1.1 Activation Maximization.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 1.1 Activation Maximization"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Introduction\n",
    "\n",
    "**This section corresponds to Section 3.1 of the original paper.**\n",
    "\n",
    "When we are asked to describe a certain concept, say, a chair, it is impossible to come up with a precise description because there are so many types of chairs. However, there are certain characteristics unique to chairs that let us recognize chairs if we see one, such as a platform that we can sit on or legs that support the platform. With such features of chair in mind, we may be able draw a picture of a typical chair. In this section, we ask a deep neural network (DNN) to give us a prototype image $x^{*}$ that describes characteristics common to the set of objects it was trained to recognize.\n",
    "\n",
    "A deep neural network classifier mapping a set of data points or images $x$ to a set of classes $(\\omega_c)_c$ models the conditional probability distribution $p(\\omega_c | x)$. Therefore, a prototype $x^{*}$ can be found by optimizing the following objective:\n",
    "\n",
    "\\begin{equation}\n",
    "\\max_x \\log p(\\omega_c \\, | \\, x) - \\lambda \\lVert x\\rVert^2\n",
    "\\end{equation}\n",
    "\n",
    "This can be achieved by first training a deep neural network and then performing gradient ascent on images with randomly initialized pixel values. The rightmost term of the objective function ($\\lambda \\lVert x\\rVert^2$) is an $l_2$-norm regularizer that prevents prototype images from deviating largely from the origin. As we will see later, activation maximization is crude compared to other interpretation techniques in that it is unable to produce natural-looking prototype images."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Tensorflow Walkthrough"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 1. Import Dependencies\n",
    "\n",
    "We import a pre-built DNN that we will train and then perform activation maximization on. You can check out `models_1_1.py` in the models directory for more network details."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Extracting MNIST_data/train-images-idx3-ubyte.gz\n",
      "Extracting MNIST_data/train-labels-idx1-ubyte.gz\n",
      "Extracting MNIST_data/t10k-images-idx3-ubyte.gz\n",
      "Extracting MNIST_data/t10k-labels-idx1-ubyte.gz\n"
     ]
    }
   ],
   "source": [
    "import os\n",
    "\n",
    "from tensorflow.examples.tutorials.mnist import input_data\n",
    "import matplotlib.pyplot as plt\n",
    "import tensorflow as tf\n",
    "import numpy as np\n",
    "\n",
    "from models.models_1_1 import MNIST_DNN\n",
    "\n",
    "%matplotlib inline\n",
    "\n",
    "mnist = input_data.read_data_sets(\"MNIST_data/\", one_hot=True)\n",
    "\n",
    "logdir = './tf_logs/1_1_AM/'\n",
    "ckptdir = logdir + 'model'\n",
    "\n",
    "if not os.path.exists(logdir):\n",
    "    os.mkdir(logdir)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2. Building Graph\n",
    "\n",
