!5C;W>C2rRJ4`r zx%3O+GdUmNjr0A#GrakaT>VZJ%99&TG(sZ_)^yLDx(3XI4FYa5I7CLJc~{sGfJ&(h z!^=_CEL$AGoyVjNU(#JbEV@Z;EWa63-GQZ%@tN1U)it&{Bd93 z;Tyn`%|}QJQ0yvrZB18dRqUAK5$8FeL*A(^9%$IIjT9fCF51Q}+_15iHxgaOQ`vTM z{AaIFs?s&xu-Nc@!)NIytMdKbr{K}hOI1-v8yr8Ov5I2RR(?XueWqv6B$*}M*!9Zv zWs*e;CSpv$QMcy>lfK$N0hDek@7$OA-t|0cG0nQayKr8SPy3%Zp+(7R7{9Bfb&Qt} zRhZYj(-CG2N0<1lV!KMTZa>cCtP*kvdgjRtLWY_JEf*hOSls{KT%j!X^>N?DQ$%mW zJCOZ6tk5JjDdlx}R)$HPH&%%M!7Wx=JN0DFdWx-_Ex)oxHN5Ck`y;#b*H!6=YGl-M zABMC?2`^1_S(^ZxVfzFV=ud|YSfw%Wq)WRcel{M(9)V!z7+cnq+<-JD&zE}9@yDz9 zwo8XT*6Zy`uB{)QAw+%DcJ20et~@c6)o(V9JZRXduQ7y1eYtFjo|IBsn^vE=uF_o)?l%HZ8Mg|nC81bO_Wj)y~wpjZHUgqv{{;=4KOx9@QL z5h0RC&!={Zmsk=pmSGajUIcEJ^{-9?kUuawLc$SMMUM0-~X@iGe!n$pf{Uyfhc zFbgs%i;a)HKW|B%-rDm%obAqjXMh!3s9*JRDmE4+lBYc>KbxBio#<-F;IysQVc+8t z@864AmG!G1o<2y{46+aa+U41MW%P+#M{W982Z4w|7D!27RYYCYGhgNiVyba(2P=l+ z38%EfP$tOQaI)6JT7}t5tD7Z-&9hCLNMM-Gv{CBxjbXI%qh4zP0`#wF77Bo-V`+xXMC^81n;$3>-rz$mJR86QHf^-Q# z`?`HB58%aos4zWX%_aJ<8>gkS@7R>sK4U*=He;!S4+ss_@k#MJ{<#XDRc`lC5t-1B zM*4JxY;LER2g-_0zS{Y8vMZhUeudxRi)3=KaL}T4AP4vr?quvNzJFEhe#S=Sz*hy5 z-LbyvR^(VGYiEvnmWY#ZezpJSEI`e2mZ529ik2Ifq zG=I%=aaP-VzK#pcihVkI?cWEeB((Dz&0^ie@65CEeVqbR%L$!?NlZ>M3vzPfV{l0? z*VdQ)$^vnXut+?p%HNZEzrhkL)4Y>WH;HPTIf z0)nWFi6+M&wY$&G|N4H;U`#sFRe|N0w(Hi?%|1Jxqy~x zoxv9hlJVv6eAEsz5~gH0>NH%Pgf1-W?PqNvLwbn0W9hmsc4pt#+&(!PJRSU)TJm|y zfw6rI`~PD$%j`dz9o{nIY0NG=R;A9FYB|%9*X-GyT!#;n7tV{n;GyVms#xsN-VT|=ICRA)|#4?F86c!}c$ z@G8Js=m!rbo&D4W8k-c*Ae#5o1&z8H(e6?+AVgb>?W{C|%1CFvxi!Z0PJPPwe_{E{MMnr3MSyCI`R3n= z$1QWVdtCpHA)BKkZbvn!u1;hSjPVH5$bb0me|?li%#GWzaQU_4rmrQCE)0R%Sj@>} zse&Yar93B>D6qzab1C47vG;*{BkDwb^}bVq()ZD>E^1ubVtq5fds0=ZmW~9;BaqSw zN?>8L40co_%HVLJ!AS)TDd4S%F)Aisty7BiFYrzFclt9Y=z1=pR^A?WNhQ}qVQqRv zdQVChqY*szrnzw5^%87NJ_^JknlxG@>?2B~I_aEv=0uDJrzu|J(Gwvi3q!#_)jh=; zIZqq3c0$nqgnk)u3_=^H!PxaAE zg3fQC(6g+ehpcQI>F9lyOYSi@aYtSMV>dUWJb1@%J+Y#__}4Ls%HfAGso_m*>2N~r(;7Ghy2nnl}#%!ngnQMik+l~ zT0=K40gNiG5ert`63t4MCO}ZH{3mUoE_ulM_l3G&^TJcS395l5nu$8S_7xF(N|H+) z`vrOI$;1_jUsxbP9DU5A2abB91?^NK*oZvCO3qYAs+TvMebG)(ebE22MG-Y_E?=M% zIT-{igesR#8~D`nI$YYt$%TCl-Wo)w>%Z}|YxX}bO@rmO9IilyuDf^x_e8=SyF*Wj za=Af3LCb@>oj=g9HZdw2hR~i`9a#&TP>#%uNzGsN!(=*5q%J%$+k>>*yu}rfbi=EY zH1+e6z}exwdbuZV=P&|gG4ZnvPoaFgbuS-v2-WU`?pLQjL;MPCeQp& zh+kGoul9cO9DhC1elkvXWj`an>@#Ozrt?2Un*37{#DY7!xsK1T%peoFuN4P(1-N46 z-gN5my(_FxWXxm?RZIMwVuoC2Kuxnx zYEPX_S9`_2%_d`*Rq7oJy}WpK=R!Vv6DtbfYHrx~Fo!cCJX(AQRRg>PBml}0p=?g^ zXJU>Koz4QEqhZE=}OTa#T;6d_#%B+e>vWS&4u#;HMe4} zR@8LC4&|2M*-+w`BnPXk`p2SiZk(G2PqO*FbGGI7TR!?lnfU=J%;Tl9!siyiQ1Ge{ zHueL?Od)i}IAVAFLwLp?N~R#3l0nrdT)#2>!h~tJu>3oql~Jq|aT|WIUDqo;Ddj;} z(Z~PV-3_HmK3;iFk)`r=nba2rI|#qp0Q4sT*bMAsO8I9A>He>8C~k*{!(@K1#WBAR zy}4?!%CU0L8%n1q@1|BJqc(jkmv~bLULB9#D$k@Jt7{s@GcSJ0F{8(&a7-t+(S}wC zcBEvfm&eioo0CC2>vig$EuM&EJl92E0-%y@?Lr?E?C}RbjMA8SL)^7RePR?##lgx1 zPTai%|8XsNm;B>L;tPDCN~uB~?@wcHKl*s8YHf-r7M!_%NGFxYo?a1U=~gXV9b6dp z`#O;q`Kirg@-0kRi%=a?l&nP_AnWP|$;c;3!b5QS%-f*hnKO4@B(rIwPsNQk`lpF! z5N&nLN+XZrAgslC`s>tW$ow6Rdo^`hAJz!s<_jr$gs}Pxw$y7pQJR$*&Y>;W zan7SCxfXTPFJ`{M-vs8?yOo$y;*Uli@&>GZELT0wS~S-+ToCb@e-5FcV6mRa2a9QQ zvUO^)o5q1Zy3MEz*Bjy!%rOv%&uZtP--Z@lzev zPtZbT?bpS>8rqoBCY(Z<1}f&e=EwXRyWZCXjMpgiB8vMaYxsWf8?gmhXNYBe(^@gF zzK-p@LNF!e*9@}j%}w0+C4)O+_G{~7mw4M5^0gH>h*_-=Ztl?|2%CvTQwA}Hd9WjB`D3&1TT2U? zf%OeLp8RR4414Fp(K7a?PW|~6n(O21A%Nj&x3 zM0Mq5h4#H@T^3esPX`&hHvF^t+A;dp+?8j2JGc5gEhGUq7<)CrbZNJ6rZT*8VSc zFVRs*gHz7JaLu<#}whr7M^7>;x@3H3iZ! ze4WcMNcoJP`ieHR2yi{x_BL|y*A|_{=_<<(!LSJclc!z6eYj#zpF17N)3o$n>(Wk^ zKBXkaPs(QJmZA9{4Deg)u$-&hOsP^4w>IZrKrKUlZQ*FPm*|Wo=h1HuITt@Dj;<+j zx?YHa{`poaZ0LB~^7ia2bR|16x+0*Xp?sR->e{fWqQQ45PVEQ4Yf?^9XRGG+$wi9= z9q-jfsIv4&HHOxmFx?%Y=mt*mF})8y!YK=`6P9`$H}yh7a<AW9gq~;X_pM|iHahzCPm0EEIx@GXVyJa z2mD6kuF&k)FAt?(fRfH*YZ{>my6^A*|6Tz33TCF*D~Efy6D(8W#L%MQ@A{7F=8w*i zvnA$HMHaPQ-3CUSzZSwaJ~huwXswNQ^+)J}=zAzcR3SmzaDm(da-qxV%{>d=5^K{L zms{w;9SkJO<~*o<^~!oS(ZI)@l1UGO5a;Ur6C;zf4vg&_=ZcX&p~g3m2j))odj+&g zXI_&-%E0w}pMg(`urW99Rm@`FoI7{RlKOgGdBE7Mdr3jR>at3M_Xb&JqWhd790RkX ze!Cfo$>=^|{a-KPycAZ#Ci%K!fHPkH5%S)u{?n?B={IZhPtEp;wyjPf+hw&IPXkO! zQT5EWx6|bw2bx>srR8h;Mj+4*3U8Fgn$RZV^u3H+n8`55#}I?yWhCLY27`U!4E9>z zOzTb_gkqb@I2$!x?rdz_t|}=?l5oyUb|%=xfS_hs3yblHo;X$HlNVTwF=V@TIPA9t zne>rhNWw>&_=XAIrarO`xKbtIll?_vEvd@8ugTT=Xe0z5NED)^_Zty8?9wEG96J*e8jF>|kr%O7kOOCy2d4RCrP$zel9(y@m zOAFg580MUy>Exm2Z=boQ7DckjGuYqi6ooyDnx(%4`HZRRWF*edm_UiU7wqFyvNNd< z1ap06NAei!*+`!u32v_Ih!VTwhk~}Dx;0}1@w18BqoQBLWY%tjVZcv?d=?_c{k@f& zR+6nXz~eT9n|6dBgz!#X=#&Z#3{tF@#;lGMJOYz0yR5NEb_iFbS5+q_H~7ERXPtR* zr!cjDa0ES4wH8xZKddqnTj(>iONu($KnkWUbvn5oleozjRW=uIx&XhiifdxAEF%e# zZ>{WRuZ#e98JONL_+n*=<>uq(T!oXQQktKWY$>^g$+{x(DKfT9jryiC+41Y;#=yw& zBd=jo0%ZE~?~=QP!iWEJ&hksr!AoWq~orKh4-9GeLF$f4+DYKXtTYE1&>xsFu? z5*5=%8q4MNC45Zr4pTi*=x(%F0{U5Y87}LuM0nq|w~dNk+L3K=-CA|5#s&A@$p5!g zA}d)N)szczyy7`?Zs~ElJOB5bhRF=NC)R+}sL3uM})Ft|_$)g{4 zJ2^Z2QJXvW=bMY!^q$3P#%|>Yu*9^pG3f{5XD0pQ?m!KbOVk5qR>q7#CF;`nj6bi~ zz~{%zzHAXnNnarTT0A#(kVq?vKPM_4p6&*Y!BpIp;d(c8*8gAQThQlZ^RAokg(RHfUvF96*owDEdbikI&QE zL5^#wP7>B6P(cv4RqvYCt2yF^iY^vR>~1MK>$t>kw7KTzypaK&A-Wcdt$NyQ*unzyd zH-W?*kna+xaYpO~jPOmLEJSAgJrf?%Bagtnqh#_z!SlI1&Bqqbb}|o$g6eBta|%N5 zPPJafsb;!E5igLi1vSPJ*x%B~P2`=Qlo8-lOxFJJf;r<&J4VV8^LAkVHBBx2qacf{ z`zgnI)tc>~r!S|aM=3GPUz~sx)6aRuy=FLjuP{#^PW4^feT$S@!$hK?P)IF?lhBZB zGZ70kI*7lI_>3rIBsbn=mT=_)V6%Lii*_^Q`~I1np=iO7uXW#F>XsPMgHd6B>MG2V8_WSihSFw=4MH$z>0EuX}X)m;hkK zMv?pzO!a&!Wa;9(#wr=dk~odJyZDFw?X`zwZDi3=>Zn5Av!MrA9lcO5F>j2;s@QVi zyrdgucRW`>(6!ffDc-HFRXzrsZ^Vv45+O~)E(QqlOS`{NHI|G7HTZPLX~P+88Mvt1 zoA?!U_-kkMt6NR=GGgAmlc7vBjaWU8&ig@mXY>u^_@>vYqSyaMIBxo#PR(fxvV%e2 zNRzX?bkEt)4=)_&F|+|Ct<$SU+J-Xk=Kt;KHBELsFv~_Bu_i`WrNxPAZs0d0djr_i zfVbP53%OtKUA1EQ&&E1~FABhy^X75{kKUS5RB^Y))A1aJS^hJGMX&eNz~tBV%C3CUs_lMJ8Ku(&^==)c4s?PH z^*p_y-9jzbcxANHl-atnNtUl480Ts6` zh`UnPn0#jtk}^=5bhGO*E$7(}@kfT6PD{G4x)AQeY5AcfQU9dgjp8M|zog&k<8Fww zdN0g#f$iZzeY3mXZ7@Fgr3kVOtdc0!R+Aqr6i6Z6@J{Uz91lh$^s0`_!8qlQcwT{X ze>k6ld|SPwU$=ikw(4%5)#lg14PyKLK-{$X1f2FIbmMBp!jG=nxJSF7AZ|EKqow9! zy-a(Zw031@ozL-L%sPR^w;u~NY!&e9!EmT(7B7??4-sPn0$rh z0^gryMoOW~b{+qMVuhb$?TNd4fROr+x#fEatifq0CCzB$vt~uF*W2zI-|6SyD)&H! zipkt5L}4dqgL_wE9Vp4M`{@HaLHA}ZJ8X!njcpbF3EB>+JCD@sW=A?@dc>V(BaKB< zxkZ96fXUTV5HE)i$2$%TM;ul@)n77!*PSztqUIxOti<4Seu5Ob?LZD%v@sq!k~209 z_T{P|3N71$NnF0~_U8hu~;Qaf8w83Oir$JjY&oXGZ^n^d?^z3M02htfy5w3*!10FViD<5NMG3 zXsPu2d@#FOYIXxee`s>8l>R<3vx04w6Y(KBJEm?weeB`^B~A_9>-IhYb55#gpu5LH zb+5iaS^;HD1y3loeC9%M>j~SkzwpDpe!Z(xB zvY}Xz&aZN??)XVyt#@{~KHc}0e*rGXVa0CnZ(rF?qvYp1f+5+=7LB4(Z#2g+d}>dF zPEHySjO&!_q)V}ib$rKEab_7Rk}5${c>-B$8LW39gR_1W@U0J>PSZ;c3{CH*!S;a+ z4iMF&bDDuN<&eVPzZCZ?Iq7-w^cnu?u3Px+Gbez0?z4)*5__h!Iv!5CO89%^%-iSN&#zF=3R`-(_C~}tmpRqM6WxEI52DuNS}SCYdZ^k_s!xAf zJ907XAIr(!$w!hdjqHWL2lKckK0UfXV9sPIRkCt5sKM9*XJU@rqc=8iCTfFi$LVm_rPUn#EFn@UUWtWYsS%`DQ z9!cQPihFA^=x5{P^pWOQSv|_j1|A*q-6I=Zcwab8A}IYRv?v-=G1Iz~KR6{VfH$-9 zrCK(K%C(&adPkZaqQaly&{HpuYH=ab;T;D0j3qw?OfXd*cjs>Cs@X2#3gfbl(NGs7 zBRvK``ez#YtFm2Tsrs`li!kmLJUvF|#D>TZY8vm_OL8cF!ihu=V@>JvxM$&0KQE=&`bYE1)dqBfz4 zh`B(M@RN8=;{C1jJO=D?S|T|rIO`OI(5R1k@>v2rO%hgDIL-^ATI~KKk?Z~WV4%|X z@N6JZmf`;?D5GBJwWsRQ>`u7Ok0<6*yB(K2zslgpm(OS$QZ0}rBqT-*?9UIT!9#8* zE0fPzKZ<8F6mvq|4|do0UT@bb;^~pwp}7LdF+A%eZ692~YZys%ziX&_viB4HE2i? zJxa|?iJP`Y))M}2*@U9^uSYZVaXAp_$$?%h5tvw3U?gYV>ADyI_}Ir`gLNkH8Ea8> z*7=JVt+K|vsobxyKx2@MdM`*5z+)rRvyLwkBe}bFbH5;c1gK1V9>X*Jvq@>rHn?n& zNs)WLvmWX|egKCq2iY)74iKwo$jec-^U+_s{sNJgljC8K0`7A79@qCC*N*7tXpSfJ zH?2h{WxQm8LU#mJrp%x=mawfS<}HNtKCe{_fhjunRqM<(9BvIzVQiGwv)ZV*#?c5+ zHUP}f!Op`u*qoySPTd7_>i_kMeIIS&yRxftb5AVaWBm$N>WHJ#?N+i>FX`hoR*|1S^&juQmq>Yo6c6C1 z!Dp!cj4_UkI=0S?BMUySR?$hI)PLuL>z2 zRUOj!Q>MohoESEL0bw=NHhOBx)g^mj2=lJ9m51W*iZmWVvw=@tcva-xZSooJ`b_W`gEBg$q9IVI zSQ(anY^p{;zh0MJ*vOH?!f#mgh~nvG(}|1d?>N7$sz&Tfy=*QnaWIwKaT-X1ePO)NIT*G} zZtLZ=Uby4XCD+h=nx=Qy^P`bij6|1t8`2(~fs~b&>zquaZhkH8sX$VJhi;M&kqA>e z_SanbQSf2wLU&12-t;hjz@et|<>nO~#(P(MZvMD)sc8#(UHTy&QYtA3Z581eC0&7VJC{);*? z7XjAdGl^;H_jNBrD&nq5x-V5!&)(ED`{-e)g=fyf+lM{NwYHTYDhKXMt;NA6VXj5* zTC0f{&Gj~I^6Y)s8csjzxk3#$dL_-~0tA3Raur?9n@!@ZqPjntvvcV zp3gPwOmxbwy5*W1zGv_jeCkv{Ju2Lrk?g4`3mYg{wfL=~A?bB}@>qk=yo;>%|QzYOWw_ztc!FW6% z<=)V+Mib*BjB-8Pon}{K!B}VvzhHr{Ki>9Z4>%chNc1pAe^Tr4R2MmJ$}{L9=}SQL zeT7zZn8<5hBmb0^ewSQfF}s6tl8hjWAl*JE&ev#s)nN6OzTR~UOz|e=sgygCEb%M) zg)Vn~pYf#|>)jt0U9{cX^O0$~>qM@qR=^)T`(NKZ{9jD}v-j+G1lKEUuEJR6$(JC^ zFReE`ZX7~=ua+&pj)7D6Zl4CY?urG8i3?l>3Qc=QiP>D9@TL<31j&p?H zzc{Rn-{|ljnmnFMz$GRzX}CfLD7M{!(h6LK;D+2;_5Kk$4pf8$4DATqHCkeWIgl2Qm&goaiCX zQAKa~UV{1np}?DTAj&m$I#*s||3|=iKKq%sN#i^3pb0 zmHNc*?eK zLMon#ur@C>V@7Fr2l( zkGkqT_vB`5#QA9Rnskf&ipTs=7?SJ3RCZFmhwv4nQ@Nu3D;a$rrs$$8Jeb4^`b9ih zXD*R%u&lJ=sO|kxkg1IUOVEko98N|U*vrvw=9f(S;+!s&3B*^nbmh!8ekm5-I-PL7 zJB89?O?hkBMy4@Jk!uyzz*mFoKmNLnmHZ}mCusUV^qUv&j!Sm6&2IlLt0B-#)2rVIEd(z;Cird~11!#{3^fKxnQ3D*cn zkYDZWg8h^V-Dny3GC`wZ*k20RPq|&oDAfVE-{kqX%dqf)=fuI7>FBQ|t@KqrnpM94 zoRGHku#^N<_T{_wxA|@Ld|uzctnmClneJMCr|_IHiFtF*T64qkaU2Y7htq3@l(+K> zcaXzJ!y`vY5m658U-@o9k9by?DSAKjq^)Gd-Ic#x2zwl7$hsXrFgX{r@+MqDlzf3q^qy@2=i!{Sy0pNjZEnM&MU>bW`EcSEE%w-1CZXSL3C!B=?$-B0g_A zp>)a#X%da<)~FEIau(kDitZ9@rQO6xn6Cc_+CC!^eirM_zKNW_)qPXy?@KANTKi~; z#;6!*@)u*#L}QnD!hX+gV%-AgCjg6-rAr-lHlie-%-Hc0zd<7i{727H+UmWFQ@7oB zN4O-hh;X!OK8`s4w3Q~`s`GTiVRNa`SAOMvm(Uth+dlBW;+}!@a{hjH4K0!H=KAVg zxyJQl^LpN!{6l{msr;4!(m#X5!q;Dxb|KeNo@mD9;Rc%guN|$bPN4qjsF!~C#%qE- zt3OylMB*;ZvoYN|%Kn~pyc=3735$4;k$J;a;;a|?K z{K;D0oyUDvU6(pe`+aD#PXrC&&%co_zYL?P9QPByClnhc%bq#a`a?azcl_CSB}KLM zOapK}eKvY*+=OxtA^$~KC6`z5Y(jlXAl#V?`u3VpnbAU0VnsnE7d;EL;1qT#R6Q*; z=x#;$(yPPe#^36Y!pW$l1wmFv!axnhcDGpNqXmrqE_pCxs_C|#`+2d-@KK3y=hp4Y zJ%u#LA$gvXQ_sN`d-^^?H${@2qDHvBY!f>ld@reO;UeRe07*}q2`i}7-QX78U-{o{ zk0I?U9nQuJze<4_AFqwMJ&Dp`@|*;mTHAg;B=`Ba&01l6OXvwve+~l5jZd8;4Dilt z%oI}P!#NE&ZeoN50TOtC(#AIcr*t_&h(Hm}WvmqL>}@*$Z`cIp)=8B}sRIBXR=Ax5 z&tYgB>s$)BS##_LFaoQPdclGLywP)=XWC>Qf+k2X>wbqoPNFfMkV(;?JIMWy$2#N! z>Ys}jiB>LyYDS|=$Iz|7ZyU2serwW309C!8?!v(HfVq-a0>c;8Zy*AND59KPPF#FI zWCAEe^c_8xGt^U<(Kyw0xGbj5HG&J>Ch=*}Q_L8x&-qXk>bmrYa)#u|D< zCIJCBORN#S*IA2OO%&439PoZ*BLJOVQABlBVjqU=D`}bge)u6sPcA7^cH&&cM(p;9 zp2mD%jD(Zr=GTr;1!0o%2kvMt=iZeE#8=k53^DdtY;(|EYlw=_{SVX~V}tqC49y6c z{}4Z-PVbQSaHlDs?zl%!$NqGY2VJ?prx@{<#z(I+QDmLK*>ZmvkIg*Zw{&Gyc-q?~ zO2Gz9tWEiii8BoHbcw5Y;SD?W;t{`o6@XBB<%6&1J-I6RBDIBUaKszwnfu?;?l^3G zC`(x|1x#IpU5=aUEp08QDqx;9mb8GXd!R%s5BeTi=T+_R8nj*4{z|jp~yN9`OPm% zoAy*@cN-l(vu-4l&a@SERCFmojwi!Ru5Wu~AjbSMDLc~o>7o8gpAU9VO=Q`RvK*xN z;A`6fTkKpF{ZFN>&Gbvp*)NRrcY~X&BYz08)7p9-vSE)Vg?cMC*9*TZ9oD{wO&gND zF)`5J7SrWw>&2T7t*%Pf$=CQ>8f-?EuoxHd5I^ca&pQCFu^JuG8thvJ!f%U7NJXRcp3 zd6jQ3QHlk2#IrWa4q$C$tSZ^T0H!{03Kqf4>)&U!SI4(XpM4v#VtBiAjpAZ>rc5v2 znrP5O;?rN$S+{YHKgzFV=7miCIQqQ+b>Tz9DhRw?9=TCT!&T4v=AE^+LcV+rvNorG zbOdJ*^&CbZPVf(|hbLbzw@kZ1_icD~*rGdZ(i3NJO-+n18)XFp6!p`A-K3ZYs9DYC_$kS;m6T3{m!;OR4$_iHB zcHMu;bo+?$QVP|q`Dy=6v-{NRuCFd$>kZR8?P7G`)eT28RW2ZBy0!+3r@l<)&g~@k z-Qn5VYYwF@vTs_hD(z#{$d{g0L?w`9g62<*R|(RJz^`4Uqp4$WE(8JK)sp5KoV1E2 zeDG)vcRAQu#jmw#;&n0)O-?`S%7kPx^cjHO--0pK`;rLmcb*!{$Ed|9{u&KP^=Biqya<|W9)VbVsVkC^&lRm zKpCmA9EI;nVR|6?hBhv&2dG#p1dN$l`<5#=k!L{n#xs0e{&McXZb~hP1YLpK=he?L zo>htr{jUDAW*$KKpd6}S^WJsh=Ww#G52XBDr#0PDu-!!+X3lfB#5O2#hhXJN1kP5$ zLm!(i3Zq_GD~W+ON;~a?{wE#!KOXkoe?9E`^}ps46g-L+Fuw286e89XLWI#0A8uL8 zSUJjA1viye!I@kH-nYCyk1eV=RaV~eGk{uHpp4SJcZn!JbJBgvF?9eL+aC29P&#yl z1Pl`P0j;8hfRginx1TFc8bf7(w}3YZ-#lsoUt z9P4pV2Mib1$!Zk@@6LN%;c@_)MLq#cAHhX9>AklM`Ab-wlKG2~hBelC_lr`+h_d-b z9|-b)(x((jL{ZPZuD7IW-;JP$H+@oXVE9s&O&nyhz#M~w68&u&(qx)#V3P-BS?2Wd zQ)pbUH6Bd-;JU7HpIFwNxOoHa*!s1`eJ;&N7fq_SNjPR#!@dQjHXLJlOuxhk_O5ih zI_(cDp<%dYp>rEhw;NYo^+pz9{s_%eSFn_Y*|1k;MHG;aCFV>z`8Y^yO`dzngOB1t zTvlKeU8vzJ%9#CX5h2tsdTy1cViY=)nkPti-4A$op#v3`_<(H;rFNY@4WsmS$Fjqu zkd#QCI!0nBc_GO=J+E?upV4}{)__52o@4=oJw}=eaW*mje2jp^eMc2 zc68Flq*Ir$*5jT?suew=3P((*W`5_vk2oCs8bY`Q6hwS~zLJ$Ol0w`5xFcz_Ht{45 z+#94@+OhA3el}2=SW3U9LqA9JYNp(-YM`4ozZKs9x$i<(bhk5*sLWa@rSCkbiowI_ zZru&gk6#+u6ur9sG|ceijf0A7p0zs6-?sF~zyrz~seoU! zrfUa87qP!n7%OuG&*yC$*odypv z>lK=c_`tn06A@^}Cv&xp4_>ZdhX_E5cM7ZX{ z735Fv^{&;@hQA7Nji(bkJ;|k;*LUQ%UV&O&{;8;`8hpkn&~~C{f>JC9l{a0w=+IF$ zebPtd%c7tEQ0!d!a$MKtu;}EC(d52m9Ag;&E~P*bArs&~#~2%~|NoN^(skU4-&bGz zWGn?gLEY~z9Tw=_pi);$SqoNu^x?m9}^?{bK>sS(6L5R zN^r931XC@n^UikXv2pK3xp6rkEYFM2qYIABmpfk%HdG}xk8`-ZSUmS{Bkc9_fxT4( zt%TVCo9e*y<0F&Uei^C`#4^v9ko)I#SfWq@bo3kri)%ib0UXQg zivImS)Bol`|NTn;?eWhDZ;xVoANW210@a18oFYwDI4*NYDPTGzu-A!G?*vGINQH6d#wW(sN8NklgGNo5(bxK=j@DS$i!cHG^W4#;FH- z`<$yxJV^;Z_Qt^X2?WgeDAhj9!QwKdmL)g)mlm?v9F^_?J(k|nHCIo%b;UU=I#Bcw z?~g`G{d~O~;F$C@R0VZ39lwewpmn0Y>d?7rjuGGXnkBR39PaB*PTC?J4?p9ZRZg$} zJo~bI+T#KyB>f`%ST_IibWaDep^p0X03u7iGy-!XUwI2QTo8t0qMaxkqoMvOrR0I@ zOW=Z{Xf=Mxy-AK8M!)E?YXl}^YVu0^`q$vcds?hfZp4@XH$ce0d>pGrzwdN31oYsb zCEP;cT3GX?GH1hZM)I@NS3juGB9342h%I!r!Xw36(z`f)}6nW&i3{JB@l+NF{*m z-CAF*F&vjZ)nJmLPoi&-&)F?jZlYYTpFBQ%l&A(f8d#F9An;FRWvvtH=Snn7@~z3l zM?%a!Bia`o=K2^McjyLng-dOC1r?j_UfUkM5GG#CY>&G(`SllJXr*gV`mDMrxwO}m<^>La-SpYl?GIFl(7`f6tV zEVQFE>9O|;q^0*G-a2>&5ChzBh`PvvNt2JINY{}dhjwaA20-e^bO(-lq^0yI< zp?G6rC41CNt_>1RyvV~mN@MXE3`aNd;xIXen7qL$>(B~=#Q_Em4ZJt7 z{5Ij_pHZIKjFtcBB{n^S5)tvpV=8_Qf4S2pw?Fy@+aoeq!{e@uukyj_5bh4Y0wPcx&^T z^1t`sk5_&K``#5h8tl}Xeg3oXTjk`p<`aVTqF8jV0^&+3%%wNVDkN&{dKsQ-Cic)= z+LJ~U*Js~x+=`u8q?HVOC+{r%7njlhllI$^_EBi(yM0V3cM`+}jKxOQ03jXn0a#t^ zcQYI|fspQ;AAee1xySYEU8F>00f&~tQ__ihKG3T|A8t>aICpUY6poby@c481b69OW zwg3nXz}`^IMA_E|BjMVI`;G0hpyu8U+jKyVfG{0!iew(fUKG$R5k3IId_#U18;eL8 zn-i|4TFZb_)@|$qICrpe?DG+lfV=kjX8~#ZsPZQe-b+J&u}#L}1l~*#A=&?!4ZrTO z<-cTeqgOHo+n{OKt$H=e@2z7%WC&Leyn1JSzq{K*YchVDlVe@@o*0Uz-9}LT%kjvQ zuOB$D33$CYoAI2`Sfpf=WaO<}CG4KMer-=gx%4FJmB*k;cr z9+BGj0%|xZc>{?3Yc_KzqHy0fSy1@~_7;o&!l=m|_{2DCZYsV~uX(MfgqbEgTg6|X zZ`}xsBd6i%JbTJ#DaD1qF;}JZ7u7V0H;gvp!3if5ir4{#iAH^2h8ozs@b=#CJQRN8 zwI9@SxMpM8;L%jP)$}H@1W}664;q{wcJc71>*)`d;N@n1Da2J{=Y}6$(673AS#{;- zW1xvbzGUBPGj;nvZdJxU6_#{{(#&E3wCMXPz0z_?pnV}Byk)4Jek#EvZ0N1Y3;-4L{5! zh})$MN%FY_9mA{CRO?|i{bVLEFEarV!aCl5G5Lup)H(D(`K8Jt-SqJN(SP3b9+sa( z^Fud#C_8(&W>)`)MK2<69lm(3888sAvb?!a=&Ko)l~~_9deL0pL|DaLGRt-XzI0(q zs`r{$@^a(ND?_M7L^~;B(|dC@W%V6PTyOYFbcI_x^`DYM&yI>pqDr`1gXxb)MXr9= z*~M?rtXp$>bMk*h9l;Z%J2J6JZDeQ#mYL?nRj*(54*HSPW2Lh^y7Q^~d9v%aJ8f0f9sT7_ z_5g~WX5FAvrs5h`Q?c-!@esAB#j-^gTZYrLB-RrgDQj8|bZc7M9o zELK!Qk8-Vq(J5!y1WT6&9iF0rqeLn@_K;DAcmb_pJr684Ci~Kj<)g5>Aguk{USjyv)!jvZR#WBTs^W+9&c9YqQ{<|^{Zp5X%^{vZKMuJHKIvf1&3MJaFJfO#B%0a6 zwYkDMDVZ6Rq*#_4F!Px5_x9h$S=f^V9@I5s)1p&0)f#7Wh}lGQkk5CmD$_^1)49Y3 ztY&Hr$33i~vwW2o?GnM!W>%OAC@4MPN?4xbJ^|Q+5bl_%Ur$ZJ16`c28^Go$Zno*k zSXnyeYw_hrTfiM-P12?Su)`6kSu;iaYD9=vV6t~)=w#cfZQZESjhcMIn-D;o@wqDV z%J5JF>#+Stviux? zQ9o7Qvsu6iGK>AZOLa>}S*K@2WZdk$RZ|_g-_>`=x}89oJB+IqQeT$r0?FjhTQjXI z7udYk9Qhm#tc-bv$&c#qg3z-U+CIg3hj&&o6=$o4R#sNpNRFyAIW>n=B$C9WJ%8Z9`ioD{5!|e7d7l-X%wordCF-k*%dW?wQjKahZ z$oK4<{jL8@XA^&lhD+AA1!jFlyHh_({YAo?!abB%4A_KN>UyoEtajKXUKQ=B$igHS z|6*uaoDoY@PCfRV#oEU3wQKpe*w5Q!9lG*e@il!nr;^vb-{&2reb?RS*(sPw^uZy> zP~3m?_o=HsmQdw4DV0I4V!n^9+lypfPfxS21-CXbE}G?J2F#1U!BaaZKO|Vw!8LlNS`C+em?g2;Rx&ZFlFU}F`YX+PW)91Q(&jE#y4zB= zZ&J##&A7KF#X6+i`%`gBegYAOyP%kpo>4GR%&6)8zB3@(WsrZX%NFtY*P5$;LWAb& zZ4;jTZy7Qv;g1kyL%Cjc?%Quc?pa!Pv|!d(*?D5JrICB$b4ujEd2ROwg<6ecmJN4f zhh`?f)-Hcgs{$d|IOBjG8FhKygD!~o4P1+)Yw3Rfkel%4FfP#om4f&inm&K88#i^| znKLY~`plY%7`=DNDUe!cAh~K8VMuI5fW5LX8bC0jvsZ216s}1;b>D5Q0)dS!w(hAZ zX!>r-$)HPI2zTDpwxZeEBTeUZ7yZgvk{D71Ftm(_k)!(2>#tSPyK((MnCXf^U^ZscISx8EDcQ0gYo! zVrxmdn*WPNvFmlF>A==bXf0Np&#F)te^4i!3(uU zn^&yzAN%l^^YC}n_^7pgMY%+vJKj;aL;0q4bw(6Z-ny4j2PuhHuTE zZK{<2dW@sXIZS7+dTRrPlfQR-n;n^$zCY3#?65s5#_Ic$vQ}383~if(?%A}wTgQAq zKVIA$J@kL=m#6=3Z?@X&H0~zz{aQvZa)};$l>5Sxtl$oG7Osn*_yz_9fyxDhWcZ|j z7S2N%JJ^2!Wh4Qjl28fWm++5sYM?vTw&hSP{hAc{rf3)%bOP`qNhbfSLIgbv2*7F` ztV9_YtJ{!XA_*(n9O1?ysTa~LiN|=q2{VUNdn)nAbJK=P6bs~d3P=JwQMjlFah(f= zb9EmD(8s=fvJRg>eMy@sN%Yc8wRi;i1Q0Slugfozo>0zsZzq!`z-c(!T`Iu0)u>_8 zXsiVZ8tD(ne~Nm6myv@XTV8OK8HJqPTuFlO*HIt264s3COmm_VHWn(azX#J!R3Z$_ z*4D?GNK$RUCAU(oqZFH#99fy`?HnWyLe#5`5haH&JkA_cLQ5mGSJm^p?sWuCxpw%v zIb&BezDO~@>cYILUZXB4rElpwM_d^2zWEfqwSk)bd0jig4UE|}U(=M*TaDOW${B3S zdQuZ2~vL9Cst>8# z%XYGdCB$+G&(u~6XaXN>5xLqe)nK<7C-aCkJzqhrzAB|~iZLDoJ?c9L1*7YpIOv64 zL4Macq;_Rvz8zf{Q~Rq>w#&AH)_N}6vLzxa(o(i19d4v{U%ZA}aMBK*n8OC)x`e#czO%~&6>e&43 zw$%tCPjXbtUT$J6W2)D?^`muVkgr=r_!D7lSkG|KR-=BGcG%~E8wmHOCc@~N9=Bn{ zVa6#q-etim@NXnKYiwxv)QvNhewxR0&t=SKfZLXzZJ#!J>!Ho8>Thfr@}`;kYBx}K z7;@2@11cAoKdl?` zA4-(xrAgCh)`;Wt;q@Jsm4FMmzIHW2ys`8%jK~;Xo3c4I zZaj3timhGRQ?ncYr zlejeMdwBTQgTx_v5q$SLJ-TqW?`=$Fmn3V|Yc;TSSf_~`)ppurYz4wPd`v<#JJaxY zQ;$9^ow=Yn=E3-=G2zYnX6@SGwO-7; zPQT(l;rj>kUr7GpS)@@?{a$0fy1K9WcUNOq!^W?lR&^~#LZ9{`^%nn(ijhA*hPC0Sgv%wQ6WMh!GLz zRkazHe7YSt`$`f|4*>0v9)7V5%ePx~@N4x_FKCkT{twOWQI&Eujk8uNlPotD-0OJ@ zKl&{`7a7V91$l{q6?EEHGgeU(6#348heFb^b?bbzQGrB3>&4+|E~gRLab%R*B0p|5 z_^^TXxyG%-D?A&OT(Qu7GgfIh2JLL^rFlRV7p_AsUOdBJrNr|ApfSkx z7;7P1rH?&E2q z7p+N7ree7#u)>iL!=eOa4R77zlmc%v_QjvR>~T`ErstyR)Z8UgsHX|^$c+a(%yGTW zJa7xB%CYM&lgo2%IH!?;b!2he<>JaR`J>Ab?O7Y2q}goST_2SOc!d=+Ue&)(FAa0Y zUPOo1CJzur>uxJJAMDR~5ACCGrwL10Q@w^tp!$V$VPo!1^o?Oz=g(3Q8NzwO+LoX% ztDR%hZP-|i&gncFUphxV9Yp)6<}G*s22szc?V6`;YB@7LE0QuYQwp<=szJ-GpC5W-! ziovRFZViSlVF|T}p_Iv#Lb_$br%~R+4Uew#?tA zCW_N8oBCcX^-g67rT6+KSU{?mdg+e@5PvJu(X;7y` zE&meqQj=HLHH5Lp+^Vb{j4s1F_h(Z?s-_mkr8xQhbwxv z-qT-H;D^24840ZIUblA{eMJio@JjDdjC(u<<~o)YXU0qJX5OTUiZ1R z3uZT+rbd_y|Jk;R$2%pveYXYCD~YAaUde08)MpOKQ)l5{b5|W`5B0Tm&Uq(L7gYmxys#=VRoIbSFLGZ0$}Byf_K4zRH3Sd{@STdLodWd>rV64kb}Dh1}T*& zk)44ghIjJwr-h=f1X;s{(fM(yLoto}3j(d_4pKFz%!L}G;>Y-9l^i4il>?*6?-}J# zMp6WQiP@}Kajo~WVOy~+=A|-4>Nqb?=Us|y`GO1T%lGC8OY#PY{Mqnz&T#U6W?X<= z>(z>;iQxTsyMwV(TlBE-5d~8CBG1KaMUG(y!SINF*P2foGv}{NZ(utG6DuTfebsH6 zeYLWikI|)Q;G{{g!f6VjpZtk!$bsN?cD$E_eXV^@5^e8<p=P57Ns4PXAC9+NPNx50DR@Z&xN^SuhXz%MVG_O z-2umC#DYK^97bGjKSBExH43L(MSS@NNO*(D11~C-hLa?U@H6xn|KvzRqr{|T z(ebPGH?P{I55)o+L{ItqNt!sq79AL+uHIL7Aqo7E*I@0T2iwaDPRi1)ZKh^>a)8WM zOd2Abnt<#N@i+IL!^;%^>ErfMxzCvfs-l18W+)i^meyX>-*W@`f z@33e9)iDH`+Y@VD%xj&xBr&n>2+g8=HKP1z=h)i#6vq?d!4G-OQaHrY5x`H$Nsa-^ zlEg(C#sNwK|A!Is&BH+CC}mcgj@9I?E@22$`CM3PN;m_qzfqaMyTlW=D-hy`oWXp} zmq)P}3~i`m-*0h}&LI;TY+bCDn&(|5HACXV(Q%ueWTT+P&b}ho@}GF>`%LS%-Q1Xe z5F>sF5>LMeB%HZfX*^k(@t4ws%+a`@BpwSCSrS+5kcrPhxR%q_lE{41%3xeVVL6U0#*B*F)s*m2@dA-Slq*P9cm^&=wJ zL{{S2NH-a?q%#LagA%r1qHkpdwlsA`IUrS3uxDczp*!;-fw8vzR633T3S*#1+S=)o zz7eB#_A)D@e>kl}{dfsM+be*du6!WU5HP=|h_ofNGk)o|-yX}Wo!amqk1E*YfE_&xN*42&Q|F-v5?(2Qw zUItg6!3d>Z-7@xx-%U>U94eZ8*#o6(S}&7-0#1C!SYa(dG5TbalTVe_BF+_8bv}bL zwSD5vV_!dDfz#;qf=^nV{ zQ;$1_b7~lhLmkwIBkdMwm_`tA`&!W(2|v$CO;PlkwY@RR(j9`CAHn`O@8HnPH>xy- z5)-~tl03tub?R7_wYHrH8^bS8tz-t^FAV$g^mdecv1SsCVs*N^?&)t7`w^Dpf;J

