From e94cb59ba580336c9a96aee43d5cf5e5b2cce656 Mon Sep 17 00:00:00 2001 From: mhjensen Date: Tue, 6 Oct 2020 22:32:36 +0200 Subject: [PATCH] updating project 2 --- .../2020/Project2/html/._Project2-bs000.html | 435 ++++++++++++++++++ .../2020/Project2/html/Project2-bs.html | 435 ++++++++++++++++++ doc/Projects/2020/Project2/html/Project2.html | 376 +++++++++++++++ .../2020/Project2/ipynb/Project2.ipynb | 259 +++++++++++ .../Project2/ipynb/ipynb-Project2-src.tar.gz | Bin 0 -> 192 bytes doc/Projects/2020/Project2/pdf/Project2.p.tex | 417 +++++++++++++++++ doc/Projects/2020/Project2/pdf/Project2.pdf | Bin 0 -> 236422 bytes doc/Projects/2020/Project2/pdf/Project2.tex | 389 ++++++++++++++++ .../Projects/2020/Project2/Project2.do.txt | 42 +- 9 files changed, 2319 insertions(+), 34 deletions(-) create mode 100644 doc/Projects/2020/Project2/html/._Project2-bs000.html create mode 100644 doc/Projects/2020/Project2/html/Project2-bs.html create mode 100644 doc/Projects/2020/Project2/html/Project2.html create mode 100644 doc/Projects/2020/Project2/ipynb/Project2.ipynb create mode 100644 doc/Projects/2020/Project2/ipynb/ipynb-Project2-src.tar.gz create mode 100644 doc/Projects/2020/Project2/pdf/Project2.p.tex create mode 100644 doc/Projects/2020/Project2/pdf/Project2.pdf create mode 100644 doc/Projects/2020/Project2/pdf/Project2.tex diff --git a/doc/Projects/2020/Project2/html/._Project2-bs000.html b/doc/Projects/2020/Project2/html/._Project2-bs000.html new file mode 100644 index 000000000..4ccada940 --- /dev/null +++ b/doc/Projects/2020/Project2/html/._Project2-bs000.html @@ -0,0 +1,435 @@ + + + + + + + + +Project 2 on Machine Learning, deadline November 7 + + + + + + + + + + + + + + + + + + + + + + + +
+
+ + Project 2 on Machine Learning, deadline November 7 +
+ +
+ +
+
+ + +
+ +

 

 

 

+ + + + + + +
+

Project 2 on Machine Learning, deadline November 7

+ +

+ + +

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

+ + +

Department of Physics, University of Oslo, Norway
+
+

+

Oct 6, 2020

+
+

+

+ +

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

+ +

+The main aim of this project is to study both classification and +regression problems by developing our own feed-forward neural network (FFNN) code. We can reuse the regression algorithms studied +in project 1. We will include logistic regresion for classification +problems and write our own FFNNcode for studying +both regression and classification problems. The codes developed in +project 1, including bootstrap and/or cross-validation as well as the +computation of the mean-squared error and/or the \( R2 \) or the accuracy score (classification problems) functions can +also be utilized in the present analysis. + +

+The data sets that we propose here are (the default sets) + +

+ +However, if you would like to study other data sets, feel free to +propose other sets. What we listed here are mere suggestions from our +side. If you opt for another data set, consider using a set which +has been studied in the scientific literature. This makes it easier +for you to compare and analyze your results. It is also an essential +elements of the scientific discussion. + +

+In particular, when developing your own Logistic Regression code for classification problems, the so-called Wisconsin Cancer data (which is a binary problem, benign or malignant tumors) may be studied. You find more information about this at the Scikit-Learn site or at the University of California at Irvine. The lecture slides on dimensionality reduction have several code examples on this data set. + +

Part a): Write your Logistic Regression code, first step

+ +

+If you opt for the credit card data, your first task is to familiarize yourself with the data set and the scientific article. +We recommend also that you study the code example in the Logistic Regression. + +

+Write the part of the code which reads in the data and sets up the relevant data sets. + +

Part b): Write your Logistic Regression code, second step

+ +

+Now you should write your Logistic Regression code with the aim to +reproduce the Logistic Regression analysis of the scientific +article. + +

+Define your cost function and the design matrix before you start writing your code. + +

+In order to find the optimal parameters of your logistic regressor you +should include a gradient descent solver, as discussed in the +gradient descent +lectures. +Since we don't have so many data points, you may just code the +standard gradient descent with a given learning rate, or even attempt +to use the Newton-Raphson method. Alternatively, it may be useful for +the next part on neural networks to implement a stochastic gradient +descent with and without mini-batches. Stochastic gradient with +mini-batches may give the best results. You could finally compare your +code with the output from scikit-learn's toolbox for optimization +methods applied to logistic regression. + +

+To measure the performance of our classification problem we use the +so-called accuracy score. The accuracy is as you would expect just +the number of correctly guessed targets \( t_i \) divided by the total +number of targets. A perfect classifier will have an accuracy score of +\( 1 \). + +$$ +\text{Accuracy} = \frac{\sum_{i=1}^n I(t_i = y_i)}{n} , +$$ + +

+where \( I \) is the indicator function, \( 1 \) if \( t_i = y_i \) and \( 0 \) +otherwise if we have a binary classifcation problem. Here \( t_i \) +represents the target and \( y_i \) the outputs of your Logistic +Regression code. + +

+You can compare your own results with those obtained using +scikit-learn. + +

+As stated in the introduction, it can also be useful to study other datasets. In particular, when developing your own Logistic Regression code for classification problems, the so-called \ +Wisconsin Cancer data (which is a binary problem, benign or malignant tumors) may be studied. You find more \ +information about this at the Scikit-Learn site or at the "University of California at Irvine":"https://archive.ics.uci.e\ +du/ml/datasets/breast+cancer+wisconsin+(original)". The lecture slides on dimensionality reduction have sev\ +eral code examples on this data set. + +

Part c): Writing your own Neural Network code

+ +

+Your aim now, and this is the central part of this project, is to +write to your own Feed Forward Neural Network code implementing the back +propagation algorithm discussed in the lecture +slides. We +start with the Logistic Regression case and the data set discussed in parts a) and b) but train +now the network to find the optimal weights and biases. You are free +to use the codes in the above lecture slides as starting points. + +

+Discuss again your choice of cost function. + +

+Train your network and compare the results with those from your Logistic Regression code. +You should test your results against a similar code using Scikit-Learn (see the examples in the above lecture notes) or tensorflow/keras. + +

+Comment your results and give a critical discussion of the results +obtained with the Logistic Regression code and your own Neural Network +code. Make an analysis of the regularization parameters and the learning rates employed to find the optimal accurary score. + +

+A useful reference on the back progagation algorithm is Nielsen's +book. It is an excellent +read. + +

Part d): Regression analysis using neural networks

+ +

+Here we will change the cost function for our neural network code +developed in part c) in order to perform a regression (fitting a +function or some data set) analysis. As stated above, our default data +sets could be either the Franke function or the terrain data from +project 1. + +

+Compare you results from the neural network regression analysis (with a discussion of learning rates and regularization parameters) +with those you obtained in project 1. Alternatively, if you opt for other data sets, you would need to run your standard ordinary least squares, Ridge and Lasso calculations using your codes from project 1. + +

+Again, we strongly recommend that you compare your own neural Network code and results against a similar code using Scikit-Learn (see the examples in the above lecture notes) or tensorflow/keras. + +

Part e) Critical evaluation of the various algorithms

+ +

+After all these glorious calculations, you should now summarize the +various algorithms and come with a critical evaluation of their pros +and cons. Which algorithm works best for the regression case and which +is best for the classification case. These codes can also be part of +your final project 3, but now applied to other data sets. + +

Background literature

+ +
    +
  1. The text of Michael Nielsen is highly recommended, see Nielsen's book. It is an excellent read.
    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. +
  3. Mehta et al, arXiv 1803.08823, A high-bias, low-variance introduction to Machine Learning for physicists, ArXiv:1803.08823.
    4. +
+ +

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/Project2/html/Project2-bs.html b/doc/Projects/2020/Project2/html/Project2-bs.html new file mode 100644 index 000000000..4ccada940 --- /dev/null +++ b/doc/Projects/2020/Project2/html/Project2-bs.html @@ -0,0 +1,435 @@ + + + + + + + + +Project 2 on Machine Learning, deadline November 7 + + + + + + + + + + + + + + + + + + + + + + + +
+
+ + Project 2 on Machine Learning, deadline November 7 +
+ +
+ +
+
+ + +
+ +

 

 

 

+ + + + + + +
+

Project 2 on Machine Learning, deadline November 7

+ +

+ + +

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

+ + +

Department of Physics, University of Oslo, Norway
+
+

+

Oct 6, 2020

+
+

+

+ +

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

+ +

+The main aim of this project is to study both classification and +regression problems by developing our own feed-forward neural network (FFNN) code. We can reuse the regression algorithms studied +in project 1. We will include logistic regresion for classification +problems and write our own FFNNcode for studying +both regression and classification problems. The codes developed in +project 1, including bootstrap and/or cross-validation as well as the +computation of the mean-squared error and/or the \( R2 \) or the accuracy score (classification problems) functions can +also be utilized in the present analysis. + +

+The data sets that we propose here are (the default sets) + +

+ +However, if you would like to study other data sets, feel free to +propose other sets. What we listed here are mere suggestions from our +side. If you opt for another data set, consider using a set which +has been studied in the scientific literature. This makes it easier +for you to compare and analyze your results. It is also an essential +elements of the scientific discussion. + +

+In particular, when developing your own Logistic Regression code for classification problems, the so-called Wisconsin Cancer data (which is a binary problem, benign or malignant tumors) may be studied. You find more information about this at the Scikit-Learn site or at the University of California at Irvine. The lecture slides on dimensionality reduction have several code examples on this data set. + +

Part a): Write your Logistic Regression code, first step

+ +

+If you opt for the credit card data, your first task is to familiarize yourself with the data set and the scientific article. +We recommend also that you study the code example in the Logistic Regression. + +

+Write the part of the code which reads in the data and sets up the relevant data sets. + +

Part b): Write your Logistic Regression code, second step

+ +

+Now you should write your Logistic Regression code with the aim to +reproduce the Logistic Regression analysis of the scientific +article. + +

+Define your cost function and the design matrix before you start writing your code. + +

+In order to find the optimal parameters of your logistic regressor you +should include a gradient descent solver, as discussed in the +gradient descent +lectures. +Since we don't have so many data points, you may just code the +standard gradient descent with a given learning rate, or even attempt +to use the Newton-Raphson method. Alternatively, it may be useful for +the next part on neural networks to implement a stochastic gradient +descent with and without mini-batches. Stochastic gradient with +mini-batches may give the best results. You could finally compare your +code with the output from scikit-learn's toolbox for optimization +methods applied to logistic regression. + +

+To measure the performance of our classification problem we use the +so-called accuracy score. The accuracy is as you would expect just +the number of correctly guessed targets \( t_i \) divided by the total +number of targets. A perfect classifier will have an accuracy score of +\( 1 \). + +$$ +\text{Accuracy} = \frac{\sum_{i=1}^n I(t_i = y_i)}{n} , +$$ + +

+where \( I \) is the indicator function, \( 1 \) if \( t_i = y_i \) and \( 0 \) +otherwise if we have a binary classifcation problem. Here \( t_i \) +represents the target and \( y_i \) the outputs of your Logistic +Regression code. + +

+You can compare your own results with those obtained using +scikit-learn. + +

+As stated in the introduction, it can also be useful to study other datasets. In particular, when developing your own Logistic Regression code for classification problems, the so-called \ +Wisconsin Cancer data (which is a binary problem, benign or malignant tumors) may be studied. You find more \ +information about this at the Scikit-Learn site or at the "University of California at Irvine":"https://archive.ics.uci.e\ +du/ml/datasets/breast+cancer+wisconsin+(original)". The lecture slides on dimensionality reduction have sev\ +eral code examples on this data set. + +

Part c): Writing your own Neural Network code

+ +

+Your aim now, and this is the central part of this project, is to +write to your own Feed Forward Neural Network code implementing the back +propagation algorithm discussed in the lecture +slides. We +start with the Logistic Regression case and the data set discussed in parts a) and b) but train +now the network to find the optimal weights and biases. You are free +to use the codes in the above lecture slides as starting points. + +

+Discuss again your choice of cost function. + +

+Train your network and compare the results with those from your Logistic Regression code. +You should test your results against a similar code using Scikit-Learn (see the examples in the above lecture notes) or tensorflow/keras. + +

+Comment your results and give a critical discussion of the results +obtained with the Logistic Regression code and your own Neural Network +code. Make an analysis of the regularization parameters and the learning rates employed to find the optimal accurary score. + +

+A useful reference on the back progagation algorithm is Nielsen's +book. It is an excellent +read. + +

Part d): Regression analysis using neural networks

+ +

+Here we will change the cost function for our neural network code +developed in part c) in order to perform a regression (fitting a +function or some data set) analysis. As stated above, our default data +sets could be either the Franke function or the terrain data from +project 1. + +

+Compare you results from the neural network regression analysis (with a discussion of learning rates and regularization parameters) +with those you obtained in project 1. Alternatively, if you opt for other data sets, you would need to run your standard ordinary least squares, Ridge and Lasso calculations using your codes from project 1. + +

+Again, we strongly recommend that you compare your own neural Network code and results against a similar code using Scikit-Learn (see the examples in the above lecture notes) or tensorflow/keras. + +

Part e) Critical evaluation of the various algorithms

+ +

+After all these glorious calculations, you should now summarize the +various algorithms and come with a critical evaluation of their pros +and cons. Which algorithm works best for the regression case and which +is best for the classification case. These codes can also be part of +your final project 3, but now applied to other data sets. + +

Background literature

+ +
    +
  1. The text of Michael Nielsen is highly recommended, see Nielsen's book. It is an excellent read.
    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. +
  3. Mehta et al, arXiv 1803.08823, A high-bias, low-variance introduction to Machine Learning for physicists, ArXiv:1803.08823.
    4. +
+ +

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/Project2/html/Project2.html b/doc/Projects/2020/Project2/html/Project2.html new file mode 100644 index 000000000..ac2cc35e3 --- /dev/null +++ b/doc/Projects/2020/Project2/html/Project2.html @@ -0,0 +1,376 @@ + + + + + + + + +Project 2 on Machine Learning, deadline November 7 + + + + + + + + + + + + + + + + + + + + + + + +

Project 2 on Machine Learning, deadline November 7

+ +

+ + +

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

+ + +

Department of Physics, University of Oslo, Norway
+
+

+

Oct 6, 2020

+
+ +

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

+ +

+The main aim of this project is to study both classification and +regression problems by developing our own feed-forward neural network (FFNN) code. We can reuse the regression algorithms studied +in project 1. We will include logistic regresion for classification +problems and write our own FFNNcode for studying +both regression and classification problems. The codes developed in +project 1, including bootstrap and/or cross-validation as well as the +computation of the mean-squared error and/or the \( R2 \) or the accuracy score (classification problems) functions can +also be utilized in the present analysis. + +

+The data sets that we propose here are (the default sets) + +

+ +However, if you would like to study other data sets, feel free to +propose other sets. What we listed here are mere suggestions from our +side. If you opt for another data set, consider using a set which +has been studied in the scientific literature. This makes it easier +for you to compare and analyze your results. It is also an essential +elements of the scientific discussion. + +

+In particular, when developing your own Logistic Regression code for classification problems, the so-called Wisconsin Cancer data (which is a binary problem, benign or malignant tumors) may be studied. You find more information about this at the Scikit-Learn site or at the University of California at Irvine. The lecture slides on dimensionality reduction have several code examples on this data set. + +

Part a): Write your Logistic Regression code, first step

+ +

+If you opt for the credit card data, your first task is to familiarize yourself with the data set and the scientific article. +We recommend also that you study the code example in the Logistic Regression. + +

+Write the part of the code which reads in the data and sets up the relevant data sets. + +

Part b): Write your Logistic Regression code, second step

+ +

+Now you should write your Logistic Regression code with the aim to +reproduce the Logistic Regression analysis of the scientific +article. + +

+Define your cost function and the design matrix before you start writing your code. + +

+In order to find the optimal parameters of your logistic regressor you +should include a gradient descent solver, as discussed in the +gradient descent +lectures. +Since we don't have so many data points, you may just code the +standard gradient descent with a given learning rate, or even attempt +to use the Newton-Raphson method. Alternatively, it may be useful for +the next part on neural networks to implement a stochastic gradient +descent with and without mini-batches. Stochastic gradient with +mini-batches may give the best results. You could finally compare your +code with the output from scikit-learn's toolbox for optimization +methods applied to logistic regression. + +

+To measure the performance of our classification problem we use the +so-called accuracy score. The accuracy is as you would expect just +the number of correctly guessed targets \( t_i \) divided by the total +number of targets. A perfect classifier will have an accuracy score of +\( 1 \). + +$$ +\text{Accuracy} = \frac{\sum_{i=1}^n I(t_i = y_i)}{n} , +$$ + +

+where \( I \) is the indicator function, \( 1 \) if \( t_i = y_i \) and \( 0 \) +otherwise if we have a binary classifcation problem. Here \( t_i \) +represents the target and \( y_i \) the outputs of your Logistic +Regression code. + +

+You can compare your own results with those obtained using +scikit-learn. + +

+As stated in the introduction, it can also be useful to study other datasets. In particular, when developing your own Logistic Regression code for classification problems, the so-called \ +Wisconsin Cancer data (which is a binary problem, benign or malignant tumors) may be studied. You find more \ +information about this at the Scikit-Learn site or at the "University of California at Irvine":"https://archive.ics.uci.e\ +du/ml/datasets/breast+cancer+wisconsin+(original)". The lecture slides on dimensionality reduction have sev\ +eral code examples on this data set. + +

Part c): Writing your own Neural Network code

+ +

+Your aim now, and this is the central part of this project, is to +write to your own Feed Forward Neural Network code implementing the back +propagation algorithm discussed in the lecture +slides. We +start with the Logistic Regression case and the data set discussed in parts a) and b) but train +now the network to find the optimal weights and biases. You are free +to use the codes in the above lecture slides as starting points. + +

+Discuss again your choice of cost function. + +

+Train your network and compare the results with those from your Logistic Regression code. +You should test your results against a similar code using Scikit-Learn (see the examples in the above lecture notes) or tensorflow/keras. + +

+Comment your results and give a critical discussion of the results +obtained with the Logistic Regression code and your own Neural Network +code. Make an analysis of the regularization parameters and the learning rates employed to find the optimal accurary score. + +

+A useful reference on the back progagation algorithm is Nielsen's +book. It is an excellent +read. + +

Part d): Regression analysis using neural networks

+ +

+Here we will change the cost function for our neural network code +developed in part c) in order to perform a regression (fitting a +function or some data set) analysis. As stated above, our default data +sets could be either the Franke function or the terrain data from +project 1. + +

+Compare you results from the neural network regression analysis (with a discussion of learning rates and regularization parameters) +with those you obtained in project 1. Alternatively, if you opt for other data sets, you would need to run your standard ordinary least squares, Ridge and Lasso calculations using your codes from project 1. + +

+Again, we strongly recommend that you compare your own neural Network code and results against a similar code using Scikit-Learn (see the examples in the above lecture notes) or tensorflow/keras. + +

Part e) Critical evaluation of the various algorithms

+ +

+After all these glorious calculations, you should now summarize the +various algorithms and come with a critical evaluation of their pros +and cons. Which algorithm works best for the regression case and which +is best for the classification case. These codes can also be part of +your final project 3, but now applied to other data sets. + +

Background literature

+ +
    +
  1. The text of Michael Nielsen is highly recommended, see Nielsen's book. It is an excellent read.
    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. +
  3. Mehta et al, arXiv 1803.08823, A high-bias, low-variance introduction to Machine Learning for physicists, ArXiv:1803.08823.
    4. +
+ +