    "In this step, we initialize a DNN classifier and attach necessary nodes for model training onto the computation graph."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "with tf.name_scope('Classifier'):\n",
    "\n",
    "    # Initialize neural network\n",
    "    DNN = MNIST_DNN('DNN')\n",
    "\n",
    "    # Setup training process\n",
    "    lmda = tf.placeholder_with_default(0.01, shape=[], name='lambda')\n",
    "    X = tf.placeholder(tf.float32, [None, 784], name='X')\n",
    "    Y = tf.placeholder(tf.float32, [None, 10], name='Y')\n",
    "\n",
    "    tf.add_to_collection('placeholders', lmda)\n",
    "\n",
    "    logits = DNN(X)\n",
    "\n",
    "    cost = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(logits=logits, labels=Y))\n",
    "\n",
    "    optimizer = tf.train.AdamOptimizer().minimize(cost, var_list=DNN.vars)\n",
    "\n",
    "    correct_prediction = tf.equal(tf.argmax(logits, 1), tf.argmax(Y, 1))\n",
    "    accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32))\n",
    "\n",
    "cost_summary = tf.summary.scalar('Cost', cost)\n",
    "accuray_summary = tf.summary.scalar('Accuracy', accuracy)\n",
    "summary = tf.summary.merge_all()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3. Building Subgraph for Generating Prototypes\n",
    "\n",
    "Before training the network, we attach a subgraph for generating prototypes onto the computation graph."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "with tf.name_scope('Prototype'):\n",
    "\n",
    "    X_mean = tf.placeholder(tf.float32, [10, 784], name='X_mean')\n",
    "    X_prototype = tf.get_variable('X_prototype', shape=[10, 784], initializer=tf.constant_initializer(0.))\n",
    "    Y_prototype = tf.one_hot(tf.cast(tf.lin_space(0., 9., 10), tf.int32), depth=10)\n",
    "\n",
    "    logits_prototype = DNN(X_prototype, reuse=True)\n",
    "\n",
    "    # Objective function definition\n",
    "    cost_prototype = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(logits=logits_prototype, labels=Y_prototype)) \\\n",
    "                     + lmda * tf.nn.l2_loss(X_prototype - X_mean)\n",
    "\n",
    "    optimizer_prototype = tf.train.AdamOptimizer().minimize(cost_prototype, var_list=[X_prototype])\n",
    "\n",
    "# Add the subgraph nodes to a collection so that they can be used after training of the network\n",
    "tf.add_to_collection('prototype', X_mean)\n",
    "tf.add_to_collection('prototype', X_prototype)\n",
    "tf.add_to_collection('prototype', Y_prototype)\n",
    "tf.add_to_collection('prototype', logits_prototype)\n",
    "tf.add_to_collection('prototype', cost_prototype)\n",
    "tf.add_to_collection('prototype', optimizer_prototype)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here's the general structure of the computation graph visualized using tensorboard.\n",
    "\n",
    "![title](./assets/1_1_Activation_Maximization/graph.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 4. Training Network\n",
    "\n",
    "This is the step where the DNN is trained to classify the 10 digits of the MNIST images. Summaries are written into the logdir and you can visualize the statistics using tensorboard by typing this command: `tensorboard --lodir=./tf_logs`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch: 0001 cost = 0.235915433 accuracy = 0.931200002\n",
