From 47fe396a585dee650f2e8645197d8889a29443a2 Mon Sep 17 00:00:00 2001 From: mhjensen Date: Tue, 1 Sep 2020 07:23:34 +0200 Subject: [PATCH] typo in latex variant of project 1 --- .../2020/Project1/html/._Project1-bs000.html | 617 ++++++++++++++++++ .../2020/Project1/html/Project1-bs.html | 617 ++++++++++++++++++ doc/Projects/2020/Project1/html/Project1.html | 556 ++++++++++++++++ .../Project1/ipynb/ipynb-Project1-src.tar.gz | Bin 0 -> 192 bytes doc/Projects/2020/Project1/pdf/Project1.p.tex | 573 ++++++++++++++++ doc/Projects/2020/Project1/pdf/Project1.pdf | Bin 0 -> 281744 bytes doc/Projects/2020/Project1/pdf/Project1.tex | 541 +++++++++++++++ doc/Projects/2020/Project1/pdf/Project1.tex~ | 541 +++++++++++++++ 8 files changed, 3445 insertions(+) create mode 100644 doc/Projects/2020/Project1/html/._Project1-bs000.html create mode 100644 doc/Projects/2020/Project1/html/Project1-bs.html create mode 100644 doc/Projects/2020/Project1/html/Project1.html create mode 100644 doc/Projects/2020/Project1/ipynb/ipynb-Project1-src.tar.gz create mode 100644 doc/Projects/2020/Project1/pdf/Project1.p.tex create mode 100644 doc/Projects/2020/Project1/pdf/Project1.pdf create mode 100644 doc/Projects/2020/Project1/pdf/Project1.tex create mode 100644 doc/Projects/2020/Project1/pdf/Project1.tex~ diff --git a/doc/Projects/2020/Project1/html/._Project1-bs000.html b/doc/Projects/2020/Project1/html/._Project1-bs000.html new file mode 100644 index 000000000..4f89ee7c5 --- /dev/null +++ b/doc/Projects/2020/Project1/html/._Project1-bs000.html @@ -0,0 +1,617 @@ + + + + + + + + +Project 1 on Machine Learning, deadline October 5, 2020 + + + + + + + + + + + + + + + + + + + + + + + +
+
+ + Project 1 on Machine Learning, deadline October 5, 2020 +
+ +
+ +
+
+ + +
+ +

 

 

 

+ + + + + + +
+

Project 1 on Machine Learning, deadline October 5, 2020

+ +

+ + +

+Data Analysis and Machine Learning FYS-STK3155/FYS4155 +
+ +

+ + +

Department of Physics, University of Oslo, Norway
+
+

+

Sep 1, 2020

+
+

+

+ +

Regression analysis and resampling methods

+ +

+The main aim of this project is to study in more detail various +regression methods, including the Ordinary Least Squares (OLS) method, +Ridge regression and finally Lasso regression. +The methods are in turn combined with resampling techniques like the bootstrap method and cross validation. + +

+We will first study how to fit polynomials to a specific +two-dimensional function called Franke's +function. This +is a function which has been widely used when testing various +interpolation and fitting algorithms. Furthermore, after having +established the model and the method, we will employ resamling +techniques such as cross-validation and/or bootstrap in order to perform a +proper assessment of our models. We will also study in detail the +so-called Bias-Variance trade off. + +

+The Franke function, which is a weighted sum of four exponentials reads as follows +$$ +\begin{align*} +f(x,y) &= \frac{3}{4}\exp{\left(-\frac{(9x-2)^2}{4} - \frac{(9y-2)^2}{4}\right)}+\frac{3}{4}\exp{\left(-\frac{(9x+1)^2}{49}- \frac{(9y+1)}{10}\right)} \\ +&+\frac{1}{2}\exp{\left(-\frac{(9x-7)^2}{4} - \frac{(9y-3)^2}{4}\right)} -\frac{1}{5}\exp{\left(-(9x-4)^2 - (9y-7)^2\right) }. +\end{align*} +$$ + +

+The function will be defined for \( x,y\in [0,1] \). Our first step will +be to perform an OLS regression analysis of this function, trying out +a polynomial fit with an \( x \) and \( y \) dependence of the form \( [x, y, +x^2, y^2, xy, \dots] \). We will also include bootstrap first as +a resampling technique. After that we will include the cross-validation technique. As in homeworks 1 and 2, we can use a uniform +distribution to set up the arrays of values for \( x \) and \( y \), or as in +the example below just a set of fixed +values for \( x \) and \( y \) with a given step +size. We will fit a +function (for example a polynomial) of \( x \) and \( y \). Thereafter we +will repeat much of the same procedure using the Ridge and Lasso +regression methods, introducing thus a dependence on the bias +(penalty) \( \lambda \). + +

+Finally we are going to use (real) digital terrain data and try to +reproduce these data using the same methods. We will also try to go +beyond the second-order polynomials metioned above and explore +which polynomial fits the data best. + +

+The Python fucntion for the Franke function is included here (it performs also a three-dimensional plot of it) +

+ + +

from mpl_toolkits.mplot3d import Axes3D
+import matplotlib.pyplot as plt
+from matplotlib import cm
+from matplotlib.ticker import LinearLocator, FormatStrFormatter
+import numpy as np
+from random import random, seed
+
+fig = plt.figure()
+ax = fig.gca(projection='3d')
+
+# Make data.
+x = np.arange(0, 1, 0.05)
+y = np.arange(0, 1, 0.05)
+x, y = np.meshgrid(x,y)
+
+
+def FrankeFunction(x,y):
+    term1 = 0.75*np.exp(-(0.25*(9*x-2)**2) - 0.25*((9*y-2)**2))
+    term2 = 0.75*np.exp(-((9*x+1)**2)/49.0 - 0.1*(9*y+1))
+    term3 = 0.5*np.exp(-(9*x-7)**2/4.0 - 0.25*((9*y-3)**2))
+    term4 = -0.2*np.exp(-(9*x-4)**2 - (9*y-7)**2)
+    return term1 + term2 + term3 + term4
+
+
+z = FrankeFunction(x, y)
+
+# Plot the surface.
+surf = ax.plot_surface(x, y, z, cmap=cm.coolwarm,
+                       linewidth=0, antialiased=False)
+
+# Customize the z axis.
+ax.set_zlim(-0.10, 1.40)
+ax.zaxis.set_major_locator(LinearLocator(10))
+ax.zaxis.set_major_formatter(FormatStrFormatter('%.02f'))
+
+# Add a color bar which maps values to colors.
+fig.colorbar(surf, shrink=0.5, aspect=5)
+
+plt.show()
+
+ +

Part a): Ordinary Least Square on the Franke function with resampling

+ +

+We will generate our own dataset for a function +\( \mathrm{FrankeFunction}(x,y) \) with \( x,y \in [0,1] \). The function +\( f(x,y) \) is the Franke function. You should explore also the addition +of an added stochastic noise to this function using the normal +distribution \( \cal{N}(0,1) \). + +

+Write your own code (using either a matrix inversion or a singular +value decomposition from e.g., numpy ) or use your code from +homeworks 1 and 2 and perform a standard least square regression +analysis using polynomials in \( x \) and \( y \) up to fifth order. Find the +confidence intervals of the parameters \( \beta \) by computing their +variances, evaluate the Mean Squared error (MSE) + +$$ MSE(\hat{y},\hat{\tilde{y}}) = \frac{1}{n} +\sum_{i=0}^{n-1}(y_i-\tilde{y}_i)^2, +$$ + +

+and the \( R^2 \) score function. If \( \tilde{\hat{y}}_i \) is the predicted +value of the \( i-th \) sample and \( y_i \) is the corresponding true value, +then the score \( R^2 \) is defined as + +$$ +R^2(\hat{y}, \tilde{\hat{y}}) = 1 - \frac{\sum_{i=0}^{n - 1} (y_i - \tilde{y}_i)^2}{\sum_{i=0}^{n - 1} (y_i - \bar{y})^2}, +$$ + +

+where we have defined the mean value of \( \hat{y} \) as + +$$ +\bar{y} = \frac{1}{n} \sum_{i=0}^{n - 1} y_i. +$$ + +

+Your code has to include a scaling of the data (for example by subtracting the mean value, see also homework set 2 for examples) and a split of the data in training and test data. For this part you can either write your own code or use for example the +function for splitting training data provided by the library Scikit-Learn (make sure you have installed it). +This function is called \( train\_test\_split \). Similarly, and see the solution to homework set 2, exercise 2, you can use the data normalization functionality of Scikit-Learn. + +

+It is normal in essentially all Machine Learning studies to split the +data in a training set and a test set (eventually also an additional +validation set). There +is no explicit recipe for how much data should be included as training +data and say test data. An accepted rule of thumb is to use +approximately \( 2/3 \) to \( 4/5 \) of the data as training data. + +

Part b): Bias-variance trade-off

+ +

+Our aim here is to study the bias-variance trade-off by implementing the bootstrap resampling technique. + +

+With a code which does OLS and includes resampling techniques, +we will now discuss the bias-variance trade-off in the context of +continuous predictions such as regression. However, many of the +intuitions and ideas discussed here also carry over to classification +tasks and basically all Machine Learning algorithms. + +

+Consider a +dataset \( \mathcal{L} \) consisting of the data +\( \mathbf{X}_\mathcal{L}=\{(y_j, \boldsymbol{x}_j), j=0\ldots n-1\} \). + +

+Let us assume that the true data is generated from a noisy model + +$$ +\boldsymbol{y}=f(\boldsymbol{x}) + \boldsymbol{\epsilon}. +$$ + +

+Here \( \epsilon \) is normally distributed with mean zero and standard +deviation \( \sigma^2 \). + +

+In our derivation of the ordinary least squares method we defined then +an approximation to the function \( f \) in terms of the parameters +\( \boldsymbol{\beta} \) and the design matrix \( \boldsymbol{X} \) which embody our model, +that is \( \boldsymbol{\tilde{y}}=\boldsymbol{X}\boldsymbol{\beta} \). + +

+The parameters \( \boldsymbol{\beta} \) are in turn found by optimizing the means +squared error via the so-called cost function + +$$ +C(\boldsymbol{X},\boldsymbol{\beta}) =\frac{1}{n}\sum_{i=0}^{n-1}(y_i-\tilde{y}_i)^2=\mathbb{E}\left[(\boldsymbol{y}-\boldsymbol{\tilde{y}})^2\right]. +$$ + +

+Show that you can rewrite this as +$$ +\mathbb{E}\left[(\boldsymbol{y}-\boldsymbol{\tilde{y}})^2\right]=\frac{1}{n}\sum_i(f_i-\mathbb{E}\left[\boldsymbol{\tilde{y}}\right])^2+\frac{1}{n}\sum_i(\tilde{y}_i-\mathbb{E}\left[\boldsymbol{\tilde{y}}\right])^2+\sigma^2. +$$ + +

+Explain what the terms mean, which one is the bias and which one is +the variance and discuss their interpretations. + +

+Discuss the bias and variance trade-off as function +of your model complexity (the degree of the polynomial) and the number +of data points, and possibly also your training and test data using the bootstrap resampling method. + +

+However, before you perform an analysis of the bias-variance trade-off on your test data, make +first a figure similar to Fig. 2.11 of Hastie, Tibshirani, and +Friedman. Figure 2.11 of this reference displays only the test and training MSEs (and bias-variance analysis of the test data) while indicating possible regions of low/high bias and variance. You will most likely not get an +equally smooth curve! Note also that when you calculate the bias, in all applications you don't know the function values \( f_i \). You would hence replace them with the actual data points \( y_i \). + +

+After having produced a curve similar to Figure 2.11, perform then a bias-variance analysis of your test data. Here you should use the bootstra resampling technique. + +

Part c) Cross-validation as resampling techniques, adding more complexity

+ +

+The aim here is to write our own code for another widely popular resampling technique, the so-called cross-validation method. +Again, before you start with cross-validation approach, you should scale your data and split it in test and training data as you did earlier. +Perform a resampling of the data where you split the data in training +data and test data using for example + +

+Implement the \( k \)-fold cross-validation algorithm (write your own +code) and evaluate again the MSE function resulting +from the test data. You can compare your own code with that from +Scikit-Learn if needed. You can alternatively write your own bootstrap code. + +

+Compare the MSE you get from your cross-validation code with the one you got from your bootstrap code. +You can also compare your own cross-validation code with the one provided by Scikit-Learn. + +

Part d): Ridge Regression on the Franke function with resampling

+ +

+Write your own code for the Ridge method, either using matrix +inversion or the singular value decomposition as done in the previous +exercise or howework 2 (see also chapter 3.4 of Hastie et al., +equations (3.43) and (3.44)). Perform the same bootstrap analysis as in the +part b) (for the same polynomials) and the cross-validation part in part c) but now for different values of \( \lambda \). Compare and +analyze your results with those obtained in parts a-c). Study the +dependence on \( \lambda \). + +

+Study also the bias-variance trade-off as function of various values of +the parameter \( \lambda \). For the bias-variance trade-off, use the bootstrap resampling method. Comment your results. + +

Part e): Lasso Regression on the Franke function with resampling

+ +

+This part is essentially a repeat of the previous two ones, but now +with Lasso regression. Write either your own code (difficult and optional) or, in this case, +you can also use the functionalities of Scikit-Learn (recommended). +Give a +critical discussion of the three methods and a judgement of which +model fits the data best. Perform here as well an analysis of the bias-variance trade-off using the bootstrap resampling technique and an analysis of the mean squared error using cross-validation. + +

Part f): Introducing real data and preparing the data analysis

+ +

+With our codes functioning and having been tested properly on a +simpler function we are now ready to look at real data. We will +essentially repeat in part g) what was done in parts a-e). However, we +need first to download the data and prepare properly the inputs to our +codes. We are going to download digital terrain data from the website +https://earthexplorer.usgs.gov/, + +

+In order to obtain data for a specific region, you need to register as +a user (free) at this website and then decide upon which area you want +to fetch the digital terrain data from. In order to be able to read +the data properly, you need to specify that the format should be SRTM +Arc-Second Global and download the data as a GeoTIF file. The +files are then stored in tif format which can be imported into a +Python program using + +

+ + +

scipy.misc.imread
+
+

+Here is a simple part of a Python code which reads and plots the data +from such files + +

+ + +

import numpy as np
+from imageio import imread
+import matplotlib.pyplot as plt
+from mpl_toolkits.mplot3d import Axes3D
+from matplotlib import cm
+
+# Load the terrain
+terrain1 = imread('SRTM_data_Norway_1.tif')
+# Show the terrain
+plt.figure()
+plt.title('Terrain over Norway 1')
+plt.imshow(terrain1, cmap='gray')
+plt.xlabel('X')
+plt.ylabel('Y')
+plt.show()
+
+

+If you should have problems in downloading the digital terrain data, +we provide two examples under the data folder of project 1. One is +from a region close to Stavanger in Norway and the other Møsvatn +Austfjell, again in Norway. +Feel free to produce your own terrain data. + +

+Alternatively, if you would like to use another data set, feel free to do so. This could be data close to your reseach area or simply a data set you found interesting. See for example kaggle.com for examples. + +

Part g) OLS, Ridge and Lasso regression with resampling

+ +

+Our final part deals with the parameterization of your digital terrain +data (or your own data). We will apply all three methods for linear regression, the same type (or higher order) of polynomial +approximation and cross-validation as resampling technique to evaluate which +model fits the data best. + +

+At the end, you should present a critical evaluation of your results +and discuss the applicability of these regression methods to the type +of data presented here (either the terrain data we propose or other data sets). + +

Background literature

+ +
    +
  1. For a discussion and derivation of the variances and mean squared errors using linear regression, see the Lecture notes on ridge regression by Wessel N. van Wieringen
    2. +
  2. The textbook of Trevor Hastie, Robert Tibshirani, Jerome H. Friedman, The Elements of Statistical Learning, Springer, chapters 3 and 7 are the most relevant ones for the analysis here.
    3. +
+ +

Introduction to numerical projects

+ +

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

+ +

Format for electronic delivery of report and programs

+ +

+The preferred format for the report is a PDF file. You can also use DOC or postscript formats or as an ipython notebook file. As programming language we prefer that you choose between C/C++, Fortran2008 or Python. The following prescription should be followed when preparing the report: + +

+ +Finally, +we encourage you to collaborate. Optimal working groups consist of +2-3 students. You can then hand in a common report. + +

Software and needed installations

+ +

+If you have Python installed (we recommend Python3) and you feel pretty familiar with installing different packages, +we recommend that you install the following Python packages via pip as + +

    +
  1. pip install numpy scipy matplotlib ipython scikit-learn tensorflow sympy pandas pillow
    2. +
+ +For Python3, replace pip with pip3. + +

+See below for a discussion of tensorflow and scikit-learn. + +

+For OSX users we recommend also, after having installed Xcode, to install brew. Brew allows +for a seamless installation of additional software via for example + +

    +
  1. brew install python3
    2. +
+ +For Linux users, with its variety of distributions like for example the widely popular Ubuntu distribution +you can use pip as well and simply install Python as + +
    +
  1. sudo apt-get install python3 (or python for python2.7)
    2. +
+ +etc etc. + +

+If you don't want to install various Python packages with their dependencies separately, we recommend two widely used distrubutions which set up all relevant dependencies for Python, namely + +

    +
  1. Anaconda Anaconda is an open source distribution of the Python and R programming languages for large-scale data processing, predictive analytics, and scientific computing, that aims to simplify package management and deployment. Package versions are managed by the package management system conda
    2. +
  2. Enthought canopy is a Python distribution for scientific and analytic computing distribution and analysis environment, available for free and under a commercial license.
    3. +
+ +Popular software packages written in Python for ML are + + + +These are all freely available at their respective GitHub sites. They +encompass communities of developers in the thousands or more. And the number +of code developers and contributors keeps increasing. + +

+ +

+ +

+ + +
+ + + + + + + +
+ © 1999-2020, "Data Analysis and Machine Learning FYS-STK3155/FYS4155":"http://www.uio.no/studier/emner/matnat/fys/FYS3155/index-eng.html". Released under CC Attribution-NonCommercial 4.0 license +
+ + + + + + diff --git a/doc/Projects/2020/Project1/html/Project1-bs.html b/doc/Projects/2020/Project1/html/Project1-bs.html new file mode 100644 index 000000000..4f89ee7c5 --- /dev/null +++ b/doc/Projects/2020/Project1/html/Project1-bs.html @@ -0,0 +1,617 @@ + + + + + + + + +Project 1 on Machine Learning, deadline October 5, 2020 + + + + + + + + + + + + + + + + + + + + + + + +
+
+ + Project 1 on Machine Learning, deadline October 5, 2020 +
+ +
+ +
+
+ + +
+ +

 

 

 

+ + + + + + +
+

Project 1 on Machine Learning, deadline October 5, 2020

+ +

+ + +

+Data Analysis and Machine Learning FYS-STK3155/FYS4155 +
+ +

+ + +

Department of Physics, University of Oslo, Norway
+
+

+

Sep 1, 2020

+
+

+

+ +

Regression analysis and resampling methods

+ +

+The main aim of this project is to study in more detail various +regression methods, including the Ordinary Least Squares (OLS) method, +Ridge regression and finally Lasso regression. +The methods are in turn combined with resampling techniques like the bootstrap method and cross validation. + +

+We will first study how to fit polynomials to a specific +two-dimensional function called Franke's +function. This +is a function which has been widely used when testing various +interpolation and fitting algorithms. Furthermore, after having +established the model and the method, we will employ resamling +techniques such as cross-validation and/or bootstrap in order to perform a +proper assessment of our models. We will also study in detail the +so-called Bias-Variance trade off. + +

+The Franke function, which is a weighted sum of four exponentials reads as follows +$$ +\begin{align*} +f(x,y) &= \frac{3}{4}\exp{\left(-\frac{(9x-2)^2}{4} - \frac{(9y-2)^2}{4}\right)}+\frac{3}{4}\exp{\left(-\frac{(9x+1)^2}{49}- \frac{(9y+1)}{10}\right)} \\ +&+\frac{1}{2}\exp{\left(-\frac{(9x-7)^2}{4} - \frac{(9y-3)^2}{4}\right)} -\frac{1}{5}\exp{\left(-(9x-4)^2 - (9y-7)^2\right) }. +\end{align*} +$$ + +

+The function will be defined for \( x,y\in [0,1] \). Our first step will +be to perform an OLS regression analysis of this function, trying out +a polynomial fit with an \( x \) and \( y \) dependence of the form \( [x, y, +x^2, y^2, xy, \dots] \). We will also include bootstrap first as +a resampling technique. After that we will include the cross-validation technique. As in homeworks 1 and 2, we can use a uniform +distribution to set up the arrays of values for \( x \) and \( y \), or as in +the example below just a set of fixed +values for \( x \) and \( y \) with a given step +size. We will fit a +function (for example a polynomial) of \( x \) and \( y \). Thereafter we +will repeat much of the same procedure using the Ridge and Lasso +regression methods, introducing thus a dependence on the bias +(penalty) \( \lambda \). + +

+Finally we are going to use (real) digital terrain data and try to +reproduce these data using the same methods. We will also try to go +beyond the second-order polynomials metioned above and explore +which polynomial fits the data best. + +

+The Python fucntion for the Franke function is included here (it performs also a three-dimensional plot of it) +

+ + +

from mpl_toolkits.mplot3d import Axes3D
+import matplotlib.pyplot as plt
+from matplotlib import cm
+from matplotlib.ticker import LinearLocator, FormatStrFormatter
+import numpy as np
+from random import random, seed
+
+fig = plt.figure()
+ax = fig.gca(projection='3d')
+
+# Make data.
+x = np.arange(0, 1, 0.05)
+y = np.arange(0, 1, 0.05)
+x, y = np.meshgrid(x,y)
+
+
+def FrankeFunction(x,y):
+    term1 = 0.75*np.exp(-(0.25*(9*x-2)**2) - 0.25*((9*y-2)**2))
+    term2 = 0.75*np.exp(-((9*x+1)**2)/49.0 - 0.1*(9*y+1))
+    term3 = 0.5*np.exp(-(9*x-7)**2/4.0 - 0.25*((9*y-3)**2))
+    term4 = -0.2*np.exp(-(9*x-4)**2 - (9*y-7)**2)
+    return term1 + term2 + term3 + term4
+
+
+z = FrankeFunction(x, y)
+
+# Plot the surface.
+surf = ax.plot_surface(x, y, z, cmap=cm.coolwarm,
+                       linewidth=0, antialiased=False)
+
+# Customize the z axis.
+ax.set_zlim(-0.10, 1.40)
+ax.zaxis.set_major_locator(LinearLocator(10))
+ax.zaxis.set_major_formatter(FormatStrFormatter('%.02f'))
+
+# Add a color bar which maps values to colors.
+fig.colorbar(surf, shrink=0.5, aspect=5)
+
+plt.show()
+
+ +

Part a): Ordinary Least Square on the Franke function with resampling

+ +

+We will generate our own dataset for a function +\( \mathrm{FrankeFunction}(x,y) \) with \( x,y \in [0,1] \). The function +\( f(x,y) \) is the Franke function. You should explore also the addition +of an added stochastic noise to this function using the normal +distribution \( \cal{N}(0,1) \). + +

+Write your own code (using either a matrix inversion or a singular +value decomposition from e.g., numpy ) or use your code from +homeworks 1 and 2 and perform a standard least square regression +analysis using polynomials in \( x \) and \( y \) up to fifth order. Find the +confidence intervals of the parameters \( \beta \) by computing their +variances, evaluate the Mean Squared error (MSE) + +$$ MSE(\hat{y},\hat{\tilde{y}}) = \frac{1}{n} +\sum_{i=0}^{n-1}(y_i-\tilde{y}_i)^2, +$$ + +

+and the \( R^2 \) score function. If \( \tilde{\hat{y}}_i \) is the predicted +value of the \( i-th \) sample and \( y_i \) is the corresponding true value, +then the score \( R^2 \) is defined as + +$$ +R^2(\hat{y}, \tilde{\hat{y}}) = 1 - \frac{\sum_{i=0}^{n - 1} (y_i - \tilde{y}_i)^2}{\sum_{i=0}^{n - 1} (y_i - \bar{y})^2}, +$$ + +

+where we have defined the mean value of \( \hat{y} \) as + +$$ +\bar{y} = \frac{1}{n} \sum_{i=0}^{n - 1} y_i. +$$ + +

+Your code has to include a scaling of the data (for example by subtracting the mean value, see also homework set 2 for examples) and a split of the data in training and test data. For this part you can either write your own code or use for example the +function for splitting training data provided by the library Scikit-Learn (make sure you have installed it). +This function is called \( train\_test\_split \). Similarly, and see the solution to homework set 2, exercise 2, you can use the data normalization functionality of Scikit-Learn. + +

+It is normal in essentially all Machine Learning studies to split the +data in a training set and a test set (eventually also an additional +validation set). There +is no explicit recipe for how much data should be included as training +data and say test data. An accepted rule of thumb is to use +approximately \( 2/3 \) to \( 4/5 \) of the data as training data. + +

Part b): Bias-variance trade-off

+ +

+Our aim here is to study the bias-variance trade-off by implementing the bootstrap resampling technique. + +

+With a code which does OLS and includes resampling techniques, +we will now discuss the bias-variance trade-off in the context of +continuous predictions such as regression. However, many of the +intuitions and ideas discussed here also carry over to classification +tasks and basically all Machine Learning algorithms. + +

+Consider a +dataset \( \mathcal{L} \) consisting of the data +\( \mathbf{X}_\mathcal{L}=\{(y_j, \boldsymbol{x}_j), j=0\ldots n-1\} \). + +

+Let us assume that the true data is generated from a noisy model + +$$ +\boldsymbol{y}=f(\boldsymbol{x}) + \boldsymbol{\epsilon}. +$$ + +

+Here \( \epsilon \) is normally distributed with mean zero and standard +deviation \( \sigma^2 \). + +

+In our derivation of the ordinary least squares method we defined then +an approximation to the function \( f \) in terms of the parameters +\( \boldsymbol{\beta} \) and the design matrix \( \boldsymbol{X} \) which embody our model, +that is \( \boldsymbol{\tilde{y}}=\boldsymbol{X}\boldsymbol{\beta} \). + +

+The parameters \( \boldsymbol{\beta} \) are in turn found by optimizing the means +squared error via the so-called cost function + +$$ +C(\boldsymbol{X},\boldsymbol{\beta}) =\frac{1}{n}\sum_{i=0}^{n-1}(y_i-\tilde{y}_i)^2=\mathbb{E}\left[(\boldsymbol{y}-\boldsymbol{\tilde{y}})^2\right]. +$$ + +

+Show that you can rewrite this as +$$ +\mathbb{E}\left[(\boldsymbol{y}-\boldsymbol{\tilde{y}})^2\right]=\frac{1}{n}\sum_i(f_i-\mathbb{E}\left[\boldsymbol{\tilde{y}}\right])^2+\frac{1}{n}\sum_i(\tilde{y}_i-\mathbb{E}\left[\boldsymbol{\tilde{y}}\right])^2+\sigma^2. +$$ + +

+Explain what the terms mean, which one is the bias and which one is +the variance and discuss their interpretations. + +

+Discuss the bias and variance trade-off as function +of your model complexity (the degree of the polynomial) and the number +of data points, and possibly also your training and test data using the bootstrap resampling method. + +

+However, before you perform an analysis of the bias-variance trade-off on your test data, make +first a figure similar to Fig. 2.11 of Hastie, Tibshirani, and +Friedman. Figure 2.11 of this reference displays only the test and training MSEs (and bias-variance analysis of the test data) while indicating possible regions of low/high bias and variance. You will most likely not get an +equally smooth curve! Note also that when you calculate the bias, in all applications you don't know the function values \( f_i \). You would hence replace them with the actual data points \( y_i \). + +

+After having produced a curve similar to Figure 2.11, perform then a bias-variance analysis of your test data. Here you should use the bootstra resampling technique. + +

Part c) Cross-validation as resampling techniques, adding more complexity

+ +

+The aim here is to write our own code for another widely popular resampling technique, the so-called cross-validation method. +Again, before you start with cross-validation approach, you should scale your data and split it in test and training data as you did earlier. +Perform a resampling of the data where you split the data in training +data and test data using for example + +

+Implement the \( k \)-fold cross-validation algorithm (write your own +code) and evaluate again the MSE function resulting +from the test data. You can compare your own code with that from +Scikit-Learn if needed. You can alternatively write your own bootstrap code. + +

+Compare the MSE you get from your cross-validation code with the one you got from your bootstrap code. +You can also compare your own cross-validation code with the one provided by Scikit-Learn. + +

Part d): Ridge Regression on the Franke function with resampling

+ +

+Write your own code for the Ridge method, either using matrix +inversion or the singular value decomposition as done in the previous +exercise or howework 2 (see also chapter 3.4 of Hastie et al., +equations (3.43) and (3.44)). Perform the same bootstrap analysis as in the +part b) (for the same polynomials) and the cross-validation part in part c) but now for different values of \( \lambda \). Compare and +analyze your results with those obtained in parts a-c). Study the +dependence on \( \lambda \). + +

+Study also the bias-variance trade-off as function of various values of +the parameter \( \lambda \). For the bias-variance trade-off, use the bootstrap resampling method. Comment your results. + +

Part e): Lasso Regression on the Franke function with resampling

+ +

+This part is essentially a repeat of the previous two ones, but now +with Lasso regression. Write either your own code (difficult and optional) or, in this case, +you can also use the functionalities of Scikit-Learn (recommended). +Give a +critical discussion of the three methods and a judgement of which +model fits the data best. Perform here as well an analysis of the bias-variance trade-off using the bootstrap resampling technique and an analysis of the mean squared error using cross-validation. + +

Part f): Introducing real data and preparing the data analysis

+ +

+With our codes functioning and having been tested properly on a +simpler function we are now ready to look at real data. We will +essentially repeat in part g) what was done in parts a-e). However, we +need first to download the data and prepare properly the inputs to our +codes. We are going to download digital terrain data from the website +https://earthexplorer.usgs.gov/, + +

+In order to obtain data for a specific region, you need to register as +a user (free) at this website and then decide upon which area you want +to fetch the digital terrain data from. In order to be able to read +the data properly, you need to specify that the format should be SRTM +Arc-Second Global and download the data as a GeoTIF file. The +files are then stored in tif format which can be imported into a +Python program using + +

+ + +

scipy.misc.imread
+
+

+Here is a simple part of a Python code which reads and plots the data +from such files + +

+ + +

import numpy as np
+from imageio import imread
+import matplotlib.pyplot as plt
+from mpl_toolkits.mplot3d import Axes3D
+from matplotlib import cm
+
+# Load the terrain
+terrain1 = imread('SRTM_data_Norway_1.tif')
+# Show the terrain
+plt.figure()
+plt.title('Terrain over Norway 1')
+plt.imshow(terrain1, cmap='gray')
+plt.xlabel('X')
+plt.ylabel('Y')
+plt.show()
+
+

+If you should have problems in downloading the digital terrain data, +we provide two examples under the data folder of project 1. One is +from a region close to Stavanger in Norway and the other Møsvatn +Austfjell, again in Norway. +Feel free to produce your own terrain data. + +

+Alternatively, if you would like to use another data set, feel free to do so. This could be data close to your reseach area or simply a data set you found interesting. See for example kaggle.com for examples. + +

Part g) OLS, Ridge and Lasso regression with resampling

+ +

+Our final part deals with the parameterization of your digital terrain +data (or your own data). We will apply all three methods for linear regression, the same type (or higher order) of polynomial +approximation and cross-validation as resampling technique to evaluate which +model fits the data best. + +

+At the end, you should present a critical evaluation of your results +and discuss the applicability of these regression methods to the type +of data presented here (either the terrain data we propose or other data sets). + +

Background literature

+ +
    +
  1. For a discussion and derivation of the variances and mean squared errors using linear regression, see the Lecture notes on ridge regression by Wessel N. van Wieringen
    2. +
  2. The textbook of Trevor Hastie, Robert Tibshirani, Jerome H. Friedman, The Elements of Statistical Learning, Springer, chapters 3 and 7 are the most relevant ones for the analysis here.
    3. +
+ +

Introduction to numerical projects

+ +

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

+ +

Format for electronic delivery of report and programs

+ +

+The preferred format for the report is a PDF file. You can also use DOC or postscript formats or as an ipython notebook file. As programming language we prefer that you choose between C/C++, Fortran2008 or Python. The following prescription should be followed when preparing the report: + +

+ +Finally, +we encourage you to collaborate. Optimal working groups consist of +2-3 students. You can then hand in a common report. + +

Software and needed installations

+ +

+If you have Python installed (we recommend Python3) and you feel pretty familiar with installing different packages, +we recommend that you install the following Python packages via pip as + +

    +
  1. pip install numpy scipy matplotlib ipython scikit-learn tensorflow sympy pandas pillow
    2. +
+ +For Python3, replace pip with pip3. + +

+See below for a discussion of tensorflow and scikit-learn. + +

+For OSX users we recommend also, after having installed Xcode, to install brew. Brew allows +for a seamless installation of additional software via for example + +

    +
  1. brew install python3
    2. +
+ +For Linux users, with its variety of distributions like for example the widely popular Ubuntu distribution +you can use pip as well and simply install Python as + +
    +
  1. sudo apt-get install python3 (or python for python2.7)
    2. +
+ +etc etc. + +

+If you don't want to install various Python packages with their dependencies separately, we recommend two widely used distrubutions which set up all relevant dependencies for Python, namely + +

    +
  1. Anaconda Anaconda is an open source distribution of the Python and R programming languages for large-scale data processing, predictive analytics, and scientific computing, that aims to simplify package management and deployment. Package versions are managed by the package management system conda
    2. +
  2. Enthought canopy is a Python distribution for scientific and analytic computing distribution and analysis environment, available for free and under a commercial license.
    3. +
+ +Popular software packages written in Python for ML are + + + +These are all freely available at their respective GitHub sites. They +encompass communities of developers in the thousands or more. And the number +of code developers and contributors keeps increasing. + +

+ +

+ +

+ + +
+ + + + + + + +
+ © 1999-2020, "Data Analysis and Machine Learning FYS-STK3155/FYS4155":"http://www.uio.no/studier/emner/matnat/fys/FYS3155/index-eng.html". Released under CC Attribution-NonCommercial 4.0 license +
+ + + + + + diff --git a/doc/Projects/2020/Project1/html/Project1.html b/doc/Projects/2020/Project1/html/Project1.html new file mode 100644 index 000000000..90e66f282 --- /dev/null +++ b/doc/Projects/2020/Project1/html/Project1.html @@ -0,0 +1,556 @@ + + + + + + + + +Project 1 on Machine Learning, deadline October 5, 2020 + + + + + + + + + + + + + + + + + + + + + + + +

Project 1 on Machine Learning, deadline October 5, 2020

+ +

+ + +

+Data Analysis and Machine Learning FYS-STK3155/FYS4155 +
+ +

+ + +

Department of Physics, University of Oslo, Norway
+
+

+

Sep 1, 2020

+
+ +

Regression analysis and resampling methods

+ +

+The main aim of this project is to study in more detail various +regression methods, including the Ordinary Least Squares (OLS) method, +Ridge regression and finally Lasso regression. +The methods are in turn combined with resampling techniques like the bootstrap method and cross validation. + +

+We will first study how to fit polynomials to a specific +two-dimensional function called Franke's +function. This +is a function which has been widely used when testing various +interpolation and fitting algorithms. Furthermore, after having +established the model and the method, we will employ resamling +techniques such as cross-validation and/or bootstrap in order to perform a +proper assessment of our models. We will also study in detail the +so-called Bias-Variance trade off. + +

+The Franke function, which is a weighted sum of four exponentials reads as follows +$$ +\begin{align*} +f(x,y) &= \frac{3}{4}\exp{\left(-\frac{(9x-2)^2}{4} - \frac{(9y-2)^2}{4}\right)}+\frac{3}{4}\exp{\left(-\frac{(9x+1)^2}{49}- \frac{(9y+1)}{10}\right)} \\ +&+\frac{1}{2}\exp{\left(-\frac{(9x-7)^2}{4} - \frac{(9y-3)^2}{4}\right)} -\frac{1}{5}\exp{\left(-(9x-4)^2 - (9y-7)^2\right) }. +\end{align*} +$$ + +

+The function will be defined for \( x,y\in [0,1] \). Our first step will +be to perform an OLS regression analysis of this function, trying out +a polynomial fit with an \( x \) and \( y \) dependence of the form \( [x, y, +x^2, y^2, xy, \dots] \). We will also include bootstrap first as +a resampling technique. After that we will include the cross-validation technique. As in homeworks 1 and 2, we can use a uniform +distribution to set up the arrays of values for \( x \) and \( y \), or as in +the example below just a set of fixed +values for \( x \) and \( y \) with a given step +size. We will fit a +function (for example a polynomial) of \( x \) and \( y \). Thereafter we +will repeat much of the same procedure using the Ridge and Lasso +regression methods, introducing thus a dependence on the bias +(penalty) \( \lambda \). + +

+Finally we are going to use (real) digital terrain data and try to +reproduce these data using the same methods. We will also try to go +beyond the second-order polynomials metioned above and explore +which polynomial fits the data best. + +

+The Python fucntion for the Franke function is included here (it performs also a three-dimensional plot of it) +

+ + +

from mpl_toolkits.mplot3d import Axes3D
+import matplotlib.pyplot as plt
+from matplotlib import cm
+from matplotlib.ticker import LinearLocator, FormatStrFormatter
+import numpy as np
+from random import random, seed
+
+fig = plt.figure()
+ax = fig.gca(projection='3d')
+
+# Make data.
+x = np.arange(0, 1, 0.05)
+y = np.arange(0, 1, 0.05)
+x, y = np.meshgrid(x,y)
+
+
+def FrankeFunction(x,y):
+    term1 = 0.75*np.exp(-(0.25*(9*x-2)**2) - 0.25*((9*y-2)**2))
+    term2 = 0.75*np.exp(-((9*x+1)**2)/49.0 - 0.1*(9*y+1))
+    term3 = 0.5*np.exp(-(9*x-7)**2/4.0 - 0.25*((9*y-3)**2))
+    term4 = -0.2*np.exp(-(9*x-4)**2 - (9*y-7)**2)
+    return term1 + term2 + term3 + term4
+
+
+z = FrankeFunction(x, y)
+
+# Plot the surface.
+surf = ax.plot_surface(x, y, z, cmap=cm.coolwarm,
+                       linewidth=0, antialiased=False)
+
+# Customize the z axis.
+ax.set_zlim(-0.10, 1.40)
+ax.zaxis.set_major_locator(LinearLocator(10))
+ax.zaxis.set_major_formatter(FormatStrFormatter('%.02f'))
+
+# Add a color bar which maps values to colors.
+fig.colorbar(surf, shrink=0.5, aspect=5)
+
+plt.show()
+
+ +

Part a): Ordinary Least Square on the Franke function with resampling

+ +

+We will generate our own dataset for a function +\( \mathrm{FrankeFunction}(x,y) \) with \( x,y \in [0,1] \). The function +\( f(x,y) \) is the Franke function. You should explore also the addition +of an added stochastic noise to this function using the normal +distribution \( \cal{N}(0,1) \). + +

+Write your own code (using either a matrix inversion or a singular +value decomposition from e.g., numpy ) or use your code from +homeworks 1 and 2 and perform a standard least square regression +analysis using polynomials in \( x \) and \( y \) up to fifth order. Find the +confidence intervals of the parameters \( \beta \) by computing their +variances, evaluate the Mean Squared error (MSE) + +$$ MSE(\hat{y},\hat{\tilde{y}}) = \frac{1}{n} +\sum_{i=0}^{n-1}(y_i-\tilde{y}_i)^2, +$$ + +

+and the \( R^2 \) score function. If \( \tilde{\hat{y}}_i \) is the predicted +value of the \( i-th \) sample and \( y_i \) is the corresponding true value, +then the score \( R^2 \) is defined as + +$$ +R^2(\hat{y}, \tilde{\hat{y}}) = 1 - \frac{\sum_{i=0}^{n - 1} (y_i - \tilde{y}_i)^2}{\sum_{i=0}^{n - 1} (y_i - \bar{y})^2}, +$$ + +

+where we have defined the mean value of \( \hat{y} \) as + +$$ +\bar{y} = \frac{1}{n} \sum_{i=0}^{n - 1} y_i. +$$ + +

+Your code has to include a scaling of the data (for example by subtracting the mean value, see also homework set 2 for examples) and a split of the data in training and test data. For this part you can either write your own code or use for example the +function for splitting training data provided by the library Scikit-Learn (make sure you have installed it). +This function is called \( train\_test\_split \). Similarly, and see the solution to homework set 2, exercise 2, you can use the data normalization functionality of Scikit-Learn. + +

+It is normal in essentially all Machine Learning studies to split the +data in a training set and a test set (eventually also an additional +validation set). There +is no explicit recipe for how much data should be included as training +data and say test data. An accepted rule of thumb is to use +approximately \( 2/3 \) to \( 4/5 \) of the data as training data. + +

Part b): Bias-variance trade-off

+ +

+Our aim here is to study the bias-variance trade-off by implementing the bootstrap resampling technique. + +

+With a code which does OLS and includes resampling techniques, +we will now discuss the bias-variance trade-off in the context of +continuous predictions such as regression. However, many of the +intuitions and ideas discussed here also carry over to classification +tasks and basically all Machine Learning algorithms. + +

+Consider a +dataset \( \mathcal{L} \) consisting of the data +\( \mathbf{X}_\mathcal{L}=\{(y_j, \boldsymbol{x}_j), j=0\ldots n-1\} \). + +

+Let us assume that the true data is generated from a noisy model + +$$ +\boldsymbol{y}=f(\boldsymbol{x}) + \boldsymbol{\epsilon}. +$$ + +

+Here \( \epsilon \) is normally distributed with mean zero and standard +deviation \( \sigma^2 \). + +

+In our derivation of the ordinary least squares method we defined then +an approximation to the function \( f \) in terms of the parameters +\( \boldsymbol{\beta} \) and the design matrix \( \boldsymbol{X} \) which embody our model, +that is \( \boldsymbol{\tilde{y}}=\boldsymbol{X}\boldsymbol{\beta} \). + +

+The parameters \( \boldsymbol{\beta} \) are in turn found by optimizing the means +squared error via the so-called cost function + +$$ +C(\boldsymbol{X},\boldsymbol{\beta}) =\frac{1}{n}\sum_{i=0}^{n-1}(y_i-\tilde{y}_i)^2=\mathbb{E}\left[(\boldsymbol{y}-\boldsymbol{\tilde{y}})^2\right]. +$$ + +

+Show that you can rewrite this as +$$ +\mathbb{E}\left[(\boldsymbol{y}-\boldsymbol{\tilde{y}})^2\right]=\frac{1}{n}\sum_i(f_i-\mathbb{E}\left[\boldsymbol{\tilde{y}}\right])^2+\frac{1}{n}\sum_i(\tilde{y}_i-\mathbb{E}\left[\boldsymbol{\tilde{y}}\right])^2+\sigma^2. +$$ + +

+Explain what the terms mean, which one is the bias and which one is +the variance and discuss their interpretations. + +

+Discuss the bias and variance trade-off as function +of your model complexity (the degree of the polynomial) and the number +of data points, and possibly also your training and test data using the bootstrap resampling method. + +

+However, before you perform an analysis of the bias-variance trade-off on your test data, make +first a figure similar to Fig. 2.11 of Hastie, Tibshirani, and +Friedman. Figure 2.11 of this reference displays only the test and training MSEs (and bias-variance analysis of the test data) while indicating possible regions of low/high bias and variance. You will most likely not get an +equally smooth curve! Note also that when you calculate the bias, in all applications you don't know the function values \( f_i \). You would hence replace them with the actual data points \( y_i \). + +

+After having produced a curve similar to Figure 2.11, perform then a bias-variance analysis of your test data. Here you should use the bootstra resampling technique. + +

Part c) Cross-validation as resampling techniques, adding more complexity

+ +

+The aim here is to write our own code for another widely popular resampling technique, the so-called cross-validation method. +Again, before you start with cross-validation approach, you should scale your data and split it in test and training data as you did earlier. +Perform a resampling of the data where you split the data in training +data and test data using for example + +

+Implement the \( k \)-fold cross-validation algorithm (write your own +code) and evaluate again the MSE function resulting +from the test data. You can compare your own code with that from +Scikit-Learn if needed. You can alternatively write your own bootstrap code. + +

+Compare the MSE you get from your cross-validation code with the one you got from your bootstrap code. +You can also compare your own cross-validation code with the one provided by Scikit-Learn. + +

Part d): Ridge Regression on the Franke function with resampling

+ +

+Write your own code for the Ridge method, either using matrix +inversion or the singular value decomposition as done in the previous +exercise or howework 2 (see also chapter 3.4 of Hastie et al., +equations (3.43) and (3.44)). Perform the same bootstrap analysis as in the +part b) (for the same polynomials) and the cross-validation part in part c) but now for different values of \( \lambda \). Compare and +analyze your results with those obtained in parts a-c). Study the +dependence on \( \lambda \). + +

+Study also the bias-variance trade-off as function of various values of +the parameter \( \lambda \). For the bias-variance trade-off, use the bootstrap resampling method. Comment your results. + +

Part e): Lasso Regression on the Franke function with resampling

+ +

+This part is essentially a repeat of the previous two ones, but now +with Lasso regression. Write either your own code (difficult and optional) or, in this case, +you can also use the functionalities of Scikit-Learn (recommended). +Give a +critical discussion of the three methods and a judgement of which +model fits the data best. Perform here as well an analysis of the bias-variance trade-off using the bootstrap resampling technique and an analysis of the mean squared error using cross-validation. + +

Part f): Introducing real data and preparing the data analysis

+ +

+With our codes functioning and having been tested properly on a +simpler function we are now ready to look at real data. We will +essentially repeat in part g) what was done in parts a-e). However, we +need first to download the data and prepare properly the inputs to our +codes. We are going to download digital terrain data from the website +https://earthexplorer.usgs.gov/, + +

+In order to obtain data for a specific region, you need to register as +a user (free) at this website and then decide upon which area you want +to fetch the digital terrain data from. In order to be able to read +the data properly, you need to specify that the format should be SRTM +Arc-Second Global and download the data as a GeoTIF file. The +files are then stored in tif format which can be imported into a +Python program using + +

+ + +

scipy.misc.imread
+
+

+Here is a simple part of a Python code which reads and plots the data +from such files + +

+ + +

import numpy as np
+from imageio import imread
+import matplotlib.pyplot as plt
+from mpl_toolkits.mplot3d import Axes3D
+from matplotlib import cm
+
+# Load the terrain
+terrain1 = imread('SRTM_data_Norway_1.tif')
+# Show the terrain
+plt.figure()
+plt.title('Terrain over Norway 1')
+plt.imshow(terrain1, cmap='gray')
+plt.xlabel('X')
+plt.ylabel('Y')
+plt.show()
+
+

+If you should have problems in downloading the digital terrain data, +we provide two examples under the data folder of project 1. One is +from a region close to Stavanger in Norway and the other Møsvatn +Austfjell, again in Norway. +Feel free to produce your own terrain data. + +

+Alternatively, if you would like to use another data set, feel free to do so. This could be data close to your reseach area or simply a data set you found interesting. See for example kaggle.com for examples. + +

Part g) OLS, Ridge and Lasso regression with resampling

+ +

+Our final part deals with the parameterization of your digital terrain +data (or your own data). We will apply all three methods for linear regression, the same type (or higher order) of polynomial +approximation and cross-validation as resampling technique to evaluate which +model fits the data best. + +

+At the end, you should present a critical evaluation of your results +and discuss the applicability of these regression methods to the type +of data presented here (either the terrain data we propose or other data sets). + +

Background literature

+ +
    +
  1. For a discussion and derivation of the variances and mean squared errors using linear regression, see the Lecture notes on ridge regression by Wessel N. van Wieringen
    2. +
  2. The textbook of Trevor Hastie, Robert Tibshirani, Jerome H. Friedman, The Elements of Statistical Learning, Springer, chapters 3 and 7 are the most relevant ones for the analysis here.
    3. +
+ +

Introduction to numerical projects

+ +

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

+ +

Format for electronic delivery of report and programs

+ +

+The preferred format for the report is a PDF file. You can also use DOC or postscript formats or as an ipython notebook file. As programming language we prefer that you choose between C/C++, Fortran2008 or Python. The following prescription should be followed when preparing the report: + +

+ +Finally, +we encourage you to collaborate. Optimal working groups consist of +2-3 students. You can then hand in a common report. + +

Software and needed installations

+ +

+If you have Python installed (we recommend Python3) and you feel pretty familiar with installing different packages, +we recommend that you install the following Python packages via pip as + +

    +
  1. pip install numpy scipy matplotlib ipython scikit-learn tensorflow sympy pandas pillow
    2. +
+ +For Python3, replace pip with pip3. + +

+See below for a discussion of tensorflow and scikit-learn. + +

+For OSX users we recommend also, after having installed Xcode, to install brew. Brew allows +for a seamless installation of additional software via for example + +

    +
  1. brew install python3
    2. +
+ +For Linux users, with its variety of distributions like for example the widely popular Ubuntu distribution +you can use pip as well and simply install Python as + +
    +
  1. sudo apt-get install python3 (or python for python2.7)
    2. +
+ +etc etc. + +

+If you don't want to install various Python packages with their dependencies separately, we recommend two widely used distrubutions which set up all relevant dependencies for Python, namely + +

    +
  1. Anaconda Anaconda is an open source distribution of the Python and R programming languages for large-scale data processing, predictive analytics, and scientific computing, that aims to simplify package management and deployment. Package versions are managed by the package management system conda
    2. +
  2. Enthought canopy is a Python distribution for scientific and analytic computing distribution and analysis environment, available for free and under a commercial license.
    3. +
+ +Popular software packages written in Python for ML are + + + +These are all freely available at their respective GitHub sites. They +encompass communities of developers in the thousands or more. And the number +of code developers and contributors keeps increasing. + +