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/Project2/ipynb/Project2.ipynb b/doc/Projects/2020/Project2/ipynb/Project2.ipynb new file mode 100644 index 000000000..b0c37c9b5 --- /dev/null +++ b/doc/Projects/2020/Project2/ipynb/Project2.ipynb @@ -0,0 +1,259 @@ +{ + "cells": [ + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "\n", + "# Project 2 on Machine Learning, deadline November 7\n", + "\n", + " \n", + "**[Data Analysis and Machine Learning FYS-STK3155/FYS4155](http://www.uio.no/studier/emner/matnat/fys/FYS3155/index-eng.html)**, Department of Physics, University of Oslo, Norway\n", + "\n", + "Date: **Oct 6, 2020**\n", + "\n", + "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\n", + "\n", + "\n", + "\n", + "\n", + "## Classification and Regression, from linear and logistic regression to neural networks\n", + "\n", + "The main aim of this project is to study both classification and\n", + "regression problems by developing our own feed-forward neural network (FFNN) code. We can reuse the regression algorithms studied\n", + "in project 1. We will include logistic regresion for classification\n", + "problems and write our own FFNNcode for studying\n", + "both regression and classification problems. The codes developed in\n", + "project 1, including bootstrap and/or cross-validation as well as the\n", + "computation of the mean-squared error and/or the $R2$ or the accuracy score (classification problems) functions can\n", + "also be utilized in the present analysis. \n", + "\n", + "\n", + "The data sets that we propose here are (the default sets)\n", + "\n", + "* Regression (fitting a continuous function). In this part you will need to bring up your results from project 1 and compare these with what you get from you Neural Network code to be developed here. The data sets could be\n", + "\n", + "a. Either the Franke function or the terrain data from project 1, or data sets your propose.\n", + "\n", + "\n", + "* Classification. Here you will also need to develop a Logistic regression code that you will use to compare with the Neural Network code. The data set we propose are the so-called MNIST data set of images representing hand-written numbers from zero to nine. \n", + "\n", + "However, if you would like to study other data sets, feel free to\n", + "propose other sets. What we listed here are mere suggestions from our\n", + "side. If you opt for another data set, consider using a set which\n", + "has been studied in the scientific literature. This makes it easier\n", + "for you to compare and analyze your results. It is also an essential\n", + "elements of the scientific discussion.\n", + "\n", + "In particular, when developing your own Logistic Regression code for classification problems, the so-called Wisconsin Cancer data (which is a binary problem, benign or malignant tumors) may be studied. You find more information about this at the [Scikit-Learn site](https://scikit-learn.org/stable/modules/generated/sklearn.datasets.load_breast_cancer.html) or at the [University of California at Irvine](https://archive.ics.uci.edu/ml/datasets/breast+cancer+wisconsin+(original)). The [lecture slides on dimensionality reduction have several code examples on this data set](https://compphysics.github.io/MachineLearning/doc/pub/DimRed/html/DimRed.html).\n", + "\n", + "### Part a): Write your Logistic Regression code, first step\n", + "\n", + "If you opt for the credit card data, your first task is to familiarize yourself with the data set and the scientific article. \n", + "We recommend also that you study the code example in the [Logistic Regression](https://compphysics.github.io/MachineLearning/doc/pub/LogReg/html/LogReg.html). \n", + "\n", + "Write the part of the code which reads in the data and sets up the relevant data sets. \n", + "\n", + "### Part b): Write your Logistic Regression code, second step\n", + "\n", + "Now you should write your Logistic Regression code with the aim to\n", + "reproduce the Logistic Regression analysis of the [scientific\n", + "article](https://bradzzz.gitbooks.io/ga-seattle-dsi/content/dsi/dsi_05_classification_databases/2.1-lesson/assets/datasets/DefaultCreditCardClients_yeh_2009.pdf).\n", + "\n", + "Define your cost function and the design matrix before you start writing your code.\n", + "\n", + "In order to find the optimal parameters of your logistic regressor you\n", + "should include a gradient descent solver, as discussed in the\n", + "[gradient descent\n", + "lectures](https://compphysics.github.io/MachineLearning/doc/pub/Splines/html/Splines-bs.html).\n", + "Since we don't have so many data points, you may just code the\n", + "standard gradient descent with a given learning rate, or even attempt\n", + "to use the Newton-Raphson method. Alternatively, it may be useful for\n", + "the next part on neural networks to implement a stochastic gradient\n", + "descent with and without mini-batches. Stochastic gradient with\n", + "mini-batches may give the best results. You could finally compare your\n", + "code with the output from **scikit-learn**'s toolbox for optimization\n", + "methods applied to logistic regression.\n", + "\n", + "\n", + "To measure the performance of our classification problem we use the\n", + "so-called *accuracy* score. The accuracy is as you would expect just\n", + "the number of correctly guessed targets $t_i$ divided by the total\n", + "number of targets. A perfect classifier will have an accuracy score of\n", + "$1$." + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "$$\n", + "\\text{Accuracy} = \\frac{\\sum_{i=1}^n I(t_i = y_i)}{n} ,\n", + "$$" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "where $I$ is the indicator function, $1$ if $t_i = y_i$ and $0$\n", + "otherwise if we have a binary classifcation problem. Here $t_i$\n", + "represents the target and $y_i$ the outputs of your Logistic\n", + "Regression code.\n", + "\n", + "\n", + "You can compare your own results with those obtained using\n", + "**scikit-learn**.\n", + "\n", + "\n", + "As stated in the introduction, it can also be useful to study other datasets. In particular, when developing your own\tLogistic Regression code for classification problems, the so-called \\\n", + "Wisconsin Cancer data (which is a binary problem, benign or malignant tumors) may be studied. You find more \\\n", + "information about this at the [Scikit-Learn site](https://scikit-learn.org/stable/modules/generated/sklearn\\\n", + ".datasets.load_breast_cancer.html) or at the \"University\tof California at Irvine\":\"https://archive.ics.uci.e\\\n", + "du/ml/datasets/breast+cancer+wisconsin+(original)\". The\t[lecture slides on dimensionality reduction have sev\\\n", + "eral code examples on this data set](https://compphysics.github.io/MachineLearning/doc/pub/DimRed/html/DimR\\\n", + "ed.html).\n", + "\n", + "\n", + "### Part c): Writing your own Neural Network code\n", + "\n", + "Your aim now, and this is the central part of this project, is to\n", + "write to your own Feed Forward Neural Network code implementing the back\n", + "propagation algorithm discussed in the [lecture\n", + "slides](https://compphysics.github.io/MachineLearning/doc/pub/NeuralNet/html/NeuralNet-bs.html). We\n", + "start with the Logistic Regression case and the data set discussed in parts a) and b) but train\n", + "now the network to find the optimal weights and biases. You are free\n", + "to use the codes in the above lecture slides as starting points.\n", + "\n", + "Discuss again your choice of cost function.\n", + "\n", + "Train your network and compare the results with those from your Logistic Regression code. \n", + "You should test your results against a similar code using **Scikit-Learn** (see the examples in the above lecture notes) or **tensorflow/keras**. \n", + "\n", + "Comment your results and give a critical discussion of the results\n", + "obtained with the Logistic Regression code and your own Neural Network\n", + "code. Make an analysis of the regularization parameters and the learning rates employed to find the optimal accurary score.\n", + "\n", + "A useful reference on the back progagation algorithm is [Nielsen's\n", + "book](http://neuralnetworksanddeeplearning.com/). It is an excellent\n", + "read.\n", + "\n", + "\n", + "### Part d): Regression analysis using neural networks\n", + "\n", + "Here we will change the cost function for our neural network code\n", + "developed in part c) in order to perform a regression (fitting a\n", + "function or some data set) analysis. As stated above, our default data\n", + "sets could be either the Franke function or the terrain data from\n", + "project 1.\n", + "\n", + "Compare you results from the neural network regression analysis (with a discussion of learning rates and regularization parameters)\n", + "with those you obtained in project 1. Alternatively, if you opt for other data sets, you would need to run your standard ordinary least squares, Ridge and Lasso calculations using your codes from project 1.\n", + "\n", + "Again, we strongly recommend that you compare your own neural Network code and results against a similar code using **Scikit-Learn** (see the examples in the above lecture notes) or **tensorflow/keras**. \n", + "\n", + "\n", + "\n", + "### Part e) Critical evaluation of the various algorithms\n", + "\n", + "After all these glorious calculations, you should now summarize the\n", + "various algorithms and come with a critical evaluation of their pros\n", + "and cons. Which algorithm works best for the regression case and which\n", + "is best for the classification case. These codes can also be part of\n", + "your final project 3, but now applied to other data sets.\n", + "\n", + "\n", + "\n", + "\n", + "## Background literature\n", + "\n", + "1. The text of Michael Nielsen is highly recommended, see [Nielsen's book](http://neuralnetworksanddeeplearning.com/). It is an excellent read.\n", + "\n", + "2. The textbook of [Trevor Hastie, Robert Tibshirani, Jerome H. Friedman, The Elements of Statistical Learning, Springer](https://www.springer.com/gp/book/9780387848570), chapters 3 and 7 are the most relevant ones for the analysis here. \n", + "\n", + "3. [Mehta et al, arXiv 1803.08823](https://arxiv.org/abs/1803.08823), *A high-bias, low-variance introduction to Machine Learning for physicists*, ArXiv:1803.08823.\n", + "\n", + "## Introduction to numerical projects\n", + "\n", + "Here follows a brief recipe and recommendation on how to write a report for each\n", + "project.\n", + "\n", + " * Give a short description of the nature of the problem and the eventual numerical methods you have used.\n", + "\n", + " * Describe the algorithm you have used and/or developed. Here you may find it convenient to use pseudocoding. In many cases you can describe the algorithm in the program itself.\n", + "\n", + " * Include the source code of your program. Comment your program properly.\n", + "\n", + " * If possible, try to find analytic solutions, or known limits in order to test your program when developing the code.\n", + "\n", + " * Include your results either in figure form or in a table. Remember to label your results. All tables and figures should have relevant captions and labels on the axes.\n", + "\n", + " * Try to evaluate the reliabilty and numerical stability/precision of your results. If possible, include a qualitative and/or quantitative discussion of the numerical stability, eventual loss of precision etc.\n", + "\n", + " * Try to give an interpretation of you results in your answers to the problems.\n", + "\n", + " * Critique: if possible include your comments and reflections about the exercise, whether you felt you learnt something, ideas for improvements and other thoughts you've made when solving the exercise. We wish to keep this course at the interactive level and your comments can help us improve it.\n", + "\n", + " * Try to establish a practice where you log your work at the computerlab. You may find such a logbook very handy at later stages in your work, especially when you don't properly remember what a previous test version of your program did. Here you could also record the time spent on solving the exercise, various algorithms you may have tested or other topics which you feel worthy of mentioning.\n", + "\n", + "## Format for electronic delivery of report and programs\n", + "\n", + "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:\n", + "\n", + " * Use Canvas to hand in your projects, log in at with your normal UiO username and password.\n", + "\n", + " * Upload **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.\n", + "\n", + " * 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.\n", + "\n", + "Finally, \n", + "we encourage you to collaborate. Optimal working groups consist of \n", + "2-3 students. You can then hand in a common report. \n", + "\n", + "\n", + "\n", + "## Software and needed installations\n", + "\n", + "If you have Python installed (we recommend Python3) and you feel pretty familiar with installing different packages, \n", + "we recommend that you install the following Python packages via **pip** as\n", + "1. pip install numpy scipy matplotlib ipython scikit-learn tensorflow sympy pandas pillow\n", + "\n", + "For Python3, replace **pip** with **pip3**.\n", + "\n", + "See below for a discussion of **tensorflow** and **scikit-learn**. \n", + "\n", + "For OSX users we recommend also, after having installed Xcode, to install **brew**. Brew allows \n", + "for a seamless installation of additional software via for example\n", + "1. brew install python3\n", + "\n", + "For Linux users, with its variety of distributions like for example the widely popular Ubuntu distribution\n", + "you can use **pip** as well and simply install Python as \n", + "1. sudo apt-get install python3 (or python for python2.7)\n", + "\n", + "etc etc. \n", + "\n", + "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\n", + "1. [Anaconda](https://docs.anaconda.com/) 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**\n", + "\n", + "2. [Enthought canopy](https://www.enthought.com/product/canopy/) is a Python distribution for scientific and analytic computing distribution and analysis environment, available for free and under a commercial license.\n", + "\n", + "Popular software packages written in Python for ML are\n", + "\n", + "* [Scikit-learn](http://scikit-learn.org/stable/), \n", + "\n", + "* [Tensorflow](https://www.tensorflow.org/),\n", + "\n", + "* [PyTorch](http://pytorch.org/) and \n", + "\n", + "* [Keras](https://keras.io/).\n", + "\n", + "These are all freely available at their respective GitHub sites. They \n", + "encompass communities of developers in the thousands or more. And the number\n", + "of code developers and contributors keeps increasing." + ] + } + ], + "metadata": {}, + "nbformat": 4, + "nbformat_minor": 4 +} diff --git a/doc/Projects/2020/Project2/ipynb/ipynb-Project2-src.tar.gz b/doc/Projects/2020/Project2/ipynb/ipynb-Project2-src.tar.gz new file mode 100644 index 0000000000000000000000000000000000000000..60a31d1a3a5e2599ab7462c2e186f4ddf7363fe1 GIT binary patch literal 192 zcmV;x06+g9iwFQh)O=q61MSbv3c@f92k@Qu6nTQtuG@MR^x#1l;tO=HbDi4EwgdO} z?gR9sco`z}cli?%LbBhi*1JvQ?k-piAtXx@7?Wq|lq8<(38fq;?>JD%rx-j^Pdy&Yiy2mk=A%UeYN literal 0 HcmV?d00001 diff --git a/doc/Projects/2020/Project2/pdf/Project2.p.tex b/doc/Projects/2020/Project2/pdf/Project2.p.tex new file mode 100644 index 000000000..ac5ad4ad6 --- /dev/null +++ b/doc/Projects/2020/Project2/pdf/Project2.p.tex @@ -0,0 +1,417 @@ +%% +%% 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[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 2 on Machine Learning, deadline November 7 +\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} +Oct 6, 2020 +\end{center} +% --- end date --- + +\vspace{1cm} + + +\subsection{Classification and Regression, from linear and logistic regression to neural networks} + +The main aim of this project is to study both classification and +regression problems by developing our own feed-forward neural network (FFNN) code. We can reuse the regression algorithms studied +in project 1. We will include logistic regresion for classification +problems and write our own FFNNcode for studying +both regression and classification problems. The codes developed in +project 1, including bootstrap and/or cross-validation as well as the +computation of the mean-squared error and/or the $R2$ or the accuracy score (classification problems) functions can +also be utilized in the present analysis. + + +The data sets that we propose here are (the default sets) + +\begin{itemize} +\item Regression (fitting a continuous function). In this part you will need to bring up your results from project 1 and compare these with what you get from you Neural Network code to be developed here. The data sets could be +\begin{enumerate} + + \item Either the Franke function or the terrain data from project 1, or data sets your propose. + +\end{enumerate} + +\noindent +\item Classification. Here you will also need to develop a Logistic regression code that you will use to compare with the Neural Network code. The data set we propose are the so-called MNIST data set of images representing hand-written numbers from zero to nine. +\end{itemize} + +\noindent +However, if you would like to study other data sets, feel free to +propose other sets. What we listed here are mere suggestions from our +side. If you opt for another data set, consider using a set which +has been studied in the scientific literature. This makes it easier +for you to compare and analyze your results. It is also an essential +elements of the scientific discussion. + +In particular, when developing your own Logistic Regression code for classification problems, the so-called Wisconsin Cancer data (which is a binary problem, benign or malignant tumors) may be studied. You find more information about this at the \href{{https://scikit-learn.org/stable/modules/generated/sklearn.datasets.load_breast_cancer.html}}{Scikit-Learn site} or at the \href{{https://archive.ics.uci.edu/ml/datasets/breast+cancer+wisconsin+(original)}}{University of California at Irvine}. The \href{{https://compphysics.github.io/MachineLearning/doc/pub/DimRed/html/DimRed.html}}{lecture slides on dimensionality reduction have several code examples on this data set}. + +\paragraph{Part a): Write your Logistic Regression code, first step.} +If you opt for the credit card data, your first task is to familiarize yourself with the data set and the scientific article. +We recommend also that you study the code example in the \href{{https://compphysics.github.io/MachineLearning/doc/pub/LogReg/html/LogReg.html}}{Logistic Regression}. + +Write the part of the code which reads in the data and sets up the relevant data sets. + +\paragraph{Part b): Write your Logistic Regression code, second step.} +Now you should write your Logistic Regression code with the aim to +reproduce the Logistic Regression analysis of the \href{{https://bradzzz.gitbooks.io/ga-seattle-dsi/content/dsi/dsi_05_classification_databases/2.1-lesson/assets/datasets/DefaultCreditCardClients_yeh_2009.pdf}}{scientific +article}. + +Define your cost function and the design matrix before you start writing your code. + +In order to find the optimal parameters of your logistic regressor you +should include a gradient descent solver, as discussed in the +\href{{https://compphysics.github.io/MachineLearning/doc/pub/Splines/html/Splines-bs.html}}{gradient descent +lectures}. +Since we don't have so many data points, you may just code the +standard gradient descent with a given learning rate, or even attempt +to use the Newton-Raphson method. Alternatively, it may be useful for +the next part on neural networks to implement a stochastic gradient +descent with and without mini-batches. Stochastic gradient with +mini-batches may give the best results. You could finally compare your +code with the output from \textbf{scikit-learn}'s toolbox for optimization +methods applied to logistic regression. + + +To measure the performance of our classification problem we use the +so-called \emph{accuracy} score. The accuracy is as you would expect just +the number of correctly guessed targets $t_i$ divided by the total +number of targets. A perfect classifier will have an accuracy score of +$1$. + +\[ +\text{Accuracy} = \frac{\sum_{i=1}^n I(t_i = y_i)}{n} , +\] + +where $I$ is the indicator function, $1$ if $t_i = y_i$ and $0$ +otherwise if we have a binary classifcation problem. Here $t_i$ +represents the target and $y_i$ the outputs of your Logistic +Regression code. + + +You can compare your own results with those obtained using +\textbf{scikit-learn}. + + +As stated in the introduction, it can also be useful to study other datasets. In particular, when developing your own Logistic Regression code for classification problems, the so-called \ +Wisconsin Cancer data (which is a binary problem, benign or malignant tumors) may be studied. You find more \ +information about this at the \href{{https://scikit-learn.org/stable/modules/generated/sklearn\ +.datasets.load_breast_cancer.html}}{Scikit-Learn site} or at the "University of California at Irvine":"https://archive.ics.uci.e\ +du/ml/datasets/breast+cancer+wisconsin+(original)". The \href{{https://compphysics.github.io/MachineLearning/doc/pub/DimRed/html/DimR\ +ed.html}}{lecture slides on dimensionality reduction have sev\ +eral code examples on this data set}. + + +\paragraph{Part c): Writing your own Neural Network code.} +Your aim now, and this is the central part of this project, is to +write to your own Feed Forward Neural Network code implementing the back +propagation algorithm discussed in the \href{{https://compphysics.github.io/MachineLearning/doc/pub/NeuralNet/html/NeuralNet-bs.html}}{lecture +slides}. We +start with the Logistic Regression case and the data set discussed in parts a) and b) but train +now the network to find the optimal weights and biases. You are free +to use the codes in the above lecture slides as starting points. + +Discuss again your choice of cost function. + +Train your network and compare the results with those from your Logistic Regression code. +You should test your results against a similar code using \textbf{Scikit-Learn} (see the examples in the above lecture notes) or \textbf{tensorflow/keras}. + +Comment your results and give a critical discussion of the results +obtained with the Logistic Regression code and your own Neural Network +code. Make an analysis of the regularization parameters and the learning rates employed to find the optimal accurary score. + +A useful reference on the back progagation algorithm is \href{{http://neuralnetworksanddeeplearning.com/}}{Nielsen's +book}. It is an excellent +read. + + +\paragraph{Part d): Regression analysis using neural networks.} +Here we will change the cost function for our neural network code +developed in part c) in order to perform a regression (fitting a +function or some data set) analysis. As stated above, our default data +sets could be either the Franke function or the terrain data from +project 1. + +Compare you results from the neural network regression analysis (with a discussion of learning rates and regularization parameters) +with those you obtained in project 1. Alternatively, if you opt for other data sets, you would need to run your standard ordinary least squares, Ridge and Lasso calculations using your codes from project 1. + +Again, we strongly recommend that you compare your own neural Network code and results against a similar code using \textbf{Scikit-Learn} (see the examples in the above lecture notes) or \textbf{tensorflow/keras}. + + + +\paragraph{Part e) Critical evaluation of the various algorithms.} +After all these glorious calculations, you should now summarize the +various algorithms and come with a critical evaluation of their pros +and cons. Which algorithm works best for the regression case and which +is best for the classification case. These codes can also be part of +your final project 3, but now applied to other data sets. + + + + +\subsection{Background literature} + +\begin{enumerate} +\item The text of Michael Nielsen is highly recommended, see \href{{http://neuralnetworksanddeeplearning.com/}}{Nielsen's book}. It is an excellent read. + +\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. + +\item \href{{https://arxiv.org/abs/1803.08823}}{Mehta et al, arXiv 1803.08823}, \emph{A high-bias, low-variance introduction to Machine Learning for physicists}, ArXiv:1803.08823. +\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/Project2/pdf/Project2.pdf b/doc/Projects/2020/Project2/pdf/Project2.pdf new file mode 100644 index 0000000000000000000000000000000000000000..c44de0bd3f6a988be154f7907e2fd89e697b40b7 GIT binary patch literal 236422 zcma&NV{j!v_wKu6n-gsgdaViF82jKAP1=az=o;JHbdNgPaU;Q9FAndHpvEnTfhSXsF_N&cS!&m>`O z>uT;y!X#m9>}oD%Zt7rW4lf`8@8argZfpndxe=?iowV74=Ko>{yvO+E5(FEtC%L0S zmDVQD_(L~GuJi^%UYKeHvplG1`N42k|0;;g&BIQTZNTKdc`bgI5x15|^Zufg^B!ac zYVvOE*1VUo5Rp!*a)%*F`8!PeOM-&d@cOy3SF*R$zI&eay|fydgcXCvrXzL3aMA|p zaI~X&=yf=?rh0R<^GVR%-+(f5u0qRB?Wb$M_l;el;Lhu8^SWP0X z#H?)@fio~qN&~r5%(tknB;cg#%}x=s4z)tn091(Jhq5BM%mz)Ra#u5&u&qy zC^6-`hUTN^8}qV=HXN^;3QXNAg)098<#}E^eXhB#Gr`H#aXT_>PFtg8DF)M%@l0)nLD@Vu5&zIGIA%YBF>L9_ptL!G!Wa-O5&7eb`6e&o^QzQ+ zFu_7?T1R3w85Y7E0m8e0@Ei@7>&lHRN^geoZy<=06~`%a9!aQVHkRKQm4bx;&xbBd z=qt+4EbB&ZFm4AFxE&h@N6?+`<%ULBwe5i1|Fo4`A_O@e@pRo8Y#1CK&=dJa!*T!h zH3-p^$c1G?6aE0TBO~cwh zk!P;GRoL!@U8{A7V(}OZ&RC4G_@R-}Z85}6o$V%rzn#ty$jqrZCr(ym9Q_wfn-3#U zh325bk|ku2?c*F^($~`$OP^rIOi=^PhE~x9ZVsu>gv@pn5%Gtb6*AE0r*jRY5g?N9 zK^>{Hy*+H6xJylFswQBEk0h?ZAvAKIwM=I&E1t18?!dyq`(jpaG1}pezXL*Nw|u($lBpO34)^jYK| zhb&_|<;{G~%ptSxX`ucOL2|C|>OE-3j=vPICAx9&7TgUt3OoU#-x163=9h-Y-=Zdk^SQw_jkcCj#8Fi6^OAF3c)EWOvN=v= zXwjDrA!u->A;=_!`<|PoCbQ%UnUtD%K_SvY{csd?dB141uLm~xQ>g0f;;{+1>cUOO zy%VU=USb?aSmvI(2IqN*7zrDdw3A%Dy{SjTJT9-9t*tds&TQ?Q%#Z)kq^?gtYO4NR z>Hv>S2BP&r;do)LB*^ zmF3PM0FijmS6k$OO@wXYMKvjdv<(Ql6P)LubF~XT5d)<1V~rHI!#oKT z;Lo7Ia2m_RuM=`r@vRh7)W1YZ{KA4$gO9fgW1PLBUvo&QaJ7;>4~K2I7UX^x>L z{QF%oUXGVefSRp%!VReg7Lk~kCR3ly+G>luqj#1P^d&z*b0w+N@D#xaK|gs$c0yU2oX1nyJfx0Dpi;iicZe5Q*;9GhcT~1QRD<5i6oF%P-b>) zT}%tDTUTq$5&2yI{k|$3?W?>+ao%9><`K#aJo?&W$SMI+0HU%1VTu@)#KisiKtK>M zxRzF7iBt(}pqF-Bc7N&x=D@%WhS%*h(BsUYR}5S!sljf7qfAlJ2o7^=`nRGn)Z8GD z&-9A>LnTy*EwYUO3?P(D?iSxOP%fAQ+X(JlWj`p7%greJ0#aKY+uku=AW*ZKGOEoKK^r-yho{G z%Wcg3GCoL z63xY033C{_*AST0pLEUOj}GblEIk*p%8vRfjarZ-2nTNZP7x8ceWWyYI^2*yIvet! z7-Hl>vg`M<;=hR>_TaI3ee8^8x*gfFnP}Qk&EI_IPaEDs&=Ousv^qdkw_$tf?f{a{ z>p<){nvQJo#T@-56z8vzm)Y9GsG;P*xo4AF%i4meAk^a>ULD!<5)G~CPlN`=fujFW z!wV*)X6BTh8IkTotsqvzDv&N)T~cGa$t2rA?BVbYv2`|l32$z1_Wv=^f1>}*Pi!pz z4N`2Z%*_8&kjl}KbvYbC>$%jddp4=-!ZU^rPE!@&Qn<36IF@lwac;LICP7U`4F*62 zRHpd*_T=7xfkh=Zv`3_SC<^x8D7SSOhF0FzioAV2TrW?1@Ms^?D#|rJuWrTr!($Sm-m^o>VJiLtMU|D6P*zIc84 zocobmR?7Rhb+PgjfoS`Y{uS6+3$rnU)Lnb?VS{_2ATeq08#|sA72cX-D_qJ6_46A~ z;@RtK^LiI&Nisnu$u?z8DuZKwV8*-bMN)a!?Icy|x>7(4}YQvvgx^25VT_fauAX;%UBw3OnIlXc*Z z*W<0C`j(J4F`$+hrn5lUaDD+ujuIQLzlxGC@j=7>IuzW1#u;{BS6&&bW2l=2C0NUb zFn@>Y)VMtct1PNUOEf4ARKjR^IXB-}vbo*2#oTf|qK_VgLP(2xXR+ zD$?im8cKrQh5mA4Y(U$*?2j+sX6|G%N;!Y(GZErdvOjXTv_15mYdAm;Z=@n^9!|!KAUEe@@q$|2VvMVv zMb;=Kt)vs|I-*po4~C-1DXw9W+M8qSU}EhB|G{w6uprnezG0oh4?Kqn;|yR^Qbha@Y1w+y zibrWkj{L_M*CCcYl=`_IYFm-Wb%21i<8k-%6?dY;gh@P1RxOQhPee+Un+j4l#&Beop}|#dtc{xP z!#HhqA(o8B8c#T1h&Ags_HW3q#cTVEi!QtvSjv4D&H2B9&-CO zQ_V8{MlpA5FyeX0RzC7s!v|50Xh zh>DIu1S)n`3R9z$JL$owZg7diFx(WqJTy;_K-vpl<}b>ZdY(Q*kQd7D`5tnCRJQv; zK{&fN!rTCTm_#wf#U&Td^-3QOG&iOt$c=gT+`Kh+*IL#43RHvqx?8rgBKWV<}=yra+C`-jaY@T)5lW^t1meH9LN3SbtAfeZr7~N zGrx12I^ECBoPF{s9uXTIyNYlV$TcB8I`<3FfeQgH*yL}9&`o&1cpK77-7NwLW)>N~ zq)^oZT|1h9-wjp@pRwajHNsM~P>y=xJpkdbB^8}Y!tBO-vDq*N6HOFmr(y5V_U`~m z>=o{_pi}f0UFyqe^)mcze5Zu}^PVJCYENLPpA#vjNSt*jbCs`}jSw}q z-W(uV37;neaN6NMfd_}N-V_~#$ez+cX`frv$TNe1dpUbZZdwjx{wT($fHSI)DyWt~ zmZy0C_GpNtx47AB-AQOChvUCs^Fkm-)W*RT#uO+R`OEkg(XO7K1S!YCaf|W8l|=6NVAaKnaE^8;Y9Mv?|+F zGTnYOO$^*Os?H|wPC@&1=k||~<90clToT7AIMJ(O%uVf`-y~zSe|xmPhDjh~X*%_sYX5ag;@hz>XegEX6b}-W|x6QgcKUhK2?(sz6DS{2)-Rlvzg->RqleY4 z?c4TF)oVlF4v92je1t8yE|RHSL1I83ACfBAZM!TGhASp=U6~t?qHfr|XT}~>^HcJd z;p0t8aD!O!*bDaWY?$Y}Y=H}RRq`F?sHK_}s<*367YolVq8=>j%Oj$9^F$Iz+%y4F z>yLq9ngy~`{hGF-;`7QVwo<5`{YRM9xhP_1O(Y(X6LM}CHuLd@--881Z zVXz3}&g!od($+U8j4Q56Vc+#cT@to$tEb5c6p82av4y2=YQ?N0)H$-~`-ateXxmT; z)i9h(lzU-9q6AR>-s{kE7W%MS1a}~qjUVTjMt0du9ygWzMkK`$+-a<*8iP)%JGj)# zTU|nJNTw8FUhQ9&+`ZAa@OXZbc@0Atk+yL822!VG9T6KW%%Op@9P>p}ss|(~))VRq zbc9uTl<>{IziGJfxz*7Asecgxxb?c6EDQ|qPP8QlN_ny7W# zXP#Rkv9jm`>aPjJdzUoWQ2&=1cdt"AX#Mu!KE6Y!Lr-f0(kNd4&6KlrnkqW5y* zbLrjFrNfK^tJ*l(A7^W%=_xLLlZDBka-&Q#05?{qRRojAgIIL`^Nv)GwdAyt!G#{xLN+k zdPe&{dZrDnXSQ~ZK`mvP)PN#k^NONT+J++~(`|WW#U7a*YCMKUg|zIRr&nP644OEx zD`{eQAR2)8^>p~#GHdYXrVohmaQBvIryE4^FIL`;@nL$w%h9_(1VVYt2by!8Qn_F*pcqp&I zkzZw8DK_@PIJRQg?