    8kHd{@U)w?z15qJ5$rSFKaU@B0Xu zG%H=4O7e6?ZcrDHAsPX<3%($)gAj!-;+cCH zl&Q0iMpe>|QCHTlKI42Khzx-=(5ZC)^MO9e@ezv2GFT|&rGK2P#dr1Bp6VR1`0a(7 zC>Zz{9_o2vaE{*b;q;dqWvoR^Bs$avqZ9S3#W_Hyj1y8qzs`W2sxCX~SW<3L?Jvmm zd%2^|J;m^_N!QBC+>F-BN007xr4M4Iug}BCti+hV@qP1^+Kijh`*-U6*#eNn_He$H zAzzj?=>V2@9TMPcy8hq)g^e?}l*c}bkA)VYqrfxW`z4jLFBSak5T26-$>n=E{WW_4Ohn|AvE)!6++u8pQBOK$oCH1b6g`Y|HGw>}gs3M5uB5Fql&`-xlA-d^ z25WYFTGFcRls_vS5O-u4S{zqPk8Ua47A(|uSzP%L_+hlV{>|({b0`y=83m&#zl>d- z!A^Fm*7%#AkhXu`+Fq+~H&E7s+1>3#P^{8tL}*#Sxx^^PT-RMDh_Xl4;PwWVf zV(co{RrMPdA?l1%ZCb;o)?yFmZ1WEvBVcKVgD(#GEs0zK_0r+9>H$p|vL(Mce4|3p zB}hm8T+zMlk`nJp!GdWi%u6$?7Nwm8MkUW9cFYUez>J@I2aZS3UFVJ2Eak!9$$J>D z$x2l<1nli0c0cNA#=OGmg8{J5bpY~*K`RGIHL>K-e-}Luu0zkfykhO}BQa5b%*PXDRJGEX{y6Y_xzAX%Yep8r z<@u|i@66{OGsR#xd7sQu3d{d(bP&dEnX@E z;_332XQ!J%2E^|5`Ex_YW!RBavGdOWUg(C@Ig;J~>hv>&3%WhDyx*8t>ELi%-o>jl z&#!x7ZO!Z8);jviz|ros7fH*cmH6XY?yJoa`^vtFn<3fW&f^X@!K{#&>8DrQb#P0S@y4_ zNPfFB{BR64xrmNGTmY-`R58t8)GUg>cZ~}A85?XbnNbgkm(AH&)MBby{<>x^sL7h z?+q5NRlq1y4w4jYWH;T_?u%atOu?*Q3AVxNax+?SPvZ2l^%V|-{2QXLSRLP#qtk%V zT?^rNMbNz?tvIXGw{Qv!WfdHbUC^gX3;tvUYh>M|f8qI!%``bVc^G_lSr*LH@9)Ru z+$3~R<$9x;`$F1e=d@^PgVbKj8O&^ZLrbChzlMm?_`JxL@S0ACIDVlQ+bUmg^7P9D}bqa*STt}KaDBo7<58>*bYuWuN@leuk3+42ef@_5B{Y1SR2aR zmE|l5zW7Ts{?0p=yYFu%8iYTzi^)#uyPmD8I)7;{%uLr`U@brW?1TU7RYemv%%YeD z`g?-uq(y5`NeR-(3*Pg9VW&OB=eF>!kVUj0l30#iKZjqHrMATOdme-fnzV5T(ryk9 zFviwCL+DBMvp6@Pj+t;u>xA}j1dB5+uhPS|D$K{3*zSYi+JeoCX%@Pwhh|`*^Jt{_ z>BzX%puWyvX;iH7_0sa|ut|8EGRdd%n>V>@ZX`5?$<-YT^d}cKmCyG*3H0jMTevG% zWBspiO*B*W)ldpSPXIn|J3&Z0ZNFK!Ew&<(!^bY~i2Tk@ntNonN71ar42V?*dR~!O z{~8_bv9VmEJCbN3M>?hEGQF-ya~jjuiL?v62g3%`ClGmTWOApw~DD3dPZjgl9JB2gPEPO0gP?QwYQ71Xi`h16^cpJADKZb3v5t1ifQrhVO z32b}E3YMOLn&!hI&K3qDIsks{mrq1J^slGwM|+gN=&uSvXyW#sUnR}OLkaE@L$Ur; z71xg=a+vMJ^4Ud}`B3v?-n1pC_p;f6F>c<&)%53U(~EP6dP4kU?-uQ*URR{D+=?2& zHF5mW?(+=qKpdiIoR%B$sA5x8R&x?3rMRMJXghAtuS6zm%e%dU z1Wd9HZh}KOkntLg7~ZTSzi1coq6ihW%W2_1Qj(e8@sIra*Q~z#^ClfY2#+QPX_&Tl zv>h8nnx8CgC;? zgKv>wO4)7yTDopU5ATF_d>Fq)06frHuqG)qtER)YW)`OBz$91b_@${XOfS`I>Ic?ORL0Q zC4gS(%jt2Rh(cM@KQI}ZXMV7gu&jRt&|&6lzH0gK4VO~ki3VhVX;%FKbbVb!apdSu zpz))9r)y*{ds+PZ{k=m>KT$Zykdwq+W$*;cpLiXQS%oq&-%JPyw5W*s7 zyw+qhZTAWoCs?^m)yJIIo|JL$=dhFPc4k6AiVV}u0Vo6UZ=qO$S#HJOFSNbwN(jF- z!$Yieeb}bL?Fb*tVuKrJv;$@(5LE|Uew1$|fffq;icff>p1)@XJGP)vssLC2#$Z;b zU+%PD;CssHfINj70yZ{1BVC*P1zt}ayDYKCNg5K%l9xNG{}U5nu@><$kd$Mxlb#zT zE);OfEsZVa?-hvsMFRv{2D>mVXz4m6{VnxSXAHC&oU*D@q5DDDdQmqR)bX3~PCm#4 znnG^`%cMcTGY^9r=;*nexQg$y2-Ld};)oW<6MT`y$#G6P+dYmT`r)T%b5oE;7Cy+y1!f zke;W;cs$%_pZQPQr6!XL%$rrJsn!E{^P4-PBOpe9uqbq1+;f06u7X+edekd$+Fkojs*xIyqAz}Yu3N+ zpJP4F;@(38#_~&Twh7_TK`XfAAD->F8yrN_CH)QO@LX4PBmf9nNH|QA81(E1IL19U zbSPS6_3RSCwR0Vjb%D5cVAdu*>2detxvxHiSSaO>utJSM4KGH}|%eBsRIG2imh%C~~dYt>~T-3LE6RO6=5vpMlo9UNfT zJteCA!5=|he0PKtZed=~y|PdcL$JW?Ka-g1`ktxVF03Fo1hl{(sy)rvcYGNN;uRo> zI2?G`k$SrKl8F{bgx1fwkbY8jB)AS$ku8c9-5_^=S@^;V?znWca9HO$)4tqIzV(sN z#s~FatdE(~c{vgPoDa8i{}ON8)wA`lF)`$X=Wtt1vEBQVsDtInB_l5sL0o{)uSx#e zY1AD~`!2vs@j4f-#AF`kRbp;ftQ0iqI7CL13!^sO@4yxOYC)yA{*M#$lRdrAzXI4Y z>ZSHl9;dRW$A?M5woj!?6Qj@=6oB?WHG?yC32ei@@ zsFbHUupEw>nm@$giK8;(8=*eiQXz6fbrKWAmRwf8cM_C`;;fH7so@ydRaTTbM9EWZ zsq6Iff!C;o?B*$z>!;@~oV#7ZxADQ=(Lhk(jL-qs?vTB_?@*0akJH(lUu}c%5eVTH zlElMNcv!`#Y;Jg>^%7moo_Bk?OYi!`#X`-ChBgKoF8_ z>Obc(N-BS<1vwSi%~9^(GkIY$a|=3S>6U&=*~k?1;0|k1t>T>?%L&iE{7miMmdx^* z=Z76~Xa-0#E3Rz!ji}DXxp-^$Hl1P)T!+cY7&|!7uoWgzd+ARI#ERAJpU%|ES*~5o zZf(6(VWDWty>0P{Fd%;s+_H?V_c!bMA028K0AJWImw8W3k-TYsvo%V3=8Ai`G_@5N zeO#yw*iUSf0dV_G{nq0%ey3Cf^vx`H3HfV+dkFANL)}9v6dwe>Z`Qif+D=Q{Ckg;pZX-9V26~aahACkJ%0V#YIurOCb zQoiF_Deg!*EjP>ccg~M2a)}a{l>PKE&g5+MkDH0rU(DE4HIptWE?VBVzK@stRHgGC z`XQCfa6uU61q!>hL4PwbO~ZU+L0G{~*oAc!6or0fT*NRZR`Q9>>>^|K;4jF#CxAyx zQs(o(u@`y#x7DZB67ePs4Z$4E^a8Bb6t@q{52udr!e8!WjyFxpVaj6jt`r~NHWIF+ zZ?+;Nuu-%5W;E@C-_nlMFxgD6o-kj?Yvu5?M_?P*-|GbnO`*eq)8%QsA|(_QcQfcZ z<8A2b3)iP?r<2v&pOit4Ax=A07h`L}{FW{S)m&T&JnUlP~`U-pP_D*4@W6KCGGDy?wx zd&h$7h|MBKB9rcm@3=0a_4Ft<<}*3aM&f%HrJ^cDQn;IS$ybzX$SC?kiV$xa8@|nL1`K z);}|m$!mpmBuMBE4KNBQjj{1YjsMf13a+JRvEZ4^~IF!)@p z=Gg@0_3=Z|fH4z@?dWOmQiR}?Cra9`UEmb2W$A4#Tb1FI7wfd>c>^LBC1;2J3{@J` z80g_9_bEqxDK8br$aCkd5))k<&7N6HS3m4*e5QyQ)H6DGv~v}*{#My{@VQ7yFVjMm zYAEioKLUmv>r@JLK9u6^-MbfjMxNc6$~kpO(@)PUY-|TP5%cgcqD-!AQG>Crk=6oxrI~SV=p!U6{{ta~^c7*Wj=oNnts*`978#L?J9C zKe)#|vq-kIvW1@e^-QGvkh~3wD)6`V_)vWG5At*Z(uK}++o~bH%;{zQ*YbT%gr84~ zHWNCxNkh_G#^}A>qGRL>tlH*fP|x<4v9MAiIZ(=r3pHlt$O!| z03QJ>J_$Jt)$x3y4{=Y6=qpFfBB;bDeYE_%M2xe|(6rI}xp#UCHtKGz$x9A?dn0-K zuxrt(l24=fJgBOn$gWV4Yu#P!@EPe?NpDqk$5A1$jH)Rb=J z5)PbU9rzX~Nv{{e)Jw1yKX}Mvhm)O4HQ#=>aP6bwgyppz23a|4uBySezIIZ6%H_{s9?AP~KR2Ph4++bx4!+eip z&nwevE2q&ByUK-!0OX#3Zu$gjvre7&+Df8?mHH_;UXRE={iGq$py}}U7kPg}Q;_jr z_`r>=oG)g&OQel`(h<zX*mP*XYtDf2^zl3TLQ=$8OTly$% z8~Q9a62Y1%>~meC`c>E$4YgoqNR@8p6sY1IYPRP-7-l6FV=_pX9?)fo1vN3ybt?oZ z7!5k>ZG9P-n3DTxV)f6Z3wMK=s7rXb%1ltoi-CR6A37_n$+2T-rtA=k^)Seh zvBY3imX{sGd5o*9{NDQrDVZfEgtAk}72}~V?T=iSce%Hp?X&|R^zP)H8*wWad9=wh zP;Ovp)7kO6P>=CZV~~L}!|^|tjG1ggTwG^1V02xvQ*2xk!QzkuySHI0J>oyn+syq~ zElI~4)Z?{M$j*a`8-ui&P+&}-qAM#y7Y=NNNEMH%nGQVpy1Vz!DLn?h(g8a|T+V2_ z^)rvM;TnVyyHe@IItb?Ja+>{UiF>l}6nJY+5V^~e73smtuEp!Sa%(rNmRkXYBmBGd z$vO5h_)-@s`CrsL-;PiAuh!NllHw&%`tI|9sQDFo26$c8u7+9q%i%F0k<@L;aOh_J zpMg_B<&vY&#*z7v+Ur~yC_8CmeT19n$3qn*Ui{6?cEEUU9{dExVbL>Z?5+peBbPOe zalxl9Y%IU@HshZ|NSqO-gx^RZ<5qhzJ#W<<%6vRj&RpHL({?|Bd;rmus1-?punQ~A zO&#%heb?%1o4$Ei++Bi94m3u0VGQLh=SaDH?zG46Eh5ztH3KKRQ0cUaPe60&NL{bIX1r+KbwST6@JjF{!%9B%u zl8ex9lB+AE{AP4PsqCoT%cMA~!nd5}33j#0Do1`shIigfXR*S;m_n>3sewXSDhJ<@sE)s_>A?7}M8Eg74j z-!bk~%AJ(ondXD6wsR{S^KQ?5jG@i<1@2SZvwUU_A|Qc2)E80ltOB^-9RU{%t)A`< zL2D-5ry^**ouokmH&e_#7T<>E3zcQbCzk|-+$kmp9F!Y()St?kCC+~R-1OO59tIb?gQ&QEcLleEg^z*MGOXa2_L4V}{?|X2GON75JY*X_ z_wE)1om=;hOl$Hq>frV|cu_IXVDD>c)D6|Q&Y z@>m34-JlZ-PNA)9AgYKaztnky$Xh(Yj4p%!@Z?$Bj=ag1kPkM&Hmck*OSr1k0IUgn zR6*xT&TDtTebGh)bTJeyzCJaFpB>JJ{-tlaYt4w{{xgTfg$cDxKlj{rO&Z5K=>SN`ZF80 ztOaerM&gZtLtV3qjtJM4u#d{!^95P!$`9>rAEu!QA196&obLqo6F$Wr&n;RHg?^x@ z8QFF)UiEEz5J;{qDF5aY8Do6ko}7~rtu1v;pV)Hlbe$Uvp71|KwM;5K!??e|emqAV zMQ}3LcD*8Ae%`=PQxEy(MX99?vf>fR0T4LPRgt?9G9J&rdFMFpsp8(^2W4VkBVFSr zONW7!Qy%=7cH>}sZ8od8oqOC-#5;p8&C?1|)ioSmhmj!;<|d*!azw|6w7p$(evU1s zfhUf-Sos}KY_$0+uAZT1&U7@9|1QJbilTLIH5_FqRqZydpe;YmlMB;(rNE@0uM=35 zoF}u$cYb_T2YP=Vb@@3k7ZNU2KVB;Z`Y?rJ1P(;UjHyd>jQnJ@RW~rzn z>?kvxbDH$o6IiXPD|@^uNXScpFpWbhL|a>VUTr=Q*;!w9e2M%tAZph6S1OhT`zY#} zO8W5^7AtT4`Adh>#rcLmZQ2{wC(O}cSZu3cS@51v?VpY`j{zN}pQ`}ZlRI9ALmn>= zWY@T0DKnBO%;90`v-C|mbWcr!ZntouRE8(O;9w@JJ6z*TDZ6gl3)&|fgG>Eq9a+~)r%5xKF3t(J2lTGA7^nBz{Va_`S*C>ui|Y{v+O&tms>{>XXmtK zIN)O^P~Z@BR{4j)XlPpcL;by5U{Yez#@H3;fNYtBPWO^W)Ys7x2X()^9I>q7rhQR;8q^q6k3 z^t4`Gf0Y#cWdW}8ys-NSRvF=^tT31j-2<)wSHF1?MkisfQ{AxzhPz#FW29Uw)8g_c ztI2Oc%|?xp@H47`IFr?CVN^qccY@YW+vU6&hO=IO26}oS`MEolB0uTto8#I`e5#wB z=s8vPK|C4cF9OXr2714M7W*iL`YxYtsOHO1NLsE?WfKhaP{5+*Mi)2fAtLBH&jY|G zG99WCo?21pzTA8@6@^W)X*XggokUEiWVNg9ykKk|2I^uL=-^7*KdiJg&mT<83G=pJ z#+U?yFlXG{sq{1dh3E4Zi?FKYt3hmW^O)(ADY{;mV0$g?=k%$|Al+xe{zzD$&E&bl z!)Y6lR61w*?2ToygybSmfZqy=F<$8?*4*Zb@#t&jBjO{)vilH%c5j}@ZmN65U56l? zed+PcqN%SN$~x&Z*hXgC(Khy!Q2xlQchE$EW!i^F{cwJfUYk#5<)(ME>&CLNR-oMz zMSkqnYPDqc?KsRmD_Os1F}Tl4jpQ`(V=S*C(a@~=9gsJDO*AJ>ai!e>9rn=xKS030 zEkgNZyA+V<80mC8ei;3wuNam(D447Z!2WT(yE}3NGo-I5;KIk*k;t=FsP0#uAHge0 z4qWL0OU;Tdb$w9oUTM*!soE=lX?mddml-BdHxiZ63!lKNRKdItxRHFPMOtW6G1GIr&4U6`|-e-y9ZiU>AquBDL^f^fFu zv)?R*5eKi=V8X|w?a3RK*eqjb^}ztJg;z(z;laHvQmuM~{-I!2+2ugu^irK%Fo)Dc zj_%pA`n-mK?*~!!R%0uG*D=3~?ZBJ2S-(F+G$1#Mc{f2=J$ zz3yG?Ve@&M#yIyp=SeEy@NE$3!!79yDC(W)L9i$FV^#I0>VkUm`oah^^nk@;61bqwpw>%9bD+Gcb3JANSV6Ar(LKiQURv;_9z(CG~fV5`8VHe_kFwwD5kUJQi<$ zuhaoFaL=Y&PT_sDwnU?6ml|NFwWdtiG-0_i^YYFO;!1Y(?z(Z-dVWxOJN*AmT*Ur| z0+V&iq~ZqO+Ep-rp|K@u)PhceMX~pVZMi-L3R`n!;t6U|8hN0DDyDj|Y?EgP9Zh(! zL^gd2Jty_P>YD4#`}ApDBQHd14p|3HbfjLS6;V3$n|$4c@>ncSgcr;S7p6Y1%TGFK zcg){{smAuA)RF_cS038ewS6feM$#%g0uq^&P6?jkIc9RN)VQxdhEuDI!fR|dGK>H+ zm0VeLzngDe1MMXUNKl%&r${v+zOP>0qjC6C_!>3qA_CMHtjs6zm-5Qbxh}p< zU~C3C4JaK<_;_q&Vm5jGkAh2Zb30GHNpU?Lv0;|=@0_b|b;{x>(Kkc{x^(md#`f2% z|Lw_IUp+VCh5U35@!m%S=EnAJIidhCD^c%1`d=99_Y4YvXSCSxdYjMf$h0HUQ2c-x zDN$FPz^ETE7$4%{dG!7a%|g4os|og7Xirq`$47GW4i0K&amrIUrB}LLNxAVQ$7Hji z2y_PrpTYs%X9=Y)nO%aiGT|W*(vS!}4%`$!ruAr|!*GJ8Hr??!a9hZ**oeqcsVMF$@a1iZaPE~qxT)Nhp(}d8=ynYP`)wp=78lmEBb0=H z1GI<|M8F##N{srWb_WhMe@ZT9zUMgV1)yd{_DFtTt6SO&&`?dqx7gTLXY1^ Qi3 zJXUTI#pK6&$~e?`ndfH65kI@RS#yqcLoS=6d%upYDw4o;JyO1 zwXJc6ASCZbWT=^IqgMmEwUcyqcF+WYaoYy!;SX&f5?2n-gZxXN0bQGgy56%6lqDJh zeoPpQ4d@%NM87(hjMzzXm+8xqI@u|XhMmotRS}b8=9(MuMrWNlQu3y8)&mHzFRH7T z1GF%fjkz8a_Z)*{D9e+$fLG;CG1yq((ze= zuep$FBU&A@a=gb}-?H}V@+YDC8XUt`Gr~2g>qdU)ZYl`P(=jKw{EC}*_MW$vgTK<# z=BhueGYW_L3THP(Qdjs&Y#*O|^o2qe5rR_6^zYNQ#GF_CgsY%djuxO&Nl|92p zojA}(^cUkQBZoJ#E#1PlSkGWhtcfFllOU4Dbe&7Qz9d7o<&*iaDu#;3YHoZO(8BD+ z93!+9mWj5TyL!v?AQ?ONY=GETTUk3gMgdB;LSV^!n)y0ZNL%WzI$Ys1{8~4Wliicbev^WoRzQ_`#AorGVa|M ziB&4&KR;##H}H`)AD~qLS_l%GK}u|8DAurKnW62K1_fA-{>qc+&m=*~KL$7)s+sHp>1}=_?%uF0!Q1+8nT z`EbRlxB60+KMgOt;5uP;QPV6V%#yXLf5la&(AHJz-Ih&K><~MA-Y6rT-{+lx9d!`r z-Tt-SbJ!4a{0^eH{pX`DB?s#*B}YK@aPp4a8IBzo)3y%^h`PG-E@`rQrkB|ltd7@TS{ zTlWt;?LY(Z#OTfAwN@Clmjm^oPBbTR*B6#cMS*Q zlp5*|5~IzJDR*$Fgp-8lNSa;sU?vS;k2I z9ir*~JF&5Pj`0s$KDfH+dNDX&`DKt9PYX zvu3mnVYM68m%|$Nk?HBzx6M9KMVTTzBSw zXK*03%y~2+9mTyIbTU5Oao?gnQR4fLtNLiM{Uc-aHWU&+VgPY#ZbIiprK4b9lAKyW z{({3T<&tyhdUH?B@JE51c~1`55vR8*ivOpnbIG++YYy}OOv>(90ZXqA2QC}l8!$5qo<({5RJMGi{uGD8Xt9*MF`89qnxX6Y1qeLm$Cb-nR7h#1 zgd@%KmQCLPxBj94R%65SOu&sVs1+Z!XG=rxkJ^a(`Kp##Iwo9p=K6gG3!-C>;TwZV zM^~4AfFA1cC^zsycXndWGIPkQd4Ro4A@V}QGyZtSrix?F!VP)x|0Twv|A=wx#=xgC zv1h4dG0tONF;{}In3M^FK#cH1uFS}1L%@H_zdFt0QRpYGZ>W`qu#5&?mjsQ@TmL*X zd7!TE!U?ds#pOWP^N^K)=6Qx6i)-*L-oGLPwXSG_h>$(Mfad=ZPg(PKIdSPH(Ca0{ zXQK3A?g_oER$YR>Xq8-;1o3O;;RJ*Y;P!Sz)=rUXwKvuvSCf5q=TAC})w+(j5#f^u2AC{2J?(DO__>aQ(ydZBY59K4$C=Q0E z^B+pxAd8g8!j#BmN@j@zp>)^KL*ds2_VW2Mj=r{av`3~S(3>DOh!3{k_cz?gV)GoF znv*7xeeYTSB+tZtJVPz6QC}zy8U*=QJ&%oWuc-aFvTQ2$5+R~!S506R@J_RfVm>3u zy)1x(rRuat2-aUBOK?61WwX%_`RG9>?c$$kAq8jcPShW&Ta~Ud6ie(DCv-`3n3|FnA$bA@%d>M4`|I z#YUn2ujlOq50M?FvVn5c{D{)fh8(9|esAznC{<+wil>AtE zmaO#*3C;HX0$9n9_#?c*`0DWez}3}O%O}nm!^1rJZ~@fL%IJej(x@NbLYNz?TE+tq zsevTSMP22C{)N_1edn37e89gleGxw=UD`eatoKzEcqw-|203NBw=2tj!6U{`i`$vY zg-RwmyXK2+PY2r|t{!ybb+7MwI2&k)$=Y{?!0JIZZS75Gd!>?kV?2v6~NZ5!_MpPEQm8JW^8S{mi%L?gB3MkxR6gIi{n*V(Z-#SYLov2Hm6Sv!*c$AZ~T85;!LmD_G7hNm9dvSj6H7G zZD%8KBBgG3gbyRmcVqYc(W=bslD5h_NqLe*nMF3iu^%w;%*-;&k=AZ~NEBD1<%Bh# z^Wi#I)?wTB4fxAK0p2KyYED<0`WLC`pWKujOcYPTQh6#z^a|;=i0JsUB|rEG@~3;% zw)y{*ajq}LB)%V?r&iX!YEedq}0y6J+-TB zPPU!L@oZcz7hmz3({(mWzq-QDqTc-C6-W9EllB9=9|O%{^A99w`4w}ZPae0Pfj+Yj z!o5UCSae;*I`=``6Q>)B*iBq{SAwxva!e_&5a9*i{BtVbYYhO@wxJGtVy7 zh)j+79QM<@0^XCJ#pJYr#G^~pVov1N(Mt4B}P2>=&|r{}Jm`&FB& zvi|t&4C7yrsbu_Ca^B!PH=?c$a@oM?C;2RARNM`!@7vjvUMeHq4H`|3xxe`6F z4}?vyuL5(sx%Gs-@HeCIeO_GYVHRBWiP`vx1(&(1OE1-a7%f@ z_;H9zfJHL!>wcxXFeU6xwE5RP9+9y|Qt0O`Jt7@x2^}+VdR{l0LGa@6kEfa$1k#-_ z&ukF5SD%}zW14{ql^KFlu>3KAait3287Z47oie2lTC0LjfcWW!UA^T@uYWjfvF9Fa zs|k@0X3hRV{^!wf&UuD|WSqZeiJsX`y!yn2bwblx$?y$s)@ zYfJqXMVGnxL!@LvUQ(Pw_4%+rVh$Ne>u;N6yW7gor5zc6>&p&3C;sPjC~=g!fq+d# z9~~nZ#VDV_GOZW%HB64UxOV7ZmpEM!fL8_VjO{_Y@-H}53n_pXZe{sx&{>`_+tZr_ zEM3XypgX-)ZGUj;;|5h8yJmqK`N-k;RpKN0W@P)}*Wp-~82s2$bnO@HGjoSz-=c4# zi$UJQ%n}tnC5vE`{WVZ3<)*Y;*NWvrN(MQS+jLuy9t6p!DBpSOu3 zXClB^T9gGr=PnJ%@*!#qbq`F)=DxO2{w+MKv5DiA6VcdMvcWhsQyv`eEnA^kMv7$?r&R5=f@!7-VoZpXK->s}$ zxS@daDU0wp;9A~P=*Xe;fmaXP<6S%p3T>;oMLa(^9FdbsR;z50r|(~TtI7ua71TYR z&$a?@<`kqCMXM{jNqo+e13LRvq_&AGq31eXpBy8w{+yPUs7K5+idU(Z>&~E=qT>L} zzZp(26kq9)o%+79W-o710lihmn7J_Jyepkn0lBumFlE=jeKQgxFE<&4C+JHLBg}Wx zBAju<2)3a#r7<-&AGgr;7$1s+K)XDyqFh5bZEm40xqp3pK=-}m1C8DpQRZ28nE)@5 zlB6TQfgXY5;m#!!3SBY}W_9v)SEdr}mPlPRCNM9yv<-TT_*=L@VtZovbV23>3)zjy!WjIqO`bEC)?yHl6aoj1MsZ5x771` zH!UTb`=~d7L_#}>otPTvn0~efEaU1ufs{GW3Gu_QrT6q4YqfZeNQ_u7;fR2M2<%=* zzT*m%^(O4x&v#qq2hO+a|ELI+04G&OGYQ%*VUe^s$JTwC1MTH-Eg9ULygC*!u$`^F z(exL*ET`7k&^)y6MOBg9-YgT)g<_`?TpdYmGRv?8!W*eSe4H#oJ z41<1l$O_n3)C*~x$(ION(VizV1SW4^l%=lIRla4l=hckZ0P@=HHu){1R0=%mzvVgk z5g@P#kE;C2Wg#cmvYR?ZaN2ww-u}~%E430)rC4rm5>Ufc>{y ztg+iWQA~-k;z!qghfCMI%>OrcG5+7&#XB0itCmPQK_1G=wFVkMcQvsEU(JTIfFId^0mjX}G3Gu3!hV83> zz2W3_R6M0iBE*(88uu^LAt9gD*Eu^|3W zhm=GS^xU#O!CtV6m^U^n?foGdz6csUQbRho)02L*>Iam!bC7PwiesE4>_jMnyWEY6 zugJ+kRf0gwojehRyNFRnHr^FY*y)HP*l_vu;k-!`pj;5Z0w58YH?bS>UThS7;nXZ(0B)1 z2FkgrhDA-7BHHgfSl}^lI8o-o_Yn554RX}gD>B#0%$iY^)P+s9-T#w0L3&mI%|zXO zk*91j*oi>T74|c7`7P0>0XmQep)e>q7?xb8wl|m67q|Y3Yy7_K8xlGAbu3v1a5};xBel;n*`wx-6h0@_f%z1 zt5z(X1ta*u@vQ#xcVlH!38|tr-kE6c7qs? z=!*Z_x`vKAI~EHGKgX4 zVpoX4Rg(Au7~P~pUl1s^*8-i>$ZnuGDLqj`^=g`Hf?dGguNooXlh6HcR*!`!5=ZZk z->t;I%oM?n9rijXOV5o5d_QSH)NsEzoMtX@>l7xoDFwPu~up_O88qOmo({nqSIfFVXC8yjrYxj6DynBQ7hUzb8!;CDzRjaZ`8_Me#)aWR3b;5D3su*i|`4LNVaks&@KZsFq<`N z+yk34Pd9S~dUI)pEDi><`cb}f2P!=c^ZKprlhEQNHK(<7FLdYL8>wZyHlVg>f0Yaf zjfdF3?ZLzk{cv~o{yEuf8))5f?H%{Zj1{$u7h2J`GaI&*o}{(#w8YgeW|81{oaY>= zx9}^gzhu_Khd+=*eQn-Oq2kx;+1DFF54f5f<1`?4)JuG%j)>wec7!nFHKHjpfM1~# z5ERf3XjlnTp<3%oq_@_*`W|$7-lVO^&p!l6W0%GXZs3LEuUg2rY}&B?z2DlqnpKHi zNhL8-y!nnX*gc2;Eoa}X{-1Jo|9ulYJ%}rbg)mnXCXRJ)j?@}!``-U~dTzL-r6uLx zQ`#GxuqvyDN~+dxwZwrfdTG`;VI8gl4H#E})}(=EnwYTZsPHBy@8GDVRQyXUG7;^y zI`RTOJZNYrK5Kxqb3g6!&Cv?7iY)#=Y`uq9Q*HMxN)Zq#@f8gsB`N|UB@_wL5)=iI zW~ECgp(vpjDWORdLz9klP(hIvnzR7YI{^_w?*s@)AV3lV+}GdtoqNtLWBvnsuQB%A z&suBF`2;KYIdA`5qDHS}T2dY>reh)bl5?JvXHxkia=hQZs$_8wdiro-K^J2>% za?GDD@-|;qme}Ko0oqE{zwR=$W^z&gz-?^~<4f8j z0#N%lp$OYzVv7ZlaN3>MQwC%}=F{G%JK5Ghq+tTc-C99(svr#Br1W0N@59$f1HG=zFfl~2+&PyWfUT)U7T9}XWY}u=+H1je|eRo#=}5V<;XWAl(;?6`OCQ>GSOf# zlIB?U@?bxoMP)*L2_#pk0Qbq|Wkc|YSxl(&!?tE5VsOa&gWFET`)`&8%q{!hshk?c z&H5}zOwmUNuU`ms_#0NSByksmmf8zHf?_Qgtc15H0|J_MCm8Xq*_hM#Zho-hq5Q z#OiLRNuc$>!;{$tu} zeh|4D{%=zZvt^X^&;0cRsy{boZ!(|tml3lht2QdtJzmK!O=P}?&NLg14K{+f)tOln zt0F7R+b+7s<=93t_ivV8{LYc_vtuXHbTHw;6_ee{OW(UfIDG#ZXi$`zJNnIe&zK?Hqa3u0fLSF=rDgpoXzH}97tWY2a zDG%Dz!OA3JTy?u9mF$Anko{1#uJ|0i-T5tcc$g0E^<fqJ#r|*1W zGmqm|HE<%6XeGPzVehnb(_Kg?9^ujaD2sz1%7y$@5xYE}iGJD&iz7E`4%#h*zl zYdV-@T$V~IgdLQ$+GtF(dYX-CT6+5tA97xENGXm_1aDH$Wu3XJ4BM zx)MiNoJt#JEpZqOsLlhKbQs-Ea0|StQTQCzxJ+;aagiwdY*um*P6qR!8Inb})D`;mp+6L1aHh{!)m_yH5q z@b)U};BgAHGjm#x!iEK6&jWh8BO$XBT}BqP%4)9L^r;&K9)a$2>KQWvWL!z&T(>1B}O_ zUX+24hGR0V!^~|VW1SP$l^#=~a%u})m=QoghSusG)kfE!vB(VgarABK^f(pw<ZB!C3Z>b385#k|RKx|@n!^nT+#Q(v1d5hbc$K^B5;g@uF zxV|&_|GaQPb;SsRe7bL);l0Bb&5_ZJ?YmJg? z{ESsrE#?oafe~5;Er{3or`feOnH5`|%hUH6CPwWpHtmRuY0wOFRAZlR34HFweMm|K zD{tW26ke|0=N&pgvb?FRus;vR`a-G$+afRu3E8i3R$U03cMYH4dh+p(2)f_EXV2G< zOea#KB{y*AK{Cw)YNw_VTXcWm%Feg4Q@YvZA@1^nzn6lyYQRr|vzili0E%3CPyF-}f|fXqS06KnD7-A}WBLo;5|-Z)^Wtwxq6XSGZ!$8AN@($9 z?*avyd~M|pszc;Be1Jc7CxP9%2f)-{j`kJP)&)69m^|j?`*1jiP9PN5E@5_69?x{1 z!b%Q~xR9QpIDFp6ZRS%k_zJ_knI3k1y29Isk>#``Nfr6xsb({TRY0nSIax}kGHVa0 z$0&+U4*5Y|0yD4;TZu7T3RnX^l)BC2NmKeZ3Y}tt(Je#KIn?a$a_amJBdmI!yZ*OF z3JdO>*Maoh*ve;b#IRB33L@dvk5 zhi{UrTD!M9(Gc>mSd0EDb~jIje^l>jWVZ0!vPT6tiXT%DUgNkW@N#^$NLg;cv-^d= z+>HY4V(SlYm*KcfA?%I>Lc!h3uWV%0C)l;}HU=HRDH6Te3H7XFe=Y zcEJiKns6%X{lm$Euq0GgBI_tw5eNP<(Q@Zgs9OS{9r_Wum4;f*2c2LJJd?&QIc_8^ zlr%eH?`w&&@&CxF|82m9I8ra)BgtioI(yNkR3Sp|z| z!9$MQNMID>v3WsUb|5{5)pUJ2f}0;_A{{aD6#4!cZG8%*ufO}M?}XydZ*l}lPO58O zvTBRUnrFboM+Hzuu1DC&+B2V%I*6b6|b0m^b$o-RER%)Vf9V{5{o^Y9M) zUm45)hJGYVtV&KBmn)O5(Ocf6$Y=Acl;3?L*;{HJVPMr9M~Z*cU#gAqBzTEko~MvI%3rOER7^#nWf z1+1*JqK=mQU}Me70l4*#0r@J29sJvsr6cn1&xrI(S(# z_ea=HFHBCh6;8J1$(if~d#PzWXG3RCS;X!!flw}jTG2k)<@hy;yQH>S3-z%%g&?0m zSI;dttB*xUvp@ENuxBeTOo>ycYy^937qIHFby!wV5&K_4%J@r2{_O>#<~(J{1UXW2 zQaauAvH~C5H)L>{45JLc4zP&3_5obIiz&N9%;ksKjhf4B%M7sD)!MqfzgwwrjTOaN zpCyKSqO}IFC@L3;)=iJ$kKiw2b=_!~#RhEyX|M~<)>WdzC>loU| z@qa2@!U`tg1i9)WK>#1dRM$A-v#Y%mX+kFJQ%uMwQgmT z=ig9+)V3onnp?XZ-uKQ2spJ(3>Fs|;jO4_6UduSK(7f7k3jY>6P5dqetE;)@kT}@K z(kY=*U%RlB{jL`36rEq#uZC+fP=XE$qFI#p(!J@{-{X_PI<)xlOl^b1CF?Bi!Jb!l zTbmNit$VJs(K_=^X!YO?utic7Ky_do@_t(0$@BH*2~xPPj|b~okl1l<%fB_UX-36A z*p9hI8`@Z9)sRrpYEO@AR`l#p45X1V!uc>i#xaWg71cyjmEF+{uJ>jqNR7u$-kyd- zImiJ7?WgYDi-N1hy+NnN9+ptjxd<`b{R z`&xkqqGIU_mVDdKxDVGgSjXMz%K0Fdts7n|o?N4tmp?8A)$fR3S!|(kW^RqE%s+;$ zM@iGEsm*j-I-Hfsu#uT_%X-VA^`h#wdeef*>5_UiqR-L!z**$G3548}O_#eDb{GYM zLGyiQTx9#q%+UOt;Yfpc2%4d1Xgc5xf^<7YE`U7v;tq_*u6mxX32z`y|87Q@U_ z)HsG6jN8#{R|t&L4j;mssutwuwdXE?9N=|<4kT@KbKt>nem+=axQ$s4zbG*|tIWJ3 zrK?hF96()3&4J$l8adFe4_kjb-MVi;bJ&bgK;ip-$}35hkHlp;LEmXNnyS`{TP;-7P|U~2&CL#rU~6fa*o7+CmubuPkdr$(8FtidK8QxU#>j$BuN>81dkor?j#mRUi924irZAuS7TtDp#KaMD$Oh?5S%1F zA7MTAystPb=$HYnma>O=^$cke-bvVQ?4DYY zgX}>7K3vD#L~-mX?t%(k8vTzaRevNEzMdZ~jgdP>T|>0>o<5XWP0F)+*6{@7HpsSJ z>vJa#Laj7Fbr(tIjAk_npqo(Z4|67Sb%qF2C~JB6DywLI3RbtPf}k$ z9r`ft;~FC^gm9O}x43~T(>3gt>g(z8s>>oOLO$ArvFzqwd&~h;oz9Gs*L^9FTTTyg z!l&_}#L2~LiR|!0Wj|AtYnib9Qmcl95Wy=an%6gTkFgGczOwpzJC$(6ZiA`sa*!y{ zh8!wX6oC2gnN0*VM5S)6&gebdb`q`a-T*gkbXd6W4}UymiNR+c?`Rty_gwwTi;k6> z>B);(esEnbYif4G*36xJ>OA9*KU0QM?81_4mttXAqfyz2H>cF2xl<_W->;m~rpa%D zZ;0v4aISyV{Z`Zc{WG-o&ilHSCafE2SOdW2U&ZJcdMC$UKHa597bxc|pWfUUAZdqU z>Grm`X$62@>|4}o#-8_d$y21EoKN;nT`5zPyvU=F6GUZtOk`>a!MeYqvT+KiOh6rN z)dHyPS(v&Mq5h-O>Ig#LcCVSAvDmQY1#f0mELU@D(ZxHgs%3AM%zo?~pcK ze@=4K%I$k`y9cc{AD1Rd=)T9g)KnNs`+bz^bHMMY%c^XlWjHFL0`R> ztZ9khorKqa=q0owvseEa(Kk7VR9~YA?d3)p$hj$8+b+&+WG{)r0xj)|+^Z`ZCc(rRB2EUdA8Dl%(X~?gtay z_$NwglJ9A62d%%bJYmT0-@pZKRgZVtRdpEoy%R=P24ABs_T+yq=y3MJlqpnIYd;Ev zV4{9Z0iOh@-BVmah?Yg3lRM<8MOcaL&P(`HT6o^EK9^z`&5#N<^0pNAM5yk|1CvJ0 z%udD)-7Q`thhM6l62SYzi}lJYIJeRqKfmN;ZBlCvXd=bA)E>G)jsYC+49`VK@bhPs z;)lPm_!nmQ2V)6n*m3m(Vd}5cNLQEM!hy`5Ya$IEyLoMfJ>4}iyiTN);YkD9Lys(v z6;F-SIQ)Y;e}0-A=pCNe7D(x-jwwNtYyo?$NPFO^)^X=-Q)3|M%dZJvPMh z4@3^FL!feQA21b}k<#Eqr@odk!O0DFYr61XC6<*z2bg`9#BD+Qax^vFk(c>KxAzOc z4VB3ix7OrsA>RmBV@2r0zYO?HDL-=NU)^n~J2E=7Ih46g=+^QXj*~>B?AwC6w&CtH*xyQImg7>RZZ86K0?lg$w9b_(&LjU|u1fjuxb}b>`HD z<6nSm5;?bZ95Q%RDV0rS_-?e;(h9MLA#W=ZMqiMH(XInvfj>4&8(85r>XD6uqzb&4 zCM8tWhn6tR`ngVB1;(`k{Zsr1o6s%UIq^b;L)-HP51txs&>ysV15Uwl1>6?EKZRA1 zt9{Mo*l^=IM%GjJlA!x*T?l7-exFdshXPXSkQJj_>mc1y^$Hq^7){8M>K|rr-c~@| z6P|flFi_UZn@_D(6~M)^90!`D_5+EG|M)23=1H|NQe)d77TW4>(`+ zHxTNq;!i|&EAm@w_|ACX`&4&&n|AdxK-mkeptTDY_hlqSmht2{#7=;FMynA1p?2~F6*U-q%(UqgnL1_p~w*#>d6 z9%(l(Tk%+)^{MeGhGhfa=@8M7)BHZR<=@ihV+@>718c=+83gAmTfW8)kr;nVqIR2d z9_?vfiMQY5-r-^1=w={9)m2ViFU`5MzY?zOEyDW|_^#tqnNiP~cT4M=&{IW*0GRzv z&10fiu0$vFgcKL2C%Lu=DD#*am$WuTqK865Lvg=TgDIHQNCtXiOt-h^;c0TaVh*Xe zrl$)=UX$=b{L)f&?D|6fkm=Y5bI?XgeXjKLn*1g#X6i8lggLETn5E$~1a@+;IGn)| zt^Rj*c1N-Du5ZKRpqk2?oDOC;MFGw6Px^nPtiNad=%EkMYFurEubINDdel=c2rhRj z@eMx(>IHm4qEl#|&mzBIo#Hn&z*`fK;QPc+uo?0j-7?C9A!dMCd2%F=+U zCS#5UDcj+Y&tnadN#r$+nt3s0u~0XHIX-Z)Y>`C|YW>E$f<3H^E6Ax^&gENPVL%kh zv$q+woj0BTHuXGV;lxU|Z8e8R~C^2qElx8QAp$fkjqhcrAlPYq;p+bwQyFL83D zC!(MR(xoc7X#j$ZMhx_nM04YB@xQ+-H&L|vGu~eTv||gPcpkhc`cbaV1j-%1Y^5?z zS+He{%mUP=#tygY&~f;_%d3~_f0Rim+7CZCovC4WbhhGAPULffdd$_jIWK)gj=y{} zGn~4;y$-y9{Ij$>TA*l1QAKxPcBNK43jqdb5=8mJL|%17e`Dy@c19@q%Et z$Ex`fgWY9nDvA4lJp;0Zv*z{!zRdN$;gSx7HY~!LZTteZRD|>)CYkRDx$e+k&r1bG zY#&;WO&2zSuk=V-P+!o+-W&ojDyeAuDxKcA#O~(*zd`o{_+CAy{kGV|V?nyi&?(K2 zkuSUEfFZOxf?Z*ot!5g6mVXTsCD!MRTuLFLng>vVkWgxZf=bxe3B7dT?_G8~SF$XvGhE9>qaB+f zr3l>eZLSuU}^#y?IosZ?8V)b^u${9RnWVUPz8xM2F zAVYeY6f6-RdJU#b2np3sxbJiae6_8(qR!AM>1TT1Zy}slGss1PWkhnsq(xplaV*Rr zFt<&X;ZrP9WTldOW9@s^$?4ZH#Rd@#Ou`epr2@onHkoV15O;0CW|oh`(L(31+cO8Y znEBUmd@3@4I3?sQA%z1-zw{S*_6Z9y+&N~dWd3a+SBO43BRHi7sCK-}S1u+b`DxUE z_5Sf6k&~Kqws-koo?uqwmS`~LS`6YqXQt?djK+iY)=kEi~kNq|n zV2O6R_g?X|Au4+Df)A^Mlk-G$%jjDVxIAMaQGa{ZxS*KD#0wjHJ%>i`)WpbM|7y>c z3kh;4%RJVhN;gOh!HB}68Pmk*IbC5a%ILd8S&s6t9jJ5spDfHK#c-AInSo;)!bE60 z=)CQZ{zrAE1Ykj!2YU(d`UhC%ru@H`l~uA7>y9i4tNe zK}x*|?k4vNE52GNeI)S|I&r>o5xAcUe$OKJeyda93Zc$0GUARxM0@t-{k{H;VHxgd zkwa6!T}S}AOSrE4hLY&%i9+CpV0HRb1&4)$XzBoT@2%c-RVO9|h*LH?2oCXekoQJz z`;U%B2u{>57}LGKBDo{6%0$(g3>?M&*hb!NEqj zr>XtRrFtQ0DX8(;T=axuX>}@Ytr-GY-#+>2vN4P#{fwx0(>U1Z4&%np2zR$*h5TYv zdL}$T9)Fer0eB-bdR67u0)T)1n#IbdTSf?_9zOCc$Z0}Wg|C)2oJw5vI(+Qiq9Mpf zkp6O9rs6cApZ3i224T|L8|+@zp7+(ri3Zw6 zab)K5R^T+f58#mg43~<(1gKaypX!tg6me^_ShaSujV%PghMB=|O$?veh*j-C15rrr z#3S2DUKNX}%|s z;3iK_4%#Kwv2HVxA~RFac(Z2PT0g{bv<%bCXVVjTlj?L z%{T140N)Sy=r+{n9sf+6H&s%r6sP~9n&BP;GFlUNzSwJkKZx%&o@E{ZXoj|y6t?td zk6GTQIfQc4uYelyH~Oyx8fj{-U;@>6erRf_Wg}Z*h&@&7P$1+$wJilZUn%Y|b2cVh zVv)?X9Fuh>C z!KfqZn<1qhsXf(BL`x>-6S;1nUjtVp(@X`b9JVOa9z$F*YdUv1%z$FYzj)*GkC}wH#EA4kGy^1&Xz?TBV>K3UpKA##ZF>`!HA+4Sf zcNEmQIB$&$Ik~@DvR6OmJOd?TUIpibYpaj^^n{QjgswnVB)w}}VujMD?vOs`b?5;& zTXz0l`@$G6Rg`KGus>E&6w~bOP{592jp+EKr|%afUQ@UDULu0NYzuM{!7vjNycz1; z)V-smb}Rkx#y0+_8@Gr1YfG#1GIkk|q!?DSyb!D~e63H&VfeA8T|oCgCClh$;gjAN zmzGR$LP(IBu`#6#pk(Vr_OO__qqGQ6`n2q(-${0p$}5Z!06AqEWxJRg*(1lG0o{tD zPliN^hW=HEz!L%mXJHOpd>~h@}=nq}L7XJa2g47t0XeODU~u8NV+S8&G0Q zr>x(_P@j(b$ewhs-5uz=E;Naw)A89%p-|P+c zJHHi7N92Nkk=^31Rq9PZ{N7(WScF5|VZ`jtH}p<}UFX9;slZ`Na6r`>Z5eYdZzGw`+PPXjX-jw8RBh|e(&Hc@Kzyfo_d47jH(n5y)oCcm%ga8_km zn1{F+exZif5vdMvFE(&)Y`W752dcMon58>*YXc5|j;Lt;z`{b0`r?8NswO8bD8%~x zce)k;qg*e_BRCyNveNPFPp97HzfBDm@bPT+&;#x6B>L9H_aJq zj!)Qe9IuBM%C49`jR+bq*m-$^d-K^gmC+FropBoA`+y53O&3}aXAeRNwtaCJNe7T4ZTG{)1gph(&5S|+meOk1 zJYSbTZEWa=ANREkck(wed61CvkRZf|y7V{a0n_#ee;>5#j}4=iu=#c_nmIl4yat;= z(2{=Pgg+*$+yXb02Cc=4VEWffCU3$2*W3vCKdEX>hu%D9`HsR0;$Gje1={7>zv_Ch zziuEa*c~tvuSkOtA)E}xKQNE4JZ8`gKF{$Q_(PWtjA3ME=o{*xjub~-`Qv-9gMUZ% z6|+NvA0~bXQO0d7E}lC3sB7{#NGMTc7N^hJRS?WZ#p?u^%n|NrhQ{Q##R5SgGoJP; zRSXK~l9o#_hDS&2@GZP7mf#2zquaXpExhRxe8ZVh@JcuyK)OG)SO>uhe22G?{|1Za z4n)XeF%Psd-Wb4+$qXS_RhAa1ex`6l;`;T$$xnLQ4a(}F%3KP{6QmcaEyT#ZNC zFg@OX%rG(4aKqfmm0n#-|W|nXMT=1XFRe=52k(z55T7g%oEe{Wgw1jpoP^1o$ zjI-phR#xOQzsqm1K(ggSUUG*BKZX4splDlgm2KZ4^$LKpkvgy&;ys|Vame#mvw*&I z@*r-|VO|IR>foiUgZ}6&Md?=)>IQR+}1O<(@HF{=y7*_9oO_V89G-Rv+q5 zpa?fIb}}0C4-xCo*q~TZLGd#9T&Qai6_c5md`vnV{bc>yTwm;vU|D$7k~~AsXfh^{PCy0`KDNg-pz`Ilfac^qu3aGv=WsI;MaCx z+p9AyoN=RFYZYIC|9R`NgQk0oHhF8j--mY-sa zzlV$#G$3wmoye=s1OC=F`gCdg&_!?^VF??a;+vF1?Y_saf~lPNBJ6MG(}p%!ySDR zbos2sNEgROIZ#Xv}(chSoPTlG5pU(tJiYpU6EfpN=J#2E;)#t*TmQB zHIL+lG43S-CpN3SaJ$l`b3$Pq^CoPm7@54(=&>`8kzJ+`0uF8Us6Yh(K*SA`@>1mb;hIAJ32GCBwd^t&xpR>*CRp- zkV5H1z4{FXw!oK*rwvNR=Q7Xy=njWo3C?y`XM0z^9tN&s2oAnaQ>@&jf)C>u1Hec7`l_5e3LYy3R*rJ%Fv z)$#&4Dt!o)zd0$!ALwZ~ey!MjOX}W)xG7L)ISr7x^BAsPC<8aLP#zzM?5+BzGSqLk zBb#xfSR;-IEZZ2Yx`mEAD&3bI_f$7%S5c9$N$pL@TZ<3OEiKA+XX8M&$7b6gw}s&D zd0&ODJPlw8g8bHO9^*Dx3O5M2#qx4#SB`(D-+`Z|@#7aMoa?&3nfNH? zS)M85Z{wws_>2O@ol%Q66}nhX1s!jaB;=IiXRfD216_fH?6uDFU2b{1jPjVP z17EiHQc(xjc`|x=yT3u3!t0>;DxST(lm1z9FE22;^A>G`Zxc!6yK0~|l<6R`1Ri>H zU?abjUN&q9hhmpU3@Bqi@?aLYZ)*Gn{9NPNHl0_G>JOqvW4DejLF%{F#PFg=qrx+I zMi6b{xoc61;2X%0=WTqvIdPG?kgdPZICNQgPyA>!e0KRc8<%Q#QM2i^%JUI1+=gX5 z)N@5~$ir!D6c|8Jhc#?`D`t%yuC4sQb+tY|Jkw$Gl)lC{09xf;>b#qEvKq5HGc)tV z;B?qIP$f2?wxG3axN@EG=eE>9NP-iuo0aEUV-nN`6B;uqqrD$=XTWxYwpoM33vTbWOn6uOn!vTB~Le} zy$7%bVemLXqW9Mrcg^S;a(XzinQ-^d&;Ei&9__@<`x~Xo|Br>-Z>ELVQ1@5RNbqoQ z%u1_B`8x)VOUvk;WSVOwM8?kX(u9VC)43xGu_5)sg2vWZ$hei&#C?mj8SN3foO8)?&TJS1C_rk+ewLU& z(f-PMihCo2!T?OH;)3u$5j_=kV_>>dzHYY!AP}C}&63%!=#JPsLq3ySi>xoy;(wo> zzaJ+#@>WkaBm1<03e$H;BRi=w3;HB--MRP7m&e&QxNZF$UYzgIQqIhbk)df8k6wX7 zEr4*7!&o!!lrpiJ{Y&lY<(+3PU&15m10aB#l}hJaVHMPHfuo~nemFz&zpov~y0iKw zm}{81y}J3|j=#&yihF^f4N|c$oFnk7^EBz>k@YU=tDYv*aIW(c)WIz{@??J07Ap7+ zwyv^)3wNf2O$CM-!2w1p%I~4PT6Q$1Y;w{hUGrTJ>Dv9*uIkc8+?Po z6DbJ-CwMSRl0h@~>$!j245 zb`3@9>EU1GA|*P}21jZz1>0<{`09G`k7%Pa|@nydP|WJ1{>` zCH(}LPEs$*2h6i?k)6o!@e2>qa(+43|KIz&$Y#^;5On zacX&&OvrR+xunh--ai*EU4h0tk&h=poFd90mJ?Ov$=a_$@pU)ac@rv6~bzMtvjdU z-O()p9@ghU$@QgU2c{(9#K z_9@5RyH{+Zry$f=-9XT*=d!3F_tCkroT>a5CM&`=|J5qpc?&dVdmB0ZO7F4Wgx*6v zzp2HHi7NKtG*q=$S&6o8wlYJwYRfAX2Jx0>`~`YG4*04naocB?;~6cHNG1`8oZj}V z!>(O}-oVLFhTd(IUL!ztr)9#^+wzPj#WM|l15eR43~p6ud0hTVTZ>`ElSLVg56o~` z`P^IFH@o~o^wJ18qnFt(sq$LDgTMqfx&w0~@Rx=>e9wxUAe_5pQ3qviQQBc#MN;sW zy~=|R*q}QhR!02MC%m7y?uiHAWOzsK(L%9&e{FAL9`pqCa2&Q+cpCqq@hNShyZwLs zsJ9utwvZQ(;7RHgdS=J0pyMbtWN&_}G_AB7LT=ounTo|veEH1$ipuo3aHy_}Ie>o! z;SxH!7z`L*lZUt)H&rs$G1Q$zfRu^-DL&}kFoqwFa;u8u31%l>R4!NN@6V`q*KEZhj}PoVUw~CT{uYly4jFbTWFwj9(S%n<@I-Gx!IigF zjtA#epLMJ3=>f$}dkb_@cB`Ifiu#V+Ys}A!J|RCg8U|GHb}vz%=9=x3U##q#VB=gM zC4y$V$t)LVc3$LwOm&MgP(BI&IAMG(OzEVGHWRrbbhG==i?7Uzf$d#y`!DzDcYd-j zs<6hbw)9Knc~6d5VS;yq%IH*P63;PmCLo0}4NTmJbZ*_;>-$xt{MZp|VN-^sdrXJuEM2n|jZv3g>RTqO`u((^Ckq17es>s83#V7DSc#@8)TVI;l-n zKZ^@|i}FQy>aiY9{`gYMoKGwPV=C;6pzX2baR4R98LHH0Py5Jhd(wO`>DZOF6J*)X zOUyMOk=CGxbs-95EBB>Hh{{=nI)@WZJ_?yVko^p5)ehL*53n83=)4BqZJb|>H#n`1 zVobb4#8B_S0)m@%c&e@buz^rJ#@W8w{q`HLiY*|?T@m{(RN6l;k9Sl1AsYRBtKBr_ zsSQP9Ut~7V>dU|=V7=!>YHD<+5X~Vg-~g`y-XHs%c+ks1A8NJ^;DAv(t4R_HrFb2+ zkSxNMQrq0(_V)IUOO?Ux0Yfp%i&1!oa(y=%JaWl3y+ z;as1yOipzR3$M9Mc|~=uH}ahSd;GP=At>L1BtH_Yq`X)cC?!xEthNGpmkh|XEvrhi0XL`{V1xV-gY1P zl!d#oH3tRZIfU?6sE6xGMrv@M4^Y0Uq_;}-xJ(rB)^oy$ol#0vZRx*9Y1$)%T*0pZKGvDfM<~XAf{}e zDF(%qjpAi>*pZ0Fv0>fKqEMr+s(y=FE+L%DUo@N&4c?bNgeogp-W_5Js)nUe9W(L!yugql$F4@KUZ5rE(R`(#RY;~*r43-!D)e{ zcz+8-UP<;^Rj#Va579z*##3&PROdcpB!Ge&_osq-u7SFOY6$OjE;q5giDG8Fy(QY1IZ-n_EVUVM^#4V-=-I+`Ly?XPaT zMLvGb4VRtt-24c`pNHMXYbp7zOOzOFnDwq3x4_f7LRIOMd)Uxda%k>%pFHy#)}-px zK#c^Ap`$;|rtg-g&v{Nt^RtYCJSWs2!-v1f?BS)+fab4angt{c*#2~-&I%5`{=fHsPcSV9 zjXvk`Yt$DF#}3Kkncws%g1@E#j*PSyc?k zPA14PE_S#40=12C%2;&f7l64}2S_HI(NUsQf=2?+Wkw32&NV#R3TPePdN9$}&AqI_ z`=w1M(9f9X=PTF=DK1iNqSGjQ-pv2@a&G8zq@V`G+YJzB3Zpz_qi>(`Ei#Up99DNe z88>}kcDa`ME`4Rm@ulqm(`bUi_xrPmZtOwYykFzOb5)WxfbG+=^A<~_$=cav^*B1- zYzEKkBhK@SXL0jm7D3&}duzqa7oo|KEgH99tib6%EaLkIdCi7!u=#1kiV{)rTTe3otanf;S-5$n==@Z>=(G8V z#_PEv@QWkk3hKhc)eK~Y0BCJ15XtJw6dkC?MrQm@x7C=lyIS6rLR=orqW@o1y?Hp4 z@7q6a2^B`AOtPm)XoNW;?tkt%j_W?|>zwm^U9W|-w>NcSa716e5#Zyt@gK+E)m<@X zj>aKro_q7y`oAbD+oR?)kX>0Ahl86w2a|b{Nuug!f&mvd9VHI*k{UYKy{f>oznhkf z-Opj_mw$y?b!M(-u9W;NHti&{=}q%)d{rM|`@Vn4?%v4HPD;t4mm0qIDM^K6&h7jmrUcb5lANiubfpJG@ysFi0;?jjOE zTXsPJ0q^XqEAZ#@XIcC*sqPR!j}V1b#Y+!K2l%nI^Oe;zO4gMU_z zz^4LJ?I^`DWRXPEZZ!igwz$6U^(*$_D+N%(^U5Y{1o7KRgQ&H%7 zvjHLZeC;aKvIb-0RE$$3yHEf#_Qj$~0#qs^1Fx}CXb1Kgm<5`9%eF-Tt*?7i`K}6e zIU$!;0DcNGifMg3La6h?Kai%l3z1qo53Y)Bv?ie_@%#Uj78DN)74qGVnar}?1E*pgR z9<}tb2376x=7n(23ZfD07osCT&A!UE10~4FtUo)}Ks#p4L)lBl=;{G_xYXsYt`wUuf~A*E`Nj`)Wi&!_HJ- z&mcrFEa;z~9aWFz>O~g_IasktG$}?hjdggjdMxwZA4*FZ>n;9t6+mdh?P` zlUCN3-`INJyYS8NVqk zjDdNd?oJi@)0RgPN=o^?IouHg+-X0X{^;L^?4c03#{UF!%WG9~_0eyK(6(pn*9C>B zr_Ley3p9>aY`YY9v7SE%8??zNHFq7c1Bh+y8F7MDs%GkMe3xz<0dLpbR8Ha=DNmir zz5QJ$5Mi0v`m}`Gk{dP~`F=)xUpT5cbmNmxs?Txsf4J#(-!8Vv`OUF+yy0N%?@9RW ziA?XdH=}l&W0bBc7_|RQnX<90prq`kUJ*Om8c~u7B-lJVOw@y%eGBy);1n_}o|~3kY=~T(|$hZnxw_LcKHh}TSeaZk}kw)XRvV5*g zHe)i8lKhTMG?PvBdNg&ZMnx@B*-b-@@Rj%056u$$a(xSt*pz(v3nn4WN36uWt$@R+ zYAeG0avmX9tn=hASa`Baax@@GTsIRVjmOSPniN4KN+Hd5x_C!Tj9hh#!&X1L-~i7j zjwe&!P4G?}gy!M^0EgR#`s6S`o8($XRQO@2**v7fB~CJ(ex0YQg=yX0R0-txyz~aA zt)Iy^P@M&sP@GmN*z6c#G#~bk0?oh_9)21Di#Pj=S}oZOStw2>rP>Y79hG z`AEmz%W#__^yvRuHcY|K`mGk^eJn)vm{I2mnBQY~AK!aLjg8NvpXGcD0K7VvN0Y?7 z$eqjzN3WmAcKX8U#lr(fqODelE0Q;btpQJ1)X_yvdOpKH9iNy|pS?%$H<(`19JAX<=X@7xc`Ak$8{_ zquaz9vb}}O&bxYThnN0?I;7X;6(SMs<0}*Ovu-g>7Y}HChKG&&I>&K~eL}`UA>@vB ztQ>2p=Fj6!apo%xZ0-Z<9b)l|M~E(fHxk($6^E935`Cw#BMybU_*MR)byH3_i&2wi zywoc$)J2&~zXK(JAbnh5-T^xh{#n&lk2c;0!%P0K1*z`58SF{u`A9PR)WE;$5uK5| z#W5PmpKw=u#1|e!x_DP+g%zP& z1a|*K6jgsEn%i@h2}e;nWWh=_l{qnfUjnwCDR(F@of>lG9Fv2QoVW6&gkfjYy>pGt za~Y@wvs{8@xU=-~z4w7A(*pw-8^ zb!)lSd$*YNrP<=WTxw~HZ4jPPE{z4W+ad#k;s|bT@=z+|lg{UvPJnBUwKj4o90j)6 zJn-&X{Hd|)(k^Gq=W}F!Hr^vi>v~D>Cfm-Fty%hq(RxQ4X}>8$zip(=jWYk+$qFUW z@IQMuwe~W3TrHf&%R{4919!i2vs{7LV=g`3TqxE@!nbTX+OcX}3cQbMlw zDXa2i1;X~f{De!7ZN!d>Uv$sler){R*kkM(qvrX8Ys9azZZ}jnd&+CLP;7eHF_%U3 z*#rh;D*Z|0_RpzYTF4M~2|#9uKnQyX+oz$o)%=%MKTzC^WH^PnlDw!c>*9!(n_}O_ z^L6o$M=6;Z+cZt9(C6NKs>BmUkNls8QkKi|P|O;j*k@vA^$Rk#DdcX*36GpqCOW^x z9k3np49&^Gu7R7Gk-(I(Z01yjS?BQV*@)A3_P>7SHxLP1-t&G14q^5$++#ClkYpm2 z_QKcBpKzHI>Rc0aoFZ{dvRS!V4H^WOmWmu5x(jqD zVnF;vUJ@DGa8tt@^2Y7pl9S1H7h*{Up=~yOYI9V-8^(OH-~YJ?Y##$7lJW&{I^Imd zrd_xmd}*3aUzWu&?OWPDS7tsBD)99leFUL16rLbsu^JKSXn)znY z43~qOPSEPBGL;=So9H7J9hTbGc+SXc(9Q4V{s41S*6)_L&`-fDPR_i%sDGjR@(uardz(a z^yiz6mndY|)hMa5>VZa<+<^QOk`F>WXwKN0niT~V=B2OD?8EJwMH;{|pZ1kw+=d*0 z`-#CgdZAMc+TOZ(P)m9_4og@F#+%goPbH;WGTE?Lfn|ytRP5Pv4;$J%DA->#HgPmG zwM`imHsk7bW123n?Uh`7a`%-iZh)5`R7hD{i#^b@hSH8*gi?j7pw3wB3G=dJ&A@%- zd%?Zd^irg3-@5m#g~m+Vws&ILH&^IY8EV8H^C|Qqu5DpKeLvdRMpEPi*mEqdk1%}K zDv0npI!2Zdb=^5X2hCIHJ&Ye4vDmVgs;hZD=|`4bdQV{rAI`~s)kV)Uk77}4+V92m zw#uE@|7pU+SzUv?dJ9h54vWZ7pxH&N#OBv^!%gY%id)unF1O=dkxeYOfaR=*+VD4M zRF~4AF*lQ75o`}bKS+(ObWo$yWBxtdx*Ne2;4AiOn`NYVtM1{;$1_S6Y$F?jLdP7fivL?F6NZXmNb?1$jHekzgu0iC4Gj-4Ph4VbTe|qp+ttYlKB9tbp+FL z(u0d=TUK(SmZX>oYR1z3*$KB0Wi5l&fQSsuXsBF8=(@|6WE`^UiQ&#!>_n>o+!?HKEqT=#xsV9q}8tkZ>Li0K|SuB7dOp@Y`Nj zwSipieGexWV2|SDaN(88g{$*=E24qg+sTDVxwK#0*gzV=AhO6qK9F&a`-6QPLQb62 z4N}4TH2jA>?lGoV0yrX27kRfFwZ(mGWY2hfSS~~MW>TqSaX7#|WkP}DI;-rH+83km zA$%c!RFKVyeeD2$OZgwVYzKPYUCr0!K zzwDAEG%|H!2hSjxL}yN;kjiv5rAvt!MQ%?M`Ah6YsvIqKXmYl@O5FjH z^8=o%?49;E(;&|r_Or{bp%!J%c@lcO0#f(xjBo^WGrFI)Ucc>#`>ico5(mbTU4b~~ zl(L+VB8@+yq$IcZx(rJYqNE}7ikM#6GaPO}8q;`k9hq3~T<}tv9i+48oU9**@=AuS zm^A#V*++ewY)CsK%}=L0g0lnnb$-u9kV}(Xq4v5}zwo~`ZzV0;QR$!{Gl=p`*THVW z&#j@>_C?L4iLv31m15EaQM6e&uKl&%qG;#&&vD9CH=Bl}LijfU`HadLl@_d;AGx4C ztyB+AlBs6l$b;APecwMy zt?wN0i_h!7;up5`NW=I66+{Cy zrC1acygQ#2!+Yp8f@!b@myH!4{u-95W*n^=z!o1-E>jEaJ8=2vjEM^r>fbI>K=5T_ zVC2#~Mo%TUU9+*hARK}*STOSS@axszB_tGE_RL9FofL}?;6B<-ck|5e5C?k~)veXq zG09Eco3~*;^p|C1_*w+=UPD3+LwB2;$z^dnhS#@R>N70909{q?51m;W{fO+}#&WeE z!=9g5a`jI6BjR3WZ$iC@F(pZa@(&5R#ywx1UqsM1FZZqv0f!t!n0|omo?9U8jlq6> z5UmSey}Y&fzs(QUgohu32aG=-MhIQ+6;cwKIr;tU6(!ut%B*^snYiz32;kfXLHu)b zTV;*-mDE;UYvnC@&D}HD^2PhwSm|=9+ODIo;g&D|i`sigSU*)h?0HgTgFa4GUM<8y7_8TWKf%WQrn zj5YX_m}marYd&kz(R&Gb#!~f1w9lt=`p9^$azI1*>GHaZ-dy0*U7_m=aLHE+H0P~2 zH?}%7K}TZdhtAoM4lK=K z$FvDn#s#j`gjfCKoR~3ITb(e~PXauZ63?AaZTEU_NK`NjitdwdGyH+k@-8ks_#n`Q zFM3es0&F6X*@72*yNVbeC2Af=`uCbP{?^}L)$&a zP^4cToSCRL?ce?-QV6%4G~T(Tke9bfe!LpNW)}m|kn|%Z80b$l^o5=i?%H9FHlam- zn1>uIj~#d1qTG8263*pf#{aZUPBtEkvPy#Z=X$4nl`VPe#36F$H?LnA=acM3F$o5& z5PK4krnO!l{g*+VCzifjyBHo=)s|{Wk8Gz^!8h`HL2q0mnotT8;x^QYG%VRdY$xP=W`f6>=zJa<*9E(nf~lQ4^;>f?ltr8XQdmmBLvHMTD5~T@jajV2I0Vr12bBE5+GKCfu3E&p6B6bEpdbT;(sTg22I{rNaY`I8F>F>G5-Hhq>v?E2;|@0M z3hALtitIB)O%+SN)2*atoi5Kp|V((IQ(;KR$FsU}4=%>TmwW za)bjvcW|rMd8478zKwIyyt-5yaI~_+ z_0}3DN5Y`nAu|pj9}6=ySy$qmpXfm=E@n6)Vg{bpzogo9}rK?4h4pr=BUha^`OOiOgF$HG$;#Um7AAgr~Jzn|=_G9zv(uS$h zbQxgQ*!^Ar#RzzE=3;Rz?S3R>*(`F6gE)M9J%K@z@Y(O~xxwKlyXC)_q2`<<%&(kX z01-Y;bsQf~-!t;b?Rq$VJ!AeWA6%*?7pb~;FE9W&>?FX*Eh4bYbUpP7ykWIohxw?w zgF-K#PZY%w*+HBeZ;F4>>^wsdvLCssRlIRi9<~XZ?mcj^JZnN3JKtWZyb3uS@OCGR zMxFM1LRpfwWie39=3;I{Q>42t^_k$w-SpX!NlVNyp)WL3_Dckiu!IdrvP7lNa|uqU*g}$ zp)mpX$a26JI?S7Xdkdhx0XFy>BQvMnyI8a|yTK1qF1g*}b6;Gubde%h^?+x+dnKy! zp}w_pDZ3TzWRNyarX2Vzn$zbdX!<(eHHrLiq1pxK)!mSKKkCW!-7wtHCb+idgyI2s zf%UFWKM5W|eaoq%I`$USnTz113!qj6QsfrB9Q+F8<3p{x7cvo%dY$du8(hTbUOs1*ttNDUAZiBOX{twQB1Kg-{n$a0AtnJ~5ByxB}p!r3HnlN*=d|CdD zrc+vrb+CjJ&qL&Hn)5yixqoEdBjZ{a9TWpa>2hCODg%?f-2_J7KO!3~grkY-fzb%RQUbvXzz1{Z3uNg)= zZS$DD0r}(Z+H6(HM<6F%kWPdnx*{8+=V#vE9RZ@nT--Q%T%?)FU+M8cF;L^~AG<(1 z=b4^Z=hmvV%8PCPzJveM0!@&_FyVKfk7J@k-;P(cj6=q|jc34LM3u*!Oa1Ltcr!~A zklA1&hpLP@*dVS1FatRIRLqIZq?aE{w&bGuf4nLtve>c`Sg%tC(y!Z=EZp_ z2u%e=$9NkU4#^HB!AHDXKt%YfQA?*0Y2j_d`7@Qa)vNdFs7~Yaw+Hjk9q1*=Gv)cU zM4iiPz@dT|Dl^nHxbptmy}mFf_wwe!;FuQ}JX7@XKvhYpv+ZFQ?7 zR&2y4_*uTQ=JM3{q zADTM3qJnj0VssyX2k~F?jG|#UnQ;Ci=pP0ddtmpG5AxgKAazZ0B&9_b*=!9!*MHY} zM}(gtFSImTFCl2DuUXgfPfBk|OVNIUm!H6dg#SGRS>)(?m=p642M9$qKq z325m9JqPhEiVD9%KfC&Hf#=LbuCW$~k8rdVt+ISfbPgdYm?n)6yMaD$QD#_)Tb*e{ zxuqN%FJ1a;C%5@0Hb%>LK@+F<$hxqJB0DNs^{-HZ#VbS>_CIk7k_!0cIBfNIt_mqW z;#YLN$Ms{g2J>K>r8TOV;94>Bj-0t7UKN9O_2UT};TiE!N`sHK{mR}B11{G8paiey z^Ke@E&!=$CwYZX7z5JRlJYWaaUapP}+gqEkar@pK+*ka zksWq?{7+P{hEq7X(EGp0@)z^J$nqJun5VkiRC=@<^P2jpnuuGO`DNf)wj-n32{GrB zsu5X4Q4Z}8AjX_s8gq4-^M(w4w0_}Q2jEDkBnHOzjwwTFesf#w9>+NH7dsocC*(-x zd8LGpUEhNI^Ln^wIpzF1Y zs%DRpQTibvc5CL?ZppXK=V&pz@(~d?R;dswHuA%wToJ)w#_M$EYQ>RJfK_}0J zT<1EODA)Vicw0#uo}(P@_J>=zL#+_$sd2W5{de6`cSHth?}QGBr$M!4{O4Gb@aaD} zH)`cZt)JW~UUSOmNh(W)zPl0>LUo>%mEnYC$qOir2#n#x!O7L(0n0gRgFkrs)O)G@ zW!=E8PVQZTiG-d(axnnMOU$ZIpX%L`9Cgl=?Ug(5nyRwS+313GByl!)FKLK>_+>cZ zCamSO8@j);zFcmAzdSeK*NaAd=0_~M7sa@+0-x8!Hl?1AR?lAq?pdS_3_8hgj^J`B zvD{4}Ta4QXEHIb$CbD+vuCLoopF3hDN2xRw&!hGgC7(bI$qD$h2ac1%IH8P&lLbwK z9kvaN-w|U(ds#bW7mQI1ukR^RlHL2_J(Q$_0OLW6kJQA--6s+Y1(f%Sy<*s_*4DPX zJmHMi&@&nMaaQ;EpEP)x1!^NDRbee9*J-rNJW%d1{+m;X26A9*#J4W^^5COrvu)Ey zZxFN%mo)-n;Jw)!>!5UF(U4y_6#07mA9=hjeN#Irnwf8E74!;%_51VoKBCFS36%M% z02nO#`meKkIJbsh;#={x&Cx}@DF1p$H9-~#TXJ_!j4)cl{C3*Qb*)PHn#bRGK4ARl z0`P46=jg-fL_Q^bp0NGe#>M*T8oCH5xc<+X;vy01J2*N0;=sD_@M8_-|0yuCkDEmi z15MR0-Z4b_Pb6dJ+T^#@O$6k!$^lh!Fm$C=w-vo7@Bud8*mn= zzr`?imEP?G4b8xsF8wfQ*i~I{s^ztauAM;*ZaA0*wDKYTP!^>qd zq5VLnQ*grDpHkYkzolj*?7&al|4 ze1SsZt%??X?xX;zbaFB!C1d$CnFBd~o$Jw9i~_hHd5ehpWFa!zoWOqxJR)AgJ`-ZB zeN2f%W4PR9!Nd*8CYi0ssMxq|(ggNf)SOmF6@%g}nsnk6 zkg{dKj1(>{!XnT718Tfp9T;Qf&ci&~U5j1mbc#dv7lqX_HTTCJh(YmCz=w*`7k2<6 zO%pnS^zv#|Vzrv06VVYuWw(y8@ZKoo|7Tu#J?@Z0h0A2j6U2-6vz1xLCswQ03nE7* zfO@LVt z75Z%yF%@4PUb|Gphw$!tnw6&rS;%xl|Iw8a8??G3(3mHPwI94|DkA}N*e&6-E=mcE zwNMCb^z$0@%SmlhiRu9;ITC9>b3!S}MU0#muF*2?tJ{XIpfd}Lx8^>$5tbQ0KMSSo z24Vx=@1DMl(-_LkS$qRO`oy`BmJ9kFNAHX(#{2ArY ziR~}XSyc?ab_}eW5Umzikp_qQtn|=>_8-+8#Ju*6U=5$QRL3lE|$M7r|d0p&%dbcGGwQ$kvBx$k^I)IXqsI8o{t2xnFV*e{RUhgSCaGi(55#%&ZkmT{r*2psA&nw7#k~wM6-0q~f{u@E|+iFu-i>Xu9 z{)_29`&n5BTiBiYO+Lxg=GWu$P9{Mw|F@1n+yOVd&_!aOet)e~@JLg%QAnHP3aTet zLs3#+ocgg>n=f5&fsIF=X+wR`muS4Y-c}44qv%{WqH^RNLY#_^0V=pnfQ_MYYv1Q$ zOSxZ#PqP-(;vm9^hH+3lyG#~mA-ik~;vgb)TS=CwR!Jc zm^REDOgu>h3&>c0%6`?{NJbqyZFoC(+98mXDR8!JkhB^TgxRc~ zM(>UJZAYCJF=Pu1SW4g{Bp=jqOc3G@ywh6ZVlFdip0_v))o-~C-cRqlrfRv7(Q+c^X+yMZ$kTnV=x~&)_yH8)*8Y&6#ho*{vsI)(6{h~@KslnI|lS; zIo8n95ast8^#w5z>3u=_dNqA+J~Z{9n~j-`5qfCPK0IKIF=7$*K!ic*NgW@0bmxl0 z(*qm#(RPhSzW_i#I?d&#ipBd!2I^0cZ?0(UF55W=ZpWM0i}-Z5LDBT&)q2&;zWUqi z!p{@7=OhmO6vjnS<~9t6q{XVmO_YaD)njOVqKQVT2+w>>R3FY;5x2S|eTJ9zM|1A& z+FE%yDgl>5NLtgxIAy81F8v+HU3v{E&!L-%ubhsHs0#SMKTei;S=%N5s6BQfYx7=X zZKcUV?bUCEn+9e(<^2R3lhv|sDAhl4R{}2?fDW_gYfEDq7r$vNT%PYvG;Gx9Rv6pU z%P6%w>Hd){(j5~8KErD53my3@P7wQm9VF*u!9MU1fVP~}4=cU4%p@?Y2zR{w*yEe1 zbMll-BiM}l4o6eAZT|p}RtbJ&ne^LlNBD72i)A5JgZb|2TCV6;g)Yo@>*I;e8M82E zqjG%%(G3d5YVnYn32`X`9?xi_NL~)6c(iSG|(2XX6QOmQvD7<2raPvBX{d{bfw5FJQCPeEjQN z(n68#ZNH!A2BWs@Pt7QU@)^BLh(qS+rq40D?Pb;6?buJ&mf=peifO*qqb*wYwgS7Z zRadQ!ca84r437L#aeT!~ocUX%%*~pk>?L6eeSncv(!ID0J)E=8Jy$FJ=g^Q%^19@i zDlEdN99tsb2x$QT^tw)3kMN9XAD=5K`R}oIvO)>qNL*s|0e=~XW@zUA7*F&<-%WbO zRrNdaoX^=flAIl6TvSQv-}QV*>}Qs{%C9oFZO}F_)sP3PAs1#|i@ANNOy@T*71YqS5wTgOK0L_SJJx7_ z1%5{+V7A!k^uh`=!W7EP4tb4YebRHVKbf7&k12_PM{1 zDKQ8?ObR99t#09%pa4p8^selov(o;3r>0D(Kk<#|@~}j%TbYx)D%)d+&7QU2Il{0} zE>%)q9E*~VeWNd>KBm*07Z1yZ4G!J)hZgs}{cL|KQBK?|FaT+Q8VnWxS(C9sPR%ya zRGqX?5|+n8g!q|oUjBoZt4flZox1pj5VgSD^pp9Gqb&wJVL~!Y%g70&Rg=goaByea zd-CJ&Gj>|-#a9>7^Av^4jKSWBTW0y0e2V*bEWnapp5OD_RhN@Uo0e2jDPW+#V+EtQ zL8JYXUTW%zv0<#yc#XYDP=^#&a{7su!!@HgZd z80-jYos-~wk$^WyBZ+`rdi>%ypmOK8XbCqq+Jgu@#r#%a(J%Q4*>mGd7vLE;G!o3j z$~rM_NyVGM5q|wXzt~}nxcVPa+S_(u0YU0@fPi>134aJ z+yxkG!9<(<)lpo z!r_4Dx^^7WdV!FN^mUGlLogFkpQE`hor_xDS$Pt~C*vV`bnYyC6cN0CRk*VVeeeQ@ zBg@T1#AQBD2&ypojQ9d7?vr#_8#c=eAkWM-Lg$`9vTD6*n9OjeD|^l8eGPnrGWr3$ zq9p=Gk#&F*{KD5srCcKA4UM{y_8EKXSkz8+)4tmVjk{618XJXu3x37Ii+w+S=sodS z{ai8ZzhEXlgi9+Da8Ap@%v@*}UKUbg-ULns+Jt)Cc8E!Ru)AQ-IquiYOKw9GQX;pf zxJRGNT|aRlhaL~oeq%p;SLFM#{#VxuGwRKAdqb@%Y>uD&9CEcGOMK<7zi5G`0z!My`6qHJsA2k)wVV!2G!_ zK5ws!V}&J8QIlO?4m3M*?3EltTStfgjS-RnVC^f`!n#_cN8XnxYkkWg>*)q)T#6AA z*&;%ACRIvgaV6E^<{&~_yoW^Ko`$fs`_r@YLdIXUIR_qS1J9PUgy;1siQI6L(V#3D`}|W{ zBt(j*zVGnzqmX0F$}}|;>)cPL-~E%Ub39|W1?Zh8JUA_1btf&Fe)i@%L5y2v69wA= zF*xvLb~WbSWOyxDnrL){JTxlqQF>b+9_g)Nf{s+vnC%vV>*{y(nJC2B8i zQ8rP(rY(8>-VrgAq^um&X!d5D8k#HM&2GWq*ig}ntonP0T~(E%0Jl;n?Z#v$sTKd_ zKEr^c<6rGw=!56p_G5b!t0Ik@yuu>8g+y0?VX^mTD9-<#6DC?IWcxg>ou_`DT`+oO|gxQR@!kfe?o^@b+syB}rkr@lUDIPfqcmolk;zB{ZC(!J8DjZ2c+BrIgcr6@@3hpR4eA(;uBTh7+YLsSmkbV!@x6^%evd9 z$n)>E$TXs={6NtEIida%{u(>|zbDj>7M^@BVv}zcWY(1_y{5MJH7N_(0MD|uhX~u8 z^qvL!E9trVA%se{H$Bht$ya^mQ%kT4Bwzt3N3kRB=1!Rfg~0nkp{)m83deH3xZ^n5 z83U0})!jk$$vfg_m;M@WIjo7BiUmOZ<_g zl`9mjiIiTVj)lW`Sw_`HQWXzB`8b)7VO4K=99Aw7INtFL-k~F_$w)o@0I|M7E%N)D ziD|J%bJ@M_|JQTrK0?9-C|0jm6}rPJHej`Wx6j5UXPEsf$++{MfvC&EKzl?6ng90w zB&e63%z1{oz!Fw6eFV`i@C`W@zq!+N7=$zpls|r!ooxunCHX@?$5=;%IQ3O6UCo~p z!Hy+84@(vJ@iydvYFDn6!}76^91ZV#S2&ychxA-7$;KY6oV-@&5hdJpx>vYfYi zbi@aok_zL+23j%JyEQ^qoJ#f;UBw@bniKgA*^-|(o6fv9y5XHI40rMQwEWiV3raJ% zXwb)}4we93ex_D!4DhG?g0L-3g}#E4LI!4Eb7GqzN=3zN3&c?4v3AF6A{zE7{k(vn zh^7@z(!2FS18AyHh2UUlljVwRWz2h&+om+mypDX^Rc*0vbR0QK1i>+6**0(n1N~XO z8ng0|OrOLmPL8G$JV4Wac1i7NKRI97f-!BbSH@p`Q zk|8MxNvQL+SP1gcYDf~Qn-ql{{Vu7Cz9qyXh-9_`8$D<$6_DbJ@>|K{L)A7__qLE; zC+H*9jXqjYtZ81~AJr!$yse^8H+X#J1FxJLL^);kOwsPGZNx%Rj-dr&O=aIacWzWL z%*B5n8AZ*ATdj`uUz;%3$~KVYdAQqCFgZELZvy&CmK93fZRZK97ja^=LiQd$k|+`< zl8YL^5#1@r2wxiNy-gksOwX4Gsg6A?L=@n}#@LJl5<{V+=IX{={H4M6EDTpY{7g;N zo9fNX;A{*C&e{^<)Pcol1ov*yA z>lEqtH%z-j%DPG1!H~XGS6;5-w^x5{FYt=G2axVszh1jiBqb2E)^vEVX{hH8U9rQl zNV1WuT6Sb394eBVIK!?Z9Po=YKj^_}b)2eqpaZv8+3o?X`*mo(hKXSb8F;c#1_eBL zA?MAM*7lN2rgT~H)wT4NDblr(v72|ec_O<9RHg`X2Go)mM9#7Wx%Ao-v7-SZ2pf?q znqxzsEM9J#0EBM7h4W1;V|XKujQx)n05ViypR9Lq2K$WNS2MjCi#MH-_8h_u%*$)h zBF_9zU>TsAORFQoA4|6cVnBCra>l*F3;tg9HRml?&w)-Z+~c`h68nlh$k9Y=y?%>( zEhM}(!h^5Zb1p_L7Q8JqL~r7;6&T|v3iEH;hyd}8dq^6+18FFcoM3yz7m7F(ddwec8INlJ|BT@GsJ$o6KLDm_Ka6SGH*h8;>dsGhdXz9; z%N!@b;VAu{&lrB4NE^1)qw#FkYz89O&~tJjT&%a`PC<4y|4?@N5u2lzjvkRSd!(sc zkfTV{^A8c%ewxEuV%otisr^(s0ZPs79s*Xxi9uLLBYa1LaAX|=wGohLk>=Ow?p0j!?%_G>%q1q#o;=8aeo`{_$KLmLCrVA(uRO`H_lk*CTC9uHcUa>> zKUMQ|2?vQ#A^w{NFvi$>{>qby@S2tJq?7!SK7df1-=DWHJeE~T+vE8Kx5{;Og4p4{ zgXp!r_AknZYCqCEcb6UbL5vb9sP(p_z~OQm%dtFUdjA*5B3>*g@;ktzX@AvS0wee< zX?-bZVc}weStB)N&MXr7si|v+Zw9WsG#w*RF{UTBe*>m;rE}3{Y%Ep7nHr_N4KFG= z>R`XARw7_Wuj;1E5pm*w~V-TOpmi{_p5b|hsEX=y-hlJnPb(5OL z;1F7;96ST{B-lofX^_P4);o}n-(BBd4?Y~T_ssl!C7me!#Qvq#Js$W;iN=$EW^1X) zkk3$U@%^wE(WAAFaai_cUUDl8E00w{>WsD*#c0t6Y+gog1MC`%wwpNxqWPt~kw+NU zR07|UB%PFV>RL?{mG(CCFT#U6oMZ)QDXaIeXe(_5sr*;>zuT)sVPo1Yv?NxYZ{(O- z$(`%>q4mdHn*}}OqK`W@S2R**#zk}Zqw^YTC0H{&9pVD48Q1e>RzAbcmplp%dHM0z zPrSNqqkjEydSz^C#?(&Z4nLAB>H_{-^^}I+^|3HI#phMZ%*FbdXSHvvlgq%%z_9KQ z z|2m5gyp0*-0F@aWCK!kdAFhb_K3b5!=9lRf$8UUaZ($k`-&PrHIOA}2<+u;%`%IXM zFL+NI;Lb}t-ta$;VTpfbmiMV6iS#rs4iTA|<}g)l+w%U`rzIT*iFqg3*hG9oI0ray zvYSrPd^_!uM%tw^xsF!8;Nq*8PIJw4=NNqsfrqD%=z(4HBm)>3h1OG>jWNo@%U;LHST(t#q?6QTPz z*~*&*m?KDCpN)9gb^&)CYq#B7nLS2!_SiqLJPhstG?0S-(T+n?u*K0Y=6>izA1rx^ z?{U$GD{mu97c_7p!QfC{q5G&+GeeZ4PGsJg)ES&P@@ynCp()p-MO-@4R z7OqD=fqv@}3dAfH>4HTrLi6(BrEd|;6hG7+5YqW@|4Go1t{7Ng zLC%%^o48|Y>wAY%yO}95LHgrcb5#@1MPy8-x3G=&pTh&2C|l`L`$7n>(9)G)&d=q^ zZla!nudktN(9z5}mqN^wqK4%dk;UbwoSu|}8H$j1Y{FFTKVjbGnBG}9Wbx7k{I z*%Ua|A!}TjmrAb9gde58OBpsRXD4>af19oqifI1`wRxT=kWObmJ5bs9O}x?jXA-O; z%Wyv}mieycCZv&ic!X}{Goy#&yY{H-Q*pxEhDT%591NGkUvrE7`}4b%2ZyKd?6wOc z-+tA6JS6>p5`n+h0+)GfJq*9Dlv=~Zt>V<2Jj>=K5PejoQ&8-DOO2BBljX?AqN1%b zjfH=_)<-S9`pgB*d3I4Xl?ez|OHD}zuxsW{lV!M$je^`H2PlEq^H}K$@|?u?yTS?= z;-{@oK$lxlzrt#lok6eS$h|Vu&4`uc{DK%yuO%&ftL1stV>ErP?Jo5ueWB*MXc}u| zbA-*BBGcB-tupOBXsin~7b|wojZ1|S`M^%RIFpO=4Jj>49^giRgRR}l|6}Vt!>lV*w(f0#X7}mEK!Ihlq3mQR!Vo0qHgLE)aT?5(5%K3lJcL zcJ$r*>}y}&cYe*k=bCwDX5DMeJ+p2oSeEdc!<*!jEHPldVfCYP5-!{j`kUZnSU;+7OPcxi79AbhmzC z+2}7(j`kWV-vpqaaMY`Jn!c1%;yr!XZ`*lU{8G2hjhxq-fSIQ-bdofwaT0+Lc3};vombV3dr1tWf~QKv{OZ zocG2kSC>8kijd^aOVt|{!^aM7cj}TFE`ktXTnv)z^UOF1ltD8VBt%Vm@$h~RK8=5A2p2ag!ds82+yfRD|CH_$F*yc?MDDofAY{emk z&S$T|)Rz(=ET;vO$0-a0>-P|iPx!lq&HMXYor+=H`B_a~EYtoJgPcAy#r*3#%$-A6 zxj9@1-x8@uzGv{|9{}l{^X$`D-#u_<#>IDARSxE_7vTMqUECxL8#Bz~SZubUxk|?T zZ^)qK?D%yF*>>6O!~K{`O&yMsp~o;)PED^w>W%2L;3m&uMf-`+(9eOV9!%zRZ>Lnt z0~a=WV&)}Fy~Tg+rIy2OH-UaXm{7`|`%k4a98qDW1orR1Y)pq*-O!Y0v_c zwxfn#M<`_#ekQg3rweabm|PJi8Z5oW~0ST z5XgqBQGaJXbTXq*l-up2ub;WrJ#QyC8~Xf6?cE^-%BH7It#>QXwkm#>C$r2km~)(~ zPne%~+)(k16qPx> zKGCA2asDJbqb50_nCL}b_Szd@FI>-Z#*IMbe8d|=PP5N8?eodyK?kK6Ve8(giXYZh z(AKCgOc;h&{T!~y6ndT_nH*IlIl-%5t z&`3NIdzOX3bsdIDX@Iv*-w$g{*}IS!d4V>BI88R|hd<+IUp!iz6|%X~hd$_=IsJ5! zv@uXf8Z*&K__^0jvH^Sd_0b}xX7_oZBjx~qI#XuN>NP_oTJDOA{ktf19Dk~i>2__y`18L=y?!Uj2xm5*x-jL;lJO|14K;Vs@SJR8|}%Q4>^j_!bP z=Q?_!$47X}zCc{yJco+EUdfC+KO4BvvjX~5N!Bgbm4r%g-!Q$^G+4!z$Kub>DPwe@ z#s+qt)E@LZ_$3z~>(4Dp;`DSNL|$Om+|U85%{V^IQtRWGg1&42^l*8j? z1NTiqqgl;z-an5m^rhEMobEN8K{fosGnf^Q!{Ykw!l|H$E+Am`DzDr%Y_AL6-}-YV zb+)p`xRq6qM!HItJAQq~A#oC6fFKe@q_Q1F-}JuN{Nq26eGbbBSh)qf;pLW4enM6P z`QZ)MAIE#$Y9((*6@84|TML=mr6l`3qWL&b1zDjwi-aL)2WD3I5m~{!)Ey@9&TY9B zC063&XNWG-(}RA$+k@l%&oeJ-z210yoTcbT2Dvc#Z%met>X(Y~nCC!LF3moD8OB_JN-px`o4*Aij3X9! z#m|88ob=NkBy;G!asr-0J-yWXv3z-sQY`$4|D1pfwUuZlbQnZHmD=1R6=4>hxk_b` zc=UI)Qm(h4lMstbl?B6iioE_v-tl43#W;i<7Y+D{C5pn||Ov_Ym^s*E@x3c7}p4d`;>Jz(zl)H7l;uA-Q zrde6ua+WQrOD<@;DS4YFb__h9@*8$3u`))bHr;S~fXZci_Q|z--13kW=-vjHm?Lg!aVh-NUS?XSHe5ZMqWbTGa0-i}p*M-W+pgfpAALkMKGEpeKlNu1rP0dV0b4p|*&ggL;fwt~?OFD!)3cO<>=lOVv z5uJQVh2`sme#`5L8V%j90l$p`{hh>CeZ3D1Ll?<8Obq8^X|ps4kYO4^+j;cKu#P7z z8YT5A-kQVp_Z;J{qr8{1O^*GkNguZ+Xgt~nnLRKY=In&RmHu(2*r8N@yZEuk@FI}V zPKRV#AC{AYzUlPc^y$8C#CU9eK`aLIB1czjkj$LCU)t%(_2IV=pW|?1?@hB{(f;un z-cu%>Na7!sJscI0)C)d$@A+ruv|Fmdtu;3~y%uG0L7q_#~a%9AfEIc-*; zjfs_-pcE)nmSl29+2qhv)lP#S7thGe9t-7}8v#vZIsH*QUa>Sfz(K*S@e$ry2->ZA z*H%QVH64itu6?Ev7?4^VoZRmDO1UERBYf^19^mi{g149{&u|)8j{q+mh}NR{`!N-% zLRl$@&W_>Z({NyfmqIwmAp9 zUMl^vqir%XaUs!VR1p5;vy0nP&-~^;dd=geO@DgveQec{(*?ISOTOonsd!cUibESx zn6NT8RE~Rjn3O-Kr9P(d`bNjoxEOqKMwQ(@JC$@k8Q~pMqcy|D3$Llms@Ntk9--u& zi2vT5bIScOLe^PG{80|l_4Rmn`g7`^%XJP5^y^>{$u?W~xpDu0jWem@pIo`>@>0Xa zw2j&|k6mw!@`*sbyok|uqtZal{m}WD_jB}=N=|qxn~;w%;Ee23yQ*H=Y$n|+=!zuV z6MZi9%^EKIa7I=tIRZ6&?$i_52&GNepi}jqqC%3 zxfCHcRk%G44y0}?2;0Lyh?E=tD#$`_$mf=vCY~>_Q$eRR3gjJEu`a&% z-9ybR+r7=k5C)^rI({5nsfeHcGP*QNiswI@yJWn@CkbU2$=f7@OQ38ZjPqA& z+S2JI-&t6Hfu7#*CA}pp9&35{FD8XEJC(Pq znUpX8ae>dGoAa20$SEG4QE=JWiTwzs6s$dFBPWq*rm`Wo^I0vyQRc|P zLLkPmkztPhG!KXofj%>88^2|m-GhXam6RK%bhMH|@E)Ck(d(m6bU@O|B=5I>F(>H9 zFbou_Wnn#m5e(}2zE)P-6DnoJER=ALWHDv$4=)e0a#>th;BD?-OMXn|;7%`X;aOfr zf>Sn{Mwrravg7L3}EA+Vr#PDtg~YIXk0~BLnvD!688j-^{vTT!_7Hhw5X^CHUJ9 zeMhIVCuV_OxPqTGjoY)v6Y)#yz07jDeFFFsKSfSkD!TNVP8fP;6l;yjwQuZrX7i*@ zN;o0tkFM>jy#up5P^mM*b##*OGdr=|ejt1xVdcY^CK$`j4T7r!4@hxa&bRz&apkUl zo!mX@L`T@koPhjs-z^(?A2Un|d!I0L=xojR)L%LjSiPFMcDGS!0`YPGu4e#a@`%KL|kgH+| zZIqU|x+-%h7D%)7YDBKr=`i5lNs0AH2mUTyOji1=ljL1-e-+QnD$K5y+^VXhxu|PU zsLxg|&^)nBTt#QF#>=Zx81ZLZrIY#bnKfURkFQN|=L1!BWhXyJS&MJh(BN&)Y~{#s zuaEwr2EkoM&W(q8H=8D!Df#Daa|g{N58MS?v2s$=hT5}HVp-vG2UCQA%b|P(E+#HP z)q3c4_+A9XtpqEBK7!@7KKuHaLMyAc?p8@tYXd6F;yEVQY`Uc6#O|3mRpZ|5qLZW& zHPNOPvA26}YL*?=WV&PfjMdX~dk8;fqWIyGN-OS`4)f_pbjpHH0%P@b8~}m+UNcD< zb4Zi=q!M}#11{TLfrwSBdXeJusyj9$rH)H3q$P>De!(=*lc-SP?Dk^UBX8uDFRpsm zQf$?vClCR{uj(Ta0A2hXreK7a8>vI?{P>H7U4J(hNZ`nxudVYY1cIo{6|zHRnDr$m zI1Y@?Y~WnQiqvOIA8AZEUuHzr@cQ-?@4jh1=120#2w9Nlg`X4mt=hYw7x&WGE~~=V zJ+YG!TG=;xd9FD$NGuz5P5Y=Qx;lA?iH4Uo&euU|%mZYzhYgf#VuZ;Oal;sfi(zf@ zLW^!=;1vs8L0}P?+@asV0!BT&w%;vu{^BbCRz}187lkD@~ zUh5#vrJAZ6zs?dLPUx@s8Cf)q#age;fyB`2$!_}MBeePy&220|p$lOjW|2m#Fjx-v zoHk6^D_o2`d)xeCOpqepE7syGm1ffbj@>@;;aBqrMP%&}m8@lNNrCI8MIM;cA^6fJ zyd@{o&*OWL|G);2X!T5)WCXQ1o3pVn_g~E~$rcOxXqoPI!pP#JGTz+o7+W7j=dUVn zWvS8uxrib1L5ZElJ*D(?r|kxf3uZ)G)eDBakzav`?OpOlIqa+;m=F&;%UnEWEPzU@ zl5zRQuz*#?oo?=!dli+E`8!*;3Yy3tctg_aTFEQ; zOQog!D<%$mW>}u_x;);G6~A$lKyv*_xZ|D!ot8Wo2rOSA94xQ%Rt^!Q0LzOkU5zG$`Xj)0b0<5G^xU7h5y=A>Xxj8W=lXN z?@Wu&$tRA0*`fU|&nQGcdkYy~ns(N~xASTLUwY|%FJGV&A~b~8W8-_BL}0=Kz5#n& zz*I|f^45ISQQqO>8i_?s?n%?M0u$i8$z=WqA+Jn#2uFl%dI;slu z9#NfFy7Fi@V*yM8P4?CMuK+&=62h+K0V&--XH~IJZ$qvjpKhqrFIy6hSmwcZ)v}KSM(GIY1dna~0 ziF9wy8pufgJ3>zCEE@V)O@ayfVdw7)?GVgu8#i0o&9rCn0@eoc?U=qV^Acg()9%g` zAEeq6t3vDr5<@MezSW9{NmVrK4d2vE-nAa>qqJ_%W25T7#fI(m*NN(uId-`H*ne#0 z_>>3vmX`+a5_eyRzJXKYmI;@6>{V};P8!5P%JtM3J7?v~cn7={ zHZ~9NR6i1RE8Mm@T~2hq1~@n~jo}oed)Dd07jFsnN~^y5SQ#JP>Qmw=C%gy(>elnxYtklP z1+TB)s60^{jF!*F?_OAWF-vT>IJ{Q=M9^bzoI3lU@Am2)dB&#Yymy=(%&hJVJ||&g zAn<$83l$bKM=Ed+b?okP1I1%9R(v)*)00LY?`LA`C|sqqSTHp*48YO*h2Agb=oS4uCpS8^L)3|IM`%k(2ikT1&I;T% zfgDlBP@Jlz%N9q@6^k zQ7If%Ys)X^mN}l4#-kdr@yb(S^u2jr0XgCVstfpVBKf9M;q{62#bUk%-L<1#W@X6M zJ$`4u9XjhPwHd!j@`uYJW3S3Uw;aHzCXf!-iWd2Je{_+VkIDJJ! zi{GDBDQ5K%rV)0?q0oHPP(Ll<;L?T$X#F*x;2fqm{*_UsYRinnVmS}D2El1ho6@IM zX>MObN;yVm;QmRp;S?{<$|Bg!gBDhg3bObWADN9hty? z;&|QUp9zIOthZtqI zOgS}>Qqy6S);g*>{7PJ}arb&JS($yYcVD6`7u0-kvY7Yk(!;7tg*=^CM)0}eu3K2e zhRkFG8;6HRwM3g#;gm8JzEtr(Hl9)=ZDP$cPIZ>sS`*FJ!n$GPJs6YS z-R4j77b0hb|6CcQZqLJ_sFVVU77rV>IjD)u|w3k{l<)NPTpEUukl5`3!Z%sfxoD3q&tX0 zfbrbR9>$3-Jj>e3E8*A|ER&CGNze7o4a^}P13;*%@yF^czn>?k3kv`NO^~Knv z{kTO;w2Qm!$S4hskKRT3C0)2YKzAE;jF;-RYU2|D7d4!rd=E=nC<9uDg{PgWp&JF%`sD_Z>Il{$DgqlwXk_L+aL646hT@D2LO`7l#UaghW z)`^+zFRYec+E>)(=+)}U-42Y*v_bMTEuC^8lH!uTcjz0IHE(%VE^d6{4Ic2*Im}2I z8BTSv2nCsJwOK=|QuC9xdjCAl(qk6%R6I;A|NNy~+;+G3ywg>da0kq{(r^eK*_4Za z<|CT^R5@#dtl_^(=)%xJtIx9Y*T6jQ<5cGC1@Crd0Vi6^MxP0v6Yj?GO8x(J1qF*O ziv>#{M4+=x^H=z6XRz|9QIF;#m01fVR^2@6?OrT>kGicGBN-kJWs+3l3}P=9GZwL? z=(``z2+SkLQuY|@uu^H6(B^30(aEBD*h6c>^Mb+SP19Kw&~)YtHs$D*gZKR-PB&upO%3Y>Gkek+8%*U7lakeby;U!wqWvZcN6sB)8^a{s>jVZLC6*oCOf(CJ7I^k^c$@(T?DT#oxW{g?%h{h zPHZ}n3)1_|yZ)_KJK4bcGh66*SFs}&@ST#;a%JeH{*w`I4`t!b;a#e1V)m0Sab4j?bNw$q1LaKe#K%NY@_gwkmK5i zy+OOpW^=WmJ)TOT__k^V{a$Tdi*u^tCS5jVLtjJ?kQ2wf_3mKd@LX-WJ4~9;3CV>c zzKQ>kdUe_s>f#xn#Pwkqd8Mc4rZHf5FF|!;(Jnafkc7^9UjX zACZteq=ssBsDX}^VW&`kaIrPy!e}7dg4|%ur_AKvjY3# zJS&d3&fiJKyE37kxEl+Q@HdSe0;xC2F9tP(%^uXoeNs&iKY8>~lZ|y>pto{-R zwVtR*&7SLPnxUBXo7OiaU&8Z-UKMLY^r>_rBQ14!sm5xw?*PneMvSd=OL`v+wxK^P z-#|!NBv0O_iZm+}nf8b47%FkepjVLD0g-d=&+M};RwY&hxv;AKbrG4k)n37t+tT63E>e#pkJz?1FS2WBQ|(fSBx9xUwK;xsjQBcL}OT<_57q zZZ}E&<;m_rmCBn7IBmAER0F~P+Q;@p06Rd$zju8jt2w^`0f$j`eICS~qJwU{!cw;v zkKM4i#s(GhB`mRPSj4eX03Te9d*_%;l?;d(5c2G42nuHiP^9Kv4`bj`mHc21AU{Rd zR(t6Lk8vEc7p_0h2S~qx1lP^PZ_m;mIP#Ay_+~WqE>j>vf3A${!yg&)v*X|T`4mi( zk!%7B^1E8oh$NMG9-|c*tK?$7MnZ_dOFHE>@f(tVs6+En27eX1dnz6=zPT$*omE|R;O+?j;QTD0|z zBTXY{k%NQ#_U{?0qs3x%<5j`G0J!GJOWwOTf`Vp9m78z0Gv?m-KKcBAhMB0=RyJ&8Y2GQr?hsn#~>PQR48_HCG*3}pVAy|#Gn zlQ72IdyTguq=1pD<1G&<|1phhCy=o}Ep~4LiIkmDe8UavDLwhJc1@V;tokMmfcfb` zm)BGCgFYze%n7N|;)ms4svj7|nz}UDV%o0&1P^{^#wV&PorUp(`b)tn$woICGO(Iz zk8yJia9LIR(dm(%;=m}Ws9JkQULP3dZLf@fTq&Mf{6KgF{d(JXaP?oL&vlUNeM3Dm~CJH}?3IMn4raQe!}9V@;ON&FM! z#C~|&G-4rCSa!F#Y{pETdv8W`xH~Iyvm`4#EArJyU%-o?-2JR#KO3w~dcQsQCtngH zIJWtc^Fqwf3ErqRpeDh-_Il92xxv!X)i zD(T#5^<%7&s)E*kC#}pRDG?Gs%WlU&_PBuYdneUjx@mPn!G3nn zB-yzF<{XS@t>bnJzj3zOSP?~Hbebh?2$`dmt%+B%{RF0Ko!0v6RPS*Ijg5Yf8_;fY zJRbRz*lfjlX!GbrOw?Xoj7G&(B_)=5aVA7t&IFSDLXBgHFxAUw1{h>dTu+6s`y?3W z99>W`+UsA*4Pv-(cx@W{SKRmSWnPHIlhA+IS>Qa~OL7bT(|(%yvZPOHHKIKcp=`r> zerz|G*-0J)%j(&m-DU;o!L&_#JtFq>;rt(b_Hc$XqJl#agSXqt?)l@J68ffTRc~^j zJZSHgyQWsn*WFxbX)6fV72e7c>#{B3zTgj!ZFA;x&Ma0mawWJG%qnEB5I>1kz>Zar?XByI?g68IH19&>w zN6A?#erq!;tk0km(}WID6NLrB2*rCUQ$;WIAJ+Sa>m?^mfwZU9>cIu=6snC(kkda7 zV;@3#{cxa5u8nWtDzzuy%IZFB$kuRgCY<#n z9oKtqEl#$B8|DPw2CsYR1$T7Lcd|1p6iparCl7}~l%>~eP~q~wGCNb5VU3=$V+Jwm z{R92CNAjrQnB^Pbiq3Gyc#(L8vl8jmA$d zIdpte+gT|Du0xLdIlm8Vu+%3~wK}Y@%4Svfm&phnhy<-CsEoi)@fQ}9?HpW=P}qGC zA#Z_@Q1|a_LQ>gOW!XPNjjpC-$x%6T9v@WQ1aUs8YW`ox5`I1vLo;tJcmkF~>83Jv z=N!w;YpLzi>|^^tvc4PRsZ&w#Xf>k87yH>_&_Ri<%*VQBVj}@B2oUQUDPr~cE0DR@c8AzD@&A>?UnwnyjjexRyqKC~kF^9i46=(OyYdq!* znd3I~|5Cg-|2p2V;g)ZMQ!VR7>H52EulrueWeZZx52zr`{2+$h@uV^KCMK4#T900% zy$%iYB)ynA3$@9j29}dwqn>;%VIVuIjTl^RxwBVhRZOX%o<(yihuvr<_CHA+4yC{B zq8T0NKx(pNt(Y0=Z?m-IG`{_+z#q`x@oA(;%1+BwSxQMvPQRYSn!~R>X0N7XKOu8_ zomumJ`wm+gNV9{G>U(xD4LsXj@*%!Qi4!uC7Y`qoKdHMo>K@O}EOMHYyV<`GFBN~y?{#s| zJNud6CLs-n;yTsy3?qapu}Fm{9Yw9Wwnw>W2bFp#L@t-tEc5 z@aq}4&$2+;w$I6*_V22;S*y5BV3Nj}nXKFDBdCP)lvU$keZ|AE>V>R8j z#~b*M`yelZlXvd`{rmAiKA`F8wwj`H6g zHlwdNo|XH144k!^Gl|O|Q*hg<(dO$UmocTcZl`nO~dCJ|ck;=4*vAatbFO z%=6==%VzJ-ddEAkPuTC+7&X0sskm5jI3mw_gKU;uwjNdweJ`l`{zjC00sgisVQ6(# z&Ex$D&w5uhTMW#5r3!GireLyK<%npAP@(+1G7eJde1`Du=rOPuZgC5H?sMNs=)e2N z6Zbb%C-m5P(1wn)T8c?Ml~&`_i)DVzJ|=cq+Ky9zJ(Dw#eM{)?Fl?HY*lPi)T$Su1 z;y+(%)fzjyj}?UGijDEdUyV6nm)0hCW0J09s#~PTk~G+&1|F1e?cFNto95hJrO&!sMQ_#$qzn4co=H}YH>a1qvUD6KR91r!fh&vaC|6XQsX z7*Lo-p#W**7Xi$y#_Pg>i!s7u*reg2fj@$hyFPljQ^1@Vyw^?#U1#ChjRyP2yG_^f zs`gw~#rzs83zRU&WQ@xdTe}9lS8W-ms~DM$G^*}wzi(gmgdg*`Q4LlDL9UiUBy6jZ z_Uoh-o<9z`gNr;-sR3jI|6!puKPG-}^xDsGisn`RO`a&5i4S)jn(X5}ReV@>Uwi|N z@d?wB!)WMYxP-R-p3$-N&;?av_Dg#%E3R~wb25`ihuitLvB7-GK@ zhka3=vrHCFHVJn-?W~f&bY|HmRHxRy1C;fqMgM3-`}H?(APUO?Q3^_2j=W`I52mMiTT-rrQ1%i0iBKj6<>9C>v%N70zP%(Rm6GIOJOCJ~`qe)EV&~&QBYfe}xmdJUXL0 zeADeRMHB}*V_ z?e3oJ0Bt9*vR6(fGAv+f`EMGJc=GS_rVQP!Edws>P61-+gNjoux!vV0%?_@=wlZj7 zh=Y3!Xu59Pl=PQUbRS_|5iQj}DCmUnWobFnliFO9`CboPUg^r`R7C*m%gl&xMhw2_ z^$xV%;zP*)!%iLn-6Z+&BtOUgc@^5@2MY7f~Ds~YZ_GY+BoRUOW^sC5>} zfep*>vfz^b{3Y2)nJ3l)RN*;H!dP}DXa4Y7G+5#?B>O8(#OcHkgEse zv3kV6-XI$<8Y4qovs{ZvMM`9?aCEevwJG|iT*2l#Hp>>_vUBvwtREa0=0>5yz5t2Q z3Wr;n&Pm_aZ}AjjD43+j%EvfH7g%`4jZ{XByi<Bb&^mu-P=Mai`VQ#uIDO;47xmx9Z(Fc!(GprQG@I?iKn@!+ zT_R7}+M729K8!(h)xPqym9GH*q?p*fnkDTd=Nan60ShA7a@SrrwdJmqs@gZ>s1BtK z6neSJdo-g$tG705jEkAG7yTqtQ9NYu({sfnpP(VplpOblHylBqPO{F9-qhTO1LJK-8}wg#lLU6iG`?f!Iz44h~8(?3$8YqI4`E)tX?}=22^W?c%3B$n(uD zb);5LjgpU(amC7QndrWUXlivX+4PT2^98-^kz514qk8bX1m?ybHTA}d!u0%Klns@! z!Na%NE;q?2cq&K`N>LBfBKI8@0(V**Jup?yLi7%4#or=f$pDDg1uo$nHxW!FnURZGwZ(RJMejgcomXEtS(Htl zlr*oQm_%_!vjs?q-FJv;`VAtq#xbEE^$d2gzK*kz#Sw;da6Ci86; zr#kEUOXOYGXaCu;PxAfTjpWbvngp{S14b`{>lwyau;nPU4=1ElP_cC93-cQNxftK_ zi0~c7!{#gBzz51s7@N?xt~nA?Gy~RrGV<8CQ*JXh-)&TgSA&Y(O>NNcDnZ?ac`Q8B z8lZQ|4Irl(>IG($N{;X3fs8&h|0~qT@xjR4yXd(p|Nr#@nEY)NQ2CNaAY4k^yer|f zhp}<)S{wxgER|%^w$5TWIv6aW~HyQs4GAwk)&wFM0S_P(1=uRpAPREalj?hp4RuY7DC@Z$Fzff%9E9uO(GRs+|I zE*K+_Qr_Lyjl?X(f41{&G5k-!yXodG*!IFI{Rh|VoReMcnfJ7L-6q|kzRxvN>!l(C z2ttnjrNxJf*+O@+E{1c{CQTmZ(4X%S~c_NhX!6Hs`JF-=3?C z`RQ_HrXk7@(8FG(4d~5)FKWmt3UH%%w3FpA=}V`7JXB0bE?0Pl-J}^bVBq&GIvLM~ z7dQyT<%Ol=T{MV;^b|!S2&xjFW7_Jo!{FKL;Gl_^mVAzjp5@}37=2w@hK0NwB6{sKV&7by%~b=3U#;LR3%45283ab$)qhX~p~lVYoIhD}xU4C0L}^Q4akOrujNlMk85iICJ82B- z@2WMNy%r1t&I+7m`~*?z=q67@tH4lrmg`^!g-uZ=aD|N7&b+LXXp){l<$RUZ%-WDtZb^{b+x(ni11r;_ z{rmyGRDF;Xs$ZZBvgWk+&;4Lfz`+;xaXH~^rTZRGvZK2>@~kgP8nQnoIR9RC#3e#- z9^m7oQG18GzV)#F&u2$=u#@(_yK^g+cgJ|93Oj^Rq;9e7pCaD62}0ZMN3SSK2mwNl zx&CwhdQ1IpaN~FgN%+_c2}!aVjkfBnw^VrO>6JN(F;iTR)$U;mclbvy7g&#|&RDY_ z<$#4)+0Qdg+iV^v&%M5#v+60h*i%Iqa4Yhxacxhw8bKE|ah*U5&ty{7vpYlD^;-2T zGsXkkyB-9NiETGYENJx=v|T)(om~TSiM5%^+AYhMygDr%T(`tzTPJ$Z}U2p^H<@|r=Ko4a}W*y&9%9piL!mX==91M^#fDnf#Z>YN@)T! z9koioPO*ZmViJy%2a{^u(cw3IDfzj|2EQ|&>h(ZRUE>@&c#dsMd2NhPkYO#|4JOkd zbSyZp_vP}Pr%s}zX|_B_9&YQfcsma+Hd2Kl+Wxgd=sMdh`dTbkbvI$L)|`8to{vzj zN7N>>viH9O`FB#ck}qH26ElNwQ2onZdR}_(S3P@j9V^YyP@H>+=wL}POB@G<-Vs&2 zTCZ0O^KXd1HM&fKB>N+(x4pgncce>OVlw2UA|x|Np7yLR%UzwX@*O#rN3nUA^6_4$ zc2`WLTS}SA=I-X(ERQ*17RWvSI8%URg34asz0%X2^X(mppal-I4J;*jZWP684zXQr z4?o@3&WKqV=NMSlHopQM8yNR3i@_C-;2v@G%{6%0cBrF3UOSzlJhAd;JcxRZ>hD3E zxlCa1(QA#_Iilci{gzTX=GYoFpU|wFxUAZIah5GNH2>WA&REKkO6C6EV;FhtWkpgs z(^1DcM{gc$qfNn5#i9y=l5z!;m4u+F<#G4{KomjI$PJ)V5i+MF)@YE z8kQ<~)NUN$giTmq%VT%$V84qa!qgA^ks;Uy=ZdP#3h&$Dd zX*^Puol)9+?<9Zd5Q-?^Inhy+t+cx5StLmAn4O5``=nKytoFjIRwc)^YDqo3*7wiIEa|f|J}dXpRW1mj~Y#jkCl^~hlV>t zu0i{s>3&T9+@1a5yWl-tc`%dIGU3QKBH%C_0r9-z;k4DYl7aniV8Hylbm?X$QYzC5nA015%W2+}!G3pKqa>Q;vuS>M(qdPemVJ6L!?6%shMchPMO>sW zU>77@{4F{q`hE0={-oOT(~%1}7m0^bh&Hc|Rs7ybLGi;&bnh!)M&b6gP)u!$5pe?`;S;N8Rp+c0lqqTOt!&@10gPvCu{l$Cs@twKMdQj|G@tT0Jc>>Z z7b$?eqKGNZidG5IZlZ}(1PXq!XPgFO)XXaJ2{zqn9+V9!OhD{=7msJ}$Jcb%zj>sY z`^^3}@>l7|?Bw6sOSZqVc;;XPb7;yb^HS(!K?FGGwwJw@=)p8Tx|)Z3H5#o|{*vaE!Y)3Owk{<_scg^B!?GJ5;&fQ$7Db zY`t|@)9o8JE{)QngbWZ-P>CTRY{UQorxFuU=>`E|)Q}z>A}~ZsWORulA~!B+MO!J?Fa^&mla z9Kd2?=M(F&JmaLtP^g#KUHiTAvj55p7qD%OxMUFUqgmj7@q-bj)yq- z6Cn1Bf};teO3;L@@zT7STRQH2CRv#At=K~JQLfBF$HJ!Z3#-y5Ty1a^?s1R^_>b~% z^5L|kC1SP-z?S0lUEZZB#*p>j_aF86uM>k>NPvSH?@@|NatLr*u|;M}yyW0ErwfQo zyUfl@zN@4TMbqW<$8YOTyT*x%zvbra9?a4i^Kxq*nctfZVk(@lL+du2tu)lLnR3oV zn(cA!?Y`=HL;Aq(<%JGc4h4T%wiBA;$o1)Evt6)Q|{F4&e&X796oeE zWlnooApx&d@8vO(+f-b$c-tCk+jZNbU0`tg%3rI@Oo#q_b+-y+bM9T9c3Liqkvd9)wTOfrMtHwS+VsmHD$Lnp#y`Ons;3X zd&AI{Zl$&cH^94z6W;*=074XBtd&)dRCOrTsR~Rnt)MN+SD|E#J0ziEP${{E#n)!$ z&v2&pt?Eu2aabZ?X34OJ$+CsQFfKmV2)fj!S`!Y$a+QI*BdzEtNP+4X+68QWq$?u5 zg7>xxMdg-cM$4n3+t3yyQ>8~3a7AP?zJ9;Fn@<<%GpjY=yJW*wR?XY2<3$`rS-;DU zjcQ3%X4Z7_yXYj?{TTb&^K*+{aE6!1y>ub&qAtfRl(g%7{`d1eGfgWb6AY-@!M-P- zv;xIjPEli=%ITCN9^Nt@%nBN9i9NoV(wvtLj5RU9ZdEdD{rH2OZ&yBj?!7zU(P)jM z^StG=BzF4FFV3-`s)2LE$xXD`L3VOR9fdrrsKp-rKUse{+G4eeip!WVqvkgq0v4qu z_l5;m-`hVSc74Z%zKD)u`D4!vd}$S{B!$V_=|M#=Cl;-hGOY5;mGqZE*<7~*pptc# zOY@I_1GDiAHV(h3UqX?<{fA*$>*5mZX|bdQfl%wg3JxyNAd=+ zG!L8BiajPgNaYZXjK6rhIC~BFd9H9CqNuzw0Tzjj;Q}lu`URUD_dbqlHd$oQ`Gv-~ zKH=SICVQQYhzIFqa;q+?7_LZX#CjDPQErn0Z=}hlR)(CIs?=Xdn%X^TW_ep+MKGw{A7Np=A*V|j zDj1gDNgl?V6c6PxVNy8RoMl}Tkn|YJ^-p+8(@sge*`84Yh}^dZ-%gs(m*px zKX@BdBQwNZyc4P%+`zB?U75FI=oAO0fmLB1E7Mb~bxb|3U6AW>^W0)~M3D>%43Y1B zX9x~D%#rn^!6H!~pSRjvnJ^a@>SIEr?`}oDv`#tDCYfitx}Fa>!d=e#qfK=i%+`Fx zq4yhFs^=!u)9gd;6Fi6lDbrsiRsFD9py+HMu}w~p{rB23f2p(eta{kqLz>YDlatw; zG_Y^~3c>k21LvxSanouQ5}t6c{perlx;EAnRu4O!)~Mkh`7{Ccr4sMeIiNXNRzL~( z+?e`H2+wXN!{nu~do+N~K8`?c8Ip3as!Y0wcoQSfdeUfHn$2(0$=7o`6;{R>lk+v+ zm7PAj?{-l5rwNuc*DD(yNc)4O+Xr-vO0_j6<6bL2B4`d&$@aH9icnJJa~y?zy&=-6 z{rvOGI92-t3LV*bP@Bl=Ft=J5NI4!#QE0Qk{HsAuYpQYkV$AAon7Ym3hCC^=pp{n+ zgwN~3sdl5?U(t%~$zqah9E+pHLP3Zw%4RG?R=cC+Ik9$Uh}VL*2q3=v zp77$aini&Yn!B)zxc=f^nz?1kacgGHQwm~6h5E~~7QFzVE&66)jYpZ7gZp)&ymhF! zxMf$L zRr(!Q&f7dBiw$LmXXQ0)sl3K-#)q+^6pprtcbOoIHU__dU`k?NaNqsPt7#|s0GU>S z#1AjS)XVvdj+Z}#VC^bvwmX4Ov~_;zNR~0Qlg6v1X7X_0FSP+*V)c!XFeZ@tn%o>A zGkN>%*n`NjklgyN?DE6^vgr zPGfl`P2Xt+K#zk@#(5(x=q>@Iw>xrp;{rq`<~M8R6AWH+0dxsLcAuLyzBqq_RI3C4 z`g(~s`_|d(H$zm%3N23yaDH>rz7{i|g06@zX(2;-;STLW)xm9dAp@`S9=ztXwMk#` z{R*7m?WB>G#dK;?^XV@(QE=FovS0F0!5p@Cj(qq%uum5!jZSck-zTXaeP-{M*^v*}Aq zdrMuALsS&~w{1>8EQaSUkdbncOxX)loK)Du{vgYMGhpWaAC<;mO%HDb?(RNGMa=*8 zJC)Dv_@>AaG97%sIBncvSO)6)R_k}1QyMm}`P&bh@ULJ0{K#NZzvI|6rFl!%O!)rT znnj-_-{h?yDPnB4oTTMox}%;e3NG=vfHUx|o|7gf-(mrl(OVKnAda}1o|%*4VMj7O z9_#FI4OaG8Hk8V+)AQy;!*X1P=_3LnXfrVXY4{U2xhx_;18;sBEIPR-nA(Gl(dCtl ziEH-^NU(^Q7I;dixEmd+>Kt**w$~YqKZ_*zgYvZ-i~QYAFpPfh zwt^L0;L7jR5I+p(JB#X>;YZk+TffxseQ}qi@sA|U_&Rjl4NMG{@ADgR;>nM|e-^KV zeVenur0A2_;n-X?cSg%t!&7zGV1Tnrt1Muk7ln8L`6S^xs3hpSmHFGZ@wmbP{@U2U zHL%_W2u**m08xbX(bVXByH(!sCS+AEXQ?u&o;t|&f-&c_O=GpYAzAkUVB<(@*%#n2 zpGoahW@-pUK`Kss*^l((Y_m6&iXuiiQk&UGu5S>+8;M`nI&Hse?N$}a=c6ia3ysS| zOHG@f+>k&gO_7EwUr&a!3MK@N6E#&2ZD8+nvwZpZNm!vQ^x1Q2HmnHaC7s8|Zg7VqGThfTYgrwK6LI zNMR0Ua=ivR?_n+Rxd4l&ebM0d6KP-#^p{NwuIFZM)p1Th#=bk zTfk;#BDWW+pfO06Z<{oimW;M2D)!92CuFHp{9u^v(LTeNpto{;J-%-QwJ7QO@)b{tK8;<_G zngpmMU3^e;aB%_rBtl5N(V|>1*}j?PwrjOc3nB9G#w;ICCv0eJ;i=fD`SVnn|DV9@ zk^fec>4@|*4g~#Y`gr;hw%F|N5;L<({+hb0gBwe2uYuy`e2)~aEI&+o9=#J7gY8)O z#X`rX(n~xnx%4(*B-xGaG15WV+eUX)epp0ujLMuP(YfwnLY6x^>nV%grp-^&tHIW_ z;}^9ub5o_sM-AK`F7y#?<)({@x%oxT^cz*(h3w`vnoV=yaD9SQOh7zL*-I#vPUHjk$dn+MZ`1fz$c+TpY7H~Wx^ziAy`{{``EJWUq5j%{ z*5e}*H>B&Uph?5R&0?FWN7B`GsNDue3h)S*K@9}VJ_(X%idZb<6dqhFp-@}_U7+>?)`)-2# z28={(_rXu2c=sYk;U`P1;&-j(X4>-tz@NsM4Eryia8Uq}OKOR=Xw|k1Niz7s&9ZE8 zbBWx^HDTY6Bi0{ho(SlMyQMVuptKJZfOn^rke_oKIy12qYv`c9D_g@4kLDN{4-5S! zWNSYM6dEOW#e_@>5zi%}fJYuNIyc!1;`Z{b9lgeuL?*aq)^mW9NUx#KSLZTFC=v?Q zqfyqmKAE3pnrkGJ@4x{Hl}=q`zhTqcFnM3vxVJG)uB2Y31QR`GyQIo|pe)1TunfXa zQB^f*{W3WK9EmlW?EN<;`9C;URWX~g{$+{}R>}A8a0Q687@OfOTdES;6Ie>$-Aq8X z2*ilK@h-L-t{GYFDuq=*iOT>t!|?VKzUYeKar;;J!^W|+(~-0vxPULxb({4I3l;SD&5(X<@|0V)l{xuG%%YqXCfp z3$B>L%!-l4cqGc>JF<3#|8)B6$*^Pl%g}tO3!`xTqF3^3jd!p-BY9+GBqp+0B!xZ> ze#2tEco&`o_FLyu)Tux=6alyC#^jdFWm-OqM3rD3zU9PpJ;fcC_rUqkJ6*xe&2kj_ zBS9076v6^j4IfvgAUOLu`wWa{i&`B$$DT&{(B=0DXG)E1@xEfqR@lpj{ZQ3`lN?(n znJud%dP3mdV~9ER+rS!)&-?T34T`?8RQS!=YwzHF=AW9vGc@3c&!6pg(}LLA;++Xq zVx2+JKgU!pBq??8@iS@QC)hXNng^u`yy_~WErgs2zio++eY7i;b2+XCz)yIX^v%yD zQh!|PY4B1GCdD4`YP+YBpAKS>+MnXo2jN)}!iRLZt?8zL5=_4M*oIMA-FMeb+TLGr z=$_RB*Rt(<4)^76^jp2VU*ty#x^4W8p;}_QVA;@ zOQ6^&P|zKN0a2RS&s;L`XcW}jy24uU@8$G@8{ zYG)C&52DWxqw#e(QO5)6!)c0*#cOeR;GVR6KhGJL-(M#cP3Hk?SbBL*HH_qDB7HHm z?4ij_RM@nr%*#5yLn0}VMnm&;hugiAp#o&le|<#S9rC3wQWeDVBt|#k)8e1=-0#No zoL^Qy<^AW4J3OZ?zN1nRd~CBHWOra*a%^|-2A`AhlC#~O^=O30KE1G4Czmwlscu+u z-eX=`QaQX(sOpyC3rQN$=KCK0pQIr&HQ zH%i@2d=~7JJ$L75C9muMPJ%jJ3{rzcV`cv5olD8d+JJLkW*jR4b5qHl0ChH)2D$Jj z(r##o_SC$!jYI-BXTJtz&sBhinE3jtZV0KZrY4fYXgH!|cROCQZMMWHVx5CSJ5f|l zo?V|z{I8QX1F6EYCFj*m%d+ar+{j!)bsH;VVAdDLIonRu;em3H&{}}^uAHf(H-m7xH)z%- zTVZq2a^y8)+UzFgpeKUF&J^t!*1peUbo(~Tz#(FGd*2PO!aZd?`u;>XMaV z^x#Z*{ceP)`sNBG5Qb+nnKtg3Dkd=7SC-YfzK~rGbw{ESj+s3dqo{7mgfh{ACXxpb zjIc68sPcI_%LZX;A)mq^qm}323H-YnaPdJw6ckssp3#<5gL2#X0O%zBL9#q{RlWJ$ z>yZV*=x?0wM8;=oFA0%x7|CjH?PC(snM^`0{ue-1iLp32T5FdW0Wb&FEgc&WY^ZkV zTIasmcd!ng&VSWUN-8-#s_n6PH23PN`F_&2z1Vy&l>IbQe0IWps&4V{+E?Jl#o}dM zC?d;aFv0CNT>%Apd}RL2{mZH6&L8Qbqk{ddmZ8#!;v&1lg0Jk=g_!jkD~04}+O0XX zyxhMcXc|*?xX#+xrH^ zgo3i(JojEa*=pxCi4`sb=}hCA{6o3XeVfA$!Vyz@A>pIW&98Ik2K+>aGECa&h&1hq zz9pp}l=XU;T$}yqKJ12+k|kvNveGSG2@BgY`Z%M!(^sRsXmHaUES?Imw zzAT#nQ)u{zivgzHa9l^8^b5oOeOSx>oAI>xhBfr)wx9tH5G&tN`0{Z_=$n<2xnX-I zxy?p>ox@r7hN@NLol7gUtIG>229mYtwp@!u1bKyWkDw5EgJ4S3%Re4aQ zB+=(zz*+}2f++o@su&0VN8o%HXg_m^6(j>d;_IBdY^5U1$=7ss%d5H%D;nXI=#=Yh zA2i50mLi=aaFO1umm)Sb6agV<+C>AxU;=E>lXZ)0>ab0p`Cvc&X&6p4)7JU0DnoUr zG(TF$M{jx|E-0Y&Eh{T46P8g9FOIoT|&Z=+wE@ETH(j80My;VdX$ z^SfFG{}jrEBG-Jjn{7aV4AQw!d+bk)}#Y_Sh%85eU-?0A8(|Lpmn`C<1bQ*=C!UJ zDN%XHTEr*MxdhhOtRzk6Z$0<0Aem>5OSuTl9}tdPqoebUOX?Zi-Lh9({08f~pzF9A zE#R9+)#}gggrGn2v(Zlp`5A8S9%4y8T8A+V@Xw)~sUQy3bNNLZ!}OWd`SWWc z!P7JwY(l%qA#{xGY>=L7%cX#qZB}Hwx8xREFUGpXTr_!2%>sfLrqSDg0vV&Tpyqdx??7&*CS$)dAC`+g-R$0~w z&v;sy3~)m}euY^?4CV6^Ub?1E<^y>*$PHUwa^hz)Q@RZ3GlkkdADaOkvJ*-1^*z_U zRzohI=o~OHR=F305zj)kKBl& zKHU%(zCa$mcE;#0?>B`!_~Ciq;P*N0&!UlEU3SrbzKH-|eNofpwFyhF7)EU2wDK8~ zKXOL__rDSK_d)ev1OY_PO3E*3|ArhA7s+3VW{ISK>O@tvV!52ZIX4V~szYAU&9k{R z{G~HRswR(RTE=fZu#S651`SZv!SsW&g^TGHCDzONTqr+FJ}LFE51cM_U~>mI#z{Xd zzO>PBDb+#D8Kl{0A!p#)1`Nw-g|j9HwpwuvgRKHBdPfMIw>xfA7P6I6{cO^qcVy+n zFJBXOLR=vY7|T8s>6r^qeyRu6)W&MBeW7id(wg+`srV}I^U!$==$^ac0ntv1kdKD+ zvpcr*UAGA3X`0j&LMrRj{(0M;?0Tf(R5@29R~50;?@Z$3`G$c*$3`>V+p;fv04ay> z5F2&|w8IMN3c+Te&{c8mREa3oUamHDDj@t~I{VAlgIWJA7}Rb`o?4UbR+-w=yyx+7 zASE^2+ln} z-yXJoOW2H@NqnaJQFtpG3#@A-W$k@M3d+(62#ub*#_Kf1=(Qc6*S=W2Z`fna<}pyJ zm?iI%KUYg6pBzLuW>XyQ41vFldB=)x6%@GD@UAG!EQ+{P$p~hVE`*5&kumk}_cHd& z-|050`MAGy=sq^v6M!>&UtoVpxf})LY%3a={GP#B_UQ;3 zUtTd^x=~j)FLZNNbGPp9!3j~p4}8+|i}hGDm3`><)8Qq1kW^XXw#u$|YQ~>+?FQ>u zIajas66i?9?0e3}b^g-fEOe)c5*@j`=6`Gof6s(_DT!U&&!#&b0o$)Y8^81a_5SxAto- zxn=V`TpTI#guMb<%aMmRoV+QeCulL3b^Z&XVKS!om`#(n21ZTX?W4QYMh(3RiU~*f(?yvCV6!ko* z{6?%7NI^vL*f6biy=8)-9&IqG07gK$zk2kuS_Pw0 zl>hoqB?s(`0~=P&{i@T^Kx$9paQXI6SwyapvU^6yRNDu?ziWCT z0RK)LkS{r{ZHN3Vj}n)Fy|fG3yf$b-XqD*bx0R zM0HXvZ9IJDJX83D6pSWCx;&PtaXZmnlqOSB-mOSi^>w6o@U$&C1cl1Uv|m0f`v`kO zce?3xTd4kgrQysDIob2;{ckrv=j97UnofIZ4c1t;^eteN99VT)VAN<{J6qt{t$MI7 zNDs^e5a~de>Jo5dFSso2$zS?ql&mQ)WxB-6bJ-81ILVN2!hbFes|?4S4cGnhYzh$W zV|Q#=hIu&%#%`Il>&kxYJ0&rR46oSD817P>3Z?T-<-Ak(1BHz9ubxSqfj7EZ6jFA> zJ8hhQJ;^F1h*3{lpb@v5+x%jmA+Qe?C!3$<@0-D8jqjjvQ}k{Q9;k8M-p^3J@}-b(Ly!&x5~*+^o%*;bGS>_1>sAcEh@`0ZPL5ujJ~FpC*uvx~M>=l=A1h%vVFd1UMn zHYNT?PWzS9kC9NfL@Vs@;qIC=ne}`2g1&Ul z*X-E$j|x~Q9dCCFA^&R-CQ=7D)24C(!n*!Ba+6WZ1=WN+r+*@$#LU^N!2|lmF`YSa zhixdnoYwx+^m#w%3g^~h-I9i*K1RxTHTi`jV=-f->li%KV6C*|yR@6b= z2C^G$u4Td*i-NR=O*%hfjW7M7)&{uRX`y6N^ORn%?w)PNRI?03WV2!URgr$|JuGEx z^kEw`_D=0*pA4Lyr8$8$)`@U!>Bm!9fthAKvST!}ha%Twn*jw4|F_!J=ANUzGM~RH z%$(m!cT{DCK~c0yCQkKE!G}!GG#Szey}VscHqHXx$MIO&HUPKwV+nQa2iZZow}*deJ0>ZBeTey_6?udc;%<{5AGCIbVlTDIz{ER zv5ZjGUXN5>qa}@IcKMyO`xUTL*U$_fy|4UURPVJFj)(V9%Zx_Qz?)ftQOJ1$dt6Lf z;ge{h%{vtn@WD57l8t^r`0M%3XQF$inLxALjoy#Md0NNVG0yd~oPyS5C-slgmM!t_Owed=YsvdcuOH~uH=IgP5lRO05bgzO?sX;q*Q45hG4llD zZOQoOC=sX_kwZVv*7Tsc1N;7sVg%<57U6R@A{58x;xJOz(%~Kz8Xj`~tTY|CPn>Dk zk!>gi(74?=GE9RF9W^|lt2um*&u%+9lA11uo11Tqo0r>>DPXrOtRq9-Vgdjf3Btay zL6K}7YcERPi+;0yWWL>UkCLC2U)BT!nVI&(uR6^DewwbcdW_vFTX%sJy-CA#OLGmw0tGQz$qxkWojbYp&ZS;A!zqDzkiR|g|S+wQdIz9hda(lW} zcPO;v_6>aJw@65fU~1e7j|#PIR|Iy2&Xo$v2(?0dkl*g$?_OFu*R4HLpS;M>su^QX z51-TU5>#o7N7?C$QDLLPj^5d@O?M^%CdgX)j$cS|i(joET@C3Km&mUwotdXMWCBSZ zFV0FP#wO4$OF>dd&8|Fy__5g)o~P2gCIS2$JfpqTNj=wh@6p`h-f9j%+v}t9S-Vnp zlfc=B)n`o+`G4hTup;EEy!*x#yvyrHA?xSs^Y*u6)>_|KI9@-Yk&ykUu|E&=JNi`Q zGcRPfJ`3JXsI0a6Ll$Rt{1bO2Rm8(|?&lb6{fsSG6V8xBG&!_KUQpM?vlbbF1 z9Sfhrc*YpzeHM0=q0ga4S`$d`D0+{_65fbc=>GyFMBoMIyd0@)`48uezGQ(eK?nRP zc4*2AF~i>qlCwZ!HoGoC+XEpjX<03r%vjNgCbXg3_DM7Fq*=zA(6C=#WE6^#IF2lxuaPOC}lv0NFd_)CY zZmi4EIh1K4{;|UJZ#^x)&oBG=7)9n%%7NN%J0STttm?=E$ zI?yM4@p9PGCGP7Lu`U*uc((`gR>&cd3^LCk??B#NbJJ2IM9Q*`Y3*pjwJ2xFDAofO zj(cHJP@V}Qv&z(1`(k4hf7fRThuU}aumg?*?l}f|M$b8HoeJ{c#z~A$@+cxTvTyo3 z+{&oZrWkooky}e|cQZh+c1BrG_RhS)r7gDdl#iRexf6YzuXw)&G5NF%apYi+j4&R3_#$+v z_-D)B7rR0|;;yP9#UD1m*R$);DO;VX^l)FfD)zqX7)CH<6bte2ahPZo^ur_%%+EBK zT|7hZSLKf7k2)p~nz6r>`Recy%lXK&oWt)j&J+H@-i$918Oeu&Jf zE3oZomL}Oac~#1U3Je12;yp~6UbCl}yV<#=NS>6?bP~Xp>(6^~zoI{`t;%`+E+rmO z#r!j5hc>t<1UPH~-u`oxrsTj!&PQ6d)gq{+gT6edXXrBMSU`WAB6mz{)W$d0Rnl+N zAFsa}YA$Rw>?NKH%Zxp+xZ23#(!wBZxSiZ^lq|%nvCFN&lNxUN!c1w@yz(+EfK_pL z-NSa9z%&HDuDQ-uK68&WE_lQ3Nv3VUc=IVMOs>RDDxXd|Vob-_bN+3Fg>o>alSwYc zuC#Uk5WM>JT>U*@rJGc`-51T;lk}HX77rD*Mc7=$D7;w=tS>NGl%n};gn)z}rkx+9 zsZA|*l;e#tyEDJSS?zol-It1SS~+JU)RH%dOV@6UK_-aaW7HEm=YHo0eu+F^-g7my z-5qruS(g<5aPd{c85DxOBiVxuRI#sTwr3$fwh8s)FF zV-&8yjrU?#M`c^AGnF6erE6{Zsvqatjk3dZ6Mfg45~HMFx)BQEYteAH1Gm!5bZXmD zcqJgv?=Z}-YE(>fqD44#*z+s*0r_btWTc-9-lE7kKi5<(&ITi7;N=_1qtdze_9BeJ zH_!Jg8|oG@MfHR`z-q!pViw;elm~C$9c1CL5wJT5^S~2H_(wfcCx502jJ&&M+x?{V znYaB)HrIGULS-x`m+fR9m_38c%PGg54pr8dqcPD6XNS#EzfYyVCqCZIIUA(@Vla`z zWcl75G#leX1#eWF3|5-14p#K?i&KZrJGS>|y*ucS*0xS-Gi3TFf`5!KW!idp?+Ldr4Ge>W&C zXueVsyMfuLx03kTI$)Ok69-#e0M#E(tMMH+QY5bx`wJYSChp24N>z-T*=UUZLos+S zNT}cMyw-mV`ekwfI{K9#e3Q3IU3)1XL6i+sb-z5NR<5_37k*>#tD@Kr-*eGGJpZJ) z*#2tfs>0cvf@phsB{lNa4i(U&X|%VwF*~)30 zoJ!C{9^KlNpL9KWDhAIFp2|0=T%iWgo+<>w^1jfYj^j?p^E**)$?2_c@nS57bK1XP zKAl1nCEGcjI%lBIZ4@_HeFHBUYNRe9Eoe1K`n1I2@32Lt)9KCaeUbKrrx@IA7j!Qt z7U4L#pe{ZXnF5L3VP#Am1^TM&ZCx!IL=_cs$G7Bpl_wPty1;Ls#i|r(bh#Kd4$%2h zc}Y5eMmCIbMEDj5(U@K=H%@g*cb4*&wI(0tT$n@mbMeP?Bhr>-Y$9(uS`1Ean6SOv?f)B`V?5@6-(uCf*o^*81+x{( zb7O`^+w+lC2Lllh*wxI+1V})whxdLG{3voG8bVt18u1U}1VS3@OIjUdbeo=dDsgi} zkjQxY$^w{>cmLdC);lHUKweGN0ovSnYZ4nSiaVqw#B-{$j-=EWlP{N&9~}V#(BIE2 zSg~v|a_r-*$PaOo>CuA2pyo$^+Q-$eQnUnACB@F(U^)1S(b#8qLbJITou(NXd6)^(v`iV-#YJFY(Aa~J9l zBs=>3JI5bRJL-K_45n)Acg|*;cLGE@d8~ndA;ViAjO=>-c*E2>F?r0RGa{wEzCXch zdbbJO0&(8^1_|A|i;-x&A3W@qNtry@`EZ z;bA{B6=UYuq@SufL%-UE5q-u=HIH2X!^oHRxGeVZ>K|oIX1d>QKDxVmyl>t5Myvvj zHNL2&ynTx`U;CeP^HxEzDBYk6BA5Q}(mQ&gNa+eBEQkZ!0Jn@=Yy5?l$cmAYHA6V~DJL3~Ng zVd&0<@#b*fm-vSTpux?S%bjL>e@)`)H3xLXp~nH{FSzW*l$eG#o#-uGxnjH0MP_L< zyFL^T@}A-{#XWSOsixL^XDW;@IWX#UD-_7N)Hz>mPmhRolTl?8Z_5FcBvMiE zVvtEyn^*VwBOJ=%vkPsriNP-Eu6jS0%O_riF1sDa^Lm3#nR)RnFcCSCm;iBXZ_={LNxQe&57Hj#!#8 zQ$srX4SI*i)x+71)vDHx_w;qycs-^A9%{|V6+5jd>=f3MmiIbY&X2AB`umCOw%*Lj zGHdN(*`DG@Xb@kV* zy7-cm^TN$t1$dVlFr8nLZ2fLje@;3>iK+dKO0Y9pI%X)>pN)nEQmM&qrRvCoFn4}? zMdUsQeuARo&UA7;g6CeI2;DU{dY?y$65Yr^3N3G-&kP-YSXwg<0rU^w9-a5cxE{DY zu)deaPFe@@l6t~K!PUpt?&W20yR7p{mu_d%k)CCuc4Rf|k(?}~O6vX;|x6sFsh7JM@x zCoNhHnNd3L06DF)BOUKiyK8|T&IT*w=FQS;eNQ`}ke{Shc`g(UMjBI8v;6IjVA#$2 z`1Z;(K04Aat6%S96TPFrpT@LXeqpm0DaJj4YIbE%IBXSiCr3nv*W?hkQI=FYBlrI_ zd~r~w3egsO{v7Ni=$o=X4D0x7QcH;7jgzpwASOuF>ir4;F}-FL_B^0T_B63e$W8iM= z-O{`Fo-q9W8R(Erq4dI)#mKel{a4}TObJ~j!4KCSiZdm(t6yC;^@&={j5zZrKV578 zjX$U)RW;NcIQS_{;by?5>(xVO3lzA`0=qY2fjant;NHR4E<9|1_#+LY+gGU#d>6B- z?We~(SuwQKzZ|_Xw-m9Ai@*BhW&ovVr6O5VH}d2glKk>*Cp`Hvbw&goMJL9|nWcp8 zTF8xbH=QVqzB+Xxm7KchV;}9_4JKM@lE?#gTCM_V>BYJL0;Z6VRTiLxWLK!y%1;sX zC~M(YLlQ$1>bni;L>4UYFRr=<;u$c;V>TZGcJ-r`xucU(su)+PxSL znrk+C-r?`R^EQa;cs#|}kc-yI7<}Nvr~yg-RQ(R)GzG3YLIrz$(d>R@(oHEw737we z5I_+-eLya?Tr72knbSrLWHzm6>d2TN)J^pBQz#1$7q_>2uB10Q=NN_|?jQ57|1LMa z*^?49&KB%-(dtRbApf|ZsBsf}wrU_B4`$uKzdnA+zCp|s-QV$UIPiuZ?E~NEs5dEj zate=xWs7`5+Q!;Gt|*y2^NDG!teLPXK%vUXZn(elDvL#m-25uihsrwIpW7Q++gL)K z?%UOo^81JM7U&q5Ca-*Zn{{&qGj+N?=a|?&e1($4wGfjrnlEgHhDVE6n!M9TQbtoL zpE@jD=;T_64-=~BzSi_VKQd&{yl3(vL0m_t8@A$OxtvD%zgxTdWcmv2;zj{eK2tfK z1TkJ=UsCWYV#;MY&6TcOT5R!j39C;OH;mOUqW3h?UNKp6uQq_C>HGC1uI@hEliizx zXZav)O=e$Uac8SIp3|0#su)zU2j*MAxZeO7J={y&vzp+uNfC4$PV~=S8)`Q=Tey7t zqat}CEZu*4UiQr(=ZJ93`amAcbLo=)92b$)ihtHi{vDB4I#X5=LAw2~Yn8e9e^8x;M3 zJ-R91g&AwDrSs|!w1M!~hYNi^x$sc5pJ^q*?#6rHbhraV(yo&|VBeyPqumE53L9Jn zxfp}b<5rq|*r(|hXob5hMy=sNj1a%P?)a4*`)axUr8=46p{c+@B_3PP`K_eQ!&~m8 z%hX;!>dB2qTN?Ra3~HFFo8l&ds99ILD?j|0GgB}Yn%cb(A8R+I%^U1=GxnDJM8+6h zdCg~gbn5m>Sx|Tl&TOxuuE8oV_iR0Be8#@I!HoL#fGXD;56LsuH64ZSHMJhtM74Jv z8P~fKg?7W|y8hZdu71G!ZJ~zkDpswr)3rFtKSRFrY3ttYo2xEIYj)kR=%RPar7|44 z+jkf#z&vXN>>GCnnRA~QWm(PclG9K4Wz!c6qswC}XOQ|exT7rU$J{aZ)E0MpV;@E3 zndtgCWqVM5ws5;&kvn?-{C7#F-@lN#86b2!ovGY+$hHByvUj{^w#U22=|^<;C3^IQ zPTj#r_u;Q~#NrbVhWz^QKOGcoGtw3!j7i<8xfu;VVD*F=Ba_YYBIMduj%tA7sUj?@ zhkb~5pmM-OoA+tmUZMzoL)+Tw8@|4$Vmn#zFH|lY8V8UECLRpG+PMI1Dy5+Z2LYdt zUrKb{H+uOQb9dOk@ZDsxymM^EnyKJ8 zH@-Kb(FF6(#tOR93yV_y6EQ1U#pgs6m2nT3_gK*saksg=csc!Yn@3q@gj>_CG`cjV zATl~h4J~jN``1qd3?A15DLo$lwqe`HlcvLd2BDy4M4uP=ovuX7j#38F-}s!m9hV1q zw@r0vte|d`7?TkQJu9?{R*K-f2a~6iS+h_QICeYu(TJ8E{!ANKzk%G4UBBkqo_tX6 zadUc6Q!?OT&{eumqqx4a$0H5M<1tHKfsLI0{j@LJTJ1Ugb8oAE#+w~Ep0qgvrfS@L zD97oGX{lxRc@GI@=o?joo)^Sp9a$Fupof~Jn%@vIb z^Xbbcj(Qb<^aOFX^*IN;vrIKX?-%X`|JrLg!# z(L~2j)8zRSc20KBZtu^o9o#ylHuY_3`mFmn=~Keg`cRs2gE|E1)NuK3k#G4~x=&!B z$=SS#uc5_Pd?68b+(~MCaas|S@jvJZ-FdFK`HUNTBlhoueZ0un8AYFIZhUw!4R`6y z+;L>~Vi$i__FFZ)+Lm3SyL7Ggz-wdv;iCPKs;2drwU`l?X2a#B*=MqBlBPCE5oLd` z-^ut;7_^_q{s*FykWBZjy77R^uHNW4!62IDoqIA=x_&bRD7XIg`peQf_-~D4^j!;L zf|Y^h$WbM3X29LyOPYpkBGv3|EBGqJ_01gvfrM)Q9f4TA<^}~WhN2SLNMC-cr!R2X`^i$u47QwHi-`0?wuVLcx zMd;JDzQ=HF>?fI{g8S}!#p?sEJgBw4rLAIg$hHPg9f1+tKHz+^|CjoBz1-3i&5KXU$h;fL9Yn+UEz%ZQtWgT*^Aw; zzm}Kw6}Y5kT3@?2%ieJ4@wTPXccqdC`R)|67a%_m-Rxz(9hU<^Syd`!uFTKAxl@A>{OLCBWwn`vc#Sl#?X+lZ;}tZ1K!6=U~-UVfj6@CAojM^dh% z47qclvl`t{ewLT>kY;y$mc#n~EB5aBk!<=sh4vS)qg zDZj!tKlF@SZOq=Fh;w8ka$xCVIcU^H!UghsaOXn8c1u@}LdH6FExsLU-2U6Kh6Mkt zK*~~(u`l1aUgXgl<#-tC2wDdW2n0dx!N7ZJ-UHT?37t5c3o#JUT-iwlpz>bpzVrzi z?Ik=|ns&N6r#sSC@yP|hex%o6Syv`!Fe&#hRIXOtCnu@?pZ2~qs;R5%UqXT)qBx?p z0xb$60t%HG0<9Ji8Hxi#fS`y-nL!9s0>oNvsh~1MWe5rwm3fdUBmt>LnI%Gi06{Sf zfe^?D2}#~W+phO%pFghk|L}h5`Eb{|D>rAKz0V$gd!KX96<9ptVmVFqAq>v0L_)sk zDtP&-J;4J!%uUs?oN0Fuj0>JE%tO2+V@H2B3{hHM*=q;w7o=(|r)m*&B8D)t`VIKI zkHbZ2gyl5i(ccW=IKedeJUTclhcx90xbvBCWqk01C7xAX5jha3C^@UjuH^=AqCfzx0HIyC3dzB znuqmwt_mSw4r)HWiO{)~mJm#+!9i>GT5`|U6T#+c2%0;Brg->Ap){z&#-(ZH`miJ> zAHg(9yc~w6w?dj);GGcj1*6ePbOE<4aBtfsuE!n|==r=s_0FCzhEVIusA);;Txt@q zIbT`f=9gfqHY~(RFSmO*sw^}1P=+abIdH`V#kKg~ssVlk-#yRi^n~zX$$Xf7vWS?x zO4LaBT}VbS$m%YMebXdVzdgr`8t(w@vw)9q$>Jpq$W=qT~5kGYB>|@}eqK^ekL0+@Pb3kC-#~ zG50B1OKZYvc%_PcNv+$`^n>SN=ZTlh4Px1RZE-a+n=rLWN_6+@JC0A|!OwMNWBqb_ z^H|x8=e}-uu7W&Jl2?k?okv0!yYfQ-m$9WTe%*!0nED^Y1~ZKCY4Pu|3rvIWg$?;h z0LXZ5ZyPiJajx8EN@MQ)QbzLX(ylHYt9MG9K}yfw%}+b*=3%;==*21{yqBi#ML%LUO)Ic!-;+<<&ks0I7j!pJJa8U3cFGE`NL1ADdW8Q$)4k>&O#Gv zb%cFMjVI+LRaxG!J|9K=b;FFP-%!+lMfnB`yEsGuA8 zJ-wuh(H<2LRxwZFc6j^Fo%{qWq+b2(Ru9QZi7Wa9)T!UJ->_CJ`||diUeRmx>g!2| zUC1j4uhz_Cobr9}W1=ct!#wAiVdN|-;raEso9;OJS$he@vqJxr;5JX^^5u%I^6H`CBQEun`gFS2OL-;DM=3~+bxo4%2Dfz-2p z`uU*J7v`|CXfbvdtRa_Zh<~RN-uP~doqGQ4LJd=hXIdJ(0!7#382gwdK94oR!$kf~ z(Swu8z`ZI8ic`BbsxDH(Up=ALXD2&M^c=^FlouKuT}1lP1)Zhi~^5UMz4O%7Rl|XzgbukL-9q-1qQ0NPkP*g&wQt0fjl#EXlQ1^HsjUy zTZt2ctuIGjA-K?7@IVs7tKc3(G{w3H?4JM57S=fmnsK%uo*mmO-!oo{O3D#h7U^TZ z>oe(yf}u^6wkNLYKr)W+IQAjZ#N4@Kd&osyDr0N3@Wc72JhDwE1G<~Om;*_*z3Pz6 z(n*d=>h!kd$tpB&oC)ERcT01P0z=UqSy|Zn3IzQ!!Lq;?^;c?9dZ-G22cLt#ol=&9o^9brELjG3%i90i{@cuITP3G8H+ta&4tCR>4kEHCqe<#U80{Qn|u| zosaZNQNHVBuqmXrCAX8He9h-fh-=(jm^*J4q0S)F^)7!ERFtL@8YNeu2NRKmuIYN?C$4T*m-~{iG*tl(Xm&j1O7yo24Tb_vB7aa&$(jwbke5SRqBeG>jDP%9H=%782LW8?)17d zeWZ!kvO})?Tie4`S&9ASDRe*{y{b8!$}1{k1jcw<=Y+HM!TFWtdpzPtC)HtP=W+!1 zO@`jbElo5LuB*R0=%;sJ9`bG~Ju~V{wU|t3A^V97oUQ`%u`oec{@^)EJMsimXENMl zGpy@~i1`7tTVv0tz#g8bC7yahlD(bLPB!ft=nY?Q;1?Qk(yzZLw5g(iUKJY;zyCG} zPmlMS9uR($T)FM9sn;x&6eaJQWtT~s#7pB4<}{EKluXAU=reOl3WA%L?l=eY zazFVo(tIIBj+HMX%XY|o%wFZP$>fCUnl$8&D3c7n!u?SHLQe~v+0a~hCe$}t%~nkW z_|_*}P1W1)<%=xK>=tUFRfbA`Z~4lAT0}*oB5dBNX9;eTH{Ap4gliX<8#gc@VPn^5 z4Iqi^tvnA22)YO|UF{(%`z|6U7xd)R+N>#wI3qC|EMc3ihbza~<;m(J(IoTc)rI@p zUTqoL;<5dZv1zNHt?l9dF<*LO0&ExwA35-ZXhT*0sXTPbW<9Uen=$zP)lIB02pV0R zEdRkj%4@K?9U|G)ScS747MmQYPRi(5Wl+V8qH(ma37yzpcs)haG4zc>%rGK5aztHN z%_Qy1Xbwq+5LPcN*JUhjGKz5BiWb8dfRllvkDfE+BRmXmo(;5)bN zKDr{>T2pg*8+d7(J+#w*(VBkRwvYK6#aF{|cReK8&0LVAE=Ql;&6GM9^@Qe|_Puk; z!sqwIIYvptj=KxqS0f(8Xm_IOLzwR2odfMHSs!T}1#VFEy*cTjCsR7hogUMOJt2q5 zNors0u|fr^tv5Ipg0+j7JC_sFzmUF^Am6D_(y+PHFsff&0lOF}NBj1q-^(#m13d8b z#$Qntm=xteyX<@t)~g?Jk5$VygZ8*&uSf*bhe`^iUm^1P4XA5rtJG( znE-4%jFbjF#6p5ZNE{V1+Seh|R%IEz6(9w0QF(N^84KH&lU6#R zh!&pjc&H8`HnIv=HwkMoXZP}+{6;bag(Na~3Dc=YIMbEf^OnSR@!s%DIZdTMOfcM& zx$YSzL%pR?5q+{c58>{b!rv^O*qq+>)k^H4FI|W~ES@&k>lFjxd$nLG&6%zg92EW* zTJ>hExJRQ69K*Ts1O%i7F7V~doA>m^&untLdXgTQP$t5;Umsv@r0U>&^5ch`?^$j_ zXElNT3ikVV^x5|l`U%#I@0e*AL7?D_c1A1NRQNzc@SsRz%EC*dORa5G_c0SwSweVn z3T`1f09CIow{L+|KpS){WFMh{Kdx*M&+3*4W+zK?YX%&_PUtiFgN6+^;htd?Wlkfm zpX`4WU>Yu3l1Pe-14sHb1jd>;C(TqzZHhAuS1eW^OB-P{m?v&@4sCEICcjR|Oaskc z@rzq&=qY5~33xAU|SIcu3F3MQ-{=^!e#mHX4}epR95Z zgyy#x&_gA%RG2skA76fHs!j!p-y4eG?}4(9LZhv;`{T7MQ{@P=+IQLM(yY93IaxFZ zO>Py2D5)xqGwJCDqLzxA7<5pw+6TyRFp80lyCM)= z85^bzGMKM}#?DbggVBjn(hJ);LvLk=E!ekqrre-X**EEx%fD8d&5!y7Qfh@~e{Qba zYyVCP!nu8j%4xs8idECH-=eY9DjEnc@ZFqFeLtkA&v4?f0r|*KKDW{lfObLzR&a z%zNyYdy8c+VlC0u!`5#Y(RC~!O@>g7n@I#lXJxu?ubU?_Z*$Ot*T<6J2jO49jo@Gxl{coGWf$lIi*^XAuq}@bGKSkh4{x)2+rT+)*cn2g z<`2S1;q3i(B6T6E#TVR_Sji=FPSZS-Vew=`=y+&p9v5SR!hCWD<{+sv=$SlMzOAti zV3EzgLysFz^P)&K^7ytPA2)Ti+4mCDWVUI>x`NgM%h4pgIM|3Fah5+e3et%_p!a&J}3f8_SRlHF}%66KRCnp=nYnfJD#;``}rT&y{wblT&eOh zMDwD&v8Bu1&f19qmf?*TgGNDecAX}RwW*mZdTPps&J+pLJbhkJ`2%sb}ub8qqdBTgU!E~DDCDnZHvZ|c?cg4a1%gXFB; z&e>#SCA@a@je6^E_dV=bXP}o`!~U3|x`62;m68bc4u;kr37EJNAox|FKbK5LmlI~Z zTBG2R)n#^t_(}C(oQ)5ms>j5zy%yVU{QnB9#t}=`n-;PcqyoY4l z=d2AlpfFiIek|^#aX}=IzdAlD+Y&LhEVH1=rE~*UsPGze`I`MiuhzAsL=V5dF>eU- zA=S1W;H&m}0x|evLeVAlV5(7DEm%YXTLYkn8XNulm>UH3+*!$2!sHuGa=U%+FdBCs zTN?5E+0k>Gn7ioy7wizA5A%c;7Gw+#2^*D)4AKymACAB>uNaSs>S-s`8dQ!^KnK zHVdOs7<9@zDPZZ81Ve%IrVYE0;U zGpAD6<=N`2>EelunUW!-B&n#YABV_WEc||2L*9%hg$Q8Du2+$#N<@Umqo8P&`Ldfm zJ1Vnv%GG_b#(gPwcobBoi ztE8_ELEzp&j08!dDjS^RSy`X*28=Xsn2R3`jsaxcZw(6MgdP0hIybDYBi9z7=F|!z zq$SA989d*g`UPSC%xx4C!N-PM$zE?z-|}I>M6N_bSgnzGS+&*Q&UzDDR$JTR$9n9l zCc(%nD*$s$;;Q)Wkrge4sRP%ZjiIXKf&d2gq6|7lNNpsKxVPr_A;1y-$Oec0G9 zfx>AUev_#>zmYf=*KxWaDr<6i7$Z?ri}2QTh3}k+ng2lgUid7`&T{UMf1wYe))k&T z`=vm((yHb9qn|@-Khx_}N%m1@(&8ddhIJT$_@K z%Rhu{@dyS9|tQPsBJovanMB1dn$&Dc)uE!QT_h}w* zopx?2-&RL4tyZaz(;xO-mysVNarZRz)2@`;)mE?aP|>zS$Kw)&_`0@HA%IH`t9w>fnj|< zI+--hh=d>TgsW2kXN}c~_2Dht@^EW}VPAd#lye+<0%MxnWp;BFcg79?8cg{%~EmQzbc&%;J){D>=-%-Y64^ z;Ahr>KxsC4&OuD7x;Urf7G6HC5Z6C0RI zs2zvW2)RhVV78Tb5-grvT=a0>iS$*T5p`2nyCYHPwoGfIw?*HEPTod$ZG5aaZY^HQ z0ITmf==u`snVxA|5Omc-7-~y=ou87P3EwkuySFth6vrPq(%@a?Hl}}iVZq`|MeF#& zvC*clQ=a?=N(wFY86=1p;b3(a`Vwa3-R+W*lKh?rvd=XVnIHkf!7dQU z`@1VuKgPSPzdZ#XXA3q@>yL~7)W0ww=j`D6x7WSE9Z{KhvTtR0K-eC*k?183<)RVc zkQ^!M$fn^8W6HavQHUybzs?A!9d-GLu<^1Cn^|2+Q?fb;u)y%|4Bp0}LMhs^-wpt< zelVRvLh!=DXiSbWf}l0C<)OhmD7t`GhuLeVpQ@Q!q_Xuud&+YI4viW-XQCRJZMdqI zMLM^0%6Q^DE~w2cqd$xIVZl6O>yDpdz6ADfe6IL%`1^vf_}Ru?DN7Kfo@3#g;Lb1A zLl8Tz*7K7<00IOUU|j4b-Y1AaE0u z2wVb%u7V>N_&r>X3;2wNtHenX;()&5Ac=w-Q7XWJ_s9@!39DQka>)P+*Vvm^0K09yj6)UiG&t&v>NxB zUA#lpY**M5;=ckwjs)ko$e@RNo^+|LFC}f=1MDpD z70>L44C?lf==JDHEwrU#_PZ^G9Y^%d<|=GqTM#|dNj*!ssffupdw)*J3Nkl%K7E&eR>+~QnPt;m&c?zTrU4E{n&W%+M zx*D^Pez6GALC)FVxu5G*(|vqJ!msl!rRH}To)j^QD`*aXlzck%3JH(U80Le@fP_ZG zE5Xshx%G5;!NVN!=&VVnX?8g!Dq1+^Z4+NFAz6gv(md~Ve?&|bC|Q(jcph;Z9)t2KtgnmxTzz#C?@^oxtt!d8wVN#zv7+uT|wq%yq& zrv7A`Jf({nVjD5n9UAR5-rh7`N#n&E&RWgI8FGXFj1`R*fQGrs(NO3`m#2O1Btr=G zw>vs^GgNxfFzU6Sn^x@$4kfS8(XSMb562AtQ3K782{5`WI0&Hjt*t%Bw|^}5Au~WD zRJOKmfNfcLICiZu0AOS=148Wdi*&yqfV5Kna%Iou5vpr66V}Fj)^aB%2fYbk z<;F;;>)5f27W2h~@#Tk%J>R?$H+BEaX1!`?Vd?fRMFbZ5p~=XIg)gq~r&!7(yC1Kh zl2=gISrM*3W4CAfSG`2c!-&uExSc7PE%m;5Qgd)7gW*!+?B$cqTtXw$ToU>-H_2z{ ziOPBfWs`Z;Ju5{9!E-;fz6SGfF&%926>yE$rTkcjPVn_-(;o4gBtsHuJeB+{o^tLD zCVM7dj1KT>;7-l?$LpgJL##t+nl>Dl9bg$04L6(XLY-3+SyK*TKP89OvgELL(zkS* z6hBLy7-d6QZDIo@8!VH$B0$Vj5-aX&QiY-J?whIG*E;#!r}g#mX5ud$hP_gk_G>Xk znH7aCw?tB3q80`*)2u_Pl}%z892t(o1& znb1L0%0B)}T^b*Qr{5MCwgxWjI&eDWjfLGQnl9D$w@F4^0?wYe;k{>=Wrp#y{FLIF zvTe)kWw#8uXWho=l`bvK%lCL8w*rNIQ~8NJoQSz zU5K!{{9=9({5a(gg~GIULnJ|?6mn=K#=YG6Q)eTJ^UtGwuK8SCu)g6F5*Bt(^IP52 z(yu3yOwA_0M75YIadtOusWMjD*<2LS3GA}!KU38=kw!hZ-TNE0y1I`=(hRJ?icU(j zv!NI^&KNhOetp$aTi#4rJ^b)xG+|)r=@2;`_}A*epB9lUjX+Fv!2bQ8s){AR9tGgh zg>QDvh=SH#s8m+GqwBLHm6S5k6Um0BL0Vl3dq;iX$YkU-WDI^7x9%bF*2v57bq0^8 z8x+R$1wZLny?e0Ax|<}H%8;k4Im3R2)#Q6xmXps$R_^_?g!R`YB3Y_7 zAg?LUGj2bqtwz|{22>oHV?G4gxLM6?R4}FFeJPa`?PKXo zMI*l7`2o@w9uHa$@gK3Mr133(0w57C5oiJ_rH}>3@7&eWl|Epd`uFMUahI~QlF-ob z#}1SHR-{@k+Qc^H{S^xz@*J?@VT$b-AZ}!u+YdkrPvwp42V+5^9O~pT+Lm2s(y{B+ z79j-`5U>GFt7Z%bK*(1AAhA``)wYWL=bcSSD?n4|WXweIW}bH|pFH(wHV$urHw{CMe94Tf>6<@O z(B*xpARtJ)#YRQ3xz^#ZKO2BV|4v3a8fZ@NWGkYZLGqyS;%=8*oyv;~Y!+0)g_`pA zAeC8T3&u5t1)^8e>^ja9DZe2jzVI-2Og8DCG%I9#aOdEw!G9aYWSRm4;q?0o(CW$&Zwl#HZ~l{0g_IRnCfx<3TKG+gD}Dzj1pz3ZE*2;`g2 ze?0S_7lQ{<&2fzDdm8mm`TxfScF@&!(Yd4pwZ=0a%lX@-Tq0H=5&T z7}+XA@;|LFKd7_j`1VD}?1p*fY0_sH@%~n}BVn=dEbTLMw{S+vN&d-f!_P3{qL{Qg zsUg{ymTMZ0=&wnulK$!lsqpWzDzklHs6(dDlm4ZD^n%}Z6S zbY@~R*Z$KAXy+>F4v9Rjrds@IeP%5YE9WW5xj=)Z4$LvhuQXw873A(T+p{XghQr;t zrpk)G%+tk#Yfp{$xl~ds)i=Dt$yZs?hF6_|yX7=*3N-w1{)yY#lci=i@y#N}CXuj` zvqQB-(Z1AT?YZ3-u^te36^_ZZ8f^w3F(+}Kvl4)^m3JTWN)sIom+VF_?N+NfYi-2& zRz2rn_L_7SFDo7DYiLTy!;cF|0Sl^!G?y>$t>)!^#(I4elji+3gvF!E;!uP3**R&f zIyQ*>oDX@_$^alsXMn`z0b58Vq@7;-R$uI;pf8SN!gyu{_71d{|lo0-@X_oTRRkOd@Neq ziMV@7N1AQZa@1gLvduEG3xT!_mo>-N*e@$xa^96ku1%KgBPAOHmj5w%ZAY5DQwknR zHizF_TP%$WGP1(muEK3=%DbXY3U2h^alzhelZ{c6k{#!~9d=18RS;yore_L*kEiZ8mmwgnxAXMlb1pNZ2_2J>a0#26E`wi4}&(3LUR(SELE zZMkl3G?uCxZ0bzKXU1<(BE!k?Lr<@K242Fnr8v1KuVd}-e5n|r0NHcjiU0rr literal 822634 zcmV(#K;*w4iwFRI#=~C#1Jt{9P#wYdCwk%H!QCN9aDsE;LXZRx?izvycL;95xwvz2 zmy5f*OM<&Q!JQzF@4l_wdiCnP+F#ZFvuCRMGt;N%^rySJ=bSTBV`gFQ;%Msh(e&qc z8%}mhdsB13|90Wv;Nauoq5co@^YQ*S9+&$x(xUXKm;NnAOI2H3PeOiBxH0H73F`7z`fmyjD(Ddf`UnjkAqM7UzXP{01gWL zZ+Jf-91Q>-2M&k>_u2=b007_+fNyO7rvMQUk>KEwQBdF1QrG}^I3PU2f2hMF!oTr{ zg9ic-a1h@k;c|*0<0*fqaqy2t!RN{$pcPlC?*2o_y?K5?XM{@h*^!vuIIf0A)g&PI z%@FBZ7eKiGM?Y_t-g4@K1Xjn-s%5$D8{2Qt65*yhzN}nb`=k zlrOn_fXJOc2nDR+99*ace>6wXCHG?5o1)0CaH6u9BeHZFf}&HoN^j{-l)oGe#$Xj# zJ7Rp+vidC>sbcDd={H@>3n~P3_Qp3zua!JfZKR=elWF|HG(pd!ozbYFibQvLE*0mG zt$n|{a{F&8={U5fQAeLctx->-TOgDc@Y^yc;d|?u_HY3Tou7yQQXnO#Sd_9BQ)WUj ze1j4Z*fra3$cnL(P=a~KLnG6uqN? zQ;D{wcMmEG0=F~;7@r7NF`umnak!7LIbU7@Hw7xxep^XZh_>GkA724WWpTbl@G?gE z&bfiDX)IWC-*UqlOQhclG{#)N{}O=8brLkHg??diJf?XXyIw5`W&5EH!{?YzkFO~2 zq8LXRA8QCR$TkEoNa~Ut30qmVRRtA1ZwrLegKo;$lYPcR7g0fg_~W&h>FaxUY5fj$ zS_CF6CCpXJh`$NHwc39}m-9yQ@>FWpQGPp8>iJ3^%B(C7zTe;Pn)r8{&(4~>$?;VU z*hFzat|xCNT%}tas<}*Cd{>;;`)IwnrD;KHFUs&JLt$|;HjV;J#(;EDSa}PoQiDR# zrykMEn{>eP!Z0?iqyeGcxpawM8eSybefIn~?#BBvy?3;j8rCHz^7WyzOL4zSRAa*@ z^#f%c5a8dxx2bLy3`$%Lr1+<%khWzUJmSY8eN%LM-kJ7JR&=26j^Mpkk+i9aBWLE# zdpIcr({B8sgB2*g7d8td3C3fZ$PX^c~moU<9=foem}1ICPvOGyro7?_+&3 zgMcTLE#xcU4Xx`4%l(41=%O4}|`Cer`|)7nfm1aGvS@)t(iM>;?T z7Gc?jm@@quyZpH$@nMYxi_wgOI*~+h3ZKoOgFeL~DSnuKGY_Ao#KMkPWIX0dxQ1IA z))H&s?FVyCGzLbrq+Qmk{;m=gunrZtF8r>FH0Sm1m(RzDjbEzbRv@L#Narp2m3d=f zXzauq5tGNWl7ITnmRWyOwDKBTMFyuc?GsPD+BSt>xI{*h&kjn*WnFqbArF4r_=(H1 z(c}&Ra+n>fhc?{u79Rs!0Gx3SigffRsSLnzQC%R+Y|}u<9p?mUqKaX(>0G>~9+q!y zXq2YCZg9j$Y4YO_^jg@mk-cv#HHkZz&VMSes3oEH%j68U(^5ihk}7uC)mT>X)EUX*2?!m$qByu0nyhP{1*imt9GG5dP4@m_+C4Y3BWWtjhQJF9Xji3GQ;a)CR_SgIZn_u@A&XIt&b2J`yDpzsVIY&hWtp? zy@HQA;T@+r`maVy+7xr4B7$-E#%13MIu#3&s_Z>z*dp>3QHpQ02ShH)>>i0}ko(eH zU}B;8AN%MauKfs_Q%={z^r&2KOjN7K;|$6bEg%F)#Z@8D<*KF)t|8_M<5jc}8 zR$hurN9;&6O{3kmRQc0fVGMd{`GU?d3A3}6ZjGi|S60?AevF<7pL`fD*AOW`$PH&! zM#NYC68>LxFw8^lq39elRRYIz>Au(jlEeUh564Uy!OPrj_0-D3C@u(kxYP|cr!v2nH0y5?6C~En z+PO+i!ZoAsJx#);t>yL^QQ47bjz|+yMeXcTQjhF;R-Q%qb5oK>T@=xZ88WQD7>87_ z3Y6>E@j5mavo85D9Ds4|sE4W=Eg7rjZOQgAs9hyV3l@fCO{T<@2^+YQTdBIMtfj;l z5aEa|osQErU~P*r=ZgqY@s@veaNov;M=oLSV; zZ2tXcxJG`D1YAH+iZx57!$hvF9b%iUy4X5F|D^ys_i|Z8)Q&{OLjkt{Li}z1$B!vY zy`_Ow{XCzsmBQr!{No&Lj>J!Nfm3aNo83+pIW97&bSK`$(+0>aQzp`*w=DpPZ%cRM z;;HwC1QyiBaPhm8(;^*L8GNQ>hDkA^46;R7Sx3Y7x6WA~g-6mY8GYfXUd&c-kn zo3-e$5deqiaoTS_tlsIWe6v(h@*-e6U#5|tAVsU%g9RboD}Zx%qhrHX%45-mAf7o= zrEH$aR;>!uRVaC>w8(Ix-uC#cd0cNz6wj`m!WFo1rf}jyw7aLlKoPA|fFZ76uM{Jt zepABoE+>}Z6Wz-a)JH;brd7v)wVYrG-2+Zx*$DrRjI_95cn>o___how-K+Kst=&V0 z*loq>CyV5-0-&3Nr;LXD1@wJ}L*4929V+VB$P<0Mf5bgf$L$D-2Z*Mlh9^sLh+$>J zDpXDvHaaI~y0OPv_m*=^ZZ!xjcu>eOx&Yj%@Ui(7mk&0Q>w`u`QA{;>%Ja*zlFTFr za2h`$C=lvDZMnv7sK`GR@3=Wc`wAfNNw$;$;{@t1?&uB1i0GRtFW5-5*|O74-lY|4 z-iOy>^MA@e(aJ{m9w*3%eA~5*p>KYSPFVu~ZS#u$FN4^q#zx0OzqMIxNSA_CyM%kI zDV>UNiS+3hD}I0+M(E);&Rf$5gT5Xj8`&QPyVg4%)4QFJE_S~L@C?RN9+KYoGQw$O zKwN(-oNDy&$T#zO=DvB3uL2)Ju{|~XPfiwyv(*_2SD9H-evKv}P(WB~0r+v!6Z~b9 zv5C)8^LR)X=U9B)ONsoO*o5(b6J?@1c#h^+B+)_rqmP=+318F%i;dZD81uuyj~Q^%&s0waDsMd@-ysp$NW7$1pphp`#n} z+eY*7Jky8zejHuS5{a%UR(;eeUeX+*~PogFGYnibk9<>N6KB6Cb^6=rMyK^uIkcE>7JG6S<OR4Si?^IKy^LFpCup>Q+tNtet5Zy|7V< zzpu>crD_WrxGJ1i_EQ-$3vLo`E@|m91vR+i% z%qozU;E~;o6B}dBcKH&Fl(FDg;d&||ZNFDmbj(4*W{Vv}$1$-Hm3LdQ#`?G1I?=bP zJ!*?1D1@V6b%7#kR@YxB2?IayOeOZ4AkH4U6wCwaRWoU8eSdoQ{xGJYFO9~9?qn*} zk~FO!hY5%)*=KC51pQdao>Qu5O_oyV9(vI3%G)4RbJO5P_E^gu;1oXOD2TSG?C#!* zYwvBvjN}UAE0|jnbb|bwlEg-GQb2>^qO2FZXLbA}`u$iw?@NlYb)bxNeq+Td|hAI33RvRaA*sBKFpKi(THHR%}Jc~KMaJX zErYiaD!9j&(s;%RqD&FsMUiuE%HsnAkL!7UC~HmhZ3KXoqydB=%4?F`$Q@HR${C8r z0q+Ae781}T6OEq?a`kl~gj>QGj2?jHa3}_%1vFNGmCV9>IaJ_;e@ymAl6%}{|Ht$} zIqEe*oG6`hH+&ioWB${9m3zMWvEO)r!(m#FB7O=d*HHVSbkEaa{!wCXZG&L=JcxvW zGWF~Ww*WGO1o-GP$Qm@=1;$Io2^g+C# zINtqv&pm!@I?SnF5Sz6bTZQrUkd!^o*Veo)GF#uDwgNmm<%9Q!M4Y#H!Q~ZDtEzFX zyR=k}U5u061@J)oyM0e|ld1Q3y>s_RU;8U{y9O>NF=#0vPGaN(_f!iJp0@7#;W&A5 zX;%OY(oKQr z7MxK0c^(j*WtrZPioT|t5X<}mYam|m_E3|;lq@SXxO@+feDAPQO}%}hWjttSwQi=# zCkaAOslqt2AagIAUR1^&s;s6q2dL*MPG4u3tGXoJr@l9y#Gy4*K(@1N)V635u1VRd zgM^5W3BFvnX_}5IxE1PM!Bc3VpT+7}YD(g~kEHX;uX2FKp6Sc5UiW_s9>K|&fO`da z*c2ZKD&$ec-8vMyu{KXWRAjQVfX$O>!wy%(2Ik4-emFx2xAq3J@bn@Pfa~_Yme(4I zoS9QZv0t4oqg*ECOOyMOh*}5}5gP&lNz_uD%1vZypXw`*NJ*Dgm!-Fqq_R*BdOLN_ zwj1;Xn(0PtX2+OyGyE_PSTy{hJJyI2_Kn!uR&z#2b6b2deg@cXlBb-9s@Wa*RV22( z^=_Lh_Pea*QjJHss#Ub!$8^z_n3zbt1O?Z1?KGe;l9P8>vU`tEi3n5$02u9PDf)el z{Rrl&ypyo%=+iQb$yvR-2n%IK8;Ns33q3qD6#cL84|*s&(zTWG2=ivxE*NcUHlyJ4 zg#Bte%W!<7-IlSUCV*dsZ!I}V;N62C`lGTPRdTR9=5~Rt?#D4cf(B4+##GovF-bRt z6_JF1{Vx-jazIyH5oJ^1bQ$uHQ=D7B)$X6%trZ^I@hDR)M7gp6iDcgAsM& z@<^1woLc*dQwzOTx6GRa6Ea;N^4JTR-JQk;O9&v6R;i-Q1)@h*Rw%w4;)rO^%X3X9 z4TpuQO55_)yM|ATn2z2MS>a#F#1O4*iXS!goSjK0Ss7mkai!o);}iU6H0}*6m?xTT zC}7!;A?gx@f3sLUY(zSL##iQ5Wp+lvHSiwrebEcEcLshdl}TFuS?xBKg($f6j`o6v z&M?lM+5%e^-k@n>wOXiuyUb@V67ee&%)K(Di7yNY*QIxa7jow>5I4^q8OBg2Kneo@ zSHi`?QP;UpRe6T(Z0c*?_YjZyplfdt#hKF&L4J5~CszWX=i|Vi-?oK>q)LUGry2Z} zu=~3J`#Hz@j_camC+N2rd&4Aqj95zi6=QC!%jtrhG85vL@I+v=k+R|yQoEe>hMdCB zKM#yu$qO8QLrU(wAa@lHZshYVwiH)W4d|kiQwg9+**S6eeF9Fqo-YX>;(%L|mTGaU zZsj<`t+v9io4&3hlG&}upN1Oa097#)Lu}_Ap4V`ky-0!;1cTSwN{^E2&^`r%9#vv*P1H~~_> z(^Bf(i7tfiL*`sQQMYZ{VzhY53r>9|Z^Xe#55VLSd=loY5~yGz4BYy9iyUS=T_4g2 z9J9(EOnwD4oxLRNpa08Qa?p%?#7r#C{K?KP2Xcy8Db+>(%W~Lbam%qxM&A%-lj<%2 zj}8Z!Cq5M0jP8xzq6%&D4YjFb=jgC{{;(%T5@&p&8(%fRxdPYWG2eQU57~(NnXiGF z9gO_m|CD7fdCOe9db5)BCc=h}RjGbBkyihgmPo#LS^dw(Eh2LJv(4$=3fj#V>elJK znJw%IKNb6%MTk2%tVa8)?TOIW>!GfH32(Kc z=-jcSCtL?mJ~dNV#X0OF{tVe<)iK9c>_;=}pe|`lsU1*IAm&@_8kdYMnj%zL?2df8 z(!FBRpH6V2JJItzuimnm3?oO}x5SApv8bSKc9f(D9TG8O9=;EpsjTxw0+p#xv+vZk z*Hyb4rV8f4b4WUZ(`0g;J6Hx32>yiV0XwYB61fO6IXuzlLwu1tXf7IY?h9`PUjaA< z>x=1G7AIO3KIE^%1#GK z%E_rB;#aU;dADBy19ub`cIZ~*EH=H5C$qmab*y9r`$YHcyWkx$??1!^4r0=7p7qiv z6|tAF&10_*eGlIg7x*LNFu?&BqQ*@4Q&gpMJFJ_qDw^+t*fscj>Q~?49yL0c3%b`~ zw`?yGDx}{v-@&-Y81_hJ5&Zu1loiS+|2r-7zkP#J#Gj6Y*mWeTLPv1U@xRj zNwi0^Ifh;fGkpl79C8a5;;BKPEbqcvs}ImBaN1cZ|Alj228{MD+dsp-u%o;`+p3?( zwD4%IuG#Zt$9b<_C{8JKBH=30Uq zr#8~X$f$xFSS?O@)Wk2{?Eo!`E-+27m<*|OB~fr5e&cQ9x=8Q)M7m8T(KK&o%Qm$^ zKVF_sFvx+9xhlC0^QQb$DVbBCZ3)09;-KrIdLF@goGN+re0x^>_iRgS?4BXy9h`fy zxa&bAKd)z4q7vha7I0JSnP+Wg2){CUBriMHqRSu)`&e&9j~iLRU9Y59AvOyVfhKA& zfO>DwWiKx|%mOwpun_XYZe9lR{(*7F`%F}t$r!WRt6|%NW7XN|YZ}^Gz72EmD{y@` zXcv6+x#4uXh-Rz7jyySDiVFtol91laBuPi19`RLtjHgd`zSwoR&OZ7er>h&9fFpxeKGA~n(?3(NBL#RldcVL>aQ>{6^3z(EzT@cS(To#rm=OCK`nd z#p71tobm)dI-It5dV4+8aXcS!SlO>H_Dt?`ah*w(sd6OtAWU3;?2wFWp6UW}fBUkw zVn6%G4_RPW1P4fb5~=wO$+M;0oip~J4Ezch`$Ed`4j3mGUg0noO;}*Zw!&U5gALA% zlv39F*CYHn+@{GDbe|ES$3l)X>o=(P?4kKNQ&-+zU7qy|HL|xq^|Qk4aum0% zH7s5MZAG6S(E;9)Pho=_OUTKKVaDtDw-3S3j9iD^{PkXaiIb)mYwBG;I`l`kO8kqY z)IK*abrJN>=08p2cKX6bP{1$cuYh&go6gP7o|Gq^&6+!0_q6(>=<&|mYtZ45h{tJf z@%G5=vHORBUmK@LR_|hU_h21~PgbmQJRix(O{}dl98AL+Gs1CVTq~m075Hj3!V1xy zI;^61Hrree@9!FUW-|6-l{Wm`@Pn$I+a<{SJUmo;@pvK^mW3_+H0910mw7i0YCJkF zbYebIn+v=GWS2xK_-vKs%jvG*7#yG~r(-80dGIaOvyt>L<)}Vrv`#PHw?t|@I+W8}u?8qKEN0L(Dz0@2ths%;cNvG787vac)&A%EB;|DJ%X0igRu(kzX;Owc zC@XgbbZr43%*~MEV#hzFPK0#U(gX#4rRTnEn4D9M$!}|1O>f97rb$3qnttju$A@+M zP-lg8065fTGnJ~#6ZAkF;53)p>%vXZtu;hTQv576D%2k%TN9poQl>QFvqBW`Hp%(c-$)Avi5*9#$^~1h3SUi@m#XH{Dl^y z#(#OP^f9Ug-Apu&$w(SaVXNJM$#FU)z9_E=tXnhtH!o^2<`ee_mfoq)>?h91{X(f_ z*v~j$lKDX1lk>ZQZluC*eDoSO%Tmt66bnd9%OKnw`oFWi?UQ8S@D@X zj&ngJ2HRZE57VM~A!$OAkDxv2>yd7SCEJrk$rOj=b8)#{R(AQUdTuylHZy-%-9f!K z+Le6j=Xas{fIDWdR;v7TIxrf zSOZAJMRLJM^k3;A?5}{xlp)q2uk8B)H|u$y3&oxn!VAnU-Qzh~`j_;Su6^aGos-P= zvQUwY-k_yxE(ghCHo;RoX!I+AwPcd{IFny;k;V^r3Nw@v&$0N+yQ<-GBKh$`r}H z7^j3ZezWmpXuOD^-U@({j#wa@Pee)%Yi=daa8xn;fp78lB52K2-`6>20za~ zr@sO+AJKL$A@WsN6=ZjnUiX|EFAwik|mzr6jz*yr{ckS#mMB@n1SueKNZhQ zdW1vVp3zKB1tV!HPO~4fABg; zoz7crW;P_`g}@TZ#oO-LAAF@P04a7fk$jm#i@`)Tj?}DqQa$es7LuCAo!4uCn{n)Q zBEFwLYlJzxzZcr9?_^w&D`U&z1H9*w(({2DrSl#527N-?SR>hJc45byv*QJ`$+zwC z3pJ}Cvwy+{v-{Lffwn~(wmNWU?mZ4g16~$$41Mcvs2skwfi_@7eodBaSdx$=l;C%r zM^5u#n>84PsA@A+@h#=dV&fPDs5vKudy!;8*YL^a*Pg)l?@>R85wMDBvg?ieNyks- zLxY!;t>?RVe^a4Ji!q}Yw`uGCWB)*JjvFg(cnnPr}n{&m)4^ zyx48w_;%|CLvvyi+@XD05jM5Wh2w2xTW(}lK3onyVU|ge3;_hS^$?FZhs^yFlE-*4H^_V1s2GsqR53fs>OudfILxP%EB6y+{7%nibu}0x@ zSD}#sGDC3EwDZEpW%t6H{K1b2ZQ@T914JziFgpwSHcoy9mL$LxJ3d7y&{wDVcs+0e zzip-+M8CX2= zHo23lK!so|>!2HlGf4u^(OkYMM?VaRctXHSh6Md$(uyKuG>|}p5?UL zI;Ev)2zPfFj5`1##!4Yy-r_(*vFGAiKV+ypFu$BU1Y**wPw7J!-J(2$H-1 z0lqQ8$IVy3uNC1Q14_AWCI`NpqOVk$a)}0B)pQWQ8=l2>A;o+9jnFowi%;4&`YL4V zy7BHj$mY-_R?&q>Td@__$~?ZLswLfS1ZtF%Ev+qJjUV9IuL7gHGnkTW^Al@y7>^JA zTu!FISS}&XFP8DRR(1zS%e*H-67m)^5)x5{UpjX6B~*n4t;{6?`%p z5^z1q=$y;5Eo^Fy`D|H`5lT#%anxUAaj~p}_#<#cQbxFpy!tW?W?$S?*AAAhmOk0i zh=n!^kVVEg$bZ~Zd|&gK-0ej7f~OEk1r;ca1rwd*=K1;n61Ez8P%%8&O6T-Ud^#A7 zk@~zaeQ&$CQhEj%CeY~2czOjyZ$^N5nQV&lZA;_)iVU6R+D2_xYblc&V;ArRx^Ie) zTMbL4(MB^v=4uH}u{jNACgfF#KJb4<0+X&s$Bpvz3VnA=#zwDFY1$oHbgGvWzMh+B zW90|^aH9oLcBzgor_=lUr<6Iy0cgSKcNL};7p8e*f12O>_Y#68m)fJFlUDSm6$X1& zl=IVPTW3z#<~FN`ofk&VnkPJBPu5zZGQ6!kNaaX0eym36V~Wu_Vs(ymSl#FK3C^U& zHt2=~EK~PYbM~U2+$^_?KxSx{&UuOdQk^Uw5tUMYHK5f&qOBLG;d`GfXN(*MrK9TPB?OqFB_3YEYsHSuF%f* zjHXT+Zc;WjJ;kpN2LCB3?&BR`D=Px?BN?_5puKC8})8|KPn` zSOjT!ulgTIt{ z7FU-vVZ<3A|BYnVIkIL(bBqWLAkdFxU5LAm%}_D>hi4p~r6~| z{V`4SuIoYNg}wIS9z9{>9uLHhumXlebQCydB&`j2#b#QMxT@-!DnH$YB{4f3DwI&; zq6&*HkdVKR$-Ktx2#(1}V24B6AQOBNMx4}Ca5pi&0a?nZ8%>=Fqx67OJ7|&-MyX&M z5FAQ_o+M(qhDhm=nLc*_hwx9M%6H;hhDHhWD`lsc1oDk{aj|m10<~j&=5}^A4&)rX z-SjY}$&a(tOE=BW>QDyeb_SrwU&9w$Y*PhY8T1{RaNU zd}nTtT!8l3R%{4t-q_AiwF307Ql4!;l=edbOn@=Bm;6|oR&Ivx9_BP@W?4K?(q1R<}2r=$<>e z{Lb<;__f~}{>9#(< zd#JrE(ABkoCCfPe^(%t+;O9dvW-jT)>KPZxn>0oa$&H;yujhk-L1e`0IKjCe*|T{X z2pq$-Bxm91hS>^%RP{Cz5)zih1n}%#8t$4<&J}F;W4%M3Ev7B+n00hdKFO)GjTBKx z+3wG^TrTMX$Qk&PpS$cXVY`jfc7|-i<%*Odav0r@nu=7Ig~#baPh|lYWZFk%yKK_~ zBV6PZ{Z7wv!S%Wk>og;$?<$=9Fx@-9hhOvKpMQ$?`f{G_2i^H73)!52*XP^6diV1Z z-Vxuw@kDn!qvyJycZ4yM#E0zEWn98Vfu)sW`OC;5+Nmxwc!#2+{#Y0+*D%eN+fWPt z?guiBXf`6~>{ty#qg#f|hR`nwpw5V5mf*D-W+1P{ZtJVb4B=fIyrMg5cG@^`;W9Ld zZ8scRM3W{^_Ey+UwC^)Ki&s!6pyjcNL-ttnf?9!lHN1jEu*U&AvAJUl&g({pI@da9clsL`*> zodz2g*3zp%%!2;)=4uIIhJQo0SehPHf|P-r*Gpw>GCvQ0Tr%3KaH^*)k}NN0j^d4$ zjJ_W-`iHkvE@*q1i|hUhXvzKa?i1#s0G&&T9EbX>7@U;TTzc6kVFs`CH=PFi9Rz7Q zT5HmZgX%CE08WhH>}beuN9N4beyeUsv!q^<^Z_ZCzhN_IT$nat+FKYCmg<~C(d^938K7Fn90aNF=@_fyCeG=vl~5-p7Za%u z9*jz0xxg+KYN8}Yfx_#INACx^F4vVlVcu~oZ%zEG;LU=~ag~4*n8u`fJBX=sq=ONeBp1Cs+*)HiiIqH#;^62F$%jUCJ$o5EuXG+b zVB^?(e%SX%-idx9r=Sn9F3M4yn;9-kEMDxegZskQ4P@%iEuT zyIaq#r|tU>+n;tibu&un!@GP3!rj2z%8WI_LnvzBWSQkBItNdSX_V3BQs^j49N!8~ zq*tj(ZwF)BMkk(%4@qZp(vac~dPafdy%YzTF$35`$=Pm@ANNz)mKPskOHNZwzzQ5G ze-~_QakArPt`q4AwcEoWxxhg<8b{_+Yq1W?$|8Kz*ZA|;$PH{Poh!{PslXc_tP9dD z3InYj!GIUXPvX*OeNoT1yfeNSD+HD2QNO&jTACvdl^02jAtcQM3UJgXGvQd~4pfBa z^XPt$x6M{2WY-U{7g2<{HtqrPkA2NH$Od&CH`op9Mu=XAzw4s-E|&W|n{4dd_fyBF z>LSKf&271#ehc<_BrBhZ0RT+tWbfb!Bs#g$US=g zL#tG{&hdFxNZK8Cf1EWNr5EKrMf(BqMQ7LwPWM;5P*A|!h@%(|;V6FxfTsxlKCxTK znv{}!zPxf*>1)iF!Bn?9pl>qF>tTbt*UMqhZK*O0s^L8PRcK_QuMHMEl5H4iOxq|M zl%*5s|Mb=B1##Baw!FMZo2`%z_yHjDpFbFe4!;~;w`y)H5KY;oE6miF>3TK|Xj4=z z(1O4oI4XjfjrkU7j+UB_LG1-){-bY`IW#mlrWxc*s7KbvW$C*1p~pO4eioKv6|6yf zdFgJ+;bChpJzy1o+~(Q8qM3i_#p_^G(TF-cx2SJ8oNg z=6zc@AKD4dkcy<3z~c;i?Uo3F1!>m#iCoxUo71qKfxilhgm4Y<$5B6}Haf|F)+6qf6|GAlw#c;iFLUB@4uiUGkcuR;#Mpn{ z{eqpIoD|NNC+P-#;2^)OTBd04|GIkQ%6EI{#`~9ZIdLHwPK)y4wlWdwo9lpueDU>3 zxI&J{t-*x-)J=s8WgSF|pqVe^IlVL8`L2!Ifnp|txz03IZzn`RK-dyXQESm58EtJ@ zZ)M5K)v^0Wm!6=&?&l7;<%KD7fk7y3rM`@Lp4Dw~+`wRla+Jr}3`UUGbv*t*3!>;GA)bu3DZsu0>osDRiigQU*a>#vuxwF-1gpCGxpq^6IB}PCG{dOq( z(Fx{ohIGK66*JqEH1aN)8XC~bfHLT61u3q$J+|XcNdJ+a!m(N98-)S zNB{vR#%G|z;9kA(;BheAHHqBQQy7+iSu^q1ew@!9e}6U#!<%Q*hHF~t_lY&n!@H$^ zG&=eyDQW;Z&PljRfs{ep;(<2{uOBBQbW&NnSUkg`a79nLkqG$U?QUt^je_=t*>Pv;bgZVr=gwJ7|W|FI81;g=r>Ok2oEpK<0{P4~LvHP|w-{R_7g=Y6GZ z4(9|pmW*}xmFH{VS6e#U6!*NXP~H-XUA9!#OG5hWnbbKW=pjezI^otbip^j0s`(%JbQ4i8px$J zR}3R(pLct_SVfT$G~<8n@!OUpZQ`18A|gA@)kFLwA;J8|+Wd@|0D+srMqCllFo1He zW;9;4;Jnw`V5R(rW-)$Z{SsFU2j^GzZ%HZ@J0;7s)hEh!Z2fT)W_yOdGN3M!3g4RL zBB}X;B$nSMoJhmq6GdvH%2UhsoRoa6Q-KI>_U#5sv!l)La4QK6*%Z$#dFcGa=2Lp# zlea)TbrfoRqZts+sU!;<5nfk&&dkpWaf)NcG5AWZ1b;ZHgKAzVPxc7pyU(=Vs~Ls< z2EcP5+|bmhG4*0AI$D_0|FCxz2G8Y6_hlPC9a|(O_@&U@QIz29U2W{2D9kOp!!{Er z?TPVLRIu`O1TgRoDfjUYvh_^1X+Z_Fu_1>f`Ej!aEn-JP{5Ie?V*uVcRE4Mkeu)ku z))Q&sF05_fjs$$ZPQWbfn;g(s=pbUHV2G;`MK9dXIK{B;q?>3 zNgC#3-prg{Un>yubzz!ElMKq2>4EMbs;u<}7D*!PuDgHjKP4P1XvDq<2MNqOAC{k) zs$el!53CjOK+52lgNIaSG>9SimQeLyeVMNHh?jwbS9$jG2ieJ+*d@xA!xG6=<# z0{D3)2V%t3PgiQKR48qH)y+hFZSRYu7$=C47#sp;(o$7fde(UFANn#s{NPK&m`k-O z>IV|W+5}SZh5{3=aP2v;jMEM}Q18KImYpFJADT1Av)gBL;AaejnaeBqz=j+1Me*HP z@3v=s!#^Tf^jxKJ1DV15gN2jI_4Z(Fh~~zy_vTyW~vDwexk93;8ih z+H+_N6u9wI`u(i`FfARHZT*VZ$8BHa`gOePRoaj?#=7Z~kKbbSzLK|8A^L+HIjc_Q z+7YOS>SQ_dY=#PRx}0Zu`7dDLmzem%SP}+W$=nW@|^!&@$vjDps6{r zDL#a?p3opzNsLNK!lL_y%>E;a9`7K=`D|1=i?t3%U+mlH@S+kH8{=OAgkHS-X!itN zZNeWpJqa%_qJD*dpO5xm0loE8B3M5Yq*owM^18(9lq!)=we=q0@O!!7`wEooEee`- z6xHuC0z!>MN&obVRHGn zTqCPbIj?GrEm>+f=!<&fRY?7tCGmQ99%~nLdCr%@F7VL1do_L>Wvy!V$2VwPG3cff zi%4yAjjZvvf#&!OHVfiD$-5{c#Y14bfspKPk*x~ExqFw8Uf&<*=C6PzwSnZKMGMym ziy~bK8+ZT#jy)37Fd*ZuaJ)nJa8839af_lqn(X|_l2A!n<%hepAs43F4*K?byGsYv z8*T8~$0_TdzCAzkg7-|K_T>Efe^Z1hXwnMbof!;mQ(@h6NA$d{7^tvmiDM*?ImY2i z1+|czbCW+rh})yFg$mW;Ruj#UPP8xJlS#Ps@~}wcD*&gS?8L=Oc$zK0Zp!%@QV=FN zl_WZP{6{NlwjUJXcV8Q5eH}&JUY&+=hCdDrdIo_=9&RcNbmoIs-odP}$}QV?OxEKmha&|0?1`6+N+@HgF&5R}YQqd|d-=Nr1E-IU^8AO#N5*kcnkDtuPOEzj$xZ z{Np53URONxBhNM)Gjny zyh-=hy;D$m{#zjPVO_B?A>g796fREZ5fd5Zp;}iC5U4%D-@^2e&<_sd1kD9&Y z7(IF%>>_?YG+KT(6Q_^t-Jhe73=4-9S8sj{fxoHr-DDHFVtb!396W5jFYK|im(F(; zy?uK~6$1K1z-|^tWI8C`^f@aPLBkixO9AB6!%o-vg6HJSk zXWa0?+xp37IDZY+7YD2F`kNx7pa6jg9$57@y(s4d5 z&8-d7ml@V$M4(Dy5&F2fj&R<;ym}zSZ)^;%&N=zutnJg}JYUrz&sb+lE?kPs+P_Rn5xMXdZ3Oua3sR$_-FhLgJoos@+ zZ~NE`IB8wV_gCLWpOY%@uP=}X1#Sa>R|~uX2&VTkXdFmCD#!V_d^wMoA)M1(ie}PG z6VJ2Frn=?z^goFRjD#BQFSTXATZ$zI7TqW_o68Phy!qtme zx<-znHEDClrwvkqA*)uqb#@pltJc zO}}zQv|5Y^I8|ji@(KYCe8`3Vinh$jmRIEnw zY383|`5i0!9XQ4?jgQnVMq92V`L@bxOUQiH+;h=W5&VwhAmJaYaiFB%fCR)<6?_HXA!O#x%YVE@{dyrU0AF z8_&&9QlgQ-szdEQ1$%tBDd;js(0K@%E9~AnQ9-ULGs?QLzhM9|K+eCmg+q9NdoE|I z+3?NZZDpd)#$YepRw_~|K;6GDUJS2FT4{Lsl%rY3VWVO|(cS?BR%?V=tZip6*ryo3 z$H(f(37Rty=hOdKz_@^J#ewv%VpSFKqD6#*Su|oePSU;r+$IfX1J6%|EpT!VnGReD zYFoB9si!a%kO1Yf-J%p1DoH)^6a2jOZ6T8eHZ>0I)F=Z|I%^AoJSMLIhur^*v$qP0 zgNeET2MGkX;0}XJa2N(B$l&fFxceYskRXBJGWamKy95secMGn;U4ugc$$tOVR_)8y z*8cm_x2qqzs;jzBpL_e<+s#t2^WzvU0G+jQrGdA$0x1#Ue+|d~rBna^ZVc09{#!|s z9o)`V47DjWz4zE)-mx*_*P5H;soXFdz}=)y?V*}+pG{c6QQ&@{hzsr!#DSc8Z40*W zEpht*{f^=IxiM^C7aGGe%dD3f!P7EqsaMO!c7679t(0@Kd`JsC;fdVRm;R9430Hol zq%sUzU|BF};|n&F3D2w;LnaI}<-G%qU<0tdG;3rX3~^YK0WzNP{{T*D{;TXL01nVb zr+adG)z`$fR7hd09m8R*vu=PJ);d6h;!;05yPXEf|7XT-|7HS;*!@F2!P{X}`}T7~ zy&lD|`JH)}=fwJW%IQWcV*3Z_6nT63c$R-Cs`_!duIes#hwwN4@&iRqeA&C{8uz}{ zHpJ7})3`VZGNoQWr@U%N%U%8{#Tx`WE=qW1o>Mci4UzQ!QWfHdQ_(z{a}_=5zA(3nryl<$ijVCRRzr zuFSH_3&+-nR4z2-)ROnXjriY z&hPtZ16TFdBPgaq|)Nv@S@QA+U`bM@oKN3@pT3q zXq-*L)48g$7qTE^4J2IEQ4Z>%Ykt4Tp&HutmX{+0kQeAMv zNb_X}w2VER$r9}2^YL0y_O|xDr9uCM=7cj4WC=a~3^onBrk>*)wirWqS<&^3aK7=phZmZ?|dB;p#Gls|W* zNwt@SYhDXjw|ST@l{l6nSo_mj4GwOL%TiLeGlCbo_jsu^px)4bfZ&DbYFo>RSyUxh z_*ahQ?!Mg&Njyf0SZoUbbfdWYi%`Qp(uT-n+KP`nDdkkYS=Eq-O_Vb?hcjnDWJyOh zA__@$(s;bSg2Csz9ID3a4fJ6sIxGj^>e;d&*jgojun`%(E(VH8;(@y7-`Ow*TDW(q zoJhR&KgeilN`gb{Das|CBkm4#u8iC5D0*r6&6P&C+TF_x>0rX`3;NcKKf zfjd{GhVmw6iGFj9XcV`0#C}!f?-0i_f|bpm>uDhrMMLHW%LLlyZf2CK$17jB=gL;r zOV;Y`DQVy6;yIcyr!LQ~gUf;q7IJu*Q$OitfuAdeir$d0|8+{g5UoId+G$Pgs%%`J zK75q~Z==Ukx}vT4*1VjVi8iM6!Ks-wx0U!a5smlADZ;&a#dNhbMKHwFg1VfwB9S&@ z)=sCr`_+jXVj$h5DUL-WL1U=HH?|sBx6i%G(S+cU+^7ltwN8RRbA;0!VNG3e5jtzT zHL9x`)iMVoca?hN+@VT+{cAO9tLCpa_`J~K6D`j*z}6C_FWi3oSjvDcM7bwv;VM|| zb9xo3IWLZXIBnwZ#a2EObH!>+eo0MBQ?rt|y@q#GaW$#*4B8Kt%sYL4hvuO7Cg!v0B7j zqAA{GD3ImE4_C(5cgKM~zcZRU1;iY5vGJcI9j`ibl#O=x&4OP}x-G|M{B<-y?e*BS z9S@Vzg46COpZLv%#pp^KhsCqw?$Z;VGIsoiB-I^+8#6F>b)7mo*(+*G7CHU%nT5F7ih6#8% zah_ko-%9fAy~YRK4|?xj3*0rkQCv1Z#ER|e{iRJsk|hPFiwy<1-qe2h2Ot^C@S<*E z)kjCksh6ELfj733+$i~?uNN_Ax7H8p1-{=7i1PN{3F6%@;wPHKw8Nx4zqx!braF4A z*6ESML#6)|S2B@-Qga}kpVNE!*Xgo>KSjq^VH3?PAyr}rSk~o zQDa5wrK;A$t-{~8hPDTXs;*BDN!v2rWhez6PBnyzT_U;4`z4rwDG1pHx&^D&D4zL` zpvS7bk2XKH*E0P-(~!*88E7(g#I-2%m=6!N?A$!rRW6(=*dXI>HyTVvD-Y}cBT4N4 zYm)SzLFWdBUU86q%Uq*&mrWu{_{l~jPl7#l^X8=IY;L*hb7kgPLw+ay&65aBK$n15*Ny}&{&UkI-^UJp zhrXd68&d7#7`0+GWxLJqed$)zB@Vp}A-;W@51wyw7g%)Xll3K)`g2g9 zSmMb7SgauI0Q48AVcUH9kgpKM1JK1YX6nlN6IBpG%XbQBT4+=u9aK+Y(tc#FOW`Yy7+_mEN*TtMr6eoB^al^xX$zgX-8%*YWC{ z*~aE?*TNtY)qocZo7`sgg-(uLES}x;G0Gh{L1@~SQ|38(@%uHC0!|dP)IArGn*E0O zumRxwhWkqtX6xp#PbF3uvFe_1eb3@>f!swVDr}^4d_?aS58F z@*phO8Aqq+i6}&zu*fNVu}w*Rfr-Q45`7Ft{F?bZ=3sp5?ZXn%W?>sH^NW~NaWCga zQ{mQ2uG*&ele}Rr-!CU{Sfj&|tJ~Io63UoT!7`#~{Si8}>NAiT#uWwlYoxXpbaiQD z`>Xod#B2QQV)YuRhZ0K4<|t;1{{`eiZ)vu8@b@1;rAq`+OyB_N`+_O0u^`=Vri7M4 zW{oYv57pm{IyF2 zl=P(&{?v}HSd5mNz37J{O|1dZWiw0{-H(szI}0NCv;l8ln(*YqOe!CR%gKe`mWs>{ z0=Kav@xIY(tF?8mto_;_&V@_3yzYzd={?p?5OtWE%LGEpX`Z!$>ojOAdpaI}ijZt>G6T8OWjnAKCx7la%)-z({G@XHovN3+Mca(x zNnfileJz{X+C9xN59^BS^caES-R*m7J>lUnn0JG6!^s0XeHHtMJJPPl3{jo~c>6_srRg*{ zEm^A;f__(F9SqO4^JqY6n_8*Jl)w_1vT0qkZDXqvL!-Cg7%JKts$t^xbY?N6PTpKA zG5Ba70R~gvA96?XSA(Y0GG(hx&D=f)3aq5Jp&>xULW35sUe66CETfc1P^EgbRdV z5NL;=8_>P^%O+W!-paI_Ys0Tfs6EybhVKPNb*@W{*){gE<*t+d=mE#)ie#j)O*S@4 z$VV}c-43F^zAs7F(#=HDeY>w|3+P>QWx7_^nED}6O~B5u4@K@?8q~av5C(rlhp!3M zn(g_bf%TunYIXQ`rWEo*)2z70$(BzIUnDBhCE}kdYw}GcODfqT!f{dCsP)7849K-Z zkx<0>IeniJ=99;Azfmn)JR6s<3sZAFV5{3tpzl%?4-a^WkhaM7pn1;EIW&qs zc8TQMJZyxrE)P1auN+`@@r3(QQ0JU#yuVp(Ie6tq9rL4s%`G)e;JATP+ya5}Rz^ng z#bNlZO8ly8TKs=E>39-gUq^H!;R6g5bmAH_grRP+MGP6V*)vd7iHpC|=JcH>#PCc$ zHtGW?PieX>d=Id2SH#ok_lo~UikH){+U#IC7MYlef7<<2CT|&TWP?6`VtkZdMYV}{ zY_i4gM8 z=W1M%N3h{A-n#$XZC3EfA7Rupq6Ll*IoAIc|2B#zy5&9ZURW^{Ai8;IUZ!d^!nmOP zWPsPii9#eFarAnletc&FVLfVY+SuJSHxqx}nXFtDRxOHJ{kA&xk3$+U8*dN0g@x2E0 zbk(R{qTvMXBmvQfDgpv0)hE?}OiP~@%+x;5)Xj=NBzG0Gnjt=?(KRJ8ww_d^4DnOD zm}U9Zzv|FW$+XAbvY*4z&*tz+7 zcG=bv+WtT!e$zfrX@VLCqlfcsf1>uNlZMo%x7cItC-{EFX0{{tfp|S;*3u>_ycROR z#cM8-Un!;&R#~6&XUNCFa7?xP;#b|7`}I$>uhB4&tTGbAZg&Yf!&@%cW>()y4auaO z?UaAidFT;4>o1 zLv!&Vn~S?qH7Z0}d{Q^4@;`r(;eT7d<^L4Pm+0X(T~&t8x9T1zH@|)Y)az^cTJ@vD z;!h`jzklgJ`l8;vq#Q9VIAoMYH?S)?4ND_3{|CTPHRc`p8W^HK%^@w-Jad=l3ZtHg z|KLYdK#xO4A>1tjF3m1yj_0uNhjG-rcq7q8_iGSd05C@g%dQ?Ntc%|y7*8tn@A zClaqp<?t0=z4JCpIjyF_5g2kdOz4KCZ@XiJWQd)u*O&;r0d_mx3xtCvL6lq1B0jI{pvo;NjW?QE$P z=r)A1XoqGv?{-Y7nYl{#lE=jcYLMx7Q@->n$#aii)EXl{*098gvCvNj+p%;#Ot-51 zH_^F89Og|C8#vH3$k|%R0!qoOcQMgVI>2*TEe-r-w4N5WwLO!$qwy#?tahG|I;Z9E ziNPiQAK;q0+(3+OaPkzm)s}NrCv9o@VTMd55`uitkjgTHoK8iU9survvjzCme7@=k^QU4e*jL( zEsjKC0SwOxfnfbzH_F9k$J&@%)aC7DqPTt0WnGzv0DKSvI?$>uV7ip2a3$K!%on>o}89&8w z!k1O#`sux&UxmbxiE_{0V^j|a9)A+~$3b)0Bj-s;?d&RG5az1dKsH~ZJ zkU?+lP(DsgPY>RNm**Pgpi=X*YVc}&1%@9@89rol8L4|Jb@&ld17Xl4Ar!}0r-%J$ zZoqig-WuqGJ*y(I^z?Wq$xHsaTIEXw1Ngb9b zE!G_RHRqC{KaDO~CU@CfTkzZN4?o~8p|^BAcV+j!S}$)THkP*s-ko;C46s1l>hMXX zVa8W!nOjNFr5bZjlANDt1v+XCLtrvy%zX%@{wL%82P@HArna>C}%1?F2T*Dh2&PVoNMfFnjQb_tj=A2(7t_?;1su|=aV(63Dw z0Q)wbII+iF5H)IcA(aZ8X^M+w;y16v-#yfxE;t_6rHX2n4~V0b7Zs6=^jyEMDfn@_ z>|Wfhs_;CKS2Ip$y;MF0&RrOX)p$xlcJ0}LYH|C}wwlkC(2n@lQA|==$3i3H^>t16 z*BDy|G7u139c(Z%qI88iZ%jT^`g!4i4fKRyD#jfh`H*9*WT5k!UVKcGTIm3bD5F$% zSwXioBpjNAoSWQ*u8oM&_2VDLsxGCk`LxKy z=y`4)@{+whNyetuX-y0_#Sg3Oj(s&noazME}WS)oICKj^Tf!Y1nv6u2h`8i!}@u;S_pDCSjlau#T(oR^f(r`#`2$S0Zdog1W>bskWM93`V8 zm}~s=(&AztM=6CJw}$3Oic$TKrc=e7?GOxP3J0zx|L;6YStj_m|v5ule8vJdW8xufw(;(d8Y)Kdbi=JQO z>K}?IH+<6QOg~qF)5-QwQW9`xT=2 z+cZW=?+Xf|whi%zpNTInnecqR=>_*mEc1^HAQN&Q&)Dgos&aLXXl`}<23=_J0T!km zVS?I?M=+Pw$`kk+*E-X6)eUVF#W6B?9uG$~;jV7C+2|r$Vyb!odAI!EAhDuyFQDVF zfRZ}!R9nbwz&}bQF&0y-lXS%o_8 zCJZtKJ#momDSAJR!%2wx{pm@+Wro$FE!zc(lGD0T9xYK%&$%<}J=7aTL043xeq<;r zU{=XR&RqC#UyG#Rku&zkR$ifNYQlpOvQSu4ZXg=kdk4xlc^iyBh!TFTaZcW=PVLM)7>@n6>D+ezvr3lyR)qw0 z%ZbEV7BGpnFLW%p%Sp?TrHJ2==^geLTecmxCD#VksLP`3Cq~ZK2al0|aL;uFl_Sth z6n57OFUHhACrLYv6ru3!0m9HpheJB6GBJyQuH-J?penE#`G z?I>x*RZE2$g>1Yq{=^f0+k(rS4wrkSkH6yJtHp_phsUr3bdsMK=RD_)^<`s+xC zJ8Pk?A_nVw(l@;3qD&%s>>h9MyJ0D0wdhGIAE;HQf5?(OG3R^qVka&`)-dDE!#`D#luQJJTaKYG~n zCO-WG`&({`C$b>6d3VUhaj~1_XW@=C^s$zxtfm{1wCgqW!)XEqTEgA z3`s~CfS?(J`z%W@zV8-*5~o`do(3_m6o9q_&+mXSZ6A3lVDT+46%I}Q2NvNI8NuPy zGjCQ!An#>3GpWNg?s$qJPSHTGwP8<2N-fBUv>c~`kn5Xw-8uE0F&-ieyabopbEeCT zNqxM`$pn2e6w(qi7_$?sJRBTOvim@iG`g;1BVkpr!?il(Vz&d~9|7ezng z2*_gY@oU3=`2r(v>ZCJcr{gsv&`||ZVl^67C`&8wSlx`amAH^t`XheH4>ZM%9)pK?5Myx+vk%nOvWBb~op>4wadON%mhdR{f zmxynu(}0AMO3m`I5MCB8>jM*R8)6NBW$w=)&QGCk7X`aP0+a{F@@D0LyAw&yEc1*+ zw>lXxO2dV-a0I}D0QynCu!8fbjCDPzJ{EBlUmS!Id&82~>qyWT=Q=EM?wqEXEDNRIVd72%>L z;VL~tJ)zo5Fl4UR4Lg8#xvPaX@~vW% zlJ=w(g!jST5Ej#cuioXUZrPEbU@K;w(<{m0AofYeMoRq~By<8oTcJuz%z-2M?A13W zA|qnui+QOH@?;hHODVW^6P~7nsj*P_^NU>BIDj^);*T47%my0N=a6EwuP_Aq@HF>j zwlGp`!h*A*2i?8ZZ2sjxe@-<%rVk@E@h#Yol%v#Wk{=raa0@!=L%eTF!om?iZUHcM z;Co>U=*UJA>ZE?xA(5Ap8+}vdwc|SU?o5H=$Va2*esS`53}X5EMxGV2jI_|?PBn*B`G zjUqk*4@>8bv>^Vs`D08r{HtMJbpOZ1{NmLUc2)M23E6tIW{0Z=&FHW^v)f{VeXFGF zE!QhT7y_nNHlsF9k&J6=bcieZMiqE1-ju0#PpymP&Oj7WJqB$p;evY4a(fm>WKf5y zXM-YD{c_%w!wGvrFdk{0iC5NC&a>spvj6^>B|F|9YXtkVQIq_Kq|4zf<0G?`aWuOW zgF+?6Y>gUPF@=&*pgT!U@gR;QekEMbdG9g-XOB*o@Es8Dh1?=pMQR2W|NU@1Tj4mh{_<5lGdo0QA3g3TmY8%img}4Jz05aE zx#n&?x5KrL#Zz&-=fdRL&%&|9u-u}34eoQF0MPTUC;~e~?-m{J9{`q7kHC4upA^rE zVlX;(c)!u4TIJF>IspL_ts|Uj9;3d68+5h-r^~X607-7;@khA zGUWeH2$k;bkM2y=l6C?99{VZS9~%AARU5KNGzBg-6iNmQy4#hDTd|#%j=9>hR`}be zn3Hxde&ks|_}pUds-`tdTnbjHpz=VYB4$6_Ln8>f+TmfbXi6Q@Fk(cAVZ^bH-(;#D z*w|?LevpEhaXqK2Ib)_F)i`ef(T)$@FXmRkkBS-cd zoP(C&RL(Qu#j$MP#2+Q(=DEtB2gU@ZK_ZFf+DgvPFVMGo-&C&j-S&d2%(Sq}%bC*pqm6y#C;;Ok0l0hmrwzEG5`%6C9AmK~FMDc0Y70eLA7 zQ^)BZ?$AoA^34<30Pqz+^gD%@ZD|ZC+o>AHpHz^!IqFdR@i`ZdXh6DeK`leG*{h$P z%XpLWqcsk-r4a6X?BO&oQK0vw7F8nYirK)1>q>LFfPNvHm$?0Dm;Uy`*}Kb$qrvqg zENRq~2I1ihmQDS+1d0x+oT|0=8m`2XST737>-PTuq{4WmX#pq^)dj|Ho{cdX9%wz2 z&2kT;%4TVV@Lnx^O&&SXP^AlRm5*sv7+9jxPwTpGIg0lXsO&OI{xNqx7F?0>rba9; zq(7Q%u9jAYX5>p1&1HOtRe8$sYRm^!w_hX5(u9&#HAhHU=PJ|B#Hk*jG))T_w(YbMZLrP_4(i; z&j7LlA;To$&XZtLFj|*TS9`8jBlxwFoS~QIcjs<=;}pcz$I{vlYHClMQJXw?grgH_yDs6z2JH7hLX>6)v|6D=QlQSs7hSR@~o65)Nn6 z$?vZjeP{}l2K3zjET(Ms#b(YJd$zkKo~u{R8jCC1^i^JViG2My-gGMA^yUsYAGA?H zjRhh{1JC1~@2W-O_%XA2akjF*t@7t!kY!ToUeA&~$Ow{qd85FjGO`Ac8FDJ>M^q!6 zyu=8|3r;n7?@o_Pt?frD6;D={VKM%)p5GQPr8y|zjBbRkNZU$`Qyl?*hxzcF(NjdV zx_=;>!6C7-raV$-iTEVUFx`QeG44jhAJTIilM@};mEd5yV;4wc6&rP66-%cia3_pL zJGqrT5o(>vxS1kt-gG~$HDB6pkVY#*jZ0vXrC1a2b*=LXUPiIWc3&O4Qllq95J`)s zyUp2LG;qJAd_F-De?KzF=;=c;POWaS9P5JeHTV|CBIp$Q={Hx4?In^cK}nIeoeAl$Et?a=u3VLSzGEPm5XMhW-3qjjNcwb)Pl>-_8d zQ~wc4{S3Tf%g+)iZ^rr1ZWI5I(m&P1i`>JHSM!Wd;Hz20ln`UtR{{S3D^r|WI`4&` zZ{ZWhQ79h#-}^}!`%Qud>&T8PJU}Yz6}FO%ikytKKt+7({UPz(&%RGFsoQJGY;PQQ zbSHr+ie8$3xAjK@>AFDZ@!!oCVtTalT~Lmkg-0`5{sF>b*vlCTW7w1bx4I+D%%Md4 zkIZw23vjN$FUht5o177J++AQyz!2*t2ou;UfbLH|ibn@p-@UwSTQvgJky>ABvgBhAV zdN3>?{V0Dq>+4Mg>uv}!US#ZYj5863V^OG8Kb|W&IuPnAA}A7BKztc3*xqwh#YfLR zZ5+!=XPt7l22(inOmo`BT~C8SI?XOPzOuwWNC-MK%IL6sh>5;iW9SnvXTM+% z2PepKE1%f%zLnb!XJ~eUv!<5XVa_s0`5Z@_zf!0kjF{)=y%Vx_T-J_l7J_*83J1EBi z-tT(l5sEd6NsSLHj%4Cmp{ayhp@zCpgj*O+QC2Q9lWZwR-> zl_r+S6&1_88`-u13jIPWY}ZVw_{8MZ^Zk<67&ES>7$xTA`zo~sB4&NFCfoWN1bmVE zOn4kpUAsj>hRF>WJM8lL6&U-rRrD?X?!); zzk1nc8Vgx<6*+_8dy#$%o_9o2R>=zwq!dKUyjZxdd79Tsv)VF;{W@I*a3ccwN1x1| zzF;!v{{sY7l9;JtF(cd=kTg)Kwe#cm;fmWs9Y0ohV@6g^l#Zp+Fl&DuYjdL#Hz+;# zY1vDDTR7y_xoX=euyyug37gf?*9D;k?zzrw6y2T+Go;NIHv&8-WF~#Y#OzKkt|^O1 zhN(FyI04LTK@nKJx?QvHxgBQgcW_@%uU2FLSHJUq{MA<3eejHST*&jB+^nSS$h}Q+S;0(c?jxM z2*C$3w*`0Ah| zwP@Bs8i={=AD{%>G=H8ukyW!OOpVEGeOI96`FiAQh8s%4eyL{U&#Ra z%iz+d|8!b}kf(pbue}G&|G`7 z^A8Z&wDyZij7sbY!rgtPi2;e>{?1ewKN49B7WEghAp`*v|+|rwWkiJ zhboUoD*_~^6cmmTU?aSnO85ldt}jnVLe42TAn`0H`^ojk5mjmF7*G}`dFvBbFQ+a6{gUDFpy~lZif?NX-Ou+taIy{ zy>{l6HjPfwwW=rL*w=qNvOcfOjXtD1h0i(#)U{a^nz`JV~zbM!Aa zUB_Uk^d6;k$kYAi-`Tf4j~7C%pJLb=UF5TehQ`cPl~M1nZJ}uRUFlPrj~k!6s|O3N z?FV4)oz3-24nZd{4BzQhC+GOw|wny$76NhJ_ zko}i33N(`LTX;;ZNrQ{sY2nGbLVv4nZM=P@F(?GJ{)@zf%=&qYm*8T4TiDur_-G*& z?L@t@6qdFu{Xyq;@wnrmQhD*5f)DMOAZjo%C=?{=0jN{`HN8VO{H4Zqs2tc4s+-0f zuUFACDr&c&GEZe=KFdDP$3hGZ?>$C_*(Y=f$^ZeaF^^s2O|cV%_)cMh3jLpmP)b5LE64V(ZA zbDt(N;t~zW$it+;9lu}MYuBRacNT<~-(jxW;Fdrq8+d46YAC)6pN(Pd+ZVK~bQ;>K<%%3cXLah3eNvsZ^Mg!eqs0F%;N_ zzP@u)_El6)K$&`+TCV)rr`r4w#zVv!Y40(^X~E?UL`yzBHhAr$gm~HIkF}_lYrJ@U zyD2;&q;Ad{vw`+;{I%AgEG4(^^kbD7`APU#+|AaW08QRT{ZxnT@yuDZfmSO`J~0e! zwwiEdlxoNth#_Wut`|&^YtVaB%nn0~>0EYVVdP?ct(~NRZV`OKSBCXhrF+BWXi9yFk7Zx`n?XNuJ?w<_bwVFbgRcPrJxjBbWlT%t%27bSL zrSc9i-7GH zf}eXgZ`{+53QGK-GG5qdsSLOXa(cYkXOi&U`hAH#USbZ^o@7=m%S6Te6EKg_ZOxm- zi8G4VF*<{^AiAwnI3Y90$dBGNx8QqCy`UM_#*63m3Y|cJ|0~gAYyaNmPlGGhVVew% z7$N=%-fT)1js8QV4!Pekjf<7?5_79b-eMc-#e8}M%04CV(DPis>jNZ6c*TupFZG(d z55J$Tc?ML(-Bw*5TU8EF@^yp$cw6dMMkST(Sa*_ayet^`Y2HHwtlz4#P3hJo@AG5< z1P`@=a4s)rtDhMQWD9y)zZ;nr`cho#Pv>ebW?r#)OTGHx%}d9((N8-N@@14wa`Zyd zv2n7@(O~k4%Ym7D4OIY|qI{Lr8Q@S0Oe=~H)v5FYfpyivpPRv{sv~w%ri~HYew&M( z7v`q!`rjqISjg^;>+(Ybdm8aAAT*t8K`=yK*k7Z4{`R=aT@OvmZj%~?;;w`Wq7L`B zsm}yC34ijEgheYITMq)9C;49u*mJRoc&#vdCiKI``~#dTijU@b;+q=z=q?C|0d*`Y3RpA1YNJH3mIywpntYby0J+DpT|+% ze2Mj4q{=N6rZu>_DO2$%$K>B!_g?Mbn+_8Es&Bj?=F?m&BBAxQrsFHV4KR#<{>5>z z>{Gu9U9k^|C+dWAo*xXEky*4Xc;f$DOtY@{6+XLgKrM8G0G;w z{BVu5I&XfI$A@~8@X-CGiQFD+ZZyYqMJ*3BAh=2L~^gT48IhMcfC(D4upiwE10)!qqx6}-! zb=PSQ*Wtb|$FWY8@{q(I`u4HmAAp?9=;v4aPvf51Bc;ii&)CD;JZlAagGL|ab-RK= zb5_nQ2dd@mp-1f=_@b4;MHO8WTzTUWc)#NBhvhi_0Z`0T{{iTZQS8$V0?ONCtq6x5i(64yPr6LkO4GK9fv)4)z?1=WO z=BRA+jT?~bXguc+$zp4&9g;5XCYv)dU6cj>u|f!~#`OA7^7<=!U|q@pkrwe^yM!DO zTv{ix!&-i{y${tHS~8NJNs^8Q2rim_7ed+Niy$tcAME*ut}Xxdbe1hT%YbK+U5-WS z<+Y%p?x^o5+v;wYO5fA{d26WC6%wgw(Ben#Y76>)FSHUm7$64WZDIRp`7SnSr|}g< z5OH=&;2CstwcZF7jQ{yJy^6z}kU{5eY73v>KY(2e<_{YCx+#PZdyR45*_w-nlpe;b zuv?%_z?@>Xwh?vQqGOTV|NRJcP1gYBj^^wqrszwI?N+H}Z?@V99H>V_n{x{An&nQU zio;y4A_>(dsW%46(b=B-M?<~&T*oRzg;b1S24anUn`YaX4ILDQ|0B+sv+$MH?Q+Gv zCiyj-qNy)$Y=-rtDX<_xJU+T|fBpl`YH0soszJwsF<_f>ubp19ghv&RKSrVUc2DxQ&>mWjHd5YL5@`F*$TW>IZk^phgq zd_US2%2Vycw`nJF#+UAujPlH6o2>66JH^6Zo^vxhAH2G`pZLUiu3a~*HmAc#efXVB=-4(N; zn0&qxispigxVMi~cXmbdvMkx7oDN)i%gUF2KK|Gpu_5-E?8L+XT|`N}Dr>xOwDhO) zoG5op37G~Fz`Q$VG1l1adfYh6F(BuQE16F-^) zzFc0UTr-7`jVoX?@wcP9_}y1!;JzPs(*8M)yudnBfS$8%-P9+KjvhIUmuUk1p&THJ zp@r$R8k>aPVrQ*>;&Ntb@yuX`1MGr&aGuldJyn4rV{K;aaebGf*Whe4H{qN9_mqAJCc0paAZAre3&8?S;!EGu+}yuWvei_oUzU z?S0a|qgPq~z-JT9d_4+I*{J9+s5vz0bu8(nDxmCWxAMaH55zts9V@do1U;LZ28ob& zCebh}gtx3+#^l_Ul=9E8<`nL;poOm+*ru(7n(->B#HvQ03mxIzRE`h~5t!djIgc9e zCrdoNp{b&=ra0m^-iOW3QRG;&dwAgy3pVrs>9%ya*e?*Q$qUh*|DpJKneRm$M8Hk_ zu%f!v-r2GEYM6;|gkm=^p_!!S2xhZi=8%+>YfN$G$97*~y7CXei}satT1K2IHhHfN zm&RY4u57UeBv83-8F4<(FT>Ch#{pnxa1kkIZ8Oc@er+|Cjr1;~+V96505w3$zjSFE zx2V6oDMb*q&P{wBix)ZzYB0$DJ_n5Y=~z79{KwS6%2&ZXT2s!8b94~a{p%-#ybG>i zVPXW#d-GJ;_=L>!gPA%$5-wrQVhXmg$%)r%d(h4j4$x8y<5=d`qXfjccDo+ieg2+Z zL*~Fb|CN7AE_b;8_d=cKl4Z<=y3}85YesViGjEhg&+kDBP=~t0j;`ir{eh@mBwa)h zzGUTLWfu|yWM^T+F@_o!T0bOs#QA7y#nE|Y(fIaF;X6kg?g5K-To0Hb!4>N4rcUR# zrg&xo1`r!KCGVj&Z4kECofW10*99AO=nJ4uEvI3ly_Hb*AK<`lPxSww?5cw5fSL8d z-Q6i#w75IPi@UqKyE`1*wYWRyaBz1m?rz1MLZS5bpZV|8eY%;l-CcJCqB0GmhkwD6y|dIG+WN90Z>52!{RA zwlA{0vFhYb6h9qHDEY>2on+%jN$Fg(OsDene?OyMMCkDg@UG`~h(A$jFA1A23p2MnCVaXMIAHSX)rF;FQ}PYD0h{0M0sjJ`m2hN7 zm2V?qG1Lgnx`{^Ec`^=XHg-H4k{{oBk%qK;)DqLAw0KtbhT)}EQXV> zgx`UrA$}B+@Lum(m{`yUb|~g1athELTRIwultj#)R_pX*Skc|&eR9lz3^;f^tH5_z8axE=#kcQC37wvu%4G65n z4~gKM@%|pU6bJqi&Anu9@?h1dhVk5Jerdjo-EI zf=WHhZYxn`xKcLs&(3FSTxc9)=(nJy5UL(`hod8Z<<1@97)9Z+gmVEU$nM(UgYDAh zg}iT9byl_I!(*;8_mA!m46Z)W0}^Vd{RsB;K7I|A9hlc6-W-L`CPF2Ow~5gcO>FkK z{sq9-iQV-@JziwdRJ4*RiC;ty`~~1rh`m5Y`!M+Q4YTA{Jl9xmKEdSZB(wEn`uSmL zIjRFe!e7G2@Cb_lvDgP2k9sEWMM=uvb>OP-Mt{*-psM+Cwu>#Lu0lorx4V%$A;ho z;Zn5*|CW>ZIRBe=!CwGeu2FH7t5FAj;cUZigWI~&--hXM z)NO%G1#9biE)bxI4Efo;vMt^jud70{=a)kB_BCK(&rgxu+f0b%aDcF7-PxsL&yqdW z5fB(6%Xck6oKf@7C)P^Laoq@6Prw*@b$flr40=4ShP(tKw|l<_ekq`lKnZQmgq9@rrpymohH1F< zAnrH`lvkVNOvNSjBSlVy^oCGcFb;}0&mM}t)^%TD;{xHC(O=JWUM(Q{=HXl@fy7c2 z1@s8jV{h6H`?H^-6=lN=kVqZ#pBktum0f7=SJ(#DePd@gR}J@OcqTdM+EJ(}CsUP% z{{q%(tR3U!=xw*<5B>s-JJ|-DIhm5{_8dC+suC22oG8KVx%!p%wQiy6$w6toW2Wun z4b?*sD#q2itDEo#Dd>f5cx7!3JgODQBJe|7Sar1H1mv< zHjcAX{aK=JaQZjA4DkV258<;ba9IkMWjQR{I5~LXX(FHBz)`RNvjtG z2DPOk;->1{H(v@fG4XV_ePS)5$bHzSss0Hk8s(uAl1oS3&Xv&ZcoLarsj470zco!6 z1JAs+3zlZbAL^oZCyQ)nTP`s~kz7XL?yZVaGI+H}NUD_2X`hn4qB^-}IFN2dTxnz4^apc(!Gta2|3M3P@s?P}gj*Tp$&slF10AtuBNmL5Kx zmZSBy)#>{BksKnZai^ZYHqYd3d7IZ~S&VF1eUC@s`NhVJMmrk52mKCct#1JpTEDMQ zlnZ!Ytf_YeYHS#k_T_LI%`aR-0tfEJIF zP2<_e>6^JR9P14Fp9i-PI4U{A*F*a$94}FGRfa9jsy0uZz@`E&5NV$rv0i@($(~dqL(2F?%3iInrM`sg4 z#bD`^N0=a7;=-rX>T%(I!IUR%n9{AxKC5=NhMC=H_+5nL>>%^f%3$xO&G%*qEZ0O}6BerL?{)4OX76oShxnA78aI z0vfN?focDU0Qa}%jR%I(z%RxgDCE2Nd*mg>FU3x*C%z@HJUmo|MrWhq&i?jDj73?( zp&R*nbGS`SsoZWoVzc&S2F@~?lPvA(T1!6ZhK3{A9q9aimKj{dD;82jKmqLLVbnOo44 z!|nOWI^Tn`eY`mnNxFt5iOd~(pI;&fn9f31xJud|pxEpj7*idD+^5q%_ib@yh6Ujy zIWyqsU2^NUhULau&pzSW9wuj-|JxsXyJbT5by? z8;p31CI*Nd5E)7hs7Ry=Jf0cL+ou2a3?E-qFtufHTm#9=iF8n<{>)Oceo^ube#IMd znIy?!(G;VG^AaT97ft+*a&5j;bhwQ?U&xO^dq3E5WSVwD;S~OdY3)vutkc((*`?|a z6*mCtQb5Zpj43;ZKGH%)u&f4ARKB3nIe!P~MW)fG?KgP)X=K=Sqd%Z4NBpH6mRuE5 zT!Xd}I#ON#1y;`9&r+d5+)iUp4rt*uD|~TY4=(5i#QRXi%Fbdz1Apg3>7k(D@E15- zZ5X^O9<#ciD{(63xi%2oDlH7%KKS;BQYWb`IdZ!4k>tq1QZ&Jo`|J!}3@gL7k6y8i zY$puR@oQ8L2tXrV`iGt^^yY>DQaX|h?>3LWJ&}>fFQvy{iEQkp&D=>N33ui?Y@VE= z*#!6>CdJHeZNENrFG6t-HDJT_Z(S{#u01uDk@A*}l-l6?DfUh{Vpk|jhW3hdlsjTGUCoi`UDTr=F3mp2f+jNYFAN(>O>B^E*xDXnbJpi<3D5#!hVMf0YeAS8H) zlJw3&@Y^cl7f*>bxx=@dV;A6^1Ej>~n-uaYxRJl%JSx?x1o zr;3?g9oLaQyt&4?z~Rham8u<7GPdR~iPF+8!t9TY5HrP zb6kU##LK|wFE>U@0!H6C2~r2wyYPH8_5$vdL|L4E5PntlrF6T8BcxDDUx&CgURJA0m=9)rrHMEuhWD${ey!!410eEUjyDk2|-VQ@+`!{azCiXTLvdsi8~aM4yx zs@IzbRiG5aKtMYEowJF8@serxz3U3kw7J|lZt6%*3esGT-^zyf`&pD^q+8k1{@8+MjQYv>Xd=l28g>B}i7xr7UTJD4KvHBI! z&A9+ll-5c;$302QtH~T-dOUlq?Xe4i>;L`!wz9O|{4Z$y?{#_cgYT2DKyvl8b7lD< zW~}@Gi+q!3@{UUPDVp;{TsWBY!z-QpSF`)&DECz59uKxJ6_S?I!Tij?_PKiiv&fs; zyQ+KHfGNW5cw_Y8YJRYc@lJv%_Q#L1kW>vgVf4Z3{PV_W zuyGe*dZOXy0J!~hIpn_pobI$)MviFGn!vUhedhp1aNM$0 zOJ}N^k`AfkB~PW<#gwU#Qk~eux0@c*mr51Lhqzei_$N8Ro5c&=trX$QzW{|j7N!?9 z@DHAa@HX}|AWr^L7MiVq1;Ksk91gSyrW`C(!H35rn&rl@0g95Y#%VfsP@tl^Pf|Rs z?ocs1=eS1o+Fmlx*lexSNv_;tk8Qjqd3IIkK*l)g8#Fxn+D(bVLfXJAPa+tzY&*rea+N> z)kb}o!yBLG8XjIv>!lH;kM*kgydyM*DE@Ws3{?g<0Mt|PMdEUyGKI6f&9Xk zin+=$8^v`AY1cmSYdb6EMZ}-bbu*KO9k3ynV?-MjkJT*9hc>Sb#g>Gby8K_w*RXm8 zylyld8Ofwi{#=zt5N4+~Z1cM=1xho>4X0JLF2OdBVz5fHx%Mg=>q+U|q{;MUE8@4p z^{i&ymgc*YJ`$2=px6*IRVG*^RA?V_!1D<|+!x(~L7n<`tA6<{(qpgR+M&OphnobC z$=gE956d}Q^F?Nnt`X5MHdZ`JoY@X*do7OPLN%1meMSh5Cu7pjDQ)lI82Y5p=6q63 zc)SU@TB*r|RA7<1A{ii@(8JI+C*5h8R_&m`nwnCK4i&@*?Hb;xx-FQZbi`M^*A}(M zfB4ySB~_H#0J(EOqftXb^omUx2j0%?dbY#hWm5i3RQWwj-j- ztqj}W9CuM;kv6pd+a#IG+?IuN;&PsS*gX+{jB04}Q`NSDp-1Gr90k_E)0nv4zV5s` zUbk{)yB4h&qufM)|MNy^ql2S!LieW)29{&&noAT+H_l62{wJpDedFmNY|?jT&Qu1b z4%P*mMvcZSaul4U!+)G>`OCLUSw+&eg*loUJ^qjhvL;>%#+Z&4P%nspV}L{f2amS3 z<;4V1m>cr#O`wpJ=R_L}LoCWe9K;(7L%Ufsac_ z>9NUARVQv#GElYY+2G=mSNN$q9&aneyqaxPbW0nx6-ZM!sp7ykiF7DxjdJQWxyR+4Dv zrG7P2c{`~|9rN}&nm|mIkk&yj@@n`%7bQGCnwv~s{Bpcd0*?}zknMp$dD=W9C|9~v zDVpt5l0`ITe)CjMyPspBDkw3Dq51f$PHIps^kl^MB;^nPmcPf9A?iYMc3pCMrUVI8 zDtm^=QNi1DWdI9p{Y^|KGiGswi~#mL>VpZ;n*XsX+mV|@OP#EA92c0_v0JGJy45`5 za9v5IX~I{_Npql3y|(N462I9#^w>Wd1kQzl2b}T~&4Kaef-e_-n__nr_#xD%r@F{p znzV;v+5W9S0WBLS;0oeyb=?W@he;%L5%of=6j9LtAmMjn^p6!G@WiGWn~YhYz1?T_ zN%=^ceA1&|&qV+Z0@K2az1gO2r@UqG)%F^ph{v=>6#|EUlf;>{9BNH^OnV#A6Ic&9(!$wf0#19(|kZCO_? zq-5?B!$(fny3`@p>{THI{RPNdx>GULT@QDx0nuVMI&r^AR`I=*Nhawnu*h4TesU!y z-75s6c~^sMpcu8=7=@px;O{ZlCt!++D%z41;6{F>N!;=z(BYw1 zg_b;{900>`KM~|dGe$*|^JQG2vnNJnRNq#=(2>|DDn)#1fB%_b>632mG%d_HVo#Uh z5lUyXqDk_A1L^y{47k$-FUe%?MVdAdj#ou5XUl4gBhr6rasODJvCIHX3e?XQS8Q%% zUtXK1O2}u@Ja2TY;>!_rVbi>?0+P$UVaR?}*%wGyn_+vY@Dg%uTxhjh>#Re;nlzs) z@z$(+ELakJkS}|3El@{Y7_nP=5y6lI)a+wWYxDRIZrh{%MkLYrT4Fky3}oya>tugprVf!&cu)QGDp51 zR)-_oBV?JM29FH{_CDq@u{5Y5SK&XrDyQo2vF9OjmC1F(|lnvD)%PAJ9iok4*CEv3)SwCt2e)`zF zQ1VvZPa0h`Xq#~dl{K(x)e*0zb7xx|;Bg-ZEe=Zx3|=W}pznB}+AtXZguE78!G|vF z^?Ruj^NdsCK~Rdhb|&_EoajqRq4*-TdHY$ebJf{~N6Uz_YmNM55#g~5OeTVx3p>zH zo!JwA=fLWE(T4Ac1CIN?dx2@Dzk`4Mh4d>`u97u3yfhcHQKw z+O{H*yGpoHV7LcKY0CpYcPH+LYe3jsssxwiES zeNQYbX^~*koj)GC&cDLD%`?v2eTGfN5bHC|O*kwN5SI8c;XFen;Vb?VFzVuZZ6_F2 zmU}=8iQ>mweN2Y>o&T%uz5tX}Hr!c1(lrc0qj8>_#SjOYm2o+-XD;PLuoKgRwf}bGm=9N&7cqX?l_|Q@;w}k27Q#3$>}hRHS=BBRL7FF4b)ZL%d~pX zt$grf65l6EzKUd-Li12qX*m+yP)DszxtBIAoVd(B*r|O-_Dy1eQ6l)n;$IZh%Q$6Q zaqF2_Ok1}Cef~3swwoNw+}(`Oj3&-Lio)0R{3 zvOSufnF8*I&~5)h$Oa}xFRgxFuD=0A$I@q{H%JuyWb+HOD=)cwuC1J;iY-sBMmAJF z!}D42jA!L&>1bLyX8TseA=^Gd=|S|mTvlU;f+nr@v+igrG3TjLZA0<18fj?SG=gfH zimfA0P7MY-*=sgzV3XJ;w_Qsn1M0C(Sg|3r7OckM zMhh{qqBe9TSM}ClmJKEs3tMz?qj0hr;T8o2dueN(h)Wm4*#6gpQuK4@GRqQWxBWsc z9^k^Gs7%XP|l;k3bzl<%aa3R*~TWI;^G$lY^(W&ezhpl5qYSul|(g zP;^O{Q6`$Ji~pr!Vu2Qp-hk|nl(;@YvQX zJhpF??8QcH>RJz95#V1#ke@LD2msJ|N31$Ho_FuahBt3lYDE76(2f5BLL0M$DC#zD zZ{nKleuZ|&>!)545dFD0=f!UN3y9`c;T28G68#u5X7=${UVJjfslVhF428@8RcztD z2>bZ^v1I=2FW@hLsH$^4md{Zsmm*zO?m_;vnsjZ(k10Y(cE6W;FMMQ0yN2rsn4^G| zosF8|w7yHCoh?degzw`%u!U1;7r<+4Uuh)NtD}>j#t*>BPwPeZ>Tht|FRZPV+`}N# z(HE+iP^B=x4YLqa@_Q^gH*kJwUv=`#arB{&-pidNU^$i70fkq{ssl9_`@awW=_9bi z`e~qEIZ|=X8hw9-CDf9PG;l(53V3eh2_pB&i+_}?S--W0+qi?f8Ww7qoYdV{6C)&o zY1kHAN=(~3jsKlRQc>+4mQ5V#?9v|Mu6We7C=?NEr;GTNf``8?(#O zLJ@i|OI~7oNW%!9xrNO& zR~kWgl{<*I`i$i6lO*)n6b6psXBY&*n0~=fE`>RsubQ1Jo`TLIOwyvrs41BP2t?#P z7xA{c@rm56%Lg4Rd{z?C-oncn?6Gah2fR!Qv0#sW)Py8J>My0x1N0Ppl>KpW;R7}lk?(RO&nZ5<^qG855q zj15YR#+;gH&hM-*2I;!~o#3E0p6cL$OhhK4QLWzr4gu~?=1$k&ahO1W0>)@Og0!n2nSy#K+qP}I!Xm!9Q{zaG2!J^NDNk!Yj)aL*XW6+Lv6 zK)YLT@Ok5;mP@JF5;pr@oEct>25VQ?E;#0${w@vZ(|MuWxu;(*aBvk-Ruid72>=Yt z{_#`s`!e!<_%Z-)?AI0>*Vk@+XXf$&dUP@vi@hYH!E&F&JHbH#3Pjt^ww%9!69UN8 zG>gVZzQfk$B~*j{%g|mgdGlT%hS4(X0qCGSjISJmaL}%)fw8$|Le9Xeb=fq87 z@4vo*;+4#tiUYDQnA-^dc@we2d)95c{ z;up*P8^YV`;wo3EZh~LU<9MGknvX3=*#^i%eiea?gV-O3yFl*{G_K5sj_hWk^G%0# z2Xp?5D?>q`kpFzU|4_XkuyWoL2ENEwm92ZyJ>l0V#%wqLOX&U|oXBp~We+FP4*diX z>~bP+WF3pDS8SPu9)y>axX;1JTU@b|Nbk+$T_H*$Abdr^yWk>`@ZdoIvh0QuO9g!k zn(m@gH>6Q`co+XgC+K&hvyFz9A9o_DosKDly8ASd^p(8qg6};S`5!1kU-l_5S=uv= z=vozq?5Gn-({9~sEu#>(%(;h;=j$!*l#IyZb67-y;AG2lOnDX0n^5udAm(@TO}+C9 zG8i!et+-=7xe3&M(tiwM6miRXqsuv}koER2fW#+Ilwc#rByoY)zSepyJ3k~<35k%8 z#yQ`!kvWL%o%5g5|L`#jB3(3V$~+*~oy4E6gAN;`xt61M(7d^IENR4gp&i!aO3Q{j z6++&_h=8!oDZTE5^^NStlhZ7{jy%=@k^)?;W9(4C4IwIMqt$fZo6aZ=`Zor|JJ>NsPmK)KB)rZ`6eL<`Ro5S_$ zn{|sh6n2NziQpf*F;`Ebn3k{9XYJ$#jC3c`OsTr**kY94ui8pEPLvjH$i8F>U+SfI zC{I3_6Cyy@InzWc-}f6!sy>uV&@&E03CmU&__JuUYCHJ)r^3mc^B|5BlAc!K*>*SC zd{A~&m@8(w$QR=&(9){!tmu5tMIr&r;gbiKvOhxuN>-M(Vh zv$(-b&s4BH0sFmNY{FHzbv&-D+Hvgpy_UJIhi9zO_LVAg<8ZD`71c6`#_%=K&sjRp z&vI!Rnkc^XxVVAXk9D1Wh1Vt@~-oK*7nvptTE{W)Enc)SC zO7<6UrsQE`28JC&4kxil=KaTG^gw5R;;Ga)ZczC}!fh8yEH9=eRI&z{+4n6~<4?1H z0gO|uHJ0x@`;yF^z!W|G*70f7fi_eZ9`dyjWqY3ap8Q((sjA^qA3u;eNkQ|EtJNN} zK;i>OX@s`+dsKTMS(34-x?F{_&F$RL_@wfKJU}wwL6tsT2y~R}l8`WSKH=~Szq$3k zxgRF4nL<8!{slvU7^Iz$N2=GH#*T%(|*Pyi}>Q7lPUDNmo#{C0W;q zv%iPgGRi?Fh`x7--U;!oFzk_67 zyoL9P!mu_7)oJOdGZLIvb+~kOb^JbLdsJvCh9t}h{9{{*C0Qiy4x7&{Jp5TDr=$}A z?ZWxnEzpkRFF?Or_XwWSj4~n0wJ^J)lM7;XlN0IzNM+-1@^=RNy(r2)yne}0(AnU6 z!QA2+4t)o~z-+qa26;3U)zAEh`WC=0(tmZyncK)y>3=HsD%6jASqB&=Q(A7%Q|+Q{ zXoR&C^Mr93c7i|bli#o;oK{ym`yH1}#WTEV&AbM{f&_u;dyd!)F(brkh$PVNT|>O* z^sHeFjirdsd0jsr)2cK>$&#Q&yNB3H^LX9CSwT3i($XMRzT=rAn-|Es5Eo=_Go=f-$`ifwT>3W(L&MaResd5mc`*gA)4kH3`t;nC(?fPN24l zK5%sHY-`!Ahn~XOLb?O$R@{Vx$*!NYG#!9oSj#=7KPP)lohS!Flh~j{ZP}Ess$FW9 z0KBHfy`B>6?~XVs>-f_CaAF>wHBh7+tR_hqwg0&f)trn5Cq631ccxemnMIj{7P9GC z(N^mD0p>4}zF#?4)1wryus9`unem$&OoH+FAh^xEhyYvbPZg+QoYJW9YwTg-CZj_E zfWTD;z!!0G6&c0qR%e*=LIj_&{qHyvzWr9uFF5ac@!V?DU3+XM@cxe@Mnd@eqQmgT z2U>Z<-NwFw&S+gJZdv_I0w&L@i4{E&-u%L@aZ3PnF~A{(yu;@BE-(q1Z#5}EViD{_ zykOZ4PEy9=q@2|e#CgLqw7xwZ$Zd$41 z>&CM?;Zh(;{x`ISgX`|Ex~rcIt59H-(%3N+%+W1rFbjMdeQ z!_g5Bf}r;u?G+Cc8LrS~?zi-F;Q4#ujE&8WB9P(P1W++DU{Nu}1}+ryd|U4{_gLo0 z4P=Jj*)HH~q47E*9^3h8*!$_TQK5dA_^skCw(H-CI-BrW?U4DQu_$M1lpyv$&GI+* zHQ4VKiwUg}jdEV;uvb&=S%Y~T=YlbEF>r5pqci=S@ptN%&#wuwWr^rvlX-U<1qy%y z3BVOzS#tg#rPV}m%7)cH03I8yTjYo3JdXRKn86rE)VDGO6OBGWzX;|*6$;*MDwgr| zd=4S)2qIl(WiJE~eb@%#V**Om!>#9;)#e;csV|I00N`bKa8;v2G7H#aDGNZEx8MXOt{|(?U z%nLGwI2ClRwoNi5%WB{QLIM*k#Z;g}02MRcUHiNdq{?+ZM0K~kB<0I>^`Fy}bwfUM zp&$tEA<8{2@kxMgv;FH&{%F*)OEjU`)!kq|iH@68Uh0ir;GQ`lVQIQGU&TzXhpXs* z;|PkQW(j1WVo9@9+cu2`dlCsYpEOKXTs-m)!P1(9aUQUv)?x zREs9GK6z0vQ-D9cc--y|8xh_P#d>rUzI|P?7F(fVd_aIogcm2I0j^f9<3$ldb9_D~ zn&thsLI_14WbYb>g)FHzK&ssFSYNBoPLk<3Z>A9GL%uanyz3y$d6WHX;>AaQscEld z8ljoxuqq;vQ#@vPbg`W#Y1fqR@(EWYYoZV6BH-K!{@s;k@q?~SUKM<G9VSEy{*JoQmA~xVqh~gqlZFn11;d+WK(1We_U{E-6BEuW%- zr~ZoD?5Bwrqaf##vtKEh(YZn~tZ6C3@9amer!OnEczR-&g4GQM&0Y2^??#T-Qy$x##mBJM=+di?9rr4QFr)hSGJgp=?eioMlS#IPXsb7$1 znRI84g94oHG##Hq20bqI0%yX%C6G>Wb$3{rhrtKdjMkI(>cy_ZVEkyOyjk=+BF5QL zw&rp($}YX1klCbnz0wl?#}`z*!3vjC-(#@d zNSfX>?c>1WSpT9+lQ<>UagLEF*ZiwQP#KT=g)38#Z;=fK1J>RHddF5E6M-Jtf&}4- zO66;Xco+H_j8k^~eWALYLK$7|!>%z&?3VOG(`Td?1rG`Xb^kx`{9!oKFSxT(l zBNMIrpEs_Y6OIA%Ri-h|Ylp}3w$JM8A**dI6GRAy2_Uf$>$|mEw}q#qiDUsOA%Fq$t_1<9x4Ni!=FrT^Yv zK)pdX(T`8}3n%9YwnS0xn}Ia}GkQ>S4*yqF*Ds39Uj7{doD19lB>59IpApyRq>8#$HlohYz6d5|S1A%DkvHFTr zecmygSqPFE`4^Cv6w1MaP(1#ssd|(sqw2O0D-AD3fiFlL3}{V0pn4vW8AT+7At`(k zhP%0O;Gc{wEhs}3Q*#5#s4qG^zN*6~eIcG6sj#^PTm#nr*yHG&L2z(zqr~FV82Cm- zj2YDq?vWx`T&{#%8yTAbXi@BKXi$XhY>Y(1-^xN~mJZL9U)Xa+*cZMgoLC(PeR{yd zW^JjTERyKxSDctL&-ER|A~+PLvWTJju;d~7*jG&v?wLte>JQ)LQcafI6`kza%9tC!bUIT%|BykBP#N7}KK43yUweO3{|t5%FDTsE6Kg2fRe?rdqddat z_1BD7n_sV|hS6xa1>MUiFgJS9l9?8st!K^!162toQQh%*3{NuJ8uVz8i&|wMiN} z?D6O-LdM=P!9^U@S8X*}gaCDqC1-E$ben|&j%|xSLPv#;J*_}Zb+y=QOAe6zxY97% zh|=~DldN${3_UExgItgyru)NnaOFZ?NoS`M3KE+xwx+szg2V2xXoWU`XGR&=BV zYzF0$<+KFBDYDL~>4iqC-IQkhrW`BgX)0S^Et#SOSFG95W|*w7qP~-9qKq${T4d+v z8x`&67`K>q?m``f7trj~2G6E`AybAGWwRi*pSJAnsiA$>4b$|*=#XHV!Qe@M^@(ZR z;$|63sdz5jmpY`Jf zr793czr;$1K!H^SaLfC1^5PhF+1g16dIyQ~+bZ6zyv1>Q*olY!yA?LZ`G1aBF7x@l zRysZVvJ)4B)eCbbHm`cv2ZOOPHOfPHyistShNg(}3&(^-bogtDWKTjPrU}D1)^}ZV zfnD!)E5FK2^;A#T$@`J@Cbjj=IXaBSBgFz$SfLcDiirynR$Pt6z+l*w_IT{zgoylptRF(GiHwC z$f3XD1vpF)>611ZbOa2JK4aEIPh;m3jKWJJIAzEjzSRmcnW0`Vc(B}tMiHI|>GoG& zvneoT#x2}$mvlys$An26oHDABD!vxH(s4ts1?n#NqD7z=z(cnqgZWUwpUCFbDDVvo;j{LMb0~2Y^-N-h1(uvIX%9n4rxQ#Gbg&x6Ni?;blEX$ewlnC zx#WGUY}!)7xxcgHsP9w9fg5{iNNAb*?Tnv#GbNnzejz(dE}3a~*Xtszxg5_%RqDKT|;J54V>-|AkLqmm1brJPb$h7sv|g&N@ zlM;02<=ts9TGl8Zax7?a)lraBL-z!>ik=hg;(^-gBZkEs*hu)f37BA56~E*^SSKI7 zEu`p_ack-d0ATXb$kjv99&YMbZ@8!6uYFKCys873Aa~v(j-ko7~T|*isy+ z+Ai?$81iZgrrd2>Vx^2OewXfQ+9~HL8;nfxk(Ah&&@vt4Ra40ydO*y>DnU#^Vh}{` zGT}Hfyteh*3>`v0d)pJ95E3G3{;UGlRMKMNlc*^vBgq^+4Jyw%cNQqN0O)|>KC>E> z-qp-uSW81mo1G#2u#zS{Tvj~lJ?pm|r7uf(-=yU%zLMJ;Ozd7$pG&&@i9{`u{u+RskxbT&@gOP3~|j{bmsaTfFch}+L?=-S66a{R#ObJ zPt>Q&9-!su&lY0pJ-0=6Sh*SQzI9X6$mHsx#1Yi5y7nPoY$A1$6 zqVZ#EtniNh>UF(qHg~rLIyc{t-RG){wq&W}2U4805XKy{?h%a~)=Ot3NOlOirqVi| zDQI=!No_@iGC$bJZ^@~gAXk7#+hn`HYQpUPb|^cgUHF3ZHY31od9l;SlKL+I$16}m zE}-o=M+FGiFPLvZUcl6Qdy3WF*2Yd`oyf?c-j8&r$SbUo zqyJTzO9w*lU;I6YCmWB`6b%Bi>tk~fOz}V-t?}c`KBHy`GPzpF0PThZpg-#iera`1R&K#qS z7B8JtjKoi636OP?k$z(t;zWOOY;-*mW0|Apgee%^Iccviov2xtt9&iDps0=^9d})N z172-pbqC{ckzp@0)9W`ficRhXICapYvWEzu^)SBs$4n8aQN#w3_q5GRcU2PU)B?Vm znBV|8K*qo1lYM9&%znx})mbG!xRygwj>cGZOj6G9Hy=@etah_y4jFaW%UMWS?xLNT zT-{G_AHDUGV@_i8VqjzQN4@WqtrkZKKA%jU+21Cf4=F9xdkn+S60N-zZB(?wk|D%d*81gb)J=5!qs|ud6Ng_3`@j*GQ`Q1p#5HZds!g}+3Ii&Zl zB7`D`gGTyqXRVu5t+)_IeL_2>^Lb0tlhps=>@A<#aNh6lAjMr;T!Xt4+zJGDhhoKp zO9@h>(BcG_AjK`XON+Y{2vVT96e#YL!u|gI2lwy$Xdms)oSnH|XV1>;wFhYc0lnL5 zu5O!pwNvnK=fN`29+s`n&!3s)XGSmy<5bP@vSAK1RLlyD{PSPwV>^;>+|u9RcDw*i z$z>sB&0wzp&+%%iFx(e_kvE2b#x;iDb`YI|&Nt}EhCkVuc#i=n69A8gm==2n?NC#3 zrSiGQK>yFaz-#8bL>0jPO(_}t0+`b($}3FdzemCPhrM>N2%msB5ke=bqdlY-_dFt%P*9l$c(LQQ5RZIjy+BgLMF z{CufO{2QmQ^T6=3j;%`sKTRpc6LSZphoXjflI9$QDOzdaXGhpn_a>k)JUpFpb?vq) zJq|qT5bM0->5t=$xnAZw(F0`@)b^Sn){pi>7htybuU745=GFm*y#QF!jolTrLa$xK zI$e6hR#h7t!zjW4GCj_&zidy9O-|X`!cjNVsITwH>1kVf1~=C?z??Ku>=+?UEu5OA z-EGkVQw-^=g|u(kgtUd_hA=R_33d*8#Jd|>4opsJaq);iz0~0y%Lax+>60Yig)gMs z3M=sZk9EY_6Jh~(ay0IFGbfru)y?ly^az|Oe!&*d+BDBm$S%3-(XxuJh_q{88O5&- zRB$rHJS2Jpj_YdML>pmV6jiV zU_9??el8NLb-o2;Je|O7GNP}0*XpD7&0KZbU^hIT5Zbc-R(S9A`pPd-$s!^9r%UIl z=%fJ2(RpWE3PZsj&ygQAH^YAy9v$}IK)EZ1G872gD}s9HOf4hvc2DJwEkmxwb67kk zu~_ZGad)y#eppjE2lYccE7NO$IAgG6p#n4T>x>bq&N)0?!3X=3L)|x>>M8u~kH%IV zv#fGb%M>)$5g7Z$9==3J+RTJ*hHl#FUYhEfAHaB81b+?e7yIn`hfSWUtB)RmYRbmj zvU9)w9T7n2Tzz3kMz>%ASc~3PqZKkl!m;s-2u$=UV_nQ1UbsOc)n5G&gHP}Nauy@ zDvthMtyNYeZByH6cW=vfq4|zC;eBzmVtL=tiMDAhp6JiUkQxzW=VpKbtG-wZ4xL-i zXo^MO?Fc)=i)r12=Z8W4L=>I<{(4n-j&9R$OHi=bb>g9!A4uAxw2$2!nv0PfVLQH# zasCKAS7=Y!+-#1NErAE^9GY(pnZ6(WexEvYess>0nFeHbi}2(RkPb4I%>VJAr4Xh} zfGNjm2r*OlOMA^+%P+B*-~seXGq9UkPHb69l(5ll>%-|L<(txnn3sRAa?-htl@7n3%02Xt6>D6f$zBK>1 zAAu#g30rAaXU+Z@7|9MoTSt@%-a5g5zPc3nCP#5Q?ffr&YMet?AQQe>fhN0Me`}@p z#A9B0KEG*Anh$_n7`}PgUL&*(Wg-y#`@4| zSD$a3>QfCcC?N5(6T}*=;86z1-$l42x5r?s#SYty_6s1nETMQNQuXKD!6Qw2mXZ+e)imkn`ZS=I}xT8FYs??Uw#aOlTv8_Ng`Xv1m6N-5+? zl6Vx->514DJ8wr+vzm`hMmQf*_{H(GUH~UrgL*1NxGqsm9UJ0zVsvD1hN^JtIe;k+=zKMNnB#l^X4)bhO#XgKZ-JE&b~t$3g@ZCOUz$>a@B| zpt39!%!3VV-i!0GD;z1;R?8O6mMwqXG+CzjD#AW7OVX~gQa#T{H&-KZu(6_d6h{K; zTagRw07t0%i0j3pv0Km(X6zylg5 zQNBtf*jXudzkUG-Fj?Cr@_SB$dugi%Z+~CD-cK~?W8i0_sGxPYSjd?o%~P~KtA8-` zt(^@JV5zBTP;D9>aj+aQa|LBi8Ke&~G%(O@##)FL2p^OG%vm6#=?qVZ{uOdq30vG8 zqiU|_=pS94nXXJbD;ZCnw@(=%>AeI-76Wh!9vAW_zO={@vwd^@q?=q*h_YIbQRU^5 zxmX6WMLwVa)8FffQTtCxG%wt|>Q3DAP#=SLa@Jyvv{z;!%O+1))-sturGn2zg|FUS z7=_X%ip;OSTG&K=D{g9X#W7M}q2A+kL){L@G!7N@!5-vDBNX7)b7go{wU`@8es_+H*=2p zs>5iqm+%76)rT95m6*y>ce`?1*-sj2TmgcT%=+eEtHwzUDGRlel^-w#qbssAr-u(? zhBnuynCcMorBk|F;Mq~kKy!&#dPI|l-|=PFuSUoX*+Rrr%sJdnn~^!clcH@^(f+91 zSDI$X4XSO6&{E_#Pe1Fl4w+P<6(1npiadU`a59QD`OF9gqDmyWK>W&9AMo*Tj8yc< z>-oiwjVn^-^vZ6nmo1Y~QhQiF;(up)z-n*{B}K+BxM{6r>isAtC@D_)UI7psSxnz@ z{rin_v9QZL|3h{MPG?IF_-A|jfpu(L$My6J;I4tdJ!fZS< zL;b1+(SM5$y9RqTazM&SR&vEZk@_fyV3GZ3zfz`1DNH78F(Xy$S|~EvYBfrR6H@|H z&r(g0?OIprD0OZn32M(7z4TI%Mkzz?hVNP<<}T%`-Ts^g_)(+XmCp^lYA5x~{W^G8=-sfs_kB}RRC zD0I28i>EBf+}Ujr8rFg&s+g6amE<7b4nJaxF%N-hWfYyz!kqQF>20jtr@h$(j8Fy? zp39YTm!D?MVq5`aaolgXi*3CLmaowel3lHN+k{>?(fgx&Y?67O$8%S$W@uC{3xRT* z4QN1Uj<<(Gp%<4L7DGiJWl<*eA9edBoQgYL1<_~K-zqbdV`n{_Jhq=OTLt=cb_IZ8 z1M?sRH_u#9n7acD^piLs=l~%Vcw+mkZh<{(^2ZZ!OL1Kd*37lEa4}{8vrLi*(3I)} zVe@RvcHdp*KQ~ureaIj7kADb-61YK<`{AZww_s$#PihcKFWR9#9B30~)to5G#jIMZ zk+lb6I9bnzo(TTxVS|UyQ!LMZH`6q6ya_g;0L0U$qckK_WSn+EEKIF_k+_SU$$@@D z_1-tikz{tU&~B>r(-@{olJx|B?fPfD)2N&y`7p2XWf4GBB9Go(l+ zHlx8{bIo`b-k1Yu-}x+b&hkEGQ4SlU&nxFf5Gv*f{kMwQ54hk~?=3o!tJKc{Z2mVl zDDnErx%d}JrH+{52s9s(!7xd@c-tg?J5K;D0rbC_G?4#vBhd0akQh)52D?3zu`xOu zo@JU)>U{!5=&JXf{%?oW%%W-vMq>j$`ZtY04@KT zU2D3L9#-n)=!19M2f(~)clZB_S_mca=deuLY}v1iV-s!+Z>S%FvUhk-=EHq8t08X3 z?*1}Y=t0Q!%s&aFF7i;fX_BlEh`9?2H1ZbTg+ zxOpIcw^Ge)y4_VB?&DMC!sA|r3U~uV*>I^{!95n&TQs#7Jq^sz{gDi%uN*SFR<#j5 zpYf}t*x(K);15j1bScx@P=T`$k{xHjqz7~I4&|^6lf#6T;rgefJCY0h;7S)U(k4of zK{8-QE9=m-t96d&x+_k-*1Y<&TOk_&RqL;9WgCwVPTd|ZMOxcHcz9yhP37=={PE*Q zHe!ok7GTbG3j{JyM-YUQ7k)EJ-ukuL)yT>>@?D;_8wMJBgw_taZToX>Ew$-jl>IAh zA#_Q)e@E)(+uBoRZ^EfWQ$5R9p}B5sZJIwsa2Ha{%a`a>RY#4OC2vk*p-*xdOHCC> z$BcCo|0{vyWe{$XX6F)hQtVcHL-h1p7>mZ$s3*!Yl-|W_U}ap%y)bA0XR%Kj??_fW zJp8&!qxSk+=X`^y<&hYT2%adP?=~3A-x~!DGAz@4Y#5%Zl_;6*AW;2^R}~M7d9E6& zP41-YY?KjKIpD4}HMx3CFQIR_ZAJGw^EIMIo~h;|Hn+EV$YT)ByvseEi*bOM7;c$! zUHCI3(dp(F91Wm_mDo>>#FCR`R}WTR3W$dDP?7L9c_Di4Dt2oZt@frvLwS*sYBH- z0McxTA!;ML4$9buO@4`?!$h2Qdz-+34<(jQLoe`l#tcF}v|oz<_^u0!nb(5Ej@m{Shkc57q|5;Rr7@GZmicHm?BTTz-mPid(cEg|+j`pU#}v z1V8_2URM*qDxF0;`P+ID4AVp3D7H(BjUDQi`ex34h&1{UzmzE`a!vZgLz;&SvZMaz z+Hmjc+MlyTh>iI5bZ0R}`7s9jkFuAQ^L8yqQ$x!WqUP8O7`fFqhWR|C*PNgo;}9s7 zr{HSue;?*%ccNc1oj<)GoKH7<{Wnh0f6^X=GYN_rUAEh^I3CPxdj09vjAiY1pQ0l& zLeN~Az^9cf#}lI`uzU^tzO0%_oe*A zpp;^e@aq@Aa4nhll_=I20G6N52B|%oJB(kVDjok4tt!4oSZ-FlY-$+_;kPtWt*;I8 z$We98BwDjj@m7k;wWljaqi=Nu87=PrBEOk3``ss%Q*$1*+!ziId8=XOweYPV2x}3t z*AtKD^YuRZpI0(xrTa`^d6iF5wj2(o$KN5An*L*VruKyQ@2TxXBLlX#a9h2996E9;jlG+Hw2e1G~oB#b|i#BmYcD%5b?qUg$|Jozg`O%bvo7cb{&|zwunRmk>jQkQ9g@#v& zBfA1`+gC-K#HV*aE$$QJXec``z%z56Hx6y{UbESx z$N_0i5)LPmd}s=lj&l8O$_3We0(qt={6`0fTX-dMRnkNwwKhku{pLLwS$xvd$MBby z=jNQGNCU}?!juD5IeX!7JIalfoQ%uE;4e7vI--vVybdnZ+5s4pg7?(@^|tXFJg~AP?|$026(6+QTXKYO8X!9RrCmvr8x#hfSJk%`DF12^q-)GyV9QkW7L`F1KD7V5t zk3YC>bR~xlt?Wh=Z-Or86ZnCE}yTBU~W-GPHDN z;JT8E{eKF0L?q*W?*Slz`JzyS$P6sT(1&dc++@Q`Ay^n>G!{jy!WB& zLF|a`hvLiWK;-E!=)r-$u#VD^1bNSH#^jY+#2SI2e!~dDBcymJAjs8we*Al;bsyQE zkLuisrRy>(L4w(Af^6l%_w?WPxZg)bQ zy?|F<8HYANIdG3_8RYB`gJFQxPg0Q!4`qZOoy5X*6cSl6rN5jDbnd#Cib zd%k5{FvN<$Agj#3KOPz-d^G)#z9_=|5sX$by+BHS3!}kL=kgx#{`TzwAvo^Bk@z~_ zX>r!Y44%3noA&f6jL?%1;ix?aIX!)<69>$MncFN@3m!GDT96Xs5GqgKORJRKd=XL4 z#%62`Y{UWvluRqFy~*5C*FGq@1SS-G&Mf;mq;xDHGDssP-51Lqk#7LG^z@2#EIfjK zGS+|Py>3fHo}>}k{ui14jCjzb;=;{}8tC##9z1E6{!%$ULCmt6clhX)&eEA3Ta{cx zSSkJj5KgRiS)9i*)Eb2UC2rB433P{PQ^tpLVfNl>)W)F&}xC=cC(*Vj_)Yj9VL# zB6)F>3c^DQ@S8(D?2_z-pOo41AYk61_V|OCce=G5SD55oxoBp}%1TeAZb}DcD?XEM zozpBExB4We&MRgBY>znW3#fOcW{n@F(HH}8Kogff0-w~7FHG7C;Uz@9Gq31ysSJQw ztf*B{W@fpeFr~$w9<9GOYWIV+(i5-LLR$;)GLpw<-zOfoX?87qG;LCJ;O#&iqwJme zn67LL&P#3YT%1LB)Jsu1oBvlCJL2Qz6XOmHe$0ykn_3M$usUb*Bt{}wpDNu`ciH{> zFl#x#JGva!K<&LOxEciMMnFSm&C(235(7kpz)yTaZbbU{GA0@*dUKAHeYTi?k@?v{ zj&4r~NzRO41fcK{wZ-?dmI?>a!VoUy#t2Hi{XYeqdgY9s*z46R$A?Qs3D!TtNGq@Z z9p#GzEeNifSqZ;gPNvyy7;*2ZgfBsRS}AM|-|xHK40{<}Dlbx)D<*+EBOagR1vxGa zcD!2uN&8r|88ny&wIy3(kUC~!ZtR^HQ`GvN7;EKl4&BY9xF9#2I)2S4+)X!|C9UJBAXlUVYWNQZ)9a<5gY$S>HM+*co2+Y>cmY^ z)%XQTJ9ZRO&E=k2Y3@RGkd3xl6tBLtQaI-j>laL%xOrWfps3NRuAG5739STJ<9t?m z>s^Ztuy&|yP0{3Tu==T4ESAkKh|+lA9l{ktW$vFI{D5WgM`*69gg7d}6Fe$>pcR<$ zkq1*MPfqx<9$A~|;?US0vy-Ng;#U2e&@7Y>*qZSchUZ8*^WCL*#3zu7Yw}rFH%#dJ zT)v4*W2WE?8$9Ts?3b*|H>njzFDusQbt}5|C8n-+^HoQ521slenBpNEEMM!=02dx@i|fbn{x69BW4{c{?)(M&JK-YrC@?)hmF&OQgi#$AYJH@-q(_4M6^Q?F_yIN8~wR?C(Xs0jBGl zDS83Ltll^Hy$9~n5OJBp7gs)b<9c-HhE<#N`9$J~BR32M%&l zr;&tQY>4z;CwM=ba}1#- zGk+qcJEv97vlgM0jA|r$n0KKc*h2cD(?7`LwTLf@JA~v19_j#_thdWmzs1KQmtjfQ zpYf!la_Gruq*mT@bxb`SoFkKtkVyI{Z~wy{89px$uW>|5T@=GkT4V#=0%>m9$9EpO z7j&>-Ytuh#jQ}j>;7E(rB;hGDcII}yJ&Lfv0EQGHT`IO%iD#H?#(`j=>J zT&acG#q03bMJee7BUv;gmVb>ftYFc<*xdX*kI2Z*1B^C~%&x%^Q zEM>&|QWQXm60jFQ4&rc4y@aPokpwpotB+Gm7CM~wq47d1KjTu63w!}WD^pfktV=Lj zZt)tbFtxWWVu7KcLn}7!szi{2Pkf^i_H_GoPj-GN@eGWE?#ez$()Py%lFMyo#sPq! zr_c|7^5Dkb!yk!O;@4_%cZpZv%NnTmGzOtHL;H9X#L2I?AP|mEW6W7gZYb-G-7)Pe z@VOSoSKFiBYhOD9R#JO9+q>IcBM^DxbuCgqAdmyq3}6zxAV$k4=^euH-NNRayBWn+ zELb7Em(4AnDtpWsXk2XL#Vj>g*HHdyVOGJ>uHmL)){kXs0Q{&EA2qzdQm)9gnrVM8 zzdW~E@>3YEcRD6=2^xSH!NFP)loyEK&LWAvk-zs25Mikp!wUCCvF0ZnFM99TBI;bI zy7-tpKqGi9Yt}pzVk4|fscNdTN#asU+g~!lk5VrXJ9nH8pCG;h+NB81)q%u(<3zvV z9)smtOvy{+qbRS*)^*tCaVAIrpEB6?3?u_f=j)6ver2Wjr2c?uyNhJ6_sy0%2njGe z8a9qQrKHa9yy@)zg80(+TSC993kWg7HH{H?jpDV8BuEmT@O}Yg+G3yS9r}1)5=!vT zP~uVCkbkaj`ExiA@ikX}!sTa90aW*N-LYcn#Yo!;{d1#g@OUX~O|4qO_*3FGX@QWaDf$O6{x`>~nTYi^-Ml z!8d0*RIb8HF125iH{nl5zkEB@<=PV{@|NvSL!1T8b&Z>lDUy2*=tQ32-!AhXTJJ<% znRn0)SX*2jg_+Eq`Y#E7^;AMgoK7LJQvx`VG@pCkQGZMO_9z@8mE9K=V{wbv(EJQ+ z9u{S;q&;$#Pvr8R9u!1N#+SJ|d5ACJ(AbRkt!zIrcPW=L_8MkVaEO1zQ8-p_aUzEc zd+&JtcKy1-(=+&fk^QLwnOM}^{aGbvMQ^lb{Bv4f4CuDjPv~bn)B3n^-twJ)t z`E-0ZMB;jkgu9)Z6ghWiEU$ynCcs(A5PvGa+Qagp|I9FgN5;s=o^o#k)5C0hjL&r^ zQ)x8+^iRw%s8Q~^tZmH4jgl#eQxch1W1M0!hThPS!SeGB)7B{fa|WAiv!3NqFLS zv+*7O=7v{pmTdIIyTkphreRsaVNv)Q%gO#!IML5_^}u-LDXq{qf7 z$C@n9(vn*rpWYzOcXbaRm6W$gNxINdygH7rs>v71;;aeWwsds-IB-3queFUir_*w#}z@Hq3--ySfhgx3imz9Nv%IuPuit_{5SL9igqC`fb1F^#-h`iFG2K*d!PThf~2n{BiV7V%6UK{H3*(_uR^~ zJ`TvpGrj&xV5RUx_#Jt>na*MHTBFZi1S%Nw2W|D5*Q!~!oJJUtWZpNwS3?2*8TIH% z6X`t$s49hzhUK*Dv%6QJ;e^yjPLO8P5!5or55Pt&j^N;{^=c^ zHr2X_Ix$}8$#aWJR{Ym` z^~93R$3>?NVCwf$@)?J!|2x`c2Y};!gVvjWT->K0_m8$W@{AaX=_yvJV`H6a0k{nh z`V;0Ldku6puN*ZK_~IEAmqMjCi(86kKovYwaQvqc@Bm1#An04h8;k_nEv>Y+HWQIf zVA-5gsqM%k`XF z09wu<)kPL_UC3?*;0Cq$1;9wf=$9-ertLCm&J!3~}* zS!T+ReoQ@|$sj$kJzkAIUE?L%BOH8XxqOSCGN1HwEfpvtF+dLsyrRs#M+=(~Ac{Y~ zvvh*a)WP1ou;7aEsDGL(Uw7MMjT|x2+AFCk4T^<7>=k2LSCsof&bE6N!v#D5b) z<9{^A2(Rw=o|b9s8QcuHcgC2Y8&m90lReJ=y(Jo;-4Qj2r}y|fPAi{&$N}tbL!GX!0aM0yCp6R-xFVRd%xsf)E^mC5i`QKet`ssMBQA@&w`y9dJ-5NM!`@Tj~9SiLQDFxl*ASt&DLt4pBG-hU7uW_Bjp-?C*-fe zesPs_&swTNJV25YkKy{R)>4dmZAzrZOjtHYm*Gn7R7-W_W1hNeu>OO2*W0cZcu5Ut zJUpt8a{Bqke2isoU3b3SF`my@09d7)G&J>t?&gQ@ak)C-BzJ2En>aAc=d#wae`N&U z(oA@^SYVo8ZcG;1>fEjAXl*>+v_HC}#+g4p{JEvb)J?A>eqF()P3T>bEA0--uchBA&dtTIH-v=`vP;Nj*%PZXYP+H&Mj2CYK@lscE%m~L_F3h{V zcj@c%?f6+kZr{`bOi;+Ee@0ujc{ zz}w0S-+ef0cPz(``2ti~5mmO}kK5HOHg`-#fq5v_gRmYBQYl9YwgdeD3G>z7+7i;( z_YVQbu$)%C0$-u@;6dP?Cco%z>NjmSAuN(h$(T6aG8b3?@(-(ltj=bb|&OsDf ztdbW1i$pYDVd~n#S4y2LigSTO%{}|{M!;%|w*6l0<4P-;q<4}#vSEH)W9CC(8Ic)Z54l0h-A1JkBD1wiq zAZFf*qV4krDLkVV*Rl_V+T{~3fH`(T4ZZC}K|+3t885SEO_XUAtL2j6=4=I0kw&E- zB_$yGvafi=S9G=ne|H{3rp(-$J8HkDdN?`4TCq@gj8E~-Nn=_{-cZF0I2*a6;DZ>G z>U$-377;zHYC8^7@;Iz`gZ2O+?dxfxZG7##8sF-;DC#|f%?P>C$DG^DS08M8r35@s za;Bw~pT!j@;r5uAz%VRY>8d;Dy|VY`q_>eS_QkS6s00GB(=Wq8ad>C;g7RtP2 z_zEsRaBLqi4T$0V8U>8fbfMgpRc;={)K?~I5qt;I?6PUEe@k4^!r3RG@dYMEXC{82 zQlL?$@LNgw)zq1Oxp=3IY~JoaR_ki92MYKO3oP{rzghGE9B>nO!98*wV=A4@=qgdv zdswHpv=n`YQJ4>74=l9rL(b>l0c&G3^hl@n72%P8u!pRs(>kxuUp2%Uv6tumXdMix z=rScHK+)+DSnLrP`S^(zY^9!0zO?fGsQwqAB#1!wxn&?&?tCJto$gnv)eqL!N1nXN zBz33QF97$2g2aEt!|az98jXNyb6O!}aOJhxg($}Gd%TWO+W>NM!Dc1Zgeh3N^jksC zGd2M?oFwc_uLZxFH0E0BD$Y#8#n7(07;RNcp!mXLT<%5MgK6;BwLehOO zECzy4Cp~gqVmTpG?!b(-Oe68-?Yv@L7M{V3JG{g~Y^{TN*@v>uyunodUZVsIz!<$5 zBp{GwzE(wkF}IUW*^cqSV8SSc=ZUg<#EzSX!U*L6!}lnXsfHm?PQ}} z@G2cwp?A3J6)o0HbwXJwRid*#NRnPV-7}PM;q-dbI(vp+swDq@Rb>7;K}pjID2OTk zPo$PAPdSycT&@LqB9F!(Ku!?u0Z?y771JjL$u*HLu zmn*ONj5JzBKlMjyxc-XaHHk~c#RN$vZ&w8PNZKVFe-aOr(1+Q2BQStqcD)q@cK42c z-__@u%W(c(SXIthLK9H)Ilk>*Y;E3;&=uQ@?iawSdaC<*CC#3f?mmTSBH4}zXfI8kD%p5XRal&j6GLmk zHD2XsiO{No&N8sAYLWe85sL3?@Mt+nl(` z7r;+SZ%O=~3PTh@M^GxM_<8=g6*iJ5XFJz0Dq>;_bN(^X9YbYfzC zi{?!*vZtLe3o9|*Pq#$bx9rfK=uzDGa$1zI9hdAV&g_%dR)bhAr*V#|7ZY>ReyaB_ zmW*w8p551lg~J^is;(u1b;P9fQh7Zmo+D~n^@>cCO(WSs=zhZjJ&uC8&N~%3^A3q? zs$c+q)1m4_s0^_oR|z1xduj@RBzY4;?5 z;_BG6dBLvTyNzuEM##nLInk#79NOHWsKvNlx8)Oc^HBC}>u-3`FI8-N=1tomn%(fr zI?#RgedrD^UbE|6xQCk^C7&C%ZsKk-eh6MiLB@*G)p!Rzwm1;dK(?^?S+_o8!#F}L z;{6_sBAEBHZHSNV0rC5!)YRf8Vn+1H$i(Fn^L`SM8R+%L?RULj`^&6Yw<=a3lre!n z@Hc^}d0lSW(5o(UqF!>T8c)RXejTSYWx6xA&GIe6AbD9wQE%r#Lm^pCoaw)`k7|r2 z-MLU>P}&NVD);69pI7u+$Lj}ym3-btnD*{|&9 z@F)Y`RIBpA6Y~BU!By?KL_SdAf{mRCX~eEAFOzrlAD zz=m8ZYG?Y8u|ipEzw1xQCX01FWu+FEv1sZ>PVovoIbcZb{BxFg;v_)s%D8UQi070c zkT5-lnv&vF`mN;F#E?5nlVLo-fjYcRS3k|T^V$Jxj144gG>O>&yhXU?|A&JA7*_Fr zI|=$L+gfaCxoV5@A57{A%?m#UefXL-0@N+Z56fBjGs-Z7{vK=8T5F+Xy4zCGhf45D z895yI$v%3*`nFQ+(}No%nf!)wC3nRA1U@Kq6tr-vwnuyx!VzENq$-`@N|uSr_h20n zP~OzXE>*-YcEn{L1?%^YAatok&bBjHccKck*G!&E&~q+;OZOUKtPBLjA|)A#J^rkv zb}Y?-NW!uByW{8U&chFKvYco^`=4W7y(Sw`3VgtwYwcHb4xEJ7`Pny>G5IU%<&)*C zNaCBa^C4y}JuYi)#r%i@b*wA&j|Gk_-UfEex0Sq(oEU_Y$4UBt7Zt%+Eo*530pyk_ki~ZfNoO zwhmBN&W-X&E#!DbY(eqmRjB-)5^h2(HFtL<6I5zwTzq}-eZ{jg)4fb>#kklqjZ?SR zrqusAFXfRx<-aOS{m#~JJ%MC2g`9HIU4$JqUi1g$^C#PFkz)oZ@ZAHM&nIB##cgVs z$S_%BVC}b(lcbf7r!J`YPQ}2A7UX27%`d{@l{YqUr_N&0o#A_ZzWImklz=qv>UtFA zhtVO8P@74C391@cYKEK)gBNn{O#p_SJ8WHL{B+-cW(c!in@_`&Bm|6S_|EASS30ZeHK%e~zj$*=__e}}vPM$XJv-0e-uKh913_;<8m_ZrPsZkW8!@MvUpz0h|w&o%*jQUg0tK`bVm zIW6=c_@p28HKnBIzR(0M)-PAu_ zI6ee%Zm^z8VRBfGhnZzv9d@&l&Al|kv^frgi zz71fIo*XP+KT@xGyhMchp@@m3s_OVWe8c5avH&c2jq}QnUV@EOGBbc1qn94)d#>ng zpE8Qpl-CH=k$XCxTA+Q zn>m}I6OFivR(jDQci4Amss`W>)`uDOrp~rJr6sIbGSoU z&ZkEiabMZTZcO>dGb|^j3zdIvVz6ZPqS0tbER1zG9UavxDtjtVAr!#sfBt%`sm{C` z2n-6>?{p7kR*{i=4LKFhTj%fETjd zs`|es83~=;QSdk6nV8lpX1?y0Yumh5%XdWrT&ep??aXem-95}9*@&T@`Ad=ZcNz!( zj=^6|N)1hNsgp4_Hv`HEt*=B$==6_-l+66L1+1(w2*A=DoO7W$eU3gKE6i6zbr|y^ zNV2&xYySq+8GW!E=wu^w(lH$oMwP?N8O-+w;_Y`|^if$H4-yHaPAbt1KK_RgwUb4& zvOH$WJ*=&NfTj5E!9aQ+;X^v`p#DwWWNj~o{*6ZLH_sOU(cxn=k4q|7qjF*V4XB0> zSf}Q-7NR2Zt+}T))+y(r9JVJJ?-5Uf-$~F?lc}|2#QEht?5lwvt^e4}=t5oGHL^b1 zydq(pzTU8DhjbuUTa>f{M*u6dcV3xA!~A$?0y1Rb@)6<0_GBOf2>m z3~jO6Cf)y#e6c!;oyZ$EtIFoldOdus3sEsAUX|o%vTghp`dVBvbaTPKt+`ughU4Qn zXE!X^mQ}L;aP>gFpDC(1HLIEAKd>tV#Q%NXfR;NiWeT|zJ6MZp9p z(P|I29{gllESuokQaQmtl3$aKk5q!DH`F|wgT!_rd|+0J+KANO{Qec`ule{O!b8Qa z$*-s(KV&WbwN}gIY^zi7qYB*H%Wu@a09y1~#m)sr zd)J^Q1>+xUxUEWk|M>_1A*;w^P7ZkR4V&_>(Og?P+S?B9s4(fa(w!Kd|8kQ2@?S@& znwSs=KIJ~)M8Urn#IVVXiwFH^r`x&3gM1c&b@Ag&VYSTXO4DD?x0ebA57pgmwM9Q)KrYPZ`@JraCho+h#Gx{?-JyHGJIPY&7OuqFi2-Bfk$=s zW}9tbw0+`})$KA`HH+a!gPz z{t9GA|7%3e2WOa)5S<`KON(&M`+d24?*sRUUECgbP*yc(6G!V2BSzijLUo>Xe~*Tq zpI&$Q0?=l_KQ?SA1AWBD3r?}GWTuXV1fn-`wh1uMX3mu|C#YMeR6$|Lrz&%etvr20 zo7V-7z%L&whcQpUvSyQ2$y*aC1pZb8e)VH<8Uq1*0efa%| z=xwt&$UsR6Nqp;TMRh)-GS^WY-K(_;;Vz84#Uz*60lTJHy=A$GXq` zKGR3_>j|5uEtg1KDkTgj|9&W-VZdhyliNq0?Bg<-sxl|8|AMD@3Zjgh#VWAV`0b3Kc!qZI##A5SwCOaM5)q%n zmd}i}u(lD*ardDcmb30g0A4_$zpkcXzqI7Ur)8lJej`-hiihFmHD0(%mm-r0&rOd} z?A|)aooy^!*e6UYp-Uya05QCc*>1 z@I+&x(e*B+e8lJ4YKFfHer5@^FvazIs!|(7;}bu{k`@AE)2l#kra0zi&2A}nJLE2y zMRCy%Z$*H$@q#rJ&`6jS#*@9*JHmBWC`$pB%Y{0AR7K2oOZN9dO>N+IW?VPz z)%_3E{(XAfbd^zjoF92t-SpNpDF^jrvHz+ zcM7g7YP*GZY@=ga9ox3OW7|$T?AUfX_KuD2*v1Y!b~<)C`1^Z*T>Mq%)cG#X#Tiv= zj%T7)%{B4N@!ZS>1p4-*Og-P>#RYTUe*&=QD^#F_7*+cUO1@QdM7H`9`^&T~du$}T zn!AMWr3c9YlJj2w*gBd$E-Ee3{93YB@V7y4S?KK3B39zGdH#bS1b3lJh343NUoUb) zy4OAUjZy1^q|xq6KcO!uT^`vnF9q$wkTRr4X_w{e?&(7>t!C6HqWesEWHx3(S@0@> z&B9$H@65ZaE*0{@J`_#oEXwop?l!vpyp5kXy8f&`nE1_Dv+}=_Y57LY1AWzHLJFxgaRKAD`4! zQ;6dRZ8xx@x-OR#y&r3AvxbrtQP~Ba3*PgPJVNt88v?Fq+UPJ^%V?Th4n!ft+ARv}0asNMiQr zR^6ugaF5V{A5l&_x*eOoKJ5^J_9%y^D%(=(0;u%V9FCH*3};A??M>39g7l{5`Y0FY zu%#Anm?}hkd@{w?n-9+E_j!BR9Vve-r!HM_4Eau0@((L%<0Mc%JTwg4<^$;PX9AzE zX_D;E)?c~UrQ9xC1+C>jgw~3)vqLK_jdnaZ8+|DNFD3~u6-I$y(xl;fdD5xV{yZDq z&-bEdNTqX3NbmE4kmU!n-72uJZu^R~Tm*7t6SNORa5(dGsiX{uLOF6gzh*K(B&Z9- zjKyQSa}O@GBU0hTS!kAP&$o!PC<3+$uk@jj;mOPioxEQbl15@*(^BQ}x_IcvdDYJI ztG}w%5iib==Gj(JpCELlAtH<_27#o8s|+`croD-3%Qh$j3zl55r>(7+NGPFLkB&rw z?yGH<4(!WA;3&J6;E$O*|9+0G*Td%@mQBam3cJQTJ z>_snKEBb=Q&GnaZ+AmyP_Nil=k=PjZ)oB zVyaccL8=D+0v(~Y8rKcI*w2;G=D1QVByyZIjV;fhTdV^l$qCp%2afgPG8Xz z8CJd*2PqjDe|MTXX6p_{XXrjrk#^I?6D4^$+pn&sL$A|ajWBh5X2ylQp8zhWjT+j0XfM8qIzKu&PCLaesQVC>U;?%l3D2}D-0e=l+cIodSwp#rjhM}Z6-Wb#CX5P`%YZmyRvY)-$I<( z1Z@0Zt+~Sk6zBfIZ_CLTQ!!RK{C*+0%Jmd>hMB3vU}0EwZs@oJ_2d!gbBadt5m%+U zB?Z1CqL{)A%Fc_wIu_jtZr&j30XeiZ3D!Qm)lmLryA8#bw>!>cSXCF@`2MHI8sd}e z9mGXnxA2MmDjdV5o__jd3XddY!Jti{peu0WgxAk6hSWz9hHT_HqzSC7gM>tRVUHEK zb3=E`!@1Z=mt*``iXPv@5tyJIQz5z0c+^%nACB6h;~tLzUN{c^8Z1?3dP8S{fZ)7z zAvBOL+6~r#*J}`=To7(Pla=8{vJCm@Pq`WUl{rHhCb8_!RP*HZ5}(P)THZv7jD;iY z$!! z=loUaVm+vUlPYVB*n7RuSiGCOECCL&lWE8SP&e+4_l?KN_AAVWga=fk&kJI~ zydX#W5*x#sY~#a`Y2ri8F)YyKeJGJ@rr8<0u?^qgzMt>nLVK*9$!YE4eobnjT!vG3 zWZSMRo^5|2)_nI0MHlbDrzFGjBhLaYUnL$-X2K+f4V{n-kJPnSQOY`#=>{PEJo9f~ zIIU1uuvRC>M5It~!J94&amWb7c2Xk|!NDlY;y?6F>(TUf68K{2SLKTy>yB*PQwmi> z9;@^?D%we8>=xnO&PqW$K5z~k$2=g}@56!|vAq`Jq;-?fzB(i-Rh!H4tB~?78_BL| zMY5r^anEprsACpvhQ6EZ=GEli#yaFs;Tk*eVy33oITQ3f=O7iTkqeJ6juoY9BBPY)+kaE*A8y?U4D3HS^4 z04o?eZ{Hq10Tjir+S5WNJFv9G_WI*9-ls9crmrDV*ceGNt?2!<0#0l>cPsn3@hWJ~ zCY-#ev|tO?P*g;iLF)q-DL)(t0M>g)eSt_k30cpcGXcx}xhqG1IfhMRu=uEp{M;h@OrmLac_+? zt(&`iiz};v@JAGygi*x&bKQQ-CjgpA%9fwQ(6%kw;ZNV4Ffke%seumH8?utS+{uQWf}MP2x_{IwIXyBBTfuKIhd zO4bzftTs8GdBrb~!YAMcW!@8?i};I9#DW~tP@sZhVG=2m>to%wXRG~L^g^)57W_PI z>a*wc=$&tpTQLsV!9JOzhI*Cj3WlC5X@dKipwSbit=WC_JgcrWoD_6OUWk&Q*%B|7 zHu$2hhkwqw#a_4g(8od}!^M3%FLjal?BY56dNe9Pr5yjH$7tCGH#RP+U%Rg)q~ZUqt7#>aM_VovDX$EPtZ&i9+WE7=CCC z2jLnKZux}lF4CkDa+jHZ&BwN5*98PZ+|hE9@a#-04CET4)Ym@zMp9k!$@zmDdd3Cg z7zGUwCeQoh4O=FHm`M=hQ-K>5ddPX9NeKK7_=q5LLj7G;9})1)b@%&97fWy7BY)iH zyVLUEU`Qwc8WPR$*1k88v`2rB*k5hMxpXHMXvmG^?(Q%r?n`L_fIbqqO7S}io7v)X zkp}I`L|Kycno*_KF*%v@nm}1R5{A0omj~jz)HUv1MJL&|=ftJTag&^#jxjaXK~Oqa zKni$OWO4@p_j%%&?cJ<$ZXZuZj~WD%$Q+O_f(y-gv9~qmpC|i8uu4pts2|co79`;& zv%Pp{IKo=tq~C1Md9)1_%C5P}Hj5|a8gciK=kRVf`rMy!_B8s<$}Ewb z9ut|K4+22kTU8>n>AGW8QsTp7zlXBbbT>_#8SCXxJu=7`D(z?9tMds!DUW04)oKtZ;9Q^-R#I&%D<1Ht)+qxea=O+E(rc zXH!F55x`rBeu$Nk1M(&XVHGRKPkIC2{D4VrwD4SsFnL6?VJo9xxt&5ZY^xC`E$+K- ze7g&bG&mk9JWxF(fj`C8^r+!kPt!brps!Wb?T<3-VVg)Re+c;jZ&afIB*WsuB5_Kb zVkP0!GhBei1yyWrY5dBx)}-SGl*p-Ek2eLNKo;o=oNrTg7;Rw1t#uxsAXYiMSP@3N zT7C02^@J&3i_1h5>b>Dt)bB}1o<5>1jCHDQ#HtV*mvC9!g1viFtXM0T82o2&_oVKZo*o zVg3#_qa@2O#UGLKH4YXOZ4cd3PxzYx5zG|JxMCjJR1moa1e1K0rfd;&^uV4JLOm0J75?!!$0w9S}2r>zQxvsLSp z&|P5&gAIrAy1S^)BW(G-_V|K81L{O++W^r_W0WeZzUNpIv~v1edv02UrX*IkZ|BK2 zeh1$A?g%tFOOicdSbTTi z^+ripuF%S;ttLsXAEc=dGDm`6keVR>ctd-jDRXA|>QpP4fd#>jUj9{;|M9W<@5cI> zxI!f*hLQ9!xnC&4$hQWDuc2Iwi>_2(ltow*MElR!{r=vUjw-BmoDd0B&a(S@+m1~x z1yqkZ`V_HEA;1B&bnzA}b8`Xl{p$t&>|NW765slzcPsm2oxFm$X zD?IzcDBX(}Wgjq5o4^@9BQl7?nW2y4UgL{WhF_aP{)oPIr{q*g%6{?vqmJhR?(-zp zL8zQ4?rK=uz*&2S1Z`1YF-oV=C9P5yn%oMT)YEmT%~$^p+uc>H%s)O56JQ11k84EtoLd;=V87e4jehi!)Lct07Woq+lx zryCMg6F_7-T`z8(jB>_ zPf+EC6A?+5>c=?@Hf9IQA&fyps{5JanC^gV9eHCGnAO+Jvk%k=cu!Lp7#Npqz5wE9 z07nn^pc+jEQ5P3oi^>b2wrBo--K0ui+GUi>RIcZ7E|Z*~z{wydT$-F>`x_fiN>jD6 zNE0O(4{8HDd^tXS61KCsJwLdQU9wj6CCPRe8)ac&n9PQzJO%(h(u%a4X*wY@f$8=CrXd=Q;0Mu*lz$g0BOw~U+}gWD!x(E*M=7K=$NHBbzF{|zw58KN_8ZLt&x|s; z(a#dQEJC8NUdIMLD7aoIUyvh9(QQRbgqB89T}bdNMk{WcwpZs{Ybm;QeI`L&pbumw z6m))nad2TsSc7o4XK@UCzDlHrz}>8G>H5UlEL*so2uF;!wYj-1eCYZFdpU9%J&{ zPMkJ#%U>kxW+z9} z6fUyrCP-L*0Q`D^>No$#LV55ycndw_no1sy4!%gHLQFnB7$?E_JB$BOn~)8g6t=1l z@y-EWH_D%p?RrbO?e2^2}>l=3#7O>@Km9zgj>N$ z;al}Qf-@-r0hyou=GZhirs>WKku~#M(?QJ)QTqu>3CaiG&Mf1vpM8z($}+-hjLF#6 z8dNUHmJTbR4P^^w+2XvRJ^?tkt(}J>G-FVBsvPRLU#xRIsQg;_J#XIlUaLru8?^-Q z1OWU6M?dtgL8-trjR`n+Fvx9^C@eW$r<2T4Uz);1%dO8@edB|LO}S&Iql_=+J?ehr zy~bYnOA3jPWGxh0)Q)pZ=RmUiagtSjx^|xl;wwSSVqzo_4p13C6dT42lXde@`abbs zVv~Q|oJfKeV_6)If|fSS92Ip!u_IezAvx6Hy0{=1N0b-L~Wn&gg0{{BQ395_;w zGYdHB>QO8oT4N*hmEilEbE3m4l07|{oVSnD=-;layNf8~sx?X23fh?N0q$(o6O}6o zQDv~Q7sjy(cWZH9g*{J7(D>F`-A0P`<=3pZtt+4Dj;VE-mDIOnTSwK#5i1X&qspw8 zj9@Ve3APVxy4PHT8?o1N)&<|>N1eC2Su*#_V=I;H9jj{T!uxDx@!^P-`P6!GD z{$ddBTLBINAEZ_SV2+5^l@5YaROq3goHOjGZtg*Ce?9lBKHPOujw5nVq$8*Qgj0#s@^A?MWJ_XpgsV&&!rZjwW5F3%rPkHv$1wkqk zvq7MtZ{t#m)jtb=mcLoy%Wbp}yg=CB5TI?Qi$}bCJE2UkP%!rwh8} zq1-Ub39PK#WB6AFZ9LiH6yn7ujgvNczs>n(whopv!zdPW?xq7;4opeB$*@iv`~GSf zt-~+eOQ0-*dN!&Xpd(JhE*h)!K!v&T1$+oh+%7}G6|3=gF%!G?Dvg$@x{cj?beyqj z)GZ9YSAgGyaCp!uE@E)4`dt1XVS!!&#YGHr9#RbQXx%wX1n3mMt6w1Z4I7%o=$3q5 z{W0^Hxe)e{i8EuW6z?J>CGZbJ9S?Hod%8ov5S{1$st}GuF(L-5=xZR5R+yhYd z#WVuYyLHr zotGOAQ?0jYYOue@IG0wKA<(~xC8-B9_1c_|>=OWzRjR+fXU`7$E;x0QjzBliNC{VT zq5YgY@z^fRLGqZ}rI)kj)KZ~$K}FOTe+@r2JGaC<*bS+dmayZ+*$#O=Ufwh56-|=w zBA+$$@8pE!___9Qp|%OMb$4{Ts!%vym-kK4+BWZiyE4j*A93oWsf^l)R*^XGm6deV zB7{L~8}i0ciy76uT`LaHDmdo6l4|}XI&dkiL+g66UUQUBt&Kt0)E`><2{3+DU+45X zj%LaolgX3qWD!x`FXOUzPtZ(`Ky^K{lZn(yMAy4xMJ5hJ_K)$Ek6rvl_TkP$zlj#x z_*h2AcZce_SXf+T&;MT$sF)PC8pVJ6ZgxtSc4YRwvt0y5YbwW>hAh2k1)y7z*U<~2{;V}cO8Hs^3EtdSD*?eY zRFv3nIZ-ryeU(QrC1fejQK@dMHCu7s9g{6Z$ktajVJ+zEb%a2O5{NS|)MWxYd4=~% zc|Q*$s?WevxD^5dF-DA*(p`njO?Je49VEo~niV~00?BczoXQ6{DA>_7;DDwO2A!*z zWrY?x*~%K!0$8mF{Rkv-ixB1GqYFbVs3RCX*eVZa+?O<|id&3~1!!4`MM12KL=0J@ z4o_v;TJbKG@K-?mSYd34b~{_#90eQ0{_(zzb6~8`BE%T}X^sm<)REPR5@uH>ytE2P zCWd=ccL*yt4{H^j(U?Y?s7#*riH;ClgeVio#+jDrR;e>7SYf7O7tgd{ip(OO`SqmF z{IT5K$iSl3L~1p8>HXXtxkIq=4)u@RdQu84fDWIe6$s-U{p>cW*N~@lD8s=5)iy-* zII=v(RfoJ(Yi*4#LI#bX=~XXEXh|iI_iqbOd)s5E91Rhxn9*AuRaq6dwEyaCP?<2G zlr8dwh5hZGst*40xM1-6wD~*ai(3U1x)jId?15N# z*CJl;yMGPcCgRA=V|JDvxdH-Jj8p;q>_-{V?mlJ5sDpXcm5L|uv}C?``F`z}0FlBC z9hB>t>*+%0pG?2>^tqgqxWn4`_RH@0V86NS`oKU?30&vU;dP}JR}^B-)H4c~h1&hi zeum$vceu%}*q-t4I8AzIu~xsvFR|ov!!JWDWD!RbjtEF9L=##ASHXV!SK5L2iyq9h zqD>uLZ5XmKf=z?&sX0fJNEOQXPD2&pf*O@)?8C6_BI%5@WI|yUD7W=f5ZCV*^`0fe;wDak|D~x7(h@!#8i1W zK}Qf+R;>%Ug3B?&wv_m~)H~9ZN#M3QQ5v=3;uq8O;k2MOfk6dTCtBOW%|0+;7AHN3 z-0^qCig$(UzU;TFP_<;CK~bnU1pZ0T6=xPW11Z^^s@jxsqEmni_BdYD_y6{SmoQ2M7WCZ-qR)Aq5ZGJV@JC*MBn27g?J$<#)YKk_ocC%|Z_xHCmN zDs+^R>VRODMTB7mxz(r%eAFJgue{7``UN-ShEpOSR2ppz`Kf%pR)=5|opemf+0tgB zj8%>-`IkZbw3rWlz0I4>Ky29-9YLJUFh^-SJPr#vuaZOMF@ zKLXVuC`mb#>n`qp>%%_)e_Cm$Q+Nh}#mFE58zgCG$I5MsU#(jf%&pAPvO>ioSFu83 zP7e-l&-fY#CYqV|fQsg92k0(=4}?>9&z9m%+gS^igGTlgvd7V%iJ>blra+`CF8lAhd~_FYf=ysE z^UnfHrS<0=N2o7V91%t^ z6x-=St^UDPg?Lv+#JuRC7&)r4?D#*=xgv~j;FSL|{@iVt(cvK9Qw^^Qv3(S%8LC^W z=uR1yR|h*1kEGMGhk`3mIhLp`b&x}9*Z@3_d$+~-#z*r-Q4XJ%l;JE}fv8PSx`Zo5 zT7w&t*-4!)9!F}Y z&@yy*S%p=il$G>V6=9mQd8xQNq`W;o8;d-*fuhp2rwd$u*Y~F}F92`>26aityET@_csH-CU@o2N;frI9EWpUp_PBmI=O=I?cmpBcfwBZm7mWzw^8~mX! zW^);w8io+;h;_fi^dd<&tXnrp4u+r<5^!n=IF z%|do;^J)Vd?9EU?mc;L<@S!e_m_eQ>0hF?t#6WhxtJc*|NI9d8gOw&D+JR8JVp3r$ zaO_sBWs3lW+*WlmmyB6K1Nb5ISck{oD#t|_;#cLcfm=k2T(-U$GXgGN)Y9%x)o?`D z*clUw7?ZB}PDRo>7wg@*r-i|TJNVOuMsAJ4pPyUa>u_kT1c#ASQTIoQi?S|47Pw-ueUe*$L+M z!h_JPmQ`%rx%}OBCcU5EGA(3xfdY?@L8n#y#yp5{KRtnR7v`fvgzyu9xx{=7lh_AX zla~3bQ^NAVXutM-L0|T20z#b?w$Dp=Y1%`<-cVYgr9oGRccA|LRU^A@M@g<&iY$Cq z#ZHWV2G>VxhqHH9S-rW8cz!^PlIBn08FjB$q?cYk*oYG3U}RKGb52v4zKige12$g~qcCBTZrvJY zLE3r*3XJK%XV0N^n3Im(3psO7K5g?*A$~;PLYM7me(7#kkQ{UGX&2PA_)r8H5lOx` zP1g|92zx(z!cMAaM*w;w4phgfPG-xpf?Q_YA1fLKx??OXGE{iP$qoR-Vt`i4`jcK~ z1E!H)sg7GCGuezLR0sY=e3f+Xj~&5L_?`Sn#Vh?yqvnZW-3KbDIFsd`y7VQp!~M+k z3K}_VlVbQ*rb{rZdPEqCXzYBCO-v(s2V$kA0sBN zLNrDZKzWFGiosl9r z!`xGAPmeBwTAw!>eaF?2cTIqRyKE$7juyMZ$~_SGT_Xw+xUCs-*x0YeX) z?hJ{`7U5t4$r00IydmKa@VOz!C%?Ht|7J&~8>N7cARG+t&atfdY6I`+yXsacMRkAy z&Yg9}hMB@hdfChgswJcHb#pVhNQhj|npxm#rz(tpr0kVRS*LX*N|0z6CqDI$Ct=nF z^`63_%&N1+|J4T1%;~!wKuSY!{x;f;WoQ@nOY+ld>@7Z%o~;1do*%AD7U&nK%J}J1 zhyp_4@hAnf)p3v}WKCd2p^v$$$0%JAIz9pNEAoGu$9&Uy|%QdqVeYBt|{NnEi~fq4o7x8%j4DQ;>ViQ>!()x#s$n$d6}dR0^~*E9J>O}sMCS;mf_i`@SBcu=@}B^xY!A$Pvzxu z`n!d(*-`lFy=W5|l4hN0T@8Qub7IjoF;zbSOqR+BqxyrR6OgxY1Z%kJVN)ms3EvZb z-tWQhC%HvS@Hl93X3Ls4g~SZaZ8VwQ{P6Q&-hi$G_%Hfiq=>3zg$l&rW=ENaHk6hm^e3OHH_X7LwO1)}z5qbMXuv|mIys}fIr8Xw37KYV2hk8N@n4N9XR$Gq zu+_)%)8=0c2nZ06ZQOXJQZ04}E}CJD(>%UN801M|@xfR6xzR63nD7E7`6o6lnWCt& zeJ1=1l62x{E@%#m5{&ee_;6MFU_M`VB+TAOSEV2|FIR6{>;_06{<#`jnbeFChf0Mv z1MIF6*HN;;ZGNwkpxMHzEpG-g9UPRM98J=t`LryqtqP+wa8rYxnLvhIfsWX#YAM;V z;a0K{7ix-coTOmsd&=m!ufutd6 zmPsDsMp*pOkGeJKomHbw+T#3cS$r|v#1FWp9 z+?oKK_WwK2%F4#Z&c@9~#>T&RY$-@pHWBuRKoBtm? z-91d*{&mI1&eF-<(&>NL?q4%2tN+P+{_`=cWd9%kAA)FAWf>Gi0>pnNikz&ZIsgDM z_0MiWhy0h0eAS8hN6^+{N@4&&V2Z7Ee+Ha~oRSpW z9yA65E$M!EJyP3Gd}Y-Q_UN#^0}VoB!X=mG%v zZI+@Z-l+3DlshOeMW&oR|pF_nbP#<~h}LR$ppW zS_@xV`m0Ym)wosRt}M0fIf(aAR2`G)lQ7y9{kMz$RhRFkvI)ff=FOMkPW%EC46jxP zs&~%CbU1&y{Mx+m*_8ezFjwsHDcP!4uk5Sg7@Ff%^gjJB^l=nbAT`Z|ZL>DLCEk84Nw=xo><*!m4|#do3Q z$-}mUQ0wsd1a;+fyS9I9irG;ocX@gzyQ_!i0}LZ#W*p<}sUE2zc)lI)a8_v~87Ser zL{aePl4yhWXQ0C$h)u~)oxkFr;m#6W%%erQZN|F~l$_7p!5{UGe3>~ncT9E(OIwcl zoorAP0@W>-2{|8^%p)659gn@?yDtMP1-~{_j;(Z%>`>@7nQYb%MwnTqcsdC!SYNKV zt;5p#g@SjPHagN$yLJ*y9GY5D+PEJpuV0yjEGo8_uV#sYEKYcc$2LDHa{0VnRY3f9p|WxK?{tIu^;E%~mWGaE^56yNP?r3=$arLQF4+dk?lqW|XE-AD;zYHY<|lcXlq z@JER`2^ec#H=7^p_?c-*E)aaX7wx{B-Dp_>1gH zT~rph-b%`B_miY}yYhYQ6ZA0X{aap{$A-!aA*mWkc9!(aLo>7OFa`Ca_ow%IR!25} z4dc?wx%UWN=aXW!D7j(4E&`EKI!TH;E>it+oBpYmWe6b8RMF?}3q<>oGe<55w89jaLkGJ@$Bv6m#cFw7?uLVtU#G@4<&R?OA3g^WU48S5P1eoJ)ptL( zF=igKn0hZP&(tio#QlfGSN}1c+=l+)^KcNTL+P;gS9{GLi=0uoSU3|M9`+oT>5At%XxtpnPcAU=iF-tWUh;NT)=h0HdNOnRc5ZtW4^kenP ztAtOjq$CW-`Ti{ZMLOS^AL&ratX1_$ed_E#7YjdR$DTTgj?WrrIUm93k}CA0(4|9_ zfH7)mkKvtY9?%ftu`t99DVFv)CmY3X-NG(A%6E`yWPr~?B|ze%_f)v3w}3sX4AoJ) zq)bp)DG=WhAfkZLRm| zUP%*%DI9&pi{$rcHHkNAbA@Mxj-)j<)LjbEQ@!INXoUkdH2Sw-FxUjpb(xl^$>D!D zQH?2=vl`5^XL%Z+ELQKyh|niq8Je`TV-4X^!jSHZl+wFL$B(5omWBju9qaQj_p0PKzk?UY zh%DA<4R@C%p2uB~vd=o`S@rd@{=D@~bL`SXF*bz|+x~W;6pItix6^BLGPh7ipTIf)l2zCrh%A^)pOW421T0Km4=y7!c1o^7pa}Xc2 z6A;|tx%y#0nY%F*`LIu0Zc6E$f};;uvr!Styzq4(9ed%AlgqxZbxHE9>2<1eaJ>_# zb<|B=VV3-$E!E!eBIzgAS#)as6g_~IndRb9&+Q0#mR5vwSpTk(oZZu15c_#mI4cEO zAK+eYJ+hn!ugY{`(syR|K;$22DLCNFuNN{R(ID?X`6ZD+%3C!MCBsz4i*_}jMr6IEOqsxyHngxzFMvmCVtGxRHD2^!A5Z&i7-MEmW-f z?DWD^7TqFzh*~9~OYI!ulbddNPJGJP$+bW4176^jlKG3~`TFFc2g?yjd^(x6ZHOg) zdtjtoKC^ec<6@GpT4#?%+v!)^R2SJ>3JtiHOYGNpO%Nj_UkDvaEvgCE^q7T6d;9nl-Knvn3?}5D8Eoc3<_Wf3_=p+i4|~k|9w1 z&1TNyj|cDe%jyP`ajoesKyz?fB;i$#xzW&a1E--WjCh$&-%SRYMpL1{09o5Lms$Id z5wWm6!XMTZhs`JcjRm+)NSm|h8VzFkQ8Jizfs@n>>XYV>hcCyZn6qCBZ9L#*?7rz{ zz@{(7ONInv6ct-3>;?;$TNUAr$Dncf&?hagL--axz`!RA1>g%NAxMNM$5rZrrn>2` z!SMZZct(+Nu_+`s7hSF``9E3~;6Flknx)%Tw=E=%A3`O=Mxr;9_O#6)_-G?gf^zW` zLy#DMMc^R31#d$&He54|m+35v87_{t=yM)4CH`58Bu26~tLf%GBOy*sPL zwGUN7tXt+1k`P|9X=X&`#0-^&^R~cpY(e)LR%qSw?_65VLk1qzXi?wN>Zi_IBS3E^ zAdVOJvE5TG5+SE`MEAjoNtkaZA)imR#_`zC$LyQ0>&mc&<679*EWwm_fz_jqI%!eR zF$zKp2_p6yQ_pds5p2xpItHQ;Y<};3MfGw@GHev5kO{f`ILp(6(!as8RpRs}FCU<^ za$3Td-K}RCU`b%-mLhuqze4wgE;Tuq0Kq><<3GiMR$|~;o1n`e4Zk9-nn%=7vl_;V zdPyEsZ|iI-$6l9?Ulq@yCd=Ispfu`_Avnbo8|>fk7}?i_uDbaO*O&}dDSZi{O~PF6)*Tj@>JaLlT}Ebv7~PWC5?}{=mgD>Nj+XAKi9R5w=#^ z_*hacfFT&0;3Eq-dJ|2(csqp3$UJg}u;Exm06vw5DQ$G+246pMGktRv_)RzD|3h40jY=f>^m5$%AF*vx;Iv6w|ktY=2h4-?u z>3|n2LyQd=*!y>!bhq~59-@|s)b{aHQw7T`=O(y}=h`xRUO=Q(eaZojAL2fMwh#0$&hhM(ZRNB}9z^1D%u4eN_%#K9x@)9@(?z4;*t%)|GW{h{bN*=L6x*RVi zckJQjz)*=UW@4lX$Es#O=}|w5&S2so{YmopRiIP`N#y?JVa8$>oILO@!v)*zauT^k zTZ|x)Vp-BgY+nOVM2ZOqui|1%w&N7Zdx!RcdjH1x1*XS-s8Y98@042Lv`Vc z(h>hbM)$kT{(|NJ@u|VmFe)K|iMn~wTymF~o>+IY4`o7)iVug`emy+p`hnkWUIsV6 zNFSaUxxB!;rC3Y<0=_(!%_EhJ09q*R{!Fbo;ul;<8z>=zh}w#f)eoK%dGAXj&ypJG z8nS;y6<*1^JP=*D9q~`ywGwKm##VOOP76IPuT=(voRVv)V&^c4ZC9HLZnRH~iQGLd zzxuxGd~zGu!NS~*@V8FOMJ0wd4E80w_Iw6RYi+$`Rpy(W4E)y#pP+~wN!)y zZE>Kmql}S{zlDwQO^Vl$1ReCo=UF;JlI`zFkf`7Q`{xWZo|?Q_E;2*l9{t;?ZVR|+ z5wdSJ`cE2^6ehDHyjYH^GK*$phA=fN^7g=z{uE?vP?E6FA~^0@j%Z5WX6m%y!l z#+tpACYbiPw30Onc8ZG8oPmvzjBOxw6()4U!?^hP*lQZb;EIXE?IkDTRnF8z<(x&W z4OT37N$-#4w}@eQ$zlgj-@0-ML12e$5{sxgE9FjbT}k2vIh!)}83DV2EJfuz6D?z4 z43bryAB zA0@3A+gcHEcq7y)3gY{Q^Qyv-kLu2w9OVS=MdLBIP_>FJ7!@0J!?C)=s|OXBj1BrM zY4_nP8-J=C4r9Em47+RYko54jwe+N$LdqdWX>19+(lKA|Bot0|S~})Cs5jmGU(9=h zcO}j9cd#4V8*^i8W81d5v5k$bjlIc@ZDV8GwsB+ML5Z2V_?!2P=)_wto)UC-Phz*Ls+FMsh&7HW7s}> zDwrb)AY;TMaWEt-17u_=O2t5=3nt-zWwxMH2_jD^HF2(H|DHpDh7CsNwkF_0{{HZVMlGn`VWG*+7YU5RZxuN_;*7%)6l9w^U^h;5vNHIsn%5O@y zAL}uOm}1HTLNaDoRfVMTaH2+pru@N6$%Q!cZt9`aq>-!jjaXtEox+|P_I*|b}hOkY!>gP=hf#* ze6hgKCyEIi`D5O^>v$K(aARO^n5dQ z64TV&LV&M!?GZ|_C$$QTkMqR5iH!~k&Ir=@--iYHnLR#XF}JLuYr_XcyY$SMd*6P0 zN{y1A&MhicK61Yj!+&tji+$0B{R-h|aa8Y$voL~AW+{3^%mOC7*exEIm$vdlvDG|q zA#N-PO;De1FZ`0be!GJs1j9k$yEg54c=?jr7jvo|7aQop@9C{_@+G2u9XKX<_xPP> zyQaWyZ{h3u1A`yom@*SNFX&mUQ7P28@YLi^c>S;Q=BrJ=w4`izgqh)Ijdnqmy&es0$ah^nUBH!(;zurTfdu_Tv;r$^Y2DDa>#vjhD959#c=Du`@b!}{6(t^E~8YQQ@)uh4VV#IrnFrw zKKp@xJwxLX`7rTyMQ{Sv=mvjBD{b5a3R5GX5=L|eFNF}1*tK4r@BaFB51ApF6!3Lr z3*1CL;-Zq`V&rhSu>P#LZnLJ4-y0tN%VaH?_bmi6csZZ*5oNHdk80vgw_!(mwh-nL zT9jxJK^BKeJ2_ZHSx{xYILt&p6$3==zM~`fRt^p4Qp=vO`|$K#!(po?MgrbT#$X^k zT{iJ@4b`ezcn`m~2y-dTpEJOK{)wYV%hHX87OJ$|MPnoD21&JifSBX+SAFD}HV^n9 z8{AzTOXLXmVkCC-l)OLt6)|O%O!HO>?;iozs5Gs-(LRW7^AR0jT7SoXv%>1Nsb=*r zn{c&Uj=OyqJ|HX4Ntc^X%@+BIUIIS>K6OrWKd z%9?P68BjMbGfcRXq~~{tWg{gdV%~v12A{=iaB&KGCSm-p3RJusjm+auVkxg=5L8}Y zKGJ@#O;gqlZe%Y4l-lS4gJ`J&m{@@LLBMI_8SDU=IQ@!eE09UWO224mCF}kVDhMA_C<}h<~D>v-$5- zgHn9-77yLetUx|j)Vk1J6*fzJUJJ#;aOn(KTp9>wIC?bvHJ3z>!4Mhm*~m8}-y3bc zGQ}iqjOmM0Io2Qa&Ijui)qFISGDB0!o^gsWBRmsehUsh8tdfaSbT*z-%2fjLs$|cw-}jnn(@H} zoc(Ak90=T-_($YhO9(SL?YKxTv`cSmp<_)dSId%B@u?z6om4ouI4X!yJr=)*Ce5;D zr9~9p$K{j|{Y2BHjf-oLssOrkx{bR0$bwpwJQhmFDHA>`n5hr%ScnA1Zd3M2T*(%y z7G4FQf9$zibjP+9$D5YQ!)kA&PJw)oDqhUy4M$h;{9btL!`g^)H&3JNVLK7{XGnE?wgBEq8j@wt6UL44``Dz-^pqAQXI6fF$8 z7^-eU5n7I(*z;2gb(1LHBml*%BPnVXe^ss|6%p@Y(s0?;RB!4WBTQjy8&sxML2!vi zRN;IFmLjt13j{hnRN#kS_b!{Rv7ok8uLB_X_0%{2 zq=d>KL*+XJR;idl=8G9vuKCQtjF7oqQ;2FpPYLdz3v^ZZZ1}>U@U|*hX1>jqeT)HS z6*%>Lbm2xaS?S6+q_$!CL5r2*7SUY6EE+Yekah`By?FGwx~<F(9otPkG+$dyZ-u@m{igw4{*4SOuchz5mVm(rbDDu<+!40t-JvZ zjBru07_Oz{c<@9@CR3N{s<|6IwuR%oIkeEdFg*{T!^mnub5Yz$D&GWl)Hmue@9)Buips1grZSzQ$zVtAp$+LJOI~h%+-%fqZRG~Gd12MWk^tp$JHlYN+o!~ z*adr^z$DFjrLMf(_IEa#ym>%99?3_0j^SLfnS}7&l4a7;ump`Xyn7j@RZ&H!$UX@B_>GjFWx30}X z3k$?7LWaf0_;PDJcI;R3DC7d+k&b0qb-bYbAe=cx(`gwK=MjIq1GEdNCr3x8iA{b6l5GeAD%m*G~!2SWW zTvoTJ778td)Ve;ej0E6cTN&}8lJXqLEhHogjC{de`b!8V@tX%K^iA6)N`{%yTHE_K z`bt*9Dw)0dD3bTLZPSS~Zr-%{6s0e*5kFNj=|%jFj)c7AJsVphxMeDI5?#d1Nd_NE z)R01PMYlrBgm;I|+o&cjg}}@)blC)bqT-zlzq99m0kQ)!XcLjb<5Gy&&v~51pgmU# znY}J2Puf|BI*EU}K|*ObD%)!eQ}O(na+xTESr|CLo4>*ax-|r5B0L4V$F|ar&t?qF zNQ>Zw+EUGEzzCT$wr2Mj~R5s2k zW$C0v^{hlBNo=aUDZ(9-9XVue^Sz+t=*h4KEzj~3 zC)j=eI>)PI#~Av@MNlzV*}r6wTD)(Qls?K;9hH{J%)QMECR1vNv?z^FbY%BZ z*X$N{xQag`Ay9-p9xM!HU?9zlhq@4SeepXT;VQ>~6}s z2&&$GgUe|ObAbSkNU1-FoA_jpa9ES7oAjXsbX*IPlNP$|^&}vxAWN%E{0VrrwQ|47 zsi0rll#OoLh@73b#@Vu!7bVotl9EPhFwl&{r@NuxZb2?{hsPf1l$ z^P?}AQ2^EXd68%%givjcK0dJ()dCdLRle*_Jeo$HDn-=_ zLm%&zB3!|F6FF-Yi;jBC09o#Kwf0-)wK5al7n;XO30wg`o7XqYYmILClHKUERF-UY zkpI9ksgN8?tx*1nH)UYZuymi{X=pRG$c8y(qcOF4pu#xEU?w1!5cw;OB%;gH!uu#2 z9~MSz=R5;1FAMaxK` z<(Bq5G_*CEsvS}cdIoPE<)Cwr0v6&pXWbdv2yPG@r9~Vy5aP5=2vOJY%vZwZ+D*Eu z_OWDG!BcTY7$%vsYO|ulvDqnE*!n1DZvBhXszviiKtx+sQ-MKG5UNkf7GIKVs&=l%sH%d{TR-U0S6mwM zceCs%a=5~Lpol)__wF<(qn$1a?<=&?Ou8q4W*L2tGB#UL-5=p5fXa5UQRqBkTc=&b9g|4}l`IYs%UoddEdhjrjdA~M#on2)=J@~pCf}bLy0ewxu{Byu6Cy=8 z>@da@ZOy(HkUFbiiMt`q+aSL2E=)GnCkCN$apW+yoTHg>5{(pcyejIF%C2B=6G0p$ zAL&1<=EbZmxV{i-HJng$lb^`lij>72 zh)g{*UA3&x{$fwYK*zHrN-Dzo7OP2c)@IqmZk*igU*Bd$WP}!S#v_x>BYg+MPDAuX zm>WHj_eiJc<9xlJpWpVY5$x*~O%R?DD*+~BfTIFKku(+;wf>*1_!#`)AWC7c z_2Shm-9cHfC{$5Pv3x;F4pAtlw)FsqDzW&VU1lzQcOQ@xabtnAW zKtPC3AX1NCWw0Xlvjy5d*;a+1!dq3Qo|0@)=&v}_$ag6Fp|}fN`|wO#@NSK$$xt){6|b5i0KsU>9BhXnH`Sjgs@ z?U(5Tbz-WG^oeQMsOLEQU&}2D|6;7pNl;b8g~kIEu78bfmT6921PZBM7lBnofGT+8 zH}7JvS5*|fXLJ>;1fpg%sEG%sf)4TNo?X zj+oQ`b#_dFzjgOvE|4A}>jGLWj?(v@aCm`B%KGK1A#VBLH8j7J;Y|rzp<(OL#GoYr(_QgDUOv@Q@6$iwzO( zYB4t=kPKKhb{61%aq%i|dXY1MRpo2p!6@&ek1S3>6KO$;oVr_+&GWL&gO4HMtOyg; zBmC$I}dd2IouS_s(M#&i2b==59*ka-c6kv250VrN}Tk~cQV>14=#ti9WYmZ}p zwHHwh^|U%UE!K7f>zgx0-2xBtXvV+n&FK4GZ9yTBy@YEA-NYc2H|xAldt-EDxz?i+ z>#YzKlU0(~0kKRcS(^^s@Z{vl`OzGg@et6H$hkwfbN}61cV6 zeeCL*i;^`B@hv4Y_kGw~q{_O(oYN^*t$nuqbyUD*Q_d-dcxt|+B`FBKjvAR9&+kEo zp60b$`(vv$IzpMKV20zta7%0D`qFcTIAPqT?Z?DIr0b z8(i3~_u@}y%++-*_)FwSoR!I~by?k7?%d*snq9G@ixXPT6GSMjM-%Bh(7J_CY_sp( zv-nQ=L06-mj}8K!7E@|7D{e{aNxnxLjZ@Bbdy5A-ChlX>auj+zhqI+diaC2hbWMU;{7qsn9oh;1(v*F%HY}li zJqJAR+kPH&7=2oJNGC zg8kvTi-e#b=sm%%a`?5fm-OklRZe0iGH5CzYU#1(k{>=@)R2m>nhHe857uLwl{ zk`^~^NTvd5;GNR9FijMGktlYKkW=*Mk|C-v9~cbVU|-g+RXfD!EV_1kx5cQ*vR)0* z3!p{Ls2TCgletNVVv@7CZ;cC&N^Y`@d@O~LPaBkJAH8O^Kr=kBeP)Y zh=_bx_5k2Ugpzb=?+0a;Q!#!V5iRs8Q9|dy3@Hlr#u8-eUbK=lc56sbt{lZDiQ{dl z34v)M$!uLCQQGfX8eHg5YgpqA4N8~qpj&pEsrnT?f`uqNh#-xc(mSx1H)YApv-L51 z!`T3{<^_FT}7+LIl7Uqjp0vcIkTpVG#hm#^>#gUH{i zJQTHZ5>4&IY#%XBjJ<=$m2j<&?j?Ncf1KM0J{(S& zi(0$m5nnE5dm}FZe3F810bSIa`O{c;#LY-+?d{3Rfl3~p@r9wzONtZq4ih@SFCG>; z&?@7^Vy*~Ks4u+Vx(YCn>hr~|vK+vAD<oo^MkAlGb%`vS=!L;vS#l#3K=X=^@^u{$Ym^tzo|hOr z!Q^6j-mlJNkXEm`Zc+2|=}qCa21x*032w>b7V4&G?_qKShQ_hzm877p#8VZ7o*IVF zR-IyZK`uwDky?1@e8P+#`#c`HBI!m?9_=&Dqc{Tcb#$es-xlQhimzsV&RASshoR9u z-(rmI-qh#O#P(Di&$5l--+QE3;qpVrV+r&gf!(;f%_)rQDh#9QgOT@gg^Hx^3?Jwq z-Jo?P0J48d+|Dw5dUaJ_%EWf#xa`C~N_Y1{F?LNu8;Pi1n+<}oc}XJj0E7qzS2+mL_3RY z=^jGRz|0Ms0yibb($ewaXVuqn=;)K2v-nxG!!tnGm*W=f*#R+J8Gm8|md-3LqdIrR z^^2D>j~^`ls~Z~(v8VK{L4wOSg-Av!G32e~&Rr~DlSjCvYU6jlzUj(udfnb;gLhL) zazo{L*yMV>oNm?YI&A2Ir`{(+hkBg8Jjq73lWzKa`v@C%-UC;;ki&F5xL@MENG}a? z8wt4E46pT5{=LxT0UGS4(sxBe73OnH{~e=T2ZH19fY#dQ!-hDdda#0%#n}4-{U%`K z;%RRj7q7Sa&4J8ic;9F@1nHRktmiLDg%yz}r0}~?6yHV@fSSq&RJ3eNF9Pv53RIb2 zrZ>9Xt%7J>J)-OR>9VFvQav`CaDuyx(QjNhCkbdDG|~M;U}WtFV#Gu_9Cv?1m`o7- zR)*Ugiu@*a-xq8AyF)?rY0qQq=kSq&T_G#&S6_&N$e6MRKpQC^=uHP(a0&*(>1?WB zSR4TMT*H{`re|j)1_kj~+0IbKZjtVNX&(;-Jz7TNte_|S`rN69+B{#?;I=oM#sJfC zpi~gxmcwQCPwU_z5^*z4-Gm3tcc0=w(EWM_vb19Ak_Zx2Y4!E$DQ=ygQ2iMH`x@>; z4}s6z%DVkRO`Ls@o-!QEtwZXGtlbj=FCjAGR;gikr|Jv**OD`fixAdeeUf{nzOMAl zArK$AWhlW!vSM*ToB~I7*C6RuiQ&g5vt@4K<2CGyYx$r&=gj<7ZHXU1JZt7$Jfplb`bsh#0q=fr(e0QoPakZ$OhCEO=o^QbjktN{tye zCUXUP3TGN{+!Py`lEO2mglefEt$8O+`%j7wI>t&m!V~W}gQd^pUvuzBk$n8c z+bNn4AK|^Mb6QTQAO1X`EUWnxU^hu)?>+__I=&C9w#>9hD#uI8sc;!V5ow=lkLNZT zfkhxgpmBqt0ei;XsKejv#8ndLL3Tqg%PhcZbQL> zEK|bx`?}e}goTdFQpg!!`o?Bk0`Qz$x3%xGo~)N!0v=Rc?R;EGN#xl&@6PnGw%t@Y}_man(8K$-TC93`6o?w*Dfa zFm{xC^^ohQ{T01W`o4-svj_hxDbGdrY4VkNB+BOilvE#F;^}Ip5_ou?KYx||sPaH` z$$~(9GAS?%U@91h+a~I+DYx77OZb*-wu2`Ttb2XT!}FF(Sa1gxpE`i9uQ9bM$1I{1 zt^t-_LRD;eMH?2(3!KRK)^#VrxD`&^E`cSVh5Fde?C!>KSjg144jaBc4;)+1- zcz#=mX|0b``M4TfznKkXLn2{?@^*sA;W(6E4F0$HUrC#pid+QuQ(~p;Q$Vk=ZDwwF ziz5u-vQ(%x~JrhnJ zy5Gw4n}SVIK!VjOw|Gnn4u0p-gd_bpXC|qoO#0d8H5_H;fAlB*UG9pO9wLH z6!>e{rDO4H!pvx$zDC%a&JY!~MSaLlutMZ-MXJA0J6EE&@4UVeV~wU(IH| z=S4!-_U4$*IKGJc+U7EtrNO%I)Mu<7$z3(_Ps(v1^N#X*OK5yGeVjZydTnI9)eU>h zq_dS-OX6tnYK44#u6#J&4CRHRM_(O&0goMU{*?>cPihig^4zr$yIR6N4HIZk{)ZauUfe#pcfO|Iea`H>9)7(FgwmzElt ze1dSRn*d!eCs&A8>n0MQd5=l>obG$`o+Ybe_I{5Y!$*W?^v?>$tWAqPqx+fFM>>j! z{v{e~OPi1%^Q0?L56BeC1V-4tqw)LhHudV8M+dYYQSMp$zG3ss1U6r0M^{ihe9xXv zqhIyCgbKX-OGG9W8PM5P!+q%i&X65HR-WoC=C_?0Om|P;buHo3CX8qvQ zz`}_&+8Obly#Ru(WtCfx$5YP9nM{62qr=I6Q#wfZ$BheXTg(WIfS%binw4AmF}VP| zG^iec#p4T*+KbWTaZX`|HcsX4M*AvlCBGKy-T7!y784e<<7}-jKkaMAPwO5NSHzmq zA%(#$t;^ATLNLlaa0vby=CQtppV!6s#O<`FgyK_K zNVmI@oa=j-XJOv9F8_f%4RrbpV?tjUi$8=nw6I%`OI&eE)v5D48XO2dN5c_XDSQ$) zPt?ziIau^Ygp7>JE2Pia^Sapb2)UI#yk_j`ea2&#G}7uE zby%~I{oqY|?QQj&S_|>Air2MnSjbOyUj)V;n3)XSTJ%z~OKdphUr9|U#Sv7NbDg%% z+@90b7cOCe_HW+S{C34`9_PsCCyW-(uQo>A{VM8Xehi!0JnH=PNiQbnlkcrQ?jt7Dra&}~&= zo%Af%gwH%G<;S)}ReS>yc#c2W%pQ!h?mKdw2D-izeLD~i@m;18F(|mL^o`d8R0Jp2 zM?0C%Y&R%P)|`8af8$Tq^wMUVqf+Ts3I$>ub;DEW@CJV&|plJ~h=`4(9T1$GSBPfiBaq8u#|f5Z#f4VheaTgN*) zZ3N}pH={e>6cAD5wJn9#s-6P$xDMn>Pc(--fIp42pYm(ma}$1R=Sc|}6Zp`= zw9+3i-31JtF19o%$kUyc;leI0yd@6qReXnl^U)=ujtOO7$ZV-*SP1D>ujnrrf+4wM zRdG;8euE7;bJWQ;nIC}sl{aW#rJ_LFIC_fkcfO5(!@dIRKu#H0hUlzHqhqZG-(u`s zNy4?6v+7=s9i+=LtZ=9`17mMwS*&RSaGZ4x5j(RwSHBE-;^EZBEBWZJs+6yO|GcNL z`57;%Yb=8P(Y~KFzt)F9rFD68_CL-t{Izp?6L;qedV5oT-Q|&+E$HPVK<_nr{H>y@ zQa#S!cxmwap}*deI_*syzQv$%cyznJ=xp5(x?jL{dTc$7Xpv=afzjYU1 zgz{Evz7}g%mY)6luNdhA`*J0=xd5gv5)4d5NlH{m)noM{+d>y(X=!Ll9354Lj>=fB zKahY-lZhrW5(keaf4LTg2bT##Tv17Api1f7m~+vvFECQ6R>rs|67egorlgPO&z3iUksoLlu-%xKM|oqGQwtmAdn&0%r%-G}Z=P-rU0#0>8-*d5$n_t!TxEI8|0S4c%Ks!HJ&PV1a=c?+r-0seN%5F;!^(XrL zmP_r3sxZ6N&$Z$BT|KMpwwsy$(?Ha&k60Mq1~m3i)2Y(v(pdXYdB1@jNL> zT{`uur0|Z*ed;JVq9Zm}H(SqdcG7NOiU<-}B(_EgzGQ|WY3Cvx^zKUze!KSGeLQYw z{37&*$NTG=smiCH>)P}_xd`uqFegLLYGkbafEK~dWq&{anKi%L`-j>VABA;1`!Md8 zgmxwvk|?T$QzImAD@;_sD43a0qU<&9W_BXO1k3A+w;*Y$WZEf9TqPP(r8YC$Y@H|l zgmSUl7qtWS$aIjc1qW+Kwo&a_A2~ken9gv*TWx;w-o4O)>tE*czdE{(U;R; zzaEk1Sl6?DR&x&=_&CqnyTY}h!afir@V&8gk#cF)qWC6K@E83>&56NL2NaK zE?1edkfz_;nBb70gi&TDvTbQQ1yiGy)lzY+ZB+BORr3=3*Q{^OMLp^6$5P2=YAofd zKLytLT6{4NE*XUNt{hxT7bsqgNqAj=9UZ`|o<`yu<9FJV>ZZ1s%f^S^wOE4J&UPOp z^okckbgiHYp~d0O?eqa|yMhgd&|k~D9xW-`!n5|sGo0#5$1!x@RN*UI*Lqefx}Ujj zRo|y30d18x72g+;+PnKxX0ZzOuBkg-R!SoXBM5)M(_22Sr}R{!uQj0NoPCh~5fX&V zigF7EXri{4ZI=ZaYhYMwt@b+kQQtvd1aqdVIW2NM=#;lFYDDh4zHRRx?hjrnM0nub zFo`~uk2FvH?d_WBnsE{rwM@mE_>E_lap%c-<2Z|wJUV#pKycv~>w}@YYo2hFJIC+; zRspR2{kic>W7c7EEZrVy^~+UZ74kO}9w)bW47NZooF1#YQjvTf#&vwD)o=}KIw1Zz zGfz``5d1{FP4GoXPE;mc6Z-4j=K7IwuHGBH$?5Y!&fVV08SjfH-PJKQZ_#D&>YBBE zwPx*2A5C)dk7i-h5cq@ky~=QqEbXJu@~zR2_w5{!Vx+IIYmYw+KgGm?Ek*-LuZ_u*0UHa>CW`SnK6buRDeh_&UoR(j2FGP60dzrJ&}`_EUE zoEy;p>7Ua1t~!(An^M4dUSAXEghl+M1#H?NZItP>-{^hMit*s5s0tiWm+iosHvCll)*3UsK5J1w2P7omxt{i8`UqN>1?sWrPAK z_oZC5%!EdvSWdn#E3{O#jGu<#1W@U$;jVR3QwP#`SwHyTRB`8{eYB~OE>td+zilw! zEqyAoZ;72p2!^IRtt0l@(PHJfJu$yVO2xNYq|ta1+mQ(YMrMCaXnQ@F#O*G}lGAzH z+a*es)hy92iL4K&CDg!o>i*9M#UONyym()j4QK8b;k?u^{UooGhwGHtrQZHC*ErMa zW~^k@1GUh{=C&Ygt%IB_09$c`Q@ln_`Q285U9SbyCKhq@Y;LPbjopM9%H0~KQzLvq z+g~+a+MsHeWhBqWD&KUTP4N7Lc9_!yu^E$hOwOBm z)#hrB;fx^t9>rC0z6Fh>CUieHK3dzKKahN8IVaW%Lc?6^7mEIuCB5!Tns3Uh>qSOQ zQwx>>*blzL?e6IcUfp*m(~kvqRn~umv1ka-K&b7oLp)!;Z5k>fkDn{tLYdPi|8RdC z%Vzfz7|n??nGT$dXaK_tp^Fwb6SzGZvrqqEiPiNszPNUktBJ{Rzq#bC95XwyCLx91 z#Y~6BPv!L4nojlwu&}-Umu-=!A>*T$gVN*+;`*a09*)f^_;WIkF`b?)7m9mieP1Dynm)CRcF{1-B@*!U0o6jiMB~A9itQMX69h>ncadWG z?ZjBMoiuEh8o*An;QyyxFk6L5q3MV+K*ldhOX3So1*ZW~Gecyu!IS!lzL6MQ(+8}p zBE00XbAm-`7A+JXWIB=>pnBhUTBDblCaeNyg%%upOOHq8hHDEZm43lvJ>^ZGR0&XVG30CXxl{-EeJ-v)v$knra{K7DkqgD z`LsmMGLNaeVgi-9kyBTxlvqlYCb?|BLAkcjI9pAgh7uE>P5SXv_;esGikZeBIjF4D zS9qaZt-4J6=PVMI_GfHZ{m4YUs5qT3QE|I;(_7JWie87mWbcE)@&=c+dM)Xo^i)uO z97=I!c$%V+gZjE_b6RFqZdLAxRivuFDI<{6WjD}pK$%*8&2Odqzio)@FrCK61e&Yp zFPoUj)c8`W`r*f0#|6$rS*l%RQ*HN=<3s+On1!{Bp|_nIY#dQ$VN8*mRMe^SN^)^| z9=qas{`e77oA{<)$aFjnA^dZ@U5pp8NHretce?Hj^`GO~w!ejmjrXb%858fd&-#i~ zCs_;V#GYXm4#9jvLYkypF<65gJi`A#AlaK^iOq>em<-Z&qa?vM#568DN2rtTGj}b$vZ-Ut^v(Dh7ZlL{G~}$o)G>MV-7F; zc$xPyyAQIfQuFw;u=xfV#jV|rExGvEGJjI*~9r}*M-{7?Y%F6DS`;M5gj7uR)qfvQAsuCOzel62ROGZu?&i5Fp=XHPX;ix zwwZ2bh|8Ut6R~q`7KfK+T4TcwAbsO9|CdKh7j1N!4YPTgviq`gxhJrA&`Mv&s+u}& ztq08alyHZwZOU~*k|@)Qx(tU%DFqk={$XK7wMbh3-n)&5i!nLt`3=0~yT1Sm|vqX9u zWVg79BS-jIOt?(-HzWuh;3AIQjL_WrWrz<`j4L#>P0*w z88w!B9;sU2jqfqPt7JcAOH^RXwsJo&8XU-MWG87a+@pFUnA&X_w0p095l=xV)4A50 ztRJk*A4wlcM-JQK!n`k zE_shPbzM@_>aTh}zAQj!=g0OuW13g+PMSHL>OM}mK$x8T4+6}7tWeVNIf$r<7YY@* zyAItb=+?d7<8AKvV%@D2x2*=0$U z{TXJK>ZI4>p|kDIr#hy0E!#7q!aUe-dADr!yDK)Iy^iMGw;H9i_gscthDsTBV0*9@ zEhef8Hp|(UtNK^qKI+6bTf~kevdhwk-qekq4ri+{b2XUC4U2$X)`|20xBiRpC|*=o zZ~d9(6#J=;4OLmSs|5nJ1eLEcSs%T)*(|FFAXWA{)9JnKqnFD@j-G6$$3fw8u~c{6 zbuMpvMJSSyD^-gs?sv?>a?wb&HtIkrTO5M!~=hS4IJstA6`#m3#J3a_uYNSgp zms7>~Q=QE{fXxy4)I&vCU9|?isoO`1o>F^JxiDqO={#peW$m923WMZJ^Lne1y3cL7 zIV~qZ-SCSThT$@%8|bk`3N&80#2z^a#ipOxevv#0U zfJeyJ^;FkytF9}ttSe{_agr&S;2JZ1_h{$xOe`omxm?<}`Xf+BSWB z*V{^tA!(~BsBYqj=EDae4Cr3EUGGz1GpKK@0NCuAi2XltS+{f&C!C#TSo5TEDz z`}BGX(NsCpigC*frL`*O)`Npdb%1gT~9Yf32xv$P_-I8m^(>GB%vm3gHn`8JWGAf8*WTn4F z_G&*pwMhwOhjaIsP4E7?QAx=wzKVO^8DL;gpGvwB^ossS6Z|92Be`j?>1j*h-A{Ro zjEoJMx~;~S;pY6e^FR$E^U_T9ZRuq3%hx;K1J(+w6DyMooxQcwiS7pk1f4%7W|uA@dj6)%S+WVYJlV~@Vurmj%)MOa|00- z%ZT3xg;yi^-K))m@~)QYi$I|!4ET$I$M1#dUJ?}Bw?&jqyqcq7ue~e~$UzJgi<<&8 zTLWJFstO00QB8_D)Oy${*VKJA=mp)qcPNMt(= z$VHa*C?C5vw;;jE$`0Aig}oyZZ+~jneVX1ENk(=X@{1pf~%ljN-uNU-M^w$!jMiV<|Km1*YN6Vs;)=wu@#nyCWy7(kl;AduhiGv z0U8(|K!~ERcBT|`mm|1m)*k8#t=-!~ThOubHKST0Ck83g4ClMW%UHFq575KbafA7C z|KbtJXy}~d zcMt!b4~ZYn$}Y(k2h!4cX(IfroA4}k=C-EwU`UqKr63Rn>p=vTqAjAda|+w`Mopyw z`Xx#9cWBKFl!)~^cfFrn?KK?ydzi7h^)DT{4kKe2D)Tn3(b{&NIssY|i^pAev@v1` z|31lvJ8fPD_Js)4=pBe8BPF}nZ@nL{=P8AWpA)+2agf8)YLbu@gQ?%8;c0bC9E5&ahi~AaWZ(G9d1Qp_VwwoS3~h5l#@@kis#{Cs zZXF#j;Rf#n*;H#Hh3uBUT%yK=kVFx~2nTaN;_5kh1x-zmH4|<6X?l$RG4uN_5YxlF zj77*=bT(xNPPx2Dx+$8S#)-Ve?=Ej0zxc}*I(KqOyTa0QhjU>xj#l7Oz@cSh$4%et zwd-UngbymW5@OB&avDLRj7NOO)sg<70^AV1Nlr6j2=_FWxy4j7s`QYz@?;Zt_7MuX z24^tb5mqA7euWHi4)=>t3B2dOObXZQdgrRn?E}2(ypqiMsq$p7a{p@c+ovCuls>_I zCa2n~%21=F@1yPHVG*=eCcYxpGTsqn#&*Nk*vkTawa{Z|u;7s3?(V_e zU4pyYG~xgE-ZN*;zB(6so@YKTzHYi|m91J?s|Y#*T=Mc6-2L$8u?Eb&$zn(r8f zw?li_jlgIy863B&V-QvHu5cRZQMQb1b<==f(Lh&(V0LuWLHR0cOt&~$%-VRNAMI9e zmD`YC_q&^g9kxeyWr#E)xD=r&c3ljVCsG0T)Z~Y4tFy}Zg}L3J8w~U( zx&AA$^hn9iSIa_XxuIy4^YvAmn`dI~_sZ?wXQIwZjI6LI)dvQb?VXp6`3yE~V*lo7 z!w7yjt^pdeivu&!Lmcae)tjS$57OU{%tLt#~NlYK{S0eVWZ}w zeob^n95~dN)b}V(#u2uph*VwVb}7`8*XQy6<{B3CmSdiFZh617`t*AvE*Tv~YFt}% z5yVma>mNbZ;2pwP&zW5En%z)tR;+AnRO3&nF9YeUN5}<~bX30PdgclZQ*K@p!Hcz- zee4)o<3A;g8B=?dxxu9zZ?NTk%?!z#E7@nHn+)Q%=nTY&&y9hXOM>n!=T;YIRfAI( z#$B}D8u$GO9lowAoIl$CHjVnXX()+T98;xReRvs@SCZ*DhxL}(S^lN1MA~l*gs;td zPUg`$LFk^(s|2`l}21 zHP9a&);5swlRoI*YVL!?iq3_Zbu^(7kf}C{KQiD|1Y-NGTr&ChAz)h?fe<;62|LG7EtlV;ebN%7<#d%i)i}m&>UIMwP z9CqJQU#R!rwd8vqQi%{L2e&W7ahb=Bdwf4*7-~;e!6nb=eZA@q&9@`FSL&3+TJh$> zdy9W3{_S|_+YDd!%m&k-Fakgh5k*Hc015>`nDwtw2|HG>m=nWl^;09(^SY+9pLYLq zXwh3<40JS%aZ4JHv~QZ#b4p9;r<}dKDx6dL9y6KFXKGIsi0!1M^eJp#4S_@(3;2WM z$%|&r0gD}p9>!8wowBn@rs4#&B0*Q(eQk)W=qVFGQoubqQ+x#+EG}ALH3_K{S|?8F z+K-P>78nOt$LX{_bC~a>sZ8C%tDL`_+n(48EseZmn7S(}DP1;6%HJ{)L&yv8tIUT* ztGaiy*wliK+kzh^W;hqH$(d291Ou5TnmA_!5spG0cqcE@`j|Z?SyrpTd=9Nr z{bQDK6{Lx9H9f3$^wnJ~TLeL!;yl%nG@rV`5hcwA#)--lh>gTVf?y@|>>XNOLso^% zXHHo*_P&l2Ta&&%1^ep~O%u;0X029hcQ!gU>Io*MZ~jHt;bM_m@UfOt1YS0fwyS-VqgT&V8 zZrevi1pntb5#>11rDT;OZ0A_RoV-wYWhIjwo|A>%kKp@IxZu|V_PR5R*z79|!ODEa zg8XaItuRAFjM*D}T-cx&%VjgN;8IA5YOm_5lT$?6$TT#3#NbX;4>q>!+)GO95VyY#kctSJJNVO`a=~8NSgs6i=UCd zSr%NscgW0828C2#23JkWc_W22@j>&3YfZQ?s02eoo*L{;}q?$=Yj{*u{xj~yta4y`B-{Jo46cBBW; zY`eWKfM$he%U5uhV>Pzr zx3{3GGW%dcJZLZ|i;%^1Q>0L-GpZnF8b;V>;(;|a8w+Ph$6mZ~9;>ARKE`1|s|avD z`j1ARjgwx<1BTD&^6WsoCIcM86)p z<+YjNJLvQyZ=xH|$e9@HS4DDl&MInn%9?k=bUL~r9tQ$>T?O?b_I`r&-Fe!$qF|r&)mh#q zqs|Y$=FLyyzsGr^%;r-l?}Q;iTk_Sk%_- z>)v&BytnkaN583MId#;|@F06$NZzk>NVOSsle?8%paU+|Kv1_PrG65e5kW0% zzG9d=QyIqE-*v720eGtSn8NdsYq7THXXu{9lP0;)88&K$N1<}4h<_J8{SLduk}q@DKh ziox%ezVFR`SXYS=bgf(Yvm7m@B*DZ?Ye5P+HAAUtqZ6`@hWBf?Y5)y@V8=t2G|tA zl8Gv~Te`;ZW2FH+I^kr5yD+=Ax$KdfOVLpTU^LM~QB)$0Ydno8!;YvStJL}LsHGB> zuBwkhIobUQrfi?7uwtnAXlpf(lcE)80uElF0B@gIMwpKyTvR)^GB?|=!nNtYmu#pe z`*ReHzLqN)!Kh@k#3;%z-jBINtMFE>j6BKoiu~#m$rXEmj|^5h$4DQ0nAmrD6*_D6 za++!BFSkGX!2j|1@7kDgSpBqyNHr)V`@%MC%>0>LVdXICcZJ?Y+yTzGRgf(C;2+a)|R{VA_X7g&R_M@!B1A+V2aONwy_q(O?yp z7fPBW(98r_y+mFv5mic5o5CNskjv`K++@eJd$_kg(BG*ZC0F|>TQ<$S&qxs~8H(f{ zUm0k4K?GzxTy`u5%$w@jO)u#nyUb?{?gZs-ClLvZ?tLYuNOQZtR_I)BvBQ2(IL{(V zx~9Rrdt0&lq_OPyf@2VSYPs1=o?d2KB_?zC*qi!1Hf@7Y5|q~X+K2uAqOYu?2OEzU zm~JT$WUpGW3|l#1TeQq`YA_fvg!7?;;+pP0^Ejv7#mOh?&Al0CRRi0uEW|A9Powzp#^7ENgip_@qDhAU6Q^l4dLHhU-C^EjNtT z0yT@1)jd%^hl+x7keYvKZ$8vq%-kc+TO9m3yCYzJYx-@9X~RQAOWyYth^y5PwkNUq z?w!B=kZ#!slyi&i?*0CRXp|_b=R8ZTlR+}aehuIFA$fItHPrzPAI^0Pf*INffPO`@ zT^C5VkAHwaD4;Kzp3(?4~l&(uPfMudsFUjKHqEXwdiB zR#;SZQ(dpY&(LF{E}UiiWpXT+Y=Ecv2(yUG;UI7k5=q- zO(Q^VACzwFD$DI{Ywuuh7}ZIc=4CKU&39#o;(@I>UITOl7J^(_)~X~E)WLrm^z_7s zBbq!l-C8ca2M-H)GcGUZ=8NVuIjawo6iLX*H!*#{$1KeYao&7hIzhciIQ=dmN74T) zDnTQ*xpY`~bl>;5uDlI!Jh3((c#Fs?d)3a-j<~{k@n^fNHOq8IA%`O2dv+lCY}vYI zyP@p`MNcxFg*$N*y3|U};gKRsc4eK6h-*U-h$s-0DGJfLuDLSIVnew;?e^q@_7O;N z8-}qu-*g}*|JdH)?8Jk?CRXVRVw#kz;HG*SKp42WoAps%kqn=oiO?MSca5$s;v z|Gw|ikVvW}(sC7?LVGF#nZFmn5LF8G4B5kd5T;?2!01KiJ;qavGmwEajzmeR0)333 zXaS&$a=sYx)GZiP=!xylU|l}LU^qL^V-pePC0m2UPuNfXb+s-~)Iodc*h@sMvE~4E zPR-*;w~Ui75e$s`Nv<+J2pPXh<`C!Va^M*OHSjT_E%_wTTuaXSGqa7E_tYANfklCB z04P%%!Y_eg%hLi2j-YW^kinD2bY5@}`^c0NsYc*VezZ+jNt!Z%Gj+{62=zuLnzLK^r3w?Mo-6t#yQ}am<7F<5S(7J3*0S z#`YQt4Ye!fCcgtx)wt`}ZjqM_8bkC53p|agL7LC=b8lT^q>o0+JgJi4{GvN4@!Sv>vs*@(U&eZ!MzK;9+8@K&o?e9C^cZ?9G zm?}L)T70PMxLX0O4}p}6chN3KEa?*P6;&=^fovV;PZT@8Jif5Y#qc`wl}O{ygN>B# zEhp-%qMRbqn}&%g3Dpuqp;Dba@~*4sJ`#V^(Z5vtjZ5C8U&~d;GHX+IDK*cjj?S53 z&Gne_y|i!8V+DUaQI?~?EEu%U6(JK9#HG2xJTi6U|0Fp>RK8ZFC%CNk0V!AoZ8tL+ zqOD8EbIt8`ipDA`LJ=_TX#JQzHKV0{a&YovdxMR6U3+P`(zCy-vD2itASZ{GMvGap z8!ln4kPJ#2Wy`A0(C8RB1U@|GK-7gkf36+2UDtYe1R-@wx;C5DNmPUnCQ(?P?cDZu zdOBqkFVF4hA07j%wXnk9vC{IdC2IS;CG#+hqcFGdVxiNn5lD~sBmpNVWH3=uTtHkvN`dXD?#f&`&A>SLu9~RA2l@cj^ zQxh6GD!e*^%vOUJpR(HKi${(gKkKWO#{Lm?dTV8DX{iisjtDN}y3vrqG z1|X-%FUA>n>7vG3ZuZyo(MYXn9NNYw`Fd19i#O}j6!bjZv~e?3S(-c?G&V;~6chX# zWc$~pa-3y!Fs|+4o4io2ODEcDh%=Tby_6(gd}-Hw#tG4*2w523v=cO?=xIlmCeU8| zCs>%}lF(TZNolYG+0{RdR=X3%@oopkp`LJadKmIx$-Bk=;3x~Ml_2Nnkf;WlC*OLY z&hihaa(bajA(JP5k!%o*Ud0lwx4=m*l;j4{TLCXC`l?iYYICV6vpWxgI+_7FdIQmm zA|ZcHm(U=5S`mzxOe6JYpgj&x7MsKC{fD3eCWv`Yzz;B8m3WN*KIcye+fDB5M9@d( z9F&Op6V^NrpZ2gTB}y_5{^`#ksWe*Pka@}kOZ}vnD8GTEOMKVy#|}r#N56!lP+rlw z`>V^Nqb8~L8u*5zotk;Eie4%6(nF%K^VxiHOq^@Br}5$R?(f>HNfsd(5G7u{rjgFjUVB;qQOQ!A1Ti zD2_|2D7eJnP_g@Ye={JUFr<*6oW)JqzGxo5ln(!bni!scK;qLiwZX)Sf^}%Em^{Jh zo2F8kUzhiBjo#GA!Z80}n|P5jjO7i&Oci+P0!qU>z^L#9_NHf8B(8vYbt*>7S3M-o z8dn@_<+vLDBpZMme|Xl*2|f+7%3Iwc@f3(S0iBU`N-HLzDS%2amye4%TkVALe2enwGrn@w)9Sb@#^$_>$@(D64wm z5D(ia0|ibPO7gzZSlAg$x>t$An*CDJvg4uU3m(U5CxXkqUD4t9kzE|iikPa3JMn|? zIb*scq?ygFQFv8tjgy34WjYP~n`AwYmPn^JSkC@I9&0o8E+|qDT;#`u`^W{rD5x}_q+U2 z#BkXYtt*i0N}~z84sMW{sMUUc;*5?a57iDlRBAZOE}uTEw|s<{R#PjVVi_CXb^dZJ z@MpT~3-JH--kC?>O!$L0cmnz7L-r*`nm?4JI<(+WIV>g`l){=m;xhWGN`u$*N<+-d zFrx)~1LuZ%0I}5~{#+0XF2c!QU%XIIOzP06>%O!|>!* zE3_!l9gyMcqdAz11bp-66M_X-7ipCaJ-D@OyN$Q6!f)>2D}vU}Opbvq$u_QEjP@ME z+qTBV8tGt`_m0&%^wQ&BqxyyhUg3#FpqUmr5;ATUKF1#I-_%PVe8@P;Mi!ER^!|4` zM?xAKj@06?A+YTIja57U<%4U~g`MPAnvr7mP9!9f$c!F_^qL16Di9FxHYBEeda;J< z^6XJCE!@DWWv8}foNQ4YT2XiLT8A6rVfvKfV(OE?jIX_p*oB!o=_b0??&^OCgn?Z| z>}QAptp3JO2Kdm+Vae(E51h;#dQ>jgdh^bqx}aL?iQUqL>T|Szt4hN~8@wetnf}Y* zP&(K`zJ;9=l$XcZ`IZU}$cS{{se z3<}S^a*EjlVKG-w6D!VT@Q|3$#x_HSws}Gfk}v9zKtYV1NsjV_QuQ#gpW=i&p)nsz7v( zBov^k+V1gQ=OY%xIs3_LUYx{EBFaSi+PuB^kExyReDD4pao`789Zj<8@wP#DDMSy1 zx14r5y;Xf%T#BsV&{uDJZV{MuRl9{IZbh#Sr)WiHplHHB^-TQa&H3VJFA>#+WY+uP z%2+4ZStnNqyi#nqN%!%6eAnqSd!c2aXV+)rN##N@qoDv4hTKivLap93 z*!TddgRn)eg+eb%8oUR+XFm$UT12jDiK5i}b*p3fYZN%Qtv{=o*Cc*p`wPm8M!!9t z>DUjeM#7@`mQYf%c)3apWMIvaGP`_FDBAjhj@UH!Go4eeK!frI3oZ!C3Wb)AFaMc8 zO^X0CmK%_!iR}mKIvLsezfpcKTHvHs@*V@#|%wOJd0Q1G+T2 z!Hb(oU5U%rfOxZh)p;azsP1Wi+N|^JOSXS_anQPn0xX+*i+|vQ{|c!#tnFVBC70dK zlSIqiURpv82V{8^w0Wh%dW(Sv?dIcWGK`xo@p_O!5rHvpJIaGL`RGvmYe8MKM%WZR z%vL&gW`j}e0&DgA8y7DqhZ$k$+bX#G=D$Z}HZZizc8PfrF2cmf*T@h_P#9@g5)=vZ z609bz3=WawjdP~pW3EE0qKJ|~!a6o6Y7@vQ#s{~)9=i&zuik9oq-qtUV^BqaLW$R1 z>V)R)VlCJ93-<43uP!f(8TjiMgKg&~ zu5emR$t>BBNmt6c|H2ryEvwa38-+~}*EQ^PB(V2|l0Q(w5nl|l5rFlFs{?^=_L56= zKUM$mtv&?y*0!d1x18#A_1B8C66twY%0PT&=Bhtdo<|sVlr4tCiY}5moX>~nm0E&+ zP$-cR!KQd$9JI+6(&Ys}-8;GlrqnS-*f&U}Mb?+v9hGp;xY1QAH5ehfgr*e#95j)f z?t~Owd)mt%5(OrtFm%zS0_iBu$d8h9I2`l8zPVyS^|j|_xm(@LZB-#wCtEt`?xt9C zP5G)Rs>}O!-72h(1xbOYw5BMmh-!zdF2p+1X2mPiD=>s_$VD|L%IOYj@0}cB2NPAO zt$?dd8H-RznELwHrL+$(n17}_`UWaVI?C2s^2f~{c#9gf{JWq7U4cu2IR{FZ%rQ5k zgo=OqtN`Ojnb1^U`k&!!tsSSUF{|<%q+*;#A}6wj5h@EMAq$fEUnZI3AQ_qE@68m` z(v~tU(blkEuxfQzzN|vj8%5?1L5;ogVq@|0<$gvgUk^jQq?zCt zLj;#nsXcRwBlOe?GQYoiM%2Y!dj;L!yDbw#{SWpbxl~&6FM(M*Y!*tAcBj>W8;UcT zm9UiZ%V`m%T~sg5{gBl+2F8RN(A@akKGowXsr7wPzkaZnBl!1@PSj2Ri5U20QJieSH@fSD&#tOjp1Be3Wh?{hV8>^8+OHD|EF+jS0_0hMrO zKzf9EOkF8%aOBQyg;pz~vrlJoKE~PBPB-3t>9l5*(c;Yoeju5_9m)0*+Lza=u{LhQ zX-&hIh{rVZS29lwmUq{dSTm0fAtrne(-1v7vcI%a6Aal=!XIAZfZ+!(7k+$RLgu-m zd?;?X@Aw{H0*5<>8v$u6tfuM#fpkI}>Lg?{;#-67YFMm|&8EJ~hw4YwAN)WiB;*%& z66miE=JzS)=3H3-5FpmI5<$21SHry_!7Q;vwM#rKA|=R@2bfRe&DC>u79z#GZ)A9C zB8|8Ez9?oe)ml4&RW(-yhsp<0>|nDgybSt*+vIdsIiL;8uZ=>7+rx-Qg(19W70}b( zkIdK34i>-uHa)NhjhZiyCb)n=+%!YxIkS#Ih8~VpNr72VUjui`7*A&=q%|dfxa12A z%0%<}_yn=s03<-7V&ouZA5WA6peeyEU|Zfc1P(Ktb9 zV`3t(q0(U9H`QN#jMoYLJ;2)zg=8>U-Dyc{1x$U!b^k{Z6jFR! ztF2%gtPz~x1)bLwmAID3GJ~ll*A& zwVH;k>iFzk#tURBJiaeSpNhsf3+K|G`Y>a%9*|JcJr{u=VdY7|f&Ca{&`t$%D)Ku) zTF-WJvq1}F7AAs7I=^0Of0DcbKZrDjC<dUV(On|8;IV+!ji1ZwzbG85@bek8GP;K=^~K)xvaBzbXzj zW^3osvPW;h&t>0u-^a-YZpA?eBjy5o{(Lfbeud#9^osc{=Zk6WWDbPPps)h%GTv=H zCqg^bsEZ}E5ZavhV&&6rB6qax;xJL4-cMgjZ25}LqhLcIGn-Vb&9^DQHTxzerhJXB zzja}TMw%w~)W~L&5L!3FvzAfFPrghSNZ|L`^=^gK^>N&#@a7}{AOEn9_3S7MX0t~^ zI>wQgpTjet&!@FuuLvHyf)m1JeBPkqZRsynv7rVfv5vDKZr6jRNX6(0WgjOD_?Rs3 z2SDx>>}FwGKz*jF4w%XdvfwAe%n;KxYvq&UaQaL-C30H^oOpXwAeo|kg(B>IeglKn z)znH~CJjH&Lu6Zn}9W9m%0>7rL1 z^d?`wn|>P-QGKDSN{zB#{c*|fsn6`LigqVikg@1!if9=s*ZctZyn;Q@6k&ykq3WpL z_n@=z9=!fVd>ByYpLud=v_R(gkcTS^#->JOI!PhE z@1nYh`86_<1djYjB73~Az`bwJv%EuSpLvE_ZJ=diEJ{&c;v}HxYh8HGg~S zG9wNufRZlkc@zuCW_ni1#f<93Rl~!`JZ@Y{(+J3VY^6Rq&xW z4ZVd6>cYHPx>{mXKjrVd{$v{O1lGVQ*ORTy9kqy!1;`}r|J^kHOjQk#ra9bnuBt8G zOE2n8)fa+iV^E!QrR7ez=hw7n9cEDlb`)W6g?UaIJRy>aWdvsoqQZ^;VvYPDO-sd4 z5P$bX>#r64-D!m*d8#GSvYvolH9U~^*<K41dr`mUy8LXd zi+F?2(c!m=W8VGkmgmXt3qXMOatRXNV6bD7ZX=7g;qnzJA0WWWf5LE{HsmkleynYeg+o_R3qx$-PQ1Z+l zr85lNl1!#r^H;Z7VagcbV378x|2fS5wD;tE!>hpx78KqwiPg~|X;4^>IAFc=XdCW# zA7}Oe9OYKB^N73Gy4L>jh<}1ZZhQ%m4F#VAlmwo!Ye$)n51d*Qa!Cms#w4C)dbdj> zS5%0?3>z`z2w5RI%Rt$>6z|Jz?8_Kbz1OsJr@-Hw?8^hFn;)_pTlyr`U3Fbv&ot_f z14J3{asy!@t^?s{!}5`LNijPLX`nB$EA1@S;qgGyU+jt$#ACzZltx6hg%@@eN-P^K z3#Ap~DWaH9`O1^GPR9`!{EdCL(D8a{)x#_5Xivh!LIW}R>T4tG>Bm;Q=u%f@zn zX=W-25!g)brCf;3`Rx(NCG?(ux>j&>8NFL;ZkkR*w$g9x4tiRn47)U21KHO^tCDG? z%#WF@%Q6m5o$Y3n484HEyVPGjw7=9DbQl67`I))|oE8v7688-Z#g(b>G27l*0oN}E zNGB(<{ad_WUzju{pxKJKtGiTwc+@xTf50SW2Dj=($iMZ9`pBQnUq$Dt@$fe|*?d|P z6^+v2V?~Bw{)|=rh?+e2FgVItvz%hsf)!UfN3ZgO5u(_M>6!&rguYj@pC<$fhLs`5 zZHN>~X^ChVzdjFgg!u02*Nd|ivGbH63qw&>H)3e#L6v!lVpO16j{p{2G;ZkAUi84x zajpVm$6pbyfl|a_@M&_6N7hX&C~h9?%>s3tFS;DG9u*jK&v?nT zEHZRozY|CEg;->`Z&AsJWxUzk2CJQrr|NnbxlZ=_=3R=E@Nz1#c^@14RhE3TN}djz z<8GDvKdmKHb=$~}nyLj`(w}+!IV$P=!nOq2|CegK-)yN&Jog!4ee7^}a|bh2QMF)0 zE?%2s5fmRi(!kEg}36D@5kjd2K0dQ}*hE7s+bV z@wFVkfbP?`V}m)?3ezH_Bn@qrxtdPbAn^~3iJDJ4dNJarn5e{h_P=C=lBfR0 zh}$=7Sh4j#e-h;qc%ITdeH;&C}%05r@v1W6{hc{k8_Q&k;XH zgP4o?qj_&b&d~yENYiI8d{Vws(7-xsRC?iI`54If?&T_M(pjO z#v)=>&Gb_7QH=RA9yTZaw&o$#zSjFpGq+!z@5l=o7mA&l#x6Qt;8~|kMYlM>@vcVW z<6Bwg%oPe-X99WNx_5o6LOb&nHk>#_NfkvgF7C}eH@b80NWFT%`mip$J6AlVrqm; z=Um_0DH3~*XF(JIJ-)d~tStFcTSK!w5rx{ydVC}^SVQZKc&W(U>&oeEMj-4RA*cll*>fiWRVT9JPw2$7E@U<~XAbes385L!rkL-tu=Xty;}U29;9F!h?|o z{m;`~df09!Y(-CTSaJ-WxG4%WS&Z2~)swryKoK(cCyU>ci7#13vJOkW?5({Xts$>O zg_t``T=$2c^O;z->_$2w<6%~7Uf?P2zix|CYqpmEo-yFRgl&R{BaLnL8tFN1j;O7b z4mGQ)tSuMg_^)2Ba_1I4T#UrucAuu>X#^Bz=zvqMZdmt+(JhOiCufF51X>1Cr+N#n zf>!}Y*HyuvCS!l@{QmVB(vvvEn_DBsc?v!7MHjG1k~ek`jl2q`eTgf^j5}FXiJ%4h+S-?{ zCIzn(Oq*TNbSqYb{JoAG%^t7+Hzi8XoU>Ej^k z3PkT-uJJ`wLHGTJQ!ls3JIr?v*!_UPi=$R~&d4@g{sujzY3D3=E!jYxy?wn*;3cuj zI$q=+$7O&V?a}m7$;}-F9D#rPBR&pf2xD22GBoUnnvX9yfig0~l1`3=C)#qIce>&{ zudxIKe?@a_FD@W0)7n>=eeg6u^>{DGN4vx{GbLXky5A&@*9g6;Bk2vcFaENZi8|A< zRL+;!)(O~?f}h78>+T6cjz)qd$F9HQb2C12j$Y;X4&l>L-v*#Zs@ zYc9kmmKu1im!Z!4tP86WX>ha(n$dCOw^r4Ig@5UlKD_|iR_H}%On#$G-xiTEPBs4V zTs;#l(ewW13YQ+FhuN%r&ES#k$O|w}v16FCtKC4;qh*9D3tCf0L1-Xl4u}Fa2XKDl z{^UTo{q3%DU#rPx?opxa!XPd_v7+JfP_uUawOYEanS0vK3GV26Pt-f2r$w@?5Ggr< z`?tzS#RMt^By<4+Iv_B@WDh>kU3Jl{#EI6w`E3v>R9{Xxf=nF)nOw;DVCI^To_s8f zfE|8+`J(seX<;-bg6!%Rn`oCcJ8h&#+tle`VyV)$>%4J7vt*JfoAe=a6lh%(c$WJJ|cpmJ|kM|O=9AyE7W6gC|D{{)5g?NqQG z*S%u`q$#*soY3DfEYXZz6oOS?(kg@&_IZiH@K&E4Ki#%}K)$0g^-Bq7DhF1)o%`AD z&r7*A77WMyF!K;xryW}eLYQiOTDm3Te;x_9t7<#NH#JP&Se=88Tk0;~$Czc2KqSGN zm7CuT8^TY_9eqAxRu_ZnBf?>t+qTFQ% z^ziGyR=V#ZzUOlhh_aw%qBz-t6^$)_TdZyXiIQ#*5$+F047Bq|g z_Fk+|8@x`;T?4YT72{~aIGj(NJQNSB%G;<4E9QEZ?mDgvd#-9VBk8!i!PuqP*vaQO z$!Kj=OrW+en(AzFLlV@i#nouE>|rfX94Nu?_iOEa;)c}hpZ?J+_WIzoFv5U0pQ2vN zNoUB+7_uSL>kFA-;Gc(@{yNv-^TG&da)>0qDng@oHCmkvFywnlI&l()X2Q*29cXCt zlI>?EAlmd1yK_lmX2+>e1rwZcv8no4RQ?dZ|Gt@_PW^ADkkW$}j8hsj6kUyf%fid$ z;N$K-3VGI)81!S9$_o3Ig$(Kj;xl~ppzs>29ZkA#DeZGAbXG%pgpsC32UR7IoZ#K zzV&4e*Fmn~D)j3$?AZh^b2?%*>r*i+3bSVZy_t`BH;mIF+eM4DNc-Aw+^k|}3=Qes z0CO1cs~d($@-N~4D5wp9P!uIjni(!H)P^oxGm>i(yf}x)a%Wx`YI$F|8w#L-C0o@F zUJhRFQcKlw38i_}!77??_TBSn2icNIqv5}3+yGa%Y@1gx zH_q{`Z*i*4C)0e$6QCahNt{^|n~Kh=aW6&Qle)z*y@Q8}kinswdVMKEx3+CW#`|hX z?bxL6Q6szx(RNSdHIfGc>iJ7jql9WK53C(BoWC%?*EaLDJa>sGyktoT+1yLpfu9m) z9oA{n5x7V7cqG*qVoQ8{#*=&J?M{iVF`VuI!E@ih!@oEh0J%c>Lv|(}{Q@I*&Uffrt%ij`=Zo8~XZOxY$tpGrD-GX;GHu9Z%!nKN%93CLxsm>+4_xsZL?lr) z3xm~FxfZWtM=dOBZo~PN+`I!`d9`g(k}> z&()J-i<1AEc;Hx|45=w@Lup$zz5n8Q2+oy?kJ}2XasUJ<}tK4hP(hQOw3&LOl4N|N9D98RC7A3UB{bFyt|@v9|&_0zkI5mga^a zM{^r%Mq48jBEO0@NFkEvLgtp1P7aRtAV(W}9wkSRy`z(@yuFQyxur2c-rn5WQN}XB z;@`56syq_ripCDYHcr+?=GLadHm(3R4klM-E+$&Y-|Sc6L`Ye(f6GD$895mm+XIA* zYz&M^G9tQ`;R&{_R2OkmjCqP(|-<_ zL_~>onzt|pQi0~*3NpryAjnLE{`+RawtWcyYu9A{F3W#k#~IPD2U3UgpE@!!Jfg<- z=FY}OAMI_d08UQkMm)mI>>QjdtYX|k%xq#@%*P2oFEQ*6GLtT zdNvj|E_x6vGdn#8HzFU8e3E`R;Y!2cAN0lCV9OpTSy-61?_ z10n3dfRkqcDTMl83duefaR4x{J4&WQegb?FlNBu%*7pbe|Ih!V$Hd&!$==w3QP$YN zz{t&>g~8U^^#61lOiWB1Y-}XYZ%z*Ozi%cMwtwCn%xs)YEF{dVoa`LzY%J^?tRzg# z?Ch*e01~GE(;fMLdmyq0WDn_znT4^ngR%Ahwj89GiOGLMhw1M_@*nU2k33=Wvf@A_ zJS50R;3o-DMF0RE1^|G*M}UL;M`9fb697PYYc3)p|4BrIUH;+a z4A`YVPAL$7>@}Y@oq0KI4<|Vc-ES`>CBFr1SZP2B!TkgXQV59St{!zoM#$;MJ_1+- zk@Epk`hy0h$b20gB7pd^$0HU;Xp*(ZpLJYQ0zFT^sO)QJEdkNCp>bTZ`_iFNP`Fx#+>bRzCZXM1BSD|P)1ZW+ZKLd|CVZe&k0f=;PD;nbgDXSt+;$@fYr z^o>CAPj#zY2(Ok};hFpps^2rgXTQ=l^$`)cCnr3CrXzu&_C+2LXE!7&i+%IK^P|7# zF{P+>H2>i?`P)r@?ffT};)D-_LghwmYHKmC(}YvPeyv;AA`?NS6P0SLf2ke+My%v* zshY_+8iFf$PWC+rD`ja@3L z@-3Ng!h2YrZRkLY-X}$F;P%e*k{_Z1-~HRqbSyLSxIXH}IEj0-6T2rDW>7 z&Ug*J?1n(!u+(OD1 z&h;EGI|pz*VghbQI8&GKYn!H^Y%?S1(YGX>YOjR=m?Wq!!FcyiBhcMUlwf^SZD06f zSRoP=Q9oBEWL17FBjG8BHmFieOh5EeY)cre*4L%@%6*cYgJB`}P^2NW-)C5HQ=&p-8pHGLu##(02EiN)}>+{f};GrUEVu ziS6FW_9WGEYp1_*WgF_QI2OC6@kG`QnqIfMR=Fqr;Oj#^f`A9Z>d#6VjV3NZIzW#3 zrXRrru?N9)b!mqjf)B9&KSfP@Sont;`K#tNFtGy zNpNSs8#5VuHAX##rO3Dgt&vzLQ%!LfgVF10|EeysG2JuPGw42&k1mVEkYt?>gQ6oQ ziL&CONtATlfN_#;5-9v#SLm8#SZ20DqGCUFReVLfM|@3rd3k-gZ27p6Te+0cZ)4QH zw69)NcQHRj*>l{LOy$KUCMUcmuy?I?weUWhNSys3AF((4WjziHaH{rhVlaZC%{atTS6e*Q{PtD<_ge1@&|YHs_FcRHhz z%+kzCb_r|VUC80X2@;aQlFgC?Il?)MlBV%QbD@{7P3jO3&vTHGbz|aMIa< zab)%bFBk6(-Wj_y*wGBd!ovI(jKszXP8;2yHZXHD%Q3ea;VGERlbLoN-!iuu`H=JA zU(kMTWWw}2$!g#gAuts{7&F9>Dw_(@AzL(Cgj(cTymQCo7jH6Z0=gHuL*LHbDqZ0G znf|lP9L;RYCc!>qG-7?cG}6m7Wm^XdSal#%$r8-+%)-m^=w>r?S*Kfg3mnH{Pj$zS^aFZSL#D(bC^8z)8_1cgC5 z2Sx;y?oI`i7LgDHB$bqsE)f+51_c3WLp{Pbn@a?(JOisN4CmN*% zPjg6Ldc85IJ@9I-0$JP~`Ymd{q$j)CusLDw?zf{B`Oe2P`2#br)1O<$qHsHP?nd2} z%Qd~bT9r^mJ^{IO)RnCSHkCPp;Hws*({vtq;Kr7&a>Gez@NIC2Z;vuD^E|FUDDbsYj zy=&j|tfxL*Tg8*sgd!jMM)D=2sJNr=tVX0xZ*0cL?oZus=KHDlVrY2#cvnrm?&6vz;FNciQYmXpG86pr*xd) zff5Tduh2u^Y0KNxSuX`+$YY8G@B1yck2m_Xu20%bZap+Ce>ZB`n$!BOwNuMLt3Q3~ zT}67-JM?Cy`yIF39=tEEZ>!WvlqaI4UP9tDqpx?Ue3t0YnNHsv`D)eP8rQ?gbDo^g zqaeOe$l@&c@e|L&7fY{)LPguwGJ2T#C@~j2Viu$Byf`qi`#cJB4|YGS40BK0yxrj$ zwHkel^zxGO^jXA9IHgP_VK86rFu-D_`Q=T2lSc><)ehIX(aqw>0s-%DtEZw)y*srU z)x>3extGJ>=I$=rangg}+2Fuy!qFc&Ci3x&>h5%xf3cYR`pPIow24ZYL%HpSRPMdp zirvD#&N*qs2erFn<)3`o(Z!9$hsDDE<+)jxwK>|ky(@hd_V>?zKHEe$ZW~{g{^^L0J9P?ag+_rG;*rB`fLewI~T?&^wf+abl`#yA*Bd?YjN30)Nt*{TR-)FN& zWg8xST)1-f3exrbBh-pd1)GKX0e4upfZg0^Ub=}#`^UcAFFZSKH z@y}+Sl?+`E9cz9ZRw`63z;Mt$UFzK~)=t58OYMW`bx}Px$Jtv)nRS%1%E#X#dk}Xl zOUEV8Kl-XaQdH7WSZCRpv?=~5ZLAQ}>prmAP8H8hUo-EPKlKvxZvWiJFw>e5jL(Oj z>cXRfhPh8%={adDm&Vv~%Zx<5`%nwz*8T%rh9p(v-mv{o*>U`K=+Ws>4X(oFQX}+A^gzOuY6+;H3v*5oTgJc8^RP{AXcO-q(h(U%{z;d zw53aZ#z#>D1KaoZL&l>6q8l*Rd@moBZ@*g4O7(R=?BgXB?mE!l^_XpcU;4m)r>f9< zK#W>y*O=YOYS(-HbTNrG6N7Y#Z!l)d#%;y&3u=S=qET!$fp3-YOHn=+V?w=`sW=35 zvGK&Uao#vx(^EsZ<1_4zUk;+Ll736j45;<3{~{iHbqRM6g)6f~h9fsgV=it(K&oZJ zE#;%=j???eqB8V`Ye8tS042rvJus7k!uSI`nFn~2Ar59{ZX+!%(URdywmA9)Z+E2f zNvCLh(+j8#+f4nL=8VW`Y+h9+@M9xJLI)*1Cmb9~R_qV%^_v&gaBy&w?%&dN)>Tyz zGqtznGQp;MxZG?V{s=rI+{A!CZOxrcpl-G{c1~h$k}Q9;5Ci_kzRk@7{iBJqwIqwK zss>cv-q9Q?$R)tV!yoK4&~?VMnL4)ULIu9`cUI^K7{Cgh;l*CzMuU7RIZSpI7B z=f6KEGIzWG_e^$9e=Q4GAUF0CZeA`P?*ERB9Ul9xn8tlKa~s{O_ice76PQDaS6Ea? zg8N@T`q!>{C%W?R@to|c_m8d;+&?VoPkwR##W1Ne65Rh~zSJ2A>y>q2Kj`mYRlf!N z#``la#KmUq*#G_@JfqrViaE#und&zSAYwkik9m zZx7nr$*I!#I}3eTCaOy@cxTwVRPd^ZL2=(4M|A$R1G$v~E`FUw=%C)q z{?vbK3Osr<{rrs+?o!9S!@f|emB~}+N>upwRg%wh0YtC7+^NJ);|-<;_cy!bQT&XK zXhLeRD@`^9mJCTbU}Qyui*%Vkp~Kk#XfWf-i2Yn30H6tGaP8n<#|0iXvb8b}L?S+K zv5ozK#^(*pK4$!T<91Gf0iy#G<5)5X0><%wM(%mA^9AyyzAT;dErZv91t+fqe5O05 z9g+WcW@yOxGjbTmj6CK}R9$uT5A*W1a!$pNKnMcHe69@R{Hzo1sbR`_rQ~ia6NM2m zpEtEpr^GfeDW4P$&e4GF_N3Oee%`S{I~nG$kotsX%mY`#9YHzYvh(zu@bIG+fVC5! zgzo-i*EEv5Pl=IP3Fyak&kqb9xg^LaR!bM9byv$11Qf9s{8OE$8P9+`@D#4eVYd zTm8}R@%W^x@=v?*g8t$Qwa%Bj&QJf~OW_}U8GE?nuJ(&9y!01eTyK`g;c=hG?!*n+ zWpOj;Zo!>~vp-$lnO=m9B3`Hh5&r_}!Qjipi`XT4+x$VQ`bQ_*+@EnqMvKt-LYGw$ zPiiJtS;F$IKg3NY)%z(iNPN4bfbZv|{_$8)2DGa>=V`X)uUY{d(E?;%{8dleuZJVg z;06rAeg$L6&n(4%G(?R8utm4juKaxO{+zcOL=0M7^cbZ5%|gVqfrV&^W~=;W-hDD2 z(hs?dCf2`M$owr}A!>6)|MxP`%GqBU^G%&$VfxD5_26?1C|N*UYSCwUwV}ITl(}6p zq^mr(7rGuYzWs?v0u0sZ6Uf z%RqOVrH>xQ9kkU(=~QpL8u5xhjr}Bn88E#x!Tl*eo4Jb|4-~+Vu-tJ;K5m*aKjKX9 zJHVJ5-le2__fum3R0kB`d|bm)i+|Au`_PC6sCRMKvx)rWZ~&<=fK<`2{x`pw*AQ^T z{d4B|nm=u|eR=HbW8nad%J=FaiQa}yZsxb zgaFghOR_4)|To0hARr#X#=yYn107*d0>;z zQVajo@BYOVFfeVX1I6;=->}&hP@)pt8H?X=g%&_td61kp-oK_!Bl#>K1wzK7AAZ9X z7%;6+y>=1FZ`fRfH6H}uBATF?gVT0 zHLu(g6C8lVHtfHJ9aT@faOw^TOU(7g3$2OvH=Z$?iQVerc2JFxe14K90JR@;4dI0` zT{_7`ne1Z*!^O?w?&Y)*8|a{ObfvY}5>=VCLALgH+~J<|jV_vKlRO~}O8)xC;uZZU zq>s?SuuR_2s-N0IjnSl9K+$(FVm{CB*zk*q7|E@_W#-;rU9G2;KVBa=3lg4MQ&l{5 zi1+AL$IN%TR|q;>ecAD{i1BZZS&(IH*Q8o|S6fs>-C4-;pNw)o{4j>+{MITy8yr?Od{c{yT&a3Q&EM&a(4wm(-<; zSA6uyIlFT}S0`(xikEb$EJ5)$=nzkb$LqVeYK#S7mNU^uHtoNeF!i6l4czMKf)g@% zJ?wEGS!!zAy)nyGC;b-`=-b<%op%DoN~$LJABW{^?|BSw&%a00f?Z$PMD#6Em9=t} zjcK3Y*11yvx5O&hg|7eo-pZ*5nALgxq}&y5nNV6Qs4i??K|0iEQlX|Et$sof_o>!u zloyU^hvZJYcm6L{@1I{yq@)46B0L)s`nVk$3~!2`z%w$4^FUvZQP#q`zlnB%YApIH z4(xlm$QV$m0&p+{*|kYGMOB6}lJ`8RDC*dWJ@zSzFef<3{8#5O6MY0qfb>Ii)|XJ* zjgZ06-9-IgY1BL(*3HT+w`$XZ(X&~WBfqj<{?SC#;Of!0nWcb5+@34l{FS-EMz3ex zG)vQ!2?9n^QTe}Aehh#e`X>ax^<$ZKrG`!CSC&W25HRpP*F9aI5Cljb><@YT(w-Fq z_Q(3gMPGa9;H}LnmNMlEZw5ZnQ7LJ}FHPlwDnP|}j*@U`vrJwbw_p0N zY1RW;N6315xBxU`&J6qfie_UiU{h2jV~-O2je92W@P5U)#{lPq22YKRuTX>Sa+8~X z>1?-IjiI_m1F2m4Jufu!baL;ybi3016z=}<$jbzalf?;@fUtlL;{9H#k=+YkZ* zgwfuHN569$k-)ae8>)%_ED`kAqc#l?vpA1kwcognARxf9$w1z>{0(Lqyj3gXphYrq z>6g^8cWGdy+eLI4{u}Qc2LN!pg(>mhaE0~;ARZ}P=?TAy*Zz_Wuk_zIQ+cdZ^v)IW z|Ax&rmyMxinNL^$ZMF*C$#>Cr7rJPGz@RT>MndVo-pCc33#`Fks`=-j8m7Q!w2wW- z{?pb2c%4>r3iK}%Jh%c#;~o3XwEy&`278eF8v60TY}NqwuUfHpZ2=2c^(6xT#ssur z!xk@><+xv~Vh`Xmg)p4Z-)Jlh+wA`s#&wC})f{71kx55e=>-kv#3Q5cdR&U4GV;nl zoFegM7hkP(a&NkFR}(M|6bFav2w#p|uemFxcUBH+jt^?qkF;|=w3Q!=B%0pcj2$Sl znV~0NDE!WB_XC<)YB$UHPqv=(pTMj$`>Ub5h?BPtC z>}iHu3Ssm#sb@sPWL{`K7UknnB)M60?c_iuT!5i=f$jX$v$=vv*UTpst(w8r#29Lt z7E3t!g_arqG<{`fN=dYi23iOF6+ z=;g^{plUQcG45`mpMP>`z&`}WIar1u06q(Agq6*>ge}cn00mT9&Q4VAjNhT;GSEGm zGdg-Fu`}wlepoAiPV{ZO*GB8yS$j8mIN59rF^!XIj1}}5{KUfi(SV~W>1?g~r>DDu zM{_vmbnGYqeG2)iIcSO5-!`9Xdo{Q6ag~dRl7n6PXdOWW0~>j*H#x4=osK@QE|E|F zX!KzmC_qnmTSShB9V35|O@Mxt#I-T}1K%23EJW=AD2t`d=7W{ z%-bRz+b*R?O74vEceKpcN(1`GsrB@0WJ<67Nvz@tVQylc02(u(3iV%O)l_} z1Y7uk$|*6^PjE2EvLl{`8EP^r)#19P8-&rB&Wz%u@!0-Vxp3;F~SKpr3? zrx-N==z4y;^mn`V3ys4`VMyKU&owf-onxS9(H^T!$*E7CONjGm+#Tgui+lQN_`t_U ztJG7?TS+QR6=tn3H(J$w3YIDsbIj+vQs$NlJGstZRAI*_Abv&~?r{keuqT$JI@s~G z{G_E<1^nn>QS|tar%D19GGT8`gWEZ5U>eB#!dWBk39vOs&zzz03JogO0YK0$!OvYF zHORkr4HN(;Gx+FgD>@r4&ob(XSu|qm@#8aRKggGso&NGp8F`J$kUYg?qS_lNii+2| zs1{ERBiI}rO0rQbHY#r~J~?frL9NW%+(o+|@nf+Balnw(G1h^AQ^2z!c5`0uqmg2YYQkrLK-V=}egTr+O6mu&n3Z?e{vb0FzSUHA6S??3+9lhpvXynmK zL-)IA^I^?I@fKdn?=~q3a6AX`qd%T?`B$7D)Q7gTh~tUl2`UUh-aw}FNe>}g>#=gl z(Xe&LcGtUwX7h^BX6r;ErOB2InG#6<`l!qT%apKF~~Ip#2@4r7^-ra+$Q8pB8)^y6H=s7J1i=! z{1{Kl7@jE-CLPT)kq>+r=t~}EdX))wL|v!o)?Dv&7{z&q(z;$2ZE}mzT24+*GC*3c zGDZIBCu=-eg{8TAOYi-y=lhNxT@q|A4r&3F#8>osq|)fkohJ<+jZt*6NNjbIq}+8w!tboTOk4d&m(TS9a}3 zHDF{8H_MIILF;QFTc=kI+P^F`5o7!0OHW1<!>^m%ISJo}vGpNb7a!gH(as%# zz>L)zAWKkOi<)L;j&{9bLydp^r};@aS5B^*)Qj$W(MiMH491ketm7(2B+a~ z%Pt+KWoGT9sGNPN%tCMW;ARKXrzm~6)F#gQyF@~zpa0}n!_oIY%G?+cX^Ht{RA_cJ zTM_K-ok9y$klU5pAJ_wJ;N)f=Y9*bqf>+Kl@LQ<-fzx=K>CgMqnJ}c{5stfc7*p1Z zcgXTYr{eKXT3;bGJ3fSKa3hd1Jc{JiYxP8P&)E{$`7Jv*00gFP(l~jF%0R<)%w+)U z^oro5vt=J(BXi%S1BbExd*qY)k4`kh8aJTi0gjSs*6Qms&J96l=f0&UaRu6fY(N1L z{0@5?bNLp%&`x5BrM!l=anDr-Yc%Uy82AF|hcHr=aDZW+^`cS4m<_!E8>En(U7&H; z#*6_unl;1gaJ?BBDm+fTxY;3|yHqk_H@BVZ>#c0Kq&()%IR4$qU9v48fF-tvSO4K8 zC|mNfhg%*%MF8`_XFp!K5jSyoF>TVI%)tcNQX%`8py^h;Wum)<3W>+;3&j$ZRy}V8 zrfu^sMQ}{&CLbKxg%Dbct@q)ek4|B7`HEVSuYwz2AT~2XkK9jBV5z;(GCSnIGxo?x zbiMIB5^M7S|9ifjF>vrP>lOSz#?up*$A!>}UCweLBk}M6T8C*AUgf4v;Z0I_~Dpz`hIcOG$2RTGs zR1LOmcuHcQ`Ej^ff{9{HY@|ubmI95ODIfto&sJ8Wh&>=9{UTgqD%(tVw+AFqSPEm{yHiJKbH-^vj4kmeTZW8y3=%{^c zjqS{W&4N4JgeF}E40YUYPXcOj13)P=qJA*NFpetE9j*vAb*?~vc-bQ0hg2xqoKuE! zmee18_;x`Edn^HONkkT3q}b2fyyL)BK@ zZ}^z@H-86^~HZkICN>_1?osJAxQIRL1Z@}8QN6nci8Xk@3h9_Ig?7jZ(vqK zJg0lUH{YS!%ZI5G&rD#_6d`##5s0S11b>9Tp$y_e3Iu%Z7km6RKFHw7$W=>u?_XsI z4A3jc9_p|mS0l6{C_-fV8P*g;W@@D?IJ8M|d(&`p!t*1ypR;@qFozfmGf#+<+|NFN3`J0t$PYe1q?+J8YMf zYn#?p#M=%STnCx!rRzVABD~#l7-6_n+;e-Lc;c+eusjz0Th&Vb>&B_QJaPby1nTw}5Iz^9_uA6KhAO@h04O+Q6mkGS}1yvnoLbvUYX1}PIR zq&*VuPiJd!(62ydkyLRnA!_Jif6>ztZod*KSliViBiAHxdfD{2AJrnf!h+%69_{p<>kj~^x z5V+Uz)=0nD-Lu&uLCIljVK`y62&lu#XgM$5T7ENQh+rLbim}HHy*yG%OvCq;e z4+Fa_O;6wkjaJZ2k^>9&)~(dOS_Imh&XH+c#|R}=KGDD5JCtO32hf_+dTy>P-VK2o zTVIs9Ik5mjf^HpP<=F^8;{`T7itpIZpdaFav`EIdyhIC%sDb_r3YB6RZyZYF}jId1(-ut8nw?19_tyB{YoTnd$tjfbBX`m&If z?we;!Cb@mqLkjH1DhiD20;rdg6@A^`piJdF`$h zG&6cOZsInFF^rH6LqrN@%fW5Kmx~^E683^I88t4g#^EkO@TG{{EX)IM8=cC|NrWIO zZESZfQ6>VqHf)_63U$Fol5aw|E(0FW{Q~6getfhFJ{DnyVMKyZczhIAi%^79 zfmwJnFX(9`G09AO(y_78nFAEconz4;CdbK1b~-8R+ZF}pZI@KN45fld0RxGFef8c~ z#|iJ9SAZ$`XyDk&ip>NP6{^|8DNaPW|2eN+e)+wtia15Bs@Ph4Sw3%}WH4ppv~2(W z_9BriWsez|g?Vz5W$*RD$keYH5iczJDl)?`gKW}Bv0b=Ba0PX_f^ji6P4G-w0V8?1 z)__e!&Y?%_5ik4#{Y2;HChH!js;XehN9;aQ$jKx=QS1`5_#CtQ!G5ATo@F=!l~PrN z5SkTp9S)MtfXSEiduQL4uPLY+^4-@(%F)e>PhEzwtxt)q#V*Pvdo-LCj?BubcpvYx z|81SPu4Xr+=9zV!AZqu+W5JP)SS>aTi*8e4zjxNVRMd&s!Pp6Ve5m#ChaA34DIw_@ zO>o}(#RY3TrW-B^T5e?Myo7i9r&sySkPh%2kQ)3wjDtnsUd;pJ$O8+F=H4XGEx5kv zg~_&8SCEH$n|Bll93{3%l^_}`0K(7|?)*3czzP`g5)e#-g5`sVpL>I;3Z;uKjkJ~! zvlepan~EQ|RPfz;YfIj*XVl**gtk;YcJdTgFE;!*=dL#c=Z3MAb5*hhvXUNlBYeoU zo~k8?6%*5o=X_C`_1_ z+m7Z$o!U`-^=Mnp!CMGq9^BUbZbyUl{-EJ0@!>#0TQlZLq|if2LL>X~xhQJZmg4D6 ztj|R=YqZbkYVn!P2@ZrU4-M!CCB#ph$bZN@0)VXR4#nEx4^e6ym}lEENFJn3)3Pom z&vC9L`N}}xjitC0br++O^I_&zjn;B3XX-Q)VfaO?%2AE{OW@-~-=n=4+QRgBZLPQU6S4xPSPWbyLgdgVh@8fWgy@Zpk0+R#l9;&4fNA_UCHOzm;n( zbWs{vJtro868FlaVo!tiDN2057fX`!gMfJ}is`zDUrc`~3YS)#bmP`QpG^2Ny?S2c zL7wlncaZ5)vz2UzGWB(lYtR;fSMVFKT^Rh{E$E2bZRoN|FH>i^Y2-xX^YcWVxDsDu zOj%lB$!+`mv#_N%HeB#~69yYj9WQc6#JDcoG&5yqZ-f)K75H>Se=S$%>^BhaP>qSR zZ3~YyjvNTgcS3b=l1Of8lppOcBBz@|V5DYmjl)*nX48QA&J^l@@w3$)SdgkdDT$ic zI||!v!`$Edou_ARJ*KP3_;jtSa$cJGH3`*2 zfPn}=pk(m%QE9w)xcCG`AhW$uF$Mgu z83Jjztj4pF+x+$6^?Ub8>hyygywqBc8skEJ^QvgO=Fp=EG--#eUD9Z2wx9`9bD@S& ze}oAQ%7RHMbuH1mf}i+vwXxn=(0v&0m4W$GCXCe|iAFZ*y;iX0&WL>t*ruFOTX?_W z%_w5=`t{PKJN*<)-(hrB7Gvn( z1*&1Y?~}Ze*OF;xNyl$Rhi&=Ggj(X}7?c-$PLi(qoG2-tC@pN={{b9AZTJbW^xl9v z`!a(*zwnO}Hwuaf)C#=KG4x6HF+tT*IK96jY4m|PdyFiD%er1z2XdgVjgL6XZe-Yc zA(OHr&bDRWQD1vFT6sH0zw#?>H)DU1zn{QJw0A~z&F*xFXHd~_GJsL4* zD+BRiWHt8qMnIBN5B02Df2-B$WL{bwaA4ccf;J&PcrX=0LU{tKKr@;%`I3)e^#XV~ z#G~K{x6at`uw?7Kjkzb5d+YHsiPxL?=Xc>BZ`6NkX&(a9+)mm z%Nq^hT*5-WQI_yGoW0kj!AzTl3lH5TtV=Adt{uCpDh+hk?-BMol0#j-d`fT~&V1K) zNGvav8}x)zyl?VUkOBXP*9(Hm<0S6flMj1TgwB+@#!N+@TB#xfoPlh^@}qXfs2 zXb`3+PU_wb>63X@fe1j+_zi6jzRwhH;rPGiw^XQ&lBLEymqKCE5lnIAUN0l;xIf2m zq=0lBf=70kN6MA&OJJjddxKO2fKCOSQU{F!5h{=yA_|ynWjiGF`3}gsGX3tE5i8S@ zh)ysmVG5rva`Jg6U)ft*8 z1qNL6VsC1Tk&!#TmMc$*1|nS{gBB*Y*p~0z@!-B&DWc$ojj4GHHUVfGwkYOvxCEbVEBY6%is^Va&Kt;eq2 z=b|3?I@X|bzFDhv$J$Qorg_Xsh<+h+%8|AeN5wl$)AIAJWu%5()=F0kT~uMRB57Wa zHFXQc$4ph{os?)5K+xNriuD%D{R~$BR3lud>ta90{YIB5iK0-5M;|oQT5eE5 zU_wR+Nl)05uxH3)kT z^}!GKP?Ll3K-b9wsO6{?{PNc-#9qRAo_Bl0onei8wWp=J_-@*4&38Y!56yg>94zzH z0iMZCh@n2!Y%P~boHdBRdB*cBlUWb7=cbx+Hqc<{4nrs6a&wB!?aIKsB14T++0!+V0;yoO;R>w&r|vxM%Vh+42Hr zTJxAhN~FFAKp%>Eu_@MkU5vxmJgvCvm~E$Hr&KhQj0O=qTkmeqz0b-y6LOl#iziKX z%zI}Hc|SejHo;k1^5x`g(5tOz&twnv{`aAYiYFAqgOrE#!_zEdy6>}b1mrD~&%&R{ z8AqDv5IEbg*|3n$?)kwf4?or%d;}8^o_>ACak}w2Zxve#6Z0)WQZr{dqHSKW(0b7V zGaTKsnczZpkV7Q#__S*1QEiaJWqg6*=7Z6@X9vLYG^|SCT>xb@vdgA{&7Eh@t4lzL zQm@cJlt_00&5*#W)}&LsMn*2RHyiPg9+zc`xt~6S3YV!Go_2xyf|>aZZ8<5GA+tzk9>$z4{0-af-cz>;}!0NeRL)R7}r-HThk8WM$fV8i6kW^~Lb&zV)P2$<^re}29*?gIJAnvpIWtRZl1TK@i8r3MTakMoSbF@Zv$QlN|I%@7v?V;TX- zt!(|`_lHZC(pQwDxj5COG0r9n8fgt)tDX~durCU5Y6z@GBxo%p@ABis2kcFY1_dKD zlm#8=IZMN;j%tfI$07#n_<5oyn88mzHpf#WzPi=NnAK)0nE=VO?F(zLGGwNfdwa>l znNLmYTil9G<_=2XE%P#o!Yx}V@U{y|GfsE4L{Naypi1Qf9M2NCn>zD3_qO@7>zJ}0 zsgitB#)9nIPdwpA5=Xl;bI~=2<`T;5YYij_o=w0)g_d+(+*cg@oKaV){-95x^F*1i zj3oA^_V*S2*thY|kXmUvIx{ls5CqUXgIqLc)?hJXsWq7$z^0_?3#ZalzWLS)J$tVBHX#Qt^psceW!qB$pgX!85}gV>KRx6Qe|zD=tQ$6|ZXB4QQXDMrVXVM_kVNYS`jw74D#Jw@V@;|3 zl`wG1%C&kTn1k6{=(1_6)LSOox7h}(5<1xuqMg{%G}D2Ym;d=LfI41WKCNfx=tX!G z1bk*;OwTP(li~4EaAGswvKme^JdqrbaULY|R=;}3a9+h#ua%7o0sgg`sa znl^R7XESc15nP_0=VoUc# zsCTAztMzqhH>qg=nYC%_7*hB!9Sk;4Gm0)`q_S9jO+<#rPJTHV&huj{mwWn3a=+jF zv!}q;?0tDTVq%C+zRyUAjg6@QRWs=@hkg+|-j&28MWv>qsK7g*TPoI~_rJW25_2pv z+4kyHG@(;Yd^#{x{lIw7*Ygo3iRRPXGpXU)45j*NBg}$M_bN6=y?Ho(gf3N_$lo2` zsn((7&?W83*!WOmijWjiXylDW7?X(0NiH;?RqZ{ru*EuOU|>-yt(T5=3Vn0VZvvmG z$hRqde#+!iHtI59sCn9d8tbT=TN)b9iF2TvMb54(>jVhNmB&{XFP=JHC8iPUp-hXK zW853?#S8$f0UYdCu)MnHTx40a)g6KWQ9{GzdbxubcY9_U`6wu*`S!E$LdHYtonp=sV`Z5nd20q-TEG6gTu+-!=IKCTl& zBT3-{V|tEfloR)&m)W{v#UI(JkhEcK9o&+QPxZ2vJK3@TY$sZCw$U>(pLC?5cZyi!>H;7jlO0O(?`F+Q`Qp6ew{V79K1E*9-tz zpQd-f>8t$Ua4rDa--mNak*(h`7->U4Md3Z|{>a@YacXBX=~yC7F+te=t|(x*X7qfL z_C8zc+Z=65AM+<-cI=PQ_L>wX^mR;A&=e}z#6^`DQbyBH@aLM5ppYebUBx;Gi5zdq zdSbVq@R+lfMB$4}`PTuF-=w9F_d1&eQn7h`Q)%@*(yV+yO!zsJ#^C|mM^!Y>llsTSv{ zuS6%!v2&qe0xD_aLa&(+wMm04BW4^^BAOCBztT6?^tbz34|6 zeH$d8C{Bp9NOG+fpI2JJrY|~oVIs4VcU2nNp4Bbp)<`KZjRU}kx9=kJ z3~RIjnaf#5i85BD#*mpZwW0*$7lO*!L<9V*p$40yt=d_)Umfovj=|9Y6#rPc9%>Iq zUHYB^ZgXV!7TqNCcr+&h;3qvL<%iC!!!foc%~1K0LtS{`)qcLzK*n+ z*O80JrE+&V1j2H=q4lux%0ST*scx?Q)?4WyYit=9{Ih8!V+y1H_-X0gx|jv03GbcI zizcP6%fnzL_mT;0(W1KVeh>I8;i=<=+@tqAmmNmS9 zp`#3Za$m~n>u2g0z+UenrsPK4(9q7w8Jf@{a{wvY4<)UIn(~jb5`V6+GQBfUcEzCQ zd;>A7-g$C9%;g&9fOx&Ef*zM{9(2Kc zmN}I(GXna;ArcaB|3>l+Lr_3(J9?*jr)vGM4x43XM>5L?1)fei5B~Br{&Ep*&U!_^ zbXgZ=4jYQ~?OQ_|*!)|2Yzc|!g0{iYe(tdtJy`V8)@iN%cw3ZP398WX{(kV{k11ZT=`O;xzJ@qua?&O2h^5hloW5?L{t1yOW^C0*7|q-jE8*@q z1UTU~LbmuWJ~1AASvly!x18>~TaVE_j0)?#!|Me|t`Ot!0SWPNjR@2@hWc>64^K!$-s-6>vuhuyWQhfg)BQkYF>IxQOm z#lSZ+5i)$ar(~fn>xritIBYCas*0W8CFzyfgG25+36fVW=6r0!7U*5Q)bpM!2zyxH zeH-sqSnBd9)sXNg*p~2RImx%^zY{mfPN0p6k_4@he(R|Tx2RA6J3z$0{&btbTwC(2 ziXW>2?~p=!Q*PnQ9`(F>C-xrE6~trN6ifkmfQ1m)VljVcO27E{dYId7=(c(CHSpE0 zh+|O$BonxRsocUDwv0JMNi(I;;9rZ9Th}u9kh-bGh$zZnH~xZ43#Ji*Wi#&ze1fr$bBnS(BZoU;cSmWwjs zVuQD_^%C8b9j|xr2cliiQ(}|5JdRt~s+IDcaZlv8^rX8nN0BHMGMsx&4}0a4GT9bc zNX^WkF7MlEi22ae>2(RoxxCa#n=~h*#CB_i6~vGt}Bci4^Uy3Yf~FrGQCh}KwLN@Gd3C=eIUz;|Eg`#;~SnK)$kbgcEOqd5IyD!bmh)bP1 z4Xyx}Q`21UYAT9>6PvY1ru5+;f<+V24r*tbHxl1HS{>))GC+(*m&{zVy6w^sg&Dky zWHQKWEt&O}ksef-a)R4cu77d_0N_>Wn1JmDrXvSXs><`&PN%nH^okQb{)>%5pY=+G z;1Rojr_*bN2xn_{O1~v950*zAOO1tDxbCG81Q@C<5&m!pE_OLw_Phy6E}{7@$v*u% zg({jY$V*88V?mBcCmlWyStc>a4rwYf^zpsq@FkrTxjVo&%dC-V3s*=b*_rV5Rl&3t zV)}Co=~5X8*WJdM>d2oksZeKslH+Mv4}9Fq>$Wz@=dm>()y4f-7O>ttO2$gw@ra6f zVdYpdpQ&0ak&;3Z6V{Qn4%0zI=VsvpXp2;2QK+!s?qjJgh2dMaOz9Tl7Vu;q3}l{3!@XHEp$I7lV*h)zn)dw-v`&8~+iTIl@)=5VqtHx-i@vEPHRi=* zJu-PER7U zVwKl?@94eO>NvIDk{v(IY;9zH%?yW>{6Fp!CU=1Nx-yTRikqC3rUw?c`=KH))HuXp zd7$RVrTk#!BN7`O`r+c=d(eVa5=b&{EEixY3l~H)QnQo+1$IE72*y1Yvmpi$Gcu{} zPRk^3q5a+(-(#>WVfs8%QXwfAdmc{j+$GEVB^1M>TjS%coP5oySxAm+KQ>HzM!;CQ zUS961jmX4TIq9(MpvQfS;Vv%4rv2)^?0rJUiU;%j7`;+ELnMmnc()k|02$aV^9Ow} zgals{uj!{osV&Y1>4kU4pswC#QJ(zoy=TcE-c$2W?+F8yUPF5SP_I~FknGaZK*(R) z)>#0d0Rgdh$R)woV$Q=w_aR`iGTu|OdX@|c&SpX6#QK5{0Dt(N1RqeyQ}EXwc^lw+CN@> zHzvZP*)V$+VSR2&+C+~JYIf)7v{cGa&{?{Bpuw$gJ;+pq6s?vcSp|d7?v#0NhGu>T zYSX-|Qrt)u2VI9aLGK()I)!qjRJc{hafEygN$0jK*Z_xNN_dK>>)_8j^ z>>Yw5KH3(3XszCgHphwjL?1T*#SBK-S{yG_d-cD1Qf=Y6F$`Yo zBDf1ty9wv-cy>u=2hhL4s1^E7E7=ZzC(eVuim~edu7&nG^tkGFmek&A*>Qvr>4!cB zl=}}RZJ0IOgj|~3tSi#`atjXoz#h{s$e3QaOvrw`(k*NfTj#8$z_$&9eFLJevie&u z;O9&)P`~T%e3Ci_QU&={dG3Ts-9AuJx>nXCv>%m1Z);BXj&{G>9-T7L+a7DX^?HZA zquDxywYR`*fNy>$UhKk-`A#Hc#?KU&3DgQ=TZ~1`ND`F0Z6Du-;b|{r@dHa4y9;R?9M9 zpLlqjZVS~#e(Z2M5q9RC@;@|*GaNo(;#i74=6%&XOS1RDsZqt}Q*~v>&K)Z?uSge^ z&$4LFHm|!40Z|~sl9u6M)RX1D5|ryeJ4~ZWc*Domd$o9+&Bf6-v8nW4Q?MhFv$Qdz zqM6CJ5D2qW*%ITe zD4HNZZxx+GbM|^1c+IM(D@Fd!ks8t`{j7J#lg+Ov8aUY!B$!VuvUUrbtaP7GdD@2X zJ)GJ2BwWp#Ml8ABcpgUo$#3RTd=iGxPE==G>KVh~-{pwLai=yW^hdmmPm@eeG*G4! zZ!?~>fOeqW?$=M4;`d0R4yRrZ?9I3eMcJ0mA}04{q?z%z*W`ty-xNwp!@$XR$4ie8 zJPX!_8FBU1?$zzf(JsZsY3F?4={&{E%dl|ccG1D8?Sh0CTJYgE2i9|2B&SMPYO9iV zg#GDczi6pTYTwS**OtP3Q&-GrXpu>KCIQ6r*gG0L$2KJ(ZfL`uiM1Vqm~7zIE1O(YrlG6ZCMFzMC8 zHF*J|DVdzjLW513CroXHVWUaQHn8N$=H>QN#9oQ4Mui>!A6@Sq)Z`X!4+91aMY=$w zgeoc>q?gdaLlu=GT`3~Hw-6vyr3F!nbV0CCLvNu9p@SeDLhrqX{*gQLn|nNWzP~cV zFq6sq?!BM2o@cGK-48&;;s!s?4}4L@s6I!Ks?GISpBTweSvprp47Le^Gyk#~!XQ4q zidk*yWzNtAM%QZ^+6v^z^7q|2#pi_Vg}!(n>| zS$Op6EHm)x)a`~2(`Rt0TWAcN493A8XLO)NKnH=E=#7t~x!i_Mh1am6|79*|xG93~ z^w$6S1Lc`X7AfX0DEsVTJ^d{d7a+XFjdfdx1qPCiQs=9nUNIP+t>s73IpvL~{Jstw zq8*iT;K`RsBOOcr$K9i+N>W={eI{cOa?gHDzG}^^TCyJ{ad;Bs&!aJ;5zTL0=jB+l z{MnwJ!K!VRN{xxCpGs!wq}yo@O-=OxXf%miv@#G`ZDtX3MD?fT0K5PspUryR$SN0KVFd z;JC0H28Eqh9bKkR*3cjLQ%P5yg=EFP90y(G5UN;N9n7ivDb28VG5{we#WBP#TvMMR47(%{D5(nGt{645L}L_8 zluUeKbIaQ?TW=$8|4Xf*tAxw_c;)}-->O&6rnq5VT9{9tMbK!?;$Dpaq$7xKv89F0-ebk?~-o z9WQ1xxcbv~h&_2p$_=tSS=EEu8#+AFL*ckaYpBPRJNRrMf}P3gqZ!@yUVYqw*;0U~ zG={sxToJzr*?lJ1zaz5Qy=6!!cw=y(|>)@Y(ZNoT=i@cI`$H z6(uzb3r zrg;K>h0A&6y2zZgwNgM#oKUCB)#81S>kDBkG{Wr$T2(QK{F7jh_DcJ={^JlA>6ziT z%J?Ry{JHYHu6-47%KUUd#?Ig~vZFvH(2TL(6ip@@mVH-z5}Czi!k z-2%hXFZI`#J8@a>@E(wMxSKDZu(m=f_dq8S+2Eb;sr@y92jkv1zw->6=6|{Zx%S0b zxlf1lT6ium$Nmfq!Y$z6AwaP&BxR8UJs=-2(tAuydJ+ zAn=9zq6SeI2yftk><^HDp`oP|K;=m23DVzf6+5Cbc#^b_jYTcJW1l#Ory`~dR%CK?kZtjP!dScwa{R0=KxE}n#yM4pT8IL`SO+Lkm{Jq@-V}|8g_s z;Bi6YR)qv-hF2S|py1Pj8O zI6~^lo=MObA9Q%h^*ffco%A0s))sC-wXq6p<+B~$s*8ee*K1ZXcy@d`q`xqSvTPan_}FI(k+hH<;E4yU-3xV_Db=y&`s)03*8t9KCUoZk=EZRiC4ru>Np6Eoacrfo zxB|FXEtbg$_>Ek8vjliYGPcr<>AciMyRX-)ulIB1Q#G9j<&wpW8KY+Yj=r0v9bHw8 zms>s~3zIX)mHV}$4g7hBJy#Ao7_5JMV_@<>i4RR1Pa1Yqs{L=4anc}T<}-DR=!c{z zTC;z98&}`qO3&VpahrdY9@^QCT{4`WkP4{hlfKe~XC8q(h!=n3J_3Zt-6ya6e(DdC!@J;a*vd zONXEQ2g|)t=NB91*M8BoX|n!a&??{t!)d!~wArGxqHHv9QKYZ`v)pFG$&B^&<%s}y zCe%%nuNLRbxH=S3duU_)EiOzqrczH$hZi?#zv#b}Wa?!ED2pb;O`gF0rkl}Mi{|{?hAEWxpmj?qaVf>QPNrOg z%t4RH|M+ZihDU1tMQQc34;*ap+bIFv5+KgL_Q9 zr(a~i&epJd0lTx!ONT=`&c6CDw;N?Sf^MF9s!mWzgnx!0_m0uFd8Np!xOg}Q+b%xO z0@c_4{4GP3PLR zyu_V}ApBeN)nu%a;`ppv_uKjv86>Q`J~A+OczX37UCc*j`x`s)w~FL%KD^!Qj!U$+ii$F-`KApFs7b`ac$1A%xCnS ze>hK?jrEUviUeqF9^{EPJS5o9iIw_c{iBZza4_yV+*~@W0sQrOHfHv{`|jmzJ=AtA z>0CyG2yS2=;29muRlXvv7TEzbi{<#1PmUVBq2ui}-!o639jCFvZ)0|V9YdLKS=D7} zuhj4-;-8V)O0WB-!x$ExOoE>DU<@+PhV$olfLANwuj6DZ(>#7Z;p_Xu>S27|QL!TB z=45t#u>y%}Of9P0t`&x2cd>N`!_RbLN|MD~wGMXL9A%Tv|E-Rl#!Z0(2xb1A!4^n2 z8({~VT~P~2WSV})9IuFIkGqU6;3%=k#Sn9+>5@Y(cK0t1`a5i8yrXOnrf=%Dbz3Br z(9(3>%u$iyfeCL$>MxXxK-3C$@0N5<3{Z)|$S|EfK5vi?Cb%L&K$*8&Q2B0`<5tqh z#`>=%9}-KF%Uo^Me}wL>ET=lh9#8}Kr;5Hm%dfmp*AcY9pB8<_@I?&SPP&5E}GTdeRd?%(E7Wz61`ysoZYH@3S1%Il8X0Jsy6;|qhZ(k&ShJ|@xcjA^A;{v<;sM+rKt*aRLI96DV?JuG2eb* zaGJ@L+>+dQJB9!778omkp+i%@0vnq;+KLu4{I`DG6n90#pO%sT4VN8wO^9{<41QN; zHCFKz96FwPX}GUu1HVl&3A~HVmOhzd;teGWy~`U)b-~&v=fDk^?SLhc3G+8z?EBmw zj|Pd5^w+&Q)5$1Q!IS^i^d<+FI)sa$$S_iUgZ}mfH=>18!lV6n8pPpB#~h3P-NrN2 zmS`MR#J}Ze4VjP!c6#LSI^;X%Trp`1evVs7xdRgBtjuRTw@|ru&1R8%E$unb@jnpd zgZb5d5>>q*ANx?+ep-3CZJ^l-nk_a*)*+`6^j?>Hg@ehm_Y`?2R9b=>YLjaCEa|H& zU_T>Ut$rD{Q2Pw%G1h*mSb}nzuHCOF4&YPFQi-Cwo1t^E?cFGy!G)G>yKI&Hkc02r zUpoARHTlH%IcksM_-l#msKWvaXVXd=F5Y3jUUN(O?ZQdKdHmWxl3?iy=L>W;~JHTJ(?KS`xYHo0D~Xooxy=M=oNXj?Zjr zz|4M%jz$P3pk_0IXSr_x7`ABtWQuWSH=JSc33zSv_+veIRe&_{V*2z`F_a7bI?v`l z=d#gHlUrA)*|g1$7XKoFc3(&?+tw>a{Kw~m;D3whr*DwZSo&GUzp!NyAyy~}JXp5d zD)9A8$q$xjf82kG-OB){oz6)FhW3L$jz@%T`LY~y1aUab-gd1#P?Cl0(@-NzC?@4j zYWgbxK-{ntTtG|*yp*^NvATvFZyy?!{gAqay}Y&0Wv=(>hoKqUTTOLJcHt*O(OohW z!oUY16G6ZUbg~Wg0A-|vVE>qAZ#174$DPL_`1yE)oL---AA$GT0u^U4zgI@)v_b-$ z#iy$T>Ws#N;_Cmp#X^{SU0-KSIe^@6d5Fyx52fcSovjy9goVsrPb6KhcRY`w9ntk z{>5c?rCtQD?Q2dC!QC@qRTvY5n^AHnS|Y#ESLqrnH9nDx_^NDf``QTe%dp4# z4B2d7+@OD~0DNie;9P%$u14-zUJ8a+3_c5V;bzR+jpurhkkD{Lkj-B#mv1Rq-~X{v z0+TqT;9;=WFe}MPvK=uznitt%9NI3XdIPu@teW-zK7a&TK%#eY;si1B90S0||Lvfxn6M>ISQ+t48l*wc0Y8P$5Dd2Zb#^JW5I^$ITsU3H6ky-|Y+sZ@g6-~8%01Q3 zVy!8Jq{kCb$YE#Oeioyo_U5Am^p}U`e?@69yHNq*qonlwye&Gr_gufpzCv6n{ch6t zi=+iThR>l|RSobDd&MRTfC|O|ZzCdEQ)hr=Zda$7*EMGfOxm&P%FO{O!6-KDYEB#- zf4c+ev&CFL;jiTpF9>$6*Z@>4z)ySd)!<^*hFUwFa3Se3{7WW>U;J#rs~GDA<)FBd zhrh!i(qj|8>S)&V;fg~3`4wZ#*+}TlkYn9eCCyEd+XyZ63>eZ=$o-XcN^W$~<+=PQ zP$XjLM8dDK?Ng2aTo`>KX5If}+Pios5Vpi){{8fy#M7AE$F|nA?Cn$$Hv|?M^8^e^ zLiGe=1!QBQ_#>UA?vBb8r0;89|Nq>`yumaFL@ zX7Q;f0ZxCt%QpZq&YRciq<&bp^Y6dj3pR?8hu1Cqb5+Xf4DW$Yz&n7C0C@8nyDY)? zOL(!-Jv0)DYvG7{M7va=GxiQ67qcJ})gfG(hH``>5#hBpQCATAt?Y;%>{y4K{&?_D z)s&!hj@m?CBfZPKKw-zY6I_RjrV;x~VBOzF`0xW)^5z30eH=zxAD~Eg5Pd{5C_Z>7 z@W#ns)^>-AwmnPWWF_6(EXG`HP7QK4c($FLbsl&*8u;RVj^<}vK{(>8Iz3p?K|{-B zzm)Bj-3~DQ$-^3_O!ZNXtYv!d2FusZu&G;;ndAHIut%894Bo+)HyBIU@qK@0D#ik6 zCsz2(;eWSMdoTJL{Y3@XC12)P@@6rOWZQbWUG6w(sBJTvt z)4&-Cj<(DG+Q$|-^7283E`IgVw}TqMax@RL1;?bDtwD}DvPaJeH>oTl`2iuklK8f{ z8F-LacY;;IEgBCH=#q+TvWA-~Tw19}%S(?{i?mzWy(4El)j6WX&~;Dh@^EN_`AcxU z)pfkHzl;Z1u&t7T(`p7a>e-TmFKpPNx=QiD9?+!z}aAA^Pr_@~IB zb)|^k3mBGUq3?1M>gnF6`>n(r?7e`jU?&(jsYP%JQe8ykXUU|p@=$^aPLM?iP9!l+ zWYW(LS$x0=5>sEQ<=Zt4tyJGEV0K{;yI_h_wa^pDbg*l}5A`scu9wm>sx za;1hCuFw%|a`@56k!XVmWNp>3*26BovOO-kOstc&@_v&(W52Cl`5l&B4`*3H7>w%G z$k|$hgO=Ci&8i}O=2KLJpvgRdRl|ylU=aKN?A{Ujo84O^kXx7e_e+-yk%x~Ebg;!K zGD%7UeiMfM*^7};SV}U7)xgK#z3PwfofRD0=!7*@!R%xTtXXkP^3OZ_lvIJxpBme# z&M2eBH0M~hHA;MBDy1{QZ=l+E1o9H$3?!<>fnF6SA04&Y`*o{bJjJ~G_-qk}N}Fu= z*3J(yuj_&GoLRi`#vj=?5j4m%{Q|?MlT5H}_spMLsz>(zgY;%T*)ts5(XG`~cLPtn z0}X1eWsrCNR26=cmqtFn7S4wX49dDu;?E1xjim(~kBd4Pr`JZrpbkTVXrm*T^;FF_ zt)~dCl~P^%rV?b$_MHr!>$Ne?J^*2G$JU}G#>K|mJx=SC0F0UoXM7)-aCe!}6O3)T zeP3H@Eq5R6+4Ivfk3R5sM&{Sz>^Q)?)sV$9BJ}PO=Au6ipYmjxJpjGZdU4n%y;DGD z0>P8_IfpUBGo4bZ&dOlB-o_^EgpQ(bQNR0{t66jSz65bpIPqa@?%qJ@zI_s!4}3 zEmQwIexzV%mb-eS!{~Pwy?*Hc4q_9IEeAF1J00j3Bz83y)U2vgYQ=gelo=_Cg|t~q9BnzCZZERXrIoVhar-%QsSR=k(nvcSyfiI@0v? z!>Q|pUH+%c)TH-kl4uvL%CZlrqXkfAmvpWyM%3V_vnoDV*sjyXhXo{T-7AT~Gn zt2aM%bT|v#VQ=J#&dFR4Q5t`}J$a52HmWy@E=pf80e|DGp0|K-g$2u5bHu1L{*eiH zWZEnYK0)*?!$M@f*FSi!{F4HAkid|R?n_;n(5z^%kxHa8AXCYUS+S|o)9OcDTZt(v z^D37FpI}Un^x0|-&(HRQRn00&l0yRnqG)GgcD~iqFac=xnIie`KO$CHFiLOuG`)$7 z0u?dxmDMqyS$(xSsc|ywwA;L>@G9DV9BF+Npos5bBnPbJtC~WptGyT5N#?fQehXNwA(ZT!1pGY*-LI3gXb zUgqUqVaLy9G^XYH$pP5MjZZpFqOVg$99CxIE}BLvkQUa9d(H=a-h}aX!d|`jYF2v* zJ>Svu5gG9-vgS;v9QsVHaErr*jfp)>kJe)Dj^B*Okd|mB)q12LM*MML0inn9@)z_% zKBK1IcWbsM(+*X0$79D{T5LUq>l_~OhGt34NXuw2dd_SRh`%1pzZDLAzJL=|Yj7X$ zZvASdfYdMZ)ejp_wp2pH9mh0@$uC+)x^KW~0q+*B7ZCXbQdA;`EIW$mn~+YRf_7$e z$kaXl4@7394!#c%=RBMC{(H|BsNR&~EOCVu%OI24msz8d3Y(U@;FM2Z-)R&T1`PL@ z*ut{p!`b?w3(jS|Vn3rDA}<`_|Qt? zhb6TKSdpW;9u@$W!8k8EK?VwOAFb5Ncx16^{5gzH^o!X_#c~T!{ z5Z<;YN;UJ}s<5WF;&X7ETQ4CB-yFJsOrF;~UdR9UD~)LiSI#1CzVi-tGX2*^7W_#? z38f3wc6n)GfH0vtRFV3&z(B_dhHI6)WK@H96Lg?MElW&yPXn({0=YA8Q2RUo?V5b} zRb{QR6Y--BwQ!aXJyo~TD|ANEt|$oz{t%3+-#5SWv|eu-o@g4!1;<$)^xG zmHR_U$Lp+sILYN?cD(p^XvKv@M_?SHzLbZT8m)sD7&(NkI5V}OCrC*he81D4=&;ey zDhk$};wTKE}hD;+wwJS9TXO9M(4Wm0PwFLWd#+A`p+rk4Sf=~;Tkkqrw-MiJzE_r@lb$>%ERSQ$Lfza7nMdZ@o`@7zm zl9EZazLSX>EI&5GC}4anaeR09{1sexXg48oP4q7#>H@nB|t$8E}U;{4S{ z3W58CmkmnjYBi!U?H6D{RhKR@#Sc>G;dZZF;d=qCl&9QLB+Hwwla5t?8u+ddVen&N z!5GGNRB2=rCuZ#MyK}P23E;^6SZ-9(VChQ4o)Rq+t++f;T`0UWyvM(o;2>l4<64W#{(aNC|A{dd@yRoX_{d#sG_RiCGPxZW9 zG6nlAur^li;g=z3UmTCkm!GA`hEim_|32;L9tT5dShO`HS>pQQpxbN2WCO_SJ#L-! zqt$7II>Kx5HYR0t%<X~D-JAI5@149tNts2*;H4^0OnW|);kfBg{P33c zz)P{utb%a7(y$VeW={0x%UyH%nUPO@#I)Q%rwtur2lkTtYAH`p2}{f=6F&u5V>hX> z`5|KK>^8p+K_YGt@q9p5lz|){k?X-#{I9tU4)b3w<0rFGW?5{ z<2TFWV23#sp9C@&Aj?QEayb^gcQQMtWxj9aIQ9{n%gHCpS$u(1Ug}(<(3!d(A@|=R zj=k@23e_WyTW*NmzX4d!DWKzSKD1VUol@;)sYXQdiWb!y`&=Y+?e~O$m1((n*d!j|Phx9^j_R zqZCX9b{C4KMren`!pL!|@V&UT_0gmMlH?PDo63wKeQs>?pUSKvk1#-Z8Ol5h?ET$2xB(wo4;M)gVgXG~|>PohY1RyKYN*;nXa)mdx!7G-6#*DecxL8iByNU?1 ze!#xcc0AX)R%0dc<-n@{D`XO-Sv$G%S8Ar)+M^uz5Q4dHLUvJ`|DWNr5?V6 z_^BS_t<^;}K&0eR?=`(S)b-U;*C#G+`-mNq+DiTbHO7?A6#K)+c?D0}zhqodZXT@k z3z%9x_>u4h6CEYDcUd2p?=Glynf45G{&COScbiGF54k_Kctbh(nF}Mvg@x43=A{3JQ~&yhTudc)Q<_KjfnDgT`Y8k~vi{bL)5Vdu!lOFON1SCOy$ zB;t~?&3sRsr1G`@PYb0&;ikk_C=^eM{dM3wJO=_dNiIm(xz%|FqC%+wI=VAi2*Zrd zl0j`3bJesU!4|SX&a9nOPoyH?Q|G56B{Ox!<_iSA@DepIo$`J2@OL5sXv)Y$e9r1< zz=8IF)USpnSAsEu(;4odOh0=!ZM>JJZ6Ra|Aa!88d!J-J>S#rB8bMy-hW>{(SM7{# z0H?mMRTE^_j#kPu+l3vE+eRthRnKp902@86#A@2avO*p3)hd1VJGpt3a3Ru<(&+eJ zK39OY@ggZZzYC1VMbZb5pfkND?7}j4zIV!Yvxgt=&MxcPC@Hu zDfz&~X|GAg_mt)Q(ey@P%d&20x}3^w&p0c^Wj4Sx$6CC5q`SMp$d9w-0FwQ8je|Nu z;w*s|d+%^5W9SD5JF7WyMMv%Wwad|a@y=SiA^29GHw`Y68!tv`En_ENdf(700qSX! zaxpt+Qd4&-mi2Y-)zNL{^`@@ure}>Ka_$D_bG6n?r4eFEDWy7nR{6Z9wdF`&fyRsd zmKbK-Jr(czq01s%N|1KZ@_g887WnKzQQ5Y?4OxbD!D6+!={pBoWX#Y6r~MEGTLHHk zlWSYs;+uwD4)Wu#34RmR_b2DfO1GD!{Bo~l3ZvPf*9do zZF*Fl!H!MMD#LR|DpH$x(bXGpD|j@!Dp-WFw#Xw)b(w=BKC*}KIugREo+4K%;m9X| zpiatfbPwo1<|WLExlxrS%9m(D@S%&e%z$;8NQ&=sW`U8nnmyg}jj4l1V*8v?r#99M zr(T)SOrapzkf?Uz3BMtqVQzX!wlYM4?8#J_`g5ayT!iQr z^uVg?)oPk=y_*T?pC2zgRbze;V7nn#TZ-@fYJ~0oWdZCjSB#qR2vaK9ck5u;0OKm{ ziCa&sZ?Th)?}jrvJ?M##Y`i)u=J~dhUciAdeNC}ApJ?o4lxu;evh|Usyg4zXygnD* zslBRI3j#|D+s?0C{NuWlE@ebNCLp$0jO>7_J!2gIio2;NzH76P8S)f8rvudu=W{!o zurDqIll0s`wZ&ustIs0q8wnWW+lc454skCt1aDRY$mJ1wQ?!uyN; zq{f03bukb^U1`5liQqxIvcMKqWQpA6;ZRgi1=oWf1U(QU)ybfUWqU%&Xf>gSAO|XV z#y>X4w$Q&ZL9r%kIASfs0a|CwT5fgvjH zo5zl!=tS+uely>tPbzYtTY9<}k3qUUNy*>S-r{Sd+;SPEdSSeae|FPK$&Pv}D%WG_ z`Q56<$!{Oa#%1Q zKRqJdOnx2v{WL0T&-5QySWuZ+GS4!KPtqKAy7X?;JTQR3PKBJ^G@QW=EzRk#{s7h8 zh&2MTT8RpGrg5a28%H4Xc5AFsQW$ands@Hwxc~Hffh1=N8DvYOG3^;;67V0)v@&g% zi^vXQWpQ)rrrRiB^r2m9*-DW+*vMmu&h;~aTBQo| zEdI7|GIIe)_){UrL~Coj@Fqqi;A^f~pYP1de_5#`u5su{wR!w^gQAQDzgZvON5N1* zU%{c-6#wlB!5lFj=_^tv9)cQzF8DQkIPxpPW`ZOyf|GQ7i4i^#AMC?Pf&Y!d>XU`= z7cIdI&0Kt>?yLmfyO$Q$pPo z7XugYwnrYlCX>f&Z?PR=X-dJWXTmz+2c30$H7!Zj&S8|whwKOOIvG4b1>+H<=zh<| zr>`fwtrVjb4q{gh{1bg2cCc(gHNm`#;>-ArdpUTV88F#|C#&NK%IBFf*43&tRWbLO zaguyL+bed0UWztQizXxTXl{b-7Tr5s-|P>QcmHUwkqYx`5pWN^J%RgKp%)&NadYH0$k)+uz;|xJ1ANIpSuF8uF*+?FEPU0($A!;by1=o;Bu=6F!_I<(4mqg4+-CZ*aiGYN;hM>oi9(z3v9QgLN zzWl6(8Qw#D@QJfKY_p$H@coQ<#xJoB0pD{3{Kj8WJtUydB_e39UQ_pSQHQiIna*&? z4!eSx#{m2uh=4E;C2TYCAY5I6NHt=qJ84uupF(C={zyQ?_U3fplf|jfk&N@obl^JB zt-3e1sVZ$9xcpBf_kNrSNN;^m8U#l_9Ygbf)LNCtBJuZ2x;N!n-9HEJ^KiLytjNT* zJ4AZSXI{{c+crq77 ztn|fh>rV<;Uw-Ey6Px%Ux7(VA*?UK&%jDU1MrU4gGtn;1^677zR&Lfp;n3eR$>rBK zrRxcTgNKCf~~B}%rW1Oj-5nj70ZLc{O`dL(ez$f{kwq! zKItgu{|pap;PNzZ5rx9Z$Nz`ckVQbLSMPGtC)4@dl~G;w6?h|{&M&deXp%8btz9xY z_tF^OyV=ypBa#>265qUeg)(%8SR&zJRIl{p*_wGOpu96DGlUFD&FXzD!`1JuL&V(4 zM1(rHic2TVBm6UDd$=2IvLCUiOt8E7gH#C+!AEELjo*}7H@}$U^3BSOM5k91rtGkg zT$2fMKHg8_eb#BE=$Fgtco>rvA~fj}hvm>|y23@6!|`1o`jhOjql>0-RyTHOXb5F| zi&ICeu-_rq5S1l-69-dFj_PgIQsj%6PwBphncCbYXg#G4WX{Op0zXxB+f%yg*|GXo zdVRW*Y2FrdT$v`jnicX$i4woUXK(%`2MpT^rh8J|c|Ucy>H^V+@&t`nqRR6gEaL?8 zPeJ*>_Pad#wBJ|Hla>X{Zc5<;a@Q*-*lOKB#*cM9ktDG;7#r5HGL!Y*GmB_!J$9+0&WrZYEb&v z!C$^(fgbDb|Cvh((%hhpG~&0@dH+!nD9UHOPz}#2lrO?7%_^8JV}BF%oSL)hSMDbH zVw_)R8dbf%n&3T2l~{)?fLEtW`lirg!WqFFnvj!~TE0YT+~sT}Ur~G~9RF264X0Nj zSxBo#b`FXLeHBnBdd^dQ=XL=3(p$&+nL8n885BpZTH&?r+k)X~&mj006UGk(xv>nTP`nS1n*)W(X86g2#ru8;FxeIe#rRonjMCK0hs}kkoxoOI(vSr{$J zRwGBYA%y96d^qxXq3lK9g#>ms#J=vORd&T&iIY`(enb1!CNLFdypnd#I_98IHdlgO z%im3}JR;22MVO*D;K-_f@O-~pK%nLVGm_sPnme3-s@U(^z#a&_Q)0GCK^H0HW4Iz? zaQHApFFZ%mk=Nmt$mQncE)Bj{KZQ5WkyE-IW=5kcy#w5@DgVn2bR)ff~`~S1Lu$z|R<{6466F!^|^8 zz`8^4_znEsXeg9i3&LB>2R<|^Vq-bIEY0Dw?^KZ4_vYFNK^1_Obzr&{y$W({+@}@3 z*7ha`gC`ql5mur`?M|9~i`HgF^rzJhDs5SOzP-!=YywnnR`KuT7p~haouvH1M$LEi z+c&eSu7r~;oifg}SzmqJUPSN==~Wh&8Zs7S3&vff9gvNHC_rai821QGZ7jp2d%vuK z=Xh7Gyx1gm1RO2Hf86zf=tG731HIz?UzFBKTUzPocxK}kzs4!4Qa1?H zEPm*Elm?l2ubY^6rcQ~pVq^zQPl^pzHLkuduf{p2&C!1nO&ibt6eq@fv3)ohCY);w(}<<)*;|vTI@xceo=g*xl6YvW14M%;#Mbb8JEH0;*U4tw(L3;)otbC?PkaVBw*XT3xCq&JqF)F zexKg(<`l$IG(t0e(4HKzQV>TAfx57+QZA1?K~@NfcPdA=1egTIO&Q2(X-q*SPi3%tS}!;y7GC zaIu4K&p`Mn0p%vv_kZT^|9>lm1ORwTcuM%jI+=Lb|GuGs56>=-lBz0#jLk|O$gSWO z{4xTsj1{TLc!OW4AMdT^#ErNIbj5x#7kp0tR4I`i824O3hSh?S*#mJ$ z!g;2C9sJ}04+(K3(~M;Va~B1+d)zkDw6^njPi|b!BPSn*lR7-4u94b;Fp}ZJEX;Ph zKVm815?2^9x{ zX&zAQDsX9a|-j0Q}tH*rOl6~)Dw`FSAMAkm|c!XMb zHP+`v+$LoX@aY2NY&GX+dS!bTto;WhY^$I9*@q z0KJGEKnS4`p)p}lJOqA%<{W2*7o*ugJn-gmmRQZ(zcozNg!`?MiAhKmNuP6E0P+aE z2>q!-k?=!~5?(wl?i8^@cPoeX`IqLzWE3G?t)+E+5~e>-oIL;dR=px43DBc95ohWD zGng-O8SY6qaI}-+O>W3j^enp9{kA_VKLW4PPMN)lf*4O=@t??hWAO`eZ4rF*)|N;b z;rSabk=CMTk(oYMn9eHA-sC+wmVQY^b;!!-J11>D!@*rH7i~QIdzg6%lgD-YK`CgjdU{ z8$CQ?wt+^OxJy3ke4R6Xw4%_DILY|8=&lc@nJE2XZddQi%h5qL$J<*vcQSHy1qH~#N;66~o5fT171~C>0 z^=>S=)rV|}^FTz8f-%7)zbEd-vIcYnX7^k8>%dtpI>rM+tK6fGYpTk!nbVP+qd2Jj z2iKu@4ks=-jJ$`WItxdd%qQfd@2_S-OyRYvvRj|h4yGhu-;>+e_#52Rgxs)f_l`u_ z5C6r%{4@hl?q4(|9@zvD;kfxGIY^itTcd@0F79di63fWaFP_RpX$2Mn`#&BXAix4?Nd;iW%QEj zDXXF5PkE3?Wb@b&F19dUVE4J2ps8SHA>s90pO=g6PG@H)$v5uQ_e{ss@}sH<6MrKZ zG9PgdNZytyFD+dLV!yS=t?tYN*}_sAZ|vaiko&!gvyUpSEvGscMGI}==IluntX%PI z8N^cu3SmrgTAAfExjy|+91C0P=N@Ck z*CJ6<&xxWkz2@0-Yoo7w9hX5aq3_MWHc!F2TW@kdDAP_5BlW&P-Nq~jA;d;J~f}QObA0^{p#3*Q@ zdyv_OrmD%JBDP*SjlUmEA-)rBj!<@n(d%-Mc7<&=n&wk< zBQ@CWd+pwnF4I(#)>tgXi0b^iH6)W1T07QJrjTadYyKJ_Isle<44s+|hloNR_Us>m zWpGQa>!6zWH-t;_qfZk38HfjhGu=OCTY>J_FwQ9D7!lcxEPgIqy=7=f!R=!#sxkwG_E`sXhZKZ9lnhU}|IIBkd;LQ>f6!9LA`q{`B{_kMzxF zo+`bg1Wtzqe9E^^4#)gywM=ga@5#u3Zfs(V$+IimCRt>`&yz8Uch)kiVBtb@hQrJG zn@yuXKz6dgqYv~w;a5EjCprHkJ?*?p?%t$%uHFA%6?t(V6bcfA%JC)n=22eh6}p~P zpR4i(nyYU0Fz_M8yT#(~LB1-RwaJU+C9s@E z?x?@!k0|%T3(u8G=pp%GaXq;CKnF6qR@SfoIcgVpGi;Sl>~J0tYXx@2 zr|dR`9|$o2@uF$d!f;b#9bA(hb?k?`U2}Uw?{{7Nk*h(`t7F12@A03@(J#v~8^ws7 zlCPc6U(#+1_%Z>|w!D3EeHU1wlGt%AAO8T|F!8{jMC_(R=zu)Q{U8^1&&va%j zV{%sl1Fiy`@h|<3a|1o447k#hYL{t(q;fTS$GNqL?GP3ENX4}Cp>HoL5$;Omof8+# zApZq_Ap|<3`uOrkn!X_4$N+(1eY>lt7BVLbj-ZtA4X$U+mmJ?o)q6+4(Fu-Ng9Op; z2xU|-{jCuC2fggyS8e!EPPOrW!xZ(yre4&nJwAckwM6NuO%Bv7<)*b$lj5VNkD<`R zKFuu9tsne;;9mTloST3}JUp(zLnG|@Yqiw0;UZQm8d^n8cCK&(skOO4^>`L z3t0K)(x-X9b$CY(4N0A~`;ij0qjK?m8o_&v8GbnYIE!#GHDI+FceoFyG^Ooe@SRK0 zeojwtA4%U{=*b#BFB+bsrZ<&`<}ECDwmt!n-%hs?)Ilf_tmfCJ$HyIDTUi+5`IC)%Ad-N(9LAR@Vm7B7WON^pW(B1!jrc#L9>ek)P={ zGM@7s>s>}fR%xUWDo`#W#qacBVTGD<^$x?#r9A>@g+eXfFQFO9Z{O){g5uA|4+)GM z{#9Sq%oVx5t)#ZMyM~Yu&_>?O+U=Cb(9f|%NUmj+C;Nq^f8=<2C%44EOv+$Km(i|g zac0WEq<^YDZoq}J=H`SCv9_Gy(0@K$3s|vBTOAXa)b^!zGbgLF@B=X+uUBSk*Y3E< zYk95~Wt^YP%v~L~ei*a9(<3Ff%=sOTlH!4dPfVnaZD1!%7ZU35FM(+g%HFg~1+7^2 zE=>(_1(Q~$pq7tu+7Q1oU#e*X2v2t21hFzG*RA80;Rpfj}=|}#oV!?$bR-34F#TB>z7!A1rTy4|Acl#fs zVFLi(-oYdSl4wY@&lsW>Fz@dajVJQg-;iuBM{_p#rz!vnbpQ=_DJ>{nLb3ykr;4E8 z_5F+2&s39yjWFGNOa^4>_zX%r9ymzRcSO2YY=hy@}vx z4y&Ya$MhG|dbY4boJA;laLc7OI+PxY!7^Jao6$Vl$f|h`5OWS=8cO?`TCZe~*U3xK znurreuGi_Rmq(sGU|;;m#D6`CURBR79IBMMkm>Uxfd%aprDpOq+2Ykw^IQbyAsNB_ zJ0Wh1hSKY>H^Z_8fdZe{(Nb~>@<&ZIYtbH++!UTQs0g#w9&UQiOV%K?li*RcNPOxT zx~GD0VI%_&(@5O|U2#(a(km>9g)g|zctw06Z&wnS)_m2$sZHx3L~~Mk#N_agq%%f8 z>gf=064NvD57YsV;)pkV2RN02gBIL3qlpa-Zm*sx_g-SMzf?oS9pxU94`bfwzx#|w zZZzQ&uW*}dA^rDh<^zWXzh&9Pzly@M`x9JQ$~NjAuJu}o<(E!8nM;yx$&POI#!Awa z@8CeD#(5X_u&>^B${iPNy)+8!P8k=24e(!f9B;4Gdd$~fU@UM_INYR8Tehrow@zGQ zpKY%SuaNYV?y+bM#&3w?C(ie`ek0 zcM|2O0yPS?v==#>W*|UY;jDp2@nD{drEr0&r6zv+ZVr;61dvJQa4T9dM3weIqyaw_ zSOagP>V4|gb~QOEK-jdRVD3$A+qE=S6KArD`C7`_S^Pb0KQeYxZc-~JUpc11c!~AQ znK4+@Nt?n#oc2X=hol?(i+Q+2tMw)Ea3+QMTC_May^xDZPCPj>lt-fq+0jd0<(QY9 zvHPQc>tk?*H%Gu4h4l6+*H=IH0j5C-F@-; zXlI2B{Xh|dGi9mEtyA75zV|i$oQ0PbE2OS?a`z(iBink*9Y$gdL@VN@+t}l;Xjlia z=XAAaIkjBzlvRSEWaW(7^1jn_W9VOXeA6%`rGqaac!`}8`@u!f{R*8{3sFfTo^h;V zz$8ova+7^h_^;6Rd{^2N$xjyo_-*VJwhnY@eRnfXw}1Yp4NMd}&qaI#MryRj{t;!X zLMiQ4`&cQhZ8-z?{5g5pAMWZ+8L14j2Otjn;+BO;L!wic8Ncjp5fI?&+{Pph{99}OXWdjx%4^X6^fZ2GPH2w3;|GRHK|)! z28r=xSj*8Pn;vq9rjvB%2BP{PL7S{=<0##0Y2Md1vX~);jFfHPkyQK)V98MnzuiGi z4Kl9uGlv{MkUZKyFSO>iuLhlv5&w9N5Ex`yyU*~<-K*R9lWtZ(90i)^7;q~7(28@W zX&bB=hd6yXvPND%g-{NMu>{#DCkEN8ld80Me#Id{uoJ(tgt`(0VpG{~SxkARpcrA| zd({tmv={=^Ff(`SlG;_LLPf7Z4p=blmqUfM+D)?lv=W!5iZeABWqXBi%ymP_s2@4J zTe?a)0yC@UwOQdV)BTv28aEm)s#YA#rwth&_i4}lU{o$r^2)O7{+e$&Bz`Ay;;Vrt zZH5wKY$gig-M!46c++%pIvwICYnrcQFiG&~MF}kyG!{z{suAO218g5?B?Z>h7eR{-G0m7FOeUZc@EOoBH*!P2*W%p= z`DHj}k>CC`&t+yzE1%EmfWj^>pu0u2BCx>!$4_3$Jww+pw<4xm+Et9h#NfF}kc!q# zCKk-`G+V7vcB?l?-H(@yYKQd@*ub#Kpr3oAD4{03i|_Xi52vUR#yIP!`&`D1$|Nht zOKoC)M&Q1Xf+^i&G5<_A-~N|1=l`FCg&(D?FjNHd9%qchb$Q6_c)zpzTImHv|FEXQt->KiH%`wb!>ubRa^*`gOSV+nkokA1~@E3Y_m=QB#! z{@o1!zePnAQ{%8In$0hTxPrS}-|dPw^!l8R%`$#|s=1)roHdZVdBZ2Rhpge$ya?k@ z$f`hSJ|l>7t<@@a#c_0L*EP>czx{S`CVKL1qbN{L$J`v!mi;P=`Wg1#J26X( z9EPU?bKS~EtA;;%Z=d0*#;9n&{2*B3DQ%Xve1Ob>VT@@45hB+wIfe3$I4Uk?$^%x4 z1|uh{_7#bEmoDi!C|!Dag6*e`t6ZOP3iF+#X|s#*V5cuJ!BNF!%cfOjx`utl@0e(PJS-N4G0C-AA0hEgZbCKrx* z3|@==ju&0Ui_VkDmj~FA>MjDhqlI>4G{F&98dNY-psId%hplcfR_tqQ*2fD78{N~X zj6;bnGSz7d<(K8$NV980r4d`P45L$LUXnc3nv5Q-D&Oz)aa_{iAhLQ4x@aJJ+A6}@ zvvV0A#PCc$(0JeOF74^K_z~$J|6M6nF!M8Kp8vsq4Vxp6`!!TVnvWmUe#zFNmm_ug zC>nHg?#2kA|9A{v)(8G8)Lok$F1kB|Jjz<*@Ibi_r`|Nnbos&$)f{=Z8X|QnZV%f z6R`#xBFmyctCUF*+fc%qMWEr7aWRKW_##UghdX=kUTw)@;4P@ZwJ>P)c@m(36S{h@ z09P*rph(paJ6t=@%idC56VI^j{>ZrY0*h&Pcr2G}0C zh4e`B5Tw!b?Qyu%rl6{)U)y7>)MsH;Rj=>`Uo)Q{?gwX}km%qZFr3sF$C%eY9xDYa(^{%<4E%*u>7A zc6#rxSU~mhv%Cz?Z_0}={QAPO7JU{we4~msKad=ktGq5z+wIGJtlV=iqxwvwX?Oj* zEk62crBEQ2*H*s6KNs%$u*Cn;-+?B?@usiw&YF(jN%b?K9z#XC^@bivwol7ekf@_ZhW(2U@OAPtb!#fW2TnN2FrHIG7 zRxQ+4k-~dth5x_PQT?lw_f-F|Qd&_OsG?P$ju>ID0HdfjsKgwpmxaNXnGFbx)op=L z?F~bwMa^dD#YEl^S+L&Q&2T+%pw=p$L_R$G+8O+Z-D>bVs^LikigM+xkkLB}5T&Wc zH57byaa#dizrLCX-Hq4S;JN=sq( z$$Zm4B|Cs-5s|M+;fcm5x;5iKX41BHrgva&Nay!W3dEn==T*+FQJXlWZ!xbT5(n&b zJ?-3-txro!eWa~RNg1bRs5HC#NiTexo?x64yR^i%m~Ie?j@E0w6BrOzWOoM)JHCTO zH*c?ic*Br6%DfYO{g*jve|s(W@*n=?%}~q>_Zls_oba~du|<`$b;gM}hL7B;Kj~FH z(-GJ;*Y^_;h-6&b%csc_KbP|`+YYikYZNv4j`$if`8Aq=eE`Go;{|XffC(d9i$(mi_Dy= z49ue_%l=03tu*`k*1g_-N!5NbNc6@%(L}&axS^KM35h^BPa7?Ro^6ps3@rnki3-U} zgQeY}D;{fcbk$G>6?f>|$!&4g^Q+wh6}w>$F-x&vOb*@ie3lu!kH6sXdWK5CaZJF= zWILLc{FUlp1JIkq*7;2Gg;6$@fX!bV9UNe4{Bs~2GKB)sYg&$Oav>!VV*zI;9Zpme zZI5?8P5v2bI^C8AJ{8)-7_R=6i|#T-uExkr7>8ubU3WvDXSn$yfwr$o;Ayg`YA3vQ&|C^Oa{RL1vt&-tzp6eL*B6dXxVqR%MUYxKFPqR3E`MBF zADIXI)`!gg9iNC9G9N!FuC1o5BPYx|6;}na$k(6+$kzZ2d{=ECLgvwFkR}>Y?8S0A zL5J>I)7x2x@Q$isAx;#^Je;2wT!lx3h^EI|EW;lPXuI8cR_HPDL}d1z1!2DPZpG!+ zScX3JfTn%#?jl;bSz`G9BZmv8ZN`|Z@tW| zmV=%Zc};!lL0G1Wz!pP-z?WEhRxo=(5)HqK$kCnR;qs|33rGi{MM3a#M*Vkx;1{Ze z-Rab5UXL9S=wZ9i_W?pSh%d-6WR|Slq(#zYKA%+q(58?F>TJT5=GT>g`L@OJ`I!_2 zfv9AUU<&TM`efeej`{V2IQiA3>V3h9ub*Bp5J}p!S*jX2@qZ^87|Sbdzo4(Bexs>? zzd1SX+rq3fUWM1E^#-U*`Ja4|>06wz79_nAdnFqzkN z?Kqu^2(>$mUoPxYok2dI}(<_cJfe!k3IpL!y4*S#i$i7H4q4=7%39 z>nm5cf+>6Oa@(Df0@t*!3{g_^+}rIZ-b>zJmM58?S5l$LySmb;L=ddNcHG47?eUk! zl-ZtQx^-ul=-HHorc(L!E}9kL5yDHL;M4KLSRhhz<=eqi(PFLwU%KBTuU2bby0`zh z^J0GOUh)n^xOLJyb653f#9~o?$8UBs+OnbV5U$IPi%{i+)0VP-BdR1d5zfAl4(P;w z-qZN#=|9lONXpxGr~6~~AZ#!7_R)*}e-*JxFRuv!e2uk!_qvmI?7IUI ze1$`?{BJfzS#S{p7w0H#&hNPOhrAhTDSkc_k++i~%7i^fu; z*Lo#QLUh!3muBTtg?0vHKIV1&P!IeL6`Wy7h~|vw%HT!U2bgUKeGmZ-vGuT^y&gjn z65iN>n2GTNpzC&A_Q&f9+I(M~Znv|-;}~hn6n<)RS~@pSdCY&B`-;?6i439$JUQOInax5!P!apgofO&m^n4TJ`&rm5)wT8jn@tpZFAhn&Zza zm<-r83V4M2j>=(;9vSeFj$&xYFdI&bYJ0UUd=+Te?Y*wovVrC)t=$kx(b|ueTCwsa zq;`ePx)o+tY+wo@lEsleBZc$eH(-9@s1YwNyDgh{i$=71vpvhJw_@ha%m00>@5w{; zyY)X_{hRuYF{L;_@7)^?YIcK8_$KPsKXJBn?)R}%y%*$THv$Kwm^#&Z_z?{P&Fjsb zjVMM)J&H`tjqo!$7c_Psw9SruG*s(S5J;Cl8yd4>_W=AI*;Bh+`x zIqGw{g?SEXRXM%=EEIrKLgL*OO)APqv9n!8%cVg-Kn=rT^s4p^E*_-h>>3m@)~kqj zR95KtV&66yevVBhnSmO}OKNwNv5}Koy2>x7Lpj(>_-6&@v~;c?F2$XipCSOo;R4)^ zH=02_Bg1=%I&fgv!E3Ntnf?6{jCnX0rlR>_T;Vii@}5B-wSw%h`OQhU!X8qz7l;kF zPe#go=BVaRDvdqJwW_Es!$B#T%#oLFKi`_@cq7G{>AYR5*<2Bge#JklFFPFqD`86U z{hd&fSL-H@$wG0yT&Fjpe$&QN?D?c8AxHwq^aZU^L9BYxQ91lHDzff$rA`#qo^BmJ z>?7+mtqbu>a1hCp=`5_H^-Ef`lnF*~8Ss7fU$>hbc{#hk)A3_Yz5@S4LK!!(`R7bx zcckk~qJq@zg){U=kD(@${y0{S-%xkEhTL5~B%DEn(1V}<{gqG18QNjf51)Ue6Z)`X z=^zy@A(=nUO(FGi4>VPb1Bc%E075YzFkAdDDXs@I(ms)}$FC4oC+ZD6nH2Eu9L zC+#gdMO45qHlu+(_cqv-fJwi75m%QXbFU0+479>=6|HnV;zqoP_!{_@rZA5`X`0PU z6AbdD8~#Q|$Hi_=Py+CW1IwuD4Z&J8?(j=InSy9`*Ps|RvtlKPtEauXvJdtHswX^z zvM%A*^Obv77%lPdKJ*~Yd=IjqKJHV&*g{{0*d8e24_r*`R-bm3Bu%~DvD-l>!@QJ2 zF&sAKVpTmW(9sa z&x?4C(4amk{cCyr%krW5ZkGSy(pG|P$2&GI;q77e57?WmjtBA?qaYRL{*ey&+^6Ko!BuXUWp8YoXk0U23 zojdXhf&Z+|vH>J70m;o;HWe+}^FKY<0dSE@pzYGgd^GSLk$fkFN}cK()dH2A`W{}7 z>8JJ{^a7F@=1`X0rhXC7qA+3_p}ju}Jw4!1$VoVWct${qZ9oA3S(UP8$S|~mY3eaTxlwxegkI+gK_b2P*7eDVg=JA-(EDP_5>tIFRY8Wno#IXcOeaEiLl{Na=? zrL&&17RdEwDQaT7z@a`TS{jv4CBpyAkY_1F!_ z{^4FQB`Ua3ZKT_axO2CX8sb}w&0rD*=}g)jk&&${R5krnTopLFx?3_N$*fBK!&81W zw#dkVPeRd1+V-E@>a;iB^BezvEC3Q$jd~EZ2}2GKjESMzYp+x8G%g<2Ln>^Ia8H3l zx=j4ySBlw%ietM*)UCzbqSu9tiiCeNMe4|0DR!wQh;Ef}k4kNi%MC4e&NJ6ZUi(E# znyy-I7%q)A@Ee*l_9BDl-gRX-_O%Nf*HgXo^~@b^*6*{u+tcTEcg3srKfDfgGOwfQ zyXpT*8#Zv-R=Gj-J$~zMY_jYjtm5xfuf7yjSxTkIox(zkr*P-B30ML%ro%nLfawtg zbTX*gr>bMpah6v%#9n4)=Vg~oRVq^zIm%qWE~M4B?_ylnYMuAqN=KW3mJ6f3OD>An;Gx{to#= z8X}RwW}`X`LLOzlp(w|dd1ZX8RT06cf#t*ZC#o%Zr<3;lGWh@zmXt0jv`_^#&W0z} z3OGy(>8S4P2*wMR-MqA?Yrd-#CyblryN%hYIoTVF+9(~oVyOav`{I+|R-At@K$mt& zvrL|y?h6^-d6%RjkLaT|AM|=WB(>F35k1P!F^i~pTlUuW+$zAXY?O@^t1MW+1RLI)1}8V1*P5JcH5t58=(3RMI#hu`zuTEVW-oK0mfGrI>hSC zPH#l9dEYQSH{1OpNmyG7TI&6pd`#EnOYkkLC{X9^d`1{%upyZ9ko z2N%zBcJXIch}WvWmz(56Z$108#Ln51(|k?RVeCK4X1Q@HESUb*0@HupKRuY7-GPnb zji!uMOjIm3ewW4vV0tjau|O3najJoGUGOa~BMbpEfZGidP7=(Gt-z#X@^utPLL*d~ zqdx{9bE3t8SDEik@l~sft?^*K&9ymyU}07XKrZQ0m?=;$r2-#O7w`S0pq!&!O`xTL z`h-0UR!GCj!9e#*&fPmZzwk<=0gY!V*U|OSuLikroGrH|k`YVEwK(=9;!5*{BRD-7 zQZxFTeKUkd;~Uh7BvOqw$Ze#Kzkwk-{oHJ6CKt=d&-6l6k-rv@EuP*AfZ~T%G(JeM zLcP&DmN$1ixdBlk1#}yp8^LK#xa!sR3Nmf{#JQeOuxO zEmQ_3#oaB>GS1!htc2HaNRO#lS1DTVeMJk0t0HF=W9}$5O4>6+?ZxT$FGSw)MaHpT zQ5+m4pT+i`(e3Ws4mi=^lVgy83lV=$qLty=U7?HVZqA)2B+h6VYA#&$;>(>r?L5NI zetm_)c`BY#{YDo5evy(P_p#c<)Bd;p|ABZB=*bVAKCkFsdfNsA`JrEj7Jy>ujA?b| zkNn6g_Qp$n0ydO_0voHZw!37q+DYIPj_h{_(E36s1CThFyaY2A5r2LC4%iHVo4kgP zp+BAzK$w_f1>4)`o6e+<0@f5`d?8*GF z_2JRV?qMzIrTp)>Ckr;Qa=8(T%GnGGrFV5Wdj{+TfS3gx8dCx6!ai=l7DkNab-dXI zAw|s8m@cHd=^xNCGrHKL?u>jnPVI@m&2BlES5W`xQ*=66iSoSTo<{WMhOOg)!}+7P z>b?=dIH$mU&6Fe3){%^twl#WRiC1n;td0TyT*^($1!6zxJx#s08oU#cp~hK~g&P)a z?n?^n3oQ?3@Qp?&Z#J`B4bEVC?%xdUPmBdgBG}{GbOcU4P7f9g^kT7#XznOrU(cZ^ zmk_4Sy0iGBkn31gfrBqIQ|#QfI*oGo#bV7O_`;in>hn61HOPLIpt&}{`|M!T!^ZPv zcg04hq#KwIJ(Ie>+}G#Uvvq+&x{7?n9ajO9YDK>O)>`Bi12$uUM4E@6MOjm;|2lh| zT8PcFe6{rXa5Ca28PK_W1`lAIcXmgNORW0HJC0QRr&oVSj`+Lb9{d06N7uBC@^~$-qGq~diR>nT}EQUJ&;zA{fPY%ga%62mbxy{u;}2UW_wL_@cD4SDFm>2 zvqBXBH;Wh10cZ%rE|>_w>2A7__jMnKlSzh5oDwy|35*U+nIPV|AUiy)8RMk$DvtAy z$B=mxt6rd4*a!NrQDV3%-(Q{yI;_aHR>bM%^HvVl!^3CqJGp_dT4d?wP31;`rQj+J z_E6{JfC~W-gm|2_Q+H|ygWGFH;U{HZ=1HIO&m-~Fo9!5Ub+5tJK3!!o)oJsl1_cQb6 zj``bQf4R33{_zDVG32VAyYRRCKWb+t{^DyvDjpBSC{-x$Hh2ITo*G!0Q#_$?cTKtt z2d&j$064$^O(5+ndEhtOu+V#~$E+Z&3g|PSa@+4LtDADmaOZ*D?#fo6rtL0`=ZZ|I z*;>35;9D#{>5;l!&KtYAwbCBtx4bqrcdpJx|EzY`pW;$3aqQ0I@>!o=h+cepUpI-7YPAH1jTQiT0 zG(A>V*xL1Kq!Abx2Ck811>12wDFJA!%#tiSBXf^OwyszX34#~GakHHRNePPsw~)p3 zg+i-=*OQoj#bWQ2#uhu$j;tm(ekWlU1T!5H*CG?_39KVV?z`*f*6G639C7;!-BJ5R zO+_)Ys^Vq-g$JZy|FCB900yoFUcU_9ev$d8UAiSlVJ^3#@s2OI$S0(2p8MTJ@K0bx zzUbAr^*y<5xpGY|wSQE@z?D~hJlNKq6Q}DpEA9U(&9d7$tkz|?VsMnsfCQ@{g?ZnG zAYeCp24#8{rJJQWwUVRVq%*Cp$IFy|rv+RT*yjqiE8gsUSsq+i-g>6A!s}b$g_S>; zEw*3i?Qv$HUpX#W8QzH-fcxo57JgWBZEm~ zs(_+?bb629qHC5Vlyx)I@xKwY#wHbYX!BX7gE(@S5+LT_Ql9}O#gBCMKpIcdISrjk{NgJhXMA+ zy(VE7{p?;C$5)5*xZ<0o$Iyn8G^|&&S^buDEpH&K`pMT)a0cO|4~YfWPvAapK36?0 zD;I&B7a#ryp`8y?paNhPq=kE*=3gIj-#R9g&$p|FijI9i3PD~BBL!9CPtjd*eBN^j z?SQ1K>;WqHu~)n2^<{oJU@0nR`vRrE0}ftBij})OsU~fE$)xro@i@t14)vYHe_mRk zF1yAAY*6{Zzl?6HD?rJnCTsJ!Xjr4n-;KOR()lA3Z;LJPaJ9PkxCdFU+D#0%3_Z?049VLk zyi)#vJL6iVMNBG{vDXg_7j)^o1lInDU9>+gt*sxDEf6ZHm7C@|3AdS&oTom@F%CG6 zr<4=P@4swUtHJP%(U8>e`w(5Y2DuKb{YM`!z7Munb6y`Q)BY#WTvS!Dx_>)+xV+xB z8&mvR#T^7zeGx*(Y(ODPX&>a);Hl~{7AcRO{gPS8neY;WTc+FlmJj{3^WBotn$lox zXOk>)0AzrGzyykuW2OXnJOc4_F1;*-!pr_2i;i!zM(gn0fB;47H*_t_$ zcwc#q=SX5?_L`{GQ3O}z?j57n#1A9U~5t1-1*-7V;#=>dNsnomDkIxR{O?*a;l^1 zr!46RkV&8zj`Wl%4kY03dbpG~^tfE&ck~FSq5#JxJ|NS3K6Q)hR9j4ybb(Q^xJ5D9 z*Ty`SDi+-`KhLhGw5@b2XFZrv*%;TDjOM-}sdtf&+up7SCI@Rumyt=(o|9Drdn=EB zRLa;Psyuv>j|^T;>{&9LhcTO|>K!CB^7*XZCxRuH1$XXDqmUaY2ue4u{W$kK_O`_DMIs-GdxW8Yl8$znKg_h!TpD~lW0#)ul3*dw^I9E0tnWd`->kC zFY|58#lt&qo@jC?gE3M1*y#dFe;zB5*NlBm5&L@X<`Bndq1kTEU4>?ump=2(P9i%^ zoT&q+;cc(WGkKlsDQ$JngboTCyEhZ?-#AY4{_Qq)le>-Sk`9bt{_!zUfl%8!nB-YI zB1-{ljz@+(AOY}GM?uw+^7cLhmkcKLKr|a#?Wj!UTRHkX9~J=|p!)n4Srjh_Z}fD- zbRcQ)XAmE+(;TDwCya^#VRiPGY}t$>g#H3R&BcODfGgGsP<|LxOM{)?KJrc^1{#6= z0)D9C86Ox~5DCLf-MRn7_%t5z*T*u?w`c#*_u3mSwPzA?<{oau{>r1BrG2^a3kxUd z{@^=?7reE@>PxH?%z+b3QipWRaO(KDKK>BsgL_&`m^%R@*JIOm&=&I=+9yzv=C*SMalZ1D1-S!!riN@L{rp0N;jvM#UAa7`!Z^^nhwpnC{tI7*D_0MqQ;Z2g*E z0D19usFLj=4F+x*{Y!c|W^EsG;s7^5$iF$al$T`@|t|m}NQetOFd2ZwIHChc$L+!e}>W(H(rQoji?N;wClc z0nmU2gS7-E*{0HD(*Ed>-AQwa5Ws>XM;%AYzG6r^RIU3CLLC@=z;D6S@6i0}?yJra z6S@-y$!)gyXNI0k_a0+{4^v-TDXU#7xn?i29dXGzt|0OLuNUW~N}w>uT;wdpdDIzL z3ap~%oSQt{U5SFK@bm_+mDvi^`Gxr=@Lmy*9hTzb=I23zV8Fq|t2OkueMfAQ+1hlI zT~wJnWHmM9GH+@3znJ)0E%Bbucca09Cz|rYZ2w(n|1AC$aES{!9rt^1`m&(*WaP9y zxj;esFMN+$BXv?;ed7i9US#~e*fY&3SRfVwRIljme&2bT6YJ|bTjNkyy1!`m9h-Gb zYDHLJ{YZrh{vC0?E>mPeZ8VbW>6nDriyn%|*g}>KwkpjH{v$I$Kav(i=NooCiGK#hO|vL-if0D0kvx#;&~K~! zHuL*m;`I-qA9$v(_k&H~X>?x~jdfl@-M}&uEj&!m&9b_TC{IHhZ`iZS#cdjn1Jy&!@*fv|^P^m{ z%Zd-Bu?vBw6<>ngzc&9~3iX_CNrB@;qG#NjF_#c*7clg(=AeY>=u2>x_tD4DlZ$f8 zPVf9Wov5*IPZPM#o~t=~Jt=GbXEf$d=Jfb)68rwQtQwwW0>`snXSaXEe9O}ac%zll zc2_h6yrn~krYJ(!pTEg@pr~y(^l|+4(Kd!R4x#o;b36LQ!^4A`Kio=v1{Q6@afgoan?Tq0jwq`JoOrP~?+GByLI5YgOb$+{YHjwsvXg~5< zZibcYvF#;Vd-yA+x zz3MDH*Fa&dOg{I1?Jj1u{fR8OtsU3F_t{|YLxJG-GZQ67NM@{fG5%yu zGC+`=ytYoocfv-#bCvb<}N_I(VuBMLV4RX7tu z7UeeTFXWY$00Y}xI<0op2%ziS222Io6!Na3b>ml2)P2w+4T@KK;`p65-vM|g>sHqV zc6YN?sy$3QO_`6@D3H5HLYivRwyiDpsk(A}_N}yrJK5Z^MDHG-8~MM8lGvLV?G`%M z?#(oQVZ813|Fsud?*~t8){)QrjKV#c01}jB^kq#ipIF6DW8m;`{b)=`9vJ&2M@`%^WP$$X>zY5@<*(xlTU0fACNIvYcYgNbsxAPAM?vw z46s8#6JE^j@0m(;n(`0%jvFTLHPD|n(9R2vsQ1UE4{$ZvW^_^+a5WLeuO^#4_|X^V zm#~QSS;{qTlJHDs;Ccu0cf>9%&Th#x{cP3Nvs2!K$eRZEsC^JJ>q=N(y?N?$CGkCE z!vAQkUps7cXlh26l>nrFJKXNJF^BR^C1Nvr{-d>K7a-e;e$8INKRjzk5a=PTL~L+` zvJlN>pLf$i^-OqeiuSEO4mkniSbL^n4k*WtJO3Nz=DQnDz(!UNDGVPoJh&(RFkv&? zIf4ILl^|4jvUL33lN zyC`orv?fSMG$7x%;}X6uo4iYT6$-(>_+0A!W~Elx@Cq`P{%`*AE4?aUT_9)j7r~qTQ);e36QR;+&opNRDe0TRuJhjs0wKxUsGn%fv zYU;36Y7#l3P;x5?-kT8F0XXeNi5%aRWHg>5UYVRVzzx?Et>&EsvoMKucId*CN;`?E z{NX&sZM&1{q5sG+Twvz~(o7ku=iC2&9}iQ(>^QM%XiPwfzz!zO*t@I|?yKAnw49Bs zl6g`eWB1}IK&&Pfaz!kgpXw+|?3%$3v@A!Su7e;AwnQ=!k=7mdsx0j?{sS=ogdu&Z zcojYil2))_rqO|@_PeMGgSy|BW`D~@WAbUjP>Ei%xTsCJhNV=HND%r;*s5py3J*&F#oYh?#qzre-r55@AW{isga7=6s{n+zO`M z|Kn!BS|s{XO%J>r3Q*!h?iFI$;+DIJMJ9Yt^!v6%$yzeUY8^*2g zuMhJ!JfD1KZ`Syk1RHl(MB_VDbU7CoE6OfsZhyCg%JszzsvJ z8a70#9$|x(+9w;q9lkwJ@ElL#{R9bdN;NbM(BJ$2>kan!fbYS7e9+^Pj{jp9QzEL+ zpy_xuKw&k|lFEij3X>dr=vMY|s;GcZ)wMa2LZlJ&0k$Cm63(ZTOHy&eLXSPupv8d# zzh975$Z*&$(;{p3KDSfAL5JmYMaripvJ#QV`!B_eg*lgvrA%Pd5MSqnAb+7=#puZ9 z+Z$tcZV5`$nJ|jG?2Y`d{vTOy8V+UuzmKDlvCoi@Z5B%+Yj$Sr6jA98w>3r*qHJMo zV-I6cmaIjz$TIe17h_9QvJ^sN8~Zjh2LI-J9KYlE-2INuSzi{-P8GS6eY2!o&$oKddw%YqSlGg zNS-iE^LQ*AGM6ZIkE2(0kKuJq0_NCFa+*Hlu`xi%2S|{QIjT=Pj zpfdBLm#@3$^@R-__62zqJ0@ani-Voy-6_NSQC}Nw`858D>-=_3Jj>i^OxIWB&+!)v zb`hCi*;sN>_lthoqNj8CJ~wIa&ecx~+%-1?8dS>Oh9$07DR4*RqU5LSXr@m{{ec6lBQRX1>=M ztW-IaxN=~T%aq1>tjrCgO;M=$7rkp%3f9g}RX2)wMq>OaJB26qC}_Iz>8yFZqv^bP zJ?4?RfZehpe6ci`73*<^Vc&uXu*GZ>m>6&A-2q@BZw?5TC+@BQZY!q(t`8{=%S!7C z@$|WG3`#}W1SC~gHMtx!4VkJNR_Lmf^933%LB4;jxv14K_A|8Oehk|o)d4Y_pd&7? z^l0drirWl1dMkOTEZHosUA=i=-#jz9Sc5|No#o^hvXw}8Okg?Ogiza-fBtotcP%{i z9VOE}*1S+GYAMDYR8o9ia|Q65xpjo!F_i_snVKGia?6dhD(hsg5ErwX(p!r;Ao;GKDL+Ft+Z z6?Z#wWs6zQtCke59&sU$%_-!qBbKh2? zgpc^}7=wZqA*adZ#Vbha!o=Z|UAWoA_2$}{k)Mu?xu^uY=rj*xWHCwiU|eheW%Pqz z91+$;fg|UfHY@+nLC^Vj6v_rWW-bf#6)c|MFJxS9x%cf&|BKzA>!BZKwAH&)zxGz- zv?ojTPQ< z+<)4xi7^=~(rtC`9-s@7@5!8EVuau{RD>zgS_{gB6mzf{K^stfpj3yuf(?BVkU&#z zHPe|eZ8zZ3D8UEf)OvUZO=N*e1JgVN6YcZ!3>R)BaqR{|1Ga^QamH1;mt@D7PJR4w zTXw*5vp|IBS{#cP_|9vCH?ihIEHC@?#-Xz)mX1P)@0CtYfJFCfwtE?@t}$jzlbAW6 z5lYzs7eJ!?A-!!Rxr{jZzH{IBFVCyVP9)@0#KO`dP_Ue|Fud&%q6ng$o>)Lte~&}E zZfR||EFy`$mALT!KM2519MI|B6~`G6Q{tX2puF>@=V;nAcCcKelsOD*vVl zgsK$D(Rnpjm2b<3yLzT-h&zJZg>Nh7p)fW-%3|`{Cq}FosHRSptO@<>n@vSCygsUL zXnD5}@MPNRHnAe-*GC+!I(Z`YyOM)f-!+_Fc6G!*wN(EVvmMZSn+qMgA#H1x)~tL$ z4jM~pu6F2%vd;<%j(d7=W%_s1h4=`V$&~gKe_f$jbuY|x?fd?d@p}`S%>`q8N&kJ|o7UrK%Ib8+fJAYpd&lwl7c6{`nDRB7T=UOScOl|XAtQWAW+Mc;GQTG^a zqIv=Ddwi&`pA>zSUIG;BzL=;c>%NP;!{%b-%hLgOmL?Q1RL*|}{9slE%VSR`b17W` zC1BQ8cGI4k@}tBZggKYbw-Ug2O&BOg(lNz@!fksJ&HAUW5cz)c+zpVj6r3-1>^rP| z-Oo(5^tVtD3*1#b0G-Oo2qFo2r3qmlxWN}Vw#~d$Cyw11FC z5qz%0gx$Vu>eQ!-Y_uM(_+I!VJ7+)yORxOoL<;r_i*{b&T_UZ*0x9#@A-AkoyHJO5E%BI=dYSSwWVRRqqu+& zg5NBQlV^l@9EIoUL<%dB+4I5}ou_awJuYfV-V!=Ih1CCi!GBhY3X8vy{8Bvn!k(7j z_pDZ08V&b#B}`R*J|}m0WjRkJv*q)K?VE#G*6<#wa=siqzHL{|N%He7gK|l5*>T*eHl$>cIn^replN{DKh`9sBj_;PnnQ&8z(-S@qLGNJMAE*JZJY&(?P8o3@8TN%E~W5hfxz?<(u} zpM@R#YzVg;DY!HI)yKG%l)Uu$Z!8O^W7&G2;fnu#o^PxRxeh*mlSkHn zxP}VgtqYQ9UzQ`EkIZ4HtzQA5F+VL46cqCibG^usu^e7O$;}rUhj+txawzTdBrg<29uq)^hm#xRv={N(}t8-Om^bi+*$T87j zu|j5{RJKy@fxQw$Vq|ZaNR~ARqGNxM;dsKhso2r7bj{sE-(=BT=sGR)>}bjiG?(&h zMiS*@TM9(l#`;5r^f*|=lOZ2Su=hL2n47-3(*jOH^XXCsy&Jh(W+mrg?5`8bi4C-Y z2bsXj6F#)FX)DUwXFdBSiQJ^l%q`Awyp&kv$+%DWLGuh>IKh6r)9JLx%xopw1pX3u zWU}kb#GpH_W)gpO63=h9opR>*H6-oNk#$P!5=*Su?+LAtXEnjnE9p9SwbxosCcn4d z^3%nP{uC9T8Yoav!9AyDyls(@P=B=8uP(@rn9Sa3bfO-Ftplqh&_AD!6CA>>6#QJ{ zBdu+hV^5Zg%sca%rq!#=F5)B-;ZCub^G zZ&>94e>EgYrIN4Z2sH|S1QM8PQ$Nm z=lUe@M}?!zAIq|eyF4c8F+SJ^%rYaH2bgzf_9}wDkuj^2%So^Fm^5D86brCxxPMXU zdXK=36}oDtGftQ~I)LU6)z#_Nue!=G@q@QJj^7+-kNGMm=#TlLAQjC`PC=O5);xEX zaVfvOA&^uy^GwC_g54({$71cF>(})7lY}1uW+z4gyTc2Gs)l|lvBB--6v#*T*$$%kA<_t_wlN0{F_+I zG*+ut=ON8I85hUJD!jZMdV^}Mf26!b{2l^fCS>|xSq{oluGyZ%;RZg$D+?{Z&s>T3 z*R!0$gNO2mxA|IskJ}$Nh3u`Y_~e07FOVfPq$g69gR+Ar0)SOQPH|ev6B$?SuN=!^ z!{En@VUx>+;rrG~VGp8)57RL%^LyW&cLW+XLQ(y+U^k`+v=@NoLC zo-KOz*?#nY--Hu=NYvm1kkzbm)ZocIZ8Ye^S#iF+Ghfeqeo`%^Y)0e?R-%A3nGWeB z?(sEnt(5g=p&Vru+yzsH91!b!7$Yc@z?9|Ou(kWeY17Hq{|^T8A~3dg}*#7Lud7h2ZX@P(Thy*2o(3Ia<1IAF!iFo-cA$`#DYJz#Uz=`SXzo!M9wgfvA zC9RUXxC!p^If}HHf)#QHqDwJ}BpWKZ15I5gI%_MXwF6uHhIJayPg9{!^)Iot(mQV6 zCf~GGPY{U7_`|7Egw!(Sh_onfbWu;n>pQKOeY#IYr10^p!O0?S2k^hd)4aB^PLzJ@ zH)5Uciu{4jP77z445C<@FVMJH*@Fbby=F~s2iDx07ziR17GT{)o>rF%(Op|(=S$}^=^n`A+I@fW0Kn&Tp?*J5wv+RPvNMW;!hwS zW*QBQ5wy)NmDcw%l^^g*i8y7ZT_);EpFVP4krFJn45_>IokWpco+{y zQ<$|j8N=_)=q6N4+kbBpOCVOK<3hQ8+`xB1$TiM#%TUjAa0Q%ksjGY+*XZE`*rnY_}J z64eevhn*(6Bu?SH-G&b@TA>=ly7=$Wymx~L{3vpLam`U!?ZzhacHie2zTD7CbF5ek z_3x?h=fp?+_r$*j6u*`OX}J>h@qzD%)*3)eg^}@UytV-~6(5c2!ea55aN-q^d!Y#14;Qk1Boz6oiiVf$EP>2;jBDh<58YK#ii~{f&_}^2rvyEAxMUchdx2}lfO)dt}VYG#^rysbZde=BrEaoYQ&B! z%e?@BAAi>882yQM?vejV;oI)_35Vi`Z0m=oujxB>N9a4c4Ogp?kNW6VqbkEE(2eo- zkp#q!bO5#UY$BnO#26m`;2y-k&YJ9CYA{Utbef79yr)i;I@Wm;($!0ti*xRb%}~qBPf2d(%^GBzC1Fiqhd*zXP=!VQdlgi<{o{nR;GrTMdK)!%`JjFKg0( zl#AbL_g_}RBCr4*@C&4b-~Z1iZw%49+F@*@pcS?Dl~w;r!*<{0Mh&_r7vF1XM zXW3TReG$8#4UAq8J{c@PHUPC?J$w^jB_+x{4GNv|Jk+HEeec1e`@tDZFT#tt^u%rC zHWNvDbLO0ek%P#NG0%#o(n)wMZHxzFGWMK-_U90DV)V~(Gs}{@h`+WUyTy#^6V8ah z?CydRehRS(#T(xK4yaAzDd>&5zmE}Y-afT&k`BEGYN&1rLQCX4SVw8p^GYXI1^jrc z^};+U+fSCGSz$9xVCbF1^t?VacM)~#did>`I$HQeD0Kw zFt6u{K2rvt)5GD$R)DfbY4uX)!9Mwaq)kOqQ{fKpu0C8a-Tu4E3tVE z+y(vcHxUg!`@tFqFG$nFP%s7f>RNDP=1$%lcGE+u*qv z2A=*#_k!R%U>z1=UtMM^uqk#DZkF#lhDA`9V8FeWd92ozINgi-cRoJm<1^=vl(0?< zfS#9$JjbI~WX*7^x#eZd2g1(}Ix;k8lF$}=hdUsiL6s?b?N`5Kg6M+SHb+#{rFpd~ z<6Ys9rRPS#{?2F>jW*hE`Bm}+>ghSBqC6)EnPA9+zsM7NwERLOm`0)tuV(a!Ljtxc$Yofxkb`IWUdH}{c+dT7FWZe=;1r5s%jEDFIZaa9j3LJeCtmD zR);*a-q^~iG7>W>Y4`VcrJ1k?A46LQjqwbW4HG1Fq(5`tjU`Y@Ik2v`yFrPoIixTo z9MrZbb38rrS6q-TT!Fr5QapF`zh%Ddj@U>@v5Z6^v!0F=M4lk1WO&U=;l?J|c5YlN zDngKB68J9+36jS*+yE&v9hpWF9wwJLYx1;b_Y6EJnuac-EP>e;oZ{e4FyrqAQmr50 zm9rN9GR^^VYd;eb9G<$o*~KmIzifRSTkGIS(chGt&W5${$$2Ys4OHjeI!n& zAj=#bCYx*OT$R!SNU+(V`vk@Pr{HMc*s>tDL(TdQ{Ih@ro{qS0xEgA z%K`}5f6ym-=B`O8MJU=oZ|lYHZjsbF1_hg1##!}Nd5p(khy66e z5F0;&w!pt_z!Pi7V7F)F`3SF`%5nA8bTPVi&rPOVAh$#5_!*pUO>FDS_!G{$V$12Z z$9wrlNi8WKdsaXuw8@$#=!cCDJD*@L+$dWYy&Je~M<%n2Rn9#8dDy)E{_V;y!}%8Zi>2GyT-&nQZ1aS+(OfN>Ja)8@| zNK&0F&n0!2)At~{a^v!*a!LBcq;RI0>4Aq}Lq1(T5Wl$5IbdXZ^~2x&>4~UaJM)9- zP@lTu0lH`wg+&lb=z-4DBU^N{=B&nz^BqtH3X#8;88gGfmLFxgn30KMJkW(CL|_7V++hmIR={_tUeUDJoa zv_S~WI)0O`_J(BfM!_|6!5lLf~zEkLIrWz?GEb{t%F?M+K;_wQ%jcsbv#2T&HeVL@%6Xazz>~~Qx zb$wsfwx}O@yp=k+ikn1RoPBj<66$nwllgQB_$bYwHD*@-#*QsN$B%b>hhnmzxdiuB zrrYoDeS4EG?>Xt{AvQ!?K6UEoJxLQVE9CUM5+6mWlD!=i&J}8nKN`n{p>IbaAq{15 z96`sO!xZP^Pbmkst{;)_;Du{^D`Y^#jq*zE)wb+|3VcUrW5J9+_ImcWB<9Z z5(5CvIIy0V0_8Mw&8uQ?Ok-H$6G3Q((Pyho$NenV&p34|donBzPy<%L+tYSP}6W_=kUNuXyQe%jQ}zU|*J;%X7Hkx4xV)yuRQj zJ3Ozsm-7~ae<7z?5=Kou^^LTyV+Yl#Xd*s?DU7piK_Mx$Irt>RmwaljjptJT|H}f% zl!qwqO2s=t_P+{6>+Bk_;oH3U5k{(`hNmxV3oc`{^+7rD2R*TTp~g;v>R7+Z<1?sh$j5W_3t&hed zVNrtzMbTVu2CK~Eno3Y=)I(4;D-boJ=OP+-{&pfi2KLhYk|{?ESZ_yLvwE9=${fFVfi-_>J zxAGS^?Yb-AzDGx{mK)eq0ODa=CVA{3|Z@S5QA-j!!QHK%_SW;w#^;ykQ$;`!s4@^^Zb< zwqbY01q@9gruxAPZVFW4tI+q~Izo_CK;n>;o>ybLGUM%sJ8~T^50wi7L@ z^4j&*Pd2*vDOuI-$whe`xdEp`R^J(gw1Y3}jWx9Hy}S2zC{e1wW+6U^3R3~yrM_o> z^#x2!(6aTXr0BZhq56iAVN<$|8dvszXE{|AY2N@jQbUL=s2sS`$Joeu8B+-=kT}qz z?*iXK?7gveNbFwclfLjX=#%lBw*D>=+cV~h1x=sRJ}u;kQL4?xh>yw# zCkpOPC+LKV%*-_qYi@VN!J>sz9~PH`4JNs5v3GrOu%ZIQ)L4(lH z*Hi;;omyaTY8c5?<@C|A7nf`Di^E4(-SacD(kTlCRGqE+j%ayFc?LSR1GzXuICeRgI9()}q6wCt@#5~fyaDEbg zil8d5#8>D*O>qm6FIux`O?NNc+Y(f*5MJ!RHKG@IR>?drw_fgO>B4}rt&-9fs=Yna ziJqk3Cf2d_@|1@eR#q>}?F5Pe=_`4XPca#7{@98)RW39@17GFJPf=U+4Pp2v$XAPR zsW8;IJY$3n1R$cEbS=;jepHuXbYb9H&slwA$t(_Izt|>E+4#2hboBo07a-#v{-0m0 z$^h;ey)&DhF)7I~!%O(yai9VUv<=CAX5hsH-&Z&uFC}_ftURrL_jc|gt=E2Og-b}C zQr>PPxcU)kPyBs+&~_5`UMJUMz}gveP14G)M9eO`fzoLa$2T*_QWc-k)Vu?>sW$Pt z)jzmhIH;%g=NEd6N$J%s2RT?`SzE9)E!W8@=x2b@=dJhw`L+-H=L~D*H6hqE^M3lwj+vk0*_@tj!RoumV`=VdwNX#K7LN{Q+R!=Nu(D+%q>YKFEj63>$kkvt z!wBhp@We%qn3pu2)6kUNEVDltawT`RGm~;d-xL`UG2TmUKuRxEMIWQTilW*wn(kNP z-M2kj1%JGCJ^Aw28#l;H&+z5l)baXXGRDpq^#Jz-%apUi!@)1h7h?ioea3d5y&VQ% zMlzeGo%S$^8)RK`yLq&g8c|-PrjPSNCCkdhh++U5i;r|3rt?LkH>X?)QLm$P&w#Qx zi&I-MtNtS_Ig#W(AC=gfN(iOCw2WP--_Q2}M$_bS;AUwUrqMso8`?`_&gQ*l%_0?K zgN^>YAjERN0(XM|KN^zp29)e^Uw&0!Ki^T9SN&-r*z(RwabYvNs&7mxhk$w)V5G_) zvXdN}oVY%@&Dz$LwI$p@nLcO@>Z7X=VZox|EB5#P)5p#5x>LDP|oe_6Dlk=cyb=ih`CC6TnDwnA$pYDgAhr={#)`kP(VuR zwT4z)B$nIASOfcn{&>jqEb=KWTUWV#1^MH4(TS5<)_Sy* z>M-?-WkmhvOXg6p($v2V$aGfz`oOE45vib7rQEd~e9lCVs9s>zrHZ%zKdN6yUjP*@ z9gP2H(CGpIIA%HK>&%xNScTv2$liSnVf_Ugf8$uBQWE=`75`?Sx{#0s!eD4@qlUUO zjg#roXY{kxiQwC_54#JB%HallQ4@by@JswhfD?0z$kBNQSp^w}^_&W&G0~8v=lzix z4)uS)fB!46A^(qW*1j`jVs26OViSl1CHS4zC&2RU8~a>*d*th;S3znKt}De@xq=;w zd{xYd^t80^ZgQk=S-fVLvwr5VDN){u8Rgj2D*H1rUZK8jE>WQ9&jcqMF1q zyS!lGh#p8(jZ=$R-oZ5@6}m{LBgK3a5#c3IIrN2AIrfVJIqb57BcX`QIT4}v z&gR&1=tP{Uw~dDA)B6d}`5hnK?i*5@u`juxLvI4>%&hRzkrch1?hw9Dk$GurQeqZyMUG_Xq6voW%m`toyzyZ+J)hz%y`Y%Eil` zkB2ERX0bSU=cFhKK?6o{a8n@@6roOS)Y0vDkz2>E7hf63o%>#pj5Ahp3|W3@B>8V4 zv1p-yGNN_pK*R-3Viz+J?7tBO?1hh!pC|I}T%b1Md&Zw#qzDRABIVf#_5lqQ6Z!ds z#Trs)tJm)VAYQwrq1XIuAbbkpBD?liliMYx9XglTNeoZ@-)PEqCVfc9C@2Io!YaU? z5GuBP1@sS&R}btWX*JRt%Dk1~YQ+5Bt+tW9mhC2-2hh_uX7c%~fT{RGq%1$9Z5E^= zu#ns>bwL+<9^#}R5G)~MfJ2BE=*s43oc;n%BR+RLJK8+GS8>jgL2#m|0Kv4o0yu`t z_)SeTKQJ)u=*r{qwf2$Q{Kto5?VHCu5Bip6*;kK@B+)7*F;&Qb1jD-&%6aND77ZW^ zxr0ZLOdk+^`%wu7u)7=ifsfrw&MFOFFIqTn%F^*rsZOr~`h(_Jtm$X)c#~7lZv&z^l(F!CqFGJoX_ekXJ_ zmhjEVuB{TNHBQ#v=cb+~zL++OdyxEE$8LR{M|d!w3wL?8#3;3=*ZoHASaY_U$VyWOr(C9==~jkOPf53M=n ztgd`}>se0XS9oA9S6El@bw{Ul>D3%c?OCx(ZJ13c(Ew5Sbvoo%5LUSEcyJ?z*Wx_V z0Jst@`*pgN-K13yiINaQ@sHEFc&xPk$n@Qdq5N%Z#yOsd)UAP#{$|SiFAwm^tL081 zU2x4VRX?j<1Vvcr{VKa!$;B6EqdXmUtW*WcxOR}clmb@Z2mL#oB=$#oXe0`+Jg3cU(H$j=H-@Nzl*HV^#F1hCOE$( zY3uoRJdCpQPJ=R4;@=Qd3W}k7K`{l+Z(WglLC_;y3}%6;foY-Vxf&0nZAk=ERRRzo z*vXn!7qB{Z&-KS2T>Xu#jvGnsujaiGnq=qS9b~}5p{vT+i_0UBspl4sm4X9=k08@8fZPZ#|-ex$9nV|}0 zIacXnJ=hQ7SoIf#(>C|G56;+p(TmhpU$W%x@O!xY9G7|*jLxzpL~!ZsewH|)Eh3Yu z4A0iUb34t5R=bHa)y+u}+43~eN3&7eYL7aYNAQ8_N>99cFLq+4{YOMSyoANih&V?< zBDp);ZNjH;TtwZ7tl*KWB2F*lADWfEoPirs`_AlA_NRR0@xAKw{M@uBrd?zi*)vK{ z$A|OC60*yBPJYiS_4(hA!ZmlA$&z7Gd&OF93Tithb6xD031#FLBpLCe$qow|?xM4T zTt`%1vH^=V?Y7+NHvwCxg-3U?K<{ZgFS8uq6em)|rhLY$o}(oWnZx8hY<6xt>=-0O z5S0sG9}W5BZNqo;?ComCGIYQ>?k!td;a;*EyA2hDOFoC$0ZD}Ks`&D-RUG$WdQ9~a zp}~2D;F$Ay6U84m-z!b*L=ZW8N&i%PPY!+07wK`^t_dRI_R{OO`22Rb?oi*;9Gu+) z_1j@yzu3CgPA$I~_XI^%4P^=K3*s5qM7D;?8Ew*XRh~WA4zEDiKHf$CUGptRPvdt0 z-bwn8QPI*^SHbE5|LmE;GvhY_@0cA7UH~g3BP$%WI_~J%EHJ>Bw{~-9Xr-8Grin>t z6_5DUp5&s2BGnroH<&W)=;Div=B8@) zrVdLn6x$2k_J7Aff|C$|8!pzGOVP<7h9Gg#&Zi>6PX+O2rD^Tz>Bg#354o0qJu<2H z4U&b1rilb6h|aUW5Hh=ShR5);NQX=kXv8>TdkJ;>}SfB>4MIq@Vw_(xwMAwhmF?5$HLn4$;ua|fQqIwWMPM2JS}Kdch|Q; zl=34qpQm_skL9~i%^H)HrSbyK4+i@aLgez9ty(F_@=)T&zq$YCulTah(9IF7w9*qR z8}=mmw;*ee*hD}@tkX!&dcoS&p1EG^?~+Q5%?_E0#uv1!LVgR2czc5jwo+KhQsj%_ zEJ5eRB^z(@X66kNNze6ck1bvg95LPk!3}hKY&>AYW~l=mXlRgSJ?g~ z?ThrjxnF!)7+czKy+X7nT5ByO%PikLL_jU;Z|mB*sMwf4CoX&6x&Pky>?B0CdoblO z<*)+Z8tejbmOL10N#Wqx?u4SZk#CC{WD$)1wgPD)a~7o>!I%?MfuI0We$?YFz-t42 z)&;g#hgVfQre6OVR~5$~2(o%MVfK@PLVff1<>>ijFq3_7KHw;2BmWpfTLFZBb3689 zU3uo|d;3NSlLm2py+73TUS&$8KTz*z(7;a9=ne|_5KgM+Y_lTTSe7^@@Q6FHX~S|k zvhkey#_d8)3oT|i;oH$aoVygklDe(KIOsf&c)Z9BjDzsTYcYLJ4HJ|1M(%<+UN1!} zNG9?}U?1Yct&jX!C!eDn30I)r_SEINN8{Kd$Z(u>P%mbn5WTFhx7KY-1s`@n7Zr zNSXs1^ll3Nb-=x=U7rew_ggwYYQ{fR`NTazP`@&UQPh#?xSp$Bmfsxmcfa#p0FgIJ z60pd7B>JB~dFGhNSr5z`rf2iB5-we#9f2DF6J`wuC;&k~zQ1EZICU9zD&lZz=BzaA z{Kcg=z`P78@`i1pW!Q=ppScMD#t1kD%VV~HBCf2TLQ+xbg7bvy!y&{}PuT&WSLg z46gy^JotLugfE!cEXp6iK^q6*9VS?~+$tdVG~z z1@v@9D|O}B+u(sq;_RPjn4v-MdytYo=Y|}#+k8)o(c^xAa?t*-ARt=+QJ)`iegF(k zjLd$$`80Vb&Q&bu^^x86UPph0)}&);gL}>G{#fr~TKU4H*}GsHSSLlt1t;8= zXd>BcV>t<3dgm$*J!#;QG^@%lpUE$bey*rvIc?ROdZzx0*rN_!#T9nxR)NlN3KIr# zylykUFBQrqd%M4Pj-4iIJn8<^jl;V~Q^2YD!|=1`@h$$PVK2^XMEpJKe$q!>GwYIQ z>3;?wRu@Ahu;^XYyT)jdrCSeK%m#p*jh~RiE-|iOw5z|V@b8_RKorCsp#PJ(h}oIh zp81&hb{*89$O&h+d6qHSm*)p!0cUUlB-gmQWQ=-REV2ApU6(^r-2y3&kp(e^=-DvZ zw#g^l)aPKsv3J8S;H;SD6JQSqt(>*>LORHRNA8|lQ7-D8QReRfNWWWBjob$B+kv9% zu1ZnKOMIh{K=4BN(cbqg+8=->kGGIa4}QeWJHH$)3^7rd8ee#;W+`4V)1UFn#H8b4 z$>dkvIl|<_T4#KTbKfVqZmwpi&~BPHq18)LsGq!!f?ireeX1CgqUNv&+a|C4Rpv z*TlwX+*^s}FKak1J`_&8!eaB~3z=}&CN1`d9$RNd3SBP!vR*m z`iy{$m@TqLw6c2k_mNjpE4nWa<5Ms+72d7(&bA)qq$#tDSF2VN`z2J+_w{@A^LUX! zo4w-P{T0fU$JIlz@;j~@+p%m;Rgf>5b&_asXLT<+(Q<^@f(DA4Xe0Shq1|-T&-Zh) zp$fI!cHh@G_eleM(`gw^&BN|dPt6j;c7kcxX$6#21akI#&3D?G)aF8vX^`LmULOAp>I9YX`*zQGxOk{5fKQ2 z2NHnLS0WmO#4~FSzS-W-*(p>u5c~EQCtaJO@%#df{!iK<4yGxl5ms?2hIcwSavRIc z{4oP-4Kt2yV99|&?PL>CLAekdZ|FG#sO$f!VyXluCgEGNa;|Yhsx7~!k! zFzZyfJ@b!=>^1s2yI=Q(^_3n;UM_#5(ezkD7pnyapv(4ntYtPY)2Z8042qJ4X<%kM z>VTxO+zHH1Ckz~BYF@&SJ~CcFaR3Sz_9-3E8)S*jva!M&%Zb;FMe$-tx6}*^O^*-+ z@vc|B#1uHqSH)e>7Km(n?tTPGUrNf3Y*0~D0j2LM-+4TIWrY}RGiyTZ5S6B}t?%}K zEWkdZw;m@c)Dw4K*6b@R%6N5uqVY&S-CGROo5)Y0r0EzU`R#B2QmP1es$#6>CO5t` zkX6q5XJ}I=WrR089MOqKOOF-GqU)j(=8C=Ug##OBt3&oy`WaOCl_-i%3ZfcB6ndwo zRlAwoaI{Cy7W_#OQ4-3DMb(~NauMDph)ng^FS%SD>E%}=%jY{(xt~IKR2qJGW1uwJ zXVinvbT(zlK0mKH%yxp|)6Ze$S4s~tJjKTkAzg2<2Aj4Y#MF&hSrj)4&GlqGAcLQ{Q zqww7T*EB24;)K^~a$YFu9qv}uH7TH6LWPAf4+rQ~6`zf}(=dT;CnLrE{>tFfb=0|t zlcR&m$V2MhY8k7`dyh&^d$-l40R%b@-3lq<=C40k{}nGLhwpsk?;mVw9cT~(Q^jg; z6C}irH{;FZ^iMfgz;uNKNlQ23bl!PR}3_>R-u0rk4psYgC4;>N8{SPt}M6E(ce$3ETl!&yglRV zJiK$`F{Z}ac1|d0DX8Re$Z=G*LH(3_W#ioAf8k?sBwdribGdCo`~_%SD|%deFy69D zW@pw1?>j`di zrTbc)1nsfH-_9x;u!yHPYv^xLS1T)>oAxow{ydu|hEOOVAq_#aL>8vFI;0##UnXJf z#%h|6nRYe_d3ErlPfSO^)9RfJalpYrnFE5P<#p~L{E(n=m4P}TkX6vR^2*HYyGqT% zWmA6F8WHoQ*dInb5Ka)!LYIYIpqmW8$nKOoWSEd58`5c~Cr21JSkRmXeY(;Uon$5! zQmyfmU~&P<`LoYY_Dvx5LXUa-%!gmSLE}${z?zdg(yGR1AxpFRr6iq&1WHcja8J|j zs9DkzEy_X$J6yfvYomX$d|M>9j)4t~MljjjbOOY9iHG2lKKve#LzM+|s5#$92Vgm3 z1@*rO6r_~pQ!WIH@;dLikfrM5-Ksbpr*o31ge%CPu+g!_KGo?m?7O-;3&PD|lVoVp zs7M8x_4d81)gS3O|MQPj)gFz+bEs1Wv4yk+sicyAdu8LEASt+JhMrU*UkJrqQs>(a zqXc3p$6@^IaY??$`Gr-XrWH5jfc;a>p*vKOpdSxZtoIFGRX2xq|5Pko5qT*yKccB; zI`Z*B4~mNEe9R~FSm3wGSdmVY(~j6-hWkS8-W@KpEV*}e1v5;8(`JQBqIt7-96S0$ z4Fzq(t-feFi?t3qZmLy~G^nGw`DB+X3R24RRaRFvKV{P{1#W%%s7f7`mpH`jKRzDO z*&jwPaaW?HV@1GGBo~p6>-id+Iy+Ika~a-BmLZK#245mRbRctv&4&Vav9yU~#SGH_ zb5}g_1dor19vc_$A0Hb7_YVI&vbjtD0zS)*QvctIG7tdB2NVP101V8^-zbF`T58nn z6SLWCT5HT@wS~08NI)ln&^Mi9ah&ZbD}<4 z9|e+tQRIA;cK5j%^aVX%06P!Jw(Ee5C!KFSh!&$2VXqlJCLBL7`HX`nm?2IUE7yCV z)?=a%jUw$gZ=H@O$wWCAa$&kR=2^{!;)SUrX4mX|kN(WSe{|T-u%ikCCp}&!MQY;X zTMJKUi!u;dcC#uQUz@;G=VN*b3ntcnZMk%MaY+Py>1O4t0iD&SSYlR~gi>v^5fFl6 z)W?ZpPX3U8`!v1@U&-&xGOCXj(o9yA(^u(w82HKTWvQsmT_fC*m=0iI8A@KiFRKps z1B`(IOsj-b;xU1~D9b&PVFAYfB`*{6{YjIc!9#!HdI;6ha z)_%WPfOhQ`0ilc zYfZO-)TK_E^^JTYZ|Xm5pKw#m^^R9wvBxMWnyCp zir>2~Wi}!b#rR#ZNhmj`n1`gj!%A+d_4`;jF5J4Z`@o~DB(+{TZMDKNm;6{WTaX(e zS5m9-zNYN~ncIoYgW1REq}nn)sr%a}hfFCnK|zkRj5uz-<27$6H9V3bx)PlmvH-Tl zKjj-nZK~w>PTC}`EW1}iqmi+?2JHhOm*iau+a*F83ChRbGh0SLM8`16%ff39s7tNH~y3T{}PG99sHaBJ1^t{Kf$O6_3j$tU;A*{^L;q&vXafekxZbfvV|r@ zuV05nEFjN$Q{Q6bOY|Em$pB%M2l#I$@n>qvj|0&{YfteyB0%RmrUx`)-+ASY^E z0gj&zh>EGL#tAMi=%5sY+htyj_C^1QVVOcndAmhVum!IxTjh-btA!xS?OQx>m-p;l9pVOCZ-uPTL$- zZ&On=`AcXTXi9dH-GSP}&ozbAbBZnQ8M?lWua(wrpr05dKQc6ZMc#aVKaND4#9-q$1Ex(IXHSJ zDMLU)o=#(7X#W*lcD(B2)~&pvQs>2FkJmx8llRhzbV5Y|rV$n5?^11E4Vkz;*qcwr z^Od{CJQOsUh4D4(x@3@UqD6x)}1_Dw3BYqSi-r-OruFahk z@;l{P`@^i7$uYI65-R?eOE|_>Vd7v_s|E9pMhgB{3wFA-lD#hp+vk*-P08(p&&KEC zBq$!LNJ&?&WCqtEtj-`z@!Xe+KPm576|}9)nWCh8!lH-iOtGL6AlvmBW~5Wz95hB0 zOX+&F7;OnA8Vk3gFSfvm&DwsP$x_JuB-1rrj6Z^uB;B#FH)||bLavcU>)rwT25D-n z-)c~tK2VB%JhJI1$r6E27Im|qu7^D&X!sh(!JtdS%plSMaEQq}OQI}Wz%;OYfqaoy z($H=0W427La6(`4XQdyfQ$SOh118IGqxe|OOU7S3b#(^w-R@PyIsYzZ9(}2$>uHNY z-2GZx(JCxpXPjEG%(7sA+2bg>&(|!ce-IDJoPsc zP87+RM;_8FDbha{#Jz`9T}XXHb^D;~UXg)vXX0E`Vf$lYqxT*o^u3gHcEdoQ^#}zA zp$#vNlW)#v1HfZu02$Xb+3MRaYXGpOaLzHQ+D;&}L#C|4a^Eq2@NchP({p?7D zlG}TIt;_@MJ5niy~4Dj5xu!BXOj?z+J;q!$vgZ@#y7LMB$)<3TJ^?~6e zpVmt!53u11KL_(gGqqW>oI5>cQ}%`sc@y*}KW|LWd= z&OYC-SjMP}vim!TogwqlC=_bv&efChJ~;Xp?WLH~@eZh_YRRplKPnHg6#5Q`Bt_`jIbBmM+J0G3E=Df%LGd_;=x2U|3t(-qq8pVWH z!0!SUXp#Y|ELQ^2fluiA>0W=^(fKh@>Jexb_?)_ZPFPt8x9G*A%qT}^MW^mvcveLQ zCjffJuACXf`CFVZcXH{eds7|j<0z|gZXIi_C<@9Ssf^oA8nUOqh5NMTTI z?0Obm@TBzXdTU1fh2m9(9X9sLS7ojocfJ;SP*0OS+PRoM@g}9eImjEguvRomtVQ2{ z7DTC+NV@H$hkJ!j=6n4q)4$>r>SwKfImB_Ss`wBdH-MP+c@0IG2fRMIvD7Pz(>x%T^5cek*!r=y60*oJIdak|e($=09#pu*+67PV?4Od6KFjszY)%*{ z(iGT}eW}E)@nrwjwuX0>$avXoWji+kJT65`>_}7pYXDmaeXzv9=HaT&`s8Aa>NBli zw(Yk%hg&EFYklmuLH||Z6}Bfq+OG4Cz^Hba5a*?o>Ya-lGv;X>P$XhD>C|aoXOv}^ zzLxPOb)N9-Ds!{^R;%hei`q)_!YcrodP8+rcz=9Zb&wGVD>X2#cp<6a3>E1 zhxnBVO1ZlCDMo3*$W5`)7F0vI_^<2u5XJ2)`_pP+8;fttf@sgOxs}$mc7uFK_rYt# zhw$J#+|c4xfu%yFllJlhoLV~ac$)c3_SRCXudkN=rQG_zxK}!_&0HKzQ!sqpe*6co zrr(A{)4TpjGo4}n<3y$ncp-}V_zHaL9oonpnwoIi`0?z*0l8a=M#%bFs5Uw|$XoOE z-Z0zWrO>#TP_CkTSBWRz&ikuUt8Z`GKJwHeZIGTp)}E-hP5JId*SM&vBYtf~@8aJ7 z3~+w&r1&b_rnhr#-SEP&gydMLxhR*Z)>tzW(%%@U*?_I&Ne{kB!QIOFtP z!$!B2!E0>c0cCNw`MchxQ6Gr)E%#V{h5Y!N&*=|837p#Pe!qW+n|9rF0(WgcQ89UX z;4$H~5SMG2-}U&N1G>Jsr6EAOh$9iT&qf5!2pC%~)T|WL-n@9Pi0?z#-4-L#Gfa(5 z!LWd}GnD7JAPEt3^tpa^vxvW-h3lv_-yB9+{Q@A_av| zDg_IO)*I@2oVK6M6t2roVq816x!>7`{#J1*j@{aO3(uRHCF=n~yo|Yv`V*L5+ZviH zt1$7~EoFdBr^6W@cBktm!@&}lQe5%zihhm1CO9XI(Yu-1?j&mlqZkS6i~D5-i(eR% zim&)7fB2OkcF4_KA+}?UbhW&NZt7OVO^jGG_{6%AW?%6fDm;CM3=Mw2#%(TTD>**R zu<>qH-|68ZK4h=~cCkMp|H=x})z=x|dUgfRL1s!YArs8dG4z3_;eGW;CYfg&i!4f@ z{OISq%T$*kyPd?-)?Bkg>qc_eP`lQ7BUs*)rK8>8^b?bY_-ij+SWBJ32A2jY-otAn zEt0X!xutcL8E*6h4t!ussSZ=GR zdIU)C1saZAZhCp^hN!a8CO6lFAW>NJ23dIVYyQ-iAL&~(9e@{X+5*W;u~L_8kCAsq z>;}GoJOuzVvd8joH#R>X_pQ&tHz;3}?FMoFRo$G!R)Jm1<RKu3ELZ)giltYk8G~3U}?KWM9#>G~5J5;9n2NK10%J;jY z>)7#5hYGJAexf9zV{#8htj7h5@^p7y4FfZ6vjWQ_a{}qd^Bk{#)fTYhLmaAusqEce zmredi(V0umv=_y@a3;xSNWGny`wI`hf+AG)<21M{a z4n5I*_;&9%Nd_e=V<0+fC@=ZliP?v56|bMcTtlS2cTj$`);DMjTva0G90T1&ehxd{ zJoLuFToNmld|pCCI)v!)DB3py7Gy1Z%lhXF+a4D?R4R&($Nle#?P;PN%A{>iKy{b^ z)3HATLD~)3jeYT+rEeosr&rz_E@f+>8W=p@#4UBEQ0S9ZBX_ff^U9#~SHmQIIFLTLw>IMW$}L?_?jw%we$ zy{*!=gKK<8c$Ss$oZ!*c6~enookxx_=Q+en7AC~zKrB85`}1wgYfE8;kMZ5V@~BXU zBcn64_@uNVaSj@$WtUJ|E|KNwJjU8vq}+*?^xyFu$;q^E=-#v@OmeYjVfdDHiml|7 z+>tG0-G13b7)fOk!oflA|#H z&LAS}^}DgNo7TTVc|6b>NELOeVzKBS_Mj=XPO{(Ywublq-{D~PJ6F{;SUWzdFwl)% zeoh})OCM!SJE-of)xQ-JxQ^LNaqZ1TBiZI2TuA1NDN*|f`3MNoMV{>PJLlbNfaTOZ z>-unw!)vy(g2fT2gzozbaIyr`OXugjtqiLL-YdgUwx1>p##Tg>({m(9!v_bB@=!-q ztSVolD}+;DKg@SF_j&R+>FwP8pE0+)Lz&X1 zn_d~3INyzx1WHt$8%sSZWNgxsfDu^ZE`QjU4yHYVh<8*PzECT0Z%?-FhW7D@MBGpj zf~fJ#E>}GUe61k8G%MOwT(ZT>xqhr`4mgfN7fPEqd^jPW-;pRK7RZ0y#;GxhP~#Ke zEYv`;hGjY9Q~XuUhn_|CGW)w^(cyi&VUyFVINL&vH~34ftwC3wj5N~+L^G&4<(X8M zoTd?Mt#2&RYN@kPN(kK`OFH3-0H_k#=a{LwOLsU5YvkA}6|)s>SukjP9L?Zs(^K?T zX7Hm->-Isf(qX9*SYxpCeewAqVs?}9q)adL-^8-4A4 zy+X5s8IjxbKyrV&&!4-Y_#TE0hQz9fTa|Up`uzr8!@+JB}y0orK zThTuFXDTXp&GCW>{>inU68t|w(htO_6iW7_K3Ftf_<#ufK>dmuM;%9Pak!EYMXmBM zz!nup_a^rJg`5jSIhJqcuXhxE4Pl4r8Wyy`%}M8uv#FZ4LUQ4PvoYQaaeN+&6j(#O zM7?rsD)jNwt#qc(O&c953_xmQ#~6EdJ&wFeGl2J%wK`^f{vw#?E8e(Ne!FnKWLd|8 zDT%o5kDyhQ{}bW+ud}gJPRpes`Nm16?Mw<#+C1b+`?&rHFpzD|5Ut4{hSkSmCbwup*-7o#+ zb$$a<1L^Np^G5&O5==EE+k&sruajW6+Bw+U2WVCNfhw$}R=Qr;}* zzSk!-Obk2;cTB>*X(^S6YAGaNs7=?q6>j>Is5)9K5_qpf=iMN;rMza~0i_K=v;XeRMz0*C^p zt0SaFq~uwdhhbD+i_Gx$YQ3vqfWXTD8-tL-lDR4qj zB7JobYDVl(w}zfVrs>{brh}SnG_6jeJfP63*a@)Tv2*NzJLIDul+q*UaK&}#sM`S@ z;lS@N-vcIc9zjES9->{CUa8Z#Xpl-sww_!>s|GJd5YUSXx64tS)f) zuxKeo^yz3*6KdR6z9})}qgErq__hLhRT`?UUh%{qE|Os{|Eb|j7(odiyfHGOkE+^! zI<|-I3C0c)%&abd?kLZD)XJ;4{Zu>JcyiUN4_x&XoWYt$P>zb?p5VgrI5F=J>ZqCn zAWgwRw;MXBYPs@HTv}=_bNd)!OW}Q)FiKXcA}M*?aqxu|&OKO6z6(){(PUyXZc(2& znW9DJl#lk_ug)|h-QDI_g?tXyymhdbo7ONVOVv8;E!Qc$w>yjL+mw>onEvQHmHj|` zHla_Zf^bjCK@v8{GWmNaDbtY%?;h&*SsMm9&PGDoIVz?uOUbX6zQ4UOpVucvTrxDr z3}{Sa8z`a&bqQK7ESE4gx7^Bl>YPFt1I)C0w06SMDg1$_^Rd2k=7{9AIB7BLS9Uw+ zplrnXby?@V6{18NkAYp{*0n}n{AyULbElMib3su4 z#%#zy(xcVZ;R-n`ad!E7U!D7HqujOvnSvc=vw|0yX}M4pJr^do-27RtdFTnU^%;*< za|nVx@&BPSxIWyApPYUZI-?aU|A;e+WLci8I5~cr@R9lV3eW%u$z_>tS5{|}WPS0r zUoB&VE=XCG041zh7a8~aUHh#jDL=-%WbJRZK;&8q##yPQ;vTn~S$%Sx5^w`0h=fT2 zBxbHGg$Jw23Dklz3IrZemMsPXE{r}oD;{0YfB+#n^ItHJ=f3h|A8wDx%h&my9I}eC z6agLJ9dm+lh+4}E`{9SK49<~Ldz^CE38~QEYR9bPImrMOm9$ZiMwTRajYWmkHhaW~ zTy?TPs`VT)1G*FjBFTW!eMjsKHMUG!_-hsoY-z-|E$4AQX5SO4SIaY$Qo768ts)Ub zmA{e;Noj8YGnuC1&vM&zx!v*1 z#jTd!J6Ul1aHd;^1!KiFf@yt*j6e zAPe&iv_efigC*dhzl*Vo?O-n>?w`bD- z5uEq|<6Ip0j~76wBPcZNaciFA(44tbym3Jn3XZo}{ixu(gPT0=4<*Vs8+9ux#D+aL z862O*A{&;Acz*`?3-MX;Y5Syk^6)nTRotU&6s=kjer?r)8Z*xko2UYEN*s5=a5d0? zKf?V`7ixyTlfgq3%GoM=K<1ztZtbb4s3lE~-=~(|btr32wtdxwG+OFP+=aFR z8i!RxWthXOkLPcJ>82alE%?}u$h{Sc5?AF`+K6=4L#7_4;{#@j!&ZhzYt6HdEFH}_ z6Np@F+yD*EtyF(TRj-E6~5_rU|c3(bIg% z@^z>ZXes+ptpBl-`l6G&u1_Ehcr{R;b~vYgs9h&Z(CW8c(`9PLrEmF}S^}GH!0V3l zCil1Hsud5?WPBY5FtN!%OL#F$1JtfuMaH(5MHaVDSyOIovpIM_F1wryoBSiN4e{c2 z2l0ol-Hba}CZ1gVsh{SZ2ZLS!<@<$JzmMCRQ$j}LhhsPzg+o7~ERq-OU(R+HS`ffP zc~m0m7kr`J(b9o5TaaMS<&w<&HzJs4Y?mlj)Ko#5(+6z%v;#Vs1NuoMCf5wlX z$&nGC1zGl%&<%_Qv(B;1r9j|u7>&BNZ}X#9TcOIa65PzOy&I?1jti_?@t!%1JU}^- z!+H3Yj-N~0n^*P6o6XF2^$A+X9Nvg+W-&~BI;OlrgV2xlc5h-8KTdnPEmoc9SGeb( z4brzLybV(KR*gH2GidF)YQknf=4|wKAz12czBgYN#LmTK)N@u9q3HhpqY4<-phZecE4 zc)8XcT_w+wo)5=|69)~3UdwMtpvfrXu`&CCjHq`uHZ| zlzq>oXV{zO64Raudg$c$176#M-_Bu$4P}phhsFlU!FmTd5Km=^9>iEJmg{d=i9X@V z-1~~(I3BA%seaxA(v^2wuU@!XV0HOcx-cpD_~!@RB=9lF98p~e;3)hvKK#UI?Okzx z&G}IFpT@vb1dKwmWm7@ZWe3WZYW~$M9sr)Qj3qC6O#YkfL`L`{w|8j0Pp80*C>iK1 zP8WC^!1=3xBLm7r=hSWw$Ql=Dm!F$7F$U(cJfHPGbb*W#QNBxRyhFHZS9-sPV9Ssk z^>+f(ich~zn4NZ|wk}wsNyd9GRc)vq%MQ!*@VLtLcN<~E)!t%*2u(F7RUWouNm#d%HGDM8 zm7iDT5}jiV|0lXWUdLDZ=~eWgR9_eWY>nM@5vjZNtdg?7{RdI*my=kTb`^77zq_P`h40JE&^@- z6KXsNtH`qrW}f-jmdo0xc=NXypV3yb(i#!k3+E+)1FxLc~HS~G035~f@%EC zO1>b5Wj&0$HJ|s$soAv_6afVKN9TjT0# zN$K*4UEyu>Ifm)qIx-H^TA$DZ>vleyD|I`)--ig-sF?%gv2kV1IP>lI4yu=GXFjwA zZV!7}-q(DuRs{7wG)2w!%TnFzeR1oRmCkim<65(DXzTqOB}eaJMN^q|{@n1@rFh2K zzLY5O*)*Am!d+GMq{pRCl((G0v-q-c_ z@&pFd>5UC=)CMb8HUDKo=GOqo?F=V5Wi@zjJu$y*-yPjy{m?FSGD+Yy+u$yjUyq2x zfX?V#P@j%ZL?))JR<6ot6Hg0QcXainY|p=Kbl#3;q!yo?i$y(W)elc2g_Z-ENCgeN zF8WbQ%ZA;=XSD0$Gqx@$lej|6;Dplbbmmd<_f)&yTsWe(S9opyB6ml^%gq+#vEEqU zR!e^8qmvFABl$U4?YS>rzO*CtZ$1h6q9X!fMjTi1cy+ zudn8Jm3;AnG)x#t*@q92@g-F)kv1l|h_S294G|6%J4iOG)8vH=GG;OWxwj?W6Tf@ggPdaa)9G zW_=DM%u3C0OFU;-RuIUK^$}UU(TkTkNQpKu_JrHf*NPp1sa^ObQdUkp&B8Vk^(JcE z?B>(k-tbq13y?zO19B2%+kL$51G_Q01C%miPV>2- zS)j2&+7tfz;Q`vw2xnthD*zEyZ#yLK+M?*9YboFwr`Lkw=V+W#elFqhr!R$J8f#<- zgpVM|Q$bobHa5#$)W~amwwht$z144(pD}{8WC%Z*#0o7x^HR~qZb0gC`~0X`srop{ zOt9I3kaT;H$8KZ08Fb*7$-0syy?) z_|ETHvt$E*&8W|z+4wHo&lT+bA(uWLee|rZVNpfzoz|-D@kn<5kkg|nvB@zs%qt6v z8O%*SY)vzfK>STZc&vvoaj@1KX{}5GO;N<81U~wehN{g0G?b#fD#6?x`XE{6&u{@O zoo2ZN&rGp8-fmVxzt&xU3btUt4u57=O_tc|NhBwEp?O+=eB~Kvgf$bV+hDGoGOf=U zTy&x=o^|8@I!U%l+PVhX(K4A%{4R?MTLo zXmdr%^X@_Q+Nr49#ku}ETi7kw)mAXT&Jqj;kBB1ubMN$wa7hHdQWVW7SgQ_-h zB4_;QE8SoEqpgxsSRt@p|CFT}{7P?#Ua*%g{sOV;1ak@$@xsrXkA0 zAx7)i)g3~_Ww-u7S{IoLw?7FkpA;*VqWu|?mBKJ5G&uwIBhOp(x!DqKG{sGdKrhrY z_s$l3F}69^m#ytIVEq+8(2@d-IxlY2e<+<<2GYg!-}PnFo=dLg9}4rgZ>Mpa7b<3% z)-_T7RjWJ3Uj!K!P#yj$ln#Eg_?Z{GEI*C6G)Sw%8Ag-8eDovXf0fj&ds$h%7JJqJ^54He5Plb_Aq z{*J_*-lkIExl8X!L5jM2?Eegt0K~nf&&t}x@O!6s^3&03p z&f}d}86wYsOUy(qSylA9%Y#Bo%R`lXvQwq?F+l1fSM1d6sr3&9Y^UPib#$u%^8q#U zQ4aKBuJ1g`Y1Hoa9(g5g8` ztUmMUqo%pJx4Tk)U1}(DXZ|X3__~La!WzKNhp3#QuYE@lss55y%(-CQE@-X_LE!uC zsxO~m2EVt3NURKfF4|iDUijE=`S4*NC7n0$Z&v0H(6GvsV>`*ag4tEriE#)m4|l{= z@_~Myq$-<)CZDc;j2A{EFBlMw@>o;iP+WzsR+fBnc~o9=Mb-SYb6<^qfErw|VXyci z!rhkI1+b!2rik#=x~2a%b-F|CeJS87;DDjPq5MB~#}NNdsxr`AmixfFgzJG<0uO0O23Vq#0Iw#eYrh;}=ulPojEQb%;s2 zBV7_0-^{tY0%_mu>%BBr*>-_{0ypk{*~SocNUAK3p_8js!_0!WqPw-Y;&ER2sY6^I z%llqu>J}-?E{|iXNsux@T_DVz0W`ou*?!6g`4zi~;&CR#c&Ui1q#zH*T&2^A{4a;O zG;tD+EUoRYcLJ5vOAdina947GlaS0e*Jei9zg~6fpFl?i8kEzJ7;A%98sv6GS=)z- z4KY>ELyh_1&w0fM^QcHz-?p7Wo~<-J+X62~o)O|*Akt0^)1Gp#(cI&7!|%-4cO0|R zBlwV|V#+e_I6&ZYi@aSxL$}AQuq4P4c*h+9oe1Y8!Dl9FT4Rou!(%0khxiX!LZ-yh zA`U@-d^P*6QKYqUH;-EEldOiM> z>0*Ty^MpI+&EsONd!pxg#!rWozUsqg7*ry{Zrfl4AWK6vIp7AnTOKJb`n{lVx(DuA z*9ymVw(o1NX}m)YKBY**!#-|@5z&@{I*pSGUy>WI&SAMtY3HVR5bh%NosmxUDFPU)KS{|{c*_ZeRay@-tWFS zdWjfZGxI5Dg0mbvN(*e&?6)(z5H++poxG=Rx)u~9w>3nxkV*4HtdI*B@iB!iCfr10 z<}go}BpfPHO31zJr|@wKMX+&oT}sD1Eg-u2Da4R?M|l42I3XA|Bjr69KXu{n*Fs8r z?(!r-5598`?9!RpwqQUmH=;o}$>`9yi6FM7YA>ta z#L=bfYkU8F@uMtIG21ox zZQo-Bz|~Qpr8?vn=<}LvHSTSghV^_^y}Rq=j_{m~n_LTWRRsDpn#0RRRAd?Q*V?wq z%h8spgG6$Hm}rc`at{B=+Ol4#QeqX>=-EZSRnNa)MnLnJhkI}lFIM_=*XvAGS@BA6 zooDBtx$l^G4)u#ov@R*+tOC=&033mzs{M~O0A27}6`+u&6}*L|Q5_Pj^iwr9NYk5| zUB8{DcC(58mU;PfqyU!tQK*w(LhRZhDV;TWR#AGRd?DE_0H%p?Q~LPZOH`-E7GcY@ z+2_6I!br(4T5*mn4az}hC(COnb{_jbA%!23=*>&7KI3bC(#W-<1*`vy`s&ZOMHU<> z6){AiCa(Ccl{S3Zk=1pXW$jBpLl&R$kQ71e6Y8A4&utT88bF6u3m$xQ4*}w}uzV5s zCHY7;9`+f!^gOniZp0|2$h%u55t_pikpFpGgS#U7_6n}5r0YYKotn76Fvx(RPR5MZ zoZ@((A0YucIge1>4obZDe3SbUjYdrQN_rgBN&^Df4CTt(jPO4zt~V9A2(TK6Zt78=R1s?=zVN)Zd3Vmdnf&@Zu8D$ zJkoBcAb%gzYb)z*Wou$zT~Lpi=SbduU_XZdnPQNq$6g|@Uo>)y5FIR`((J*=&rTz)}8HWSj zwf+nE=s?wozed%?YfV8CryTJ6&GA-Sq|xsLQ=N7MoCo)5)#r+SsA!~*kG8RDn8K~q zC9GeV1*(AUmTKJyKF~twX-;I8x+cH4+J3HM-YhTICYtTlVR+vNfp%faFqQeNXyDnI zy0T)6;rjK;lq=Q+EA9jBjDVUH53-M_YlNI|+dLce;{|1Z3Fh!!pl_DEN=QGx0wIu- z6RBv+o#ZGcFf%ORf;n8uzDwvrt5W|mYizgn*)&ZnS*kyFdw^te+q(-<9`kmv{jd*^ zK>rjz>X@B_*A`@hZG$VTX}GD!&b^#m!{lNSK`$iv(IjKc?}U}Z1vca0zbfdginhHS zigRz*4hs6<2&3T}x`gu1KKH+`>K?wZvK=Vx+$ zOs3(k0>OzMn|ts{icY?CurW0l?u%qFTi!G_3xj$8U1=}t^@j>n_<8hkXBb9}L6I3| zZt=2-W*j(}!@j^D^|q9!@hC%vl0^b$F0=BOB~9l7IgKk}?FiDtIYELgh*J4x zZ@H=8GmrN8FS4V_AXml}sw4=|@q{}86JJ}ZhUvLw7rGTz&g(bB2p^i_1m`7@;@cq2 zT>^2u9Hy`J_mtLc2CIN+O~gGPd}*qFElYH-4QBeDbu_OfUz=v4pY|HtS;~k z^HjG0_N+1}E0#{5n|#s*nz2c*@+2|9@8U)@+npQXLEg5fYk3pB1#g4ZzAlAMYU&Mv zaP31Z9<2)qem1(c$u>Ukyw*Y`iOusQ($4{t?X|dX>u{Xc@m9v%snI)TU~sG zCyl!4t7G0BkP)Xz&4ZVk5)`&mr!sFJxw@t|Vjd3Z+-yYJoU#qBW|f-#z!URsB^kbG zL&>;$7Q$pwhj+bQ)S*osuraCh9bPYG`CrJc*^{Z|Q0o%5Co;o4;Z#OMhJm~ijdvRl`Rr`G*LctBmTE7m`5 z4C2}1=r--S`xCW6axVz8^&>Vi`= zQ+`fJ*Ea4hM6xY+%EN(lhQW8ym7^k!scm42hTLzys|2tSxegSh=<=H!rm_9@Y%}P+ z%%Qdko$btuQg;-+*mBtSrUFj6?WFg8{&}OVgcu&{NWX&ILft)f-|OHOLsKh`v-!`Z zCfC1mRgOCsg_aw2a-(IgA{J5F6jHq}p}lG@!@XeVwo$tZC*0ULkID*cjIjb2^s3v* zkh{RnmI)6U-`~sIt1`++sXEzCXw0k0PIZ4u=i{OZCgo(LIuo~E2upGYdj{4O-*{p? z^`#vw*Y4H8S``g?94zi^UToD8cQD~yXx}MVY*ha8QP1}zm8>*f(D#;`Pb%jgMA&V* z{7@D9Qe_+E5|!Q%cHL%>XQ#C;G<)wy0)o(b93ZAkU|r&~ z{pTlIAa%P4fi+GBkli-s1#Y6tO&9_qt*`Pk>xZ|ivGO;v;OsqlXZ|n=0I=V?}xS;x%`x&c_;Z2x7SU_4sU3 z60aaB&KqbQ7;d6|H;V$xN^`lzU1!jqI)h)E7J zOpa9xv_dL*6H&Zko`>6DbC}rmCl=?|Dfv~sP%v|V+m#S)$0aL^Z2E`moZJzW| zvS1Z|1TNsiAE394RBPX@&Ka^n%(#&-A9474g@m-4y8LR6$|hnY|GkPi|p-lg3CV{aTOg9{<=@x)PDd2n0Jz z+qU@o%hvwkQPIGmF z{DN48!&(CuyU;(wxHO*ta<@ju0&A)vPv!Ok)z^9a4MqHK=WKLMwr)ny@s6B4 z#afK5l!wpmUR7~zDDMvLlb~adp?c|Lo2wFb^m-HsKxqmYR~&9>4l#8)oOL97C& znVTXs>r7QyYjRhe_RO}P>h~$OyZLP5eNcywQR((=9tZXAx=IK0VOz#13o4Jqq@88q z)@9btviFmq-1$H|5(Z;yVC-eXWQ`boa?2fHyd!Ki_QW=|oUox;vyomWX zJ(c8>hpn&fJmK3M=3q>*Oo7QK?ubv7IyZG5>$hb(Q$!QO!nbwcE2Iaq<9`+R9zLAX z-O`;AzZ>sm*kdBCL!%>NhV|)CfFL?($xT1U_738SiJ30rw&B$ROa=J9BWLi^_rCv2 zWW~>kthijo`+wJf50DCQ)rIa}sCJne3)kB`4OQqg;&3sOe-TK>1lLrx51CRu<}X=Q z2jb_ISlKMYvdS*D3hb~Lw_5`~?YB$$*kv~$*f(NLp(YC#xxF!NgB;%SR((oyDW~Mv zQOqjCEIcMUTjz_G1Gbod|>TBE3mzusV54=8jQ}w1Ikao4`MK=ioQUh@Z zCa76@H-|7vA5fm%mUN_JnuvT~-W5~6)@)L#rB_GXm3vSdZ?XIL`(()L!=0HJ?h_32 zd|~@QZJlj{XEP+h=&L%rgqd8?!ZmM79ZO!s`t0p`jXG35Y5={qSnlH9oGMMV-b@$>XROO#0-#Ru4;wmd18?#S z<~=bGD=y?K-pq>ExrrpOgq)}Kg<}zHdEYUt7deYSKMkXzlf(tH=S>C;5k0pLyKBVE z+pOXgC11@x)D;zE1 zrTxfrzN;^@`TXV!ed+!b7}Js&1vbWao&xTXEg?PB%{6OCUY$rk-OYX)N$BAztNlyDCOzJ2!MSX6w>IZk+Sfhm0TN(^ z?Yv`#-l+^6t0&JydJrK`)FL zYJ&1$oHo*OklIE^iy%2^RKl{l>H{=2QF~8M?w$ zSi3q2WSM7~ZTFJ6;t}|k@~&;H1MW42Gb9b2-(J=Rgb9!UGd!R|md`zc{Fg05rh>z; z59IHO-a{mRKP-ul;BAWavNjWBPMwFv3O8a=7Kse1@ux~LaS*G@Pp7ND`Lhz~-5EPw z)qPrHi$jhz4r^_h=F?I&2=wsi)h2``J#szHp4S@H6wi~zUxkYkESzJ(t6V|kZRSfOI|-(Hggl-^}mp_ z2$OhMst%puJ&PVrP@TKQt$r)9nV2ae^_c%Un9_{$+y*z0K5%w9p5A46)|MjvVRxh3 zyu~flxt57rHA}weS=WcANySL!B(%6|4<}vCHl*p+q?Oj$jA%@eRRR(t?S9ZLf65XO zeNQYI?R4jHI*0g-fH%d2gsdx{rmtQ8zThg1s47V1bvyiBDY0UFw1sg=#~w?|_AH$H z9vl{Z_3X9_mOC*{wxK;t^n&?xaTjd+rmReWVA}XGxxcapU>5&f1m*p#HkVKielKe~7=oYTXwW04Uy_36g@-ZdX$;ocyJ_t-uoN>v;Jy zQfLwPISi>T{OgzZ^)EZ79^;Q>Ob;|A8&SbzH*@o_+l8R;34zO399&_(>&r>Jw6)OR zQArPcs}ffJW;_RB+UmER1s!5XHIMobM|nV}g)&st2+$O$W9CR=w70F+kb6~V-T}TLrs18b0kX>@^pvy(y~!P3hzua z$4LZTfI8TzUuWwn55eRVZ+M!s!D4=z8y|-_KhKX~!$;rH@ce6N+5%g%9f9|q;OEwQq#@dQ;Gb=@ugnXY&Z z4Nd0}0k8>D7VRf~_zZk~00-C*i)hdYHq+sGIQ~g`uss4GWjpKqAjX@Sd96h(BYJC| z&jiz|t}N5HYu7#UeC6Yo9Slt9vc2v1d-{91>f~~A5Gk-PeU6tQ>XBI~XzY(`5A>}2 z|D)_Zqncd1tzkukAPNbffRu!UBAYJKq(dlf5JZ$By@?8lz$S_bL3*em0@6XSU_pA7 z4xu-xp$7;>A(T)<4DiW0&ojQ~J?H)N{>?o`#{FYm>zZ@kYt6~&FdiIfmnmUcC^;y| zM&q_*sOjM9apURKL-Qow^z3m`T#SaEEQAKQ7>|M0)vrvvI9ylf8xyIaHJxMUilB44 zkScZW*afKMa2PKSBc4w-f%m)WYbvS9Fssj`3fK<2M)YA*cA4_`jk>CKOOyX4&NrPX z|0(%nEOT`v;lI0m#F-)SFs>mv#vBh!dr}ny* zVIGM}yva26f56d1bDpqh?gmJ?SD?Z6NogfiiI7LGPuYrDUh!+AXAmu3)yyh(ozEUy z6^N_nrWXEIzGAa~x+vbwjCBg6(UN< zeEwz7$lLf`Hbf+IZuRO#yv=csioauP1d?XlNaGmJv?JvEytu@^9>QsR2r<~G+v)GV zy%yRLKG|##n>b9n4D*{H*|C)S4y~jQC@JpsWVgQ4tSIA5Eh2NavY>rq1Ztp_+r|x6 zMrzVywp%tK%gG35*5@VMN<}N?)M~5q%g=te7vCZ_8ydGC%!{>e#CSy04W-2@YY6`% zYW~!$f}imL(3U^5_TQ<~dOBwUE)QLJ9F4kQ9sb_vaLwW5fhs9sauPyMh(Jhwj;w<} z`~VRESplTr+(My&jEmb#JswJ26dhInk-e*jygB2}iBbo^unia!$&cnI@GHuAw4G-G zyDvl>Ie7VWaI86=#)g&vd@omGItJaI7JCa+&17N4`0NRE-mU#RAsv0fn)a0)z2v9~ z&(Sca7ClAj$x6(wIUos{<}!#8ql+DicAKTe0O8HoC(pRSk^HY3Q=`lZky&aL=LR-E z9L|{3v}M=b?Wlu5I>i?Rkzp!620u92d)4>GMeoAHq9M)k{J0{7c!2S>5~g_#d5rk# z)VIXL4sK^9pJ-wju-2En42AxRDSq2fQnG2dAtDwBTX=Uoar2l;`ua^{n9cqIoeL3? zOPFl#R-nF6bK810bM>{#8O~!6an7l}*l$1^|0-A@2`=H0;8X166&gh}8p)S?QKYUq z%h>_o6)bup$XQ7)h}Qd8WX% z6wT+;J_d2aFMAq((;|bmLbDDgf2zv4M+UX8|6u#ro?plxB9S4}NcUYzmbuX>p^Be_ zDbmBjH^{W9I#Pm5WQ*p$ocwOV_djsza(NU1#(o3UtOsh7GKqqXKyp$q(4YD-t@ z-tDBMP}zU_)P#Uhk!o%o8%F-~c`|{jwxQ=WsjPk)Q{sl@*FR$JZ0p%kUVWyu_Z^Gp zXWk9u@N??WtdzI+dF4&(Lu$mZHqKa5i=^m#tqzgZuJHjy@rbRtn1kpNS3!~G)een| z{3stOLPcLi>;%lGE$oFSbK^|+b|Sc4Jl@xit>?Tlr_?BklNcN7jS%DuReOfbnGkxk zQd-frd8u1Fuv_TplZ<8HAx0pz#afUy#9-sB?!}8mPFC)G)>PtdKdQ~gq`a6_M()3+4L~5R8&GD@YeP*&uA$3^VT`x6M}$Vy zZ*tWzof#0y~~0I~L0p8h#tm)G}>5`68kw-ndg!@lirV1Ezl2VYk02dH`-Ri;uV z?Ao{)zs#_Anp?R&52ew39=edlrV52DCpyPLAEOT*ucd#VEtA_Hrao8Pzdm##$-@xU zXoV4*Zo6#^O@4qsZ682j&axHn9dJmk)97?HE=1N-;NA@)4@SH%wE zi2R5dt3HRF;8)LDSj3W)4%3{Cd=}d1fAJyh{=h~) zAf%tIQRve*yT|RW7x9+p9KU`Qqlb}HPUCTl0ji_lANHC&AehV|e-K)wpr4h+B0_HN zDKv1o;t|-=%gs|z^^98rfoa<+7byf{~*w+s@4)ZR4T#j4$ zKc>g5E;26fR9$GKZ6>#2`cr;C(%$Mx-Tk*tbHsHvC@i+EIw1k09rOr7znf5EXr=0v zFfJ&7_?a6iU6O&{ln^=wMtH_|S>Yjfw@>NQ_LZ}Skm9MO)1_LDEIThC^zoTt(f8hk9~%4$Ra8R;r`Klr@_z^L zXUU&XH+>$_4bYK4m;J)HU3B^-Jj1j^@LR3(SBZyf>aQQz?boxHbVq!jZIw75v6`P_ zyScL=1lixdf3eNN=e8|}L5a?@eBk|Zp;XX$7DU)hxbl##O-Oo;P)c7)7ldiqtakM? z>?7u%8sU@Ouco!n0XHXD_Il+9l0L0moez7~`cY#kO2w-xse>j&J{rsZ&d9eD0q$zW zu(jvO9LUQYYP}ai*KE|I#~1SHiLJ#$-o6}9XASBGQxrO0^n&&TCQEVC?nIBH+QJ6n z!J>@M**EO~X1~FwhT|Y{00#ojpKGm*3% zK995&XCv1=>K%qc;v;#swJT?1uj%fF)*3jJdROr~I6hlWbH9REBNP#XL;YC=w}z|a zJ~U+wgz-{;x?Eq3haEnuYtHb7XXC@iJ*UT5#=nReRFp(?&5)OO+<=6oQai`3H%R)t z%a~6?$XK&+K;Nk zmXyU((B$YEi;8?FpV8iQa+tJuZ&DNykn_8r z=67v&EfE~w+qW2A=D|jKtqa#0kG|xS;P^E1ts2IQBxgD2nj$u(aj@PmJlE}B*_PDg zk)IRra5jdQJh2Gw#cbf8U^PWKI9U#bH+#*ke(Qvwob&&?#@UC@Vv{%<*w@lTmUoN*sr7TA4O*yAvK)A!lWV$}2(PVJ)g z7@_OiRx+*y9IulD5#`7~%_?S2U-;!eZ)Zfp;I2|3SAwI8Zx80QL;}DV9khNiTy1+a zum~4R8*cO<$9Zjsnqo$|TnPG8Nhe^+>T&<^vkYY4MjAOpD;s8Z*w2jxDlp}p4Q}=b z%SZNmwq?sN+~k4UIECMcYuspNi|?e(ojKP=<=S&x`2OAQ%9XPrpn^7u^FD&xc6E5w z5XQm3`J6*-4c^A!bfrJ7l)QA7vR{s#F+_(-_$eG}(H%1cRA51)a?i=WI@`!;C^6vx zaSU{-dq{MVUApG5`1%}3#z0=PPX%3*e@Qc$VtbH{(Nmk$pzHI1ex%Qtee?tkQ$-^u zLRjY86asoG*Dn0tk{3L1{EA-w_3t!&kvX7QmpBCi7? zx3y8P$N@HuogNp1wtF*+AUU>&8s-b3WiPi! z#>*|i`zZ747Aogl)+CL~`>aM41T`3S4-?mxu?HGN>Lu=?w*K*gU(d=-PLt(DAe;dc zLs+@~y%-{;gny}FYerGLOuxKYnESVnUvc{nEV&xENgxP^u*BOy^I`bE!rv}FL`;L?uf?*(%MwA%J+&X+_ASG?03jA)7s+xu+_XimTa3>mZ0389k{ zm$XJ8&t5PJ`T`$+zQiGaKH$?9>)EaWV4cwGbrbEo?GAf~z4@_i3jRqWe8e8GGYLUk zK&~zcg6sQvdVZQ5YsbT|piGQ8zVyHo+B2brYG=!vMb=7jX*9P4)|yd+QQ_zEZnj)Q zZkaKsY4RFcEC6kyeZF$$tSxM@*PfPkSY;|G^sa3pD5ynl zpbGXy&Jnmnh?%4qI;qh^d^<^o?O-Y0fQ92Sz^fCb`e{)t_^w&xwYE1N?N97iBZ<#m zZ^i>c+CQ@g^10Cb`!n|^X)RbKBC3p5-v6lMV~IzRhNs%c(LeaD@20DLx)B|{lx*3} z=Gm)6SfAcWZ$y}HOZh0U`B@drx!aXW4Bd3L%rNk|33v%N7~e@02mxf6s!`xe5tXwx zq=E`h;5`SUyN6)tqshnUYTQe++T3I#hRIX@7G@i){&ETf4|rU&Td88 zMN|Ir1mXbs2qegR97M=k5_aIo$KSW}qV{u>dQ*7GlmBQkB&y*^vZ6WFwd=g z&y%ugKkvOZob`Ox+HblrXf0aJd@&N0(zqO_QDJEspOG=E3fS`Ph&NV_{;ZNOLsEm%3RktIjuvx0Xu%ONAt# z13-Sahb>UqV-q%yt+W}wuUy^Us36!&a7=9KgTxb1J#VBWpd3Dt#cq7L`SnD@9ZYf6 z?8~p6`p(V0y~dzvPYFt^2mB+)-O=Avt$Xl?(D$Gk#_MQt5cdqB2}2U<+aJ-W(o$#1H7tLZ8?pgF;-DCr@_0EZfk=z zAY=j&VF4Nj$YT&b+tgb+Niz@O!p_EFYLyj<9<+o=`!$1EPymqZYh>Mk3h29B`*dX3 z;!kl6K9J^!fdZC&g45XBVFQ;5b@5iB+#&%froz%*85>={oFB zppcIGk;m7pJkz2>?E>g_;C7dohz0LCB>V=c*Vbu@?M5vk3Ye(f1Y5MO9S_ie(rd$y^Rq0VVzz$x`04wL zZ%eka9EubCP(Bd`3my9O5qG+{t^?hLj>w8R4|;%dlooCO`yc=1TJ)aKCbe&66May7 zpBu5gc^ai`1)?vcPXEwt_r+Bv=w zOQPD}C$g7S7i##Ezv7>|tITH`2oA7?{B^O-;KW(HvO^H*+1BpPn-kYe52;b+SPOg_ z-N<7fbhM`+dPw-Ma$lA|=$(=NPi;|>iDBo^CfwmqJ9mb!^aPeesBXrzzyGB=qGQiM zrSW;1(mD^nFWa-ZzDM>Ox?$yo9qV@bRrSJP1(rqjcJH$YhIWjQyd){OREYPV;x|Im zNnmc=TI`(eetZmioB6K&&N<#+>T*LKcxj3Rl!dQT&mWk|%YAob4fv&BoVW3&XZ+>| zlYUAY60Vd{(igvQHDtsee|;#uDReQ6C6DXZt#a*h=$DycL6t88)7LE+zajz8T&1?J z1jK#P>;5tzoy#1&W`^C`>Qq>#^4j!A`}||g!-%&5c~HCEs#T$qkmk02HbvF9WD^+t zQ`tYih6L8t9=!embzjUS4wer8I{lc)VK=?>20Ac`Tk(A*RkRdl_cY7^%c{T~bq>w> z!&;+DQ^ke(D=I064%;2L2vs+WY0^96Xep;yB#+IUz!}jYspDps=X=;>@S4q1dUlG7 zRb`h-1a0jckU^An}(+E9QOVv#3({&d8<#f57Vdb{7|K+6Y!QbBa zc=%&_;+A6$n06lzT=oPK=Q}^%Y`?`{D)EC&^eDQXK}p%IzxJs>Sgipor*>>Jo6i2O zQLpgzRRcQUh_;jZ_l`!x)+I$YLg;4=J5E^ST45Xax-H@PnCLMzu$#OUz1hFI9d}#B z(27)>6+;h=xk+9p7?iK`BP43QGr6l*2Rs>Kui87h#5ck-8-a4vC}we3@JzN#PQnYCrcge4pvT*}lQN3cDob2P`^lk+nj1L|8_;tM!LLH^<&itEz z0?)p8?lj(j{e6@>kL4nO(6g#QUyLqeoxzTf<9dSDvfXJCbiqrk75ofzCjo1CAG?A) zK3wtO9;#|QTAvho_3ELy108@ z!=AXqQf{TnTFMDNm{K>YA5V;9 zSflf){OJq+uSBKMY}pS%ez;<07erlE%k;K_mKj7)yvg#>gYQ zOPst$xr59J~4u@=;1)nK7w_h|R6Nle8u9j&zrF^4aMy~$q35;a033z|v zV>}(|hT5CuW(F*#*!p(GE8F>x-rqx}cwY1ETl;fT`YyFimo`lB8ea7bT*7E-{q@F)aJ7S}&{e3Pu4e@O_6bVo3Di@!uHl%rrnfb9 z*3B}$qw<%7ooW%V4*`aV2$U>k$~{!0@uiFiI1INpcK|Dzg*UTwkD<2pV#1$RD&(LM z^$191)`VU!cAeMzZ-Aq)q%Y6s3zjb+jd9#^@v(YOi)?F35leQda%O<_zupb( z7-$vXQcIK}At^QSl0&~ITJOXJz%7bG{x-IjP2EchLX6EzQfdY&j zic~uwnd8$>yBw6dmh!B6H%6dga%YgrUu+UzqL9L%r032brT&EP27sNX z-fbiVVj8Br8v~g?V~Y`x3>+D7zgC0XPWcv8dln!V#tfIq02sRZ#8XM%b%}7rlnmnhcY;M{=!Ryu^&?^|C;hcj6`0dd@@|nQ}0zTMp zcV?R56Q!J)a%~&oV4X|QtY-rTJ4`s!B;RH3?HS(Brod90_@?Zv2PFfhdkdif-@px5 zhxg|x=#5{b(jMurWxrP`M)u!L4TmFMBr>ShJdVtk16fno(#t5#`cCG7?yB+R{g?X_ z?nYYjMA#VcWv{6RqYr=y9Jp)NcbOrf;=;oPQp1cPN+F2lZZSl967Gm95I9is$oW7hX9CpYV^aIE`eB?%u^?)zIoX{yBu#N9 zBHI*-;#3<{C06x-!aAHfOSLq3E*V(9QF4yYC8T}^-Qeo1l#cLa-q;SVChz^h_VzjQ zw4qt6B?&q6*or|7mt=0|YeBZ586_NjnG=`!4gxniFRw`l$?FI5f<6D7OYeLs&*H~q zd)0B4%4c(57wnfsbIb{Ab(W@?)AEWG*Qu13z3=`Uy&_-T_?)2Q!wRHB!`9i2$zMqK zkhcL3HJT$$8MOoQz|4(h8(Yn`2fP}pfF64Tz&`bUWiPv`=lYdhH?W3Xz*6#3rTtjX6D$$LeBUwE(lI;pPrKU}Fq~^sSL~B}QOhRb_k+a0 z{!yN|=Ti8+P!NZuk|0;Mv0ER$5pPt)4A7P+nRo#QlV_=>TZ%I3Y=a&xAWyua1|~+X zh4~HsQtgdzEDKHr@NXF1{?(g&i2>aPq;31aV^9a(9zwxAORnH%LfsvhWHt5Y%&!S? zltnBf=>*N7B`l>|lVVX?$(Pb_x7^Ip>01?(9+0146o}yRsBFouM45uz*OV#DCJQh& z-yu5C2;Ieg^pzx-fp`ycQ0E||nCyz3Y}Tw~KN0I7zj>4LY`C$ZFTD3wAZL-Iqrs;- z-=CrCCCl=fc$S7k<2jGNx9@_TbG>~Y_8|8WRd_LUGm{{lgdH%2ejv1ek#1c2+=`6+ zExz1HX{4=wO|C-7nMTwdEL~k=!jiPfsJ&Ch_viYc2exN|XX`<9kZg554F$I{AKm2BKw|xCO`acVP7%H37cz<^= zsK_{Gw`Igz`hZ3yb>{O7$=%X3Kp?VQ0;LN7pvz;Ad*Sc1>cuGWwqk(hNJaA%!-N5>Glg=_9K zet&bDuj|v%_(R&f1eC+&)~hog*zeM^}m7; zd}mcy+z<0HyiOR%{+Nw-va;U2*Y3`#D9Dd4F>IGSbHa-61GdG%983@cZ%Kl(aN}aA zEy#eOR1x#SZ%|q>k0sbvJyHdlzm7aF4Vm~&xZovU(P??jv~!D{QLVQ-=ntxDH5j1QkmGAi%POtzVN z9ljIH>Ge}3jO-{IRi_QC@<>b-T6>I#)qHPPv!5>`tVT)6DPu2=5~iK&tcRzKruULt zMCEQtE-?nR2bQkgbU2i(I{aXV0O$i3Rhwz?svDnF&JDO=)ho9J+db2Q>|+SaeHmXJ z8Xd&;W!+nIrHP|UoN9?hsFhIE9S+lVu7lV?B;_hL|1a`_=Hy>w5hhvx8U7n7_}AX_ z6f7zvPLUpw22;dKi?n<`;*P~d_0`X?vVqN&u0Vjmoo%TehqWml4GxceO7c+PF45b`Nu}flmE{(fBnb_ zrzx$oA9rHd<#8~gyCGy8b?MZrZJj6J)FRtaYf%U&fVtHsEr9(zRsPr*;72bTN-+{!r~GRRrMzDiBimE{)m5sl|; z#I1KC5Ae5+Q)PVnHcXrot$@#PTb-zVwFpZza^%xZ=QYZA!F9vkAMCGZP^B5N0q#n1 z=tbm(Av=3YFmV4*KOqAp`c+mkIeKvGbci&tg(^l&*q^L1Foxv4AQuS)v^xn-rX8HR z?`~x96|N2RtA=o^qTIy%Y}E7K*;bmH2G-Ur*3;To?kpD9xpBV%J(ydZd?RJ^H@U1V z3e(J#8@6?@m}vS2^S7b;>ZfO~Qd6XhhiS@b*D;22+z8zQHusdYKO;h~MPb#=d>kIFuS&PMq$odnKlZ4+|oT zim}A_#v2v|QCl;E5Syj>StTkvmk@&BZg8#Y=wJl!QB|$XeyY+=v#iVdPH>dg*t^I2 z!~n%h5A3S`f3%NaK9zX#8jq1~&+p^}OX8 zCA{t=kAAfz7>?=+L$9+4gBU@9hBmH}cTcxE;Tq7p6|7Oe>KE9-gv(d~z2&D3twL38t+;5wC{thL%969EycfE+7S)=|v?XzH2%YgjdI7xC;V z819yr%@3923vep@p!;3hSGIz_5!udbUe%NU`$$ zpEO32b3D6k7Gmo5XESeU6uTdTnyBjuJMd?<{^aAE+nq3vjZ}%W47A6<2Dq{HV_b(6 zM;JAk3jDle4?p77~_<`-6=M%bVrm+oW7=kD^O z-{WXiQ$+EwusY`^?S8qx)TUibLTVkUOztPrR)*(ShsV)-Z03aJ^zfTY9*R@Gcm*HD zDtzDg1dOZ3U5&BpIWe$m%y%ayXu01Ty_3=1<2RsQ)!xjk9DBq89xvH65?RoCb$5gm zStiHCyyA2C<)EfR-yS=)vFf4RB^E9~nj7XcRWsA*X08z%2(7vqUD@vRtuqDAVM$Fh zhr2mv)+Nhp6p@SFEj@!yeiJQicYpcKm@F(_r*q|~6n<`PAT7km6b@)3`iwlXRXV{Q z@04kLO=_eZfxPy^7FHQ@(-f&HYQ~cyro3#M^}hmq|4ZCIos*Yl=$z5}m!97ccV>wF zRt)1m)R~4;6j?; z+n$}T*q$ODR|51Kk&Eey7+u%zM7U1>luEerb0Fion%1kAl5-#felz$NJL?HtSW>xB zOMKLBq^y4F&Zx_+^-t04Gx8O75nYYDFwq4^=1X#bThf+o_xSvxAR+I?<|c%$^)Eb(gf#bTPJ;+f|6Df$ zdGa_Op+?lK08~nC=gWmeMMe&1@DzRJ zoq&&T{13NT;5&e_=fj3{P2a(Hl4Ys@eUw?pMC$s+buh~%ylS2{Y8m^T9{=K;OOYM7 zSl;KC2q5UV&&`kFm<`Pf!_8_-hKLK@3W7sYzqkedX|go`mWMO>zb|Km>9Ahrt!?T zPdPU0M}GFXkEkK+G!Dz?5%N|bvsmmxhRe2+-?00c#@D*-LbL4)2i|Q1?q+nZpd*UW ztsL6ziHS8z#Yo1cit@vaE%g9*_rf*A{sVAtg%8ob^^Snc4Rj4|p+TV=MG3eyU5236 zecx(fHPYf@uC-mia)tRIYDE^rN4Yz064d<5UGukfTi*%s_}{wh>i1NyCXmoYog3%hi!gB36iDX*gS#3Co{)=sYY-;L5B_6Y%(cl}})Eg0}{E|yX!X8|_(a(A zCwp2PWNhrL#5$;I5;{>IF?@V)Zthjg>mAED?LZOtqG9P3%F&1)+w6v^xnn`SRnu&| z=weleA4e9RSa^fVIY{5Q3eCZYf&rHJ_v@|kIsq=v`OE-=@y1G;7`s#NVtd%so_6^i zEaVxT(m^@X9R@lku zV#p-&Mq%Z0>a;vqZGS*ny=7RT#yfMXPg)(?nxBJy5cqQXk)YTolVuDKmE)`@%jK|K zwbwuD&^vuf9&r?^2}#~Q6?0Z-n7_EtD)G;aW9)bK&9eYEK*+!C)H{KcG%3Yisdd;L z9vj)aSVNs8=T`sPOgu~W*pvqkN1kcUdsG-u$)>q z!raT%Q1cK)Edap;|A=v) z>D0#N(n#&VzlY4&&N2!*k2zH_wamKEfv*HlB5h-}a;2U5n_rw99dHN7EZS5&xsR(e zG<}~Zzs=!MYy+1DXU6AdT|z_y8E^#nnMK$F&#eMM?X!oN-tw{w zC65m-Z-biSA$zsnf^qFvY7GN*as`YfEDAnVqDbsPxQ))+d%MjVAblV3t<_fsn15V+ zul0M118j<%fE;T6pcXj0GbcDMrds!(?+eiqUmOBqbrU}Zrrsl7Uip92o4J<(4zcq?${ny4`-hLzsnD`}TLgx|-NGI~a z`=^rgrJJQ{ywP338Qu*w+ctv0h!NPuPD`wih#+ulTFeB5JG_>^R5t?`_vO}L)Y5_Y z;4|lx&Q5zA*LcS>yJHy<13ZaEVSg?`enqrSEn0vUj3uCpVv7OX1QU(JMUL}zrgBXF zeG+;|7xx_Sa&RRLDoUHt)kt4qgYAPAGvdnZc%592n zX;TP1qyx{@%Jd2N844U@f)NWV1r5`@E05Y+2zC7v6*E#>5s@W7i7)5nN5%smpWkob zf4hY%W)iDuHMRP{qsb>LY?1MG88^MHcShNuGcw$VuSlUoeGPLwua~57FzR9o3Od*; z$~x>b-M5R(A3*Q-vX+!ors*Er&t=H{GzxT?dsh)s?cGm>;;6;M|83ZGHMkA__Iz_2 zy!Q8#ew3pp7#SY;Ot(wcK|fLF3X_v9{fZs2X(g=kr2NsG%;b|`4USlPsGQIVi}im z3_2(LTZ8aQI#-+1h`@_1f%w%Bh*5${m5DUoTIHk7K>xzVgtWLAkQof7S z*e1~%5bG+>iMtG89ScJD2?f^X{AN)aq8zpQ36L1*;?WnSSB%afB_BX;u|)%UXqUJ7NQ+) zKJS_qD}+(jfj30=F=FkkMT-pq9hy7%I-!?W*-w`0>7krH!EZHr@k;)53!cWsM+=|q zmd{+Qkav%patUNP57(6QW8Yu|FsL|%J#s1>zva;YR9$~Rij`sG0hms*c(U@Yw5Ne zr!mCYUbLsZQq7EtU2vBYp0&X)mNZbs9*)5rt`=u-j$J^~M&4is3`k@79nO4-8sGmX z&N+oPsi)2csgIsJ{9EFYWH3~7&4n!%b;N^uF%IaCUt6v0Cz`@v>{!qzY>zQ?`>LmQ zEEqr|SC|Y^$N`inq-f48_xS_^z?drnBLJOTExC4zhHypTYm5c_ny)}q)J1(RB>yKn z4RlI0u|*ec8I$yMF# z5~-ja)d*j^;0%>EvAjbvBznfTM@qq@p?Fi47)Pvoh0Q_=`aKEcVFiL^;!kIUT+CPL9{~8li2HI4zW6?`F8W@$;)nV<7x079Jo8ER~#{-r$5OOo(g~Va*;RJS<<4IFT`3EUv$9s zKDS}qa+frQH%cCmAdtT|$#uU^5cYNMqVPhQxza~j(jB)FrBCp7?< z5K%{m13|MCnPXfbJdGLV^R6>v0BL(&U&N?Eg?B8EKf9Rd z0TM&W$C~>aY_^^;NayT4BM+qd^9tZ7g8$Qt2!zH$2*Hw@3A zxF@DTZ4ul2Yu8F6FGKlW%56%gfj{8EShM_qnz_eoD+53ZMPGQX_Lv&l#9c;{H>sSi zY5?5eRXS_zfLi1vvRnXF?8zLvX?Ab-yaPosT&P72TM(o7q>Z^-{$?eQKB%*T6)FSh;QeugI`|LxD#82B?tRqwnxbdl4d4GsF2TDCMe@?uQuD- zF+13aB|T8xD#;#+yaeu($|1KcW9zvl-c3$%W(h27+l0JTPV(}yZI;yNFz{+8P0gfEw590-Z4UeSoR8R!yW< z3NHzl8OFRY?UrI^(Zc5=bzVqWw4vhp?I3_Q5f>Lda6z+N+YvfxXV4Bq;bJhxVTwNN z;2?*3H^Aj5T*tD<)n++VfiJ$rl`17nybx6w!3MLZ-cDx%ncs8aIFze zZCv`~mTLIss!CZ*uUqT$qwX&uUk&81a@MYmeUp%Lip}wPg3Zj5q+3uA{I9iVUyn*& ztMRtgSV1n_&5BiUCy2%%Tz)4WH2$nr2~`Toxa23=4)DT13{z`v4b15i@y&Xc>wry7 zN_&EKYZNw$wd1Xzi(UX?%SB{MB<$N>o$Hbz8R!={X93@yI?sAxKbND&7jk$m4+tiI z7NFcRE*Yp)&}J0=(t9y4b$q|th_vW8*1#hYMln}+ZLFj#uA88MXgQ+Vu;976%bX^z zm2}PRZ3&Lti2tjOwDWxwdyr7;+0h|iYe!{!zBh7YiL zg;tfl_WJ|mbILif(Kl)FUPSpfKFD=nyV~6qze3dv4{Ef+ix1gIa$Ccu^KHA|RXpZI z|H#9I#DPM-puh3Pl_T?-#%*;5anat`8GjklkVPQr6_Or(`{3@j%8KoYLuQ|`JJ;0nUOk5h)z_bk zx+B{2>eBmiws<$u_6yV!>0-aC_3b_8l}10axVlijjul7L6T zA3F|&WUJV)@DWYueG9?H*e;%jOxk$miNBYVJF-C9pzj+hf-isOLXYDD$T<`qm-Ym7 zY-shOoh2PN5k`$%@h8|>DuT+yJ$Xo}Kz57I@32L>^9TDV|3DN^(Q6+jU4qUcwy5T~ zQ5vr07qO)uui9JXfoCfT=~Y@`+YvTl(O3w#TFeUSk~Jr~%Xy<&{()DD7`+)uvl6Lv zEfA6KbEg1=UEm6MqHIewS=mtQ7~$vJ0t>UnamjGYf|Qy{U7HUCg9hmkogVzV5t zdEB-m-HqI2%j+3)S_gh9;B26@;ST-dgRBR}#3}dwKgQlP9?Hgj8!x7eh#6#O%#0;T z4A~jmSW1#q$Ua6CLb8ut%vc&ymMl|gQ6XdB8r#^n7L6fUMp5?N7~7NY|NcGC|MU6X zFTQWDH*>u>=eW+}IM4Gq_+KHqv(|8iVTnNzJS&*C6`{aQNTi3}#Q8BC_rBu9&vc8D zm5{j7$|W&3TUTG$C!W(O%>{;QomT|Ax*w%@DtyY^V4F&SMAS@Q9Up5vy^}rRzPQ(X zB!^;5Q>T_V*tGx~Xj>jO`>lDW-9{-!FB*f2d7OG5RC0>YaoeYKmlOPz=(MzN9%Jj~ zgXj6B1b1>(TJj%hhz&(+?M^748%gZx3hos4zcK2r8&1m8bPbzTUw5BpDoZbXsjdP) z-@QfSZp&41i|Bje*Mbruaf$fnlit4JuCX(F2#hInZL<F<82sy1Uw^-j@*C4q!!eTX@3TR9u5s+63mRO>^JPrDmVuFTOy- z&-Lc1-@p-MBbF#HKgM%Y!rc)kBE#+0hFRx5sf_`*7biNAd zE*SMXNf=V=_*)Qx6;`>G{t{Z?$4Jj4 z;5W7xkIX4jFkzCHeJ5rd`~i3cB^?RQ1G-Pazd()IrA%#v!jbT8!!tc+(5y$_Ik6eo zMdBQgGj#*=AQp@X__}uBa?~Wk6A<;QCv%8q_~_k0{7S2&lq;9G#>~Ef&RSR)`IdJ3 zTlJZzOpVc4H?q$_hUP&M>@wfpiVV5-7P+snW<%q}gW5-89(|y9xIa`N^okeW!tVj! z*6)ie4{8J=0Vh|Yv<1r1*15llsR@N!9{PUr(kTvsPi%~iu zK@_keFqWFk^s;kDAsNkIby(glaoNxkCcL#N8}l+s*(o}N?~g;=>=${_6OvzLTN23b z$6c&m^N+jyaftgsX5Q(8f~_LwgHejjT`O}=2X&?8nIMytwp|_{%CXViX29oh$dRl_ zMizXlH)UY=oVrt<- z#&w@P=7x$I>TNO<*_y&VT<@x|m3L~PeTKPu>HslBO&vE|@y_D~;6Kw76?WCmxw}@x zQ}V+UWAH4xQ>U!ZAw8+yo?htOsI&Yj#FLF3CGOt0efr+m5GWCmSDN;1Oox1?KQ(Np zNEt8}R)-WO zXv}p`?w?5YH=V^=`?#N4O9(dUO@dHrxAh9)TYd1fGvjU2p9LqEEF-{wK#Y-WJ5Q4v zsh3W}ln{yXbe^=!IobNTXk|<-g0~t2uFDtv+?cG^a>6Cm<5|j9!pd;ysWl28Ae^46 z3k)?7NAwF<@ng!sGPr0jTc=$WZtS&==N7Qsk9xVhP20aqFAowtrnH+6EQC*4 zc*^yI5bZaiUbg9xhZj2^5B7H+$WP)sJ-D=Cv@Eewl^I2ZBg`c+4DG%raRnTS(ReCi z;{g|EJPsasS_P3sn`3FQ^D-q0c2z~wabo^crWe9yCQ`$gJ)vE;ja(5FHd-yWvirEd z!sb_GK)AJurOaP=uJOw=k6>JG;%A%A_&^uVS3ED%KmA00?cBN5A1W(NOGV}ZwCf!^ z((dsau1?~aw6fRR2z=U(>~^kyN#v7qL=xOjK(rx!NVrq`R&Kk24Zj>c3bNc~-&w@B zB#t=A>M6D2?9fD@b;f6HIP_v7c3;231?^{XGc>;Q_FiaT+Tt2RkHFL1T&Bej`2K$O zCJCOspKE9G?GTtn_No1esLlm~FlIG*ah* zS5d(wZSm!sD0bB8nDhahIYp-ICgM0UZr40=q^{&7I4ZL9_Ry)8SIM11?NI?q4l zq6fnJu3qU0E3zdbPFzD1kCwDt=cliVLf;L~uzneZ_B%gZzcIS?!zH3PA#AbM{kzUk z@&lR-6{8D2!lm&ul?S;#nXfJ_4E_q9j?;V@!ki&w1>%f`bV9FicK3gK7>oi)TxmH? zQs62ElfjZev?+G3CzSWAt*hve?$L{!Wz8mnf<5giB66O0H)g-7@O!6(pH-y6E7G68 z@!+tUYW!P&o{gnc%-61mKNTSCdDK{w5N5BMfs|JfhhU9`%|PklLRKD;86YRWicRHt&kdI1x5Iry* zk;f?4(p5f=_D%uhN}eEEH;NRzoGsXr5ytF{I3l3x+|C8=7{*FfG43dAo-^9Z;tdo9 zPt$txJ@*}k#%xY)ePP@AE=PKz@;cwpE@5C^H zibY7rM8ew?h%QH3NBvWx&i1#7*=wKSqb|F2xcpFWZE|lk+T@u#8)m|c5zyxIrt$5J zfWBH`xJS50)wr_p=uc$y$0PwHhS=*vuDo5;Ar>@RRofYwIMJE;Uv<&Xzz6vK0H z7rb?u9Y$T^-eK4|Y#a2%)BsQtbi#Y2;_cX+kwm)2ONpJ-Qq!Ktpj4=ogUNZ=skCIQ zuSC~&^x$B;)2uo+sPmB?5;nOVGl3aOOFpr1PoB8_*sp9R0wqr7VZp_C6(;2xXTh$2 zRlAoNSpqfKrfbzxKp5Fwn)uGyo>U3nzwiF|3VL2i?k`vk@F)$cad|Ty?PD#X33H3@ z4+74$KgZ-EMTqp9?V0}Qt}Xlb-3JAc(#P~~rL9fd^d_;PQ(7{6kF9jyOuhbEoM&tV z>KJUDPXe*H-@0{ixkor`Dn;$gTD+ctpvH0&tH46tTuCK=vV~Sa-o`7R@eO!jkr_*t z>X@eq=88!;TVxy8WcD+?2KONqSbB}LG;AFY6i1lanCZ5A?EWoNCr#}Usya;Zx1ml# z0A+v_RUQD+2MUSnTuR0e2X>Yk4^$6|ID|~ghzvIeuy!9CCwY6HSDYkE^;JwY-N{RKOJhbuEi4pvKdat%b%;j<8G-Mww zVI;Oka7{Kx$hBUJ7e=Uio<%mp)k{@2bpzUm{E6e~Qd{-{kzB6HV&l9r?Qp|>?fT0V z{O@=D1f3E$+h5BcC>s;;KJ>mRTyeDS?j&~rQKwO8*2dcR)cfs|JpSnOgtCVNvmH(W zlNZ&i*-F#Lb+c~mb>;q0(ta^0tl}7!S~Z{KzQGrWkKkI`N#<^Um#6dQ->`Y@a1&rH zI&^FJ|FS|LNoWQX4)jE`Sb}VEkBwX48{k(Xor1pH_|vR<$1OWMY4>hG{P4^78OhI1 z`W2sC9ZZWEy~aI^dv-r=(C8s75xl(Xt3#NCUwR$)$td_tXD=V+ebqb+aZ2($TA6Um zaIB6%;WX}_hk3=aLMqLxpfC8hyG-1t);&xTw&Y<4;PIJU|H~g`lAVuao?A;e;?g-9 zj}zd{_V-uU&_#C(rx}?&mY`;tAYyAWyG)Ltmo~rqe6UemqDT^$>zB1w9-iKc(Dalc zy=vq`NR#oa^m)8YKFxTSw+iS&jTSWg;wg8lt$G8_o0j!j42pMdlJV@n=_fzzL(TQO zEt8Sb1CM5xr$!~~_`TP$j0VBm;asB} zgO5jV9#fYqHVD!<5(po5RF*UiuQ&q+A1 zenq-qN4lS`23tmgCP}4SEN?33#VzL^TU{QGe?|JvakqYK{H*-^&4^1mNk8Y+`Co85py<- z{wepk^%PonU$8pFB0BR?$kfheqMtf)Z~g02&ldH1I;ydq63BF66DP}Qi1^I@yvY>V zS}2e0<6BV}E6Q#(7HeW*LnD}o==Dc`;_5U)&QG@_1B-9~=LBxNaad1V0N<|H6vcO9K3CoEB0`grl z04DGV^!#B1>IAHPJYr{ohXa%cjEwuE{yx^6EiKw%D{a!$ZiA+O_MP&kk;72xi$o;5 zI2pbf?rtXbU~h0s$OUgN^YFI?zOTXs4mgYE8~&qLR<3_mk1kwa9xERaFK9Fh(*jR2 zF}%<3*e`0xQLe*cI^w$DNnnD9cv`Me34spN+(-Q+&b7xyqTws(#XiKUov{#jBsYw7Ot(cufk$Y}hRY2E^ZdSen287=Pcd;+ZN-_cxtohrS;BBEypJHGV&^Ub|>C)&*}5KB_i=>0OxlAX=wx4 z4Tpn8yG(KL?Mnb2!nJGGVn~e~NcdUXvX|7{v!`1gY|N70?>_G8k{-nep&YUw#GP`D zsqTRFJi51Rcd7IK;MV-S*CdN8S1h$&3soOmp2k0Mi3@hH{1^@05#VmbpoTGPSF|rwbwoX%Xf9ymSHnfI#AkM5#YrSelxgc9HWRlL_ z>Hh8c9oh#{wCgszLnyuIY+GhA8iV)y@W6Yw}?!%wTP>tMgN}; zY@!^C*z?P~!5>N(s*xjqw4@j>!Nk zaUjf^ZW)fx|MoXQpRL)b{g=tRnEv19S4_Y7`#<%JEdc3#g06z@11-kIfEO_km-+Gm zqidGBFYhxy^WP?*I1#z~--5<%4`j)5amRY%Sf8eVfE_m`81OqD7|kvs(G%iJRCh9!8J4im~)*u59Qi@1~Xys|~|bV+EJieU3% zn}Ac%wy%>DwolU~0r-;?Co3+0=OmAOa<&gZ-5{mXxX@SL;9N38j<>s^TqY+m>O#%0 z*v;|%&1z3hltDsePbTDm(V;Rv9p@027hkT zCSu#GKTt@if<0QP>?2#+mEYpu!Ll1CKQJciAC+rdy|=01{~={yN@Zd{Q?}j5Ni7m= zztQ~KraVY`@e4OmiTDQ|F`YZtKUKG=^&ca#r{{y1q?x&)kIAtXzLkOSi|rYj$Qi*c zwM{6I{nkj9_1e#<###Y0W|@Lb!gQRo{9XUJ^ zR`&G@@%>s$4c?7O)FXZa!pgS-7;+Je#ffqae}@%^-b9PYPtHpdtxNeDSB4;MHfDgpITNRg=$m{0t$b_>V*y8U5!2q9>bQ@m<(3_bSV?kv zERoP_Jlkb(A>J8xW9u&jo5T^7U)(SIZKiX~GU-0|TT@X(1DN!wjr9j{kD(9d*MYiZ zsHEEoDAEloVZ4pBt*~Tx+dpIoz(kHFGEh=(@aq5T1@Iaq$S~2K%DebXi6N5IgN3BwABtDT_pRTF7}hXp7N86R1Ts8#b1KTQpaKFcFdH7Cyiwq}@+fFv^_``}5_M(JZgBPW-2| z%C{vo2?>>5-NrW(7yB&=kD$%hUFovbVYpoi0YWj|#JwH?Z$5(7xPhb=^vkx_ zm$;$HnGUw0)EdOH2zj#rcy^dI_U&Fr_zN!mKPw#&g_Q@9?QeWm9)ASp&l9!)#FU&u zWo68Bo(J>L=^^og;*~!#g0=a9D@Hm9iLoh}diwf%{=lhN#N7C<2+?oWRr@tVk<1UN z3X=&wtLBuPg^ORJ#{V_m3&uqA5GF zsI5?aSnEetdxpcalJ_ zPh?CJ7|wQ$X!$9At&$x$B-~bcGjSeD1&5d#=v^wu+w0ysXpb*Ban;W3!djV?0Sg;g?>_P^a=pN^QHbl>Yq+S z%8TC9RjGMUu8vrY$Rw`9Xsr9S#LdoJ{UZ1&c8rn}PG%RHFToTWIn#Ngs=@^L-ca(V zS{X9VgLz{>1>H)M&( zfS$G0qEz9I#_6rd1U4L8ew~wb&*}JcFPm)7xZ^QlSZ6yYlmcM4x0I_`fNYn9?j3RI z(M&&9$c^Z?c!_)=K5SwCsXsL3^iOc}?na4rl>(VTPxM9<6DhTUT%Ih_Ww+TgHhBrx zB{4iu_N?|9rvHGB^}@DM)+C&yKTef8nkx8P=l15Y7@_1MqmA_JH{?n#xvh1dyEG+Z z@3jt=X_Kl)D_I<{_z3v5rdxZ=jbD(R_TQond!Jb?)12q(;9)7-{i4R7{N#6%`*A+-gOo6Qi5e0y>L9?D9e+PHmH<3_QM z{7p@aOPA>v3*MLdUJ?D%R~zCAXC7z>K}z1nY6}QVJBs28Rx}@VopDTMpp*1usBN;7 zF@?PKOs(I~VrTQBz}@W8BRyoOBXr0u7O$LLhwj$*gJu9a;8n7;Cen zHq3dwV{?7ri_-qZf#}(HFB)r`%^zq-UqWBf3ggI`Ru6k#s|Ja~M2sBpSa~Ix_zOcu zcYKU;Xbhi#gYSFJr5__=!S=l}$yL@0;gpP}ZBCJ?xNJ|lzDox)E@$H<&cct&4tB=t z>-f)FVeCG4{U3#$7@b`2D%yR9!{8g6D*iUMvtKhbTra!G?LGXEGGz)u!FK*ZOiNC{^~+>|S&JPa-nuUQ zaXR~bX#%a&y4$NfFe(kag`MUQ-!9G(jhy}0HaG8q9!HE8$8gp6|F52xD-Dk{&*-se z!k>Wyfrb>KYmbCDh=S385;M=*UNuO#bTm7o;K}Jnz4``#7%MyaNi^T~g+b-ZOLV2X zwtVf=mI?m}9ANoF@c8CPecUUnnyH6c7iq9S==Le0KwuC6 znP%!$=CjbtIjlJg{swCgS$|d|ZF7^GnjRo^_b}n%&P?6XCd<|YZ*$WK3>bVL8f~r% z(40%xn;@`ljz=DJ-$SkIG30GXEJ;+8Dg|y`9M4xn52O&~A^i!`CiJ>J|1|VRN zcz87ggx-EZ-uhr$P0b?ZhR(j-7a{kT3={f@iH*YR<%tb`!n{S~~fwzEVSZUM)fNUZ<)D9C+(8ZQ+0C zm!~wH2YcxIKK*}-qa4CVf1zv9g~N=RO_F-Q{HubXEmQpm+giTX-)c z%jf3n5#>wD<5v%*9GIQYU(|-#;7_9106rWs`nUJ{!OE$B60;k0NQA406V>M z>k~=pVVh)(gV;im(A0zijM!<^#Q%nxW)pBg-dU8>pWhgAQh-|OWV_QHGq^9|ZhL8A zIuftLgk#2;Dwne_jd>O}9$p77N2Qz4`}_N;buk>{j60E~CHfaQ!R3gW2}R|3^J(>A8|Fp4ylRf0%` zH2KI}=Z`i?t_MW{bIlPG)p<=FT+g#XzpFz6Z3gyBjdx|+F?l0vKwSGIw;%=f{PFmv z7VJ>;QxkK|BBj=;(YlJ>MmK#~9`u;}o zyZ!as;xw4eYa37B|Ll_U08lZog9%r!F|9k163IM@3sT}=Z;z)`0TbQghv3olI`R#k5zVzIOFOx{L_rq+IpW^CyrQxB!)l<}Ipw>=vXn5En5$I3i)nY(2t~U!}D9t(R{->a9;OUkAl+s5s#2??3&p&r78#YRFRAuLKlNT%dG)B!Ksd65kods2uK z#T%y`!{;Zwnl7jhcx;4Y{Kb@Rh;8UBv?c8~)F<41VCO_yQKtnus;C(B0Xp`fhMov> zcW@U0&-y(8UOAS(i_x;>D+v>Y>8EDlN1vof@>>3_-d+V3`o&!MV%z(yD=(tqxoW18b3f)Xwj}!V?(-F{lLn0Y<`9790zAT2`EUc9{oU>lH_)#E-Gg18 z3q|g8d|UuCM!Ej)LocaYIyZYZPYGpVB{G^RbsKljd+#1SLS%pK`}N_$G-o_h=26yr zt)+%?Up_W#hk)zsKIO5r+4@)ua5x1F9<=`pHm>82f-V;wVx9z+fW_GL%TBe$=x{z< zY<*aYSsKh_QC^@p_$op+-RN&Z)STL#332cDY3NuT+KOSARMpzYbN!RV3ZfT1lo?>r zs8K&-O01oU7-_P~1ICa=l3Lx{)CX7c>vi-N(-Yq?RygnCJGH2ZtGAiAeu*u=y zl;GT{n-Cy*+2!sJPoEOn6tQa7h)->3I``8e0ZnqwMQCFM&EPMu@;uhuZ!nVOBU=_Ap(UPjB(u|76Z0S)lrc$xQJ_ zDsk^Taj2HoBn7gKqzsECKctKx7zJ4GmoN~3f!N=ld_nY%!=Utk!Vcw&pT==87SuJbJ5leoW`#*7&t_^mE>KeI9s&thU}r3g%;YFMj=e{{(ZT1Cq%w(EgUMr>`dk z6t>IUHfqAc=Gz*0l(g4oo(lE^&4^OjVjRzQYVum4V@qZ3VsfF|KekLwCCM~UEG}}R zn#(p4eJz>yUXj>~j7zF48S2%Jv=oV~efB`~Q+dClb33z1Zu=x}@84gUpaO~+07(Q2 zp!8c#=M^rh34)Wb3WM; zk9#dI+8d1=R91_t!|9u`fd%-LBu-6POxi715tn;Z!aZAcU6+51s?R@&yBbq!VwA|; zaz(Lh2ycOqK0Z0XwTq<^=XTY$$_N{Nf9V2sSlk<;x;0YB?Md87Q@^J}RL)ljqh=_K z$hm5jbXaVy=d9J_sn*Vr!?;}i<0Hu;y7TwsTQcjKQ*`!P%pFJ~Ilr2gEq2W#`$v)o zMqd8z2={9la{f&o*ahpZC{)=e)MoLP_fuI`J_!qaw@XmxRi@hSOSD6He}7HWsNSrQ zndV*Dy(K#<8Zc<75)m+&%6bmX$-gPmczw1LcfBLRnYfZ5=ksynMll%iMo?hBq6_|U zyF_2O?Hkg67EsiuC?C;E_MRbPS28@aEBOV5WUkqwcB=l&8*I;@db=if4VcB}m&V4~_ zfabiG2LdF?OC!q`mI+syq?M@A*!A$K_7tPkHFm#;ASvHFNxP_@u$$~rI5Y1bhQ8gw zdiNZCn~zCpON^!!+OhFuhLZ~mB8T$8)j}3!ByI74zdemSY?CoXP_V2NU5exfuz(@G zA8XC9+w)J+d5n^prTQ$As&=>4*{lQ4lG@iO3~5%(j$**Vi_V`|(1hw?!Le9(i`~Q! zL}yjR*J@=RM^!-={UUk0sv_K$A-7Hi{km2Fr8?^d1%1M)sPfR3Xm(A-ljuWD48&no z+v*G>cgx6m-=OC}83gKI&HXF{mEgt z=kZSBo=|5GC)G<{FRpp_@4biyq( zZlYF4ee`i!X#l zU~wVguWR0TC?g3w{gCjQ0SFspkYs#S$%6VEwqJaGuFcWqoZEPvF$mM)le-%Fmt z&1Z*^medGWd_*I3uTt__$#BPh&AFl$<6NN}LlO7DBB8xwY_<)|b%#;AFuFyrmh*zd z$efc`=`kxKttC}S)M~3BW0w19kzCb+q1Wm5+a-aL?cI6@U;l;oal!tgG8fBG-#=we zwgpP8OLMHA!@MZxVL`-X2Z=&;z)=7LOaZIy!z^wRo-(;%^~BR{Ou zfKsNP|iX`XEi)M?GN(?HoEdn zM7rZr7alNFq=(W8vYU3Y ze=RxfuoZ$GP)VraQ;vhI$?-m(fY)Wtq`q4xJPA7NTsB_8n~AO1(?8T=pT-q?bp4|YrD)fYc!zc{L&YM%;TsrLv&n=YWy?R`% zml(nPbLWl4)B_bTp$VGwMSap`+^>%*b?)Q#bj~mCNlUw!=cKg(M)+0&m@7gE0(sQM z7G~4(V|Ume6At#m*x_YgBqNg8keG9{^K2uv?g!ukMKt%~Vm2=ek3NWs65eu226M zc`4TjdNUID_#__x`+iCus1kV3J9P|rlk#z=vwzSfNp~b(;+s2j)%LUFc-a6oK+3;% zh5oZc0{DH*D1QFhC|vwF8;tcJt^Oi&LnG$d76HtMOLET-x#ecF`l@Fa7WG(d29TNh zZF&H@NWt0MQ((b`Aw5m9%J^|kIA4T>c*PU*Im8nHG_Hn`Vf$VI5J^q&9I-V79GkmP zqya^j@u83#+g%w0im~{Jp$|@t=E?B5E;Ri;iy8V|$EiE6=AEFy zOr+TC&#%WcRe3pP$S?v^wkY{oG+M+~z2mTCk+TlIjd;_`@L(UjJ8%95h} zlr1K)_KLMw4as}J8BVJgZ)C?aH%<6PL`1Ma1C&ISHkG%3SK5mw(AmWY3ah_Ml$6$- zZzj)KG)^Pe0J=s*#dwvYSG>ghDBy46$4k#!<8>Kl$Zvj@hsp|NE+OZvCtWX&q%>MM zij4Qn)aDK=Z!-G0T~~j&>~GIJ4f7ppPYc6xasAO^AuhEqbc#PK8JNC?wX7UGXtxU&oVAK4noQ&e zj@)+ZQE!VPz3)Ln){nvwKl1}72zIy-lz)sC$?0RoDQl(BZRUvQhwJl6?+=|%sF7GS zJWQl$z17-iH0a+76AALr_eEY=LjIE;!Y|4Kot>r&MkhB(46Zq~%=99vK(*8OXRVTSvSE|vjHZFoPh}BcsOlgT3A0VBXr~~b^GxEKIN#ce zojmXmxH_ofeeb0Y;Qc(z)-28#Qxqf&TNT6VmB4JnRd4KdKD)vw<_-imnC^b}Y5Z8~ zBC`ZSfr7G~05BGU>*rie3hUMH+IU^RtSr$m`PV=sKrGt2-4@az&GUy8f74=3^ZX2VQEinjjA>*#X+U`%58GHiudok!*5YVkZF- z5j;7$=UD>2<#HkjiQHXo-Eg#GdC`&}x~ZaGY+L@iM2Le^w5+^WWc+HG1<6~KB1R{D z{2D})?6Jx~98GN$C!NW_X`>ANy(6V=@_5n|jf3kfPRD&zop0&3Uo#Wi3%m5L=$ueq z|MVDR@=X(!EaTWm*tRI@$k5~(e=^u*Kh%bgM7z@Oh3~xx`~H6ECX=7H{moKQpP$Ia z?(srach{I>SS9g=FG#X_t6ZC?bX7`TAH{GMr$ZN8@#!fY4~vxa`WE{}FkS4uLRDX- zy3FM@g2M}SFynArX218j&F1&X9}NI?dVLGT&m%ACSlB>az}!0vq0Ya*G!-%!Ph_es z7Pq>3r(AakR>vIRp#;iA%-n7G!N2wpa9JfabuF4i{k``>)$L zJ{t>H{bzMu9RLdSm&ch%YNZ)SbD%}^&$t4clanQ}5*&sn>IguiWSAc~9cPGop|5va1czJ<(kvjlshL1;JqMH z8dn}P-*pFOygUw`OaZkn?kqJr!QOGo8;Kd5^&20!yaK+iU-cQ2A9_(6ed#tlCSLC2-y+dXHSpibQ*joFh_;zJ#ohF< z?UElqD_1^Z_siVFh2c3|G>Y2RJ)+K#6+PCoHh*Bdq$D&>k)e^~63pOq@;gobVNvzE z(w$ty;m*i$mm+jM`N(FrP;`nUxlf5wJ#_76kD2buqQs4hc|QDoEB~R!be8x2u(BwY zo>O$S&8F8KFq{0j#<`V^FiRGLWN{ZO-{mk*oXP9s4tu4U_>JTFpW z>>=ZZiL>c8dQqt^Qf45H`OjfMi-_4wLVJ2ddt4#n&{PTwKi|n2JupA-1zv3*{@avx zHbA@kGPCffl^$wsjGFB8VkjlGst0KN6^VM8n8%vanMZ!5MIMeN@!fc3@3z#e6mIIH z7m@TdFGEyg`B^CJ=;z*%FB#m})iDj*xU$ELv&gkEO*#~e_YuEt2H=U^{X}E}j*@ya zmW1T=@e_JK>FEK^MF5q+V)m){=%DV|_7 z&$_?Bq*oslvU2ROYjN!ha)8@*o8zJko#7*bAND6^6mLX;qbMr0P`l#v&PSO5Z#AKhcp!HeY-#pv~|iQ>4AtKafdp zKL{F@2-U+MKo*;!QVXj;Lc)gZ2|mck;YT`|gb@oSynWEJH*A;VZmp{LRE-aVkr&g* zT0++=O4MCB#>+NIRq;RJ{Ao6xgS@7#|IDbwyt2rc2o|E+DCHbu2-pq};p-kWZWTRgJ|9Lyrn_mP%l6dJO zJ3w}E0LYdL6tChZvkz30#rdP7Og6f>&Nq^i42N$5BY*Dq zN;odm&g4rqb)G;Iya{ZB{Ucz=h^-8h-vlg%dChZ&lBfO^8FP#Tlf(ifSYlOXK0iW+ zRY^od8d$t~ebWabmxGXliNp>@zWz1{beTV;c+b821;w*9m*0 z%(7TUDAE-vVAQMI1KQGhoEVe<9jum(ooBIb%J`l+GFfSCVln