+ + + + +

+ © 1999-2020, "Data Analysis and Machine Learning FYS-STK3155/FYS4155":"http://www.uio.no/studier/emner/matnat/fys/FYS3155/index-eng.html". Released under CC Attribution-NonCommercial 4.0 license +
+ + + + + + diff --git a/doc/Projects/2020/Project1/ipynb/ipynb-Project1-src.tar.gz b/doc/Projects/2020/Project1/ipynb/ipynb-Project1-src.tar.gz new file mode 100644 index 0000000000000000000000000000000000000000..2f9e4fabb171d1aa41717738832e6db601256c27 GIT binary patch literal 192 zcmV;x06+g9iwFQT+D%^o1MSaC3c@fD2H>uHia9|^nm*QoUAPd6c!AWWHnnDIl7hXx zeSoeMH${Yeo1ZX0Vdju+w)-rwck8W&5Rya@OqnL}oFrV{GfGpSEDafp;$&$cyFnFD uqA=4Cw7B(B3*c%1UY62IRJ`>&jc3iTCh&iKj^j9v^Q8y)_NuD@2mk;sSyzVu literal 0 HcmV?d00001 diff --git a/doc/Projects/2020/Project1/pdf/Project1.p.tex b/doc/Projects/2020/Project1/pdf/Project1.p.tex new file mode 100644 index 000000000..67151ae52 --- /dev/null +++ b/doc/Projects/2020/Project1/pdf/Project1.p.tex @@ -0,0 +1,573 @@ +%% +%% Automatically generated file from DocOnce source +%% (https://github.com/hplgit/doconce/) +%% +%% +% #ifdef PTEX2TEX_EXPLANATION +%% +%% The file follows the ptex2tex extended LaTeX format, see +%% ptex2tex: http://code.google.com/p/ptex2tex/ +%% +%% Run +%% ptex2tex myfile +%% or +%% doconce ptex2tex myfile +%% +%% to turn myfile.p.tex into an ordinary LaTeX file myfile.tex. +%% (The ptex2tex program: http://code.google.com/p/ptex2tex) +%% Many preprocess options can be added to ptex2tex or doconce ptex2tex +%% +%% ptex2tex -DMINTED myfile +%% doconce ptex2tex myfile envir=minted +%% +%% ptex2tex will typeset code environments according to a global or local +%% .ptex2tex.cfg configure file. doconce ptex2tex will typeset code +%% according to options on the command line (just type doconce ptex2tex to +%% see examples). If doconce ptex2tex has envir=minted, it enables the +%% minted style without needing -DMINTED. +% #endif + +% #define PREAMBLE + +% #ifdef PREAMBLE +%-------------------- begin preamble ---------------------- + +\documentclass[% +oneside, % oneside: electronic viewing, twoside: printing +final, % draft: marks overfull hboxes, figures with paths +10pt]{article} + +\listfiles % print all files needed to compile this document + +\usepackage{relsize,makeidx,color,setspace,amsmath,amsfonts,amssymb} +\usepackage[table]{xcolor} +\usepackage{bm,ltablex,microtype} + +\usepackage[pdftex]{graphicx} + +\usepackage{ptex2tex} +% #ifdef MINTED +\usepackage{minted} +\usemintedstyle{default} +% #endif + +\usepackage[T1]{fontenc} +%\usepackage[latin1]{inputenc} +\usepackage{ucs} +\usepackage[utf8x]{inputenc} + +\usepackage{lmodern} % Latin Modern fonts derived from Computer Modern + +% Hyperlinks in PDF: +\definecolor{linkcolor}{rgb}{0,0,0.4} +\usepackage{hyperref} +\hypersetup{ + breaklinks=true, + colorlinks=true, + linkcolor=linkcolor, + urlcolor=linkcolor, + citecolor=black, + filecolor=black, + %filecolor=blue, + pdfmenubar=true, + pdftoolbar=true, + bookmarksdepth=3 % Uncomment (and tweak) for PDF bookmarks with more levels than the TOC + } +%\hyperbaseurl{} % hyperlinks are relative to this root + +\setcounter{tocdepth}{2} % levels in table of contents + +% --- fancyhdr package for fancy headers --- +\usepackage{fancyhdr} +\fancyhf{} % sets both header and footer to nothing +\renewcommand{\headrulewidth}{0pt} +\fancyfoot[LE,RO]{\thepage} +% Ensure copyright on titlepage (article style) and chapter pages (book style) +\fancypagestyle{plain}{ + \fancyhf{} + \fancyfoot[C]{{\footnotesize \copyright\ 1999-2020, "Data Analysis and Machine Learning FYS-STK3155/FYS4155":"http://www.uio.no/studier/emner/matnat/fys/FYS3155/index-eng.html". Released under CC Attribution-NonCommercial 4.0 license}} +% \renewcommand{\footrulewidth}{0mm} + \renewcommand{\headrulewidth}{0mm} +} +% Ensure copyright on titlepages with \thispagestyle{empty} +\fancypagestyle{empty}{ + \fancyhf{} + \fancyfoot[C]{{\footnotesize \copyright\ 1999-2020, "Data Analysis and Machine Learning FYS-STK3155/FYS4155":"http://www.uio.no/studier/emner/matnat/fys/FYS3155/index-eng.html". Released under CC Attribution-NonCommercial 4.0 license}} + \renewcommand{\footrulewidth}{0mm} + \renewcommand{\headrulewidth}{0mm} +} + +\pagestyle{fancy} + + +% prevent orhpans and widows +\clubpenalty = 10000 +\widowpenalty = 10000 + +% --- end of standard preamble for documents --- + + +% insert custom LaTeX commands... + +\raggedbottom +\makeindex +\usepackage[totoc]{idxlayout} % for index in the toc +\usepackage[nottoc]{tocbibind} % for references/bibliography in the toc + +%-------------------- end preamble ---------------------- + +\begin{document} + +% matching end for #ifdef PREAMBLE +% #endif + +\newcommand{\exercisesection}[1]{\subsection*{#1}} + + +% ------------------- main content ---------------------- + + + +% ----------------- title ------------------------- + +\thispagestyle{empty} + +\begin{center} +{\LARGE\bf +\begin{spacing}{1.25} +Project 1 on Machine Learning, deadline October 5, 2020 +\end{spacing} +} +\end{center} + +% ----------------- author(s) ------------------------- + +\begin{center} +{\bf \href{{http://www.uio.no/studier/emner/matnat/fys/FYS3155/index-eng.html}}{Data Analysis and Machine Learning FYS-STK3155/FYS4155}} +\end{center} + + \begin{center} +% List of all institutions: +\centerline{{\small Department of Physics, University of Oslo, Norway}} +\end{center} + +% ----------------- end author(s) ------------------------- + +% --- begin date --- +\begin{center} +Sep 1, 2020 +\end{center} +% --- end date --- + +\vspace{1cm} + + +\subsection{Regression analysis and resampling methods} + +The main aim of this project is to study in more detail various +regression methods, including the Ordinary Least Squares (OLS) method, +Ridge regression and finally Lasso regression. +The methods are in turn combined with resampling techniques like the bootstrap method and cross validation. + +We will first study how to fit polynomials to a specific +two-dimensional function called \href{{http://www.dtic.mil/dtic/tr/fulltext/u2/a081688.pdf}}{Franke's +function}. This +is a function which has been widely used when testing various +interpolation and fitting algorithms. Furthermore, after having +established the model and the method, we will employ resamling +techniques such as cross-validation and/or bootstrap in order to perform a +proper assessment of our models. We will also study in detail the +so-called Bias-Variance trade off. + + +The Franke function, which is a weighted sum of four exponentials reads as follows +\begin{align*} +f(x,y) &= \frac{3}{4}\exp{\left(-\frac{(9x-2)^2}{4} - \frac{(9y-2)^2}{4}\right)}+\frac{3}{4}\exp{\left(-\frac{(9x+1)^2}{49}- \frac{(9y+1)}{10}\right)} \\ +&+\frac{1}{2}\exp{\left(-\frac{(9x-7)^2}{4} - \frac{(9y-3)^2}{4}\right)} -\frac{1}{5}\exp{\left(-(9x-4)^2 - (9y-7)^2\right) }. +\end{align*} + +The function will be defined for $x,y\in [0,1]$. Our first step will +be to perform an OLS regression analysis of this function, trying out +a polynomial fit with an $x$ and $y$ dependence of the form $[x, y, +x^2, y^2, xy, \dots]$. We will also include bootstrap first as +a resampling technique. After that we will include the cross-validation technique. As in homeworks 1 and 2, we can use a uniform +distribution to set up the arrays of values for $x$ and $y$, or as in +the example below just a set of fixed +values for $x$ and $y$ with a given step +size. We will fit a +function (for example a polynomial) of $x$ and $y$. Thereafter we +will repeat much of the same procedure using the Ridge and Lasso +regression methods, introducing thus a dependence on the bias +(penalty) $\lambda$. + +Finally we are going to use (real) digital terrain data and try to +reproduce these data using the same methods. We will also try to go +beyond the second-order polynomials metioned above and explore +which polynomial fits the data best. + + +The Python fucntion for the Franke function is included here (it performs also a three-dimensional plot of it) +\bpycod +from mpl_toolkits.mplot3d import Axes3D +import matplotlib.pyplot as plt +from matplotlib import cm +from matplotlib.ticker import LinearLocator, FormatStrFormatter +import numpy as np +from random import random, seed + +fig = plt.figure() +ax = fig.gca(projection='3d') + +# Make data. +x = np.arange(0, 1, 0.05) +y = np.arange(0, 1, 0.05) +x, y = np.meshgrid(x,y) + + +def FrankeFunction(x,y): + term1 = 0.75*np.exp(-(0.25*(9*x-2)**2) - 0.25*((9*y-2)**2)) + term2 = 0.75*np.exp(-((9*x+1)**2)/49.0 - 0.1*(9*y+1)) + term3 = 0.5*np.exp(-(9*x-7)**2/4.0 - 0.25*((9*y-3)**2)) + term4 = -0.2*np.exp(-(9*x-4)**2 - (9*y-7)**2) + return term1 + term2 + term3 + term4 + + +z = FrankeFunction(x, y) + +# Plot the surface. +surf = ax.plot_surface(x, y, z, cmap=cm.coolwarm, + linewidth=0, antialiased=False) + +# Customize the z axis. +ax.set_zlim(-0.10, 1.40) +ax.zaxis.set_major_locator(LinearLocator(10)) +ax.zaxis.set_major_formatter(FormatStrFormatter('%.02f')) + +# Add a color bar which maps values to colors. +fig.colorbar(surf, shrink=0.5, aspect=5) + +plt.show() + +\epycod + + +\paragraph{Part a): Ordinary Least Square on the Franke function with resampling.} +We will generate our own dataset for a function +$\mathrm{FrankeFunction}(x,y)$ with $x,y \in [0,1]$. The function +$f(x,y)$ is the Franke function. You should explore also the addition +of an added stochastic noise to this function using the normal +distribution $\cal{N}(0,1)$. + +\emph{Write your own code} (using either a matrix inversion or a singular +value decomposition from e.g., \textbf{numpy} ) or use your code from +homeworks 1 and 2 and perform a standard least square regression +analysis using polynomials in $x$ and $y$ up to fifth order. Find the +confidence intervals of the parameters $\beta$ by computing their +variances, evaluate the Mean Squared error (MSE) + +\[ MSE(\hat{y},\hat{\tilde{y}}) = \frac{1}{n} +\sum_{i=0}^{n-1}(y_i-\tilde{y}_i)^2, +\] + +and the $R^2$ score function. If $\tilde{\hat{y}}_i$ is the predicted +value of the $i-th$ sample and $y_i$ is the corresponding true value, +then the score $R^2$ is defined as + +\[ +R^2(\hat{y}, \tilde{\hat{y}}) = 1 - \frac{\sum_{i=0}^{n - 1} (y_i - \tilde{y}_i)^2}{\sum_{i=0}^{n - 1} (y_i - \bar{y})^2}, +\] + +where we have defined the mean value of $\hat{y}$ as + +\[ +\bar{y} = \frac{1}{n} \sum_{i=0}^{n - 1} y_i. +\] + +Your code has to include a scaling of the data (for example by subtracting the mean value, see also homework set 2 for examples) and a split of the data in training and test data. For this part you can either write your own code or use for example the +function for splitting training data provided by the library \textbf{Scikit-Learn} (make sure you have installed it). +This function is called $train\_test\_split$. Similarly, and see the solution to homework set 2, exercise 2, you can use the data normalization functionality of \textbf{Scikit-Learn}. + +It is normal in essentially all Machine Learning studies to split the +data in a training set and a test set (eventually also an additional +validation set). There +is no explicit recipe for how much data should be included as training +data and say test data. An accepted rule of thumb is to use +approximately $2/3$ to $4/5$ of the data as training data. + + + + + + +\paragraph{Part b): Bias-variance trade-off.} +Our aim here is to study the bias-variance trade-off by implementing the \textbf{bootstrap} resampling technique. + +With a code which does OLS and includes resampling techniques, +we will now discuss the bias-variance trade-off in the context of +continuous predictions such as regression. However, many of the +intuitions and ideas discussed here also carry over to classification +tasks and basically all Machine Learning algorithms. + +Consider a +dataset $\mathcal{L}$ consisting of the data +$\mathbf{X}_\mathcal{L}=\{(y_j, \boldsymbol{x}_j), j=0\ldots n-1\}$. + +Let us assume that the true data is generated from a noisy model + +\[ +\bm{y}=f(\boldsymbol{x}) + \bm{\epsilon}. +\] + +Here $\epsilon$ is normally distributed with mean zero and standard +deviation $\sigma^2$. + +In our derivation of the ordinary least squares method we defined then +an approximation to the function $f$ in terms of the parameters +$\bm{\beta}$ and the design matrix $\bm{X}$ which embody our model, +that is $\bm{\tilde{y}}=\bm{X}\bm{\beta}$. + +The parameters $\bm{\beta}$ are in turn found by optimizing the means +squared error via the so-called cost function + +\[ +C(\bm{X},\bm{\beta}) =\frac{1}{n}\sum_{i=0}^{n-1}(y_i-\tilde{y}_i)^2=\mathbb{E}\left[(\bm{y}-\bm{\tilde{y}})^2\right]. +\] + +Show that you can rewrite this as +\[ +\mathbb{E}\left[(\bm{y}-\bm{\tilde{y}})^2\right]=\frac{1}{n}\sum_i(f_i-\mathbb{E}\left[\bm{\tilde{y}}\right])^2+\frac{1}{n}\sum_i(\tilde{y}_i-\mathbb{E}\left[\bm{\tilde{y}}\right])^2+\sigma^2. +\] + +Explain what the terms mean, which one is the bias and which one is +the variance and discuss their interpretations. + + +Discuss the bias and variance trade-off as function +of your model complexity (the degree of the polynomial) and the number +of data points, and possibly also your training and test data using the \textbf{bootstrap} resampling method. + +However, before you perform an analysis of the bias-variance trade-off on your test data, make +first a figure similar to Fig.~2.11 of Hastie, Tibshirani, and +Friedman. Figure 2.11 of this reference displays only the test and training MSEs (and bias-variance analysis of the test data) while indicating possible regions of low/high bias and variance. You will most likely not get an +equally smooth curve! Note also that when you calculate the bias, in all applications you don't know the function values $f_i$. You would hence replace them with the actual data points $y_i$. + +After having produced a curve similar to Figure 2.11, perform then a bias-variance analysis of your test data. Here you should use the \textbf{bootstra} resampling technique. + +\paragraph{Part c) Cross-validation as resampling techniques, adding more complexity.} +The aim here is to write our own code for another widely popular resampling technique, the so-called cross-validation method. +Again, before you start with cross-validation approach, you should scale your data and split it in test and training data as you did earlier. +Perform a resampling of the data where you split the data in training +data and test data using for example + + + +Implement the $k$-fold cross-validation algorithm (write your own +code) and evaluate again the MSE function resulting +from the test data. You can compare your own code with that from +\textbf{Scikit-Learn} if needed. You can alternatively write your own bootstrap code. + +Compare the MSE you get from your cross-validation code with the one you got from your \textbf{bootstrap} code. +You can also compare your own cross-validation code with the one provided by \textbf{Scikit-Learn}. + + +\paragraph{Part d): Ridge Regression on the Franke function with resampling.} +Write your own code for the Ridge method, either using matrix +inversion or the singular value decomposition as done in the previous +exercise or howework 2 (see also chapter 3.4 of Hastie \emph{et al.}, +equations (3.43) and (3.44)). Perform the same bootstrap analysis as in the +part b) (for the same polynomials) and the cross-validation part in part c) but now for different values of $\lambda$. Compare and +analyze your results with those obtained in parts a-c). Study the +dependence on $\lambda$. + +Study also the bias-variance trade-off as function of various values of +the parameter $\lambda$. For the bias-variance trade-off, use the \textbf{bootstrap} resampling method. Comment your results. + +\paragraph{Part e): Lasso Regression on the Franke function with resampling.} +This part is essentially a repeat of the previous two ones, but now +with Lasso regression. Write either your own code (difficult and optional) or, in this case, +you can also use the functionalities of \textbf{Scikit-Learn} (recommended). +Give a +critical discussion of the three methods and a judgement of which +model fits the data best. Perform here as well an analysis of the bias-variance trade-off using the \textbf{bootstrap} resampling technique and an analysis of the mean squared error using cross-validation. + +\paragraph{Part f): Introducing real data and preparing the data analysis.} +With our codes functioning and having been tested properly on a +simpler function we are now ready to look at real data. We will +essentially repeat in part g) what was done in parts a-e). However, we +need first to download the data and prepare properly the inputs to our +codes. We are going to download digital terrain data from the website +\href{{https://earthexplorer.usgs.gov/}}{\nolinkurl{https://earthexplorer.usgs.gov/}}, + +In order to obtain data for a specific region, you need to register as +a user (free) at this website and then decide upon which area you want +to fetch the digital terrain data from. In order to be able to read +the data properly, you need to specify that the format should be \textbf{SRTM +Arc-Second Global} and download the data as a \textbf{GeoTIF} file. The +files are then stored in \emph{tif} format which can be imported into a +Python program using + +\bpycod +scipy.misc.imread +\epycod + +Here is a simple part of a Python code which reads and plots the data +from such files + +\bpycod +import numpy as np +from imageio import imread +import matplotlib.pyplot as plt +from mpl_toolkits.mplot3d import Axes3D +from matplotlib import cm + +# Load the terrain +terrain1 = imread('SRTM_data_Norway_1.tif') +# Show the terrain +plt.figure() +plt.title('Terrain over Norway 1') +plt.imshow(terrain1, cmap='gray') +plt.xlabel('X') +plt.ylabel('Y') +plt.show() +\epycod + +If you should have problems in downloading the digital terrain data, +we provide two examples under the data folder of project 1. One is +from a region close to Stavanger in Norway and the other Møsvatn +Austfjell, again in Norway. +Feel free to produce your own terrain data. + + +Alternatively, if you would like to use another data set, feel free to do so. This could be data close to your reseach area or simply a data set you found interesting. See for example \href{{https://www.kaggle.com/datasets}}{kaggle.com} for examples. + +\paragraph{Part g) OLS, Ridge and Lasso regression with resampling.} +Our final part deals with the parameterization of your digital terrain +data (or your own data). We will apply all three methods for linear regression, the same type (or higher order) of polynomial +approximation and cross-validation as resampling technique to evaluate which +model fits the data best. + +At the end, you should present a critical evaluation of your results +and discuss the applicability of these regression methods to the type +of data presented here (either the terrain data we propose or other data sets). + + + + +\subsection{Background literature} + +\begin{enumerate} +\item For a discussion and derivation of the variances and mean squared errors using linear regression, see the \href{{https://arxiv.org/abs/1509.09169}}{Lecture notes on ridge regression by Wessel N. van Wieringen} + +\item The textbook of \href{{https://www.springer.com/gp/book/9780387848570}}{Trevor Hastie, Robert Tibshirani, Jerome H. Friedman, The Elements of Statistical Learning, Springer}, chapters 3 and 7 are the most relevant ones for the analysis here. +\end{enumerate} + +\noindent +\subsection{Introduction to numerical projects} + +Here follows a brief recipe and recommendation on how to write a report for each +project. + +\begin{itemize} + \item Give a short description of the nature of the problem and the eventual numerical methods you have used. + + \item Describe the algorithm you have used and/or developed. Here you may find it convenient to use pseudocoding. In many cases you can describe the algorithm in the program itself. + + \item Include the source code of your program. Comment your program properly. + + \item If possible, try to find analytic solutions, or known limits in order to test your program when developing the code. + + \item Include your results either in figure form or in a table. Remember to label your results. All tables and figures should have relevant captions and labels on the axes. + + \item Try to evaluate the reliabilty and numerical stability/precision of your results. If possible, include a qualitative and/or quantitative discussion of the numerical stability, eventual loss of precision etc. + + \item Try to give an interpretation of you results in your answers to the problems. + + \item Critique: if possible include your comments and reflections about the exercise, whether you felt you learnt something, ideas for improvements and other thoughts you've made when solving the exercise. We wish to keep this course at the interactive level and your comments can help us improve it. + + \item Try to establish a practice where you log your work at the computerlab. You may find such a logbook very handy at later stages in your work, especially when you don't properly remember what a previous test version of your program did. Here you could also record the time spent on solving the exercise, various algorithms you may have tested or other topics which you feel worthy of mentioning. +\end{itemize} + +\noindent +\subsection{Format for electronic delivery of report and programs} + +The preferred format for the report is a PDF file. You can also use DOC or postscript formats or as an ipython notebook file. As programming language we prefer that you choose between C/C++, Fortran2008 or Python. The following prescription should be followed when preparing the report: + +\begin{itemize} + \item Use Canvas to hand in your projects, log in at \href{{https://www.uio.no/english/services/it/education/canvas/}}{\nolinkurl{https://www.uio.no/english/services/it/education/canvas/}} with your normal UiO username and password. + + \item Upload \textbf{only} the report file or the link to your GitHub/GitLab or similar typo of repos! For the source code file(s) you have developed please provide us with your link to your GitHub/GitLab or similar domain. The report file should include all of your discussions and a list of the codes you have developed. Do not include library files which are available at the course homepage, unless you have made specific changes to them. + + \item In your GitHub/GitLab or similar repository, please include a folder which contains selected results. These can be in the form of output from your code for a selected set of runs and input parameters. +\end{itemize} + +\noindent +Finally, +we encourage you to collaborate. Optimal working groups consist of +2-3 students. You can then hand in a common report. + + + +\subsection{Software and needed installations} + +If you have Python installed (we recommend Python3) and you feel pretty familiar with installing different packages, +we recommend that you install the following Python packages via \textbf{pip} as +\begin{enumerate} +\item pip install numpy scipy matplotlib ipython scikit-learn tensorflow sympy pandas pillow +\end{enumerate} + +\noindent +For Python3, replace \textbf{pip} with \textbf{pip3}. + +See below for a discussion of \textbf{tensorflow} and \textbf{scikit-learn}. + +For OSX users we recommend also, after having installed Xcode, to install \textbf{brew}. Brew allows +for a seamless installation of additional software via for example +\begin{enumerate} +\item brew install python3 +\end{enumerate} + +\noindent +For Linux users, with its variety of distributions like for example the widely popular Ubuntu distribution +you can use \textbf{pip} as well and simply install Python as +\begin{enumerate} +\item sudo apt-get install python3 (or python for python2.7) +\end{enumerate} + +\noindent +etc etc. + +If you don't want to install various Python packages with their dependencies separately, we recommend two widely used distrubutions which set up all relevant dependencies for Python, namely +\begin{enumerate} +\item \href{{https://docs.anaconda.com/}}{Anaconda} Anaconda is an open source distribution of the Python and R programming languages for large-scale data processing, predictive analytics, and scientific computing, that aims to simplify package management and deployment. Package versions are managed by the package management system \textbf{conda} + +\item \href{{https://www.enthought.com/product/canopy/}}{Enthought canopy} is a Python distribution for scientific and analytic computing distribution and analysis environment, available for free and under a commercial license. +\end{enumerate} + +\noindent +Popular software packages written in Python for ML are + +\begin{itemize} +\item \href{{http://scikit-learn.org/stable/}}{Scikit-learn}, + +\item \href{{https://www.tensorflow.org/}}{Tensorflow}, + +\item \href{{http://pytorch.org/}}{PyTorch} and + +\item \href{{https://keras.io/}}{Keras}. +\end{itemize} + +\noindent +These are all freely available at their respective GitHub sites. They +encompass communities of developers in the thousands or more. And the number +of code developers and contributors keeps increasing. + + + + + +% ------------------- end of main content --------------- + +% #ifdef PREAMBLE +\end{document} +% #endif + diff --git a/doc/Projects/2020/Project1/pdf/Project1.pdf b/doc/Projects/2020/Project1/pdf/Project1.pdf new file mode 100644 index 0000000000000000000000000000000000000000..4ce5df235b6d2e1ad87c609d4c97ee8b5a1d6645 GIT binary patch literal 281744 zcma&NQ+F=F!h{*@*tTukwyhmId1GtGwr$(CZQHgzXJ*aa`7Tahb^n2?dRA4FDT<2I zG10R@lg%v+uR(JVG7{PuSwZviLNmyk0?eH)2w51}SqcBY1I-|AY2$3_NXQ^=W9V!u zYHDm}VhYXA5AEdaXliH+?Y?oWE#q`Jg4A=Vk@tdJM;pV)z`K$hc!Gd;p!kd7 zp_073eSCGAc?+5xrdnh;DES>t;+kLtn%4$R6x;?fHw}V)spzZwMhbe)I2e%BGjFzjd(fbIF>kijE>aTnNz8N)~6L_ZCPFuGp*&*;| zf~uzMR5Q9|0oL~A0sf{n-h>RmYrGEJpV2z`u$?tL3>jtW5VPXw_`9XeI--TnD|sBD z%B_&h4Jgs>W_8LsibJfG_*iAX;doFyOt_;Qr5Bboj*Xd-8%_Ig4i3IsOX%N3$v!y) z%>|j^Ux0fAri5*4#(3OaUO%5Dx8`VuJj-`#&B?QEtKQ8QMelJ5;EQRpU~-b})h~D6 zQclZG2izJQ*PfHflGAp*AsbgIamdm@l{U{wR$hy4WwZD&TAJ8Eew3Wqd(i+{^{6Nr zj+-YZbfX1t@b7QSv;+6-FHggK=IW{isUKdRqf|89Zpvh5fI!UM$l7>B>pZx!?ro0f zUu;+6ps&j=+icQk5r^LLwCQLmE!8K7G4zaigMwdC*wS#orYdh^8As9ityGX&{n2CDnsO9&e}I(PzV$9QT6H; z2{0?Ep>e@qCQE!ypk9a^is5U2o|_$(pG!S8y9w{@7+@!Wb_g{(S+(w+iNuwA-uC5t zL(nZT)XAj;tGk(%BWb<)$#sjivwWm%QV0ws{GX8-%ZWsYTih+0g0AsO{qwm41-jau zTQ+9^C+GxfDN<7T)4wgVn6iCt4o}Twm<`wALA4u`aM< zgXVok33BWS&SE{Ht;1 zNVvnAQQI*CtN+`Ak74PD#{EnlrA8WPc9Wyk)$SIcdE@*?@nt;)+HEeyhC0cgn$ zp7Z?H%JfL?l>jT7kH}X}e57v#J6`x7p(xYEmbuFr(;?2wr5*U{9+8(DFtE`*VY1?tPqNrj8!C;^RnqF z_W`H%_9ce!h(WagtCf#tdGuJMgTOXrMCVA8wb!F^m9rLx%(gh_{e~Pam}P(C)|A>L zJZe^#X`=Sdk{ECV_4-efl-n5cTKB4z1EW{(>RgsNKW<6J&=#AP9plzFCV-4TUY4rg z&;DO#pWvY7Ja=2q8;?HwNZXxwJ+d?o{|qOqz5bk(V8vdk6n+ePVaO$Le8NN67%88S zv`0mwp>Sl@oT3?bP*efb51}~+2{__lBUDT&ar1r7VWPR2lBdDbk!8*5ASDGHI5BEN z(8JmfDVBVGE~-MlxC8=U{4m}6jJcqCTzEVRt_ObD=1VEe@bPi zR1LQE>|Vu>NLxl0wHiYFLG#>7>S z(fJ<}$>-3t$`B*VvWK+y^`!2SHlkeH+I5LW0LM7D6XTijN_4tW6DrnwHcQnsbb0Q6 zI1_kkDT#67d{Z-78t34~33r093LXqr8T&m+AP$qP{P47@XgYk-yfE-bBXQ|2G?hfj zwczZJg0cAn*WbTp#KQQlF}m`j@>}5JlL%ms0ysR+=EL%u=t1t_8na>9&WUp5?H`=3 zBMMqhuBbeTjZGi*Ku4;K&BECd*XITot=P!c0z_#8V!S0bW1w$iobV@!_QnJ`X*TY_ zTs&dE5puW9v%Hl);*#YM8*a|nQ;fwOzrYu`91mJ0XnwE`vs|=1LR*r3;)3S+crSz; zT>j4_?>)Rq!mq9NWN5432(9+z5X~&u@6hm`x)k?;Vk*FTS}kDVDh`j!XOODM(UXokuk8{YE5(W-AJ&unfX zY<}1Xf2|oEpOKrkD{`^ajSFb6GVoTBpxig#`PvMRUWUzco`x+iJ}$a>p0K#ENW;UO zoAn2$vRO+H8b!Tw9tE4HR&R)gofG5+A2?m%l*ds zMH%W!)#Qd%iJA_lclpeX>(S9)1hnnz>f2*PIKp1}z0RWZ)|@elCGAGVIbK-EXW|G5 z@?vb<+f2z1juJV(7zf_S6mBw4qfy&+vTz-4lA?(-E}z0YObX}re6rbGT~^XzS9$M> z)eVMjUH-O&Iy$y6P`t8z9CCn#UmBu%6!X6Uc`n-eOtL-SgKA7X+k4K%<$E1EvhCA; zRWLjz)*Sm*hW2DsEco#8KXW<(r|tWg>vOk;j#lykpQQa;BUV}oB>w)RZ0kqdTIs2B zx2XJB`Af=I|J4x}eBl`Dxj@o3_Wfb2!>D>Qsy1%wj!i3L-{sx=3@m#eNH5eX_L>x% zq<)yPEO5=ayV+|m>A7j^XtZB*j4(rEltbXVdczcOqA(;lq`O5WA(3rwJdxnKcAs&& zohDr+xjdhM0+20^qqT8q8i5iFaXIwZvvg)2T4SLQH8<+5l#9dd$q5ltXjmi{N3r$`L92ETVl(&be26c7 zEHJe|h`<4X2mXNvuU+%}Xz>|Eo|57m~3yypo; z-SvL=_%HP79SZP~)%(cn8wHtE8)eF73Fk*Vq7ihB7*tsy2nwJUy52|r)o^PUw*-b9 z%;O7D(V{5x5pwSvOi&wva|=mBLpR&2z%<3w>0=oltDbEi=2wK$8A2ea;Ts`;E}uYn z{+8@5R55Ms7coNB03ITK{uAx@EfgW2KmyJW6D(Tfb$Alv218EI?f2p0Vt-%AW>Q=A z8LeA9Ma)UfhWcywNH5t<+wrLJ!7LOelaqNUg@woy<5f!sS~~ zxZl704<1jSLpRkpbBjo`8{bX3(}RK*P+!hSvx2e`O?uL%4ucr+6;_Uqat-WMo3m#) z5kNZ=eeX*rTH8?n6Q$e+(l!{Hx2!CqDrkQHN_Pg*wLm&L9g0D~iX3zpzWd|PDCJq105k@g-1y%8g@CpbIl8jmiA#*Z9)MOc&Xh7#h1|EtA zQb-7WG|ypf*+v`Y`wy?(z~MW}TTn%$xX_F*t7|D4U>2_ktF@6-Vd`S6Er#YB(q+D| zBfH@|Yz5L!(30}r!wWj|u(!k1Lgbew<+}sZ!bMTuTcpAPnT$E3i?P0Wb?jw7DpXK* zMNoPRM~@GzpmX4B^E~VYXtdV!lmF$0i<%PT6uFsCE5|Mb}H`%&c$(YD7#Pw zhG78PiJM0j3JR^kdw9LM$?7sD%FQfTq4e1(pTgV721-B?bj#;$D7gip-N@`rx^~Ej zv$!c2c++KnNprVJa=RxcXNvy1Z;nMk=)uif8!)c7dC@wS)TY%%x3S_YBSEh!fssU0 z;*;OJ5bCDpIj1}Z6)AS_68--v2nKV>fTIpTFDKhCq7GNK`@#xH;=yohg)Pl6DQPvZ(bJmkLXNnAL zpt`Zu!-$>024QGEZr@chRk6NY&CMy^?`?tSk|B~ZZdwvhgT91Cz3^uZEit_us*!0o z2emy92*yt>{NW0$!abYOUnNsFjirAaCBIQ63+}08F|_tMG^J^yH2iJYV8|QqIyGr# z$Q+!jukhGpkgV;}d6o0vVrD{ChM8uRUqGxbtKQYe50n&aSzS2RbNe>7x#v{`0c)5R z?mBo^+ZBwAs1tA>78ojgqS0vW!8T8-^WYr98Bl{m+S_a?mYnZJ%uSkm z3-Sb9+zH6(6$L}g+N3O8wP+vDTs%%2PK6FWWg15uI|h~?bOh-d+MQYcELAW=9}kDT z#mSp;_LY!^*ZKpos+b$DbcktsU!$KtKSpsyuC7ig2$eElvfd{NcwlCE(;?sjzt95@ z{b0V(?ub+6lYkI4@V+J5`9C0?VU;|x33Y9r!CD>B_)V_9L$PD`btJU~;p8_1n?WQ$ zG3*7nX>WAbbjjcP3gnPp$qvkt#?P+PTW$o2{7dymNNs>m{K($;cC_icuCw-<`fDdj zK2FxVwF9KZi?XV8t7HijJ)4?Y!xd?ko=h4j^gznm)a3f)??x*Z$_U1~Pm#R2-B4+! zr1fT^I56hj9uQ8cK<3jlXB~F7&ERFcuX89Cqw9S>Mk7M+CT>*J?MJ1$e)7lUxxP9s ztS3b`?`2`bd7}7D{ql1F0omM?Q7enE$PR4`%s3P9#d-ahe^jI)Fkn7+VrBEj%bM$$ zwFhEXN^gZ5iU5Lf^J%*(Hv@jU=B9K-R!C!oteXi|TWriAq-d z6|MJ#5O#ysvYwva?S=ERLvqcJck;*XEuf~yx3ED?-f?*8;shYZMUlQhLq=-o>C&jv zEfez@)Aag0_^DcQC$UNZ_<4B75vQ;?w;E(AAalX=({V~cm22IB^cF0#Tfs@bI2tw6 zP#1G@)c0i+yY#Q|JC7XMW0AXJ-!n{95Hbj%qL{^aAW z=Y6-JUPR4yuyfgc3uT*32%YG@im0)WJ5{{3LQrnYHH1hZ=0B!*KC;gS#-b)aoz`|5 zg)2b;wIfcSV%?68OSrRX!QmC}Ez;Ih9?Jr9VVXItd=9Cg<)WMC)?z>rJ^mW^boM`| z8h0_^hRn{RDd&o3b<=zioK=xBG`A!-&f-dbM^!?ZK-!D|(ODZ!BAgB@Xk6c|=3s6g z;>suu8s{DHe4di#vX?byg8#Np*QYft%R+*8Snvy^9%N?1s3G%AR}Pv@>FJcEpTvZe zh z1ShZCz1<=`xo?B8E%|2-2RH7-jbzr8hfEu;`WGU?N2d^JuC8miVRjgx!HqKFk+S6G zyVN_V(dZ;%3$Z46j4ss;k*Nj1lT}Ntc8J}s1G+(#5Qhmr^nqEK4czepsp#NB^)hEW zISRy5EI-|@vFR-7x~2eOTc)EDq!gX2?xoM7nj&)tN`stoMwfSHmu(A|2j>(N`uN@00t5P)MFH~WcyDS10 zyr2RfV}~T0TDyf2`0=14leD!vp*M{~pzmL@W*{)rA&4o%mfD6Vrc3NB;onP?gngIn z(T2o5GbxpwDXE^@er06(%F5YCE%O9tk$8Ivqa$+dyxZilMPkHT*!bMHwYA($=URV# z?&sBzzfmcJb?J2$NN^sBfKff9$5;F=f$`;A44Q)haPofVZc97T0_ zZab8L&)N)(P|V>nslY^X%EEsc6H2Z64%~gE!f?ZcJ&@VHk_8>nEur>M*&q>AwEo>Q z)X{+KVOd^@<#aih1%)O=9%afp1%!CeW*is*)pTP@n+Ba~$=-e@MNDp@QiH~p8#U=9 z4J;OvqkG7+zET87K)P-~0Z9Z*j_`3O113P&H+_#sU}343oG?bY|4?w-MLUUZg`V)l z&a*b>BErpYUal4rMW@G67YO;X^OAAD620lze_H0~4u+o{JqLuFA=Xu_so1FIe09|c z9x)K+?^!Xe6;SOf1Nk}8;a<&(I5Nh3chvQf2hZ6w3@o-2Ng&sY0J9V%+HEsgN`}2b z({j3(&r|4i_q2Z;TS4Bbz?;)mt>nLM;MeTFMo^OJafB;TPi;s%l4w_k$A;MoPaA&-aK|N+)R41jT5TK4C^fcM*);cpWZZ#)*p| z(uLpf-I$4rNOWQVd?-Q-KiaMJ5+Q4>w69E=Ao4N!*B$lXONnx=auxF9h`6s;|cE(O}Me*p-M&+W)xFg}0P_sUn)YmK|%{@ot?&acyeG zpiDKGGfxxRn_d0eFdlA@IH5-tMyG+FwS_;3d>C>X4b7i#5NSg_o{jQ${w8JQ;s$49Z=8X|{Yoq# z_4oci%>%Lf0SDyHXHq$XNn{F{G;t-f4?kp(jVfeO&hcqrADl6)E`&CLW||k3dmBfz zn;k;Wl+`44FoX1aJjIUboqH$|3|xA}uOfrEMFtYP_9DjSodre47!xSo)|PX6bbyp2 zc4S4BEg9KRE4F9~rLR+PaU-yDZ|fDXS*CTB3Jq#3={H`HcKPi>iAv~nkQ@g+zb zeL=NXznzw$ecP=YxfOE@rWd*whMkeu!%X4!XLz01dC1(yJ3iUjaUo3gb*j$NQetjh z?{$WewTx^JyX-LSi#SD*3W4`$@t? zOL`eY>*8D%a9E*j?BOC*dZZk=_d2?7czu;ZBE+ivU0*%`j$q_3O98ytm6Oi6BcsNT zY8! zMGTG$HJ3c6>N>N&a4*)Ozu%Cc&t&@#|%FpL^s?tV{8$RMZ;{V6c`+uvP`=``Uk zkfP7sPLFWhWAQ@K^v58zmo~@2q_3E=%`B1HlW)t?grbTZGX zvumuz$<4qfdABVKImeFXiHX8hH_{4io8C4w6L=nMmrRF<+b929iaASA1Z>LG3pOxdrq3%Gz zmOzH6C0>XQ{Basf5eCKiW?cr!362Y!wj=lf!2&dD(WZ5h|0)5__Q(Fr#^Robe}32pDl_t>_h6ir`;ILH8G*&}rj%#Fil9l1 z_WIy{5fi@@yE^nOI!?~ZvhLO&-F7=2)~N2YgMdlE%}Tl5xlWsWO$dN-%7vFbg(UVk z|7_}KsyWHu8F(RxEh)^|wHyE~6~DqN{`S(Y#@MY$D}G>pbrO~oxSd7RACjvDZ4zs$ z6-7qh!k#?+;JX=u8gkCkGJeWldE~Y87Nh=n6neuqRb2?3XeP64uXPT!7!+W`L?CNSxZz+9 z`OSpi-*$%>^;tCBg&Mlf<0FCw%7&PL_1bG~n0^SNAPD{JE;71v#CtOu9M!0#_AX;L zmPonkMLQ8Z9lC9b>qVCYlL{5a85yt|`so-6SO&ee@wTegy#(_8+TS?Tnf=$*S|z%< z+TkB2n~#^CPO-?oQ7magje;GUYR_6b?2EW#xb}Je|>bS0quqXjShGWjY0PZS%|7M5 zjN{#ea0Ry%G zT-iEM^}xxoP*r26E|q0|N2wK_EV;OrVPo19ZCatR8oC^-vnVPdCh=?Q zv6-T$!RF%Kt}ZQpn<0}f@hYkPbKtuvgQN%CELkX`bhs$ty(&5lH$8tV-BD)coiCuy zH#G`_`kX=lkm)^ZxnLiyp?M##vGE8n+GR@>(Dw_>9DnxakRD|&oVJN*5u;beV{0op zU5b8!`?h-lySgeRRm%k7Iw%!;%`cQWfTQ0TM31c5K>g&Kh1PJs*Ap_Tyhs4=p^I#( zxzJH?+iX(n<#Dbz)^#sYL)RhF08djOtK6fBgAVJjPZtOKwuRX%ySl&Duv_#ln3o++ zAA+iFy?baCMLN?Ztsg5cUE6!oo9!wtB6y?tE}EbdYcv@*Nx@YtsLDOg70pjb4tL5l zi%DF=Y}by|Jf_|wM^o`N&P-npI(r@{`DaQOG7La3Q35(LTooYWHq1C5S>$Y-Smb5) z&$fpxm~`&D@5LxlBJ8o5xW^DJG8HT`Ia&48)SMjJpsgj#(S~UupRtkqJnm`Jg)zoB zMIoM<>RY{>Y;MwOBBe1j&W9M7wG@9f|I+KuqlBu`=D8q}LQm_fTv`kZ)&eozp%>&N zB*f7!T^QNx6f7d1X*L`3kLKhg;1OoQEequ(&*!p4eM!;9WZaal+2 zeiX9zT4WCm*QzK3RNTai!|y5df!R!cgK+=3gT}XNPO;32M2yQXNowUcS{fA;g3N0l~FGpW}IY zD^2u=d|rs?4`P>}t}9qI+irrm?NW(i@X_fOfZynbK#;TIOA?ba?oF{B#I8-Ry4-KD zXuYBX);nfE3yuLZqq{OUBCwngrM8C)2QS9#CrpV%&y$7EBPi!g8Vb)6L$Rg(BQ7?7 zWZfE1Y86ALCumac+rEUDj}IdWc6Orp!Xw%p z-NFh4C%p8cKRIg5#$EHs~yPL$HBQ1*~t^fFZ-^D2Mz#}0PMFX8-N?^ zKc7TFd)Fna4-roek^xcl4_`gV{}dQ?34Ctl{}$)FTHQ$(FLn!v(JZ92&^_t({k zjiHEG((r()bIhcnu&mio(DI##Q7L|FI;sbw1}C-3Imao5rlv4;kp)C(Bzx8noPKq~Ku=*`Hv3ItHHELt21S z!(T229BRn!Dy-M;kUS50hP^dY>=5!PnZ|ad0NP*^TfJLOId3G{B3WHCG+Bp`?DGZY z;7-xtTQoWby?5EywH4Bj%oEO%jy&32?%n6*{0|pNGp|`*PpO!OBlNf92HVV+KQ7$B zi-FmYlVzZ#T$8+=X3^tpAgSG#CFhuM9H($wW^ueYchP0;t}*&_W!ocK89juo7g&3z zz?#rZW`Ac*0tz$CfY~t|(G7r&Yo)BTvihUQTP(IkW}NR&J`N6tfuIvdpdTEPVHnT6 zo4Itq-kcgX21NcCKvTB46+uiv#^rurghu))N@0dj3aEgu)M)xX3xmVh^hXkheoTO? z3ZVg&XDq=$8DicQu{H)Ng0aIIwWMXp?C~KqGnqn_?RB%6=Ax@TdDU{sXPEf%MYKO7lic)TPfPB-4cmnl0^#l{scT z(z-zHu~N@AJ6T6e;zJAQ193L^Wx!}kps*x5n%RJ*Erp?+HvrE6hrVcEc6Ye7Oht|! z;(s?dlby5WWhV!$-QQ!-#I#9TwcoH~@}PezcyJrBSZ#dYD|6C9!RM@k;Lh!w@jArB5( z3F$X0xuady)RRuTsrp^LVYHn84KQ*VgRzS zeHc#rhH?t~9h@4pAwS%kyDtw+4PHiv|4Xi#E1;M28KFS7XO^62bUHn_-D6Ldx}Zqj ziC6)f%FlQ-i;SAArU`o)bZL2cVd=imb_3F-&UAkF@j2llw}*uHr%fznXZeesu3NRR zm`f6jQGsovID~4mqKpG~Ce)Y_FbhkjI}8T{xJ|m&(|NgBD#58RN-dQ)O)?`~O)`C2 zbr_k#2u|O{*1ca?htd$-M&ESY=43;m>t#ubzKMykWX^4Ga^~P_cX`b`v2SG{sK|-f zL)ppCVwdWWrb}8@|8&%)0G-saNo~LM)=VHcna|d#S{GQE&ZmA+{n`hW%=?2^bWGBG zS%il$HtzLG{2V;u@tyw6$TZp82_3}Fd(&~jm;KMonB@8oD0qdg6i?;^bOYDbBXHb1 zt3g0$&r@-SITXS^1b4`0Wh*nKT~v%1bDQv*5%ql;q~9lelV_`eHMSSF2-HrB9`FbD zht`iGvZ&Kb5NoiXrKZd677_`bo?|GYsW}RNJGC9Qo0-uO>0xjtkLSjKu?;jPQadVU z4ysGN-&IyV7C#Q_j|n_JNx`b~A07kDTnJ2XZtg4lwM8eeV)7R=7BxlwG9ovtwt`8K z`+nP~n>)ppXVqa&S1+4%5FPU6{0$7>g^-_J^;RZ5;HP2KFpsC2vyX})Fg0X7jWyb9 zS2dlb<>b*<*R|OzR_0`CdA7^LYtxd#7V6f}fQ@PNRkEF;I~o}&8kv0>SWL^w%m%tk zxE#WHNOvmxt{S{)R=MjvhHhblFjI1{Lab2jrxijl2pY_{n^2DJ^L#chkg?q5GYAXh zxiXC=Xf(dy(5yibcfcN!RopjdRG`Q%OF)IEj$!Kt6GSXzgvqAtLQ3n|iy%~vHIsm~ z0*YiaDvSKUQ~{sXV5J_|?aqxwoKM};&Cv6L#8*LzVE=E&6sZ=3dSvotw2X5tBn8DD z@4#U9MPMVFvIzRJoB~Fx=`BY7@55pICY$}@y^_P?ZHS0E%Vzvk9$@mV$jQlOJnJ`z zka>O=M?ZnPmxIRaww((psTq1=#<+9?X@uq#I8`Mpd0L2+b1(5kM z`^_a2w<4%K%Kfg05ZtQR)iSr320n-i`#Rd-Rq>h+R33){zL;BSXERq^lcJCwDq;qg zo(oOks$PkNkl_j^yITmyKKh*k5Zai*{ezGBFBsr%S2qp{=foj@=5RP%0wFkO{F^)= zWiEn^W<80SHat`ej;C9(>*h^9r7~(Ra))Gw?o$MADw(pJ%FzvTzQ#tiB|&}V*Z`TS zU&H0Nm1N@r_6;{@wutG?cC!#XTRuCLHQRL{HCRT*vYQe(Z`Abij4a6XAQ?Oz<}HbE zeYQzhAn%Q{ldlp#LipnZt>nIX2!-^)*(JnB{MSHl_15JU?j?X@JH{Yzbu@-*rU>Of zg)1X(MBMt~(SfCjF&I@pbbu83`UvqGX$2?QMo*EMrk)RP$~{PDIL~WC5B1LvGAE6; zv*d~>&dZm&gnskh22mWQGxPFFtFZrAyJ_Vh=@@n^tE{*TkA$g^(k2~>V7w3)T+Je` z=K6jrY8FUo+Vc^4L@t#l^VLC z;@s>oGuS!Ob~of4I*hn8!S1G#P7H%5Xrz+_dn{ohXxb2`^y0Y4wP2u<5L|M%<{=ls zMpI@6FIGFFv}g}o?dh1V2!og4^f}9P;GcxRXRPQn7dyK=ZD7?U6jNX9udh&!shF}% zwt#wPLB0ek6Bq>-CQeHN&zc_AT+^2J$shfXEB%frq@$56*x%=NP6}4iAL7T5gD#<& z<;p3#V|{rzPFa`(P7r(vafEoC)r*?9;RcIWwRK172R?56+pDe5rw`Rx&MRECnP>O5 zgYOst5w7xtAZcU61wxop@e@&OtFHwM+9>atf17_c>a(z=aww~oG&PIo?+>&q_C zl`M-W2>r3o`Up$%`IEBOQSRyaz%J+>v%< zXusNVOBdrcGjY%5L&xru>H?*~e{=4fK+(_ntaMkvwS^YcwePiaKZyi`27#%T(s`SO z`$|pd9jue_rrSG!i)Jm1KT^!oMsIlZc% zR4OO6PiVZx(oivjiL6W46s-)H9aFc;Q`&00hA#K6&T>ptcdE3lTfo6x`Ph{=UIWtUL}%Z6BMl*n8ioo0?wOLYiBI z(cc|C9wjAOcUF7QIdBp?pL+q&V66r{G(CdN1l|^v`R^gC51;0|02BCO!y{@A^|FAM zAN&lW9rZ%p2vUqdM=DKv&u<_jf2I$CCy}xk);conYcXC*6jQ?m!xBXK0Mb3h>No1& zGx$;JAy61ZY7XD;ClmZrV9sG zlhu#o`0(%aGTpybj%2~ELtix_Aw_bqLNwJWm-fSyCYHhUS4J#|S6Cb- zI8qnfQg~4nU{VgBmqKx zYivbZ#+JJBM9E0fw?VpAz>oSY-VD)n%$h%dSGXvzZN;?-?4p8C$=9q0rrzwI;KSkr zKc2_~`&Yq?QzR%yIIG|ayzPghdCCOhW4?C1mqXNqdzXdGeGxh?if11P5O}^w7?8={ zU`$(J70f48(Sd2A=)rLaT+`?#pVnoIk??z5tl&>`f?w3wt)2V-edd+#fvXOY82ia{ zfc?=~%yuvM-5+c@_ozBt%#j2z`HuYRy-AZ$-{(X#$fYpN7SUrLDDCo?F8U!H&Rn&t+I^L(I;|%!uv5 z`c81%DEE|aSp9BsO#L11RhmUq=A1cuX6kU*qNQ6&=Oi>eLQ8$!knKIfgDLB4|ORBbi|px_e7a#x=`!K7gPw9@0yXRh|So8NWkB5-@) zlLP6D7|DBt;Tktl=R8ia@f%O_8U8oof-D@z-$`%no3cHw{8y?1{ly;4ny=!?&mlhR zl=Mu#$`e#@qa zh-XIgzth;mi&9QZ(P_OqxdU8@3W2yy%|Wsd`^DN!9pQAC-ZII~#(&`(!wk^>2~xOq z1cfn&2)}@#QcrKNpgPtaGN{FlgdP3ojRlDR1x(C2s$xL~Z9i`0-`3+tZWJs4hAZo^tIsbi%)7a}gBM&`n4FBBOx8bkWr?pTb3Oq&Hq1cl)OGk4Ci!HHo#1{g510{%2g zPJz}w7MGkUfs>mGVDdWBVFWKw@*V8NRpMCAc|^QH2b9k@Nzsz3{Yz52wPOedt7?-g z37U8i8o#hFwVE(f0K2$rtK%os0s~T z5JQyNuEH9vV%h;>+zmSQ5J((8pYr__;Bd3ZM{^k= z2hA6DJDxl!t>s`4?OUC-(5}y&mxV>dqr+g=6R!1qO4OJX7XK@F2rQzEOA9{h>P0uv z=NOkeisf}lm~SZRn9}$?3A8X)B4jUs39nO(+c!RxY9BRi#t_Q9LT^I0j*Nf&FJ#A_ zW<^kgZhb`orcqbCmM2b}F3niEdYmTBn-&W-lX-bgBtX>V6~bbqo>f z6KH(l*k0?2Tbhr?GH(JRbxgwc!8sj5V)*Ah;6>J{VoWrr@#L>&Smq8VNg8(!Wf{g$ zTn?X4?5&FrZ7iwPR7omf&Eb~kYJGi&LuexnQSmw(Xgi&3AfpPKP4Im{Dsjw2{e>Y} zF~yEn5(SdL6Qqa1H;!&LR3mh!nw78j>&XpW-~M<%lFX#41s~jC{qOk=yt2U6$6?!T z0K&o2t!MpZ&-Q)I6V^U1`jNCG%Z>A=_IQa6eU_=7w*`t>MPapKtZwO^vdaTCM_qd1xFqXOb zu|fiIPL$aw`Df`S75ni4drLm7%0w1p7wU7cF^f3JCtYzX7wQVOyMh?WC;8)WO};C> zcuUO@pot`koFsY0aVI~;kXl62*-PVS2oib2aS!}yYDwYgPWIEl_2&Ct}Te)Zm|Sr zRm+Om4>T-FC@lT4u&qN50?t7zXYU^eI(L+8ap3>9kLyJq-yj-7=91B@+9T)=21W(jtis;E+8%aFzgC|bu_u}K3S2$g zU5vL!5=$!Ek2RP^iD_5A46)uX-#2Y1$5D}&=NK;fn2e|R1$!mS;%b9v--*fvrqPnC z0Q7TcCg%H3R-j>}iZLeU7%IL!evf|5uc!F`out|Rk0i~+#qvK7ccV3A>_|9}dpmXWzGA$Fnd5;){n8(2rWf{>KJIDTjY8Q}>d@E*D*C*@us#O(`%HW$TOiR@7Jgd2+ zR3*~B-&221ymr1XbJ$H%TZ*bwFsIidmp7M9a8pG}Q|ggEnm%}4bX+saH*N}}S@vGG zcxhwLN^GtZJm&iES0@!vz0F(Ua(B38Yw2(dO~Up=?3MdTD@wbgf!egBs7t`D2RITp*f|dr7r`}@dc~>@B6cx4p7EHPY8pbb< zJPgDXI(*B3yh|hRx0d#NB%!T8Z->4!0KabS=Kay{|8c>D&@h6$B4ZPLp@l@_9WX$m%bkyYuT~(rqppo#J zUtVmT)22B$<`vSRcz_60_Zu^2+tj9Q%&sqH6&I^>6CU{2>TBL^vCsoIF2Zf&U19om zU%61Cq&Ktk6#z3`?ucNJTO8%yi0+^i0!vA!Iij*-!9}vMIc}NK zJfH{6vp#;TGKx#f*2`4P2f=VF%jG}&&P~`GgU@vZ4AAXhIer_%q7=&iO4?<2UkOu( zxla^a!f4F6Hkxrby~I>jPfw28$5!EYk0UcaJh_11G0ufk&IItOET8{C% z)F{gJ8oNz*bsUMG0p*4|ih9KsW@;tyfx7CWTdStw>w!2^K3&$^pr?E{0Vj9c!IH_C z$!C8#k5vPV?Mev?r8P_hvg*Gf=*d=O);{+QB5mZq(kDr0t%zi1-rmVcvCmo!zzjVI z9u^lu*aV?x$cGP9T`qLjqE~cS=;t=zvN8YJX~W4L5)J1EHhGbAVt|<8Yc@jccVg$w z{FQ~@P7)r^s`A$i;77!}q!5f$RQK?dkV*hejW?Pmi-N6<(<6T{(WP*B-yf|!dKG*# zSh-w@7@lLIlSB0OnwfE_1a{MJvm znv$aGY5zInE-;4$1hjUF^k?8_UJi#iz<3FiFKVPWV(VX&Fio=HGc;X$$IdM#JijUT zEwE-Uct%HAFg;r#d3s=kwowJ|F_a7KyW{_h`b~Xungf(nw zw7(CDX^e6a(N>rUuoWbMnwK`B>RCAPsl>5;0kBM%MCPg=Ona~s$_COnO>yPx$_EDt zGpK2!grq8rfWf=N^B@*cq3KtP*8OXy5Dw8XdF6cmKA9;V!bA2t6D%75>{PHEH-<;Q z|9Tr6uvD3Vg=m&-o8G(|j-Svyo(=5;9mfC#rV*E{`xXm7?MVNpPM>$0gib3j z**yelpX^Y~x6wQ`*+U4Qz&cn)`}wwhr@2K=Y08ImCkx6hF925wMM$FBSqhE_!Gp)c ztf5}Qq;#3jV?p6yPGI)m-$_|iu1|)idML7_zGzRD8r$A3A^Fg#VPk9uMB#XoNKHtK zrJ>gl{>XVbfPV~pR9{cT6%73lOBAf{>l^10^Uc5Q8nHaq76P&C2Wjx#! zC+GH3`^xk6%wz%c>Ar^FSb;McIs2Ejcu(H^A&Rhty_C&mwH08m*Vzb-d;#nr#aIZ1 zfGcNaSWU{ljl-sI&tGuBt-*-6TPWO=nu8TZmsK8uQY9Y}c!80AM{F^05`hDMu-7KP z!i+^!(BFhcu3qQ~EJv)_R02GAEbyBjEm8!e?%+BNZ2maNh~MEQ)1}ydvjJpN0>Qkp zPW`G?C8gmdoYYJE{xd`szk7|`*@v7w8;o1Z#?