_i}yD{%vRXFC<`M5{y9;OGrHkzJIYD#BX{R+WlKxIT@dek@v_S*VNm{gSmLp zrA$nV}A#;T;UXy3mHaKiD&uh^7l=Rup+{%P22lS zc&wCs`WDYHuiL}HW+0jNSJ1C{*O2Y|tP=QW3sv!^q;dn`mq8Y~BLWi+cIHrgk?`vY zR1J3?4qo1e=;3Hd$qtSOZuScW{^RAu15e1xKUvYRQC5rN;=iF%@-wvejM+@s4j3I2 ze8N+zjUtDK7Ne6yr}d+3w~((fhK+D{T-(SJwQ9Av$#);WOI){n-wOm|r_ugmh_n8l zwTY2`h*^F{_tA6lg#UYAi7Xl_cg0$`HXBj*2I_8WDL2lKX-t^i;FOJp!Bg@ zp?W3Dn^fgDV*9*cg8MlgGg?Qj&Ia$J&DKe|gN@|Zp1Nb#OU%Modtp9VqHDt+LZ+(8 zCRArr3O&7tP-4`Z7}yBnXb2$AU!}hjxNtu%;Vmw|Jx!|3=8qU4IsEOVYqgv3aEbD~ zX6c?q{fn#xknDNq`pi^3U+9}PTfpKqk_pCAv8m=k62;^`Mr;A`aCiOEpryP`!e+_v zd;6l9->ArxBRj6%+Ga(tvzN3xrQ95Lzx1-Dwi_j<4M(RUaFL`EHfoAr>+N&1?ggZL zY9VpuXjb-KuPT6@XfK#UJi#T{8iB=@J1nfYr%@P?B6SL6RVC*}(3F|N?aUC4Y3V5R zd={&VS>hfAV91=1DV}y%4(uH>FA!Xai{TO9-#`3G((XA!XeirYe+cIEdk5<4V_R$A zE&5)?lRhSobM2RB&v%NLgmTmuJ_9q%r)wxt2BVfljfY5tmwfb#{nyji{kWwK(FY20TCoYQt~Cb;Lx3HG$Bj)Qy?J& z_q(gzsUtG9DQn8768~XW^ok{ue$EUQbrp$l1>omYcEvMRzp%CWj6Ds^8>;vc(bhhq z#o{NGgnvZP9$!Tv{`TIX1aGF1=gD3y*c+0~P2Z+k;o&mrJy$bWhwv>Xo=XgT>sAv7 z`reRCdg0mtt_fygzW(Z+*C3tzTL&@q3!=feUV)0GwZ*Hxz#bKL@WS_&inpUTXItE7 z3#C^+86Y2Rp@_9edM=p#oVipQ4>jt9%Tk7U;nks-FaO_2^NySUoe#9G*XxH5+uQuO zp`^~%7P6a8YWdia3h2zlX0-)@cZtNIIhyp${}=_`xt8z&rh2m3Bb+|F()Vv0qv-jTn!Bb5gxYDoy> zVZf&5-f6q&D%=rY$9Lueybn)9-PlyWRw-w!@F51J?)F2_DO0`CIZ~>V#ZWX7sc@%x z!h{t`hh9^JkAXZ*s=>WnCdPS6*m^7t!(uv)U(7gR5jjhZie;2f$J8qc0W#-^M3ZE8 zj|d|&BWV!vRBCJ^Y&t6WZV(WhQ?qV}x(4gg$}YoxkZCj?;>qN$F2P}M&0Md+1T78M znR!HcGm0jm83W({p`<_#iFcAl6c)CSkG@kyA#4KdHjRo{hpRVOR({XcFkx?(C2zeJ z!)o9p~ruIGnVt?$h_}C0uV=M&OhH4u8@C2+a=DXy!ali+96elHg6+qEL>!` zId>T;VCgg1V{B8Gj?Ft`(v?CU2axj~>4gnyU_qoxSRV82rxaqFKy?}x?=Y_ShA{Hfk0zRSCw= z9($L+P5g~8LoAPZud)K$=^-nK^8U~4Z0a|koU{#ixON%=1p(EmnUR~}U$~ztk(7O@ zc#f$a4(%#5Y^X7!i@@u*pRd~sC|koaE=}a=@k7L$gxbi8+3}K(pY}X?c!sSOsVOzM z=EvW9v@?vC<*ampU}^IWaeC5G%c4JQVGz{VHpQj~Gs4^g)Z#1E#fn6hD}I#0DbQ#X znv7RL_lqvgjVmf5nU{8%DH|YZvxpmmEY6kJ2og>GPSR&0hFp&S9!{uP>V`XKM~Khb zG79(5-(pG^NR2gY4X>>0saEOUJt39EJq`V6bQD6>*z2|G@~%{(kb$74#UWzUfPgaV z3+xaov=L=To7|F@YiDf=f)=ZP>}AGVGo2%39AOTJS;HMMRx&2|n;2cg95B`h_Q&(; ze2Ih9D=M*oiYPQubtr@={I`{HTbu+6dRPE?T8%6EI%c`j6?`v*1YG>95AM6moETU_ z3s_}k(4kI4uk2<@IW;`29VJC-svewpcLtLM7^M%N=N!R57GPfz9@gpVjE&196Q6!eBU(I%jHe zpuzJVWQldk+ZT{fw@uQIa~pQttx|zhMozKu+G5juG>dZkjS%B*=KSAC`-}7cg0yUG zT>n#~?beb_+T=j*`Ov(>uCN&*g}Y11;=~_Spv1XVySC@{?o+}=`yHhwobc~2z;mfaw>7Jr)<73PBk+0P5Z&s zlwjkTtKnck@T(v3sEfIL=3F>F1Dbr{O)CP~SGnv=;|iZxTMy#T-Sw!8Zp5?DmdKu% zVu5N!Dl8*FVpQT{z!di06l?ll_r8UmMu;>bSc_s~gOKpa61MAyy&_rD?(VxVyF`i5 zAQVq!TB5=gUxRwKf^U&0F=0BI?@Bu<`C!>#fg%Tj%;zwu5zr*RGYgF=*06a-McP`> z-m`u}F7MlVhqbzxdUkg02lff%U%Xa_mX{gCHocO;7XJ+ayjwX2D=VmJ9JyKhk?^k=f*!x%@**Tl@>S5)wNnq~>=FlWqRG}O$Y>iwEMMpB02}jz z^XjMGbXE?J#YBt!UQXVqgvB6#J74{2Y6VNFKKbY_L{>iQ`h3uiSN|ax$=PPhDBzC0 z4$+sCZovEY3PQH4dOP6|(0XsZpZX0bb+fc66e2lqt3ID^&TSLv47^S9IOH4WgHqg9 zjR51$G%;WPgK#JBHmqA07SF0100^W7ZxFz4y;ZA^>PGc)tCR0%mLBJLWl)HUW~nWp zp`wNu9LO+I_jAS+bkgK5Ow-(-yYtI$YZSfSRt8MSB3p8YW7E1?WcQ&OR4{GCudVHh z>K*Y6`d=uA5e%@{VDEDU=IJ45cT|w6A65MdEoEy3$3$_r#U0HFDX_d+QY0wG3vQPD zY~12Qa(@YTbg1`4YyC!Pb5j=Q!Ds_1IPNgrl&J;^(?#fh5D`>y*L`OE+Ze)9>xIQp z4TM}iMF)7vf&os}w0BH4iQy2o4m3nXl2i+4jRPkFPkeJ-wh*=-$QDupzIChE*A+x( z@cvwP+|FPIM;r~sG_>Y+1~<@iA-iupHGO3Y@{V~Hna13Q@DOr+@lB#znJbqJL&mh~93rW&PPb1po;u%7Z}A|v_`Z7ikN3_P9+c5CrKoEsQ=x+d zUEO%Eg@8@8+Ko*?59^9Osm=@vOzM1oYnoM8+A_Le59tsh)rWU5@IHx0hUf9Rqd&{? z+0^?AH%{d)I!~D=4YF55E~?-mKqQ3hxxnpZEi4<6bjZ($X6YW2H02T{0NWd3zzIyvZ68! z79-vRF9Xf4uS;4PUssPj|KZ}PtGH+&XbKKtEZLvNO5aIDy`oHx(MPG=ra2xy=$pAY z2CWSyo01MGMpIm3O_9LML@o{{_2u*{f7sED-AKuGUM`e4P>R^G**ngRHORPjU}dXO z)w!c>4dS3S3bHM#XGoAi+JOtHjEAu#>dtP{8Nr&_F@qHg+YD#yPcaE})6F$4n7JcWjM~Up2|j-D-~`E=h?M0+~6KB5ortnR&xe|vSY|TYhngYUvDYYEW$K>V&F60w%(Nu%a!uj9 zhF+lvI#gC|AnG!eXqfIpzaZx8*3!vmQh6^3 z;CAX&eEyi`2sb?@;!+zA4dzRA*r(R_(FhJ_I1T#G%l96XOHs#0$Y=WxTSV95ccd&e z>pudv*e#DgB5ehBHHGXLrW;HvgFl)%WmU2o19=YZUL%ZjJIYFbJIU6PiV)IBlNYA7 zqD}Sj{$80y6`DZy^f=f<*t%_+T*aJOkQxSOk|aj5Cy%l*)B0+1uqaa^i+kFR^h)bf z^iWA$LZaNmjL&X#<5qol27?Mlo~!UTnad9E04D5wRlZ+v1rb%;hyv5>7~dGT#O<^xk?HR&KSXi_X|23_(xk)>nIo<|B^2-{ip74>O` z4AqOPK<)JtZ}(%dyJRlKEX+Y&ZBu$Q#NMAeQ=HXTLAkeY?Ub=T z=o6OiGcT_#I7aWIQ2+6=VK^(t8oq)tu&|!PNc^LqmJuc*@}dWs|2!5{7gLXw#v9E^ zlH=j)5}OX>PE?W9{&`XtyruTra^pEe^++xj*k4&IX9 zh=;AFfxBCJO>$Kc0wH2Z73En$aPI@pEMv7rsYsb9THb3Py%duiZ-#)F8=ybbcueMZ zF_V({MsdD|JmkdYmH&?{g&YY=l4)g*Z>-t&XNDMKh}qQr*{`n$t_e3kVGa!zJ_zbN z;N&dz`C{zbF#-ZFx#7lfd$CT=V}j>})v*Le_#OA_!tnK(O~zZf(?2oyYm3lxj@m@5 z##8P)={X%z8O5c8Q48g(p@5T3CJF(s0IhELN-;^N3~1f_9UBW?ujSLxPEHU6piceh zzln?E|3+NwENuT%;%e5Cjo%zY`wwwZbHe{tzJeJ)*=caMNHPM$zsFrO%7Kuyk)Bd2 z>ia2K>7Q3oV-vMB7B9=(SGTJ_Tic`2u*Q@)TT&uuigGaw<&EX;kN+*(00LdRlD>!n8`?{u+7%acmMVUCDZUv8IFDo`gX zXhBF4)&_=6w%sy?JGki#%*gs%{mHlX{tz3T2U%`#(H5v;eJ#)L!NuIf`txZ_=Z;)7 ztfMd^r=rg9ER>%X)05%5g4H`Dqwz3edWR5zeW<}hzX&MtbmV96xHEV4&jdm}N(2ih zZai~&jQFo`us_&;-IV6-sC*b(GoybX)02>8P=t->_S_e?LS2+2!GsTZR$s0V$9NUs z)v8HRa7o9|{-vUI&?qMJunPB*wZl&4SxxF5(_MJ>sp~?E<4GaZ#_7NeoXdx0FPoU9 zo=Z=>gh^2xB&M%?GNB|D4F%eMCPsW(%>Kvb_sc1@GWmND6ND_#$Hm zg{t6;tJz>{AB)pBWkma&eI$MS=@!VFy;o{>TCC~D*Hjs>Zr7=T9T*d8Szk@zt;)GJ zsEBpDtT5cXuA`X;{vO*(Td3|e6-Hy#D7Ma)`AY@SjxPEyyYf50(c-5@zd^9HuIh)& zQy^Xd!nO^OTR#J!0o(~^)^>s%21b}%Lh*cS-Ehv1bnDPTd;&hp3ifs4ag+=yNqc*u z2=Hn?19Yy~be657A0~8&6d9JbIqBRe>DdbWcW$S0?(HPCgrLUmr^Q)&v zT2&8Y;|t1EP_&3lYe3yVN$*6cD7LsCiF7ZuvuH1kLZ09gqv<MiKjj|u* zW?koEnPEE)5OP?$ks7yZR5Gw6+e`@=zNcVdO`}!?*=gMOFFz6WNNKusL{7pKfy13u z9{jAEm1C7@ZI8CrrLW4y31)Fc7}G&ENvGfAocU9(S0WyJs%#qm^! z1S3nm`s)`TOW>`V`wu(1*1+Ll7XVgVE4z<|G;v?5pC9xnu&ZM!Haxt6s{2FEk(ka6 zVQB-iUZE}7>-MC-b8c+}a{P7^z{&qhN>(wzf$oG;#O!u)ZL2>5z^tqa}=1S2JArgEQm>3lsm zJ$DP_9?f*1RWlP^#81Y8Um`Ldm=}Q;H#Ck1p=5VW<08%F$ha>F)@?-pI1{?%K!FU7 zV@A`dgZ~5gqVrPoMW}%|6RMP!&Y0S#o8rq3y(fh`1xH%f>3P1`uN_Vgi%M-w;;23a z@D~AFDTI=#o^mIwiYDyKyu$-w^aydW?wf<)n5#U=P*_Nhg`;pv>Zgo7J_)@jaz*#r z^(PTVE>5>}EOrB+Ufb@BPqml>Ql{;%@?+Jcg$3#e|Q6}urju9*2wcClE z_l8^T5173|F~&13y}PNYhKOf<-t4vDJ=IZMbFE#nL^`*$w~B%YmubGqrvNgf330|% zJVXAdVznz*xmF07l@LhekYXE)xV5HH)BHIMvG@$7DWY7szwLY+CHcvpOE+FVQn~lU zz@^9eHLcF(t+27{{)^o0uRe?gx|{)3R#|RukB|xif3+A1rGLO!a^klP@$5B2>kKl} zd4YIY7HcIJSwycb8Zz-v!chq?{!X`Wjpko&d)IR>T;6aQh*<9uC=3UH)=qRlF>OX9 zbeImLHjN7NZ&aC-YkEonq+A)4idod`8-~0-cw_GZRHE_Q4!SQS^|4*dm9>xipV3_M z>yHoN8SzO*7Ba%|W(KsGzU33%wra)oGWS1;d8s_Hcj1D`Qe{Ga6=#77EI}{;|D8Mx zken1BF?y&-qlJ6G?_9yfhbI~#oqc5E#Zftza{k>^TV&*W4zzty>Us-VUnGL&V=2$A z^n;=V9`^5W)Li!gX?jJh&yS@#Z$--KonLKIalz}&4s$E+S!CnsJbiX_j`|0MIiFIz zcr85j?XL(ym^?E{#7^dyr|HNBa43&+FlSD1)TCOl<=(BG$9QPL7d3izl@~dm^YIYcP)2d=|>AAl$V@?S0A917>bbJ0Jd<{obdYk4bbJ4ab zFm!y8p?Zxb8UaJ4`IM&E-$nU?^v!077c;NttQKe@qN_3N+1B8S`n7pGm>70iVC*A3 zs4T|5yVks|-}ix1YQ8J7O(p&eGV<1`Z1e@E=z(r$ zEGU93GHGiII%d3Jm)1}=OtAcN@Pxs7*bP_ROyw;5U)EbJ4mMa4dIw^qDM!e>cm|7%>Hz9{qS-3#wk-KJnx5yB}y z@Ld;^(99&z%5h*vZckjmvJazp`ki3~TJu92R{S((@50<6<+CT3qzNgCWYeT40P4TPn1~VGQ0cXrzg>?|D6A8mA$>!q0HAE{S?_12ttf;s_~OCUc4_?JO`ooUkW?zR7z zf+g>l$z?kzWCaAx?m?3``-4;yNtsK0r{nf{a|I+pns7`|h;*{A6 zk@0F$c;<&U@=1>6yr<{I(ll)5%_sFx+Q_d{ADza2Tw1HB zpU>~?8F03K#rYHIQFlQd>sldbSHbH0h%EzIV?_UNsmuZo`WJWyHswq$bjmmab)WUu zcO925*0PRWPDY*u4kL$7au1Pf+ca3IMt?!HZu~1sTz{cA)C`&>@jJr&lby2+d|+C; zJJ^~yQ-F(BN8VJkDX|mlL_Y*fqh$>RJC$05nQS1#A2Vo5P97p52`!Tjvmbs>ywa(W zQ2N?vj1wW2$;{ZkUk}t-*0#sjfx7_s?yAs?me(|Lfd@=Ih?jwKBsG@UsN_Y1y_n~j z+*Af{b}+>L{!o=2`{>|W4~odX$%>y<6=<*Xaybcu`J>2=ceK~XYtwy_Tq^|>T)1g+ zr7bN9-6FztIgjM~(|!30_~zBmAs67g0s9ioLI{9LL+lGDVx9&;?(MZvD0BT=Kg6Q7fIJ>WmYFUBk;ah2GLqDP`rldPh?nKi;;;T6z#%cgTVE1L8=#k6 zFW*=n7YDlaYZYA%1o5u@Er+B-ea?uvW4`jC4e3K4u#Hr~X4w3TX%trVC$Q`G;X;H= z(FU6S@s-4ymBfuH&j z8M0;y@f-^@HBGs7=vj}YT32usox+E5y;$FFtlx_LCY|*N*8ajMXrNJ?jdSN6%+noT zs`8FHi3;v-bN1XO2*S!x0c^|6gxFCNF*-E!7JS4Xw@xrQ!TlN-x-|E6&$=KXMogTn zj>YH}YJ((blNh3}U#cyNKo@(1AE+J7ztm7U(C%!Y>AFQ*RO!2>S z9}Dw;+h15nSlQUQ{!f=AD+wzn8yoBYS^g(i{-2P9m5Y<*|MmU<|MyQewY#_;={^e( zxLw61DO4oI*<}%KR;;zyx!bx{#m(2-ZMNieo%8Nycw4UA?s_cCsOL_fUUkoMKWU49 zs%r#h?)K@-Dh+K+1ZL)j#X*qJP&(MbH5HT=V%N1W1DNOL=2`mDF%UL>fD_79SBC^3 zhYV@qZl?iQnFCNu6AHkj zbb0+@<-yEWGywdf82O=SeH*a(R%P%?DyXTNfQX4n>T%k?i$0JVj!pITEgyo6LGV#B z1!@3NKWlO003a)tkeamc{cL9ea2dwmls2I0ApBb~7iCcJMr9!CGs>o6;IYYn1&|NJ z^s$Ima7#)c9qha0dEds&((X+Y9FP$i9E=3RP~F(b7F%D~4H4U%!x9@iZ1{;O zFLw<#v*Lad5Up>jG3XEVx`-Z7RNezj>E0)%Lc8>*VGxvg+kYhW zctxp?7vcJhw7zF@3Frpy`xmo-ZGevHXz1wR>L4?9xN&vBsviaabAHHi&w%Y-Gh+WR z=pqoA6)dW2Q{zXH=*2E=C#dL+(+%u0KE+z6%^N%i)M_e3C+XelzjV@@R_$PC?qG9k z1geb)1CXmC3<_RdgC3`rE|@YfdmlZ}jG@GB#^d`N3sM`t%i^$NgQEuxSNsp|5u}{~ zYiC-3$16ENM?5K33oB|^hM$b>z*Qd6-Z#pK(0e@vU->dsW8)6Ma z`EB&iVHwn%cU3Dd^Ib2BR8>u{2L>G`QP;Gvf=oOp^bDFM`pDBGi~EnkN=R0;0mOI) z(g9W{vH57ME>5C`p|tz$!~O; zuXRamVu%l?v-EH7M!?de96hM;BVF9rg-+Dl8!*|sg(zrqC+H&x4+>`p(2y5VHhAVe z2E|MoN(^YcuRxb*i~yi5umN*3$Y=1fFDwvwDd&$lqLgq~~kVUNkpO_+b^^?3rht$(@o`DH_bM*QvZ z0)W71pup8ZcC&zGfVPUSQhnOpY-FP1&=otCIWAwtK|N}8IR-&2E zia#rfCImVLa6KAO6+>WW*bcwt?~e~Tp`Cp;g_vlV^u&z6_vy{?D0|tMQfo$rWK#BOIty^0&dzb>uWRRb(32cC9Wcr@ z1@UwV7Klp%I$v>UO59w6nZt%j?RYj5mT8ilvD|FvPLL5v$oD4 z)|PrGwTB0*N{wj?2S2#r)E68jq2qZ2+Ve4f+iF#wd&;~tV6Gi-K8uOFioWkI*Jv3} z%dGupBPjijQsc$wGb=Pe0XoV@bZ^dRh%{NbOcw@q8bHoxShp%Al+TZnwn-C`1$dh) zRJ$pOwRTlPua8-_og78Zbs9$=!xz3DRR;b$*Raj_FkYdm7c#|DOQpC=HO)!+ks25t za+wgDZNea>1Ay+l#L-79|1&w`85?<2X~gu$o-+W-6w({0dd5gUKc_Pc0{ACmlJU8HJ;M=j61;sq=n#J*-R$IdYDiPMgm!rie%e(W+r7q7fnpn^QSQQsfNs@dh|1Lap>Yv%wX=QhF!&WS_Ih#rg zYcD-i=Z7Jxb4~BBWf_6HQzi0qTS=As;I{p&?PI1l4T)y#ob|li2m%bqE%Ul&c8-9% z8|Xmo8EZfhlVvf9J~o{QhCPk>etZ{P$)Y&LpsG<;K^GSmkhtDOVb_~_I1wmh)w2F; z?iI^cd%-<{`nGBu-N~)#@l@9A2QTzR(%R5dCsc|c+r_^XOzx*m#J+Q=@Yd}7fgen7Ao7wM5mxdX}I)9UCX;rLQ%qRXty+w5uoY`9J*qAhT za3_K*C^(WE*b<@>(4&;wk#1^&qD|z_$N*0<q5M(-HZBrt*97uQpkGp zg4!Dpx1Rh@Rumo3*7v$&iz_rCM5*zs_zi6I zpIMWw^IK&gyw|ez`ST6OadAZTd`2SLAf+;i?Il!U5ur2E6*+kt^(Qe%yGps;;-vd_ z*SSHl#q>+1j60&0Tv!2)2iF6@;g}7E(`Ir(iW&v(Axi~e1)Sct9JROW*{5?y>Cr)= zs7jAw5^>y=Z)=&W^>(}C2D=aGx#Z;j$qy<@6<%L@h*H|JenlP#4s;>q@d1!7TOiP0 zk}ocp+0+TX=MOG|!QGMJl}#z8h8-A)eUWi$j>pAf*nVv<^sJuor_I0>1Lz&e2yUSu zigA^C?ROz}2VBxQWjM}ZgS`(kEjg{tD(CDwrQzeQ0?kT{+BJ+7&pZwh54!6Qd#ndq z@|f~Tqz%BE_x)Mm668q&R$3eUwUdsGenbl&`HEl&@soHd;(w~HEE>(OmFG*3aMG5M z?6^{t@^T&kVMG&vT9LQ@Ru$Cl82waK@3M|*_8l*Uk`6eJPgTr?9c0;>u!9HDj#>`P zN-yG>D&fnWmk}m78l}s#>-*vgpyp^Iuc0`B5uH3~q>OonxOC8a(We}LTIQR29K7nk zQs(#2mF&$MNrFO=niWNj<`NhGMplOx&%wX&5S1qoF%#;3Y&*zrF%Yt;vF;qzbL8*6y;mf|T+2<9E@c3^?%M`7VYS^$fE&#k3CQ0qj4Ag$m$d9q%UEEggUzoR352N$IJ56md&#roaUY zoBB3$@M;}CO)Lk#K-6e%OEnwa|IW2f{i_S=fojxG_U1RlEJB*4TeaWZLq+KxbrgPx zltUb5@6e9DFV2Wq)ux54#Ra%ul(#~)G>l21Qm&}Fi4ZWf&0<%z<*1sN#cok&S$#i7 zZcd%Y@dA>u41-aO+#r4d)#^-vL^B#3aqilshr4aRKp5V?kVQ1($|`&S6(7ZxDx-S! z6HXwcm1LMdW}p}>hw%!bRdtgAsjE!Y6helK_TJdd@Til53x^!G57|4FNHUUu9}e8$ z6ua6&lU+!s^Jm}brQ0g8i{!Db92Ii~9vFNvKqu{v7rmlOCfFMb7kw+{}{E-kv3Mmf?_8qyXeK5{Jp#dfKsb@ou{NKmw7NG0P{U=MVyb3o%K1H+Y z8Ta_0%)uvfUDWRiJepoyy*+ULbyX1i-p;07q#=yrv4aF)W5!yY;vJ&f(qwK)nyV2~ zc8IMF{g6`Zr-Cs%x@MO`PGDjjOgL&eP1?DEgEIM8ap0OzWGukR(dTYS)rdp{!Du2C zPy4nUOZ&E>Q907xb`{XanS~up$m)yQB?@0Dw3W}mkp#tM8j&jC&i3L#-5D6cIYvF> zd}v*H3TpO}l8I%g|Ko}%{+w7?LEwadV7my2+>+X??g_(4z0ToxKa50Jn4iJsmORUb zO=Y2UUqhHUm0#QtK;lgCJGxo)NGyxgs=tAEURH*J9`EbA!3OmbbzANEP#xlGkiI2t z&L->!AL^@}Wf1jm+T_9t2y=ZE*)r&Eh^#kx9vcf<2&=|6&17KmN1k6DM*5cZXhroI zxiYA`>#(F21X$oCt>keMysSbYPy{_VP8#&b(}g_Ixuk`R##d8=0P(=Rg%&)*fFd6(8rz`AgifXkiqNKDoe z?kD?krV)-UOBuEAwc|GO?Nvwnte^0d^wcLG8y^#`3=0RvqnME6`4pB)m8+oC;n^4n zq2wnao=}8N!YIIPT*MA|XXa@SLA`$u+1ztyhNk9fQvDd%0kzQxgw{F5k4?nQs!$Jw()6@`a)WCL=?mH!*Ez%^cD$jo+!A)|rE zn;a%Mc+oP#5qHCZ!qXvCqsMb7nKk9wYt^>FkYrqhfHsgo<1VOE6KTglWs zR5O~6=^-#T<>+4ZX=BVqUmywtR>9z$z?vhP?}e=B0Yq=qBE5K{%ySrNF>g9^0I*T7 ziu3LGb8IvBppm7`D&p?0K5oJ7L3l{kMAq!a^YSSqF-GEAPiSSa)gSB{V2`Z@nW zME6mt?-3?pGgMu+_c(a?eAi4Or=6bvAl!eOv!C-70qv_-)79|T zi+A1NemWlqdlk5E&FJ!mTSpFH?U*O?sMtAH>bTftm1>TO1gm+;^HlU4XZY`Ec19%3 zGp+ly5*V>oWchw7)#I2@iSo#pM_0~}>j(rXxn69)V3GTv`@c^%%p{-x3Sqsjmo0qx zCW&XGY@-YU%ilWlaRtR^KR=5v%`)qwe_o2 zPbcC3l}$~_KD7wW?~Tkn$XR@f=7D#z_p7}?Ps2Bdzx&tZ&?jg2IukI@SJ?gtw)DM- zc660Wv4_DO>#v+HNgR)ksUsPa5=MA~Q!w@P*H2_seQl!qPJ(*5fRchJs~q#LE@DN(VUJoN5^(x@-IxVCVW?oB{Eua7TU~wtF^Pd5!rq9K38;NVU~i zH4ZY=+=VWp@y8=Vhlj;HD?5Y-U>^nks9kd}OEvkCjs6SY#$>IRN;*;??ek88$1-VY zo)0ZF-T~O=X_0%xH|x;sW5UV4*D70qUl4AaD9Km63Lsy8tY2^JjM5S49E9vb0b1Zg z=KH;?U_8uO=Dw?8yqZ=y?LYew`-z%3`O)JOghWaH>hvfmQXCX;=%H zn+azJa$dBnu(gag=TRQFH_G956?}(NK}WbhkT@F-d%rJQ0M_cHJ+_GEFVK zyjo21*Jgqe0|2g8%nycuZ67G;MMv~ui%Zn5q6}&wU#l1g^o4?QvYOfrmD3Y_dxicP z_Zxz?0UxnvHRW>^ovfK71A|u{cRUv6SFaphxv1`^+{=5qp>X+i?*#bGRsl~)S>ECb zr3^a3dDH-3bm+Nw)*%qpJuxPSJG&8+2#(Z|FK!MYsm$?_IVoNltPv}doFH$?5$WF3 zI0GOsDA;S7q~GE3xg>-A2WgHU&gIuE`L&Tm|{dIeN}4`@H9Hs#%`iV$^OWIm4W`ayL}SSt#ub+$L8q>-YKnwo!ej zl(X8)xucj5na{WM+Y~j|v8r3V>lF6GRjk1Eq$=`F;5GjmIJT5#GOSMw@oiW_*P(5sk@@%EcsSpXsh#GCQ}aGJj>w>x8}{{oseLR|E86YF*9Un0KitrKoWGg zu*nO8fKJoZj8Yw$0^WP{g-gB_u2E!ub{Jpg9t=feVx0V*KV=RPaZO+wE^3bdSorf8 zvQ`Myt14@m68wASPu0wjz4?G+@~?Rc$(apf1Sc8vd*+q!FirABJxnLYfe#@JQlJBNMN!Vp&>Fv9V~ z;PeD>7wjY#6}>K>0pA|Hn+x$X@84oL5*;gac(8# zx)kboZAW)%!6Ug`lp`3Tx3`0YEvr&hXz(cJT*&dW9e6-ESW@4kb)E%s`r~I6THSma zJ^V#N}q#xW+@wLKO7DATO4HFZe18YfNl}UX3Yh63>VcuVnMu_ z=?8OW@G4)h;`|%St{$W{Xu%SP`JINKj-Z{nUcJPUVsD23TinKI!%_mhN&LbSB6f7H z$BAg)uvp(bf#q!si$}$I>~1xEBm)AKZ;94O0F!O`=+xBc4+?BUHMJqv2C;;JaL40{ zAU{MXR5WIQUA>G7cKH4!R&vc=nPXpsBDETC#uo8#uItj7I*U>^cEj+~bK za4ee_RaaJHr2!%p08e2ad?dBY6v))bYLRMr$MrVbOoiWOs~D9 z_P3D=$1b-c3p=*y?$I3Vu6jaw)Tr(O$n;hN(D;LDaf@AJ@0J?i)yTp z_|3s{nG=@8cEIm*Uq*uEus!Z+e+<1&gA?htK{ZsiyTn-WhC4TP-H<)?MP$t{%R7T` zTYnKLSe?b?(!+`{nxCZFbh%t@V05E{|Cw`G#~ipre~54l^qQ@RM=4q}8L9B8e1Gtg z6K}a~bW!*jAA6ooU!k~=D4cky1k-V7%}@&67`{qyS+x$DDw*TF<8QdW(u3huKCWHJ z)gz@2hXGj@Ufn<{HgU@PTVQ~t>itd>q2~b&e0msqy}d(&0& zqdt*LTw?)Um6uL4Ki$mimwjze>jSLS&Yp5OQKX%)LcE{{hr6g-m$zde{M&46^@Uxm z;klJZ5;I&V&mD#1$PMSdJ#cIZQZ--HH#7JW>PR64E9|XINK{2ginJ-y2jBKKmytTv z15GrB__@bwWqGWRxC25-F`7K$8X$=-gCrF(k_>sW1heWj8vPM+#3w45 z?00?4QgMOC)TRS@^GN$qmlC07Q_Ps9k7D2)_H49-e|t+k)q|T9+3!YLO~Ru!zI`<% zxk-)9xT~Z~IYJL6V#%_|9$@2}0l?7W;w5q%G4D*S5W4S(XIqH)<#c%Iy7&S1(eKfa zm-@VbMZFn43XfshHr=l4Io!{oRIwnCyU#^1x!||J=45(75hQTEuwf1f>|25)H<|ssaOSpqJyIOtK`qciB=;#`YO=*LRt? z>Zh-Sw}Y5=KV_S}fQT^aB;~#Xj)d;xJJ;u`AWNjYf-N~{yuN2-@IyG!PNrqcTe3~H zhV_Q=9pkOPHi15JK}ZJIj3)dt<;@}-O&Rp7{~0i?$mgS$uB=&+=O1%K5=suPYaI&3 zXVZmfqykQR?-+QKlS}g);UC_Cz?7gei|dXmX>bq<_S({G#TI>u+>E^$u7WVIZ)7^T zsEAsCKd`&ONq8?ZAeV>7JHjJDH{E{>e&W|_X69ltVWZe7X*m1Z3FzZ8ThVo9{{a z=_Z-(R=i{nK@P{{=hcP>E!!!JPR~a4PKSVU#Bw&g@G%%Ubg&gfye~+~q*OuVuM2@+ z8*E-vt*4Nk!&uQdud8Lnji$=(UI=9aaS#OXs#u;g9&0_q6f0pS93;rgR{Of4Ds+=) z+#&Hu2ATtLMWwf!xg_xEDLJ17s)CX4oGlz*0Y}&u<)U4F<+kO@o30QGS{H?jyP$QM z626{B442M8KX+k8K$Ql4$8kg~lyP^$Nu$lmLhSY<1RY$8cA#F3N8mwB+(~GY zj?HNiS5i6pa$`=)@?K{LpaJZ!I7b!Oi)HgQ`X%R@c}qy4LB}|!nAcqvOuAA?`YYqp zi2^tiQ*{aA;AK7gHohS(hbxAY* zHIFB(7Iq~1E2AV100)A)k>?zUxx__Jiw^Bitx<%_UJrHKxzYaTT>M~dt9<4$t_hH) zPV|t23ssh|Wg+OVi~6X&;Kdpo_BV7oMoDXWtgWa-Pc{v(%7=+cXpU^i9R)6WzVZLZBS1NrQ`{4ag2OEH2%g)TLKk1P*MphqJ+2)o+F{~MV?ZSyZD4wL~3W0$7zt!{F^b$ z=6J<6s)iUj&uBu*P(vhYarTV&GPHYDUQt8naF)f}IvycXM7G+__BpnsWPE+TfQeq>_gf%=71qMXCW` zQh<-B*M@-cX_ifnH#5^IkHaa?%Ny#>Yda5%dKr*WrOPnX<)4i}d$sak?5XbJkT%zi z3~JJRNS~L3iQ!5l5iu&d+H;AC^72dZGr7tnOjHqvqu92ts_8yx-%<($dBJyuV$oWNK)GIaM@BQ)oW(MhB{;QXk4D*!{}|n zN#K037;;&Lhmjbd>A7h6;m>j0RZ+~Do*jD6fPkBpi?+k#-}%xus=}IMQ|r_EtpMb_ zWU?CY7$?(GN6ak7M0uVnMk)KyKqxUdm)oC}J)c@QbUp8hBb?j`o8~B6&3>@K>D%!6 z+CF^XVOS~lMtBvg=c_D8^!rLik+#TB?6F3;ZYx9za!-4}>D1-;aJ*pDk2p03*J8{x z7PDx;r;{;vnniHIDIO53VKN$SL$EF5U4w_J2va28LKkB_f}WulIXIw#ONohoPPnI?_Nmk9fzc7Y)&! z3{;{?;qdpO+v{LrZ8R%GcJ-kob?iQ(>s9Pak>26*)HlCq&}Y@-c}4$X7dG8qH)qHy z&Bh{O^gGs_5-B+ZZ!ft^0*w32OQC8KG#Jaj18IMN1hNhJ*Scse> z))=W6*Q3sCF-~5-yMk{|O7E%gw4xn8asZpqyR(aRm_X(eB`pfpbVO3&8HVfX9Zt2J zPj~35*=CA|hJ(i7)2CCTnRVS75HQmt`%2Zzlc5#I9my18fA`|s3C#0R%NF6pyk0*j z`<6P|x~I!hFCyH80s*t#*0g*?%vNK={TIysQ9?P4MjB4oC7s9QuV;5>yi7 z)p8l4z8iaOji#S_CyVcM_KFuxoMZ6&q-$<_TEVa?);!f}%r5oQut9=K1kyGQ3I48= zuy{~**X~sv)vQo~re(Iyp`=d2C+W&TpHDi*sOG;u@gz9}nG@)6kA!y^oDEL8K39Gg z)n}gCKd+tPQ_vlNM)n0q!os{;qf%oy)Kp51)AqTgs9QIez@G%3WJ~k&GcE9fZ0u)U zE5*u8pGGTh@no6|grBF@c){F$KGR32%v(E+5Cfk}E<`}TiJ%2WXFx%Nr{Q$8Z;HL! zmVIa5&tMQ*QY*cD%JI-(zsZH=KEcM|_Hc+0GVPI-e>3Ys%|!%jG~ZP_EXAmyM!3sh zdvTO_@DX@T?uLL?TJvYG)|4am^pR!c*8`J}ATv9-k{EgR|3UREG$Xt+Z9EAAegf^2 z(ctyyh?_!y-7GCl#Xzv27J)s2<`P0+w4%kzDJ_oO;&aD_9=U>RxvKe;uOBo$0Snk^ zVUQHm2j9{G$&z}Ba=5F&@TJHPxdHQ^y2<;n#f(bM$z8nsZWgV$75dPNpN8&BY-Xxb3a5kjw&gP;{)? z7D)4kVj=YUfx@Bw%IK(C!XCxio-7k5X_Rb1bC{3(6XkOty3U9y8ErM2@giA{%%G47 zY9#PrU8cle&?j(JFzv?9RP_wuSofKth9= znoda}M^a5TjN%=w(@&2MHq`54#_!e+;9cLdiIdyV!sSL+-#v~TRMxe7#9cOQ zGG33Vhzn&8RFTiX({s%u?!4b6#FaRZV3Fj?gvP;Veq*Y}pHNII*ON-c)T_5C&WG$E z_g$U;mP_hFnd>3+nt}-&LVkWe0DFU^Sm>yYJ(&=ROKQ zkz|aSB|dCR4UJBO%X~s;?v8=+<7w3Bp|-Yzal%Us!a^)`7pW+@=G6nM9Q>T1ul z4{87!tZeC0Q36W5vv@ebh8hE&Np|X_a-B}EKj;`e|6;9=#S*YyHkgmk+K4a$4>Mp8 zbPHV(J8f1OWrvms+jHuS%RuX--OS43vCMaDf77NZx|yb+<6*7Ifjeai_C@+7Mj)Qy zDWsNA9ghf);f(+A2k3(rlBA{%AlT-}kc<|0rNcR{G}h*g_eyFni~^VVSAVAz8S9FD zH5uQ7Cz5IK=O)XisMAv#5V!|@=l}-tVP9#X zjO3au5TqZJV-e(D(y&WSL6-WMgRM4D;7iMyCn1fQTCPPRFaAp)lDxG|E74;76eNvLa2oqVCo0nb8(#88=n34RJF{tJeoQ2 zsUmQdCe|ZrmKR&R>l*Y%fOqSQ6bbR($HRbXfJ=Kk0fCbYS(?660C(&X2;mQspLS~F zmF1jWz{xki=uX+K_$9E%D_}xDHh!5#z1PDFiLfqIV1FHz^s(rZ&K#e8CIQTBf1ADY zR9x1#9L}A785cHb@9C5g(dKWt>zwLu4Vj8=XFCViaU&y+8yiVavBHxAi6O)J$qGEW zG_0=Y4)Hsry|*FR&R8w_5HnQl=-$P&d(+`Mn}=I2Pgkn&{`p8yjHL_)b=u%%K;OvG zRWr(0Z|I-Zh1rx{34sr;KYq=COx@{iPz2>}H`0m=re0ju&S*WnUYUgeOrNLuT`FM@#e?$2C?DFoFm|HRaEcvG1 zmtW5FjEJ%!I8-~~+byj(yJZ3$DOv|evUvG=K~8K+6}fo}jL60Q3b!Yn3?9H`Ww~a8PJ`BqE-742Xl!i8vkI`-<(ezQ z=7)4%jvn1TpE9#;K_CjGse!;{qgGTV#q@BI?pn)W*l~}EEzpShCPE&%94Sp;xpbrb z{H;8+zgqpqeO;L^Q*ec)=Vo-`SEPF6{E)v&`a``<5B3|~h_U)QN~ittX8gu|IMB+f`xCFKldTxHH+9X{&qkV)4qdU~dkUl-%&R1Sn{gWB~g>I&8PNs({%2s_$q)h|6)M$P1cklXX6@fhHCuel&22X2wF^pIY zg#)0*PYvmS(w`1N6pP!DAsL_eLqbJmkg?m&9+$>-bO#ofwO;%{OE)Zwi7|TGcablB z`nV?8r&FpHdjR2U$wGqs?A3(k`N>duyEP*q5ptc(x|gDf4Vb5-$Av*_KPN|7mUmGQ zzp!*YN{Z-*IVqmgdR>D#Z4 zf1JRK1(K3FB5K*_y0CLuXkF&e^_@mOQsYo2cNK(Xa0vYN1*MQNnhl0@_X^2y51Ca4Aa#fiRDz;lRCRV8266<_N7wYn`Vxx3 zV0aF7TPZgC69J19GySf%RugToL8vu(Es*2dWu?V4$8@21RFc8h3@fYi`Fdk%QNd){ zLzl3n_h*4xW&qy`>w*Gfgqd{ki`GR$Del0^9^nda>4FB51FVW$>BRn`)6Ju4qOiHV zj2n02W4``$F=384&_z>Vs$ZnoH{OLhRaSxm_Q#~)@%v+PlUdF7rr%bYZ`J~Eml9$F zDMl@N)xqb90`28Y6tV<_yCUnyY_B|^V|`h9EvHHzm9VxSoQo_6 zHBv`~k_dNKonnq(81Dz{j)q-5E$>0OK};coqr=dWvxY4 zKTsksF45SuTN6d3hNlVlOSZPQ4DnV%OR@Zsi&AUfp=%K~G5Ga>CK1Bmd%kK>q=2~v zgjFhZbfGFzD5S%9`W>}vGxJx}5VV{Ity$aayc0DWuWWeEEAcF1#n?pXEZ6<-Qnhe! z)4HAwB1ZIn6qmoxYe4s6V7uzIk(-XQWWg#>>3UjpoXWp4jYus+>knXl@xunbdpC^^ z95|81MOq>RUHdY|b<>i;rOZL84ExH+(ZTdQ93grD5mI`+U0+1M_PM4SH1krEBc9FG zn;giQvsw%+d3nq)b=iASQk-d_zZcl`P7C_tVhCZ#jy*Upx*B;TwxLuJ@)lY$ zGb>n|%s)Mp*J*Wd^;kH&u6hxl|6pVfLqleSiQ&!~tw>U?$@v}={^H{nmjvhmXPmE6MJsSWF0a)klrdmRZM z=OuUj8$Tv68mfSg&TPRzz^Ij-@Imj}1jM#_W6>@6b*w%`6Yb`G6{o>og4@|`r>8%W z{|L(N9Dl6p#%%+cd5>qmvp`V?%Aj8OX}a5jfg37~I7e$%bSmw5V(yzzrG-&%VRBKU zaaZbom)&z_vI?_>^w|1Pg@YO#_Y~YHx~Az1)Uehq45!DEH>S_p10;fH?rtf4N~`cs zJ~vTM9M2?@@0YtJ*($^pUaQ>7vSap&j_=k)-;~v`9#090@Z2o8F{0CFedf!Ja@Rc5 zfBpWXS87RBuz_4SKLzQvYwlb-$nty>N$QVkI+-Nzx31aB`voxr;nTXL&=eP>2ilYP zSlzo-5(pd=8zn>-$zxX}n-~30%=lYmiRZSxdRzuZNk8ZOAYUHI7zTJX)en__K+a3t>iFtepy^c?Q z=#rW(DEH@(D0sBq^WFo$UW4nU!%a6M#AsFSGzJ)vyem-8aPYu6dF$>~+F`4yi!4R2 zRzqW7F*-2(Bsm##iJ}#4aipm{zi$gS31cDjvl`&)p&t2)ps_+9(=*h6W9bl1&}cj8 z=JJLW1fl3Dqouq*tgl^$4m@eKH)~^1VLfT@%Q?@g)Cl&ly(DIyFPnv%pbQC5%tUV zIiO2_A*Fa}S3KRLYV}E07E9w#xIyOy-ZAxk z=Z3Ca!-!x?4C3HUkT?Ia;!pu`= zvL0r8$R1N&8)U-24q8$VJ>N0cOkYSU9k#~dE3@?E z^(@@`-mS%$+Sb{kA8gkn>83@w>&*Z-56Fyum76}4mrZ0;d$#ug4 zejLLQb_qg~5}D1filA$g#F`JT{jAS>JVrk_f+JPN!d}7>J(Wa6B%JzQ zBs0gqd55t&J*9l9FkT(yIbJ^Ww9F=WxZyWN#5Tv~`>J?TE{;y9kUMA>LuXCNw-q_s zjp&7(^L)pPrw0F!;O(QJBk&4?;RfumQ&yNBv?OYB zQ**CWZ)<;vmYBr1C;g?B$X&EB{l+Nnl~*HFUCSJ$VL!vb+G`r%2tb1%&v&coHj4Kw zQv5_{yC9*yjR)H;wGWa#S*LiiNntaGegUxS2H3<;vgXc49g~t^4z=P?Im2 zkB0~Rip53{NG*NDOEN#S*q<1MIPDmvroAaOFco(i-W2xLEvGrb^#=<+(Hmv}Y>4^Y z&*cIDrP+}!so48S3>fAYgjU@&7bZ}F{Jc!S%{?Rgttn+$Sk)F{ym+p#YuI$vU7m~Y zLwZ>L7wO(poV#~l~$wW?QjLhe>PhQ+65e0n07>;_gS1 zdgS3y@ZD&Y_IqG>x13y7*XJBj^^{h6H@ue3a?=VPXD}J4#ajz^pOf-uy_KSg$s^*f z*pX9n!Fn6{QHW`@CDCcFv%1mBN_#cuwLvf-V^Fny4{xI~d#K;aQQfca0;}jHxih3bagcEW8k|#2Q;{JW97Bt?_k-`r3 zyf!<^8`6_{^=?sC8_UoJW9IEkXRFsXja<|lWC#lY28*a2(NYhn)?6v#e4mkRz6-EMpr<}04w5sO=#DTJ2$Yq z`v8ZlPxtI~xHKB?(HIrIw8YHX1YuksiOipKzJ!vxEi9=rbZbBNs$Gi)EZaE{`l`cM zyt}Oe^M>gLG~)Qnu5cg>_H2Bse1PgPrf0k#jUzaVnbLReE%C2DmX1U|ppAqTrLZbV zJ?8;5YNBbk8g;mWE_ZbaJ9EqZ14mWlyT4WGnhwH*;2GL)%Y-xYr!PEX(o+exG{C4c zHFoYT=F(4ST~MqBR5{v?Z5P!j-o6+_u*_yo6GvF)mDB0ExYHl~))I45K!@T@h#u|7 z4qb8em)i}A;5|5NCmFkPgfo4v)g86%VD%(9&Q9lL$~@8HZc-e+QGSGbLG_iX%X!-n zRn;s#>u&46FC$YI7%j4oNpyAa%SYc(w%x9x53Ow3brZp%-pG~};=`Un@euH$Fb=LQ zktWb=&b6um<)31XfaewcbHJM{J{DMf6X7Kl)l1<7^JY9IxZ1F5~=22*p3a+)rqo<7+of$zZje z4az3T;UdBy`e#qv>>9f`^u6=F9*?++t_zMo{bdI!nxltYVDjm)U!3z`=2W>Tl z?81zgKH;UcFVKE7`;9T8oydtJSsvpv`;7slHSEIF_HDWvXad7r2ORCSQ{7%U#i9(mZ9Tn{jmR>!S5T&|&aQ=1&YxB4KLbjF|+M zKQYQIu9S(A_yU{YjO^d$Y8=)$Eb$>@Qud-zDUOt8LgfbgfAq{YU{5nm;lv6bR_mJo z(kV`fMN?~f<`DU<{mY?M>$CO{diHQQJM}~_x`F(u?1r!ya`0N06=D*J~I%no|8 zHuvgloW&kkZmN_41+vE(yP<~)pX0tO>b*) zD~bm1`ZG=2P^cy6$#LFn@+p{xaJI;F%i}O%2s@m5s@4cI*%vUEu z+Y{m=V9`YU1ZK~+qx#Q*=mPd%J5OXjV*Fi7ICXVnwO|VExLu!@?ZmmOZfox_NAFY$ zHC+vc!6t=badZ+eSBP-po#x0Cj0aL6Aflj}?l2e?)k_1W&h|p5%nOvmC{l)mZz)WBmSh{F zS;S^aoxdp}@+c!3e$P9>}O7JHQT)FW;kgv1#SEs zCH$61o;4R9vVLqMj@_Tc>daeG*991RQXm&z^`^}kK}D&BtnbKUH%7JE0+i=Rt66y& zENiRzWA#T#t{3b8PoTN;udR!o%uM1eSn@UB0LL4*jQKGY<9W81Gh-yzNc8b*fqR!K5+c#~d{+ z6O>jn8jolO-{A~eTm{hqC2X%tu4Il!{Y+Ath4va5!*dJX7649aTolQN+Y_zw78dne zb4TAuybH$Ar_N3iY&xAkrYkO1ZaM1()!Z+v|oms(~$YPGpnY- z7iF$1Lu45}>Myk4A8^rDPaZl34!rI6f5eCBt3BKuFD)`p=0qV%f*+{TS&i?GsL8$% zke-66jnDN&H4p&lT9+vC*dLv)Dn-uKqpG?!BkYNW_f-nP(t)Gu0>&NJQV*i7hO2Vi z;{9LQSd&=?@mlM`s{CR={SJNHUw%_>0@iokr(+{l(|bxRC}{6)W|4nWuoJ|4>Y-Z` z7yYZ6x;!Omhjx|)5JkBU$Y6NwV{QvF?Yx4jh%JTUpyQ${b#BOFuUpvrHwPJ6wA*DD z4gGY7Bnf_6LV93r!rQ zB8%s>s947mUy2;gRwn=MX&QJIcp><753L88wz8ap|2+SC6D!Vn1CrCFWmKC$f#G4D zy6iLnTk=2nt}v5-R=?{L>=nn5JE(61Sl=^BBiOrZcRn*$Xo!2HFaqw6jC|&#{Ndn} zO4{UKgoEz&DiwES08_0-l>8-%&y%CMgLqvq77wN#m?@Y#r@#N4^PD@wTt_cmg|t z*9B#woVQ!B+KLUTCOx^xLnh(UB$uU$${c3964v&~^OA&F$iKza5LOjZMw2-9h&Uek z)*ZY(BReo3`3=NJ#dkItyCD^}mc2I15Y=fhkz@ zbS5jWUs;J@!>|-8e!Jc8#DLDLOTEN8_0W;lU3152%B-qsop{CVdFCf=64mNMG&MJTqxy`Z+@{A(QJ z_9nJcjvj2{g(P$^e&WH`O<(+(a*9AIU5b;oRUX%nb}!4Iu+4W4b%{SH#bYPKPSB#I zW_znTsyvUB{s@ZHqElMpzDjltjq@iGKMhK%7QUwHF+17xTj5T=VcprL%2!6k%1V)p zyk%JBOqal#_Y4_XtIG-*!O{fIfZ3UW4Sxp_QVxS>>NvD#9YDC8Pr0F2@Omgc^_%Fd_cCu1>J*I-f=h zF1YusB0kMR!ygbB=u#KPgK^39qL6)E-t_p(p@EF?SDiRYT*q2F%+abqES}Um^Vx&S z%y;}Z80_txG8+~UoXQnX5JKH4H=QJg5#M3o>m zW6!EKU7V;m>K}=kZtHNpOY-p}{c39Z`-r-59q?{V=HTGL(}x-c2Oa*UR5T=|9geBy zw6fpVC48& z2dw57a*_m=J@XXeh|zG~F06rOpe@skIl8L-hA}j>m+uJiX%~g9=(y1o&90?~WQA8a zSkEKjnr&4%Y)n*{;vd^!s>HPgD;t4SCasVmK9PIKciX#_)vrCB_ig>@?3*uP3T& z_J?`8V_Ko{`_Wu=xwuyCgna)X^xY&ikJao37!MzHY!vo*Q;4g#KJ>sh%=@QCiY?wo z{uiWn?_Fl&%|Vl1ef}9PuqIzS_C6)mDbs$$5_61qIEdh9n+&<_?BUN)7@hl4&#nv5 z?I%|Z3J-#XQ|*46UBBZNm*x@v>MI4O)JQ@epNhX(JQ0O)Co&v@R6g z_ni&TnS5cvpNc6f2~p_aOv-?k-8h$uk9*$0o{ zvy?FZxuR1bc#t`;Ss!B4=cpOEwelQ!1D5;~+~>oMwj3*+U!9XM{ChlxNbTnszgPS$ z%V-nfTFKG+FXC4CX0v3`=*K^F$bs_)5wMXKfA@z~dpm9Wu|c4D={i^qP4HO%I_g8P z?~r=OFwD`W!s|asW4O?yuBc6hkJ_ypG5n*?WbSJ9{d?h6^Hb2`)?%jNQ80(|rI{HG z(uK`Y_FL2c@A=`AoZr8V2sVQ-bOoOV?i|TYvX<$?ULTPhX0^q&w;G9Gt<$Hh_O-G! zP=CWw4hzj7`UxSje*7O;H`<)4Vf!?9*c2wMha))H(}oo}5mg>R_OUZ|x3!`>5bbluBnq zJSYO~g$E>t?Yt`J6Rnv6oA^fipG4N6eQS=>%5PrSvIae6Oy1wxga8Le*RO`dk2x<4M*i6KpLizUHaM4Eql0V^FS#)j*4iBGI;l1SfExcB5J3TO!NAf z_O48D6NiKY;@NwOx6+7e*l^aEcS$V94rZQ5?A^>rAZ<#H@$x^~>)hM=)yucN)MXVo z#|fVZ;riW*W-MlMO@0bL27$`(dT1%*ryg!b0>H<5IYD) z2g;87ih4CkXH`u16f6u-tPXF_Y^|X|aN5Tg!OVYuinrdHP=#VEJsspl?2L5{ ztE_l7ccYNURhni%c6S|*#7>$;=@5wb1igyNk=CmLLGc(Cx4Re`yECG?UdV}F*Dx9` zgSnMn-v@JGPuZj1+t2|v5hkE*HYIva&=V5>^HTb=p+&d z=fDrW&A~scvn5wWI=5X*K+1aGDSTS}bxI+|iZlN;S6a%U(g!rUQM}`F|Eq|cE7d8+ zZVQRWd&5KR=V5i*p?bu6Ib2MK2FK=@&H=ANLHl%nzX~4Z>KIR459=rOtYBat7P5bn zDff6fz;{Sp61kK>V6&}3*5Da=^jgAIAm1v8DSw%;(4*#HWCdkio)sO_nU!I9Y*BWK z5EN%K6}A4U{6PE7p$Ts2ssqu=dA?6_BZ3L=eXI2QWj}wm8z*;uhpP!)HXV(0d}vkZ z9AbjGswz&cSX#Ak*@x|m#+|ihJRCcu+j&z%8hzdRdEI~*idTmA*n@h&5Y@@l*6xHE zOJ_9SA|ZVq2G+;K*DN0s%@PxYg_g)DGa`qR4=TB zKQQ$jA|A(pCT*_jCMd|@q+{kx(Jx>L#ds)sox&*t`H))6YkWQ-VDqaR^Gq(?GyEdv z*NeDIbc{CcL)zrkk9v^)p6%@2MjU`T>}cPMwIdb()Y>AxrS8m()1O)!PVgKWu*_J@bTq$qa>pF2&pL%dj6T&&vmRZ0BrFp(g=4_bozooS zl6R(L3c}bQZdokP2Z3?a<^GP0RtKT0%gC9pTed-}=V9PsK=QU zRtz~2@B~$f!6G%-5#`k7{FEmcm$R98K$8P2PGfD7eA2Q;K5|-qEFsdwjv3b0V#wvk z=H=P0dz)ABaR<;P9NN!QBk;WT3XgpD&XATq2M(a$lke|{ZzdVt5p!lrSTeo~m)+YG zD_$rpv>f(RB{|V^xSFgm()kuIgC@GI`9o)UYa1ReKJnr)Y6wpm4rSId{k@xvOoB2; zlK?wscUjo@(bZ=+`iGI@?ki0up;dGO!9&B!-_LNBG5g~&S>+|ca2To#SsODZ2j(OlvffE;7kPI{Sth8Ih*q1CoU2 zLGGCKv(3Nr79TFV#h?e?PX8KKJcwY2?7G3D=q77oVNDyLPvJ1EJ0*`m46VrCr}|C0 zLP>;*M`^pbR8lsOCZoY&v$Cu9r75dfyxPoCcoZ8rnv!26FY>k|oFGXKW%siM;Z!o% z1yrPam12n#I2xX5)}!4E%fEZ^l@E$6w{8Yp-0_gUY%+dZ?2a`4MrFI5G6?Y#HwJPm z`yjs^61%nfJq)BWK_JrU;i>35%IZn_(Lzq;%0lMe_?umcWG z31Jo~b%<(LAMCEVz3jWHIS|GYve?K}cn^6u``G>6VMqJtpSU6^hoaP#c=YYGLr(K{DYyerXh z=xfkluAWJ&HEN^9KJYn#+^Bg1OvxqU=6t)?8#uV4F#Gpi5W9z^t_I7A1-;6+B|t)e zC4OoeMF%Zer&6qHy49$Q#dNd9JBE7C=6Cn?FskT2{;gkTv}~ANpK5fJfUGY5!^=qn z<}BUx!6)zfC-PwgfQtT@zXM=YeF-HI%7uQa&Y;K_>(gc*&R^jMi9-(2ZS;XeYCw`H z;{)kp>}{%Q1~o@3Ahy)0Q=AqT@C z6?VN%bd`kJrUmyK{!Ani^>EDsk(6a_*O22fnNL+924haEzfjG4J+1n!k0E6hwO1<@Xi>qC^qMMpqmT1V@C3Qut#uM=#rPG_ zd~wXJ)NjS>A7;)#;^$sWnHMe&D=HjcdDT@Qz_$C$zs}(ocdtZ}N*aG5y2(~m45(V- z#ope;3=NAa_cUtdXDZCIk+np9B%Zu5m^omzMVbmQz|9*m^CEqwzz(pyu$J2|7vndN}#MMpV(S#jsgnGUEm~ z+>UT+lA}C#v9O|Cy~S#1wV7!OVP%89)%k>0myu#S`Ee{}B=!cNx_o*l-pPk+I zlQ5JpGLzqAbm^^FMS$I114Xe#H6jcY!lANi1pW=`qsyG&i3j}w#6fl&R5ZDUDKq~i z}Cw|S=VE*eDos_=nD0R4Pnu(pp@TL3$ zHFtu4wwfkd>bUH^V*n=5@yF_KKz1Keye}SwA65px`k*5odDr z?Zmp~-}{kmYcjFACfgX|V!Frb_}%vhS@m|>M}q^FORkcXAt7>ru8AMa2y5J+^BVTb z*DTOi%_?5>Zh?7mszNKZc=Jvp=1)S4u}?gcfS!^%6XYe$^L>^c1mE1pvSSfY>`HF? zNZ41IzbLe8Jk2#5i#X4{`kr}Up(nLkxId5i;FC#1cq$<)`n)S1GZ=VWcRO?2dj->Y zV*)W6QLslanEnAky2$Ni6qRJYMA6~%E0tb_wu|_eg!dG>)+xWesb`_6I z_L0asH?zORLYMNS?>Fbh4E*@z#GtbDxwCH&^!~-ZeC3ZpL=HTW=x+5Tuvt+pzPJ-U zYE@HDy(BYU8_n)gYMTFo?h}X)z~mnx!!7KZ_0Jsg&R0r3g9?v;&Gs7N!y!E6dt1yk z03=whgmkNpDE)YRsD&bKhztqKvn)xCoi+2wp16H5h%#`9R%X$5{+1)rd z{Rll5slnqE5-I(U%=_*Nsb3w*vtS%*M1R_{hP4;COV+~C7+{eXAD6pwVDNKmAhW`D z$;$*p3TxJ0noF0L%w9J0EG@c4W>fCIAK-(I)#cb~b%K#S$feJ{V$CAasJ_+j;x~?8RaZ4p09s-~OOKJ4GKzW5 z^$sZEhcr3eA7{pdbrlD?mU*+Hg2o{Nj2?tIq7Bs%S^2<^v%W}>C{aSZ8gF36NR@0O z?;cZlPN4y z`<;yEpd;%pLY2h1jovzy#J@_K_g%Y;y|$5)3|Bs`m-g_O`c-7rx6a8o?FDzPnIvQG zIqPneIs3BN>y{-m`K`QVx>2Ctm`^$&)|^XF7mj#zI9mYXx8_*>rE15v3X?tzlmxT8 zCwfzekCi-~WedEs1Ie4MN1XJKDH8`#q-ru7ro3+O&RW+!;gO5c%jZizOs%r|hxYV_ z&waWw9SQ>G2PfL&Nbd+=kk~G~FS6hYRJVAhVIn5Biq3a}hT*y+kiapsT#{)orGCLr z;f(z|7#9K%(vj$c6`ltj(a9o~CVc+O^!RvSq8=0mf9nta=dbPP3BHMqE5dQbRz54O z2qrs!tJnzG5Ek{rU*8YjFoS$WMwr5u%+xG|iu-yRmpwV>=p>OXsYcXD zh2mReao?v#qFB>UEw#9{D6CAk39= z4Ju<7VWb9Rj#j)23>(FHgIKl$S0jo)mrCjx7%V17sBbzF+o+eC0iYicO5FX(mwBs^O;u8-Y2_)0ahVr9k=wAh*IHsJ)ZTi7t0@ zS`?i&fS?;92wUfg8y)mGInvy1WwL-tb<2akJ;uT+LdKGVA=2u#(D0iw1>7WVM)AiF zdT|{5Zs?XMCwWQD_DpO>;QrQ*5i9jPPfR^=fxGk&6tnNM+G_enS=}I@dr5k2OOuO% zjQSntu6OiR*x);5I4f?5BW?HdY2c!SrA~`8(}n%;<|3=$V!OxF1JO+n%!=jWSmlci zDZV(5mz+QJcPRIf!!3UjkFVXfMW+_+J7$?UlD!?M)*~FX7`@>V%>KgBzu_T+3hr^3 zubx=1U{I!8U%v`=W}pf%KlMybnblBcwl)z;gOUaoE%!l~b)@S^4;1&st7eN!(G8_K zXFI^E3+q$SA09`l@VggAgJ!3mh@SyMm+HT6ARefJZMdV1wSib#)_C~@pL_K!HF6`8_i#k>)(9j`cx$n6qD3lK39rM0q1co# z3-wD)K!=W2%I9q%Rs#O>R8^rr+H`V57}|0Y6ykPM{qf;&c0cCP|9vO~Fxz>M)Zioj z(yQb%E+pV2^kY?58ki4I^JxTLi=-~RBgN9w62c=*W{dd#UPjOCZ(lN09UZH!E zV;9Sp_FH?ay}Rj=R{AnG#oKg8N8WAB*!G}1$P+07meEd*bluPu0-kGjI#sfGyE$-> z$^J~{oLf9{!HCDoRQ!>&85k14wo#yl(#O>`9p%o}5tm-?pc1JroDse?d1rEE5t@@} zSxygj^7SLOEy_%EgZTx?y1o3iBkyb(;AL<{cIZi(bKy67Q~ewWcpbW}W;r0#g=8r~ zaP!r_`(Bf0?`&>ndK(9kiedF`LqO{nczDdHdE)M}U1CJ#(DwLB58zrJtlKjoqTW)8 zWABk6z|4fIk&MdWp+}56)I(fuvZ{*$?5+{9i^TFvX=55vYnF-I-;QZPFXrkHlRgaF zy*z|vd++1l5TdYBn<56P+4!IrW6tF>zr2@P@`}Co+%dn6sc}K)|9%ciNt)#9{dM91 zm(I`S#4A3Fv&R=&E z&F^Ju-UM1I3B2KN6`6ak*45*brB61&v8Z%XgSUx;Wj*krR8lt^? z319DF=BS0b`qYs3I%QVwtJ-qJkK!H0O$%BVp3?KPDasD&nT;)$u{=Zv$d9x4-3e9$ zA$A}myM?s0QPs@+;;j%t5ze$jhiTLSz%pFbn$M_^hy1TK=aRpFL$zAg91A;g4GG#q z-C>R>0*Z&Hbdu3ez%L>?@UQOrF&h)Om-fs3K$NOUx^sw@*+`Yz9wGp^FVcZ@-kbiy zH(bKT0@%tCHa82{YH8O!$Ne`O!2yWNiJs_Co>@hbcJy8vB_5J>;nqKR(G$4HbX=4q zut(kkOr4!AM#30~(McH{cm;Df`oDKw=RtCJ)EA454GKG@kg&HRp9Z9ATuX&b>{EAn9_|YZTF%+F=*4FG z!|Y<^MICybu}vm#dIYYfy5cd+^H(G3zj%bJYcl?_g#WtC`C$j1-74ZBtQU>kLdcvk z@B1?d#IS7(UZ(3PwI~Wi`je$I?>+Gi(j{Jzdh0&z&@{+A9>DCFZbeb=lOaa%=9IKs zqRuPvpgjiLJI<}myJGtwq(nZ6z8We;i77jfY#Fl-jj*$*kpLPG02bpImBW&%UrE)z ztf7)u2k(hdj1@B||BdJpKR>jXG)A`~5#885G%ko}Ag~B*jH}ML4|vFz^ihFQ(C9zC zvB#GzxZCHT!6Q-gI~QKxYMRv|-6ZKF#adh9Uxn#8NZ7K1zZpP?+IrR8AChKPDWf*y z_1{OQsSTnxmLG(v3mS(1T#sU$AYJlU9Ke3TuNx*#+`uO^h>)Rok|zvTVZZkcU6aHz-^=3Jm8ue3al}i)uBpAiJ%99wpw%81SLON#!mW zJ*#Kk4$eZS(&6pl<{DK9NGCsWv0aRp|G{4s1E6R6LSa6 zxcaG(quw0^{k0-Xk3kW6l{q)J$7gmA!o@f8q>d}#C5Unvs5oATrsGQ&MWV%jM{oRz zLqub}gOz+>nd-J${2dx=kg(D4@toC<;AvL>Q9c2H8{y_Qp5vDeGJp;W9Y$J6nPd_v z0SybL%-PTM+1VfqMFQdGVI_<;C{v7#rjA2k<>`=kTG4=jUNk6syCqIsd|jqYgW`#* z$lYvax#GGip0` z6*Z>t_xR#7_Yd(az;E>I1646eaHI?6M%9>0)NgClW9RWNB7CoB;B#1t0efALg>O<50Cu{qpQShf-zxLt2Kp2(I8c#> zu?XlNwCn=HGBJ5PNX6moY!ChT9d|;A2f8as&o@pq1SVS&@W+=zstVyKyN>S{3lXZ? zWOo8cNkq0eLc77m*hk zP0B?HjhL#Sn?JvZ_M}~6O#XPV$>7uMT?T_8WImd%(XPyGe8Ze!_SoTmB!8I9@dOJv zBb!<{2+WowkE{JYvJ3}uu-g-$0z&8x@1}f`bfEZK%{L&i`d~19(Alw+A=&^5qd?WR z@Aj6$8G!@kF#=%j^A|7V;kE&OfsP7$1ad)Y)9&xi`aV-B{^bBxk}hB?gTu6#0@7b(%xF+L7I8<% zi9k3b_Wkfz;oYs{tV$ z>4bvI`6{=eF;&mo8W9%l+B!x0j$xc%flF+1y;W}hX;CGGeiuxUx%K^@S*Q}m6bSyATbsb0>pn~#a=+oxYwSMGbI@k_CVP%vS;rF=pFyB{R|3RWf(y-%|OWM@NMT`c|o&?ZNyIvP6%E*LS5wH{KdJv zL7r8V+xS0^M^I?yxpgl$w