    zVWa2U4H!Ap4ja})_?xc0HYkEz%lE*L;5%^?nk=J89>GqTx9zJ_Xnt%Bc;{QB(CX-0V)2(%#TygmuV zzci1;3Z=JsNjw10VD%chr*S^Di+-t{_uFj1Yl$m@IZUuA)_7fbz#-_z^%lQ@vG!{38&9qw3LDI>;X#8yEI|1Mr)935iSLDQiiTyp1TIU459bOy zUQf?DB3(rT1$ia0sZ#Dz6SJVyPNn-PDL}|>w@O5!!;2^ClBPIolDWOa#ZZj=xc$&W z=eFZH0Oh&MUWwP2L?usywzs=fSJkj4J(1b_pgIS3gg3b5X<^=F>L6@a?^w z%GI!F_^wr%-h4?rUkK&3KiBho*=e$IfD6%+slYQxy5Qc2c z^xi2$tSL*;Uxh9|_!v6Ulq*JJ(KRU9nzgl3R5F=v4X3Rgr5?IYPr5HiI<`I9RBGM0 z5PYhhO8nRw3KGCB+g{>HUl8(IHfk4drfDi2-z#<1X|)&I7pYdI@=a>nfClgh=pVsi zN-k{=g4||vK&K0w<{4r$zuTwg1<;ZA8WD; z^}j6H&So)L95N<)zMy`ITyJ{srK%F|^9HZsGNYZvdY`o1+4!?mC)`Nt11L8*-3w-# z_6z{d=EMcQ)U3yu$Em~kWs`tLk3x5$EP1r(!Awi&4p1YMF_|EcR{#BO{V5?~I-l$X zjODudB?(DbL>jqFf=^Pqj2cZ_-^&kMlyzwiqa=4kzq}3<>9mTtU9f+sP`-RSPS%vgRxWJYd=~-4Z(QduwQ=&BE z>&|f;e=4At`+0i}-k_|^jdP@6!u?!dENO-H2I|6o&0@XIQ`0WpO}gQ7>CBy+;H%a} zg0UYHn+zugRl{%3`O~l4h3~(4F&24dzQ3|`255Xx{4qGr`^0EWxt}_gpn&eqZEDY+ zXbb>36hqc9dr7Wq=oCA{2wqyAP~0^?uCTSaa+}L(&9YWQJF=6DgiExa1>?Li(B=^a zCGVr#*VpD7wt2i3+9Y%iZn4_Z3*5EwDw4EVZIQb=K0=4rPn+*{#cyc+v83)+RAK(C zrqsCa?te-!Unz28v!nKXvvzXGdOO3-vfd^T$T()=pRUxdI`pqC!v9*a;g42;-D&K+ zy^xy4m90(Dcqz3in0Bd2?+y$Mjy1jY6K3#5h3Mek5j1>49B%3iWk3ZDmEc-w8C@(W zU$|&matT~&11~5=?rc~VN5p9ytVcJP48s0=Gm7ZGulE=h3*P101eId1WR5Ex%S$Wn zysG-cdVIKGFLmE*1+Np3oV3(`co<=DGjBNh8ggUMs+wJTMGfVtGjJygN98m%m~6=c?dQvbX2Ja}6Y8D(>EP({Y}Mr>Gh4ox|H-*jRuQt32{SzL$}IzkV~s5*=QqwKJQ! z)R3Y!in3!bq=c;$$NeqBcH`N=xb!Bcr2^l@=9DjHI#q%|VtCiYQ%8!-Ql`>SZn<{& zf`&o+o=xW6z`>i=gAFVmTRnszpP!X?3l5$}JJ=HK5_%(b)~opEWSj z5(MIrM8wM51&QL`I-BT{PnmwpUfepcTj$EB^J)gNCaI`P>!ycHS1Shw!2&_kgLdnh zo;(yEvlPsc@7pQ3kU1H+-(3**sGO2k8{Jr$GxdBIeD+(yJLQ2s54syS{95we-Fnj? z-s2n|)bCx1D( zR#`WlQj$(3YF3<}DfJyNOt#21DvtYfqe-e3@sHrQMZR@Lu~MV|P+>Vig!*#RvtiTs z@|e*`dE_Yn2EfSw{!FUIP>riTvo~vSY1=l%K5}PHz^Of(6>~s8zLHxxzWSd;mbFD= z1OdV7E#8g)|76UpyH@jUFC%0=*W)av4Kv_6*O@^=%pzPnug$^&)L+2Xd22NV0pJb1 z4L~k{mymSc^Z~E6y^(r_P|UI4X;+N1U`$dyus40L6xtB6pahdZq`lbwWn7(kvGv4y zrW1g5^a_QO%EJ3l0rMqz5gH7BgTFm1+ai#==-i=6TEY?JOKYC#y@eYQUhz@|xE{ueS%n1S#wYllrbcUiQ+ikKiJzg(yy^8C1L45LOXMVWATx&0f)M#h^28_r=54 zng!ktorIBv;g)TaHl1Kq=-%d|Ri+0qYADG|gUI$L#0qin#fWE6o8tOFdL`v%UCL-p z+u7ZtaU($jA32I1NqD8Knn>SIKh(o>X^k$#nzuTC7U}&(J-^JyYI~A zbD4WrG!Bf%r!xDw8FE{IrJs2+pL2NhK!{C}e=j#xJEmPHWgCi-TCr_xcebA>Z;#q^ zS*_FKFs49Ta^r|6{RcfFe#qldOj%*$kUt&lYaSgFt~^~3o*X;ORJM9Jwy7m&Bv!$G z)KAbFP$>H(1U)chpgfD5{Q1EnFKTO)AA=L}zgS%E^tno1sbMKGf|{=TbrY{fkI$ia z5U&glJ@J{K1urgjAZ%fbl>YL^A8WH$1GQQm>dTs~XVfJE)5ErIGuG82^0$j2oA?f@ zbR&WW+U9pBAL!If5t`38=fek=LE+-McQ=PuXX4K!gfc58Pjzk^U3=bv`TT7n|r4x zdq9}>*6N%@?;&nccJcdH_9wuasU3L-j2WUQnO@gFoUHiq)gD;(3^lg|82ckR_xtjH zjN{`+aeU&9cjdq1IPqA9PpPq}>$z(@qqOoE6mS#;v_YQ$16VA3^c7*3K3ItdizPc zv{PFr+s1H(MiAU7R;!_x7icZXPAVV5hcseZ_d0LL6#udc!|FJK8^S-ZYvF|@8ZXjZ z6E%%Gm9rPY^2@q@XrdW>W<+8^*<{PDm)>QNGsDg@WG`A&o@*MLX(M^8*2M2T57>A# zt~O);OK^m-SXEKySJ&2<5V+5xTe2)6@9w2EHR$GW-3a^e)h+$E+2?<4HzbyZg$P8ysN~O~W-i zOLaBfj;FLY_a{KdO1w7~@-lYfp8Jnha@<>O*!3w9Lbyn5vIvbN6&7&(+~TNS=8zCM zWwkU3>1M#AT?rebx`Bp^W0*oBs(^QJ43epYEmc%Sb@o2u%jbktDl6#qxR@Ghet^7j&d@P zb6y#p5XzSg42n$?HvI^va9l;fH;Pi&c!+|Je$M4-K60R(pcagnhLlB)ZvaiGEG*Gi z;&X&laiX>_2p2a?L?b`SoY+~Y%dv3Ud(T~Lu)PBnoDl3Wjl(cDe1zO)p?P-{ccmlpYbuVQ; z;c&j2DU(Mxg3_AX_vFGKy08TY@&v<3bgk}&xYv<8o<4tE!naIYPDi;M^O&hRY%+-O zZ&uRG_Y|&lLyp$?V|Jz>ftPJQKatCd1^;Nr*v}K>=Cq3^!Z&)oiT?4&4M@UgEIFLe z=3(A^yY0S;(;PfY&9*Tevn~@V!F{iVE-kf)9ZYA&y>k2184@Sr#gRvO;NOm9%L1ru z@FA|G)XKQiCdXPWvhDzEW~5)RA^o2_Yz36NDbjetV<6|;Jv=-0R&($|{N?D_2*Q*k`Eyw>AEQC~00J_6lY*>OGSO`S9>?@~(CZ7<0A)>G@B&u-g= z{>&7xiu53C9QZYjgbN*~vZ z;;L@3)89HR+Z$EC3`sdX5TmR=3x6f)PRa&aVbfbXuLJ&x@x$(U@N)piQv9#j8?rWFaS9t%qC5>X-2PlkU@4}P zf{ec$u|Xh)4}QRw66}KPjfU!!o_B|agOd-KiO4su20U5MX2c~XqES~Z0?7ji$GXwNeU_!7xe4 z;0L63irT?W95*$u=V~J-+6RfK=Wq|!z&#IQWHyw7!ouGQ#-?yz6)fOOq~|8Uf4 z?w`?bD-6NJO??Ui?JQfN3&dg1BYurwh$Wt7Ankb`!TFR3#sy%O@_E6CGI>h|Nh`=f zd-FfZbSqyz;=jp^{j~m9X{d3M1t(_+Lg0THJHk?R+qt2s43j;+FZJ3y1Uw!ExG;Yk zl%jnNuhRL|+9~3_nK?=?_S^GeokhOa=PE&6%V3{9>7F&ji+caAJ;{^8q5*=4M4oRK5^SqKUGMq0F0TF;Lf78? z_L5fvALJGKLN=!J0bFw}CLILr&~0%0#3x%UzI0Y^ZPH%Eb7=hUeeq$LP7b%J_Oc~t z0?uxJEec~|CZ{hT#lt0^t##R?+(fEKm+T8UhPVtACQ5w#C5zm-s2lrz4`g|*%Hv!3 zS-(_aQ!{9Ep6Wi|KR^o=1&8h@%};u!@mBy)Jzif1x9G-VStFw_c;hlHV}MIOH;4Aw z5w6FF!&`%cI5YrKfm3-q&Cp#Y5#K-M!K`1kcj9VHKf*dUkMnFtWPOS$FrMO70F`M>h6g?-b*L!0 z3N@!iNQGwa7)#FTW$2JeqiOts#Wn6LV$jZB4D~!~LWrcG??_qU5D_?|@zb9VsVcAJSW5c*51)ck#@VUU8;gU0JY5 zP(b~=eirx)rwr=WKB&0ZVVmHNJh#>bxhgjrN3#k~ z%%`vC>IQZ-M3`!aK%0kN*{0!e+z6|xa2vT;tlT<%-o_k|qfwztSI z5;Pt<=hd6v>S6N=?#K~))Lb0YkPt~5>RX@dmcNyv%eY5wA@`6yU-N|*T&HA)%wzr= zfmDVG6m$(-zMx+XyZT9<%Yg8LC->vM!8fq?q(-w=gcififFtJ6ZK=r7%)9}7W~ju@ zdR8&RA)&LU=Mt&V?7oP8xQR{oIYWN9G5CFF=!{`K?+H+K=&JC#-GGp`nayT;B@J3e zdK&yQ$a*t=L>kn|1PU%C73YPkrB>%aq`8b(p{B_KT z<@dwpC|N^L84!%)nyE4Ya)2)sxCy;2`owi+^;7a%vlC|R&UGr`0uH;+N1I|%(&K$P zbvF{PIMd>>vG_%3uQ;?ib#uthw`9{tV)OqIK|g(;Y1E^!@0`oLt}SAHrcA0dhkYBb zlM4w0#awjH3ZrNwzH#~VB`2M{sGP3*e2FS29*gfePflD&32(>eY}xJ$`yF z;r$>vBLso;%Gnql9nIq-evU7id0q76EP2Eu6W`7qrStK-O`8j#`NPD6w*aqa+I2(c z?WR_?ITAQO9c0zc6xX6KfUn@kF?Squ2g#Iu=wXc@rm!k|H9iU@an<8mv9k4TnbA6 z0w|LkrSMhoZpsUk3iB?d=cP1Vjc!kyZpqo zV`V?d&xi~DT=!micZ-pJ zL5^ymbTxP^6Zc_vX2g8&m6B!p8a@I>4AKX3YAH_OWdH^d;226uKaStE;lg5N2hVIB zgPP2w4#7Rz9;*Yy4nUWytg}I;S8DdRepp=M`Wi0ncPDOkTQ`=KZeq_H3zTEZo(E}( z0X9qGlr3NJ?q+)C=^N3mcyvpm9#G&{V?!>|(&-&m?9<&gK$-S8-fPu@5siP#k8bY> zJ}a@swb9dCBW6TsJPQlX3}cT|zzz;;&@-lj>v}nj1Z;-5O|#!WHJyJ0+d>wdfCnSR#5T&y2!{L^MiH*$tu$O~JIkk;WV)7|333fcP^ zaAW{1ceR`4V8qY_u+H?Wr~|ryV(II*O4v<<6h6Boo1n?uey<(OshR2g2V+TWA#aP6 zhQAL8&i7GoL@oZd(({pgOd3yWEm2tUOY5e*h z_Fev&%enR0YB{?RKL25*@%GWKt-_1Ce-kf##aPYsW_?dqjHw)#xK%?L@60QA$H!{hQlPEk>21#q|RnQ4APoRvS9k+V{jJ^Uc8l* z#xhJ1#`XlPTq1Yocs^IF_Vf2)o+!ky>dan7w-v3A~7d zxpNMYkobr^@jfu*xt)>1{&t zD;#%+NV?HECLpOuI*TCqfwPN!>2Iqp`;hITrJ>25vUXP(ug{;0xbieYX{fJ5#i6|{ zmWcrf^S&#$<+;cW8!VFGsacReo$v z=V{#O_5Cd9f*Mi_ClMk>jB(=eXoq@A5NK$qh`EYOH(M8lNluwvY z-beiGhsnq~91j4w{_XiUpvK~v_H&u+qvks4{xM$CFC}v^9T8F$ObL-B&=3zg+w7~` zw4`&Ip;uy6pO#~iQ4kEv@M$3ZSgdeqAZ-JKo(a&6L^turalmIyM~vQ&?hzn05RDK> z1sOs@--%Z^LRR<+*J~_sjiwr~#Zk!3LgS@dM~d!9=Ycn^Z75kr$$}RSd^22QxLeZGKqnaCvSs7V4I~ zY=zK2^)556CB7RwP5j7=E39nYd(|J6xHX^HB-=eu-%9pXmU$r3apd7PBl}B4m8`O0 zk*%meZc{}JiruzagKzzAJ^h~8-HFL=^vXGoO&J{}G0;}D2g9YhEtogcZU=5@y9(yj zFF&;~Ht((g?jZ=(A|Ku>KEbb1aros1qam`|=}Y*hL#pBYfo?x--l2Q*Dn7V6 zeM_9w_XcL$@(qSCOd(!U2j9u(=1+2)ujLkhn-u-O;X9IhWaOnf;QixYh1a2zWb~}j zZ5V?T14NFRm|0U`ynMg$1@?8=Wo`^F%)j(RRiI@4j_DU8`wQtScxPLyJy|))F!0az z=z8f3lFvEv%owi6pyx^3w1g(p8eVAn4*=JG!X4v#P7#}f(kE)kUg@BN&dK!Gf#R6q z!%3)en(7A0#vsH)Z*x9yasE;W=s3W?g|}t09=>3uyq{Lqo%gf1e9R*A*)}}9k6Y25C4ECyqcJAzhbVp>BDM-VP$lYb@7D7YHQ3ChXk#UjoUAY25(WFL8bWwK?-bz{~nEN!7gw;)KiZZ~wRA@P0b`<30hrAy;s zt#WnHQPTUkGTD_OO=94mS7wuK3o+~xa?@Q9lM$#Ht{jtkYza~1Cmqr<@xW#HAAf=s zMYyI}+c0yyqO?MMYY3(;5w%CJRL3{AK6nWtZ+5x+qcsXMl%@3i1cic}K`B+TPel&@ ze6xDDw>K5Xv#5uV*Os1s)_K>a+4jkucWGKujzbRnZq|8fgE{NjAq(fKO9^%LQ~CW& zG-j;Rb93(OxQbFwBJeUw$3HGp2g7;QZbD{tFD9X3s@sx1CPVx$0QMw&03d6?1l@_u zbh+VC_rA{x!g-iX7k2r{Hpk*Wt?HPv?$zWiiaL>H@LjzdG45TsR34wbOIKw1x1!4N zsUoDz)k*sVV*h+aNaJ5jcyrP>Q;Sp!u)|c2imlnPA%p%{Nyg`kw*?qH$$iJfHklLy zv#B-M<*_7P(M^Tb@LLQPtH^(9imVdDlz;!HW%a_NPJ|HwOZ?YtW9<PT8%TER-5C za|!8=j$0p#Bx2&9qVSTZC0FJMBV*z+lN89bsnCCKW&>i=)J`r-_m+C#bP|ef7#wsTdD_ zZ83`cMs!WHOduKPaS0|U%Hcs#q{c0I-M@T@y$Ak{H=0kNk5XhOdrdds8$_NpwSL;e zifUKchmnK9O3(209WU$rXA0y$twQ_Ef`dTy>W^T__k%pUz(D4Q1d-$>OzG~WS=m=3 zB4#h36*ltUp}>mV9M|ANEvx8(zB>LkbFTMLxg({pcv%+{MO!0|0KZBZa4!BSbBLWL zuskZ+QSrR{BkT2%kL^5Y2z_&Bf&4B9^>ALaag*ICr>0$IsnGNrFJ!KG3w4Pe=qiTJ zttTD#f?f#KAV1VCiTHodB4TwZ!cMneu@9LDrmOa+ZwG9}3pu{hIZp-xR#wF%=I&;v`-8S|QnfONz1Ov)x0Zicqge*h z0BlCRe?%Y}x5k4YH5#U*piik*A*O*$b3QeZQ0} zyx25^a43eJjN;JY-m+bwsqih-AMfMVH{>ek^%*i+K9Zv6#r^45d{%(WN*4-0f3Uwj z>9pyv(Rq#N@PwZ(ZB*X5!#O zm-tqV&mNvYyZYS)*(fFvCci}vU%xn)xp{u$7DTT*Y9l#)nf?*W^hW5!*_>d0v!*PNpVN&cg4zQ6O_}#&Y%lR z>3^o5jL&DjCq&^a>G1?)K`8Tg(b7Q6SjhNmR^-7VJYN6p9eGp6^nc_99rDHq0uHl( zZIArx$l^2$3SuY(7FPM9hEik!5C~sP0$?~2vGZD8a(df_0@Kk9>LYnEwZCuVva5aq zOx_6fBQ)lzNipwLP<^M{?8`6;|BDX}(Vcg-^%%)0?&@jCpuyRt<64cL!=;NI z7&G1ZBmderyIxD{ji&lOQvK~@+bl6557szgzRR<<9f7>Lvk$}8C&$-peReE`uONvp z?kiErI;u)QL0}~LQ-8nB*oBPN^_!7-))b+-7Q86%0i|9H4uo!ym@}q4_dGt3JG4*d zH2S0*lmOI+usnC!JXXaHK(1A6{Q!v!}0D|eW!D4$2R>f z%N9=;UJGN_X1poED#LG!9M0$aDjqIJR;1;`ja9JtP{1vM(~qoB{AgjiIaloCk({`B z6_%@rmkr+&ozTT;L9G$-YpgsrpmdI2_lS*kv5nl;o;G-OGclw%bsQ8{EB}FzIdqAR z{i)ghofy((rPZla%W_?_h2>!EF6QTb;o)#8sh5zMj~$TBwxJ8>ton2|dU&TNLl+(= zy;jSAdJ?&$GudmcIi}^Q$P1o>?{WNWN4#tmUaYt|8P0p%Ukb-Z90Z32#wh1)Z|5RQ zwZBXA@@)Y9@-=)F^EZopw@ml4;?}aQSG1j56>V~b=F@8}P{B?v@6bBJE$xt+J4T&u zME{X4_Ht`VJH~P|d-8na!n)bSnv*EDfRLNsyp!dNDTc_22Vsy19dtnOYmQ~mby=Q;^rW~bb zqcg0!j$eGvw6LkDY3j*w(V7ci;MYWB6nqQB1B~lQQz%Z4*oUeHRs)G*=@@P; z*av$!FLVREfiCq8HA~Ns4=_h!Kh_p0?YwCy2gL=loWej$cE2Jxh7)?R|Gs^ft3Ny^%}*MY8W`y1 zld@mgd>b7A#)eOS1|fI4tQMaXRJMyI@=&Lv6?Ttw(h>YTTTNFTG(T(CFyeSa7lW0z zc66^gk+Ew@$>GjM^F=#We;(RV{HgCEXSpU*n)N1AEv{sVh9inn=AN{M1l$(Ye%O{# zBNse!hijW+G1*?7oEhPlShT4x9txNJK1e(epa(hq6jKiPskKpZb{qL0Hh9EX0k=Uq z0Lg98zfPSl13*uK#wcjUO@ow=3Hm5aSe{41;del0X>%#|rniX%&smKMhjXMEoG3l>ItBDBeOx~! z`So^t*KstvL3e@v79VCd8Iis^&pF*H1_ko09))&30rJcX z4;b~qRtZ*EG4LaqI8EbosP9YLi*MhRYH_(lc13Aj#!F5>neXaWJ}z2tn<~>s?VamB zdVS#hqKPwy#dVw-YkN4Z;e?0ckl256HXKP~ zxn-N~J*axVEtmH}StUWQfDnwG(cS!h-3?jDua4tcHLFxLALsX`puNbCtX27mLr zJx0kcWrm(hUD8nAX;Pm|^;mc;{&79CL{!7)oX!-+C!|M7$HnVl&B3wLTI4$6tGEDhHtWHEOxHJ#()FwK75{&y>j0qS zW_kn^PkL&m=X!=JU_iKp5*(Y0(?_L@g)Shx#-J|-Z-x&Q!Dn57sO{^MYBRH^pE>v| z%vP?JRw~U_t+xB!Qliv}#w%x}aU+po0t;l09G zY4DsEn4Eg9{wl4{zar`Wv~k($u^nc_{&t6b4%PdHV1Ob1lRYCNT}Zg{=|knD*~oEF(g~hb+0KUQ6*jUT)_Po8kQm ztVApLeTSt8q$ZIYt|Ph3cJ-6WO5!T8|jtX z*)=N6C)KM${%~0PN1uY4JN~km99nRuM3&#TazbC$ky!^D%mH19N?7#=H_q?%=JtT8 zPsVR+=m=(obotEi{@Ak=Ze@7ID)?vM&F>^|?2Y(Dl%?VG8)%k;+>v{3&mTluJ39+K z9bY#+6F^>`ug}SsMf-?pHJj|qQGZqXZ%TW7+6Fb0+KKjDwrS3f_Kwgx+}n^R5*|n- z-|GWYQ@s}|%8RX-iDyJdT4d0#C-qxOwUkWOt#oTN?Yn)Iwe6SIjR6(zQS`-_#^&n( zyICP7{1;;Lq+A{TpEqJ2b<970%}5Sr3C;%Uxk6#Qq_j1Y^ohvbW#eBT6a3yMOU<6A z>vLSigN(1hxkxY0Ktmijcjoe#J89p0JWx_llS3N)r|{UXR$&@MXnXKmk+=ci@n5K_z*=74EO?nM_PC-=CF^R zSORTzz(lRwoH_6q0D+ts$iNPR?K?qiAXP9r`+@60n~Q^wjMm3=BSKS@kaL+I7r{n&mWf0YdnvzA+RN6FMPpnpSFGuL#bez1 zj~Rhc_=n!cKMMK{-E;Km2LEN?Mw&|O)ADHpXUT_6z4bwB9i3}~L-ld0?zs{p zW+Ks4#@$pJipRHLmjs&6RC^xhK|mQc9TY3w)gAat%^w4%ufQ1p(Q=IGW^<&rD92s9 zYA;(Bp_VH{a-N_c5gwv@?-5z1#Sf4_;6$BkTVSCQ@wK(m=3k5I#eU$(kp8n=u}oDeg(tLK&=mU){9?AO z!*j92W}edt4r1?I|HM%h2LYM*88+2O6X;>tYhH(@G>Jyy{Uz`1K#W-X=^)vuywN08lMTa)|u+;(A{VFp@41>?Y>P`n1;7e(M6BMqvxh9EXOC zK!p&uee$lO1Po6U63l*`j^a265Y_5EL2SNH>_xF{jP@zkUTT^~LYv^+pCG-#)C}=5 zih$x${m`PSJYKQ%?Z$5ug(S!nsJVK_S{;{JR|s zR6ZB0czDVg%QI!)gh=cXKdQauk!{Uz2|_GZe}~&5NbHCb!jXh?DwpOa)FJlh(Zcz| zX9(~pe$2Yjx#fup&y+hxWmT1zCcx+0P`VMH!80x2NnuCDoHn7k>ms$PVOshfFlgv> zPTH`cR)4V88>=9f7Sl+ujB48B*F%z?XX(T!Y##U)hsjE_UilXYa)*}A$DHdgb*M9U zo%?VAquN@J%4H4N)pCBD6*b5c_9#;P(Q~;azec9RS=``Dy{iz?1#W_Kr9QIh4nn_U zT-dnCU8z>v>A1A@y6DO6&jdCm!-F-sK-Amh27d>K$ROd_X zb<*te-C>6>#c7?H5d z^>L1p#LQ&kPW3YLe>YVE44<+!&+#vBz5mx_{VG`BUd!+moK*lIwca00Ca~Zmwa=%2 zGc(gZ1>`V8&HfYn#hg4QnF0Q%58nn#y16sCxuNB;`3DRWVj+Zohr|x*Qan4Is_RTB z23Jg8A$+{_8<>~43_in>!PPVq?B9<$3j6;0<*47SjLryLg;9j|+4Q1L52_qp)>S

    -NG^vr&lFncT31|8}vuaojK9sZ%$rc+(Qo!1FSv@=H z)Z>CVp*lQur{Q;8j{8AKk2KB2z!?_ZZ0_*)k2ClM|JtnB!gcA-Du}aT#G#FS3v`Ub zhI{|6sMtumA}lR*@p1gn$;i<;H4z+YX6aj>zDJ_NjUg5-eVl6@5?fYEAow$$`C^25 z`ncPIG%JSdMrdNYb4pRR-kZ}OgN!FYgzvHwBd2n0?i~Mcw7iE78@y?Kle-hCtx!^n zr8eu^5$9%!;-NH5QKY!-TZ}3Mp4H1$t3;kN4al9E1##Md>BTO}1MDf^G*TE!vVSl? z`7K}%JD$0xW^*oTGH6e1I{B8~Fq)4~U?Ug1Kb~7D4H9RY+kBtd+kKAVt zHX4~f?u3;8}D@80s;=#a#|&D(AFNZgDqoY`8_bj4^VyU|aIdaE1_#a)!Z4 z-=J(AI6>G=ZItpg1OUc65IiF(QN(FDxSs+sFm!5n;ei5JlDtmsBvVY9VCJO4naA9~ z!SpGh?+c{>hX^*8C4?7tK(db`d<j?>3xkNWeEgBfV)1T!CmepWZ!KR8YQoCd|N?+$z(C9q?0ES+2`+c3|P z0U5P?fzZgvL@NzJhxmeLe!MrjQD%W}Dt#!j87$5B`?z1hbo0_2p^Cc}CfF_2dO`x$w*jNWs`ke8*kPm>Dr zkE=N@6auowgZG@RvA_rC&s0xMs|VmQK%1(S`{mF&R%y)+VZ>AAL0k{+%X!-T4MS;JnoSvUhB25 zoz1W~JF%!;A57xm-f8jOxqB}3(iOE6dnv!S+e`IZYISbVlpiIp5aP$dJWHwrRtJRg zrAQ^lyC~v!HHkgifx7HS?=9_fy~wCPE-PK}-HLU^L%?THyS+4g(S!BxnqzM{t!tVM z``LPL!sCs^F|44kuZ(ZBNA^k_$!tfE@ zy{0JeN`=xUB$Rm4=nj4NcVMhty-JR{Yafoc%{1xWPyOFZ zQK3h|S?l<~uz&fAt{vj^JLO&ZGx~!mFJpl`{`jZTnjzNUfm1H^?!9?15mL=J*S}F7 zMTrJ2G9|Xj2-rQzivkgDMXb+k2P~sHZ)sbA>VY?h|B&gU5(XfSq%v1|YTmIK3D!JD zi#ls^^%!&wyw-VhMuamr1?2|c)xq+Pip6|~5375R-Du&gxb4(@71G?V)fU8FCjbb{ zZZnWP62d5LKCMNxmo3l9zy3 zTrs3{BSNPpG`CRjmR#`<0pMxhL~{A;n}V{ibYbexM@QS_767Q9v8yC4$BA|xnQ~dK zX0>7Wp56ner}Z+|gpCG#{rql?bU@Hq2eNT;&4Mk}J4LA%Xu-m!KY(VX##QT%Cd0bS zFQqbpY2J$?bL|jVidKirD(j7Y-DNzsJrH!Cl4N#<8)bk#d8tH=iDR>C*8}I^4XG$8 zTP>g+1>|4Bch#rqi52%!PJ@obpe8^|vx)n}3V~2Cz;h|(Z3$Ikx?2&cE#>mtA=)ec z`DS}wZcyin0ZeWny+3^CZ-s0AXFbx~lh@!%#5RX1_dyR=)GKFpjIwjw1%Q**y=8yF z4NB^W5M-p!5k@r9DGnM^@f1ER_FgQrh+2n6{c;xPx2oy9%l4%G>yU@KNQX`axm|Lk z8Ik}nV7%#9_s_bH?p5;i1L%jcct#NT@_WFc=*XR!Q(CK6bJx~QMdnos2f|!9g6ux{ zl-P7a^mqZw7t|UR8yM8n-}YucOlpLz#W=SsVYRyvWS7@USkQP09L+5o*_zTzyJ)xL zUXq^8#Dds%T7{xg^wVW-R2YT%zsK;;WB>fN&~C7OyHdOG)Hi_zF)qG^rMgND>|W$VL8v_T65n0;r+iGHFHPy z3f*VMmOcN4WS+r6>nBCxt8Rj#fk-K(SmED^;V7RlF^J)$;*!NkoFRhdxgMTt~y^;P#@?(-A>zO(&I7s zbZXp{pr0<*w*VrWq^i!}pn3dNDaEf@H7;T`O)P|iAv<4z-zk(&f4*99vad2!BxnJ+ zSq!F(V}4Pt*4l|lHa5^#Y}VVEtsEh|8|i`*11}X&Y*Q+YO!XMFkJV;d!%k73aBRfc zhme5VK62|T9moTJ4hJw`E)y0!_fj1NXo>%+$CNz>dWx4=|1o_vmIeU{vH1%GUPa-t_Zbvm&iR!9Q zD&qA~UtrZB>rtZ+coy@;^vbyb~35g&OzZcd}Ev z^w)O1h^E=!MPx;Lym@;W+q{GH5Wd%TAY+mhnE@MPyzs>2JaqV*`7sLhUo zOQ159hXpbbxv=+g*1rLT(W3HjQHI7OCBmhENhP7MesL!C`Og<+(tf=Az}WVoA#Q_4 ztH`&Xh97n+Xi&PDnGMb*We!~tK~wK)79I?V4}YB){q$$XF}oie-D`MK=l}ZgHpPJN z7_2o32mMdhceVr=u0WrO;Vsypse34H7#jQ#=Ib~oS4C61)5P7ufDyJ-_1BvkE@#HbFVY7;9) ztEH{BC~2%xwTcQVp>_~LwP=mBw52sF)Ye!L)G9S%Z;BW(BZ9l`bA7+}{k^W=`JMZm z{F{^W&ina#J=bHD>Yo|Wt_h4wy()l=2AoRlqv6hSZWo0eQcc;t{ya))GN>~Z^pC~+ zYkv;IW4m%)Gso}Z6U13((stf2cy?uK8IXntCE}+dnJV#mxAlJU$ZzVt z1O#u?S#sJ(*ch#})gR$dgPwSvnfiI@Tg_dy*!J?qsJLmp?{_Sq#cxwTideIs^j6<6 z)+#Gj+!(bH_ofIKC&)X`6dL(d6^as~f`_DVb_mz=a+}AXV6UP~;vZCc|kQ2Xu z>gIn>dd2ljVIB*C)_1Gh-aPoyY?#VpZmaiP&_NF?kuZWBc7M!gmz=>6pGZAC9jKMu zU9s}Ek#SDCqW;*Ux$+lp+ZG%9{g9tM%dBEGpY}ZzFBFYj)1MtW zaz?SNUn(7VbV>K8hcQ9Qt^VYBdnryN#5#$Jk$V2rS<|aXUy)FB>8@;`1)e#rG7K3Z zXe?bFj+<;jyI#1=HYcBuUF1{^&tZmg>9=9!Jp#wguKs~O+icybrHXja(JCHzu* zc(eXI+B&U6L(T4oq~60cw1svp&m1JgEdK%LuH(u(KkCa$kx-`i>sFA@yM)igQU4XI z;VqMqus4v3l4kn)g<2CX%{86DMXzu2FO_sn4{{s*f`!GMSccfL*G*T7R2m3gyH5+B zu*$7&y*{}%#Y#&(dtfUb;B1;U?*;-`cY)a&(ngV!>s{$cGto=geat{|7FlCA@)T$9NFNXpRVpArU$I&4ej0sJYwiw9#U|$ID55M_UwKXKQjJo z%YDZu7-MA+F|m^6I?Ls>1R~M`bSp8t#nv&82xw~Y`KUsxlRd30J`f!GTs7#GVpTRwECtk;I0B^F z-mFPvt&ZHW{dwO$Je3wUNLr_q#XcQ^*15O4eb z2eZE)r1HId!>AMBj+K#P6Io9_W_X5>{Zo~Cm5+DV?a^AI_IfJDHr{_Vm52p)hm5Y9 zp^;TaIu5ZJfpe!6+HKyQ0Vh{F1U+q(a({&sz=qW6S}*Fu zt415LI@6A4pbPj>88;

    ke(NHKKAg5%Vuone@^!)eGZV1ju=7CoMSHSG_$GDLf6a$!}u5j0(dz^1VY%gjE3bGaPu_ zxTObp2im#(_Z-e<@ql8_!a6n|$3HIcyG8AhOWSmn{usD;5$c+qNo=a`wA`hNGb^1^ zAl8sGRZh^KCVGf)r43U08vVXc>K`NjYkoC604aGsi-2aSW+Hm;hK*!a8(QO^j=JWb`pO?EzqjV9rg;3QDgta9}cqrIU;u#U^iF^xw0`{g&;sFdHk?tLk!j@4JA z@7G&YgJK^PJiJe76H=w^eTWkMO7TC;HCi1SG+~Al)TH@rbtJ;$m|a!3>-i6^U$z-A z4U!H-DhN%?<9ej704JWGm0@9=-w?-=MF^NM(H z08>i{I$hi~pO6f~0^g4D8j2GdLorSkEGHgjB#;<7OBR{51v;6#p86cN=$p-DHm+JVFJVp{eY`Aw%zcf;3eF3So_my zsC{O_G}^tz?x69X%A3D*XTaYEKg#U@zf;F5eVqGdhaGTQ52kLl_?0b62Q?o|ELZiw zZ0n0I9Ez{bCf9`2s{Z5ov3k$gZzl-?Q&AIdW<#x~^Mku)C3+C~Y(DmR#71Y98l*8g z3tVp))Ur!;d<9wW(!KVpAHvH15BaKv@M4-A>`Z7(Mo!@dq#;JZT+<1#>zUbNvbs9G z=RYQuQaXTJ4QB}w#J$v21=|1$Cz6zO(LRZC^6gCIxQDbJrn1!#CmoJ!v}7o7+PyLM z6t?FBMiwyorL8lz5*wo01sdXwtlpJFd00wS3d_hKmYA>|#o|887%276PmReQXYye2 zW0T4G(ICb~^Klu%@iO_alZR+NPrRL+5%WU`3IwRv#ZDzv5tpXlI%mpgH>Et8S*v&u z@CD&2LyJL>qa^cxWG5Vk*RP^i10~-9JHLaN?Pu&kcV5|E#LtN9u5JD{ldB%|xalfl z{hbFgy%y$9RZ3o6zqU`ili6?|z0)pK;W`a4p*&s9WJD7x1ji}ld(H1Ij^~#ywz4;< z4qxT!04@uou6l@OT>Qpuf(z_Re((IFEa|FhQIltw-{^Ycm9fSPxH9+Q2OGa?B#DDG zY1U2Gs6yd~YE~4=iW7M% zPg$lA7rPges)kBo{pnBak2r8XA2OjFI4Q1ub2>Poqw3+mk7KAu#MgnYSVuT-{q;!Tq&1<sY&U$R!))sw+53~?usPZeb>IOui=w+JYb#zUc^2Pf?Z#v!0@|af zU%&62?dO9(@6KO4?nkV9^9{-OP{TI4L3y7)jsJ=8)P?$g$U}o({0g1I*EoUqHS}`S zJkj8F>+Fd+(Zi-eK|-+QNsK>>?(P)Snq1j{7$CJUK4z{m*K%*~8~o{B9V&I@U=1R4 z3x3k(9|GR`1rp`oV4%#+v;h=zOAqvU^5WC z^zpu5eTW}$@JTo?eL7q(MT8rSa9xTlD{PDVrMQKTD+)nCIzq&)^=fcvv zOuFrl