_0tu&}BxA?yU;_=x4#gD`Hl901mC zx0Vv14%|7&P>(p%j9t<5#U!(XC={Ohq2xdPsrNkeg-;gL&Uc;es;2#U} zQ!F*AGqWp8$PU8(& ziOr2P*d=9hd#jm~q*k=jBtOWKrEd{gNUSknmq`3?r` z5%{3}-xkmDe+kYDm0vQU3Rv5mu=98MYUTJ@p9S%+b^ zN&h%<^$?K^)JgdgbGcOA+d`GQ}87_JGbRf9vbRY`cmTewiy6aQs z8=b{vf0XDh9Qf^@DSbU6sll$WbLfsL`&V#&q<}vlBB+vK1hWxq{1FGr%*?b!y0!>1oXT=IF4gM*pxB=YdH+5B^9LwauqyCwu24 z|4GiiTv@j>Z$`#^#+bg+E`$NNl-DBU-Bee>})AYRPg)}{EH^-F<5s!u5{wzp5~?e zzCM3p`{LJ{MnUt}%B6?oQp^-zA@=iwa4>9xDD3-V1#A&E)?(aP-KZYEmHY^>={{VqNCU*6H|qnh)6Cc&B{ zKjy^A?U28I+gSt)p+-fa?mb;ECB*mf?J-*95(NB`gZYT?mJgo47kJ!v#GOiJ!I7S$ z*eFBpmwyK^g2;K1sgLN#Z$;%(=Z#ZQ>gnhL{~~DQP@(En6G-Il+2k`KN*e-lHk#IAse|VqDZ!b0KKd z(M#UGCZa|=vb`qLFvIwQ87##5qt0Rw@rLQ04FWrsu~%NKs3n=+7{YGtE<0t1QexaQ zNo@<@V&<#C;Txy!L9n1#da@x|H4`J zW?EZNrz*mLNbC|?71%#g2rB9H^0gO`R4+T)?<4DsPrg)Pi`5%d!s>f;r6Ne374Mp z+t>_NhOxh;K#}y&?9Il4$Z&-85o<|zL!(~bu|7^F1epv5r;-%PYQ~(ZAdqp*bZ~vg z=PUUday7k_oO4KxCv?yM_7TUDc@a)SL^A27xs^7pw*#$?)hBk0-~i{dB~llf#nX<+j-4^vbvS0u_Gle483;cLwFE*y z&}9xLnNR~r=p5R0_;EYJi%Xg3^l<_~a5qf`b*-@fQ;?B}Wstbb_T@{)v>vdpi`5jo zb{j#{DNmS>eG!=d93EH~*Yhi`^NwBMsnypHA-#aBuNydFxq!B`LRH&!n>s`yF2cN3 zHpug=7n9Q#JMRGNonjDK1ZS&18q-=mHtP3HB=UO+E}1lF*uge*gPgH;72T7a9#Dg4 zJbMt+2=Xc7AdY8z97dn%f#uB5?4*SPQm!+P>x!YyF_lVa`Dby#A|8vZ_RJ*DnDbts zVCaju-918vPqJ=MObCKH0d$o!COxT#T!x)AjY)KJvr~78M%Ux~{*t1T)QN7JTMwNR zcgwF^Z%eJ3gl*|+WQnsw5t?}3>i|~4YD|ipt>}Q46HIPmg1m2`J*Qug>ft(QyzCLM z${Dq6LLm;>NiQBc=wqM#m@0ob&;%g}{8+KpstX<9{9nK(^KL8%hjDW#sj!4zi-@#O z92l0x`c38{d-Iv73Md7#>cy4iqBMi|Y)fMt5@(FkoRL}+PT9;-Kuk6Jrw~yVL?j3C z`S}(wwxcGeRWb|EBPCB5gb9IG8uo!Bp!o#eUk5=H!JUjFb+BTb`nt3Cy;%5^mhZD! zQ&HoH8?d3w@7s<*8+f@$g#qS5{Lz_R0MePWo4dg`ymVvV(G*fDr2{8kd&Ic@(p+47gIP?PfDz8!HoAwA44y&l1rXkn1m5^A7WH58G0Z`!d8{8b2~=}yq~>LTVb_DL zyMGV8Mb;3smTDIR_A2;n~$CPyj9uaCDQs;S*T5 zrNO6*#!S8ShiyGIHhpU1_N45la1Hz~GW^~l2W~W?7^PR7zAf5XurZS{ChLSbR z2AhZ)VknmpjuAC5bi3w{Kd^H#GERXYiIdoCsl*RZeg!%iUlVSALN$w5a&m+XM8RD~ z75M6=mTVRolWvut2 zvpqbo$TqWoTxqIJH3H5|A~1n=A{fH#0sRb9POeMRldlhpoqixpA~UD|Df&1V|C4BC zVrBea6Rjy4(s4&ENIkc;XW50|EzLybr;K7ea=;+)SB!goSwVk>Z_w1}!3I`Ex3#WfNPu=iyRU9Y=h*Hk?-y9Jg@XUnp7D!-9q~K?l z_h0smNg59Q@_txvKki5thW}8|K!c&o3*v|*hFPTY!2jO&*AM$cPt9{Mn1nic`B@_r zqL^6g_N41n%B$Zc8woo`yYg`*i!U!_+Fuq&gb_=g0rBx)AAoTPD>F3;7SlKPl zy{t6h1BqR}>n?2L+t~9B(2GBr_JsMz_VN!-f?hGhZX~gsPzJM)`$^9E0R1Q3E4{xF z9-AL$t`W!cc{=$Gyw362kSRC?mlYOICvjlpzsGEpA_CSMv>JjAdM}Wx71=8<0dhy# z;)0cTyM`JIA4?ZYk2#vR=1pf%bgT%n=nO7NDnzOQjVvfIB5-wMr;b#uA=*q6Ce_-2U1#L((% zNoBdZ-JU_^R|7lorvUr{?nipTRAKo9Hj3Q%qKCPw9e z5Z;phWy3i4D7B;v(zxH#eqy@Iu|SbiP_%zn@ycFiRofCVEt5<3iz)RGmpdd0hGB^# zGXKZ3a#o^0rsH@lORDSqfl)FLx#?ly)v(8(BN)ycF~#A)8OM4|K9VL{x$t|X)sv$3 zo!7B8up9&nwi(?hV%5lD(dJwufO9Z9&+6z0nx=0kHjfpqy%5vl2gC`#{}Y$VG?E}7kw(W%}PTO$1L{< zlU4B$JrGTF2a2pHr~RLl=~K$WQ3E?%s;6vl8U+GYW|!%*rx*7k>HT#mF<=G zuh^_ljUEKbySSAET&~H3Tq3YoV&prXv_c+9Ru5mCr%}m(La-YH32_xiq2i79`^B~! zQT?jK$+-?qcG>Zf42(?vQ5b5s+sq9^`;F&F2V@dnCxLvudIcdJHkFesv(Z)9AI;dq#F5Bq9Kbz@DBvqNC2SL^*j~?g&zKlyEbzp0QG{z zWag=1GZl@=2Btns9z&<%yaWl1j=r3K`5BbZ@JO(ddoeD5KKN)%`j+;3CuBim#4y=0 z9QL5Wg&Ud*&{D)C$IQC=g~S<^=0eu+{DX=@3iJ*q`B;L0A&91o3;DQ8Yui znjvdli8mh5bE5#`K&#ehjfmkYM@ulB!wEd-q0v7Rum|hWWrja*T=Tb0Mx{MCgwGk* zNzxK5>hnBpcdUrKNSt|4um=VL(`4z7`5Xnk8Z$6s@Jw4HK=$hUq|hB)F}rocS1)WI zi}#e8mAoiof(Q~T8@(&~{mBm$a0f`SX%Q@zPU3lmcgs7FFbJ6|GVZ~;5;uXg4fGXD zqb+Gz(;mUsiep@*t*yWkXSD3}!C1P}L5+s7g zHKAyT$+IoEd{G82=0tL+%AJS8+1-yHzJS?!lifLk0=lK{TEYWQD(#PJh9O(TUve>f z$Zb0i04;)ySW&>ffW51%Us)k0fbGx$``-JFwpjCS+MpBM|6AGbUo6oR-LDxv{8C{B zDn^~cwnlR$(!%(Xm1z`%+#oB zV1TR&yvv$B27^|!s0K)mX@-7R=~d-QBSG;a>@g)V`-8aridicfb%&T7@QTj@Ejlrs zVI_A4Da8;jiXsQQdiR{P!9BYL6;@}*3Qn44+G-Xm16!bYiX%Y4G*CMps%wEvq$wrD zU9vg)_(2yU0z>=-fB3YQ#qTtcW`j(D`OC0|!|}VCy5OKa818p#6~wb)dG*j z)GbHqr!0_Iu8)OP?`b?+T0{${OiJ5`+9@tdqc)e2Xyvz}y16|xOnSvg_VUU@kZEL( zGMk%r9-#n3ifBD4vpm6^dp2p?bY8vIpfPWrzrTk5 z&60K-0bG;7hHC5k{#L~7s19&*36rw_%DQ!NGS7}SU6l?H+@JG1qna8CL6SHJU!zA> zI+k!qwIdkPd#e4(BGyNgov#Q8epU z&Mo`(#eRZn8KJDIO4&8iVfNB8ma3^Nv|gcWT+e}xT2*}pT~L$e4Cm;~BOkTpC+e>S zs(87EsM{nS3W!AV%5HUIZ^$ZfL|eP>-)37`{}M=X2B<+kcDww{-tI@!w~wF}!2%xr z5`&foFEI_C`@(~g5dv&{K)yR$k1;rSco6L5Z$9!O$J*euqiI>DeLNQhMR? zS=F3sMKH^Tgn0CDDj#{}DfD7CUeZU=vfX?;ic;Olhu{2&ANf_zFTaQs5l2c$L{(>u z8*{C^Rph>~xB!EQf)O7IZk0>+L8Z~{dW{VON(heZpDZu=&*CMt;Z=^xsbOa2GnZCZ z8`hok-`l1TYZkRQseuU-?F{8 zCwGl*2;e!AtZUQ~$9yq`gsdp>Blzh9sHAflk9n=0Zuj6zz@5G?C(A^@&cgXWU(5sy z|NVo2gN2#xf7@OX;|!*cw@pe3w=l@ihseypJS0I4xqu8z71QJ_ zAps@fCgD~JwBY+X(*4r?^P{!fYF53Q-T2b@!uN4?%6T>QF7_z}Z|@b(g%f&!jtr0j zQE^V04)F&ND2Mdpo)@2DHNl@XFz7c|{%NQwgl{234_p4bs%@)hy)}gbkOl_x}cxTaeR>PtDN7&YA7d=5E5S=Tbo|*A7eXy4*r+) z0rnu@8PW*8lRA+8m%AdZz>z@xt@s5$%-cVZZ$I^)ZmB;hXFonfF7)iJ{GU30z~A`% zW4ITg-!Kl7I`m7`(9ZjnM<-_VBFkk zn+~+1fDXCz0N~==p#a+~7y!orQgAOh6$uSIqKM!& z4ngmA@Xr|Bvk-ypujkQ6JD8vDAK@9qaC%S~4rVZj$>0u^5Vky3iusQ_JJ|7v|gW%U=CHOn`?qRedR+g#cb1#=S{1!ei%C$mfszZy$wDV93Ar_ zuck+OFZMgRN-pTcKp?&K2Y-H=OM5NCFLS`B)M%e+z8yL(lE(c z_HzCg0HNR(n65*|R8BfIUw@+3Cio%moy1ME|M5=lTrPAbPpBn{`JOB3_MLV;nD2k; zTya0uo7{-(Bwv~p8?(_aCR!0w7+@!ga6#4=eq#x-NR~v9S|IgmQcCkra@$UTCDSk? z21kcK`Iv$&(uFQQTs^0S7>0HN4@>?igg_iD3u6>nIZOby_y=)=j+cewqi+CZE-P+_kn}BepPw9oFyr20OXHz`g zZ}R4-!v^Xjm~LA+-P2(o!*e%?sd-{G=)@)a!5eo+hT*g@u8`;HA8P!DIOpf_jZp9- z{!o^sF1`5M4Ey!B%wKO$PqFUYJ%qd`7OnQ0qG!^5nu_q^q`vi827XaIJO$l1bEZDP zysF9~@XvzE+Sj%dr2=#AH4*t}O^0Q7GI*uMVyBa+-RhX$5(E${_zyX0H5%s}I}@g7 z%1~TW1}G{0nMUlr_;?f^)wEQT4CPp3vdPh(#r$Xer}weDXj_rC2&87@^2#KNtZIG8 z5KLZ!p7VhTl#}x9($O4@rlPpiWVSdv*tZiiB}e>Q;WD!s1Tg75UA?Pf>rCLSf2@u?_Vv$Rv zUgQVhPJQ%3U0|$nJ{6aeOQ-L5^MQ#@xd&Fh-D};=C2FSdaG^&j7#v*_BV1IC!)*=s z;TLHP?(90Y!_OIc3ilaw1O%lsv52zQXxJ6WkcTWsI%!T07{lB1s;i1BxI&>kNxP&P zIu_^f&iQ`-XQkQg>a4ilzt>PRXg!UdQ)IroiR?onw293h!cDf%MLu+i`&2;7mA`WS zE*oi8qU8P>ZX_ql27ALH9zS?17#*N|XiKQR;^@*D=hy25-&CEipxcI2dom#HbX0sM z8yac4aV3$^7VW9p@~l&9b%9OEYmW@bkWbJQ`(BVBgP z8$&`0Bk@IZ%z50%*>KgGO8TAMU=7KcEKz`+DP})n*!MGjc#d`o6FbJD8%q^|6$u`x zP7zfiZA}eo9N(+aVA~JDn0islLwjse9sSYR+3;&#c)X&1e%V`=YZq77JxTB8rP`yr2u}s$2N_9`qST+e;$BbIV8CusIl?@=U)~Uvhs-9Uo%b^z1JiTT(&KnO|B=K-tl;y z7g;XCvTdd)$IFO55JZRgf`7l1m96^DijJJ4(Z?MX-xS4m9k{?z`Maw_DiL|oofND3 zg%!4L#}&z;R#2};6+Tguz$l($;3pF$98Z^vqHsFely`oB9RzTnhDLNoT17WVp*U(} z`ru5&DCsCupBI{W-os+Fh??jk^QMi}qgS4c4l^HJ08J|=zLbafk;t9KS(LG=Bl{8K zIYBbO8Zi@{<0UW>LGFgW=haQfFKpaC5k8F*PnK!!D87PzXmQl>eaKO(&-FV;lb4Q- zgmzv4B565_%RYhre7Ks?`_8!i)YbMERGsTJWfBXi4EG=pml{`M5;mW3Hb7?8CQ#0U zZOcc#jeuVovu9k>0Juu`4HQ z3S^#1R8Eiz=J4b@6(FHf4EG8|}?`jeJWMV~5R7d6_~uFw0|& zD&4?-_ZN~4;7xMWO}Pba5@+)H?6OWp90ikWKC3Lf0+tIzCoS|A3cX;Np4cHLb08yI zDRnREfn@3`n(XDag_XP4U44mVmNfw$k#F5LeWuE`+B;Rb4??3{Q|{w6e8rK3$Z%{$ zif{#XnSK+6jA9W{SZA`qV(f0mr=W=D0nTmRVt$+5vR*A5kD2h``F{ zoA>aB+(*p1J$&6+YMXL}nWcJ#Xw#D(s6g)<4WZI#S0A$`-*YSHYB?wq+VjZ(o;y&i z=|gxavx+xk z#seS3$_Pq{$TR=O5Met=KbF@$Q#J~VG0EfmCIbmh-FkuCk!WC3K*QY4RcL{g_rY7D zWgZxE8BVV>qz>zqT%FZPQLn3^#|ZbLa>X^4un~dS!+}lSNuyEPH+A!={&h?{!uZ5^ zgN9MA`sC{S6^*;TTJ3A&JY@I{*ff$rZC5iHrXbq=3 zUBfhyK1M1Ea(jd4TiMq3|0oO_mOuX>Orn1Zn#}Ok%@<1n4~w6;pY%oBgfTCyBcx zT0dZB!euMu&)-1u$s9K~1K+6erT#PQPLdM0QG07%a?Jd35JPIV z$S5*f4odWcw|sA_yIiK#&LRZR=deJileC3b$|%t|&9Xq$X^3I>4)&k5gZLU3y`9Y#-mn5i6) zL77Fen6XrkKGWUbY(K;T@V8f&Py;c1L2pB)gPa0@pG7IH=%&!D>LVKzJ+esTtfKFoNn;OeQS96Pj<&o1$&P^z({P1$7 zLMrl@kxcrzV>@#b&`qoCtj^*KKBCP@d&VoKajSQlJbD;lRLaLW-TY);bZuc8$oBet z1-qbdLgiIX;V1{J#=)^Xrx#o|89(`h#MIr%(aC3Yobvd$~6osZx`6yOVRK) zTb9`^ABFWpSXQeZW5t|EGXS#0r-`tB-j#b>a|(~qtNf?A=yHf@gv^!jZMkK*Tm+1= zN{PC^G$&V*Eu+Ane~E=C~_Uhj>cnIL0-6e+yq{3_2-TBPP7}5 zzzqWb@%p@Dx|43+>O#+9XqU|#1b>+5t3fcJ1Iuix~b3%$@2#9I{vyhviY ziMo2QsVsWosEfrGyf~QGKyEB0c?O5u$O{$Sdj30ZDA8^8t+=R_>i=+cWHs-g-LG zfWH@bk^Bg%Kjbm}+wXijmG6lu#8O{9ke;rUMQT*6E1>xbw9Bc_frQF_ z;yxz(bTjVBt(i+&N;QM_M2oCh_1tT;@0#DOoz28)%7TA%5xUrpIfiG;f&j_Go&JDX zOd~;UOL7#~C!#UfWaS&eA+!s>kBpi$dSrw;T?cbZ#zoz7gALy4iq(c@*EYHkw5cSDbYUixmyxJM;csK9&J(D(OxWuMk6i-V5Zr9sH~)%Ng@B+hJh+Y=vck13 z8qrnfKSTm0zwfdw>iDg+l;{HO<(d#<4SB&`kWIPruI(0|M8>&a>Jpw z=&0>@aVlyk2UnkCa?TlX)NSmOkyN`MeBQurhrU-1ajd{4hmHKrhJ5(Xj^&-WLd?fg z62CG5H~xKE0>9{1EAg!%b5BXGeG5Y)lq37f-TW6}n=>GDi*;j)=tMuvn@yj3owIglyOz^UiOovWjq6MPN04mG07s6sXL5Z)`@@wVkLoiG+W(LXAZ~l4YxQc(- zzC-T$H{6{~z9BL`p@#l~+`|Y|tq4cu}>K;x6BQ5OtA?TYT=c5*fm9!jyW>->OIuWfa;fphc{d3 z77DQx5>7Ys?E6l^Ksp97>|-IriUP&KI@jI&$K1-_H!qzJVz0HpW+oyg`qS6X%M&+$ z0MRnmb)kg^dwBd~R=%7K;gnRONJrMb%`aV@Q5Mto8%Auy#_tq20XJStSEzaO^+~in z)h;$@3^WbaY5GO8-hZzopV(?c{eXR2ifn)fv}6yDw~)u2ZfvTQy#u)Wi(RVeXkOHz zd)U=OEK>}o<1}$UkR$VR$}GZ`i!cXYn4GZeP*DZ6@?>Jh)okb3Hl1$$mbs5RAn1Bq zDjN`Ecl0nL?&1b8@3qwzfJ-qja*vU`25{QY#!q$%7kifT22>%gqobNt4(d*O`CtFrK)!@@01(H2Wji&F<*uP1a7*YheMY+(TOoZPI-iZ$RrD*NWsFu*yN-yvj z8pO!FSce*70+^Uc`&bGCLep7CVfoj?L1dHuH52br|ZoweCASG@(tD)OlM~@W%?xiZ%`C$SD*nsxMvD zLfDniO{w^*`|`5nO?-UBzX9w>ELM0rlD@Lt=g=l-M4c!7t`8i#7LkSem)30cV<1ai z=aOlwB34c)kZVv=t$qZiGuS^ymOsJN(VXCg>N)o+&6_T&Haf`I%M3|CotDd6l=}p+ ze~BU3?C+?rj&|6?L8q&9;hGMe1O_(TWylp2 z@AirCxwEh6)!*P%?16V%79a5uq;_HN$k^@qnsTjk2uZ}oVlY_Faf33Y;QfB|w(xud zQ?->$h)l<%ds%Y2xmx5OW6HtFSoy9ZV1@RIMmM47wNMH1rPKiEHqp}q`WY@-%5u<)kFcJA@ztZie zC#kZAO$_JxNoZW@uVca&hBMDI$E2G$nd@1cE-z=aqgnSx`UYE7IVh1CqUaB-RWOH9 z(@uV4$kRF66CW0_S;RJje$D*q@qixR+UdlYWvCWgO@x-d;&EX-^E76SFDT<%$NQe+ zF(di%ZQn&@Qc-ETYwjK?N9fXx?KFw#7{2^9=7j*AqIH_=ZJiJsWJePttMgE3U+7lL zmW^<2UCw{!+}rml*5?0XAEF4dV}#vBg=-$ZDcoqvJz?#9lp?xNZOOve2XX#DmscqR z8ixgrl3UIe&GWMHWTU?dt4vXmwYLcRL4O@B(?KFE){T!ZJsxg*Zv$@paLiw6GR^cg zmiOdSWp}NOt-gw+?OF43&whhC3Q>c(3DshRsXHBMJt5o9gtI3J!J2L&sl&M8tW)~a zy{03(&W@vcf{S9Y{X7&Cg7jd3A#F-J`bIq`hn&9hJJ@-gGl~g73kc z&il2Fnj??E+4}0V>$w3X6)QrQsf@WnW%U?pq+-q%jray`H?0=9dves4^~0lRy``Z> zyUwf)2BgoYs4lGTyLba?huJ|(YG8C>wW3T(IO0Z4Qo%IP;oRx?W1(~HXD9jGt z;f`T{2VBbV1SM2k2$}043wKN%LjLSP(otznz}YXNb;t52@WkqV4UjYKc_TPln=})D$D1yd#H}gFLVn^56~B78c7irj8W;^*FynROY_U~QqFRCP+Ws(j4LAD zF6!uoIbTP9__gbig))nei_!1&&jt0!QtgDfjD99yOqJaG)sItBoDs_uRYs_WFx9K8 zxAg2S4~%EGOyXt!)6*mqpUz4$SMM(v{KXHH2l?dNi5fo{k)K2ABGA=y_Hovc*%rQQ zsk>#BwbWpgI$p!7rs)@gd_S)dm{j_Vgj0RFefKU}ky8ee5sExo^~pt_%&&9ng}q^# z{+nceM8C@;UFW_0hnHpn?RhbFpg*3QAVC}-5#0A+e!aG?aZ@p`Adak!18i4x#pS-T ziY>1*2J$DIQeOB!3TQR;ahA4*G#1=2s+o@<=c__hp~7ol6|gwjVwvk3Hxl?ueuP~{ zB4A<__-)X`o}mu65Dn4>W|Q9E5J$e;5w4e+AuAbn+W3)`JQ^ zQ)1^++@k!n(vQ0Dz<*G#ZZV5li;i1aKr{tq_A=3yl!WsNVLus=^c*`WlaG_lBUAThaoUcR$L*k=gQkZ>sQ6$|*m$oi}8Ka z7odpmBy$JFCCMt-&gnUZ%al<@O>PYooC9SfafOeg5M&r1HOt*6Vk0{SzPPFJH6(Ps zkAt>@;7?Qs&@0p=di5nja3A}EYBaMJNARCu1aj%2!D?8e8Z2>j-Wy-KjW5f^ej&79 zJwHQFj|S_D89`Zd9w?zq)Jo1(zVQLUovy%F+R>Iq z*KmaMFLUpCs3x^znFq}pz!ErkmdyT*ubLn-V^J=qgeuzlXLjp5byx=FvGU%perjKy z8YUpLg9<5%S1!|>xt=a9#6bP1!AjBE;j`AyoskCO%|eN_CcQy)!9w7ly9DQ?sL`9w z756xJ)Mrh(pU@PmiYn6bX%!R+frCjVzKvXRl=gNR+d|_)2{v=v!EEW9bS^SX(IX(- zI#fp7p&54{Tc5NK<|UaQjvr6BDQ#8w);BM=^@M$KRWbUL;)E7at8lT|B{OY+A1&qo zo|-u7SipvAkiFkl>eE|;Dbca2$3T!SMqRV8dOK$Tv6kTG)9y~_iF+#3XnJ#2wcziX zv87q+$|hkk_$%!wW&0PJvh?H!kT)VZ_WxonnEnIV`NvwYFf#r(Qo%&P#>x3VfW`mK zTCj7n{;%7yV%*$Rl(!e@y373`CCHr>2%vC_U*@8SaD^oZ5RoBuJ?Vi1m5QMx#6e4l zWd$OgDWy_&yu>r{-~Z9#-Y+#Sw_H{^jpipA>&#yKv-n|Ava<>T{ouxM>cjcEV10oE zl>Uh=Y$8ao!2SXQfdui0nCPp*#njN>wE)aCB%I5@q5P6>kiy^)VclAv358XwVm5HN zIeTHCz5zg8I;w^aG$=4+fI#98fKh|epJ@r#0%Q&i^tZvBJvtK|X><}gnuXP=XZJk6 zG=M%D0g#%Edis+DuV5c+oWMYVl@I6A9K@}63mH@h2*toazPWhPCv`8m*3r!=_7^}v zKoBe|iXs5FqMjT@xD)8u0D|44v#;Xb0>AV6?GL|%^gU}GXrkjE<|>iaOB9L`TnT{+ z1EIsMAc28%?U-{I+7e(4w7LlbOQ;3VCFgjjuUpXv`1b9Vf&gFwe^hU7uXaP?&-lj` zFklbQKt;U>=UV`v$Y2m>6_!vPo*jQd@_%RT3Q$adIprX@gb^t!Tw3+z?Gyl1!F&LZ zoLT4&L2pb^0&!(pVMf%awqZRAt%uU=7M zdUALCf_HW=PpCaRav0=*15d;@dSsy9?|6?+f$e>OdKLE8>IXl_k4-{AKLDF#>X!jr zt7zf)x9-;locO1yzA_-{5i~uZ1``t8-6;|u%yG&P_D5zf?f?90#W zTU|vF=@9A;4G;wk8Z06Rs6*6%CGhRO`??kKR5;p=`4p!Ejp+lBd|$l0=YOBCf7xN( zebXZI_4%zX^y@UD!|#2uuTKjW60AMJeE!J3@l*b=|M{u;`$P8eV=K7~7xE6W_l14$ z2gr|=NM7BIw|ehr*OL~+cA|xR{0Zd?{UJ3Fg`GI@8zrcjvl67px2ct+1)w2;MtmNN zca#zF?z4GHsHbo1H@!_FxU{RqUFN~O*$mb`-S2{_&M`z!+Yv%VL;~e&&@4q=y(8Ds zD*Cg(YB{jk&+@Z?0l?-lRxeTeTrPV6py1NaA0s3r0D;5Z8h$hlNI+9)r_j$eA{by_ zZuNa2zyBuy7|1{O$1o5`pqBr{uNmBF(5TkyM>zJf=coP$5)b|*cyMf_*U$h#=xS0J zOKz8%!4s0^2_|CeVl9q~#G|WB=(eX2znJ9fL4f0QJ$;}w;n-GA#`~$x4lVC;520we3iY19N!7i7I_*CH@+HelQBkEjurWuQkC;)EX<`UM7%XXUc8L z#jjLLa6Sf8=8lbfr}Yq$bk&VsKIYWZphL?PGD!pTgI*8Wx!_{5AE%y&qv&jT zZ&tyYsXpvyfaV>73BmR|+a;&r)qBSNiinWw<-A9TKg)&2=d?+iWv7^wSC}qk>m?nF zdC<~Qw;uj=yr&T+pFvTxq}w_yQq%h8S$1=>PVAwolC9j~`xN@t*@krmOFfgd7s~$X z(?khqA6AN2yv;ga({< z(oi2-q8%!v)ty1DBb4=HjNB1-X6#9~$C<+RCdH;8?JN~^Rj;hqo=Fgl zvKNyePS48k8&f`zo}U>(*$IfI{p@T*an78Wj7=>k%g_l7=m1;aU4m>HRjpVhf2?#k zM;@YRp?j$^P+ltKFVyuWqu3KJU-cB143#Wj>vUaXXcXA` z4|EO@|8`gKi$}_)l#c&3+7-^^joDZzbTV4@~^ zf56b?VmMi8`%hRou&Y=aX&a_3VA^bwt-J@uJ3ls7cTEPaxmAnEY_M#j?5oyJ_Jr$6 z2G-p45bnUytQRHu_?2fE(Jy#SYXw?FzwpB?gBK&a<~~nzCi{YpB$KTP00t9Juu2f7 z>)dX$*%rd@AE{$9Ep4o>-;!DAG0r<;E_oD(cIq#RwSP2!WYq}`ENBfS9ilBt5{%j{ z@5xW0U7mGOQ`JH~T;Qi-wx&QpBeOGsKcXp(~Y)&5l@zGF;8`yn*3 zd}+*Q*ko$9=#J;;*HqZ1UTarfsAo3*>E8M!8AlS@1`LPS?OA(-^;B>(ji!hiXy8av zj+^Jqn>s#o9Aq3Y94H{B*gv*kMlKM?VCDJmdb9VB3!Po|bk%qC;DsT-_;0U^4IivM zPZyfeAX{{x1>K()Wx7^T#AA|p!;XsjX*U2&Lf2oN)z}#=n~!kfpA7t&5bj}zW%*vxM~yRpHPe- zoA)Yai>pB;!$i)4TANOi0%GxUOX1`#l|TNYmXAvP8<`IrSGJT}&4IPwFxLDpgdR^3G71`#-<^e&_M5W@C-R&P1rf1j zTMvwUDfmenAk`i5(Q5UOI%rdqx8>B;)R25bVn4&KsDVtwA?(EpCi&vS;T(d%+*Yls zyD6TlJaLq6H`VDmwRP%2hXy=V|9S#x9f^s8t%z;S=`2Vd3%2jPyI$R;J19A>BJoC4 z;c=sVZTB6@A*3 z4SlVn9m{lGzz@W^V?@|qt32O#sa(n0NWH9=EFgKRk?k}U*~}p`o%sD*727z7T^S0= z6idmwvb;Sa<=8DNTD!PDxlW&dx2~n7gW<@2nR!^e+#{zhLfPQRH-FvElIyCgloL5q zx!gk95qEVSY=VmL-0C~8{5urn0)phXRYwG+uMcnF9_2R|BcJ1m_;U+tEe?Qj|@GK^AG+WUsk1WN~d`a(iaq)t?w8)dD>Z| zzM6HWVK!V?HDscmsRol7;elxcaw`$Ru^WYLMSVOxD4Bcj&Pr2cNU-pvT4SwyRuV@@ zlRC+iRLnFp5q`;V zj9p4%1<_||kf!kO-EPWB{oRwLg{Qt5sHq`-@|Y9Os86%#TR6BvYaAP zadu7{%D2Ulbx6Be0XImTp6N|spJTAAiEik^)jDK(e%Q{^L$4zFR5W?y-Fk1@HUqmR z{0y~KkJj2R+v0R#I4#%gWkK;VR~K2AWWXj*g}34(5Iw^Fy;_D=ke}tJ3wt(wMW;r@6e}4?w!*Td|+m9I!cL6N)o)!L`s{H@1-yL!2;i zO^c77CA^iYYZ{Nm@SP9LpvB|%V4;}KGIMLR_1;UkRpcayvFlcJpW~ma%KR;Og19Mp z=Hy^(iQ;2B1XO*udnV>#3XPtJ*iU0b9@y!1sYpSA(jIC|SD>;u-?fy4PFT2`6fVgWYO8JTKna`BQj4_Kc8{*tM#p z39>|<;c%{2>*s#>l?*#M)@`-cWK4QJNzg`{sntj6n#!&ubVr;P)7Y|;Pn+YFkuCAl z++gFL@iumZhiqoGV$7btEQ@GbJvUh>WfW6%f{qY3p~t&*wKscfW;_#R8guKzF`61o zK#GnLBumymAo?Ba##~QJ^Cu-Ggy{HuVvOPaMM86vIxkdnSDy6?$9kZm`+6?wRulNv z@7^1D!5tRo)A-_U;Lg1&o<^=Cr2w`y@X&4u)~-;_EK+&*HK@8%CXzKZepKt1gw!_> z-N?Grq~0PtE2T~2UtG37X>_|wJQCLkx<57!eYqLMNpax0)k(KoA1*v{3U%i(Sza-Z zh*m{eZh=HTWr^m@tdolsO92&t?n(Zw;~vv8_WIMYaJ|t*S7B$alG0hza8y-`@x^%q z_6W6WS?{`$T$~lOJhAc!yAAvF^*-*!+m+<6lI}Hp1=hqPmq#2gqAHR-#yAUd&7vF&K?z3YZBpmlIA-KCmX z$#ut!<1m)4$(owh!PNCvnNSfAUV%oUkp}>|Ze!zyG&#D==`3^D9FeV2k~h?UV#pAf znG+5xJLs2=Ej{W}b8b>7Z{8LD?rLYjPja0HSYTd6J#<(RSn$yW1C^;YE0!418?RrbE8$WXm!yT$6kWHY%|Ld z2mX^MYqDR`KG7^{H;Y*XPWK)Q%gq>YA*2N{l^|?j@6rlQ%+qeATZ_+pNT-BnuA;2k zfPM{cRqqh9N=hWjrs#WG4)gMc$F({rZYTtlLAN>c1jxOAb(FL5C^VJH(H%@HO+UQH z!#!Gkb>rSp`yMwJczC@y1wAX5T$8+o(|X})GYgt&t+Dd5d5Pz{q1B}ivg+K25y*Q( zP0l7RDPBm}kO<}C)Lj?9#vSuVc-_9xNV!>|^K@q>VMT0^!fq6tk?ItfzM@_2${~_c zaSfRz+HW{n@EAeUAe;_c5?jM#A}Eftrwt26sb?=0*qKh04=xXS+N~a5El^O+j3ak7 zsCtv1KIa@#n2T|?*ZGj6x^^KmMQnbv3w0`KK04oB9<6r%dC7}(aDyZo4ns#r#n|Cz z%{oFpe=hd@xPb-N%EMJ7_Jnnf>RcaFzO@$rYlQ(myB{^nIuqI4FCC?!LI(*S8vJhZ zDreJjeblz`H>@PZ|r^89yI@sIZW8Z$KrmpC7;sCqCSW@*JF>IoTv}g*q;oH=sn*6 zw+)-nHK?p@x7+7aeX=TH{`riT5NKpuDU3!vEhLwvS&hp3-^jlrHV3_i{Feb}RAe;* z(imNX@u_=flNNO)1%8Fdw8^g=0CL7Q9Gmf#9OyK5y~@ksa_M)wGzJ2Wl!ila_8c8L zvhMn26Firtw0oj;@}?b#oB)WO-5*OCw&86>Wu{R5Kg7q z|FTXggT{wFT`IX*KG15o8>E&hcD~bT$j!A&m0p_&#~iJD9x73?;ALsVA>i8@73>0u zILQz=C!O4HZ?^&oGGBiW_{u}8hx_6xJ;JrCc#h?cX#cFVFbvk_g@?mo9tSsbv`0IK zFj#AeVxNcQAcQEsUr;$#k)j7$r=L#rLjf9$K-q!UIgG=S8C2Fg|? zOXtWk0#oNhQLJ_{F+87v;75tgS61@23&Gjm0x2Py3$|REfMTu^0Ob{qQxdVOd4uB3 zz;z0ui+x}a7?;bEVYhx+i@Og^_0NBxFZk!}Hm&U=*}kVHE)z6VQ2f20ZgYH0f%9B~ z)3jxlxHkLfnWbEnxc zX4#MyOT27ZIkxz22im)&-ix32o&r>-sps3HNy|gbXE%uQ3&{Pz&3+zM?V=QjLta4H z1e1N+8#1i-I;-0Bx%!?rblxu$&pjtQ(r*}f#FD|bt)~pBwTN4F4MM#bOxHiYhQP)rGe|5z;5UB+eLLo7?giQ0R#@fyaEs4*x8SFX}X48ystVBSd?n+ zVy^tw*4b(8O}{n7wrT```pwd0`2aGnv}X#_5Rwh!&F_ypW8=`U~%e@VNixZj5buk}1 zqjW2`$$2=es>&*-Zyo650;mq_Bk9VO5ifx$s=|kqP8w#99UHqb_p?@nBebN~U9~rC zGBE{bZ+=OA4g;&@)Lk6YU-zqX(u)s$CGac*vEfwiWRy07Lz{5mYvxezTR)+>qunMJ zF6!PC4}kO^UJlxylOxh3CG1{#T+P6V(<>Sax~AUmAWK)UJ|zq0HF@1W%-tjSiK7(Z zItr+GP~6K-n6|jBp!KmH%C4cLmepU(ip~@{}VhtuKDJvF+ZC@((OlCNz@QbH|2VJu_Bjsdz^J?n`^1!WNTtLwJJ1AHbIwt2uZ5 zck6?o{;Gr~cPPb073w#S{y~Ik+Gk%+Qfvzr8@{ET@~`gzMy%e?2fHoh`fIQ_*$JH| zgi)6qSyT9LsyFG%t!&AkLW-z&O|#8%g0z6l^&%JeEPA!lxN)xu$I&34ma)lVeZ&$L z<3?i|v_~1%!_Tq9Zzwp;WTt?V&yEMv0$h)*=O9jm5dvqjv5Jc-ANzT=>&5g8`D9uO zbHvR4@d$rstns1LKh_*DRo(o{fUTm_2_Rlct*H5p=^~+sWL%daqpultKhZa%uQ*BS z5J1$qRMNr4G%VMg{)=V3DzS33cmFI%PFJN?(S(zjD-cHt2525DjnJ0J1XCH1A=WhN zbl#w%yw)U0nV;dHur==l|5PD?_AWBW!FFkw!-^x~8q3H~yJ=Ppof*1IGrS+7+j8hu z_c^~RV6y!&W#t@7uNV#s3+$9AF5Wpn_VQ;Q<+)vBn<00KO}S8;Y_0^C1k`hEn(fuRg-Z8o5~+)B8ihNTlg{f7DKKgWp;!*U1%xFq+6JFpq{w zAskJS16vLyHHD^oSw@5Lv#eBSG5W7WwK3snW0W0)-hTr!q)gR z2&sU!4cX5rbYAcpM2($DLN!eS@4!Pp%d=265`H@i;Tk+SNm1a1sjSA0W;c?SaTW#l z=)!bfYaS!7Yo%rD%~T58+K)W5!%4-`@#DWB95bJYi+GDxo&dKhjw42qD?r;*FzY+~ zZ!<@x;kGP?djUJ4w=dQ6SvzNg7n6m_e=n{QWn=Qo{&f`))U3ju(Pp+~R=J50k^?xy zy|a}OritPh3Nos7yZ>Tku`o;5X5;Eb@3k^mEGwJ~wE1U449ffR%*3x2Of1(nPnX`Q z&au0d2>*6JmkKU5yCX?q)Uyp`G`f^1V%|?lU3nz&)N#G=K~l=*{FMl8KJ9fVBuy!SAoP@npNIMbij5UWMmAA00* zTLO!*!*p1oXkPT5AT>U-0ApQ2O1CL%FpHsJ5YE*A8~5R=eD-J+n++sfD9SwB3(C7| zwxV7N1@PRYE_g*}o>$z(HS*CLsTK^zeG#|ODfh6~7B=olX?}GiY&m0#?ZiQduP0f| zyG%1*E=j)i8~Gyv=~VT^#YT`vjb3Lt_re&$(!1QrZK=(s3kB~v!7*0&Xt)3p%l`=! zlHDikCwPiK!IhXqs?fI0$?_N;{%P6w?{?$9Y>37r@#@;ZbjwqD48t3?Bp?&RlBZID zRbiyaXE*+2^AYLQ_KeN>2{!TG84bXst|c_7X{pX!#*=ZxnE)`RkwWComY=~|S5+=n z=smhmX&r8l+E;*%wmy(N7onsqjw@H~t68G{aOPZC?;J~boJ^Vp_; zEJ_d4B)3F?NKfWhu4vRwp|CBQpt+LgF)FpH@nYmTtAiG752ovK&b0xXbr ze5)!IlEEq@Z6)PqtHr`LUg-`7iQ46k1)H@VLdCy*MJPZM4*ly%P^hpUN%bQx_f4yB z8oLYXERsgnMY=QxQ3uq*7xFv0@(Xj00G*4hv{?W3(D;XaxE5l1TaW*g+l4L5QLwSH zp8If5iDErg=PSyrQz!|8W0l_lnu{*@EcoE;HKa?V*FB{?O}kvi3+TV7wKm3B7E=Wu z*6&kM)fl~y=nre()D|(_oPtQI($ZLT?Ocr7YwnlB`KU9z7~hs8O0wXZ^Mj|Cu4GrZ zc9e^o*k^CXvv9zxT5&j(d{&j#Y3%ws#A&QHobYf1lA*OZuvex#)7ZS>b>OF)qz2xo zb|;6{dmq}9l^Q68csuQYof%nYCY4zs_zNUBjBOH%C*uIlrJT%CVbF-AyZOyNPI(`L zM~jSA-`jL4;1|o1_MqFD+=58x$f%OpQ1+?&@rvfNu;m7T%Y5l2J{@E_znD#d_g|0; zk@_?O+4=pv_2qXDeVgowrmx5vpS zy>di*u#lw#p)uu<)GqkjyZiMaaz#^lgO?|a!SF}~4vnLzq|)oX3?zGgNF-?-b1Z@^(qIw4%TLafk#D%A*mSUH-AOkv(ah=O3F+!S#j{P zrMfj1_&5FcbLi3xu{q;{-2`!|F;a!%*Dpa_P?12|98MfbGwuKmi8DwkmM`zz^VW8vj~$HY`r zWpxwb3CXMT$>A9n^Fbslt17>J)mfS#?%VP~D9Zz$`X2pZ{kkTjs-UVYX(Cuyero^p z!`ibsIktY}yy$OH3Qhs}l?Uid=ipxbNCD+Hf&$}tU}9}=ZDG>tWCy^)%b~-@w))*G zH#Iwgw{vl>0}1BInYEhvv5k)Gpf1Vfm#pvp@nwJ#oB<3^?W|w5B?S-dO7SSr2+0w0 z-SrI%#sK}qT7I^1^{-D|`7LN*bNwFElogXg{ExdEB6A&QJR+<#x6c`h%5 z>UbA)8mN$7POi^s{8!@mX_xi4fc^leQf78#a%}v1m&VNG_TcJ0Zs*HxogDsUMr}s$ zw2vh7&s5F7Y2PNdpB^){7r%+45VT^lVb$fcW5FPEd4^k{i&#F8IlQ;D=UewdBS&5! zdUALm77iSLDAwYGm(oIbS69N7nY+GC3Nx?^l(~V4pnH8~8{^wOv)gC8ijI_?iuc>v ztWHkm%1vz_N2Vn7PXXBJdlOcYb1NtsHJZ_P=!C zOz&|Qzo~f_b+S@$8wI8CFRl1M$}CRq;$RSj7MSHfzw5@o0J41}>nZ~ibMfasxbL=< zgq)O`vEPmFNzuhC4%N!vu%?uz;FPY6ibS-z;fLGWAG$Pf?0`_cwVlBg*k)qCh$X*U z679MEKzO{WU2}B*+>Dq*eeOBh%(b;FE*=ohA6FKT=RJLIp_t!k^!{?v>f*v8vCle( z_esgEo#BBczz%T#;LIOJM@Qxde{R=L+D{w@K#Y-JLfJs#^8(>8{gf~I!xNwe$!|gFf1(rp5NZ4sKM;UT zE07M5ne2r%(e7X*kZHPinsC*AZD zROoT_BV-Am77%6%vT}G!f;G7`_(iyxB5aKHTOu@;Ael4#eii?%!)#uOHIq|=|2;!u zDKC}(_0P}zQDkB%2Raq_A^e$X6`x7{WdsWOO7h1`#;`ayd673FBuM*j|LNNQ=5lZS z2%;-Y-~89K$M+lf-42uLW61BDQ7;w1dHl6K>8q!nn;00CSp=Xy)f^F!KZpxb(75T6 z&+U)HCI;#KzRUmmA;O#Heu z0M8-(9q6;8^v}rX%7FTM2n>R+E`xLYxVeJ8`_*XI zRdNdS5hU|#k?H~OsM!6<2L+;h#Q7CQ7gTPC7NrH{(Of>xc+C2LAMlfx>@| zCjAG7Yr+El#EF?@dZVnrb`jP@pyhG+QI2gc{Gtsq-37X231=y-)yIMzXj4oaM%?R# zT?_@j7Gv6(RuCY)dK>VN9BV(k914D7YEq+X~T48EPD=bgU34$Oh_^R5{xSsK)KMg0Q>}Q^Hj=tU`j=i z94DJ5Cg5j@1qWXH#blfAHHARo+sYCv(ici?+!|MuZp51NxZ5!L1NcI_mgrp5Fxhoh z)YIa#v!;Qf|Fa?N>EP>qRjh;Ga$&6=z(Y{o%BE~?s%E_pifLd7g`}#Z%n)U@qW*EQ zlD%|b*rLwOywu{2dL0ey^9>XKY%24b{YS|wV7B7T~JcQjt?$={FeQ7 zc+8Xgz$4iMUnU;fKZwUOiVq#L4M(QbTiZGSy#5a+4G{%;>`iEPb~jH04$L}A7XmK~ zk7CxL}_+}@jd5I;Dxe77; z%*e<=*EffZlqVCaYyU5N7|Qh1a3g-_ODN&Ig{-Wj{LW$e<1~w(RZ&!s6)GBXpaToN zm9p|!Kp`;AS4r!l{oh)Fpp>6U2e7$(al=vK_{x!(cd(~o zQB1_WtQ~tVQG$)(X^KG!UptVvS`=9KP-h9C$04?bf80?`=<=I)rpP|=zA^SB?45h7 zea)ExQoWD{6EqU%O=&^7{%~otLC>@Bz(xgD#Xl~2G-_)aro-Zp|BlQg%N!Q9ozz*dIBChn<4mP+RWfC%P8==XOjH$`fN?+ zp?s9Okf4pw&L^0lqHG70fiQ94KVF!q=(Wa)LY`NnS{#sPwgtGswcRsCt9rW8_(PI9 zly|g&(QjT{0mF5SAcSn`2zf8Zz3Ro zDGuq06JR#)N{3r?0c>b;57JHL?U=>NQXgKn1QJzO8|7}N>C63`_MsvJ#%O9d2^f=r zCh{hqNcukI9seo+1zexG!>p5bFR(ABv8LU2%K8Ymy%go321he3)9pg5JjaWGh~R}C z_n8Z}<+uq=fihjV86VVvkhD?(w628%+((9==@*;C7nd-Aj$CHHN*q#ES(W#vMox^9 zWZ}7}P-itIv%79{jz>)fDV+@sqVZxjhn^{J1nwq_b7kSJqFF{fD%hFKyFKrE1I6v0 zh-bbJ$J87ZW-1x_n4UTk1uu@}mv%=a^ZlAPFKFh*5kB}i?be6n1DXvPG z1z|`%NF94{C)Dx#-FSl0;^8szjLDms*+k{}R`o>f3yU;=gmm(n-#`}Gdru~pm$iI# znKBIYHeBe_C1k~f;7mGy%I5_vr*>j0iunln;W0mgS|{f?F)Ew9DC}paGOx0GURhAuS|%g?ia%k# z&;s+L7~eiZJ?+v1b}%;}qq=VYA>wS& zD5Zvkaj6AZ=jLDVNtUIgi6jIDu4n6Q0=;nH^!kTDVeI8z5^qxA1`rrVjiHKPmiZ*z zOPd$=0{k+FvUt25!qdsrLp6wP8s9}Z2w2`m-30a5L7e0{@XMVSfw|(6ZBI2(+U@lf zTjdF4G0734$Ic1XUEtps$yIx7s!Tpnh#5EJTg3m&G)nTRD~%~lQFafx%2!>thh`61 zt{bSH5~_I!N%S96uv&AmSE0684kCAccTk7u!Gb;vws;ZvjR_TS8Yv)!nVh8f3?p(A zZBTShWxA9j5Y4?f>Id2j7w>t{u1H>4XUbCXp;np4GouOI$Zti*c2#yhQXYON_sst- z&JPFm6~xTlU(vI2g}!thf=t)#s7d|EU~Wh%vMizolAr`TGuwJLRdbpJ_B-A)jzuB- zY%eSA)8M!hn-eLNm^|n-pp}Uc-^mK~Qs?>r%68NECz<1QG7(*te=ItN{?Ezf&Ur*c z;n{4s&s~q=5rYWR>+D<@;XA9+KbJw@k3d+#hwUgJFH04{6thZZixKH9IK;P2ChZ6T zjC$6j=ULK&;^<{m$im}>WqwUpVim|YMR$E5xcXJCT>uq7rZdEdEv=$$8eh)7@!jC9 z;Ri409I)NriK+L?hxWZ0*;%6NJ~xy}x*xcdHW*KZG%pw72B6t^3@xKIh69&a2w3}j zXYWzK?h0IGHhsJ=Gi>;;8+$fa^7fzxvdfbzn(8XD#d@5ulc8#GKn@<3Z(raw>1E^g za}jUi`YQc}U7SsWUslFj??WlbI%Q-KJ-D-- zKTky-O72wikKYKRz4ZDQ#fDDS6;6g43SEZz-MqU~$bPrHU=8llt zId8?m0-Y69y?QRxty(O-DP^{;y?vti~*WKkrsM{~Yjh84#uF#bX+ zL!2!do)`lLDjL9gM3T97j)4s+;Z`9s&>=QT!q}IJG^YQhxo_Fw>8u~-I)LjUVgR=& zY+_z6srpde7e#*Q%s%j1d#(clCA?TbB>kIY%Po zHvF+jFKt%jPr#;Qan`6?46}nCKMH|ayLrm;5JMEgWbJvU%l}P`aytJ8rqjlTC&i&p zIoTbYpVyfp+pY(TvgspfdvVWDGgbrThLT|M&s}v9h8Q9jv%tLa;L~mNuoa6;`%vLi++Ir(r@9V~+HhnJG)m?_aw z33brpphV+Kn%L*D;%PGR|Z4~f( zS5+0Z&^t>F5(gVzSUe@M2NOk{s!(J#?<9|AyV%)^;~*ozxTPK+d;S=G?pZm3 z`lEO*tr7DrIt`(c9+DS2j-D?OOmDX-CTA;Y*Z&9mz@co%w^a=Qg256nx8j^rX9g|> zdAMj>_)2Z6*(*Ro`$P2hqS>HMnj!MmhLU(5kai(YoD!eAX+fnPSa}QCYpZ8>9m|4* zg&TT30JMQPgbwjtx!-NWiC)~&pkH07i;xRE0YXNn3z7j9F^HNx`ew#Iy5Ea z^Oq5McU^<)izVfS(tH5$3(<*dPW<5on7%H3{cEP$Njtl+Hjd4XqUmEUB&jixd%SCE zOb}1M(Sc#aSTM@DUGw)1PWR7Vo_7e+R1TMBz1qFp)$P5WPRN$BA?V^v9%@!f#!gWc zeoyXfRgmr9D~Xo(=mr1kfIuZq*N!2w+QD|4=7jW|ABmsDF(S`O5*%snT=epWxcB{) zJmf0qllc2pns0+4tpRIq5+s-u4h){pJgBnH#FHA7Y3vi~1=eJzjVa@Y)!t@TY`}=> z0B$FgcW1nqa|y>hvWV{tbrXsfKX9q|65c%STt=AXEu?`_F#Rx=>fNe;Ue8BOoz3(s z%kZZ#5SgYguNt3A2WunAPkFDjpr(px9JB^!ZCRUSi?i5D%5wL(kr*D{z5A>e2ci&1 z41aTN5;u%pl{?EZcrp5L_eW#G`vp7w-n$gfMup|s$>=w&JwbUFs60h*VQ5d6Zf0j<^4ddKGN-0 zjns;r(;~IX*7J-vNR!Jh2O>wAG)=QROvD?7oFcg$>7XL!TE6Py2hBE@RvA$chHseb zR;rn(X5=wwn~^6C8vRp2|C<<4TH4Q#eK5?HsV3v-mP3|88YsjkwhMw1Ahq%GkJIjz zg(U!7ByX3PvY2dai)i&-LJKKkfgy!4!FxWvro{3T}(^g7w;c*^4@%Xi__Ff_lt@z3V$aU0_RQ_TaNS ze4B=1XUOZvfm5c5O>UZQXYcgPCAmIgZoB+T7pA^>bA{ezOR)}jQ=QY-AiS>jEqKIU zVzz8TSU;k88KbuDD{vZX#`LeDpWa@3VYI*3_q|q|7g4?jcs8t-lTdLkl+(nB?0zHY zL#HRQWba&+r?y05`-AfNv9KfeBG9IMtx`>pzW*{hTwP!LcNGeuURQo24oCJ4T8^nx z#uoFB4-T1`nf$O_<9D}nKf)tME~jvxgWSRVaXfr4>L*tR>`hAhTF)NM$LVWMd=LF; zo^aX}9oXq}T~|4o^9NNoC+i(|G|Gsbn3`p}bd>11kGpCwnE^4)_@uPv4h6yM>bW)u zczz2yWXqW4O!2BhqYC8?aL2M2P3MfVt*V4;=Pqhfa>KvIfK&IDx>P$-QItWBYtm{> zLlP07T4MCbSqbUQz}xtGOi!(Ot3oBBk2t6|l?|j!QuI>^C^Wx-qH{S}rC($@T|x03 z#Pg(^1W#Iu798@G zdFbVFsk?OM&}g)q=~S2yVsoD6!iiT?q<9_|n3QfhEGXFp1<2`2$a%f<- z`ixHvrmPWD0&fP7AKj10*QKe+-9P*g_r`CwM6`2e_F*l`PDOhoDDO*%?(5@~g4ijA z2E5^+_~!*@D{_NwR4dO55yu;$@(vx$L=3q(2;at;(TrH6s?vq>uvcO)FTsVd^XZ_kvhykJ(H7%3t>7-I&V!dl2diN zmr!BPw{0qQYFSGk+*}X9pwtZp!)k~BqI}CAxCD$Yz!UP2uhvL($-)-+Zr9}uz)pM? zt(DoIeVuO1#|q4#5Jpl?_*54QZHp!)ul=?E<&arhB=3!+nqlT^ZRhFw(_IMOZ3kTW zq;@n0{2Q;`!2sMwMs-o6A1och{XPW*-Qob|1IOQRm2-ez%adf6IXl8kG_<+Z{!Xf9 zMVqWj;Z|^%WZxVmI*K0aCZ82I$}M5!A$Bvm<-FfBdl3Co|c4Q^Cz{G_{fOUktq`bO`~4d=37gx)=36lgm-`Px5pn#)+U zPGYR+*%cwo=4`;_N>egDWWTr@9Ww6ij1^$Vi5)Ym;|4D`g3;AmeFcc}M&6$SoRjO| z#iftq;J+5WP{0ESjzHZKZ61s1m_o#G?Vxr2K~8#~(fDn3Yb%MQpK;1duWY`fMU(f- z#SVaP1tLMNpZaLQgh)qVVkU&?c5SWbKYu=i0=HK?!aUCf`Z-4Uvb1`(Ai~$iQBl_4 zkqz7Q>X<}kwQmQI>6&<1Qdag&UZfiFyDC3XvaNMtlASdtkms_L|Gwx~CuH z3gL8LG>Ip8y|V21%#sK_+~0XXSdGor>LZ9DUj)bdXJ=*6r;DW~QmA-q>4)NhQ*jR( zjnx^9`+~p_N*qogl+cz$RG1!0SB3%T(Qd1xd~wC_l+IIPJY-I5V9oL8&RBzl*2_B^ zBIhnhAosBq-??ToUQe1t1Yz*I#nv&`YALUM`#^T#TpJuo>{#LJwN%!0sAcX zT(md+dwGwU4#FS@2C{T^P(i^k!3TaE*i?NInof4(o`)R`d8R$iI{ds1z#%VW#a|epo7S#Bx@d*?w&qK^>OU&yQV7t@ZsNqJ zmcUu2)L|ev1ef_4%{J7=H@Pqv=V9D>TGylsvxDxy>53j0On3}1skXe zYmdY?yRglXQIwGp=Oo5=fvWOWO3j#>D4sUSZ~IlMMb%v8Z?1eUWf%3mo6dOn5BA+O zG4e2Ae$e>LGgUqRpRdzC{XE*l{Ii&|2(Fztrn$9Ri{<_p18(#4G0zm_lsi@aX&*kzHUH}d+Z2U z^vx>sjZl_HaJYVkxKPo>2KISuOw<6KOu&#BAh`4z(fde;tn?z*JdwGqyElWMyb*=Y}WRI0y@$kyUU1S0aMj;_x0YLeI4qwO>^ z%%D0meptFZ*vS{y1D#1&jvo0z-m8RvpRy-$4I%)awuOQ6OsS`*?OzXxOmy{@>d?|i zMozY9{Ut~4(ds^qv%^pq^&MG(#fJUVjZ>I;sv<~Ry;Xw({*ctOM;os76 zOFWGD5#n;`W2NmWH8Q`7qaqBmHUNZZ+NNhP0{lLBJ>=z4mUOr9)qM);x$HFrVsq~? zpfN(rwfLm!_5xQfPvbO^s24G!f8i{-R=g6;cA}uL}+y zyAsPFB9Sb`;RwWsINpvn$?oS*LgId;TUQz9==f^3dT#4LCLpXecyuLON?N+0N$lC4 zWQ1^~eX@e}W4gI0O3Cb-CKYnuQIE`eOHroia^k+@*$N!Ykp{0uA4_7&6u*ZU*COzn zzjo*{!%+a+hId{B9sdqlUxRqiLvO2oPEA9+hx1n|%NVW>jU}_sw#boGt3=})iJu9t z52a3!7O&3am58ejiPDE?&&!XnAEa z_H7VW_ofe<*>J{?@VyFLjcfTZIvIKPBa%9uO7&LGX*diTUFN@V#(#v0;I?^)LLZCe zPMBF#1{e}RP6NIY^qxAHlNMkh)WfcwW|JQVcE_ZuLE<1S&)SwG7i1@=Yu#W> zy3?!zxj?~me(jceM1X!#G5ba^j2dKUB}fn3E(2swa@2@T98KdWt+(zo3L|i*=|AMR zX;92z-|LkBG?|^V`-0og=hEkBM_YknR_xCIrfr~HXH2*-6t%PsP1KKDPSDpJGyC5M z=KtPh#HycI*Bm4@JcbwnEjzKr5bL6bPRhKcHGP#u&M}(t2>_GPK4e>Wc&&>4ORf9h zk%47GrHjKtT6>~`ZEG0Tw-y_II`zQURJjUy95m_(Ye~HPh1~u*?AiZ zT}}A%dU4gaV%|of@)|e&$Habn&jE@3=3(PNkreCDjns`fRRqZ38g|W#pmph$0+fCN zo}Uiwl992s;3fi2D-NYus%;|jyZE>l%8^QI>H3+c0QP2Pp5sQ|N0!(6W}m3bKCfhF zvKhAh=h@ZB_{qOhP*J$@^kXU*yS7W)qpJEk>~4p+CkM4;1+K=Lpwkn#`T>vISW7tC zs*muJ(yP9C-Zy>RgrV{NdG*Kk@%CE>I~E5sl|_Ay+=djbM(Ey$Zjh+fDtKWpA-3yZ zA_kw=-Q&gHRq}YqP42?cash*bV)q@1%jH%i{JWffE%}Ur=lr;`#yUo=^FV`+Gc9BF zVe*R&z6nE!uTrI|(e>`SWk{N1L1(W(Uc7Ab zr|477>mwbq?#o^Z!0qvAgH^}U%0I;6YHbUFcMJWEW3i~M2mDSO4R-5Y>f8H2b31{<2<@t#jRXezayu>+wIfT)$fQfWt|NcE2xDKa zbB|6oh|FN5)uYRxb3kmEq&+JVghI&m=)mV*)}nb{GGGiO+jK`qm|&fK3bA(EXPFpc|Yt5OI9) z4GxdmD0uMSxU#DZwtAo;FZ0+n^|WGIE|WY!aANT_@KE5(UNLxZ@G3k;Zfi4jCr2O4 zOV!r!R2BseyJ3(rTBN$SruW=;Ba}3`cF}wY>1@OioV|peO<5LZ6{8=Y!02QmrbX|GJftkiwOOBxKZmnQqZm>yry2hdu^%3%oRWMLP4X09oNuz4V-D(G`~SIN&*EVK=CsRpqhOV+hK z^Cn&)Km4aDmksatyHcW6keF5{lv)a_U-2Phl7f4JnI=|)Tb@)?kuTw;zbkI)z4uw{ zvng^Ws1m;4M^_N4?MySmSZ;ZqU z@H7sbmD;0Zjb1&S_k@KoRI@VEfhs(Ak2=O7zQVW(2Cg^pBI&&su$L@Z4R&e0@?#eJ zzHkfo#IlCIXqRTk7S}ZklJI5$BUfHBbb7>Bw#98gb&R2Xu^?p;_?DAOTH~1ArVnmI zZ%FvVE4ke3hC8ndy<1=SnMV5VOZOGfP9_B#wzDH%jGzZK*;2tX7E4kt$5W3viUD(I zx+r>f{yhyN_qs{xyjRauVb`f&(e#O+?8aot4`krT{83)u4dX0{H%sOQ&cn>(*6`^@a$GS5Q zR?aI99)^EZXBpUAVqzzb+3HlZr-_A!yr%EXWzcX zmQDHb+-7YUXAsCb^3q1w&pb;@TMzS%TA0PS<;e7&wsrVWF{6el?hg|8_0tM%atF_6 z8#lZ(P>qv1Ey|L)j9F)`p*T{wPbOB)Oc>%DV(478eAq7 zswOq#DoEZK)+e4&XT|eiTo}LsfKFV&9T5<2kqeVJ%qX90=w8PxZ#TVMbo{GBsc7H5 zN!f7VNSMJ%mgLtje~PpJ?hq4~xgIOy#zHU=vXx5 zom0{vBFl{KkR&Ac!|WsY^$N6b6bbXQpqkJ&M)Lrdn6lHTeT4(pVmGHp= z)Z`*D>V9`GJL1Rd(E9+@sKpTK_}vupg$3WBg7$fBK( zzuSG|6ZT}^khb%5!3iUY9HK$*q5V#mNvcGKa}KBWl0u!H=>7bCjPm=5O6lB62?=Bt||#E9`>Fb(_hIm9NUmrw`f(bRI?Cyh2nketpFdca(kRB z6O7rThz=Z~XZKj_A!z?=psLj5qw&DIaRpSFUw0j`^v@w(pw8ORj~XGDoW+zWP#c^X z;MyKTOAV}wq=l%Cili47M!yi*XOIRgBGK=TT%$vGQ(8N}v;!5(uKI_HyCWnX0hh)p zmB2X<-e@h?Dcuw;hVQsIybj$hv+_->SOh~|B!hBQN{W!$wRxE8bZ%{l-jbT+2>dK` zBZHM@Js$9w+)(8QoXS0kLlrKp6 zl?8HgyaBb+Q!KhVfm?yLo8wyoNK%bF{|iGvyuVSjJdhevG_=u~>TMHaGNNd5H&yQr zRRAsiJR+XflD6X{(a2aVn@;LZ&dMiPxhS|am!(YW@FU|b zyoud9LU3aP-ius80_032s7*Ps89r!l%tv{BFv}=xI=|w?wb3O?b3iq+tuMrDVMR+E zcGq68W}y{2H9;vH*1>`hssR^{J?|3EAT*6(UknR2C#iZwv`cryt1MFl4_1E%N99w&$nw; zspjARI;au+nv{ay%*u@D$UA)$n6ssC(W|XZ@Hm|YHr(q_xHkwp+`ASZx9tj>Nb`2OdvtGCWVmM> zlaFuL=+5ntB$s3s4XB#Ws=aMc_r^W6_;s&2^teS9P6Ibd7xWA|cKeYXcmvs*uleDI zJej1k<(@7FBg7>Iw% zIPJZPY%6iBzG_yrbXn6Wh;j{gXdhuzlkjTDb@z_GOv={ia50Tfew?$`W?Z(Nvf+bG zOh*WXt!Ws8n6vHUGbnJHeMJv{svFYZbh^E(B4nyre)(l`*%8(QuU#Q@P_%aIc12IT z)6+4}uqriG;vvds}(vKdgTpK2u7hZR>2m{4npt#r3F4yWACg_M{L zP{JlB;23<7i(_rI0663j_8~ztQYiB5IWtd@n!bkonLK7d0+A{88lLXG4}yXKPH^w0 zkDs&J5X?YZ1#&nn!*q!AUDWg2!l(0Qxh|I`tbu%W=~L3xkgq&yD%ziPsd@N)gC$Mx zWy6+}TlBpATHazzmZI_k8Y5K+Ym1HiSWlP*9M?3i&%>ReOBiW8Lo`x&{4=b{87B=s z^z72*mo?ovWgu5S9d==SEBLNgrx}`rJ)-(6I>so?PWI42(W0geT1TUdv25~F$AvhU z!cpkP`cEghmfo)5YtF^{X}^-1JKB%B!PSONXO$~Ljn0F{9(e`Y7mGugvp;L8q!HDu zX&39ewFc~l%YVT^2qVb(Zlwx8oC&p$SN0%AZ0!>iOCO~lIr*P)B3Z!&M(yM0>#MtW z>|6#`0UaQUsWtgLOd|vh)>v?`DkEhpOM--L;bCTY14|Xoya|DO^9uU0>TVvuZ%beN zG%SOBKo2yohEBT?NAyK^Se2ejDZ^t0s5+Kjz?JTa@Mbgv5z=!j{lHYI0l&sz&Jv4H z(SGX z!LI1vKJo64uwh*X}cr?j~CR-5IEGCxqr{i4ugwz$47ZI7rxyq(j?>Z=L4Z2 z0OmAqB8{xGe2ktB9CVRm)INKv$XA1{;Ju8{O?CDM`l#IGz?@J$+h}0bkS&s+EnY;# z-E?D-kbDrkW3U}kj3Q#7;8~>&S9z4^cx_(L>wAX7H0hp5dJ`S0)np;IPGS2P+odAu z13?!J50hw8!yafbD(7bV%nN~vf%)cd;ty0v9_3j`x|dyUo%F_b#P8*=MT&354nz5} zf$wcqd?x8A&I{`y*vY#W!-2%k9m0H{Tt8}Ae-Jb0!dO-SFSLwKHPkkeQikZ{qn}OM z4?KaPeZ`0MRr|JP!l(@^XXvV84~HJ`jbdBwV~m1FpQo7m?$?2!^pFxn(tDb>dcPFN zAdIkHa5|;838Bhqll!`Q+;k)4I>H1kz-Aqt6za^Y*Ohw>)@qtjm`>PIOuyH8&LHaW zSP4E43ja!Jy128b)SFAd$S@KnpMk@70U1I6o@ZKX!hBsq3F9sUc(DmpgY8c&9}ZA= zO|cL|+t(re2I2{vKShMde3(OSxgt6_*NBR-Xv%SSa~4T9Bmam}V51pPrf{hqFUEkP zo-{EQ!`&e2!7PZ$*mO!7+?(U>VdJ|jiOwWRTAA_j+2F8fK;<%zY`ln(tr&4*$b4vcNy!@NgbdHjW1nDiG zB|J8O7x4<hJ>kAe`YUz_l(yHQ2C^Ip>dKcl~EkC0C!!4|&qd37h3d`RJ*6w~2Xao#=-; zX|fwl*yN^XYam`-Fa2i-3JrCcNGSGro|MC#eN`I1Vj<(9ADkZ!p5%!GRd34F#V71T zgW;j7$Qk0*f(H4`VzmY=6<~MQ-;ucaNjh8PWuD5@a%esrg6lg}aY}cy(cHXCk@cNKJhU6FvVwr`%1bx|Qd_1xsT?jKH(kD4ONMuFxI9&Z5LDjAvr?L$y+9Rb0+Y*v(fcc9d>n;%%k&Z`gMk6@v zTJoFOE>psA7-7Ud;|%&e%i(-Wofqt`r%1S7$Q^EU$0^$&SYYvwlyOHNCh_ziBhAv~ zMEeL|99@lWXJFrR4B29>3>PE@%$bl|@%ZCbu&c_=k)7yPkbli-y;&f0bB7k$*U#F4 zaWzO$oml0u+Y#_c<#|8+kZ{pd42I}Bov`2Jda@J7m~1Ua49|0xv>eSI-(Gg|0H!1f z?TN@@BXXmEC5ZX^8?L`rLssi4e~?QiW-Co%aR`gQxW+N%#YC) zt34gL1ruU*;EgP&#&)>g2)BT|CwX68imio3kbhL2i@RAa`+t2q)wFJVT zm_;E&pfPot$A{+iRSjh5KB}ay8@qfF&~E238iEZOEQEA*cr>d}PZ}ZgMHuTF5Yy$6 zlQz)O<)kkqrfUUw#iv2+V6 ziwn?UvO-F^u7Q?pyjHwFks$S%+D$8x?_xS)!X<9H{ugA$XiwO#k89HTmk+ZOEl&a# ziM*~^yF`Ur?DCJk6n=^Q0vEACfl!&qi{K&aecW)a1vMiBmUB}sX~Xv?^tM&R? z-RM(aMzC!=52Fe+?cE*s=kIhzTc4)dvHx_t^!_=7tppd^+h5^<{?a_vB%&$mc8!V5 zn2TN+mY7z+E(o^=C3(O2c;w;5jAhe09@0Fcqmr)75VAEOI?DN;`csamvybw}!IPD` zO2W~cfg_YqdU1|xBJW?z$8F6?2y^`ut~L`M&dR_rNhuxP4^d^!n{#Gz`AgHs{EXoC z3|Dq3KJUQhC_m3^cq7LAys*)`7vy6Q%wRc|;p5PDB}f|`qu9dfRIW-5?n4UJ`zFRC z6JAI{o1yJuv!Llnuj8pEU;ox}oJX$WBZL?uhI|YP2+&i8Jt%!jJrjY`KmG zChDoWJlu$m!7U2sNA6=cX0-B-Z+48Uwg;1o%uUi_1qei%b!d8r=E}3|MBH^9M2>7n zk(KD1B!&^YWPKz^#MiHn1swtN)z7JWeb)-MspKvva4?xl7o6A4cxU5%3U>k(2~iYm zC_&7C4gP%a#YTF<8=H-%aaEQcb1!D2K>I`!YcMp@zVVlForyUTxvxGfML~P1c{1#t z1c$Fv(0maDlkhMOcvqZ75H)8$bqT{gZALh0(GtPS|j5?%z5jwx$l(rI|S z)E5znd0sI^da0H{>e5wd5%#(Kh)P=aKi(CK3xw}A7GQKQU;5L%jZU2!L*BuDuNE!a zoZq`uE5aY!Tv9+-r55J@w3u_a81nw-xKcuB`W_;Vj^2UU<|{t z?|4rMoac$q&yWSjLa8fCQ_1p*?ah{kS&&YZVls1-7)5G)&TRyK9&Ame_srUTo&y+2 z4^8wE%yQ&pvUAsE(=cgE_Zi{1SkV2KR1JRbAvkWifyfB%@rfM#Ei;3@05i0I6Mggs zR<3jYMSGd3ad9`MQ?2GeJ!?Saq{XK4DaSqg@B&1Q`^HixZxTrr*INg8%HSj<*!JqVH)A{M~x z_FIDn*Mypb8y~hl0?k57Z`UmejS>rz_=EJGW}!J2Qmxxj6P<9Un9;k&s`TyiXq#)Z z@Lx#CAHF7{BE0jE3phd(azI<K6Q`?9wLZ`FoyorTnJLpO{nVE!oXfPW+{QFaP_ECjNtZ+SxcN6t~HbOeG z{(@0?n9TYTm$H9ShLg-o>xe;I{o&D#cB8ZEltfVXK8i-Uy{%o!8iQOuO4SZ_WdBTQ z+)Wf9IMuf^6*Gs#`3+%`r56!T#fLYb)xR|;OLunWarEI($SAP_Iv2(UhTk7f2?&{#7v z{Tk@yQu638F1y3h8<0`>=i%g6urm#+%{;#>&40u==t`@$TEsRZpo_|%Uw(Z&&f-#u zk8guKDKP6fj%|_1uEU1St(%2a_}mdO!CR2Cd6{|okdqNwWU#FK5vO;W&#h-VZ+bt=dHS83dQvq zL_DX@h3K_a;ikk(Xt!mFy&PAuwi9L3f(LBpJnNqv-!k~K4}#BaOdS~ozcvJTBcgKD ztOxbC>J4BMxRKPJe)lfBhbvz|5JQl~`eav5Q#5yc5`5$;DqId3zJR8CtzZtxw#VcK zuz|L+{Quha#@M4>%G>i%jEfw(ORdLIh8HXGm5>;x-Y=szW5R zC`kL#;xaerA-*4636$jGq{N{+Csw<3-8)I04qo(#oVTS5j-9}MlG^nAzf6HHlIkK3 zDL?vgMCxsym;W4vW_gcIjwW_W7F6H*rel8bac zxoOvXqn*5xb)H37#^?r1E^W1wLZ6rV2J2w`xX#Tj@N_tl$;AW0HY8IP;YnPbL$_w{ zTC?%U_F7Lt&GJ?5_@KrU&DK0qjAg%oO=#iuu>5RC1}gbSP~Dd%&4P3}=jt70zNVng z{A|rcdP+DJ0S7U9z~6Ngn^Pqn@B=9+v@Um@JBN1E4VM#5Qz39s90jS1j@{6Es`Qt zGo^PPNQ)~Pdmew7p&?OaJ;VZ`kPDnpa?X}ouy972G#qGb`%_5gsi!mS6vq+BDIHs^ z4yM*xgS?FtRnAW6yQf66y5#yOA)l9=JG1yF39)E=eKSZ?+%Z@JskXA1?S&NcnZ+Kc zGJ9Q-Pma*094q{YgzE90-89I=d-+tjwt3QklLwjm2Wg&!yL=rWj)1XO&Upq5pEDKs z)>H*ISxvxb>ag={M{5#m^xHeAe>TtD8;;H9Ys-mQeA2f>i79I&Ho+L*ay3Hu3%h04 zfHkDJ!LTgxI1<|UBCdvL`O^2!MZt*W(1JUKHFu+ej3AK9+Ai9Rj>?CxLPAI8(B;q% z?ri9k$zI0DWjM3|d_{(h%*)@wpx&S`%YnZh)=u6vCTjRqBFcHjqUSP*uS)c zN3tu;x_aK9=VLOsD;t_Py+rDQ5%1aatL{sSeB?fjI7o~sA@ z;|ux22Bg7tW36cO_z3Q>$=wmRxVLhb-(yYMx+u1OR(Z_+EfihhR7r#`-|Fk2EQt|s z|D=4;RjJ-m5$(G*C~T9?&Me+31WY0MW_!cMk{5(yZo1UK?;JCeK}xd%fhN?VW!b+! zt26{Tnfk^Sw!yBn;@D0{TJGT2qUpJXpVH^`hYC7)cJH;h3eOauWwtR_N$LDu%eu)% zc|#(w+OXrOHnX~DEr~ugx#uhF`tNC>S9G3<$pxIaF;H?QunW{Txz;4w0i_ftRZyWZ z8l1cZ+ViSSoldF!dkNc@LcrvBmleU5@KtCe=Pp9ge< zGtZKJC0d$I#je1w4#Jw5ZwiiI_ zX{EJ4X_Y(pG%hyqsUe}tNV=FwniUe~fQu>bMJ;ynU*m()QT~1yl8bU;#WaeoGhqb$ z{z`$>C)zGqD$^*YnUA;Z`RvY#cZ$J5|5oeq1POi1-zj>ayuu<`jWA(>%eK|{)b`>I z8zkiiqaDvwtBE6*g>ax)EFq%={oQW~6Z}SVA(g{WkSq!e?hs2ybvK&NEPE=MjWy{Q zXG8_h6(DE)v|2uK{-;a#=rX^#1Zz`x_>PoSwyu$4C|+;GjQRnG=%iG4kcb(V?`84D zG&}NF2|6vxs;+!0ID!mH;Vz2Ey2r+JKIRPkL?>RE}i>MdwxMvFW??>}tgk$xAj>B^|JI0~Xf-ffOUq#Z#A zf^-aXoOfg9O^PP)t)p$@83|EA$Xb z4zyYFF5J>B*sV{p?S5Iz%7UF}hE6!KJDlVM1?Wq=OAz`Sr*^)Jh{VadSl8SK*Dcky zfGEoTL$bwWukAKe8b7zBpCCBO5rMC{-{)rgDvEN1=_W=spBp@4Jc;F)YR1Y1@^6qz7z%(;cXBgzu`GbrIxo@^0?yyo4H zUN?%`YXCYir=i__ujab3aMzjfdC{EjPBTrHDJWCER2;KTT)A~|_Z-$LyDg zRF)@4S3QN3-1DCOdF^t!3co)#2);e9%wn4y6Dks;6hTpDU1xy6jh%bH`z`h zhZ+LL*xIdtta!p@&7p2l!p>L_lInxkr?3-r|C|2<7RtahVoO6vs|t=FsO#sTE0nX+ z5L15d2k8+zyCyxHp@7Jhy;U4mo)lDyR!*O@nyjM9)v~~N3mxqpE?uEcrRQT>D?tIe zf-dZd3(8r7Z+}h@_MW=CKQTu4%|omy3<*%2Bl(b?De>Vla6udMm~Wp$qPlT@Dv3^n zNBmv+wD>BH#FUWASCr}1`8S+xv)9lo*yXNo9S+)weeMNvA{fPVd6PGR8|5wj*?4=wcr_*Vt|_1i~l+OP$BOGuITQG+T0+zse=m zak$DoCY%k|m8xPfMQ$5x7@jF~F#Ny)L6#F>GqPA&+0)9@ya z-efU^eTDPj%5D=^9;!G6}~xrqfn=>d>Sq9N~&03zD<5o_s(OhboDf zfbQ)qSF~&cWs${6az2hS*)h&l8`@$_e5dj;*aqhg-~zA6%tM75e^6w~zg3ewA<=ue z;$dx2!rrPUp#S!Qx;Hb)9;E}a=8}WyVVF0U$dVRzBZcf#_Y`x|LA6fzBn}b+@9ySb zm%(*JZ1B$OcTaNgm^;*Hpc_Fj-v7Z9YM#nLvljJvGGSX>DNsDlRFM4fAJGHVNX`Wj z5p9E5A4$2s67$>+wV(p39KCjqbT4#D%WVyTX##;i@+ffe;s}GAB%7M1DpT}jzf1hk zYzT7)Tvwbb)4jjPM(`_~UV#6t8QNS!KBK^Pdmv1O`ndhvZ@iRmOTeLKsrFe(7 zJ>%iRD}%ne$T>P!04iz7qw!UYKa-Uhc>cui?k=IB$@7o$>WyAjva;N(Wpx27_iiV9 z-4;?7NErl0dDUu&{fwNM)n{z6Dzj={OYyacD`7_JYA3-cTHzziFpE^s5_%{SRmFA) zCi3UUG??R~;x@gQn?!sy&S+9M)mKTIDVgim?!YRLG-7%jmWrQA&WwVvDKwY0jsYE2 zc|z+YM@a_Oeo}9zlu9M>_4kBkXREuG+@dHKlYu<&mhW>Tru_62Z<%(8CsWf2jMu6d z7_GX7!~pUm6sifF-{}@J2p9M8wTgBuDy+ofI6!W%7ws1EW^{-23#xZ6F5e%13OsSm z6Pk)?MnrAm`_fSn*;Cpy4xVrA`8Z!-Q*CEDBm<8MJCV5LR{)}OaVvYnc3+TzJmxPx z8w=u^upfGk@tXzVVo|UaPlb;{30G)ss3iofS=c&y&I}Pn-#t-qPTdUH1|Q-99G^i4 zBgEztyXA3ihloh9VHFi;Xx~xalV0K2XzZaz_t}q90?>8~r}HWwM>uK;I54^H_EL^D zNmKZr=U)@>@wL$Pn`~$irTZYus|T0K&~wG9DAhhFQ=H|=W?z~39^*|i3H{INodnjc zTf{^f99FX=Cyx2ZZlRN_Q;!)sp0A?TE^Jv3)nPH@dxbNowPo}^FV_Fuf=)DVD^9Co ztYV2%N!WxHNHqr`p|4%8W{%AR{1q_L{-Kp5&_w+=0$pRJH9zXQFou1VC2(nbkCbZj zkxn(bk-rh%XYo4xemW@Or`t7Vyf&vn=MejnyFeF_yeE?MO}IH8VdnH_+v~nMqHt)N zxU*i=-m1HKL4J9~<-nRe0AFSz$4cO>@-&Dn1>yut=Y={eo*YB0Z^by+z%Za$&`H4o zXAAZZ^|5Ov(zQTX5l7`YWe5(A!85AvWh07&JkL+S<3Iex;QWcm=7EZg?aLX|iyHRy z`-V(%diT@JMehe#uwLZL_0Mc=mh*j}9J_;$Er`e2tOEoK0D66~6iR$=zB(YOL=A3z zhw2Ztte6!L**EI+$s%gLS=b3np6%eT?$$+O5_amz&d2&PdvhdOW^vvL(XCwvHnE#P z=v&bci>)jK_IM~g6QMMCtq7T2S7#ib$yKUXBTrO8IRMaW-WvaC{*c=(?l_9*Vc5EQ zcz=d-d)3U$HTU%6fL$iF%po*i-1Mj8C${KTiY^1tgm`f0VV{Fx^imL?M{=v`IMC0Ysa|*-5&FVl9vnXQPgfA zGfX@e+o|av>bdjx{1^dq@ZzXW{Q637xe;Jr0B_s7sv6)A`jKd_C!FX87Sr0J?|SX{ zSn7yZ1KdGbHNl5Z4ckMUaYLSYC$mRw*! zFNE?o`nEqJr}J{Lo@5i@x?T_YT09pYK(GnZ69f_^!Y4pR_;l(zps_M3-z5eM=0$vz z#=*5z02yQPM$Eoa;R+bN0WWavX%e#RUgX3=VtF3|9WKW?K*TB3{2=*NFP=>}*$QpD z?sTDVrIDU`aW%9s;71eA%hcbpY-YxF${yM!{|6ghY-VwrpF#a9^_cv9#mtU}bvMp~ zJMC*EGb#LP@&zBc$6BF}AP}fBHWuZ4Sm~~`zIf*sTBJYs^7kIwbJSH{^|%fyoV`W5 zxtrS9>%kNtz#Gf{*NgO?r7AQ&?|Tuyi71v?Ej!A{6G}tX$6>Nj+tUctVnzsdBOoD+ z;g@>?;h#MO+D~c3PtJ};p5^~Kc8Sk4oe4teYLM0N<mQOoA z;PC8Mm$2CTA+<&o*m_n~2p64AkQsez?&dQC3P|zXZYMNIST05AzALA~eiE|OjN6xE z5+#Y5j%XfnXx_t`Zvo9~quLvSIJ#;K*shUZetJy$%ykcg@{U?KIB4tkT}4+&QKH#S zBRxvx<8s+=chUB>8S3}NUg_xS>3GJ9Ro(poL}1DYLL9oG)c{x!)wsY2PsuvKAy1=jK#xz44w|v(GnbJ`MTj?zHIg1JWjm8AAwNwLf6(HR|wk`-dCI^S0lMh z5j4aaR4;&JBg+zFof9rf>_J|Em~ea7rSm=>>SF9Cu4^8fYx^~leQJSIud3e+j zrB|MU6^t*~7kPEV<-zdsV+1m$*g$aC3N#`|dTws$+lOzs3v>Tfx&6RxP%CfUQG}cf zGlr-(^OV2v@_~=mbOd$$1&?sK2$l?f{xiURS3M=7tp{N=vQ^Jy%!KiZ;!ns4A6j#* zxeg(;)3no0O+d3K`$H9E#v=`EZC`P*)l6nJe8a z;ie$d?=CShr|r5U_CBm9Qb37S*`^#KrC50X%o+Sye3BG3h19~&}Qpom0Frv2Rm)(J%w@ROP`@dqIwmh{Y$(gb}B6aJsX4v zll|<{%GJBNC9_nz%^eU!1%%M)2h9WbgjyRPEZYW{d`-si5k_}Ijw z0Gp@oS{3+pc<^ip@6}UwoW=O&FIgMSJlYjrbR;cK61Smf+r@Lphe2(B?@mKW-&e#e z?lHwiXoyc>jPlVyTC1i)H!+#-@5vB733p|Hf4pAk)wN_6vBbAyUwrYy(;#b9`i|OO z2d^LBgc8e{IOyjRsmFj>3y=h))s#SmjWm<+5c!GReV81Xc0B#=s3sY4#bWG}p^ecR z^xT<7a7)-fl*dpo8Yflbffo*%8wc?^&fWFp40u2}B*!_1|4hSq+91OzY)uH_27GBg zUa-#w6X5AxP`~zLC3@iF^ML%0fv=F-%jTvdG0ucloWlr-IJZ3pCn6GMBvgy!<$Qxr8y_7E20M3GD)a*+$CZ&XA@Yz zN(-@$YZ5*jA+-}}N;OFC4we9;uuVKfAg@ArYyr*Ghlj3~m!6DQ6%ZBW{2q>cVu6+@ zOMamqsVtr!73x4k6!HHUfg7ouTJCN>;5k^lM}9YC;lg7a#D!1~xhEooYE&4`8L z4l>!tsSJCy>Ai~Nza4N_8Z=pyT~~|Q`&#Fum55PecS8a_Oosmm4;G>?Jf*BZG*rSx z;lQ0C?@cf$=U=uzWyA=5djsJa+J0y)NRF%rwnKgY+8ePK!4tww0N}{ZxI1W}uYfyH z`b`wepAXP4B~_X~5#A!y*JZ16Sihvs^f{xb{@|b7d#^LVxx@y zHefCAz!Mg}OFT7U9}f85yW<(}twC$!gjqv)8&Zv<;-m^-wXo7O>ap*ULDs{A?M8hR zweYMEbGz`jI&kv}os_Zkkg!b6_W8yPs;E8B&h@Vr;`5d+a^z;p59abia~>1Fc{p~C zunTDlO*C3zx2K@QhW4vc9DPV#rji_K0{_BLC+pd~@J|(3_{jJo6;N|ttT)6##_n5> zorxi}1Dv}uam9{SPuX(3fgdR$KWujyZZlLhzwB$gCkt1cEDhgG8{P2&=d;fMFfp^+ z6c1|gGh-u?Y+utnkL7}u+r!7_s0Rwka(AaVggh9Ax!#8aZ79Y|7&hMpv;)RF*}EfX zt6R9kR!Et#!?h2(8%w-bUN!SC9qXppfGwMJj=%j9Z~z1TJmq@ogz5?=KHM;79?GNt7#7Yqj4e)=$UR~5%b|DJ z#h#eo#!eliX;Dn2NTPeIuWI%kXhDj(;3p3X`_B*w*jk*tCg+`_(%H zDP1Q#KwvcLAM-Q&*S3+Bsje2bRxmbWhV%*_D(vVTHI9|iJyEiPMJA*2+kUVX2q{~v z{xUCqB^#EmWj=3Z)Zub&V?Jq^1;;dw78?FC7RTvo&I#?LOC&QRkam(f%o4GhLiu*)~}Kf%!(} zPDA%Z$>e>&oa6&nV2%h*K4^oX`qcG9a<8ID_4Hj`EhGDB1B49&%*eqGfdlppX{*WMgc!f2|%-b&;+AGx~I7wphiITG=ZM!|k_wQWZ1QzRy>6 zEYn}*Q*5^6<+CTlnoZ7}Lt#N0C8pJ{h12 zpIJ0Ub0o{dIg4P}lW2ktma7kKAO1~8w#rJ_Y(~^y#)GG&UZD4(sT8-FO9`tv(C;0Aru+gk^5W|1XQtf=VinxLl1;Q@6m)_;Cdu<(x539ts3s?RRWp z8K$2aQt1+g!r6Xb_W25;p+5$Y+G)B#aMX3CGrCnhGDE-Bc<9bw#@;p};?ePVvs<2Lb%7{@$uMyq7i82hU>58{(H!_l(4)`1lbaBY1p zR;QY(X;qbgI{;Qv=A#wPRDqE?_6RKgJ!)^RYq7Q`<}|JO8I|_|(3p@j$H?!N!E!!a z3-Y)+!*d8tfVKVE%X{&Q-dK?3Eh0=vi%h$E3BcVRrV*+>I|A*@oW1=m_-)D}o-wa= z7dy-^@IVyCz|1c{Hxa>_T%ZB9F6}+zlACd-`Fy&!$m-nmb#cvTNeXR;5DfN6%Zh3w ztkH2*N1$H{OxQyU&)$UWZT-7?lnd9+PWQG)f#z67`kJ+pQYYR4AwB+^MZBd>3JZF> zpQ_U^aHeS5O;J$-c6yJpPNp6(pHqVwO1%EEj3O1NvT*~iQy_M(x~R%@iA{_dyE-e0e6hjRqecuu zhDP{FmAjsU*dCWWb)Hs^cDEe_fp9;j)tl4K|0PA(=k=3X)JKjZba}pf@3+Ru9D$2u z0e{eHqtb7Yk@!$^c&P?kGNV54%EbChhGwH1hqyIbE4|L4xr8xh&ma`9uBIcQLXm}qPsuY`W!Kir0%PM#!H;v}`5gK$GF$$sM~X&Sc|%$El69jq8*4 zg&XaUJ|itBjVY#!K{28lf5U$fZiD9Ne1+s;D8qp9S>I*sD6rLnEpk?K>#S6+Ok|^$ zOaBRGjM^0nHlFnue_mpr^)jJSb&JW(*WGR;=(tO^oCt5=Oa51qEQ*j2SJOR~lj(C; zK!Tqf?E;H32zr^{>V+6x<`~w8_#VMyBnAh>&D+1*cIlISKOK|6@wKkQ)Q;#gHoS! z>SW-8T=3ck;C;Xsal?b@>4gOhxLAoTq;*r2&np+2lRpGjk z8OY!pC8Hlu&edxd0_?y;0)yfN*ao;%b?O_;-Lbxt9XCbYs?A2`*V;T=wRrhT_Ol~s zUX@MXtRaGhWPyHNPyJQY{F^Vekz*=HN}d1;jE9wXJ)(r=VC9g#HcyxK_!+NjY2lyr zONgT90RKAV@9BUi{X*;Qpn!P)C^^1}NNqS|kihU7LSr4f{Qp_`=- z{fv4eHSG27fAtN{qr2#bf+#d~7CoF;Ddq^zL1YEkQGjrw4+0V#5zlrluZeUSm0XkXsc16 z;U;;H7uh-eQ6};juS+{YQ+%$$Cx5V;4K%c;lNbv4QNO4qqh}(tIdw#3S3gFf?91 zSW)vD^ullkLHzX3B-GuRvPd&pBK{!X(PQD6lbS8G-*Qe&;Dbst5vyPDUf)9B$`8ry z{ge?hH#ti^eg7`=Bo~~ekEEz5g>$I|A=Ka}Kg_ITmq_ve^(9S8E(97wBbe2&3Ub!$=tl+-eFXYAl#PZcUe3X$j}%y(s0PHe_>Dif zY{XYH6dnQ*{gqx(wW`u_Ba5O{orgf&Tbu~YhVU3;5hUOLZB}t#6Kj51=Zq_c!Qm3K z-;bj8Y?q}!uY}&8gzE^&H(fomO(}AtD68p=o-pIw#%@iKbmp~RgDKwN&I7DOQ!fZA zA@);dTvh|Jmr>y+s=oP>ja9GH5wH(XA5vT)Fc*36-1(w%=D_Im&ybng=?VhhByo1E ztgBtST=z#>bbd8q$Er?F)9ALTkEsN}mG_7B*DyO#k9zKD%MtZdI7T_)D=`uP=K1|GcrOX`6Sq zlbm#ZaP=c@fWKM-bKa>H<%+ZBtV(>iqec7Ey8=EeeN%+2?!n7qG}?9h3<9D z_gkfo_;3w6Kxx<3Mr|P7-f;w6EZUpQHA|L|1)b4B_L+%_qK5YHDG~mmu?+V@Wsx|Z z6uJnMd$x1e;ZyP?`#E-;v@{ar zy;}?fb~=)fO`AU4r-vUDztt}=a*a0QmvWnsNiO?5djfnlm0-LAjzM$n$!Uba(U=UHsEq<+{Oz#=m>+MON-WBi%Pf3|YrE0Dt!v?&qON z4<#I)_S`fL?%U=}S>35Wl6=e|BjCVVzwW%F+&`yVVfkt&yI;&fwty{zGG8>l)3-rJ zd`U~c$pW5p3-RE*jqJJQ0mmWeqT_n5c&(<}TCqNBKUjGra<>eYVj5ts@z=wtDM1L@ zrfFn=Pp^XAF$>H;K-~vQYnyz9KNws@t}3Wsw4}rPHHuN*E3X}4h#M{(6W3*tH(|`NvRI|H4lasV zB5zFG#o@uh>!Cf+73k4_A=ea_X{e?FO?