)e3ofn(q|-R;|3Z~AOWcfx_Tlc(_NyKPcfG-W{~#9 zXm;O;w&@$+?$2IX!sN|u`?lf+iV$5RK+?_toGTys1Kazp^^0Pc>7fd_|D+b}xL$Ej zy0s&P`U%$Z4RQX+kHtX9HoV}#p1%cwGTGKhCkPp7f(41$E(Qv8&+Mbjn*ukAOBWUN z&!=P}EOdy)m{1aez$&17SqymP58D8NzdR?}`i%oO?~NEDmqo!d%O$zvnQ`nptyuy87$^E0GGB!rPF!7mNRely zozBBRMUpGEXOUT#Nml$EUS&?*YyWqK=Or|_)%mCyLKJTkk$!3~hZJr-Op*Pmos~;j zZFP+U86QXDBik_$tZ!`)9X#o4Ulwu>+5XA3j89cI->;55(>xXp%Y^34{e2BXXN=wp zqP4+~1r)l!*bZz)xsNl4YTV>uDI%dqIL_?z>0PYh-TW*5*KEpT}_-*wgJQQk+Q z4_eJ{XVhT#Q#X-yO?bSf;|_D7%Zv@~x#rLrc1)crjlH>;nqh*jb74dKM%l;}!dr15 zhWUi%y&52G(yK*0V-M=1G6{@%u1-MeFm3X4IOmXqDHTC^nM83_VL`59kF*aqkJCmz zaLvO;PDM4(**|9nDOa_-JH3ihoNV-kDG;Hp^pz!y-&}aXl5Om%2Td%YtbRQP;_N7| z3OFPf`280ro}Q%s;7XLgIeS|cLd&ktliJ`a^z5+-sbHwO6;>tqG915oV=Kl77e~CI zP6jpZa*!(0S)ukhvPQ0#SbKw`Twzq=CNllgflnmJ zPEOoeyzV+j8L^TeupH6LI@=whr=?qy>Gd;xX9b-DzbtM+-1uUKFsLXn{?`0ox!lE^+{lSH!@#J zZX5dawzj-z24-i`;5h#~&Vz54x`>t*NpvHrMl9*Y|2X3ufx|uEX8MD+Ib7d z*GDikHZj}>9iu7dTEp6GY+#DyT+0MrOIKIt*rTEv&IuSHr1|w3H$q5A;E!VrCC?1r zO~D1armpV3PX(+f;FbT_xd^L&03{C;7`@Dk7f=m^ec0>Y5CjuGKeTqd4cF9K*YInh zr!1VC!;?!(Lp|_kN#D%)OX*y$1rtSJWo>Dse|vZYTHo3ZBHlz%AB2*ttrw>Thz0%y z2qnLyzJ{}}A2%O<3BtTwqP!BgSUH6ydD%Q;2d1i>i;If`K!D-E>HH-BM zp9D-;Qc(JHTE;c^lhX?hDxmPa4Je!S+RJkWhjLMH!83tUJ>j1P(g$w?&e5sBAXV|8<_`>X=?A3+9lJTfwMb#pUouy+Aw;^fd`zwg8KB29=er#J6+5#gJB3VK&(53KIN2}FBK zoKnY*;#Sw_Wp)h$F!Ls;**OJsw0|-FAyC;|<8IsZ0@N;VP3%m4%`Q!A<7Kd3)nsc2 zFcSb6Ubevit;-{my z>6ZQ9YTa2u(T89OGgEa2>W&IH0pxMN@OP8y*w~=OT>G;Cc;^5y%q@)#&tJD~0H}G8 zp90B8fOv~zLem3KCY4rzG#!9-*>BxMVC1$Cqq(&XyeDr=k35F|b?il_?pA8_Vs-|E zvBA;lC+=1!s+8?T00RREV^xoS?-TYX?DlgHPzTOvB~e*1Ry7ZZ)Bb@@tYdd!Y;0%+ zs&jAzj?T%6;oaBRVeK0khq^QMNo&BGJ7h=zO3%2$6S$=U5>tz3=-cYY`k58%VK+nn zH2T0efMARQ{BmdnN+$#SxdJlH=$-e<{n|a_mD{%)0@YXjLOAfKE&O6P1ZAxHiP^EL z{Qh;{qsH`u)2+tzj&l!^E&~YS;Mke}6#eAROgZk3S?4FD@yB{Yc-vX|vKxERo!+`2 zxN*Zr2l(F}(*bNp-eY<2=mK5<;qI;9zX3kjEB_`C3?C8Ren?yXJx*W$gjU}U0TeQx zbJ_;^_qZMY0fBC&GykSow0}T30fhOR{2mQGD4^eXcRSEWKeHFNB02!z-H48F&>j>Q z!06mghQ|*$K&y1Z@8wbNGVIuH`)4=mlA#*H(X|fPQX`|gw+Vn$s1NJd@-S!Ov5QxI z`||pa5G)1g!BGNqy#QnAC{A~Grbq6pL?k!}4E`yPIziRdS6e-R6swO~w=q4h%kRK% z9Dv3i5aArCIdpH=p8f=ZdCmf+_60=(zY2ut_@uLH09`R1&ad`tzGYeePbRZ4SW4}? z=65Lwm4Kiwmb@Ol1+`CNn+XBh&DEIHp^0&;-)IRsM*QIXAjVMh^VTxtZ~)kl5(bkT z93C|e@fnW?PuM{-QM|0_w@d{}kI2;rIMZBqedU~^bxF{fd6n&nm z2AD6;sSJLW#m&Q2G^(Tvz>65`Q;KuDyO#c5e`Ek0nu*?&TDoG0-Oza4t_(t}1@gGb z6K`r3v+IX|h2!^r({MT=a*pR~uHoB&REO+*ktr3!SD*=&+&00#8egp4q!K z+JNOC2X|{xF_P1>D0>5EftI0^`FcTPbUcP?-Dtpyud5M^_=ZeHyOE=em3<#~3;esPKB;$*4#?&DI(hTE56A7fu1tYNz5 z6{+w;7x6@y03z18?&{|9F_(M>KBH!{-D@-Ei6mJs599w7t&QlGfYXXbfU?Bg+KPXj z?(N=JjHB9VcK#Tyr`Lb+_@2|@^U;8pea)#UdY8HcbF>iLglWKJVSjkDx#jnA02DM#NTHNwV+pQuAO1-potk`Tye}cDm|q*cy=VI6 zI~vb9U3&6M*#rs92}%)eu6%x3ua8ssi2?PQ@H#)w9uk^f5ukLZO0O2lz2GO6BEBHP zxN_D_>eSABc;&OO+!+rbiKp-@pUiFCS0a52T@RY@i<>qt%$!`{j^$jNdERipgY7~M z1BZAfZ5zFbW=&1X;H`?1)dL+&&^PukBZ}ETvv?3mG|*MCiD&&$(Qd@&Gv24&=PC&d zQd>B+)jlHvRypI?^=QQ@W!SdLT99bhTuC ziA`m^fkVmDopQr9)D3iiJQ6Ou1P$|C!LGFe(+U&4Kyl@qf23ULr6q-TK2TzDnZ`Q~ z%XXD{SgVmJtXZAbns+4Pg$;OTY2R1Uns2~TJD`0TEKCiP6tU-fU#HdLeG5OmrF6LQ z4|W^7b3S_#)tYCWE5V!qMDAnlf{ioE!*Ht6ndap_#ZR00-iUH&J=Dn8x`F6W#W(+qFpk)d-;t- zR0XqzkbOEYU-3nv=w3RUN0(PlNFg8FwihF9Kl8fd0gq{J?*K(n*+x-B^dAkfH7kgi zrMqW&6S1dHMmt6+L_Q>w?dM`q%3U%Z3ITym9^zg~2|2Bv;) zd=z!LOi2g#uI@mRI^_9adQg=zuXkxYp7j1~x3g2WBj6;%)(yM$lg4@fEnFaz)Y0zt zvi>X}VrRv3<{N}(j`N9B)$u-nm4PaDOOz0T#RnZpFd zAJqI8|A&r$il{+HQE9$Wb^Qow&u}PLb{%gz&fQ!hAQ0E%EA>tWeizl8Ud#6Q&tA%T z&XS;3Qi8w#k>7e&&ir;^`8o!xWXqG+&`L6ZP;>NKnGsb*JlceiqVNaGp~zh}5hg{U zlBXr4-CvqVq5NiUwuSwoE2h4Eahs%nb59U_j2O+%ti_S#VHUq znM;^3Rh@gBdilb3ZOC~*mSac6v}*maLV5Q=A2aq$mHu7H4EnyIt_)#F0;uU1jkro? zG1bL$$qP!F2wCmvlnd4j(6D%pt$c(3HYeJ4`wvsbC>n55>RqD8gCpgdm_dsK&2}!r zJDMnwoXe^+g0RC;kh}f{L30qDrs%(#{uTlcKQ+q27)UG`FfJ}Z8^4*^~8mVBs) z2F!sER(9w@JlJquLeJuh`?5sL?Bt0_8t-I1b#=EELImqzO5p(bUA^a z^tYqGZBi(*1t!bRGY?<*ytIJ3|Idj*?Ui9dO`E<-hk zDczsvXCqen^V}KxOHet!jB{bWc63I!hp~fUZGnvB7E}VcAZ!vD9#-p)zi~VJGq-5n zF#Uw1i|BaN1}+jZr;=_H6i-sYH!3HkwB)J>z0G}`M=I>KHFFEU)F~ZsBE4t&v+(l+ zI9CqFoti^*YxLJim>C6QBb#aOs`9KAAG^+=7qdn=R4Zr>%5=7;h)Pd62t-c9d}A^h z6o~08f170z{jqIdz!_X+dHX)r_S9V-T2z#{;+!W+qSMpta0>UDUt&+{w<+_hCcNT5vbN0hNY z^{aQWf<=1K3XR3z>)1k#8I>w_*d&CWBNr=Zc}Bt~Q@&h4|1occSk)^8>u=uoR=u{^ zgCQJSu;GEHj#g=&Ceh`n9+~Dm?3`nI^+PILr(W{%W721H6l=_W$`;WvqHVvDhLvS- z-YTYnB_2Mb@P@xw9~{UjVAX0moWJZ3aQ?jE)y+2xEvJj_2C**| zTv`c~V}qldd|8*XS5axiu3s5@3xRbJROb>G!-J}j5;^yV=nRlxA_*-9zPYJTZk=C$ zmqD$xRy_x)%_i2T4J&$Dywz-H;3#*gm>eyk<`F#CGnyZI!#;?JmHntuYtFOxbB~av zgi-$ej<(&AjpL~OG)0J6V#H)4eWJnD3M1lMFgT!r?>?TE7~VOyfLy82t{@@&(Wlut zpCXqZO>I5#C#rhFpEW)|@^5&hV)SsmpC@Gqxk4*~~Vw(y3CfAKn z>(`IrxjXLfZ!H1cyhDo0)0WwJIAbB;3et0aMtzK~(W|<|aH!r`a$dz#KOZWXM5fsm+SOLl zV&gXovW5TZ8lR^wS&fwp{qZx@PC0`({cW27ZUd`jaYbe zd*mUIp#?Ex!&)we{2mgt3bh?wyhWpLsh|@H`0|o_jk$EGIK?><)rkvRUzCUG{$xg& zxDu^a{nQ+L=T+EbX*Mnwbc^zT>1)>XD2Wmq4_DZljzDlofNe3P-Uz%E--!59)63VA z@lLJP4t61$cAbC;CIB8NGr-hZz=GJ?@)cZ1;a{$Ulc&I-Av;>wwgE@ofAOF+$|2s7 z-#@#8bs} zgxTpJ3@ayUR(_0`LMS~D1wBw(k!p^Hs{5_va9Li zaiSP7T=cg~Jei8y*Gvjl*%(JTScN-|f{Yt5i z_OX{$@D+V!^5L{{byj@CN0`Usmkfe}Xa_~oF;77+4~e1rV&*_{uJWV5ME1u@KRrxS ziNigrv%i`{%oY{gZN`{&`2cG;QFL6+D!@hDUc@kRw5fhE)xw`W7kyOwMIH+24%RZJpI) zpJDy}M&;)arnF%*V*QX&F1A4@lf{Siz1?MXy0URu>A>A}rb>-I9a{OA<}~rX`{ywW z242~@m@Y?9<2T=PbKny%WB;L2-l%y0nw2`9=UcK}nZo*8QAM_&2)1S#yk-c><6Wc- zoHR#zxcQikNV_$?C9(yq9aOW0ms6ipC@76u+`rJ=U))?;C0%@lq2fR7y)7#mO&<8h z#1k9hIyV9VO%Iy4sF57d<^-h|I%|FijJ!Wb;*w$-vf1lv;Wx-^S(1^j?6ZW#chUw0 zX@r8{#Kwjl>5S=$;aJ`riEe4-|#Q)U}TeDPOURLe#lZemiCiqLj9wj8q?emG9KdxzU zIfsjO-B`31A?bjE<%6J=l&@q#Ac7J*C0FT&FRhaEi->cv#16ehy+&Q)dw-;CQxr$F zGkc^L>fdJ>MS@whjulQ#EffTbg2MHiPh!!tx+zTqf5VadIjTzPaE>Rn%qI-~N`(#` zDWNel6>vmi{27ex+4-8=J$U9^qCODE48r^ynYwkO@EB}U9`w|dQ-;=`kadoTJjU_HKUT<)0$r~sP=x## z4-*`i#`*A)P6=w!nVLJhSa~y2vN9{ou4cP;so;}e*{lu9c;BWYnl?cR5JK|BH5=KTKUaKNv0&%C`on) zM@K73xy-nZWrFCZM6;)PKg`|8x*SaxY8FNB);2v{a`nYa>FgziA?twj-w4Rlgtz8HbK% z8>w`_OBkUHB>e};KiRskxns3b4ogcd6~E>bcWXyTW4sf0RCd+@#GvN4b4KF6CDBYMO%%?{myyhNF$>s$vmZ2o<#CL1K@Z#?JK zgIH-mZh6QSBf5r;4xY~LNoeNGi6Z>Zg%SZz+nAkmrA{hp)Z~uC$Qyv;FQ$cg47{?4nF~GUsfyatPza=Wq55Ps zb<(9c!K{^0oHD;@2a2PJ-_7euzqRw^c3jnVi~q81I`V~tPvV92w#8C`o->X`&lWyYu$5GLZ%sHZ z<8oBvUszd%ee?!zTWGRXD;I)|h~Rva!#{VvN*g!!|CNP~3Z9Y^L1wwq;<2eHLIR^& z&iNsaAc+|Z3s81nx+Zn*2j||a=v@(tM-yTyHTA5;3{kEqzV%$}O7}PI!=#>p{fW&9 zzuN=mx3&&fmY2F>H4d~Od{dbRHOmi)V)0K|7QK7ioB^gio$oo6^G6CZiazljl}2Zu z2mkP_AAS7}w?5y8IftYWu!zM6qK{trE^t!;HI2`HIWM$hnQ4mPD2smWx?6CuU1P4t^%X#F8@GrvB6+T9JP}IEX43gs88qI11*_uDjWmJ z{p*!SRUSk_9K6$Kp3g+`wCHuOo&q*~i2hf_r^_0f+i&}PB>OUfs~ASRfM8~ZqAkiG za5APs*;;f1A%3 zqhEyRiG%|9vxCdILxz_LZRIi=#*BwgsmiV;B;vZZ@kD;zDp|$rG-i-@%~+OM-Cujd zrk7S_)sP+4W}Qh(xL4RP?Y@x-w$Mpaauh~=?*Dq*gv&A0W&hYa)`ui?Rl5U=f*B}U zW4b(iXV@BFKIdg{5y>as(v&~_3*&(ItL;7l)awto2H`c*4H%7|2RHC7I4EC3RYaQM z3kf>R^s(>&Yv@*s@F=Ez<^Jk>@fcS{239dHg|@H`%MTMhnwwP_>wd;^Z--GP!IHfU zL`mH%%a?^WVoT0du_v;Jt?W7#%$pb4n|71!LqSn5QBE+fPEaHsI~Uo*d0*5hA2@^1 z2&A;S9C5%csTR`Ik=Gtz)hhNOR;f^?ZBcskkVfTo+N$^v`Q7XT`-xtR7H`X>5Z^>O zXzNCqD*4VMpjRrS!=CLGG5Ogyh41 zhiwsoY@s)UBVr91dTSAGR;G)oGh_bY!YPFc!d7>b72zzOM&R!UAt^91DsR2R} z9iC5CgTt`^D$IIbD?6) zTP^zXK!FfxvN@xsBXYh&PtETQHf4!$K#Oh19s}_*Ot){(ppWxK-07jJ9TmyT4imH2 zr-NF!b2@-{RVn3up6V3Rgk?kdrQ1oxSqK?x^&dafg(H_A88Nn;I^k>~`>u&`3^x%^~u$d!n4sVM5NO1X)OMUtOshC=RK6&rz1mtOTjh8kJ65)z zjIvGeXz~k*f2mOgCNNBsW_fffopFNZtJp)=kE=+l_X&r60KW_^LP#ppD5UOeFC2oV zazsVGr(I9VU!Y-WBJR>SeG84xCl8t4VD#?YUs+C4qeg{v>*mM6hkU$zLzp;*He2yN zEym7H%z&koO;t;s%sQy{aE)`a0aFmSr5=;~d5@Bt^Qi^NBUrqn!e}znzYy8hOoYc# zCZ|`2G)C2OI4^W7=ac(+Qi*c82xg!qIFWq%xgy5Tayf*8mvv(y;#4wC*y)V7W_M0d zbvzejYHklkYhwk^(v~<{5yc19sRe8a11SeV6ANllV(kMliE7@hp3~ZvC zw9n3uaQH*&s~qTE)9I6;kG4RKAUa&B9xe{Vd5Jx^m>L{EyMgBH8OVKd1FbPjnS9_w z30sHU$KUk4*8#4@i@Lp^PpJcYa`jDy=bj>iArbEI(=)iz?)$DzJQ=?OBc9y@R({M$ z(@PmY2D0q&<$W%ZqZRRNctWBv$?WthnyamtSht)FE!dq)w4{%@)fd)@5xpTHYJp!$ zD&(N(_@Iiwla%nK@wZBO2V6(yJeZP%;g%f}*F7B1PO-1D*-u>HFh-er1-vO~vCy2VLeT@#@fsr#9BygyK*}ihwJ^MziF?Gg7JPA zS*=xsLt5$iq%nm7x{k5nRM}UEDq_HY%??CpJbh$AMnFqa;&7lYVxOiOH+}gjy@R@A z++%-{LvB)cK19YrZ512bCc4J@gL4J}U`jQ32%gxqYq+=%iyu@#1ZN4MF781;^X*AL zfc_eKwL;!d!(iCn4bfjBw=!EfBKI=XOPJVcFC81%5oQCUFwzkH!utxFgRHw;n+WCC z#7M5O!4_7%BaY`1h1AEkJ-r4I8mfL;nbxFmMMbTP-wQmpkn&OO264L9>*F8a9#IpAZc(*kTS2O{b3YtMtjil%z z=c)K-N7I4>YV+|L_m;n~qSsw$E>6jqre>-2AFj~5of?xm`Nzwj6+g#fG+nQo?t1KdA_Vg2-Sykm7X&N``mECDN#~o-Z_^ZY{tAzN9#Ya~$pRXQ&PQUt z*as~4TaXDouhCr7C==~~;5~?$%OO4R=6nXNRCZ_FnI)SQVCML3MN4z|N5$K%{X zpBpc1OqiZm$M|Wds0~mbG!m~aFhbEs^l*hn{mpe$IH?iV+~fGpVbd=M#?l^Vp3B|s zkDF%l1t&A+EtC*pQDbI0>!UgBjzg@hNHPyAb8Z%T$^%-$Byt}Y@5sO^Uoc_SF7d*> z6SXl^LD7!&6Co3{Mu{qv5gRT>(55D<6p{u{l^aPAyZ2NdcF;rI@LhsA^7J)_Qn&X| z!q5^LEk#uaUi#|rZBKoRCBLAFA$5LG_|)iuvqE}7%F>kg!;t>* zP~?|W79sT=fiesK6MLuf@fE+YM8WMk?QhddJZ0YH22UrpyM=SYN3|Bi3+mSl(R?{3IM6>O&-!Md{ONDT!EG# zd))H%+2S+QG%m~U<-6#WZ?45=TXfQcO)E_pGdY`l49bIEu8HlF4%7+d!9uf3mk~Tw zWb}pRE*lk$FQpHqraD6<%p$>R;<2QL=cc8KCy3G^%Nk9-*ls`PK6eRpal4W|#4(UqautCGGj9#FPJ>85ok*g(7SZ_hPSW?$i zIdQkNw)VE18M5P>SqTc|N-Y)%=@z{njkUm#WD;`Rkz1OTxC!r7Zg`1aylMz|pVGN&iM#S$RuJb8zrEX!Psav#${R zy5BhEUjf8Sw}A4uvX;wzYHq=5X+gxZmFSBc|6#$!d$r>G9dpg~a4vI!^~=RvmYJ>4 z6T!NMO|KtB8KDv`IOiy~4{x7nVbc8SC=!9K@}WZsvK=7|v1#S01!-jdvLDwYGQP|E zD$VpAKzZ-q^y5D!bqCIJTLt{~Xi<>DBI@=a$I9Yf3x0-!tAJ$Oc;?bbX81wO zntQ!pO20WzVmM2iz9fvG5YE`%BowpRL(rFwUp`3@<^->g^co^@Q|MDouiV@^S*HWQK8uwfojgO{LRHhu# zIxBFUu0>kvijS|()tAZAU*Atd6Y2YmF}Zb$`fTbq0@eV;&R}iHcDIqXt`7rqb{cBA z9r!O4AP#0RfeBiiPsqM0Az~AU^An#$D?iJUu^cw6dWw0=K|NNS0r z5T2=p5?$0ucrT$Hpi=GrKmLDGxt7~;^JK7~Rad5$$TZb$Jm&X7MWF1e?HGH=N6+|s z-UPgxXX&NJfEj#yC8|2=C4-sW3I=FO8*Jht%VKOPrBTiqlN)O-EVTkz!vT90)FTMB zW%3pqp)IJfCU$vXT^23!q?E3un;9^fL|Y1TyMS@P4efqAlg<&btb|?G_0@+r-hdk? zuy@nk=pj35qHwXl`suTggnvKF(zQ3$$%hXIYr|m; z?&znXxf-#+Wrhps(0GkG{Ndu{A80drWvs5E8kK{#knkK>SzSISG&Imxl8 zjJxI?U0^;=e7?yczv?lw@n^rvvYpOt@Um}o#nJ$_7eKD##UgXJ|v)>I%hj zT#1xR$aGRL+l#9A&Qyx3M0)By7r3$hJJedp>T%r@lfxp-F{RzvqG)VqcIqQ- zC~$e3<7MN;UBp>a8L7S->1M}m@rarPq62XNf0pAk%yeqhN5?6@aH{P)mlYth%p=`U`lTB7(OT7MvpnFdk3uJ zUQ@x)N7&bK#(mIx%Nic_{iUs_q4KP1>e4^pSJJxXz)%n9ABYX{^t(+2u#KH_@A!R? z%o3{hlTBp2{m`8k09i%YxAUZydTqv~W7QNs_>P|ff7-PW`oM#OU8Dq}#-CL3=r-%S zrZVIB;CA$y%1yLGko3Lp#2ou159)1((N|UuT4Pa9aL*5A5CJ{I%|=Zkm-X{}!J73a zoh(V!OnQ`G?aZ6n&pU~x8s?2Qeb*d>7=||Gu!lWMYKQgjR7#9B4Q?-fk!^3pcN4;0 zCQc9xik!vfxGBa`FA(l0M4%Hh#(xuUEaC49y+OZ+c5;=n!U1RQf^2TE)Q)M86spp65pRRB-C z88q48Svpd1>$F>h*aouWl84)vG2Vr9Lrxa*7rDQ3Vca}Ih>w*p*`6I8lj;#kq$xhXPq-v0&6Sdx7lso zJOrhzfNPpGGRN%?nN^1oj2J)(X!5jQ!(phMXDQ{Z#uIm9%S^@^#pj&VJe5AD;m5q| zV5cgC`Ld47A2mscP6_i;6b_n`r9HaIX*yT}DS~8Q_?B6!Wqdq4aK1NKr95w#inssH z)HT!*`NiyHR}plZ9*{j1Err;53lV5dGyw}g-=VM0oK<^U9<%a!Jfuz;`zwnMM`5$&nrbr?)SVUyCW zSiprrnaZX-XtdFUMkxQq``73a&p{ave5%$VTOP+ggt(J}2<$gA)I0JEYD`8Lr^Lhy z5z>HDnbX3+4gpn3vnu%Jtlz-~n0efC!~Jl+WmrkAmyex3iRU$C(c%n0I`+J@kQ&oD zdMbj4c9Ue@-X08?2rnG}v5S z#!Jb~O4_H14UKVx*fB8C?{lYu7bcE$4a4$KyirF8%B}oqdc(us?-p4@-Ce?C^zFNp zVB`aRawZIK0A8R$OeLI%$Yle~xb!%4MEpq8$0@?ONh(Hq#azufi=P5)XbMGF7vkM& z06jF+m4%6~wQrKe?d>C=##zidzUoVCT;#lUeY-kH!q>4uOYMl+7?hn0sJ4mJn{ee! zGitB&7joW?3KAEskc8EK3z3g87N|Ei(5Cd?kMYHVDRgG!E^mj}emPQ1k35KvT^AXi zgpDplE%w-fC%c*szHRmVOcV=EpHTsl*~lzHA$_cdp^ta*4CI_pqtz@iYqWo@n(r|Q zxjBb8n{9&`gk$8M$b{{9^2>|iaZ%jP`yqoOSdQMwWLC7I(KxfKj{ZIw za~acH;lr`^&x{Z0lo}W;9k9xK~#xsGt14{=O;;;;DEssxm5vM*^8k-!gJ*fp$7Gf zVrZM$xYubW9h8qs&Gra3w*;IHf?I49l*VZhSB|{D5F)DCqT=T)dhr8+ zgqzMgw-usf3D76pEt~dHr-Zij8^Z@DO9u>n&tqP46@nalREx{Yu;fPELwt@2Mu(8a zXS)fG)$kbHwD6jA*S3eAIaPf#GA&hQNvOU%y8StrdqSw6q0u^&Z;Ds5D1WzdoCIY& z#7%L^jkNgYm*Az*lOE`}X-DL5+3ch4lQg15rgHn+BJ<3CI{tYyRTMs@VNFAXZ71;= zcGuS{LNu*=2irJf0@|mib$e*bRW++_?bn?h2qMSuLPClKnjE-gpA_ zy$J3|cV5Nhho?#CSJn*3rUO!Ia zZ1FX`e9h{Gs%1wjq#2>B=d^#X34heCDOpP9|DXLUwY{*U5@Q)?FQgWnyQplLY6pRd z*`5<~I2u{4Yin>7Wzw-c>F9jvq?jTlKkR#s-s2C3#5MBo>V5qepY%n z>Jd%TjRr_-<`viljKw0o(W@jT3?CmbI`fd9C81dI**s~?Fg{@u6WQ2(NBIoR_b@gG z%^w!n8=9T9QU(Y+zNI>!=$^=6;&ZmNoE49rF{u0;QSPLW0e9;aTR0Qh&DyMn+$xSWk_s3<@{E|2`@Cjn zZaaMe5drf<6P83bhKp-#D=*;zDn%B78{)p0E-^Z^u8w?b-A|V+Jw$DyqR$vC-5?f- z&}Su~V2*z&{zJJDk5g)+<`S=!3qwC+21GDlfKr1{+KstF`tfOdsJ&qRXa=PCs^Z3H z(;-@18Bj|-eih+9T4ObG&~Aw40jy>Zd#IHqA}t-a{v>MuLQcI73B9<5>51d?lPkmQ zz?QIw)x9 zSeeg_;bm}BckE;?9c9HFZevvW%-n}~-3cVk&iP*br?Q6Thc$J^(Wqw^uRfeSQM>oT zd3j}z+wHm;2;{QLlIGIYqF}G|;NQBC*6Ox9quF?$Z+GGOIp|$ryJ1Y6`a+u+%QWNg zA1qo^*lLeCU8-enUV18Sy9I!`t^y9fp|XX?ILXRwq+CK$?j~PTC<)%)&rR8*}xW%S;Rf#E-q*x$SWU*E3jUp@aHA|y3`QQ{ed>k=eivwwv z;$Q`rxOPTt$IV9}+^eC~bEl4dRa4yKJ{ohhDd?TP?}5?;YHK0hruxO*PDr-bE24QI z2SyEmPY4!+Q$^xPY&|I&d8p4a*r^@~Q+S`_g*CfvOD>((VS0$e!{7v!|HwQ$2lLdOO!zZ(m}!R(zbg>4!-5 zC+B~lQYY+|RCuW+kQcJWG{Rq?QA+=eUEUwC{5rEpiX^actpqKhwR$x)X?s>^-DBl) zABhW#)=)^v4IXkQjZA=9UVef*6~Dp|pmBX}&>L5`*$=BLnXPaRmyG4r(11a@Q6d?49w&vTyk+=J<~TJGa`h z&vng4A9NEiOQd(Ks4oV7K+>alFn;7nH3EUy3JP zKfBmpTM1wU)tkgl7{FDh(|}G%$s12aU@=A_2VJ#bYTxekvoxT%c3frZwsk1pZ7L(@*R5Rl`Rj^|{1|z|1lci4{+sdA%dNgO`kyj8 zR*JpIiWhYwoT9*Vv5}uP9hV82wX_KSd0@`kI_)$5o>8s=o@tr6N_ES0QvuAmLx!)Nwk#jX!qfN^u z;qEoVuG|mBe24!AHxIdM6w3_#r26YeATiNbr&{!eEKCb=DuN?{z?bU-cKh zp3Rqc4ql@q_9Kt%dW?vM8d}qs4+JHeEq8}GraI$v(aZH;vliS?>t|o!Qp}}xTcY8v zdIso6Vk$?hL5L=Vqsh(*_Od5m$#r=$W_^H!dS;bvQ6@+eZ{=TtjPa6_^&GxDe1>QG zMNs8hcQ?tzrTD zExq;CR2@NuTG8^lbNEiWw6|BIsQ$(M$89;#?h>=PyfCuHwg^t8H1;I^dYeiR1%wsT zHYiSH?s_aOq#;+)Q{pkGTf+1;QEdp6%XnK=;56pPhc^><$4XL=5_F~koT=B%%(G!W zowpoKuqUOo-Ja;}RD_8*i$WqnLEVjb47Wj})*OT!OwPDNZA@uz?8^Tp-w~=HlURI~ z2y;V&vTubrn90#~aa}53LvJzsYd0?e@uQr3e1Xx=frhXPC6YU>#u0O%5qUVe-%;09 zIy_TG^$T7P>s3`NZQo&C46K}qC!CDySB{I5Z(%}h`LH@p=gTYYR*1!ozg0%amC6>= zZK^@dyd7W0%e8`!9$r9gR4Nd+!|3FGZm!=Lsf1y?)30896Zfh6Wt5t6VVaY66dP(% z_F!ua`xHK+3tOLU_Ek@E=Jxlrqm_wd9abgzwhC9sHC`+iOzQ+C53Ek7O?$E)!&gsc zXHk}e)cZqtRs1PqDibH8lvOtgk=kQD$C~~78&z-9>kpam)e%~98gq@TBFI*=&L?#X zgdPKZpqADm21c>S8^bkBE#27}yROBPW$rLp*Kbl-;OSaF&gMl61m-N^%RVjoMNTC> z)Q#sbA~sm`u~kmjpPY#6>4>`ZO?QK--BJhFQ9N!Ki*i?g&~TGFw!@u1)&1lNNrOv! zh3KSC25;6~)SU4;=T=tkZCbHIy4!YHptqax<%udvosZ^4vOWURND=xw-UxjD-cY`L zIieMTb+>S8$-W3lkyKVUOM2K~D0d33hIDj;venGM2KNEEm`(U383Ug6InrjK<;exK zcZai8%x&ntswDcVN6krB`Rwx>yY$9cC3kz1TOq9Ew5pla3|Fsr2ZgrrKHre%Hm@1v zH$c0Q-7bWzeh04u%`vHTuee;r1SqiO{t%OCl%DPtL)O2V>bvIa{T&@)VW|RP`j-*Vl6*qhA9Ddfk0yD62EvAiW-^7mS3U87#f8apZxS@t zQCNUV@Wto{1JN+Qj3${)d}7!WqI+biQNvJv!-Xo9%j==7FG?ZB1E{iL0T0WuKRHr@ zFSLq0sGWGQlSvru&_c!oAc7&$st}z?So95uqBKJnSI{abxVYU$HvL=@?i$lS(&qt< zm8O!`-Wp^fn%Tc({gBQrgDQZ9)#>^rK?UA-OAM&FN|?^tbSQ$8(1rHf>X+H@P3ft; zx?zAGx^&7LEVI&8t;4q=eHPt&ipowg_LihDP>sHaQAyXJx(jJwNxo;xKYlH(rEU`@ ztz$^-v~`OjfK7@iD!)dWzOpl7u*ORx<-XO|!-wB@ls6H%+ysaFN;F!rf2YChU(`EE zxBc_?EqC}rw?fS@=O>x1#AvlK=4Pq%Xw`^pgH?%Obu{e**9Ex+Axbc*W7HdM52=>= z+jj8)cd1()m5MQ;*ECF8hopX-z&iyWY$trWy}IYzq!swn*Gito%WChb<5O8UFcNnn zd7Hu)M+dpsPzjEthf%5gB7V}LYKgepm>+G{o|R1dzZfQg5-WFf4!_LMEqPHNdH~;T zAden>W#wN6)wE-zzU!3K^17nsNqh|b+IhsdR#e?JV(#t{l28X3B!nJBEsg6+)kc73 znl?3!QtO$zlkDdJ?cfKYp;w=Yjfq|xrBEPmRDR=%9}AbRP)E!x& zB+Chd6dhr) z;Gq_kzwkuJb|Q~k^BUew82>{xp&uJz+zQj3MpWsJflqh}j-WP6O##ORRyux%kKnQg zs{q_827-}kQc)m*=FX8na6VWPtA%81!JRL2Y|8wrAgPG?;iB-0xY5WK7mSVq)xKgP zQpnZ;=_;eJ)(WjCN3CkcI^^`VkAs6SjelUJER6QPupO`gpYq#hGhQKH=Ef~<=}z=? z+Rc)W1GTXF)*tmrLfkkju>gr8H=J!ybhcODSoM;cC{R2gpV*yZ?X)1i#iR0XBphxg zYRY0FrBZ;dpGre_2EWZ~nm^fWahEZDM<@hcORMTw;RaQx_N@P5#V$hoQ3G{i2|0qoJ!v40(6qDq^jLf{g6s0>a#(e> zTLLm^O7ln!#w9UNzc`g9=lz!kQsFC!U_mV52ZI+>T0RNTn^L!wrSk2=EZ1Xu4pVfGbiO<*V zG2HEe{Db*GH3C}YqCLJ#bfz*M)I?z!%)^CfJ9~-1=PP0im1%!FJk75QZ-`dX;u~O6 z7YbmNq|A?Ml{5&5&uXPqz+loBneHE6V{CLc$p?J4hMD_rV%KyZ7>9}E!_eca&+c7xSmh2tMH8M+u?=cG|C_41>rrKL;+T8c3(;P5{KlZ7Af z{=WcmB#zrUnoSt*9mztueqM+uvkmF(BZjLAgU0hM*p&cov&>a){F+zI2{c9r6R>lY zk(`wX^$2{wi%dl5gGhWSMAK9l>FHW38((0H{Ddp)`uC?GtO}~P`8L`!PBfzovcYTR zdaifHEUV$0VIXsUYuE(D{Bqz)L}EqR7`D*B(4RYZq5H5#y=I>Y+7gt0Q$#yKtkxe6 z^i69aKQxAD-S^t5*6a{-U4L3x$}ndE%h`lN6XYAnEs0RqKy$${!@(~5NTCUI@!4R= zvAnp)mE^!`m)IUmkA*)$)FlfaD>6*L^>J)7y2o{|Ha&i9M%8G#IDRl|{f@7Wwp%`| zdmSN89wwhY_y{-Igw-jUA{-55M6da(e|2{t7496BTyO2JdsNm@VsecN=(m%yhhklH^xcudv|VR9aVCMUO6X4kCt8 zQ^~}=(DUHx#V{>f(ofnbhkI*=b`x8JSXemuO1gOJlZ){$xo|_n6c{t)@IbG1rF0{U z*M41d*eS+RNYnRMy+rsJs;G7&nj)td2EL z9O|5Q(GOF`-Ja3zBTSx)hff3aBSoHB({sTh1@pUE9&7kX4WPn3qV|hPlZF->OVR&( zVoeOIb=(7%;D9R3?;5(A9NBq(1C8WOIGjOrfgYYD_*#p1-D=_%A^Br(i}vlr2%gX` zR(T_?ZM)2*j3jCZ015gqmqP)tKCRCnk{NVrl&5*_KpKWqO>gd`%DuAFS;s!J!K<3@ zz%59I-UGJywt5Y3xOuZ+eKi}X7`+KF#F}&~&C_7w*eVdd|FQOE$iA@kFbD8LtQZ>xY(pYajaYQY?%!@`=WGgw&^W$?vYaa}hS{$JRR&cjePlnqM?l)TnCMr(p_o)Wc7AiVF5jQHg!Q&9G{$*<_WDiTS(-Sr zT5@#=wN2X$_mgp!_d^6rozgEeTb)Hr@0lqi>YZbxl+OV)7>TghBV~F_vjT z=F7XXn#G^p~IA+i?Y?bsr2w4-zl4JW+b*F_1h=sDJ5^*jT z=;F}V8U5~m?0;GD@O{VkN)06h4*>KL0jKn$fUDuh_crd4YR~5sq}eY8)_O-f%jAO! zwOK^)edCD-wT%$Mx%UbN<1|XJrwIJ(#eL{Ct66}f*8P=x@t1S9YLd56_(FY3!_HZg z;dXGG7pBBHwWpZ?UX;BbstXuPk`5Su_mbp~hcpM+?84C7LVk{&f6Ra6pg8fp!)Uv8OhKXzBvRG3{I?J426 z*!DQ5Th_ZtLj%d$>t;xmgi4cfQ6gwvqZmpJd@6J4Pf&cx%u;9Z9WH~?B$*cUgffaQ z%Z>qr)u@kRQAF&f179rKW$-ViW#d3rp@YF!ZaGr+aSkH!`G_R1^E+r%oa0K}@>mFuEy0)_IRyA!j_g8~LdkD!JXel| z>eW#~-iKvC%Zgs2$)72-qAi}9(KqkZhWlUJ?wBUp0o{+K>^M%Fc=}p>_6pqHOJ)lK zL4+PJd-M#dUE`-PNMq_&$ssUXMX|X5-YN;tV*&Jq$hX`q@&|gjVBsi)9uoZp&@;EY0o@CfX~__`r`v@t zTs8U(uCEG`eO!Iq;62Q;Da1y2L@;w&y$PQyHH0eirq8$U*0py4u@XwZnXchlSOtVc z8b{#`6u(Zd#NAtCK%Ir*3yk=)u5 zaQ+$*!KoVKIfy!$S>iW*8!Jxx*{DDfoR3UevEc&EvnuklGw_7|h!k91jYwe3&yA-F zFAgI5e~?MlZzS7&go{c%&ZnT3%1)it;%IUlS{*Kp5*bd+AxjKE*J0b^V25k>aGM)s zgoY+ykEKNzamg$erqwKINo#!U@oS8^9Rc#XpYE}wgY7`%Z4ez8DyCY|_G9}h*i$XH zvXa~rEyS#`-EB^^=p(UM$sB8g1J|;!pUco62CEnpz{5LQ?o4pzUtv&*9F(N8wHyO6 zr?53ba%9XkC?Z)dRp9)Empf(Qur7_05OP&9@EwXl^RPa6DNvxQwx=YEH84snkU|rk zUVYR*I3Y*5Kj>dZ4<6AsDukh~N52Wypy`iX9vHtS(Nq!5NX%?KehxwE_7RcvTrF@D z8dbM`Jg~|V0S)V06ED=o%Ym)CJ9)UB=jCA;Txe6$EBT#;nJ?e}tMr5FyoU2jABhJyQD3*)NhW}pYb$uJgu>{^Y{SB)sT7b+&?dIlQ zeSGcQ4h_q%9xMy7sKE&>zkJ~TMPD@dmt-P8eZ^a*X}EbmNXm4i>V#`v6^d20OCnn;Z-i{-U07)`kLEz$Y>aE5lj~Z~V#H>H zs2bwBgT5&5KrE8Cn~Dy3UM_q_m?c1}T6IkxUx#G---`Juemj*FYC3QQfVd`}Jh+%n z<<>cQ0zf&mf7?3Tx8WdY2w^o@^nj9t$jOov-p;OcQxFAa1DNx;9Li-?+8QEo3Wvg( zo!(cQU_?%+H`_4H#d(AJ{D}61=Ayzk6hHYaCPOB*Mh-!&X`@2MW5x+VeX&jS2r5)^ zfMV1W=bYU){jhN{{lp;eQe42bH+V2P0@8&siq^!o!_6f*lib_7L&@4y+Z4I5DS?zJpe&fiHc6vyfh&68bR|&H+obnhl;FPELLu*l^fK?<1632fl^qA>KM+Gq+%Qf*2Bgd zl1382VBOUhEtuS0->40j|BO*4a{Z)-{Fgvm;-xZgE^lEG#mcK55~Wg~x6R|Bpxcuc$Ud=4)jH!6t8LnS7VkU~9Z>Fbg_a!%^HehJko ziof6~0;fN=(*Yg3G1TDO$yR)RC#e>pV$)S2>^L|o;mhQlXKu#jDx4+w2^_h&qfNn^ zS*KPfB#*ysQ7GzVArnM7hITK9-R;L*ArV@f08_IPVRg~%0dg*bV3=cS_hTj>fbcC< zb^I1spQ|R5&0n^Tb`|`0ST-UMmMb&>t_33R^e=g#p+1joejt#ZSMO80|+19gqn>J#@#s>Y7cvS_1i29em3*u)EEJ(f0CwYX3sDaVj+Yp~9Kh^u)qy*}5&g z8EJ|iUs&{E3i1$jYVFZ*sB+Qm+#@S3dzb5;*Atu8)KEEM4H{R(A(Bz<{H3v&7sd;! z*4GwT_u!=)xR^l2qF2y_mj%&MtL(_l02#3h_y7bF;q~E~TP^)(Phd2#4Gq?C@7m5@ z3+=I@H)Sa1*gURj=`7AcxHapYE;D}!1*_~*PRl63Kh-B}Y}fO#Z6hUp?C_UHcO5KLPQCxPZcoB-gsxx5)k%7ddS8y2U6$b@UGV$|Cs9QU4 zH{;U7R*NruwmFlTCbqM~AIYk|e#)hA8x19OuZo41B~FwmGVUA+^Y7IBbY5F;42Zg| zW5f0|F`e4J`@*Gi@LY9@ zgU-^%d%g~#7sast@Z2IM0JN&z4_-U%0wG|(iT-(u3Rx)k&PHp!+JS~vJbqt8KB0~; zz&C{j3{*+hQlz_j#)>9nJSt*d+C4nv5sAG{a>d_l2G6Okmp`^Hi z&cVrhSenaos+}x2oWByswq_`)-&88Jsm6aoH8$|jOrmV$rDYdctcUu`u#DPvh)&no)dBU_fG@ zqh8Nna=a1jRSXr%Cv}nJ){-jC3GyBqyJFWr%$D%l!)whSb^BMj{-U$YScIq=2c_@t~5whMzhB0;oJ_6Wj;FrCH3a0 z!7Aij#oZa}{r?>v6@yqX8#G0<;2&q@DNQ_NB!FX&p4J5SXH3)LjoD6m2-OF8r4@!1 zw3~~!kAeO|+ZZSLl4!Eis($fV(Y9aHQaC1(hWQ>G;-QDy5uVb`*)N+q&?nSuq7+&4*A$4S$DG) z#;{@NBYE-F|12}H#q5y&;|O~`y=HW|zTFs1Pe&C_!Lti7iD!2c`yJNFa^*}@a)Uqg z_q>4oA{k7H9Zb$E0MdH=jq?`1erXut9(t@$Z>)Vfl&lCfOR9%bC_2~IN^Xh;=^Xuq zZrVn-Dd#Kad5W8dVTn9>@wy|^9=n88lG+uroS{O;$HI~AJ(o)ox2Oj@+zKady}87^ zNp|g6W}a1=f%r+9Z)7xrg5Lhxv5ar)lkECJ%I-9`La=p2K?sFJ849WqweUd;Nk~@O@XT+$gj^`ORFDq0Qv|xMIp?R z?AZ-dM<42vEdVFqITY7blrw0lRZC#13Fwmf8cQ;f@)%W}xWwTtpPrZH+o~vWW?q{t z42UueT_n=A5T08p8z8t>k^XrC%(VLbgypM@%5eLAl72R#Lqc>xOG9C z$$+%o^)NFHCXj=jz^dW$sXyE%pvuj!B@ji@!dHbK!JbOzs9zF266r=bQo!+SUmdT# z$h(+aKRyD%-nm75hh0cLZ?Ub|NAFyDz<$b7?-uE~0n;+T)1l4-Lu8D##J;U(=SBgEZGaw?l zm+&I(*$mP%c=`Q55ce8e2g#NXO8NzV-xfWEPWHp5l)uEq^}noq(4m~ThX$A{W`{Tl z*ufLL(XlxTFE==BRUN_|GexrQ_vky6nKe(ICh@Em1z*AjuCG=Vn{rEM>nS(2ad?+1 z&gD)!*wkQrn`*GIb;7^RT|DN)SK#*kYlH}0m#|Gr2s_EK2s>cZ3{a8YvG@WrC z>iQr;=CE+4rjV9z-Q6y5+*G2XYaRF~io>dxuxM&T`g0poKWuHiYLj#N=y&AnS@ZR3 ziI@PUJE#$a1p783dshaY8iw@JQ+2XM{AlWYS4wJmcWQ-x{Pf?(|H5EVfw-b>vjIWU zHH-{DN3ZuZ)f15CX~cB!*P_#%p6BQ0vl42tV|xS8V7x_6cLLP^9&!8uRn&M(31}Xi zScZ_IwGMNbRPgBi{yfp>{+NL>!egTr6dus>T0~XNuA79tvKbg#LUS2X*bQA^p@ErF%>|1zu2Yx~mk|{L?!;x@b!JJD>+qIfbI1o~1KS2BJ6<<6@@L8MlJ$HN} zG~ecPgy!XX$RIyL5?Urx#0+9XPFN7clkV()o9T>>R?mjTk+K*8)xCLYx-&swB9({q z+!6Ndm}>)S;^*X0OAadAnew2`L%+7(1zp078#;THKlYYXL|0(Tq0N4{jSwbM`@mCn$rU}oE=a;_=cpU)E-EUQu*A8-5f0PzTq3X0EQ;_j6*gMG4G|f*^ zP|({zNpX|(N zbQ81QSW5wiJSGV53S_$Z0xjJ+4)kW!3v2*ZiEvLa;ha+3p*W0xZi*YtBi>{ZKqR}G z1x68!o3`o7c(1d^*m@iUKA`)G9>8d+mM5*)GrYw79vq`m@j+-`I>YW+-y;f*5(3JuIA`P-G2(26WsjqVGrxLKDF7 zC?zQ~y}1lLDs|I>$M-fG-ZL0GUl0LxkDZ937h|b=fI#&@|1I$6tsa>$P_`6}-f>7( zK?plo!+57=+F3AYVPh~p*m7bjASy6ojJ;Ekra>2_`;~2VSzWg6>auOywr$(CZQHhO zvrDJXO#El&MEnsmmpd=stIRj{TDfyQ55Lr?)tgdzWpeAcsR3CQ2-dm8KN&7U(c{xZImzF$TPT65=7>_dnF>J90-$v8XGrjt#8z0l&n= zmqm<516YRQnX$Bm5iDS2Ay(bjPh|1vFZTIsr;B!)=cCeNmy+TZ?C&C$LI@4bdtHWa zSQc%hF(uu&HQNu3)Dp&(n?Amwc4*sej+2VfN=+7ok?^vGxL)TBgC*J`_^|!QeJ8go@`*TJeSjP zX~UyKN0@5!A{vUObtQ^XV`O161M|P^$_(rPDY8+&pEZaXw<($x zdMo(}39`(uZ|FU=ct1p1#A%xUJzUmy8q0um-TB-4Egl79N6L96Ilw|hgRNslX8>Xr zl{iQQau(n9Hp>lwVJ5SQeOAL04Q*g!qaOt0R{p_LgDV1!92n`$Jw zN*;aSg?SKly^Qz48kxgdNH+7O6~kPEM_o};owJVEKo?f&N9b}3_@xNj9~GX=ta3RE zb!-?Uoie4pmDC&xS&OGLce#Bc6L}HgQZ1Z&5@rz~u8a}RSWkUe8$0m5@`#8^(%^&Gm<{BglKQDCKRcYu&=vXUN>s5|o zn0ifOg~9l_Y1tR1Y?6x(AwDJ95)=s;PXucrA_@H~wHg_Byj2Jti}swaLmH$d>yscA z-M;Xxd=>QT)7+$$>GD}H!~`;{{mtHSw&?_a-5QTKiU^P0F)QbVTUg}8)Z;_p`=l90 zH-k{AJ?gWdlWFk0tl%IfnjSj=e(iQ{3;giwA_r+8%U)xUSSmAK4|!&ySDcBOsAO|B zc@;-BjINe?CPSu+{WOc_q6urvXx8nbAq%A#N5`N{Z`5mdkVJOXX=l0J(LSNBeI5p- z#sc32C!#{s%n5 z{iioXS+UI|Gd=TM*5I+xQi!;;pK#c|yYZ|n#)X^x+UiyufUtLB7Mzo`yU|pgWZX}y zU_j)knXmgQEkWPI{tP6041u{1M>|IP+-fW#g=(a9Ytl00{ai*0nm{2tL6}?4*G`I( z;^MkbiggFKcFQQYed8&I9jCP!1R;ep5f5}U_)9Gz+*71GOsZ~{*g1%h?FvL5*8u(4 zhXt)p9su-e`AV7$ON51c8PZ31>#&3Q6rYWMrBSec95vH(su$6{zNr!Yh7)5mZ>W=2 z9s%ci3->DBgn^>dnqY6I-NGEwysZiT!Hl7x$_HKXy!RtAh6bNjH#lmDyytu#_3He#{A#I zIYJmxEu}4DlZadieo-0{yyp7P&m*?7)tn7IdSAkd58B>jlMP6ugIlmmHu`wSuekJB zOE_7Sg(C_4&P*rXUNzA6rmgR-ujTadwi*R(S%W&;HFGkXDlYhotsx!yxa3w-yAl6d zP(};fLz;X>^x8GFzeV4oM?mn7Os>Lz_9FKo7>dIkT5f*TBZq)Hl2m(Z4p}dn5QG2A z*6M`o-|M$zyn2BLe&?vL;8Egjrd|3LBj|(_hYvTK?y`@n==CU7DvJlpobq?4`!Qy8 zMHVk**J*_=dj})XKi8|#og)uoRvPz8ru;RTeI^d%(!Qb?HaZcuD9}7mze0|+*j_v+ z*+dpXs9i!?MzD#?*npANHMPR813JJB>u7{ZZBx`WbPatG!cWW)d5VqFr4Q(beEziZgQ!8zc@$dlsNFwT<6 z@w8PmaDh(5n#-m6X9&|+3fL|k;gZwNF?r()06U5;B)&UrU3kvKRA5@vw_oC;uQLlI z!Vk=0YK7$t!z>ZSeaRWjYjV_oq#q3l?~3U+$|rYt&@{iDl98LiCmEhdl^y`?n*AYl zI16j*oG~k%*NX>cy!TGa-tz&0o%@}l-N#b%v)mdS$5K@ilp&Lqxag~po5R<)0{hdc zXiPTG>y{?yWz#q8qxQxfJjEek3;fd4`L}>IMcgcdDN5)N z(|L7*{bjd+3A=NBKl_R{DJ9w@p^RqQ>fL?ufQAcRI2ZJ`L}PD>v%}V^lS~D&_)PR= z_z`sS(DY+n5No?%t_Y+vl|xpx!+*mw$SYKX+DB*}$sOYutY&0rQ^LwZ5qG|M4K2}$ zo)u)J(e5OyxIIA$(`3HUOpGy5us~|r9e=u2F;`>$RWYV&sYHY3{5M);H~ID-76BJ^ zy0J)73V+w%Y{&MgH3y_}qMPHa@~XSNc)2tH|J6)4OA`vi=MY@4kT?cRs<0<-eiM4G z;zb~>wqPi=#>!pjxC*hLWtyl+7PmQ?^$RkG`f0sdB?e*~i%J4`7oK_-rvU9}!zQom zgpY47W-ILrWy!!J3!G(0y#l%~z3|#ENgU|*E(!wq^FSNwWUKVe$V~EIz=}{&E9?vR z)vnpyir-jn=sE-{O*)%!K@bSQVbY1!H!#Ys>oEU?Hp4*wUuZK7Z1l|ki#Ef+$ieWR z>i-9uVPataAFLUI|1)f+S;brlMIFnz$U2>!pbVtx5(N_cHl5a0GG8qD}O+4P4Aw7xV@t&rB}| zNAko+i|g*!&8tjXp|OKcT!d7J{_p;~Kftj7MB-$FD3OF2{(up2!K41zKnf9D5uo_K z#lgVA;}AJS!#FBbLp=auI3`@25D3WgcpNbKQ0^euew-ibLnXfhc$^@(qS3EB$UFK+ zL6HSmKFq>+Y-9>{AZ>XGNb|A0J5;#AmAiw-bqOn%I)Jdg1Wme#rM({#B_c^2G(UBE zNEG{CJg8U*mQ*Qt5f=F#Dk5F#KkTF1mg_xZdnEuZ1se`L!6TMLxF!t26(V1t9vaFo zaa+KxORL|9iopM_v@K8#?4j`^kDpbWQXZ6|Wa2jwp(i*2up_}R<0FI>EhOd}BwTR? zg@Wx4;EoyA5E2o3abW?JLL>tzPAQ@rBP&EY zBdQ5)=c`g5%0OZ$p$0JrSV)Nqq11$&4UZSpb0E?uA!Ri1I(zjT zgn9Wxt1^oHHS=OWxrh(S0R84M2|5bA>?U{10cKXNF>_M>oj(bnEIWZT6bBa%di_c7 z((w$&bnZ}-LU7kTYHeXW>zLT_eVX zofbumF9@FZYRHrfZwCnpdCF9rn0GXSxO z+pmd43hVHTVLy93R!{LWqm0eI<0M99Yc>LScmuqA^V3%yLWr8_fmHJ^KNAr_vj!GR zJqez};&8orKP5h`KyF0$eG_nTq)JThdAfmi7j>g%j;;^kyu5j&bBifGYtS3*kkav2 zuCzcP7xNPqy|R^JG}$}L4CZI7#LW%W`D%Li#%((LDCDVnEp&F>ywZfs>^q6G>R3u0 zcED4sjiiM64P5l$?_Vjlf_;~}tv~Fab0q1M5=nOk8;aKlt4IW%Gs$2MPHp3MvQ%9L z=AZSKP7=yDyKAl3J08x&5OsgZhko3gReMf{U0Y^X*k{POJS-)9o2*xbMMLB|1kTy4 zLsZ~8*VUfMPPA9!j@{SJ^GsSR#YrNsE(&{CvnpEBzdhWX)-$S}OJ;hU+TPxsu{%2V z{;qL-@V7*hKAu;|=X)M>T4-DkdPx>Xmf1!;u)$q&-m=JQ6hxc5pU`|4KgK5-fAKsw zaTzO0l4_JCPxIC0q@TKxyW1^<_f)rHk90RX`5rUV+xB0Y83&kHm7Kp*cdl=v_^c&S z#rS^M>2s*rQ|o4KumAnzFEkuZGRhU|a{d~5c5SP?9>Zhg((?JzX{FWJ)Os1YJ`D=% ztgJce-q`0e5|f)B%VXl0_`JAa#psO?&TMVO^1{mndp!VmsIu&I9V_+uspd~pazD)| zbME-U;=I%sxju_|Fox^O{=8g~JvHL%JNd0)vfSKZRgj*$v?}9?UKHcNnV^vR>&EKv zhx^WdIauWN9~jkmMRC;+jSfRYLVGij^tK+4>tc$c<{TsPOUg&olj78;kNEV438;2! zNputZ?YC2R*<}cJXV<{?3!VL^g>b=gW?oTW=f%yoV`|e8RW7Wpl`TS2Ry8iCy8Dk7 zVXiB^tE6<1t&hRa!v!{5fbCb|7+sxW=}bjh6RWAj+kT88Q>f!|B|+FPNy`UBj>}*+=WtnsY&+4)<7ZorCz=G4=t__o8yd#*^?SdOCBH56+1luZvoVw^KN0$`P-+#<^arKKW`zSSDkt zMnPiRwb*$yBk|; z+k_r?oQO^w@zY;)#XHv5ZEaV6m(-_gQxcQ@s?{<}&MnH>NccOgEklm#FR}U;I3M7A z0hVg?s~ywQZxw?}@f94@v@Z@jZGFebiSXL_d(l2qdr#h*s!2B~9krh5b7`u-dkIfP z5vZC^ojA5_DA{%Do)WgtGlX5|=Tf##%L&pnyO(?Dlj2*+mqgKWA%30MQg+L2`JavAopKN z&I%ImO*LAEB**eXu+IyJwkj~UPiD&JK~d!K`5Dw>CkTXE_`if%3Ak&f1-6R2=_>N~ z#_Iuj_erXmVI^t~>{C}3!@J0Cy3(tU{B`<`Zr}R`-UQ9rj1gNJ_UyY#pZ9MGJ3kDL z*D;Ig2E?(B+AtKMp3w|5X2b7sgD_#QcBVg(7Od%VYgzjsBn{a}CHTJ`DC ztv$W;CKHaK#COfpiHl0y@ncB!Nc=Koj)2(Cdg&X() zz2gKxTDDS%LBX5i&me5b^rawbd*1wEz>I-WVb~*pwq!fRs?u)|SR2sf^u)S-csc&N zR7Rg)6CPE;(a~Ute*Hp9hy(n*b;M{wa0LLJOU^f-`UPx&06$-tl*vln=ce#?>D2Dr zbJ*SZ7UVusTHG#Kh<^aNKHu}ns^JGBA#{#lv6T@h5Kaz&ab7WvX;A2rQ&O8skywb! z^AAj#XIudZ(hYo(FkEqsC)49OmYNxmnY1zhj87^DOHj?*fFBT*F)(eF3xs!rk}Ot2 z_#L?SjmQ8B{X<6ukDzZqMWGic1_=R){|ht&>k|#6vh8R2JwOQn<>wFhb#WM~m!e|| zuNZAmv7aIwlo<9K5K1W52gwZ|q`S3`Dn$<|H!h$j{5Bh+OJbA6={Ni#r@t^5IoO@E zjXHn?G!4fRb!5A!>l7!i67VXHuA@PXL^KDW4m%W4Ln^1|F#7g&xoOo95daN9BLyVA z??!)ZW25+y&3%Xd*!sv2ptMGz#pC!9ewqv_M&0ETtboft-*1{dg!m9A(;i6zfc_*^ zAS-}Xcj?>f*9~s#(YnMGp&r`15=>#qqW-r@6UXO;Nd}_aQW4-*xNBI zr2Zf4lP(m;CHdrM#$rb;u9`@`$Cc{``_I#Z?tt*7sI5uEvQ4L9`EkM!n{{mMPdTlZ!u8}z+Epd4vsZ=NaA0`p+$R=G ze3~428}-B7aa32T9=5niB!#))j?P4>HN1_<*3;x}xlpaN@}ScK#o2WZ4BIA67e`<7xE_vGY)TL0ij-C+1Hx zhX**#e_=oMI3K#~t<|((7ROeD<<{5YysLU@ZuIGmKGV3bHKntsHHJydw$65=8D@7+ z4>>w2h4Z-C7pp;csw1RrYtkv1)7qYJ$%h_ax?-!iwA7U!J-UjIYewEGt@Dq;U!VW( zeh*a7+bbTfOzdYOaMD5sq)W9=5>cK}7;dH|JgsvoYh(msLwCeTW@2Qv)gMNufZ{i) z$|)C=rF92*y0bMc6O53L{kF5g@s+t@HVzBhoFTQx4D5>hYQ&8kBcq}pou1{`pfsM7 z*qmVl6J)PB7CdMFu33F{HwQCS6}PRIRo`}W`0)%Pa#6UrDL`CnH&f^j?Y~&Km(-6p zf;}ldk9sJd5R){q39wa;rk@Ay#BPf1!HQr1`hhBOMcN|VXJtZMt#sGnJ2;_XZ{rIAegl4H+)Pb$sVk~aa%h^h+2!b5*)*jAidi0{xMN~39DOKe@q?Ro&X0Ud!Qc;QNIpE zs3XPkj88RxJls^`n$&T&7GLEUe!ZQf@}TK)Sb0wF5!A_|f~9C_BM-{tRY%_7-cN{~ zl|R9lL^1HHK9Nn~q!j#j*lYWijm+V)&(g=9EcAD2ikdxwW}TCf$jHns^QH@l5{mYn zjIF9pE{ZXK4y8e2<{Knxtgc*dqO_m4KL6O6#Th80%b+_KE~R(kdL3%JW)?2P&^-@6 z=f8KRLD$|}iatCGa;k@S*2?_2uVv;AUlhByygo}!Nf8y^^j+4;D5;cc1p9jrdF^fw z%M#TpWVw(`NtQKG1bN@VNze7ru6sefQl^`sWA|hQ?DSNIsdC*XxgQbH8J!5D7vr(2FA|@fu z983R)e2_Vv(58uZmW--ni`QxQ5^>(slqvhZcXEwCj5fxIYRk|h-CWCwoT(H!Wn*x0 z<>8TL*Iq5wIpVo%HYS4ZLb}{4lC>qEQIOf9qss>L>gx zK6br=fXzv9Nxm#lMW2rxIz|cI0YjZybU#kMMCU@o3ui$cDwMTOt`>1Qjb=Tag6ieX ze%d++dm-*UqC==QObyIQQ(pv64*funKW(Q1L7$f^?8DRONUBT3 zJp^4@O!YNY2EWch5#nDxn@eJ%3Uw7gjcxW!*Ym+u8DVv4PC2pM@pxV-^4E}&Fug{r z_E#A`&?I|jg;QkLWf@14@zPA%l&_vU%%D+JEuf@gAyA1qn2n};v*)3f5;0O-;x#K0 zzq^(wx6K?pDe!|i>((~MYV>IxHb|+EBNo74C#9>6_I+?SOm{=U-t2!C8q;}a;ITXg z&TMl`g)Z#`xo!0}XQK=G{UFng@(SNe!jjy=q88t5G#P_YVpC)>NF$r`fOA`G)*`;2 zxy;_(`!jmj_Gbh?(4rhGEm6V1<291QG0rr%>F6eQp=?}@{sy;-inll&^=;b8YlK+f zVn1%lX|It3rAaX>OMn~ye67r?{zTL&>(_3BcjhJ;!$^Kb7Xfczxovmp2p+8zu z&6Z-FHP_rhd=EPx9;uCSBA3PpgVeMrwyLoqO#Tq!UEroa)RA+Q(#6-fZ>$kYuMbAv z!_#Vi#WHoHKJ8Q=RFF2K-Foa=7`}ute2?bFEB4~$BY~o8YpB5miAPnTZ$TSEuWxNd zXuXu%9<5oLz5a9-xfGv1Q3`GI8-yufLQE*|+`d|p)kFUqjV)Q=rIj*l(nWN7Aj2Dt znt_zj*ofR?JhB!xGzJ^aZCVr<+RA}f_!veKSiKS{=MeFRGK~+CHb(X(Q&7B%R)HoW zIYsz8X0sSr1$Sx+qnrOR>E;H^({*)fifMnY+@605JjNc1m`n&G#%(P*+dSyMvPo+v zlUqm^xao?w%@rbbxBLwY$!Foz*1^f?)z$2^tFvG9w*o~8eNYYlF+Lux8opZ~eQ1_=!!vo7H})nUi>?~d|j#0tB!7ILOsLXaEIYfJrKAGFZg^g_4o zcZz|QgXyt<@oAXbT4j%$E^L{PHud!^#$4;jF|IThZJtEm)<*rM)6Y3@G_+3$&J8xV z)4d~P67dOts;>c8l6dW9ZJN>is*0_`*)po%(e>_~O}Db7pyc+TaNfa>3mMjo$EWKt z&@B=TQuWVk?SI18Ay<8suWT;iTP`a)2|f+n^hZ3kcn>&xjytFP z*CFyS!?U^ITB|?;u$TRiQ!IB+yTJ2rrMv-TskT2qwJaqpN z+E(NEob!S;<{)DB2=n$|QFZImYaN+?-Ni+UAXlet%q;d6;~grNVvBO}@ro{t)``AtdHtTP8F43DS1N|RCKYa9!=x*M#9|0vS9r6`KI)gg=b4D&KIBI z&1cDfa}JpQn{&X*&i)^-;Gc8A!1VuB|GRU*#K6Jyf9)J7xpOO7;`b&GuSRL2zBcHhAuU zXYjhab?4*z-Dk#hCjGS4C-?U8i$|kUiwE{xD+!Jx57I`g@RvG}DFy~QUIUQnz5WcA zkSjTR*6aWv+Q7<4cwwPokN_RN5Hc`WNX5@2CKP}m<|m|Y1qmV?Mj{TwE(jTfMjCBfHh}4O{*nOXftL41;rKwItW=LTC|oQKn@EJ-#^SE1q(NvJha||x;JzX zR1iNP_p1(()~=vU4uBHSNa-~8mD;^#{cBCzkBMc>o!&jViJJofG3oy42-Qw!O` zh*y9a0q{Uj-9KAytqIC#Pu5JB57vaYjIm-l! z9w3Ga12%*a_%=t)un(E~FACoO1Ud`>f_)GN5=rpjH-ys+rih&QZzPvbzCK7IJA3JG zEp|k~IaO;xwSjdlQ&m+{#E}3b96>e6P{<+mf3u;W(;H#E|0$XJ9QASPI-p264ItD9(k1wcF14>gWrD3np% z1GJyt=P)Bh@P6=D2)1Jc)ga?3EdICvk`LOB5HEZ}*j^B;<0m)#qa;;`J$Ke9SfImq z^0KwRiarhhd3S#|=$+qSspDT5t>Jy|eZL@t8mPUteijO&=zW%V4Kgm&iBMmhi39^9 z16$6cTR2&?tXM`{ir)RO$Xj&JGm^z|5K){Qq?DAYHe&xI+g|t}q7a0VDZZ#GpKV~T z>Vj1j&wxNXJu+<6^b`=xpql(|VnMtS6BB|E5wjeaaFb_{GE~sva8v@wQ9~evVX*_I z63lcUO6rzr8gmi9h$-6MkR!^R74>(_ukia=ODmUae?v3!BM`@M$xk{$P#akEcWk7M zd?c5H_XpcYRk)__q%YkDYioJ=C_=bjSeZzf1=acfEG8oLx+D%KHHqr;$WlJ)k*1jZ zagYaboSSkn&oeZbrwI{4vffqfsqD$aFHf4IT@*T8F>OvmsY~UTa2&Xk zk5tUvu0^EPB4}>gSS^iRN_@|a_uJ{kwg{CFNlVXiq)ADIOycf`pL68M2;jGMnT6q4G=c^prH<>xL~ zFKJC53iDLrH_mS=Af`ueGQUeAMke@}^w?o$e7*1~3ra4SHmr_taLzwKO{E!@?uH8` z%i;GsWapzAGp_Rb^fv9)BxlOPK7Fc%@CP~BU+f#goWJHjuHa185jr)|a89KWvg>#| zzB3#gBfsf5Qldu6X-jb;UJ(8iwqk0sZxX)xW2Z3Qt6Ln^ICM%~qDR)Fn37%}&|RW! z>wHo%e&+pq1JFFYk(XBaaBnWX(3|bu}ARZF|mxzN1##E~2wXX8<_yZk6+5}K58Ve3yB$#umiy^f}6)Xe0q zyQQUW<^ktD(e7H?PoeRl$mzk8p?kXetGVDS@ceWJ0c4$bwP0}K&jfCD@=5Ob+MQTb zh;_se!V2s2xpGUbfVtCK0vF>SoyYH94(v;jXX02PgA16uc^>1 z6|EJjLiXXMt_&rSZJPL(f;qm+xR;lc^HS-A7a3Wfr27DHsS2x!v?LUOE`;K6&f4drr4(z4 z*Y`@Bjk4kiF4Nr^|Kicz%h>9SiWj07UFa{LlfhQTl~qVdR6fWa4Db-hCd8X_Jd&lh zKP0nizzQGES{$Hk+l%s41uPfF?OAzVPTI!}<3Vt~Mm&Y`j6BAV#x&Qri|dpo(MFos zTy{D|9^*B4>^IbhK`cRPxU%wYjAFN}@S|eZihnh|eUGMV%NYuU$xJ7D-<`QQ7kv*V zh@A841oz_^!r|gpCyj(Bn+LClRDpTIJkAs~+hdTvQA-+SvNnB+5c}k|DQ<}FkpuT*VPh6^SYh^X^t~;1{j>N5^g)I1|W#+=E-jkE1lqHo;VyF`%E$>Ql5+u|KAXLYO@zAO07)1g4S|1+PL zgKv=YzeQ=bTFyUmnYpG?lHcnMeSInupTfH~bhkZ|oO!!EQN+G(()Rd`9zrgRq&TVm z$S@j*rjTD;q(D&F!2Kdcmy(sT(HUZm`Pd1}q_DEccw6m3ix`DKQB@oF+?ryESyLJ7 z%ZL==)-Y&BAc>Qv%RATj zXR9-mx6GVYyX(^!&DnJJGaDLSpcR6!8gROu?(-${SnJe+iB)vNLbcJ*h(D{c?iBcw zZ(?FTqqN!rL3@#x@(xZey_-L~*k;9nR%*8rQrq4`aJ)7r>;CPr{A6H#Yg{mUNWuE8 zl=w~FD)keMaRh>v!21r?!JTcQttYBW_5**cxWhGmfqOOy$J0_ds>W98Ey)8?ZS z>sjsdOq@E;NluZ#wd`h_Zm9Z!cdlFre)t{*NMvg>pXDn^Gaue^8ZHVP^P_P$?|rD5 z?`~$zOd6OKXw^(|fo)rJ9f|ppim)_V?Ka z*T!HdZ_`iDiK=>YP+NS6Wi3&zR`GgkVuz4pgn&dp^aXk zL_Ecd|0Z}?|BK*ZWMg6bUxJ5;k(K4Y-jy;jGcy0rk+uKx#9C9Uv6|L&3rhs3tLs8r zXQz;zYwAe$NT8-2Owb-;KxZce!i^0SZGU#{8jJCiyrXh;36Y|_ zDRQG7BZveSM?l(In)*8sbcBTrd`XN|m6QcPm8++sDu9cLii(T~Ejfs44%O)Jh};|+ z_yNR?e+6Ldh{%q9)ju&2Dh3t@#0jk3ABhfNvh^cEG^*IYhBX42%lHOS%G%^_jQ z3zNsrD_{FUa3+w?d~odjsb$R1IQOboa@VS9*CNR9nuQoEAy8c zs18XA~iV0kn#rZ|b}5#evM;4D6NnmlV&B=cp9!