      "Epoch: 0002 cost = 0.089025966 accuracy = 0.973109100\n",
      "Epoch: 0003 cost = 0.058212864 accuracy = 0.981927283\n",
      "Epoch: 0004 cost = 0.037602341 accuracy = 0.988036373\n",
      "Epoch: 0005 cost = 0.030371556 accuracy = 0.990381826\n",
      "Epoch: 0006 cost = 0.021637062 accuracy = 0.993490915\n",
      "Epoch: 0007 cost = 0.018915237 accuracy = 0.994000006\n",
      "Epoch: 0008 cost = 0.015711658 accuracy = 0.995127277\n",
      "Epoch: 0009 cost = 0.014997141 accuracy = 0.995127277\n",
      "Epoch: 0010 cost = 0.015114903 accuracy = 0.994981823\n",
      "Epoch: 0011 cost = 0.012736740 accuracy = 0.995909095\n",
      "Epoch: 0012 cost = 0.011209372 accuracy = 0.996545458\n",
      "Epoch: 0013 cost = 0.009346475 accuracy = 0.996727276\n",
      "Epoch: 0014 cost = 0.008826889 accuracy = 0.996781821\n",
      "Epoch: 0015 cost = 0.009214011 accuracy = 0.996618185\n",
      "Accuracy: 0.9813\n"
     ]
    }
   ],
   "source": [
    "sess = tf.InteractiveSession()\n",
    "sess.run(tf.global_variables_initializer())\n",
    "\n",
    "saver = tf.train.Saver()\n",
    "file_writer = tf.summary.FileWriter(logdir, tf.get_default_graph())\n",
    "\n",
    "# Hyper parameters\n",
    "training_epochs = 15\n",
    "batch_size = 100\n",
    "\n",
    "for epoch in range(training_epochs):\n",
    "    total_batch = int(mnist.train.num_examples / batch_size)\n",
    "    avg_cost = 0\n",
    "    avg_acc = 0\n",
    "    \n",
    "    for i in range(total_batch):\n",
    "        batch_xs, batch_ys = mnist.train.next_batch(batch_size)\n",
    "        _, c, a, summary_str = sess.run([optimizer, cost, accuracy, summary], feed_dict={X: batch_xs, Y: batch_ys})\n",
    "        avg_cost += c / total_batch\n",
    "        avg_acc += a / total_batch\n",
    "        \n",
    "        file_writer.add_summary(summary_str, epoch * total_batch + i)\n",
    "\n",
    "    print('Epoch:', '%04d' % (epoch + 1), 'cost =', '{:.9f}'.format(avg_cost), 'accuracy =', '{:.9f}'.format(avg_acc))\n",
    "    \n",
    "    saver.save(sess, ckptdir)\n",
    "\n",
    "print('Accuracy:', sess.run(accuracy, feed_dict={X: mnist.test.images, Y: mnist.test.labels}))\n",
    "\n",
    "sess.close()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 5. Restoring Subgraph\n",
    "\n",
    "Here we first rebuild the DNN graph from metagraph, restore the DNN parameters from the checkpoint and then gather the necessary nodes for prototype generation using the `tf.get_collection()` function (recall prototype subgraph nodes were added onto the 'prototype' collection at step 3)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Restoring parameters from ./tf_logs/1_1_AM/model\n"
     ]
    }
   ],
   "source": [
    "tf.reset_default_graph()\n",
    "\n",
    "sess = tf.InteractiveSession()\n",
    "\n",
    "new_saver = tf.train.import_meta_graph(ckptdir + '.meta')\n",
    "new_saver.restore(sess, tf.train.latest_checkpoint(logdir))\n",
    "\n",
    "# Get necessary placeholders\n",
    "lmda = tf.get_collection('placeholders')[0]\n",