    *^foX%Zy1o;H~Sz?VFnuV^J(tEH{bY(d1{RY^!rFHAFdu>an#lk`W?l zWh5wPc5IA=)&1t=Eun=|wzDbS?21$=38Ef#?(w|21VmCXe9|O5(2q&mB&3B2a_#8x zW1mq#-@aM1sjhY4>Phc=E!Afx#T&MzraLv6`I#a1S8h^D6~)hwSkO}n9x+(atI;u) z^h;@q`7kqke->NxO5N28@e@EWt#DQDR40$=Gac{w5O7uF!;w^Wh#Oi6%!ez}%lLDE zEpXM&J&f!(WE~aa!;$zk{pwFr;uw|IL&Wo|0$|y3&Uzf4spw67MOd{L3dPGaF4+z) zY=L5^KXt zNJI73YHFCntGYRB93mvEm9BH?$gDoWWCVIGAmTe~>^H`l2R$1gP%iFCY=?$wao{VB zu_nEuX87?GZ$Cc@TIP+-G+dvncx?U!6*6`ayo773D{@Cqxuy0QHg;mBk+=cl1v0PY zHDvyv3w#CpAh>^k6|cUxWR~M@J9xV|#H+WcsbHtV)7N7?6Ea3B!1JkF>MRlKG&eL{ zFi?DTfGn*{rYY!!(jIQ$$-Kg@5~WdmWCTa0hrP4TuP3UJ_-s-%%Nl0U0k{Qdg*UUcs^t zfhQ0QMLwC!``cX_{J;8mav3?xi>Y}#W|uOTqVosJFA^6zau0qH^Ua_xzyu0U$m%`m z=8W~dT*`9sC4^4QmO_>;Bqarb8ug^2ww{_&@LdTuczos(qzreX&t>hYrk(V>bpR*S{1r?jCe?|~%cqm(v zkoT=xiH6;YxJh`i)7ecED;4@?{F+-V-`PZrkBmdTK7}f8zEmLOn*2 zLk28MsNqomh*lrL(peq;5cS5%dVc)X&aH_T%Hk8Af;E+pY__h(1C{(%w{jV&mBr}H zHSQc-f7L7ZMbC)C$*rq(qz?{PtXFz8nR|PrTin@l#;s-4&-@osGm5ix$X`!sdjp@^ zL+`ac1GLp~WgLCyMSHQCfzMPsGt?Xu!aY4bhYz9$0-=wS??rMY;6k`&uIC~-;qs^d zlftwOIc=z0Kcqjtms(u=wCW_+!+FlJxXEsIgPotSTzakhJC-Fl{|_!FePdK_8bldW zr|{U?D2HQ}1rZ06Vguhv#pC3P(u>G{uI7GB^2@?&Z6>5r%0g<*&5BW*Yg!k3p#T@NHjXQ`@29N{W2= z5c3$2Sciuhy2dY8DmtH}T|B99-WlX|%Wol`p)b|DK&Oqu^I@cHPhVH*grZlKLm)ZX z3|u#@DZ~950O9^s)hk8cd0R^R$i>qf z4={|DY~2_!2LJ;HN(GGFBk2;>d!yf2wY}n>IfALJ3j&*byZj!Z$@0y#>uR-;>w;52 zwgXdvGb@++0@>s2ls|8$M7skMPO&`UvJ=ECs>myU;FbW#gyFX0*RZ4cfx5b>D4{0o zZwFz%?s%{Wy-veor{Ji5Y(KSYTD_@L32i`5Fr|@~1x*$VoX(X7z)APz;RD5+x%44> z`zp7XNj~`Q3Yta4es`$wJS^qdy#6U&9C`7edo;7@!}lSZyOOQ>lV~Y<9;Y?&<_gg2 zxOxu*Tg1kt8H%qS@XYDLXOamfvD6rG_DKysl6a06pS`v31fkue_i zKxh5arO?M~Cz#uc0aM_EA`cVn6Q$L+A#jH`hPL|z5sG7Cf@sLms-y7$D_|#-IXqHz z$MhAp?`qDblkZH>aza>lp$?*-IJvIlv?Td_TZNwp8x=62|AOA(*!*;e2UY4X3o;BaKc1UycXJh!{(V zHIl_F^_3i2;)6$3OvATRlLp`KG!@^TDp%hTO+Fzgc54V2Acr|)dz;Jfp{OLMOQRg1 z_~@HEFLtr{JSW^^A7Wjzyvf34IyuH>>(F{zYhm)45`Q7~A;pQ>RhL|&v7uLCPZ_5d z?|XF1W#KhaTY|jc=bEmiP8*J{J??(NtL{(dL5BENCM8ri?TN-ajQrKiA48&@9s5QP zn(;X`AuWB?TB5d9FJl34mU(p5)}wwF5BI``&u#vaXKq{AGIc5g!qDHJp|Cm-DcsOc8}js2bU)zaVEHsha}=15^cwBNCx2jLt}BePskqacoE}1) z{ONkD%ER65F4On1B%t~+SRo|BmG^#?H2XC6Lyjp8MI*nDtaA*fywDfW_@(4(JXSfZ zw^T#W2aHp4PH1l`32^#IpRCE^#HsmGYU@^lC)n$&Z|yJnb+4-Hb}y=N6>89|hF#+ut<_QvITab~w-5uDo{r&`Dh#2pGRB zTV3QtJR)&T@2E!+uLIQvY92vksn{8bfk@NizBZ{$8Rco^MvOxJT@SLXr_;J!gz7sn zFO0DmdsesZV4@&#>^W7hp~?IFuYO+g=vAp^Cf?xqPDcrQ(QCCCXQ zUbk2E#3nBrGYmIkI2#TphBI86qsu%cu63Qy+o;=XRYXG_|9!M_xPv?m1#_G==y++e z_sV`+?Q=oT=ygj=d(WUAYn;`iP1%m%!(JCn`TT@%#u$D?wLlJfFxPFP>ZO2d{@5^@ znBsBQx0<#l$1&q4YALbhXhJD*+xgP&eQ*Z&sdhjG?|&-KjjwjOWi!f^O_-OR4pqy% zE|1s6KO!m;g9~HPm^kSsogg2%nZh`6k*A}kIXqXKkzQN8 zhl_TeekIh^P<6G{eHgssGS1?A?*(IF1kD3kIwQEV&gCTck4!j08$$RwnsUXClIFJi z-e@-wd^Z`PDpw$0N&F|!*)dh4DcQ7pYPAS$X5Za5Y!NoMX&2&{MuNULHx>VcH^0d~ zJRhoWy>2@ORRtf!02Ss;CGf|?IeM!1KP<)chFz#koC(@hR0@=D*wTx@7QvQmUjhj| zAIgPYLR*qv;N2MW=Hb2VsQU?uFx>q)ZQliwPlXiZ8NK#4H3|?~7&cn*7RqDR`V-~t zH*J-RlXa_HH>%4J~9#G@WXr`q~Z@tagw^oV7|QAl$% zhw8kjo+I{d%izQwxOL@Axh^Ew3N;ynOSB9P52O{X*yU5?>t{MH6bu> znW`vwBq44Z;y-~Q$$H;BzMI}aaKEpNfMQvuXOwrY#2fo28uvfi!`l8X(qo~;FSzNv zM-SxCI7?}I){GbD05Cvsc|-y{cThJr23gZT=K&r!12ub<#a&$x&9QPE5TZ=(H%U)b zq%-(fQJ1-~pNB`4$0RjwMlA;1*uD>l?@r7_F4^QwsSjxD=Y=nc@vYQyecsPT2-JA9 z8Bz@*nnf+7$JeR|y8R`V$wtYBk-x--aaLXl0<{Ah#buGCaHMQm4 z-c(*=ZNMZ#7r;`{o5?x_SX_R>m9DEdZ#TNOMq2rtcfu_EtyI52^6q;0@h}kHvC(`k z(XCoPpDb)njlaUzyL9;O4)ezYm#fx}6qxeG#+~MO`A_mKFBSlnEU2G8nUqLUj8%)J z20Q^vqwhNwS=8&SODWWjk#zf`d0t0>_9aCWzp9?g+LwsXtV``aIG8KNiRWVgG85D&8K__dqig^BBfprWN>_Kme>)MX{B{atdmow9hn&4CFEhTF z_m9ylW~avJ+Vlj^wR#bm>L2FnlvMEY78dCo(Q~^YUdCU;yxa5vv1V9)uqJJTaNO3` z<}2N_n6i~iQ!gstPEeXrrj01eTvk0FZG8jg>x`!RbAK3|ds|H}gb{x++)^uef@5C5 zS43Jr;g`;v?C6?f?kde-2WJ?WLE`d%OP*9CrRn*{xE?<)nNW^DG;nS;a`Kv@W>BFH ztO}F4tZ(ZZEA18yxn86(?s~^WmuKIJS&$DH4%V7z8Svj07q03P4;%$pZr%GTK}Yw# zXraBCJ&^%LLiW1_>}XYPUQv1{fp?QS0$0x=``B}IK9IF#RIyo?G;S9%hcQToa+Pj- z^oPGLHu`48kJJIvD4`3@kQk+idS#_P)>jIV9Y_&3-;@NojFmK}&ivGt-KC6arP!I5 zV|YGJd-Bt4wQ3e& z>}Tu><#m!R4a#3I3<-UY(j2XF^SPbHbNu2E)jx$_paX=b4eb6Jd88+zixnUhYQfl( zYU2a7QQ@{iJsc>;MPXa7^{eap!8@N~-Wy*Z?c84FLEqW>BF&g%L6~R^zmAX->!e09 zLv5pz!dDUDd*o4bT+2ldN2hI$p=)QZ@-eO<`&Y}6P*J{#kKIQfnc)@RsAOS1@oU#g z6{Ia<-Xf^tGlJifNBs3k4~3)rh4w#Z{rcJU7mZF0SaP{ddU+!DJGRm}O7m*ww31nI zK|^};ckhFF$5bkcw!4rFzDN0!3F#_PIK${$f0t|NopeA>WT9=jrMzwAdkb!vx%WXp zVDhKlGR|dJiwHr*i&fgBz@aNUeaxCqKf3wF#`I+^r?sH+E6dbw5;CF-xg)@rJ7D28=`P{ z93_s2mn+XWe&xW4aEY@#0&G{hp%*1ZG?KiHfPqtTH!(V zuVZ3>tzFo%AFBZ;93Sb+@hMCiciNYl?;=Kq^nAZ6%o2UaMoYZ+W+P6|oNTPK#Id9T ztq1voU8Tlx>bbOX`{RGVcK>ascT^sy?OEV&FvA(~(hPXT;LBA)VGCUlwf}C^ojZKt z%0FY#@XQ#G$>J7MCVt1!eBNYefz3#G<%I)QoMa!r>Q*AK z`uPx174T(0n$@N?6{bo_Sao-;T`ltT|tS&HWJ2FFJbZ~RDpP)Lja2Bzy2?N3-FA@Wlxg?T0n{`)ZPGh&)BdYP#J zJEc6-5iIn+!pWmIhyKj=ZfxBvp8}b zP&S=r7W?R>H0C!~(FK_a%=S;AaOya-P?#2P_E+A(eJ!tOozuRo@cpMrh*2Ft?2>>- z75&X5<3`Sr52M6Z0^rzl7>cxe!4Jz#Z&v-I`8&ZKO4CwyXnKD8N?N}E zQfyuv`gpoWZfR8JA1b`UP1T!HoB>4QC!n)lgSX{=d61^3Ud9|68Fd#1#UY#$8!ifX zqfiPBBS?1xC&!a+)ZN%6SzG-Neq%Ob_Q&-1!0#%nWr_x+ z6<2ng1)4Y5@5*>u6p@{mUw{Mi5AJ~=Z#BP(@+iKKQ|}qp&>Vh`U_<#6j;OF&NLts` zbFGvl^VMJ3A5TYh^wn&tPZlfN&&8steC3$`nYn=PTK$FM`??o18<7HeXXd?L zDrtu1-Uk*NMa*J$@YD={T{8VuL;_( zV4JB#o`ByVqo69wYRf7B(3+o|I;me@!$m{jQztltUd?bAze$c041^cNxY-s~^yLVB zsNa8GgRU&+@yvmO($f=Vn5zvv`jRemH&koo6zKXs1NOaVdgrFC9Q=ZHGIyMI_FA>Q zU$^c^E7_$VPQ@H`!cM&VS$WjG5>JK4O?j%}Xji%ZCDZe^vOQzx{q!Cfk*1@N4J%mv z3@n_r43c-zrj~O(Ayk_mxkp@Wva9auwG-z(NZcPiL zeCqPlYs1(xce!GTpHL!!(M%;Ng6;QKJju_SXOaGC!xcmA8CYd^(}Q#`H~_W zsisI;#t0VM&yNx~Ct*q1yru!v>2>r$K`fjqd90n!flAW`qACLmQvJ@`ujK_qJt_`X zC{^0*dcl$dc)Oa^e1Q-(mBE*-O0cat^lLqSn^ue{_WJvGPQBU-^Y)e@K3uy@ zUKpgqXuEXZ^(Nq6s2L2CHN zPUU^FM(E11C#))dYdewELV@rIHt;JTFpEjR`IsUdPP#9S#njLr{+Wjk?pG4yQi2zb zvzeO5L;Sr`xtsJr#FZjj4Rxy6(4w))#Nv_^qpRB@RPy)QrxUOAfy{6haY7QbDX7A9 z6(6{tCf^&l8((C8WL%q4<|CLoQ>L^RC=XyXnnK3U`Cn8ih}$42i5~gAuejS1^dz|O zp8aa9dR;1N6&N!9udGv5x18XdunyyY8TnSX*X<6o zGO12>1{CeRU;FvXU3K2o0Xsrj<+OVOS%wUS8UMiSILi{+i`cThw!3339H88dxqC?( zXWA4R4qJ}j9MCT?x)7OaYtU}Z>UzV_aa>`{!OA+Hj!SK)KLj1riKVn_29D^}2Z#nVvI%CBTel2o z^Y-i~oygo#|R8(L%|mV{=@*i3uc` z<+xP*0jbsNzDU1eV(e<}d4H8#40B}1^U%b1#;pw$_SamQi+lnT>`1nFHbQ)!0IzJ( z_5G2ASo-oqVP;j4p{S+IlU`8z1o`Rl&>a)iyy4jwDfxBpm;6AEHRbj>Qi0?-sQ0g! z^65<1!3Mh4>7-Mc=T3n@nCn1i8>?8iKWDFx{~Ua(RdA)|S44)^&YcW;=5m5&aye7t zr}hvcE^&?Sx#1o@;$|+6=`pD`cgO@WUw9{|Y_wCCT#(bZZ9U5v{dQ2t&jCr#+`h>~QGUikj%luAYKy<1{)fCvCsJ?5cyF5>aC7 z-*dsg^W}dvZ4}g9Gc8wd6Lko;j#$z-E{UiPA!f@iGfMOrIfrPeIm6tR!})Z<;?z-) zfXS&~q5|6#dl6wF>zQpd(HCIkS9mY17Nj>8mQbEJz1LpF;C#3+@J42R>&-4rGq^(1 zhw2lxXjP?OUqa7_X0LbWc9YC}t?7s9gzB%*91(2oUA@5#Q1)GcK+B}R78bqk7Z_F( z1<>Zq^_$ir$zm@z4OwK_H`xz}H*3|tVsgUC%zD%Z?{sUi#wRP?vAGHF0AWMdy~dxf zJC(~=7EAszt3MA8igbAF^LiU7QuL$N{nmSR7tFpR~09 zXF6R891Y}9Yur|+zX_2k$E4@`32<-x4y*v*6+B3d@6rwNdFuIfSiP_6>!mQtg|GtN z4eU)JBgAm=r3srHM}|Y23i#)x*^X1o2End=-zVKNk+Z&_@aR8 zJNhJ#fOGFv ze%*|C4!`;j1C>cmMvhouY(1%9*5-vN;pPG>4A(O;B0))o*1wt@4UJ3^+ZGY`{dgA{ z?4!i9)CRtQk-NJd7FV#d89pWswaD<@hWc9eMAEkGA<(jdz z;opnWK!6O*t&yp+Bp2$yM;_2~sVVF17>x{^!PZEo0bIky230HK1=at1PiS1z7Wltd z<#kRkKIJ>;=NG7Sn1g$`M;m=~u5o=G8VpVW^O^uUr4X^Ha%w8ze@pR$3kaaVHP8bS;Z~>v{0k--)5m=63gAfF@tDWe?n!AAo4djy z^b3Wk?KGs){1Fyb?R}qz-QHc5Jpulh_Dy0SDbafh{$i*MYSTHC2Zq;z@0us7TaZ4* zG3pDe4w4BH6#1d#aJR7+#uAfKN!qLU2vO$kwA4F?BJaa|^&-d#r~b=K50Z@>kU-8f z&Xmi_FXdf4+6v)1lmu5s@%mr)0l^r14dh}E3=ndCbgus%udM(3LGy^Uan`bP*tqbO?#2 zJ$|tudOQSQa1OA(7r}MASJlm2Q|HJ%AYMhT`-VL?qa#rv(`;;C@e1&qqrRIPTQ^bXUo_#8to z>`PnTTwQw~Pj1d$hH6m6UH~j-j$gt$j%7Q1+qmGWv={a@muY^E)Az(I)>FsJcKhQX zc$fs`^h0$N84B|yD9vnXe;_HnA~W$w?RR4>Bn0EdWwMUtlyGGkTX}+*LKC4>%`9R6 z{ECaj8qJ!1_*QiG^HWr z=%t%_Ykm$82>vGs{kQ8L7Vh4@HwWB@dO1z^CC_T7LxyyqONMq zWYoQy_8&{mz*KnCQtgt3bKYw@nfJQ*^7T30Y?7JdZdoGcTb3M&;o86sxbiKFNNfb}01MMe@HE=b{jx>PZuZOx=5$43fbFa%)5HUvJ zHLS6RyHwt9Jn4cuK-p;{PJ?mU{|fDkd1h+hXKwFXEF!7;_LZu!Cbi53L{m}%supFT z$wO8+b57$m(=zC`izyX*l0s>P3e^6FMWOn@smwN&*Z`hHvr+RLBcQEh&JOv9aGn8* za3&V*)rDn?lUl@C@K)0GU{j-x3Vd!hvcr06Esx~vV3sVt`CQHSHVhb_6g1m=uf3C2 z|JW1wTU+7}eIfCShu9_8*;jbo=Aix%*Yw`6%CHYVu*jjY^liQKkV~hY-TdYSxr;CR zA`Gal;ktN|gCLumf0sES=LE0L1?opMH5H(6QQR&Jg#{*=umVFfMp=Z!^dO)Q7YRBt zpvzs0Zy`anwG>l}@{ z;^X@%)VF}0E5~SQ7to>Hd0U1EuOp>Gaeh|X70}Y;+D~`OAAWBg%={GZ5k^zw74`Rm z!bb3-7G}PzXeoX!G4<*Y%jYO_tBcAY(JCMt9%pH<{x^>SHfPCY?LWWVho3!oCmH3` zpKM%Ha2}zvl;)+X{7B{MhVvdc32X%9T-M78UI(j#?rA>*5&uviWRYER$0Qmola+cV z!&n&6^&&l0QSF|JhKiu_g0?m9%zpX(&>aOgl_0T& z_J3o61qk52s`dkJu=dR;17-;*A?|#%j7<1e9KdF`>cGUjmg{lNP}{Mm4p)AGhV38f zHAU6A>aG$TzvrA7nPjY-n;-lzGP`oWiD2*fZ)8tRU~0 z-`7MZ<_4vCRdD(AE=e!)R(6Fv)mm@vvFAGB zm3DA_^l{)wxt-!@VclmGIDeEU=UFTLCr`EJSAM*NJ>*)sUBT0wt>Qp-E_a^~33^CGXf7zi_hr3UuNqgfO5*5N-JKkyW@K{{~%$EMAC1GKXKy9Se}}j4nF*wDhSsl zG;dYM#h9I5BB0qe&27AYQi(_QcAAry@~N+V^BKdXz&6DsufXfLiJjKs3>>4|eH$Zh zJ@-J;D$VS7FTU54?w6BGrfEAyUzufmyscu5APWf}-f;WRW8k_Erp6w=#d1+l2E9K& zutXgvBIA zn!eH%e?)y5ZsW3vx#^6czHyObwuZvI3NPJWEKCH0hgagBc8~kG-2*A(K44R8yIu8X z28vPPUiUiqY7Z}Ml1k#zS8A1J_>PKrkxqa3-py7EuF*&pc_l{e=28a#SaX7)@*?~{ z4tnL5u3*ZE>y6Hf&Apdkqfr9_uco}yb|nMcvS%zdog6o!lXS-__(O9y-hm13(uVtT zvx_3{{Z-8(9J>;yY)4F*j(n3s+`e>LA}jcQu=MHYuGpV4hDmvff`dYePHK zUX1}jPt+l8_kWk_2z>o^2fcxN#JJx{Fi|3&$KM7FIuypJlb9bJcwL&Pu%wdy%}Yk+ zf~nv0*EZ)L+aLTg$H`b|?0!oUfZ~<$exWbLBjg%9fa~ty7V=DZqHR&<{)3EGEQUAN z)vP%K`_x>GzZs`wYyNL%%)t%4^-587@}sA`f{spm)e48qMbD2dFDe^~{Dr@icXsa5 zf6bwdY{~%epV!&n)L;TPXnSZy&FH|Y%td;sE`sB7&+p}4IaU@~QIRyDst7|f!2fun z&boIj^jys`ICZ+F3HoqR@2rii?|JM8pWtTC_=^J7kK4U#{w6eJlPDh?_-q%)Gu)-f z9C|HHCgLvBXZp{@u}W+G`qrTDFLVaDoU?IDW!tA;DU;NZl^?}g`C1dy;QH+j6W6)! zOfri&-!J%CT)kjx`iU*eVm~gKyD1B4Cg-ssIWi>b2;Yu(DW4#n#r-Sa?x%g;?5b>( z9<;qr{P4vFz0}nGvqPM*hduF3Gtuj0zY`{?oUj$E0lHb0QHibKHXiWQ>GO!L=;OT* zE#gqbp5T5f>;*nyhohX7gb|-0KBh&-xdg*))Je07ffMZZ;kz}+1c!E^1p14P@+6xF z-%B!WCV%ud8|RbnwQGr;(8cw%nK&y32uG1mi!Vb64zST-{9pzef>z$B|& z{Zzo^E7w`UJmv$<*sv-hDc6=*1G^|Hw$m$%=l1Sa4#{9&x!q$Lk^kPDI^f<%xaX!c zk?_bf7gI{N*oG-nBG8HGEd=l0;UN%(Hj`m|QqPsp#dAK@ThwiSZUwjbZ{LY4&ESOck|6F|FlWiVb#+qG; z+frv#0qXj=cDN15?G-(K-BI@%xQ55iC;sM|311onqsK#qbF8JvcvV?Zfl;m62Y% z>ntA`2?@|)p!IcvvuA0-)K+|idXYG&fNkg~gX~uQl&5_#(66ObGE`7Tn53shESSum zoM2k*w~`-ouL^Rre^#LjnOGdX=XND;>1~@N|4=9vp}8lj-bKkhnak>=*xM`oauv$J zmAkclX4T_2CByUHzj511GB1{Cqud2VyGJQT!9+tz03LVFd^7+3cpiV;mz;y7r_|_Y z9@KRektNb>Dgm}bIt>XfQCaL%3Nj$&NLE#-gl5GJY0O=kT@-49$^>@AIcdH)py+(~ zHOiJ>p>!Ar`49VERYgNKF5MoG5(fk>6dpv`->_~b z|BnCbK3{7%`goRTCF%4zNoJg=;u~qQ+f>)NVd*x{xw7Wl>+SZHeg($(9Cy{n_qGyG z_vPt*Exqa?8H@Bl@99uVD%i%5xc>)q{-D{tixz-rmjNf4rx}0abVe+>Z2fPT5|!}B zc%Y{RdarF$Ag8fZWBsLG-hGoxfH#kBfO;s&K5{mzsjO@H3qRgn?4p*8YwS(Q15cIj zqe7-vak>J6oHiy`OG~-WwyHqqCO;%he{d~B)u;EXb1Ll75Q4;`As*GvvfmFTsS~+` zD>)~AR~r1je6oWDKdh#t3|&D;L(5qn_@&8jy39;GP%?Rd(XB{`OGpxWZTvvtP>xFj zIzMQ{73kGp%Lejf{Gxas-&$??7174|(xXAb_5wz9%zzG!Sb8A)E|-G9eJ&MWVzu+4NhtqzLMCwdeZm?S7tQ`#NQgCT z!4T*`kQz5Iw=G>@G+%A4fwFPIpLhS#7P5g(6yq%8E~*klKtpxVuxXUv3xmjff=08p z66Icidg34=W-U*Vg=123fIaw@#U$QZmG2L#{fr^KUbS#mU0U8F>6F(=b}dxBY0IQv z?xU2P(CnCI0-0Z}Qn)<>J1jbYlr6YG8kN0N)y*V%LXn}dS}kMXMw^<(>p^EMsKj%c zX{gr;uVfJ>)zvdLTvkoGtm7+ZZ6p?Q*0*}=#I~HzSrhRp!MfLE4z~>+WI#Ae!qyJiO4MtHq|Rb$RM~L1gun#{nKpKkt10%zRgLE7CPvxGf%?=6U?z0ncx@ zJ%T)mpo22_GGL4PvsgrGH3-%>(S$3^}(y@dRoG%q2ibDKg`7)69nP zTF^kUCC8(PVD_txatGxxp5~5{r83lx+8JV!lR~31~`msL4_bcGC+@1+4v_ zY#2%)f=D&;IlVx(w!6fr0;Gk=tkb(!;kG}y?)~HsF`NY%WK!O zBvqB_X3hWuZ4;t6^mBFVC)~K6hW3p%uXx(L;~7glE?_tIwXecF?k^n_z(1EV-_A(y z(B`+02Rar!iukTE(;*D`Yy`7xiho6Bk)Os#@Ed|;)4LZZ#*dvo%NSCc{yf<4aJ>P! z^kkPq(_)cJ+73*=4C#{8B=LrS<44*Y98&zo3CXglogdTQf{9vGk(iN1X16} zbO;BTPzzfYZ_X7SIvIyk(`XeQm)(yCJVv>^x-B8K?`yWwf|_*Px7TX$iE@8V$}nfg z30F_!#W=V=+mk_?5}Od`=<}Y{R(FY)Ysj@ib)RdsTIc_h)U6WPA9{M#B1+Qh z7UpH{o$+nGrs0t=_>!c}5_F=1@|_g<28l7p7$(vlKQ+-~!(BE1#kR-(>nef4`Gb<& z@XVHdx{{U|K6ojN@{)_qTy_$|!5mH%AW)Unyo z2z@cvvoL}YU6z{*w<3P^d9o2vL`KcTisey#LurwxJ7Nm$dKvMhJdL**w~gpU;#cFK z^6&uFw^T4U^sGC13+4A91Y}3jqz^x01q8ls2_N_bhC3?^gavzsf)FdUHD%;qf(y1> z4hm8B5we8!|2zVTT@HY?lArzyvG%gx@rvN-44U=Oo1Q4zkLy~R`DlXoyU-!?fTp}^ z7`o=JJu!+~tjhguNy|iTT>tBsKG*OVV4U@Mfu!|k?tQWW!j^d9_?6J2*3?oRa_ZhP z0Je{P#s$qWAm3{01f6i%K%dULY-`{Y^}Z>>w+xpMy)G+MRuoADp{6?A?K*%8Iy2=Z zTgs0wsR-)p0k}q{hEH5Ws>AJ6nPkUQJ~?cPf8aLsnkz8%%@0p0|2D^!*t=2Jb24{( z?q1o6##o69eZZ!D;lUQfNx=&@Gv?+kq5V^k(EUhr0ZaZ#8wdC19%-6@`X`{dniTjQ5kBX593@g|D6rA`aE`eR=#j1C zzxK1ZOq9cFxlwe+Wes23Mx5k2>AYP&cA)Z3TGSjh(?@N5odXI-_Vp&BRT1(i2qN6tO;@$I?6gBk$A&;`vvU2-L5D9 z)K};^xrQB_B-R{rs%&RQ>pKv4w#(KLFV-XP_5;qZ}zL#b}mioAU+9L zjq3xV?Jnz?GQRW?OK{k7KaTdf4a-uh2ZXI;}arB8dlYrku#w1H0qarJ(urx;4}M}3IApUj$(;*I_44h(GG{9=kVQI?2V zp3)^|OP>SQQY7w|m{xYzJ51wVJLH(uCHr`4|8{xIHl5@Aung6A2~RWdLV655Kvak}WuddC}kuMt4_MeFZ6@ahZZp^p&~z^T zOt|m=&v6M`r5xHA2?-IJvyn=2s#GJ#9HPuw3^S)#a*U8Osl26}&oSroIfXfHj&t4^ zW`=Kl9>3q8a6j(H{kos`{d!&3)zDWRpQJQ*0K#=Y&z9wKE{xvVInNZdnA5D@z0Z%0 zbDGC_#Jq$kw@fplULa^Ic^0Ea@-x~6st7rUeb2H2L%X`Kb{&zr9ZwOS#)#HI1-~x( zPNYqqp`dAu07Nl7ZqL42BbG9xI{#g|KI|pLq4M=cRx~*tGXyPBn8~;Z@8x^zYY*Reg;b3H((k#Z z=F&iw^f|Xb$Ct@utS-i@FwiIE`Ea0-7*oj4?1pR`vAdxuG=FoZi*apS9zwH3eFD2x zARMK)+0BELC8H3GCwdxnVTC|TRLe2dBpLKJminpC^QfsH`>+EB)Qs2juD#YfYEfBi zyzMS#w^)SdH(mYkvxTW%MbZs|=ze1EexIuK$f^ivu)A(5__2K?>x&-*wUQUR8NaxN z+5b6m$g=b`-L-BdFx_E#DQY$h_@+sHH8PecfzQQjnEm`3)w3d{u{cOB<%@?rNcBBn zT3iy}pr()`X6h2kU$e~|wx4r0EIo`|6npD!o^8*>Zgvmv&y4=f<^;Z9FHEVCrRmPglxZ=J2Wa3%g`}d{m zks+##rt$hcv57CA&+MpnDfjr4epnrN5n!vGC}2amzO>QzS*}mmxikbd?HH)F{cjokajx)Ipdfu1ARj znEs__J0gVAnoHupjM22Vh!@-(acb(rcg<_4J8#G@MOy;9X$kV~dxeD?>P8N0yDjs< z+#uCkqtv&WNxuv|?nY*HmtPF3^rbLZXo{-5h05KYLgvU%a`dzaSmje}R$ zpiB9MG$HG))lQ#&5!#4&O%-9=nCOVEDUfmw3(DHvl@I1KZwn)fOSlYemfS@vzVfLEhgpeG8aXV7n1Zet2OXj-1lN)GPVSEQOB{sOvEyf0i2B77To(&IQ- z@dS1$gKpC32MLTGR$gt2fb$atjGA={qZ{c1k{EzjddGq*#09LYwtrr|OPN7p!`#uS za4tlH;k)wF?%`1$&99%lkX{@nnE9bz|G6u8 zGfw@52tY<#Dtk|uPk~@=pVY3`&YuPD4s%uT^jylPY#~k1`5@dt6a5qr94nbVcAq-I$Q< zaHNEo>JK2-MxM4tmXv}O$85zP5OTv|9>E42w5 zTPU>$9C(Nm=TpDh@98c(njGjN{(^UKCyH^i@y-d}PgdcA=t?2?(OA~jSy0LQ{uc_s z-xh-3R_A9C@#nzO;j`%Wc>!ye#ESyoMB@^0Ww7BYnH?$sx6svL?YEdj2{OTpnQwct z`yajgzz^v6o$*^PG4muz8gsB{f%bw5w4yXjA$BjfF9sR;Z zC$9N!Xr_&=^@Pm#gq)a1Z#CIb|2&8-eD>YD7-Fgt%o55Of;|av7^`_cr97+_$M;mNz+zvxYp#r6>I=uQai4`A(;q${W37Utyh4pL^ zd)_h-Qh1%2ggT-K9$0Q#`SQ8R(|GH%Ke7pv9p{g#B(Ge>rn55% z9(l5pCC#>!yZIVn4yByzlfP0lx{&{A&6v8MBbh&_aZ9{!sgm`H1IQc8%TBMfIb-ZH zUwC{TJ=H3qIf_vnrz=#Ouvz&GKv5mwiYAVY4&nI(|Ba@EF85IthgFA~g7m{exWTWJ zACJ)g{cu)~0Zm*jHfoSZpx#UPuY73#7=YL=dB60^qbQU6x@x%WF<(M4KTX4Z4vg89 z0VAlb5gCqPbM)2=v%X1#JWk%sj$AcHx#quvqiYW6=cRXeH2wI&9YA0Ffu8S2yCrhQ ztaeGKS>hv$Qz7a-HfG=HL0S{wsO#ou!8$=yR!-=U-TNO?V?s_^KkcJ`Y0DeKovW>8 z%^^7~_JeL@ZtqEjfH`y8&A~^oG?~?2E2*(DaZt9t(ovi(eI-AOG#%-nkSR1krW%7{ zie0m;)yIT@6J>@~+{wt1cUwsn;eZ;v_TsvsH9WE1P1$~IlLbm^Dv3PaPdQ!!mJE$# z*c1&z4PoCKRG=7~#j-Q9C=WXOiBp}!n0&@6!T)tnG|}v7F&8|#vy*?)*alCyEY-~q zbeU-KwmCh+tu!UqaTyNA%)zZM=l((5BKv$i-+~h|oq=I>`#2I`hkMhb{2cv7+uc|4qk) zuyL%0SpfeS|B|YgkVU())nmoRE+DAQ1}FYpr+Uh>HwTcpehaxYB!@w1Hw}iR^?sxMh?x z7utQiX)F@a#cZ@W93IyZ`C=o%xF!rR*fTxoceaMXWWK5@lnkGE%07xI@P9o4{ZNrU z=~`U~EPAx@b6}(TC>}FirdVx%=>{ULfvug7+M83i7 zggf!^k)Ker;h#_FMEKF`vu#DwBZqkQXr`}~!tkFzf2w62m>xO6l?=wzEG|lYpr(BK z3z1x#KJTEbQ-iAKrM>S{?)*d*LDPPiK}vF`nKUsah>Z#nyFr}Qi<2TkQ``W5Z~2WE z&~MkD{RWtr$9EgwW8u7N1t6al!_lX*R)}#mnLH}m8UN!p z|KPyi@~_(dCDCsHDZo?UcZl*|?R|hJW*S0xXfN>bRW%plIqu738B~O_CYA&@&voB+ z`=wN{mLF3{=$Y#aH8IzIpDkReV)g1*cJcJsiw0adMEGZ@sDdvtaI>;}BD8)95gn#B3o!7hMT3#${wT$gXat{Z?z6}n#(YOh`0j=hr-bH3tq zk&Dsk1-9RWOjB6Ae&`bKsbOHUAy1TCbGxlmErAYTEvcgDV4Xu`c0)H2fzE4Skx$nIUPfNl zr}A?nU+)!ku<_v5lP$?08)xSMa(jYG6F+d`N(te##9r5O$(=yjxg*L@#tS5-E#kdS zfJ;(Ar-D~pKp-L4z=B=YW7?+$bky&G=A4TD!Ie$;lFhPiWZpwMKT%Knm~J>r(m=VFrn&$<-)=fe9@gIX_gqXu0JDkhNeQ(dW-wa; zvtT(v^ufE<-nZ*4P-+Zl_vhIn+Iue2cUofOJfi-AFuNAtY`b{F?=b5gvQps2gU!j= z<2~|ASczcQdp>BM2449A3m;_u-`MC@==kli(PlV)CP5|Dl=9?t2eI9)K}s27R;SNUM%Hto;#9m=3a%oP3?R z71lUiw7g4a2aNy5i*JA8v(lP0U&6pWjCCP${dSxtN$$rBgduzS5@pQ@Pr&eMl zyF30VtyCK`jrfU-l!@Z7zY#u8h*jIG(lGN|!%$x94Q9i70Q^=RC!ERk6GoWp6Q`S8mT_ zgIr9wb7iLGFYJ4tWh*)f`qIEAQ|EI3fZl4`^Xar3rJRF%Q0{*Y^f`rQDF!o4-ba^MO0zj=Nl0e zbc5EwG3a#aFFsFk9&978KWvQVxCoJ57 zXYk2_0!iqoOtK}rl;t>d;Mci*?+79PQQ5%Q0Q3b*`zQviAu>`O}hCmo4?%r zSDG?pKIW5(m;P2b=#~2ZU_f z70wf9V>Ohd62R|aM@E6gPvER9Yw8#H0*f!ADna@KkxymrB7W}2ZiZ=dF3ZGG_BG1& zg9`i+xfK*AU*+7tq|KMZy(U^=~-XGY7f1!Khv3{G;ROB1%J-W1ys_r?IXJZa##@DL9u* zWwsEon134EST5w<7&z^kVllb~#?X-m!6ytR%L}@PDPq)}HqWyQ%U|4Z2TLQGk#YBg z9MuMV4($hHN#?5gnSOjm2^7-Jqmg7y$wUojDOV<0D5CEN{25Kw>r+ilnCd?Aw=xpo zVk_jo@vrWlJTyto1df=Y=FCuE8i`!j>!W+0HfT8+>}VX?=v$F_ewLiu3&jyyB1_aV8Sk~P!2uND1nrFateKVj`eMQ8*%hE>i*78h-2Fc<0h%hc3F$1V z0gr*7cbj<-UTMnbt(FAJJaVvn^cG9C#+>$)ce8yj+coTqvWl5aAk^H&Ld2Rh>g^&q z%UIlQAM;!UY9v9rwWi|5D*(>sqFm0r$kGxSC+55JBSnPVuX+(Uq38<+f!aI76q^uG zjx6fX6w-pco9(crIo)J8xbni?a%6bp@Mt3V;Zx(~XzR4+k|Qw9_&~jbw7*K&_RoM8 z9{9faKV1v}o<404m$74dNm<@73@Nbb%gxErdNn7br6|HFex+F#%LGemIJ$DLRFWFw zhPJC0SS8yAv~2nvssu(|^ONa+vF--RgU~rT#1i3y z3w>^flYhFUe|m3;5I(!)dk~9M5W*|Z&$>vbLi@~Y6}>5qw6C%m+m@u%mod61nUxXc zJeva+D@jZzr+a@YcQJq`G23C*^T+XLs-(i)Ki0i)T)^8m6Qd_F7;zFOI;@VX=0 z;k=b)r6l5O!)Mv?oq9(uPx&ezQubtP-ZG&d@p`d#Dp0PvH*QYhAlIDm8MzR$8S`Co4SZB)~&<&u?M)N?A2Y?hw;Fna!~>&KC~j4Uo7Ke=PoFmaL&dHmiv-Mq+ws3ddVs$b zkJ60yh@p-ZO7oE7CE&Q3xV~aVx8X7~t3W-??jIUUfGzRSS9^5zHp}64gVrkRKCeqm zGX3H;t6V*Yblj2x%tG8&h0ktGNjedFKrp1dk2Jhqjx4+>~ zEvSk4Ld_*Ci_jYHEZl3A82(y>!HQC9UwQ zZ~t+69BbUs0(bBpq}lvhS%}fBeb6$D!<`UFWH_qM?p5LE5^1j1U5uDSl`E^B#<0C81 zMx90o6D5>+1lsz|Z{ZF%4~v^lWvKKI?i@|MXb1{`2Y6Pw64{P4N2f+ z7TfR0QQ*B5dXK4s@Zc;G^7>Fr+fe`#NMc1zZ{>@I+RX=BHypdO4F0JpavBwKCA?}3 zSS{%=&!jYq#Pv(0KOkR}PGPuRO1CNMsHpLbw-aUm(Sw{@b=#@254n6Zf+F*lumDXo zKV@ojdOgNbTf}BRQG^s+9sZ%6(_&b7s5;)Jn)jlNO{HTBvu{+`(RoZ4q3)8e+JPAl z*w|)#y;Rb^bk6WGxVfVfK--3jk-JU_eX)PL=_~aGM4Wq8R-1`Sh`e(%YmQUp{ST$B zq`cW%JV!W!h_kgpGV)Vm;+^6`L+K65a4#>f6>Nsw{q<{gfJ;K~V9GPC1i|S~fEir% z4Mz{BhxgrxLV{V%p^oypoc=43K2DUrrw4)b`&%R~8sV2HN?f$8n)bzR!Pxi9dZ#Jx zkBLkR7$2pd7{T%wTrvTktifLyZ;rEG+eNttZw{OIv~E=hv{g;~&XHQF-ph*89zT=$ zshn!AA;_A@UHfWqt7;{@+}pUv=vRA(M^ba)2C0Pc+yq@v+UzJg!Jp4qNHmU(G2997 z|2*T3%{RQYM+}ZTp)!Ya>v)}pw|tdtp#IVZJhO+{?Fe$)OhnHeQcfEjUx8HIFt#^A zJ_thh0B&2?zbm$Jl6uK=GRO3hT+!XNw1_)B&qw{!argHNJaVCs2HNQA3oH}%vb#qP zl|XMxc6>qw1XZn5qUT>Y9XDYsz6XyK| z6%737+$Kepc-Gynfl)biVN$v2YY^;FcdX&Zu{LQ`UPJ$r8^J8KND~dPNBc5$paGWP zk4oj~zONF9-I+NLDwWs;V#bqBjLpi3K#3F1X{NB+arM1}!5|EcR~b$x<=LTMyI0_m zJELgSZxgy+9G+EUMz$TX-SV?nwh=x4*_!8UBUu0^^m!RqqL{KHhQYC<*aZqK`t>%UxfGMI!}`m#tS~ns_DgPw6bk2tcMPU@zB~O1#23P>1+8e ziAUSTwqCf$kn_?dy*+*1MG9`?AIe!KFiTt$kfnrMM2A^>hX&4H zGQ2uAG`>7q^OfPLpsJ^@viry*_7?V9_7Ou#nGh zwN*pSNm{J9&^h6Ur|s>gFoq)V1+Bpv_ZzYTo%%p@`7ZQineDMAZxlotE5g=ecUds4|r8D%xn6VmE#qj6T9D`gb+KjxDHDh}7(4?WLQQ+=21 z!De9XoTE7F*-&nmbhIJWkWMe;BboLcZKpIs02hyav8OiYqGW{vE7h*D`Tji4+1~J3 z+cEbFW;`=E{q;Vx2|A1AJ2h13v4qpbj~utyTr4?9n>ZtapOjpEd$EQyi!E_g@Y2OL zu=yurPjDC=pnvooFx`5tZY3Zjf-oH5H!yQpSJAEY?yArzH z4W69ZQF^LNS|PC^FL}nEfAw)UL`*r`ja)pYa?YT_tebtczV~mIk;ELW9ZgNkBW<{E z|2-is-$8$VjO}XzXl8XfahCWf9S-aWIx2?SUuIZWQp0O7`&vGBv(f=HF{Di%|3+(% z@b!nap3;@awZUFYBkmgSz9;r6jR$+Uh20N;fdKB9o%Gr0(m&-qd#@7Pmttc?F$u;c zE4HF{mjK9LdEs%@FFWBMkekcc3Zle;pJJ)hadzBu=Ou}~Yt2j;G;=-W9=wE_IH$Wu zx<(9#Y;?#gm3jbzo!)|IQJTkbpc{knAqOpd;a2vPKiGqN8Yci*GBL!B^hjJl;oxr6 zwm>%R9Bx(L=2evO>i!y5Xc3+3vFLq8xrid6saRCCxxc2c4}Y~!d-uXWc$uN9X%F{) zKUGow+_^}>`0a+q$j$A`CVR=!lxB6duPV%NJX-*4zcTWbt4Rv*r?{P`0(7|0{_R$W zLXk>*AnnWFve7H8S zXJXf*PjM@f0TMc;Y$~N{+gioNrQ>BNZsVidd4`^_HD|5B?%TCGWA>YcI_P~K|DFZw z6Z+WvvROB`ko}eo@8E%`Dp43_o@=}y_VlY4kcp-TA56*orAPnFU7O0+Gy=*1JwU?0 zvb|1uU(>QwTWI>*|7ODo`w<#yO48e(Z?%~)XN35^#1$hOa4VvN;e5zU>lLkdNW<0n z+31hpcEA?v?7>+dlTQ*W3)?4!(#vjQMquQHN0S@V_4fb6{_{T?m&A^s=`~R}{mXLOY zn`&yM^)(ysW+@zlL%oh4#?>LT59NmB*c7LBHGJ0lY?Ke_51%hes`()@#1VQz>+IA{ z)iF4tuwuH;TcO{ms`mp3m8vBx>DF>Y?9i?2=rm%e@0!R7_F`k72hDDia{djf z&v^QhZpMAbZhjf`N4}?RT6-K73j4VspzZQLG+)|C_rFHNc^cVV0L%w8?b+X5-I?#u zDJ)!YOz(cA0e7_;0W>haBGuJXx}OGl&kG*e*!53#M3motik7nc#7B!%Dx&>q@R6?@ z&KD{cL+O>H?>2AirjrU9>TI4Yo4qO;soJwlSNerOeSf*wgn^iXSoi09o>lz|FP1K8 z@f7R*S&+JSPV4AjL%fhT+|Z~O>FSt?`$NM2Ii0^QckXep*uAr-*Or5|$p0b@F*ZHS zR_WsVNBq#K@*k7MN6)I@0J8Pao3U+a(4hUZ)Zq8^Y)L&+Qasc%)TVP4Qoh#l*Z%XC zy~-2(?DLRJh;x)+6vuAaF@f(?m8mTGF|c}dS8t!{{4hsDofYeS!(Sbo$BKI#yn?$i zJ=?q72t^>>u_fH z`1lx&`6Ul)VUEY5+hYyafOlo?QaG>k8uOlq!{nBo3 zmzZ*+xlM-?Hv?%u^tK3~i?jw>#XBes>w$*!8dp;K1C>c@&hGy6auK-dp~nEA`U-D@ zgn5l5z}%c3q6E~iryZ&?{h-2*TZ-RQZ>PmV0i^SpWvRE^`W0YsB-cUrGfaZ3{(8x0 znasprTi)3N1$IHU=itjf%W3kM_@ghy^jFJfqqF6Or6ec2u}%fS7dq>Ar{;gSxw||o z7@0lkx7}Kg$|Q3#G{9QmfO3k_7C!!D*rCWh-ifPfbHlwV%cY|B^mXu`udIljaBe?Y zF#%oOOVe#D{UbQ|l8y9Aq56CK* zwmpn~8^jjUhzaPeQ9^uQffgFfulhPN&LSzl7Z2(zuYl6gQlKlN{m8BB>bss#csuR9Ve_ zlo|*JEo{lfAU7_OdC*u{lSW#*Y%G7O^laqqG5o29%bi7=M3L8-op}m77A=mF^ExZ? z<`gjjjn}l1w3yjB&Vb=!>{dNMN;GLwj=AK;|y`efx_#a1I4QOc1>J= zZ^8O-loGJ^dwTJ34*#-Ck7r@SHePViA(6iHb-ZgnQV;ASwVS$AM9S$@-vcuFFoeC&a2%lB| z&mJ^)p@wF>!fda^c2HLLY|3QMf!5JVIG@dw)N&%88%>UOZR(UAaT1l1Fvjt?HQ!Y; zOHlJCF<&A$W>p%6N7dOkkG=DwoDp(jQOdku-F@qkwGdk`bffyUzGd03EOm9n=C`=d zI94ou7~p3p1=yQ!@)xdwB1X&p87;9I09H0Jo*#NohORnoCuI|~f;S7bo7dG<(!cMX ziw+ijTQV$0qll*&vT@!!V!P;;6xuNDsun+9WSu>j$r8FiiydT^bYT9D*-LqOZwJzkU)5!>%v(hUFwcnM{<=7xXbwu(}6^Ll1rMt*kh)9lHudquoy=?)&Yt=-H_ zfzvtNZc{ve{*Te}t0vAKORfdp7T#63>H>v}y|NFL}pS zWyvHAt=b3&Mdq*4?yH8M*xMNZf(YIA_Rr5qc(O`0U{m(oqf*FsIIR-%ws|{9l^)Mc zX0Y4u_peRWt96Fk2Nta-IO6k4xO2LsSP^FjG$S0=eCv^$N8>sG%p@yx7z zE+uPQD|R`nVjwogjF6bKbxn=KqPUbn!X#4FKFIs-XxyL;*gI`Dw;UiX!q=&^bkhoLf1ME3&$HSGa0$hu zZNrdQmx;3qSn3~6Fgsv+;=gjHCsXav@p~7`DY^=?WcC=+un@$HI}NXdA*L_W&i)vN z##tONZ+mFjk`5=L7#V$L+{>pc@U=|xK+ClEE_?_?sa&Q*bYahuSC_jBug0h$; z2F}9y+=NGR&S4ArBc$hVYUtr;*Z7u)1L}}7JfLAe)jRO^40}NS!YxEo{=U@%s;7ux z&rBtYxw@_9P3tr}2t-^$v6c8Q@3?b4bHNUs>XCb8M`HR-fTRH7HOCe?erhQ0FG9^U zDZz#jeO8wV0knGn1|!iggMsYXl9n~0bw5zQc%m}i>MTz!N5M!d9M+^ekM@j;mN3Wo z2t{>t_~$-T1`GUYx>Rw!XVcOMP2Mw;2cMf5H(%cN)H~uto5!ND+*M$HJ5$UP z%Q8dcJc$BKYU1`L$EemSm={Wu8?ie(eR-|h+R(CG(9$dTu70OrU5O=tz6YVD5gG+& zMnvjvZ%(>@GjaRAu%5Q%G;UF7t$Klu7C&Ofb87M9$eSP6AIZP) zimd8*a(5PQPTO|w!PksiKx8|fS6h8*DwahPIYqBioE1BICCUYSmTkHjT#lHJh9%Id z+d>;0$aJ2q7Rteq6Zxhf(mV-b}IHm)w1TPmTM^W1k)S`>w#HGQP!15x&Lkw5P!J&VN=lyy&~b z|FdK+IB{HKeJK8nQ=18IVytnK8S1iGyPUP}lH4E2gJVk<&SCRw{t?_Q9$y^(%z`c; ztzCM`R_$gaLK8l?ztd(txl7+u3GTFCz7%KKZJw(2@RamvF}G&>*X3wZbcUwWq=NJ` z-|e3;*E+{x7En}Kf(^noQ7^tY=9n<@gh`g4`L@60=~$tX=)%U}QCp*|-KTJBU5jNv zbDdJUU`CF$yQ>0z?p6~N$ji-)43TQs$}v{4h?bJ;RAADua7iPD<-^_(=g#Wc{IFyw zB&)hV`)}15k}gXNSIYX+YAif^xv3*E9P1;qz)B_Y^>vEc#fL{5VbP@E!q&SkGS8w# zz$}gAu3N34+Pt%4o!W^|cR|Ks~v|gC_iEWX>p6 zriNb;GGC`SOERi@#g&UJi!R2b79Vk0rbLD2E7nMYhUS z{g6ewJV@oF-N%N*v|ZkoN~T9M#4}vl%GIdeoxdD_Hena}k<^gb<*s|yCTkms~#we-L$agI#*oZCqm2xWs+2MRSeQi&6sMC0n)Q z(~Vo^$HwZjry@)r*H)2*O^#N0PU2l7{UE&GMk1t}LW3$+jR(A+EG?7RrUc)OtMlm& zLvJWqFWk{kf4_1LMT@x3u^xu3f>BTyAa-?h4Gj6{=WUu5FQuY7*F~n(wAY?Y}uU z>wt~Q5Ydc=snc%GV6gb_QxpZmxT8N~%`P^=Iq68MMZ6|qR zBWP=tBegC#K#TGvvtl>$;?VVNk=8)kZ}1qWIGwIK$oLAXos06+?9OP(Ujez*FYmN* z0gVnIs=Td)N}ZY|YWUJu#~v+b-$k>9!dPJ@^uJ zd=UhJu)8JK;u)x2&ziTeqqdhIID?!cD1-WB0TI{aahzAGX_$%_LpgxrrijwKe^S<0 z${m%7Vn=e=k8UFZ^{oJ}H-&p0GDY(2p6xbaqWYGF*2PkQ zxxci@FEP6IukuC;VC0db_vg*^^cutO%zT_OZYCV2()A2n<0%uEpSg5Z4J%Muj#Vaz zK9YS`*!ITN`14o#!Eg?(?EjfKkc`2Nu|uzAEyhJut%NT}aUsu#Goky*hKAc!UGc2u zY&L8jS&1WD`T^vnQ~l*0GS?S)FHt+oRXNe=zA~Slkw@C;9PiddPla2%vlVvyZag*G z{I02zEiIV8$SeLDe||LvSY~J(#T{7R{T;46VFszVfh_NBX}+jlUm_`TcJJ;_c^^A% zcE=OnJL$|`jQ->Ccj0k&I2iVU-r6=euZsp{cmYy)cx z2A;nJgs$M&iuyrrFQ1v>gGEPZxPCC2(~;4N7)uBUB)*i6dQ@lY@dKI`)A{0;w%l&+ z8uf7pbOV1`u##lRhv!h+S-M6S^d#@n4{Igrwgh5mml~7w>Axh^t^TDT13x*hzrzfB z>7~Z-XvSi;d0^}-KA6^|;!h?kyc%67be11qvWm5k;`9y)Ai zv2E)r_}2Zc{zs<`z5Hf{JAhAL?!6=8DWgx=CLyHQFdKOcSrD0pp`fS~=M4fZR!90yhlw0_5LU zvT+)@e>UGtYG%u%#tUBE>686fT$V@>YL~(g6ta@L&TTm9U8k^Q;=}xXtmgu#F^&s! ze9ynn{wAjMJ6P}SDR|Sa0IU7hIUjk(x4&KT287bqsVrMqXKe;%y^wJ?FduR%GsVV^gk1t&fczn$a4iAjU`^VIccx>eVWn z`wDTuUQSrov320?ACJ@X%Mi+FuRU76LB26TG)UyVrRILg{O{|tMM3g?nq{f$Z?hhE zU~7EI$iR`_PfPG)fZ$RPOt?d?(}h>VQI^n62Xjh~F~pAEaAs7zS9jKN!;8@MZAEdv zb1w#IE#svid&#fs_S&y>xNV5V%0>9^^-81{?gr9V@{Q+hxVwQzx1#+fE-6m+gVy7J z45)Kew^|Ih%eM1*X)$7+DoO_J;T=1hQ(h;sg6N-u7<-8wwi2k;l&xU%`n!d%w~ha5 zNi%c@4u7lbq>yj(5mMF?MRV|^_t-!cRCDUq8!U^F zwe1MlxgQ{evKhKA)S*+q3_Y+?FBz+@CJnq8HlsVm?yoP-xaSn3i&gCim9@bcBoWgk z7rU0$au3M(HRsv*=yDBO{^11wq;k)~x=6z}X|SgZTgv)ag(tz_-*4w&TV|!}VpC6u zT{ja?k?@^hu|b|P;U2i#{+-D1dab6<8`YBy#4AIJE1XRRa|s+XTSpb)iM!O> zdY1NZ(H(r;uQTf{75mXvUxW}j!5Qun`eWAMt(%?tSfG4n2Puo&zm+cIwy zjf1aVFPRqZ1}mkUp5ZkIUP!Feq%mEG5;v^mhE|sm(pFs$YWts8B;?)!J!?|T`e4+T z{-paDZ8;1v-cJsX57*Re=}u2yJ-C_vg|-3hFdxDb^5dK;ofailM33Wd9{=Lr&IPR| zy}uE3Izs1>zdg$ZP`8X(JWKrtNB#APfdp>9_^=xk{1fQ>h+jGnvMqx1=u~UUed@1| z^cU&$_-q*UtaM%fSF(1560LjSR_gbLQ@3I}Dr&L5O91)oL>x)U?Oum`QT^BXRwn=3 zAQG40SAchqMN=~JTDgvY*XVBto+9Ks!IF@T4{$)xo(_`sug<=%INRTybW{C`+)iE( zfz4_su{7zWqe+D3P6#r!!5IwMpTO6Y`wcAP$p_SYxJJi3C+!aSF^+j*8P6SHU*2<& zCRJ2pOMAma`wjK!gyC;wl}bfz&=gU{(J3L< z(Ae`!I!tOMFFUGFlcXjj0o_4FULOKcxJ+#b7ctIT_hV_&$xQ)Xna=b=?$lwk8)pvx zjMcckybk{Vy#VN{{8%MSm`ti~rB~Oz0jf!+N8jsVy;vKP`I)LZDz5IYq|X`z)sens z;68a6ts;fy0p_AX=vY~W>7#F)$Ax3bKxA@_uMyJG_u!WM*5mJQ2G<)>y4whIX2<#l zTmWz@LD}MpTPbtxb_ixk;c#9bZ(XW-)N9&D88MR;RSRJ?fqfiB35?)V#kS zK}5kgW*#B-Dg8%jGYd_9c}n_&=uYDy5oc~pn3c+7%85*SK7%E2DX7l{Wz6^4o7gAH zaCGsyr%U_GslS*jQ8AO|RRy*daBroM9c(bd){FB-G z+i@|y>;&Pjj^%UgH6`ZoT`8DvSB2TSUcKRLrPtZ5Q~7+<@hoMCadG1@{oL$)0mXhk zY)$O$aiZ*E>wyhiq^~jHeRo~33gVO-ym$?DI8qyAge%$ zDw_BgY7l)Y0zZym96?_y&sBTQfA*zD7?)Mu0zvOp^$^2nmsCB62ALF(3Gko)N7Gxl zHQ~PR-*k;okQ|{>Dly8@3`9Ys^oMo^i_bxdXQwGm>1l7L87pN~@lQ`|Vn)Jh=Eg&s;Fq2>uAG`S!zQQO( z%-)So2&Q~U(Fx?yA-+;Rf3Laii@2}j+FX%M@BUA$!libS;_X4pjaB$k z$I<=#$wGRHu_dGB)tllGM(Pe(77ZS2bvn(T&JLTAHG#%cpb~m3=tlO$2HB8CmRK3l zz?c$=Et6^EGsUyrXDp1y%(FBxrs6%A-{LK&3OQ}GrffrH)#P;fAG48H-Y=CIm;8~_ zq^&mvb;t??wQWq?MP$hdFcSIvz?QJ5I0nX2|J7Fyk8 zF@(yt$vRBQSX((X9G9mQpGDZBJ};{QT3BHGX>r=CmmzaKSSt_eTk0 zuE{~vuU-*>Z&FvBzzlIpN$Pv8MouQr@V)f4bIjCa+Lk_Ng(*5FoNGxq*qm=Lx)BW5 zvvicnN*R#vxdv4rwD~0)x0020!rzBcZ!h;m$q#Jp%k97H3{ImDjFg`W*4WAug;$t+ zZPpe3qa_n$V9goL?uObidaB#iVF?pA%PSq8*4MysLA5B=ykN%RRkkY=6ZK_}NY=t>RN4WA+_vXBi3yb;%?J-^?H0NC}B_VMf zCE9S4r6oHe^6T-WHA1UkvvnngA4{8hrD#`X#vob~?y(Lfj=J#_5k1#miN<>tO+evs z#c-!doGg+BGwz4?*y0E!=N9ATfXs-a`94}5c!n%q=*)LUD)h5>KzPq8m&pCs(3FYp zy*b}9z^3Ou&)mK(v4Vf|*QSR{61@-?m!!-6Y)o5nqWGzF-rBQMaej@p4ebDdpNHQ+ zPpE6mY3UvMpSl@KQepCB<@UfCtp}J1X37mu^Fe1>#O=zqOj?m1sM3|RbY2~xvHe;0 z$o%L1mrRS{+KAqD&mcm8d*_RL!KV#uWk%7}~7h9`vqnX(+f8zx{VD}y8}r;`excw5v# zwGYQTyrn(q_sdy()m2uKYhc2&J`UJ+o_mqtq5NCE2zUa=($3Q_HxKW$+cr95xAta?r`ga|3EYZ5;B?{lE`Z*Gl7sErY$F$fy7h@K^8?Tl%^+rSr;8^g>~q4P$*qu{=BbzVCcTZp<= zteDug!Z;O;7%;0=J$CY=*GxaQ_S5agm6{_I%!RU25`m}r(IFP6v^q~uofF#ax}7C; z#BJ`*&e@goioojK;VT@Wq*#bjrCR5mr1n^h0`z%#ty;X#RRG9mPk(gSbkY6HQ>TyT zOqeB+@${RN22mH-pi$qckc1($w&EF?>)Ds1kP6$h>ODV=-%-3c_hhqfL&sECOAo%v z2q?|1Hp&PcD1wV`<$a~dsbr*tqAZof_MbUlL0q1VbTSDOoX=zKqUefI{54EKgvcP6 zS?GF(AycvNw?A^{SD8nN#@dPAR#eFfmozNg`_+7 zvi{uZr>wuV$BWFZ@gI{GJYH4RmvdaaA#sf2GdP(65VXJj%xILd-#=FqPwe4aFDyW02n1aL>bgHU$)VMvzILs|G`ga;^n&n-Jb9^_k}uu&=0 zba^p%{Kz(05c(qXFY=0fgvs9TyJD(K=eLo7k`Dg$x5KZ$e9LJn;8!}RLOFi%+&Hn{ zPIP>0vR&|{n1ey@tM20Ufw;nZTG|ag@jUjPk2WIYY}#0twrfhh6{PUFM;rfPAmXu3 z1-0{@3kFe+77O5qax`xQL0tj5uBCWmt~Jl!GWt->$ZYRvKSWbx{r0fi8=Hsfx=lB) z=ko3JnL=nAOrPj|E%uJ8lKt3OZ1YFKR8)^6~e5=s@M;k!D^2{qUdKp6pC~~?bSrHepwnm3@{qi<#kifunA?I6 z0~RCXKF%>v0*B)YUD+wKV;v>_;|j?2Y3@Oy=p zuq-#W1d{7J@b|U|gmVheVcfml9@b&tLs8Nha7*Qg1%+=Lm&-#{1B@3C9w@1wyj68P zRLb5MC`}&;`Ot4Rm#>&o?x0EBR$U}ieBlV6sKI|u8Mka|jv6dp==sqS3V(qK#^Kg1 zVB9~$kAC*;;zxXlGT>*p=m~9B81VhqB#^U(gQ!Af6lQlW%~T%De4&0jxWVdl!{s-# z*eRWU{`Lsi-FoxNeq6{*=PH1qj`$R+eiBUMsI`9=G=`TE-Q(F=UUb)=tl%#~Gj>zB zeJS5uTjc)2e}6WETB$FtB!T^T)m(QN@-{}p|(&u?MC#ZdPjET@G!mf=y-sd@f9xwJ^;8*FY%ph_Zy?6m3 zFkPPDdA>hBIB|O+L9Wsk|03sSMuy)=jMo(ZGmkpdrN0!mBGYYYrY6Om2z>JB-pVVKxyM!SC!ao+W^hrFP@JLEh^XS;9$rE(=2;*(+O3Q?N@mJmEomPl{ ze?DAZp3my!R~qc>z0UD#O%3)d5q*8MjiycogU9!gk8E}7`}kezL8MmeQS((&%C0dd zGrljaT%YemXeN|*_m{n7{S9FE>E4+6NseRTHkTipZ_9uLP_?^aD z&kJuljx5@-V_!Rtz~Xlc^dHQLT}aj|dd>M$Zv{ogA1BEX`cmJVOsAwNr#%HGa_#g* zx}*+D?>EDW*0v--`gt*Dkmbw+*3x}e4<@0nOKX>+vd6}Z266hE&Pr-lkC%ww)>gRF3-9hJbkvN3= z(T^wy8k~+lor(b=z8<>Ls(QLnzV0pf#-*KedQY^UiV5Lqhv2R-gj2w;PQowZgebih z@$6q`;cK}+p09!1IXNdJW^XtQG6>>dVP1%hSI3?D{j1E$J?-qu?Aj>FfL18On6;Ho zhoid(WvJDV2$ao_GD77Hq2Vt&xG^j&@t0Ok!$whKY05NZMa3v$ar5>C;xDFrr2`7Ub3f2^Qf?JJTvQB2nWUQw%12X)Ya;$e;*__ZE5#W13_rNQ~ zCHdz)NXklOy;j(7b0GR-q<|Nn5n*abfKTApihEYR7fZAp_nVWa?Ck32QS_iFm4f5% zwaDAln9{C+sN&S5*7>!p9%%aC6wM^=q>c&ht9^PjppqkY)dno{-w|VO*;d+=V7?G| zyg7J;ICbZT+b4aA0(m^kd2>~@%rVQN++)~j&p@xil=sSq6x3i`cZ`0#2cHiZd9ut2 zf5nWCde%4Z^f-wQc$e>OOB*I4w zx|Eb6lU~&h_Ub1Y`AhL6@c;|>g2^LMpyMfCwG|nt(U6y$^|$SEb~(zOnx&v2E}xWN zm1faA!uX;;ZSm5JSaxWKVFYQ$9Nmi^WE3I<09^?Ukq!6Pt46FMh(xh8bX{qb6p#ac zkJC4y+gBW_^@`udIvv;uiXo0)QxL`$m_$3-Q9gkLxx!sLDgZh3u7wWMH-X0FDA;y# zN6}B?q7T%H+6yC!T=$<&t z_6wss@^07h=;G1W_M0O~0Ypo4b9`d!x;IQv4n7_)Z50r!uHUW% zf|oFx;jxm`(#cMIeo@r+IYPQ^nqCngbo4ewbMJhGD?*L(3gY^F_mWHxR7>(lr6EaV z2^$mPnmW6Eb$)p2$yb}r1(jJYb_CYPY&)*%?p87@L_>DBB1Ekw9C%qiOrI<9k6Iv$xNrnGa`*FJPo)^T0^0y|wk~Q=YQ#FT{?<(E$)omDY)&z)`SpvzQd=UQ z)xO-1&{a?B?`L8yD^)jw#NC*RAv4S-7{@3o40366pw(}8>L0dXfAfp!jZ(E4 zvC|}Jf-V=Tly({Gsr9GlKhX0&hwQH}s^r6KeyV&64eM906J-(gqpc?VUx?~l-LXNL zHGk1H>#I^$bI*gTv#!>@FZ#Y}_{i9pgq}Q&(2I=hbL(JScAJZ=7f_zdhi_f|dus7D z^)>yvOiy4oqo1pE36FJFMp!$TZ)rQKAuTY zM85mbv+k0J{tAlDL$9XiKgYigm*j^(#M4Cp--3PQv>5w{G!|1b zIA(K&o3Eeocz}!8A!J-9?;Ak=fyHi;4e$$}poeZ4qE7MV{Z_n(-}Amp9LNR+7$C|+ z1&rc1v3tD@JBGF~5|NUBGZf?r9DAstk$+;%Dz%C}xb zqOto!oS2oJALGooBV831-V+=H(C>&A^3n3$8{~hO1!`X%Te>a-pf3ob3-=L?W=-W(MHCt3zCXc>;;VzP2xAC}^5<{*?nSuNo4Pd8`it}XT<)VSq#D_yv#=HH+I)zj zZBo9x9cQ|veOW=^(&{@=DbJVWrN*$403*S|`&%->&Gnvr{R>lpA&rzPzUz&qUjEx- zqTz~nmirQrp+T|T{&{lfS*bVvLC?xxn|pG)nmg_d%{S1q`b0msL%myOj4vczfKkVH z4ttjOs(kUboLVj$`pNNb0laEE-;z`i%n_W>GaSTve!YB( z`R#F{T}ruEr%NLuF69oWu*vV2@3a88XM6#SIJ47wdTR19s8X?0;-mKVIOM19zo-+YB5H4z?BA}VCA4%~ zuOXOul{uW9IY7uEwo!F=KNf4~Go>GtjivviRg8HB_1zf5MuEubHOCMb_q+ZZ@O6MC zRdTR3$|HvN&wnZy|2tW)mj1k#4PA=??-l@uAAgF*JrVtG#V%vsqtS zt=@vt2HF;nno;}SWmW5u>VtUI{?2TV3pv4vA!tYRs!kox|2Ce~UQ^a2J!1=z;OQ*r zo@{iQV48!RtOYl{{3~O}oPNr%Jj67Tz*Tp7vSK({_-xrqCw}=-z?5yr=i+l%Ue^o# zB1B}+X=V%!Y2zOQGLUv+POi&6#*=E!>FRl56S^rKPB7XRJG<^#U)q6aQ~KZav>Ohp z<{#%JzY+tdSp!^)CsJIcCpv^J=SAO8vK$4iEH`hVzvBWEQZ+oB`2)Woog?Z~LXt5m zX&sM=!s7sBC;m@N)1Z`ej7Ot~#uBUE_;bcw%VgbVF>|wjY zJwgRf5y^B7SbNe(0>c*K+I?;*$t-kktsVMtp znCAnCjqwttI!%o|tw$~7SCR*1KW1CpCLd~dETa-_m`c1oyB2~>mt9;-YyO~iXD8Qt zO3;Ns&s4y|Oid{MLB``e@l`2X4@Jb#x20bc6eXM_&w*S@4B7RIJKwXv?b9X_08|p1 zKVtipVoytivoJhQray!b|M%=@8R8lvdLri_ddqST0y=inYBdWa##^}tZ50NmA`wx2 zUWfP2;U*|j)Zch!>TC2+IwAVZLh8*3};b(+o$L(Lv{W^W|~($+EaBx~xHr zU33ywUf-UJ`_h>g)-xCf{ICRu8j7(I?N;8`q$sPg;+tQ#Wt+`6HPzK$*m&7}>auQQg#Q3C8_~_Zi6<{Zw zG41v^GrG0b9%m45Jh;#I<=)%k|9U|-5U_{%ByI_}e^5@ZPLXNViw^Sn)DjYdne$f@ zuyXWrVKC(jiuON~<(A#+;R^(;do49NB|^D#eXmCp#Chm;Pw3GG0`C*9X`YfWzei|O z`8e^R0b=2;v#Aj^gx?E@OX)mlnR{GP>%grGP z#m+Gv-}&uHQCx;x#l1_NPJfC(E4C9ik<*+JI8Maet)&KF3zHOX@QV3Ll`NAoXTfbL z1eFbVPHc#BmPJ}boKZCE4qUc;ThV3R)c?+-gp;|X1L6hJFE+V4tE(Yu$OjM9i(fAz z54ln|X%9w|ua!QxS<}Qg@HJd z=)M^YW$D=S^Hl*s4skAh#Vhpcr31pZq#jfR($y%V&J88z@!ox-Y82J|fJ;RAQ?%z^ z>M}9Ve3lRkmGVuq^pY$XLujFHLuQw?%$Cc{)jRAZ5!c1)M$4U#%Tq{wrT(+jPkpKw zYRr@3`GSBE(MeYrXEc{)S#4&0ne@VArWRoo8c}Ock_+9wCDJ29{ayJJSCLcL&jS-k z`#^>LmJh+Vvom#XzW6z$S|?uLgp&YnYL1M~nmO?C7`M-+?er5lRi}AE8TtL#PGVIf z^vxF*w{n7mf(+~ihS0MY#cRui>e?rrSDuOIFPAPjR`CD5t=`Q;c9nLg_iQokxYJEw zahTEGHx4qW`E~ZH<=_?6bL719)yh4FjvG?=xGcQv$4LA0rg~_TdejBS? zXXEXs?R~d3HpW05#ur^XCP%{mz4pWb7JD3}CEH4GYu6%By?RO&K`SLD5gKweOSY?5 z0#Gl-Y@>42=i7ed!Q!g(tBOK@eIhxRYJ^T05cR^^!nPlR&G%e@bWp$7ZcpxstA^7h3-ac4j6E za-HF;=y@W#JR)12mDZvXhSpikegEhY+{L1NdTqlPII%sK!lrIF^@wm%oX(P@2d{g2 zHPbIAJ(qML6=Ly;P^IoMhv89Fl;5!y;;Lbj204p38@l9>m^0I_@sCrAZx1@W+5Z)m zDf&7I41i`IYh;^E^9Im32uf3l@0Lmp*`3}M4nP}ko*o>iY7j?tSn{UaJ9Z>rwDquF z^yqRSCM4(669h0iue3`84cM9AB1ehQDZgSM@Ov%X>aH{M!y^y7+?Bw)7mdL_n)$rX z=!d$EL)x=A%w)wDX@r2tN7=46F<%|8tN8gf-k{Y2? z;3;sX3-osWlKYO)F4`?gmabtQz-NM^Q>BfMRn0;f&HqgV#N~f+&qX^7?Po<^jYINM zc^-GX6~Gp+moXZmKbM%%^&1c#J=`dA9MeWaKl}@p`Q93qD+#XEWfx(z#nPzah=4;J zFf7_4dRdwLOQuKe;T19~nQ;I2;gcp^|Q38Ac4B>9)u=M3X7WjKqh-o~@iaDVR1NwR#GJUzrR7j`*x_rf3Hr#*9ZAj&& znV}cCkLbx9&yv(g4mWJLDl-DmLZSx40cCU-(f%D4$1JQ_u@Qz!-}xYC-E2t@;ChDS zXP&+To-LbCom_nD_%-Oo(&mkJSDa+f{l)I&(n$SX?%N5fYCHXOwD#C6#VDpV1Erxj zJZ=1VgDPu@=__s#s-rS*-JFt;7Jbex`nsXUmM40EzdcK`4aQtep`kZ-G@iz_k9yBV z52yxtSY#FobPrY}`n2)-K6tmN&n%Pzamf%d?v-%=YTE5f;8`ZcihbB3)?N00Q>6VKh56kxH^dBw378{S}y zs_*@qR{#qXwB4Ci8^;jFmI-}JDg2D`3Q0**D{?R|*ty^qf_>Th(IQ4Q#LdprbEkk_ z7;9oy6jhtaMd3Rl#s{qxbJY*i)vDhS(~@L19B3QTUt1P1f@e ztZcYY^mXX$Jz9qQr4WMm!4pr`=|rtP0y&F+y^|HeTB3_j)-A%$eu^6&U=77YLRVEs z#(FPNahKHLk9vVSo7PIvTtEMIU!3Q*4v(K=T6NATYxMkxK$;3DUr*z)fOo?d^IS}B zwCOQ5p&EB-LwbAAKnM|-^>e#m0>7ojwip=i6Utdz@9ChLmhe{uo5qD9aDp5nNe8cv zvrMstbly~Bci4D{jzhsHA~o&>X5`*?F}v8#t&p445k*aZ&=NV}Ewv9&Qzp_WGBjaK zWyX;5JCpUOjgB(j*;SH0UQLxQO5)J9q(ABP8bblFyJBmiCg+j0j{69JwYIy}z)jSf z6;S7m*|@6^pGOYKcT}k{%wWiP4cV#IaDjEp#=_GsB`HkkROp&^Mw1=a&jW8 zIF^zhX|0LRmHk9i6y{u-2;Z{|Rg@3EyoqQXxFBL<%QLll*xZixmIf%TWV@T zhswE54u<5Bis7!xh-$?AWvvFwexJ=y#gw+!7t%nj+zXO!-1(D)W*P+&fswnZ53i zW(t8&6hj@85rXIC0{D{MM zRTtm;nc0)sW1k*5Vxr)5M;SZGQyB35eI*MM1B+A;;QEqtPNWrBFml`~hd%~BwQl?CPATJB6r|rO|iB3f^hs*e`*+vWmgjz~w ztxJ$bq>dl^j{LXaJI$PjcDY6`^0{8J-LQOZWo4x#ZRa!o#vq4C^c-bCYyrPN_l99| zH9D3tFunQEk-%eu$dK-s_EedwI%(8qr~6S1kfiMZPbP<>`f$!%uqcs`K_+`mE4#1? zw_uTw$D{hs$o-UmN-Ho7)qZIojXN0oOF@^7P z&x$xDXu+N8GkjH`T8a*ieZcJ^P+!XE-gT;hcfhvUt@P-=2Rs3R?E-NjH{tp76 zA)7Cdv&pMLJ9GP8-~dBFyuZCk_~r%WxzfX?N5cc{Y&nMrV`_cwfZ=iG{AC@L++y^r znVekLXFQx**_c>!VBH>%hY|m7Wft+CW_KyPsEcR}-`gjwN)fYSy@09oT zYdp|LAy?Y4EMg&9#lJ)ao9CXdU=gxl30r@K%G{i@2uX0Z$S-qM$Ir0Gm%q9O&siFV z-}?D`9vjd=D>K(NgtGgs3^A_Q#Sj;*LYJ^z_?;utP6ID>z+FR{`Sh!gj{DlG);&#X zpvsS7dtSf?_XgEePcAd+2yUS=;y>bkqjrGP>$?0`e$-uOD{{f{BP zE?SD2Z~G8GASz5*p+(s27Hlb`F#1!2m{?D@s#g<(E=G#q2og|`mu>}F=45srlRY4Rhx4g_8cVE1PEijwX z=C?MOE*3h%j8H)PXzM}RGfp=B6^JPYyxPQ1RIK5E!PSUqK)3%v;H2dma|$eBoOspT zWeZ$`ciq~w0$&?)(~&03xu#Mhpw;DReThtBoL)$5Q8Ibqwf1>=*Ac`rL#CF{Tpf4$CRf4yaR%dlION7t#^zv3T$)z$1B5bgN!f=pUP0|%zMt1`N}!v0va8Js*EIgmORws>$A5s` z&X60pD?O}i+fRviF?8~WEA6^QFC)5H-ZH+Lug{2<-@->dg6idY-nRW_u4$C84kdcj zn?HonKe+0R)pfjqZ?vHm?d0G4?@x>Nl8i9#9XvK;p< zinsTLOXQ9$UyKaPVT`Kds#Z?Yw6k8eA7>wG<%(W)W7+ii=PQv3Ihu;QK?>gE#7I$ZQmGan!W z{Egl^25~KeXQuEI2D9+>9NZlr+Q#P^2U>;c?JcOu@`%H03Fvsxqv5qrl|@2i<5St( z%&hbE^ktVceAzyjOOtwc=agLH`mK8qe#X;)W$AR>Wd z9&ISzAs1vUm68WD2D%7*;XA?I`k5|{yERv@3LN?B#eT3qq9}P}77W9?x+x*r2F#bYKE> z5&kSy=$!Ui$lRwTHp8zIN;;#xm%Cpq)!|=@p@!c1H_1s1s754A`l?%D?_8wT ztfYY2e(JZL)XZdq|7E?25CpLJz9pYBFEZ)kP$>rGYAj0O9sBHB2qH5==C{9vL*w^G z9Q%y__}ZTj_--}7tM%KpDlPKO9zRY z)!_`Fd^=M~s^pV7HDF0#HQbF8vpZdS2q!Og1qMt1XtE$*6VFDHf1|+KF}3cy=&d5Q zxL&Eu#+JD^?guZCd4TinTuvM;R#w|s-gQQM&%`qIg=E(YdZ`dAa%;PDs=KE1T@v@n zUgJyR^BB(Xv;KErhfJO{k|63%v%1HQL0D+)j5g|3cWj@t(nW$HCfU#>rEm}9eD#R8 zQ-fI=RUR$^d8^x?QLWWbO#VGj2{2V%)vLb+kqP3}Z1;k97$+um2XHzp>?lRw3_vOF9KPjt(pPG|K4Z@?S72hfnnocm{E3BD)S%@m~tHJ!Q7NWTZJ zD>iG5wOLFI4T^~nHwGX06F-P8O^jeu9T=6st}LE-pp3g8&VLe`2jBjx9IcgPe^?o8Kymk@PMJzNmV*b0d|`@HF0=Og#|X z@?x41%%5W8L{2e@wMYJbn)>GkUtfW68ggKLQ1C)J;&8QwUk(QE8#=_gYDXoL2Y758 zxU<5et9$8(YT4vF?BgooNGZw?Uzi!Rb2^VQA2&Q) ze|+s?XIA^H;rZ6DTm1TTp87;*!t+P%5_J;3{1T20Rd;-p9-F|UrJ9W}uWk{!eOWB! z4s*xl-mr8sa7cM19Y>o&xUNnqVr%`6Bp*87jC}Lh_`0k!Ya)X)orPXW$^FYj)};o< zsBCX5*Q7zeAV2y) zpckEX)8Cc}0Fg1ugNBEbvG&%9Yh(19BkLVW7UqTPPzV=4$AKZ2&XjYXQToLXTOSUD zZ?uf!z8!H}2)TCWA$#TZj4Xz%*k5w$8ucYkv6)o?RJo^jlYX8h)*{N2_`|pH;gh z^&QJG2I)aHb4=6~{g@IcW0HBP6#(MVuzmCSPcMyonp%Nu%=tJ8_~$S}h0JWG&E8vz z^_NUpv{41DuzCtqguk0bmm2*xlhr6D)(u!s4YTD&HkG9P#EY2#go!1Lc3?A*$DK6K z&`(p+w)bkmgjSD9x6`{fL-F+#^x@-JsHFIajO`4QFwyvZvX4sXT(lZ;MC*73t!|fk zOMuXF$;>;`0W9}?8u~1r5ZH|zl$sxAwex>6{C(jgN{575KNi?~y0Q8o(hR728l^bN zv|VTq9~5=h^zFJxW{BSS_@dkz?kK%~@l=G%j1*s8{=x4Vd+7{i-5A%<8&__=027j{ z`QH=o2H+*%)ns>_s4iQwg2Cbp7gvg69XfUh0Z6UyyAWD%C{kb!Aq@&_h^;(cc-4>| z+It7MM zMr0ahw&t5X#hDZ!jp9vZ39=%XuO-*=)ZMu}9&P7)4*+(?GPL@su3H;Y^(;t#(df6iVg@i(mc z5HYQf6Sb51y)yFxyhNJ)YTRm16WV&(ZoOUb^2-Zo>Ek;6=f|sOKB=$sX4gJm&(bd* zWVLoolr-eHwRjfTgjO_~GEcOG;?7Cur8>|C_=WCkM&5{dQnr;#HSiziJ4 zNccsSp}&UlHS;yg51_U2T!z{(vf0+^@;|Czxrn{6H%Mi^-&uy-hr-40!v=%_@nMPk zu(yyu4lSRxJ5TFV=s-dm`Hkc8_#LpVOE9AC-B?IFPshr`MloV0SI1+-i#60|FG*;O zR6|eB#Qpu4h-^pMdY5|5}AFz#K0a3=A0^i%~9UN>_&0?$U%`m zDDuZVLzKs){V+Sd?k>rNndsh>L@cPSh!4kM9uBi{uKhI5UcLNh-3c7l5AmwOXUPxu zwP%_LqfIN;^(^UQb}^V#D@3o@NZo9&ME<+>Z&;BCc;|r za1KZB_G|ti7O9XVtHm7m(wm~#$@5=Z9bp7lAXc_}b8RmPaYdsa>G5P6opFAp&CMoi zPOVHFc}WNSigW9<+%ur1oIh)-QX9zMjW((jLlDt*%JpdivBKNgB3g$ym-23b#AJ#o z?eOf`d)Bp4erlKSTwzn=mC1LnMJcx057k6>O{T3q4 z1sW_9@5&Ry)rXW&&i=9ClwHo1|E%z^M(_2 zpG>C(1Xla|N&HV$gHmhwKKy^b(R6uI+F7OlU2?wTMR(=OJ^MoKw{O55eH{T@k=KV;J4Fqn;}zd@xoRuh(2A zO_SpT#|kmcX{(17_TfKgmbEE;77=alWTNk)zkr^2Nf@!JVidt3j)XD|Rlr~B-fx-*xOYk;&biwjYSZgS!7s;>>ZX!%>zIQdevoQhDK?kfY z_udGdN|aK`G}?O)(5TC}@rb5CUJf1>_TzdZhbra3g3K6y7Wt{fhLqg7ZLjou*r3U3 zJ2w2Y>lml-6FBKtQOiy|ECt|y9f-jxE)k7)k9>T$Lt6q0dBCHbNEMLzKfnFca;m`DhE^{ z8~1i8501DYO+!^ES31>Dp@?0X?&Arc!Hb5|e}?3p6#H6#Kgk{>?(#`lK+9-wqjAuN z;8*`Yc79zKzhQDL@rlwxRrAFa0`!O&zvxtlNCm5=5=TV6r;8r2$#rW7 zr#^4{4PTHFpqNS;gbH2^*0jKV#kgGQ;sVZx+FIevOcC3n5bqkac=ufcM{CDfQ5Xi# z0FdJgpL(G40DiD$NODET*M5^iCJtyB@>s9aSMKlLHYXXcJ7xS-R$3FoAMHDJne>^5 zZ1=6c8wfwA{ez;J9=yYS;jwhU$ORSp)2M@u<>S>kT{V(4Mzmp$KJKK6{R1Mzr)}`q zx+?wMc0u1ycF|)8Im%3Kr(QJMEUcEs%7*F|Pv{1?w(Jdyr?K1DiRl!YQIrA?_pLVN zQSaC{IrOM^WUO5a&D_o}3btGYI7Z%x4$r*w=bxSGF8fD6f1#%AF2`58ORgc!P zyAP9nJcK_0kZ~p{GJDrAV8sx{!#VW@D=%&!0Fio+P<&3V>?=+X>!0FF4o_i#EN98f zX3i8S-e|#V9($}k(W)&m4^wF$szj2phT`7M?XTo9{nCi{wN7^B038e|+A#2|^bh+D zS5{YgX-L<;y-TuaioKG%Y7N?ArrlKjy#E5Q9{j(7bIrryM#~lbihvH+h)v&EFO^Aj zr4{vM8pyxqY1-6KlHlZq zgo0NomH})RFJP++Y8e50wXH3+B;04MVYqG;$bCbg2$ z5#MQLZtp}aikpWwSuKi7`%6MGS2Qe|%FRK~sV}B)BPX_94J%*(hodu$dS_-ffB10k zU=9DQEq?V{rspw-o+cer<`oi`kL{Qc=a9Hx?x z{!Uepb5fe&siZrv%#6%p;uc}=j`v!zS0(_7Zrr3{o=d16iT&ATIr4}kE8@`Z|7={` zclBkFbmpT_ClKj2O*Mqy;Ggd9{rKW?$3NY@#H}!rCt(^oD%^H^b*MI*A)S9xtuYXcnx=NaW)Zwr`Z?WjE5wi%L736U7^* zC5k2_+!g){yN8W8eGgEXN^muZ<~AjkDi>;gb}XMOXp^+byb@a1rn@y?FMU{4cpoM) zyTlonB$6qve$o&J^SXpj6D7qnk)xUUY?aS{#p~SD^di=-_RVSr>TQC7$zDeM)RD(` zq-d+e3SwhAWp@P4cetQ@XGC@}9GMQhwAm|~EU8t(Dyx!J%i#83I>YC?qMC}UNV(|}1*)!l z9@yiw_<2xJ$4zr%OzxGn$;{xw0cGpt*fyvZKNNl{B<@(nr$gZfJmqUZ9D6*%P$F$@ z{lAHM{{cbg8fDZJgwk5r~M`|>1vtTE$Rpoji9#Ap$@uS)T*!}<%+D@XVt z zPnhOBSTPw3UKqHtbC-Qf<|=e?Sa~3Q=MfrM{9*O-*{wF1dBkkhqCBm?rm-S&IDbOI zXP;hh=IhDlocM`sM;z9Jh5oA>|ZnYohC%4tK zJlVNQA-NH>@V4kYHhIXy5ClgE33|j>dS0pRI-DO?KZ)MoFUIg;OaA{ZfQwT?l{Vaw z0A)3w4c0KAN8E}q<#J`LZ5!5pB5WEwpCCyZn;e-|49f&Gi2HHb)=x%)gLKmn8gSR> zp!&#Ku0IRj-D%y0FySobRq(K>OHP_Dl=jmZUS zu0PYME9$ip@Ik@f2SA0AYSFFNepwf#_`r!zcDC+G`*G!WwQT?vS93%36+|>|!6zq0 ztlHK7c|iCLt&*lCqa>pO-snb|@Q8e)%XokR=(#6Gr-e=rKy&|O`NbGY2N?mJH86MZsJ!?aa!4bpwySQ%xDYMjMaGJK z;p_8-(tNJQTLAT>gr?|f&xyBcWadA5&|37k2MjsCqk;Bbox0*k{iLu(r&ss5>0>m; zBgmAK&heB*MG_co30#PF9H9+mOmyt1By7oWL9Co;8&*A8y(EMTZOal-OP^(k1N>S? zUIlROC%)*u9L0NPa7M!(Z)LM7Ie|EfhHEyx@r<@q`ttSv{@WI%%+C>(9Y%(kib;u( zz>HE;na&wThm>axmvzhU@dnQ1T`gY1KG>Z{XLzFrPbRz?)3uGj;S(PEmoV`BKT_@Y zVmKbK(#$g)^GO`&M>ex?=EO<>tTF|SXqSfX0}o@tvxz~(`)pHdosr7{ zhcvCRenuaV@TLTJz3~bf_()Mol*Kho1&MDI;zm)Jqfs2uDZ?fr=mF;Q@6&41|D)+j z!=d2*wrz;96dB84Dl}yqTVpqtkc4(bhzVuJ5?ROCiZqr`+4o9i-zkQ%49YT-%FJXL zON=Kp_QBwNp8s{dAI`V))#WGFr*DXrKLHiaO1(qNVw(IFg9|5CtZMwA8 zRJ+%FK~c$CJ1P8qO>w-N^!TdVEZr@ijYZTX7jSA*N3L-=RFY3zcS2_&g%2nRGGcWv zXz0!?f@f~+<`cCkL@lWbI@DZ~*Ufo{uqKF_qlHzz;B#3mhaOCFm&M%<;+$e0D_?YP zyS*!o<&&9nYWEfD7DzcOpIKoSOswz zk!Db#OS8lju>8*Pyn3lm4JDX>0QF!losRf&IJC7KnKv?H4vSj1LYW)CSs&kp>Tx@3 zOfQyG@+a*0vw{v!TxqUHc)2L2(ir8suXP4a(=Ys$GAJaVQ@d(3ilg;XZB>HahJEf< z8-BY0Y*zlOohU6vgh=n+t0jC(7{ZZL9Kh#YxhBvXIc?vF=V*TDyo-#vKYiB=*M1=H ziIH0$r5O^etnf76uVl~;>2y2%#J>J$_Qk}t%767vMIYZIh4I0+A;=`Jw~_R@tjU2a zrmt2;_0nX}_I+&qY1UI1<{pc`zHup8WFso##76k`Z&-Brz;oMHZ@-#iAjFlPKKGBk zWk5;V1nUAO<3yxc4OV=SrQuL|D>7W9jbH)7yPTX0{~!+Ao5wME-U-yqxil*FJcl!5 zxp@stD&Pw!SmEJ+-G$p0x7im8V+j52+2;P$E^TZ`o~=|1 zipL0*lRsZe6moQpxbMFkHJ2iY68#3f>xkeOP0+qm=}f8h$}601>BG6Q7;hCk6`kXL zh0i64XxPnvBJxS%6%s|Fh9z7*PDk$DbH?7{5~1(rT+f^ExMk?5e99T{cjU*W)*p8_ zdn4!mNc;cT^wij3sgJe0HC^t7iC94y-ArR}HH+2;qaXL-SK(n~Tp;ApO=}gHTOql0 z^V)l&7iI)N-~WfuW*j#IZp_|rS331_=)IBHGXG6QCSWN}BS4^aa-3+S?~o7V0Zi}3 zsh)dWWW@iC=Pnn*3N9a5U#n1j1K<^>sxC|QY6Q9iqNm>TIB=lKY&;mo?Ng#JmoK^U z#8IJOHGXcLJyjED7oN=FkD}bQ3Qlpi`A<4>T@7HyBfyzcZ?bq=x3%tb`QktUTkpU* z{I!abAfZ9g4tO@tfQZp1sg;ORlB%cJy~gV+cyl(;oiY@jHG*`QhHV;2B%9Z7bvD-XLqv-4a}TN5H9)7wrUs#0=yw?blGfdY=vfXiW8q z&w|@hiviGfD{lQp3m~2;$ZJF@L%x_c=XQR9?EZ_7Gv16R_-%Q*;VCw7pR>_PmA%)m z10l{P2&Zp(hkLWYs@;3RQhY?)r|k-Q@TP*fqISea&ZyvF*M)aXb2X5l!rN*i?$IFl zZyB`^zbkjX53J5YY<&uG*ysE1_1vF|5t`j15nb*@6)rCPq^9urVWx69^>}YnAQ!K} zBc$81<#01gX6{+RMoR$O;AvH)vrPU1ck{q)akz-iuiiXg_-D7|6Sc|L&?zUZqIMI) zIC>l(k7{9wVcU;5j`}*`UaU*lUoL(=8U1Xp}DLu!br4+S&HAB^Nwy0gtOkNY8n{%8ujG3L@Lrl{MGnJ9dyDwUcz z20_!!6LrO>kjO)=qm3GR#s|IWwm#EMh@6}-!q@nsKfrY}HwfikI+~A8HN!o`(rtEE z!@l#X1WA52{J^XOl`ny1^3qYmZ1a(ZDZ77;ittvho8}?QO&Knc#M@-ck$GF!kv>6o zt1h2?OS5CFE8~UTaFwf?W=oDkSBy;`Y>IG=%+>slA^1nRbX34OMrE+^A5BTR@3yZ%YAvwyVTqCRM)Xn&8rQjlZsVTr9;K(O+)8%PmaL$jcI!>6^JC`GTH}k^XZ_(S#3b%uYjGyglqgq9>e)ybD^hEbp%UiXSHTYjP~ZJ7ZV`+} zJp8~gv@zU!;<*lr?qsDdM-jvyX zZUoO++=8``RChnu(9`8)#x$JXKz8bXlkXk0mH*EO@?cA4_J2l@1fDAaXMHoNhUDQ! zpRp$lymdflN77VOEzdbFVNR}WF8vkTR)B^U`!P0D!Tq#}_!V*vlF-y^%W;k)v;7zd zkj`bt`7*}t>N`^TB9*TrUeT#MdQ{;)2)Omw&y-Wpf0)1CXe>3w(Q(ZFxL@W)`hr}m zCS})1v0L;OAL#MsI)zITSAoyb&DJLD zb^$O5`47Gnsn^W)_to&8_}Eo zr5SILPb!(QuQRTG@`Nq(L+CMoge)2_pCOfqu#vH_0H*ov)pTGkHB+Gr+zcn73NyqG zGCqHiQ<1h|?HcrP>Bw47%)HfFp8+b4EFUwr)@3sitq{#Zklmg#h~5yE(sY?Om1qeN zLH4Q51R;T8=}-PKA7RzH_m8ORzEIT-XgEE`Ws4--9(Z}kyx)+yI-nme6-^dbB_h~+ zaq$xE@F_(y^D=GbEli#KS8|l#+>blW8BuON7mH6Z{ZxJfz9i*rGv1nurUb1}1RXIk z%}U@KDn3YO5GSj640Ld-ag7soLd`hN|3asuN{j@AR&w58QKTUd>vCaavn9+UKgT&4 z;fZb8XH{Y51mtv60zC(_=~*PYUgtkjZH^`ykc|(_^LwR}M>ASkEKU7S19lf&$(1#A zPegP#x%mV)GYZ!Jw3XJtF@VBvDUo|V<+B7MyuCBNU#7cgzv!-eZaK++M7w31nmRbD;{}S9m z?4nQIkV_5h)9~rSbk$CnCEa|GY=m=jsBx98dt#y9I0lTMa1`;$wSppzHpaqG@uNS4 z)iw(q2x(!*s#5e?2SuZhr77LbOZh3E7VrF+Hqu-4K zy2OXfPHsG1Hjg&EJ)q+hDk8^Rs??Y%)F@H?Jm)VvbK9jsb(@2~oy(al0T z2P9%&`&f8~TR>g;(g6+OU@fT_2ug$}EO|z^+Fa?oH3J6WfUIgr-oJ zWYKLuJkyNeNbAu@prGOZ^5kbFqHqw>9~ZZv<4#H5U9~@BTn)?2?atpO;3VK<&UqOO ztIwE(D_!DN8D*GE$i^8_U3deXd-~`L>L3Fq~7wG0Oh*OU(wNQ%!iV}8Qmhuk)rKp$@z$f?m6PzzO zhK=}>J!94P>d)8m#DZb~V*$s`7TB!wRL8mUXO_83U`$oQSm#Ot-HL@@7rEC-otBa} zFfHrT=YQpX&HW)+{4cyrg~89**Sq3h#E%3J!>$3|f|Ns@uL`vwO(51%dg&!c&Zp?S z5Ojde`at(t!L|sTE$R8HDD+WQ@(ZX7k|zymdEH7PZ2O33X_WYCF)noQ?41J@f$+GM zT)d(@5&X{ElC)zUzCq(JL?^6Z8%_RW>$zY$w)@5 zZ@eq%HH&)=Vh$$YX)j-PwZ8Lw1mNyONffNs$?x^S4qmkIa74{RKX8~w?xib~iuiAe z$Vm^Hu3r&Gzl^ncmtlRB$s-87DG)JuvxybsA7KVI1T`}p;$VZPMc`LjGS040x!aMK zJKf=hycu(dd9$6u*cw%8v24`lj4}9>*gE&uQckCy`$0cvIq#6%#m1sG{{Ry#_#c&d zCshd4m~C%f>|e5hSb!H9PsT6gWE$92i+%%RgoF`kvI@8k(gQc<#uRZw84ks z-*$@omS9IrZlFhdtp%3L#ph~ip(Q{CRn?qB+Lrs9yD8p$>&y41hdCvt?eGp2!#4eR(k!7Z}_MsJP|NXZ~H z?vK=FKlQ^Drw`|b4liMlyN2FqpduUcc`Egk9= z9HNxS-Y&Z1P^q>=V(?53TuJWxmkF2?uXEmRyK!pM%GBZG9RuF-vVBIL&FCsm#iH!u z;p87(MBc8Do95+rn;-4BV-ko(`_-TM_$HM$8r*&P%Hj0%z^5WY_Vp4rC03Emo;TjM zX4{s?e?Fn?tj`7ml-PSmn63CJI|vlH)=d? z0T0;nMFTovoTCy&a2~3{B>7&msLFQczpEl>j%Ek-H7{_taZI1VzH-yewIK|F>cIgU zf%NTs^!J30;b=Xq zQGFU419W3?<0kXWJB}j{>(g@uF?tsG`B>MvF9}Iu)~=R^>wv(7L!xLHP02^wa)8@N zptCY~+HfE(uZZ;24*VGYdj3M#D_D+5%E4ZIEFW9Js(+r*7S!4xzLva88h2h0hvm0g z{blrz&&mAgiO5UM=4aslNYRurz(F+w>Sr(T4UB$|PQC;r4 zdBlE+7ctiLfkxI_WNq<8m}+&bd7~zHb8F6%p`8j=0hyu3N|ELwxwq4%fhBaGF~_}bL#E+{yTm!xFF zwWu1KfB0ZcEz+o}_x8SGtbH1f03#4{{puO_^Vu^tHwiE1E%xK?$EF7LbE(E=(P6D1 z?xW3eM9L}q;b2bxmPyc(0{%Bjt=`o&dJBe>c5 z==YanW{O^-X9fkN-h*5eAI2?%UUjHG=ktdNTt<_Qc?_PM$Z+&$hX46Cp1|d)DOyRZ z6}w{7t8ijOQFZ>G_fh^#$&IeEI9dc z*Pig!FZmu1j$EiBJ%@)`L3a0dMf8#KNWs+kV}6LtauOovMN`Uo%@V{46^A^g_6cAU zrx;se)ksQHNED)jh2f=;Jp~m}fOYC~HQ=~FSb!woyl|lfYWoYEadp2(b!GxR82hGs zBg-}7fr4O$6z~@+$E7K;r&K+v);7vD{P(MS5YbEerDKbQk$#i}1brE z#E+(Ky8-U#Fe`&!F<-lLn)jf5k37oo*&59Qx7P1Zl!4tC%CXwbTj@iyGye+hf1vo- z_O7>(TDm2lE?sFvoY@!(=~$lqFSB#q0|K9)d8dOWSHTLYPeey$`-kTcdp93WZ~c4Z zqeD!pgDt9w@xJ=n36z4T&jfy*vC%j3y_dN;@|SGjLcQzxZOeuJ#?xwh)<>}PLc|gn z)%-v#?a0e&-{LdM^XR?mgnL{X96vw=xE-tMPLtFEiKay3O`Mt1dDM2_Ao)E4$Jf!@ zcapV#QuXvQbq)z=%s2jsAqXa3ZvdzSG9&@20j4Xr4hFxQDRs`JY|SS?`!u9{DwI~! zRVO8bw&s9Mr_=TbfYgt&`bzA#`wUO}g=em>>ohJ|Aps(H*A8|o8FU?yW}RAe!U`N> zavM2XR1BD4#E6A8QM3)$5?kA3v&1*zY95So zEE1*m^6T!$E7%2GXi#O~-=n5nEgbpxt-ab<0mK)0m-=G14mE}S)PKA%YM_CI*+@R5 zR^xqg`uQk~k<`WlH2*=E#uTMFsQuww;ojl?-?VTS?EO`Z#jl-ooth1FC;K-n5rX`i z0?bom0D2@z_G*2hUlv`y1`cJ?2B||O3&IVnowoY$ce{aJ3<@ZL6J(PKXBwn#Lqr`GgnG<7loA~dU zGt0Si$1pVikCOnj?aP17l^I>uqS*&utnp^*=iV#pj=8E+y9>J3E;O=xo#1}SW!yuFJ0=pH(+-WY|X?+&; z85HX~m$8i!j=y9wP~`@=*E?viU&f8~US&P{53YA6ZP>WEsbw*DAVcxr@*Xlj|ICUx zaHo%37T7#!PEw0~B~!=Wd@u0zq2ijTQ7uEX3Pe}5>^Oi1jJg^*$V_sl9i+!wI5b|$ z2e9UO(jvoD+j@cfryG{}4is-Jo0KMq)GCODLFJPw@WIi(`$fk^`TRBa^`E3tGkKn> z_Er>K%gm+-{&YKI{Z3R#({Jc4l$~BSD;Kt*wWIis7;kYKq4Nb>(9`1wUPLDLPr|(@ zPmi+hRWj_ekPK)jLqM?p(X^RNLD&s24E55b%hn#1KhZs7T=1~xbt9A;eUtyf=eJ^> zUsweavunp*$3>4n$8Q9>$Q)VQsre!hZi~^*=iUOFdqcXlg^YXt@a-7IFW?qFWetdE{K<{agXo! zrkv8Tqm%%IMoGQ?(sLlg0hK*w2IL#x6?{T;?P*3K9E4z<2#WPR#A7iEPT9_>%@P_a zPd3OP+kq3wlW^>yy`q8k!QoCP1WUD1Tum4ML`o(iP4FLQ<=WXN!6qrUy3Dg5zy2XP?Qhf~LOO{hhlW@w)n zPz?A>i(Gim)LACU%vA^mUBuP1VRwf3W;pqT;qM^34O#~{RffcoQZ70CXxl=!(=S!r zldc?|o*w>#SM4C~eyTh-@R_?LlsOZNU;83ahU?&|Si8(S|1sr|mfj!!+Q-ZnaCeCx z;e#AcCGt#W6#2H-XQs|S7#S8GZew*r_D{>#D{=?b1$w%kfr+o8xQs@&99}W5*fl+N z8%|x&xb3WczItaSnXf-OBH^{cE85WL%uU0J1OrD|-^EL=s}1S2em>p8y3s1wParoly%`fv;W&U32F%}zJbJB@yFb*4EG-p}a@Wfc%4cZE733`vVkBgA2McB9 z@x(vvk%+QK z9Bh&lG4SssHHTh#$Iaz)!rMcVV{TX=;+JUJTpo&fKm_{>=5C40FYy`g?S}bk%!Y2; zW3lPHM!{(?!?DaYWx1aTm3JGg5JnVv!Lct9iP8B6R(8_9rAgNkH*G7({6Wd zud~Y1jP-p~v5Z}PS7ZK4A;Q=4F<{>btLWF=gqnJV+W(w}P=2N=Adz+w1w1?H00`%D zLl*#3c^<~8$G$%Ge%|DWeJel~Xa~3(=iXl!zf@@Q^@LhL7wDOQ2_U49e**A~>(c1W z;FxK%wx}_GiNg+%{cq;nX z@jB^(R^F{0vVz=1gb8xM`^>X_PpCq`^#9zRw~{vi>! znx1oni!pwLx^ewfYA5y|asoJ{8=GRrrIbVKBV4W&(Ko_qRHLgX+(av?J$l0PCc<5XL>N`32zRBw_Hq zvaq;BIV$K*`zs~ux3lKf5|VnndR{pU)dNI=QT&M z7kLnG**=prY|ug}E=Y02(1%O4b6PE-2R0GZPe&cG-jARLI9?x%JS169HayjCll%%` zXtgpqJ_UX+`Uw67gg&-?#Tf(k9I?+ZUzaq~JIr)(IoqoQnkDM`>i zW267-;jUX4h2N;5lb9Vit+jpeFYSEw5;>@q+=Z1j3hG=;T@cv(^?t@+_jMamY`>q@ zDw6hPq|3i@r{O*bn`=V}&)ZQ|ms(`>NKikFOlt9_oc34Yn|AnyL=Qdmo|q#C&5fla zegn~$?{v|1cC}W9xd%@}e0qi{4eq5W&(+vcR?m%*RZ4~F2Cc-0>+nM`jy5BY8ByM= zU~T26l6!P=i`KlaCH!sRe5X345=^^tVPW@lV_d8&@B+I${G^&X`Dr2QJ`s!{KZP{v zmMXfvvEMMDuJ^DDpE?~TC6m9M5pNw7LT{Se1;@b-f3cB%TY@Xgd51LQ{4c7o?jXa* znd8a$xHRK56n4)sctcT8jvDV6i|&!?47=zszHo1RzFgIbT<~W@jeK$Yxo^D#DJA;T zA~1hZELkWR>g+ix8KeMy^g}M0RSuoAQuN<8^IEu3Z?}M)Gu}7FH9yn_XnhVXDDbxO z`RR~Hdn^;vj98l@M=jFM%NF#kjS67B9fLaeM%5$jZWoS4uguCv)Tb2w$s5UcXmn`A zy-r2W$`}W~2+|Z#%f7|d9Ul$Hnjv-!#e7LEYYLVie_d}0wg59g%)f!Ng#6!-&ohfG z{)l%Q46c5?LRs;-^M(B>TZ_{fCXbASj_WTJ^(Ho`B?oSI3p1u<6vk)7=9QMV@}}7f zM>43k^rzt=3wlMr4FBbNSK{MkrRLZ%!6%dR3S+PJTM3{4sZJ!lNiqEI3v$Rt z{9)L-N6N4}9!`_I-K?Vp1X!R-s!+Ni1zRbL9}Sy^o(XS=k72Za2z^EOBs z{y(M;_U>;~-+H@^0vL8ao4+(dFDY1wk*|&&*DHwh3KBBjBXS#@z>U?^aHIe{!6MA) zHJqDFGY%5?Vhb34DzY5Wx@0KLtk`=j9n~&r1|@=}`#fA5d(tF2244B~t{_9Fo1KJP zD8q(kiF|o(#gX7%t2|q$ezN6>v>C7(wih@bX4#{Z<@}$!<dsqq>WaEb zt2k1CrXT8dB0!uX(XOuqXw^FKF^6Fo4uBUK9g`rFvTKiK@fWYup7Yt%Iw5-A-=kCr zd%)9>!L!4&$aQVACkdh*1O2cG-W2_&ck1NkN37RHH~x}00;s$=QEePi@qL_Ko-#ab zfHIoXWpg|3CNK#EAYZg+oYLZdrQ@#Xh}`W4!UPTXx^OxK6(`z@dUwD< zi1Tyb-295|W$SdIE3C%8NVZKO#_60>Ahs2KYSyOS93*=7fI;N?bn#L^8q`!TY#v>g zi&feDS8TB^R`7@yJV$=-dl*MRf9#r6P1WP1vX(chUSzUx*@u zagqD;MTvaH)2WP$B6*faRVx)2lU@rsC#BSIE9z(1qHp@QKa{tw4@+?M=kD<&10Xb6 z)ZmGZ%fVFT&n*E$iG9E%S!zF6|NExpJNK~uy~DdyNtyd#tO*i_X(~KPBy#JM}(#AvC}w zXBHd=hk5gdZGI2yOv?ZxT+hvz5BpASpT*7d>+Hs*%l7tP;k&aYKJBxTE4+)`$cFJ@ zIv9Gbn6saDuZ28RVJ;ybtWsfNXS8rpZ1}#PdF`&CA(LFG8Wqi&Ni0iigFp)RA2dzA z^9RT^(km8<(SLzQ$xlGS8+HEGjBZkLTK6(^#Wqxz0n{WTVC8*YTFdK zCeUJxiShItiE1R>irPsd&=8j9>ti+F8loJlXM*#0tu6P$EhPv-@8*mVgrdiZI=TZ8 zL3Z2iC0^lCnY)Y5pBB#@&I12HUBBO+SZtG2-s^o9+~2=0;W}Gpp!wU1^qzh~(D41{ zr!Ub~T=S8%40OSFojh$UWBMIu=gvPBHNqY1i>}$al-V0~nvz(jry>##zYO7jYoYjD zNK2i2=1}`EmSq)2xfRk6C*-1^PJYTd7{+{ZdyLRUr3sZC{PUxt$A7I4^%COM!eBNe ztgc!6cytD(8|26vjg2nw7LD#K^+XeD{_8xG!;&XrgtfC_zYE}M>ql&;Q?5gbvQ;C+ zP&ysq`gBq(S1r=$+wS49-# zt1`>Gbtx86BB9I3NM=Phehu#nU*K(gkKr^%y~v2iFI;zvi2AMUCKdD~f3)397;87l z`*D{uYW}h%D{Orb+VACdbjU4eNH*PjWl)yxvZn(5MBqjso&~3)xTO@v7<9U|P z&m!2#_>(Q)z+qUUzK)ux6<_l#pWj+*-gjeD=%$gQN>rGEX{IlF0(-aQlmrVO?-3e*Gd&kMgxCc=oqiEB&OH)1paH`0Z$QYWg1!iBIn1s9wv;?2PuU zGB-nAZ}39Y+&=-%(WJzJH6c_V?U)<=Z9s6-;VIs?-01e9%Q{MK+^_56AecMgU2+;$ zdzLpZcJ^vSf(>M`0;B(5FDF7a0`J|1Dn+QqHpnB;(?78Pc^KDiicM4?^6wXQcrsS` z=>iIqp=xxAWbsWsiR8~An=a_Jl#}tl@gZ^Q2Rz+e>IZj*$hW(oBw!0xt7;W^>IW#g z)H#MXK}m{$J0<4tE-()m^5JkW6BXhzh~im%9uh!My_CGY>U5E7)Uu%OHld@-?fRU{ zaTFO;GmibZ@^ zk)Tb$8ZRG<_ARUT#u?~J@aG;#IU6?09-( zFdhkZ4Y^H-k5pB-{9Qq1UPXfwa=6@=ewS=yOyM**qZKwtGEz74-f-_Ki}_u3$2x;& z+~cAr&k;y-(9X=l_%nJYYuK7TcujLCf&QedR8>yJE>pVv2~zp9hZqK!Gu z2bsdVqS|cmFwyQ)NQy`;qw8#`@iMQRVWxAE7@lro1s5$~eH%%DMz)-bfp{u>J_UOe zQiOIq`GP*{lrg9_z;mR;4(77laZw)Z@C?=+S+6)?V+xLJxsZL;W9iVrv}G2j=1B(f~{%jkLL(;_1!)R0jaRZzMy?3x@hs}|iq?o|XYOOFQhILvT2T2IkRst%ii zs)m@2Wn#ES)^>fIUQA7WtB2{=uL>ZprD0-PmJ{pvMtYVsgLG7N4>$~?6Gdn)tkune4vh{-k$|eDUL*7XpC9vo7cSW=tBuF09N;g7|NPbe{0H~wr~!9Kc_wjb zNCL%?jQ=vrO^LSY%(9M1I5Gtag&N62rhKKcNwzd1@u{DO?Kq&M}Rds7=G zhAGF#zXlBjt*gUPYRB+(B}IXpFk{T%{Ir=5%ofRf=k}+M4vSw+pfDL4y2yC% zPbMu#9p|Pqz>#VQbDcx9XhxBX1fsA`v28%i-cZA5t>HED?Ba#AJbC&Jw!J?7`E(f+ z&ih)A3$lFU7rb&hYG-oh4+^>AzubHF8q&K*@D%Ga0@#%-gPb>y%%M=@zss@ z>>sI)FZ0RL)_$zc42NrecRQU%%n52LP058=)WEJ+T?^B&J4E-S_;2^AeRlUm_T%Yu z(w<;2qjm>kPYT~%24n3FohQI#qc-qAR zi8lhP^dBQ_y(R(2UUywC{>l5K%p1_)L(c%qx~aM z80Tfpd8qh`MlC*JSjI2Q^#~Y`S37RF51khk&r7C}R1fIo$m1hO7?qSN`0Mz~m~U^7 ztYQ&v_ZheaB$t{bt@KA;n5fd8S-d-S_Rh+`oR{knZZYF&8Ty*6Tx<(A!-BsGU~gS; zO19JI<&_v~{Br7~ z@j7|@LiTdkg~G0%ud#|5itV6Rs*fT+zlfsqAECOBOkQkS>l1kUXRF{D-s@e7-pjJC znXA9gBIP^l}v<&cg~F#eqd6u{pn%boBQ zuF!#g{-P&eCcGghU*RTg?;k#F^BX#!Y*?AI|Ajy1{j980@i~|!G4$n#$E$6#yBHaX z{%9|Q%Kt!pS}hipp_(acl;Qtrwk*Z8KL2r`!Ak?T*wNQ1ZrfkBDJFT4hsqF&y>uA;1lhr0boC&)eHkAPi!_1c4$#6w5F4^9tYcm~3uWF- zfeIDXMq?n^)WQ4Ol*6#p7Mj7xikBgwuk+hhtL|YY*6>ewyHqBjTGny?;YjF++%R<$ zDQvyE)U8V%x#oge-f)H<199Q#+2#qMKFkX!u5vlVJIoFB&DuQ4$W4vz+j2h9NxQ`M^sL!<9Ycuyx zhd(zY7Lqp+nFUb?dxBd+xGfcPZ>!Wy~I~IE1o1@^XCYLO9^zgWgM$SI{%}i;!DGTBCyDJ&}#W zE1RerqAPsPsMGc*A&StvC&;OzG==`cBmnbX=9Mem#`UIone!%#$Zj2jq64|l!y$`7 znjr;II*VO4R*xLn&zuK^Yc>V(dfyh5jVUmvGU{3QeizIj>(95E8Z&B+hOIGM%UR)u zn?k)qbKeNMR@yx5lvr)A89m*|`CiNNOIP~I=nI9^6JUwYB{ma-KYKmkkz{X`5}?th z=M7Pph~h{?rGOY4crgCU@7Mpi9w+uX7Te}tj9$;x9{m=&hIL+7(@C?D)nR8b`Pudpm(0R+1TE8X! z{btFe#hDKB(*R7L6C48YwKMdME8>{H&3)SZ$!BJcQ>cbP-^WiF$Ch;2Lt4pB#VNeH zm-9*T**8hBb0=S9q;uRhsg5SRV3r}8Dc9lv1 z%QK+B4{;GmvUr8sfcNvfB9g%?a+pmH5DT81pbdS0j5~=931DRg53- zAtviZ_@DO-rNJ=v#E6uk;x>KvK*pzlH*8#f{omSGlD;cWJF*dW|wTh`3@*!WE zRk-LDk0A)ra{=Fj|E}B&kx%zfMu}iBA5pZUsZ)HoFXD(u@HDExj&Zs40RbSqU7Yey zm6!38yFEQ}UU?VO@GN~=h8j3!-us@62m>mQ&d)S*>gQ{F&s5sfg7o*TL86qJ{zk|m zUUNiWg3n)CAf7(0CwIFBEhMfYn|TXRKdA?65>z zo7R$Lm$Oh`<2a_fq4_rt(~X8kDaY?n;8Q6Kmk4DBn_GV)L${PepM|<2$5+s?XvJ>z=XT3Geb&mcxS^Og2Xoo`> zW&fmsR!AkXca)a88xG9uSe7-rgWz9-u|_5RHV{>Ha$UH%J>&P-AQ`;25EhUdDgr;^ z<8^Z|t%$SSECpAQ?|wcg3sIu6RUF*#@!r=`P>Z{1!4%IT zc6O2ES2uK~1Yp))C(-Vtpe?NO@J_x5H(N=XEppzNG|{zKZ)B`%M2dYVFYfy2Tj!C^ z?+KTr>V>Q*DYi~^=6RZ(bXkB`^gLBB{2z13s)1MZYJ)R#8!a47>#peVT_gf>qz01e!3<@X5{ESm^a|n0kRHO>=GL}p_ zaYg$hJKi@iQNw=>n)6I*;8_89(KW)oe^zvyb(I0Ko72LVdjDHJdh_L#=(vrV^o;Fw>&2xMWnofD|2S<(YU?H>hZ@UAQ_8Zyo>} zt9Lyy|1Cj&;|8#0WTsJ$y>y`}7E4eRJ2Si~T4JZ6EvtsIj$ zHQ~@@2F(49S{~w+ltvqtI7GrZOH^7<1t4onh2R+kTFxW?NuFD)x7aIggE zTq}dL5mY5x&Rg4!SKkF=w#HNfKo{dn5V}l~XT-ap;1{v;SBw9muQa}Nq_!&!W~szj zz(6jFCo^v7CuAnmD&Kjw$BNreRlYiv&e_pmYFH+&N-MolTEi+ZS6lzufMqLdE{fesHaa^&>=1BwA!^-eA)-i^%1y{f_#7G@WHw6W$y5=?TbC zkWquF2vaEu$7m4&ky23si6N~HX&A9l3L+&S5~EQNQ97hXkB~-$k)y`w(PPx}^Zz{0 zbzYt~=eo{y?mGAPb6;)9FI)?DZV3g+m!%*>Ed4&K_KuA?N4ny({T>gVW%4)=j3i%! z?1&IdJ-5zF;l?M9K_4VyOwSD^InU{6`ODX_Bs9vYbtS3n{Wf}qh<--x>di@(Szq$% zGRnkPdU(bqd7xB`1r?(C`xBkKHZneF-)eeX87w|lG~$^hX%UdMH4|6V+YY+gYZSl> zC%wY`wh^$HZI?^m(pj*1^a=3-fE={pQg7=pA%i7pA)z36PN{cBd#wKgaU9Ja4|FvqFdGb7K7=$=+`GKA2_nIG(+o1!y zwho7>%=8s?8Y;``$o)c9)i}uvhvlJ2-HGC@pL=8|xjfQKPdh&$b7F<}BVlZ6mqwR= zt0tbWg-9H4v=%m0eSh8I*_NMDK4_Q0Fv^!FQp@&Viez~9-Q%E%vGt<9pACXXK0DTm z+<)7*F3)GQFPgset%nL53Ylk_!8*|vFXl*N*K z&uCY)nyO*q`8v0!mKM0b;XGQFgQWi2Oz*(~jh?COHaT6Dxfu6OJC{X-2)y+w@*G#y z%a88`kxfR?0K9GNc*3UA->>!aCcBNsn>`GEeQyN(et!S*moHa)q`dh1Qo(l_-qR-2 z<&FE7`5|o4z%rBC9eoAQe2d3oK)H_X?7gQ5sJ%GyE@DkwDg5SGEUrW9i*>dK;($=% zLtrv@lNL`RxE)ykfsDwg^P7Vm2X2ckeEwxQ8I1st`1RJz zQ}OAcV$0sHIWBqsr9$Z)?#52O)vPb>ltDw2!;emP3QvDXS$)gb%a_~}`-iNBDsIA3tzzgEy{ zy#1}jW!`wkudoto)-r>d{IJ5=J22s=zSUYEUwYAAJ}s-PrvF>4g6tDRJld^wly`v$nMk;;=tmF#%K^VCXC$fZq~WTgG3NE2V@CNVt78 z1SP8}HWee_ZLd6@2R1VbR5FSSY7DEN9rnNAhH>xR_Pf}zU9PH!az1}QZ@!)8@#3Gp zRr<%PO-}rIOvg3Gg9rc6jCu@5T(|OE=b|oxQ}7=cykagv3$5SIxA9k)dX=SjOW!C| zU3#zC;k)>UH3seo{uLq=b4ll8=n9)K;7?-f_q6gJ#rYC{(4DlTfon~~bKVejatQmI zzH}{FjYutA+7tN55`0du7}P5qtr4D`O-)vkmlyD$QSy<4?uy!Axh^q5mA3t~hzvM7m!`x6AJ>*fv zg;vdl)&v(w)s#Z85ABHJkl@Rv>3lv9j4fd{Zg(CoJ}?Slqm1+mlRM=E^Bm6_>5Erb zL7hx0y9507cGr_XR`4P#u7CY`;8bb7+K`pfo29`;$J)JnqiVal@x+RVKtq}M^XQLA*->$==BA$Q;D{69;IPm@Azk?N3_UHKaJJj(v2!k!=LM;RCj|M z2oCRiHG>5jvLGt%UwH$ZLSrdTCN@cr?JfN`KTJQ~^S2s35d9d{neX;3XyOcO=Lq6q z8e#CNePWWSDAwGe)^nR!CaO5$p4e)uKM83ZlMyJ3Oaa1vd-osRb*&0Mv^qLe6c~F_ z(bOeOkfinUlK-Cd+&27tl!}cC;GoOs!9+_okFu!KQVhW8apgtOeSvTl1*?us7`2?5 z`%}gw^D2N}k{k^|4S>O$vmY9ZR{BqL6qW35QL=>6>!6X8tlr?saNg?co+4 z)>{A5k%UYbxoRmGf`{zCd~@3{HyS)Fn17S>&XN+Y)&CkyZ4$gyfBhXZfl zlnj};I+#xTiiNMdb1X6;1UPGc?2S)wqE1 z2Q=;1ni}-aR^)Y7L{Ke^{uGPH&thCEx}7320+?K98L*8EpB-6GW*?EPAU|h13Tyt0 zeCRn93*>%-4Up+kV`cvHm-xDp&#OYEWtUC3sBF$r<^G-t1~<~i#h0K>W)FC&CJOw^ zzdagOidWP1tpA=}86fAXJ*gw)lhb7$9xo8zt98cmK!;ue2vJ^ZVURS*&@RIrk7n)H zFj>S@IaGcbBDzkvk<8^Uu?IxYU7fF~u{gi|57vT|&-y*NKQVLaEzg*V2vyjDDDO{) ziOr->g@}iFj+h|%wql(vvU*MFVt;kihe{(-QZjoqX!$k9`b;o54egPv_~gv+uWPDD zQ-<$oAjSE|t+SB!YLg`4fmlNmmD#vW+oG$n<<1+`HKgyGxg%jw(Msrv`Ul}$icX;p zZlWo#T3v351w{gsS_Cu5Ej%X+x<+K5SgX_T2Hp7U71(}=|5~YrItJ|UGzmv(nz_T+ z_{HEl;w!(-@|6>KkKXh#*pGO`KN_8hSB5Q>d^0stF2%t+q(whAN-MWNAg$p=Z*AqJ zyL*2T_KlDm)Y(9P?F6(ndI#`g!|;}{`M}xi@E#=X!(seI-!dPHMVHPLiPxUkJ$)Ox zdp<^X&p9hDml5IF1tTQ-C{+Ers5)KFij$f2yk7Ls2z*{Dd8IZr^k+3yDc-U~6CF7f zp|PA3jdVVoUYK*e!MFB}!{#_Q@qE`E^&5w^+<{Ja+4)t@j(2QxM{3zlK8$va3dG)n z(Rfc+)SJymBgKO&ub|w#{1L7WL`AXNHQ-_I%KsI1#<-j;QaU2QTFuIUhy>ArK8+_P zf1i;Ccs$+y!b{WO8f*X@Gd+FK<+q{TaiofZevJk4B3%pM4GQj*CKHJP2jojyrblV= zc1WsOG?RiRd`T&TEHNPwX*L>Q-*ZlAgM1+Y%Kng^G>6%DEU=TuXuGo|cI}{Hh+KmSS%DRkP%ptLksNJgJ&NXidl!@gyQ0#CZi3&MTjjqPDdN zaq+2*&BfG8YO_({vvvh)4Xp*L7F_`0-Ii>ist0jIWa3mOw26Iln*C-auPG}WO`zu# zsALY8{21)-HySh9IfXZ_e-&z^;+Qa=6JePfnVvJ{A1qHx>|)TNFsU$TYOteqiEI+P z?CxR4u7i1YvOLyr!>3p;HFF`WnSGmv!pAX0mk#08K55kl%y$aF0Rqhg;;Ab~~ z=DMq{-2^>A>OC~v0Ihank(OPL9~(8&hyNJ(;Ip|!V%PPeaAZ#M5x+;PIfXvUOm;^2 zqvL>GVLe}w!>VKP-aB3@lg9}^RGtN58aLGCkl6Ff5t2ht)r1D)F}#Ymz7aA>^83WG zYt;EYK$U|BBe*u(yq3lM&7_UbbXWJ_clB3t!yk)Q>+nXAhWJJFlz-8W)gGOBk>u%I zdewzo^|*+$(Q)cvT2_C{<^W)P%()^Q06v(8hSL@pIaO}0tcq`aZ+PN=0Mgif$qpv9 z!Xi3Z42s5IG6L1?7Mr(z7;fG)X}kw?(&b&tp>=)R>NO>2{7mLusly@V`jTTM1ZJ@r zCq~!j-)V`m!$|>%<{kd1Es_azpjnRB^qgFNa^!L-vP(c>H+w?m@~kMP7Vko%M1f=n zKdXk!FTevpq^JVWwuICM3~^`35YpU8!{1m>Y+e+$Ftdy>d_Wd8wk@yGuCrUa5HCGF z_$>;dB@F885nA9`45Q9(7KKzlx}z%cV_j`2Nx{lJ_!5 zB8%#jwa=u5=$;bejHwlm8ZpZ?PgG%yEH&j{Y=hVjSyn7x0_%shaQ$E0l zA2V+@=ry|1YTJ*`)PJJ=XJ&D>p zcKFuiP?2VneG5gt<-0X+St|H*qv&ZhzD!alk?8TBM>=9VE94_aj>U7j_4&#p-%~3+%*OnatPTD|VdFjz4T0b5NV*-jQ@(dC%I;hZ($^_VTn- zc<8ojIqSYub4T7tf@*o?$rsL51nouu@mszc3fZZ=oBjJg?_^e*iOZSU=^#`df(=W* z`Sv*EDPG7#4fLtxs^L~_$+oMMqm^#$Opa(9gvVL5b1U;39Kei>fa$mC9xrp5lk=V4=1cV#ng;QK~twK@2{Avh4^&axh4uikX1R{29(_PJnmD9^@e7?qaSy^pg)z;kiwddw;K3(MyanO$C?Y&>PRls zmk*9#{ucWbX-0rOCzjsWXj<>;>iYPzrZfANN>=aj)eRAVHy$*gg{PF!t6fsDBi$Sp z#NOld*PSlv-kSY@QCa2An8F@eq4!2gOUGZ_TtA+R^mD3^8Yt!o(%@Z?{!BDq?m%ju z{3GdxI;zC;pHA!!=kIcFfAKjPP*&NQXlp$-*j}uTD?R-dqZqb6$Z7Ub)VwV0j+CqN z`<^A6OjqTx4#PpyyVXz53IZ&pi~dz-+DF#W&1Ru}*=G*;9UyRuNcUh*%>q2cRxT)koT3SBD!T=e z*bVOYUHam-V4+wIHt(xFub$6qKygc!{7~cm-F{E<+j>>I3eYi<06T)8G2D)+HV(0N z3lKHDX63%ncy_YM>3e0)C3^z`Pg(oGcLnUh`8cx|;?v{ij@C{W{y3X-=W zgiBB~grzE95^bNb);f1(;tn?9o2QG+ovSYEV>j!D!il*q8kdy@!hN>mmw7ml?3^LI z0VN1miryRRQs;n=?=V>nz2UUHVrM3!|I)U4)=#?^j}OCF2!MJ!$U3Ezo?QsWX&(%k zGHuiL_pF=K6O81`X#UDul~eQ? zexqLb>D0mGZFmuJwbj+HxbY*0xq{THIcAPkwOjH&UoM#DdbW^Y1jIU*vj14)-W)GQ zmnp=+>4Ev}LMl%nZx-YIZ`lE~q6BewVPau{jBo?O&1Tz-$&I2nFE3PoA zJaE0OM&y>Z%Zb6P3lS3DIeK-@TuqbFh=HBij)0Mdf6k&KVL(Z0vO?6V;k2M;a|M6i z^gj^?`Bp|eWRlD@s`9iY>1}?|)&tx|WrSk@ky{n7ir7ejj`<6QEZ5vw`j6rt zEY7~jpVj5X*nAe^udW>AC3>2&O6-lvX7uuc#w47AMY3m9pI3-@AUFo%6q{9+?{EqJ zsowPZ;Nw|4W3)9`>1U_!O|>bW?l1Y15JGr*_`}FcWpfmO5UdqBaQ$X9J*eqenfb!0 zLwKb3bgcFikNGs5E{kdbg!#t5(VJZ%JZ(@EXVm#S(-fLtwQFF7iZG-)@{NBgGx2Kj zWM(hME1Kvko-Wh?&q8~)SqnKq(_ID|4Qj<|M-Gu|mXHH0{@Ohs*MW-&r(|Es`Hi$W z6EkPaCHKv_MDW1XVo#rICDN-C!QIcuA$|7V20C+Jj$fVL3LK=5z&m&_1rClm^zCR} z#@Z*)MEQz{!!l#h=(UzQWOw<&(e7yeo(x)x#Yhz`AW6EtQ!I{nCg|B3kwJ?(3(afr zvC*zC@v(<#q&k-AIum~No7Y)lZpQlN=g&XZz!uR~(tkF5V`(LfP)~=`ve^!!3bz7?vi1EUN_K4HLgX!Y+YO-D z_ri=Cj#$^Y6mM4n=xjx{lm3+@WKjKxIsq0%&V|R~XfJi#IQc*PMso3sv@aLI2o*_( zypHDe_E90c7~2G-u-GcE73u|frdlKV?@%f|- zN~{ZPZB=dS@AC4pq_6nAN8pm~Ek^#zH=V z@a-$abAUN*kLTn=!(#E@*!t1HPBHT<)w{oAn)bcE;a+hB2{#>_aqa2ViipdCbP~d& zk*;Hs5Kx1iMi^$jX~d5c>*xWTEd|s<)di!7so`5uEa-@$L#)m7`L|z$5bT>U&91i? zWn_=`aD%3^({5u;SNcpx_{}i7;;pwT*PKmi$Nq=dHE+!H+YmqkMLjtT!yh0=5L~~& z|H>#?dJLQ~>96NmpwZl>#I}ngG4owr35f;$v!9yCmVF}0)=wK@&2yLB|G>S9(o8GN z{ycV&W3WKM^NXxqSKuC?P9m?x^(9v&-IY&#C|8eKMWUgv+a<;L=G~0kp2MTj#?#xN zSVe3%lGp_Gs8YC#gc%udOQ2awfFPKvkH7RW}ivr z>dyUW4YmqSv+g*u1K_;V{F~oNGc&8f8sJ-Y(*MOcoG`(`H-z?nIUVshn8S>%ud^vnAcf4CF>B`- zBJPf_KBa@yjzQ~Bb+npqVrH4l)j>%13Z}gfTX=kCA(6cwX3SPX+?TSn8!mF%JgXxs zV@NCk!Aj{_W*0sHBLN|B^wXemlV>ZwnY}rdswYfV?k&0Q`{%_bA3N8^v|s~=zILqi zZ%QAj9+pJz{{Fi%v+#9h;i{t*z8b(`u8jM_KFnm&x}+@S4Yz?0C-ha{Eeo|d>^EIs*iWxCO0^` z(~W}Z@-R{AkzpjoodmU4DrNV>ph~xs;(3&MS%Ia=>`_U?toHs}&B4;U<&w~eXtd{O zfsZYlmBdEy9(6u8wif@m0*YM#sgw5&**GURM&fE~6*h*%Pgb~{bc%gjTBHQFuFYPQ zsCld1Kdnn4atm>(ZGG0y^xvqFDGJ>z!FN?P>AoRJm?=j-j&5Gi>01VX;dZvQH6PS^ z9PKns0y4~&LO=TNLXRE*x`?3z^SsjK`XeK1DZMK5JN2~jZ~2z{vub8>h;`zpy&Lu( za~^L^YP>zil5-`d%S8H*>{k-0m+=?jBOV;#rFc0kJtxM5V0yLGXx_m#L{>?gw+?&A zkZBN7d{!I6|K`?YE|l;VkiOE&tZs^wp0F_n#YIa-kCGj$y9ee}Q30}LadY{>0@zJ(;fQu6!9_Ek~lS;6ET#C$}`;@!-y! zZ15xB$y96+6V2=ratq1{(n@>w z>w6PEGiaZ-G+H`NY=enW%|qv;rixm0CcdLDMXPL0geA6NiE*Ls^X#VN3nt0j?RqGM zP3k+)dXxd_yJIHWyHiscVArb)3rCXUuBZ>q$|8(DaQ7_Z1!Q5L7aFq+>M!F;6!DxF zjN#|s`JB?3B_9YVcSfrPEWZQMO6cv*)4Gf>vXrcuUk+(r@S*w{7NOcd?KA$H4GOMD zhwl3ha$Uvz@1uXHmhe^#=8LN)@T%-yn}%|4mMLt-lqLurj6ktj{@d0|(>uju7yP2? z?7j$}wwcVq+?kU$=`#Imjr&sp!r-faW1EJwd$c&Iqd4yXcG%i;GkJxX*>+VkC!-|K zcV~z*+NfUWshj{k>hQe>~%J|)VTJBV-d6tq|)s-=#Q3k-NCBCMl(gCl~X}!s(=0i zu41S?vFQxS28nCy{e3HF%z2J{=v~}VRcA>poc_OTfq%=F>sFFm0NfZ1-ihn+-H8cp zE!{TUj$stNJV#~?LQc9}uT&&}O5i$R*m@Js7ar@NY&41(<|xH|4n?A?1z>GSHSbIW z<;LBu2XlacxSd}U3}US7v5+qtDDCmi;%68`B`+ett&?3vHX%puxahPC-tpA6 z=_N1SHw1Rj^(X=8PpZ$?lr`*M@y$oa6diQi>nd*>Xg9ZoD=mbXsYp)(mIj9#IGSohA zF+Uac9-U*;pr1)kMZ-5ah6!n119L#ZoqNrk6xh3Tb~2$e)5Vu2La z2FsPsZZ{}SG>Jve@(7KGG@njpW@G?Q1~n_7%Kg@dw2`ptAb#-TTGS7UB)F%husF3wo?~wy5pto1Ov@i4>uP{aRASNoY zTo*Q4?!B>>Bsh`7W4=9|lkj>sJzQG2sp-D}QDq!o?9;8%%B$`a|0nu%da&5v^9jh1 zofi~?G%b}wSAN;J(sefpMq{9sspS+ek2-0aZ_@bn5ueH%eC1@P)_-f7?}W&VbYWoG z|LvmspAAHV)}v)mDs{fKQ==bq3|T}nEY3ED~SB^4tk(JqxT&|FTi zEmJD`;(CJ?h`-iW86dm&6KY=Bmo&*^s zmb)e@JbSl#O(o;A8?D_qTUj?>UKMaw+guOb*A4aH{KmdBsWoA|5z3g-<<8DcptjjY>8if8vN%e5;oykX|^Gp0~B zWxdQ9Mi>EK#X4G|(e@Igjxxa-R3F$8yij*Dyz!Jy7%LCdqwtT4+8>e$Z_vmmzYd|} zg2v8pzFcR+ZOH}&_25CV=+e)PZ+*@f_FtX& zDC2PSj=n^h^1)c5jDtHfD3yW<}gHEVOvM|kv z$u*R+%Unmf;j}aWMWC1L*)IY~1emej+D?CN)mBfKXQNx^CGE2T+<{ z$T^8}uL>hW8a~PeU{XIJ$r_nl#j7GKapv-5z{!7SPFz)hu@K}10@fp>WSu5=JbtWA zo@IO*qbD>ONO5I!Y5LAjqDoIKOj!enVOh-~GT?;ITVRpO5eHrC|1F}SsP)ENy&Dn^ z|H>fW2LTK#dnEpdVqAyHz@Mf)i}?amoB`?Sr9GuMUrL4A_sg0(ovV`FY@HWcif(Ox zRGv+jbmxOyR}IO6?GGCn-mo#)TDlOVr15<+L>R32hT@7(Y`((I0xxwXBsaef=1l`X z-7b;oG+SC=^JaQ;ssVp~JShma&SkVfymc#5%(S~{eaTbt-SUE9vXRPGThL5|+;LJT zefkRPD|i>t_sx|{lZv8%fcV0Ljl|z%HUNCOdDgOaORb?GSH-NLLow;5hZZM1&6z&7 z>4R|RHNtB_%OX+g^Y;())J)@j#=SCswsD)sK(E#e5XnZsy_Emfmf4(Z%YJ1Q}64oR_4AE3qg=O=*EW}93 zCJj7bA!{}&xqcU!Dg3k^#<}679osgYfZpAcdcGy1he|!nj-2+|!#y##r~H{lZ<4Vm zrR)cTOe_Lp`M8Jf@rp-Js$%R93y=&AgGV`!C2DfsIW*#RMO`e@OH-n-j!7pD6 zzXkjRy?LG$+;7CCH*MB5Cudr}r60EIr0>8@FIK_0kDy-14Ij^%Fb%ZFRHz#DGj`HEj70C`AIL6u zbX+?z8{)JpaI0TEe)GgMaN{*A-|OW+s9{vxhNWQaP3-#Rx(y}vyD2?+W4T*@;jh-@ zVpEKqQz?jco#-wSKc*bs_sbi(@XaMd{25)xEFTJd=Kw!Iz`qJ$ zvv<+@FbBAK%v%v!7;~t7%uvNqZ1;L{^lzi|z%)1>LrFaDI)Apza319>zxOWdu+9Hh zuio5C?5nweoveK#rF6N7L&N>?8{9Q+kNmNx`m-h1lV_paWx0Cc`BYKTO<1sw8oTNT zqfSe(h671$OQ|LzpCA#KjbgHrj}MK$Ay^9|_e9{RCJbH))c1g4{VFj*t8U z+-T0^U+#Y;p0|j9i))xSWqSZWV15c?NPozamBlX^B8|;t5I%N)!~6~-eCd45Pa@FL za5LNANACXHQC&GNiq%!kKKV)NTt!Z%1Y3MzBvi%qNrQ{tYa<%2sm2J8pvg)+~Mmvn^6Mij=Fo?r6}?GoNic=Xr^?xa46*CpXX+hiGEprd$*cE! z4zrru@>+lTRG4HS9(!h{y%qY5I#`!E#ylZ!0w)x43Z)XopeN^~s`@`it%Kf0DfT*o zxUTKOvz=eG1=F8&+(4Cm+#-K)F@eWlQK5+l3G7G))ppc*$2-ao8rY^}R@$6_qUxcs zb_@K#-MaiyMPFXSlaR@QuTp&7l!As@5=vdyUb7FM(JnrplQ0@Uzw^z9gmKqN%H$$J zsHxNaclo&`|Mu=&O{KW3?HCu5toF587S`n(-`!-lwlfeTuwin!!3Pa`K%|D^^DU_V z(g&!l=={G1DO;H?xcy}L!kOcVOQWhiRjtt+)k$6I(b7|3D;VVO~dc;lUmfz$O47G8@T zZF0|}mhg+OM~k94&h73*R4T-N;HQXMU9t5jl9|+wk#6XzCZ=a(#3+#-RZy1O&pXLE zX19|fY9a!tV$_lwDQR1<-;NvARGs`l7dkQ3Cd)kaTfP3k!<5ioJ!*%X>PwN&PMQ}i zJ##W@0BGWX*49&{YyLEiLfY{K3{PWN;Co)$1;N>f%`AO?%r_e9R&}I*+m&|D6-wi$ z$ln9OR&JO425w{=>?|C!$|D;5yxM-wNw_DRd{IEeO#XQxg5;KPAN*`ICzJ-vBA~cJ z1T|?!CRj6{^ z*Yg*pu%D>=;upCrceW~)0DgJwekOe;q{1wA75NA(Atls?T%l_Ku_H&`ai2BH)2W($ zh_nszvpzYRIm1=vhbtSg8(Mx#5!!6R>H<^^u9xsAr4X^4XXf96q zQf7@a7{|jdhbQeUrW%K{1YZhIDDTkld~hBKl!1ev^?juaE@x58|I94~k7IrWHrzz^ zWDSnIG<0^ZYWfKjo{I%Dv8A;otnZ49!~eJzbj1C^8{HbeUwXZxW)bzEE12=-JTnLR ziu$IdeLXSh$KYMH?!;L=cZ4wW1cdUX+p;pDi7r+|FOS0N)Fh(NFUtq{H27oP*U)tv zR52%w%Vp*m#LyyU!4pJih`Tcd zdcnbj;ZeUctTqHbP;t(5|A-k)_JP?jH0-ONtbts(!SD^56mD>p zzj(Xh0r}&KIkh+n_w~QGn${c3HUaUf^!l0Zvm6p#=Lk6}pb?zc9aLyQfv)&~G4bSh zf9&uY7G1`DX8g{o`P}KZ`V+<2{plKq1U$t10-`+XyHO^WK-1oum0(2+Izw%+2x9tv zkW(TEoZ56J;ZOroPFFqwx*4^)7NcrL4$Yj(Wwr$EM~PF)RK2H_cN0=DWH!w4s+{$6 zi7SG(57A&w&j75jPp-OSS?4=>X;#$61y4Bg55AF$+rTl^&(i$2*awyhpmuPN;v&*K8{#s zx%nGHN;L)`ys*1YuDgeeEJH&5?}dALR5{!%r+W9;$6opV4Rmz0m8^W8>i(E5W0j23 zp&6JZv~Me@!mQgyd!#$ zB>Z!sYzQh%e{yOi=IxDyOt!Qw^3^r7E!?-a z$KO1anrUv}pqsCg_N3--^h#Ts5)~Zy84qVjT=P|Kx1OIP96vQA?@7YT5MC9V=^QL> zST{t?AtR2T^ZCFSKFn=|6j|e3eMmoR#T$AFPnAzgs?ymY=f zH+uX%o!XQaPHlp_%rgSld$?Cp0=Gngi@6CrrgH1!-*PdsB8;p*m-`$f(D3^?MZfNw z^Q(3fqqVUit}XV{qH;TZ(d*(xb4xZ4P=vR>eUYTJ zy5rK>mxc%a8GZ%|rxKoqUqTN?c*`;*6%j7h1D79=^iTOAVVvQAbs%WZfd`0%SSwlg zr=u)ZjR8T@GQ3L1#}3}-6a?NNWv_^wi46HM&srXa@I;Jc`l_Q{dU%x76|&#-mRU1yQkLk z(e;D6d7iPHO4;mH!H#9aK$Wcz7+tY(sLw3{_9$I0)$zAp-xBMo3CNJ1t@}yVdITZ! zXL&#ChuZ(AR~(~RPX0eD`G9rdkJ}qNAPr}n8}n(W?Zyu^z;A!&Paer|_cX~ELC2)| zd||*(pKe97G)16T_=kias^KvSix~-y9$Ly?miFu~(z-xK&2x0G9qFUFJL#A1`9W_i z?Rwm4e#;37$qr%l+8Smk_8ZUE)Xryy#asn|`d#M$Xa>>doXdYh(lSNp8Dm#mYZWuW zroj=P>W8)Li8-8&8_m~8@E*0~DGN}K-Q~;CCQ5f>gJP!HKaaGJrV;fz>{6$1VrS}f z^=ape?;16s1oMv%p;atYCoL}&_thkH??G;F!Vn z*8&w;k1q6BAqPs$7yfQIt^12vUM?|69d@--k0?>f zx!Y~`!J*7yVkMg+`rzhQY0p`1)vp`=9hkus)xvy(?PZ+f=15ZWz(y+ydxSX2!F?$X zvbCVow3Pj%!v2GqV0fqe5OK0%QjO>Ea!3uSL2U3xuZ90C4M#rpU!Qwq$*RgFFgL79 z=yexN%u}~EJ6GbHD{X#3Nq!d62MSMnI|`M2m@-OKAeGAnWJOG-JyFnE)P%mqGr?`QVGc`(XRM37XZCR#?jRCV!0Zi&31GR7HHL_v~wT8PD!LO?3O+ ztb+w0ujycx`Y^Uvd_KZ=6faD8)vtZJ6&rYdn&KG8!s$4{kbHd5c!rMR0qsgi1P8Nt zma4N?0_CX)6kc)>6bAn8+7z}n7y`_n*>bV?!fC5}9DHN;XB+909m|iYy0}iaWXr!1xlQZ87mI@8 zJ9>m}O_~_vrRR~X7OWPiC5sv#lFqH@Ond`*r_bTT7T2585`&8v3|5>ZyHis7xl!b= z=S#yk;pT(#^FqQus@sM+XYhd6Jhj_5K>;dKRxH?I$(&|#rHn~NNr(% zb>fTsFqliR$mkg23xDc@Q9_b|E+BS{_2`O1#ZG$LGZ^18olC~D#rc)$3gh}*oz zUG*bGAnsgtf!T#}r&H~Wj-SB~<&mr$&gc9bq@A8h9YSKqSBwHYRK~9{Wkzd-X1-j6 zEbFbzGaCii?dLT3a0zeiw@YKTJ~w_zGZ3_!js&D z@^cw8In@JX=#5T>4cHZCb`U!=HyB+%c4_Xd^3-PJ1w+Nu+4)~fQEX}zJd~{9`gl?B z#$dcuC)?Qy{vaUQNadw;aCW_4H+(6CBbY6<$$9v(Yk}1B2P;90V);o(6}{=FjY^e_ zxjqYc(129%qsJNP-_(`=dia)*yq!|lFJtvFo*cprEt2eynIsL@*+9(KR(>rj3i5(>G*Y!ZZu3?Kx9Or-Gai zuig!due6J)%(*D8wUzmW58-QeRXedQ9@W$3%n_V=Yz;An(7R^4YD=a zKeeRXTXLoGd#p8YW4WHsUKJ#S#Cy}&a9?&G0P&aWvl)A4PBwg6F+FDE%Lju+MmYwm z_GM~DMboo7CtPZTX#&h@juUA6(Qo|YJH`IHk%PoLMcD72FraelRu}7QG!rlI;AaNtpb6Hz|csA_{oNeu^iY1Te1Ubr> zRoU<2EZXFzx4svhjlOy@U=)^Teq3-63OTmk1q7sM3XKf}jws7-#gVsD95ZTev%otw zlJLy%u3-i_{->EP!Jyh72tLQk?7d9KEtF%NDH7@HcJlDK$w{3^Fl~1#cz^I= zOnXPsgJ|fH$)VHt8=$FY^1Fb6A|S?DjW5Bn>!ET$vSyQUEVzx&=79{`*@q}f zn>vd*TJX7DyGsY}w2=HYD8QB~4AW~?b?M6%-&PpCEgffE@+tT5%(-#;h46va)kZ`Q zgOq@%>@D?)wF=#rMKaK>)n=X7x<8_Dr(P^zlJU;ddCr-4$44ftb~bOygCArBaFl*y zgxOpMc=VV9LG#8(2?+_ayknDh<&>_$DbOSJ;Dt>(F=g)JFD8iZ1)(lplos_bjE%c@ z4edgK+eG${a)Be|XZ}g+N6Wi^;p7D89qiab{X` zwVO_(bx!0)7X#Qm4^WZI7y!GQz;N#hrK@&?Zdr%(zPxP-^9nVT-V-~Y!?0;)<&B{BV*7b%taU&8JFSJ4`5AHjX? zpeN;7W@x^Z#`~(38uyGpNulj8w|}TP8d}EW6grv-*{4YX(;N+z02XxZkHRxO*1Bt$ z5Tvv|Rw?-qY%YF_^lSY5@8Crz_cX;|W^ct+u=%m;6J}5FddN-SHOHh@Ru;#M`5Sa4 zB~F?H4s7{Y50_ak7o5YxOp)(EE22)I>^-@T#QA6)xZ?6c>STg+#v`4G zFV8IW+5_>&+%F1A-xSN)F{>=#T7^8C}fjFgqk3UZV#jKZ4e|9H;Ib zxy>Y`zm0K+=^5^7=-BZ|uZYhWP3sXo?UTCgUAq(nS%w>Y#);|zAzyn`;R{T79p`g7 zlF2Jc#KyX3a}u#o9r-uG01wvk)y`CsA2y|6Tcab>S*?s?7AN z5Au{xkR6t4FKF$g5VHm|*RE^-fC@K;`4ntk4%!+y!wtqRaQW$tx3*&l5zM#BO8G$? z%Mn(4FYB8*^O*0UdL2PgJ%W{*g+88N_5W^?|Bl&Zf60{J15V)u=XLjOjX5DscXCf% zddf9k{TRr1jw%&mHv?%2sg%li*_$8B7Nsgn{^3`M@`4qP)xEoH`=PxMIhK5>R;$|-9%4}3&?cl<}h_|{me+vM2n zYt<4v*+2Ws`I*A==PjTUcR%?)QNG1#fOkVVpA+!kYBRcQ%pO7F-vJ!@Y6JdzaJYJ? zLj8z~^OrJ4Y~^mD6dD??`O;!*d|C zun%~_j=%4xu)+WOkDRb?fB2RmODoC_GmG7BIJC!?kY3J<&ctQD{t@3RckJq6Pz9)( zOmjg>@edFD`L&ep?Q}EyqTxJ_M(Ue6J3R=6VHxV2g{!;fA7+8AROXtgD|+HPz(yW4 z-f5TEY*CPL$UfujfO5gPYN;#rXt=(w9ticuJ|x`^Uqrfg9=o}77v6VT4uVXsU`>XZ zsib7#%@SivQ_Tv`(8%YG#Go4>?HeMB(m@O zkU^F)NoEvgj2IF#42>~(?)&q3p5yz&_qhIp>$r~ddY#wxe!tJcxY;TNu@-wW*4a3s zn5KU^SE1t?oGdSl`~_gdG3=?QXTokGu$YUox;Y zeohlXa1`@I2FzJ$cD*x|5$J6Z(0c%I@jzF}UFDL330@m^zkXSZqxQLA)5cFB@VIT+ z&c&+1mwz~Kvneoxl`eW{#9otV8k6(k&sg}tW^rHU{_ZTOelrKl>V!2@`U=I?xm-Kh z$clXS8tcZAU(TWRhhpO*DRH}+4UkL?$ewu0r8av(^CER{mJQ`OlBaj01gc20{~E}Hh|xwqKFrBtv0$rKr^#> zgH)~w$U{j&G8AepLOZg_Hl@hQ&;X;HY-{+qZi2<8lNf|(JCI4rnZ^w=UY7OqhBg}f z`okxj6-4v!3=pxNIcMd5&%Nz0hNFIcwJbXPZKc0N*C7XkeB&X9g8L1C#1DM|J1OdA zf1Gk*n_gu@%j5QBi_W>{{w1#If^XT zq;J^^bW%v6pELbyF6wqR_-euYak`GXcpk|=O~s&M5$<|3V@kw*(f;DY=_LfuMhuH^ zu^I;#djAtjm6Z@HPbmG_`f|R7O`Z4>EKQzr8nfTN|MDhAGhpmoL%(9v*d230p5q%a zfkCK326XcyaPT(>j=ScB)hgJCk#AD9cCV$_Mqjb)w&shZ-w4+N2w-Y>`$YV406@X| zdU}Ol-zi)wgKv+bI}mXnUD?ufL+%Kj)A0NFdhD9CJ7T=kZA-CT7*H6!;glaZx0dsK zA*r(Uf!EL7M;#PnZDK2>_Hb6%Q|s-BP-^+qOHyX)Gd$qOy){L6YwX{66y({0grZeJ z?i3Ma3An-tbni(q*P04$J8c@>(X$LrsSQt^V@Qtox`_*LkPpY1O?9+So$Ib_E0*FX z=GPXmBz|In#qJ`FCLIM4-po3c<6xBQs~4B0)_qM>Qt3qHYOF;ok-^Mwb|5EH5xk1H z-g1GxY@T}*Ns-rghdc$i%=<+Sht~rx*olS9dGd+9zQd76BN}F8hk6jXh8ed2;Uh*Q zq2O1GbCt^dmrj!%^V4w_5M7y-8-@M&!B5)Jx0%zUV}b29T4&SXuXa{u_`7DZ^>qL&)^t>OjOA(MI|F?BEoLHvPO`dF^@B#mo3sgEKpAzVlD_WkWEZoF7S;1dZ^C zVWni;5i>nXYP376)#;Z^@_%Duqp4qWqCSWTBd+g`RE95LB0_52D;2_qhb`7CV_tF` z`0b?}Pi(y^R%2yv=*E+BsYr#N_GlO95caCn(AE|<+E|Xor%&)VQTrlv%o=R(0`j}i z{tvX@z?oq~{_LT2op9Z3Sxt<0>lI~4QfRY`$$^}K%qWLA+=Xgem`+b%Ww-wX#7r<0 z`VD^VuMe60hpF&OVD|rusp+h0Dn*bqbEX#&=5Ic4T)&>H_F4V*!k0M7q&;w$v)FVo zB<|f8^{}Fc=3r5m4m{~1Fmc)<_pG?wipIFtaGL&?nm2^Wz*ViFkJ`a2$bSsGl%9Dq zr4Kf})P1IwCHgSgUo=W3@MMNtwc2s!15cue*%H@fJd(@1ShA*1>D2eE=(LRO&>w1L z%-Vlh>mdFAlnzZ{an&9sIxZc$vN23Tk#XVygT(v6y0b4K(o{Hy%+mOK@J&`3D&F_f zu0cB9@(jN8LvV)rbU@5hO(@uiWzdrxpf z&Z*aOAjld%*YHULzoO|_X5z?*k>>Ikv(ZMzmKotcDddqe9(Q6t262i)%C&9j%tnfL ztG336LV_;%a0bC%JsuU8KLZEndFXFBWH9p{a@nyyVKx%s2~nJ_=8mGzPLDf=@GWa4ay45lbbdXU$79+n>?AXL(8)pH>2PkT+X)LzWmKvpH zyg>2|@}yz-nl z#o~TG>f2q_l2(v&vJv%M{mogSL)6vXU7ccSZ+9shj|Be^Tb~L|t!piVQkl>r)@6rJ zC2R|;F?`1p%o0oA=KNE4=%PBjW~8e^r}|@}v7lzv1C62=f^>=J8bR>HD8GLoOTHGR zYvC%kKL3ZNyA!?I2QJm$aNOTiCYrtAo}<{An=*jf$=s=tHBp@g%6GMCi+#C%<$2-XgaXQAS&{Udp0~Yw+)v>t zeRC6rA3njSMSuB*5~WMymuq-5-r(-sbozpTDW>G`cdjW@hExA2zNgGJ1bqgnE#&i! z8Fz*KV0dY9EHk)rOXJemTZf6Kl(@%}s_DYj14oNHX%UK)$4^s$)IoPABi_FFA4jVf zj_Wrg4AgVtsUdc1NsJV zLD>Q@?o~tc%(W!d$tg%J$?zu&Smdyci8kkHxpJ{79&U$_Yu~?ANry5BeEu6tS`9j- zdKD(N-0nnEO-h=EPfBR#pi=e9oRB{(&VR^j1HTSEJ$xlO+2!DfFyZjuDj!g*>8t(u!QS&L)5sqHO+btfy zFN6A;;aD-}#6Q24`x#RD{UTVu>1GgZW`skEMQwapyD#&l{ye?Z=JjbjIDCFow8-6i zb;5gX-gZ`kuSvaNbt`UZP%fM45N-Vo$07dz*}{^0q+|bECelp2rm+URd`tl!-*u(v zPcW_I1;W*yfB87^U}fx{`n~PGzms9{7AsqC`&w|{*3Kn9j2X;ib6$HFNK$L{d6>KB z&4hkrwVdN#%j?ZIEqXUVZAZPa=z)37T^;js9u^f&UbU13gJd%vb7-|1#|4De@Jog1 z7Pi}s;!!jL!So!20CIQM4{$3k6FBpqt{wyw8)xe+@`0XSp-EY$)kT+>y=^XwCnhbE zMB)}tRvWW8I12aBr$D)?m znD|(EU8Gi596_uAQ=MWe^>=kUatFU(qJYT`-j0M2E^GaXzZ<0lC-F;!RatPI2{CVw zmzWXLu91yTmhr>KOWM$M3te(k&%tnPV)on^&=|jXs7v z{vH+-V|@|#IAjH&v)_Dy-Z#$~MSuPDFEv&s`JzuKKcp~3-ji?P+w>!KTvp9XcK6k% zze`clOpxP7lSWdUS!cp<K3;AgawJh}z%(he5GvC^QP30al8TxYh^{_+^Ka}?=*SENHJzv43 z%%ZvM{YO+<7WY#Q&8(Wg_=-!5U*>X%!^{bRYz)*3p9bDVpWirPK9K!jwI!AQ{yDuU z;<$j@bFZrjWBnM#e#}>oa%|0ml77aSCOZgP%;=wdDL&irfu#X_b|yvD50}KIOzdjV zote;)m%A)ERp_x6z^j4rk#p(rQAs!o;xnVOA6;qr&z98duE|SPGj+kFh-gGKhN8r{ZYAG1R znhpcyrkGkgj?}v1bZdu*b0`z2WEl4o?RqO~BZk67<8g+FauuMadIT=A`=!aJWjCd`YE~$Lyn|Wy+ z+(OfVzWzhvBWK+bCw)HsLlkO)9uNaW>{j{nkr<&?+KSM1ZyVy(@o@hhbkIwgDlI5sal#K|yJK%}qP_Lv-Qnbcb)2fp^IejCqz5@zf&nBS6jEvB(%h>#`E ztMO^=3y#RORWHR|Wrzm&Nz>w=9n<$IY`kJ;t&UAO z86o;rA-bU+kIru`Rmf_qzvpa_a3p^&&&8y2Xa&=YD~O8r4y?H)mg>F>eTQA&yjaC6PV-kJHS%oaif70KYi+oe&lP!Z^+%{K{P zdmIGWQEK)eH}c%Y;*agN?@y2Ha?M%2AV9XdV7KQk3o%z@xi;b+>orf^0I=7?Udx}; z?PHlpK$c&R8jJTk8=A@&0>*kS@81mh&>$}6ZJD^zn@z_Q{^@GIr4jrv&y>^}b(8p! zhrT#lQ`^2Z+4)PVw<$-PE6Y!=5!yXxNU4i_ysi8NzwhL%YmHvY#Ni-m!p>NQe6DwM zvwXetxhABK(81p&Q{_{g^$GxG$mAZ6&1SsySA&OdMSqU+Vf*&_SDm0Y>v!7k8WHEq zYbaMPkB<)J&Q8F-bb4dh!1&?ozUxnB|5-j5FqygbH%%KKt3bIHVP$Tx8P37sXaq0* zRb(dOPxU?K6+>wfW4pKFN+xUZ<0y7q&2z|#tdf#a8XHh6^w+nJ?vN*lGr3y$)@zfY zghoZf(&nyNxTjp}2lte(s&oENsn23>&Y-Q(p@}+vq-kpqO%>9lu42P|RhG+q4Gj9S z!E6?gk@QBvE+9lh)Deb_6X>UyObgDE&T9*HPVUe!&+!;`r&HfAt`#1v9Ef`S%<@g+ zQg>On;>o7(YJ~gImov*WkWjO<2^G5k&2l(5rsGBEzSz&3rxg$gQZ$?o$>Bbz*Dhq=V~Z7Pg)8r8BZwetM7KCn3@@ zX@azxcQ{*hSkpBw{N(42u^&{C9zpP9N$iEranGqZ#fbId=j@RS z$w#zHB;~(_g{@aAeVF(LJU;M>Ty(BU=CJL8Y!WiX=W^Su4N99ENHh#@e1*O}$CW8# zUR}b3=mwXb1_b(+>^4mqh#XIdyc6G5v_v)ryK(p(-taD5Y`mWM;+tW0szIMUQl7>w(9AkuRsHd5W}@QHiItUe5Kw)d6mAJF8B*#CI_iBZX^J+ zh??t8T@ehQbVJ)SqnLIk+w)&F^_U_D!Gn!VlJMpv;{+f4SoT0S)vUgu?_eRLmgkDn zg);mjMUZgTOq@)9JRitVY${H-ow+*Ztz33bJ9#C(p5QJd=zlC)>T;iK>_o~!{==u> zM|q>5J%NjV>2ombgyd)73)7?;@t)TBKR-C_`mE7a`Lg}^Y}SSlZcqtp9vfSnaC4)r zp_ickV55!fwb%*t%EcFr+Kf5xL6G5rfewFuqa@dbh1iKI@@Fp$=tBzoA51|M?wp3_ zL97n6r^2FH4A2WvU=4)n>Ppj^eF!Fou|l)}uEC9j;}{bwpKRPY?7uZsAEYUmp}t>N z6v+w8Il7u@jvXuKjP~FZk19R`0=p+gUC;ftmxg$^x^NWnEY*wpg<8+o_jf6x6z4~* znPP_|nIC|f_-fi0dsg%#+AH{a=v}(*GLCjJmDXq%uXPu;L^F7Oy6u@ji;g+(b#vCE zu9=0y>sXXTWA>-NsSjZK_8}EK3Xx3dAx@|9&Q=_v0VbnpZ z59b9l<~#J|P=4eYc8+kWNzhO-kD-#1SEHr7@^N-;1CGosUg)g}A<13dm=)xmQm1Ub zgD%Le*fN^bjGyv{&#`dm^xER30DJjC2ZjcyDsyU~Dn|b)ub<5%W5Xg^V>9oE{K97C z38*IOZi$bCbG~UMA=Ixn-1|pwq7F?eV8+b>WW6g9tRZQg{CTNw_ls<$Q)KMcZ+G)= z0AJj|u;7VM)?F)zOTYa_cFv%d`6Ac7w6>`Ty&prAjV#^v$6An*-J_IK6hDV#N;%5k z{q4a0R7a=H6F4jF+UWAIow=Qvr_6G&k7}-Gnij9Jc&^k zfCVQvO(TY_JFZ+quCJH*Z3n4>Pr)%nk zv1`{j`d}fSLBea0z#I7s_}5H6bR?ddL22Hi!>9W-TJHH@#LwO zB#e{KjYp{!bMenwNwT*IT3uGS^XlE{@T0M4&WjHxir-f z@uCON#vo?aRJz`&OjQTQetVZ8Oy9j!%LWu z>CEHG96V-4_o-nDjl`5EqsScf0#RFY`qN-H^;ZyMHH6E;s$Ng2&8(n`}5a!)?Y?#^<)Ic*F-Sil#|scmBnEN~&oq&Z`p!?e@Oo`uiBR;)8} z%9SvWo}ev!W4~5{%cqm1G($2oy?@cbP0%0?P$$VCp3N80oEyT_5J9(E&XtD!qbOg- zY-qo`t6V@&9b{Vy00DPnfLZ~?=I9e#yOe1x=<-eVtGdpO~xkU<1dMf(dqGE!UEu`O^% zy8vq(Ie7~=+e9gy`mNlr_0=F^f0e(aRr0i1i9E#u_b?%)bNG#WI$Pl5*@T>sXDoNO z7E=J==1_V@6V|~~?(6p+_&M`NhrZthC0HH@dg=&IuuX7f(*x)BGfYvttiv*DnjyPd zPU|lrG>Y`yTv^Ag+}gvm-nqMlwJF}6GrvB9?N^&KD*P}1{)UNeQ{~N$Ad1nVcncz% zE(K?w0L7_xHcCSF5-oY24>FeI`rm;-b-08PfM1`aDjD5Q(O+)8?(mDdhqfC+?)0LxU%&buSN);AdQ-USE!?MLHr{x09Cg&Y$@Q?O#X1u~gPs!wucV z&FY?#o|0<+#uh+u-py*Z5O*Vv+kV1fsz+6ufav2LOZRNba*;QKUoxzkTU*%-51$4V z9xA7m3z?y0-6MLv3mw~hO<4rAG4Fn2{ii9xsbAWNnmdqaA-^x5CAK|@u*U35b{;zp zIoZjb$)=ORsdGNJys_MuM)NF1KYW{}`B@I-q5WUZ$7=31)3Seh3=FwWf(rSU@YcCM z2lX{=uMgKoT<=c-kBd#w?rpCn(^zx+N0OyfO0Q?fVMP0EO1KU8g=bN=7Y{{fE zuuxqYnu_=<3uI6D{$>VC?|YXOW4F>Tt0c>95l0xul&m9+lYy=xpZ*@VV0g({VybuF z-UMcE3!S79{M!C^7)R$MKVCHOjQUG;^|f=qO4L(}#5~Q90uR)Z)D0JczO5-uw@wmP zV)FGJ=GZJ7KQo1(k$+1hM~oA4hq+~{Q?sX&e!tGYuOjQ_A8I|zj#T| zL<5}NV#Y;E1Z-%rLQ&9;tGefDHYu&w#={Q!V-f^R`~B_(77(OWc~!gFi~=}sw}QNR zAt%{AN{uryd`jyyu1@l%uY^XmCbHjU{P7*F$NEA`AZzoE?&3}&Y&b~}E~POgKBk`` zS-y;RT_86nff@x6b}?Qo@MVit0}hwfI->qn!|vuTl1eprcV<|8=C+kjI{P!^{puGo zt>+rY?n&JK9_1$Bz?3%PN>%rhGm=!*yBtn=O27)qZw#%#obBWCdGYK(=<1Ca{uIuX zjbFcVoy8&E0ts3h`T>bRWXHim*js50ezo6RTgC$(hJs||rOx>@?QF4#2dA>mhhG{^ zIw?a>$9Vea3Q;&sdyaX7_B$mY5XhZk8!fY5Yd`qjc|oNtnyP)Y)#C0NBRBX0a7?=L z+!$(Zy&tV%8SO-N@u#&mm6gcVSeK>ytF^LNq7pSc%&h_@_7j(QT%MCbd6+k8lw|hf zJ$id5@*qX~OEb}ZCaX;s%PyyVpq!5Uoa=CvO9ir>ec?z_nJD!mBZUj{WX30RqJqhZ zdZDqlTET{2YiUFo_Z2LT`?pt-wa)Fox0Lx{r`dN0$MnN3*CpQiqvM`Hn2m(X4|;Pb z>`tKwFmH7Rt20bBAL61TRdnp+nlAX5&MhMofKyo`+>y()O8+rsi|z3n;rLw45*#M3 zT%SG9vB9fA9NeEuLvVT(F$u}-3D^Dy4Lkz;5`_C zS#@#G*2Z#EMc0BT;Uun+1}o#!+*~E803;Bhe67 zCuaQ!;#KF&{ro7k0D0#xu4zUJA)CR=)RsP^r zO+f9M%CoU^sjCK25OtF2;_*d)mjvs+ZGwKp?raI;n?K{8`2J(X*{;y>A6gcFdlruX zRglSELogJDs1WOP)7fk7Lakl1eZ14Bo?7SGNXQ<%uuREiTunI4p33VPd?!#?hu?20 zTm$b7IDa5mzCtpJ)Vi|8mJ^m;pDH9I5BRt|-w^jZlv^`$zP0-cs}1GieulM#8h;rW z1QyWd5OPY^=mc@`Id8z~h_x5sI}>?=&zODdHQ4=6j5a>kZ;(%6am7S584da2=i*gK z%OrRXGgQ&r7~SQq?|r_2IGDlLhW*D{zx)kKJJZarqh-G)e`Wn#Al)1)^g| zE^wqMk0d>u=Df&?JHN`hrr333J?47pkDznvj-u+*g6lJJf_b7rp7o;Ev_V}CwIUsK z5o||2w_emSt$S6Q#(cJVpfyL05D>T~Zxc+$%bj@$a#A0qyfW{-C>PL)adS*$%E?y9yCgpI;qOzcd*@Ws8Ss7uDYZ+O_5mt&0N%<>a##e3}P+uHb~id{_Xb zK7YJL)TdUZvUzBt&b_vEqXQ$i{l>deAMAg1N3xBImXZ=m-Fbz=II&~jYM^@D4nEA1 z#b4873^WU^NxGVSJ?ZS!vfL76xpSk|5vPgsqOH&R2m7J}sL(h!qVjK#gtoQ4W!=HyBgLQOr5DNjjH6T)-Z_^-7Wj;5zBi_i^lbI zh>fRJ{2?yCf79@S|2bayK-?eI+=vraglM`p@_4xe6tP}Ves{?$S6--5T%GSZe<>Vz zbGlb`PkT57O`22vWcqmDuFcBJGA}YNY1tq?@r%DAg6ldjCCo| z!g--G(Zb!7`2+M=cn63eBU^7lGo{o)U^=a8n5HDIrO6vvHY^^VkVfACq;1qt)sUV_KU2V+31{wjL(Yr(*nEz^hfFS8*ynZ7Rh&Kyz%m(thw5xmd7`PC7|;mMGGthaQ8(ZGWBSVbc!lH!Xfz4^qnk~28;Ddz4e z^#ad4ec0q~$TUtd9so|nVL)`i9lUS+VvjK3IaFt6RCdY-_rF<6&)f--4gt_ z7#Fx7Q6fK^qXB!koEq!%JK3xS__yMHA9~w&XXd+h^M`dj@tf!Jex*4_fAiF;C7zh(#E7JOhdfcQl+_OOCcXy1Sx6pWHkLky*ye~w0?A1 zRofD%ofzs`sYBncWK#9}_I%AU^#X_OoMh!XT{2fg_U(Rr{gIR4Y6AhJ(3bRN=XDM( zRCw>V&*rEWvjy0$!r$pVU*Zw!lRmOihgy@7jL3nHgJ<4XI0`?m%vf$4ZmTHqjm%SC zRa(>W`=zmZ6)0Bdi(_?|O!l~$UKsDMiZDB9-Z?j43xbOC*ALAA{$)gudSvyj>x4pt&2$lTm3z84(pcy}9+nGuhR6@i;2VMz(J7!no`yjxfbtDyf1FC!SLH*pMdu}Z@L8K@_Bsoiq02xfaw~pCVe<4W zv=NQ2-L4I*ZnBqlDJ;_qE2UE_`l3S??Md&(Bw&5X?_c$pBT>g9Q9y2JWvaGaxo~&;Xx# zS2srL!$r|^uP(-xdEC!f>vwS&^hI`eOZ%`@E?i^28~avb);8uE?IHs>${w?x#(&H2 z+lAi8L^&IMbfng$)hD0;fkxxYtm%R#(mYwf*y?^M=eU28OqlBEJ}VuAR?G=iN6dz? z4F}uyx_4()d+zzbR*gdB3Raj(8#D8a?y%O!xyIJUSOsi$mu~D1BwqBh2#iTgm5Ckb zmlhM;VQtPji=AQ0rx4dAH*R0}ykZ~2!%`@(!}V2~WfhZ9v^_lZV4NM%Gkoux z%{)&ki*wq>{aH9zll#EURx_L=*qmj-jA-A&>0$H>4I|=i3*)0@ZQF>*QRR z&6(WuEr5UT`qy%PAD`6rvkEqBIjM7t1{^BmGtI%$OUj(mB(#^1ZhB8(8vj#v>s0myRQr_M~a#N2O2J zdh1MOh5Q2VeD(XLKupVQ_unJ+g1Pi)rOt%P4z~>!NI>G=Bd$6*ML5gbQ`b{!=AXN& z{p9jSKszVEro|P$f{9Fq3d&>4roy{Fh(*ojN1z^zDF4CCf(q~FZZ8IZu!QRPT6oiZ zcwI^_^{!|Z4JloohK&oyvccqCnvGjidHQE>v)X^R;9{o6)Hv~7MT82^BCgU=Qb7JH+Sf$1&*1m`<%2)Vg1F65)lE_f z&u5mf*7X)b9klRpjgQD-*8P=Fi?NlRdSkDavcnS=xpIh*yN1*k(^%=*5G+}{eUc2- zNV$9KRFk6=wH1A`SBjffwcSq~g`+@Xn}AaIYVgLaR#Mf768(;eb3i!dcZ%?9kD!bQt>3-^ta0Y( zzqHQ#H)q;cYMH!MI@9gG>teru<--GHBvOV)2tqqGqwA&{+E6tA^CG4gjL%;8lkql zxK2;4p0gTTi8^?_>~>M7>j=JmzZ7llknOw(gE)t4D10+S6Y`^6tTfT%3*{y1LIYx( zo!`Ml-!MQs`1$9AsR6&na_Oj!B#)Ea<(k?whcDV)@ye&7zz_*uK^8;nQlSkY=%rq~ ziS}+jm&0s4_6II_|MCv~UCSe#XJns}Tj&UL!M!Y@2miJb4*~}?jt8JBQ?gV{2k!GL zk4*Gyu_8hlR0Yd}U<}x+r}9A1gw_yb2{bMJV!4>3V86a_HcRPgWtW^`h|R?Uaj+1sa!VK z6rcSW%ngg-ZG{3U_RTBkTiV1~@6fTR&qt}gw84o5!<*&*w!xZcS5(612KWuo#q~y> z$X8~RK@Q|}L0Wv;$Q-H+{t$2z=!m)Ull+n$K;5mo#eMvw_QR;)pRUVzfkiGhn`;|1(lt*oW52LQ#GZY-=0C9|APpOd5Rbae~fU@MCY)N9eTVRq5Oq2J;kEHk@t-Yx{6 zGS;CW?Z|tAurp!VW`7=N#_PpNF)TwA>GPj@2{yI(7Z^M>lBAAkGs zS>G+zTw@+A_f=J0>J9ywh2FA#5!E;QZPC{Bkg^DZ#gM;NJ|FkGg+;!b^h0G@`Vr|y z2>voU;AMlqL9^^KDwAfEGNbesk&IelTEv$gziGXv8|hWOHIB5JsSwCBl9~gRhNE~= z#^dX|;oc@bg!CfVeBK~PGvI9=5*>N)Dq7>A^Ga=+evseft{6XfS$H3{<|_62k_x^a zD%>K#=rZIDaWNg`wfQi-MCO*vGt%`?#5;8YZI>N@KCP6!K2|8c&xY~so(U3le^V1$G&PR1XH5->^g6E!%^VXf794H`h)9#(}JXUVD zL4#_{0U2*?bA@Ty$tCdBF&c#1NzogZJJ!q#ClF~uMq{ztcI0Ytv1V^B9r*s$7@dtQ zf92E8ZXCZinqDN2iyJ7$ItN zd0T!dI-~a+pm`etX>(VS`+Fq@L4#~*5PRIMWt0rw%{6PWu#)EC7bz2uUJy@cxYR>S zU|P=FCK5GDTqAH<)7x`tQ=TByjmX63I-^|`^^(Vse(iO4QmKD!otBUH&kvQ(2e3NK zsh!1-t+Bm(QN8HHSG0yy^cJhX?~%tRF^%fugRbDnKYOS0%prL?#bVp)Ske>stvEEr zOhhLRNVGs37RjJpav5js@e@;;h(e(pn!3)5WuoDvetnB?2yW;J-}d#zU{r1e*bnE7 z_}$SxKIi>xQCZ=evS5~G#NFJf%ZUr>6^=z)wrUv{%ya~xEbKiD~LzdpU?vch5wxxEDGHT`1J?d2AvcZb!5VvC`7ED1lp7N0bbYD(|WdpKaHX zUE|*0FpJa@;>%YhyfOb;)4qlLDGOCbKhYPDY!vCjyG|;vJy(zWIi}tI@PhN$rKRZ8 z&i(aY#7@Vct;`l!%Y^h-`p$Dc3Wfvo4iAIq@4;~(o=J{#@ozJJ_EbydLRF|kBV z?uUy&h?ivkpV?}S#@$J^n0VR+F|cgX19G}^^Uxb;Y)cqL?K^0Uus+ddIKNh9!!J(W z56Tt3UvgtQR=Mcgb7#5Vr~cNx@O##fn*E8VxVCOLy=ENiXQ40{YZjZ}=6@20=T7*c z!kc641YIy{QX_GGN5`L4qEVH#Lkz|W`5zK*yY0eL#HP!9T`yB#cfAE~bRI-$6yDm( zzjM1V&RDN9BCeAe_IKud6kIi$-;*ZmUTRGDEqhxtm#UekEwC|ND$&|9_6k9SB-S&Hs<|j)P;7siB~U?2StCg^<+n0CsO00)*z^# z%*+R6nvl@|p%W$EA=H!isEtVm=Mb1y>w{I(!US_bmVMf{ZdA({;e zGb*=4lS&=3Tnl~4uFo5*oKJJjZWGKblGw$DSe@wKI&?N1BeBTQV6AH&AQt16s0PO5r3!TH&Rm_^0*z9#3%;(yOfUkVLKR&;q8h?;C&4@*8aiAcf)bipfBV%z< zDa(G&-vzyw+RLHg)hyUSa|*XuQnxoRz4h#w++ks8zxJ5n2W!npPd*xIfF4rQIl7L| zP9cAVM)`Wj-mQ2o2y-Q-7zm40lA0y=9e&GR-fVC;TrPPu>W!W)M)N(KvBL5K@$Z?# zs{oAB-Jtw_{74n!XBHbXmk{GZWP)tCMzd8>$$m_6Vs@sl zt^1uK{*1BYmK=(v!*7pPWsc+%$zW8^%a5}wfNn%9k6xaJ3pWQs6_{SE-xNNhwv4pkGOynXzHd!aRr}>RG~!w zwAu&=q-Ko1S9;1w+ulzVEcJ!u_t1DkNB=hZgMDb<`%~{|5Rzd*UF~P9!66$7j@^9T z$#VfS&kqKlBay+OoRZDI6M+wB$Q5URuvlI&d89fXL8#KP;}ymkbqubbN@I#&>W2Dj ziIn=>N*_)JUr{x+*3|0=3J3DRLprTy2&%ZM+}o2Z^qWup+P}urx6bISfwYEGw&3XJ z8B-3DIx9f=Yj&Si-Zm5AJx|xky(tcjbG#4&aqpgQpMkwL@t!^WXltt_{%KKdAL>7yuOd? zkAj~SUZ%EfUfm2?wSCrY0TSDa1rnemu<4aYI@sh#2y*~G?6%y<_sI??H5EfUn|1q0 zgLLhc>aaV`hqH6ySm~3C5E3bPudbiGqYJi!2U-CBQh&o7x^jyv*PTKr4;=Ral_&vv z$@P*gxYd%l*m&C}YwQJOUt@m0k|1S%GFQOF_sIo@$N-RQY#VsA?xu-sB zMZ`NtSvp8o-AzAz9;yNve-X;)-p+49Q(AUoWV*&&PeLI68IDnpohq&UMCl<`Gr^=E zOaOJtC(6*r&SrmL*1Y%4TRm!`Cw~F10lQaaG;c;D&&WV7Q6zUJW%{%}_IJkWI1n%S zf%_UQv5h9d`HXuA_o&3{qOM5xg7|6)BHwYw%b&PdG~9YQ^1#qqkv}|GL zZ_Ie(g7eCi-5;OdvH0??DBSeW=%^oTkH|lZ9#@F2k|f%1l^5ICl#+-R6dpsR z@++myw#n>KBSImm3i%s$y=LAgj_2qhCZ#9yGGdLYfVW#QX06k4))47_%u-$ruu0@l zJHA@YQVhZpt%2e@dRy80mI)9l3)&Bgcg}V_9T_7JYl@5g%j62TUHs2Fz4+4K@h^}6 zSKZU)n3+MbCAvn9k21DOV%j7$nAl)h3C|u_;SDT;j$ra3I!uaKcH)Ri8 z0AjIFC)UrK<&rZOO}o=`Ez(b;K^xf@ZHiuUal5=P(fzLd2fy+oOi+jIlc1tBfZ6UD z8pRAdHiED&u(h20ec@qr-$Zbqb^e|o8FW99O`L=Qq%j8@FC7~Flhc0~nN8A^Rbi!*4nF$i9H=B4kksdgdqN-eZxdBRCgdt>W_4Pl+vYo zh;7@jY5}d^zC6PC(`x34)XV<;P4Ppi&=6mD!w09|jSl2>E5TtACx`3n`Pu+3dPpNE z1PARixrYOWhfF|c(o<~#edImmSV4?@-SDMt(oEvJf3)ab7#J{hfJt?3ol$Kc8)^TsTE)_=*@}5i>^(x&t+cFfYERx)U9F7XHFDXof>MHT zCuHvDU|}1%fX2JVdv^F0n_mf>qPV&vpF`7m*r-}t;OFU70qH2~lHv4{9qG{pcKJV~ zlvI)sRKDo{u=VEuQ26opXtsoyBI{t1t+EWs*arz&qEaFII?C80jb+F>*|!Q=#!^(W z@7vf1g&8DcUq%d}8S5B>`+0xw_j~XC;r!Le}Vs!VDk?KDxCh7`@SL<<~!@*1$K7M3ps75 zCFZT`08z`LIg9ZBqDAm+U+W`95%FV{lJl^^>W z=HwbyhyTOEeku}+nHsXE!l2o>TV|BSqi-BEU5hUCsqoyt>@{CS3{^EoZaK${9x4EW z3HMAkHd8KuYlJ|dtkbRr3thj5>d9+LN#lH2N5k#E_Kq*7?f|#~%?+nm=AKf|A1-|d z(8|MJ<&NJ7d3hgKKmHUL?CbUY$=;s=pU(#)0YmeIraS!`ML2gt10}uz<{4fAzB!$t z2XrehHOEm?e$Yb8!-G3$$8_II1^z8oUXnio2Dpa9_c}`m=7AsbKC&J`!e+bG9I)QK zFX$H8pai_ijrt z;-x*uwDC#P)6(kr4`F0MB=Yv@uQiil@Tw512?W`AD@Y9wn-CtGeH+GtY*Qx}ho7kT zIw~|&yU1o+wB2>Kk7rfvfaDjqjwMT-j#}7SKDBN&1g~BQh8SkGlI%^*3=ydo1Y1wD z>-|KHWnJwR#kIGGNb!r19X$M@=~a1+#HKr3j$AG(FC&Q*)ipyk#?8UynZq4yZCYve z%+RapjJ%|@{s;$pf^{A5LQV)eXf<06UHDbRe#!g{%XZpx`HY^WzMa;WRxxl95cnC~ zAO-Ep^r?2MT)wc|<{*UU({={#bU#~o`9ia2FyTYh1b7o3`LT8C-2wFu2<-YsaK^Y} z@A`+(ll@7UR*>KcTp;)>2X8Oyj%O? zzcXsPQ_a@j3jcZM?9^1PK~c%Us(f^fj>>iTo&Y;L%eF699QfKn{{uEZO4MZuo2Ho_ z*g+Wi#r}u4gYUR4%v|`txx1kdLzi1$y^dwoO;zq8w$7W8^ZC4W9I^El=8jgrUy^ii zyWAf(5UbE}B` zKId8hE^!g(2(YYsp?gN&Mv|EJuh*LnEMyR`Qd!Or-#Yp4Tea|+nmmv#`UYmG33^T; zn4{lc&s^N#0g=0S1jnk+O7DU(X^(D=o?WjgD9CrT@8H@R+`_EfDPj+aQ$3~_c@vJzJslwz+coXu})U!KbSDR@$JB5A-z-NvmGek*B^O+Mrx1zZ-k+M_#bVX+dC1 zKPBdJ8*LHVe1t|nCyBLpHvC9^k@SguqI(nmd{+PW@!4TM0U}o z&xnfrX!)FUoo5oA=)nAp?2*vGra%YoAIb2isVv&0f`0=CFc{H1E0}M71kh?#;L=E2 zQzEIK`A(dY;c(+GQ(`RvTfuUh5N!*Gm62qj$~~vuSSabET@`Y_&c!GGZ?docf^v)N zV@C&y>3ZA71Dh>^xc%>=ds<#aUmegI_!XO@u!P%Q7rKS)ppo|}UiO`fjeiYo_NEL4 zssxb|DYg78R518ll|K4;T6;lW;zZn()BK9w5dYpC2RiFn_aU$~R3WVWv%^vpB^IzlWa!v) z(OjCKe@f~DaZFcn!MrGc2kag1Ln*QRDUNQ*D!A!i_2;mqg}Nm;$*;aYx$S9@TV4U3aQaJgoMaCor{|K=7|K0H;)#Gz{KUL+xaGpcWl*+1 zH3Wee9iOs_$v%j0#)isp@lO(Y?M`|`J4O(A{_L?NrPGPbkmoL##mc_&`f~985y#6o zX}mtJ@b2p{>WaC*1}ie?@|n2h-TqfugKBu`(pl?SqMKlKE~m)8u&>`hm!lZ}PPR{m zAxSUfR!9lxwv4kJC0^62jX|Sk?0qgOe+f>+YD%A0t@omzQg5)@TzP#^#$=!M#E}6+ z&tF~%a_QYw*xbxtqq+k(HQH&<{j1Mm_6olg1O9Rau;@{)&GxDVY^Vh<%eq{+PtPzi zi-!vKX<_qlBQGUxO<$}pyq9w7QxtU1+k5wZKAgV6&$aRPGF-j;quugfY3-xIA=J_^ zA0e+He~B0A2R;(Pw!ONdW_7rd;H=QnABA{eKj0$}uAl}|lnPpZ>{pq)*u0jXgY1dq zjL+E{4V9Q*LG=bJ;8ofZmFQje=v9?hQy&TMw7M*0S$+1VKBM6bzl$wuCcpG)i6MLaWRpTz*`4 z+CP6ybSc98Ff`-QHOci(R#^hXrYK~ewkn;on+@OSvRgJ5jnfoLpT$NdNd>20ni9^a zGXAg*5ZQN3h02XJnR^e`eTtLiyH2S4bUqoPF8P;a?0luhlf#fzkQ{?40;{Tz6|Rga zCoXwdTDp zPq$i%rNEpxGg#&JW8C{9wS@_qxsq*&uL~P?tKUI^K9||exP0(M9ehwmd*4ZPAlKo| zZ6HVbOX{eKM$NtA_O)rtWsPUPcX4U9SVw)WzFk@cruGiFl;)(DAJTBJE{efNMo@-R z1PrlqGdr`YFmw8i?8Lvm1)ia0We%(%kLt<^63WR(*9p&rzlCX1rSe$fAk3p!jZ>;R z>{Y!uYltM%OuzU-=OoY*llGQw=P#a*;2i@2s&ivMk1@K%Uc$t6wrW40zcPWSh; z`D*Z3B9FPI;u?Uyj-H)L@s~WM32#FWWr2YJ?a=STIyKbi%59Xgy#kQ>ZN5QwE+ufG zr)?LXe3IfG{?OKjx^EgjS?%hdnU(}_f4#c6#|kF?MEb|z0v2nF1$0Fkip?mO7b2zT zdG!{!?D;7PEfT=)m8M}2)wz`2`q%3e&IAW7qJ^gaqQ3TC^6OW5>vQ$B%lV2PZ7sOK zSr?@ZY8&m57ukja|Nl5!bo%#(?Wy2Eo$w`GAb~aHLBC4rRG|+zPB}s8h+mOG8)q2Y z^#-s6n~D)ooTkS@H(0?4-w^}KvgqlH%)O-S{6DWf|2A7`&ER6EilCa60^sRpAJYP( zgMG*Td`nh`>E2qu5^W4GWakp1({KhGPjenw1%YYh{+B+wMx;QT-dNJVW9;Var5^Gh zEVoTdvL-GqDj1C3T^UcsP0hOZm~5wkC6%T| z4Gk}L8Gigrt>_t0T3YIq!F01hoMI{wf#946tRve?6Uu6P<^V_r)fcS1lfbpBg=~pc zF(vmJ^o2SZ$v1qeI9-|8K6=Gnw=W$ikBRJ{^Kt@vY5EJe>f)_@)#&-fmYe2Tu$j86o=NhE{1xJ)GO&)TAdgeG&aqQFZ8E`?%$q~(OYjZ&rpY-UNp5ll|epqK^_TP3uN&!ewUjxuH0s67~$$F#6f_}{@j4BymK>* z1?R@E73(}rl|I*(Qea*n_q?k^hoGg&cZH*5^l^XpeckUgiiBdPHjUf0jpOa$jF#@V zPl9qBfh(B-8I}ri4Bcf(!ZoSD1sd75BmIUC#@ec`AWx-Qq!hmC?AF~ z>1MbAjg42Jd;PyO^6xWeY>q03LZ3KmI*wZIpB2B;&8zlb9Q9{FtPC z(%deq9A6XsJaB>U7#>{xIUY*Z3)|UOn|A~l#*84xhZ$MuWFx@DglHbG$rc_5DBR0&>M<1g>6TcfLW-hHT?M0o(3`FxIwf28mjH9MW z(OMmgiTR2@J{(zOJ?j-O4I`we(;2@H^jlipyS?y@)#hc8H?*s@`SIDbI%=HRldjmX zKPXE1&ULh&U->#KgXB{`1&>8WYv-tfz*nu`jOTZVqtTZwQWhZL*_K`x(qqEM(Q8$c zDgL4994Ee==AVyUZT^nK{Ne?AG~Lq_p7|sTXwk*&Sb6dEO0t_G~xWPV6bro0xYPOL>IDjA3Z7X-c`PbN^iHw^v<>VMBo z0zeF>Y4#)yw21>;8ZFik_||XOJX9#WZ@%|V=(h&UguJ{)@{z5rtGp8j{Exl9c06=< z5b0hS>g)gcci zoc-Gf!yNNv=BC9_bHkr*EB5w@83}|TrIDwnA7AInxE4my73mb%=92*~!x~SM17qBF z31ExLLeK8-`j$=nQ=A_}!6NvGKhvljK@#Co$qJ1}rj}&dmQQ$yA=>_t5^fMjRjGN) zwxUIIF(grLtX@;?{?+zRuwyklv`6ZjrcNk;W6LSRVOJJTsdILO%6RkozhL#r!uunL zt&B4X^arZlpFI7^&+XcXOQqT;>+h^>Nqwe3=a-LLdZmGpz~E&2TW)%!3@t&YPoUsP zR<+&&XF7Z+%QGs=M;d+BYzR46R0^522Q8c)G_Gk0e9Ry;gfkSL=J;ULj>k*a_N1hG zq&5EttT^$|=;P?q+b3QB^*}6n z2VJ*>tCALoOXt{WEPWouF(th$3pDOa95pkkc+_cKM&0{@rJ4F;g{$aJyHuIWJOR_4 zmp`!|P{dXHScB?bdGTz8%0iyKtFIPwM#sFIwOS5aNvgy8f;Y6mR9l5N%#mV5eGQ)iwwO7NsJ%T-)t*(zH!Kf;O zyd=npR44L!p;Y!rGv=@H3Hlri9TLB7B0Z>6j}rJ@-w)h_NYtjkwRAx4k6goW%C-?M z^`TB1j?iGi_4vY*S-vCyc9;5tP2#ep0Z<5@J(X#my z9bvT1k9onX10djWz$)+qu#>F|Hv9v9@B+{uI)&~U5xA&A1GwKm;k zV4?PTNwg#Gbyyd*UmK{Kp_Muu`FP>d?ekh3r8K`)e#$+|jj!Xe1uEYMUE3CC(lkTg z`>lOuKBVNdj~z~rEXNm4{E={oG4@6YcN z*2Xk1<_D8@)57Aj=ejOHNS(Z~!-h2R*EK{2zRBiPN(~D%UQJ}xnGqqR0$-@F9tb2| zM}Fo|9an3|qm*nEQh?^0 zRo(rfq~zwtOdCD_?+B`G%S`$n6}FynzeXHtl)J#F1^nYV zu?8T$FCSBEz0z1~;m5GB?r5@6%N95wzaBVIxyBn--lOFuyjha&rCv;IhLEZ(T(&}2 z_^7WI^+I+9X8OwfHJ`&wzYs-3Vq>ywmH=A*|geD>q^8ZAwTrCaCfVQ0*lp1#0BVDbGgwFnwPN`+Ar5AZ_g_MM=U*(@G z_?*0!|9mpk=R3I?5}_sDcr@|T{^hG3Ja5Pgs7{XkM2t1&Sw6FL0NIo)>|Nq8j=Y!4 zMPcPor%YM7AMVK0#?K=)1sYxPAIvj}V^{iS5EC@nsw?IB|8hJ zFIXw7)JVc=YTeWq7giRjPk!-_^HoxKeAoLPHY^G)U@NG@KYOSd$IEe=)7M9~l`*Ts zCou7<|7BwKc`BvX#aSQVX2fNMeMC0FkYWP4#x3=I~Q%_VNH*Z&yYsuo%)VIIoW^x~BkjPw+)<>YD)r(b503jqK)srr>ubUV9{kLu z{cO?EbOFtyl?;EU?$|G`HE5wUW_*HhWnQ-;re7bkFn& zw_;Z6Bo05C)+j_NsW{%uyBP2%+`GypN0pK}_cZ}BGa@xt+?qcsW;*p}%4*2Bq4W7@ z{~d!Z`adPBjf7Ll`*1x*;}m#*qk;@%iF9rO>zJSnw|r@SKG66nZYq|=|CVdWsB`i3 zJa+MiMpL3c1=Iq9LzB=~&M};Dng@XMg zC5$03ROlb}iuKgF4#!=Kpg->|x5;m$_0g;Dz)Zeyc%)G4G^H|sH-eAPi*)><3Z~*c zo8>ZI^uS-m^a`xy6AzJjf&5?*g$q~}>ey(lD7Y-r)_qZnHGi?6zq^O}`ilz>&v`m$ z+prQgqEGLZfCWZ>xIdRO{3nL0sK_xrJ<$l=8_s;(QVKXx%h?^etfj|f;M2bUbXbb> z<(NLw!-S8+FSPhfsf{L=wWx1h-K@7KWGAuac9-1;d?w(JDes&g(j9?wVJC_m)E#Sw z7oZ7ogsPx-NCA!Qujp{6N-W#Q06$vdiCbyKu>R`H&_ct^%|4}uw6;~Q$*D>uARg_P z{Q7Y3uZ|uP@>9B2q+upOZ+f-N>FYxMUh^&u?x3y7yi=KEqpk$V2CZD#tO@On<(X%?TQ6+?d?3AlAKf5bVvwEVu88sw@)0((x%w7%Rs~*ZQToT zyTb4ZF|Z5HtDwmn1E2lE5AnFMGAcc@YU0$R9U||pvexQzs%X&uW)OSi4}7}4#A8?8 zbu=I{oRjp`UO-bc#ecJVQ0Z3rj!I?|3a9W0goL4%l@jaxzRb*)AY)APRZs%Q)ZP2g zpv0iTS^0UyYBsDy6I&DLF?}ieHQ44`%K4AWFX&|bjrSdYXBx$w#J5%Y$%ZGt+OvdG zLTzfgb57!9RyZM?fS&WCsh-tJkQXC=LWasJ)}!GSqc2GuXRAono$w7rNoUeU_PkN; z%^J0qQ8T1KQ>E5NxbCSuggdiJTQTkcQ6)xlbGTLJQ}N2(BuSfIZP>89S!??)+!EzI z%tV{?2~6ah-H-EW$Crs1lSH)UA~@-VUAkjC$$V zhF8vBXRO@lN*0FqHr}HcK0S~Kk9@z(&x7YPSzc&!RT@(k$*^qe#Uv` zhzA?@G30GDPhR~Z%Nh4Ms|((P8<&2uTeY*X0_r3(BH=!B-xAR5*}r%Q-T38d*EygU z%R1&VOl-PQzS|y2=rTNaI->EhNeuU_wA*`is-uqau17uY`5E8^+?ZreGQo8ZW#$MuVRv7`4h|oiWb13 zOB}nR`EtER+Ix0;(cl^jQ@G{k@ImhO88 zNj|`Ln@p7f!7j39wG1e93dM_GJ40G`H?Ww~QoGWH=Nw=D4Mb-LQ|vTxLV21-@w$6! zY3d~YZj71ad=+KNpKcT|z#`*S8`9LoC4Ta}7&o8$sg-Ee^akxZgng?hX4DP-xN=Jh zes3zPM5Vz~_l-RKHVGZpo_8ZVeRIj8=xvF_NLFksI%5=F>^w{QDWHq8O*e*k5f$4s z#^g42x?L?8lOE)To1ED|3!FtMT_)L1!F2^{v)&^}SlVt-Px7=V0goADk)lgPuy$Y7pq=+tTypL%o-dP-T8z}rIGy$GPG%TbKBu^^iWmR2Xw z=BLOHkrfbRpgD+S?R#%+bVcbK57#c1TD$Wmw)hcfqe^=33i^odkv35~4{4a{3!~nF zkOYsfX2zU%NePpCS(VnxyHkDR04q-wQw?7?arg*{Os?74T+16(4!lW<_B$L{3=6JS zb;?!k!Pn~xSqX*nt;v2WET%%P=R>t1;YpkvdJtU-vy96i?p#RJOwq;gI4K7wu^;9- z<8ltqxnd-(>)pW|BUCMy%N0=KT*Pl&*gt|UjC`TrX zz@=Po$~gdm6gB@8J{~Th-grSeb|B0#R;8tnDVX%4G90(Jg=yK=AaKsh`F6+PRM6yC z(^;&;x-b_dJbgP{<5}m+)Tos2iw+BA$aUOh2l!>Pu!@9JLMb&;n6}tT%Gj#m$O)T% zMhL+BadbPJ(I(eIOTwCVJ*g0!-P)b`pL|w56U!Y(Lx_PeLad_BLMX=oJwU?0OiYF=D9GDX>}?OH>K+$tMU4`wAR-s_dcuJ=C>ik&*6E<4Rae@9n4~LmzxrYk?=@+M)&v?Vde@3pKgt|ogjBN-sX^DY8E zDqEVWq+Y^vnh99u84`5%KFIMZih|6H#-7aI&pZM@`kGNrOH2;&cs$!v78pkpSsxP6Z~l03`JQ>jgrYsL4B@ws zyPeJCvq9oXVLydCQ*4}XgW+*1g*Jq|q%J{?R1 z-$M0#lUDt%*nR}$M8VekVYCMWCT5UjeYPXCYRoLT(dhYjeBKqArHG3j(I*m+Q-p!R zVLkurAMJ2>&{O(KZKmguE$!ZKk7ebV2wsi9Eb-gUA(}9t5Z2B@4)kmjGj$pwG#9QJv`djJ%5~*%JUj=MetC0 zOx69fpU-k=hI*jw!U({B%1*+irnT5bea`JQ`{ebojcxV+b+P*27=XySa3T7Nf12aK zXq&`!8R~X+XXQE5_UUHJbB*VJFv_#YMO2=%U6dwVi8i??@gmZ&tMZ)XOKG-WnFXc` z(|i&yDgikB>!t(YlZ|2-Co_YHK;7H4bXOFn{vkjPsn+^W)Sl&EdTHcCqwnV)IR^=_ zB{?>JiZ`9+uwu!Ca=5Z=K%()#(D!luSQQ2qtm}t2h*cf5EcZFKj$RYTB%g?iPbT42 zsg73Hx$A6xr9CjQmptQ^=g{DmM@zD;M4+SZIk0Ejm1+dPzrAe7o%lMgrtX@yf?%;`+dT8bK)fD_r(OsO0TH^&z73bznnH4OuirP8Nov>I!|^BLy#Fi zkvZurg&dGK6_I>`1o8aP5)RdSErc|ERih85ODwE`LBctKe1j&r1Q{W2p}3n`Hk}VF z!#&17<2*X!+Koa9DLr>0Sgrx@c&yuq+?KnFk2$7Epk0j_U|H`C&Na;>?EG+n&uRi!!2pu01AHv(>R*Zdv=BasD2z};J2IM^4pIGx)A6Hta zis=~E;Q-U6H0$C}wC%)LNPdPEd?01P_w54hamqgnl+WP5*3vq1sp$9I(U@taKcKwOMU- ze=6;#D}UT)T3AhkFw6PYymHCV0{+jGjcwdR2VT-Dy}Q zD7GJK)TU14(F&cVuEemN*}QY8f6$w}o_)3gYNtDi{d5=GJi3MvtmKoNY6%N95wLqT zbp@0=lDA*BXhCTXLFNNKXjeTj3^_0aIGHrs85)Wl67ILWW%cLmc%Ix@naQ3g<8pYg zHRCUv4R)kb`a!s)*SqCG5vNoNEh+^g#1!yHNdu3*<3q|hlRZ%Wx6B(moyuSrKn*)YJg0Q{b;k5(BHb`A6v)~M|BSG+RoWIDNAv8(uR_tmPQU%XG{ z3NX>cC1MU}N#{B9_->8(WEnUxSz^;v%Md((As6OzhgZ~FlFEzn#luU@{ZbWs7l=k@ z9)Fp_H&S&?i>i2tRqEJp!K+sTN5T>x*0m8_#C|h+TNNeX6n^ogj83u&^42&U-YK23 zsVOncD%Wq{o)}Ssds+hb*K^K(a4p`Sq-!tqX2y@M%i1Lja-n*LF2gs3{z4pw%;^ks z{U|75`bX61jU8>OEAg?ebMo9CX{~qfFa? zOK6(-3>H9xz6L;dSY0yzC8F$a@h%97uRgD%lyzX;UrGb*DUw*_l%In`Ow%2EVUQNFp=H}R@1k^t=wM~^kbS|H$#&M7hZlKX7&HE3@uQ1=5EgB z=japhKnBD2#u)O_(f605xCYrZUZzj#+V}ZS9{KOQ?kp=pqCz~zj399a&7o=yly~w2 zZ_JNAR5R<*uZLDb!8i8=wKsFjreaR(qgrvlp}hbFEQfs`j8+%tA>5;6O>x`owgHjv z;E?-Cg4C`>1+3QWzU29(jwJCm>Xey(7UDiv99L_BKiGYJQWc_&SY>%eU$!_)UezXY z;Al#(UzFw3=;?mX-~;Z2%k6`u^}Dl8-I)R(^C|J#o_qnECcBrdlVmW-NnIQzzcpw1 zeMVD`5qPtaD2TAv$Vc2sJdZ4)a( z_WoD%;b^^crhE}~VHC%RXCXjEz85T`G>M+1YJoDX?7HfR5_45!Z;1<$uLcv(l&tPt zU4MV6i~FhMn(uGvc@e0F#CqiOvjE%lt2+H?T9)gMaqgDK{hDJsyQEh`#$muzgQW*I!8YPxt}>vd}O~jDW#6jnDF&u3di9 z&dJRYO2vbQ_HS6H*Wv&EXW#HLbT$q5AT+&dAaR83bdb9_c-Hvz$agcjIS&`(Awuwn zjN3uD#h4FJq{dJ;3O8tAsX#v8^a;D%!F;bMmOp@S>P%rrsy0TOV~AVJ z<*WYPz0nL=YgGyU>44tzZLhX-K6s2;wpEKIT*LB&==GbIrpK5$mM!XQ)D#b%C-o3? zK;f8IQBb>{%&lzh32RRh9Hwo zVk}?r=LGL3N}e(X)?1kSZR?mnG+c*1L+s{;)oO6y`K}QpLybSKb31I`m|g;Cz0|PH zAii6ha->JFh7X()lF;7KH=b2}DPTCJGtCzdVn{=gfIMoToElZcYSG+AKWAD$hih+5 zcwxWTTOX|;meFmn*I$uW|0kht)Q81qdby|8w_99rDqUYrz-EXDq6!j_)ax6N!p8Dl%g$uZcmo8&u4Ss2mB1?VgZQ70M zxR!uU?Cro%W%^1!;<&&m0+U!tq!3TV(Ph=^^cxPt{dHmk=P6v`0(k!e@@6e=B+%jL zhvh0`A&Wkge_<6kiP`_@0Myt-mCrchQ-a8<5SZMBmfxBBSDSXnotXLc&$jo2=G$=E z3Q@{=1!?6TT*S>0!DP79Ng1wj>8Sm1K4%xK)uVq>iQMws9SXNzA<7*2aXfC9Wzyak z+0*4~!KPH>m5^$)YaqGku3?qpi9^#6m@3&VfD{gn&*O4&8WA3lCl48fCfJGWLYixB zkS&;s81I3sk)%9d|9`)d3FTQPiFZ;kF+V{9M2N-IS3f7&8`wxi@G5r-xfXs`B%xo$ zAkn~*w2Im@KSUIovqCE$k@cal98t_smpj$1p%YdkM|>|r)`ANyZ0c^V4}lA!0D(We z=_|s*RK^8``mDb;Qoae^{&j)0U;GYt4D>sUPu;?rO2SE3xQ?zn*Hps6)#I-OCX&s* zaX8@0{thKm6=)WO&w83S3_fI#3*iy@Uf8!8gqt>38pkHxOhT(tj2j|Bx5iq7xhSBp z-MswTS3Aw-SXW;!Gh*f$wO@OrTDro$j8p}*BH==B%72ii9^Hi`1fIkcL+eyB-_8(f zs-5KTnqM4R_3|#>QaC?427uPbycb6&7#{ZJI^O_Z{}a9=Q@1ac{{%-zNhRKd_FSlZ zPiM+s^bo$W@dxy~MZq7$NncJFT7ET2R)D3>fU^+pQ7QjmOySfXP)Xjv<$Z?9o-C!_ zTx=uGzKI8Xc6!)>4gzX-e&{XzulopA(S@loqK{TL&(o734mUb(sVrO1N*dO$yCkFD zH${l9FW1E6?*TZ1ul;;B(9VaR=;z4m$@V4y(d(1J)aNZc$7vJ7W|c45VEs?n=gDQR zIB1T>5dV&Iiv+7}FBDs^+B?f=0}D*@K~G4V>;%lw<)tJ5LaU+;k{#|UFQHJ~(C*k_ z-qX+LrXegnT_z2iR`YMT&XyAaTVzk!f)r6E)|`6*d{Ln6wwCPCu7)ppeDM_q~|CZ_ptE zs)sr;kN7X?2K-;rO*?)H_rJdbM{!>OGxMADr@EPrW^bhK@qTAJF9~63WO=^v1&r`F z*X^mVz4}Y)?VCsBNtI-YoOI!cs?J}|sFk(43&NBbH0^7$%7xL)m3s}$LyUpCRd1ds zw1kya`f%_4{uFbqtIg3mHI6l*G4dft52HnnT4d->f%J7=(+s^ckLCN0a@p51^q+cp zao0BqadBU~EqgknxyQ+%s>a;GeFf2D$jAz0OVtk~n7qT+JaiEoE5c7vt+{BFj?c`0 z(^=Zs3Hvdx0|wVH1rFhSAMojqH}n|4K!fCpOk`yXraEg8#*@@^R~b@%%F2gCNYW0yPX4&$TGu`##~j>nxd$V`&HW@mHlM_UG%D+})h;u(^q{)u-nG0V z?zjJ2F{GNhrY1Wp&3|;I;qG}lOusfRZb=wTBf@-q|Nm_8Y|Hb ztbo$etpi9Ss@+Vdqo|a?_z+ThjF+I{iY|Hx;c`Zm1ZZ4C^^~{;qWEl-xA#pd^Vero zC%C*TMsInOmoSuQbk&aFhxm|CZri!PIevVP@u-k}&vEMz$lnLEhNVS5xaBT{Yd_yP zg21%v5B_f`{#!TTUG;0c^qV95sMx?gPkPt%HO`Bg7X0qKEssG`9(RAf0!vw4!dDKn zhK`C0Q4A`4ZQR^!vTz(8kSbEYNB7Y8HfJPLt>Z1{c#ev(z9^}C%5D|a= zeG)fE!YVT%q$Igz$Dz~Jgcj9okVHtqw5W(W~l0DjE^woGFEMN z4=qo=EJ`XYOMP$WSt*ZZKN5jpEC3G*{39?J`=E`!K?;mWB(izt1OAfB0n8j>Yg_B-NrH{f-3@ zS%<%_JZ1ITjSHhVj9MO&Ag#u00YlosVgs5(xED+S|F_NK-DoK3M^=H0e^!j8&vC6= z0&=^|#GSQ0&1av1^2%Lp%ncAdKc8SCOA#OHji;a|>oXRzXVn*;yveK!Kwm2$Q> z`)0eLb)R9kucVt3^pF&*iHr^?Y;fEL-AO(Cvcm9*c)Kmu6e)*;94!tl-m(J?gFU_M`d?&e? zGyqtsi*ZAt6eq%0)jV@P%i?UX?_ATXa*rA1Q?gtsH?iAB)(0+NUWFtksjOPvI{REo z{}#3wnQ6T9SLLh#6YToN8Phm!RD*iap^y8w%tC9;m}i8Cl2yeoveHRX}@b zEg$B4S>}s*Wq_OPpbBNXHnX|KkagKJpe~~FT4o^5A5s8rzy<#5$z0=_4A3??aW_s-a)t&)bsg&qOGR>z8tUb*zZ1+RHe+f-mOYoi~_6*xcqt$ul+OFk*RJxsZI-^oi{Kbx?u~~;TEXH_kU9PbFJ!J? zp|W3>yw0ee;0}2v8(JV!AbaU<2h$wwrTJO=mKWu2r`$b#o)o^j>>umLF`Z+K_Hoa< z5cC;5-ppBMkhim;Q{vuJ2OY2g-#HCiPw{gfqSucVxZhKG$>|O+nJg#MRT=MwGR9_SMx=iMJ)w_Wl_Y~>`pFbSECYUX8^VB@OuNsqg!Gd~ydim!lE|z-q zXoqNmuME#BLYeEfpXp0RR1doD^89P`p8MF6?s`v${LLJq$D01cDh=RJ<=@b03f+gV z80AY8A$rQ9&mCaAuCwh}htsOoJ((u47u&y7wnr13zeTf$`$mKEPc}L8z8guOlBFVQ z86u!Zqk5T}o4oSdMjw}!ZEml>mqk1^Jl;*4%}u&*%@K0CCaU3`HjOA?YWP>~$aYDX zki>)|Kf);uT5DPn9D&r=1#utljyNa40^&flsM=}|9-hi6ksva_1~|_y6d%l8pDs`B zLLqHD$md_=eDcU(a^Qmyw2?we2p0gaYEQqrgG+(h_JT4Qif^iVGQa1Xd93I#wR` zzryXIVqxqW&u70SWQJ-(>nu8y?0-F+_1~N7W&wFR;g$=k%b>K8jpe*K%04_@ehmDP z_GnB5I=E7M-+v2!LC8PUWG9WuOFPFcv`Rei$A>yYm7~4P5j&!T9lcAl)m?mrG8Q## za@JZEMh);6%PBIqTF*c)!xBwSO%|v>Cap*-rPQ$7KqA#2lv4FFDHhEx*;@Vl@^Gwf z0}JpJ*LZP^C4Ya-?IIG9-REAQvU9|Qqxx(1*tb?iXiO_~^}BJXllr0#=1llqBiX{) zpnU!p|4$Zx(kWJ8FBjK|hG7}eq(5xXYJo1*=V1^_@+e&hXjoa~VA$}-w1I1%5Ucoe zVP@|$5Iqs0h9Pl>ISc#qiLQ@f+gSp0-YyHA$@FWk0iC9pfO-k#_LFy@>Ylkk5IgdA zLwer)a2%bv&5Qo6vi!{97yUceF{=ECMb2w!a1|;4fm>#&YnQg0ySG%qMk zUD5owF#Mk8yq0wNq!pJ;<$#RH&q=d}IIUsD1*x16r+oc!7;K+nX3)^DQsw7w+>nIk zWqrBhu+mx)GekB+m3<%9N_s62Xooz!8L;|A6xo&0;pDm32f(8UL$F;22PfN*DOAT& zN?#D);Cea|KlghpKWcTSjB+;&SL21A%ErO677z5o%{ffjR=iNaXs9$L$5NZ>u6__-&(e5I077CIbyz< z3tjV}eV`0n<0_ut9~D&_*}wPWb%Cesa>P8TAJJWk+gOxQpT4<4OuGLHhuqWbK`99i zZ8^(>RhJ)be-aJ3TnQXk;7)jJ`Rf%pB&PawiNyw`#^M9XbMo+y5v4>!g`BTqRupzD zfQ_WLDpL$L?s`?fRKj}lM&xbM>-|=0#d88UrEzmE{hRT!FEztN%{Je1k*8v>+PIid z?=n?Zv7vvG;{hyY#>7kIp~{=@XvUe-LWjP^h6ii16+gf${_A!~^-`GPdyJ#0c0j4T zCL!DNX;Tt@Fs0j1YJK4!R)NwLk`j2I6+q4!=LVP8$b@g|<2XJ^5XuS!mPLzN?T~L@6-f zSCn!-erKxsr@`X2|Jb|!m(7s>NAqt2FiS{9jx(w-WObUpk!g}my^>n4pIVmp=JCry z`Dc;D$eC%YMBaj67mITn4-5+|AN|R9tX!K8ek#JTp03UDhBi8XEsn)xFekbz);fo4 zT3w!B?`D6>fgDrYv^FTJ@^3`MMfAhaWHolfhbx<3! zmzz7=FcdBmhMG6iXScg+&w%GO>LdcV86#R)Oc&sk*!=st9GdLeUf6pLa`GlcZIais z8U*zl7(?4kWEa;yUQGW<&AgZXWTxldl|eDTUX4&W*&|SBe$N1~JK&~MCqRUTgBj7hcLEvjB%jbB!0Mb@2aCjM0FOP@!P3H_=Z((Lsc2b$H4d|DFVKiV%}zh zAWu0l%mN!$+zt5vwkGlWDHZ6Do|F}kW!+m#1=~I zgdigL-uLHtKHs07=Xn1K*Kr;1_jz8|`8okYzlVz1fqsO6BjE+2;SE8N^a&(B^~!#U zTcMz4%;0_~N)Pt_Wq$nB(0h%9^Z0T98Irb&bC)rONejd207=0gY1_W zLmWKLCFMfh+sxmFwlmkCq)(tE{{8(Q^L+R5-yH1|{OnyBSWKXJM4$ zw5I#sPAD*Ei=W%@F!>z(c^uWK<$W)_9(CQf)OFz z1O_s@(r6_xRO8;bdzo4G7MOJ-mtNNVdL}CQXMHC%usq_QV4dxq#h?ZrV985!*udJ` z!>eDLf91R4PXtHuH=y;pw^?km4m}FN?f>i(ku^T{_mq1&v$VHJ2xCnk89B7r0CM8D zD|go0=VyLdfNLq|Hu2fd+u2A!6)|>@W7XJW6(brqs)muY2{QR;F;>a1GnsyVbV2#o z_m-*c@8s#1Y*WWxt{RW2E|Zx8`%mc92Xl?hCA0nanyNWdIG`&E5`H z47c*dxh{!={swjfSUT{(u8R8sYDjf1JLtF-UGxa0xoyD^mkwH+ zA-7RLiy?ro4~v$R~@Uun>4)k*02q~>P~0h6u7EjG+R zN2`XT4)X2mt&@yl*nwAe+qp8oj1Z6QZW9&OH=5_Pasu(!(}5_Mf{k$0jH>IU1C~st z_nlVedjQcDrt7tqVcq}Q&rNhoxMe3F;`dI`)ncgTvUr>Lq+=r_!^i!oY4x@%j`4c8eNwgS< zf*8`1M^UHdoWQd6kRi`0_d}Y{07X#RZ}WLkkKw>drd%{|N7~a_Z{w?b$#}W zufunB_IgJ73Gt!x9~-565xTl<(qr%H&rl>oPni-bZrSmb#J@409SbiEed;Yk6gZKl zr3O5r-iM!Rh07-iN`ghC>Z1P zjXOTY-+B^$mGCQ3e~fev7w%kr&?-PVvqqEB9kIJ@70NB{hc=wI__wU$`fXi(rS614 z2<9P|oN>yn?>K?*<8QQeEi{o^W88}JC+!%_&W}!x#GD>rzRA*W5C3c?>)&e%0U1j= zfrdS2TD1MIy*`&`{rI|f$Wl(A%R^bEP9WM%I&4_Ll;ADMSsyPqC~=2T0O5>ZxR^UW z+jslw^uWNtVo~g1itP@;g+@1mF^7JOWsIHH3`<<#IC2}RTgYsJo>>O)>huWM$!5>R zawmz*AYH(SZ~9%wzZ|O&g%SZZa;Q5C9@G0pl3-U?sEov`m5!hFm%-zRq@PK&|FF3e z1(gfEE}xaZBS1t28>+bR8`YzDx#+w|y=YPGB5QQD!!0T10|*FoPBIT6D3zkBr9toe zAQ9k}gyq6BePr#2*mX!E)3dJG2Oroo5E?f_hxzN)9Mj38)r+b{Kw1T~77l+1{vPF#_>_*VsE8z-I>dL-O$x^1SANNJELDsVxLB+FuJaT!NPQ9h+zFJlH7jq--^`}vi z04;v&_j?yJ0_b6|K?Hd@mU~WM8niWlRZu-9vTM67T?#<_7z{7y35jt+6?SXI{pnL^ z)jDiscSotkJnHqe1=@(|>ULc7b|n6L+Fg5B z+3k4zr5ue}cdlmoFOw=bnqrOH!QudNZYzm!KNtQ5r$Oe*v+S!NhdNZ0kJ}_Tiwwpz ze;r$uYO{y*lCOj{h|Flhsj6-@FV{ukuI2M{nDWE$O8UYirz5*eFh< z&u4U^ckkOC{bJDN;G=dA=U}AE!B0o)yt{1eKF#WRG*Q{h;Y)+YHEgBp$@aG_tF6T3de+JT( zEjqddCDOik`tI}xF5a6-AuBIPA_rXa9PIqevvMQrf(bLqxQjRDhFu`q>ApFHRUJ*G zy{uxIfT8y=EUxP=^mP|lG3ili8TPQ8dOf_NkJhhp4E2-SF;Nd4FFI8OeE@9Lswnz_>&TX-a}ouOAdDwi+YDTGcO3rQ ze*Y8Xu(akP8&Ro+zIw*Byxag;{-AQY>9w!jYNzXN+Ca^d-o>&13y$7ioC@dTV{raC zTYozKyeyEh0X*|x(`JQpY^0+1bMEY>;wdcybgc4u@X3}oD?XWW{`#|(B1{8w2n8r% zvL=Dn0qV04hr4+E9nj-{xG`S{vssY0vG|(qFp<2(GN&$ z?v{EWmR46oxZ-AD!Y9-A;q893&SC0cfp|Z{)GGab?<}cCM}!d;;k7-Ne``z|1!RTn zHO~BOD1S$dls5ejQf8hjfk6)4D~K1!a&nH6+n>{&_B*3TVAq;+RWh~(f-7yL+ztJ$3;$&XVWj@P+vR|G z#Fd)@Y9N-G3C&bjl!up|zGOz1^)tTQYj6)nC=k&&f0mO+i)v6z{Rob(V+RtHxT-kQ zU%L8C<3&xY1Z+DwJlA+YZsj7V39blLEm8|E`+2Z4 zG=cc@1MN-jN1!<(s+4BPvt>Ey#u~E^P3%fSPdLl8zk*1usl4v}I2-Zp?k@+kGz4%4 z>Nu959PBuXDK!&m(Cr*T7Kzt)reNmsSW`)g-#rN*deVxGW8#!UgixaYxwjV)W4nd)!bwI~PT9qGU)D==T_&dnk}9jNE0ep_ zTeqd}={rB|%;Z{&MpxO!dXHa6jzC%vz2z`3r)@SR`RT8*xlq5)<}(L``#xcVM2D?q z3;2<$5Kc`9SU5K;s@|KYgJzTt#^}(Cc_7ZJjShp3KQuR+lvO2}8Fn5)v0YMEC$4;@ zos07B0XB8tY*`n@$d~Pt<3Nk}`3p*@|UD&e?Ykv-G5vCgO77h3b{=r;-h}!+2VMl%LG**y5c2|C0 z9j@romWWWy17q}Y(wwUts{*5;v9l*rDPz^J{zL+ zmtNTYIU0vudz(Fdmb}MGH0!NlNLo`*L^qjCx9ilTSqtecJlNs@N3X@0;!pc@lnE!v zyhgjU{-$1O#nSBkN@r zW8_{b048_hJ#jmF0C#xLm&%qmX<0Wd_4QcEG zm7k@Hl1Cn#13C~E<21>Cm&eKlq#e5MbwBrd{hHc0_~geU`sbV$;xJ?aubSI^i#DRo z!o{NmA*dbtIfFF@%YU~)SaCkj$S3`O*;X<6N>Ui>hJNSSLmvrLB~$e-4eQ7`=5;S zEx?$n(}m0TsY103{TUyGF@}C2=nD?PZo6aSicPqSnNZi*aDUXOGt8}lI6LD;16va0 zP=6o?#f8rF6KCK_$5iWvw4NfpY6kF7Bhz;bg;hOBnb>06mmq0f?L6KOJ)-yRkzxQ# zV|kc;Yzv`C1dh+bQ2JbXD?TmF+1{X3`Gb{ugd?UJ?mbmHzlpfS?Jn#bn%c4qhWM40 zLYLKbezgZo^QpaQs})0@dA~j$)%L$)=S#}VT)PXE402C?XS^#Ba30AoJ%bMTKsTlj zCH*#;{jG9Mx#F7D2V#S?`r6g}KUnt_HR86dHKBx%_U5;Q(Tx{AkAJS!`^ z6D#>;Ue5UZkpD*0xVPSL)Ma$?uwh}h#902{i>c+Am|5KLw{FF-`XSjieTKdPf?gim##pUS0(HZpKl=|zsw$&FQ{08Rjzf`7% z3hREGQf;f+ZR^3rs#UZC?wK1}qPe{O$u@6TLH%C!9*xOX5iwT1S@E68P# zx@$g7O5=Vw)^Y0)dUv)sKGK(8n>9<~C_5ShS&cPquASnB_VbLX#Z9Qu`OBnUmy<(4 ztsB$Lnmb&)-EVkA1luI6D^GNnQ8^yfp5{O5Iuc=u@wU|XKAtoJ<6;Jp#mdabqeYn* z5$9Lt=slZZd{~RG&Z;uZj{>(<8`6Y371MtNbjcl+%1=DRDuC&44&S&&eXbz2JCNJvDifoT!1$H6IkYPjvQb#bSi=B> zFIW~N0;{S^CG}Qao$+FDQPMyuiM)z zszPAwQ=!6+Fy$(;U4aSuAzL5(^$ytMt2qbiOQ+6$(UEYuuHH}xV*ogx*s?n43ia%t zR~(6&pGP4cHw58e}c4Awt zKDH|~Ah?P^;QbU0#}>=flB2Y5jIbihxp;jgb77cJ)<+$1@jK_`$(q;OXOCvXre<*{ zMcQJ6(un07fvg}9)q;Q^G%OcY>83 zXis}FFvznp(+YC}1#R7N!f7xA zs1Ap7x*=T98s~u5R$s4qc~LW?+FlV z?$`&+3n8}3QNN-G(`uB~sk!$Ey;9Uj9V?23 zmY=PiG-ht9Ml_V)6&`Z9lQD6@JU(_PG+u{?bQaO|xK?`fed5^K5Kd`;Q8BT;STC9A z*oT+b>Ihe88?g#Tu2{>58o7>cWt4YXV`QLi5xhq~2XeYCeRQ+`DP);*=}xf0$Ft+p z7g8V+-$!DtRLDt!(Atz17lUYM^)~9sa!`ST3liF(ZDWSmwXc=sa%GUH55Q1%R zC;XwVS2g_Wh0>$*^SU%PbC!hg>_-K!>sf4D zzW!8)HF59(_MfBvzf}9FwhVMI1<$&`yM-^tVS*Nu@zyP$nzJ*7|Hx3g%NDkvf%`Xb zH!Vy(QK>yaC#+RQ_pD46?eb;J|GKY>GVK9 zT;Ty8+f{q|yO}IMBwF@JB0j|$S7QM|6Y2~!{3NSW%u)^arczDa=Vb|gcf6<4(hL9j z7VbqTZ=uf0M%Sc^`@DafI@QGypJKXgqIkayf{DJ(hn}Q38`khgk#a5EzD=B&p69t-HNbAnTb2b=Pg~PpKkPrEuLTgPg z5q}Lhw=KTsR!O~pY$N#J$ht+u60T26Fc&=M-(loZhSv`^TVbn70t%GZ?j3Kkw5T<2 zi{Xp|(!=$MuJFB1Ebo%`;S1kD**6*Kvj9V;sxDzs7Mfn2ly97YA$S#hR0wcWZo8eZ z>fP-G1v=d*DAdWGU0%PrL)><5C1_5LR9!Qbucn;_^V{#9RYsN@-x$gXMx?b2EbTj&V! zX6mJDI~)wSw9*L`2^b`4qbz9Q_^1 zeh`YScQ|jR3N@VPH}r`5ST|(7(;u~h zP3U#x&k0c@UL3A&cc(9NnKQiuy7po;gI3DDaNX#DR?h^t;yfRXX)UVDH$h1qDKGj+ zXE9lH99c~?~}H&>XsrNdwm5=H4xxsYGCFfZtB z&rla|YAXcn4`p3S?7B%ORx+F(M55bbnvg4t0p7zjHn+wwMb zt;edVEh7HRSZpD=rtbQ@kSAFY5^03oJdkR!X+*SxGrDf5G5U{|&u@~ysH!iE;vLWA zQt=!#6Bso@#8foSEO69=qu9hLAsEzOT~m?1{I4^zZrUSAKHA-7q_P>N*@WxH{=sBaSMq{Dq5XZ4Ya0n;F}5 zxYF=b>hB4ZQap1>aX+W8voz{*Ma75@IT)eQo z>Xudj#X%G*d9Te?E-l%^BP0hlEZu+cm6QxR8ylhFSnu;p*bf@;u8eTlB{qK%w$$S% z5fvj%PEH4U(JtoVH1)>_Fe0!B8B&)d`h&#^Ry!K@-*}}_Rt6&a=Uk2l zX|1D99lJ`&Yxg9aGI6?;^*1ek>hH%by~Fg;{NjX$ZeuSQ65qS+@6tzK`pUYk)TL}L z?p?fb5xA^->?bbzNk+8MD=xC;b?n$JIJr{I-SZ8vy4+0iprBs%-=u|MT*XK|S<|-N zWZ)+U!doy)OW36;sVt*-QV2{dFM^>MT+$lLm$3(1eCf+35c+LV8W)6_JkfS~W^=RcPURrU5ESRZ0Ixb$3~_ zWT-d-A>2x>wF{Nv`kTr8n~K27kuRDbBIIHCx&zUNn&A1ku^#FD9A)@O0=M?wq;C~` z=Wmf8yr0H*33pMTx*Y-@yw~}Ti-n!^=4+&`j{S&g6FXhY-LCfuF$z8Xzmg9PJ>BSR za{el(%*oV`#%&TcrQf9;Qcdq(sd?0+5}kkAyz;8LznL}8*6Vf^7zj!+<}BKkno7A% zDAW;7dCXZA9buek?0)lWVMvpviD=G_M~6dlZSl%8efg6gP$~RgD|a#_J1Hr{Z%k(b zdy4pVGLhYEI(Nf&3ca@4+qJ6+WKI}e7-VO?kbfTC-P$e8Z)aR(Tw|Q~QJ6awCYW)? z@=bYfiAZMd#t06Vcq3ocXXM^1&y&v}DKCI6QZ`aG3zdS;&u@D0@!JJ4{?xUosBX(< z4?QwBu&xddbJUB-mbwJ08rq0AMQ6Cp4j0IkAjb2Y^oZKk%O~#TUazM9m0qO)v;~t$ zrv+LUUb6Mxu8B^Jeq;PGfRc4TwOxWhR>;zuu~zhRzp>Z{+O# zv==0|tKS1!GV7oYNG<;`A00B`w>|7=a~3s6n1;AzT+RPWE^*R(qvwERYsjR;g0ywzTFAi_7dEg`-!}P%ZPxEpwioQQB4h z*t(ZPdN8grH4^DsuJDl2DEUJ=n~}!N{ZVWS(TA~i%X+!dcPr_8Fg2ha-maiUNglLD z6$YZ_je|YNn$Ku?2`yhDjJnaN@6sEoqaBx<q;eg*|;9#)%hq4if$~S>@$;Ssv8mj!&Bc+J$qo5N2#q3*|1#30q$G z89yEGP}8XjyCym$l6B=PF0L~T;Qf0D#66!(8%cNJHwSBk*D#q*%AlbMK14etfTar= zphqXD4dQ1{`=+adN9V3{#7dHH+WvwC54&ne^6S_(u18`ny4F4pW7nMJFOAil*4R?c zggTEYp|`l5-Ue@l|Bxwby3ylXi`EdmLj5Bg#Ih7M228gNm$@r$`nvAU5@Eq=wdErV z1t!%iJwj%qgyF1)(h7rsomKd85L4WmmUkZRV)07IhnRKSnIF!m#UWkcP{VGX*VhhG zM#t_vpP~WRMk(m+stDfnDwDtC>32Vel)mN7q z_+sbcYFns$i!G{ARWklx1QL6)<83MBa_s0)VwYZJjMEyP4ep!-w16l4xYu$pwK?E( zN7g{k)X7WH_us6A+ZmOLVBO!CvJRxCG{$s07SM>3y4lL|mCx5{)t(|OZL1}M`fn-BZT?n7lo!w9CT+SCFN&@o;X$wUt;|)EY<7108rF#NjNmr&R zP_3FkxQ(}M2vRTP{}Ja;pGr}{o~}7P6ljcFc$mz1B5}BK*I5#K-k9(;I@1|`$~Ac# zVd^3pC%q?GYW9)%j;lc0_?=>AsPRk5FHECHj-L|V5cMBbx{EzM^(_STS=!+}YehhW zfZ!}T{rJ_3f|1idCAytwoP{7OS+f&OFCRu4Z54XmM%-R{;bB}KuMZ3#uNhmCo~KHN zsn84u`2(c#ygL$|Cl|~*WIwAq)Lp(X?dGm<4WEX?ZPOb$OQbDob{2AsamUMsE;n|% zO5ccnz&tg~@vrna^+NA{!N$^Utfq~8)+>3A8r#b;f|yuQ)mk<-plkbc|?Eo&169uI}FBah==aYa+s%H0sDa1 zvC&x_&)!{ z@I4svz6E$^?f^?bw7-^4w^gta>vM5^t>a>*##;C~=~ftTRut#4FHC1=y{d9|IKABS zPyHql7vNF;b{>J#S_m)?Q_z6AP#ts=W*n`8jy5*JLg)K_mu0ATZZ~G8;TE}AnzAs2YSIWUr<-G!@XxZ+SvgZb!WPPre>?FH!6!w}i+`=1KC^g?e^wE(UTKEe6p+4v zS`a{N-gsqpTXliAi4vC0qwx>Gwc-sAqe2On7z@}Em`_4=b*N>X19nLQa;V#>MqnYb zG%H?&bhcsbZ43G-B>=LyRjs!te@aP6_-kBu9Fmq}`?hS0{dMtJS6#1IZUwtn{uWzP zxyjAG!fIVC%lq8u7{uDy-CG{Bt;xGUcup$u>C@^|lbZQ3+R?=!sDa9W#jI*OM@d&t zb!uk)ou_uPU0PJZKYjIV~Jv)_%p6`<9YKPGP8S=>1p zfqtumz}ieVZU6c08K;eRsCX-HZ?wl_S;UL{r5e|LafiAv!y~x#^R>NE?zH#f#6_A( zF!Zh{%gxAT?8Sc@nC>wbER4_^EfZNk^(M=|y$Y2EH|K|tol#?|kS7*k|1Ngd zJ*{j*BWdvgn#w$<=-)5C zNVaJi65H!3rz|FM^eOZ-k65|K8PVI%8*9bN|Bt*+o|dpqQ6`k0w!E^KdBrOA%}%pE z>9i4;)5dtq+N)Wm)nry8`MfG1Q;Gv*CYdUAif6P&h4g4pqH8WV^}7Fs!_mZOb{*3M zQ@ZJfnJD$eoY8r?3c1(%iNaD|F@xio$Y+zO)#TBx{*Jn?DQAcX`fj$FqvWisQ*xVx zS$h>G^S+7Y9IYJy=|7N4f=)JFD-g9Y&D%+YM^{GI)ZLoZ3%DtZeC8IK94Y1}`zBw` zM>V?oepJH3bcdX%8J_h!qgxT!Aw^)8rXvbBk11*go?^qNA6N3SWr3J=YPSK-RhN0i zjhl7mUMi8g2K-;u(A=)PR3=EU)vi1{t&*(7xmSH-O^%$&wK5SddSn(#eGHDJLrpwB z4Jly|=Q}uBp*%F_Ene|dKBkY^4E>{MZjC?ja}d&n7eQFyXk|k8Z};iP#3%Eb9Zodq z4ZBuOWIs>ixsm_k(NE{T@MO;9;Bprkxq&RC73H#WiCJy$r9@_tJ21rKOeH(CP!5IY zpNll>7`zh7uD|QR_FzV*Gehp+vr5Yw$R5$WY`h`rp3Y)Pb8#@iN$kxLg-b{ANz6&g zTozZ%CP?vcQvn*YCzY@-jq-< z;IsU4P0qY6n!nh!2rg9@<&I8({mHA`Wl1CV_VD4%dEN@UmGAoE9M?| zh!9o3SV{Gub%dV|#L6~#lng|I{L3x8zE&up<_*AGZ(~{f7)0)|ivaVRJY+SPMdF)HgF4!O z5+Mnb7429dRvLAe2cdB%5hk|mi%i3w4xWA530?y@nLBF%?7eM^vNvLdtvNn{-os$o z?_=`iF|zb%Yfy0SkN?5*Y_{p=yT4im8kW~4+ehevUyXI0_o9C0$slz6IYeBUMr3El zMM7v{O$$!e-M777T2u37)@kM-!1wnohZ@}L`BNdpPB1gZ;NgvqsD8Yu;C`vRb#+)U zl&l213tw=^YH93onTD>F#=<`0Y0*HRQgTLd%1G6s!)sVmk z4hn1SjGni!E83vXl>W%37(k za0=`eJY!9#`dRH41m+Bq{`?T{7jcvnp6xYS4x8DsswL+#+6DqeO z^g&RnJb4bRivW>V4|k&yEEgRBPH`29+67`!kQt-CNP7ZXC>*v~VXMSn zR+|4D5w1M*URskL>xyweKw-Cr9h41%F}h`rY~TNm^Z0f1q5I+iD$i<8CQt~&8+Hk zW&VpeRaY@hjg?AnbR~BaU0Zt?8$P;r!E58I$?i&ji~x%baC4b`W9ys`KsyKua28** zWt|;dyZyjuhS82C$GME!cBVB@f9~40nufg(pT5gKg#z+2e<|3#^Rqr4*zh+rX{20o ze+hKmH4Q1hi(x+O5j%SO2#O|!+L(d;_hlyR^$@AHZ9Ns>RnSWQoTgC2-gNM$S;z;c z0`U*|dh$^C%t~jsVAsGVIdq=%X8T+DF@M633 z$~Sh)Q~|=LBU7$y^yaPBuKOW6CYeJ@VqMBXQo&NCW<2cloi=64)ltc1K^P`^Zqu&*L6c1>6y~`<} zOp|4G^$GXFHu4)n$~MXJarG_Fg7p_qQ(ICiyxoU7I?^LWy{4=dZl2py_F=w920JF1 zVoqIrL#AjvnpL3e4}R&G%WO7cL9rWFXpl$>FOj*20K~_BTAv zwL7Y#GZwN+lG@+IY*2xLbXj@p$%ZEb<4@e8Vo$U_7g9s{r8cMHA)u}vNB|Ggcl_`0@j8< zri$(?MY{xrhQ;3sczz`c091kwiAZ6QigO>W`H}VIPf|@#h$yu*fH@kVO4jitYR*-bPqRzLCn5;iJAvDPG7FT4p=%ktu?ZyE#DwT7C0`*Q?&~%H}I%6d=iAyug-yQ)X`+?Gg{O;j<(}tE)9QMj5>g28%>I)i7n~HJB0x z+}Gv4Ph>{ft5ITqLdd><_cbtf{;$p*jRXn;MCd@xfakK%BEew$>uE79L1g9NunIjc zK6eTyru&wf%5Hh3lS$`g1RO1q#3AzmKk=myMm{@-1Ga2Yd^4Xz(6``GckAUXsuZIECe3s`Jn+~kwmMg8(Q#v7weDx*j zXWrH|-z2KZa3j91oeGSI)+Ju{u~X-Dr$5bHsyVJM!}T0X(XjyMJO-y7VRGY)obxDW7{5upePzCYY$EI-rmix)L9b?Gz;`MvDTiQcJ< zt=ZUptv4%py3zvEmSDz}BN@6zYJW^|Nu95oM zuR85*re$Wj)ta}y#SH?SIL9V+P3kW68S$oi>g-)Lxwh-$cil#+>+CRz4(6dQjr(69Db$k5Mc6Hk4} z*=*p%?~I=Kc+TrWsy{Hk_f|qzFYCACIXXPpZX9HaRmDp&GHzOTm4EU)f0tvOsQAJ~X5m|#!7(U#Iny?I z#`#Q(lsD)C;$?+U#W@A&YZAMV*?5<~*XOICzX*Bj-cX1JqWKgAy&NGbXY+Pw<)#af zw`reGgp8Z&jE~n(Tys_-L;*w{VtCY-W8^9<*qi+yKKFgE)B8N`;mk+gd(Kjj#%<^) z+`Dd|-9Epyp*ijQDS`T5ka0$Yw(J=9G$GZ?P+ZdxYhIz^y0^2urU`R0nI6~0mZ12( z=%Mm9v^R!+_5+}nt@mZ7Pe)?CfEEZ+5fB}%zM)RBn4$U@^4itx^mFC4y8?aysiztZWCjFwq=&bDj=SxG(Df8Ww>EQ{m4R`F(e|UoLZQW_J zp1Mi7lgbz&9K{69blfDH@7gE-)^7RIY>5@H%;VGxdNe4N7jJ}$c|rb)T>p8}>r3^C zpNq5s^JSwZ@~}6RFG@|~Y2-OD|EBEk3LmFb9}b0aqJd>>A-Rs09*2->qQ${Ix>ox@ z>&utm-sk6K`(qCV@D995<$)}+x#_uLhXiTCW@-m;yJR2&z`H2OsYABp-=GOGV-CU<>L!~l!PGlkDxMt)iN6Fw zc302%MG4UN9cs%&VFv_r%#HbTGo1oW!$yZ2INAqDVSEvE;Jvf$PH`zxUutJ6;v`(o zoD0)5@*aYKfW zaBf3BHyTTM8ZVnWeYTAHu(?t3bGYIYFF>mbOv^2+Fi?-Bg_~8XU;$DzZm`y@^_t$W zn&;Yky?K!hvM9KJgAAX8zF3yZ77oQ!(X{@UC})@+9R%sX{aT7rl7L;l}Pi^<{aj?|Y0wF=K*Lm6Emn z*m&+O6P)&MdsemQ7046Z3`0e<;*ISV{OB4FO}pj8!0^SKDxZL5K1f{WIN|+d4;CvB zZ~eD=nMM0&>MG(4OY_u3)NJcKE#p|~vR%=I1^MCKhPoGgiG%1kl=%?{(m2XqK1{1@ zZTPe3v!T^-f35$qkF#Jk73AH!4$n{AY?EpY;&{MT_3_8vUH@I5XRlYbGn4X1{i~m9 zpg$nKL>OtQUw5A>1BcY5l zlQX;faYc{wkVIcc^~Wcb*BnNsKiX;o^`SY{%A60zD)pbXO4;g6AX8YGHY$YLWtUUM zkZQ>eshaclmC2lD{5oiwm)4g-oY{iJPq$3{81`w)M?F{6)EBu}z@uto= zJMu}hdXl9?Yqt?FsvIkBeVKpFBQ$0BbZ-}QqDLlI?o!gQwGUL>gX3$WdBbxoQv54t zm4QouLm0c^g8pzA?=G1rihjjR2=VXb5l1Dba@Q9r#ImCz5=b)pJG#=fiYDULZe(dD|MmkE0bqt2kRB-Nt7&H zjnNt7%w!3mz3%_UqXEIcK!+^2ZLJu-d=pcOuTcybbk*qOZKAKwvbj}gYc1>LD0;_F z+jb@%jbKw7{CL=)_(jr(AUF3y&n78ltLi0Rih8d@ho|n(qH@7dY4l^_?+^C~ogrg+ z$O7})*dol><46yUUNiZI1D9u?0Lfh>jRCw+6o%C#p|q z3z=KY$2M(j1GhjxZN<4<)zN*}RAHIC<_H$RGfU=crJWLJ*lmlh7_)S2e|PN4U;cBK zREVKG72B(j%*i*v`CTb$-$%q*lk4j)&pIO>CpORWPWwf80ej1zThpDheXcg1+jDR+ zbbgT%Yg}Dv zl$xg6(&7-J;Wv1e6W1v?;ql5xP9g_Ho>s??YA=I#0xKav%B0z<}L; zY{e+UN$Zgjy1MrZk6rI-!=^W}wRU5qtc3(@>3+5|#ts+zS|2MN5aU|$z{Xb)J64dL zESk!8OhcH;!G}K_E(3Y-H*U_`27edML`)z%2jZ>9U+{`IrK)q~4EYhCd;JqFK#1UY zMy^w`5H**d|EFm};U$H?h;he!1MI|tSV5g-7jcVs$w}e|H2eGWzPm9zR?C+sq$BiT zrE`!)77o2K%o5<<>plAjnwOxI#N{o`59vwOAXy%+OypA@h2)?OC(ws$Y2KDE9%FM} z6b>#p$oSpjs22f|i=JdYSMcMq-t|N!ZAWt_u*N3u zPj}Q&I?Kw4Ih$Lue*FK{*xj_f0@nJk9udfz#2Eg<#>?)wLJRec1Vme}G2R8R2^M9=)zohw~=jR8nS z`o|JJvz?hBdtn(e97UYe)T!M+f#%!8>cz|gG*j~!7F$ey{}kh@n(h>MwdReLR_x=9 zBOA_*B42_DZM5$;mgkGWb)J#YF{s4`K;+im=qdF13Ck>=@N?G!dn!w%l_IU38%JS2 zF3Amhi?JiFxCc<@yW-j(#$7u2)vc>CUJR{vneH7NxwVYCq;nRkzWu1Dl_m=ak!od( zuY~ED_!<-L=6(6gHg!5)<+~()9TV^{#7I}Fi{E6rwZpBrVSjMdsj}&h0*pa!RU?+q zUxlb5EAI~(Wg2$wG&Fo%5g7RU)}o5el>m#T*?1u zYGaimTN3>6QQlUfUfL->lLs1B^Da1}M*8Dfr#o7FF9n~!j{Z?FZcjp{Cc@%tobGiv zjir_Fmyr!@Y6~PZbKUZLB`;&?U;Gh!hA(9=(@tXM-)UvM%^N(8=_jTbJ%~RGX3~&7 z-eRWIXat{FnkbC;wFtZ8*?%uQ$SFGZL4TG$CIL(eiR+1otdbZFuPRVzBu6QrZh@PU z5}Q~&c+-?hFDt|yY);!7ifdTk`dIQnu$hk^VndX{&ufcT#`wcJ&e~S~R=8PpN(T%{ zvfa-UgNeV@2eJCxU;SUYql@kTmD(3ud~a4boWCzUv0fG{ZN0#MJB^%0#WeiY%WN z;`iyE*#370+%3|mQOm=8f6T3!pnQMyi}-%UhbKk2OMC5Cd zF-oLQi0n@uF>eKEjuPv6^S^I2$j zmw5e<&?UIo#=R)4K9hHa3tOWxlcCGXL2s3?5|$&S04j^LUY+kCey)mZ@KxY#j$w{x zjwI)wwxzeRhO9N8R>wrf)FPWeM@;JeOEXdlXUxWu6M$gg7tSH;#|lHIe3~8Pu|6z< z@Bfh}J=Kg6s?VYj;8|`$6#ANOb-0mcLi0g#U@wAp0=bWXzXy%9acl+r~Mt(vWB_v zJ3Q*XU%2*H@gW0GNA?1>sLQVtWgXs@<@*tQ_ow^WpMPmrm>C%cW@qjdS8v%%_{pxO zx|PT{o(gV}vAx%EkNf09Xv)zl?Boy$m!{kmgX)L+?YudDqeF`_%b1UcX2vTmEy5^; zMUiqPQl6&P8P47Wm{REpV!2o5U_HH>8pd0@@STQ8SeSKxL~#!CaV-uxj=Pr;XSvln zb-TXvHu*bT59TD}GXH62@ITs8I+9%FK!jolVwg*n^SlaZmT*UUusIi3J zG@e+})oCvDd}QM;Djnk>u(Y@GJ>*iOV9w06(!B}N&6w8oR@DTAUm>0~yRacO4L=eo z6s#q(Yl2c;OiPS|t$*$)VOdnGSvp64WcMQ(-ygr&FCOfxAoO$f*DJ&U(IhHlxJ`_!W&5jT5+Vc1r&ZLS^L_U{_0`}X_YhQL#HiTy z7HLuEaigElZ&HW7-3(mPBVy~@xk)Tsgr)KsZeX!aV)9R4Y2%&^#LYpf>sHO3NQI1Z zKV)K`_Ujrwd68%D&^DG?X3KZO-H0HGXAHNEv31`|=(`R z@88;kb9ryB&$d~MPzE>v><3+EpNhNwFn*7tw?5L_vwEvh(_@PWd|JotTD=*<4X3HYtk5 zm#WRr%vsDn8O>4H-{KLvHP34XDhyZ!q3(+})>4t)NuK8E?N3+ug1vCPt1wWo#m2TG^vZx*G zTQRC&gNo6}92cyi-KcZfB@-0tyaps@uPi>~(qddg4I$L`gz)4++v8~YUseb+#$P^K z$(6b&=~_bt`VfcpaTK`5%AW*~1Gy+Vo-nZP9g%{$ML`pF#>U4ATZcU=<07-Ka^;X+ z92z$z;anqKETMk+**wh)k5CI@yu`>2${mCvF0437V$!Kwv2kJwu(HxhV*eatnXj?zO+x%8!DH3PI$aWb6Jn!>yOzS%h%_PnJuNij9 zNaXlWHpA-Uc1bjBrc2Lb0cMSJbo`ug^&^D~S}3n63Pq_3fGFIo?icGPNCzf0t};Gd z1&K;acUz-5g9~iwIy&_(vkUT@YdYG)meK_EvBQW=ep52_qd3EE?s;|V%z-QI7Zb<+ zQAQ3+>3fxEsdy=-b6FY&!cxBV2iyS^IV7_CaN(Dh-IkR70@?kJM1JND zdHFZsrFT057+E4g5r8o=)^(x^3lylZ~dj!@(4Ar~#P;@$ed-jk-JhSv8z z0n*ColE()kNG0N0Skm$(q-jF*B{92`t-<^wR5o*W)r$?YulNnP|J#Z;-=)2BUmCZ89@?u({FmmbmXEYw6kXA~ub8#Q0K#RQEhs3excZ7s_bl-e zGd7Uc`D#x9*hdk;Qgp@WI*o3$HVsU3BkqG7&LF7UQ8o2s z5ONY3l^2R7+Nv(bOfk+v7u-M*qV-q!0*PW`gD{)JoWKZ`RV^CAF>&@s{I?@7&0Fn+ z5QnG1=6_^H+YuhCIPYU=>AT5R{aOc8_2BHGPm$AO{Gkr)A}z&~q&~EM8nbR|AbZQs zdAMKmF1G-D$$2JoIfkb4?Pq=wB*0F3Aqp$KQn=(f9sR4fG}t1-tXUar2%i zuSU?Fm~5g&?G?b$pg_$?clMI_Yw9&;eOXL!nE6d{%DG@PO9}ag5P&v5JNV}k!m@=`*Rg45TN@U?`jUeWtnx!4i?w<3TCvJhSr-b(H*pkcg>MJaOd+}lE1=Df)}8GG zdIliMgGNi@x$1}idtoTl-Q^#+@lHf0?v9^j{1qfw;GM`NZM=4K)4YwU;F|?n>p}qrc)JqgtEA?J z3z`S?Y&>@#b-j-6YRdh$X#TRJ99uS#UR{WgsH;Y+du+Owzb{`SePEZfKwe4xU!VLeyr0 zHhvS66YqOu3}U0KQP>2P=NIee&liUtcH4q&z&~hJ6fcBqdFvnDY>~D2O$o@%H z@C+o+{tFa8H`gm1?ToM!``tZUw}&tX)fl9^Kc(jZ=cqU{pA#|q)7^SsPLnk2cqRRL>wmK5*~1oKR==kYrUx4jf%=n~T6lM;|%qV;KCx zhpP1$RDPk@tqN;_iGeo#c9#ni{1xNUwn;F*eA4i3RBl1T)SJ!CuBFV`q5Rrq{)spK zx9nGxnL9u~$6>BUe*E}V7!bs~PK@lQNnunu`GQRBzIjmB-t$b>mw6HthGd@4^BBY@ zNw*#lRgQ;;CDo6ebm*4=orc8>F`Kni>(+#XB^<7AOO8Q2S3XZL-pdWGZ<%5RF0f(0 z6+%P?DqIWUKqyI!uvTI8Kq-MaZ-87XW;&IbpVBcmAL&S=061v+e98Y;OP2QqDnh-( ztNIcydU61_Mzye=Yyw{~IE~7?!r-3IgiYGmJOrQhI|PEnU7K~qN)D{WrLA4yy9?|m zt{ZpK}QHZ`@y~j}cX11GeZOU^SU44s%oBa<)9*>ESiN~n05gT#v9#26`+e=S1 z(Q#w+yxuHjNNXP;f$?HA1<2rQi+}+;EQ_Z>i}{Sp3ix0dYvYaY#e`Rj*}_kSZwURf(KM^XW*%rn2qzx)rDz(@<4pt{G(NW6Uv5w$+;^7`cl~dy7HWNw%Ji3qk*^ zcWaI)MQG-l{tKHMG2xuHt&RI$+=7E@A;eUP>~V-XJ|6^R4BZZ?`cSByLy3hoeKr=k zG(H?t)^ZQUFEL`EkGyLABU!%goc^F6zs~sGnhWZ#WM`#OP4oI;KoE&+zj!lr@=tWr zv#TZUXKbWCSB05sZ<#;7d@kAD-(RN zFO#;?IhcK0`H}q}Ob`A8cd@(*c<>Qk{2kNz=;&?r_Do2MVd)D$g6Jg(>GF|x@0(Ho zv;<`7vElj-r!TSb%~P+Jl&Q9Mm-k&Cx_)xv@pQyLuo*ank*bil{T2`shL3tab_xCb z`;IsKfB$U*jAez;K;yq!B5^RiSvem|25x>uP9RpX7W@`$4DO?Ko=A3qBeLgicYVlN z3`aRc9nsuVx+R4}`MIcxyi=0v*;7(wP`|oqltCN9^9*>ax!om;{VA6%>GNDH6I=$~ zoRP<6{1o{$xT7ogiMou47^xUP1w`k#MXd5@YR+Rbx8?=>kWJbMX-8cDn+XIn4O z&o4~fZQ)WdybQif#OR^Yf^KGfclB^&31#;Doju;Kij}aMj?}1KIvFhvxg7`((aeiV z>V59p_L_^!sWhO;#gI$LlCIb^X*#qXEA;j5 z>X)`y2m6qs`$*@CojSK@fEls@hY#_s-gA!^&)+pm&qfGX>hSTDTaEF$Ll^9Hids#z zP{ZP&N7R4#{Q#GDJ_eWfGE*HRs+_$)imt^$CfJv`c83(YR- zB6%L(-O+E4DeU^Ct^n3{umsU=9bZe&XZ5w_!_x8-_aJN)F|4q8E?%y$eF2WZiPSC@ z5BA=zUM_uxy#%lNFu*$Y1!OM)yTB>jxvtS8TiY^R{aw`ZgZJ)(faZRV-|g2H^Mtdw z7v@89C89Rzb`lGxh1=|c+9KWP_7bG*1j?VQHJ;$h6K0(f1k&){sBxhQv#+S2b?^CJ z@y4fyv=8Wckm%BD7X|8R{UDbP(I;W~!dz@i#fK9B;f3a7RHiwmCo>X0!L3W|L_*M&(EQw1LOs9@G1`GAYf*gY$gr%^deO zE#(WV2}MU+&taUdh6Kzp@v{<8E<=4s2)zLpL-L8bpP8thUxcs?M?o*{T%UmcIIX<1 zF{z#XIw{6S8>(3dG>LKB8rH@|_b6USBc`0x-6$@3kdkfp3wpUCQlXA)lCt@ydJ{T9 zIW-yQM)RLDm@?p<71_EzG59;<@o^vT;5VS&cFMk_5tQ!y;f3fmCa*^Kr>S}BK22?3urb(w>qtut2lVWd(i zfb7V@ibic+3TqmRt?0yX73z-0Sb{dH3G}v)eg&fnuUaRfu42^EJrk2PXlHt;T(;$H z!HDW1W%N5E0;#uD0pK&18}R*y25_!_6pEK3=lL&b zw72zIT_t6}LPR~n6Zfuf?VWqvB-NgK2_AAc%Wx%tq+afu1f9vZouzl_FM4a(ikv3< zIswT7t#MlTy8f+kkO)y9?Wdi1-WXJ9WfRbT|EvUY)N`{ zljk(~ah5iX;Vb4!pzlfU{%Pw1^-=;*5ZmQbnIasXELKC(gNowBj&IvXtC%p;NyrvgcOd$Iv&FDGC%{AFkkq4>5;~*)_QHo z1Zgqe@PcZsu1eOJd~KFB5KxX9^ZWl!+QBOGeA43ylF2-=;zgM9AFcQ)6Mbvn@t2V6 zfmCY{mRPH2H(T*TL`=tWp{vTU-g;3C{1c4Za_cWPJZ;ox=yPc7W3|zgegR8+ofw*9 z#F6%dPn_>6dZF8768y3cU+6Y;;|Ik4S^w3bcvf&CX~$cT%mn9btpVko+?4xoW703U zr7c}JbYQcv-#}gngn)|!bX%PS@arw_8N8bB9zS`N%LiV+9ecY=BE9_(!Zsg^;A&gC zHCS=y-wQ{sn?-ABNHBqIVx|4oQ~T{7bAE$dH1ctO znlVB%6U0Y1cDU5T;!Zq zVmAh<>^ueV(?1~%R(Nt~WegOA488_yFxAQS33kP=zv`=6WRQ0g`lzfxK%M8U2T#R- zv~N1J`DWk!aPTBmw~k%Oj~j3>+AgN?ctBj@_%QsmhUQVD9c|}nXl1`mc{_an8*Fy) z8QBoNX@wd9fSU91aV$Pu3S)6bB$}b=UjFLmZMtN9%Q4DtPFn<~^dp;H+I$%Fe7%Lm zWBs}-?gjd%Wv&@YhP+e$TV3(@yW1gCtvZaKIw$|L?TjpTb)7WY7w}k(a>@8oa&xwK zU#6ji&|EeF;yO`1adhItQj70jO+c~p`C7s>d0TrJ%5?L4))8$Evze9^uTR~|4x;@S z$l7*o?upic0n&fXC+Tw)yJEAmjVDAUah~qE)tRyrB_*EM`uUDXl>faFdErFjV;6EV z?%mr-jo*$_LD~w>=dZ`{ErekY%Fq17dqx-4S8TfL;`L2WX8DTwr4Q@%><@;*L!NJj zca7oqD`Hu)-nB;F2v|~46faW_^_tbWNy7!6l^##&$4&lo(Wq~>!YW2%x8v~kjT4Wn z(L+KaQrPCkjKkxjZFtt(c7kxdAcz+2`2e|McCbB*rH6@6o-dSjvCNF9QT-=vx7M$n zL|xPnDm8Poz}?Gi(mx=xAcDqu43U2(G7BvCb3_vgffA)}Qu?>9Q%;xR{fmR9F_(%C z#WHqGDt|V;n{o&*6KwTJfDzdCM;D-z_9p8??i6v&phv7LT1%}i73YEF%3Bvt5@Z&b zW&Kc%MLE#Cf&)Irg!D)x&aT zMQRJlye<8=rLrl7@>v1bwCmq_KdP`hE9hor_SnNKsqY zxkziF4f}X_2!inu_(Ff5n)VCQKyK+sG~3Y(`A>4Y>aRNdmJm}e9*!|DP7Ij8oIt9$ z`)XWpk@yZL3{(t{3~c|OftgsO>b#{LLwl62Fd@75=(Kg}=zQp8%Ngrk{(#!qd&10U zn0lNW@~UO72ez9@N|?zSeb)IUXZiV9*6w!E4}ivg%|skjxP5uk3axiU|Du1h>w)HG z7l$V5`&>A%IjfKC&TTp{0I(p2ZMR>>+TVj{(vkp(4$~qI7FugscHCp?dYN02Po6$> zZRTgZPik+T(C@y9ymrXoYPt_d4)Jda!R1y*({cO{lFg$2mx}l7{>%k+w*j?#U;10Z zB&%Bsa(J*DF-rm??4F!@oMl1Nwy?TN8~<(ahwh29NXZ}B?@Oz%dcOCmmDh%xdYFnz zm;D^jV&&yB@PYCCiqFTUhc-s$AM)gqNDPiQ{&Q!CFzs^LI6F zn!D~0b%_pdI8hi+2bM!Q4ifMLoQ68;Ge0-b2&L7vI+!l(Tg{OP7W0?X_p_f_tpY+HNa#In z&scyS7DcBneU4%;cfV>LhRfKv+(5KDw_E>0+k?&=YdpS9+5h`BN;Liq#L8XrCqQ~{ z&Pc^|PQS5ahyT4gd1xjgQtxJvU;n*N?SM42VBkiKRY`L(tB5)U@uo4HpS*og+7l_5MgLW)B`)wLL z!y&?J_EnSU@;Mf1$S?Vs=`rb}t3V>AXRxeY#LJTG_YLmIja*MQs)@GUnK#qAb(p9J zs@)agiGSVaBLRK)pfTXObgcieEzGGUjdBUCJ~)F`=SeGRlL)fb$G$?A?CKv-j2n^B z3?G(YRAF)@)GiO2y(XPay@cMoSiL^x@Aan}-Jg&52q4BUW12!4<>L3@iowet%)kn+mEvb4^o|zZd){Z9(@!vI zJYjCbCKAesV>%8FS1tEAa#i1Qzz#wa#>1*Y8CHccGRW+Oa4`h1Z|R_MpJZT3F-7|o zg>VH!#t*@h0s>ugCnn7p_;9{XPjuCWV*UZ+3MOCUfJ7ZXZL7b3HK}Q?IiNe)xa~ZS z6$sJIYL^Ci64-#j-^)OhK7R;*FM*9#rZoOurNo-CIO%(^mZTMsrl(b3qU=lJz;B)f z^GB_T&8Mb*bu#zB_6p&&=q=Y&UN zBi8O5w#cVNi|(Onr;oNzu5?J((#)W8pjMUo1uQBFKit^gS8_!rdmil4YK)gKF3Bet z`4%Q;NARjG|8@X}I-8)LF)WXsHR4hGIM z(}ZU`Wy6GlTaqu{9K=Tc-Y`R}o9Q1f%ZA_(YOID!L|2XRtd@HtQMo+As^e`D z?D!E!6_lQ}$C({vFYYB_^am z*xFHx=R*IAf>&;!1`z6V0H0Xt+P!yg1;f+HU%1?t=1EyeF`)^G5TV1b8Tr4yjoAPKzLbNPwEACRe5P&c>-vKf@>B>?Q2OMx{<6vYiH zjwJgH-nE6je%3i0(9E$P_%{6?BjC&c%jC6thlVN-D;EsK&Vo!3y=$w96n%}op&$Jf zCIFHBFAe?39`t@9u$;(k`P~{d+Tk~IuKLFRT^jE0d>pt=5|&8{^P7zSq7x=i-8Ff~ zXzTRxW~CmABOyinztWKpW6hiOK&MOd2R!+k?5$>;nZ=B{5K zn#pZ?^yJmtdqp>e2OoVv<_2CyS0vwa*o}b_(0=`=2R|=m74UhLMJ*M0Yvxpa;0cev zdIee{;VC&u$r8)>u8s-2&1vVj{<-?$MI*?lzVve>R%G&GH@w_29~`nc$SYJ|las_LGi%b1 zfWE}Jsqlx8-LAeUbq|MFbOE_gi;LT#Cr3+z-@-W4W54T7r@rttGVfY1Q)NS>xtAN- zOe=H#Rrw{$+dTZn$ohf$<*nG;eIyKZvHA38d?T%0!C0u%@#=k3kTRH(U0y-T5z~yb zp0V))=hJDP8G^z25nJT>P@d;V_rbIBkWs^y`MUs;`FI&EEV-nW>sZc@I-HHvMb}nb z%MoF~JCn9s3jfcHcoo~`Bm!dd7Wrxt$YBF5IGn0t1<~<+VY$#U3 zF%3~c@#m0#ubw)}G1w^+s{ig`IFMfTmD~m1)8ntt>;Bfc@VVVRz`{IS{Kyo`Yu6bw zg^=EC)mn^K&Js+cWgVke9bcEMEfkD(E_W(S4QKx=;oF&bI~tSPP|D}`+tsa^4NkA1 zwDWg{&5qdwyW5mCw}3mVS3qfTAZBKO%(9AIMrF=W%T?*IaPR z=i57_Z2MxMpsD6-Hdag-gVYz0aYv-ILpeAOINpSn<;`eq8QxyVZHEbN@XhpXs^iNa zXVwwS*2Y#;GUbABk-n5L;j@{gm#~!OL>0}{aD_SRV6-LjNr%6BxS3{72Z6=|8u(~C zkW#s3g$AgvA8^I47dV(ke;8mEl;&c*e(yh9s~+go%8#mIT{uY_Qq_n z^(AHp^?F+@>>2SQM{e)XqJUd5g~&XDgs13K|fbQkKr3}W`kG`#nS z1H9(-MkA#AIzEzAWf<=p0yjPzlLoFT)hDWLCL|;#zIFgHx&;kPJsw{M#Hj@IDU^1N z_8+R+4iw;LgSQ*5!(b6YM0-7b+O6ZG7pinzu8zkRPKHXUwQT{b0jJGkP<=73n^lM9 zjcG!SeY_c|$xyo$4Xb%qoVD-*j1f0I#D!|-^Cb!$nP3}Qyi^!%2+#f+b3H$F^TXLb zUM15_=88(A8uO3fN;9}J+{`Z###}eCExO%@<2w*`d#sQAfpv&fx0aC65Rc4Df|3*E zp@Rc3n1JK>kNr45x;ASF??JgT?^`j404dA)MyQ)~B~2`+S3H!lCBE(RqUO`(q#Ay2P%_wP`w)-JE9FSMY z-7M=*TwNL!(&qXYqGaSeU{>jgpKP!e3^iFIq*-_5j?5vIl#%NyXcK`|5I}aZ$ zEjc-yg?mc{bUi@D{64%cd|BbIIjdWBN-HrOf+TuGiTG4kPU!in@A8?ymE*-y;HLUf z|K4gV=Ih`tZ?8yuaj1Xwrhn6+*C{V}eKn7}S6NCo)O+!j6vdH)W}EKh3Nm-+i)eKf;h{VkK)10P8Y)h`56D(bgEe*0vKVrovXl7VfS? zerQ*(rF&U8`NQ{90mwScBQY><_=NqtE5%n7v{T>SeWEhz!6un9(>?2 zaadY;RV^ZasRMH}JEv4o@)LdCtWiA;&G$I+t4_q%A~wO9b&G)TI?3s#HT4E_ZpD(J zaAlF(k6O*1Y<*@iLnC6?Od`NaA2(Btu)i!~Y{ORULPE^>xKTw9t?r-T^!%sWNAd_wv> zF1epyOtXdX%aD_bIDb5-2;xsoyM%SkOyDf(`WfY+n(v%qZ6#5KH@af<%UXHx-q;FT z;F^G~)6K3g%4v_w+i!8Tl!WQfK7Yx}Z&>)1MW>A*uk!;ZqX>k7uE z6QFxc;2mbo$KD_MClJ<7gvqe~7)Ae{x+80sKCGxrh-;M8Dx$zk@Bc_jKPE-Po5Jg< zz`~hZQU&c?i!4190Bj%rN|3A)uS>$Q20algo**YPF1Y)_3+Ufi<4yTxy7T~BYgIZ} zr{Uy83ubEz2OxxU`XmM@W?aK$lTZ&Z2z7ln&>^!q5FwB-Z|B|Xt4RqgMXYXK5HXJy zqZc*b)$VpktGXn|`x5Tw=isT&ocKPg8ZUYj4s4XdhKx$-YmV{H#`z%yI8a<3ba5cM z^|P_;ubhsVfxpG-q=`&t0cvxyEVX&S7xy^WXDFs}KK9!YWmTiG?2)V}xrTbGk<1H2 z=#~hYPHD)Kduh zkn5tQHIzaq*WX3lqSCVDgt-p1Xp9}zo>;8JscJE(HPWURtq+ktnqUme!XGGEDu zEKY;tLQ2idV`DdGz^Aes{&h+>G;0gIj$lm~muBuN;j(_?c=kK}yEe$*F4ed3(9Z^A z=L-jvtf^?5`Z+B1sinug)K-u)b?sc^A-*~een1TOd?u^FOHonk=Fn_d#7s)wP-AlFj?AWvWr{@e$TvvlvW%kB?u1)6kXAVgV=td>1EJ;knh9fV2^GW1j zst!2R(J7$~<8Lz%o3|C7j39QY1}YW9>A?X9M5a%(%eu5hg>S=xy5-k{qSi3c!B*uO znyMy}EHq4B#Rn}15`VkV2K(-DagU5;iV3<&L3AraG+g|9GTyBxF8nA#l7UcfHe|SN zD5P4?Hwr>!nM6&Ssg6dx)mS=oufxSU=BwbJP}+M6r;(p?kBn)lP^CV8K~`i>X#aIs zp!s-!a?y8vaKNN%nf)pmt>W-axq%GWYZ$#j2)d@DflkxQwVYjsI-Oq~90|p-_OGk$ z2&t=rFku8u$4tRwUXe1uezWlsX~_jnJS}S{dlKAt0AYrnn8#G#L7z}s*8UcwN> zlwNnSM3aH_KMF$r4ezv56@cJm&^6kq1G{xEf(-e0o{Darb zMP#0YZ;$(_kq3Xf2GRDUhR=8EPBw%-LoFEaf0*r(5SpLD%OF1 zUx96zyyTInWR>A6>Es;;(0sEiR26q6^~%FXyf!{p&fULXJ*c6ePSN|DZ;nkZK5_V} zGSEbLCuEz2_# z-$_ECha+leMAu!ycnK>}bb*JzjYB9nd_`{W{w|}$Y z5uT8hKOPI5SbmTX@BsCTU2gMzGm8DifolkxMq9;JFW zq3%{q!z~$8PLgTt4K=4e22gG(YLF#wiVTS{TMSNojP+>;BJQ_HvcX?5Vnp0I)$?zR}`pJZ^Rgn`i}xJE@hDCcBf&>6D=)*S3#KD1t}e>5{~Bk zb_qss55I|o8#n{bh6|ZPSx0oqc=NXKd}srD9cukm8Y7r`T(?>Wog32fFbpU=^&gjH zH6P-!SfX9c{~|}UbCclu_yoejhjNpa8N0Xdoj{A0m~OF5VE2IJW%b&5Q9>t%DsI-7 zyl|Y!uv8qL>tfoe%#}IfHt2v^8F!I0ZMHQt)gBUbY z5U==7*a`RPe@bPhF@#Ca)MrRu!*?$Here`-2g?X!WvlEn1`RKLl{r~FXG2KD4{00@ zz9$I*N!cpRl?fcvpH?*ua}>&npv>Mb#>Iwzgl{)PS>c+&uVciT5P%4u_~=?>Qw5u& z`a#;@3Ut<9-^G2d)q}q{ew0{k&3k;m=v>(X@gw^Nx5aZA-kztFC=@;0w z$&4F9M;EXhI*t~yrw~f)=ffn z4u6>D5(_%ml~Aj43e@SHyh^0$%(UC_p!P{B5SJ2E?17LuVby*dy&BS@d z2-?MZ@%2JC@w_@AE|~FMwROHkQbN(M>Zg$BSb0H=LT1vy=~T_j+Ac1?ST4Cy+(upo!_8?23+jXqR>| zGQ+*PbZnF5l@k?mJ=+ehPrj%BocV&0Q#@54Cstol74U?^IT9R6K+D;VZ;N^hJQ!VO?v0YdD#@=rY^a6%v(uj zEz{)ht{Nj*(?*Hjb6XZ?xY%Ud*kPx$PV2$JmgeS9qm0JDr)KN?rT;LUHLXWX*%~AM zoAr?BF203`Yb@`us}~dlPygak4F7QDJ16de$%*}oR4dk7XmeW4>W$Yo$oxK&zXtB* z#0%`PWkew0_c!{l-nk;Zs2`?~^G%tr^mrLeqYA-oqi0R=;-|tl-->*cEnPzXJW{DE zb>=eKmw9w6dt6FGJ|KbLGJPt2YIT;3lb206zxb9}=b>bUMt7pJ)gget#cG>^>Ya9V ze;UbyIF@wd86`IV;<@P(81~A2cWu&<`X}KV+WBk_mDwDdCLr z0N^KxY@T=K6IFujLkZ`t2V#x9gL?P7By$UskXCg0+*oNxU0Wnq|7Ok?;UvZPbC*Rdm>z`^CBM^rkS{QzmYHgbUQGyf<4y{CKt{|_ z!cc@+B7)9~^0-w_v9QH2)@0~pUK28t@^z|?iM)A$N=L@wy6U!S-To9MFNovQjmIDE ztAo({fNf-MIcV{8Bl&J4^nICa0G;UUI5sY@D|q5_*X*@tjs$dJc%Z=X=1vEsUij>e zxcloA9%~qS0wff$KNr?}7%r^x|7QW%I5ydffOa7lEEvG2R3xMI?vOx_d-Ymt^QAyL z_k`gCkVm}|nmu&>QDhb=UQ10<`ETb6&+Ywq6~CpI!zm|AtL1^zL48CpK5WycEB!Kn z;Z?bOBGxj_3PQ%bntKEAqI^YpEJiRpm&8FqMO|u9%(zfeGLJgdr2~2D6|auHPK-h$ zV`l`wNJQW};HM@Y?YVGMt zz^2Dy4PQ!kY#D3bKT9&%%Ex_Uf z*m?;c#QJU;_n$SzEYq3W(}NZC^)<_2X8MTo-w%r^?)mPSI4Nix9DxaVtPi@C_^kd< zuRzHYeM9l+`LPCePJ`;pg$|Gt#QWq6vt*;0b3L%R^Tnj1dcy9>&P&aAjWq!ICEw%( z!q=G3R*5etg1_E(qf`&uS$z6|qS2!5nS^U7;(J;rZF!M+PE(2rA&?iKR+80o9Ti`M zLq|Pl5eFJeD*`m6Cx+W3ge4Spi*d-OBy_yV_16E;>v+s^VTNDAY(Myy0Fz>*m`db5 zBE!Ndk@D7Q)1$Pe_?XLw5W$jNEOQg#&AYOjaD7oeSP__)FOn^+9=~g!kX>`0KoyUV zC)-V3_rCHUaY&=czPgH~=!>nW?FeF#12n&WL`L!HmllR$!y!GU%~W~ zmhvyNl4^l&3QoTxoI0V zQo6Be@C~r%ZJV{CHsdd&A8Yl~J03K6ITPxUnz_jUGtFxFHu9R~k7>W@bJbaAkbsCP zzFwKTG6+9WPg6udw%-ZXx2s)y(bxCaRyX*$vAMh_VN^w^GvfM>EB%D1J9oma2tInH ze3$Ihs&l7ru+GCv!1Z${`}GU-2zgCU`uwb~DgD=sz`PZ+$abVCC%+I9uxmYVb*;NC zM#0MfTDW+0`Oyp8#v+;MvU!?GO_$N5C8P7<@53y}#V7wO%i$>Dm!{?L9SjZ^|3a&_ zMINnxzR#XV1<(1xudLPC8Po)yHujN?VQvc4mi;WtY-qxJk+^`D5k zb$=rlGq3!Oxc&?DQ6R{9>-Phq#q#=ZXSceo;(sLLtck4A^k6s4hvW|-2mR{>L39n{ z!J;HDS|Qg)vE$h^f*|)?8va#uEG74MPKr^Pt^1N|F5h2t5o~{Kb@RgSb&w3f(C;Pg zZv-JDRW3R~>KOhBXBHfzT0m?cs~8m0gw?eYy6?pp1Wv7L9b$r^;R<|1iFjB55uxa849DuhZQ)_Qzoh#tzcoXSisHQ zSo#=mF)xs8r=u%`9@5C1b14fi_D_B#!nRQS1}Zy_?Y$rW0wf!dLjHejz56>8{vSV_ z<7ik>4h499<0k>>4(pF44sj&fB&0^Wfyrh6exRofQKCqBt4mjh~M|Yh5dne|M{2 zYXDNz-mj7#QM*U<#f}o6ZB=M7-z#plzSZ*>OxIC!S`}o{iur=p1p5T8x`KzGuM^Do ziXxZ>V26WK|GZ=-n)?&|W|5!+y&!+3{4_IKK6HE1_;lXD%SJzDZ_sY4e)B`;;~id4 zR=HDsM-i;w?imTlWSq}T&~l0X@qY2nYm4Rh1lO-kW3_X(x`$t*1;EaxOZ{ihr5z2Y z4i8CsZ@f!&I82cEr_f`oh516{nnEfw7LAm8$mzw=WHQML<>g^-AB4pD#n&! z;7*TyIWFBrTvxLSxzgxzNHZ`O4`RJf7NC&RY7RpRS`V!q!veS15OCke+w}kKNLStu zXA3*x^L9OtRn-0CIond=^e)L4RcI}-rGGG=?q6JpNRIeBKK)23sk+9#OI|^~LDV5% zavrI9Bws)`32)3nSaX}=!erMIK8AB=ZzL*g^F(aEat{-oUQpVqI7?uvxB&+O+ri?0 z1SJ!BVIz7@6=(RXjixa_5^eo9V&O;2hzOlk%vP7YC5ROJtWU1u_|I*XMy!3aAVM%~ zB@;1kGHjABE`V5`x(}wt+4p1lAG*8(l~c@B*Z3j?zzOz%4(3nc^u)t@lYAP@{z<}t zc!a&8|Br;Yq%I&t0}n*wPy0tkOJb!1X61L0J)f2;=m@|@cCgzXw@HYbV!-UBipPOR ztVMXA?-K{0(QeNsYFeiQ_h%}o#^IwS#ReISVsHB6(f=Ss_qT}-RS08+Xee13YS0Mm zG!r=Dus}_*XroO>fm;iPD%&WSf$sXiu$69r-phI2rvFe38WDu3+v(B*69wQ5@o-Sy zQU1lZils{h5vi41!3lTW7?~8Kta8wn^+MxjfLpkqauY)Z^LuvL9^ydWr?_NpE zPp7pwpPr3eH=tRao)~jH$Eoena#6m)qCAIydJhv&GiR*;c`7r^nK@NlrChlDOMgK#5PEbNZ#dxG>agfn_yY)d4KYo*{KENDCgO=X9 z4r}suZOr!S6RfA_(pXVy`!iN$ANjnz{wy&tzvu5?Js0(P`fww7T05cO?sGSeGm=M} zCt+i?HzLd#!b{X!+2Kj6|2e9eXs6j77VF*l+ob^;bSGLkeOB zzdTpXePoIL0uPJfZ)gD?nbJ4D=56@|XFMq_{*b&fYxzL57NWfGxd>|?w$3t%(d~{nKB&lQ2X|Qz^or6_f_yl> z7r0Bl3ABs1jK2+VLbcnhN%g#UXquWa0C~X@^2^k}@(Bp%2sc727}HZ{=(&(!niF-} zIajq6l8>BS{LTZ^C$|^`EqVC2N}oCwi2h&SeG5fAb6gG`8~gmnl##&7j0Q89%UAXO zvU1P+ru;!!xTy>34g1G!vjhcc;scZJ77VX2d7RwxHt;`6E@VIXf_d(XaU6 z;Hu8^^bntYim}$)yq=GyyGp|c2DrYG9fKOE`b?rKA-#hnbo|uCI3mD(h*L#P4b&gs ztG)B%Ai9zNQ8mC43@$;Ba3`|PO_c+fgIIHk%e!$Rj5U(p+U=V8_nw%RNfK~M;%L^? z7Ha?n=u2!M2U{JdQDaU*lz~xs6mqHH9lzYR_778ZnYU5NQ&?&!w2wb)q51;(Uj=7Qy2 zhw~=qER9Vxf z{4A=Fy$&ttj^0Gr{^Zc!4?}KC?4E!>f$L{STBd&CY(;UkbFS7Y~LKsS2hp%rzfzH)Wfxlq3t4v&$g5E7vKjy zWoB01+so1V;0A#k!%l@zI}i~{5%~cImVL1FkH*hPorQNheNd(L>StBxH>P-L{YDzq ze4Cn(5z0Q$sID6OrX}d~(i-bshO}HNzoM;;`sm*%s><|V>XxtIpt3`!&bjspyrQyM zEd=7`*6vid^vZuT9mV8-+hn?JPQCJ#h3gTPdmr_5EC6);Qoc57^>y+fTx=e?eGuBx z`s5y?$Sk6FSX@A-jH5Bfhx?142*guCOW#Gc4_F%wo4@e#R>;yRIcIO$AF5O5W`9f~ z+NF12(%4xhpjF^D>EyUZ5W^Gj6r=pNz!cg;V*D3ACR!w~Aqqk_j=b=$ewo*1OfG2b zY=wGH)T)Cr59E(pweOqQlV=sCFnh87ar$rF|2JRIS34vs{Ee>VoHFE#mgYX4&Kv5t)AlOkgUCAhyd$wD6$EGLOS*>roIeA!QGFTHgI!RnZ1eZnHl-N?`3badD7&6@ z8pSS7WUE~L5bw_*wlR>=&NoM$mE16R^7!m|)cmBVF7v%mCGGA#NW`3q{~0}g5%m^{)(u>w8*O>O^nm~v~i!72D~Aha!OZC64@r%*)33sP5jH^dselZ9p0~e z1xifj|LSS>#JlHVX07V4LR(+xB;FU#1s;SS))(=q?5yc#&s=-zlC8QGfByHA@N2_* zFWp%~LOMRF6#$#e8R-`3hJg)oP{y*gEv5{#&w2%r?8%>B^zPyk!ABmvvG;mxg#3pY z4e2ro9tSxuFT#`tJJ$)hU_%2oj@E9@2XfE$Wn1p#hgqAPka|fMl8G{5bL?`q2Djj( zI*g-cjMXld`8d+%D?*~gTJogspOXFjpW_|h9vtlC0;ew85TS3KZN{atax+6Qz0$Hs zNz~dpzq_-vL_BO?w(@u1#tqfH37PzQK_vf=XeX?<01IY#_AI&D$#{nQ5kBkzWn&87OO8jqZEYseV37Ry1-9tyC^p8sx6& z6>h1~2;-4j9&$FdCzFUQwWAs3HCpVZT7dZiW2A&RACp`@7W6%D_3i7-{?y;09c~?g z!z1qd-=^=^CCGo-gi)<^A3O_uJ%F#Dx6VVA_)H8go<@SpeK}-~jH2w1BG+&2$bQL$ zC8%}Ts&%TJ!nRZ_z z_g6Bi4E7|vm(F=YV-Mm;l43zFhF+M*n~i?Ba*}%XuBTo(m~eFmFYO~5P8O}4b&`JO zr8GZc1hf|uo^{eZmoH2q5qcu%iQT9Vl$pwwC#k6IUve)v#E!;e0hF$C2!OW?e=~FE z^wX&%C_f^R&A8i3%Ip* zRh>a8t)JAl`Pf_N)NdcG*pWDjp0N{;*hih(nuelwqGO5wfd|mDmVcLJgH9O5hgOc% zL<-2*WOEE8sbX3a%rAg)%kNHy#YTpA&cG%Tfm_WFZL7_S7G$6b1F6Kj8XiUtgckCI zt4pQ1RpoNL{Yu}&^H2FIg}MPtoS*a)H>AdmUvg+`r1r*gy_J0F6imuaMAwhN&Af(Z z!LTwCkVU_UzX44aV39fgRCjHL6#5R z1L#}ho&;PI)^Xxkn*D-L+<066Pc?cL2kQ8Bz?%V4h{Mm{EXw~DBmPLFeKjiG0qJVL z+U_i{k2Al`xBxhs{gzjBM5zthI4DkPn>x@7lAUGyRD%BR*Bc54{9TF0&No~S?N8w` z4GT9bPP7KO(-)!slsABBom2B*B$bT=BUUJI(yPCPExFeat$VO$(BINf57OW1*Rv`A z!YvX=;hF9_IxvErMUXh3F&<#glh_>7Rfe60=6rE+-QsEVeeCT1a&u&E{##ED@g$~O z#S$2_f;SBBM8tk2E-)0)Gx6(UwwQGFm#}98_cB$Mo^SN0btXix>QyU3$;2d;2~&|} zL}<|YZ8pYElgRRj^jBZc1Y++e=sEOkC@n6Z)Knv%2&ioy5REw89hJ8a=j_vH>XfVy z3NgWAd<;ixV{r%E#sL6udGm$wD0tYb_1(uwcd*GycwIktMKm&h&BW2ee&qUwt9f{k zVRB=rZU=cgRt59J&{S)u&x&#w4T5Peo(TCU!+fMgvqVS$j2F1@2?tRQ;z~o&n<&d5 z(bi2}@x@K?+VxmC)FDx|-tWrM(mn94h@CflHSbkju0K4Kd9c$y)t0PEpta_BQqJag z!IC>HH0m~eI)b5|;&{iS4s;4by>Y7Vdh2tLH+eD^|UjI@(FLXXtKYaf2+MCVe?QkYpAk_Bj|}8ghVjn4w(b@ zYnITl^5yT8Q{U~f*@`HpnWMU^6BM^*TUbvItz)koHsdDW z-{97!>7qy8=>0~oCD;YrfG+?=#k_xOm;G_1oj0BT1wec7=zE^R$JGw!fuC*X(mKN{ zhw{Q_zxUk#_32xDgvPDzFgY-*5-T!JOZgDm-Lw9iN7BPx-ZeUI&vxwp`!mq&f7QFp zTejfR8m>t8Y4)h76GDXMYObWe>@U-#4Em*cgzr3ie`iJpUFmVJrf7@nNz?@#Z~>E@j4x_I+0;?0CA)!bgA zwx+w((d*eBkpCEr= zruLIUtmV>B57UiXWAC|&m{Q*oh4oI(E&qGTz!ex*m9XZ`Y1u}=Akw|wjomAlfS zMW*$I?qsV^n*lNMtikW?w+N8!Z{ecYrpC$i4AT-t{`|isx$~iV*ZqZ+BAyI=UrDup zFrBeq=Se7b+g*R<9bgw$lQm+sXK%GJ_FfKFy&55~k93#m%SgFC~?i7=|HKV*k_aZf7X`vm~x&z-znJ`J_ET zg5*3}dwC;4lWGZJLVwX*i_hVd4JIqdm9^L7#WfmNhptiP#vZ-YL1W_@`%T6f3mmB0 ztp`(lJN63=t?lf?@9TW$i%?s>^VgbSd(;%5gR5LLlv0x~qyf*8N3FaC45xa{)NKJ9 z@ajbq($W+<Rt zB+d{41!T_~ilD8p=vRKek&5i1iBG3OPP{sk1JY8$$sd$@vQUdjqN7|kGc>= zOEcR6dWPN{-Es-$upn2CMDxtbLS4U*)^qtQ!&!kGSO&4*=$LSAH% z#tS>Y*IJ+Ml>9p(jP^);jkAO`gk_oPe$u=Cbl6%Mfk7$ zW>jor|BwUt!BxH)b1SOmTl$7>!%$ul<*cob;#Rcr)HbAqB2PLPd7Jjn0*3SkU_S>?BMq+|~0*pkii@X2VP#N6T-nf$Th zFs12go)Xg{n329C(Kr8F9yZzViq}fxYI;VF8ovy5)v0J`c07^pc(j%2ua~viZcNIK zz@M-My=gVI5ScToJ}*!fsq#s^Mzr*I$=U;3mlF$;Q%=C&ZsAtH4KPb{`A)k*x#>`?<>O6Cow_8KQsL0d{^YBY z61xfdI)5r=`t6VjSCRyX*fa^{-L~iDbuSTWdEg;*oAo+fA+x_lwP7b4HatQy#EI4b z`1+MBW)~0C@EJPc58v_=r)Do7wALRD$bxkLtX2v^=5UkcH(UwN3M7hgt691)$zCNMA=<#!+VK zf=J_QYQDhsU3nR~ek?h%Z~K&_tuXGKkn3n>kQ1K}G|oEfvrL03-aG6{^`zy7WP)qC zjZ2w&|(b?X?$R=0I*nH5|#6OF? zaE@I+Q2&Rw%eNX2P9MkDoKTO{_K{%2S4#b9I|_C3+LqRL&Z>-n0BrZyfz@ASL~mG^ z)yP$HmauCZenwJ)5izTlvDwdmeTq$dH|ZW(iqzJB7w^dC4wU~xe}~A*P{O=Umq?~< zJrMBTH{G?&-)O*Euu0ora@P#FE@@a^@Zdy(h)%0~4|48_;mjo)5d)y8+kQDmV3(IA zf09S7$akMTd%yK5g8I>nwV(O#lAX=)tpfqoFuThwbOJBj=Y25HklB{r8?zIo^}v4F z3azIak@>w5&xweiyvOV_o>|>~z>LEC1wCaECG;<2tmQ&I+t|(EP^6O2!FixR5TTP+ zbp@6n1~i;rCcjJBgcem+JRTET7xreR+ag1pZEUBNPwhI*dWh#I^{T_jt+|fH)ij6OUliKlf86uZ_<28fW?z<*J)M+M-7Hjl2m95$c(nSP-2qdg}6vErsq$7}qYF zE#=F@M9=>XZrskLl=43*&yr6FYiw@SaC`}fcfuM#(O#^Xal}TYb_b-|?{Vs~;91g7%f)9@uz~HvlOfN2|MlkBm&J;Fk)EW^0MYFn zMYF0x8NylzskTl)LT6^blLvu-5f38)l3CTaxY4sXl>tG}m&cQ2Ra_HVj^ZbhtA+|B+) zW>1)~u7OdK5A2%+90|{y#Hps!WkDm75|rR|3O6%g3CU1fi#%=th<&Jnp#L_4sZu}E z#3b-s^(3pfoTZ9yf_iT2A6fii!BScJ%~9)xM%H5kLSc+5pbc{)CGSY0mL9(vS-@{B zO4#aF+*X^ky@Gk6rA;%IPrR#5Tb>Fi7Xt#0uC2{C4;ss$4tT2#RY3e@aBwFk=2`UN zF>|yjPmicikl>t~Y>{o)7nag!5IlFE8_T72q7W1(Eod!8i0H*cy#)1|$ zf9j2sEhfjWz8#R~Dhke=f=1(y70{x4JVVpRNBjQfQNm`(SE|))CUG0RH$#7QHFhD>g2PlQgcW^KRp7zj3z&+S~tI z!r6|f|LWmldB#l+g2UQ%!SL$Ym56~W2jFp!0)_ZYGG)6D4j*k>xLTW1^ja0`wMUhD zZ6E)ItS+Ws;H>x6NG|+s@~Qo+JoDZXd@D`G$@hOVSzoR7=&AvVTKlzF&{1dY<;KZq zt|rg5cuyUkg}&jls;^8EJgKK2?X+#4qpoz8$wb#qimxudX&=HEvo3ZJSCU`3EjPJ3 z%xsfyGm)V*=lnVIfKDuw;C3W_q&Bm7D6Txwwc?InZp&eJtlvVOn%ug=s05rpI-7}i zB5N<*7>bn2LuO9KZPUEWy{1xn^*wGg`~L;^AKSJ*nhrBO2r5ugZGMqfoYdI%pvuH_ zdIMM5t~Q2;99~t6sGq;%!GB1-!_qKSEc!eWxfGF~S z%++MJ@9~PG_`d7(tWpUv{c4+D%Wql|#J`DNQbK+WX_X0Y`@E9!9lxw*-*atrM?lD5 z&tX}3nqiH#kBi<7oDSPJf9Q!J8n<#wc+JE>nlAm(eMa%e_IrH!nba%00Go~%YL#dj z=(8fP`bL4S_R#0NVOc2JDMad|DxQ|yu8bZVjBWlO+-@LsPZ5gWjx%6d>``oNSq28D zW-6ME3TE@9OeFR>IoJV_k;t~#TBHr_;*#L&C-OdL#BHS3NwG;%rl-zdJlzTWr6;j? zNo}w+=B3nFE5(@)!n}q+e)_JICFsiYMk$0dk{z z|GIeojtX+fK@O5P`#d%w&+eY;M5QctX7H?4R1i)e%8sojhBJ?A)hjC0&&~&|yt89^ z#2gY(ab1@jLwox7gtZie&Kt_S?K0!uzwP>_kaGU^A=Ju>H6a7A=M{F_A0ANETOM|2 z2FApH3t^`26ecSEW7JSG&v4f-Ir?n@wz_ZSK1dE&X~+%B-@mGk(x&%U{`+0^*Pi~R zc*~|yET=x^sDyt}k)?XNvAui_cvCZeQhOi0rxIZ#%BVNc0=l z9{g$~C;e43SH+0hR(1A@ypMddxAMd#)zh~fI`G8qOS<_bA}}{d)LrN~$j9e%4#b7Z zD(quAuL`3{>R|N}a@Jo=4$ResKOg$^HI+Db#{TjMo*skP%P1D^CB5mr)u>c_BF~p?2W5=1AZy) z(UDDB2+ZH{=_8U8ik20DLeAh=f>Sf>?qGe_T?u=ew=Od-{-?dX&(Ors1Hz8N4Qdtm zYYv15W#hD6&F4#XfZzwVhlTFZU9&2vfsdhE6oUdVj(xtLG=6=-z-_~}KUm5`r5VIW zb`I*Tc}#?fvsdVcEUMBB9{J7T5Ru@<&XI}sH%Nnh<_Jhqd87AM*?SYLt83Z8hs3?5 zn`>=oX-GrLGCMQ?5D^AWC&f4RKQDxh@B&5Cg!nZ+FTRrohXs5`Y!{g;@5#u5C}eHj z*A#GQSsln*d*vr7C!Cfj`~c~mFx!Vfx14|k^pBrnI{kGGJcSIep3VB+46quT8xM99 z)RVm|E*$@Ccx0alg6r+Fd#=+2Os5kg2o|Dd?_!aErt+jiQv2%Y>In$$%Y*(F~M^a%`hRG&4q5zO? zB!~Mt#=P~cHghEK<|DI|QA6&~SRn#KnK8z*8ziM)t2!I)HK1?%PwKR`VAJKmaIuIT zf6+$rtPK27Ps1XT8?iQe{4n4IQPOEcey0CL;tTiv*?IrCDH^|so99*7GAhhPqV(X5 z7a0id7t!r#0BABST*k)bBnfwtpZ4%#6ra)F!Cme;3fN+To1ft^{_N1hxASb~6~&NqD?gW?6k+ zQXMc@(T6fs>VQ#Zs(iJb98BRm3d^J9w2I9KK1py_EMQFIEk$Fz?OQ%y;8LG)msihn z1+?CX1C&O7{(~6FB0clwdE&`e;SYEEHEomJ+#Y@MqsLHk-}1MuQa|%W*VoIv`YuIn zoAL18(iPEARKOqk5=!pP62>+-rD)d zIED|pv&v4>)gxwUuf4i3k#(ykaOOZIa665h6>+%~UD2&)ko<$bV5qqV-~d#whXsC0Th<`AY~b3X#g}p%fCB1n{~+Dg$5&CYM)({yx%Jd(sjyhb>Y;0 z8^`=Rqw-Whcv)1RKPl_ za%|10BW#zGz^=*S(dN{Na_|~XrOTP3g)uXC%ibX(bDHnpQIN7W5%%C6q%DL-dPO2G za|j8{8`~CKxTCVnp;~nVam@&9&GIzi_Tu`5#e2}~rFc~%2}siv*{pV)JZ|=p6IHTq z>DnuA0`aC})G}0WNzP#xAwqxOsIi}>|Kt642Mt-kIoYt+Pl{l97;9@t)P0zcjpf;G6OJ4)yjK@2ZH9I&+37wdIh{|YC_ z0wpU0;SNl@2Rl`Cv13?;M0rD6pMH5my<-^ob|W<-(I)3x(WHN$zLtoh@HupB#5#@e zu}|g{lqA|i;mToP*Dl!;N5cZklTlY}=Yt339R-!4QdPB7>vBt|k~>cLEqYefOKKcV zfAQ@|`Gm4cVj!k$l9c5a{~Oin-gp|amXq5JMJ8T}_lJ5rq3*`ZAEtFJsG|BS86%nI zr*_3tE<{xKz-1v1W{|{Q=jpx&m$a(s4kl0dg z53~AiD*_iSZZ;6Db)$bM&jJau)xdS2@|Hxxm&6t^1b(y|RCw6U@_&tAcQ~_L-PQ7@ zEm#iyn!#EB!Sb+qJoZGf_|3{Mkx(%d-xCfI@+!`}HAYpTh37cH;Mny8u)YL=Q(({n zE`!e}ZWd`YZlyy#mpaB1o!PVAzx)0I8hTzNhoYH-x%O2z^p7Z(IWG7(8F3{Veb3giCTdd0Btawd)IBytU(dFxs)^d3ZCPr6YW$qRY_E( z-1hj19(I5XIDrDHz+lHQG0$0~63u+hWPQ>dq|w9?Fg&c;UvB%>$&-ns)0IJ~jQn?!}AJcsR&8M(OJV7y3@bcQ{O6O-o4`f21 z)*m4u7OH5?r<7_aN$DI^UlzJiHfj7mh)FXR(v_P28a*Oh*(+By;CRSPD5>6A{VL$t z=}_T)#K*n%c^;W2k*!OP_w9rx-uFgsJwQRsZ3e}}N*$YEe-}f0|1ht>c5Dd?%`LT8 zVv%en-g$I9blxkWqG8_2dzWZTd;7hubn9N${^2G2uI%Cx3OmjXurqV)@DM(j?E7ck z!t<6ll_pP$Ci*Zx0fH_P%^!i)iXvP_IRth(-ws_z{ zXDD2`>#>M>N~F}$K+&kHPV!VjZ`rM+Mlg}IR2shf?O2khTz{`GNY-n3|Ho0)z#ED9 z9|?}|(cYr8@?o-r#7W<33!M;`FZ}^+AFh=;mb@{US(1Gn{wVs~Dg(vlD@(hdp z73Nbyw#Tp_C@zvG699%-RB*R>fk z#_&DyQ=orXIECy zLri&icsLPq5$QEb1%jPfZRT51Ck&Hg%1-p%zcbgR00cciy0z|KX!K;OX}$^3(1~(~ zj}(ITIj_Q2M}<0%reE=&Zi#KA%aMhVGaYj`Xn$@g=<%9yDYM^gNqF8S7d;~g@t8sM zUpQ6%_+#D)>#IkcKI|U?hmc%nr-c*Q#ls4*ne5){GbRFmq7)2^r35$s1*lVcOfw5V zs_N9vA=xF_JezBc(c)O!ax;RV(gJ}BKW4uf=?B#Hn@v@ru|xi#g>o*%yZ@0T-5vJ66Hg_fPzGDQc(;cU#wrfu+;ujC>G?T23l{>JP4eM4zfTFw()8Z|< zIv>%_B51$l#yrgTSC(v>mVPJiQ5^CWi5xjfCAQaRfAqN~NZNlX)k9UK>qt34&=;`z znw8p>JVT`kDpOA%6{V9$cR=H*`lnC&i;7v3^bWfy$$oB)KVe0tqC=-A-E4*9R?2cw z(qZo%--E@J<0Q7&p!S5>Sgm-n@Gk0d=m&o=1oO6+9rGA*Ec{itf#C!>aPf|(X``dq`@1fvTh1Io<_j3x z%GcxQ0)O=K^rMWBg{QS+oUTJ2#5859bPj55tSmAa;7MS6YqxSvl?1?5J-ZgcEgW1wFiF2y&WsFM0juN}-WJ;A~4{yR=%L zq_}g-q?5{+t4jw=S1G+;I@%1XR{@}sT2gv#nI<@K&%sh?87V#7p8iDhEI(pcQ7e38 z1OhRIQXRs&?pJ4DJbJ&08nlrS!L>n}zLugpHkZpz`_T&nI*t!0W`u3svc(2+oYn`4 zje$bnYY_r%ozZgn&p&OV)sw_aUSTO$iid?vz@8X^+$}dj$_0XO#J*+zT5}l1Sh`{7 zxWBRkoFndWleYCHL*pn_AVJ;BDaNX?#fi6JH$sTs^5{a~Y~Y)XnOi14h4R$&Qsa)p z$t(a852<$SuI?vt{Iv*X+_yMR)6kx)Xmf9HH*j%89{P?kUE5p-OS?0ym|{9jYH3gN)o8zY&wR>lQ~WtD}-C=Uj@< zISPoPxNf(JL%Q+?=BeT#Zvm{FixB6q5@m-3 z!>#Y-dDhh{>CqPdlHCKbU>y-0EptK41#z!Wv@k+rJL7jf`t(p1kb%VGHs_!)8dJMp zYvIq8RIXmZF^^lW(GaCcBEUi7GW{`2i!iJSn7JG3Ch!wk{KM!9l-qMVw#R|N<$EeO zHF_G^;Ouz0U6?rnf9R2AG#(M>uJL%$(CBiT6X|8$&m1oYWTEq=?aL53*j7-X?~P-Y zNZzo}VzseKUSxWZ+NsgO^tzk6yX~DNOk_oS-x|PpBhEIFgYc+>=~O!3R`HBI%w|AZ zw;57QE?|=OKoJZgrRKmgU8ax-TpP4MCz_*2wX!`J3F{2m&hZ}TA-DU(K)gBB6!3``IR$}G~vMU-@2=P0k-8rf2_X z{=L^1TrZa%+KntN-mD5t&S4h%hqYW{S&7lG5u#K}duzSH(+>S3!uo;fz3zdEmKzmJ z7?oJHnSq7q48*@tbl{wb3JZ#q|L0SS3rok+??Mv4KPWrQ<|*T4*jz6+0jla;Nm?>{ z+G5lvxHi%yH(?u?RliEZs_oNj&DtwjjfIq@z znAc4DOGoG>+t2hC;doRM$1DWn450V%x6>Tx*b#UGLAi)|aw8Whw<#8a^2OjNXYBtvv z$H>Ly&pcgX!aG~~W=-{RW>6c{(M&s7A6;>Gd=>@HV|thNIGY;j*ly zA#M%W9u@gsd-h|d!&{e^>!L*^P~5<9vhVzfR0&-|EJR}2(Juze2P}Wmk(Y2k>tzIi z_jzoh)MbXuA9SoHSQv4`KU_3o=Cwoth{LOsD#fnqQ8yQv+me%b!!%+kg=-4kV?0yv z{cpe%&06|WUWkDEa@(i#KgzwS!+M)u66*)jP#LWwnjze$$fTU=c=!xwm}5zWFYswCRYc6+|={oA8!NQF;AS?H!_VTCX{x1D-&(qJs9uXy5?V zcnXc!p`f)eF`$mmM@08scSPGvWI$(!X)|57~BbkWuoTi(SZ}#cKd39; zX0+~{?$lACr#CC4@TS?E`*limsHE%6{@kW3fj8lkFDXxgvmYc;ZfRXz;GJ!5cRgY^ zQ=c)VU+qDduJv_cj!53r2qKF=NW8(o-*1dwV--}BxTbk+;klkick`PrTQC${mb&Eg zbGGqyK=ZBTXnUr`@A%F)%0#mEoF#pwU{v%m3JBXFBIwq3qx0$If52+s?Q4y0?gJU$`rjhwxV}1qN}I;>y7SHAteVV2wU=N zmi8c?<@AUsj>CdzKoB3`GQ$}L0L_Ag4slxj+Y;t<7~~fs`?yUKM!Y=%^rYxlGUGxC zkCP!}s!7C}NC`SEH?OrDePuO0ccS**SRWnRL zaM$7F%{dvHSA8@CNI)b^zi!2PqCK|jd9dH0n|jHC5WjcA7MO8roQOL*SK~91|14Xw z$_2~VzjI>p<=qjig6$nhDe5(B+hk>?C5Ni%K#w)qBU%i|>4N=v+%ZH0lA;7L_gB~> z)3M9kbS|NPu8mmyAtETz+UqN|8#L2ls3lR^7YoC0S#=tV=1uSHP_c`Yyb|IWb3H!BupcSX#>Ks=S zyB6@OVNT-t{T=1cr=`;DQ=Z8DbZ79)zrVq4_~6S7AV~Q#r;X20W-%l(pR2gpA{6{9 z%C`LRsz2|sk5AwI-%ZzpFuOzNXcNii7A>Ef?;{m%#`5c-I(g<$jnuQrAFKM7tUn=k zsijg5bx6CqP_zi6g~y9Px%fc+lA2wq$Er|RewDm!yp8CXUZMZ~a608{z3^={D`Obb z2w^3;iQc*@zRUqZ_gM^&TG+c~$b+*7?v&&my*^N?Gltl&@1{F6$QJ~IB(Mv4I&LDU z0R->hJG|_>jyLpb=c~XP_>cfux~W+}Ns6MW3Z7OB;d0}$X#xCx8gMOB!~g#lK=$!q(PsAnkb!O})smZ&&iB%c%PBX%A7m!!GSE65Eh}6R z-kqfXH|~7-C3(Q%5?DC=eJ}a9e3Q3~(Y-!7Zm}K(fzzn^Ducb_ntmy8Q*2{n=Tm%d z5m+EK+PLykxi~lc;`cas-f;s+zu3gbiX{Ki5zh8J7uVbUqn+{q%Rx{D}Os zD$PE(Ax2E$!OO`O-1H;MHLQIhFf;IINOfW)5rE|*P5rNje*E+a*(X5MRAsB*91v9& z{e!^w2d4CK4zm>Hjm8I)V^N*=)_BexbZEpKxMXdgM2gb&kV_quOHKU{|ovb>Q2+F7MtW6dI2xJ@I&#nUgXe6re3p~hK zrW+kJb5n~HHJr73HGBEI{;R+QMRYwPbERc0XQ={T$mdyrxQ;-=$Z_)JEw+VF&z8e$Qbz?Ahsf0fz*kY8eas7nJ2a`>ka#LgN9AGXm2r zf5e*qaJ0ndMUBO*fNkeB_o`%UyN;FTGCNy8p~AP$27LY=zAOD0z_A~)0yz$lPx`2P z^t|uNr_OfCL%t0xBwYLNY!l^8jonre?Kx!3l{VzKW3(?YL1PE`iNQ&lq4!i2u>2@G zM-BA;KmbH_s#z^|rX`kzQm9`^F}KBY=?2}&1ND1nwOn^BqV_5@o{s1qi2*(g_!M5< z_sEn~uWsPTGzK1PwllBdQbB~x7slhME%6Y#Gx?jiq1!c&%WWyw0 z(z~F9#6WsSm-Y@38x6$*eitqOD&?8!3&Y1%sp%_@k}?ka3ws(f94{|>_AZs_*D|{9 z_+L40^DYyyu6AwI3;jr=~{CDHi*-pI-=vjoZzaohe!P4<7e(dZWYcwlE_V zYrbMsfA;U?5_+_C{a({Ux*;^=LcH&o_MPd86OLJP0BA{os>3O8oO_$qf|ASS$yw_| z(%Ya(Q+7%H83?KptQ>%H2*hw_e-rk_o(Tk9A>U@SfDu^3iscrA}PSya_NY$i%v=y&}{Zfcd$g zO)tepGR{`o4@6l8B=YMGVbxcgZJmQ9`Q|)XX(xYp##^h)Jfk^Jy};$y#Htss0=*84 ztG#~x{FFg$I8)6S$W*F(d(7bUwyH085-r0yJA!UBR~%T`Qibiw3Pc&qhm6u^E`KYe z%l^UV)~WxU2#R|Tg?dY{wZuQ~NpJt(0V6&hP8oJJq54}6WpY40uB#<-7@R8mUBfR0 zKigE*PQA22^1kMb5q4nr04s3;qns`)4|1(SH#xsXdA7vQAPlg8u#;9ND^DVKiYD%> z7zpycQ+|2YEY$=lYjYYsi#G@Ep8U*tb3pLL$*2jL_!+a%{K=uxpp|0aSvCnVuEg%n zh1-&bAHV6i3fcjyIcw2!nQOXwraLn2I7J5}qX>TZc(AP$HZWYI16|?la4nMGKh%qB z{CL)Kx=XeWWJ+=8oo=LRQiu+9eY1;yA1FNBxd#~=sqS_p;?dd4*>?wtGRyO)@~!B)u@+7s#3E;c$uSn+dQXb|FT>im=Q=ZnlZm`k?2#p* z@*L9}%omT1^s@aW-Ay@$fEuVhYF#g-=tH_SeULq`<1#Pi6aC4`h3;tF(Q zS%3?zd>RGPStL)h7S${Lwn*)5Sj%f#P~Z%&_#sw^Z?8-8+fw~i%p;~Jgi@x?XnRSV zZ|iZT74aS9J%!mmDb^^vzO?N-o<@ojwLp}Z;a~7^zsow3YV2_EosWtrf6b@FBS9fB^!9?pl$Qm3Rde|s z?{vwPgBsoo439z%ZpqvCrXeO`xLBpChEGtR73SXb=IJY=E=WTk(MtW`*W(13uc>-o zEA>Uz8eyR)P>k_QL-3~NrgFbIL)2KP;ZV!TJCFI3}^@ZNS9|6&boXybNtAING$Bpu%)9v@YZy-S`eA-YaS`t>x6J2STJ zHNzR+Rw`4k+lpRW{__p?UgeEI8*wn6{q|V70Q4m|*qWHsg=#Cp34Nx=;Y$67M)K5nZ7rr^}#4qp_7y@pZJQQ8gL z$H-m*)`g}Hr{&v&5;Y|oq7!Rx`ovF!u4d{oA3sJriIGpl_*FLEnV!gZGYeL9((fd4 z&JME!7q7!EP9?eiY~#*Z!6n^A*9Q^$#`cG-;ksZm%F@V=Z{q6ch1BVu$kbtdULB9s zAoL{pD=mJnO{7htIO_Uz3+6qfw}a|3r*ct9H|_V*wj3+}kysZiS#_<*e);9*Ok_{> zVh;XS*wBv!^WHBWsfn7Z(LZ{{IN51==q^j}VsI5f|Nh|?&o}2N-s$H+dWWN%D7WQG zylIE=$dHN}e*d&avn)dUEWIoWCT36$Tpy5yOyoo6#NiZb{6-osaU^c>BCY9oNpbO- z-%aKVNzt7IIBFnY_k`A;WSo(ZasA{IPbNZHvh?tr>Uqs-wbHyKxw`gmth|`Z%&tb) zOyAW`FO8uz^&_2!9xSc0Li4BWQ?n_DWS}3ux*U<%##<9~zR=&zXMC2_uzCVu#e73} z;aufnmO>}qcfFEk8d(ABzvlPKz2o^}*wc~sR34C)0dEODBc_EJ{JhJyU-C@9H`>#X zsor#{AcDKvO|Qz;w}GbW#$MCf$2uJq;omQyJ3Bi3BaScg&5&t)j)IMKW?~34S^!ZD z>S~b~2((7>dIfOfrmWb=O6TY^=eG0rHS zKOBaQydJN4D*klVe=D}O2URrdH%mL*I<4XhJH_Z)tk1ityTh5#-L(R zP^A=$tIv}=9Gd?n~}QhHGulUIUtNR|cnd97 zbHN-c=*n*LqyIPjh?a60TsIci`yF4HvM zcTXG?Mp*2XS$BD&7KchrAXa34fyd|4;|pKs(#*L-KS4CKdR$EKyoW8^mU~@13L>9N zodfey07!aFa^v65(xUyOb6ceItzd^dUiTr9Kn90x@RQ10212?)E5tb6h>zN=yjc0e z$KRCePODlKi_4w#Zo6q_qw!#)&YrSqatN#pLLyuF zp{k{yW4IUS0CebmoWzHYK37V<+#X2-{P?K^alGO-x>@on&P+YjrVxx9*dggwO%~Yh zQ9(qNXDPbZl*eV_D#P(nu(6T=awSgx&)dkQ5Jb2UX`o9%Zd06e)Xm!wnKNfPXJD97aGD*TS%v-2gj;57J%(xi1Dbqrp2w( zm0{a*h^{9a(bcpTOlur54_MPH7RANz`n<8t;Obtdn}re zI3n9RXUrooddRm7A)*Ig#8?ZrvfvcT3rJ3AG{R6jVz~9JFp*2O3_$^Pn)9O6baf|$ z=`V>zSl=j~aTSDNlET|^a{J|*b{YK=;&v~y{u&xKwBVE2p62W)4omd#TXdV5p8mnG)NBtv~`mBTR?Z^p+ zDsV_Z7saA=VrwuhRce@WQ&Kh|HxaUZF_%Jo=axd(SJuM18W;=gEhi+pE|it;8)>5k zBsIGpW$$r)fR3E5C^vwA_f32%(qiT;XixLW%@b1nC;S8G=%wX(ZLA3>i29|GX5@s{ zuVY|)yMvzRV$QuU5)*qL+5bpB$@kq*JR)~D2mQL`%UEx4m}_U?#*zW-x( ziIQ=JQJhx#3H!6w$}L|yUdzjI{-$HN>5rJF;B;Gwzrd~=qkGYMzQeYSaypfK3*)jt z*rZ|Gy;r+thIQFFRn$Kcah;dZRcUkrn$`lXR(N?&LSXG)bnMXr(^COf;Ar2ni=dOh zRUW7p0}mR}{9f1QTD+j1*!S!Enf%tvgBfi=)U1s|W-hJ%Qbd&OrC`uC*q!+dLs$ix~Ny)r+ zfTv6kQ1-ho9hNzvf6pf-?ri?tx19XVwY--{9 zHK{kvkV@dN-i=j0u)*K&AK=yvE9>4B?#fwc*2opw#_v;_>A<(bj!;Bey z+`iO;zq`pBvc1Ee44pPBk?raw{@gjMnEqM7y~9`@RG{ttv$2vdycis`$_-b!NA%ia z_T6vIj~UnjUZX+=aI{Y4_oc&_oH{EK1Wil$!u^hv_(__FZ#N2ZIF~8D3QpPhCVil6 zPUul~v4TExaW>=LF!z2#Cm)5ER(uPTjkFv5Moj+}Ak7unyy`!ZHHqlXGUng-Z>LdM zS)X$jgc76VtpMWn_HpE7;hwJ_Yn{=q46` zVM(Yp<29SBPCK1V{U8#ODDidgUdg!Gd_PPJww>fAE}$Z6%}=ZV3tYIlLLHtjhpBV@ z(k=S6)>thQDHH5EjqGLmoECrko_mH39^4~=%5aI8GJ=L1m@FPpmqul&Es>1J?#x9v zT|Wb(x@otRo^uQodTb8RO6!PS^zr*ceuLQO4!>;@4MY%JAAaSX3!VGMT9X5};I{59 zveBY6tsSn-D2fh+4)h)O%vF+vj@xV2!Zbvr&9yurvuwMajboo8)+`biY~`p$5V*Q2 z0R2jW-l4o>j-DnhpATms*~Ag^ zeYdArzp70^JBpXoZiBywdKn;YHcQg@BcFMAh4;`N()>i772oFWBOa^|;PFM}CE3~y zNy6TcH64ExHPnoQR;g6asvjnQXU#U`so#J5Wym)R@<`Up>_1^~*6gl-ITyU^3_28rqTf8EV5#nrt z{Bxlnyk~jM#FBaBd|yqxJ2!u-b)=a~>2Ov+cX~k6s(V;y>&Wa;ju<`_~D6 zB~W!)_vbSmcRIjcm8t@w0k6CJkJ^$-uaa9%b$`9tG+yQqw+<~N4SXn z=Tzb3piAHB|E>9R^c}exv|}@~s(m!iY&WkkHb*Vc4baFBN)A0;P)!Q)!!F4Sm*Y3s z?=aSX8yc(J;Abm^{d4{2Y6y-d6#_O6DtEg}GK*QeO2XC$23{VrveznRN!ixKaIJJ6Hb7;*6j$dg*SH(ej@J4Jf?d;_yGW=ByKOa79_x|>O|)Xd z@UvMdj}Gvqu~&5e@e)deEbx-!B1Jx=$Rh{Y_bDVX40l5KHQz>R*rJ9&;g}tzpRG}) zwUp1@HtQd#s^TK;l6G0ULavL__VBzRka)x`j1HgZ7U?Yg$?Q~@*fGU~%Eid1yCyql z86%66gs)6GBdI_N+0!_?9wJez+Hu)g&pZ5R)D+N9h zGPpCm$OC)higAC_49aX&-wfr%5Z5X3nW$ohyZU0;AsQ}_OMNNFB;`OETTM+2yKSMY zvU0b?mXustf$n7aL1bkJL_1w<7~JXnY0>h*)C0E#K9F~6qP~lLMPhr$Oxh#q2^Ze8 zvvSI99nrF&_8RafiC9K^WaIry_trB@9>cUH!P$)19c1F|8ReYh-l)dJf%uTJO0) z&L3%P?fyOn3fMs0#1tF=uno&rR)-xQJ_Y=Q%DelVUNIOqE#m zV&wK2{?OEnrhXv^T(&2poil5&(KUVD)nLuu{{bCYZeqX2ypHyX)V=mtnkJjWDmO2f65ph1?% zQJ-u3r|Zb(wsz$X#J?d7InBfrY4uS7!U>4)!Q#(yARlqZRCUrEFl%^tQfd7kb2(YZ zw*FuD?&C(&ll}y}*e1qH>o<-X$Bm^s(p((^ z{>=uvjY{gzZ*?ot&*>=mtG*5Ai8@Z_<#Hnri@*sN{f73$pfc@RbKv*PYmPfY(<8$I z9S*d?1lSh8>fhr>=hgd*u&{^uVAxvlr~%j3qo=AdxlXV-Ubiy7YmlEWh5g@f?|fIr z(J=LPK7Sv{2e!pJ?svJ)Ye5IQlsEot(oqedIr>w2S8OelWu5Ge-lcVdtwp=s#mdv# zOJ+}qb{TRr2q*#XOvc z7G{8kaF8~v1*?T617lTMUSev_aIjCO#$7&|1?P|jcr~`z8l4r9o4@uJ?YhCs8_!c{ zw&E}|Yuqy4lZj30Wu7`C=09IbU(WR579;6LFk9Y64jCR91*7yD-G3P#^^FAGa_*Vg zV_pHP?^f`Rvk0|xQo4AOrGJ4s_#?Y>l&V<2qF=Qpq97xAoWQxCnym+-C!zY*PWYfK z8zfWgVqkT@Zp}ohc0b-{ud7zu{cS>xRl^Aj<%sW}7&n8hzZbaXq`k^buML*r_7K7m zdYTeD-Ed=Z;JUGM5|m2b_ookQWb2y045*iSNj$c-P5aPJ8~tHd>nG?<{`xepQGwt%tn*Wxl<#+G%uqkrDI6 zH2bIQjS|8Q03mQlvs5#KX3}>ge3~7^6n0Y!uxc%quRN zlCD9N4R!T!z3aA;blp@woA&IZP*~eSi`~O?v5dc!x>JI48w1>L5j-G_>GSAhBFXWp z;~j*cy?BkIuhCmy{UH?VPlrJA#=yYSt+2(eWWkAHD~d7LxndymcuVuzFwr?172JPg z*?yE>EIv+2FAesP1NU4}p=bN**fhDhZBTvL4&#wi^hdSTPIPUYSyw^WsM>q8rS%=; z51pcrz-D%1dN5<{EaXaA5C3p!?!mxQf!tTCPfyo<;P&DCzOk&fR)o6`>|=H5R;)w@ z1ya?BJ4wUA;}1ghN=-i=eY4uB8s$F7wql*V-+1ndu|(%b3rFJ4J+DBS7=|9#VYV~5 zGm;Qj!z9I|-1tpG!D4eYj7+?+2qAauBkXIFG&lLGg^c%9qG~RZ&Hb2t1k@$F7T{#>NMSJ{TkaW>#`u3=$?%ybMz# z9YiX>m~roVa%-iN)W83sk)+N}Z0W~}SH=Ph5^eo4wa`XMGx)QyRs#uAmz5_{k^fUs z@OUM+E^@@m5q(P+siYE*qu07F`1j^A`QhoMq`dh~0C{W*v`#g`DKY9fO@1<)?7b_r zCHSX9n}NYAtGp@;qeuf3%Es{%?z7GGn7K0R!xAhfTYMkA?s@}I6C+T+b$RE!4&jU~ zy~E?@OCeS-tpI%Jy@E!ebf3vVvIf}wbjC#cyzM9jz~h!7E{-!*h{0(-a%jNy8vP+XH$CjNHN5gGmb~M?VlCc z)X7eie}|>Pb_9?pg+YS=Rd6JV&2L+zL8;lN4IikCvj0nNV#TM>mgrlWhlJUS+gJPo zH<#+APgw6>;(CdC1-m$P5SQY2QftCawe%b$N&KiYt5;|)Kk56d^zrTf;3PC`I4&>a zHr+bFAU8JH9m}V`yN6A~YX|(wc+?((ey7;`j5-&Cj6I%7GCo@0cTCuw)d&y0-_Bl_ zc>IP(KA_aExx1rF&#OWUnRUlX>Pald@}zs7A4+t4MC%+K)F@o-Al*<^Yf z1|6bAA9Mn{g#_4vBN(?}AHO4zTzrs%OC4`U zbY~FC#LYwqU8a_0v*F7Apx#N3yfk~I$;~SkE&@!nEulfJzG5UqlV~c`Gib&*-H?03 z>4~&@RvVu8%KVFzCH{Ba_vor3^-G)|k!*2{_#(HGMiXQ7rK8Hru#)8-48b+0wFMT~ zfw&30MjC$?RRbeBA$vUEAZ?tlP5n+FcY_m?1n}Wpnq)8OH1x^<`xFZdz^KOU^68_- z>^vSm=R4YG1GT0!dPD`F*!lMUtHZ>L>Qbz&bcAusC*486842zxSNod(>U-*te+vxs z*90U)R(yXNyNR{E4N2@uNFLM6n4=0YD+QbjO@jY%v?@`8tL0W$$66?WMb;nvFVP5| z9aHCkv~&YM6hu8-Zupfe&!u_a94L%r}_O)ya{ynlQi%DqO49 zyw~axaUVE&on}B>TG!obyU{AoBdvC%^7^Ieks{R0Og4SRG_|_Gx+2^vBtbei>n{Ca zg+d*6rj_8T|H&wgm0R?UAn*W_R_w`(tTv3vUXzi0QGm<4Giqe3&Fi?5#*P#0RFSD< zkJPC4E>3!ZU6b913J+f2uaL~tl@<^elauz=(&_Bv(ARBTZZvwJOHiMl31M9ks~Uj* zq|-8wf+R7mg$P^0?>P~aZ$vqgSkw69gwdKZ^#`Hz@|4dY@vS*>F2|(IkVh6v&i}tj zUh=_Z_b~8BOG-@#jBJUVfrq68OV6~)JUgHOHCGnrb#sHg^nq1@`Q0Ff^k-_DM4NBM>k0{7oqqhMC5_IqiTR;7V{%F{) zMM2>AVD5A3iu(Alw#l*uF~l#G$jc=4Jk?UoBDqNG!i0N8mA zGFdNclDt;wMU74}fbnX|9WgKIEI*b`6tMGd!>y0K*c@O_rhM>Qiv!ssKBwPPENkJ= z_S#`s#;zgo+>VvMPMYsZ%Ing0vDWiT0tf>bvFpTBje%aNvUl#}B=H&j`|fE;mPo`@ zS^lo~mrk!lZjX|&9^+2pgAbco9t;PK?Op@TR9Jfn-XAsgpUpgkkCL;i=I#{vjYHe=v#t{$Q5r zLl+0uYn7)tXZrZjpYpg(pt9`t{dZpRvEH9y3GdSLvN&&I{IbMM8n@T|A|sfPhM#_TuyrA<5tT-_R^muHJkMEm}XTd4SOXtxqm)`nSQSDG~{~{pxEN4n( zs1GqjQL9@M#eS3jtFwY@l+%$ln7AX|D968?vXS7o?+9DI!G?B6^jl32lvdsAw9_k` z--*;!`Z?kKye?Q+rz;v`fL{L#xo;Z+p9Y`&27drSK)%0O3$9E-~U_M(Kq98XZFgmjfq%! zR7@!=|0t?xb#|X~?vd%uIv4rJoG?#A7W){G*4OgO2*(+$bcMIe}bG%QdKeRg_jfepvD2`tc z)Uvq{D9~FWGO#exojTr0h$3EyydSo^DjxM8miVGjhSac13g@a#;Hc>N)oNn;S?n{c zcY?9t&89R6DfLWdmH9&pL77mwv*}i&>HMZK9H+&r9H=UF>1&$idNNUhr!hpq1VR2F zGozpKVTP3-ClAN54z$0hc%jU}UKCySb_x9ZpTf$Yj`F-rTCzmga6*&TumCVV^F^;{ z?nc&GI!Vp+!e7Fb{MSDo#BW4 z?Fl6w6Kt^n-r;%KtPcTe{ei3!uje4OX)&^b=boNdvsVka@+G6H4GVsp5Z7_)nha4+ zHvKV?&<&te7*-UNdKdh#w%S{MWU#$E;YaK}D->3vV(;7PYhdDqx;QO0gnEM*6rEhZ z-srJdNyX!mJ=91w%N}z^&U+2G_d}hoak;yFbWH57&+Qby*7U{@B|^3?NfM-?pgS?x zC+9JR1K*#Ssk?Dvl*Na7_x|%N54*4UhL1qRHR?V2 zGrDTzxo-gO?13}RFubim0kMo5UNQE4^MQ+vt&W>Il9Xy4F(Dzr$PC?D2h0zjE3o*n ztPkY{x?$qKM|h801l$KY^Rl^d;%fb$*O)dW=?R6wJO=-sOxBPpNdtS%mk&g}J@T7v zdZHs*d;hPagU)Qx%3u^C^d;ZClp)&EPlwDAfr24ipUd035QB5$Mg9BZzEgf<{tNPuy#sj|U(9=9x>(Fi!K=@EZdyS;48qcijCBQ%2fL;6Ri@<{azdfv+>;Tj-a49NjQv;dV>@^VV+YeJ2T zYht;bI});?mL2z)BM@p7Z+)9VgQ`>Ol*94qOq!5kqUQJm8uMD_z!jaCBn{sii?zTb$5L%jTP+#A^v8`}wAVTNXlSMp9{jAnNC{mnpT%hsM1OXO zoBoVZmPq!P@q+nu+AjM%9r=)b!}I8${Hsil=Bzb_*7(`>m~~_g&84mDWIo!1O`-=1aWv6 zPP|M>+pL+bV~@+(FG8u)Z5rwOpyZH7czSqKr#cn=?Ai50cE243CKT1mw=pl15m8IZ z-#ysZ;b_g#ZBTK=8p5v%RKF|T9LvUbQkCi|1KRLT!Na8kgcJtvM7+>=m}q-~3iGYy zRd?=j*4@Y(=8ne&Z_luZj4!#C+=u!Pfm!XRRNrkJ-aaFriBtst!Hd%e{|V z5tbhzB3xr~5qJ^EgR<|Mx!c6MF zS$xp_!~xs>H}xXFlP<3*Z^{DKXurL%Erp6O{y7-9)%;v3<rL>n7xo;;2zI5LM-{pvUFgic+UcCiRt|j()zd1`TR~j!4 z`Vb|Nm`%QMpYZcVS}veV_(x{@%{7_ozRoChEn3_gC_yBcFsMeX4k&IINPqne(-dD^ zyMff!uekK(UKLhm2&jaC5+olNVm@Xnr-#PZ^W6xNnOrN@6*YHilb?{sE$JpSd9nEU zkN1g@a=NBLARfA&gLcKuhU$QKKkEEk&-+%GjCws|$Rc#+_T;pux@0O%&mW2-q8kaa zk=ODAo0D*9`lxHQ;c09D;T@`4raEbhk1OG9W2S++5dRL@Qi5KY5doXF<=O0`JSim` zhG0w`D>B92XCD-%at1d3o|-!FQ|*@UM7T_NC=(jCq(8nOm!^(k4AGkz?|rgt0yyhZ@{*rvaR zh4w&J6ZJZ??Y5(x{rnKHYbN$m1LJyArHWTfmsEux`oj{4+pXZJS=mkNkgxU?n0?K z>qfGgZuR$2u1I&;rr4wp5;Vc0yF)3!O7VK~ws>fI%wd5a+qpt!3~V!2d&kAT1#FCrK@(1;xB+^0go6ho zbMN+#z4aKMYD)Q+dH{9g6NU*Li$K8JWhvqS7-g={{|WLk%fnqdI5s7s>}&{Ka=Ppf zQ0et*#B6y(>i2f{#`}o4re0L&mi0c*v1(tJJWXXI?FM|fmfX!H%$z5#DEJp;d6mG; zb+B`H`u4v)4S0S?*N1rhO-zCWSm>xev766`*-{FFMT|3mM8>IU!oLz#5;qHhO0++w zJ$y))0)fv#c5IG3v9}uNaOU{g^R`I~Zcx!Ah}-Goa~Q^j{e4d7zzvPgWLJZu?@gJd z2{A#Er2%W zb!BR(z2N&%kfC8f?+CLL_;##)%um`F_BAp=-0dYYS>TVr;pternMmmi2OXXP7nH)z zBYJGw&l5X*%bOQ3*CmO&VeDkNQe)P;UV{e*>|~M;F59yh394}q@m{fq9nTu(Ymnzg zI%LtSdEoItqaX^Uvh4?gHB01ok0Vl3y;AM}O0nxE9Wlu_TTg=<^{Yx?xGxs8A zxM7v{9nVbmX7c&!NG32Jel}3he&twKXJR!BOIlJ|_ujH?+`MRLLfqRp96jM!qh-k{~9f|!SvvH>MneGO z^rt-X@4cNJ9ebflDs2xP0?m*9#N4Y z3|xHF?z~J)ytL*druL^CD|GJ+WG81}_e|rswhSrv0C1wg;*zDJnB_%UX`@eg|2^{1 ztEC{RB2IqnHQu*D6^y);cb8IaCxT(I=FXdOT zj!o6}Zc*k(z^m(#3Zm-!Dj=i3lRHLz(bZ*c zG3Sp$n`4XVf(l8IvzOrpL-|y{hNR3SCBQx3({`Vwz-K5We2R)Ew?(1N;9Mw(suI`y zqL1|_Us-&{P;TyO6P?6WUa5MDpUL}h;Wprr>w>J}uB?EwrwJqdMs}~&)K_`!9^tDF zD%I~E(UKy&Cacsp@=WEY&$dihC(U~4YiDf>p{CWIRhMgCnjYF~zlNNl3Yx3m!<*<1*oSRTUP3yT_CJsz3xAE_N1RiA9Xrg=SU)!F+J9-(@pxF&P4 zNBxiL=#8NS0sM)8zg$AoTNcF&K8tO$p90>OF3;grr_ooZUSj5LDK@ym-ChO9zq69Dg=vmZl0@23S3u zEkI@yzkCy)N)Evvb1cTVNkmZHgSswptE%r|7jnrRyJc51=CSmfTwV?Pzji&+Yc(2P zbbBrMvtF~-qHGTZxVD0~ky$QV3ZR^zsOtzi+nn6=tbEYQjqti9%> zG8tJ1Oj$U+0XNZfVO4Bx<|b_aVh##>Pq`yyprwsr0fA_Or00p_1Gv9IpPWDW-5=M1 z=Kj`n%}*AyaTc_lm(87cQ58if>hhSo6W~ZrS^KJ_Gd7`N2!V*FMmZ|XyVoMMP&p)0 z)KpCslnlHv<5kjeOiL@+_S+qlAe!yJj~Lh#&Au|>H91YxKKUR1Jr3e#wx5uB2TM6V zsSS_%n7(YcF_5u;%P_4#oHF|(Q;5DoI2mX9i|37p%Y#f zlWz$qCL|i&l)5u0)AB~cbXLYMJyVV%IPURQ>;I~6<8<_s_&ziE0|)zb z!uIvFt;|H(yW~7HhFPNY&(FADZ60@{;HdJ2^p#7cYt)C%Ssr&|bXHrRHYWfu_ZT$YPb^ehi8F#w6EjIoHJXVY= zm!G~=#|Qs8XZSX#Q+xlfEwDyTT4e?cg>;airysD+3P?ZWH@-G4@_TBdaVRoNpW&x{ee#TPhZ9U^BrK_n^yMFi_F-JU^0UO5-_?3_YxP|Mq6F%>x2L>;UIW9# zfg9MA52c&S6S;MO(H(Q#K>yx;v1(Ly5}SuAfPqwlmeSQhok%v_BP-)|suU1G$xnto z5=_91y({s+q*TW0{QFZ!B?b58W>(D|Z~cZ4InNN=m4n*leW{-mL}_Pj+N}aQKICbw z44Ui0lI85lFRV2vrPOl{6d9Pd;#65_H2eZUg?tmh>rB0+`1z&UN@}m0<^#Y*g8Y#K zU1f35y1ua&kM?f0*1cLESFbu@a1lKfD5srGk?b23p7Qkw^>mHmQ)>U9Fru!U7VGDj zbmIai_{V3H5|0@m=3E5fOA_a`InHDKtcTEa*C$aNBPy0KkSY5%qR%|p61gAlSo@X^I&nhlyCxMg>%O2bpNH*41o zGy(dy|IyGfd`2KdXKax1&)gg5wwMOIMuJv}?*mci#9to4GlUSAzCF^{M(@8E43Q~- zOir0AXZQs45#Mf2=dP*&eh`+6{U|>3_gpw))Tfb8yHCOsT%jG5Sbzf-bKItg$ezS- z!Z#+~BdU|0hSr%hF@vu5WB59XnUdQ5j;M8NbOAeTL)dnpvdp{1YO1>t&LRsGrC@n{ zL+AZgwlP~5j!MNHXKtc37;r3Q{6~k)L)naXlGD-ZqRaY#LoWUmTG*lR)SCIxie}re z?)nvfE8D9jL}cUl7jwSE%{f@zNZ>2^90q(EkTB2ZUv|sBX=mNk&fyZUUh(~_8-W;B zf;egM*WOgtGx|AR>Q=#jlWm8N_hC)?&JXo|i^bx$AMbir`qxBB+qPt&hA~Zdrft-J zbq!W)(VgM-wHaQ9CkhcBX9h~!G7cBsO%;x2+`Ox(Ocu?QvkQ7|dPasMT8trL@| zy|R7da8NB^{ForLx8EMnerv}mV!e9V=D!AcdB|$lVT7ttPk+CrI0<$XxX zu$>nuZDg(_g50h*bd?Ac@PHvxd2?!0tee^A>Id=@LKj5Cs(@ZOt{X69gNRgNv$KVX zVJ^P&{0{X^Tb|bxMYImVrNy`)O1nq>7CKPTAnwdcQdqgrRJ9O;zwRzv|67;v!q1(P z$>c#$`;HSMlY_P)C;N5I94|O~r)F6&R@Ap%XQlm~BWFsL766(1GWAp;=nzJCE}`sR z|ETrAMRpR26pmr^HsWtLezqoLl}IeTAIB10l_8~D-*3((uvAe0nG%*u#}ihW8ja%v zNt;-GFa`&_gEW>;oHq*8*1V{@0B0G4j%#l@eHDNI>%58lC>|1a-6H7*T^{o=if_Q( z=3Oc&4Y^T3z*KOGP1*{0u36njjeAz2TOh7U+|y_Gx}I|%c@Io1Ar{0u`-B_b^!=Ot z+Wh~8vPo#qk5<#w z5j<+}L}x^1`Fzs69q#+p?|TI#`-yR^Fj|gHP?z`Z|B&^j@lbGo_;4l3Ix!h!3u9Dv zYLsPUiR`yji0m0<>^n0fgk;N1g)C#KRI-wJWbsbRh{h0kU388B^We=8_FhR$u2_858ujV+B>@`#JGM?+l>m6n@ zI7L+l%r*4Qor7CxR(X^KNgzC-yu>}XAwTr!vtgv2oPLvAU>@!saN-&48tqo5WL*lJ z;-<*uI$rrkSO9WFxpM7%Lhp?1Z&nA0wrCpJBgs#a9C%WwTC8|2S$ZyzmLzxd0{lbm zyeaR$6w+$!lwLCT>l@ao_7HCTG?*Uo;uFy%^QiONaV^pshj81DNWjn6X;iy8?@e(r zn%bCK02V{6rNkD{?^U^*7jCZ4Z>&0rfMiGm#=Xbk4qZeDme62!LHdEt-ipG8Jm@w< zLds^%cQSTHQ*V$wTkJPgxR|2V{POeGU7};N72dC%1!PeZWnip07P1DuR#96W(}&vK z<i+s&{+TqrEhF_Q;hblJ{nEik3`4(9+h#Ky<1=-Zm;V zz~{nD%KDnGZ;#P@ZH||a-y*>yj?*<&fo-6FC$nSSo>e0PR5E-nD9Lr`wWPB{(N_+p z>jYzN6ep)>NFpAma8VyA;ER^de4X{=O%(v;Ds7H_H9lwK$x}eD4J_Vjc~?rxL4uTa z`GciRocB(o;}uV(9IqD4BY3!DH_0YP*~U((*uoDMDgjzc?ltDPgbL0JshZU_gNu%d zh_#gL>264Aq!p}*F|UgJ=e}rHXScRveTDQFZr#S4JxL%v0>v$+Wf4^&{OCS(f zTKAx#6nE9#4l7CZ{FEG{@w_IXs$^B+IKjh!2VecCLMr@GX~rxYgAXn3NVMI4y2|cf z6qoW5&Pg7TRsyCnDR2E)uq%d92d_tb5uVb1gSX;1(w=F~6fL=k zCOP`$+x&FG>91coG+$waWx&!F!PZw9xXAae%#CT>;ZE7~0Q1=myC7%RnjnE#?RDv3 zd4118Y2D;z_IQ8d;aBgCF9B(qtk(_wQhAK(+cvmgjiA(F&EE6qi5$y4weTd(_S*6l zr}m#ez8;+Kgzx+!AODmt85|fQi(mLL?GYc4_AJVy1Fxd+KvR9-UFtA!FC3)$?TX8^ zC~yUVWPWd}*;&y)V(xGpzLko_V*eOyN^4AAObLG!FwIz?q%1t%D2pClD&zW}c~YixCuDvJj;Jw6_~OMNIJcvDB`n<^^8 zByjZ4NedUsj=neIuMDU#Rew&G>hh>mGM^rnNmCP;GBe~F=bq1>W#I=r$}Nkz#1;jC z_2m8=;`K;u`xf3E8 zPV-FI`~wVM`HHx})#a4&XwGj*C}8h`{!-)GXr7YKwd67)+gq89;Tw@$>j_c)pS;O_ z+=-Wo!Gdemcwfrn!KppcPq0 zFzcJc2KJ3Zn)*fTphFv$uUS?7;)mK%>T84$#!HhvI^t2rDZyPqu^xBSsDs$oEi&pL zszM?I4WIsg_uc(Y^fCK#N#3`B@Ck2pOCdKn$MQX}Vy$SYnHU)C0#tM8>F!Ph+rt?oUd$6F=X$bJ@yA{sqKc{=-@|Z!;ZxazFRV ze06nX|0#?c(o&m|Owee(dDW_6)mQ~9T8rt?+i~M0?hGjC=J$NLa$qH()bzS{{Oc2% z8hhiBC6Akx|K-s)T&%#jb{3`Xxl(b!xR;?D#)F4SfD21DbTLvh?S5#T=V@cX zTpX&;>+;5q*q)$~rS{RmR;-Qmr1W%`@ee~DCpv-pR`F;6r|YA=a~$@tdt|=7=yBOq zD{D^G9GWft2@+}Cb|$3v67YelI5mcC7CSjVwkR@!EW|xuye(Mf7N0EorlZtKg_ls+ zbIf%K%jpGY6q}jw9s|jK8?fP1%LV;q)Wqxv4h+H*jul4IY*?v+YW1gAtRC=V*EVyRKMde;$HrtSv6Y77bS$6L4XSvbh6*JM|D^Jzu3hf&S5 z;j`+wY>;2tz=(%Nb(q7+c-%h4 z5HPoWiuWoq8>wa%8^%;fqk;1rzRpwlHqy1`fAws9Yn&zwyz1AAzKZPmX~;KUl$r>f z+DVZ7)T>S+7tGOirWxG!IoBra1oIA5FGf;T*TgsoLt8sKti|J=VMr5f)v9=;gPUy6 zf40i$q{_hY5iJ1q7ywy&@eq*+nc4B4!WS!#kmaR(%ccKF*Dvrn)hl_i-l-H2$WH3J zAq{;!sh8A#O{k`2$1AniC;uE>be;LHbxP=84m_jDsKfVxSlZ{PRjD>Jt6zWd?A#tS zwc#!G7`ml-fVZ9?~s#5#*XUI*;qyHm*bVNOuwE-Kh<8 zS7^lVT+Yb7THDS>#F;NI5j(rX{G!hblzRbaXZ+%?i@JS0+MSuPSgupm;8@WQ;thop zJt4c)x1GBU!E4hWVj4f|hv9S=8&Mg1=D1G?n>~b$E1u}u7Z}C4KPDDVmdIOoV)5jy z{;j^0K@ULpk3bbkQUJWn)~^tDGxI4a&pBAD!9y#ox4godR`JFz(%xQD!?cmL&zwH9 zAk9vyQe+#?|ItVmxt%80@4NIIXGH;e#0@`NE@a2e7cZXS>Qm&3Cf^J{tHaE-l863E zVCl^7OgyJXUc4^g=HmNjushv_wL|RKRd@{yh|20;_J4_cTGk)sAU@FNN$;~zBPRvitoa;M*u(c5A#*tRqwP(9 zEX6n24*!q7h!b^LQ+uRep3^$`v}(GK!U`7mUe&u?RjP(JceO{rtHz}SZ{lhn(H@ltcv1B&1dn-}5C-Nb(FBxB5WZ!UJOB*_|P^Ma%B9;5Olxv!y z_-SxSymSZA_5|Px9jUd8Tpw5?{nL+?iX5MU*E6*Xmp>GWnKFPdXvE?GtRMb+wWu?v z00}L05C?KFK1eiO$qszXi}rPwZhEeD9~C~D*}k~$EFEFxiex%P@%;Qb!8Q0=KgFPr zZ|M?j)mn-uLG`!zO)Gc#`|U<*{{Jm-NN#igO#;NZc0Vd!8Y?vop!gLY)phVKp=&808x`a)ZK8hry5^@e+O3D2`Q3Ubp%Qi71uw-hL$i+6;Zl-?mO~A>U|u8- zf(lRGhxH;{Z(%;Wn=CyIUWK3|bE(A_b4R4A?Pe8z1X=n=d&@r2pIKYDA28odJ&$f@m({ z^g=?FS^f=p(^8iJbr#;`y4u#hCkBHMApzBoQ#pYwXjDA+!Qg7T6};(>8G7UGF~)1% zKm08Vt3^1?q0t?79+AYMs)w(>5k|^#p!Bd!DN-Bo}yP~atyV+{Pz46 zZJ&tt!u!5J)yvNC2psdB+zcbBkyGZd#|4tcdIAd-rQhlR>F++=8CP|rea8tSN53ib zK|*^Ku)jS3;$+*{?-`$vKdzpCwfdEP!qxMML*=}fX|MS%qc4F@?wUCod;-5rxZ%D1 z&03NxnpMJS7&IK;jy7>Q;EqM}s5;O+c@;r7>pw4m-#t@b{C(}vqm0C5-_(UU41Jd{ ztb%%3<-5;Xud>lodfsFz)sXx)K%`Z{M9gM&;~LDRp*R)q)-@KQZqVh5ljG21=5U&| zFg~T1=A?4HW>Cv!9KnMPeZ}lav492-8~74Hdn1V}|E%aeXJI+_H6@kSEuSH8IV34t zKS^eGVJLzkz)qPB9X~bKC!(tG(6=TD)poh7pUfTA8=mSpGmU&{^i&Jf0V_Vrijo&=zcMCi^j+9kgRGikS{uJ)L0`c~nG z!SG$>*HQin85wKu*Q07WKWkIY>+5GX62_EC3c-0M5L!56CWlt?g#EUx@$EJIt~)5 z1}i(de(Ir)4I^&G?scaUoz&XIRKJUw{%mw4!jX5$R@SWTE46>b)zf0@enxcc{}i+F z!~9U9@S;sz#SbIG!mfx2xu?V0N!YDF@Ix7qxPsXsO_((VZGw&*#VKtpoP&h^i_IHP z@iAV1*%Lgk%f9&iwVAxX^+5dZ_w~=#Q}pvO+CD~rJwb)_Mi#hz$JIf!H5zjGRM%YZvaa>sYmlRVj$s%>O+eqmRox zraRq`5ZgK+LOD-vvf^24gvp*_K=8~D428Qi?)gI@ogMeX;*OdchqqfR`cME8#H*(@ zPfwetw)!yNWvP$ly1v$l49eUh6$zEKdAWy7Cl!h4559fgPTG5eXMw+sl4Dsyx(po@ z;*9A2Q1jQnX|8H!r_>Yk1@b;_zk;4z1e>J9xERMLYj3kV4tJ}t8F+Hpr`qlkhm^;j&_kZVtqabk_aTrEk zjzj0f$f-BIX0>vTmu0Z;&E3FS@^HujMvA-2tI|pnY5tyiQxz-u`}@gLdVS67QA_mp z-ROPil6xNIQY}u}jHCdf`H9lsN(OJLjKr}P;nhvWda%%!JJV^k46p6uijfn%sFV_I zP`;A3`ORqoP+{(c*9*DE`)LA&LZ~KlXRHC{iwW(sgTGuQK|@KnrD+PaE9p}-EU-bY ztA%`CRAAnrMJPXf$+noaYGR5uTa6&|2wv%PYx8$48DIhi#T%lQmD7`=1knebM*0#? z4v0VTF@Hz68IjDIfnNKSI(NeOhhFL~!~t?l{yTMoUjRWyx4e3)Gm%-iVWtms4&S~D z$=%fiYKGBdJ8+$GSwz;vvlTyw*;dm~8y}Z^2$@l}YsWMYOA?6wM*-BX;)$uan z6C({TmuLF5Zq2k2wh)4*Ae|n1tTabwP`4FPCdg$SkU_!XBlSf79r#{+yCKwXH166g zCeM;G0VlDpdg#T#L{lq3aOM8ny@Iv5-=<9RHBLP@K&&gI?ZS1PK6z@^@C}1RceoNa z$okhLA6u3DY%rg8gXt^F)5ELC4@o;d<)Kn8NEm#O0nm zG=<#5($bSk7x~WEv+rO{PXA56xbdx6*^zpF-aX=g#)8FK&~||x4S8#N+arYi%!rLsLGK>lM0`?B@jbviCAss_e_PXE{myrn z6{vX$s|+aS6=_ZsDc9Aqu5l+Ws)DNDM&tdu(HUWgBByPv)9OYGyg+ZAhEydc9Xlfb z5XQMz*oeygx`K=kpk{8UbO`Pr9}fPeaC}SSNbjfJWABx4hS*j zM5|50)Bbqm)HzUL_dMP^W7QQ9|RRU*@3(SpQ05y1ZS}FF;z~4NstD z`wI`m%T|m5@RzT2%1;Ym&&pfjN&$`bD^&ud*Jr|NfeDG?nl- z`bSqJ{8I7Ik6e38)*&NfENY54;yg6v~-sd4)EcEGJQ1Cg={M@_Ua^ zn*kBpyR%{ca$Z`;NFDhQ{y$vl{y(@fv_PzD;b#@BwzycpkRe?rFB<#IL11A_+7(;O z`3r|4soYPFe!7#v7i$elDic+?%W=1#-!uzncbvn960!-+oYwlL5B?U8t zh8bHP0a1gv7B&tpBjW5!j2na`xob|EX+Xw;_nhP9i-T^DcSW`DwGk`FT!3I=$({C* z$*FfS7=k&sZ1h^^*IvUD3XX|bTy5y9_huQ&vacBqjN2jsteZTNg%-wEhO0^+uDEMX zcX9bcA7pw4{qk-Mqs- zts?{A7N*co$!EPU4xtyVKJ7zKw+}wRoL|hQkrOisFEy`~a#~GCYB#9&9IC%pWD`0{nkwPa7n&L?(>(fa zs;7{guvKJq_p&{c`axNHt&`Z~-o-TEXlel2Z|{`KfFACk)UEg^Wjm99 zd(A`Z1?>o{pI}D{$%$gAy3B$@H^GsTldV`zKiJC3hu&cI8@Pt2W&mB3r{mFiIm^lc z@for>U#~H-wmir{BHY`kZBYVUN$?Ibl}^meHX1ZWv+A!w{O?k~czaQF)ULR!XV2A} z;$p{AhWbzL(zerLMN4%PB!4^&Vd+SJn2H0b zD7Of5P&t#($jCO*J4}N}cN$rSx+>k7he&PE)Ij6IRTmk2;?2i#6s^cV9N}v>E|OY_ z^KQfQr1S%c;1g}TW>*(h%UaU5K>ti88rAeS29Rz_YA|AxgwdHWG-KO#O6_)b$1`5m zr1T*znHf)-Bn86lkc;LDoGN`Le@82d2+|}Qoi+*f&qf8p-iXeYrF2bP>Z?)-gV?y5 zYfjxUVQ$aS!%c^vg;D%38)$-f3@I>_7mwS|Nr0 zm&;PAM_+3BB~Z%%w$-NvQ)D4OiBk$>ym)uRQzys@DX6eO6?@I9XB2IbbRk-8yG}vZ zQ6zaFv(`?uf=DgNA742t+=$)cI1XLlMcBEz1E#7Yp7xN7n58XbOpv`lMxQ40`>+vnsOq4LgW$rb+e278bpD-f9yFHN^O^e+6BNJ49TbGVfOcEQfW zxjI^WWg0Qo$myn=;q!q&@gZ9(7$=L10-|R|6Y`hmpl9yyC%HxDrH=OLo;>C9`f|gy z9}XHOkq>QMEd(@HwFFSvE{p=F!FN^S3B3V1wH4}?`C4}g;=K_agXcvw(X}!$hlhVV zaI8VAZynLPdv8jEZWhY9qyiCqB7gkphRCiSQh=Rh+SN2*xAk=^1 zl&k%sFqo2!UQqi^PyTPbgazm*mYd$ytnDmEL$b4-24hF9T-k2?3H`V57j2d;7O#gN z^jtOfM@>p3Zf!qIOhf+XY#3O)4re&Xc}SzMhX@`W-O=;BF<*u$tUv`WB;^uxt6fayD!|dgp|0gNAF5aZx`LEdhWH`_L77^*3o--gu`4Z3}K8<~03;j;hfx!H%G?CPWcwa+3m1ph@I zoaYi+@V4Sbv+s5Zb?_h2lMx*TbLThfdcR*+dHAt_a&LS$Z=UUKJtVIUk}waQ^pg?v z@X3I~-v7SbPjGp^a-JjRAnuR|K=yS(=eM;4XYFe?wHY`nwN#j zdoo^zVt(o|fJo&I9qFR^{8dZPEkF zDSZginkI5i>snh4H(k#^$uqMFN0z4@q4BevNH?>>cqtkVRVgQFYmp8u_^HbZ0yV4*u7Vu(vq@pS)Nw z`C-(lICW*mw%!S4EFQK!?PILzqS#?ZeWm}|b-i>8FBeTPRryN?#ts%|#7sNxdtjqC zQ-vWL0;KNz`N$D($N=q1kk+-Nns|SCv1xEKMi@|AqZ8w#v{>qsQvUhQyn%lr@io{f zkjD;t7w<89{p$&i14$5#ktQcL*s*4bS9{cR6M1W1Gx}Nf+SMJI90%jm$!J(^n$I6p zP%?Ul3}n40hN@N4?L*yEN6t-I;@ux^)O_$>B6t_VSs!peww;IKWm@Y~yy-$HPm$cQ zvz~508o!S2@(1wpraOV+dUbaY_wH?&pAvn-uZw~Jk7&x8!Cmw5SSaJ8k-q4Ad!;w3 z8-(|THN9Qo$zwLcEk0%Zvu+Sb5n;|l871Mh%*(p01ICHO(Gu&Rh^MDMjdMyv`*3&) z9V4joArntB1i%*4Bg5mx4kS0qW8a&q;!Ocd1(DccLVO|qgsZCJXlhl=DUtBvcN?rC9N3vmq@}!hn9So_2S>jiYYq{P5rd zdJyTwujF-`6UzaC-{wf43LMtXEkoJNcnR#*J>46r0r%c7nIcVNp(-T_DyVLlPFDIM z?R=eP?&O0APQAbGJgUUAyQ_3Q9XO4FwR_DxWyDj`5s#(jFISn)>qQQJ{UW`sAN|n=5?at# zwWF0mU(AHDe7>9vfh{+;P&+43B#eEB23)SuS!Q}Vd5#u}((HGO7T1~PzPk;Ct-ZbK zS&LVM^hj3G2M{UgagD6(d^y3&YOgmz)_(xW5R2$0u?G zu*Nz0871pQfP2eO#Oa8=KbE+%gmj0d8nL!gk-1gjyOdS!!_dyWOr0GL)$^T1NO`o& z3y7A)?wzOyK|)aB?Exbl*4^*>n^|LHyKPtLT$Mc0|NqyU6$ewZxLSen`w9(SB~MD~ z;FX_MFJ5fMj@}ggp7PwRF6FV-q~#~DBs~jR;=Zk9n?+?hf>Lr4P0XHw$9wPVa5tSK z=DmQnSdwTjNDjthBTlPAH?T4DPimigccl$&=H8Cd$xOqT&SzjKW?mMr+k_#-+11$l zB^TT$T!O)ZMmFO{{wZhfL9SH(#4;@2f$s{LC!e%2R3?N}@ZS@%Yg=5YUnPupRSiYFr8m5QNX$_PdLE*owJBC*li0WfMa2bw+L1(3ooE?<+d%y;WP zq88z~Pm9l-SCAGbdF`d!rH&)ssI=4iCstT|2xp!8GDYbnF0Y!$2%sd?YcC{%O`;6u z5!uIM>ko+BjCHNZtJE|tq++%0e4M^sx2)sG^&uN8HLJ9RG{`yS66}O{4^v)}acMs< zlobWK>JV5|3LBDUKG1khq|Eheh&%mr<)JK&|IF@=QA@7#+MkV*O#jptVbReFvLiZK zAs=-&f1)Ggw{Cd>RB?C>d3S|Un(B-G{W@(#jVvDODWV={#pLYZw!srzCu_3abkrYL zI5nxTh&e2uV~qZ&VF65mwLFfTdVO^=)QG?W^lCK5ONa z61YZ_iRVi()Zg6@lvu!JLQnXle0HOcvJ_1R#iPPgUMToqM?kArUm)qhgJ&P6BhSH?~%5ZMc=9msvejqHJH7xj0vrcOMb;r z6o`UAo^`p5EO2;1VjZgA_IO}#Jc*b`MW~S5=>6h#v2?D(Ea?wlMS6|3pHJ)UkgO~p zSw2aG^R&6UqSSjdSu;g-jCiTM6z;VWE)LcgF4|l1Pr5+(LaeBGIJ-MHKQg7&OE962 zx2C6YeKejk|Lj4h-XY-#V}@va>YE^H+bw{wJ`)f%qzT2AO=9+4Qv{gAKcL?$Bv6igu z!4jXXfZmPL_FCZr^L$`w1fvbFkdgY+%ZG0iz({AYp4SLAAjLuK_}LMfyAgdb43(PE zueCe+;47;&rh3!`#dBPPnodlhh^rM)skXtm5#s(MoxRXA(Me) zjc%K?VE)#p+`={78-KzZwJW0fneK;5&%jUm@_iL4?ZDZ{ zbv7Wn^`~D3d_nczM5}st=IDPh+L1p{julqzyWX(0zY(bkD(pMsDc8a_lOD6S?$e@( zAbA03LgT0#9Q8;Q2RO+{I;0}b{nk>Ea7O@w>Pn1P9vQlotd!Atqv9$LLFRe3zR$r*&}q zBVJ*oOVBC^6pKnpm++#qXUR7Ne0hSw+g&du&((e5pl&Ky4acV2?OgVccI?%xnI3C7 z5jC3S8Ws2T-V=VAmK6YF=ASop^oB42(O;Qs-BJ}uK|dm@lJUtOMv$)4b5*OX$>LCM z>Z*rsj|AJoF4)*kox0rdPV3reW3DMceH4C~G(tL!<H?Rc_Knm3 zt6A0m{4@vTX1i9o6IM^oQ5wxSf`w8ssUt!cZ``uQB)(d~aGAB4B~n~ZJnXtf_PVuL z2(}ag=Vm>&@K!!=_QFcSQUEdQV65)gNV_0bFA5oJxIk7m3ZO-$I^B{t*tWA0y2{;P zmXgwnQ`p_ZQmo0|?wua5IJLMxfCr8Oidq|dvc{55P9wZ>uR7$^{NrQjUUR)0E|WD! z*?yP9sgwF(Cwr%h!(`38@>`fI=0?O{FlvPIIK}K#Gk677EtW@Tmi?V`=A5wrv#d0zCp6~X@9!{26z)Xm16F@xVWaa9E?ygfivZ*Og;a)XY zf5={S&hzPff~L8dHeGYGaDnT(OWEM`t@6B)G?aal46xWC zz)Qd8C2rV&{?d)4Vf1fXRD#Gt-teTsd`Stc*Ogw@c=WZapM`4D5DV^wKm zkpQBy)bn(JSR>-9aWc_oy*bEQLTGC+1E$oQhrGpi$G%N%+3{)3>4#kg(s!l*y&gSM z)KmO-7Un@syZq`mj^Yy4g#rTpz{P`mj_t9|H1~z?_ML>DNkG06*?(A$uam9RR@!_T z)E~C=F`H+{$>En`=-{k`nQpDr zbS|%)ouI#ZLth>UfHJv zh2HJvTn0I&CHOo_hU5=EniZOojI8F5B02aU4}jib$j*m@OCznlTfiLnNV@7a>84&< z4tyg1fUqzHWW2L7NPVI`UI=8W9ggSvIrwf&l^$On8POU!2^ z`BPE;goN-4j5_j z;y<%rM_^bh!X_=~eS&w3$=66ZSM`Rh1U%(UtU8Y>>=Lf_c;Ct(f5cKH@@Ecd_KwN~ z$toQE1;1AJ!MDl5^^8t>1be_i`N$0(bXy1CIg+gIU)zp<0b?bj-RIux9bp2RY!o!U z{3d;e!!-uqR&WnC(#E)U#r76(YHFa)Qy>A#W1@2-SI`zId)w< zR!qo#>I0+@@s9t_D%PeyK!5B1g$~{SW5ulVY_f|ybrQ^Bp3oLDm31F`V(Jcz_Qg21$xqGPRFWsbVjxUjo8Jq^ zJauQ`M!!D?&CA&Kel5}9Q&O8S^GJo@sjttB4wsy5mhVEi*UF{5dQtM@q@|GZ#divp z(^$0D#flOv($6F4OMz!2Q3SOH<-NBxFdx7n$dnEjj~ zkjO4CP$t0hb!WHueI#}DUTnNUa3`le`Y}#unHo+jW2?_A@HW#v)7-Zc?qaX&S{YmN zqipk1c|3 z>L(?V{^3@P{4_-sXFnD;9)*BttKOv>2}=zqtC1r(}`gyn>-U*{-W_5TH$%-xo(8;)_+8`F$8yfI;CR8ZfZ?lKfq|Y$8n?S*(d37dl z=|{E=-hf8Fw+P>H=dgeoS^M=}OI3$v^;)5LCFjq)sS2MdypZEclY1)u)XPcrI@SV?xZQQ?VbH<00QI-+JQT_&fI6#JMab&#{?6YCmr3MfM8^< zR+c49{i)4ps*>c)J&Mu1jmDBpxZ~Mx{oN|}HrDb4w|O|NkfdzDwAXGfO1O3+g-Dmf z+rZUx3eN6v|3PoQKvF*#^&H;Me-A-$50K{48bPTYr80lq5s z-o}?Iw@M~=tfi7q#$MU$j8`o-(ogM}p0KdrN#?)*?JFF)DRpgCz@p7<#D9h^KjjtN z5srK=K#Nbxb+?G{wp5 zH{t)DU+Cu)UzrRXohxUp!z)fwoMpmo*}h>y^ti7(u`*6n^PbpnW$)!4^cgiLGv2z= zHJJe1St1G3z9YByT#l;Qe^He%C`r0`w7qST*xoA`avBXj%rc5y$lI*Z)A#E5=`Zv5 zD8~B;(RCNLUdOuTL{~_ReVWNOK`t11)Hap1AHJ7P1@@msg9VBSB8*hL(%iG(Ncj$0 ze&$2ZkuNp{%reG#=+jJ>UX2FkcP8&ro}u6AZH2Z=$l)M=nL5K*2U>(icQ4X#>#FMU8U3od3tis9r_>#tS2zlD%e4o;YNaLi>xN^D{%{i+T#~a| z-xuKgvgm3kT(jQ?XH&`Q;tOBo61j1oC_yWc+dx8eI>JbZ62pEDM+u{64s?KU<=MPP??q2qyO5Qj-D~vHt^avZMIz_vV%ak_RGz8>)->R=@%A9ivKl#7f6!|aI5>21|@Kf+O zchdOzt=X_Q=J$}{6*r;IzVq!EDWbW#QsqrNdbiwl7<^j@@}c^=*nMp9U+{I%8!%S? zVAW>jhwY5HA?oe`xVz+iN;Zz~a}9ar71&-V`@ggwoYIilxO|oXHi|pbYozfUg2HsQ z6mkj9dtzqjr*Y5W^HZM*o04yh{K}+z4?Bjg;gg&(pv*Y?p`n##rJVpu`UHU1`d7U?i!j z>Uo^awYtpc(BGQQN}!w~20?GTI~~WwSyjsboS7|VXQ)F6+i~-oN7gaVp$8u4t#Cw}D|Lu#DKon8NR?NY zlr-y+?f?c+6(Mg4IDu-}M$R!-8h#+tTKKmF`|>@+={)nW*F%`Mkpq>HwVbUzEiB4E znV&%Oq)gZ*lL`6V3xf4Y?f&YQ+iqIw6Y(Lo$ zA}(a;EjdTO7W|0$xu4%&ez(G@Vz>NbyG=~b2Q0p79#rdWiJwWQde+ z+XPo<`y%2qd_7W;o9+V5cO;oaatcmOj`zTnOW%v)+h16dq6K}CISb^icjx#$dZw(t zOR?kpx_d8|pYrCyRB=Xw6nX4jx-$BKv#-U`rsM7HH&+iFs|~Q0p=Q>#i(A2zaAfh< zyFQ}4BcVyq*ih-JjiE}f&rglY7$uteWy;o0kzUfC1Va^=5||ukqqClhQUc99EfUeU zRl#j@ZIkaPX}K^8(893mWuR(?Y#v>|9r^(Xkyo@xJ!maBbzef65n6j?6I zODY%G37N%K5Jr~jxN6jBoVq&}efQODwY*ZV>gU9OvBUrTf~(KNvX8!B5C1(h2b(A+ zZgf?$ou%yEH1~iOeW#aR&Q~+1r>E0+D}T&t3AESHhe!-9*Fo2U-K9dZ=D%W_SxEmW z>aH`NnpM|z7_xNEO%9f&jEoSiLU>9%BWx(@PTya!OqhP@2*SbY zQv#5lwR`y=R+GkM9HxAlFbpKSYT#)VAoWc zVlW|UY}(QPfv!yaqMQ-@7XM=-S@YLfW`Zh1#ajH&a^QDf(m)TaI)m#pwg1Q6jqWnU zF$^dF!Duqq%idER3n$8=sB$Q*qoXIM*s}+f*uJO=&%>U(G??pKCk+DJ=c7f#&UrXt zGWepa@cwvXw6HqX&;2@U`yUBM-QJ7(-EpjMHo-L=x|w)>o{tfcC34!p zfL$)t!3n9Ee=^a`_7-%-F{rk9&18a6G*k6zoqY4Og`}sQRJFia z0PRd`^_aVx?NV-A|6|WdU@qx4-gYN>o=37dp#3xqsQSfh)#Pv`pgOiep11XE>C;fpBh~Wf$sJR{~b%57}EI!$iepb(nue4`&00Pg8GCyq7%peDzhrCU53HX zo;@pN#L*dl3lp&!;>D4S-~V*-vSqa5v76UVTsZkrm)~D~$|KvC#UW*tc1-RVk(?3# zDkUKt@2Ex#@qze&IF;&8^j<%I#yji#$x~{d&)>M)$*m9J z>dfww3bib0Q(lpBj0>}Ab9!DR9$jqzclEdf=k5EY3R46tuE%2kloeTQJytmIR>cw0 zLHC^=%PK}A(qiie`A>3eGB~zT{00O)*p9>+cgjqbc$kVSYR6Wb?#;NT_)EDET7oMh&Mw^BXs zF*W=5=dL+!;$<{UasB@x>dnKUQ2#gJu})=~O2#r6(xDC#GWHn~QkFW(z7xtw7)*>U zX~>?+ZY)KsJ+h5`P+@Et!q|;{hQZ7q%t$P-qmai)1r673Q`LPF?jhHSg=qv7 zTfwogN`JR0w2JBCIyB`tXraItTrlx0wWzy%V51j~_BLsp#j$ymH)eBAD&_C0cyE%P zr8XE}4k4yEg0IuNjWItbo;m34e@1A_dF&5Mxv(?wsfJN<8SI(2pTjWRP&o#=&}1Pc ziSD`K@b-$JG;TldPE`PyBwFG}d zqIxs9r+3qpOZzXJO6QE=Uf=hvzNMzvg8VT zjt9Wpo0uo*ZJmTjo@>Mnc#DZDGks4 zBd@0-Eu{F|7}u@XJt8@4jc;aZiy-)$g9OzXswD`We$c!4821_WjdJhE$kG9SiwxtQ z;1y6YbQ+fge{wWQRH7PYz98Mz3nr!wPTYWQ{6Gxd9_&<}&m#Rj(8)2`AitcJ9C^t8 zW&{m$6{@8tgnCymWZ)J5e%RTnsJm0&X`6J`Y8B$gQY~Md`7OR@s4N(`+F4pllik{t zxk`NZXECjFG-!Do+>y=@NO1B0(s8!DQ`KkAtp#qOLjtUplM>--Q!&6i-?v z=3)%fcayxa?EbPYReCoVwPy!cBUc0b3%(;5&&*|Lqh4xJ*G zZtr|U&lX+0O6qI2$g7n3^<3KSoG)-ASX4{GIL*61YL$Sw%UZm40$a`%j1)W8Ip(PvUFSt}R4m8$hfeVLealk*p1s=s?M|uy zYGdigU&bCd{EQXcRpCR(c95+^fxQK1-+u&(&>vy zl^P2ML49^E?PEEH6w0Z3>>Vy1T>0v6bymPH<5s~23(0@`Wu13U}2G+_YbMb2n_Ga(uO z8ji#$<`0);y*7gJdBDRWq00LAyzv}QD4f?}g~$^uTrk)#h7daQ#Cy1|3hUxZ965@9 zDPYk2l|IJG>5MB!5L-NEU+?v9va23eiCYH5V4yGgF?eyIv8(piM-0q$@G?j(6;E z8(kW`B`!Oa70Kb4h2&=X40p2mm^UOu!L*PE81J8T%|okoIkcQ*`i z(O3Q*cC$3*F@U$vq!{xgYKO8BvTJ3{M<32 z{AlU+{^^S5`?K;}ig?{O^D+(M{bvSrN_JqB57mHJh4bnjv_T^g=3G zZ70lXKo6z&SeEzVM$wTy<&D%-Nf3KZz*$QP8 z^Lx@4>-zs{rsO`FFf*&YJ-Q4oAiaVoIpE)K!?=|$+b{rjy{n3YFMR!8Rv9!77bMS# zH#4@5ZTmO~d3gp>eo5-Vv;-0DVQjePrneGP`KY(-a!Z^0hC7m^sd!w}SBG)IoLA1( zJXhzY7MZ&_d81ZzTC`PyIT(Ss2jsGKyzh@Gf5(tkhA4wScZ-U48Y6J4GG~s*6Y+tR^(uOM<^xw@bE=+Kv(Hjx`SegLPW8$Mu$pEwka0&)$ans zRBz!$a}(QNfEpeqa*LAV^n&K=do(*6H7Bp_r33e#D;R`ixU1`)m%8`DkO0tTc?KQ6 zFh@ECxU~0N#uf0AdP!$%y2`#*rlapV+HLRIC7t8yQr(*%WrfDOVL;3Cq3&4Sb}Vre zm(0WLHO!coK&WgF4H&}Mw;f~d&#BXFFSELpBUK^>9qa-DxDj3_CO4VU0Nw6;JeKWT z+f!20sg7Dbr97W5S{hY#vw(C~U)3NOc zU0A7|UBA;~QB4_?k#F$&EdCq$P{_*SaauP!X$qwfT2&FuGfpB-br`}a7uPae@Nqqp z)L@rKXL^>RLf(}ekfWjqlB&7bJRFW3;c9a1IhLtbpnx{6S*OQE!RJ>@nB2#GAW|#ZP2@bx$ZP4TpyFHJu))=H_)oSv?M)4_byV-LQ20Rq9D0O)Zd&$oSqQ0-UIlNUJd1Cox+q@rPv;I?zxk^k^8#xj#3 zmd;#7(L&Djg_#NT*`dL5+jE?wQ?m6&&TVARf%6K^{$#wKJPgJ)R`b|v7>MLZ0pPpE zXN6DHR07Q=Dr+@Q&~x=0mLKTyo$?&|wb;JQ9btF-d{Y7kF&{vva&F{)!um~rj{rw#hYLFV8*y@YOv+!`5)X#P-3p>`9>HR z=}tOr%SU$Qstfe;@EGd&+pgDA1fJWQmVQhPu1@>y`J>$nc1*H4l)H-d<<5$`Ovbmm$D2nTbqpvyOndh#e4P=u_rmq0e+HkQbTRso+qq^>mBwP?3fON`@Ojs-W=;Xa$Gpp_?D+EDP8&vEK!Tvu zqWB?ItXYc$Mr>)>w^aoY{E?0yurNsa7-yDdF=)K+BQ(TYx%tJ(jT=~kt~e_h*!(Fr zC455$+UK-YCHDFJz;LTdH+YGYuNh3&$ zgakZAF!^82$13zSeDCOcOpE7gxypKei}9uJBAX08TrG6WlF*G%Y79(~{PD2VMNbSmXtD1Eh{Reia)=10_3{S|g{iv-3b zAQ>mxB)sq(FS+p!uU!^)FExMEF!lIjnR)Qgc)e&*N+YIYE~oM4!J54DK66>U%@|35 zZ3NiwNRHREh~92u-`~4LA3bw`xjZ0VA>}Sa?s8MP&i~Z(@EIz|1~GkN7p-pq2^cWc zT{e+I39BIqeLS$!UGfda{PM9dX1-=RIsbb|>o&vsR~>L3{jNZDCGubCH`p%r~)! zXt-{nE>gzVkfq0rHJ3;ADiCH%Y%|ak36G`>q2};;4zEak=UpErkMS%2xzkTZy_ArC zo1MYT?i{|TVGZp`6Da5k){&61Bj)lJ*i7Z{(-V~)XSsY2Y{HLUlMB0^+WHGLO%+~O zV9n_;0nInZF}IG9C5TPEXwSSfs|c*t`S5+|oFi8iAiKL@ZKPE|+hb%xW+o$7X%@RY zI!V<7ib=dCG{INzY?X?jX5~pE+G$5g{`hh!{!KFz3|4}vJvs3;l4{+$*u|$a(yCj6 z=d=v-8DxpK38HLi!!u3}k_ zw}LKM1I|fvh@Kd0BVDBvw|}3m zmFAYVm#+2!b-njL&%*{dfvwFf%u4Ow*4PEAfL*zQ^|kRa?v{+Oo1gnN3nikt)&Q2X zKR65Ow()VDT-(9zTG1ipPnWPw;@&ENl{Nu4WFEed5H1hdzFK+kJd}-Hx%MMVBZD1K76{? z1fyMZp8YLGG%|Dwg*16vB79uPmwvTU1Aa9QxG)%zaR)==>(+}ju~~Z| zytr(`mfSV0P%|aonQ!QH;pKvjPYN zHE2vX^`8C#Mt7{=+xenOZwR5^+(r!;Pe@6s*tyK-ZckQ6MRdmoltg48=}F2ViQbWp z30VAsflTP(d#eCFcz#4E)c?g``9jZ}4o+lU(*4{-m-BeHOLL=lp4N)2&n_ew$&L@6 z`Fp;kT)2jAXraXe1@ls#x3H74!cc`RXwN+HTj+%Ybi1LphYcj19>2WQ8@^K+P`F)z zSyBip+1S`X?tH{^>3Q$$p2>BXv`xDTr~A_8uMnOIoYw8LoBX2cyM8)Ix6;F{BTM5* zH%%Yj{Boi}^SakzqBq9cu?SRKV^TK)cG)rVFDvQRAayecu~X5Z3waw`G=oDbi0q`X zymzYrHLE@Cz&{_(LKR~1|LT?1TaRYBz!_bF`|}C~L@-oIqFVTHZiG>(pEsjKU;Cka zs!)EbQkGxZf0k!(t7v_rP*!*M61)@Gt*U7-aF8!av2y%{<1)5_I?+ju&EB&7QLrlF zJ4oZAzxxFTu}_#OnUzX0oj>J7;UnT5Xo#v*mpfeF->2$Q%U9CHqYRtY$q`v-J8sHU zFk%&WoE&t(Y<7uswPd77QlyTV@3BN(+qakTuI8xbck+V}1)ttk@cSW~)D)_lS$f}7 z*&NTR2fppeN~1lHmR;dStjyxW1NIbDujWkDl~U482+SlSs&~44H(=cF)v!%Knm{Pi zmWv(lM0qb^P{vN1N#1*5D^vPyTPpw0#_iNcPAgH9p{awcjU1JgS)KB#mZy(4M@@R> z8XLbRW=7zj2vmxMSMoXhDM#%$(3WyyqqInAsrO9|h=F%|)Hij+dyWtv0=8u!=xN;> z{ArnNO*8!0e;e`zrZBt~ZU};N>9J+*k5ryYfUJ58iMNa-WYs3! zt2Abo{tB36o3O2FyAv4?OH7^o`jfjYnAHO9?z#2AKG$+Td{SxE`zV(A_CU(U(o83q z)oB?LG(ZUCXhQeK`Yi%fxUGx*>;5tM-F?kDwL+;Tn59l4rl)XUL zyB^OfyyZm!GnLk{A}UST`UrxWm87L4nB{L6rP0|x8B0)M%Ja-Ro&j-b>SdX*3!1EV5aMabn0uIw>sIbO!t!Nxtu^gz9xkTD|g8P2w#AZ44zK8;i@3`Ge~< zPesQ2e-D|M99N$FQ>pwj5eRQH-uP}I7~%%HI<%vO_~`HF6jVt5mV}L%`9CuY{-2q( zKC)CwKOqkAdex>6aQ#d8N`O*Zv84Xl-XPF7z-SojXD+|r3)}eWiyNzcfS-$PZvmCu ztZ11;ys+huKaUJgJGBARR7^P+9l0I?r1??{*s32Y=YBy>v}NHQ3=3A)FrG&hPG9(0 z8r!Xb0bsy%jGTycf_Gd0+n+59fG0B-YdOR~TU-pKi!2LyeR;8HKS(5pEn<%|@5w(u zuHS&p6jo{izZLU*dth=o<5Jk}gr4+SP5JtIrM^?td(@L@oWD?|Qm*D2`j_Cfz%IY6 zZ~jH!-E+dDFGYW`BuloH7a!}xBhv3gG`e?2Yq)+#0IUvePw-#IB^&uP?p#%L%GS+_ zpNolLBqs3YW+$cdfY@As`r9A2J8ka~j$G3>M}MO?zMVps zq|zj`s8ogPp8939`6MQ|e{U0-0x(&9`olLm}!^jkwQVygQ=NM7jQwUA-{N|7Il|WLH9?xm}Q4 zC9pZQm*9Rq4nCTx(C4{e0ZG(|{3ja}&%+vaPWpMw;A})mUmntf5L+XA|E;kwdH!XV zjFeanW7eCVR2NB`jS&i_D~aqVF|I|ejl{w8C!7`k$t2V27e}Cz8=FM=0^REdQJAMm zkvD(-mP8TuvFR0#R=9qK=5d(^Bz``r&Ig^5@>8i`e_zsR) zRk9e$&6+b8UbucVB-n~d6?~EbjA~+g9~rFBYnH_t-+tP-X0y?Qm6*0lZGE_%^X>6H zH>d~s@T(y76ZHWQYMj_Y1UoqGMr4+ma3M>m*g*G{`MWAu9JK;T!TU0;JZiR)<@l=g zx!cS?|84l>)xTEg3IqL!R~ewb6!qFQFQ}1;hX!4$0$5ZAycJ|I6>z%|$%dek`??C0 zd&Gs^z+%x&7gU>A;2ViA%0>IHb8o{T*H9N6r_g#$#hD3Qdhak8`|!U z9!+h&Bwttp^*a~w>67G?ep#eUx8s=ObnZ)-@w7(4=+aw% z#{%YK>_k4JEa2n3{+_|h zQLJ|g4b!%p$6^jPb2X2jN6tz?cIUNQ1=WTXv_?lL*^2JL;qJYI;NwG4>n5!%!`~Bk63PfA z;i;-y#r6N+K>grX+sx_a`bYcu5Ms1v1vj^aK}>jCLdoEOoRmIy<)nK-92n3)D8?D8 zpP((60hI2Ye||Ey|5Wspo+}UWF4)-sX7HRu?%`cLOELXVI2j*36;lDqxPaOXXz>O(=~A8RKYl zyrs3{aqy8mp}xudqUHjV>H1{)X%I#q4N5hqO?G@f32dMxKi5Y2stiNQ;JKfvP$7=Z zc_qoO(QQ!S=5#w;OtAMJaxRuE#<$V^Mo!sh zc< z=6*BWLmP|v*hG`7FU$Jw9yUU^ox+E`S?!>R1^Lzkv19ZNFx};n2lY)?<)g$j(HJvZ`p}YxH)W+4><|r=$qfypXDV z;a2rWY(dq^=Rt5o*JZzjdzck90swKLNh768ttw>@QURmxDZCL)oRPaG62W&2U*Pdu z@q8_BMh|b(M}VD(Cb9f<&!h*katrNTVDzqWG0Z(T1hrCD#CTTaZZ*uXrc9&nw04W} z3GSRNSX8Jo9#A!Q>TIoEo0fYSB84)r`b?8fk|;+6FGy(&4fIIabP7+g(XdWB(ZRcz#sgHysM(jfNB+1OTx*z=yx+PUiaDsey zH%bKZd%ZW@B-s1)t@*4jS;QOQIn@MHmcEtQ&bekri83Zb((7y|gWtI}h$xmL2Hn;U7bC-qzd=T^!h)ZRFjg+4LBrNDY916c+(QG!Q;QTCp19M~Pv z&20l4C5Wqit3{U~89T2r2_Rc5C)>0wCSU$q8O9JEkTRo3G0KeG<0KQ_1aw=Uni+j< z!f?^bV!Np|WGYdNnlP3Q3-P!gdIX+7=@|k$aT7yFYU*uxl9@_l~h}%&=q?zNG zRhpg>y2!bj7j=w-{MJ7k?Ysd*3DH&yh?z?l%chD z-Q4p~(6jjIdHAsb_9xoVK`;jR^Oj>qC<%+%NzzK|hcrazJ7NF&S^`81NV5K+Iy<&Q z^fViLDi-CB-(1(*8zdO7i}b|4XL;~xW&d0}p->eQ*?%y{r8UBJnH4M`mEOtZYz(}D znVoY~`N~xhsHvw#iFKBs#oB1ylXw;Mg=;9raZRTgpfFSqoLLNDugAJ9V5963WB8EG zMMj}(_RyjrEo%#FY7q!~L=vx4A@qJMhl99Qh-aFJ=_=8ofg_(@f`|xcAek!&b#i^A zB+tCnr>&XYseB?$=ku}@;^%X;r}AM_SsI@c=9_W-+6WhJC{r69?L<2H%T*&_4oNCd z^!#B-3@j|N47|8iveY=pDv{-Jc`OQP1V`;IAJwQ={tu6p{&T6uiwWk7chrg8(rvR- zD?Xgug74I#0k&CFoaX@}OX$EK&)U7-06{>$ztr^IN_;zW5m0L884WIkRC1pm>J za?{OS(7bpjE^hN*07uLNOA$b=X(24;TG(3;!1#GtalzbDmbZuHvcpy0>wEXBJEhQW zHo;~gy*GLH@AI*~nl42N&GJxqNMm0w@_uvhw4oav3kK7AGJE4(y49;So8DB+_5}-_ z9*=pX{Ww1r&@Bg=7#jK_GBnY9Tgq57#UBvHKxowc5)N_x9Wbfj}-f_C}HJNfs*-}K_2-nm=S*fzBS&uR-bG<#jK(PUXzf?j7V1?V3*uk?yu` zrDwt^Jmgaj72OfR`TidlZ!P{f7zx!qnEsbg-u`AJk1rOh9Y|1iI1XydT*yNfzb3yl zuG)J?-XT%FvCMU=cvO@?GPCKSY;!)*dv4WOgJJtBugnN^D5>c7Zs4o^;Q zZ)G@1w8@GS>Wt2^s%A^^NA&gEAWOkG2~2yacAJ7jO)BF%41J+2a`6h&CDIz`za#2U zogC?wgzADXx&_hhZ44VDP_sE$J5tK)U)oPB+Nn0bb2Z|*?5efkV`B}cKc6iid^^ZzxMe-sb{Me^_fp%jN94d>P92;s&XCdkex_D61=UcJg<3Bj5^ zm9;c3#MTc(Db*#L<;1n#Ot-%p{*lynGqD|#_@3QDjVXvV?;^1UakTYady?&=nkw&t zcmC}6JzJ}J1-dt$T|R7!KwXIXGi907Xz70?1Nd@I7*m-LTr3DTG3nEp2HRabi2lr$AikJ zaibc`Q{v$Is;()B=)I*(mjeGrySJPdl^%0e03OT^&D-<%0VcTH#G*?fv%=;(P~Nt8 zF62G`qRQ@jg(-@#YlRBJDE$l%wXcrHEr@S0gYvJJmRR6EorS(6LCEd)hhZejADTfjkwRK)w% z!r%rPtfdmqVJKkmh*Ic%{u*#V9}VAvvo=nRj;KI#_-!$Avi9J9jzCCd47&OvMo<~` zL2rgWhW1S#*V~JZ6w=(E9~2xPS^~;Ff!$n9z)f&_1$lImhJKM$Xcy``Od?jSHp-(^ ztG>jc8?QD*{MoEuAuTvyetQ1kE=qtFNP1ek=y+J^)xJVy3NWIzVQXUxCdhgC^oYD2 zGXZxlY=6SL?6P@^g1J4lzn%g%uVtD8dgs&iXrHZZjd;~@)4e@3a~F&zXeMZSu!VHftF zJRFDW6)xmNGV4Si(K_@Zj{nQl!-zu&>e{fg44kboIKas9^sR2Yz2^Q+ZSqcyQBT#X zjD`IxDlv)W-hx%?f+VW5hjM)g^2diW2Y8JhJlb_b%yRy{{D;0$JZ@GU*2NQc9)xFn z{w7`eT^_6`C<}Vunc|*}b|3liO%*Cpbl-SMYcx)))I z6KG{%(I?AhLcK3KP*ms?R0ZZL|0KabAz^3ZzjS)L9%wLW%<7nt0r{Rx{{-@5R-1oC zdqH+SR9NoS2F;oj@g2GTq;kLl%$@l1)^=`%8OMGZT~fOeXXN!Qo-H-t?yWnNsGFuV z+_^z7K^$KHay;@EKK%BJid8H>UEY$Y1E(9xH3&v7B@(V6%Ie)?;eFS}ecqb@EB{hl zU19g)%&BI^5gD?_f5XLV(8XK_`NErwkKuIwxKKlIQmhC~jJ~ z?hdIcd}2I!T%y|pzI3$aup>OlE);FkkeH8eUNNu?X3Q~}%qjYnDC6G{w!0K&Yf}jn z^`dhgi&l~&olud7VSmZBjsP1Org9jB&k#HezH38Ekf^L|1_ApXNI3@ME<;4-GbI#p z*`XaKv}XsMZQ||<>Nj+QjY}i<9znO}JTzbt?5-e=OMlYg`ds*{wwxi+9<5RjkI;)m0lSLo= z8$A(0HyAtum`3AFwC!QPIxxSop#ct!<(R#ubq~I1#d5dIipIU1mT)2O2!ziMMKqiw z+@YSzY$M^QqGcgU~KahMev&nq*zyiY{}9KEB@ zr_Je5IkNy?Vb2wBbBgMpb@d(^8*yq&Fba14p7HoaByz|0qS~3j!SlM01x86zfO8z) z!MZZ5-UXbSEFP~t4iuAl4v?pd_8B>SvG(*@!WGM>TzV9)H~MOYhSq#VfVA%0fK#IE zIZ%|Qh-+VJRG>YwruUuRX&xk3e_Lqb$>@U#$X_uAd@j)t;*F5P-xoNdKYx1^6oUs? zAOzM-{~^1CZ+=yl;3d*q$0+T3`oiBLjt>vy{QzT6<*AV>k^84{3~dNPa_&-)>gsrh z*zY-k$Lb&6LNfrmK0OvR*Llyg1{IV)PpmUE{S?Eh>y1>J(||6EQLFDVOOx7^9IAO5 z<+wLYtfgn(Vx1|54zMx>NF}2yDCD6~;^onvpti|u0UBAen%d((!3XU#5AkEF5i2aK z(>WxuPX#A(&oy~nOKYvxYk4@hqNkoMCGIdl7e8$xSXj|Yaod7(J+(Z>n&4XEST%F@VOvhzUxT0XE&VHH&qsaQ3sPs5tl(yK?jDAN22(u?b zbzimF!gs#DtVaozZXnsHVnOvzkDtOPE_i~0lumV8VmZ>D zX@;2LuDSGA{ET5{Y46y30AZQkll{Qrg=0b>^ivw8}n@h>hikbrEF~9c7H?3|riTzZv2v@s{ixdJ*^eXo>kP zPh=j!)xZg(0wuy_^jimOoi6lEsn*kECL}jQKr*a^3magBNVd}saB_+q1cSFJP7*za zux{WVyA7BZy)#qS^O=9G--)%sc{Rk-J(5CGdiLR1QN@2NmuogIfekpZGWi1%xA~KQ z*mW2C%URV`CZVYAR;usDgECO?ESUKM^oQxva7XZ+4k=?}QQu@J(-*nkXc&NluO^vK zyfI^#MQzn^9ae_=|C%=m-y5=u$_)3FcF-Id`>3qF(HMjq>$AQXTH)MS7D3nk^ofCw zEcdIOsxUv;XL_lpDXpo8qqbQ0RwlK*e|&Y)y5=z69X%P3CDhh)bPJ=0JD<$sB^2NE z?FN34@RQnbyTCLv(}o0T<@(L1n_u+3tN-t{aCV zO#j{UnH?QPHFyNfzqkRA?KPriPF-*zCP*M7Zwj&#MwVyhn?lf773w?EU7@D_rv5SH z-RznXVgB}`7QIW*${N>JbfTA&6zTbg`Bk zB3bk+fv>0HZLW4)iky^mjxnb=SUeT(bU^-heX@>A+k&Q1dk#jbslyTc3d2vreONYSsgUXbCuWDap7mQIg!v!TOO-OFM105EX z>FH@xH**GA2D0Q*sA~I#$I0x$gRk{8a}xrOmeV+P5N$Y;o{zEaeM`^xpC7k#7^iu% zDP+TW44-vLR0Q=FsW)pr7jsG(RmB44q%7u1rXm*a(#KSklMowr2|Yd`haYz@<$qe+ z@cqBp2X6FVB4NLU!w<+8HggsAPu&Jczv<;r((eiex}J8?=f4ejWf#wRUhl!9GpBFg z5g~GkD@EJ=1ioVbj5jj0yvku3U1-i5>>PEtcM_%a?1MjzxX9&aImYGx-*=`f$A-h1 zM*p;(5LFXXjoG=HA17NV;Bw@jZz_;ez^m4<6LLNpSjg2G{U%Pf8Bn;!6+RGMU&f`x zX{UKj=yv^k-ccROkBkF3A=uxm-EJirf-Rfed3HC{CMs|6=6y0GqZV}{hRS&4SUhw( zkAAqe1PI1UTU)uWSX|sAr(c+?COzq1X?h}!a>=-Pa`vhltyb67>~e;gD(AqAX4UK> z5#dZ}^B=Bx-JdNNIa!%0Pnn=!0qi4=nS&pG{C5DMB(Cj^&*ePY|H`a;WO{OWoP+j# z1S;AgtXlMadECRjvHJ)hS@q6h=GsDn^B6FaYnL?!b6F~ zK2<&1my`N7JfB(^XFRDJwMmSZlPFLBQYEZB;L|9`(6Dw~$wYQmnBVMM|NIT>jTO*V z9T9G^y<1Q6> q-iGUlx=g-Yxq>wE5|UdOY)~-=2Pbi96LODJ10*tT`cZn^FTwks zTe6Cxy4AINK=3KGM+&GHpq=ViS|Sjm-Q++2NmPXGELnZA;Pm|<4@rN)kWRHzI>&hoqv%rIfBRjaUP`gu-ibFjd<3Up|*M4CTQZ)OCEoTM#B4hpXbqXD0NbZj}ih zj_|b_gZx>;-eKb+_;V4T$G;|nwLWPmY$HO04TVt|TiQ(pPRW<>J&1|8L7D*U*5KO5 z&ZsWuIpgw1m4!J{NRPFG0fQ7e;37ei?XRFXjz7MK31p00F}<^*@-w~go}JjdE~xjcza z9paEeZrF{=0>ulK%QWJW1W|n-TfW_JxdND;*a$RlK99vCS$!T16et4X1+Gv;PCC|8 zo|QH)m>sz1un6K|JFh#y^U9H{=PX&jQY@5gUz~xh*}LvUb4!uZHhZso?X40FK<>}s z`FkaL(4E<2_jarwCHFAG1OoDOSEK!Yo5)GgT6iRUR4?qBoaSGdS8b?7p$-u0t{veM;0ot`ljIJ7Zfa;njy0X~8mEXogx4ayY)@OJZjGg;LW-n`~ z4fX}BEhrx+AOfw4?NEEvN`Wtnnti#?PSIT>l7Tckt_#(P2_Er}?@>0thnji}UZd z-OUfWKt>UpAJ=oj1p`h#I<2!X9ecW*X~ad7GROJ~`YpR-3b}9oDn{< z&l8uV(#~eM1w;N`o?JSMx#YuAbeZc6ZN8{5xZAdgBe%7>$BzSU$g|) zkrcgW$5xg#-BoA0qw0$IL}LcrM0ZVt0vLp;ILwvT0!C=#0-Ew!#|)uWUh86=+?S7E zMe6AilqmvoFGXC5ffk~2)4jNz1`TpeL%pe+`ca1kmE#a}Rf_)PENe5wHx0sBD^YYTsBuntv55SF zC0Twy6>hUl&qI9iTbTG7{(9~0kMObI6wDVaXHx#eoj;3X zGWv8x*b6~Q^=U3{j_csh(Xwq$4s8R*I9zY#9rt#EmzciKF35*-I9i|YdO_uI>oK3} zdEP90UCzLmNBUBtJf*#B-cEpfxp53GcwdqThBM;GCD};7t$>YI=Kz$hrs6TbG84F^ z`6{&(IdjgS>X;p zo&od!2T)%K32ef^q2WS2F|@9CV>VS|W%2nedL z!(zB)fcKQ7cCA=MK~ZlY-tHKE#njVfhx1@Ei=Sj>Zn7a6ZQF_UHOt^y0a27(f3{u0V}1XNT593`%=wjNw@W?hY>A}-V%2F`yu%WIaMYu=qk~N z2KoL)D=RN+$21&X^7$tVlWDF&ucU@{^OG1JBL&pAd~8VKJXgM#^3_!is8c836Kct~ zN7avp00M#QhWsj8YH#RgDptSNO|F}KP(J)(B*}7zHp&7U{#(hMz=AY)OzR3Q;dka7 zkUmzK%86x@h3($SVj-}|(W>?Vmn7u{t#IuhP8;p@3WP5zvdYX2#X4F5;uxuB=HT&i zd-qPmXLGNBJj>Ymvf>OFJQ1b3M`TPh$8IcH2H9p}EX6RG8H4Zhyg#4s=XL)9_Ye2$ zKIb~uxz2U!xG(L+$WL*3KROXI%HJ|EWMC5Z3L<7J`Rt;?aS>n49^De@epG(yRQA+xZlTqj|7LTI>px99?s0CsADv!t@}KAo-x-dlQwGh_WK{ zfSRdyr_0DC7?X5*e^Qwq^#ng}&@lALx<|mb6XZRxmrI|0Hp|OIp>m=`S-qHy;!9x$i#}gyU7q z7DtTrmvVU*dTxg{#n-Pb*oHFuqOq9s>9je0iy;g-QR~T3jLm>3(acL(sPM~zRjDyj z@{CNVu;9;-slO5g4hkZL!WL>DeP>D>(Tud7^E6OzI$$F7p978t_p7-TF9vi3(ibZx zfZn^qDVndLT+Ll2c#&itrwnpmBs6rD@&$RgM(5q(Ye*errlv~vM8ee?Uw0oNie|fo z(5_{Duxn+SXMB5ud_x*Ul;7>sHa7BLy=5R)AsGJ;9M-MOKDB3ql$-$R2aD@W0%JT7 zkhRtBVj+Bc0_K&8{uu$SifDZSWNVhM`so}vl@Bgi_z1qEHWf(}@wldrj=3_XQottDQNB9dxL+(J{2 zeHU)ly_~1U=egLq?cISmkw+P(uL7^3o<^OLPG+ryjOg#C?I5^5s(z`M@vewtvnbGQ-FJXc*MDd;-rI! zL>R!`u&+IryH1F^mOFKv_f+JS{4PR_hfvJ@VcwVjY_8tqQ&(D>>bX8f;m)YN<9?Cj zdF}A>v8a8;Mb0wDJP$EQ^n_Qaa{p2WHq2b#S_3P!(uBHz+UtITy`+=$Wp}Y)HoqrR|+V|yw53f(Vm&&PVFgRz!Fj0phzOofyGo80XX7?l1aK}Z;X|*8)57=O zPqY!P$iYO;-I0ndguq?}34PAdX_ujPxR~j+wOswmhrMLsp#QhU{b1-&fH>pB&_Oa5^%zqAnN(! zX*;Qkn=-FO>SZ_Cui&&rC~LKg;X9!+eaq*4;<`%sFmLnve5e_FkZ?f!0SN6irc-W) z(@nxxQ_6+(&?#Gxqn=_yvIl#xJ@%U$Dsr&4Qd`D(>h9#wrKUt-|LWXVCv@H*x6J&f z5`HTEq5WrVi96o00U1**1T;T%I=>LdesFYia~m`B^kodmF_puYzxO`%lh|AmgQP}} z?BA_~PJfU&#MT58-l;y@5J|?0s$eEtz8OE^GogJ#GTHBKoih(>U0!~$Ln)?3hmks8 zzq1;_oY_o#%ACd6AYekaKYa4U*jeRZ6X4>|%s@;QmICe+avM>0Y? z7lI;{Jxx;w3j{>R15GD2V`!$~iK?Lci)p2lF;GaQro>xKq^j(LSu$#^jdyyT6G2^1 z|G5AV?^jD3iJK#?|nM|l@==18e1#_JLR@IHk0y*BNL2TTe6>LEeQIS zbv0$$YfPw{aubiIZKvwY_)t_atr)Gc!