_O{pe!U^9dtE+=kqdiFb7Zow0 zp1C2y=k(6?YtB0Rx1l1cVnT9?!m6GE03nAsu>eI4SJWg9;|%|H_kHL$+8a<;)fRx2 zR|Y`kCRU6?^E(R%r>|-o(RY%}9ApLR=;&hl!53n4W`2Be{`JSOe_>>P;Gy*e{bq(b zBZigshF^XB#~vaR@n@~dj{ragI0FWFq-!>Q(!R}heX=%wu?7a;JvzHMfU9L}bOL=@ zQTy*92HXKr-_?Pv#L|3L!>g}K*_>c99ofzg|fQ;wfo)O+hS ze2mRaDr`>-jxI#=%Z|j>yST8ph#rH^0dt|H^`paftE2JL`GJM;3tpdx0)+=+?+p9| zq{!5S^|B>p!t@cnY(CKq!RaCW1iAsJk$n;A{Qf*39Dvk`ePQ3%0jm}L5U>Iwr|^g1 z_1s?}>;Y;$Sr)(6QUmo|90|E0X^CXU&OS?g?~#0pSD9VCj}1>{*{$q0(%!1 zzKCrdBlw^-HSZAqm*zhLdmfzsmU3=-k?_|RpAdjOGrk1$WFLKrY40-H6M5Wzf3FUV zzi%Gw-I3))qe<-^&_S*szmb8cCHjs$&GLTgxZ1Nd|Fi~!Y!SN+UQ6SCPXMd*KOh)E zEH*T~Cng`VJ-sQ0M1x2TJjhDa?C#6l)cBHshN^vm^?R%RoI_U?HMcgh%-z!=)J>nz zLD~$x{QLgg4IL!<#5Wye$%iVFm7asIu?Yy(-cu>-+Uodyt;$;8R`w5dpkcOje`*xZ z+(T?wmer4;^B013e=JQT)7ZiZ9icB6P}-7{T3eIENpvbwY>!F-LJZcRZX+_4tr9({Hwmne`gD& zk6)j^XS;|0o>qP2zUKJWagHsI!=2*{Aqx?^>pPV3bBT783xnWA{zLce{R861PnXAW zu+Wv*&R=$2I&z2XTTYi)Ot0iMHG14O!08Zfo3cQ*8Gc9N*gEb{1ST-;kGTXU1AY4j z_oWzs1LIsKXn&=Mu&ZM0oUx8G|LbDLE~8ic7*G?%46sCAz(g zX%lHn0x+{E0?FG(=Z)d|s4fGo!Z@6Srs|^}ckfICi+<;Zp$AO?iVGr*j0=B&&Tajn zp_J!v%9*Lu!=SzBv0k=kN1Y3yVNV42pT!%KyIpO$kL?r3>z)qxJ;9}eo&9r2&|lqaXiUE3w9cL&%${nk7+b=;~aH<`M|;of1oi-+rj zQIpk7Zwy(wPdDfI%}bRLYlE1Pf2exUT62S;cUCJ^i z&k@sFtgdY<{R1we>V2H$@xpaIQq)9A-&YcVzJ&+tf@a6+&HctH`Lt;D^tjT$6dfcKg6|0Jo07GO2<0&0Na$V`6Vy+!cCz*ds^E5 zg{f@|kto5+Z%Y#z4U} zED;OSX{1!yIjZJ~@}echsBG$(V18-~f+5|B(oo7y8XLWq3jW2Mvv!(J&thgRZbb_M zqW)_GZfmFaU>#}JY<#MKA>Q!WBAvQf(;0O_R6`&fyBQr?-jJVHTX>l_S_1q7ZR;e4YV3!JJzkp|LvXl$ z?tOT7FQpdvSvOD=R@t-%5{G`zeetwrVW!{?deII+9o-3<_Vp!hiQF_XJjine1msBxr7bnzJ>K@kh)1UGS1rqvOk}W1+N^o#|cHms73`ySRP-~oTQvNaveQ|z2{N*b}kfe z!@KwJ&dwqO?=RLPFc6F%zeR$Vi~240r@*B8lKeuIyEW)gBxzc= zU*j)f4MI-Z-&W8c2`tPgTt#htE^Uz$s_zC*%auQdGlWni@RFG;criZ@OAk`dY)Kjq zOTnyCKjH~n;#VMYOZavvd9=9;72lVRtK~Pb+Q$vB567NLWrORloK?WA7D9BaH+^sO zHwL_V7fIIj{?H_>K;`$L^@=h|#v?_KN>EHf4>J0w1-~UB#2cl8itB~6lR)Zd{-Ree2QNi|D!d{R+s$~75h1+oiHdb70J z7GOd%FZIT~`+op3K+V5WE^GJxcts?g(vWgoOp2LM&ySklRGtN_auH;)Qew%qjly0# zG?I_lJn5bKgLW|le);R*e4=F$;XO{-IwU*>+2>m!5;lrBo-wHY4AHPzs1lx)IAFhs zdD~8oD6K6MuC_!Rz|-j(z4ylOlF!}hb^nzc#rD&nQ z@L5GV^g1GYT2h2iBfiK`DkZe>fGhtmT?z?Ar}FUb*(t?VjH9aEKZV9RM=+cDjv2dN z1BT17ZZL-*La+eTPrM|t5hN(7T2834=u<9$(VB_5?4zRVRYsy_Q~5hJd2AX1csppO zHKL5Ab;HPtG(Z3)^*D0-r~B4x`>R;z$F#naklWshYT_iPxAwDJaV5U#+IEsk5ZyJI z+3r}}k#SFRG9Tk;wk{rmUNk!Gwa;Q0n5EI+P2vp*9^Hwg2R|_^?{rIUwoSfHRR5@X ze>Qau#qguXh{S#IOA7R-A7?hH>Lw#|8S(<0iPD|8E!+^lsFYiOY%1=Dw_!mMRt%@0 zGXHwC`N4%mOI$v7KDwy%oZa)TN?fXIFK}C+j%-sQ8|Cnh35iGi`bNtB4#iUoUeej!1VFS1Sy6D=$H%|!> zVouZLAXDD9p=KT*?!$PCVDgO!Z8U`%Q--s=QI~e&7oRTDwet^1#FL+&Y<@v4qnnXV z%N9%*4Rn~lYjI>q54TY84wJ>qL1VTM$cvrD3pCrGYIqIx5)ldWWLdkHUjktyFCrS+ zIxv$`gKR$*d7p|BMjO@;3$a=gVntNoX}*1E!?-?IkH5j;kv20&L+I)j5X@J%q((JT zyUQHh>6I-VIJPyHr;OpLm^XC2dlaPe z%17IsD$@>o{gN5mfTFRd0`K04eiZNHM&P0>ReJn2Iq#k5mi3a*`!Ow_-Sd@{V~G!| zHyv%i@xy2)A2rIa*yZY0J<*R2nn+%3`KQcLAQ{Z-TM9i{pFZ3g2;Kqs`n3g!AcQ7( zNY}^I$6Dy~hA6q!3^}cVL{F~%b}1JMc2!G9+w@nUAj^!Kazfe9nKza1tRX4IcI=fN z2P>(r3UW5m90E(Pk|NiY-#H55zsiX*w`{^fFEg$RV4s@NnQNa*&XQ^p*UnrUbPbJ8 zEZkQ$GRXH?UV|7alZC?1G_6C%2UP}ePDaNzID?Zv94`}(Ul`>o+^#UdjFogW?|gPY zZBqx*ObL5sDiR*47Ms%2uFpvBu65FZFd;ANsQY2!DS4L^(oqVd*WsT&w$-n6Z3*za z^a)+oiVSw0eAzP6=(RV()f>o!mpH25+R+CBqF165!Nio%dI zL@c9-x$BJHR6pLZ-?K!Jk@yD{=Fk3QFYzTnzQTWEr&KllOn z2Fu)~RIx66ldm8>Be!YaMH2*zSk<2Wd*z}8tHYw{8ea5bc*`azirRG(js4(Lcz(gA zI{TI-xwlfZPoOMekBNQg(+9-lP8)Bj&%MWJZ`_T@guE-5GnkRS18py{4j{;?g)`a0 zLOVIaIH;iFb+IT9wBaQ+>;|{Lq0X*qEAy)`>9(+vKKMj(H|L4YPKh*(|x@jTWOgifKdrD1Xe$Ts9HdKe?F z{^IarKM{H(C+VBS@Y@|1ZT+JkJ8p{s7e^u{^%=Bs~ zvi5TFOk|2&%~b}HQN+81SksUZMP$ILEo85TsJox=&-k*wmi`1F(l!sy%1XV(dIlxJ!lPNy+Y#c5vDBGh+QAu!>9egX$9=t@w>z^(Bj!^(?!nG<2s=HN*3Erm?|m?M#*`82G{at&R)#eD^K)-& zYk!`9%Q&OIFw`v$F6n!}_A$KV$^ryGujIpbc3RhGVKKD0XrCC5e7ABF)Q?3O?B?U> z)v1Xc)}U2jmNcy8J66-!wCDAezfvXsU>c6~Kn*pB4SDO)iIUFsGr*?8>Z63vO9QTs zn87U+r#b9?piQnwE%phk-ysy4Nmt~HcQm3d?d`a)y$I*cU?GshM1i!#78_D?CKf&p^WTmpo^6;g=fx)OYi7 zFOOcBwuxkvp*L0*xr;c1I~q*_f$=J(T(Pf6Q3jDRDepY&!X#qQ#h5dRYW+gUcpjw? zTT9Rc&Y_qssj8B_NvXUVd%Lg%(I)?GKV3A&qIrhXF{-px)2J6g zNv+9Te}XLPk8OHGct{e`{VTTK$EBq|e|9Z>@yqU=zKwW7Sid^W=ZFPpG&lUH8^N!& z2eu!s?(?s|&i;^WN>9L8V6!XLIkb??JWi{pN{6cE*4$r7Oy^&nU@EN$L5J9OgPR|v zFFd}!r?cypY)qt4_6@T!HU@Y@AP*E4wuaw&`lnGVW@eH9aumgv$ax0u2KdFPot2G0wV8)f)kwme0sGLq&M{MyY6MeTXO8h!>aK(6>e*daoXEeU z(PeuIzD4Sl+L%!=k6?j?^0r-avqA51>D{kFSGfx)mOZmoTCU^_(@%SNnk;+LF1(Zc zx%f|ob(cO8T+~)0Gu>0V8gg1E%^N~^oQ)+m4hl~2WPz&%8B8|s-NzYB(cTg(+@XjW z8f96o(xBA*g4hn9^j@)hm$EH#?1!R~<=aF+-v05bW9hL;6u&of(I#*OPCm}w0jR5I zmsF3yD>yMK&L{FFyQDSjJ+*dBKw{o0%lLq|*2f_Zfyb>ns+xyqMQ||0SxPYtAAW3S1NOZ$A3m{{Ueq(H z@OUZF!qdX8XfBR+mJ;#N~(%DP$n(_tcSwzQkm> z0VB5DVxN%A3pGy5yA%>G#lSnr8-CX>Og^4m&?Ckg5pSJaOm>VmJ@OZbs+x!^1sZHT z&-bSjxqc-y%V}MOsumhH`ql>L!cmb2#aDzgY*yd^jvxU3nQnFmq8w135b`24P^GS3L zOg1+Yq&a(((xk_Ea5@t?29HE+TCfVI1Iayh5bpvo-3Y()>`Z;YH{IsF9`}bACu755 zW6&jogveT#Xk>&=|* zit7-X-yhd%Rw&5nr^-}8`dO-n+H$&mh(WQqJn0=>Zx?AuED z#^kT4l=oDZ#T#~mZ|@Q2*wTP-C=2)!qDmTYxrFpxoBqXfKLWl%L7zGfnCW#6$!^@m znigZ%KZRtDXEU`y0agup77z=y6b6RWqqn8Xe|&<8p;%NvYi|9yUzfTh=}cT*v61YI zHYopwk4Ks6ZS)z2#x3OKi+UL{MUq5sF1Leq`UhhMzW^mq@n;cY1-Pz2u3@(yBXimE z5V;0YdnIj|DX&63%vL>+O}OH{mEC$1_rGE@3)2c8X=C6xD%gq^)3o3TH_iGt&c!+@4v#Bu?Rqjy4sO;KjW1UsCStZ zlVYE;vj$$R-1-oBOKvT;4}CX9BdMK<;KD{AGc6guc6BBc46;!Vv2sMi)pX zT~m{Re0umXqaP&Zo0b+YtuJw3&J@iJoOYxd4(LW9HwJbqMa`fK7!zL=KrU2VvDM!4 z`*|+u3FG~(dX6m4pH9B2Z4wz$W-!V5g0MnN1z!4M{J&!)c6>Ljw4Ss6p3e|1F!QD=HGlz)Ff^~vsZg8i=s%}*gDX^m1W za~6=gyPv45AKQp0pO6xVbF#l|PeDf^X|Kd@M45c3Iy0VHxZ@T!L=1g!2RQ6VtMHPh zF3I>$jdIWn)w8g9vGOnme3rggDR>AuxhecEFj`#(U_f?iQZLjT)(GDvfzcU&Mkpko zf~=1(8tQvJ0iqv3ygo&b332Hu3;H&{LSo$rnqy-E$Y|ow8e>|#<(a}qBAb-eI3Mdz zz3gefOGLhUn3|2xMlQBvFfnQpGOkFMy#AbJkTygg0LD`&=Wn{VJo7q7rP*1o4bO6h zZ}_&uG^LZKUftYs7gXQMh_R5rL15~QT?2fwXX%}3XBYbteQ@N0j4Y;;5;O1YFZ6Rb z+n2oWqKQtqCcFwJ8InXY?rzE~IR-;*6+910V3MlZ$43HVjU zKiYj3tQGd5rSO~sh3iHWp<^jsJ-{F_8NZlxF1}bdMiU+pb}piGY$V-%Gyc8;QwaaF z;PBf6nO<~P=*(bP8{x6*##;!kqBJ)6q>{!=YOCDm(-ehcxq=>^tU4mz*+D#~ju zhr8us&f$Y;*UmX8n-=A7&91RC=(U6-kq1VOQfc$qYgUDybBlTqSalZ`2vu>nkn|^z z-g#OPD4ujzO64S4ivIf2F&?q&X(@n`{B0o5$cXLN6Q;mh5TLw)2qHD56+oqWX3XVc zOS7B4>LJPDpzIKvM4Jjx&iJ(El_5}MX3V6nZPk0b^}5@VZ0ek{kA<6K{ee`)EZ;a1 zN!ik4hog(~YouHOX!tyC!?Ka4C25G2@~rqNi6=S!&Vpf1b;FV1%N)zb@@P=tE8&25 zNbQdv!O}j>gOn1pr6doh7-acMa~iXOy--`IgOiTur_xdwNr+EefGMd?t|9l3L|wU8 z7xkisS;@%N7ZK6qF-NRs=g?i#ZGJ^-v-8vZAhC*~GdCmUvNs{9=bpt2A8ILkLODM9 z4_Nyr+f1Mg!kucP}bxiT1* zl>7$MM=2&aTpB!zsCrX6`L&v$Voxq=*jKpr6-xfuq`M+ueT8BO=920=fL~qwDtDk< zNePEjPhoWqIE#Fv?f&E|PS#WCS&wy0*_)MjF~chMnwL0fF%?S0vLhmNO--M45bvYo zc8ip7scXMrLSHassVAt~FjpB&6T3OWlUV)SV5YJN;;#r`tniw@WY&yUMMfqg$TrFK#b+sQ}k}EoM zi)($8135n8Y!>$@jI;fuE)dEo@;)k69Vu^u?yew|6m|HJ2@lY&rXwyQK>xHs@V+1q zdol;j5T+O0-xasxa7_@brMs_D+{l7e#QVxpi^_$#te3GhD3Q{@{xveBUX>4Y zW?>h8`D@(Wv3oG+!$c7m@PlgJ`q6{LPyU$KAt;~6E`y7quGMw=3`eQ`2*Df>zP*~? z1S9onVFNT5RnOE^1nA0y2Bw+wpqh^q(>|qJD#-^gM=pEMoa}OgD3&ID-r94s+6W5p z7{8unl_XZja$FKuIKe%L-?$*Vl&bK95Npc9asP_BzqgTQQ z-HeSFnmTR}&k}+Hia9rvZ@+F&Jzm<*N}`zdFMjSOM5ma1KtQlz@L#k{<=E7ZESuq= z#n{Pe(RB3pIf#_j+^aT8kq9`WX(p|aHs660E^26=5I%>GmU@Cx;OO9wfJvy0W2bnz zV4PJ;@qaLa9x6Ne@J@r04$G4tz6jFvoo%6t$I)#YY)tyZE`5^TENw`71WGXG@hvjt zbr6wyuYkAFUJtBZBKzG3>axReqvCR#iY{wAG;h5ypMZI-GJ07t5vq}@ZlJne9RmxM z5Jj9R4o2vTac-@DAU@RqzC|EgT&J3ulYzyD*`U-uN94q@L>Cgw0vx0H-3x!0w`!VF zitEgrGl$BGP%T-W=1!9>N+SZyk2qQyKXKrXJ+&lo-7ZF#(D}2eTzZo^ARjgFso!|_+RQ|#mROs?QzZg1j zIETun$d_Dst3*5pZ&Kl_4D;pzCQZiomM^7Lfu|!@Z`hk_U7liW-EqLRSXCkSgXOh_ z>Rl^6sw>a82Q)}+gbQ*vVpWkb)2};>G*-WwS(8N~DaVaApya1Yn6QF4v$sfc+ zsDaT`0cE`t9y7)jeH!D*HgnAZ^BjUt?(%nm;?HS)Z{Ri8L-=Y9RH^(zwoX-ei;3m) zq2fNr5P94jY!$HwjEx>CnE;sU#Po*PoEq04%!|feunF?1?Oy-tG zayN6E0Vt9y#2o5_=@ce4KN7~^BJbnYJgXwTU|>5H-E>=!nKFnn1Uon8REe`HuUVXw z*u%B#wzA-2Glp5~m_ZE<8F2^EPgX7QK@{&NXxE&-s3{vEo5OAklFn`eW;(%C{NS}C z(+y&#ml`dVBgl3lh`o7KIPNX2H(FE`06r$O09&Kb@O5S4ZJn)!|1iQ!HY_15TGNM+ zQ$YDO9_j6P`4y}PQTX6oe`Y9@cDoHeuEk;K`jWprW_IkMDqh_XBSyZA4$}piWgDHU z1q$d#TH;!p6g3nxM1y1`j4l!7$#r7wPs(>pBe(9nH*8+8f=0~KQLIDBO#PFLqSU12 zm;T{ek<{;02RYMTe^m2k)Y?P~cE(c!Bs$O2K>+`OhGd!3IjN)-_grjmSR=g6_MV20}Rds~2wXM;g)%b?T5O}&cxT4A;xmLsz6q*RTc`Co# zbfzyf`}YXlHd?jb1#Zo8;UL=9zB|?hOfne5daZ^F(He3vK;<2~Smo(bYb0DsAryYJ zF*P~v42g|2if+@I-zGq?o7&8|>8z1X8}qDcqn;<#bUqYSR)dNfr(Mdz9J6?Uv=hwP zEqkX%_9)UqCyK0TbT;5c{9N6zC}vqNGU@^CK5Z7>3OU%l+q|rNT*i$rn-W~YX??*} zENe9XLR(MpKqer4^9!M}Myz>@b2Vfko?sB~k*P38|2@{Xbz?W{2x`l;9PS@b_?m+@ z>3uil(pkhBr@d1wT~26MZhJMa(!EAv)8+4TOPLVwjyK$qI!C$ECMr!*UvIHPXp;q$X?pM!*5>0LUzEu(Qgpfx^)|3}WPDdme)hQ(I{mdw zd4_Bgl0mDc<7FU^{OzsTy1EJ);)ap55;F7WgetgoT=dseLso66_@)@gQ45p&pkg!? zDvEX}I>y3kw|w<))Rr1krEk{=@})ty#dC$s(b3`{Cg$PjU%Mv?8?notpI_;ts-(>r zBad7bM${h+;~i>?ghdObwK)JDNdGf*NPg(V04y$ zTitOPM+?>i)-qeuyLNbdnMT4mfL|IQFpL$>Sbww>f;O}*PpYyKW*7J|FN8psAX0Cv zRYN>T8kTlDeQ(#KY$O8N{~Ws-dqfdMrBSH(>K3ke7v9N8_UcAEU7;5M>kHR!sgB3vG3mFt(WK{RRP$|J*jxRZ@Bp#C=lT4y~BadhMNdwY8E9$0}Y z7Omtj78ny38!G-$4p|&m27{?%T_HM>wPA z6ufZfItjOx{4AOmRC;PVyl}CGQx+)KO4hNRYUuxFBdJE}A(3G?;>n($!;>3D9Ig?6 z=7rvhkpUCyjSBnY`K+uvUwHa9Uhvhgw?L${&yHaSnn2CEQx-WR0wPo(-{gyoH8WKG z+S-0J{Htz4d zooB{b7GC&oYwyQ%{sbzmxXv@8j^)PN(M8z9eU>ORWyOZUMN5b2it%dB!t^(6WDHL{ zF`INqpOh#}f=&wievqp^|1z?uE3k24JR2yz>a{bbc&u-;w#VS%y}F9q2Rm9W2y31P zI6dx(3|JnAN9X9Zi1+B##5T0rFpn3oyseL{xPpQ12oQ#;UXS!naLq)Mi={~XuHnoC zckwnQ>Bx_>ry5(3yBL;J`b#__z7tzVB#VUqW{dZXHT$Fn z*ta)I-X`03r2}` z_-NQbqC1M>Ix1D!uMGf{iSa@VryXX1*$D}^dsFH+6N?ZPdq)h?)dar;F_y`DnGkX^ zqa59z$*FyM{Rr=6Lmw;lYsD$98=%xf>p8uO)ZbjTEjmi{X~+zH-69n730kwD;tqOI zW?fBkPlinY7S*6`OXZH8ih`C{`$;kdcZLrGZm<_>K2{RG_1d-6}FcNm-%oea$%53z}Gk@)Ruz8Iqh+>?H$+s~+1Ah?W^#y!sTaCsJdmO{MztW@^+?;$>pB-FV`*MJM}L9HAW7xeP0c z3CNSQz;;0Wo@0Le8QVUzM6RfGU(O+-h%<~AYEp4O*J0k!2Ei=Hi{&0v^%Mjr6Hb&= zZU~w}ZD?CBr9|=CNg>;U`Y=&%*`xJ1azA&)VXj;y1=TF3UUR>>(0gBm^|8LIczm)S z*J8t0zYWp==a6#-@BQ_KUYVbiH#!+X#x_&rN2D3O2;xw@(JBf#Q0-O%3!O9d}gzke;p5qMl+)?l2-3dKD{qs00PtrRoD-(*KS#SECKf-*`RL%v`OP`MufZ_uoKj&a?BteBpCSSB(EP2V zdjA#m;8J8d8{G0cQ;4SAx#cgSM7HVYIW>7mYHa;bwib-mDkjfMi)B1AF zN1dxuj2{!cenQB;%kiSrnA@`nNi$qe&m?{5^ddf zJ{7AN);w?mQ8lL_4`Cz=`-t`Cs;AlDy;`-LH7+Z+@BUWwDBov_k&3s)(6pxf1$&;a zkfL7_!x6qA@0_xcH_&*>fy(c9=s>n3==_|0QU2EOTdnW_lhaT7t;NI~1eQZ^5tl|~ z{2#|!Vy@i;{c67Gjj<}|5p$iy+BF!87>{@!*Wt0K>7F%Pj9Z%Fh+MqXMnEL+#~(uy zoZ#2c3sy-c0~Bm_T07UjzoEs|Nw(y3BI zTFt&a1{zj>q+!e=bxn)WjQZh;Qup zF-rck!DS-)jAcA}Jhh+{_3eLA6*ovn<|By}JhGPzOIYUMzq4qt!_doN%`~f}7dM~`ILeQA zFsYc_>=$Wwr3`CTfg?xuY;|a+-HjVbz^`F02_#JZHGQbhd-)B*%*-og{kV=3hjUy_)syaI*B;mg>yf9lvV!O!U@Xy3}Z9MITWbKNO{? zshr2f8%&UmicT#*ge-n+)J}-PU`XuVrQSi0%ti8Eg+mXMipQ2YchE zH1}nN7BRUiGP2hdiuCJT?N2RevwH(5;!%|ztCI=uHJOsC8Yeb5uiJ^n{#<3RT)y-v zu6To}AcZ>5p$Ex`s*8EdTHnhzvMbtIQaH?;EF-T0s>sFR6YoDm9%;`@VLn>a zAQ@v{vp-5_He*dm38gtwh6#%APeL{g&5&XcRJ*h#h-E)WVb+lQJ79pu{Dib?xhz7) z8Ur2Y1Tu>3gFO$SKF_`ev_F`-VG!)bj|`nS4MC99*$g4j-hscG;4w)~-rLj*pmD!l zUe)+c813R?<92br!hRSSqc}^9cF3oC#2Aa$n^7Vi6fdmS`zWt1>S#k9`W^jY<4yg> zXemsIzkV-~c*7NGe7M}Vr^!!#Y-iu4Cnk~I-Vw4W1*&{r19WJA9K<_SV<}>&NGBl* zZ!k0~Z(@w_$={Lig*Nv0pCl~(C^_<8JqfaN)8e*SFMrTxS!pffUQX(=ZM6H$XUOS9 z;eCF_jZM7*g!H;zYy{V_osk&d&?KpTXm!(2 zs2G+HCc!)6L#I&V=ygrb6^xD5=7I*JlZLc;R9djY$G^*AN=!Q$9d#dhhfbF(Or(0i z@^qn#z55xsCp|a8s{)`G(^@1E-}w zIf3|t9MBtiqoYX&Otb_A7*L!s_sR%T;mH;=2dI^o>l0LE!xe!XPjU4?J!uu6F!BtZ zH!_eDG1#?iw5BdRYiZk%wr;IxiAaf#VKL&}dCruzX7~0GLWmu-v>7tnN0KuMBF{XV zU5Pm3ESpSGBE?xcaD0mf@13c{SNm_6X2(Mos|#yq7S5O^8dMmiH{>wNzbcXwNs`O5 zPaC1u8*fQI)NTb>gT$%nxB1nBVmuxcFL$Z%N!*GVzAh6oDe zohqf0{Wgc|J1>r^IYPGuOyf&TQRdf=avM2W_)O?0VGoEz3|fjuo3xKQcR6Dl=AS7Q}vwuJ47r-eOLVcy(Odt#sXc zl)v&}x9z;@s3(6|DON%p`}PI~zwH-NkHqZu24*0Z4koAgRi43xwJZh^SZFU;Yg$68chwY*1}LKnx^0#G$O0aYq2CC#j@qM^Tp?%sA|&ha!^-@HO^bd(<$OJO@h!E*^D7u*(1Y@WO4!hf4vJS*DrUE2 z)PaHM);9R8D~$R=EN7Hi@oQ6mFpc`%-0&k^lI#(h87`-rx(5>1_wC?cXmZU^X!xQY z&Ss?O@YP1~Y_(kF1<05k3v_{A??=YnE=t)R>DddqxdQ~3D_XudlSi+4U^pn;(czw# ziPh7MUdos=zCeRMcaO0?1q**pxVJ18Ke1`uC}*5ui#Xj?qt&vxqubfv;h4X4+4!FX6~bO$BUxJ~WN_8ZdQA8WzE~Lx38h{u zMY5|5VZJkQ$F9*sTLM)Zxy__B6!k3ZAvgi+*a(rW;nlTJ_?}vc3@uys9}RRu>&PVE zPW!2S_;`xv=5KIA`e-E-lv=G4YzT$oWQ)R15$8)?#FTaJ)4}M)Tu!U47h`9bYpdFM z#LRhH#`cr%ZLdh_j$=i6`V2Z!_?DOMZS!M>&Jd+us@;@7<0>J!^IS-;NukeB@A}e< zMR$lD!T?CjqcDT;qNp@8T{@Rr#W0SbKl-9YMZ?|Xbh-GdMTMJ~+CJ@dQm)Ll0WFV~ z(5_0`2g_1C3KEI1d21%5p7#m@r#7Bn`?Rfyb`YU+j+1uF;@nkpUUjuy3015nm?G2p z7Q>s}0LdR`@le~Ze@)z>&-CCm@Jw~~3G-d0#cbRPJd^9Eyhp(+<< zeB*8knfZ<-|7-nADESzfGo)>hQK@yHgMemCGYt@F>Et>Ye>tJHa_~F5NF!#1$^m<5 zL?}*$E}@hTvlOR$zN`=J6y-hMk-RxdV~{jXBA>5L(9;Iew^!fgN*)~g@^C|?$-wUy{c79ES?5Pn^1OJgR=OpV94;j?f4 zaNN>~mioak^NU$ZsJ2B-rX$$&GE;qXhuBpe^=J(Ql_2? z2mkulo66n;mh>|6$aPNxMp{JF@PK1GDX|V)l#;ZW#cYIhaw2WD9)G#b~~UeG#7RFH);4rgT{O9;c|pRE?#L3hIoIS~8*+Z>QAYrIw+96++y5 z;cvWN{<=2gM@waz(H`$%nJ#T)ya-s?(7f$ko#d zsJ}$bGq>vA64{fOZnO#tovrC-hzcy9%)a$wSS}LrSEH_+Lt4H-B+FwApt(a9gx>e9 zC3zM9`r;)zkT|<>+{-a*u3`Tq2oabn0l9r3;P-}ZkatD=`sh-RevhZlyAeU+{%k1L zbCQYA-UQb(aEvRuBhwLvpDI=hVLniv z=cEdcp~0%-K%M38Gd+kV^epXYjttbTIMop?m=fXAmbUT}&hmt~ykvjd#K)C6P>8<% z|FLlm&ABklGPZ5ow#_HDb7I@JZCfX{ZQHh;oH+N}yu14YdNo}=)iphh)mKuo&a+DC zJ1qMEGzDk3I5|VrE-NRvfElT+8SOJxjB~?#;V3@2w$Mq$FDU-$t}F|`TT$Ql$GS>p zolxITMCDKoA3i@FOw|H1}7hhw{(=1V&4PcLqM+Q-Ph(v$(7+KJb54WjRx zGd^v7*)bMuObYSS)yUVRCDr0)v?1n$9BOXTRQI0L(1>lcfGRl6dcTA6dQENN=(L=p z`!o|XbYGh<&xx%BedNxH(=+wCq(f(oIhWAmByifPzf0N$!aUV>gp)S|zwbipO;no? zFTnhSIxP4oxjP-DRhc)Roa*XR#YEV9jam22j*I}4lzj{z>5(hbWnMrm>LTsNwVX+~ z$osr4qNx+6o&H!_;1%?}WF=V@z2JR_SmH$yuD~kkOQCesiaJw=^pi~h$ z+M_e>?Cwk}oY22Qr6d{dLu-u;2y3MwZj1i(d8iZ}$0x?+@zG0@QAXg;A67_d^>_qUl_-{!_Z2RA<*rLo+oE?BKDZme(jnAGx#EgYl#3&j-0|Xys8=zSdY7}%$_Mq zBKi4^DbEsB$l z5-ZuoX1%0Po@BKC_wku>U{0}_J25bx0vSzPlKsg#cdF@@i%e z1F3PRgebC!Tvkfb*2EwTM9`wJN)qD{p%5X~#e6SvT>K`9k?Q@;Zn_G!`~{7SKjNAh zr(b|*&>>+>hlYqL=C#Zo3xm(#$2Ux=)?{SvL{J5Rc4T8EBIQ+a(4w9W1bH z|D|Ne-&7>_cxH1RLT}4x3Ttrq9j}`L-jX$Ow|t6XxdiJtiaHX*+jo}a^8uG_+I-p7 zLfsU%&CH!YIYUe+Lx5@vlFtL#xPib^dc^PX`Je~-e#WPoD2rH~L-f%1g0WkMG?&!- zl>Be1B3DrJ{8wv|0z}ipr%Z!!fFZsp)FU1+vwq+VTaW{u~SdpI6j|>mzzhNxkASJELVkf&hc1$ z)<;vMD@odgQeu@r6YPOZj}UyG#j_0lRGR%VhVLe2BNR`2v3wNHP=@zSYwx-+Jag`S zA(X^VV4ZqZZi>36+YFILM%c}nc9}#@AA$J;W%j1VVOBOIe__IO7jx*WEdpBzfzY#F zS)}jn95ryOYF0Mq3;W10Eulin=j_Fn>N&r4xI2Xl`%4LwfmzCUik20`oammhrj-S; zu=dp(bX_I$=`vmo4&$mfltn7#qs>@p@G@?UE7s*!^IYXpNAF5XOIHQ`j*xkWBT4TXOZAx(QqPJD$qg1He2lG3ai4mj$$v;qE@5)pRpU%MwzQ-wz&SD9 zcbpkSlK#eT9?OjSdj7S6i2|j5HXg`r6@oCKDOlCI)Xv`k6TV#1d%y)O513wlCH)iV zT%41gGjD^YE%u%fHy0rqsIadXy}w!Q;1`{t*}r5$rfOM*5--96<|?U-Ku_{alF*L& zKN_fdV(uX|RU1ncg9bg;99c(9KBd9unvUkR@1Oy79od2z03S$2sB0223^tNCFIoIr z_-}@(%)@Ek0jGDmkex}DPG=Uz^GUz}k#OLoS6$uyM z4?7nd3GA2$sjg0oGFce7z2kQ1maoH%SksQc>cBq(;PJLK=7S4TJa%7p+c%CK9a|gn zvF4a4C&H!=bzUl$dx5i)`wuHC4!)ul)qNo*Av3GA3O1`pj+`@Q64R*SVHr)9{$co{ zJUKxF`=4h(6E^)tNttFVIX{wCT~tZ8-DanvI!&gKRmO=noZ>S@$6fXsZ{4;FyxJUA zidKOF^XZM$cdXnjP$oWVHXlZK69@5P#TOK59K=I0)n-!lV!?kOwZbQNd-_NdU3({N z=9VJs7oatK^_k0p*=0WizHqD~p(Atk(vXMBlS_ErcH7e_tmT5;20wXvwi2MZA}BX; zT8VRv;KT2lY4Kkz%iF_OrIZUc#+Q~pbN3L5@%~Y( z?La75sHWn1sIbb`bzs;h5w(gZsaBnvAG?24?uBYm2EhWK|`{JY5Utah;j z=WPPHMEgn?eaSd&p7$a>=>e@Zq9AhwBuieoHPrRN-4&M;jb4ck=Ra+kiG``0aO~LP;aRLz!0vN!%zc*F#LxK=vgE@dMcCMBb3l-5HjKDo zqXTT>u-kP}b_4b1v2_^Lf5ZULmSNWDHjZ89-$j1flG=Ri=J?zr>D@tvDzY_nvR-Xb z1~<6}wgeS!g3$Y2a6|2YY$2u|M6ZOf@ZBEOwo-NGa~IS|$QsfJ4QhSz4Elp@ZJ*1| zUaZ4^H8oGBfO}AB$wxvy_5fr+o4-f}9|BFHu9A3DhKHfLDeU=zs%*_!t@tmgLfXiq zlH?q8n%UqBlqs*pfcwn!UR~PgAlTDSdCabD%&6b`-DYWi+S?t6a9-P?6BWo1*UWkb z?IgSgOkk+g?@!~papk{~LgI!jb2w-S180MMyRMs)WX$S55@py_+1=2`~= zllpn@%#1;=l>;t)B-p?<4(5BHXy2M9@@Dp%dRQt5b+-vohQU)paE_H(%P}T$%frM` z*;XhVr5yKa!dJe5~AlG`u!#*DeQOp_=^KnqbE)C}jTo*gK; z9LH*t>#P+- zTyQ=Ml^&Dj-}QrTtD4bepyd`-(onlerUxG=^aX)H?_xjD@E;wj^b5QF$C!SPs%rYZ z61Eb5i#*5H%5ZrORKTQYYeanqyuy8fP-YZ;+nyHP%rS2htaX`ZD(K$qfq_@|N?QOf zC`Kkzw@{-v#4K;~^3fem+cs6Rlg61<7aFX5KgR8F)V$R*7nT3Fkx zm|*=iTHq)6w?m#Qng-cK@%ryEJ4wIPyeoWjo9eb1zC{&^r`oPG6bRT2!P-a;PA~Rg z;wwxGZi`l^4S1+u$hvXXWu#yEqxjQNRDOa_a+k?YSPH&#c+oqJ<>btkl6gGBBzd=kbQ!5oa0l@_8!Fr;+f(XTxUHHrTO^?{wLtmU&1Drq(;hY$6EN#XPKL=7Y_+8cE9y9gcyIU+d1>`hF#rO#$vPS)EC0IG=wD#`LjJBDIt*;JWR3kd zgET+h>D=N-2U8?M43roxrfwQWC_)rY)t!BANPn1zx;7bZe&Qzt*iMwW#B%G|c6QLK zbxy9pJ~vby-T%-~8H7sJ0JGvZaYO5(6`014nP+a&aZ-E^M)%!9nL4Ee@Pc$7ixSp@e$wGzTwz zb>Gt?vAwAW|AQqN{#g1!2(G=SD03w)7Rk*p#aOLk%A5AyZU|i> z%gl;?SD6fu(gOJF68sQGI%Isa-^(&9Ttv)3Zh zjjiH%$-48Wk7`x@L-4YjjBF6fk2VK0T3$YY;GEd?J_=TF#UBLSCl8S6Y6= z!h#c)HVYO{AS&;NR~~2LoX6F|gSiC(f@~G5>UBnuN}i4Kg^uHJ$Lq4ChH~t$dZ}$S zxdt8ctbFGz+40 zu7T}AKmI0vsGWvuVcZdnxUJ`MJ- zcGR&0p6b!D>hW6>N|R%4CrXXhU-8~nRl4U*HM?_6n+H0v_Fmn?9K$?&Iol`?U0=!O z7tXvTL4pexn28XO7~T&SOHy7xXI1we9vZo&Ix!?xT!P$815C#rz4`8ZaeT+BEQG}b zmHzHr9*qcGNaC21=+THz>YOViri#kcd`<=FOxWOid}bw}9&4GseR*^q0(3FX_>59k ziX*CuXNh5SqzaLhy^kQ$dM2lro=R(aLYNH&R}a7nIpC=uBiUMIXdG|j%uXSlSnux& zycF8T&h$CKJIHBzQCW_upy==9qmfFyVN7z(z1UFpVyR)9a#x^8Qq$@=16V#_+t*)5 zf!E*)D@RTLt=ruG_i_r!;_vY3|9^NYz~6vE+xs zI%_~>!14vq_;GEmp4THiLTy-JHwuOsQkNQaipg`3Fg`P8AV8qT*rLu&T zVm2ulDbX?eB_UPh%QJGOVa-@d8nM$N>GL3#`B2@+G|ugAEIg?UG2}h%D#3%_r)b$Y1I*UDms!tKJ_@Vqi zBOQlt4Rxix@g35{u_DC9NCT>QKzDj?{MVnrL_P>p3vnOmrsm!x z`4|0sUSy-T_fPCEJB2b#3P!&yPtcY3`c*PN4L)S0?j;#1xpPjyB3tU&`cytdBc<{l zi#UlhvSGfIRz)4h#JYVS%V-wg`~D4*vHp<_0Dz+hKWDkjaW7e5G^-B6TN+!yfR)a6 zoU{>uQ(lbw#~HNR>Yr}S-xa~}rS+T+xK=M*m;#_UbXs7t=;W|52){lyrYaxCv6Dfz zvfqvFb7moEG4F_ullo~;WBTphcE~8`0}6Kvxh*9>&n;=)FUnp`wMlWucl9SLakwJY z4_Pzik3jHo&}4p!A%Y_5g3)FtdVsWbAdrX^U~IZPM6Rh{395tqwsy)q4&Ma~Y9KPI z=Ubycl_w$jwPFL)%)uM*{4G<>_3O&-N!(E8fbVKVPJT#P#um7nYmGR1Z z)#>3}axJdBF++=2hVaN<1IRgtyaR%>5Ttgdu5@D(1D-&%wgA^VRb>B(6@fU zXp4QCsyRARMEUEH-kf%rGM+0%4N2_Vr5`cKcU1e}vi1o;0P z7ut)7zaR?P@tKC{p2AFkJp;vZATUE!Y-pb~2nNv?daj55vtV(c=fddtDIuy&Kji^e z)r1CHn+T(Mt1JtLc&Ysk?4*f6l+4>jO?RVun3oY{<&@pXCKx5hlfF7vR^}vCV}WUY zI;sH38eZGHc;NuOBT__1aNn5hhm&ldmbx+MT;;H#30{ZybMxcN(+pRd0tDqE(^P@- z9H`9>8dOAou}%&3*%;rTGQ9DG(e)+qUZ#`a4Gp*cEHK`lRs^prKL0y;DsEbQm5<5t zfIHL0(~7h^r|k`WD&_l@MR<~nJL1C@sVZE%(XPH0q6IY0=%528f15JXqgm}7c$%s8@CCOyW9}n zZb_a{vT-*&!oXo*ZI8Rr0Z4VXWdLfLd-T>il?qik)!+FC3Kxk`&unYOX$&DwiCk#P zZa=NHjsB6(NaIsg@rj!JoEE^z>89&mLQ3x$+H&qaFjv(HUA*@b$t8cL1G0gafNk24 zT8>%=LQMjRxZ})^0x5Ycn127TYB38-<@9WGQ;$yXX`B3av3UiDNk71pS#iZR|Yb{B|raDSe%Ag zl+*(UEr5do?7PN4W9x=m@g)UvxSPF6nw#@NQ1I! z)^+g31=J4p3$HY{p1~O(3;JoY&s$5Di`K-B10yT2AcsY|5f%=xOW>zFP-~s_z(c+O zm^OGGUD)e>o4t^wo{X)?CwJHYuS6vC&aWsNRc#l8>4Tn|atENv-tRt~EA+>bmgL*E z$edfy-wmej3bI|KN%4$vUdXchyIoAN>u6;*gY75);q-)5JFszU@rWAF#U)SLfdWgozMnP7a1`djk zZkb!8b!T;mnL5+FBhFu=J5B~xl&;b%c8Al@ZD8>)#@UkJ+o4MjHQIz8=)Mnvdh^sM zD>QIFIAn{+Mm*Euu9w5saG(82uIjZO&Z|d)GD3R>rGamYlwrAA-sNW9?Mt8{I#_ja z?M9O?D(o(uZm8U>L(%tKsqy^_o|t6TnXesq2iN^Ql|5f>6Byq%QkFIEY370MSx!diAM=2VGBX1P6TagyZ_bw zi53zXLFadB##S-Z^{k0Ft3>;}9b{a*9WbCv1CKRmrHA5Wt0#;|7&#u#F;OJ@>sEx; z4E%bq$AD-7XZy7abMfsb`QSubR=)-G^f%9t6yAd`@)x0K!Y#sP#>M{$7#(H=e!959 zJdTU_k6tv+GB9I_L{M5eDh)r?tvVJN$rb{}>lF;aoAGWg^V8;37*)xOA2%I0yoys1 zeK_qt#on)ma3FA>KRy=Z5scHlDjhfI*dOr6;vI6=5XD?3_wNtecD`P&aEjGRwM=(N zmaCQ$$s`~EiovbepE8JT)h}QnM~bC%!9%*GY6bf)n6(Z|DQ^!|H!~oQ;(iXgaAH{- zW$?9@Jjm&<(-F#q^u#O@Rc3`b;#I6dy!eyFYUWRo4@VEnwwqRyTy*SLIuP3cZMxlO zW=+F?N@+$Z6U+9=#tCut$C(mnwPuJlN=GWCXP;U1_W{{j6T`;#BIy9w&s#(T#Kne0 zo1%SC;v=ElqmC3jN7LFpjMLfjk`AyEP&}nPbZulCv=_J9SG%w-JaqSk1O_^kD>(1Y zjrog7D>Y}W^BQ)1#|pJ5*5Pw`34>GBXotw*jc`01B{aroNWk>x(}Y$UHV)5$&pGMl*K{8x}&#ymxAqv%3535EAKGW ze>1c0O23KSvyD&8b3<>%u;5a9*kF!;9>asBh9s{A);i=PauY$1c`|IH8zj_JY%wyj zw@DJjHN;-3%X0ia;B8-wCxcC5`wM-Moq922t=LBpdS0CyrOYwR@qXRms8t3OgMQGp zzRD#Z7yv6({F7o-UpLoN-D-%54Jx;COI{vISqWXFDs$Rrd=WuJ%3^~e2ej{Pu3Yc&Ht8SR;5UM4sx7tBEE-2FowYvbydXDBelf0)Y6u{f97Ats zk;SQhpIV-G2EcT6!u)@ySP~Ai;LhzD4Wf)<#dUa5E2|!RTtt?Vn?&*Up5rCvn*NXl z(LIcxi{Xy}c2_k=|9X6}hqXDBvIG$1mY$>Bx!R@Oy4)C-faAAiLC z5H_d+P}9MB`H1>Ntc{CD&190l zEAJ6aL23h~o*JyCRAK&=0*?fy_Xr$6U7jh)Uw5i>O)=;)353l4tdSr1;t6*@X&B=V zoq;l59VH~onr2##niIl#fypnXNt{^{J`fDx&r=hB#|R0({X1I#%#5nf0A2Pc4=?I1 zBvo$E>*>l2uQz|x_TK`brrFki58Z!^_2uxfXC*D9L}9$ZLKa+7V@;C!ClOCX$&{ak zx{ag*XGxnkEJl#>?>T;I;iqmrUJ?E-i+l|us9_R}2*WZcw1l6a5N z{U+>dQ<2I!_JfAmc(_sU$_g}`7K#o>*mf_bmrx9x z$M*)eQ9WLdQQtLH69X~i4a|7}AZxiDqz!=JL}IObGl8<`y#dL0^0X0S2I|QY)wJQ^ zK?{rt31lR*_tKUAqRGkCvXzg`?H5|-;XX-5OV%!F^&7IVdOsoTpu>?i4OL7L%uI54f3NzZTaom zrD~gQ;wPz8Ro6mH@V>2!(18-MIMAcf)V2WvM2d5Y`E-Z5lJ<*V>R(%B?%vFO&1513 zt6)Cb+Fv36wdJ)iz=y=>rCGerDR4Ahp7>BUY_`27Dkr9Z%W^_1#?x%;mw#!s2rKce zenWxgJi`#`Bvka^>44ZsmW0+DW1DG=)=!3zY$K1(R;_Lz3vTyz2Mb!Hj%?-)5A;z- zicqk-x;cqfiePmxu^`dYyiD4{#EH=(rb#@c?#QA2H#^Fke))=b%%FD@iaY*_Q86zzFYcTN_N{k` zSnoYq7*6^Q+-G2IB$RA39GTgWY#xAHnrDVM(%5jAri7#y>&#Sz1+4#e@hC^7a;;ZY z1;h6mXsMW1I;x-NI9aESxw7G>yj~=6?w1js0EZhWCH%?@=FUAp<XI->&Pdj%5fk#ySa5E^>Hr8G@NdUX1`1iBB;M(nWQoLy~ZYRe9#Ovvy+HD^W? zBAnJAk@ojn8nKG8C(k6k3vYgxKL!I4#myxTnxIki>sUpFn2ZMh2KMEoVGeRev+Iw= zPF%^H^W(K1@2T2%C(78rd!?P_i;d4^0hQhW~L#*xIHMw8bNGJCCyEK4avr zKw;{NM9m4Pry^D`!Pg-V2LaAM`KxEXoyIAJA^#UP?z?bEBTaP+^0@kP(iKd`52vW4 zK;LZAjl;Jk#cI?uBK?mGaB<+r*%LHub%MIR8oI)Gl(qZH}X^K+z4;Ka4qVJ$d&85z4!GjfVh%r3esoN&cGKTo4-5G6z% zyM2Ioi_y^~_I3K-Sg}7HsdIlk1EJ6HFm+vmv5l*QOvLkc`6lt6YCY0M6o;+0Rh^7IYGy#6C`kNk z93x(==sdT=1vh2u2gv;oZL|%JrL&=D;!(`1^eu1efaDjF zK$Xf$v-8qW@dEXx(V}5P%a*7r-gAdg?WF%y8CM%vD|srmeX&21N6>(>nMCR{{*@|( zul03OTdZlk^+R>p5tneae_IoKN6{aizeBQ12Vm>IPcTd8Uw#)-OG{S|W9LlNAg>ir z5mL2Cp$%V(7#I4)Voq;CRU4d%ttQ>^pI(-?WI6vl*{ftDin!lV^_e`;|kX!!1V&ZiGpVMksd@8xb=tmm^7U9bBoAbpZ=szAcgc8MWx7{)8yHy$_! zv*D$fgr7)^97!0 zJOyMAB7h!ma&G$norkbI^lr3OEzkRhr9u6fia{l5jL5F{l3BNBgGm-1{p;lrQXiVX z(m#KOU4k+TK4FsTT+gtpOxpBSewI$Hne~Q$uNGba5sWdvImrLV>@4Hg+#D|`A2NU$ zH;G!O?dRsGT}cAWuH|O=ua0N(O*Ybjn2m35tZwYM(*;6 z-b+sHQcUg6Q`f&#tRC`4TK)WH_PGZ!uYRPpA~uQZ+Q3crxzwA)>iIE@PTckU^n|1q zshBKt)Yjmtz>f-EA4TehWJ=QN=wUKSMc^^2HenZS|ZE z2|;3bgoWUQi<&A1vCpcpwT!|a(Y7z|8f@#%`ZCyTim>wpD-a@(=ld|W@tci(*0P0{ zT+*}uGBYN?Ko@uDI_hniNy<4zRlt|&FGgF@vMT^a52P*pheeL1l!qI=@z7I@EaTZ2 zXg3Ol74Qt}F(gG=w^S}y%M?f6q|3r_@}qp{Z{GwzGTqa7FA<2NMrr4mj9tq6F60G6 zF&&!Sh=O24vn{VtTP6QwU52P!mzha}3#+L1vV zL0iQSt%QB+^ONCEj_A7Ow(9;s$Knx7e;sXGy){$@mClBn9$FBXw6DwftSJIC2gUwS zO=2|jzUjs62`5nVS+^n($zo)Cv@sgUdDTb|3O?RF!vBn>Ffe4lBVN+);YiB}22QNB z)QUVS)FTHe7qQC)fy9p3s6hcv5*B+jztcmyw}+N9`nE6yJ~TvQuOMF)dc2)Ca;lb- zzO0|r;IuU6%m*zxW7{;^-6AX`nZ9CYB^tJ}1%gqg_#*F&g#B-3U;iN^d3+f5sodar zrr|}QhftU-J+SF+m={Fl1nFe`kHPtdzi(qKL3#zIYZF>YhCZo3;&GDaL7bDhn9juS zg@2lr5n!TUyvSu$%8xukaHa&LoGW$kw|46Y!WMZr2&Tu8l~Ge9=`Stw7`9*sBMi?S z^v65M#f(cBh2hH_|HOF}f~Jb&R8B%;c;RU34o|NeFi|r@?&k|d+t&*PJWuF))I@at zT>0?@vb?)A+H%plaF@-L@Vo}py2SX|U{Vmmyc8OGk80bmcl9bTy?#$%66(tdpXGk|~A7^oZ@E@7Df3DRFpUD-`XNG*8XV}94Afw*l7uRPE`VwAJS zb3s#{Jk~8NeN6G+CiFg^Y7S~{xEetf<3py8&F}S$8z*iRb5#~$H)Im_bvf05*9hci zYRI%`RPjG#4XDbFi#N-CC^IWXVe=>NPB^mpY!OtluYwSR(>4@l`0L00^|8597)F_f zt1P$+$*yOpu}jRAK06kPB6mFPok&CVe5}a$^M1~dqpo4`LB#85XhQ|YK$6S{jiQsS zr+@+^uu9m=a35=@M9aQRmoOE0DU4VgO^M(iG&#e`m4Y~RkrPZ_g(GF%-x)vPYRoqV zzfRi`5rllzGe8F7ttmwkHHqKr4Bq!}K~K6AV&69(;;}`Uk{n~%f|b@`$cxnJ=bHVc z#^<8(7(LwgCxnM%3CCetxw~Kr&M}DLulHYbTG!VB0RatgGBIvI+(I;r3|p_HI6hAP z%g6GYmUFa^5v#!m19xd$8JnAppF-rdX3W=v3JzNU;jhLA13sM1!ooDInkX=VLj+9J zO~p}86x+v0+>VEpU19kX)EeB`yEtv^KhO!gd7YxGQ0Oe4DIE5T3G#cZm{#S=PsSZQ zAx8&BP_HGA?~7em;802``J3$lXcoL~WizJ&aoE91f$f>g^mF%I;M>dPwI6UvQ*!Ce z=11^YobwsfO|$MG8n*ZiGawOi`3FqdMkdu8VBC9Nl<9}2SRE3QWJ&%7e_a{XbT7s9y7lcjjGC}tawT8H^Uc^5>M-dFu4UE4TjKhn%}!n&{M zf_7Pa7ylvuUol^XwzL}aLK5MsY%rsW$0Soltak}-h$@wKu&8793<|AMA3V4u77`R( zvJT%iz*{R=_GquuY#_-QrAVqp<{c+72dq9;GYQGjH^zQm$Gl?MRHk&VO27o$)$Im) zi5GtMf_eB_dA_J}8E3wFVIjcxOxTzP@r!_(*w@L#F%7s_d~ zbp6%O+*@&bH7QiTgXQwmqb6ut0SBfH7A5@Asz;s7pE9`rL>`c&>+o4!mEa~5z?PAt*x7I2IK z3*yy5?G-pQOB!ld-*Ylym(1*5;rApxQ3Iv9X+; z(;-45*dA$LrjEb^<7Qm{$Yli|1_M1lU zy*GZo!%*gjX4N1%F)b>lsn0S@^XH#!m{dz%+;Tg6PTZ$KC^h#8|7=DL6;=~E%6r~x zC~CS~8wqU1{iE42`SnDafhxCUPf;C7yC={)1j~s+%%s^D4^`I;h{YrH)i>k92 z=Q$wj#lc9CJOxsiiz3C!Xgy&UyyaMVMbD+Qi(r$3KV^BCUpPJ0pm)#O3o#t8YyRqrNG}{_4o>D)|-Z6 z2)Ub0cVA&GSzELu1q$uTf!b=N;?T{L)!`#ffi42%3Wx7)Q9m9^7XIzpR)g7Ti(AZv zBtfpKI7+@KFg=m<)`{v28sw$9ug~MW6hXwA_Ft9(1V!j3+cP}cAY}XK5FXRPj~H|4 z%u`c$nq&dEA7<#A#UQZXHMpELc`KK0u7)A{g1xWN8dL{7NDg{+saIYmKDR+dOP;%O z9MV!%=(G6rmK({AOBK4Lc{T(p0dDmu4myxm%t?>8y)=ry6nJI|lk>MzVZ zQM`45%Ye|MWSx+Xzr}_m(M_hLE?-|2RY>J8XU?Ni>SS|d8Nu_hiEC@XQs+BXl4&%J z>=)d)Bs~%Q#87~&eM2z@NT9HgR&m%Ji!7w-h=ox$V(9a9Ibe>lc4&ozWXdH?%@-1r zI9sE{zQR}xyE->{<1k~gw4p{sTlI@7Q$uP#7Y`=8(A2U-g&4qoX%z}T7q9@mgONk@&u1Yay6!#li%_IOq2H={Jx-aTa@5Xd~-&v@W=(}*ub!&C6YGH z(KkKbeqgLiVEu;$Fdlm?m2<%dDu#SJHRy{RY#>Mi`92gV{w)xjjBU_y|NEPtQzW zP1%Q-S~HX4?H0reXE%dt)u@p})FgsWx2ojbB9-yt*613k2>TLb16`^C*V`t}RSy!4 zPE#@&`0d)lYu(_=-Z{{&eS zVJ_z`ePfNrqLh|TkfhtdkJ!@XZ$^#6&?J0n2Ao*Y4Plg1z`g~h`ZOGD04Sr+0xh{P zz)gX+nq6^GY!kR+jj5VH&GLbANn!F!8=y*NF;HKYK}~ljNOw`}y0R=7L-erG{5iGZ z-(T*E5=N1XhSD}TLVY~^M=dHjM9s=})xusp>cpX>Ne4AG5oYBz1ZC{E{AK@}jgvc; zCCB?x=Bm9QtyX8Ay`di(mdR4*llXuRpem*)P?YQKh5i5(l`S=XMFL7Su93Q&j5&>B z3zSTRbHem_FC@+-4HZ}XIP8peO<;Kg9tM}FNx1j)-yhV2rv1;yDOpVCD!CVcDT2ko z`|w@z%wvC0Z843c#|tw~xIU6x#4RM^6ylO8jPwD)@0Q3;$lRotIX}(`WfIk!o|g`nAH>kkNR^>H6SDu|!Pm)^%B;_>+Lbl6s0q zbxuC20Tqd|AlZQm5O#N1R4lLj1Say@Yp>Oin66YgTG*8>dp3mX%`1W^Pr>@JjOJ=q zeQUF1-N+RP(>`)~$yiS1%|4s8_*TBt&c-rD;CYV-M^82P+GNkT-Dmanj!!>grswjX z3{Xq8Id-{It-;)x_g&CV*=S@l z6w5c{S_G$CW2$2V9&+Rp#Ty`H6{ns<>Ia+xDXq#YkC?7H^QsQ{l6yoTJ`P%_;L(!j zW7*|9Kw~L+pa1vZikU8c?7Aqw*dT-moS+g|xqrw{Dz#*hH)Ns^lvX@Ap7I-#{ulb< z4%wd2dh@B4!%V=d<4Ihxd5NpFH`y1-@+d39y&t}+;Bx$tZg{>+)^Th_D&9|uqo~~T zs%2xr(s6#H^@BQk&+~)6d($uIWG3_RB;?oRMd`XYiD^ad>Zz}%*#>jJWV}Q-jfd8w z&uba`gajdm7{-XV0SUTfQb`HB+P|I^HP=ALAiuP$d4*e@u{k>94IL3w6$wU@*m4rA)Fy zeEd@giCyN4+q9Vb-%zc1{d+9Op3t5wY1C(-Bgp`XrONI6j~p|5@FP(?&rsZHD?wfk!bK_K^`yP7!Pz)36_NG!4 z;R6|Ou$&bMi>*X^E=OOI)T)};sY}jQ9vemGJ{Y4+_gQ{fMsuj%)YD5$5%=7zeHtjU zvvOXYHAYH&2!7}|$p$3TUr-9`2V8hj1TE`pZ?&eBOntI5(OCtxC-T+;S38Ji(5v$* zbg}qF5|aC&<8CESN%=Ho2?eAtEv`@&aO7_^(c;jIm&067loeTcQYm=UbK)-H!4rephCxUY>&4hcvL>+a(*swOdp_iA~bn5PGal7{v+7*;!{AGmUmG=q<3 zZzlBA*W&4b_8~~|+4MzB;8?wqYtIZQxt830o?=FFY&!2jxu=BkOU0#$bLPft>>FTk z*hfGM^i|1?N)ND-=gMZ{uXdo4xKLQ9&Y-!`1zt^;??Xca6)BkJS-0-Xb5`~bH5sT$(wUXkaucayG}SKZen zbXE&5m%-xB3J`Gsfeq(H;4mVutS)&hGM%gBA2g^|ghaM_%UFyeB(#BA=8UZPo zP_f`~H!EvzEJLq3^+Pp+9?XO2;JaSYId&Y>4g$OypoA=UQn!RMGABg3ZR7utxrt_3 zZzS8|<~}3w zF9Qr?fXAl3OC)&6a=u(!C~Sv8Qu3+_#|Ju*KWOn-x|n41mfY$uzCuo#=#OmEz(~Ki zb&}t7KDx+?u14g-kcFGC{T6NfBRkO=9e+@mKP!CfKtwbTPXs?oV%&0j{cHA#N6{mK zv47lsHceB4f^uWX+p})jo46fwF@K}U+wMqj#@8B9uQ+gOU^QUCP#DlF3%m_f(n0@{ z*$K4{&Ew|frU0R?&v}JhTNU8Rfd+N z(W8I~eB}nH#kUBNy&+5rqB7dV3QVmeGfI+{8bA{{nHkXI%40LamZ$60e$fe}Qheo; z;FEEORq**+se2#Hzj|{V1)@~C5a1((R-SbGFs5Z=%T#7oHaaXuc0B0A9k1;SQ&TA% zRag4N_U$(U4@$rRdUG~IZ?n`#M z(wTn?LNFw`qE&;5*?;4`n3&aSdj-kJ=(B1>QF=&WgMs7PsNBMCZDZV-@^X&^e1v%h z#C%x8j;wRK;n;bjXFb$3upWs4|G4^~-d2d!oPm4o^rqHP35W4gRrNGY*(-Frg+64+ zv>v4p))Kf<*c>sj=o*(@^WLhzaulEuNFwXbS(TXVPJw-7=!1RN0-%jml{`!?ev7dR-8!Itbn%Cvs669T^QZXg?9)^rbY*fNFGg%(bY(I55Ntw0w|e*K0-e{SXg)wC;`$y2cXM`(Hvmv1yBHjOw_%cfUE!-lYa)v zj;GP|VoU%z6Ej;!4_8}jfQf@S zK#oa)383ic@c~)`XdE2?ra&tbI}3oL1wb9B4bV`NR8<2=t14+ItI;xjBvx~Ca&mO} zFD~M0>Kf7v00~h=bx8nFlK~*Dp{D-#st$Dc;BUzQP*nf0|E=@k_%~cZQe9MCM_H1U z`R^P6tN?eQi>vkDvj4%2>Vp~Jf2e(gTDUmc|5E@!V+8^^@i8-dcz7^by19aw99=A# zob3MLr*38K3h;1ru?2kGU4VALe;VWFVE)lgkQMM>3H~k$K;GI6=->+c8zklUFQffO zD?frhaM1rD_R$E?-<)>;atF8qf&W&<%Ea}bSb1e-d4Ro%wF3y~VB%o*;RrGTxw!(2 z|G_@KKy&JU5d;Fn-CSJ$)=>Cwlgs}p^WW6P96x5JZ|CD@;_;s=X5!%H>ixfZ^Y7C( zb98XEb_KcqDyJKx99_Ja|1VwJIyibb`263v zg|&mZ#ox}GyE!pyI9NNo0c9or%liXD_#ey?2m-JGfX)D*rMs|MWAlx3=^8H_!j@&;tJDN~7rLVsB#iA2w@ODQi!lxw19L%<5l_{TD6^GWpQH zsDq^)@MBT_fi(VZC%X^bew;nle~&HzBP$Erf7m`GX=dvHbae%=bNpihezf5~2!Hhc zFI@n$x`v9lj1>LhF?f5gV} zKNSD#sQPED26Az<1!`HFe;m{Q%SFKi?F36x?ckvIN6gr7COx9dbhE~J zVc4T`Fizc_pZ2p_4u!TI{U?-RHIAPKT>Y$|#SD>j z?(Use!=NxsA_Os(Zp;%w1zGrRUUpU`Bhoi36+ACIgFj7Gv66qw1q?!Ab$pEhUr0(~SHVq(Gjh*L zuAD`tRy)|I8)#>zk}>I~)f=1$e#FY+ByZ8Yu)UFhV6B;ySguss`H8$XfuHDdBzg=v zA+zD!S-!koN8709pSH2^I-}?tUqGm;9@xsc8jKUYfi0goBYg+x04QP$gtnaBoVGxXg(u88DEmtWxh4ovz`{({AZR2t3-$4OpQK%^OJM5ym52Z!r z=7X4c?6nRx2Gd}X@MtAOCv?1Z`1YSD3H%Ke$jdq573~nKtFJ{G1$g$amf>AC{qK~G#}IGsEdj1 z_IQ!MVje>H3kYHiB4{M-zRN< zCkH1T&wj_SkG-Gu%o=-cK0x zqL|*A@mwPrE8AlUYikVUEIf8%;_EG$cuTLYCVz6@JH{m^`{GY)8?Hj;` zj@jgI8R8$ks2{~kHs>tPVm3Yz?r_*&1tpxg1B8#OqnFzKvw>5r>Yp@FD+p+s!|+>* zKFLuyueNc`Gv-ZmKWWav#4t?k1J{>iD8SY^xKN>bUH`24#9#WwsU0)br_UX)o0pI*t>_g+P zBMf-MF5&}3`{?%v4He>vG~@;Su;2$6Jz%X> z+r}*5?fK5Ey%-Ze>cGvin3>k@oe!-NiyI@iKgG^{cXNde{+t8sU1N2qPb_DuRoH^pq!vcJ;`l3 z;@D};22(`o_qf7(Wts6Tnh-Ny&@&&6`LLA>x%Jg#{C+wGOsY21Wzz|`(bP+>RU!hy zJ|uBFMRkF%@!!f5YB&|IqJ{gjw@_s=8O2SoVA3-s4S&wR$BA zd|<=7R?-C?c({c(XR7qoFZU&rwxSPM2tvBdJ{rJw+B}JGLbHQaR?OYrFreq5Tep&M{QoTH7Cd zV7jF#S?-O05GR;k=w=CEN3I5D5JuP}(Tl~z_WTym8GDFg9d27FmUdzDUo0(GxgxLE zQ`s7WbTBC0&pLQWyFNBnER~^A??c8DR$R1y?vvqa@ynateZf|Rb#e-;B5`CNjLn&V zgUbc;E}dVjU3bHI)#%v)@d9dJ0IxHOqWx)t^J?rSObd2R@t6y^>(ONo zV$s(TzN83hSAzpbwLqzN=k^2C4yy+9MbzR%4}4LbCpa4>eWJT4(sdS(v(&y1Mg(IJ z>Yl+62lBe}B?dl+sJ*WkaV$s$FiE~+3F~qOR+yHJGAi}NNT8xa#~4i`Np2hZEEPA ze?!vaxT-Jm7GH5R`!7n+X19PyG>dlvV~-)r=0j4TJ%;<+K9d>KQly?lg2o5>X+zQhw7uA@2$hyBTKd% zc0u(#dwvVZ{Dz8e@QMY=(K2iNH{Ty`G*M4_HY{P}Wo%H)yravqF~$x&I&8>DO1M#T z)?OZ@kg7-U7uHPYAHxz0PFGU;NGYhR^$wa_4fEp^QIn${VhP0gRde3`vSy2%u%*`u zTh7#6cGfj{16wl+{Q+h}weZEcTL^e~N<#|R;9#f?${>3Az-1D3%XYc6W+>)R<$tSk zq>S}0`GZT`t2|$U6RL}`b+Ec&ri9<;uMy){mZ?mygh0ZLl%4-MrBm}0#paeAWp#PE z$ZP((BxwYW@4Ptrh|DU-<4m^B__EcTJp6KDX)X1y4moRbr7=+ldDV2H2hMYb&4cG^ z%X?b`K^lMqQNv)G#liK~%PL7cg%t{_;&9eeCL;OFhwg$5@nfs6F5bAW{qt76`cAZ# ze~@s2Mk~_P*6Z*kJzT8<;iS}}wOk~0q(f!9x6USQ$`QG`l6RG1k+6@~55Diie$yuR z6{|G)b+dEkxm=n7S+*BN&t^ruy}k6BbtANqS)7{++boEmHsZA~&5IvRmJw5eg52kT z_LqjHsXqQlVT1<^K1(he(=s%hQQ{+B=@Kl^OOV0@J45)aso%BNk2)F|<-B!Ykz#p5 z_$>WGsu@OY&bK<+W$&&Jj0LvqN4md(22Q3FE^K4Vk5yOY(FJ$$;8@LXuFDG!`<>i}*%r_`Hu>;x%lgn zi?}A84y~kpXgXsN#fQkREQAv}cKENs-vtL?H8fB%i4sd<{c^vzb6vJ%{yH-0RS6B? z+R-f6>{U6GW6 z0g?RQsH`_oQ<@r7Jk80?kk5U-HJV~_c`$~|cX}7^|LSb8t<&=1wGYR6Y0|v+4it=h z-PRQgih-!u-wusq7~-I|#}c2Sv$%?89lu*n-+xYvvT`xrp}>Wp00`G^Fs5MB6e z5hU(UotM=lS2w4qfY)Sn(4ythpl;|Bc)F4`*>8>(*H`Frn90OjZ}1SomlX2eu&W}@R-pFc`zCrBR zQ2`4Uu)8i~_A2(JoKSaY`xOCZb1kw2?BlG9U^R6^q6=;QBvQKiMp^!qN#HhzEXEV#*9n<9@uEJM_IM5S=Rh-07G^w23qF z6afPOjBWF9aMMGRPmbrpfU_|xuI$eCQedi!##rwPpAl1(V1-R&%@s8|H}zAWw49KJ zW%udOGrf09;9Heamm{<@-NdpPGty;7JC!bIeYxpf;(Z+I_nH+i);ypTD?fBl2q1Td z28t#&9T7(`MTGxJJNm6Ack_gB!N(fY6aM!nT2p0@OK!PmV%hYp<~P!ls$taezFEl{ z`tP8s&x$BtFE~lD$@4iXx=EV*s}EbNU3wYKkZrb0RjuFbY^D911_O%kt4?{>vRlBN zD=3{>QuhFA*5h1I|5pQe41>mZ55cRGy_OL z9s|MI$2LIg91KABc4G40!j=6~TRYA|B@&srheY|HI_rb80-^r?HtMpVimB%v?3)|2S=>fqMPj(Rzg1_d<6ck>mgu+#6v7h4cXaa{ z+#XOldS&nf_^k4A83-+@7)}zckT%9FV~*g1{;!=_2wF^6b?wC5j-jQvMBk803pBrw zyFGK9&jzKBeab+fbjarT32jn?>(?c~-v0eg4D&>&QN#EB9k%|{4z{gKqP$%^RwN8o zXqNFhbB>i>Fbp{u|u!$zdyP% zEnJY^xs1F8&;`azH6W#x60o+`8Qj{^ zQ6c~jt79JJ9M$8#F63m#NpS&WQC*5J&-Q0E$s4FqBhl;G@9mwlH?Vs)GMbP6@6~EQ zS_5jZ{!>n~ObJY>9YK*DAt9HcQ@ySr{77)ZJrmqs{%}y%xa=uM{5{dge*c{ zf$9+Klvvk2=ihkkXkt#pgzkp&5o|u>!zCf|Q2cW%t*vJ(v{=n3rw%1Y94M#1DM2nx z(j=$R{Av=k36#F7I-RnWdlMmKd6_dI`yQUoY{+7~R5#^(UY z&hbUC7Oa?;UAojPnxLFODniM1vKy4hlbzqQqk5jxT|ksLT9MmCy3*GaR7r!sY;#%APvh?k}+mf{=}Lye4*vi zD>efni?C#T_ljLNu{PMVh0l-4Xf34$Vf$;y_g;8s~L zjYa%Tc5&)5@nLtR-^YMj(QX++EMrzTJ&aHd)B&dj-S%~X4dUVb2y((1*~c8w)z_Z4 zkws3~4M<(c2Zji`yaEMxszHwD)$KW@+inCsiB*T5GU_*^2$2il+mK2>>Mz_62d}gl z9=#g3d~5zKRphkQVH^iU7Z+rSdHe)X!cDs`TwMyanT}=_o+j4(fQ-g^`=b8*&zA-b zsDef6;0O5)-EaZJO3Se6quqF>{``@+dd8v>OevgL1W?1zN_X$h^aWg)m>)BR z+DI^KW$#$~BfOBXKzFLxFN2l?QN}Ot@0&^Ac}K1*b7hC@5Wr4}@2xK26TO4?