    "\n",
    "# Get prototype nodes\n",
    "prototype = tf.get_collection('prototype')\n",
    "X_mean = prototype[0]\n",
    "X_prototype = prototype[1]\n",
    "cost_prototype = prototype[4]\n",
    "optimizer_prototype = prototype[5]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 6. Generating Prototype Images\n",
    "\n",
    "Before performing gradient ascent, we calculate the image means $\\overline{x}$ that will be used to regularize the prototype images. Then, we generate prototype images that maximize $\\log p(\\omega_c | x) - \\lambda \\lVert x - \\overline{x}\\rVert^2$. I used 0.1 for lambda (lmda), but fine tuning may produce better prototype images."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch: 00000 Cost = 25.786497116\n",
      "Epoch: 00500 Cost = 2.708985567\n",
      "Epoch: 01000 Cost = 0.190258697\n",
      "Epoch: 01500 Cost = 0.012256578\n",
      "Epoch: 02000 Cost = 0.005008637\n",
      "Epoch: 02500 Cost = 0.004844339\n",
      "Epoch: 03000 Cost = 0.004842392\n",
      "Epoch: 03500 Cost = 0.004842398\n",
      "Epoch: 04000 Cost = 0.004842440\n",
      "Epoch: 04500 Cost = 0.004842412\n"
     ]
    }
   ],
   "source": [
    "images = mnist.train.images\n",
    "labels = mnist.train.labels\n",
    "\n",
    "img_means = []\n",
    "for i in range(10):\n",
    "    img_means.append(np.mean(images[np.argmax(labels, axis=1) == i], axis=0))\n",
    "\n",
    "for epoch in range(5000):\n",
    "    _, c = sess.run([optimizer_prototype, cost_prototype], feed_dict={lmda: 0.1, X_mean: img_means})\n",
    "    \n",
    "    if epoch % 500 == 0:\n",
    "        print('Epoch: {:05d} Cost = {:.9f}'.format(epoch, c))\n",
    "    \n",
    "X_prototypes = sess.run(X_prototype)\n",
    "\n",
    "sess.close()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 7. Displaying Images\n",
    "\n",
    "As I mentioned in the introduction, the resulting prototype images are rather blurry."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAssAAAQwCAYAAAATjRM7AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzs3Xu0XOV55/nfo6P7/XJAd5AMAhuIb0Mgs5ykSWKy5E4C\nzvKMA0x6cMY2TpaVOMHtDnbSmCHpXrYzttu9ho6j2AQnHZu4nZvckUO8iLMYZ8VEcuKAJQwIgdGR\nhKSj+1060jN/VB1SHN6nTu2qXVVvVX0/a9XSqaf22Zeqo/2871v7fba5uwAAAAC82pRu7wAAAACQ\nKxrLAAAAQIDGMgAAABCgsQwAAAAEaCwDAAAAARrLAAAAQIDGcmbM7LNm9h/LXhYAgNyRA5Ejo85y\n55jZC5KWShqTdEHSdkl/KGmju19scd03Sfrv7r6qwO+YpI9Jek819DlJ9zh/FACAkmWYA39M0r2S\n3izpsLuvaWUf0L8YWe68n3H3eZIuV6Wh+uuSPt+lfblL0tslvUHS6yX9jKT3dWlfAAD9L6cceFLS\ng5I+1KXto0fQWO4Sdz/q7psk/ZykO83sOkkys4fM7LfHlzOz/2Bme81sj5m9x8zczK6sXdbM5kj6\nmqQVZnai+ljRwG7cKemT7j7i7rslfVLSu0o+VAAAXiGHHOju/+jufyRpZ1sOEn2DxnKXufs/ShqR\n9CMTXzOz9ZLulvRWSVdKuilYx0lJb5O0x93nVh97zOyHzexInc1fK+lfap7/SzUGAEDbdTkHAg2h\nsZyHPZIWJ+LvlPQH7r7N3U9Juq/ISt39m+6+sM4icyUdrXl+VNLc6rXMAAB0QrdyINAQGst5WCnp\nUCK+QtKumue7Esu04oSk+TXP50s6wQQ/AEAHdSsHAg2hsdxlZvaDqpwovpl4ea+k2pm9q+usqpkG\n7jZVJveNe0M1BgBA23U5BwINobHcJWY238x+WtLDqpS7eTKx2Jcl/YKZvc7MZkuqV09yn6QlZrag\nwG78oaS7zWxldTLEByU9VOD3AQAoLIccaGZTzGympGmVpzbTzKYXOAwMCBrLnfdVMzuuytdJvyHp\nU5J+IbWgu39N0n+V9A1JOyR9q/rS2cSy35P0JUk7zeyIma0wsx8xsxN19uX3JH1V0pOSvivpr6ox\nAADaIacc+KOSTkvaLOmy6s9/09RRoa9xU