=uOo;84toaaXPWdkVMA=w<%4!j+`r$YwUY<1J=^N@PsPT%#V?mP6?^E(lad2AAcB| zQX=qL9kS$VCzytJh8Bh1_-n{w3idz~k`ES1ys6~hkG6iax1=sW(!I8UOLB}nKaAZ& z%rHC`Cg8`$``ET^+qP}nwr$(CZQHhO&A0g%lgwgv?V_6|ZIhGJ`%0G3K@n!lU@4Pf;b1p}HV0HUE?3 z3GsdlxqU}O48YWHD$i%GA5>!k=Z*ItwvwWY8ojv-hY3J2LRMEQ)$w=jB|9`ksg!#0 ztpl5Sj(>-A#IP$5rPu>Xo2E50-rt+e%3Q0%#y!{o%_u3dn3-ZcirN(ry#@rHhHjmG zKku=In91cwl+2kwCrtb5$0Ws9f7>{PZHQ#f%#ApaT#y4 zd|XY}Q4<^XyXg`qLlStWz(%X67cwe}TLzlQ6@Gg{v>`I?tHS9<;n@ToApU86@8FwDS!sYLm28WvF zDTv@nnm67%fETRBQt8&R|2)CZTz{A;V&q9%ymWEd*b?Q@w}sqgaHJMI2pXWPRF(U7 zO3=4r3G@BUX#H~9AqHd-L0KUbV{YsQsXrO^$yQyw`%i+F3|;WB>1#xeFu_Mf21lNy zW#&oS?+#xp)Z!SltmW@9Q$*~juriQ-|J@q|NAeNx9gi0fBkoVW${&|MTA5G=gybP;%?tWtZsrbJ}{>t4^{|8 zNfOXuG)itJ)6YT|6cr!+rY$;gKgomZ#EOqck?qwGK1PHlMi^Bj#rI5(tZ3UJj7}(& z?h^~7DVRx~Vly?3o$!r?;9qEzVB}^JWO}&+q?|DEtG`{JQ^YQ*{i+7W=K`PXZm8Sk zGyc5lNDZuie-)=^)_S>*P0wXb6)vvmuN1_{YqDnCmj&_VBAOSlNH}C+C z-BfdXBjuZw=gpB#MJJ@|~#*)0y}5DX}MIPQ+`eL#!ZMyu|$Gbz{Hnzvq(#J&d} z+1)fEdN_e;H>}+xw4a!`AwV~d8i}#Lh=Poz`Wh_ml~~4rreCLWJ$DA$%-hkK_z99k zwVl~6YqQbK|HN6r{PK0m|CCd&>X(^U@XdYTN1;kBmP9Gm&TKfaUDa)RjDMvy&buBn zX;Y>+tp}?`5W|sz`ar1U)xw%!JcU}oda-6^+-QKM8?6i*a4z0D9o8Bj2XW8~1boM? z@2^Cxhnp4_vOD6VLP(KPfTrtp>FLkI=!@bW+(!d%0);5=TJC-tJqiLcg%8Wt?spyK zjCPGm!v400PXOD(XY~ojOZcwc;|$?hesX{oNX3U1rwHHqQvT(DHoDd{yPjAaP zMm9YNKzb7titBBJHW^7+>%>uxi|B@IFc#eP=h79MfjGo2hP-*oa_ej&v$PI3{-y|- zQ1fmhykS1EzjM&*i=cP{?&D}KidAKdKgh;JO0wE~H*TD7SwvSTtkd_@*hu5eD87>%1(fzn;*{MGi&ukV^3lR|P-;Pt` zR`?1E60hI5ypm|CaHIg7^Y&b|s|$5rMnCT5q8&rTENY^_xm z$@u+SUG{+;$lKdnWVru$S$haXi2ug50)bouJ39p0j-NNYWnSKXclyX9+MI8jPIus7 zA_9fwf^^1Ej9`&L*Vz{t80qc+jgFJ}T0~r?sGbJ#F%^U?5P^7bW5wQTMq@=d&Q2rqEaOdy0%%JFA0LT8bqLsL>{H6dc3wrzr$H&rk z*2mAcV4E3Qn}B>hW#RtUJlHfdG}(X4`Xyz5N*8ji^uYX^lQ;!NWV z0LKEhccM}NEaPzcFwOLh48Z==0G0eQ{+W^KjZpl88Uiv3CrCqH1}JD=UhzkN4hRYg zDoOdk;ZPG*k%0b}&H#`sC@Xw>Dgmx{-!m5gkCk`7d_KFtf5lS~6;KtH6p;)}zZ$^) z82eDp&W+zouYT5|{B2_u&J=tNV)Y z<)4udgET+32=o36c6MTX3WtXt44z4;$(g18ZB_Z#D8+%*6Qw&oyAAfJPOAWxoME3RfzT0AMXanx-2xM>neFD@wx9|7f3;zqR1yIzSTwGts ze)BheQispLg%n>=lEdL2k^`)Nd~SXWIQCo%i0bODq!*q9*-!q|oDSMi89{G&T^ZZ{07&-{zkv&Q@ZZ)oHtb|-dqXKq^h##;LK-ysp*=iB$v zYWm}UeQPyMt^IrM)0+MR+`H2H25tmKr}=NVX8QvFF1z)Kcc0TX!?(}nu+>NK#3j?^ z*YB2@-lTyAD5NXy~Brp7V`e5%Iaf5mmG)mgu zw=ZpGFSpnuaQVnjRm05JXJn>aHbiuN)n&>=EB>h%JK{=H8k$zC&i4o28s=bgvkA>9 zC)GiR(c3bKr zTluJsqdsPw%HE6FemGBmEbVX7zT$p$F$E3{P0bs>knoxC z|Mpi&M|w(f5p3zyL6%Zvp}KrA${FCzSgSGWXQ`f-XTkg*&c{@Zb-mDXMS?{L-T4)$eJtV zw840k0A7VptI*MpAnjpN()b>8D{x2?y`e{kx02R&Puf$Q2cNFgeflzj{uVzs3=r8xm|IgNRpeAX7-{k%;)g$ zwhuOrd29n=Y8pS<7z!3q_BTV^XvZEhRm34;;uY@b6=gBkclg|3^`R)((3Hz5CB-S% zaa?6Jf1@6H0r%4J(Xc5->y^n<$Lf@pOS%<_u;KF*6P^H|*kPtKbtC^Xu?4!sy|xb` z@;b1G*eF{O!4%+ohTIGSXXc#qc5KJC2d4Q({#0NW-YPiUma%}Bk_0M{vsMIOJP6`Q zqK4Fk2{hm3qUGdxtltj_BfJdGAg|asIeyz@SaZL_CVuOe1qch%fRgS&4PYu2=^YCd z_@|f9?Ilwy48lWY&nqa28jI60g47GOYKBxFl^ zlunb$Fcb6L^s!zTpSzp&t*KS26O|iAEJ{8ezMwBVYx!CRdB-4x62hF$#Uu5^dLtv-e^czk4}Tjj_L%M z28sGSx+alrT-0_dc^qR7C4kw&5|sq--;>Qmcv%((ABrRf z#R!-#|9*91-`q2ip-_KsqrGM$;D*Diefewgrr}J^KCTaAWh*? zE~!aLEja8yflvN7T1YlnqCx?2RpDL-9rQSkIX9%snWz(O%|8XUSr(|-A23iK>I575 z2jIJ_bL9p^GPF9}EI@K`2-C=mL5RRRqWaEDmnUTbiTPg_bx2LrRJpcP<`FqlLOSTS z&)B`f3-D$=v-s zV=+5-4*PR&6H$ZJ&Ltzt$)SsY#@?e(NOL7Ukc3=c+B@(Ld}vH%5{1$Cv!UPU_4d|Q z@)*YSU3Po*4AxlofQCxU-gq2Sq;2Y2+Pe>xg=5@fv!dxRDd{i_m#Ai)g8jRT|6GiPIm@Fqron;q7&C?ywFFDbPIZ} z#$ODB8gtcb;1JzH>edj3Jmur9vrZqqtlsk_*3VN}wl@%~uQHj?%o=uGwXE&cUNV-Q z)sw1)u2SvlG!RH~l}>%>1J^^8$&{P>hl=1Di0y}G)ul9}h~JU|d+UN2?8dX%J7Fmx z+}(XK0lb}zoD`CXtU|xORZFLE6?-q~y<|i9#JJ@PZ;8Vuo|0mY6)s|zzEQLJf9lLq znoz(9x>baUWe(7 zHkZZj-uyFcSYg*VYpzLqrLp*H_12swo*se(L0Ndb{w70xJ1cA5rN5Wy-0C#!6CS>jIT-|3xf>v6C8BqnD!p)*1CFnpx817fc@)`5gS zvxO(F?5;5!6&j{|zo`&fO8Two8$28lRC7OG6;`27qB^#dmu;NUW2|X4> zi#`3vM8D`Z6opQKlgYHW?oqKRt};h?n>2k}jNhFz2aPsvNH)`1ntV4XTqdT$NhP=U7K05JLOn4LBjIAh zs@ItF9zA%2-2VU-#(R?8e~vbR2od@Si3(!utgN`#N-n3{Xn zs&=YK3Iyxd>Re3@sR<*{8+-&8ONVi!=SP#q6)Fc$EarhI)^YJTv@44K=JxQ5({2D92Lu|^tquxbggGa>0<0TH z2$gXNk8C_ZnO^j3k$+%}iGQIcZ`sw!d@P=#5wT9TD^JQJ4u!lGsHo;{Yy@0G!ObFU zT{~3Fvs?k{{>wE#DcLs0PWo&*iYQkb(GD|hk%apVSihw^c@CAE4j#IIN~&AA)`c|&`W_!OM?0T?#^JN7`v?B?<5$Y9UFB`r1ZDj{0!5Ek$;{N#4txsFez2$p%B^{CcO6(wN}ho&yA*TKH-oh>A5 zj?nN3f@Bp|vT~~P-v#I#iB4bcLiQ>jKgv9l8g^&5zFyiY&(WBfd#ja}g&4;+KGliq zwseL^(`+lHJf z0KqCZbbfX0;D@ZOa||V&@61H)y026Gif8NwK=z3wEXt5l5BbO@=Xd@vWs6A}q|u`Tg}FnaCD zmbc+s0pl2}f#MjwQhq?C=3$mTf=Apf~4(qN7g4FSRRxv1$B^5-;gWa)kd6ns%#>yDDfVis=$3r;g3T z?2hTU1u506q?6f)WXyQ8^EC;XbQVB_Eat6Egc)x_4%Za)u#4#=K#JP5Tmb#>u3zW2 za?|xB?Cv7B6Vxf1<3*t%Es(}~EOQbcbKD{d`L)Y!c)DpdT$R0>xcWB z5$kf{C{^f<((vNT0H@Fg8n5@&?60vTq2I|b`r@1hasclTM%4bZm|l)VU-&j_6s}Im zrL{N+zSJ%?Oih!eF>PkJRV2m_@#mqp$GmlkE1e@OuxJ0^GnU;S8c|%Y{?_V5 z?jBU#z@4J7uElt~vJX}2W7H-rSP^rQ2LyYY^B?-Bp9GB|Qvu7ayI|iyI`3?_yP)!F z(lCBEhyNZTNYGm8zFhQq#d`A=Vt8R0TZyXdw#>++G|o-+04PIE#d9%4yWqvtbGpWr z{<|lUW(ppF@_!LpN{Kggl=fJ$0u}~QK30QRdJ@Rc-+P3#=6CzHityY zr{`;)Xs8*$cT1HgYa{Jud@88&y(-WRXc-4xcr3h5Y%Gw$3*`mDtwxN&^H}q4%G7?l ziB(?$VVP6~5`GFH3ih9{^sfm)=$dBHcJT;Fe-kVh2u@&j+H<~Sd3Cg35)0euo3s2} zT@h|k__u28x>$`Y&%V04dNR@ffaww7L34l6W?fm3(&(-=X zE7GBBPN}SQqw?HnF^Y)11W&|T8T&yT?}!d)YAP)ov?*U6%msXTKhU7xwnf5s8eVFG zGI!c$`7jauVQa~ky(tm=E2R**bT`(6+F~>j`9?9Jj&&sJB&MNNL+Ig!zOKp2|1bP1 z8r1JA(?dtelsS9a7d8{TYfaG_$BgNC&H9po2X&VhF8#*%8C~Z4= z*zn*3-ZR;*OoS;yVU$ma<;YPbRY^zIW&0~BnBikJklCJ78LuD5k4O18)0?Ic+2dt; z@rab{<+LpIPV7?j{S;@HBG~`ZBg;3qhd?=ed40>auT)K;D4ASMj-;EW3NNR*o7w{k zX)<(HM%HlV_O9MCS1tl=Y!2J}I73VGjG zicxj$a*FXN#}N%Z_bKp&8tS9SmRLU03U%lHk#pl)5qk}j2(fSr&A(AvoaTIcp9 z(Ay$S>)tZ~aRe_EYh5K6`mJbih^@a^;ENMU2@~^M)}{odd*yL!&I(x3Mb;$Ujeidx_`}UyZb6Dh)3c-uG2>BaA5)1 zb5oR8ZP6m+)2TV~TG}h}6>y^i&Jh*l*9K>LJrf_8br-boD#3H!Mz&fao7V*uN59*V zBcBAaAp49Lq_{2iZB-ML1LGyU$Ets|xm|r5W4(QyF1`CFIg$7c-uAMk#i zi;phJ*|ir=Sr8-$c&)Iwr0zYGR&G`(rjO(wa(0bn&=|w zX2&ZqJ<{+yP-GtvAw$*-I$nOrog>@H0b0YS&LOXnWUGm7o?lrjyQXXbJrxOT&aA$` z+QWkXS$B9W8M9(>;hbym>`&N&j%sfc(KU6Q=XU9Y-0cd^V$Hso#2^S`?R=P=#q5P8 zneE>zM(=*w`^|e&qmg|FYCK4nilR8C0UOtj7p`qg3q?+W1f}dQQblx!*J(4pNsFR( z@u%iak*_Pu{wu@dj+Lc@Gv)Q|>H(0mj7Qt>clo(OyVmOS ziW748%x;+ZMO$%0rKN_}g=7vgLHHKk{vx`zKqLKTVu<61>1V3n-wK6YSZ*{S_^XH2$ z`?ov*gVMDTuDh8HS}t_~4{A)^rKe!PZHA5S*if6(E=B9y)+I#|aIC2n9|oAX!Ep)& zK~Jt%#+5ZMi@BE@Cw|7T44)Gt1AOJhRpL7vP+M6e=rFQ7x!7v#;YvPe=lC}NHia}B z=$*BgN`++vgwN8ZT^$P{Gmw$ZH>ac^)dlVEvqrdeYth-{0v6h)beXoQSVyDxh^Ut8 z?*2pq5i9&lE2YE?qjlgV?Z$T40bHIskD~i|2$@XQzwoeti?&7GN$%|P{(+jSE}rwE4_y%Et6RjAW3O_>YWZMk+&My(6&(ul+G0beUDe| z%+eM4SxbmCow$1*uAm9;bz}dnN?a0J31s~`lSdB?&K052q^2!R2Vq>cj%NeL7-GsX>Bcfm}F@@ zuzfv34PH7y2;La>e2&PI=k#`&gT0g?h%dqU5bMSpnNTn{w>6oWJZz9H$JAuY10ETF z$RNBN4F=^ZWr}Q8rdf)Z80%9G0@d}t4%NX?k8c4h0*@z!caOw;TU`~nc~mTbCT*j7 zNq%S;%yo0T!Rv-?rX@P?v;fw&ZD~~#Q{GCc6Zr&)$A@<6q(^$K4eFaRAEghdp7VE} zc&;#w5AKFqmr8ui(h_+!#*zgOZbT{Ftyq3wrFq=i-kT&G7i~||mhk)OMa}i%57L0O z#M107i1ulofHW0d`BOvq>;Es6nf${9W9p_nwbB?eD~36)MQ<(Hh-Amo8SAyfLeFP@ zcxI-y<@MiPVT(8!sg&}$=wu5nW^Lw*#$hy7)i4d?fRTK92}wSXMM9DaElX1 zV;uuWTI-_{=pp4!q0RFTb^x+6OYUG3N_30@Qh@xhc(_)CYWCa_rWxlEt)OtW0HHO* z1yc&&M(vA+?(>&KfO^;HX0iU88n3-?^fmdWMt3Zn)-?s@!@&_lh&cu08rHNnGi8V~ zopAEXrkHwCH88*6(q)G#j^}?VGtMdsnKs<=lu*mc8sd2r1)w8nBSCBsR9e3Jkgv+( z5Rgd2aH!V3-yT3BIVMUTG9a+TQaTG?(}YbB=4>gGCA(sham>w_Xz+}!~w?n$bW7W-0)9fl`<9lDRwz7z?S8m{C;P)bmm z*0|%YyB70s5^|mtQ2nM)l75N8y%W4gh)-V=pra`2<|JSB@=YgmYzs`^tBCFFkI(M9 zhc@!ju)8!9E74C88DM{}3SpuXFdq}qL=PG8h5|f66h>Xv9hsD^KJwR}0xKo}epUTI zTty@U-Cc7lg+ylSs^vZFa_LPZz*bjd%Jw#S;Q*#yG_gs4?pZpX!DkVLbmdN}2OlrGKg;fHI#W`Y5J1IPpk8>` z_MhBVeo@6jejR{K7TIq!sMXT^VAP4)iq)kN#5x+|+r*I1!oL4%WfqO+uWz@eMKw6k za7wP$x)**19$Ql-m{Q;056a|r*Ax}#P6c~9SsIY5huEW8h!xbGJAB>Fs>MXnk*LL% zo*YXwX62a+oaD55AR`XiRAi)OiQb1txdcWE-)q6>F;)zb4(aR`&Q3RD1uSOc%MbcLOrC=1AE9rev{3Xt2QGKy;R8NV9KEacyDS<^Mc49tNC0DD*uNc3}$ zK1eeuA!Zx}lB1y;eS8P4*FK84X|^|tQV)75XKp(W%@+4xEd=aHBlK{s zg4tWjIG0f%&_sMt2?jMVUZ)@|NkFeWxgYmhHsW=m_l8j;o<}u3YnpJ+z{1&x0a)i@ z*k04o+cybE*ps7XE`W(|cF%_Q_H`hvWt?XZUk}|^fExcx{<>O}uk%G$8mXd~uL%Al z#gUgQs{0@|tQA~NK%?q{S98=OR%>-IN2KPv^rPn}6@uc*pZ0znHV2|2;A{CW@%&_(xJA2+FxQzY9O&^-6W9{bn*M*U#dKa-tqfaM#c}ABCzaV^2G(?+LfQk>{2M|+MOP%7ik8}-oXdj z!)Ym-sHoQuBb!ibtFcmU;tpd+m2Xu#nXMc1&%C^PsCcWpa3=ke zkJ_|OpT1YJU4mF~PlPTg>ODyU;!DLC)8In=kl1gK)z}36+xT53g3xhh;=U;20ji_@ z*J(tPuu+R@2`#sIeZ^(YYK(N=_EFg~FIQ|icylpJja?6~Axv3BWr6AhGGLn zi{nJuY%AkVq{U_1_u8ps^R;J~Je4W`#ysZ8Nw3k!!E7Zc@j+VYAPCz=?`-`f%2YaFvq z1?@+LLq!WC^TK7+P1Az*=WdV0`a4eP)%}}ws0F&j89AxS1d5#nT+1@w%$8{`;_1RS z$9jW|MaldO>5UCf;vWI+f1qe;OkSH$7DGB6iP1PSw?N-w{dtTe;j8K?|f7 zD_ubhO$DF%lk?Xysu~RSA@(Z;A>=%12@ijb-Rx;*z-MGs<|JQmw?9i8t{qY3>z18! zlDBdp+oup@FN#wa?^9Smmla~)M+0+X9)=D55##a?wyidYe>N%ExGoS}>{A1kg;L~3 zLFFc)Auqv{|NGLUVR~*_(`7b1aC9yN?H(sV%6xKAiNi{AMe_Q}=$eA^#=K;}XnR`k zuY21X-ol&Hm+3gKax-6wximfK)g4Ta1W$G6cF7G+m2(v~RX}1>v|2X%N4xN3lKfnw zvjNtK*|>A(oj}(p$7o8Y9YobfEw_qVDsl3Q+|}i4HeoZK_P!}ePIw6`Mg-Ln{6!oD zBB>ct;LyZte9uf?btIpncNg(Y+@e118`|R`Gb`5nqPxzKi14Zv&Go|BQP)3O<*Z)e z6y!^LPU(Mnfi^O^v3lTAqk8d_5c1kM@%KNCUfKKU0|WLP;em64G1)rlso^J0nU`(PC8b4yXP2!f9RX5TgB-%B&XL*WF6 zIWIEO>?8S`@-({qQ%&YTWpOWW;Y29s%p2v$a z4_sYk^;2}V2#&V6i^@=#OVDA3jifeW{}kLP>|4}0%s&7M*UCs~hmHKSTJ$@pw#j%- zS;Y#PFG_PJROYrzW(@DSj*Y8wZ*;qOp~OPDHJWsXRKhZfQYxt@**Gb!pEv6!@b~ic z(vC6u6b1iwt>q!9i%t9nF^iYoixu~r0;70`k!rN&a9RuKVRH3GDXgCPMYJ9AVh%jX)& z8>x~1t>tdM;BC|KNrEXzwfp#5Ow~Xf^Sn{uzM6I-1A9}Sj0{0Nn0S4SkM65IYMs7w zDfR`FbvijeyK=r-LhpE&p^evDVxC7t_3q?Iahu1cG%zxvrd`#wFYr0crm)D0IGE0l9IyQ z_6Z4CmjPl$VRBN+h3Bo&(novP*bI8|)ewdb)=;>kXuUiswUoiC=(Th{8&o)ADh)!RL5ZOw%5r^g2%e}G9yL*Dd%>Qobg^v26-cLtwvLb)p<7rf?3Uj> z4g1XQ2Pb;%eqDKik`h&iNM5klu&u)9U{9$F*6#%?vvQQO`MS+4rk;uEZ+6Uv>i66x zDw&u#5y#3l`U-_WlbagmER3?Cw@Uu=cmg@2{S8V9hahK4ug}Sb;VmioUxhhRHZ7LT zcNanET%0#F8ZJtST&2i!m#bO0$TH5-+-b&}#TbG?4(@p*%;ms{PT+Klojg*u;3Gt` zdh4blO7;yM%;whh-VHvMA$Eep&e9mP{3Ndt6Ic0!P30Yg*lAq4`|z*gSw=0A_=UUl z6t9Qwt}705KWUl`Y?hdlPx8C^7f%F_whw>aDNT|+D2F)l6ayw%4^T4{DSoEUeC4ZWA$E*R#h`gNf5Y9<6WvNj?+Qgvn|mxnfQXlCXSVeba?%Z0J zzx^TJe3cMd2vyvm3RuPWth%!vug&YLahKkC36#LCu@2_IWkqAFkN5NnSD>#Jfr1uq zJ1b3Vl2V@Ozq0)+sog|w6M0(5wR16=xxk>md#YdT`95KvmMD7K_0HG?@{?+za$lKF zxIGO?pZ=;ZSgkt_Bo?=4%6J|}Qo#?+NNc})C(kKW`sGF>Jp~p?RJIIWb%>y!K?1C$ zkBu3K(0P2Wkk(i0tHD?7Rb#Lq5_9|!AjDs&`Bg~-6Q$~ zaXAe?qG_5vs_L|)^rc0b)SvX!0V?m%{)%bu{0|S;$ucjE^2^mO&NXHxywII)_g=KA zvpliwBK<38koaQxcuc;YsRt&O_&dcK?uRd~C#pzU&Aw(?gbF`nk`XW}_3=#PA{q@9 z2daxad}KkgnOQFwq@^eC^r%)J7RLrhLOaY=)v!F`e&rKh)Qm|~kVVFRx{k_mmc8V` zQfmGL@D~^9nN$AC{4jFr=@AseJa!>=0=WMQDn5~k`M5Jr+lQn*(<=r<9TZPeaN^4v z0a+eAzZ}MXWhO2G{?%7TgAENCxnnG?<=x<=2@(Xk`N%}S6m^0@5T8q9kwfL!xYzEk zQ0vxx#Ep-)wu{6+dvu`rErzEm?JBXt?RDQUg!>I2^@uZz;iW|f+s#*Dukz}3`_dXb zb}bjb_NR)1x{HmS89KO+BC|wGWFMLDP6$-0ER|DVdus8wD2|!u{69S6W?LRH1Su-D z(wdW)v1N7GG29vQQEWKp(+cX_qm+DNOINj9l;l=~vq-~ji@GKcF&O!;{jS+LobPXw z`9-p@m!@*PF5vFfNTfTmf6jxsb&*APd?aTA0KWU?yjZdtvhLP4RZY7I``~RKDZKwA z6hNY8)Mv~X`b-%qlxq}PD_QKbZeo^`oir@2Fe7Y=OPJbhgL8|+FQJ{M?)?_K)QNiX za8a3l=qd2H%YW+W=$*I~Zs!ZM0^m0)WA=Tj^pO(N{#6x}@oB}o(uD;jk$ZusQ)qoW zBw=+5vR(Az=eYbc;su^nbz(-kbu#x9(ydp(aXXh|4>GWGXG@Bqjj@#+Uw0h3bqpg& zV{51c#BW2U0m+aL5veE5V$gCM$t4{Gf@-o=hRPwtX#TC9`@XYZ@0}lk&a78+RA~n_ z9GfpsutB|?>7jveFk!(BmtjSY{KSes8D}}p{mfb2ZzIP>zrFgh{;Miy6bw}>vU*71 zz1MYKEr8w_lvE(>WBhdNMT$npe`fM6kN+1?^!gAWWkv6KiI~MG*xX2SlzS{A zJQA!3s2*$!Nll5YiK-d~_u{#v3>b2(8oF(wIz{&mxg`3|Al9e;j3HzgnjRsC{nbc0 zSFo{Cq(qCRAGV>+`{k9BT3@ryjuEF<@c}(o<(KlpXOc>?&DzrQAX77$?~}n zn)p8HBcCg;oIZ0OtvuQN%4}$Ims1!MNuXs;4!l5aW6Q^yct)SqTMB#KYlQ{ts!TSx zt9uer-{{8@l=y6l<@{C-lL+?Gg9l6hlyk#yB&q^9!4`mhT_6ny1^$Fnpv=WVr&1EP z#XR;<4|JIpwyWYZP)h_3?9D)NIB4CbF1&)TGOye049=mIw_hg|vzEh0W~LO9gDk5n zl8{v1!m^3TQkyKB0m+u!R^*{_+9t5!W(N}RaJ7Te=&Yhy>fF3w)6pgSiQ2Vxwrmge zzOGe`k;S)BHc@5HB1=@xj(HX$1ITW1^7^E{*l_}kK!(Rd@+oz}%yEJ**;d+`Wf_GZ zaGBoZ{zs79%$ZFqiMRM7=877bgO9ooh}M~-E>@+^wDI_(noQ*EPF?*AzI_Jw@~TUN z&3mXJ-@&Ur%aGUwQ z*O@QjDwF99_Zk!S(euUN-EqXUcYkq^;!#-azaGGD*AK3JhtA2oGRxxPvqjciCXs~E zMK#fxyo09JohoSE5;%xPTNPA-YtDBi%>3qX5zQ}wgnp6Q6VpIQP{d@o;^L&y_fJUS zxR;asfK@HiTZ_f`e5X5S#c@}mrzseNh}!7=OCmNkX%5nAiU&1KEGgL^ zX?Jf*%@WpZxxO_hF!N?VE;C=V$XI>H!J@ItZlRBAuajoL*TtV2P*b78s~~B!6zhd; zTXok4zzR6F&*jB`rv?M1=&>gA=u$DchROs|*O}b)$wcBz<@@whYTec{ zi_u1csZBbg9CzIJqc#x$Ck667R}nB&T=fN8F2n|=1D%-zZ6gY&`zotS&5lq)+0KhB zli%VP(spB@rM-E%X9UM6$LLM4>*~ucSM7~c6;jRo?nfUQa6uUzf5VcI$$5M63jQzyf*A$yWSV6f$l$Xo2Qa!lTcOdYQ z$aAXqX{jXh;Yl9utQQ95@Y4pYs$1;d=Od}3NSs4E{bvFh#4=c4Zn~HtxnN$v{*_DS zy}!^dW5>_$M(1Bzc19RPOa*cV7vJfEq|h}gR|-A3R+vBp_Yx01dwz9Sq}v*iE!R#8 ze>j(l+@DVED%_a}zBe73oT3)s;j-9y2}iehMP=<5<1fUW0si+t0&1GAgY z3&Hp5!77D@+rNjkb-GerGYb)}ROkLxr^ZPduakF5ks3iY$$oB>bL-zb5ZhL?w(2$@FAH(?d;{e*+Cpf=xOC?3A8Help$R@20HK)Nbw_oyPs}S8$Zu|9+r{B)Q2)x zw~Z>&9n#GM>XRf^({c-D{!2qm8QdFm2Igdk--HD4)hJrP9+eO>J4t91md_Oq)<0bN zt_0G{Ku9+<_!Q`HC3mQRXHu(X@wRU>kpdcWrPAr+lm1m90#xQylGbS5A($vCQm|*F zN~tHQ;^5Rbzx+%UjfpDGzJ@x-%0QaiVhT`2R4Pl8S#Qz0Y#sSV$aFhb z#o9G{l%`YGT>=0b2|fe!4O)PSkz!4U-U53G`c9XQbZ|&qP!V>8-)lUBar#EB)bhPo zUI1t^LU8e#kItHQ%S9?o!BecBW7P*9=_{cTH?^vr_f+E2ou=@W*ld1OMVXyJ_9hOR zcCPeLP8u<>RA(INMKjcck3kvpcOoocxRvT5q_^kSx4Ax+A+xlu@6$84Bsu^hewwa~ zc05EFoQG_&mmZtw4^h2I%kyPT&>ysBYWRyTB^+7JQ0dVQ17~oi(+t47vhr7I_`YE3 z^@)`!GuUR4-J6UIi6<32&qZIL#CfQfk3>|NM`ngO%=WO|)dGq%K;g~fENMb4Z8OBq z&q$|s_`zI_;$0l5W6Wv+lWZ(vT-+G=KD^RAkp4f6y;FB4T(oZc#`?_pLr}*DNi@U)dOOT z)kK|&4e3&l*a>fvK6Y@o?IM;f^YHixCVwMIG)uVftemx3UbHUTD_cjjKW9{_r*HvD zu;)9j7c@6<*pq47p-qv&-&RLX*EP#Zkr~N1p&QG9I1htxQ7uC-N|iZ-dffcr(nxr^ z3?D1Z;KGz<^uC00^}HN(ypjp`C0-s=jFv(f`B%QNDE2&=Z;KjbOyt-qJ3QZe-(()D z+M7|3*$Bk_rj#a+M-#E+jJ<7yK=H`ZI5R2iN6N*=e*ZO8d}6m}eBnRZPvy65a&^#} zY~2UjORQhz3=&H#P~o+JZrl{s6Qc?QSOeTdRxQvJ+Iq&#`6c~kZlugbQ|%tDS6E-Y zM9%H&e8T9~y^SS6N!|X@d&!-lNku)9MKX&CNR+>?9BlT|sKx+Zdoy)ghy9T1yL89<;<5&yd7kpV9d8(vq|be%Q-?sX&h@l&e`*nt?y4 z75oD8K$D6$dA=ByceB0xand=3RKb4NxP856#+B6(pn$>w!C)^aLmVBCNk&0>E)=rg zgUl~UxbNhJFGu`JU?Q@P^~H2CWF%PLY>c9%baX8&?@gWx#gsO~)YCvs*Pxk8EB;r8 zyYrS>Vu=VB=(`u)UK&5dlSk=)hDVbUz9Q`(sLX?#*>zKz!S@yFIn{+oZr*9;$#|Aw zy2;(s{=SU$TQ5_ueXjBislEJ0GK!Gm4d`SZ;2wYFR6 z!=FIE{n$w&)}m$;L~Q#pG# zgWDpWsbrX-C_>_4ATd`VfGG^O*sp!l0TRch^?lWhpxZ5Jw;w2*KBw1?E7~NYa|mrU zN)c)pBx}F@h1alN6xLh3gS*;G$Y68(F9z%IGFQ`<&m|{vy{p!0)KdP@>=tFW2P7#` zV$Ha?xkE;x-xPcQzbe!L&OJBraHhxf|CKaQS`oF(k|PN8eSjY_rI+TjuPRYci#u#xPpG1^?C z)IYTjrIg+WbQMA_zbifJNO)DS8W)3xHe5~FD~g_JqUa<_H3VM!x#DntVpT>;l0lm^ zsmD`)<&!?FAe;%UiY>h=lKe5H>tx?_I^#{I3i}6AF0zEqr9sALdV{4V$K)s23z=Gt zq>gm?jl}Ns4*6`E*>MH^r(|~zCpO50fZUa}!LD>_eidSAGB|DjlJ#z(W2*Cvy-#-E)KgE5zLLexx$)-H&NYdcyQhUP?wsLfM z_vQN(AXNTyka5;39hcY?#~#yO0T-B!VEnfMmIDXEmXLjiGwC@li06CEuMwXl`vZyl zSd~q&_E&EXPMy}oHMu#Qnn+)>#MKzyeTGHs>-}FlDhJvVacb-q2e}%@bY2mTzRqO` z#iw}0a|4l;y7e#uHU1oB#JIkS&RstR$11mX>^RBn3q926gqyW{r=r)go&sKiNEPVa zsZ#PV0A609o?9$8ESQ?rh{v0xbC|oblYfc<T+T`s`K^{;c3(5@a;~Wbyf9jY&YodU7N_+(>#rU z#+abp?^03~2a594;kDNo_fz; z$H=xBj=GStoY|1{O8HJ+GASW;)}D9hn#kNAWr==A%sd_%4#?dv~1LMv))EXitZ@_zy-I2p-gNO~xVhq5a`rs|;mYai3VLgMD z@5Ovi_RPTf{+W{?_#}M0RF#pA6K+wHQC^D5cVMc|#E+&P0x!tNz@85WAOh zUJYehYRZn^4^jC1871*iNdpWec-$2_cR4zv#4Gbq7HBE{SA;F$>`}q-Cxi^LYSz53 zD$OM{w|x*O0U=_L_+MQ-`_IY0h;&p^08gx$YSeNx+dAyjz_OI@<`&h~LPMR;-S6L; zi_m84v)b6S0<<}n8BSXI9gw)hU4MyR%hZF_`nTQvj}dQCjcZgDf=;GpciY z-S{V|Y2wGg_U zfJ`OM)D4Jo0o4>S_^Bl8xKu7|Ar+_^`N8r@K#TTK7QqRl|p_a zWpz*H3`TpJ?uD8{sKQDagGY+eV${vKd-zRxf{J6AN`dg*{9&Awd-~ziWHm@?jNvbG zG&>JFU+ize!432k)XZ~p6`0iN?MJV;KA2knPs0)Vfj=rXz|AG$%r{7c7lnL#R?HWO zkUuJ|n#g47g)^C$VYT%db}B8^&|qK_d{D`X|E6VT2=AOtF#POOrr3+m4#&xpr2m*} zJBAw2%ZQ`+%np!?5T_=U`s+@$X}^{R9}y(P>aa{KtUzS`1C=$)M%>z3LyM>%?r44;~?XtBu~lm36A&GG6h46c7HySWDnWY4g*hjjfy0FoZa8 znnh*ga+HSS(aP6vqB&$YhK08FfS|(yvYJ>ra|e>9xH2f#66gzS3)$SW7V4^+F(I!& zvuY7pRuX|153^lIC*Xg)g7i5y zQ!j;4p-mWY>ZRu4f9(~8tVQh*FCdIKAp+ma?>&3Y{O!_;D&wnyz=&88VK{vB*t>pM_ZVceePeYRH6U!nqRNt zu3HO5a1e#^T8hG@rTtM}AVkV?`j}b%5HV2t$+b&{XPw;}kju*QdKqI8SiZV%Kd&}E zVJawLY|))IMUF`7cx!>%seA-)CPOrZd(LF-Ov*Pds@XvqWJDCD^PJ2kagoa3aSr%n787R#?1J3{hg5*KGPe3Z+nbY{-U;73$a z2Wcf>yKq>9f}%juExS`%PhU6loV*ZMxNt8kMfh23j(8Ko2Y2v_Fv@M~)4J{nS6C>m zltzLl=Qu&!yydLhGwYVUX_cG>PgzB~Cq1*9+b?1TOKOLr0UBTaBnqp``HC_1z|3;$ zE|A(M^I&0%@PjH{%Qgsx0~6Ot5eh6q6<@c#(en41DiG;qCgk7Rz90YrgHdD?pKuvQ zG)DBBa`Keg301vXGkuBDF{zJoOStyr=INkp(udC@8uI|}Kb_{3!g$zJj@Lt!#lk2b zV&~!hT>?HdPTQeNe5lbAOAgPO7NyRkP4r~0uuxyr^>zwZuu&ebU?`E7IC$?L57|h! z<+Q6rO+t{_a=o<^Z;mfa28Jef(1zAk4+9?E;*OVBCTOLlOQBf9UD<>M=3j^eE7I4p zvxG5stonRAeweP~u|e5xzT&34xSj$4W<`8)2kS~}c6tMeyf zryweUf+%_)z7JLkvHnAIl9NpK2DywP%r{1IAu&0D;yNmlud9rLjm&}YRAJtXF44#A zAZ_UYr%2mN)l>%00G{EE$6|U@<#$Pez~N1i@!Q0a8QrQf_3sZ!QbYSx9O-y0P|cs4 zJpPjZ*v6BzZzp7f9B~zfMq?i&_4-yz7*`nVkLipqMwd3&DUYcX9^BYV3T!S|@t-Yw za=SF&Oj1pV=a>yw^;usU?D& z`Dz=k&g+@tWJopr8j+3y6RH(uHp14&;vi&fL(u5-x0&6vZQ!mm^n=duo^_kF=do38 z2>=gHx=A8urJjpKP86Ep^E1f#)buAwJ&C@{oVNmNcfS0NfXp(Rc~k!KK=u1M>$W4h1u!Tu@jE1?RvGnXEouiBL9MboPOt z05RzEIwB-n40idQRYCRhG7|aNOraV(K>WN^%9mcRl8IJ#j~*(GV^ecq|M4O>tgs;t z2kfeu#l&_OGO{5;Gv;v0TWD!S0d&^ad?d4_I1IOaLrlj}%tGdCwiN8HPy=Km)T0`f zY_|28Wv@*l6!l8F1fD>^oNZddJzb;Fo9rGve7fg#;cRMz-pY^s2pT=<~5ssu~lDweKBZ9J#ZS z4@84cGMvF(SN%IPm{hl-=47x%7@;;N(*?zEvS|rEEc)HDIicLEDyV32F~1<^GqkLj z@_rvAX~D#23G22L?u`klJdJg?u%rN@Yxqf<-^tO<;!Gz(S0Q#5)`+fg_n1$TUcNjGR+8G{TH{qdI6fda)FxY@SPjQYG2uhBk1DN24w^4CoVGVVimjkpa zO71$jveKe;KaNjFdI{hPoR!*u=_|08CHf*`M7a$=@bp5?$o=@>LGBB3f!Tz%5MEl< zWu5p7KBTt3G{{{gY4|_2GaT%u zL!DQm+82-nme3H0w*Xxl7bbg(=dkKZf)TyDAn2Uy%g`}Z5Qs$XxnqV2^pe~1DOtp* ztFWkuvJvmX9V(N5i{t)S&+OK4jQy4u%6}#bo0_a&0*CU5s_is3Lbs=gyD!^esk{9*QZ2@_M&JBr$|X^B z*%E!>KUT+WQ;FU3Y4vklw>i0NYL)_C>Xhi1#&)ver;IU3?j|+5iHX7;G5G4fdxXsN zR|czzry=PzCjKWmXU*_7{)AosC1Mi$$Q7fSU1JiU);vf+)(lU_F7h}1k^N9b;e|oY zU~iWtAm*|;efGmwyG$_?)mC+d7N~B8irku3q$-%gLZSic(5|A`sqvrRZ3nOhbh9Bi zcRx0=MP=);oSrF2-<6mkzh5$PXArOb_s(eNnj7Jnqw*5pR^eEcMv->BK|?}#5Zzg= z2jvvhOl9%ia5PaIu#Mv?rA)cE71o;2z#VNKS=CP%q|VR(SkikB2;sfq z7R)&&o8t>{V!6IsDIb8V)7xj?MJ0+CJ`Ul)j2jIRL-9Xluh4eY?FcH@)#@N7Lg|6p zK(rG>z@76Ic-R5r#CTylt`h3(;wb&L$uLd9um=}5WKVAsUr?)i>-Tg4U-2%kdosd; z$olUg7w!x3fwxJtY1dfH76B0l+$heAME}&e?cB_$z;|G;)a2cj49bF=>;KV> zhB}u@#!;U%LkHt0Z3z}-aLq@SOm4+ETT>1hGHrhY7fO6Z5H&!vxa!uSm)_zaP6pJN z7U=k9u;N@~;xU7wJz_9xAu14&s^eYc&79dK{h>R~-3cFNB}I<^2+Fp*Tyf0yZWA3lo~!+M|D>A+(7Sl&qU?Mm>+z zjuu*qXmGb)Eo|h^@CB_H*s;0FzXT41` zOoF4-^zqsPba>@E_l0nV8?i%4m0;%dN3fC_AwN87J~+-czgu#0iX>-W1cEg5X-;bpI) z$PcMHT|t(agQ3D!`AZZ0OC#L9!e+){a@*R%)9Qu`@3A+%m9*2m2ORU8@Tj{C6MGd( zsIw89{Ed0+9#|YPAouUzga%F}VGwCgdxA?G9jM343Ej!6TnN43#&3# zVz$wgJc#~o>`aVcRAFaDYvixzTX{p`PA7DFyDv#KqY2T95|BWY8z6tM(&V3+fg=0H zm?a;c_L19JK7E({^RB}I{ik21T{2=Y4Wz^e2iEvL!M^<(6Y|we-!l$s8@N7IA|Dc7 zzqn2jN2ii-`tB=^Ju}LaU8dq0XZ|(UWf{*)E**AYQ_f>={kuEm zZ{1h(*pDuiHmIOD&`w4>x-%n=)mcYr`E{uMldYN{P==e=$0Dk5GDsz@*^bnlYnJ6c4`nFyG{6foYXskff?ulTgHn!{Yw#ViHO(-LLt#)*C%M(L)Q zTZvO4lZWlL6Zl6R+b*fALX?L-fFNQdbxPB9Cf(VkI7U388+43%sAM z9L;EYBda(iQj#xOlB_+O-CPTZ{3PoY!70;soKpAX0Ef~sg!^7L z!fVX9ZU`30c+_=YkM~E>vDA!7m|-r$YzPnjn@UyQj?Yc(66IKhnXb9MC?W$nQAw@B z;Jf2~U`GW7f-L}$Ord9W4YxC`xJrRfkk(9CzT`XZ%k|{@EslC%t6Cs zH8{Tr5X-T%kyLd_Pa%IO7XPov3@;bNv0)g4sfeY5EDwU^HX|TjQ-Pg^sS6B;bH|ID zKgE`c-OuE_5pJ%R6kUShwC6Od_u0YhNSpG7N*vp|dHV%<*qN-oBV67}6lNBF znh%Du@c{l+;C96A$lWv#l79Ey*iTD!x)c5jzrD#b_p}@I-1(}5$EjFFM$^0>Zz?Ui z`md{NOB`RQipt&Y%>s7U2xrkW7f5-<`3^ndB3k;Tqwc*P7rleHqrYS5N*I4yx2qRJ6JF@l!d;^h}OzZAbuVe^2&f*?Wc9@i?(7_&tei(}6Qr<9JSGgPYusIK@Od!`RP=|XwE0vsBIaFbW~xY3ul`n z&t`oN8N|LGNc^hnT}B^S4Q~f`HIIG6ks?pPJYCEC=e;~<56z^P!$B-!g1~Y?3o;Iqtfo$&x0w&VGcw^T+6&!tgVSD^G|Bb0kEDS#A@gO%5 z51O7c2p6mQY9==-G)BZuZM3DA$_8su3a7D$-rJNrO3E3Gs^PKb*`_c;i8OVM4X7ea zpG9*sW}sGd`fHVj{^ob!MG@BJXOHXN-0=@xO3*UdpKh8ynW-+{Df%^6t{%x#1{j6~eB}nzH2P}~O6BzLO*93| z#c#Bz%A;4L8zX$ejrc*DTh8=>Zem@4;476hY2YT5xm5T@YT(<>Bje}8L%8-7W=>P9 zQu+m%ffvloe(KFiP0&>5tDNZvlJCML$3qlaGx(AV2;aTEM{-));hsk~#;g78I=Q+4NT$7bdWC_(bsTQaAiZCQn&tX_&;2t#$FfHaX+D0Atky44Ott{ z8$;kH(jy!3Fh6X9i!F9N7TqExPTC~QIfmhWRHNZV%b>>rrn z3+mt^?GCfI(fMd`9SM$~5!rGttycb+f%9PYyRYbA#w-i6R?P#W%gtyT8(!olG?<_R z!%<8oy)=V^AP}5rX-1HhQbO08GJYO}T4aK`$J;e#rt%lb;XkD55|{}dXq5I!(tUm# z3}7L&+%;}O+qAEX^>+bYoVQPuPQVQ$&n=gCZ*@t>OnGrsn;j?BaFv9qTD3o-&l-`# zKNtaB|9v_*J7r*Qr03!!7gj61`5eMKf{aDS>j}o1IRQi(R?=%B!VMZ!Afx!sC~L_& zgRap|X?tOFS4Kl?19@q+fJP9ge?b1)&mkDyD5oCR2VwMxFI9|&L&FZmiE=rjJA3hV zV+m{o-y7TxoX*gZ1**H}4Gak@t(|Sones-=x4N{K4urszFHs@ON-!i|W~rd2P`Etx zq_Xlmhkueud^hKfY08L!{63gKvz`0hEQgX58hv(vEWd>);Zc;+^8&IuSTAXRDC^ge zRGM#n#%$Lb>*AMUV`WoeDPT_>vrq){HIY<=!Je6R|DA|ZW>dqAV|C_cRW5c7q$>-B z=+%XfuYss~v2%wXzLaFv#Ds`e!LZz-dkN|l56UlDGo0LS2B=y5kobj8wAjwcotwnN zxK??saNa793`~r2tacxHM{McGj6dbBzD#Ud)7O!0pxS;V%gl{%0h(0ist1Ez>)7E; zHr%}b*x}ZASwau=fLZ1a;h}lCfdak!Vt*`jrxu5!>fz=X-$qm0*cN64&JVejp@Tf@ zvVZb#T#@=Cv(0Gsez%+W{u@=|t&AEfT|@Ux=HnOSgzfnN48FeGE;nslt5P$-ahl#2 ztRW6csTu)RKIWiCd=Z2R2N$uk^WVKOsIc$J=K$_imR0}kC#yKFIh|_pyfkg@#nxsJ z6h`bVxr(aR8tHHT%3rP1>#1Bt8;1$pMS3~3i=nn3F{TsQh=3d6kPV>;Kz}obep$C; zTxDFX|qlK;81PWsDGF8#@?3J5uj2!w7m|?tM^(4qcD%Q0n;b~1zb6UKosU-)M z?GCBSd~N%smapb4u-LA~Xp($rHi=L1=21~!27vr3>b?U6k?YM^Cyo61VJTTw&MinR z75ppWy|-PEDs*$R71!S5Myc-o!(H_hN48Ca%eKf6XXzG-2ez zHDNPB^>xtnEEbVF{Zk-pR}_g_cKFgg6dxXc+q~1epGW88)+P5gaGEA#`K2&8-LA}%q#oa}jP~Fmi!n0!fH@-$WSH26CHC{`ph zP;2W*qDCvc7WaKOo>!wkM?t)FdH!M=w)>zXZ}6wBc;5(K>E+l`J&TVe@WVy93Y*fRz-ar16!mhO z<|jad)zLn*40816LHdLX(-3Q%bmJ)1GVk26v||62mB&*EtNo zfO~jS@1LPwr+7kA_=5ATf&&4`uKRHt;=zXY34%Js2UOt^@9o2smtr^>5^&=J-#hLH z^OxhHc`TilnZohDJ*7Zd;h$A+Bu6n>PkVyppodRx>F6Yus>hl6SINH^w`4*kVe_u2 zeerNi|7?Gw##R>`dbwn?$LosR+Hi*?Y-XIn*&#TSdz-J8M@K*x8mW(Ls`8gnXvF=F zW7!U`gdU;}^#KzsX&={};In3pWqj|FyaNCwF1x=J~&?|LJyNV`t&{KWZ0ZSSAT; zJ6CgOVkQYYV^?!Ab5jR1b65cZSQl4kb7Om0&y8kvFFv)84+mxI_YIk``zv653JxTh zY(+nDOmw}$MFx*)SfewL@j)b4IE4_@i_8 zr>E8{m*@1MEC2MZYxct{Z??a(Q^|=O5+8nJ06BuJb41*!+CSP7VBQZr z0MJ6uqwOIC z2okg(EQGNz$<@=#0WiiQ6f~C#yDp&=kcfaFvzv-CB^2<>B7;|j{^{{Vn&67~dEwkG z*X=*ceseN8uuAOJh=GKGnEx+C<-vB*xM>+Y!ibX$nt2^xY07}O1_v*zURi1;+iqOV zBL{862RRP3=v>H?95(X~^OS6zI}?o3VgvxvJ>)ABhP}d^9^HUxG<3gSr<{wN!L;ym zN{AG^Cq;eI%gwxw986OKAv(7B60o)};+Q_W{n<5uyaEc(Ng%9CL_jnp_biW$ z203LhO`Ys8w^};KGE}5m5*^0~7RF5)0K&@hKEoW;5;5~*aqqW2sWGJX3Um2`Sq`X>bj$kIU19>u2d~GKFdQR4}i?GPG|59 zk&O!b;PL0L9iSohp~T4h_0P4GgbM?O0NAr(|4sVzGE4p?IYBtv-1Oo7vgka?-?Sdd zho3iw#DfKakE8^|&Xz?oLt4dSqecqBgAf29u{jpS0t5(yP}sruU=q0lU>;ZwMDS2R zzKjHP(2(~7L4QJUK;HRqe=YSwqQD4(35H<%i+_{h%->=i@tR2q zNkfO|T~A2{_ha62v(N8R2O2$sMFpG*MHl2lxS@-kEJCi`sD6kxrDEV?1hcV2I)+(LS_#KQj`VFbhdBv` zE?CYEXI6;v?r-Y+p#~wwhFD-BKr-$SSFuuF4+Ew zR)$mZgS`)76xF8n$c7*|_j3B{@P$%{s;gdllE={ddrj5plx)!LkWGh=!azJr&W7h8 z@?|Yc)~l=Ju`ukKE}RTPKzzi*<-&dS3-8v-%idnY-Ik zDku5WB~5}qO}z(r-*z>iXXY*Er-^$-TVeXt4Gg-1Wk7`U(qtTekNf*8?K+wVo9maN zktOS=k_L0?oum8>4Y5HQ27gNTLg|2MRe|ihyFH0HufMp$J4tH2L=i(u5Odi~Z;bp- ze9uaXo!7``ck5D_(cvG;?!)c80QOk6{>dQ5nG0{{$%E+P9V+?ErXeTE4lG-F*DXh& z$k|d@is+^sUr(3JFF3P2MWkq=E8H5p#qv8u%Dj1=x;~S{g~hsl&o?FB^Fd2h{L#9& zh4JN^uFHIjUHK0+J4zm-w(2 zE*)BBR_xa9OO~cN`8g)8Rjl)Ug-)cB$i-Wk#u`PurzdPR>zr;*^CcFbK44jxrVMMd zIxN1eIT~v$VdiZwO&`!RxGcZc&_VL!Exs!59uWVDBL7Cp&g}P-dmcgl{da1Bo+fFWDOrAMv3ngg+CvTQo+UQLg zjHyH0*gPX{9O5yM?~c%NUA2UucZ{!3N8_S+Zt`F?8){3rT=tBGeI4^Nn3f6joJ1eE zM>Q?+(YWx(jlqMrRL{oe&`C`%=eT!IXv7j6UpByWtE6tM%PI8OV|w^c?s!f{JEL5p zgu+((b|fs}I}*nWtBLpd1U-?v88!$)=VuXkEpW`pfi(@c4sS^%KU_CEKDB#Y<($2{ zbUP58`7xxHeLMBadI^1;O!SDFK0)B7uBDWEBEa-;h6q+Bb4BLhuD_1`>2z~P5@6WO z_1oVMXUD1em)v%T&$m|Npu(A*Sh@WC#VOMBFc-6t$7~Fl`7cn$8#Ctz3t7C~)NE+A zLFe#u?wfl#%02J<`ihzK=(vA0L2(EzLrypbE8V#zrt45{9=YQ_R)Ot2t(_(V;hG1{ z#lzD!N<1Ajp>I9SIRm=Hf?XM%k5`6dlWs_3XabFn=L=UjfAgx+gP@kuPdC~c1jRbG zaVq&Pl6&qn=Ab&7tUXlx=D)%T=xRnF$MpQ67;~p|{Qd8MN#?-alfh&4>XU552*m4| zjB-D7xS2>9gP!8tKVrIqMm0-OMF^HO;~phu>`zeU<#;!h>*T)L?rm$#zqrdOM6QGl zP-Z6C(4956!%D+xcyKTCRKNjQ&kFb<)EmEf*F>X}Ca(MO{|f zDI>n#s>BtphG!JTZzrEj*0KU!q{>;|Ik|x-a<%-cw)Q{Pt}`?om)~wlL+yiL_WV>w zS-C|`E#zt&On<84V7jIcrDl@b=Dnv%y>y$XAKOox25Z&dfiZeQOW@7BqmK^WaHd)MsBy0F9>9(*L8zm+#>`Q-> zp1lNafde0lG>s2}0*rDVSZu0}mrghy-(m}Zv zJCSJ?I|UxKubvvwx*SAlAYkxBUv-y>f_Xa zUlJiDa8ty}E;{1u4&c}Req6>bP~CNMz~Zr4%!^Y{mQ`FiglPd$SH~g?Lc?<~T0$IM z<><@0XL0E$jiqTV(|Bkk!dznSkdW$>$$0~6?{`>PefZd|2Ue9NyFp^CGeI3pK|7dD z2q61YeH!zS>!P53;M zIq!WxpQB}3kZ&T4tb}9ja@j;)O^_nAMiXX^j@geRK6?o?~WVun%j){2X0N7I=fdhO-qmt9-F^el9v*0r zctD5eQTwm4IgM#IuX6erRT?#1_keDU%SItRP2E6YL1|-A@Ads*TX<3*=4eV1#nIt? zC2crs_Q$Peg|(Nku?zlNGu@@Ap3C9yLQwd*cT#kMSLeM{ThV(?waBHVXZ@>R(MQxnb1-_kl4~hTg_Tm&q1KhQ%LDh}Tq&&PKlpfnm`hLk{+rz>Nh(y{fU>_gy`IkCNRP)G6=yb3-#1C9H$u1vsZ< z-i;p4MJctdWl>Hv$v_~L8-Kc`R$KS+pT1|iwW)!X8O`MwrE3MG^Hl8m$sSFNpSuf5S!?HRZS;BUXNl^+QZ$Sx$Qi4_uNRVaL|BuOIdM^+b| z&I=8L3Hn z_Mqk41ie!HL|I)F@sm)THxYf zK7llW#q%LSc0BTLqsGh5sC)oXLw}M}12JL~2wm3%Hv+9) za8s8@m#0Ti3k9!P0U(2V1}9RVsc-r_9riK*=k7#EK933U`LqG-1p()wtwBJDh=_pE zSg;5oWcmTnFNU9T#7gLWP;stcMtM?oysxW9T}BESPK6|UyAU<*P(Uy@@Ij@$zvLXR z4rNdlDH-H31YcMzQ^a$c9NYfkU`j@}f*V3Y<%aFQAR&JNkdV>CAP0q*CFO{5L!L!I zF8$Y;9qQ0d9fFJk8n95kmPKEB8IKI@zh(7d;Mb5G3DWK9K@D`}YLu0)CF&!Ul43YTf(C z;Yf^NgT1{2In=+CgIQ3%?Oi~MK!N~uRDjD1Xp}aaFIgT-B^HJMWDohfvIfQ_46N8Mn?>&X;yoU} zEhI4GVJMcTNx?Crlc*2oQjQu3C22a;|MYV@_4PMBE4{Oma13nf$sg%?cb*uj{J-J^ z%14b$pIqdEnFlhg!oXXzmHt4 zLN>3&O2%IpZ?|^+x8a}TLF$!x1W1VjKw(g?p3#)q_b_22f#wznsD}6XxsFcPkEmR4 zkh*c4BNg=&1Tj4&B7h{3SWrYbMkHi-2mm62z0Vuy>n{yW3<&9;&p-zZz!88uhLOS4 zwrJ~t5(EF3zsD8>Kr8)8U}6+sn8<11k(uA|A_3qrKO!m_Qgen<*o{A+h#7S6Owr}k zLVYMMG2boFFt8xteg0+8#+AV+)aAVRl#i0gBj=w`Ln(ouAj1IAIYA(|P~Q_$#QxDU zR0=qB+t2(BY!DxQ&Nq;p92z;`3xs7F#h(d8l>kISWM9Ew%;=T)CjeWZyjK_6w2QiY zk|-K;SV^QkSjoRi7+Mwciwk{3>GliEB<$z~?pu`vQWNW}_Cx5_Vf^zn{TPc=f^7e{ zkxMz+iN{ksa!gEulA*2elriSO8cXeuU*}6ayJYF%4c+%DN09Qf|K0KzEFjpoKQ1Ks&`~lIs-~hD$q?f231Uyr3?sxAeL-U5 zoXkM(6t3A8gtzTy<~$cp}v5d_oUa4n8bhosqxLl^@IMZPVjoU~<6#DYX_? zMxvT}J<}+>6>eI^$J_S>rpC2fbxwL(XyIUHtaR)YV6g|myrOg^*>gWoHmVz4BYQQL zjj872C?0lMnPl7H-H+1J%Hor6Ehe^yuRtaSteH;XK4z`=kJMtB8`%7cOnV4HI+H)) zWOE8lTNoNty>ypdI>j-=%hBwd0a=jrGyaM7%>9e)1geUO@eOk~2xAlW&{eDbei_f= zE?2y|0uS2#=IC>2U&r8(@XAG7B^TRVZ10ksmaR!f7Zb7@-IvJ1X|!B?*4vT@Ti>R? zFYQ>+DIBlfWH}poHiyQw#_3dYz6*?>90pyP>jOfnOcJsn;SM3`?l@Z6g`GZTFqUQB z0uK3%$q<@TB!4A@$Kh=U(+!V*ZiZ_W8{9tm-xtCjv*#Ut%h#Bq7BL=DWhtm>?d_q_ zW$v5RmiH=J{I~V$=tlVhV8c~}yv)RtkH>k&Mz|>_>Pw<0PkPd%CW)kXk&CTprPu^&}s++sm2QDm{&|i$~M$u z5WC|qHuH^RInkn+JegOoqFL?0IwP!>5E`tWHLeR}U~K!ma*BOLik4h1zYd>8g-YSUc}Qy+d#>TcPUJuw^$>B@A!Dd1o89 zbV#vcZ~AJS)OX|U4Fs1V{s`u48+~2$V=x|eGgGq}&6v;gPyWsd2VtSjoPm`TTM!>5 zk$|vTjC;6l(M_4oKtqna7%5;FwFxCVIej)+jTN0%IC%Lh5!6@~EJ3U1^IC~j#oojv z)?P2eql#K=Jf<}@D-Xj0?{B0&r;f&j*r$u2YsPjQOe>;WTavzcT;1>$#1?Kg_d2OP7s!&Q7L;?0Td?sDVfQkb0l zBv!%E+;}jeE8z-FR3K*Kv6`YUAu#zZ;YG9dcx#&0Y0YB6`Ni;vez7t=`3* zYPlb2FI!kdS~FAdvoMYSobcOKkpkL(Nst(&^fU!YT0{Ww>}08lpnwh@!rzu#C6hj5 z>T>%0Sx7kPFk5{^b0aJ#T7LnQF&qhl#rJY6oBbLI6?t<(YRpVuQ$Fsie$v4CatT}> zPsC3QZ;tS{lFKaCgD6v}@H(CZ&e?>%G);AGVQP#azUyeV|Ew~nP1BL-B}EvXR4yBt zzMU|+menphk2w+)ZRr!$l_cGSnBL71JT(J26t602lNqsxhozQV1oM_d|E{?e2Uk4! zt_}jJtzW)rvT_eS;dOtRHsEJcw*NKWnEr_uT*c5T5bHg%U{x9u%}_vIe04>316>vR zS+5m~c|OUF>1#^K~4tuYnMJ$br3M01k_VK8e4E~!gpGvX_Bp};`e z7_p>b)b`o%ksVV=(J)7>3}HLZH1#A{hTwXJB`<5#Td zwq$wa`cfe!dqD!&gsIO%E~1geyKL~8&Ra0=>c&A4f{CP0!#nt{Kf-eJ zS4Hgo3hDmKlcoGdCXYbaFQXZx^9^;CZR5aqK$7`&hnuX!6Yod19p;#7$6CjZ3!iF; zek`&Joa0p?!-Ku8_*z`=d7sndC-iMsZk#t&T{o0qou65-D`^n#@R_MFNqu3dBRH>XO z;(?1EtXsVN!MCuIS1Fe3`P|4PkE($J-#^XCPCrI1jv9RN*|jFY8`;JCh)GdQ-O!)J z6eRSmK&qg6NoE>cjc9%M4cxQ$(llen!AV*5byY%Ff2ND;TlVVcV*8B%CM)r7~*vfGb2bCPg3(+l|qS= zMnnx>IWJG36CXqoQ=H5zAVexIkgvc%;*^b93wTmd`iiyy6&aWpr;|xHCLX#aleLV1BV<|GG#$P<8!%w@C>QWe}y}p@?5`D>JSe<{ZmgnOlShamt z9vlkFU+*(BKp$Y@Eu;>A4rW$%wE9~2W>4PE<27i)$5~YR6Ak+1$8La2$NHjl%sQEP z4}TOCpU~@EA#CQhW}rbstb`Ka#&vT{kd~s7qfjPcT9nlDrtac@Ycbp9>Hu4eJq7!| zU@AJSh#R!M-5>0KAs|ta;&W$kfbBFAp)I`qot{Z}xEUqi2b~eD@bIQq7c@~A!yjbw zu@EaOU6U9gjPcbt9f7QhM6XNTrkz{ArV2qTm0ITeAn_)g1da{on1wPaf4xFei78|m zJf*6mRWJJBQONqkj^{`MLUK4Fd&1KRWOClDiVX~x4aGlATzMmcQ zFkPSBI5;`TLVs%8OHwnu=KV$p96kTACq!X6Mt$g0*cl%ixm-{K)^hSypB$`Fr~!p+ zH9(_jk%Q#YWTV~;60a96#wM8rxZCEX5?%TXH}qQ*au;zqbW9Ex+e#YG(Pdj0hmub2 z>+#CFt6A`l!0}e#V|X=lCg%NMc4J{BcHQ;S0ZnmxF3=Mjse3CoVnnDDx$%~#wMVed zah~pD9m6uZt@XxqtEZ;izpvwfA7kJ*H0S=&T#@H{ATxXf%cL0iTv|U}=OxNm8Rm22 zMgY?lB#Xw^vvNgIy2V>QYG+G)h~Ive(N2M$ql4XV>u-l`eF~g2GLZ0bGU)eJK-rsh zoU2M#FxXy2W5ctuA*L7^P0mW^`4b-=;VqoAcxJG;gFrr{S~l-k0y{4W0|9LYnPuL( z%4JNlt469@FyMPyqWGvkNYlJcjT^XT`v)R|#{ity9_h9Lj9O0?iZ?Qo# z!Y9-X!-j(sI^Pr$hh^tM=z8#V-k)jW6IPRydS1mrG^Fb}xZbx2CacwRgHO%W08J1` zH-2h-jan|NMI79o^@sPChnm=%$gBId7raciP$KjUM$WZ}6QayY>#S~a#;6K<+a;Xr0B1@S=d=4WlRw1#m8X9ATk zr@X%)mnn~Kjc`iFbY*%ue3zc*uKqBj&>Oj{W0f&ag#S*ZCGpd&YdK7TRo_;Zn>yBsc-xdg z*9kof3v$+Q7+>{$rnarT{>_nt^P)XxJ6r;EBvE?l+VDk&y!SqYEjjEKb8L~)J;25n z**^c`QgEnUH@uGTW=*Y3xw4^?|1cJ(RMRQDoZ5VWnI^i28$P*|Q3odk#yhW>Y1oEiO;f2&+J6#`pb3 z?O4u6x{#9LzBlqMasyhzeXbB(Ff*5<(_H?`wm?6OMVEmtXmGnp4QvpW8)%qR(>hU*SqhrfFH3tRTGNR@6*+l! zbjJHJ%`Z{e=?ufIU4LoR24AWTabi+T8QPQ%o@JuQcfp>I^lEE&Zo0O7-)vQ-J_)HR zgG1ozuKPhGR-I(FYGew107qLfATD*j{`$38}V2*C2*xH!mw7{f2%j~JpmrdkZ zWRvY*Mbsi82>+C;2L8RI5EEdnt{LK%obs`5Duh3S{jXCO{?&TK{#!_IYe^Vpf;rM> z;WDG&=>bR&CNQXjl(!aVr?|KiEw-J%Igu^PFp)s5(frpvQ3YcAB&pw2X)Ey$)M$s=LJ4 zae)tWY+Ov6ZilxK-)iMjAFkN->Z=%6mc**ZiBY%#ffQMpf`l}Rm261p{XqxL*#wq_ zw#%J;13gof-eAq=0M4^Cdh=-ucm|6qVI)EtmPYb(CFzzeU#Ajk@mC;d!<47t8DbKH>ijyV&bU|OLUTQDD9rZ zR6%(09d>=d!yxbF%9wGO1s2cuB(r@@TDQrOeQ4WAho!<+Jk>fQ0eGkFBwBrjYxAwl zJJQ*cBikjyDZRw8y##>OoD0|!0qFePyd{(3Pjs;v*Hkuw?WEj3_TB)|F^2RjfUy9$ zAy-x{al1wTP%z7P;%9^rX*z{7XQ6Sb&K>S#Phi`_&09=V9nGjcvI{Tx_1Y81kz1Ow zd*xHA5&CeNsYBtw0&_2tto}S@HY?I*kPrIq$Wk7@$k=_k)5Tg5;u8n~%oa!CSO}W; zWlhA0uxHOGE#%NRBHwtMGyAMt7$eG=(LI7w2A{{%e8QL7WPVKX%ojPqPQ~T6% ztM8E)NbjcWjLdggUdtU_wEbi2Jvm=+yKFM7R9sEP(RHarOWtJMaczSp#P&P)pcQsV zopVDySU1AoRqmCrdy^GAm&>-W%svK{dPEKEiX)Y z6}Iu5)o}dY&-kfwN|nsuM;F#Zv_?P(yuIa|%%2Z7uBKR4QAF1hg_KeIMcj#xz7Ck5 z`*T0gf(qV2Z`6mLl;3u!H>}lXoaIgv80SmwOTv%lK-`iig@r0iiV{Y;w`su@J5;?vmG@oW9sB> zHt}JE=fBmjQj4bu3iBM3H0u1O=g!~TEE~I0zDt2sH5J59R9GDm@)#4_Ew6qDb+hQo zEUiJ0cAGQj?3wc1cjD}w*`IYQM7ilwG(bQ?s%8l85!+~1NIDZ|UrYJ~*t+xO4$O?! zK{qSM9`5%>Z0g!gFn7^-FEWEKV`&YWlyCKa7s15oF+7i=i|+TBk-`%!o7tJuN<@oO z{#n`5qhmZnwG9D|wz~PV12ouk)fO=gzb$4;mJK3mFM^gpKEY$ST+vQsV55~I)k<)| zU~!0_gYpSmAV~mxL%{i#zxy+|?FbcSxUP_y%FQM(${ti&?$vjEIqYVel1%-sxKiA+ zfBe-6I%I_F{Dk!#a}J++y@53%L(bA*6OxA@J1)6{=S)9qnfg0sdmw?|4svI=cUI_% zj}_4zX}=E7SM8p$l$mth-Zrt-kM=0sPn#9O05 zBpgh6qm&=!)ZYK>-c>?P`3S!(*Vco>V}HfAQxl>a=_Nae^NgcQT6d=@6QxXb{@l#w zo7s!QaGdK{mgV#}Qeabc$KK3VO7)qm@A%H9{8dkX;rak3u2ZYW4U8a=N5Td%dXM1S_Tb${B3U=-NG!Iva=QGmSVEV z18JPoi^s281?>}}-dCzic&(dxq8BE)Moz9$&q6TU{kYbK&v)XARD2Ir?Z#bN7M)d- zt>m)8!BNv(IFm5>)CfA>>0?8f;}(J^cQ4DJ16z4KUyiKi_&)_O_~(g+5{UTud@f?MUO zZ2Q$uJaObIuW{wVdvnyu%Hjfy{O$SEyG;gT&RcR%}^ z3<~yEu@t*Khl~7Aw$XC_SWD81nc}B+$ClFb~ zN|mZywrHo>T{oE~xctw+4$8t-`8vIQ$yQ_5$1A5*^5d$neOPvE-}-;NoFH^73{74& zD4A~jv(~H8k}%5HK%%-*I}#J!#(mEdbnF-98PQjNu8mxbz6~YNQi=;{y6FwB;w10I z=H@GgHc?fdH;=qeO6qd?w!#WALqB?-X4uV$*Cv=NESydE;@m0Itsc``k`}sG4Wn*0 zRPlLkv3Z9JTB<;)wLW;-n-bdUH{OFh`g;BlvS2=_YZ&as?jkGad^`o*tnIyZ2%oS@ zc_O}nRNd4z`c?#_-?*+#$yb=W_ur>2YVl-5BkPyc7M$blk7%Prgu-KI79FTY;cSHG zj>##R@v`13QK=9w=mKmXb5P?l7X9{Vbz!}7i@j=LfadN6Nrb6rA%{N7zOKkd#48*t zZYf{ITyfDWT9Oa&?;@nVf@1jy6khHEwJ_z9Z8-`e2;4&FWS3(?*6(LNvR> z*4C62o$VtO^g`K#JK5@!PjVt& z_KyV4Rzi+O^?pn`X0X9PaB*~oS^5+=Jku=w7@CbHv9Mz7k#HE=h6RCSqe^2cdWdpy z^IYL8=Ee2W?bnYt-y_3eN6+d*cipGg_QvFy?#PQZF@&SA{!)4{nIpdftm;PFF%VBjtSy}r9bPftKHIe4&bDZDt; z7o?C7g1|cMG2I;)G9vy4x7Si2^BP2ZLw66Uo^Aq+Lp*Dct!@VG1#sx(;>GfvmmObc z$_RW8VEh0Gar`66jh`VIa3ioDr9njwpkTgjfU}T%^K_N~1We2bcr-`_ZbFEWu=~6L*EHg+Z1%5lu-iV1nJwhTe2O$$a9yUp8@caBP(jZuX zq~(L@KmbCZuQ~DnitK5}{JE|F?9l)@2$KOCS~#OZ0s@HaXFXqd4VMrCkZ}S({<**> z806393APJ32-mu&t~52%!zs4W0HMLtuIb}pg9Zfk5tPEdvE%$ZN zdS{($v8M-ir@=;0j0FViAB?!j`E&DsO=po=Ec99ifa0ZS{Dn%CQ=~-Ryd#?7Nzriw z!zaOnf#d^xms}{Kg!}2g;79fbtYVJf)(RK_fKG!VbCwgUC{rPyH_)N-{2=n|{zKIW zfyMXDw#2rHLou+#tfVlIz&NCG5W|if2rofVh>S$|{!qg5)>FFycr-ZUNcHX1C`sIZ zS=-Jpstn+Os!g{r{&Dnb3}+d84K@p-s;M!3UF3bC?CgPDfrblnpS#vua@{y0+__vy z@aNx$+J11R5A9v$zt}%Y>F~q`HK(xMi9lbv zaVS})X8II9%y{WNy@b&$lG;PP0sFlfj1mKV}G#;R8Jb8#n{8^rN*ncMTqY-ylR}i9bCe7BqV} z09m^c(mW1A8`7@YA`k+r51WKAzkk^@9_9kcBv*~-f!l&hG%9{UJbYSA_qiW* ze=NaX^Tjl@;e5gtZo?6ERgYXeXX z%xj|Aopxkh!YqZqyOUs9@3Bj3Cy5$SE2W$((h#YTZc-K@*_dgnf+l^0H+WN>;BqJ= z)6-+JxioOhmQCl5N;hy_R4eT7|3=F2paZyh4#$Yaw#CLc_tzz>-_mCUUeS3HIy?HudJjv#eS&E~}cjS;Bu9&pPV<0MDsTT1IrXg zOpyQCi~c}Fem6PYM=@K}j1uLcXhYMY-YOQM=W^FwK%JWye=cj=yWOUj75G{fRSJSX zKN7cN(A6bfbGd6>w`tHUOm?`{O^|4wI)OP-Wx6fjnl9a5<4u=x^1&T{EQvnKMX7M! zq-sd7L+A3u%{#79bT5tEXo|KEO35S9XwH$V4W(n7yn6YSDQ_gL&O{TJEan@scgc@G zNgj4q=kwwnHy068*engM;*ueIdNQ^B{k;!#toYwx5 zV!EUbtM@()0<=}t)4`tPVX2pD?RgHCI-i(vlf0;o)ZoJ+J*4%)@D{ zxv*3d{Q>XFC*uv+v#&)b6S56|!suzdFkk1C`C?+>a7k2mn9Mh;4#a5Xa8ryL)V;{J zSlA;1*@b~r;d5&pIO2=>NX?W z#eMv}>?^K@I&68`aPIVU4XTU|oY|sL+l6FibPFxL05djs;X|89SDk9O&muv8t*46m zMyr+FefT&i{mb$ks<_PsZ)%=>X7*k)qj<;Bdw93qe*4#%SG#kbwAP-^Do75Z95Ysw-!;=jBKEcK}rOjyN>Pc>0MFy_&XR`ca{%Pvs7vIeMvv!MjE3NG<_Ns6t6 zzN+x*si%ICBBbHf7UN zJD|SeWE9L1yxJ=r2aZc4sM%M0o3tAH4a4zKCz*PcDeG=eRP~Xj;%gzuPZb&0R!|QQ zzj~s(lU>PvE)lrNPBwNpUM1Jry6mK<-};$mp!U${JM`0S0`%<1S)HWEeq!4O^Sa%^ zMbhXja`wycb6K0q)0~RsH~8bpo76LDM*Y7krsk~?t6*Xn!bN1nZpSB=kukGvvNx_2a?a5ERyRyI+Z zGBkW$+#D>P#D#}*y{bLGS5qf;lFYJYBp)>yn*Md{NhRmpFU5q%4x$s4OUMtW$9N{l zbgZh+Br42m7&O*6n7~B~0&3OFWPz!z;(WSp4QJO^_spsNBm?pla(COp;3Aqw$Bl?M z!{cf@oNekaIg!b}tl!dhZ6yY0AE>OuqKF=wAGF$ymT5IfUel46vg?@6oH^B$FpTiZ zvAdPow(gHfk-H%}pO)&=G#Bro5kj_)SkZW!Sk!vBRIHEcoxut_G!OO75_lXY?sdkN zdU#0=m)%BkyW`C2$vItK*?lQs*pq&<4{Z*#vGXjle7Tc4N%T0g9ewO^*)WT(I|;Ov zm5z?l?atrQ`}$_1O>+%edXyy->gQB8{f{`vjxrM=0IGkY?h5;E>=DnD(t2-8M|P8O zpUKo+LNt}rHJD&O@1c4DH*G&tsW^^7C#}s&$BlMV9`?Y zDvM&1^tz5)K`2}_3Q{;+8;bsBZhBtoxmm$nVN_PQ7Z|Mg(Y{>gBE6%wWS3mD2Q`_b zHMnSvrf&9F`lxlim*N(L|7)~z@%asvhw8RzSpQKjYtv{U`Ec^AMYr+gZR7B^`{I#J zukvz%E}6uiX0jPD=&jnqJWZce;=&p0_|*A!;p9*j(1zpd-LNoQ400hYJ^349A$KKb z?(Z5mMRt^+b2-dQ+LTFU!!}kSfVz*UN_xQiri4*wZ)oz)}uTKcn= zzJEhAqDIH4`QUUQ!9yCeom>$cTMj!IX=Z$GdATjN8@*A5x%`OD6(%!9dV~8_U(TQV zVp_-79%^3XqVr?0lL*#`cWkne9^j!N^dnpA;yhBQ!gbbjJnAZw=*f)k8&Rc3>*GA= zwi;xrK^Ol$S@jD5i_kOuUowvQKV%#uJInvEK{MhrF|o5U{-=&(VdkL!-$L&H?TT*x zG*M33{osK^UI=s^E$!?qR_f?vo8EDDb90k~0$QN?v#^a16v(W&aKGz2^NPcB`r73K z7L=|w!?N0PY572jM2`kU)7IGv=xVH9|Pyxuuk1&UM&M#l7bE7{vLtO4*)dg7`N zkIsedPca6BKU+8kBoE}pAI%EDNe`h9*38<{3gS-#h{gZMzqBJclapUmMng#?Cl^#v zeS3Ny@#@P@n|O(Z*(4nx(NGdg5g;E8nO{^%=kQex79V0@Hxe?x81Vl01{m@yr$!21nMGx!ftH1-$}VObT+ogS#~EfjzM2o8iJ;E!w~ z*ze>yKmt-CNJHmO7=|#!>MqOmFDjfpAkc4BOiRTiUpo{fB@_U-5_UWlf-w|h7)VEE zM@Kt=^iP8TZva~JFCheocpC@T?@hs*UD~7H##dGcJ4PcTuooVmQrz!0gHn7})T`dc z*K6Z+KR!A(x>}l_3J4Ij%~e3}dgs>;6ZRY*s|W)LhXsWTtSRW&UO+l(KY|NCb!~3# zukddz8EFMRz@E_|VEv;L;2ORVC1_(j02e#IdMEyGs@wutVG`BadermpA8RgzHQ=Mi z-$4Tl;Kur|n(mM+qAJp{k#SgZvUk6I1mp+2EYvyFJ}>}IK>ys6mFiyXA11oq>8svp z1cEPboc>qVDzA^AK$I?EX4S>YITpeg59Bw~24M1JNj3Yfsnwl3e#*Io6iO zzoP?dtOFBlKEz2pz9uw=P0=2`8$EvBgX)`}9=zk>8B!{jCSo0Az3T04SZADb*dbj@Q`RI{|yE7XsS^I(wh3 z2T<4G*H;sV>YJZlhS!I575(Wh;NAjIZP+dKBftfydczxn)=zm0dc#-#!S@BITJl4n z!}eGCfHwrMpZF5M0Z?7%3rFswd&Z zDRhcsZW-77{9|a_uPkUg`8BwEbu4Nf6e+QFg|4M2b7?2^Ba^vaQzpQ z!1-4bhAVh0Z(=K68R1Cy=QvCp=tl@F5$0!dx*wDR&h_6F8-A!j8~I21DB{6lOXH)i zdRa4~zJIHKyNBb&fUAh+*1^8a7TfR-*f)e~$M~KJ(8~Dk++vfk*HGVfOV-v27=K6@ zxrCeeMB4m_@;~(o9fbyTths-mLd5TTp@W=17X7gsLDa0w_Vj>3H7cvX-|9fbVO#kg zq0irEsY0VT*52(R2(UW--7S{a7mIFuVeY>ZQmgC57q+_P%O*(a3YhsdyxWDBPuZnI z{N0MR_RB5*eRX|?S9Nm%+XQT}p~kmM^$Q`-6qGF}8e+6E(mi1EP zqTqMBk`MfKdey-hJ%Sz);SA{?eR&pSk6Z9~L9;gVlAbkj9*`|OK9@CLP2;)>R$t_8 zKn>+cDzx8X@Ngg9v911dwx*aciqVmS6QD&+(ZN*~qcRN)@=cy5i(*Kg2taudr^*Ph zfdXu-9e3p`(#5{~?(nuqQN`Vu%+G1L2hr>v^gucfTZP(zNF7SZ8$shoSdBaDu`GA~ zJ?H4n>hHe9vYgQ41Jyfbw&eK@gsMq1$O_aqwcY0=Hlh3NiY?u?W zqzMM!ILB{s*FlvR%pBPV-GdsJ7Xr2O5{B5x_6PONLnAcB{%F}u`JW5D26J^_-#!}V zzTq$EF%xpy@(M-|eWyhBkOQopnOMhg9?L!X-WpUAv@CI^FvGp2HQX&Oo!r@Ze)lRN z4v9=P<5T|cOM$JWH*sxBPwP+oM~q_9Kx+p^P0HD5`%#rj`)T0Ts4}(Q%_lwj{m>hQ z2ssdbR|q5QNK~SV4I^qU^7KsUjiO%2npdGo>#1djy@Ey}Yz<>sn6HXp-F6-!yPZXJ znuy7c#^1)63jzsDlojV?g>HMGsvy*rzw!5$hoHu1>Aw0!_!={KpmFDelyEiJ8*TyN zP4)^GwG-RLu4t#R-|HRAwsws~+75M2oY9g9OA0D*d5fJkUbNg3Q7zl|9RR0p~X|s9aXaKmlxq6v9@{YARPxOx|zf)(CYQ`vLrYUfAm8lW9|;6 z#R=-k28tiHh(y1926BWSX$C*;oWMN^Vd7Gtpo5Omk4h_#>-coI1%@&KNgJ~Q6{93% zc<uBC&%T0&}Y}t-8#A>rY3IqkbOWfk-Fi? za7j1Pr7bhtvArSg$FZYR?55-C4hURSq4x~Rf189*eIz=SpB)~cyOE<7uk3(Oj$o^#ILe~q_c z_WjyEa;UxOX0)YG4=24cyEAo!$|1EMgJXL?y=kjt?p|Qf>bpKM&JgY8bQ4ixAf*>s()sKQBS>EPqx05AZEEJ_b18$WPA;DlG z1+uoSp6H_6WXn*LM1$@^bkIn}yK+ehznbr)U116IY4Q-<<55`a4*qm0yq|jlcqZ@p z;jM+@=s-3>F$nfpxD|n-DU>)Q-`)amuTtB3+tiWVV~DdwGRPGl^PY{ zFcVo%bO)RcISLnG>5WVs&dlk>GtDNKP|zH4-?}3eF?+Gne*6w)8$oOO)P2ZvwDT%9 z%a(a%=wy#CqNRK=44X+k;lAo++vf!cK2dr(pScR7MG zD{q$svs}$>->ZS|S8I4;8G3G!c<=c|#tu+I5f-8%RMYV3);huAzK7pue@#7! z#q=!D`n8psk5VEozpNryll4Dr9FZ&5i_P;2PO+0(_KYZ>>ELdAe~aJq7fa$3lXlaN$FVR@^w^5Ig7BzA7TA-$+p$*ZONgJxxY3tm*`#E|7UM2;(-efy9 zYs3_?b(e|sQaXp!xD+4tZ+dFZ(SqQ4C@%dr?K#5Tw6EAQk(#_&BxpQQ@^*TNOGj%E zMTn$lt%4eP$)--MA5>E`(tITuYiO_VCNFeVix=8jv2R^}B$mF6v+FFpUqHaM5_va@ z)ymX{%9k$Xkt08)s|#hm{TazeCh#(BXt%^Q!F_h&-#GZ&=L#=r3#ZeAXv&SCZ8^$7 zM8b*7A?3l!`J+YQ#m}FI8dZ3&tW#=#zbu+49+44+wiR37JIBih$zR`H=`;BJ?t_q>vz<5 z(~BXIRIm*rd--$M59p@+|ge^uE06depjQI3HmM^QDK8KEPb zxez{!Tj%VLdJi;wgpP5XY_B-F^f)|~t#JBjYQI1YzQHRmPLW781<)c)Fkq-3$0R~o z2-AjC%lly~43~k(av$``W~vQbd5;A%t*MRj0c8F~qlSSa&nVpe>jsmel}Vg^pZa!r z*jARXrohxaWPp#xR(M}6w@G{>?_>0%x1b0q zTH5pA&Im@E5Su_%r$5Jd@5^71?>i))$F0i4h0@_hOy1|G z)nsv|Gc+K2(^L|{S2@ak_s^MlPSgca~32CK*F2im} zK0U|kNlBi5`IZs<5fAs)wuP$`iuIfsACz@lDXHUD)5t};&WeSU1+gxptlk_$Zl_t8`3#PLfFJGVTny1zzR*ko9Abr23Ut`viT^4(HgFHoeZ) z`mlx-a^gjol{`q1MK{JafVcg1mx=<5`t$eM)n7bXU8hgqyi=koKulE%U>POJk+_g9 zj-1R|+)6X6Ax!{x6DBJ_55&Idj#twIVJsDuZr z%vxGO;a$0#%69$e+m?Bpv{2zzw_0+6z0d^%(NpuE+EH{gx5EHduXs%6fg9txEM-H# z1|_((jeA>JmHO9duX;1CzzAj}FvhUW9Jquz>LhwE&cim=?Lyx+4b~qBxstPCHj>Tu zI1-Q>Wf{KgRL*KTafT@aWO)ve64)b7Dvo~&IY(6Z>&wwki!sO(`x0r_gR~SijDi{o$h*SuA%xI3d}SE8)(xmP*-t54Tp#w zQdeeNwf)$7vhPzwDH?R;@2qewH=WZaC@i~<9JfEsh_ceVWY(djYzSFW;=W7adH|Zh zU0Vd9)qIm!i(Mff4K8X#Z>fG$P}>;SaL7< zNoJP^N29PmPn`co*QFK~nWV|~8tbUR)}2Aj>aAKDwpt!Z9Jl7ZSmYFG`o3r{_Dgdf zzY}I+dX};_k^U^}TCom7Tz@UJjgg;@Z)9`ld@EAd66v{L556D9oENaN{6dj7^|LtcHN?}EQb?Up4a1_Iw^Q?0)w`CeOvKUz3h-_DKC&E;l zA@?U`I?W`)`Ce4Sd9csGDwk(gv4B>$%)<|eIeXHe(Z0Iaf}x7;B|Of^HVOK}o2^ph zF8Y^AQCsW8{C%56uCmL<0+~kfU1)v=LS@#n%egTv-l=1KJXgedCGp299_GmHniEFJab0xl z-E)b~WOsv6-0T|XhVq;kzPq+cpK#sobiub&aOSCJk#*%Z+XZ@&$JpjZ^-gep=t%f* z(s6)u1?`BXje$BHseU@meHJdt3prZhJE5LXhCsEtK77hV45zo6zj{fXoZTn1#1nH* znzS`PY2sCKyzMJ1T*ZM5W9Y+26C#NvP^l9>T1-pQ=Y~9&PXQy=-^L-*K69m$w}jlE zVCdkFn`(v>F(lR5z#~TNgz2V8sGm8YOjtqsj!3Hu8>E@DZ+D*=u00e?vR*rsdOGcz z3J<#6VjKfL{kl(|d*-kbq14}Zv_G0{tNuWZ{QMgBuraWoTaN9;UF*6U(Nn(kRBz@& zTujoklzm_=V(zTj2DU%yn$Jl z5$X_^iiv5cdrZAVo_*gc6LcN$0`sDaPcEMt>DPM)iJ9e{Q z*U{u`8qR6=>swbh^phtVQV2S73(mcQe}_|yxILhWE_#~F%Q`UrHG65)v@;S$ZGWK@ zE9?0+ohu;ztn=<$EvwP)TybWxz`KsRkDCnacw8Gg$uYX+ZmakRP<B05l5&W%O`RW^nReu>1I8`cDI;Uoyk7^I5K|X2pSuf4hmN^= zYVv-<$-O9sU0)XMSquJlC~=a*aki+xUrb|yXk;OL&sgS&o;v@YG3t9sN-hq=WXJ*L znD=bpSyo1y`u*}-L&N^jG{Huky+V;-J8RU5WPc9KU$7W?y(=UGaImN;JbI}ls?KNz zYZR7bc~)*}ZG^cu$_xZXpH5eCN5ZJYz>QaZM~6H7Y}OQ4@&;A)%%eV!-?B1cmc5_x zeo=LwEy3d>OuB2R3dJIWg=av*MU&UA#y9S#&RJ}q9-olNj{VEePJJLNr zZRo7~NJPKKzbS8Ex9ydMcyc1g6DrX9I9!TghOdwFd48?}o`;vr(}JE*P{d96XQsF* zTOmi6T8I{>G;lHY#WI3UnDNp%5I1zTpIglS>x%5I@#bTIv^)y)+nbhGHx@*F7IoCm z3@>)O{)n_Yj!T6a{^J23dpmEc%N@CQn=i({9f`6*EfGWIa{JN{SJlMWBoFVt{PNn@ z1^yBZf@;KV1AN*nWJGM9UzF^I9h!I0R4?o#e4=vYVm_^_bo^FO{d{Ev$iO3C{VK}> z_z0s!}A22vrA&B0Y6A_({-bRGIW| zc*)9YON(<*w)wXQJp%87G>pR3j1SCax5nU#?G_}id8XMO7}1_qN}Zv~>0J-c=Imj? z9oHk6Uy^*C4&rh=(b{{%H&}>oWyO~dNsw3GJJdET{Brc~Vt)>>{EkvhEi{`#ua{HE zA-sXIhKcAfLub0WWiwc|W+cs7FeH$gtlQ@%W&l*Ll%|qZxjOhVM(#C02V<1>Hu`2{ zM?9e{aiT(=O?AIYSwgNvzWN)Rf9M%~L%|Hfx{fqy_fp{3J-u**#}i>mVYUuZ#a&aV z@2|Jf(Z7Ejmi5g(F+HrzKbpwpM?T!%2srJNw4HP%W`>{2%+wjsOVXFQ*NLXOKcujz z&h>5y?_e8%a^tg4z!VQYA6rr%+``};0uKZaM`Zhei?SjHCL6ujvJYH3vSK=(6=ZFf z^D9wblTVlF@X2_s(UX%zK?U(&NL%)~f4n<|IU<%plE&*my9;!kG{vp@X%4f4yqM-< zl{7T%vH30VQ^(!vOe=jlQjgfXI_cI!O^eQ?O+UF|3Q6wUgrsgPM`t`0D3*{=gRIgdgj~63@sWBqnM)sfUUGa`7FFqt!v-PlSjt8K3kkj7uz;B5P zSJtPx`fXRQi3?iQ6nikfOM$Ixl${jX zQXXYa2-|BXER%SS9IkP-KB9Xvdfj=qElhR^I6BXiE2M7<#Z0a_zJ0HGI24PCJeqwy zr@$JCiq_O3pc~IZQ%AVKfX#*O%>fk{_$5}{A1)$rusl?0XMv@`uk=elt>zIwi7?DhzSm z@MlvftX~!)#+up3-}a+b7#eN%4`b0o8_Z)P`8H_}PsN#_hI-w)PeOe* zV*9|-M_Z4id+P9*`*7lNw1m3FC2>2UwLhndDnQ@G5=JhgJZ8N-Z&E^J%9FEt4NK4D zXwHnz3&1I#D2hJW0Kt6s?oMBH%uI7=$RLnDEL+7PnxBx}H8RtMr;kN@l1=e1`W)8X z{NdI6KG&6;gvp#cZ_}J(`>@;#S{6fj1wfF5kqE?t%G@;4x@wx z6>-%0o%S_1oW0eXe)d;@H1k}v%bLs(^B;~!8EGukeE5{=bwAYQ_2l%HW@~8oj5{_( ztKeTivnjsfof`Ovz=`Uy?#H8OumeLBw`@!rO@EFS$$-MVcFgZ^=0$%*Do zja-??7sQ-jYJ>zB-fxUthzCISvWtNxwGM9R{LmzAGU8K}L;zIKX`62v(g?i*_^X5R zn%e{N_d{I3`>FPC`XDh%0uy#egP5qZn=PF28URo`jYa*Stt4OWTVmPA1D3k_%ff)b z&a)tgij1^TIxpJv_^|$Gr&o68S0a_<-HL==uj?