+<Rf|XipLkHiX%tmr{%<86`IP!=r!HHg_D>GqGj1_gFEFp{pfn^qbtGOTtn`<9 zRQ;G(TaBPNtLT4tMPW5r_as6gY{{(kz=Xe{Z|$=rsHE)~H!fr|E)%@DP3jc8*~cpr z8N09L58d;Zju}tk{33j8O^kzPvV(}5IA<@kY&QGxnwrC(#|~qK8G&^0(oT)T!o4GM zWhTd+8UJAy^{C_@Nk{3T&M*>`uCpc$N%^CtB@fhj?;P+Uf^q z&vHe5|DAkEnpk-CeruNg`E|n#r!Z}}vf@RDJ*783)E3Q4PaJez zy;0)?+AmE!HZ}k)5oS+b2BNy|cG9~{(UM~nB#B4-i=!quEqLj&X0cl-3tuJ>>ylSa zSq|~H`{@rcuj{yYl>f$C@1;}i@BB-k8!mR|`lGv2Y1M=6R_@sbLGsIEFTid4a&fURy-Mg4vI;EMR&Di64}~ zH``#NAFKhAH5$=Xea$$xb!hHB)>7#R%{?i~`UpSTSUsvMf3u#UyzEg=_ zfZA34Yf#|^B1&G&N%CvN^B{QiMVpo(*tN3xxpqtrEkVDdS#9*^Y;aU#g9*PHQnRn( zr;)|^o)DySPR3w3jy2yShFM5t<>)Njt`Y#M>J{kvFpOta!`wZ5gO;pn)`DVaL{5l! z6{6+p$v$8N5Z&V*X}lBpnu_RDqDHFkE6Si3s$3s=Oq&X5c ze0PhuDF4hdRDde<3u7W@yKk7)+<4&M_|0%K8bS!*T26E+jU^TkWei!Xp%U3Tgjnm$ z(~BX6VR^ViZ_Zexhx0(5pFc_to-FMCu4-x=yBEUdA(Ng5j~W%r_@l!kPjV|JCKw38 zx|gO?yuDA~7#T3mkCvXHPmID2!${PYb}Bv1yqZ_2=ywex3%V4jb4{|DVz1W6Si*8@ zY@b5(4qhBF*%ku?;-a5e*8E~qh^DM??{_i$n|k<@q2X1zL0@XbuDMfcRfZ`t7_G;+ z2FmvDCM1R%xj)(GoP~awdRJBmE+y|)X5@}qfmMa@=*Q!#Ezk@tOvA9zOELi~C$&?sc#YesCgceElJx}a!qsj`XI=}N|52=v-E1|IKv{v~9S@74t z2tSyo$y>q^#r{b`rypS$=ni!N%9o?q(uwI;LtESwnn$~YAryof~B*2 zF!&*bba8zm{fLgoBFMy!V7k$&X1+%EV(XyXBx{UYBorjDoy2758l+EEFd$)lE&0Z9 zHBgJEhCAx>E)O&ym9Y8%V5AY$gHt!RYwWy6v7<-Z%t6@ql8liK6QH*Jdmi2un|@H? z0&MBPCiXtk59x!Yj!=5VVCfl+@d@>xYwE!&V(GVyM$!S4tzF_dGqedq&|6;vm8m8D z`=~Q7?~FS}Q@_D{ifXZV%cpT+rkOcU45P6w*z|@9*$^-17B3T%pH5>A|AxqKP^i3~ z73iT!68lvriO;0Cp3*r_fLP3V5K{B3;X7KMnAR_KTyJS5jq(e4iqX2 zXXsh5sJ^qu)OqTNb5t$|moic{#NA#p_=jBn>Q^nnBq31&P#p4+04rDcbflJ6e8=Sd zE|ZD-D(&>7za&D6RWmXrE;qkkhe)%rbJv0y0GZlY$mjTY8E!g5NyC(2w*{w0{+^{s zJ9OVAZv**Xknuf!pKHI)8#d}mQgJA?eA~FvryS0_=s)^{in6|NceD{KWe=^|7;FU= z;udM6`!dQT^0*TiiugjWeaYDIIsk-mOy6v!51aQkHVGt;{V1nibmbJ^V*qQ-+%}%I zY)zwgV${oSlKA9mXK@_qfHh_N3`|Y)T{ewULF;@+ZZoERXtJNe-MP^!p_ktZg~p@h z(yf15uMO8H5*>I9lGA(eagGNivXVt-KyXEjoRh~A@2_hwjrOT_cf-Ka1WorQ1ch*C_fwdLDNQ4eXa76Z6a8Hfm8?$E4F} z+GlM3`nafS>%SF;O~Y^}t3j)^u+m|?9;9X{{fk`n#VZbs31nLTd3*+b1_o(Gmmo|* zsVD|?pEV92{|RVY+(+SGX7{_Wg;TRT=yQm!k~Yps^f}j^m%{O+ z5n9fjZTEDaw28#?Gtp^z;a#Tze|sW%m2_XJ5Lq{SUnGIDqDC6@AW+r|Xm1@Gp}*O! zY3$T7L*3~HCVRf{Xyij?k=2-`X4WkiveUjeBloGQ0MLSmXkPyfm z)fAoageC9ACkX2cBpFeoeo4pEy!{IyzCv0##S0|F35HXT)4059wpa3?V96U;_-z9Y z<*H5R)4k1^b9BwUA(|Oi6zX!Z5_4u|P!XWI`HqQev4YNhlx+)``U^vb&>N3RHMRZ$ zXuVd5X5(RZCcyM7w8%-)zYrNgBkD=e{iv6{qc{ukD);mYkQh!tib2lf;~9zG2nIIIYN zh3Feu1>9ZO+Y!{%-+7)m`)fk19j1w{$uK%>thSixSmdNQZBedNlu+fhvJx8#91llR z(k^_T5GKh1$k&&?IcG1I)+AP5IC{g7iyFw`e~N~W_gkuVL=G5x$0!&v9(w8ME*|7q zD67BJJs~vQ&6uWmK`~-SeXHXpTkE3hT>g?e!!O{s&=v;EHA%&g6AiZFw;14N{(@|Scbl!VEweBRQgiC5s#?qG8!_IY zWwQ0j$Aoy<2N=E5psT>fYArT3?mx`&ER=T4@gCzIB-hM)krML8qPBUskR;Dn;jui` z3&`YG{#-v;7bhbT+?@-bfud5IgqlY339q9><-lo!e}vkO&;oIkKf)GL>xmJ5FhNMZ zrrwd#ex$&Hz@Tzqh}`sUF6+LuZJt0}cMtGJhXLifQSHoj7AyZHiU-XOvZ~`CG!Kn< zb{PRBDhX7W2)qfy_@4g$Im!H`j|@mKGqlqT7CtW?rGuB4?rOf zWS|=}y#pA!mV^8J^5)->>qQ;bo-4JeWIl6r!PpWxOF&8BHRTsS%W81;~@MUPKr5~F>FqVv;g~{i2J2sb33wsXdu`TQYi$->~vmeA~#pGx{ zX$o!gIO_4=<#fy;&!2b@x zpi!20Yo)0h=S&y1vivOfM0Z4%$%vGcW|j!2izxO6r0Vuci2J8(Hczo~%{oOYMJImT zwCFwh_(WH2;K4~*VEQ5)9~-Y+90%x_hC~CG{fMAl(I*C?{}BvECIgnJ={Ps$l64CO~m_ zfV+>)6xiA?O`{%mB7ysBFq8NvneNzdV0dZ^!72MPGXg_mSZU~DYM}I^a5qEmJ5yc9 zO(Zr7smu{3!^Fp=bAPGcu%hUqdDxo8lioT0IA?09+|9`6&N_E}A|>wW%h=}mS7YpI z5HkCqtX%(tV(aaOr8udoS0&_?^uBQP!v(yqC;^_{bYvS_@r!|>@}ZlCH7m^%TI)e`H4SH6u$*{j zp75ZaC1GqBK;f^Es}d7t-_+_EdvDGBC%rtdFM;7GR?rK`9bAYl~yv@l1LL)Q2>g50xe7Mt9dfR*^ot=d{y6yjg|}V4GTsfkg4?S6b-)d za%fucHAry*G<$=$YBH%DZI9>N%UWaY(;-BCYkh(_I&*b=_<$C%J5Pe~A7F*3b+ttS zWC6O_19hym28P*a`$d#L*?VsjAX}2cxyx6KsMT5{8&(G)CC>pa_AekZusm%n)0gJk z3r(VqY-oq%FB7r@vSfKGDJftO#OD*%0&1Gg9e<0I@0Lnyw^>USXQN85o}N!kfscps z#E@V4V#ucM6kNb<9ot3KKbXCba!s|?$!F~eItJq`+Mz|)aY9yI-1nn$BR1)M*)w-q}uRX;kyIE&TI?^Q9`DOQ26R1 zUk1Nk>&}dx?m(5PUvntH-k#`vu2@7Tp0|Vey@%X+J(5%3_4h99AnFo??dCn45J6y6 zB&zQhN()zsgjWsq#FD$fhSE&eaq%rmK8fz#H2OO>riMSB;?jsENZ!BM0;bd$&h(OA zqBtQN=Zb!4cKD{*jASlwfwtVpQrYR{JA2AGzr-pClHO~NFI|}Qw>o&A>O`g&XF)WY zlG&X7?@jL=H!VA)vCK42HNtZ~4=qkU0BzUd0Jcm_IJqB*%m=ow zbDSti^~5*FT}*0B#?1|F3A|t6qLhunKyp+g)1u12G~5*{1Y4G<_1*hE{dD|W5Kd!gUlTd+Tc(sT7TYQA6|Z@_o**~ITnyf0MR zB0+!IuhadPIG zF3=G*iaH*?L|8@C!NlQ9jA+A>^tEB@;yi5lo9*N#&a(|V!DH}h8HKT~ z!FHSaq}i>goFCo-d0kJEz5@%1NnkgEq?l56s^fS9PZ-(!aRKpERVKFIF%1(34&c8}``hEGNI#?ul%9t$1$@~{J0dqbsyd(rM zxZukZGSL#gXm_Lzh9GggCul5#dP-Q++h~RK#?c#<89x3bwF?<@bINx9R|{dgo9GSN zx~a!w(_;l3>h@!eUymaA1VLpM(hxqTS{+d&Zwp9}|1xORmM&GFXZU0~{v< zGi)+0k&=J$dN4)UJ@IRQiedmAKjw9%U6&4u8cmX*p4q7mce`CdO|HLBgNr%B{@(F} zJ+t+chO~#VA4dpR>OL~-SN@-ciskY-dRS$i%tSev7yj51W7S%g@!m70nw7YOpe*0JfUb}C8p8^f$>BWIkswuNvlmhUYEy7 zuS?0}jB+`U7c8Fd*=-aAFvF@SMKi#6v41-PVH!p>%fm*lVhn@64?Ju{4S0?CUJKy{ZD zb7tpdY|&p%^AmRHWrs+^Q%uSl))9#tQByB8%k2ha$6ykEK27t6`TV*DXZ}WFMpGNT zGPo*-XN;e5lsPUq3$||kjC}nq5L_MQj8DII`g33+E`zxV=W z&N00wl#*mUBXHYWns=U~zsn9bboEqtCuHo7e#L##ltv|wL+?G{puDxKN3L~cnPz+x zVp2aMx253XW1plyX4`muY{00r>;Fs)C$g^{PISHAEvFCO%ZTn9cKLRE}Lv8B_wYqhu%adoScR_KL+D(DN=&k;h<^{CA zk?0h8l)aswjr(X3UqY&1H*=%U#HIFvor{#{*wi^)`xFo~!Ksd-F+JT#x+7a&&Va_o z-S6FOk#kk(#Z;>2q!$=DIR=XL$IiPjh3qnAz#D4vzrD*u!Ak5;84*kaD-vWLO2FzB zS{VX*G0d4WlI3F~$JuZ4UT8TCr%Dnl%NR0Wh=i3A6(FQaHH#CkrA?xTC7R~4&BEG- zK9zgjyPo-4z8vtN%xhSCTN}VOw|PM>gj%$TH?xv&#)jB&&aT& zt9RZnE;I*{HxWtGe!gn#Zdh~Mz1p+OjtQ|fN~G`GVKo>l!^!z@3qe)M3*etzB@5qW z`LUFk9lNbDtU+33ek??jHSkGo5G>7{yLo>=v@iu9%H#GIobE*`x8X<{%R{*;JGnfc zcqe2*MtZSXcN#QPiE0VWza#$IsX-eg9o8vf-^xNi@B1VqFkrrjA+@QDo5oYhKP<-si-)B6_W!QJMm#lLj#AC6j!tnUCyERc;2l+0dg?1$Xd!M^@ z@DZrIB}~oG?D}E< z@kEwqe`u^tajUA!K;7_XnWvZE(r)o%pXaEerf*%tI7q^~5> ztx_C6T3ty+CV%|Cdien(AVFH>x%fp_GT{dc3IFv#)r`5I6O7Gf#n)D&a#5pEwON^yD9a}N}<#db+9 zz4#-C&Fd7K@#wgUybi)#WnS>d@5%Z5ykdeg95MZ`Y>Qa46s*`u|5QLqV{8YXP$Y4%h@57*%U}^fcM@sPU(|D=h>w08Z7FzGXs8 zEw?5A$^dUOXbR4!C{rS^VS{21{~rF$%lHl{E#g{nQMR3dLW-@8$G6Z`kH{Ad*|K4* zPgqE;R{eS(xmNBtJr@IuIuC!*c*3&W%*~g1%x8?*GThWh?VdqSbr_ump;I3JrLy68 zz=-V(S^sH}`IBPCzeqXRYcn0Hk`@wnw(Q~|vFwm>xgpu@iWwlmk@9Ki1NMMDlk6_m zS}rO*;^<5JqIwqQY2@u{X@88D5E9xocG$ds@v`(dX(hz7jkzIEwdBdOG~o&lUlcX93{Maoi53q44DSFunh z$ru6Hi5A=G|Aw4FJ$cLuL~@BO2PH!JS$Q-ipw+QoF>p)QTK{WA!_8m7PSqyzeQ-4? zi!1+dU`5@54?5JYQI9jxpI5=A+B`RJ6_?!2u;SNI&(;y{VFuCN?+mlHVR7byA$R)c z&0ice*JXb^EhPz+kXD~u#+ONV_a`v@HRtV->zb*05ON;jDF?X+~5|AiQP1tVy2MTE->$Sgq&*9 z+5XE3z>6))F+kV6fHK3?bfUfCvocxoh$m2oXmTR*$zPIcaCp_!lfyUG^Vy?rDD%#!VqFrj_>ey2 zw*dDstwg;yb_x;Xfb~`FY}Q`NCp5GJr?qYCcr_9lfnwE>>4T~#Z+pqjSE4ek<^{Xf zT=ehN1al_p8#_oGmZ4A0NXusZ_-rBi#tRZT8@u1zUDLJqJBfbNtf~%mXLtZBFt9Bo zXHB2*8~r=bWruxt$(%+zM$(KmTC-HTUTn1^ae$xZ)20mj^RtFjC1Ph}2~Lb>{jssb z$uMFQ9@p@XoXj&WVpiSMu7@6ejOW9m!RW881^m0mcG!)<#! zs2UHno?e%8g5AXpbjLw1d12732hvQFjByD~>l-Zf`ra4Fr}Czb!?T9(sfBLA>m$ys zFEz@zBWX!=MCI$nRM{U&#gmP@ufSvSxceTw=R`%ZGjL%pF!@PYoFrM6C{BcM;B9YA=wy9uJUJ5#;oXdhsgN1C$W?DCZn?V-a^Z`WreM-Dh+7jk6;fSjg3|dXNXT z)_V3h1sNr5h)ZXbkDUIHeiUi)@EB#kxyTmgsY5AAj+^q>CK1hW-uJ?uuWH5bMJdF8 zHsypq7!sOx#8P&z8Z{m$6H-Kd;E#{o0Gi(NibI?+pzO2oEKHYhO*qcdhj0vo?#B#~HRa{;E zajhpO#rj`wL5EF-RY(h<&oWbJ|$~9k`?4{&o?T@Cnp!`L9n)6`|B}R5oTrXtx zll78)PsPDsR=#$Z?#{p;ip@WqFd@}(zkfdhNM^z$t8Jt9J>kCKkW&9ty-s!KuYILf z-{oR{?ANNZEUi}179{V+ma4kQ(F^<;n2sInO+8I_Y5%qgCxfHtRp=axNOUpi|2CM$ zmGMUx-5unUc%uB*Y5W;ntL_Idc1u7O2+W5vm9+J_kOjthBx z$%f5VE4)NkZLwNdKLD!LD-R4tke%xTkv`l<^Lq*kYxv^T_#Mc19pg&@v*kP^52&e4r$ z1)Kl@BkaxT!Ew>g6|7MGD_$)#W;*@($5yc`2&McnACrMO!zNkwaTT18WS+=cElbq9 zqCSdnGw#GPIH+r*5^9OrKXANfRM1mipHWQ+ZeVjDrTR60Y2$}YF%-vn9j=YWLE(InR&?IFxmw%rX-+VkQ z9>v^cEt>oDUAb~8o_cNt`ObCF71A@UnwO9Re<5qs-}c@FI$3uf$pZF8)O3O8{*l(z zfk(ED=Ew@9T+^W6!?8tPcjb72{Bq^XK36c;(J5jsXGiLLnhq^OOe3}ggQ$X8E6iH& zbix112C=i)C%&`H;}Dg0qTpjUus~nLEch)N(hyN zAvXj#MI#|1nO?Rq@-p7TufHJKk}`1YefqxFNy_URt%ip|1g^4j`&klv^~_UOn#ePp zL;_*qehMz8xrBr}K5SK9te_eQ5DrK@)Ns>tin#?q>A>o?kbV6F#sWY2XetufzEc-d zx_Ln}1WS_zE?=F{Q$uMCw}Iv$Lrc5*C=77})MzVs%O-*EL}{JLQUuJU8MwR!mzb09 zI3=zX03AB1V8{JL=7rY@?g7bYzav4yyf%HDna46}aH&*(*lHcM4w7_ui)){cUe{}; z5it@y=;7gA0l)5>Jc;B;`3)f=#!To%8cI(X2vg1*h&z*H*oXMqdQCB zSn2p#=7x*~P9oQMql^;i4}XNnTzUo1hFzycI0lqn)w7|~&+ z5wGVbT>F;_9DA-lV7h!RsXl~%s?*NBQhdIl7%1J1QAh&hspla1V6>*a{@Fmj;-o*- zuS0+1I{QY7-B>ZkIt1|35W0z6iiJ~|fpDO$cEL6^w3H0ZZ?`y2Xmt9zGt&ThCL{GF z%I+USlhhxyr9sJVy42DfZNZ2gDksj^4fTGJSLja*H^c#bow|x(!bLwcmL7CK1bP8V zi80TObm9NG`PYa`mX7>a73eVHb_24?Krxb#O&KjP0c7luWz1cQ1hg`8{9foBnZ(~ z?6&T&S>ZO^lgl_$J)?^9W`lo{tbG36<3vah+Y32y>`a{x(voe8BjeA1lo229qx-O> zJlQTCA}p^=j&!&)mK$`%NCy?t%@FIn$7%Iw)~fWqL*daCXw#T>2V0DNRg2ZXRLjX( zOOs!+BdKzO!p5J&1NihI`yGbOR?lFG($DacwY<*CqYU3HjGxF&Y@Vjg>cyTdKMler zEs84=@eBjMiEcu62B(iW5`>nw$4h^qdOYXx8oYH-UeX%+Z=1(-+nUNCN|(hq?^;Qr zI|nQQ$JHpS<2m|UhX?Kozc*zxk>NCtwt}gXkXuyDEdd#UMKY2_$p2M1nfS=ZR{Ou! zaU~G0-$~C+`(Hb}ML_lxlJLTHj+df|>$u#ub71@l7|z~39fMACeSmK%u>N!FK*7zA z#mgZ<)zQobk^tXwjW7>|rxp~ddX@p-0gZHgeh?(R{)#D)Rk8Mu_Cay7vmuf0%K4!d zZM+zuG`W-8=|_j2G3~h?Aj@q4&;k8jXXT<_bP8mKgLG~?$)a{4Ec%bhg`#du~KoPe zDJ5-iFh%5XbUL%Sp$o82*H~RN%7-TE#$&d#MuoAfIo3z)b=;tx0hQcJ22{FTMTynUF7;PqwTs8Eca1q1i!DFDhs}f4byn$Zz|XQ ze_%dL4F3Hm;DOax3E4D8JRXZn9*K1`e}98CYem{0qEn9o)#Yh+iKGN{{s z0}@^k8cLxZ91f45y-V00@)m7-XNR@@DZ%Hq!z<5eZ*R@K{#^L$+G)C@5|uO~lB=jZ zBL_-h68?Dle_K8Z3Mx%atl$p}l|Z!G*3q~?)-p6SIFQN7hp_@iX4KsA!41s+Sw1(` zPzr1=zVbeC`UVH0`JfUYf&yc1WtjcLAo-BaY^(S65RJg*(f@#{T;*n?$wuYwYzf8( zp+Sp%z#j<|Htx?FX7+#AIrx`*@~N;+=|+taU7Oltauz$q0z~UT@%Rs zTHp_Ien~^j{{lroy1zNV&K#ft)COp(ORA{@q}7x)Rn%#j-V3X{*xTDV{tp*%bq!5v z27rX9l7=Jz_>lo1t*Nf@=dT73^v>Uc0idMu?*G&0J@8MuqNIkXhK`CPEAyW<09XO8 zKu0I5KW+bu8`V2Az~9u~Q_UUiZ2xKips{pzw&!DJc5`!MvT$*7X0mg%V6wOQi=T$2 zl@q|t&e0n1{&WP|0RIxk1!VRvr?VySUj_bb3P8ci6bNzx{z;Ou`DF{{qdZ z|3wf85O;BO{L@46-!8}hZS!yHVs`H<)3^EJZS3~nJ7x@Waq|3!HvhS7Q#+88m6Nm6 zzbXO&=2kYqKkS|U?3oqlFPoyMlB|@Zx(1`d`{02X746>D0WmqdJO35^r=F;U0v~{f zg$uyS%K>0{A5uw>nYf*;?K^8Hgg^30SiS4yZ0G2~{D0=!8f51N`tpD6=2jpx^FPX) zx!5ynf~*`|fU*+*6a8*N_}gXybOx{hfDQnlyQwAfpLTyu%O5lAAM?8m-e2tP>;dM+ zHcmiqD|6uc1L2F4u`3YZ?C1jY{_;=9e5yj`WNB`FpK|bQ= zUPbY5^j<~je-RHGfLZNt#13Fq{~K`tm^J=JyZ~mM|3$p-HI4s9?=?;SM(>fP|BE=^ zeWrFc@4Eln!ol&!Vr%<%J}b-nV$A*lSpm$zf53N==6@G_Cuwfw`nR9+kK4}0@gE8A zAr}9D?+q>gZvQTorH8#G5cH3TcbnBe;5%oVf57(z+5Qb#-^u=+%lWPZ=zWC#;eTgu z_jkMZWV`=xztgq<+x@;(_CQA~yT6yu{!Z7y#qRwMH2DV!)_3bamdN_f*y-;o?UeDw2_V0Rm z0v-QV>_3l<%AeQ9UpFGl-{#@9FXRPn` z_n&|NXYd~lDF42R|B)>wX6OEek>mXcGP3cq0a!WS=k;B9PVfK7HT~DC?XRo!{lNbl z|9P(ffIxSkDZ=8Mohg6d=cM}J5+BLDi679Eyi6x$I6~TTF;I)C6Rm`J64_fHfWrCy zjecoV`F092d25 zpgok!%GD|2m`ZDoX(j<^4j+m0wst&3vW>l=UnYLg|IxgX1MOBY!`gu5XoKOtfJV?p zxG>QM?)>(E6%}FJExJfI{V6vXXIKu#zN-Uszs%4QLpbAc5*zguQzxvOUXJMc znRzZ;z3C@}E3Wg#KIXBU@?c*`9_1iTTzRX>I*^vwu%RsgUbcsG9gZd=ni(5|L4^wC zj9z*}yzJpbcN9r;cy>sq)tNPw@S8yfK(C>E3=z!34ikAhHhA4X;UW*>#VP%>;!Fa% z?+x0_bOCMfQY(ty+Nku74l@O*BI()jf^a8;#@mxt9{3)L9%WTcX{3>?ezs2G3Fg~Q zK(#Pub;3ekC`>cd2*d!4lUTbrAfM4nTVt6`MtVN-=&*svV?JAAw$@fz=J$e(z&YEd>@;gv_Ei;$rj1iHfLtj*Q;6S#%Fzp4w-G^*(F(5 zNcP>WNd0UyvdJ-jOWADBXI#-67f}=;Ya+aHHZ7Pt5z!wNTnxJ()0+N#tRH!e_Kle} z9!1Ztq;~#6@hNwI4~fv&0Le`mm%X15X%M9_!&^Qk7Qs~nP!#Q%n~sB)ZFJdyr3I4| zj}SLiE4rVqayK=Y*VR@;)@Y8O$A?OCyBWc093P1jhIeH2uv|g0*V#cb^Dzd+>ADM% z_R>k2b3>U_6AAVL{a{Ow?)GB7j)BPtK>Rrc^_Qoc@XaQLxfS!|PhbfuIwU22X@4in zcWOI)RG$@D{+t(q;r33IyOvUrf5Q_%<6tHBMk=98``geyr8^*Wo2DPS9XB{k!zYYP z8#f@bqjB9$1kG!cIEy6VLT;!8s$wX@91igHCcSZYzr^mvv?Ka9h4pH-?w&}9To!8f zRexv^mj}mwQX|sw)xqeI<+ZsVTU7{}5-KLNX_4^m_)S8hv_Ka5SHWlm;ng{&Xbx2} zl6$m=T+q4AHsc4+1h#sT&6h@ju~1o+Z$jBGrHAApbyTNFsj;jo0qc4v;3p$7CUej@ zU>3)e1mPmnAJXyOfNFs06~_iCys_lbhHd-oW~l79-szsoayTn6{%t;_entO<7UYoK z&WbpCad;k|j1H?ly7nM2+khqrq)xG=grqS721p=J;)aI^guz?{`oOv9*fduQ`sdnq zW0$#AqDnt3#o0xúUAo*Pt4$=jz@2%G|Oc_Z|!Ew@*0juiQlSX8*_+RyoU;<(f zPtXVw!t-W&^{`DL${p!3FN_=(dXbnWUF{@K>fvAJ>{zg>NlfRFas(6GbT>1~-ODBV z2QwCl9Y*McQ0_!9kay^>U||Lj#ag&=gW!%0tC->ZR}_F{mXu0ANw?p1=Bd~m$w7`-B`ERM#E_&Yi@=wV`yMS>C)ER-J|AY>oy^~2aC6rnO{T4%3q_%RtP$)|6+OOV12}_WuMEkpdVGlznurunTBs* zM36b5AfTzR>J3*}efUJ{&AU!-AAbHx#pMj!nQ+6I&f7Umuh)Ahn<<$&bFv?-0_%=m zQpMRK@}+$qS;M-2s}E&nMXPqfMRhEAW^36M_5%1g1m0mLgVdLE5b4#pgEZIsXf@9F zVM?u{+1y-KVi~f21xwAI@YeSV`EARC>aLY*3ZFeV&hWEUU+_1oNMa#khk30vY?k%m zyC)rwb0ctuN>RbmVYcY4RHT!9t(g9q4QQva_-v}GM>tZum*yJ+SEt)KNdHvOZxDlR z$>g!74%|4j&uH_Mzx2{CpGFtw@CwS!IAs{gmcYWQUIX!Prwrn(?HV0o6{LMp(J*NTqRYbIVu?mL*If*j&jWM9c(^Mstr<9(#ypt@@EFxv9Y^eC1MwOSWLUVDb}Y+G;nB6#i8@BAdQbja2l#- zIh0WT!`;~-eUa6QkAiOP4eDkVA1~C20!Lm^l%BZF1(>Sb_b-J{dJ-r2jKZe^J;bSH zNJy-!>E1mG(t)ErDgy_^bK*w7in^N!Etqo!vqJR2NL-KDSL!G<+zUoPdr`c_B52Pc z_eH>0nUAo&F8J44R#}hc*4&D^+-sIY5J}j@^el{YwJ7WH(ILV9B}5TN_iRXGu&>m} zYCEHrTLF{7_W=q+zGaKn%g@~H2a;m=-=DJ$fAwti9HH4ix}8jxVd%T5OufXB&hcGZ zu)s3mtRuqU2HHgXURr*t@Ma+ z8#u)*cci?1u;~F&;24#9-m@&^Uw2H!pXeqzM}*rSpW$qdB$+viB{V+rk|jP7B&+Su zBj#Zw#TQRUM!fE;>XCD48lxs`p_eULypY-#?gl`B|K+ly?x8Zxb|7Q%vUiR6{h=Li zIILR^*p}d}X*FLdG$VeoOFCF>`LUsK1iT%p_E)Mm?$sEmw zQG~WheEv{)<||^oc)-_S$RZJKjM3&5N$D8EM`ZfGFg#m|*{6f>PKbl;ACs%bORJuw z3(QDN&I&#ChNs*as8b+ZjEJHRgXhN!FDmxn(O*A6qK#1fz!1erEOXWMq36mT8Fle3K zH-2iw$r?Jz5f&@MfhzII92$APWnxGUqZ&Irsie!oEN|;%WZX83_V}BKt`U8@7qr4# zK30+PuXsVQxOD_&NX;jyvSlSd5)mb<)L~<#F49QD4^nl&8(&<^hW9A({h+n^@p+sU zxKCFi^aHh#{nGDh!c_JpO4Xr}ws1Ppsx5@a zwZES%jZ&(!m=zExjW+U|4X)G7)LEgM&t|0sL0PdZIA7>p);38NAE(J|c#UX%M!W z5D*J3(w+BVD-KQA!#Y51)9{+0m?{arV@g_tQ=ZJ&Ux53nm0o0Acq zx-twTo1esgUd*&@g$5svIA{g-claXQE>m!1`3LBTq68^CUHFTg`EIN|+d0GJ?F#6 zP@1Cdasx1*yPvYhNFK>r@}PN0&*7|3Mo#LI71l50BpDVCvY%ZO>IU@V*YiccRxb+a z=V+fod!;?#pc-pS!JTn|*m8T9DIt1$O9=ZzpD|Re2lNI~$-BpRxVjffo6HUEpiV&$pU^w!!?+e;SaYQ59lNDXWi4+S;QF8tzx?OOe7VSLBGN`CY1VYh!4pd{o4h zCl~pNx^>|WN{UY!r(-v+^|+zR2zs@LxyR&^bYvl?vRA(0nMt z{Ehg*+L{x>*_=y;IH-t|E8YV$i~N{0gEmfmbTF-Cj4{^(Mn(`*I}a#@s9TJctAC@t zyyj1Z5q;!Gn6v}kDcJ)(m;|jlX1ZhGeBWT`D23Qb&JXsAAl@>ZmPMcivevmimNphm8I*y9OeW`?JwJF3Pm*32fqFVfry*h)OnTGTW%c z#HnqB1J6mqjA|G1gDkN_ zCM_k%Ng}GqqziR5?G?F{5oW{*&hG&R;`jmZjE-OthsY~2o`P$LkZc0tfsT@RLvi7( zoexC;&T$Ht7et2Vd@gS7I*(e%qR2K_;S%Y;*S~Y_Pf6Jh@F!zLjeQ`fv)%{C`j(2_ zaOXKLE~q3U6*$N%U&1Q9eG){N0jKM|$uGJ*^~r_b3=&pHnlUaSu!sm+c3CGsDe&5xaW(7%g@dZBK{^bdKA)EO6yeKSR4UxAv&<#( z|CZ6Qejyc#B>jTPNi6Y%jX7LHH*iUVFIaqA7;-92=rw*9_>D3MH7`I$&D_;{Fhf90 z6BGZ$Maz3Y2+XCz2u^xhwI~@*HFGSoxAe#dYj2TktqkQ3qX#l2$pP{-Hdjj(EKdj` zWY9MT0tZ{{i;Lp{>Tx~-d%|icuuY^4w_GPvQ z_NWw$tt=*$2ZGMfi)V|8T*&Ue`y-F_+|kbAta!z`ZkUu9*g4<3f#h^m#%%MswJx@P z3QHAs=_3M34mt6S6?Zh`*Ca~R7f@?+wAiEEqB_ccNzo>8+G2%pRR`Iq?W}}*cky=@ zCx#?KF=c;tbD*{S^SJ;c6=&RZPS!=Ho z1<(GWy9qI+-`F+jJPXR62}}! zUfV6?ik`*-4;MCt!H}_;#FmLjX5qyn#20u50L)l!}U^N>*6eeg}zvZX$=y`ccuOZpxXUAj%5ryb=S3$G zY>;%;bUst+B?_s-0u{ooHuP*1%?JW}ByM#>Jq-UElItIl^JDcZYKl-pX-?EIJU6`2 zj*O!^h7zjZ@rkDNh)BnHQxT2MhkEgj4)*hz;3uQyvt+6z@YSe-6CR~B_eWiiDv&yU z5n7OQ-;;Yg$`d`ECo`WM%%vlT$k8?&tfxjnHBkDEb-A$zuEa$obY7&Ngwyor-32&3 zM2DT8{z%a4wlT*Y33u{$%8}=eu6R z#eE_*`>eq!?z3Hd^E}@gD4|K&LeBIR69M65_a-qoWER^{5~i=kTo3K$F%9W^iH^s$ zB!2l8w&c$!(H9&aH_3^f1bX)m&MzNX4+puk=s83x8aa8z!y7qGS3%y;4t!~>Q1o>< zc@5$T_vN0BLS4`$UVHjYxoF=R><*|u zblAeVbZ&GCcBLiJV-~)%f7(xwwzS9jM(SQ06DS_BOM{2f%TX7?ww*n5}YHRTR@e^SL=n`xj zZI2yuu{JBD(GFFe8)sDvw3J?!lZ+)$t`5pnqMyH!j#ZT(Q2P{kdg)jg9^EUmYg(6_ zV-HeT6`2Fcmo3XiH)-{ZjktgFYEy4z^5LRC(c8T(DomDPtwoGW5}CJ_o94Ue5-K-naY$I68Q~!jbqady>WV%E#2H*{U+| zeZ+ST*eV{;H)bBO6IWz|Xf~g|F!HA0(&>cP%b(4D_~b*m!pf0^()b(3vUp@9qh{RZY`N-;XAo|jn4|}Yo?Tg5{%YNh7wL|roX;6+iSnktAh4&Zt3Pj=L>AVZaK!- zC*OLCRxxgQd0_p%C;>T7hZ$FngoGD?gZskdbJ&g#O7V(Po!VtR84|pi-tY!0iPB;B zyWbJi+8c!<_h+qYm`CBa8-B8Z&9Jx1h*YF~a$Es<=#8W!7i`5F+g@|72VpP4+|X{< zr-d=?px>YKm9p>+iIN=-0#voWN?;oqNZ&We9{1`dSO&S|Pld?Uq|nE+FN6i0Jl=@= z>6?+mr8P^Sho3gmwQXZ-XG*Z|Ws>eiRoS&^cYI`bT%JZ8cJyLpD0@77!Bbv3qpxGL z58+{@0e0PS=lHRsf74cM3sF@$P6fpSUTzS)H@^DjTI1M7g?%jkB;@D&AXy`8fAYzF zH@j**Li-78DP+x|oUfuIB{O*uvy=;!^>rJVZ|@&g)OJ)eO$U;atEg4cy~$(iF{4UF zz@uoS9d+FK#ie?N)!HG91zORLM_sn;VJ((A|mfSRkx2rAgN(ZIaS zCM*%wKm@L7z(d98hKN6iuNyXai6+m12!@*9<6hQ9V0>Q)5ml_ck=GXTSLtMwy}R5? z7HfhaGg@7*<3V^C3f@K9b|tgr3L`fv57A8%#o+sm9QE` zFVc)Q>Tk+9l8qY4jc4{yuNBWkC7PRnvj?bSE1l};yd-l?*g(4y+}eo;(=Y~4a5=vF zRSstB$?GLM)B$lR4Hm*n?WBOCVhwPnTUW(F1m++&ZWM`zhb4-cw8S@0e;U3ogD}0he&a@6>!JdS9Z0LNF)D+;fSDc0TCiE(9(3vWRD)W&*2+7d7ug)@_nBrHVg)4|jfehFgrN9AwI_Z`tKrNsv$DDK2+}cp!GI1GXZ#A|UTs zlGD0RK8tCR5k8AGBGP&aRxM$EbNx`lxoAS%@4@}FU(-fV7r%|{;o^UjeLzrkS9aZ3DimES^Bd^P z=56)u+Xu$oVDg$q0Gs?`HZD?Ef@Stb#2zO@Ln7_4W>p!pP9Vrn!|fL;U8I2Dzb^|; zVR)wx?6BL*;rZZHls?ck$g;{q40SX+2Rza$^w&91tx`$#BqFrt`5(>iv3Xqfa5O9V z@e&l<7|0kIE>DE+i7@5{NN}||EaBPQ$d(G7EriV{p=?FgfS(45M zNF1;snT1{h}$fz5S+a{h{iukfppz3qyg~91UDl~(6oH~{jv@#(cEb7u_#@q4? zGU8qRHofz?N-3ir&Frk)riXVfhsDAl42v{7c_>u1(YlEGMOCG`$iC*!rW%lX~T0zuh*Mn{@{4IC&cn$ywX5bM-w z`g;elJCjY=x8+cd6zQKbsscU%OifpMH0QzHtSr+#$)zN%o1e<+7w8(X0&a3*bZKBh z=@J%iDdMVzeUF~gDx(`94461iy6H%X$XfEd*K3J!3&APQA`6F?;_+?0nF^KoJ}vqj zf(1C0Om)(kZtai?C5T_ktkqXfS%7`PjVvW^?saLlAFa_;E|)1MYpZ*SM$q>$kEcJd zBxC+QdRt8dBP$A$5aGA{#ks_P2j0ee!|9xIy_~U886dri$!&>F)e#=HtVz@Vx|ysm zH!pUT>X}Ux&f*M%On~HIt(h8k-^BCjK#vk>O5Km{f*fW=o9_5|Vi1>SQ^$$hMf+)m zKt+5P-`=|DO02wC?eVsBlqp4((OrUyu51Sd>ax|uEo^UK@16>B3nRppz$us`*Wlwo z^5za|sCC&fyVBYoDK4SO{y89MVE#nBPvGGwK=nfl`zKbrM2y_5LxD_pnNPb=(HbVZ zpw&&qYAGh=WE~!P>`(@4lnwmmMC< ze;NffAwk%T;Tcv`e($x&+*}~;ISegYY!NR7S)1-cw!A8Vk7U8Ix`PArxUWW*daWO@ zsRcgAL`mFUQa;v0X?@(B!-`7Vf|27kvyekAGHRTyLT+ewe|pwS*tGaH-jqvucMb6r zTAU$ne@C-mXRwqkv?BuWF)7H3IZqTbdclejxJM!~sfsY zRFGuR0KfLp;u9Ta@+jCbZ2$!}Uj-(HWYs7_Sbv>+Jj@TPWmw%(RADj^+9JpDbPQYz zcM+qd-$CJHc{u*0#g)bWElZVob z<*4(XMZW){+(ugx@(FCqb1~YF4&DDI(4f3dmN3aVfGjW#9KFiaf%kNWw{cG+eG}+Swj`q!*Jearlaj3@2 zb0xf%-VOe9oC`aqIQrD=-7Pur?-f&gQSbjJN-($5@1n^paFU1VrL_z4EPa~c~p z8}I}ic`=cMU@i+jNFt@G;i0IlzDL25$DwbYI-!i&eG)im#mTQ^XXEko?C*1ETfUKGb}R{`YvZxe_`e-LjJEJ3Qefz-}W^fQH3qr%4d zmaly@b>YPvA1_Rh&yXUoENjPX`=c=tEy6qkzxA(Z9aPEI7aP5(9TTUz-^9QpFEW}Aa>z3xJOyiM$ekDHXi?w;}X zHfvu|+ICs9VHf6~S_L#GVfC!-O%O0|tgExP!um^6^4XDQh9ZCH$wh%@sZFy)XgSIA zZs5F}7Rwa!dzHJ7x6*Z`1|#$))nHg@Y6_Am=k;B<>#sA09XreqsXrP-*;W&wWGEmH z<(eITXh`yNu>Vb>;oCEv9c$b(?eTqy^8y0fE<$V--*yv1p2g@)(Wicyhq9EU#pmL3 z;L!F0B9lBxxU0xqn;oI8VcmZf@4ruwIyleAm+x~Fj{rCi;Ca0xrIedb>j~m-cGWx7fv}#LJ zg_wsV9a3D1>EFNkK+)wnY#8@gZfos&M)+b{ZS5++{Dgbmiivy)j_j2|ZVS{u3Ljs^ zQTWEV4Ep52vq8>25L@PU&G)KwpX&6Jy+M?2S?O#PcLBxEa&Ql}rAIw&Z%c z*XIu+l}}71(Gct;yomJm8%{N%v|7^FHG;j6!7nDdY(GZ`a;^`2N%52|@vo=MaC#n^ ze+rTP`)~YcXvrxAnZS@v&TcZ{FP8q z6>7iyVK{ecD)1tiYUy5W-;+7*HjF)?_k+?;Tt!+%l$;^XUp7wXQb8YW(_`pHe|ny5 zKe&UWc|&8m_Z3AAmc!GKBLiK{y0gOT_0SEaR_02G@XPxS1;e7W;1GTeZb^+f3cBQO z+Um(AI2zX#{{TpsnT+V!YrxI{KSA}a{u zJug&hkH%`>4rr|j8cnRoMGJA2Bl|*<;N{Et#2hUjK6^hp;{B~m&e=i#*|pkr6HQ^4 zMhba?)@|xH1qrXI&&vRGjrq@bFkESAp}B#;G)>Ggzmg#t&M;j0#_XB)q>k=LI{gxw zD8RK(UWY<7;h8*~OJ+i`1;eP|XtJkn_=~fNj9k-++i>R{+{Ks{OB~HvodL*9+&a7( zm^yFga yArZED=YmY`^pM2GxP+5E&yO3`Q11jwSzEJcdEuCi zr&(#&x>$*V<0dxIhJs0Wys4PWSu7DmzVl?PMz5;=c`hp$IUl?#f9%K6@YL)f7`BI# zA8pnCE$6S^Xs0iA4RdEUafwDTTa|>KC}QXm3|$5NdE{=+SdrxyL699dpR+`N6o22t z^W+*QRrn=LpIO3Gu~m_hae;ghjoxS{rMr8Ccgg$P&ewD|NnFGh~zl|LkH0 zW1h+JD?8ZI?>iD}LD+X1*;<4p9*!VLgtNHH z=9aJCDWH!$j6`2$^Y}n8a$S^&3Lz;eRq}V^-4C2t>4s1%(Vv&p)LoncuEG28ut==0 zCPa|s7?k{%&losF!~G0zi1da11N%iNA)K8e8kUt(v1kc?>ut0*N2TG!KyYlR zvY5b22Q|MABVyCENhnTMoiE*5l;Gi_TP>KU3|T8Mol2y{idnF1=Cem&{SINRPYg)f zoVpUGvm1hZ;yRCdYI7JX@O?QOA33v0^}arTSjv!68#6Z+#bXZUVg(PPGbf^~cp%?% zh37mrOvD@BB#o>PQkOP?O}YRid^R2w>e3zgzI*K*f&>qBD7n6|FQZ(GX!14TL3A6z zfN3M)oNfQH6PA+-tFyk3wG)I(c@>LoB@;~(q}yH0XJRH|&6N-Gp12rjJoR%R^RSlTwc>T5DK>s|BY^d1CIree>LP1GDX&SH@7Aghg*r zII;(N+5?)ck>`(MQdzR^?e^Vq~XGsrY}%3zRkzgxG(bK@TrS~Q=FjD4vkxRJ$gwwlfi zp&^Sgb_vZkrfVs?G~J$%w$n~)RBz`NUi#|}sW}^q=6h(t+H5g}cfmN;NS>)Jk?BUP zLfVt$p~x9Ot`;IpX=FKxNe@6Q7>;aoUY=hfzy6A6VoG~}B<#En%^`j`q$GmYcvPp` z9=BHLjZMTsH7z)y_VZVnx(3r*WcQSKKl zi`8vm{W~roOIJD?+_ou}G9HkG_hdSgfYVf^*fUf0XlgNk@Ym1G;R40UUe}7cG<(`Q z+ATb6sHDEL5Z0y2779s6LG7eEatnKq>Tu%0M06;y0ec(hx8OObEIdN*%d&h&Hk{eCa=jDZ5I`NV61cLY7yet?#02*;i&|MhMT128w(7gy9$^$8$hH_AR46?2Gp*Dy)qr(?BpCb zIz@XeaL_FKJ#ufZlkpuTQ6~8F(x-JfR99el+YKC_X$X0 zeF}ozB+_CJ^aRv7!|3fXK4wb=*h>g1gpmmHd*(b0wvG&>hrXmK=9`$@aNl^Ai!_cO1nn7DiDkQ7WbcpdRqKkK3V9I@F1`}#omp9+8D9U|f(N={gF zOlJ36$Dx?CAfV3&FD-X35^dB=EtIM_<#?PSZ$JomezNail07X^3-^u+p*UKqK4Lul zT^kVj)|V;%b>1s?1c*xgBFTT6;JmM*c`7fUQ&a5m8nl%>Fy*c#t4>n>UD));O%5s7 zIhY2-?NUkU-s2QXgwfiV8w!Lvl>;23pFv}ZYmy4#XOX;$^dxGZBDc3WN_c<7kUll3 zddQgB@#cD>PI_7YZ&)N;g~vy|$E$DKIiueauacPZHDo=#Ae~~sh!m`iV`{m|l=aFB zRD`3+_}>`ob;>6XQ-mnt%4MSIA{j@!+sS`&XiF=qqQpOMe(hrgiVBgAd`Z4-@C8B= zc`JwVV2ir_w0`ZLRzAVin|B;h5L!6uxFm(MT`Z*>@hiJe4RP@MnpjQ{yaS5Lc+d~ z0!)Vd17^;Lb;T3_M;1FPjbT-gp0ibQMjNjWfeP|ZbORyeRFB5ek4K14biKYzVjO5a za>OA;Yb_p`T6`J@UGdA}z+8$!e*s9FqqNp1FVRQCi9>HJ887HJ4U-%GaC_YDo;(NA zUayJlY7V;Ug;56uIYrq;R~#>0(i14PWk|BHDmYRVed`MN!!T#d81cvm#FdxLDEERQo>xOi=a)KjSsHObLNoDr!s~ic-UsPWjz3! zm%&UhKF^xDUtYcBkhW8~@<$5d#i50cE_uxrom8Fso!vgf7l9}7@ngekf^>SHm03(H zvX&T}AR|E4T2@L4aTk=zZVOiSl}>M$V*-}(L7=lY$bg9;toCA72W<8xin}q`_TUN{ z0Q-QZfQo@wAJmNnl}u{};fCB##o%z(k$H&I1`*mwatPn`ONPK$gm9h#%NCQbuBe}r zG{SwSDf>G0jgTFpP>gL-ceCBMR+V)dwRseim$Hjj)Q9U%52jCZ(R^RHCGdVYx2fe`1zTmmA zGta4sOahueYeI&}4_gj(r@F{%FlMN&8%JMC4y2Yt8k1OI*2dR#m(CmLb`z5nri@nJ zv=Eg#4bA7Z8Kc?nIPMe)bjmfDG+<}rPJ2w&Qg^lq;B1XeMBQv|YqxpFy29rAy2&^0 z*}O7eNkk?;g`10hQFlpIuca;siXQjUx4bfJAI$H87V9BAY9P22k${gi!ejCNy+~c#V&mFj;kYCSCCl-0go7@*DR6 zQf3+jMPNTO1>rGDHDy2=JWgWpDZ62e!M_w_@ft`@ZTgq-x_QUVnk5Cdkw%ZTsHCqi zv3|~SR_`GTtKWIT`qsxm7sp0$R~Tl8gw(u;9#VPc^n2Rs6jwMx=}4m~Xr=R2LZNH( zoCs&|BfCZQT2}<2uEwIE?6)xzcnn=uve>s8Sv6m(J@Nsor-AiUY`$hejN!TfMYg(> zufZCp*>ljW|D$ zd~%E5z&yupK==wVQ+_G?JSNa;1Y#{+?{nsq`eIz)uRUj-MY<9H+HFXiTUPzbZlRfn zVXC!_-Enl;c-<&l_~@~im@1N7y8TEE)1vA?M3;`bLE&@nNQG_z0j&EE5l?9Igkp+w zi@{w*q#$TUCS=j5&k0}5!akPR{4ni7c-+YYzCi>$@zlQp^k?w{j0pX=a6Su*Z!g*5 zeQq$S?1lA4hooUKXb}k8eTz)9+a%>amQ``DP_9V7eV&aXAWi}w{75m8(hCcR&}4eS z--jtfvpkY3GpuuBJm|4*p;yeyB~4NlNSZHeMlNDZOi={nTC(TDkJ^z%j=1apHPjbC z47UPk&JtXX113;rLUaK>(f6~RZOF)7qR;=ZjMepuFq`uuJ^Zp!| zuX#1C5fg5p=r#5#qyDXcY*-~us(F69t~UunHI*yt&J_xgSbbem-G{qZH6t1;BNQjs zM20!2s(raIY>|h`8*<^ma31!F$H&$ol;(5wicO94i_r?$nE~AA*fvGFmLg(=X%EqSg6hwckv5_B<5iB4GR=%aCYF4=u?NhDyn+N*dCC z#xB|TJOMYe&}>#Lxb8wSgbc`Qxef6}(0leF&XRiWXkAq7#)v9+#OG6n!{4 ztWAaZWOtYjXPLLmyTer$h8ky?UY}V^#9KETbb@6>TXD7wAAArLN(yndkG9bRX>f(b z*=$LU%hG@8-%ksMQ5v?b>R-b!4||PcSf!kG`6X>(8&kW| zhUaTT9BHFFBPBhnux5YrRwbL$tVp>Z)`ONFW6%(@?)QmDX4!S#;y4?>Zk>p2=4Q zp%hHqo7WwL)eG3g4Bg8=QYSCqYX@sKjx@C^ReYLzhR(htBIP^Y>aO5%tj-|BaQtS? zOlrJ3bg_ZkabR0(s>gXx6nO7k5>HgBAZDE)#TyuQtcUUSzT`SM9SZ}JiGw6;W7H~c zW@_Tlz8lSYsFMU4@}T6@aU@O#F94ke=Col26RBDj8W>q=xARq41%(t+6f2(^&112O zJI|%9xSg`S+Uba0Dr%~9nKT^#ahn3eu6=MrWS%3=?^CF7V_Dk(D6DnmXjZvTl(u0V zxt(c$^BEVDyUZ{Y_GQr3wl|~J_BZfoz|-mk`LaN&-F9+zydxv9>d4I%TLE>hy%Z;L zW2+0#7B;!7Cl7x(TY_xOu(>G>-L_E0_4uv@CSpI&N9s1UJYDi<672rC*Xip%kza{Z zlXRD>UvzI$LvjgTovlx-RQD`ZqU+4DD)i?gq;nVsNpgRV?o)%i2AGOj|J?GLjXsN# z9bW$iWJLz-zq7R2u=3Lf)KE8v5}0q!9mm{u}inm8_zX zV*L6Tf@8gyr%!94L5V)o49nWad$m$faq{FJeP2M|o8AoiRUzO^R; zKj>Z!{&w)6ZEhZAp&F$<=wH3flQk`9V1oUPm2#WSmTj%9)|P*7o6hcLn@!hOnU_Vf zPkgeBj45XA^|e7KiRSXSvkMb@!-E!U=Pfse?*fy|LK=%b^D#gPVsJ>YXB$VMz$H*_ zLp_WBdmX>|cS$t{2=9z9HX;((&Chv}UyZ_GrZoHPI8Ka@Wv8w9G)CakH@5D%a2ci3{>yF&8-Cm~-QT zJ!ub>=9Q+@5*=WCI$(3oQ%Uo3z4G(UJ9N9hVS0UjIG&>AJiX{|d|r)58{&Ex1Du4; zUP9;3m2ibbEy!RRFhg;B6UoQeJ~u>W}5^pgIUFf`lQ%3~Dc> z%rlkO<6p;8Hf72{X9J!(K8i{uZar`$obxvwA7q=}^ph(XP!xxo0eChucn7=j#ZViT z>Pq$i%q!io#d~5vPi?#Jx`%=q33Dr$gf<*zRLCXYw_SaekFd&I%J(t2(Kn5+v?$(A z+7UgAW);SjAxFa+FHyr=y8Zuh=oN6byqxaML`*TEhtLF znm{7NZ^=?Gu@FGJATJcjdyWr43z$}UT58eZ4`5YCPt^P)-_(;bwLL;`+~R7SAmK3T z`53-t4*}L4F>XwX5To3qF9xd$iwM)&i_`%IqzXT5bN!-jko}eqJ3AnMiw1QRX)K~uT%vwr z^nbOTkGr))NvIU!W?9nfOEJp^!{du#bZVU+whiKYJM-=iB{htxQb@h=f6;^~scfUO zVJ*`}Ob;*^fnDq_u#uE$WR|ry%toox#}O6l9;)gnXyI8}93u!H2z-AMv1NW6BKFxWz0n65?PMrj@Zw#$8)Agb(8UAM|~ zW9r!XW!vrK0(RoPT*@`7hw;bgVn6m@kNFeTl#g9M|jC@1cW98)>H_4?CP34;&Vw zzk%@HUVg;O5MU@b%ZYq|rxhx`{+rJx0@F}FDbo{-jMmAz*MLWx98ws6c>CO;)#_<)2JFA`B8@4Q?8vQRZMZG z|70fp0c5H+$OJtV^VvvnUD#UF&~Q5#hnHb1DTy5Qh=2<-{p#k3NZGrLBs^{KstyTj zRRwnlhNe6_L@Vh z4uoO^AwxlN5Db0 z?(&}qRh?ei)CoVlyLdO+Rz$;uYi4QDmZMAn`~?qVJuU9SrjGhDsP_{5Za5cAW%ap0 zJyVGIjd_77as08?PtU^sS~P#2HTm>$mOzcQ@M!-+)5^tv*&tZErUI1TVYCRA&uH4c zgvi|24v$Tp$q&ySC(M{tsns(8k^H74!yWw+22*O_N2iDxnR>iTfBz z^9X_cChOAHoxz@^xa+6B#U6B00*C>JihZro0}r^biExogQpdS> z0I+mI{^BsZ25a6)Ff}JT{qfNOo@vi)#eDX+7BN$OB+zK^hH2A>>TwLvHk?eyP8tOr zJcu^JR#;7+V({b6yuC#T;EzIfD%sgQ*y!kQ^o%i zEU;g1hDcX2xS2P)0K{VH)%W!Tz{r3E7_W`26QNoe10G!tArtfPh@gl%dk0M?6cff( zEmo(ZIG@T|EhU1b`vM2sIwuRblu zbbZIe{ygcJ*p{EF)45A@>87q)EXChvU+)Gm2r=GaYjpz?)u*=R(-!HUatH3UjFHm? z-P!ojDH8x9pmoF5&w}FI%Ma99Sh8-Gms^A4msY{Xhl*)aSjqOiuJ2mgy0_f)6$=)* zuqQzW6jIl{SUW{)wes$~2vq``Q6gSFw61{Aw4p@1RccvULx9KeBS`wFZ=*rXQgsJT zDE96z={e@`od084alHM9?G>sz+Dvt|MEdQ%BCfdJC)SaSVSazo=&i}C*tGcLHK2WR zPE?mWWITITYW=(l3n^CLL-mU#bIU{}qaf>Z)5R+1hzww%L3qvy*R4Fku1_I#TGG>6 zV&Rv5X?{fhkRaX53=`OM21LJIgu_4;idl?k=65xB?aGNQmEmwM#%;OkH7?Upc8x4& zLkOlQ>Y8L;LJ47T$dT||v`3Z&rmsRX6kM9$oEOXx7!4toPQR>7q(F_`?^!o$W zcA0_C;lkR|<8&cVt^4c_*{c3{kvCtD#F23UUdhKzV+^B=3>%iw0`xe#5Nfpj)4fY{`|;x;&PQ9SKOWU)Ficsd4goN zM~?~!v-hvHqw!dv>Z#LpqjYjSn%G}{kH5Z!s<<7H%JkF=4-8De+&YfhTo|>HUY)p` z`nXP9FZ)%aUQmbj!YtFb#DZaOv0))FfACPB83BI@y8FI;S)-h9bo8}sZg=TlT4?g~^e zxj^0#VDGqG9EG)~(&bj!e07%g2l4nbKr}XT6t8qHtT@oj!BNri5NOjM)S5& zibow`K&O%OXvi{|Zww!$SnHg0QAY_WU0NS^v0Y4k*dy8lxlI|STlKh~?yTdL3|~jj z@2Yx?+g|~+)N?!eCKx0RL1pc#Lzp+_7(Ob@8djcEB`~@nRybBXBKtV_OlB}+GE==! zMZE%F3t9mmis&e){KKBHJJx`v5;yGbItPOpIvbtarF{&*5g3g6b=tS)6< zvLD3Jh>k#6U&Ovep3@phwhj@DbBVR=93{@bs%>WzGmGWOHB)iM^-m&rGj55%Z>XQUTKH!Yyo z({5&-c{np!LgbYfR@9*S#zY&+;uMn{1TiepG?Tt1e(a65U<$-4)m%bi3Vl zh%9Z#TzK|@s9oCBxhNNaS%_4BuIGc>rgm4w@yg^XFEF#pLz$^jHC5DNeXEtetJN`b zL(v3D>(hn61xr$4p`#E#?Oe+Vc{~jUM8d#{B+R5bo2sg|z=+re$#um{Aoq%$4hK;D z!7YK&VMnT2YDS7!$6=>2UPv--QH(VhpYNSGFWpO(_-pS>BXYE=xYku-_56}Rm=mSF z>gt~!T%1xN+0w^qWtd--;q>fW)Bnm&`4}RQ;q+U400eZN6Z-hZd8_hDsFLh+R4cEg zs}5772&%fLIZQimBgkm2k@4RuuCtcwyNIR2! zBo63o^L!`fBd5{Z8w@Y)Bez*+CFx$#3rv{tNjjtyQ)(_ku69g^k{o)N_@%SY#}>m% zG{J#kIvwCq(Qyx=ot1`;8k<26X7QR(WyPEC{{VR{h|0t0LYT#bYb^OAvS77jQDroV%fvL4}Z_I-K8lU^v;}`W}V_^*bKbzrqmqa2u;sz zbSb-dn2@7+rZtx_Jtjc#C$8$OxE)*yXRS_9J0)^zJqxKHy7KBabOgn~u{$tU-`hHG z6R0Rte@kFNw777BY3?rry@KSq>&;hbl@>(X?-?~2g4-dVBXzMGv?x4*BCMrZV>Ur` zcB=?4Nx6(6H7XMqP-V{cnD^jkqiZewX6ApK468l7P|_i&-e2%!!8SEm9#>x!-W{1A zO@;_@cNN{MpxstoG3-`OWN4aK-5D&-tUG}Pz8dQ$@_-S2>Es?f+BsiDp8a%MOEB=W z*i$wXf)15%>bsj80Y}=*?H6hy1ijV@dM4hvU-$dnlU{hFH`^8cQ8Zc8X&A4J<5z^# zb_7TG$(pvrXoUW;M3F{~0KC@U(M#o2VPHF;#$)cFhR0R8-{EWsAO_jNj zO4H7n=`w&b4h9TixG!oO-WRO#5UYJ=TU@37bZl964&IRtgK2O@z>C`wU(a0C@#~7o z(DZbjGbE|*drlw}%Z1I?SAQ2)*ri2?toNZJ%Y`BGlRK(=j}sXo<6yZ+9P(|-0Xi#! zST%>ukD%Ewcn~9?OWd&|+h^Q)1}m_u*W1rA5p7tx)8vyh_|pY-iF0x(V6(aTSDBq1 zjMXlpSS(lZ+TZ3OMF8YTzSPxe@-3YU!ezAf6cy|sUZtZ4=Hk+E=rp=oa6o{Tu=+Q#8$bsDOq3*6cV*qjBVbVOnbiG9iw59HysX`=c)t!VEP2#P7UVBC)g5{i40v2n_44hwv1X2BPJ^UBHxFZm?kLZUrJgm9 z97YcJLE|%-ZDg$BEDJtO8`toLByqF^JV}TuVIAH_Uaz8V-np=FJoRi8l0`(NJK4>4 zJAC;~9(uPDVz&MEXpfJvT8`t=sddT}3LwSy0}f;CTSkl+Q)AmEg-yhogCicsjqVU) z!p)f?QD6(HF)KV^NG`&7A)n9Ar9ufv6h_D9Pb@QOQAA&z89LfzYr%PD=Bh-Vkpy4g z@aMnG+{R&sZ@s71xok@x<&!@xKM{Rl0Lh`yV z3`yzinq=$Jx>f#xN%5(OhB&&B_i`^aug+VfDZr-Z=e%Q%dD4-guo3IBU1R$he$n|+ z*2VXX9$eA5k`w;1kD5O$$;o0M(ndGKx*?Vs6fN9S1zmu4ZBFB z61e1H95V;VyCgb|jgdAyu*@xHxI%5Qq~4AO?i3OtvddmRquR|#7D9HKL5)=v==#{CIHmL}<>0}DH@U4u+K}@w_0d;Mr zYZ!=Wm(0)TX&0*Z5s>$;ZiN>juMcmZ;7K!+!y1V>M- zSu#iRNgU=&u&NAKT&>d$wZrMiKqn!zLf7FTOVBjv)~navX!ZK^jlFL>=vsMPkfEt* zCk6_*8EWnOg(WTU6ZB!A-t?-!~R=!%F8U{7XgyKA}dQT zM2Z6sdOm7gK~m4e$WL0;r?JTw+JilO5(cL$(tpv@KtHCG1^6#>>;+{s$-*Sh8rACw zTBO)Ewe9w(c~XL?#jsQOs4a>&V%C*c^}46%F~M9teQbwkIcDd@5SZWvwfPDHRr#H0 zKY-o6`wA;oB}#8os90y&nGoRKEUOSmDp$x97HkkcV^?%h{TNRr0_Ug`;vo+nI&$3I z=GJq=>)1{52?BPAqG>Wdb5OUOvmL$cdmiuVD3Crlhw2bqP#aSoRLotBt!$7lh{b=R zNM!FDb!pl;ym?BMs>Bk>QQA3NZ$?)|Nyv zP#%w>mp8`hvSF*15b~Y+gOsinXAW@NEwT6M<5nc4w-4xz?hnceK#DSOOL#-ML3?WxsKTLAjO z&)iAphFkW_)@D2RZSG7>5*0<(4~p#{;*ima6T|duT(a}wU-T$r$5@X$lNdAOU>$A4Et#c@FZ*#5cj- z*?@YYjhfhhL#60J4!sImI_hq{=gAOU7j6TXtNeL&@$Rog5fcj*1eiIy(3ZTWP%Hq* zKpW36!ANdivNo_s<)}Bu-SZ*%LZ{Cvjpk+|*;N^mm;;L?37%?tR_5pHrcd&tZz zy|GrrKuxF`zP%oD-#{(KHG&Oa%TX2zB+#rn-Tdx?Ir`y~v_rk`JRY20xN|p@BoWsH zRg6^%Psak?WU7qMVOU*dG2R2n$#R!g`%Z7I{P6ylg+?Q&PFpv%SE9sr9igY%!r3K-Q+u0Q zIO|HV3MT#uw2ugipOipsK&LwxpW#*O7&l`MuSu$vY2L)G(zJ&r&(EdrRQ*0MpgAok zU0{Z^;2YRL!-QeL5?a*TN6KGq?-foKi=rR`6XKUX84^7buNaHsT)(9s=lgbi93L)9DQXY2tOzROHY{iV*%tW z=#I{?5AXEZn=G2}37DMsT%Lb7i!9l6+kD+@O8aCbaYbnpFP#g(RhEV4Cf|DJclH@t zDfqoy5F3zi-BiBN&-pc~dpIcyP_}4L{P*x%UvX8c581Ct_^St8f)%6IsajMxXQS}z zKq{7)H|uY2AYdHdd83peQ!a6qxU>64D4Vgw$paL9BE zsiODDH{LYRX`)8VkC4?tQ?5>X@|($XO0qbu@?TnP%zr zG^5ziL=$eR6Ex~zXO2Z{fxD09a*%ogI9&zuaCSOqjvWNp? zJWo&!RYMG;t>V`4+u!7|FI3Y_i|{M=I*? zkLZDuBjjw%%o-+S#dy^k*fdAPnt=Z zjzxLdldwF7??Fw6(O%*c!1=b+!1OS>=OmtZ+SASAJtGaa>9 z7XNSz2j9q&sGhjU&Y~~_4&FL(A?5ztaE0sKToy*enpVzo!c_}-KO)1;0cGt{DE<4Y z;Y`^j89G_mGFK9kPp>D9!g!`&F-DQ$Y99CmTf_t#m}rJ3r&q510sTx9ESffT?3@o+ z&DLRg@SCV-4>j(9p@Ru1Z|)-U#>l$6K1pl*$G*cJ)mLJOYiU{lZ+a~h61J?lfk&-; z#%6jJ_yCPO04J%TG5>Z`lAoHP0(O1>RoRu6eSx?4AqeH-9IBDB_M{t4^L|fkIap!J z=isT2^pTirXF-Lu9&x>(Uj1I^C!&gdXAuX%Cecr~OY$*PoXm*JV0Pry;Pix@wftHA zP!*rXbfdLi(b{Ycfx#aUg+-h_!Yc$w*nGjawOwvSu%1*tE(uJ|&AemD#(DMUno z@N-ghllb!X93C4~ZjCD07c&}yFi!z0^-qadyW`n#Na6K1k#{sy7%aNAd9T@qdRNHoE8YmalyNT# z1eVMijWHU0^6>y>uF!SxNS^0T%tg+Hh&8AiF?EhrQ&1&%Up)TSZSf z$$eZ6IBw(P?xm&`j)h3tnSS@)&+7f4o=@GEs!GX9~be_hq7{UQ4IE+em# zt?ehwN<~)3wzUr}xcNEcO)+&Ez$=Ie^)M+>baH`PYoEZ!wne85ExA84>zwIw;1H!2 zGFa#VuFc-_*4zG1k!F6-LlwN_>>13xT{Q9`YaP(5TKaC5u`~z*J~i4DIHe)7NF6P- zwuT^N0rM&Rw(Jgh{YRT+EH7K=3DiF280P|8waNP26u3q-eh>CDh*w`($3xxDdQ5)f zpVDy+R-&{k3p0uy8eiK@w1-H~i<+TSLKL$g2az)u1~}lrO%id+i1Zi@A8c--;;@C{ zqT>RZRM8jx$yDNU^Xm^cAnw6&gnU#jfnBAUJMs7(22wP*Gc3B;!{Iq)&yYotL9>}6 z|7|;K?mDlPztie?9DDOBsFIjUd@pl6xAw*HPGk$E8T3oUU|WK#kn;OMASdph!5cJ` ztBg%PuN3)^?~5}#eXY2t&_whye~ek<;7vK-uTmxyVBFe+OQu*&2vl(}(NK0tK6eOo zS#&IuXIf4--Ysi!Pp6fXKyz=FD_P%wur>ngPk}~C@x3iC5>)>!P1tTwju$9Pa7fg9 zQ8YF8f2rB7WqE-)mn?Hl@*etK?`4-NZqzO2uln{&O%k-OUWOxtTe(c+Po#O@WodC; zz5XQ{aUp0`E1Sl>JKA|93$f)!eQ+3l&J%Ymk_N>h@tLIz*MDGs`KNGyyXg?$CYfaD z09FId)vEWj;y~~l_Q~2*lt2)wEPe$=VREf#l6lHqDbDMCO`fq)a372j+D3LS+(@5W zI}!t@l8({($--=y4!YG~Dm6LfuXy%pMuEf<;VADcac1%K2FBAo8QK2fae+jZS-eJ~ zzD*E0aYq~dl<%>z0)Z5Uu`l^*n|V-8o0Y>nyPZ!r^FuO1R)|QPwb{M4No^x@yRz#j zY|<4gE?Lb-dqB<*2vL@Ufs|;H=-6$$XG4WgS3%lOD6b4p@X}tUj2;%&FEAd`iCv6& zymK1(iq{?esFf#lBH)SE^6&QPd@Bow4dkwPdDGKED~u$RLVd+T4AGnq1=W64NXpDY z+-y-&snu}wGBZJk4WBHMqDSGOaVq7-E@J?o(s_*G|h*r$FPmYs8^7KA#lDo9K%oi6`M!E*27QIG=-NL+$!R z#ZJgnFyp5!=*~o~&zMgL4ZyA~{lc23r1VAj6n#AaYUS?9E-Hv(-UjGF%z)q;;37V6 z*5ZOaG=>;RR#b1+)TJNO0FGrGi{-;O?}6d{)rQP@%APpKo`4$SkX|)$^H)s1$&YBq zuwMz7Bk?a7gxK$HQpl#@TQj2x>5P?t1vmc za23aq1z?8V0r9}2Aw?j-4u^^zqoq|!U~D{n9uBmqdCd>Uw?!h3bNU?`voN=E1+=qX zbU>%>lzm-e^Io_jAT+M5Ls7uO!7Can%{)MzckmU54a2~I z*}~9%*I?e@B?c$FTgLh-ERo~_q{GHx#zC$wd#UtP;l7HFqIo=R8aw(?=@&#$9XgjIM*X!$ZC!p<{5 z+lzv6mXj3$Y`gTqcmBy1Wm;3rvNXJ5N>i;ooTh?OeKpCXV5mtp3PoR=%RzpcM2^JV zA9p4xhtgCeZ6tG{uYTh5l+b@YXlu9~ys(jd%5P3plN0u&J(^mhoJ|`)>aYJwY8t!W z<)B+VbWZbpAZqd|Q2`YUZ{H*wC!U$$nB84P29%;UqR1l@O zzZn^sKwAfvlMLwkC7<;O=#2rlGpQkBdgx;Oyg&v;Okl(16W!W$+pR39kQfYJ)AQ;X zvH&qT+N)K{CAlsteXAhji}$<)K1ru(+YStRr@08a0gSGSC1tt3b1+s?HiXm^CG;uv z<>QI&=m~hmGl>#;QiXnNjJ=#HeeHt-{*9=k6DUX+pL)PnQDlgqn&gYQzV5fU+FJ zQasWuNem*JJAL@~M*8j4wKL*5d~Ve{ws5iDBR_8bJzuah zZXiGjs^Qv*0UArB&)aFFTk=Mmkt)z1?GP$E+5mzh)_U3xkHmKdCtl$uw0IuA8Ijde zV-UyFB*C15z_`e~6!vCvCCHhh=3AL!Xr>4gt~pFakb_FiD9wXol0Mdi-dWP%r|ODV zwDaEABklo0EH>HW1${hrPRS=UL+e-I(8X~{Qz3*;&G6e~Yh}b41G#Wpd2VYw)al