|O>+C-q_xORw#&m}ck|$;9*&EoHJl=l+|+ zwEHX`WfT&ZQbaz+8kNP%boo@87!eJ4k6bu|>^9V6%Bc{8CUW=e-6nr}e&4T*0}r|V z*zeC%o0Nw#W%_28gKOiVM?HlgHnk&ApUJM^4pp64^iv&oqe<4iMRYlDkH%fx&l>A{ zx~Ne(V|wBAL*oxltFwf+y1qKumDNl==AzVL5Byp|Ol&E57+h)&cYkeiczjbgwSyVW zb`$6vVw=AWr$&3yvJ)`7XU;!h5=xky@W7tl?H1&{>$dP;hM`aSAZ_(ws~q#&0U(#=t#-a@d}Ek&ufg6D*)@(= zJH{Cu#Mm~y*`48i5P2eBW|-@W4#0fjZ(L)=l``e$ zs6~!p&Q<`%kWk3ya-tRJ8`e15UJy6<-=5)yxPs;o!z$ci)z&DYnyGU~V=i?6XM?s| z6jwgy9XHe+OE6y!P{y0^nX?BtA}i*4I|6%%>EOQ&y>n8^$NhB13n_b!e?caa_$~N!n# z*|lx3RNkCHP~7rEG-f0nwmM8mt`}8|s8+Y%G9#{OACysqBLm%dn-}8}&=Z%w$CaGS zyKcrVHZN3lsrm@gf3_~giPIpB)d#a$LCEX8-@Hdxb>|pG!5e1~K&DU*>6(t>^r&>` zsx6qMDKLE{KYYJN)#&+s94Nar>$X@p`=xWD!|rO7^7yZo>21BV=aZ!SA^tyV0T$2~ zF%q1n~Z$nT~F+K z%zSsWc8u)zrzpiOby)iRH{9>x4!w{!G1&LscCE7Y9ZO9I=t7Buos)ZNd!z~3t6`ow zDN6XUXehm)IU3eeRPOZWE07##Md~dWmY>npV+bCMRk8?@hhoQAM<-e|amyCh74Uni zHEj#p-&{EmQ1)0MEg-Zalh+FEdY+Ts!ApVAoE-1n|Myov*?~0t;CMMTk@0I~Db3_k zrIDoJHGpn14lA+sb?k2$K zH%F5d7*Q9ccGfQa!?-+@_Lew&cECsr(_LPB8 z6BT{ByYg?((&3X>PE_3L#-E4sSh}h7{F$9Gt3w7UWKUC>{@D1o53XEybgYRsVt}Ke z2MWy*{%>qiHsDuD40F4eo5K&ajV17LTEAy!9(%ZWEUZpOsCqm0LzZ%pBqk*C>4DV@0fsXsznT{{rMELHXRXN-Y@@brkPo}={epJ__I2WIZ zSLhmG{sJDKFUhsFUo~%GJd~^I5yg*vn&q*`1ZSgqY#47R2kS`|irp%Dg+Ag;!tB&* z5&UkOHOV^fef)gTO0Kz}gm#DiO?sazWAvtVo zEgun(eoQ63RvjqL#d^#=j@D?|;S(fq<(Nut7s(}e%&=(v=UI#KYYW2RFgHQprAbY8 zs-SX2B9$LD%=07)+4Qs*X@K^>*PFbR2Kio`%vzMA4A);T-_|vVT27r3ntj?Wij{H%A04xfxbf<=Tn0C)4n!ey_rD(omL z>XEk)8JuylN*L}wz+?SvK*NNoaD{vH;F64#g(Nr(I-xiR2x!fd(+BD{1&^I={ZYOt z)QJ|d!&IMfYa^O|;x>pc78(Q~zy|*4ogd30DeJ#=BhG67Xu%nN5{`G2l^1##zMY1% zY<{{~0%f!1BN7AQoaRJ;au&cJ9EH(Xyu*6}YHpZ~`}Rb7;E9qbNn#TnE7w@0?(I^M z{>nc-{b*A<(@J(^Q9<^VtR9rBcgU6z7wt3^bxh;mt|&6`-K)+B7suZe%6_{cP*c>c zMK>n0BQ~kYO(R4(&Viu|znO)8E?K1lt<##i&v8z}C)J>MAyB-frQyn?-Gbpo!BqSc|aYQQi#o%ucS5 zb0Bwp%*6L%LFM~8#Y(TmGqxeEUANI}y&b)e>xRdXYYdBzPsr{Ib! z3mYMPB>#_iUJ>ej@k%K}q-@pNm=ZXIGaXG&gHr@|vn8D->(xaAncn=vdIyya#s%=b zk6LD^$DT@fNwbYDSz&Mx-+PrRt86lq#t?(AgbOgsxKE*M9m%+L)AA#_Lz&<_sdDyp zG8<{2(~Ef@y>%J#PdOZH`>4E2*2gqq+2K)S<>e%?c*+#&yZZzus=jInDX%>u{?c3=J5K$su^x zUeI&78qf)MqKU#}dE>R#8fCkyiA*N5CRZdHS@)wJHjXl+euY(Ysdc$EmAioVX+>c9;Xz0aI z6fh`6{KJiiB;m^?=^#yyA27tV%*?LuZHZvemWXjHk+?&XL@m zVm^PJmUm{9dRG>+%Mydv;`QF-+$kN6Ir;@Z>==5KCZaSEXSMWYikcqKc`qp6Q6_2K z7ThqyslNQ?&%D_2DNYHsHzBMaW;8I8@&0)DjM;ZB(a>19t7(hf+V{UUij!-p7V~B6 zkiAs7)uUTlQQI-dj*zrYJ)pUk{gHm~qV|}U9IpT3 zS!iS!^?G8Y`Q3EFx|4RPy~2p#50c(nJ(h!%^bmgk6)qn~%1m-y+I3+u`C za@13rzOv?M7B}}|v`GLDP5eJPo{~G9A;w!zYx+uh#TA445bF8D0}b&uJF;K67S0_ja8hrXMSqT>AYC2k>|-r1_sG)5fc~d%qhgMy=wU)ITb5I#o&U6cSuCT zDg^+)e=2e2nyimdAK)&6Jkru-QT1grEA9%u=hxPaY&;F!-tHxP9S|Bn4fR_6;LP!P z8bHn~)DZm_J2%Wnsp?g$xRfL6cI<&zW*G~;R2{%TUw<66SRI1Zz`?kiu#(>N0Wank z7R^B->`giQUN~!n2Z234@4xQ^#*LGcDfJ?Q5NgsfSfyK4UWzYdG%lEJiG<$VB`!~I zhyJE{b3Cr1tUT^*!=wA;tR}2r33_UkxhAL=Q#daxiWP>Mqulc`kw!ENm4o2aGAH2@ zBUWx|tWt?E%cI^KP`go#jnv#h}0sha!U1ZnG*g& z=-uHvx(5luCWp0unix9A#t1sjY0~GGZP^$&{QNpICSwS?)NCTAnm(8y0c$UY-bvt# z7se{Fgly6JBH7q@l^tPxPvGzZCu1RyZ~?iT{N-?Kfku}_>sA{`?jagErSMtr#EGh8@ZiSx#9^V_bm56M1LER5%sKCI2r1Ox|Q%l4jg27 z6Nip<(=l&xAA7GKD>*X^%G?@8*!S1_|H>PL|JJ|M5kRk;zCY8i8rJpTH${Uz!k0-a zr?{2%{(h8;BNqqa<-@B=OtgJ&HM!c*E-3vTAH%&;4N4aXMG!Hll}A^I>w z^Qzz-oF^RtVb$b+Z2o5aB&<_5>1h7!LAyUa7%FsfRrO~If3mxn#hbcQN3`EADlZm9 zQARk-BT(H$;P2W`#(S*wak6JL10-1^e?at`WAlBy;^~3z}McA=SHOv2F z3(E(hq(3T?`%yhi(#q4LWzFW`|Bx_V{O`1Ky9_JmWMduW0%J7^^;3_ITm`uZXM__2 zE2T=G!LF7ul|N*xzmXGQnw);WnNw?HmSc=8ZHxYPaVXa2&9|c1>ZO!Dh9K%!@Z>K| zPr8C^;Y#F#rtI!ELeqB(bHsSQ0XAC%nxtJ>$(URi)6@}ol4_&iAL$nSR$c!J#psdm80a)$C!@Rn>GI5W z2JZ!@63{Ls^N3i;PEui2%eDzfrnIFM!Veph1NF&c(I#`x0v9s%uZe~+g~2o>HFvqm zf5^8dJpSZ<*p@z)f*jiIw6)}vL;N5O_CqfgYLOxNd>Nd@#_=sdc4i@%k-*(%8mOf; z&i6{FsKSJkOF)KguCKPiu<${J4rBLw9Yikxb_v&=NJ-=q+PrrcK zFMo%(9PB>}9N%RHY&N<(j}!5A{~P_u zy?Whc!Iv-{+>Q)MVi8uKLV3IPaOUB4$o%sv;*(NB$|gkGP)OXXrBN* zSqOS@*zy^xP+={rpM7&%*HauL zbaSOZvK8gsUx>ijL(6(3X0+-*um_ad6J8HTE|$m1!K*6Qv+Ca_SdqDjl%;t>mKF)S zEQsy6bi*j!dUfJqbYvb{zeGk&kJ_mjqdoF61{7p-7tZS9b5684iVoNCd~o-NQL;GI z&VRC`vW>^zeBR&E8%-7qarY>C1)9DjCwD|Ia`!Hc1OAk$*2TUr*QWm5$KEV;PpGzi z@dJgd8cx@OtlDE;RA|KORh4$VfAH%?ON~c8zjC}^gyY~70PyhH?0+M9zQx|(8C%ug zGcHZqU3q9$r={HK5BKP`h620N3D4@z%y(WrB2IQ%z&cwPUm4Ov@p@Q8OWlOZ^SKzG z`)2y7#4U2cc{a^Ql?OsN1viUE3pr3k`7GVHZPdh!302?Ri%x*tL!R@380%ks$RAYX z3B1R0!jMl|p$FlC7ie-1*7_m|-jz{#ilf6VCc-+Wl*oDfor1X4jz}Ko?=f!ic}v>P z7aAX{_;`VTQ7EflNs(c)Kh^x{UrBo=0uNXzZA8Lw}PL^j#^NNZu@k40Kd6+-H!=S$$Ja5yK@Q)tG;iwJ+#q}Gl_(zwf*>RGL zD@XK)58ou|i%VO#3AtB+!-f1{jD=`>WY}Aaktkk9y_Po9g1scFgiXyYqdH;F%{Nsh zLY--&g<_ez4sV4O-R-Xkg&)3qO3O&r#oEqJ&S^Hi%|#AtZL}Zv6vK|#@8rT(&my<1 z!BK=Hs8AQJcW)I8&3EwcPh_g0#|;|pIH8>3r{<_l9P>oxCnlT{ zP5rrlEw&NsHICX{6$>BQcm>%mST!Wx85;HapDgAW#2;Hk2q%;5oi7|TFDijFyg&Hq z_dE@gt&d+Xi!*}7&pfK*b?iJ9)IpLC>?LkTHEw^b^Nl>}lASX8xs;W0@z0f|G{(tv z$E1c#M=rCTLugQq4f3?me`)X=I!*n7W5 z*@=exO*Wf%{>_>ZuVoCVw&A^$(!UYBt37VNe~Il*MiG~qb?H3uBgsK`vfv1PqzYkQJ z>7zAr2B{};{(zPC+e4H<9E5e5H-jr_EZNZ-IiXe>+PVd2w1J9`XyyEwHE<=Cw46A9 z$W)EyGF7AnOPo@*W;BGY_P%(fQdKRJ+h{FE0K#A7p2pmPs63Cajt+BN0bx5C*KuVJ zt;izK5rSK&{8nlq$UkSd$a@~FzQ%CWo4+tA9-0S{+(u{=`L!&s-9Ppr{61t&?f+jX zjKcq@^$QlxY2ufMJudFUdjU!UQqXHQGVw~XSA`k?))Ac-FB^hlfo`(ut#U3=fi6A7 zhgHkWwoZu1<|DGE$YayfR5@g}f zCcqEFW)rpStO#X;(XZz2hfg`1wzVdieBf+AwuU{|6V5|gEji-`B@Lh+Y<1hQTLmr! zJ|KzPh#8}&R>q0swgwja16r1YwIt{YRqeHA2~2U&*Et}hHPktYa7SaALvAYV@x}H! z_Bkc|Ef_C}ZhXwt(h+C*?}e)~)xcF}aer++JJ+*`7VC~Sf6r};LbMRlPhOa#p0Uy& z=0q+Dvb3IdCF8fAi4RA6U}F?9yAqol<{&G=?7Ew~(%#xU`1SFV&9DHpLLV$CL_S-O zBrT92)85K;{3)>=5EW&l+7l_G8gcy^PZ1}N#>tbi0y!99sQkfSrHHnDJGZgKrY19A zZKwBl%WEh930`B`-fIbd_I2F9K+(x#?0DhQ7yPH$hm0b;i6nMi26}_ezY@hHNO^V{ zlT%b%UqqgR79$U@2YKF|GazL|32&zwIE}gm+w!VRdn?|xW^&X~h`iL3R;brLN%&pc zwsR=)2G5tnVCI(?Th9WA?U0gn%~sf7W=9{CxrGy~Y*B54UQiWkwpU^~yDe>kjU#>y z%V}1cBc*N@b?=i%#Sv;l6652zcPnNa9*T?V1V9Td7IH*LN1(7{$BttN31$U;QE-TG z+(G3fNcqLY_@ePQS_TDQB+%G>jyLXy&OXW4`3q?pKfe^op6)WsPF>aw{h&zEke!)Q zn@f>6re@kc*8j){i#OA5b6k1I9eN#pDLu?3={n+A%PXWK@o3iqS^y8dGjAHLT*K!r zGD9prRf%)+TrE*s|MXL)Am+b07fnb47M>*5C4Moi;>Be2ab?N5OOxTUvO znO@W~;oWi1YQqocV$#7&adQmHc`O{cCOI0k~-<+7VYqlq))l5xebeA73=hZP>9=mWH0YdUufr{+?%P027<>cUowk`+& zV%TxUy}AKb!ml=>^s^rtyBi7~z;=t23Vi=ZMkhuzafVtA$31Zi^AdOCWGh&xP>O6> zK40gP7*^h4{n!&uFI-=I{8MeT&&WDR`qZ}{0{O3*ZN~`@tPW9%Gh9YJL z`w5IFxkw{3Q55DRlI@)I74ne+U1PT@1`KtG|KXGN*muyC^@+sF* zQuMG);w6{R=CUUVNZ)Veo}gPIUp)7|@q#nlh9mFI?iiK@^%`I?-#T=sn_K`)K(oI$ zJO|cet;r6PVz%Fp?kQT$Qwt*VC|N#DA&KL)-tCU?3u=6p}|^aqscOQI@2|y zLQd_j%ZebDyCC!-3ZxW59cZ#Fac_N50ZC}*9RN>QY@Fh}+VKY+%iGw^`#q~;nC$&0 z^I$gZ(D|2U_jhaAeF|k6G2JSR952R1`S~d-cFT3hv(MwU_!Z?lq)H-#eiq4SPc5HCg}8$ksq3R22Z($QKAL zhBV55%AS_y%dPynjp?_|QIK23RZX2}SiK%}bSc0nkWSTtjYh*tbHtTAM<5a8bM*XtGZ`S zTmz=T`T;kYTq48Le9N2&K$%pk;pOLlUE|I;>v~Vfjz-s&lI`K1n$kY!FW5@YywZ9T zaK;Mk4*1xb>;kKZDXZH8&OXq${{gV%@D$Pl6u1apTh*0X5j)~~%6$%KpMB!bNoerW z9jViR>WDkGA%^u`d|~L~lV#1vM+ZB3Qe`eFh6RS7>T0E*ugLdwoS0ksITM#81n01r z14zpldPsk?0!OWJ#QK{e4Yl&?c#z{~UsjP!aciewRdFSoJqqq7Vc<`zi{;^m>s&vf zQHo*_R^CEO-KJ;H#GA$6*mh6zVo^kM$0AL@;kV}mqYFJ(u=%fec^j#*lZYAw_Y{t7Bj@Y`0)6yC@PKe`&kW`;KdY zs~^}Y(uO@D*DH<5_M%;Mn0*`5d}-g)daXstrRnoags7+5w(aiLmFI@C`VFRGdv#ki zm4?vJ%;U9sk|xgI>LSQKQHLrTcnb4(QRu1$yWB+HsWko;{-2!fAyVYva4lfM+ZDT{ z<(#9S@GaUKtm%1ukvp07kz3j6vz$qExy8u(QPa0$OMX{)w_2{74F1uEaK;jxAdkP) z`ea}M6a_#J@iMPMyf%itckPe9AVjhmIkYzM603l6PqOfVsP_m8g^H-WE`c&4=cs}W zVEw#|4|w<44mo_yE8Ea>AgVBS(T{a9k5*nT_MHpNC=R!m|^g1f^_`Q%`RD_!1k~OAtO5h> zXYf(`n{K(m4@kt~0SZA6w+LcP^8Cc)$D^@xGjd7j^p>g%El_!iePWjAg%WvPSr z4GPlnO!PkbxdJCCx45c^jOj-pJ=+5}HWSVLWJSl{Zhbx8md^gP%y0i)GNC}&f5F<1 z3;Y&$JaQJ_vm*8&b-k>&Pk}7?T|{uy^RdMHxV-sQ?ZboSv5D4+RxLA5R#Vg7Oy@a{&8jL(Tf;ggF5Uvdt$uPUJNb?vaaph)okBToxi} zHS^jZODp@Po(`$4K7eSk7T3%?Nw`%wJu96QI$ z(J!fl&_)=?WG1j7$2YzN6n63a`MNifSF>G^@S{0QZ>eFOqMDr2`!=a|aZr1Qw>g72 zI0@~hCJugWP$opfmNkA2+;bnkkh7@eP!Jf%-Zb(1%BoFE$|7!cSyzRq8aIN7&+}4{+&JJMu_vhR|Aq|4UL~?SwsZ^9c6;$85IQe>6M1amvk@Q+A|Wojb{Lsx{Lw#p3qK z6Lg0Z8^E|t6cia@1F*RLcO1kpO@-%gZg+q(?vh}}1v^e>uCYk7R;$+jB@XQ>mot=Z z-Uv<&dA?ztX{pm#lDps~uGhe;0BfPylb96F6Bp zF(KR{o!{#zRAOOVnG%}|z&*~DAE^?2fm#$kStBjwUf^bYA;Qh(7;-ZR$NR|v{08d8 z)yP$}nkrzB`C8b}$*Csdx$jMMKToQb+Hg&ef1+};=1t|`b`=SXN6IYVngQeG<%S-2 z={D#PGg?NMdXnQo&x600x;Z#h)IzljqY?3z!25Dm1qH`6YmGwKnwF96LHAWm&?9aSMvPm7Q3RMCx2q!VYvgEyISY4 zw$Cp};BROhrE@vi=byZPs!o#79)-qSO4jwCn#r0zL$yj4GGpa^Y%(mx|I^}OGXu-E z2xN2Ns&px5OKHbn^X`62w(uVAb0m6bZZ$r4K#o1v0X8^&J!V7R zrFDM*LNF%dL)h`Zr&<{)$ZWp^l_*X2Yq%JSZeMe2`u%Qb0Ln&wnwjiF`j%+MmM_^9 z>1Mw7No-kgfKX5bsK$s_&fOT?5_gO1^%)Gs936T)yiRpxER|r4N0^5FbNWHw({y51 z^p=J5udRprS^_D;5a=C?X_<6Yki@UVSCm2p_K0v66+AZb$!}*!ov5$geZo)r0ouh` zjYnIodkT18s$A95fgpJZQaVNn%x@IK4l6~OT+VbjIkzqm{AX;0hRIQDlVbl1d=q>e z52giO&c#;A+u<&$WVtG=PA*IDNa20#8*~m@Mv@5TEh|4YL+&T(PVh=oFXW+<(pVFPPj(CY%>iS`sYG1A_+FLb( zm?Wbsgua_1LQyFL@Q=%i{!V@rxIeLuBJ17z;w1m$U+%iMoXQh}eEzo(3q#e+-|T0F z9@^owtVI=O)dR%9rF@CDlCElX9ee~Zs;E*dK=qGkdV(|of_m#Seh2EDjjZ{Ur`wku zlITHD^()kj)9JD+4c$?aT;$r#&E`xXE=&Bv0tw>iqmz2z@b_BKHYLIxk(XG>sfsZ5 zlDf06TL`L;dww=5qDIZ-b9KVT{b6}f<)TRg&niCqOWQcP;J$!A{pb|^_inZgK1W5# zu&l=YWyrvFXAj_xNQgs6&=|`R8Ukkd`|iaQS52kcxOs{p@&fMo5=l*73u07=hx*7*gE}=xM&{H&0uIs&+vS zDpFn|evK?-ud4mr*4lq4&-zb@UtGqh@Ob_Te?81@JX&{oH#MfXmfJ7Q=^r9p{)Gr) z{t>6Sj_0p5e-pXB(!F~EJW+D*+w@L<%qvx7PNfc1Nc<$7E*549S&Wra5eL|%#v71n za$suR);p*2FMea1{;nFA{Xed=F6dOfDst4>-PT(v3-mrG9{#MqGN7?O`TyzHJ^sbv zbLGhn$Nd!Mi{dCTL!L9BZ=~l{C(fp*z2)5GkTFaz@`!?7UO2mTA&2uDI~?F*ZrJ@K zlRGvfLVOEV3A_X(07?)+9F8$(r?}n$-qIJ;1p!$m@oEtpK3sGzaPx0UmWd@m@41Df z)Ic@|)N=ZCjUBBxoBKVh7S$m@WORLbC2SvoYJ<6ucp}!JiV={;oe7B)r3hDJn~F^F zb6awTt2<7u@9MZtaxbIz03`omw{xrK0ejHPPQ7!nXV~Yb?%}+CRA!8Q=z48s8Gq7f zVy<)ReqO%J5#SzQRp5MRAWuE__Lj^|vvg=W7l+cHo{<+W_j!VAt-O=gs!jWNVkEH26H#v^IFf|o*|ANGk1HF0@)jJjZK8sQoKRX&;g=LMZ1TX+>Pe-A z={qxMY7V<>OdjHEl?3^{BfVYNi@3DSfmaLJLBLGI!ah5(MgnKMLku(Z{kzJ4_z84Y zXC}b$`TDS)61)kAGxuX!R_Sz=sOh{d>S3%~U7Pc{?6~MvyHbu zoMv2RqdsHI6oRIVL$^mihol}*(*)tvRGLQK+KtKACM?_e zrQg0y%mSs*+whAmy6!3QiI0Qxzx>zkt}9aVbkBASE0V8HqrJ}ELU>pCqQ45jreMbt z%0COq4}SeX@!E&(C-A#3jCgqU-KQ6f+6J}@QUgPsriWIW>ej4+7Ge(nCs}e=A z;I!R+2Dv2a)UqgBw|xF;|NNkL-&l6o*Jju84=`mdLPcbLf)=HhqO0pKBOfmb55(!S zZi5D=&fI;Sz@d#k5j|Y*lPsP}w9zpu3OkI3uovbS?~)QAb9XiFS5|9%UL}Z|&nM~; zf@`Mi=jrg2fpLrt;dtRPFMd6lkr%3Of%GzN65$zBWav{PryL#~gS!bp~!eF=_feQW>lh{kKam7w%d^!D{ z`3?-jdRc%ulCGT$*DOnQ3TnKLa~ej;HL9C_H}eYkAuzqxp~R9Be>(Jp&v*4piRw}M zg1N5Yyol%AD+nD0i*iFgUPzvntyYT~r7u{S!w_1Go8|tkFoMJHw|w6J8D%T>x6tqD zuJ|Rn4{L=(-ImNE&N}&i@Gz4IrnI=fcUmc0_nsX6{@Dqhd%AbTI)#s zA^(TPJHAG;>U=)K${_3OKheApJI+nPUNqV>)qdgF&cU9 zL~*YK4X@AMLpC7mRI&6QTFd4Y*RgF^2&Uwm%6?A0>9PC27kWeWb=rj`bFLw#YDV!v zt=_2SMt66IAohaNZ0T>p?Z+1M-Ft#rt5bIlwx?x-CPMsI>(}20Dla~6E*5Wal9M@F zoo`yqYUp_)sqdov*v(Jz;EIam;@I!4SHVjvsl1`2zNTl+F*Z|_Ys+vDt3n~f+_h5> zHWh`Y_G1jQVTaI?XJ#{h7UwnnYU;M!_>+^Ec24`l#hl;T^w;8V(l2!^_QfN#+nup7 zhrMqgBEtq59p1at9-SPrc+$DCin7ZJEjtmqY^>Oh4l-(W@Mp!fBlNB5D=)jZZgsm} zhKK`;TGbgww%^%x?-079CXFA7=(5s-wQYxAv*|fwMF+Ij*DN*5wieQK;yaI< zsd4D%c}&V>`sK&7>3Yd6|6=z-9ff2#Rem#Lptk;lG|rc#$9#7W?jd*dp7hki z<$q`4$$wt`*)RMQHZ_=mj_;qto|f2AkVOmJ@>E5CA4j zI)`}jM4dW!D(vL9#ZOw7w$k;fg^}J;ceZXBnjc_*Kbi*RT;!&Tl=6ACx&H!c8S-lj zM{vABr!G1T|9HZ^@L6$qRf*f>LOAr`N0G3h!)?pkq_^lYPGUr3$7NHdKNwL zT0)Z4ZonqZo*y9d0u8-cG{GvJ!bScYd-U@%XAp3xVKeRCnWQpWjO;aQkPw)Vxpx>CasQeRUjN!KHKw&X+};za3u5e`5aD?Q zZNqs=C&-y0r#JmF;E!0d&ZyjcCvLAdPB!y#)!Vn$q&Nf5N7OWiKZH14?GqOkueE1v z>o8q_^b9h-f!s59tl7z>mpSp7>{ENM=XmyhRfLT^^jN_xbWcCxZCq4et1j^!xphC@ z|5tT-QNT_=+e~zqJCLiFH0-^d8kd0X7S{ju2F^!i$8J!rJNP=`<)0$&ujs!hU!Q!x zI`_hCmuS=E7`Rznwf@4_qzF~RYW^@;;<~5#XSAqfb=L?4+Ct%tlG$V0M4Z06kuxg+ z=I{(+5U_+K+}2=n&Y!_v>z-=b%7##Fl9*@1Crg}+jayVDMadFQX$ejQ+ei@9EPZ|< z2GJRqf34D85byuIBSGXPd6ivu_Y@HSe^Ah;dI>*@l&5Qzh`I+2E!vVU`fmrVE`=p@Tt+HPV=sq1>ln@@3~$ z4#`&G(vxsBA8`42w2k zjgW&Jh)AFX`uM0;#$sOSScM`{k_NNhf3SEaGPxW|XR?>=jXj%1@B zyD-JiXskAFIvU7JEi_V`4Fv{UK@}JI1V{j`IxYg2mnBQqcLCzG3sltLwgFnWv*HTZ>v#6Ak%i;qLGG7oEs_@QoEem$fv=6i46c&n0VcYjb`nAh+Z=m$TdOyBZzn%6X92 z&e`dyb`>Qfh=h(rIPdLbDzkokklO&f2rkgmt&BUMq#?rAcPEj!ed_&IS`3`Mh!wo^ z{5|Ylt(_6?!V|ZE-Qy(cVs7AErrJ{r7h5UEVS#Ha9&@s1K3-_Lfm43>7y^5Vgf6Nu z$RK|!xa;|keiCrNKd;{JjVzim3Rptf}g}l`^YcL60wC{X0Rp@j)7`?|MoZU5DjyI^XP9ZI^7= z`V2PWYwfR_xV4l_aiLc14g0;?Gw+$1RTMvKl`5Y!l@!iFA+tXDOS4ZX_r;N#F`Vbu>v~)r|guSsA%fgg^d2v_l_CyB10IC;7j&rSP zy8H^14QIbX4ucy;9_qoUv9^a06_%7JEl8|QOn)Zg3n;(ad$^tRXm)d~-Lh7hc93C%-Dz`!%ctGsuHuHI_UL{Hn?1t?70~7_e-aF?9QOP zy4W&YLtpCS{C{(L4O0Y1(+|jF*5uf-q{y&}E2v<>UJptd^mZ*!$Y{5_Y0dKg`Pv0^ zeOMkw>NagI3TRrrnZ9g-T~792s1IfjFl0gm=N*l`$WnScV86=$kF7U`S&WQnoBJ_9cvc>@#LCros35 zdcQvJ&oAH4b^izVb=~JW=iKK!PV?4s6r0wXeoS$<%JcDDno0f(J@a0lse#EK?Uh~m zgjL&vqB2sa3*zHCNEzrPJk;~tLcI%nW`kTN)^#cWh&=*)OQ?J7Rb%U`r-jBNokoma znngVF;lK2mBUoE=m&BrH45hmStLt82DJ8@^Cb@RCWHckXCx~6wex6W!?-M5_$GF=s zC30CECb>ZGPfdu}|D%8MbH4XW>x+@rUmgqjTesLeHLQj+e?q6r^$%Nu_Bak+=PX(M z*GmB30d?6dWmTVRP=N^A7Q`FZHKyDf6iyu|O}^drl#=`6x8M`qEvKbxZ@cu|@#*>D z1pa{J-i_iVOn`|0#j|dD>GfX7V6Gkfpti|B zxW_~F5_=WLC85kxr)+P*Z+<&p5dOJ(UAyk!oJ7_AKC7)CgB!T^gMs*I(+L>md-%r9 ziiO`@wee5(Kq0I!ib_k(mwNH`Ix+Ri&N|<-p~!VCo#P-5qT9;l--BdQP%U04I~zHY ze3XZ#{cJgUQ(1pMI`fF5u$=4Hi>zo-gvp-cF({5V7-f&!+XsZze|b~BpGY5^hLFyDNO{~A2qr9*f{`Gy{J@oi+Z&i6dFAJzw2D~|MB6U z%>L=qv7}?8M|bZDRRp%V5}2@?7_-@rXPNWfSWZ@O+yRyt6cTg+HrPKOeTq9^E2GtB zEb2D_+yLf}zzU|<2GIumX{>yqa$sUL8N|-S!}O5}$rO*0N(&H4WOrwdCn@<#7|7C? zou9#Z?m7@d7HtfMP9+VDLj71Oa6HSlU;>LDF&0T<;s*vyFn3JzNycwtQ->Z;6EE2b zO97-qnU6-t3N_CW&l(8)WYen^604J~l6KzwET%JA)Tz_f%(e`|m!1m%U7Q^W>VP}L zNxYTN0KntspJiBKz?BDQJ_9%P)L4^-GsEVJhFP83Gyv2QmPmkh{)c6M@Xf=#TUbXF zFPte6sxxJlmc30D#eoRDYrI~|!4NP%zgrry`@>5F(@*|0F43=Fp+f1e~J1@52SN9(MtWjZ)J~JPntU*g%t7;^+xR>Qu*3QbbDXOuUJ4uvInQp{{Fd>FO~0Q zW!JQ)HsleLj>?EjqgaqCPe>E2Vx)%Bo@C~sI*;aCghhzYT1w}$lz2sTzCA+wbna}ug~zi6HtK8by%?*Jnk#oZ4~)- zk2@^qs70ly)W@3%51ZOop%Rk^xZ^uTI;o=cVokp>MYKt#f`~%MR32CMS|l0S$ajoUCMm$#RDAlGuuN~Hi_{Vaxck+pdOCw|9_;4P*#Gysy3d)@10hXOs*iA)sGBB6$D;7q(YX$RHI_29MW-SNJN3%GBaLbM*dDi{#3S0FO~oI;0%IG*MLi~J86iN@M2JK{jI-i zhRW6QqIdHZnmhRrDc798ebAG9Q_1$^4!oi3C-)>fRt3{Kd}elbf9{TuSMu+aE~g1b zPEvMq6Zv^ckuR1TEleNM=3HJ5R;a+1*IZ-_J}%_d2e7{t(BEFJDsE zC7Z)b2ymnTIDivOfrZ^oktbfUev-hT$>fH+AMUO1zuT#m!BC@j!rySEykc7?sQaP= z-$4n22PB&&Jtpgg1jRq_%pk`h#m%1Py*`S;W90qD+&2`QqTC0Qsdm8Io6wm-9uG|Y!1+dwO_pD=z1PJb~-K;MO zUjPczo@X#e0Xc+pXB%`5-z47yzgZ7uASZz78WT}BTM`f_qsz`zw)-VOyhe~a-Zc_k84M4us72)>igEB^Wt6-K%RTt3R5N!8%yZ+Gt(CARo{0Cy1-OT`>OS$ zH8gq+P+_2-*R$Fvv&Pg2kkz1 zGntb3XeG9EQ0tK_`lmk3;cV6DPEf*1*HL4_h z3`ZHqP>JK zX7nQe0P;UR$JYsYgCq~&*TENv{>(9^%(JxnsSz}(cFMtP1IknRAl&8DQ}RotXoNdC zS>WJ<-9(bk3#Kr)NHBWqH zBvJGD`+amru&!3$`rAU%7e$A3&eZ8KX(yV^F(9Ib+(u1HeOmi5DTwL_C9$SK>t6-? z5jEY4gRSSqOrlf{)&=@Vrmk;a-mj_F#PotggCf=5mfi?<447E?%D5^ z^ok-}HK+S#<^*@6~wE>S{8FPvP4fmc=YhyNjI+AAR}@E`Gzm;3!(>MC zi*YcS*l`g^fPSaF)I7x7rL^@jTQA&m=#p$`zDUtJ>G|D=D~5ZQ`RG!fToNuTE!RAs zMBe&Q+EW2115eyU9>evF>=-}ZNR5C`S{J%YV)CZ(m;r~H&evPUnzT=@nBFMll!1Os zIO6S|XfMq5!%1%r!>V2V@aD9-UX={yNHNJ7DXtc!%pKxsjDD}yIuX??)vfQiKcnF; z6;b^$DJkm9i<-ZGzyB9?#5Dlcf-_0!$`7yI5Uz+f6LDXvsGhy8YVyTHR}FJC3v0)F zmaFZ^3M(ACFSQnj8b-JleQd49X_#tl+2q;#GE|&?*RzD{ZuE+n%ms1*f4`A&xoSE& z*x6LY#aQpSAZ$VPD75nE>j<7})|}{+Sao}2s{4t?UhuV38u6rXe@3LIqAX&dVAXt6 zK}E#t*5sK6o_QAu^~q5NS7k@Is)L-lNx(ZVqkaRJIxdG|8`uMG?1&v4z*og7sbk3- zd)B7HcS%C9j=RwFonQf!4!u{zcpUeP8jW`%c8FYH7^ji~sfABmaZx zfA?Q(MzOp_y^$WvI{zAk{G;}s&5ent@9nbXk1=rC{@sfJ*FAv{0YR>tK%Qxz7y+9b z6FyX2AUBbg#KI$UlR4l!_z#nn!Fx^iGt&e!13pjYFcx*7)Czz*Q540@Q;(MjdCeV} z1w3Fx=|+nKY=0!8Qg2u_U+NxXe+lg6w+dlF0=(KIFxS)iq|>oHMIhg+GPs)%=-`<; z;3Bpm`771Cb|CYR$%z^kj3|1~@fy?z2nXJ#0uesO06J5q=Tqw+y!k~^lYuJGUeb=j zHFbcSHuC|CZuW&LzSp)XisWbh=O6H2DK#pZIvqa;7_C4Rmu{f6y!l6K(W|8A|9r~I+zuX)%ONxS6{SrIhnfr^P*pHgg z-3)a|v1a7J(4Z}5y6Xa=FISXouE~FOXGsc>VoDurbjG(oT59Rr<3-)ipI#N%0rhO5 z?DTqvvW(eC+4aRvdc%(;c@#X8plx1C@L#D|nK0r6I@5k8#Q&qdJx=J9eqNiNV#$?{ zJRY6LHgL?%D0QsH?H2H5uml6;C-KP&mt@shs3`4_;&a3vc-Q zS2FuNjJ%7C*^o&UR1FMKb1sQvu&lJ=wC&Srh>?vBo!g1#97)vEv6rOW&M%qvLp#Zp zaV1o?bmh)AelODQ59}{n=tjxJlyU3hp~ob!zkIuzewqXPpvm*NOSkZm z=fvTd(a4`AwTx9QidBBV9FMvfUQ~!I@%m%?hy1pB4zHhJdSw3JEO#~k3mE2CNqKY5 zYIFF6cqW>30(bCg(y{{={l9%%7NJ5*6PqDuiLY5F}{v-s*$a?{UFP@{yCTxc+p*i>GE@-p_RrI8@5h3zy#*P#B z6B32xJbjVcR_{}sw&T7x%p!u)i}Y5^N8`qyw^F29HJ@)dY%MkVNv(Y9;#oUVw-5TS zxTnFroPS+j^A^f?bA9`y2Wm*Rd<0%%RkeDTA3) zO?S22uL@K~jtE6Mx9(K#OQ#E;5a&s`^-K(b=bxgkrHU|8m9V##ZQ|xbA0)ReXwcqr z5wz5gpaqrJHdy%&R{nR}V@SP9letmuPbo0-i`kglvlz`Iw)21sYdgV1lHbnStQ97- zgrDONV8SBYIFy+p03W@^jD)LvnWq6~4fRnVKq4DZ%)kP0;TlsI4#>y6j1om(zH0|y zkC;H-Ij;~cdI;b^@wPKzm~;)|oJ#?>YtC2zqfpXma^^IkMb9m^X~TJ8Z>-Q!-Da5Z zd2ghr@Fai8J@~}M#G4b^J}Mwf(Q|*Lw9c$=iZe^xVP5qGy%p~Tyw~M zYheYRho8y220HVrnW|CZ{}Ml(X77*>zSD@~+PFtg$H8=w2UWhmr&#Zp!a=Pwl%bzP z+p>NhkIOpSw`9D^JMH5VBW(l5)uwJD<8?zkUE(WV`amzde8Q<+1;CaXfA;fyAXz0< zq_kiL(|a#A^YCZ-J%^3YWvL5BfGG{=jrh6V($-=XBYo@pfuu4kXoFS>WaP_V{;Yr! zOjr&1tODkO-_Eww5L3yTmR&Rbt92*sjaA^qDxx%Yzoxy){a-E-k5(DX;(}; zBRoL%+u`1YiER53x`QYOY;7lSoAE|L`)g@yGxd5f`(qDxC7uC zr_vFt!nmVDe4~lOMm0Nw>S)cI^moud0w3SA9cMqPIf;+b>x^Q#gv;3%YQAII(p{Y} z3_CH2tRos9*qgYf4aj@fyxKbt4;qm9NZRg*?R+yMCfsp}auvRpQ-*m;=@vAJS{3ml zCMJ)p=8XuJwk;Sd=6~M$divG$FouVC=Dg4v^Q+0bp{Gk}TL>SdtCF@1_)ig72vlFG zkCU2%M;@)1p!5kS0}WJ zUH%ZYqIS#N~aMHl@;uWp$4 zk%An2rfYk!cATI#a{uZQ>*7qcE z$7|n`_%>%l@4p+FVPZ&IZf^CpMWGaN@Vj^c_Y_c~c09;g#k&Kl%=zb;=#yj53YprV zEDC@>>6P6cRvQYo>vu&+_5gHJ@B3M)4_rzQW~KI<2Xe-0d#P zP*b-1CAJ|+yI3nv9B{S@7XH*|kr(mST2=tOQQB!A@;~X&|M9Tz|Hs3ASpR1(QQD(u z0qOTCT{>z_I*iv_=<^+GaVtl0tI(#>D%cSVmi+^}?^E+CW`&gxoHU>s3Mel7;A0ZP z-<0r>bVeCKM776!1C$OK6M!MSzMxeE4^U+Ocqdq9(f}e3yaT+A{pnE)01L6-o-?^v zCEN3W;GwF00=U+9)l`hC`;Davs4E(+<~%mTLj`4NGi8xc9@>CM(eNTG0N3kMyD)$= z%Mid&oiurh18wytL7H!`DMcOq{s=~dcHTy1woc;}ZFZxu3)VW=I)g$@N#?T={9KTL z72DM5#wqD*rZP~bsh$#9n=?HQ$^d*}orGFJ=-#}CF^dDxB>EX(`V_{;Ozpj+%UMEa z7Rg_X)~&J5dsvhvfRo5C`izzOn=vIr!0~(TcYPpK`)!0gzU`Z4f#gVCHgpir1~Uy} zOSE?=aKmYaj!hnfZkgN1Nusbo*VvE=gX`A}`UJA?#n0=o#?`Mi9BDorlqRiPJ=21NT{8v`lDn=kPX?fgK*Mq>1avg|> zq(=+`2)XOxMJTDaJB|?{3MWOg)zOl|i3`a-8F`f(oV3=99Zz(Hba0a9oZJV`U#yGS z@5|KTT-3q_4k<#aMX?h*c;IOUZpDb4D@MD0VAkEcaMOBOt>Kz!DstM`*1^q92FxSa zIF;hq`1$;9?8u~zVdpjMT911Yp_c!Y%o{bGmi3DbGwg8qXGqU2upsJJ z@Je>(a4Kcz%dUw2+QhSXaBs-9(vAZ+?-v85Nu^XXP3jejS2O8uRRh(i`2+94=z$AW z#@$Xws4{z@l)C$<>JxjjP-$fsHuzY8)NR(tjDBaJpQQUkhJRh(~t z&sxo+O`u)Xy6}8li4B8Mc{Y}X=SR%_HAHQr6#n~K{()SM70)ay~bnTbH( z*KQx($$YO1vH*~=9S-+9_OL!VwU-SR?e2;~1> zP}X~?_^ulfu_+rPDSgXm8Xog8wLnHsJTPF6 zHiobNKS>C&I@Y9LtM7a>mqMQ*9(EUx*LSU-*Y;a!UJKq9ER|}{xBKezC#zrh4yCwQ zi^6z_d?hhDaesU0Od}yRG(~aZNGYQ8-cIM4ai2xWaYIHAR^^k^dzD}1cIE*{#v zQT(cX|D?}$g?u@8vDPbPiYFp(Kx7KIyW?frg1FUPQ+dA{m@nI{h)lL}0Auqeq)xgg z2#=uhnBwvSVoW(=vsGn9`{Cc||Kvdb^OXLl$3MfnGlJ@UoY)qv>Sa!N(GNmHnl)P{<8|&n%uef&LfZ$s%D@f@MAp7=n1d z;4m{wVdkAh6N&LJJ+5vFZW7gUo{-mn;!xiwPWtY zOimx_|HHh1auB|@0YuQ)N?IK~jI3}uk-fj4>%p(4t?@qFLcxUtX?0DdTQ1C)HkKbJ86k+|~e~wK00VH$psndvpy{SW!zRtlv0|GH^7sVt9dHH=YX@G1QJU)OTZ{ z00`nH%Pq?>2Dn2V(rRA7WAhOjZFp;HC z64rL-r0=&7lyr&>8>b#=9WbvRVFUXv8M zq{G8&lasb^$CGcEW`&FEqnBToPkYEA!!k5rXL2}~r+Yf!4Rz!nhr$xX>%&kdqVWf? z?gB3a8S6ya7zqzZEhP@zS^^go#VT=%>`yZ7()#(AU89hhQpw!D?yJ#9SoOxF z;^{ToeHUYeL5~hwBF&}EBATz4IqOE!QeLE`tPgb6ENePG3^>SZ@4DMOKEH3v7?EvD zn|Pd8_^uK;63Z7AAiV%T_M`iOhk7>}KmG~LLwtr&ht0hsA5uozD`ZWBOflYP)f?)o z2jXZG=p;k_uk>P-cTQl}t$)9OVg5prxUdXvw9$KXrFPkOND#r`{aIy?tsHd-T@3Ea z`O~d*5%IW?j0M@dwZ2`WIWB#zL8d^S#aa-r*ezCWAzW{rKZ`$2Qi7fiEQwWMIj6F- z*Rl0;C8{O))Z36W0S;hfKNR~q)P^_E>$(pRdvdV z4LH!01&+M`;O2W?+O4pR5R)pVx*}HGD4+CRnTEu|Jss@*tDr}sExoHY?MjM@PhbIm z%S!=*M@M|&Z)esoLpn;6pZcr_xAcC&SclF4UI8~8Vl?PbG2)q2u{whAiJj6B4Iuh^ zx&uu$nuQ2Hqc7{*ZY7?+Dv$=E}c1iM-rP>&uz`$s6^tjA4a_rZf-j?qAr;w1b5? zyLx9hMC4h2U25Lj$e}UK(PsWPE|SQaM}6zp$vDDkw7ajZD)j0iM172KZzpxx{=TIQ46Uluq~3kS zg#M<_F`gD47z5u7O!G;<8gMqCG-+v-m%{un&n?qxcDmRuj!UiDe>W6*1DxZ_R6fl& zs`qh<%XPly{tdX`N+tV>RD-;D?d?^)hAvL#%JiQSAJl%gdQO3Ey%sCfMIg8c(FT4$ z*r&mEX>7CSBaiQQH3i{w%t=Y%${;7Z`MbH1*&Uq1_Qf4MdYa(e2T|2r{kYh#<4x0_ z9lhVjHb4hw60A*Y%8wtwp04~3^}8={I@qZ;8$4S0vvTrh^EvK%ew25ww4QM()TKAZ zDlBI0RvCtDBJkK$%#(r>)Mngs+>V=Aq?8Q&BJM8z7ni;NC+)W_<}2OKarX?*n@JE0 zFb)-60~GF%3PfE){W3wL60sT1`3V=5<$GKJQb4W0f82))MHeusNk1o?d*BPX$@BT{ z#JMXP3!q4p2!JhsDS*jpj&N49%_>b zJIqY$yblBr6!kW&;xW?`PmX@zz!u=$;%sIxwy{XoCRyJ{zKYjj$w(Y8oCFXRB?;w+ zG1bvPh3i9Q7Ii^`*)0^p{T+G+Z1kj3)CGt^`(B3JMfu7 z_S{rLrB?G=Psvfb#B3F3fwpxcG@h7_p|b7EUnUh7ZX$1rYA-6O;w<#H62OV)6N?yu zg-QB---j9)?6CIUUu+~!gSFq}a+qpk`rzqQg4OgEu0*d?Pdj9A9`E82K-JR5mtZ7k z{z%7Hqvr5WXQFKX*R9IHx5AQ2lbu;CfE4{&rB+%lakVcb zMz##KQ!gYMMhty0oI!#en@c5nFHgcE2fBO8^|VhCkSj-5!up0&YeKLR=*yK*vS;gu zcE6fp_oRMEPeS4R@PTBGu^R)9JtO^Z|;KK=Jo(_KaLuazmqjna@0`30Rt+2oTv^^)DD+c}$WE&7 z9nmn=Hsn=s7sQiT0Waq%1eyHXcGO>a`Q1wtTi$Q&&II1-0C*XW13s8v zN>Zvt<^rYod6I|xSYb6Q>_Dx(*`qk&Q9y#&(Vy-X=z1Q55#KsT|0Lj=sWLPDCRhWeTqGH5CiLXpa$#YIIxgV#`R1n$UWpBWX>| zfoct|m|=Fec1;d_-MVVVID;Za)d0?2T(x(vbBUcX5z==S#&{OmoXF@O^`0FiOX*g@ z#9UAr@o$qpiM{-+Kiv~+MyJO#^vKr=>0fY`NVIfm&}1tbI87om;!bE~dM_b0^hesW zLPink9oM!;9H-NWsW7Q+oRZGcIme}}`SY;h=z)7?q5GoshV6aBQY{OykPuSh_FMna z_mKT%x(37~Ug^q^2_2QxP_ychu+tpR|2A?&KV<7!V>^xw)dG zP$2~}U}fO_`F_HJ%*sy?L1aJcRUi!e%IA&nHtC=EQQfnF7X>GqEnC@!#7<%~oNxHU zgrfG#;Fh`)<2)t?GNKtQgBkB8*!uU^qsH>R-0Z9zc4W&c_#C~lpZcoU`dMcdlG})U zU3{g{as|b)PSpBn#fT;W>-vQ&-wqk~t^FL2L;NiZHkK@>wxT`S?dBFdns*be5X=HK zvPox@s?RYpm5LxGiK!vvd?a$nXje>|uf0kS8y`*{N$#HZyVAdJ=U+tR)*{JhaHiVghQPBmPYL&~m zxE!1*$oH38mC=*E={L9rlxkWH(*u-@vs9HAC@4K-NnD;|JqOqq#@;hg zzLl1W0lGNf(t*yA+-x&aP!d$+kK!9owt>3_s)Q{rV22}6wPp(ULm!(UeZ<%mrxIF$kq@>izJg_)Du`UWTfWh|;UIiiP=?Z`s z-ejTX=SH}h)R^DU%Pc13&Dy{i1|uVt#<-$$#^#Jk!xi8)mMR5j`LhV^mQola{gQL< z+~C-YLDSy;lP#P^n5i`@f&8`dfz1M%`>5F8r}SFs2>s#=pSYWyk7Al5>&N=;IJa{M zQ-^WILh|d9Js^?zZF{D5MUKI4&6Lm7K+l|~8UC*R#SOVE*Y-8uC$h8hNM^QbXk}%k zjZoQ;ZOSXUvQ&&QkY;dZWkxF=Nk*C4!rdJrFO4X?Wm@9rL71@7E)Z9|(8#@B`QnI! z8|KRI^hrQnf@#o>tOW!PVOdOW6rOjQgRJWxZzvR?qn{gjil66?Q&)4GA{6riSEhm$kky$WMOsnbrwL2iu zrIUZB%U18{pEcKj#0J&XyM}BBKQqNsBcJG%4ZZQIbKm&@a?e(?qlD6bNX+9>EcM-! zf=ST>SJmAcq-#~qST@{`8=9H?QM>$EsS2dWKpO=1h$~C6A9e}*SfE=(TuTr7hunC# z@aQCSM5^9#c*gvLZuHbcXXc2Y>Pu^e0@U6mryz2jj>xKIlrFAO5A2nVQ~`qZI(wDY zjbN&{3lH7KD)dls#nwGF1x>$uCRRPGm;+RhFK zPBb^>Y)&8H-_YGNg8KTF$hWOm^@AidLtWa=>)fJxa6Kuc81Psu8VXZ#HLf7t6yqT- z=Qvg=q^hQouz@D6VbD0zFs_!s`evwgx0O%vKR_M?1BzrWWJ#?W~f6Le^|EVu$IU8q3jjvMM4}?pUcgF`(AtraAUbfaXAu@X+ zE}Z&WX+m4O+X1Nq&-APLyF(TYSbyc{as}D>M!B_t#LU?{zQc%4$~YMA40YHU5uo>d zPhBgke&KDK>)o?udB5)H)BJdGZ|uUv?NNq z1D$#65+;6v0U@ArE*^0XQJ}f=Q06Y`A3zy_3n$A{g7L%t=Ionv ze2GkfBwGQ2Yc~cR^C-S^p>VG53zzqq?_aGWClKG$XG)U1RMX6#2!93e7+k%^$(NB> zPWxmhp3cRrJKJ5##j)L}V%TV)CLA){ADI6f@e(602|KgA;3z&Ke0ggn8Fo-de&UK< zGpIAljY-^CsI>kSN;y}lr(?3VKGsAKZ38a3m8zYl+O*_Kh~H{wA~0cN-fj%bI(%ny zW+G!-8X>)ko>#Bk(nHJF;%}MK_V^QuWb&)z<`uOXuMtxFmVPnCM*tt1PQh9m$T{EE z)uY_N$UW0FRZ*?gsGX(U!M5!8<^Bn&9M8&PA#}bF`PY2{b>`3msr*+!;WhBIUn<56jhO)x+CIX zk=CtYhF);k=%rRz)@0G*f@qhfNEddyR8*Ukv^>;wo4b{I62;)Rt*uOvx9^yQ>fU$K z3dl;RH4-3I@ah8wvnZc8oSE^!3 ze~~k^TXxiMmQ;!a($mPX`NdtUVZA(&5jA_siLuP7UZ2)4)|DZCZc&lXcu^5O_>k>J z?Jo6*Zvz&3?#~T*y=!{h@Omei7ho8d1*@RrXz%Q?A^Zi4OO^hrXRckzoX-TeEx*{g zsQPyixX{zyr_@rs-OK9QB68YMi zOwek)Z*=32{o9!9?QOTO+%ygnifFDrGJVp@ac3b@aWt^t?3UZy(#!^439_JBqly^s z($?$GBgGShv{iM;EMU&`ZsC6^QJ$A7MWtCKp2L^j4=*7E<8giGs;~Fn(qA_!6Tduv z5-)y2=5WBFU*wfjU&~PUUHso4L3URk9g<2Fg(CmX%v+&r6T0umtF*V$787gbA^`^y|_kIn^-_+831xANo&g*9Nck;-l*h>zA50n&P6u zIwkUOevW<$4=F!rBW;nDrUBzrWA_QazsUbW@^8-~m6GaDD)ZIVebt*?ja>~Je@3n9 zS|FnB$%&tccQLMqqV=+}4o6+%kwZ*NLk~|GPVQpa687s|c-<#rF-^Btk%#^Ht4fq> zTR?}9D)t5xnA_E=q2cfqpCG%U&A7<(ouJvbA{c5QXrJ)-yJbYa-Kv9stCw;?lc>+X zG`mMp)X^y3TDDB2+<<$(=L76?Q*bUioDmN45&%nUwy$QcA|^;uoq>;e#Nz7KIVdAs zN!-?p_-PiWVdzfvA|YOz5wc zp0%5mv{&}HPE9>VmxDI}1a~YpUvoKc>b`>f)V%Jyf<|A-<<8o-xrFHRnLMjh4?5jb zx|YA&<11Wnt2Z@;554ZcVm*@)E_Tv#F+XP97e&JCC(~Yw$A{SL6)Z|bKWX#)EOK*LzMLmv;61<&vPG*L8NqmV25#LgTJ99 zVc-c*RgpL9<}4PZ%lUY^*_%->|MumK6H-(? z7mcRot{Xu-4I!s)Y$!a_tvb`7ZJ;94UV!);wk!DDMjFbI&UBxJCEM`tHKth4+Jt1) zX4~%im~_Bflz_pi_CsoEggZ*ZJG?e!0LNcUc$%hs=EvGg!PusNe zqv!*a8-Pz7xtizkMA_C^01d!kt1M;xG7M*8mYIG6og!ltoml3)RTpyg38|1;wU%Ds zBc$1edGv#|0247z&PbG=&Ic*;K8Q(Fky3sh@+CM6QT5rVk@_mQ!#r#Zstx6uO5#3> zVth!hWT?+IkH@g0GyG0ebl@feaF+Rf<8AYeQnuD%+0%26>e)vkBkqSr)MMq)Mc|Wo z2z>z*V7u3HV`%vf8d8)fK>r{Bt+u%{7_o%H*6IzVPNo)8Efc?vu%B#rbX|4d=S;LS zbaPEBtZtgoaUPVDtr@5Z}6{ z6%58v|M^+0{t*Ly-ZTK=w1EH9X=w}_P)22nVV*xKk8Tv3_RvmCG4P7arU{tTYRE5wNMtB$zSg> zqptoI3n+y@n_+k)Nn~y^a=F}*J)VRPjr^4WxBe3ArL#wMPZ6bX;jI8#+irx9lT7?x z!0L9;iJMq=yQNX&=ywjIkn`xMrG#&Fe^>K;2CGxsNZ(HJ`X^Zu#VTdV%J2uiL#!e* z)d_x7B9{pi*Vx-RsF&|fNs1gT)8A7N&Z1AZ>|%6g)Z?*tr!NbmyCA0{`crdFdPd+F zhc13_dsSlfBVj4;TOq>`BoHXdSR4>pT9BsuqYLVLCwfez+g%{dk3EFnQ@_H3j8iQJ z*ZI6LG2p35?5X**i(QcHQ=^$r)Rz?)ypKCAk-pRG_A#@sXyFk?_ERom&)D7PLv^Lk z|1@0wd<42CB9g8iiB>%5zOp7~a@%QY_=pxT+g9;(w`8yHF1PneQfZ1;%32Efg+t2J zW!R55s}7XM+UlBDJ$BTU@+JsUUXQ_`&U4Bv3?m#llCnflKA-^kDLU&;0zNR&0xWEl(0X%g)w$pCnPf;DY%6 zt2xS&xB()5(|wnVPdUhn50q@ZSNeHBT8XWv-lg8~6;v(l3_Nzsi;erw|KpMf`?EJTma9M} zLq34DbN|oHFUsP-mX&DB7kQ+KL=FX0af38LXI_9DVX=@F1dqaKrI0LB3b z9L;9{9Ob}og>MAjrK1FjZbX>61I~&Ign*cs^jX|SK?fuy60>AQLiq-WcLT=;D=HO( z5rm2)1ws)I*<&~gzZA>+{>$wa%t_~+F&C{hno7J4@3^xJp&-s|Jj`|LF%|}28oWdB znsY2`fE$dO*i-*FfoK>5;B261aD!FY6L8l@>quQdxE3USt0H(V)FUR`husrikg9(| zb11F;@>m5h8h(aza=4*UXwtIi?9KYyH|;Wp;s6c&7Xti63>~414zyBNpPPHaiJZ}9 z(Dv}do#jL)d9l_uBa?keKvpXB_b^~ zk_OY>s7z#EVvE@03iF50Ab;dbA?P%kI>fPWQ;?u}as&yrE>=p*^QjV@A#k9n=q*p8 ze#l~HUy*D1D2Dtg%lbn%EApQt&*m=ajCpUw@Fmfrb-lN4N@mT|YZ8xA;NkSVReV1GYWJjV3t= zMGiVK;=!B60q=(dTkt>BV3Svd`J-0U2A2j-ajwD?7Pz$BQ4N<^Pl*Dw@x=|zty>e& zx}@x7HEBY9HnX&G)Q+qd;RB>b(t*n^5eO-(ZzU-n;Web$ifXYvWb#l!k7tGND*wtFdi z)U=0kMXEJ+o+Gxf{SBGqgm!@Cfw}g7-`2Xwc-LI8_$XFr9+LhjHTW8bJ_Bfl`rGL> z?zajDH!W~R_$%wm7xF1~m1$mdd902Q|IzdqQa_?1k4Cpx38b8#8&SZx9X7|o-Ip2R zZTw);#qt=Xuz*<;ImD*RlLs8DvUrCu{1bZdA+Nr0_*Jc_o%r1G7dPz_3W+hZ_>%bZdSmV_ z|37;J{7bax$IE9--hQlCEdnenq@QIz$#m!a?o~6YVXB_*Y>6?*ov%q4JLi1IO4-D~ zRQxxuGZy6qco~Mj@BpL>G-oVY^RNka#_loQ3(4hdU7UiU_LP}4n6_e!Vw9&@2+aF9 zD4tVNCl{CO(@pEpY{FTx70AWwyQ7%3EllMUR(e_cne?a&Hzy48>27>m7w(?FCIOFz zv;qwJ&SU|zJ#V_PIH9DXpCQvU9lsg{2N-cf(@D#XNl-Dmery0UZV~8tW4`246=(VX z$Hae!v;F>G;J8szFRP^uilnWgN~qDO8C%tQ>9C0rqZQOFVpgrjR;xBuZS7G?2}%$< zv?Z~{h%NRAzJ0wvzu))zUg!V&dS3T+-RF7Eea`)0AkrW<=w0?cTU`M3syM(EwCi-} z>H4bUt|jGm)xnb7fR{V!T+?(Kt@<`LmKM|wURq?oJ8cLvZBr3SW+mS8jsKg^du^C$ zgMX(lo-MEwSu{u;bj7mFsfW;_o8TaS^Y#DU7q-maQ65hd9}h1>#)4*h4=QTsUMd7Q zXnRkUq$Ate2JznTsS}~h?ANS8P+{RS`idh%P`jxENgEF z+QJ!FEGQ^N`DOI#EPAR(wIR^_l(_xl&h}ap)<{_g>Ts`Hn_`naD@@H}pHGQ(E_B~z z04ocYx)b=mfC(8Id2ajK?gxdH5VN42!fuK2r>MP^-0H8sg*s8*m4+j{If;bEGCq(l zf$z+@pq$mrUvb;&4CC=mIknqqpHFauWY!|}!5LpJaM~p7gD^ape+@j3!UK+T>&aKF zg7tFuiu4slio}j_2>PCCebsv*HviZ#WCh&Wuuyb=yQ0E(N}yy$3iZ;$rd??#nO@2Jm=*OxHaO>p!J+dp zWY=|LE>C&*SNcB6XR20JO&j|52z?OyG>4>c_HYQ~cM|~rX4Jt(R86Tk3fx7KK#j=R zd))@qYeRc#QJ|6bR=q91kFWrI+5pf#n^jw;|F0PlZcly6eo?b_ltxG~9QX4^nAELw zXFm?USRODH?U|JYb9(j zTFQwK#a}8Q_UF#LN*B8SDrLRgTxsY{$vLaL<=4%)_l2KSLwOQ}!ETXz!evepS)bKQ z#`CfhayO4@bI4odU7q7_W7b*wuJxbm1`j&=+AWjr>B8$W1UZsjmIh|?{xQV z+B4OeP$i1amiL=_iVk$MR?uH?pHSNw8B7uYQU3tpFIiswWl6!cfPv%Qk^`!-vEwdB zKn=7hJusdFlVc(tJo$9_)3dY95F0TV@a4NT zE=!uN1FGvH=%9>}vkPa_HJPeR|EbBDY^yRxXpLJaRlnab1IG1 zEg!zz?UipbIA6HkZ4S?LZu1$$>3}l%S8f=5v_Aq4#s9hL%_Jc|BB%%``2^E;)g@}R z&a8B`tb^l!3SjstwP`-ALLZr#^f4u|UUkX%2cGXO(~WcP){g%FIvBtHl5${Ax%geX z9muThPpM<3$4m#;G975`jg>ZH$v#$Nh$={=Q^k%rY1A2w5i z@;kp5W6<-6i)da(ab3!q>Yfp;Y?j(@KZjauZ|Er0|JM*vnji_UT+7$s zvZ+fdjzxJps{h-@t>287uqNnbseBn4A=uDmFaU6>TFPsj84P?Yc|iJxo-9X6=;fbUNg~lq_uk!3F^bf1h|kX$xS6l2O1d&1VWA($zgC=m z{yy;as-hVSYEjGz`8C;m%Bmxzq5^K>1M7Q8w}TDyyDPLSXcZ>_Csd=?FJRYY_u3N% zy$>S=%sRP(skcXm=o7v_(>9PAWO8jroG@UN*74X#Z6;StQLUGKU4)-20qdvDxdkPO zX_dOGhv%RnBqUt??DvGtu%X^?Wo&}!&C2SV&?#7_GSRR0i!Zrn{&#o=gS#i1J&;`5 zT1^^w66`Z*uy9YV!S-LFhB$`$>){l<0Y8jnKZ(ye!`^P(7F!W6;ANF}hJR%x&Ofr) zr)brphQw-vy{}2Ee~ydu+E{MT|D9qcM?BjrWO!YZ@-CVbLV z{}%POK~Idb+=?2&J!Rt2?(-bqV2pOz1hp{yp=1|%;}N*#l4S<+Z4aYvJ!=P`GvphKyb8=gM^(T39!v>;1%?J(_ek|2Oqq# znl%{$R28Ma$4A7>`mD+3+V7XpPcn0wtB<>`J*i^jD_|u$?97G%73pT$f)GaHU&7J+ zb6kqQUg-KdRN#MUMTXhv`>{+%I^f@1B!sri>ITh8XxANb22j3K1Y0Q_C_dqdeg2LS zoNQh4Cs-3<{Hnx02XRC!PhRf0=}&x+ z)ml_qFtNaFC%Z6KTrlX4M;1%`-z^aPi)JvS3VLZqz}kI8`b*}K-Z*eIG-Fk-M*qE# z?V^4tuFaSGRLkO`S`XL|` zaKASHdVb9-cgY@$ZA6{vVn#FWQIt@C18q!Es8Nvk-sLM*yg4iP=u?l9Gg;asQKNlA zG(YI;1zkh=^V9i}U;iFhdcNAE0a0dC;K^fJ$uozx;=L#s)poe*?q#GuDlSj#E&U30 zxkT}IweZAt0`_s;5iNIv=|rT-0pp*}E3Iai7&q%wC&sSW(Y>k9C#)~WMw(*gWVH=p_AMWkMTWkdLCBqHZ$U=8y41gWD zkbIOXG3-4Ea87(~>{PbM?A;@b!Ezpx_1Q64_PkA6;^W@O^Pm0j2@uL3A%#R$IEfMB z*_c*)s`v@_SlBhL*lWU&L=oi0$Hk3z_!(`54W0EdUOKmX%)5NN@}(sAMtxOS@8ORP z)x>G!TtU)w7aItAUujSN@Q(lwt~W{wvp^E?tSuEl;jK^y&m^XMzUJyX2w)g$L66~NZ9FLjrSIF!A;-cJd1eyCiU9D~Fo0L=FX2X0XZG0@6MT~|Gk6eO`K zlZ&PQDWWa!JTO+pNf~r^&!F34nZ6D+QmFJ43Q=;tq@OaYl$a#67c%okOHee#nIC)a zMWUeBnGxz>C2z5%p0kgKK4Vt0n`e8RKfHHgTv!SJmWTVtLm|Plf`^>D*M7`+bn^>X z_vmUJ)Of^2U2N!zE8-r-3+MXT^(=xu@c*=mh`#! z!rDgtetx%LAe?L-IPWz^tbVEkKI7jlP#)Yjdto+v2Qq8zk$p$m#2olggt_dyqDY_h zr1wB^u5N#OZuRW*qb@ll9k`7dQ?>g>RBz)#yrXxUMzH{<$KYa$9v*7mijeqz>*QqlJN+v1ZE?0jV~>nfK1U(D-&^!6eExY9wn-1};ZIQP*0FIZ)akv0yw4=lT@r%&03+^f(B4kT(zM)I5K?dua${Zv#v-4YmeI|NReWHv zxJ;iv{1Y7g1n`JK%96xB{-TKQuKJ8x3eJqKIh3u9mY>;{;_-f2WBTMC?B!1GMC+6s zsw$!ATKUOc6QNq#HXD30%U(Y3td?WwTk5eIDxcxi6UIwL9c5;R*_Zs||RQ{w`$oh5J*Mv#I*+56VF2FqfUW%Lxq;@>$=VX;mm{hA2fy+p(2j zGlCfhlytrGE(K%SF`-t8HCG%r$ZY6F8VB6GA`u8N#z8D57f}1b7*=+f+`t`N_4mts z7SVN~(YgH>JL>F8GNFp7qn+-Kx1Kc<(Flnh+TYFx2|poH7K)5ru8AWZpORUIU-k)e zC`*3*nucQdy@qin;ZqCJMky zDN&j!z}DmLfso?ddGm20!y|AW|8Ihv;Np^ifN$UR*gQmh`LX(vUG5xN&O?st@7Q@L z50vY?+=}$&%$@TWo1U4;6m>wmk|p#2GZIq&S+ zC@(X(4yy4;B|LByC$u7J{f%?WZ^T5QCSOyc*E;Fj z>(XZVWfJ>+X!yBW!?Q`s>k|#ppm8&>{n%OmQk1~7H$vJ0%YTN`vG%oovMto++Y+4NMLn?OX@1 zzg6}helA?m&#+LZ8jd*{jDo_)yOqLSkED3|_wR?Eljk<33(nlK46}<$Td+V^Tvi1% ze7rZir2M<*#~)r)vPLBI%igyaktL^jP~vXsDTBgnH~Vxhv;vF93h_E`l7M&mjgFe& z6s8lquj7Rw6#P>9!~0ybi)3pXd&osV-(-tp>u& zf_~)>~t8HEd_HBR4dU#bGswyZmuHtwRSPq$* zJl)=_HACjJ)bD=d=f$JNr@%+ydfrbA!Jb)B1J#H*?L9)QAyS?s5$|d@GGp>?KH6Zx zPTiv;eaR_c|98;=^hTVjTL^5Nvj-X^s=J`1$qK zo*T+Iakw^~j@0Me_sYE9#%1icL+yeF0KOksm_3QuY*gpDv63QTqkcwCG9a+dJZVlb zYCZb(Nj}in9BBF%elSxT*UPz{5?SLP^!W9yg%p{?g|<|#KG5?PLZZ!uq*4ls>Zf7|0?2>rdlW?xK2NJ8dwvJnCp80g4&41n+@Y< zhVWi}+^#n1up4~-4l?8G@d zS9S!!{2JLKwEj}ETw=H`&&L7mI?h>Deg9pQl*|%?wz5muHPew#*hlWmyIkAPcCY|2 zttWZsR^rNKZe8*$gp0kh_536{+-qXY6lmm1ck<5_QwF;*H}}~MC{0hoGz+Ihs5to0 z;cdi9pZE{tHsc^#N7DJ$-pN`ec;{iwtzqhHID7nnqB}EP4+dnTohcsQFcWhI{_i0^Dv0J$7L?f8uMh~Df^vy0r)OcUW^wHs}7I< z%AMVa?_3H%4F2D(53UK1L05W+>Hi`~ygPpRKRY^}NQzg)8hVldv7{ARI#^@guBJuy z%h7Q`;mmExNXTZ>pP@4W<%;9*mfxh`-*0l}ARMGk4YfT)(~eY>cyPBjv4DxfBG@UE z&8lzS)YAaCPp)bm=Y&mP+E{++Yr!|KEpd*Y7J4HEPh9QG^}f?^B$IZeoV&X1pzC=G zeh;K7?bW0PLocngwRR;H4cut@-ulJM>Nc}NG9ihINZF}G^#prL;L|s7sd2b_=V%?* z1&))W=r5~#l`X!`iQNzDxAJ(z!+ z({s;&sJHBNs7sBzdNPk=?~+&GVXq)WQi;RY`2%>m*TYLLzrke&QzX@z=F02&_ghoo$@Hnp z$=Bbm3vbeLv=(qeeD_CK;Kw2Lf&E;VXRpBzE`G4T6To!zUFhBs@J}7hiCs5vtpDA8 z)!m*00`Ea9I4qf(A*1PcYUNI=MP_)xt5`0D6Q1q)w0J6MAoySpo98!s7zGaY+j|k4 z#LSQR)fIHf*yic(2&7@sb2^IJ-%T9Gb1}rcuIR4xTqLH@2296F|3#cOer zes!Hp_$eK(ye*ma(>=Jj-JDf%AYreYD{q};1NHwv@EBA9;k1m>-~(t5$C4c^%- z$t~`rHz%g9z}70ibBI#Wc} zBlCL>{Y&6iF(QpaFQn#$i_~iw0#w!9J4G*A=qcf6x~3}A0B~&k(&xHNVeb^B%wdg* z`5YV^j#dUTLBujA9X24ov{!{DLa&|B#t1oXD8Pwd)9GKh=3c$Xnc@^lroda2MIy|9+D%As=dn zZc({ok$hdLnY|(6Q4Nhdxd`io`J}5I)I(Rc`1;Hec78+(|4ZLAH`=t*2hW{S7ADnl z13U|{T2!`m`q@AifuwazQ&rWy^Bw^$P|zN}+Q=y^(Rv~1{CK%rtvc_pE_6kdPdO2S zsF70+t&{y{MeOv85yg#SI+?7`TUUuWS-pCJKnOR{e77TVF(V7l+${9!?}{2=h4@EP zo24$Um9FBw-ivbWBJWv~D_Ey%*SSCQaB z{D*{-`9<52@b?ro6Z;9b&gUR1Zs=xTf#G5{FBp2kw=}O%&B(z_++~fj*CIin9 z?Ndt6P@XT)X-BKV+)7uH4Ea)uDXByIXz|9LyKFs}ydz`S-nJBG{E=~F{HdD5i9$(A zFmm49v!efdSGx6i>n9Yhy}doZh#xf16jZ}g**gh8%Htmp@f*La23HvK%a2RgnWu&L z9GmmWJLS!X1rK6B2}~8*uh%5WlZ=eD4B&5GR9f4?YaS7u0Kp{An!=5+i6p*Fk(0!y ziu;T2l?ekaG%cGr`!F}|1unmq1d=5s!^)8(tVevX_y*VBUkyA1Ovi_^c;e4L|Hx7)gcwEjRM zmuB}%frvj|Co?IzPUVqBzI|3__x(8L_G4&1EK;gz;=2^^{V2LJuA#%#U`Kiw_ykei zp|v?B<-h%{{Ui7Ur=ol^k>3%{AGGgueiO_){i=hNv~1JT7@ua^mW7Pp6o4wOsG8KL zju&dUonM=QlA7PPNt_TuQz}{xh#oMCqvG_*#1#iw&ls1(YlMT6U45Sh#$S#>s9q-PgVu+MHyg*32=>QM+f(-&4&wnB;U9)X zExP|o#i~$0MFVrm0KU?4<*h$|>2RjJ*!YJXwt0Qh5($DPbO=<1?hAhZ)0O2lq^I;_ z72tj<;&U|O_3}`5jT4$NE1AI<8KFK$+p0(N)GXv~J10VAbP@y#Wv~(PiT|C{{$TPe zYZ+j)hnN^neG(uqG@~5+uZcp?I9-qq*mS^%x7k)_oINt~$9RwK*ZW^sD&>FrzcL-V zzSlcVPO>r+Ic~~zsOUG6qf&yFMYiloxaW&Z|IWsnq1FK z*CmfLf9Yq(_ljj__3HG1pk1x>O z>v-?}^E}Eq~?+>J-r4^K)7pWEg zLEF@ph^_FeZ*!sLP&oi{=a4_~x7isP{A9N}Kqxfz`1L}xUWS9S3I%Ihp+F6OtA<-W zTtp4+kel2O0UyXTh+0@?O{wQ{+x1KYI>QcY!b&`inpDZds_nd>ZyROTM=#L8l(xUw z=xAL$oSYZp!Cpm~g#uCMTwIy7v;T$Xq|0S!)#}v{mP8V2=5(5-A1Z+TPW>@+<~B_8 z5r6PIBG_)~LP2B3PB@dsRX%@XSu8od%o`A}f}l@QI!>^(yJkA}n(>(MNU`bx7_Zx3 zB)psHS#!@R1mjqFGP`K*?}4yQJqxvy*><*1IKx*zvgjW+Q(&6$<5oW+Ez;`p%B zjdR~vHq{Asc%sOMzFx1E&bpn5x^E*J@GKtlQK^NTC4PeDQ6v~!)JL=PWUq-9WGSv- zosbb}9#P7t+m(P6=NOlhiKDnr1Le@nVS#jg0Q!&fz1`opP$Pzl{BFD)T`AmarRo9I z#Zf$xEWP7~$6G691>}!Kg*+zqA=Tp-e$Grj3bubT0{>aceX9rqeN{IKVwGx2 zz{d5E}T_5V|5E#o7up*4kn{RL7 z0BgHC@H%c*LJ7k+8&Hws(vIW}YjmEetNL&d$jYay`RMTe7V*1!l;M#;Ue(oL!pu^m zTqv8=WP$$qvN}nV-~YX+dWWeEz~_X|&3@?3+q_>N!E!w|EKUIJW(doja7q1hpcQpS zIdHr)GP^N4;i&B*QFDTe#BrJlIC>jGe1Ate2ZD$eJq-2UORKBjR9#R{Utjpm2svc( z8YQ@{i;%dgMXJ9X5nVCb;bjO%u7g=PX^09ii4l=hfgWHj6atZ^H3?iz?|0pODNha8 zBS6F>KaZGlY*mJX%gJ^`+%w7Kjh+|Q?O83c*K1}Wys*{r$IUlPsG;>Jnje0gx$(%d z1S!%Zlrj!Vz4^s!u!4}KO#G3SZ36wL9_F1~C@uac&mrt0k)d#Mb9L~^a223gOBq@K_<#c~sh)n;h!x|@#Ua&L*#Dcj zi2aWO)Ag&Q5{KV9)G&UcvZQE!2cCwCq8|v^bAAXGvgORh;niX_i-3o93{7JBX3q}0 zT5%wWeA*0Jj=cw}Ywoun&}Q}gej(g`=KC-ow$jM8P;>hCF7#AJOcv|x$9H2ra1 ze#%9+i*yI7n$V9>OAi1`K(xQ^UD0rC?EF+gh@sYa1*I@3o#DO33oPVbsd3$W4BM+S ziEOan$T0!P)NF2ePi`wJ~JPxhE%6XP%;%*7^ z_vjf0jUTMn|J#?hzItK81ODL}=6j$WT$s?m<%|G8ZA5+l7=EH}+BYg;Kc^-{Hrai| z!n2NvBS}MI#1wsTJiTGia8j6;_wl=PBop=Ct`;a@0h^*ch>PLk86MWkqL5u|yb6ajA0?D}WOnh&%6JWSaB~XeBzRN&gxaTtjKBz(+w~@4K%I3Y zb{(D;EDfqi6BIN>e~kNfjwI;{GS5S?1dW5|sTsQ=_rG!V;qvo3wGbthroONT#;InC zPNVbD)0Ew|uLdxg_qz>Jl%n!$U~j=1M_AzaAT==;3XC{hV?+OQT4iLpS$?$D9e56Q*ccWlS~0K}Z|J~7~PeLJ=U3DHt~i%#fp zwaw2cz+#ZKUms;N`wl zFX{dOG*E)HwKvTX02keg3Ab=>@o7eObQ90d4_m=dF8g2u+>sqv;@Z(gNMHpdsAscO z-*?W5vP9K}o#2NPf(C}Hk*_W!qjpj~Wd;hQPIt=VpyzWIb%gZz`L<@9$$58yl)QPO z?T|LeAJNm#23#1=N8Jobe2#+CRTW8G!l`m)7#*z9Le`Tal&ng@lDxBXI&RpTGumjgP8t&CZNe%s;Gj5$;I1Y~^FYrWMu-EU4 zNCd9+33~%gj)x>{$)rVT&yM+c@0E0+ipHf6VXp}$7CyPED?$aS)A3R$XSiN`;=4K&h zp`s~5HjZ>@dWpa^psv(Cb(q3O z*o|HS2Wub&&2Fud8KI4CT#OfaXKi}pAHgGWw?;&*nY8V`yLrm$QE7!DUnx*bnh6GuU{WlYlemw;} zL#a{q*qm&v-qzUyfc$K>RnTsBIb~$H&+q4?NxSL@59YBw^}h7$ZKbT*H=j-35I|m< ze1{5%d0M}qaB$KMJC_8ee;Zv%<)ams{`xCJ!%X-i3Tff|4GxQ6rg>jW@@ z`x%{67~Ea!x^){}m^M~Bxk2+8<$IdlLSZhra1dy|oYidue;`=KT5A}*lcFx`DWq;6 zZ#Dh|g;7ms>HX%Q8*C(=61RD>)&bq?XG6R%m0l8bAQu$31lVV<#{rtS8>mI)(7gsT zo0K;0d&a|wO3jUjDRGu36cG#}`84@Coa#_NoJ+-(xi0gk@CU_&l=cbrW}C!gcF76x z?&s|ZA><8LtNPdOwkF==)*-<_ULa`Cs4_lwUON+GXVPW3Q!7v0<~^-ioU6a}F+m@+ z@e6$`OJGw08KxEZE2X9W_ZN3d-j0QOJkb5{`oWAV#lSc#KEAgTXcib(4$eG!#=EOH z^F4utZk1D~S|sAr%aezGG7zv4lA_|tJQ6#z*e>X`sUp{rl_--#zJSpv{vge0b__A! zogTl$0h%9H;#ES^5#Jdessz3_)95h{w5=#8<)RULSalpHJNjBqY@ zGeHwxzbnO@H>+z5{k~B@5g8&jobM1mt<)X#>*24+vc0qptK(}BC}$sGBJgcVx=)kD z(aXf(byvPj^jz`Sgz%e(}|g`2Uge=iLc+T8zRLHj!lu<5OC6OBd|wXE3zmy z8v*^4>e2xW6c}xs`0O1+G_jgEvPV+ zYBm!ZR94cic3<3oj=iR^yS`?b-MdzVxcB7sHESbS38IuG$NRG$*nzS)xlSk#5wPLp z6ptvk)jNnP#{E-L4VRkJ7MF35sV#A7YAef4qLy+UaeZ2n#?<4Po#rEyLxjPB<&|(@ z+o#{J!#0LfkFPI%18Nv>D>w5(c6JiZa|_6;MS%TWLGnWLGrlDH)|wOV(hYg>|0Bk- z|A=wx*3gG4v1gfNF^&@+F?YPFn3NeEf4q$$)YGHp=%Fh)EGb2_!f z_G7%~${0|SKw}73OeUm@0d3eD7~E^nYgz6X|4!%?HxF*|188zOtE=C^;y(&si$Z*D zyp)fLV;CrsCU7KkgDhN~097JaDOscph11+Xjzr$%-!JCPIsV+)g^kHbrZv;vAUxcD zH`shDkHvdUU5`qcyocvE*~p6eJX79!j)WHWpafd+E$J9%G_gARAb54P!}^J9&gdw2F^nIvvoiMZiZtTe zmoUbby7q}6?aW{z>axD_;ow3?xS{K8RWaaSnSrPuQ*NCf0k#J!3Otm1Y{ML~z1y`_ zKVeZ5XXRMNYQc)h?jBOP{n>CQ*xid}qVe@TFIOW?F`_sP0dZhx(g?Zu`So$ zQO8)ipUV)Vx9x=U#I_yR#Sbkww2DmW6|Ic*aG?@A3!{{O48W{&s~a6yToQ(;9B7XM z{=6MZ2Rlp=7|<`Bced^n#kNP`cqdy(uw0)CI-CmU0)HTZQ0vvZyl(=38e`0_*$<+1 z+?CN+y-dCCHg4y`F~XG|B0@(|7rO}ufk;(GR!Mv1ozx=9vfMJe(1iD>Bt}M=Pw+M<& zH8Hb$e_!o7i;I2tNfHZZ`{h?WmNeaM(yy-ZF{!t`c*T}I%b@!Z7eGgK+WZ3!S$@SB z>{rBPU}VVXr|nsxCnUPA;`Rc>nd8z5m;LMM5$b@!Thd~ST%0yy8N6%)S*$8U%us2{ z+g3bQ-NL&^H7Zw=RKR-nyTj;IL@sqX>kL{!pf~ic^Li%Kk%nln?TZ|2xyGd1k&5uN ze#5n*533KKag?iDGgpFJIPIw2=)a7vL4V^ieU{eq+baQD)s)_+m{66Xyn6g(9S?AG zdV1l`aZtCZDjSH)&oTW8o=(SYr56p8xU?HP!6&`s#6F_{1>Ug`{&a@4zYu3+E@N$f zx``Ds=S=ZJKNK=UzX~qwrsU%$7u@FSuDn$HW<*c3ULPT0 zqroq)O`8_v_3a3e42_}V)Bj9!XWl1vQrt}w9JNm8G@w2p6GtN77%TZCw57}C~P zSjAwK-%s~7?t5k6qUbUgUzn6k*h`8_xFIk4Tl^6{as6$pY;R}vg|svMFGJap=Y;>9 z4keM&I25$0=%;5QqZsQqT&44Zwt>MJlZb^3_lVP!0eDnEuIN6bJKus!y`TbU;Z9z_ z295POqa&?Z(9*S>E}FAD^^S*UejX6z@f%i{-)U^lpC!`Bw_~tJpGOni;&J0kao<0o zpIJJk`bIc^IVFAi+euT7M%B zta$P+@}g4|J{twf)1fQ~xb|p*myZxzi2ERXKG%(f>MxObm36o;)|M#0{lqnAmQnoo z`@!2O+qF8$>#M_`T$uDTXU4-T{J5P2>ovkBiOK1Z72z*Mtqp@NGNDmx)}2)~3%B(^6< z&lY4((clBwNa|i9iXmr;9ajNC5u2ZOu~{d3u(?!cc_bW`DDQu12P-Y^G|IMm3#VM9 z^Z~pZ?l1Md-p$I$=Q{2WA`+1O_4x`kf4F zOJ5z28rsg+-RSu+h|~YWF;yI&!^KXT+10vjB9=ZQ_40 z&q;S5N9va z7&h1Z?0V2RvvCCEqf=hcfuccJ%WSbk(26dJKo^|8eOY#Iou>9Jvm=jI)CN0`!)|N9 zGD4-qtLaOTO8_1Wjq<81u3Z*%aj&|qR|cb!8gY&vzFqq+5nBo3_jX!Whx(9vC6i|Wm@vw-4Rz8U zeX29VkxTUdq(Tqz6-;+xgj)sx?L;LG{41_8-!EM*QP^tsI_jv6bGq$etMqaCR?R!f zHD#*0VKXRjyWJMOy%Wojk|2J3(|@#b&ByZpau?J8%UzuGKX}?OUam1kB+cAT#aU%S zg`k{;F{2CR?4fowNKTbZ9G6Y)!^ODrK8e;qy{|N?ocT=F7R++J3qq7$y<9&!={EHZ z`OPfjAIZ=soK-F|3(4}hYR`bjr?W4wVLHGvd+ect!b+*560>QRZSQ=EkM>vmuIc)m zkmxh29rEeHGGX>Nqrfy;%r$=7*RNb!p!q=MeyE-ul#mSoQK?h~56 zyqmw>#p@o+V9d@)b6buelm~d37jk}VtBbA-VtoTvppEjTc}0i?$6Y6;d^;2(;Lx&oSf^2LzRqlkr7(N!%`2V5V` zE)V6Y84{f6V#z_c)-^Dx2~k7`U587&NXC;@Uc4GmjZW}!PrvYd2P1P%U1|?H-TvTD z?j-S92_zSB??sWa*>JZua=vtso-<&HHVe=NKMaRLkQt8<_ub++d5uqYc>+8=v@59j z){YjQ9}THiDQNdeg~wftS+y?8p1;_^0V9F|@lqkPMRDhGoI!*%u}${Ph`^3N#3Zv2 z+>xh*xX`|;>{;E4wW~lBA4s0Spb!61uy!8Fjye8mq?T}IIeOIz#3|2K6F8GC$D*L4$wRBdn%TC-SQrroLjB5{Ra4uiJOEEUE{T}UKO zjSn~5c*W-sA}++P5kl)EaV1ciDW`!DU_!qYvY>_4NO4MfvSH7sb-opP3H_k%Hy)RM z5qP_PJWP==_F&>(E$(HmFnau`-$_||ej@1WX}fj<*NdYW#tM&aAwqj$?D!_;X8m#d zxD9*oKor%KgZC&Y;;h&i*bqm=;SUx<<=XikRwc%O7PJ{`9(&h z+qewrRsotBE!s5igDe?m+Bk!KIkmzTheMeIC||jPm7YfU{L=MHZugOz*IBwBzH|SL z)G|&N@V)F{l?(_^0z1C#LnV!T^K|w7G1X=tY}kYw&oUP7@nqY^$E4;+6sPZ0G zZ6W#_1ao*0pF%evBnS&=UWriIv(=Z#?r3=RHRLMEtg|m5FpQnbDvcJ{zzHQ?w~}w) zv}69ezu)g^)ur^LR>Vv36g$VG_nrQ)oPE3gf934K2WB{0AZIEQe!eV19PQZ_qch(5 zb@1cN{Ahc7d&a+K)HfI*Rc1|b9?VI=NWiH77;) z@igMeC+ouW6KZnMds=6pH-wKa9e5^IXGZo~P(XX1$Q4_iKC5_H5qk42umOfqe*eTmNkWS865e=T0%kLcHJI@`i+t=!2`5 zOW#2+g#}l?#E(RFHoA5UrUQV@koZ*rG#94o5%cWDJXb5Xyyz}x9MA#M^s?L7j?P^r ziOtRu#+y!45;2I$t)6QiB&Ls}bIz!rDYIxbQn*228x-REQk$*5j&`UXgs3}QWNLk( z&gku+Eakf=64}Xo{4G62wb6^+2q}@0&H}aMUqJ&bVnojHf?WQ(Ak29Cy!0zpjnOPc zKT-t992X}fSE}CUuOs%DxFSl*ne4@vCU9$8#6Ir^9+&yz4JCqp;bYjSdcHL}>+zpF z6=Owk=}LlB1iU+uSIIt(|GjQ}?@GH<7Z1VGf~7Lfr5PuD+BDel$EC6bx-eVW1+fn7 zT53_f=Xe}h9vc=ok|C__>-`bTn;+BQ+neJeS~7#&S{HWf6l+wh3P@CAp6b{rt@fH2 zuciIzJ0F)52R=Y2Tr9U6cR{kEJ(_S7NAQWCvo17PReGW@-E!%!R2`)zCV)s1(D&s1uh;tAqVQ28D4I5?HGC0o*U23xMDh zwjNjEg>6iU#^I3~gPX4F8m~zMmNxy-3Ku2`(|&WJ6V%bc+Z@5p{|&1kiK^D1A-fT$ z${1@JTY-(E0X{9q^BR$j>7A3cO7 z#@X8jtZ1<8IXS-4Yvcc6>pk3>>bh@HihxLouV@e{QE8$;C>nZ#q9D?&R3#vxN$5pN zXwt;cq$4F16&2~d1&|tgK!kww7ODvZNJ4=7`uo0f&$;Cp|G?hs*?X)t=Nw}!9O~rg z0PrDJ&Ar{;fpRt0ubr`ePgveg{gq@MKis3vPEx%`0yt6hF=tZ&{>miAQ7>+rXK=4r z-@-$x-Se~H5&e}Qh*AyzwwNG@D;NUcmNW}QLq27&f`{c=L0kf##+BKg#b zgs*pHtH@)8LKvhXXhRPxmsH`V-#w{f7rct>hiY^uU<{7Vw}@)YsqA5g5ovPS$A~Vq z=PEWbK5b|Qq9Dzl#lKvMV0DDB!q5x8YFtv(UeuzFSE~Hp^6T-ut_I!d5xsi}S9-_q zR}i_AUnhV;`poe>o>$$-Wk}d7e5|;T{7k)5Cz_X{E9swHA@#!K|1I3ET1wRI6BAcz zGml}(36V)EQXxq3nBn6fd3Tr_5De(P?gF^3sN@lvn{0Bt;LL9jDmY z>^xWq#d_u65T=Z2KQ^>i_@aXrTKG=R%l6PcTpwnMO!+El!Wp=ei;g>Fu^?rd^FU^uCN~q^gRW^6K7}2zM}~+o5+&H! z^)86G>RI#2Vv+v{X(2sY>$Co8^F8B4_UIri-*jkZo6T$Y+1X2*wwsB6Oj_t;6aB-G z0Q;%1urTuQ-rxDtM5qQOO?+$EXW-XT0zvmh#SI&st*jzZE$pymy}!>AtB+#N153L{ z;2N6Arr;>C4koVU<6YXx=Nx2b;k*`&4G+Se1>kz3ATtx)CYJQdYMz^nDX{|2Adgwi z%;_7umj~Im+qXlvF}Q^6;cHmjc9x6p2g!-sgRqxxze5Iv)@-jt^h3fsTI#1R>e$yh z@8{if37pmf6R@afrQpNi*sPb~&ut-NT@x=WJ*OlTH0F6KMgW1CIxEWRO>RHpkeTqK zn47fU<23x|3&)exXuTsOv|!mjh8nj?ZC8dJtr>MUqmL%gP1O1*L6e2az=3o=KH&Qp zBQHWumZjKyt)l%Pcfj0Mp}eznZ;5kcfP2TwWFA8b!_~s4aXwNoGMg|t^srHq3JXKJ z|7kwpt;gB8+|m0jsiy`x$$$E(fl=u9Vx!gQiZ89 z!Q@ynwGwl*$uMa}IWrTj1pAOv&D=Z{=YwN)QOW%LCE^ti5cfF(v*WfKVU@!rc!ACf=}S7xgK9p@Iw-`6C)EP1zmUJiW< z9iH8;>6I|n7Ol|q39G7JBp6-;Beo7&ZeA7q&8f4&s?_FEma)e)F=}_dnIt8tMK{K% z$35N>`qYbmpPU3%T_?2Pda-hse`p`c_PVm%{wx^l2dNHfkF2B=t8lPQ^tuc19`$QWY|A3BMN+^rYVe&9Uc@TQ3UxSpI^)nC(~OoYd>$ z=z)%nYs{FAFfq6?v>ncj0g@ zy&x#Q_Rd%tSK^bmD#&M17-mW6c#9-)q~)~w(>cf@a>sqkj@T%jkVW(GPT?SG zMrr~2p0Rr8*UqPceyuW@ObNcxT3gag7FPC7f+if_F$ zAg`Ot#`Po#uPYv);R*=&&0(!ZvZUPvA>9U4MPT6pfF{?Iz|vid(W&xzaYI6TcoPfA zK*?m^I)0Z)!m#sWL6mMLzsR-DERS->tckM-&;3=!Bk~|W6>pUB9cnR`!IP{q9w~%c zHGKfJrTD-|bKybq^VWEY#1na+I%iAbqDX16ypwc!JoxiOtMbP%_e4Mk^aFA$9krAX zIf>60L zw`P+#T|EL$-?EGmKH$2E1V%F-JuisQ31Y;uo3H(jRb4(Qp(Sg*FtC2SH_N-@Q_7f(4`jwZ$XgEz-`dr&D%O9;A za3AgQz>H~}E)*?E(+#G?=t8R9owSC3gkX_xCDdw&9=Mv&Wn2E8cZxoHElizd`bs}=q{v`x^CVm^cw=kz4y$ocGOcUl3^qF>JvY0cH}?#hWPzE6gZPlFTg+Ib%ToJkAl zO7)8lTJNkW@Yk(}ZeE8>WW4B1*?IXl&F3Q5?G+cYt&qi28!(p^eFT0LpsH}`@!})l zoO5PkTN6fI2g;m(Wi0<2hEZ&BYM6E&Hx_-9H~h&_Pv+REzxqaUwlqA$!Ri<%uBkSN zgCF~7H&Wohn_QA@;cn++ zq%2C)j;C5xnh~irXr;A8kr6DE-q0_4m zi0y813UuF2e)g=r5ss5{lauX*lWlnlW~2~r4Xvjf=$t9bxLpb8Pljuy2r?*A|@J*D|yv0DDfi$dGh7W;sDU;1}N7dvMKemYhyW*E$Qanu{DuOt6`i z+WOtUQ>h58Wu+P4MW#CvGn$&MZI4*aB9fX!>-C4%Szk^vTZM!fWYVvv4X2I>#8Pjz z#j~Hx!*KURRuS*1?bS%@jNdZ7r37nEG&9r*MOV?s{h?^{gJXXdDQ3x6_~hWQZbgllBr$HpEkL1;0;J72_B` zKSGg)5=T9H#x@Fq9}5?;!pV4Hp86;GK=q zVvO3op^H^();4(ojWtMJJMzNwm#znQeeywSd4(bdd!G;^m^iO1na7semm5#u-{O96 zewTsO*IaQ(8th~1x}nxkJHMFowifCflV8}cfp0cafes3z*;IEkd>AjkC!~P&=n3Ol zx<&_!FSB_EagMj!nv1N6G1o`{Z_g_Qoi?cOXWK8)O? zzDaX%`_6dFv3Jv5ouGXQ$&7g`f$b-}2WwjF;~osve30vwnD??5&uGPqx(mS#B&kaa zt#t0Jt#P%vhp@G1Sq3ewg<;D;eDS~8RMkBww?1EE=Q!X{TO3S|xTQBFpW{g7aUQZK ze6aNJ7(_a}^;PGF#g)hkB-@*?Yl{LF$uq4qSp=oRITR0^{+Mq__DLNlGgM2%I zP3fy&+}J5w6#=3Qr_>j+1RCA@Kcb$N5VdT*D5g9cYeAMCRWP?<1P4<`w63XU~Bg;PpWcWLBdSqvW z#p*E%i^EGB7PN$4EI;-l%b0+KkDNG^lZv9wdy&4xZ>Sh*FfmUs zeguh>WEB($Pi{UPVL$S^t286*lnJhuv4?r%hO~)qZ`iJr(*4H3lX19*85;<Xy@O;5wdk{bX-#I%`6nBC@r^b*)|Kmk797%((<;TcYC>)`#AliFR?#rzt z=UG4Ld<1eI-sPHP=pt-jDmb#mb5xJY)&2?%>J{ExW~r21bJVlDL|PuP$*cr{(7^YamUp zE3^1jUn=Cf^L@PNNkZ7>U1tIH* zq-%VHyZn_O9j7pj%ZpvQcU2*KYG&Qm!h>?+G95rTRfja$qwVqT30jx*uD-Py>rNil0`LS>F*}7RV*(U2Tygq9 z1wX}%mZm^yI}}@wkL5LM0Q7v{f=&zete0z^5*_7oynEtCoucMN9flqwDl=lE(u#>M z`^zhvrhuwM)ZtbwfYy->UAhr1!1xUt(Ls__pyxn(4(~WoNbV09=Zj8#GuT&1o1f{L z9M1dR$wJjKyxB!*!`)9@7_slTs9{|X{&P_H4+`5^2FTqhb~)qL(ig96ayAwK_}Yx+ zlI;ym+qHp6+X3dma9t~lt?z>2Va{iMJU6lQTW%Gh=5g}1={ekJ=Y!~qR5&bNf~SKb zy}~``t%e*lBu)EBkIlhbX0C5b3S#(YmW*_ib-#)5@@VB1V8UBVeb<7GK~!z zHa2evc|)FaNdMh%MtZ~SzQFe=Z3g~MV+W*!fEFTuf~S|)&>zHYei_|$zVMv z@igTAV-DoC?zEU*vD`aX8=(f+Ne7E;tm~%(2d_K9MD0|MoniY_rI7n8ONe*@Q({pz zWOL|3Ko;(bc%$b|Ub``_rzV!) znVdR2X+*#8neDmkrIi*>xK|$_NLK*8C2Y0_QE}DLB>}eb8ACENNl61F(ZYi7v0`?_ zLPKN&$Kk91qS9}`WOF-|Lq^_!SMbWh^hy|H>9!)Bh@*+Nd~WnvTYl~l6L;9`)R2#^ zJ!}ei2mXPx*(`g%$)$4uRL$)JrXe#^8=V=nS8`@}g~1+eSHa7hrKQjTR^LS_ThN{Y zT|9{!*2Oz&@}IL~l`DE)-Unq?QAyG6 z4|s!eF|j&Pe1gwG1b9{oq*z;juo0!fIp}5O|BMMmZmDb)N=PReC#6FSeBzj+C47^D`I?lx z^qYWSnM-D06y94{v10WCWc&C;yU4?udU!J zGOQ#hL~_bDn5*qjw`Iwi&+4>qjc*Yw2l!TR6Ad}Z@8ekdC3`m3$Qd=TT6CI8c&@Vb zYupf-`Ij_mry28LSNl?e{Vp$wk9EC=i5Oj9Id!!Jb75~eLe)o{{{!%C=f_eL+^M&V zYa7rLC8j`_{Wa~QP08FFUC?84e7u45>H?tDb81}rr8yEk6c!eS|CJU(tw@VvVl>6} z_;?+hq;x1@$VD}{ZWv|thBxA;j=EF#XUh95r#_g2E>h-GrN8&&H&ID*&k-QZdHLK7 z9j_%s!eH@uqeFVbubiCDBGnzg#)rW*mDji(EUrlaS`r@h*QKt#WB=f32+?U;X@aks z!>e$bspo{3x>N*)9|H{nKO)hobgw5-pRvvf8(QG4i3jk#&5y8Y%4_{n>b)UWfJIqK z6raL7wWj2}qL%dAv#f;>q+Qe3U5Rhyi+TJSHaPU91vIj&B)wK(enahGQsfnWEcFD< zT(Qh&>zU6nn`o5f1+(MxUzC#Qv5P4%lJ%IBlk*8ou|uveZdfp!d2L-iSp<%<;uI(o z7mOgSIbSsmHacYi|A@VA=yIF>ywTI#mt$Z>9t`{wT=m|`PA5BDkr|0s38-KA6Az-z| z)nx_nyoX(Ola+;7dTV*v&EJMW0?=QBM&trqP0npA0}55hn5Ssev3CAxLw+~dp%3wT z(`-PcF%MUW%o~~#uh0QcSHZr)kKfH4doOR)NtgW#l4ZsJkkiboa{xEx4`mi6v9lds`QjpAAriI40S3lP6J=VtiC7~W%Avj6r*}%2kILCC%a;U zYP>@H#Kg4?VIj8fcijv?|D#SS23Qj3z}`aq0m1796Mj{<_$@;V3poP53@UkT;1@6? zd4SWoLhX_T(ZNO(*WB^>LW78`4Ak^=E_y<#q&f}1+5&;BZ6E(|T^~l0e?&I8YwfT1 zg!2-nMSD81BL1;zxCu{?=bt4&Ai>0fQC0c101%MBYPr1Oo*7DQfR8*0cAk(|6R4#R zr){oyA3XGF)e;sU%6>j7Rdb#&%(%zD@9uh8Qbb`EA03E*x4v^aSVQT#dY5!I^W0Xi zZlsLpbr0!RVti0n(BklIVHdOrmfjU5+Z8r&V>jBgY2|}c7xkew2W`9Yk4Qs3rt!}q zT#ShC`wTfES6rL7&s(Wa`1jZnhVhA8(XK!XFFyJudw5B)OHG6N=}=5{!90-6BuZX= zjor=O^|>51(MVq}ippBr3i{3H12|+n!KV=}0Ls^%Pjx8-iMzL3uDo=&jVlDehFQUI z?Fs>l5$oE4#!V58W6vCCp=HF|@kWjH|M#3N#P3*{(>Pt8h4GK^TM?nBj{$1;=|AM$ z-mm7;{f?u+&0gGG^b73i_soWNIXE~@DyPcMa9`NoY2smR_b-d*-dH%9u!+y^e&D9M zIjZMk=^K_e*GPH>z8m4$V{FJj{%Ld0Tt%Z&it&?Xfqw|dY)c}2w$}o`m)dPQ%{l_m z4r?neZ0*k(v${)n2;*g30yPoD`Y#2UXlpN51ZnWqX=`cZAlqPwUG1N??SC=~S%WMJF({*sQ~FWEq+ zLaa|5@fqa$XP8ck6WIGV43x4l^q*A-=x1@#A| z-Yi3~JYyGQ)|2qdl+ldRo$A;`ODE-T@`#~d0hguI&4sEQwy3{7hj`>x^=@-n040xp z@+aiShl{9$peb$e$08WN{Aa5T4U-^S4rimQ^E}VRw7NGR`1rwHtiE)PPz<5|J@{hS z)O}xN9lrFjR17U0twEjy?EGk8=pQXrQ@`-;MkHg&7UV2m z!AeZ@VXAl6^of?)sq`n9+617)ZVva?mQ?3u?l2+AvFsLkp;%G)YM+S1@I!68z@C4K zm(VSu$Gx$xty$p2&|nQyQ)($d#nzePX*sQ|vH(!|xa4lwMRAwOD~uHaIcJ&VxIQs!I^$XxJGihKN(N`ncA|GNwO3vh@BY%{0pHKWUD$K~Kp8gSSO99T65 z%jn3ImzJHt<3Fd@Z0jM)t`vakk&|CupNz~%Sb#52(%X_YdRbXqyhuwgK*%*NwM`D9lMI$ZXk8|aBmzAk~c7N?$-cIK(xPI zdLP4fAS>8Axv&vTzn2+Hn0*w3W$g5|cLlvN|6yd|%5~^J9r-EN!6sUxfuByhnhE!k z2h&s@)#UdzAIzvNiSlhOM4YSPcS33cJc^85nwpi{;6Tj|E{hDO9$mmb&-02Zz3V_nn~ws8DT?;1m8GMYh)S>d&Cv7Q9Id5%TqF@iYMKkdk~!w>~Pr zgt=P<0iD(fznW)FwZ|vyxQ^CBjpdikA4dj{7m!{Y<6nQW4gKF3wGI_Hc>eaTwS?)9 z-@JT1ZudEZ5~?C(?7p+r`ii`XuhQc96^5C+;Pkkx%>I1cg`lul|F7IZp-IC(LB?D&-3*K)5nJD{CTfrdQiSeDuP5@hJ+!$w8g(U4;Afy2zNm{0oZU_ zF^6B*f`#(~uPd-=1U>mDUi3qX+I4VaN$_f%ct!tO@#J;*|LPkd|0h+g`M`(IBHu~$ zmX!B*Y=Lf>?$3H0_NN%KoYSFV!jU`}8OqIMT37M#(nBWgkh5H`fOYx|U@S8$Q{NDd zHc}LQ>5t!?PC=5qBdbHAe?`I&qLkNEN-AyULHFcQut<{l4Bn8vyC8&vM$ij1n^-K)yS) zP!GWheTTPF{sxQZ4n)dhEAHuJzBYm#QJ6xp>TInt{VWlPq_wMqlOK0)MMW6@O$+*H z`nXhP_y!&qc{u@T!}4(NA=AWE>$P$lu30Y3F)7 zA(`dE* zfbEs)zdH+-m{d-URx($>Ci+*Yd-?hNoCnkT3NkQoURC|i?kVfFgYxq!*sTd+Ek%xvS( zGQ3J-X?Q>*J*%R;lLH%6L7uxIAxKj8bTSwLSnz862}FsBE1+cS)?6WV_JmTQuk2G`2xkAs#| zOyXkg(JC|^fPeeBZSSt|2<0ViiioBORsjEtnY(&dm1a@9tqLob z{WD#m)HVQbgYT~>8H6nnMj6%$<4}9$np2+PAatn@)w>dR&Z6Fhk#|p`yWRPAc=7+3 zKNf+3JDe|?TYpF{{2DTu*MhjWcOfsk3=!QT1Dl= zXVCx)-*&Xo>doB@zkcVL?JVg3r%hxn`}D)guAJutg!`Lic6gMB?`c^t9+MvuZtD6h z7~i;IfJ9^x@*K_Y!Nt>-Bi&r@l&N9s@BOA>x8 zSie%}A39qwPtj2ew7P0~KVV|0l2afn%JrZ;(LbF(^;6R7IbQ+oK7I-VzQ6q)SCPMc zpbY+0cW?(J$}M`cD%{P=jYaaVE%3ta^Ez3N+%?%wNowA8XS#R$eCI>0CX##Ij`&YK zmBXYcR}2F4YV*}v%>zZz3Xfu;W1E#;xLrx}8IkbLIWvwlN=O0WY_LxUe_+!`#N)0K zO5e&lP7<6`rhfF@xu@U9cmIuu2n&XqCSxgbz$W26r@t>b^w)2{Gq?TA*4H5(`r=*- zH^u2TII1!j?3EP8jbf9u5GC;mtZBmi-@mW|9Z>ulT6w2#tIhj5<3DS8f2)O?&wOe6 zZ^DJpZLaUEZCcw^Ot<+A4?cBZN}uQimPe0(A$Hs}moXqQB-TIIrbX-}`Igs2=vz?psaLS8e0 zGy&?rhLFD-ezCt$o$x*N>uBWk*{Zf8_%MHPs4U`YWvq3@M6$Op@hm!yMQ&qKq7M4q zwT5h3}TKL%3$49NffG zb$lSIx9XqDF#ny-9Om^Rt@uq~>H1*Rb#(k;$)5bUm!?sNn%WJUwBE$L)r6qjk}o+P z99+nbxEvegwg}uK@2kkA$AN6YkYCy@W4uO-5k`U6*`9;@izT?w_Ib@&i-4uG7~AHjtYF zmyI-rvK($Kf`=aL+bAw(lnxuiq1dGnBkEXP9?TN|O+&CikY_x{rt9)y!+y+Y+}7a* zNW->_R3@^P$>N z=AiEd;zQW?**v}A-G3Tsa0tcoORo{KlVj6a$Pui3>a*g9^`at}^3AIEFL_7*B;F1{ zQ8-oKviK7>7QNh^ckjXGMZx358@)frcxy&ik-vwNT8Ov*{OB)e;?qsqxVv7W`u`Zn z{bHGq3-fSli3^Z>>wg+#!NsHo62@!;`InS7~)2!o$`Gj3|fOLssy`YXuAll`{frO-@k zsL!%s%}7_@G*s{v&s*7UYV(f7yhH!imPDz6*~9N zu4R%*k`TEu=SfojM2F+cDc_jkv9hNnK>toG?<}7I>lL?RoFL)NBPbZKRy@RkoL7R1Sx?B zirFJ6Y9iD&QFcY7fDZCKSxA11VRAUDpk(7!OCe7LWhSj zf^epke_uPEowV{gglCwwqq^naioe^!ns=V59a6q0iV^zRb&~wy@Z}EqtARGuc(&^! z)WJO@>UeI{7ApJ=wx+g@k8ojt&4q@U!GR`fs_&ruI(BrH97^&eL;ZxY%K_rx@5LEO z9Bq{}ZePm%1r}p+eb37>b*9pQ*}OFJw;7kgPt-!4KD@L zCf^d?2(5oaSJkX@DPm4CcTaZml}59++GeR6lIujS?M|RRT?|W)dYaj*-uZ)et#=0l8!ys$Izm=PkdPS4u zHsCFKvK5y+yNaQP>|3$A9IBLLdj_tM%i?0uaZBb9l^DG4=cy!^6feyWN0Eo|xBMnt zc-ND3Gu~n!!c!C9LwoNY0fl$^nPn>S3WrXwTOf9&G?MG>JZ-vs2MmcH7LzwXr~(bk z-iiC~*v(6v4ZnaZ1R8h#eZ0!bO2wQ)GOk6K<^A{qrV07YmB`QAPbE}NSc{+dtF(Eh zYV?SS!)uaOoQ9Nc;)uH?L5}NYM7M6{^kZDm*m&yT(&Mg~$-WK0Lf+|(gUvpPoMy;B zP00DOey|yS|M|W;`3Jy!l6GD(aE^0};!N4Y{58cFz|?L^FLC+g6*VA@wD+|uS)rIK zdvD&Yww&Df0^jCunvSV8;Nciw&3-*oR(WjHg357Lm1{hvZ$t(5ETlzTMLmICHL*ct zx&|2q?Mi}ZAFCCPa@&<{->4)UBkkinT)&TGPE$DF8^nT_=@G{|{8AXa&V5ke)2AjP zpix0kOV95qb^HElI`8Zn7scar@8{Kj3!dE4;DUd$+m=Ft6SqPJjQ9S};_itRbKM-` z&aMh7l>PaY9!&Y$!}SOApgrqYIbH86zBw`9b5QzUx6Q8|Ubc4RSX6sesQ3Kc!;5l- zr}^6k(k%Vg?$3KDbIP`xba2Sj*`;Q1<01>4m|Na7m8I0+Ed9RvQ$jGsM$g+*NQx+X zVU{5k?r?i+8L(Uup8*4xQL@Q#V={niRwb*Tl4)7ZT0!b7{+hH_t((<^7z<>O?Z#xQY+uDsxCY)`)7U;LjPo%dUHB7 zc#qS-*5R8`;T`%%jE7IJBF_AV{f)z497icoKLRAZ8AeIi(3;OjuBs*Y4`mwML>V*z)Jd%qUOrZ*yr^F3a4`Z^-#Dc0R_ntuM_nDJ zWiK{WG$AO{b@@|oQQyqcGl>f$;LKiDyX49%f%k$EIT#MCO~9X8itt@)N}_1)mSsJZ zwN-`0yn>_>E_#=R>~la#q1GmXF~|HLdG1JsTw{96!0DjazQ3}!c^>=-bblPSPPTaG zSU6PQ%^E1UjBpJbT?hdTuPQ=3Oq(m2>zV40BSEU0{i(j_op7c)CxsOy$^@&kA1W8L z%cc77u3GW6KnUc#0iZVWPy7Qv$PZxFxNWiQkS754m@B}lAAL(eA%~2+Zsj0Z=g>q) z6L^x3u<-JmDyRLk>Q8#qb`5}1=Dh`asXJAVv?csT?lk4+#T-)}nhXP~_^d zD9@Jn%&_sUkYZtrofNk7)1+q@kh%VsOq6fpKh70?mgWp{dApgy7j%oq(6g_sN`NfqA1bXjF>7l$i^*8?k9&1jEaaTi3F z2JGbNNH}XuRX>RjdV}&qcp0!COxAs_WzF9#23M5Ze}Q(yQN{sOj0;qy&z}B))%Lh$ zf6}QtofK@<&%c>#L?*9759&j2QLH@{qabRh5t>}ic*SUB&Opu+sC5TmXD`rpK&$Ht zbf;-DczBKt~B~TFOGK7 z`XO5V0xLaq)~R)+&AzA{zLgh2(ZB|;^R%>>E)lvzcHlli3%ob>DQUl#i!s#lGLQ>K z>#8Q-xK%>X(+JHbZmG1-E^KdalU%EeZVnhrng@?r{GWnKZ=gd7>{`@hgpm`trD98+ zRu(<{fr~ey%?iSy!@|v(*?O(Hj#BIz=wxNQWhJQ0DNzf2lVN?siI?UOF*J9@x(nn6 zG$TEX*T=T_t`G7|z&pZ~rXeV$kY9*b&!4~-=xwdekgZ=z z3@yh~SaOg4p(jMQd}scj3_`YIiyuS^zZKrMQF8q6#p30b0E{BWsESB&0yA%WGDdM_IH z6#79zPqY0l@-Z84QyT^a;X8l`mTN{BNJnY$o()vJtYWZIf_pE_a>}9L0@Z_O^R&e4 z+c`CqKHlwRG+ro0DY!<`w~1?=%X7xZ%)8epp85>B`_%=SXSB8ODcgxYP}QLAx<8Lh z@?}vADOPGkUQ`Q7>K=jsnN?k1eg!=IX1^;;dNQk>tJ?Z)K+_rdWvA3YAB%dq%cXV1 zR3{3yU`@iMNw;3d{BePt3su5Wj4%2PCrm1jTmI}OdW{+=@dzfcp5@1{xE|?hcnN8< z7OG{ot5f!@y;=NAguN1@ghH;Vgu2svB*S7YnkX+6D&+lrci1v0SB{=KxaX<|ue8&% zA;43}eh^EJ?-Y|#>Uz&UqN+Lthsuj>8_IJ|4t6>5D6{RS~-Nh(gl>PI{LVKx6$`x?N zUFc}446VPq{W|676)#+V(re=bjBplqlc1yGw|1l0Xx*ZB&9oJs-W{gSpx(iTIVzxe z-+uJUYh0D7*8nvUwT2G=w47W7Ez_5$W{>wDxwAa;E>CZoT_SnQUyO~G!5S4D*G@VslmG4v^)6z&)kUJdpM2OlBSkf|He5#CH8k^ z7_AIzTjF4;mF7%)Q%t&GZ&6X~K_d!?(o>8gcS^DMeS&%i5%%z-+PSld&*_Qv`9XKs z1(i4NzdmUstWNH6?JbWo|Do&e6GZ?!jY?f4pOE6xNw=#2Y?o9ZO-2CqEk#;*4g%j{ z%ZbP^@x#r}RIA?kx=|k&;Bm%lQdW>{6y!Cb`4B$*S#Fmgiw3lOmDDaEYr*z@SL!X} z;cNeU{`aV&b-&5?EMb-QtntVpWjyPf0af@HZ*63--S2fawe;zevM0Po8H@OTsZ5dM zjYF%{3PN7J0%Ru_>=YljQ+AHl&OBu*G5r(3+N%eo5Krl;P%FVBL1(g}L{MiMA8ZA- z4R76>Xz$@&(&GQzt{3EQ%J;(&c1(_s(wOKn$(geVxVe-Y_7o|s1@UnQ#GAvYk2x6I zCjtx1qh<%yU601i?^#@|WxdTY5u7?`6Ynkt zxRxLP_cExSRevy}%2NTUhlGCkea4_f!eKo0{5Y1AH;b7M!YvG(+`NTnRSn5d3i_bV zEXr2_=SufFqH~2Xv#gvAkt3i(1$O3*QkR%M4|BA>;y^gq{4ilhiL3dqB|raUuwN-% zxN+&r*{KeRCv%ZaS98VT=SRkGX^IY4Gf|iVp|x#5B)c0+OppNwh50+fR%_Poa#?rk z=F(_3W9jg4=IrP%?HgELZ)ei~B>rwug&kP1KSennF6T7;?V>Unw_eboO2SxdowZr$ z^nd6}JKDJrUU9M===fezbI%{o8p!-~%bJ1frOc-FkcbtaYjq1iLmWN(W6{AcZ{mHic1y|xiFn{)*%H-If646@u8_OIh7oON~{7L8K9 z%5>dzrNDBTx<0y;B+-cd7|Xb`Q+@G?A^bgf*Cvn`;FT}GVdSzrIatxTf%$}x67xKu z=4BY!)2R*oMJ8%e<_L+*4djEtLN5FVzq|s#$D*g2XqgFQ38;lyOEUs`J2OUgzq|UC zyej0SFO3CL034oOHE>Ee>_u zwqFMrN{E)twjd?mv#`ne4$PWdu-0PI;Ci@IZl=1UWJVSX^Qa=|!$#o!VUrGEcuv61 z8G^xsUk>#qCW1F7MifH^gwOdAqM*L{%W|<=)7Zl7q}UJubZ)uJ5;E?)`Jat zRx(SkdT(E%>fi@leeL;<_1cjV0pENu6A+cPPCKp?K8Gal*r{%>(JioN@XZ3QHvS&oLEMsg*WjBN-GqNO+eaXI+oiSMl zg=~YdjTxD--S57i&-ecQ?)y6boPW-_uGhJ)=kq+5$6^nV*xEJbhN@Og*I)fA(>Ma% zuDPay;~6PWozA`aRWAr-mDKvUgx87}F%$K6T5?}3x;bnk$2ZmYIQDGogFx5@j@ za(2ApV(ssO{dYlAJMGPwo#t5O%Zi5We_f_*ttx0KJE@l>jyPY#m|V5eWh z{Rg;3jf!Vys-%aW9WcU`jZwO)L7i6f3SNMDd7@X~3;i!*)+GA*#asW8`~#H#9c9^V_c!TP$@a*)iP*H9yO=hB_so6F z0ua!~Jl8VuJFXp_gp(6wXNRLNy?%vePJbfK} z>&~WX5ZA}0S47?GnF0gV9{`iQlc5hXC6$7Cl|s$I9HvZp{9C1y43JL0=HmO7cw?3r z7%HQ>%h^50H8&t9G}Xh`&YeN!j7EP!y`tqbQn72r-A7nl2eDxKa<+ps2)2>L_k!Q0 zTV!)6l{~qCv&kmoR-J0*+;9zxfhj4Yr)9tg;{&frT#YWdsuKw{weC>!DDiRkp3_Zv zMr6_7>Z2w=OqH)pylIB}FLKZIAInB5q#6I!g1q;Im>zTbJQ@FUjO6QQTGZJ1H0DY6 zmq5UaGkFXt{IlH24-wdP;VkD*+}?bANHo@Zg|Z@bP0R-HXpcU+sLjl01*DUbQtGpI zX}`z4-^YlCb~8g}MZkQ44uQi3HtIWKTZDZ;0$V%=>9d9gxQhq+OI~4|pdz2vK2HmW zQh4B}HjE{MRaxC;Hn8n2bXMMF^*{W~Z}cI9Ht$eyjIW<;^!K{OG<_1F^$7_v?&lKE zE0KeagTtsD?F9L~rJCOlJ0;mKv9;7w&mN$<06u7RcXT{f`ccfS%8qyp`ut~w z`!-G45qqqfG?S%XNzpFaZ2B!Y83OC$f$|U7Lr71mwt93)ws?N(yDeCC=d}Ou;^=$b=1)0>)OJ~?ztmM3v zPbDmS77W7`(%j^!Tqq!u&o`Bb z)E;YXx$ox!#Cr;-|NysM0pGR9W%y*;pPPQ`szl{91 z(Y9C10&XTNmc$@`?_Sf{&E#{lbRI7ci(U=d`OLfL8f>_2v!Fk_f;wcEs2i_FgSM>( z9KEYD`D%IgHnm)i@;&Lfjcx&~1+{kpS?4Bys+vi_KXm++WDooBMSUp3e8$Y}xLT#N zS?##Ol_x4tcK^XATzF_Jaa8iWM+WahldmS8W9qD$r+3vUpXJ=Is%`d^ z*YIFC46K?q#dvVu zbk}uBRLeDqFXQ?8q=%!l%#3Y@wsqK39|1LruVPvOEmC1}O27(pmkCN@xnb z9V+aZoyx}Mw|D@yL!V%|xj40m)6-yl*`DpJn%LeM5@#0btOzqwu z9h1@Rmn|)J0v48f+6$qg@7eXJqmO7*(%lx$FoV29@^XT1-Bqm{@j&X#XsZkSkVOL1 z2furVDDb|NXp=CbzA972X|strV%cG(YeV9WQipGTEe`-#U_Sh8c@57AS#fsZ zr$ir6a1B;+bt`*o@6c3#8(dx>T8)vf%{1zL}sCLEY`uMc~D1YIi5gX2qBr(2K>UMTd_F^1fgY`J6z()Q%_sQ zJR~GQJT7T8EVWGq556TZ=m|uz@pC2zqp!H+XJh#aeTGS6BbHkZ(sea2C;h2%OK)jxvBO!p&-&PT z_E7?6P4}&Y!B)95=f6dm3)xQ9;uY3$7Nl%$79O^{A-HX#X2sGbMdS=nb(w{;6j4h+~QD z-b&_OxXuxL%Lz{&id|WWaP6E@CZw6b_%{-6pvqbXZ2*xO+A-d*44!}GaWdmWo&KI` zNo|unIGhF_?sSwQXU~8CDfm`aPW#qyRK`*yoZvq;^9xmk`7t8%7m*Dg?34 z@V;}1N6AahbwgB1z779_9`_tmDghi3s*Ac^j@jZpHnM9nJ}jRhcMVr6RU85ENSRRN zy0TZ6qw&f3Yp6izZ&h@2QeQhD0R5X&H#G7!7!k^Qno9?*oZigMdza27y;CrC#NFPE z!!M~?O?c$`;HRBAuvVsC+~6rRo8rQ499o&Kp?o1JqsaYnl3M=^I;*JaV zJ#7Z3OemEPvC}ffu?W%?L}57|1#@vtYkP04Xks>xHvjS_!DdT6@JRV@sKOl9R7>dBGM@uaz2@S_GXs??W0RAvf%jd?k60})gP z!cwz5RO=4V-0w&{Wp8x9+za-~<~+UZ7G_!Ik_R^67m_x;HNq9x&FX&KdgZ1Q@u#j_ zNj#K5bpsMzQp&PJi?n`=&*9wP>a(mMsFH@vOA-cUPl&_;8GPf3b#zj_OTlv$PKe&F zOY-%2jCV3(#jN2+%|7PKWJB8F-27C!6ErJmU+?E^B()Uh26xb}`a$}seFL{_PiI1c z&0#9jT?adf-?xTZ+ZVNQ6Jx_2E5&mY6!B)U`1Y3ui{hPUKgO$6U27VW4i(%4H$quG+sh`;JJp$Wb`_FyTGv#Djvipt<*+djOuYg&S-xkP~n+T*_ zLnH^rmE|t-)-8B5dR_gK(!a!T2iK5{*uCfLE*O#^EO9m0#Q>V^ZUlI5dd%us!orrp z+%RcC71cmbDHex>?96|NkelxYO@9k{-C!N&)Q z_HP#{q6D(=2x@5_tEUp$uHD#P5COv*E*SfH`uASnAtx4F^~}mtosdWfP`@%;#lP~yErN2$h5zbC5x%s6026>b@n9#KgAEhj870o0*pT$Mu}eO6;&3UKJoSRC1v8u%8X{2xuoAq z7~sqXS@L6ZTV;*prPNk^8kI)lf5Ad(^*W($VKZuW_`>JAAzM+=q90UPif9l;7TZ7~<7NTsl5AAXXFwTmL#6 zXAfUdw@6KW@Bj9;N;`0Uj+Uj=8n_k2n9! zN)@(q(Rzs30WWS#QKHt+vnYWQb=q3)I5yz56Q&#zQnC^_Qbzxu(;D`1t>|ui3ZR|r zaZ3_;H+0LnO13%N?GDTpuWph?O}8-l3|&V5LpjGA8hTQs?3j20hotc|v=n`k^|7|^ zlt((ZRhFPK-Uj-QgjfFHO97j?qo#>@CermsbkC-9`^tK)a>2s{nF{)gK0MHqU195r zNU0Z!4418VcaAzNSr0t@P49GQ2Z3Q9f-l56&L|H+M6QMM9j$O(RJs7?h!L0C;6L2R zj!?kc?{GBJJI5v*ex2TuKi2#j+?fgC=ttjl0==ITp~?^;3o2SY#fK$Peb>pL=NuZ% zMo$}4k|N@uMT2;nk0)eGJw0cx)o{ufQ9-@>D-`-L%!FQ&5mY~xd}+CR5E?SvkrR&b z{dU!tN?p|^vTZ|Dh(T*L5mn#0C#Fp_RwwRV#{nKoOXkj}wtK%dq9~dN$Mh+*8GXa+ z_!Ji&yc6mo72Pd!1vZhX93cyST}7<-;F^a~0loJce_r2T)$v_I-6JSDzw@~tMVfaO zT3u0Kr))3hw*lH;+)H>v>u741t!?+Jz3^PZWPb=x0DOIciDY4*OV`mX!ZoqT%(4hQ6>f zVqJgOqfJ=xZx*4)%HzhJwrHkrAY!>Z?1b-j$;l>T(bhOvK(0^9XStHs&Rn2dKl%O3 zxF2OLN`P62Lc$3^n$CKC%wGg`o<#ai?P5ewRa>eRGpe0Yh1|&N5g*8{3yqWV!1CU6 z9udjd>8%~MA^z6i&L&1cnVoY|>64g?rzj_b512XwKo()OerQ^3$-H1$9!UA_om2>z z9*Y!oPok;W8JV+nS{CWg(6L!+dc zZafW(VcjC6U7EWun?MNqENRUcC(A7k|Lte2Ob(sDAC2uYC>{vDk8oPjI}AkJ=)jEL6zRcC^Tj+g|`1 z8B|y|o%&O-u^iKv zPgcMy44tg)iM_Q($x#UScIdPt#MjasOVtOT@fSa6CB_a%MouHs`WN*m@7J*EeBNzA z4g!{~p52ruZ`F%m54QMmI`rl!0lV*=i3e0yo=;nhL3tW|(4M�fQ~AL4oQR8frEYeUw(cu#1>$kl zV|>CJ6?&&#RP1nauq5*{2~ts+mz1Rsw7Drxnq8<1h)k+fuaiU_*YoCBEgw5@=umql zn;!hcM1pvg<8+v=pDo0rVS-kD3~d&@O1UmJ4h>{=y_DHFcI3%QDkM~vxN~1xxyv~; zClmIoIhVbMLv)^(jes&Tf(I5p!+8?MUscH)KL0}d)cbrHfEw=(C`EV|RK1G-Ke2>GQ?4j5|@ZW%H;tF3RxD^+eVj*mu9X=PH-K+*ZJ1 zhK37HOi(4O048>v?leA}zH988+jW2ZO2+(W0i<+IE?RBZG$;@_>@38}Eh6uk>w6g# z_#kS%5A!i~2Zi3gIW(maxj~{ke@Z~n%sfj3wjZ^sQ@n9a0kH|0>OF9^I&DT9JKJ8V zvI;vK@bMszMxXS5L|c+0VYkaWUe;bli@U^s+?-CZ38GJbu`cT}2Z95-soy*vGdG7d zl->=AEOqm?=PZ_#j5)_tvO#cUtBXz22fo0hI0YT3spwu%??1HwEQt{!T* z#7kYda5h}5ub*JJx2D|D9=?m(`wh%KTnS;dZY34NLMX?D_-Pp#8TOdY70sPyZXtpC z(A33}zNEjFLt{c7QRRS7OoR{f<`zJ618Vr!MrKyGcd=+`WMq}U_eyl-{p&U=rJUA`6T!Mf*>d2M7;fKdkf|#I>frnc(b@%<)t%6KfBK2^op9pN zCbYIjSm^+|u=l3#Iu030f6c9@Huf6QnTz6Q3Sm}+QsfuC9sLUw62fe{7cx;%2A%D^ z8$6Vl5J#Fmq9W$J8iR*i9(HIe)R{b-utDqScgayei|x&&4T|y=$#|;o*C-lu#5hDabYJeHSN*E>}|!EGpAQ) zz!0uKUUKh`bF|6k$@3sII8SnEC0qw?}|j30HTn9#ei-etNc&ZV_#t@3=^f4{;1RRT@SN#G-HKOM)%guNcGY8i)(cbiN@KZ&c1 zxs(PtsPbo)CZe;T6fQMc8;-Wc>tQGH)xVbiQp(yLp6a^;I8se~rgiI}&y6S7$O;MP zYdAsV7sy|<%RM%AgDF1@r1&3l3fUsD+a;RgFQEfB5%{pVkMSN8+=kentyM)~z)aug zTfCr><*P-!G?Cap)&Ojv8p!01dK0$ZKf7?lul7lF@vEqz_OB|n zGn_Bv%<CCE&tHNoZLtlG$*J}@aqRYu4 zrB0xw@$|E^(lCZ9hKctvG#Zi{!XZa|S|AkUi%~1*5gD;1_-duR`akHiOJ=`YavfQRJG!h|Xgk3lKxGFEyNl6_$SG9w?l$w+g zQ5*LVh<3!oiasoTaz&Nk#>VU4g$|NF=NZQ!h_VraN3h=vGj^dKBkvTpp~0Hk)F@hu z9J<*CfUW^dMrA8QH6Wt!NX7KCTTTPYbXT6VU7yUQI z5d71knL%1f`V4W=J`$~Pvb^PrMt^Svw({20Uz1}U?wu%)G51{x?bj9jt^s>-ozV}7 zHJRrbbs1it%M;SM4)hu%wJ0h62>a;f%LAP?2dVEZP#zE&D>`NQ_?T=mE`%XN3crdy zYguMgNnD+7#JHy%8!uh@D<`-4J1$noZ$X=A@W7_9i6%EHRrOz@#2&vWRV+aGA6P2j zhtsh2-@Yoe_=tbel^(bE&06e(?LD1Q?L@bV={MBO70IetteZby_z2&KuW}l4wCzXM zb~tdc{u?c1J)e);I$%D9d$z@m+Uo7!eC{qMr1oNUT=?$VgsuD6?r0y+3p5ez)lrBD zVsAz$9wrI7mTdq&-5c23Gt+3piN_vn5fQV!ebK(-Cm#Wu=am)EMUy@6OSIq1nrKvf zznXU3=G^nIpXm`Fshcn?I*j0^lTymrK>1cXbcPe^jNG-t{08D_9r26^rI{Ko0i=1F zP~&#^%wLuNgY%q-jY0a;0N|N96vSZs}W~LB<==1IkHA zZJFSiJt^eW@9e9!@}oA7ZWOOMXY}C8QsHkd1&7jIX5?hK5g!zUlt+Zdh?3Cc>WIMQ zY>mNhe0`d|^!~DLU{@#a4%rNBV3=GCAo5c_)TjUI-I5x0$&~ArKk)ulWs|+ph3LR> zH~1`RNxu7GG~q6$m+>m9cImdC`!(NN5+z6J3{|hA4izOikcQ+$Qrcao zNim{mM#G7Mroj%ohQ+U_F^Yqny^1T|IF{eJ?zW5hmYJu!DR%8J$J9R`_i0p0zB>1}F%AH>|qioTZv z3S6pReeSrGCeGz?@F@>Wske7-680JB3GojUoBoUMAI6#ALe}p*`ETRo{DX_&$|p!L zQusH)$USTpM-4PpKYPOh^$RECXWQf}waIG6A7HewDjnAfJK)#8!~0Q=g(CHH%%bNF zD`QyxNnOBc;Pow*iJQz$A7p46(RAUPVZ)Bvf^$7DuvKnqf9*Nn1IAf#MP3nKDXMqo zwxg{xpB;eSx?NHgVZ?{A@ndAYLx#5AZQM?_8TmA#zwzjbc}=R6T>#@+7_#h6)dF|m z1^f!((`xm~2Lr$DFxG~&!h6}N@i54=bJY8}=OifFPUwb?cX`KEEcCt;QrJ%>57{)n zOJBJ;ue_o@YNa~ze0fG5b#Xy1*#5t#&UOj&mGVE?;CIv!8t_iaefiyclO!IILLHo= zRQC}DuxoHw?iP4@;y z{tQdJ$QLT4+^A@|&Wj6_PNyc*QZklLQn}FMS9l(b#VSJk(KjfV981t>bE4n{=!j$q z=X9uv?lEO9t>JRl1v7Uvhg6mln@1Q#`T7kd(CaEwR)acIQi@U^^Q4XX%_(ve(7LKP ztGl_D*heQs6;fqI&V(sVlPb4e`ll7%jaQXY`!u$>ge6XDRczceYl8YOYENlmiXjP> zO?vT)Xt^?AMhcG(d694aE)2?i^6N!+WTX7CEz4D;9W)O zvs(aA(}Z3Sv%Ff3QmvunOmTwIIc;JseKraO|Ls>^jyq=4k+K={WXa_n;k$PK3@7qPORxy8tOtZo~*K~60!-k5#oPF`kx z|0tTW6GRAnyL0j)QEMnOd+`L5-K$~=wBr%ohiZmqLccq6Jy#Ju)AoitSVU=c@BMop4z>t?>Nf?XQk!3nD>$14KmR{@1m*#_>WwXu$ocuXPSG<>$yPCK zmM6HLY6Hhee{$|8TyDN_r3En_b*c^X?mD>f@_Jh_V2q}B#hA{OcL;MXJ_e}ZH3K$= z$*+B#jVtAS5izy5pb-xfLp6*;5;$c)a2ImQwV)0n!?u;>*vgK4;i%p5(d)koQ2Wba zf;0M8ZU&_dvj_RnCFk~dWLmv#TC?TwUKQJ+Sr+m7ph~w<2LSXZx~n`f7B9|*#Lo*@0Oj~>lEd2J3&97 zIU_*uy@NkZh?0+0qc%aS1L?t)Gx2Ot`Xnh>=ADA|e-KE=|AL!$ zEONHJm;U_ryC>jrnB<_-73 zJL!GuYF2-mw)W2QO>enz)|PJlC0X7Xb%JBjcJ_i6bbelt2LRw6x<;eLv^7KbyHZvw z+_T+-yi)`@or1$cD%0I*HUbtmBWr9b$O9ge(SIol$d-0kCM{KmqxRnuZ`be#fBSe@ z_$TM(5@YzkLG$rKv0GqPqBy-PKQ2KK2Mf5B676kje&ojHLUwwT4mOQBNLUZ^<#-Qa zKKHTeu&;8&l$}SM8ZtMR2(#hYt9v>yHEq1}2MxESzwNkduMzDjd^72V^Kp=xP{7;cRN3dX zT?!A{<0d|Anl{!}nl02`{!+MUX#S_XpKNQkTJ{B__B;Mk&;>)tVb*+YX>8--7p;Yh z^W8~CjauD`W4i_!rS>O0-cv!{vC+^|d#!z8BY(*W67L9u)a(z4cYI`^9XI{mO0OLo z2i+?o9&bPP@EYcf0xi=RG41h(t0~K_e*nm+gg&ss{q+AM_At1`s*s??esg&(SA45N zAK|z4Ubu7GJe=LAQr|#vheHWEd{lN~JjP*Ym`~!!8<<=Dwanrcb>)8!T;5Gom@N#T z0Y3J5S8K~o{M#UoZ-pwd>isbJl42D-Z@bBhAyor$l9Y*ZeHWen)vveSb)pf_`RzHXB{A$n}FIdy;P=~v#W7Yg+pd?C%!$_8mX2QR@(ywzWy z$ENxLHd`&mKhNS8itKLse?K!Az2)%Fv11(ILf-J&82@e@SWp4wcJ&H*=eg-1B>S1QiXDJDu^_h zXt<-ly(@7)J1*?K+V?b*sy)T(A3N>s|Fw={%NTiU(#E*KD`jl6gsmlh*4aL57$(@& ziseY`Q^803Y|ASI_;A~n@zs^_*Exw~_J6d++9K3K@2-ZPn|>+b{;4uu(4tgCOV<`; zyG*}-kg=71B7ZY01SeSEAhIEWwB(o_xkDG_{X5Q0neYJ0EAi!FaISlqvxgeT zL&wdYwV&Byh*2K3xxDy2T0Y^`b!p8pz2>|GL>6Lj=ym|SxbO8xhkugfCB1_J(T148 zFv;&V87tJ(EHiDjNlRrh1p-V|kd5RQJb1pUET!G4PihF&2)fBUk>5DlV#pURD$BNt znm}7OfnGpEI@{h-AAX&-*J&@lypW!!Bvxhu^+DY*&(9Q4+P`H9mGbubn&+XmjGNoE zqKiudg94l?Sj7!m?KuXislsE!&N~AGQ%VnFzneEK-`6(`EW2NigZ#?vx()rXyByye zeBe?XYk(qMm2Y5iqUd$bB6mQ6uh7Qez#V47;uoMw=a(3;I|t)#B#CBmBdF+y!i3zJ z@uhRfjH_CS<`QL{cz0a!CNOzLh@O2v*pYnA=)&q5Ff-D>-}eV6g0);ibn_GWEu8|n z-|iHIkE}yI)Aqm2x8zL-X>aTA@4pOs_;m#gJx7IY&xJG5%YRlL1q;Y}N*$d&jT}XV z>|Yk^EW#cvN?%pz?-B{V0cwT}=Y!Pj%D2`v!xzQMZ~9O?VA` zu}6~ldi>C5;-lu7V#I%jnWRu2ohZN=9ZPcy(H&%2XpKb^G!mHgAEox+~nRxi47i}ybI z@CZ}lS2@I?_CP!ldXZCuy*n9E3zeZ5AE6G7N_v*wypD_t&=uDnSFYHc7xWLnO!uil z|7h~qcB|Wd=S6q%E}z;ECa*lQ_WjJ~a;&=&ay|QAjcIV;fJDaXff@dX!@$eKo!i=h zUG}QbVF;SOqyRVUA-Qk|063Y^j8PJYE#sQkOdD{?^x+@2GIlx3(zsvbCJSJ_4cVyE zVP}e6cpxc+>5@CunY=nA|H%yIx^Ed!mWPDjdx_9Y(zB_U%PX%^J^kS2)7hJK5ii)> zwTy%{VA6T*jW4z3f=~okHxB6BpKXTY1&Am(18N zE>i=(U1EZ3FKp2^F+ZlP`2F8d@RM^NvN5AsoADZ0p1@Z-1%qQlMbAFen;v#mRgMDO zOPzHalbxkk0+#!X0*{V=wtsdVI{Uhx&~q1h!!P^IDAb zm}KtT<#GA5dlQ#85t+j8gZK5yXrvf-LmV8O;chy$Z}TMHRZ2UvcIMV=;xN*dPb{0_ z8XK*3<70n^iNdo2{VxFL$?A>d=^*Hq(H-xa#qv-h)*g^0m(HFF>lW{f26PQDH4Pyh zC3Vjn6#WM&W0j@9g~^A%>_hg8Q(>LBJ?mM+xb_-uO&_Y8jdI*RUDdibj~$sA2frp6 zH3@hH6S+1!F8c;`X*_eo^@2BmrDHj?s^sh;{ioM6^pnT$z9J9IEQk-H(gHh+FKO~a zKaSvrIl^OK#3N1fBVHa9O%)i$v&AeOJbd|CcWRX5FKfAF6SId^BP7aB)REi=J|Rc) z2AcUAT>9Kw_1i&|Vbu`^s7cg8fB#qrXDong^Z5muhcY-vmAVYKKmvs3{Bt|6d#GPZ2K(Q~zf}{b1qI*CGyuW)XH>net0|dtZ}^ zs4eg`M|-H4?FpY5V1TlLn?FjlWP8)=w17g@M*)pQ>mV`#fN_#I;$h*OSx^YP6CBoh zz@vC9`;!NetDQ9v1y|b{)SSE}d3x!u@RsA6ghNXg;On&E*O7O9(GL!CxmjQCB)6Ds zB%N%0DYzsURa&`1)0s%=rRZ5Yj+bRrZQ!c-1gXcV;oHi@< zsN;?VseK0e<>Na{-+`8 zvNY5kkwxdfzB38wWhQfleJ@Di_-Rg#As~;`x9iy^ zdLYWL&l>3(0o*7~!rasFRH1LLL+`3}FSf?!*S&g z0W;fsc)jd{%T|w`qasBWlg9>`>8F{NA`ZNzjNR`DydL>9@vLeX_)m zuD&_Tuf0EEv_pypeSPZ?iO}UI8r3F%0NM{2$I`E`7x1~zfti=wgl3p>Q8C8?CCp^3 z-ANs!<&e_P4+xHITH&61vtDQjPZg~Y84PQ(T9K=aeT#A5lp$KwQE$3wEcT6#qem$a zB%UhQ2F+k$KWbLvSKd>ZlLV#7(NwZ0WXj(j*PiyB3$Y>Abg@7p=oBNeJ6DuL)T;em zWO5L`6|Lm1Y8lepbJaK`(J%CAsqK5Jephq#&PoX3e!FK&cz;$*;Sx@;>)&tTMS?)? zz1;Tu)KGQRXYn8zmXe6WoUJ9m&=*!iahUG8X!Ph;DShk>Q9cngyA{~z$xy9;6<3ts zNFE=mwyiSVLVF7{N2(irb)xsC`2D_B3xj>EqcK{D93kyr>9_Ya#qj`@|h|pn!3}@7hDf= zX0^h0?>_(+Nm8gq4baH$lw;&i4fQ@|4+f^@%Y)U%?iZp8h!SHQCV@#|@VVyd#$3{c z!S)Y$o_ge|nyOdTo0-LFQRoLQ#c(ld-*dJAB?ipEaJ_Hn+ed$v!1`bDRHWuD1RiR? zJ!I)g=yZOLeHQ=rP|y{79(ZsTyrhZ^`%{+MK#NIj+MAaF2m9H|sG&bogG!b6sm4T16fb3(Q%@}L$K3p&C%5%+y1{`S(m{2*2e9tnq5Tpe zK_F+4sG=D(=-|2RSHE<&mt-?#%8DVFN(0M$G?9gzV9rWFtmxkZ#W=@nZD@UE{pYqfd?a$>=h?{-Pt z3(jCCGoAJNE#9@zh}K9?fnKlKSdBR7w&)PEiO)`GjH@U-plKr#A~5bLW&8%Br99`1 z*rhyEB4TL2qb+E2wZMmV63{mYfm=3YEzyycf-@r(OCy+k>)Nt{mG0WqC{j$+z#Zg! z*U!;Tc8I1V6K#w8H4PxQ$NoPLpZMtBlJf_&S6|OI57z$-tNR~6XUL1e*rJ1w8ZW)@5reJqm*r{{JL0W0DqVH+O}_a^JsMtODf zOakyPu=%#$z3O5V`?oOGT%VRed(3zd!LtafmsLyK z69h%J%JubvIgx&Y*tOmEPb!BR-_pExmK_BltP*Ls&9;=#;d0xaQ+epr{tt*{f<$oC zSAb{J{;CHUFY*Joz7)K$a6Zwzk)ASZ9)-?n>iQ!vjZ|5hiUn7U8A$A3MJQkDT(lh< zOVx6rN9%4Qi%O0LR6b95U1V}Ba3+E<9YQHBa2`Qo={QyFpSr17R8HB_Uu6t(9;0p# zzb^hR@#htNT+1L>rl_5Jmb!!ZZ1%#WASDKd{7o?Dyp zAy0rCPyC*#rK3YX!gVG0!(+vd);h%#IG6dUtq6hwK^3hx+Flf^!x*rA9<>dyZ!q3& z<`#+(l=eX%VX3PIy`GbDR>`hwHB(aF-ON9a4C!!|6JeyRni8PS8Wx6@dqjrvdtq9AJ@dHeLV8z$prSjt zm))Ir^^3IKJ%9FJW$}U6v1442GQ-0}LrJm26_DS91qF5gO!s&}lLOO*DL_J7Wr)$V zm(zL2le;i`VnU0r|&KjnDC|0;$h{iRvnp^qdn(|EW*veV7sYPxpi{Vz{SISx|t zggH1sexck0+}Ako{bKla+T%vrr89YsRzBkqsF>lV)(D;Bkf(DT(Y<*O%IBM8gyrav zJsH4ZQdY~y7JtAg2`wMsHLU=0cIpBiHN^7X;~k>@es^j42l*wq!9DF_b-;0D$~Uw~ zoC)=%p|o}Zu77T7q?8?XIU`I5bex$UdcLH-nsr{bgeKy^j@M|ZIJG6MU53+GLIb!n zq1sF!sCpu7{~AYmvk-d(t?#>$AlEMBp=aa1b0f3I*xrHg8OA=5Xarbm@W?5flOq;}^YySv5DpIO#>@jY*#(TA)uyu|vPV8#`?n zO^Y=d#kkl1E|?O_?q-4+F{A9*dmvJ3YZqG=brl(;wqj3D5(5A6S!P|6We-?BoYf1e zrAKKk$(}+N7R&~G21uqbV`hY+m7|;hx5PUw{5ALdCnkPD0n-w~wa#+0M@`C@ipFyA zyDvX_1$S)SuX%;oA5u7<2W5y5FJF8==5}k2@if?BlHc z&wZ6V{(N?KRm%QsVoU&u#0^m%COOy5>* z&mG$JiMyT&B-$y-a;`)RXar_&bHJj z%RE|+dMGa5D%)82+k1V~%Dc}(#DZ@JQ&X9U+H0vP$pCgu-)gdo(6d#PpX7ohQhFXL zUqYV&f4wcHcrIbeMi{=_iun;Egud9U<(l~jiIV{F>$~ET?k5e{E zSr!jhIdpV+6Uw@H1$;euF`T@iK*IPZk5<03ld~d6Ik?CjX4+Ex1E#>QdTo}Epjr}p zH@h6$jw^9|zkVjuRiA(CE=qgU#2y5rUdm!?~N|ObFRgNWc1E)bb=)L z$j2jE$GW2pc^C<>^r(OEh_Gvov#OvvieftI?FWdM)ZRdL`kN5QlG5 z;Y^TUM_r@LsX51{qB?T|GRe$lbQt$^CHGgpz)nVgg2aZ8b%;;#;Te%;ZD9``&K274 zRB_Bi3uoN~ZP60c)}9xHRgK#zd1g2H=o2eh4m_T$0aX+j^8JFOnu~g}L!adqJ z&me8;E2UGn!>i`D{qp);f_q-72E>>L&fL*fTPl_~Ll*mXdjlD)49ZF|AYc`y zgdMz+{5iz5^6P77OB_Ij0w0Wm$2h^vP6QuoXTk28juOM)7o3ZC%E#AZOjuYYGpM%f5 z7%XVsO(|CfFKqV4&5M`&ivHeDs|>N<0tNhFKqz`2JeACHMnst5*}j98Z>MR|U1m9Zb-q-%Oa2_>Qct*}>lhfUev>9zfY0LvEr?O;{ZMX`j)A+N)pG z2o@=4`}zC#1$7M_&Pt!D9RI)Q#TAN_0M@1mPU}Wvhsarnv`4LaYFwtSA>`6Bhq^0Q z0L*+xLkwImJEE)r`7jtmMF%Usz@pQJ=W|D--tE5*BAWEnYhJvD4ktwuoJsNc%r%56 z-kjo@ody@4FE>Pm@;l>^i#eIR%w9ij<;SJ&kx_>Ps zk(czUHKp`oqD?{V;z@2+T}omp!H2Z$vp>jIypiLI9Wjvc6K#HbmV3VCSV*c2JuF8H z+V;g%{jjZ>Ci>jMep=332#{DjR};Q$lI=|pETcHuvdDGMw~1VO)z4ux{XhYCV4VEz zpwUR*%?oQ*asIhauD3GO@#h0sI%HW`U#(X@py4oz>cLK zb=W_1_USZvbFi2=W~!0+bH9gZ2l4Iir-n|=9smtSEC7KtW{m2!>qan?42#mGsKP8Y zw`$AQ%I6*Ij9rpdQF~xCckqwMZyR5;;#XD$4fC9nw%q(XrDV^j2t#c+J4Ta%vDqyx3LLT%T&to#BJ<4B~3yNp#*}EM6R>&o4yxYYk`Bg7gl*eE4M*6 zeLNB?Pe}^k0G!c=4&3KVRmNbA_6o&$v=x+kaAt)UniIxN z*<~7n+>u#PU?%81gLa#p9S#KC85$pWo_$&GGp{B1^>SdY`htzur{MY5cyrBT(%k<6 z2@&g1r{TB8-=Cj(8C2~Yf5>3RS_@~;$wbJ0W3YbQxKxTmKL;W5sSao>&=ztOGSR2s z0xj_o?C~I$07Lo<;!j7g)RFJXDP#)q^h#ew3hGK?}%=eMbAgC8FTU-`LPYy&=`LUBqtvukeWYqGS?%8g;Sv!;MZUc2o${1s=!)d!4MGh`nD?OY z)Xwg>$$b+3Q|%%pr@})aaI_4u*1V@4b~5`x_XHV!s|E4zHl?){B04oXe9vC@NazV> zHl|KJ*Ly_Sf95mq$!nfSEE?ASqgXeo?{cs$V0l(Wz?PZUhiIhet9fy!a-z)(z~G+w z)txUNddOu9bT&uu(cx z({=R8u$H$~EJETH+?L(_&m8@pv#gJ+U7q8aX+M`bcs$k*mOD5b;p&3Gman;z?UF0L zTg>P+x(vp5(!dxthGk?7-XQ(A{CaK})1O#e5=tT5u+bGe7$XPwmrhzzW8`h<=LD?C z2eTZj*g$xe@02MLMp$Fo$5If8eUJ;!-hX0Fd*y0e8grx5>oJy>r0GPYV5Z7e55i?g zmULCYWvdEhLa0^;ry3YY6HU*_Tbx>II;n8uqFMR5W8uKL5%5Hg%bNVjinZ|})&SBD zkMh-kQg6?@w?pG?X^2!In&)c4L20F7DV^T06sy8BLgwD%08Y=eah5ZcSuTScQILg0 z;d&JB0JDn>sfqO(~KGCs|?625Kb;>?PI0zg~%EH zcAlby=8)3u-VXe#k58z*R7iA3;9W5Y_faGi>!)d zA80)j=pdh>wfua|#+zfMm&+Y!;@XSDb$(`STz*E}EQk}k3Gzdiohk1Ij&x{<$60Z9 zO^DxVb%A`nQvPLE({yIyQmos!DDuf?H;<>@g{^CPt>b1bYkjzW*4pUlqT5?#-}5Sz zd}{qg4VqFJFj6;_PWpfhil5Wdbf|p3(Eytl!!FM$ayw_Ik}oEseBKbyY zc1VlI2w4}xzjx;xGHb@L21~Iu#ZX;;ulHv^r~av2=P?<49SSAd=L$YI8ThYtCRhCn zSFX9eRC6=ypmZ-_(;Fjy!q+G(WW3X?Fj)5>e17J`94)zm3y#9>ZH5(SM*68kO&@hG zgDwlIDjEAkpA&Vfj?*!co|!@hPZ=_I<_&5#pw84HW8BA7>T;{7pA0*wkCbpG*?RvL zkYDVeu_j-+5+yTLyfY37rtBz+*vCDLmKpxa&qQm)nu4K~c|_FYDK z63LtJoM@J|{SCrtMdFUQd1SXu%8R5tXCECZHEyXc1>O;`7N!k`7mtE~DnPl7f9^5LOUr z71`{^9F^lt`ePt4&6xKxYZDOi#VuqO)mp$1N=o(ejtP;Toj8bMNX0m!H}jGhW~!U= zkl$J&cIHXr=aP-FL6D_c$ zBGLEVU(5;pF>D2n*08jlK=X(8eqXPs?+ur*VH8NbfH9e|4MbLk+PEz)EO56Dtf%OZ z*ts&x+knfM)wuXIRvy?UU%=c3#X{+CH_Kn{B($ z-G>`nzihn}MPBhL90(~NI%httpn`s?blD29=`Q)uOg=lKmL~-X?8U;siQmk-tvI3k zJx;YJ>PsQ-KK37mrvdUD_7pUkbx14J?kL6< zk?+{t^~~-`gM?sW=$fwmtNlasy9N^H_#3EX!DkL4`2(Pk!Nir1W9kqL7Z*506?8~U z*mk`gNR6#@4?uGDsuG;7PUrY!Px^1$$@-b2%h>t_4St=k`=13$hJ$KXbDi+AR)ND4 zvP9fvmA`v?i^cQvzEl4I8$_Gr-un#Z8^vy@`lzvwY-V}yL~kU=TuHmFC$92Twd!5$ zb?Mb7$?8H+?%jgbGv7PZcqSc#eGWQGkiEFkzW!h%_;M=rEL3)3$<5EdrB8RQX6iFc zqL27fV)~{x{K2&PW&G>a`@vZ+>{9|J`iINidTK6Bc@UQu&QqKe@}=MJ&)KUKuQ%LV zQ*1MJ^B~TuC6sY`_UfwCkw`Gr(yI}fK9{4Q`=@2LBVD+AG;ukZvo7NIL<5z)vun_M z8ZxWO&KAP%q2Yep`5=p=3Q;Akp*kO*YC-7QxDt|4=iHX3+t1%F4Ea!5RnaBDS=#cO ztwGq1cdlY|q)$d*xM5hgv1{{D!L62wR&w6CJ6xeNDTDVQHq0E9)Zvb--~<6GsQMG?>udir`89RmPPYSuiJG?$cWrG zb*aU^*+V5uB&nm!EaUI=-cqkPs>}Ao^qZ(==J(=%%*619h%2mkT06~W9@8lDKM9W4 z({cg?5BSU^XU)M(8$~ z-H&}?EMMI9uBY0oNKQb5hF>*CLjk(DIdstoAwOD+gv|KO#HPQO55lwO&eu2i;)B5y z7IL}aQjGfI6YPh^=XN2SrSg>LOBqyV951tC>bU)TOZVQip76qeQUaEwdBNv|1Dnon zgUkErtXGvoHoP&DQ5v~7`+(P-n#7ikyQlq>nwt#bB7HT>*54S zVhO`&y31ir(n6bVb1;h~wkWuSMCy3}O|X=nynzzM$)5ZTjBpzu!r{_Dw$#?YocT@l z1kmbX?rHADAD;~{$5LI*&EMxqk0$ij1B@+O#^P<)=fEN;)f5kX(GhBWvephJkiey& zAEQvSO@vh**PN!6qEDm~sUs;j+ zmSq8i*v0?SF0w5zJHYFEXyD)`h+y+fk!WmSc|K=nX%VO6b#guci?QkjG;G_eX=mEY~*io0J`Enc|0eY>cI^pX2*dPDp9_U32)Csu*f#HcUe zAGcJ&=U5)i#I60pNH(@qZ^$3iK{b=_JEQ6kb`nHtJdJeTccO6+kBYqT6m#)^=4^Zt znVn4mQHV%f>7y{+%Msxo{R+ORxiN@jvh2+VOHcJ*b%d+DUa2L<2r32Fhmgi|Ew!Pg z@%s(Gy@tkG*mwP++2Wf^U$9!C|^YajV>a z8I`IvGrrXtP|ZEl=6CvuJ!tmVLAQ4dbbzgmgey-!@8a40bnvgd^r4R@*aaH?mfLId zdxKbT;sUM-b5g`mPj&jvV%1sJDPx_;q$2aAWm=8_aM5BqZ)x|67hMdJHG%c>_2=w| z%u}z<+Rr*G3vwS*TvWRQHCr(N2EGwY9;V(%i*s;QyozJDNZ(b8 z{B;t2)7n;otSB5dHoMi6?ZRFx@{ys_NR~hATG;yoq zj#Zz{39&4_dos(>74m=7m-^DZ#zy>67blD)WKlW7x|Yipd_e6C*n^Mh(qm=5i3nsq zH1$~LxR<25gynY!c&W@R6c`X(Ny!z`YupQpspIZ*%yV#W4MHSMm{dm!^5oj`DwxMixi9|{HDLczk$gbefp2T^yG zrbhPT4PFzsc&fPl14d;Dnh=z1E^hDA#)naCr_JfL;wQWwTl3WUM}3c1?@2RuZ5Msx zY!F6OSI7ks6Q><=-{4Y-$=sO&(n}e?x7Ei+%ZuDPeZAAhW&fk{2 zm95qErF8ZAb+Ue!&3)=C=F$L1u|FhOqp(;sH8Kpq(guV-C}r;x{yZl$I<-sCit>bN zib)1LAphh9@0e;IlgE)=urgBHO;;eEPaSZ8*sSq=uZ5V}0z4{9*t6mZiOSagSr2KH z4i*%DFCC_z#->xq9oK5gF6UP`pO?cCO&GZ1R0M5bflpALD4+5IE|NgHwC zXF+%Uc#lz0d;31GYrrmzElYjYAENBhve4M;5N|jy8=f%ZHma37=h|m8STdGNj7SWo zpy%@6na6taa2fefeI!mo$Y;kg=Ms+L#s1GMm0MAwXP9iU>x3KU`%98J3<1+tq*PYp z+473!H+yi95=J<*K8X95`~_wB`9;vzM$|Ee5?Uz64MNF6HP^>z6tvhkrVN{n&gB{Z z5{bWj!aRhws;MntZB>F%b%bGrP5Ve-KIYc|HU99*rW$zTH4pzBx)1)!xLUbw#%Zw< z$fbsN+1DiZYgd>%P?J!MlNx+*8f)}KzW!?meBh()P|b$i_Zcx@V)WYm*-%^{<9nF$ zX|;1|rS8ZC?i2g#mcZ-~8A(fd17+`;-Rd8Ej__gXn%-!GLmP9mfcZJ20Cs#CgSMX- z*Z56t#he75k*u<<$*dY&hVi?Anl_r6B#F_wT6#(*H^3adfG=flR5dPtJ$$&shS}tI zO@Li#Ui2?|>1{I(HJHS71i7u2vKFrb=WFc!zAHB7UmQJ`$Se8PAD%82yt?wJ=1MUT z$ufe=k96P0$TwxD7}_~KGOj1sr3t21DDk9;_Ok-ZjTO^8Mcgy}^0QMwj#syO#P3)^ zcg@3GJv}NAt+ziF!}tT!Hn8SPwafhyF5$`Qh0zASb;VYk@c1bm|6EgB;5fy3etTUk zPaE^5vF}h^Zcj%bF;IY%75Q^zh@3|xLG^XwHrpFCxpB(f7)ZQzVf-GF@6|Q?y+xBT z|J&Me5^q?R3Mv~E5)@OeNxt8>O6rwXzP2*?5z!SHa?GZ!$`vqXjCJwV2z`w!@n7)n ze+2qXaWm6NSQ`ZATJ|zYasw`FDy~FgUNBAS)DxfUTNqksdkumNlua^fKLH1g;VuzA z`!BPmbDA&uhM-B_n%q!(@^9aM*th-;Irx=+HduK!=%pJadQnTouWk4avf8~rDA?jH zDG2kU^lB`{Bp)O!qGR1WrFGI{!hFof*{2Jf$)Lw!qb8+XgkJ&arQC<;N}hFCg9*BW z&jeH zJaaFkwN5oy0@2!HBcLFIk7~BM2o)M@2N90k;X3!N4-1A-O^BzK8uJIAY$xDinB0Q_ zx&5sAhvNrE$^no3nM!0=^8BlA2?wH_RnZ0)D+V8de*K<7K}Pe+D)SdQQ9pMK{XB6@ z05HDt(&nquqI|PRga_?o?}~3GyhWuIhf;`Ti}ooR>{VRJ`rx~PVv(#mi?`u1Kr0O~&2(u?CG83pjC$84)2FUoZK63TWqai{bj z!8O(~?zVSfc?SVYJjQ=gAa?hCD)gG?I7;S$10+hj|a_Ktw=E0A1Slb{w}0!zjDU%onFBN2pj%!T{rjDb<4$ zryqi8$C7ef(Y8t1oiA*bUOHCQ=jqk!$=nH!&bEUATb9n)p~(p;-@EjUDq6R_s~0yv zafb~CXdPvxjtr-{S%!m6w>xaLYtjmnxBJ$f=IAl4p*5*TYsQ-?fMTRMk8uzL%Qkb`qV^l3--tEWJ_Nv;8(33(U4H(1~ zI6~P}@#MWGTFew0rqp?wceQe;9`#aNavC~Gy%)LpQa3@XRiN;w z{~b>tzKvlYU+C%;&4SSil+4jnsbv0M0b1cV7ZpRowl)Aj%^0iZg346es2v)g`^LP+ z)ZBqpFN||1*Vz&|%NAeSM!l4-zg6arjXdJ$@iWho^L(`tsrl+PRkK)ciwvIs*NM&9 zi(_-%+*kx`oLzz7_5=_uK0C3uF1S_t85K#Cs}9lXp|1eN)T{X|C6B zivTHG=~)D^2ULMbX5>AzVGSeg^|Tlqe%hM*L7~%%<=5VZSxHYW{7Kv;F8x8tO_zk! zmQUX?wD9dOtt7OZ%7p6u;okUGubEGCVjskk)n??LQl2KwM$*VFg>{sN`(6kim;49Q_B>`2_QWSSvy$8N70Br^~9vTpo7g3&x8o`?)Q4OITOi;<*tGGe_ zl((`Nju(D4^2hRm3i;ofeHcYsH>HjlMywPkv8kt511HepgrmCC_v>e_4GZ-F^|ULN zI}kg$_xL>be#}kk9ag(*MV)X8xzanzRkZu{4Q;Mz@>?{y7EElJW@M91o}<%hs3M1j&L_`crxe5VYJiU5Tk`;2ZKAVi`~Gdccz0iJF4%W^r+EH z(4)wUwzE}`)#}SUi5RTNf-kSbB@xJm*a$g(`yDvIYH4@F9I2mYT(LI`z4tq0?73C) zCmyayzxQ6DB+z5xa`aE)FyF%$d~fU`=?!J47TX3 zu1acSn^rhe7v7xbv8T=`TXw%XsJ>eV<7E_=VO4x=F={L^GX{L;!ga(kA8%3mjXyJK zI}%hMH{)G}d2x}Gz}*n&BIR2fDR{sO3P)#)-px*V$~CLIVh_`iO_wq_ep-7gv90%_ z%l2T%MVLrCms*)6DSmtd;M%pm;^t<@lh3eypH&Nr#>8;Bnbdy`WYI-kOPw|jSQ=Kw z-(`v!>@sdcdYDG7IsO-3kp068HlML7Vi=3H_&c7TqTK=aq?nHqIHw1%4o?a_K2~AB zk9p@1o^11%K&bS_M637SSXU26ciytSCH^v`;Mc2CO>KP&t>|cLEpCdjI?cNPi@FgL z8{M+LheI8xkIOfq5|$~G4=AF|i-o2GL$r((IHgc4u-u^NInQT~xt6P9EBu@o<-i6> zw?O0Ju3K3z2`Nm*9CB<0oF#{37?VrMXPk+o9<|boH0|7?OrQzvaa&MaZhVL$KS!mp zxo1xqqF3rBp-5&gS@q@V-eHZ>o68XDToZ{VeBkwr&LmHLW1G2vK|ZHZHhmy;U*1U< zF1OU<19TV`RokRsyn-Qg4+}Y0^Wnm3vG1KzC{h4%g96^YO`(x=LGqN`8xeGz%Hkg_ z0Hmj=`dS~Yurc-%w&IP4`T)r{+F=bd@SR!eLucNR1^=v;zGX6K_|KJb{gB5-yll94 z0e(f(Bp55-g6y8gG&ETW4m4hovPmiBX~w_hdr70XE_zdZjWWCtVYs#!ylZi|{TV|q z9ZT>l8_1%|c}s2D)mpnf+9b(64#SnjF)kKOg3xw`9Ys8U_sdZJ*sb*@a*@>2;>jRL z(x7f^9%>2^}6jaQr}58!Hm83s;8x24GvGulVlW3>yXnp!$HOIK?2HX^yw0%|L5Q zIrNtrTC}M~L%}1kVPLgC$j1UTVpj~A8z0Gu?L`TK_hRo3!N7$4o*IM(*GDJxyc83| zp%-n8Vi1){ncx?H6U%dcrX{%eol8N;$w0BUADJ+W$sXmEQ12x1XDqw$dG2zd7d~XN z+_El9)kBBo3<6NRVsY*e@&Sf{UyS<(m5&^@pTbWSOwrg?iuEm;voGcD{hQ{egW11l zuPfPuM2XX*WRK6 zFg`u(_IYY?*lz$ncY!Ij1z@@|KXo(c#^8Zc?2(bpd6B#VtlhH z3!|>0gPm&%kydsbogV2e4UUnBsduF3_On92>yrvhXkR*QTQY-cxe=YM!f5A4FJHNK zUD2D-m*Wl>S5Q7Iwc0@bjXv`Af(+*gSCui25RR~}qtFH`hYFhT|6=PrluD49au=?H z$Cq)xE-)zD^%|BLjvU^s0QvJ7;T%C{7H^yp~^sCzq@A8$0D_(Rn zny?n?!gh4WENUTKPq6YGQ=D;oP+g*9{f*FneTgbhK7>eqDt9#aqj>RozL71hQ~T^OD^Bp- z+HC+n4wnzV2#>7S2>D_7Yx<40>7f_nhaWNuBNxB_JmF~}6V1^5F9J0i)gecjtJn6R zH7DjRO*M@twSJr}OgTRIOR$NsRqAl#y+$@p{z)5iadN2G&x$*7+WVX!`2K0_mmX@Z za7ci|GjTS~pgAXFYTJap;%^-7b~Xf|IIUK3JAC$Nb$b#^ZUEnOz03MQgYtc@05w3$ zztFMK?+JsNEzTz+Ye}s(97lGKU&O`iH^iw`T~knCnipk&cH~XKNH0{_f8nS4=*)CV9t~6!ebzEcu~ymrk$E;{OKs1N+&RB5)#Vjg2V;*s~ zTuWmAG_~?Ac7zx8{Yv+=n)!xX3vC@mk-7q0P8^qdS@`=CRx2OZUS#++Y-P&PIMQ{a_x9psC!}eP?_Jo2k6u_8avsUXC|5FJl$$afp{*#nQHO|>{hf_W zW3+1amL4;VQymx_xHD2f8G>HE38_LxI>UwFa_40haax3MexL*Cr(qx&k4L!XzWjR| z?U8Q}OnxXIL*MA7cF!YrLrJ5V7da3~bIk*G9cb%wjN zB=2PBbT+Vh0LB+A;1e7E<-&v?H}%_3PZuEA1evZRUW^!KPYPg?}ofnqZbn_ubblU);6&aY?Ln#S%h znvMBS5r}hSz)18`O@b3Ouy^!k)_d zC*StO(fcP}Gq=-KM1v`jvn$?XU*b3CK)mzWBmc#Miu%m{hf-vJJi=jAewUNh;W-Cg zJCHu$Ca|nGYvg!#)m9RnpqFwju47Kt^87Nb$NG}ap4_3+m~)+KJDQIs*3%4<{K=f8 zY84B7C%`1&9ZfQ8sgj_JObN2{xK4n@K9ZA*$yD+|Pe8S}Ss&HnMGc}FnqTp8zb*6j z(X_{vo>yBa{^4TM>oDI;$L589l*G!SBWScEPV-KSyd}SaVi@RAQmS|2EZ*!Cl2Q-e zAw1-qXqe{aaQjFmNNbiSX<{X-!lY4q$dUfAY~(cGw%VAt>jRHlpCJ)vr6k0g8z3mP zZmHUS&$%=?#J#<_Z{A*8Htj*O2r9)RwS+<18YB7&0AORlKZN`Xbr(lH;k=AOXL!GE+1}SOx8NRgD<0ADvu(XCX;MU?X&u+`+0Ptf1t;-Iu%u9(M za%91QJ4`r_Cyyd;WMMx`gQz<`r+hlNr`%zy;;jHbN(6G=>7Al+V4~k2XXa$LZXi%$N^y<;8u&q9Ka} zZMi5=v-k&45$|eCh3g3pkmkR$80C0;r}0AyVp^6&mYBoBR`R=HYxuzh-fu{>L1e4J zDniA->%VK1|7_Ttw(4Y7W{ubKYVgwOVnddxRiZP$w;26| zf}_?m4@u#6P45}0j5w>-ifD;A_ZA9vrm((QJ;nIqRBDXR<#P>yq|9B2M7Q6hRcbF{BU=NysR|ee#)tt&3EPjJT zgEq3AmF~tl6QTy?rV$80I_X6aBeThdAmDNguVXc7v}hOzCFg349Y*q5Fhce_X$&@) zfP1l!K)AnEzEI)xB^^r${$(;#-s(23)cR(x#i<+GO?7go=Hf zupl=zQA_*gY<3*U{M6?K&sMy_=&%%KDC|ao^S561&$w}$5qqL{Ycq|>I~8q&0-+d@ zhn({Xiw@`h>=Rq@N1fXQFsH`rjxRE^5rUl8e<_4(rsCtMKlO{nP$-ap!f%fkq2IDM zHq-nmMRdj;h~Lev$36@J6}0dM-N;#a{q2!=u5`}DKIqcR<{(D4)n+5|R7_dzUwC}+ zhdm7UZiSD6+n?!JNEGz?Y^eQg=%(=8(QB(l%g~cmopf|0L}@vb{xDaX5CuOysTlb3 zEhCGh;j;DPS94+j7)x{fMqlOZ-;JP6o?g#bLDCY}wxxJD6prmdo&EQD{x@hwUI~ue z?k;gy3ZRSsP8jw_c+WCey4WS&Lps=`d_m^eC)Oxy9z`h~uxtCJTEtj79o{Z-mbn!i zqnF_tB@?vuJ;JYhYhh0GQchbYr=w)NWE1pQ5d6!%+H-QmPXHP9AI?u5 zUdX}$TOOU!9lqspl`Mvi;*p5Bmt-i1CR;9tWBd(30y)y5t0($O_C0|>Y9Xlg%%CHO zqN9)5ZwEUwN{4%=yCC}sj5N#XMAl`}tB`rmWJIx}QyI)HkYJcJ{ICQmz*F3ZF_N#R zb+dVns;J%8_@Z<&a5F4G{86>_5Zz-pM(((JwkZp5l1Y|e+z93*t3S=yvPt0jwvyU( z1tdk%-8l$}QrrF4KkLC!f2^R7JY!vt1bKs*dz&x%ap>#0@YuZ}8|x3@A7iOaYwuzK zg#Pl~S{uVAx`Ys~L3Q`dTjGHd@}48iE5haaheb$jo*WHVT4INL3eW4ot1I1w9Li8o zV}&{4&4}R_y}rSY+dS4$Oy~RyX1UwL-b>?PsATfN2TIO?K{^$;>HRF^;~cM*K z?BITKXEObj7g~9cw&*uwuNyK#J2Ew)Dc*4>SqfQK=Ee~rq*Y+M4<9`ndch#*YXb4D zJw_8OvJz{@zA30LU_T*|gC1v;OtF^i$1Gz!FxKG!L7#IsrKG;gPn~ZMB5H=a=S<#G z{jQDVS=2a>X2(Qi`B-vFe*ThTtO$&^1Xp_Mg)B*VNVdpRXy#v zA4Pl=Y9EEWCkV1gtFZ7%O=I*|m2do_E@1oKJJD#8W3>Za+qR21eE|(A;Omt_B&aGB zF!j@O$>r;TfHseLH=3m3!f{fBb?f!0G?-Y<3VT=Qd53&pDhnpxxmCIZo13RkV*cnv zH#ce^=noJZt#Z1Z?V9{;<2JAuO~xQORyoE#y1)b+H&z-o_Dwy8d?%oHT6Qfr!tgX< zRlV0NRE*?<;&@d*bA;gw%Xb=pRaL#L(8G_7Kd?1V@TppCzG<^e0jz^dAk58?{WLwA z4N9=J{g+8d#?H$X7*gw z5;X_r?6u+4!2&N=xsPX*sP)!oO|Wruj>4a$s!D$u{`6ikEySydwWK7x;f{pTW)f|4 zW4F`~LO=;WXlfXK&HrSn8*zwK1Xr8RI^%R?wH-_6pF35&(oeN`XTXx6Cp#H3&Fc>&R*ymR{>h5^wMw9Rw)2&pG$1w z1$H8gLONO86T?B=a)h+tmi*>jhvd6D2lr^1QaA~Cim(l^O+mch07Dq&^H!Gd%6h!? z$=>S@ww%4vR=UJ}Tw32OEsZ`|il}SBQk$AZNZj0`7Xa+w7ePG{E3*I=bj>#FIg8$* zrYg6-V2UuSEHQarO+JT^7@DNTD4BtMol7s^`l{LS_(Mn>E=xqf5$j9?ne*LJ0<%dO zT-bZ_NIaW1r;txfjiw>oW;qI{}Pdg*4KQF<~kZC_fJEa(?;p z-7k#mv=<^g%Oiq!<&RofzCjKZUC?&n9o=(8m~fU=>*7J!XEw*G6>$NtgS!nfqPGVlLr0Zjfe4yt|$ z!oj=yPiO%9U?@*U+qM*w2y@^w|3?vt73? zVZd<7*wt@7<WV-Pbwz-?61@p~*LYX1{ozfBs4q5t(Y#mSs@>S}F1{kWR0k&q zTGsw7c^DY+jRfy@NRi{@^`$j$^5}NJLlq1f9^YN(Xd5x zr0;j#%yy~7P#Y>k`_l5`wHYzQqK*ENj@|6BS$EArz26f*Ca(337_1g+;#jTO%}+D_ z7b8h}Zk9-%5r+rHh?BL<>7k7I6Sl>fDZdw=GrwTm*}GW%wtjlxpR#3+noIKh^qk2+ zyzCs7Hn2ekKYGebydB^4VLu=$_6um33T(i>R0jFa1f$`9Stf|UbMqp{lahloUO6T? z!!t%=n!Y%`gl(APF59w5gkEDKa~|kV=)O_$>xC>A{Y6j z`|hmoEmctUCrZyjAz?{kg1qA6BC{`tY`5g)RYgsPWXtd$3#csJtr2SvDox{lizinq zw-&4)+?lJ6`{~9q(-dO_=w++X1oUNvEUHP%^Kl`7nklmA%%!t6FC|l=8w+sQLy}I- z3i5$TE9=?t0z1C6vbcP_n+kfEnJRCrji|=unYH`v(s}neIjKXZ#h;{Ou8Zlt(W_d% zr^mTk7d>2WCI5KYZQsErm+=FqyF8$H)A^86x?3$It;Br$88z=%ecV2=m%V8+JlBgv zxl6p5n?}?NBmf}kDE7OqjeFq%*DcBnczn^D!4jjiGqS?@zm zs2~$Uj$4z}d)R7yA78NhmlAJKBJghtT#Y3c<&(Cuy4pjOlC4^9BZ^}yHmq8m)$eg% zY+PMTMe1PM_b_^PwqxhymL7H9Gf=dsd*F}|RE3M%wQ#cTXjxtCnA{%Ep%Q zGA_FRchl%L-q)(R`Yad*ofo-E_cH>f?RqJ4y2y{J3e9j_jsy9wgt12(gZai!EI-Y~ zxo>yqX8JthUhS{o9+vb)6&ma>80_hU#k*s&Ec-VRrHp;wf}66VMQ6Uyn7IeF2o{9P zDi|qZg(78b;$H9H6NUhbVGV^}cydxF{fD_H>9kkzD+*v`L2S}#Btq8mH}7vGo)fx_ zl=zKth5uhmyVX%*ZC2GYqqb$}cjGi1+JSf#RwZS=J1^}boUF%JJzrxpv;IpWzbsVd zU18p^p^aI|LE#`znm$+pF~HZYz3#IAPyJws&&l5^V>$7BrRP3KysM`*`n*3zQu|PZdr)xWgd)H*92J;IoPuXTYFh#h13yiWq zh+UBv69BwD;rvhj^|tC?=f?g>J29hA`)#t#Xsivg(OT}2w@>yc+FX7kUbB}W(&-<) zTwp$?IA_j%oCgtLX1mBX>#%#MIQRNa-l{kMVs8z8(4)k=&b>3mW&~Bz!g*>?d@hxy znu~nfsn@P&oi!fZ+5IqhOk}4;Y(b;HsN?ea?Cd&(Q>4R8+F@C??A2NM(1tZS*EY$! z&n#Sk3r8z(2dJ|6lF(aVkpg7*Ye;H~cEOAdGZQ@#84t~&hm!zxxq9r5)AV&%Z1(dx z)kGI!*=0XH?BBzSp=9!~I8v2>f9nQ?>sP_gXP>UPvf~c{t@ZifiHZZ9@btIxXXlJZ<>lh+kT@p`bgp?MA!Hpe=qxITt2 z%CeR0fsm-6T9zC)`tx}%(k2m-RNG!eFOLl?xPuobE3wK5<@j1Ie1mlsbv>T5wgr0h_vf&Nt=Js-UftKPl&&ei591FYv;;i0lp zrUZ5}y<-ZvN}o?0;~(unTU3P@QTj(rUuS3MpJ=y^q!jJbs<+vpvef4dIi9LK)$d91 zK>5}^a-DsoW_Mh*M{0%A*51~;9IrV+Chhxy31$HCM5X=y`{if57dyLRp$qKhn;3G^ z+$e(CLfd||GxBUlGb?UooPBUv(}D#uHaPBI5r-`u!9Hg1pKJ25?@~p8eRh$;zyv#9t%HIs^f{9Y?As1vRq=kxqDV4Oz;*}lGR zb=G4gA|e-F$5bPW*h>JMvI^>JcAgHjAkIdgl9{)j3#R=HcOrY14r zRu$)rs7u3`+GAzu8HKG6F0w~X;m{)Bsg}HSwatC+5`I$G>_jZjCyn|Pl@~tsN_p-T zdyB^i|2FzwzIVp<6AxsiD=&KfPT)|oko`^OZuJOH@YWBVP(mT^_rTIXn%18`>eMYW zswcOOjCOyyhaP;U`7!x(Z}x}(g70+o;Y@Pdgfq_wpVM%Zwl|BH%Xarl7UsW!0sZgN zrH8qIFS8s2Sq!zUo`{`NQ>a3(31N-+TjiF;ExmPk5B-q7y$&?=Q_)x^GoV zU*3I3BVC+d%)t7eBEL-xoZHsvGd3gD!)e(l7?<34M;nr->hMD|DCN36fgwciy+$pe z*h@fuu(AG00x0h17;1v|b~DwsP^GSGx)#_lsaL+dvQ3lD3=Vls?8}uQuydlnLeJ3Q z{8UT8YsC7RrwZ$dRNm{jXknhZ?}!7R*O%`a0;U&@=VQq>i{s94%S2ZD~46sV^5N-d1XJA zNAV&ghDHV(%?D$VN0;WrDurscP$kHN_`f*PPeafu=GC}FyPk9}@}^V<5axrM*Rv1f z>$)4?ywWZFX8#xmDs^Qed3X0y99b66os6xVTJkD<*3xbu|;X9Fn{v$x!g@X^QGs=vM0aGzS;`dw*>uUd} zyzuYX_QYK=3;*6FaKB1omT8Ml2G|m1ku18~nwPcd3auAv>D4iIaxPSG;mvnmpM8w; z`EG7+{Kx3bX7f+1-WP_Wl){XJfwcok^mfA%!sQ`o_%p=aicuqS;kD()s*YdY$h!hE znC0!1ima0og|+@Qyya8-n$D5dsDu%%NKyG;>NA#qfug+w1*8jGD zxYl1Mrfn-FAK4*9DJjh%z-iAGUnKR6gZmXAJU*9>otK=Ts;iv^C>~GSf4S_FDkk}c zn{#Ng(0D$`uWNSoa5<8xV!=7fr2S&E9l>VJiH^59&w*vj+ub#j3~1f5tfr zFLL6BMZaK7O@|-!A~c=is*e}^m&*3p4!7n0c&O7@jlSwvyxj3`Z&?Jc^xsPK@s;zE zQ>(R4kK?vsJFc@;>*c#QQbbllOdf99Lo>YYb(d7!E)Ik(&PeOZsVbBYF1N(@q`V~(;hMl+C)(}swkfQ6FIs%GpRNA1^x zcWnCvCik!YMP=@F><6P<&-kt`vu_@x(lAo#BsM4;VZB17EEP5uviLco~c zwzHx|c#5Cq_I;)r!X0p73gVf*;-0?t#H7CdJ)g-@SeBh%&16hn2cVW5)xtv4YQ@?)Mm`eXg1Y9^8;)Cv7-roiZS(}0ttK{^nWVLVFcrN*fAubqR^RS+*o=RkarR z`uNOg#8KT4p9wVNkKshfrKmPocE4zjC=oO=g*UjJ)gDaYIEJn6c;G!cHTxQxGxlZKP6h~ zzC^M?0h*oMM3v-K2+n4j9_>=Wq!D%Zn)#oc(6QFIc`I&$?m`U$;70dRI-@OYM z4e}Gey&>-=TP%~$HUDL4vPWRNb76HO@1~2VREv|+)T}lNNvW*sJ^DXce`UsEt;)*t z=m@ixSDm8PWhJ*p#Ma(AJ|uL1$A&+Rjp6w1$PRe+GEQ9{oxj_Mid{)6S}$W+6HzW1 zEQfHpZwEu<>TQ-6?g5785?JhzJq$X);4wMBu$4t=V{Yw4X$qG|u*uyWM%?8;YCx_D`I zg`0p@rf;K#f;n17Eb62m;P0gKNJS@{yZzt(ZdX~~|h$@DwF`02*ghW-?rNNt+jH8c2d#?RDci>Jl!(B?hziD4P9cR!_d(xd>|D8pu!08o#NvL#nWU-1y%vtwg)i|Y zb^D_noL7_$X~V@LGP|fFgpv~=d@wfEldU<hP`hf;B&3fnbgStiKK2M)* zc4rjHq|g-kHeeR-WWXNXK>AZ85X(eSiXM1#HwCb3!Cs={Sn1*ph>mA(^-U) zubAigInW=bkUl7~u!7ZeJfJa9S^#!Pu8jz2MF2GtAT&Si7TMbxmG!2ZcFxAEUPox# z9d4?SvJ2aVl!3VXZgf6Vhk?&P1&v>ME)!}tHiP^GwDY|PYk7MPnv(is?{T0>>W)lM zsXJQ-SLO*O#uWvlYTW9G+CF#S?vN2tBlthgNFN-32K5Wm)md5nd~VaAmA= zCGB_%$V3w<|6b1L=cqiYyZ9)zOv{w%7-B!FF)s2sVpO*~LT+oGXEr_hhSdAu}X@cvt z-Y6+2$koYm-1bBwz<44dpJiS4gm)Taw$~P`hDni@ zc87(mg^KuPR^H;DKDe)Maj4}b;U;6eM9i?VDLHP>u6;y9$g0$SR^Dn946skV8dB?B zF74!XnV@1DE@SgpkdqJok$0c#gTS_JmJoLA%`FDz=>D%6pWkE>C_$B_!~!?bk92N{ z`4S|Zt2&-vYA98K$5jhC{ivtURQHS)9hSkULX;LmZ! zW=JR-Po;)C{b2rqe!|s(4eo}mFwOCwHsnHV;JiYB_3X#c3sTE^$Z#RJ6Gpt|O~*~} z(DVE|FL>?kGFJn>0w#sJ7!;JyUApwb#!JmK0*2gu95HwN&#zQ-x6vSj#BEdk(vELL zRJVUVtC50ek^7g;mxiDy;$?vx7#Q^Zw2?wA0t77od2V1<0iOI;sCal?%=agq5A~E* zYyRGk!kzo2nZg$aX6WZPytuM~fO@ztTf66@Ron6k*Wy9H+9l3r_b!6oJeEV(Dr@yn zTX001=l{HB+*I;ZCU10_@+_@OZj87%vVI_o{HVErzGqa)lA`lIF-=BSB zeztG+q)#0>2KkW<)OO=kFEuZ<-%r{!GX$k@s z=rJS{I)n>bUlHuO)ht6RGg|4=^)aR>ZSHsFuCSipAn zuYjiGN+#Q?-(#$)-tmQtk(#Pv@G)>-NHO zZu5i8qivO*f+GB<^wZht!8C>GSecbT(nHEte>xq_2aRZbb~ELHKVYF~+=d<*bJNzd zI$YRaQ^p<=pPwo}V|8*xHZ^&gG+gy!Dw0zyF?52Ut9fV#eVdmPAS^=a70*efJfY|6 zm1I5N%@^e@9G7#PDuefeA{v9d+yvLSwCg&9)6YQ(U2`{oT!lUpf-Q@fHGTI4{`!L0 zP|QCdX$$$Tj;Y*Jorl6M*FqNDm1+=QNP*mBn~5ccsJq|=e4xcD)6OiN%DTgNO^`u# zU+B)3)Xj)+nVo+2+*?sXq|tizUT+bG^gVIGNj??sJ&w_}bBmwy^hwP=RO@5$d6S2& z5q0>{eIFtL$oT&ju({c&9mOjdOj1-jrmPg?Vy%mceZJfhw=pQbGs1OmpJiOkPotq} zO5M*Pt^|U*u*99}T%>hskMXwHWyH6trgVc$);Ti{+uRP38;A@O80~hAPi2hj7Z9j(^+vOxz^uJ%<)Ri zdBkC~Wmjr(@r`ZW#96J(%0hkWUZcRfGkrvR<(Z;l0TD@xag(N(xWj@@i$yLRZj6`z z5S#$j@D=~idVjp}q^GeCzsQ3e?J@U!Z8>EnLl6e_DV8AEuQLQxDvt)+>y5A~xpm7v zx~9R&>#-e_*l=k`@4*pC0Nnjm4A!`Kwb*X@of=O_v%xYnvzTC(U2T0}FJ4@Pla4qg2M@6Zm-KE8ukmZikN` zkQ?UZy_6DCJ2sIBh-LkHoC`qOFM}d7OJV{uoUdzSq}}He2-fHZ5fbcLoSy1t%w+ut zovhC``)GsV6uX=}VGAriTmJYQXCo*?T}#qd&5R7Rj!CZ^?DO_){K$al_0Bj?yhlUW zC*oj@=9j#Q-(kpafM~?d0Gu2nM2s4PpDc6A+_Y7m#pDNrK2ETi4xT^ZqX8h7wUX*m zH9NB8sNe@z%fEnHN|aA7Nd&wfwS7PPP}DHeGp(f$rGKCbxH+Sa{FK+&mEBvpo*FuE zVSD86(L4+5;m^QHrMgeSKh09QKZH$*6HaAg07u>*46bq)#uE!{oqflbB`5i2H*x_} zNZ;X47w5A`C=v?Qr&HdwF;$RZk!Pk@;KTzAS4dytzGBzkICWdWyuT?!xuij{1RXnW zzpTlAprOd)v;xFU(^WU?wZRVfkEEMm#Q%*+{sZUgYIY0GzfAGY8uk7SzF^5#b4#2} zYjq+fk)zDxY9g{#^n=tZzhZ}x+R?S{GH4})u)^gecbbab&~8go4SiuFSA9y}QoxFM`&P>F0T0_-r2D=%9qwtkX~DM83dB6v(J~l^Y>A(@6 zJ~{4XgmQHxxZa}B zA3^u>wwGb@&i<-X&zupcj%(j%WT0Sk@TJG?qCi^UWed3J2#DSf-2p5BzFn>WJLad{ zp|SOi`#Bg(tr!DQ&doRR!J`wCOvnAp_g^MdCiGXRsdlD>@|=pUky=V!!kix!)Ohf=OCNZP!3qo~aCgc_4z$>!g?Nv%BMCQ0r z`Q*cIU3p!_0|t-1TMz+CXJo^1P!I&39n5_+$@FL8`mX(BxuG0Ym;H~yirv{vNnZ7< zW>L=YW$X{s7lkb~gNg32%1WkGRCmh*P`}Fk6uc*vK-;b~w)NyxVAC5@3t5_$4m)q& zsck;OwRDMh;=9F?ehyy$Aolby7FUmzaz0QvoT1rVx)hHC5EWDg1u1-if1Ok;lQ*<+ z`PnJm2vUHJ^x4#=k0Corb<4UcKj-)ciKI#z3okGn!T6;>M5$u``iP7h)XSgcD+v|J ztez6&;?e~HkBNNOXEl$6{(a*PPZ^7E+^>9dY_}ika9~w(>~Qc3mz(vB7vsozG%Dzr z`Lo|3k2LP1ZCY~LXH`~GHM00q(=*H8GrM-DUanKhN+C=9H6)s0Kfz%?5wTN$O5TLY zi~8Yk!gY0oFa_AN^W2qYa`okQ8^1afU`X=nU%NK^bjp`bX%~9B2gUU>y<22hdyJ4m zjB%oq&|6(ozomYmv^{0!K>lCmZyc@WcmKbWpe|>F)Ig~?#eckW88uZ0VE*&0a~1#m zbczqZHWyTfTH+&VFFZ_tdcoFCHc^0gP=~hfBEOEb%*KK0W4>Pc4ax2@RvnX2^zF z$7+ina4|$vm3}qxE6zGiy*G9-#VOXDPht9_;Q`buCbHl8F55)*wY5E^H2R=C~gA}G>xLGkU@*kKPQtN>= zFTH2=AFq{V!14~LC)x~-f7r&Rz4Xp3umhhH3$ER$-)LT`IW_E?nW-bB`@J~5iEFuV zr7IJ@9HRqq%gHjkLy_3H86~B?wF(Y_;<#Wl=6%z}cy`CC@;djYN-N=BNL1o6yU$V# z-BrbKHYVU?$`FDTT5bx_IL+kP#7{34(3oVk3m!a#f7b#m-6@QLV9PhMI&y1Kp1be( zyGXy091q+zuYUKvXN@rX8|S-F30pYIffb!bb1>}#Y~lt}$*86O1yI$}94^kb`Xy%k z=mXo<&P{M{xPJJ0*S_UmX*(eoEmAx-MU=mN#-GudbNw?ffT;r@2(a9LDwk;7r(SMHjh z=#AQ!swuII+w-X^%KsJFW+p_WS(_2D%TM_KIo4fmV&~@V)?t^R9sgnF6((SJO4&SZ z5M}I6W;DD`_-gGUJEU`sXS=_Y2kzGN6`~)ICBtQ}BPgNk&>C_Sp?We)e2^^<6yV)Y z+L$7h0Rtw2V#`>(<+_?g@^4N7B_!rjlq07;URR2s3jm4F%*8;2CDKw%vqs;D*J*$3 z#--EQf4Xbm{|d#1f^c3v^;6sE^IPC6MXeURW7bVs-#0)AEIgh;7T#H1I?H{M8lZgq&Iv z1}!F-;%&Rvna;X!maLja?@O9sBs5& zP>cJ~e)+^6YF0Oc_6s{z=fFo%*N!1BeaROa54gcFi7(g242d*y3*2WlAEErcAIb`0 z!w*E$zbivEL4^5#73qY&P>yRSznAdN)(3N6*`)XtA5UE?J`z?kwyS3AD7$9!Bh4n^ zrv#F7SsbHG$iT)AjN`pdcb9KsQ^bqv+_qowJI%XxTox2%rpC%qcMy|X^p0YBWdBTJ zm2}gPR}d=yWsw6ow)ia|k)v)VXpHHEfeNV&&Ha5u&+(i2jLfDjr z2c6-sR!inb9NCn&nv4w&=eQfI*UWd%tun5yEZ#Sft4r<3vra-#S8DW`-oF}^bhu^P zVAbHj+S^=R5n3fj@c$36wxLZx+Cceg)<6FcxPV2*Pdw6vDf~d04c|L1+fPHMl*}?p~3C1IXOAmdRdi?4zYSI_D!=E1|qy&Q!NU{cE)vy zFTjNs9lm8dfQXJ z`m4gMJFUfQ?YAV%Y148JgOI;Vj4$WeWy-(+JuoqVVwV8lE4!&5_KoxSq-w&xM zX4@dc+D1y#z0|!cFQwvKhxq6-pU4@PldS9dt?yPZQ1P+zavlQx8;BLq>Fj#tmQI0t z+4SqnT;bdhbDz*+v7!xao3*2{UrHubs$4e@qYsyRQXJg&cZn{DimcgN{ zGwMxZ%G)eZOuq8Y#0n0z4UW?O?zc#wz901S!txe4YS{&{Fv;mF zxR#_kNYsO;7WCp-{*A6vg=i-Tng=V{iIWGpwg$urmrJKa*r`Xo`-+y>8p*CWZp7SC zd|!3E7o|!-?&u>u6EaP)k&D!^>lD@?mB4A_!4IF?CcjP@ON&O^x*cMFf0G0}|Ey&w zWEYWHIfB^6>J_l2ycdW8?0+K|?*kh?i}4eDs%SrF{1{ zF=VF9iOUPr6tD28_}pgWxpXIKSDUy3%F?W@k=b+{z^h7sLh9kh#)b$-enujYmchUHxAD_0NpW(|S_@eU)ET{O`Js z1HAH9y}|m)Q7WPfy>t6g3VL9;z$jHy5FxcDegx)E>-hI%GFWJgRUfD!Eb0d zWPB{!tK-XgZvgGc4MOAYkbcBZrk@~7VECGhe!6T7XFp#@YC3=9*>v{juLpBM+fazZ zv?zTkq}x)mcl#G7Hu3-i54)q&R^+(tq-RtyjZT9Q8cPT@t^_h!|#sZ)z+221Dw; z#(}osd7FZafl#=>!U85@=bD5iHJi+A&%MZYE)Hn-XxjSQ$~2UX3jh*3e@Vz?nALYD zAs@3;vv1mG%jG>(rk11PUoc-spq~1JaQ;GbxH}B`Jnk1Kvt3x|Su3=vp|~XJR;?(O zLpl>C`h)z?@RpdhU*Tccq~-7R%&F(tk|+vi_s(MMt_UQ0 z4poB+*S`D@uCTvw9o1S>Ejg|`4v>$L=`CYR3nVB_*R%y5lzbQsc7%DR9t8jfODdFU zz?AJ+K+0PdbEQY4skn;Dg|f~1@&)m$Yr1>&HxEt-s)3-BzBbNd-E{8ZUynwX9f9)Y zNjvxV{L-_2Z|FDL#wok|Zj?YqE9d@~M@7W`wYON6JdJ<(@b0^F4414Ys2!skc}JpO z(^q5HjNBK_9ZnrIRC%u7EaDHHm@BS2SbH;fZfhDQO0y$zUEIN~Imi}2PP7*L(>2+9 z9Th)KS)Vp%I-wLX>3GO)@$RK%ut&BGvfMQ+=`_L6IM-y>y^k`K8nA$2_Ysisp`<&DIw=W|Zm5IV4NqX%5NLbXtO z!rlPWxej{BMTo$dId$f)3;rhl3u3xo)PKOzy7?Lv;kI%GgV;%jY(ND(_0oh_8-ECw zY%~B(RW(n?#sAPQi$lloF6&+_C0cC#X3wox{j~C2dK0o6^#@b!5l~?h8;)(3?}zsR z#x5O2$OwK?>sQl5`o2su;V~zU`jg_SPd@d5%3sY}|h^$Pc zi!iA2cH8Y$!Q7(0hgIJQ4Wem?7(qL>_3qbfP}IGhrVr^n(Co>T?wPdIj`v~NoRp<+9r_DodBL88u!Poi>f$!?wCxISL(Yz)q2-G((LsE2P>yza2 ze;c+WXiIZP>^8dFrVk$S=t>lIr%tIR2tO9E?Qvvio=8-Cxk2wUs(ehhn z$87sR)ML}sM`4;%S{W0Ov!~e-C!{y2q-eJXinX36hD!=m>Kc2MnVJF4%uYV`C5OOp zWkt;S!}9mgS4`wBm+RsUr>l(=2jo;=+uL8BfvziOiZor03Obx|T$$T|7-f*=jOduz zg8moL$JZJ_hCm|_8^2^H!om=brFzO|<4FB1Fso!;V>#O`L50t8D9uHYdJA5K%$n}eki|2FrgJ^C1A$=^B5M>` z3TU&tRW)rdvdhl3?O{$CUYdT!Iu&ufr6VxzF{1a*(p1Z%Ji+KKyOZ{mk_ePSOfGMn zr?lHo-y<~@)SYAf0)HnydIM%-dmj+UPwB6mKEb;hdh3S)1= znk@%?`hlOj7HW4L2ggLM)sJ_2#F77N5GK$EyE3Np@kez3dlaCeR}QTWd&2xsR-K)< zUq=ww#xs*Wd4p>>p@Pxz#Y9BEh)!(wunr=}9OKvr; zJTEeiyVU?+K%l=%8y9=n0g1a&_sKsCYh+`E=Ztf~Ut0e0NJ(_I#fa)SmEBv7?}6Qr zs!q^r{TeHuu>i$SUsY#MuVp%GazY^}Ms=7=gUg#kw#T|G8TfvoZWlXOQNQDaUd9f7 z0sRL^Qohit0s*QpQz{lNS2KgVF}z|N%eA{x}OHVu=fGMwt0v@t6hLNMURo{yme77TU@IHFQ zu_y+33nw53xq#=6|B&(XVJyMU<30>@@Qs>eXIvQhVxjA?6wx9ZV41hs|DLeG=o~lB zyFtkkX?wXwB(3w`nC z7TmY7lti3=rxs$EsqSH(z`cGoZsqjdz%851-(^bQ~SY&s4yzthOhtDjcX( zd$+A^qr+aKgZXt5B?96?qq#cQpO(Cp`eu92YNz!Utq7-x5)25Qoe9LPxya!d!Q%~2^mDQM{+{z-^eE*br!j@Lh|x4*J> zzI?(UtMp!He*qA9^s&f)LEK?u4z!b4Rrm5YRlMbKDfU9Tq_@R<={R(Q!u3Y-HEXC? z<(2VI74?|tcCI=5{O?M@=j#cTJR{_p zih$d2zh}_2G8Rk$q7tq=-tFuf&Ni9wKy~JqkzU}ZXM@75lJl)8)%Um5fPduD`H1Cn0+%b}+^o+D?F{9wQiG#e6d!{Q9Hx^9JZF5zmz2EB3n;htR zF7z#w&A)Y+$AF8iw_@J(A-(btbSd68=?=KgE-2590%a4#-&0U|3Tcpg*)I~|=FScn z6UoT&jMv-GiV!r0FF}@zOIwMb9ex@S_B7RKescxBnA?C%+Z#-0Mg}NXa}Qibvw~=2 zI0%OiBBZJz@1_KyffjGLWwONDs&;LD)VstPj0}tau=d$X{u>s*W&Dy8u`D035EVm3 zh>0poAbgeYP3(>TuY>?RV02u6w9n-y!}oJ|ByUtVouT{fT7#nu3S61az3;|Y+8=*# ztte@~F7u_6^M-imAXstD9nf(!$B<&4vL+9s0|9}CIByHK7u*?Eo(`UAawjDWU3ieq z#*@Chuc`0X*OYyJl@X5UKKwgmmuhlW2yoiwfBo-KftCjsy%25FQHP*c2>pDgfu-A| zb20V(G_`X^lRmDczKVIX;dtZuaLdnj(|*FKgyQ%e>x)esZmlc|raLK(M=9b$I(q^- zg6WYKPc7BQtg6mKgE`ekHoWb3@NC1N%eos}6|=WU6Jl39A7YJ`1m-tTo=CyV#V|9Ln1F4?$~RPqp9julA72b_QZO`;!0C%HrXojwri}Xw_Fs zAq|DFC3%KVW(aWPVaDlEhSu~_X9dn2y*Jwy$?4#~vX@Wv3?0%)PTPMsKK8lW^vu(`EbnM9}Yn~QHO&FZybSy zyVo~;@_V|_%&&V66R2R!z7trowJsPIu4ps0;-GZL@-cWpStFi2T-8vKijGyK9Ja*# zA}f4Pda##E`9uHNWHOh{=B*cS?t?!ayh&^7jrvT@8#Uj+cx}jnGp0|^<6tmW-!`Me zl@NbjJRr7(!e8HieMd;xW+>;XO_;3*PJ$-qrGOUD~vvs;4%DE{65nR zb!&td(s7l-Ur|Q3S#0jT^L5IZJna0V5|q>lxIIn2Vbsqh`Kq(kgnR|Y^=xnOZ*Y$G zUifv5)97q7y2pL=cDUe`Sq6QddoMd#2>c=4Eu3upiKvap{*PksC9lMSNlU(?L7}_= zaHC^MyOW|}^FtqX0iGxlmEb^mFdOpb?`!NvWI`_Z`E)&hho?|&QsY^1hqR1%LRa3I zoc;m!`CQ7qBmUsj?-Xm!Uak+y+!LJ0ckxr1v0@{@mU~9(#(QQ;Nr1Kp1h9J?hxCb{ z^EjiT^t4MvPiFp(-I@UcLX#jZC1eUh76nJ?5laKyLhI70!529`|7Fw}vh| z(=Shd!#$$VYtMnvmm%iLb_gXskWkYQm{$T!U~Y2W`JhqT#oghWJTRcVLhy%_aq$=Nm1a7RnopsAH zV^iGEmIoP4CVg1y)YapJ;$uDYOg7M4kS35uDBy7RF1_{$QgQgrxUF0Y?I35|T>iza z7k#HzwI3;+P=kRFq|yHCsS@r&jpI#5Hq0}VHq*Tp{#RN%%CIc1-fqRN%*Jn}kE#%C z_55k6Pr2C2iK?f&sy)H&prP#W$xtpSVE;C0`jk8!mDf(+3tX>(7*-!HDm2zeFJuzb z@a}l`D#Pu$N!!STP~qg3h}5pdiI&KKXSlnCz&~59=esP4e@!C!l2dBsq4xp$FSs0~ z)!Bx(T$ruh`Qo}WCFdA)yWbW65h7!=WxNd_=@z!al>4mDc+i-ZVO@R0RT|`c`h0-C zk9X7tY)k{@XETJ7L#m0L4Kiu!3mHDShebJka${_Pne35nY7g?ceH2pdcGz`3-FJo} zAMgg+lpE3+^iJ$OQ1O^E$YAWyAifp@*CxW$2>D?yB>M%?K=x{O_9aS45^^#r5$ud< z)_a+oacgZ5s6X1p_g{L)mvL=+5Y3P@Rqe!ZJ@?`e9H%ji%; zwPpHmE>=KB`MGnd`)67JHVmbWBSxe4YZ9KhMNcyIQ`YP}$I`r}Y8+pBDvxupH1j7* zl1pNdj`dqh3t<1(gC{{6?_;J6q=)nrNNdUq{moG#CeDGBoDG8|I{(kaHM}KxzxvhS&Qs2fErb*AirXXg~dz8Vb`M9_~+F#d>-$b!daXwRY z3VBWKlL*(T&2tXQqx%wVdeVzg5+%;+$2rsjhU1Ngy5}wNv2C&%>MV1WP6YyG zg%8DmRzMfku6(;87IC#90aHa0W+Lrz2KGOI86Aa`nlo+(M$Db17>S}a+#0}0=o=3$Ho?JJp{C^t0cxcnb8H;^N|8(JvEx7MSbpAD|B?QpsNdz&J4P3o;yON)f zS-S@GvU@lYYiZZ}%?z-2yN7Ex1fWDd+Z1x@usLP!e7r)6-Ta~M zW&Y@XI8uj0@|KLw(9O7;WjAjLnS1ryuM#WSvWDtn?=IbyVN1qn zUtF{Bk6Fr&q6ATs*D=3v2UVo%#@Yj?K-FpNEXZP`W;oRv1?sTwC5~F74n89Uc5!u! zcQJQ=WS|WPs`LRKAJ+5(jfAEuhnELeVpr#vqgJp97av{?rj@EvqiXI!o_s@6pTF*c zqdun3iejSaLc6$flrUWiyOQZ?5Q8?>rca_%)|PtUuiwAP#z;?+yu;1NSE!({R3D5- z|HS8%2dg8wRU7mQ(xkl0TLtb?^NiI52Q+gAY&>Yb#q3FEP|)t{P`cxZG;>ovMi+C?fj_GbIOSuF2ij#CRDFbc z!tC_ym4n8?DWzoF_dzvU?V9*0y^|x8cp#A68 zE?-?Zr~ou!7{)v#JywjQ&Ehq8hL^Lr9jcO7e#_m}GxM~0_E z{-_Ju`z&lHZyjFqBAuuA{n0>eKGxbK^7N06g|-EDGL)Wkt*7eUk9kW~bMfiDGx0Hu z1!Ml7E>CkmxsOz=u~nD+cgCi#ua<{K)?zJ*mGzA;^YbVh$rH1VHI0_^FNQSvUU|z= zIM)qSd)Br3pp!Lz^;CQwC8`+HCx$`#eeQvPh8^+7ooY_Kaq@aR?eF1$g^Ue9%+{LQ z(YiwqG`7fNrA(2>aOVaq4M=bu-}}nTN%7SGgNBlBPYL-WZpGs1;@HagDh1rI4tG{U z{g^)%nBEq^G!4*HQKU9bX)&P%UnDSrMP8};TrlXcT&XuLgl=vc`$WI(!t2{ z-7~m)-fL^O@&6 zyvwtrJe#j&FlDfXQZdQt=z+SqzkbA{ao9d^+401;P5S}C3Y&dKB+1s0yv(y0yCKvFE{$rc|YoPmCcA zo5)S2jZ5yBl!FHEt20Zwa=~{)FEV``!wyg$j5Z>V$8Cf}H**K~Gd}O=VR8rOU)Pi- zSROf_bU5=**LwQXPB0fU(ktx?9X2tbtY0_1^wM*7fv8+xU>xE7I_p#uBxc=w@7Qh1 zr9}IqEsb2|_CsqQLnDIsNQN6flh3F(7Mu3o-_|+kQGK~O($MhL#%5#9!p6j<&XWq* zaYY1uUcCJdUo0xZf1rRc=359f62saz{BTbHdn?!fxf(huWG2yTSZKg2P6C%S(AbGX7}Zmd0hTWGqT2>Tehccz3#wwbK&li=43^sT1pfX|`9>^RXRmczp<1)|Wf6$Vh=_!KtyC$NEH6nc= zr;C{~2P`&3cGCecm591Je7mYOGeoUMzF=^r%7gxHVUn}@gGPdFedeE)8I;1KfW}s_ zucTkL;N7ngG6_Yg;?%Gm9l?4$D+n{>db0nQ z`nE>qIJ=MMH>G3jHh|+*z9UIHf_0arJt4vxJ;>JaX+G9{%Nj-5kAdSx`6pg=^8yj| z2<=h+^y3MF-~#Y8{?^r{Uajft88x13BcmPQc*$FSMJ(6OvtIxzc5L`R=Z4F?|6;bE ze*#TdtgETZFH-$e`9teJOF%1?5)_cqfU+m=S2ULrA;@HH)#I~K+t`3q2SlS$5%#<4 zD-)zNc?{X#yWemrKjSN4S<9lKjyT8NcN)YKmMiYZo79d;woY&)X0SEBu zeYnC|bxP$Klx%KK+&5Oq9Sm3^?PK0i9-Te9x}2Jh7OfkvZ}kXn6QQ@#B?XoBz|#6oUjJUs~pGn z-r2m7G5uql@c4bN166vNy%+k5@)IobRXta^T;`1K-6QZH*O$@wk_-aNY2wJ~z*FI|cJ7=o@D9%DdHX7j2>D(TMim?mT&?8`m=i`VY_W;M5AaMx1NB3 zm}7o9*OK6$R7p9iijEbVmy5jHW1J7eoq-$tL!zM&M-bqamfw)=RALtv>qZDcv{ZG` z@zeRO_ni9(iS?EEQQTosQl=L+c+|6w5cywHknfX4=NX{Z&Q;rNs4(y^j=M+ zfLT55ennj%$fteF^}PTma%IG#T|5>ST?kb;E{gGCy} zTt@PjC94`feRU>BH6BJ`?nVQU^`t*V4ch}r4SR+~<|GG_y=al-U{c&3i`lVV6s1l^ z>tDWMa>h5n0e`?}Kg5BG`gCsFvSmk4yIWD&}zQ^3p^cWHbzOF+JVw1iTw^f7HLvMscRmw^hY@%Z2Unjv2ec|=X zL*td>uXYo}GWOBdy2Xx-`M0lrjR_EYALZ-NSXvJ&6;L8Ruf(W=xC0gp0u~shUmyQ< z>RwZS!%7Ax|ILGdO?f{%r;L>*lbSkYj|pf)W@0`gomglq(EN&M$R`i8OdS^N>Af<@cytVVtK#tU|`2?GZGIsQ7=Op_+%}UEAw1 zGCA-HJn(`9k}-iN7$M)bV_d*GP7vzXM;J^Rfb1Dh`#gg6DI&*+ILfxr>Gsk=Aojo? zyGQAH?@ssB^mp#V=#@ET3X=tDc+SApB z9Y(Hj=606qE<6ZWm3p1>4JRTpqda<4y!(p#w~_+hNLqKMWAg&#?7ldj$N*;)!2W&) z7JZrxQTxajzY@?EDcIDVg!LBlrwv4U8rsv;?+v%sLSS3J(j2Mn(BGjJzcWKTD)lH* zOY6=AN}K1#rqqMusx*IUlc^MJ^Jd!D6e_my>XfzZ*uax*o?l}BPkUb))zsDfEg?Y= zQ4|!d73hO9h=8IpL!i}CM26ykjKLy}lz9wO0>oNvsUR{$WC#ivm3fdk1f&uL86-l0 z06{Sf2}vL!2@sOJiQ2CBY5y;*^?vw&>iKZjx+^zlpS{l>etVyD&$W0bw-Y4y?A^kw z-F`0SwF9+SVT{{l+_u!{>%b2TG4%OC4dLNz9Vo3AaFgypJ(~2$aZs48^UdAalR^Am zlx_a#(^=9aw|=_sOtK^2h+G|NTTLSZa8`S5c@V{-F6AT&%29JNI!R+zo)@U@Y zdXYI^cq1yU)j2z8D5JL(1W;Iv8P9}o%e;Ttz`xXvPjuv%>h;9=o$ZcjYo!buzsh2! zmyZ?hkd&7a$NlDcKW>-NfqqLZc}?#O_Y1CAAaJ@py=iAZ0SmELzqtFd=&Z;Y{T;+% z(0I_GRw()M_M3jeEA;xSX}f*MNQ8TP#_8qqL-5mrDs02T@-u_5c~soouGcyI5yu2eASQq1BL5@A z1VyeLW)nKb2msid_s{*V6EL(8UYyGImiET&DhkvEq(i*J*iQrQ$C7WSpuJjyj*(tu zQshsL_dWD;b@G|LmGT3jZ|m&yVYx3%VPz4*mVK~>T)aWsJB5(Oce`v<^XC~gOnw{F zT>m8~q8>{>#4Pc8tQHb1@NJ41o=yZFRFILK*|%MBkreXs3AsKi(Qc~mOq)QS(P-}^ z(2FSODV>~bQ97rJ{&ff+JyvIp=I~o}d-xOSP?42x{e(4=(_MeNuq2A)ji5oHik;J! zbKrB|sF@)zj2#`BV_2uXRK62GHQe4c_7cH? zhQBy*P`Yoj6qS(8H!sp_`DVbVI~;~Kl2eXf*YZd^v*+}OFe6jPZsowMI%N9p2>ypF z;dw->9y(+{burr`+4{O&B1ZOy4WbHG~GCyE8Le>MIb`YdG@) zZ`5C@Md{J1w!3ZFZRtscR_T-qrK3Dj;+i{{7XgMPJn2!=Dq!c$j>oVv&h8FC~okytBiB#Qd zU-=iMXa$8!l`v#lu||4#uBcCh$J!HZ+gHCDdDj}+#*Fyze6j|<_truyiVVWGOn6qw z%jV>nGa}>12nEp=&3K<=c7UJUb3IdXCVT4oPF1NdaZ585V83SjbIy2Qo1g#);4Jcc z8HLRGU^8hcq*=DyRsN4vQQT6!NBLwc=A?lM#*2Tzhm_B!Ufq4?E}oTBc5a7ePp=-N zro7H?_<(k`f#2A-xpn8aTo@osy^Pc4zf_6JfE2QyIKin3 zFfTJBg!%VwD43rby3Y>rPbO9>{WbOKg`%RwL-Xu1QIl|K(t|k*T=q|- zVi43hnw-qq?H{FV1Yi;kuM?6%UuG_!;iPc1$eaq;LCD--C7;0*GKwJcJXwx{!&*Z3 z8<8tG%DH~N4)G|y@!7h0q#4EC3=4ylAzG|Ivj9Qu~v zR(Z31upYQ(arub`x<~ND4N3z@BzY_EvIqoO1R1aQ;pM#-k<$!aDU}XOQasjB$Oenp zCR^e1F*bRU`bY%9w0WIzU+LwpkzJRSkDf4Y_p!Dvtw)7M$Ey?4y4UGsA-!l4&2K`- zcNX4EQnwF!BNH`>$O;=%1u2zcHD)GKW>zYOFh30x7i)VhS)TW5nEG*AB5dS=k;zJ&* zm`qLKc*|EAMl}?XtHM+s;m7n1`Xy)f8W)JxXVqp?r-}s&>yy>PT?B6b6?V@3{z00{ z=WOy&=n!zZWw~}a9AUO$!l?cJg7?*k2T__msQN&rYe>&fXKUt18mEBU zWW8@sU)Gf@f%* zRF!F243naKeb%RI!dSlzcz*kDs0vJyJk%yDpU`qy<(}JJDEe?uHa?bO!5UqeM+@gQ zw5J_y$g;vUvE-(CixxAs122sL*4>73!e+7A`uh3R^b~Yc*x}bD#W9FaQO{)46f8$%a_t z)!ZxQ_)g)$kRP*~O240?yC!m6(~L&?OCbX4bafuW)j5f`Q#iFVwd1SR=%Zgc;lE!z zZ>rlb1i}w$z>=CXoJm*+{4J#F%|vmZS_e34`PLHQMn{&@z*Jga-b&vQ;>*Q?fQcU7A}n zWDj;gU&tReXt)h`3$7@07<2w)|KkAT5W$j2RCEG5HmJ6CLLKX%o-C?Oa-?93g{l)N zWAp~o`0b8C4UYK4S8*9Bp!rCjnAL{9Le^csM}!VRdCOz=GpMG1tLOT)T6t~hG5x}$ zx?WCFv$rCy%)Ycz!_59{k*gh)->Ody5=l~F{4{)WrOQ~A3>3aM5WYVEVV{CTSZEH$ zYE~vo;pR2(u~WradE-i=fCfoy=LgCu%8fPlybclQbWSBt-GLwJY-%wp7@n{5_C^a@ zD{f=Z{)s9dJVpah^ek-T+FInqC}o(=eC0oJi5wJwju(?=C@+t^l^nKU+uoCOi%e$U zrdF=}R%x;@?gb^)@-P0Sk&2L5=OcYRRqUzKS`&~o+iG$(x-J$!Q_0e%Gj1>6u0c2am_yu9KBVX!DiAu>vN zY|sg~4dj@T#oLDEGX_0@ICB0lj1a;;Y$H(RqguVeuj4B@_~r8yw?tSh(Eu_TRGP=Z z7@;tqoPjAw>tU*EjS#gx@t( zhejnxe&|iTs&2p*J4=w1rOPF&w9L3y_TJ%deQmylo$i6UyEGh*8maS}JzgmaRqdv0 zd``fGm4LP1*7RtJRCGCR&b>Vx{s)K)luneSVsq2;v$24mlL`^{E!vc4di>UI(0W#X zm_jAILVf8JGtf{wehNwFNE1Wcz3Ti0#<8vXpD?wP7*iA3jYKL?^;B}yIc7Z|>N4|E z*}7Ql>A-dR3HsOL(+JNI$@V!*{bf+Fq#i#J({-XC49Hub9G7f~m{^fm(DYKOJ}ZcS z1F~|%cB)_FMne2$pMeQa59UL%btk}E<<%5o_{Eg0Q}U5y!;V_8fB?1xK#n#x`VKI+ zt<`hpMPKm~Z#7Bn_r6PS+<$s$%;(pWgsf@`OJA;DZYEY4vn9`5U?H>Y#f&5Fn#ywZ z01GTJ; zTiQD0H8tXsZfkEDN=vntwquox*8^-n38;7^C=si9RApyf)i4lB9l&v2k52YQS$ITa6vUB<^D)uDo3KooT zyf~}HoOi9>Uai@gesDpNI1wgGdNSK4PHIzau|`^p+S^VR^0ePK(15#R-iietVbdlb z3JZ+=eOeO8)yRt6m7WmcOo$a@JRE~gdM5@f{c2LGFqydGBJ;*JYH1ra=IA5xLFEhK zuu*_%%an1Wy5d!c>#3lm7oLlml=w+@fv@5Xl#NoVgvaV(q@c5&Z|kAaK#V- zEZX-n>|BWe_jnu>p|DVPyKhfrmR7l{cgu-^q`O=as@_D{ZCH3UzuCXG#`JL~5k5h* zRL=B7@ZVmVO@-gf)OZzfzf4y>=ExxJFJF=f9A**|S#fYp_F;jc;HV84GBZe4+P54h zFeTA)bC&HmLj{qvr9$tPZ=6K~bqE~iE@I4I6kgfj7|Y81ls90+c>^tWJRk~?cE8;p zx*UAu`8oydC z{+eRDuZ`snwxqVU%ZK&YSw)1AmX`r$8^u)dT*E4`0B;$W?N4Rz$tA>=J>9b3Rq8zK z1$O0yffb`K!nH*GzB$$6x1obk4C?u8-$RPpNzglI6rBc!8ma}~$_Hea%OC8q96ayF zg(})Q_r+Z|W3XyPTZ1Qlh4N>ucumHtyhi*)O!xVM@XYCzQH)4MCDc>h8NPQeYT*On zTmG|P8*|!G-$E}$tus7p{>wGVN~?zRPd;|-gG`5~1j+HsMf%id^}6TBR_8fXsG2=r zv{JXjgZEt@4Qz3@IV=D3$^g&a8vUXUkr~cGFKZCfnkU+pzvA$?bvHSdoTlYnSR&Zd zrENZ%QbCS%-%1)ylJ7@(sb1EVKayApwCCUQdI? zKq`?l;ET!ohPtFjX6fM%A{hl2LT1y}zPQ@#l|$mqFL||@=2gTWf&@%si-&ms75%(H z!)b=7WxrvAH<*lNOm>;E9*)8a8H=aEbsRGhrR>%EmWlPtqu`GsC!Q@#ntEn7 ze#$7>7UFYou%-N!Ss$#$>!XuNQ4fv#z72PM2H>c+KD9NZl~W#Ki7*(*_k%2-fkYln zTpSMhbw$u3)|D$W4bq^@@%`GXxz{`hG77-rqY(A^|J3o*pI%QYl*)vAxaYbZ+w);J zFcOw$z6IoJd3UIx0!!BTChN+6u%FzNj?+Mhw1Ir%Hre0-zvN^5-4@Funy?TvD>N?_ z4Fk%6$o1m~fAj9qr3beVd?}p1t(6TsS3v;OUd2q;L8GYU6&D+48ylSYlc105THy=C z*rH&`(F|~)0Z1x#|2Y@qAP@J^nRsl}Ozm-Se`HJZ_$y(5O99LU!}=3Z4k=$wJR|aj z>%bi(oTN* z4E+i?{1uke{nj7rQO(j4BLd}->HVOG>t0Gf?=IfG9^ZLjGSuQ1RvKHcA=tul5u*x=|4$Neg`E4m3j^11oRNFsuOhyGxjdMWUM5=?}6lV zjd&(#jc#Y-=f`ZV!iFC`-}cA5$f}=WowlaWz$e*jJ7@JKg@5W_FmUI&As!z67m6aRr>nznn-&rx3j2e&_$?Hc{I zU?O(DabMDs2U6F*@J&F^mnu>=X;aGIDOXt@@FazKYyJYoFnFE2os{rI!Kz@fbZ9BkosQ9)J{O*e43s1$1 zxq1-i{)v!;A3Fw;Qnoy3nFqe)ntY!|Rt}8Vik{X$o6Bas+f~?oT+f77VGY}b=$lRG zTgpvFOusq!OX{xz)4kLm58yiwJt;N2AD#ChgIo>qzWz0-%gpAi)VT~i^64y}EE3+N z`h&TG_Hy%cAWg{1a}`xiw5tF0C`RhlB1AVa`*6=;j(bh-nN<<5&bySH|JvZJfLUBY zvHM)|>Cr93Jwjuc4=VlQ8WAtoP7TwxQl-}(W(&vXje3l;%1Pl7{0UF1*m@DcEHIbi zmeczYF;$=>Q6j1}APA7w-0rU>4#l;p%!$DBlMFd*kCF|gQuKV0hN^eh1fr|~FL>x7&cpjhTw&w$Dq>oT6khxGdQP*(?)N&c@ zSyN7)Qrf=$($*>*s34N0^s?<(c0SW04_Y25b(+Gcu<(noD{b<5;H!JBfvl)(ask?5 zMl$^ckWnwt4>b`TnPpq%xA54qq5OE02VIvxVr8=`zTRgOr5y)kbdCg8(~xLFKpJ!hE|qN|57Nl z&<9OKhBCg0aqYKS57E5ZAqIf@S1`uAyi zOdQr0zwNzSuzA{vXZcCRHDyXG>=l(5tUC)UV}m1Mfij12g&YT&#e(ZenTc zD28E_?MhS|EWSeuzJ0ADVLUE$NG*n*ArVA z@;L387!k7!(hNTwNA-sjy6Xb@)#X?7{o!XwpA`zz-s>UqA14umDlx9*j-NUklU;Zo z;dR67D#P-YS72~(j{4U+$)*39N-#E={u0${EVsPBaaYv|xxLLrp*_IY7K0b62BuQT zN0dEJs?^neERv?RtXWVA@itZ@gT}cN4axtxZmub9BCi^9>>3(3wDfd@m0e%7MSb2Bd|UJR=| z_-6_0uS;m6SZyXD{bxUIl9=k2J*wU)qc>N@!>|d~?~uwET6})H-=5H5`2karLm#si z4T3GLw!YND6&CVlR7Fi&3ZjrVM``p361}h8Dj*}E;J>MfFK9A&{BR9yM_t9tdT3!SCm?dykE`m+yPa?`IZj;f!b!}Bu ziu$a*pnnLNguKC5=zt4>dr%*qiL!X1QEs1i>iyTH9k@-_@nP)e`*i~;sb2}pPmnv2 z2kdvMn2fK@$a(&lOo;F@cO;_`-|qe3F%S|9S_$+WGpnTV%zp+T5l*3K93iQY1#jEC zul056kY)1Ur?1ad%+6|DL&N7CCi(41m0Yxub<+DtGcO_ySn)8)dIAtLHp>|VAo=I= zCiMbZK!R-Y^l8ejeHT(&wyG?86p%o`1~jFbKI-Q|wD<>!-RjQPRqQ|StV-Ge>U@V2 zOcZzSd9VE0bB|Up?z&v)7{7Op9~o48yZ_hz-}}3))apqZ-(V;65#7&jPoe!-2Jojz zg+WH;WkLTF1(QA6>M>8>{F#ET97+ZOL7J^r3bM_$cE^0#03`YkBGO)8eTpkt5#0%r z22B?CI^}9rUS+UZ5D^Dr%sqgVXH76p)D#v7Ue2=XmY<0E4IJ}^2fJdj2>+y6A=!gV zgRcjCRs{4V9~lQK~N?x{N zJ4MHXH9Ce!c(k5c$n(%lTgPPhaceEQ$yQ~$hySh!)&Pvy#+**Y{h9FXlsl?^ML%Cg zE!%F(56QbD{H4+;{srxRDkD&z*%kkIN3a$~s)Nw~qTXW?YiwMezy>LzU^H&inCHQI za0L1VAqo?<5cPK|a)-n-74avB!MasGotN~jt9dh+}v+!6qP`@RKO@dzWM@zG7rJ5!s^E3+uiuHWWU!to6_!$$0s-bM9hhK))B}ijG&ywPUf)1wTq@3 zj!!+=Tryu67mEERjO>;m`JcCzAJN)ye1}3McEh~zG+`4)JinIgNLVbqNZDlW7#GBx zrE!+fuWQ=k`$ORsi%m9FuD?-V8uu&SE#S5`esg=KwRZiHe4c z_M?~ftJGYyG+h2#HTy``hIAIM$sHYNXo|~go8%Mx7>Y;LSFRnb=H_lgE~v!UyMK$QQt zFNR6h4h7pE3pREluI}QIW?8kK(%+bDvxF=IYE5_AaD0t}lG5e!yYjG&$r8Q9WTU{+ zKP7JLNVE2e!9&T;klP!JrI8^a%ir(JSK3hC6?I~8qYIA-@Z6Yel!}<_M1Rs4dSm}v zBB^i8cX~{3Jk1p??Y(K_vx-*7jd|!bZycWQw3kS5 zvd&JAtj!QWpE&WJ%sJn*nS3)PIGJ*#ba0d4JDwxK$q`3fH(>q%@vjQiYFhR#>*8|W}#!)_wW`o z(J3C;8&k3d@U$0iiIb>mD_-5`=+w03p22T7QFl)pv4OiT77iD0V&P)D4c*)ucR diff --git a/doc/pub/week48/ipynb/week48.ipynb b/doc/pub/week48/ipynb/week48.ipynb index c3bf31609..2d26b01b4 100644 --- a/doc/pub/week48/ipynb/week48.ipynb +++ b/doc/pub/week48/ipynb/week48.ipynb @@ -2,909 +2,1582 @@ "cells": [ { "cell_type": "markdown", - "metadata": {}, + "id": "c3193a71", + "metadata": { + "editable": true + }, + "source": [ + "\n", + "" + ] + }, + { + "cell_type": "markdown", + "id": "a0b80e0f", + "metadata": { + "editable": true + }, "source": [ - "\n", "# Week 48: Support Vector Machines and Summary of course\n", - "\n", - " \n", - "**Morten Hjorth-Jensen**, Department of Physics, University of Oslo and Department of Physics and Astronomy and National Superconducting Cyclotron Laboratory, Michigan State University\n", + "**Morten Hjorth-Jensen**, Department of Physics, University of Oslo, Norway\n", "\n", - "Date: **Nov 29, 2020**\n", + "Date: **Nov 21, 2024**\n", "\n", - "Copyright 1999-2020, Morten Hjorth-Jensen. Released under CC Attribution-NonCommercial 4.0 license\n", + "Copyright 1999-2024, Morten Hjorth-Jensen. Released under CC Attribution-NonCommercial 4.0 license" + ] + }, + { + "cell_type": "markdown", + "id": "6da3d550", + "metadata": { + "editable": true + }, + "source": [ + "## Overview of week 48" + ] + }, + { + "cell_type": "markdown", + "id": "3de7e991", + "metadata": { + "editable": true + }, + "source": [ + "## Plan for week 47\n", "\n", + "**Lab sessions on Tuesday and Wednesday.**\n", "\n", + " * Work and Discussion of project 3\n", "\n", - "## Overview of week 48\n", + " * Last weekly exercise,\n", "\n", - "* **Thursday**: Support Vector Machines: Kernels, Classification and Regression. [Video of Lecture](https://www.uio.no/studier/emner/matnat/fys/FYS-STK4155/h20/forelesningsvideoer/LectureNovember26.mp4?vrtx=view-as-webpage).\n", + " \n", "\n", - "* **Friday**: Summary of course with perspectives for future studies. [Video of Lecture](https://www.uio.no/studier/emner/matnat/fys/FYS-STK4155/h20/forelesningsvideoer/LectureNovember27.mp4?vrtx=view-as-webpage).\n", + "**Plans for the lecture Monday 25 November, with video suggestions etc.**\n", "\n", - "See also Geron's chapter 5. Chapter 12 (sections 12.1-12.3 are the most relevant ones) of Hastie et al contains also a good discussion.\n", + "1. Bossting and gradient boosting and ensemble models\n", "\n", + "2. Summary of course\n", "\n", - "## Thursday\n", + "3. Readings and Videos:\n", "\n", - "We finalize our discussion on Support Vector Machines with an emphasis on kernel transformations and applications to regression. The following [video attempts at giving an overview on this part](https://www.youtube.com/watch?v=Toet3EiSFcM&ab_channel=StatQuestwithJoshStarmer). See also the [follow-up video](https://www.youtube.com/watch?v=Qc5IyLW_hns&ab_channel=StatQuestwithJoshStarmer).\n", + "a. These lecture notes at \n", "\n", - "## Friday\n", + "b. See also lecture notes from week 47 at . The lecture on Monday starts with a repetition on AdaBoost before we move over to gradient boosting with examples\n", + "\n", + "\n", "\n", - "Friday's lecture is split in two parts. It starts with a summary of\n", - "what we have done this semester and continues with perspectives for future studies and\n", - "modern research projects in machine learning.\n", + "c. Video on Decision trees \n", "\n", - "## Support Vector Machines, overarching aims\n", + "d. Video on boosting methods \n", "\n", - "As discussed last week, \n", - "a Support Vector Machine (SVM) is a very powerful and versatile\n", - "Machine Learning method, capable of performing linear or nonlinear\n", - "classification, regression, and even outlier detection. It is one of\n", - "the most popular models in Machine Learning, and anyone interested in\n", - "Machine Learning should have it in their toolbox. SVMs are\n", - "particularly well suited for classification of complex but small-sized or\n", - "medium-sized datasets. \n", + "e. Video on AdaBoost \n", "\n", - "The case with two well-separated classes only can be understood in an\n", - "intuitive way in terms of lines in a two-dimensional space separating\n", - "the two classes.\n", + "f. Video on Gradient boost, part 1, parts 2-4 follow thereafter \n", "\n", - "The basic mathematics behind the SVM is however less familiar to most of us. \n", - "It relies on the definition of hyperplanes and the\n", - "definition of a **margin** which separates classes (in case of\n", - "classification problems) of variables. It is also used for regression\n", - "problems. I recommend you take a look at the lectures from last week on the binary classification problem.\n", + "g. Decision Trees: Rashcka et al chapter 3 pages 86-98, and chapter 7 on Ensemble methods, Voting and Bagging and Gradient Boosting. See also lecture from STK-IN4300, lecture 7 at ." + ] + }, + { + "cell_type": "markdown", + "id": "bcf392e6", + "metadata": { + "editable": true + }, + "source": [ + "## Boosting, a Bird's Eye View\n", "\n", - "## Kernels and non-linearity\n", + "The basic idea is to combine weak classifiers in order to create a good\n", + "classifier. With a weak classifier we often intend a classifier which\n", + "produces results which are only slightly better than we would get by\n", + "random guesses.\n", "\n", - "The cases we studied last week were all characterized by two classes\n", - "with a close to linear separability. The classifiers we have described\n", - "so far find linear boundaries in our input feature space. It is\n", - "possible to make our procedure more flexible by exploring the feature\n", - "space using other basis expansions such as higher-order polynomials,\n", - "wavelets, splines etc.\n", + "This is done by applying in an iterative way a weak (or a standard\n", + "classifier like decision trees) to modify the data. In each iteration\n", + "we emphasize those observations which are misclassified by weighting\n", + "them with a factor." + ] + }, + { + "cell_type": "markdown", + "id": "3496f6ab", + "metadata": { + "editable": true + }, + "source": [ + "## What is boosting? Additive Modelling/Iterative Fitting\n", "\n", - "If our feature space is not easy to separate, as shown in the figure\n", - "generated by the code below, we can achieve a better separation by introducing a more complex\n", - "basis functions. The ideal would be, as shown by the code example below, to,\n", - "via a specific transformation to obtain a separation between the\n", - "classes that is almost linear. See also Figures 12.2 and 12.3 of [Hastie et al.](https://www.springer.com/gp/book/9780387848570).\n", + "Boosting is a way of fitting an additive expansion in a set of\n", + "elementary basis functions like for example some simple polynomials.\n", + "Assume for example that we have a function" + ] + }, + { + "cell_type": "markdown", + "id": "3a5f4cf4", + "metadata": { + "editable": true + }, + "source": [ + "$$\n", + "f_M(x) = \\sum_{i=1}^M \\beta_m b(x;\\gamma_m),\n", + "$$" + ] + }, + { + "cell_type": "markdown", + "id": "3161dc94", + "metadata": { + "editable": true + }, + "source": [ + "where $\\beta_m$ are the expansion parameters to be determined in a\n", + "minimization process and $b(x;\\gamma_m)$ are some simple functions of\n", + "the multivariable parameter $x$ which is characterized by the\n", + "parameters $\\gamma_m$.\n", "\n", - "We can achieve a better separation by introducing more complex\n", - "basis functions. The ideal would be (see Figures 12.2 and 12.3) to, via a specific transformation to \n", - "obtain a separation between the classes which is almost linear. \n", + "As an example, consider the Sigmoid function we used in logistic\n", + "regression. In that case, we can translate the function\n", + "$b(x;\\gamma_m)$ into the Sigmoid function" + ] + }, + { + "cell_type": "markdown", + "id": "88356147", + "metadata": { + "editable": true + }, + "source": [ + "$$\n", + "\\sigma(t) = \\frac{1}{1+\\exp{(-t)}},\n", + "$$" + ] + }, + { + "cell_type": "markdown", + "id": "1d8f6ff1", + "metadata": { + "editable": true + }, + "source": [ + "where $t=\\gamma_0+\\gamma_1 x$ and the parameters $\\gamma_0$ and\n", + "$\\gamma_1$ were determined by the Logistic Regression fitting\n", + "algorithm.\n", "\n", - "The change of basis, from $x\\rightarrow z=\\phi(x)$ leads to the same\n", - "type of equations to be solved, except that we need to introduce, for\n", - "example, a polynomial transformation to a two-dimensional training\n", - "set." + "As another example, consider the cost function we defined for linear regression" + ] + }, + { + "cell_type": "markdown", + "id": "66b214e6", + "metadata": { + "editable": true + }, + "source": [ + "$$\n", + "C(\\boldsymbol{y},\\boldsymbol{f}) = \\frac{1}{n} \\sum_{i=0}^{n-1}(y_i-f(x_i))^2.\n", + "$$" + ] + }, + { + "cell_type": "markdown", + "id": "007ebce9", + "metadata": { + "editable": true + }, + "source": [ + "In this case the function $f(x)$ was replaced by the design matrix\n", + "$\\boldsymbol{X}$ and the unknown linear regression parameters $\\boldsymbol{\\beta}$,\n", + "that is $\\boldsymbol{f}=\\boldsymbol{X}\\boldsymbol{\\beta}$. In linear regression we can \n", + "simply invert a matrix and obtain the parameters $\\beta$ by" + ] + }, + { + "cell_type": "markdown", + "id": "818f650f", + "metadata": { + "editable": true + }, + "source": [ + "$$\n", + "\\boldsymbol{\\beta}=\\left(\\boldsymbol{X}^T\\boldsymbol{X}\\right)^{-1}\\boldsymbol{X}^T\\boldsymbol{y}.\n", + "$$" + ] + }, + { + "cell_type": "markdown", + "id": "2a0f1f70", + "metadata": { + "editable": true + }, + "source": [ + "In iterative fitting or additive modeling, we minimize the cost function with respect to the parameters $\\beta_m$ and $\\gamma_m$." + ] + }, + { + "cell_type": "markdown", + "id": "20fd7a4b", + "metadata": { + "editable": true + }, + "source": [ + "## Iterative Fitting, Regression and Squared-error Cost Function\n", + "\n", + "The way we proceed is as follows (here we specialize to the squared-error cost function)\n", + "\n", + "1. Establish a cost function, here ${\\cal C}(\\boldsymbol{y},\\boldsymbol{f}) = \\frac{1}{n} \\sum_{i=0}^{n-1}(y_i-f_M(x_i))^2$ with $f_M(x) = \\sum_{i=1}^M \\beta_m b(x;\\gamma_m)$.\n", + "\n", + "2. Initialize with a guess $f_0(x)$. It could be one or even zero or some random numbers.\n", + "\n", + "3. For $m=1:M$\n", + "\n", + "a. minimize $\\sum_{i=0}^{n-1}(y_i-f_{m-1}(x_i)-\\beta b(x;\\gamma))^2$ wrt $\\gamma$ and $\\beta$\n", + "\n", + "b. This gives the optimal values $\\beta_m$ and $\\gamma_m$\n", + "\n", + "c. Determine then the new values $f_m(x)=f_{m-1}(x) +\\beta_m b(x;\\gamma_m)$\n", + "\n", + "We could use any of the algorithms we have discussed till now. If we\n", + "use trees, $\\gamma$ parameterizes the split variables and split points\n", + "at the internal nodes, and the predictions at the terminal nodes." + ] + }, + { + "cell_type": "markdown", + "id": "a5253a01", + "metadata": { + "editable": true + }, + "source": [ + "## Squared-Error Example and Iterative Fitting\n", + "\n", + "To better understand what happens, let us develop the steps for the iterative fitting using the above squared error function.\n", + "\n", + "For simplicity we assume also that our functions $b(x;\\gamma)=1+\\gamma x$. \n", + "\n", + "This means that for every iteration $m$, we need to optimize" + ] + }, + { + "cell_type": "markdown", + "id": "82a331f2", + "metadata": { + "editable": true + }, + "source": [ + "$$\n", + "(\\beta_m,\\gamma_m) = \\mathrm{argmin}_{\\beta,\\lambda}\\hspace{0.1cm} \\sum_{i=0}^{n-1}(y_i-f_{m-1}(x_i)-\\beta b(x;\\gamma))^2=\\sum_{i=0}^{n-1}(y_i-f_{m-1}(x_i)-\\beta(1+\\gamma x_i))^2.\n", + "$$" + ] + }, + { + "cell_type": "markdown", + "id": "38f0123c", + "metadata": { + "editable": true + }, + "source": [ + "We start our iteration by simply setting $f_0(x)=0$. \n", + "Taking the derivatives with respect to $\\beta$ and $\\gamma$ we obtain" + ] + }, + { + "cell_type": "markdown", + "id": "8885d64f", + "metadata": { + "editable": true + }, + "source": [ + "$$\n", + "\\frac{\\partial {\\cal C}}{\\partial \\beta} = -2\\sum_{i}(1+\\gamma x_i)(y_i-\\beta(1+\\gamma x_i))=0,\n", + "$$" + ] + }, + { + "cell_type": "markdown", + "id": "bc3b9f34", + "metadata": { + "editable": true + }, + "source": [ + "and" + ] + }, + { + "cell_type": "markdown", + "id": "308960c8", + "metadata": { + "editable": true + }, + "source": [ + "$$\n", + "\\frac{\\partial {\\cal C}}{\\partial \\gamma} =-2\\sum_{i}\\beta x_i(y_i-\\beta(1+\\gamma x_i))=0.\n", + "$$" + ] + }, + { + "cell_type": "markdown", + "id": "4e6d45c0", + "metadata": { + "editable": true + }, + "source": [ + "We can then rewrite these equations as (defining $\\boldsymbol{w}=\\boldsymbol{e}+\\gamma \\boldsymbol{x})$ with $\\boldsymbol{e}$ being the unit vector)" + ] + }, + { + "cell_type": "markdown", + "id": "93f28313", + "metadata": { + "editable": true + }, + "source": [ + "$$\n", + "\\gamma \\boldsymbol{w}^T(\\boldsymbol{y}-\\beta\\gamma \\boldsymbol{w})=0,\n", + "$$" + ] + }, + { + "cell_type": "markdown", + "id": "c95dde03", + "metadata": { + "editable": true + }, + "source": [ + "which gives us $\\beta = \\boldsymbol{w}^T\\boldsymbol{y}/(\\boldsymbol{w}^T\\boldsymbol{w})$. Similarly we have" + ] + }, + { + "cell_type": "markdown", + "id": "01efcbbe", + "metadata": { + "editable": true + }, + "source": [ + "$$\n", + "\\beta\\gamma \\boldsymbol{x}^T(\\boldsymbol{y}-\\beta(1+\\gamma \\boldsymbol{x}))=0,\n", + "$$" + ] + }, + { + "cell_type": "markdown", + "id": "5049378a", + "metadata": { + "editable": true + }, + "source": [ + "which leads to $\\gamma =(\\boldsymbol{x}^T\\boldsymbol{y}-\\beta\\boldsymbol{x}^T\\boldsymbol{e})/(\\beta\\boldsymbol{x}^T\\boldsymbol{x})$. Inserting\n", + "for $\\beta$ gives us an equation for $\\gamma$. This is a non-linear equation in the unknown $\\gamma$ and has to be solved numerically. \n", + "\n", + "The solution to these two equations gives us in turn $\\beta_1$ and $\\gamma_1$ leading to the new expression for $f_1(x)$ as\n", + "$f_1(x) = \\beta_1(1+\\gamma_1x)$. Doing this $M$ times results in our final estimate for the function $f$." + ] + }, + { + "cell_type": "markdown", + "id": "6994caec", + "metadata": { + "editable": true + }, + "source": [ + "## Iterative Fitting, Classification and AdaBoost\n", + "\n", + "Let us consider a binary classification problem with two outcomes $y_i \\in \\{-1,1\\}$ and $i=0,1,2,\\dots,n-1$ as our set of\n", + "observations. We define a classification function $G(x)$ which produces a prediction taking one or the other of the two values \n", + "$\\{-1,1\\}$.\n", + "\n", + "The error rate of the training sample is then" + ] + }, + { + "cell_type": "markdown", + "id": "a367fb84", + "metadata": { + "editable": true + }, + "source": [ + "$$\n", + "\\mathrm{\\overline{err}}=\\frac{1}{n} \\sum_{i=0}^{n-1} I(y_i\\ne G(x_i)).\n", + "$$" + ] + }, + { + "cell_type": "markdown", + "id": "4b3dccdb", + "metadata": { + "editable": true + }, + "source": [ + "The iterative procedure starts with defining a weak classifier whose\n", + "error rate is barely better than random guessing. The iterative\n", + "procedure in boosting is to sequentially apply a weak\n", + "classification algorithm to repeatedly modified versions of the data\n", + "producing a sequence of weak classifiers $G_m(x)$.\n", + "\n", + "Here we will express our function $f(x)$ in terms of $G(x)$. That is" + ] + }, + { + "cell_type": "markdown", + "id": "bdcfb41f", + "metadata": { + "editable": true + }, + "source": [ + "$$\n", + "f_M(x) = \\sum_{i=1}^M \\beta_m b(x;\\gamma_m),\n", + "$$" + ] + }, + { + "cell_type": "markdown", + "id": "08ec0a91", + "metadata": { + "editable": true + }, + "source": [ + "will be a function of" + ] + }, + { + "cell_type": "markdown", + "id": "0cf3f4ea", + "metadata": { + "editable": true + }, + "source": [ + "$$\n", + "G_M(x) = \\mathrm{sign} \\sum_{i=1}^M \\alpha_m G_m(x).\n", + "$$" + ] + }, + { + "cell_type": "markdown", + "id": "4b8e7ed6", + "metadata": { + "editable": true + }, + "source": [ + "## Adaptive Boosting, AdaBoost\n", + "\n", + "In our iterative procedure we define thus" + ] + }, + { + "cell_type": "markdown", + "id": "235e8777", + "metadata": { + "editable": true + }, + "source": [ + "$$\n", + "f_m(x) = f_{m-1}(x)+\\beta_mG_m(x).\n", + "$$" + ] + }, + { + "cell_type": "markdown", + "id": "f6f51dc4", + "metadata": { + "editable": true + }, + "source": [ + "The simplest possible cost function which leads (also simple from a computational point of view) to the AdaBoost algorithm is the\n", + "exponential cost/loss function defined as" + ] + }, + { + "cell_type": "markdown", + "id": "8890839a", + "metadata": { + "editable": true + }, + "source": [ + "$$\n", + "C(\\boldsymbol{y},\\boldsymbol{f}) = \\sum_{i=0}^{n-1}\\exp{(-y_i(f_{m-1}(x_i)+\\beta G(x_i))}.\n", + "$$" + ] + }, + { + "cell_type": "markdown", + "id": "52acdfd0", + "metadata": { + "editable": true + }, + "source": [ + "We optimize $\\beta$ and $G$ for each value of $m=1:M$ as we did in the regression case.\n", + "This is normally done in two steps. Let us however first rewrite the cost function as" + ] + }, + { + "cell_type": "markdown", + "id": "9add0067", + "metadata": { + "editable": true + }, + "source": [ + "$$\n", + "C(\\boldsymbol{y},\\boldsymbol{f}) = \\sum_{i=0}^{n-1}w_i^{m}\\exp{(-y_i\\beta G(x_i))},\n", + "$$" + ] + }, + { + "cell_type": "markdown", + "id": "a44a0cd1", + "metadata": { + "editable": true + }, + "source": [ + "where we have defined $w_i^m= \\exp{(-y_if_{m-1}(x_i))}$." + ] + }, + { + "cell_type": "markdown", + "id": "3531c7dd", + "metadata": { + "editable": true + }, + "source": [ + "## Building up AdaBoost\n", + "\n", + "First, for any $\\beta > 0$, we optimize $G$ by setting" + ] + }, + { + "cell_type": "markdown", + "id": "633d1c21", + "metadata": { + "editable": true + }, + "source": [ + "$$\n", + "G_m(x) = \\mathrm{sign} \\sum_{i=0}^{n-1} w_i^m I(y_i \\ne G_(x_i)),\n", + "$$" + ] + }, + { + "cell_type": "markdown", + "id": "ff836742", + "metadata": { + "editable": true + }, + "source": [ + "which is the classifier that minimizes the weighted error rate in predicting $y$.\n", + "\n", + "We can do this by rewriting" + ] + }, + { + "cell_type": "markdown", + "id": "f3f23906", + "metadata": { + "editable": true + }, + "source": [ + "$$\n", + "\\exp{-(\\beta)}\\sum_{y_i=G(x_i)}w_i^m+\\exp{(\\beta)}\\sum_{y_i\\ne G(x_i)}w_i^m,\n", + "$$" + ] + }, + { + "cell_type": "markdown", + "id": "8b4563bf", + "metadata": { + "editable": true + }, + "source": [ + "which can be rewritten as" + ] + }, + { + "cell_type": "markdown", + "id": "4420620d", + "metadata": { + "editable": true + }, + "source": [ + "$$\n", + "(\\exp{(\\beta)}-\\exp{-(\\beta)})\\sum_{i=0}^{n-1}w_i^mI(y_i\\ne G(x_i))+\\exp{(-\\beta)}\\sum_{i=0}^{n-1}w_i^m=0,\n", + "$$" + ] + }, + { + "cell_type": "markdown", + "id": "d79ce5d4", + "metadata": { + "editable": true + }, + "source": [ + "which leads to" + ] + }, + { + "cell_type": "markdown", + "id": "6ae1d2b3", + "metadata": { + "editable": true + }, + "source": [ + "$$\n", + "\\beta_m = \\frac{1}{2}\\log{\\frac{1-\\mathrm{\\overline{err}}}{\\mathrm{\\overline{err}}}},\n", + "$$" + ] + }, + { + "cell_type": "markdown", + "id": "adb45dfe", + "metadata": { + "editable": true + }, + "source": [ + "where we have redefined the error as" + ] + }, + { + "cell_type": "markdown", + "id": "33cf4819", + "metadata": { + "editable": true + }, + "source": [ + "$$\n", + "\\mathrm{\\overline{err}}_m=\\frac{1}{n}\\frac{\\sum_{i=0}^{n-1}w_i^mI(y_i\\ne G(x_i)}{\\sum_{i=0}^{n-1}w_i^m},\n", + "$$" + ] + }, + { + "cell_type": "markdown", + "id": "70d46dfa", + "metadata": { + "editable": true + }, + "source": [ + "which leads to an update of" + ] + }, + { + "cell_type": "markdown", + "id": "0812d0a8", + "metadata": { + "editable": true + }, + "source": [ + "$$\n", + "f_m(x) = f_{m-1}(x) +\\beta_m G_m(x).\n", + "$$" + ] + }, + { + "cell_type": "markdown", + "id": "783db749", + "metadata": { + "editable": true + }, + "source": [ + "This leads to the new weights" + ] + }, + { + "cell_type": "markdown", + "id": "cc77503c", + "metadata": { + "editable": true + }, + "source": [ + "$$\n", + "w_i^{m+1} = w_i^m \\exp{(-y_i\\beta_m G_m(x_i))}\n", + "$$" + ] + }, + { + "cell_type": "markdown", + "id": "a0948190", + "metadata": { + "editable": true + }, + "source": [ + "## Adaptive boosting: AdaBoost, Basic Algorithm\n", + "\n", + "The algorithm here is rather straightforward. Assume that our weak\n", + "classifier is a decision tree and we consider a binary set of outputs\n", + "with $y_i \\in \\{-1,1\\}$ and $i=0,1,2,\\dots,n-1$ as our set of\n", + "observations. Our design matrix is given in terms of the\n", + "feature/predictor vectors\n", + "$\\boldsymbol{X}=[\\boldsymbol{x}_0\\boldsymbol{x}_1\\dots\\boldsymbol{x}_{p-1}]$. Finally, we define also a\n", + "classifier determined by our data via a function $G(x)$. This function tells us how well we are able to classify our outputs/targets $\\boldsymbol{y}$. \n", + "\n", + "We have already defined the misclassification error $\\mathrm{err}$ as" + ] + }, + { + "cell_type": "markdown", + "id": "842d68bc", + "metadata": { + "editable": true + }, + "source": [ + "$$\n", + "\\mathrm{err}=\\frac{1}{n}\\sum_{i=0}^{n-1}I(y_i\\ne G(x_i)),\n", + "$$" + ] + }, + { + "cell_type": "markdown", + "id": "44c88ed6", + "metadata": { + "editable": true + }, + "source": [ + "where the function $I()$ is one if we misclassify and zero if we classify correctly." + ] + }, + { + "cell_type": "markdown", + "id": "b3a7c4c1", + "metadata": { + "editable": true + }, + "source": [ + "## Basic Steps of AdaBoost\n", + "\n", + "With the above definitions we are now ready to set up the algorithm for AdaBoost.\n", + "The basic idea is to set up weights which will be used to scale the correctly classified and the misclassified cases.\n", + "1. We start by initializing all weights to $w_i = 1/n$, with $i=0,1,2,\\dots n-1$. It is easy to see that we must have $\\sum_{i=0}^{n-1}w_i = 1$.\n", + "\n", + "2. We rewrite the misclassification error as" + ] + }, + { + "cell_type": "markdown", + "id": "ad288a70", + "metadata": { + "editable": true + }, + "source": [ + "$$\n", + "\\mathrm{\\overline{err}}_m=\\frac{\\sum_{i=0}^{n-1}w_i^m I(y_i\\ne G(x_i))}{\\sum_{i=0}^{n-1}w_i},\n", + "$$" + ] + }, + { + "cell_type": "markdown", + "id": "6f0def3e", + "metadata": { + "editable": true + }, + "source": [ + "1. Then we start looping over all attempts at classifying, namely we start an iterative process for $m=1:M$, where $M$ is the final number of classifications. Our given classifier could for example be a plain decision tree.\n", + "\n", + "a. Fit then a given classifier to the training set using the weights $w_i$.\n", + "\n", + "b. Compute then $\\mathrm{err}$ and figure out which events are classified properly and which are classified wrongly.\n", + "\n", + "c. Define a quantity $\\alpha_{m} = \\log{(1-\\mathrm{\\overline{err}}_m)/\\mathrm{\\overline{err}}_m}$\n", + "\n", + "d. Set the new weights to $w_i = w_i\\times \\exp{(\\alpha_m I(y_i\\ne G(x_i)}$.\n", + "\n", + "5. Compute the new classifier $G(x)= \\sum_{i=0}^{n-1}\\alpha_m I(y_i\\ne G(x_i)$.\n", + "\n", + "For the iterations with $m \\le 2$ the weights are modified\n", + "individually at each steps. The observations which were misclassified\n", + "at iteration $m-1$ have a weight which is larger than those which were\n", + "classified properly. As this proceeds, the observations which were\n", + "difficult to classifiy correctly are given a larger influence. Each\n", + "new classification step $m$ is then forced to concentrate on those\n", + "observations that are missed in the previous iterations." + ] + }, + { + "cell_type": "markdown", + "id": "cf8a89b9", + "metadata": { + "editable": true + }, + "source": [ + "## AdaBoost Examples\n", + "\n", + "Using **Scikit-Learn** it is easy to apply the adaptive boosting algorithm, as done here." ] }, { "cell_type": "code", "execution_count": 1, + "id": "fc01de3a", "metadata": { - "collapsed": false + "collapsed": false, + "editable": true }, "outputs": [], "source": [ - "%matplotlib inline\n", + "from sklearn.ensemble import AdaBoostClassifier\n", "\n", - "import numpy as np\n", - "import os\n", - "\n", - "np.random.seed(42)\n", - "\n", - "# To plot pretty figures\n", - "import matplotlib\n", - "import matplotlib.pyplot as plt\n", - "plt.rcParams['axes.labelsize'] = 14\n", - "plt.rcParams['xtick.labelsize'] = 12\n", - "plt.rcParams['ytick.labelsize'] = 12\n", - "\n", - "\n", - "from sklearn.svm import SVC\n", - "from sklearn import datasets\n", - "\n", - "\n", - "\n", - "X1D = np.linspace(-4, 4, 9).reshape(-1, 1)\n", - "X2D = np.c_[X1D, X1D**2]\n", - "y = np.array([0, 0, 1, 1, 1, 1, 1, 0, 0])\n", - "\n", - "plt.figure(figsize=(11, 4))\n", - "\n", - "plt.subplot(121)\n", - "plt.grid(True, which='both')\n", - "plt.axhline(y=0, color='k')\n", - "plt.plot(X1D[:, 0][y==0], np.zeros(4), \"bs\")\n", - "plt.plot(X1D[:, 0][y==1], np.zeros(5), \"g^\")\n", - "plt.gca().get_yaxis().set_ticks([])\n", - "plt.xlabel(r\"$x_1$\", fontsize=20)\n", - "plt.axis([-4.5, 4.5, -0.2, 0.2])\n", - "\n", - "plt.subplot(122)\n", - "plt.grid(True, which='both')\n", - "plt.axhline(y=0, color='k')\n", - "plt.axvline(x=0, color='k')\n", - "plt.plot(X2D[:, 0][y==0], X2D[:, 1][y==0], \"bs\")\n", - "plt.plot(X2D[:, 0][y==1], X2D[:, 1][y==1], \"g^\")\n", - "plt.xlabel(r\"$x_1$\", fontsize=20)\n", - "plt.ylabel(r\"$x_2$\", fontsize=20, rotation=0)\n", - "plt.gca().get_yaxis().set_ticks([0, 4, 8, 12, 16])\n", - "plt.plot([-4.5, 4.5], [6.5, 6.5], \"r--\", linewidth=3)\n", - "plt.axis([-4.5, 4.5, -1, 17])\n", - "plt.subplots_adjust(right=1)\n", + "ada_clf = AdaBoostClassifier(\n", + " DecisionTreeClassifier(max_depth=2), n_estimators=200,\n", + " algorithm=\"SAMME.R\", learning_rate=0.01, random_state=42)\n", + "ada_clf.fit(X_train, y_train)\n", + "y_pred = ada_clf.predict(X_test)\n", + "skplt.metrics.plot_confusion_matrix(y_test, y_pred, normalize=True)\n", + "plt.show()\n", + "y_probas = ada_clf.predict_proba(X_test)\n", + "skplt.metrics.plot_roc(y_test, y_probas)\n", + "plt.show()\n", + "skplt.metrics.plot_cumulative_gain(y_test, y_probas)\n", "plt.show()" ] }, { "cell_type": "markdown", - "metadata": {}, + "id": "269f1edd", + "metadata": { + "editable": true + }, "source": [ - "## The equations\n", - "\n", - "Suppose we define a polynomial transformation of degree two only. We define a vector $\\boldsymbol{x}_i=[x_i,y_i]$ and have" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "$$\n", - "\\boldsymbol{z}_i^T = \\phi(\\boldsymbol{x}_i)^T =\\left[1, \\sqrt{2}x_i, \\sqrt{2}y_i, x_i^2, y_i^2, \\sqrt{2}x_iy_i\\right].\n", - "$$" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "With our new basis, the equations we solved earlier are basically the same, that is we have now (without the slack option for simplicity)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "$$\n", - "{\\cal L}=\\sum_i\\lambda_i-\\frac{1}{2}\\sum_{ij}^n\\lambda_i\\lambda_jy_iy_j\\boldsymbol{z}_i^T\\boldsymbol{z}_j,\n", - "$$" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "subject to the constraints $\\lambda_i\\geq 0$, $\\sum_i\\lambda_iy_i=0$, and for the support vectors" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "$$\n", - "y_i(\\boldsymbol{z}_i^T\\boldsymbol{w}+b)= 1 \\hspace{0.1cm}\\forall i,\n", - "$$" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "from which we also find $b$.\n", - "To compute $\\boldsymbol{z}_i^T\\boldsymbol{z}_j$ we define the kernel $K(\\boldsymbol{x}_i,\\boldsymbol{x}_j)$ as" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "$$\n", - "K(\\boldsymbol{x}_i,\\boldsymbol{x}_j)=\\boldsymbol{z}_i^T\\boldsymbol{z}_j= \\phi(\\boldsymbol{x}_i)^T\\phi(\\boldsymbol{x}_j).\n", - "$$" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "For the above example, the kernel reads" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "$$\n", - "K(\\boldsymbol{x}_i,\\boldsymbol{x}_j)=[1, \\sqrt{2}x_i, \\sqrt{2}y_i, x_i^2, y_i^2, \\sqrt{2}x_iy_i]\\begin{bmatrix} 1\\\\ \\sqrt{2}x_j \\\\ \\sqrt{2}y_j \\\\ x_j^2\\\\ y_i^2\\\\ \\sqrt{2}x_jy_j \\end{bmatrix}=1+2x_ix_j+2y_iy_j+(x_ix_j)^2+(y_iy_j)^2+2x_ix_jy_iy_j.\n", - "$$" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "We note that this dot product can be rewritten as" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "$$\n", - "K(\\boldsymbol{x}_i,\\boldsymbol{x}_j)=[1+\\boldsymbol{x}_i^T\\boldsymbol{x}_j]^d,\n", - "$$" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "where $d=2$ in our case and $\\boldsymbol{x}_i=[x_i,y_i]$ and $\\boldsymbol{x}_j=[x_j,y_j]$.\n", - "To compute the last equation is however inefficient from a computational stand.\n", - "Instead of computing the last equation for the kernel, we simply compute\n", - "the dot product $(\\boldsymbol{x}_i^T\\boldsymbol{x}_j)^2$.\n", - "This leads to the so-called\n", - "kernel trick.\n", - "\n", - "## The problem to solve\n", - "\n", - "Using our definition of the kernel, we can rewrite again the Lagrangian" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "$$\n", - "{\\cal L}=\\sum_i\\lambda_i-\\frac{1}{2}\\sum_{ij}^n\\lambda_i\\lambda_jy_iy_j\\boldsymbol{z}_i^T\\boldsymbol{z}_j,\n", - "$$" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "subject to the constraints $\\lambda_i\\geq 0$, $\\sum_i\\lambda_iy_i=0$ in terms of a convex optimization problem" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "$$\n", - "\\frac{1}{2} \\boldsymbol{\\lambda}^T\\begin{bmatrix} y_1y_1K(\\boldsymbol{x}_1,\\boldsymbol{x}_1) & y_1y_2K(\\boldsymbol{x}_1,\\boldsymbol{x}_2) & \\dots & \\dots & y_1y_nK(\\boldsymbol{x}_1,\\boldsymbol{x}_n) \\\\\n", - "y_2y_1K(\\boldsymbol{x}_2,\\boldsymbol{x}_1) & y_2y_2(\\boldsymbol{x}_2,\\boldsymbol{x}_2) & \\dots & \\dots & y_1y_nK(\\boldsymbol{x}_2,\\boldsymbol{x}_n) \\\\\n", - "\\dots & \\dots & \\dots & \\dots & \\dots \\\\\n", - "\\dots & \\dots & \\dots & \\dots & \\dots \\\\\n", - "y_ny_1K(\\boldsymbol{x}_n,\\boldsymbol{x}_1) & y_ny_2K(\\boldsymbol{x}_n\\boldsymbol{x}_2) & \\dots & \\dots & y_ny_nK(\\boldsymbol{x}_n,\\boldsymbol{x}_n) \\\\\n", - "\\end{bmatrix}\\boldsymbol{\\lambda}-\\mathbb{1}\\boldsymbol{\\lambda},\n", - "$$" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "subject to $\\boldsymbol{y}^T\\boldsymbol{\\lambda}=0$. Here we defined the vectors $\\boldsymbol{\\lambda} =[\\lambda_1,\\lambda_2,\\dots,\\lambda_n]$ and \n", - "$\\boldsymbol{y}=[y_1,y_2,\\dots,y_n]$. \n", - "If we add the slack constants this leads to the additional constraint $0\\leq \\lambda_i \\leq C$.\n", - "\n", - "## Tailoring the equations to the usage of CVXOPT\n", - "\n", - "We can rewrite this (see the solutions below) in terms of a convex optimization problem of the type" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "$$\n", - "\\begin{align*}\n", - " &\\mathrm{min}_{\\lambda}\\hspace{0.2cm} \\frac{1}{2}\\boldsymbol{\\lambda}^T\\boldsymbol{P}\\boldsymbol{\\lambda}+\\boldsymbol{q}^T\\boldsymbol{\\lambda},\\\\ \\nonumber\n", - " &\\mathrm{s.t} \\hspace{0.2cm} \\boldsymbol{G}\\boldsymbol{\\lambda} \\preceq \\boldsymbol{h} \\hspace{0.2cm} \\wedge \\boldsymbol{A}\\boldsymbol{\\lambda}=f.\n", - "\\end{align*}\n", - "$$" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Below we discuss how to solve these equations. Here we note that the matrix $\\boldsymbol{P}$ has matrix elements $p_{ij}=y_iy_jK(\\boldsymbol{x}_i,\\boldsymbol{x}_j)$.\n", - "Given a kernel $K$ and the targets $y_i$ this matrix is easy to set up. The constraint $\\boldsymbol{y}^T\\boldsymbol{\\lambda}=0$ leads to $f=0$ and $\\boldsymbol{A}=\\boldsymbol{y}$. How to set up the matrix $\\boldsymbol{G}$ is discussed later. Here note that the inequalities $0\\leq \\lambda_i \\leq C$ can be split up into\n", - "$0\\leq \\lambda_i$ and $\\lambda_i \\leq C$. These two inequalities define then the matrix $\\boldsymbol{G}$ and the vector $\\boldsymbol{h}$.\n", - "\n", - "\n", - "## Different kernels and Mercer's theorem\n", - "\n", - "There are several popular kernels being used. These are\n", - "1. Linear: $K(\\boldsymbol{x},\\boldsymbol{y})=\\boldsymbol{x}^T\\boldsymbol{y}$,\n", - "\n", - "2. Polynomial: $K(\\boldsymbol{x},\\boldsymbol{y})=(\\boldsymbol{x}^T\\boldsymbol{y}+\\gamma)^d$,\n", - "\n", - "3. Gaussian Radial Basis Function: $K(\\boldsymbol{x},\\boldsymbol{y})=\\exp{\\left(-\\gamma\\vert\\vert\\boldsymbol{x}-\\boldsymbol{y}\\vert\\vert^2\\right)}$,\n", - "\n", - "4. Tanh: $K(\\boldsymbol{x},\\boldsymbol{y})=\\tanh{(\\boldsymbol{x}^T\\boldsymbol{y}+\\gamma)}$,\n", - "\n", - "and many other ones.\n", - "\n", - "An important theorem for us is [Mercer's\n", - "theorem](https://en.wikipedia.org/wiki/Mercer%27s_theorem). The\n", - "theorem states that if a kernel function $K$ is symmetric, continuous\n", - "and leads to a positive semi-definite matrix $\\boldsymbol{P}$ then there\n", - "exists a function $\\phi$ that maps $\\boldsymbol{x}_i$ and $\\boldsymbol{x}_j$ into\n", - "another space (possibly with much higher dimensions) such that" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "$$\n", - "K(\\boldsymbol{x}_i,\\boldsymbol{x}_j)=\\phi(\\boldsymbol{x}_i)^T\\phi(\\boldsymbol{x}_j).\n", - "$$" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "So you can use $K$ as a kernel since you know $\\phi$ exists, even if\n", - "you don’t know what $\\phi$ is. \n", - "\n", - "Note that some frequently used kernels (such as the Sigmoid kernel)\n", - "don’t respect all of Mercer’s conditions, yet they generally work well\n", - "in practice.\n", - "\n", - "\n", - "## The moons example ([Adapted from Geron, chapter 5](https://www.oreilly.com/library/view/hands-on-machine-learning/9781492032632/))" + "## Making an ADAboost code yourself" ] }, { "cell_type": "code", "execution_count": 2, + "id": "bef1d223", "metadata": { - "collapsed": false + "collapsed": false, + "editable": true }, "outputs": [], "source": [ - "from __future__ import division, print_function, unicode_literals\n", - "\n", "import numpy as np\n", - "np.random.seed(42)\n", "\n", - "import matplotlib\n", - "import matplotlib.pyplot as plt\n", - "plt.rcParams['axes.labelsize'] = 14\n", - "plt.rcParams['xtick.labelsize'] = 12\n", - "plt.rcParams['ytick.labelsize'] = 12\n", + "class DecisionStump:\n", + " def fit(self, X, y, weights):\n", + " m, n = X.shape\n", + " self.alpha = 0\n", + " self.threshold = None\n", + " self.polarity = 1\n", "\n", + " min_error = float('inf')\n", "\n", - "from sklearn.svm import SVC\n", - "from sklearn import datasets\n", + " for feature in range(n):\n", + " feature_values = np.unique(X[:, feature])\n", "\n", + " for threshold in feature_values:\n", + " for polarity in [1, -1]:\n", + " predictions = np.ones(m)\n", + " predictions[X[:, feature] < threshold] = -1\n", + " predictions *= polarity\n", "\n", + " error = sum(weights[predictions != y])\n", "\n", - "from sklearn.pipeline import Pipeline\n", - "from sklearn.preprocessing import StandardScaler\n", - "from sklearn.svm import LinearSVC\n", + " if error < min_error:\n", + " min_error = error\n", + " self.alpha = 0.5 * np.log((1 - error) / (error + 1e-10))\n", + " self.threshold = threshold\n", + " self.feature_index = feature\n", + " self.polarity = polarity\n", "\n", + " def predict(self, X):\n", + " m = X.shape[0]\n", + " predictions = np.ones(m)\n", + " if self.polarity == 1:\n", + " predictions[X[:, self.feature_index] < self.threshold] = -1\n", + " else:\n", + " predictions[X[:, self.feature_index] >= self.threshold] = -1\n", + " return predictions\n", "\n", - "from sklearn.datasets import make_moons\n", - "X, y = make_moons(n_samples=100, noise=0.15, random_state=42)\n", + "class AdaBoost:\n", + " def fit(self, X, y, n_estimators):\n", + " m = X.shape[0]\n", + " self.alphas = []\n", + " self.models = []\n", "\n", - "def plot_dataset(X, y, axes):\n", - " plt.plot(X[:, 0][y==0], X[:, 1][y==0], \"bs\")\n", - " plt.plot(X[:, 0][y==1], X[:, 1][y==1], \"g^\")\n", - " plt.axis(axes)\n", - " plt.grid(True, which='both')\n", - " plt.xlabel(r\"$x_1$\", fontsize=20)\n", - " plt.ylabel(r\"$x_2$\", fontsize=20, rotation=0)\n", + " weights = np.ones(m) / m\n", "\n", - "plot_dataset(X, y, [-1.5, 2.5, -1, 1.5])\n", - "plt.show()\n", + " for _ in range(n_estimators):\n", + " stump = DecisionStump()\n", + " stump.fit(X, y, weights)\n", + " predictions = stump.predict(X)\n", "\n", - "from sklearn.datasets import make_moons\n", - "from sklearn.pipeline import Pipeline\n", - "from sklearn.preprocessing import PolynomialFeatures\n", + " error = sum(weights[predictions != y])\n", + " if error == 0:\n", + " break\n", "\n", - "polynomial_svm_clf = Pipeline([\n", - " (\"poly_features\", PolynomialFeatures(degree=3)),\n", - " (\"scaler\", StandardScaler()),\n", - " (\"svm_clf\", LinearSVC(C=10, loss=\"hinge\", random_state=42))\n", - " ])\n", + " self.models.append(stump)\n", + " self.alphas.append(stump.alpha)\n", "\n", - "polynomial_svm_clf.fit(X, y)\n", + " weights *= np.exp(-stump.alpha * y * predictions)\n", + " weights /= np.sum(weights)\n", "\n", - "def plot_predictions(clf, axes):\n", - " x0s = np.linspace(axes[0], axes[1], 100)\n", - " x1s = np.linspace(axes[2], axes[3], 100)\n", - " x0, x1 = np.meshgrid(x0s, x1s)\n", - " X = np.c_[x0.ravel(), x1.ravel()]\n", - " y_pred = clf.predict(X).reshape(x0.shape)\n", - " y_decision = clf.decision_function(X).reshape(x0.shape)\n", - " plt.contourf(x0, x1, y_pred, cmap=plt.cm.brg, alpha=0.2)\n", - " plt.contourf(x0, x1, y_decision, cmap=plt.cm.brg, alpha=0.1)\n", + " def predict(self, X):\n", + " final_predictions = np.zeros(X.shape[0])\n", + " for alpha, model in zip(self.alphas, self.models):\n", + " final_predictions += alpha * model.predict(X)\n", + " return np.sign(final_predictions)\n", "\n", - "plot_predictions(polynomial_svm_clf, [-1.5, 2.5, -1, 1.5])\n", - "plot_dataset(X, y, [-1.5, 2.5, -1, 1.5])\n", + "# Example dataset (X, y)\n", + "X = np.array([[1], [2], [3], [4], [5], [6], [7], [8], [9], [10]])\n", + "y = np.array([-1, -1, -1, -1, 1, 1, 1, 1, 1, 1]) # Labels must be -1 or 1\n", "\n", - "plt.show()\n", + "# Train AdaBoost\n", + "ada = AdaBoost()\n", + "ada.fit(X, y, n_estimators=10)\n", "\n", - "\n", - "from sklearn.svm import SVC\n", - "\n", - "poly_kernel_svm_clf = Pipeline([\n", - " (\"scaler\", StandardScaler()),\n", - " (\"svm_clf\", SVC(kernel=\"poly\", degree=3, coef0=1, C=5))\n", - " ])\n", - "poly_kernel_svm_clf.fit(X, y)\n", - "\n", - "poly100_kernel_svm_clf = Pipeline([\n", - " (\"scaler\", StandardScaler()),\n", - " (\"svm_clf\", SVC(kernel=\"poly\", degree=10, coef0=100, C=5))\n", - " ])\n", - "poly100_kernel_svm_clf.fit(X, y)\n", - "\n", - "plt.figure(figsize=(11, 4))\n", - "\n", - "plt.subplot(121)\n", - "plot_predictions(poly_kernel_svm_clf, [-1.5, 2.5, -1, 1.5])\n", - "plot_dataset(X, y, [-1.5, 2.5, -1, 1.5])\n", - "plt.title(r\"$d=3, r=1, C=5$\", fontsize=18)\n", - "\n", - "plt.subplot(122)\n", - "plot_predictions(poly100_kernel_svm_clf, [-1.5, 2.5, -1, 1.5])\n", - "plot_dataset(X, y, [-1.5, 2.5, -1, 1.5])\n", - "plt.title(r\"$d=10, r=100, C=5$\", fontsize=18)\n", - "\n", - "plt.show()\n", - "\n", - "def gaussian_rbf(x, landmark, gamma):\n", - " return np.exp(-gamma * np.linalg.norm(x - landmark, axis=1)**2)\n", - "\n", - "gamma = 0.3\n", - "\n", - "x1s = np.linspace(-4.5, 4.5, 200).reshape(-1, 1)\n", - "x2s = gaussian_rbf(x1s, -2, gamma)\n", - "x3s = gaussian_rbf(x1s, 1, gamma)\n", - "\n", - "XK = np.c_[gaussian_rbf(X1D, -2, gamma), gaussian_rbf(X1D, 1, gamma)]\n", - "yk = np.array([0, 0, 1, 1, 1, 1, 1, 0, 0])\n", - "\n", - "plt.figure(figsize=(11, 4))\n", - "\n", - "plt.subplot(121)\n", - "plt.grid(True, which='both')\n", - "plt.axhline(y=0, color='k')\n", - "plt.scatter(x=[-2, 1], y=[0, 0], s=150, alpha=0.5, c=\"red\")\n", - "plt.plot(X1D[:, 0][yk==0], np.zeros(4), \"bs\")\n", - "plt.plot(X1D[:, 0][yk==1], np.zeros(5), \"g^\")\n", - "plt.plot(x1s, x2s, \"g--\")\n", - "plt.plot(x1s, x3s, \"b:\")\n", - "plt.gca().get_yaxis().set_ticks([0, 0.25, 0.5, 0.75, 1])\n", - "plt.xlabel(r\"$x_1$\", fontsize=20)\n", - "plt.ylabel(r\"Similarity\", fontsize=14)\n", - "plt.annotate(r'$\\mathbf{x}$',\n", - " xy=(X1D[3, 0], 0),\n", - " xytext=(-0.5, 0.20),\n", - " ha=\"center\",\n", - " arrowprops=dict(facecolor='black', shrink=0.1),\n", - " fontsize=18,\n", - " )\n", - "plt.text(-2, 0.9, \"$x_2$\", ha=\"center\", fontsize=20)\n", - "plt.text(1, 0.9, \"$x_3$\", ha=\"center\", fontsize=20)\n", - "plt.axis([-4.5, 4.5, -0.1, 1.1])\n", - "\n", - "plt.subplot(122)\n", - "plt.grid(True, which='both')\n", - "plt.axhline(y=0, color='k')\n", - "plt.axvline(x=0, color='k')\n", - "plt.plot(XK[:, 0][yk==0], XK[:, 1][yk==0], \"bs\")\n", - "plt.plot(XK[:, 0][yk==1], XK[:, 1][yk==1], \"g^\")\n", - "plt.xlabel(r\"$x_2$\", fontsize=20)\n", - "plt.ylabel(r\"$x_3$ \", fontsize=20, rotation=0)\n", - "plt.annotate(r'$\\phi\\left(\\mathbf{x}\\right)$',\n", - " xy=(XK[3, 0], XK[3, 1]),\n", - " xytext=(0.65, 0.50),\n", - " ha=\"center\",\n", - " arrowprops=dict(facecolor='black', shrink=0.1),\n", - " fontsize=18,\n", - " )\n", - "plt.plot([-0.1, 1.1], [0.57, -0.1], \"r--\", linewidth=3)\n", - "plt.axis([-0.1, 1.1, -0.1, 1.1])\n", - " \n", - "plt.subplots_adjust(right=1)\n", - "\n", - "plt.show()\n", - "\n", - "\n", - "x1_example = X1D[3, 0]\n", - "for landmark in (-2, 1):\n", - " k = gaussian_rbf(np.array([[x1_example]]), np.array([[landmark]]), gamma)\n", - " print(\"Phi({}, {}) = {}\".format(x1_example, landmark, k))\n", - "\n", - "rbf_kernel_svm_clf = Pipeline([\n", - " (\"scaler\", StandardScaler()),\n", - " (\"svm_clf\", SVC(kernel=\"rbf\", gamma=5, C=0.001))\n", - " ])\n", - "rbf_kernel_svm_clf.fit(X, y)\n", - "\n", - "\n", - "from sklearn.svm import SVC\n", - "\n", - "gamma1, gamma2 = 0.1, 5\n", - "C1, C2 = 0.001, 1000\n", - "hyperparams = (gamma1, C1), (gamma1, C2), (gamma2, C1), (gamma2, C2)\n", - "\n", - "svm_clfs = []\n", - "for gamma, C in hyperparams:\n", - " rbf_kernel_svm_clf = Pipeline([\n", - " (\"scaler\", StandardScaler()),\n", - " (\"svm_clf\", SVC(kernel=\"rbf\", gamma=gamma, C=C))\n", - " ])\n", - " rbf_kernel_svm_clf.fit(X, y)\n", - " svm_clfs.append(rbf_kernel_svm_clf)\n", - "\n", - "plt.figure(figsize=(11, 7))\n", - "\n", - "for i, svm_clf in enumerate(svm_clfs):\n", - " plt.subplot(221 + i)\n", - " plot_predictions(svm_clf, [-1.5, 2.5, -1, 1.5])\n", - " plot_dataset(X, y, [-1.5, 2.5, -1, 1.5])\n", - " gamma, C = hyperparams[i]\n", - " plt.title(r\"$\\gamma = {}, C = {}$\".format(gamma, C), fontsize=16)\n", - "\n", - "plt.show()" + "# Predictions\n", + "predictions = ada.predict(X)\n", + "print(\"Predictions:\", predictions)" ] }, { "cell_type": "markdown", - "metadata": {}, + "id": "9338a775", + "metadata": { + "editable": true + }, "source": [ - "## Mathematical optimization of convex functions\n", + "## Gradient boosting: Basics with Steepest Descent/Functional Gradient Descent\n", "\n", - "A mathematical (quadratic) optimization problem, or just optimization problem, has the form" + "Gradient boosting is again a similar technique to Adaptive boosting,\n", + "it combines so-called weak classifiers or regressors into a strong\n", + "method via a series of iterations.\n", + "\n", + "In order to understand the method, let us illustrate its basics by\n", + "bringing back the essential steps in linear regression, where our cost\n", + "function was the least squares function." ] }, { "cell_type": "markdown", - "metadata": {}, + "id": "6fc86b75", + "metadata": { + "editable": true + }, + "source": [ + "## The Squared-Error again! Steepest Descent\n", + "\n", + "We start again with our cost function ${\\cal C}(\\boldsymbol{y}m\\boldsymbol{f})=\\sum_{i=0}^{n-1}{\\cal L}(y_i, f(x_i))$ where we want to minimize\n", + "This means that for every iteration, we need to optimize" + ] + }, + { + "cell_type": "markdown", + "id": "f4fb65a8", + "metadata": { + "editable": true + }, "source": [ "$$\n", - "\\begin{align*}\n", - " &\\mathrm{min}_{\\lambda}\\hspace{0.2cm} \\frac{1}{2}\\boldsymbol{\\lambda}^T\\boldsymbol{P}\\boldsymbol{\\lambda}+\\boldsymbol{q}^T\\boldsymbol{\\lambda},\\\\ \\nonumber\n", - " &\\mathrm{subject\\hspace{0.1cm}to} \\hspace{0.2cm} \\boldsymbol{G}\\boldsymbol{\\lambda} \\preceq \\boldsymbol{h} \\wedge \\boldsymbol{A}\\boldsymbol{\\lambda}=f.\n", - "\\end{align*}\n", + "(\\hat{\\boldsymbol{f}}) = \\mathrm{argmin}_{\\boldsymbol{f}}\\hspace{0.1cm} \\sum_{i=0}^{n-1}(y_i-f(x_i))^2.\n", "$$" ] }, { "cell_type": "markdown", - "metadata": {}, + "id": "46ea5f88", + "metadata": { + "editable": true + }, "source": [ - "subject to some constraints for say a selected set $i=1,2,\\dots, n$.\n", - "In our case we are optimizing with respect to the Lagrangian multipliers $\\lambda_i$, and the\n", - "vector $\\boldsymbol{\\lambda}=[\\lambda_1, \\lambda_2,\\dots, \\lambda_n]$ is the optimization variable we are dealing with.\n", + "We define a real function $h_m(x)$ that defines our final function $f_M(x)$ as" + ] + }, + { + "cell_type": "markdown", + "id": "0fa310b0", + "metadata": { + "editable": true + }, + "source": [ + "$$\n", + "f_M(x) = \\sum_{m=0}^M h_m(x).\n", + "$$" + ] + }, + { + "cell_type": "markdown", + "id": "3638bac4", + "metadata": { + "editable": true + }, + "source": [ + "In the steepest decent approach we approximate $h_m(x) = -\\rho_m g_m(x)$, where $\\rho_m$ is a scalar and $g_m(x)$ the gradient defined as" + ] + }, + { + "cell_type": "markdown", + "id": "40c8295f", + "metadata": { + "editable": true + }, + "source": [ + "$$\n", + "g_m(x_i) = \\left[ \\frac{\\partial {\\cal L}(y_i, f(x_i))}{\\partial f(x_i)}\\right]_{f(x_i)=f_{m-1}(x_i)}.\n", + "$$" + ] + }, + { + "cell_type": "markdown", + "id": "116dbbbd", + "metadata": { + "editable": true + }, + "source": [ + "With the new gradient we can update $f_m(x) = f_{m-1}(x) -\\rho_m g_m(x)$. Using the above squared-error function we see that\n", + "the gradient is $g_m(x_i) = -2(y_i-f(x_i))$.\n", "\n", - "In our case we are particularly interested in a class of optimization problems called convex optmization problems. \n", - "In our discussion on gradient descent methods we discussed at length the definition of a convex function. \n", + "Choosing $f_0(x)=0$ we obtain $g_m(x) = -2y_i$ and inserting this into the minimization problem for the cost function we have" + ] + }, + { + "cell_type": "markdown", + "id": "1873e19d", + "metadata": { + "editable": true + }, + "source": [ + "$$\n", + "(\\rho_1) = \\mathrm{argmin}_{\\rho}\\hspace{0.1cm} \\sum_{i=0}^{n-1}(y_i+2\\rho y_i)^2.\n", + "$$" + ] + }, + { + "cell_type": "markdown", + "id": "a488e80c", + "metadata": { + "editable": true + }, + "source": [ + "## Steepest Descent Example\n", "\n", - "Convex optimization problems play a central role in applied mathematics and we recommend strongly [Boyd and Vandenberghe's text on the topics](http://web.stanford.edu/~boyd/cvxbook/).\n", + "Optimizing with respect to $\\rho$ we obtain (taking the derivative) that $\\rho_1 = -1/2$. We have then that" + ] + }, + { + "cell_type": "markdown", + "id": "754c7e53", + "metadata": { + "editable": true + }, + "source": [ + "$$\n", + "f_1(x) = f_{0}(x) -\\rho_1 g_1(x)=-y_i.\n", + "$$" + ] + }, + { + "cell_type": "markdown", + "id": "040f985c", + "metadata": { + "editable": true + }, + "source": [ + "We can then proceed and compute" + ] + }, + { + "cell_type": "markdown", + "id": "2095bf76", + "metadata": { + "editable": true + }, + "source": [ + "$$\n", + "g_2(x_i) = \\left[ \\frac{\\partial {\\cal L}(y_i, f(x_i))}{\\partial f(x_i)}\\right]_{f(x_i)=f_{1}(x_i)=y_i}=-4y_i,\n", + "$$" + ] + }, + { + "cell_type": "markdown", + "id": "eb32c8f7", + "metadata": { + "editable": true + }, + "source": [ + "and find a new value for $\\rho_2=-1/2$ and continue till we have reached $m=M$. We can modify the steepest descent method, or steepest boosting, by introducing what is called **gradient boosting**." + ] + }, + { + "cell_type": "markdown", + "id": "1dffe706", + "metadata": { + "editable": true + }, + "source": [ + "## Gradient Boosting, algorithm\n", "\n", + "Steepest descent is however not much used, since it only optimizes $f$ at a fixed set of $n$ points,\n", + "so we do not learn a function that can generalize. However, we can modify the algorithm by\n", + "fitting a weak learner to approximate the negative gradient signal. \n", "\n", + "Suppose we have a cost function $C(f)=\\sum_{i=0}^{n-1}L(y_i, f(x_i))$ where $y_i$ is our target and $f(x_i)$ the function which is meant to model $y_i$. The above cost function could be our standard squared-error function" + ] + }, + { + "cell_type": "markdown", + "id": "a773c034", + "metadata": { + "editable": true + }, + "source": [ + "$$\n", + "C(\\boldsymbol{y},\\boldsymbol{f})=\\sum_{i=0}^{n-1}(y_i-f(x_i))^2.\n", + "$$" + ] + }, + { + "cell_type": "markdown", + "id": "c5309d69", + "metadata": { + "editable": true + }, + "source": [ + "The way we proceed in an iterative fashion is to\n", + "1. Initialize our estimate $f_0(x)$.\n", "\n", - "## How do we solve these problems?\n", + "2. For $m=1:M$, we\n", "\n", - "If we use Python as programming language and wish to venture beyond\n", - "**scikit-learn**, **tensorflow** and similar software which makes our\n", - "lives so much easier, we need to dive into the wonderful world of\n", - "quadratic programming. We can, if we wish, solve the minimization\n", - "problem using standard gradient methods or conjugate gradient\n", - "methods. However, these methods tend to exhibit a rather slow\n", - "converge. So, welcome to the promised land of quadratic programming.\n", + "a. compute the negative gradient vector $\\boldsymbol{u}_m = -\\partial C(\\boldsymbol{y},\\boldsymbol{f})/\\partial \\boldsymbol{f}(x)$ at $f(x) = f_{m-1}(x)$;\n", "\n", - "The functions we need are contained in the quadratic programming package **CVXOPT** and we need to import it together with **numpy** as" + "b. fit the so-called base-learner to the negative gradient $h_m(u_m,x)$;\n", + "\n", + "c. update the estimate $f_m(x) = f_{m-1}(x)+h_m(u_m,x)$;\n", + "\n", + "4. The final estimate is then $f_M(x) = \\sum_{m=1}^M h_m(u_m,x)$." + ] + }, + { + "cell_type": "markdown", + "id": "ce123557", + "metadata": { + "editable": true + }, + "source": [ + "## Gradient Boosting, Examples of Regression" ] }, { "cell_type": "code", "execution_count": 3, + "id": "83f66b3c", "metadata": { - "collapsed": false + "collapsed": false, + "editable": true }, "outputs": [], "source": [ - "import numpy\n", - "import cvxopt" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "This will make our life much easier. You don't need to write your own optimizer.\n", + "%matplotlib inline\n", "\n", + "import matplotlib.pyplot as plt\n", + "import numpy as np\n", + "from sklearn.model_selection import train_test_split\n", + "from sklearn.ensemble import GradientBoostingRegressor\n", + "import scikitplot as skplt\n", + "from sklearn.metrics import mean_squared_error\n", "\n", - "## A simple example\n", + "n = 100\n", + "maxdegree = 6\n", "\n", - "We remind ourselves about the general problem we want to solve" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "$$\n", - "\\begin{align*}\n", - " &\\mathrm{min}_{x}\\hspace{0.2cm} \\frac{1}{2}\\boldsymbol{x}^T\\boldsymbol{P}\\boldsymbol{x}+\\boldsymbol{q}^T\\boldsymbol{x},\\\\ \\nonumber\n", - " &\\mathrm{s.t.} \\hspace{0.2cm} \\boldsymbol{G}\\boldsymbol{x} \\preceq \\boldsymbol{h} \\wedge \\boldsymbol{A}\\boldsymbol{x}=f.\n", - "\\end{align*}\n", - "$$" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "**Note**: we use **s.t.** for subject to. \n", + "# Make data set.\n", + "x = np.linspace(-3, 3, n).reshape(-1, 1)\n", + "y = np.exp(-x**2) + 1.5 * np.exp(-(x-2)**2)+ np.random.normal(0, 0.1, x.shape)\n", "\n", - "Let us show how to perform the optmization using a simple case. Assume we want to optimize the following problem" + "error = np.zeros(maxdegree)\n", + "bias = np.zeros(maxdegree)\n", + "variance = np.zeros(maxdegree)\n", + "polydegree = np.zeros(maxdegree)\n", + "X_train, X_test, y_train, y_test = train_test_split(x, y, test_size=0.2)\n", + "\n", + "for degree in range(1,maxdegree):\n", + " model = GradientBoostingRegressor(max_depth=degree, n_estimators=100, learning_rate=1.0) \n", + " model.fit(X_train,y_train)\n", + " y_pred = model.predict(X_test)\n", + " polydegree[degree] = degree\n", + " error[degree] = np.mean( np.mean((y_test - y_pred)**2) )\n", + " bias[degree] = np.mean( (y_test - np.mean(y_pred))**2 )\n", + " variance[degree] = np.mean( np.var(y_pred) )\n", + " print('Max depth:', degree)\n", + " print('Error:', error[degree])\n", + " print('Bias^2:', bias[degree])\n", + " print('Var:', variance[degree])\n", + " print('{} >= {} + {} = {}'.format(error[degree], bias[degree], variance[degree], bias[degree]+variance[degree]))\n", + "\n", + "plt.xlim(1,maxdegree-1)\n", + "plt.plot(polydegree, error, label='Error')\n", + "plt.plot(polydegree, bias, label='bias')\n", + "plt.plot(polydegree, variance, label='Variance')\n", + "plt.legend()\n", + "save_fig(\"gdregression\")\n", + "plt.show()" ] }, { "cell_type": "markdown", - "metadata": {}, + "id": "1a395403", + "metadata": { + "editable": true + }, "source": [ - "$$\n", - "\\begin{align*}\n", - " &\\mathrm{min}_{x}\\hspace{0.2cm} \\frac{1}{2}x^2+5x+3y \\\\ \\nonumber\n", - " &\\mathrm{s.t.} \\\\ \\nonumber\n", - " &x, y \\geq 0 \\\\ \\nonumber\n", - " &x+3y \\geq 15 \\\\ \\nonumber\n", - " &2x+5y \\leq 100 \\\\ \\nonumber\n", - " &3x+4y \\leq 80. \\\\ \\nonumber\n", - "\\end{align*}\n", - "$$" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "The minimization problem can be rewritten in terms of vectors and matrices as (with $x$ and $y$ being the unknowns)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "$$\n", - "\\frac{1}{2}\\begin{bmatrix} x\\\\ y \\end{bmatrix}^T \\begin{bmatrix} 1 & 0\\\\ 0 & 0 \\end{bmatrix} \\begin{bmatrix} x \\\\ y \\end{bmatrix} + \\begin{bmatrix}5\\\\ 3 \\end{bmatrix}^T \\begin{bmatrix}x \\\\ y \\end{bmatrix}.\n", - "$$" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Similarly, we can now set up the inequalities (we need to change $\\geq$ to $\\leq$ by multiplying with $-1$ on bot sides) as the following matrix-vector equation" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "$$\n", - "\\begin{bmatrix} -1 & 0 \\\\ 0 & -1 \\\\ -1 & -3 \\\\ 2 & 5 \\\\ 3 & 4\\end{bmatrix}\\begin{bmatrix} x \\\\ y\\end{bmatrix} \\preceq \\begin{bmatrix}0 \\\\ 0\\\\ -15 \\\\ 100 \\\\ 80\\end{bmatrix}.\n", - "$$" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "We have collapsed all the inequalities into a single matrix $\\boldsymbol{G}$. We see also that our matrix" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "$$\n", - "\\boldsymbol{P} =\\begin{bmatrix} 1 & 0\\\\ 0 & 0 \\end{bmatrix}\n", - "$$" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "is clearly positive semi-definite (all eigenvalues larger or equal zero). \n", - "Finally, the vector $\\boldsymbol{h}$ is defined as" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "$$\n", - "\\boldsymbol{h} = \\begin{bmatrix}0 \\\\ 0\\\\ -15 \\\\ 100 \\\\ 80\\end{bmatrix}.\n", - "$$" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Since we don't have any equalities the matrix $\\boldsymbol{A}$ is set to zero\n", - "The following code solves the equations for us" + "## Gradient Boosting, Classification Example" ] }, { "cell_type": "code", "execution_count": 4, + "id": "09b1de9c", "metadata": { - "collapsed": false + "collapsed": false, + "editable": true }, "outputs": [], "source": [ - "# Import the necessary packages\n", - "import numpy\n", - "from cvxopt import matrix\n", - "from cvxopt import solvers\n", - "P = matrix(numpy.diag([1,0]), tc='d')\n", - "q = matrix(numpy.array([3,4]), tc='d')\n", - "G = matrix(numpy.array([[-1,0],[0,-1],[-1,-3],[2,5],[3,4]]), tc='d')\n", - "h = matrix(numpy.array([0,0,-15,100,80]), tc='d')\n", - "# Construct the QP, invoke solver\n", - "sol = solvers.qp(P,q,G,h)\n", - "# Extract optimal value and solution\n", - "sol['x'] \n", - "sol['primal objective']" + "import matplotlib.pyplot as plt\n", + "import numpy as np\n", + "from sklearn.model_selection import train_test_split \n", + "from sklearn.datasets import load_breast_cancer\n", + "import scikitplot as skplt\n", + "from sklearn.ensemble import GradientBoostingClassifier\n", + "from sklearn.model_selection import cross_validate\n", + "\n", + "# Load the data\n", + "cancer = load_breast_cancer()\n", + "\n", + "X_train, X_test, y_train, y_test = train_test_split(cancer.data,cancer.target,random_state=0)\n", + "print(X_train.shape)\n", + "print(X_test.shape)\n", + "#now scale the data\n", + "from sklearn.preprocessing import StandardScaler\n", + "scaler = StandardScaler()\n", + "scaler.fit(X_train)\n", + "X_train_scaled = scaler.transform(X_train)\n", + "X_test_scaled = scaler.transform(X_test)\n", + "\n", + "gd_clf = GradientBoostingClassifier(max_depth=3, n_estimators=100, learning_rate=1.0) \n", + "gd_clf.fit(X_train_scaled, y_train)\n", + "#Cross validation\n", + "accuracy = cross_validate(gd_clf,X_test_scaled,y_test,cv=10)['test_score']\n", + "print(accuracy)\n", + "print(\"Test set accuracy with Gradient boosting and scaled data: {:.2f}\".format(gd_clf.score(X_test_scaled,y_test)))\n", + "\n", + "import scikitplot as skplt\n", + "y_pred = gd_clf.predict(X_test_scaled)\n", + "skplt.metrics.plot_confusion_matrix(y_test, y_pred, normalize=True)\n", + "save_fig(\"gdclassiffierconfusion\")\n", + "plt.show()\n", + "y_probas = gd_clf.predict_proba(X_test_scaled)\n", + "skplt.metrics.plot_roc(y_test, y_probas)\n", + "save_fig(\"gdclassiffierroc\")\n", + "plt.show()\n", + "skplt.metrics.plot_cumulative_gain(y_test, y_probas)\n", + "save_fig(\"gdclassiffiercgain\")\n", + "plt.show()" ] }, { "cell_type": "markdown", - "metadata": {}, + "id": "ef37fb74", + "metadata": { + "editable": true + }, "source": [ - "## Back to the more realistic cases\n", + "## XGBoost: Extreme Gradient Boosting\n", "\n", - "We are now ready to return to our setup of the optmization problem for a more realistic case. Introducing the **slack** parameter $C$ we have" + "[XGBoost](https://github.com/dmlc/xgboost) or Extreme Gradient\n", + "Boosting, is an optimized distributed gradient boosting library\n", + "designed to be highly efficient, flexible and portable. It implements\n", + "machine learning algorithms under the Gradient Boosting\n", + "framework. XGBoost provides a parallel tree boosting that solve many\n", + "data science problems in a fast and accurate way. See the [article by Chen and Guestrin](https://arxiv.org/abs/1603.02754).\n", + "\n", + "The authors design and build a highly scalable end-to-end tree\n", + "boosting system. It has a theoretically justified weighted quantile\n", + "sketch for efficient proposal calculation. It introduces a novel sparsity-aware algorithm for parallel tree learning and an effective cache-aware block structure for out-of-core tree learning.\n", + "\n", + "It is now the algorithm which wins essentially all ML competitions!!!" ] }, { "cell_type": "markdown", - "metadata": {}, + "id": "e35c17fd", + "metadata": { + "editable": true + }, "source": [ - "$$\n", - "\\frac{1}{2} \\boldsymbol{\\lambda}^T\\begin{bmatrix} y_1y_1K(\\boldsymbol{x}_1,\\boldsymbol{x}_1) & y_1y_2K(\\boldsymbol{x}_1,\\boldsymbol{x}_2) & \\dots & \\dots & y_1y_nK(\\boldsymbol{x}_1,\\boldsymbol{x}_n) \\\\\n", - "y_2y_1K(\\boldsymbol{x}_2,\\boldsymbol{x}_1) & y_2y_2K(\\boldsymbol{x}_2,\\boldsymbol{x}_2) & \\dots & \\dots & y_1y_nK(\\boldsymbol{x}_2,\\boldsymbol{x}_n) \\\\\n", - "\\dots & \\dots & \\dots & \\dots & \\dots \\\\\n", - "\\dots & \\dots & \\dots & \\dots & \\dots \\\\\n", - "y_ny_1K(\\boldsymbol{x}_n,\\boldsymbol{x}_1) & y_ny_2K(\\boldsymbol{x}_n\\boldsymbol{x}_2) & \\dots & \\dots & y_ny_nK(\\boldsymbol{x}_n,\\boldsymbol{x}_n) \\\\\n", - "\\end{bmatrix}\\boldsymbol{\\lambda}-\\mathbb{I}\\boldsymbol{\\lambda},\n", - "$$" + "## Regression Case" + ] + }, + { + "cell_type": "code", + "execution_count": 5, + "id": "c2ae030f", + "metadata": { + "collapsed": false, + "editable": true + }, + "outputs": [], + "source": [ + "import matplotlib.pyplot as plt\n", + "import numpy as np\n", + "from sklearn.model_selection import train_test_split\n", + "import xgboost as xgb\n", + "import scikitplot as skplt\n", + "from sklearn.metrics import mean_squared_error\n", + "\n", + "n = 100\n", + "maxdegree = 6\n", + "\n", + "# Make data set.\n", + "x = np.linspace(-3, 3, n).reshape(-1, 1)\n", + "y = np.exp(-x**2) + 1.5 * np.exp(-(x-2)**2)+ np.random.normal(0, 0.1, x.shape)\n", + "\n", + "error = np.zeros(maxdegree)\n", + "bias = np.zeros(maxdegree)\n", + "variance = np.zeros(maxdegree)\n", + "polydegree = np.zeros(maxdegree)\n", + "X_train, X_test, y_train, y_test = train_test_split(x, y, test_size=0.2)\n", + "\n", + "for degree in range(maxdegree):\n", + " model = xgb.XGBRegressor(objective ='reg:squarederror', colsaobjective ='reg:squarederror', colsample_bytree = 0.3, learning_rate = 0.1,max_depth = degree, alpha = 10, n_estimators = 200)\n", + "\n", + " model.fit(X_train,y_train)\n", + " y_pred = model.predict(X_test)\n", + " polydegree[degree] = degree\n", + " error[degree] = np.mean( np.mean((y_test - y_pred)**2) )\n", + " bias[degree] = np.mean( (y_test - np.mean(y_pred))**2 )\n", + " variance[degree] = np.mean( np.var(y_pred) )\n", + " print('Max depth:', degree)\n", + " print('Error:', error[degree])\n", + " print('Bias^2:', bias[degree])\n", + " print('Var:', variance[degree])\n", + " print('{} >= {} + {} = {}'.format(error[degree], bias[degree], variance[degree], bias[degree]+variance[degree]))\n", + "\n", + "plt.xlim(1,maxdegree-1)\n", + "plt.plot(polydegree, error, label='Error')\n", + "plt.plot(polydegree, bias, label='bias')\n", + "plt.plot(polydegree, variance, label='Variance')\n", + "plt.legend()\n", + "plt.show()" ] }, { "cell_type": "markdown", - "metadata": {}, + "id": "9fde7dfa", + "metadata": { + "editable": true + }, "source": [ - "subject to $\\boldsymbol{y}^T\\boldsymbol{\\lambda}=0$. Here we defined the vectors $\\boldsymbol{\\lambda} =[\\lambda_1,\\lambda_2,\\dots,\\lambda_n]$ and \n", - "$\\boldsymbol{y}=[y_1,y_2,\\dots,y_n]$. \n", - "With the slack constants this leads to the additional constraint $0\\leq \\lambda_i \\leq C$.\n", + "## Xgboost on the Cancer Data\n", + "\n", + "As you will see from the confusion matrix below, XGBoots does an excellent job on the Wisconsin cancer data and outperforms essentially all agorithms we have discussed till now." + ] + }, + { + "cell_type": "code", + "execution_count": 6, + "id": "8492209d", + "metadata": { + "collapsed": false, + "editable": true + }, + "outputs": [], + "source": [ + "\n", + "import matplotlib.pyplot as plt\n", + "import numpy as np\n", + "from sklearn.model_selection import train_test_split \n", + "from sklearn.datasets import load_breast_cancer\n", + "from sklearn.preprocessing import LabelEncoder\n", + "from sklearn.model_selection import cross_validate\n", + "import scikitplot as skplt\n", + "import xgboost as xgb\n", + "# Load the data\n", + "cancer = load_breast_cancer()\n", + "\n", + "X_train, X_test, y_train, y_test = train_test_split(cancer.data,cancer.target,random_state=0)\n", + "print(X_train.shape)\n", + "print(X_test.shape)\n", + "#now scale the data\n", + "from sklearn.preprocessing import StandardScaler\n", + "scaler = StandardScaler()\n", + "scaler.fit(X_train)\n", + "X_train_scaled = scaler.transform(X_train)\n", + "X_test_scaled = scaler.transform(X_test)\n", + "\n", + "xg_clf = xgb.XGBClassifier()\n", + "xg_clf.fit(X_train_scaled,y_train)\n", + "\n", + "y_test = xg_clf.predict(X_test_scaled)\n", + "\n", + "print(\"Test set accuracy with Gradient Boosting and scaled data: {:.2f}\".format(xg_clf.score(X_test_scaled,y_test)))\n", + "\n", + "import scikitplot as skplt\n", + "y_pred = xg_clf.predict(X_test_scaled)\n", + "skplt.metrics.plot_confusion_matrix(y_test, y_pred, normalize=True)\n", + "save_fig(\"xdclassiffierconfusion\")\n", + "plt.show()\n", + "y_probas = xg_clf.predict_proba(X_test_scaled)\n", + "skplt.metrics.plot_roc(y_test, y_probas)\n", + "save_fig(\"xdclassiffierroc\")\n", + "plt.show()\n", + "skplt.metrics.plot_cumulative_gain(y_test, y_probas)\n", + "save_fig(\"gdclassiffiercgain\")\n", + "plt.show()\n", "\n", "\n", - "## Setting up the matrices and the problem\n", + "xgb.plot_tree(xg_clf,num_trees=0)\n", + "plt.rcParams['figure.figsize'] = [50, 10]\n", + "save_fig(\"xgtree\")\n", + "plt.show()\n", "\n", - "We have the general problem" + "xgb.plot_importance(xg_clf)\n", + "plt.rcParams['figure.figsize'] = [5, 5]\n", + "save_fig(\"xgparams\")\n", + "plt.show()" ] }, { "cell_type": "markdown", - "metadata": {}, + "id": "6d840ba2", + "metadata": { + "editable": true + }, "source": [ - "$$\n", - "\\begin{align*}\n", - " &\\mathrm{min}_{\\lambda}\\hspace{0.2cm} \\frac{1}{2}\\boldsymbol{\\lambda}^T\\boldsymbol{P}\\boldsymbol{\\lambda}+\\boldsymbol{q}^T\\boldsymbol{\\lambda},\\\\ \\nonumber\n", - " &\\mathrm{s.t.} \\hspace{0.2cm} \\boldsymbol{G}\\boldsymbol{\\lambda} \\preceq \\boldsymbol{h} \\wedge \\boldsymbol{A}\\boldsymbol{\\lambda}=f.\n", - "\\end{align*}\n", - "$$" + "## Gradient boosting, making our own code for a regression case" + ] + }, + { + "cell_type": "code", + "execution_count": 7, + "id": "736c2963", + "metadata": { + "collapsed": false, + "editable": true + }, + "outputs": [], + "source": [ + "import numpy as np\n", + "class DecisionTreeRegressor:\n", + " def __init__(self, max_depth=3):\n", + " self.max_depth = max_depth\n", + " self.tree = None\n", + " def fit(self, X, y):\n", + " self.tree = self._grow_tree(X, y)\n", + " def _grow_tree(self, X, y, depth=0):\n", + " n_samples, n_features = X.shape\n", + " if depth < self.max_depth:\n", + " best_feature, best_threshold = self._best_split(X, y)\n", + " if best_feature is not None:\n", + " left_indices = X[:, best_feature] < best_threshold\n", + " right_indices = X[:, best_feature] >= best_threshold\n", + " left_child = self._grow_tree(X[left_indices], y[left_indices], depth + 1)\n", + " right_child = self._grow_tree(X[right_indices], y[right_indices], depth + 1)\n", + " return (best_feature, best_threshold, left_child, right_child)\n", + " return np.mean(y)\n", + " def _best_split(self, X, y):\n", + " best_mse = float('inf')\n", + " best_feature, best_threshold = None, None\n", + " n_samples, n_features = X.shape\n", + " \n", + " for feature in range(n_features):\n", + " thresholds = np.unique(X[:, feature])\n", + " for threshold in thresholds:\n", + " left_indices = X[:, feature] < threshold\n", + " right_indices = X[:, feature] >= threshold\n", + " if len(y[left_indices]) > 0 and len(y[right_indices]) > 0:\n", + " left_mse = np.mean((y[left_indices] - np.mean(y[left_indices])) ** 2)\n", + " right_mse = np.mean((y[right_indices] - np.mean(y[right_indices])) ** 2)\n", + " mse = (len(y[left_indices]) * left_mse + len(y[right_indices]) * right_mse) / n_samples\n", + " \n", + " if mse < best_mse:\n", + " best_mse = mse\n", + " best_feature = feature\n", + " best_threshold = threshold\n", + " return best_feature, best_threshold\n", + " def predict(self, X):\n", + " return np.array([self._predict_sample(sample, self.tree) for sample in X])\n", + " def _predict_sample(self, sample, node):\n", + " if isinstance(node, tuple):\n", + " feature, threshold, left_child, right_child = node\n", + " if sample[feature] < threshold:\n", + " return self._predict_sample(sample, left_child)\n", + " else:\n", + " return self._predict_sample(sample, right_child)\n", + " return node\n", + "class GradientBoostingRegressor:\n", + " def __init__(self, n_estimators=100, learning_rate=0.1, max_depth=3):\n", + " self.n_estimators = n_estimators\n", + " self.learning_rate = learning_rate\n", + " self.max_depth = max_depth\n", + " self.models = []\n", + " def fit(self, X, y):\n", + " y_pred = np.zeros(y.shape)\n", + " for _ in range(self.n_estimators):\n", + " residuals = y - y_pred\n", + " model = DecisionTreeRegressor(max_depth=self.max_depth)\n", + " model.fit(X, residuals)\n", + " y_pred += self.learning_rate * model.predict(X)\n", + " self.models.append(model)\n", + " def predict(self, X):\n", + " y_pred = np.zeros(X.shape[0])\n", + " for model in self.models:\n", + " y_pred += self.learning_rate * model.predict(X)\n", + " return y_pred\n", + "# Example usage\n", + "if __name__ == \"__main__\":\n", + " # Sample data\n", + " X = np.array([[1], [2], [3], [4], [5]])\n", + " y = np.array([1.5, 1.7, 3.5, 3.7, 5.0])\n", + " model = GradientBoostingRegressor(n_estimators=100, learning_rate=0.1, max_depth=2)\n", + " model.fit(X, y)\n", + " predictions = model.predict(X)\n", + " print(\"Predictions:\", predictions)" ] }, { "cell_type": "markdown", - "metadata": {}, + "id": "a016e0f1", + "metadata": { + "editable": true + }, "source": [ - "1. With a given kernel we can thus define the matrix $\\boldsymbol{P}$.\n", - "\n", - "2. The matrix $\\boldsymbol{P}$ has matrix elements $p_{ij}=y_iy_jK(\\boldsymbol{x}_i,\\boldsymbol{x}_j)$. Given a kernel $K$ and the targets $y_i$ this matrix is easy to set up.\n", - "\n", - "3. The vector $\\boldsymbol{q}$ has all elements equal to $-1$.\n", - "\n", - "4. The constraint $\\boldsymbol{y}^T\\boldsymbol{\\lambda}=0$ leads to $f=0$ and $\\boldsymbol{A}=\\boldsymbol{y}$.\n", - "\n", - "5. To set up the matrix $\\boldsymbol{G}$ we note that the inequalities $0\\leq \\lambda_i \\leq C$ can be split up into $0\\leq \\lambda_i$ and $\\lambda_i \\leq C$. These two inequalities define then the matrix $\\boldsymbol{G}$ and the vector $\\boldsymbol{h}$.\n", - "\n", - "## Setting up $\\boldsymbol{G}\\boldsymbol{\\lambda} \\preceq \\boldsymbol{h}$\n", - "\n", - "We have two constraints, $0\\le \\lambda_i$ and $\\lambda_i \\le C$. To do this we multiply the ones with the constraint\n", - "$\\ge$ with $-1$ in order to get $\\le$. It means that the problem $\\boldsymbol{G}\\boldsymbol{\\lambda} \\preceq \\boldsymbol{h}$\n", - "can be written as" + "## Summary of course" ] }, { "cell_type": "markdown", - "metadata": {}, + "id": "db7fd202", + "metadata": { + "editable": true + }, "source": [ - "$$\n", - "\\begin{bmatrix} -1& 0 & 0 & \\dots & 0 \\\\\n", - "0& -1 & 0 & \\dots & 0 \\\\\n", - "0& 0 & -1 & \\dots & 0 \\\\\n", - "\\dots& \\dots & \\dots & \\dots & \\dots \\\\\n", - "0& 0 & 0 & \\dots & -1 \\\\\n", - "1 & 0 & 0 & \\dots & 0 \\\\\n", - "0& 1 & 0 & \\dots & 0 \\\\\n", - "0& 0 & 1 & \\dots & 0 \\\\\n", - "\\dots& \\dots & \\dots & \\dots & \\dots \\\\\n", - "0& 0 & 0 & \\dots & 1 \\\\\n", - "\\end{bmatrix}\n", - "\\begin{bmatrix} \\lambda_1 \\\\\n", - "\\lambda_2 \\\\\n", - "\\lambda_3 \\\\\n", - "\\dots \\\\\n", - "\\lambda_n \\\\\n", - "\\end{bmatrix} \\preceq \n", - "\\begin{bmatrix} 0 \\\\\n", - "0 \\\\\n", - "0 \\\\\n", - "\\dots \\\\\n", - "0 \\\\\n", - "C \\\\\n", - "C \\\\\n", - "C \\\\\n", - "\\dots \\\\\n", - "C \\\\\n", - "\\end{bmatrix}.\n", - "$$" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "And then we are ready to go.\n", - "\n", - "## SVMs and Regression and multiclass classification\n", - "\n", - "More text will be added here. See also [Hastie et al.](https://www.springer.com/gp/book/9780387848570) section 12.3.\n", - "\n", - "## Summary of course\n", - "\n", "## What? Me worry? No final exam in this course!\n", "\n", "\n", "\n", - "