&NzuIjA6N+4YwE`m)4({SXZcuPM z&tmUBj?VSBq=nVa;o$n=efEZRh(6KK#dNCb4NNsDUl)T@s$7cmdU1Y4OTJ1JdV&iB zOZm;{T)hvTJk5sl8!b71eW6QCgXJ%$`|H)nS{VG7^v(i+1& z(%Cg72vSw%*KgIe?4aimbTPg$1-SMjOG__1J1mJj4pOwz4v;A~^60pZqY~mXitlV< z+jWfxHLqfty>tH+zQAZ%rmW_OAU&z(>UcSU$GbqMT_jSG@F`FdFPFcIDU-0$G54zi zD%0pLvKrwN^19Wk&1!wFoL)c0HXr6ZR_EQX3r^(6 z{g>qJZ#mpoHkDt>VDzfrF181-$Ei+Q!MQY%Sy(1hCQzLT^S~}PIR1z%MW<PVN#aPZN#|tBL zUU5^H-^)H&X*cy@P6u~4%(_^Equ<;;elFGH^%-KoQJYrH)wFQzZXM<7BwvyU}1Hp_Z8dk$sh|lMuN_nNC$jAu_0(nB>fP>G9yvm;PkD z5V@8u3*USEjWx1m#S-x+U5^Uez@1#1G_&du9^f%qEZnhEOI@vnF^5c53R@}eK{BhE zUWb7^8r;i#aI?v2B%YBhHfGKKUFY$>qpfpUZq+kBLv7i+6^a8v`Ohry6iz#ha>-;Q1``n)DSWt7FaU$ao z4WTk7auLFtYt>i~EDN6!jaj7JYVpuMU@7`VbmmQGL{p-6g*WSlc@Vm;k%~W)ST`6% z^V(YtcFX{3QB_Zj%2^;y)L8=b&E)d+a?PFELHiH_X0xaMY3b?QqnvT%4oTT-b(e_KoHk!@!Z2*4FrGbN*e#*_j~ih0J~LI&^Y> zSjh%Zpr%&mU3xz(6EXkG0g;!4r>#i;MK_XCd_A`N2S!t>ZAFIX`X>qR3adF0iW@B6 zw{Z49im^cSVAt+~@hO zgq)N)I|FGyEh%~E;(Ca?sMoh^asW2`Og17HohWrE(W3)aCaGl!iaTS#6*;|vHiqOV zI4;_Iv+ot^?p@_Y<5(Epd&J0BC|fJOx)fV!AN7GO0vjY$f z^NzWs>x6tw)o(QnWCy%n)362Zdjk}}D)46aUlCm#W0ADWl>n9JBz)HM)X4ZU$Rjr$ zwFOSQTI3Vg3Y7uI3)*$6yoWS}S|y+DVd1yUpF*7G2A)eRtBZw^fpzs?7%Qlie+2Gw zy-P*8{P1lGlNG;m9F=Mup7BOcGo%{5D3<89Rm)d`*4lu=D`_y zI`nB$)(EYyn`KS!rwGAQV>59YGISu@uoCaW(xQe%8aC!$es<^UA}Z{gQv|UuRLEXb-hHqkRQ7h` zKXPOu8nANlbl}?$`7pW4d_s!Mw*REug<~4r|Dw7`*F_?*nY6QrUU!b|kdRVr7+r*| z!93){g5C^pLt~QM9m|_^^`{5>yY%EzR*0~FUZg1)Q#!uHBVxm1YC$>@p5Y4IVedt0 zT#@J7*+w9BL(X7ml(_v~&<~H#^EN__YXK`RgJ!1}VJ-T$Q+@TC@&v=efkHux76aQ8>Ti=8Bf7m^#|jHynx8=6;%Ea$C-!5i6i4x*2merw1ruPcXqT~ zFUE33@)gnAK^Lo%2ASC+H#@o>jYK&!MCB>mi=FZYfds!M1Nz%omb<>hD%!l4;PBRo ziPa}wOmXuS&=21(`RTPdPDGjHGv2^fMj-IU*Nw?1am~EF>e*U}{`J5AlHvNXf09}| z%RT?()jrg`^_34I0`5t2rOa9rsEdDYBL_@o26s7Qc_C>8gYuF7l20e9EZs}ILgKz@ zx85$yYCIi&J%mDzT+cv0X?$Q?LB)=j6^v?ax1vAq3h zPcs^6C}=QI&xERz1MzuY0?exefECM&*N@&rwBzcaYUx zO|iGj`-kvW&i)MYLjx|0S}G$L=(o6*>-#E_OYlE+PSjzl;&-*kY@@jqzeKxkgxJ=1 zWv=!2MLy9)l>kxF#rewjHJWuyrzkng&um>xH_eT1;xMVB~QsMxQ4cJ*}8}LNjyV0V~Y(x-L#L)WJ>zx z!_q*Rs_9#Y)uQ)1E^+2eCP$K&ErxP+;csABLY1n}DCQfX7$&0OFY+4-rl_^HqnUp; z^I}s>Ogo!H87w$vu53<;!3K77RpHg|FonLaS3ox(!uAm%;8ekx0tUHpoVqMw;ht5Q z2ZYU4?w^WYb%C33*crHt&tq_NFfla8p~I)H<<%3 z;Oe61{j}Z%g(sB*Es`I#*xZD1l|Kq-QYlJ0+EByXKpp8#FxWwdN}(E&E$N}8FuO_2 zttozlk{ORQj1G)6;1(|Z(h0P9k-_;N=1PHB%sL#=SUDWS999D?O~%otaZ>t%SQQ^H zI$g!*QW_+P*J@jhuiR`X@%ne!hNkC@6j;}+l?OjOH_=7bIx>b!TqCQNov{-XDb36@mS@-5SoX~}&I?>I)!fYwRbRt+`(6&2^{cb%Ic z?-nM@eLkJSVW*Q89lD0wT7?ioD)u}wp8nHd2Yt2@Jttb-Yp-)f7rvjAvx6$` zwXRm^u1#pmEd3WqU+Ge9B_Po9yA|6r2kw1maABF^)0$_qEYaQJVD6pOj{VV+)=Lz6 z15*goAoF|{4IbO2jGQvUHcPyFJ9XR)BA3V$zF4{rPAz2D`vt!o&%v9|lmN5ETfnoX zrZU#TBS^OxEGoBAA!aLuvdKKAnHbM|e=+vN{4rMP?@^Bu!M{xZcY^zYpEc>8-bR>O#zn`)D|oa!-mOGP$!4N;_6 ztFV&%9$3wrrTr0VyKJ=iifvyHV!zL=EM0=0OfUgU=lwK^mQT)rZ&)}zR=G&cqa$atd2Gq_nFZZh>u|IdStn|^A{o-iL!!XybfywRk zSjooKnrN|4q{$c}>Kh&m-V)|Fe|PB~CMyy?bLAX`B6tF_nC>LaRegrBD2&0~&pK>* zj*}5qOg2?vt28S+yKjBSst;^`FH|ecI6W>(4WG=z5L#BPL@32iWAYQ~!>hRj?Q=V5 z*csfZPSs%i&;^?PMI0gh8lD!R1K*JE@UxVY9Ne@Y|QnEe#xqH0(yZ6xzcCRpe33f-GgcQur>QFM&)!xcf zl4b~%{o)0g8xhKwg-ue(eY0Jh<~eJ8E}5JZbgxEp}QHS zFZl~YR-dh01_@4?e}NhU?YiRpu-X8Qsu-Z}nd@ui}SKf}|2w#EE;OC)JEXOKpicw&Ew zIopBw9k{~4Io0a-FxR|Cy*m_B1~hjw{*Y5k}aYb;~*7|Pv5Q$*TxSC+t6RMRr%0&_?sLJqax;*nzM+0 z(6&klPT0h}yvu~Ey+2Z4IXMdU53Nk2EpH>{1Gq#xR^eu7ugrMmG6n;wbF zILP8P-A-R$Me+}9xo4Ug3?d8HPLSN~YPrjB3+8C~xe1kS6t|e!2yW$lbv!j&_FXc) z+;VyYq!!k9fZwNtzR1jXdc!Xtzk&LZ6?&|>*%}vmaUz>RuWy+02d(3pNsdZymKg)TQr(Z&ahFU)y_=i>$NFH4n2<)P20-=`a zuED;+^BXtBb9h{0=Mv}x@ZzMO&kY#jqb%d_nl^~y^u#zKR<_Bzt*lq#J9{&w>LP1j z*#HRGx&%Kaz4~_O$0Ib_E&Q@Tq>;QP*ByyogR-vCL4xRtNV=^hjIHef_w(mPtb{x1 zI+|taCe7Bzh*}r{#IPaVb-^T+pJio@12@xF4xKN2o{(R_9{fsv;kS}im*Zn|GMce8 z=z~Au%hV>a8GXkuP2y|TrcZAEa=i(+NC0=q5i;^c^ruJ@J%y*V49~mAk77sPCrjQe zz*Kc$TB@zy7(QcdxH;PsIQy@ z>s}lkehh9*62z~2>z_!}6DkK|_;6NNXTMi?XPd9ZC57;9pqL|WK6*tF^(WM}k^85? zde&(dM>2&rm-A)M7`qJfwE1|{RzWs5xqe^Tw|LiB68ADny#9iSD+lmZDs$k*I z3YulqmYV2x0q|-baEdE6iY3%36_ncA*k4t@av7bwd=|}P^)$nN+QQQj6Xhe3xV=or z-}&Vnlwd$X3V(|5?j*4fvQ^5h6ZD|0GjmXOVI--Q#X|8t5WR3w{)XVIhn4_H5ps<78*?4foVBEVRXKmfCXm<(ynxIIO=?4%q}{Ck&gQgLN6~gT^67a)CZvqNY&Gd;@E04yu(fybEoPlX z){H5>$-!Uz}h50*ki=*Q9wD&b1{Af(;5oS<6|Mh-3Q>>)So12Em8t*$l57fj|~i2KWKxg-8#euDP$7n13(vG+xQChiJY`eGidO3D zC1sXA^-Q6JN##`-k(eVLh||)?`Zizp7~Q?r*Kzyd$Et+j%?oKVS!pr;{l5X?F|WVu zytT5%UF%1!ucP;}{r2TRXNOCa*@yd)qgUzT0yT9a5!GtscGp(@nesycX>#_oylcxN z(&*kB9xjLNDQVQHr%H)>TZgUZi#gY}(VzMi+-{SLHFdD{S;TUZ63zSxvatqJ#qw;N z&{;aJxdHS#IK=ni_1*7$j#pzC3e(ZDZk5VZXz()5qH~TJyqFq^H>rwd!y9XApP%b= zSOo&=?Y(t0rv^Ff>>jJx&;os%hIq7hoXRuW0}vPtL_W!pw>o8I9Y!IyZ>WlZl2*X$qA8nbYQpkT(6STL1mPJ z)e^dx+`*D;!q6eP?5BmntJGBRm=p=J<-PsQ4BgT^G2;Bn7!0mR>&OZeIrNXV(XYb$ zi#H}Qspj4Gs6`?>ky8qKxw^O}n-h)IcCvI-&{k%HO}o+ zUqP#1EcbdP2nKv|WQ;^qymLw3tjX_1TVdJu^Fw}5#W6i`6zO!wG%Bo?-?qDX}FFTgnE)8AnVL$hxx^|VkTiC&NQlX8FV%-W)fYrf5z@N&luZb9!;-I&(R5HjICn~&3m*AO%D zPQLvbzw?!ngi@iGnkjL>Os$5orIu*Luae}8XZ>67D33%Q47aiYx~La(-4_adcJ0=@ zYASf5=~i*Strgc=wgHy*!f>XcIj8OhFhj`6BDyl3;->7Jz>5Q`VCq524{6YM$Bu?8 zt0xxf#Mh?LeMeDt9AyR_p9xp?qXQZB9UP7Ob?F$-aw6C)^iPTwrm_MaJ`x!yJQ`Ux z6cvoZ^0P+ZmBiU3+^k#f1rMc}5c^--f4)F(TBKg;I$@_C2&+?Q+K z%bFK_WG4Z9=?A*k4kP8TpuNF86|3kpIdLDXyOyw*0W7#p#5gL)>wA#DFaQdDjXoZ1c3k-?{>1`S<9WTQbR)3?cl@7AC7X7|Ym}@7z!s*iI;xM!n zNjk_J%sYIrl>$Au6KKeiBd%)AfT{$tYG3KVCI6@yy7i3_fM6>wjVk7kt=K@zyY3MF z2iG1~DGBU}M)+R%Q5~z-&_%=?BUkc1D=mJ>Rn8 z5hEIzG@d6U7|o2RglZ;|V_B~CO82SbB0_>ss8_Xh^ZpE$JkM3<-TtVA^93~QbPoe3 zJ2oWvg)ZA{tn;9q;M~QT@N9U^$UY)d{4V&?VXG$YIX9i|CRQzO>+m!7+ir52CtkNw zQ(*JEu6hrW{CE`+>FaZ(PjzCgTU_fQO9=#D@E(}o@4`rKTr?7hw=mjs>daz|H)@e3 z+I0i9F-q(j^xjIxX(h0RyR@yDJJj00&3~jp(^!^V`0)P$Awb^0z-u*$nKH+p|9)Sx zNqst@P}j{qAM>#DZV-Hj&<2Qmixi5-R2Ap^*j1V(gyA@d=vbyA8hemVe;()~zU&;h zZqUmVThP>q%b%Wmdb%=Mp~18tyk`LV@pwQ0oioh)-mcV`$!wUFRV>g_<7zH8_dihOyg z{Y)kfYJp>Pqy6(}J?mhCypLZCVUc)vYb9*WM$v1=Vv^YxL1dZ83VeFp47zU;6BPCB zcX8H%gn)*;2U^uu1$xtvU_gW|W59EgaGJPPOmsYbiV~!nS8fkWoXGO~Ia(jPMF^2v zfI248)zM_}q(!n)k3G_9euQiX(6473teF)l5B4{<({cy=bqBn&cRif|1V@RQ7SWZ# zhJ0v@DDyur0uQng%+@Pf*tzI1?U+MG;A;v#AHI_KwBz4)!#F)b5FP;bOUb0BB;YJr ztX-GT8RRqHKQ@=m>2&Tu&s}gar_W6->Wy~v__(RG%mTwYWVvA)NO{hDkKh(9*~iPDZ+@l zGLD?->AspvmtCw<<)Y0O(bZx;lZ$e;4=6``A1~uF_^4Y2MfDtYwNv$#ek$N zE4@>qv=ZW4Lz0WU4=rsF{7pLBpyl`z`JRSbrsmO!#WdN~RYJ?F;Mi4-R3ViyB5u)U zoe%!YXq4g$XL<;8aA*th@EhOXc1EVLBoHn>dHvtZ#37>Fq5P7i1Dow%A|?y%-1~bx z-6iw#pO&L2%*!a-jVxh~jqD=b)JXoAky6Wq5EI04_kUz^nU(MygrAiIdIkAJbm4WH zp9M7?2cbc>v-i@oW2ooi11U+&qpll^)E~#6L1kprTRDYxg}F%Vw~x^}eK9fV*II^3 zM|aEB5YIY5w-8D0IrAUvnOe^wWp}5(c7SJ_Qgkh@UToD=V^W2E-Uqf+;%_e``toAPaRVvIj8fl^1^_|dr2H+ z7g1XPFbWvEfxDibNKWZQr#KDu7IB3a@tpJ-KQ`~MMNZQkfC_NKY|w+2q-m%C_f9|- zH$!l*wr=p889HerPAYawUlzD3g`h(uYa% z_4&D%E>=8JN;YZej!Cd@ zRRTQRtg8QK5!;Z|D0 z_@hrcQhUsgxZX{iv+I2{wjS3~u3nPfSNHx!4bYK*2-*b#w9w>0uG@~0HC*GwqDEcC zr-dMB&H?TL($kirxA9)nvCkGGKkZ4RZ-n2SkGCl)`JnGI6>UteFKxWMDQ%DXVqi8L znTpW8V4vTbjY#z?!@finIFAW((LVV(Q-%ZI;u`SEyhDe+oasI8TKVR)c=YL|fn)gA zYC>LPebjjj$fSkM3syq9e;etJbUl9K4i&*=nR1g~?&hNrFmWE>$pLTDFjZ}{?0RzV zYU#aU|H?%+K9Sj&t$IuI`Zz|*VZTl681tX7BwN_VQW{Pg3OslV> zI)0fGU#RD()VL|}c~yZS7!^b8aTn^ih9n^YF9eH6V5Y+6E(hVe6js7M>^$7@tDd3^ zFd7ZFyuxwB0DG82em!_pnmJ*J%B3X&uL)?ah)Iy6&aPiW&4398E^f;aP#C8Z& zBy7^aof!W(pDw#7YZ@*zXGqbz>TmT??vz}sN@rd4jfKi`zNd6wMX=@`4^r4}m~W)Q z1_9;x(1WjYJXB7%@(QN!gMWn9m#yhm8^qQNxQnV-Wfo}-EG5jfcfn?sCY-C0EK#D( zmD|*m1M{f#aIomZZVqT9(A-6_61g1|zwXNSQZw{{kFeZAq&GHR5DdQ&h>b?YOP+gu z1~b6@3NJcJLeu@O8+@mDsIViCWP{dMct7YRGFIXF2k@&?MhvN-qVzljK%@{`-Nwy+D{_-zPf<}oC6aoDCNZzMKE}(fIy-64Z)}|!2ByQ=ri(2-xU_lAPyr{R96c>_a_+AH&0W_A#k^fj zfOn#big^NsJ=10BxKrse1;_?+@2MT87Eh`-PCSHVHItKA)_v1CleJcV{W;1&zD~XY zbh?6meY%x-|8+p_#B)?`H~ndt7-R!T8Y4p>B!DbsBS6^kf9g}&OASC)<#Py`$ZSXF z>nO2H0shc#oyFKS$Ldbxz<*Uhf?Tg{qC>J+z1GUhgv90My){@s`z~n^cMzs6vTTqv zwPgZ3-bCm!eD8JIq4y;*a?H9rZzK6WhL$AAp-!KG=`DzL8ufpYjg<^@iwqEb@M;&O zP;*ky!9S&#Mm9{??W4)NZMO`F;w4GKKn3_XA{)Iquf%%iz2=C9c%4G>};O$CL!T_qDkUF5nt6E%4+xO+ZCkz z@7Tr@h;FMIj`%)2nS7sq%BdCsEs&J+e34wgl2aDywa!TPsw0%_n$mS0dLyOpqAMq{ zT+7iGPf_-;A3VamZ{h2YeD47yb7k=S2=C`;v*VcU!(KE%PtG0sX`YOok&5rU>>|@^ z4%XN(@l6d|#ErW;VDCE9lVLEsnO;p84uj;-K75g#`~f=)lI0K`BD%~gl%Pr(T@`<2rNDF;N3*h z$>xv@>Ct@dS(GE8{VKdporZf0w<|bb^vM}pt?62zYEacwDeOGu{;O8cGicyG?UTdq zgUfH9&!31VQXeo(v@~3VsS>_!|71Y~4A0=gN$dSz$|mLSHCYTLE0NO8H}B|?!o|Tp zer3n~5|};gNK}B2w4I@OF-}}|{!udR8*R*W2P7vhGFA8=rXB&(UFS=5!B%nNe`*Va z_m*1_r0J_JI5<=Og}Q`aZYUOwr6-weQ2JZdILgQ)$?pPgza8 zti}LA;5oPjy6KCKI$9r2djSmKnGY(MV>oun?#5>1T%kax; zec6B*?vw2^_y}~dQ8B54Ohnoqdit;shq8NyA98X-dvx3s6*eDUhgSk$qG$+t51-&=@>A^(;qHQeynTGiS00#9U!cRNDLeaMEAh-CMDnUIH zF@-8wYq(l{XPS+{A`n4%J&P)o`2l~97%!khh-yNIfU{-k*@6k`!!zz7`zKXBo@o7! zKtf+_ONCR-o-1SdMNmzZ3=^{)x3qb~il{#$T_O0f-GnJI*Lwhm8G#_5>e=)tY5~qw ziTS@%fPk>1H{KtsUTMU^uub7-t9vw*APTm-LnsRVjxz7~Eto2bL=;y`LAH++hPrB` z6jw1;wpU$GG5F_r~u#)tR@LN6gZ49g3Cqt|q~UyHKXm9I?pGpG0tROcpsuPUs8%2K>2 zT^naZTImx0B1q0($8{Y+chVTc`DDU`oE%@%X}{7hRneW!Is-D>Rc@u)o5uNdbv2J; z>Y!sIj20DJ@&Twve&vq|n+YecnU41+M{{0FR!(jvT1b}%?|fsw*H*^NXkTNchf_+7 z#;?xCWvzwY{rYevaM#kgHsw!8x@*TmM4X(FzeboygZ;Z1UJ)9*HW52trhOi{&)&D< z`VO$)YB9zXN$0=>u;(l%x&ANf8~szT8aVDE#!|!YIVZCN?O=^TDZ<9m6&-qQfgEPX zDS?Q^<6E-A0A1Z8;RP!2(K4V2UpZ`t5@38??P(;4iUHMylJN+=s9>5ULDRNrlC!^0 zD1g^Q19JcZefN~YpUQks3Kgbq8&R@axSHyNVoi0zD2Qn{5%y#uGf6r2UDBQtEkOgI z@wI!Q>kDIUkQ41UoUMQhDDr3HK^}6dLO`qmT4{d@|1BK_5{6@f$~N6$r=Y~qGl)Rt z_^Xrjq;(~^XZ5@`cT;1EmGNaV2_(4oOQWvXqx{L4D~ZK}QYnq~(5tEgOcgoTc-Qu1 zkSgLlyQaA8+-EJTZZToDF-89Kk!5~NdPI8bCx(u%65TT(^G~h~bYO`P`#vfAwi_pI zUH+*6(qK2a=8LkmDRB)?JLuPa<{80vVzfSJwE~oO#n|p+_kZ?q4KyqG|0qfQmfp_b z^M(Oy$D4#>DS8ZhdNZH5kJDhfgch9GF4r(C8(lR`Cl`?xlUn$3bNtOwHbyR zr=yQSutm!t_|P6yMWQ;D>uq1U*=vcKwP@oCWCBoEDZbyOBbT64vwqAe(%`L-kS>Qd zDIYCIur9~>6@q*Q5$|Yc()7vI?|O3bBd}`C&bc_c%gG$ zdIARmbSyBXcJqDi2{7C=D7!2&X98z3aN6HWdOw0^sAW0X#xmG^kUDudqp(Z=6h<2^ z!Ear&CMq?o3oVMGFssy$=2EkMks9FPYaT3knfYo?#D1VM#=alilIPYWEa`cy^LqsM zaeW%Og$zoXSZN$Lk>`r(e>jjXh7xPx9K?H+R#p7sU4D|se|UO^M< z*($?O&}&x^>OENuw|l}Q{9shsBK4aTkSCKFyX#QI28tcsKd^x@oX^CG%&~(Qxv`fz zec4k|zZEj;hum{U{UzE{r3xby17A|XWu%cH)iWv+WMQ&?9cKC39QqCqirRoUrvK=MAx-pq6URT=V@)^#pM9z|=r2dg2pm9s zV})_K&+l+wvi*Vss-|6m7M0}VZr2#>QPujB&lvPbclCyn*3rY;@AOtC<(Wnr+0I{_|=p zr>N+Fhh)cN_oFT<;k)S&F$YK@A6wttCXv6A2GtY3zdZI2M^^=7{=WgEo<-aoppM=i zwT(mO0b|Hvc+cY%8B|x})BqXNLyDMxLYq?>`4%%PVyxTda1(89^xHAxM=Q3%^R=Us zRbv&Sh3SFVT|Nsl&$}s=uDZ}_27C+au*o0Jzxa};)e9FTPHicSEY5*56B2gpKDQ}> zx)R!(h4Es8#A9lgOS$9`H$uVRHNBV!pOAD#Oa~Nq&~p#FwJ3Y_N4^~>3NI{MQ!rmJ zmH|H*`P1?MhU5wTr(678^;cIwn8}Qzc$<%d3DJCgVXEfxJQjwt9b*AnfHkSV9i#X% z5l&}`V*a z>q!;p=*5|ubgJCBz(1{>%m~lMR=BPSwBd`l<^AT4%`K{y_IRx7Z~!t+e|u)I$7F^A z-D1NmRA64}O*u88{Yim)fk!*+<4}A73xI_V3AY|`K4{JSK_98lBiHAn>l+Z*@{(oW z0HwFIF+6-qli(tY0Vj2I0_Zf40pZf3p9&R&g({%RcW7bqhcBP*!VvO2JQh;`n5mL?nmuRAB> zV^a9sodvF^;NF4pF`twL(juH0$n_kx0*Y+;;grpnq39YxbN{-I$lheJ?tDw_R&&A2 z>$HB^m}b$PkgxWaV7&XpH7&R25>MQ+8X{=Jg4Yh{<5W}x*elfUlY%AUA=`kOO=PW#5>CNM&h zdTV`l{I+XM=>!G{S$O#B?x{H*NU3k)>KTybC0a(1CZG2w z;~HQPc!?P2Nfuow+I}kjt&?itW~=|s>ZXYmu~WR?8K02fTQh|LKsFTV2j0p7fxi_% z@6U1T-DxlcI>6gBKQ{Q}*x^u)$q;f?kJ=X7l7P8Lug+R|BriNN+h`$7o)k0A4b5y9 zWUPR$O1Vf4J267pDZ#tbZlYHD>OxTlTRsD-FzCAuELr7V_bSXQWEpudFc}d|6x*y$ z`E@O6i+7piSJjXtaW-?NChB72Z37ooWSL(o2RSlT*TY2wHoq%wIpsT9-)NhW3)%_| zF^Nn(O}eT@w=D~InA3ofqt4su8wQOYLn&U(7bXijN^$tPyF|KQ ztFN`A&8*>*rd)D=(ovGOqMp1es~n7{xi+J+q=J#~QQInt?n$YCsC;!>k;1{fH2vHY zTcSn1K@Z#gkE!DING-WaW*DvJA^$s~V?nA~!rbDDeNrpWv;dkHaSq;mEvnihlpcKt=wkQF zv!Ue{@liLbr0rQ(AG=>G+8?K{&hMLf!)q_a$lg6J3EiVkT*41O@sz;Is=p%q(XEcH z&+dMbV|@S#b9W_v{00(3b3UzE}o2j zG=nc)ZfAFCmHY|8Bz$$K&fiXS}$-OOUY2 zLk#iTIljehLfi3m(Td@skF7DOfn{3Wad-8WZq0KU%r4Aq9KZzf^@tu!cQ_j4EggXB zchXX9Vd}1cPWqbyH|tf8P(hNwdSpJ9n``GCZAFnq)z@kFIe#?Sm{BihH;-2ktCfq8 z2SsHKkJ2$8l8-m>M$@=iRzHvG)VNv5xNlN*mol)bC7fAWQ)F1yNDtiP^8GWVl@|ai zm9UNCd8|m_q&nJhUaYH`Y?`~kXo=?p z8x_2k?5e*Ei;=JPrL^3 zRKr;pcdy^=BvYDe&Occ|I~g=h;G%LKOU^=Zu&m>ix4Fxi{!~T^ zz&(;jz)Cv4CE&ygILW9HndH%5Bmc)<;y`r4j--dqTbwP;?jGH+w(xn-EZXVuFjrX+ z>@rU-WzQy#s%1%hsCQ;O#a`+W7`6pU*Juum{4WuJW}41EnT2aAMm$S=XT*{GJGO?( z$P&(;8AL7nja7yBl2z>TuVhZ%JodYCY7{Ps=IW6A8nb39cnr<>5Kqpf8>oFAvy`VxVK5?1uOrZ0@O-ecIqnVhc(;wS47M_>c!I6 zGf8&lTk?{4X9NW0L3bS7rM#&0lEm#Hx-y3)cGXGYd>+e0BSyIS!j$_`Kj>@*GN)T*CWBd`6pKreS* z?^lWWw!za<_#P87Kdx@7!r~VBJB{@|=;RXu<(IKsFJug;Zz&f*Zc}%7^EAZetwg_n zwY0@crT|Qa}JUNn^I?~sw6aCRLHOA2hkz*KSg$J?- z>O)a)O}e=c$skCep=VDF2oR6bzOcD=q2-9E$Vr-s7s|3f0K5`U*)W#W!VI+(f}Eq- zY`}R&OdVw70vSm=B5A2_a{bIC?teb{J+*hz;w7~u3ysEe%xJlwJO$c`U@)47|KPb3 z_Z<8rei9R>^FEu<>Kg)PlWP0-;bfPY!qF|Fv;Q2@^jPVsXrwne7?m_Qr@x2cFDMEa z*fmZ{S+w{3>&R=(hHTpvG7>B6yW_9}#K~*C)U5pIb-bxn4j|sN0?AOdDwtPZk7sCwW zqbs(rgMB=F1|lZ3dz`hjWPGg%N>T0dr?YlWUYwJiGWs7KTR}6g*`MI}pZNL+2lfvnzSA3; zM*UmV^AGjjK5fv`94bL}r>mTU+eP+iY&SwCL=1owkR`=b9KN&Xu6dlc#inO9iX9CX z=IMY}y7DAOFY`)5r|tM*YVYFU?uZ+!I{=$#t0_3f(cvdVs}z9UY0RLdPc7pS?hGIt9aT|84YSEYo;tya@3Ex(=f2 z`^YQYe9b)kbxZhwLj9P-1PE2$gE_KXQ6+gqt8p$f1>TwhsI_QI?}ypaG027d0deTE zB=L7FoR%`B7Y98~jvOFWh`rX{hRH}u7|Zf_fioHUuWV))e;@;i##lQs9ay45%@A>& z94Gf1P%u0DGrm>pQ(P?3N9%$30{#0UamjoH%TWH%Lyb2MTT5Goh6$t`0@GuBUiwId z{EHZ7l@CxFlSNiU?D#=XW*M5gH$AbgtngN~*LlqI)C0(wDhK?id|8kF_2g?Ib1fEZ zOTFLpNng1>UD<}3IA)j5vM?=*&O!*Tlxd-EzgaZ#K&i^Y{TeA^Jh5G00Z-kYg!%|b zI6>fl7%uMyW^)wN#ic(hE-ITO)e$OJ(pHsZNz0*llno6hiRqJ@+5!dnxTTuu`+m(K zaTqztX5D@36&uHBuUg@c*WL7Kwm3#k{UB5$bv3Z_qlvFKa`f8ld*yM!HK$HQJ!l6R zu%ew`i8hyj4FsfpCLNBIwn^!KY$EIe&ScJ?oxW0(e6LBTj*Tu`(L-J%_c>T`tq~QB z{G&Vsto@*9^BgZ7t~g@OZ^Xr#^u-%MeMCl!Tc6TbQ<^~HfedQ2P5qOwWEOA@Faz@T zeTo3i&l==0ocfr=AK9N4o1~iV-z@u_`^*r^mEcv3^QLMb|06@=?ym7zC6BNhw9_h_ z*F^O)bKY%Wqb+s|&`|JAOrcX|&N-muVqF1}9bpbA6MRQ_bX4h7v6emAD;d$dtRDIW*$2qQWA zmijYii@48FksSuOY?L&N?;UwWCIB|wI%Aq<3MHtHUcr@f*v^-vG|lp&tjg$j!?kR+ zqMwnT-zu4>bUws_?>c$O@Mp11K8izLy1f<5`;4jugBp64TOV6KAJ9%$$Ax68gH{LP zwrK42kf;BXW%9P3%rk%0`UW@hnV?f4KGmhgkmIMt{v-s~bbxj6ef>zz_86ZK+{MoT~Y75LdHJl)H<-qJ0n0hZc%*Yop&$h>WumJW0lroA2eDefeu*&xbW-z>!5^l#LmgX;%o4PgF5QA%a8#{ncL?9q+d zg>jyg-4!!Ug?@YtTFR4`lo}fTui5ZZB38q3_Q_LRk3@CF#-B1%8Y8K6aEwlV+k{}6 z#(n20g#;2);Yp$=V`m={{3=CJSFukLxsPDJ&;)c<#?k3PiGNdnoE=TubbGs3?LZLs zwdddE8otqS%PwpWndB2;eso)A;~5C)3dt@T49NO)=Z#Y8FsSYpF z+xML;&=1PS5aP;J(y&uM(y10X~4#;F8#}lHy$>2C0yJ_~$(%-?!!N zQx(G&I$mXr=tR`;-M7(sDDM}k4j;bu<9ChW3Xu;qY);cXbvk9LasYMe)oJIx_#1*v zrKZxd9DKBq(ft@|;^+CuALzvS@$#-am0-!=U*si+oPUtXogW06JuR?R94M3`SsJ0e zPB9(9>OwbLxK>nqg4_USNB&)wg;TKe@8T*8@8gSDjATsp^({{q8p&y&hR?wj&uu_n zZ0W}`s-&FZ2gehm%;TT@xRVjc(=hO@ZaIcil~ZUa+KJ7Sxg;1$e0^kj5pCp+VQ+5s zRwBjC6ngoQb;N9COmgGulpP1Goy{7m(3c1m`hjw$sZ7$OQBkMezF%RwF1MjbC>G2mArZBq%B0>_7HVK5pnbpGa>}@^p z*__~U!*hN_xnOMl3_g5CVDPg){~y%MSa{#FRC<{)H$@H!OnER0EdLdD=`K ze_WWD5vhJGyVCdo}JfPx5j_wv(IZQszVFx`17 z){;g8+S~>3zuO!{p#CEP$1K|;G_GwXdH|cyp%oOikEOIohW+Bs=t?vqC|w^EJ8k`s z>6d83mih}r&`QN7dF_11ltc)8ke~j5Sjw#iUKmSylHM3Z)wEiq?3fvvy)TT&wCQ)} z$)3$y_w^sRTPo|_KkRyW+E? z@Ad~1_u?1SQnN@UQPrOBhhO;l=DJDI~n{S*EaJJkyf)+Jiv#7Z= z0%EAcWryAboGo~8(j~%$4y1?WVdfeZ9?{i!vGHnzxr$vi3aSV(hh_2jKx%R{s|*jZ zv>&MJZHdmsfT?D$T}pY?lL1>yW8;oOtGf+8b&k>Ry>n}0Jc_?K7Z!*py!DGJVHFS) z*0`)Lsu_k*dayOfHOQf{x&XdWP@iZ~bNm&2rYUw3jY#J#6Azm{kZ%wga@eRZeAf*t zHkD&3V+uP~e6w=AiTx#Q_+!>K(pLVi0%*JYk32togtj<}gfCK|Y&A)_&qv(ssPgzt zEU=)8Ca?m4{#W+P47V%7Zt(2gD8M8f2PhlQ_n&2p#B=;6F5>;_(JvH?^-07A7Nz(Y zq7vF~$~d4VQQT|$VJB7;sZa9)Z*{D*8hRwSy z`0S2c1eH1B|8E)ZSdNt_0zPQiDDY0vOW>FG}`K$vcANrB=-#s=__JWKRn} zlyWZ$8%83gm+9wfCyOmr2h{Tj!99 zPWCbpXA&WgmqDq}yeqq4!)e4AKWr|*GnX@s<;h{Qs&LvO;!j;)eemvRs#?f;SEUeI zI&>m}J-Hn1?~hVB8|^?K-ay^y#dkwL95i9tfnhki$hbFk$|yh?4fYKYvCzPV!;c4cN`f{OI7t?-?X%nm54=IZr= z-8R9}k!H=srdaMR-QNZmfkXv0(sMWzliboMxI3^?s zG98s4S^+XaU!#U>`?)zT3LE;oz}z{(_@()Cs26n1`#bqcls(b!0SP=LFf+59!a1SBRd=If|nu8yJV9r&#n;f zO??Jty=Av%u*(&}{f| zY^njG4ay|;oXCd^b7XuNd;ObVg(48zJzTme(rFn8gkfjUc53xrn?xHf(*8{j33PAH*qzpFG528Bg$@x-YIOhaBl#A9WYUgY0 zO8~YcEOs!mGz?aSLN{o*jVs$s1iu~c;|6!(1O;pzYhy{u##PPFvb#;VRdJ*NmGp& ziuTZtLB#l}p?ELQf0W2mPJAFKa1$C;w|+dZ$`Sz$>su2q)Wyq;KzZO0ZMM&s25R}I zFyhM&1x(p7KbjeBNF+{W_oYr0>OJm8A>%IM(t$RA!=-b^uKA_q@J4h2?m8xF{w>3) zhX!(rzQ28=RY(_4LDz}!;dd1bmb*s$31xR`Sd}rIV(v3 zK8qTEqsJrK$9rV2p$BomfM+}=JqZ;o&!!)tDqMrcUS6r6{k-o>@{QR{nG<4_bW3J3 zeob{!C-W=2v;s|uEyK@yY@rT-S$u6q*>zCH6r;~gDQF^( zaZLk&>?EdSsQ`jtY=0ZHvq{?46q5bg_7Kc~%?XVjmWvqf39kv+?^ZS^KrQ{ii3&oN z^XahE1?k?uI(8vQ>@FRtVpgfh6dj9vfMBeHZOE+|4Se)>zp7L{+oN@mdt-IW z;PxAO#Ipt2#8`U7l~)b<#PC%mk{bV0(Xk-V$|s-W71e-}_qf8Dm(E53Ug~JGY$kAv zaz`rsO0+1MC^sXWb|!-8{Y*#M*9NKh+OT6=zilj1ouVpQp9=PCL^Mww_JJXmCc?n% z%E?=N4-5B~MCz3`KUAf$pVNN}!;qFSz~Mk=EM!G>ZcanWil)gl?Li1lrK=7f=q`51 zQ6z#yM1|3;*D{$~}2}J+?Fbf!T-uj9L-{2y`!tiW|pN@O` z=1?!9rw?5*^ylAK z_Wl5E57tn1^?S3Mp8aW6}sqRxyQiWZTDn`FZvz+dl+_Kjg? zm8{WAHI#@eNfb`~J3hoX)8Drn0cwl3vlZTQh4I18j0;Y93Y%H;7iwHu`-7_>tdj1# zBWue9H~3_U0JYy@!Bih0^pw|*@1Dk^z%`rB1dyGC`?2z*fF)j(t9eORai_KlW*c(k zWr@rP>WA*n-!`}zuyY2@jJ?Sz8zhAWnskRh5D`nugxsQczI=>fW>C9?O zK=S(7U0C$;4Tu!O#PdW0#016`x;FbdunL~~$byU|$K63KxIqj7QQ3Wx9!xHkq^PT~ ziuHAD+x$Uo#(f8Az92evMf0#2l52m2@lr|^{cWd;EUD8*uGbEwpR1<>nn{-l_sCxd zabJ{+)OP(Y?saC7QTgJqwr2)#BII8Gmw4-jwyzd2ixilmrH708jdRx*ON4lTYGhxX ztryR3+5d5X;&&`@$`U0)(*?T8QVb;JJ?2N8t<>f(`(5Q-tcJo2xDA9-^wLDSq2scG z{VC=uf;dJR!UQKekbN=MaFT?`A3p)zm>j*sU0a9bjtmAyL(n{38kTN?ZQtWSgz+Ga zh*#3tfAC!gx)gz@QIDeKsZ4&FNOXcrGU(}FW%6NcsKN{>MuY! zyg3KiyRq*45r7R|3XTKs$2C7+z>Iy-Tn|Q$h#kA(`#7HM6T)OO>W~nke|~`Y)a6qV z(R1qOyH%{Lbsoi<@AOWG;#SVHE^8$VXf~POr*fEF$=YHoG|%7NDuU_7qv0}nTCb(6 zEL3%F&#bhKT$p-Z4mwpJwN)5eq-euR;m~bS`)9XH=w55%AX{7vV3wX5mK-f$l*9#{ zv01u=(8%((My?akUGZ`&>siX>=c{~iv`0X<>0(K=x(jBksVYI8a=wypX=dIf*IQ&t zNXIFd(KR=lkhwba7+MBV3N6)UD=a*`be76VFmkVCpLwBCYb#pCLdQWSL4+s>%k_z? zl-qL__GmCFZ#xgcDFO#c;~c6RfPS?M<@Al*2jwn>p+-^5JZ6?Loi>=_4kK^HGLAUX zcrm)F{=)1L+N$cMMSoUsJ0~A@hyx^rt@ra+45!H;+XQmq`>uc$4INU7^<@HK?ELVe z*NY04=uf0o;OyQb3INphE+2@i$sr;~*((|Q#c%`d=5A$I=e<{MfmLhPgERopU~X2@ z=Jb7*BGyB>a~!dbfgBiX8rJNH8Jh`9xnPm{MNLVvhfjaj!3m-p@1v-GvItdR&gz%z zZ>Vb25|Pibm=hw#%Ck}^ezu~=N6&E+A`c1{kqDNywagONlFRhE5Qf}t$nCi8!CuDf zyKoU2JLeE*I0<<_i@!5=9+L{#uJbNSYY+6ALRAN{6*Euhj=W^}$f z_17Wm=bB<4RFA8s->7>d3FH_ki+!{U<1s0?Dcmjt{2^OgH9U*4xls)P1Aj%p@syPt z#Dbl{qucAUc_1XwhUdHM$;lGT{A8WuY%z?_X};-L_1U(;(mVX8=Bb^~r;=L&tI-Kr}c z^)}DXZ3Whbah)o3g<#V})IFr3UMk|l_p-B)I9|wz)DUm!bm>r{tOd$9d{o8ywU!Dl zK9SZN^`AzmXZ+O0&q|3Zco%0EuwSbU&r%sR%kyI^qx`;(ptkJu1i+L1B z5Nb(3f{BXYAT)qN3S1=Bzh&6!*n5LWQj&*LvHi~cRsD@EpS}OeAh^h4Sdud2iugdr z-a$6)T2tRDl;72N3SQ`M`Wm!ZTUSL(kZA>i2i$fld-AzY5}WvUNt|pg(I65%AtonN zbw;QVY(jdO_z?6!-?0~i2jjCU0cyGH9$3R(N`W8qg5EbKhFJggIJVHFzVy2w1Up4C zttY#6&7K|24W{%6Wu}K_;XAvsOEo`KND)KYVB7}A=h6{y(W$T{Rk0F;Ejfjb%XML< z2(C*2pXRB^o{uc48ki*kKHD+MBL#St(Hum+LRkUKdNMWtCqHNVx&5z{Z zX3qLTBXBm~Z|9h1Yi0&FI9c$uP*-e3Qv9Zcdxm?@Ln&V7j|dF8_^voQP9pFWzYw*QVGeU-sE7D{VA@k3G+EPfUS8 zpJT-()^$11G<=J9Wk$zvZsPl<<>x8MhKjo^<+8y=cS~YizMFI3!U&vlZqy*C-m#j? zl3*Ck`KB5UXaFglLn{tfYf$mOE%0E0f}UeyG8#xC<8a*fz`JbwRslnYzN&`xp)-T1 z^gfZNmR*Jpt^TqukzfX>S{l-)5m3eRLbO@#dmqPkt9n>>djY(=xwjy+cL!FbTCzlg z>JKQlxVRNegruS;_5deEm+9CqLfC|39t}qKiF7XAsDfN_VVcpaKX;Dd6#c5SmP57d zP<125@699UfW=^OzLgtR7Hii^d=#$vu5Nc;-uKgwDOw2=Fr$+7oU zj{xVj5gBD+;COz_Q9-aVM|~rF_-Z|}#5#V~)Nkqds1&+1Js&B?+VjCKxUG*p8H0nh zt#3&5m-)5lc#>>Om;jY$?&~kzCH9R7le72^6J&&f(1d&WG(_5$2}|TT=@T(=LCB}< zB9pcn6ceV_H85HifNl~T@w0QbLgJa5FYW^WiU?I`}}mqNCi5kyj@a~gs%f! zT%EK>{lJZflLd8K!%|3Sq)vE?V z&=Oj#eYW7J6WW`a_#zJLx5b)A;$#+giD2vb?iAxM&awOpi_9<|=zf)DXCQqlqDN^G z?svQ%ag8>N76T5x6(esCY3?oc$FDT=mPjtthDpQ4%T7zd*TN) z2f66bvETC+uG(epT#dI5mslp#7qz7-wekRK;R6;71R zmk@B*sEk^<$ef{uaB{C8NRQu=Ewok9`VC`i=?=5$AyUlAR4%KG%Ag?um?f(?BQ_cq z=G+&5`-p2}0X-``=#d#+NKh!t#~-$Q#9+N@y!pHhT{58^iE(4DCXX6 z)nlL~g>ByVsm*e%)P?7TyM84-3f}??#P3=r$&lg``gTh5FX;a^2=`RjPLeU7F!Aj% z#Y7WtrcrzvxIPQM_8W?zBoxB`VpxNZ(1}KX>q1+EUt&d9V%GF;#JE3<*(z=5w;)HB zyEDT!M0h0tA_cwlo#YAV3=*k{Mw$yk5?V1RA+=Mf&Km-Tqh;F1><9NT;1|E#F1KYX zVn|yJ;oe`9!zOX2OxSl$kCT!RTBfgk^x`x5Tge6cc2bOzyknV)JXU;Lj@twZH z!)(0?lL~-|xm|Kj8d1(c0VZO2>Q)X9Zj?p8q(c-o0FgES!zJE+vp%=nHe*YWMlXRQ zQnB+=anTyKD^7wQLP5Th_;1~nM+#QOwxk>pDJZ>>bk$75f!oxY^?7e5t|=+Sn1yq_b$csU=`VsXjVyp6+~iz5U5!SUV?1MN_U@@MqWIsfiwWr>Vnr?L7Hx5wvGAtUNbAW|KhH|MT`}P_-_UID9PP(EMT?oqUXIWC`CZQ)i zP{a6|kJsDYF_p*^Ea-v3%YNV{f-`E-n9lXRB1;;~&gnk$4%F0@?z2A&?uWCIX!Kky zZ&Q}GVnzt~`x!ebs7!q1m+6rE_GTTF-^aMb=kNKo!Whdq&$XnWSF<)YSPH&MMz!k_ z$6@vNO@y%V`hGrUBT^@TiW`e!+A|dO6xeTwfT;Fh`B` z2QR`FoPIWewGtgA`{4dn@o*$NSKP;^lNc(SU-3AFqJ}rkdREYBTm{M9GX8pnpkbIB ztrwgjd3+T}cJn_!pT|eFsLsTtc%niv=43)hb*BS$H7Mnzo{FRpFB!u%j-1DIm#7nI-bg)E` zf$uyrn(eZh1lTrS091>2(pCF4Q#K{-ZX7WpfuN+RhV!w2ln&@ri2EV3Z1R8uCW5oL>y@Jbx&c1?6F;!MNWAt^E`TQ}P*#-!!cs2ELN39#CjA)gLM3 zAhL}6`2OJ407=5hkjXYJ$D~wMzku_|CF5K#dm&_e%!Y$y!yQn!W@+df}VYHk?3D zN*}|wG~zrs)!cD6d9g{A=jm*5N1x-Qr5M9gjgfn3CkGc*LUCLyh?nArKwp+YKKi^- zYN!!S=@lBEyEF3ohet<|K+ajPZIiIn#vpHtWBqO2eM2BNf9_BLYm_*X(HTBo7`YNR zlitr2$hX={-05#*n|%>!0b%H*;9fX^F&;b)^Xv+_`T-ox+Px_H&t%eL;ptdsIsTxHZD^Qt2ay2wgRwVt22YKMJ> zpShi^)EQDxe3M>UV<-KC3OpSk;phz4ZQ{7Q`%(DHI}S`@e**#J_nh&~-wpjK%?j?) zo~4`mw17q%5s6&W!UL|wRUe$HH=P~!E~sXNXC1PoC|9?2niD`N86rpcW2MNi`s zjN6!Opk!q=#TGn)i4!!uXf9x~4OQySmCIO4zTx$Wf7yE)#Vc;6z(es9EOePhS-lTY zaU#o%GXd}zV>7ZL(y^SN!?$KD=2Y_sQ(ix&zcK?ERuzs~BHX`hOs69?#zBaPOFvZ2 z?|3s1s&lo(>cnW`sh*m#Yc~qo0<0%-%3#Mu7^c^prKj>PTwk`~!{ZJ=Yr$u;tq80h#mUMiH>VLf00PgZ)d#^~a z5Xy?v9!F}G+e<`zVS9k7g%}9>Wzrl-SA6wdmcs;T`tV~*zDg{$lBuTjXheq;Y6c>)%ji4}F6g6fUh+rx{#jo zSSxYw?JpOqURmw3@c(eGrXCuO8e!h^P98uOHZq0Sp7rRLKvGgCbX9SwERCKD{F*rE z5ZjIcfz#KyAs-K4v01UUENj+g+BGX z?B$@Zt>6oBJljmBhC9p#f(~xQmuKQTouPXR!cj?>G0caN%QI+&h#6W;>Lq45N?Le= z-?jyHy>W1NFPaXqB(Jj@&r`!DL}dK9foOOF36)T1I7JXn#$aw_e0cE~!SQ<>Q#}*q z#-gWzj=%pS%x zue_KFd(XRA6J*}Pd@ZCpfp%hA_j8RGITlqqvu({*1H4iCCQjutK(jcw6dzgpaTmR& zs#bf?{9H@m+fmNcIE__etop_S66GYR@yDjz7Zxwyo~>mei-E)+-EaGv7BCU98=wk zHXtw{Z(?c+JUj|7Ol59obZ9XkF*Z0j3NK7$ZfA68G9WQGG&VL0FHB`_XLM*YATSCq zOl59obZ8(kI5jphARr(hAPO%=X>4?5av(28Y+-a|L}g=dWMv9IJ_>Vma%Ev{3V7PI zx@A-x-MTf3yE_DF+}+(ZxVty*G%f*x1p)*Q?k>UICAe#FcX#KKcOTv7`+G-^uIe=( zoAs=@)~FiQl%&dPjH2d_ra&o22M{AG6AK?eUP0B--o$~MQ59(EW@q98U}Iuo;Xt6I z6n6odfUF%IBuqd+J^(k!3ZP^L`Uw5-U}51!pae(*9e^$$Mst9v7eE0BGEw((0V zO#T@tJGz1xO-)=s%s>ZAYX>0hM~Jwilb4INr4{IJ4t7SyznT6?{C|765giHb6s7QdJEgt*WGm z$;r{>Ke&jiscT3x03<{e)g=KyO$LCphMM}{qdL&xgTEyMKvDg}{9)WXHl{+|K>8Y>XUiI17t!^4Bg(#;jb zSAd73i!I>e>;kj{{?iyY2lJ11f~Pb8$y|`w!Nx2!GorVg1o3kfVzi^Z(Jct%IY7 zgU|oPEvy~PE&g`i+|7wu!@=6w4Ja$|AKo7j!hd6yKoEch0CWZbJ?UF=Ql{>^6XDrM~nG*`9;nOXg-vH!wl zK_(yC7j>|-1Aa`(Kaj@XhZk z8{v=M|D_9HmQ_-a)>fwfKbrZ+P13>4(cId>62QjE1u$`OG4VoR`49pdCnvy%^+VI< zK+k`a0$^rxa0GpX0G!-FegF$c7lgmF$;Ap_20FW$*!_k7K{$8;%#IGg{~B|$0GL4@ zj{j}v0Wf<5UH%39Z_3Jlm+U`kWclxF_>aBppRpRq#nBe1Wo`bkkN*!B1rv~qwWl7- zhm=@9@Q=@b{xbMq1t|Y}75tBAF)>F^A4U!yb^s$AFB^cBlk4Na%EsaMzhce)wXOUk zW##G=aZIPR#I_IvG)9iZdD}W4qu9pZ(XW!o^h;Y-a-cm5W?37tTaLym$*6< zel%6VVKLx~Xgoh`7(4lL`Kz23D_GLGyLV0vgTgS85X4xrK1T!gRg&1ls03=@a11h0liCZFl3c1+$5rbs+W)+j}6Xmt$X*_@yWKK)@~l5zLql zHik0Sf8ZmW+GSS10FS%z2STJw;_GlFB*;;>l%M$$!s$TNJMNZ+hP@T}kz{v<`k zO8zYeFaU+!@huvBJ~5G91uqTG$UQTuVg{L7ZGVsMcRM?kj7c}G-oSX!BX%Yyd5h+` z?X?61YxRW0p9-bzU&yQD1PLyOqDPS9GV9KrWq-D6X&dzXQ`hHTrxktU@(ESd16ny( zf^ehOapW_mrEdWp03{B3VwBHtAcg@M17WJ=H36Hpv_d7_vI&FE+An)oy0_9?DJu!>+7qOGd*Y?4r64C52_?16cU%H4fDVQ(zJBXvIXwbiB0$_L!8P{0$Yz z%Q)c`?GUS~u0$FHc=j&;z`Jbt>%YjhP_U5%J3lo!MLtc=)s$fa{-jT7!RKb*MD%s=>`h;y|#ZlPO?67e`o9q2+ z(S3977=5i4@jh>qq=(px$hsy$v!m0~ zvL`n+^rXQIF`>(h(5!?wzPFv5CW9u77DZoi6vvA=b_6KS2}KE%ps4)cAyIB(B3%0D z3;3}<2az2xKJ6`xEF&>fMK*Q5(@DD z;B0k#Qg(s7lTJ=xj-!r4DxDSlBy&O@?spY<+M7#}wF48SnhE$Mxhl=uk0r^x(Pqmf zt)f0WMw#skdMl-+37|r66yP>&3Y>GKM)~eBR)R9{G%t06JEOUWE3dC%P+@r1jE=Bl zH8ra3?aBZ=_H~W$(-I<=+~=&#@&V@!`pLr>1dmrwx5+-r>#@oqlVCfkojqvhL%P22 z)(0cY@vMf(SrMyo4mk-b(_rVbDpWz3Zx1_cNA-^{bk@1u*D|5aIPsa+3kWw#=LE5* znotU`adawv`qlwl^FgXx8mlDfdx*=uh|Am2gK@~*hyERyOqgD1v{vHZ`2t>7wH^3W zl>OBkdFp5i5y4y7X&D3yKfwlWo+%Rn)~_y>UHR9d|^b z?n_>&)4E1;{YIca-0vb|OW&u<*&v`@qyClnSsf|lPfg5%dBI(2jBiNKWjn;AmW7q- zb1R=V0}*yBLZycL#y4hh58VEM<)uAb9*E3Y8#t}VCx^=g$!JVFzZePx!_Uwk(bKjQ zz#TRlB`MvpD*Al|7Z!tp6v=!FysJL|$Ko-tRoxn+*~k5yz2(J7=D!+Lwu%`+gHdkC zV9X-onMAY)EIZBjA(eOil@e4Zec}9GJlyk2Kx-;cO=JDeFGo3*OFYy(#7Hs|giV;| zpAXyiJ?#g_u~))6K`D2t>FG13XEAdzC3+`b7>_qCXZ)o_FW(x6zD+YeX|oWS^;ep* zBt9EcqFZ2BbbsBNKt-$Cs7}0Wlhxu_Wx>Y>TuP(5TS0{D6r{BsoR{rqj&q}0pQ zZfKBVLzY(W*--IVyCVyM60p5f7&~}dFDVN}2Cu_qQo_7X0tF-14`3q#bK#UfI zQMNhv(aoxrZb+|iV5dx_L^qS$3v(0$7$Ftg*B#^5+Q78k=>a%Qx*I6jQyX^OdC**~15-(<|S^Z?$ zH#d?iX5#katLx<`xNs|f>a~H_n8gbU0nPw96q@YM-p=q4u&{`kxmJlyp)TD*4R~Vb znwK19lq(hVlM#EwgEntLF8_P6EmF?5;Oz&OrfP+zTv<~KPwQ0=r2~DV(s&c z-9wXZO}MD*vTfV8ty#7`%eHOXwr$%sXW6#-?r-mi6>%D?ak4QY@(*NVWIoreRj;ZP z49I*g7p<7~2}PnAdymjD0_YELK+7(B$v$_AT&@2DM%uy-5OYb$SFLD;AC*QH?7(%q zXJ3k+I=ujg!SAX#Y1!77tQciy_;gnqwsw zHA34(0;~F^C33cE6|Ynpuh8E3u+Gm(SEB5l1Z6kt6pq#t%obZRWp>GNE4(dGKjgI%%NiWP}9saS$v%EX8^$%dfvT|a_((>B$ zRE@QfHY`N;VZ$KAPm&ye<%BQ&chd-!M%)FRhxFMENV_`<@rfrEGJk`jL6E8-yySZh#!5m>66^MAK!t2|y5OkZvfu<0{O|>tStE5AzTSV8 zs+*Yiu7R3uOvwx2PmU8xzE15@mHZ|BPVuux+Ap3fsTz3tO&Jeux+fu?OT9t@BTpC4 zyoLfl*wA}$tAfjSWqQiTptQgL8d%WR8Prc2?~Y~o$vj{_)q0WWNVB)Ax?`k0H21Fx z5m!}U)~0+q!VXndFi#mY{Kav3rN-1;Un&h+yf4X`KoYtvjXk4qD+o4IY_PrU3}lWz zUD-G&18cy|S>5PLH~Mc{%?!YpT;cLxdE4{e)3bi2>AQYKKB!}H0158dww-SjuqLiAJahg)GMR2{KkFj zV@XYlpB~#HSb!n=ZRo;V085SqSB9Fx??{$RQ`77(@%Z2)H;gVvt|se>3%7B?BXH>o zEZAF*;v)u2_?&rQEgwJ~i;`+V?zdPmA~}2xdAZaAGcWJQg8C|N&$?bQY3sabeMJVY z>(6ZFAS1HdGwkkO2{R0vutFq?fe4?}dS4X|qE%Vs*GRu5T?u7WCcNeu-| zUl1Z8W`^!U#(tIHR9i|s9;(1axuOtnQ+^HilGBBiR1m9JJdXGj<-I*uM)!u$Gh$*y zoUfCquwjsD6$?n%yLw*?$iio1>lhBKO+sqH)(k7!qAdE)S* ziYX9I#hSnc$NP-GLK#G}lR;z8-lrqFRtyZPV~*lV4=t0eW$pcf@;`>|d^ksl}Tuv~{A`lExI($KpGKV>YPJQbNpSsT1UaNZ~N`m zg|6-5I%SC5%Me#FsHw~uk|Q-2!4r*@HCmh-vl2K}FI&XbBQ0Y!(O+~#XURI|}wB_+_EYAE*G1P#a8 zgu!0v#>ldRofGcOQ*GUObOS`E1B;X+f}dSb$MWMVIv(}ZyotLyE`3H-9kIcWa%c+) z3Xv3T>U8CP7G#J@iO!n2*ThI@OM8ekk}`rz9IzC}VzuS*in~(Er$@g*b(vSxl{9f1 z0aelr9}>R}>ejSoH$2w%2e+o&S^_#&v~`}^Yn!=acIAUuZBBHQ^M)DBQKmkiu&5E? z`1j7Q4B9ew7>l`&8}f&mJKdw6e1?lmJ8u(=$ahDZ*R<1@han0(VFnEWJ(Y77xG=N^ z%p@hYo^51%cUW(7NDO-1%XV{`;1>Qk%Y&1ROk{3fCl^B;l(z?F-CV*3*7u=0xzm#mBob)Sg-_u5@^lc0P}u zVXYN?t~C|j$(6IOIy>YqCL*mT15%+IWMm=LpHb01y6lpY)8~t{97_xb&*QSPdq@hF zT{~D;X10KuT(!ZuW<#j&nQle*$UDBhSti{$3eR3zgp-}%fv-e}SSGFQUZOc?MIBK} z;JDXxw=-Psu9UUPMy@O39f5U+H(si@hza%(+q>S&*Soz}ns>sio=56&^AyeBh9B*} zI0aeb?;^hj^Xa?Z)XCir5Fc*W8n1^5kt~&A{&3lC?S<)~rr;Ixw3twKuId}pK}D<} zMDNq48Kg(FYz*c7&wHYE`r8{H7(9vbqO6VNYAXr7iB&n;O{F{Arc2MeXPkE}I07xM zxjY(=((nh1nC0cU@uBqADL5pua&NygFCpOE(|ekm0`oeaos`tW<^S}qoE`bgl&Rws z-jZa)8tslv%lQU|x*03NYbKsij-Lp#nYdr;wv%!rA=$P&O08ge4L@#%(==RU&3A6XZppdU#6_6ZKapM5Bwy{hTP91D z4@?wOWm>eaZ5^9#-r&--be7x-g8g0eZfrfsksh{ND=#j9QRnS(C#Nk z#%l_V)j;U9JRofZMYkDF=`pvV6=U#LS(obQa3bwZg?Rhp1iyxv-}*;whF{YBu0FrB zepc_c>YQ|F4JQgS`Tb4W6Uoo&ruoF_S|oy(HezQ0jpBohIz}Hm;a71jlmb*m@6u@&F%TiYRxJFR7PcC?T^W%@>S9n^6 zyiMA@$KJW>B3R2~D15zd$rLQCoo%)bpIj{XHJ8<~fJz?vao~<&3hOkqD6)hB?GIN+ zE4I|F;~1HE&iv1g;Q2e)J@+{M=fL-RO(4C|G?eh^RwcIdne3jh#9nZ)>xkLkUJ(2! zDEPv0*1KeFLgGGXpOaaksmGn5L6x6aT%EM~eD{{jQUbA$-jBxB;V|3OdyMn^7Lqt| z5a2ZNi1{wJ)13uj`fC;?QNN3S1Uyt6>@HF2;H+f0CqrlNk;gu8lYtS(vEuY|_n8R+ z7(B!_1_oRI$_$zfN3?UMQYx)1&JfrGKCJJzW-P2xf3;HhPsFnv z%-wY`W^47Cx5vFD_t(8HIPShC*_)dWKlZjbElBGc$)Ck6iP$rTow4d3n+F6k1I|vx zLaFHs(cTiDPZH=d!WPMBI@r|C%JBR*gqm=zGsPZ@gd)9_!q)6y{+0xgdZ(7MD(L{; zij&K!86IW-^p{RKJ&JgO&bI!Uqp7n|6}3^jIe&g^pQo%JWyXf#k7o1~FVsZH?7^Hb z35EB8JIe~=$UhpcDR^zp&>{tD_~LFWmS=J_CsD;4r=kz3H5h1Qkw!EcP>ZI<2trc} z4>@MZ8o%sQ88p*#e4le~%PmC|AcD~_$|&g%DHN-pP&-B7ZbkPYd*MmE1qlKyu36|$CukBYg zrVT~_Mtx*S5f)-}?rusY!eqP3E6k{OZVe0c@boWN8YGVzf$>~}(zP=!DXz^J9bV?F zlG~p2GS6H^sd$JWhdW&q^!Z^0885BovGt0WkylvsM|-Qos$ha@B98ML<^H*I0mfp- zA7V3v7Jh`IS-L+(8aNuDVQ`u@fWQQ38vj;5Rio^-drI@Z^(`K@QU4c2t$VsvF#AeJL{S z*F_4*LL!n5eBjKGmFxKB+|?Z1+dkZD{6%F01d{i2zA-Hs!i2R7GktHV`OYQs7^wvq zzt4)|?)=wFPY(EV$;_tR05hkUPb~x5F z>Y{-$gY5AbAZU7BQV|h6H*Ms^4Q>VStQv@HmR1n~;LjE3PZo}&_czbtpLWE{r5 zY+aLRL;seo{0{B$jXb(e`256Fk{-T#s-WMQ(#9>l+xh&4Rv&g+O{~0T73|`uw!tgJ5*wE6!E`DLlT*K)@}l9@4Kfx`qBt43hrGJNI{Y z)oS^ao^UspQT*zfrT(EC?9|jU4ou_Xkzqt@kQ{ICRe@6V5Mi#-KQ}*!b*7tyP#Pmu zFh2F>6ey1+D5H|ao3~k{B2?#X`F`6X@c3U;K=i=vN2~X^5Dfv zo;YeVe_l6eN5Ao`WHL|<6hk7;-^7ctAGJ`|%(CWfEN@eROB5mWE+R!!^U?3PVa3q> zE`5V+)!5Z(@9w-fLo21wA~beg*Q;Kj%1{D|2?O|AeI#Ab9G)aIDWL|y@6ZR~y|@aj zoLPR=RAoF9R@5q$89wad8Tjg6sPM}LE3pr z?;uIjYzgj=E)bv&lTuz-5x>v*pX~2j?dUK43}Ha%gSCchBv>K6xb^dU zC(r8?yZW^(97X)@s5zK0K^Xf#Hj%y21!h%_z*Y`zpx&axkcYl|F~rvYt~_H2yu+Hf z%05{|to(M6OFF@G^~pYD1-D`f`TW&EXKPQhdCpx|aKIm?aJ9o>PN$%PLf`~ncYz;m zC6Q=QB!cZOI;2Oi_X6%S%K`kxiOuVo%K1TO*uv(P-^fT`C_WuSE$!G(r9UP_R-7k3 z_^V~kMp^z#_Q7=QhyXVML{(gmpHS-mEXL%NN3TzFBu74Z=$vA1Na}F8D5SF>XIr}r zj$m-_+V(|>*pjZ~RjiOwp3Y!T{H7{wQi;8;2O3aFf%?ZJ|9L5O8$)n;0JWUg0;c6# z$LDT!V%?auS(}ZO8pc=5Gp~+GxuP&rP3L5;jL@$jlW>=kSl5j+hFh)dE3z=Pz^xa< z+CR430nj}DDd7r1xAe2)%1u7MK3^q&2!|sCPBshwIlRW(KvQazuwKj;rL@S=fJ&>m zeam_V0+Z2FBINdT6>B~axr{P_Iq3p{B0V?gP%&+~3G6S9 zACRw3!{M`wPXfu~Jj`(rTRkQ49K=z#tV?7sU(@WH9`m%HAU(O*S(rdR;6&d$gHY3a zmrJEl)H&ag-GJ{No*JO?aA~qm>=U#;X7p;kbRU?}YsEJjgAOzXBjo%0WsV0WuAaeg zRAg10jEBb&@4Bw{K__QU`->wo$K`yd*USO!K2l%L0TeqhZXu7f7Ihk-Iq>Odu;Go< z*WTx9VJ-^K#2(=0jY@&JPTr+gQYM&8@z+3tl1zK01@+o$WuMwi)*2auw(u*|B7MF+ z;OE;{S8p(&rPIZgBPKtC37B!4En7*VIixAa@YTgOZF~uav@Y`RgIq~wQbUWSiQ$hl zuVpEWncVqP5Z~Vi&LE49I#pwv4DNC#K3A%;Q6y5~;gv{vJwvXonhV)AlLb&!C|{JB zL5K=qv2eZGB3XLxdO}AL1 z@8V_@0bKqkYo67cGOZt1l*$oPI$PE3fawE_J(OSmk_KmDL8fvpLLFpK*`uEJm=v)m_15BDr^1? zPKyyFz1lnh3(uksPj%c4@DEHqqG8>n*JG{>60nvXW}z5;=8Vc0S+Vz ziWy$gfpX3ce#k=Q#Z7dm!C>qcFg`k#A}Bdc>74%={XF?*rh8rpD3qjzQ56J5<~FSt zauZ&7gNCsXY+z^wT*PG}oXSVYGF0yHRsh0&R1zBZkqtec&E9={uZ zl5m+$ExZ&A1cG)+3OEe^>u4;kiFa-q5I|Bm>F9f4$bDSMPZAgONil z&cWI`w}}+S2{q7574vwcp#!7ur2{iTcznfu!VzAFB?I<22JN zjJa80FQjHAuSMjBYFc0kq5L`7s{FBUi$w`clg(xlwG!aTOT>4Fr;E-4;4g>cvjjoTL0%|a2JG!%bqg`9{E{aqQZG7y6uYt3n?uQ)xNVuu9LmYIl4Np$tU3zA6C52 zAEVKx#6qr~!o$nYbdcW?KZATYI`MRKzkq;;-ieBr*kHq&^7FHq1Hn4>biP8is**i{LqZS(%kaxHrCKrtdgyH|Ppa1J2tY zmsK>_cw0H|vD7s!?($saTjgm(Sp@y&B{?*V8}AU5Smd*-!z7F|vE-Sq@Y?qLNPn8A@-x)1;zq)&Vy9X2 zFhZjY>@dIP9;|FV8NnV$QPQdU$t6d8Z=3vfHkz2_82VYl!XV*acpY6&jA+GEg5)Ut ztZw(})wfo>aH@AHUVb^Be`Y-)jpF@V%vQ(!V2ht-8FE_qvs|H5nlNy>>uHHQ2Jf5t21ZNX2MMP8B0%;Z>jdVM|#Z|Vz*q>RSkU8 z5N~}-0k31J%kcg7A3lIj3JQyq-|-X_dnG=m0);2>U1h~n}&vc6R=*Cpud=X4a~)XZy@l;9V! zxuILu(}PSuP{9$o@Y)egG^Y6D*uwBWF66TaHTB2fJ8d$ivuY+RbP_qet=6#Zo>^qy zH9sM|Fj2IKyl9W)dC62`WH({*!5WzRZ7-k@9z4Z0(UeB0PI;V1l6|j$`b54@4YAkd zu^8lW=Lh}-G{dJcyiAXWA@{W+tBrSSszxAE#qcsr*uCP9-HV`x?R)ALGy9~U7G0=y zOJqFrR|DO;Pt_o7aFV6A7K)=JtuLx=LzS$@{@{0yv(-pafcnZtnl%$RVW>$xLG$hs zqzxsVb=TeudLhui#~KQ6XZm5@d4&Sk8h5N@mcoSW5#i@h_N+1bO%$V+B{w>!U~W&O zA>#LL6g`u3FHWKi|D@|J-Hh(jO}$-(`-XI8-^yX>!I$Fg!ej?EP<)zL=@lkn;*X?| z)h=l(2*tFBK5J}<*^i*s`@tUe*d-^`C$zb&_hpPJgBeiSj|WdVshZoGi} zmDHF^+b{h<;c@J?#&CThnk!r#$z;TMvRhwTsCANFuDC7WoJ%;_Afy)MTFR)m+g90r z0$$)p&mV`(R<|-NGC|A^SK)q&-bQRmWvhn@vD zb{qDF((CM=?Oac1k2jmr#p&dL0lz+OTky&w7dMdg8wH&X_wQF@YLtu5W_qAf#I6ds zuL=0*IubP8UvdTL#Q@IqLb#%!28@8wf*d+Y4AwZ$lf?s08{TyQDCs+rtxr>|0YSh# zX76<^yrpl#Ldv};#vk~to={sVc2<=+(!S-y+x3F`b)1OeLDQ6w^P6M1F~TNSPk8(3 zmS2AgAVK6tqnN?Zz_+j=K4p|?p|!25i$~HFitd0vGKkCyk7UJ!f&_8?T0p*2A1zPw zYbxJYO+`uHjYb|&$yKA5#!L{0W7K5`$Nl!(eQFFRO6vAmbLCHFo!}^~=TQ7Rarod4 zlqY@~%eHdx4mLrp7@wDRui^6dI0qPuF6PaEuq<<|zCg2;?T~Ns+`dX|+FsMyt#Aw+ z5pOc(jHiQ=>GIXR%d*<4I5!x$L7-~#FViqY7N;EhxdV4WWGL06d>RV@oQSk0%#)4GY7iJn+!r@2`;-&kescjN$@Rs z@w|mAX7n1p!5!O!(9-8XS4w>~RVl)euo<-o+smqI5^oXhnyQ2gSv$ALY3PzAWYqp6 z%8^#Owho_~>IY^*xph8ol8D_iGw~?Gjm7TviFr_w0BXdf2_cAdJwHTu{4P~c{B(qB(?1%IlptQ1{cboU zCp@=7`F)35`MpvyKMl|>oP7-mA`XjQ&In-8DKb^}%cTYD6vAtNkHivsd?Tsl+Xcne zq#vY?ADRMO+cV=HuJEZPQY7v>dBbLuna=f7USR~`S{I9d>309qWo%_HaiDhGXi}I2 z^qbqOd0Z(~A*Fn_9&f_28Fy_NJ}p4hC(nLlwWYST1~!`Cv9cCKn%?XjjRtKfutSzY}!25_3MizcSq6C;8;84y$qPF#hbR5q&Q<5lvT)u!CVX&dHKhV=1$9o1klSlkjzBCBY9hk z|GD3~%SPk%3&&Vf8+_+Z>+nhqHlwN$=NiQ2V2cC7LumRKUnEJHy2Z9vX;9zxnCg=$ z0zZ4YR7sU(Sv7L8xX?gkmjj^p$V7EKFa4optJMw-S(~WS6^S|zh2&MCxR+nhY737z}5M%l=w7=s4CG$9M2W&vAgoL8haD zH?e_U#AK4NSD)t-LeaRIIXS6XAAUI(wZDzjoI4{wN^e{xW!(Q187K+B-Xt!j4+#lC z!u6LfF+gF&n&!;pU=jTIviOIPo+nL8!-7RA*s&Q_CC?%?RLmKCE^#~D;@`GN(@uD+ z?ye!gKxNGlDHOlRa=Yv8?yxrFn`hAD`c6Vv;ZVi+)rP=*ohMAOD zU~3qxt^fqD5#y5hfyWq&xKU>;Ch)QC>du$DjvU>JdxDIxTQxgj=b!hIMpw7fYE2mi zX?45Sm}dlv^T)njmYt_7NIeSzb?a@#s*b;?>)g(ecb-ER7ys|}O!t5SIE?K^p+~M_ zuic;Ty%)QnYQwFu%$ObPNLWs$%*j1%tl1x&d5VSf}qPSs)u$~HB2q#g9h(s_g9-Hb8 zB_#9_qP{!?`jcP}$#Y1yh_BpdARWf{5yzj3;?7;dOdPr^I3W@4d~9)OBx zA%~png*UK>8Ygib?Y539WlaHu!rB}8d$qPi?l%7UBYwU!D>kAHTIqAW-)zqXuBD#B_tld8fzGMty zW_q^q8Vy-)cU6tDUl4z^>T`e#pL2t=mQ#!e9-A^0sP+bS&u3a>{DCegUz1{f<*Hfb z`V*y)i8+gdOO)HK{!3O3MUa@K$z9lqh=(pNWbVy`{RQSji?CrxV zv@D7BvS=~}a%Y2YK#-f|8%B{5!GbIhWf4ox<`r5I3G+0{mpi7Zcr9T@^WU1T%4%=2 zqg2rhUCuc!J}HbMRdI@$_Ept zD`7=UQ(n*|H(~uRPx0MfFu#O8=2iuE>mCZ4XLXAmEM6s^V14>ABF9~P*tC=wZ0xI+ z4f|D*YBv{IaT{CCXlmnoDn(|2U=0A7#KOZG5%nriV}5{UV5ww~5f=S~Z_LWPN|urnGlVf*oMF&Vj)aswuc9a4vYkzuu7zQp zWp7cQxS>Ws@*E{JR~H5jMPD1K~Pv&uCo~7CEQ7+#VJO zhWU`rpe}D>>kNn?CB5``n|IG76`#Q(x{I`+{^>!@rr{2~@ZMkHY)TJdI`N-Rsy+Iajm6r9kokCHAbgt#`sHZG!gav=Fe*G#;Nc#3PctrnvuNi;PF=6qZe_GGgU!fjV-^kp1;5 zB(?s2;OST0-DRl{udO69v!J!4LZm)oOF6GW)-B!bJ{y8B-IV^_4|cfI#HJdgeNgGP z;SpZ7Uo_Aoi%olGyk%6g-Qcv8t|65pT=$V&*!+}BNS3_vTHPmB6bQ^y>ZCHSa72GL zgj7P0Q4s*3y{y>!96A%qqUz-poX6U(NA|N?vY{M=moP-gLyeZ{4@^f)`V!PnjA}e) z-q)?j3XZq&KbRyn5cOg!p#Hw4;|oNEn(2z>Wjag5XfwUv>1yo^P2<7 zt1Ijm1+=0Ex*v~NnKEr24G7NL(vzy5pfl)h? zqqs3m2e1uLN{`NtA93y2myMBedWI#}{k322qJAfT+STmnf8>tEFB23s4}{4jz;Ul+ z54v@3#$!Sul~h1^q28>D4p#dxSjw zVl@`*U=f&5!MvahL!I~#Axv;D@tbkwAlMy`S0divU z&?51^703(w*vy6-*=92RO*D^(BIFcLTu)!C8&2kSr*o7O>Wu>}OlzAe)!q2XoIA~!F3&Z)u6m&qp-tZL1ux++mHYddN#@x~$ zvOR1#qqCiQgq@2Ia$0Na`nuL0n3Uq5ph9qSoqiL-6u7YT1>&T zT<>j4m04`Um!6T;h^c5994}zya@YJHOySq;Z3T_yyDpT%cLbq`oaSg(I23DXc9|`? zclA&TeQVzsGsVXiY>nZqY{@Hs0e(tqR%+8;1WD|TW>?3>#tV;o2tmD3VkuXPn&)B2 zht0E)L*Kq%Gxq3eBQAh!4+fEhhivnfBK?xim{^4>{XE3lt_8o(&FsL@Fzqc`0iv2d z{td-E68alMx;@o>m{ggIc+XHV!~F;M`nF$GYHE_>upQ2i6F#ib7 z>H0ieVs*l2w%aYqBafP|07xIlqxMJA7`O7}COmHwjByaDyT$=~E$`hx={DDMsM zpy57*E#Et=YL*Zs9Ro)VyaFHVX{XxLNJZLssc`tale3_}J8ko$9i(+`%W zXUFBY&UMeiJc{XTTHYG>!1cP{H0Oi8(_9KFHC$dm$eKJsLY+8)DjBt+QB3GG1Zgcf zsj)28jIK~~hS*Zjt5qoMWmqNVW#aZyN3oABs5xtOy-$;aW;UNIm){6&ww~kd6rrc~D=Nym;1Il|<{3Q_=!}!xpy*Ke?_2VP z!{AL^rUt4EWx z5$LVI304qR>JscsDCLh4UPqdP#uRekFJ*^!8cn6$t zqh_8=XS5PFMht8@H?iM3@&tpw_(NCJ`Ne)lsigXefQd?>Qn!JBicCWb)r3$6;S-^ys%4{2*NHK zfp(Y!91PgTN`986Ax?nnFnDzjy;Jy0)RuhW+KMH`kDx(_JM z|852`bn9x#Og1(UU&jc=kow#*0r*vhNiA!4K)=cO3M{0V^u2*N(Bj^6Qh4x+c-0w= zC%PLOzF-@3;G||>CK8$kFf9tf1@0!J?Z95K&vk4bW6=tJ5R)S>tI(UniB6pBo0XME zX@6@pt#cg@B;mf+H4v;?XR;ty_V4C6iLF_qKc~Jqa?T z{n3WC>PM3_z?qx?Qh`1{8X-+ek2^4#p@qB{3z{gH9v2+Un*_Y8WP)7Hg(Vc`+c)&( zs__SR=P&{7Yh1De%HRPL#-b`7NV)It$#I<6J1$-xS^s z>~FRI5D|ObbcuZSm!qQZ=+0gwT^{!97ExcPyT3S3wl_jfDy9571@eUo@x&5t#N7%c za|4V|fxUsbNHWeF70hUAa12KtPk1C{JoWm)rz$NyX~HPl0(v}G7k&th(>Ny_NOK}n z+Ufw^W`sqn)dXv|UQneYR7_mE!YL5WQ?l_o%p&+KLSh?fR}bbB01-a^H4at2H>yH; zJ+D6w`@z@uv*}xI$5XS?^eHSl})FYX7{e9Du<4PC*8p1uF-<42I6y{j*5La9-n* zw^^hXd$+}mNg(~++-c3g)Wd-_yH=39H3|L3lFv9wws)6JBa&7UFg8mig!XNhbJ2VL z8Kgx4b90Qs+f0L3l~l9Er6BFwoCrw&$i9SbcGfDn)31mO!`mzei$2tLm1z54DlfNB zm9P^M>{?=PT3qN|7vf1K-CE~2u<5fK`h$HAC7$1NxK;i07d-*Qnv9Q_muWs3@=t#{S|zb8GC1^%BU?tkk9!sCE#l1*LCsML`pB zO)fzromYPdlWHq8?f0+>pkr+uy-xT&g7OyVHFN3g+g4-}_d&}dpUaOERB#s7HbRWY zPc{C{*w1SvqxeHI(zMm}t@mDvU=Jr52B-XcSLM)}$||n7Y$r)R zZ#>8=G)HV|$64^*7Hm~D;L*qQ&(!2pKuoeXtm@UHr=isvAZjpB{ld(18;HIR)_S+2 z9=(Y|s7#r4^jDp-)|ZENIf7Q70|9spVv5tZ@TKjqvHN_rb<{-4@~$?d7>;Kul|(~r z9>N}`8xk^Qpv9DDhyEJ?_|Z%21Ce1Y@Ge2p;fo7wQ|9h0Y@*%C39X>YKWEh=m#1$ zaZiQGEkWEyJeNcP_P^s)K8a%LY5Z3WTcMY_NFfr6(`YhQ2+U9991qr735rW8lJ|f5~ii3BMH0d8tLr1<;Txn zL?qWBdf#kfh%&|`tHWZIM0uH6uUBy0`7ASp+WHjG>bOC+oLH);+ z;Q*YR$rTnrr%{tPv}ApKg>$r;nfd1Pau93jw!%toK^rwUQfhz$MZc`(X7exe?^LO4 zc8B#S{vEZ8NSbhN;yRI-N--CsE++q#hS`9`HFZmqh~2~|2O)G>wlq-SXfzE z{s%Z>A!6ZT;$;1w?f;iPV&P`v`u|NIb(pH^ZngbKA9bsMx&ODI5QJf%6+0l{35Ytn zMIE4SG5#Nj^pxm(+v%O}ytlV*(Qq#Mb?rRUS%prX8O2jvk(mpvIE8RLGch$X1gWT~ z+T6?z@xW9COs{JbgAZ&YOGk$Xm6CE8CuD3+%bO6=$eMsSaAO0l$l>az;2Zxx_K~pA zxO+MFzzApolnck&eFNlw_E8J~gvL#NE{0-E{_ekr0z)wW_52i&+ne7UTACVMepP~V z6CV^x1viakfw+xTta&*pt!V?8C*X+2OD4b+o1DElwSg_Jg@8~CN*iOj2j=h!fL9_D2pbwYnAs|9TV(YUL;tOI*1yDGEzH#Pt( zpn^shx3)85GdE{v<0c13C$lE!7Q?3YkLdic(p1!e*||lW`(GEZe?S0~F>sSthVo|e zdjs%a8yKO%DKsc2kYA~!rmytnSG}pX68G*W0E#!}(1(ItwtpySPQU=b!r1cqS)(o} zDhgy^Z3E^Q4h%b3b`LZ+HdikPDF3&q7XUI_0T4%mLUMO>`Cb-%*kc$6*gn}`8oAxk zF1LRi+Frlz@tVOoIX-@ea{xMRRCp7TQ{uQSol8MBKjl4?>) zx*&>&zV-vL&~J8Eftk6xc>(A@<#`n$AcO(op|Ar})4+zFgvwG0nH{;cH@w!AA%c%m zn!Q)YHfQI9-~Olna7@l1J|F#WPSuH6p9H6}J2zOZgKF?}1Da3(s6Tch?^|ZTE)Wes z0SG_^-Ie4|`ELGl(}b*7giqaYy9U=c*TAP^9P&c;w9flqpa;>(jkHMn*XJP5zmE#z z{v=@Pb^)2Gl}}#<{e&1H{Z-RJGdDmz0G58%kM=8niGb7kVCgSE>Dg-`4gw(70&bAz zq}-=3p@naN^W1NBdP;O`Y*HhA!f8MG&mEVs4lQmxAmY0oihiH4j!oJvu5FA1aMJ9Q z)Z_%2RjJ9Ht^F&r;lF8`8@p$2fM)U+)MqcHUpC7>E_wEAyKj$Q-M^0Ezs?+?7GU^I zAgOGPs2~|NAoq$9^Ik6uK=NJpL)ZmmhAnOiuE+ifU}g0Wl?sx*nW+)N0OF8m2xu%W zZiEDx{vv8{bOPyX=$miBUihh(0%l?jyw2HE0eWSa+XZsw{SzYCJvj<(i1B0miFpXY z6b*>=0GdGoM0aQb%_#d9zwv$aBVz_=9oRqXi$DC|&xJoSB5r`zjr|CMp*jH95h#QG zSE2{VEEgc&bH^SFkho))`<1v$ECWPy?fI6uaolwQwuS~D#*M7M+y!W-zl>&nfwh2V z2K)!ylTQ6HeJCyMPCnc7pW?UXMjr^F?}#w~0eA5906PFfYH4^cInr}{gnRa?;6AT; zw*SBz_~tLX*52fA?(|0;*tu`W?}5L${hNFDn@Kmop?|p*u=UfY_A2y;w!1%l{#Sl| zaP;=)zhNLq=r#8%!L*6bxs8iEnme)_o0wcW0P?;5#)o~{T}}=?ejq^c#ut6g&J7^x zQGf7=_&&FFV?XgqaQln+u0MZt_j>2$34#w$yB3U(oyuPyGvFkd-a9+@-!1q5uLAUk z*Sa7rb@R(S(m#X>hnH8o^Qy7Oz`W+}^)25YV17>sPX*|U1SHZ^n4SHGCLaAju{1q2 z0c&#ap1tUK$ln1b*nifmZ|bX`e+B@!0MB}$03r7R>Y>A&Tx$RQvy#t!6vD}#NW%Z8 zK$-qu1&UcQx)QZE!~;*y8^HD?rr04j+m0wst&3 zvW>l=UnYLg|IxgX1MOBY!`gu5XoKOtfJV?pxG>QM?)>(E6%}FJExJfI{V6vXXIKu# zzN-Uszs%4QLpbAc5*zguQzxvOUXJMcnRzZ;z3C@}E3Wg#KIXBU@?c*`9_1iT zTzRX>I*^vwu%RsgUbcsG9gZd=ni(5|L4^wCj9z*}yzJpbcN9r;cy>sq)tNPw@S8yf zK(C>E3=z!34ikAhHhA4X;UW*>#VP%>;!Fa%?+x0_bOCMfQY(ty+Nku74l@O*BI()j zf^a8;#@mxt9{3)L9%WTcX{3>?ezs2G3Fg~QK(#Pub;3ekC`>cd2*d!4lUTbrAfM4n zTVt6`MtVN-=&*svV?JAAw$@fz=J$e(z&YEd>@;g zv_Ei;$rj1iHfLtj*Q;6S#%Fzp4w-G^*(F(5NcP>WNd0UyvdJ-jOWADBXI#-67f}=; zYa+aHHZ7Pt5z!wNTnxJ()0+N#tRH!e_Kle}9!1Ztq;~#6@hNwI4~fv&0Le`mm%X15 zX%M9_!&^Qk7Qs~nP!#Q%n~sB)ZFJdyr3I4|j}SLiE4rVqayK=Y*VR@;)@Y8O$A?OC zyBWc093P1jhIeH2uv|g0*V#cb^Dzd+>ADM%_R>k2b3>U_6AAVL{a{Ow?)GB7j)BPt zK>Rrc^_Qoc@XaQLxfS!|PhbfuIwU22X@4incWOI)RG$@D{+t(q;r33IyOvUrf5Q_% z<6tHBMk=98``geyr8^*Wo2DPS9XB{k!zYYP8#f@bqjB9$1kG!cIEy6VLT;!8s$wX@ z91igHCcSZYzr^mvv?Ka9h4pH-?w&}9To!8fRexv^mj}mwQX|sw)xqeI<+ZsVTU7{} z5-KLNX_4^m_)S8hv_Ka5SHWlm;ng{&Xbx2}l6$m=T+q4AHsc4+1h#sT&6h@ju~1o+ zZ$jBGrHAApbyTNFsj;jo0qc4v;3p$7CUej@U>3)e1mPmnAJXyOfNFs06~_iCys_lb zhHd-oW~l79-szsoayTn6{%t;_entO<7UYoK&WbpCad;k|j1H?ly7nM2+khqrq)xG= zgrqS721p=J;)aI^guz?{`oOv9*fduQ`sdnqW0$#AqDnt3#o0xúUAo*Pt4$=jz z@2%G|Oc_Z|!Ew@*0juiQlSX8*_+RyoU;<(fPtXVw!t-W&^{`DL${p!3FN_=(dXbnW zUF{@K>fvAJ>{zg>NlfRFas(6GbT>1~-ODBV2QwCl9Y*McQ0_!9kay^>U||Lj#ag&= zgW!%0tC->ZR}_F{mXu0ANw?p z1=Bd~m$w7`-B`ERM#E_&Yi@=wV`yMS>C)ER-J|AY>oy^~2aC6rnO{T4%3q_% zRtP$)|6+OOV12}_WuMEkpdVGlznurunTBs*M36b5AfTzR>J3*}efUJ{&AU!-AAbHx z#pMj!nQ+6I&f7Umuh)Ahn<<$&bFv?-0_%=mQpMRK@}+$qS;M-2s}E&nMXPqfMRhEA zW^36M_5%1g1m0mLgVdLE5b4#pgEZIsXf@9FVM?u{+1y-KVi~f21xwAI@YeSV`EARC z>aLY*3ZFeV&hWEUU+_1oNMa#khk30vY?k%myC)rwb0ctuN>RbmVYcY4RHT!9t(g9q z4QQva_-v}GM>tZum*yJ+SEt)KNdHvOZxDlR$>g!74%|4j&uH_Mzx2{CpGFtw@CwS! zIAs{gmcYWQUIX!Prwrn(?HV0o6{LM zp(J*NTqRYbIVu?mL*If*j&jWM9c(^Mstr<9(#ypt@@EFx zv9Y^eC1MwOSWLUVDb}Y+G;nB6#i8@BAdQbja2l#-Ih0WT!`;~-eUa6QkAiOP4eDkVA1~C2 z0!Lm^l%BZF1(>Sb_b-J{dJ-r2jKZe^J;bSHNJy-!>E1mG(t)ErDgy_^bK*w7in^N! zEtqo!vqJR2NL-KDSL!G<+zUoPdr`c_B52Pc_eH>0nUAo&F8J44R#}hc*4&D^+-sIY z5J}j@^el{YwJ7WH(ILV9B}5TN_iRXGu&>m}YCEHrTLF{7_W=q+zGaKn%g@~H2a;m= z-=DJ$fAwti9HH4ix}8jxVd%T5OufXB&hcGZu)s3mtRuqU2HHgXURr*t@Ma+8#u)*cci?1u;~F&;24#9-m@&^Uw2H! zpXeqzM}*rSpW$qdB$+viB{V+rk|jP7B&+SuBj#Zw#TQRUM!fE;>XCD48lxs`p_eUL zypY-#?gl`B|K+ly?x8Zxb|7Q%vUiR6{h=LiIILR^*p}d}X*FLdG$Veo zOFCF>`LUsK1iT%p_E)Mm?$sEmwQG~WheEv{)<||^oc)-_S$RZJKjM3&5 zN$D8EM`ZfGFg#m|*{6f>PKbl;ACs%bORJuw3(QDN&I&#ChNs*as8b+ZjEJHRgXhN!FDmxn( zO*A6qK#1fz!1erEOXWMq36mT8Fle3KH-2iw$r?Jz5f&@MfhzII92$APWnxGU zqZ&Irsie!oEN|;%WZX83_V}BKt`U8@7qr4#K30+PuXsVQxOD_&NX;jyvSlSd5)mb< z)L~<#F49QD4^nl&8(&<^hW9A({h+n^@p+sUxKCFi^aHh#{nGDh!c_JpO4Xr}ws1Ppsx5@awZES%jZ&(!m=zExjW+U|4X)G7)LEgM z&t|0sL0PdZIA7>p);38NAE(J|c#UX%M!W5D*J3(w+BVD-KQA!# zY51)9{+0m?{arV@g_tQ=ZJ&Ux53nm0o0Acqx-twTo1esgUd*&@g$5svIA{g-claXQ zE>m!1`3LBTq68^CUHFTg`EIN|+d0GJ?F#6P@1Cdasx1*yPvYhNFK>r@}PN0&*7|3 zMo#LI71l50BpDVCvY%ZO>IU@V*YiccRxb+a=V+fod!;?#pc-pS!JTn|*m8T9DIt1$ zO9=ZzpD|Re2lNI~$-BpRxVjffo6HUEpiV&$pU^w!!?+e;SaYQ59lNDXWi4 z+S;QF8tzx?OOe7VSLBGN`CY1VYh!4pd{o4hCl~pNx^>|WN{UY!r(-v+^|+zR2zs@LxyR&^bYvl?vRA(0nMt{Ehg*+L{x>*_=y;IH-t|E8YV$i~N{0 zgEmfmbTF-Cj4{^(Mn(`*I}a#@s9TJctAC@tyyj1Z5q;!Gn6v}kDcJ)(m;|jlX1ZhG zeBWT`D23Qb&JXsAAl@>ZmPMcivevmimNphm8I* zy9OeW`?JwJF3Pm*32fqFVfry*h)OnTGTW%c#HnqB1J6mqjA|G1gDkN_CM_k%Ng}GqqziR5?G?F{5oW{*&hG&R z;`jmZjE-OthsY~2o`P$LkZc0tfsT@RLvi7(oexC;&T$Ht7et2Vd@gS7I*(e%qR2K_ z;S%Y;*S~Y_Pf6Jh@F!zLjeQ`fv)%{C`j(2_aOXKLE~q3U6*$N%U&1Q9eG){N0jKM| z$uGJ*^~r_b3=&pHnlUaSu!sm+c3CGsDe&5x zaW(7%g@dZBK{^bdKA)EO6yeKSR4UxAv&<#(|CZ6Qejyc#B>jTPNi6Y%jX7LHH*iUV zFIaqA7;-92=rw*9_>D3MH7`I$&D_;{Fhf906BGZ$Maz3Y2+XCz2u^xhwI~@*HFGSo zxAe#dYj2TktqkQ3qX#l2$pP{-Hdjj(EKdj`WY9MT0tZ{{i;Lp{>Tx~-d%|icuuY^4w_GPvQ_NWw$tt=*$2ZGMfi)V|8T*&Ue`y-F_ z+|kbAta!z`ZkUu9*g4<3f#h^m#%%MswJx@P3QHAs=_3M34mt6S6?Zh`*Ca~R7f@?