{%DK0*2WIP14d?*oH%9*xHkkRU;`t4uIJeBFyAC$!HKYS!?2nO(1ZMmny%IJk+ z;^&9#_H)5eZ{Q6jmYYyS%$shS`)Jl+XS}$47OoHe>xuk>^LEF}OxJJXIt~NDg z&(SSp@=&S2)lPBEXOM{my3~DZ>6A4(cdhr&#h-kT31iSr0Gn)45$NLP`1}7rK7<+9 zyTVLt;8J{D@aJ9?WCD|e8sBIvWfK}*vP-FOMu_2gX8tA}nORus2mfR3G z?8$q)zQ8y{ z&a@tC++VSoTj&g{4uFg+laiYPNRJQLrWi@`Nyvk!zst%r-qBP`ozSNP-jm|hu#+k; zAAGIUi{uFeP2R>3dF{tSuW;;)TTyn4MMk!zRxv5Kcv*a&EOoHKM44Uz9`Ms`-ahiaeH0n0J}6duXP2>=SzrO2F;mmXg?STmA^7M@`KFcOmM~ zea)T~-~lfV=y7vX=JJ6JMnjvx+VR!tM5FMj>}MwmZr6(PANbwozYlX`vqK!WyYzJyJ0rbwMT@<+-g&O@9`N447U5g`8lOX<<@a3LGGv}WQo^%VCtP|Dafs#ET`crH97MsZg} z?V_huo>H5?r|yz~gZ^dQ0)X(eNr4nT+fh?!T5|mvJi@TQ6Cse`4l~`4x$!54))a5t zVh}KV>!Lg+e0OfO!kGYnPn(R`#UoLui8~%)fb8a0fcC;7R3}R@cSpZR;8j!>X_2}g zE`b}UGXnl_u`+>zXd?``deE*^i_Z_(;y1kIIot829}~H1kM^xljfWM(v5G-e+MeYvLT z!E5`cBWYWj+iEd~e~;|C)l8b|DYopqpDF9z?S5CFMk1+C3a4uw!Ksyt2b8e-K7qvA zQ_|W1!wDk_E@hbrKCpoEuR+tLqlPE9WGa(EZ1IcZK9cN9ed`V(}Bf= z3lHI0E`=>sEo(28s5(S(KRC`Nu;wePqugG5kOD{+#rHJj+Cg-eL0$zR7XK`nho(PF z4*o5fsS<0WS=TV4hg$|sP1O#*9(@Ac4Ghv5-+ln=iJG$ii|AzfA4Df3C)@u|bTTur zGyYGalZEM@@cG{ZMK5Y$?QG&mKrd=-;A|pnVq|A*0>#S<<>c&WVqgR1zR~1vqOPsg zLJQ(X4nq46LaFdx_(cIi_VxJtXT~#D z#^%lkVVhVR>p&HKiy$`Fy*M<}Gq`^>f-+KHW+nX^Le2mbLq}9yZDm!_0ED!V{ZVpg zaQSCut_RHl^XeKwF#xAwt7QOPO8`*w!1z@(lqE0;YDj7;@}wnR>+2e#y1R1@`I;f9 zE6XAm0fa%0m4yNMS+V@vDjFvKs-fmxhxsyN`Dagr{ld?R?frGuP(T$xl>t={{os2A z&dkYhK(fcQK=D(*o6L!XNFV??2FgkzU;xae6{1ZC0Y-o zXtJ|&e(H%3^eqiwzY2HGXLB{yeoT^3lao-8RaEm+-p&7LX=-%FU92tqAJJ*XBY=oW z2LKL??VpeY+IKf!RM0{;;L65^Uy-B#n=ChUP$)a&8jJVzTUlFeeZ6Jf{nNU=G_W%C zBzG*MeRZy4L;KI@OeFe^{`uz)Z<>NRgE9P{(OJ*=kFo6SIF8G>k9#kK;njU{eF0Ds zvLerm3!%I_1iCynK7#_`;O+qM>H4{OlLMI?gJY;?aQ!iViZclKwsMisQsM?U<(tBn z{5k$GgO>kdEld8l_+@Bjbq&Y>m^rX8xz6dw0y6yjm-CaJjuG9;ic<)iVAwD1srL}Z zhNg<=hxnnFXeq$2NpWnQa|1*3H*I!6Vrl@_yg+~7+WIGM;AfET-1=?)GqA;Ur7uqA zeX{7oEW3*D{l5&>S-Z^LO%fK5;~~g`b=pfO%ApZchyJN9s=O`gaU;Zjut3 zdfZ(8w>I~kT6$?^ZDwd-1*UIk0*t}gndQmfA0tJ94r#k0R z461KucpY5d@)Y9JjD4L0;EBjj+K=!Bp5)bT2r@727k{W9AW!X$-4JxH@(2FF7<>-< z2ixZe^B3OG;GG|T@!pP)7?{`}?~Q#7B#-?Y`%9eR7yC=U;SKvM-u#>Wt>5I)ZtRV4 z;l=JQXzb>{^7CK0|F4|e#l!O7e*OFC^5ehqVmJ0AaCDFUH7le1<`n+T`WNPRXW1|J z#?QdyAAL!k*AX&4O1`-(XbhW!OS{#K$r~DVKlpS1w#vQzOF;1b+J4t+O-RV?Mz4v6 zW(q)yOpffI5R!5{JMHxLC029K2e;7+_F4G-6Y`tw2tY80YYx-drEfNcYn`!yqj^HD z$)E!#Fd^Zl8Zll&Ue#X zsv@(4RRLv<^(p0fSO5hTXI7z533e)3=Yjr~sNhfP#P}+WLB_PTocSntS?pg3Cs$X# zVFsC=LhVw7^W@grV}Ul*X~vd3Zq33DdO{<=QH{(WxQnl!U_5l2HS#v;=S2cO6en${ zg`QSy8+ewE@LxBSVO_FS;!KL+q`{HM3&v#E;6*nMZG=C3NL!A~4gl|Jxa!ia(Z1`m z{8N9zLQdYSU3skxjPc3rHX%BsLhBD1QWv5M8=Z3{L-JSj{< z(_C*_TORvcZAsN`e7#edXk`C>TG0#_VV!z6PBHHrWxCRp1h{dR8{TZX>@XRR-6rI+3)~sGQ;B;(Lhf_u>O7ycq6rwS;QTfaN z=^)-T_4tE|@(6cBs$9`#Q6Y$5YLT#N+Ho}8kf#kn5Js1*YLza3 z(1ce&TR={s2Y7XP1!CUEv|H~8Zc`8Bk54}|#p<9OJmyIwR~T%{wr3a!0qBRnF;Dx{ zxW&r15#r1QJUDQXJK$-u*wc&TU>x=&&#^F4sM_Cwr$ihgk-W5j!L_4SC)%*dTZ~S# ztlFt;?-jUj9J*lW$ML5rOgBN`7rpGbP};fUOfU-sx9vxmLFnk z7JjnR?LheIFCiR2+%nq!F%{g>5?AoP)S-{t(~I)h4>ddP5^e<&1{kolr!je|H;3Fm z-91%dw}8LGCD1vYuGyGRl6!kd^0lnibGU3M;h z4-pA_Y~mh&Yy)Dw_1V>0#TyW3uwG6ho5&1XkSA`5t=Qk)Jnrz3%!$wzK6Xq2X%!k{_ zGO-<7AXvK`Z&1KvN;QbI2-K$OLRCD?_FR^KZxFAaO9L#bp%3PmpHMSrDLN3B zc=|BV(ADT5HoaOU!#d!?bft>yRCSRd?}Pg(_|$~`$T-ZgswF*e$yQ-np4A+%1q&kg z@8xD?wL~lRgXjU}GQu6)zMj^KTyn&1uKR?VNmUv3+orrOLqe-8@mM&k7=j&me~gmA zxgV^G=Wa=p!}d|I1(%j#MEwa|_aGNqLO=gFDbC(?VZ#}=w3}WBcn$yy*+rIt55Ux{ zs3Ai66~J(%y4SSRQ?AOUUnOQZA~@!EEb2mTWPGlCY_C zA5Ca)#MCI`8K!TD)VcMzXgaDSy3>?065)(=0v|Xcp*c4&dRo)nYL7nE6%9H!=Nbnj z1+Y@}mgTrOp)K_kkaW4<^xOfgn^Q9~{|;Ukg5AY24!^w87JOdn8$%fkWXG2}i--sJ z5l3zY{PZOvJ_U6zZJ4kcaMp!HM;rh~JZDmgmgh1PSf*?;OLRzUHH4$cCpNCUxg5#4 zG+$$|IP1v~*epm8E9eIiKuM4orKJY^8Y~BvagK!mq@j>Zx1cBy0Wwh4Vy*91Fn?hHfG%iEzyd2ARS;4@6!=s700$o>IMW=%Ir_3 zqr@U%C4`%Q><3*f>`>^1+;;)OF6xG0Hn0Ls#GEO0g}nVNxm#@1=$uL|gcSz*>qA_o z#AWtt)d%|*E^aXJq ztn$r%TrR*_zIoy1lB7PvK?~tHT;SPY4``vTbI&^*Ki1C^FRS04aFz!e+7MDlM!rrx zG-kJYT-oiKe(EW@6#wqIr?S2*kpkzamQoAmXL_r-4e2I5E{;9#1roIa>Ia0C7dYIR z$c%F^&D#1v&J~4hkX(I0tg{-lydOvLOR}4aHg%UL{l;QnfI4_~gGV=O_;k7sGYgI0 zfrdXfP9zMHXBp0Lj}3u*0E+oC2KPbpFgG0`3j9lBAA9qOv#5h~aF&NI7Jp@=5~QTC z_}byYC*YwAU%E&&iH^+u4+ol@mwzSW)S@$v6$f%*0?%&-EUN;c)sXk=ZMNr)HZ%Ze zO8Tp>GQ}1sFJ%ElQp5w~df8Gr|7exAc=)Z&%Mvs~ajq#9XEvivCAY3- zb-U=%O{PVNIGgEh?H=8yj+LpMX&I*wM#seHuj}XK57$X&FTz|WaWH+w_Aj=ZRK?_2nauml z*A-?YEEf*RLM_-!)FMBnxK*lT{9x%-!>7QJ+->~d z2-LOtQ^KE)FDbPb(;koy7>^IU`M!FrhKewxWp219Mpn(k@;@I8j}A4@jG*<)VR`%U-LuqGNA|s&=eHjO!xVkXysIU8%O|9&vyzpyLw--|XMeldFt$xI7I>D04#%?s}D^H8mKh?;9E8#gzlb$674* z*T8ZNxOts7_^u|0#I=+Mw2osu?Z>{(H!5wCE|E=6bMdVhlGN!=b8VU72r5R!r(N}> z`9nO1#t3m7jIkAFCn#vBamIum$>@hTMTQRQ+KNGGAy9q^$&^$74?RG_za7#svW}us z4)-&^nC9J%V|#w#Tx-jKIDKTqP1eBXb#1!|Tl`WBqD)*5JD)i(E`ab{#Qp`zpV8>S5p^qv_K!)+sUroPy^UM-hzpQ_`zeFHJM(Ruk&Fjy3jAK(Pa5KR2ts z){KAxm! z5TyfUb{IiAq1P;)Y)@?|-bm4VGveU{Cn4BoJQv~L z_p!6T5lP(QhVUU1yPxgn?FuMi+!wCsl~b~(%62sgZ6}{R&1G*I#&;I!^o&LEa7 zD9)nLvh~);(S=g-IISXZSx}DH+fa@?iRC~kVcQ9WfBHU|z@S&FBw2KQMCm;J z9SlY&Om*br%pTeZC#U5rm0C)NN(GN1XDo!$I%=X)`D@(TGq@6PUh4VEoNs! zm$qCJrbG2t_0Kk1y`l&gVTkA1?Lbj<0(ws;sGm^eZF0o-5d02&i`9h0u;a@}3lkG9`c9@JocC zTZ~OMD>l*>oT|WK%dp>kSAE}|&W{}QTLZ0ZfDaq&z|r?qHWv!H&w|lD=5Z@)^st~$ z-D_WBSFA%GtAOq5=gva1j7kpF!FCp|kCe`D2q^w_zw=5Ij0#tizFB?wDai6)AtdvpD=g~QD3T*auxm#Nm{KI!@7E#|Pg#C~9XMsJIQ9alZks>RqS8Ev!on{q!r}JfJ)yP9hY}m0>+ffH!Ql{Ub`kbLe<=8iV*%K_V=)R zV%GszLUkdpe&^bo6mQv?_c!GSpshgkdL&OE#(dIH8_Map)NI1K6SsXV-de-jQtHZr zPs|cAj5U3i_je6}OD`7nn*cVdovb{*+3HU%QxGD z>VpI_DOqANxOEOhK?787)$Y!1_Iu9q5fdA?+5%QXW@&?%^kU5&NuXj7Fo~#K6`ThOq)3%>@2)W zvt+y9Ix);`OLz)#FwK^@mf53=lp#l^qhQ@~_zVq7w1{Dt?nkt~(V0~-u<>wOLgT{{ z>Rg|Pf^h7sk72*-a}rUH+(3Y89B%fC)zo`4efQ)#fE_Q$XvTO_P7gH`5GD1xWla$0 z58>ah44U{#pNY*V?GvABu8tYl7QE9Rk4^Qf+Nde=6L4@Z+b^xirzDi;w(K|Aed%D2F@|{c-$G zwP-|_N3pFnB@H^=A%D0J-TbD>j1ou*Jsvcad;Yhm2!OFp`~dn9B8CN^rHTZ*g&~RI zNBD}8B3bA;*16VUyFpv36fBLiM8CQ916^8(ZH$~eH9rRpA7A&^9C3{uAyR7yv7%2~fuFM(t%;`GzV&0r52ree z;tkAxPP_2NwdfoeI%`$2qGK04L?#Zfx2vJKcdnwjXRF68^+WumOy&Sh>#y5d=9;R2 zV<}Kjo$e^0uULRE4*M_~UO{?A6*jZ2NdG`^f1$Bs<~daTD`c)`l**sb5I6c565WpH z0y!eQD{L1=-O>240TQQK^V-IZyI2ZD zOChi%5OQt4Px5&{5cyjbE+Szc4ao1}UikMcB?~I7^3nKJ5F3A@R46sU;KVk_@ST!> zU}WVTU)_-th_mzSKjvNJwlJ=^?o=FY{0@q!5dyqYh-Ve+64iw|4))KgG4S%j6y`~F z$nAyzbsS*Zd4I2e#t6@KB?$@#~bugehxcH!Ave;@Pj74LgOb8FRm^hWPnKX{aJHJ;xUe zGF_atjO9(CPdoNX zQv3tJHw5Z2#;;+B@{kK?BIAn0^u{Jx76jd?3E8{CYHsoR+;-4H7BxjJ zyHezI#+neRM8YWP2JPF0YpERB814P&{9q_*#6-7h6;|LPP@o2JH+z~@-GuiE_jGB& zapa(&WiHux;bbMw)`q`^vy%;nZn`+uSWn#?!0mKgoo%B`NfHA);u;Lw>~y2^8Tqc+ z!lo}CyPjFpEakf5Y9-D4l&A9>s&S#%v0>Y|Jw{%7nAcrrVE)B%7^g)-%Ap6jfXzO} z<#^^Ch)p|JLr#x=ac{yyDG!s>FQW{y%n7%8?2i4zEae`Av>zmc2&K~S{_WFr*Qo~? zfrUPvB9A``)z}Y1^H-D^>r$1uq|sLw9Mh|{1>(FJi`WGA^s6TQJdgGAADr*P@wtAn zmRT=xFpAT0_+Z_h(&3BxBR}nO)Tm(ElDi|QRH0bC7pXy-l*Mx_e3Mu-aqU(>-rrFy zBvcnBNj#&`dYBcBop=A7<+tW{TdwShdKG4 z7MT67>xI`226??6LK+L7glFT0{bZeT`8OoUb$z$tL~{vU&eYEZ)#x|D^$S@ph1Aa8 z>8PeVpMQwGL4#WJ4r1asZ-B2jf1U_}^Uywzh=a@jcz-$PW<)1&}iaFZ_ZS6_!(v*4hw$?UCX< z56zs8aeiw^%|V)6V0IBw;$JxXcW0X1CJNWn))nkB9^4yCE2W)(syzvQy>PIY-(guh zeUn&gTC*zWXykwZcx{LK3>%PiDf&(gdhpc^sE$TdeUjT~p+}6@I1I}|J_$J#6r{Z+ zkkNc=t-it^o{^BoT?DJ3?VQFqbMh}gL1dv<+2-Mw@Hdrv8F;(&o2S4AwOli(G_{ah zgMua4cneo!#&jShewi zdU2C!g?`7RDTONY)#ZhsPJO?S++O|mOrjKw)+8>c@eY)j1-1nJv{7XCRIp4T7KyS$ zdOhts)NOa)nguxq{Bes7d`Ss+U|I62U)il0x75-sujy5#jo!l#FiG$U8{opp28-`D z5+;>ecCVHAgXK(R$UVd-~bWdl;%gWz2j^e_4r1vkFe^MwA$*lN>Z`}UOl|0 z$NE@|~yenoysviDwxGPuvzG zGmKD|gxPYCng?5+o2b1*1!8CkGAFY=_7)s-@plo=6sF+$t&_|KD7qH&(GUk`{sF{`@AH3DNgGKfA; zT@CE2D#3Ze812JTy$?+%3~KIZ&QegPM6Z{g8-oE;-hVNbq%|Lb<(6etl=>8CD^rhQ z4kC9qjqk6)zA%+_cKvv!q}p#sYC=m9(9FuqeQv#_jTe@Z*u72X-PGUwI=X{n;rne9 zgU~>F>q`+q+)0V-0{Il>VF55F@@df;en=KkFY0~Myy_*~;n5s-pT0;cuc3?F*I&je z&yO;}v64%Qe~2ay;-QVbC^ED0P?UWk!wV!{2;fcO9tFqnWOF0}7iA4Ar2x*QdDG0c zIa?P&qHfW0&hYygnHY9q^azFp=WbsgVUQGYCJo#mdPwURWQPTu;LZ7ko7cei35%=NObOKNbgUfC$qHp^NKs82)7JeM6*FfdIk`q=g(6 zRzB0u$b$1x0)7SyFT{Z47@r{F5rGN2rHQP^>I+(_Q`V^!3GZ%3@lW3o2vIwt-C9K3 z_Tdce(jF`oh*_%n$3-9lLcIq8K?5EqUs@hS7a3O zJlmkx*;0%-oFnPaV}fq*7R!r$BBuL-2(uBJ?)|K@UGYtW0Sm`QyQul|Rrlcp-%Y34 z3EQ)9{4q%*H9`h(3F^ zE$JY|h+LZ_N5@<2ow+c4jvBsZQ+EyTx`v-I*Q#c zC>>~Zw}xbMPrXisrPP^zy($ARy%h)TCvQR_5DZn{GmA#$eS*eN@?`^kQ=IcNNK$Nq z6tAo(Nurkhc}iD{IQT`NQO-c8`R&2R52~+?fmKi4W$i$o65PxsS_t1oEUkk{5|;ezdqPWEn)h4ZuN3m z6OwcZPJ2*83T%dE3*I&Nm096|@K_Yyw!(>(7a6#5Ew5jb_fZ*fOz=0np!%@=W+(l* z0NN7NU{9^^8s3KN6p9yq2d4@Q&YM|w6(u)I5J==*sUq>$eR=0s4oyW5=OhihG#r!q z!0mjpIR?VZGR!X;nY@I&!x9wZehDi$IDxNWsTZe|!B#e{vX|4t*Y99#>geO&Xqg$R zUL6ImL*`KmkqajJX|-N|nPT!5BQSp9i#5p#G>at!TSz9}R zIvDa78gD85ftJe{8L`dN{`Z@H&UfrY={7^%4ztirAW>t@?VN3tf*f6>W9aue>_MaI z746=}l2k2{9YfhV#bGGpdNX74GeX*~N2E01#CH?HaxAQ9TrkzDI9a)~ELJyC_=wGD zXNTpp?%Ne7(WsRi#;fr8?_D)UzYy659Ucx?1xtHo;Neo6zKIuSvakt!03E z)7Ab>gJ}0Ny~|6zga{KOIA`AU=L=tT=}=PFyAw%$#qEYtgZT3rw*rd?`U3qatOz_E z7@KK?OcE7JZa9w{qO9{|L>Z(W=BO2=JQZs4Kv|MTiGB$=m9zC zeV?{hOhYBE{hFBds4tOynLk_C6Xr61q|o5L-aZ!X^WN`?Yi1W^cV>^v`ZyPHL>D zX6jEc=E}4ZtQwh~Db_kL)9I*E_o}h@3Sn**0#R{PYhI3hlS?$$iiZnhyfI1&Z@|JS z-m?tSaeq(0@<7#XxL+pw*_2G3w%)FA(DcySxv5VcAQUpIW7EItZyXAlxKxNFc#O)i zH?5kDl~_g9pcLyDg6m@19)Kx{Wgh81uKs`gW_`Dk;qK`LK=GxlOC8yLA~N{>hLSU9 z=%(rBS9|-eJRVLc{_qR3&5hXzDK138Cn!nglz=Yv#tDUZt}W4c`?_H{C=N?d@E>}!Ls8Jp zOa$ThX^lN7g^d!SuRD`RQY=Iix&-LlHm`BVUy<`nC}Y*l7Q%5&x*U9w%hi2<&TE2b z&Or_=-j4&*;d~t8EHR&$*ViKOvw(uEp^5oncTiH%pzW$y37;b#T)h_9(HHE${F>~; zJsDx+%&j=%KA7C0fBC6G=YbP8hh_-@nZyjC>T*G?G`q=G#qG$xdPK~y**e;*(?Fb_ z#K*rN--3sO%tOZ90wM_q>putTLn>R$$HbAFN7O%Jg;T#Myps%BTwjbz{Ou;9OF2Y7jCfkH(X|#xd5gvwcFjd9(1?`spg3xjLVQ=XYW_T z%OyEVL6wHgXr&F~Tq_(`0`m(--pUl-{cj0to)2NB@>3*AU?XC%Fd@k;UAhNKNK?e> zEte)Z#)tXe*l6pt+#kuuSv77Kio0qB949WMtuN>U~&X=_P^dN*0*Kr#OUPHG&!U(!fC7! ziO`I!&FRGWEbNI)U#7q-GvF&splsVtI}qQHNsK}v(~($ldJ@p$l%l}AXEcv{RiSgC zg(1=f-K(r`CLtnu_4}@vB#egl8gkCl<9_p+gblk@lF~Ltm|eIaK=|ba_-&G|2G#pV zci!8Z-?|tOOoVN?1s}igaGo#7YQ82Nq)itqK;+4eORSRJVKl2-9%)gw15pSGw>LI} z__ONr;FK5#D3`yTcC(BE`0TLLOs2sssK+JRh(YN@mw#QTwi{knS|||UKFVY(e7jFp z9c@9iZZdTyV9f;LNoX>$!M(V&Afu$3=NUfsxf+Tp0dEwPkXmKRQCJMXiPIe~LOVnUbIzIK?) zA6aSaMCPd37v*EGDQZIbfGJ^ZgNI+rb_6vHJ2PDX2n0vC3HdrY5v?QHJ_jSmVXgB0 zu!yvD7Vgub#CcfDAv%r=fJmqvB{%E@EFV0cGsN?BFJov(kA_!#Ifngxt-9{*PokS0 zgO)SLHaZIHe}QmzZJ>TCGqZ6;c4ps@8Sc-XxFF&}Xl-+o9avlwVhB`Or;I*oRDjUW zeq=7JYr^URvsegci(vhzn`73p2^L-{cTM0diFFbCA4`f#ZW4 zpmK*MJ*bgo!Y!J@*x8jk0$zzp6C6DWvs2H{x<4Z@yfi3{6Yv9O((67j$ zEBkndZc8WGtK~XR)+ss}4n&1*55Eu%vXeHlT=iIix*fY~coz^qkV^oB$%{(hAxy}k zFEMBJzz;wCb)Ck?RD=56Dy;AtLV@gbw&UT)vSJF;+avYc6CctaVUo4(F8)*oNqbPG z#9fWAoe7ec%q4y=Ef}b{Jl@<5IrZ{cm^E2)Z~OSyX-)SKKbCc^{iMAQIkMDvQ3x}_ z{Xj#wF4GjflEpw~cYgc$BDaeGzA!Prh%vv_8e}TQ{p!^b3~LG|)ExN&-J(TE4WBc6 z4V9pqe}yPE<ZEY#oJL?Sj7CAT}`^5ub9C#v-yIlQ8y&R~u@a zz~+^P0&6$%x;Pc++t(BHk#oIt$qC8z&>!E&AJc7}T2Lh%Dp_Ap)N2IIVa|-ZHJA5v0<(SCh+sHw8=?SL}gr<63+ke zC*QGcngF{yw)eAVRwJ{nxu?cG7RZa(MqeRS*}26+ne#$B;Xop{*Vj9$Q(f254u?oY z1r~i*X|@-G-cLr8>lWfj z#B|VyXJSB50AIzKF63Ihfh8Y1EVPX@7cxT{`Bp(?cH*AMWOwo&eKABKA zNu_J!=;e3S#CiUh-8ss;X*=Epg^DQY;8%b$f6IL0PNWFqc0a5V@3>?sX6JaahNP$I zq@yI76RWUm^#i-q_AE#Pk6x1@mQu~Et4g8SJ5NFQkpPy&&&shA9MxuHwlaVMRyFQ4 zTAopq@gb+x3fbeVr&vEk1~p6oPJ#Z3*1a+`h?9;g9y`^1NQYpAhwkvoCKcvxN){P) z=OY*1s=^MFav}l>RRs~scUN`lHke!7l8-Na{a?D;mK1`^30tur*-PLODHrcsWAh=< zZ$!DEuOi|YUXsHZf|}m>BLYg#(dNMq+M$@+pW46Tkt*HJ@sZQb3GuBk`QssGjBo#5 z_4WQVYZ$P`sZ?YxQ?5DS&HT`??b-U46N-fagCif{vQJEFDw0nI&3@(RVh348Zw9sB z!vid%)R?6}(>xM1~=@1~=XPG2vraIoWu zacSbSbJ?mQxOvRLn8ly^L?rCYcE}M{^7hM@O&I--y`gXL`RXO{&rTJ;df`)-Tbzt* zFTd3LhRTPz3clWV=j3A>kOJSsgmY6F@YN+s4Ed7zG3iRSTb$`!BOqW$d0fA6b}rO~ zOmh@2?ECH=X5&t(8_fB4-j-@bYt=)cyR??6P$sH?gHpCI|NJA2k~7Kr?EG6z$!K7t ze^|)5SUo^DbL=E!K$yLF`>huIj&qPY7tpVEHc!g_dJ2(?AxrhUXr)11}O;7N^4W8Db3aF4FcgVN; zZ-v)5PkdDkvgjAJN41f43>fu35blrx{hbsbF#a53_X31?1SnGXLrqfi6pJ<0QWT+g z=WQg>hc&{<#svDHrVE#Nphf>VxuLe4PN8Xd~3kDp|fczT6ry zyP{^;0ZG`RRjgvk0ij}$f?p8HeDzV6WMy4CRvEc7#}`qYxhpDFq%L64rN3CBXU?HW zs}wEbTeUSEHKh5qmoHLb>GPI<+WQ0pEG4*xxX2M1&jVRUKBvs{Ld!%5LO{DICVpwk zJ?wDOYnRJ{Sh(hxj2`_oGj4#pmY}Gd*XbRZU#=>%3ljPWc7=Q9-&9!UEzN*crF`#A zU0R_l^Cm~=09H^Sb@yTjeDOy6rcZ{2ulIai&YsQLeWlHaiUPXpq=U-wVy*5NJ*xg-( z;+FJ+@N0SBWI%3|Nl1PeYV81t6`F{BsCaC~KOUfwN2wWFu(=FgTB37r)jEK7tXJB+ z&ULV4d%7GcM8dHIjkkB|-dm>gcDKN!9Ecrihc=l49#fxT87J+OjUr7Wx!^UqhNWa~ zT`RZ^CnG)@4APL(HVx3pJWp$w#RkIwCmE2{~5#v_3dce+XN3G$X zUbciUz3C3G3(K=_nVhW2)?~VnTRAro^(t_{tVy<{<3RzxPi?bf4d;FKE6<*fZzh-WULVeeBxKQ zgNFExs$;@DH?kJ|wtd8srHp3ji;WgIh3Q%{&~)=+Y^Oi2P7P`ER6%VAsd@j#c0+-6 z{NnU^(-4G|l6;gL9`flZ*>7ncf> z&HGLtasEmvq+LxNU0?criV;=L*%u&JQTwTEpFecKV4}$vnqIylT0K-)I9X0~1){*2 z&z`iI&B<1uu&_5yK}Z;_AmE1~A6~taz(vZwKpr>1g3@=yk%itR;9do=@4~46eEo#R zD?Z+{Rar+5TvcOO+dD7a&t)KBLZHy#AM#0VGni>iE%odu~F<0Ix3`e@D0mpM++0;DeIPb?;qUMUL zb1p34wX{lxvR$R3e<_K`mV`ssRzt|lA^p;U^7~%0_eHSaccay?CnoQK5`VTS0?klU zj%TSCB*~pG_%*p5EnatWwtSETsIDO^hjMwbY`;F`7x9$-OcGq%iFtKv{%yYmuPTam zrrea~d;Dfl_jK81p2lw=V~VO`T$l09T&^xlgI&*E9t4A7;3}TBX+B<7X5jeLrKFor zQ;qeuDNwG8(G1nS{3CrWm_F4wD~FB0v}I;&Xso+;e$Y54RU$FC2d1Zvk`NE7mb8a0 zhJ|y)q>{BoBH!{SpDyLEC`(XB6fCcD)!bhm<8O{HLAR&J`jd+-bEOkSi|$T>npts~ zd%MQ;i8g&T9cGxwODfnugwW#EGEK66Q5q}9hXrj%ME5On;>E3m(uJQBOzw;)+s@|| z8|4*gl5sR?I{Xy9Hdfr5fDiU#Oc3AUwpDL{=N0pvO z0(OkHH|lizx~c(JnA+1S{Gr}XRm%mp{gHMd4-^x0VC3O_{RkG8j!Q zqRwXFOBuM}PN(1=#`u}2SadyLC&XsquFkrI9(l}2gH$M$(1i2?Mkryae?WLv4o|YH zdh4@zpOWCP7N3MH%GtMZ2*(OjKjt_4Y~Iy#T{4U89<7so?5Hy-XL8hEaG1p&F=htK zZ~ICq{McSM>?={}oS!f`l5%Rnz7^dQJtDjTSai_&TCejXPV9L|OnDdkz&H}Nws@=i zq@|Jds_-H%0v-r0%WKVu^|>YzTqyGPo)EjFG&Vpd0&81m`Yh7TRcTKZlY=shWA+l> z_RQ{MNXHa0n!MmCdP;E7($0UPlzl)psXK zU?^>8`y4qM^KL}(FwAx+PY9^+y%Xv zI%N>O+t?fm^{e@*^+9hb*R|+g`_J1;v+@X@!Yf)*nl&|u?R~J>C*UKacM-yZW6ex| z#MpD2uwbt-H2Nh_(5;r#S9i9#_L|0^dTP$}M2rhcyjRB51r7d)Wo9Q)w?$Xo8_G9} z=CERBztZ9C_alwt~@2(Ue2(_@ z8qQ{W(iu+v=%AD>Yyea7ur5$EU&FEJDD_?2w7{ByOwR6DZgqa1LjJ*>tw^IQ^~^F2 z=jXhXq_v#66LNaI^vS0=OwFBX98!Q+YJ?GOB9!iOB~M*(A_wZba8RH{(4*$F^ga#A z;BENqonwdEAj#Wc|2*1J8-sP*6#{-U2ZF!X%R@C*BCnj>QyHlGI<=t>=N&=0Ej$zb z#Q|fOTZTg*0~#mX1`{zEwy)uiI46PVIjMEvn1Er0V;%1GR}bN2?K~IQ#qaMXfkr0G z(s0w!KSrj)Zpn+W4*R2_jC6t)?@{urvcZM;8@LXC5XGhWW4~{G8RYrRyiC#%|3*(s z76QhE!K-70xjfns+}oYFYa9SxH_Fg@DLQUkDxeaA6_6D4GUtJWr&wzWn7jJnsDZg! zFLD-_@5<^C6?t)CKG>*QXp&n;#%e|ouUzpfeA!cB&*<_kA!S*XpkP9M;eFmfhq)Y) z{=UK9*co#Mf>Xs_GFYVAKC*gixtiq)bSxw^GE@?MaS2yVT`H2DLhINln;wZm&rd*3 zlLfvAulr>4dLhIvPmArM*x2qYhy^qQkG{YT>a?1MpK~@*TB>jiyTzjKk=kzb&;2Y!oc#N|{0m*iBl79lrW{Well*s0?sy6Sg3m3U$QtcCZ1uL4GRSzN59oHHBP+7y?!$rHihbpoU5|2O0*3LrXnD^Hxul(oi zP#c8`M|npF8O;^&0HM-0Z!;nu_pf!oO^`ogj-tjPfSb~NFBPy64~ ztpZ0hy|2p*nSQ&i{zHHGEa4~O4&5bB_{`3EJa^x<|B(JW^-4nB9FDvi&E(tg?zScL zuF~?99@&t~^@MjqE%O5xH9?@ElyL%v-03H7r4T(}UT8i0ypJ!@4!;4W=>n_q9l<*Q zs&>hzM(bW}%p&|bg*Lc0aH|}E<{w|^vo}`uV$G(jN#jUy+!_X}mi;}f4@4M`VjuWn z_?@M&pemjX15j(6d%iuqwy!*FX2-5&z*exF1hm3iEE6X#&C!j;5sUt|3C!46!c0qk z9&;EwU*2t>isJ7qQ^}ZgxeH@Cr-j83gKUrv3kFVHC742PnlN{iF_mHThUXK3Nlk`p zZv|U^jeZDZ1nl5khIMs~SvIU=DkbJbn`lyyrq#NbH6n=eESy2xi zLaJ8NhfRwpL67Nn6Yedo9aI89w;vnT&6Ky0CVoJNrnpE_@a&&P24`}(WA8et^rROIlj~2>W|<|&7)}? zY9ZQld@SmVVTUo$Y8O^bRL6;&gZ+u%))G!!N@Vp6BBGx#4HMZZuGvCX(DLg04X0A+ z`w;Z{IlAS6C4ILcAOpvF+A=$nlw7*4jnA>e1Bb`s(svLhr$v&Lp#l3(wg*4pVbke- zbM8Mi`_)8-&RzlMdZ%RR4k1gqir+zh`~;x;2{(bjTFyQAVMyipr`zZQ*+SjIp(R~Y zB7;J~SW2J&2P9H=^;Zkc1K%7YdQ2VWubs0j`A+Xc8Zu?d zngI;?zQj*=Lh6#ZEklj};R08v4;yfJ;uL^gnVOJWaQGkm1P-M#9C^*aI4#WObi&!A z*=v?-F6z62X@|iDrb3xr^`wqcW>tLGVg>rK_DO^?#4ZYc8*~S29jf4RBl}Wn83*Aq z=~@VL);x#&Lnz-z`S+>M>g^lOO7cH?BiS2_udaCamxQj7_Y?1Emj$!?Hb92LWaeIuuxq3(jl{eQ7O8QBk(1ZH3C6 zDpyEp&i!Endns*C-KZm$NgIjewGS)tIApi}Eu;T7yFRC>U}Z8LD0^JFo3>kvN!O61 zjcxVH6PRo*KM{2y<-5SVh>w%r6{^)@XpvBd0DXe|M6Q&Lmn#E`D{`TlqOv?6o{tj^ zw!zA?D}*`*`Z<8YS(vZZly(Qkv=?v#YL$316{rX*CJl3D5rLTV86=;CT!8xml1kdn zUF9I-&^pr~S47uB%}ajuGtUt@)Y|`3c$X@1(Hd}F8`{3`OcJ0TQsElkrc%BZ>KzJacinH-$!;O;*G>~yXuZbP|ZZApq4(jNTrBG08Y0N z^l3&;_bAl&B<^i!Si^Qhq6N7yJZg_|(osaQbgcRS<;22#$Ybo+)7nYpnPu0W`s20A zINgwAc0>)ji?RVYu75d1GT6kaF3oXIpiAxO$%2aTg%SBr#eXcfV@YIe)Z3mV1_Eo$ zIxe!Ef?Pk#yS3>&R7Ahve{$Z*i?dj+Dg9xE8OvLqFdHkN4-net;C*CaIEtZC8pK!H z&>w)&9B7d_mv_rv$2iI5F3+cSREH7&TV6ljO^xD(n-wdgn+;8yA8c^5RR?D6y~`Cs%5y%f(Vm}?VqhnE@zG5Si6!Hd-%%0~;dsi| zaDUg7s0@%<7gn7H&cu11-Lh)wpkA^oE$*h}y z%tobj8cGBTlpMn!$a(dlm=Q{QIt7~St^^C|-Tl1yzF z93_Dvd7-a(ITAmK)8O2lAeQ$giQY)=Z;$3S?I)zw2&pW*1I6_iFnw(X#C;56ZdPg2 z2-3JLwsXh7I;%>jW*)%dG9G_)q4}{#KW1EabQgrUo7*Xo*&E@r=wVDN;|yVcX}C(Wi6D(Q@5!{|%8!wa6%NuS-1g&^)vmgh@iG z%@oJiw{e&yA(K}kuRy!BWU8+di;Ko-+Dr!3-?P|U%{)T#@?brvXn*x7NXRl`ITycF zu0ef%^kVFfYClx!NM?I4J${`Uonn69uPDGzxOq>4&wf%PM~w~Glg;4y$ol|OV)uZ} zWbOQ^1c9QkfZ1S~Bceh35)rTE@maf={1{tb`7J^WlB z$soqKywfuBCo{0N96g$8&W}CdLTlbnvd5G4)8Ar~(qWg1VYF>VPO!HFyL3$~5jqjW zFwk*poMf+Syqjx%X!iZhlgk@o$krp`lcRDLPaSjvQSw|jwNTt@gLM4WcLfq#z>j|wVcuq-5;xl(-PC6JSFQ9Z%QQ| zw34N`s4*JS6^Hw!BBV~9)+3*yAkm62Ho)r2WDXbZUIk@Xu*#2k=;jNOU$$H~uuPNv zs*U}BQ~E;w{@!@-trT^#mxE#ztzvf0+T(GtBGQFG&hkXtPC#$bzuAS6XgVNIETb1X zqG}eU!2Q0bW}b!X25AZUQ=b7ECpXlIE0=w-gHP3-#u_E`mHx)K@KBrXLovTyqJ)g*@y*=1QT52y5N5|whZkcOd@X@Itb zw{y1u)E}yU7{?JQtefYwWLqnGJDjyk!p&V2?X5I^I#s0SC7q|Hp^fQK5L+;++OvnB zUsCyeCio@5$rPj&%p`u#+2JLE@otqqa!i?dF=uwFY9sEJSzj+ZV`uYI%{4uR^1hJ! zpGTA5D?6y>c$2HIT<7t3ZCGoZw*Gq&=;T8D?%*sg#&UklmR?ke&^b<5PROM-cfYdk zIGcdrpy0O<%avu+toSDyUJ4DPbM*Y9MLnSS>cJ76h)MpdwLZiS|D#aaNq&%xQ_Pt! zkIMCXY4X08?2qJA+pDZHZnyZ8WMWwVZNm1spuc%b8tG0kdfH%_sFo)BfyzcRRrwTzGTRV4Z$ji)WVV-hPlRxnEO)M| zgD6|DmT5{9d3*<5eef&ua0n^DJqvJHvvZIX_|gJWPk+4G%aj=$*G;1*(UBsl;ZMGg z0>S`u`e^QUh%TJe;T5WYar|~F${F575v ztV{j6>foJ`A}ozbSreg`SRR|pLE+bwQR7ZNyQ!^f1MgAPe0*uyYf4G#wm-dA3>C5Z zKs~4*le%p-%v-MyaD-B&WqKjxrXyZN$-~j0NgZ~L#BR2`HTk~bHW(np>N=((RDuzv z3nGQs1^`J#ixHGs1adAxbrBI2Q7CCxc4VO7AWryS4#!#CB6hxjV-M)A=KE$EE?sCdn=Ju$*6-qhed_a_VP!By)mndnT-<}ow22=FbD)w3?2sCW zJ(5VxQ_n;ALd{NxcCCjPsXTQM&hd^*&u`~(U2tQ5m!T)I)5E1u-qI7o&94Se%cAE9 zRdOPsg{ROAell~5xJKVR$!cjnd$4wwGU?{b`X$^`yfnb6uEpbtb9hM8wtH9E=wOXV z*k-g)wl~TRwF~`gYd1ogb|WIeU_rM$**fwbJC2AHFl|)*vKhQp%J*n4^JF@LXy}ty z%KhC~nveA;X3J)qg}dn}c&ZtgZ&Uh`@eBr^s$ub`VIHTv^rEXn@OA=(KIua7BV=T} zNg4{s;$S$*45(|tE-kfix}aIPD}FcX{+c&2Rtm?=eOhHe0=|+3vTJDHI*|phZ+}5Q zxvPon&t<-&eZ~+fPUC2pTGeNnNiE1+eoJGV^LuhKklf_fGDPF-U1BZ1D|a|5CvC1s zG0FN*gu^a%{~o-qx}Xp68hTqgoZc*w6Dg}mbzgtD*KKsNpFM1VSk7!VB~@R%q;~@f zC5moy`ryJh+b%Q*^#)V5HS4`mdU@zi9tyUHeYFt-T9X?ev6|5iN@pJRMvEAwH1xbe zyMRPunM4kz9tqY(3;NHt)E7&;R>1bfNXra^#l1rJMA^Ze`QyxM)?E>sM{kuOsl<~N z_#fUApCzp)B#d)TZFq35KQ_r8y~O?v?IMlMW^TOje0Ff_*y`p>e4jb}#$KgpiS;sX znW~FlW`JTF@c9t3!-)RSj2Mbixsra9^>~C~T#;Y$0BEmK)f{mAxWX(J)2m!iBN2hr zo3(lX6AjD4pMm#s4?q3GO?xxO`G>YEREnXXDY=jq)keU8JT*+cmk*k74%gvlMB9p$ zF#>XABB!X!p!9O_Wtb1kneBCs^O2=0U)`>_qrjVB>Hr}xB9#eRhm#OQ$(M^L?cUFv zA+O_RgO7dAI3!7y5#8^hXpAC+%VUEC69p!w0OY9Lan zxvfAT8=^ovu##Em1UKU3H7C>wcsd!l!9dpAdY&)gb(+5sFXbk)u330Oq%|a3Jk*7N z%bBTGnj1|fkCu9ycF;(J;xsyK^A8_Io@M*A$Xq!AIL>6MQoAant%HA0*kw|sLeYBF zr?^;|$Pfvyag0Oy^83XOGZzosh_yp~tMy}AEVX6%xr8V<6LgM^hs)?J9bHCE{uRRw zD8yl8E@>EGHghH+zj9-|BU*cpA+-Xu8WdrypXKI6fzd$RMyx-C&Xt_rjL2>kY>~Az ztVh1!FbDy|y-JqHEOp!w)LA-x3z-#o0CpYou&}f?E3OsKd&dzH5Y{vMnr)9fr*DzJ z92g3D!H$~aRp%91QZMxVl57x4S(9@`;FxIG2~Ap?;pZ*z z>|xlBO1<-!l`Joe{gdn$zp~#rLvlhz43H#G+y*xJ;=AtofPH7v=UIGC_lUCO-ftoi zb65V15Tuf~p(c(SQdly!mSgjuc>jKtE8Vb+22v4i0!iYk!6Yn_6voCU%(fM)YEF8wa(fqhz7=gC^sE{z4 z(ZxLng!*%Ao}oq72$F?{TXm>T=6`MvcUSo{sNV(FkcGsXgO36LTINfRf%{qg*eW)2 zG^^BOH}YjpHYRw<;>@p`35gmZ4$MmZR7X%f zF{6mD)E;TLeX;}e0VD~B*2!)3ylB!lKGg7nd0O&D?IqZ^KKhZywbW9-BPPGlP#xM5 zWH2ad*j2n5S2uif$bN?B*2Frv;7T$VV=Ov(y><|DGFvH-H4ZGe$*0re4U~tOCP4Wi zH_b9}$BX98qJ8R<$h?gP0Ar^YA8uC;) z5;meOimTOucbv?wLg932#67fWo0?X>dVBh1eF)UcJgp<^midQ2dl`7V%Pqcu&7tgm z@*uaTi=BTIow?FKYH#LxFKX4n236qAuxzE+XgK%AmiFJ@gVKYY%XJTws^*}mv2;Xp z;*&71n~Pa0Bx{tj6bLLX`l!Tl+CZJz-+q6*f_6}gd94s&Z~Sc_eHj4gQ)dWbcm5+_ z0?~y#m<{w#I7%}zBsJ=Yo1|k9jWj|KM7^?wJ?0YaQ`sJ|3nbh{_FPg(ZfJRKQv6Mo zu?lTB;-9^WIAqW66Fqx)&ySHH|3-0q%*G%yuO9>^-0%dxiVfMu*FL&h0rHj(VwTVD zh)-W+Sz&=TT}M85XN04H)`eT!qou+znyag=!r_`8hDyinS+ODpX6ZI(_2rJGN{q_+T9mw*!!pD!;)ezM+omKk zp-vapzYyhVRB0TcaViOM-+p5I^LB*5!*Y7=2Jdb)p3~E`#^m4DYBM^oN+XMpe*iC( z8aD0xMO2&Qx$4j{vf$cx)*jY}#=NpqKfS=M4B4Y!cyOs;G)9FoOH15F5=!;&{A|XW z$Zot+Bk8Z6)y0gsd4Y?nF|_Ial^eln=0NDJ_r2pFT>{l0kO7KtJwgt3!Gy`baG^-} zWFCK>)YpJ)F`Qe_;`L(VQc1BnBTBBK0X`;>oq9yRk^z^yuoLK$M8_`|_Idi1tYfsz z*NbpG#)4lB4xP2^WU%@TdV~w_ipEP8P?BXiuJFs~t-`B4i1eG=LjBEf8zjp$!x}ma zNeLCvwJL@|AB6H_I%If7Vz4=}GdB}_tKuR|zE!mN(SEtR)AhcSZ@|$9s|*f1O%wUD zp$XbDWST4{ z912sl@?f+m?z*$dRQ}wNp404UqPY@sKt-4&?Q{|@R8bx5AF}mHyXmzI*3T=O9`93e zt;KMGZ?lcqMv960>D|Nk!YtAXyeq?Uh0cvIV+DtOox=Flb6>({I{s#(v0x~;EX&pF zm32miC-Vi4O*Gsn0RB2TTw@JAL*<^k>If>9ncs18uJ1>RI=&P00<+zyWJp;F`)TCZ zSZ?M(D=MMh1WH%1W22u?n?H0M#&exEVF%m5)3y&0ZTZ)!>TaWV3Nve*;Bv9gC%d=3 z2fH`y;3<18pP}c`^RXCNSC6bhCwvX}NX-|l*EfF^GTL&?!kk~pAXqE+ec2UmJVR_U}n1`xwzvofe^EVHZBGy5eU(DE65Qq$dv8l@(i zUFc?z6Hz0EokOhJKvxIyavwMz`vy6toK&<&1}B04MR{w-bcH!F?J|HGP1JAxDTrfa z)i8@N1ep(<^+TCAElWE|>^hmSFd9xLbYrtv0C!Ktu)TskLq_^}n0DgG)a&OnfR9rail7`veNg zU?A@{YC(|T326Z=e_(z}(4tbr#^C-!LEHQiidw0Q41?w?Z_V3(sJig<`l_UsoNYdZ z+9kb4F7+83fF!G*S*J`eIYn0YdELLLWtj#dpu#dT|5pH zD8ch@Jj$&C0~X$#AjfoXJ!AtMDbfXot%reI(e|0jNFtPIRA}sWk*_V+_oERwsMM%wEnH8BR$Vs zuepuC+qZjm%jDBxk2#kbmo*HPAab<8ljhx}UkM`24K z`p1CrpqyBh?(4u9fXShLfS_DtW|~AoWpHfstJ8z1;?D_$)@HW{<_4xWUu9rSG}n3m zCl$Z>|4+pic~wx~7}?#Oz4vSXUn;&IU^v8NML3YZr5?axEd!Rqm zko=>A=;oIoUrEHKuWd`GLX&aDa9xjm;XV)tUuC7wdj|ky_;-Ap=~$fKDO5%I_`vlH z&Gjx|o0yxJejuD!oL%gI&_9hn+ySJ>en=od!8+JDzfbYsdTB>}n_qM;|5SXc|5ou) z{!_&dWBv8m7a%VPbWAiXoTO}K#0xT$qOy7| zh^5JC>Uz{$@VodEIc^Wgj=l~MolOl88b^TA!h+D+)bOr<4e~0+ix35ulYNcNbM`x@ zt+lb)vElix*HsZx6LF6BzoYm_-v)u7Z8Okk(E0!XH~{=^Mou4m*FTx5e5Rv(r*IhF z-B;IFfTf|Uvix{ZO8aA2*=b1yRPg&(XFyN)A7w{;2-rki{lkOn??&@{tWeu}D{Yl#UHc&4;DWY{r|(_znZCRsPkR?GJJ=uZvZ?o{)%1!cy9O+ z%)%MJ*$o5cvwg4~g3M$6LN(NT=g0Zn- zU$=X>!BgCqQ~oVKVefV7A3(l#jbAKA-=`lj`Xe#BF2DY`-Irf~ycz6YKt4*Ye}~SF zzTn;nBBuEE#q`$v`r#C|e_%3rKRmxcf1PqrezM*CY)nG^=)QipA$|rYXJcJe`CXur z;qP`jEW+sWKS&;EyGMF=(^I|m63KK zPD!fM7-Y;^_-V-j$mevWp5EAaL?2>)k9EuutXA6X%>_SKWu99PyfKft=#BpBfO7Nb zgj#<8M&O~{tk<&-@tpk+Ctn(JvbPTWKb?G%NJqM$1NdM}oZegmsz2CH`OKx?*P0I# z^%7c|*It|49W)n0*CfcCeHS$r{5JtRa@MqL0H!ewBrw2Ja+u4pNVtZ16S4WsQt%x& z5Sw+>*=c+NF;UB6GAmdbehEd`kPwOv_I|8td?~5Kez=o-@NnM7H149BLD)mSF`1E{ z+DZrOQ`Gj=ttaiJ9x+kVEk8R{(X&l!L?SQMZk;%hu}H4&=j`YopOF0b>?19h6!?Rw zYWT*I+VXXD{2he57e4c+SDDm*imy&D za9n93v)H)SABf_2LRR>2a2Ll{$lWk6`t3dq?y+z%7N@3>bs`;BB1%M_g2$-w8HLa> zBs6DQnZj=mU4oQq?8Kc}3d?JQjN9v+B3Y4aFWvS$WmE#+X^ zcIu(fT-oExew5eU?AOKGW+ZIz(|K8BdR6zg_HX9rs&4N_-X%ziyjbn>UV>4Uq5tOw zkCB!PMITWBHdvr2zI15vG_goN`=i6K*8lwgRigOaSrXNIw&j?BvVYTJ&2?tn69% z^7d%upm}iQCUVtwNaXFIS5>SDt;7coDPV8?UT;=YA!B!`Drh%Jlox(;5sPbE2y}O> z9kTS)^06A{vgk|*yzgckNhU3Tz~9%1cB+WB4n0i0T(*?n%SO-0C%w$O&xlnEbfKaz zre;YifFh&@IWf*~z8CqzNXad`%TIFVkKE*e^1#S8ZqRw<2%UQ2_Ny5&b_ekZU}evx zmOKbc<#oqQ^_F379(hffO+qO_hASzuv*XdjiuA;(wOm$&0&VB4_Fk!Feyk}1Pqaadw;xk&p8lYL9@zJ;NXOba;(KVFF z-E3fKa*ip8{8)Nvggn>q6>zp9IRxyy z;X0$4e;7)37RWmBr+Ck1?P*Ox(Sq6nj`DJh^`v}wzbkc3)KkPb_(V;O*`W1D`=h3h z%qMwm?rkq(W!*+k$D23PP2kk0c~e?X2dHluX(8c>CGx64}F*|BM`oQX(f4uzGCbl znx}tQAR@~xaqs|EdBeUuxTTk@E6qxY$aU9IMTJ9Xm6%dEQ0cN)CLQnV$m-$DuT#}0 z6fzj4tC!)L=-0JsIvza7p+3z`#ek#df{D+V#z@SxaE!=LH#dE7q0zxbNV%7}eE<&| zDQU*M^4aR~tHuduntfpgqRfcZ>dSpDvsr-I+B$8IIhLd{@%1Y#>36O=d3N;I+5A(c zXuS60<1ER5$jsz#AtLN-I{BU+`sw0>GV*iH#OV6<b4^|R=y?i z1>v8_Q!7kWFzQpa*NO9vqQWCLkbM@5(+1blt9-`j4CKO%Iv1OND!d z;jzx(1|lg^nQ_QX9V@WVa}K6GQ2Z9=BRwJsK<&X$V?`UDnI`nbkg7n6|Sga%}h2wHC44wkQD4a2Q5dwlq zx&cT*tUGg(ZunhpU$a0u-K1(DZUIhFyH}jZwk31kg6|p>-`q_Bnq-|ugHL7(@1@|h z)zLS55oM7&4p8me$wc7KYFlu*={<%XfonpyK^^()639F@=!mqSCiq}D z&fvlYv?0*^;qDmX`54hn@r*zC98U?8#s1lX_u?*pbwxn-+?MOze=6x~<5hj0?p zqqJCj0&fo2gKyT4B^&&0yauj|Fz14r_liHUi%Nj!&(+MFM>Oycv1?pvL@dCx99lJUi1%53b|F?VE-Lc|7iGI$?F zi-{@gXR{nZ&bsBri+%5jK}UDFgAcz2J@90TEHbWL-h94l) zWpEUL5?gBRcqw78S;9)`bLuoF$Pb|YV!oL=s9Y6Fh^0WI;|v)Vl1~V*v^+R2a7dq zIccWtZuJS#em7)@k0J>JV>A7t0Es;gJYxU6VC8K}P-tV}a1a`9qAQr&=;;v4Iz6&X z&z+_8iZ8*c7sGsbs7iD;cqvh%FuNT3YJ9;g2PtCArjMnl9Y{}bY$9(z2@gG8uF-ft zoaEb@8VI_%8qcMgoFI7YZH3Zqa2qT5?)$S^jlCjd7I6!w<5_XW$c}ADt)WQ=G07=M zpC+;Ivr;|od9GE|NT|4bXriv=NGQceUJOG!+#t#xE-=`Yrip$!1utL}mtOSr4!O2vAH?8}=LU zmiwlT*FT5WUq5B>R*F$9*BrYki;UV8bOJh?N1z{%W@6O!YE=rrzuPQM&nhT^fU4p% zu2S|4Ki)FdO3NYfc>uwQ+bnSYM$8Q#z_Wh6b7$HPvVbtgg6MC z4`n^srGRWrd+DH{ILK7!AK`mlI-$Kng(DoErV8j#eb2auS*7koy3YDCC9aa2_2}@{ zLAwWMCHvHRSyJKQ_fKrj>M?U$)}v9mBmHNF5PevRNwBF*A2222(`#}NtjW8hjM|e_ zJi#|Z{|tAZQ8ZlNE!zsi%>vZYErIXrdm*e* zxnHyUDG(PA0;8ZD%IZ4z7lIo$5b6+UpFM-ZEDlh2;b+1UrS z@{s4A-#l(JxQ*LTsb%)I9xN$oXt#14y&?`(L%Sst0Ev?5aTI8t_O^doYxOUSRVGnO zzhb-Q-U-m<^6{9|Pc`^7o%S@Wh+MXdi^3X8JU??GPmYZjOh79r?1gqe?m8JI<0wj>cOxugq<0|X_jk zy20iOde+aOEjT>huskT-#}&uxg{9Q|yvdy!GsM`h;v@IqGukf4f-7k8dLC9f%0JCi zRZc^j1W)&pTrVJH(@s}W1C0}VE(xL zqP?6E`;7W7EhjQXLzjT8*-8HK zMah%G^z(sk;jYrZnz(bW0kl#kjAo+2N#BvL9!BERXyV9&X5DeY)ql{a94G#NL8IR) zWOdm|V^1554nY6X%*lutfyWjwA@PU=E~=r_gdPUn{X@g$%oA>%8U`)7 zc=btGTLBH39ZZ!Yf)aqsBz^JIr(K)I@$sun&6*y@+-ucW? z3ywD=$z3Y;v(+mondVn1r>2G5O-D9j002rzVm!okcXP-?rtp#U1U1; z-^q*0r?ITT=v3j^3o`|fCbKPXZfEeYYVKNcjSRGbBE>wfIklN446}hjg9n~Sf2junu816 zs^o+qz3?!xF!*eoCT-dd@>X5y1@SYoA+ZVk^MslQM6FA13Kt_iSwbhJDxfH1!F6sO zd)E-Z{LMEemHGvsu!z88#kSsv`sND)|u@!6-9BYNwOx!L; z$Z`g|L^#dhG2a$!9cZ%JKnVF#4B>l(%Em0dp0QRa%?C%XrDS2r&JD@NDA%;NojA%l z#q~ZXRllhF-*?2dETe^SrVu-avZ~dY`v{gvp-@{qqI~c*z{W38fVS?Q>ZwT0KPNhO z=Qb*x;2OK_Pbra;5@k84HQd!+MeTo&33BNpn;Avvcyr7H(oGnEsrHCI7pOwJ0b>x4 zy(ca5W+kr%NHLNtT43t(WDx%dYv1Ff)_6y*7d9_xy!j|R-4SEE%gmVT$b68S#Ksa^ zQ8=>c?h)X?F@$R?yx|8EyfH1u1VYgsdvprgDSn|yPyvVVcR0T>-if*DMQJ#s@la^e z;{Fjh%bMDjq0xD!qDl;0smur(s6!+_<@CHX)K# zVhFSTICZS1cSxSyA*IMUjD2bY07eEuL3^wSP>1dm?>mDkcKnhxh2rCBLlRo$1TT-A zhA`txOOzhW9S%L+2t^;C-$tn$pkqoNhJV~bnfd9X&=cMth^D2yDI3pyjkJWQ_c-oy zFqRdl?5Rlh8*@>QaQ>Gi0$r@YJ)!=gdP!MyZ^X;j%rU{%yG^P>N3{hkAW&E!H?X=h@QLk47s^qJ+BfWDxVX#zEb}>4`rxXeJ-dTr6hP?>MrW!UgkCb=%}0uuC4PxDsaB*<>T5}!B(JcKXX|`%QQh81YcWl1+u2s=HclS%oD(v zQ4NQRF=_7!Xdf5IU^-9ta}LQEWEjQFRa!AD7mEtkw3mR^Oe7FtI!$#!h(P2z+#^Tmg!EMtj?;NF)tGR z_`o8Rl(~IWXby(~-w6+xy?`oJvKkdP^E5Sjc$Uty+Vzv-WIw#wAJ=!|(wYrPTLaNwFZY<4-WDAUTY0OSh$wj=J1N?v z)h$Tnx#Ahke!ZYR9>7&nt%-rl;N-IN5rP14kTum=HVFpM=q1%|0o;wV$?85+UttTi z_PP-?)R!zhB~?qa21)tYiz)>%2oau|9qtup08)(_^3N z0Vd+C&rO^h_rmssVG#*<($$n_k`k5lz!2llaPD??xkkn+c_AZQP8Y#K`*xW~NKF~Z z4M|#=Mny2*0{i^D?W4N?uw zH8_dY)JF7)wbB?mE9{c4Mcg=OQ4t&pu&~l!C2h*_f`ofnVGt`05Hc~JgyhnRyPBLk z>Q3wk7G%xlZn&e~^+H>i^j`B8<*9wrP+P2Nl|m8zhP=j~B0Vu*g~ zuS^oy3Zc21^pC$Me&6kh`2H_UMof9}BJbQ|lX2KE(cvdb;G5-1fkR4`1+%6w!;agC zFGCR~@4lq|4hTl#b7ill|4$mz9t>m#12Pr6b1#l9)hA}VQF>)Mb%j#xLPSjO>f(+^4q{eKe?grk)^=`Cl2{5LTQ$GT*G3?x@nFGk>2dNcM7N~CToJu> zg|&OxvLB%(&EGUT3*et`S})Q%B;a%pGYiI#YmQ<)uj{a7;4$o!Oi`$C^MV|5`_-Q# zFY#9SGID(#?bNHh0$-9k=L^Gdu$0S3vzg}x;RWFCLJDbu5XPYod(p*gImO7e2{NHZ^g{)U)R*rYvIif zE0`hb$?+Nijqt$(cQu~rNNQ~6R2n4ARnRx8_39P1ajo=hzF|#7z-vnfhZf99wPy8& zwkjvl<&nJ57TxzxzNj+|k|yk)NhtBC*>WMU;QQN5R}<4UQK}m!)dx^9<*Lkd z?^r{AJVR48FrVl}4va?6JD17pQ4ND2wjQQ7T@K7<_X|Q(@U4*^LmgM#du&CaH~Y5G z_)n~Zk+S%>9W;AUUt z$u+4Y;FpEp9KD0UR+mkKz>`n|^cs~#goG4dV>TBpWaXu}(-FaL`T3r(N*qrLhTL%M zQj-2Nq|r30IlU8khqOFvhj8l>>mo?{*gcGTQNQZ&Bmoqj4QqmF%sEo1|C=j)O@9Yi zyrJxl0Hm4oSOeY+55-HBmaA((B#x+@bhL@^qv|D{q@*%rw~9Mk9!BQdCA#vFr|*xN zmqP04>)+42|E60{Bue4WVdcleQ27m`h!Y{YX7Ak9E>FgKIXV{UOX1h%`RhLp5k><-^xh6 z8E|mNM7(@x<+6jIk?!7AlA4|=Xb%gQs%XgA(7!^L3iI*-sW_=PBKffyQ3|u^&&^3q z?6}zpt3(REVEzuzs;j2y)Z_^6wp;XC=~PW*I8VBS%e1A~v7Y|}<;Z~Zs{a?)ZyD_e zsBjWJyDNzT;(47fU8Dvs7v*+wN-0RUGE(0vrd^e7URexgH2a;2l!Iv>XVCV7JrtdJ zZ}{G2@G)lSmsI&>DxbzVT{N^C8;gi-rZBtaLwaPJt_r=lg(cpp2j_W$)UaD>Yen|H z+>#fnQ%1}pX~{+CH!8*INbq1K#emr+c}_DvVN=~RPKK}~ySX{L29_ zQ=+f|cs2cL3ytp&m{sSSQ3TRC4}RYhXKQ(6q`}!WA<67IKDu9T=T;t7m;OYvjxgjS z8a|{GG>LleNqVV=B}^bIsFPDMMgXg1096iW0Wl!WHVRObpj_Jh%PaQPW@YlNR*M!d z4?Q`S2yPdfRnPghc*HscYB*1d{=R=x=9Yl2pF?l%x$8GGXXht8nZ0s(>9&GU4OeC8 z{PLrUPeKf#5<9LmS4CzTQy>-brAy`E)Vk`LtCLqyg0yq1@ln>taJogbJpaLl&)jjK3G6`gQR5 zKj4|e@Xs~X7KJ0@-~eJ<+H*$~k?4uj-*Tu5JmeJ8mK@}S$Qv9Z9$#;r+`Q=BglbxP zj|(bY3y6RJ3K|SskKvJ?Qt@qg^~ATRAq4HkdH0Ig~<7S{PMc!uPVC3MW3L;^sN!#L&j*;cgLChC)U zB_z6Qe$;Z=b62N(uA3?ibUy-gnh%!#jrU662^auZ62g-;>}Ib*6}Hdzl0+_9aBVIp zmo2*okP%0`JuyuRlTP?;5}esF{cEt*-tEcav8c2m3(@$9h=&3Q)dA2dhKjM_kLk_Y!ed`i^1~ z=Y$do^PM6HHWK2Z`jFI7hK|RF$Y8}|7J|7wKsnLWuQNQoe9}E5)g4lJgSlr7tCp!T zLB3RgOExd?s_Ugx{WF|vc7YF*DTV{6()B1%mL`^rU#d|=Rm5jaY7?1e*W{90&XvJK zravC=%yJ&Q{nu5LQl8_)l;r8eMM$#^9)6fA!hW;GZ$)1Z>u*I643Vi(D-*C|DKdnB zjazoF^YP>08@(Y1F|j~ztrmDE zb|=1NTr7-j5R9zf`Z^!Z&F6&s9;CC2Q=N_LtwpDI>J=HZ^!p-{8&++!Dlt4ADX|m> z!VR-Z6tH~#CZk=nf+h4g#2WYr8o&q-IqbG|EyVBEWnQq#1{#ILCGm86aV0y?lV40q z`$JF*VYU-nCK8PGG5fFx?=4G&^+p4RS_)B!n`iA*dpRKCMawO_-*;+chz`0p?1x@R zA!GE&4R?@!xiM!ClX<+5@doRA3zbO!KIf!A%%g@r_gxT>wY1k|`WcDoGXDJ-U5gs} z+bsh#k1geerL)7d4Y%;(W*IQ>$J;TfxGdO&69QgVP*y-3g2DQI+Sg}!2E`Ss=Ca)* zz&5Q0fN|ZKbOw%^kRqWK^Kn?H*I*d^8hJNkkAp}naa!dR5X@Wc)Zlp3wnCtAVfkpI zW?+a+g)bLB>uYJ)ZMI|Q32(a`^snq&5+fd+$7D*U!DenDTH|Rx$QNvpZf=l|JE0M_ zeAq9YyQ6OgH_dNI(Ox}+v3;j4B2GG(gBy(;^HysmLp^fx8f~*Wx#8Vt&3Q9w0jiTC zxZD7xQ%m#wlB&OsLbb3X^Gf?l0QqNc+ZOkCQv)?DZ=$amdk;aZmlM60@Q_>QB6XW~ z%TJa~dpcHA-iw;5!``s#5$o>s5e-K+ACtuCHW`A3lNF;(Ze5cQA(YXEE9;RiOO^w` zHw`y6KI1LbL&Ghw_j`SjS?ze^Tc2{GLaQ;-kUxdD>*UD!Q^<>^&A3jy=Il)`K_A^r z5gWAOx9VAp1UMm)RTBGmr%d5EhnJNF~{ z%~d(s@21VsyDlV|1>cYSZMTQT~K*=)Dv%6lfBYbbVm!FC6JzUF3db zV;DqI^Xdk8|AAzxJj>=v9-5gurq{sUw&FpSS54FV>`EnHkF$|n@S$Mm-u!^=AT(Fp z(4aj(T#XA&W=iv7Av&fX|KmAoOCBLH?hCVfq29oIT&069{Z{WeUNt-6pIRRvDH+X-EB-WM3M--!-uG>1b*I;+Mq}Lm!4?ms(F9HkWIRmr7S_L(L(T|RQIS^Q>G8sA zFtId?xs8i*!^QjCbT+sm-X#Bp4(hk~!cTd~QfV{;h5(YBGpm_6`Qd=DYD^W|1H6V| z7Z~{>RaaGNL@LdU5{_cqyM+1;y1_9uj%pDAQWumcu-9P$$jkB}8aQX#{EC|>6;3x~ z!GepHRE$i`v?Wsl1eVNZpe0I;iMiyf$JdP0YN0J}{o=q#$Egygg}M6y;GN;{5?4I# z1IO$2&G7H+$7!A(vL2Xb)NMUO3`*3p&f5Y%@#F-)w!s^}v;gIu!jcosr28vacD02c zC+SFvsSti{urt;=1mZ74ADFT0KzS=~<7T4>us!^FX{$4qH`HMIq=Lz9hB7r^ zn2G!Fi0m~n*qKomWBO38GL02CJq!4i+JV)Qx94`qg#d+sidELLb#^ntiE_5=_ZV~U z*jIL$1i0K*TM)Ow547Zmw>WP1ljT`czukT6$Rrj_$i9Qo#j3#plnv4QBM`hXED9vG zpPHNC&6dMs84jld=s2*SSPz5cBu4UH!_-3<d;}~rUi;N0B2p;F!kxzCLf|OUgo;Kxgj3Q z=?^?Aq{EB0^@05iC4+4Ld-asv2F{W8x>;>cfQ!y?ns?0ucgjO7Jidk z%JMhRFZMBpW)@$Z?q85i`eyrnRqPri>5_`+UfEyDs`Z5@~G2Nv2*6-I#DiM7ETb$2bv2~7G zSo(t4K2dC{datpMs>3S;&^@^cgKHSy1U~U>yGOX<6o;jGJ>Z8Lq^p!kL%ULzQjjMW zF^@(=q|X>sw4|_I9$^&)-|(IY11rwMXj|_xvD^jk?2&Nq|L|3XELR5c;&E_9URWW~ zAM5;6@55azpAvvv;R}PB;@}eHb$e+YMPKW;(c{acI}hb zjsKXNO&sRodQMqe_CP25WtQW&yY&kq6gzK<|{{-w~+ow^%$deo0_B!RHB z7-`&pd{%p!ZIU}}`l%uQ#h@RVA!WX8-@E*IFl9(vQZRSwCK6{(t-R=Z?3?sxy4k*V z3<;M9h=jPll%=5RhX)Nu=kqzZNMv4Zw?>c>B3M##*6B4@!_(n;+4$fw5;QKC>-rJ%l7IsGcQ~j*(PgdIs)&tSppO#=W zp8ejZ)a(^pl6Yx5B9r==m`h}%DZ2aGRxTqv<)9GsB)%I1R$wG7>fCFCuL9o3T$oXa zG+GLkChNZBf&rreWAvDvm{2wSc?17l#gdVXbC&X8W6j(-8k7AMs?V`0qi(d*+0z!q zPQYpaW+#`rmcYN=XLV~CF_~#rWRnp{g4)I-U+k5l$If4DYDtQE8UD}4%dQSQK0@EB zbkQopMC>&ZtMyQ8n+2hWFnmn(CE(60@moo&=fPO9ZtxwE3&jRSEU-p(%;GtAPfn%; z2uG_yykqV?9RrC`P?4Ep0`2?t@UWb&zct#9AeB%RKJxd5SiGMUj$d~8o(})sL}}!