    \n", - "\n", - "\n", - "\n", - "\n", - "\n", - "\n", - "\n", + "

    Figure 1:

    \n", + "" + ] + }, + { + "cell_type": "markdown", + "id": "ff2eea1f", + "metadata": { + "editable": true + }, + "source": [ "## Topics we have covered this year\n", "\n", "The course has two central parts\n", "\n", "1. Statistical analysis and optimization of data\n", "\n", - "2. Machine learning\n", - "\n", + "2. Machine learning" + ] + }, + { + "cell_type": "markdown", + "id": "afa4b12f", + "metadata": { + "editable": true + }, + "source": [ "## Statistical analysis and optimization of data\n", "\n", "The following topics have been discussed:\n", @@ -920,8 +1593,16 @@ "\n", "6. Practical optimization using Singular-value decomposition and least squares for parameterizing data.\n", "\n", - "7. Principal Component Analysis to reduce the number of features.\n", - "\n", + "7. Principal Component Analysis to reduce the number of features." + ] + }, + { + "cell_type": "markdown", + "id": "29f93398", + "metadata": { + "editable": true + }, + "source": [ "## Machine learning\n", "\n", "The following topics will be covered\n", @@ -935,7 +1616,6 @@ "\n", "d. Logistic regression\n", "\n", - "\n", "5. Neural networks and deep learning:\n", "\n", "a. Feed Forward Neural Networks\n", @@ -944,7 +1624,6 @@ "\n", "c. Recurrent Neural Networks\n", "\n", - "\n", "4. Decisions trees and ensemble methods:\n", "\n", "a. Decision trees\n", @@ -955,16 +1634,22 @@ "\n", "d. Boosting and gradient boosting\n", "\n", - "\n", "5. Support vector machines\n", "\n", "a. Binary classification and multiclass classification\n", "\n", "b. Kernel methods\n", "\n", - "c. Regression\n", - "\n", - "\n", + "c. Regression" + ] + }, + { + "cell_type": "markdown", + "id": "3eb0cf47", + "metadata": { + "editable": true + }, + "source": [ "## Learning outcomes and overarching aims of this course\n", "\n", "The course introduces a variety of central algorithms and methods\n", @@ -991,8 +1676,16 @@ "\n", "* Have an understanding of central algorithms used in data analysis and machine learning;\n", "\n", - "* Work on numerical projects to illustrate the theory. The projects play a central role and you are expected to know modern programming languages like Python or C++.\n", - "\n", + "* Work on numerical projects to illustrate the theory. The projects play a central role and you are expected to know modern programming languages like Python or C++." + ] + }, + { + "cell_type": "markdown", + "id": "0e7a583d", + "metadata": { + "editable": true + }, + "source": [ "## Perspective on Machine Learning\n", "\n", "1. Rapidly emerging application area\n", @@ -1006,9 +1699,16 @@ "Huge amounts of data sets require automation, classical analysis tools often inadequate. \n", "High energy physics hit this wall in the 90’s.\n", "In 2009 single top quark production was determined via [Boosted decision trees, Bayesian\n", - "Neural Networks, etc.](https://arxiv.org/pdf/0903.0850.pdf)\n", - "\n", - "\n", + "Neural Networks, etc.](https://arxiv.org/pdf/0903.0850.pdf)" + ] + }, + { + "cell_type": "markdown", + "id": "ef032896", + "metadata": { + "editable": true + }, + "source": [ "## Machine Learning Research\n", "\n", "Where to find recent results:\n", @@ -1022,8 +1722,16 @@ "\n", "5. [Journal of Machine Learning Research](http://www.jmlr.org/papers/v19/) \n", "\n", - "6. [Follow ML on ArXiv](https://arxiv.org/list/cs.LG/recent)\n", - "\n", + "6. [Follow ML on ArXiv](https://arxiv.org/list/cs.LG/recent)" + ] + }, + { + "cell_type": "markdown", + "id": "d8a99fa4", + "metadata": { + "editable": true + }, + "source": [ "## Starting your Machine Learning Project\n", "\n", "1. Identify problem type: classification, regression\n", @@ -1034,16 +1742,32 @@ "\n", "4. Consider your data carefully again! Think of data representation more carefully.\n", "\n", - "5. Based on your results, feedback loop to earliest possible point\n", - "\n", + "5. Based on your results, feedback loop to earliest possible point" + ] + }, + { + "cell_type": "markdown", + "id": "44bf42ac", + "metadata": { + "editable": true + }, + "source": [ "## Choose a Model and Algorithm\n", "\n", "1. Supervised?\n", "\n", "2. Start with the simplest model that fits your problem\n", "\n", - "3. Start with minimal processing of data\n", - "\n", + "3. Start with minimal processing of data" + ] + }, + { + "cell_type": "markdown", + "id": "3bd98dfb", + "metadata": { + "editable": true + }, + "source": [ "## Preparing Your Data\n", "\n", "1. Shuffle your data\n", @@ -1052,21 +1776,26 @@ "\n", " * Why?\n", "\n", - "\n", "3. Normalize the variance\n", "\n", " * Why?\n", "\n", - "\n", "4. **Whitening**\n", "\n", " * Decorrelates data\n", "\n", " * Can be hit or miss\n", "\n", - "\n", - "5. When to do train/test split?\n", - "\n", + "5. When to do train/test split?" + ] + }, + { + "cell_type": "markdown", + "id": "6ecf1ff7", + "metadata": { + "editable": true + }, + "source": [ "## Which Activation and Weights to Choose in Neural Networks\n", "\n", "1. RELU? ELU?\n", @@ -1077,18 +1806,23 @@ "\n", " * Terrible idea\n", "\n", - "\n", "4. Set all weights to random values?\n", "\n", - " * Small random values\n", - "\n", - "\n", + " * Small random values" + ] + }, + { + "cell_type": "markdown", + "id": "d872afb1", + "metadata": { + "editable": true + }, + "source": [ "## Optimization Methods and Hyperparameters\n", "1. Stochastic gradient descent\n", "\n", "a. Stochastic gradient descent + momentum\n", "\n", - "\n", "2. State-of-the-art approaches:\n", "\n", " * RMSProp\n", @@ -1097,12 +1831,18 @@ "\n", " * and more\n", "\n", - "\n", "Which regularization and hyperparameters? $L_1$ or $L_2$, soft\n", "classifiers, depths of trees and many other. Need to explore a large\n", - "set of hyperparameters and regularization methods.\n", - "\n", - "\n", + "set of hyperparameters and regularization methods." + ] + }, + { + "cell_type": "markdown", + "id": "f1c129da", + "metadata": { + "editable": true + }, + "source": [ "## Resampling\n", "\n", "When do we resample?\n", @@ -1111,8 +1851,16 @@ "\n", "2. [Cross-validation](https://www.youtube.com/watch?v=fSytzGwwBVw&ab_channel=StatQuestwithJoshStarmer)\n", "\n", - "3. Jackknife and many other\n", - "\n", + "3. Jackknife and many other" + ] + }, + { + "cell_type": "markdown", + "id": "d0f27c93", + "metadata": { + "editable": true + }, + "source": [ "## Other courses on Data science and Machine Learning at UiO\n", "\n", "The link here gives an excellent overview of courses on Machine learning at UiO.\n", @@ -1131,14 +1879,30 @@ "\n", "7. [IN5400/INF5860 – Machine Learning for Image Analysis](https://www.uio.no/studier/emner/matnat/ifi/IN5400/). An introduction to deep learning with particular emphasis on applications within Image analysis, but useful for other application areas too.\n", "\n", - "8. [TEK5040 – Dyp læring for autonome systemer](https://www.uio.no/studier/emner/matnat/its/TEK5040/). The course addresses advanced algorithms and architectures for deep learning with neural networks. The course provides an introduction to how deep-learning techniques can be used in the construction of key parts of advanced autonomous systems that exist in physical environments and cyber environments.\n", - "\n", + "8. [TEK5040 – Dyp læring for autonome systemer](https://www.uio.no/studier/emner/matnat/its/TEK5040/). The course addresses advanced algorithms and architectures for deep learning with neural networks. The course provides an introduction to how deep-learning techniques can be used in the construction of key parts of advanced autonomous systems that exist in physical environments and cyber environments." + ] + }, + { + "cell_type": "markdown", + "id": "e28e0a40", + "metadata": { + "editable": true + }, + "source": [ "## Additional courses of interest\n", "\n", "1. [STK4051 Computational Statistics](https://www.uio.no/studier/emner/matnat/math/STK4051/index-eng.html)\n", "\n", - "2. [STK4021 Applied Bayesian Analysis and Numerical Methods](https://www.uio.no/studier/emner/matnat/math/STK4021/index-eng.html)\n", - "\n", + "2. [STK4021 Applied Bayesian Analysis and Numerical Methods](https://www.uio.no/studier/emner/matnat/math/STK4021/index-eng.html)" + ] + }, + { + "cell_type": "markdown", + "id": "3fd96a2c", + "metadata": { + "editable": true + }, + "source": [ "## What's the future like?\n", "\n", "Based on multi-layer nonlinear neural networks, deep learning can\n", @@ -1155,8 +1919,16 @@ "\n", "2. Recurrent neural networks, which can process sequential data of variable length and have been widely used in natural language understanding and speech processing;\n", "\n", - "3. Encoder-decoder framework, which is mostly used for image or sequence generation, such as machine translation, text summarization, and image captioning.\n", - "\n", + "3. Encoder-decoder framework, which is mostly used for image or sequence generation, such as machine translation, text summarization, and image captioning." + ] + }, + { + "cell_type": "markdown", + "id": "bb05416d", + "metadata": { + "editable": true + }, + "source": [ "## Types of Machine Learning, a repetition\n", "\n", "The approaches to machine learning are many, but are often split into two main categories. \n", @@ -1176,11 +1948,16 @@ "\n", " * Clustering: Data are divided into groups with certain common traits, without knowing the different groups beforehand. It is thus a form of unsupervised learning.\n", "\n", - " * Other unsupervised learning algortihms like **Boltzmann machines**\n", - "\n", - "\n", - "\n", - "\n", + " * Other unsupervised learning algortihms like **Boltzmann machines**" + ] + }, + { + "cell_type": "markdown", + "id": "21d8fbea", + "metadata": { + "editable": true + }, + "source": [ "## Why Boltzmann machines?\n", "\n", "What is known as restricted Boltzmann Machines (RMB) have received a lot of attention lately. \n", @@ -1188,9 +1965,16 @@ "\n", "The original RBMs had just one visible layer and a hidden layer, but recently so-called Gaussian-binary RBMs have gained quite some popularity in imaging since they are capable of modeling continuous data that are common to natural images. \n", "\n", - "Furthermore, they have been used to solve complicated [quantum mechanical many-particle problems or classical statistical physics problems like the Ising and Potts classes of models](https://journals.aps.org/rmp/abstract/10.1103/RevModPhys.91.045002). \n", - "\n", - "\n", + "Furthermore, they have been used to solve complicated [quantum mechanical many-particle problems or classical statistical physics problems like the Ising and Potts classes of models](https://journals.aps.org/rmp/abstract/10.1103/RevModPhys.91.045002)." + ] + }, + { + "cell_type": "markdown", + "id": "7107efa8", + "metadata": { + "editable": true + }, + "source": [ "## Boltzmann Machines\n", "\n", "Why use a generative model rather than the more well known discriminative deep neural networks (DNN)? \n", @@ -1203,9 +1987,16 @@ "\n", "b. Generate a sample of an ordered or disordered phase, having been given samples of such phases.\n", "\n", - "c. Model the trial function for [Monte Carlo calculations](https://journals.aps.org/rmp/abstract/10.1103/RevModPhys.91.045002).\n", - "\n", - "\n", + "c. Model the trial function for [Monte Carlo calculations](https://journals.aps.org/rmp/abstract/10.1103/RevModPhys.91.045002)." + ] + }, + { + "cell_type": "markdown", + "id": "74dcfbe6", + "metadata": { + "editable": true + }, + "source": [ "## Some similarities and differences from DNNs\n", "\n", "1. Both use gradient-descent based learning procedures for minimizing cost functions\n", @@ -1214,67 +2005,89 @@ "\n", "3. DNNs often have several hidden layers. A restricted Boltzmann machine has only one hidden layer, however several RBMs can be stacked to make up Deep Belief Networks, of which they constitute the building blocks.\n", "\n", - "History: The RBM was developed by amongst others [Geoffrey Hinton](https://en.wikipedia.org/wiki/Geoffrey_Hinton), called by some the \"Godfather of Deep Learning\", working with the University of Toronto and Google.\n", - "\n", - "\n", + "History: The RBM was developed by amongst others [Geoffrey Hinton](https://en.wikipedia.org/wiki/Geoffrey_Hinton), called by some the \"Godfather of Deep Learning\", working with the University of Toronto and Google." + ] + }, + { + "cell_type": "markdown", + "id": "75c514e3", + "metadata": { + "editable": true + }, + "source": [ "## Boltzmann machines (BM)\n", "\n", "A BM is what we would call an undirected probabilistic graphical model\n", "with stochastic continuous or discrete units.\n", "\n", - "\n", "It is interpreted as a stochastic recurrent neural network where the\n", "state of each unit(neurons/nodes) depends on the units it is connected\n", "to. The weights in the network represent thus the strength of the\n", "interaction between various units/nodes.\n", "\n", - "\n", "It turns into a Hopfield network if we choose deterministic rather\n", "than stochastic units. In contrast to a Hopfield network, a BM is a\n", "so-called generative model. It allows us to generate new samples from\n", - "the learned distribution.\n", - "\n", - "\n", - "\n", + "the learned distribution." + ] + }, + { + "cell_type": "markdown", + "id": "1e734682", + "metadata": { + "editable": true + }, + "source": [ "## A standard BM setup\n", "\n", "A standard BM network is divided into a set of observable and visible units $\\hat{x}$ and a set of unknown hidden units/nodes $\\hat{h}$.\n", "\n", - "\n", - "\n", "Additionally there can be bias nodes for the hidden and visible layers. These biases are normally set to $1$.\n", "\n", - "\n", - "\n", "BMs are stackable, meaning they cwe can train a BM which serves as input to another BM. We can construct deep networks for learning complex PDFs. The layers can be trained one after another, a feature which makes them popular in deep learning\n", "\n", - "\n", - "\n", "However, they are often hard to train. This leads to the introduction of so-called restricted BMs, or RBMS.\n", - "Here we take away all lateral connections between nodes in the visible layer as well as connections between nodes in the hidden layer. The network is illustrated in the figure below.\n", - "\n", - "\n", - "\n", + "Here we take away all lateral connections between nodes in the visible layer as well as connections between nodes in the hidden layer. The network is illustrated in the figure below." + ] + }, + { + "cell_type": "markdown", + "id": "c89378ce", + "metadata": { + "editable": true + }, + "source": [ "## The structure of the RBM network\n", "\n", "\n", "\n", "\n", - "

    \n", - "\n", - "\n", - "\n", - "\n", - "\n", - "\n", - "\n", + "

    Figure 1:

    \n", + "" + ] + }, + { + "cell_type": "markdown", + "id": "e96b75e9", + "metadata": { + "editable": true + }, + "source": [ "## The network\n", "\n", "**The network layers**:\n", "1. A function $\\mathbf{x}$ that represents the visible layer, a vector of $M$ elements (nodes). This layer represents both what the RBM might be given as training input, and what we want it to be able to reconstruct. This might for example be given by the pixels of an image or coefficients representing speech, or the coordinates of a quantum mechanical state function.\n", "\n", - "2. The function $\\mathbf{h}$ represents the hidden, or latent, layer. A vector of $N$ elements (nodes). Also called \"feature detectors\".\n", - "\n", + "2. The function $\\mathbf{h}$ represents the hidden, or latent, layer. A vector of $N$ elements (nodes). Also called \"feature detectors\"." + ] + }, + { + "cell_type": "markdown", + "id": "29c228a3", + "metadata": { + "editable": true + }, + "source": [ "## Goals\n", "\n", "The goal of the hidden layer is to increase the model's expressive\n", @@ -1289,8 +2102,16 @@ "\n", "2. $\\mathbf{b}$ represents the hidden bias, a vector of same lenght as $\\mathbf{h}$.\n", "\n", - "3. $W$ represents the interaction weights, a matrix of size $M\\times N$.\n", - "\n", + "3. $W$ represents the interaction weights, a matrix of size $M\\times N$." + ] + }, + { + "cell_type": "markdown", + "id": "a9297455", + "metadata": { + "editable": true + }, + "source": [ "## Joint distribution\n", "\n", "The restricted Boltzmann machine is described by a Boltzmann distribution" @@ -1298,7 +2119,10 @@ }, { "cell_type": "markdown", - "metadata": {}, + "id": "b6d45ca8", + "metadata": { + "editable": true + }, "source": [ "\n", "
    \n", @@ -1313,14 +2137,20 @@ }, { "cell_type": "markdown", - "metadata": {}, + "id": "a794db10", + "metadata": { + "editable": true + }, "source": [ "where $Z$ is the normalization constant or partition function, defined as" ] }, { "cell_type": "markdown", - "metadata": {}, + "id": "5508a27c", + "metadata": { + "editable": true + }, "source": [ "\n", "
    \n", @@ -1335,13 +2165,21 @@ }, { "cell_type": "markdown", - "metadata": {}, + "id": "074eda44", + "metadata": { + "editable": true + }, + "source": [ + "It is common to ignore $T_0$ by setting it to one." + ] + }, + { + "cell_type": "markdown", + "id": "ee390a08", + "metadata": { + "editable": true + }, "source": [ - "It is common to ignore $T_0$ by setting it to one. \n", - "\n", - "\n", - "\n", - "\n", "## Network Elements, the energy function\n", "\n", "The function $E(\\mathbf{x},\\mathbf{h})$ gives the **energy** of a\n", @@ -1356,7 +2194,10 @@ }, { "cell_type": "markdown", - "metadata": {}, + "id": "4f3aa1f0", + "metadata": { + "editable": true + }, "source": [ "$$\n", "E(\\hat{x},\\hat{h}) = -\\sum_{ia}^{NA}b_i^a \\alpha_i^a(x_i)-\\sum_{jd}^{MD}c_j^d \\beta_j^d(h_j)-\\sum_{ijad}^{NAMD}b_i^a \\alpha_i^a(x_i)c_j^d \\beta_j^d(h_j)w_{ij}^{ad}.\n", @@ -1365,26 +2206,39 @@ }, { "cell_type": "markdown", - "metadata": {}, + "id": "62947b2d", + "metadata": { + "editable": true + }, "source": [ "Here $\\beta_j^d(h_j)$ and $\\alpha_i^a(x_j)$ are so-called transfer functions that map a given input value to a desired feature value. The labels $a$ and $d$ denote that there can be multiple transfer functions per variable. The first sum depends only on the visible units. The second on the hidden ones. **Note** that there is no connection between nodes in a layer.\n", "\n", "The quantities $b$ and $c$ can be interpreted as the visible and hidden biases, respectively.\n", "\n", - "The connection between the nodes in the two layers is given by the weights $w_{ij}$. \n", - "\n", + "The connection between the nodes in the two layers is given by the weights $w_{ij}$." + ] + }, + { + "cell_type": "markdown", + "id": "501d4d2c", + "metadata": { + "editable": true + }, + "source": [ "## Defining different types of RBMs\n", "There are different variants of RBMs, and the differences lie in the types of visible and hidden units we choose as well as in the implementation of the energy function $E(\\mathbf{x},\\mathbf{h})$. \n", "\n", "**Binary-Binary RBM:**\n", "\n", - "\n", "RBMs were first developed using binary units in both the visible and hidden layer. The corresponding energy function is defined as follows:" ] }, { "cell_type": "markdown", - "metadata": {}, + "id": "a15140cf", + "metadata": { + "editable": true + }, "source": [ "\n", "
    \n", @@ -1399,20 +2253,24 @@ }, { "cell_type": "markdown", - "metadata": {}, + "id": "26cdb304", + "metadata": { + "editable": true + }, "source": [ "where the binary values taken on by the nodes are most commonly 0 and 1.\n", "\n", - "\n", "**Gaussian-Binary RBM:**\n", "\n", - "\n", "Another varient is the RBM where the visible units are Gaussian while the hidden units remain binary:" ] }, { "cell_type": "markdown", - "metadata": {}, + "id": "3bd28b90", + "metadata": { + "editable": true + }, "source": [ "\n", "
    \n", @@ -1427,7 +2285,10 @@ }, { "cell_type": "markdown", - "metadata": {}, + "id": "9ddc6c22", + "metadata": { + "editable": true + }, "source": [ "## More about RBMs\n", "1. Useful when we model continuous data (i.e., we wish $\\mathbf{x}$ to be continuous)\n", @@ -1443,9 +2304,16 @@ "\n", "4. Rectified linear units\n", "\n", - "To read more, see [Lectures on Boltzmann machines in Physics](https://github.com/CompPhysics/ComputationalPhysics2/blob/gh-pages/doc/pub/notebook2/ipynb/notebook2.ipynb).\n", - "\n", - "\n", + "To read more, see [Lectures on Boltzmann machines in Physics](https://github.com/CompPhysics/ComputationalPhysics2/blob/gh-pages/doc/pub/notebook2/ipynb/notebook2.ipynb)." + ] + }, + { + "cell_type": "markdown", + "id": "bf03ef62", + "metadata": { + "editable": true + }, + "source": [ "## Autoencoders: Overarching view\n", "\n", "Autoencoders are artificial neural networks capable of learning\n", @@ -1474,8 +2342,16 @@ "\n", "[Video on autoencoders](https://www.coursera.org/lecture/building-deep-learning-models-with-tensorflow/autoencoders-1U4L3)\n", "\n", - "See also A. Geron's textbook, chapter 15.\n", - "\n", + "See also A. Geron's textbook, chapter 15." + ] + }, + { + "cell_type": "markdown", + "id": "f1ac866c", + "metadata": { + "editable": true + }, + "source": [ "## Bayesian Machine Learning\n", "\n", "This is an important topic if we aim at extracting a probability\n", @@ -1489,8 +2365,16 @@ "\n", "[Video on Bayesian deep learning](https://www.youtube.com/watch?v=E1qhGw8QxqY&ab_channel=AndrewGordonWilson)\n", "\n", - "See also the [slides here](https://github.com/CompPhysics/MachineLearning/blob/master/doc/Articles/lec03.pdf).\n", - "\n", + "See also the [slides here](https://github.com/CompPhysics/MachineLearning/blob/master/doc/Articles/lec03.pdf)." + ] + }, + { + "cell_type": "markdown", + "id": "72efdf13", + "metadata": { + "editable": true + }, + "source": [ "## Reinforcement Learning\n", "\n", "Reinforcement Learning (RL) is one of the most exciting fields of\n", @@ -1517,8 +2401,16 @@ "\n", "[Lecture on Reinforcement Learning](https://www.youtube.com/watch?v=FgzM3zpZ55o&ab_channel=stanfordonline).\n", "\n", - "See also A. Geron's textbook, chapter 16.\n", - "\n", + "See also A. Geron's textbook, chapter 16." + ] + }, + { + "cell_type": "markdown", + "id": "b6c3dab1", + "metadata": { + "editable": true + }, + "source": [ "## Transfer learning\n", "\n", "The goal of transfer learning is to transfer the model or knowledge\n", @@ -1530,8 +2422,16 @@ "for us to better solve the target task. Transfer learning is a hot\n", "research topic in recent years, with many problems still waiting to be studied.\n", "\n", - "[Lecture on transfer learning](https://www.ias.edu/video/machinelearning/2020/0331-SamoryKpotufe).\n", - "\n", + "[Lecture on transfer learning](https://www.ias.edu/video/machinelearning/2020/0331-SamoryKpotufe)." + ] + }, + { + "cell_type": "markdown", + "id": "14b0cd78", + "metadata": { + "editable": true + }, + "source": [ "## Adversarial learning\n", "\n", "The conventional deep generative model has a potential problem: the\n", @@ -1544,8 +2444,16 @@ "technologies, generative adversarial networks (GAN), has already been\n", "successfully applied to image, speech, and text.\n", "\n", - "[Lecture on adversial learning](https://www.youtube.com/watch?v=CIfsB_EYsVI&ab_channel=StanfordUniversitySchoolofEngineering).\n", - "\n", + "[Lecture on adversial learning](https://www.youtube.com/watch?v=CIfsB_EYsVI&ab_channel=StanfordUniversitySchoolofEngineering)." + ] + }, + { + "cell_type": "markdown", + "id": "2082df21", + "metadata": { + "editable": true + }, + "source": [ "## Dual learning\n", "\n", "Dual learning is a new learning paradigm, the basic idea of which is\n", @@ -1556,16 +2464,31 @@ "many problems in machine learning, including machine translation,\n", "image style conversion, question answering and generation, image\n", "classification and generation, text classification and generation,\n", - "image-to-text, and text-to-image.\n", - "\n", + "image-to-text, and text-to-image." + ] + }, + { + "cell_type": "markdown", + "id": "0191be7d", + "metadata": { + "editable": true + }, + "source": [ "## Distributed machine learning\n", "\n", "Distributed computation will speed up machine learning algorithms,\n", "significantly improve their efficiency, and thus enlarge their\n", "application. When distributed meets machine learning, more than just\n", - "implementing the machine learning algorithms in parallel is required.\n", - "\n", - "\n", + "implementing the machine learning algorithms in parallel is required." + ] + }, + { + "cell_type": "markdown", + "id": "57579d7b", + "metadata": { + "editable": true + }, + "source": [ "## Meta learning\n", "\n", "Meta learning is an emerging research direction in machine\n", @@ -1574,8 +2497,16 @@ "itself, instead of just completing a specific learning task. That is,\n", "a meta learner needs to be able to evaluate its own learning methods\n", "and adjust its own learning methods according to specific learning\n", - "tasks.\n", - "\n", + "tasks." + ] + }, + { + "cell_type": "markdown", + "id": "5ee1418b", + "metadata": { + "editable": true + }, + "source": [ "## The Challenges Facing Machine Learning\n", "\n", "While there has been much progress in machine learning, there are also challenges.\n", @@ -1596,8 +2527,16 @@ "forward to the future, here are what we think the research hotspots in\n", "the next ten years will be.\n", "\n", - "See the article on [Discovery of Physics From Data: Universal Laws and Discrepancies](https://www.frontiersin.org/articles/10.3389/frai.2020.00025/full)\n", - "\n", + "See the article on [Discovery of Physics From Data: Universal Laws and Discrepancies](https://www.frontiersin.org/articles/10.3389/frai.2020.00025/full)" + ] + }, + { + "cell_type": "markdown", + "id": "147876af", + "metadata": { + "editable": true + }, + "source": [ "## Explainable machine learning\n", "\n", "Machine learning, especially deep learning, evolves rapidly. The\n", @@ -1617,8 +2556,16 @@ "transition from solving problems by data correlation to solving\n", "problems by logical reasoning.\n", "\n", - "**Bayesian Machine Learning is one of the exciting research directions in this field**.\n", - "\n", + "**Bayesian Machine Learning is one of the exciting research directions in this field**." + ] + }, + { + "cell_type": "markdown", + "id": "8c4fca27", + "metadata": { + "editable": true + }, + "source": [ "## Quantum machine learning\n", "\n", "Quantum machine learning is an emerging interdisciplinary research\n", @@ -1640,10 +2587,16 @@ "\n", "[Lecture on Quantum ML](https://www.youtube.com/watch?v=Xh9pUu3-WxM&ab_channel=InstituteforPure%26AppliedMathematics%28IPAM%29).\n", "\n", - "\n", - "[Read interview with Maria Schuld on her work on Quantum Machine Learning](https://physics.aps.org/articles/v13/179?utm_campaign=weekly&utm_medium=email&utm_source=emailalert). See also [her recent textbook](https://www.springer.com/gp/book/9783319964232). \n", - "\n", - "\n", + "[Read interview with Maria Schuld on her work on Quantum Machine Learning](https://physics.aps.org/articles/v13/179?utm_campaign=weekly&utm_medium=email&utm_source=emailalert). See also [her recent textbook](https://www.springer.com/gp/book/9783319964232)." + ] + }, + { + "cell_type": "markdown", + "id": "ab24b71c", + "metadata": { + "editable": true + }, + "source": [ "## Quantum machine learning algorithms based on linear algebra\n", "\n", "Many quantum machine learning algorithms are based on variants of\n", @@ -1657,8 +2610,16 @@ "machine learning algorithms is data input—that is, how to initialize\n", "the quantum system with the entire data set. Although efficient\n", "data-input algorithms exist for certain situations, how to efficiently\n", - "input data into a quantum system is as yet unknown for most cases.\n", - "\n", + "input data into a quantum system is as yet unknown for most cases." + ] + }, + { + "cell_type": "markdown", + "id": "38ce95dc", + "metadata": { + "editable": true + }, + "source": [ "## Quantum reinforcement learning\n", "\n", "In quantum reinforcement learning, a quantum agent interacts with the\n", @@ -1667,8 +2628,16 @@ "achieves quantum acceleration by the quantum processing capabilities\n", "of the agent or the possibility of exploring the environment through\n", "quantum superposition. Such algorithms have been proposed in\n", - "superconducting circuits and systems of trapped ions.\n", - "\n", + "superconducting circuits and systems of trapped ions." + ] + }, + { + "cell_type": "markdown", + "id": "03b3e934", + "metadata": { + "editable": true + }, + "source": [ "## Quantum deep learning\n", "\n", "Dedicated quantum information processors, such as quantum annealers\n", @@ -1682,9 +2651,16 @@ "correspond to an adjustable Ising model. Then, by initializing the\n", "input neurons in the Boltzmann machine to a fixed state and allowing\n", "the system to heat up, we can read out the output qubits to get the\n", - "result.\n", - "\n", - "\n", + "result." + ] + }, + { + "cell_type": "markdown", + "id": "9677955a", + "metadata": { + "editable": true + }, + "source": [ "## Social machine learning\n", "\n", "Machine learning aims to imitate how humans\n", @@ -1696,34 +2672,43 @@ "evolve by imitating human society so as to achieve more effective,\n", "intelligent, interpretable “social machine learning”?\n", "\n", - "And much more.\n", - "\n", - "\n", - "\n", + "And much more." + ] + }, + { + "cell_type": "markdown", + "id": "cc3ff097", + "metadata": { + "editable": true + }, + "source": [ "## The last words?\n", "\n", "Early computer scientist Alan Kay said, **The best way to predict the\n", "future is to create it**. Therefore, all machine learning\n", "practitioners, whether scholars or engineers, professors or students,\n", "need to work together to advance these important research\n", - "topics. Together, we will not just predict the future, but create it.\n", - "\n", - "\n", - "\n", - "\n", + "topics. Together, we will not just predict the future, but create it." + ] + }, + { + "cell_type": "markdown", + "id": "26c891b9", + "metadata": { + "editable": true + }, + "source": [ "## Best wishes to you all and thanks so much for your heroic efforts this semester\n", "\n", "\n", "\n", "\n", - "

    \n", - "\n", - "\n", + "

    Figure 1:

    \n", "" ] } ], "metadata": {}, "nbformat": 4, - "nbformat_minor": 4 + "nbformat_minor": 5 } diff --git a/doc/src/week48/backup2023.do.txt b/doc/src/week48/backup2023.do.txt new file mode 100644 index 000000000..6f178bacc --- /dev/null +++ b/doc/src/week48/backup2023.do.txt @@ -0,0 +1,1346 @@ +TITLE: Week 48: Support Vector Machines and Summary of course +AUTHOR: Morten Hjorth-Jensen {copyright, 1999-present|CC BY-NC} at Department of Physics, University of Oslo & Department of Physics and Astronomy and National Superconducting Cyclotron Laboratory, Michigan State University +DATE: today + +!split +===== Overview of week 48 ===== + +* _Thursday_: Support Vector Machines: Kernels, Classification and Regression. "Video of Lecture":"https://www.uio.no/studier/emner/matnat/fys/FYS-STK4155/h20/forelesningsvideoer/LectureNovember26.mp4?vrtx=view-as-webpage". +* _Friday_: Summary of course with perspectives for future studies. "Video of Lecture":"https://www.uio.no/studier/emner/matnat/fys/FYS-STK4155/h20/forelesningsvideoer/LectureNovember27.mp4?vrtx=view-as-webpage". + +See also Geron's chapter 5. Chapter 12 (sections 12.1-12.3 are the most relevant ones) of Hastie et al contains also a good discussion. + + +!split +===== Thursday ===== + +We finalize our discussion on Support Vector Machines with an emphasis on kernel transformations and applications to regression. The following "video attempts at giving an overview on this part":"https://www.youtube.com/watch?v=Toet3EiSFcM&ab_channel=StatQuestwithJoshStarmer". See also the "follow-up video":"https://www.youtube.com/watch?v=Qc5IyLW_hns&ab_channel=StatQuestwithJoshStarmer". + +!split +===== Friday ===== + +Friday's lecture is split in two parts. It starts with a summary of +what we have done this semester and continues with perspectives for future studies and +modern research projects in machine learning. + +!split +===== Support Vector Machines, overarching aims ===== + +As discussed last week, +a Support Vector Machine (SVM) is a very powerful and versatile +Machine Learning method, capable of performing linear or nonlinear +classification, regression, and even outlier detection. It is one of +the most popular models in Machine Learning, and anyone interested in +Machine Learning should have it in their toolbox. SVMs are +particularly well suited for classification of complex but small-sized or +medium-sized datasets. + +The case with two well-separated classes only can be understood in an +intuitive way in terms of lines in a two-dimensional space separating +the two classes. + +The basic mathematics behind the SVM is however less familiar to most of us. +It relies on the definition of hyperplanes and the +definition of a _margin_ which separates classes (in case of +classification problems) of variables. It is also used for regression +problems. I recommend you take a look at the lectures from last week on the binary classification problem. + +!split +===== Kernels and non-linearity ===== + +The cases we studied last week were all characterized by two classes +with a close to linear separability. The classifiers we have described +so far find linear boundaries in our input feature space. It is +possible to make our procedure more flexible by exploring the feature +space using other basis expansions such as higher-order polynomials, +wavelets, splines etc. + +If our feature space is not easy to separate, as shown in the figure +generated by the code below, we can achieve a better separation by introducing a more complex +basis functions. The ideal would be, as shown by the code example below, to, +via a specific transformation to obtain a separation between the +classes that is almost linear. See also Figures 12.2 and 12.3 of "Hastie et al.":"https://www.springer.com/gp/book/9780387848570". + +We can achieve a better separation by introducing more complex +basis functions. The ideal would be (see Figures 12.2 and 12.3) to, via a specific transformation to +obtain a separation between the classes which is almost linear. + +The change of basis, from $x\rightarrow z=\phi(x)$ leads to the same +type of equations to be solved, except that we need to introduce, for +example, a polynomial transformation to a two-dimensional training +set. + +!bc pycod +import numpy as np +import os + +np.random.seed(42) + +# To plot pretty figures +import matplotlib +import matplotlib.pyplot as plt +plt.rcParams['axes.labelsize'] = 14 +plt.rcParams['xtick.labelsize'] = 12 +plt.rcParams['ytick.labelsize'] = 12 + + +from sklearn.svm import SVC +from sklearn import datasets + + + +X1D = np.linspace(-4, 4, 9).reshape(-1, 1) +X2D = np.c_[X1D, X1D**2] +y = np.array([0, 0, 1, 1, 1, 1, 1, 0, 0]) + +plt.figure(figsize=(11, 4)) + +plt.subplot(121) +plt.grid(True, which='both') +plt.axhline(y=0, color='k') +plt.plot(X1D[:, 0][y==0], np.zeros(4), "bs") +plt.plot(X1D[:, 0][y==1], np.zeros(5), "g^") +plt.gca().get_yaxis().set_ticks([]) +plt.xlabel(r"$x_1$", fontsize=20) +plt.axis([-4.5, 4.5, -0.2, 0.2]) + +plt.subplot(122) +plt.grid(True, which='both') +plt.axhline(y=0, color='k') +plt.axvline(x=0, color='k') +plt.plot(X2D[:, 0][y==0], X2D[:, 1][y==0], "bs") +plt.plot(X2D[:, 0][y==1], X2D[:, 1][y==1], "g^") +plt.xlabel(r"$x_1$", fontsize=20) +plt.ylabel(r"$x_2$", fontsize=20, rotation=0) +plt.gca().get_yaxis().set_ticks([0, 4, 8, 12, 16]) +plt.plot([-4.5, 4.5], [6.5, 6.5], "r--", linewidth=3) +plt.axis([-4.5, 4.5, -1, 17]) +plt.subplots_adjust(right=1) +plt.show() + +!ec + + + +!split +===== The equations ===== + +Suppose we define a polynomial transformation of degree two only. We define a vector $\bm{x}_i=[x_i,y_i]$ and have +!bt +\[ +\bm{z}_i^T = \phi(\bm{x}_i)^T =\left[1, \sqrt{2}x_i, \sqrt{2}y_i, x_i^2, y_i^2, \sqrt{2}x_iy_i\right]. +\] +!et + +With our new basis, the equations we solved earlier are basically the same, that is we have now (without the slack option for simplicity) +!bt +\[ +{\cal L}=\sum_i\lambda_i-\frac{1}{2}\sum_{ij}^n\lambda_i\lambda_jy_iy_j\bm{z}_i^T\bm{z}_j, +\] +!et +subject to the constraints $\lambda_i\geq 0$, $\sum_i\lambda_iy_i=0$, and for the support vectors +!bt +\[ +y_i(\bm{z}_i^T\bm{w}+b)= 1 \hspace{0.1cm}\forall i, +\] +!et +from which we also find $b$. +To compute $\bm{z}_i^T\bm{z}_j$ we define the kernel $K(\bm{x}_i,\bm{x}_j)$ as +!bt +\[ +K(\bm{x}_i,\bm{x}_j)=\bm{z}_i^T\bm{z}_j= \phi(\bm{x}_i)^T\phi(\bm{x}_j). +\] +!et +For the above example, the kernel reads +!bt +\[ +K(\bm{x}_i,\bm{x}_j)=[1, \sqrt{2}x_i, \sqrt{2}y_i, x_i^2, y_i^2, \sqrt{2}x_iy_i]\begin{bmatrix} 1\\ \sqrt{2}x_j \\ \sqrt{2}y_j \\ x_j^2\\ y_i^2\\ \sqrt{2}x_jy_j \end{bmatrix}=1+2x_ix_j+2y_iy_j+(x_ix_j)^2+(y_iy_j)^2+2x_ix_jy_iy_j. +\] +!et + +We note that this dot product can be rewritten as +!bt +\[ +K(\bm{x}_i,\bm{x}_j)=[1+\bm{x}_i^T\bm{x}_j]^d, +\] +!et +where $d=2$ in our case and $\bm{x}_i=[x_i,y_i]$ and $\bm{x}_j=[x_j,y_j]$. +To compute the last equation is however inefficient from a computational stand. +Instead of computing the last equation for the kernel, we simply compute +the dot product $(\bm{x}_i^T\bm{x}_j)^2$. +This leads to the so-called +kernel trick. + +!split +===== The problem to solve ===== + +Using our definition of the kernel, we can rewrite again the Lagrangian +!bt +\[ +{\cal L}=\sum_i\lambda_i-\frac{1}{2}\sum_{ij}^n\lambda_i\lambda_jy_iy_j\bm{z}_i^T\bm{z}_j, +\] +!et +subject to the constraints $\lambda_i\geq 0$, $\sum_i\lambda_iy_i=0$ in terms of a convex optimization problem +!bt +\[ +\frac{1}{2} \bm{\lambda}^T\begin{bmatrix} y_1y_1K(\bm{x}_1,\bm{x}_1) & y_1y_2K(\bm{x}_1,\bm{x}_2) & \dots & \dots & y_1y_nK(\bm{x}_1,\bm{x}_n) \\ +y_2y_1K(\bm{x}_2,\bm{x}_1) & y_2y_2(\bm{x}_2,\bm{x}_2) & \dots & \dots & y_1y_nK(\bm{x}_2,\bm{x}_n) \\ +\dots & \dots & \dots & \dots & \dots \\ +\dots & \dots & \dots & \dots & \dots \\ +y_ny_1K(\bm{x}_n,\bm{x}_1) & y_ny_2K(\bm{x}_n\bm{x}_2) & \dots & \dots & y_ny_nK(\bm{x}_n,\bm{x}_n) \\ +\end{bmatrix}\bm{\lambda}-\mathbb{1}\bm{\lambda}, +\] +!et +subject to $\bm{y}^T\bm{\lambda}=0$. Here we defined the vectors $\bm{\lambda} =[\lambda_1,\lambda_2,\dots,\lambda_n]$ and +$\bm{y}=[y_1,y_2,\dots,y_n]$. +If we add the slack constants this leads to the additional constraint $0\leq \lambda_i \leq C$. + +!split +===== Tailoring the equations to the usage of CVXOPT ===== + +We can rewrite this (see the solutions below) in terms of a convex optimization problem of the type +!bt +\begin{align*} + &\mathrm{min}_{\lambda}\hspace{0.2cm} \frac{1}{2}\bm{\lambda}^T\bm{P}\bm{\lambda}+\bm{q}^T\bm{\lambda},\\ \nonumber + &\mathrm{s.t} \hspace{0.2cm} \bm{G}\bm{\lambda} \preceq \bm{h} \hspace{0.2cm} \wedge \bm{A}\bm{\lambda}=f. +\end{align*} +!et +Below we discuss how to solve these equations. Here we note that the matrix $\bm{P}$ has matrix elements $p_{ij}=y_iy_jK(\bm{x}_i,\bm{x}_j)$. +Given a kernel $K$ and the targets $y_i$ this matrix is easy to set up. The constraint $\bm{y}^T\bm{\lambda}=0$ leads to $f=0$ and $\bm{A}=\bm{y}$. How to set up the matrix $\bm{G}$ is discussed later. Here note that the inequalities $0\leq \lambda_i \leq C$ can be split up into +$0\leq \lambda_i$ and $\lambda_i \leq C$. These two inequalities define then the matrix $\bm{G}$ and the vector $\bm{h}$. + + +!split +===== Different kernels and Mercer's theorem ===== + +There are several popular kernels being used. These are +o Linear: $K(\bm{x},\bm{y})=\bm{x}^T\bm{y}$, +o Polynomial: $K(\bm{x},\bm{y})=(\bm{x}^T\bm{y}+\gamma)^d$, +o Gaussian Radial Basis Function: $K(\bm{x},\bm{y})=\exp{\left(-\gamma\vert\vert\bm{x}-\bm{y}\vert\vert^2\right)}$, +o Tanh: $K(\bm{x},\bm{y})=\tanh{(\bm{x}^T\bm{y}+\gamma)}$, +and many other ones. + +An important theorem for us is "Mercer's +theorem":"https://en.wikipedia.org/wiki/Mercer%27s_theorem". The +theorem states that if a kernel function $K$ is symmetric, continuous +and leads to a positive semi-definite matrix $\bm{P}$ then there +exists a function $\phi$ that maps $\bm{x}_i$ and $\bm{x}_j$ into +another space (possibly with much higher dimensions) such that + +!bt +\[ +K(\bm{x}_i,\bm{x}_j)=\phi(\bm{x}_i)^T\phi(\bm{x}_j). +\] +!et + +So you can use $K$ as a kernel since you know $\phi$ exists, even if +you don’t know what $\phi$ is. + +Note that some frequently used kernels (such as the Sigmoid kernel) +don’t respect all of Mercer’s conditions, yet they generally work well +in practice. + + +!split +===== The moons example ("Adapted from Geron, chapter 5":"https://www.oreilly.com/library/view/hands-on-machine-learning/9781492032632/") ===== +!bc pycod +from __future__ import division, print_function, unicode_literals + +import numpy as np +np.random.seed(42) + +import matplotlib +import matplotlib.pyplot as plt +plt.rcParams['axes.labelsize'] = 14 +plt.rcParams['xtick.labelsize'] = 12 +plt.rcParams['ytick.labelsize'] = 12 + + +from sklearn.svm import SVC +from sklearn import datasets + + + +from sklearn.pipeline import Pipeline +from sklearn.preprocessing import StandardScaler +from sklearn.svm import LinearSVC + + +from sklearn.datasets import make_moons +X, y = make_moons(n_samples=100, noise=0.15, random_state=42) + +def plot_dataset(X, y, axes): + plt.plot(X[:, 0][y==0], X[:, 1][y==0], "bs") + plt.plot(X[:, 0][y==1], X[:, 1][y==1], "g^") + plt.axis(axes) + plt.grid(True, which='both') + plt.xlabel(r"$x_1$", fontsize=20) + plt.ylabel(r"$x_2$", fontsize=20, rotation=0) + +plot_dataset(X, y, [-1.5, 2.5, -1, 1.5]) +plt.show() + +from sklearn.datasets import make_moons +from sklearn.pipeline import Pipeline +from sklearn.preprocessing import PolynomialFeatures + +polynomial_svm_clf = Pipeline([ + ("poly_features", PolynomialFeatures(degree=3)), + ("scaler", StandardScaler()), + ("svm_clf", LinearSVC(C=10, loss="hinge", random_state=42)) + ]) + +polynomial_svm_clf.fit(X, y) + +def plot_predictions(clf, axes): + x0s = np.linspace(axes[0], axes[1], 100) + x1s = np.linspace(axes[2], axes[3], 100) + x0, x1 = np.meshgrid(x0s, x1s) + X = np.c_[x0.ravel(), x1.ravel()] + y_pred = clf.predict(X).reshape(x0.shape) + y_decision = clf.decision_function(X).reshape(x0.shape) + plt.contourf(x0, x1, y_pred, cmap=plt.cm.brg, alpha=0.2) + plt.contourf(x0, x1, y_decision, cmap=plt.cm.brg, alpha=0.1) + +plot_predictions(polynomial_svm_clf, [-1.5, 2.5, -1, 1.5]) +plot_dataset(X, y, [-1.5, 2.5, -1, 1.5]) + +plt.show() + + +from sklearn.svm import SVC + +poly_kernel_svm_clf = Pipeline([ + ("scaler", StandardScaler()), + ("svm_clf", SVC(kernel="poly", degree=3, coef0=1, C=5)) + ]) +poly_kernel_svm_clf.fit(X, y) + +poly100_kernel_svm_clf = Pipeline([ + ("scaler", StandardScaler()), + ("svm_clf", SVC(kernel="poly", degree=10, coef0=100, C=5)) + ]) +poly100_kernel_svm_clf.fit(X, y) + +plt.figure(figsize=(11, 4)) + +plt.subplot(121) +plot_predictions(poly_kernel_svm_clf, [-1.5, 2.5, -1, 1.5]) +plot_dataset(X, y, [-1.5, 2.5, -1, 1.5]) +plt.title(r"$d=3, r=1, C=5$", fontsize=18) + +plt.subplot(122) +plot_predictions(poly100_kernel_svm_clf, [-1.5, 2.5, -1, 1.5]) +plot_dataset(X, y, [-1.5, 2.5, -1, 1.5]) +plt.title(r"$d=10, r=100, C=5$", fontsize=18) + +plt.show() + +def gaussian_rbf(x, landmark, gamma): + return np.exp(-gamma * np.linalg.norm(x - landmark, axis=1)**2) + +gamma = 0.3 + +x1s = np.linspace(-4.5, 4.5, 200).reshape(-1, 1) +x2s = gaussian_rbf(x1s, -2, gamma) +x3s = gaussian_rbf(x1s, 1, gamma) + +XK = np.c_[gaussian_rbf(X1D, -2, gamma), gaussian_rbf(X1D, 1, gamma)] +yk = np.array([0, 0, 1, 1, 1, 1, 1, 0, 0]) + +plt.figure(figsize=(11, 4)) + +plt.subplot(121) +plt.grid(True, which='both') +plt.axhline(y=0, color='k') +plt.scatter(x=[-2, 1], y=[0, 0], s=150, alpha=0.5, c="red") +plt.plot(X1D[:, 0][yk==0], np.zeros(4), "bs") +plt.plot(X1D[:, 0][yk==1], np.zeros(5), "g^") +plt.plot(x1s, x2s, "g--") +plt.plot(x1s, x3s, "b:") +plt.gca().get_yaxis().set_ticks([0, 0.25, 0.5, 0.75, 1]) +plt.xlabel(r"$x_1$", fontsize=20) +plt.ylabel(r"Similarity", fontsize=14) +plt.annotate(r'$\mathbf{x}$', + xy=(X1D[3, 0], 0), + xytext=(-0.5, 0.20), + ha="center", + arrowprops=dict(facecolor='black', shrink=0.1), + fontsize=18, + ) +plt.text(-2, 0.9, "$x_2$", ha="center", fontsize=20) +plt.text(1, 0.9, "$x_3$", ha="center", fontsize=20) +plt.axis([-4.5, 4.5, -0.1, 1.1]) + +plt.subplot(122) +plt.grid(True, which='both') +plt.axhline(y=0, color='k') +plt.axvline(x=0, color='k') +plt.plot(XK[:, 0][yk==0], XK[:, 1][yk==0], "bs") +plt.plot(XK[:, 0][yk==1], XK[:, 1][yk==1], "g^") +plt.xlabel(r"$x_2$", fontsize=20) +plt.ylabel(r"$x_3$ ", fontsize=20, rotation=0) +plt.annotate(r'$\phi\left(\mathbf{x}\right)$', + xy=(XK[3, 0], XK[3, 1]), + xytext=(0.65, 0.50), + ha="center", + arrowprops=dict(facecolor='black', shrink=0.1), + fontsize=18, + ) +plt.plot([-0.1, 1.1], [0.57, -0.1], "r--", linewidth=3) +plt.axis([-0.1, 1.1, -0.1, 1.1]) + +plt.subplots_adjust(right=1) + +plt.show() + + +x1_example = X1D[3, 0] +for landmark in (-2, 1): + k = gaussian_rbf(np.array([[x1_example]]), np.array([[landmark]]), gamma) + print("Phi({}, {}) = {}".format(x1_example, landmark, k)) + +rbf_kernel_svm_clf = Pipeline([ + ("scaler", StandardScaler()), + ("svm_clf", SVC(kernel="rbf", gamma=5, C=0.001)) + ]) +rbf_kernel_svm_clf.fit(X, y) + + +from sklearn.svm import SVC + +gamma1, gamma2 = 0.1, 5 +C1, C2 = 0.001, 1000 +hyperparams = (gamma1, C1), (gamma1, C2), (gamma2, C1), (gamma2, C2) + +svm_clfs = [] +for gamma, C in hyperparams: + rbf_kernel_svm_clf = Pipeline([ + ("scaler", StandardScaler()), + ("svm_clf", SVC(kernel="rbf", gamma=gamma, C=C)) + ]) + rbf_kernel_svm_clf.fit(X, y) + svm_clfs.append(rbf_kernel_svm_clf) + +plt.figure(figsize=(11, 7)) + +for i, svm_clf in enumerate(svm_clfs): + plt.subplot(221 + i) + plot_predictions(svm_clf, [-1.5, 2.5, -1, 1.5]) + plot_dataset(X, y, [-1.5, 2.5, -1, 1.5]) + gamma, C = hyperparams[i] + plt.title(r"$\gamma = {}, C = {}$".format(gamma, C), fontsize=16) + +plt.show() + +!ec + + + +!split +===== Mathematical optimization of convex functions ===== + +A mathematical (quadratic) optimization problem, or just optimization problem, has the form +!bt +\begin{align*} + &\mathrm{min}_{\lambda}\hspace{0.2cm} \frac{1}{2}\bm{\lambda}^T\bm{P}\bm{\lambda}+\bm{q}^T\bm{\lambda},\\ \nonumber + &\mathrm{subject\hspace{0.1cm}to} \hspace{0.2cm} \bm{G}\bm{\lambda} \preceq \bm{h} \wedge \bm{A}\bm{\lambda}=f. +\end{align*} +!et +subject to some constraints for say a selected set $i=1,2,\dots, n$. +In our case we are optimizing with respect to the Lagrangian multipliers $\lambda_i$, and the +vector $\bm{\lambda}=[\lambda_1, \lambda_2,\dots, \lambda_n]$ is the optimization variable we are dealing with. + +In our case we are particularly interested in a class of optimization problems called convex optmization problems. +In our discussion on gradient descent methods we discussed at length the definition of a convex function. + +Convex optimization problems play a central role in applied mathematics and we recommend strongly "Boyd and Vandenberghe's text on the topics":"http://web.stanford.edu/~boyd/cvxbook/". + + + +!split +===== How do we solve these problems? ===== + +If we use Python as programming language and wish to venture beyond +_scikit-learn_, _tensorflow_ and similar software which makes our +lives so much easier, we need to dive into the wonderful world of +quadratic programming. We can, if we wish, solve the minimization +problem using standard gradient methods or conjugate gradient +methods. However, these methods tend to exhibit a rather slow +converge. So, welcome to the promised land of quadratic programming. + +The functions we need are contained in the quadratic programming package _CVXOPT_ and we need to import it together with _numpy_ as + +!bc pycod +import numpy +import cvxopt +!ec + +This will make our life much easier. You don't need to write your own optimizer. + + +!split +===== A simple example ===== + +We remind ourselves about the general problem we want to solve +!bt +\begin{align*} + &\mathrm{min}_{x}\hspace{0.2cm} \frac{1}{2}\bm{x}^T\bm{P}\bm{x}+\bm{q}^T\bm{x},\\ \nonumber + &\mathrm{s.t.} \hspace{0.2cm} \bm{G}\bm{x} \preceq \bm{h} \wedge \bm{A}\bm{x}=f. +\end{align*} +!et +_Note_: we use _s.t._ for subject to. + +Let us show how to perform the optmization using a simple case. Assume we want to optimize the following problem +!bt +\begin{align*} + &\mathrm{min}_{x}\hspace{0.2cm} \frac{1}{2}x^2+5x+3y \\ \nonumber + &\mathrm{s.t.} \\ \nonumber + &x, y \geq 0 \\ \nonumber + &x+3y \geq 15 \\ \nonumber + &2x+5y \leq 100 \\ \nonumber + &3x+4y \leq 80. \\ \nonumber +\end{align*} +!et +The minimization problem can be rewritten in terms of vectors and matrices as (with $x$ and $y$ being the unknowns) +!bt +\[ +\frac{1}{2}\begin{bmatrix} x\\ y \end{bmatrix}^T \begin{bmatrix} 1 & 0\\ 0 & 0 \end{bmatrix} \begin{bmatrix} x \\ y \end{bmatrix} + \begin{bmatrix}5\\ 3 \end{bmatrix}^T \begin{bmatrix}x \\ y \end{bmatrix}. +\] +!et +Similarly, we can now set up the inequalities (we need to change $\geq$ to $\leq$ by multiplying with $-1$ on bot sides) as the following matrix-vector equation +!bt +\[ +\begin{bmatrix} -1 & 0 \\ 0 & -1 \\ -1 & -3 \\ 2 & 5 \\ 3 & 4\end{bmatrix}\begin{bmatrix} x \\ y\end{bmatrix} \preceq \begin{bmatrix}0 \\ 0\\ -15 \\ 100 \\ 80\end{bmatrix}. +\] +!et +We have collapsed all the inequalities into a single matrix $\bm{G}$. We see also that our matrix +!bt +\[ +\bm{P} =\begin{bmatrix} 1 & 0\\ 0 & 0 \end{bmatrix} +\] +!et +is clearly positive semi-definite (all eigenvalues larger or equal zero). +Finally, the vector $\bm{h}$ is defined as +!bt +\[ +\bm{h} = \begin{bmatrix}0 \\ 0\\ -15 \\ 100 \\ 80\end{bmatrix}. +\] +!et + + +Since we don't have any equalities the matrix $\bm{A}$ is set to zero +The following code solves the equations for us +!bc pycod +# Import the necessary packages +import numpy +from cvxopt import matrix +from cvxopt import solvers +P = matrix(numpy.diag([1,0]), tc='d') +q = matrix(numpy.array([3,4]), tc='d') +G = matrix(numpy.array([[-1,0],[0,-1],[-1,-3],[2,5],[3,4]]), tc='d') +h = matrix(numpy.array([0,0,-15,100,80]), tc='d') +# Construct the QP, invoke solver +sol = solvers.qp(P,q,G,h) +# Extract optimal value and solution +sol['x'] +sol['primal objective'] +!ec + +!split +===== Back to the more realistic cases ===== + +We are now ready to return to our setup of the optmization problem for a more realistic case. Introducing the _slack_ parameter $C$ we have +!bt +\[ +\frac{1}{2} \bm{\lambda}^T\begin{bmatrix} y_1y_1K(\bm{x}_1,\bm{x}_1) & y_1y_2K(\bm{x}_1,\bm{x}_2) & \dots & \dots & y_1y_nK(\bm{x}_1,\bm{x}_n) \\ +y_2y_1K(\bm{x}_2,\bm{x}_1) & y_2y_2K(\bm{x}_2,\bm{x}_2) & \dots & \dots & y_1y_nK(\bm{x}_2,\bm{x}_n) \\ +\dots & \dots & \dots & \dots & \dots \\ +\dots & \dots & \dots & \dots & \dots \\ +y_ny_1K(\bm{x}_n,\bm{x}_1) & y_ny_2K(\bm{x}_n\bm{x}_2) & \dots & \dots & y_ny_nK(\bm{x}_n,\bm{x}_n) \\ +\end{bmatrix}\bm{\lambda}-\mathbb{I}\bm{\lambda}, +\] +!et +subject to $\bm{y}^T\bm{\lambda}=0$. Here we defined the vectors $\bm{\lambda} =[\lambda_1,\lambda_2,\dots,\lambda_n]$ and +$\bm{y}=[y_1,y_2,\dots,y_n]$. +With the slack constants this leads to the additional constraint $0\leq \lambda_i \leq C$. + + +!split +===== Setting up the matrices and the problem ===== + +We have the general problem +!bt +\begin{align*} + &\mathrm{min}_{\lambda}\hspace{0.2cm} \frac{1}{2}\bm{\lambda}^T\bm{P}\bm{\lambda}+\bm{q}^T\bm{\lambda},\\ \nonumber + &\mathrm{s.t.} \hspace{0.2cm} \bm{G}\bm{\lambda} \preceq \bm{h} \wedge \bm{A}\bm{\lambda}=f. +\end{align*} +!et + + + + +o With a given kernel we can thus define the matrix $\bm{P}$. +o The matrix $\bm{P}$ has matrix elements $p_{ij}=y_iy_jK(\bm{x}_i,\bm{x}_j)$. Given a kernel $K$ and the targets $y_i$ this matrix is easy to set up. +o The vector $\bm{q}$ has all elements equal to $-1$. +o The constraint $\bm{y}^T\bm{\lambda}=0$ leads to $f=0$ and $\bm{A}=\bm{y}$. +o To set up the matrix $\bm{G}$ we note that the inequalities $0\leq \lambda_i \leq C$ can be split up into $0\leq \lambda_i$ and $\lambda_i \leq C$. These two inequalities define then the matrix $\bm{G}$ and the vector $\bm{h}$. + +!split +===== Setting up $\bm{G}\bm{\lambda} \preceq \bm{h}$ ===== + +We have two constraints, $0\le \lambda_i$ and $\lambda_i \le C$. To do this we multiply the ones with the constraint +$\ge$ with $-1$ in order to get $\le$. It means that the problem $\bm{G}\bm{\lambda} \preceq \bm{h}$ +can be written as +!bt +\[ +\begin{bmatrix} -1& 0 & 0 & \dots & 0 \\ +0& -1 & 0 & \dots & 0 \\ +0& 0 & -1 & \dots & 0 \\ +\dots& \dots & \dots & \dots & \dots \\ +0& 0 & 0 & \dots & -1 \\ +1 & 0 & 0 & \dots & 0 \\ +0& 1 & 0 & \dots & 0 \\ +0& 0 & 1 & \dots & 0 \\ +\dots& \dots & \dots & \dots & \dots \\ +0& 0 & 0 & \dots & 1 \\ +\end{bmatrix} +\begin{bmatrix} \lambda_1 \\ +\lambda_2 \\ +\lambda_3 \\ +\dots \\ +\lambda_n \\ +\end{bmatrix} \preceq +\begin{bmatrix} 0 \\ +0 \\ +0 \\ +\dots \\ +0 \\ +C \\ +C \\ +C \\ +\dots \\ +C \\ +\end{bmatrix}. +\] +!et + +And then we are ready to go. + +!split +===== SVMs and Regression and multiclass classification ===== + +More text will be added here. See also "Hastie et al.":"https://www.springer.com/gp/book/9780387848570" section 12.3. + +!split +===== Summary of course ===== + +!split +===== What? Me worry? No final exam in this course! ===== +FIGURE: [figures/exam1.jpeg, width=500 frac=0.6] + + + +!split +===== Topics we have covered this year ===== + +The course has two central parts + +o Statistical analysis and optimization of data +o Machine learning + +!split +===== Statistical analysis and optimization of data ===== + +The following topics have been discussed: +o Basic concepts, expectation values, variance, covariance, correlation functions and errors; +o Simpler models, binomial distribution, the Poisson distribution, simple and multivariate normal distributions; +o Central elements from linear algebra, matrix inversion and SVD +o Gradient methods for data optimization +o Estimation of errors using cross-validation, bootstrapping and jackknife methods; +o Practical optimization using Singular-value decomposition and least squares for parameterizing data. +o Principal Component Analysis to reduce the number of features. + +!split +===== Machine learning ===== + +The following topics will be covered +o Linear methods for regression and classification: + o Ordinary Least Squares + o Ridge regression + o Lasso regression + o Logistic regression +o Neural networks and deep learning: + o Feed Forward Neural Networks + o Convolutional Neural Networks + o Recurrent Neural Networks +o Decisions trees and ensemble methods: + o Decision trees + o Bagging and voting + o Random forests + o Boosting and gradient boosting +o Support vector machines + o Binary classification and multiclass classification + o Kernel methods + o Regression + + +!split +===== Learning outcomes and overarching aims of this course ===== + +The course introduces a variety of central algorithms and methods +essential for studies of data analysis and machine learning. The +course is project based and through the various projects, normally +three, you will be exposed to fundamental research problems +in these fields, with the aim to reproduce state of the art scientific +results. The students will learn to develop and structure large codes +for studying these systems, get acquainted with computing facilities +and learn to handle large scientific projects. A good scientific and +ethical conduct is emphasized throughout the course. + +* Understand linear methods for regression and classification; +* Learn about neural network; +* Learn about bagging, boosting and trees +* Support vector machines +* Learn about basic data analysis; +* Be capable of extending the acquired knowledge to other systems and cases; +* Have an understanding of central algorithms used in data analysis and machine learning; +* Work on numerical projects to illustrate the theory. The projects play a central role and you are expected to know modern programming languages like Python or C++. + + + +!split +===== Perspective on Machine Learning ===== + +o Rapidly emerging application area +o Experiment AND theory are evolving in many many fields. Still many low-hanging fruits. +o Requires education/retraining for more widespread adoption +o A lot of “word-of-mouth” development methods + +Huge amounts of data sets require automation, classical analysis tools often inadequate. +High energy physics hit this wall in the 90’s. +In 2009 single top quark production was determined via "Boosted decision trees, Bayesian +Neural Networks, etc.":"https://arxiv.org/pdf/0903.0850.pdf" + + +!split +===== Machine Learning Research ===== + +Where to find recent results: +o Conference proceedings, arXiv and blog posts! +o _NIPS_: "Neural Information Processing Systems":"https://papers.nips.cc" +o _ICLR_: "International Conference on Learning Representations":"https://openreview.net/group?id=ICLR.cc/2018/Conference#accepted-oral-papers" +o _ICML_: International Conference on Machine Learning +o "Journal of Machine Learning Research":"http://www.jmlr.org/papers/v19/" +o "Follow ML on ArXiv":"https://arxiv.org/list/cs.LG/recent" + + +!split +===== Starting your Machine Learning Project ===== + +o Identify problem type: classification, regression +o Consider your data carefully +o Choose a simple model that fits 1. and 2. +o Consider your data carefully again! Think of data representation more carefully. +o Based on your results, feedback loop to earliest possible point + + + + +!split +===== Choose a Model and Algorithm ===== + +o Supervised? +o Start with the simplest model that fits your problem +o Start with minimal processing of data + +!split +===== Preparing Your Data ===== + +o Shuffle your data +o Mean center your data + * Why? +o Normalize the variance + * Why? +o _Whitening_ + * Decorrelates data + * Can be hit or miss +o When to do train/test split? + + +!split +===== Which Activation and Weights to Choose in Neural Networks ===== + +o RELU? ELU? +o Sigmoid or Tanh? +o Set all weights to 0? + * Terrible idea +o Set all weights to random values? + * Small random values + + +!split +===== Optimization Methods and Hyperparameters ===== +o Stochastic gradient descent +o Stochastic gradient descent + momentum +o State-of-the-art approaches: + * RMSProp + * Adam + * and more + +Which regularization and hyperparameters? $L_1$ or $L_2$, soft +classifiers, depths of trees and many other. Need to explore a large +set of hyperparameters and regularization methods. + + +!split +===== Resampling ===== + +When do we resample? + +o "Bootstrap":"https://www.cambridge.org/core/books/bootstrap-methods-and-their-application/ED2FD043579F27952363566DC09CBD6A" +o "Cross-validation":"https://www.youtube.com/watch?v=fSytzGwwBVw&ab_channel=StatQuestwithJoshStarmer" +o Jackknife and many other + + +!split +===== Other courses on Data science and Machine Learning at UiO ===== + +The link here URL:"https://www.mn.uio.no/english/research/about/centre-focus/innovation/data-science/studies/" gives an excellent overview of courses on Machine learning at UiO. + +o "STK2100 Machine learning and statistical methods for prediction and classification":"http://www.uio.no/studier/emner/matnat/math/STK2100/index-eng.html". +o "IN3050/IN4050 Introduction to Artificial Intelligence and Machine Learning":"https://www.uio.no/studier/emner/matnat/ifi/IN3050/index-eng.html". Introductory course in machine learning and AI with an algorithmic approach. +o "STK-INF3000/4000 Selected Topics in Data Science":"http://www.uio.no/studier/emner/matnat/math/STK-INF3000/index-eng.html". The course provides insight into selected contemporary relevant topics within Data Science. +o "IN4080 Natural Language Processing":"https://www.uio.no/studier/emner/matnat/ifi/IN4080/index.html". Probabilistic and machine learning techniques applied to natural language processing. +o "STK-IN4300 – Statistical learning methods in Data Science":"https://www.uio.no/studier/emner/matnat/math/STK-IN4300/index-eng.html". An advanced introduction to statistical and machine learning. For students with a good mathematics and statistics background. +o "IN-STK5000 Adaptive Methods for Data-Based Decision Making":"https://www.uio.no/studier/emner/matnat/ifi/IN-STK5000/index-eng.html". Methods for adaptive collection and processing of data based on machine learning techniques. +o "IN5400/INF5860 – Machine Learning for Image Analysis":"https://www.uio.no/studier/emner/matnat/ifi/IN5400/". An introduction to deep learning with particular emphasis on applications within Image analysis, but useful for other application areas too. +o "TEK5040 – Dyp læring for autonome systemer":"https://www.uio.no/studier/emner/matnat/its/TEK5040/". The course addresses advanced algorithms and architectures for deep learning with neural networks. The course provides an introduction to how deep-learning techniques can be used in the construction of key parts of advanced autonomous systems that exist in physical environments and cyber environments. + +!split +===== Additional courses of interest ===== + +o "STK4051 Computational Statistics":"https://www.uio.no/studier/emner/matnat/math/STK4051/index-eng.html" +o "STK4021 Applied Bayesian Analysis and Numerical Methods":"https://www.uio.no/studier/emner/matnat/math/STK4021/index-eng.html" + +!split +===== What's the future like? ===== + +Based on multi-layer nonlinear neural networks, deep learning can +learn directly from raw data, automatically extract and abstract +features from layer to layer, and then achieve the goal of regression, +classification, or ranking. Deep learning has made breakthroughs in +computer vision, speech processing and natural language, and reached +or even surpassed human level. The success of deep learning is mainly +due to the three factors: big data, big model, and big computing. + +In the past few decades, many different architectures of deep neural +networks have been proposed, such as +o Convolutional neural networks, which are mostly used in image and video data processing, and have also been applied to sequential data such as text processing; +o Recurrent neural networks, which can process sequential data of variable length and have been widely used in natural language understanding and speech processing; +o Encoder-decoder framework, which is mostly used for image or sequence generation, such as machine translation, text summarization, and image captioning. + +!split +===== Types of Machine Learning, a repetition ===== + +!bblock +The approaches to machine learning are many, but are often split into two main categories. +In *supervised learning* we know the answer to a problem, +and let the computer deduce the logic behind it. On the other hand, *unsupervised learning* +is a method for finding patterns and relationship in data sets without any prior knowledge of the system. +Some authours also operate with a third category, namely *reinforcement learning*. This is a paradigm +of learning inspired by behavioural psychology, where learning is achieved by trial-and-error, +solely from rewards and punishment. + +Another way to categorize machine learning tasks is to consider the desired output of a system. +Some of the most common tasks are: + + * Classification: Outputs are divided into two or more classes. The goal is to produce a model that assigns inputs into one of these classes. An example is to identify digits based on pictures of hand-written ones. Classification is typically supervised learning. + + * Regression: Finding a functional relationship between an input data set and a reference data set. The goal is to construct a function that maps input data to continuous output values. + + * Clustering: Data are divided into groups with certain common traits, without knowing the different groups beforehand. It is thus a form of unsupervised learning. + + * Other unsupervised learning algortihms like _Boltzmann machines_ +!eblock + + +!split +===== Why Boltzmann machines? ===== + +What is known as restricted Boltzmann Machines (RMB) have received a lot of attention lately. +One of the major reasons is that they can be stacked layer-wise to build deep neural networks that capture complicated statistics. + +The original RBMs had just one visible layer and a hidden layer, but recently so-called Gaussian-binary RBMs have gained quite some popularity in imaging since they are capable of modeling continuous data that are common to natural images. + +Furthermore, they have been used to solve complicated "quantum mechanical many-particle problems or classical statistical physics problems like the Ising and Potts classes of models":"https://journals.aps.org/rmp/abstract/10.1103/RevModPhys.91.045002". + + +!split +===== Boltzmann Machines ===== + +Why use a generative model rather than the more well known discriminative deep neural networks (DNN)? + +* Discriminitave methods have several limitations: They are mainly supervised learning methods, thus requiring labeled data. And there are tasks they cannot accomplish, like drawing new examples from an unknown probability distribution. + +* A generative model can learn to represent and sample from a probability distribution. The core idea is to learn a parametric model of the probability distribution from which the training data was drawn. As an example + + o A model for images could learn to draw new examples of cats and dogs, given a training dataset of images of cats and dogs. + o Generate a sample of an ordered or disordered phase, having been given samples of such phases. + o Model the trial function for "Monte Carlo calculations":"https://journals.aps.org/rmp/abstract/10.1103/RevModPhys.91.045002". + + +!split +===== Some similarities and differences from DNNs ===== + + o Both use gradient-descent based learning procedures for minimizing cost functions + o Energy based models don't use backpropagation and automatic differentiation for computing gradients, instead turning to Markov Chain Monte Carlo methods. + o DNNs often have several hidden layers. A restricted Boltzmann machine has only one hidden layer, however several RBMs can be stacked to make up Deep Belief Networks, of which they constitute the building blocks. + +History: The RBM was developed by amongst others "Geoffrey Hinton":"https://en.wikipedia.org/wiki/Geoffrey_Hinton", called by some the "Godfather of Deep Learning", working with the University of Toronto and Google. + + +!split +===== Boltzmann machines (BM) ===== + +!bblock +A BM is what we would call an undirected probabilistic graphical model +with stochastic continuous or discrete units. +!eblock +!bblock +It is interpreted as a stochastic recurrent neural network where the +state of each unit(neurons/nodes) depends on the units it is connected +to. The weights in the network represent thus the strength of the +interaction between various units/nodes. +!eblock +!bblock +It turns into a Hopfield network if we choose deterministic rather +than stochastic units. In contrast to a Hopfield network, a BM is a +so-called generative model. It allows us to generate new samples from +the learned distribution. +!eblock + +!split +===== A standard BM setup ===== + +!bblock +A standard BM network is divided into a set of observable and visible units $\hat{x}$ and a set of unknown hidden units/nodes $\hat{h}$. +!eblock + +!bblock +Additionally there can be bias nodes for the hidden and visible layers. These biases are normally set to $1$. +!eblock + +!bblock +BMs are stackable, meaning they cwe can train a BM which serves as input to another BM. We can construct deep networks for learning complex PDFs. The layers can be trained one after another, a feature which makes them popular in deep learning +!eblock + +However, they are often hard to train. This leads to the introduction of so-called restricted BMs, or RBMS. +Here we take away all lateral connections between nodes in the visible layer as well as connections between nodes in the hidden layer. The network is illustrated in the figure below. + + + +!split +===== The structure of the RBM network ===== + +FIGURE: [figures/RBM.pdf, width=800 frac=1.0] + + + +!split +===== The network ===== + +_The network layers_: + o A function $\mathbf{x}$ that represents the visible layer, a vector of $M$ elements (nodes). This layer represents both what the RBM might be given as training input, and what we want it to be able to reconstruct. This might for example be given by the pixels of an image or coefficients representing speech, or the coordinates of a quantum mechanical state function. + o The function $\mathbf{h}$ represents the hidden, or latent, layer. A vector of $N$ elements (nodes). Also called "feature detectors". + +!split +===== Goals ===== + +The goal of the hidden layer is to increase the model's expressive +power. We encode complex interactions between visible variables by +introducing additional, hidden variables that interact with visible +degrees of freedom in a simple manner, yet still reproduce the complex +correlations between visible degrees in the data once marginalized +over (integrated out). + +_The network parameters, to be optimized/learned_: + o $\mathbf{a}$ represents the visible bias, a vector of same length as $\mathbf{x}$. + o $\mathbf{b}$ represents the hidden bias, a vector of same lenght as $\mathbf{h}$. + o $W$ represents the interaction weights, a matrix of size $M\times N$. + + + + +!split +===== Joint distribution ===== + +The restricted Boltzmann machine is described by a Boltzmann distribution +!bt +\begin{align} + P_{rbm}(\mathbf{x},\mathbf{h}) = \frac{1}{Z} e^{-\frac{1}{T_0}E(\mathbf{x},\mathbf{h})}, +\end{align} +!et +where $Z$ is the normalization constant or partition function, defined as +!bt +\begin{align} + Z = \int \int e^{-\frac{1}{T_0}E(\mathbf{x},\mathbf{h})} d\mathbf{x} d\mathbf{h}. +\end{align} +!et +It is common to ignore $T_0$ by setting it to one. + + + + +!split +===== Network Elements, the energy function ===== + +The function $E(\mathbf{x},\mathbf{h})$ gives the _energy_ of a +configuration (pair of vectors) $(\mathbf{x}, \mathbf{h})$. The lower +the energy of a configuration, the higher the probability of it. This +function also depends on the parameters $\mathbf{a}$, $\mathbf{b}$ and +$W$. Thus, when we adjust them during the learning procedure, we are +adjusting the energy function to best fit our problem. + +An expression for the energy function is +!bt +\[ +E(\hat{x},\hat{h}) = -\sum_{ia}^{NA}b_i^a \alpha_i^a(x_i)-\sum_{jd}^{MD}c_j^d \beta_j^d(h_j)-\sum_{ijad}^{NAMD}b_i^a \alpha_i^a(x_i)c_j^d \beta_j^d(h_j)w_{ij}^{ad}. +\] +!et + +Here $\beta_j^d(h_j)$ and $\alpha_i^a(x_j)$ are so-called transfer functions that map a given input value to a desired feature value. The labels $a$ and $d$ denote that there can be multiple transfer functions per variable. The first sum depends only on the visible units. The second on the hidden ones. _Note_ that there is no connection between nodes in a layer. + +The quantities $b$ and $c$ can be interpreted as the visible and hidden biases, respectively. + +The connection between the nodes in the two layers is given by the weights $w_{ij}$. + +!split +===== Defining different types of RBMs ===== +There are different variants of RBMs, and the differences lie in the types of visible and hidden units we choose as well as in the implementation of the energy function $E(\mathbf{x},\mathbf{h})$. + +!bblock Binary-Binary RBM: + +RBMs were first developed using binary units in both the visible and hidden layer. The corresponding energy function is defined as follows: +!bt +\begin{align} + E(\mathbf{x}, \mathbf{h}) = - \sum_i^M x_i a_i- \sum_j^N b_j h_j - \sum_{i,j}^{M,N} x_i w_{ij} h_j, +\end{align} +!et +where the binary values taken on by the nodes are most commonly 0 and 1. +!eblock +!bblock Gaussian-Binary RBM: + +Another varient is the RBM where the visible units are Gaussian while the hidden units remain binary: +!bt +\begin{align} + E(\mathbf{x}, \mathbf{h}) = \sum_i^M \frac{(x_i - a_i)^2}{2\sigma_i^2} - \sum_j^N b_j h_j - \sum_{i,j}^{M,N} \frac{x_i w_{ij} h_j}{\sigma_i^2}. +\end{align} +!et +!eblock + +!split +===== More about RBMs ===== +o Useful when we model continuous data (i.e., we wish $\mathbf{x}$ to be continuous) +o Requires a smaller learning rate, since there's no upper bound to the value a component might take in the reconstruction + +Other types of units include: + o Softmax and multinomial units + o Gaussian visible and hidden units + o Binomial units + o Rectified linear units + +To read more, see "Lectures on Boltzmann machines in Physics":"https://github.com/CompPhysics/ComputationalPhysics2/blob/gh-pages/doc/pub/notebook2/ipynb/notebook2.ipynb". + + +!split +===== Autoencoders: Overarching view ===== + +Autoencoders are artificial neural networks capable of learning +efficient representations of the input data (these representations are called codings) without +any supervision (i.e., the training set is unlabeled). These codings +typically have a much lower dimensionality than the input data, making +autoencoders useful for dimensionality reduction. + +More importantly, autoencoders act as powerful feature detectors, and +they can be used for unsupervised pretraining of deep neural networks. + +Lastly, they are capable of randomly generating new data that looks +very similar to the training data; this is called a generative +model. For example, you could train an autoencoder on pictures of +faces, and it would then be able to generate new faces. Surprisingly, +autoencoders work by simply learning to copy their inputs to their +outputs. This may sound like a trivial task, but we will see that +constraining the network in various ways can make it rather +difficult. For example, you can limit the size of the internal +representation, or you can add noise to the inputs and train the +network to recover the original inputs. These constraints prevent the +autoencoder from trivially copying the inputs directly to the outputs, +which forces it to learn efficient ways of representing the data. In +short, the codings are byproducts of the autoencoder’s attempt to +learn the identity function under some constraints. + +"Video on autoencoders":"https://www.coursera.org/lecture/building-deep-learning-models-with-tensorflow/autoencoders-1U4L3" + +See also A. Geron's textbook, chapter 15. + +!split +===== Bayesian Machine Learning ===== + +This is an important topic if we aim at extracting a probability +distribution. This gives us also a confidence interval and error +estimates. + +Bayesian machine learning allows us to encode our prior beliefs about +what those models should look like, independent of what the data tells +us. This is especially useful when we don’t have a ton of data to +confidently learn our model. + +"Video on Bayesian deep learning":"https://www.youtube.com/watch?v=E1qhGw8QxqY&ab_channel=AndrewGordonWilson" + +See also the "slides here":"https://github.com/CompPhysics/MachineLearning/blob/master/doc/Articles/lec03.pdf". + +!split +===== Reinforcement Learning ===== + +Reinforcement Learning (RL) is one of the most exciting fields of +Machine Learning today, and also one of the oldest. It has been around +since the 1950s, producing many interesting applications over the +years. + +It studies +how agents take actions based on trial and error, so as to maximize +some notion of cumulative reward in a dynamic system or +environment. Due to its generality, the problem has also been studied +in many other disciplines, such as game theory, control theory, +operations research, information theory, multi-agent systems, swarm +intelligence, statistics, and genetic algorithms. + +In March 2016, AlphaGo, a computer program that plays the board game +Go, beat Lee Sedol in a five-game match. This was the first time a +computer Go program had beaten a 9-dan (highest rank) professional +without handicaps. AlphaGo is based on deep convolutional neural +networks and reinforcement learning. AlphaGo’s victory was a major +milestone in artificial intelligence and it has also made +reinforcement learning a hot research area in the field of machine +learning. + +"Lecture on Reinforcement Learning":"https://www.youtube.com/watch?v=FgzM3zpZ55o&ab_channel=stanfordonline". + +See also A. Geron's textbook, chapter 16. + +!split +===== Transfer learning ===== + +The goal of transfer learning is to transfer the model or knowledge +obtained from a source task to the target task, in order to resolve +the issues of insufficient training data in the target task. The +rationality of doing so lies in that usually the source and target +tasks have inter-correlations, and therefore either the features, +samples, or models in the source task might provide useful information +for us to better solve the target task. Transfer learning is a hot +research topic in recent years, with many problems still waiting to be studied. + +"Lecture on transfer learning":"https://www.ias.edu/video/machinelearning/2020/0331-SamoryKpotufe". + +!split +===== Adversarial learning ===== + +The conventional deep generative model has a potential problem: the +model tends to generate extreme instances to maximize the +probabilistic likelihood, which will hurt its performance. Adversarial +learning utilizes the adversarial behaviors (e.g., generating +adversarial instances or training an adversarial model) to enhance the +robustness of the model and improve the quality of the generated +data. In recent years, one of the most promising unsupervised learning +technologies, generative adversarial networks (GAN), has already been +successfully applied to image, speech, and text. + +"Lecture on adversial learning":"https://www.youtube.com/watch?v=CIfsB_EYsVI&ab_channel=StanfordUniversitySchoolofEngineering". + +!split +===== Dual learning ===== + +Dual learning is a new learning paradigm, the basic idea of which is +to use the primal-dual structure between machine learning tasks to +obtain effective feedback/regularization, and guide and strengthen the +learning process, thus reducing the requirement of large-scale labeled +data for deep learning. The idea of dual learning has been applied to +many problems in machine learning, including machine translation, +image style conversion, question answering and generation, image +classification and generation, text classification and generation, +image-to-text, and text-to-image. + +!split +===== Distributed machine learning ===== + +Distributed computation will speed up machine learning algorithms, +significantly improve their efficiency, and thus enlarge their +application. When distributed meets machine learning, more than just +implementing the machine learning algorithms in parallel is required. + + +!split +===== Meta learning ===== + +Meta learning is an emerging research direction in machine +learning. Roughly speaking, meta learning concerns learning how to +learn, and focuses on the understanding and adaptation of the learning +itself, instead of just completing a specific learning task. That is, +a meta learner needs to be able to evaluate its own learning methods +and adjust its own learning methods according to specific learning +tasks. + +!split +===== The Challenges Facing Machine Learning ===== + +While there has been much progress in machine learning, there are also challenges. + +For example, the mainstream machine learning technologies are +black-box approaches, making us concerned about their potential +risks. To tackle this challenge, we may want to make machine learning +more explainable and controllable. As another example, the +computational complexity of machine learning algorithms is usually +very high and we may want to invent lightweight algorithms or +implementations. Furthermore, in many domains such as physics, +chemistry, biology, and social sciences, people usually seek elegantly +simple equations (e.g., the Schrödinger equation) to uncover the +underlying laws behind various phenomena. In the field of machine +learning, can we reveal simple laws instead of designing more complex +models for data fitting? Although there are many challenges, we are +still very optimistic about the future of machine learning. As we look +forward to the future, here are what we think the research hotspots in +the next ten years will be. + +See the article on "Discovery of Physics From Data: Universal Laws and Discrepancies":"https://www.frontiersin.org/articles/10.3389/frai.2020.00025/full" + +!split +===== Explainable machine learning ===== + +Machine learning, especially deep learning, evolves rapidly. The +ability gap between machine and human on many complex cognitive tasks +becomes narrower and narrower. However, we are still in the very early +stage in terms of explaining why those effective models work and how +they work. + +_What is missing: the gap between correlation and causation_. Standard Machine Learning is based on what e have called a frequentist approach. + +Most +machine learning techniques, especially the statistical ones, depend +highly on correlations in data sets to make predictions and analyses. In +contrast, rational humans tend to reply on clear and trustworthy +causality relations obtained via logical reasoning on real and clear +facts. It is one of the core goals of explainable machine learning to +transition from solving problems by data correlation to solving +problems by logical reasoning. + +_Bayesian Machine Learning is one of the exciting research directions in this field_. + +!split +===== Quantum machine learning ===== + +Quantum machine learning is an emerging interdisciplinary research +area at the intersection of quantum computing and machine learning. + +Quantum computers use effects such as quantum coherence and quantum +entanglement to process information, which is fundamentally different +from classical computers. Quantum algorithms have surpassed the best +classical algorithms in several problems (e.g., searching for an +unsorted database, inverting a sparse matrix), which we call quantum +acceleration. + +When quantum computing meets machine learning, it can be a mutually +beneficial and reinforcing process, as it allows us to take advantage +of quantum computing to improve the performance of classical machine +learning algorithms. In addition, we can also use the machine learning +algorithms (on classic computers) to analyze and improve quantum +computing systems. + +"Lecture on Quantum ML":"https://www.youtube.com/watch?v=Xh9pUu3-WxM&ab_channel=InstituteforPure%26AppliedMathematics%28IPAM%29". + + +"Read interview with Maria Schuld on her work on Quantum Machine Learning":"https://physics.aps.org/articles/v13/179?utm_campaign=weekly&utm_medium=email&utm_source=emailalert". See also "her recent textbook":"https://www.springer.com/gp/book/9783319964232". + + +!split +===== Quantum machine learning algorithms based on linear algebra ===== + +Many quantum machine learning algorithms are based on variants of +quantum algorithms for solving linear equations, which can efficiently +solve N-variable linear equations with complexity of O(log2 N) under +certain conditions. The quantum matrix inversion algorithm can +accelerate many machine learning methods, such as least square linear +regression, least square version of support vector machine, Gaussian +process, and more. The training of these algorithms can be simplified +to solve linear equations. The key bottleneck of this type of quantum +machine learning algorithms is data input—that is, how to initialize +the quantum system with the entire data set. Although efficient +data-input algorithms exist for certain situations, how to efficiently +input data into a quantum system is as yet unknown for most cases. + +!split +===== Quantum reinforcement learning ===== + +In quantum reinforcement learning, a quantum agent interacts with the +classical environment to obtain rewards from the environment, so as to +adjust and improve its behavioral strategies. In some cases, it +achieves quantum acceleration by the quantum processing capabilities +of the agent or the possibility of exploring the environment through +quantum superposition. Such algorithms have been proposed in +superconducting circuits and systems of trapped ions. + +!split +===== Quantum deep learning ===== + +Dedicated quantum information processors, such as quantum annealers +and programmable photonic circuits, are well suited for building deep +quantum networks. The simplest deep quantum network is the Boltzmann +machine. The classical Boltzmann machine consists of bits with tunable +interactions and is trained by adjusting the interaction of these bits +so that the distribution of its expression conforms to the statistics +of the data. To quantize the Boltzmann machine, the neural network can +simply be represented as a set of interacting quantum spins that +correspond to an adjustable Ising model. Then, by initializing the +input neurons in the Boltzmann machine to a fixed state and allowing +the system to heat up, we can read out the output qubits to get the +result. + + +!split +===== Social machine learning ===== + +Machine learning aims to imitate how humans +learn. While we have developed successful machine learning algorithms, +until now we have ignored one important fact: humans are social. Each +of us is one part of the total society and it is difficult for us to +live, learn, and improve ourselves, alone and isolated. Therefore, we +should design machines with social properties. Can we let machines +evolve by imitating human society so as to achieve more effective, +intelligent, interpretable “social machine learning”? + +And much more. + + + +!split +===== The last words? ===== + +Early computer scientist Alan Kay said, _The best way to predict the +future is to create it_. Therefore, all machine learning +practitioners, whether scholars or engineers, professors or students, +need to work together to advance these important research +topics. Together, we will not just predict the future, but create it. + + + + +!split +===== Best wishes to you all and thanks so much for your heroic efforts this semester ===== + +FIGURE: [figures/Nebbdyr2.png, width=500 frac=0.6] + + + + + + diff --git a/doc/src/week48/week48.do.txt b/doc/src/week48/week48.do.txt index 6f178bacc..920ee82b1 100644 --- a/doc/src/week48/week48.do.txt +++ b/doc/src/week48/week48.do.txt @@ -1,637 +1,830 @@ TITLE: Week 48: Support Vector Machines and Summary of course -AUTHOR: Morten Hjorth-Jensen {copyright, 1999-present|CC BY-NC} at Department of Physics, University of Oslo & Department of Physics and Astronomy and National Superconducting Cyclotron Laboratory, Michigan State University +AUTHOR: Morten Hjorth-Jensen {copyright, 1999-present|CC BY-NC} at Department of Physics, University of Oslo, Norway DATE: today !split ===== Overview of week 48 ===== -* _Thursday_: Support Vector Machines: Kernels, Classification and Regression. "Video of Lecture":"https://www.uio.no/studier/emner/matnat/fys/FYS-STK4155/h20/forelesningsvideoer/LectureNovember26.mp4?vrtx=view-as-webpage". -* _Friday_: Summary of course with perspectives for future studies. "Video of Lecture":"https://www.uio.no/studier/emner/matnat/fys/FYS-STK4155/h20/forelesningsvideoer/LectureNovember27.mp4?vrtx=view-as-webpage". +!split +===== Plan for week 47 ===== + +!bblock Lab sessions on Tuesday and Wednesday + * Work and Discussion of project 3 + * Last weekly exercise, +!eblock + +!bblock Plans for the lecture Monday 25 November, with video suggestions etc +o Bossting and gradient boosting and ensemble models +o Summary of course +o Readings and Videos: + o These lecture notes at URL:"https://github.com/CompPhysics/MachineLearning/blob/master/doc/pub/week47/ipynb/week48.ipynb" + o See also lecture notes from week 47 at URL:"https://github.com/CompPhysics/MachineLearning/blob/master/doc/pub/week46/ipynb/week47.ipynb". The lecture on Monday starts with a repetition on AdaBoost before we move over to gradient boosting with examples +# o Video of lecture at URL:"https://youtu.be/RIHzmLv05DA" +# o Whiteboard notes at URL:"https://github.com/CompPhysics/MachineLearning/blob/master/doc/HandWrittenNotes/2024/NotesNovember25.pdf" + o Video on Decision trees URL:"https://www.youtube.com/watch?v=RmajweUFKvM&ab_channel=Simplilearn" + o Video on boosting methods URL:"https://www.youtube.com/watch?v=wPqtzj5VZus&ab_channel=H2O.ai" + o Video on AdaBoost URL:"https://www.youtube.com/watch?v=LsK-xG1cLYA" + o Video on Gradient boost, part 1, parts 2-4 follow thereafter URL:"https://www.youtube.com/watch?v=3CC4N4z3GJc" + o Decision Trees: Rashcka et al chapter 3 pages 86-98, and chapter 7 on Ensemble methods, Voting and Bagging and Gradient Boosting. See also lecture from STK-IN4300, lecture 7 at URL:"https://www.uio.no/studier/emner/matnat/math/STK-IN4300/h20/slides/lecture_7.pdf". +!eblock + -See also Geron's chapter 5. Chapter 12 (sections 12.1-12.3 are the most relevant ones) of Hastie et al contains also a good discussion. !split -===== Thursday ===== +===== Boosting, a Bird's Eye View ===== + +The basic idea is to combine weak classifiers in order to create a good +classifier. With a weak classifier we often intend a classifier which +produces results which are only slightly better than we would get by +random guesses. + +This is done by applying in an iterative way a weak (or a standard +classifier like decision trees) to modify the data. In each iteration +we emphasize those observations which are misclassified by weighting +them with a factor. -We finalize our discussion on Support Vector Machines with an emphasis on kernel transformations and applications to regression. The following "video attempts at giving an overview on this part":"https://www.youtube.com/watch?v=Toet3EiSFcM&ab_channel=StatQuestwithJoshStarmer". See also the "follow-up video":"https://www.youtube.com/watch?v=Qc5IyLW_hns&ab_channel=StatQuestwithJoshStarmer". !split -===== Friday ===== +===== What is boosting? Additive Modelling/Iterative Fitting ===== + +Boosting is a way of fitting an additive expansion in a set of +elementary basis functions like for example some simple polynomials. +Assume for example that we have a function +!bt +\[ +f_M(x) = \sum_{i=1}^M \beta_m b(x;\gamma_m), +\] +!et + +where $\beta_m$ are the expansion parameters to be determined in a +minimization process and $b(x;\gamma_m)$ are some simple functions of +the multivariable parameter $x$ which is characterized by the +parameters $\gamma_m$. + +As an example, consider the Sigmoid function we used in logistic +regression. In that case, we can translate the function +$b(x;\gamma_m)$ into the Sigmoid function + + +!bt +\[ +\sigma(t) = \frac{1}{1+\exp{(-t)}}, +\] +!et + +where $t=\gamma_0+\gamma_1 x$ and the parameters $\gamma_0$ and +$\gamma_1$ were determined by the Logistic Regression fitting +algorithm. + +As another example, consider the cost function we defined for linear regression +!bt +\[ +C(\bm{y},\bm{f}) = \frac{1}{n} \sum_{i=0}^{n-1}(y_i-f(x_i))^2. +\] +!et + +In this case the function $f(x)$ was replaced by the design matrix +$\bm{X}$ and the unknown linear regression parameters $\bm{\beta}$, +that is $\bm{f}=\bm{X}\bm{\beta}$. In linear regression we can +simply invert a matrix and obtain the parameters $\beta$ by + +!bt +\[ +\bm{\beta}=\left(\bm{X}^T\bm{X}\right)^{-1}\bm{X}^T\bm{y}. +\] +!et + +In iterative fitting or additive modeling, we minimize the cost function with respect to the parameters $\beta_m$ and $\gamma_m$. -Friday's lecture is split in two parts. It starts with a summary of -what we have done this semester and continues with perspectives for future studies and -modern research projects in machine learning. !split -===== Support Vector Machines, overarching aims ===== +===== Iterative Fitting, Regression and Squared-error Cost Function ===== -As discussed last week, -a Support Vector Machine (SVM) is a very powerful and versatile -Machine Learning method, capable of performing linear or nonlinear -classification, regression, and even outlier detection. It is one of -the most popular models in Machine Learning, and anyone interested in -Machine Learning should have it in their toolbox. SVMs are -particularly well suited for classification of complex but small-sized or -medium-sized datasets. +The way we proceed is as follows (here we specialize to the squared-error cost function) -The case with two well-separated classes only can be understood in an -intuitive way in terms of lines in a two-dimensional space separating -the two classes. +o Establish a cost function, here ${\cal C}(\bm{y},\bm{f}) = \frac{1}{n} \sum_{i=0}^{n-1}(y_i-f_M(x_i))^2$ with $f_M(x) = \sum_{i=1}^M \beta_m b(x;\gamma_m)$. +o Initialize with a guess $f_0(x)$. It could be one or even zero or some random numbers. +o For $m=1:M$ + o minimize $\sum_{i=0}^{n-1}(y_i-f_{m-1}(x_i)-\beta b(x;\gamma))^2$ wrt $\gamma$ and $\beta$ + o This gives the optimal values $\beta_m$ and $\gamma_m$ + o Determine then the new values $f_m(x)=f_{m-1}(x) +\beta_m b(x;\gamma_m)$ + +We could use any of the algorithms we have discussed till now. If we +use trees, $\gamma$ parameterizes the split variables and split points +at the internal nodes, and the predictions at the terminal nodes. -The basic mathematics behind the SVM is however less familiar to most of us. -It relies on the definition of hyperplanes and the -definition of a _margin_ which separates classes (in case of -classification problems) of variables. It is also used for regression -problems. I recommend you take a look at the lectures from last week on the binary classification problem. !split -===== Kernels and non-linearity ===== +===== Squared-Error Example and Iterative Fitting ===== -The cases we studied last week were all characterized by two classes -with a close to linear separability. The classifiers we have described -so far find linear boundaries in our input feature space. It is -possible to make our procedure more flexible by exploring the feature -space using other basis expansions such as higher-order polynomials, -wavelets, splines etc. +To better understand what happens, let us develop the steps for the iterative fitting using the above squared error function. -If our feature space is not easy to separate, as shown in the figure -generated by the code below, we can achieve a better separation by introducing a more complex -basis functions. The ideal would be, as shown by the code example below, to, -via a specific transformation to obtain a separation between the -classes that is almost linear. See also Figures 12.2 and 12.3 of "Hastie et al.":"https://www.springer.com/gp/book/9780387848570". +For simplicity we assume also that our functions $b(x;\gamma)=1+\gamma x$. -We can achieve a better separation by introducing more complex -basis functions. The ideal would be (see Figures 12.2 and 12.3) to, via a specific transformation to -obtain a separation between the classes which is almost linear. +This means that for every iteration $m$, we need to optimize -The change of basis, from $x\rightarrow z=\phi(x)$ leads to the same -type of equations to be solved, except that we need to introduce, for -example, a polynomial transformation to a two-dimensional training -set. +!bt +\[ +(\beta_m,\gamma_m) = \mathrm{argmin}_{\beta,\lambda}\hspace{0.1cm} \sum_{i=0}^{n-1}(y_i-f_{m-1}(x_i)-\beta b(x;\gamma))^2=\sum_{i=0}^{n-1}(y_i-f_{m-1}(x_i)-\beta(1+\gamma x_i))^2. +\] +!et + +We start our iteration by simply setting $f_0(x)=0$. +Taking the derivatives with respect to $\beta$ and $\gamma$ we obtain +!bt +\[ +\frac{\partial {\cal C}}{\partial \beta} = -2\sum_{i}(1+\gamma x_i)(y_i-\beta(1+\gamma x_i))=0, +\] +!et +and +!bt +\[ +\frac{\partial {\cal C}}{\partial \gamma} =-2\sum_{i}\beta x_i(y_i-\beta(1+\gamma x_i))=0. +\] +!et +We can then rewrite these equations as (defining $\bm{w}=\bm{e}+\gamma \bm{x})$ with $\bm{e}$ being the unit vector) +!bt +\[ +\gamma \bm{w}^T(\bm{y}-\beta\gamma \bm{w})=0, +\] +!et +which gives us $\beta = \bm{w}^T\bm{y}/(\bm{w}^T\bm{w})$. Similarly we have +!bt +\[ +\beta\gamma \bm{x}^T(\bm{y}-\beta(1+\gamma \bm{x}))=0, +\] +!et + +which leads to $\gamma =(\bm{x}^T\bm{y}-\beta\bm{x}^T\bm{e})/(\beta\bm{x}^T\bm{x})$. Inserting +for $\beta$ gives us an equation for $\gamma$. This is a non-linear equation in the unknown $\gamma$ and has to be solved numerically. + +The solution to these two equations gives us in turn $\beta_1$ and $\gamma_1$ leading to the new expression for $f_1(x)$ as +$f_1(x) = \beta_1(1+\gamma_1x)$. Doing this $M$ times results in our final estimate for the function $f$. + + + +!split +===== Iterative Fitting, Classification and AdaBoost ===== + +Let us consider a binary classification problem with two outcomes $y_i \in \{-1,1\}$ and $i=0,1,2,\dots,n-1$ as our set of +observations. We define a classification function $G(x)$ which produces a prediction taking one or the other of the two values +$\{-1,1\}$. + +The error rate of the training sample is then + +!bt +\[ +\mathrm{\overline{err}}=\frac{1}{n} \sum_{i=0}^{n-1} I(y_i\ne G(x_i)). +\] +!et + +The iterative procedure starts with defining a weak classifier whose +error rate is barely better than random guessing. The iterative +procedure in boosting is to sequentially apply a weak +classification algorithm to repeatedly modified versions of the data +producing a sequence of weak classifiers $G_m(x)$. + +Here we will express our function $f(x)$ in terms of $G(x)$. That is +!bt +\[ +f_M(x) = \sum_{i=1}^M \beta_m b(x;\gamma_m), +\] +!et +will be a function of +!bt +\[ +G_M(x) = \mathrm{sign} \sum_{i=1}^M \alpha_m G_m(x). +\] +!et + + + +!split +===== Adaptive Boosting, AdaBoost ===== + +In our iterative procedure we define thus +!bt +\[ +f_m(x) = f_{m-1}(x)+\beta_mG_m(x). +\] +!et + +The simplest possible cost function which leads (also simple from a computational point of view) to the AdaBoost algorithm is the +exponential cost/loss function defined as +!bt +\[ +C(\bm{y},\bm{f}) = \sum_{i=0}^{n-1}\exp{(-y_i(f_{m-1}(x_i)+\beta G(x_i))}. +\] +!et + +We optimize $\beta$ and $G$ for each value of $m=1:M$ as we did in the regression case. +This is normally done in two steps. Let us however first rewrite the cost function as + +!bt +\[ +C(\bm{y},\bm{f}) = \sum_{i=0}^{n-1}w_i^{m}\exp{(-y_i\beta G(x_i))}, +\] +!et +where we have defined $w_i^m= \exp{(-y_if_{m-1}(x_i))}$. + +!split +===== Building up AdaBoost ===== + +First, for any $\beta > 0$, we optimize $G$ by setting +!bt +\[ +G_m(x) = \mathrm{sign} \sum_{i=0}^{n-1} w_i^m I(y_i \ne G_(x_i)), +\] +!et +which is the classifier that minimizes the weighted error rate in predicting $y$. + +We can do this by rewriting +!bt +\[ +\exp{-(\beta)}\sum_{y_i=G(x_i)}w_i^m+\exp{(\beta)}\sum_{y_i\ne G(x_i)}w_i^m, +\] +!et +which can be rewritten as +!bt +\[ +(\exp{(\beta)}-\exp{-(\beta)})\sum_{i=0}^{n-1}w_i^mI(y_i\ne G(x_i))+\exp{(-\beta)}\sum_{i=0}^{n-1}w_i^m=0, +\] +!et +which leads to +!bt +\[ +\beta_m = \frac{1}{2}\log{\frac{1-\mathrm{\overline{err}}}{\mathrm{\overline{err}}}}, +\] +!et +where we have redefined the error as +!bt +\[ +\mathrm{\overline{err}}_m=\frac{1}{n}\frac{\sum_{i=0}^{n-1}w_i^mI(y_i\ne G(x_i)}{\sum_{i=0}^{n-1}w_i^m}, +\] +!et +which leads to an update of +!bt +\[ +f_m(x) = f_{m-1}(x) +\beta_m G_m(x). +\] +!et +This leads to the new weights +!bt +\[ +w_i^{m+1} = w_i^m \exp{(-y_i\beta_m G_m(x_i))} +\] +!et + + +!split +===== Adaptive boosting: AdaBoost, Basic Algorithm ===== + +The algorithm here is rather straightforward. Assume that our weak +classifier is a decision tree and we consider a binary set of outputs +with $y_i \in \{-1,1\}$ and $i=0,1,2,\dots,n-1$ as our set of +observations. Our design matrix is given in terms of the +feature/predictor vectors +$\bm{X}=[\bm{x}_0\bm{x}_1\dots\bm{x}_{p-1}]$. Finally, we define also a +classifier determined by our data via a function $G(x)$. This function tells us how well we are able to classify our outputs/targets $\bm{y}$. + +We have already defined the misclassification error $\mathrm{err}$ as +!bt +\[ +\mathrm{err}=\frac{1}{n}\sum_{i=0}^{n-1}I(y_i\ne G(x_i)), +\] +!et +where the function $I()$ is one if we misclassify and zero if we classify correctly. + +!split +===== Basic Steps of AdaBoost ===== + +With the above definitions we are now ready to set up the algorithm for AdaBoost. +The basic idea is to set up weights which will be used to scale the correctly classified and the misclassified cases. +o We start by initializing all weights to $w_i = 1/n$, with $i=0,1,2,\dots n-1$. It is easy to see that we must have $\sum_{i=0}^{n-1}w_i = 1$. +o We rewrite the misclassification error as +!bt +\[ +\mathrm{\overline{err}}_m=\frac{\sum_{i=0}^{n-1}w_i^m I(y_i\ne G(x_i))}{\sum_{i=0}^{n-1}w_i}, +\] +!et +o Then we start looping over all attempts at classifying, namely we start an iterative process for $m=1:M$, where $M$ is the final number of classifications. Our given classifier could for example be a plain decision tree. + o Fit then a given classifier to the training set using the weights $w_i$. + o Compute then $\mathrm{err}$ and figure out which events are classified properly and which are classified wrongly. + o Define a quantity $\alpha_{m} = \log{(1-\mathrm{\overline{err}}_m)/\mathrm{\overline{err}}_m}$ + o Set the new weights to $w_i = w_i\times \exp{(\alpha_m I(y_i\ne G(x_i)}$. +o Compute the new classifier $G(x)= \sum_{i=0}^{n-1}\alpha_m I(y_i\ne G(x_i)$. + +For the iterations with $m \le 2$ the weights are modified +individually at each steps. The observations which were misclassified +at iteration $m-1$ have a weight which is larger than those which were +classified properly. As this proceeds, the observations which were +difficult to classifiy correctly are given a larger influence. Each +new classification step $m$ is then forced to concentrate on those +observations that are missed in the previous iterations. + + + +!split +===== AdaBoost Examples ===== + +Using _Scikit-Learn_ it is easy to apply the adaptive boosting algorithm, as done here. + +!bc pycod +from sklearn.ensemble import AdaBoostClassifier + +ada_clf = AdaBoostClassifier( + DecisionTreeClassifier(max_depth=2), n_estimators=200, + algorithm="SAMME.R", learning_rate=0.01, random_state=42) +ada_clf.fit(X_train, y_train) +y_pred = ada_clf.predict(X_test) +skplt.metrics.plot_confusion_matrix(y_test, y_pred, normalize=True) +plt.show() +y_probas = ada_clf.predict_proba(X_test) +skplt.metrics.plot_roc(y_test, y_probas) +plt.show() +skplt.metrics.plot_cumulative_gain(y_test, y_probas) +plt.show() +!ec + +!split +===== Making an ADAboost code yourself ===== !bc pycod import numpy as np -import os -np.random.seed(42) +class DecisionStump: + def fit(self, X, y, weights): + m, n = X.shape + self.alpha = 0 + self.threshold = None + self.polarity = 1 -# To plot pretty figures -import matplotlib -import matplotlib.pyplot as plt -plt.rcParams['axes.labelsize'] = 14 -plt.rcParams['xtick.labelsize'] = 12 -plt.rcParams['ytick.labelsize'] = 12 + min_error = float('inf') + for feature in range(n): + feature_values = np.unique(X[:, feature]) -from sklearn.svm import SVC -from sklearn import datasets + for threshold in feature_values: + for polarity in [1, -1]: + predictions = np.ones(m) + predictions[X[:, feature] < threshold] = -1 + predictions *= polarity + error = sum(weights[predictions != y]) + if error < min_error: + min_error = error + self.alpha = 0.5 * np.log((1 - error) / (error + 1e-10)) + self.threshold = threshold + self.feature_index = feature + self.polarity = polarity -X1D = np.linspace(-4, 4, 9).reshape(-1, 1) -X2D = np.c_[X1D, X1D**2] -y = np.array([0, 0, 1, 1, 1, 1, 1, 0, 0]) + def predict(self, X): + m = X.shape[0] + predictions = np.ones(m) + if self.polarity == 1: + predictions[X[:, self.feature_index] < self.threshold] = -1 + else: + predictions[X[:, self.feature_index] >= self.threshold] = -1 + return predictions -plt.figure(figsize=(11, 4)) +class AdaBoost: + def fit(self, X, y, n_estimators): + m = X.shape[0] + self.alphas = [] + self.models = [] -plt.subplot(121) -plt.grid(True, which='both') -plt.axhline(y=0, color='k') -plt.plot(X1D[:, 0][y==0], np.zeros(4), "bs") -plt.plot(X1D[:, 0][y==1], np.zeros(5), "g^") -plt.gca().get_yaxis().set_ticks([]) -plt.xlabel(r"$x_1$", fontsize=20) -plt.axis([-4.5, 4.5, -0.2, 0.2]) + weights = np.ones(m) / m -plt.subplot(122) -plt.grid(True, which='both') -plt.axhline(y=0, color='k') -plt.axvline(x=0, color='k') -plt.plot(X2D[:, 0][y==0], X2D[:, 1][y==0], "bs") -plt.plot(X2D[:, 0][y==1], X2D[:, 1][y==1], "g^") -plt.xlabel(r"$x_1$", fontsize=20) -plt.ylabel(r"$x_2$", fontsize=20, rotation=0) -plt.gca().get_yaxis().set_ticks([0, 4, 8, 12, 16]) -plt.plot([-4.5, 4.5], [6.5, 6.5], "r--", linewidth=3) -plt.axis([-4.5, 4.5, -1, 17]) -plt.subplots_adjust(right=1) -plt.show() + for _ in range(n_estimators): + stump = DecisionStump() + stump.fit(X, y, weights) + predictions = stump.predict(X) + + error = sum(weights[predictions != y]) + if error == 0: + break + + self.models.append(stump) + self.alphas.append(stump.alpha) + + weights *= np.exp(-stump.alpha * y * predictions) + weights /= np.sum(weights) + + def predict(self, X): + final_predictions = np.zeros(X.shape[0]) + for alpha, model in zip(self.alphas, self.models): + final_predictions += alpha * model.predict(X) + return np.sign(final_predictions) + +# Example dataset (X, y) +X = np.array([[1], [2], [3], [4], [5], [6], [7], [8], [9], [10]]) +y = np.array([-1, -1, -1, -1, 1, 1, 1, 1, 1, 1]) # Labels must be -1 or 1 + +# Train AdaBoost +ada = AdaBoost() +ada.fit(X, y, n_estimators=10) + +# Predictions +predictions = ada.predict(X) +print("Predictions:", predictions) !ec +!split +===== Gradient boosting: Basics with Steepest Descent/Functional Gradient Descent ===== + +Gradient boosting is again a similar technique to Adaptive boosting, +it combines so-called weak classifiers or regressors into a strong +method via a series of iterations. + +In order to understand the method, let us illustrate its basics by +bringing back the essential steps in linear regression, where our cost +function was the least squares function. + +!split +===== The Squared-Error again! Steepest Descent ===== + +We start again with our cost function ${\cal C}(\bm{y}m\bm{f})=\sum_{i=0}^{n-1}{\cal L}(y_i, f(x_i))$ where we want to minimize +This means that for every iteration, we need to optimize + +!bt +\[ +(\hat{\bm{f}}) = \mathrm{argmin}_{\bm{f}}\hspace{0.1cm} \sum_{i=0}^{n-1}(y_i-f(x_i))^2. +\] +!et + +We define a real function $h_m(x)$ that defines our final function $f_M(x)$ as +!bt +\[ +f_M(x) = \sum_{m=0}^M h_m(x). +\] +!et + +In the steepest decent approach we approximate $h_m(x) = -\rho_m g_m(x)$, where $\rho_m$ is a scalar and $g_m(x)$ the gradient defined as +!bt +\[ +g_m(x_i) = \left[ \frac{\partial {\cal L}(y_i, f(x_i))}{\partial f(x_i)}\right]_{f(x_i)=f_{m-1}(x_i)}. +\] +!et + +With the new gradient we can update $f_m(x) = f_{m-1}(x) -\rho_m g_m(x)$. Using the above squared-error function we see that +the gradient is $g_m(x_i) = -2(y_i-f(x_i))$. + +Choosing $f_0(x)=0$ we obtain $g_m(x) = -2y_i$ and inserting this into the minimization problem for the cost function we have +!bt +\[ +(\rho_1) = \mathrm{argmin}_{\rho}\hspace{0.1cm} \sum_{i=0}^{n-1}(y_i+2\rho y_i)^2. +\] +!et + +!split +===== Steepest Descent Example ===== + +Optimizing with respect to $\rho$ we obtain (taking the derivative) that $\rho_1 = -1/2$. We have then that +!bt +\[ +f_1(x) = f_{0}(x) -\rho_1 g_1(x)=-y_i. +\] +!et +We can then proceed and compute +!bt +\[ +g_2(x_i) = \left[ \frac{\partial {\cal L}(y_i, f(x_i))}{\partial f(x_i)}\right]_{f(x_i)=f_{1}(x_i)=y_i}=-4y_i, +\] +!et +and find a new value for $\rho_2=-1/2$ and continue till we have reached $m=M$. We can modify the steepest descent method, or steepest boosting, by introducing what is called _gradient boosting_. + +!split +===== Gradient Boosting, algorithm ===== + +Steepest descent is however not much used, since it only optimizes $f$ at a fixed set of $n$ points, +so we do not learn a function that can generalize. However, we can modify the algorithm by +fitting a weak learner to approximate the negative gradient signal. + +Suppose we have a cost function $C(f)=\sum_{i=0}^{n-1}L(y_i, f(x_i))$ where $y_i$ is our target and $f(x_i)$ the function which is meant to model $y_i$. The above cost function could be our standard squared-error function +!bt +\[ +C(\bm{y},\bm{f})=\sum_{i=0}^{n-1}(y_i-f(x_i))^2. +\] +!et + +The way we proceed in an iterative fashion is to +o Initialize our estimate $f_0(x)$. +o For $m=1:M$, we + o compute the negative gradient vector $\bm{u}_m = -\partial C(\bm{y},\bm{f})/\partial \bm{f}(x)$ at $f(x) = f_{m-1}(x)$; + o fit the so-called base-learner to the negative gradient $h_m(u_m,x)$; + o update the estimate $f_m(x) = f_{m-1}(x)+h_m(u_m,x)$; +o The final estimate is then $f_M(x) = \sum_{m=1}^M h_m(u_m,x)$. + !split -===== The equations ===== - -Suppose we define a polynomial transformation of degree two only. We define a vector $\bm{x}_i=[x_i,y_i]$ and have -!bt -\[ -\bm{z}_i^T = \phi(\bm{x}_i)^T =\left[1, \sqrt{2}x_i, \sqrt{2}y_i, x_i^2, y_i^2, \sqrt{2}x_iy_i\right]. -\] -!et - -With our new basis, the equations we solved earlier are basically the same, that is we have now (without the slack option for simplicity) -!bt -\[ -{\cal L}=\sum_i\lambda_i-\frac{1}{2}\sum_{ij}^n\lambda_i\lambda_jy_iy_j\bm{z}_i^T\bm{z}_j, -\] -!et -subject to the constraints $\lambda_i\geq 0$, $\sum_i\lambda_iy_i=0$, and for the support vectors -!bt -\[ -y_i(\bm{z}_i^T\bm{w}+b)= 1 \hspace{0.1cm}\forall i, -\] -!et -from which we also find $b$. -To compute $\bm{z}_i^T\bm{z}_j$ we define the kernel $K(\bm{x}_i,\bm{x}_j)$ as -!bt -\[ -K(\bm{x}_i,\bm{x}_j)=\bm{z}_i^T\bm{z}_j= \phi(\bm{x}_i)^T\phi(\bm{x}_j). -\] -!et -For the above example, the kernel reads -!bt -\[ -K(\bm{x}_i,\bm{x}_j)=[1, \sqrt{2}x_i, \sqrt{2}y_i, x_i^2, y_i^2, \sqrt{2}x_iy_i]\begin{bmatrix} 1\\ \sqrt{2}x_j \\ \sqrt{2}y_j \\ x_j^2\\ y_i^2\\ \sqrt{2}x_jy_j \end{bmatrix}=1+2x_ix_j+2y_iy_j+(x_ix_j)^2+(y_iy_j)^2+2x_ix_jy_iy_j. -\] -!et - -We note that this dot product can be rewritten as -!bt -\[ -K(\bm{x}_i,\bm{x}_j)=[1+\bm{x}_i^T\bm{x}_j]^d, -\] -!et -where $d=2$ in our case and $\bm{x}_i=[x_i,y_i]$ and $\bm{x}_j=[x_j,y_j]$. -To compute the last equation is however inefficient from a computational stand. -Instead of computing the last equation for the kernel, we simply compute -the dot product $(\bm{x}_i^T\bm{x}_j)^2$. -This leads to the so-called -kernel trick. - -!split -===== The problem to solve ===== - -Using our definition of the kernel, we can rewrite again the Lagrangian -!bt -\[ -{\cal L}=\sum_i\lambda_i-\frac{1}{2}\sum_{ij}^n\lambda_i\lambda_jy_iy_j\bm{z}_i^T\bm{z}_j, -\] -!et -subject to the constraints $\lambda_i\geq 0$, $\sum_i\lambda_iy_i=0$ in terms of a convex optimization problem -!bt -\[ -\frac{1}{2} \bm{\lambda}^T\begin{bmatrix} y_1y_1K(\bm{x}_1,\bm{x}_1) & y_1y_2K(\bm{x}_1,\bm{x}_2) & \dots & \dots & y_1y_nK(\bm{x}_1,\bm{x}_n) \\ -y_2y_1K(\bm{x}_2,\bm{x}_1) & y_2y_2(\bm{x}_2,\bm{x}_2) & \dots & \dots & y_1y_nK(\bm{x}_2,\bm{x}_n) \\ -\dots & \dots & \dots & \dots & \dots \\ -\dots & \dots & \dots & \dots & \dots \\ -y_ny_1K(\bm{x}_n,\bm{x}_1) & y_ny_2K(\bm{x}_n\bm{x}_2) & \dots & \dots & y_ny_nK(\bm{x}_n,\bm{x}_n) \\ -\end{bmatrix}\bm{\lambda}-\mathbb{1}\bm{\lambda}, -\] -!et -subject to $\bm{y}^T\bm{\lambda}=0$. Here we defined the vectors $\bm{\lambda} =[\lambda_1,\lambda_2,\dots,\lambda_n]$ and -$\bm{y}=[y_1,y_2,\dots,y_n]$. -If we add the slack constants this leads to the additional constraint $0\leq \lambda_i \leq C$. - -!split -===== Tailoring the equations to the usage of CVXOPT ===== - -We can rewrite this (see the solutions below) in terms of a convex optimization problem of the type -!bt -\begin{align*} - &\mathrm{min}_{\lambda}\hspace{0.2cm} \frac{1}{2}\bm{\lambda}^T\bm{P}\bm{\lambda}+\bm{q}^T\bm{\lambda},\\ \nonumber - &\mathrm{s.t} \hspace{0.2cm} \bm{G}\bm{\lambda} \preceq \bm{h} \hspace{0.2cm} \wedge \bm{A}\bm{\lambda}=f. -\end{align*} -!et -Below we discuss how to solve these equations. Here we note that the matrix $\bm{P}$ has matrix elements $p_{ij}=y_iy_jK(\bm{x}_i,\bm{x}_j)$. -Given a kernel $K$ and the targets $y_i$ this matrix is easy to set up. The constraint $\bm{y}^T\bm{\lambda}=0$ leads to $f=0$ and $\bm{A}=\bm{y}$. How to set up the matrix $\bm{G}$ is discussed later. Here note that the inequalities $0\leq \lambda_i \leq C$ can be split up into -$0\leq \lambda_i$ and $\lambda_i \leq C$. These two inequalities define then the matrix $\bm{G}$ and the vector $\bm{h}$. - - -!split -===== Different kernels and Mercer's theorem ===== - -There are several popular kernels being used. These are -o Linear: $K(\bm{x},\bm{y})=\bm{x}^T\bm{y}$, -o Polynomial: $K(\bm{x},\bm{y})=(\bm{x}^T\bm{y}+\gamma)^d$, -o Gaussian Radial Basis Function: $K(\bm{x},\bm{y})=\exp{\left(-\gamma\vert\vert\bm{x}-\bm{y}\vert\vert^2\right)}$, -o Tanh: $K(\bm{x},\bm{y})=\tanh{(\bm{x}^T\bm{y}+\gamma)}$, -and many other ones. - -An important theorem for us is "Mercer's -theorem":"https://en.wikipedia.org/wiki/Mercer%27s_theorem". The -theorem states that if a kernel function $K$ is symmetric, continuous -and leads to a positive semi-definite matrix $\bm{P}$ then there -exists a function $\phi$ that maps $\bm{x}_i$ and $\bm{x}_j$ into -another space (possibly with much higher dimensions) such that - -!bt -\[ -K(\bm{x}_i,\bm{x}_j)=\phi(\bm{x}_i)^T\phi(\bm{x}_j). -\] -!et - -So you can use $K$ as a kernel since you know $\phi$ exists, even if -you don’t know what $\phi$ is. - -Note that some frequently used kernels (such as the Sigmoid kernel) -don’t respect all of Mercer’s conditions, yet they generally work well -in practice. - - -!split -===== The moons example ("Adapted from Geron, chapter 5":"https://www.oreilly.com/library/view/hands-on-machine-learning/9781492032632/") ===== +===== Gradient Boosting, Examples of Regression ===== !bc pycod -from __future__ import division, print_function, unicode_literals - -import numpy as np -np.random.seed(42) - -import matplotlib import matplotlib.pyplot as plt -plt.rcParams['axes.labelsize'] = 14 -plt.rcParams['xtick.labelsize'] = 12 -plt.rcParams['ytick.labelsize'] = 12 +import numpy as np +from sklearn.model_selection import train_test_split +from sklearn.ensemble import GradientBoostingRegressor +import scikitplot as skplt +from sklearn.metrics import mean_squared_error + +n = 100 +maxdegree = 6 + +# Make data set. +x = np.linspace(-3, 3, n).reshape(-1, 1) +y = np.exp(-x**2) + 1.5 * np.exp(-(x-2)**2)+ np.random.normal(0, 0.1, x.shape) + +error = np.zeros(maxdegree) +bias = np.zeros(maxdegree) +variance = np.zeros(maxdegree) +polydegree = np.zeros(maxdegree) +X_train, X_test, y_train, y_test = train_test_split(x, y, test_size=0.2) + +for degree in range(1,maxdegree): + model = GradientBoostingRegressor(max_depth=degree, n_estimators=100, learning_rate=1.0) + model.fit(X_train,y_train) + y_pred = model.predict(X_test) + polydegree[degree] = degree + error[degree] = np.mean( np.mean((y_test - y_pred)**2) ) + bias[degree] = np.mean( (y_test - np.mean(y_pred))**2 ) + variance[degree] = np.mean( np.var(y_pred) ) + print('Max depth:', degree) + print('Error:', error[degree]) + print('Bias^2:', bias[degree]) + print('Var:', variance[degree]) + print('{} >= {} + {} = {}'.format(error[degree], bias[degree], variance[degree], bias[degree]+variance[degree])) + +plt.xlim(1,maxdegree-1) +plt.plot(polydegree, error, label='Error') +plt.plot(polydegree, bias, label='bias') +plt.plot(polydegree, variance, label='Variance') +plt.legend() +save_fig("gdregression") +plt.show() +!ec -from sklearn.svm import SVC -from sklearn import datasets +!split +===== Gradient Boosting, Classification Example ===== +!bc pycod +import matplotlib.pyplot as plt +import numpy as np +from sklearn.model_selection import train_test_split +from sklearn.datasets import load_breast_cancer +import scikitplot as skplt +from sklearn.ensemble import GradientBoostingClassifier +from sklearn.model_selection import cross_validate +# Load the data +cancer = load_breast_cancer() - -from sklearn.pipeline import Pipeline +X_train, X_test, y_train, y_test = train_test_split(cancer.data,cancer.target,random_state=0) +print(X_train.shape) +print(X_test.shape) +#now scale the data from sklearn.preprocessing import StandardScaler -from sklearn.svm import LinearSVC +scaler = StandardScaler() +scaler.fit(X_train) +X_train_scaled = scaler.transform(X_train) +X_test_scaled = scaler.transform(X_test) +gd_clf = GradientBoostingClassifier(max_depth=3, n_estimators=100, learning_rate=1.0) +gd_clf.fit(X_train_scaled, y_train) +#Cross validation +accuracy = cross_validate(gd_clf,X_test_scaled,y_test,cv=10)['test_score'] +print(accuracy) +print("Test set accuracy with Gradient boosting and scaled data: {:.2f}".format(gd_clf.score(X_test_scaled,y_test))) -from sklearn.datasets import make_moons -X, y = make_moons(n_samples=100, noise=0.15, random_state=42) - -def plot_dataset(X, y, axes): - plt.plot(X[:, 0][y==0], X[:, 1][y==0], "bs") - plt.plot(X[:, 0][y==1], X[:, 1][y==1], "g^") - plt.axis(axes) - plt.grid(True, which='both') - plt.xlabel(r"$x_1$", fontsize=20) - plt.ylabel(r"$x_2$", fontsize=20, rotation=0) - -plot_dataset(X, y, [-1.5, 2.5, -1, 1.5]) +import scikitplot as skplt +y_pred = gd_clf.predict(X_test_scaled) +skplt.metrics.plot_confusion_matrix(y_test, y_pred, normalize=True) +save_fig("gdclassiffierconfusion") plt.show() - -from sklearn.datasets import make_moons -from sklearn.pipeline import Pipeline -from sklearn.preprocessing import PolynomialFeatures - -polynomial_svm_clf = Pipeline([ - ("poly_features", PolynomialFeatures(degree=3)), - ("scaler", StandardScaler()), - ("svm_clf", LinearSVC(C=10, loss="hinge", random_state=42)) - ]) - -polynomial_svm_clf.fit(X, y) - -def plot_predictions(clf, axes): - x0s = np.linspace(axes[0], axes[1], 100) - x1s = np.linspace(axes[2], axes[3], 100) - x0, x1 = np.meshgrid(x0s, x1s) - X = np.c_[x0.ravel(), x1.ravel()] - y_pred = clf.predict(X).reshape(x0.shape) - y_decision = clf.decision_function(X).reshape(x0.shape) - plt.contourf(x0, x1, y_pred, cmap=plt.cm.brg, alpha=0.2) - plt.contourf(x0, x1, y_decision, cmap=plt.cm.brg, alpha=0.1) - -plot_predictions(polynomial_svm_clf, [-1.5, 2.5, -1, 1.5]) -plot_dataset(X, y, [-1.5, 2.5, -1, 1.5]) - +y_probas = gd_clf.predict_proba(X_test_scaled) +skplt.metrics.plot_roc(y_test, y_probas) +save_fig("gdclassiffierroc") plt.show() - - -from sklearn.svm import SVC - -poly_kernel_svm_clf = Pipeline([ - ("scaler", StandardScaler()), - ("svm_clf", SVC(kernel="poly", degree=3, coef0=1, C=5)) - ]) -poly_kernel_svm_clf.fit(X, y) - -poly100_kernel_svm_clf = Pipeline([ - ("scaler", StandardScaler()), - ("svm_clf", SVC(kernel="poly", degree=10, coef0=100, C=5)) - ]) -poly100_kernel_svm_clf.fit(X, y) - -plt.figure(figsize=(11, 4)) - -plt.subplot(121) -plot_predictions(poly_kernel_svm_clf, [-1.5, 2.5, -1, 1.5]) -plot_dataset(X, y, [-1.5, 2.5, -1, 1.5]) -plt.title(r"$d=3, r=1, C=5$", fontsize=18) - -plt.subplot(122) -plot_predictions(poly100_kernel_svm_clf, [-1.5, 2.5, -1, 1.5]) -plot_dataset(X, y, [-1.5, 2.5, -1, 1.5]) -plt.title(r"$d=10, r=100, C=5$", fontsize=18) - +skplt.metrics.plot_cumulative_gain(y_test, y_probas) +save_fig("gdclassiffiercgain") plt.show() - -def gaussian_rbf(x, landmark, gamma): - return np.exp(-gamma * np.linalg.norm(x - landmark, axis=1)**2) - -gamma = 0.3 - -x1s = np.linspace(-4.5, 4.5, 200).reshape(-1, 1) -x2s = gaussian_rbf(x1s, -2, gamma) -x3s = gaussian_rbf(x1s, 1, gamma) - -XK = np.c_[gaussian_rbf(X1D, -2, gamma), gaussian_rbf(X1D, 1, gamma)] -yk = np.array([0, 0, 1, 1, 1, 1, 1, 0, 0]) - -plt.figure(figsize=(11, 4)) - -plt.subplot(121) -plt.grid(True, which='both') -plt.axhline(y=0, color='k') -plt.scatter(x=[-2, 1], y=[0, 0], s=150, alpha=0.5, c="red") -plt.plot(X1D[:, 0][yk==0], np.zeros(4), "bs") -plt.plot(X1D[:, 0][yk==1], np.zeros(5), "g^") -plt.plot(x1s, x2s, "g--") -plt.plot(x1s, x3s, "b:") -plt.gca().get_yaxis().set_ticks([0, 0.25, 0.5, 0.75, 1]) -plt.xlabel(r"$x_1$", fontsize=20) -plt.ylabel(r"Similarity", fontsize=14) -plt.annotate(r'$\mathbf{x}$', - xy=(X1D[3, 0], 0), - xytext=(-0.5, 0.20), - ha="center", - arrowprops=dict(facecolor='black', shrink=0.1), - fontsize=18, - ) -plt.text(-2, 0.9, "$x_2$", ha="center", fontsize=20) -plt.text(1, 0.9, "$x_3$", ha="center", fontsize=20) -plt.axis([-4.5, 4.5, -0.1, 1.1]) - -plt.subplot(122) -plt.grid(True, which='both') -plt.axhline(y=0, color='k') -plt.axvline(x=0, color='k') -plt.plot(XK[:, 0][yk==0], XK[:, 1][yk==0], "bs") -plt.plot(XK[:, 0][yk==1], XK[:, 1][yk==1], "g^") -plt.xlabel(r"$x_2$", fontsize=20) -plt.ylabel(r"$x_3$ ", fontsize=20, rotation=0) -plt.annotate(r'$\phi\left(\mathbf{x}\right)$', - xy=(XK[3, 0], XK[3, 1]), - xytext=(0.65, 0.50), - ha="center", - arrowprops=dict(facecolor='black', shrink=0.1), - fontsize=18, - ) -plt.plot([-0.1, 1.1], [0.57, -0.1], "r--", linewidth=3) -plt.axis([-0.1, 1.1, -0.1, 1.1]) - -plt.subplots_adjust(right=1) - -plt.show() - - -x1_example = X1D[3, 0] -for landmark in (-2, 1): - k = gaussian_rbf(np.array([[x1_example]]), np.array([[landmark]]), gamma) - print("Phi({}, {}) = {}".format(x1_example, landmark, k)) - -rbf_kernel_svm_clf = Pipeline([ - ("scaler", StandardScaler()), - ("svm_clf", SVC(kernel="rbf", gamma=5, C=0.001)) - ]) -rbf_kernel_svm_clf.fit(X, y) - - -from sklearn.svm import SVC - -gamma1, gamma2 = 0.1, 5 -C1, C2 = 0.001, 1000 -hyperparams = (gamma1, C1), (gamma1, C2), (gamma2, C1), (gamma2, C2) - -svm_clfs = [] -for gamma, C in hyperparams: - rbf_kernel_svm_clf = Pipeline([ - ("scaler", StandardScaler()), - ("svm_clf", SVC(kernel="rbf", gamma=gamma, C=C)) - ]) - rbf_kernel_svm_clf.fit(X, y) - svm_clfs.append(rbf_kernel_svm_clf) - -plt.figure(figsize=(11, 7)) - -for i, svm_clf in enumerate(svm_clfs): - plt.subplot(221 + i) - plot_predictions(svm_clf, [-1.5, 2.5, -1, 1.5]) - plot_dataset(X, y, [-1.5, 2.5, -1, 1.5]) - gamma, C = hyperparams[i] - plt.title(r"$\gamma = {}, C = {}$".format(gamma, C), fontsize=16) - -plt.show() - !ec +!split +===== XGBoost: Extreme Gradient Boosting ===== + + +"XGBoost":"https://github.com/dmlc/xgboost" or Extreme Gradient +Boosting, is an optimized distributed gradient boosting library +designed to be highly efficient, flexible and portable. It implements +machine learning algorithms under the Gradient Boosting +framework. XGBoost provides a parallel tree boosting that solve many +data science problems in a fast and accurate way. See the "article by Chen and Guestrin":"https://arxiv.org/abs/1603.02754". + +The authors design and build a highly scalable end-to-end tree +boosting system. It has a theoretically justified weighted quantile +sketch for efficient proposal calculation. It introduces a novel sparsity-aware algorithm for parallel tree learning and an effective cache-aware block structure for out-of-core tree learning. + +It is now the algorithm which wins essentially all ML competitions!!! !split -===== Mathematical optimization of convex functions ===== - -A mathematical (quadratic) optimization problem, or just optimization problem, has the form -!bt -\begin{align*} - &\mathrm{min}_{\lambda}\hspace{0.2cm} \frac{1}{2}\bm{\lambda}^T\bm{P}\bm{\lambda}+\bm{q}^T\bm{\lambda},\\ \nonumber - &\mathrm{subject\hspace{0.1cm}to} \hspace{0.2cm} \bm{G}\bm{\lambda} \preceq \bm{h} \wedge \bm{A}\bm{\lambda}=f. -\end{align*} -!et -subject to some constraints for say a selected set $i=1,2,\dots, n$. -In our case we are optimizing with respect to the Lagrangian multipliers $\lambda_i$, and the -vector $\bm{\lambda}=[\lambda_1, \lambda_2,\dots, \lambda_n]$ is the optimization variable we are dealing with. - -In our case we are particularly interested in a class of optimization problems called convex optmization problems. -In our discussion on gradient descent methods we discussed at length the definition of a convex function. - -Convex optimization problems play a central role in applied mathematics and we recommend strongly "Boyd and Vandenberghe's text on the topics":"http://web.stanford.edu/~boyd/cvxbook/". - - - -!split -===== How do we solve these problems? ===== - -If we use Python as programming language and wish to venture beyond -_scikit-learn_, _tensorflow_ and similar software which makes our -lives so much easier, we need to dive into the wonderful world of -quadratic programming. We can, if we wish, solve the minimization -problem using standard gradient methods or conjugate gradient -methods. However, these methods tend to exhibit a rather slow -converge. So, welcome to the promised land of quadratic programming. - -The functions we need are contained in the quadratic programming package _CVXOPT_ and we need to import it together with _numpy_ as +===== Regression Case ===== !bc pycod -import numpy -import cvxopt +import matplotlib.pyplot as plt +import numpy as np +from sklearn.model_selection import train_test_split +import xgboost as xgb +import scikitplot as skplt +from sklearn.metrics import mean_squared_error + +n = 100 +maxdegree = 6 + +# Make data set. +x = np.linspace(-3, 3, n).reshape(-1, 1) +y = np.exp(-x**2) + 1.5 * np.exp(-(x-2)**2)+ np.random.normal(0, 0.1, x.shape) + +error = np.zeros(maxdegree) +bias = np.zeros(maxdegree) +variance = np.zeros(maxdegree) +polydegree = np.zeros(maxdegree) +X_train, X_test, y_train, y_test = train_test_split(x, y, test_size=0.2) + +for degree in range(maxdegree): + model = xgb.XGBRegressor(objective ='reg:squarederror', colsaobjective ='reg:squarederror', colsample_bytree = 0.3, learning_rate = 0.1,max_depth = degree, alpha = 10, n_estimators = 200) + + model.fit(X_train,y_train) + y_pred = model.predict(X_test) + polydegree[degree] = degree + error[degree] = np.mean( np.mean((y_test - y_pred)**2) ) + bias[degree] = np.mean( (y_test - np.mean(y_pred))**2 ) + variance[degree] = np.mean( np.var(y_pred) ) + print('Max depth:', degree) + print('Error:', error[degree]) + print('Bias^2:', bias[degree]) + print('Var:', variance[degree]) + print('{} >= {} + {} = {}'.format(error[degree], bias[degree], variance[degree], bias[degree]+variance[degree])) + +plt.xlim(1,maxdegree-1) +plt.plot(polydegree, error, label='Error') +plt.plot(polydegree, bias, label='bias') +plt.plot(polydegree, variance, label='Variance') +plt.legend() +plt.show() + !ec -This will make our life much easier. You don't need to write your own optimizer. - - !split -===== A simple example ===== +===== Xgboost on the Cancer Data ===== -We remind ourselves about the general problem we want to solve -!bt -\begin{align*} - &\mathrm{min}_{x}\hspace{0.2cm} \frac{1}{2}\bm{x}^T\bm{P}\bm{x}+\bm{q}^T\bm{x},\\ \nonumber - &\mathrm{s.t.} \hspace{0.2cm} \bm{G}\bm{x} \preceq \bm{h} \wedge \bm{A}\bm{x}=f. -\end{align*} -!et -_Note_: we use _s.t._ for subject to. - -Let us show how to perform the optmization using a simple case. Assume we want to optimize the following problem -!bt -\begin{align*} - &\mathrm{min}_{x}\hspace{0.2cm} \frac{1}{2}x^2+5x+3y \\ \nonumber - &\mathrm{s.t.} \\ \nonumber - &x, y \geq 0 \\ \nonumber - &x+3y \geq 15 \\ \nonumber - &2x+5y \leq 100 \\ \nonumber - &3x+4y \leq 80. \\ \nonumber -\end{align*} -!et -The minimization problem can be rewritten in terms of vectors and matrices as (with $x$ and $y$ being the unknowns) -!bt -\[ -\frac{1}{2}\begin{bmatrix} x\\ y \end{bmatrix}^T \begin{bmatrix} 1 & 0\\ 0 & 0 \end{bmatrix} \begin{bmatrix} x \\ y \end{bmatrix} + \begin{bmatrix}5\\ 3 \end{bmatrix}^T \begin{bmatrix}x \\ y \end{bmatrix}. -\] -!et -Similarly, we can now set up the inequalities (we need to change $\geq$ to $\leq$ by multiplying with $-1$ on bot sides) as the following matrix-vector equation -!bt -\[ -\begin{bmatrix} -1 & 0 \\ 0 & -1 \\ -1 & -3 \\ 2 & 5 \\ 3 & 4\end{bmatrix}\begin{bmatrix} x \\ y\end{bmatrix} \preceq \begin{bmatrix}0 \\ 0\\ -15 \\ 100 \\ 80\end{bmatrix}. -\] -!et -We have collapsed all the inequalities into a single matrix $\bm{G}$. We see also that our matrix -!bt -\[ -\bm{P} =\begin{bmatrix} 1 & 0\\ 0 & 0 \end{bmatrix} -\] -!et -is clearly positive semi-definite (all eigenvalues larger or equal zero). -Finally, the vector $\bm{h}$ is defined as -!bt -\[ -\bm{h} = \begin{bmatrix}0 \\ 0\\ -15 \\ 100 \\ 80\end{bmatrix}. -\] -!et - - -Since we don't have any equalities the matrix $\bm{A}$ is set to zero -The following code solves the equations for us +As you will see from the confusion matrix below, XGBoots does an excellent job on the Wisconsin cancer data and outperforms essentially all agorithms we have discussed till now. !bc pycod -# Import the necessary packages -import numpy -from cvxopt import matrix -from cvxopt import solvers -P = matrix(numpy.diag([1,0]), tc='d') -q = matrix(numpy.array([3,4]), tc='d') -G = matrix(numpy.array([[-1,0],[0,-1],[-1,-3],[2,5],[3,4]]), tc='d') -h = matrix(numpy.array([0,0,-15,100,80]), tc='d') -# Construct the QP, invoke solver -sol = solvers.qp(P,q,G,h) -# Extract optimal value and solution -sol['x'] -sol['primal objective'] + +import matplotlib.pyplot as plt +import numpy as np +from sklearn.model_selection import train_test_split +from sklearn.datasets import load_breast_cancer +from sklearn.preprocessing import LabelEncoder +from sklearn.model_selection import cross_validate +import scikitplot as skplt +import xgboost as xgb +# Load the data +cancer = load_breast_cancer() + +X_train, X_test, y_train, y_test = train_test_split(cancer.data,cancer.target,random_state=0) +print(X_train.shape) +print(X_test.shape) +#now scale the data +from sklearn.preprocessing import StandardScaler +scaler = StandardScaler() +scaler.fit(X_train) +X_train_scaled = scaler.transform(X_train) +X_test_scaled = scaler.transform(X_test) + +xg_clf = xgb.XGBClassifier() +xg_clf.fit(X_train_scaled,y_train) + +y_test = xg_clf.predict(X_test_scaled) + +print("Test set accuracy with Gradient Boosting and scaled data: {:.2f}".format(xg_clf.score(X_test_scaled,y_test))) + +import scikitplot as skplt +y_pred = xg_clf.predict(X_test_scaled) +skplt.metrics.plot_confusion_matrix(y_test, y_pred, normalize=True) +save_fig("xdclassiffierconfusion") +plt.show() +y_probas = xg_clf.predict_proba(X_test_scaled) +skplt.metrics.plot_roc(y_test, y_probas) +save_fig("xdclassiffierroc") +plt.show() +skplt.metrics.plot_cumulative_gain(y_test, y_probas) +save_fig("gdclassiffiercgain") +plt.show() + + +xgb.plot_tree(xg_clf,num_trees=0) +plt.rcParams['figure.figsize'] = [50, 10] +save_fig("xgtree") +plt.show() + +xgb.plot_importance(xg_clf) +plt.rcParams['figure.figsize'] = [5, 5] +save_fig("xgparams") +plt.show() + !ec -!split -===== Back to the more realistic cases ===== - -We are now ready to return to our setup of the optmization problem for a more realistic case. Introducing the _slack_ parameter $C$ we have -!bt -\[ -\frac{1}{2} \bm{\lambda}^T\begin{bmatrix} y_1y_1K(\bm{x}_1,\bm{x}_1) & y_1y_2K(\bm{x}_1,\bm{x}_2) & \dots & \dots & y_1y_nK(\bm{x}_1,\bm{x}_n) \\ -y_2y_1K(\bm{x}_2,\bm{x}_1) & y_2y_2K(\bm{x}_2,\bm{x}_2) & \dots & \dots & y_1y_nK(\bm{x}_2,\bm{x}_n) \\ -\dots & \dots & \dots & \dots & \dots \\ -\dots & \dots & \dots & \dots & \dots \\ -y_ny_1K(\bm{x}_n,\bm{x}_1) & y_ny_2K(\bm{x}_n\bm{x}_2) & \dots & \dots & y_ny_nK(\bm{x}_n,\bm{x}_n) \\ -\end{bmatrix}\bm{\lambda}-\mathbb{I}\bm{\lambda}, -\] -!et -subject to $\bm{y}^T\bm{\lambda}=0$. Here we defined the vectors $\bm{\lambda} =[\lambda_1,\lambda_2,\dots,\lambda_n]$ and -$\bm{y}=[y_1,y_2,\dots,y_n]$. -With the slack constants this leads to the additional constraint $0\leq \lambda_i \leq C$. !split -===== Setting up the matrices and the problem ===== +===== Gradient boosting, making our own code for a regression case ===== -We have the general problem -!bt -\begin{align*} - &\mathrm{min}_{\lambda}\hspace{0.2cm} \frac{1}{2}\bm{\lambda}^T\bm{P}\bm{\lambda}+\bm{q}^T\bm{\lambda},\\ \nonumber - &\mathrm{s.t.} \hspace{0.2cm} \bm{G}\bm{\lambda} \preceq \bm{h} \wedge \bm{A}\bm{\lambda}=f. -\end{align*} -!et +!bc pycod +import numpy as np +class DecisionTreeRegressor: + def __init__(self, max_depth=3): + self.max_depth = max_depth + self.tree = None + def fit(self, X, y): + self.tree = self._grow_tree(X, y) + def _grow_tree(self, X, y, depth=0): + n_samples, n_features = X.shape + if depth < self.max_depth: + best_feature, best_threshold = self._best_split(X, y) + if best_feature is not None: + left_indices = X[:, best_feature] < best_threshold + right_indices = X[:, best_feature] >= best_threshold + left_child = self._grow_tree(X[left_indices], y[left_indices], depth + 1) + right_child = self._grow_tree(X[right_indices], y[right_indices], depth + 1) + return (best_feature, best_threshold, left_child, right_child) + return np.mean(y) + def _best_split(self, X, y): + best_mse = float('inf') + best_feature, best_threshold = None, None + n_samples, n_features = X.shape + + for feature in range(n_features): + thresholds = np.unique(X[:, feature]) + for threshold in thresholds: + left_indices = X[:, feature] < threshold + right_indices = X[:, feature] >= threshold + if len(y[left_indices]) > 0 and len(y[right_indices]) > 0: + left_mse = np.mean((y[left_indices] - np.mean(y[left_indices])) ** 2) + right_mse = np.mean((y[right_indices] - np.mean(y[right_indices])) ** 2) + mse = (len(y[left_indices]) * left_mse + len(y[right_indices]) * right_mse) / n_samples + + if mse < best_mse: + best_mse = mse + best_feature = feature + best_threshold = threshold + return best_feature, best_threshold + def predict(self, X): + return np.array([self._predict_sample(sample, self.tree) for sample in X]) + def _predict_sample(self, sample, node): + if isinstance(node, tuple): + feature, threshold, left_child, right_child = node + if sample[feature] < threshold: + return self._predict_sample(sample, left_child) + else: + return self._predict_sample(sample, right_child) + return node +class GradientBoostingRegressor: + def __init__(self, n_estimators=100, learning_rate=0.1, max_depth=3): + self.n_estimators = n_estimators + self.learning_rate = learning_rate + self.max_depth = max_depth + self.models = [] + def fit(self, X, y): + y_pred = np.zeros(y.shape) + for _ in range(self.n_estimators): + residuals = y - y_pred + model = DecisionTreeRegressor(max_depth=self.max_depth) + model.fit(X, residuals) + y_pred += self.learning_rate * model.predict(X) + self.models.append(model) + def predict(self, X): + y_pred = np.zeros(X.shape[0]) + for model in self.models: + y_pred += self.learning_rate * model.predict(X) + return y_pred +# Example usage +if __name__ == "__main__": + # Sample data + X = np.array([[1], [2], [3], [4], [5]]) + y = np.array([1.5, 1.7, 3.5, 3.7, 5.0]) + model = GradientBoostingRegressor(n_estimators=100, learning_rate=0.1, max_depth=2) + model.fit(X, y) + predictions = model.predict(X) + print("Predictions:", predictions) + +!ec - -o With a given kernel we can thus define the matrix $\bm{P}$. -o The matrix $\bm{P}$ has matrix elements $p_{ij}=y_iy_jK(\bm{x}_i,\bm{x}_j)$. Given a kernel $K$ and the targets $y_i$ this matrix is easy to set up. -o The vector $\bm{q}$ has all elements equal to $-1$. -o The constraint $\bm{y}^T\bm{\lambda}=0$ leads to $f=0$ and $\bm{A}=\bm{y}$. -o To set up the matrix $\bm{G}$ we note that the inequalities $0\leq \lambda_i \leq C$ can be split up into $0\leq \lambda_i$ and $\lambda_i \leq C$. These two inequalities define then the matrix $\bm{G}$ and the vector $\bm{h}$. - -!split -===== Setting up $\bm{G}\bm{\lambda} \preceq \bm{h}$ ===== - -We have two constraints, $0\le \lambda_i$ and $\lambda_i \le C$. To do this we multiply the ones with the constraint -$\ge$ with $-1$ in order to get $\le$. It means that the problem $\bm{G}\bm{\lambda} \preceq \bm{h}$ -can be written as -!bt -\[ -\begin{bmatrix} -1& 0 & 0 & \dots & 0 \\ -0& -1 & 0 & \dots & 0 \\ -0& 0 & -1 & \dots & 0 \\ -\dots& \dots & \dots & \dots & \dots \\ -0& 0 & 0 & \dots & -1 \\ -1 & 0 & 0 & \dots & 0 \\ -0& 1 & 0 & \dots & 0 \\ -0& 0 & 1 & \dots & 0 \\ -\dots& \dots & \dots & \dots & \dots \\ -0& 0 & 0 & \dots & 1 \\ -\end{bmatrix} -\begin{bmatrix} \lambda_1 \\ -\lambda_2 \\ -\lambda_3 \\ -\dots \\ -\lambda_n \\ -\end{bmatrix} \preceq -\begin{bmatrix} 0 \\ -0 \\ -0 \\ -\dots \\ -0 \\ -C \\ -C \\ -C \\ -\dots \\ -C \\ -\end{bmatrix}. -\] -!et - -And then we are ready to go. - -!split -===== SVMs and Regression and multiclass classification ===== - -More text will be added here. See also "Hastie et al.":"https://www.springer.com/gp/book/9780387848570" section 12.3. - !split ===== Summary of course =====