+ zwAiEEqB_ccNzo>8+G2%pRR`Iq?W}}*cky=@Cx#?KF=c;tbD*{S^SJ;c6=&RZPS!=Ho1<(GWy9qI+-`F+jJPXR62}}!UfV6?ik`*-4;MCt!H}_;#FmLjX5qyn z#20u50L)l!}U^N>*6eeg}zvZ zX$=y`ccuOZpxXUAj%5ryb=S3$GY>;%;bUst+B?_s-0u{ooHuP*1%?JW} zByM#>Jq-UElItIl^JDcZYKl-pX-?EIJU6`2j*O!^h7zjZ@rkDNh)BnHQxT2MhkEgj z4)*hz;3uQyvt+6z@YSe-6CR~B_eWiiDv&yU5n7OQ-;;Yg$`d`ECo`WM%%vlT$k8?& ztfxjnHBkDEb-A$zuEa$obY7&Ngwyor-32&3M2DT8{z%a4wlT*Y33u{$%8}=eu6R#eE_*`>eq!?z3Hd^E}@gD4|K&LeBIR z69M65_a-qoWER^{5~i=kTo3K$F%9W^iH^s$B!2l8w&c$!(H9&aH_3^f1bX)m&MzNX z4+puk=s83x8aa8z!y7qGS3%y;4t!~>Q1o>F=R><*|ublAeVbZ&GCcBLiJV-~)%f7(xwwzS9j zM(SQ06DS_BOM{2f%TX7?ww*n5}YHRTR@e^SL=n`xjZI2yuu{JBD(GFFe8)sDvw3J?!lZ+)$ zt`5pnqMyH!j#ZT(Q2P{kdg)jg9^EUmYg(6_V-HeT6`2Fcmo3XiH)-{ZjktgFYEy4z z^5LRC(c8T(DomDPtwoGW5}CJ_o94 zUe5-K-naY$I68Q~!jbqady>WV%E#2H*{U+|eZ+ST*eV{;H)bBO6IWz|Xf~g|F!HA< zNgZKvelW|CcVwu_A?3B`C?jnTM_1#&7o}DO?^|z5Td??2TQLIR#^MR67v1;W%g!<7 z5|Za{u|v1cszwMXu6!E-g5!N{u$2GlFzScE9RMSl4#6x+eYpj&_xm)tAC3>0(&>cP z%b(4D_~b*m!pf0^()b(3vUp@9qh{RZY`N-;XAo|jn4|}Yo?Tg5{%YN zh7wL|roX;6+iSnktAh4&Zt3Pj=L>AVZaK!-C*OLCRxxgQd0_p%C;>T7hZ$FngoGD? zgZskdbJ&g#O7V(Po!VtR84|pi-tY!0iPB;ByWbJi+8c!<_h+qYm`CBa8-B8Z&9Jx1 zh*YF~a$Es<=#8W!7i`5F+g@|72VpP4+|X{YKm9p>+iIN=-0#voWN?;oq zNZ&We9{1`dSO&S|Pld?Uq|nE+FN6i0Jl=@=>6?+mr8P^Sho3gmwQXZ-XG*Z|Ws>ei zRoS&^cYI`bT%JZ8cJyLpD0@77!Bbv3qpxGL58+{@0e0PS=lHRsf74cM3sF@$P6fpS zUTzS)H@^DjTI1M7g?%jkB;@D&AXy`8fAYzFH@j**Li-78DP+x|oUfuIB{O*uvy=;! z^>rJVZ|@&g)OJ)eO$U;atEg4cy~$(iF{4UFz@uoS9d+FK#ie?N)! zHG91zORLM_sn;VJ((A|mfSRkx2rAgN(ZIaSCM*%wKm@L7z(d98hKN6iuNyXai6+m1 z2!@*9<6hQ9V0>Q)5ml_ck=GXTSLtMwy}R5?7HfhaGg@7*<3V^C3f@K9b|tgr3L`fv57A8%#o+sm9QE`FVc)Q>Tk+9l8qY4jc4{yuNBWkC7PRn zvj?bSE1l};yd-l?*g(4y+}eo;(=Y~4a5=vFRSstB$?GLM)B$lR4Hm*n?WBOCVhwPnTUW(F1 zm++&ZWM`zhb4-cw8S@0e;U3ogD}0he&a@6>!JdS9Z0LNF)D z+;fSDc0TCiE(9(3vWRD)W&*2+7d7ug)@_nBrHVg)4|jfehFgrN z9AwI_Z`tKrNsv$DDK2+}cp!GI1GXZ#A|UTslGD0RK8tCR5k8AGBGP&aRxM$EbNx`l zxoAS%@4@}F zU(-fV7r%|{;o^UjeLzrkS9aZ3DimES^Bd^P=56)u+Xu$oVDg$q0Gs?`HZD?Ef@Stb z#2zO@Ln7_4W>p!pP9Vrn!|fL;U8I2Dzb^|;VR)wx?6BL*;rZZHls?ck$g;{q40SX+ z2Rza$^w&91tx`$#BqFrt`5(>iv3Xqfa5O9V@e&l<7|0kIE>DE+i z7@5{NN}||EaBPQ$d(G7EriV{p=?FgfS(45MNF1;snT1{h}$fz5S+a{h{iukfp zpz3qyg~91UDl~(6oH~{jv@#(cEb7u_#@q4?GU8qRHofz?N-3ir&Frk)riXVf zhsDAl42v{7c_>u1(YlEGMOCG`$i zC*!rW%lX~T0zuh*Mn{@{4IC&cn$ywX5bM-w`g;elJCjY=x8+cd6zQKbsscU%OifpM zH0QzHtSr+#$)zN%o1e<+7w8(X0&a3*bZKBh=@J%iDdMVzeUF~gDx(`94461iy6H%X z$XfEd*K3J!3&APQA`6F?;_+?0nF^KoJ}vqjf(1C0Om)(kZtai?C5T_ktkqXfS%7`P zjVvW^?saLlAFa_;E|)1MYpZ*SM$q>$kEcJdBxC+QdRt8dBP$A$5aGA{#ks_P2j0ee z!|9xIy_~U886dri$!&>F)e#=HtVz@Vx|ysmH!pUT>X}Ux&f*M%On~HIt(h8k-^BCj zK#vk>O5Km{f*fW=o9_5|Vi1>SQ^$$hMf+)mKt+5P-`=|DO02wC?eVsBlqp4((OrUy zu51Sd>ax|uEo^UK@16>B3nRppz$us`*Wlwo^5za|sCC&fyVBYoDK4SO{y89MVE#nB zPvGGwK=nfl`zKbrM2y_5LxD_pnNPb=(HbVZpw&&qYAGh=WE~!P>`(@4lnwmmMCsfsYRFGuR0KfLp;u9Ta@+jCbZ2$!}Uj-(H zWYs7_Sbv>+Jj@TPWmw%(RADj^+9JpDbPQYzcM+qd-$CJHc{u*0#g)bWElZVob<*4(XMZW){+(ugx@(FCqb1~YF4&DDI(4f3dmN3aVfGjW#9 zKFiaf%kNWw{cG+eG}+Swj`q!*Jearlaj3@2b0xf%-VOe9oC`aqIQrD=-7Pur?- zf&gQSbjJN-($5@1n^paFU1VrL_z4EPa~c~p8}I}ic`=cMU@i+jNFt@G;i0IlzDL25 z$DwbYI-!i&eG)im#mTQ^XXEko?C*1ETfUKGb} zR{`YvZxe_`e-LjJEJ3Qefz-}W^fQH3qr%4dmaly@b>YPv zA1_Rh&yXUoENjPX`=c=tEy6qkzxA(Z9aPEI7aP5(9T zTUz-^9QpFEW}Aa>z3xJOyiM$ekDHXi?w;}XHfvu|+ICs9VHf6~S_L#GVfC!-O%O0| ztgExP!um^6^4XDQh9ZCH$wh%@sZFy)XgSIAZs5F}7Rwa!dzHJ7x6*Z`1|#$))nHg@ zY6_Am=k;B<>#sA09XreqsXrP-*;W&wWGEmH<(eITXh`yNu>Vb>;oCEv9c$b(?eTqy z^8y0fE<$V--*yv1p2g@)(Wicyhq9EU#pmL3;L!F0B9lBxxU0xqn;oI8VcmZf@4ruwIyleAm z+x~Fj{rCi;Ca0xrIedb>j~m-cGWx7fv}#LJg_wsV9a3D1>EFNkK+)wnY#8@gZfos& zM)+b{ZS5++{Dgbmiivy)j_j2|ZVS{u3Ljs^QTWEV4Ep52vq8>25L@PU&G)KwpX&6J zy+M?2S?O#PcLBxEa&Ql}rAIw&Z%c*XIu+l}}71(Gct;yomJm8%{N%v|7^F zHG;j6!7nDdY(GZ`a;^`2N%52|@vo=MaC#n^e+rTP`)~YcXvrxAnZS@v&TcZ{FP8q6>7iyVK{ecD)1tiYUy5W-;+7*HjF)? z_k+?;Tt!+%l$;^XUp7wXQb8YW(_`pHe|ny5Ke&UWc|&8m_Z3AAmc!GKBLiK{y0gOT z_0SEaR_02G@XPxS1;e7W;1GTeZb^+f3cBQO+Um(AI2zX#{{TpsnT+V!YrxI{KSA}a{uJug&hkH%`>4rr|j8cnRoMGJA2Bl|*< z;N{Et#2hUjK6^hp;{B~m&e=i#*|pkr6HQ^4Mhba?)@|xH1qrXI&&vRGjrq@bFkESA zp}B#;G)>Ggzmg#t&M;j0#_XB)q>k=LI{gxwD8RK(UWY<7;h8*~OJ+i`1;eP|XtJkn z_=~fNj9k-++i>R{+{Ks{OB~HvodL*9+&a7(m^yFga yArZED=YmY`^pM2GxP+5E z&yO3`Q11jwSzEJcdEuCir&(#&x>$*V<0dxIhJs0Wys4PWSu7Dm zzVl?PMz5;=c`hp$IUl?#f9%K6@YL)f7`BI#A8pnCE$6S^Xs0iA4RdEUafwDTTa|>K zC}QXm3|$5NdE{=+SdrxyL699dpR+`N6o22t^W+*QRrn=LpIO3Gu~m_hae; zghjoxS{rMr8Ccgg$P&ewD|NnFGh~zl|LkH0W1h+JD?8ZI?>iD}LD+X1*;<4p9*!VLgtNHH=9aJCDWH!$j6`2$^Y}n8a$S^&3Lz;e zRq}V^-4C2t>4s1%(Vv&p)LoncuEG28ut==0CPa|s7?k{%&losF!~G0zi1da11N%iN zA)K8e8kUt(v1kc?>ut0*N2TG!KyYlRvY5b22Q|MABVyCENhnTMoiE*5l;Gi_ zTP>KU3|T8Mol2y{idnF1=Cem&{SINRPYg)foVpUGvm1hZ;yRCdYI7JX@O?QOA33v0 z^}arTSjv!68#6Z+#bXZUVg(PPGbf^~cp%?%h37mrOvD@BB#o>PQkOP?O}YRid^R2w z>e3zgzI*K*f&>qBD7n6|FQZ(GX!14TL3A6zfN3M)oNfQH6PA+-tFyk3wG)I(c@>Lo zB@;~(q}yH0XJRH|&6N-Gp12rjJoR%R^RSlTwc> zT5DK>s|BY^d1CIree>LP1GDX&SH@7Aghg*rII;(N+5?)ck>`(MQdzR^?e^Vq~XGsrY} z%3zRkzgxG(bK@TrS~Q=FjD4vkxRJ$gwwlfip&^Sgb_vZkrfVs?G~J$%w$n~)RBz`N zUi#|}sW}^q=6h(t+H5g}cfmN;NS>)Jk?BUPLfVt$p~x9Ot`;IpX=FKxNe@6Q7>;ao zUY=hfzy6A6VoG~}B<#En%^`j`q$GmYcvPp`9=BHLjZMTsH7z)y_VZVnx(3r*WcQSKKli`8vm{W~roOIJD?+_ou}G9HkG_hdSg zfYVf^*fUf0XlgNk@Ym1G;R40UUe}7cG<(`Q+ATb6sHDEL5Z0y2779s6LG7eEatnKq z>Tu%0M06;y0ec(hx8OObEIdN* z%d&h&Hk{e zCa=jDZ5I`NV61cLY7yet?#02*;i&|MhMT12 z8w(7gy9$^$8$hH_AR46?2Gp*Dy)qr(?BpCbIz@XeaL_FKJ#ufZlkpuTQ6~8F(x-Jf zR99el+YKC_X$X0eF}ozB+_CJ^aRv7!|3fXK4wb=*h>g1 zgpmmHd*(b0wvG&>hrXmK=9`$@aNl^Ai!_cO1nn7DiDkQ7Wb zcpdRqKkK3V9I@F1`}#omp9+8D9U|f(N={gFOlJ36$Dx?CAfV3&FD-X35^dB=EtIM_ z<#?PSZ$JomezNail07X^3-^u+p*UKqK4LulT^kVj)|V;%b>1s?1c*xgBFTT6;JmM* zc`7fUQ&a5m8nl%>Fy*c#t4>n>UD));O%5s7IhY2-?NUkU-s2QXgwfiV8w!Lvl>;23 zpFv}ZYmy4#XOX;$^dxGZBDc3WN_c<7kUll3ddQgB@#cD>PI_7YZ&)N;g~vy|$E$DK zIiueauacPZHDo=#Ae~~sh!m`iV`{m|l=aFBRD`3+_}>`ob;>6XQ-mnt%4MSIA{j@! z+sS`&XiF=qqQpOMe(hrgiVBgAd`Z4-@C8B=c`JwVV2ir_w0`ZLRzAVin|B;h5L!6u zxFm(MT z`Z*>@hiJe4RP@Mnp zjQ{yaS5Lc+d~0!)Vd17^;Lb;T3_M;1FPjbT-gp0ibQ zMjNjWfeP|ZbORyeRFB5ek4K14biKYzVjO5aa>OA;Yb_p`T6`J@UGdA}z+8$!e*s9F zqqNp1FVRQCi9>HJ887HJ4U-%GaC_YDo;(NAUayJlY7V;Ug;56uIYrq;R~#>0(i14P zWk|BHDmYRVed`MN!!T#d81cvm#FdxLDEER zQo>xOi=a)KjSsHObLNoDr!s~ic-UsPWjz3!m%&UhKF^xDUtYcBkhW8~@<$5d#i50c zE_uxrom8Fso!vgf7l9}7@ngekf^>SHm03(HvX&T}AR|E4T2@L4aTk=zZVOiSl}>M$ zV*-}(L7=lY$bg9;toCA72W<8xin}q`_TUN{0Q-QZfQo@wAJmNnl}u{};fCB##o%z( zk$H&I1`*mwatPn`ONPK$gm9h#%NCQbuBe}rG{SwSDf>G0jgTFpP>gL-ceCBMR+V)d zwRseim$Hjj)Q9U%52jCZ(R^RH zCGdVYx2fe`1zTmmAGta4sOahueYeI&}4_gj(r@F{%FlMN& z8%JMC4y2Yt8k1OI*2dR#m(CmLb`z5nri@nJv=Eg#4bA7Z8Kc?nIPMe)bjmfDG+<}r zPJ2w&Qg^lq;B1XeMBQv|YqxpFy29rAy2&^0*}O7eNkk?;g`10hQFlpIuca;siXQjU zx4bfJAI$H87V9BAY9P2 z2k${gi!ejCNy+~c#V&mFj;kYCSCCl-0go7@*DR6Qf3+jMPNTO1>rGDHDy2=JWgWpDZ62e z!M_w_@ft`@ZTgq-x_QUVnk5Cdkw%ZTsHCqiv3|~SR_`GTtKWIT`qsxm7sp0$R~Tl8 zgw(u;9#VPc^n2Rs6jwMx=}4m~Xr=R2LZNH(oCs&|BfCZQT2}<2uEwIE?6)xzcnn=u zve>s8Sv6m(J@Nsor-AiUY`$hejN!TfMYg(>ufZCp*>ljW|D$d~%E5z&yupK==wVQ+_G?JSNa;1Y#{+ z?{nsq`eIz)uRUj-MY<9H+HFXiTUPzbZlRfnVXC!_-Enl;c-<&l_~@~im@1N7y8TEE z)1vA?M3;`bLE&@nNQG_z0j&EE5l?9Igkp+wi@{w*q#$TUCS=j5&k0}5!akPR{4ni7 zc-+YYzCi>$@zlQp^k?w{j0pX=a6Su*Z!g*5eQq$S?1lA4hooUKXb}k8eTz)9+a%>a zmQ``DP_9V7eV&aXAWi}w{75m8(hCcR&}4eS--jtfvpkY3GpuuBJm|4*p;yeyB~4Nl zNSZHeMlNDZOi={nTC(TDkJ^z%j=1apHPjbC47UPk&JtXX113;rL zUaK>(f6~RZOF)7qR;=ZjMepuFq`uuJ^Zp!|uX#1C5fg5p=r#5#qyDXcY*-~us(F69 zt~UunHI*yt&J_xgSbbem-G{qZH6t1;BNQjsM20!2s(raIY>|h`8*<^ma31!F$H&$o zl;(5wicO94i_r?$nE~AA*fvGFmLg(=X%EqSg6hwckv5_B z<5iB4GR=%aCYF4=u?NhDyn+N*dCC#xB|TJOMYe&}>#Lxb8wSgbc z`Qxef6}(0leF&XRiWXkAq7#)v9+#OG6n!{4tWAaZWOtYjXPLLmyTer$h8ky?UY}V^ z#9KETbb@6>TXD7wAAArLN(yndkG9bRX>f(b*=$LU%hG@8-%ksMQ5v?b>R-b!4||P< zz~Hp%C?0apjYTxp0WH>cSf!kG`6X>(8&kW|hUaTT9BHFFBPBhnux5YrRwbL$tVp z>Z)`ONFW6%(@?)QmDX4!S#;y4?>Zk>p2=4Qp%hHqo7WwL)eG3g4Bg8=QYSCqYX@sK zjx@C^ReYLzhR(htBIP^Y>aO5%tj-|BaQtS?OlrJ3bg_ZkabR0(s>gXx6nO7k5>HgB zAZDE)#TyuQtcUUSzT`SM9SZ}JiGw6;W7H~cW@_Tlz8lSYsFMU4@}T6@aU@O#F94ke z=Col26RBDj8W>q=xARq41%(t+6f2(^&112OJI|%9xSg`S+Uba0Dr%~9nKT^#ahn3e zu6=MrWS%3=?^CF7V_Dk(D6DnmXjZvTl(u0Vxt(c$^BEVDyUZ{Y_GQr3wl|~J_BZfo zz|-mk`LaN&-F9+zydxv9>d4I%TLE>hy%Z;LW2+0#7B;!7Cl7x(TY_xOu(>G>-L_E0 z_4uv@CSpI&N9s1UJYDi<672rC*Xip%kza{ZlXRD>UvzI$LvjgTovlx-RQD`ZqU+4D zD)i?gq;nVsNpgRV?o)%i2AGOj|J?GLjXsN#9bW$iWJLz z-zq7R2u=3Lf)KE8v5}0q!9mm{u}inm8_zXV*L6Tf@8gyr%!94Lmk9-Gn^dJ3UlGu2bL( z|6kkY*joz&AmQ59scqY~-CLWtwr$(CZM(g-ZQFMDP4?3!`yVEmnand>+1SH#Nwuku z=;~!~V&b-Jbli6^{i0=>qbzCQXvhIcBL@jb^=)!Pg46{&Vqa;&@m+t>`ASzX2z;6R zHXAiGM8|ik2x6n)pNX$2(^jsb1`h&k>j7BgNq97)LF12uEr_}>ENvV zRUqX1clE!gMHZERead)3aw&j<>RN75-s+V70lLR0`r*_!V`;F?1sbfdLc|A z#~}gQW<#Cb8d2V{gn#C5M9EQIU^o7*N9-I1GuvseJ`=x@>}zL)Z=Xsc!`NxB94+8k%e^FzD@89#Og;sFn1qhPY1wlH1&c znv3b4*c!j$m*X9aI?#>h*~rwAkgEGv39&Q{i=9Yald0~K?Y<&Wd~ew|Eunar0VpX1!SSW$P>;#a&u z&@12?q<_PtT*bf}6iKcTUj+&n|k&7(zSXv(fc(ETDi3FgC$#u#3y;1U|iHBSS| zpR$eeGNWjFlG0Iq>-=qtfLg1;9wCz9;TWt*DCy3OyLccUS^;W##SA70$2-UNVb@X9 z5Y1fED}iApQT{g2!A5rp!D{l}-UOed!(CD=X^d5Mi~fyZ|6FwCyzZr z)mz$U$19l05mK=7tgt984>Y!Ni6PeQ^`tM^9S!ZMP6?xd0!M&D#Ne!6#)Vav)gblG zFpMkzT=c#UPahI$Zn3b(e9iYRz`m3SsAmtXOD|`ctTDim-rO&2-9wZhOV=*ovTfTo zyKLLGZQHi1%eHOXwr%tFS^wk?z8vgP4r8s{84>$^pdV#3JZ;eJ5?VWzvF!}Xu6W#@ zK1DT5fKvt-YJlrh^r(g(NN~am-ZRU@TF=sz$z|hXx1o4Uoq{nCaSq|~yuKpS%>Mv1 zNW6BKFxWz0n65?PMrj@Zw#$8)Agb(8UAM|~W9r!XW!vrK0(RoPT*@`7hw;bgVn6m@ zkNFeTl#g9M|jC@1cW98)>H_4?CP34;&Vwzk%@HUVg;O5MU@b%ZYq|rxhx`{=1(| z1g4>UR!Cq7M!H)_m`y{mps}Wk*Ca5f(=??(i;N;ywPT{Y;Wceqt*vG@y&>~KfvHd% zZqLENC@%Om#1kuzrcpIK@}ms7r(8Qrs+i(V|H(}H1IScukO_J!=ChIDy0EpTq2YEg z4ll!2QW81p5djxw`qj-5k+OFgNqE}eRUH!6stWE93{81zX4=nyXeS5^OT~ISpz=eD zBov>^oLonHT)jZ5MjZ5ZmO9v+9t4zY;>ZSd zU2LgQLVqm7hK2_1#2~XNIdzz;h7!dj+G`H2IuMExgbXRItkkuEF|^_7z%b(XJqg6C zlfYZdzyWwkB=EMRrAUl1k)+*VnU)KH9svj0y32neRCRi3Qz!iJ?&95OTM-Qtu9>Ap zTaGdT@E1Ic^|ZJPn>y;tpx#UHyWw0gmDT40^-Lk+H|7PV#PP>kKRpZkYtj69*5uR2 zSpqfI!lV5QO)D4wWrJYtnhH>UhtVQfKBH;(5+ZY7J3KaZCPPemTmp_q0&v8y^0n}M zD49wbMJ2bfms5C1&}pXX?amrzQ;C?u^T%4&*vkQbkLu51WNUF_LjLetEEx|6@7j{5 zxn(nnwuiH(8%{FabTrG;)6*KC#hm>w)rd z!+iNgDBA0?A-iVv7$a@wAr zwU4p!aHLigyI)&)BBg)pPyZ)&l&=$mp3p%0#B@;U?(jm$lfML424Jhe%Tdp7YCXnk*_43%#vBuAn ze4l72FZQ6555q>N@JxGV zE9SGmwTPMOBY{SPH%yy8RF7kTw&7$tcG4*5;6bzzw!&)q6oVgkE>AF7<_x1I+*M+d zyF=m^7oLUB6_!mGCNv)`uqw(cNr_-)GhosbL}$%1@EdRBpOOpSpDO;JV1fO5Geo+I!OgtU1t1nnufDG*07eER zz<6zBoe0&+81U$F2$`6NM+8OG**j=Dp_nkPA`i`%f2}YVa!IjMMO-_$b6Qv&7P$?i z9eGHHnv4$xRm%w`NNtY^HC4dYB0%-0Eh6bffAGeF0q~x)!`ua!+4yQavUB=W!uBa$ zARWM>*e*8H7pdlm18Y|3FLixreSE8}MK5rJg%)sIc6;57f}PpS%5Vk=-$z$CQ>L7b zvu(&<2Nr70>X|_52G5$C(p9DrOT_r0^y<@cOxJfj?9Y>qiEa6*I-R>jmu~8+#Zvrz z_VsQ6gAn5#wpKSVQGIG_K5dcyDR-w&>t$WK&U$J133wsiDKp}PAi?vg`ZgNGELC^lgktahlAdD@&-p)=7026u*j}Nkqs>%T zOQhfKE8>dlePSKS80PmUjozBPicO0@UIW@E=R|e6L&mdbrPj}@u#jT)JygF~GPg`r zG77RjH(ji9j>rHO8iePZaNWui?D`Z^rzJhDB^G|^m*z+04++w}%rJpHXF&AZMK}yp zp_s*pW`0+5*RGt{QW*~CV%(OiUgI(yW!K1ZHiTe`qOM8iC6o{bha3sdMSEmfVEQUF zL&2r_&3VBLfzc3R>GaFWL<-c{{hoC*SJK!9V3!&A94@RaJx&(_)w<8_kge*E7wOzA zlt$9u_yJxbA(I!&=gF1Jcx5sqQGq__P+||JPyvUNry?9o(7sDCoWg%amC$fPfb#Lm?ubPd-SM)Fnj-6I~tD#s-8MsH%ceRqlx|H z_xS5usEXSGsZ3A3@W8+X%&p_7&4p1L>D7t5sgLW#^|D_z>IHRZFYMwhlJE7sgU3a} z>gDIea(Q79fQFu*%0)%ZVr!&I&@UyYcSTWWe}thbJ@byIH~iZEf^}*q>mJlD);D-7GE)CM-wWfKJUf-OdhSc=Ev{N&bLQ1+XMl4IoH6j>HahvG z?YH4eJgYKKG%$h7Fp^+&G(bVq>DhCO5Gl!VNWEQ^v?sPVG@uRY9<%CN_jFdludwlUA{ay~`%;jTahlMCb>0rrl|#Zg#$DqU`s%~xk>e;^OA zk!D$5KL@^Osg*}%YqJhW?wI80oQRb++>CQS{$?$db{I06UxcwDCOFg%fZ-PPM5LDK#I)r&+ zj^U%itYPIzRRW_MVufSHBeIW!&twKOCNtFwRWy93c46Gjty6fG*@`r46KRXrq12@W zqxq=zqlz8WS*>Zf!4}da@PUtW{}-Yhd7&BK&3zFB0nw!z@&IY%C;nU}iwBL!@QlIH z$pG)*8RF09G*d~JWe&ld{aC>IQ?JVMh(^3CB}(~YOg2@$kx zqHCDgx0@uQ>+~Ar0{L+&FWI-CHp}fjpzuJ1y+Erx`&gmxYxaTLds!W zP{<1Z?ufK%S+*NRT4UtH*|!d#6Z$SbI5ELK#X+rc!*twm6K^+4=V-n48J72h?!WnJ ztKP^5DPup4TPBhlR@-(7(XLbuy(hse@)%!Owkh}xx1or`h-n1x97=XyT4 zZEAN_9Is5S@&YrfJd~LlRZ~SR*0);eyILJ1Hxx~fv_4%3T(Be+7CH*?)6TV=kjK+t zKqL&DNWx62v#F|T3yg?ukX%>H1ahy~>2LtWAKVfc9d@LerDmjvbsTmYsUnepTT_Eu* zp1rU=EToXMH{IdctZKKr|JEBve9{%Zg0wTqN8*6aHqUorK5`npy}|I(K60CNR+8=& zy}*PSpQJ-dF{S1r2_l zRaU(D{tuAXf~Y)faEsxDQ3gg5eQ!^}Vh0Hi3#V^|u5TM2ib2nCAX6&?`uuyWV`2 zR%t=B{hm>iA-Em#IZ_w9L5so@D8gErHD(i3XSa&*l9bCBQlm0)0afN~k9iM%HoDf* zZ)X0-$v-jUg^~_I_5Ok<3%04j^0@k<@b1X`Xfi~IyQ}D41?{%#iea~MB16->>ds(s zX59%a@YPs1kq3(a!lI^6aP6T7rR}#h$XE5Ok=7Q{Ua(2sqMaZog0yA?USM z&@=JQ{kq@pp7g>az1gnlkD|$%PQ!R*9KRx@wj(&gPu8?0MkDl(C5kj^1mLy)j$SIK z3Ip2#H6C*ZH9W4${SIeC05Ql8CP%#ZZ&s+KSDJRtOqT(iaWG&A!+lZP@V;P;hgj`1 z+u|zqr(?^qbMTII7)*mJ0$$vf_I~D@hTlXFc+7OL#NTzl8e3O9>(*_XdX4lv5rChCN7yZn!3{_hvDmV5Yq}u z?B@_rnWAadLj^4Kw`Rvl8jT|d4AWXGkO`T@JY?X`feN3Ras+y9aHWz*x-VGJF$2a5 z`o&^6Z!#SnESbRH>_Ree%j%- zPyi{mA8;6B-!fvvm>SzQDQqIv931g5Zghtb6K>8Fi2_?ljalIVLvj(u3;BF@E)_~Z zqA)r(e`1+Qiz52!%+S#$TMN!JGgl?@j3oH_{?YcX#f`%Z-+E82bJ>XNd zbweyO$TO}G$fRD|Jt7{s$)4SRwjO+Osa#J4dkgsKqH91v4<&Er(Jj z7eh$Xj{qK5lC^c#o((XsAR4UJn}T2Y-h2eGW4`?4 zw~T$36NVN`&&W#P!Uk0-(l1SIu3T6R8g`LLC2+~ZIA#uzcS&>_8zXIaV3}LYaE01p zNxdBn+$khPWS6~sMzx!fEQIVd$vtq;6>v|-b8W3L;G5C``OTOoQ4!)JvOQGs_Fgrg zBq$U#&?{3ZZBh|z)5#QY;aeGtgZ`mF0d;MrYZ!=Wm( z0)TX&0*Z5s>$;ZiN>juMcmZ;7K!+!y1V>M-Su#iRNgU=&u&NAKT&>d$wZrMiKqn!z zLf7FTOVBjv)~navX!ZK^jlFL>=vsMPkfEt*Ck6_*8EWn zOg(WTU6ZB!A-t?-!~R=!%F8U{7XgyKA}dQTM2Z6sdOm7gK~m4e$WL0;r?JTw+JilO z5(cL$(tpv@KtHCG1^6#>>;+{s$-*Sh8rACwTBO)Ewe9w(c~XL?#jsQOs4a>&V%C*c z^}46%F~M9teQbwkIcDd@5SZWvwfPDHRr#H0KY-o6`wA;oB}#8os90y&nGoRKEUOSm zDp$x97HkkcV^?%h{TNRr0_Ug`;vo+nI&$3I=GJq=>)1{52?BPAqG>Wdb5OUOvmL$c zdmiuVD3Crlhw2bqP#aSoRLotBt!$7lh{b=RNM!FDb!pl;ym?BMs>Bk>QQA3NZ$?)|NyvP#%w> zmp8`hvSF*15b~Y+gOsinXAW@NEwT6M<5 znc4w-4xz?hnceK#DSOOL#-ML3?WxsKTLAjO&)iAphFkW_)@D2RZSG7>5*0<(4~p#{ z;*ima6T|duT(a}wU-T$r$5@X$lNdAOU>$A4Et#c@FZ*#5cj-*?@YYjhfhhL#60J4!sImI_hq{=gAOU z7j6TXtNeL&@$Rog5fcj*1eiIy(3ZTWP%Hq*KpW36!ANdivNo_s<)}Bu-SZ*%LZ{Cv zjpk+|*;N^mm;;L?37%?tR_5pHrcd&tZzy|GrrKuxF`zP%oD-#{(KHG&Oa%TX2z zB+#rn-Tdx?Ir`y~v_rk`JRY20xN|p@BoWsHRg6^%Psak?WU7qMVOU*dG2R2n$#R!g z`%Z7I{P6ylg+?Q&PFpv%SE9sr9igY%!r3K-Q+u0QIO|HV3MT#uw2ugipOipsK&LwxpW#*O z7&l`MuSu$vY2L)G(zJ&r&(EdrRQ*0MpgAokU0{Z^;2YRL!-QeL5?a*TN6KGq?-foK zi=rR`6XKUX84^7buNaH zsT)(9s=lgbi93L)9DQXY2tOzROHY{iV*%tW=#I{?5AXEZn=G2}37DMsT%Lb7i!9l6 z+kD+@O8aCbaYbnpFP#g(RhEV4Cf|DJclH@tDfqoy5F3zi-BiBN&-pc~dpIcyP_}4L z{P*x%UvX8c581Ct_^St8f)%6IsajMxXQS}zKq{7)H| zuY2AYdHdd83peQ!a6qxU>64D4Vgw$paL9BEsiODDH{LYRX`)8VkC4?tQ?5>X@|($X zO0qbu@?TnP%zrG^5ziL=$eR6Ex~zXO2Z{fxD09a*%ogI9&zuaCSOqjvWNp?JWo&!RYMG;t>V`4+u!7|FI3Y_i|{M=I*?kLZDuBjjw%%o-+S#dy^k*fdAPnt=ZjzxLdldwFY>n8b&qQcojY!Y6-;^b^ z&@=G#=#7r`bI@R8L%FXHl2|2XCFYkaGWR zxWaXAE(;@KO)FU9yaHi~%44o`&nJWp&r`MB4VLVf?7^BE= zH4l7(EnMJqCwKOe&H@y}L30v0Oz@t_^V>3Mqe1JwCfRohFn18z|$xqEt0lU8c zs_aV3zQEi25QK7Z4%J9md(w@jdA}#N9IPuhix~|;n5O`h z`lm#!-SKQVr0{y1$U7P;3>IDb$O^d!d2RYeg%za_QQGW~d)xv3NA`rHTG)XsPt6tz z_?{!p)8atzm`FLJ2XfFpoy@|uqs0lOIo_6l&0-kJy%D5K=0!wC%#uyDg`FH>`SLix;B+qjv<|5}p zL>_Q8mcqY(v=_WOkOG_+%Z>J@Y(5HIR%fQ(ypB`U(XA>3EncVUzM2o9{5HvNXI-yv zTNPs$&n-%qX*_^kkX@jo!(Q^hXWv)Yt)i!#WWap=%RzR#nZuX{Q}cGUPTit(yBv)l zxmI&*Dfus}$FejvT?do3KRR|u&i`->e&z09L`mXH{0tiSTT2!GBC{)nrWtG}k}#Ja zS|5#yh_Nwf)OX6ui}&tA8SK%MK3gJ2ovKPp9DR2+R4H_4;qCFb2u--tgC(NqUQ(K(x*nT@~1pF=EjYR8CpSyPiIu4J7t8Y?EIKG)i z;^dwi003W3go$KsG;>CYqFD$Ah%UL79jTol4BhL5- z5uJz_25C9E6%n)32tDUtNyNL|C(s9AaVCT28V&%l)f@&zUvHV3E&FiFCG*ZxA^V|7 z)_rOI1(Fpd{7T)GjDP6qUsrW%e~A9Q%gF0wYx@baQjyiMZS6w~Zhj7VQ%s!(@Csr= zJxodzom}A7+9&X_ZP6)1OYYCiI%m2ZI7F$13>JETYqR&f^|t?0q?sS|Pz5hJdj@lF z7ma+#S_ky1mcE;1EDeHyPmMMOPHBiNQb!A|tsw|mz&ZL&T$1+Ec|--G=O;?-By@ldz39+Thrr*vF{l_>4X!i=Ja#@BWe?IF_hqGl+S z5XCIWLFCMZ0S-8DlSG^{B0WaK2b-IyIBcP~=(vC;RrE!FGL^X8{QAQUhRycgtGb(`h9o(A=BlO4c_Z ztc}3>Q=pMjd~eH(1l4~-6Sff<tHF9~_3C z^TZvCq(QMrd}b-b^&gmDYQ77<-E@d=lT0#n0IPxKYSnvMaUl2&`($k@N+1YT7Qcd` zFu7JV$vow*6zBE6CePR?xDUn%Z6muEZluqx9f^TcNyljYWMMW;2i zv!TMLt03(slvjo)cxf+FMh^??7Z?xe#4g4>-Z>3?#p{lK)XEb&5%5H7`FHztzLkZ; z269)tyy$P#FKC<7YhkDoX}e7#C0LL`70*h3lU9c}YP6PN@{qC+&hYL7CY0NGCKYzA-P`q!FHz^!$gCCt ze)wlj_i`OlBzy??T3RUUl9{!uKb&vjhGAg9Y+-1>YcOx{5`z=oEn|HZmPm2|(qZE; z;~-a;y;ORta9>47(LA0ujUD}{^a~=W4joXz!E-aFQRESdCa&!Oh&E{mzNk97KUY_w_*#MVV`*DSv+6n-H=CY0=L@Vm*uo#* z%qM5=_TS%)+BX6k8^Xzu(v1A@_6>qgz;Aoy=%e9t=Sx;0oRK@`ekx()Hh3NzC_5X~ zLJ{lhaK1)BWvHc3zLb@PZz*vdm(>9=7Hfg>Y<7FuJgFYtQ+jQ?;x+kHO#1o-F3Q`GS%8=v?bRyfl3W*+zEzO%#d}@?pQO{Y zZ3hOu(_94I07h5ElCoUiIT$M`8$xP|68e<-@^Q``y@IMHgI}jAu1^x#4x#0dCOz#% z^mNLY*h2&me3-mf*M-g<+C6LhDmwixeHB@JkUX+V^a^6D(U!hC$cgFf z)VVi=MVYwBxwE5V2BLCoQ3eUVz)1O(+OFVP-mpXePQ~2Ixh-xJE$G7m4Z1DOSOg!^ZSbV^t&H%&2r_V$`+9 zR$=YlbN7qFG$GrYr%M5NLQTaLHR6LjKv|AqDV}dys_4WoewsF02v)3oxOMN-BnFYq zoj&|~BmH*j+8Oa2KDX)}Tew*7ksmkzo-f!LHxQr%)o^XZ0F5Qm=j}AoEqSBONEPUh zb_kUnZ2-X$Yd!6UN8&q!6R&U+T09TmjL2%KF^J=7l3-3jU|eKg3VSoT66DNL^Q}xV zG*g5L*BquI$U&uMl;*)PNgr!M?<{HXQ+35F+IjEm5%&Ng7Mtwxf<7KQr{oiwq4g_p z=;FAfsSv`aX83KgwKC$2fn2z)JhwF->h$lBMP-e>6c-;hG9Ch2K9qX;!i%vgfZwQfK9fj2y}6C{QdaoK!h3ByTVLt;8J{DoRLc!sK5iHp>mM!MBYd%;>yz&@%oM}DOxW8gEx6m0@9RL|uCM7ookRBhf zO)-+>laL2df0vbMyrZd>I-ySoyeGx0VJB5yKKNRx7s(R{n!Jr6^4gDuUg6jox1#J8 zi;QebtzuGe@v`_lS?XYei9qpj#DA@8NT6x`!O84i5Pii7zbJ>iRr3$26nQM0Fz+&B z_s~r5*eCWLlz`jAEhVwvw)_!DkD8hV?n2a|`j(YV9Qk)4p2dIZ-9l)K{{$~8yB0&88m(ru* z;X*cWY0boG>M8DPpp>y|RHxi=@mzRFjN-0{+C@*RJf${&Pu(Q}2mQ;q1pwh`lL9Gx zwxg!dwB-6Tc!XhpCqf{>9cH>8bK_48ttsBP#UNn#)^`Kp;7M~xm z#cz1abGGA4KPGb39_?G98V^lCV-9^yUL$E)YW*4s(Ug(>1H%5v8zq%C*%O zb!Skg9#r`0*$jOCgT7fB77N@U!Fiz3?}(kV4;kJ51%7O>v`-6yn<0{rEHi##X9flv za07b`KszEWcF7=6k1VGx$BSGCUnpAC;DYMxa0eSz!%qmQ-$d!)8|6}1QJ{zetSnWb zXLmJ`S?=QZ4R4bTawYd5)M5NAXv{ti`f^RvgV**?N7A-5x7A_}{~pSFO4=7>veFBNKr=+z3h7(2pXg*}U}yZFL?;Uq69d724-~zqg|)MZBLTgr zwSlvVu!)hKu?Z9}FO-wBqltkHl>0`LyNSBCRtqhN8#xH=Ka6vGJ7?fuA>{byb;93( zz~0;-Z2|pp?D%{+?|Hrbyr|q@EvaIB(RQjqQWOL=(G)=DCeBOGsb;1n#sx;_7gZgf z8JSB>QiU`>F|;{@&@V93*AE~iMRf(i;?i1MkHp{tZU?xUIRWr6I*nV3g8&1QRaf-q zm_s{z+h+m9@BlClm(0C(NLDaB&Z>&t;my>c&)E%i0banIpk}Gpsp;7Tm%pXIaU@1wmS_^XDR zcOB--jOCv_5%vo|E4KI7SwjI;096K5LG**~6+j=P4G1@f=4b9FjjaLR#NSWqQ=HYM zsqsAufZyQe%6e8r?)KzF)a>NoNZRbwT-40&JB=6B37XtLE2j{3@9+Ym3G7D&16?aA z&XFq%__x0AuMFUrItH*!j=x{x(ACd=(>ukn*h1L$yI-&m%E0?<=E^SwI0s3BzH&Y{W9&iSb)LeRG~fc+}mIiJndSo<+a zLQPIWK~_=CPkA^0qot|Q8F#U^@P9<78IJ%WCLI7cFt&d}5@_Gud{IFQ*?=n>8-7KO z{%^9}&_SW>jB703({E*Mwe|Iub@xx}_R_%0(39M;jP}*JiVf{Qqcf4{H~NR6|7X(_ z%o&W~|BTLh&VP($Z^v<5#(ms-Aq=nXi|Y%3l8_a7UR(&}-67EBvGExc2nTlufKS)Y z&6^y^zo@Hn!jnY z0}@jMxaI}=`_|S!aRWbtbm!J@^Phn&rYn7MGVhZ`A7=1|SM_gB&Vc5a8JNG5 zuk(o`sxAEF>q5OBg|iTLxXpI{Kb1aK4M^Ef4n#L zF_1j=Z|pB|hF|P2{f0N}uXyus_P2hMN4v2%!i5*RyP&a~|H{vQ<^I2NZWj;BfBW_C zrOS{1%8T9DlfcnE`q!+C@|#omH|t-R-<@T@+#5dwlYjIjbzVov_$c}2uAnh&4leCh zGbV3n)cxSk{o5+{_Add!_iOuIt2H4Zw;R1C8k#8pEiyT>e?myg@$9tI+m~3)Js;dg zFW6_{_fN=gwj%(+9IiP`XP3U&6s~o~29D+lwI+iOoWO*Hn`*>(DPamgPd4LrL=b6W zrx3^yJLY5P=EDzZ?z?r3j^L~c$6wW&o>shmU;egYzO&v}dsVLFU22qH{)JEkm|EGg^)70Y+D#|0=4XJWPn?;2nj@m3l zQhxiH@UzV%sW0siqi`Oa&EJzENd%XIz;qna6Bh;z?vCBxNEO5hMh&Nkgqu(w1J#@{ z7h)_0yG5|j@4n5fC%BP^GAJ=h)*IwT6VWiqQeXR|5xHxD^;2ySXId?C4cOp4quM(J zObKHYKeHL5P^m@2s%gj3a6_Ip1VI>GvZ__O{6Q050c`;}g&yG5D7*Q{xsZ-$sZt6Y$``Mecy7$zo40l7n&B zlRU@5NTF(f2c8mfh(z+z{sq^LTAgUaCT}r1&9Z8zvb|T}zH#V+p&!SerZC+EfnW5p zbF&s(h};Atwh0doThd9y8hP~yP^q@Odsu#mty%cVPPYT$tG|SB0CCG``^QvpOG{kA z`%;HKZci`DV?WgFxJ$SdNEl$i)}F@XsoorN|8)0Mh1~-F3YS3Vbh>6^K1uHFA<5U$ z?wb2jyx3F@+j}lX&rn*vd}!bgFs`e_cXiph_&r1Gxq$b@d3 z%^C(@+%%#cqIr?C_%4p6z=k=r$rV3%@E^aLSYpwz;e-}*La+}u3)N7)04miW(jris zrVCZ^G~07o{=GrGel88LsD?h6V}3%-oTcbMT;l1&KtorfgV^+Hl?>~E3)7V`>)Nh;ex(o@evczNItYQdu;QcX50_T3PDxSL~O%B^f!4_Ovh7t8AaNUDk zXbJuN?=|Brl}g7!a(()A zTu1D1`YB30L|$7(#^9NN`DfHsJ`SR8+AQdEVs=MEi&fy@(){X#!gyN&v9bbX&~6hm zdPsZEF=2W1tt}zHr%Ab}G6%D*_gJ#U7)rvX(tR|cy%AHRjAxj>AyVhoMSB2+(#U_8SvAWi1-xLy|iJ%YQR|+5*={> z81bA*C0d@#OkkO^$t=+!tI4MZ8Z(gjF2dCaD_`R4KDRnT`^Rgq09({;?l)wXj2>7joYP2)n2o zg4w_dG!b*A)D`mfv*d2EQKNGzwGdVq=&uiPnG%=TuZ=&liQyy3RfCaQs+5PrR&td%{^BXlO%79U1vL_0X8z>TzYaYx=3D=u-T<=bp;?vP24; zqgqNWn4jsb<~F38^td?oycbB+3aB3tR$kz6XCgDs!8B{@136a|vO#k70kO_%(DHs9 z#V^TjD%#Xtp7a}weF5s=*$p1utl`t?I?OCIdIuW*+&GaiNS)>`kxGz~!s2U(2cLk4E_~@C)g(GH_dgtHa$f$G zj8luwI943Ug$X>r8L+GhgjPe|ueaHrH`>qupegCEzRDC^puCg?5J?dakn3ej<@}>n z+T!82Qp13R6BT^2ZZ1pE2*tUkRGit2HkI7En$_*1M>m-kA?nE1LeSYO((g1g&Ia!8 zOU4V7xGrY+1Skh0hXo9HmSja@d(z^bk(xojqhA7-S*O?2S+{$1pE_2icBW;VLKqzr zqra}7mp@!5oxKQioy5WP72ChqZc-JKV`Vb$GhbJjEy-oPBG1meMHFdS;Pxbo;I`*U zO0!%zBn!1*FHwv9l;T#YlJSG3R}G&6M{>9Ee^;Qc&7TtfbbLvvy_oiZe86~o;LZ2d zV>N7i?j_#^1d`^8(8?IRgGuyemf1cJq?Pt-3<{3xR>e?cUoCUPJu$Ls9+v<4V0d(> zd1eHyUk=OLkMEwPraH3k-8{ejAQ-0TTjpJzsh5_}VCqV>P4|ccYylmgDEMaohMrty zq{HQDNJ5z#YH-)9B(15zKz-lH7%#3IFh15|vA+hEW5CVpyuo)hIV7&7JfL+P<7q$k zb-qz)lXQt}a+-^8#gL>g^hicO z#3?d#P}f!rN(+JVOGu`i!d+l&h%Hcrdu8rTKiU5G1@qn4wwmw>E=YOX?^aE^H9uQCYTh|)%N!9CiVX#M7c)rNALdgW29e9c^h8X>( zv{0!^FT;GMihR)#%lab6Uk_};$6nd9673I^R{ZomYUXJCcA?1~0KjT{wQ*|+eUSvm zip!n-S2u96ez~6odm0y0m;MFcJH~zvp%~c26+?h4f0ZWJTF!=YM||Fk`CuR9h|_AR zQ(G=BV_OAloUoESwQ}+`3Zm2sY``vLZ?{8K)h$RqNBw%ttE+ch)s!|Rc`Yu*axYQC zj%OlcabkB7P9;+6S;s(toN&$b5BmP@YkgxcqQurg=p7+)=ilqoYbXkXE%chR6s3mJ z30GD~8}h$@rcLayM%>WOkJ;XFW(ZPgC}s-g8lG3i7D*2hE-c(VIjV{5HRNO0Hsxdg zL>%)|8f-!a`+e|c^yR5l{tF`U5kCL(jSHiAV)CTuGQ;naUMkD|VxyC!P*N_TB5u1G z45GyUIXaJnwi4(tt+_GxgEj@LCC{p9iU?4fkg3R2;KhkJWef(eo?P z*ELI25!t&dc@^~4ozDI$8BRhL6^>xe%KTW_ix3$5YHkAc194Sp0<)fRO#9&|@twrz zBEZ8$Q$1v_N?{Uoz1LmIRy?dp*E@DS1~!T#wU@(8POS@;8Sa+nFsZ4COwr9@5Ce5d znN+X$Pv1`tV8~o7-$O1+I1{Pg)}2&vo>sIB8#UVkdUi#h!DQX8x0=sx&P*K6lNNBb zwLd)eRp0!pG8;Nn`)uGm?}-Do)l*%b+_sgTp<0y?UZwx_#p#S-L$oB^Qok!&O|r@p3vRjFQvBvi@{STSBM>no zxV6gd#=-TY}CIflo!N14}vxv}r??eyf%X8edN-9IKsEExSw)Tk}2{=8$e^4A#*a-3K7AvBDclrB)fzm}f?e?9pv^5wa?SG9%2FkyNWWk2k6a7J4O*l$ckjqR z3Q#Ir7X<7ln~tevkdcb-ZdgOXbcW~GqJ`50%-G|9Lf_ zzuuz`NO6zhxOZfrCVq4!N~!D{Ka4Rq(h8(7p{5?9u~qCCHvOI!J0UKl739f)O5YhB zmv>nL#_Pd(0!I-yEV|uZyCyn9)!b-`5dJ0h_pp0n*8x{Tbs?{Q=h~YTZ`qmmH{}PQ ztw8j8Bu^m5e9}-G%IUb&Y{I$|w|y+$TEp5>>dJyo%n~t-HGP-&cMX9{FBbKi05+Q60G5F*1;Dfbgp@!rdv{uV_}DN?O?;4K>Jg9I`uSzirtEWApyWV_!wG0bgCcnWba&6c>9*`te; zAxEa8VBK=~3=K-Oh+&xSN3_1tnN=~c@o-u~b!bK!9o- zZuW}R)O$01_vAW&9WTge#&}Xr4>c1ICH1;xO%Ue~;or?Kw zNtFcnOx4&TALDl>%E8C4Ujka5`rLno@IaczwKOX)BcfYbmNUj;>I5l|{vd!MU$xJ| z#1|+nt@3v+|85C+g0hE1ATK$5&CQ3p!lQSTOua0z0ewf;+Nj+higIvkg4;wL0>Kzk zZEa6~D%^eWpgsc!KhddGear{oTXhfGsv8^>F4LaQ+f4C3b{HDo_ z5=aR>9yFAD{e0QwOkh6SLdiUhlbA&KEf_==JuS?D>|xz=I3L0hU6ERD0| zun|Z5oX89ozq$1TU0R53jGR0*KL-sT zU-#D>ag7}zQfmmYqEB0apR*aQiKg7X^<&2mr#g({4a|N{yYR-f=o}b2YgMtLV;4L` zCJwN-tD(7fuA;eTtH&+%L;R#n<^WCWuiIMYnyP?fDNsH~JV7-Hzu1IU>9(Y!^n|(fF_d5~o@7+XnNP z+2Nv;oYQCnUE$XQsIVFs#BrCkeWQLVl68X0_2(s}4a~;8YbavF+#*%2+y2F2BMtBv z{o|i~Qyl9{n4g1d^!7PNAHwVU+UQyGT9TN%SPDc-A+RJ6a&5j(@_9fI`CAn(B4Hm5 z$nWA_`1dR&3o5Mg(fCym8-JoyC^f<0#5TzAosxfGWaS-S-H{WBv-9ge=3V5rFs``n zR2*&m4vMG|0=!a)XBF!b)rC3^_Rp#@@bbbG=1FzQ({i=WG8@%WjFeN9trdjBM+Zqh z5)QlyW@vJl+0giHM^rF4m%f^FB-B0(n#4L(axQJL!+i)2&5-H#T7Spjh$lm@C3+5+Lc^iFZ*D;VPN0-MX_8Tcmanxvc8{0uP%w&J?*P@}N%>y*WW zDQjXkD*ATf*{_-nJB6$nbHA*H`1wU?s3I~w#}^DTU7WRyc(!YJtm?c0TGsT|oD z?fvNdU?^(DM7L@cR^TE~payX_dzw||l^<+|c(CC&Skr}G=C zaiQ3;VcWPpMqYZD*Ij2|{>5<^r$s`_p$EBu%|6EEc;+04O*>aZPLFJ?bCGEsRtQ>g+87lk3S04*bhVVSCkp+QkA)+ z(N`B7)2p=w;=CD)*aY_Ut0w(CkM;5&obST%xqh*hSub%giqmoUVBMb5;fwksKkahV zs9@TXyCbMnp;)~asX>~Q#d9otlUOuy?N&hE-%%_iR2L>mJfqQim=%qkcmJH^RnPA4 z>|e044~+l%Dpq8|9Qg%*glE1AG5ns@4*$Lp`t#X>KokA_P@>5810ZMQ>Z;F;Gzhxq zOka2{u_1GyezGmlPjM@}J6kL44!ddn8XV_`Ir*LznEkKoh1U)SdA%M&8VjFt7LQ;rVH6)7yCk)&z+RI;!m?wOB$oPLw=FR`L`eh{_Vg zA}8Gi)!BrSikB$YH@y}*1eW;TXc1oo&O}x|9!%%=Ks!jug_b#H6zKJ>z!bJ}$*m5J zZacG1{PqVQ9llZ;{@{LID9+w0)1W$U4hj=j{Ut<6t@xCBI@OmVAiQH|b1aHtyms?+ z2nICxqVNH$6Zclu>g(^!fU0IP(855I@B@Y^%C_1Ong$8d>~#rJK&hmu0>wGme!d1z zw5RMR*w`~#|KDZ>kTV7^{DK%2mQ#<`+7N#2k>WfL&76*Lerrh0L7H4(b`etIUpV`B zXPVq53fI%t73?w|+#5?PrJaAOJqdojaIl!)VOcwUlUQq7vnuCk*}!}PetbWem)U8Jff5ckAX-m%ca9qdH0LpAi_-L2OwOn8==Qd?GrCreshCwHO-n!_I2b$R9=h1| z69h<*wWq)u*d&j~4|2D8^~R3yqM%8g-S8Gzwef*^ag%9hsiYcSQHIdMu+En z(shL1I0=Tsgocz+*zL@1lGg8uw(5P_ZYt@G7hYtRi)9hBCYaGY2V0(-sJ%nvq+12O z%qXI|JqPO$aBz%cjbu+6LbMsH+>b|APz z)^f>q!`@?s2-hI=)zVY4wE1?Rm3HM@ot|wKpjzw(>;yK_5_Ro!)+;^42$v?B$k?n^ z!)Iz1S*bVPu`5V@M@F%#NyQ1XvtHW<;7# znb~(BUq!DtOBfk~u-xm1F&AHgq}qjTji!S_JWgoiT-V=T44k)wc(WvjB;AZ`2M0j9 zfBq)2QOH8tGUAIz(l3`ST&31-rJbOYuW1-G!4AIzUEZgK$aI5PbqxXMP?Amc46Uy+ z@ok#EV_f_V8&vOi04;E{7UPwXpX@!ZR~8ON>g+CF#Z?1Sk`5MIXgLhQ7H>9|Ox$J> z``KEOW}x&%!Nr5=YZpq7w&nHdBZ8*oK`Y0!t%S}I=E#6Mh-zDm1=Xf~%=*~L=jlR} z84-6eLfi8A&y!W6ah|eY4~Rc8tGdTE0%JNdh(1qU4eY8a!Fj?M?ZZ>O4^1ZwYVK&x zQc$Nvua}-1g8@_Ce=(J$H6MZHmSt6x`V?p@Q;%T|B6l~9@2|nWFqL(7{dlIN+HXf{ zLQ4_Q%*xAsZoQ?A7nYLPy-nxc)ZhF%x`Sik`)w10&_H_YOA$fbNr~(N`4r`00Wc@> zY0(;fNET5q>V4C^>LuLa(HwW5zDO#sp^MztU&boWk21lrl1qwzh$ar=p^d#LGPChe zlzkz?3nX3$;7#Ek1;_AYb0h*6WeqE(0M4a()6BLxTNgp1Zqahi@cS8=7N}>o&!}IPbMO0ix%oNz!q#QX59`z&yLB1ldUyPBRyVVI2q@0HaGR^DvG+VCQW~H7 zr|>vY`L24;%368$-9L-HpS(+%u8l!7(r3ETKgMg9<5&i7E%7XgcqN)B)!aj0sJ`hj zxv`h|)#)BnmFzC0wE0-aVEg9C)K7*7hJRb2x;jwdZCaX6I$ad|g&<*nE;2x(E@r=N zuThd7S0Ggt?Y_cBgpKay>$X1#^2(W6UJ!vOpS)7a8{V1ukcC9ATY@+i{&*ZI%>$tW z8ptCY>?n#g_BFibL8o;dKjo6RAfyHb{WG6ofdRyJ0>yQC+x0#{V{y zDDOfaWPpBm-DzF0TUq=lPsd6br}D3&>LGkrWEAo|+o0FkQj9sABk9j$f^P5@%Zq&? zru%{jvk{x_{j9TH@lAsP3&%#gsQL3%_u&NJO{duj+p}=|F-apeK|}`8T0J-~mp2&) zsEoQ{tOBgAcGaVjSARh&t?|F#e(5HusKw7RV^%d`2(R7ZY)C70)C#d%PS$(JlG#L% zn*6%+2{qKJ*rg2qttWdnUvobxkCQfz`0udFCZqL%%6N>__G_(h;m&OoR6 z?ZL(ms;`ZKRZrbz?LeLq+{`6f2;W66KA|sLm!D}ItB$kF@E5?v3%&%r#q04i=oDpH zTGluSGgl5`19dr(6PY=&kF-Zl4?S>b{3 zSQOs2!ikj^8MtvRuV0h*Q5kVe@Hf1m`mp_GC;hnq+7i@YPp$A8-iGWHiWhzdrwRI12ExlS%r6?5yo9{N5)|Wp z2`e}_fv;hy7pIiLRyM7&m(#=7?_g}|=;PmLnHj2H9R;sL=1~ff3nuz$wO)VYnYy!E z6;O5jOcShzL8|GaY|WKFWeeekH=i9gh}D$Oy!6)g5pZF#R!qsIPx25Ou(T=pRoNbhX99^Vi==VA7L8Iyw?cTG+Z89#sFfVXtMK{nT{T9( z5ZMMD9u8OqOM7PE;ZmExewn#~7N`^T+t${pWf*i_+!HuD)pIbL(m?d_=)%!9ZUg_1 z-?WGQnw7ov(qJ*zuylOKiQ=TU;I^+9tCRl_jEK z?Y_z@e-biykcHOEjB4kfc9$EKpzf5nieb2-eX!%3&@QIYo)Oj~e&Ez_s6xSr{$0QW zUin&Bj7K_weG58`uh!&xM`9sL6>dbYNxgTiWq^9q)&5O`X!kR{%S*k42oobXXWsPZ z3tx5VP*T^s6G?r=?S@i=`12aK0*eRw0{tnh2s|Aan`wkh5*12rIFB2mtn*|<8KfTO zs1>C=6>9Q8S&~MHehE31wM=U-*Hm71P*|CW_=Z`1pSD*_LnW^Lnwa&dFOhwjKU>!m z<}!bz(BQt_J{Ils-tUQPW*235W{=C`@MGuEFll}s{qYu!@RQXE5bR}*SD&&i1{l@? zlmm-b(0{{4t%nMsgEW3q->;5aKi9RlJ-sweYOJSb>Q6A{%Cr)!8kwFc);ciL>8Mio zstfm-fGLS( z9_c=={(t;teYcX~?&$?U@ujRw9oc;%GWh+5k~3%Mrs?Kad;6|E%Ml&0?6T8{!w8d? zGkDKpq<94T^{p7or1Lr6r$muUSE~(zoU}>1ta7fo@Hp7&TU6lw@C&ldjoAn(E=0j6 zC`sm&fG+jM359sBEzx-Ux?wpe4ogt*A9}MxQP9my1mXB;jXfxZjS``+JCjFJEJPH# z1nAs0uW`p;k@HL_W7W+T!f{Qy9DI?>)qQ`?Yl3LbK@Ke5j|0=;d>rB|F`t;%*COz< zfP$={iTPo7P*TyL?W$M_pCcYzy%yNf7wo_Mn(V_p8DZnhtvKU8nB1X%`KdzZffF`| zW(fhA#0;V8azU*$yUA9??a01*M9i?+I@+t#K%Ab$$G;%of`^05L&n?!A_)iUKL_eV zDqGCQ#F3jv)IVZ{Q@<#@lMGp0UyMrp?Iz;E4z2PV%O0iHBX^#GY{cPWf4(9v!bVFb z+k|Inw1|Ka9)^?cw`XCDd4#C17?a`o`Xe|A>yz*W`%Q{4HuJ885l0=_73K{p52?|v zCny3Xk^9x4a z$`s!HZwYIj4`HVAQzS}YBVw>HA;~RWx(7-~Q^e{mmnJyIhxy;wXzR4xAIZmAHEtJ* zyJ`gfL6%pIq&| zZd~j}eE^4Gas+eszuqj?w`J?Z=;YHhIixYdX{->5(2T6j>BRUf?1@ZYrobyR;44g^ zY}-yd5Z{nVj6xyPkyvqh642t5qQJaoG>>~#p>v^yA<_litE_J(AtHJ8`>vQIjE46b za?aD^e)F1y4ZBs6(l$nzUAQ1X_~ixoZIZ4A)%!?2}Xfm zqokVW89w&88j30bb&*-oU<%S&ZNTNUydd9Y(F`Wm&S$dd&>|g6v8UP-o;CJz)Y+}^ z48bg~a|s_Uv6NVr7fxk6@3W6Nfpb4%LYLLPc9_c_S!wJ<=BU{hfd05LKI*to~NT?kp zH|zu~A3UBj#Pf77V`xZ^hF5$!hW&l5y6){yqMIFqmNUmTItuH5fpB(hpnfVdvvEas zX5Ww*?$4gMAmT!3ZF7?ySX>ig2vk|8j6P~qfY8r=WG<|0!dHXT>b^u~yQqi-zxK%y z8xZDBhl8y*`x)e?T)XkLbA^QUCfR)M0Tfue>c!{JWY;m~lc1VdhNY#MsULd9R79bkE#4mF;FNl2l)t9*%`-QBGNaodI2)lbji5w#kau`Ddwx?_8$1W5 z$T3Brn5kJ?@>Mu6fk(oJ%@%&icH_|veS^t;@@{S zoQqKh?>%Ds-5dzhjfd;Fh|m<{`#--*b}@^@&>Ly27?$69d>*p)CBqmHL%sc(X`?^5 zExmeLk*@ow+7r+acC*NGK!eXmN8Rd#?z#1YR1f0RbdgCl9~xfN71YGeD>S`J4GKP9 zTIz&^7R-Ez3u&|qGs(B#eUepYYHaR9Qgv>qD4pzpEG+6m7tq{g(x=V(4*y@fJ(aPCqPDS zfaUYt-)|Hqf_RhEk}VtbP;YzNPXK;J!ttXwncMwvd5$r!DObv=Mk;8CTeillHOm(r zi7t+89fe!%g1+1!HZdF#pK_DNBC;WqF!qO68)}@u=9Pv5Yd7(_I2GvI*Aw)SbG>xQ z3CZ=)AK%9x(`}tvP$e8HSzl1(GHwH$Ld9wPpAl>jsRx*l8CnI2c=RU zfsVx$kYzC_#2zCX#KT&53>}pZ>`I}1-)m>)0`7;C{h67_T-ObG$U%_w#XY~aahGmw z)rR(E=$w^d1Wn7pN&j0Zz-?43)0J}T3_p@hKBeSl#r^Y=N z$cxxUUm;c5xy3@6^Flk}Kq9x-*E_0HUDweLhe$*P7JXM~wikomPezmFUz?yKA<`U# z%$X`QC~48C9tg z$wOa;837n!ViSdUXW)+{9~mDvY==rr(TjdQnNT-LrEBBp<#*M@dH$H)Im)|fJKhC_ ziYV#eSAa5q%Y5QaqzL17Kdch(xMV42=XkP)q^Iemqa>OWtFUYJ1H07rEJy>7UXvk~ zQq8QZN}<_1PeJ&R0G7nh%CQq1)n;S1GJpbBHSRQ8o>7$XA*a;}+2gFISU*GtHB10b zf&Pisy)rb2la4ALJJozhhhT(~_B<##~ z$PrfZ_RE(|82ygDp>Oc{>Lu~dP8Gj;;Zv7eoQ!KPztsDN%7?iMzTS7|?CC6X|qmV;U!|O zf@>*Nz&Z1ALXFYmdqplwv{?5RfjP_E@n~^THHI9)A%aMn)hODpSuZ{ea5Tu5z&5A7 zU~}FV5k;0Ve|>>EUqiQ{N*4rEnv@9E-m^N8mGw%$M8e`s97iY$GsRo^MgA^r=xm&f z)VG^`#(?38e3$A=YML7zC?DEp4OrYsE{Fd$hZ1$h1WPwd{qsy=ohs|wUKoU81+67 z?vMffofIH2{v2ZW0)%)3C{p)BO;Ym|i#63!6rp$LZ6wi$HNwfp2Ftp-mGLmoS2JVh z59?Mb7wq7~;oPO_gYsT{odB3erd`*>~PX+m&<}!xaOFQ9{n^kZh*U% zps1YJ=^dG0t}3$&68Z>sg?r}TR9NOM&45*p3T{PrOk+n0=ny@gUaz@t^}*ru`#c*?*bZ=86^HtJDt4WP-%bFi+h@# zeXPV{aD|*^<<2>{7%X-i7Mds_$tZIU8$oHibh=kyISgL4CEWqisPT=r5r8;q zCc9Gb2D;8*7Rf#Yq7ydJi)HAlb(aEWz~u9IHqN_)!%^3f7u{1vN%=b#&gAZ8cNgrN z(fYw_%<=ufFGtBp?()9ofW=}^<;gDLg|NsR&uv`z|L$(t>K?uwuCRe=?BpoUF;#WV(=B zIX4mYDsaH8Nw%cpK>@!{ZL?zy=Y94o&z_HOCYZNXRhV3shV@_G8?+feSxE{rCysoD zTQmLekTntx1l}7QRXU*GrAZbgCVRF26Qi|)?IJPyrmu4G`W@GyTr^l>& zJ%vjk?Z}Xd8Y#uf0VOUQVyZ&Z?8@~-+ElA$#vY7z8B_vlR9m!g&I30lH}N^E zJiqRkCjhQ^@*WWN#+>kFYa@hr8Q(N)4eL&r${-IRhXIYN6#^VfLvP^~*{hVTbU;2HD5mnCF z7a&(r`>AZ7KXkxgqRAJUUcMq)JycjYSx$5XqQIHYp0t_G$yT4Rus2RYNEodk;D;d} zUcHmRMasTF9yh>((s#s>h2ACLUIno4!l?gz{e;FVKHjrcSw|0CRbyD&J1^bOWguWe zpwQqS@=0zpm}sQsnI8#+zN+$Bnj0>%|3+#L-GK-G`boo4LJ{A3{TJik8O>j$HLNfKe=@2Dqi&rM+`MC)s8;<3q5 z8X_qgalHoe66=0}N|uh(Vz{rAbSGSiS!DUm|BLHqn9?+zC=5!Qb9pN&ELelC&Wp1l z4pG(e;io?3QgLtF*!lcCWDD|=e#V^pOg+$0yoHAh&T_8>T#=|GY&7qdm-5(qS5qC zT(4iu7Mj*ESKcWMN4lv2$8%rV)I8@n@5fl8=8CIxE-c`+v`U7uU8SObDT&CIghSU> zL&(e_{nCN*`(Cp5MX=y^qt&n{Chvg~f3_(C%}`U0XQ>w?$(=9wHMt!vUUzb~e2@gF zt|2Rja(S_Azdq#`@s$2d5?tGfd39_4ZNCJsDvEZd+?3{f{AN)1blGK|#%~~FimGB< zm+{S9t}aW1UC&(}1cPDVDxS7!K3-O4;P}*~q?=DujrF!EP_By64As5-BYiEHKGirY zhmF6qWoB$>th;x9&^RYmA~CoJrl*dQ5D%)Bw1+K*g>%HDlC?!5-|{D)F6FN%OHfA? zEU$9a++QB!Z;mfPx2MPYlZ!2Lr4vPq?oNW5S#g((Gx4QvVG0W}U|Zc8s<+>U8_MssUG++S4lhq25ka z%LTXnk#->uBD`7Loq0NDeLdH=mIJX(nX>6J7)>pr&Sv3D8Mxq1r{EsO_?f6!bUk1v z#Ae~H&bovidCW+IR4A3ug!BSNC}F98KzLRTPqM3e>$7;DlHjlwpM)&R*|%{B#|l$F z<~RFn-qmwmGK=gUt&@H1s52>Na@1dNn8hA3W(Lb|`${SN*j_j6D^cm3pD;O+a%#c8 z72Oj(BD?`ubkO-)uk#~L?0HB`c^CV@I1;wDc&q!QrIGcj@FFe(9tbVVYt4xDxh4`^ zDDw865WA!_Hb5r=Yg=dfEYi(YX-^fCgEEX`_7dLq%-c@Qk>*rz&Dlrm2toRdiwrm)HxD_aHb?KSv=Tt4dF28b`G|{ zKrtk2Wlp1VHhI9FF=O@=R=q4<34jN+lV19Vr!K|E~gUQneH~CscsHF=rB8qZSqncY^ zEVHw$UEj>-JUE0%M~_N@`AZDf;FvGjW~q7+e5m*MSjIu{X|=AX)KS6EDmgR1R%95n zQ|l**)U^7oZY<~am#mK`3mVJd>`TONqm4GfN)2$sILR2CO3|Q{iI@U9>hxfRwni}F0eY>1-+R%We~mF*c=M=tNE$*L2oJ7wdh{^ z&)ZA0@(7;7D_T;TH8qIseX!Xl;3K1V5yFCF%}js9*mIk(V6QPW`Xx}%t(MeRcec6q zn#Q1dYR>dTj0;MhCj)dISQZ=X&u0t22=_^);D?q&+-?k#LYn z>1&uO)$4_zNkX6N8N}c3t`r~$wXa*qC}AB>DTr!grRl=a@fcR!^j3XhJp29B5K>O4 z$eXMMbdAku$v{>bLV31ayyrn*)*B>c2DjH|n^vAKF0+PPmv7ctq7nWD76Cyh-yVdU z;U@Eewwp`H<5#KkItToy$)ILmH7w2$-J_IWeo`G8&Srbk8BYG_pp-3a08{a>E>JXI z!?EZn^ZTRe+V~5%x$=hK6JlatkgLT^# z0)8_Gg1^_xLp4?+ubkXd8L0X?wV@B^9YMJ*JQMxJ0b`h3hC?6&8YkQa6EPXKui=h3 zCxPfWsdeC(fMJDW9q#p458-6(JQvu-@9!poMkdVCaMRI0MyA4U$&0ZL`=g*^Ft7C+@JlYW4+nu;;8~|Q7 z%Fuc#I&NGlpb~-=kQDSX=YfQ$SZfNHyZYg%fw@{Qau%2G%IXmnd2wMr*r-})l3Pc{ zYDN&RT=6S>*;8T9=<+QgWm%S>U_yQ2ecnKaxg3%HzQNwu8FL1LQ^j5~SftuMvU+Q| zn&k>~EF?5CR1$r030F;BDw3T->)0rp9*IKFPe4wS1-=Nc`(*QaA;d0Ei|wM=*zPQd z1vCSXzQ7LZw3>#Wb2d?0s&EXu#iH+#+HUmF2I$VpoSX2Ptc{$moYXiIk->dcY0F(nb|Qb@gWJPbm?8GH$^F&Nft7sCwLA zvOX=YRi8!>@=eX$Bl(mEv-I_9WNJQ%%oCO9;;)-GVtuEmBCQCc6w%*VMnut3s1F!* zURtI#4c6X}atsvG>>p<{?qWkbacZZk<{#6gbFa9Ao4P~${29qDen|pl$lOjX#Kx#i zPJPunmI%DziBI893mjcsQ-tW^y>im>;YX_Z1DPc8W$Euk_OEe|Ce-tao3>vH=IJ;8 zUmMr7Pzl3mGk3Oa+nj2$Z8w=a880!K8xugeUXe!H#yLx1=z;V0q_ z-6c=>%+7f{ci*-Dkp4UMNwk7ng((;rZ*^tZigm*$M^8*((L7<_O zaRP?i=_hWb5ItaCXg&M9k1x>t1clBK$dpHn=u$s~mvl zA7ALRH&*sy&8DkK<4AGb8V0MD{XMJ?L>P}^ANXPTou#m#DxM7kP-~ofzCFCQuRLsK z$F60-R*BM22NZhm_lxvFn5$Om0|RT=M#ZRO@?c41zUcNeh6d)?BHC6b#;wdHmqYR zCHN*2yEvB?4Ax7q{rcScsFqNq8|AO8x7Wc3UpqMtAg6WJ-Q*+N#(^6L8yr&8(r5cK*vy5)f-eYYVX1IKyVGCPx$ zT)M4|&#}V;hsWd6cMvA0MUs`F0sBw32S4Cp)9HP4?mspA)kKEQUIFKNr)234AxpW6 z-$8%;1fcr~H-W%f&OP{HNagsa+vo$?LfyilC0$b@gF?YrN}vA+BvN*p`rl0?ILOdaO0owF?YPVYk+GG)q|0Sx)R#7}oZ>XNuELyiC80#~RH z8*q5y6o6fsnvh#?_#gZP4y7_2dCkB$EzIR~!r7zQYnE#+>brtzhrtD=LYZClq>fT% zReaZC1^Th}NrW=QE((4dbO&o4s^D@X`%-Ed2jMd5S_pI2Jcs;4DBnl<_o>h7?HkTY z@;`bb*&B?nu6XyCgszeI6Q!Ud_YHQ#8?4l4cOD|XtVm9N{)Ks${!z#=sdVPhxlPFf zr4NY1vOQG?0cH<66j+T5&TI>PX)Q@nQMRXTh02{OS4e5j{b2)pDQ!>Ps3Vq18;Rt# z4=eFFWVik;qyIL$KBuW*WilNodtAAjwp)uy*N~)*ZS~3%m~1US5p^NuyTH7NkCWaN zs?}m>kx+*KeS-W%u9S_JD+7uva-o`{vOFK2j}r~H!OF8MggOTLIe@}hn6K89b_d3^ z7jOe=m3T81s0b=14RdA@ftd3dB%g#_fcpZHO4`m{W)QF%|xf5mOi&grHDiTPPY;CX+}=>DAe~P?rmsT!*)ZW1-URh zYL9W!QADwHtoi`u#KL^YW9-+{+DYY^W!IkiB;L=C!&vH>`*e>p@l*u<$W z&2dklOYP{%f{O5k5&2KWe=N6SNn~u)+nyx`0&C1VF0!41TtCaZwdp)mM8DvFa^A^{ zvskVv{b7X}%Uhl>8!Mm>5ZdS9ePm%cilI^(#8=wTAAr#uXpuRWcgtSKILYNM&!=`& zhY|l}d&4NaRLY;dzx2WIWP z%N0V(b3Uxmo}ZCoU?+L;(M=19CF7LeQ4$g1c*@ssf7g_#43JqDR-FdU#Ce|GvTEs| zUa~7K?xx#)kloNx*|k8(kf!bp`A!xiBejV@+!77UMx}HbN(2g&9K#>TdG(=~5lVYH z1)A)x1PkE7ipGy6fH)33_-zEWCA1nfjrQnx?SG*BVYdIld##-ch`XLYGgkB2Gq74o z|Ca=#(J9C!;9)EP9 z`LRboW?Xl47lgQ*+bO7SnTFqlUZINAEaFU+OM*?X@f#T2!vFd>K3T|=?{HiNiwB>X zwgMyK7k=`rqx!PIzJAqUKHenF|Hbf%#-Z*#%}LH2RH5gB%WUb+AFjS*$RGp5wIzWw zSWvJp$!@-WPoBj__OoIOB#HpAu#-EZ6(KvXAk{#`>&0t|MDQf{`U z=m#!st3Trw)WYo_f@goK;{H8N^(=6aBEwYZSnfx*1?U8SVji7iFPR-!urqzfmm)Q! znx4^`Jq>_YUrZO<*8~j%u=k3)k%WP93;%juZOsIu`gw7%BWQc;Hye8KjLa!1QcuQV z+vkzdr*q2Da@}$N4UtN<$S82HOFZ(>JhT~vNkXd46vx-MahN3`lUE|IK)bYLs;?7^ zi^gf%Oa|58v)EkCJVNsFU_Gd4fAuLy$TDI%7r#`lL4AJoV(gDDyN+lk&lBKw)F&fephx?@>q)wjJBcGxm z(TXrO!0O6m4j1lT1!Y*U%8z*H<_nTvwp=%`Oq2bpjs1U9`a=Hx-gxk>6m_zfgJKn} zVs_5j<8iSf(uF|I@=5(bD^TAlkvzAp4rq4inHfpgC0eKv3S!Xzm16ND$ z*OpdSW%baMvuun9UXX&aHNvNX+Ae-LcItHB2(?yxh>w{stVH))@id zOz=6#6E6r-I!RdR4t(0M6S4Q&i4{Dqcghn3UDtC9OaFO(EZlM3)0^IjDHrY zgw`Nt?x#G0*-sY~d^5zf|fKw*av&mwS^xSAQ_N)e+G zrAI~`eFu*0sq+&2Bj0AQw_qdO2i?HO#B?mmwm>K(seiSFkdgRJ?E3=tSp{ag5*gSe zE(I{MZ~FDsB!>{$Wmzu|r}S$Qm2$I?hM|;cfVPCUbGHE0AF6*C#}O&4o9DD-TPu1y zoV82B&0Q4jtu%f*Rix)7ou{Uujpf*B!17?;U$9c zZk0ZAOqqExXLhP;Bkq=2UoSgjXY*6dH9dv$zL5K$N0Z+xJE-P(ldG>>=ka%KSZkcN z{(BMVK~UQ~(DIZjtj$fY%Rzq0N)n}Fb;;I|OVm1Wed_$L}(3Js%k z^!%emJ)rpN!4aK^N&c&~KEw|HqfpvOevpk*%$YBb%Jq9`^1heskK|L^tE@6^xA>E0 zK5@$ZfC2rca=Vj3Ih3%8uVP+40c;^U4~3;~6{odp!6S?#$$mP9*HJbcZGavQ+$T)f z2v8!Iu9bilOA!lY6DvbwBXO1W9_d~beTPIo;{XPt*(Tur;BfF<1UX{M;+R@)!Yr)| z3gd%Q&Cx=>b*OQb7JsReq$X6kU%%$3^UU_Pt^p;lP#eMN1^=JZS(tN;Ut9*77HVZ+f>dLN{E#@uQLiVk)=#h8I_8bS^db-IHSl&;!WG*NYL`jfrkN|zb)|HBkFTl5K*fG9e>?U zlhjlo1ISFG`RXko;4z}HTyM|Q3^gLL4QCVA(5V-{7=f3gR&V=~V1%|6)T}$ff)Y^Hy^uMl(-iE3mfHj)g}HS@Hn z1rXO%7LVooyew?4g7FnseSR)Dt)?Zf-1~dgVp1Ew1$ivlq!ClKL7nRbUJ?>g z6)G(?-AxM*2<1Tb;>b03iQW7Ij^Qut(7-Jjv@#w~*wL+b&asrbZ?+8M6E@;4Dffec zk8nNUTvgPG?ovSgbbA8gU{+}6WgO93qzCP4yNgODjPyrqr9g~d*@C0_@LcHfYX3M_ zc#)a~*v9+bq=!pscrSmAyDu+q)rC>vb-gv}yp| z*zk}8eXyW(4owF>}UbsmlX)C|fTn21Xi_?VH1Xzaq~r;N`cRh^0qM9UXp{+gbnJI45WC$25~ zy&K-snlI01z)5(Odszo7?MRR$ojoR(+`iT~FAifRH_clW+}WsNT%v%kuGxjN-ie9^ z4h7qKH@Tnlx#J^I8f{j8oH2Y=sNXH?pG~=gEijbTsDO7hwB%nY&Zo|_M7%F4J*+V_ z@0GulA7R4utPUW|4fD%;UyQ7UAKigN`!9wdiA_y>C9$C=j}1B}VJxl&>uOsay9_ow zssGup=V-Gsszi;?`)ipB0rw|@I9Xipb0>t@-24pmcvnT>&zkuc`;8(}%pF-~Yp|PV zOVLg1{Y)9@KzcecCGMWCF^nDG>l@PfRJn3gliJlwGfsfXhz`@d|Ka!6br``uVti_G z=I+ucBUDvOyzTmUui3gd=Ubp(nVWs1U<<3g{DLV(V-u%t|O zsSdx5OijcFkM|6{i*FC;fF00Uuk0jfjN7LZRgfp4fa38a=#$uyk;ppB*?BtH^RrIl z_KxRKH(+q^%L`OW729M&ndj(cde%de2sF$*`refmRh8I1Z`iNlLk z>`eGhdGif2)lk!U0JyDITETMq>>8v;U#ihpB}EEf&TH}pMu(dq%@BKdh~^1AF7C%V z2`;;ajGaS}AV9W7%eHNH*|u%luIjRF+qP}nwr$&*-ppUk;_ose*BOzK@7{By81fsF z3TV-61PmxrLe+ctV2Eb%?0<%}tymeup@zq^3p));E*D;g_^_SWUS~NUSt|3??TXs- zy$C1w5p%;+m|%1`i9nQmxQJ8lea#ti+i%wS*k_G{6J;4Od>;x&DTBG(Xa4liCggUQ zC04uPE779zemeydH<5J;NY0IpL)q3nC};F;`p4hK5J3=ErUdZvXk1*u_EVSdB&qb* zTs_1Q2H}UZeaSTb3nK*1ja*d$kx9*N1_0R*2iSoX&%ngHk|eDglWOr$7XU70q*!ULH<~TTG;APoF22wq)3IJ_oz>Du`-b(5nbaN2lwXni5+At?7Na^hxkkt) z%kpyxQF6xX92pOl(pfqO# zWJiM0K-`9}J%r2_pWF<~Zsu>2H`lL4yx=kj0mHvamPIeN-x1bYI(`e8<+}rR9`UfS zv@|KM7R`CZ5)l&BG5eTp4L_%CQoQUN3VFhfnB!OF7Ftp-_QhkIJR}ZO>Dfd)qp)W6 zg6Wd36G>T9aE0TVXxIr&Sep^#&hzYI+Kx!Q^OqJc&5!;g*)M!$y>SL-hlm&;OCG!S zZ}7!+-tz(b%%sh+c%SSNXUe_bL?GoXLk<(Aak8fc5p{s78{N@>2PwSyX$TJI5QJz&;43juDN2p z*s4(m`c34#PG3c*gY?I6#- zz*$P?%aJ{Q5cfv&P>Bpdc z7f?+e9A^$b0sv^4Cpil4YxQHR*u>GKQis#Pm!6e~?k?V~VQKuBxWBe5RjadASb!)t zPaSX9Q5knOJxEnqzDD#KSVReY<2y4T?lH6QVeDMwzppjRPU_#)$O+$}%iA z3+KS7>b`i!&J~|GcYw8ZCZ*S*U3|v@c5I3=?;p$SJYdWnG;=)}o%N)nI@^)3OUP>DOQ3 z?wsM^vtUbFjdwzezvZ~#r?Fs@wv}ktx0y@(NqyWT#0`({f@2VR(?ZB^sFswYjI$q@ za_KX@5i|65G`mWn;=SWONwgg&UfX0vk^pH)t3cc5`g}5G%4=&~l!`v+^FjMv(>Ei2RA=j$GPX)uF!`h;FTJ88pN$e_=j+aKM0&j+;%SA>* zIXAYn|6UJD4|dMi-O#F<1ER*#;Y|ro!o04|W+_mtkxo(|aCjIa5=W{1wPvS%e)#!q zpcZpl!N6VwTR{3U05B&`5F~E=hrooQ^LMc87@zQe%*c_|sKaj(k3ck1i9nF_O6PZ( zi?vT=yTvY$@fO%~$e_4kqgWwJ%m+Mnv@I+)2o1;#UmNg;VB+K+G^9aG!hSTN@$->A1V@ zgA`Z#4m|MA4xa zyH%1xrFFZm5UMSUX*!hX5Hp+9=Vqj*%w{5RJtPPHOW~F?3f?FA~ zN4@aiQNe1A2xXKMyAH>f=->I;j5d z<04-IRU?uEif}zb4R*qY%D-@-N_b}+eID0UgKRRKThQY7;NVe7u{j}0uAl=x#*?48 zN4$~)mpQW&>XSyrEfw^7_!h5Ww$9awa6HC5z)0Og2fns_GLO?ctvKgIkq!56MU=WB1*bd zwD{3}xx3T#x|6TR)d#Bx3Oz{``Ldx2+%)5s@6hd$rc9V`qRhPX#BscDVig-QA9Kts%quVXi?O8XOp4)xh*}b+1*HUC1j6=I6>Co zC|sbTI?y+0>z#VjV;Q8MTRJt?tKw3F>5R~78@`1c9r@FY;Mm8qO2o$JLKd-pQgE+?WBgoU@H+ldsPo1qV) z!xnd|qBz_;;ZS+>%|>Ivkbha4qt_$rgoZ%w100iJxSkLEb$qbO8gholJ$uz2SR^yI z?dVk3ha7oyC*}!eyI#SNyd3(|z_Gs6#DQK|OuYe=reMcLKdv@+;4p;mGG)RJwvMm; zH(0d!pHtP%M(+f6bpPmeh>TQM?Zu#?veCsK$KnX^X6I^$JBB4MWF?sCiL3HDHx|o$ zLS!+4f-=RU!8%(DWWV3}VvkRDcWW1JcgWsD_F6t&&%OI&A)>YpX(8&=nTvxL#4iVZ zre{aIKQLg=C`mYlw5CzkvRcLYa{qN%tr;lnxf%MpN#(MK_BGG$dbgh@cN*R zKYzSM>FyBsK)jSViK{j>u$S$Nv_XLHhQF&1fBf{!=K)*)mq%0Z`(|w~N zuyQn`v&19oB@fW@8Dm1z&59beI*VQCW`Gk(BbuEh#Z9B1vAUC-&4wQo@^{Fw_j9FHZ)?Hm49A7QgqsvZzY0cLWw z3tRnkV`pIThlYty#lp0E=WLHqK^Y9>-9{}C3Oqj5p9K>3rx-mlS!@*k6dK0nmq^q~ zU1SI}PkD3h?}w^0PmhmEO7YpoQ;1#SYs6x&u>nYuI=b4B$)~sTvymK4OU$gyLRjs4 z?AaHhqq2rmsw{-~O_}=>rt8IHf4&j||N5ic3NT>7ZCYur2h`WfvL$UwVA(d$Diy;d zQ1Z_R?BoICWZ@iu%9l;fup6KRtpuf~>zFmj)PGz>f|rDqczCKjNWlYWS1z8MMGt z36bf-@-4hS|Ev9P16@dfqqiA((*X?8Dq(fu!AP7$5f|AOK( z{x2v#I~)CfDfo;83``8ntpD9I5&Yi?0RuB52mAjMir?&JqO5hXF5rzK1*QW~d z{|+_%ZsP0#0ex#>S>;XY^+;A!NCJT8!D2uG0r#KKg~$^uEraTRq+T7vHspO~5+`zodSb#0~Rcc;S7ee=n|K)5)+-^FfxecEIS_0h7@%BS`c2LPn7 zzU1*;>HK7+!@kj_H?&1`MKy(xMJ`Ak`fDT+(k%lv3i9OnYW#+i(-xBe>=~T`{=IVq zUen7^#ZQm^=NurI#k2mSrnl)&c%xtYOVio}sHOf>-W6O)oQbwJ zGLA|@_pG-y0^pve)%Y#(s@D4Nl7AugCiRa_A|620foyd8d0A5n z(4i~Sq32;k064pVzPo)Z-QPka2cqj67{Pg4Jf#{4d{cFBuOR{L{c3#;9QCUFE(1OJ z!u}bYOw#1aEUppu2!EWt4bXb#*N={D&bJSJqvqeLYx@_JL$AQ| z2lThQufi8?T90Yyr#9ku73>Gvia(psg#Z^jGPQJ;Pddq8-V7Ta_$3u^&n>>U>*q6b zzLzxbef8>RP)iW#V1IbvI|jaQVEorXQQM2|Gbr#ed$#vIlE0 z06vc6;!7^9UBAvoQpmi%_9-1=o6G^Z<3sy4D+=oBW7tDrN zSVFdMjExfDOC^;IWB13Rn4{k7Ye7)P8u_g9N-BVmhBvEvx&k#VpZipAL14VDHZHqV zZHy;*9%)q&Yz~Vz?#S-4EBXM(bA(r}gqhkxV?xA z5*DTTQ^|KOz}~iEChktSnU)6F!G-WL2m&)G?*o?yd0kZl^|6Dx zX!209j%0u|J9Qck1vQ|tBcxyQ_)tdne)R_fZmzcWpNL7tuVUHSi`ENKfO;GGf>mjH zPNW?BX`aj&Uo<$)io7-uwF^}!*M~bNcLQ(G2(Kftx1-Kat%K$VGHhZD+DWSA>q@h^ z8xYl#G5JM(0^~+%yFIOwCsZNQSXd zEcS_U`*H;N!-|%cRqYNyY?SV|?oHvwXiUf(@YubDHA&?(xb$_7Omh$o;(z10eNspw zX_KvnL@~!Uyy@y{TBpDFPy#=rm+l4SUQ$>;ENLTIL6q$4qU$@DP)t55AGF3-E@I)k zoGVMr*P7IFEeO9j(X+d@91@~{rELpwjn$js?`kea_rg)f->u#7(QZ_E_bZYGR!UUnRjNNZt-O@?S&TL?o- zkCu{8#d>I~C0i!bDX&xisorT_WSA|QA?`#ZE?a#<2Y=EO@$QmHF5f#qTWbnAieBxF z-zS3+ZKoLQwLj`lc)NW2(6lEmN^h$E>7 zX%&5tL>SPAwD}W?<`#F%sJFh-x2}xvMXoZdm6_1cFj7z@eDUcuhj8wt*xKtj#Z8m8 zR)5pJM^SdTATpl}MH*R=R&+N(z)xUrbVFUDNCmw8Lx!8Qy{!wMGx>A`jMj z`fn6xcEW@D6&?qFG!GlE&t!R>XUc9-lwdX^$;9dY8sx&{KQ9xOsJ%k>N?#?9tG z!AqHI!14fSM6mFQPRsrrq>Y?ohJJK)!fPlfDvzoU*Lsc3*FC9i3T(BO*oP;{;hhL( zF-b4&b6ej8SWUI81ZHkb`&sbLB{6+tw(xn_W?iIoBY35D$mh9M$Xu%1^Bz>Wv6q}^ zvo1P5lQGX5Y*MdQ4cuXRbL`6GJ#2*pwA|P(o5uxg^LKQdy9XnhY~j2x9(f)jEw-Tq zz{q%)>8r3pSAHRenuC@@UO0zzhH$kPSct~;Xas;OT~q)gfCJDe1d67NUi)3pL8oS9RaF}0ZJNa07oIa)EPV%#b)>3jyUO&qIxfB z{TwWeo7;SL=CRS;Rec^wGxMegF4u}#mhj7(Ok!&RxOpFr5lQc2SU}0a=)1fw#cZ*0z@>*< z3}XPXm1%~s^c=0sC{!RjI!6(k<`6+%tz?_hun`R8Aj4{H9fID4{m#Mo!CA2I*r?Fs zICP9*!Op)OFI*F@{n{~Z2Pya&nJ;=a+4KY^yhK^&P3W6pfqwV5*w=4=TY%*T;kH_B z?^Y;NeYK7-wy^t+IJhSXRu9@=#k-`@>336prtA^r+RWVN{O;kz;X)vqah1Y*TFq9q zml837yernRX@3nAUTacPZQH@NChexy7Q|dMqABp~_C*CsfLaCcD;tn<&T6p{Hg9L0K1GOtsDP3mWL8vR-@a5UZ4>wjt>aHUMbX+cFSh|@lsYkI_ z3`IeGVL87dg!%$KiFko?O~-8~8bPB%m0eSW8c%F=u@EV7G4v!Et^=v z;{~jJnKlLW+FO*fFz!ObKRBWSZs6x3bkV5nG1J}gr#5iD`239YT=8wbcPw9ZY?qNw zXLk7w0)g;w+xA-{F_!7;NkWM*F^ zQZHf88p(HhJ;EiIyIDWWdI>iEA@WXWb0i}GptYiKR&nl