+ zU;QA-mCMwO`Fy_Vb_YA17JCL1`YdVsDcNo`T=ExL42nN4*1x5i( z4dLj~=Tx46sfVNJLQjVBemdR9Ly9BWJt?gF?6ppfa7KxwAQuJ)2MGncqgh?$5E~P2 zu*iQW$9h%eD4sj-7RjS8FjzI*H>D>uqTijchPP;dTK|3H9Ke*aa6a}oGJ-0*IEI<@ zbtGByo53`Wy>LXS6|d6zm`O^gx1UzK5~^?tVAsf>$bSK8xRYSG;oQUvu)JylKLt|Q zGNmT*<7Ofs8-L==Gj?eM*1D)Qlal{f#*y*5t$sKJxB(oI7?m#|U>V6X~Llf>--<@lM*sc4d z!_@z%JQ1h z_Nm%4A@TJ*tdQD)BMIAI>nANJSifp3xb5i1b39chH7dA5{gf)`bV>h}Ek;Z|3_=r; zrYzg5DUj6wItgFQIzH*$;y_uafbq)sM%;)P5)O%^pHLEW-4vNgc|VwM3Y{@bb{ zOC~vrGe1(`MTV*!JQd9Mn=z=k%SuLOJS5g_j%Ct#dDCVVdfN)qopI$T)0sSD8fFU1 zVv_NyX$f>_`hC5Ne2Bx_Z?bQO&6wedqV>8=o3Z5&(rp$8|5p)OIh(7BglE${D=G~> zXGBBop|`d9WLPdzZ^9$xOJDrQ_=2Qm(>qM8aSUKiqjl9=xI^2`^fiHUpzfe#!b5c+ z57ILCe?_oSp)}g7^yh!4ZOaqz=Dk9Mf~R1h6rT7dZm)e5VsMuAw9?~V_`mU@4n}lJkEZkYpE9ng*$rb8*mjt8rb05@7c2jf_P- z&yc7$lZ)#ty{|>hN}#=qcIk*@*+an%0*9XCLtr0hMxe%kuX{8uB!~%M^#J#=5}(G^ zGxP`+MI4_fPLn%PNu6S41RPnwP8knK8nMjC3bAK&IEcUI;`W;pFe8GSmv%zkfQ`f4 zzs(9iv_D@!Qt5@iy3*{Y$$7F@Y7SwF@RnDvuvPTPxqiwZ~ z8kV+(7Ph5CBKM# z$vP&Xq|kQ17r7Qjcgy~)Qk7Ko5;IIb>Xs-kmK~ShmT-NNzLvE9Nx+F(cdM4~%pWlC1w>$MpM_ZDt8yW0id!SB-7d&ap0mYa zI5mD*yW#*3#pHB-9aO0hN-8Du*n#{btMEs+StD7#QS)D~E$VI1aq*_MJ{Y_GHN~|B zn1uiNR3zFLi@hLk$wB%_Kz2vi(!?s3kB%eKk7w1_b9d#koxQA`^>KWhB@+#AAtz`m z&{BjLz9&v5kE)er0qWGYH=~$kz?j@e*B#bM|{0$Ub+SQ*DRuhz6 zlhIQ`77KT$a$0o6h^!`l8PsZ#@KHR3xjhFpf}szClQ?m|k|T2Wwx$be$PVQ|wzQ7| zV)^xWSJe9gCyfzJN8)r)ILsdr>e3?UKlP=rsUymlhu-6qn4tlQ8w1Qwo4F7p(2{yB z@A_8o5Y+0zliMhFAe&{?4i36Tq$vii;T{YnUdrAoof`%mn#|@_{O0Mr$5)4gz?-${6dLNTvFCz7Bgi8al$G03 zFXcGqhb{Ipd%PJSN{J=oFE*nzqV+~nAs_1T3^+~SbugQ24y6(TnXFIQulU`% zYFNdM&TQa`8h9S9X=%`B+JRv=;>E3dvA_q)z0b;qu(=ikJ-YjCw*HaN-VCep5};VW z>eqhCv5_^wjp4hL>w93_>#SzVo?1fVmGO7ArM-2H9K;+(W44B_c6^7U;9vFeB^}?S$IrYv|8i)eg#1sSx zJsl&^EQBeJ8mUP*@m1xKx5j_(BU?Y&?s}Mm#5v6{y~xCqjgJY#t~Dl7Iy>BR0rTYB z!k!9&#g!pc5({yi`&-ceve&Qp`>V3!l`9zD8)a~>vyBL2j4JJ~OFrw#&4Vq94rFa} zHyHkw67c5TJwP=}?Ag9Iv~hmNa9kq|?VOWppI4z|1)?F>r+2Ow{xJ>64`k0`U}^$u z|A8X!MSO@QS^FI72@-*p*EqZhvPQX@riYk0sXb~Zh6||~hI)|eqp8Y=Ijd}CuVz76 z`g~Bcc=;ouGA5`08F*`7utOS0_aVTJ&X#g{o>Fe;Fi4%y>1pKqaH)R;?-X-}qiSKm z^ZuiiyN#XW7hVgGa3YodDDtX1Y(`ZD3|as}>rA<_-lEi@dS3Q!IY$B4yEi6#P^=Bi zI&E4t7~%F|bgX=_I$_32=ZAA!t;C zGY#YKb>40@saPp18@W1wvr+D{N&T?10~4|?NQKn`>N3+y4H{P>^XdA9!m&oi=dGX8Hw{vSNg#`=GD|1Uew#Ky_| zztDLCD0(qV8y8b20(vnULl;vKQ)7D*Qz$+@C}$TZQ$t%QkIhyUZyps({CO%rmhLJ)TPu~HPP4TB%{)3C6$kY z?)FsMEhF14Z?hFOK70SZ{P!Hrvv=M8`PcU!yfW92fD__DQ3tv>MH;%TN#u}+cZgJq z1uBt_oMWy=qCy!`afks^Zf;J-3>khPB9tiO9|%xhz;tm^AV3u*1;QYZ0%0Bu83YJM zi3LAM1WQ1Cr3i)U5FLZ~b96)w$wmhU%o>jfBJwp6h&J|SY|Qjj`@_5fGNlu{Ry2NkPUZ8D-tHaQ86sZP9nhD1_8 zta2p4AF=|RE6Bw|BrLfqL|60$+;|94QG{Ybx=T7edzA&1CCL`t&-4sg5K`*OyZCSt z5d7B35eXmL3lR|t1p*2&K^Q@i7Xna)DkKsMTj$_XS08{v`{hIsy6`{X==b*7@C&|< zUgnMuPcS09cOwWgqGi5Avu9gqM;MVwC>#2SK$KMBTu&u8GYT^HN1q|H+1LaKO+uE} zhBSO~&B2Y7u%+>9tB?2a&mcVT>I0?txpKe^k@$5A3l)-KaeV$Mr;$UDv}e*#@?#FH zX+NUvD5E4wJDd_KQgSF^Wfl88zps)lk@t&a!6*(u44?wB9Eru&b?b3+Gpm`FN=!?s z9Y_+0oRG{4lqLExE=ZE00YoYc#NW!}XVhTpxDpDCqSJYJxM6)CJTg(h@qWy)$HAAq z1XsBmPa>QP#?*t1ooEsTDT9#1lv;Q%zxkEe3)Xc~bnwC`3qQ~^a4_O16*`~&NMImh z!njVK3^@t}4}g3rInh=HBtWGjk>4z@TW5+8yy$!xGH`&=ms4&|2b{0oMJ~wX<)yOv z@)b8h%kQ4RJ}p+GlNcnD`+rN^1+%gN?=FBB1Hp0h{S7@-Vvf?EO5))LMmj)7O$Ih% z!4ZN4T42bvNapP92y&&X^?_KuPyykvhbg&0DO6R!5)MIx-IsLK;Rg^Gzm}GYaAW2R zMb67q>3PfJ|Lp1;f3yY=-I-v-N}Wr8B_m+b2U;2BTI@Jja({o`q5f(DpNdcVCtY7g zTQ6zO;RT!w7QYUiJOp%iYmb>{9{zla)N5o7!CUVz07I}U%vf}WM)#(Xe!GyAvoITs zw>A;Vz`s2D9+{7V)Z=a|d0eN2X8yLpVRLw!zckhqRdU}i5X`^Lz;67x!8mPleDJ@) zpWpL;PCd@%ginz1D;u2v$p1Kezp5Y={W`5_VrGMz4~mvXyK=Op;c9#dL5--oeR||@ zZDU`5pW`cAWHRb1PdewQ1*etz1GN#ACZ(l6!u4~f(_i?@? z;vgPL&1Q#Xvw*gTf^cqx%=MD@ptu#D;Z@u|ABOi@Th$bMfZp@^-c0MjXS_pl4qczN zPKs?-BbiyxpijlSUYFCeQF8$(M75{J{TTQWeKWRLdL8V>*1_uU#f4jEXOzYyd zf0LH!@&pTyuXE;lshGHto8I0U@M({zEVp5Z=Uz2!82J&LN?tvaraQO4$ZXl$S97V@ z^koEYQIqnTTBdrQnVt#_T{Y$wf~}Q5qe*yZ`%qOe!x+b5&ehuG_F}S4{<|8pb~;~0 zc7h9e+t47XUup@pyL)wR&Ufp3J{&IQrKs*3=JxX-w3CPS&TAB9P5o{yS90IR@B-ZypJo|%VeH4pEn+7Al%fCeHX|({tZ*mX0kCIqfRhT`OGGw zEfoY^7_M6e=GX&MZ3R&H+EE5ks;}yWlHq7S#;u39d8u0wu+^&dR6;0w!Od0RqVA`)zQtd8>qWA{q=OqyG^qa*k4-@z1T#y z>|V!3Yk-(yv7Xi1-Zj!}4thFvn62#IAMG~N+`^innCQP1bbltiT70AC!}Rx!u0ib! z-CIYQ$KgyRV_DrT|K8I?v5c8oTNjUw(y%`?HFW=~@K)B=cc=I#^dzVpeapUD#jdxA z3&&mS-L5rS7`LqJqX+jRHT9(Vc~dn~=P;(NV?D`SWrfejCmc@RMquyV{!MbB5v9EJTgFVUD|dtLPNUPjU9dxSYbe%@!fSdu{r+*UBv>l_Iu?tw+QZj5 z$^*81J%g;rwV4MRhO^gwWV-;XzCB&u96Hps8s9>%yQ;iybcS*D<#wHzS^|egjnC4@ zfOS0dU^mw!Y;M%d8XsRr;i4!%*EoDvPtTcx2mKc9-*d1&z1}SJ z@I6oOJwd2E9A>1u@e(aJXEV{MJ)AR=g?LQ1mG8QGdOtGgx;%u9;qF_zjfxz^|kfX?fwNA-`g))b^#<>EY*>o~qGckQc)w^aQH{mqXjB%#c@v6oSFS$VzJ;Q^t-mtstXU1q8 zq`Hub)a1Q@?a|BU+9mkf)v+lnkThHZmKfWAMvF4?z3Y3unY4*drFUi^;<@Pj^BA5> zy?c}JOsD_Ye2|iBAHDGcy1TbpW(Fol!_J}NczKcRuH`vV6i2Jt%dBbpc}s>by?1%3 zqri5dZe3GWvoH0yIubT&k$v!3md&4s@nyZ;vXQHq9TLKEZrU|le3VUpc5Cy`pl*NK zqwu-CUf8+w?XUxVgt(WXlAXiD?xX^T)9l7k!?l~s#Y3!P=wCv}*0&auIr&uC`x9K{ zEY$0&UeSn=YcN2mPvB{VuvEmmH%=X)KX~5bOwK;+>%S8r`+oFNujIo!?(b-t4^2!x zIHa^CYKq)WY&OfDt@^{kZ6B^$f7P!-pl<%3#4pGHU;Hw$|7Y>b%)t3S7r&k=$m@7w zIt__M1wo>5sVy?A%-w8AVU0Ppc4gwfgdIk zfG9`0VogE0!@$-1i_QCU=$#B8J>ismSNd3ixaSruyO+KDM(USGAMPDhJUb< zJ9XNSvT=ui7c-Qk%NDh83__qp9!Yo*q{|42;kG4!NQO~l%E3yoCJcELCPtb~35;H7 zLRC)ZghNfmf&)u-4Ixt}Q4#eBWKgKlficcNiT*`weA0)54AtKQH98%ee-;R^=@O+y zA3K*C5$5jzBLt~PEOKnM;j^iS!1x4&DXpSZvaLB8NTq~a&FBvxAi`)X zuVcemKzz1H4q7|TczyN4K8TGQejWx5A1~Woyu@nVMHJY4W-+&fX{lSD|yyk7z`=Y|Z0B&wVR0Q(Y+ zm?8#b@c@*=>q}^+5Q#Uc47)Uh{YQu-LCOIGlcs>I5{@fI=x-lD{YYggT$QG?i3U44 zeI6w@o!P(0J^*hwc+B;tFrxn=8FWJJ?&F0JETH7`$pw}WiMT&{f3x<=1E;}P{Znv} zgv!hv$e{t)0j@{VyW_PVckPXeaTmiP3Pi6UUWnD3NsLDLXAqe#d4~@_1Pv{6D#kuPayKj( z$3Oec1-3={`IE4Iq1gLZJT+K^h_829t7~&qw$t=+(MobnsX`^>y>+)97rml^U206{p7!q@jXK(dx+iUZ=u7xB}TqrvKuX z#_u%SUsZ=7Gk13mFGtJE;B-j&F4~=rNmxxT+PAQZ+c6aFp%i&CeVwM)mac>Kb78y{ z{~X`fONS~O4}_c?QM-)hD)wb6v z$hNg^Rj;V?Yrt|5Vt(6-Z*)bCEZNg>N9DZKaA@c$_7lF&R0h)b@pnyf(pOyR{5d$f zX3GOlA-hhz$3A&;H2#hg9ox<^&LmZ*W3_W~i93@)hR^L{5#_5jRNP1#L_F*cft%672reos8^t49&ZgP=0+nx1Bf{*P* z{@NKM9OR?5&WGRNwDKRmOgI8;R+=yWb{~DY+rtki-D^PUr1NVP^BAQ=4y4v)GJEmLqc#I>vQxBgm>wT{bt8>pEMfEk? z)Olwv+vH(&F!y@}oWG=98?0mt1;*tHLX)4KNP+)N&M?4Tbek95@mM~v4PEhQy3Mb?7q?n+}QX2Agu=fceWAC7a^H;P|MmR&q_rJcegEn^7 zC^n4l9R}Irjmnxf@3vnvlN3BzW+-)8PpBT8ss5V^J~^u7G-bWnBygtVd1`Qh8NCD< zY8;O{xvBrVK|2^sySO9Yt3@M4V{9MyMt9;_V!cciKiaC=KSZ4N?uc{)8NJ@02h;fb z9bR@PoX<_l`A&nA$}6+dToB~JLRdWH);g11IBuR$okvjEw)o*v^(NQw5yOHdZJQ z_wa()?*CrP9o=B?SOfwN?*F|YZf?>H;;)?Nc|Wf{zkOF%nabi=O zv4KaguXWBd&M`Xyo1C0p++Jj2T&4t^)v<}m0hoD-iK%HYGb4}}CPtTb8yD%y}y8ti&QpElMMZL_?xr~X;;ae5t zc9!BM{2dk@*&SS-T-qKwk%vL!>f%rQ@qk_iE{THQ2} zKoYRD2WJ4&!q~_FyqX255|9nZippq$5)jf9lvNCk`XO$JstykBUH)}IR8&+=E&z>y znyiij4EWmu0LhBF%3oh4&<+1{+5+IY3jdwYgCFL8+4BgwC%-I_}tN*1!OOp$D z^CGtv5bsW20DD(I)UlcAm$Ezx^i~hxp9PfA$^y8_3H-|ux%De;`KdSsUk=|n{fqQT z5i*^h!S>4p&H)_wn>sc$@lLO=DJrTNP@7nq3%xKmF?T$?Fgh{1ifQncesu)NQu>B2 zNFXq>I(Z=v*XduK^TR&dUkSbCpChwAwzWR-osY}h6x~`G@SqZ7yw&Ztgb< z0whCo6X3VO{b!k3y1U=RN$5f8(1@zarRiU2r#QL!uOcqD7dm`;f6ea*GE(X{hA5mI z05dalz`e1==#`AE)V9B4`Gl`{7lXT&;bYlE}#HBGBg$G6WU8fbs|i=;O)H-I~SNPi3- z>{owd08aFy|Alszmc6*KsOw26kumfnlj!Fs>gAWKhuF?Ed~Q zBekRJ4QS>n*Wi~WIU0Dt;7R<6egxEj;az`Z>^<<_By804B{p7=O|B2RM_D=$968S={8aU(gI7c=%UH?FAmhrK#l?8?`Ln zMDE|Q$(-M#J4Y#VY~r*2M#GPYg9r9E!-Eg@H^I4U`@3Sg8~?%Io5aI>>c92Z#$WQw zKW8^T(A6Cr8~?^RG_w4C?LTHj=EsL77lp%bja)ML5Aoyu{1@NqKuy0k-+Miwq1U`U z7aAKG04_B6uY{6>;3on)Gz=bUI4NN>S))p9znH9Sp8`} zh|}40p#(=o{M6&e>nT5QeMww9anZC{y}~dk6Yv%<^2GD)>8ZzSpAW%7CkrC@& z>h$+A$oCZOkRn~Bwbz{pwxLfmvlen~6nN7T9l0FR%ee(~cl`^;MKGVCY8B=&PyI%5 zlEYliFND_0B&SqYc`jb$?E2`i2{YjLR?I?W%@1fRkY-f+251@NM(NigXJ&>0T<2mN zS{mi@*iv-sSl&h$4t1p}k`@H|s#XeO?B7a+0xD<^%i*22F-QP>>#Rjhq_vlh>|Kp! zi@0a8xyHhZ#5GH|pN9wHSzSxiDXXcEs#E3NF}bTJy!{a5)KU@Hb_<}#X_DpIBv5cA z&ySP*pB<1)(+{V6e9RVZxghUPk%WkG@&$O#dSSqe#{#;HEVSJyoQ{&&KvraN)I3Fs zX^M|ZWopN*KO!}XL8|>EA?yQc#FfgYI*;OiM9ol!&SQza?+P<~_ zrGH**`_~h2&&E-ftPcc17!$_fWhhHU!{A8reqVu8{d63Dc)nr(4p0}vetr_Yd`@=1 zXL~7d-d|Ou_i|efoVa~5C_By}==XmllC>R~Z=HSpg_3=5PY}A4;bbSGuv_@J;1x}) z(0CR41#p!K^ttv{22n;lm~A5`V>!!+xH5X4;Ug@OfVm3;D$mcdA~t2aG`zp~B3-WD z3jY%40VOOv!#=F95q{TxO|~c{esuECQY7_at#!u=ggBlG)0uf%*e?AZ(jIE`-c}X{ z+Gsu~oii!I=_gaJKSYz-cB4cUXtNYp^N;%|n{zbtr8>wxjnh+ktZ58;cnY@-?E~qP zGL}uatbSeEVan)^%%c9t#2=zDSX+{VC}c@hUcN32MRefk5vU9}Bt>8Hu=K8=RlIn( zp(KTdt!uC@SCoeM=Qr;I6+Wk~3T=U2w<{VL6rOX-K=zJOupJ53?3H!^(6-LR*sn_O zSzKo740?%mz68!Pq^8>-Dm~diBimyiLU!Z&fT)N%r>OCkJ-It?MxmPzK~|ixo?ELF zF?!EPK{^{_2%g>k!TCso_Pz|3aWPQ~R-^yany4(%LHi!AqPC0jp%}s2Tdafxs0+e4 zQW0S_{wl)EMoL^VmFo$7LVD?5E$%>X1Cf0R`}L*+M&VVOHz}MB!UV6@xl;3RyDTeS z-=;B1&Syy35d|m}#IE12v1NBGU77=hhXboF#(i-vrxhe#Jvw~J5|sX?pswa*{+fYv)Z$8N!kxDiEk>VeEcfa z*n6ZaW%S=pZR6umTlT|6gHdx@Y-%K)(RnqJ|CBn-PNQd$5x!l(G-o~l5iCbQ!Iu74 z&M9(-sC^-AJ4f+4vxfQ({aS+%Ept%WBOe#y<&~LOxA@}i_AWMsi)l#yK`Loo;*#DqlR(hf_iA9P-PtPW!Hf-A=dcG?E2GS8~s}U^hxh&Fy z@RvK>FfaNu@Yx;4jek~k`1GuyFBOi8o-{a@K@sh9BF}4L>lv&wVHsE+p^_rZL~Z^y zC7UImBo#b&{zK@g?}`BJq~Q1;#%GVqNn&3*hb49mSko9WvpTIiZcuA{>d9J)z*<;G zTY!A%o*nkUrF;pYk9bHS#}%7`T*Y92Uji0)yr4w%J^=LXMe|lsbI5ygv{@BhQFI!? ze>UEh>r3kegYaVLgJwgsfg*R3(0fmUNi!sY?;RAsu^(hU7bv(h?0Q>zn^)N!#VTzP_Z z!*)}2!E%@*R(fZKK3YBY*MU**D?QaBdu6FSlfQJN)}v`KIdY zETEIE8f#tBYhnA_XDXNK9+jUQs%N<#r|hC40n-*&UF!Ik*XWZNj2Ao%Nh$rHwbMvH zhZltJ@aNqqW1k?)$63nZfVHKQ0Ql8Z{81E0_8k?!b z;}oSE?{WW}bkLp%7d|q={ish@EljUkY1Z?y*}XMyHeK^()BgS`ot?`I&OPdnV+8ab zrh%*Sy+h|Va7MCmO!dmv-GX35eK^M2zrJ=P9G~#>$1B zkrb89Z1?tr*~@`#mtb3nTVA817JlXBy2Cy?WG_aTDC}N0bk_^3d6bvPnN0lL{XoCP zx7M3XKEz#qYlg)OIrO(^lDbfhw2;#-doWvN8Kx4us+G_?u68awiDGPIkrvteW}RQ2N5$;xIhY!D$C@taD}eDT)M z(E;G5@mD343RISp9ryyF3N|DOkVix7u}$)+`Tz#zR zzcu72mk>Pr7g_feA)Xex`x((Kb|b?mvAd%)JjcHy=jUYjwzY1g{aCVDyH{ywO|)ca ze=Rxz+Z)~-aqWMzxvd)8r4|H+eC1mD59MMWrC`8D9i01?3Mcz#*Q`?mpvHYzjXS_y z5UVt$8!wv4M45fYD$+<*HX^al z^b)QrFk5&P)c!AY0>?`@j78}_*<3uM1nL&Ub@bbr(q7mgy z=o<(`JbfPTtJSWnPVCLhjg3dn`+;O}Yez2MXq`HCYL3WHZsPs*e}Jc2ucex*EjwO{ zMyG~!-?hX)STC(Q$zYUZZi1@VflI<-xJ!9Al=-BV3f8!17pD8HSEiioA`a@Q!YIqi z5=|vhYXhaVQ>PG`ucVo^-su=R8z-ZfL`h4H@uQTi!6@bUbHkztFD8*2ys$Qf8VL!! zcOszto()@zq9};@3VC)36WKfUIhBJQU{n2mRr$ADKD}l~wUZ+p4H|xwtOi94X+@%n z8kZUq{JM3yr`0tGp?6;SI>6#>(8zJB4>0B7@vGGij~v6jzh;~C1B#aQ>A(Z{qv|@1 zaJl_U;rc7ow4D{Gk-AlWXgGBD5?5=t6ykjgBA;ZcdNZjN!H~N_UAQGp#^Ute5!jiS z)E$?`jwfysCZEV@t{UyV2CZpxCX zPx(L!=fXymY?e%V3kkb~Ds|@_1N><_ozN7+POy0=zky>!{5m7Z@~FYyrwH7Kl?*K7 zJ#5GlhHP-_s`+4M4WWG7tvbVbJ0d}MN;vG4seYW`p2jR7SDWK@&vr-Z3B6{_4U9|> zN!9ZONfE_l?>a^bfm;cW6e|WPh7$CdWHmLzlq<4ZIAR zb$ZnVO4-~EpY#qJ3LZYT31nNt$gHBQDp^lQ9G#i7I}thGq%t^U*+4)9M?Av&;rjIt z-_0Z)DNdB=;}pTg2*sh}Djn3K<+fhn!i%aA|%{*Av){*VZN z+Gsswgo_M?Mo+RTP)j$Lr*^%!S2wXD=n7|uOOjxFLE%?e=~e@DoUqz1Yjk&;j3SzK z8-)d5pocu~z+3C-Qr+0e`p!#95~w|7#)3sw$Zul|G!7>aT3(uDRfHVM>#uc)^gGHB zVQZ1X;}b%u#bcjRng7zlh9U~Q;f;aS>OIb-cthYHl7SX7WIW5B8t!ArPfEIrn3?SG zl6cIl63s|A{;4vis_ZG}^t_(deKwR-Ks}yTB^$qIB7av1=qE9uNJME9mY}#9vFVPK zfuCrPMhiRuK&VzF23T?KA(*VS$1LPC%~-1>)yffVxw^yHd?PqGK1c{&?NJF^XE2jS z$m}oBh~VG@^=eIi+n5Y!8-E6Y06VlpvpFG3HGeM&5ChRM0yvVa9qn731mTX^8CD1I zU>f6k*+?Pxb+8krnWn9y2Z~4%ph2AH)-6^WopFTX$8-rbp>fR!kg-C`zH5L&orIfN z)D>}vn_R)Xj^pODID$GXOqGvY%`N_eQ{NVINrjxSyOiJh&e1V;o1q9$yH>(TBY%Rp zF<{rv3vp+;M})O+|EXh{eEDOr;6^tMqvUA2p86P4i6KTxfCHS!4=HB?LA~#`H!5}` zu(5T(#1{X1G4Sdn(Dk7F2S87LR5K|*H~cLLK;ne9FhgSrzQYLwpdW)*)|_z9;GuDN z(R%6+(-ZZsd4r`TN%=xD6 zQvtn7l*dhR|4S>5wXdf2_vIg|uTc+?Pj9ijd2=M8>#V$SaaWwP!y1<1Q~D7$El9bn z#Iz=Aw%d=nw6Vf>oQVj^AQ_y!V4-x!K9xSXRa>%-McF&$9VgU)uIsW6c@J$t(6wsNU# zq87L*m0ssJgmhQ&KUr9Om6-ui`fxM`c0Sq9$})99$|zpyJ~-v2xzEfn6>!JS%}(z81;=5M){mVu^frTu?Fi^Q_4&JDP4rFkEb^f|1#_&I-`D^@DYa6ix7CacYn!C5ppvu~PQi*N$Q$$(C%-g_`9i9X*} z%mI2jUU={~$8v=`p2i9Kjne(0oCVqtZ@rm7h<`kuKs2Q8#O6USX^TFM@zb6p>zrTG z`Rqaxj>*V0`atblIt$OkD#qK%`#yBoht^;ekc3f4HbyjnkShbkLu+RdnQ0TBm1Yij z3P}9+pIHoR%np;+<=|Tr%WL}kH1FpRva02KLMeM?xvJ{e6vBr`!f%|Ou4JuGGxzL8 z-5c(aGH>XSOw1`@{TVs$Sj!6YQz!PUd``FBh5~03NoT{e!rjD?p(6f92-B_`Y3&Nn zAFCi8eYutz1&%(~W;;d;DPBOBNa`xQ5Ez7P>28B5B^_Gu_&+-)9+zb0#E@3AbimiK zrKPPQLUv+S-rL<8f`@oGrZQ1I5)55CX8Lw5y|+6_5mKMQZUv6A6@oIY84tbd9XQok zGTcZg%YOHoC)rUH*=h_Ppm}Oxt^(Ra%1wQlz7WyM7jW2lUiS#ZUMI=?eXsr?t1G%g z1sr~)|4KQXV=?X-uWW&$AhzDWF(OrvJ^AwRS$my$&6xJEW9f3Dz`TeY-wOIN zlT2HRdrmZMtD=!un-ueM2L$y>UxcfXzpC4PnhGd@;bB({!k%!c6d=MYW~|kb)-y`M zP0KfoWjPtiDB?I`Q;mu*463qB?Fuvd)Lr(1{u_%j7CVjZ(W zdF#>1cBYA(e1oam&5?GR&8f5c=Wfc!JRV<71M7`s5$R$nUH0)ar{cEYG~^H`Rjj1# zRl6jJtzf1((guJEPN2Y8?m%J4X5EZ^^&Hy;bla(}0&pS^8EX343Bhd1_y_q<;WJ~H z6?}hvf};!1S**>zqqV21p?peyQNCVF#w-1iJ{B(}{a1NMi>l(6Oy&Vo)%7erc{126 zg|#B9F{C?sr6eP!q5%(Gh%Q|fUO75>IAyoG&Rg(jZHrN@&VEU=3)xaJ6t}jgXw<{f zVcOVtlhGT^qTM6WO@F0kmwc$zzKnNBD|@JBx=TW)_ijI#!YB`_3a+$p7R2s=mqJG? z5EV{XXTw!fPv^L-zuIn}oWrMk`oGRn*mG2a6}oilFs|l-yuaC|AuSm?$+T@jCCR$5 zmExRWTafqEQ#r>wmXyae8n(qfXW@~q=)KFs@d~nENYAiS+Uqe?Onu{h4RP}J_^dYH z_utfUmiM4r3ydM>|2|QHdm|g>>t+nMc>&YP27-Ejr;)*d(cI&?+S%n+AW*2WmXF2I zaaQAs9HxOQ57d}P;Zq!G5>(d)2u!iAnJG0Sw2x8X&e@=5u5eh3DqptZhQi>S^NLw3 zjJoprI*ur+OstQnxBITD-OVn1vFuc!n&C*#uT-D-rY_7{XN`Qr6VV7lA~ z3r}83?xXaDf>F!%sh*IRt>&Zf4%+ToUY0Stx37ug@{@GHACdT%_k>2;h^#3^d7LJM zf^aCvxrD)zJGoW&ae11wHN|0@p$@==X|_M(HgPXMGLpNbR)U&@;9u>&8lGJYX#^cw zfMR-|fdCm(1}6<=zxKqiEFM_;A3Az^XjrO!FhtaZgh;WU?F3A_PK-q2Hj&l1WgIXq zh9WD5nODOoovl>f?dR`nl`~leWSE+W+>v5O*kN$tT&rdx0p7-`X|w%9lxWf+wncio zC9%0Gx6d`NsLEb0Dwy|8k;iai0AOFsJh1;LqhWu3Z5kgU{?V_PznCc23TMPQOdK}d zqpjt;)`ZXVePx~y&{iDdPUK2VK^;L?{VGG8*!(x6|H80^7J^N+$ti+Ktfwc#ecahI z{|(R@+dhxQYK?*AWEUkyZo}2uc4K6J5GRyEia1|A{G4?pmncs%N()VTgbtG@OvV1o zJ}Im<^JZDn73BHRQ{Mvrs>-V9mVaBebq#=<_!%LnQLHjH5&+y2;wHY@;x;7Ei z#~q;uE0Zh=W%9WRokW{{E(4HyCDkfvWQ_jsf;vO7XQAmzqhL-kg#h0DcX5FLce3Cz z3K5Tz00erY0~jQaL4q2*f5XA8=*>4tenKj|i|iOt6lLe02U0+%pT-+>wOOgz1qeRw zG6I}57cTJ#n2OX5mFT`1^biz2kS*j7iEY*?;~k^## z(3{o7sf+0xgWcXCSgS^S(lWQLD+@IoKZc!JhL<)#TTp{nR1DEb(OBl1@}f7iz2~rG=^bqOlNh4k88GvYz@Kqn!N@eKGq~iKwY{43fcb z?ALeCg7(C6vHxQ19F~M(0wvqFyT7(=+qP}nwr$(CZQHhOoAb|NA|~Q)?y6p)Dl7Ax zx*n+%a7n}nWoF#9?B0Z@(nb$$L34tW&4s3#2g{{O1#P#5!T^pS+Hr96gnXioOD#an zSJAVF`O^lPy$(2`<(;cdx{EQyT^~NC2vDVH zcg(T47>Fm5&0MPbLu9M?WOQFVQ{@L#00T=s;yKJDzOIo$6X7D8K#l+gA?UDAp8|f( z(!Y$z#p{ZI6lPchbuel@Dj;4RNo7lV8F8{ehy0$hc^2@fw|>W6$mBCy`!t*SY1C|* zsc{t%9}D*<)0Wxr?V>1VHJAxn6Qb?iIKAemNGDw?wR2AjAxS4qkE)H7(Kg{&Iu)~! z2XvUXV7Y9*(stdx;PMljP!NhH)P)CoiBz5N0rAo~p~^tSZWixR(y3HdON|mfV*k8G zJnC$m1yN3wvpZDX7d(nZrY>1<*P)%on3=eeTAk>q?wGcjRVrPH;0 zM(v!Vo4p!%5{F`6y=78$E;*;xarda;)STt9E9!R%TieAWuUV~Ran1?Ht&=>dDTnj< z7{6VK|FknUL?9rBLG*qUD0z&5u#zq|oO39M5>0=i_9Djf8O){wEuBHoa-kxBU!qNh z(Ly;Wo1n*0)7q$j|HW95E_X-sfv4Sj&&<${;R%8*C+BQ_;dkh6-C{yT8_2Rv#~8F9 zCPr6XR}*9{j@5K4VO0iPsBjHCM4Ng5Q#S6vByG;n-t)jt5_}h_2;otE7Cw)G=aQ(m zx`8D4jg_#jPc@w;0bh?rt~_M?V1$J~fIPBQq2^ksna~<1<2|G7xiS5Thh2xf3y}L0 z!AlD!ZKV!AHGO;RY}#x;6!4@39uu8AFzDP>IEjbnX&=yImpwM*4f0go<;*Va47jd) zCO)wTwL@f(3b#p@7B7yssH0An)j|R`sVycziSe!vo`5oHlhhwR`l6NaJm=6H-wqTa z5K4vq>%#j2?35M(3DY|yzt>btpC?|>dl#)OUR{17i!x+ zG?>F?>?i1}Z?UMmZyla*6W!irq07j76Q(s><5Q&rP87JJkt%SMf%~NWYrm59b)P6ppHX8_B(QJ!u&55k zuX_l@!vCQVPW}rxw*M3{iN#j^n|_i(yM49#E5uW;#`gA&$x!j%ualyyNA+9_B!(*gvDaTV zjVF<*s%CnJ_B{bamXXLZUbvwz$!X1Vd%BQ2^JDp^B3!HNA1lO@4InXt(}@M?Mr45>S4E1U+&Rkac4HO(7)z)BS{n|zNY2dK4qn)|pciq-dW^RafMkWw=6tjl|@xo`qRi3Yc9nrOUh8i+- zbL&+!2AYp`Eht?lw2W2jC#w^ry?FX}4wFsE0=rI#sl;IvT2s%F#p-kG^Rb()9vY3d z(~{OB!$%9@Ji?kf%GPE@NTdlRER`xY{F!)b?LJ)vtAqfaMP;LjF`nc^L$LB+UiTIS zTq2&lzAZYc@M8mrr3CdHao5^H|CNm=QVq>Aes4ijcQ9b$yT@qH_x04s+>kT`(^o1E zkcmUU5oqBcXi++066!E?Bk(CWhd78|?K-iD`++H^_wC86P`{PHY*cFY*P#Hs&g)nk zf3(f(TvR7=#j*{*w@T`q)=T5ht9@1`9iD@a3eU!PKgwMW(rOCZvvOMLE>k9VS|Rah zKBX|~O2=^#8-Zq7x=<=SfYWy<#)#*ouLi+9ouqe+_wr$zI)5FcBHn25ax|+Qa6*w@V405Ek?)NH@M}DIsVE%vPHB`l$=Ij zzZhS@bo5M;!41EO0e7k>6nAxZlx3Zf(c981a;7uMDq5xdbEq_;oC)x>LU5ss&Kk^q z+_UF^qMlN5OYpSSu)w;WBpH+uF2FK1kZQsW%tO0YlE*?H>vG7Yw4e+nnHkQ&O>zUdCViGa zPIUMr{6ak0NhPhiaPQS-EF1$Sur^sh3`Or#m*YBzMUl4mc2pST>0$*5Nxp63g&0rB z38rU){#@nGq(|v{3`5Bc&!|9F3(^NE@udC2Y}UHt8`49X0skP)%sIJ3_q*fEl2ejj z8XX=LJcg-3DtOAFyG~P}c>&kMvIZ={jtj-RSN!cPOGsa$IiT4c{lN7nV+h}z${o)= zlNLp$c)e)X-@^m-*Ifo`#&`c%8b}&5i#lossbFDboDqNT;*CQE^y_kEKXE`@vBT2i#Ecy*j?-k-T(oDrpUJ1Z-FQP^sVb*Cf9 zr8<#+J~fU&#y?2)cKYuXS4Mu00bE}O&&mxpDPgI^>={W#;YV3Q+-GZ+wy^m}Q)s8q zg`3cmNV`A_qe6Exxt%ZH(@n7FsTPhhnipH1`}^4C$o06wstLB_bP@bRa|#Y4)^D?f zr#JxmWWk5Ks;~n}BdUJF)k;53?b;~tF3Ej+Zj;oc{r1-`TEQPzQ;XaghRo_ZIq2G;@~g>$~F0F_XpuR4m%=ceKQmbr?SFk6vC6U zx5NN1FTbBBaMJ4n=TvrT#w$!BkqCg?_F@{|)Uy{fdA0JT{7$;_mk+O(Sq*TqEz>-V zu^g)fuTTZo1$T3-fK};jNKZ#W<8#^)TvCtm)9{F~H7=S{a>S_H-R*#037#u%S;pND zR4<9;Z^^aytE&H|@ADGVm&vqK*pYF--qlF`j)|&w5wFv6J-UcTw@OgKl1i?)Qsa;R zSm`}T9H9;WwG9!Qn~SJ$nEav}n6&fb@3BT&WbQvYW7g3r)TtMKlo+I?81 z*7zF-a5Jj~^GxERTD8~&Np|qF_1%VB-LX{8dn*@+-$2%h7|XTkVNiIE*F<7T4Cn=@ zGQ#OhUm;5bZO((}^(?~T%<@VEH7MTcrC&l@-L!opitMRwTTV8=8P&{Urzig86}l<~ zqb_)tc%~Vegbw%F)57NAR2n2^2`pnP$l+qSd5f#T&l(sb&Ll{!E3#ugm~qbcV4%c?`b^J>L29;aV`isQ^?;l_0Kt*|5q;*dPu)IlX*O>YQJ9Rx z`VhR(Dk_&X!qQnqE(bB2uVsKmI9_?&g=kn+Z1w#L1`^jZ{+K9Eb{;+(t9Rt!5@Ug* z=irQnKt>7C2ui2Hfhoy=lYY`HZcJ82+~?5BV27;2PyShI_(p(Ft6m{duC`{>&sHA& zA_p>b!EX^C(or9vy8pq5L?BA?VP>`V>~0(WQ>S+P%`CRxSk#!T_i>B+0ClSf|`6&C7qpyX@d) z#kQgFj>E1(|4n*+^p86sMQ+AKr?7=bwv(2<8$75iT;_YX-$ZbnFO=IikjQZC`3UWm zy_ArDSqjp`-h8`vO-uKNTsGm#stsPTdzb>uuio4+h|jeZ_C1sY+=;n4NvD!C`Rf1q zR3-U!O_f2!K%CidS9DAE#$Iou;M~s?u&CKGZb~c2N57uki^nw@JEGUSWv82CVa)`t z>HLG_Jn{3(57htyw7{lGMV#qkr`*~MfJ%lYC&H}FTuh>{$fB5$vJWTHjHHH%xHvx9 zar4&>2!31)BlsA=EhXos;$S(MfF|s%LJB+CTC%BV^t=T%RK|A~+KS#n(>JsBb=1-N ztfhUvG4x!dmetk@dJG36GL| zHtMa9v81W5@M3B=wpOLKqz&L3r$qbkxeUfsvNa}sM!sr+RlzI-4)ZNc`my!dWs+VB z-5^H`w>*PEvKLh?0sPL@KMzIkeh5SeuA0oIb0v5}e$ujH9SjU3HkQ>#UFoFsLa1Z3=L|nx~b< z*>H;tMI6`D?aG$g9u5jK!#t7%rw*>U?9RlgL>~!Lm<1%@L*F(&d(hbw%FEsOpa`3y zU-OKhXLS%0;r|cGcB5tXWEc8zV5o#fh{%xXAGt7=@MfDP{oi$&bYTb)vnNYFSnTQm zDlx@&SYR!S>=hE8TjXD2exmJqJ{q(dG>7Rr8iBna{*5tFV7(x* z!N=7m{^=aL534@V#3RqG-R&oN{pG0U87Rd9?-%#ZU%ToOEsIoHwOiiv*if#ba^PrF zb7}{mf`FI0wmjs`DW6z^hksE#M*C$to}Fzm^Aejj%acK{ZWj3y^*f3I5!>~h*Wdh5 z!#NITsJasDO{Fh9vXdi!OT+T*mvWoTMYfD0BBx~%c0fO& zwT42zO_**Y#pYCgZN5ve9i$IU&<#dXjA>(vTC#YmAJ}Lm*V91l( zb?AQ!YD~QHrx}O*S%2D+Al69r1#z?mq-v!K7-5;s{Rj>ah$tZfyPuZFBa*n_PLK6( zAY{|Aj%C#MdQ-Aa{}qA~s$JyJiQAs!gXwE^o`rf>t+ z($2r^Nr}~IA+8jRyn^Lk{{})GxAE<2cLFbY|My9x)E0#n?sK?s zW|32@C7y=ojuQVo9go>N#m6H0T9J*RxZ3Hh4Sb^-Cey#;Wi{S=IGs1gmq!Uj*-J>N z%8|^YSu?-#k8#{zih$;_8FbMg2WVPq>$)MT0mh_ z-%H1aD|k9lz7wJv4WTSh{zXf~^bsz;`J!>y?KRxTnyqMYqQF_6GepV|m&UJ>#XMs#J?c=^-^&)Kis83t*sT*@bnKb$hDT@A}aWNcFXej=@0X@_Tb?;POIad^P zK;@{}Nr+86t9m4EO*&5efd6S2C}T%0JAqTnzhxv=anc$Nxm}Y=auoR~>1<%CHk#Hr zOTBi?y#Ql)(5k1&$xu;FVmX}1&OpEjVOxORgP0V-ZAfTww`!Meo(=x+ybdTqP)o5_xl-{ zXEEk6@RE%@c4r=2e7IwNMf1xQrP3$cs8qM zO)cmPSK`>)J4B7^L=~3|p5jEMbozv+7flySCN?JC63UO!1Z)XxAv<84qC)B%WL}>8 zvA`wlG`Jf$RZb*i7=fFcjlzMIer5EVz!?IAjMlBE?7mc3XTM2W002rOOfEq*VFrMQ z7Whc(iU=j!S%PYZ;VLVXiw)iMH4vQq6{lKeiMaj!%T#ms=0oO39fI(E5G5O>Zd|xn zWS~VV8pttnLn-E)TDwPCwofB^0SyTabeLWZblF^~V79ga zXd#woro(YjDnk{TqnF7#vvgd2Hq1{Ag%2|)sUQaEiZgbjR1)@Mnn4}#i4SY? zIZi#W93Li_a+|YulU4Y1L^#u5&O>0kL;rF%!Y?u^+2a;~KMx>PO2kNU>-5jw87wl#x(H z|IxUgBR2*<2e&bua^q&12boDxeneay8(ea`CYYSp`1(}Ao<$=^sXIV%KE77V)`pYg z76Bud;oVTBsAX%o6i4~sF%;e#gml^$S`{3XUXJAZGW{AGFR*ZejY z_3?=1?4QZFD@a60uw?rCt!;aWr0vl=FiW!wYmzbb&S;lVfd~FFS85g^fYqUlD^NI% zy`j_3KDVH@EWdPpKIWN&>$&CDR(LDTJrEe69XkxT1Qs!pnUbtRcOppPlH&3Em~SR! ziJ|le!t{n^1I>Ojs>EIByaDy4u>g4GN}Hgm;;J`Mm)bduuWQz6^O2=gigqr8&0~;8 zOo|16c_m?{P=s}Sv9`+|p$xexXjNFpEThN!Un%hjkbE^b=|iG;GIaB0LzZZ9$Zey08$7uHHb=O{o~={U zMS3hxj`CvnE$m>2X}#p`*_Eu)p(dRoG^x*qYed}$syo=E1jvMpMMn8eSs+=Z^+1o| zJZmK;R8e{PPRozmu0&_bfkeoMzdsZW}i}$h(3ylcY2-tPc4&u-}-84jj;= z^j<-?UoJ8#M9n>15ZF5s6`N8tRP2OW{e+fQHbBvY-Dn)Cgk&D5ShhVn^%dRf+@QQM zh3xrtmFHN>=_Xc_YHq1gu4$A)0p{Y>WU}azLXkfi8 zt*9U}Rb1O#SYD)h^C~_;5lMcvDnd4n7!x5OG(f;`s}`e4SUoG2sG1$=<$1^v`f*?f zn=)H56`*hi@LnTrTXp}Ix~4WmMl;o{W>1xS?u(xq2W-$2Ob{q7uJ4R_J)`q^4yKvw znbupI1lsrv+ne50`dFZ6#4@I`N(3j6@n6Oavp%g~6G_T~ zKyP!!>&}#G=@4JYr)Q?vWd<`R9^>XMwf`mbZfIU2DizkCQh zEOh&G-LEl%zi+ueLZlmN8z+_Bu2qRTh7O`CZ~JQ8m`i#{oY5)*;V_T%V{@RA2;di z(TBmsjbW|G&chqhVyWB};yCdJ2U1Z!-mMC8 zjGH3cvyvtN&n|7dFrDf9TE)Lz(bwS4*z~ng6#6yhMjF?YSPPpgklLgJ}Kv%S?( zPcm<5&Y4>?p1n@smA>jNxm_?2;ZVmjAfe_5>`==Kfpq~W+iE<~h+L{9vSC!>{9P9TVyd9@5S zzK9XSpybyn`~FTrHwpzA)oHMXHLZ{Pq9Mpvw(}reW8V1htt(ZYX*)KC{PS5kJZnO` zUA`U~ZDTN_;n}ABHz@5yal4Ej1Mt^j4`bl9IZ1;r&8HuI(qGxl+PVgk*#Z=4FE03u z_GUV*T5NqRnn?3{A9P}Jg4Grx!!$#i!$!KqDPW|1jDuzo+q*L7W=El9$Y}@u0VzP>70IjV#{@rlcF?RT#PCgXtfAVd6oUXU71kNJf24|i^QdIVicZJVn z_e@5*;SD}E3fY=Tp}E4vaNlMGNg>P-yT&KrJ`=Oi51)TZ4K%9Iv++u>SoQgtq0w(z z2Nfd&nNrz_vFWj{27C%bcC;gYwMZi_g`$;aUECv5yE{J}z`Lx{9FEx*SJ_49&U2HVimSC#nSqE@Jhy*D&Edc>q8rBXFE}AMydRpO!ka=E+*x^Dq*~N)+{?gGfRVltU5AT#) zlmxDxVNke)&y>4fYbVXp1}Wv>Jrb+NHk#Q(dYS%ef@H6%+Ed;G6G)jBCWY17mi1zI zU{U6{=6IPLq5@L-->q6Hh+u_1$u;LpX-5`KZ+K>Vvo$ICw{=MHz-VtL$^)@+I?<7^ zR*ulP$*zj*Iz~j{^b;{ixFr{#1<$ktKQe3?ioCA1vu7b}{=ZG4=4>S^N~aHIDI?32 zNCTzI!b8k^n+j)qMr|nQ>%qx@-)?`9`c2?EpUS7ijd|qXklJ?^8$Jj>!hBj5hl>~b zP_6H%Ni;k!p6AQvhaNIWNfIM_)`}AI%;>8ofj`QdsarRmss;Z_EwHMmh%dQpY;-rt zOuL-NT&;6{L!I&6P!ft&Ww-bNA(YyH=x|xuV1uFO95$ zyw%>Qq6D6N*#hL9!D2c0Qefe+*f7an*#$iS+Vj%-;6ZjoXl>H&3|YZ-$ca#^mP%l) z^Z9QepnEQ{yuo1&LG<(DlVQy$7UelMrY=Z|x$Xx`Nrga4q-@66JyzWr5;;yJwOufj zl+%7to%A*#a@s*F(t0u@YJVHF=14vg!BIN1gR5+GDkFve%$WO64~SbUO2;_fy&jI6 zi7g`j~x=>@9Zir5tA$nUM_Z$bUOntR?;6XC)k02fJ2bD zxJ4uqi-o&L{_bJB+j+BNpX=b5ySRL_0pmdvlJqMYwKC*;j9?ppDqy<5ia|=PA2WvP zOGdd;m_TcM+~qg;4jo7%*URI*qJFKIz?mB%)ibZ9I8(U0m%%!j?QAn@w^rKnVs)FtXkN_=P&4G11r zc6&CoCl?yQV95nqJ&kRbru&%L0mY;kHlFYOLoIQaJ+9J*9mu66V0DpCr97oZ9p!mT z?>V8 z(5TC((@7d@Oc)eU#>aLXqXrIYmdsk^Jy&4wlp!lK>%*8#P8?_-w!mZBCaKot00{6( z_bcEHpge@b&ybJ_@NK$Dq_RX#{JSLZ*V#9;lez?g{>IB;P_%Eu1V9-?e@Y#DdO)C5 zKQo$jeAt9EMqkw8cVR7QK}NwzU7xEGZc(4ReeDG$ zLP5nARlo%NdF#l47=!j#ly93*uW2pOBi3dP@5Y|5yIDgd`>;`MJtCsvtT?ud!3r%| z2>IN|3^1EfRwfHEnb4D|&bN?bBHf+VnzRX3@`P(Mv*L=vuY-oWDvBXFV8L?;OU3te zakWQmc(R_+P^l7QseUoF)XUywBfO^tijO0}4k|Ar=+_IY7iQv)!DQz)>RD2XRB`7v z#+`^a@R!C02dKv*Re42F0*6~qF8MuIcmARFk|b}E<4&eMP3Xvas&>yuCP4t%i|Fhp z>L}eaoAN$&aB9Z?&=1BrjOWgMA%#upoo>8O2rrzqH%BMh9*U%6Bi2k4#qX2QDjm(C z!lR<6KG~m5l0^3urdWr^pxl8jhOMc6-%8GGXLFy~uSh*+Dy>YsWnb$uoyWFMlOJ^WhJk*6rKL!XILVLvl4cZNH)h0bl9lbQ;ctq{CFZw4bb7-Q0-P z2M4YeW-;W5Mi}Us7IY+$kEDz^AZr1j$8LX<`aQVP~!cZDXGT<0w>@%KfU?!;p1@tr68M$wF%S3HD zd#X+Lctd<{suSh5`XMyAsa;VZ*zG?JJNUgCB^7*0!o9XSkxjL40gT+=4>@Fz6CK&1 zUIMr#d@VDqWWE&Zk0K9}sgQX1imAr_q=Pv$InXFXyr}9^Bi3=)T$Id(IRjp>6?jUD zf84POhobt~OO|4Yg>qX@YYjg>g?B?|2_U;$yFDmZH zkGug)H)QD?f-kWpKz7x!g;ft*5yF7sgRsvJMEBUGkj6mgA=a_HJVV zBXRQ^7PD!gU#cs8gunE==zm_RVE+jg9oX&ss8_@h93gk+s(daA;)(QAbkY8^ssd#2 z>;)9JQODW)g%?@r*1t+*nJxjrn;JhOo9) zwwuIQ;C)=?4YAIiI?18yE=??M6ixD^K3Bn1ss6x3g5!ewFr7o6A`>YT#D1F0{sx4q;Z}vdCP#y9n4!OMUD;N9 zhO1W=xl6@7J0Bp3!APC!=X=Ya!k3o z`1oKr*01B;v*e=M$CZiucxb&ApBV%v4}c@c%~`M zEMw~1p8U!VM`wsv!^3Q4BG45cx|jnMfz4qY3NJ~X4K8)8Bk;hYfLUjhra1qfd>qLr zWY7&ura4Lu;`_d{;xOTBM3*)x<}(bOBsXCYj;bX$%2afTU_YhdlE1qYo8zQc{Cabc zoXoRIQfYcWJkC1Bet?CiOr?{;{y?+=bb2>KP%XdXiq2rDmn6 z?(Rx^qU7ZVU7|vNKFO z|DvrX-+m)+po{GZd#)`Or?vCm?2AAi%yU+awIdwT!5MJpj04ReZS7iEv04$Q{E&b| zHI1_fGOH;A5sxcns#$Y&Jf0uA)mlc{1g9Qab)zPiU@Xurup7&9t4X21S08VcPV^yaU_mFdNI%`i{T_2XiJe{rK;^olt-tY-(jy5uxep z7$7l=;4Py#2E0aW1aJhZZ-Tu^f#3hOqZD3KOsr~r$_KXT-xR1+?QdUXMY~p63(B%U zK4H1T3DZi?Gf!@iKo=lDmO#r(>wT5}NEL!_vCKH1 zI()9-m^&R||G{e5ExXwk?bWNt`*@Edsm*Ta)1astqo(oWT=msRLzPn+ogS>7PyQRi z?)EzHtX`8M4G%MXLn>!FcQ7u`w!4SKcG88Mfy||&X#kECOl$=QXMo#?G{u>5At1ZI z@bk-9U&$w^dS0Va-X3(iXzxVKF(j2}rA`ka zO>L03cRH1E%m(G}j)}pOdmzBI*|}ZFN^9D9@12hTjLd2>B9mX6%>H4y{j0V1!S1Bd zlPIx-EzlrYN@HK1Xd&Y2iYo>wIGt9af`UHJ-4}vW6;m;U(5c!JtA#4e5{UM8`fMQ9 zr55*FRBpzZ(1o-;S3nRb$n!M8;6N=f7ss-O(lXz6rsQtwM8cuI6X7lL1RHilgj)0W zg%6Z>&Dl)79=WgUdY0+>K3hE7YU+zZGlFF(g&5jr5fU`Y=V-X9T0Vl8fMr)j9*v^W zeR`TB+XB>Vdm5s5kjmiaTUcUi#GE}EoD9yIuc{|E4HQ{%#>k!9BW~BUCUkKt2h2`Q zVZ`(&?Fur-$4tNDDhN?!oa7myqr=!K=2N+?b79qIl+Qx6TpM-(u6i&EIdLKIu&v)Y z#92%iH5%PadD#jkO&;LTUw@ILy)z5+Twr1vS2t|Jw&fumSEXmP6Z>n8oXdw3(m$us zs&2MXVBepkiJGg?dy$9mbQ1v$F|aQ#0SS}(l{uYJhNIEOo~QyTu}b!h*NZ|7<&qieL~svX0XmL%GsJFYK3*K4!0TRHs3PXwiUJ{sX47ypk+M6bvbvq6V=)b zLhfjVA}4{_s5_!G)Xx!i8=o$4L0hsiyUQP*;6&Q7q|0}$C}h`g(h7nW3z*g8f*;a# z{KXmo6Q5?k|H7rak`3RP2e&f%L7Cx;Hk<2JZ=4J-Lb@NpgkILs-~0YY5)dsCVxqcY z&Uf`Ag@_J;`{E1rU@04DG7O^_Le7@(aP2onGzQl4*QJPV_Nw%Am@?o&=D37wM&h1t za*~jmxKdw5nqwHT_Ob4)W1u|pu}wb>26-WJolYaaN!D!C45i)U_l)zH=0Sx*663C901+I-PRi$NT-5Wljt_J#GeE;6hYtotKgI!JvZb}x0tJ{DC{SFtpFV~TDjTr5759?f z)nSzXBgK2A&W-1ndP_pA5tZ~f_h?G0j(o`Trxj7YD;VXfLp*<31_Cf!R*PqVI;wP8 zH#52tP=JBbJMN9Ud-3B9<6SK%CtuwI!ca&3N5snj`c3)MUMwqKB}m!A`&%L?t4(_G z$IVyy-{-2+4jw|}2XgCzh=ybr5!bBFc6xVeH5i6g?C`)3NxQlu^GptP)hB>spA6}l z4$#FPz1qA79>rGDsAi^ggyD@0t_!so6ZvV{v@qAfs4GaS1{AIRTb(vGI^gMd)LPY@ zNJdGLve)q%D44RIM?C+C$9p-_(5J)-*`!jmN+*`QSE26)6szN1@%2_`5b_HYQx(@_ zoB=%B(Sy1WHkw1*`(p;8BeV=sq;veKV^BpKV^XM<_iWcW0nCnOw|Ft-1^QrIly|<% zX-<3{)ig*Yu`TKS)jTvA!H){;D!RH8RsU4qE_w1`6t&jYLNd|i)1ITLz43SNg!@5) zgpnz!gOQR&Bb?`$gR;;3!WL4Qo-m$yrT9;ofyt(3W=U4!3m%vc#i)i7W~7EsmEMf@ zc0KQ?t`!53A4vQ+g_TO9IV@VX)9y`9?*ZS;B1~NOl`G3gzYANWY~3&?l6b}F-mXkC zpK$o5B}>GanJP#DEAUx1P|DAduV2Gl{~%g-T9qZ0yXayEd5k@q2B!RsHmnVb^9A}r zI>?tqfF4M-v4m_}IK3l?sjWAso#=95w=Mv3c^@xN^JvZaW{c&c{e)e@QunT(gl5$o ztZd%m2;=l#>$5zslo(q3EcqSg!PpdtTRoUxtenB0VVpX+!9LyBsB% zlT^xWVMXTSbB%db=vMA_y~tLT%fBMwDP?D6TiKSGyDJf?3FesKFqyyKDr;<4f{Tns z`@|fw^c&-j89^MO|GB!zJ!b3kR}`-*I<#U)HL1z@nWPAQUfaB6dT#L zd$IHQ%tTdbM6O!hk>2ly3)4gGcdps#%p&pw(2jwtf(&Y>54YA5r(caD>@%V;&2R#V z@q#H=Mx=jQ#M2NW9@#%SwHkDcsntyjqx0W$m;uKcNZkva!%-o}4X^^}!VTI%+^yQ#FqPW}fE~h4%!W9hi*jszCU*JYglD5(b6y zem^nWZ;qO_0?Y?&zjgYo`<1A?Eve>?um*O4L^}q>>dER;{yu0Br;!GdB^H3CqExTO z%a4WdlnNneO6SLkbwSuMoZEsTH~kasD=R2)oVN2MaQgRmIzhS$S*pgkt>xu@=YA|H z#)v|Bv6DXtJ187}OSIu1bG@{>w6-KH!wVA+_+qb}=5Z+xIc}&3N2d&QK4q%@<%1MC zLi~r`y8>f&X8i{vZijS~!d{rjJ@GSEDD9lqAGkgR*>`1Nw~&&rFhrTD&*_3#1@a(W zIy*WSk_5OSBOZ%=-mj+(9YcLzX?IvsKaa$aW@fpXZQf5mgcg`WodRG`RWRR;72QK= zh5V$%X_h!7(9sy0!!V}99q=V64J);|5#Crmq5;XuHI;>jQBWfFw%ZLtBY?q8L{IVo z3(PcIBXJ8E!It;yAb;8R+a9nXBYT#sgJ5tHQ9!e8 zJ&|fF%GlC>nvg)Y_IYtDQ2BYBYd>uy=PDGpj6cP~3Ln2cU)3!Y4GZ(M_&##)*MFv4 z`9Ha~jQ=0iR@uYe1fNdM&{E0S28vD=pMjqK|EjVaot*L68JYec=N6xtg@xsR>k_!p zva;I}Mg6(e8;GNUa$>*B5NY5)T$X8-&1y+vjROOuSzu*rERR$~dRx2ea3ha!v=(bz z1cIj4j|&~*8v0^?V<3_=6)9z+VM>y~VTz!#VWMEV@gdbQRZcjVW6uehMrBEW zFb3%vP?iRZ5?%5{XLBzn;`le}us4<-pn{B3yFdfgbw1mi zqY>l6eNG+hJfEJHKjs}Lf%^n>e4h{9y8fh3A)Mq z5Ofjl0r(lIF4$oRUpZ(toe#xL+P|vKgi>!y(1bwZLa%l3;V%G&+`sRwrGZ&Ub$5;l zkU*&jkB~I`&I1`PfkD&aACm}3V-1fl!7YJQj17qNnba8`Yd*i^42vao!#&dkq9x=wH&Mh0YuNTfPoN{`3cUYt_%N{qlMnuM2c^sgoujyVTWj}>5b zI539y*5}WMRF?6<=&5^Rxac}26br}fus*kE=D*}^dlz}Gj#Py-H|1<4m)(4{$&%Zu zN!;1%u-!82!&$z?o)Xtx6ZY!S;EBRZNx%PKg155GE1h?QyIqFRiyoIv{RFlQ|5ui)RF_Oxk|AH>7M(U`X`+ebU%jWus|RYf`N zm^t0lMtgQdQU?@njWts9q6wFbk7lzX6a0BhQf20=TL4S{g`wt=#z| zH0C}H^=QD{`pZGQdgaxBw$Yad^9Soq7+K17_%2Uaa8-Puqf}4C_y}4I06%L%U znP+Tv8ac6)cI(h5g_X!BD0>>S75U~nf@vCrfs;G@^!he&W+UscFBr23zo%_L{N1 zpI;fKnz<_nj?bU1uc)EiO9~SX5X53__y%$li99O1Ry4DDZCO$@HJS(~-8tIRGLN0` zGLm3yH8JCCPjA}L8Gh;0vO@(Q516I|iO%2HQ;l&WqGJKcG}=?Ri2(v6g0Jtpi6diM zighPp5q<_Arj7DVyjB+BQ+LYpxIecLA2#$98?C&Rpu|)ku~}Cr zH>;3Il0)gOLydg=KP}F~qRZ;+I8$(?;oS`Z4jkNQ%l}68wUtqBmE?xJ7`ZnH>hXZk zb0Ty=kR=>>T*t&k%DVq)Ip)cd5JR_&FK!M0ns+ufQK%C%Xej?J|13IO3u~lVnf-Z&ySnE!p@mBQ6~GoEY!&(>^FB6P!8xy?tqNtl&gBo48$({?s(uI z+*g^fpwC9sN}*!zrN2`ro+NHVmmq5~>k}91Z=QP~4tb^3sU$sWlO= zJvzRfml-8cy<@sIaol*q)~okZz&N@m+g`7h+#k6SzAQfUZ;Z4V$`=)?2eIav zkuGpA8Z`^-X1mqEuf?Fp?7a2Y^St#nR}4LZEX}ZW=SEm)U>{ZMwhp!WLcAzP8jyw9 z9cb%ASl)%aFpMJzo>X1$8li#Ol8LjQsh$W&e^851XO9i;&jw}+Xp1m-U7S`ltbm;L zrpL7=?-E`+Fr};}ULR>6Ub4=s#Rbdjh4W%3OptAnGA|2D`-N09YH<^@BjW3Ys^-^Y z;d1&cxAvUZNyJS~&3WQq&-Mg(f`wBs$to<)9-Vi*?7t(QM*BPU2F$#;%U8kXm1iMX z1rZ+P75iLPJs_9&u*QE~L0g2{!Qy-3{m?}w3Q#4#s;dF)Z5-!2GW?>r1y9d1<}Uuj z5b0P8lsQ~I8=d3d9t{M2+C88x69iYA29|g8$-Rr@|CIsWoU8NsxJuIgJCj$u-K;ZN zRe5bxy$^X%xNa3Wk4ODM$eBXuE&IVKX^98Dtm74yTRk{v!DrmnZ`F}#GyY3npNspk zC!LAz+@6JPK}GbKpPr&HnAkE5k9gA)@K}iVk+A^}5HQ=P0qj3u|K_1Vfekbq~}Cu#7rNl?8~Up#uzzyS6J8Kgu8lvN!6N6Iosbg zwf~g^pF}@BQUsLpQl(cxItf*j zSR%-c^WJanoq6xv@0&R@=h;2ao_%)a>~rS)&+PuZkUue2-d&A=;FHTsJ3<(p{+sL< zo-3J6*bttsA@`;yyiU4LlV#*4;%=#2dDs3yc3B(dn}d6uX*B|FVLfH)&0E>I9lQh1yJ3 zHVLYy*3#})H^|#EA^9db`g7*>ia49Ee+$g)anq6l3{N)oLNg^&e5L8GmS>s%HDAS> zp2uGI<*UD_cHl)jZ$N&HCfuC zu-nlI$U54iDueHG7v|lv#-lZx=+>+0 z8OT4Hs$Gs0?W7VWYbr@Jdytx zId&(awWP#+=dU8c5BG2+tP@on2c^LmB32XJ*$Z{~*mjL6b&g{udE2grnEHaGVHppd zVoX<~4mQNp(hVIFcnd#$)k~%G;FYt7M$R#hf{LVQu?Nli!@y@GRUagh_GHXuK5FXb zr&=AH>V>e&rhaM|GzKFQT&#?f)j#c4)(a*4zCG^4kH@W*U#@z3&xBme`*s~`wu2bX zI4FDd-*Was27YY^m=0)2Y71!0Zl}?o#;HBglUPbUR}~CbClMlF!jsR*=OvK!AVz4T zc*TWo=?qb}Ev}|YOQ4QJ#B|EJJ9sRyx999_cvZ?H?lc8(K0P|?D-o^kJaA6*UVD~# z%u{chypt)>+>zCSxP8t_#ykLjhAseGjq;0R$$dO|TJhL^32yNDHO{E^7>5F{E|neK z+iq?C&4RHp!kN6it12*wTtgfz`pBB^Q7#-y@frCYXPni`n5nXfzckRuae@ld#QCM^ zhOOWucNbN8A_^Bj$KbGi#tL%*CI=CAzFy|bcY>sR0+`;wzZA1&|CBzWP1;>fP_4|g zx>I5#J%8`sMzHR(f5Dn8zj7Iui^jnBXEDYFD{bA8faNUZ>gB$c{IOj0a>b>!?dC*H zK|ZhY!!N5-Gh4e&*G3&L9i+J<>`n9Aap@^Xif{SPo@cw z9p%GCMV{6W&Wk%ssi(U`xf&6B)bhh&yJ(b%e;%5ax~UIa1kvpWE_C7s~H&2|s_SpE92<2D0^?dfjhX6(w{XYLU z@ezqZH4(gOeX*OvxgnehHYT`Sj?0;uEkSfG;mNJZ#i3o7&}gqS{LEHlY(q*>L&eVF zW~_m~lm(1pY-|6K5~u=JbC)YZSgw8A0wgr6<5!NoS&yw^K%Ggr`-Rq_mlmvU#d>2+ zz3+wQ)(wj#dTlE{KHr?%P!(}?+|*Tbr{sHsm9h5ywtB*?aAt#D6lpl1knm~)LKUZvl6&PkLn)0oOHU(;9gh?DQrMa^clQJL9V zn1+pwrd292XzmD6MuVf4FF#)M)S6p5jCpm7YaDqR`@?B%ta<|{8o)9DdfS4p@MsKX z$0)_$JgxZAoGW(_^ATjI9CmsVITTB${+P5-*sV$65~`JUeSG4=8lG})MuX#AZA9|p=}kI2gN_-q>s<*Hqn%BP8^_Xt6946t9Gulng%8 zBCbc(li%@Tp9)iwhbx7=nY@1<5i?GvlJTl=<5hFa)=}fR(cWip;fzLzh`z1(Tq)B`p1#oGer=Wz z^aJ}747Ppn^JIJfOAFU)J8?J1)(?1k4amtucU7e>JAh2v)|S5#x7op5(lRJWDF}al zdQcK@rH9!`lvPvrmpGwFG|Z=~IjW84v7rHViTk8pefh%nM^M6z^Mfqjg8OsxNiV>{ z4-<)>q8C_B;=g&(m3X&>@464XdP*~vl@_u_4YlKxPONW{{){S5mrWmc%$o9@l{Qd>VVQ^~v)jNB?)n+3Es2jmBR zeiNe{Sha&>i)!QdJd*23t%AH-wo(SRGNH!Q7*vy4o)5pt|&3Ep)bZczzO9mdf(ZF=q3S?l#!Bz z$ccef=UrZDmf@rnrPgVf{I84yS<4NWyHBsBv(Jslmqy@_aeQByaxACP$b9gjHh9?I8`*fmgE zk~nZ#d8nMM1XNP)UvU!siDT14KR^MAk;sYxJ?=aByZNCUfDit=otz|85+e48GLQ0h z#yGkIAkzP@_9p)?=?(HfIC~%rCil1U_8&0p>oQjjHdUB=Km9TvEBQF!@x7+=#Z}fi z-}@0kk1B)S!~z)P!ZRnIy}Z28M<(drL@+R2WaAFUFfOibx^Q@}H83KwAErfS(rZH< zN0G&1&r-^-BI9;UQl@gPE?g97S}5_>S516mGBObsla&|~<#Sd@@pipTOy_E}g7Ib| zLiyTld%+PXXOPidqx9%?Yox1d=Iuu|RQ9kf(btE41o`m+J#Fz{(*_ZTbc=1|4WEL~ z*KpsqqIq#MkEM_wE}lPU>H<6)23eP%0(`h!mbF6Lb&O92>`3=AS5IMNMc8kAFU!HS zF1&ilu?j=3g?P;W)qLyBERm;8?v!*?NfrPl9_wlvi#<1Z;_F z)^`gzl~V2I+9NQ1846^J`lFrKzDr9fhraopMPiLZLc%c$%-bJJoe;y_edi_=)WiiLD#jJ1Cj09kLoVL-Bh% zG+w#syzeUqb=m^Y$IYTIb?X@1;k+FM{CFFb|B}%-?73Xw;=T3KJgP+A((p?O^>PO0 zu~e>+<emjeWFP3Y)hDKW zV6@4BHHncd3FhUsCsZiTK0c8jG)sK2L|WInugH6O^K=Z9$(u1$%|Ci5n3cFR5k2ai_6XxQ0cvr!IGBhL(QndE*~Y>`$^EQdua)WTlTt=#%s{eia#Ml zm#(V{JFCi-*4*i}J>{d6Tt8=W<;hANd1jMEidK}awS z-B#gaGkPMNXM+GwC`d&5Rz3-#QIFSoHKk*wCT6NCJF6lQ1qAq1n$U)>G*Y=m(a5OM zTLDRLPu8;Zw6c0l-u3tH7MhodxZqO>Z*1FEv?KY!dhF3xPLXZROJS>wUiN9 z8e8gYM=Hv$