    k4c)&iz-Xg85ALZ~x&UvtbQ|$v{e9J}p!sLLgwgatO)~xa)q9K3xy&ST> zd({~E?W^;nio;k0Gs}LFON&|*h7%5nYmO;SabG_h)X$pQqPX9pw4q~4`NU?_09HuO zO&gco&E~Ax?2^`aG3X?n)qtGF)cW1wR%R~_cf;*VoNI&71)8h%>{f6OnF^c=_n?(r z<&`n2m~sNC*N<5I+nHFQ;=!fwJL1;`QXyrx1sfi7`g_MO{lDGu-BI~xLX@QT5=vUy z69yN>&qli-pNQ|VOSJS@d;aiEmLI^EwV1gBKVV=Vi)eb>dssT`nflWB2W`@(rT8zh z78=r#kQ$%RR41SO0#up7ev&QN9!~>yQFY5UPAXtuI^hohc}cnww7^jJWfh#VyWLSIj_^w<#;D&Q)*V|0hhV@|hU*un8pQO83ERtdr_Z8yn^ zHM78fI7z(V7`TOK6Z_v@UNSkTiwCW@Xkp(T8Kfe}anaO{nJMeg_u` zA}f2Wy4`}u{anLrinC2bXr*J8kga9G<9E_CCg%u*v7C89uiKAWfcEcdK!pNnEi5PX zemT|+iOB=(`oh0<|(&9Fq9dB`x}tjXz8hiMn8qC zU2HyUzj08C)XF+MsKh|4qVC-j4kLP{hX^>~D_$x3iI^_H)H@oFQk0vC)K)F3{x#HK z$iLW?N}dcOG8SUM8WQz$*qFDI&!d``VM*^AgT)FV1%hQ|gdih^?dbnL8*KS7dPrI^N&q#wJ z-|GM*<|uX_e}uuQFmIv1FH^kkcZyfvV_e1!#_2D4W_m1k_t4uzE)h(rMMq~-a_o;z zTI@2^xyCEwphlkM71tD3AVsPiBkrw#0G^Jn!T1g&?%Es5>$VSDNjrD2lWK#ua^!Y( zUM8r7vjnU>Txq-HRXPmNH4{$?^geax8<#`ZnT9t;EZl|cy+>U8<;90;Y*wv)=(a2F z=Qet$)yMs(iT>u?P+FRon~@%;U$OQq(+$Z z)a;Re#ndjLS=9PO01UE>b;j-=j)ibQRy!TScF*m#t2hAkunQ%MwS5qxzovq9+!EYH zG&VOkN6V|lfrniecI`wO_z{Z=PLJ2b|3WlQ#7PXbrn_7S=8mpZ-P4qHv$HrYM$+X3 zo`?>Bmh{oRLvQC~tRnwJ2YBm4E%m?iIIR9Kk!4&qq;_N*Z`C`kCYBsv)U4V4$X@45 zC=&{)MG3*r9`FTu%PWLO@LM%!h%oMqQ~FzHh&GZimy=@qUp;RQw1ZSJrD`mGQ*VN1 zuH^Ur##O$i7q=-jh*M{bg95K|R*R6^1Upo~Ma@A{J5*?d(D8oAeLpHiVd)YclO`(U z%9`25;NuFl?MExDH;Z>tC1lfhNNO9&6Q>}mJdGX4eIB4}#iMo|qzOtSR6;CFeVN|D zYN^`J75ctPzQpI;N<4dNg(u#mynsUx~ zMsF`Xw|kt=D5i{TZk9IUb8DO!>ysh}K8BAtr||y)J7+cg8`LKqdX@fr@MvIg3>3{G zg&F)bp@1P>!~KT619K$WbCPDBv)e@`7=(i2b6SnFj_KV+8219&7uwJtoo*PLjhFh! zDf_p7IY=xM1vWouLVxse!@w6f>T&n_IlVemB=Qf?Qd5-5X$2(PBb>(!kNr?5_lzpH zpj`o)ZCwzxi6fpvhM#lvgqB?LSA{t9;vK4q&C6&Dn^**9wD((D6iwSLezYB{?U2-@ z8;wAPMI)!?sG=vvA=?-^Xd7ZiqhSKE_B`VGC>mcr>(G{wMgjC z*Sx}uKoF@yA@Oh>_k>qy=C!6cKq5#q;QIIEuptWE@Mgm5x`AoykD2#0t*?CNIl-*Q z!=6A8cd;&K{MY_DMWwYEtlDj05IB%T`%GrZ-)dZRhm~XdN%NuR!vPK0Ad65CEzFW# zKP#^J;XBsBN@r+&UA?lwI5`ZjAi|>2aEf5&rWx9;eu86XSMOR$Qml@?<8K^+)E&s| zmA(ly;!dMKwfHto7V^wjcI%cKv70YixiFNorl=ZyrM@Q>DZ7YU<4DT4Sd)A~u-0U| zweB)d2#<))a6}4}j5?Ui7W&h=ELott7cI8f9n|#m*H<((G)jVKAPp9FWACtp3sJTU zi77)gBOy94WER&Tw;h}Fx1@b$B3f_6SzkQ%_qe3e)YQj}2e?AD6A7S@;d-8viAQDaubfl?toqq0|?qz8AG(!py z=qYVY|AzzFqU7wO?y98Qt3;R-t7Z3@lj=^rHTMes@|l!rgIG|wIy#Va1(cW(Q=-Nh zNC=lkw6pKORYaAE);sKNFDsQPX!Z1J%CA!JH9I2Sy#XcPVmpX9KN1%=rM4la?*eILh8N4?vDN890rNZ7vj_7%S(}1Iuk#%RH zXVbfMqEl5*U64QxfpmYT1fFiVuf2F-C5LlTM+nHF;CA?Fi7I~$>S^!( z2O9P<%4BeF%h~W+qr2RJn2A=uudQ62LV+-I`&;0kd23*ff z-3XU2%X8JN{&X}jSRZyifmPE{4LN?Z&*x(>I68n$e|rhnVx(%H#9ttc{9S)GXr(do zC9H@>RN1dG{iw9cW(WK*!H56Oqp@_H(7JM!zu|&1zd3rJFz6Sx=k~ycj|*Ac#h6#b zM&Cv$6jpi`0L{jCCB*)|lkt98%?gUrh1!}Z)5F$42dXeK+MvCk_4sYEb#h@FTPP#) z%T|=Qi4LR(?mrpvMsbQuZZ}&L^t0;mnJBe7ZC{ek;o`!8uOzWC@``tqXJ_l%P))#)s#;DC^TRK0}3F(s)#{ zoBhr^CD)(lSN9<^tpWC5i0!vj0h3tiQ-uV$wWZbX!A9xoTZ1=?eBc^rZkHzfH9i?-%=12HE(FBD0p9Q|M>)jx9Bo zLuP91-=SHSrbmS2^sG_hzs!Mz&$9to0mW>VdK@QsX#v}V&U^~7SkCJa^fRr(4IC~D z>aVo_`4EjY_!{}z;06Q+*QZ1p``!v1@9X=fV8bD>`sj26FCn(bP!#fXuqB>w_6sgu zNp#8ro^^aG4Q~ve7U1a4_r#&}PiQmO8t-#;n|#+j!IMT0c^(ywq$!1By|xy9)$X~O z2UCrx9J_6q-`lj@ZS*hC=VvZ;ntXy;+LoVCX|{pF@^71=jas(%8%-Z*>;vfb>;opU zl^@r`G2&z-n%n0)O~hmUwAJd6|4kZ5++R~Xq<%$2L2pxIB6E`)0)=Rk9@ zD*$zo=4LorQlz1#FewPHr z?-s}rNvAZn*p%lIf|fB3ClxN!CciZ+4^=v0TZ;mFIInJW6l!g4&Qk&7SO+)g2G3(#G4Ig6Z)_T;pXYxA}O_}E`{|r}nkK6s2!Urxk8ftFsA2C1PDJn9z zi#tZ1A@Ec)kI#otxhpI)?Rqzw(V$%4f4ho59ia?O)Lys;{;Pt11v_Sk#k528mrv_i z2By4io)8I`Nh}HavZPZN>2F{=t5XQpOU%V|CS`6#nv+-z9@* zzvb2iy>80u1VPe(`A~=j@C!0c+-lkT&fh~Jth`daO0u_e}Wo#Lqb~&j{$br z?g4L+9R&3cF)wzScy5_hLsChP%7#ZB%in0)rYJ88xngscm=f;>uN%(DYqKfB?yG)= zodXGeZ8Y5&hwwn^q*gcc_%GoHsU?cK(m1$dl86U&q8w`DqpVV6)+u{&b?y6a>>3Ok z2JylM&&6*u&U;qs(j}>>dqcTmcIT3Zp5zsAAc?GU0}~hHnWZb1W=PKJT}>#xYN;@f zHYgw#cU(GXc*{`p`8bG7lGb~pa5cDZtNDC*{;)@N0X00QYH`H7?Fw&jS&_9i=}v>n zVZr^iu(!%?(OOeg&Rv<@z3sLZV`&u-QmYB z;DlW@#M12De^)@WRRjI4F}!1tdqe-)>{yDwtT-LwIS`TI#dZVC8WKi{thg_CG}^{Y zvt^yP!ugkRw082sUl1 zN*Xbh8J`uQMz?sMiX>ekNX^B>y zN}ng2r|=s9n4Nw^#Y+iA`iI^)5e5#9%yu!cbV{KOIx%ghOS0ldBt9O&1^W1o|29>w zUcQTm*FUfx=N*{P|0Vvl1TF*?qmcqt;jwZ<{wf~=j8tish8wYm%obILbRK<&w`p_i ziBUy`!S2fpZ`>LShu4H63bQTRv4rEsTnP`{xTcnAS@fHq3SPsg5n46l>bg9(O5@=5 zx`9o;l{3|q>|A)B*!j%B!hU%@kq_4qD;lq)8K6x-`gcE!G$Pu9%-^hYW;WeVB!+8z z@|*U+j86akfg`VupTT(c>tmC-wO)_t9t9gKHyYlGbh7B8->M)pAJ&q7f~3~Ur5^sa zm0MwQEBbK9E4UUPgSv$>59h+WtH(2#L+{(2*g%Grz` zZuYg?_6arjbJ}`fI!wO1`8)r?2Ko<>iCx68kh7I9w)!km;G)Nk&gmzBLxH)V%0r$z z(cuQfp%A?@lyXl5+9)6VbeX7&0(SuM#ugejfC3nYgj|Els+V3mxG1p?6fDpzx*hpk zzw>F7yi~gQ_*Rz_>#GhrC|K)ir>r82+X|uH5^2mWxEU6Hu-g^Y1 zlyiBarL$f8fcyykq8#Ma#Hn9>n%x4XQuZ#^yDuI5d0~3c8D9{HOsVIHXP6bz-XiL+ zYeO&1VMcvPz73;8rLLD`r#^cI6`2V&jm%lBNpw<54fyX;{!uh~79K^L^bpn9{1UIw zYw~E}By`P5KW2N88`Ud)@u?h=DP1^<-WiR+Rn!~rPrNrf0GjREjzXM%*3$}2rqxLy ztNlFtb?;JTB9Y*qo)yJuBC9nUrUm*_PvDwcz2)>CZ(|V+T7HN*1#cm4ojqz-}-)39Ai%7`0c{ME0>bx*V3fTW)y_b+eWM?=`$DQ#ZM1e{V7Qg;o76OY1ak> z9a24Bo6U`WzLn-Q;J=#r+B{9gvrxDQDrq(-MW(J6Ujxd|^W7!=_;;euf52f?F9US+ z48wKp;mE^*eqJT2BZ?n0KCgE|$XmkBWieuimFT!&J z&;jE=!J3s87ZY@(6S?XPVH+Yi`u*-u8MoO1){;&-Krn;3TKLTr8blCWF)0W-b_WZK zJgGq;pE>{^Pu#1&<_HGqv0wjFrCzyO1T`R=5~oqy)mHE`D* zjhJmr>6!YJRI@*{)Em$>7dNi7I9|kTncjC$Z!lvaN0RfcBGR2UTw^O&-Q#h%7O&*9 z*cE)>6Kw(G0iPG!ecZk&Vx2KK53Sx^e5qv}ho5@r4ce}54%&T#^6skmqMuKDGIhkg zcXKZ1Zvc2uRg{OLU0DX++DTKYe3ve=Mw`k+*Fv$oo4?Ic=rdla;Z{q{Z!8{JL5nU$ z&_5!U!Y%WKMHRl}CY448GhT}!EgHJ3nhm#~7^QFiaklpzT1=r!_V>$Fema}a=E%I- zyy<>9iV1#i3vuJ8$ddf`5U&jhL@)4|{&27a(eb9&DC#{`p&1DErYNk^QHqt*6mD!?wsLWVK6KPFy9te ziOP&Ci)zC%Sz;{i$d7-L&cY1-ktxu!r(`pz@t}C|;^oC(Tm+fW_v(v0kUlT5_%&Nx zAT=z`6=feL0)xG3)j7r^o>0%`kV**Yn%39U1E_0k@)dXQe9{rbp+n2*vw}9S0#|5q z1o~Pd*Cjq%lZ1pzK$a%+G1F6*C;xb%euuud`BlpV#bdPFu~&>nlgbMpyWuyqoOePi zH!9)-*A2b2wq{Z(MH#h=sVB)ve2cWSTH;F;6B%kv=Y6$0<#*Bt(@_p-U7Tf=o|-{MTNXPN2HyU|}5@qJnO-?u7vk92!J@t2nUOR?4VI^-o4HUl}F(Wr|;wyhJHUySP=9cQi6S7l3}|8`x*#zstJhb;_cRc=>#D+*@nwEqQ(6O z;C{T5d7M{M+-WCiXSgcDH_Ng~sVrR8*Aq6Nke_{bC^$CJ;h(|V^|!V77p1o?8YS1b zG=)J`XwyVRTC*<{(i?q_EpVRNuG>GndvX=Z@c|~j3IZMH)UmxPb z;BLmIKg9;@J%^;uyQD0sRm-h-TZ$3tQ{56FZqwOWrYiCvdK&S*}dTdK`A%mmMl<&c1Qvzr=D8Y!Z{2>x`PNQ;T?SU)Y8w zZkPp$9rf2`kAt4*KbIh{QV7|M>WNT}8&tlwbHNgI&R(|^plR$@jI_U9N8q^=*gcZj z^27oWb?3Wcuwrn&^WqHd1V*89r&;@e;3lyx!MACbKXXwyPQVK%KJV4me`6_q?^zGo z?y?s+nLF&r6zhdH1iS23@`#X@u?TF%+^440lCJKEOy!xbvyL(2i{`3nO`{I43KYo- z+7z>MKB;CTjbIEYh$@A3ahSJ8%~n@Dgp>k3gU^6=YNT&tsn2*UPNZmZYt!jr?7{7J z9XwoMNwL>-bwxI;y@P2(bm|!md|aL=K5jKJ`^a~cLqFgt>&MY_xBK{)PdogSJt^54 zqj`ArwpE6fH=nXMUYjxFi8N{$p7!2aNZ2F$S>iNb3}`CID)a;@&&q0aWasZO6O+6f zeUX5T=r4>2+|HSc#Jw;erOkc3`ZAoQShiyzEPf9KT>3Y%nTg2J3j77j@WH>XY00Q# z4ZJP_-Sr-)W-b>i;+#tI*cM_P4WEOufn}WT`azXqAshhA!if0*zYLoT%hx^3GU18- z6UE5A%D=szgLntF5y?7rA3bG@v5=5$c*n&sJtW*gP&uMw^CdI=ZNwXdyl@|loFFKi z>J(jB*8l@t z9W1{3y>+=1MbZ_R&r%+0wCD?SZRy69IX&%Cx8_j40Z;mA1*{Vo?K`k48Bp9YHQ!g^TDHH zGScc!>5|3s zARL-b`P^rcl>LhkLIEe8H?XP6>h|w5pwmf;nwy|GPU4^lTT(U-T19MB+1bn{zOZlO zIsv@nGEI)VB-J=_J{H@^7qvR{)6@S*xk8fDO8ZD2^XbW0y8b5k!D5^5zF^WgwL2w< z*PZR&nzR#l0o}7hsu1`d^GEb;8z-{y6^aor-}dD5$c?Ut9GQd?g{PP!aj`&PGTqteW5$ajg-GSyu(R7gt<;Pv%x$SVI$FwYd_*3; zWlpU0-A6B@N@5PG5!}yMV;Y`R-oL z;p-_I{`c*m$^S-m*!K?2{Z~lG9n(AS5{)v@gYJy@7(_xa8wX$BH16f;&49HMiHne2 zH0SQ8A4@U}k#YUH&w!I7M(_VV-?^c$VI%kTux0abY z0Hb5wr@EhBM$^Q)g>`>!%yQJIfVOF(Ey5zd98h>W+&JJud%7+valphZaMz z&x|%FRZzM-ecKDzLf12lo1pw3J>9waRCDU?YO&VRU#A8DML@d0yp|XrQRIpGJXZbvhcf@N%OXZp5-)Zxnyvz1#yk+Zwi9Xx?f$d5?F} z0eb|5Kq!V~0E=1D%G?~P*LnZV$_`$&QDMAUbu zw+?P)KFJd@9*f*j1-_lmnE_6tJcxsmZ%0E9*~=eWV*ee6{U4Ozi6R~GP*{trqin-Q z1mB|zdfWmivVils#MpbLx>)t^`t~4s^UdhqGhQOE+62BVLGP-%Z-xn?*Lq50{@Ni+O_1%j3r%V(c28P+?>U8Nvy|mFm;M zb?iJ#{RNkLMBZi?b%Wab2FpDWyXD@12fH7L87iH~LVhtTcMIO0x3@8{2b7l>qxcPh zQ?xE;(#GT?(gQ36SuYFI%%?l^#Z$n0%UQ+4(W$P^Jiw4uZ$tP@Eq}Bew&~H~)O>>W z8n=-B$`QcL%s=m&EEQ$gnO&NZu9sWr&!mu>+Km$>8mk&t%vl2E40^ZcFvNTTAcu-Jl36Iu{m#>Bm(c!+_<@H zx^Ak0YgI})=f9?_g}gxS^4-Dm65@0mI9be*U(8ZLWJk}eI!R(1^)cPJ?r$%Ja2vA@I|<*}%) zY4xg|w=f5GgMC~B>qdvqqT*`vzTIGhjBaXjneR+tDgQKO$Lm89Z3 zX=kchv-4{NvBiMvnMyYhMdda6GTw@)H67Mq>fZEQ5RAhfH-T15BT=@s!oA3J?Y=vQC60g=b!9}D#@vQM z{*uc7xUMhzvOEOXECSv3{-t!BEPSG+4UlSlx!Qp(lqGb;y2R-gyEqBpn^3^HlSpc5 z+nf}>Y7QSyoHVvz`RW?qdpvbj4Yvw%HLtyR=1_b+(P>3(lV_dJDlG_*-(v5mBm`Ii z)&G`t^gw!xJ1sNBZ2+*uZbtb@c?~yuEY>Myj{w1Od^bO3XLeKXSV?2y1!eI3;FRL4 zcfVfNI7|2VU&XK^CBS`&3*_5kllg1b)&tWg@#wC7hCFln9kP>??kd)x;Jjrn?Y^rC z@3o7Cpa}-r(~SDNKQ_}W0VixYl;vOYj@T875_mn<#FGKnF+F67;)Q_<)P@?ApgSJ( zV$BPGLGNZrlB>4@nm^4y}!FXU%17UYms=Y_6Dx(=z9 zXe)SKsZ3F=o_|R8)`C`@^2_?i>!LeVGq!+iJ6!Q6>6YkQW43aR*bBNaUpPA_QM!}d z=6e7H&lc?!saW*h!5-Id04hW!xC&=Q`Mqv{rq1SH5$l0E5Ojm%N0Rh8^>F^&wHVnX zJ6YP`WEB){7HC-|ZMxFRv5#A(!9`r-nQeYsNuF+d`>Vl->X*yUC33lA9I~rUpE)-T zWGQzae$Y^hT4&{ZR&Mx-sCLZUt1nVp-pyD)Z^``~RMeF|wIxPgB&mIXmj1mj73>8)|3v{1M(DRo*J-nOyLmTxA zulBm4x{Ew9U!oL(?hhfovizz&{jhP=hmH>Y;44Mn<20GdgvTMpomC*+T{V$E0vV1f z)ioLeo>8bb)pN;Ea%HZ^)GV-w{ARewuDWWaWE{j*3T&JKAGU*Ne^Zm2Q!SLZC`p%m zFMs~l(dS_kT-Y@Y<#cpjoI|~5kTp`(eFpmm{(nUb1w()*fXxYln87H~ z7r8LOsfpPX#G`O;(Tzx$;SE(l(-2J6OR6T~vO(^GzShSp4d!Ur6U^C}?+g$nAscgs zMF!kycN6_3V4dYqPHv1731Be7TK@!=4!QBRY zZ{%~~%)8BYmj_BX*w05tDBSI3cY;FlMh6fK;eoBU9g80yt;}cAb{jnpO03|PpT#@u zRz<5fqb;MZP*Kte4e6ngHHnP%TUcIn z@g4kn^ab6Po*W{39a~`58{$@zXN3#I^Hs@;$YMiK(;&Sw{IPvoyoko){_C*?WdWGP z!J6dOQlsYKkn;+ zQg`&B1Mzbzt9*}?!c7kW2T;Th8PVf6ZiBB`IZzXBDtJu2b`5a65KMoz)n{R$q4q|N zEL+B%KX}E+oHU{Z=GDZ+3=tNNFsw_A^L|UwO07fZ5FH&82}J5x$%@tieY({AwRm!d zmyohsE)hH?U@}gPfN0bN-aq8j`=<2!#asKEv@K(ES_vnQ8IK$xO#pelWZtz=w))&b zT6uJiCBx%@0Z$YFv7SRKzn_Dk$B&>7k1LwY7Q1&rwOm! z$E=^A@2U`g+By8#olEO86@9uYFNGqhs$DUJR?D4YZfOnVY+O@D^>s4uc}GS}rMpfg zyFT-7fmpVtOytBXyQmywF3E3OiMgb72?Q1*$>vx-p~Z$Z9ID@$`W%knoc`h&@0kAG z9H9|3lZm>a;XJ_gncv8(lcI(j7Lwb#TK0l`iw1w|N}XtSG8o-_>qG9a(%SJRSq!|v zsuUjXW;E@pDC5xSdGnMMAN$0r%9ZM_>8l4GBC59q2Fyln_O1%|wkBHQre8yQa(oyZ zos}vUqa%C88_%^756V}S{%_kou_N0(u7~jQ|3Gt|wE3MteQ>+Tz0zr)P)93JUSDY9; z()N*1trK_R32)pveK)yO$|`cuRKI>FL9QT|u0wq|w@>8-zsU|(f8a^>E5u7Hjp%g6 z4MNH|$CdD_8z}$>(SYZ2A6jZsVv8PI7rkkot79unt3;;EYW{C0d5O8a=4<9rq*nRa6TjbG-#M9qUdbzW`Td#U zaNI|ymJ`}y5VWpVx$rP%w=rn9!)Te!=kHIO91KTG&!ahm80OyFgXw*dPfj8Bi*JF( zgzzutaDo^g<~zax(2}Gp zg(bh$xl%B7W8+Uax%mr#Zaz4)w`uny*%vhhzK;%59=&jx%)fHEzdd${*4$K~7-G}D#Z;_Mr7vq<5x_Vo8Xlgs@8v&^C z`OSrNb#E6Gni+s?e2&=I_t3O_=ovTpt5y;J2CKEAfsFrlD>&0b#X}M8S+#64&a*+& z+Mju*wJ#M(ZhtW!#Oy#1yzccfz2gj(Jc!#%^}wtDcHa8l_V=?xe^aved#7j4w#y~B zXv^HwO}}vKf0$P_>HKY1nsf+z|3|Yf*}2+VYQAvOJjxr>Jb-vG*u%{~H8 zG^(+@fdZ_Eu3|k+(;SJdY}Nx1;V)54!x-=y`%UBvAbSD{V>=^LJUGn;95MF%+rT_< zPaq=0`NwsIbsLJs0@|fmzPnO?MP7*bW6<-$LabUC18?|8Ps9aFF-%@)qYS+;N;O&o zULTx%;nvcrUj{N2t`Ctr-6KrPU!5>%mp?eNIg(Jhd+^FcRz~lg|5b9py9C-Ony7wm z0DL=cDKAC!(-@uT3To(W(^|eO_63LKgD684ktJv77J7B85@Ug)NUz9xV zcHRv6By9v4H$K%GPvLX0F~90j7|h#+5=&WFU8x6IR^Zh;i`o;HroItiG_O#qb_i~s23!*d826&>o*7j>5@c#&`pFGiW zer7q9Uv~%`0xTeDRu3eUi0(zbm!yOXV=iQI%?R8*Pg`K{m}%YtXTbeh-X^$?=u|M` zS)_VU?ts{urPJ(7ig(Y~V{3e+(m@BK`p`g>j4s#pxH^S{)u5efZqbBNSD!<)QZxU`g;9e;63K$$${ zx;Ep=CN%&Y5>;ch0LLp#RvuX>{$9SOeq^c!MY}4FN>igfR_ePgFPuD^pY+Yhs#l3qz4%oWd=Y9rkypW1^ZF97%M5pMqq3f%HrmIkce6mD@lzI!WMGR zOgG<^>k=!l!KxPHi)?{};=N|&go-sVB5vo~W&)8cz;t%dO4peg;vV&S?Vzt?4sW%l z$1i_}afkU2nmlbH$am^I(3vv0)f}%)6U&(V9qzW>p~?g#$m>V!y-sWW_E9nB1nc7{ z=wv6!5&lORRxTX^s5- z(i$-x3dn;$M%rHkQtdI+%C)23+)S%;d{Al^3ryW(xwa(2oKeJ6E<@dt+KQxt ztn=8D*GJRa9b!sL<}{eVSS45xCd{Ab;%wHq(m;)n1s;`vUu?ut@;R&NHyl$OhKe%q zhxBIc(~uwx=rShC!p6&>)Gm!--|S403P#UKGc$O%P<~hk;Su7Z^TGVf5b!uLKbNa5 zePhtiK%2ocJiHCg7Tu-^lGN%+}dbjP)_r&95wtau-qk83>-y-+Vc z|J}<}1D~^6og(b@zxaCZuq6NgUmRRiuFA^XS}B#4xy6B$vn=(Ul{vAu8t#SSL{Td% z6*V)rIHRTJ-YW|=75Bm|h`0w%6#eD*cfRNQIp>daJ^#D8F1Wd_=X#9SeZ${sEPP_C z2=W!36oZR}SUc3%bkqyaD?`^bWVv|0$U|e8VsLm=7bv{x->5$He;$z1Iwscm>D!h6 zs{})$0NE{}Qzxfk4#@()&akA4fLH!**+(buMG8M`yLDnRPWsK*>2=hogkzlyZ*pb` z+s(85kas8UOuYVFG-|ki0!l4xbic0EcWM+BJTVs2E-qi6J@v?xeh4b;x=T6?~PSSVJ-Qx|}Z7vq*X~Cb{ zE>^q}eQ47z_EJtR&Go2dqE0gq^?2(Dr`dcxB>7q1t1wIU4{Rq*<2@SKPw>y4T@gPc z@eZSZ?%l;$K=yR-Nu8_W91<@LFtEyt*H66DH@h(VL31}e&(Mae!W{M&ex!u#t9ZZUS zCqET6$ED8k)e#P5e$eBP@JA#gVk24aE+K8b9i~()-9UGzce$D;F2of=XDo1AO=HFl&N0DGh5Y(Ky2C_$)d4a4&2beZZMqZK zjw+H|s_c5{1ac?7U?`%jg&_!X$5Wi8_v_ctg1T+ebgTl_CEaMKy51v5%@p39A^TzG zdB>W;Cv4syA_1oSZ*|gEcTgVxEDd`;`=$<}qyKsHx+4A*RmS{`I~n_;20iPifbAve zQy(pB7h*p@a4=LUgj$DHX@X{T_ImWl!eVl!7wVMMba5_uB!$T>lZiB4%?fhmcK51M zN3qM$6Qzi~Ad7-zHw$c)HQ5|(3r!^F)Yn3Y$ei->(Q6@w75-iJzNjFC&)l|}Qhf{O zdYXxk@51Ez?&GI5>4n~oHvbV*j~#6wY(@!E{J6Z({|>%>Smx>2>y2#sY`Mp$FnQ1` zNSqK>2_vX?(KxZ>dMpq&TEy68GeokHWH&n9gOU2I##3p4ehzIU56~s_+eSb85;>mS zCM%x87a7lbx?>MKP=7)g_!C$Iao64o{q0vjr?Dy^7dwE3s##&m)8et~Gv_wN%mMZ_ z_gXpEI4&-NKq7m~m+lz7s0md80489NJ3*p}I{YA(nbdF-L3EYPjfTW{)QQVhFvXXlkso zNCcHQB6+Zh*F=AJG<>)*6I!URgnR%Md6Pfuk2dY9E|Dsz-S|_KIxrV8fBI(nln(DB zSv40PXM<4}Iic2xOv#d@px(quk3kngHQSj6-#uBE>o$<05 z((xr)Gc{AQO{ng}#5(nyx>w7bs!TO4`?+#ydt94rYdQHhR!E3Pjh+JuOH`uWA0=dp zENrNdqT^AbOqw2VL}rzAH~Yx1m>|k!TmBkn7OS}NIW5Q;E7M{1I}L3a6M9rCSmS<$ ztpdW}z9u2pU*=|R!(h7XX>;&E@_efE#;~|gr+q11vB3Y+*u|?A#1|RSeAXo08Y|xD zfJe^IWFh3_0+9fcr+oRGWkQfYtxLHXP0YkicZY_-srw zLq)=l)_9L@AIhAZa&8eb-W1)bY&luhAHwi-07{nHSDeVaaC=c0?u~hlbJXIz0n#1y zN3_IS@e6xmT#s1u#=sh(gG@c~P!$oCUO)}%lIJ0ydh2ZocP!TuN5*`i5`dHzz#bI3 zfD!?|jE~Ie22+V6OXrGMUoJH&@La@H*!shsGuU(oh)Z1EJWQ69(Dk%H(mKu@#f3hk zzi%GSSkDwnk{8Pk}W3N>46*N)+%Tm>ywpp?BI~&^y6Izy=Fj z!WZT6JchRQtleOSEXxm;Kl2|CD34c4{L=snz(Gi!%H(D|wqO3jBdms+%6zxVl-3a- z02wAs8sL@tn!_DOHXdE*jr5%MxLu+0eM>Sz+$;(dps;&y%ix9s{I>EUl94)}-5~u) zm6Fth=n>51$H@)o1!Nrwi&2!fH!JxZf-a$l4JPL;M+x`uPi{onX8FI!xX9!X;R`6v92{l{%!(SdQzYT-Z8XiA@2JQ4Vq@0EWm!=x6Wiq#+^POe?EGA64pvM3M$2D@ z#=23p`Z%(u&7VSAjT#Z1WS`V(2Y`fL(QKv8FDJzY* z)rIZl7#ivqe+82sb{cwI>5YW^YWmZpf1#8bwfFl)+&zsUkXMO3cysxz@C+f%Jnvlt{MVHn8m7H`XhKC zB3y9OB5hwENBHb-_kff7cB?`6RyNjOs;lTRECzJA z(BV8XCM7?zriZt@?p=^Uw(D*o3B2^}o|#pBS6-e=^+D_4#@0W(&I`xqs6KM_7G1tE zdbKtlM+kHA(DkTd~l^bo%au#6t&A^F<_Z&FLEh^X4Sm)|Z9hJ_uC@()MwV@@QE64UI;)xyb9F{S$V{`*+~@%n1QTz^{G zzWZ|hGN1~d!T7;WHJ)tj_~#Wz2vNFG1^g=>cR&GdH&7&H1ZHW#-=tF%)>9Cv%gO1^ zrf~M4FAZFK*}WX)IPT@Y0g;RXRNATp@p9D1iupM1!kxdxUoC*wj%% zw1>A7{|>HnbIwk&Jv-%y)<4&D@f*dwv;nRCfx=)Pm867Z(=?T6UuUk>Fx70vUFOw< zA$70X)Yyzv+XZu9I|Sc?(BZ#m7#?l=%gi4CBk|%U8OZC`_*rxDglla|$h<@FN%2D} zh5?OpG4%5fx3V*6H6NwuYd(^@zdndm&^ko5c>-ie9(5ugv5Xk`!>g;rvgR0Cw=T*>i}?mYPBDKDbnj zqeiMCFSFsLiVfDc{n%f$G#g_b1ub(dk9Kc;8%AJOkyZ@+qw-IdeZ<7`b&kw{g7 zXfB(hsq#1oO80$C$DkPVP3JxI*$@0{I<^mZoVr8)5!-s=;g&|rKJ1GLVkqO5R8aW} zA#AhM2WK=H(n+=(QE@*11dcKaT)!}O-xb{KH1RUMepiMdExM3-Mh%Ns&r%(2=>cF0`_DSaY!KOD-npn!mx*Nd@1%@ds6%G^#gV73r^fJ^y>a*5$oY z(~vpunhd1+FIx778sG#8WW^J2jI` z-M~f{h23`EvF;1TJnqZDSMlEtJ(NCIaPMBNs5zz4_!OeYy2fvVKnQsx)8937t~s}W zJeUvN1Nt(D8Zi2DQzwN@aLq7C1|$E$eK8V?w%VM!zFwoNcx~|Kh!~@BtKhv^)>7(& zz`sjvAZ=~BIl_ADt10)${6E@-I7(|zTU%yGSp5}H6=wPFEZL5(A}YSipbqjwP-QO;gc!{aBv@0dNV9hmN)UQ6hWguaLG6HIaa zvWICoe_Hk284o#Z>c{`3a#WS?Kb|Z%shw8|V4kD&JVCY(_vCr@z5rg0zJ<`=!>dX1 zj+8y5Qe)_YZ~bjVlViMZrrS?ON6jOYeaao!E25}Z(zA<99*FhIn%5-6 zxTNVWh3&O^KJNB)UFz3!?S}?1ul2TDGY~6Zo$JrPZ}M|6(k5HN*@{&X-IY@nSBF*U zEHAy#1H5~8i@$+;c~xD9u->1x+6>Cy8bW5gRA;U|HA?rE=(3bmA|dEoeE~`@{2WZ- zYvP>vWYmZA{dW-2#W@vYUhP>(B{jbh?i#@Q+#0yS^Tq1G`dg=I310tJk6%C9NtY5; zu_G6Y9s96ehES!Py>7)OgnfdicwG7omuo^lZ#j;FHReevae|sP5(y1c>@pY2AByl8 z3`ye7+k-Y)^IH$U6h^J6WV9};PB%zw4RL}ubzZ3EtP%`xY3(w+aDPUSu@+tbZ-r%Y zqK))AP5iyvF9iK9Mgj7)AjUe^v`6JzcttEvC;!l8^|gZUD4JCiytr*~qA{pRsL(G{ z>#{3XkIKjF_CZLY$YMo68;sYXmL~qhD3v5XM=^_7e+m~bsioYF;IBtch)*13=DYbX zWG0^a)2E|Ad5f1r6T}+)xxRhNkn=gUaBb2p1iD)Jc%0JguBMH^o+$e={^z@^tFLB{ zUzVbKtIW-%enb0~`#X=odo%tB_3lWSf5AO!lch;(W+#um8(SR7bC15iK&<`yj2`_gSG zz4HL6)7OmerIMcV(jpVD!kNObM$vqGKpaM0Kv=@M;J0-K__G1POt;7}efbKn+R?QA@^Rlp-C>rI+!#puiMSl0Js zK+>{F7R7%(60Xi*^V8LGAHGmGt*d}D3y#O(`&GwwmriW;nY*ZVYO(K(WM&l7$Rcc^xE zm>aV_`Q3ik-J|rHpiGajHw3A*qz8--qUIpNGvoTi?x3-UR$g!BdoATwqwjeo^V+?a zSsfnkec&SH>2UnqKT+pL>YW)Du>r+*aG`twoR8~ z*4qG1eYJHs6v7+|=dU3s(PYw{`gq*EJW4$ldLUpafzgC0cO(pIv2tWl=DfV?eisXM zEvkzNIUwupZ<_oFw1y-Po1voW9v$V$pyxANjM4{RR+F49$vRy>I~t0^-&+EDqgyQ8 z8J7g(n2yB6#GvRHx}VaKGhQ_zanNw%UV{G_js|LF(DDbwL52VIRC`}+Lp5^YnOT-k zPnv?}z&){zo}|L@$!^*GP^qe@P>g2XJM1eLLsyz9(27AzDtJXquxKHq3RA#2UZo?s zgAdT!h^6oMdMc+R&iBOyPkqDLU4a{pI5?SnwG_(h>4z7q-FiCl%-ebo1)nQ=^1A*_ z1ewv5F=LXU!2FB56U3wIi^p>p3Vta3XLzU{R~=sKTjX4C<9~;z!fEloW@zkhS)0(D z7I5EE`BW6kvq0O!dqx4hSM~TX(<-qXQ(?SQcRhGM0BKqLfV>3&y%e0?tfd{HH-{10 zgvFBBDE$TM7#UFv?yh)@l^9G%0jAt2*ZmY-Lt10;24G)6D|GeDzLd2oR zFOMmYKf|u}E2;5*W@;o@gr4o$5XWi{^_b^-gg_afjBU@@&q9A=_{bS=hOQE@&UQH^ zgfge^d-$d^Sm*hjB9@Ia&zKcwVcARDlCT~7 zgZb4C{gIlfrPY47{`cS$)?UE+LwhM%7IV8aPDnJrIgcUsJghSPrT1cVfSR*g-PT~%v)!()}BND}*Rg{W8--`ohR`iN~^wda?4y(?zHys&Kj0@3zdeK}T28ZQOn zQ%kA0Qrm1{G;0Z=)SX(6APZvq5vIt5gUu^MX0x6?m`uSiy{)oisO1pH?`e2G{{lX0 z;$lyW=HdGKNq1H6{8GLQG&22m`K|!pRtH}Y8?yCc3cNGIdzEB#^vIu7@ktxWR;L(^ z^Z(~Zea}q$iUO{uB#;6t%#hpj6Es~;7=){jrS_LYRX(m>hqRK{*JJ2o2iAU&#k4R`xTP)DEuqa+a~}rHLT!Gry(XHsr6Z~V#s_T) zim6@@)ZA!`^i?M(LKpZaD9w{eSuM>~2PABLQ6z^Z9lS^RW_mPo$$vh}Yja~ECOF>g zT(5v&ZJL@=*Es$&F6xojr|>2ldU%B~ZrjPUcf_VV`2V6_vW}^jbJvkY=7P@zEB756 zn}>s6Z>^Q%NKpFdDGcy&(zi>-EJ-#If-QU)_9;fnQ0)sq4h%~b2xz$!%Q3te${S>( z@>V4Ub0bY=m-QB)8}k-=5(J9}Z333Au~yy(iZ4uapSv6(VhwXNyvoB06GJ?1JQWZZ zMe5_9k{9TnnMNk1y*NxXN(EdvSf47L;bM~iPP?3`!lMKhrBVOHV1WFkdhcPk8|tO0 z!5y(?di1C`BRGY9SA`#Pdp4fm_=Iu8^lKr)oE9*%QA}KZ?qq~;x4e}AtK3K^D7HfD z{S5*O=0v5Gvmxh3tZCqaSG@oq?rNTLjctEtn>JKUYM|$nmE|85(RJkugm`Qy)-OAU zpP7w7vcwvOEL@a6%K5}q*Tu;a2;pEQM5TTXOl$dJjZWoj?Z~FnCx%N`6Gz!eRw7h>HPdZ9yelZ$Z@i>ZAKN;lD$BB zr!p_TtjsYu^|HK*^HQf7s@cz#UeHNW9dKN}xA;x=vl!_cTm-CT>mZ^s2rBI{6=fP; zGk^Z+X5r%zOHr{|o$67ZgyqK}fo}69nLHYwn?+SG8Y0aa7So$`lrfsC*Q zX~Del`Mr#tx>a;|!`=%$bwpO80q<6J^^r;HGYcuZ`OCn(yTXjuTjEcumg=pCdOfW&F|_7^zWt$ipN(8 zbORsIBqqYEX__f2FV$(Rx-vCO#Xtg(0yx2Pnmtp3T@?5f^R>|mAZpAqb7gJJBIQ^ z@TML656_=txpqN5xs_F1uHHM>y<#$W zUtc6X?g7Tj_HGwMGIW+;Hsw;nDT{7LMwJ(d^H06|PAW0@^K9n1DR^YLwC^;?H$lghin&$6Y5?2Fs2)c90k4;O&VWcBZ6 z212^~$3Gh*_hj5__s3U7YdcQz+}Z9n#S z;s!&#ZeBZ}cZu0l2RmAzkG?n|%Jy2RSL3TMOttlFG>aX2tgTI!u%s}e%E|xuePk+# z(Xwnrf6RK*QfvxTlzA=5Ru;WkUF9h<)-yCl7{2jBcfPS|SIs{^TEv28e7<|feAiPP z`aI6T^bIMwX{2gMeRLpV0Ulp<7)mh)BSh=-AA0w+n)(JFevQ=iSMQIHnRv$E)-8DZ z(?Tw38fNN8Z(+b*EplAl_1fRryQw`_|@rqP^jlrIf2l2K7~i*H4p)WQZ zE?)NON;edm+`Q>p|19uNo4Etmux2Ux%-yQP^Hf`Z%IP(W=l`iwFZDNioybTh*9X)o zcnmt&>`TzReSjP#EwMbO<}y zXw`TM2C&(0GYa;?{rETU-M%GJo@Fi2o+W!OjNt0RI8Rx(D>7VppO{EZAGU2cDQlAF@?{ zUc6|;yw=~cOkWPN%{+Q^=EOri5KNW^{Crf%SdR%&Ct$f4Mw5+R zjofuYf0I8-ZD@6W^5~HDF&+s}Iz1+Qx-Ca$(@>0LI2mdroJ(T;Bn_%{*0XZYi6_4` z3YzH_fY6>c0BY@~H}CONWA=WHe1)na|4GFJ$_el_f|+kgv|xUCsZGUlJbU)rU_v`dS;T%0R)Y)q4f*B>!g#nKH3ox3T8kLC!5-n9g3DR;|B zStXvqFGXjfB77tZNp*=pe)BU2`h#HOu8%{MLTcR}q|O zxt>UA7hpFD46CCwJ^zIwop(t5k|;i`I=#jrzH|2;jxb0Snpk`N9NZzR-q}DYiBIc% z?!5k-Y{^$!I+YTEjczMY5Nor78*G`rd6AZ;qe%_b=bdj9_H;FxBsA^@ZRUuH$j~kZ z4mk~Js40}Juisyt|ESt|?I0ojow3m0DVWw}!HUoBR3m3u1xR1fbMEx_&vTf6ThP;n zKVnHQ6HS1sttl*ulR!|=(=hGdT;OLIWd|Vl8&%`^= zQS6V{S~kjB3aW5kns&l%Uo1;Mjy0GWt;E&(UEX^Cct^{x5`hYN4c?0iE&3%hj%s*n zz10?u;-ObD?#@4`!2W0BC-NL?$s@AFZ;1x1hQ~A8^0uDF zZmJqK4;IdgrfJz$q@|ykbAd|kW%JiL08R?u!~nf_KVN2VWf6C)n*Hd>l{A~il18MT z)xZF6#rw!e9d&?S9$HB}a9#vvc3_wgYNycU?$ z&npKEeQ$(z9pZs*>fYLGylR^pg6fW4PVNe#Ud1EG;w1Dn#chZUWIHORDqE(}&u6Wr zqRKw1ZD_8MC2LdQjgB=eSyqbQtVe~=C7VINICb*_Homj#0rET9!8E{0!Dd1LWNXJG zHlmUQnU4OV-jsrJiRy_YphR;_E_cop3vvJ8&-Qeyid@65Klgo8eE(>xM#H`HX7#|X+@JNpogqHdfVRixf$3eF1F7^E z)mQ^(KBFCx67J#`vz(oE3-_GA)h1u4Q=w}`e3C=`gp5iRZc8T?8~J5=9`&QonA0cx zcltNFr`n;{O=N_~mLVg`i&}P*ANGaSQg;fE7=<|n{FjH#HmVYRWoRK!b+#|mZOu+7 zjpTz5+U_Xt@>ANIV}kr@WcVsrQ+}3X+u`&7!?yJ{j^u>So>mVe1Z;UbPJQohJOGr6 zoqEP=wx2aC-@pxf8LJCs+Sg0>bf8vIFJ1-S zo}vJs0`7&fG>!sPQ%a<|Sq@du0O3X+9*Mm)GrMN3ePMS$pfkFD6)O3gGF7MWe=Z0_>Q3dvKh|=u z^!ZJ&fp&-2S-5s4W7C{SZPq=~<4OFa?x*$`kLvG}8$Iuv4S8%S$nP84jI*JS>^^`f zid2gH;6~74485w}!2|$}Gs8t|w1}IPb;X_No?)XcqiSbl#qg`k-lA&u3m2&5Tza1#iV|u^=na=ZzZ`rut#kSSY#Ok4&)E2YkAloH zejNfEmWy8mj_7+8Nz!6mgnM1;Q9+kJMQeHy*!+;I~A(? z06$t&c^PPu@LPib@czz{==_s@vl=FI=aakcdQ-D*UQ)V{eK?>JFIn21kIE-=Hm1}B$a zz7j0BcA7W$L?yuKRO1^!+A&GX3R^q!!IXqqsrbwqEh^~Z8e#j4Ik*u*b_v3Lv?SAK#ehA2qOYJJ$^M_)%OOAyAv+w zY|<(59rzpN2|qGmc&jsYvMX?c&XI{Nlx3Y5K6#3(ct5PxE0 zR7=It$45UM0qwR98T8p2Ev>%Hx^WiDDgE$33O^CGVz;)Kr(}65}5} z;?Q0hqGdkDTB<4gNj%kuLe4{^NSS)SsKmv$0!0dnc&H2u(6PuWUG2czcu_SJZ5@oN&oADlv&G!53R_ z*j22uqHc{`sN8@ka`>gP_8=m32~_;P z?w$qzqw)xi_(_d#s)v_JRfz&0-VUA`8PikUE@lgC5xAP&)%JIj;r}(MA%e)**D3F= z8=8*JN5%}P%$pW}S*#R!wX#*3dN@Dr_k7S^Rb7;M{XflcDk_4{c$&iHKYjxr=e?SP zP4MtY*6s&HKIp4SK`(1tckMY=aljQ~C=ccg?`DMyl-YJ?tBE(*1Cwgx_mhuc(Tu_< z7WS}$`c41N@k^PoV4^OPuQjVcVJbI(m@upF#bwzTnFE6yZbWi`VY;F=cbURI3#S$? z{6VfH-2j}mX;qutBTjRX+35a;hMyyERXborpr#QYAE2U#xm%v=NEO|#kSwJ4(pot` zoP0KYE>}L()^Hyv?a=xzBB1Ri_n6m>>@s@S@c(a(i?MCIZaf;|)r$K znl9Yy$FD>>Pb($g3lq1#4Y#IH71$Ks`mw}`Zp21#fiHn*XFTk8l@YyRgM3bsbZhEa-px3+!Z`{Lou64PD-O^u{^cSaU1k;o0Feh(b)e zfjd!67a2PzKmNw5B+8ZE9Pso7I?(?Y|H&v2=v7z3-!I>R2IM28Y4_1RMe1a6Mcgkx zJe}`h#iDD>c+qa!)o~P)+JQP;C$>GSBP9@97Tj^}^s$w18ymU5lXsM2h`vtF1!Cz6 z<}GD&&uF#YpO~8Oj*i^Dk#5Kb85_xjv}={VnR@iuXUiR12&iAjvhX)(^>P5!=}uj? zx;^r7V$|T9z$Y!y8UgTLo19uMDm!S$m^NJ6oEiw;c|XXv--fN8m}%RdnXjj}zA!Bf zSP_}m-iP+AOkm|F>MrBOES5wKAp*7iuOj%x*M6nmnxFM!h*!i~|HZwBh2G5n(JO9n z3)>y3|8L%97;@B^=iu4!}x#_n{l9wlHnk53J? zr2agHQp1XKAt?ah0#Ks=3I3un0QMWy8l%f5d^K@Oaaen7B$9=qutGQYCXbktEpG3R&#$uHo`#Y8Sw7K}e~{}|(W_ITbI1&O zT9E-8t%iHbZ6}M~EAfjO(Sd3L7^%#7rDZk8m*g13(@hOgXr5Zn@9vF)YPJAPjHfM) z+sn00IT+)HJEdcwOo-F^aD%wj?#I&EmX~JI?b)b?zXs8?M2NU2iOI%J)8qPLeRKy6@dYyUGPDxoV0M2iD)lPCrsHzD}WNJt@>oZx8t$|MI! zM(`xjqZ8H>^tv`W_*hMX_%ZF1m`n4XJkiGWiCcG=E%O{k%`}Z?V;V{&}v@KX^0_GtD`4M^i z%D~qm_H*d#t66^`ROO(j?sTWyS46$z5RbzKL*ZI-=Olwf`n?AgH3M@fjRT`A zf0Mk!z=BaXnwfEUij=<$Nsq~iA8hzDEAjybX82|=eulJJyDVPTaK(E+$*8Ioxh|bP zpOfBF>ul+$F#4swYO=|+Ur#PD`JG0*7ws)HqQh-^z9GG^df{u4Zne?lAx{3mvqH3B zV)Z_0s9s4p;`oLL47J)S5cKGP9??nPU>6!@R+g9D`6L|XB9UH7sM!C3&cNpGy2dmq zF;(IJujZW@kuLJi&|okiN%f|ahIA6kRX`EUWJ>>Na;JKz&Mk>IM<+k)t?y{QgD zs*CgRVm|wyumhY@aTjsvGTK|v2cEwxo=67p%IQSjGG-kH?vC;RPTd)0-_ilRRs!A; z#-hG%!cC5u9Gd17EARRkvP%T;Cf#$3eyp^6oMYsk9a)oh*(v(lultMCUPZyEo5n{y zLc9SVu2PpIh8yjxJ-)>FFYLD>D@GWw<11z z-kGFY?iO8DW_de;Mhue6IJUz=sp0GXO`O)ZQSZKmbKt@j-RWQDf^-7L4TcMpVT=LRM%=y%F^IhhLmt8~ALv4Y6;8^oLO(!NT&TF$8}u%t>}9pzD*)ydI# z4$jypz4qSfD?Dubx@~*P93vcDLz(zvxNteuDrEA2s@%e9am+wuA(e1&_e5MhUiw%6kiX&Tya}Mh3`;6_7L{aYB1GDqy7DS_uZg-6O9}pB|eB zihag_0A019F_wz}ZooxAs4cQTyaB_b!mfY%73O6XcMA~IczT)b#pf7&010@SSlj(= z6#stdBdY?EFpvB2{gJyKD>TTkB;Aw2$p*9mb+nQdO2eNQ^4zJ#fEQR%n6wSrcb~MJ z2cwiC106A7-+~HN*!%m@=$)fYD@cT}v09zi+SkI4XRO41OYn~ft_7CPEx1B3bo5^0 z=I&Nu2($Z`nJ%)8I7q`<C)z*Vo&wg|hZkI^)PjA=-1B&yq!=h(J>ofhfGJ5vus@ z{2xzSQSRUcg+sqi``vdNRPjQO@3ym}TVG3h>zp&oTplJFm1uQWc`9{9e;M{$sr&te zZl`DyZ5B9~x}H4rN;ZX65GWGZUca{rTkL@B|BU-&X*`)Wb=nSJ3uyf07^4+SCxTzC zpC9qGhsGc>G7*NlAstUtb``m|mIG2YR#z==#njvq-gkAml_k?B`AqlMc{1@pbBJw=bT;TaH# zr$gwN+w?af%cjb~$YA>oDtR~YDW3jlk3OuE$VGGFPq+PYKvf)9`8cOKKa5TclrCR6 zmbbGm3(aN7s8;d^?zO{%TjvhDMpW3x`?3%r^S{dy@^krp^6%_~TaS5#3~5xWI*j#= zR7saLTh!3o&1l+83HYMr1nM)cmsa68xc8rIO4BiiF52N~Rrg0|j>g7lbX;WkQZaVOL!W9DDSbu{?RWJbjw$B|WpS0OX&^;sLx3sQp2H0EtbM zKiM70Yb=g|?Wg^_@a|7Pi0YJcyM3fcISY8cQR@KkqtRY`y_To-?kz(Z8_?|`w=-&* z6jtGMgLR4v{cUlY=P}%h7-NJcDs6#2|J!oPY?hx|K6}ifFrJFJi$7pQUf6x830Q&x z@9V(?(2}fEDZ;3-D{$3p{vM$T2A&%|0IGqrIINW;-^a&bqe2f^UjeAXffzB>?J9M8 z*aXv;--Zd&+-(9ZbWH*%ehV=HcG`t-dgimZ40NXN;~SpQ5h(xaFh zOl@{ik1sv?kivdxK2_ZntW}+1QhlZ@G(&h^t$MQ}{klvc#4tNjQvk&jTdOCGpm9Os|tlDxBDXtqVSk&GmkOUFmM(*_}YA?`< z`On37rlWI+2E0pYN88O>RaUJ_DMq5wwnA7JjMosp1-e~iyAyT9rO*)M^r6mk*)|?C zfr>#g{6p`mq1wj0`yMvmyZ2P-%aQE>0;ONWkff~&>rpr3v42vwOegN#Ta;#=Al%j0 z#6lE0Ud6Zy3u5cIkR$eqB6X`U+xn?#h7B>dgfLb`3zB(WGjU^beb;qJO?0{K zpc=}11A?K0;hi0%%QqQ~?nvn8cH&7t^FZrgYj=Meg)sV_YmP^J(knj$FYgQCR&D94 z`y4BpOMS?EPn58_-`8wekNtnjqsJNUqUo}X82y$nug5C7%R1X3U=yaE`Pr6tJ53u9 zactmVy*hOvPndq=w9f2VJKryIS2gJXjcQlTsYcxKT;C*#_)*Wf&fo5Kpl-WM^C0V zH{Kcm-q?C$ES4AAfAhoD(-8cY9!;NPKyzZ~B}(jxs0WEJ8n*?PYD>-T@-womRPpj? zR=GZKRrke23{LEDEkdgx&F4`5b#ts?Tc&7mT6WrXm}tRlEsevqS%>f4W1|7n>MsAa3krj4grh1q@K%O zhrstKYHi)mK67$yPuehE7nTo@^}R(^bSjV?D<41633{mst}&b`f3OMJ604{1fAhSg zNH(O?fZlqbh|lM>r+#^y;r(Cv|7KYAx@3al*9X@x77jYvZ&XX!u?C#P&>vQ9Y0^EH zkiUgrSu4h_je!?*qs8{k_g^v3Ufl|jOMyS2!UW9+K8MG3!$ zg`{ntx8RMej(WtHJ-ci^LpOZ-&kqiq3oow3?(5@#7Xm`9B9oVRTHu$y5QIIgKSr+p=+tG#A&^NT$^|3~0yqkyH9{D*TT zZq9H)w5s|4L60{L9yj!*b*nlHL6t?tDG7vT^Fd}AZn>bGc!Qe_7E4N77k4Wl!p=1E z0oe3Ut9QV?dvuzwolu{jw))C-vmWyVGXX!Gg=3YOkb|I~%g2^|fKc}cP@Lr{BpAaT zderz5dWSOXRyOfd3Y7Wu*HCKrUp#rFfjNN-AqQkNdp$$L9Y#zRXC01kDem51^apV}Zlc^6@;{((&HbaVB_ksUHOI@vRFqNwOa!v`UYq0>qk(H_t~vgD5t zVO`weCLT03j!^3<;T!pt>R#`EFozo_x+Jv*`%k|l$KGUlUt)$jEb>h=F5q)m&Xoeq zNd(TapV`I+i5=~>v~XT2h+M*7+|}4A^uNBWqMZ$Y{OL@Vna>U$8 zcCO_QvHpwmL<73RMHQAMi02TGY%CbzkNV|NV&G_oMd6rX(Bf`F@nr@+)AAaCe0O4MWPIR;!6_i+%x zv-*xQGr(!%YS0JwZ~rAtnS+TopKn{TN%z)zH{T>>t@q!hyI&w{;xZoc{4MT(s(Bj_ zB-aqky|2>|sAxs!4-Lbs_`6wV+Gt?2sj@;D{DBzWo3<)<%2! z`RyiyIt=bBlD#8GM%-D*jtwlup$MF6?S7NMapnF?#E9UAPHylH}ftRmE#ylmEHTu2s_fnSes*Z$3)MC&FN#uPr#A$50rAUW6P+Q z%SSU1!F_7u)5~ZRdSm^OzB$iXK0{-Xoo34;d!L;JbOEt88no*1v^IxR#Q__U0&9Cc zNwq4p5`qE(v(R;3u;pDc=pe3pMW1vOg`>M@3bYuMv!TMv*j~0t2emD*99O4OG zJ3hhPbEShQ?R?SvGbE3_BFFr=K%Qd$3FCbtTC(bB@tTA0tx=QL5Zjpkmi6b z?R64gAZ4c{FfJ?YTi5ar}P&|cOsNr zu=SIFYw?3`j!jwDAT-EnxR41c=Unq5w9ILp%~ngOuet)esW)@V);0O2g{j%#j^SjL z1jmx|fAd(PyI4GyqYWib{#C1LTonVzJqKQ2Z}8dH#f7reGd>RFlHD1Yp8s>@_}+E_ ztqgssq~qXQx?t*pW&}kEbR1i-wW*lV=RBY@WDEpAsK6r(ENOn`_t+8aT?*sDID~VwB4o2H%sE_ zu|V;~Od}DJzQX*eg79kE%aN#6)ZATiwt^;=>!6FLrW-nbEwLB0{YUJJy-rt)GUusz zNTm39pT)+;&D_c5Huj?4nxNp1TuAK)TBXYKENE#I5DWONLa#K%`*0j<$pVK>93ESq z)@GD#iTz$PC)J3MUYn`>w!`8db-XRVL!47e3*`6^OgaVT>pn#-|J)srDW$0K03&R= zWaZh0DFR}@tbPK;>Lb-r*aB6`&~&@;In-Zv1``LHP) z{R(!eHbJa5eWCwP)sL{|xn}hK+IIuBM})F-k}(3%0w@{Q7W~HIPGdR|5xhG(31Ql7 zrmY6ee=+D%Z5Zm^S_|KP=Dhr7ywW!xeqxmi`H8&L5Q6-_+tP1@4;GD=Lgg#PKoQz|A~|8EbQC2^JgUZKmBjX zrm`_>uHh6j_!{%?i3GTiK&k-X(5yz}+x!+0-h*F}>+`HOlyQ8t0kTn8CM$HtVLpB`UP#22!wc?tkIs=cC3W05#eyl``TFH8{U zCh>fvUYrUj;ZX7v-oA0HfRPmPC+S%hcLgvS;1DmVOBJp_&f!iV34k|UXZ-=v9W7sU zmtPxDm>e5tATu`!G&Y?DOGH|CdB&atTl!O+afaDc^g;8Wh~h9hb7zL)}puXZ~XT z#SiVMBB5d6U>pUoJ6D{E1%4h@pS~3r&{v?D52p>S`PD4hG|zHX0vu zJ2{_#DK$NNKBci%kGb6mZ9>VlSS}R5#S#aG{F_5ASf~f{g<;kbY(@0Vtmf_tq}#e&q5AnW zN@F*Ajzv}*n*OEsdN9CgbQlXypZp$yUN9pP4PM0#iGGoYpPH5=0ptw9KR8x^!Z;ApBI8>C2`~l7zzR5Uc*b92 zKe-cE$>K$kgD{WSv;ZPFu!rU-{}*|{lW|WM&96Nm8x76}0QgSaai-}zgJ1bVu_dD( zb-_Wc(FS~A9oA|fc4x*k8{5N^?f`USf3#M?hWpnGpeszuqZ|m{C`}y0x)=qHB!hNv zr}jUK2h5!O5vdTvt1D_w?y)8r1sAR*4r+8K7OJM-MspN&+wn0Qp{gV`2iYP?T9NcX&Bh`+-tCra zm5712pi@(yIX>79JOtDk3JXtbW(+5O9~f@wIPmOpbUj$AJ8-24mzZ~L#7edv6j-uZ zzIj!zfK^+2CjR)O{{d+57hnDvhlhc~=itv7)n{NbTpyRRTp^WP$2!!+2PEFlM=xF1 zZqu}}W$p~dbqdT0s&K`MR4F9GMt5>^cU+#hSGC3^{;Q`%@JuR#4MT@$CF)m`L>eg7&`=+@syO%aJHiq;u?}g%&g_CMI3k4svw{!xFnxPGSffNYfSwjmdxW{NR^ zZC(HG!NI2O;bE*`YT3&FySz)C732P~z&{}oPSeFkX%+$4Nq3m_rvcC)M0>uHz-eC` z5BBSP*kqyB!~w^f)-5}55P+TWQ7}6evnlb{W@IKC7N+d^Pi?+mq8xQ8tL};*gv0=1e##5Or?Ns zy=BB!Wv66N^UF9~k|5IHaoE9vQH8RUr{FR*p=~5bvYMK%|Go9ky7*Xf)> zv$J>cZCIdVPZYWkAr^-`t*yDn9*-8-y@6lz;1W#txw=I+aUMCwp%Y`ZemN!e?MN@& z?zmm+iYZ#ZX6D+xKmRaQ{9-tZ3QIZ3=te>6X`~73)!+tPd0g(bJYJ zt3TKeR>w~jXCk1jwO(_TpDV6pimct+jPN^kBOI!sYDBQwGQPQif$q2EkiH~(18-p4u9r-ZUp=&ZWln4){ z!9tbjCIpx08pT2U#(;Is*pNtT)aL+N)LfL^!5IsmwU@&9e5-tqOC@B7bs_ipPoqBT z+d(>mez`YiXM~(awkt5#`*qE~ukCpGEjFNhs|F92`yi+}_CN=dcUh?(Zts~0UFY=| zCR^VrZ$x0Zu(EDl(SR*;#>)fZ@ct>QR0pmbAG92`FoqX6AdTs)Z{~`5%VAJCO&meV> zKv;5Q=!VV|=w6^*6Vz-f=3%uV-X26hOoc+JhEEZqkIEyxJI(^a8N;0+7yzGzwbxG# zuA#FaWSCHk$rTfAkE^aCp4p)@?|0>(p)|D1j>kC?kJ*eqIG}54!y)0Y3lqc4-eola%rTndBPDi)4ADQs5oB;W3xJoU;fL$5Uy0f@0N|mJTVu*@>B1v zBp+ry{*~1I_1`>1@7Y8P0ke_EPAXVEme|Bsp>pQ+!_h;G!^i5xes?9*VEUS%iYvWM z!*agBg?#6Y4VLGKCts`pWg0dS_$8p6(Bu2v`ly(5Q%os2ADUmsf&qb=X?~BBRqCjA zE+6#1rAOl>OSE_PlAGh&`#^&8p#**5OYDWjL15#Kxz6tyq&Bj={XS9=(lLLB?% zCTFLy8=Un6ra@&ztM{@;_v@^9)Om_6Y|`DBbC@+H*pTJY)fvH(SPk}Zy2IUg)D(80Ie>#K`RK?24qaBZ1I)5vVnnY&1yi0U<=^4ZBBVnI%Ct|4VdA}cmw+8+G|G2nzWRn2A#mKa`bmAG_dP8U>hi((1GUu+um4*h{#iC2C zITbMF8_TB>svQ6uZk@+<|b9JY}BZFdf{WY0guMBOrfe=l0J!%*@T0J_rm0xcb zvkg4G=W)cYezXgzt)s!#66FElKqi6KBlY&C0m;E)CY8;9xO{4-!g)ERjP6xm*1knu zpHs$NVRvI!NfE=vx8g)+TqCDUkh3ptz_cda70dqEjsrd+CA40^ALx%T9w%*Ws*6?K zLqs4W>gOV3V^aL>Lr#iUm01~yYW8Irp#h%o*&A>Duw?EV&#-LMrN+;EetEg<{lCsc zkH6#~*QjZj=X{{dgC)w@6iAI^-9V*O=L08M8>23wT4CKliR5XwFcG6_AJS$~VZWga zTVa1vz#p8Xv#Z9s2H+DW6lO5^DXDtq44LWFIT=(NYrfVf?jfNDHN|k7GU1HF0*4XL zJcMlPgdvapvvHj_T(^$KVSiN2#rwgf#jBI5{I<0I)GWoXhcA6t6OPt1%IJ)p7xW?M zXV0B@ARW@P*y^E>Y|uJa_~oX`cWVy^33AhzO61dg-^S~2;wE)4l36E8g!Y>&$dNiT zxw!#N2|0*29eY}_rt>|tSPPHOlZgBS%)uXx$<8v9D+k2uNh9*fvU5=JNA3kyn44Bc zgrHg3CSLct4XZj;ijdJXi&FJ)a7b@s2)_QNgBWR_qY2X6ZZU<7OD`0sx4IfxQ6Np} zYU9K}QB#|(09}J~4Mbn#NyOi+Z+8h^Tj&`_dx#abXXC@CVq69Gk}pkQr;51B`}DX!ryb_2%;jt??m2d~9yIV<1l;_k`0N@>rnw{g=XB z8M>iuU$5xN&bYx>@Fr3X(ZRds8EX;N@#k`7h!{*u8G?l168y6!>D|s>XTCN}Utrto z&yvXy-0zY}{-Bmp{)yJ73wi$1>7M{L&j}lF+{DSZAX?2IM_%334p8fPU<5BX>IQny zUCItE!|~fC<*i95cdQvTbv(zU6m+KsEMl**FUWQJwsi=@%Q84LjMui94^o40&hR1z zQ;vAwx+$Id1Qx9pM#kh00f$tjck@h97rZ1jw#Lv+s`|TuJUc;VAYGnz-llf0!T340 zY?6HWn3B(8LgtFPhJB3Yeny)AwDRZy!!Ou#T>hMpYt9|bfb~8Jp=g5l@_BQ@lEZne zLDh}6Z2PB`X*<#Qu6M@sSUJ>ec0@;9;^W@_iNb}r#Ja(Nrr&!?&{=T`=o>~_wVi+sbT8t zrpF`a_e%-E?qj=$1!uyjgnCjbEM)taEyBlPy(lRFMVtD8maED(hzeb9PMV*Ht10Il z+CMc)8J=cQAvkrlw#HVLLf6$p(UeW6Vqy|Z7V$Ho8-|jV`!B*5rH=EhfhW_q_QC%Q zcz&r)>4$Iv0o>pB%=iTXKHEW`nMZ+q=R|CGM+RfZLlM&pV-0}-eL${_Fb(jx1UeX{ z(hnL32qP{o^4oW)cW9ejAzeWdxI(135I7s}s2SSi>CoNsQ0-PU61au~FOF^1Zb_PG z1j&Xj7(lgfj(Fo9HR|AF1Ek`6oB)Eq$&2Ayx9JMf0A3z#KkE%))51kB%K=f3%J*l__+7k+!mS#~`~*EF6M zEeX2n*&{^fI-R}$^V(&1(+;zYYixXRMy_C?+ zr14Gs=hfiVz^2k?Y55fJ!4#OYMnvCwiGo^#mfps{JWCuW10HE}+K5=ghT7bu?Zj~_ zJB~>j+0&>?Y94Uk&No`^qBJeFj=7t`>Rc?Hec^6%Y5#sB2+AN`Y^2BBtI@>7GzitT zrdpgO>_*O|(ZVb3O)5foVlKk2(!Q2UTCY>xn{vF}c^uB4PSCj}SwD7yUZJ4t6s-%L!1J(3TBWiW+F!W$ybL{&hLdZr=WS(POHK1u4v(ax6 zC$clte+hlecD+Z;Gt_+$y%oZej6|BJFFWDaLbr&Mx)NIhXXev7>BMs3cGsoGV2G8w zjg|3$*;YzE|H!?^TV+WX#*(a}QlF*s22D3)?{3sb0o|BYTy8V>s{S`?&ONfzWbYg+ zw0aHyx2ee#qXJf>p`L(uR7C}Tx;G`y7QDo|D8X}RJY;t{lGW+QjiXJ2ehW33XaI&V zy14L2f7}JLdaF8|C3u~+vu|n$BGxuf>lso`3M~T)azHLqnYNlCGsh*80!BqW+oZWQ zsDGQ!p0c}dlTmewiZDI5-3K$8x~V>{f=L3L@9K@&1d7d^6|r$K)lRj+ zys-}W^GRld`Emr%ydc}f_|6NLz@74Z8WIL$01di&toFsywCbU>X&&6sc{AHj;c8;g zV7A{b%4Vu!v_6lfr;Ax0`R4{hq+F_$`~ZfnKNl~+f55gM8Hdo(5Ur|&>~;XQNCW&v zX_C@5#*>@3oX9G*QxW|4vj$w3?muX*Op-5{nALPv9ug=M1^eUdmmvIz0pFuGgXc>q zN>=hk*A};qF6hT;(;zK<)I)@Vhqj^e`jR!i1Ws}4$Qb7jrcp{!8-6J(hK=K8UKbWj9r*kn zIm7o|DW6JZRg>C}qAN99Ace*R|902LnpMpd5h;q8kH_AedY!+<@CQ{wyKVK5MUiH? zu*T8x25ro08_G&3HHb2vGC}`bxSKrgW83!k!YA#`&YGH1H-7TXD{%AprwOK3v|+C@ zh@M)qfuJ57b3bKf&aSV_8Qg^V3w^|UWYsW9JErCTxTX%-vbHw|i-D#8YbvJ*Y12&w zvFqFb39x6J02qhh4@$DWH)wixnAEX~L-Iq3bBeuz1tN!qre6cUIv=%OyivjdSvt8t zJ}Ei%1J0zu<1nQhE*wQ{WPmt{O_tjJ5IouLNKU&j9S8Q; z016H~6Wo)Ku0pc*5F>jzkQtPOOh!(&Nq`h@!BN|7A7DIydspX}QC2PIBS6%0YS2H{ ze`Q@~j2;T;SJRE5Y&=VoFFwu)^)_JAT0iYbU2QRqZLC%b`evYTD_)(;-dwVpO<&SN z4fG*K^5D{WNa-uN*7>ruPWM-a+c)5G9W9x*bHDhGkPE-FRx}4LILT(G`%8JYc~;qZ z59eC}KaqAB`M&L#%jBKzk++c${y?7NZ;O7qT`EiCKNX=Pg;Dw<`-m%Vi6f@1&H3b+>u>27v$4-N47H#ZDDAc6 z&UyQ4?ee!6@o&V@fWS|K6YmA18&Ka-JL;X0jkb?s-;2*Z#PsNT3h1NzBbb z=Ksi$+j>~*8%uKO$A2w-HE?27>T;%p8OMDN#UV$a4t9_@p(^2l91y9V;)qONPo0nE z@=}3J<$%~X6g#hR$ua2Jdb&tSb(5@?ML$TF@6i{ki6M?8LW zr2@Hu=Q{d@fb1iCec1?A#B$^LB)(_ddF7F7gJ3(wZCs_qQwhJC*<3=+EqnE)*Lz=Bb(GRD%DjwwP>m{hqhZ(p6n=QDdF% zt7N?k`Tq=bgL`os=NwF(QhFXZ@v<+KY;<6XUQj?pwmx|Wz3R)E`DmaA=H0Sm@7j0s z;8!u;v9X7c9zHBRC|MutY4Vw65Yd1jPGI$a3KY!iF}E9dX3>fHXDBDQP7OCteGH`FFx30pJIKb@sTtH93Vn=lm77?7$2pQ@~Ffa~0gsbD$;d};olL)#+ zJyG+NUt?3aL@5aPYlq4r`rsNdr-i$K+F|m*OALTAJNXRPKt5{P&8n$z-Yd16QTxn9$puP^ zYa-?+s(SkBDfRa)+N1TBq4axJ#R5cy_!ko&Y1B;g{D-Bioj>egEupRzKdA(>>-T62 zX!K9beRM~6>F$f9TKXq@qgw%&h>|_#)yCbKgU>F$?5J@0wIXBE#1iz8wN?=%=U?rS z4Jm=_Rm#kr*FYE#WX`21tNU4{eQN)wK=-NvSwGW^^o5f#cXW%A?P+7$(+yz}_tG^V zA!K{G)^AI04?Iy?M~=3ZItNe5SHWIodth75^TD?_n$P!J_~aRuj5H)!Z}i{SP)q7Z z=Z@*9bS9GFC)R|ts9B?{nHCYDI(Bi)sDrNX-m#P(~Y&1 zb=jofvJ0;I`2IY|0Y}kE`<<6Ctyg>>s&4ztzOuqM%XE48(4I2B8i=&EgY|WNFVAu` zEl{N*E=n?kJj+A_Xhgcdi= z;R&t5W{YjJ*j($!=`y?UBoM)fVNVwR3He9P`-u~%AS?_{)c8A6P*3|{JV($pf+BVr$=}E7D ztp(NE7NuX+{w0k)_27ZZfBBc0fgo?gjziKxpc&v%fbs_>fIWQ&dzW3slMW zye@Z-^*fv^IRgsDqQr`9htFXyh(qYd@ueG`NIjyqwAG~syAo4W< z6EU?(-&Zn?yh;v&YyFVjm1IE1N^*t=_8Y-;=5iK+W4@Kz^A_Qg_WVlaI|0ewoAZk= zIyeHnO%B4>j?PuLZp!3=N0juMFK7+;HRU(4y6R`?%O*w2e^`5ya|BetOw^WfY;=t2*-79Q{IPQB-^V?hVW~{ae0w_mFIo>6U}1 zDH$alnNesFdvCH|xwGYKFdCoGw|8rD|6N9CYn$PIcdUIi)l^}9#QM8cx_waTe0r0n zhwmwTt<6vTmhJ`8U&d*k>xm0+=#SNm%l=YZ*zFg5RoI6mmU$+J?K-Oa`eqH~ zMHwAm(;Q}MqAyxm0?V6BoIZ-5yBY9&VCM|fhG(Jqn($s)(zgVI_|&zJMunOa$qI>t zxJB()XWBIU9K6eFCq7cxX8Dc91zXMSwsya2vZQ@Beix021Z~B51d%-(-y4l2%T%%# z6*>oB8r%;O(AOG(?M$1gD-68S^y-jgg3mcR4;wH`V+h*;ShB-bku6^SOq_pXK887Q zzx(cnhKi=hFbBFRpifVUAq(0ZP=2{a-9v^*+x!+H&_&v%s>xh2p1F5KB?ABm?S-(lO)Mf9KA{S3r;o-L9xu-OI#V01~#D^U^;i zANcPxg6p1OJ}SloPrD4}^R^Fch$eO!nIuvAWG81Q(rTT_;| z-J1YLAlfoG_#+7UGyuo#aTRnmgqDEkE8s9y>S|a|->SX^g);SKl#NMz`+y(-ipshr z+DKEfTNP=!%UUiQ5orsg7|F^Ow$takPAIai{>Q4`y>I9AJV@W}*dD9N<)kS4f z>y*(ohg@p0LA*eVV}6%R$1i%knorykqBJmx*>cec4R#@`sDj;ZtDJ7l!N`jbqP& zF?asTRT3#Yi~7{3WHEu)nglKDQ$fGYWKaE8g*e)Us(ipc*;wAa%&2phYv5i(De*K+sAwR*zP&w^yi)H;u ziiGZDsc2e!i|S;*YH>-~5mV5Wrgiig83g{B1J@VV0-S&ME^2Xxx<$1=y z+m-2Cftr0XEpg6~OTz_dwOqHY1?6DI3Z&-u@RI&^9dn1ke6VZ+L$6^ zsVIt_Z7gYnek?F;t)(^LyYG#MNntCk>=vQ?(VSO6Sz$?OoHrmh&av_1 zqQlsbLcH`QXlRfwKbQ~DIOaQOQ->}HOC;gX$_V!(xqL>X7y8qLDXrL^!w4eS+5tU( z$Ueb&F0!3tSGfO$D|I+HEPpqhv7w?-z~c$J;c`m=XC{CJ+$c@xW}EzL{P10#bgJLY zVa&Y$BT*@GEi4fjQVYW3F1Lct&gf4vlK?MpHY9E&`!+!EtOd4I<|y`Z9E!kx8F&S- z$g+`;!wE9zX{$lbVJ!!Q=M>geLLv*iQkf#Y0sY4wU|(_B`rMc=?FJl!6BV@C;y4v5 zIca$!w~yBS@F?w?0e=$C+o7f>2qCE)W2jr^~`iGYWwf7 z1c(^0F9%;|H0r0%L+>JGtv!Kn*)`Q2EY&GURgtpnY;~xktdWrjzmIiOLKaRc)D$yV ztu(rFksMPMMr9~tcqlx_KBWZdrZ>C0(UE*MwEaeek64nmY^GzIL8mWKV$F^ZgE#^S}IJ9;2nBE1vm3{5j_yjRrP5x z0c^unbIx?j$I0ZcgQ+^GF4ke}vB2;0a?y8@T_VA})rGo1NBf<6HP|C(3IFK9xf8Y8 zeVGdDoRh;@8BZ(qz2j|&1rZLnp$9hTcjycl0iY)SFtkGeXoPtT`{Yf3q>zoMx--Fq z$u>+3>y01ToHh;d?Tynssx6bU6lqmpCB?zN`4Lf{N-IXw7JiMWN&YgLXLL8duC=^S zk#Mu<@7dj$T7N5@ZqR7+y@h0fQeM$F4Uv;0KSVuf7ayP?b*=j{H}2bkMcq#S64MY8 zOz&P_{zt$I{DnEkrprEtvs~P`Ow3Ox}Fdp%D zXVa3k_&=FLBquD3w^hU-($*Ug5+jYI+@>@9zFu?+Rd{ZO6m)5?qxG}FZj)nH!8%mZ z5hrfkNg)Q7u%#jZNYTA~6y6b|9%*F;!iP&P78U9gW>$9)dj>&tAT8V*dvcNlD=ZqkclM6>yGOwvO z7Bm`@j|K>kpCO+x7Y!RU7R$Y^+qg_Gd!Ny{j{QpF} zlgg1K2_tj0@;VbjySNx(_-TX5W$+sJRqV+o>w4Gb#EZ|0^XeE;b@$c&+B?Be6v6&} z0eg!YoCkmv!Y9(*j)D+G&ke5aj0jf#EsD$Ubp~Je9fnT12&$ex7}xn zvR%Lz`y6{`<8`rknW3#R($U4x^~UYk!2?j%Jo}zrpMb+9e0tE}Gsd|y z#L3-NqtMk~cQ76Q8-N1*DhSQhy49t}=~Kz8yX!k%;fS+R*l;oX`=bBi8-Y`5`~k=c$$m6LSH_`e!9o#ny5;M->AFN-~S-G^%4G(f@;AL`KuteMBf;X z*`>5BtFFjUx3JdvYe9eE^y*sluG2*I*oC7nqHea^9wCVuL|d}SafuhKBx3fo>mIdD zp@BV(JV3$i?w95B>4A^Cr*|bAh5fe=#2;5Ywe~O-OReX!hzDfwr_N=^ zTx>L|7o?s16)o!{jYV>`_VSn|k1eXajUOJm0&BcNCIsx?gC4|y^1GuN?n`f@bYqg( z!W!$$vrY1B-YgnH!6Q@NXR6|<@p)Z(qNf9dkV0-6W8t?3ZT!XqNMCk`AJkc7+6Fne zhE%Ppr)g@7;d{v}KFWzZck$}HUGF7th8u_93cp18m45~-INa{gRz_}>wO@vO^ibPi zQKL$;3&kbtl`yzT=3Xb$dO1(B{VV%;i6mtBeC^P;jKs~2v!M0!ll?GUd+ktc5t}UA z**T*f+!D6p@Dtn%9vNDiXfeD}MpmuBD#MCpZc9B~egr0dD+yD(R>!6tLXH|QUUEAI zUs*z9orxW-N0Kj?Pj(A0+O38^DmC1^!1w~0ZI57&PgWk7Wbaw15B3(hV>swr_5u72 z6wDJ`bvIktrYg=Npebk5lV5Ttp>vfs9fjJr!X$);)ohB%=9m7EOt+{sBX4b0&vZXD&6VmdSnDGG)?Da|93_4FW-@>ty1z z7QWy54%#ILvNyK}uGVZ#tb;#I->iXrbBTmj&TQdBSn`?b9U<-hq#0+SGu7XcYyuru z8^C(ibE_lQcYm*%<9GU6BScElI12w*Q&FQ+_4>A`&Q4B7K)@+HllFO(wL+YC*CZx5 z!_n26y}x)riCkE}v^~G`pK3Fj6ZMGyWVTq2mPKSG&R}t@CIORMlR`+LJsEb;To8kYL4Jva9eH>_)!@;YH-r>ZfIfg-XJUNIAQsQta5S zgb~tJUOy>zMbZ;P8KV9Q z^9dY8U!Q^n_uufI^=eH<4<&0H0XX_@} zx~$(rnm}$YPy=aeb*&oF&|{&lI@7RiJ^%GOoRysp-jz!YdI5j&M}+LQJNV&AD4@rM zY!DojYcsr8rbHvRzk4WBz0?ReknpIjTf9qn0E{!YQoMQF8T^9U-xxt$tqqL13Gky^ z7iZn*YF9Yv%LACQQUo-|s0~x+NGVxsjUB>!mD$zajei|XdXQ7R^45$2uX7hoz_h6* ztelICr=!FQldt8R>Sivqi4o?f<7(mkdsJgystF-=j^|2>ggdrs3Y6iLi7JC;0=D!E84YS(fitg+2HwFd&N42% zyE-@{6l7jr8K#7pEQGb}AxC~?<&{b!f;oi$uu2a}ka| zZS>+#PIh(<9u~PT-|kh_xP63!gU1Li6?O7xU>_EFhT_++FH7|91K-8 zjK%P4&zQe;qhEp1xdraxm zsU!?t+sd5NJ!tyc_lPQMSzG5Fx(&*|PrmZ3U|xHA5Uu4WxMoP&_6L`L zdjm@e0uMEUZQR`m*~qcAL0ICfLn5V8nQJmBnzHzt5XbJj_(}q8Ls^@B?HV)u1zpSS;MTKG{{hR2B2_>t;4R=U03^ifIfx%;A095;gA-zY zI-S8~#_py8U^@#HriNHyw|*$&`cFaao|I(_tv)tIg06-gZM7@`6#-O{V!S?t`rA-J z4&4kvMPr014)haOg_EKKZZT+D7Xkg#oDA%3REYnYW4VkV8-vO+1dTEx2(q!WO8~{G zJa^&&!mX|3fELw&?;-%RwC-bW>{q@*Jvk*nf3d~rWX6ON`J0c}0h8Zlyh@Vui8S?{ z2-#hp-v+*`_>2l_QT`tsl&~?YeW#V~k=c^XSw*v=X)w&~zOBjFRRrd*?Jg_)-i_Ap z0N5NPF=w(9r$uz_9vYRBo?SFnj3qF;A`vA)1X5}tw;-OGXfeL0mob8Un0uQGFW((8 zxyVSV+fKQQ$3>^jH>Nhq5$t9{biV%E09eSDsP$(+oP ziWuAQi|Ug>>lbn$_Ix`KpT8+b?M~-hd}EEt&N9?1#p^7Zn-nx^oVoglk1M11Ymtvh zBiV$W6#IvSeGc1;7g^QYn+6Y7T;mLQtd2kpn_$qoOcOATFzws-h}V3k!_F zxjc3__=(vt)>qrin_!WnWt$g#s}>*r3fE@QdIlxhtZ(72T7AXa@SGb^pxkGEkey6) zF7`Q%om1bhTz!C9V1myC6`8s}(c9l^d#{`$*#f5$(x|QivK8MDTmh}`&AZ=gW=XwP z)?JjP8|%%?*)w>8?^(Hc_;+=29cGSF1MrEdYud#!T2kPe9Kw{Yn9mD5+Y5#{%)k4m z44ym<0_Sx@A$>7r)jraNMw5w&GfT~}B`3sUX$8}4;cPB*(;V!x4cXocE5D}eJqDN0 z_J<6x`6ThJ6#5fOIo_>hh4KunX`*b6>rfWgA&!4=a+EWpE|Hm9jK9MCtH^<8yhk^h zB6lY(_DvNh#E{S*&Q%Yg3*039C~CFyUjNo%s}$6IJR0Vqls$R3dxiWozo}DUDM<1^ zm1j4}u^YJ#m=9KZ=1U>qf2;}#>b%HP(uw)c(-V0>JKhYo_=jHnR5nz=K9~x^@kkWE z+q(XRBy|1aAHZ`!FC>ssxK5I>J0`{JS%O^cf57$x_zEc{75-Dh4)E^xwZJH>BaSEC zx(w&nDNUHfsb(O<47E7{M{wT(sW=OjFi+A25-*9%7N7}ueeJbU$D;hJ^N7^7>lU;g zkq{gL0NJ6PR)$ocO{~TVnFd1La2Xc4Y%ze?@Z9vn>iaWS5t|FnGX1lH2Cv*_`>6~a zK*`ylKor^k{}tg=pSnoh9J5iGEs-ncwJ#?R|9J(l*^r;nR*XCiI2NM|N%y3dldKSe z@q8H=YRg{6t^fc57z(|=_N(zFYt}8hz$tg{icKTvU>qbCm|2VGPNzzLxeDE@^J>95>`7%! z@j_Z0`09y+r^JY-$)9Mt1rZ+hy?Ev}p0JC}g4y37>~1C_$w#y>Pp5Aoo8NiJWf>Pi z{CWiFIE$cOe(enVE%wKIky87v=HPwqbO$4vqmhxnQTdbn$&T0FGiS<|7HLNwgr#Sv z2R!0fif-{6_Z{p>!5n`g*p^`AsbA1vea))()8@cgtniFXLSztoa=Ych{v3aL9KI`B?V~FMlM6=kV*$cJ;nD$TM8Fl7g_ufBYaK`9;@q-6@pYjlPEg6em!& z9aqIJ+)nK4cD>J>0w375oB>=R(2cW4654wu$9p2wXm!69w|55XYDc7^Nx+Ecc{pTlv|1C%DF9v$ePSb0jodU6am1 zx7(CEbJc$5P^#@iUwiK!SX4S>bfR)xO`<{_7ef!S%*2aS{UyW0Ml9ps2$X_%odY{u zLa7l@>oVz+t((K*!GDWvo?2-3>la2rug3+|sPzrD)oll;to20*hb&nIX%Kr&C<(EN zM{k_1Q`)`!x#xiBY80B}@xRSug|(Oqq)QVambDxFy7MqLTk@2v|0g!T0(qTyJ)+!6 z;m8!j;;Fj{N0?|3_u3mm2^Ua~5+;m17v&FV!SA);b75&fAnAC9P%^0~f?tX)q)a;# zhVSQBGz4Lf1Qj+YkYvnGc{YW{Fa!hz1=-c6Zr$cQ3F2FS;{`NC##KquWMw&yZA!)i zf3ST4M5jvn%tBB{xC4HRHM6w=m4KnTXZ8~bL5^YkxEHJmB#I>MDLlo8{Q*L9cyiuh z=YBQl`kVSym(6-Xgb+)5KJt0eIDY%&KRE~<#z}*VTw44Bztr%sdOA+EX5o^|+)DE){88 zZ>W~%@dt)E`8tiifD40O$}7*;ni0CJ0$<0SM0Ks!pHb^T4X1}Z>WENZPnsB!8p>AK z@Ah2|ddxf5byVAjpC9uLT;%Hh0nEdF$VNRUW7KkwP88p4eaIuUqg*9bv- zWy`}|1ajR`J_Vn|4H~k6MoYZ|z4d5r8Nz$5%eFDt@RNijS=_ zK{e8YI3Hk>B@{t?r$7+0F7881+W*Jan+8JJzwP7K%ou~gh_ag@N|LSY>qwErD0iu3 zDMVSbjGY-Ol{HJ0FsM|@zGR(oD_fQkWjA(XA7h(g@Vnj5_xn8m7yq}{i|f_r_?*Xi z9OvgaeS;|wnZP&y(=ytf=9%_Yd~65r%5%02=xh|4^Ud6wU1vBo46E&DB4!=&-I1tf zWOy28rdDD(7UH)a-A?<>3*u-HCm~quHE<{4fW9W7VYSGoxaS9NN^7Bnue>@gvB&6; zwK!?)Dn|#0VmoZQ%7&X~Gr9W~A4V%Y6S3`W?s#G-@&&d}thY7c1G*_rW9%DPYTAV2 zwH@1J7vr)d4I6{%L)n&-45JI#ZqQiG8V!%)h%g)Sc@kd)?x2|$e*z?AGBx|8eBJD| zb%tpq)I>lpU7kagyHYt&SHn>?~k3=0E-Uw z2rFi9@ekyM__8f8J7nlG6{S`SgnWl>!(3?y#y`)=n}MItNSLD zz}J_WLod@T$FF+xDkZ_bvaO}~4qTMeJW2hj&bCK;jPWScr)(anm6{(MS?v?Fyl)Um zn19#*k!xT+1M0eApi|mW-+HR`UUd4_LyHggi&cz1q0Z!vgpCK~OWeMgy`7gKPQAv< zl>(N1(G>_CW!QHy=5HhYU?bG(P+Po;?+w!oxaFD@mYsfX$TT9?3Zw~=u6c!tFib$K zY5#Kn>F;4p9y_>RAYi+7UVUAQRh!@aW@@H?$-)XD@({ky9)ZniFYsuMCkD!DpmHOe zx_4q%1AU9Cmrj?5k-z#xzKM6=INY$zB&F{#;NhU|q|hOw*Lx$qY8%v%C~oj)_t^Mf zh*|;id&J(e*!IDfpA~3JRsG}ZCg8pkAopeW=yN(wMt(TNj&wmy1$-yLV4qPc6gu9h zZD=sFT`qgDfY+BB&2bU>tQJM8Sacn3biFffwP8(9*$Lkm+0#n+PoZ-ifA@e;eLhs} z>E^D?p*NUh;}R(L9|kLb4I(%%xCqD-h7Z%+q1?b%oR;7^=|s#@)}0&o4JIt@0Pa(4 z1E6PM5!`1q3=Od~LDQ<|vTO`hw{2(;RRD>$W}evx+3&+m^d00#>=A-1Bp5sx2P8si z&=JrFzzD}h%|R&BhX=iafY50WNBXIW1l{(eA6}R+h*;g^QARhKN!eef6lw#Wj2{roi?$0b+0sZrdgE!4 z>3`U2A8Cw8yji25EkEov!@SYYG!KtEI2+GUAS6PT8N@IXIP$)A#=TZ6m@ zI=fyQ`d+&DHqFps(g7YH^jiw|X7z^{#r+Q81V0UqWqLVsMtF)(l`-8M<4Ex;mX?>5 zEw%cgs$_O`F?g&wtGuK+puzsQ!G4@aGi*wV84C%di#Ax->$aVv++%n0zt<6-dsvaTxj6M?gL*DgV>lf`3DiEAN&}aJ;NAJEo%=_zC3PDYl+BuX=GCtWGLt3=Cw@=*d7fiQt$bLReNcV7WMFz!V9P` zRk6Sc$SeFE_cGc!jdOm64?aDJFKc?GP^P#1mB?{#klFHz%$ImK8~bQp2<$f&=-_#3 zz*xN~YTE{9NxMinZv!;T%gv|+$P@*=t-4Gt*m)N|eZV#NmI98P4aNTbOUiDZCt>Kv z>g;=0(`o)I$Hben-^GJOa{zEL?5FeY;tMOVG!^pTr-a zen$_2|DpTwF#!nZc#k2GVHmvv{kfs1qgbJaWQ!UsBfP`r%I=I0uhgw~{q~$4c7Gc8 z(qwnmjC{)ZJwbHizgYl3-AS8a2iMs#?Pc;XojDBjZ)%^S!SWaL1a`I_bg4X=daxnBkI7n8rp90}FZmT+^L(54S zvS0C+1UqrUF&?TCu7C{rMhgU?TbxA@{3b9Bel;3#us0_(h`WU22Ds86zbeIxFu+G! zVRe>B8rL_?0TLJhDYM(RgX3^FaiD`ou2&J_B)BpfDGBhgw}Ofr%OHWkH0Z{-;c$|wOp z>y*WQ>`wn71b<5l;x*(90`{n&iTWE|5JfOJRCWif3fxzXwPmQN2X>xoh>gJF#X{J) zN6O%)l6eyM{QUPe`-|4h9gMxde~a>BiJ1I$?oi$m#w>SCHvLx0ZjzVL8@Z>ed@>}E z-3y!v7HP(ZtW>+SJ!#0ii&fl2(Gy;y{p zDT_g&%u7S9Y9>=qkwLSAdRg>+*|*kg>RBmOln|}YO=^xGQogKI-+#NTW0T|N(G(tA zu!NT=(OCzDic&hA2iQkT#VXAN@%l3ZQOqZVw#D@8mtqrQfH}X&lZz7-UCHE z)Rm~58xhm4i6*n>LOHiBv?18>~gbBj==*Fz^w4z49=IMald<_^~!Ph#}d z|IVt@w%c9Uj=ZWDYMiz6M7w-FT(QM-@XTnQ0!GgMywF;HCIh9jNOSGFoVIXuWBOd{ zyE8+spL=tV&Z8|+TMjzdsbgoZcW1#oWR-S(jf)&c1TeZK;L*^Et(@#x=F*#buF4Q} zleVG5b^Th$LE*i%2ZI^d$<>a7LG$qOkMT6L;Mt*xf&;Goq&Q@KWyCK52PvPK`Li%`<4Yt7K7SKkHte=_RDNx^KhKZHL z$;2U}-Cy|Cm7J-!o~=qnQFfQtIPov z(IGqeTNon4j$RL;(^}%xKnu;9`&YcdsAf>@K-|k_>8YGx=v6j@SpPu2XWxnF%Z`pk z&t7F42)P1ev9zV5G8*A*+)#Ji65}YJDw1HDuo5!*ZtaMDpn8Ld+tNM$?>&NmzaeTW zTSBc_=D`1h{`0I~dp+v4M~U;JjJ&hJ^-nv;)=c7InIi#rzz1duLJZtTx%~JJ?gj^HrGC+De1 zqre;eje_F$)(hPP{2#)DpbZ4t%-{oR1}g(^QsSUOPYkhNck$~d6*&RQ)|+@v;RmPYOv8~kl>t3zygETz0ET!|Q7 zh#F99?dD)za@z0~99rnH*ZYVw&<*U&AAP!CbuIhsYTZ=4&JCo_xr!5eqigY)8!K*JX(qdw|urd}(ls#JE9kfSS7{&F9my}@_D6!b}aS)U#~{CnZ95kiEp z)+(LwMSHNOa@u!Ly@o%cYkxp#O!iO}YAM3+Y(M$ap>aj@ALv>XvY(U;^clXd`CqOq zPzVxmiWyCU7P|=!;_?KGPJS0;nSWI10Hyv@P-FX7ZVE$q6lU>AL9*Z<0O>;P3AkCD z2bBcL^vW!-QFaiJ#9bmv^!S+Tp#=y>(2W8NMsWcfb>iT* z?jG+qfa_OuFjbfg69<>)mc**!%-&Z*h0BTjIN^KI0?y&>}FW_(!;AQ=el*v0!h)C0m#dS z7}^Va!Ew!zP*r`}kNhd9Fh-vqxOu!`%~%%gSz{?g$hzY+eE@yMc8N_fw!4{U>d6|P zrUEd5B$}NJ?fIkvQrH~CD<%XFNuAO(yfKwt`EvDJ(j)wr>6WtTfT@dhrVG2kVWmWwR`)=siv~LjX6;Y1NRylvl@0sR~5q&Gcrr0f*lf; zN*m-pK8q+_kPozk?Tu%4?iO7MDyRb<*x?d56K7|~N3RlEE6*83SFHhc= zo<|I|c-n`(>H2*)JYR|9WX}=XQ}WdJT|3u(#kc+(SWQ}Iyjdl*A=!vz^`dhcz8F$^ zg5w z`dr6IjNW@_wXw;$nb)1QbbC|zF;>+&fA^eA*O#)oKcdU=e_pwN)~4R9or+%CXtk5@ zo3A+wJ+T?EzQ$OqabE5WnCz4yK7M-CgOcI9tAA)1Y9e{4Q9{X#$N!$tw%;T+HjL*m zx66MNCYnty2zV?l*_hif{pRA=i-lzuOY)1BEr4e?kJ@~LaO^L2?=P|L)zN`d0tZ70iF2=<_bH%pYwY2C_@P;Fap<-KShAZk5dfd;`kgT4oDKX zYT9)m<7bYFw`)PBsib5&0Gk1^2ZFb1=8jo}+l-{8Tt|docDg283<)Ed>af-G_6g9k z1Yx4VYWxnugqb#MIsOrvUs4U={#JsxoQnwvd?zg_T2HX0UpS{XD95oR@D2J8vHfLK zmxsrA@C^LI0`kj1Yf^V3aykzc4EQm5kA8l;(_`vMnc?xvk%!&p7ft?mKq{M^@wDE4 zLYB$ZUMQphzL2=H72<-$`qI7DS>MRD=+BJ$<0E8ip9yc~J;-`KbY958>}pw3={ zvi&xX-f%R@ML#XsHunndo$O3k@Rqc6ZXoc~9RnilG~C3kpMHJWvZAG})-`EsGsyf) zxu?!dtI+_jcu?UI@3XsHRb+cD^7O2(`@kaiAXmsadpa?kakWaqr#)!zgT8E#>!opW z69z2f4XWFkS0t7*y28D!d_-F>vjev?#$lSEx2E}hj&?=4tF?N6XyjZsq+ z0?ARCpG|af&7-^87t9u}zq0FmRDl}4M{F-K-DKJk?;5n1c!VnwpUMX3$nR|EA;V+~ z)Fl(N%ag-96$b=Ig<}}T*uTT&6Oi@an)TPR;$P*?Iv(0I)1s}bwyqa@t(E?AOpZ|X zt`kkijjt_)Jp5mk$`S%_#aP1wlkcOA+s;$M8s*@(4CAFs5*h zp$Cx&kAl$?L6_$ZGp0Ze7ui+IA)~HhbbvT5oll{!4r{&%sXtw0`|D~b;dw*I|G+na zoo=)fF|HKRBzuWDu6KPO2hRL`mr$WOM`>Pg~X5kGl?2mdM`PPQ`l(0G( z!1Tc%Bh)@JmY{Mpa~2LrNEe;O2CWZcudfB!HJ;IQ(2zh4rJr_KtL_f->|vkYJvfaQ zETZ`#TDt?*b0V=_c2HOGBO{=PVlfP#y5S?U8r9HcHVJ4Jna`X5=uvMz&5KPDU)Cl_ z*css0eo?$4^#`vcUUpME#adQESO*G?Edm?b?Yea2nxU(-6XbjH^O#S9-QYW3BYYdj z&ke2xm4F(XPfTS5S5#|@FZe6e{)O5lXf&z^4L+p|nM$4yNe^*_fIJZ3p!#>!NA+~_ zJi79@^)g_tzKAtfgGR?GZP8h`AvD1(2xR2%YE=vEA?0gs`MV_1mxg%{G z{F}Ew-5T-e~2?Z z5q*tgGa5xl9j@i?f9Xy#9JdbNH_0I5ar`wK9d@ngmLXFbM+A6d=XvE{FfnRsPGmJh zU3S`}V>{^}54A^SYVn6Njr`ZIE=b0VHisCzrW7c|cYV6E`Y~+p)!X@z#!7~=7g)tb zkZq={b!mcQ-F)-a0%9kJo}{H2R-({oK`*x9^w>EL8{I7cf*oXMb9vWow(fTom=Az; z(on0sEP~Z;u@YjVB&>)FYa3of-Q>&LZ!|suGO2a1kP0FjmZ_K^Lw~__6koy(nmJjn zZhNK~3;BOIR^&~THolMbECgu3nElt%F)P5cMxOjR8@72~O`T~1X(_;Sav+Nu@+<%A zyGE)yM2<|R-z35Nafp+Rf-Ngm#T5?Q?}}x8K!zldaUsi}k^Ph6xQ8lVzD%2XsNMqjyI2?MWU}o5@yCUtH8BD3*&Lqlk0k0zKyaRrr0GA{{1Rk-TGCVmGX=S# zJF!w>vH|~c9VxuX=@Nwhg;O4969Gao-khB{$x>9!rV5~ zA*kB}bjcqS1{Go4*70FzPsQ8n0j~Sv3XQr`e0Hpp2s^f^=|&=}`9i}Ge7t!+E@#|^GQW!^mD(ON7|RdiE11Hhr>zgoeqYB8uafv%o@241GZ$-N6Q-&lu;rU0oFGkX-#ZG`|w|ZJ=~C0Q|%~ zy0h$v^BSp={NBF5(}vs=!Pc-T<*^Dnxo7MktK$)nMbgaZYF_$s9%kEkp??#x`8wZK zL)*pPuS0NC1f;rIYPzlvtg`wJv0RY-D>*2d$uoLX___tlbSUP`}R*pd+|th zK8Cfi{VCu=^g?fU`Aa5Be*v+Oj|h8fo(KywyyzdinXfL5&?`P-Xmxh6DLMnwE7bM_ z(Np$LKP>eQ`0393XmK_+rgSC7nm^;OKZ93^gzlBz+7Rl)bCeVPd#}Rf&s-e7y}pum$tIG({u>$J^Z4oxuxdb+EUB2U?$Fw6|_6 zYeR5=F8;5qEm$P*N+t;7Ra7dVypdUmlLa|BH3<=Z=%KkuNoVFE5@&m!?~+t2@dhM8 ziB&`1x(D{9IrFC;`1;MiOWDsomiX&`>@w>Q>qdM)b_Hj05|_B!PXB{U6_Im;b+dEwsTEOItz$k@_C!ZrC;^=z`!j|^S#Xgg=@ zq)FJU$h6S(_n69O0cYv*c2>^|D^m`(VQ|o-*9=nvJO}tZKr26YTpnuW6wx ze^4{epVEi&-s#oMiAuChUFje7jq5aydFwPw8Io#MD#p}`Ca|-m2J{)cZHu=?D3CWt z11R+@iJR{(y<_btvpSNqfBNhc@s#ZlLo>!mEp3V&kEhZpzzeg?((VZ-SC8D5XzYGyjF@ znj6}nAw30Cn_g;l|FE`f98%-={cC;y)t#+BvY){!`~}h!{Cu(l7YVi(Y%_ma&)=R~ zXMF>75r1?a_CpGE1V|BF!X1xsF7ktDb6ra9TjV>s?zhuV+~0&=0EsQxSruKZxL6OZ z2!`g}fpneH?Orhpvr4IE9BJcUl^ zqAhTSz#}ljQ*gbYd8vA+WFoYC1w;if;Zswq`zx!j8UQ5v{#lS;DoAqbh$h?P8)*{9 zr@MJza(Db!u2cANb!0T*~k61p?68^L3ErH+rHoAF5vDS(v zMLVTD>z_?};J0TYn8oIS>kXAD&dO$cHifyxT>fGb6EO>C+8<km2H>&s}rgUdFzql);vt3K1(ZB_f;69osuKD0N`3wV|7Rt9ryC!n-<(-&`Hh zL@8WHi0_@7`sq_GugdeP+a?4blmAd(Yn;k6k4aM$N&?i`t99=oBk~Tjfw|)X-)gUdAGjFJ6Xd7&Z zNN_zS{+;L+Npi|_tUaH4J-fP`l)@B2VX?!HPI6e&ZrYT|1t9~+@jnD47IqIq*}8}C zhb1WvV#EI4cJHsQB!{QAy>e?BqSPhWzCryeISbD4vk;ug{YoLn$e_YpiQq^(jW4m< z43VJemIO86z&OFlx3&8u!*Q9U@h0k5L1jyV^#?9duA~AWohLPd%~ypE6s#Z6$PVeZ5ma z@f~aWa8g+D>!FhTU{j~>KLb(YTz_y+zWw<8!vpTnk4qIC7Mv8^M5nbNL)`zVW=Xt( zo9)NZosH~u>8g+;I9OIF`}teq5LM7E;73yE0<`AlB?IPka)F`)$h%pD770Tq|Il@i z!U?oGUghzF1cCz3Ea4Tgziz#=AwlYCyd0-qa)t#dbj>7FBI{R2;g?DhPf0R5y~2GJ z7{7VaAT1m5gc zdzuYP!7{U(KgH_CerU{U%qU2G#)>SE>v?pTkw!PEt4oJe)X!)Jjo3fs`hVK!s z5MSNx9;r0z#e(PrWrIJ8ysAtAAL=%;!Kx@ow0imYG1}R=()H`Za#K>8mF5Y7o{Gly zLpl06VIxY+!Yj)QHCwVFGvvgCLWZI4Z~MCTULy{tYZeVV$v*s1!V*cQzERbg2Y>zJM-L))M16ut3Y!(PE~o1#W+#rAH@ z;g9*~?``zDUHQw{;fLln9uYcTH_~=pN234rQqs;~BxPe!WAs6^QrNuN-0)3GA)IoTuPrBNLt_{`w(X!GDETu6Rpe-xKf znVfnHZ++6sumZCKG=joN&ewm+dn@zAcHHWYKo+a3Ypv)^&>i2C1$9fRqh(54-D9Nz zTQ3mn+%tVbh@}*}4(6occ*|RJn@jZZABMdCxjAq5<3H3;Q^!UE$X`PkuhO`~lSMqX zE56#M4W;S&xi?Jf983l~uU%0`05l2k3%?!~kJofc_CD)RcWNUdyBYCqPIRmyXEZ1B z&h(#s7Wpt*{DeTb8-y}1vd2Ft(?1%}H(S^%GJ0yto1$j;Kz$)K!|y>>!2OIOqySAo zvcG$y&Hjt{(qrXiM|{X|EjAkYNS&XAhjU+kC$&H#dnu#qx6%FJ)DC>7{QTRvMVt0P zDLKZ1p<12hbk4R3GRTRE4mVf6E}6DR7TEp0Dk?@Rl(@Whswjj%?AuC}kq+Z)7%fj} z&{Xc@yvAPO`vTY$+sA*++%ybiXZ6RsxeVDm+?jIc4vC>C&lNQ77uEk`{C_4Re_nzJ zQUvEB%>`>tc9G;`9s;LURmw~(h(2lmgS4C>9yvmXF5=rfi^i9Hp&tHcO= zI?hP)*95ybWV=;5A^0~!9P|;;q{68M3OK3431N)ceqa^20Af=V_ zDEm7=9q+%eG#?ynWP$`)yO2I$b0UJRvd=E;FIW?60!x~nWZ@p@J;7v?B-L|6ad3$o zLv|m`q3vB5%0rWgkA#Nu5P>TSYeFoDeeXn7)6KiwN4`I-0~+WLKz~?VZ%F&W;zY7K38Sl0vZ*5VcG!6U-W%`%~H3?}47SH!&L zZ-~{#KV57#tNiEgKstB)(RcqLb}V+${lc&*_{OXh@{h^~w{zu#h>UF@4aYG?u9CLo zBS&Pv4&iiC<(*3Pd^?y-UjjuG{%jdL+{o6UW{ULX@t^AP3O zU4dA`#@)F&VvWe^IP%uk>IN_n&>)FITT^p#Ed1r?p23!L^qqTrze8%;x%lT@oF4In z1(m`|OSS4IUdQ)rXL?p-Ze0#9pYB2Kq+rn)o4f=;A|Z3*;-0PYp#S#N>#+^%gB%lf zmP`SqtuCcL?38>pa&VCR zHw_i5dGy<`uSqrF79ff|*y-gzW}@tb}VB_7yyu&dxuP_$U(dQ*0dvhS57 z(U0lV_}9YK=b7$REH~bNbVvrhebkgejmg96O@sd#mt+aLl4P`s8_y4%@Bh0$xrQW* zmv(q!s)9T6F|Zh)oMlmYstR5CF~GpVFsO*gb`tnFvfZvSgOdauS*Uyl`%G{){Az)Y zJUD7NTw%8wg-j$_z-o`-Lc<}bKA0r%;U*0>d$SdCwUuM!dNw3oDgk%w^^8?^2ubrS zA;APj<{h|Og0a|yVEat0876Q9m0gq&{c14~m<)nSNP_vU=pVo#x@?o62cQPPZW%NH zZLP-KK(pJWLiivm!3r<6gOf5W){H3&Igmzu7GPbMYf7`!11{n6#-%0IkSW$e)p${y zGlA<3@22U>6pV4vxE3RW<{p1hSoilA$uk7Dh>gekVVC`uM63V3%;;x7u7oJ;zfB0}R`Um?u5X4~ZmKG- zrLyWCnc6j~6g`0qL0+#peo2P4dUZ^-49zcx9Q2>`=D5?6$3i>#YmaSw&`>aVz@5@E z;1>P74(HwQYE_k2xap3x@4<6d?x{u4q*Ed_DmF5OUu#)hp!lEDA?gT~u5U;Q<$n#L zfuj~=lWjFZgQpTpw!d9lHtJ7Osl0F_y7InJBxuRV4rDlGy<=I(N?rXDe~oEt6ti5M z*6dSW@zVNuFh?`54EC;d;?^HUKU1&f(B*sXX=G(Zc^mp^<7J1jM&p5xx;xSRVP%}_ z?3*{L8Ycapw=}i>M1(=}HmxiN$goKF%?iD#%@?~|_>4bR70ygpb?$pnmJtks# z75O^FaQ!e*tEi@rI8e~=M|YvGGDAS3NW!>Dchwti_rQ9SoM-z8TwgWf^E2sISfCH$ zpldp;Gh1 zN}*WKzVSQ8bZ7J5zIXbe12|H)%HwI-e@B=nL^*UhkCF;a%YIa5a(C-(k8qJe$Lp?f zpK^^(|50+W?xMD6gAT-C1i}s=aFE024MAj*C#NJk-=&6=)RQA&uu`=;EKjtG?CE(pC1a#CjmI0=$CzL`bU-CZhM zlyk!mlH{}#$=j&7{jy6Uy2FizsA)QzmNHpjG?4AZw?dv>l+iZmXW>*^fn6rvp8E27 zJ@~1{s9$fA7%4Y4ggzb>dIE4q2cx?W^&7N|jvLje_H^Et$1ck&8Z?i8FSr>uQWP&3 z6!~yR@1WI%#=|N-m)+{Z=rmyKs+rY5eYaoh1j~``23dHJYD%6LO@=Vd^{Z z-Zp2*SNp0?@aipnd!!c!pXN2CF}63nm6ANcEaLj3_szk}6CwUnc}6;J{~XJpfoJ0} z4j*-|)oq7#6VJY1(RixkHs{WZn&tQN%97SR!{Ooi_bf~Cfe>?0!bXkQX0hZ$x8&8{ z!5`dSKX^1`E7=_=&-Zuyz~@nup(o3qObAVj4=N;d1g1NY9q`!d{K^KLbr?FEuUuKl z!DvZ6vsiw!Ra!ZB#c7LASf5RA#8H6W3%5Cr(i#*JrIU%ldmpM6?i%@Z-k-6Cg{5m0 zpW7y{ODN4|(FquK>~dAw7}U&4Q`r@?Tf=GdwfYmaEb3>oaC&0Q znZkAGPW5h54Od{beI`}r++ewg&7({`4!gEh<^Ia`n~5mOtS7hy>;``3=b)%?$$cd{ zZ6hWiSpI}{f6b~!mK(xnV%V#^rfd7xk3kO;m0K0>Gv6~D#o8xik)dg=o$3R52SGDR z^@Bzt%l@iu`4+y|a&dQ1c3^XFxZN}B;CBYvE+%VX_t!K1+FfdlyL49TeD!o&U~fVu z=SI{IDw}%>T{--?27;8qVbS!3e#Hf8FKvHpXZjQSxqP-@3gf5h>jw&D1m zJneMKA^m;di)#p0F*V3Oluh5>Ib_Gvm7U(||527@??CW?!Yp1}aKbKzU54&d`r{1k z`{GBo8Q_LsT|62>Q5aG9qVPf?3#3p8D#CH$ZaeGU;%-IvAJ;`U=8oPQQiEeb%ILUp z5#ahLov``UE6fe#Y?pR)wH@>~8=Flrgp=lQBwlu0j%M6II$qub{hVYwx)&V>WQR<9 z`Z4ga%8dNh; zuAcMu=^{`6AX|}JK>gZ;K^!27v(wp40sV3HBJtJ)UQ5V~TIqSRxBw~y9fP7k2^-Ji zWk{zHS=1K_fjpw&)b`^Ypt`K|Oydj(3DP+s5eYO`3<- zQ0=rIebrrXoLg|nKhgibt3j6N7C&tr(VrAiWZ?WyMOS>n&fo~gMY#0uTntMZ)!7Kx zx@GXFi{iZQ3ZyP`Pfm?iiG2@qw5YU?NXq^S@rA03w~JN*n=2EYBA8}Q-GS%41eniH zS;_T=?Rqb!dbq}c=MozMo;!{wUa+x`gCXpN)5wjmhB!!|#{HhQ!|eHWkm3G?aFA@> zrAYf`_;{5M1M0f#W#h=LHTDMeM5&L5BviqrKS#O66xdpU1ZCnz=- z4S4m(%`HXM6XIk&9vATsCv7MM`%{Uo?`UT%vl*fd!3xTE9><;h4fKh1shZ3 zR-VW2(faESyOv6Ccuj5Bp2C=IZdpe?g4&`qT!O^sUvxNGD(YM{Xkca~WjeW&;;(M6 zeL`di>>gZQKesIHFaUuq7}joG(FiomqYO2A)t#2EmR(*yq463jRsXbC{Ax(V1l<0n zZ&0wL?0(zF`gx;UltHDCwHoL30T?ROdvIJ|Qdr4#9MEr-%V0VwN!DIciQT3M}8X$X4qj zn6Df~rFEfMGlGz@&B42?zEEZL#%1z`WoJMLTw>#h5des}E0 z$xW?wHQhv0tx|G-_}d;K>YK4a@sNOV0ll3+s2!qFc@Mh(mJXpPbcSZnbiX(@^q1K8 zD&OT;CH3+uX+LtmI`hYF@Q0}mhUu&t~0ybzJ`W5#9BnlsFQ0)=v79GHiSi1~als8QS2=uaJp>6hH4 zJSuX#-|N?fSS&!&9M*Y5e-;sir8^=92_Yv~X|3aA<5^~f(7{p;pELB)CT97<*pPzg z5;ulQj%l#1O?COd=nP>Vci(i6Uuy@2{3hrAKVR!yd2TM<{suD+4*v7^vEp90o--+U zuB4p)p2>c-n0w<-r-LySVECCzg~UDkXehQ8#a1R$&ehDRev1VbNb$YFE`0Y0&pXh! z;1{$AKh|7h~*LJuq}a9Z9d0ks#{~()as7;$bmCfYQ2$t0gRZxS$p_W(#&iBr?GO9=kh5(iXs$!T+EVp z?=ZU2?1o2Fz-E#yf9$(qoZJ>4b>fYrB#>lIhW^8xIqN{mJ;YU4Ae|~OKCM%%rpFJjYM3C^Nl&4ep~pfm2>1*)Kz8yvb@#A zAlVp_W&~METh=bSh+S^o+SZ*HqEEv%Ao}O+`MrO7@aGFb#&$h z{baAHyo~^6N@GkNGMBHIPqk-zxodWqv2?gj(Khr{g4fiWQr>Om2}2{tgz=xBtn&Zd zqG7_Yg^7&}?_%bi#;l#43H64x#z+iHQNL%Rw$CUW*l@Go9ZEG(#;-oAFNx9(&7y}U zNsNGnLM7^FVy4$^km1WEEl3;H-zJLA@$wJmmKsG4HmbL;55?~@y4oGv@c{_4_`b=lS>ku z6M;M17c*V$<3aNF9st5y&mxxOm?b~QBhay9FL^r)neFoh?!SM35B?$GG*?&o;y_X% zfbNB|33AE5vsMC|L$ThPm-YKHftW{KmBg!Ur-S`^^lRm{QQ1XRxk9Shwh=b(`!4U6 z%M90a+0K8G20p_4Mo%!}@VCq`%f|*;BW;MP1(Vjj^_jTC4juVOS=`*+k4ZF}%%1ha zCd}p%YhR#l0g|=X<@A2CGeZ6VS*>c$$3lba?X1{ZZH>CzuNeBh&ssO6k8JgC3k8;t zo`SrnAMs0S{0GEE%4bt&@bG>dez@D^bi&I z;`f;zbVXf5xvnM;vC=6Z*SxN1%|)8gudsixm%Wmt_aBxA<;5_8ogKA)^3vvi2m{w$ z#N&xC1}7tT?omAWcCwhrnCl*`fNzW&h%ACY1GMOzkK*XFq3 zP#yzg{EG6;PSJqx@x&V-c8O+uaeO43HjpjxdAq)?cP^wzLO}dOtj*ml-CDMXg!|ai zIu4~J>1n}JuY{%*k&s*ZdfNc%DmR6s>-9xmdFa#RVAN;n?6qrGCa{P)LE1G`;76Tj z3>UR}ju9+_%q_+}_(T(X(e^Dc1vpTO+h6{~({KT-%$AwMK>_FiZxW@onH(viyhj4y z7jH$oZuZ|+Wn0w)kqC-aUxGBiK2HU_HFj-+wl$N|roJ^+B>`HOe`dMxy?AQAt)Xc> z@R(W>JuhZKjf4<{vQK7Tb4U7am-{Qtg^9YrUSvn8bzqb-xa#^THEu5%k!IBe33WoHi!^fG;Ha z3$X7f*NeV(=AK~FXxBCGB=g z)ermHl2^uFiUNlDRF4U`VNV_8Iw;L3cyQwVoXp2~i)q;=_)$xr&K=1Bx!Q2C%y%t! z4Kf@-SNq5Rrm$_*UtH@aL?Lr3uy@f{(7mxA!(!_Oi!-rTr|)e}oA>vS`;F>#E-oVc zMO=>=SH1cyiDXLwIs0F=X{3+;`mXd0wmiDKR?_RQOu6$EnKca?Q41X3#be;EmI#bS z)~{z`V!lzU3eFbEj)}-lD=!X8#Q#Ro6_}$0EKjkkKco*S8yXru zUDHtQwv2CZ=2MrB8l=3>-NXon7%@8-SBn*I6zfI>-k)OfRfURK47D7RLl572#_`1% zycAtH=HvF^az`&oc=&sys{LI&>i~~$#ZKLr7OuaX={Kn+w#Q_@&eT!D3-4xQIclV^W1lC-kGCPw=aQI3nqU$0%vRCzqIR$ww$@O=1zg!omV z&JWUDnf~W_;Hn>B7$cKwRVAxm@NKWQ zjgA??$(?ic_X0Hso@`%8lBJHQ{P{ZUyJ;R#d*!BG?oSo z75*0a7_@??yj9ch6{ACGqwPO_@9qhLwWSiGGNJWLLMXp??dRA0H!PiY`wYxnK%VzC zuJ=O4DaQECzJ!hCA~EH5r}brH(Z+z3%&w3;jIFmhs-g6z5G9AXbUG{0;wujm17Fbh zN-gSKChcDoeJTy9_CMcL=sSv!_|srHD3U7TRVX67Y)W!GFL;dh;Z7Y*@0ip zTlN5$hrCz^DTcTV-N(pv4n@5`V3VCMA32FUD`SZR6%QKe5u}mAWR9bt3dczQ?)cj2O9V?U4+g%`1>3mfv+~?GmWjCZ~97O zV@b3vK!e8pEgLjEmeFR`ULXg|MXvtMVF8L_=Um^c@@Et#i*ynNq1g)P;gg&H=z=O7 zj}}`Sks}YF)OS@ET6SNlGcI$!-h~e>t%ggyE4Dk^XMymStT?cpzB-j?{-j(1P;Upa z;=A%_Zk4Tc*&|y@j z8s`P^kyv32)F$Fk0xLP)I=tlM^L|O@pL&U!Uv!@z*yvl_qQCMP0PQe+OYGv!gO@|| zO)@MYR+I#WI{tIkO69&5Yn6fu1w8TJWHK`MpW*fvyyMUW9~{z_*4P?bYS7ap=RwHY zvY@M~`h^vUe=ZLx%~Nm)ecot3T`OPn2(45_(HUz_c@++Z(bc>S1G6gSt$)V3K5_@! zG)07(Eb(G{^@X7^4z1h)x8CE7=5^102=s54>vO@S^t|=$RoN1#-)Z2x?Gcm7{%i>> zry~K`KehEeLY)K7J(3fZsRk-j(Y)M!nLUEz&fAB&t|mkz5;?7p)InNeYmqN~NY#6P z$>Q@Z_~+L%OyS#uA2!OQ=W^9Wg1T}?k`(8=q1z7pEC5R>UrE=&`{oi zM#B9hc@C;%-o%a+R8~n`wkaH_Xqm0R`J^bj4gBK62+rN?hM(CLqn&B|`6rW5$j zm-U{K4SaH3kmRPprnBbZkBK+WIHJ-}A3WZ)s@oDQ9_(R7fB?L(zUF8=Ly2Ere6TPkkz{#Yb93C) z7J8b8-VHByzQ+ek+_3{KytJJx+DJ96#IWu=+ej6NX^(1YINHUHGFr%zxCO0$m4eJ| zkifA9ek*TB;s1N$(1(oj;(3VG6F)iKdFG_E#Ouci-0^G-032w5G>x+PNBNr>Or88* zvL;k1=K&Yxr2w3{*8RNs;!`<%fUkg*ffsp>_&`{l7Oz1TfSYnXz2@%CivZzVZN~+CgWPW$(8re0I&07wY3+FaTh7uM z0q%|4zXe}|P7ou+<0jAbbh1gl-P2y3TzeyJ{=JQAwbu)$>!EJQ(6Q&==-pD*uvg;x z;20h0e9Ix)ADLwL>U`uIK5oT}{Ig9)&)0ueV=m+)#3S}``qMf05P%iiLW}2oRZG1#w|5* zvopUe27@=bXiLe~V;pygY8g_)+bhFD@l4u)sRAK1ivmg0cvJnU0}I;`hk3j@#HKBO z|FUtVeil{(@oE;j#G&97eU6#k_FHv0-3i60v|1dAF>ZJ8G)4i{ms|9eLD4%J6ui(9 zF9c~)xG@4QGhsO{jtinM%nzA!@hY`u;cS|68&w_8wrv}??O#pn-sU(zs0Y{5)HAo* z93LoVSMpfTi4e2IDnSd3#JpYUhLDm!!&L2qK0*c&z%q|kX^Tl>jH(Yze_?1M$ymGqu?!B2*YquV?-N>AU^Mzy4EKKy~DA_s9iQF^s(Va`j^KGqVPSe)KRkvRdTAIR@2%*gn z`3zI@NLkMeQ<{ObcktPB=Je6b1=~#YF}Jqa@+#L)Flu5DfQevXo&=V+VETs&&6DH$ zj@OrG&C6N6XXwB^k&Yof#+=dL?;hgvfJg{zv_*#JrL!M$Lf)?&3pNsU(M> z0d5_G^!+84U(Rwq^_lq4GJ{U;KkS82s)Sg_o1va z{?~lbDon$!LK)yQtxdUaqinFz@1EZ)wQC<8Fn(@VyChEz^|j&XjoSG=xYbg1JWKBP zd9RPtVa7{8Sl~Q;F3q_f^jgrX^x4<1zL?$)pms4&rA3LF)qYhtN8uSRR&{j2VO4Do zii%#@8Ms$Hl-Ymm(6GAkgHoNWqCQ@qc}RTT+oC|}FvFet;p6P$a5hSHI}|}uJ2W}R zz8B0K#v%5&>JAi_$BN$$_<6)0?Rd=a7U}j=e7|iRy!X~`>98kRF#3Mg<(^%IhflIZ zmb(3(n?QXi2^_y1bF%_ zzVAW?M`t$IB_{>90D+8#{Xc514ozu|JIlW#g4%!dEIUV-;__p+&@ZErRb=Njr`b=> zXBs>mdz1V6s)xi&5>oLuP9-HrFBj%&Z+E;ko#l|jZ#y(eTaG1mxVwG#>;f|sU~YR? zh{>4I+c{^wtV@acRjm=PvQ3-zB*i@vZ3MOYO&g}YP0w2XvtSd(yr&dmQe(Td8AEa%VLs4k^#*RxZdvKb{l1I``RJ4Z>cZC zH5Sf3pP=jL_xD_M!{W-F9#xM(9;oI=s4ATZD&{`rZ_ z?^C;pfJ9%qfN3|6J$$NbW$8z!idVXB+Js-ZzEOjBx7H1?bDDMwZBFjAV@uv%Y{`M+ zc|>o0o{UC8nI>GYaOu{>HV447Xei#cCMS780(V-!k z1=MQi$&$e1I#-D`e->L`l^;_W(`a?5kc`{<=aozBqLEKz1o|GKveX3Ib}GKMJZGw^pYtA8@){_Q8MPw?Kn zO8NNU#r%K7gy%4T0Kg9F%HGnyKm`#2SNAa(t5o(u;3+ncWJu8!$}OI0$}!QAr5_Kd z$D*sik0Fz>f^7tlyVwdW`%kykx1tGhh*lZW#_c=kEkGv_P_4w4J^#=O16c_H0NKQF zf$2=eRI*u66rOXJYx#74jiaXF?Lt$tT=EkBxb{TpZW<+;wO z2j!&{2Sk?;^lS-rd$xLuo8(sXZo;EjQKl8D%^)?LlYg^lYk6$k)!uB*xFLZ^*+Xo- z_3=wMrfe#_2t4FpvkytR&?oQjd_PXzSaU1lTEmsIW^WE!KRqabDcq)uVoPBTR?&pB zh~~%W_(_qy%-sej%$u^jJk#;M3_rUdCGBLuCm7>PZg2<4n;aS2%kz|GZzG(0abr}St>~SS zoDdt^Io$ z+!1p!V^orFFGeV+S#sx?3%xxkEZb!{Qf+lU=zMKoRmOZFX+J7>+hj9SY-FIZZKV_+ z6q;!&9c9+GmGc08yhi?=3V@{A`BaL-;FJ4%VJ5jXPVL9+;@G3%N8o+q@tpq*uL?> z^aTBn!{}x^b#9z>&=}Fz`9FgUPW{*j8%UwpRYAxXEC7-_#l1_oXaKs!acle1s+$vp zez?e6w(N05_E?*)2c&E{fLH=)6N?p`j-P+W4$~l*+6i4UhP%0im>{(wBr}_7E^WfO zj!V@b@>ft%g?AK^&#T>Oiy#Wwncv0+HAoF$Eq3;S=+TBrSw%3`f3rh+I8>9!hp~d} zJ0BREZLYLE-_(4cE>W5%o{6{@&0PQ8e z0*LgZ?FT;~L1oTjqM~wk0)C@uOfz(aiUm_aPJ`0~paS!o4jBRy<1)!&myXf^SAo)) z&ZC4QfVs2Il~K(X5R$cI2r^dx(D7Yt^&-_+)d%p@c6>?_Y+1z}uH7f`!8ty)=lfl{ ze%+Ja@RGS7>Kypx*^r}jKoJPBWqz2guuV4CcI-l7UnFMIK(9VosUGZ9;~VJ%`KqqX zhk?OFF#ekM6!lNF@yk0|Qs+C;-23E$!ON(z#7}_x2JlF(>+l_epn!hu47SI`oIgst z^X-;?`xV9y>zOFO;jqbKI!~K8XFva4Gu_mBG+5p&>NdKT{Ul6|Q}uUr8k$t2`1{11 zA^O{CmKifL&c5jRN*KE`MdJ5U8Ql4Oz!=+3@#@vNA7GfZT@C$}(;5S>@w|6Yr^QvV z_t%NU*x4`Y*UrG46@W{R!e-<4kXJd~XzkIybCwgwgIgwS?TA?^oXDLEH}O=~3W^mV z<=-rb5ph$rO~2gs=JxD+#{?l+$zzh?%|{}UifnQAp(Brn#MK8~9ziDOicQ0pr7;SR zGMZp7#wa=NVUtU|+C7zuHWjHt_plrMFKtBr9-F@=)~ZHLe%15 z!m&J|VPZ0RzA|Dh=yfsL@O{bm%hml<)Zu3GC*Mf*T@lB!L_-Pa;U3rcbf5L3jq2Vz zX`WKyMjy&y+7%y?j5`RyUB4I^km)EJsRPPJ`u#Aonx{=|uJi?bYq)j6j;TR*$8DKA%a4<1U%B-B zDc$RN`iUF@Ay%T0Oicy^lH5Z{*1MX=Lt734x&Pwj+Dy;e2gyF~>dkxD-xv4SQQEGEeal-0J|s3zij%=@{@x9I*W*BnsRO)df1 zS*lu^8`TxV-L?^d0iuOQ3e>o8+yJQFx47}Zq3(E^7K^B$*&oA_alV-rMY+lHLtZ{y z4&3=~q(HTc9skP>AIPb_|NM3<>H;_dbY7kj)PdvxQr}dsgI9aXcz$ng=67gugl9`=C3rS=_{Rmefx4`AdQ(avCH z&_y7rsEe5FunEY(-rPn7!yS{Z2lJ=SzxVnI_ynH^=>i`iRU!0ct6J~__Mh29q4=tK zsoV%(bZf93NEGBP>CGUC@Bhs04FG0Jjgt9^pw&;hjDd0t_NoSOa0W`z4@Vwj0_?3F zbj2kJM8j)@CeL1YH4XfrA0KcS!Arbd!B61T;yen<%krCzghux9H~&prdyhV3#pj*H z3b*HMkPdgM|5?o=N8iprg6-t`7Mip_4a;5uoKdnRnVmCyvsY9gIr$J&TcIE>ckr0U zjXdh!Rs}d$mt9IYfe%pt7|8?>e?l~8yB+5M=Y}L_kS{G z%W@xz4gOpTFD?AZIL(_L$|(tY)zfq+xIKNA(aY{E3P6Tc2 zPWqDGd>@NSwf~^_IO+VQ2!l5xfwf3H-P5EYp)0XW2-O015XO(C?Nan$^mq^cN~wok z-DW!m`+yJI8?5xid{dzha%4nV}8!ILc?8^3+!5bjWupTLYhls>Atp~ zYU};Y@I8@BRP+9qF_Y$u8AxzV`PKLyN>RyG(xg>-eKE7 zw-&F%uO*ESjN8!j&PGJbrLdCbvAg3=*Vjg)SUC>1vkvRLBgPYIrLX@mVo`>|0jZ(* zpfwki6*WEM+nkWuwGSw(MAZr^&VcRY_-nRdf zOVD_K7JU4PKqIpT{g3+*4T2ZHT$$OEkb=K!}L69-^#%rZ1(8NI2#1BmsR z4gtF}MraS?5U1fapoVW3Zi8>;szFxzULz25Hll5QKNhL)uO-?w;H2tvxJAM{G#-sYcoMNQ!N90VoPr2UWZ6wt#D81NM2Q2<~7hV~zwWkKVu0dxG-fb;m` z7q?ONJ^8)F`rWC(VU@yDqhhxrPuh}0x`8(HW@?x=92EBKuJF}dVztAId@w-9<^fmeIaGOd0>o{H-6Sy;ject) zxJ^sS%+R;eCbJn+A=#Dt3xBcQZ0p!7)a>4|>%Y{YR_gh%LqZ!}X>49fW*Q~BX<&T{ z3Q%eJ3N-MQVgIMjT`OJwm|@Bm$55@J%~S>_{-AvCqI+cMo2tk8(iS28}$JMn}U z&=TWt$ahb*ik@R^AMI}Vc+0c=)|nrt7ILMd%5!VpD?kI$mS;$(IDfL`$5rJ;l`vwik3QjmZRj!LwD=L5*}HIhuC*?v$WbXj zJ|kMSeDJ{_N)aH6d6V?!eyo?cS&Al2fp2b7t0GRp@w4gm;8wUL0YlYy?JFJ0hG`5{ z?}e_~H)@Ra8F1;i^=Kl9ivE;@$&9lw8zL9x>s(Na(FrqYgLMC#Yu19jgWc6{hP*OK zSM-z7-RiW8upqak6LRIF6UD)_#`%}wTO|Y4j*&c#U5Nmw5cs|!uID5Vwc{)W`2E+TQK$_l6DuZeSFs(N3%Sw z50<@4FpJxx;nKsqQetEERoXk6I5Lp;&~aU=G!K^A(FtyNmDjkrkmf!`QQ3FWE{q?V zs~l5b%)9RgRK7wZq<#oi~B zyoxYg^RsJJyY%@!U>RiJ=P?-$W11!29^d!nO$n72Kb;zaMSS8=mdf83qlE;8u#M$J z;EY8B88@bhLX8%UlqJ*PfRG7*u3enqxv725of|)a@w7+g$`9ZW_I9fY`Tu>Uk2)MIqsPD=_U@OSp z_|UkYn9bna;7^!mvo7Zif*yLqT?p-SBjW!%2=ofYGfO{UF<_jl>}l=RGR0@^0@9F(CgH#!nO2eDl7imk zJ{_nHpolKT3DsY?pm=jaSqxgwu$8?Ev6-7f&ut}f*JUy0rOnTDvrJXYn%SSrazZolM~&}IM7hb)lgSjf>$nk91^$bU--ke^a|u%!*e} z4ogLRxZE(J6>%6Aq7Id zd3=T3p}tx7ELTmT*m~SniyEn`aP@|BUvVAHd3+oS-|^@<_j$)Em}zD@s~f11aAZe2 za4%ZJ)r6#gXKZfrVD`1D0y)bknA|wxCD<56_IQMv>Yup8#ze%z(T*1Wm#+3(2BuZI zY28xDJz3$Bff)8!|B%MS`>chh3M0iF)hqbRLxA zV@zYnO;soioS^&0Dwu78 z@%7=E#&G89Mgs9}DZ<7)E|$y=U;Mbpaf;eh;}9Gmn8nP}+x27E{jYbB?_y_mB(o%0 zUGra)!Kxn#$*;cWZ{DG+j2#5Q%Q~TZiCX;fgt)((Vdde;2Y)0`_FOl>;(#N7Vv)d< z_r!vKHkr1?nv{#A9@Zd46L5~X4od{)PuYR50&xImu{xlVpi;mMuojK5UcJ@#TTch? z5^`_sK9NJo*<7dXOw~lL8MR_Xu3-ZW*I)PgdZWWNghl{PZK-& zJFJm{(CKbl%?_lvDltQrrsGRksGbT~X}x&eF2w~;Z&7%E(@W-gm@2}?!P8EdVjGOk zc_#yKuP9gvzT7803Ap!n*ahm{_^F4}%$_1h5V0LSjiN=F;yX-toi8G<6e7uy3^Un8 zp6n7ig7g7-{Q07Foc1?5yS5#Vx8$!77c?CI2*hQ*P1EPgUVpeHqy?HH4koKQuQTQS z>8I4OwPvKJuwZ~)+fqsestOzLuw$+%L^pf|dOHh{_|_ce06gimiigu5Scd<8TxgEw)ys zDbj(9xE@^jUh71gO>zn-FDhd}+o3D!@@&$`qY0yh*WQV4G3JZpCik?JHY@uck=+~! z#(6sw6&V!SW*>W>#Lsj*mWw%4v&VDH^ju^H;&`NK9c?zO_OzSAr);BO@|)Ahr&&DB z5h-@p1&G4E(rX_MHwH7c=gs%CVH~i%6TxSzQ~WVjVH-&~wtUZfC4OAzR33D4@L3WO z9ks5!j!!94sMA3;{8rHXaKH#w?y$y@rH(6##{A8jcJcNpA7jjlMFyv6ca`&pF;@SA zYi5m)?)V#I%`A`@`giVTEhu}+1SdH@)ZHGk589FI!zGnF&FAojq9)g~C)M&8BPYd&QcXj~Fs%-}&2G z?R3E?1+j{jUr_iyN7BOLI!v-piZ!`Ky})#dC}Ba1A-l(0Yf^OFX-?@gy7hjo}c%8hZQVTK?f}LSDk4ZCn4pA86<>V zy5lJiE{~6pPluGeX=V|JyX)sQ9K0-Q5{t)PV6b5T^JXrb$N@EV^t$q;p=EZZK8X(m zt*YcfM@W@)l*_(8EM*e9Sdz;j$?DTaC*{qY?yRTqZ6#EcJP6qoez*?qu#^Ss%+161At zIAYk=YzBl7zy`WbRxJ6!tbmOdogI1Vh>?El@!-WcYAFdVgq&~aeRrA%p~5XYB!<~0 zD<8JsdMz*3~>llp83GMA;BW;= zl0Ix!=h9Gazkp*gH-b?O(1CexMU*9Fk*UHArYvsZ$BdaSd_`dW{nhhHg=${ByDBpL+hzdnHCLwDnWZMW3s5wE%v}Z_QZ9?6D-lUY4^1JPuL)_myKbx?G>-I~WP=$!Pb)0kC{&U!z(uAiz5%X~; zK(>v=_La`7WU%&3q2&iox zb3ePg7^<s=)!+Sk7ihr2B)z%_8~VX<#r z=N_qSg(dfpv5npOww=d7hTQ!;(BJZb3&W#;H7~>AyZ>r^kK-zrxPC-A4<-C(Jh{gh z1W(EQr0b@I4VeCB4mdsK$a!xB1hU=@eugz6T=8}*0a;6a0DJ@_UdK4f^bLl|<<1GE9S^FN&DWB_jhtOU`c{Y{56cn!62Y5hRPMKT)&kV3czIAg=s zTx6Mt@&Uame|Jre2#yCfpB2ae$xHqs4{&CK#1c911eFdI_Ya^*$!);3F`y3MVPY(z z4LrgV4Z4=|Cs>#&S69W&iF|e+q@8$*02W4`7niw}P}&z9Q2*jjXQEK>!EBwRNbUB0 zCBF8KURyM^h2xr~=#zKscr&_k#(7$bDY2Xs2<=x)WtfUhOL!2MCF`2-RrGB4LM?1d za#fz9I@P5VE_!9bh4ux8o@?61>dFMrW>{A$KE+t?dsshVtO?}Uq%obR9}Aza8s7>$ ze5BwbT&OyINjoWixmxilsA7v{Hjx?a=KLfovys}>+Y{mgRdNI~_-R}&>q+`IA>=F&y%)k)i`gGX_tfd^K0zgep#vzYw0q@%N~#9fA6 zioHX}yxVB3z#-RkQ&Uf$Q4HIE-lbqu*(Zw6y*o^fw!!efClT>`VSJ;Zh?W;J8{Nij zX}uKV!;)CZxcly_k(}xyF^5z$e#6b~^!czH^n_D(_rN;z1GI18{YKfB-q?PQ;Q5sW zZtoa%&kLuwBh+@bYaeu6@k9taQvO8D69jO5e?VgtCkHM!htQ*5y7{z+K|T<|@Wql1 zrirG)#PscovGm2eb7t@TM_;Ftmy;P0a;Zdk+Gq1rW80r#2O1HauwWO@_Bj04gUHEpCChcp?ikWYJ%+7pb*F4y3G3pdemdlFz z#cDG^oyQol7IX`pC@#l^`6lJI%ca9OKi?NC+w;A4fnmX88!Z*|Ie52ta^RD;mY!9x z*H6{i@cZtG3mBz)(Tn#ZAr%Mtni{#=hf^PItx6(p2v@r8 zx_h-cEoHFolWpPUohlBP|LS%5kDTWu{EioEDE%kRyY@+T#K8CxG4u=Abovy4-9!Az z;CM3ufE0=Ebdmr{n1veD0d%oafU|bsvVrmlN3TXEpEZ?GCP7>tes9 zQncm|Tu)$b6z$#nsusG$W%tt7i?%XS>-PX$O^ujrznVXPivnS=d00N2)v%Df5YltF zaiWo}ETqscXnV-q96BvxF|r>fWc%QoWJCry`oK#z`#9NPhv0;GDW+-F!PaFLjC(1D zyr)et_1TWRjAOgDwb1Bu7{_O$z5t?gp1X)WK?BZ{zj%%iuD!@fVs6UWNqXFZ))9DW zLp;2~6|$c9Yisvq8dGejRvl(Yd8MQK26P{{6Mf-c@FN zQe10&mFpL68AG~&-P_gGKOcA%#+EB*MJz^54;{nBCJ1*F!Q%Q)(E1NE!v z^C)ZktSi?Q_(NOkQ=X65r@V>{A`f-KH>a=9wixlWYKW#9rJLst)IYc}JfGP9ZMiGm@EI#9X5!8Ma{0TR z*d-Y|Q8(HN)#9?qsxgWL)9=f5`}ej|J`9MC&#j1VX4WNDJMDmbv=v-Os}oaIo^0FY zt9>@1O<{Y0C4gp7+suD~Eag5hsAYXK%U89MD`;1AI3Sy<=TwQV?Ww=u#KoLK4SCq?ZjqZUq zd}Xv+D@0A?hxa$W`tbZ|ItAm9z^X`V{+gGqz)6l<^}A!;melAkD3|qQ(2Pa<;WOWC z6>yI_p%68 z`YeRsV=NFDwYH~XdGJbHeaot0`W-}bvAREqS<|9zk98E+KF0IUH&N_o>N1yP#(GYl zoz*0H_Wp`Rh#HDHC|Xh0T2psodaWGSSdUcp%M%*s*cG)?c$#_k=OF9;-A@?pe$Uf` z`!$8pEt&E|Ic>9M2m2|^{~q)=e-HX?S@jqHG3;#muo3O!a)c9*L#z;_ns6F&Xr}Xk z$#E3$n9UTc12{)G3E(8p&J3q*5aOFSiyl1xGb4zUN@Yji83rj*xI8}7K3C0&1p)N{ zw;==jY?**ZJIJtopyWOoIOZ>f%)OTGIV6!^HFjZnc4%S7N*npPr?7HZ?iR<9c~2w4 z*#^u;AnEUPYeYmaORddGF(k?{DjZ{_)KZG;xkHqIq|Hd(`k}5t8#Okt8U{82;E}c+ zT2VDbNC|oc(K}*6{P@KDo7CI_LSR48xSH>QkC`jUPfJeD#<&qd#Kd3R zCTI%D2r7Qz;jeX?UU27>!tcm62*!-AgaL+$ZYdr{t7&Xmk?7ygG}ZdeANbuFk%Y}% z-Y3YAu8~yG%LtcoRsFJ?8(BWl;3$?GcxX~;Ry|IRgOH<&Y1ezH%~ST@ z_pbMw+$p$LY})&wNs`)?!7%$6-8&m4LY9mZ3<_E-UGMHsnCC+@=FO$YV5lonP*3+P z`RdsoR$-IEX?j9kVjQC%o^~j;8STVW-l6p-OG#VNZM6V=7oz_7b%?VWBmo>@J zHJgi*_3O(&NZ=<>uY&0Vc)_ppaM#_M#DTx8u+6l6Jsy(lqa^Z>nMi+f?KlrxY-38x zj#;9?kqV*FC~rSyKFSlBsqOLY%99#Y{M|{R&Fy5{fG4APM-PW4 zX(D(luYf+xd(VM#|0muN+5)aR9H>O6+D*JU;w}YK6>o9Jrhsj}4OZrWjsedD9kh@j z$04v}As%d$C`cLuNuFpEgplSdFAQw~Am_iz*X>V%l$HU%pALs2E*ZZ#dW8*nMh(eF z-ujyx8Q}*xAQ(-@uZXQK-T%Qe7PO);^}FRV<3~NOTpEH5wW-V21t?8NqW#$HfX5qu z1b+i|Pw^$!pYfIuZLlWXfp`mV8XLk?%+W3w8mB<8xWxvT(b zXrP*fkN7P4=J{uo!PhxC^+hQzhT+Q;nT}`Ib375==|x^4=AM$ZMg2f-qZ3!aj@4u7 zM`yv=Q+~;1*}4iJiE*doBFT^V6o)7F{&J_N(tCNvG6GA2ZnWm7n~A#7Cj*PGwDM?@ zimyuhm6glDsRC{5pb1Ul_q>>z^?{PvC*Fe&PHTH&fwevr1<6JGu7_T@SJ{FuX=hBV za`Y4u0~n#kdzZ%RP4XA6D^~f%;#|h4ml1Zly4rso*iB8<)KEKeulGlgxJCF*W)pPsT7j@4zUi_jkz! zzy9hJ?B^hK$z4=yM7e#DrYYZCNU^9nxkImYU4PWlULH=4>++H>IpvE8jFV4ou-+zs zAhP4Rto@Zz1^9HF8?mu7Kko2#9Mk3GY2=!5B+mP?qNG;YB4 zIgUwt4Kvj$GQi`~zy{aJVo$moLH>|L=krUAm7UO0f|#ODZNYLf`rAV1s%t2~K!98e z1fYKj?})>LWRti4Nr7K{QNm*l3*x7t|HQ!W$n@0M5-{lNBPv}z1JF-^#|e|Q!yj4? zwnw0bOEKqWBEpwf#pl1USU+lX#G-veuL6uum1tiNa!-Tq0e-5>qbf22)k%#T6PWla zaPW|*hr|>~OKx$e|3$8gI$~oy1GRSenwOQ?pJ}}69+lcr7kp64mykg$S`<}cY&v8WK&AiQz7)spzgsFO*x>)ejg8^x zM#N&6q#3ij+qchs=r^|&4!iA8Bzo`rJNFP1irZ7>EAl=+i?*Tr`TC2!ieC82K*|ue z;-N(D>1B6>F}I7WN-fT~ySPz9qPc?!y{c(rT8=MpS3kIuLXjyr{YnEgxas!2_)E{N*@Uk_eR%^a$$OvnFnMU?OR)txd$}_XIG7kQ0CnKo`Rc z`GH~OYq1$9s$)3%X$cXaHA*a|^#vGG8@|9Z;{n-8d?c!Du*$P_S08Ih>)S7Xe!e*u z={zGuTk?nuG`|Gr9n>mC?B~IAXOQ{w=ymY*IpFu4^_2W0YCRzB#Hgb(&|2NSc$Z)v zq@HWs0OADt%?LSyk+;d|+nIYU0F}{wd0+9{u?6Ch#R^aEto{*1$v?T=N;&WY=RA_X#Hh z^)dfKRGG9`>_H8TOLN7BEOuO0#GPmH?Jg=P*}gBgxgmH_8F%$a4z=2*xPgyz#1xOu z4~N+TSj{~xT)1vvNVptWM0v3>_LiJl)bTjVLQT8hsoJZyd-*pkRQ`@ zJ9*znGB|nAGGDS-sKX%YTGh9Dv#=79y`5=OIJ8rgDs2Ihse%lR>U%*#o>M4v7pME zgK6lSO5)dx-=Zzp5+KoA%ai^!GBcU#v#QB35#_~K1M^vu`c!qZq!8$!t+6bH1ADVZ zPbzt+Lj!f}m%{!(UireK4Jt4+}c56`kPI@r%wl2y;h2_fR=QfFW6Z*q|h z+{Q#-ZZw)}>1-OeU+UKOBG;oHWBniCJ&1KfhkJ`$=VcLRlw@xH;N3##b*PPLpzwOP zhOb;fvGZ8Tb*qD9uKHg_X8S+bFS_fj!!}|_S-m)ptO-JBf8CZ$daPCThvvoWCHjKH zP%H0&-Uv!cK-ks;%JnDvKFy$66m4?A)*M+;mm}FFP6>u9suOe=?0bm0D2)sk{`+gNI z)w5Ud4n@c?upsmfMRPd)_c3~7&>N9hu;WC#J<|V8ZVVvRqJRoyrHuqcHJ$C1mIf}7 z6(s=@EmWmeM?)Fu1`lKLz!pw_MjsH}GtF>-q}N8kkFm#$h^a2n9rcvn7O4B>N=ZQ& zpwt~uNtHYh3v2<*PyqgVN5G~DT|R7n<$;`{GWZ58e01+2g7rZKvlLYxO!7_)szvv_ zonQk!!&G8Oa&u;On@!f}H-7VY*6 zL#IXo=MT2aaR!UHrghY)!tw!%&-$n1xA`mpN)n(D;db^s`}1NpkG;G143AXF_Uhzn z4{nxo(M*Ph!V>A{Vm=R{lT42s#YsQ0O!qjAg6BRGYv}Mp@{o;$BW|lIv4L?C!F>tw zIU+^Ou8U8$wUm*1WF!-qWf90KuB|VnD-$Jl3VpEG0Yr+P4u+kYh6};=pf|Exau4zJ4eHMcw z69_2@i9O3EiA0ye(qhw6iCHfwPVF%IdxkN|@aQsa^AB+2?hunA|BkVoS1INLRc#%P zS8$ap#A*A%?rMj-cYT@<%5oO>x|;3M0?=a4EU4S)%(+mkY8|fDiVfyl><{t`=O>1% z?WpD?#RP9(mvx<9DQLCaoQ~!87H`TPlM^H@XFterS<Jv= zXxLS&VvOVO6LY(k-!b3AJ=f#?Z?4hl_MbMeB{AroMY>B_6wfS0|0W4JiK7QBef#8lD&97R$n^% zXkWm-&XZ*(_Zzn8DBI*)+t2#_n&+mog!8@k#rN2*(5F)AfkZ=-TVi2S!_RiLMTl?pJrr#F3THXqj1JvjK+P@%rt&~51Z{_*U2)A731`Q-GnT+ON*YLU;+iP8o=!}lL= z)*kk`&y?_L3h3xL4Y=>MrEKXPk>4DJoFC6BN6aE3L{}`1dmRzOrwiQ81QrBkZZqDm z!R~;dqgEF*vy(ffy7Okh<>cQl{?>AT&XnqQQUdXpRmkb)ElA)UztW)T;!!R586F*D z_Pd~)rL?AL?<=2Sn{?5!>h9n9hq$Q_Oj<<(yR*(tnvSQpX3*pqE22d6$@b>sfxw3c zEwqTT`VCk2DWAL?XXq(GN4f6y*1F|E-LL(M9*Onf!yYSNf7GToc52seDM?4}6!mxB z?lS&EDY+v{I?Ml>x7By5ZhB5YB2?luk{LjPSPx3MkHgdiV2_Xf|J9}Hr$K{i;&Y)|c)wa>-A z|8$>b7An{mzaAT&yAVVr1CR$xaJBD)h8ei?pQcSiE+?F&{#pnBT=(LT#wE%eDwR8> z=ndx$dF968oXwH=t~m4VTxcFS1$$y&8K?Yvp2bkY<)MXEGfKyXc4@n&`Sk0zgTuxx zbm`tsCYUJUbLGUrJvw+nzo~Moo$_y8ultfpbyOBETsNxL zK85aYu!;@KLq}+vSe)APp~aFWgBICr0axs$1roqP!TT$H|0`!S@J+6Dr`Z#9yyrdyd?DN$S{0uZj=YtUrXGN7Aqkb zDXqS9fUga)=hRx+)Gy!N$(SWRl&0bHsC6t5+u=Owvs-?&-%&z%J570KhK?ND&8x-Q zifoGQw$1oWJk4M+%u#Ls)wSKYLe zv*iQN_Rw%dsgTP5et|{PBIup0S(ing1jA(Vztxme8={|~yg&Yt+Rm;d%ccC`2AZUE zgzhn$8ERUD^cnmAV6SsWubX`h8Jh?XnDRT-va*fbvZXtVyPy_{=DG1_Ax3iZG8{hD zC^bEEJagbyF`CZOcPLg;YrC#vMHq%09>=ZxyE8xiz|>;&Pc8#=(`^5@_kdh=;3hZz zJx_=(qvB$St|_6u#CZDy^vK8Ey|*0k0WYT1byf6~FUGLePi=3+r2b!8d3rtqI0Br#Gj4`tE2zxn!cthJ62j1M|}>p>K=|x!z(xEnI*$5H|p*kT%|Z zX+P60{#lIKT>ARjA4-=FVh9s{@b}0u(}9?PFT_2730yb~i=7jrLb-psELelWLhv+f z1kkU6sliJh`2cYB+qadL)EJ#MigL#!#w>0ulUh>3Y65YU7}{RoF<_zy8ypDkhP}z9 zw8N|IhyKQMG7Sjy2$n%&B2<2)rMrz=Vznew>3VG(aqKR<+C*JGd^cbYRyby{2He&S z&u7o$zFm^`09?YVUw}(_KMg0-uOH6wG>lUrRNHSV^ zv(^qQzWNZ2`3w{CxW9d&mQD($69|V0N#0Rb@?STbomAJX)`hOQHXSba`C`(rXS+r9 z|FTb^shIl!8EewX&Vc*wJR_!I`=^+^vPMj~Xz5<#pf=x4sqpJV2^#TBVbHpaJ08py zE*G7U>E*ZbTmXXB1OMb)H0CMpJ|)Is4Jbdv25@(YS!(YzSzlsRaK9|#rnSk-iA-R4 zRX!uFN0Q+JIPv7|Si^@yE6Elmmv;yOO7W|GC-;NO zxg=4;k5DnC6xk-KXetQnY{LgNyI!nX^S3`bg0fQSwC3$>-Q4AiKgCLkZXw50HJ-i9 z3;~hUl+-&H1}G(nj;lzp_sj3VKDIl!#LO1#SqGPlh?+`d0HIU9F7B#Zbbbr^McUpk zv(j#jphY05uX92EY}Z)M7Hc0h9xu4jmCLX5fK;+H?1xeqr&Y}B^z$BVS7UqWqwctn z-5&QBbBPX4DgF-JC0g<>F$NtS?b?gV#zk-}%BzyoJ92$jtuQfAuqL(Lx70Zpi$$wF1vgQtyeUUXrTrQ1YNp1anX=0Ln$r0nhvcI+bSzF@zjdBJ_a zjNJrdYiaAY;iXm0W?X^e+!t8((#M=U-%cRM!IXP*dR~QH+*xQy6S#V#7PzShxm}s! z@Fheq#}fZO5lI;gXTj`JG>da$0FUQZix7LC+)5kwpZki)kQ)kQ1JTLzG zs~?F1!B=;~BvCeCE0``=Fj)*USdy#K`W3ne!NB-QT0Eru4(*AwiQRJ>;aVh zqnif%t4Yyna$H%a?puuqZe9mD*}u%a`#PFl9&nuIyU+g!W5&G5HQX#~V==Vyp?z)d zKG`ud+=S2v4JQGV9a9hy`>Ep}B&#hs+BR5K3g~Zf5-EC)>-DA9OI;)~##HNiHlnj0Mbt(Lk#KZ=yE zxuqQ)$X!qBMi6~a!&x&U$#;F0gjvSAjlO&^KH@gcF^=lDO9?J+%dOg(dW7Sg-T0ln z%Q_&Mf!WzK8^;_@HGtZ6QUW5HU|pEQWf3L(X77yw7he~|HVv>^6j%g_l1B1&Tf6SM zuZSE?(4be_RB4;%Ah>|R7|SaEU_OM2^?OBS@8 zdZ!dI&ZGF2d(;mq4Ah_bJ8`r662bb^xf|($KDV(hF=}s0ry=E@`x3}---Flw@<#UY z+6@ay1Nh>AN%Qo~e0R}qcr#hU(q7NOU(R^vsqECP82N1xIC{z^N-Z#hXw#RR`YeEYMQl36*d zNMU2KJ?;WzUbGYU_t(|$4`D$E{{QM3^#}Q5<(AECoi5S+!|sUg(A=)aE(L91h)Nkb z!q`WtpcMbJW_LC7YuJWP>0pAlOQZznwzV_McHEZ{n}m})D%PbZVY81QQOpaK57ncSm_3r*)c?OUWG+xoh>^1D18y(BBOoCcO2a0c>y;r@j-h4% z{2T#4v>$|e{8*UP;Bg}L`!!Vpow27x$0qfTpH*yf2%%ewQ6v&mNf$Yx0GyZwF%E!g};E^4AVk>*uwNeMhaawgS3^Jsgvsc=aA*7^R6}C9zmDNFA)HD4P(U17yCI<1WUec(x$wfdCJd zdPs%-1qhJC4%(uno|+S?u28?+rDu^RUk@4wsU3{lC1~s9Qgq-G$@Gs2dPtFJi<}BShCxW^pu6`7}b7fmQW~+a` zZHthbGe63HjM7yVx;B?~M$(L6*}&7@H!rikHJm1(3D8!)dB{*V?lC!t#*NhM1S{Xm z9#cE~z*M?oj_!A(F9YikU#XVGj7Mx;MTv@z+1)t2uZI4#F3;ys=#rKI><-;0-a{>k z*y`i-_~azH0ET<>4ew|}QknG6QS7QHQ-N8Nb}i|Fb#!QXvWO*I<=4QiWPycn3-x3N z!Pl&v9ZcQ6*yL*jR%l<6?;4V0?(P)-T5=80IXZ9NQz4$=Zbh49xad-_OAn%MH~FH1Ft=T9W+bMKe{A9@MF>{KN$g^0j}EeoVlx7S7uWPJjY_B?*^j3G zhL^W4qQ$G28Bw^Z)J}JrbrantL)Q> z>j>z+-Xx^hVmWR4w22h>7!vF+q+`V*XFGO*FU2XfA&UREdpY=&B;sB4j%!vkBG3hU z!oPfDOqIaFrxihU@%xYn0$t}OC>t^lO&|`S`gNvF7Gkz|BeDtHIh0{Y)_ttO-m_Sa zoo9sYoRBvednq-LylGYHW)J93`B7o3gEu}~Ql^C&D!9OA=oVV{UVmd&9nQwop0nPI z#mYy#Ri|d@m9fg;s~{?R>k~?Ri0|^eG_|oyM=gt&0n1Ele(05hYB2hB$`C-NN4!6`xpm_#oW3>ZFMO91yL6|jMC7%U#~d_PJWkG+ zw+0~ya`D`(U(-i?mN!fFZNb6!#!ceCNLpAy**jgs*|_v}m3@hHg8V`Hk@uB+QLhBV ziCnek;HMdqIbno0Fm>WfCC>Wv>VF*KT_b`l7;?}&Z<3G4a zAeZDWfcfKHkjAd73LvQt!*{9-`bEjr`gZiVO9{$)$+e5xxF0OR-{idVgzINNZ_mpK z1*ddbPzjO-6l|UW)2xC2jLQeiV=P)Yt-12J#~F58UwQ626X!wZAqS&0jBB^KpFP|8v z!MIyOXk-f=0rCRC&lBMSoB%iU#I%N~y^fI+&$=sv)K-Wva1|yrIRE={mdlX6F$9p_ zjR^e|4=aN4VW>AOi~NY?oxWyn-=79iLoar}Mq$d;UjQybTe(||e_WTNA{I``H_S6L zw^LoNuu;#K9vSB5v#~2r%WpvY399!r0bG~z##_NkpsEXF;&*RDFxl z_3Br-4749(zmpD$UdN!xI;!+Er%_J{jF2PF&IRXua=%KfNdI8x2%DKF`Z@2uaYFpKd`|6FdC;|P@pHI5J--rO zp0~2}pY_VAkynOmQu!IGGb?Zzq~yVj-RBWHpA-3Emru=Tv)|rvu+u?#8=KvI^4&d?k7J~_(NnM6*aI$ z+|l>(@gozIW6#=giE-c$W0nIbahzv~d)7zN$?Zp}(psk@%9Ii|9kmn+U&O!p{?-7t z`a5X6e87FWG*2_+=tY9MB*i!K=B7@=su1J~ujZ`JOTp$t+Sx%kZ_4)|^{S5%G}Q~j zqU5fm2(Mq3K^t<2VAa&6)=M%vl{Lu3B5$Jd0gm#4(QX}jN>Q(L9<*csdG9P=N$sj9S#st3|TVq(nam@z$&HAl=A*y!q2^HX#w2%k>-( zP}%o?cgZ`bqHgWC()^yPuy7|@H`@FNUz#ut6SYTeA%r&Seg@mg{}L7)Y#(3GU3q72_df~bFKL*t-v-gX7*Nc8#@p%%@JOHu+f9H%jFEDaO8N6n zP+6ZMqhXAGOs}{EP*q*9{wF|^>nnz>z6RiS#LW7>F<4wrRRFRCw0*|z(e((Y?g%i6 zF^_o`W4TDvLU+EVlzk9Y0}xw#L|D^rek{%hKLrGf<2Q`j81tTmAzEJ>)t`&kUjGR! z&aFjkR1IG=wg*3|v}2T9bIccfJD9cbR^~bBwL!QNobN4QQ0>K5V28^$4#$WTF?e(z~l-ni4JvadZ^v*bkS9c(m9tVARp%g(4CMA;wozQDD}9XagL;-aCDL|h`qzG z*KfvQ+GWlN3EstdTF<_3?X*z}p3^7r3x6KGPQJ6OQHuB_IAFQs&{Ua{IO{BF^t{;k zB#+O%#nPux=f`azPN&2+qP8er&R^ZHf4)P7LvhDh*0Bg%D=di2OG+s}Z18O=9MEL6 zWxCjT%;SPMP+M+~e*@$BP`v(Wudjz8w~hOFN5IgZ%`t^lwo{jq>3&JS;^%KY$%U)a zFq$y`?Z)vP>Pz9$*Y>}FsG9kUM_Ctx{L!$1F`)IA% z$v-s_T$FD?Ob{WGuvbbMS;-Jx!5*c0z`n`XesxFjj3S%kKOUMrNNoLBeip2C&*){` z=IU7HWL~H_#|fY2tuY@))#0df;VZ&dT46NE@Qd&|`evLaIDjw1m3#q$DOYZ|t(_lc z^li*EO;jIa-ktgd<%cgamY$w(1=AjTLiGJy5}Mkn4$6oR4UEm|Y)yRLOIByra!w8S z9~CM~!&cgbTqm$c3%CAXY9Uqwl*Y;*mg*e)-%#c^wz7{|$Z~&LrpLNZ9)V1W*z&w2D6ZwE5{1DUJh`yh}Ph?H4sswA#vIwO*N1HYS3q+|x z582FI!hQpTHSv%~pUddc6{2}+hk+GX7Gh#sgr?UW0oe-8C_t2jvAO;UaO*E^7+eam zFVG4BaGmILBs0#8&>B@dUHtTl`{*8|4`W${6m-b26{$Thgo;+Lq^Mp!x&>7PvxJ7b zyv9yY7g*IJVGD;?=cON#WV;$xK8={^YrL;1l}cde;k?l4>N2Yfct_zgI=&2n*Nfdi|sS-bf5 z>)RBHn7V4UI8B}*pJxl|a(H2um9BQ=V?j)!lk_a-LBuybUsLv;Wix-y^_0erZ3u2J1ocVxnD#?c^*8*jn!;zevMu4>R0G~el!zu zerl$66n$N{Y~dD2hQ1Y6(A&Ly>!esyg7{cE-NI-2a-GOAzQb{F6_2~Qdv*6x_ndmo ztbHED_owBzeEAJgrCw81+HZmWVcy9q?@Yf{ny>!D3A6UgO&FJNH!O)AwMXjEb9Kn~ zsGBI5*6H|Nm$2p}UV+J+^)nngv9=X4I4{_(=#`vmuYZyTt>mr~SZBv95wn9tkq?m; zrJEeIgrDR78tx}xSHi<=tXsZ_oV)>Dv>DsWq(e#F?uGI$^!Qa>nJ zKjw&P*~+#nCNsW3ITjOsHZCwr!%s{!_P+jqCq0!a9sVzk7%=2Ze)3;2;@p$o-bQgc z{yGg%<4ps=TRuKy2WT7$=O3vMw%=M5Aj5|4YS_?bpnF^m_qhfL8EU&1|rto4uV2+R?Mb zkZUH1PcYj%wl0rWKyxCDM*Sl16tBqL7nJ!3`H3X8pgg`gD5Mr>@lv2tfWu;bYlVNSHRfnT%9UC;#FJ;I}x zE$)k=!b#5dZAOF8$<+bB;P*?Yy(7k9mc;`_qbuITC#r~82Z%yh5W*$`t8j_;Q{Tzq z4xce@teN(a3PspR+`P7|FXLo)Ax6^eRQCtZ@TTO>2TGEFp6_;amaWhc==VZE#q9Oc zyup&&5E*|$>rLr9d_DyhU&wRdCj51pJms+tm8vY~z7^Dzgfx=wB&$9)>2==H#m51W z)L6{Y-HhH2oN0j=w*SqI_U|yNmEJeQwiUGmCB>u|yEQcC%zP|EGyi~JaljYt$d@eO zYgQ?W`F@z)(5JtA8@te4-VwA1dpROqT1M~9a+($LyP0C;x-7U|1a-H{QyZI5=FL3s zeZTqU#5ebPz+6Wo^|5rdeX{(O*S?TuN<0@5bYx$=Oz(r_qMX0L$y6WGSotngO1>Z& z2DA0pM?zkZG|rZ#b~E9>BMw8tmR0t$MHpohU)UQ+9H}nG-gq)g5N2<}^>=0S?-ifN zzE_=?n|j3N?5~5Z#*wpP7kk!<@`R6`9;}gdR%9euR?Qu zf6FI+q8)>gVRnPj5(iYM;~85QpB!zPr#WW;RFfa* z`E1rWFBzHIHXHlCskxUauhR8Htxx#`8}~+yLQak@=RXPCAeiLatx!by_Osj{Uj1J= z$Q98$^sRUG>T&;5M0gAbGJb!<5#wh~O#-T40xqTL<_=Z2Q1O5?8f~vIKA_i_yPPPd zZp&30435S)0}KICb;$}xNiB(has{o*FbM!PD^nNM{qGLKO#J0@)PRd(bMe&I>jaMI zwt~eJxjkt8=lKPJ?_1shA~8@#=Q-xHP`NX!0+2X)q}ntTxEPD=kO>AlQ@~XDNL#&T zb!f8y?!HE`FUxM?11GmMZ4z9FUjWz*imU+6=D?wQfN6l#xw7`XXcmm_ZmERF;cIpy zr4s>k%UNB@El3u!N5|>a!TqF)P>Pl-K_Mpz6Pr*L8{6A)A_lkL@+oMlf?F( z;VgFL#=gXSq+pZKGi#C8UzufY=d4HkdYZWm zJ$vI%sBe!rKNfx53h!`0^sTfrH!abr`qoFS`jR_z4q?Z8N;hS=92a}%IE zh-IJRoQe~Pdb)W*@}Dy$*Vw6pmAzs7n<+y0X?{0Kg#L@0?>1m%+O+KEzVI)lu4&aU z<7$tQyE->9Si#Zvl4GkWs#kWM;%zjOgDBr#T^LxV+le^lckP3mbT(M)6bJW8A35s# z-mDwAWYW~GHXd%gSO(n@`JS9_?DeAtmKmghzL;{-or>scv+4h!e2}!csFHdqzteZH zxM_B`C`4vK1lLk1&*DE9nVaLCnDsK9AdUV`+mbP!ZzYo~6*cRWfH>2K*gO7iNK#kb z`P$Auu{P%K@7-5)WPgc^(v0)joL30+vT9H7k|)1pQkg4ME2Y+|`OX|1#de;y zy_}Uw#%SLds~7p$0CY`m!ak+Ci^>HQ0Am7XU^p3#g&?e*py3yJ7ZQTn#Rz|5p5ieWV4(xWZwV6LNMKgRRiXc%2svyjP;EkyVcJo0~2= z8{vJL4?h6|@Jm8}IX`42)O!uNkJOlG1^lLZd3dgiq4k1E*3&k=98TE=_@ID&q4f3| zZWz681WLbuRiCdpj=ABTNj1%HJ21xC4uCfYSFO{s{s}<>`!SD`1AVCrrpxyyoojZE zOuI5v7ejN_^e-`6)7GYSrCtH=mxqxNx15+-=q>S%H8gL^^%pCiDvP2M1d_7q7#<(7 zwBH)OGoa|+&^0l|v4cEFtun z?3heTN%8O);bkgMAzuk2;AF?8ZdTg!r5->~@S7z~oO*G2JA984Ns;7*32uv(8)SPi z6V~#OSb8qqVjK>9Ogv=`7s3=a}VM{Y++YD>-_BM)A4GRaj`$2Qs2MQebDf&&;50 zXpN}$AYqfk>kdKG8`Lyy%yb?xceGwQn{4j8UN5LR^mL)j+x|15CyDK_(=Er;lutS+ zlU24_;HA$Gnu*Zq+@-D$<2+_$RS5H+5WQUV{>1gmG2bSgHGi$Gzj94#rb^sdz8z?5 zx{LE|N4rk2S&A}rG3NNhS?xbZ`L5LjNmdIDE;aR>zFDen&$|s&w%7MxeeU%&l&j&zrc5tO+d;W;4@;sGgNFR8Cq9Bz0TKktn5Ac>-t zVM_Q9AB#c`F8cN*wce@mF%)zOSHUmY++56`_#IaQqw1R=3~LX;c0}uiM6mw;mgo34 zR2j#X4S3GS&)g06+PSvL7}kJ^IQ@g~^7e3B?sGO{7~^WH@tUU90tZxfAUWe`cd_aD zmuA|$1DjTC`Jkp-(_-hrYC7q8L>}#gS1HWM==xxzoD)1oD~9VEE&s_@>$)CY0VLaRbp-3?f;%g}2lgoDHZV z(xpg|EmXTNGU{Mx$`};>j6UK@=c;^c{>0$PM?HUlv$*yiSO|ET$CNMT0VAG+UN|q) zx8>m+Cy5#Zqc8kwk7NK`%L~&Im+D;br_-ef-qEy(ako2amrSIwvNT+N=6k}^xqdP9 zCfE36k~FDHfotTfIdu%T$_Q27YVbC@?N}bC{erw^GD!PmSLb2LG%K|Nr)3u{#7Tr* z{YIHh>q}~8Ixj(m8r(%^*#LalYXR#X30|a$4A>eFkLmU4`A|J>8}lpuv@GE-3^m4G z2v`shz(}?5THs%uC(-J@Gsf^u(>kTF+G7EV`|}S1)gNM15SZv>e|r`JNlhAP?JG~! z^`p>2?FesFF?rLtr$|6qXUa%XODj#^!Xf zd)+5go?)s}ISW2KE={_34>(R#f$H8Hdc>fFw_}pYV=RTf$ms)~!I&vuG}M-Eyn{$N z8>oAx{V@TZOiO<&D|L=B)&H#a?KC;K(vbGtrqq{Czqibf{6kT$M_aUqQvA*=Ik=?K z&zJE^8b=v82eJ)m`6v*?CV`AB+k*Lwt3j7yR=!n^_KJv^6m}|HP~mz$#2=1jl{5)%KA5YGL z>Jtc(M~OnIq{K4W>R0Afq9%50noRUk;-6MSq~hz^A9k7Ko2`V?2C>3LZ%(|5&^~cT zAZ7p3X8qoq85AQf${uo`!aR%flo&gCc=9l!q3ANzG~~OLbdfoFZ`4(XS8+HgLZSDW zCH7g96oRiUMZo6Bj?Xu@{4pdS{|zP+I9h0Rz(D-{``L=cQAOc2ng?C)5S)1B-HwWj zmLMU;U#2>Ux6VOlq3Q7Fdc#JQrsSKX(1qOIMwVmw4|OHHZH)qggUE2)4bxhOvqj@& zHF`QFz1ADeix33&9zD$W3hymkG&%OwCgJ5)-fX7RU`>!#@NzvYA{bw2g@Sd5NY-vt#b}T9soBu{-aPdWzOG5C!CW@DJAZ$-JZ*VET&&)a$z<$s_VB9T)_k!_Uj^P=YN0`XaZY5nm#upv)m)kue#qJZ3e4>H;zZy%nCbtuatf{Opc2A>Q z&^nIAYe@prlli?6xpC`PIOVK&k~$1eIMIfU=}JAvjtiFdhJu4Dchc>R2bk~T)vZ{( z3U|8b@P^I4LRb80Qi1H4hdx@yOMkxEsmtYo@37Q<5>N!sB(sjF9fAunLq1tQ1PPo@ z3_ewD00l|mUL5AYHv&*daBi5jurY{w0{HX}rc8bRd5<75R>-@K^wXqt0F22`L2Rf1 z-|9*5Xx@ug9ehfpQ*&~FluYEq?$VOLD%v4FADJIl1XO5&72iu%FJ82=_QUWu=Qs{H z;lno^i@=YI#=nAO)1(qCCLaLA_sj-!Sp-g!2Z-gJf`*V5VdOx=TUHBP^!Zc%s`-ef z&wrVKwGK-MJV-bDr(?;XeP^3e|8-ZJ^ONvW3Wr}i4bpQb*59Qla3M;8Q8q)|dU*L` z>~&wKHl23iZbnaPkFOzTKa0q+a&py~I`kY(m+0wb)9xdOupv<)wvn0erg0=}P^3Ci z+KG#?(RQMCIhxYgo*ziK(@9~|a^lh_*~Z_hFYgQTzgcwAv>`ds)3stYCw1x0@}n{G z<)&vv%>+mSPEOFOv*--iMPXl$S6t3~|6nxaizR`*h+|hwaVy2V`z}bUH-mLooi?}K z)0Ol=;#kC8X0*t7+tbj~;w~KfeK~t;*L~)OUlO-!VB%|3+#*m^W?#*G6@StuM||#> z;zJ$Ft)|M;p5C1u+u>P%Q8W?n$YQO7?5}_?>Q>?N$NR$fR%Q&W`d719`d%fwEI%xt zZE(r;s?D$|5?r3=TkcxftDo(ab`d!crf5af9o@txX`Kx6YlrvegQu6FK7>y5nC0I+ za}UfS>YS>Jhy1Gf0v3(2OMkKhUG51W)J+99g>Tfq>HTJ(gP3@Y@H%%;ardzvUizVi zkMw&%`p7Iu`U166h(*GadyBtR=;^N|WqcptSxbJll-Kkml}kXo5$V&nLm&j`zVyQMit1kuiG=huKT_b#oYu-EjxOvl}KZ+&nP4pSvt zk=qY?L;E|ncH0IHLCQtm5a<;b@%=vS!IsUql-lg_waWhm{vJJ08msxk&3gAl)u^sy1;c2-zMz4+LI(+zX1p9lvH}<#c+jZTGnS==Ljb zL%NCC7YY9o4#H~wYGdHL$cTyGWt6`oOz#rmq^Ag_l( zJ&Y<2oCv-G+&tOJ=1{-V^1EMtBh=|%D_XE-=+oqGBRK@|aZy4w?Rb6103Ob9e815WK<(rysb)3QmsYL*mm06xFX zJy4V}jQP(K!0_&~1I&8MgFLv}u;rkRJ-;$&jwgkgZ&(=lQ&1E#;>|+~0Vc{%`wVL< z*NIwATuXd%j$BS(>(-^?chJ9EZ@&06F)w%j83}!E<|^QQ^&zVhcIWlc2T^zMUl4yF{?J=>cxL4(K?wBxmGfwn!0W@T(NT9Bv z#A`&$x?NZpV&9Hs9HMzn+eJyyU>(N5q)+AEp ztyr%bZa*(7K-+5hb~m4Amf>e(_hr-zLb7FtV`*8xLe!#@X{T!Ju@FQk;7Ug}a*W7s zb<)yEW$|52BLOBfUGT~ce)QeIE8w3_6EPRE5Dq;uJ-~ojoY25?(f1Rd$tea>I!Iqp zq^tQUWUH+!Q=r^zkKAijmGZ^Tj6v<tHM{(Ei9PPRn<9UsaB+f zaIfPsgSP3)VP%ZZaM~%c!0Yy@5uICUrakG^krJk|Y0iiqNAHr$le6y?9=jqm=ZOaA&@jb54HxUJTME zZo+y(FEt9HyZefcgYI-9Y3WfU_FaJV{n$@Mx5CIEi-}c`b*Z5Ld4ab4i*481`la|{ zOF7Ivy9;cyfNGipe!m6atj3IDouG-r&s$7FR+PE>p{f7Q1^DRrAc9qUjdZM%gFHWB zp--+h;H9-=9|QyHL{xTk=^MWtSzGYA8q3!5 z)+lg|)9V2r05obq&w#DVx3wcKT|F6pc~1IM9Hh;iJ3U@#E7g1t^!km4Gsr*lBQNI{ z9q!OmK-6xlU)@iYlkuEh*qW+K&~Sp4v!FIHphti&(ZvIB4-n&>?Z9yHwzl$AJmHMx zi*8w4K4GQtwF@BT1jn-TAdG^ZMS0eHsEAr$z?p4;lS1aeyyf~ap~;?0pTfF#OcnbW zUDZoAiSK5CrIZWzURBi~Oi-sm@w>!s!r+m{@dd~sF9zfLV2NtX|097%G+-l^6)#iV zMExovdRNU_CH>*bCB?cp`;X1YYsRA{Kui>=rh=j&)V_oZ=jH;Kc>y>B(XN8z7 z4cgjoHvdHIiCEF=B!-N<-zpQS3%^0~qGuJOUl89A;&&bjvdJu>`;F#{?c$tvF^wcQ zL{-vr^4(%9a56UVB#Qhq-GRa@%jnAfIshEWoK3sM@2&qZGhP}zTB6_iHQ zKs+ReWByV2EGBHcxst6AkzEW!KCU_xZ0arFoj5Fv^d1N6RJkc{Rww=IcpX=4sNp*R zd&l|^+aFtUj!egeU+Jjyo6~4$>IM%O9AZ_T3h>$Yt7zjGtEa*sQ+c18XKauZ;NEy) zf-Uj9`nZ@{Uf8e;jDMArI5F>Zw!1Xshly^&=x|0fEEXkbRofzphcykkXhg}vFPL48 z;2MR%P=}bce70gv?aJRhJjs!yUdb&VM0A0gfflGJ@XXMudr-B(_C|fcR%r|V(%z?# zBL3k5JpmtbXPoiI(NtAt`R)d9??8t>df`x>!!K2A85sx8IDy} zl;iq38Cv`9^@X&4$`D-EadOG{?70;b^js~?AxU`A4x*b6A)fp@cc>)-{pqw^ne^rK zKXlJ5X3f%mfw|$?e{^A73dq{M%!RsSZSu8ql~>#lAR3|dR==<_iBS~C0%y@QuEu== z6w-`Jz=zDLjE;75e`VlTSQ|6k!l_LXn^9f&Iotx6e1)sF;l)|@gPwa_&7r===S)ce zNy=uvfx>0L4LG-J)KgC9Z%=qITwg2iN-sD?)fwaF6waF19H=7+`Z@a8?cOet$?RFQ zy(1^1dA>%N@50y^1m% zd(PtL^+-Nr7o2pnLn-gD^nQR_8VRVMp`0t+EkJ>v3eV^PZ!+r2rBDR62c>C=)Sx&M zlmm(RLN#0Kc*-gwV>@n#|SL>c_#UcVJZa+&E+S*3ANhbbUKJ4)$w zJa%*O7OWgRtIt=|d0pNpiiH80l;K9!$^EMv;8sP!yHOTzgIsQ(dfazOSLV>AUI^M2 z8IA=9T#5T;3nEUu4)eNYj%8g1tZpSo znxjwrc3ej&>tWsLvad$ERV&qoUhEYTbz4_dml|D)^c2*R8EG;HfB#Ff+m|S%0}$eN zvEz}RL+f{Q8?NVCVqOijZlJsn#Tq>8rn!nSW(hT)^Z!TQ1-ZsC)_KY?ezM*|o zd}gr{lVp0?JzBYpe)BMSI+L#`sD({4O}jclj6v&gNzglwy))yq!!}w&@~$Nl$2$c4 ze4v2eTIT_|vwSxFU-qYpaZ|-?pE}eY9kRVtzaQpnH=45fWrN@}I8wWbK%L+el%;hB z2GgXv?J^>+7$5>5ekUS=U1O$Udhf;#{@d+i_u}7@zvj8`p2L4oz>)5*kY&QnK6w1vlcbb25>k(@2Z>* zhjYM4fu0&3$#JxUM=^<9RpSJiGlmTKb(cn`fJzCy zcDW-D&W}xeo$0_C{@gi$t@lB|`t>pr?>L}AGXNa_NrC=4E1B-=BaimNC$f#g3DQ$C zTZQz9X@Yno;CdNA5}@vugaI&AntcFrZ85IZ*@)CHoy`G~$d^!tv!s%i=sMsUy0s7n zSE|V+-AWLQJ!IvU^;V|as;jjq;hlrounHF@HlM<^meqnlWj3J3jyR2zx2BAM7P{{ ztZILBuQ9UfaZZ1Z>V<@Hwuyd-kwC~l1@`@tf&^!KR7j)3P-cS{hv9r3N7NR(nIP^) zkvssjv!qpYUaE>f8NQ$G+wccVhEvAzrnxUg3ng zqh2JG?|Mca$iX9JO^;3n-}lQUj4a7{ih@=3zGe+B=NVBAZl59K;#14 z%;9Jeav@9lDB}ruoj*x$(dfW4x4P9#;D%CL-*NA4T8$^@gG0xG=I6~ zx?E}!wgpXZN!9c>^gd&|vQpmp>sHv6sL&5`>!}@Osms=N4-Pgu1|l}nj^5b{WbADZ zAa4DyKEU&T`v6zhFIW7h50Jwk4X8b=1@xD2WCJ1ejjA3Gm}%%K;5~qgG+F$f^hKa0 z*K=x48{md%8L$9=O9#lT80Q0I044x==UB#T_fVZ&Rd8NKx+Klw^^uSTx+oNkt9ejk zbPUAK&@u>IL@At}y?QD&=aFrp{0RF3|0%0(L>A5n==9<&1TDtP0-!h++{4&ZE-Ym0 zyXD0?JLBv)`ZEB(iJl2pRc-Ag>-DuW=Hf1|<%=h8JS)x*{kaD_oBlLj<^**y`A+` zg+WggAMTu;VHL*@)a@Gw!3qG5OkrNYmb(#^(H&<3PvAJ#PbuI2ImX3vuAfne>VsHV zukI0C*N8%|Jsja+K$+KfgZKJmFD=Z8qoUiMlP@~6(iZ3^E<*|d4kxO;DhaUh4L1zD zvufHDU{G~&A%<++gwmy!N}bdODaVkD6Y9kPguo96mqMFfmlV>s6yMcyHP`z5dOyTB z)C5(<1|<9U{32eNG{2vTTNs(AJIlsiuizMyNvV+XHXc!{EZr{Gd+Wpp-8UqAzlaUF)|Z8NsMI{v1&| zZAq$Tf}!QRAqkr#XE3*on8k1+I7o#zGLMdJTwEm7Ofz{5Z9ua4TzbugwE^4{oDS~1 zy)+yz7g-8oQEGa%PI>MvtR33z_`|o?h*IVhRNysuwPTYa5J`3tQ9fW^EZ4-j6_pfB zmGjtm8|8vZG(YH`t_&5ZR851%UnYTEUj=@szUF}sbYGS;JjwWBX;xW_ZkOMl@{I(QZW&QU^w~@5Nw;V{JUbDv(D292tZIg)YW4R(mY$1}Oghc}Wrp%4CCC^K%o^ zO*>M3CfmJ(Kgc2riLvMSbmQiK%N;b$=h&?QKU@sFC3heaz3TOLcGO`;+qGv6C6FAE zRqchZ)uKvjtJ{tS&1`m>oSKG_Hk#vnJj*c@K6^{=Qt1B5%&3;i>^6X5I?~JVe_4fZ z{c9C|Q|I;JKUU$}9Js_f_Bz04Rm!`=jb;N;Y`X}D^#kE6SwKp)4XY5NKJR9%DBu|> zo$(F5KSMtdwPLHyaRXI`0&vw+8O0t!_(MG*BplD}o^r-(C=>f7X}4GtaEfd5m6fO5 zfTa31aroA|@}ddUKh95Wi-2yaAR^l;;qPpH~J z3D+4B44kGY{gSjrvopv72T*E*^oj1bw$#3HWIC(6(lv`^Y+DXe3}J0|SzKsFOMl@n zTeOK@tQum~Nq@Za&2|pT2do2Li9I~E#ZH5%jOTvz>9^W+65R%26YnWZql&M~&>z0#+;f9^ViF(~)}C)F**^KS8M z1D)cXjP*y*D2B#NdV3dpx1Nk%;}}w$dqVxhBYMICNqL-zd@pd8jO;n1_-HO|>kLX3 zg}Y=I6ec1Hnk%0F1spn;8^_-f#jwe5w6BH=tt4TVAU*G*vu(+sQxiIAl@Qz(_YaXa~JmxHWlx^^{vYx#WtX0 z>=?liGyi_~+{>g9%3@5>!17s7PDFvvJ4r5dLa29~!{Z9=1bKuz|NiT@9`_T}wDy{d zC`8u`2mXq*hv!2d&af{m588EyRK-?mN{cq6$U=Bc!jEFAu8!}?zT}?}UlxlPC>|Vs z`9*tpD5M*%!b@mL_(Yg5>H^CJVWPO}O7(QDZfG_Wsu)BjcdjSO_x6P#O<}wCL8>ti zMLmT?l{>l*(;7o_cXX=0Jv@t^40y|SxlT0qc_(PUUve#QNzZM-efT=U|Cw}}*Lk1u zriR%#76xW=^mDxNlOR|x+-|cov+oFgD1zHeZlx!6cS$) zK1B@6!gjEHAzJ6ff2;nQJvJsU-0X*^baa@b8y%}nd}-*)L#YXZm~TT*bsTisCH^o) zRxy5wD3y;eS_n1^|CD`K1dSMHR$tKAW}{s5yj3Ue6xzci5W4mZMs3fIf-h}#M+@9O zy-i2byR!(26j`Y@g1UV403JKiu7^A5Nj5nN$*+<`j2%lO-!^E9%*DNQxv8} zOqOCCvq1KA62LuFnHxDVIzcG2n*THdy!D%uZRL9w)4Rchtd>W~p?*lc1NgNa%G47} z#EW8mgICV>aq=K!ajXoX)WIKMjtnp4M{(j7+ zD8FxN;7ii!3*FeJwwFIE!#!xr89wvec@A&H{2e&vI@TmmrL*ABZ{4qklvwBkUTkjr zM1Qj-UTJ+ewpQd&bI_>l!V?DG_2z!;cS=3v)UALY02-I{tKZ3f>?* zb0D~`iDQ0eO#Qr>%l|VZCKJ{i+g2GcHY)%5rQ8(s<>dHttFgZO_q8{hck_#uCnEN} z9XXa-F5l_rZf%)=5kt;qGgjQaut&~8;!Ntly%6^yc1Es#9HY2N3N|@sZ);GyO6A3% zJ;$AjO}cTN*CYGBH$TLu-yb#w9j5cw*^t|G?hHTOJTy?47_Vl?Kbm~dd?}ZMPl+v7 znwf=1c4yR5`(Ug}lhP3*CPn&CKwIwscDyzc?r&B6F$&Z~nu;Ua%+iXYN4Fnaz$BDQ zrtRF?7Uw@_BhAt5Q;#b?PG~XN>u9tXN_zgK`)JpWwBAQFVGwdxT;ra2W|uMKpJfI0 z(KnH%8;^!X)rRUh?Zx$MN?2jxvN@8+M;_AU0l%(&-E#Qa@nkG}4#iXT5Zzf#U_5>7YMU5u z*>;y=T=#{^TVIO6V!<`-X!2My5IFHDBP`g&I}4KeJz@;_4CQ|^_p95J7n(sCM*mu00)8f8Lm((Ik9QzBoPL zC^0TZ-N`1?SJHY_C@Q0A@!)N=%d}>J{^x`(#X(HG5M;#!6YJ_r2~w-7iQt7)iY;(OfsZ0 z?WD`{2J_55D^bq6BIX*qkcq95Xf!jGopamU8(XC>bh(!Y=|bm-9CMV;RQ&-u71Ler z0?*&fC2LdD1W3_B%IJ8kSVe+@n4Gxv0Bg87rw{)Gd5Hhj>Gq$f?h6Vwb+>cZ3BnHf zrRl2ki4sy{ns8nB@!K2GMowTqI|B4>Bq6b~|57wGd@j}pT&%Hxz3qTB-7=AFfCQcC zZX)#8>ggSz9NFGa!Qzzu297)3<$mf0p92UV{SLI(SX#PAxTAhL+E?kQz$V7Ij`x!s zk=3lNeVR|FYxs74^R3D3IBTq3+vzR;g54xPKdZjIK;~Pgym9gw?TI_>6TQ(@{#vwQ z=k_iP-z!8NP&H9*%$U>T@TCS7p6|##=4`%f9{}z5r8!3Y$tmy|A4u@iEVa>gD=mCo zd29-qCP=l`ejjUjGnT|Rgfw>SOLjtH3uUJgc+Bk84GR* -sfe^tyb`ScPEJ+dc5 zf-|j6wN$J2ZPb|M!M0Mu7Sww0>%@_o6aH-7Fc*8A+aScEgYeQ$T^h?h9Um?FNB|v) zg=Wd099nF^`mToxzHBs{ss4DF6`9^|CMZ2u<;YU;HWyDO9@2iae6oe62)h)5G4)gp?i9iyr*Id^c$!doLB>ueV-OpAuwt#)5m0M5hDlH zmg?fc))rss^H?cgbKMe2mvBaOX3(M+J*A!J#_f#h^HUiW8B+lm%{>b+{I0<*W-OE9 zuMC$k5%?V!ujUw$+3X{(A6zqT+SzEl$B$469~ZunM_xTF({!0+x>i%_S3SV>CAupF zQ6!TPNFNv?eiPV2A%#Ss^x&5@fR9IBGRRufGl%tw=PFJ<&|di=tWLgkWr<&?v*O3A zCUQwi{zpLTn&%t8rZvJ4cpaaqSsK+$PN=xRPuEuO=dJoPLo^IP~f4S)kuZBxp^|*18lPXaY?s)!=Tr~ZtO0H^8g6kQL z$YXMcY}C6wO$LQ&-Z9*%=BEuOPjl3xM+LI2nW?@Cl9vt z>7V7pb~B@-P9|T9Ug!-?jg>tRBg-?{59r-kY&_YYtQvOpeL;KiL77SG82W^~hdRf1 z17v7O`YN~8FF5SFA zc!yQm@sT_Yke$ul=Y#~$owmjK&tsyP{Kk#tl?78RD!VNLTE)0(q!RKt8|RI(WH6}$ zZ;;4S8@k|cD!p=np_aM@7Rf9yhZGsf!D@xbor31QVxMzREW}HA-;LX=d)P20`grib zQe(Yt<14G2camQ8l1}2G0rkVs|10vP&HkMcyleWID*K-)AsJwbwh*QZ$wlHB)E%ZH zeEo63zf($R9r0y8S&HjB6P6`!M?2^*Gjp^tT{oD%m~xZV1dEEK+NOHwum%Gq?*4O) z3kMKDB`A|eRdApyz<+A}89Ck*ZeHyMxxb~4H)!grl>7v&qXM4+uALDs&n@HEzP%oUo2|crw`ts`+(oBbw%RTr@ReP>R6S(%1gW1+SC6B>ZFTU6n zTg^R^TJ|=&cHG6sd#G6Hk|5*+paE&bl-V{e7G_QC^p{6*%0iW}Bf3ej|2O=)KH!uN zAFH-zFW_3MFe{Lyk_pL_`i=64cJ9dxIN5xAO+5BZ4q!G2cN3#mUIX){z{2Wx@*6rV(AdAhW4%rU)byKh` zPJ?H!M+`MPxg!bY?uZHDyYHvNG0oeP&3q2VJjphK7vU)miP0$$^9+rIO6?#_{R$bP zJ&(3?Q{jm2cS4)E@E2a`=D*#ZiylIh`voz!)X2+Qpe=uNmXFn&7GdoL473tL!=0_D z4@FnZ|9n?k__CO&l_iCo5cb>%fQlTsJGxX+Z#oIwDAQAhM|%^R7d4^I_TB!NPW;yI zcJOYF!%Fa1K{57$n2nEzpu%csV@_{StG?5O8GmDLv0t4kk!J>(8@%&^z8=nX>Ia1p zmSQDLH0#U17?Gq9IyFDFE?l1C+mU+u>d=UWD_lDZKYUiGLGVbIKgD?_#2}yE!ME9p zkq?1d#o5r;#-?cAiK*4!Idb!SJ`-CC$Wp2Oj=|-uwX}~A&6Qcd8-bQ|g`3%|4(OQV z&Z5Hn4Jg4!v#z2XikVSIK@L)>r3wpJuYc`Gr)ALHs$(X{a_`OOTqoZ=+Ggn!|okbUGo$=c)_y2`)8JoB)loBS+eE2;Nvg4AQ!YbQONu2o|nSt5FnOEcEo#-UCHg1yYw zFDh@q#^U5Qvc_lAH8#T68OHwSO;4Ks-*D$;{?sV_-{Jm8kFMps_VoE5nr+><>6eT!Ilm^B@*mZlxf z_fYtFnM=9Tcgys}xs%~AQ&Ogvx16#bp2UU_b7~zIREu)p)}rqm9g|h`Z9Umv&ySb0 zE)Z~+e5StkEPbsE_`|ec&FTPPb;fDxf)X}V68@3^VD+=P(`7Jk?tTtO=IrCk3*4v}K1>W*|~ zMLs?6>zl@bWTQZXGf$47!_lu#S?!V4U+-g`cIjvOpCT;!+Y9uc5x+`)N}=h-*>xR= zrx50ZSdo*A9;Y3-dVyBxuT#=M{bwXs{%W16PVLLRlVHJ`(8PqwsJ5QwM+wD;USJ9# z7SzXqptH@k;b?EfNB1~e$QVZahW z^Vtyi*6FHG!E-IZsdxo zh+A!AVMxz75XNn(*{Vi7jK!j^R8MO+(7Q|3(~{%yTie;6qN2$8!l|jIw7mn?!BP{Q zgy?M}56|IWN56&Lw@l80EHK?x0gzIe{O!IW;&b&=hzK7xh}xZ^FCE$34oB2TrFf4& zQ|=y!vk2GE4ua(nUAr-Z<)E3VGaRrHM#|K?7j)9Zt&IOtjLn@I_Z1ZBOyTLRCoEt` zie+_!$UcQXF%q%F7f=$(IRi1GCGRBRRFiUI5WJTDAqZ-zGEwWd)L$O_eV>wjXf*_G zR01QfQFGZ9Kj-BVF(A4BRlq<0FVKJE;QF6!fgjK)#s^$UTrc#lcZOk8pJF%Lgv94k zb|>s#-xIs1NDY=fp^hdzea(1PNbeZMfnpsiRKXiubJhXk-j_q>oNozrv>1$1{*V4)BQMnt~jbaXDkq`Y9)xe-ejct zTMk$Xg>ySLmvI>DyTcvVTMW%LkVFAyp6_3gWpfDzM_0@!%7yZ=c5o1uTL4f%ufJ>9 zi1JOAKR6?|w>_lsjVnAKhVn=0Mvg+FM2$}9@_S#T0WXZR`#-T%o<;2b0^u$tmT4Uq zLs57Ke(z^KIGHLmBayyI^ChJ7dE0HLES@a}ZTkmKq10;EvnGPyetS2{9l#k}0`gPa z^6(ID0xVWT>*g|AvO$i$Y))>k{g4fU1^@IYo#kGuX-^qzepgk*5pf*++LQG;Y`c?l z(vrdX$K#5R^*pJTEmUci`?(A*_SGi!-^#}W&rROWaO#;|_znoG2d9!G=NbM+xoL1p z0qs>QalI>y_&L&M!Ff^Qpifj|t>$PoThbOL>vk@)M{5-e~<*#*M666s*A_IzU%#z#vksVK0`5Bqz1b2&lZk_ z1?%ij6j=3+X$r*}cDuBG4W$%+bny0`I?GouN*op>Iono)7H}~^DXOyVJ-+=h-h`HT z{oKtJfk<-k`e_GPv9U&j;KeWQb!j1j7&k)^eKa@qzM^ts(W2fp{kQ#LTXvKvRcXnD zzxL67EM#XSq2Y%XxN+McdhvRFx7@h-24`UhB)_lYOt((-W^1y1-;7RCFjh-l{M|Bv zKgR(@P8-~_fRa}ETbEqPVSaEy>P=A9xqE%tvI^Sj3BCs$-aShAg=}xi@9fzv?E}aJ zAvZGk_Y9D2?W%!*LEydS-;*n;p3%VA*x7otZyBI_wfL#cu8=*g$NCze$qk{U)N_^GF8 z_Wc5u!rBIlV56DfRp*XinXG8i66y20Q^hZWsGHvB`)nO0Gx3hmSSJs z8|=1yE^%w}$*9YZVyh*}vZnvVyk!-Fo;~~>W0gGJ*N3qM_%zTJ9cbGvCwPunG3p=q z73a9Bc zSI7GMKPuk01-6Oh%2%m3i809hlAhobX_@#jND)f}p3E!FO=~3|RtXfY z%qksRTFaW5?!bv^ETrj$1+T6%Z^ZP+EJ~~Y!ki8OO~TFgANkD$XNaghJWnfqd4i+g z34#6r^54jEQ-qjau5s8Canh_-s}T~VjKr-D7LT-hyjpVIz=$a{wR|MtQQgS{6KV{!}I_&Npyi#yK^*dCLKAJPi#Xc#fn~?6U zoc(iHhRTi1Ar8(Y(R zFS#_{4ohtmKQhM57T-Rj5!3=}Qq0#RB%7nyD!qaw1{bthbZ+qOE7;ugaMhrS^96FI z^mCF7-52$D1ST53yup7RJQ6?hD$Q&<#5&cxu>JWgusN}lbkP+jkU*uFouOMc_>R6D zyVd=lQ@bx%S?tc=I|SNY{LAj4DF*OF1H@pH`%Us?A*qy@bs0qgeAA+h)*PzC<(C>D za~BSttB^+ta20ZuE`Jvp2?r|t_&ke}L;}jBGq>kPVkOp7c?q1>kAGJ{#!y$`67WpP zRq<8n{4^PPl7_W35|~)ml`G#1Voyle7UVIMihM=CQ#>3%88-Ud)hIaz3_!tIWZ@Qv zkG~dnLn#?G{m(5G(dBgJ)@476@VmTC^s1RMnX%0%cQ+8tIc8-f%_^aMyHBtebY-tq zBvzP;+l8pIGB7@l(vomc12h8QO@9F8Xt%aYa8MyHB$Mtw;~H>{ShanaQeFyt)6EqA zp7qs3$eNru?BClQ^<(kCp41L-9|HsdNc;ERO|AS5y{Z6$u5vit2DMI)K?}eeT`CB5 zEk~GQNEuu;$r*qwKn}@eiG`r4%zI|tYHH&ANW5*3f6eH@2so6Km{zV`-+yP^ho&C| zTMk>xfP@pCoJ|zUFag>fJjPI=@-Lxo{E9dvW?K*cMlOjS6AgJ33MTp@@aNVPB~>@) zW@$zu#;H0P@(TqnPa4e6@RAbYi~(*BX0Dfo$0^ea;`)^)N^b9bJ6u$1^4gM>kAot2 zZCLoa)N0@zD-~5UC?|@^Gm|UtR~y~XuM&&5FGM_H zja~EfaKV0u*zzP^u#g(lP7!*Mvbn~9ocKZSph36>j^bfSgAG6{f!m+9V~ApyHBX~B zy;gFqJ?9YBU$6y^oE@)r2<1HCSy89OIGl<^B*Bzjz|LNw)e)0V&Vr<`_6ZW1@&{tQ zGi4exT=d+W`GY`_KLf0~yFdL-p`kGgdNJNN1E!BUL4kV?XWncB-R1B#zf;qFT5niQm>kOJB(MDE$c!ZNN5%G0 zx^i+Oc)v6}F|jbOd{`&rYG}=m3sJ@Pf%+P!Q*k^6AB|s^3~0UCJ}MI{PDr0x!j9*e zuH9}u_p)S6^2F-1XB>EcO;}tp^Ig2P_cuRv>s#89A+3t$%#D&Z+=aoON))%EBWkX#hi5x&&K~Iqv`iMrwjP!e^Na270bqN^NkpP z6jFXH?rV1O=G%`33_4r-Y7vQVk{Q@>rgir|CTU$>@TGQg&n>vOJRcnLY(3V#8#0)E#h-nwlXE1tNPyVz&YE{2$VgRdyAFX{g!mq=MO-wCn_`j zOz7o`8}Rel^D~~dlGiV*J^`OFo@2Q_Ux=uDX9e@SRnhT!(tdvCbl=4Ep*abtM}8Ya z4N!VCR9bO0>|=xzYhmmhTqk%zp3rW%VPqr3j~^Bf1mtI1nH25-e{C05vIx;ab>2axF^fX zybI~k-XxzWBUN#7h}FDTV}$g2T6C$2 zJfgjjH`kK$Y^T(Q9nP-XO1a$@duDAJ^JjPG$tsQYlxOw+nB@W*Ac=p(owxSz(n*Jt zR2?M+@>|m3vxo^&Wy-%}{-;#m+S~J0OA&$7$7Sh#ana2Cc8?5|PZFH*>@q93-_VyM zoa!$szu>KQ<_ospM^$EhtVh7IX1UFGOOiC{+&!^Y!Ld_)Y{+hrAuG3`+=^qlvmLqI zSIrcc`xlsP3^gW%Bp0A9#b!@_b_)Wz{w(Lur2vHPr3R78Wf!hM<%s&ZX&$hf6@tOc zk1v~jcs|DDuz%HF&U9FZb~Cqr8K&hj*)1^k(}bHKRi!Tx{=T^-nxLmC-7q83$`(_% z(2CebZA6rP)zi7&{8Pf-mDSVQQhRg&O!#4U$A0}!N7cVaro7fvei`Y-x(1_V@YqSi z7>d>O1Yx87mmjQW1wnajv{R}Oqs#q4^TlTrCK6G`BiCn{d6IEs{Jhdj)8erzoz35*8K{N^h<%4aztX>3h(pKjm!>RMwG z)6~s%7(z8E?<3S!M6k>TDh^-uky;2M-bthOYqlCzPMGZ*eIGpXcUHa5&jUUWs#F0< z2!*$(sqe9Yp3!&2{%+%YMh_6k*2vc8jC9WLd&5#V5E5Z7Az(IWVLA>1 z*J@es-#B7?OqWf=^tCgsF!Mk993T`0;e;#}a{o0sRzj*7q*bLz_Qb z4UE%?6qxIM;p6>n<^&;SA7XL3_NzfGrcaAHCXtc0UIS8|8+o!)ws7R?%B+F@hJj>_ zeWFY?%%u*2-rK^Y=ql_{=jUXvmki#?t&M5F7)lCC?FNLLE5R87xsQG>GzulpW!7)$ zXmp?vUI&vkWk{HTD6+_gAe}ejZSoUeO zC2c2U76z)vO(ktN*hqD5{UpAL*{4Bg-Dnr{wvH?`FQlR=3$JZ%d|2Nn6QD51Iffy| zg%2|l+*Dydb)6Z6h*pqy$1Kh5C zhU{>_L8|DTeS)wLD~^Yid^v)Tp3E>^qn?V-vaj*{ROZ2hMEB-fFvCl^(}Fw4AfsOA zCntH-m*1fj^8bCARg%5?=-OLP^G6?opjuX%+nwhOS@@~1Ws9&EHn%&}1U`>(MUtzx z)9?$Dh3Euu&oMa!(zL#3_rW$R1iZ2l zM{2-O%6wdvqbF#V@Rm*e+V+F-m$%DlQbom|P*_^9(}|W}Chs^=qUAaL;^W|ynatQA zM2A9PaYC?UREd=vJT%LF0F5f`i-%0 zU!=2l{mqSn+TU&e75J)-U~~9Z_>smZ$bSWWP*3bh{75y6RW?s{QE4C2r0AsDD@Wz3 z^37U-0nypJ7Gz~%#dbxV^Le^eEU&$Gh@>q3g{S!Z`}zF^o{i69fLy5b!{vMy@44(| zeMISHEk|B$U`GJoFxVFLliVHNIrciuy z7q%>?ho1#6jiI;Z#B3UoNZif#))QR4!7&gdS&!8}rzeDVp3vd(aP*}) z@tvk}NGY$=n-=L`Pmdz>{87F*t9?^jvBu0nLI%atZ)n+#TAGeVjED40yH>3wYHhO& zE{k9WUl7k9zkfv?w$WIdm3QC@|4q-(_KNjeaSGHWs2eE9_0}fSNlSVEDU%Kd{0rcs zb&f@?fpWaE4@>eE2xYHaCQw&0&hu59x0KlA#jUJOf??*wLF<>}M!PTuP;F7j6Kj^Zux z!Ms|ya5^6cKaW9-IfQ*g2|4BYqFZ2PRXps+z;CBrk6|bX1s4?miXb9pBw{ zR29}x*O5z8U}P9J+DB*>l?*GbpZhBkA4a*bK|_c7m9Pa#s1_YjV}DIIlp@nImcEat z{;aC3Pjhe+ufo2Sw?KB^U*Ja)Bm_ECFg5$H0vFO$Dgos;JZmEsFC<3(tvGER=$M|& z1`mCaBAcr$+#Is1^vZ>O3cS;FnCeUWw8!7(yTygsvT|ZXk4kA5{75KbB&O^P_Kke8 z8m!F9dEWeUHUBs~RatH)7<2MYk39#u%l0YTshv|1t#qn!$1;kXY#Cx2ZYB%+d2Q|q z$Wgs5OjBg&j>x>i@eiwoI`!?V+VL5lpG7{Pu|uIT=zht%WvhUVNuyNo;?ABwI=V1`(wR?RL? z@QusV4-EY`54-~WO9OP~MH2r%3q;JL`p_mhENEA3T+#}`p;H` zmLwW?Mp)n+C-PT8=d)T?CB8YG?&NXqqeNH~g z#dZzcie?=DBH3V6L|VU0(7lTl#6fW$;{11ka=p*bN3BkWU@uMsis+{R{HI2J@3oy% zF9{8^_x`a7pn&|}!LSWO%L3l*9j#}_txmxNEx>Bj?D5Q}9c$eG$JV<)GX3}Q|2sOy zL{Uy#Jz%)T>ITw!-=u77KKpVqC z0`TQ>L4C-gveU6)ut_CIc#5u$6TSj`*4JZK%?^3nxSXHjzbMWolol0;lssElmWH*t z=TEHyrjn#{gO`>EScf=cBEi`BZl0KOyC>gM`W-{Ps?hKe1&Me4*u?saPk75p5{&Jf z44uGVI^`idP%rw5U)`*P9QpL=Z2mvgbxdecxscpOMH}4u{GD@;K5Oqn!G&5JpSFOj z&pnTxzOxVppxR#eAn*Ll94eYlo{=0g{CZGgye|ZC8FROg)!^ki`(2HJg*SeXR8ZVF zPpy?n?Qk33;#&@+cE}EC?^l#vFIgyUleC+UdVQnEs7-1z<<=|5DnZdL^_bo;S>`O; z3L3mI_6lZM%Z;6H-0TU*cP0;55&MqrdbWxjjL5)hW~6HZs}KqS4Lg;^=MT1GRkw!| zw;bc=Lfc0Drp{hu>d9b9sZQc=Q)o@r`bEnktizDsew77{t?~*D)|K28H<&qUpXO@Q zCFk$PDjf;|-d%8sul#3)-R(QSuXD-pNPT99pS-($&QzaNma`yioc&fHtTv3 z_wFlJXYW5Z(5IA&5-y6V#E5xopE@F-m$JjS#2vkTG;?XIHh>ScsRL*?_J$(6MJB1)O$Z%RJgnC7 zAl|d=otf*WuR$^n31%taP$A?PH*1rvK{|#`ItrI*0*EUqM?ycPa)CPlH9J)QttzQv zJwuZ+@SEG#O?yJ?pTHLVid(* z8E*cpztOfTvMB$7!$lhGZC-vy2L;q>0s~}9e&Vly3fd!zxD>f24WD4G`y9{$jEmY=;2y=B4w!B_;`v6y$zcbug8ur|{XzoTRk6 zwE>f>!70GdA?v5IEP*R(AD|0b-7J27| z6;x@Y6WP18>Vn6`M1Y6V!L*~~j%29%a12GVBw@1LFO*M5h;RmecWgzr;<+DvV92VZZgMC%o&43tZd0 z@1uii>zn~H7qck&?6bL3pkRzXIxa-akNlFqqm+_)VVFL*6}Ieh)4$LGJKAUDT&?5~ z+za!aeXd1+X}MGAIOiGbJ&I1re!<-A9-fm7B!`*P7ez-t7~5?YKWWzuq22HLl22R{ z+a=3yXN1Z7mPrJXCVS_RVt4Jr2<|u34$AKD{FWr<@_8O#PhyWRAgkS_6>5KQ%z}ZLB)>$pIglBoY8JS_Bm~>=XlaR7f^EBQiM>_V9gVfs%~rUQhvrl_ARcMvMJ( ziuE)8odq~dQKskbf+)~EKnE6_DoE1nmT8#G;LHZvna1r3*|%Z!)rB4hQC6opoVd{- z6M(GcD_|T!4GTu()vNvhDwzhJ;oK0yZ^~Q?(*o^Rri z+QD&}cid$P`E=*+y=vggwKEH%_P`Hz>fe9;z2#nP{$>BKj(4T7KIUiHlDM;pk^u8U zJxIw1yN5GGHiJ0y35Irch4L+d05{V|rz2>LzwuavE~~I{1kELtWViv-#@aaSR|;~q zySsANR{$T-z=RJ|)#H!50?X|h9sQ@D`>rPtS#y2Yo%ko3&5?%`GE+dQ70_|k-)n>f zSq)EVh~0I|PIy&u%m|D#<|&w#RIF%Oez@S#NTqWvkJQnqg{~#JJvTVoH!u^glkRjf zd*bY+kk(WOMY9qh`YbsPbvL%tW9(C;1Hj480YMe~W>;;|y80AX>lM7Qr0ojfm=LM} zzF)I)W}G{fez~}oF64XvrefO`cVkXttAuN`1&9x7z&qF8gXcz^@xDFdJYM}~qV%XM z=oIsgc=IW=chjM|>6`VdCriG2PlF}1Zq20R?TkM8p`dO6OM83VdkKj0U&!<;x&VM} zCW!Pwe>jDYmgr~JSXd+ zVGdv9#$iou%7j%e%8RNE9}cl@F{nCl{8e7$RCxhxzwGWRP&@qWkxAyLNS<)BoNKy& z9;)>MZabb=B0egBj`@D7qAT_u^AA*9p%NQ=h};p9Bn_rbzk1Z2ZxJ1V>7;u`S;*E% z`zW2{4FN~f-b7(st{6srb-#gcGW4yme>zh^W!<2OFc*KH7%7#?)nhi>cQ#NZBpR{_bM5)5Naq>UBZ}T*jK)nkV$mlr-gykHsD|Lf!9D z=cE0-Z*Weg61t|E2O+!y&~w*{jYjhN=LrD55e6+@&CxUjc%3BX^vXHmJ`;@sJd)`k ziBKUXX^XZx`Ue23$I?DRQ=|r&Zc98c#36XIW&4S?BL5|>7ZYGf(>K6tqpFca8dV&@ ziGJ2x1Q3*d(&@FC@Q>Rj#~*^JQ6_LebtmpIX-&rMTo}m)Ax1h}a&PeYQU1BhJC?iZ z4JCzCNKSbGT0UEe%p-7)n+%QN<^tVV5{kP;7We>D++OdON@34hZ7+82m`+vyz=s>IuX_H4wr<11*c+o>&j@tJiXp zE^&%VksSEbkir9bU*Wwz?khGnZ>+ttDB?hoB{EtRLLNcwKI^F$CZhTpoT=`IIKsAW zvfv@$&x|JrR@v_^=pDx`e5Kkgy{^6O=QS&5zY?G@)oLXQi#l%UzsYbhgKgG`L{@Zo z$Cq0;(3;$wI#M$T8=d}>QNm21zsRtl09ENFwW$BXKN6qtAhw$<$5uBMiLwbB&9VaQ zgUo1G$Q``bn5_$6*Y(v@@0hq4izqnU-ruOXy>mHsQ>b`DIFt4~>%;f0*V-YS`3lT= z(FvXMY)Q#4me;NjH+~&^Z=c>k5wlm5v1@v^Ku48Zv8=7PuM7`=_WJZfQD3zEkI2!)jiZdNfk<07+ zoWMB#U(q~Gy71Tu6xP;0{xR@hGc3!Wn)-&4+WzKJ`k}P()X!iclP4xQbC0IPwve%4 z^o8liKgZgT`=`G+3BHpTl1;z_U}EfV4*dfzARtZ_acINmeozAZ^d0Ke7{GC~BLPB* z%mUb$ekq&-`~b{D?m{d9Z)qGKPN4)2wd=-yUIwrj;>Q4j0=%q^-x?AL6;M{`ne}l< zNMDx^;w$t?+)*UwgAyZFx!=@eekI6pL^d02ZSV_oLxOZ&rzUK_)7Ddm|6rDzexd|X zTm_gNh0?$ag|)d>eShzYQ~OH=2}#}Lgzx7Sm$vXE!E9}SfICkHrz&V;`+EA~Ae>Z! z00Yd(fc>yz!Mxg__kAsb+E8Z2@-;y7RleYj0QvoyV!E;pq_zM>fD0iizl%79rGw=kNX&C;@v2Dszn0yFj`u=-Xw07EjK?k^puC8+gsM93cbiV0y z2n-d^kgT&zSy?XosK&TtWqE^RopCo<%KU)68?wIp%U^eQ@#IKcLa^!(S%zQ;^RSD1 zccH8N*@aid+p9(~tZK?DmW%WyxkhKPnhs-FN7^}@#;D0v*l%20lZ^6`#zsoC*Cd;~{>JZ8+ zCgjpcjCQ&doG2t~6>{>;wr)^oiN2f#1LEnAMA+5l^84_FeEXdh@~O?|h0{sxesd*~ z-w*`i-MGJ#MEq=@vSE1xIig;)zVh3GMN!sROraF`D4EG_7e5jAFg$5RV-9?95}M2w z@}7NNKj0AXQR7z!>l}Zxa^M;x4z9RHNeeL0a`q^CotD^mQ+}d4?9GoK-gEDi=axm& zi;FCpc5**U!6z2%7(zfJN2kq0Hh>`44B~%7>Hkksib61|?}+KRE_6G=O9<-}+L3Hj zG&k%9uiu1`t_xF6wu9Xdo0OFq+RNNrK7H{%uhuei)jSK_0m?pu;6iYO^YDN;fl8#* z+@*)SGXOG6A&y&x0lE=-JnS0)K~NH~q+BbBln1zhrs{xe_&macH4*mF-R6X%z z%oRcQCFNLT#j|_*RL5itJ@+ulRfT6kIUm67JS@Q@EI(@Q1SiFQf_l<+b!OAus1K(XRS0m;JzuQAPfg#o8c}M~EYZ{cvju zz1&LYCPDUUtug*{;|H_HB@flyR~bnWOETpb3?VpkSyd1 zqr_JA_kv?<=cmC^o;<>VtpeDDviti=UID#20jlrice;q#&*3GP4*|}7c;OH#jZ?M^ z00QGyT8_*f-dm9EWv`a^6lL=ZSR*b=hUk19CCe84q2H=n-+Bnj*R zL-H~0hm3$VdgCdaMd4}Z5uQGpbvk1G-SXq;^&9Zxwes^ro#vCRVo2NU*76(?C9PBO zCq&!x7JvsM2J&=F#s#mDI;Ruvp66y_r>M!*pOe&M?PVl>571S5&V3c=#jm0TVK z-$0FCHAgb$W+I%tbc}EbJ^7Qp@68SXjhln@4VOu(6^1g%;+2L+wqv7tx}zzy=RfS7 zUu9|cUY6{%nbzOI{2G>vWiG`#A?yX(rIne87eDT4`vi7775JY%^C~=7%$AI=5wv{~2&oI?-my#Ur_r3&3f;Ozw=^%ClM^q*z8tV`IBTh1%~HPt8e{ zYN{F#m%@Wl%-+~!`C~)Fvf8;pR7p%jrWR&Ssxlu%6^+%(@6+w628FRk^Zp;BQhynZ zAJ3}#m(i6}aHyt?iOd%nI^jgZ%6@&VHM*qZq7&)aP&3b!yL~Q|gbr&V{{Y!c3I4d0 z$a8T`bHjFngxdtG1p90#q6L?52OVLgpbqy!C=n`?Bs$am(=Xxa-{B4t7LC z!PAf*SyRZgj%1+6pr<~J!b3W7<@5$jQ3pmyD@CLjje3r-mp{%o$Gnmto zN3>79R`wr*sw?)e=cU;8mLUj7TWZ)_vkL(1V6U@7GqAtXcSKYHB44ZpX+ll_GLObc3>J6eV)=(hEF)^~8#Om-%Egi{u zI!{RGK2KeV#Lnp}@1xiu{V~l4kSs{#aIvXo*g|k*5K`fX8}5&Xqus&Pjh3xvWoU;6 zG$0gy$72~?ox(p{+)!>O_jVv6t2ku4po_EQk+gY&BDA`)qR9_AUcFz4dV)Y`<>?}D6BQQ zbJRGRuY0Rz+1CLz=%`L@9bz+!*|wqMNYk^V)}h=cURl~gUsV|on{mAdwid+w^!CzC z=j01I>C1z#HqhzC6P#_98n(w}K6uX8P+L#-h_$9$Hs>aI_1y_HHiRTCeh$Jkmq!wW z9UU4WvYI>MTb%d}PGfv;wmB#i+iGmN2dGN%)-v~yvmsmf)|l3+#nP2Lq7t;06l3q1>V@ zM{U_9ebgW!WV(=We(z5Crn#8=T|-t!%TLM8iQ$*#JN_Tfht@lY_Hpi0kJ~=!k@erv zD^N&r@4-K>x_m8QdnqN4@6Q9&Uh|ZPse7-L&S_%ZJ0JhET%R2FNwCMKcs(^*IU^-b zFN}m{`O8Ea1|PObIx93`PuyoeCPa0FowsX|f0&7zO^a^mFpQl&k1HwVd%Pij)vJ|r zHhuFYo5JqU({?F1ly<9H+fFe#u!&%{p-anhOjr3|ZaCp@ZrD72tpC@4(Dj$_AbPHXt7f;jTvY@<4sD+` z)bhY*+7*t2{Q<&N=#vQ^9eOD&dp{HSIGB*81wJsz=#W(8y!KuM%Z1MK`iJHVB^*~y z6;}KG9NP#L~Ufk>*m6^tAs#1~3>h1wyGssjcI;RJ0h&Sc0B z{bT^pZl~(26C&Fhef?S?yB)L&DkjVV;zl{j;z~6L!3V^bTJ(s%eo^%6!?i$U%0a(fDDM+4Yfh)KO3QxXJc>eM{;nWkl6{IXWuqfeU^}dh z>Upx*fDZL(I{c>uhy-5}-B^}}_Hc<*A&_FfJ?2Et9d4c#;jePa2EfM#k(zUk6$`bs z+g-=h@0` z0>Vy7IYv$1230TEiC3xF|tn4E7$1jYQ!qEy(2U;gqlFS-Sa$) zuhO&N{NDGUYmohRea*SfUmIgJ7H?be2cd-@i4!B5)%SbYmG{3Zhq;M3Ul83uYmT<4 zVP`}pM;5m#Wwy!Eaz~(7mQ+H5dX4Hwb!89wfPd7hQhM+|OAq4^LF{)gkYW?bhL^v3 zbSGufZtzoJ_@<)vP&39|zWwh*bVCPhhUctwR;}CCkE^kZN|SPOWQ@rL;XosT)+P#P zHTg*CwIe#-d-KpAd_}Y}&HC(=@ke6$|AOmn{NHR6a|`=G{67#c#Dirn#&gGW3n?}= za~``xeNg7|*ki_H=7Ew|8UGQ*ULY%|YrAi!N{v8*&^O!QjwE2ocd08Ad)#X^E^{j% zZlV>-q#NgEGgzT|Mcl1qFgn>tMf;io zsFBYG`C^a20WU4JfvUYT;#^X6iU#2bA9I$^d)s!&q=#TyRes|eTWL5IRCmfOPtUxW*#&R9`n2>p2E=WaQYqMFXD=SJL z^mvW?P|2AB;7IO)yRk%YDiU-jbL$S2Kr7E{1RV=Ab1&8C#f7ZG*_-ew+J90`Q3S7^ ziX~uV@|p-AloQgv zbfI32BWJ?+TC0NGV)RX{;Lqh;-7xoCb!opAM<|POD{FCYx-4@!Ll|SM9Q&tqvB(y7X{BdIt*XmuakQerJ`2VTj$f z1!|*`&vUe|2+??O+^4s23Qv{Amv!dWD7l<$G(`+r_x#pU7L6E5>#3gkS_^e)AKL=d zLi&a;1rA8Z4i>epN5DhqR34o=D}{qA7&XahUWZXfj~gPQ(5iZCowYvge|}e+*j`z4 zpiUcx55;sjjoZ5YyZrl{vkx36*if@$yJVk~`5pe)ser|&eJ_7CA%xdSHlHTuEv{&xFs&t*8c ztv!^ROY)RY{_=kyf0%X2sq9P|Inm^{iIYkA8w?NxK1#SFqajoJcU#b#4`47sPNtQh zmk`80<1XlWmjj`AI19@ZM|eD*GQS2qPdM>U!~|BDWQ6Jv#1h^lA#a8hY`$^-kGDd$Uj1^3Mpz+uq5T+|>m20s zTU`kd2Mu64YW5o-;`!-&`p{(d`%$RGmvWmAU);_8;2ep*|ZizM(7IWl%?GzwH`*BdLJb0Azaj6{}@nTExBje~`f zD*G*ttQKk-^Mz_fpW8f!F_@CFcw2(M?g@WhI>W>H(u6aSWUP@omA^nsh~X6msO(1w ze)Z$K)X9%$@8a={`#V3fh8xWP1YWpOkNGaEH4;vFg?OG#AVpUVpGpV(5{`0KULE`- zjm=J9ztgjN%ONCOTXP5RS12#?K_KpyLptas^Sy-g>m6VdCf-r8)%rQt7tZ0%S6)aM zF6%mt_s?Rp5If;TL=mq$)r!-O!i96df)4g#E8REn*B5_9B$##7)g#qgt=i_yVYC)E z1jEg_Y|t)PyfG^zuKs60h_Su@b^GleT);%a-0aFyOv+ipDoE2MTY|XJ?7SPr8}>TP zGWcD}^S<*T{W7U~zg^PfqbdC11Zi!<;1hQ*mtDz64*@Q6lcN4#RcMQAyNRrhgzt4kXUm|TL z*3ynE^yZ%rd|)!@cOUg+_U$I8Hr>AbVp5iLK2T2Q>Mm#-9&cztg=PdSY21 zn;raMXCG8ac8Zq%g8|K#TTb@k^LSOf$zx*z$TY9 zB!iN$v-n(fyUopGJHhMq^B(>JA%%&QsrQbHYb@I(?c({U!>h4{@oKnv$0%bO%zQJG z)$trTb7o{%3(VWu=d6%ba=1!=ir(rnt>1>Zp6Sx3t+W*!bU*0AFJoYwv2H=$FXCe4`3&NE+Gvq5E=}4<*`pBH*wqstwMrA z!_a|9v>3r8(#*OANXaoC1$RRU0dE2RU=yf@?VkiBgN=032n4DFl!#7Szp5!22iAO` zhk44hNaA%h#oh;k=yMO{C*~4pJs0Qc2n9L!r^Pw5%ySQciSJ@NLF=@mPJBOd(>rsI zvgc>TLRTzr0xCoGk-Ezo6PBF&Nme)b3unP1n!guP)zNNCO8chk)G17)e99d7Ub$bv z++qnqLM(S|HKS1dNi1KvPl#?wlvz1SqsTrKa&RjBk)!%ftL4b^3ur=k@r)cMOlb8A z*+C3+6*EL{a@$b0yjf~P%)FIV7{K4lftOhXjn4;F%rf-U3kZX*7RN^E(9q-abkRRl z<>2%{2yVkvJL^7Oo`r6-Fj+1;Y2ttra++?`TAAygS%0+NVgA6pJElWDZB`*8lbzl; zDIYxO>N??a*|3lEAmujMHDfw=KHrLV-Oj~e63J&Av1BltX`WQjpL!+4#iLT434zhh zOvQv-9=$(Azo)Alxc&Tdv|a02g`X;xZ@>GWLuB*)admE0Qi?OgVV~nDr|MPY#e8_9 zW>5|}hfSF=W? z_M}a`mlUfXl2sVyYKY?vdKo(>{X+21vVn3|NgI>$#wDuia>?e1TkR6Zn~px716tEn zomr)W+3JQpp66)UgRB0lY2We3$_<&JDe+p2)4?wtSAJh)Eh^($6lgpZ1?<1o+~B+d ztihiDi?+@yqdPOnZtp;0$e^pgs0mM5^cU+(36B#F;G7%b2Wb(<6xuUT`NX;U1$b6` z5XjHqUHIo0rnvQ6OIZ=C0t8LmibkC@9vz26W|4K4){5hPS?unNa0S1jOjSo z>OnOEg3)Om3G7Y4Olo%eMghKW$%k;;)E&w;aS6Rh5Cw!ADP=w!Z?5T3IDS>8m!_8% zKga>)3@rv&5Yr&PATnTt=?By7!H*K#!~Mz$eq(u7zX2n_cAoK0ZY{Tj$*UaJ@n;yo zJ;Etct03j6#97C{#kUc^(w8m2ygu{QQY5<*zxr`XY=NpbjfwI(h7JwT)F=+-r%YZE`E!AV z06_}4N)n@E5~$C)ssE4}F6L!DwEkr_=ADPXE_?t#K)}BflIplmG6V8#7d*}C%s-!f zmNN9q4IrBUnrXrAM;M;rzE1_%}ubGZW}z!SanH%cXn|?BuYg zGZC9IQRb+s+Xg*6EC=t-F9%apH!==jtjax# z9AlxhSkloJ<5joW>go)763mfqAQnW$V_D!@2$XzYkz_D<0-m1Kb`e{;asV5+SRe*f zwLO0tv9nX%h%1!rck^Y^5UA7ihxhHan_xgchNdr%<%T1J6RDyonz_o(6HJr!^Q7L4 zo`~rX%?yr1JR?0n`quK^YjEv+`7|-w?3Sf{QBwQZMe4cULaMtlX}C)~3%c)Da&)*} zV9qDW<*uu`2HiEVNlmmoO52_G;|%2_{dak(r9-b=kKE2^t`F@@z2Jaw>TEvrIix=E z=sA@To6;-Z11lrjh2FRoo_6kV?r`3RHu_f_qcvSVXY|}EI}<3~ZbF=kbi>Aak`=J` zr?i`nBV^wGnR1ers7&>3?)ildf;SdCv^CZ|lqE6-mz;b2j-DQDgHeJJHRj zh!@HICJt3wneQ2i%YN-S6KYz&JoLVH_|PPkmmZCzY7MvEKDbOMD>btT$PlyIjV#WF zb^i0h-~;&|W}crx+4R~wtJNP~JesG*`Bv7SlBT!c>96`_t6Ou(2ga;WxG1UF=DhFw z$5tqfTV@l5)gvEk#PMYAm%{>RyghxK3oexBCz0I-!omE=5(TQRV<6f)&jnkkLacsF z58KVsUJ6BIxoOn3$sHhU2K=<%ckR&nE*Us(zB#`E3kz%@Jh~H9Ox6Ezo+K)R-4X`~$L@BBzO-X^e?ko=i+p z8tckVQJ1sNYQxXwU(MWz8%j1xheeJ&gumd>Qy-Sumsyah=}bKWGz4HsnX<&-$ENN$ zjt;&3Z>k@aE~FURxnI0)+5P*COFZV}=WdzXgpckqMI()>{CRRRA5EhqX6t>Z56`K4 zjS*#q9;=k@7!)F6*r^bnexdAAkAnj4WAi>*G80RofgLr>N+X3#<6`J%o1+=iM4G1N z=VH3;OV~*9ip1CHgH1t2v# zmfGj3anISuJ|Uv0UssHD^RCeHb@-lz0|E`0{#!P(-3HA;iq5X%x5dkEB0g?1w}@<3 znKsg4f=2YFtTP0B1LP@NBZ&wuv4tR-;aAKMFUXReJ^w!4%eJbyvE=9XpN0W7&DJLu z5nL|PhB$%dTQIvP=8}kF76{>h5D@RP!dllix$)R8DSxff_;K}$XHGvGHGQI%Ztcr` zWIG2~TakY5vrAbYFLAhG-$Y9s!2dWp|A-i<_Rx3Z62i4yZi!W2{s9e5_(k6P0T@qm@Km+7-Bcwf# zD@Wf1QFa#f!~zNoB|y2@Nj6HOTEFoL?Kpe27lYAFxiFNvv?36U4t%HO_?9ntQg&aU z?vzi6k@~Dwz{jaN9XR>7xJ0X+bBV&C?1;>rb78Vr)41#QNke|Fe>C5JjgfA%N+;{* zLu7kT3FH+wB!?<&CTi_$ue>U=G~|`}$7XaH`*vs?U-nUIv}KyfH69rGU+Z(keG}R* zoz@OXntPOK-|n#z5Z_QEyKU0iR@)j|ZqXBu9rg9SvaAQwe&ktP?qaYIb-hy*ujOMP z>v%ao$=7sZ2{u;NYk|8_?RJGYuDT=c<(w>XcObpteo){4CKS!W{x(W-%bSb+E4~js zBIUWE9(k9I4w=Okg+_`4J*(QqoDdjQrT7u7lCIt_!PD!tX;hmR%dmC zhDO4I&^+^@Fw7D6NXO;Pd&)@OH`4jbE#3gp0AMS$HSv1!(Nk+*7y2IqSn?e_J?&LQiM@(R#wPXB!Ez?l}1_anQ^HgkH0xDz~X1qw!&Hq^RF=Tx(S4V}4g zvUf2b6LP!a8yJW7I*8<_UPtb|$L$}!mFI(&#u?=MXn{n}7HH2`dQ0=dAtuev`~6m) zydP%LdQx)soOmx&B3XisXYj$Kq54A^=A!&{qMVC|?J)}13{yrp?DNK0ADz+80V5!Z z-0Gr{E)Vf{jjGPxMkA|I6Z7Vf#}FI9FSPeZ_H21r#6Fo}j$6LzQJH1FNxkpx0&5MD zYbO6;ynQkpg?M)K*y?48;1rQavW$(zA%s<|)2khLz}#znm@VTi@0hVc=W+@n#f_8o zFqZ9{8w`ZEnseP74dS@NUS##KSv~WHBlU=|!5djt5Yx|6LWw1p55qhU5@O6-W9S$NxscxJAl`|I|iIgR6eH-i@Z-Awco0(Z90tYp;n8c*3LO!=g_ z1v{Uy85=M%pR*u7vUP;GB$ewwy>N>?Hx&M}M&k$fa71=d%4>01N0rW8?KV}drdMCr zYCSe3>ZP&S{h#i%sS2+S-4wf_E37__PJK!pUTp~t!<<2s13IJH{+owosKYkZ@36Rs#X^drIDrK z66Jnt>Tx|6%swV~Z=LIY68|yf>VU2Yrdam4yv&dPJ9kv)rLG8U8Y)~2-}pBn<|*#3 zG^OW`YVth!AWW*#4-@7|b-i)E!{df$&KjRTAg`mS<8b!R1p)+OGAbn`CnS+@QYdg? zQgO#fffPjoX7i`I_qomtTO6EluMGg0-z8jT^sGphM?A!qr5fXkqVuRez3i8GtSVb&R`lS$%L^(}#x0G>^L;@AO?e9pR)?!q-*k<<1G1aQe#c<4ZifjNCA z9mZaDi4s;Np10f731L02s=VJ4TKxgHWA0E8GT`=(L(yuV5FD_tCV;j<2Ij%mz$&4` zgk$?vqI&zJVu5GT9Z)JF-wpgC0ostqHxVw#+ZPqzD-kcbx?!RYF_RLSM72KL`(t<$gW zv1S^16m5>b_rm>3j?R!$$u?AH*exy9n4+(#S8T5)-!7{uK&%gx_1(lxq89SyT90Xw zU}BQOiDp)`nrG%zDFu!B@o>N3xRf7s?k7Jc>Ac*0m~%Fiw}Fc=_Uv&BL@z$3)!X?6 zixnEw`V`lfRT{BIjl@orR@GsC>tE_A9?c%&9P|I^I6ARM#%20w{jU!uG&x1w=GpDAg+euDM;tD zwz;2koLroTn3KB=)iOcL9EoY;=@lK`9yb=Pl}2=5tt^*XRqe_3{~eWRHyB;-2$u&; z0UpI7nTu}(z^m~mvJYLkM>WZkAxmrI25(=QTybG^k4bvdl)qNCfmKP#nC{Ite4V^k zoO99K1gsOcBKJd}9^^8I6MX$PY8(`0>Gy&2OaA$JD7pQb zX7`|O@*=n>hnG?yQ2}Zn-&JkgqkUgUtk|1|;r&A!C-XQnn9ZpOFcZIB;CTKGL5pyd z2F>6(29SXY1&-Mv%+pC?2l|v7U;P28AeOcCt(l<#&VfPHJYesj;0jv69J-$fH>IY8N@j9#SOR%T_O;XE2poWB!|lk z4MsTVpyVEDJ6>}8KI-%e6#mZxAR9xYmdMZo{kQZzhI5)pVYh%yy#4&G8 z8|ByksQe>dydeqSQV)J>VF3yQJw2GUQ6%#opI5)Cx$k36Fw9(*SC~|vj9sAhUYePY z1%(}euK^f+JC0EE;fA`Hgy(g-wJ)dvr-M+TxfF(5{KRc4-;=BB+ z%>Qx~21Ja0O8x4ckZ9C1AUE%rs8e2hHKzA9{!yma>l(JQl+_c1?$K6Mu)ugOufI%t zl&bG^eNz5G`VGn1YONmcP@M$_jliv)&)>XewcqqUC@5a5X~@6$#?iA_lDYPdqbb{x zC@hfY?Jre1BrHkM#I-tOgJJ}lvGDbQf{+FKo3XRyS?%WdC7m-818R}-vahm+GUOH< zYAw}!96CSsUu4}G3s%s+RHR2xV!UXmz)RQ)Cb=nkRt=iMoMRtf$;=@p-{Hq=>=ume_XpO6jJZ%eK(Ik zK}*Ql*L|v4K-)B2;_FUeDg-{?6oB-eF+<8M&FOXnl!>pPKR^Md%@e!FC7nb{Vm|L` z=r)41;*}?^?kpt5V&U#_MJt*iz{!ns1Rgk{n+lTpoa1C3BaOI-FjWM^h6?##H6Ucstu!^psKIvl>)oYAAn zY7=cfL-q-V^$Fo%t4^S*K6~kWP>v7wjVBqW(Y-eE$>IhkGiFnPmWk}swA+24nSL6k zJ!BTyFg7ej9$mI2oy9v_#H0ibpR*mVAY_|Ec!fciqqzn51e|IrwC${q=E?(g4s`Jl zo8j9b1!yf|zBj<(;7UlSK2Eq!NKdEh?^bgbVsCr=ME0pmQgPqSspdUeTR${X92MJ? zqtQtR+}J>1)63IZcZ%KTJU(Bt6rof*K^wM=sT)7%PJDS%!hnZR^MnGWsROSg^5{Nx#k~} zrwP)W>spV`?CF}3t=ZgixndP{eQ_u1ELsn~7;YXytz0_9gglcfKvedybA|A0U)LMn z`C!|8KJ7(|EasKd1}Iec7DB`xt(F;nl4=XQ-H%PjuFOAq6R=JDmP4 zs?s~39WmnYDCynFp!GgiTyo+~*{s}~aPjr|v&10j-TvoAmBq#W;Z~T>&m>jKxgSdf zsT_4WGbQ(@_s!bJu~McK%{@J%@@WY0_?R1VD{cEmJm4X=+*RWwWg)eG>#z%j*>ukX zIdRr$Wu$EZKe0E~xIZS1Ke!^P5uWV+FpB>Ve154DdlbdYmsdEl;SCDwhC;T*Vjbl- zeXsxHEd>D20Symz) zqV?_-m154m?lE2<-;>V@)BE*XYWb~l=jpD7C2a;dR?R!q`H5-IqL5~##{OwOEcuvI z9aYet*rMNfKn4PvMmHb30GnK_@&8}Q|1)iz@b%{IzmU(0E2=-ZyyTC1jh%FvXfR4a z5IUCdtTz1T4j}{yGNH%g1WCQ{=gtnIVZmpQ>YrTN&2@|ILAG8#zB>+lrL10R5;x=wwOzx3S_BFL|b(6V<~S#tL5lOIv8j@cHo9Fa0L&AG>kW-(8niU0nutV6oBZ7LUv7 zd!P-x9nfxXPyIiHy>~bp?At#qgxHkUsM#v1n~GIpQ>_>!9d~JKM2pg*_Fh3-Rm83x zv$Qp0s}(`*R?M0uc47vxV*T3ZdA`T{JKp1X-uF2Fzy3R~&$!O>a}6<#k!4QZvhOxT zz;}p_2kW5=ROEFFOEtRw1FZAYKZms*RZlcQu7m{bu4 zMcszC>QqOssOUA0%(%(l$xyd_bV~}I2XA(GeqNT6XBja12=@R@UDA9yP#TosF*f=o zYkdsd7Zmh8#FLq_!(wWi<#Ed#cJq{rRxzwJH_$r1RnIGs@*Dk`f|x>$m{n zFrY~te52H5ma$rV#D-Wf0l^0N8V>#Ng^PF2 zI)y#?hDSQ;hIN)m)~5u=?&$NP^GV83sJ;Hi7oz~r(Xo3&LFYX`m)CYz4^RHH8s$m= zLjNdqZtERIx&LbX{;Qz$+YGEZgSUH4m5`^oYK_d^qQ=s?0!wsW? zxYs*u_a2F4Tmy?rT6bv1u~^?)UZq*3#R4Na>gvi*IWCe>D4O8Vb{fPHpm=gFg$9D- z189j}l^yWXK~nlxrpUl=nOpCbr2Iy5@kCFN_bh}l8R5BD0-!tsGK_+!)F-%fJYcnd z>w1K#r+^scXVu#k`rx(@jy^H{2ScExK|A~wDH|86-9NVszqZt}lZ+|RtU5Q1U^0{} z5<AQ#Q9tCzk@qLb2ook&7p-Mj158YgT*%5XvrpU?Ke7=4qROCML)16}-^a zVOXnr$M{etRf5wdr0+(ZwAN$od%<4g0~q11{eF_ZO{6O6jmE^cGB$zimv3_*clpeT zVT5S84v#BB+imYtwEgvnpXBy@dpZn-j)1F8k6-*qoMi!{<%r-0Vz}JEhYg2v=48yr zl)KDlxx+Vu@^m}70lJk3POmO~QH|l~d9n~pad&FPuZ^FLb7Fx~3{U&m2LagI^mXeYicA)d-&<@vm#VwK-U~Xx4QN$K$7Oca+>Kx9unQ zZwaVQ6sGVvIhfU*eV!f95nc%!MrXuuC;m@ITSHwF1XGa&1NVT*%l{oo@Se{2nz41d zMS3}Orbt39%xfkcSeO==b{+5&@LB>U@l|3*!b}ONuN``VL!D|Qs@SEoGBZB?vTzOr z+z#0@3#rk}pD*oC=rQNk^-@8-aISwbf0N{?Vx4EiBk`LO#s0~xOs!Ch)(H0y`Ohim z37gYb5J^%f;cHig|0u67Vbae?o1ymKmhG5+P#MWZ8_1a+(~YmCAtGf6Z{C^1yV(`A1s+xH{k>O+(z5hU#3)>WON* zPatc{0f;D&U8WQK$0UHWxa;c$KT<9}29Omb(c4RxEpj854qoD`!$D^H=IC{c37-hyFC6~u z16P~$2pMuEj6&l=oCK4yf0nZ~7s5|b&C849QJayGn*MjU-fwf zck<0W3x?!>Q7h8eWs@9fj*~il>9}(7FCvoQ= z+IWMWZH0w9EOLyB-|J8&vutzYDWmqckFG&r6+ViLCf5bXvB6D?E`J^RtqO|roEcTrbOE&F4XF_0D(^eB7eR7!zi$OoC z(J>8c?fnvULcwdS(B^uk$~3?rK&FB9#2ANdu4sG&v$rE1B53mIIT5q1L+Fo^rfPRt z)InOdBZVgh3{wn!Kbh>2s$w=Oxos9)1jI%#F21j&Tmnwcoh6&EYpEtpq=<=)ZREp6 z{>f+eTwi-&HIoqVAmCsm!rMzlS$W{^q^R$<63ZlRN!&LUjWv01Y~>%d{dYo~rjaSVr?#iR zm0>fD46H8hD(D9O{yy~vTkvyoVmQb6=q`@ta3-@F)aKANMUr??+k5lELm$PAq?dF= zr9UmM4Z7wo&Ycli%cSfl?!o=sWh-5Mm>21IAY~636q58)Gl@8$5~zNLzC5#(; zmH-;gV<(*3AD(dg9~Uo44$g`$n}Z(4$G3uMUkYlUeq@9-|g1ljEhX>`bH4Hd}lpp<5 z{RN^`0GSM?l8Sis`#^xVvkgH{rL#u#n)Y7Q{ak?^sk?aPKMEOu@Bt|8UH|o*!aR_) zSq%HIMWQquhG^NwM<0x>_(I8sf;kFK(awWH!ODoQtQm#jlo1W3Yme`w(07d2T>7`u zq?L1TVrb0s!EeuB*xGRof$|D*KYE&ptLZdoGzd6!#O=K%Xf4Ij&~(yn0(2i2ZNa6~ z7m2Livm)H%>;wykE@)o(Vu_bLexrHDR7#{f*?ow7j4OpdszS8-kC`(?Ynhge!Zl&8 z(TE4JH2L_a%=(lExWCYU+Dk64=E-cwBHs?(+O%#3Mq2iGZVZU*R zS~fBl9q8kyd{;be|sSK0YW?EmWmU}?5!t>(TA30A>)A;9Us{S=+OJaQ3BLgQ_z%7Ls7&f3*U<|IUmA z3}{sjT1&4mX&H%R7)ws(g)jOb1bpu`PR7^yo93I56FEL-2Fb6_VAE~U=zmSpPM7@; z*84C`XM~_ioRJ4@=<`s%oi^~st{L?Oe)CXiK{=x~VLVEe5>=Y<@bQ&bC#Q#9T39q2 zQAtEwWIp~ZhYr3JgD-PJmONqp`8&= zZ(p$$r4zrZ8Wm8f13_52mx};DK_pOLY$C@ZS@x}ef<+M*zWxG`)3s% z*z{Ss7f(-%jmHLF;rmJV8%lz_db;e(H5$CikNEINO2X-mC1rB_vk;`3uf5vtWZ`)6 ze_zvr(g`|TbxsgZyLdCyDS=L3wnv^xJ{~|g(n7U6Cv%eNbCx#j6JIB)Bwpc`?*vS~tdO5EIw=wVy&GtdBT8T&ryY=c;%*_(s|1LCv0=EH?%5zX4j;dNOOKSb zU~2+L+H14B=Zkz=oIJAHw@idhrmX$BQb2Jd=Sbldr*XZ5jiqEVJ+W2B(r0Go^sr{O z%Vozp1?RL!`?!>yfwk)82AyL`IeGTFQxze%p(aYn%M6@zdtU00$2#f~A1xQ|Mm1n# zTqEaueE}QtLk{{;`%aM*(o~bqd1i)T{HwAsnO74GQU{JVpuDqV9)<{94k3q&NNWK3 z%-!7E*_2DXL%nET+L0jIkIhX($w7xmSm3oucI9)%>S5HmQd_vTl$Qgy%ZQW;3WoY* zTWPXv4<-L2m}91d9kI$ogk2(i+@8{C=*83K1WkT4t^g~sIDqsw>DW$?szm3UqJU)ve-np!Wzzo%;cGyk|FTC4w=%f4@JXU>s5ZX37>oXfV=Y7^Lj zfH3kQf@X#fHSzuY z1gVc=Pn5nKuPri%P|Sm9!%Wp`Zw-Wq6(eMyHrM)&T5;0@HL?f9@tTL9W|xYk{WeI= zx47Wq0*!4p!F)e`LxTk-Ivr0ujI)?<}^|DkZE#IDUMWk%NSC z>r4II%XiD7MOuGwsOXGoy38R&` zGxGu%NHb(80qxyL?5SZgAKN%cl=U-i8xV!$=$;~ZxfOD<3|g2{Bh|F=1tQJFXzx?E z8Q{>KfqfKRD9`6SDcjp0ZDAn6;p8FBkg>_Ft-+#QG!3h^eopw={xSuPfOOF{(Pt7M zXvtz$bDO9#aTc!zd~K2&hAOm6hx=DRfd0p$>_?Yz2K&V`L6%3{8~RR3&U8J0{wM5MIZLh^tCkpp`kcz#_NWcqq{EW`!0?*83rr~7n*TeqS=ta%#cxdbu`UN=ZU6D`Z>56nyRom!#J`3q;qTL|mxTCE>TzFQlGy_|%p z%@$xB9_U#qy;t(?gE3vs3CC2g!YpZgTcrW;?3L}@2c7aUxsqSoktR}TE+)_L4cedwM zu|R$Bx>~1EU?=FziZI^CW(rCWKbng?8w99-7rgWKnkB0joIE(($snNC}+R=xXh>7akL{FMki95 ziOjSAX~m<1Ih$CB&_)p+P{|YcVDA{-W9=Bx812xC+vM#bo_m-0WKCa$_+I<`S#tq( zCIBH1?mC|uii~rknJ?z>Z*2^|h3`{mHPl%Wu2TtCzqw6|}Pqwt; zYy-(RA1!B`w;*$r)>vkICA6a)1Ddg6Wlh(jg1@+kaDMjy=*aVy{4B7w?J0HadzOUR% zw>3*)LiuM-scl(pq#k~xxgud^OkD!Wr*;E^?sAFv6LY;pte>|z(5+|WNSUkAs4U(6 zI;udTsKf=w$Mm@8d{~g!q(rkG91LFyI50#NSE+AGHsa0BcB{J}R4c?<$jN(-jcpq| z#QU*(R?NR$!nW~vJ<}Z;vx#e*;T+Z}WyN&XYkUJ&w5$tMH5^V-mmhmG5git&h9thBm#+3W%(BMFfR#_R1y*Twc-TB3Y<^zM2)R5ZJJ)T+0)uWX(Efc{Q(gv`-+qJ13VWS)1eVe}>L>DXwg z^U*}9=mi80&P2<2xuHzRZ$O5L`4>Y=UT`R1=)G)tuUH7zzntS0&!1S*-bwhQvj~%` z@>pQYp&~-*{VY0>Vw__@esHf6N#KU7R}jhbfQCHho36LCsPWllJw#DRurVh)E++f$ z79MlzDhrkgIUl+rhqpgc_4)!kwY3e*;iy?kHTf1hhH4&Y^*KJ*!fDKG_pAoa9?#TK z^ek=fKuWH4*;hVlF}70m*`DwCR0lH0B?~5R&xWE%)jJf6ASHbD;$kX0BTd}R--B=) z;3P4WSCd$dtxm5_8LPGlq#+I?Lh11XrlyxUEV_98J7$1V^t90iv;g8kn59mSQ{oI? zF(n0^TVef1^DDRU=G20uaK2DTwAp6v#GY;DkKa|v`v*%Y)sQcZw}X4VjIB1Ns*BM< zmV@OkMwWxuApH6$%Q{lTehl8$dg-%u4i2512NK}uhVuu1_uUM+Gm)Oc^n;%qCFCzG zDn%=+Yo9496&NGu&#p>|j)z21|5Xv*sX4xF)%%^rA6Msx%`rL= z*Btiqn;f~XGS=_EGgPEf#9K~A22y_OrCbXd0cWgnr9r_W2ixWDfgdY9lQg#Pua*ff zEs1md5oRj(DU_I4o5Xcve(XRWv(v~;f7_DhTbqe=T7}ZH1?nw?Bg)@BU^`1CZ<18n z@DqHjXS#^&Sh7eK1>XFB8C_ieWpugvP5w8%#=};)fGuUoZ>FM{MNwGBR!Ig_e2U0QHwq;!Do6r>hSzF+l#_H2^=- z&HDW@(#sE7G2txAU|#ZiuU>@syrX??|D#&@#d}?~d%EU-LuE2%_G{}(@y`GskVPT; z_93;CdyzLnbu*MtLnS9cf&mMcIT_#h?biB(Xegf`K`)PZM2!GUr?+{7@&isX{6n3z zMZ5wYPRYr}Lx4Mn@P`x@K?h!kQE(5S~rEcIwyF!zf zrFlZr-zVulfRfqs0E%Wt5rXnzBZ+3wGug6|O-YDoFne^-T(zXcL$%myqXL}a-rHy1 zNUvG4?j7lKpUdZ~T|ZSMIb?FuJ7d9;u^YX6A8r4=S1_+MPzdU5g3BqgnM!~}D z&{@;mk7-qp@j7GllHIqkT-^SHU?WC;S<G z-J5xFBHBp#xuQB6CEVat^>oE`*!lEUtQ&Y{&?R!~SJ{3kkN87(F_-MHI|`={l+lCS z{yMyS!^=kx12&~r(C1x22aU7Moovm{}Q=XUbELwePB)k)R zP(rX%AVy<6xjCv!T^1vFPOfktES1iZx~#G-)cj{}R(e_ned^phi99Zw7}xfE6cWrB zBW{=1Q*mbKD=k?pLXEYjJv%c`!1^(iQ)dtRLrBURI7FzEl+Ovlun=|^Wot(l9Qw~MHM!`}a#4{!W)ep(RuPS_CT zhXw1_0A-5!!0JIt!CiR+y|vE$DwjsbWS^PSvJH|{$kAfV{^rI@SLxWZU9ia%7vsEk$jNpWjDRzQzl)<0&S#p5(Q$6onWbEbxW$kW zK<#*3yBpDta7Au^x31fbsOY@TpgTp)Egi0*o4K1v8xc=ilnAQqlQW#@7K@mBDl-_l zDH>umsAuN-+Fj2*!Mz|+xx5E+Y-{n+g8Slb_HYtfUpd~aF@Sive?Lb28{W;)o-yb< zAYZ6oWr@fG8{1$Qs8tn2QgLZ`9peLsaIvL^wK4%}>p^JJ+{lI%)#Q9#welfMYwm|Z zz7Wd3|5m?p>n+stM%KZYe0vWMk9F&4J)(Wx0wj1lYWug2i6te8w3I-lzLyv@Z*Do; z!?f3zW$_{c-H0`4dKJBJx!6S7a|7nMXZ$VkbgfMC5eEU9Q;`c3tMwR*CmNLo%5ggG zCd&czV!~h_TXyww`c|S4ySdkl??TMnbdec!>8En{CFrE|^{dvmHG%huG&&bdLU!=2 zh9=MidPd;h58g?Yk{}ZcRfAbdIO|i3#RkiW%yrwTOwbHL%D!KRH*F=AmsDIy6)iUE zRPjpQ^*5sGYG61%6|FO0ISw;e{>SqHCSvVUJn5lL6qaO4z56clF1*K@xNAkxwPAnf z(eEld99|ac_@k-wQMO7)&_b`&qwJvl2_2t?K}VJw?7C75B50ENw>9O%WyihZ+G?9< zM8HPPLp5OYhhgaCQpbH1ih^vI=YHiWL#TBQVF+IHc3l>i7G;?|w+a7zUTH$D0C*ogvM4XfOnw);sDO+>Bv* zhZ6q$u#>2ixxA);PEKc^j1kZ(l9bGLtK@J5#^4HVGRSm%VQ zy8c4bGw~$(>Av0BH0P7b3OPH{1wSn)gj0XwK)}I7XtJMQuu$U6>g7G)U|eh zDA2i_I0~1jTa73M?y;PtNznhN_QPovM8Apr(Mfli?=lBET$G)dHF`e1Zq}RgEVDTU z@Lij+C|Vh&5`GsU)$Z#K@r`C>xL{-C=j`@PwgkEl)#V$`+!GX^LVx1 z%T5f7&liq>7KyY+=Kt{V;|rOz)3%e<8zzTiSglrm+eo`Dy3!bIdi=tL>lRheoBuME zePu5Ure~L;Cti|Iu*eB_%oHKL%~TT<6Ew8V5KB~elW44h5<)Svd#Z-M?(WekinHXm zP=#fziVsg3n+;wN1bg?onhb(NP8LL9eyt$tdykHwe(na{M(p|b9obNkNUquZERhrU}p&XorB?G0V7T>XW~`=F2`=l z{X$(!-fDt=CFiYKQtsqP$hRCfc1d=*#>*bB*Oa>y?4PJ+DYz6GgnHZooJfgh*oOb8 zQDDYycT&djS1=pWRZasDc_FeiALZ(L4v-qpd_D|)+XdasWU1X*QAYe9e01L=E! zJkigXtf-6nE1}skBZVtKjVSN_8y{p}czZLuyA=wEE$p z^r)skQ{`&n4!KRB^HDu57n(bBdppcNb$`#i(INaiDJ5Dfm{Y&nhRZ5!nCs9tYX&B@ z2Q1hw^xG3pv5?2wJDjm%?X#>8Be}s0K?4RzhL{1CO|;#ak#C{L0a>eoMV9#kNm(_o z$byzvW4}$Tb$m;#!H8IZcNDVL#3@}-JH)loHS~-#uFiwx^(1~iB1)-u@wB$AUu0p7 zmK4@WDMI_dik>=XIX}G*Dj~iF4{*x#&n2?eejKd&6i+2wOsZ}--C@%q+uC)5R1dSA z@^n?1X9+6QAuri&nx>x1nR84ZSwqUnsq(&egwobu!dt|b3^$KCl%SK$TC$n;-~|5i zfuJhHKfbUp<3!L1A=&m_SY7N^EzX5f!iIt96`BTR9%rlV+t0jaXR0aN4@F- z*)xA-iPILT&SNhxhnu^S}x7T?9ot?qr z@0e_^Ef_bFUnF^<e@S}?teBq0 z`)T!L23=ZNh@M*jXTMv@%rG0_WK!`_ei2-Lr-368=DZWvQLjd^V34jZ)eeUpP!e~oyHUG zmw6GIa#1waH_6HMO~T4Dv-i)rwY}+EoPhyu?pVr@QNHk#izS5Utw%4(&giS%w_O~% z%7g);?z!kJ^~LFajeb*K*VTk8I#BkfY8Yr(kKE(lLAuP(-mXlHlcksLs)gW`A_x?T z74NsLAoc!dW2a*ffZ#VKDDy88UGx0;8m2|l;nqP8TPj8PNlic+N_H#P5hxBknA4vj zsj7BZM4g4E93`nl?-RQ{ZKyT%T-4kr9*~XFy${C$2G)j@)Y`fGAr=tP&w{^Um?3AY zgRdgq{p+lZWo|$gV;0o`d+`NOnaXGEa7|~b%*B5{A?R&>n$eGubN$wq^I~u86SN~R zZ_X#*ehCUyb9Tg7)A;9RtR1$F53l)}ix4Dzol2yU3ctdbsigXq(uXKKg4T z@vRAa58b6>SwuobJmY@hHBM|w2B#E;H=p9IquirG6V-ed;F1BY^t`M>NRz`Twz;&| z%9U;(NsLW%qs_~_VZVJtv8`ebv0BOEvHgQenv*P&qsl1XTe9~!4J+w>L0=j*R$}!X zRS95sQEeGfhoWn~uHn2*gB*hApL|h(O)>n03u0y-?*0(A0?wH^%sYP#(LEhCc z_024lrStT|AEER={k3JUeuN|mU5 z6PJZL;*5U@8qwOU4vBvnp2N9y%)wn;$3u-;+&m|a+Mo6I^uLtR?{>=k#SNbZmMh@4nyf2?DdkB-WLO5m_Pion9o*nPTnezp}C6106;N!7rw(s1mh zK%j@<0r7&YZDVIrl{u?s7z-`OvbODm2_DVC#6@)fh16UDVl;&ohN(~95?zzBg#A)U zr>C8pvywa+h$wryM&s!zgc9)#wh0t3H?(%tMGUaXbj~XO8tCOae8DdSb^z;(zb_IA zE2Mo)71_%1P%Mo>^$wg*<~NE{DyrT+oW7~Pobi?r_SQw6_Xiz4a)V1Zo1&UNuzUD@ z+LY#(!fhJkoUToG{56&_qF-PX;)l--MY-3j#47K_FaY^zp0@?si(Gm%Og?Jk#mb`T zJeAdxjbx%V(A-TITq)A+eMAChO|@l6bMa>ldpz6}wu_q^g96 zJ1^ZOHFrLnR=3t#3x1QnQi#1BZtl|TrEatx`xZn2*58H%hJ&}iW@Rqoi?3I?`9G4T z&bu`Z6QlECu_(C$Sl%BjRV?f&tcb+U)n(+L&3RPX2#1Hz=he`88%zy7EhMrviTc`| zg|ZW`u5mnGZRO}!dl@CcbbMo*ECRTOvshK9ryXEg6PR1aP8-F->&g6K&@Ej&oyaw z(SxC7F6|MKjn~f3G4_+n*zK55F?HOIdX;QP7(ngroUK!B$_h6+UkumjLE-ov+FvQB ze83I7iK*@z-Ho+0DoWc*@9S5Tl>MJ7&%T%nXgw|1cXIsO?6Z|>Zg&d6H&(;Ju=_z(^0Ow~dn(S9KGekq6VbSZ_rY@xZxYuC#a;ZJ zUy5A>7c=##bBzjueH-lazV9M{Tdw-bFGXVL?+)C`@P0Jh%=2KtlBhLYfD68l%rw*} z+WtqgBLlsCc5v70wTZTHk{qV|;XTfn%xenRn=rh-C*u)s^c4ZyUcXRLHJNU%39ftB z;6H`B!zw-7#A4Ou7ZcUnBLdUv2829QvYztLek+8~Dlc=;GPrt`Nn3Q+om+q6R(UDD z!vvBys7tknuEg=))_>Qo!~dy3LAyrgbfsx|pZ;U~w&WbR7L*xzr6FUVe6%oE!natCf}sGa0k zd5-8Esu)hXmRX(@W>YY#geRF2qC{-q-l$ouDB6vSwOwqv6wTq?&4hMChXY^>jGK|- z&=`a|m~g|Pu^vjDd#|;I$*HOTl7;b&e!a3-hr1D6V*_g;t?jDcTpU4>KN%)yCvU_mNU8(TGieEVBvr2 zv;fD-xCySYH~%kcZW0=u_*>}*=TZ3shJW8hYsx+^Kw0i)ht9S5^Q-mSbwa?G!1P51 zjZS=c0Re$_*=LWqLFw$n2ut4Kn-1IHp1BA#_v(}q{|75kQS?}-Q+ zgpAl%tLWJrtcCAxw(AHn6d*-m*wLEkua1~zZz!Qj;;ssr&vDhBfYMvR7vQdZqInY6 z^KMbSIqqa>`pQE=Ih3Gerg~3N!=j}`EWtG)w1&)YprO|v5=n!;N#cuA4B|8EoXmUjsYD$3fx(<0{x-+|Lstnzy zpp}UHQM^D-;65}KS0&ZG>Uq^zd~3!!I9uuRl;{xrz5XAiT-3y)K2o1aa-*UeWTS4riY! zE;3kc#&?MIIrfp*R6(pdGP|ISjO=e<(FLuOYH=1IspyF7w>EFW{f6#-9#8$)FCc~> zMqX!Ec8|lge6ve!@J8?OKUhyU=7Yl(<{J`AW8E7%FO-z6wYYsik7{Iy1=`sF3q4N9 zWGn{$4Iw~89YX1XdMkd0ZtV5b12@VV=RZg>{BC8@f|ZN=QaLmOTVrQ_4i5~^b1p=c z5uHqWbJ;c1(-F)&{M-hzKLj~nFK3`{b|#+1g7>-HT3~57$$A zP+;{Yy4FHh*>w*gLV<;hd3d^Wv!g0kvVp>ncVJg?i&p*#zGV;dj3XldNRnXbTduFc zIY_Pv%1Kekh>m@u+x8oe%H ziL*twehkgn4{Us_)>&I{LvkWCmY&yk=meBKAjB1jyRW246C2&*>K)7e0A=DeTjO=h zXK6w0Tzn}re8Xz*1RQ0{`r}YlFiEOi>oM_FU`(uNi{HmHM(-4|disLfDO~`#q%6@( z>4x9&B3MhsQ_^=cW%Lt%FBbidBbDbRuATk4Dqm3dwJaJLiw%ow9=-YDP8BQ~CWz3D zG;P$jX!9Eh9^L_C@20x7@e3CI^jVo1iCuW zT_a}Zw&Sd?Dct1?k*9m>5vCgo_|g5`$t_Mc8rs3mHfdGY?)xTUE!t+|Rg_L2 zZkt)4Ur-Ni3xoTb|Lo(Zg^dv;kFw`Td?#)mM}zNxE_VE{q7{SuHfqN75Ks>#o-R!r zRQQ{Gkf1?Ungl(&8om2a+?;KkU?L*2DXz3XvE9>X2`Bl94ktY*CEpdS$tGQnHLO~P&e>GvEF>!sVT>X?eIs`i>jve^1siV96C|C(MP>e$QX zNd~H<0%-@e3w=}~pzQ=a|Jz`}r>MW4#mWa-5-2bdt`YUgcUp!2ei==gA3gsk5YTi) z%wF(UW)9Crs!tqvm(?V|E?_2+1t8p`@+3dp>M2&YyH(m@!SD4!$}wR|M+f!u888xR zLb@nR{}dOn`%^h0aFo~CKZ<6W#5gwh5__v3=-;h2L8pthN1n`voS)LMkN4&duY9zO^Sl^DTWE>M?Dl+EE+DW$jB<+` z00y30&N|bYnT|o#;PAHA$C)ys*AKym^Hc9VQ#$y=-tc%3K%R7PLzfO1RU6zk@$CpA zwendb&}K|gXhWr<;>QLqX^{Rku^1~VHJK-($*sQlI3fiZi5=`&j`Ch5JPG=)OdDiH z$H=P!ZZ<>BGs%*(bL78~d9>T;(2u;6_^YN~H#$exYkrGZ+5Uhf1hlea!mHP&@;G3s zRv|{#YZ%Q|dJcZXL`Zqo!L*{1!+k%BmxznhEZ?`y^u=k7qa3i!zai1jc5*ltXJw=? zrng;o46rw5wo8=zy)g+e8&YyhCwZq2B64ETAHxRqRunMW72Y`X2!4K@Vs~%kA_W^`6~zitX;MO zWmp(&XLxK#an!kPBR2P%9Aw5{p-uU{#P^~tadaRoq2V2(TS*^)o896y(P)PqlyQM)9~%} zWh?#NYi8*+N=#V+3h84VVdv__#Gj>0Y^uvSha9KUs>CN<-VA-W7&b`Z1wa2k?eh3p zBYC%3OwyI=s^+~tYW)39YiqJBimz6XMR1IDX`|ypo4#AtcFA3aOLXtK-&Q>LlGc4U zM=z7AV#{YyCf8%4nFaNj2RH<;_sCx!ptn<9@^t@<4iL&p z)a93k$r>&`eZu~+w_Y!}r)*8o%{3PcML<1O+=*O97_| zJ7F&gOkD1(9*gc=9t=gtNp;NXAT|s4%P+e$P?G)5CjKaM-`>>$8_%o&UMt@@$;l&c zQBb*A(yAEPEpnJ82>RZzcL;$MdM}eXEVw@D2z!XKJcMS1TNb8lHD@)UhadurA1wL`MvRu z>11FEfj={VmzuuDqPTMA3&zakV}ujD?bz)OsWR8onO?OG(gHC}1@P)MeIH4dLTP>h zS)sWsP1jOn&RU~!Onn0}O4Gq0`!ej*bB0~V22ArT_btvY)ZBmKIU@i8jSx$9P0s9d zErbnkIcEOmey?kU8E0+zLr8m}OfyeC=Nx=cNhUZ&GbgE9~ z0DL>S1jhykd~*Q8AA-;F-LX>yAn}hwdix(AMEg*>s2hok76(s>tposGHqw?V*wA!GYlcZVKsd}K@OS& z5sJj;Z@umT6=I(ZTo+~H!D3g?0eVxNH3cQM(>2cP6OSoUji`RExYvPd^N&`ny<(ig zc~8hl7E7S!Wx@@4?$dUF<*yzu`EWeR$mL_}>|wEznnd5E_bgvJCZ0`4WhH2KIB03` zw{HR%pJFRLqs;EAC_~6)s#jhPYhUC=xgs0I#`hwEwRNd|CoE^4ALPrz@1RK8F-t0( z(gy)ACd7?r4?Crs2ql(y=|BfTbu9}8*y(QedEl!b+gBPNn;#DBCTzBYCaZi!Gsir$ zSN}2_7s^z{aYsHt4`HvuXrf}I9+)klePz6>nJcoO4n;CG2W8tIQC(B5;Zqc3d z?Jf}!DbKDLA+5;)#OXWTAa;=s{-N z@@Z5)fuL8qFRk2(JXtz1IBQ6;k(nr<8xhYJx4R~PcX_ewC5Myq=f+=Gdg}bW6+Lh# z1dtXV^?f1>mf)))bo$-z`_cEb74(n$TSTviVcy=Z&~r+WmImeHK2f*-D(3i2qazR4 zVT{?OAx@XkX(MROeD-p`rO9p8M?q0SxN_?@FLu`@dSU23)T?R7(f$GAd<>w|ZzzSQ zS0>K<9jhHkxQ`-lo8qRt}7}=J=hhwh_6TV!alJN zGfIXo5hg-cj@EBR2{w#b`ut{Wb8QMM659J=2|QG}-&60wzAW(i+cVtN#ITT`4>Rkg zIAxaAJwX#G-KCqHX{l&7ArM2 z7MUzGNDO|hAS6tUFxWRo{`)yZbf{x%Z6khquPXa=9ceG~*;z%HC%Qc!E#UpoPm^$C zArb;iyrI+z?=*N~!@?7Msiec~5D|Cz1~E7a4|LV28rd$9`_(zt|F}U2^0F&K4tTRk z{v!4eB}nl4mKVWVd$y*~&~w&^dOaxb|79WgAUf@Qk(m|$J3;OdVj|Y?s{I4`hQfHP z?ADp6Jn?l7((7USpx1u?n5eqoE`uwRf#vB>zhBUWun0yYasF zmtviJh~ty$XPMpd{0i`4@X{Q;QXtW#7ChY}uR@=qsRIr&6&9YeD-u>;*RSnvZHh1b z>a188PV%AhTMYs%JRLVzVz?(eihT#3(>2|W4UIh@^lRRXtqvE;8ZYsbY#6>jqd14I zAPdsMX1#aXMORPKE~frZK|z1LGY{<%Cie%{4+w;NNVlpK+Xq-gWb4%ksRi+T(eQWo z``$0)JV51k$Sx<<4kd^YY5Yh4|d>O zJ+QxlAxLedyK@~aT`N)`UH}056QwXs?$8%d3l&@91-n0_yd7#fWzhaCPe#b$%J-v{ z?DLAh35fZcr@6wQ8^0awIbhMhftzlxhVOV@1GBHl+TW$#*8u3FF0(Ir`?wMvFxFwg ze!m}AN6VlRG(8ks6S?Db6Y}*%D9I1I+nA{;Fp5& zF!$SFR*%2me@YqtHLtkr8$=})b=`C&7-2mM3yTSC${utkFMK5-<1Rk$YKbzK1mJ8x z+yCL~y`q}xyX|3mC=pbU8k$H^kS-(;ng|hrV5KO{2QVlt5I{Nv1XLh&X`x0$ACyRw zUIH4Lgf7y9w9o?tfk^q}y#F)Kd-aWdvB$pKdyKu-Z_c&mT+45^5EO7+>ZLmZuUx#$ zqeJV@>_|c_><++Pe8+4T^&$nB;GmI3@5nsth%b-LXyW44vM2++rve`FvE?b`njdos zQ=W;g7tt=?kDg-yRW8uR6cw>^v?{m75i18#`gg@9rpT?A^$9$gKD)~XRMg69DkAnz zBm;=nr?WTt0r4t=TPt&>h({+rgNiv28!w5QYIlnCLVkr29|Ws{kz22=j?|Y}dJb!s zGE_=GN{v5SW3ylsq&5*BO0q&70dl%qs9x^?Ij9J6erY^%*#c10=-@A^c+ zHZg3Vw_GJQ4yd{NOK+EaUWafOuH&M2#U%XF(amG(o2`BUUjguzI|9HyS@3D=(vUMR z&}r>}gS%-CvbugbuR0)VgNa9+V=efT6V2F)<49PXN>Kl@vM!B=2?~T8?W@IYYS8V5g~}Va(87YP_W|%`Ll-i$93S^UFPc!K{N&bO|#Q>mO5m5@IG6 zxlqq{hMMe{xf+>G;}7DD=lb4eXahE(6x(HSTjj?K8v0Asry8Kt{H97`303c3a5t}2 z^sr4g_h!3C#?I7ae7it-x*?ZzS+2wb9MIdi<&{ysO$2TDORmNS@n`-txz$puWbGRE z&LY4V5@$z_nb=csWTyLfVo>fB#BfZ3#ud5gp`hOzr3l}o>DmW z?<0B9Pf3=r#ZD)x3SW~JSBUWi0MwdDWVlh>{h*{`nYCis7UDvA<2d+WCEz?3(@Z6BPy>iJ$B+Zb3TQP$-1- zFvuhv(gyCNz(_)*OY0`c3)F&%FK-H%kG6Zg!cgEaih4aaddc(XW%uCAPf>m9Cf@&gZYjqALYuwSpA zZ=|+x1KzyPY607BVObJ%jDLJh25|HqpAr@jlLO#ugmF2^7DNgB6d_)-F?in4)X&m4 zc$N0DTzL=@U9f9= zoz|j>J;X-8U(s@h9}oYgH($tNu|Jm>a4UzLBnlmv?7V&%C9=?T{02yVa8iv2fjTqY zh%ER9Ppcf1RxqkiQgo>~XkWcKI&rJOkf0)q+f<0(Hlw6oSZT{LOZJ<&e1WC7G+}1x zjaS*fcwi=-(UQ9Cglg>THmUjFSpf8S`^CHS=QdVW8R+W0~{dnOLg30M?GC`U`93cC_uy#NU z$P4I5y~K7Q>eE*_1MyJ5!~teST;V2feK4Nb0T+#1S~oIjr~B!LB~y& zdp4h2pM-b|`vB`ZC|xi8TE1E{=QtT_cgrc6y0%k)xz@zR-XSu$FX-(5Ju}TX5A`@N zojAY#-sNl$f7iulC-Iyl`RHz0MVWLQDgAlk$Ef8N+=;JX==BUaCj;it*2j-@FbI+c zMw=u>x={bha}FA%1iQsZ0HK<{F_oQQRaQ)Bzc$168VW1HxsxRh6dq z#%Xjf2>f(SqQC+|orP=q+z))Py*1Yn%M6=YDLwA=Rl????1SvyAIq7-21dr`?^mRL zBN;JXsL#dJ%w?le+n=3k+RPb!U>jRk120M)n}S--hnYXG-JGlxci;-}p3IhFPs?!% zNY8QXeZ?yaXJt`pq781D$^O~L?qtlRr)K6cj838_`yYU+#|6%+^GHhN?aae;JW!cA z;>l8k=b?p86o!DLrJ1Up*MehZo(2QYhfw~d{sHc2qN@C0=* zxhy^-xo~BGXCMQ;iR2E(mhv^)*1yh2@2skLynGAs;G*h213OivBDd^z$c)FWjvN84 zl06z@uc#ffc$wbuRw7_MzArJ1hEu`#w-%@^1u_~mmh%c!r8mUtN+0R#d6dHXCK3=Z z(V&7#jVaB<`89#hJ3-csF`JXT_g|8>ihlR==n2DIsBzxaNzBvOOJZA%+?DBViA zO|bP`OA5zP>*(Uh$}-n9F`wx^H6s-8y@-$2&TJ2Os@OfSc?~_!97Yc0ms=9zTPJ{X z4+(bR>Gxy~b{tfnba%QUK-G-=A$nIhHqO_BJGd4l=y6M(r=!P|0}NZoOC1L8e-;-1 z%p5N&B=tORVS}q}NMP}703%3U^>+Z5c#Fez)WU3eYTpcgnl2{j`^pQMJ!B1Z^pw19 zI~WrWi~peVR-luufirQ4%U=%1Z0qco2E*^`9Sv_lwTL>0Ya#)Qm!|LPm$Vs$BB^IH z#}ZaL>6o(l>*TzRVyk6cB>CTFAI$hD0#+}g^zf^J6;~Bko2Me-hBCFh+)&)fX-(mG zQR%Xtm#Z9s%M7hvF?5}Q4h;(ERc@9ICgz7!S`dDC9Y*sgSQ}Xwgw+mSpkCI$X{=N? zpF)VDUuqw@cphclC?mJ7Jt(bxOvySOd(q#y!_Zs(L1}-uw$wDE+@a>px3iA5c)gV7 z@Bh^p7w6m0UtrwJo8CUqA5WCb{Gak1=x^v_%nsEmM)zI7D%@CvzDl;LpsLvqt*RZ2 zTjq$9dL7<_C>mw?gk4ok@G($zK`;o&6*ElyWxGA2X)o9xp9X{wdG|3ZDrRk2jPH2H zojdC~tevVGmgfsa-~+!|8u4;L_W74(2Z& z`{&<_&Lfi-zRxB}<7rH?ZkR0RvkL3ToAMSVwHfdsPdNSgbRYu!2uxo#z{uh&aH*8o zNOmfE>32n{p+0&J5{vu2%!xmLFLCEg2=`4)1*pdtB#RI01dVJAJeXaRGt%UiBQTKO z`@2sp%r|RLVE7LeOnG?=3d0EHY7cbB-z&*yF#0P{wILSp@8{FEZnI(ZI zEuRhgHW`VbXnSJ(%W7dA$=6iz7GxhfClJZ@2e0Hm!#?T8V;`yHKO3-MtCg7gQzuA2 zm=>imSyO->8g1+#UBG|6$K3?}KC;k{Z;52^%@C&;F?;b#eYA@^X{r zRKWC>cTOmWxwqIg-RU2A*BO*3a5RlsURYBRaNue-TNb`seO!x?6$_kCCM4#4;{Q%Q z!-ylM!c$oN*IVz-2);9<2%y91-8IZzgfR}_ zsEH^cg<(r8nR@J)@{XUD(?y@le+K{kZI)UENXxgO#m%^|Qipu1=s`iw&oti@tBDkW zZJdI=lS84jP^`d|P?+zhRi@YDXUz;xrkQXa%9s3BOf_;Gg~nv$$A=S%wmHNXW#{_U zllfcz z4I6SP_YXlIXfyc*7&}l|>YKN_<(v#;lm~7YWJE;dB}}25!dS!Ui=L!xwR~ z7I`3T{oM;X3UV*^=1e;FZgt-pXxIwFZ|~4g6>|T%;GEwh$ne+SPfp=OY#x)$81O>s z^T6CmCv|0@^PYK|kiv24Qg~gx5DcTu8xYT@Y<7IKg#FT;7ZtaI_HKhL?aBcqNf%=T z+h0HaP`0{)h6}%n41GK2^TB&D?_4BdN9#4g|x55wss); zXf}uEnwF&UkRI@Zcma~Se3hh2Fv7Uu8F8}xFAOIX30X3%q#J1G zZLZcxqR^s(Wj5NI z^?{n`R(%-XS-yQj&*{qe%Qg&}Lc?XhDQ5^iarqQ0?Nw@3vwR$U(abmbSh{Bx{-#TC z{2?vL4%V`w7Ys}+wMeXoW41nAy7^9Va!ZdsSh+u5_L$D6*R!m(+dtIg8+jazW1i@{ z7)cSmx_rwcQ|fnj|LMvbW<9@UtAO!b*MS#x$%sfN-+K-=jfE$4QBj)R3FRMpE2=m` zVv?Zm1;(KZ#Dz3u)!-W?l?qJCHHLa%Yq>kx_Gdg8@in&!F8sxK1Jd0$>A!4V=p_Jx zf_ATsTdNoFxN-|L%9dCFB1RG`b}t+){8!J<-|&=AoT+{?c3M<0G-6EkFxDqNj)C#| zrqbfMU5aIRmv;T=E_Li=G{x|*UxnSL9LVYu?Xdi#iH$gLBXRL@T zftYSl;B{!3!1NTZ1q#;r4JeYZOyF>G(F!FDd?i$re~ejtDesDhB9gv1Z(sbtBNJF6 zvTTHgAe(JYCDH%&fD}#oVDW89?DshZg@OhODD<8H zDD?b((H=!^S5(B1i>Js-^!D!&ci_lKir8({9<2Kf3~%x>jP|w z6eU~Q9ZQ%En(6RjW&tp4C;P4k?1U1oF}$eX0HC)uZ7hOTh4PoMGsz>DfuN(b zFiK%8vF-jBIZ;ZMV;O6>_Kg~T?agerG9f3}iR(lYeY?lu*oS5mPI)hI`1j}MG;Uy)<>pgF7!P`Fff_Uq+d*Ix`7QGE>eVdAZhzIA}Iu-XoNrx0)C z(;UJ=`ePupv=#yv1fLgHrSAf!fnpt$B|*|`f!%t6z*3rHgcoWa>(uH^Q5&a8#p=`x z>v^!;n;2)$M-MmXt!}Lc)e@{`$mPTbb^9*ndnRtNuGV-eABHeR7EAc zd)Cb)RMjXJS%hTsc2DO03{KT;UW=5D{^Io3~9pmf)EBv6JG6J_A@^?P%JY)+S;UcNxHlR~vAIp;j zdEMFfH9fZ)sF3oIxx1kTTzyjYxy%lbKE;Lja1Px7kj1PW+H`koinu*lFvJ=N#j)UN zj5=4jw-h2K%r-3;=Pu5k3t~R{X+Rt=Q=v`UuU$Y38d(44%yVuj2p7`R=qOMdZr7Yo zgmSTGi`T-A=g-@nYl$?}fKC+q)ph4Dvx|vUAr16zBXbn%e~-JtM*Q-l7KEq2nXXhD zUT~Pmja6D^LJ}NSS*ub9Sb20$4tET7)6d-=_JTI2iVrcaxaQj?Glteom?up#ar(I0 zJL;o$Eig3DIt@rJ{$K$HMv< z`At-3za{Vg=fWW)?>Vxu@a*mPOKwyKRpb*fJm};UY<2RE{5ypC}^Sv2ZZ zJM%Gi7bND-&dFZJD-~VdmGn>~oz8z%@ILv+mt)m5!R7derIOXm`SBjKLB%-$=!0sv z%1>JaEd%T+NC#RjS%#}_gT?-kj6CE3OP2pc5|B5%YI7AIjNZ{`-eR<+ja8RLo$Hh1 z2_HK;RSaseQa{q{VS7|Hp`RaSh9#Xqsx-k|o=x^v?7!*l-O>FX`EO; zB8UfsGD0P#xFLt0yn_3AtnY75dnS%gWfcJ$DIS=>0Nf)+1CsyTTy|ZdiaQCh7C9?t zgk$15UcFcK0-$y9B-2in{~vZiWYH2?;01O3hA+&5QKyfOtF`~pyPkv=a+%8BcdD?( z;jgjB)9hBhvCmD!drzQ-RFNaaq4vd&Sq_-i49fU+tDlR)bVU9)xjdeI5=r0S3W2fj zde!;Pww>IrHBy(8B=_y2-j zwHZI{qpfO3ko^v{qk&k(602J0%ah;!0b=UXw|MhiXF@bnWcS+T`ql}CX{A|IP#=gf zap2MZ@P?iGV}Oy=_T5NX?3Q?LY+p4(H|;3eDqTv1o)-VEz14?XcXF#zwBppCA7a*g z%TZ9azVU`bmwl8OF~rIum|~-UHo;8_L^Cfi%kPq(D%I5UmdPr+or_okPD1rMA$dh8$}?!>8-708`2E8u{QUHN#^WIi zJslczh=O2iVq`Kc`z;>XZc4jgPs0|feV3m7a8>;-oB05%XI^2kuVTQ))8PMVDmLBz z8{2!=t8Eack$(G5>)S^@&aIj8d->;Fn3Nhw%BoqKub$t}OHO`_xCX(cwMQ_vdlPCM zYkl-Hf2vgphG9TMs!^;^V*{de)a}54^y-e6Np|;Q2)BK%!qO_vXHuQsBb`-V3f?P2 zIR==2_}=P$2C6KW4HQ@J)MxAqt$oPRMS1)J?I8<9h)N|gpYa4iRwBRsh2|{!r^WZj zl&A;VyQ1zkC;gz@f#iV|KMgD!@K(%+FW@(gufIL>H+D6~H?MLDjwiKlw`hP&DVDE2 zG3lS|UC37d#!G>zBX52qviZQBd=es9Eos#>EZHS%E%`Aj0u zI$QL7jk7Sn$_d7aCEwf|6lLXa{3m{{gNb>M@)W$H9r}x+XYJc$WlwO3Fwh}Mj8+@w z4<0l}Enp3YNNn^w4DHvn@1M3JA90dfnl4pTz{#C?Zn^TQW_}w7#qo$QQd!bD^8$&f ztc;#Pks<1^0)bA2l4asHb_u%}Uwtnclmp=CPc{4u4u(XnFRCDX5*2#JrFuaBQYF2p z17TX6Ae1ckW0LK9&7;qD2E72q!ysp0Tpca@DbdkwB6CGk)TksY zXWFOdsuMJ1htu9Qj-Q^U$YOd#{~WU^TuWyTKS>9~ZA?5|%wdZ3XB0;HJ_|E;xZ-q} zfXkhxU5V8d?)GZh{qgg_>0>`{OQzbKloMiOWjB%-L)wLwpS>|d#);Gao!RdX5VPU; z$;3sz5?I50MI1*J+aVLmMN`SYOz{;gk8TrwX=%qnd8u{e$YznAtm9m!fVRfgJ8!6b zVN!<8Cp|H$qb1_dCDzydqI$y2JE8vZ<=5R9!}H}^4Sip2R?j9*c(sI5lkJMPj_Z4U zMFcnIruP5LsDz)$(_i4%p6xc7;?>?sQ9vdQj+O4a~69P#h_kL^afvMrS_druKTzQoNcJ;d%H2r z&8p3N+|&s1Y|ClTy&T_~sj3y|!F;MyNcsg9{+~7bk=*?}zGvgzu$|Hm19ko_A5?Pc z!Dt{1#l1W}8}=6^`m>ikzf~k{yLRyb@gL>TYIgldu7~7LOs|-Zh>^{O?bLU{0q6C~ zmQoVV9ac$^_vx#6o^f`zK|>dIJ<^0a|9RPOgD*Uo6y$yiX`T!e@I|k%g&vjKFXX*9J?gA7lK4QW+I0=Rj^i^>7 zOhGE7VDsi!sDGDqM!XMoRPkYv&5@@qg}+uqrgtIjPXUM7X-!~6BPf%ZZlBRCA}|B1 zY9D{zbcsH2ILySeABD@KZ*|M5;LWU=_^;c`dp_*Z8QJ9DCk5Kin(b$U>-01#g(&+MiNWvc8XdRBb{wK4ks_{6F4r`iR*PrMH&gp3uh z9-kZ5e3Rl`6kL(9cCo$2os-cS4=8q`Re*bfstvntO;#bqYw+r9be?LVM8L7M7Mm9#7j* zA0rCBJ)!n1IpM^D3w+fhi}%+yFVVmLcz+O&eZB^q8s1CFhMj~%{OiK z(CMvfq9iF|yIXMICl6hm2-;bxCrNgEDDuEa?$>rC84C#I-K`pCv)(WPn#tFF`sEz3 ziu=uZ(Rb3?U(Qu8fO@0PRXyUWUc8n|+EkV>FU;jfmF16<0OCo!BlmUV$KZzlX)F}# zhvagyEp%XG8zS{P^IfTi zUnr(pDLZhx+5o=`vDZR-mir9Mm)d$Z8fj-`xpjJkZi*w@z&^@@5T=OpAV{1B|CD1a z=hB5q@G+}EXidyCTENa+lTyjD$@aDwfggzSE z?B*9hzj5%Ng5mzg|1=mXH`9YG(Vsxaera9Wb=Nw!+&g-lO9*`B25R#~2fq44OD;Hn1QWP|>+@wgl3_1| z+ET8qkDr-OU|q_6K3!Xq7kC~z3`6-1+`J1oUgq9AzIQqA6$bfHbd+)nUpFkP)%Sdh zVmY*3^KXhfS^T z4{#Ox=QkfI{L<5M_w>XpMIxdPN^(C>Y_bW8htdXA@UpVA)O_Kd6Mis}OQ|qX>KK{e z9zY5rRZ7O^oUo!W3fhpPy;Yz8Dw&otN)#rne%I0jB`#VS(bR|<^Tq#x^L1A#Htf4kKGDz?`R}OV)xEI(q**OR$}GABEV}y-te-AYD#n z8C5n>mr=8RA#Ncvx%S)M`KXS!*ntjjLRM?tw|&P>Wp}U`SljFnf;XRr@SAN+Au=!st8`%Be&@f%S^UDaa92w#o|yd>8U?Hf5gr!)IA?_DFH>L^WKON zX;i&2Y$U*eAuYImZGG(@U*91X>@K*0g$|r;O1tx>ybIj2-7O!B3v=?WiU3%ITGy0{ z4|MwN4Hzsc&%bv%L@7LL54ahBwSjFnu6-cW`G(OFUHyuJ-uQ=2JJd(75iaq){`)>x z$H5uHDDm(AK@Z?(o#&>`^~p>?5S@I@@9w2!SUyfQr$TeV3}HIxwP|19r0f66*x=8TWV4PvVY{jeYGHgVO?O z(61v zkLU0|Iiebr`(cf)dz`Ul<1-wp+_RBB)kkp4n>)DDidk}LY%GV?Y^jQ-NMWZX5agtD z_=C#1p1n!v4y)TGG~afwq&v4EKZ{4zQ#4PBoV9bO&djfh%C^H7!0!3Uof$QUKZ{Qm zZ5JPO2mMa**sGd7NMutDa$TM4AzdWs`QWE#PLB5qS0{%os5HA!Zd6b#F7IPcZ)nx0 zzImat$a_uocmI?PU2`q=2)Y%>akN`f_X4NtxcNZutNf}|X2o&Ts%^wK_!YW&ep5B0 zPRQ!5lO~n6BKA;jhq+l0_HOMTyb?{DF87xmn0p0%!;R9sYq`pVwEGz8;$_J5Z?EzD zYJgYjzLZl%#3C~oDuanLnVocr%TdjFlAw*n98B+tu5=6P;%@QodPUhu>fuOXTCan?q4Rxd35$DxiQ_8U)L+Ys`r zHos1`6}zkHI2=Q*pAQ!zHTSK2VxK2%OwrDQMq*3mU!G79^!pP`^nm=sB#0kKdiypf zz*iRL3(DsBIWclGJPTp666*kC^wF{)lWIGu=zM)q9j`*hePSz=-7bL#3YBU!i8(%* ztT@ZkT-KhOH~G+?sZ#vvzY=aaJy0HZ^&MLkhty>SsG%aO^6VA2N5LDxZt-*^fLWJl z^T;<}->zQRo~f1=eIP}T^Z-(%PFUS{VSnrsUvId=U|rNv{;4YyPv7e;)-w%l%yERc zE$tH9G8S>;Muq|HLq7)!%ev^EAF0xlf9d!Cx)y$(VSS>H)Ev%sk!AfGJ{MS3FM`a! zj>e>#)Vcw=_Jo2#4}fw+^hhKSoEospdBbLmGWjBZw&(26EY;g)hia3ysl(pW;V@#l z4|PB+>cw2VLDAJ1HmN8Avx6}bsj56Y$5MVvY{ODrE%KwnLRG5TO+l$~7w89vnEcga z98Z0#=}GnKA2pziFPbX=dnH{-f-(XJN>yEXYRYEy55qwp``|lZ_}*XmS0Wi@9WFRa zDawl|xR4cpFD5%I@>?wvY~?)K3-2OluyPBc<)9A*GwL1f%UO<%Mut09T{4S9+6m8$pQI{rlU%UjhByqD1#tb;D z$>6gf_RJ#?oWJRL2Ql^j@UhE}JI+ATLfzo~Hg)!@8pc3kzsxer1TJQ> z?Q~YGer@b)B1x_|y zx6p4bx;oc+8P8L3{9}~5l6T_z5!kq^Mi*q(vU40O)n0a_?SR^xXyp=75g;9p5^2$A z-qGIr!T8+h(bIDIzL_W;o7Vv#3YHP=Av+I)#?`ov2t(mof_bhlYM@KbJ2UvO-O@ST z$#scAw&^cDouz_*EVEYEA~NiQo}QXx&-4rr34KUV`^H zN=oQtN+y)|bI}g<(c^UnV?7 zTn;I)(5-NDuM_EQ2mQE1_ILE(m$OWC`_E#0g5qqG1pF`kF_%H51pDs>Nz46Qd_q(Y zeWEL~zn~qr7#^88pOe4_qZ5nHqTxHzC+S00d(-j6SQHfbIS!=8=jObsH%_Jtqsv}ZcvE%nG>Ot1QLccU?yN&Hof5 z`w_2_EmbYc)G$J?;vq*9_P{$N10v2V%3BNq+W}cpm7X^ElP@g@QuPj3D*ucxd`uE* ziP!mTA;4l8_Ra(yrH{U*6w3qoMV8Rs&u(t*!Gh-#4t>TisX6Bf<|`Wd zK@ZeHc`T5Vt468JV=j7LRpKTbM-l} zJqQ5(INf2H8I=_qjo7)S9I5x03DR+ALi@Rc3Uf|;o9jxajP~`eYW`%AeYU) zUzEFaCUoV`GvXOrM$J*k;Bsi(&S7k#-OB(1To=u>?yg`U6e2BTbwrdMZx5%oJJ8GW zB$%58ei9z}UkSvprv;QnMDMi@$fD0_pds(d9s1f~xSb_F)@*=thG3<0r*#mK>+P)4 z;gBW^>;9qU7ve2`_DeNHwi7|MKF2;cRWseD`b@>!W&Xdy8{d4ajp^8`dHZQfrg0;I z!K?XT@#g{}HlKVx^pdn8X|`x{{0snf{>x&VQZsms9SQR{XFt~dfB|u=TFYI z{l?nP1Cb03l3nq2$zKSC2rN$Ao(fd<_rHphhgfAyD|-#)t_%`v1b!;7n7opaOI^%VJ&JrnW6F&cTiJD2 zk#Q`N6;k$RddrmI8+S5W5Kbfr3J0(M_pmU5Q(j^JUSsuC3w>yHFQUBtmD4|-ZTZO( z?5^9MMruw`h3z-hww4sGd?p6`Dj@qh)`ER*DSsE_yGtzYg6^-Tc~k6{Mxt`IF(5}I z)s7>EbuQ{odoq(3kqO2fMO4gKd787uDElh;uJ|W~5ev4Guz+367iWxG=)y2Eu@ab2 zmHPM>`xlMLp;+V(ekY7m%$=`DUmme|QTTx+y20{EBQX-Z_?W3LB-?#BxRSMoWMTy z%gEc8BNB`E$CI(0G%kpQ!+c6S@_iDkDqb;Vv)l2+Q~W;c9BP zTUDH=;moMFZ2?0zKg=~Rvz>eC?WuVVAkkq>0CTHv_dVm{^$Lsp9- z&Lg2aqU!@^mI2^w*w}~*H$c~}UC1m9l1>-|fQMM=Jb%v-CEs){*Bigr`3!h>WR*4f z0kvVs3Y?zxz@uQQ1>#4?z}SK;efPvYk$A@4*S$mBJYgGtyVsmVZ*UbrorzOb; z^A=&-sRrR%^3D1dOS@A0<4jiCluvk7VVppd#S-24l5A71%DB*{LHCrbDYq_|BXX;R&#WLE{7}xPD?25Pk@CV z9lfc-drg@3x9$*D`J3to5Bo@kS#S%+Ml_;dCC?Z0ypu- z)|`X8D~T1p21lI>4W<;v#%vTIsFfA_AWZM8?{T^{jsCo$6*}n&+$%RJxWuhGw|%Vr zrAPA-jbsO88%J;JOWl>-Tr0FGS|JX1e9#xp9Vj_Rq%{B8(2kBOzafIY;h(AHf4kyZ zwGrv6B8MJ+RUMYvPqKxFYGsAB0bKRNqqPsO?(2(JkSt?+7PsRg6E#{p!s%~yr%FEM zcr^!7+=JRe4~_7Grg60tz94bMaU#mK4Kt$Y|IYi(EDK=n`cmP!@hC1?-(B17{EeaT z!%rSgo(k0lC%=-3ga8eHH(lWy*?RwHl0y6sWI*fh{5};g&Eg_VbeW^tWlPGsG;wWj z^U!nG#DVJE%Bjz}nMk7WCoxQb%Jc-8ZA+CmX1{a3=QN_k)XSfz{pg6Vk)eyGPdv0t zcxciU;KZVuN%+Cmj#lw@w%PR@X^toe28!*8su>GO1C4CL1t(6G`H z@5zD78|z=8qv7!dNap<6c5t_KH2zOyh9$-4w|9Mkz(Wq`!|1DW0y;%HKjYPrYOaJI z;!Yzv81ooAjQCP_=L)3FL3|eN84qDAE??bMj$cbuDt#{W8p}%kaMhRD%7IMIbKB7F zvx#^Ya9y00wYn7V>N#>cUI>uRv~PXUizF(&)Fs{|Exkj6Om#8=?r*2GM=2yb?WNk# zUja>M8My%%Y>Q)3IOhENkcp+~(5Jt6BOP9p3h@LdL&B<<`+TbyY%vwR-xygoTr%u5 zA|*(WCU79xXlck4dk55uZNe(?!&-H6-)-_oCOyUIXifjaw%z;GSn~?+RDWmXbqDgl zO4Md*8&@o|NJ169rMt}`nD+hDN>RiRHfFl*ez6jinD=NuGh}XRUfAK*anF8Ae9&Tw z1N60*&OO%cCYd|rMBY?rQ13rG3`lDIrel_x>nJuF=(GK7^t9x(=6-cY1kBkM=*gMa zJ`jTbnC#d0OnH~tmLLG`5-}}cbB~MYto7M74QEYg<F&uZn}Kj+)85dsNd zec!1g>?#flm7W;)p9<9;8CGji)_XfdQQGHJ)!-q^@ei|zE6_CZTyN_gW{5D4?q#=n zl`+Mj#{8Tda#&dgfa~Dj2_;r-*%pJZy7!?HuU-;7F&_uCW!XRt*8!+_@R^4UFa*leuhxmP`GyQy6e|bYBkR9-OxRW z=|u4%m3<>s4ST!?v?jw*UMw|a=5E21bv?{;>vdpLxdwkgg2IsK{3JLqR8v5^WuS0Tfy1wm!ifc z@V@8jBn%g|7BA5f!`>oRqP|cRP6moJmQ{UZ1G?F0uN~FB^@Ix(PJFgE{)9V6gT|xR z{rdK2YLi5QwlLZof}dMc$sQdl4V|U0SoN0ORZTkR{%ZTj|ecZkSdUVhy z3k608RV7@szXHjR2ktLl(YwVW9GPg9wxe?oZNeqq`N^?()m;l^3N=n_V3S=Me`UO8 z4V9JXd^2gCXQabbg%793g%QmL#drI=QSAZ$mU`tCFTO#>;J;8kI6k_LR|jJMZ;}Ep z>oFp*bE7u}$~zwz{vXMc;X8)iC)ax;9;?6IFsm}pLxp^j9f+4m^?);B&bq9tK4wQM zD{l+o;<>YSu1mZwZ2#;GcGiE!>)tVC_5)%fIEUvhuC#2xh@)I-Jp`W-zNm%~75Zm|!ZVj9)dy~nq5O+Kh5=yv-q1joM5~8S0Vg1BfA_SjNXFlP!&CeM zCwbMG?h9^Ex@_3o35w-7n2$GAoRaErez3T9UR_px{x#6Mtp8Edt<=d8jXLJlgQa|N z)w*)985m)=cT2NkEx4H6kH#u6#W1~^Pq!Tsb?o;-R{7*}53(U757U;AJE3x;L(-TQ z`Uqb4OA{18QBYolGXZl6h7S1Lo4a_e9q?tO;$JA#rLd%8 zb8S>*lhBhF?tJI2V&FI2N~iP9*IssWYwECH?$2}MIWjq#1E%YqaO=xv-V1n_aslnv z_m4{_t^krtw)LMXlGIiT={w`;8*8M~?Y)iqHwCf7+vm)o3;NP~4jj`}SC6gD+1xl^ zh-hr~7UfHZwkiu59& zf`GILp(}_GIs(!`MMZk=gdPP#Q$i6!5kgNOKnU&V_dWlAZq9ke?Y`N2?LF2QbI&y< zow_C2tonG+cW#!YedQ+MYPK9f-IgItrP-umz}-!#(@I)JSC3htP@NKKM&|upB#q2& zSXWnTuAgLxp)TN55$*uDT3xxgVFu?>{gGi3no$_C)lC~pL`$YT8~%e}2pQ42Xk}%U z?XG?9Tb+$9it>?n>+(flRQSeIB~24(5?}JDF^hzA$hH@Rg&eYSWM>p>s$zg2dZ{`V2?zkC=vNtocU}us|Yl^I_aeT&k5<#x*}_s$q>~|9Nh5<)WE+Kf)m*(ZQT0&6 z7vK(X*??v<&o3yS=v!oRhx6M_{{&LgZ0Mas%7We07M+7`-%2v!91%VCyUSemiyR-F zK;dFX&IsDJ24tFiC?#FbA;>%)d^`}3=RLa()+mL4e}{in?^oHr(6MsI{cA#QnYT1z zt^GsZ@flcI>!p@jT=PJ=ZazpY$?e3n0W23Zp*==j*Kgk(h>F33HP`tazwIZ_@=5Ns zls!&<6IygROEZW}7y+!jTj@BwVjAzwYGg_iiJ~a?h`_=J(*+|9a-!Ingn{zb{sdO=K z>Eclkt7oFaAFS7>r-=${uF?u@D)KI+Cq}=@lmO$)*@aHK*d~XuxwrnQClWY`_2Igu z%!Ml7i$(|N!K7&4MyNNRpLip{p5{lwyRBaNEj68RogZO$$_+ZwGluzlr(@zaZCa)= zBkcB`x-pOyHum|44$?l6P$NDwHuejHD~pyojDPq=pV2?{yxok&=@_n>=dV|s#|Pg5 zih_*1i|D62vz-~{yaP~6w?fYAqLlt4o16b<`e-}_n#kEZ+Uf2@6X`7dv|qJ3T+pQN zci8D;7kQ_JE6_*j{ct)IJ_1vOGr^Kbn$$eXtrO= zF8&1)jS^yFl&W4ZlJr*aRT*tr_mUni@Qd&bdbdp<7kV!dSS=OKZQyWGCP5euBNp$a zC96Ce(eDuapYs?z0HW%g82xfg)Fc#m9#p>Yto8b}IMc_-DM(M)q1fv3+Z^w{$h8j3 ziw>0klLgSkS#o_ z=t{qfKJm@wMq>E89jljT@4K&-YI@h{DtqUo2DeNQ#Nt@Sw1^CPBX4Y2w6jcodd98E z5ijFm!7Ii2Id=HQXkNF{tha`y(rcF1ae%Qv%dj7H^`TSDg-C= zOx3ic)=dWeS}!mE`VZT6N>&oegi3;U?GeVK^*Dhr=NXU_wsTJRv_CPmr=e0yf;6wq zC@^qIRT;}~t@zaI@^lwl*i;?6$hP$>n$%%D_sJ!)qF3{(RqLLSaQd`LsaadzPJUe= zp~2Vlz;V!jWXy8-2Kw{h@67Gi-uMV_ukXF8VEl=)j3`Th;aHsC&Jl}*r@Ou^3bhX%1|5XDy|}Xd zL~Fziin*+AHnsZDdAZ3Lyt(3>sHwC(H&fc{Pc)yr0H}XGS1-ae>Af9qfmpv+JK1nd zjw^peIfQsk$!PQh5+_PwLaK!5`nhdrz9rv>RPvP5C;7)qgi<-t+od(~Bz_0tS}j1` z4@c7_T$|~J)L4a&UQ>!ncyv)Qdio1J+;hPJ8JIon|2c-yr3V`K)B}F9;F#cWj*fw6 z)sR@X$gr20ybNotJ`)x66*#WF_<7_l#YQhSj0mg=`zx4!9NC>e0iS!y7!4c!W{*(s zls}{d3KA8R;lU8Mgz6g6l5A0#$}~Uk0R5FT>7^uAZE+bDatfz-MCZk{<<_W=3@yh? zD`ZBY6i)o0)|X=k20iY5yp(sPuXVMyNQ4A7q28fC0ipFqr1Sj!MU^zfuli7yO$K)F zD7VISuo3O6Wj5Ry{8fQ@C%Yv2EkhiBMaur@J3$Y>!4G>dL2D6fu2ILts7L!>tOPRs z4d*7y#KNaABBYOYZhI;H!2Q1w{G(GMn4=!fMU&SgoJ%J`VG`2h)n|#ee!$c7NSZtt zvL#5Q>9%qZ|1weY8x+16M|Iz1zt92j;fhP)4yZJIQC6+uUFI#)=8`BTF81WQUrxec zER%tlfqE=2^>(6|MeC3xMyo@dHk~}5B|AE^nL`!#Y|H8?*MDP0llCg*7RF>W?0~~2 z7Fr)aXI<&wwfTlId)NYptSvNLe57z4%%aq+arlm<%@vN5Y~F?;xZrahqK-fyxe^X6N?EQhga)I50$<>PDn}H{^DiwYpzhx zR^rEssc*rt-z(XakdSGYKQb~wKQ4^uDR`lc1y9EunO(f@ja)lR0PvhgCFbG-iOLjr z586gxtR2S5z{>M4gEYza8-5-SBfI2FD+qaHXSH$?_-D|(XmF*>FH*inoSu^QOT+uS z*7LB`w~{hf9^FWF?_qARL#pHS@~o)KZ4(c!Nv1p8=BPosks6=nMt z8Uf#sAWvJ1qTC%iy^}*%@N5VjH(UrwM6*)R-4Sr2VuCMic6?{{?{A%2_M1h$Y+irx z*;Ri!QiK({lxqI83^h3QhbScv)uX$*asZ1U1`H>7!#50R5bDKB>%!?Zqy;>bxxvR; zohS=Fq2y2VhdCc^x9lWrtnlpPzX%Eyq8*psG z9Nm%?Mm+?dDhU!Iu#rM$R|$RR23_Bzt<1-j5lsoI*?(iFBH#PtfDvezF^i~ z7}^lH7WfHuwu&OtJd?5XAU@;Qg0EWK}I8mXO9zs`tXhugZ9H*-(a# z%qKVDjcoXf_6MU4N_6%imprJy zBeOb~r_8h2HCN z^um`I2QA3xoR1V^y}=>BbdUEHgl>(BXG{FKe#2px57)nNfPQ~kT^1Nb(w_Y#@LqrH z;GdOHwYwEjZa#B#sAbr>(F=2@`Jw_Bs0wj{7tXW|gt6p)^a7p8Tt}NlKsCr;gguY^ zH)+C5)V+Eh)8FkIK5vNi?1ZX}de!!bQcXKTK-ku>e-qvW66+c2uy>}}dh8uc#pV~v z+H_64x{i$U7-ZXv*((-G38LA{frqxj4FzqH%2XC1p@JXTQpmdjDcG)#8m6Z{3()+d zL3er+6@PZq-cyQ+QoLMylSTgaMz>P}yBil_^E#M`G2aJZsvZ|faYAET!%OmZdQL|>1lS3!z=c~?-r zGqqFggO}5X9Gs^t?DZTfW_3EE;-Bsa%ghT)^3?~K?;yar!l#gsUUf=yQXaKs7 z?$+}QelE(Ab;X&0=Y5}udwK9GWK5E}meFKj>sxrhm@WKOwnDOgSU-|Kr91K!cN%vF+I?dP9Tj(BqBCxFj&V6WjyIo?;&`OT;Rv6vndm z@M_TLXa~)Q=DIXlHrnSxM8aHhzDczPCeo!5V)8A@81(QM_C;n^CGcG7j*;Q*SqCe0U%c zjh%}o$HHrCUIxzk^i$Y`9^o;0#1QG~!u)d=N5d;w#u07+xLbnwTN3?v>b52@hGdf@O$&86>>HTR=DSDz(2FZBv#2gldlZXXiguIj+=VZ?sbc^1CT z3HlX%laee)l)e(a<;^jS&$gBplkf#qQ|hNvIQMQlhs)jlyWQ|(iDS?X!MDM=w*9LV z0gsNlhb8qtf3~~P;|@v=hL?wYW}Vo`>8;!2fV|bwKTV=1L3tWq+_rnhdF>F6gap+i zBGEfO9fwK~2&{gxDV;X*P|Y=`#}}_21fI(7>>J|aPt7t{d}TrImdt4hY;t>{WghZ% z%+)Ml;RT3cEvDXs;{+-3kQ4xZSxs|0Y~+o3eUyyj4rDt@>sPm+#>w~jyBz~^S^EG6 zdA{GIJEhI*%fb;EIhEMuc*dro=?2~`;*t#&)jn_~3(;)e|0MCipEoF!yzee^|B)qoV>f$gYxaKI z(dMIkmJ++Qh5&%#Hmif!I72Cy8TER>7u$%40J;kpW^|QpUL-d<#-+vK9v8bzK;paE z@scrq*@f{_x;OWIj*Cszp8=#{qz@d8LERik^d-La1>yDuR8e{O4=*&&d$DeYv^kfw z51e?WU9tXz`kuBkfK(`J&|QV_Km3-|x4cxJvOIk4f4AP4YWgO?4gq?o{Q7;t_+c5) zR}N(@kw*hrG}vs@Gsc+CLG4X+{e~G)R)4&NWb@ZxJX)e@Gb6Bv6#06{!o05&TQ3`5^ z+|htjJ;igJ45_9JpPBMnvkM7T^88&ClOLV9GBgR0B1R<^qSa_@B+KGERp(dY5}V6NHFnKx;<#0zn+)NG?-jy>K6Avu|? zPuN_t!xj+%{H@1>f=NoBY_59X^HxFULawrqMI5+ix=}bW7o3j>`nc**cYqh5uN3U- zr_=;`WI%;8z4{KT@R9F`Uv0f{gUe`5aeLN|W()g1%y~{NF3Qf?X9=YYY_FrGH=Xj{ zb#mkZA|AsQTB5*0E8GTwh{~$I3vEQnz`k6Wm6E24GLm@qxHcx5}UKGzQN>HAcPmE;WvBLOM3& zRL=9Es<-+>pUfWa9B=Wf;X}lAgZ+x=jWH~OS6E@`Pak}kP_!@g#!RiBYrDnip?d1I7YnEdyX$6M%ZhM78_9n~ zkpITYGJxZ+5o_0~sCQB$30n7TCoB6+v3)Er?n*c|&Xe+f+Dc%sb=sJy6M55s6s^`j ztWKP6nJnVTazbY$6v|)zbJ{Jf2Ktw z$L|;U`QRsBgfB_EOE>m0F>|z-sBUu&MjDwYm)JNg`d*0y$B3jtH5t7k|IpnAzSPQj z7BlwpMCH7MrlB4XArvEw1lZ6Cw}`KX`mU(%Mq27&=xJXm(J`UlJ7{+T%CVY99b73_ zVJ}~md&_pdW^GmLz7~}8bI+(!?DOssI3oF0udK-pK;X~xd)?|-qNxbIy;dHnQwN>d z6Mt(!(?6i)H&ZQM?jGkXLfwpbA+U0!;?*&dz2BQJy&k@mzR`c~MRf0EP zF-YY)%oA$VxGt@^QZY)EKRvQvz3tw9W+0W%L?7pw>QXIo_2|CB?OpSf{>xe2ChK}B z9{@Cy9__bTa{X+dtcK(#UdS-s1I6m<#)2uCa!7%_A@#E5*H0K9S>l;#f2)rN;4GhMGFi4~`o^d}{Nl z%`8lKMJY~%sr6wjZyqMe`EqBih#HO=5^9>3F*Zvm{dOpS-RB?m$I^vWigU{kgX3YUo6{q0IbeEf&k5R388I z&_v-GVgHrt^H#T?8?zy+hIL;|5U+I(bE8(-Ok&*nY%xNARZ}H#s!e4$qD83u1A4%b5wd+skFD3Qxu_xAgL|xNhOBYlG|zIHn%Ar&TO%D)B!KL!!f5 z)ydevE=i4>AonLa?+(_NN?|xhRy=4#(1R&w)M%0tLfLSD_I{oNx@HI0JhBIoiIH9a ziy#Km{azD~WaVB`07-~vtesZOa=3`zEs4rxk0c^qMO{on0EaUva#*-G^N~cr7AGeM=CbH z&R-pU&?zf2+5gWGC2xYnn0Fa?Q=scj^-&LRUFE-mgsCG;jSO{&i#x6-Hum2LaoP>eKGh z?$VFE&vntBI7vwne^K|3LV9SM(L$r*1`mSnuC%F2WohXV=yEHqZgq*ZUB9NYTBEZ` zi%5Zzzk4zBGxHqKvN1au+z1%5;$pR7wmbh(Q@@Qim`_RYxGLz9xVd~_6nX#evy#)S zkF&P+64C#vUyjgu)s`}mdTHA8;2~Fv=Q`?A$HG<8_^c@NM`5KX_VkOX%TDw%NO!Dx zd9?~r%Qrk;o$y)v4(4Oj*-MF|yYQcd89;^g*_^L=^%xaPv`?;V4YNbm-cZIctizQP zwLPYJ#!M+cm2(M=zg?|uPpDz1J&O^vT^1g7ZkA_FRmqO2PfugQhl8_n}wlUbfC8<`sK_%PvMTAL1#UgTFi#rQQf)KsEGA`D&j2 z>WeJE2Bzq}Sja)59fdNy*|DLeflUD8%@}}I-27QO{XDHKUNsWF6afcP7dN^8c&v`r ze}2pj(5$tJ03KN|5Nd8PniKH@um!mgrZD0Nro3LnU5Cdu25dKW-=CbXQ2@}0b7etz zKM_ss{aO)rxcMiXWFFt?IDOE%!*P7#>l>#*HIul`)YJRY1+Pmp--Q8` zsm6L!VL`dGh-0t^(utK4u-=_&vX~3d+U*8qtB(wn7iOIQl$;ZOxtvUzv3cFU3`Ev< zS$5JIRZC`wtngISFr%!O!X*#{UTLK_nvTvGX=|6^s1o|zAyxM%QO%3a-4QTCO;v;; zO!lP&uQ#dh^UI#gCX%APi-ez(|F0Iw_#Z8ja9V`?t3`amp+lVSW*3rX3@pb!h4(#I z1J>XR3f^}AzPW1U@X)&!dwXBD9}SYe`jgR=Ml`7^`{94O#LH_UoqHH50|gALC;q`j zsWlPHi~8574IB3&mj|<;q6lt0GavkP%`{P@R@cOTmFAMiW8sdYyn;K6I!?nTPtb{Y z1~0%nz(Wv`!G;4y!81Kz(PqHoH(1%-wOYjDdJ?-WBg%xP#ZDKzknyOZHaF6-lkZ64@sE=p$5niQBN^rF&EsWI2TxzIR7HQrf;Pv^{LjaBArD< zgLpo9Axm&ZhUw8nIBO%B14gB{D2h2v!=-=T{T17!HP&kqirv!AXsjvdl!!tJHR4<% z61dc;Nj(wtrvdCj>atd;W%4olX;xPsXNyOP9pZ*Pe$-vzjY}v^rC+(MVC0&r`a=C! zeaOmhxYadpML79o`x|Ih*$8w?ToO&r862-$QK#NHTQg#fTnX}D<7<0INbku&Je1~Y zmW;S=C50&sV^zYB%Qt`LsJfxIGorgHElUEo-lqqih&`BM42Z4(L^oWRVuZyV@rQb_ z@KJl^>bso^8mw?P{HnP(V^^o$fAke5bF3)0tehw8F4g)chL|y)hn_46C!&C7Zh67& zU-4TYM8nl5-zXB(c2DHn-Fr6fI&bI6??FsGBx*(;SB+hPypQLAk53B3KUw-J*h&KJ zj4J*BoP}@11Gk7gnwuTHiur81V2$|VXI`4urkfe%Linv!@TKy2(GBckG=OfYhcT+6 z9sgJ~R$lpF`U{6JB31QN0IpypG|;5h$2u&d5&l>xiu;jFd}sb?y_PPxdZ0dAazNtP zn`?rVJ-l6^Z%-x-!lJFQDfe`WeKF%T|rkq_Od;1XGkUCLdKD# z9Un*Q^Kax+!;_T@%QC-RcpZcwY*036~n5;Do$uY)K*a(D6L1_Z#~_@uF0Wz zV8TepuVwOdIKyE0{EO>1OnO*NXXt*kNQd@3BKlJkUvNlJg7ETo+uEW?!N@FuWDT`r!i)bAVO~t+uNKXdzm&3@z zphTe<_R9`gZulov-;^2FTw3pAwdJ$#G>kI6L!iF7$6;55m14^r)$xl2b}a?W9lU%H zLBl>Rms~Hd;-npdPd?HPsghCaiAVU?%G^tN8Bo~pVbmcW;wx`QP+q}T$~!E`2R}7g zHZ6)R@0?Vz=F{~$cLpbFigIjzJG1!zuEj>u12l5%7$qNFbjpyB%+ncq`Hs!K;fo8$ zswK_+8=#ubjl93EW~v5wXlwsBf=Sd z!!qR(ov>B~jW(UJ!uRu4ZUmD`SA`OKAKdEg*AZL}xJK;X` ziTu&e4lnHoDLwZH8!0He#naK|BjQ8Idkvs1!2Z4zo-vRLx%!6vmzo?f#}D{HZ&UJE zMC%Hzq@$x&{BRKW8_)F0f>fY-&(sB~j$CcVgj@a;-Imw`oJY)u#C9Bg>}92B!HoF2 z@B(0niy2PLe@$Lp`*7^#Bu{{NOYebuN!X3e%6FS|`E5}puVmHmZ$YOmvTF?Gp<lWVpjbkVU53jMdtKW_cjt8MO&ZH%K1y{j|#%V#lH6HR4!bl zErqN72wQ|#(sKRsV|&gul=263X+cYJzZ8fZ-Dnq+pA|D-$Q^#)fB;QEvcI+<$6jEx z=S=Y9)AXL`)1Rqc)y{Ltn@PSD^e* z32G;p6}__ayM92p0Z4FD~T&l7@>+}I7 zwgRoc58uoPtwjg7mv%>%Le&%8@;2yC|9=XxXjUv(bhXVRo2a_=7BQmvf+hLB!pdCUB#pgk-5x6P3n=;42Xu*vb^r6RBCy)r&@ z8eY+{J?%_4lk=ek25I=>s-)NUht+r39&Nr&zsG%xg=Xa8{DPVF>Y)@JvsNtLG|Ba{ z5&m+-t3)oTxrCaT3EBIUyCckR7<@@yz}rNK&3?|Qa}b6vA`AcDe_xefGE}e##7Fs( z!>4w(*TOo;-wuXJ4nqB>0mKQ=xh-oKa`#dmosQO$8~k`mJ|Kk$!<<*~uhZm(NSnxc zit%RR2M%?g$}QK7wCbcj*U?-*13CxPG4qzi~zV4RQQl!{_&-`HSy;vRR1%TK@QxRn52>Z`D3cz z%#6f3^=%XYq^1)_7O(n3NGx*<+&!8oHd}GXPL&oKgOek=d6RSQI>&=e7 z>{MXr)g0?g5ng;maIN?dN)^|a8BemB3S|!6W2Z_f0jrNe46CZFnn*un1pJ721ZTxj zp_11?8UvRaYha3|1sQWH#7iLvJw2kZrlwBnGq=_rj%wXcgLQ~Y(rz6Xj>fkzG?3#^ zK+k{m+1Bg^fgj0pC`D$qLE_{HL}@Vn7{L3^=YvOJnz;L&t1sZ@^dPB-agBvqf`rB?jICn;5RUah1)gjrfSn4XCL$U({hXd#al zc{@~=QO5N2&o_&XmJIEwV~r*4B)t0D`y}iN7a+mi4=-c23SDP30=xH~qV-ChAV=lX zQhh_M7qC_pLF2#gaMg3zV>s{?jMeS#XX(8W|BG=` zn~L{!6%nQ-q;JZWG*6}>p8*5Y)l++p;i?bZjI)<<7ju{uc(f93rR!`r-9iN9vA$;z zrvsu>gYMD}P)KrToo2xl1C!a%)TznA@TTrCh|=(o!!azY4LFWrvF3+PzED40}K zDDb?jVNNaJyRb*>%YDd#tjd?psT$Q=?CqU;NtYWR%w~*mKcL!cM7}V{;F6e2*$Zpv zoB(DHzKAsDo67-x8Km0C^yoPD~H&vUV5aB4j$cO-pki=RwO@<<}_s7ln+ZnJ0QOD)1o*U|alHR0blFtD@ zGxS}9#VrSA{2<$c?t`B`yA_6eI!xBqHPV=BXbHCxpZ;ATjuiT}Z_pzn<+ZqC)pU=5 zq{lCwa`tM5GTA1Nh&N!WpnSvMHS*Wv=6)qMJ>Z&@9H%goGlZ8>tZE-+;lop2u@ClJ zHMQq8Yr?yq6W0j&=OL?uXZm@D_Y6OE>4fY-gKtf)T?xXmg1K=(C4;WavH?k2*5{kA zR}otCKdZE~)G26M&wt64gmGUl#iK`bc|m=lc+u&FwLCG(XuW>#>J!Har#HQ5kK0R% z$5I(mqEwx*gGQf-)L`#>BcqW8y^jJ|IADiOSav)99gmq=pbXdE`r^&AB$>uY!Qt&> z)>{2H))&)0^h=)o`BzQc%qJz`SBv*#P*ty$oQEZ>3!LLZU@4kaAT|=E=Q7j!1r{_KVG<=>&66KgfdTm=7E;d{cMgG~i zS9b((;1N|>;Cy3oNq0NEbW zID>3MPl{snwB^KyQs4NfJFv`#5SgB%56!PFngdq`vo6Q^jdsCKtFk8|miwvn6i^6;^iEUNkuB6duR_C}F zKXX9}iC(6@N{%3!O(77CzHQf35WArYSI?Xe_aFPbk3TvucFBi4D-~7(gpm?F>-142 z1ohy9O@go|#&*dN)>RlxxedcVuS(l79-cC29EDSes)8Pnwkt6%Z)Dfp{$DEk@sEmv z`a|!xt9Tcwh=eaRzVKe{k_XXvr>QnVLL&-j^qqc!1tYtQWiVaxs~U!EzPnrAmu@j# zuVb{&&LKcvc#_9=)~RJoOk%N@uq@u)KQ*_0N?q&yo~w!cm75(I0uTBrNPlVaQa9RPKe= zUUesdhlhWUyVY7uw)I`hnb51LM@6Y=2F`13<0Zei@O~;VNixZ6jb(<{h8MzXuesA> z_x`t?PG6-G?7?^Bx+JDap!t0t2V@k~S6*XO@3edQN#=N{UVpo(KiZNZiFkF*34QBI zk4a!ib?~NOn@b!4j`aAPv*>zJL*v^m@kaF#<>TQMlFQ!k= z=akY5Ha6dNRvl++jtLE~fXETx-{QB$`|#3LUpNe)rv{BxahnWw1-wWik#R1s5cQj; zBSmIU9sINAFK<#TAO%rFsl#jeP^ zSaD_RG-&=K2P3bw6tNN|zfK7V^?CdLAe~Ng{mfQ$=!kJ6F-&^ea<*7~J-9S6WU+Kq zr~Xz*(6)?Ntaq$ydj>${fNl<7b#o=i8*gknbR;|nV)*S_2_U8sx0-aNSoQpPhdfx% z7GBQ-+j~Y=N~br&1CZ&B(V_y2f_ydSNiti=Yw9%i{^XC8t#Ca=lk1PR`;Tcmi$M7a z6@O4t2dJ=yd7wsZr4?ma-yC0T8lq@ImCu_a)OfpTT^ypk_#x9^W9d*s>71)R)8D9>A> z*O*}ZQB?Q(wi0B>Gd?F(;0L-67Vaz7G3jiiw~72sgiTJO?5M%)Do zc=qz&!=U&Pm-~0}Js7Pjy`ArSHcY(% zd$s!DY5EcL4MYLn)$`9$D^fBZVyZeZuLHa?(5NtO55FGkQ_AmTw)w#MkSJl7Aq)Ex z4y`Rg%F4eHw9`1EIx!6?zT@ra+lwY^+r)>exW`ELeUZ1AztRA>x&JSo?(!a`EP=Xg zVnjaK!*U?{F17V-|Ft2loiJx78_mh#PmDn^Ld^iV(-GcQZC*8;C+z(Z8mh$UwZ;WT zoBo$^&SX!-0FGV=8*RyqWvhGO<6WBx0_>ah00aM#e$yzq5Jh;nTwk{t0CqCohdFWZ z2q_w${a{<2Xt&F10scPhnawF()HUiPdfYW<_pJb37#ef5LrV zrRP}C0^J_nv%KNhUv&(#xY9Uk(ur0D-DZ-Q&c#g0c#knsa-1e3u}zI~#Ia9NsUCp^ zZsmX=2-N{}@maqQuOE{FS@d2w7}@$WuvVP&U$3}UHvZS}?O2IeF-n(#a3t?(KfOxi zIF!{dQh?kAFbZK}s(Wa9kJ(Ktzdn-m_7-H$JG@3`*XTTjE05XcB)b2v`+tqa6d2PP zCas=>B56(iqzJ$4)F*)|`)o%0B-4V|AcL@2(fmnn_o>`9)f`!K0ZBLFDC4JMST4`$ zwRQ870D}3f1$Gn~QOxkVL0MdaXh!p9^(+uTS^w&`x$u6F!I#E9>@Qrv+5ON$mKnjM)9&S`k;>^+y;UM>`M2v^?zVYh_NOjmuyk^aT-#}pj5=(l z&BIOUdZ^x#njFw3p5CRbc-7@|aq#d@-|`Xp>NON26@OzV)ilVWkk=(mR;)O0sB*SM zJ>ckW7C&-%YYr9%T>S0d$K2a2O(g78Qr&2KWCY5&gP_DyHGJ2kDR@Bj_4Ew9~rYtO^7hPgbg!_Nji{Py7i7fV&qq-MRydgyE zi1GAyh~pUeiDN7K52fTM2n*M9$8BCAlA*eq11TF4Tq<|kE!QaV%;7=2)O%t*_Q9CC zzf96%3*d@#4A$M)l$4JoT1l&@j;kKLIjc(UL-x6HySIeg-ucB3$}Z^EaP!(IdZCjA zOWnXl=C;-`Itkd3hZI=bbRerc=X=7-mhr{~RF$*oggBJ4bi@_)rt z>d-$iRo(C9*rDS6RHZYEmxjswg#)XsuLIyNtOH$`F^EU94g<6z{SD~jo(3rUdILD{ zY9}kpFTV4H{2>RGdxzR?*o)vB0Q$iCbN&K2I-yrxJw|m_bTIG0T!6xy@*Eb z*`p1K6LB4Rg{sqkqeL-^5IeJ1>I-2lZj+1aB)3k=Qwv40*Ytrs{69^u;aVE`s?1s?H_q)oU6f zykQ?Q8Wc*0o>p4HH#f|i5+`5!;Gg1ey$H6k8Mv8ijIApPoHF;Gl5t}UageNaj@leB z@tV5oIdwVqju%5!4G?3k)>mJZyKS)e)uKA(_mj!v=Y}gT>QhX57(r$TvqWq%mpt_y z!Rg0>>*b5=DT;!`u?FQJv3^`LH3OH<*B9rJzyN{_ai@GDq*L0Qk2{=*_vKoOv52%G z#~9KbbqzlTaBkic3%b=t=Z5us7%<7S-zKGDWi0PuqRp7q>o6yww9UboLUosWDZi#p zzJ-@29H6(Sp*QxmDhO|1riboTh?&W{8YCzBT70Jd->snFpPq>N--Nenp@z{i(m{y z9LATYgiz@$)FMqC5TZZpl!#)!YoFv_C`vwi7BN5jW7Rw#TUOTHI+&2zvX5=J1yz$a zw-Q^M=O@tHdK<|K1n+um5+^kM_;;2-WPa8-_~u((gC!NYlTI*-@{L1f&D3DK$I5}9 zdua!yrV_(zL%LK({0Pnin^lsAijsyNf^K83{~nc`Vkn^>nill*5ma~3_`SE1t2B4< zRg6l5HWoMEFdrI>Q>{-p6|o@RJeW?R-@@P7!QIJb!|5SfOh4hpe?t9xFI(>bxgt)Yh=;FjUUOC`z&(={>(=lkPV@{J*g+wDF1dFUB0MSM(< zcW+4_X4QQ(!mg4a-}j$-2jA6`1Kd;EV6S-SZCJ7l>@vc|~)bFKruKdFAn`Hz-P z@15`Wmw%R7Z@2Pa%@Q41?R;dkWMg>3LTyvK=F!^GvdoSRkqNVY+WMioi1k{gKKDT8R2@F8 zkl)AICn>WUqAE62;U=P5tjmb=vtLn8Si#j0{$jFuWkhu=JkRu}siO1Q5larouc}O4 z7K$7pYo?y;k;Ht=-;`Yx$2z}rY(_pMV*K2o7f$N09*%++1IiK!N-J}+I zH!jV(kGz`S%fb(nUKXaF&|m{juu9c?AdDY3h{@gr1s+iZjj4zlkP&Jw? zoO6X`uyR0Ls}WY@rWY_POjuFv1xtr)JDDJ`silfhV}^U(lL4MZ`Z`aQPB8sN>e%^1 z{Za#|mSoLcsK@S}W?IFOEYRu&YxlyzdCb$|%tz%e<;(tcI^e!1-U(mdEz+$V&6376 z_3CC2pKM3$B#-CQK={ELKiTE9qf%#hgH6GXGX6h|RWJ3WQl5>awD zLgYdZ$`UZHu873k$EFlLZOC+JXMEmqpXv-0X@>>vzZ8wR{BJZE)l>vKrtCgs$J#+$ zwufB~Xe-s~#o_Eg$e)3q_dQ;tX#?>A!EEpl3?z}Hvm#Hbu6xuBXz!9BHV zj0)k{^i8R>Il4DR*y-T%4k;CSrTJ6T2DwGr)7nWnPEyoDHc}?7f4r>6Qw+dIb4A|2 zo{A0s(6~rkK~3}){!S2w%I5EVGu}#4dF9WH(VLiMzAp{{U(F=hxnELkD*#VybAu;5 zinYVjm~6Y)=s=~YMKV+P5&S@EWZ`OENNR^f_AAw2(k1hR183%8_*_O^+*x&s`BE@r zR@@3oJF!6w2Xv?&v|V$45p~ohJSkiCpiM_S2Zmuh*6YfmUKO-hej4S0?pyq=kkuxQ zpbRr{!xx-<#&{1~R$OfN$BFhawH03zxg1QKWID#tJ3NGV%PxUQx!|Bl#s9njj$Rrc za&ywpteg;)G-nJ*Pu&`LVT;U(#vV|wupSqRGy!!{;MOzT&mT=v?rLoQf`qBuqMWTX zRtS6Bs0gN?2e-Q@tsiI99W9$T2iR*Xs-1ITR8y}utdW%mkVPg(Qw1}tm-cc&NacfRIY9F-&$KXw7?a_+H>`s%^BsbwD1EV?&nUKy%(k}B={ZQWbVKCR92tF zj&}}`M&74iXKf&tw?52kBEvLhD+(WaKW9>f7ks969@9jFlwL&*$1Scjz)S!4cs^Gv z1+$s1kIMB50!%~z&u@!ohJWMqUadKO^Tupt(Y=99`dk8N@d0tYI^?EZ$^7a9rLHq{ zQ2;P;McQKK)l3-7q^u&uOp#b21AQDCV`rLp5eLZd0%nWpi2w>AU)3lAiYF3R&uNbH zaU3-pcyNYE4}fLwvgreB&SwTIim;%xb4?LkAqajp&a0TP8>MounX(BV?+hex1l=AG zF(Mp*OYaOs<{U|Q|lGYLr$wC<|O{F>B zw&&i)8@>yXsXh0(4O#ly`*wb8qm#mN>U7!NqyyI~wN^`hzW1)@0t~u;Z&AwXl{wR_jnNyu5@wDMAXJi&-wYVa{m3flV>TzJVKX5|9VCa=v zr-=tQEBXTvIyyA=`%Z9{AV9u@d2>z7EA`@1|G}h{1w={Dz5C1F>&zs7B}4NiXD*vv ze8yo2ag~7gzN_?3K3mP}ac>WC&=AWT9F}cCDcG3C=B_L}JoZW}R_pCP^2?D>z*Ltp z@ff&NT@B25U-==yuAG@MQmuOtDG7OGvn!=IA>r><7E1`nR)rjY1R}=XC$+Dh4gRr6 z8HMbG_AeKPDF7MVqN4}rFH>CPo*Kj&&LU|8I^oC5D#(tqutFeV&Vw*2ivHu_U;M4C z8+K1HL87qH_1&Y*=}6komFd zphXB*6|3!-fft)UGry}3ZO#?=EbGu0ggPC)v0QNa>j9#?NMv%;aPTwCN7HP?>#~Lb zIuunMp0kgRHCao-JZA!6Ug!kf9MMjtadQYi>yU;g%6KO##|S%%9gh!aHZ=wGGkksv zBB7$9z?>W&7XXJ96wEt7bh}Y!uYA`3gE`nFHiFfd(qT)Hb0()^2e`1c5f|$|eA2O% z>vDFm7lv-y<~fmBbVQT@sy@9vD%`md~tN# z$Y6&wgQur(d4IqPt|BZC^`#`A_UVkyrDKnEuwL&4lH zJ1z&l2i7n&>fMgSnjMpPWaQa-mk=f2)+I8Xp>imNv_Qnd~KUz9^@gFTs)v4jMfWk$fC)v(Jr%rUf zWtFX!L8BM5{H}+Ps>O5@(P$)RiV&Z(=Q?}RiOwvql~0^!1P%q}$%R_tg|$wZ0nuW| ziy^IFvk5hvCTk`F^pHP9i|`U)z+?uisdi(NM6N_hgO77wH%U&LhJzU~-#m;+4&hj( zY+K>ZS}Y|?vUK?BbC0q)`;KtuQ*?-2s?5R`>`Z7jn+Ck!;WbN-K7%{lUnHj4sOWeg z9*quT_ND9)>@G5X0S>6VHhdD2+P?69!H4Q z=n*7_B+tqX)Do^6e$tmJkD)CD&g@THoTOEq2t4=9qf=GGj@cwIAkF#>k-4=ncT~A5 z{-|8s^&&*ZUTUvpCzLBC;=A{+12KaLQI*p?1mYW$ioMkr#xQ4lRY|Jk70LB6%^**h zESHeVs25tqeTz?hjAtiSZa`r=qzDz&b{<&+fO&$5kI(8oNQ2+j!gDyy8dV#;7Y(3n(?LD{ktjS(DVV7J;WWL z6D1k6RPCst>3i_ExSm0#tGkA(AMka%5l(kouPrsU!2`7k_^4#IV%<~SHqo@8lUv9d zzE3nE`Bkp^mL^Ky6@ut)_ELr{qu2LrsFb*Ls^u-5ALdk5xyxk0Our zPI#0G&}YpZ1~m^?Cg3kvr_~W=v)_+MKp276v#>em+7Y6Wq0)So(^SumRs#21t93zq z&h1I-JE-z3Kw;mv)v3CUB5NGE=T!Zl+cv*t*h}N3qJk7hzB{at!py0mU!TU<><*Vy zrsX%PqvvY_u$>yue{ck@i`Iiw=_3gh4}(aLOa@eg!u2BYp&IW7ohN^t4Uao39Lc}; z%=C%2A~f~`@=)c7a>?)MV)l&pPY*(O^*#d#;AUS)ovLC23ev-3{$W`2+;i;EVb$Z! z2vXX?1KL>v)LP>{S35QrV&S$PYw5-fZ60mNx|$O zD&OHN>QITeB{@0IaoA$-&cz2UcVQg_mgESyFQ|lcIMRINgR|R>WoC?t9acy(2C7@X zqX8vLAJjWCdBrBYlUGOSC$4jcc~(X1O#Hjz7ml1G54Cp8e7xVV^YMZA7^jx}s1xOI zqtS75vg%Zi(CEv%_K>bZCr_nF&u7LZMALh6FBsh607-lp~Wp|{m68FgbX@&2t z8f)~Yn5pX%9j(e>++vqW`pfEv_XlbO`;9-aat=q2ZdQB>(%_otep|BE&` z|3jPjHG{bK+Czt0nN2i0-zYyIIeOW!t0gHVb%I^lU{L@5YLOT9a=E8NPFP8V!pX02 z4w2o=;7s#_er<3j7h)%dcr|1>xX8KZY_*m|?nyDPbxEW^DPl1(ywD@GxP=4`2|Y_r z6rx6-&q;m$hE32l7w_H*rH6?qeqXE%?|HJ@{CpfQW08DKFzA!CvjAmZZ`+wak@(pO z4KRh1YTs)g`mZzTX}}FUKe;S^F#W`NgLGbh_la;!Iw^YEE_4NpFL9!WnfI?l0)3#2 z=fU!t6V1)Gog0k7$LpQSy92AvtApps$%$J}6c6}EA$ptQrQn;w9$cyOW;g)0TZqQ7 zYfIuvxC;gxB|~!o=C?H9~#de?pJ1 ztKV0}TB)6b%SYlJIsJV-QnFA2j!1SbXr-Bo92%^f+bRub4oIC+Io(8ia857jRioq3 zsI7|r)t86Q&_KQVsjM+$Z@NUX#OWGOr4ju^|HpVH=vL@)(U0kt)7d8`aSsit+L|iR z4Jco3($+FOV2C-WN1iQyM zUl9RqUNNg@uo^jBt&E=r;3l1H0=F)EkydAMDg~8G5NCMM$@JztsX8-sfxas%Dt!) z0h0cEcc*_41*yWnv`x87axJ` z@589+7Cp^M1|N%PsT#T61MANHd3+f_fDjnI>jNO@MOR5-`&5D`t>- z@_RH($;=0&b^V=KHMKZwO?u(rqf0svFAxG>#+4-coeY3&X-vjNsqq`FfmEhmK+;=T zYxdgmR2KYW)>X8_bUi3m+6qVbKh8sP-j<3{sz-?z05#46kawMp=}tkEALriH6q;Jp z{l15(A+v5A_}0Xic&3#cHf3JDPIL+FvV06?i?c1~;bw{B{fa86d-phd6owom|?*`ua`_R}z2 zW$Ub3eu6hsPIpMfPmA0cX6yk{9$R$uUkf$-KOD*@zNh8szadouE7aHIqEpp5{n>`+ z57@JgJVKcwTQ{`6yJp_g{C^KtI3SxjQMVa zS+lUxSZ6Nf?th&?vHPYYBZ3x9(ralR^TT3T0l8?022+ts1ZZtbF!|g z)_~>Kn=L~iaPl(VzvG^`qRTXVjPb#deqq;*K` zi-S^W9cHk%9mSV?d`&eSKH47ceIs&1Vz5e%u73;CVLpsP`=)smW}P$m1>{4q&}=TK zS3dGz_+)+99)s0zQ5D4GcjAHJ_OfTTJlBjCTJK`ueTQ_-mDuV#-Fej*fy0fey{Mpc zCmhQ}<_%=H2Nj?B_)_6BzJj!opq5ja3#aEkJG+#%)VyTxQmTt>rk>U7m}*#R+WGDi z2W;@=MK%m{fiJ5ZVBCIAEd*j}JrD;OY?mK=4T%c03vtTC=_T_9ICem8X1_ko9lv10 zh}C;28pVFe>4@!+KhQ#=?8$5SK@TKGa_pzALp*@owSIT^c|XM|qnC7$4Q9gwk{ig% zXYi}(i{r`MIXvw$>JPPhj*%)-9piM1Uk?jh^%ZwIexwO`jlianp|um}l?U3EJE5z1 zoo3{)Nw#dfaDt^rqfb|q6B7S*I!<*O&UoC}Vr|vw7r6F)gd-@p>(yo(zbTI}KG64d zZ&*SJg(+%sRv~jt^sv+`TO;T!UMK(<-)HanTq3z&?nV-86jf4>{@9+tQdZ^2EILnf z&Tz1**icEKN>d{n)QGODU9q*XAa1`5!!8ks0f6}2JurMwxn}b|)vDe#yR6Oj*@rsx zbu4wtqI|zo&RHjF$6L5*;R5#>TDQ5^#}2hS>BWi|E62Im4`bFS2yEc`&B>vt#pQ41 z{|%zt{uzteK;~*bSwL$R)WmMyhEJ#!WtxpZ?`UYRb?|%EE_8X-u|va5yYGhDGK)vu zv(!E*{Z1UEQO8*6I-8Q~S%Uq(*j6unbHFc${6>IJk*jo#vxIEuc{cY#A~S%;DGGO5 zR^uYxuLcCF<5P0ZNxs4OY)YXPmbDvV zikg4?P1Qe`Hv=UjTxFJ+YD-FaX*V8bxTO%E`<7fgVV&od^djj(_J!31$IBXB#rNN* zuI45u*}K+D)H(`i6qZIO%8e7(s>he*Yud+_` zG(9#Y7HgsxtJ>mUyo~@H>Z>eeXuhcEyU6Fl=d)@Pff)p;OTxznrXHt`UGwFbQn{a- zuDPsi%7zRERaFW%5?Xny!{rJ3HY5a}KeK5G}K z5oYVGjZ=mIaJhjC0Zk`@BxW0Z&fZ$<;aeh2?>fL~&7D9e_ZAS;VlqSQ$)Z?{AvDR?7;X7Q7 zd*Me`?tGVOrI$<XHa%m{=q=DaUWO+_C`+lxB?xARX9h6@K(r|@U@l#2U3&C#axX1hoV$7Tg=kE_5C|LEZz!yM{4T+<)3L;!}b2TxOo;(+Hfny6-6rF-%KtDv;r{3u9%`a9#Aa<4^f+4;toc!3C1_&9-nbqlhfJ`CR6&(fzG(z zX|X-HjGYfh27;~8&b`!M%`=|7wlVtCMlTZ7k%vzZvXz6d1|!i?33$Wd&V*ox6>OYQ zeP(&C8A(UaS}9$}zOrs?*XwBif_4}>O`q4Dydm8MdN=mLQO*DI^s+@))u&F7xFMyI zxTt8J2BL)ZsZ3LR`${Ar;1tqtbVzrqxD2Dv)G@VWY3=%<5bO5dla7-F>buO>7+OKb zB*e_at14Xx>sqX`ELu%`tas>$ui#X^&rm-N z8MVe&T3U%q)x-j^Ugbm2e6{9Ri&(Gst=a>BYQ@?X-ygtum~(Jj6pE)r;zrZ{`k-i2 z*crE#VBZ?O4{HhUQ*L+Cng^r9^#0uw z;ylT|8g@D-^tQQJzJf`!1oH#DV9?VN)Gph@w;fu?u9pzvY!?Mw$8dP&AN z{}>BvFGp2p7138X0dSJ(!H84PhzDWE9}du6R;x5Cq55FT&QY(YtctzWO5_=8mIw(k zJt=FT|BTFlY%AY%3HugvSt`Swl~qIGNo%~m(1(QVm7GIbv{rN=mzbL*0awwQm3^bP zG(3U>a_Zy zr#BaBTMr!v$E*MxW4^{s;o7R?OtrgHwILsRcbCL-;%8cX$*LB_PloGD!u4Y5nH$=u zz7Zr!UjhG6Jjldkc0@|)Snn0A4HT+8flu8WgL|C1e?B*r91`{ZoAKR)3+n!zB7Qs* z_@=TeNk1gh4<9f3)=!nP2XQ0pT?WEzMpjRvxZSXGse#@z-|M_#-(1CQuPH%YE6Y!h zZ!J;Y%8?35{BU}KsD*QE@ak`@eK2V$_|iBLm)go*mAv++IV8>(F{nWAW**(wwN}bU z!E^bMua2y2UGpR}GOa%4@C1O5>ms8T_u-$Sg4b81(G{7ssxZdwFDrG7ejsM4{%AN( ze!G6-E3Tw!u4AVV85iR;NIX;WP8Hwo;;?R{^6OUbBRQK8y~lrbj8MFHCo~o`|5AQ7 z3RyYiN3$S{0gh&b(%E~g+-%IMu}jGjX967Jhg7YWfKXb*E=61c4`{to11!&L3M82< z5IN0vxBtX~k?eJD;DMB*@zRP3Z(hyifo4xFZUmIvK#GKPNK43Zmh%U3x)Iq;6{%_u zrdYYbu2Cxe9n+&L`-mOUF;A(_LRqA^Z|~ zG1ceSV*!r3tL{*oPaM{Z<$hE%btDBzhRI_7wxRU+<#)Cec7>o5cSNw+=&1u1y$`i) z*w>g|p_Q%; zPtO!aS?FJE_d4;wdu_B@Pc0WiDW{ zd3|$ND*qav)iys$Q?DYo_pFd!^O4N*T>T=2b>^(+#k99jued_VT{bi4WUG+y;5#0! zVZ7k!^Tmq=O<8k3K2v7q=^Js7z+fV;LTHl90p74)D=rjgJZo5vb8=xbVnJapvZy;; zkEaMLQ1Dx7wWnb%fMDu(re~mU+g{Izs2`!U#TtG6)kL3Yk@9uVgm1}|To|8J(j8iZ zF>{jX>MO_WIEX8>vB1!-wA=f&I|Gr-3vewp6&XTRS>@aOj1_rYNLy6Zc(2f%-U$m1 zk_8W#SACKY44iIP7?nZIsuz{!3gq{hQgm2Snk?BI9Ws~Xpre3*^&!`!j#XBXx&-+V^0v{w9nR2tjTQ~_GzZxdQ-Mngbo3f>6hGwNoEVDE<;?-RMwM~QSiLt< z=Os|W{YJLM`v;JFDo4amMIb_E-+f2vv3wLm3Ghy5sf=C9`$6xy9I#98KFwJ3@)M7* zZM@X$2iHw{%N!xV`7sdviO5YNJe?6!ZOa z7E`w}`X?M;qE=rDy#Z-2#&j?l6XTMRwv+SAw$P`QW0 zsdvMVR&20)f6DaRKMJ$bDxQ(OEUqcvQ(?v#m)#%wcrV6#_S*48cH?R3vJbP3xWCG!Dq0r8B>^Nsql~Q@E<~FSdXPoqG*EOnm#$4x*7Cx*YjHxQCWqo zC=c8|t$yokw7YLgBYX0(0hqE>!=CLdjw+^3E}Gf~QQJf7+wlNVG_lz?LvHteK#=Nh zWQ-gCC96TA7OQAD0Bi$sXxG#WcxHt#Km3ozEQkqpWDn*~Z_PX~pI`nf>>lL9_TG}~4AOh_>{=jv!BN}Jo84E_|-iLUP-u+yxPJB*2 zoZ&q@aq-s{&x}A29_G=&*`HAL;$3e0aSQ(eC!Nj2^E{o`7VZK{A?i_iMAqvIf%Gtu z&|qD=5)p=Cy(Zv1+El(6+GrV}leB`_8x@X%8o%o>HOVPHOq!(#5w{P-LelUs{XR_0gZk<;!{$1=m>BoY_`!=U-oHlD35Z>WYIrnza zvqJ`68zo5gp3>gMKN81rfJA|wq?amAzJ+Y?;695-E_fXGAqh0o#S+HO5%Z4xGm20( z`%PB>-qJVv1XO+E;2<)Y#m|R0Y=OG~JwU?0D{|lkGodp{)d?u&#kYPh_^H{vsE%C` zR1l#PRo@Nt0;5S&R#ly^Dxzfos~-GT;zsNEI7(G3cs z)6Y{Tdl-bKLan@RND8O_IgLOlB45oAYYSmDNdOA9@6gj*#rDFu1@4m58AJGh`QeaR z%#Veu+r!%&;QNzVIa}RQD68OkEltX7T)T9$8|88~*sHh1uJV~4ig9@)#)-(iK@Jr> zO7utUWUjs|nk#&^ROgG;Sa>FU4K@cFm*rHCxg?ue#!qLJM(1jdF}=gmzj*{WPs)<< zJ<{P2z=$2*R5J%Hxnfrlfo^lc>p)@#dZflC$*amqOXbJmu9*#e`a3kqj|W|PNG0&^ zwt;D|t9D`>SO(vo79-gm5;5W3i=J()9*WIqS1BhW17ArqbY&1k@svl>+l1f5Qw$zcyXgww%h$@Zl~+|AwP{(WI$X~bj~hbVbUf+dag^4?{txNX4Fhf6yhGZ~*L(lH$a z$A#MwybNcOdGkDb=7I!5@!r_m4z8JcKP#*MUmE(KCF8PiOhos;G_>!TYiDp%?Ac)a z2n%0)rC662P_DCHq6`-vP8iz;ea~XI3ghmtana^oH1GcGYkjkW+1Hv_j6~3!7HdRa zAeql`o@%50hKxd;w$+x3xnoH@qVO<&A2yA3E|Y~2He~#0@wc<_GS*b}k|^tMRcd>&z^F29=;onvOL6zLvv1zdsYp zQj@o-h#!<;NM2f(9n{K^6uroYT;<2uwh)8s2;p*mUG7EFhMXt znoFwoYJz0`sHx|eQnyll&hT5heX>B!vJ56NCJ-3SuhBhrs)1rq20kMZ0i9-fL(`Vt z=hTZ9f!vfu+)yv?0Snn;wuh+;Rcs6R#|4sA66DOWgH?9q62I~E<;rMwS9l)t(X`4) ze`bEuqGWN3wL$ssz31=&M23m8PUNJtd2x5P|10d5kxcjaJLIT<(>@tG^&|JKDRdV` z?##lyE%E)_tiV0a`3!k|`qqeq{GOLy#>N~kASSfm`{kPFfWHSKtz-#;Ue#7v8)-h; z4k3NxLUN1&r0n*oeJs%6-aaEyJMs5)PF@Jw0aO?W6ET>KhVMpq6 zQF24=?XVde8M~L@mYe@M1hq6!c z0q8t=n_7_V?&!`5;h!MHFmorVF2IwFWV9HFG(1xiPyCVYAZCAYdpNwV3t$z0ycf1F~0gSPGB|dS*uLL z&JN6CIWQ*H`pUhI*0jK(p>x@$#g@eK3rxJL@6{<0MRh@vrE-%UG6%G;yyZbkhviWv z4VynGr{s+CX3{FfH~|Ml)>+LAJ1g84I0tEYk`1LfKl#uG1PT{VB6fUVI5A?tFA$xa{sOMP4aTQ0_Sy4yFw zJwL+-uJxBi^GlDv4-dci%wlQ+PpX^Pr!`xkjG9NcyV&m%4`mX>b2< z*87@5FrSkD%7``PF9feP6HFa zHMPF%rhJ~^?audiTeuJSBu-lJO{z+Tu^k1r$OmXBZ=+0AaPLUGb`uHZ9d-Nh9qh=s z4@#HMt8yu+b*l19%spn|F-=x{ezw#(?C)12>THDj41>3EcTzloDesG)kNo*jq6gGi;G1T2lQpM!iJ^Y1Eg!Omrvzgg@y-vUhM=vII8avEw zl)8Qt$Ba=XI*xTJn|n1m?hIIxy-JpCxkn1*HgYox)eB8cT?_G`y;T^xIkjdk6;>NkFb-}Qg=1EQlh(!mNMLbCDi{2gKkuNDs* zKE|1>S9)nKuIK61Gwv2LloM#5TK`(B*>JAPS45g>4Y-Q-1IVMzUbEd_(K*GEkCvru zDZdM!Zwk1Nzx~cykF_$aXb3x-Ea*e!bz_90vPll`VNjZ~W`~tXsl@ItpI62PfY*qt}{;#+czNtdpAA3?8S&q;@5G)6sSbETC0>PlMG<+UrbHk}k?CZKA3V zcmUYxvV0hH+6#HuwaSD|Pu(;hy|T7c;Q#pOwOVfNS6zLTK}*+Vi_Et~4se#DM>+kG zmw=^1rX^y>56rj)0Bs-sb8VVATf)r2e^7Yu6sjT85_`?I)^%7Bnp8P}N5>^I(vK~u*Egkz}3 z%>dt|3%jwbLNr!+5--ft(98ePO3i zswqa3Mqr~zqpP?PpwZZOwkSS&Iah+Z6pxe@ixvBqUssmOyEuxm!*7XaYe7zj>9n-W z3>sJdM4vEleC6M+&R*I{F2@1_n|7@M?n!cY+6(u?mHOjLh)J-k@ryJ3$rd%&OMFRS z;~%ln?|HCWx6}~$foddscjm;Er_bG=%a%r_j%OIXBaa!oTpe%gM~B8?=afyu7L+i> zm2l73YFeNldP`Q1mf!potkVzvX}Mb@X!OWcAW-@|-naa2*qPM`y9nzEmY-oF-ea?& zxSCwM`n56PoCSejyJK1bCshB$uSx}3Rp>Hk_X%Oz-ZqN*OoqW>H(30A!vt&Wxd-RA z?hLl4g!cKKcrkX^x8p2h=iwDvx9K`JWeze9yKQ^8Y#tT2`L%q`My%H|%JX-z6A1p3 z$3@_4j$wN!OhsROZj+c1kfke(rw{V8~2Fo8}S_W^}PyYeZ2Yl*d+LvJ!4OUWG9Pn3HDyL zgka_po*(FHAod*WE(r3NOE(WgZlD+B(mS|&8?F672lp3Rv|+>M9RRbB!GFearniCd zY#`sH;7{>^%lIMx@{jIrnTG3qkFr*5S@cA>z6yKz3>OuvcdUFhuT289KI-1GDJ5O6aPfoN};Yr~U zB1IXsqxdB_W$~L@6yfq2^RbWYDj&dO7pxrTlUAgUGn9ynf>YANSpj&4l~RrztS=J zCcfsLkCJPF?zjYNM~`ARhywX@Z_}4WQ7gK3TvVUwsw#cCd}0~hL2tNFKQEdwH|AjW zqDhjzMWA$~CHX&ML~;uSX;u0IXRqY&GzA%?e;X=c1S#wQg9z&AqMii?;epD{;A%Io zJB!QssV##*G*K-rTO^6B&RKe_d(&xhLGi6+x56$GWbMFUk0znPNezBG`Z^Q+gqUryTqi_F1j$Hmjjyjp+ z$xrwTQ|9Enu#J-EKRtYCKT8T;G2(#lW_n{TGy7$=PB_XI+Gv)lb$|D|ytJkAGD7=s zFSD>aXNTb5?{z&GoCQCEx~;K$XTnsw=O@ESUE0kXP5T`$`ilKMjUEix@oj%L#XM#E z1mCvnk-AC&*IdNgMiyzMw%<-Z%k3zYCZ)wP2$+~B6l>E-G2#zQGKdOK5;#Vrv>DHV zyVD=9#WWGW77OIesZD;r<b0@Hy%_O*pXf9bz`ag}*K=dmAK( zdD}|xVU7%4W!#6t7ooYfe8M?qzB`n6;`dCPB>Sza6Eh{O)w%hQ)hIx+%Qu{@V2s713)l^%7!> zDEBe39epiROAqjp^mOZK1#15#i{d|T-y8*h{hK|1~-nzq+GFTe2 z{ni7qPd9{bbo?rS&5s^BqUd%-dbaAjh>x)B=5YeE9-~W-`tb#p{8H9po!6S#qKS(^ z*CGvUjXi`y8?~V@R~ld9_LvI(iN{H91b5evwL!Wu3aNABn7B&}2LVMM!zg`M-}lSteg*(dH*7!anWt-wLU(ny@};j*Ln;HOD+ zoB5*M>?za`L`KCs)Z^=V!&bk2QNtgHuaq;kv2rO7v}24kMf&t=zkOps7c=;&9_Jr4yR>!jG|=84Y8YC-kbRRu#tq zG|zF%jfNLsYUYeR2V_(qFLz}8m_(U!eM3tXum|Ux-B${cLU@_h22>zOf);!>HO;0| zGpRG=cf1uLruUXgkLMjwqc8ayXu?5ymAL4*!FacViVc54V5CT?)?;zDnd>5y?zgIY zqb%Gg*Q0;#?Pm>Fzt&!;O~M6qYD`~pJ!9!?^VloG^kRkJD%ker+U9Ra&Wmr4^t$9V za4Mz245yiaAGYE3JgO=!P}8e|&@r$I$MN0`@5DG=>30=N3r@~Bg|#lV{SP9easE}k zmL45_ngnKb`-5<<-*`NpQbq>m@NFdA&;%e*h?r3<;pAF*$-@(?Y6pARb?VFwisA3B ztV`85{oE{T-t_&!>((E?e9?o5;_Wv`0AHa_jr!lSpv~#^sMO>7wqi<_dW3?E2olG~ z&OzI}wuo?Un=sWv@1AG>Pucu$)4%Ee+w^||ACOqO%;xpf_>FNUyFPqIN}dHF#qFEq z)rL7Tm`ZB0BA9qAdSxxE{aBM12)VSh+zf>eOX(mb3sFFSep5YQ^=bpwm`82l$2$wP zY8}hO9?MHoL1Dk8&apU78x8hfwpCt@lfT{{=R)gn`?T`1?;B={f#oFg*o7y#-TE|g_(rEyveXnV3HlhrazQrTS(R9_Pty%-nu+d2jv>RZ`+ zDzc6=+Rrgt@=9qu71YT1QB`-Z_-}8X7hKBTAFy?O;H9R69^0%{4>IZzwQT~ks$7Rl z&KEpdjSu0o!kyv_kE*T!RpCwtN>U=FsXY~9$UDG=-nqd+MlA&@mQk*CQCRGFzO8oE zifLp}^McxQ+lk_(C0X-%A*)bbq`)DygWbOhUMxS8+XR(-u3w^A@_huL2C!v`nto!q zJKnW$95q#lgSC{w|B1DS>Bw4$c43MZ-a?TpV`6ap}VvKUF(_wd-{;+BZ`6G z%w|e8V$O3bI_MrPA^2d9^B|x#HI9n1a29m^5Khrf#mfFH(`ETac}Sqc$@YdeJLRdo zsDY7lI4xGu?*586vYE4MhYegvaGo#yjllTBqYS{5PV7$ME10biV@Du>rjf)6WHX;a=Qr5n9Uq?7NBB(dD zM6KqoVWFBu#L$nn!1svmeiK~PpwmaiHiKl>2tgtGkpXHk8JJvl}lMmyGeMUq9}ncE#$RRT;aQWy)H?@7UA9 zlcT|rDqGhnc&-4v)b;LY^yyLR=6HA>3|?XE{I@<`MqGNleoXH~iPpVGm2N-cFLZPO z`0GXauEP%2)beCSNjaAs1hWddmEWtfk=&p0%rtJl+yS4W!Zz?>{D^LreW!-ZaKDKi zUomU;=D_S2w1@kJ@Ms$c`-WuAnq^kEjy}jA*NVL_)yiXz83BWdu|ST@O~5*%!~E#I zX;=A`Y5mF?ZoTr*RPakVzS*cn{^`UI^pmF4Yhl4O+^Ly+5B@W+p8029y%=}^-icvN zjbzOn7?N^>&r7kygX`IXSH))0nLNMe2H zd|?l^XrmiOCPue`)plu9xpouLCv5pqp8fA=+zJ?#3qSq%fyo~{i(r*+F2?AE1={6J zXbK~C10e%K#AL4Gz>dkQglrJYq79U32~+t97~*emHRqwDD56{Gh; zWthuXS@-kfk*iJjQsr{DRtv57v$?WaStP-yM%AmdQIuW($LKl^%f{oQjbqVJxvCd& zyB~8+pWcU`JkQsAI;gyKaZLnG)NWulw0kKhXmqtAH?3VS6cTD(iEGyk3T4gi8Y@7P zvXw^O@Jn~+J_@xZ z7{<)ZCm|(L@wb7ZQ~Ux_D*(%r)mj(ON8vY0Q%tz7tw4aju+uT?m zw9OIN;V4;b)J3|nqew7bzp5CxH@q7(r^slEgV)c~P|CNB__C`C#rDk@5EA|(L{ zMXI252)&3HdI;_6_kDNnuRHmBW^&G%^Q>8G?Y-Cjy0~xpZ%Q2Czo<$+@Zf~XJuf} zC-u2E?97uI#Zw~Rf8VnYN*!WEnQhZ7Ue1qeZ>5r1NW!hjYtw$a9R-2KQXAZsv(}kJ zVTx{1&9{y;KMYzhI!;MEoZt*GUGj@2mR*5o33q4+mb=BSZ@Q*OS2LjSIWGOdvG%n5uXDWIjT{*gQ8>PxD?&C8 zkGuHxq>o=+W52T3ma9i$)Yd?_N+>pfd)SW7F~0GeXs%?KF!Pnhc2n{Y51rM4X9?$ z1bu}2g&b@hD3RidOv`i5;Y<@6sXd#?BHYJw@TAiR{WC}R70eQ8@Bd?EEbsHq{ogEr zzqW?)-_7Wgw_D1d{td!cBm9Am3Q2-T&&T+aZU>P-Q9(i-7YlGia@0At#p||@-OJzI zi~CH3H-bMagb1r0JG16zaa5nFk`px{8S7`SeD0Q4dAT`4KP!}YhIoqdN&S=62|e{i zl-&zen+Re@mV`CpJVJW%AO1dMMk#FLmv{L7cKqv(p2WrR-!{G53N2&())X5Z>U-t- zr*6W1csLz->(a>OjDU|h5hQYoQK49#pZCOh)lVuT@v2Nuz6Uwa+l(RSg2KT$ej_$} zrK57Z;}?Z<{rqkiJ`d>iAuSiy`M(RPn)vx-|2qVmZ@5)1iM`eYP+9V3G5f{dJS?jx z1LCwQxi5m$Ki5a9|UjI{bAH5B1*#)y$VElBSX z?L8nY`i%vIt*blUdTBQxV52)&@%;W%w~nB?HsCpnIjszzk9G~t#+vhvpur6=Lp3MQ z`DR2JigIH}TeB(;gdg!|L2a1`WyDELoUWZ?>r)Lavcd%Hydd{KuIm8iL(r_*{QkR- zu#Hg>wI>R@P5&HF3sx-_bfk5HJ!=_`Ez(U*>u_s8!n!lr5l$12D9O2|9eFn)#y>}{ z)_1;>d>6iz>6x-U%cD?lSi<9el=i$k-BaYKr-MNg$@z+XDC@jSKGUT=j?Pq-*z>|q zqVsA(>Fmi>ci~Nm@ny|A2hFxls7<6Tr+{budV68BRYDAArz3bfQM=CB35>FKY<_Hp zA%hvPu?+d`0^>id4|M}m+=v=yGq#^!D*%{Md#W@#!~|DA)us5VJ$G$_h^nlu8BNbI zlCNVu^s=(b523!(rf-%78cK(C-*lg$Nf95pticSo`(3ybfrsV8tK;Kr*_#Js#+Mo_ zd$HgAH#@`qjO1g^(PJUS_9%W5ZS|m*|D&`>ue0QyE7&YEQB_;na&BEpA*e%8x^(&+ zu_0qB>4nBzrfKbBPDpIlCYYQr-H_c@;?wQ5?&Mq-FA$gck@mat1kan2AEjsJVh*c( z@Y&7gLzuFg&R$)|<&+K;ssRh*>0QhmIE)_Evb!Ewox};>A8v~cb#a<7BwbTL361;! z3L-sA?I+IX8~Eu3KRN<xqmS|#NrD(Mc&BUpOZJk#OJwb`R*Sz zAm3|~mHb`hmhyw}-yu_)2r%=^b7huIXuR&aFt&Y}HwbLnnzV}}7ngvdLD z8F~{R7hEi5rlp3^n3jDi?z8i2n15;iXtzlRA%o@UJ#uJ<1z?JQxnrA22uQzHvdbbJ zguIsptH7?@Gl;6AUtU%&EAR72aaDc4j~7EA5k7c-gbyN);iSw7hESPn<7&&ATHC6d za=wuH{+;tHM=0< z9XZ*tKKG>b|LQvHz%CW$oEjb*MzWE(yV3CeE)h8JtHkOXnUsc@4L8|xL3SdDWNV~ zH))o-W2~vFR~f91oPh(kO8Yuo$#Fey`b!_kwHl1ap8J(r={=s3x2?opcsjDlYMj*A z-x3vNyb{q`*CQ$4TZhWyS8@E_@DQirMQyn|HyEeA7s~pfIm*Lp9p7uzzQk=;^TIhr z(bM*iCoU84gHl=p(B7ua;gjeWd1$L5u}%NB8s&`3i$PK_a`>})}UkCIu> zJe9kLdXI#Sexv>gOxT&@*V>jUKh;g^aK7rx>o)s_n|tGx-SO>G5vlq9wcs~XRs2)O z!P`S&S-={nUGot1+5^3jijE4?mCX`o0dQgrPr_1EMUr$*I5eRBv(gon`_I&aU2>)m zX5q&S8X)6*p3HsASm-;a0O4(}I9u&!R*aURGH&mKXMK>Q0>|BfhNWNks<6;RXAj^^ zY#O!uA-_4@awe;c;E*$EdISe=gX1W#^@Cdn8bQ$jn zq1iG#oSLT>%ro}O9nA6DZ` z9NOdE*hl+ghLq3FsNO5r=HORoXFbyB%&G7nZkC*=wC>CaF%wiMb%gYrIT10`N! z7+MIo`?MeVxkk^8s*tY{CyU*L-w&|&1O#lqHuoM{ZkwB?(HvK+FJ1UnX!I{K>uwwE zySE{sHa^8M!cR|?D5K}u(`XbQFI8tEueOR1lv3+3g_>OVooksmJW8G!n>yM_FXg(y ziLgGkWm$mx;XETvk;CN=Rn2jd3J?-&nU#4g-s^g~4o|0!*(l$y4qUwsM})9GN0cF2 zv9YTCW0Ft>x0$eY-s&9Nk)#>j-U7GFZEvml!mRm0_8f}iBIDV z{=J{(!h$7l=8I&u|8P+l^aI;Po}(Nc`NfDV9Us*P@icxAwRW_w4cX|BHl*t-cHu>= zp#XO5aLDH$fX~ag3&OO5tnAmbD-L5&(CT}6>?F)=vZud>_8f2(*$FemM5eh#ke@mV z!wSG#6_2J4*B%?f{7dI2HzEpyL&DxW{5!P{H7|0=4ye}>rg>mvxsWN>BAezSvr=4? z#nn@8^u>-Mb7qzTfl~#ZO1Ye%sh;o8BR_+ zb#oV*hk6kdf)73QHU^x%#WAfAzW0^S&JaS94=ZO|P26ETq!7ZzpUOY!-rW=m+D|5owa;)#QOf-dg`mVNhDv(?n6{AtY4%TNcBriTQnaoa}}F=&&&-x(_t>(^G8! zxX7I~VFgc~cCmP&nk$#C&7ei;H_GuMh$TVY9#pOFe96N%zyE4k)M{$=D(_x37v-7P zQj6A0dpdW;Fz^lF1C(|8h9M1ab)cf*cw|_z%xlfa^{!ThgPN*Z+=S$C#=qK|P=hQj z+Ndm5w)K(pC94CnUrgg=S9b}&9MdfeUf{ux$xW)b{N%l*%lDy=dtdfZ2;>Zu;cqw8 zW=6lx(S}ifZ*hqH$a60ysy_gd}rD5s1dM-+wTom=8ty7L6IXtkM~ z>UD+Y#4{}3L_))b9#3z)DlFYYRVKf8M9q4OCqEqiWefpOCPprU?-U!qPEc|^l#q3lDWflB%OS9 z@VDohdKP`|CU&0VO9 z|2!A<;jIF3eJ<(#eA-)}S$tOK==GeAoL<6K>d{m0U|gwovQ)qXF=QurvTkAOe8JBN z<2Tc6^4WaDcj0pWY$7Y~oG5U@fu@FxZbDe8C*vD0VRX4g+*r>}qtgpJ07iQ^N5+I9 z_4^JSj<+2QbM1A{>NGrZid-G7RytvFaLU>+>IhH`upE)AwLM@P<3MQf4IV8S>S*Ab zptq_>d)hy}W+6>L{o@^rj2Z^RE7`jdF}XQ_V|HU(+SCektf>OU)Lf;wsEyN~+ZLG7 zW!I!L75kj~DTOE2r_|FO=fLxekp&JTq1(evSmzJz!aCz>T&bUXhongKuw`Qh4F(8m zM}IoyHm=P^;|4a0?p+t%lAbQm*vLxJoTn4K=2HUuEi_%$GjUs8EVFN8`bR|`&)M2m ziorYNgVrDY<1*KEC4LdoIGf4er0tjZsnvCRo!7gv1$>}FQXih{9^33^Gut?$P4EpM zwCq{RtBRfuk`Xd5r-_axo*DhR(#p0N(&#P|1xX%ck+FR}VMOFfL!B`)Y`?horldng zElF={31*6s-rOQfM}Dv1iX=Ud@31U0U*#KfqrAIzzRsjt)*vMSm2zAz4R;24POfQ= z;|e@7O}~n^B%r54cPI1-va%Ib=-0>7Q83LLZz|V2jE{Kt|7l$Cykh7-<_{RMYFN3e z-{x7{Zp9=6QHrCbY*3aFqg^@XatnS@8nqrD8TCmWmLveZ1B=6cdE*|{i-}XY-+_4m z9FY$sW#@%ZSsl)XyKHgwQd;4=eJeq)ykX|S3zT!3yu8|%s7zL1)ANo$;C=!wI>Q1z zCVlq29t#`ee6u-#sEpu|A0M%*4Py?FL>UEoXtGLj;@hvTjARc`49@w|Rh7oqz>z@Z z)cd)5*|yrR%?n#0;Qx<;$r`Lc5K1~&r86tKGgS2$D^^V!h{3D?(l^CV8;{;LNLWu1 zHupr~1m#7dMYj!N#t~0|wz?Gd!K+=x zevw7m-W+|I2kCFGFZp_H0O3P)-OpUIFDDdofk24NY)mr3x}3Q(oy}E{T^6{vKbWJG zI?aU|l~fkH;IRQD>=PJ-I4SIltTq)F$0#3BgnoZ1A%RVLUbp(3_A>U){??bW;1~TF zol7?sX>W;U{W`l#n;()p(I-lC&qch?F&P4=*54{b+3G$Xk%mdK-yJT)cdr)crboQg zK)nbUtCq1P!?hHLMb*^H`Bu1Ylxeb}R*AZ@cFv!h0hUJ_X-$dyND2PAhfx5Rssz_s zG6cTWQXGV0z=aZsP=Ukba~kw>qJdpM*uh|mpF^O|=SxGkyk94DN7`^?g&3qLY{<8F z8eT^<#E@0bIi)ZUxHZJ2_nogJoKixZW%ur;nhfv+Hz$klIu5@gnluBl=2WoDFE`x# zlHD$6ks(iB6**dG;Xi-iVN6oLTWW@W-Lg0}o>{X6lMQ`0YZ0<#dE$<`DA&sZ$`zfL za*8%{0yh_St-lb-Rqac^PI@0~BXT2DUjp*>o~{xBuvuDD6+>ag_Jm{HK~ts%D`j*& zhVX3!s|*K7aECsENd zSXKtJ>;#{A$7#WilWIw!wJ_t*+!v;cU(2(px(J;EloMNH%$cWgiw!I^jmvcKh=T8N!b5|c z+69*j0D70GwugBgusiNp#&ld9;i*GupWDHDZy~HBJqGS;q(D~@Tz|FR-*!tGdvEke zVTIU2<<-7jzC5n&Z@jiUj{flCodj%gB8qqF~I|B1ozhcGkN&uI(WH}{GKgN(ObDWPNB^ze_$aE zgNJ1r0nxRk>94tyaYx9I&)xvl9;aDp2BK`F9OSY32zHs zWxu9&2$0z-0v~c=#)Q_6Ih&`xTvMiS8un<1&7Cijad*$({3G7nRxe(9#@81%3!73q zO8XK{{Bv?YWFfzU&rW;!I+p|8u<2?-Y=-`nmQTmly0Ei@#7wP|&6o4=X=H*?>YV8n z?8~heGXnPuRC6>tu}GNu#~ zz)(Rg&0(bjU0bbxy*qELb2fHXeY_v=V|~UiVw?!+9iTSFaBN)v!4R5z!TYRaZ{#|m z{gj&H(*pHRg@m~*dG&kFebmkI{OHAM-}38*lzcWEltJbT8~tSHZOoe2Q5VCbG393| z2H|GrD>SkkeiN1TUuWr?x3k?2ojh&A1goM*F}lqC)S7x1H144T4WvN%hhE>{^q?1} zfaVCJw~UA5LLlHU)-S9Rj$GyMqY9aEX)zkemcmV%xS0Lvh(+6={hOah%xH^Qbc+m$ zdyOU3D5J<*eN8+e-m1Fw zK|(b2jx3Fqyf0M>Q*d$D@V@X#C9K1?ZH~p%`R>F5WX?p}g_>SpS&G~gvwo2sV>nj~ zz&IHib{~_G9GL3}dt*X}l zdL=S1j%on}0$bmlXPWoQncQn<9h$N6<_iwA8NQR8Dh7ojnn(8*qK0WjRcWdde$Bhk-|0vk+=|EZof74J%k=>0|JJa zVUktc57=>i~cAZw!@eF=~svCmzzMwfet^-by9XrFIBREdOGJk#)=Ii zBk≤w`0qzg|g$P)+qvRt}yj;MY0c!%cRnv)|dwp<)ou{xqYk+!wldv^XC)e^$~k z^ae93D!#}MauT9TG@(fAL1*#u+a`h2a@K4DBhG^1CaCP9D9hSu26J z-6AAA?|(HTCa79hS4&tfu@p3Qu%K!vHte=AsF%>Hd4xZqbeu4`p0|Xwzk^`#L_qHRM|Z&f`rvIs7j+@N1fLJ_KkuM%OK$O zZ{nQy>C^o#M5A8T+E|j(RZ_R)X!svNb@tC=G1L}hLMHcQSbF6a-OJj=BY{7-ZH0zA zmq6Av!8d-;gw~dx_C**RTz=ymsn!1ozFDY&nnzOH`i&%Q)1xGv0*e$M@4~a;c3gQBx1kwKr#}et8}#*Dts8ZHQYINmSV& zyFCPV$JiH);ByLU+79678m;XfjxqExRq^p6ne)IDoy0`LwWAY!qG6(CGSn2FE!?xcfge%wb?8pXDOQ*AXa9pE8sN(8 znBxDj6UW}jDzVzd_PYCjHIS4#xnJA!TdYLp=C7=ZF|QkX9KPHAXeP(#yTdp!4lPh3`m+Dy;%gTI8E(|jfH_bBbF?r9fNB{Z-!ThI0$K~5C zNczh6!a77G!oJf4iDNY&oA-5uVrcX3h6V9y69ufotU`zj(+{q&Dlooov%!bf()#4T zPpYHx^(GvtahS*6q`IN?^P{_-KelJcLDRY#y3(%*sx2hc_iCr$Yla%;-b*=tkphMR z!!FR{sjHihG#NiFN~1px##J+UUF#!@$S>@7@L%!1z_BZ@ zX-Ku{(^C{(>HD8l$12y}{hYj2z2fRry+#V>eG2aIdtt(PBlRA-bUoHw9BX2*r+EO` zxEk1gVeeej%4-1v?Q4$=khJ!wFmz;7gg0(XfJ5>Z`}65nADO2^`ve;N*D9Wm_8T=O z<~F$LG(aR5Amh~y{WFbXN{2C@HJCkVEb#mOlZBg0V+T|A_;TEdof*dgO}*O*STwH# z5a*2{-_;J?j4Ad!$_$2<53^tnz?FV#NRB1fKj`AQm(EC>l1lJ^s+Qbu(D;{@2=uoL z6d{QiZ%?Bdr>z=OlF&%YvhYbXKLfhjDuZJK=3t9OvgoF)30dd$RfPM{DjfSiA>osA zW!N9m+tc_q=NRvI_DEiwQ8wN0qc=vb%jQ&6(as#3>};xV9+vkI+M)irm#p9=ThsTt4R;J2mFLtXdV@X^6zPQBT*V(&pz^>sKX&FwZhWo}&QO}`rDcTup zYLkAw+H6xOk(1KMgf)Dm8FSuL1Jnc6oi2g<-}{oUg3dDk=VC=WHSp;*4E>9=rxdJ| zTqqrv>B-iseRvO}V)#mj>9X$`9HVoOo+2myb(K`}t-Y~C`>H_rjS!U3K@wy*4JW zX3sq#VFe-JPH9EEh>JI_B^eb1^hj9DKC!0 zlkp35{UO_ue(xe`_fuK?%;l1x?xz6KeGVA6+IO%Pt-xeP^E}&>;LxIITxgKiP495; zaFGug)t!WYbfz+UJ*L**hWL#f;TATJCjSFQf8gl)UrR>d(Io-7ziao76WG-I)qh+( zKT8{>$enV*J2i{ty6zUlzl^Um5?svi>)m^y)`B=Kj3q`FoK{0{A;6;`-KDnT^|5MN zq7k9jCc`8H?ekcc3-(`no8LNP)cZO88w?i^mfILjhWpffC)M0!gv$tpRff zFn`(cYIUoQ-nHrdKBU0q`RhP_tdvN4J+=!|4Qx2ncPuD6|Q zL|Y9mLd1Ijmp3K=+~_PU>JOaU%2@pjbkN|pXIjF0w$2LMWRAqK7`s9SYa-%AH9T$^ zKNYrPjG(U;?wy@9pa?WVrn(l*rlF>ed$LkE*>Bde^w^VAAlW*QKrPqB3cbJ9JZD%- zul7;;iTW`63V}(S4=?Xaib8iC96tDr4hkz*vKV50-%B6qqn2DtRscyrw!h`;A3pZX z<1xkHJvQ*tos0y><=*_5BwzOjR)=mGx{MbxRtDMhy8^vQB7y{(MrV?Q#sqLE^Ewj8 zc?Slc9ZN6ZYMMIQsR2(~f51$M^X^3z@SLH~IgCj@^odB!^NuYjPE;MWzIsj{Ds}G! ztCh&k@bD;Rrm0o#AO%(qmOgk`vYc4Wx`Opy-FfG;x^#pyQ8usJEb8_ zL>c}ZDbzV=C1w<^{7s6}?>%TzV<_8G{g^*Tu9#^T$-Zx#=J6JEf!=8DNY8zS>)FWd zO7HenJDWWdcb_bD)Nz%;Qh+;7Of+Yjy7oy#P6rVL=z$IyA4IZ@C;U9?8(tU?w8z%X z5TMs{J%Y9$YPg=@g5Hdka(d5)ghfDl++)7pil6K^a{4_itzQ+Xn)yA;)PXhw*03k) zRQ|PcTLP^WrE5AXtw;`6cf&V`B<{n9U|v;+ljy@tC!0Y8WNus-#9iNPF$+lb^A`Vh zw;X@mxqQ+AAn)J(FfT?E$!bckWMS2zdC5!f3rjUU@!T-UJzh#h6X@J^b)JOWDF%;d zeZi+*yOc8;Mrq3z>({%6^OAF57T7*xkhgovaz-FC27cDQLI;V7V97zXM|fNs5OJd2 zQ@4{V7D0m&a=2D1=CO$mOFw|Y1xi8JEmPd)SK8;^GwA@(;c^_3el6|jpPIO9M~M)e z>`tCuHCAST&FLB#9s8h-iGG9?4KSg8Uc5Kuo+738I$1^g`$rVY(Bs#Fhs*Zj@TaiB zfg@L=Oope)&V}^NvdtCI@5Leio8!Yg`#Z&_@XcNR-rwTWkpNIFPyPMGC84LLUb5Hi z#{$DaP%g2Ncjpyl`iefHzYShZCO*QGZL+c_Kf!k?Lp^g}!V#Q_d(RVkbJL&f75GDP zh20H#-=R;_-YQ^3{PMrEjlH5={T@w~rps3PtcFrszf4{2aPAK+wbQ(zKILfa&Xv59 ze)T%jpl}}{RBWo%(p$BOoqp>HaXbW3ZPLGZAKY}s$M8aoYq>^e6U1WUYn0nCBJEl9 zMOjZX&fvSYf4lTz*;XU6?DOLM0+T%*Rmx@7Dm+Iyv7G!BFYssanpV{(Do=bMdzZ2& z3_n9Xkm8)TbCmupiA?^gklbBvqjx-1?Zz7Nif}?s(?y_a zF}mjCup*rtM^I;SxE1F5dRbb;Fok ze`7zO-cgZe*DvGi71_rj3fn+x5{oYI=SR9z}STIcX${l@ySv7Ag;PL?AZYHeHc9-D__ z zg1&&ZqKtUEAjRn3!q6)vgFfNB#cqk_DWm7FuMJjl0nF%n5cNu|ek^Y_fU}l%E^EF0 z_^GYqr)upYn&2;H!B|d^LQsb>73P@ZfjVQvD{BJNKloO%I*MH`V9zP9kW}no6P(j$ z1x;3T)y5$XElcV~zh$mH0{5;$P%$+}gBFF{1VC``gZ{y0u`nUG@&`VKQ*=oO|0i~J zi;&d0uL}`O#Wly?3)*jmdkm#~OTrJWX^rI_NkRO>OzGWK&X?`QSD`dR)xHl3q+n-r zInFw7ZsV0_a+w7Te&u~obx1#2>DvNu2)iZ}5l(A?pips%R?g#9zB|HRD;3ivxOo|p zPwn1A}MxPive7DYnPG;uiTsdfH=lD%PH;(gv&=H(uPT399;Z`EykG5*M>c`gwMXbO(#ms+lKYrmFIu)fo|c?zeYqC zSItJtWPYU#8um1XWo1TVRr;DG3@3O^A-6!uS-!sp*+A%DCfrhVJDIdBD|1aE4q|=W zE!HY!z8+X6Z5xbPtr*D!Zr zzeWTo+L#I_v<^&4_vc4`Lr*~CTJ0?1Cq;Tb=M3I?^kn&x#Zyx2~wr5~n|HhDqX$=Ncu&o6zF6AKy)h8Ee9pW%YN ziVSn?Yt#)J{L_=qEt_YU)YyfDd6<`c18vB*2aZ9XKAyTfzHAg}8fwJs0J5Z+vVhTw zTOYO6MuBa&!dD|5rL_x_S#u6}^n&Q@^Dk`Y@}vB3Ii~%(E)A#+KbVDsLmc_ydCKE~h8U5_RZu-U8m8h&P6}L{7;q&eKDuO!E`9qd zJn#~{FZ`;FUHOT3+)1NLEi);v;QrHTz8eqEJq=$qMNz`NKa}~=)Au~JXq*DG(8dn_ zVFpNk67H42i5dVJ9+h4DM~m{CbsbtxI|)ZxgsMG*%zDZ0CE^*Ynps;QfpXRyjdGM- zhbKaeG<@CTGI!UwjQ5y*^!D@-0N*admQ1Onw@5~R0G8w{mU3m>G$z+&T#mINoEY$O7mkkf!6~Pni=Wc%T zV9ctk0mo)8jRh@cKjsa257v0?9U#Rh6<94&eZKb)TH5bLWSO|NVCvWlMV2agOkRht z1AC}ORQ-AyxLGx1UKE~iRN6w1R+;QXI#WnyWFEU;B75Fg@Ab6qNRaI-ZM&qAnx?~F zN>Cu8S_5 zA3hxP3cvHQyny-$=xQ(jU$>a^uUiaKN=$h2w*hZf?Wx~lyG_r#1zq1d^>#!TD%rml zaBqp!Ex;S^TXw zR(Nd0$XndgSK{5Qjv0ID}Pr)Rt8KAz}1m5}SYO z$BX^@c-FNt2ZHbJb1p>a?&m2HiolNksSXaHw@R&onv|~0m}4&C*zTdl@F_hJy^8_M zO^!fKiT0N2CF5@jnyrn^MY4#4$cS<)AcVY~y4GoOKbjslbe7!i$IJ~UV1CH0;-F`b+uZP1HnO2Lwf zD?(mA(Kx2C!U#l7x(WJjiB%hFJ!l3cSDiq4k46$vosOFlu1&GU-&8atY})f2pc*&) zBgTy}i#mV-J`+a!N+3zpXf)6vbclJahKa9bZRqSn=S7&=3RWpf>&h`oyis|ya@Aq< z30jNR;}Dg#YR_{6$1285-DK1VTfouJrXvr=ZS1_KY;9!go7!EfYNHPt^F$O&9~i!f zyhu{IC*1ETJaV!27f~Z6n3`(7H1GuZNyup)V@N8v)*IFRX8HGdtm8Caqjl-SP4?E% z$vcjFf5-E&v9P0ZCr6i%id{#Mp|}vbL z)3^S^z3~b{7dG>G(?NAjgJ=6myXSl1Hp))xU-F+U#sDPCuN)r?Wrk4Cme2vDsyLG0 zwl*XwP1SOHF=#<3$R$LRM;rULTgzbD$e{Nn9VT4*dM;bGPFg7Kw5Id9xDV5pN1_Wq zne=*X=ndVw<9A2EBrv-BAD^j*NoAK8w!T}39OY-~JkWd2Gwa}Pet(HKXe*6xu7B9} zPL<*W>;Y3BG->26+g6g1)?Wy&>w%IJ(D&F8-i#ByL3)Zu-@$5Fr>sLTV*GGE8uimk z#7`b~iRALzIW)2mt{erXoy}}96`;AcK;UiSe4+&dR;~=W(erH5+R{-gP1_&6MX*1K zUKFLQX_o_Mn98u(##uPp?zPU^W@e&_*7n^d`jE8fEZ@XqFgK}5a(`9WhHP7VQDVsP zAM+ZQSyt&@(ZwxxbF#E1J%?n2)s=5&9rfH&?AwwPaUuA&3l zM13dw36mctH%)MMKlS(Ncg+KN!h-y356>%fdu9n-#YY|WTZYzKc-_7j;gkZmN30-J zSkEFHFsEf(K#=2%8m1RcxHn~GSJu-i^DjPD_(ffz&& za@NcEK)g&bz<_;;J&e7`^U^6@NZ$i1c2kiNX^MDgFbpqO3A?Xo!q<~d(Gp1J9$=O-&?NTI3S5o zz*cu}p99O@--coR-+&ZAjhwjkT%^i9aE;@qYtNZG1|_uj8q60Xa*b&KfjYUH+`Na? zj0NH2d#CuFcd06qkA9m9?zD1TazTsKPqef#Ehn+LRlbI}e}viJ$#LB4nu%NYOiD6! z$9lD-U*nC$){|{#9Zj6;DLU0q?E_0tHod;Cfy+&)KmYcQ5p0%22TK-`Uu%~rw%Sb9Fayfos{l~iLYi^{+T!1UD|MxfNhnBeu^3gZlRQ5>xB z7WQAbrw4puu5Wl*pQ=CPd_rWdYcVK^3uP<0n?YaTW;k~(5o1VDdZnsE!IZN_D3WEW zJ@L4IutsUC`W4`I`Ri~eo~F50>rK7(_P=GLbH7){?nHYt0FSj&^HqLawH@l`ZEQ8z zkwjUiONO|c?rg#b)3&UX4X^E{H66^KV}pjSkmTBoz5lHi<3vsviMI3DUMr^I)3BVe zdovUYr7nihm_4Z!f8&u;>!mM+;$TTo!KbtVFRrjtaGojatZXE4~#5B15^3WpJ{tuMB? z%P5R82UeD-Q2s9*$~JOpBf1yU2d*WIoSSDlyFFKL9;o#ILhnh{MwEE}vWo($sy!lw zon@=fgJCw2vN-K1vQv$md*|Gh@VV<#bHQd5S z`-O$KCo+G_$lNV)n%{T&wL4$ja`-l+2>oAJ5%-~y%Ehn}Y#|a06J408x4-IY{)r0GgT9I0}h9dIG^Gf&9I|oC@Y#wno#>T|NQ9#_Tg1lfE5W5iZI5z zBgA{AhhMp!syti*CZ5h|zUJU&q@vLKEv)K46PDV*@<=owyQk-T;jVvhwGujALuY@a zH`X4as4>iHfmGqV!#T%k267y^c~0<4>Ph!fow zeoFkUq+;u&?TJ%Qohyco1Hbk>qFgHl*XX7(hpLbNdH9=sps$ehxZ$ULfynUcc|pL^ z;5M2sTK^9rNx>Up=jsy2MnCP;XM8o+IbGE1GIt{5d5VPc)yd@>!G&h16Q3$;`&hG8 z=a(eyZyD?WgaEd3Py0H?s#UdR&9FJfIJRD|a)l`KsbyjNn_v^Bk?Qi>qmJ5;Ed!mD z4c1Ofp|s;l%jG6z+w0 zF-F9ZIII|nIqYn__@KYr_M&6{$jl_gwFT|2ONvPh((7B)qzvSkLO5Lu2eS{m(1pR+ z2vM%=PWlbr$MZ;}HD`44`a{k2`?+V6<=mM%caF|ALG5M}(-_fSAn8?1f@c{QsahNK&E)aAI0q!@>DRP|RTqij-rmVV_obyH96_2o z|2Hp4kBQ8QFv{8bHSl6V0(mh-4x7g#y2TfI`el+rjz-%0$e7+-Habs0HbH9viTwJ- zTkC|5pkp1d&GEv1$m!8-Hn&S;aPJEt$X4@jLmg&+E!7P~pSJZkhu0^?dShZWu~i;+ z9chF8d*k2sj-{?F#Yo+|bOKatq7Lu?!XZtK%>+-QnTUC@qEf!Z1 z4W3gOd$u~D8L#c!HiE8O9*MLJ#>Z0cHaW-J1KGT3x!lKfao!AfTsuO2NFZz7oC+?b zU_OF}I)4JaFl1XTrDG(O`s271zntMhd{V;o%>^_sdv^L-nf}IriTkO(+Bq9#<_-8WE~=-mfLCWk(kmI2M=xZ^DTh8QzJfp&8~aPhD{4Eg$bb z{tHvC!n9SN)Db9zV{w(}L3pi9JVE3?PJ4L8=~)t+=f)p-^|&gyVxaRGhw;K{6woWc zq(^NHaiTK|^ud0NwbK+KOYrGxpgdIeI=$dYL7c|KWj4-ZIL2N9dQg6C{i;g0*QTiQwGo5SAFmzT1ox=VZhAN+jtKwizCt&r=AEsN^}gt2$lKYLLAn;43pXVqne;pLHM zud~8d^Q!S@(W5VZnF1r_0-ts}C5*r+r#~}CL|L>;E2h!Ub(_dC`>z{xQ6a5nWfTJI zWhrE8H{>~g7R%JyPfSc^gaFe8bPZHF|1F@5L*XM{gC6UIKP!zGAr8P$hNikmRCe}p15okN*|EWb4MkF&;3-%{- z@sPUh4uJ|XIh@Zi7G}*5qyLNi)sW=Imz3n{SL`j%*s7-56DDulY68n=GIVA5!YMK^Sqz$KXCtYo$H))eeZL$ zs&5!=0mgUJv0c6F_x<)ite-&BsblJP0_G!#%sUE0W7P1j?e82`v=&zkn`GTPn|<4JHLaR<}W<76km;!EdO*+qqspC>S4q>d&FJ z^>ip`uQ~|gp6?ig^om(a)-L1FAnBD%6T}&MCGe*@qk@C^ zQI(Mg&QN$h1{{xd307>f(AN1Ywwi3WBO?0@B8KVP9+U3D%QSwDM9b?--Az=<)qjoms_UJ6uUmeq- zoFMFEQY#!wdi(NLgxSKzC>j@HlSb?RE~+AsF#k42SZOwU zprVa|wAz9-$9yl@jxrwJT-Vu|^Ii7;F9B-!Pw~mlNec!7EDeTshCuH{pUZ!@*_quX z{E+zgx2^eS>MV-hr)|Ucn}F8cNibXDo}05XJ6x|C|t%-j(^?^2812O%Gys)5aK zcJzIM}B14J6ZllJP!AJ^jaNhy9~m15p%iz!pL^Ryn2a!Z%viVCvFRFOvOny zNo#`@;qkM5Kft43WK3G0SDn2el+#rvhGl~v5VTI*KB1e;p@nTc$7MW^fvI*DAG0?V z0^Hvos<2MAPO^THB|hG1_*IHUl!Xfw0c4A1YHDwjTzkfDjv9EBR4QTR!MNfcw?R*4 zPC)sGhKu)&wwp~oKCEo&0tTFxZ+_y?CY4!B`v7}W<@u8KIqkNzhZin17Df9qbN86J zXbw6Of;W?2a?(G;ix{0%y8s(pnLEeyxP>Xc-!SnQ)%@Gox@p{B(< zLi4;Hu_vFg_9a|VgB#=8D&HA;3N~C^HY8%X{^V@)!FBgxN%bW`%ziYmV-^E6S$VVj zjzikzkbq7X05rxwyQ~Ft?h_lHQ<*!Ju~{4iL9zakem|Y=pmE@Sv{wH|q|)b+pZS&D z8<5>(ehZnxQXWhW5AR{jubIfh_6)CR^5EUt0kX2$gtx-{CVFEfShlL>dE~k_Z2xD% zEn!|eyV6KerO6Gk+&{QBRU)I`2Grrm0YOlDbQa@`t_#p4CrD|`_Zp0A3l+-otCDaH zhfa-kGToqLfMmD~x=ioE2?23?5K#G#z>REMkDtQj=I0?Dw0mr533j51Wffl!(i+ck z+xh$^x5d`<^B&B(zO)?*r39VC0&&qVhCp!J`CFYxYvA=$@M(4=&Nfd`A|Rk0ml_vg8XVG= zVTv0vc@3l#v`*1c`<&VNj1U9=l+;!S-GG~e5%7+$^ZPXqRCgQ#zG|FVXXW`h0cZSP z`V*u0eOhN{8~blL?%(*#;Zb&4rND)!!f{$C=lX|1|HYOIV6234FI0!0ttY1(s{Z?Z zxtHmr-=9XC_tYpb|5KNgF2gU(G500H+~=@<0KSnaOZbf7m*}ZYFO_f(Z7oiBAG%#= zb;KNw$x7pPt_c$t#Jhq&Y24PwpIDH66G$iPwgFmSw2mVsIN8V&t`h2s`b?0QLYfCGLyXdnN zm^=|1;G)uZ)n#u}xV_42srpV<_jHxeO$13^`Ere6VyKZ9p$ndpIZcZj0jzQuad=o( zyWTue`4Oz-A##hgpU;v!52X(#78ZXja`!NU<{tHsiLWRSB^j<`A97l+Wd%332EVTx zV>FbtYI=2;KiP-6@Xd7E#Hb@fYuE-fwqED&>mfq}v_w@C(Of|a8^35g&jom=)%s6^ z*OiX+4*2%BZFnX3zfpc9Otbn);kq6ACH=*mAyV`259u6LnsUtpClBQgXEI$wDtx1f zrsl%Y1c+vSCQ+(>*hKmL8zSHFD_g)%P2$)VL~XyJpJ!+Tw3mj7G`52!eMjSj;zxfJbQw8!(biNchTDdgO4;=jBTsB?=;i9YO zT#1}P&Q+n!vaQCTO2)@%qkEw)k>7ith&41k)azk9SPF~na>@-%wlI0&J;gu@Ff^MQ zrwfA^12M995_yH80UJ}aSG;c}sA`>l`wgQfGi|BIc$s>Vk%>)Fto~hL-OZ!My#SjK z0e?`CCqYlA?~T10x3?NA%Id3=HsVafzu3@bnLL{eT?~^NgVGqA3Ftjj#@SG9U3jZ$ zvq!ZpH0zL}Nu)?lT@X)ohdC2&TBYX#*Bo8dj$N)RfHV-lEi;{<(5T-{R$DF}$5Ec! ztMkjd`#Q9t#{bHo2;#q>2eOU_x#FJ}lg^%Yv*)aCj7+x1Ky7*SPx| zC9ibgRr}THd$tS%iU|lS>G}xp-@5e(U#Z$JHZuFecE);KhC3)uj`6~;{`Tz@Jf}fu zF1Jt(N6T%mTSz;85RGtBiNEHA=8F=&Vr~iCmd$xiWLlm?GxM(Q&&1`W3+&osMq|3u z+<~D1pYD!mkR8!M4nyv))W~W1ArsVrVKSSXNI#jr z-bZ6AEQkOzHR~yWH9)Da)#~YT#iP(`7*)m^cgB{q@M~w(N;Y!+0;GAx7;4rQhYj9u zS6B4obTareYa0oEl;EP3*#4OVqvcbF=RNsCY-g;mk-7dw+gFQs!Z=a7b4j0`|Dj|y z1=BBU$95zj50D`>A^o{ku2(Op{Ca?{ms28BK0~{i3FMq~IVD1wB4CtjYxD$d*Kh@V zL^}W0k8a9h84MG_R-8S1r`sh=5!L?v3ybm@+kG22U{L1!x};wOrzPeFD_Q{`Kb4!I zu`bv|M+_S8(ovd8{I*e>d&>fXLV1ag>=`CM&eA1EXdV}6l@~B zXRbLpP2i)9*OW`TJWb7=DUg`KunGmLU}l*2hnS^|0oR~nUXN@&_@vkQZ~@M5A>s2N zA3{(~>_s zG9n5-7%_k(Me;)j;&XN3d}?WsBWu$YQB?sp-YQwnaLXs1BT_2M6*;&!Xu2j`;N>W* zi=L|eUi*_k0kc6}xaudEPk6XS_s&9^(fzOaO?AXDp8J1Dl9>%nXBVs*t{G9IAI;-l z?J%!s^o(9?DBlf*&b(RwM(8|WTU|y1rVRS{n{&xO4JPPbA?~66lC!_Q2VDDNk4YgN_qb{`cA~ z@t=RA6viTkT0;!lS$nL(=Oy23lR`XJMDp)ikImBlCD~vZYNx$7&(lry2v zdWuLnPCXYH9ZLCWZk`{R#=SqL#5CWHe*ENeI4(OTX>{xP7U|6^Cqq>e>(DsuC@`*5 zWAnSIz=efPv;;w!hz*aZbL@T7Q~RoZJDN>b;)&N-%50LwcZijQxrCHgWXikADh^H# z1pXB~&Gz8CWzf$+x6ikxtY^^N(2M(*P>ZsXM<2J0{{x=Qnd~c^miC$3Z`8Av*&eb9 zv7yS(zO`2N6H98HyJ%Al&;V$Th`F#Tbu0yYmU8?66#sWm#;ZyjjXCWPcqT1zEEJSd z%EM-g5vtEacyeq{uWFpJ1;H_uMk(A+xbA$FJli^1=DdO$=RmA(xyZ6a2>>c;*_PPi zG1l8}oE=1`-vpM51q?s$ZS!nYDxGKX$>t?J+?ij_AT22mzbgEy&AE8$C`39=H{=E= zY}U$hhori@_4i79(`EL8mW(B+Xgyjbcw27l1Mk-!0>lP$z5P$K`USExJowU<@H;xb ztl%UMBg^FG9Mozbr*4&w@(9hEci`lQD9Oc~R#q)PRxglD$LM*DcP!F%Tb*red;i`^ z`hv8ax1{GTYVYW3Y{V{uu3|>JM3ur`dg1VGNu?mR!=B%KYT~PvtnkdiUPaI=FWMtx zfh>WuI*Q}A7E$8<721sR>vu=xGc`~5Z9i2zf?c4j@PDQdr4B-Jn~8#IHSZxiu&5%} zd%d2XTh;kw^(`0)#vt-FGFm^iiBhwU-y+G5!G1Fy!R)ZH20ai^>7eTe%H6LeOyv&U zqQExw`Jsg@|6GLjT-EZ{$#30{!$ z#lO+e6)vWIx7nik&l%A(XGca_c#aFN;2sH&sl}0@tad?bR%`U z1t3~ZVqxp$>0XlvpM}^=+{_z-cB%*Mo#i^GaP6=NVkC1^7cUhNfqQDRpcYMYN2heI zmp)KdgE+ZBnP?4=+{!BqhGu%eppGbIgC=Mm&XYGe+x3ppb*<=jCkM?Z4_#uTJ6Cvc z8i^glvY8-MMA+Cg2dX3IWDTY<3{X|SIyi!djI{8{l(|Xd}o6kDBMn-*CCF$^NN9`vWbTZyG;JJH< z_w=o%&s`jrJW~OSuQw+02h1ejl8E9tlRBR(X&$AqGr(%B+!WG+;A|mpb%|1+MM^Ji zsFMGfQiWHNd0v%m%@yp{xR;WkD(ap7=LeEXPdc*0U8vd(nv%O0a!WRMKWA*?%wTS! zU&1xou-WwtDG0Bdd`s%%3vH5ic6J$=(YEaVo*#rI-Smtji;h>%4y!-?RJM8G`oIG9 zZ-YOAJu&iMf7*K<#(8g7O1W#WHfb@=T~7HOA23b3Go&kaCzF3E<<3a5_uguC3^$+d z14~ms!}{&pUT(o#JE)aki^cXg2Jxvvr<_@X!o=iY|C zMumeQ&;Y0b5=(Dn?ATrHLzPA-~7qBqo97 zRSO3JguY&y7r$@~W0LT$7uNUb%-#E~@s@h@66Id!O-%>W{oWl(HYhY|sIYJk@YjaSo`~@YS0~XY<*b zMax4uL1r1+R=a05_2`*e8uqhss7&@eWQsv#=#TY2Vt!ACJxpodH=rd9J||C3X)7F_MvsT)=)r#Fp>Til9ezmy185GbqX?yj^;_~N9GiJPpji3S(U5KV zP)6}0B769Ztd5Scb8bt^KKt{A_mGhV&BSjrvP_mVhkg(gXqb2*SQ)yAM$_*1 zN#~-+yA`L@hIo`W?bHiZto{8?@^=gB6c@lUPO1R?YO6V9Mf>C zGL9d0ujFJ0W9NKQOS*GkAU&S8G`q^}m-|*e=THA>C_vXo!WnUqvc27^YM3vKU}d+P zh7M{{k+Ly|ZMi-(oin3Cb(#*}vW>Rm7#LvV6<%<4H|s$#yCD^yHs9l~H&5n^B1n-C50dv4Kh#wGXhwVp#*We*@GDRxe+te5@iezEU1#!C1};mY=lhCR30&;QD;l8V8F;ylO}@ zFyJ640%~%lZ-f)$Rp)%3%>K!@mgmjw*JX@gkm0K19f#u?&YT}6dwv&=-yF3V^;iPF z;t>+^tItZwN=7{MziPdCfhz>L%J>i!0&HTy1?y4YJWGPUEO#s!*NT;jVeR$K?%7lz z?Dj^aQoWok;33DiAKuZX>Ab_+PikdGHR3;Hof=;V`tcWLn&MU;C+?R(w%d=`pSX;i z_X3G*57r!_g;J*;L;dm)LZQdN93wZkTWaZ$EG+$)=8hf2Q20#ihRqT4W1jF^w9`mPn@-ad)UPcLi&};TgUxNp0C+-={rON zy5Q;~OX0pMu^Bb_Qj3AmV?zDiLz%ui%=!GjAU#WGu!)$Kg6~8L8|}@0re>YtaO9%t z+L*E8+U{w#BG|E<+d*6Vpk?swg~ZDIYPO)IfH(HtT(6JqbFXmQn|Vh?%;A2-IdQ8U z6bgFO6Qxv+i7mCYkzYBi?q5{;utcWxl?Ff2^27EmT?tixQUU!?cFyr78tx<#7%sp6 zvAW^63_Gvz+MgA+!x4!ZD7f?~Cb_IT*2fcmun@gNQxIh8Q=U2n4#vI=v`)@vE=U?I z(eEw+Gj16kR;OkH=`Sl00GNuU|9Y#92Ko4>TlRAyg*hRGUCp&^!)ZTNapH`eFLl;o zA^DixK*lj6EF^#O+lsFt?!!!3)0S4E2KI(~X()Qs#G@|bF=t8B+Ln9C2aDn7g@Al7 z4->`NXO{0+18xYaKkM!@U+$O&!}2Tg=q%2}IMYptm3SIy%7oGRjibW#mlOfvSu|Bu z;)kUU3i;2FeaoNIL;6ElH5Q{ah-C&t@rYct+#+!~LaL{PU+lufFF_($Lw`sT-V%|x z&l4^)%Hk%5S{^36{0aZ= zIZ)8}z{E5U#)D?uxBe30{}|u@3XKYSt5caBo~E)tYqO4`%?~@M!J$`HS68D-DktKK z08N=oV)qXtk4A4cn>np>#f05fS~VwKleoVz8>ENBLsI&T6i05+y0yICW9XuwE8VQZ z$|t;DI|L3$ZRvfYggGM&;O)b#l85^Wa!KV5?;YYCxC1tRa>5I>FP4&D7^zYCV{pfS z!m->UmF7-kgkqP5ufxEiE#3S`WAU`HP#5XZi_?S{#N&hSfB?l9#j-yc5=K?@dD^69zcre{t1^JW+rXg zM(BFtG~vp61Wz8M73rGZw!YrQ{J16jHHZd54CGJg{-)C0^s&%Ty=iL`df({s63I|~ zY@rkR%_^t%9;(t{{pHYD0R$Gk|3-1&QMXP1zwncPjV-1qq{TVhlJ^tK|5O$l?d-ZAWQ zO=}Z3h3h(K4L5Fz|84+cyWv-FE#UVhCitZPiXE#2!D|?NieTFYomF=15~+y8l(Hn@ z?^I^H2MIdTTXRY>8%!@>*~;b(a2&a@ovhxt(!RWd%8& zsJ~WVVDE04^;u(jAY29vR-xsiJ>}Cp1DFP9Ei=OGK0h-0^OM89S$53NzLevS|)=w5yYW^l*nZqqr8|V=FEzn-ce}&yNRz}7Di`ctm4le(hFS`TL z8K#ex`uL`BF~c1Md@PD&f0=x~0v1J(GHsQD((zhn%>v?Qr9f2!y)r2f)@?o*i+10r z3AR0~;n=_uM1w(o@Wl#Qr9oM|ZyE62w6=)Xes$b!yxEKTaV&oPMTXnK;@hV6fIcNG zU496vU+iFE>BO$8(XJ(zq4BP$maR?gvlCh_CRZ1$fb`XErMEF`Ev{Nj_Y0(TFfUTQ zFXuJ}{=Ii1MXq#K7%t^Dl`+2AtDD0fC^6i+?HSzH%4w%R%x~=v@~bDn9{3>r=FG!A z1a9>PQi})~!yXBNaaU95+AhejhTJwYG0_A1x$1kBo4h03N0}!Az)(aDCA7>6|A8)( z%1=7O1_57T+=Cx6+H}=L0#4goRy`ke4KMJQ41bQZPo>PNn#O?ppr@vln|0M|8M~yp z&|kxoS|>C5PahRz1QIOWgz`#mA?G+;%AGvtK)m|yIW9mm{cT?K{?(m~i;Urt+MCsV z*Q~}?AFp>mS{FXRTuiflZZ&IFwI!c3w=?mJAe+99*^E5TZbJy`+#FQeu?^)UT~!S~ zYD`CbEpgg6EwKK&DSZ{GLVl*P2uyqLMI8u{n(npm>rv9&1_{b3r|{NUUuINfkih-~ z))tLPkjD_~nY||_LDaq*pVmg*JQwm5v0oDz^GFNS zmGxl;P8?0^yITDe0Wp`;U5Xl=n}w#;{<3gV%=ftEl@uYlY@yz00H4TxgJSPFUa_U# z5~H!NkJTeD`4+n)#8LiFjShLMikts?VkM2(mNjX!B;Uto|B&PcAEnUE+C4IsP)2Qc zJ2vP(H?T|Mhj63pW{Moq++!Eq_}nRObfuS2<{H7X(Gv@*!Uj9rDt8WBdjA?po_MMy#2Zj4pJP^yG-X$>r z33v#4wi6KI9U{kcTFquJT%g);v1~~8odw-FR`?sS{+ON^6>iK1v{JY^Hb{6H`R&<( zS+)Ork6%f2i}b(_M{gH3wa`!Rojgh38xM(uWh+DQLa-pOczA&SreYv!~(v}IB8WCgmJA_M7DD`#*ZD{ z{~mXJ?cPrDdiu&Q7yf#$sem1r`ibj_!Ihcrr!e+ZJiBTgl(S#=Kp*!**LSQUYg<2{ zt3a70E)rjKbgMuE3(3_;zCb-a!6+J-^G%)3t~k1#@;JvSmGaKsw>n^M&#^ddd_hUr z1ioHuiS~^)Vipa=NXJo+J2ZiL$1H)Obz!KOoSEAH>XObm zHL?=cA|62ak*j{(*CC~a<5BpX6#>gm#c4!RP_M^tFEyci9yuI_zw$Tk)5I?M&VUpx+9Q#gTQt@XsX?0XSU62{Tt&>@#xX;>fwAAd%mf_Y!nLh((qB>z2phP&IZdF(ojluxP*(> z521!+ICAjuafntBEb@D9y4u!n_b>7YzB6R4%i>KR0nKIQVbWU}mb8sFwNzU~3I+6w zh&Os}{hob-_&!&KCZ^VMJ7HmAG)EnN;)wB`7d5!~zW1|7->v*gOk@Jm=0M`Y(`~?+q zM8+IT6f?h~Xx!Ae{q9=2Ee{Ytr(AleGXj5}GA=P1TsIz28C(C#z@FYV(v`hWUy?;m13 z&<~on`iB@7`6!=Ve<)d{m8MnfYQ780pEOTnBOV65y`5O{NNET9Jx%jK$t?|K_+<}j zlGnBVp>mO$!T&67_E0`^w`)YrC$&E$O*18PzWgp#&0R#fYT+uF*V;A7V5#o~f{uN? zZhK#X(z9N5aRyq@84J|q8<=JQK|sF0j9HJU;UkP_7%R#kIK0d`_Gi1UhU=+Y*#`R*{-id2%a0cn(B`1!lB`#v;T=lB0yZ~D(+(vqLhX;8rmQfgHGSnCarwBDWu`FSm1cTp z1G4%;Nz*)bVj)fh(?%~W$pCZmtJRh<_1B8jmp`|7)ydr1uFO|oJtF62u3!t7!LtD& zY`O|9vi(_;W-$WGQ&%Ce0a?pLA67gwPW)g%ipssZtL;dB9?*L*xF~5V`Q8sJMy-Dp zad4Kma-*MdKXiJ9!%t%K_r6D&Q+Pn&u|5OP=sB-f(g#Kt(Abqfp>zIHxCXV->T1Q6 zmxrOGSwZK8bD}%sC_|fHy`D?F`pkrHVd3!oSj&Cj{dA|QNW{uJ{$A4VdQ8xMX*l@I zh-Uc^`bj1!vin2ioYIyjnZ$ZD&y`@~M1e)Z-66tQY z*K-{n1x@b4k^0V-x)F8O55LCI)7|S()wR*h%&&3Zw%k-^|4Am2})EWt1@dj*}%`BCY>0*y@ zUZ)Pb_i+NpR+Jl8D9XVumAY|GdpcVTq^^FcWoCVa&ydo@T)49xiuT%z+Rbr@zh%{J z#LPx(B(uwTl-18(u4uHSwnkfDt#YfnN}&q6;L>@0D3_f2d)1jEs@M@-MEou)hVkxT z=7kp{QmdpkVboj>oAq1RQN%yld2-vk_xQ%x-Fn z4(!u=$Q|e!6I?sj zDltUx(!H#IyJfzfw|Nlza#9+n*{MgVu4UaS4z#r5)-qOb)4S?+`xZJyKK0;Z978QYypwNkK&>t<0w>YkO$FsCrTb3%aG$$A2Qs z;B1?q#>)8`afC@$zxffH9Ge&B>!H>8^@gONpQn0%YT0=b9&JbJ2M(6)teXXG6vNPC z#wwd)eQ9t5cx7TWL2}}!19Qf-*Esvn(!%9Nmf;%f1*zUInyD>}q3ro4ZGD#s!RgYY2*)R3!aHL!fYnY^nMbS58jc?N!bi)7Hai{JSd zea6<)6iBSk`X)FOc}{m5Iz{U4?btjAJ0<24{$lpyo+ZGKOqJOrw-Qn{4-SZod2?$q25MUThGD)p;dBGkJ;I&u~blAUGkcb1T-1-9FtWHXptBHT0a=X!C!B!i@#BWdaKSOJLDL5y=m ztI+$Z`Spmwx^TQ3kIrV!^@xv+wK>&;i--dGNvp0BA)0KV>+(^DR0*`n$OiuvWf^1a zr{Xo($vKo&;oM2`MeSyN3$bk((4pJRz$(B{-Ky`^APKY|_hqt?-X- z9qrNdrwR*bN*{l=5u(c|v}J}6|}Kwu`a(gdrN?*Pn+gXx30C947+=@a)-&oIi}KO;D8@! zQ}Yp_gkg3+?OXaobUzK1?nSf*9{PdJUnn@}4!kPM_}3BveB;|hGJK%b24;fObL z57Yyr;_)9li7ko6U23hp7K)rDsQ=N5oBwFVDek~*%YU@OhmR8MI^c$^pRaM9wVxj{ zmJEDF9djyyK@?L3U2Mj>Mp&>jc__1 z{hbIuKnDvL3b7<_e_mgSx}2zhuoTB+ybdYaK$tm8ymo{p7KFkmo;-GG4tI@CuO-F3s|Z{`-V#BpfIh>RhaD+}@((Wt_s!VkuQ`Xs_RUT#gkOnQ@UL zcFrL|lS_4SbM7ooe#JB3uN>Jl-YE2zN$vEH^088n*~*plxK;)J398d}DZ1e+6{2gr zz;7vww8(8)GjjG2JdO@uT%PYJ+Mnocog7uVEXT-e7v)@06m) zaA+{$4+mSr`xtz_rkc>KH#JHTY&LqVDM^+QYJ9>NjXuvXeNcW4#7vRI=0Fg0ow$QJ zW^sHTBl8bI-ph~~qS|s+dwbZxF0_B)sfZ=W+6nQlh=+Ghudk-bBNVz2r!q`ReO_mX zB7fxdeV^^6&^daoJbK8{rH7n_7_oD^9kfvHyF+4=E{&gdPgj2^NqG4D7k*smipB6$ z5;Hd0X}?A0wSOVN`%Q)rFvb<cN#oe|G_(1xK^iV)H_Bo`sTc+%a%~QTes&cD2cQ6W}u;_i+DyHpE+_I6SaHF+8 zu_h6w|9R&IEU3julQ}kO^`8p9($k|*YS-z8ITN`A9ybcn_%ULC26XftF+$OL-UB-u zqUA91)Tw>as6I%TRRluxD$B!EHPQ7leJo4u)_t4I<@4ml>%PNSJH-+S+H3#-~ryfd=lW;FF71W^8^`NauS5P?_ zSf$GJJrvg7q4DI;^D-P+wN=N#O|?%Dq%icC=!tgorbFg(co0|2JOpiw_TPL;Aj2CO zx4;<)+Q#Ei3$$8!Jte~E-RbnuskEG{b>va^m&d33KEZ3JyZy=hxJ7;qQcFcq%Vnj| z8%CAuQ&IiHYF~^3=d{jC;hg#>&5%0x5-$%U&qP<|DN;Vr5m&2z?!V9Ei&^?L4(fYc zz8pcGbrzv1_Ul)RU(wHSW}McRjTx#y^T{gCh72v-Y+4)a#A5g}s@=x7O?Una9x)#Q zLm?E=i@!Mh^sMzLmG3mcXYvht^k!+_NYQtq%e%)thbT21S_4v|*ltU?9SKOgLvVqx4Nvk_D_?AH&5rNm)o2KvSR-hwra;-o(ikYV zfOAQre|T;{*rKI`h$Cr^E>vN#zZe7(d~aVUqAmU(&q)AO@mYgLCA?a%rb-02YLt0n z+9ubqhF3c5>$<t9-mbM_d`lso_DT__7t7o=iM_PyflA#uzu!kqEs|Et_{xTP`C`Cabg4-q({X3aa zQ7A8A7_P^QS4{V?^!LvywV!H^x+o>)sfj!9OXhFtrE37Ezw}<}JucQyMRK)Iv#$=y0^={&LaQ3si@ z4GEW6EuN0E=%wGtU!T)GqCi|kqLSAYygO*u!(@w>ry2t{m9`w;8m|(+NCHj$JYv(8 z`;0C>O&nMZgCBgvN>P0jRt^=4{qi)%?={!D1kekb+>UE(088#E47A}q^y9xhoY(NV z5SLmuW*=D2y95~?39;RgOmQ;S|ErOj3%k>97Eiw))eeAhmBSUM{5Q$$1n;KJ^cLU~ zsWV~r`+45h8zCcb^rDLG#*@>2gyID$2x2O5Z#B2l;n(=C>Ac8pkQcoJbZ>4%rBYHYjK0mS+rp;R1E-Oky;f>={m}Ope)B(++;49<=vS z$WpSN1YND3%lit!^OwaL(JxNjw7p1&@yNl%ZkAs68|j!yx&d(|!RN7G-(w5pZ9Ok0 z?$O2F$_&=ICWr5Pe<{Y}BLdCeRcTQ;yBQ%w7HhGI?z@(GdS^IWYczw&(qSY{ORHZ~8yB=OSPGZbzB6rlF){Y_5uGIg%}u z{4ZudT)X{Z!(56JQK#pX7F(QqS2ifYk2?d&kGwkl5SUPhJOzE!^R&mB)sc?9NtT3k zDk~a>qU(8BlA5??DSK^e%mRKub!-h|CZdfX5&jJzQkQ4#33EmQ{^xV1eFl`j#Jr}w zyj`6YPT|@U$F3JTs5=2}0?v*gsusMP_kN8Of>K5+y+7lL%5?tEED;Buj5%@O_Hi`` zwiMt$(gZ)Gg^LD%I`#VG3XUkBAp$e;^%jTJiM01wIHT*JdY6iAFf z`H6{sI2t$binr7(oVLIpvqq*`CG@S8qA4aP(r!PA8G(chZuY3UK>f=ELqlkiXzHlo zXkyFUkE3b%N5K=uam3vY?Tt7RZKmwU^Bh<=Zt+A!8d+iyAz4ApxTz=d+nS*J%J$Eg zzc|0!X=@2@X0|1Y^?|RJcVBjMt&Zf_`>>Yd9(x2w`cO`{#R^A5ueR$;P!M6V^ zK{d6kPKye(RU~u80p&;6>}w9kzHa%1FAAd9>Slh^H7bOmpQ$!rohNt{C$JxYn}tUY zjV-TwkS2}1lys~6egIn(Q_O&V_1fAhZ9z|{D}Jyg^JN&OEC`b=D7qp`_Hdlvq$R6? z7Sx(aLNz0f4WFkZyZ|vYMHGc%tOI8_F73WR$YC-TD%RJ2GYNSj{s#o!Vmbfy9(w@0 z#<2VO`=woWHcq;61MdAJ*jR)HG)G1AUquOOA!OJnusk*+)(W})@T^Lqx_=0D$~%d- z*G(KUbAs1GB@1r3$?<7-Q~u-1W3}Me$!xMkUzvxLAEs$NuQ_qE3o=SIFHcY#k8CsF zs%X${>xP$eCw?l9x6BpnQWhRI+4|O7uA8YbmIFH|o)0A4)9?0d5VDD;{Yb3-wBL0Z z-qvs-RdCZtV3cc9D4mhj+|*O4_;hVWk&b8)Jo1oN&eBbF%logtc>iLE=ULpPyk;!o z@99;W;iJ%htG$w*Pp~1Li*vuA><8UeU8}KCdWC!(sM9l!ZlcJ)k|({P(tdkznsYY= zG69LAq~!S%&;B_7Kq8pmk{6;hXb$tTCKT)w7~pWLBnW0Gogx<9I8~kK40WdAr?0n3 zrPx70G;-LZp1>Yt9X)wbDhYX_=zp>)4ln}RzJ5iI*513p^%j(-HqC7$6vB;4ZF|HW z%5{(XmIr=rKF*mrs>N!?8Vf!E$S{Sa34DaNE6Dl_um5yj1BPJ#`~j~)IR%6^tJG3S zUSssIan{@@^;D_P=i50htG@Q~I9xJrKBMJ|Lj2%dn(F0X9+k6kvRRKBZ-0=+zq4?; zU*&H;6|g!RX7Bb=mvZ~3NCw-nr89zQ4>NrD7Lr{8^D;>Ef{!s>C0HfxdG+N*2f_&b z@zCjBciO@|y*|^NVdi)C=M2~kRKp)vs2dJrvu2AQ`g*vlTx@RBWIt`hm1J5yoIM`FY`9RaI5*cg|C zY3rEw@Y`hKr&s@Tt?Ee2T$ zDJkA18-MjkFPwMsB&?p19xmTy;Wyd}3qQX5j!@#F5~|}oD^{zstk#SX8d;)B;`Faj z^^#p~57C4KMamN*A7^&E07Ef#zBLrZp|rZ5j8tP!BF7YtHl0@;?-)mjIJ6FzY?Pn; zwjOWGCZJs49wmqqFWn;RNYZn8hoXG&F|uQ{B9XuvOQ-uTY9__G?7bOzOSa|b%}q1+ z18lb9Ejq`xjQ^O07F6AJo#i!~dlAjAN9-OIvYB4+wOBb^$X`7>vXFNFNtiSR0)a)_3WegJ_b#$_hxbPz2J4h4qJuldJ<0h^G>i03wz{YJ@6 zOq3&7X5k8Y!3MrJg1_N?W+!F1!G3{671^m-|DBY1CEsbGH0DNx=gpnf112heZa5o$ zBo(lwCOZ^iq%ZdgF@`?eNaTfdeX4b7q>MN6buEM%-$Q&s`Z|NUAUD*+k=IhndQs#RVBFi`}Uv*PDG=n*q&El|6E84%`#A*N7{j%5m$? zY3pNMQb}@+A#i`w`lcm%;%Icwl;FcB|hkrNWa$oHsdTLDyV*Umluv~II z_hS^MJD%NXqlcHV=H0BycqEx?o)BpyN$UGnJDbUP7ME#jxXNGXx(DK3n>EsK6r2B_ z1Tf6|r%AqiIQ-ryRw_n&ZD8#+dye|9J6b&aR*&iINAA)CdOiwi?=9<8~qV>1Y)X$s;?9EcWLlHS~!_+3s0 zKY6G)sl;kYun13ExRW1)U?E`f3m~oX8cHbWRTtf}uU#Sq?ls?v2wWGM7caKu%72l` z$5`s-KhyrK(E9g zhc!-XC91R#t-=v@S#AjVV5F5B*tc|AN{YyKZk@8Qp8|3{76sjBEk z6QZ@M6g6TCx0c$i4tumHy@^fj*-CqxHEIi0RF$YbBdIp1k=SC!D2W*)My#j({+{Re z7hJE`bzQIXd7tw+=e#}3O?81vm^v|Ry1=ah0dVbny2lq&UFwQsZg?d&?c*a(!R9YI zNp8n=`sob#qw4aC>ToB;`2UQ5A7_)YHUB33$&|>$!P7%%2;hB zo8cX!m^myR|iD=A@h_sfB%^?2oKv2`POK5T{KQ-xyC&YpRoofUx=NHguhg z)fLHdqs%*Tg0K}s%*rY3vSuV$;VLM442qCfVm-bMpM0Z zwfAn_J0z)fZziM@KkAK!QKNXy>Dj62W$&2JlQMiRPS!T^pp2PHYm}C-)dSH=h$%Xh zGuZc&XB&O=0u!$vuMzKJ)xm1oU^Ou>7F3LuHZSwY@wD~iME>E_GSEsDjM2*p79Drr`pZ%- zWNy`%GS1efZY8O-nCYcHl4uX^FvK#S859!1riv*BqC0Q z8p}6m8CpA&@_aujI?gs^NlXu{fM?@?CG;$I$&9&8gdvDv>&qA&FVo$WNMwL-0_;fhw}dIb+u4YdW0z%`6|na0bR|& z;3avaO67D%Mc28q|_*xe5U=^`4z#1F8^G`%?(W;r;gXN$#q8mI# zAOE!ZcKIWym};6?@F%0nxhz9pHE|yUap1p0uRJ_RtX`v2ssuE=YX+eDvVT%8J-gm*+xooxZpw9r^i4r>l^Wm-yD`{71QXJJ0?*-+7? z(+eqW=!!e`qA|tco-XF0dG1N~0d?v7Mt+4suev_><6~9~JHcYrM7tl)-2{Bdo%uz` z-gLj{JdAiSOuJkomffsEmZp8*v<`?LP~baD_xOG9J;lX1pR<~d>47RU)yJY* zw}0J10kaGkb3@={X1?wyk9pRk6+(RbqSk1!6IY^lpYgi5S@d-TkH$ukp++76c{nbr2=?kTU$R)V>1X3 z0=;BD5B