Tkq_1AnDfBCLJSsySe zH0WGW_fibr)DV;&Y6Tj0Z*svr>VlAd&lf>n$LEoQv<6{5QZgqWdYfr%_cx;=e;!W- zi!b)JsW%iz%}&E_?&0mWz1Gtk#}ugKK3e&rDl83L(YFoPmQni34)ZRUBYe`%#Dw(K7CqC7T9+Av-v01A4QToWFy zm8r&PZ&WuyUd@cy1L{w1nkgP$f>^hZF+}geMO^zPmxIMZ%&uxF`WX{3v#jyn6`VSb zv>Y z5nOS}``kE^AK(-FL$!Ues*~VdJGtmmRoJ2OK!!rswm-iJ?$WE?@17Oi?%7SVTYaZ~V@6*2R&m{=%>$ z8+9ttN78w|BFv!;-&46{w6v={f)i7fHnX-t1q2Rb$ym&phAK( z@M-I5UOyFI;&Uy0VHD{*PBQ`xir#t^PXe=bQ%D>hcaSm|H2dr7KfYMDEN5d3;k+DX z)(i%uoM)O)$9wR_a5HRo0H9%hFX-As7H5S1Ks?gZ+t?#^Pa*{0?^`@h{OG6v;pma9 z>y(Rw07}z|gzKzF_N73b05ol8ZLAfUiBTPgRY-!1LQlr;ie?lbMuP8=m}Qpf%biA} zyUL?PsHJIvW9_v($ouK|whmY=lw3yl9q1ZF>oYSXRT$Uss+}AWJYOayldz1H2yXRR zsa71BRO{;qB-cWQ3%{A{CIgV!?HJKpPHqgfoL`JhZ-wNoBMyqNK~Hbla(2euPemW0 z@(r!A*ekf~nWSL*77k!J-#qC{@wLrVF#+W=!**FXG)-W6F!Z_=$*V#@`1|L&{Pn1n z1gr6ds+aTJ2p~$$+0uHqws08j&pch%(PLD!6dZdaXQALi&lmvDmb^^$=LySMq?=-R+LO4u$RPT*B1lQ zfp1=(`2&;gt=@oG4ZaYN*09hSybd0Wy22h6?vx&UmQ~%3FR94xzj|Z;%T-AGIHx~8 z`V4Kz`?TYgrF5WYWetME$53gSy0V);+eY*84!lyEVh*+BXBxi%N5I`uCA`pcCQ*(Y zXV2DsyG!asG8^N3P~j+J@f`-A5YG;c49|`mz3B1A8U>o}VN>hhlHjV32saiIIkMIW z|EIoptF~Mh2exe|wsFiE7T|64uyANiAI?$>Xs`$eoc`N)Kun4f=AKDKUN)ISS3SiG z-DK)~UEbwlBeV_daS1iyhz^L(E1k_w#&u7Sv_UjQb9y%;|{dFpYwj@ih z>yXdmp?6FAZiAu<72#%t>6BX2`;W=|p0eV6pNnKMxL$37^tB=j>B@U)P*s3VoQaB# zvBY8cG&B=QSG>jV+03-2QPldCw+p+wsdv?&Wd3%6wg{yaX!1Yon%OX2lig=ff@U2)`#IqCEIpExe>jD=0U-0FvFr16C?wajz z#97t4&yv_u9?0i6{4cs`RGgHZb0z9Tg45uG+4qo1Q&@ojIdb#0Ui@2v!Sys|480&v z*Em%rA$GdBbO{)^=4gupxf3BGA#u3)e{-ihDp;~-7{}8|@Javc~b?gvm{M5bXf^zMo|G zH*5E(Y>va*Rvx#M6^Fw;5_;#!`R*zYtjzeFkbR?#(*+Q*Yl^!XZg{$0iW%5SsZ?|- z>HZ!e7^}xs5VPv&9jzHCd4}f;lFi)N|7$%eE7-3kqyU# zF(f3U>^pG_IbRyzwig2o$yYod3&g;{&%O(q=j20`Li7Q{C29y|OEtHnPlD&Q?yInEB zIL(d_9g)s;a?}%aHRFH@DtHVdkL~DSND<+U5Jz9S8Y84W0M>vi9gW6p^wA9iWsASv zo(K7k7uC8rT8hd*QJ_GMgfU?%(>1p}&uI$_=MG5BYevl{7+j51ki{9I-5cHkQV*To zfN(I!c8-|BYVsbkmmF3F$8Fp?=E;a)|7jWN!7w521e*?wJ4@~4YK5zK()`)gbGBQ) zol`&^C!`05nv~s$$X2FWN*A>hxe=Y+D*H=2!&9<#cI{<)a+Zg)0&k_XOg1~MaHN#; z!Q3e1g<-h7Xy8E%yI{fAm7HViA~~n{265P7D|@t+r^Nz zh(=!-@=4uSCYBeJOhM*>r8-CP%FPc}O7)(9P-!Mcp(e{uh;k1@yA3AMm?_d%Hsqq2 z{Dw$$q}d51akZ$E_&J9I6D-IhO`j5~xhNlRVK1c`V3`&>e6LFuE3YhO_Z8Z`n~(x+ zwiG}4Pv6JbgxRzY^V;Tg4-tgoO6Fwfl}9jSNg^X*iaKC?c|cd_Y*LTkjQcOU6Xb{` zFFCT*eiF18H}`b_s2ODmX*x&ar9vbFrqGhahuTMGktxKFqqL<(@ErS#-T7b|56(5o&MU=JW_NqsR~3qq=STXrSutDypcs0w!WmX4hTP zQwSz4-zKN_=)HJ>?P2m-91E~xS6$f8oS2U0BUM!v*ev@?xCLe*#1Xj>>7$4va1L^( z?WTa-OZt-j&EG&dNcYp%c&ImR87s!EJ5hYZxGlieR%9ZE3WoyOCo|xnWx5kLZ5o%tQ9DZy9%ZA3o!#e{cuaHk4cYBR?XWkgMdj zA~R!WV&zOYk?Os(Jk$2Gzxp#5yh=IA?CgARv~zR`Ja^=Z z%P3)%`fY)l(`YMg?@^Bs4V*vRg5p!5lH}U%75Gk0?CN`6RkDmszEm&VT6jWbnWoY5 zv*uQR*FDrr`I$IKyIXtk3y$3h>P#FQoMb7W>Ie~Csuj_*CGoUU>O$Y&BM+RO1=ENQ z8c&?%UshPk`k=+YoBy;Ftb@>w1XP|mcH%zcSh7fPgJI@wFXkd;;&To2!@uZV?*T~1 zlKzaK^=MzYyez^xSL6pP^ z*-{{AF{EAB=Fqis^et`ZhSQ7b23vUJHae8DWlXnv0I34iOO^BS_wM6I$?qZM9n;dn1gH0Va$EJ1hu)|GE_NcejA{5fQu1Yz_RIe*j%@H}b(92syzIE=2Q zvR~z0SL|dKrIX6fpm}u{i^J4G#mY( zSskqh;ORU|*ms0R0^;Ia^%`s&r&4k-GsV7bMXUHhX85p5L>cIR0I3AQcbN0!L&v$_ zS2V-fXDXYGr8IfC@**0MUyB51`=B$u%~C?}`(Ol~cuOr(LRZBfD;xPHS)_-Ca{mVnA#bF%q%=?R@vx53wyEr#L<6J`U>Lf z-RA882c3_yi;-6pu;qg&_7Jxv9yMEsLC{1r1T0z{K-T8ekr~EZ>14@&2NPAc5z8*_ zJl{!s&3FWw>cJi~9AT#U$#>gbWH#qdUs%9Ux$X64vtOFR?)bp~u+jt&oL{G0Zkulo zvoK5$A}VUezZhL!HtBI~^ zG4*|A{CFuawT~#xV>U?D*xklfR0t}Qx9oSb7*v{-Z5umZ zOA=l7LEl^~PaSny8NDLp%4;ge!ayvP)4F1sz~Pq8)LwJbHyubKY#p*v7u`q+;R;+#z>ntySiorxYUzbIlH_m!x@ci8y;w@uZmMKZVG=E% zmDD50wwmQ`DDC&G`%}A_E`oi-@BUmeLUTOT|0Cxe*8oW&cAvAeEE;~N^7 z6A`Qp4E0ZxGX#3gr*aW4a%B(si4i z0=wm)sC7Yddd!xZzV`;j7q<7)ZH7fgQ&%gZ5UcX*+2TbE8Zt3T7ZK6KG?B{4h^VE? z4G`;u!C!CJVYw|L-|3C2BtvPvLd11sY80mBuQgNA`d!oLGH+SQqSq2g`yhRbhvj%C zhtTnaX&Zen(&AJW1tvF|T8gy1{>PUB$O!)&ljK5yB zvWWNB6v2OqU0=)QXEODlc&7x9u`!8^=3r@m$YR`p%J+5{0d4%K8_yz9se z#z?VK9=!c3xOY~yLzssPZZei5P3fDw|OG3$_-j-EIE}qnGxegn#k1#mJ~^! z+IHwXP$n)_h)Z<}G1q9a0CWjNwt?1Od6<(IJ>8aMw2%IU*s?3Vc{$ z3axN8HN?Jnd^)UYr;Bn!N)zQD9l^Ap-=m9`(M>$sM)pT&UJ~tB26*)_8JLmnM!XGG zi*aIy?CL$P6{|zvyEzUP-URM@U5*%C+Q)u`S47lAF;a=Ruig$|Y|*{Pq|eM@Dk694 zYtkMwTOhtAF& zePCYHI-p9Y3_SKig|o{JQ3mjR@2ph!1PKNR7pD9-*M9-c^@L;x)-wn8wJ>$(bRAbAJuj8Hb|Icm?DGjn=u;#qg{#+ zn%*H@X=J1YD0`N7>q$Pm_eUnI`U}|>tA7G%W(;X{3jmaEJOBR4abbZq@z_f+e<-)Rj*3Q%MDh$pzT&;!O77?ZIxaZ%-kH4vqgz^v2 zrs3pg-GP~0S}dNI0(Sg#COg;J-k4f&u9KnQ{?gAE+6u5Plsxwjg^oX#`#63WWtPQT zW?X#n()B--g7W17(vN(5Hw^~6*|J*OraF`+(b=_Qt1+?|AE?MgKs?oaQ}L!L-~C=Y zye*x7q(>2E+Cg|JxTf@N4QxsIMyutL+9OR(yrAt&rOS<4wnLCbkS7JFt#iJc6A*o3 zA-aSTH0+hW@GQvM{|uoY!G<_7(E*P`p=jL?MRFl}>cEpCsA6z$M(pqqHFInFe7bfz zCRvk)!MEhjUrf~=4eQpqE+C5^POtY=E60JPyN~ZdbtKg|(S3lZ?bTO8`1X|*0W zi9SBQFMmP6Dg411_Y!KF70fr;YTbSR@U3YlV|OSXEKaf?@BpdJxX4W*^o`_jX2)D0 zmw8B>@fxxleqofxdHuN=`MqOjndg$D`$Pl4+VfSv-B#?T6u$4+vJ-O-j;>ClP; zZ;CwNNC{Dfd)EADm?1;|ww~OIHE_s)FaAP%_5I4V+>3f|A}q8_{ZI`z{n@!&204q1 zc`gdzi1|?)8+3ESZV+Q>di&|FaTH5f+G*SVy`K$By881o9gA~0yZTcu6{R^oh`ftU zk)m3V+vweSgz8c~?=k3TM%-T&$0og6p+_~G4uSjXS{%IaLcjUMHA_7X)V!hA>Ozd4 zhEoJ2N#Sccbt{rtHBQm&IZ^lUJvWK`%krtA>niA7J)S&Xs;wKxCjd<|g()%5yq84}c8)t>mW?OX6p&*hj% znK_nWccN8iAldPVZ;9nL71TvMshN=#Y_5DZEGz+pxw_AvI)5(W#&62FJs&qtfWLdi z&bX;X3;9)IuVD^E1x|5h?Ds$11d2|V{?`PaV#Q|#Tf6Ai% zy>PJ>1(Y~`!XOjL-GWrYc65!&vKq`7iy&oKFDj3Q_v`twpIj2k&!!_ta;{f4J_nu- zS^*=1ij}@UP`akZy^XcoSK~k&Mx2(*b2dvj3NFW9tXd_}^f ztIw0b*D`C%L`;c_6n@R#XN7{wXmSqnmxlmQFTodQuj^Jk!lmZ`B-GC~Uy*{=)L^rq z^_Gt4vgqcrjf7osK&IOaV~nKPSKcB0{)l33D$oG0=tqJHV}7Q~pRLpwcdzGl)>E>! zR@Js+eWj#OW-5$J7FiaS3DLR{QE8}F+cbd4 zu4%_l3AWVfFXbJyhTvMF4&FlI7GJQ}AgcqS@HFCaB}cq^%a!^tJq^{hNR^nRu>O%6 zD?4JE#vjH6idC}EW0QCYVNzs(iqjAaDxiLo&%pI=Bfz_G$t2-9@~4qCuOoPk>R=RI zZR$U{B9L7{zd3FYHEQ+nLKl#E0yg5pGkcjaF1DTN^kWIPkE-Oir}5yRf1&kBpe|Y)W4>?_PKBt(kp~q*n8g@#S*kE zI^%g(nqJg72`_&caZN-#xZ94Das(IJ>-kq|!@Bs?3M}jPWIuy|V>GaLzcdu$L~rfV zom{D#nPKvv%z0M4LoP9AEbQ9**I$@aarHr#*l*W{qw^CD<*6DqF}k8Ev{vWtJ11Io z*1u)iuL(O1yGv1T=k9=%rI@TXPR4z=CpDm0iJMsXz4z)<>i7%)cqd#kl|Jlltc7$Z zZyE5u^i`7DQW0<{(SPyU9535}{4@RosCu0bNLeb<{+=P}Bo+I8qO>F##p8(wzvy?? z1EZEnuNrZsxupf)U2loY7APrO9JG|xitUQWVn2NWAOHE&`JrsSU)FVp9@XVM?05rp zML&D+Wa*_k+D_}>Hn)WS=MQ_a+GOotEa3YqCAz2z~1cJTL~P&6PG0iwVg!kn(}YM9|r zG!7=Xb$;s>0>${FH^=Prm2tUk@m&_qdN*_i?PfaC188maSv*8@Q>b@g(6RKWt*j>^ zenrMSf!xXxVOf1M6uOu^`vXgK4AzalSGl049^sXc?x^+5QM)&B_%I!NP4Vn9GRHM+ZPvF~_b4^J%Bgj91ImXUVl*^u9$kacz_$UN^lqXptRU)-TDet7DBH*E)mR2(FF3ya9%%Zg_0tGmx7>_iA zR;T*DF4t*entxkHHe~ww&R>{K4nWS3dGS}Ff!Sul6~(dMBXLa+M!~yInP(m z6P1Cux^?EHNNfy;Asyz#0OviY40GZK=gvCod(p84Bc{3)#y3fOH_Ai#l6_Egan43V zf{|0f53m(=$WM~}x<2Ta)-kb^6^h9N5v(KG!yIY!t}9__#wTsQbZoIJ?42#u7xt|j z(m3ZB|6|J!+d9<=JZ}_~1{NzvaS>NC(y1RA-vm3L#JZRBA08r$kWsS<%@XhelrG&y zk}Q89@&!p{1(awdkRFLf@a?rGIg4y;!#uy-+he+G^&@YYZ>V8`0Ju_(t^nn0rkbNf ztJ2iw9^0aA3QF zx|J&R6LU3ifp&8_TP?!8vxyX?bk3#vOW?n6rZL#O;C2q)b}?sXIv}lt-3nu| ztSF)6Fq}N)&OapwtBWIFwlCnJ%@0qZra+p&0o%XcLOhvu8gjb2V3H`j-P&6P!$mRx zV)cJ6RA`x{IJ`Iuz9mT!a)u=jp1Y1w%d?-%*`OL|_Nuk(KJ0V~86b+t;$fM_ zrI8Eluub_X)U}NOttVl^RmWJejQ|KWK%wpBP*kxSk7F& zvEuBS9eNtx=JAB3L?u*)A#}CM+)6SAo%8FTr$3sEqzM}3p^E+;sa7|f9l{{rA5a*C z%)ZQaK^JOs{P9-Ar zDlQldpp};q;j;Z|Z$enE?em8+Li=HI%+~0U$i`(7rB@nFgt=g>Ne^2TR&YFNgh_2G zUY7?FYMXc68-!ENrrL=V9@sKsFQ76~L4OtFSuV74IGP0&iwV9*0C9;`hFeoZjJ?dUG=DXxQ_TuaB)m$SI;w{JNnY27JF`lcdSzq3ORN*koLoi!YZT1 z(3lH8izts)0@q-~Awmo9cpHNEFO*c0iwiXW4FlkVLxZpxx&o_Vni{kpSy@z&jD~<& zp3gFQMdvnu+|G6J^NcH>KP(K2Dm z!@vx%9R;Uh)dkSZY0YR4+;V!re3_zslEONxaXWy8G2(2_fSX?zAC*Xqwp^tCL)kk8 zi56{bmT8+OZQHhOowRM+wr$(CZQHh;S*I%EcK;o5AFB7;eA^NGW$bUQZ%pR_xyR~( z==EX<-PP7V-syUjAdForz!;^cSGQMDv}g!%egIOOm(WYN zu&Iub=*g0(U`JAdeIJsODcWkY+Xwx~Et|udc_OM{H`#kKCF?}?FRL)R(3@+!QiK`| zjHp7gIeyf!mm!_+Q%9on*+xHE_Z5G`Uv37AC#Up3Mjo^|z0PH?>65=6rq=)%AVP zJ;xtl^O8FEZyhk8fF|aO{C3d=WPG01#s505zyOMbVbGw1^$!5<-rp`^^Ep*>z}lu< zY=6H;JbA^ju%Vh}4n68GwPk1`#vC9O(wz_X3qVb(zME*_Y~9xgh8C`%ntxPElgp#B zMDc1|1F}v*ZbKy)YI*RHmaDUZ!P@*xYoz^}XV`s*iKIg5hLj3(Z9B6ZL&yMOAh`(( zRhV)!ors*9L~omW=&d!H;_GYeR}sDM=V_nK*|v1qq_Fnp*y%PTFe&6J2fFi69&ykl zkIGl@SgItS?Mdt82Y=3--SrfGkwU1-JC_z*;gY`W{V!X>!XLG`t7~4qVSXL5)e(C~-W07bP%#mGU|t zZ*n*NDOM5V)?5*p6}mlxT4vHx=7^aW2E@S%Fnh&7yNmpra^Kd>k1S^S3!uAi$AlPs zgS{=n*g7`v5wmiCRk)G~k0pA?8E!AvKAsUWB%8bq#8g0W4+P2zHvO$B(2IDc=o@`D zFu@kCQyRje22%^!{kNFVv(SfYX42`Zi|XHk#TB2|Uw}SIHfL-F!LIXXEQ>sv#) zZ8WR6vnPFi))-Y6B?Ubtu$YXjts136F^xqb7>sm{&}XK*PVFHyI1Nw;Dz0zHnn;GN-}#fq)^H{{SrF;rreA zQQ;m2fV*_Ew{sq{#M>Egrw$u5^xD_=BWt-@0=KBffWQFUEP}ao9zX`j0gd^+D>VU% z0a>9h`Qw1@v%~a9?CnPh6oSkGR{0zvznt*wgf`xuGrauh>`H2F6)JrRn zNBwjF;eRG>P~G0Z@YPF{}#!d`*Ly?-kH%fnlT|Ljj~BL2Ll)mHXfa0z(D3 zFr-E~sY|j$jO6=<_Y3m>ThmD5>IWRamkf#TfM!@fGo$V&Tf~STCt&o}AA^8+-~g~I z0^<{C^h^?fuQzDi2+Csjo3{3w@*dj3g6zXc_ty<1{?&3c42G8M3f$Vq$BK_MQ0SOH zt2hdOPp;S>=mlF~wk%2n#`M^%6*`a%zwYje0fC*g3Jw_ZK3~LEfR{?MibBt<06uTO zRZo@yrU5olHdMor802AHQ9m1>lndoq7a4a*@8Qq>ulY+U+7P6zH{8 zFKB<pdTCn7%=`W9RLY=4?7po9~wV>1;POhX|z4LBKtmw3O)#5i6F-?0(^k_ zc0J*@*pOT~Br_sE>F2<``cO471BOnV!Cc5R>VEn}U+G;iK_PTV+a%Ji79-L?G^jWM zU&tgP})f3zyTLZ)HbkrHhnGyK&p5r+^LOP_-)cQx}_vS-t$tv~&kk?>+>-&{1 z(T4yyu=#aO=RpQGz69}X!u+`X8rCmEhC9u^SGj3C4FC8@GbRfFI>EvK5@yQ4B+vPW z3a|D7K!^=4%7MK{DuLw1!H|N40#MVU0KPG;BL&D}VE~a=0sQ5}045*xl{3@k0T3}j z{@&vn#`^$deFLLo`O($d1sT~__1|UC#vuX#4TWww14p2hH150P4;JhV#p+)^Q+8YQ zwF{!EUsG0x27GFio(ae_vZrzCxqP{f(B?iQU}b`EDoohtOT*LBqghw(DTU)+H2V{( z`}_6qQ|@>1>ik8pv5_$jFLssz0D&x|ZpyrYK+nh+_C~Q_dtRftMcUT=n+1vv3lQKn zCs*)^u{cZ3dMB*mrB8ffWf@;r0r@a(oQ`wzLP6Ep_?PO~*xC_1Szyv}{#c=OX@;TR zQ@WdIu@D1&OU00Fq33B^wsV0|Xs6RL>`!=PQtFU8%juO+$4J(Ivk0mRm1WH4>5-?_ zk}a1?ykb5?8*6V1LwQxY)Qb0)^L8hUacx0F4!8c!d+C`b)>_l^7cE6Q<$KLa#qr!~ zJFHk@m}_-MYwPF9=p^%IWZi{D5%(!7b_8R9Xl#pjQg@~qy}H}Eg;e&N#iZq^mKNqw z#AYFcoZF<53Lgk(E#^H>mbb(AgHNeakB()MP=hA~5rq&;VSKNcNM)6!r3!u9S1f66 zi|O^^!$yIp?TMcCTt}GUJ)>K@P(BZERW-Qrs~Z9$N4;Z$kb_cw`X#iI3p)40d0KY% zisy*s{az6HXgh=cIQI2a(fgW!sa@b;i<6 z({eOf_o4-q2vWwHsf0+fWJG-IDBvqhc8C_$tAHc@0lRsE#`F?i#zEDEZA781P}N(8 zI*!kVusEGwd+(rXJP!ts$Z13zM=6k}+9lDaiaV~n4=#&_$s-(zl9lRrXAH&{oY_IR z=EfBlOk;Ja5Obm7&B5+{*^MFVOG}7G^zCkCI1l^xW`&t+*8M1oh1?u z+DYj^7fCvBc4k$w%PSa>xO9Tj!F$4XC-;4aemN!dG6cOCtG=erOGn5LOL^4-3mV8~ z_KU#^6coKfo)%q2scOl(?{M)wNYrOZu$pUO#gkE_?}_o|#4X z>zU;Oh37(K+;%JvyRfW6A7~nEUwbO_qw$sE<-82-!s}v5Q$q_Ff zcT!*i8F8Md+$tL#P_F&5GG+QUnP%#hGJP5DjfM&HyQ zpr=5`SHm(kVXG{N2i1-PYCn{olmuRbwT&!1b+r6B;->k*Cx?De}qsPAItYus0qI0QDh!*l5KA(sA`^_?!DyMAqd%BN&ySX6tS zlhkU+5>nzJgG%_pS@BFdx&$wc8^KY-k$2|iZP%CUA=E!`cP&V;BXYZuEsgT};Mw$M zWv`x#lmGm|kAkn7;C*%5=Q4C!Y6#!pF!R{NiG{1{MPFGHy!@*v^CrGO2lyer+IiWE zR9jdoVa|3xsh;zdx1l;Ge4C?LDlLas|Hk+&8x#At`z^B%dQheDM}cXw?t;#b&M~Jl zgrd0B1jg0_9U}d&gZIslbrW1i`KEM2hxPepQdjYUD|>md=U#)x?dR-LKsCgm^Mhv_ zmQlvfMkvswZqRyawgOeP1h49Z%tg~Z&#jM#+rDka3lbhz@GR9Q%D+}}_!(-cK?2p$Q@dUy2i_H~% z1o8p-L|W9!UAy5A`Hx891zONOF7QNh8QQU0rW&iYs|O1GrkU5p&-4yfkH_Lk7m>rf zCGK$Uy{?ii%noMvB^>jsZcCZN2HwxXKS?*DA$7rskd)$pwUm>E&j7f>XVLLkAKYVG zB-nhE8QV@`QMUsTMas3r#XxI9hX{i{uK_r3HI45Mt3~_0>rb(VtA3R6)gZ8~?1cL> z@j1_}H-;m3V1{K6k0tN+%l-u)5}O;Z;MIcq=4XQqqsK7b0u@|^%2FB}PWn%S77uv# z0B*#B8hJ_qMzs~lF(qo8rl*oad5^-)tL-U)(Y8F>?2Dphlvj%Uud zfgfYr$OS`HbgwizWxR?!sN^eDN!$A>d+mH;57r9|lH_Z}B||JJ3Z-REO)E5q#pCxc zD9e5J)fsi7*%2A^9p69Eg4Zf|RT;_O+K;YMg05&c7nwRA(yj9>h5Oc%ghiSyQ?aMh zs;uh6-6}mUtEK^YdgL?`E|=)Dto@TfP8xt^91_8Xfl z;9<@*aWccM5ymR46TCAK!IcP1PMm6$c*AFF3 zpQOxS7_*<)@Vgr6r2-r4zZ|^g)rm)@i9X3GLm!t&yeStM4ImsBlepDo{y@?Y*CL9l z<7zsmQE#HMZy4y?tAjMY^O9zOxxB=<>&e;}e|5;c+r$4B5mvIsx19eKVu|}RXI6po zQWBV$BzkZVlZxA3Uzr<1O;DnZ{V-M6XX$CMYZ+Z`zJkq0G;Ab=1TjK(G*hX4!b4s^ zk@rvOuCqa0j4Dcgy;ST{(ORS;lGT_I_c~H1>Ki#TJCZ>*M?b|D&sBx5f1B8YG*%J1vO2=(TkvQq7qThZ9B>Gdg2?0H`SDjqLnQ!0=`8|L zX3lqAvxa5sgo*x-_A3`>9`x8jMod5{v@}rau-1Wx)x=5Qx&oXS~BbUk+IRc$ak9K@~3yM zA~b&JqdQ&@PfmUqH3Ra>szKo5r9^spq9lBT9Izp@%5i8>&QWn}Z#6 zvgQwKAO7YxC&7}N;Bt#^rRGwLoF2@t4i@5m-Vm>13APy<=wWl7?}XsF7&=T{!Wlj1 z;TDI@3tg!*7t6=(?-1xQU5CQwdJo`qIZ3bONrubMJ9BE4_&mGOKM|s?Mbe^$a|3_p zM_haWzRfm8YFGx7d;KAL#OOXvA6B`54YSR9`+~LcimJNZ48ut z%<_G$AL{qzPZC9b^y%v_w2C7k@#+kM}LgS+*8GQn7F4DSuH&?tVX&>$@&imRN zkdm-gy1Tfyn2ITQdka-2)=B;(P1v;M#{QZmVkEfn8s(WC=UUdk^%=@~VxCbxwhs_h zZsVb3I*R93mYXCvGBDaBjg^#ylYw%+#}wTJ$;pT+%?V{OncHBycGZ=r>Xc|NBxJMa z+B1c)*nV-R{nhLG$oXEWzWoLNS)r5tZ@~oXe+wp9IoSUbMf`Ieu`#jyXZoMP1S1>! z|9vo_?8&X{j^*1(%OV&g{-(l3%V22or(lKJM4fTb$YjWXV=CRA(Qzh?BT9-~vV%Or zMv_h8Y^McEdrn1DQc@u)soA>Q_SHjehpDjTVg38@d&=}=>h^Q1=hVGtHkLb?F7GY8 z-i;kW$Q~+N@)0m!i#CW1YT@`kP5M{_GO(C}Q5^D287x|qFcCP3K8WbgM2OuQ_#Dr@ zFuoruJ9TVmvY*{OX}kbJ)PUpwE!h3JAM{}yy*A*YKXMVz9RsRtc&Ks^hG;Pa6c;i9 zU>c!zxHz)C44f;JNTLMtwHzpcydYAEDhUCA;4psBL@^L<2`N7ayFPhXC^A7hh#nGt z5RspFA0)D0UR^t3PhYwuF4>tjL-+m3gI5C2SPcXgYU1|qV zPF$Vk1wu@Q9j(|IS=5o^nN5Z(fKwb4l}cF26}b%@2-96UCL*PRfrB;RUQ*a06doH} z^Iw_dk}^BWk|ae`MjAeS!U7-*>2#2=Qlymrjgg`}gMw@~Lhb~9CLDwa;=VGd6%bGZ zB2kWcNy4zc``$3b=sp$>W1<^=&_Uf~i6G^THF;h>lraL3VSN-qWC$_r0VQR8*+@iP z^(%Hk1Ta7f$>I{Q*F=eTwRE|qZ-#^E0^@XzIFF|xd}fj4dlB0npKA~p6#_~vmIOpF zBw;YiwMu3tA$gjS=R;6%2M|F?@_e48f#I;5JY{a!RS(OD_j)03zfoy9e%DtJ%xnjF zdjQf@P1M81Rm8b?y|Wwu@Q?Io%sR9_xlOH<}GaTU=X3jn`{~dRfShV&TH8OWwUOGlCu&(2CGo+-Q=_M3!948 zla1u>!)*Zx3dd{3O!9?ETuuof(=!A}0tTK#;uj-HlFdT|)2B(tNstB6g)BB90of>{ z42C1*iT5}9fhZ*~8@@b%*yLl1dlq=M1G8}mQfHujWT5FbHvdIj;C=jZ%Ql|{(h4W( ze?(-kPxsvyv{NHAS6~(tZ~`LZ&HWknQt|uJ_|g^r{4_LAN8v=4?@D*#H>Jnuvm_ikgTzWwAS&&;$ooVvH% zMw3N7PO;}gSINyf`O+L~$&OdSU7eIg$*lt}Sm_724NYq}&^+9fnbC-GBuzyhf@!&`r z`5{e%3nW=c$|XF%+|@!zD~mGpq*_bAj{a%Pm5fg7Z(GZ+>UVJ~+Y4z=hg;LZSwVzk z&1%Gnzcy47Z+#bQ&2gH|PJ7s(i#~LA zK^Ixea?eY2_NLk&?!T4{jb?^+pIO79Cfkqqh}qEsirj=h+{~W3HB+U`P3u#ZY^egQxSrLXGhjPT z<6}2l6V;h>miR1l>!=gteWpxUsH79o2 zU2?(s+%Mu^@nKh9XEad2F?nHQ0-}b-)%QD$A>=l_VuF6<4@SjFlTRLr9%Vc13?1dQ zYXL|;TI`z#udbz*EXR$VX|dXuRZmT{;262K&kjgQ$7i!Ic%=HaDYl;rljLKMj1G-k z^PNwsHj-6W=Zc+n4Rl=(%og;8=?beEcasLq%;jRPli_V)RV{;l*MzjhgYFwX&L(f;ZB}L zOq-Dgyrap@?GT@<8q-$lY=_NzE18B%f9wtR>@!@;LmC05_}rhn+iXPF*XLrl(&aAZ zM6bukw-h9A(Wi-wi{pKzt`)HG&z8zuVd7qh&aWw_WbG1XGO6YesYu+gVH!58cG?`F zAMXc7A2}CJHM zjye+__@=h_dTvkoUmfaigY9j4kz}HFDm!|K3r@{eURUZi%jgGOVrj~^_q&aiOOI%( z2H#n#Ufng`l8%P3ZMxnos~oOBjD!BwneH4N&BYb%*1KPVLfh`MN=ykmvKrd0Y6_h> zT|K`skIJ3kRF|(JSj2SKR!+t|^?`ifur*2}}z=i#-7iT@$9@DdI39n5V zf7jI->UYXPWvCyzmb+aY3R*c^Wk^^a4n@BHrSGG@j)60GU+7&>OktK>tl+i|)IExt zjQYQqq371E-&d3cSER~`X`xSBFZ!T5Gvhe^7d0ZHTz}1=XL5I#SXRL)KdoIGqz?Fn ztW_v|oW)vwF>a?`Z^0>T+Z%1VfV+y|vwtNKaPI69{^Y4nLo&Z9UL?76&8h-wDA(WdJ(g-`H= zHGe^_)-crQp^IvfATr*$2!eHv$$$ier*5;cECk%x);8MguOCDEP z6)ahG1>tg zpXvVt*G%;PKj7M38F>{~RI4t&ATL0qx)pyc1y%z@3j%=^vV>GXm^+rB9nvr%ot+7g zFmmrNRlz}sfn_r88C7Xn>2R<67!JEsQEEi zNLX2Z*CZrKOlg!F3H>$9#CEL~q)eOv;Q2H~sgn7>*m^-w!jHt<2vQ{kL~vW;K*WP6 z(xqTUm}C0f@?%4d#`p#=)WOOpGeW_}qJjQJJNl5xw^ zK?Z8?f$ANOOh5DZS+xmMB99!44G40#fe`|f#pl^KTku%bf?&M-LKK%#%GuZeY$#w1 zA&wpw7Z%(wiKg$wB?p<#rrwSy*kKpMC86gs131k{N5snzvZbM}X!>&C!upfr@(40f zWM4R+!z7b}uBLSR5D;NBl~%Ff%pg9SB>K%9ZK9MhfuH`!(#Q2dkU$yHkBA2nH^TDB zvy`LbgXD`w7gtw0mKmrf->!2|rq^Cbh&lgEk+YZ0O;+JjpPIu_WDi_X|0g5EB6*Gbcow zAYP7ML#}JBSifv0+&HGT47+DL0pFw(9-d;Aq7;7ySmO5-ZCKiQC@v0oB+i|P<8eGB zL(VTgsbgK5d;(?G9Dr?+dQ<^DvX~#r!Sy9H_FWG{fX8(gMZV;B*OeQd2@oBLS7f9KzFxn%qc zh=g1p-M<;TrT&xP%f3lCNP;D%c4W{1Yyj6H=TV2(4{qsFVO${hwm;{6JiMMh1Y|a8bN_IWUUjOHu^YPDxXZ?0 z+Axwb+WEYcXc@?LO*YJTu=T20ZTX&uUrg??QfnsgtcX%`wi3mZsw)z==O{TWjDO2M z#b`VF-nA<#xA#0a=CQl7G4BaKN3!go?{56GuT#_)HMpPdHTmkI;$n%MUoF(e+ErnC zemg7!oLomLSU%QqDkO>tE!3dCoEC)sv8WWAy2M~>G-`Z#*`S&oS*m`qrrB}5Q@kcy zVUk>(&hg&3=NF3kN^JrJrIkY$4{av%e!f=i1f*$Fqmi%tf*y-tszO(i9-mIW3 zDlsWgZ@ARpZyjx6pB=j2d6q+4HM4G9sjf@qpk-m%rI%LcCUu}4*{;6|mdQL6Qp(r9 z{|M_w2Z^4|b$!Q-mY{J={3|EZfbG*dt$%QC%=edV>1OjsT|f6qe|$Mv?sCfc+fcq@ zmD4kH@F8Geot$v|OhY+nj0 zxFn61ocDDy?1M9qwM6PKUUBSpqwQ5?5HeF|``}`voHTZur1!kd$%wec_`GcsiAD?b%Wf8KvVG~* z=AsM?=vj~LLk_SMu^20L=LeaV=FQ3_H9mD%P6CW?Yq9l?h@k~rTCRw!mnwF39fe+k z*Qs)Ux*oocaSpn&E3LKt!)w+Y@Fddf_mXHCh%M2d9`*X=M29 z4rUSFYJG+Elzvn|4h9#sN{_SrgPTm;4(rg)9yxEq@=YX!rY=L%oHr{0MU^GCYt32T zj1Bij;>o&M)9an7gKRV*vFw}mzbo#P{S9?PJhWzc$;*eGGPa$SlYw50g)|lL=v%am zJQ(hlh~EuP;-@>)o=EV~-N;|tBLw}tG?uyW>l_xogO_oKfQ^c?rQfc@FL%3me#M)u z_nm<$u!mX>I!x=ojip$PH93?Mx%$SAzigv^qH6m2nzJ!sHRif0kM2X094HzO;0`vDHV_HufVaC@T}bxUPi# z?KZlOE+TQ;lEaE8I;QVtv(EXjbG?Vv1WNW=e>mFCWOBfLpPvY4mk+Ntyy1>$7u_CX zNbl4_Cri5D%R_2hbB9sAO*XXNnMyXeS?o-GUIAw>X;%8nS%ZObI0Mk+CdZQCe-qR6 zaOPcRMYi1*_pO7M-0Ioe@>^IQQ>$OehGE*3S{;Sna2RXPs|)h%M=*~IM&6BaJVmYN zSaUr!p6^18Us-G*ReI06I_-`F>}xkx^&Gp@tZOiZ3^qeax!uXfC^HlXyzOkgf=Jog zXyAMmER+yV(eZq*Z)~6q9n}lKrM%lC-9SdJ_U6FU|9*#-+zI7!k#M|IV<+=SFE!=`xG@tH4!AT=Yv-3dR6TqJ zThtit?M2hlYSVe|2=j@Qk%-LL#Y`8C6Y|Aa%}69HLU>i=8e4md$ciBE-@MQ7}@UJP471^FnH3aL_(Bcfl$gfz$KeMB!X^GXk%hINki^RhAfnosRNh5H>qfVQ=7>o=ZF6M~5!-x7Ygs$Se7A1uHVtBuV}fR$BF zPEOUzO^rLqWhE6b_Ow;@mY=MN zELt2t8xxzWn%_Ae>loWy!i_G7KLCX*{hKR%vTtYKulb|xMFRfV1PqSX*A(}!>*!S0 zmiF#<@w6}9CeYE5#pT7xM{NYZUvf3@9h>LsuBXZ6Q>=`Lpo)O1s8ps12BVuSH6?x+ zr)-stj?J&|Zw(niH4KIbuMYqj8JQmrV-cfliXy(47=)!-((Z%Arl6Apuz&aZPi}jQ zE08~74jC^@zlfRrXQ+gP3u z9P|X=GHO;8KY+h)I)U`h^+zva-tTE1Bfo4p$=G;5mL6o2puQ4ly8u|g&oAHUm9NKm z2N!!;1_#jjuiCIRjjn5AiBI?1D$=w~m3aMA-SI`T9wVW>uuaEQ_Z)g-wFY!Cj z4bW=B7lByr7(U>r{0IDA=way>Ult(q1aAmbZ}B7WpRs+z7ojL(2agy)Z}Ar%$?u8j z6aFA%i8(M!udcfCI~PH`;~ft~qV)~_!2gY!A1sw>;tPu5q2 z->St|U<_^T8FWDWxTsE%!E7l)+VSIdh{_7R7^(PV# z+-N@9)#H)&m?#2U-xs=QWh6Xn_m=O&8|p^FjXi+vVqxy6(4K=I2&um`{BY7N!Bbr@lGJ zEAN9!cz6od>e9Fmu6>aM-(_sI_m}ab7L3_9r?9tyghW7xmwF4KT)hqd;@;l=i5@<)ro-ng8g$NHr9KrV`a zh#{kTvf`0I+n6A)DllQ6pMguRH_el?o1`WfCXdO3oMdjr9dDT92l|sRe4(z&m;`;N z(JrO%o@H%6c@+pvf$M(2h^uL{_H&gX-h>A!GK#6hoxrUy0%B%6TirH2;j!r%uG32~ zV`egR+=Q>X_gHj%4~e!4Wp?39Ldz^i^sfOzd~z)Gbq2c4rCu(hc}3U1DtC(YT4|Kl zr6bREf@n8s`FS=P#)xM5>xeu!tblh=b~x#6 z`(lVEa|0KX(3Mx`gnN6j4=M`898BB2jN_jpaghf3;Fwu(Wmu9$*9yMr3KLMC zrHOIYi z9lD=mfJa$Kxl4uFD%F6PfY57t5W1>LnY3=|FjL|P{&KcAwbOkFmE9Shjb~gjz3TOk zJhRKL-&H|{`xuogI8QET!k;x|lmKN_i3b>aCtxMia|Iv=8gMM;PH34Iz2A&3&K&fn zdRQ%13Rp7=RMW=36;I!p4yHiKL5a?uyQyswd4T6N;b#NoqTct;V_v@0X*FngifdN( zw(f}i#HbB^$pCfULlaHH$^OB}eMBxmW0sIaT0^;_bFBA|mXcu+NEahhDY87trbWxPob1_^Q^;ll>VG-#vgCr`st+HxtR@ zy$wsxxm8bc*!C0M+5dO{ipDsv<->+hZ7i+QO#cEIbh~QDsah4Jr&a{RSONTamLR813pyq(yyHvfS&y)1J%s<=Wt3pb_Pm^gmTr|&!uSIy0Lf8s#tjfT4> za!k~pjY@+YWoDX^f4GL3e;{;rtK!#i(6|{MLnXGm$hB6k&H7IjO$=;^+3{SlaHUgA z^6tTCh;d%{G2wECIID3S>DM!~eGe{@AX&mgfLP7LyaN7e($R2^wno$2A;`jA3Iq$n z>_ce6-#5OPELc|*h#HdJ4F7RUpYEfNQq}gsf__jp=pYO>&uCwp!P_NVDwD# zx$X8BmqKEQlhq4OJ5`i2`_f1gFK8}I0YMR$!l`Yy(w|sv@i&nf;5NZi97=InXwS%k zG;qSM!x-GWH)XMOY$V=obdz3^G%VtQviBY?F5>#7Uu|T}d!Y#fqL-tGM}Ok4uto|m zYhXx6fZO(R>;sGSJ>`6GM+b{DUnJS#GsA3SpiB2WSq)0$~|rS&Gy%Nojtcw-%Nzh z)H|YpiY5UdIK##G+jB|5E&B!>x9KkVQ)UaAPOfvW-3xKJIH+(C24>udIod%c-X@{t z9F<#soGq@WiVr4exggYoUO&dD!7I(p;bsn&MD*7F#m#x!P~I0ig6<2?>LjcJ6nGgA z8;*42_PCOF#_&nHBG!mi4W`%yGwKpQe*3t@9)=`}3%l-BnOwPzbh{GE(q3650qf!{ z#6LqQRmct9&RNtaE&e|)v>#;We>{bb-&XwKrd=2-i-eTw6`oPFw^v`w%CSPCXYC>B zzDP*allGN9+x((=ZjkeRCjI=06^)IX&BKZ6@WcSuhLi8rrFfvbp=P-xh`7aJ?^$?S zgaSVkSINK$O3#(~nUdxWy?(ApInmv1#=wE?9d4NTPSCNSs^aV!#5f~@RAL@w+1qvN zaZsZc%e%l21A*EP0Dv&zOPES^wAPs`on^pXiFl}Lm!@Vifs$1U!NDSxxnA0nOUTVe z=h7|UhPZH@EeK94TyGl@`e-YnJ`0e5A~P}nFl@5HfyxQKd&$cf<$E|m(h&f=zgTiz zS4FsGU{kPQjR@JK6E+xw2tbM3r!h|5mTGroYKk&Ls)%k}o*lEXHc&r}8&g*4z6~+Y zJkC)m+USoXgQ6?RjUF^!xD-MabMHXad%L+)P$;Dt{1Nr96*4%V(Vt%b*thyPqVxNW z*)v?(xV)NV!uw7X)H!LcXIe0lT#(-gRb5^vTpfE3cE7F9X-(eE`@Gtu zSUXo}G{15@02%y4t&+$LdxoN9o~fzyVmNIrVN>wV5SqR`u+CZY1_yeQa#{E>nUG_l z<~T$!bsi)}vr5Z{@o0BrxCh|9Wj0|8McT)NG|u6)%hsQ}2~s0c(y|>EWK_ekL_J7F zHvbxCdoaa(`@VQtc<-^X#iOxrYvpnPq284jSe;_dc4H&QN0}B=)!~Lu6!=;B43+0_ zR=O5v5HMMGM6EY=AehoHC^0ymtys3MnOzM-0=uBCy67Xg_@lxK1`#%wsDX zsk0F(Lb(Vw1{`#F^IEHPB{nhRg3$SdG_zM_&A{h@{Gn~)dG$yo6b(uNcIr*QjRBgs z?#wd)jfWTLEhU~FJjaXvV*R(Bt$1VoOg{gIcRho6(e*SLr;8nZbenetg^=%sEIS?d z^Vv1cwfCU`*eDc(T;4Q#Yj)A_DkL**cQP0P&fFUJb}^{pQ!ySK#ZI??{wlU>Q!0B) zgpty2Z+^{-?uOqThRKE_Cwo2reZ5sNpPiF?C-t6z3uWWlqGB;9OLCpBQbz{Z;3O1%^u!f3h zK*o?2B71Cn?a#9weE9`b?<9+Ya{G`5eNrOyANiepOqiU^Cj2|jJ6Ro+$W$8#~4QsN*IA6b%3 zA;J;tQEvizuWNi^qFi{K9}%+5ONyP!qaoP_k-A<|3o**E@1sic*k^1rHo1sP`|MKC zeQ-$yaHu8Atx0}LF&@6v`lGYDTn?fcwR6m!8hh6MX!}V*o{hB#shaV52{*d2desx* zmCuB%?aM{!iG&7yVO`E@hTR~?d0^dTSpM*AsS5Xj0D6yLiUlNjopCKkZJH<>k0Rry zT3AgCC%AZ8zLur~b3T+8vpTg7KD}8k?@9{`rsCcZK4<+|@4*>F>5)xPaYiin^I+wxJ5T-Uz78f%0pjj?M}UHD}zK>nb>#wsX5`0jiW{8)wR)>Jj1 zp%KJ1#~8FvW0B$sFsWG}MtWTEuWt|;0H}MPx7g&TOek`B$sMkql5a(l`m@!$KKuA? zKjVCxLE>W7QR@0|O&~!w#MX0$LXke{lqIj0O0?x9uhjV2x%&$_Em&=uiU}iU0<6oUwciWneb8-!9{i&Rw4|bF4>+{a#uKbO|iLy#~tEBu-Nw4 z!P@eJptZrxPAkg%-!&JHr{AM=dnn{%#$SJ_AzeU9${TY^*o<1bxa*~+Ma_h7dM|;( zI|Iimz=zE1bSDnH4f!=sM{n9=j@m$ZsE4vSwTtWZ43r^sgkxGOG5D! zMp2U6h!V4=9bVQpoA_j$bA0^|31Ous1~V?b5Owx4uoF8ojKjv>s5Ip+#E(Dv8TgX- zz|>Z7)|?l^@YO2wPHVXO-Y$c#Q9)sOn-2rjAt|Qmx~}1hcUpC_O|_E36`=2?wmSU> zPWT{(b)(FcMqA5V(giDFJ7C8JqL5mqh6dmlLW0a=e8@Gs8wl-OSN^9J;*jy z;dLSCcUzSR@^<(g`%Q5xQTtBV8m~%IYl^O=A6B!_JjW49$D^I!jDEdBq1Wo~G`3vC zOTo%Rl0I2WSgQ%)#Fe0#GJ^7~!~Y%12h`=hH< zI_n*3GI+VP#;Bpm#ItCC>L9|}iwI|^fOcUi*izp2LaA{bJNgW8Z~X4niSJJwAiVjM`;QtO>KcEPNTvRswkz*ncI$LYAF zEUbc~a0^d$;MDwKteMJvn&2&EkEK9S*&Hu6SN`k`swDsV{+$ll3iiKS`g!`nM$z zpi{Vd(7+tTddtW-1KD!;ZH9g3!HqOWEO0dc+PRdbX)Jj5doU}s1zc@~{hEyaackN9 zv+n9zIaT`E+qeXbqmG0xL@{rE?Du!{MKKL`+1j}--RBM%J$>uzi8 zP+Z`!7l#F}B$TRRD&R_@^3-cBI5XA?8qWfpDjQJA`b6 zpQ$c%?Rzl{4+gaLkzqjVo@=BVkkvF75?Dp=h3v^G@nWQN6tCw>c?Zh!uRzK3{8Cva zI1)1Lj97iQSTp}24~TkONU)Y!$I*O35%JZIp_KJZTY%eE>#F?&^YJ7tJgPG(Ugo?W z_6THD@16GJdsuvLN#_p$V}J!4YRm}s5d{sAp~gg0E_9fAH>Ul@C9)G77heD*OOc-0 z{Zub>XRLL4Hk5F9a&hd$-iE*J`}}c9LzqDsB||94sz=F}v5X*rL!Q3s!~pBExh;rj zp)DdTYKXgFoJzVjmmHi_lreJu?AG!Y-xBI@9?QrdU)J4&r|p)SnZ z-)sja*N|#qTa8ldaPkX_bJ5cqj;Wu0R5mC;a-JD`3OA%|ge9%M@bIyUX>E~X8Si>p z;qBYC6oT8+qud+_pbCZth*FkPOL?K`7V+a3#e{}qVIK`k&cp`8Y8&-PA!|7 zH<*iks}|j(mNaT>Lm{Cin~Yo}=g^5g>}B37i&(ru?-B@~d%H&ezSIvcaVxTE>xEId zt9eaBue=cjicHzkpUr8S>5a^{PFDb*&H-Os@f6ahYI2Hb#e8Rx3xOCNckAfaTDnpI zRA%->MwLZM(v39FW^4EueoMaN&lzqxdWR1oi)

4;pat8e6p4PeGH3X6g}Sgoh8x z)-n8WRM%1IT3cSDN-1pHMk1M_qB>~?$_36SE|VH)r!!bvqMx=O zN-DsHPeY3NtshW;subXgRYC#@m(ptl-(9ien*JeWMj9C)g6DS)Dgu&Dz3UArY< zy|O`Nv=3NEzWpxZt3UMS9*m7D`o789znafUM|X;2#}o}0u2qPsO=B4$yMz3@H+7?2 zGHim@#|B*?Cxt@q=ZBJBZ}!+Jp@uqcvTA-G6-Or)Ea=k$HY+5tvM1;mRau9Cy(48D-mt>jih0r&RUfZR|(Fk<1@qbVOILdr;BW)6qxO|47xgz%Kim;QqyoxNtnoGx$ zA8y7pT!+W5r4r~E7$CW34u0d{U|@cHe}F2lHZ(NHcXQK)T9q|1-Q87S!#C60BP~3T-A>W1@(WyNK})e_ zTFi(OH}vXXrn#=d!Xa*tN=$px1wWiU-q0${92&C>hGRr5g}vXkPXPVIDlRagoo?pEyhlU*VGVGqBxr%TUm3 zPuG9oWVP56{IMmfHTjg>_8F9eJ!u6Ar+8w4WE_j?%`I7G{u6!5YSV(x1v)&;N>xyasn zHrSlJz(i(frNL40)$20b17v~K(-lz~rh^AJx{IPUuQwmy&IlGVZ(WHakbKcdH3lq} zicy8_8qJ&|U8E zhl|y=S|XS#eA%o~7m?n(b11+ePqLy6D5r90v%h&Nwb&7%jyY6zn+?;opJUmJd`d)1 zGyF2TO>P|F+m)F0rYs1jzgv3sIM#RLidIv!eCO6h>S%7T5;?$>8FxPZnoD6`2wa{M zj=JW$-qh3HD^s#D7PmR~diW_9#R@%)r|%lmv3|6teVxz&PhUrJ;gU6pMS?=q-a1#R z?h@y{Ub{0__IgRJ7c=K+yVioKaURinN4g52sw1L9m`Gbc&ogyT49DrkQ$BQ2n6OFk z=HSh60gc$<5Zq+f$vB+wSEE4-Ga->`@-vA+OYJ>oakS|z)(_#RHd&ShnT zudJokYK@8nsTW8@-=@GFn#r-E$it*!buuyZH&U|2CB3#dB%^qls0^R{qGAoq;JCWE zT^luivoY9kL=x!JITamW>WPH2&P1%0@%DHkp>5(-47V|q&CCa2CyzO0ulD1m`>|fn z0ptlI(sHKZ>pT4#pL2}Tp4o$6U~Q{-!6~{Q>}MV)Ytd3QT*+`{tKq`TqeRy59#Myx zYQ@hd?Mbw)a>CC-MOi02`XQ@=`EfsRJb4`<2(vGxy~?$Mx$u~`O42lGjrLJ{=6MG- zYK}^zFijkwzS=KO9fe`g%JIM}H(gcc4^1tECSXY*m(rAfE~Njn`r7YVv2z+-X@Jup zGBHK6g~w?Z{Q((Sb=8ka(KS+O2D6J%?xXFmS9coU`{)PF z_0yoNcX$@1xb73Z-D`)QjP!K-zESn0Ip6J*Y{#!9THoc_eDw8&<}rKP`aY&NQ5|CI zSkF&CsDGT`4C{_YU$?7_bw&K^g8G*`JtfU(4$jbd1e<Z5h zvg`D7?z^5{-1b`C{F$ltTy|f|N*<~>sDmTFuNG;Pv-Q2bzC}r=WDmELqM1Gd#(Kt* zxn$^cPvV5hyhSl}UXFTbVr0fLmE3sPAtYw}zV$Y=q|F{mFM;->xZcKIQ($G19f1{W|OCuE5Msi}D@Z@0$hEw0@l-7`AHU=h}PI*E8y) z=R8i1wFtu37+|C%#k4TpUtfZJpLy-?tHn1zW8ln7i$49F!^0EqEJnOU-T7K%|sTvaylV%(hu8y4z|a+AnImb^*{OO|Ny^oq`QnwmV`y|+50x08cs?!o{D*Zu8&e466t&O#W+MFA_akXCdrgepOm zPDFFg)V5OezM)xDs%O$pT&nSQ3!o-bM8?;FnL70a%Z6I{96EE|o+WPSa6bCC2j*F9 zNK#{QF24pfRd=-#ptH!Z`J3`G9?ouRlnN|9-X=2)YR3nt5a#L8!BoBC7Y;Eeq+~;K zJ{6NNc^5kPk-5~x_5K08)DA-3oNVS1T6dKvD_7?!exL3zq(5E~Ic~`4@uq2^!<#?E zF19zDQ$C6IwV9M_4SdY7nbTaN@tl+Td#9zqa6C6Hx}{24PRGC2~vGWcpUY>7pZiycG=vKQA@823YZU5B~}@ z@Q1Tv?tE7~@kp*Mr_7|p?H%g;cG*2?a?9pJ=$p#!EScV6mGcrDZ2PCE@c6-5Cyq$j zD>>K(nJqKep9n>VZ4NsVUne(nyku~N9MydIH@M3PHRMk}evL?VnwE2;xG7=I-6s+9 zuv_I^ZS$2ORqBi@tX*cM(6rCwu34 z|6Ubo=tvdz%!CGOK+*Y>T{Dzl+f3C86$6>U*Rsu>7aL= zIZoY)6{4$DuTlk!CDr2^x~I&|W51p<+vMycyp^YpwO0Hf+vt1zWOZrokGpdDY-!{` zFQgb>tM+K_da4#k|5VMW#vFy?Mt$sr?!NyHA@3N~C1F)Uu>{93nsmSfoaVbY2IIda zCL}aKB!2h?PKvUJ%Vw7U_1n*CFt`2wJaN4ux$DUCH5U2_$azW`T8`}I#^-bZ*@?ah=UCVg-R zDAmf{n534w!p4s}gcVqZ6Nyvleu_^(IE2SX8a+;L`QV18wBox%dqtHgqw$m$f@3d; zy9CF%vsZ1B<;xuGS#Ch`mS#v(ozp@<$NQpYMid3dK{72fz=8R$%wpBnX!`Xwkucs< zcT4vm#t7EqQ6-;cnD6RW&8bB%l7Z4gj8DHV*s_jAP7IT8zs#W#c#b|%)hkdcG zSQ5{yxlD?p^zKWOA7Ce=ba){Zl!euqjbuD*r-+zyHL4FjUepd<3*H5Mc<8m zXwB;@A`u>7_9mmbo@Mdmfyb~NQHJ530bLV+>7yS@Nn16G6XBR7EmP#9W-VyKPUFzu z0LcuAYwxkSGwAizLRXp{Kr5g{-C&x_e*9?+PhSgc9#wnaY~RzTE9@80b!FAyrLj6= z8jV#M{+28AQ`k{_A$P3Q@;d?gBOk)~`%*t|XUnf_b4oDaIiP^`kphl$i~&cCFJ_N+>WPokkcA*Nj{e7expTd>8Z z@u}BNFxHQ8TXA~hMHR(<&TmCg2SAF+&tjZv?0D)e!EUAdH-!$w!_oAYZqPbtq_PSA=FhnX)uw&7G0#++kZRG zG8vP%OTz)PDk_CllSa#obc2txBhauZ$$lB-yk3BDmIkg)OvKNzK>2K&#Yntps!&tB z|1{}G5kbPo^XRAH%dPk_5W{)tbKz~=*Radz7I(L1Nv2Xz=>D~FHrVXE5Cm*`qIl`U z6xZ)L6#4miJ}u?|`AcCxGQvV%BR-=%Ue>t#ehP{G1EK!yNw0&~Hrm*@>d|Q)UUu=@ zUDUBo)31+60{W6*u_i4Y82*)JwpOUz`W7zwRlXgHr&q7_%a5f!-M0-2v$vLZ6dsrJ z7;dhl@L~>}tJ8YF3~J(OkBMSHT}YgtIDB8=A`CP~N4`6K z@w7T=_ua6w>;f{m<*IqPbGe*n^0-NaBUggd#*ipkU#}*(d_JBO^%`x)YkLFh*8vrI z_N^irs!)a*nL79ZR964EmY3IqxKydt3kP_dF$-_Z00(WQi5+o>!Ygpk-eVD&Obn&W zBkBm2!kdEcId7TeP&!TO-Q+2L==kTi%40BKFQ4efG@jyY} zv!P<}ntGB#)30z0p8?~S&|DoDm8Vb2!8Q3e zUFh3~SMPbbUzj;T49JLZlXitd?)qcX4uY>dIf#e5&eV}K-74zs%gz%h$mFM**fvE+ z1)Rr&kL^Em-Kk)Qu;D9@@FVbC#<|{+3V>OFUaEE7KdF$1=aI@P@&=1;QzA54( z2py9lNd{$jYuCS$6%4617y2$G2Gm$%EXJj-mQ-?z^>Sbr(MOas4Zr+0S4gaj-z$f% z$gCQ;{F9kmA|k-}n#52nFnCa^*+%uSF=F`hT<0??E~qd5rrpO0JeyVwGpI};x3hE& zAcUrat}=yU>!7#t&Um0RuA;&j6m)2pS3vsZb^NKft4@IrTGl-m|7dkE6B#XyklH%|{Z~0HM*8@_Tu}OKS^Lz!M^&q`z3+PGJ z;DNVy`(iUk6JK1~4J(M@TJ%zwjPO;4391f{{ivgO;+8fTl={>ZHGMuv-3tlT2sBS( z0jtwIX8=O&dK4he$3?-RV@Jhilr}LBZM7ka&f^-EMqzT z8M-9~e`~m5U>b;Q^AlNoD0=b2loE^Xr#5q?jb8|vGoM~jek)Y=ioaLpXv?D#Ys#lP zJs55Mao=dyk_~d@d4AGxX2S9JRja8lsTjfHr$rW{tDtl6Z*^|i?{1U(s&jt9cgFiA zniBA-?C0D~#MSc1Jy!8sq7Iw@);8nY*eJ6VHwyAEBXjM+g@{UeoG_vR=>g&q%?%`M z%{nTcutwBGv*cfu8WIXIsPAbmOBSo|{d9`dbI+l45fODeDiYXI3&4pEJJ+IFiG!bG z&CYqFU9tp43?m=*6xm&uUpwmaN5Zc?e&K?RdKbYtLI~5S>5=HQ`N->D;N!Wd#V}@f z6Of(Apl<}U^e6kXsHCu(6a%jP2iJWsvr1_8R`D<1 zR0z&vY_gUFK31Zx@GMaXTCj4U_(wXviCNmobg_2uJldL!#oH`*nIFwE_~dsfL|!)K zMZ)IC^wNi2wmueOT-T@(8gj~i-HC83axGQ~EozR~4i}6Z^Xq9M4XQNqLBt|`-*wGZ z#x7|djGxy9b*n^2*tiNEnH9km+kqgH8(P5Rg>F1?Q^-fK)6tcviWBz=>=nnYRg;QI zJIFuv!;8fl<#}TC)P63!NW~5|l9hU?WO?BSX{HaT!#ZElh05ei@WMQY!gMKl1`r#! zkU-o+0!oSxtDa~9hKt0JF+*=~%)tCWAfFOO)ChCjg&v8mW>tT=yzVy>PqN#wcM-w# zDa54`%d3tKRWOI42vZYg_TG|mFv8ydBd&K)F zXwfv}!x9c<`-cU*C8341(Wm41n2F&6o)BOc@%hRY1L8-b{EEL>DIQS|gWxQa&Ea9+ zBX(^(SYYl`fQ@1dPy)m@N)#hzcjy>gDqNv)gdcPi(P2g$+`N9(wsBv4I7btfSK9rr1!UCXne%E#x0pf> z8D^g)Rc7Z+AxgM(AmtGmQSBA~$eX;`p{MFuW6{ISB`Q!8>mJ6ueQZwr@iKJN%d#Z=?T=J zU=`2XP8l1`l+osg8#KY**HmiwlYQn`x zrmh0#i?HSs$m?2~PBKBH`>PXmi?-Bx%1egt4^ni3(Ia$A@QT?tXr5X=+dR)h5<(xdo>u=M@eQ$GVqcGu*gp8BfIuz z3IA%}d+pEW+^`3d?w&%WB4NzN&gzsDXQ*795fHgew_H`R*%^C3?3?kTZDfT=0^a$U z9vj0&w|V@`0Z6~8Vu3;V+>TooCj%^5as-VLHRzGhXzge2TO-XHxR>KRvS#%aPwvDj z7PU$bX)(4RZ|#$)&CGX1t^=Sv)Z!yMiX9|A<2MS!f05PHjx@UiEPF}*Y{iswQpWji zrh3Ad%)=7g^FFvTsnbRshZ?O*Thn9Vln7sTP4zSzZ17wWZ=GknKukVTZA^{!JTs0I zb=%~w!pCnIn(4(G$5EA}7O8zw9n4(ruT%Tp-H9``K7_kQW@Tbj;W_)67E@Qk_sU_` zp;|}t;}i1D&Y|N=kAU6??VO4RZNsYWxaNx`MqRY3spknw=$p}n^Z`|IdXC0;Rjbp~n4NR{a=*|YAm_Vx zhvxot*HH7bbIh-?d<;5rKO`3L9E=mxaQ*6Tl{6qVu__CRWQEsmtU0mYG&qm9ybCPfc84-M0 z_h;dj^xQx{sTXr%Qh~Ck@ap;n2j(RlQ>=FO zc`UlEh}+YbaBwE1=OD$RCpLfhmTl}i@@vGI^5ST_3DjvL?XZ z=stdQL1!I-p|JujAfbak_Y@_Yq;7JNSPm8}v#S7z1c>3-cz(IjU|<}R3|n+bp_%+z zcf#F4H15}J%iBT2@Kuz-Bf2OyOIL63GF%cHH-SoQ@}!9e$&Ef`MJe55`(tEult1~T z%%QV=h~^}#3^I1J8{_D0s9s7NgL)fhK%PkL%EnHa1EJtya*SofQ9rYgmSy5GnwvwG zN1oRga<*NIor}6Sw0+NP6y1p2JTt4@uGlk_Jo;S&v+dVhfP`5LW2Kb(_FBsaIRqVa zCBM}}4MBt5u%g-dCy@(YkZ=Z$& zZEu?52k~U08QZ3ryMQUFUI}dmnJGS>8>3Q#%-IJwzh^dHhPK+2-fAEAVSHaY);3HJ z)UAHBfTrz2LYDSFn9E2MidFbV8>%`}xNJVQHPu0Q&u!WXJ*>>(_|r7+SXfptJ;DG) zSN4W92)ttXDS>_37!kH(_|5Un4zY=k!d5y|+bwheqr@4k5Ub18{kqYYGt}_}OiN@O z2mFMxspB$6FCvl@4 zyV=?kFr}K3BsL}&!aZXnh9BR&_SI;pZx#*gkjvAbUpPHC{GSIL?$X$lQZ)+K?)^aq0=f5jb{L>=!PC5JJZ3^Rym^F|jP7 zE3wD=7*a>69fi2@U2B(jZqs>bw7PjpJ|Xm=Soy6YjwU*m(0iCJpN}EQt?C?PJPhU+ z(RkA*J~iuPf|Z66N858ac(A8L^R01(rEdVO1u;O@^WMr)!pWSnbiHx?m|xQw-+Ad2 zf*ZxAa$vJVe#pnd-YIA-f1oVTO37**CqK^Q*t4%fLvGeAVdy_+`f^J39MerMFO0VI z66BYn=H^@P&Q*HSJfJ`e zx21qtnIB|JxJW2=>!)7=%7;fF>?AOT;8+I8GWjkFNvl9T{r}i!t;q44Pqq@kj(t~wJ zB0b=DZDHtIczg=#sKjs%s~I@N=(GLAYvzt4xGKbpEHrWLClTJr_st$TW>%Gk{<|!S zY{iB@$uZiz<5pAaqsEE4+FNsipjjdoV=4h(Os0N|n=_a9d*9Jmhw(!VHZLwsSlw4$ z_{Bd2xmVa1#3d!ce-8QOwXwSO{vk0lY1^;~@fmBrf*PIY{QawGR4SLY-HO)CQzsYa zBJ5f!)F@fz8g%i_ix|^HmW1u{(19Q3;`{{m9?LJEOM1UEl{w1g>gc+_3>X=kpJ%^C z;Q~*|U#N$bNjT#XdFp$1FkRMvJ=~Z$yoT|2_CMNi^!73D`e13F6w;l~V4TKe_^DSP zl;gERmt4oTL!M8r>fHic8+b}4uHny;^tRY8lqdoxBtSEVv2h)@V3|ns><^R=Gvc@0Ja)pvuwi~xj5ol3B-N!BeMndjARMFv!^JTQ+dr0;;nCG zgNN&;r>T1CoGjM>F=R)zcByE)94=$i7Jflou$CDd=_U0HE6@PlvJuFkzLN~XLL11r?GXoaTT%}};)JV;FY0(JCPpl0f%=aU*v zaunX`>aGKxFTu7BZ@c6B6vjZj2e6`+2z#;02`cMYG?RlXuhiF?5L9`y`{QO^6S*{m z0j!y&B&)a|f-umM^x_oJ=(j}|s&!<9jb}@EhbIz|j=Q;_v+cu6U_DIJ-rbQ=HN|CJ z-aEF9(SPBT4CQQB{+zxaqyN>9)B8f#bsez=lO;oKM4VWBI3dlDBW3aIoAWR2lX%Uy zKd%$$%-qP2op{gUFN)3<=z@bxm|8h!cD0@1iMu?@XyxNn%?rFxjAk%ZdyJdbx}p{$3ofmV*NQkivjd8%$anW)`pfUd@6LdVj3$&#OwOy5uV{|+r3hU1R_!) z$6Q`!cl7o0S?@{d(6-!s{y~TtKl6k9c6^&oVfZ!AC;}Hb&DGgji+Ms=$Em&xyKMp? zJJ({o)0?9^Y>GK?ZvD>!PjlEu%m{6F? zZV_`DnL%Q!X!^CxUnc8*q964PKj;jfFHb5bB;NI~0aPmlK?PXa0Bt3MNfCFNhkmcG z{Uys^Mrqt@csvXjGTdc!lp=8RC-Af9dS}q#+x?AA8-9n)vv>zSXEW0@h3axQTWJ?T z7eu`hisnXw;JZ6-3?2OB-yGb11D7Kuk>MuIF*)((UOuW_l&#wzV+Fr%T(;&}u|A|v z6cf0kj)L{Ba8_+)4SrqEY9bayiLz_8pXzoEqz=y~mF3`@C~tc-NlT=rgN#_<&j(}H zidw*;Gf@T9^ zGDPOKMZ0Q=vdO}RQOzz$ecFKuG}rp!gb|yxh}T9FdU@6C z@ZLY#Q9VDNJ*551G@bOL`we<^o4s#o#j|~(JEpMsEzkJ9m-L6qN?D5y?WvJ=MK4NW z6r}+_V0Q6oGfC06Z@%sUJ)@qVX8;=|=Wp1l>ThundI2`GSxv!~552dlsRtuboByo@ z8J<|)pyf}7mv%}_b@d@A)xyNWG|thMz^^bF2Xf)hi(fX2ob#$J{Uw}FL%KdxrKD7K zOU)PG-MCjc`crk=3k|rKI(~RUs!Y=@@(wQali?K^ zIbz!@aCjL`u4;boShT7_c65)1K3Kq?lqS_|sK4|MB@XYeP zXaW+={m-QiQ9q9UIt&4>a_Cd_+c(E@<7fH2rUTB?EyZ0u%*k)&b7%?-yBT7tPwNc7 zPXjO`tCpi*vpO0vPgXJEblR6*V;wZ5_QEidfXl3gy{4=ee;L%x8}iijeweCKZBOm4 z-(N%}4{PtVJ1Q{@_gBZnK|B<4Ba+fp9PH&#l@T+>NXD~1;T1yGsky@D5EHtTtA{VLAc68hRzM#H!?amh#Oz19ci;hINFk2DT% z3eo6y^4+0GPe&`ot*@c3*&u&E{yz3)M%`%&k9d@y=p%8wBU)AE`32}rA1M+*1yl4%ZMZ$qvZ3?#c?H1acE~*$DkcASO?n z1EE}#<$ev;RtuKwn%&-Krb&vq`^LWbQw@h|-w$HMA?>{RNn;(zzI|X_bP8?)KqgV- z>S;4F61~!L&yF=YQ_y~kl>`6qI8nMNi8Z~pra(S`9Xzfd%tBlZgHcQkM9S-43N^SO zCyS)i1t@hT;vvQ^^mL~e1p-cGhv@BbOb<^qwNFWHCc5+bFef=BRIGAEqDJba)}7+b zssfw^0}xuFE2kO+wFAxtg+tBUdI8iWez#3ixv1e5qj8HNt7QKe{!sKRc#|YY1SDkV z2;=Mq3^5YQ;up!&}F6|W-)R~sfmUmQ6v^ZFH^lYQ5nC6Dc; z1TTnGU>ON?Z^bX`A94sml+wFEQ~w0k4UGiXx|4%}=t6)@8u~V6&;`|;&A~)hQ*nWD z`p@?me`)RJvj!B9M0VTmX58^Bmc5gWT43=sX$iz*3m)$smCBJhR~FwyAJv+Pjqn6g zkkYZbDOr>zzYLMp@N686Cw{5*M8t@ZZ^!<}C;JQip=ysYZ3HZ96!CiFr z7jfwG<(;u792176-liZzsiriUbsO|p93G|ex!G44rZWts7KjoIJ3&5HJ|#JfJNS7e+|RVt#MihUr@$Ps`e7C!McGyCPzi<;0{Bi zJAw(=>;_%;*?#oKoErK;)k6!lHlv5}lAB{zM&Gd;{2}vU(MEz+r8d@?m7r5b@I5JN zGd@iXdEo0x&9(95;523VU6#XUedcTf7dz3&Ovy`FDv~@XuuoK%S|nNT&w5*sTW3Pc&c z+@QS+sS5LNMl!K*FiV#y%WoR~t<={e9w@V|<&>P%mzq2=6{Ya2GvvrQp5XOJ#u;^E zU+aFmrQ8lmc3bk4lVC)r*zjBWIY`SvOAJhN3qZ6zf z#*$Ew4t-Avi13IdFZ!M{tyZ))-j@7<6RKIw&Xt0rLr@2{8$9Gx6B()MJ<6?Cumk8B zF4kRd(Y>VF=I6Rcpgtt+61~#~ZVDCA)7TMG3@H9-{SMB#oH9yXe%+Neuv`1pZrU5Oh}`ui{Umnfn+ zKxs$>&q-X!2G z^EE_~vQvyN9RqxJTRQ*8_J?ah(0KhjXf(KbM}&u`t@J4&oZ<)>(%z1OWTQ0*FDTIX zxx$i^GG-kiMd0($xM?=l3i}OfAo?lKp^}3ahM6nWDw)}Af0daRpXmVHaQkS+{09$Xyt3i@Ncl}H>h`%-kSrl zxF4}h`ej&4-xp%)Lg}?VZerPJdt}{gMaUcI9ps1C1D?7<%v($d`0R>tJg{U4pZhPNpJunOt1V_nnv_1@gQbi; zN8yQ}wIGP?!(?UgFZTA@1wZau8c$m=MukFw0lmiNvf}*R~urHyJzN^31XMa-_&*W5Q9C8mKv*1NjYU#y^TtX!u~cS%3I8Bq`& zR&j~Lu#B@C0Pe;rivN4G_YV z@ptL<4Hn>l#T*HQ2z&4|j#SZ@R3FaE!r8an5Z`B<=F;jjGC1uoMtY1f%nwqzV96XX z8$L9u(_T{bILBsHNuh8Vpy7H;Ks|)6DF=tMjy^ zhV1Fy0fUJR_QU5Ddw_@}+69y{*@O}HT=G(TEfUl%c}f;K4MuR=$vO2{0O!`;NW-Jj zgEdqp_CqHLd;-zG)pN?@{jb>Sz<2~+G^%|8-Nx7EKTPG?xyTDHer3bZfIJ;Y7bO6+V5Pb`YLh4;@Z@R;GA(5zi`+uvBkD z&9_{CWW*9hd-f6``KeA(`RFE!e;XOp;&~g#w8XB>aJ*fy1HTRq%Ys}=B%QIwla3wT zcqRTEiF6THS-LWQN{12LB4w?eCXxdQ2syq^#Mw6GvsRZxLKosjt_AxLf2jzAYYmM{ z5^u*A*G-FWB*%uu!z2urnIU5ixta*of#s$!Cmsa|-$uWd zhY1%>$}D<4e-@tOlY&pk&*-8TUVy%22lH|}%+T9S8|`$bWP?;t2GWoJ0(zNHwQe;I z(@i|~M>IebIaWm~V}#9$Yfh8`tZm>62u&?&tqddA^Vl2>KlE4=|0GIJ6rP`2dR2>8 z5+c~3MXXS~tj1FHFL9C$k+&9DeZqy;on5==ve&$15@WFa)@zT_XQjeEJ924-L_L1O-LUIbl+G`3(0#X^jCu3M-Fy zMlA{-Y#+|c(FMqz{1tDwdI~EK{^aLw&n>uG{MD>jX#JdldGfpX4qp&&r5AEUPT!X# zP$WpPY@-B2? zk?sLv=;*iubupKcI9xx4aAaHMF!Pa?q=mWZg2b^=_AE8g`pw!vKK9*8(kQMaiSDUJ zMs0eK3oHr-hLPSWD2CsbCIL61>!uR<2G{c92Q?JtPA-hmU3n}v5@p?shNgbTwvOOn zaxlkKt7o?r%v8#H;#!tLEgiDN{PS%x@Ey`_T)d|#79hSpH46(5KlzNp^44k6z`o`v zD+ggJ{ywm=NF|DkN-q5e;GH_?x}76Y+X<$c;{1x{8T7P^rYZ#*n`#v&rA$bIMm6U#yq z&cs33U!J!QaXVga6l(MxiXIT~#P~nE_Ygv=6!e7gcI3(uk(z8Uo^cO(rW|}b6>FrW zS|Ij-Wtc_nO6I;Z3Uzr`%k1?mnSqvWPqx34*tyrvF#!oH88bxw9AMLN=nEKVI@VFg zaE2ACObU7ud~u~Kh5S$!QN18@M3D-45R<=gNYhz~8p=NT+cb5I{BTe4o~Bu7xI1Ky z6XN0DSX;tM@QYpqwyMp0oMJ(o%SDD)jQ8U&_keX-=^~)q9M62SYL$>XSf)@G;+yc~ z$pN;gCd?e-s_O{`Uh@j|=+}h%ExNE&SL^9U1{5}%r!Wy+C~Lg7aGn&N!q zp{wHQim*%~B8Vb-W#g2(8)D~Dl+CwNg!-yqZzoH?m|o@-aXK?=_I0!%LnZK}_7O34 zZa^HxS9xclA1GDUk-J+34Dm44YbwxFuB_-+drjes>UYqY@$c3nt$R1e_qR{6hDwzn zYnaLTX%st+S{S3zV=I&KgD%X$b*MFeUTOt4whq4Sbp16LyJyjFjf5g&a6ghhkZ{Fz z4a9|iolC5))B@t!Rfu17u(b`|PkRqOPMEa*2$)}9AexEp`4qA;9`y~RbGM9Ze2^KG zofb>@zzzD6s=pgJLMNvM{-HmpWH)r~`vQA-PDeI{VpE4RmQvN{J|hek(PQzdLaxqw z7rJw=i%TUG0Ah3=ldS_n*6qhw~Tx#L0c+?AUGAIJ|JF5Y2VQfp z|4fpsYDYdLn7TMGwc2`DP@Fi?QlHtoX>S!e7jhq_eFD!iKvKAk!~a~Pz3jC%irrQ$ zral$W*%sU|IQKVTP@8|K@9w+)&ORHmF$2kLpR@Qu$Ft$gxTRS7*@J{;Glt2V*>i`C z*;dx)qf=;B)nNEn`81GFe|q?q7qS4%m_$T+sOhu$s>uXL^Kf;TTJ}$W2Wq(MT?k%RgVw@3qln10lFCNmwIG zYUO58@#n8BlO0!3S?EYI{jra;Zw?BY8y{G-P8p&Q7NA97*E17Tesl0!_xQ^d-Mp4( zQM#-LYme3C!VN~v<0HYzbVbRys*Sw`a3wL2C1`3hGcz+YGcz-NGy9vFsm;u6x4X^k zHZwCbGc&u-b91vdce@e05fxFADpgV`Dk`(`W$MkpoTf>d?gw0$O9FKfLB*0TXwbOr z3Qkm#0G>Ovwd$-0EEYrwRE zcuKlE7j=Ay;4ked(cSRX8ovoQ%)um57oi>mL%xll4!;KjbrURi{_-(wjd;fyjN5?w zo>|`#gt|~B)%KEh#Ga$NBQAimZI)FRNQhjQBPko%-X+&Kq=5cuh1+273rq&{4lZZw za;!AKXNOXZXS>cd2LUO?;-X9S#u+LFC2?_MFCbi64!4z%#5*7VyY=}njug#>Zirxv zwJ=ilRqwe;2(+V#^e}RORWaim}9&PJxd3g2zY!NQ$6}%q5nI zPOhv8eOQG-!AQMu91@`?U^F1I+BVRnk|204UJLMqd-#xEWMvC z#rIj$RkPDSlauOs#=N*8Ym@PuTAX!8i?~jk{=(pl)}|Y)t6*w%DB1da35^oJ88LuP z8p??enIEEBCU&T?0W_-D} zoIp%$rR$wpBC*wo(l9(Bwx)W?n`z5@HknP*Yv>X0j$zFG4e+TW^*ymS>Aj5tNI%h8+3_t+(g znnP_sqQxA43U{#TaU@v8{)oCtC6I>c-5zcvFr9bc>__0~TS$7n^AB>11pg;+h69PF z?;qUhF>~Ue%Q4pe*ytIfPZQS)lbea$f+g;+|CWmT?gHyxXl@oQuyqNI&^7z^28mAZ zh#g6V=tkg8I%t>Gdk9W<(>1Jr?8k&*_>Nrh@`R>D_w43CCeWCt+_z{6~# zbzoYXp92wKI%`3&Qtek@CA$0J2Opqgw6M^=-bRdT$y=)0DAZe3Ab&b&B>ZnDf&XA6 zkhQ8Of1colDVHCCqi_UQZCJYEu(gG^4^~u(1AAh;nUF~ss{I#=&RLO%#boL-Iy)U@ zuLjTe?mg~!(k62A{fu{-fk~ykL7i3C^G#CCdcVj%)?`{Yc#M!|!!9Tig)IiX(O79; zLt&=^8I+Lp>qS61BH@Z5%wHkFjhH%>litv;_sxcE7(U~BvBu~dzPh*b_ig%y%{^CG z3ixFV>TF0eDKMwsXJ7=q@KtQabxsKWZM9r4GAGnpxz$;kap3{;4$}Om1Lnz$(ONxV z?nATvbJfZ{Oec#l=qkxsf+B)`NM%uiXb~IIA7UZxEn#$;q!kLtF{cgeEnGHrlc&K6 z|Ewx#)yxXW(#gOa60_K!89zl9?Q98mUOLY&wJz+`$ls(@cEmx!SX3ZTb1u^mGb9$n zW6goqturv#8)us?CZ+PrfQLCrp4y2z+mb@Z zEs+TMdRElE7C}K!81F30XF^R#JbTC$9aFO>*ofdi^*>?tnj$-g50$fZD&sI~)3r{x z!PB}TGc_1KSSRBcqFR10Y5L7XkubKjuE&Lklk!D)=q010(4B~nIbY;4C~JH zSA|=zH#h5rAf>L)EAzU(Y^d1`k^?Xo74E%<#WXj(js1J$UfVHj*d+&s>&4FHbe!{m z=S~K?`lPort${|ksfd-f*c?DDdAS&-2~AGw;>s>cS(;eHIKBdSsL_bE{ybDs+Sq>u zw^c{Cy&E6aPQx(zFw%!wEWf{LRK%02xjuSDxg7dh(2&@N9~9sthHUho26$;(OE?Cj zbJgc(E#*gpE@0Zw zD@Td$?M0xkbe-tb2>mRwQ!;d!?NDa)m9cKlB*r}RyWWLM&GJ%rfO}lTAWLd9yd#5|+QQ01-vU zSW^f5bm6xraWEM~Of?_z3DPJ!!SKXVG?w1H&+KW4*nV3!MOj79Xu;mZuT}LdASC+! zCx1HALe72%wPJ6Ie`V0wUA25H7Pi|v^<@98fXnNFjU=wBLPr?UP-+WCB&fKRw$hRd zS-hxrlx)2i2h>?liHHrU-yt^9!~g?3=ZsvY^8K^keeb(!*&Z?WcC82mN2M( z9;BO>GUo&7FlSY!O*Q>&;>+GC%111*S{Yas262y}sU$dRJpXwj7?0e#RiV3+FzUTK zqDZIm-OkICXPw}podPP#13VLMj6#bVjszx%)L7YledZSCkFvJWYeQ1e6C0fJ^#bgj zlg0R)%t2MLMWDRiP0kWpIYh@HxjJ}$p!@>k0FJMi)SfFX?YIrc`SN9Aw*kR|)PrO` z9a*0z)mP|sf{gA}^WO#SVctkyV$qcUER)`Ovm>inrkupu*6=pGR20<(M=lg8=*fFR zTl81p0trRDU@D>=>#3T!SWP^18Kd(@AiK+?9#IhCs+mOqaS&hv_%G4%}R zzy2gx6lnV!kQ}4W5FJW6fUDrH=wM4X<_x*QW9-L6nW7Qhp1})Q^klMP^~CXpC|r1j z%O5=z2@6d6L}ltxj4$}OQ1g;rp&8S}Sb!QFP0AIn1KC{D{X+Cl@7O^XQr`6WV>}RiLtj3xEA30N$nDUE0%t z`F6GeLY5OVcyuU8!iYJ^Is?rB2^^n|PGSu{jsOLqt6Nc>lOv43XMRj)xhOgkQL^dPx9_s~bVKB(KfC&-;9+(~NHz`>uu1CB;K*B>^e6rvZs;aUX6rxbq)j** zDk^FqeilioXrNM|*;THx(vVRVzZI(yTBJEwxXq@!uHKwjmvJpZk225c@G}S6coo%+N4fh+BiJiaj|tjK>VH;H@+xS=VFr)^MgK-) zBt~8s{%gSnpnZ^gCmY=6Z0NYZlYvR@=DEv}hzaaViS`psS#Ad(kG!Da&Fe3=Ayn70 zwt3d(FEj?~ULr9jC#^J5|7)M$i*S^3R&|r2zG=`{c({mM9hez81sZ0?WvRj%16QH$U`cJ|zN* zk{UZHBiq+w89PeUm#|{D=Pt;sr~Pcv57Gn*^*^4`HlEkpnu6+Q#5m6Vlf+!8S93#= zyYXpLm-7?MbCgsWmvLo2aEJfL(fod5rMP+(+f&j`!fzCGD*CrM`6LQrHpvM$cgF|*ta3Llg;p>3C|7vYZ5z$>vi$F&&q1Ul7yeB!~s1qyt-d2hteg=vUclF1% zx4CsHUTqN9Rnq)>QKwrtgu>?8Rr8Gi=dW?i;KJpRH@RRQZm${vu8IA^68 zMCu%kRkG9!LxR3&-Ra)_F}TNa5mae76$<())mu*#?3}8E$6W4KCDMA0XCQC=QG|`7 zK(q5v!2Qr!6>ETJIXaL#5nF~;TINkA%x>-jlxsA}OBv!qR=%vwvsrxb*HFXQj8&?u zTyc9x58glXwJe9LFhNx0Sak+N+{$B&`&AeHsWY{Xf#`qEjKYH#9*Nov z;ZdA227=w-@y^fNMGk*eZU`)BGU52q-bj4tR^UpTlL#YW6eM(rFnMz1jy&O%({0Pg z#XU<$bGA@MgJ7mpj`UMZwsn-d01J4`SDx4}dhPwDE|w6r$pyPxp!i1=nO*WtF97rP z6*JP@#Vf3_CjN>LqLe#a5`hr%A?czRGZ}Z`CBswC-<+%SgmQqzs|5^2sIVHuFhNuK zA6)-$WwSc?!>DajU3bLB^2|7IMKFxM&cr!0PF;D{uir5eD8a0 zF0bjIdh_!y{44K5rcUhrIN7P(>c`cm`1tjt)#1LPED2JO>MQ5Aa?KGTh8k!pj`L-# z8@%T~tEf5BCoNZnc3I{Iz;Wxw0Tx?UP%DEL*0x*M&!URH}8X5Zm9{8X-ysnJ!PKd%k65cOsvHPuL5Xj}ZOaN0306P?_oQxF$m z=4sbpWl^LZcHvA>G)JxSM2@S?L3S2pqLfQm>uP#MG-I=G?(kX2ygRIS0>)*f4K%C^Nv154?`=SVBE>Cs(TGRlJ1;?K`J= zJ#U?dp)%?{bil8#dgoLQTceTo0}4{O`ySL`v~Z0?A|dm0=m+;!mHIfWf64PB^7@T~ z(dtNV^FQ-edNvRO<`bmNjG;=V3f66Dk2Cg+;Eo%6D(`sJ5v;*P4r;yFSI6gM|CZoj75<_6o9!m{2#Hp)kJlZkpO z&qwH4uuFSckoHA~%eM0z%9zBg8Hk5?%DtUaF_%@3Ii8~NTn5m%H3 z$A^cDL9QfyjC*5qCBZwBx$a2Q3$tGu4C(YRq16T>7oDL!MOrGnetS6tU=b<#wa&xs zezOo5nJcM9G6CrkSefm_HHu0UhHM&X2n}{s;uIcEqCbgAN@gfRdghK_5EkMCQYT73 z-|Itcm?|{rh8ZfEvaJoWiXOkS&MLoX8a1JKA(NI2^a8R}2}v(RnL4X8`Z89oBcvM- z4~|l*sv)TJhy`#T6lr!-G586?J486^DW0#XXt7iaFB8O`1Z+g3yG@tx_oOQS`YK3HO_ROf%;#eGezZ5osHem{`@g{(LFqWl z*gCSwjatnm?sT@|3i|9c^)0&OrV5T5acc9E!B62(9`&vYi*c~<3|?DguSx#YxG}@4 zX$E;6?VZJAUfAmVeGG}$ayg8waRR&=1W+PlJDy0&4ZhWEq6wo0vi-Rt1k|D6%nCo# zk(#X>(}?I!6e84_2nei{p1p$}ZIY$o8$Q7xSt>i;Jbv}xF2cV4X=A{YJ2v)8y427V z9A;Rg`O!2r{z!{BxYUgvs4oDLX=KF~CFgETFG=|K#f}>Ss?D?g{>AQ;3q?2~3wv1n zf@#K-ar(D+FUY|Mpb}xxDKCY^sA&%?d1J|_CWw@mSvdd-mV{bbPHI zlvgas{R0vcx{4jWK0BgLwL^4rq*M$>n<`SKdP)ulM;IKl%jBLk$f{i}%)D^Q7ut>= z?$pRQ zHk2=@M#^}w1(IEA__g^HxI-Is2u&m!BorEB40B6Lgc>6JGrp(QTN;`!4SKj#5j|SA zlvE7$2Trn7oAx>wMsN%WmS)c#wl>HEMF@yA7*d)fr4C3Qv zF>G*=_>WqApX+t& z_z(&^Ay&fgXf9}slb{6cAQWF}NYMdha?lec0x-MfQV>#F{7R+XbTUM1ZvNROX66UfiKDWf- z%!ysAMaKBd74TX6V7!3NhA)8=8&llG90X1l?HlWH1sH$V!P&=3A*b>H=)+ z9T$qqMc->|Z?ah{;lLYbWEd$D>Qg;nA3Pk?)xgK0G(B`F>RmDD@2}@abVX|Uz(UhU zF3*3FH2TX407AigNV%^wpliQr&`{hcg!>avv>e_C(uo{#tbKB+mI}tkdy!b<@@LvXs9u4O6 zdle+ezugG(NNs_@8|7$|*Z9E0o1VV6elT&G*V*$}u>r)^B9*E;iDc>8FRLAS#R(CWrKHR;`Uq0+@9xpeQyHg}7gGqJw>9dn3 z$wdNdgYfZuZ~0hj!>og;x0TTsE$>d-!>Z#`A|eG(ndx)7SVI%hSi@JR%W5DM+GgE6 zSy-Zy-LheJ?ajPgSboyiMoSSXd2kSHk^EMQGgHN1%+1C;J389{cfR()WR8}~olle5 z?Af&8^^twa1$$W{0YER1Rd*S4$tI}_ZoE9Rw5lq2tuyycm92Y& zJh^vI<=Ij7REd8#wrFJD|AzSeM1;zK`c5!38F7%*TU(BI&_?Ozz2DiyL;FTA!&c+@ zoVM9W>5i_BJk31lGx5J8(KT_~Y^MtaiaM$WFS=A>n+33raIQ$p*${Zb%sI!4mr{0% z^>UYhRYxICZ_Sr+jxF^@>|0vUL{F4ixZK3Hs-IiTcz2oxY{VK8KBuU06ZeRogMj$vjRi6T*7aR%@BXKOCz z=~m7>kA5;2G#iy4LyW(i_M_y=y03Y+rY?9ER4uec=KWGy%-^HN-0n6!a*_#Bl^ z*lp@;*txy8rkZvF^+T-baTy? zFS+zj2+1rvk8$Ju5?ssta8HYF&af3NpoKX4QykU!{rXSc7o8$@f;I5lTib1oD(p9> zI3+Cw8Q=(dH!V9R1^)_0kNtg!PjguoZ`x&W(~auXe6s_h`z{?t&3p&OoK5*kz7*Lu zCDkj~E9J(ZXSE*7oYRfH0UbT?l6{(|*8GCB@Dyf`wzWShtdo?kw>_?YXw0V*J(!~d z@P*iCxlRB&T#aAhF-f{^v=5Qqs_fUa6+Lz$Q-51=vdyEqCOefzQSViQwx~x-pBCty4)Pl39?9MCU zTda3dx)z0lS}LQx2J^17$lBJyW1oiuro5Tw3*EBJPrKTBQ=O;qUpHd)mLqOLaU2bq zczfA8H}vLj8J!P@c|dinH$JCE|)D9X26Fp0BjFyx8DZvm2ZjRliAk;06nCkB89XvaFd| zITYs?-~AzRTYdfFUpA$yrh50DE`;LHOTSV)U6i~#9MIkDug5>$s)H~!=VILIPk!5obm1ExZZ zv?=~j{yRK(8@!$!NlHty>y|Qf)DYYn=In)l}a!Z8C9OrFzg zbS&=9@3IjAG;nU)~uSX%G>9@?2?k=YK!xfM#WeXCZsF*;cM~i5Ig#Z&+YdaG- zdvg@nKR^_?bNXt`YS=KkoYXK@3FT(x>c}~wgr9$4G2nDgg>$7VW^?=6rOK-lWZy)8 zbnxj2i{skEsbS2~8tpjGd5^q{@>(OpUqMyChPb2n=>ZLX{DhfKjUEd1sDb?hS6V^` z|5wwq3n6R{aap0zjsi{&9Qf9=>)NAS)sHMjpVN=5QV_KyiN#J%aHAt){d3)9WJt^) z(3r*H{1%Y}t1vnh3Nc?mP*_|Fcm|4#PRmSAHjAXr8T|(18KL|OwwMJvzA&o*7AZA1 zPPmp?L=y`14s;A6Vp=0l{pAIsYO(7O z?b%QXc?9B$8^jaRlt@@wEsP`B>lvbW7mOh+@JV_~@aLS6#!BK-x>#6AhE9P5j0M$# zz>;I0i9iS}(Xeq;CFoRfIvtY}NYA*vVO_9lZc*v5@B3n;T??w|hD8C_<*cT{}3Md8mh9eMx0*k;7^E8bT zeBxxqS%N@@T!aPi>UQBb6gmM=;$Eg1?gV`D4sZ^NR%$S#UvHp!L$84zRXz>trMzT4QaZiZDqvpO8*4oyq-?@n zjRndcjo9gHOh}x@po$vfNuAbVvPNQ7o*Bmk%Yse^Cfu6)4Z(O)M8|vuERukdfw{7u zJi3<6OI&B36)AD8+GIT^@m*0;KM(%$IUAm!nd_+*mf6%6r)V`gmY|2aDdAjr4ZQx{}39>~wsL{2Hc12XyO2pg+$?c|goUhgY2Y(9Rm=GSN82>3FB1_@qv+ zF@sotq4W0P_m?}H-OyueN3Oo&*!WS5_D^^VpXLk>I}u(j)$qTum310Ay}T`ni)FO2 zb?ddYwnuOl?__?Z1}0IP$rPNHQr1?C0kR9uNahKTrDBZL#f%y~xCW+S6N9I0Tj#CpC(1|QUEoh@O^THt>cZauRvE|}w=P5K*j4ZC*30^i-0Dvz ziEBI3k}A_E^{v6r2x=H{db^Yb3QR73b_F<20J7&qKp!&3#ljpa^u3)Jl*B~ClS#K% z-o{U;VmWTutkZEk=Oy-hi3iYgD|#ldv_Yg{^qA8*5>jIXUCB3m3{;cPZf4_Foc7N0 zhI4+@B#)+{`Kn|Mr%l9|jVkjhNX+%9vEPld_Q~&dE(!PWzDiOjT1k>E26HEuNn}9l z$}Rqp#ez<_sM(~ruS-TSA;uny$`SSRib{U{lMFV?`{rpy?RulC);3h$k> z>09C-Ab*s*erg+8GmA)WDN`SIY1SRj^m4cOp?GcjEvJ^P>kc@sQzSv8viNXq!&vUBYLXtQ(qG1x8^U zUPut6i@c+-icI$X7?b-{sMniW9dg^W^y2|@-jLZ+O$W;lhpV#5)_})i$wLKe>j?gg zIUZii!1c<5s|>&D;Z7MIt2BM31*El_?9A=YZ`vjvt>9>)sut?p*ixt+jUZr6Nk5LT z!3u2w5QE_N2x>L#k+@GoT%5ElF6JFHNbt+-7^&r;xJ;6J@0@`|65?&TVhnR3b)s-D zs>~MI(oBUWab~1`XC(9N0%2X`8n)4#Nmf-F9K$ZRD-46Xx7&KSl?&26Eu5I_}Tw~|co!>Eg( zYA9A{dUxxuS)qxHv~uGba>0J!XeM%)u++4ExN8BplPP4lwC^3d1#uZ&kyL$JL`p@z zx;H1*_0c1Fat*fXRLG*TuC68)theSQt<7n23keUx%nlWL3J~J2nz<2)yZ4@<+WQBD zM1qb~!2uPy3}f9|5tTe7s~aaj@;q9r+xZSq*AEym5Nr zL4ssFH8{7b3Hp`NbDRs{rUo^ewDp;G8C|~lPj>zwXU^Eoc9}x;<0QLyw!}}_ss~jZ zE0jAHFPYR2i@S^)vGFE*SsA0My4NPZ1_s_k)6c~xuX{Rz-B-(W7gomQXxg_j1Ubt5 zJk|7^^)P&GG&R*sa&3*NRM%A;bMLJ9pn0OHIE)B+CD0Yxw3HlfiLvfr5=YxMAWt00 zI$TPX%;2h~us&sWVdMoVLI|oVJbDEYt_DB9*lA%1#t*~QA0?%yD(E&j`*tA}YZ`6} zt3eakvQb*&KLuZD6%Sm$iC`$QCv2?e=Yq$t+^Sx^!FUsyQ)e53lutQM^T!~}82S9% z<>Jj--;*$q1iPJ=lExkAhBKFJaIg;-RVnet97neglQ1zFb6Sb6F~&BHEM7OBFK}{o z;C_~-Si_=Scu^$5lK?=U{eQQLn5&jCs@7!7$_y0j45o?vNfdrp>`q*}Z^Et0I7fkIeV90>Or^Ehp4ZpT+%z#2$;EH})vb;KZ zK10SFWL|^gnhc2GN|{T}ju#)dVU(3HwOJy(tbCkv^5%r-v;lwm+woIg!erIz&XJN= zQHm#L_7=mD_Zz|2=pc@xl$*!0nNQam-JIiiGo^=Gv6{zESkxIkC-@FG>l!~N8-z1! zT#mDtccyGY=WW#n^;YVtg6m2{q`xzV+F~bB{K4A_Dg|Z_IpaXudUW@OxP^Xh4XX;_ zypux3Lg&3!4UJ)GAVVegUNpfw?m`0t6RzcLbEHBxExt2}oQ*y8s=ExP#v?`H6|jcY zlVhv`aX!02CdJ%;ZwD5e(v_Tbq(a5izb;5Vy!}nrN^#==ZZGW|J?K@(k^+t^hetm& zyQI?0Htm{O8NWO*0Ai}N<~JhPg*MjyWh+RBE7yAya2BN5Q25UYiCv4Bt8p=)14@@L zAIPSBq3W>wgIgq(vo9sL6H_%7F~NYl{ZRB@Zu2Jd7NjesUe-vcVFrbRWf1bWN(834eSpfD;Mz!4-1`uPMZrl zYbXc8o^_7BPs6_b(1x~dqOnP!h1E$t`Z>RmL@WDjG>}NO%wP4??BKz}5O<4mLW{pf zRP7f_W6dQv_tf)TSHQ11oA851rR<(0JvD7rd~Wr1C7WfxqC*An#_>)n9>zqs`dJ_& z&s7JUSTz-!Y37fF6D}I-*+GuzFWs!EO?W1QVza{!o{4adbopn6Qf|D$?T#DBw+ls8 z;>EU-;*`dqEgdd#L%h-?BqC}aMD)y~WmN_O;hDNngO#;>BE=G)EHa(n*WMSiX%*cL z);z@{T642dM2irD9WdC=6HvH$@fM`lPeaiG`(l#)aip3#T6K7Zx!Ak}g~VBv>^Etw z%i#9zlWCd>#`#%<^Z?!K190pzg!ag`skXGe))+RsN_%I=c(YbS-udbbJ0#B}H3iQS z`P#cz*Q}Tg2THl;`QGSV@Dz5odmk8z-@6jo8@?;0UL~}S2tt(;0xp@B*8Wzh?jU|+ zA4yU3x+wsH?98Ldv*?z3P!1VaOv=-j;O5GV7QxE`9!#^K9!v2?GzV7!t7c#-BB)g` z{P=80DUeo6?KC()c5P}_`rcyiBhR^GF)tXiyPUzL*o*AvXpseV$sZLqsuS6Rjr{dV z^LanBWC0a8w)8z8nD4LUo!YToMx1v1nM?qGYbNJD@Ze4E8rLH(F7Kbe=i7Yp{_r&* z46lFdu!LD=k0s!9C(&(?Y>>+3;hhVvn-P)9?+rCLCc$^E-Ae_DvF3)Q6nm3g1!>@Y1G*5>PycZU*zwNALI6K~-d%nygjbL%FC?ME zf4)^YUBhHVK+qW~ILt)ks4BfyiTrtU0}3lZDxjPCAep)q)8M^^_Bd(IY$vJei{=pn z``ZBB*Y)M-p#MGJkSL#hxihhnP{1Ml`24YiX`f&9p^2N-0$#;~onq1dVu2ufcrS_@ zb$Y<{R`o-z19Ny!>sB%+QtcL#`h&)MPyRuo8)pvdH0%64j~lD06&dSA=p}efS)aYd z%$cToU~zL&LjV>0RtJ8t2|G%Xy;yf667LOcA_J1x;q1162P+_|E+cZZma7d_jinQJ zxv3SyXkc06*CeBIxA*JbZMG%buSGv)pWrNiUn>Ez%XVd^#4$T4Hw-S(Pi*3DeUR3$ zW^h^(S!u+9fSN$*K+xYc3Xo@tAAbnBVzk=~YQ1mwKF^I{{_Jl(oSkyL!A=&Cc;_vo zxAbp$3fN^JAunh=$SI61FW_^ixrzMYO)w$o1Z7>8kRN9>wcI!S=PE$3gf-vL3wTeo z7SGRAznr{^7)RW;cwSRVH+6yQIkjau4jG-S_2B9j7NXZ5V7mxDjxfp-3V5X5&rJCw z%+Lk2XlC9Jw3M{~ZL8OyPtb1BAI2BR{-w2QO@D4kMELY?J+68oL3CKhrYX2PMNwAm zGUEN&8od<~^t-*$J)`;~7 zMd~MQFCsFKYubt1D{^tg>=&|QtEH8Sj|tlOmmMs94qD^$$>KfG4f()ljAX(v$!&j;KSlNhFDwHqe;6C$jOQ`6zLN!j9+6QvlT9?@*wIf6W1 z0&-OQ9>|;w$<+GbyEn$6`$y2ODVxK5#%uXu>F643c4d?1+z~o~y%*>Lc>Kqx9$rowpzfn#y7S>LK|}Yy;eN*So|rP@ z%3lvW8bf&au@UMsRZ~nV8a|Ew@dzLkGJrVp;rWr4miSd2Ss(p`?hnCuOQ^3sD}mwR zQX3J_<#)ULSQNeO+>oLx{&mk7)U3E4uk6hBOoM~H^m%mT(2=2-ox8DdJKBIm!)GTs z>Cr$XzeA=lP?o==H8-iz=-k`!x-TzLx*#~J7EMMIO<$engC@&KnZF#Kf^NkHnf96$ zA2NRO^agJ|vg2Y+(n+~V0>GTlYm%AHiflCaZRT+|K@~ydAMkcozCS-ajK}rM7$~U` zRwM+0p+S|*HD&iOl#Q7P=Rb0Kr7vKxV0Hxf2c-xKefB?lF_Zl*Q>GvIeg>>Hs4eqWO}Q3q%fdsqXlh^@Z8o-Qr`3t}e|Gg}i&01cC} zi=(-l8Nh|w$=u>wjDdxbotcq^lZHtS;N{`yV(v;!1N)zMvUYS3|Av5xsl|C&nOT{c zd6-$4xma0QIOv#JshF9mzU34g&Hr~I>MkZuP5|?7JlVv~6#&bmtS+I=DCuTrXKLc$ z@QpK5t6N#S5`Q26IUjmrO@NE*_t3;_j4bRd96YR?91N_Coc}e?Z~5-cU^b5bas#t+bN&xE z@b|#n|EF%?&I}#bHEHDG?N{{s3_IHOtV9cPjZ#%W4;11G;!50p6+bjf;1X|S+`v@% zHX+Oev2b0>uMX)mp1%QaO3x<+kW`R`DJW{(p}3U0Wnq@Gb~Y%>6mb-hfIaAodhEgp z^?n$Rw%C5?!?VCo(F}(0bVN62tm(|z%tr-!{Cbp6u6+>UJ<=1H+$$s# z07@1f76km^S~#d?dZ0-~tCefGCm4-=&Nb@1N^;_wo1LkvDNbB)FWgvSI2k+JLRjnr zW|2#Tc3Qt)O8WpRB_sx99X@zF1V!3-s*hxAAtYQ%FS6o!e*~t!5CeJ8{)!T%2*0fe zSB4^h*38J_fXJdB2s;l_Mq&uzIZRCSITN$vlX&35ib%p(R43nHjEOcM9wzUHb39@Y zD;=`R9}yOTD8^%8s1a*e_O}5_J5&PFA60^IA|KNghmJ{vHHiM9p64UU8cc?si7oc zjX4mmHI#0T8M_a*p z>Z{a|G4QrHAH{`#vjD5;S+Vr*3A=0AAN+H}M`NH9>s-t#rCur+Ia08~L zSM~BocfZkJ0PsIW4qj=F$^aLYFqMb(xae|fYiLt{DycRY)%nj?nSrl{e^v8 zT7hc~w`NGGffVHM{P1M;P|aqwtx~Lkcs(|u6e5vl48K8)z&h|mJb5>@p{7vu!%wND z3nJM!kHnYfNn@AE=XlcrsUOMvGWT}!@&#}Al(*yl?M6#VtN*>-Nl zRiyVEePT;Y#5(LY$IAx62+4hm&(P~%tkbWHxf|Uco!wf*sk|tD18-xJAWUAv#n8&l_53TShr~dHGB*e>m*vwchq}(^R`C3SYf@ zo=z^xbT=)9+k+h6WsL1A>%4vWYwFyfd!~+=K4+a^xV@u?WHLBptZvINBq0<;{{E2u z2etD$x!ik8X&e%owmD!4ry%4G-BDb#Z2A1D&=GF1x<-p^bM|I^hciX;5-#=JGJJdu zS^Q+Y+zz!p#=UF9Kc?O39r?`^UFaU2yB=<`Anv>=^xZj z{b$9WZjJ-(P4NkJvbrO4LU%?6cA%&IuWCMw*DG2<&Ie7#4wpnG!;7w3t}NF8{XKer zI+bBw7Y8@jGY?j(Q2D?hQ{*;}ix!%W&gj{2nBab2%!1`iw!>aG(V<~hH-$zZSEs&Y zzrdMO^7B=9;&Vl%6_hA(6O2T#mi%mR%zq=r;WgAx5qP zDLz_OtkYtHoi$kwV8KK)NK>p?dd$58$-%1HDR{=7yx~z9F^Of;U1h11G4o+@`O_%6 z>o+8UC58e~b6DdQ8#CjdcZuO<9Y=BJ9B;Db%*Rz(b`@K-zFXc?$WTFdp=#B{CEwzt zj;TPKJ6<|k+A~)bjov30`(G#3b-!Jd|DC6BJ+@c#@{PNu{!_LLK7_HP;Jo;Nk7xhb95TJ2QoGINpdH_g_5UGTHbs^a=X2th%|0Lp?C9VBmX?u?U zrtSYzTWPBTEMS>r9Lxcp#CpucES#JMuuN*!-hlro!ZK+Q>#-2C5wm z`aJ;r=T9u`Oq>iny+sYp%p~j#ouLx!7c}^ZDpuw>FQid5%d&RI?W}b)lv^f=5@jAv z6mpSZur>7A!!T6N#*xbzNQ@I>6S4}p=6=kRAc9($=3Yh4dzL0T-t21mppH!OAa#`o z-$zlZ+60+4yyiQCL$pWL7Sn?1Y(G7sRxFRT@nEIe1b4BTqQ7Toz^oIdwm$~7P)~gz z_@VxT3vZ6KMB&ERlEy^%fg;@SW=_&9pX>@CZOdjkBwk%w(uB|Sqr*qWgqX%C)qNbzFG2bd zc(oAYPG}Cj_7bkErB~EjzOFXZf(yAbLS2B(P+?(lw3_Jn}f;I6^g{t-& zr-^gzD&@nn0ZTb~zpOrh-oN)gOXk@yh`RX%wCYsdybJfnqvz#P$wCxs)Tb7tGg}hh zgQya`z%pkownQu%Gf$66QXjjky5(#o|A7hzN<{q6Z=@^G#0BX2ZA8Mda