5IeYmavU6VRO8Pdx7q9PwAAdAo5EN3CyHLmzOxnzWyG\nmS2S9HFJX+UkAQAYBORA5IDGcv7eJ2m/pOdUuT3oL3V3dwAA6BhyILqOyzAAAACAACPLAAAAQGBq\nK79cvRXlZyQNSfqcu39skuUZxkZPcPeO38Vw/fr1Pjo6WneZb3/724+4+/oO7RKAOsiB6FfkwFdq\nurFsZkOSHpB0syr3dd9iZpvcfXtZOwcMktHRUW3durXuMmY23KHdAVBHszlwypTWv9A1K96Oye2S\ny2h/ih5bveXLOuZoG7m9p2W5ePFiV7abcw5sZWT5Bkk73H2nJJnZw5JulURjGWhSt05SAAojBwIl\nyzUHttJYXqlX3qd9RNKNExcys7sk3dXCdoCB0a8jFUAfIgcCJcs1B7Z0zXIj3H2jpI0S12sB9bh7\ntr1qAM0hBwKNyTkHttJY3i1pdc3zVdUYgCbl2qsG8Cpdy4FlnifKuna46D4VvQ44Wr4T129Hy7f7\nfN3MseW4jSJyzYGtNJa3SFpnZmtVOUHcJumOUvYKGFC59qoBvAo5EChZrjmw6cayu4+Z2QZJj6hS\nNudBd99W2p4BA8bds+1VA3glciBQrpxzYEvXLLv7ZkmbS9oXYODl2qsG8GrkQKBcuebAtk/wA9C4\nXHvVAAC0W645kMYykImcZwIDANBOOedAGstARnLtVQMYLGVVpRgaGiq0/rIqVdRTtBJHP1S9iOSW\nc3Lbn3E0loGM5NqrBgCg3XLNgTSWgUzkPBMYAIB2yjkH0lgGMpJrrxoAgHbLNQfSWAYykmuvGgCA\ndss1B07p9g4AqBifCVzv0QgzW29mT5vZDjO7J/H6p83sO9XHM2Z2pPSDAQCggLJyYDswstwlRWcU\nF41P9lpK0ZnAZc1azrUn2Q2tvhdmNiTpAUk3SxqRtMXMNrn79ppt/FrN8r8s6U0tbRRAW0SVJCIX\nLlwovI0pU9o7Zla0gRPlrWYaStGxlVUNI1p/tHzRY6iXD7pZQaOdcm0P0FgGMlJCz/kGSTvcfack\nmdnDkm6VtD1Y/nZJH211owAAtIprlgHU1WBB9mEz21rzfKO7b6x5vlLSrprnI5JuTK3IzC6XtFbS\n3zaxuwAAlIabkgBoSANfQY26+/Ulbe42SV9x9+Lf3QIAUDIuwwAwqRJ61bslra55vqoaS7lN0vtb\n3SAAAGVgZBnApEroVW+RtM7M1qrSSL5N0h0TFzKz10paJOkfWt0gAABlYGS5T0QzUKNZy9OmTUvG\nZ86cmYzPnz8/GV+0aFGhuCQtWLAgGZ86Nf2xnz17Nhk/evRoMn7o0KFC8ePHjyfjZ86cScbHxsaS\ncSnf/1CtKON6LXcfM7MNkh6RNCTpQXffZmb3S9rq7puqi94m6WHvxzcSyFjqv1y7KxdJ5VW9KFpJ\nIhLloVmzZiXjUT6bN29euI3p06cn49F5Nso5586dKxQ/depUMn7y5MlC66mXA6NjKPq3lFNVDa5Z\nBtCQMtqu7r5Z0uYJsXsnPL+v5Q0BAFCiXMdvaCwDGcm1Vw0AQLvlmgNpLAOZcPdse9UAALRTzjmQ\n210DGcn1Vp8AALRbGTnQzNab2dNmtsPM7km8fpmZfcPM/tnMnjCzfzvZOhlZBjKSa68aAIB2azUH\nmtmQpAck3azKTbm2mNkmd6+9i+1vSvqyu/+umV2jyhyfNfXWS2MZyETOM4EBAGinknLgDZJ2uPtO\nSTKzhyXdKqm2seySxkuPLZC0Z7KV0lgORGV2yip3s3z58mR87dq1yfhVV11VaHlJWrp0aTIela07\nceJEMr5r165kfMeOHcn4U089lYw///zzyfi+ffsK7Y9Uv6ROL2NkGehv7SzVVa88XFkl34ouXzSX\nzpkzJxlftmxZMr5ixYpw21Ep1ugYopJvx44dS8ajMqlHjhxJxi9cSN8sNWog1nuvo3VFv5NTibh6\nGvj7GjazrTXPN7r7xprnKyXVNlpGJN04YR33SfobM/tlSXMkvXWyjbbUWDazFyQdl3RB0liJt+EF\nBhIjy0DvIAcC5WogB46W8P/sdkkPufsnzex/lfRHZnadu4cbL2Nk+cfcfbSE9QADj5FloOeQA4GS\nlJADd0taXfN8VTVW692S1le39w9mNlPSsKT90UqphgFkYvx6LaphAAAGTUk5cIukdWa21symq3K3\n2k0TlnlR0k9Ikpm9TtJMSQfqrbTVxrKrct3Ht83srtQCZnaXmW2dcI0JgITxOpPRA0BWyIFAiVrN\nge4+JmmDpEckPaVK1YttZna/md1SXeyDkt5rZv8i6UuS3uWTrLzVyzB+2N13m9mlkr5uZt9z98cm\n7PhGSRslyczI9kAdjB4DPYUcCJSojBzo7ptVKQdXG7u35uftkt5SZJ0tNZbdfXf13/1m9ueqlOx4\nrP5v5aXoTN3Zs2cn40uWLEnGL7vssmT8ta99bTJ+3XXXJeOve93rCq1fkhYvXpyMT5s2LRmPKkwc\nOJD+dmLVqlXJ+MKFC5PxGTNmJONRh65exYuTJ08WWlcvYPQY6C3tzoHRObBe1YtIWeeWqDET7VO0\n3Wj5KD9FuTfKc1KclyNR1YvoGKI8VFS0/mYajr1S9SIl5xzY9GUYZjbHzOaN/yzpJyV9t6wdAwYR\n1ywDvYEcCJQv1xzYyjXLSyV9s3rNxz9K+it3/+tydgsYTGVcszzZrT6ry7zTzLab2TYz+2KpBwEM\nBnIgULJc5+00fRlG9e4obyhxX4CBVsbdixq51aeZrZP0YUlvcffD1estARRADgTKlfNdbLmDH5CR\nEnrOjdzq872SHnD3w9VthrUlAQDolFyvWaaxDGSkgV51Gbf6vEqSzOzvJQ1Juo+vjwEA3cbIcpdF\nM0SjGbkzZ85MxqNKD1FVimuvvTYZf8Mb0t/eXX311cn4ypUrk/F58+Yl4/VEf4xRBZBLL01/S190\nBu+ZM2eS8VOnTiXjJ06cSMbrrateBY1e0ECvuoxbfU6VtE7STarc3egxM/sBdz/S4noBNKFoBYNo\n+U40NKJtlzUiGOXk6dOnJ+P1cmBUKSN6ny5cuJCM79u3Lxk/f/58Mn7u3LlC8Wi79d7T6LVeroYh\nMbIMYBIlXa/VyK0+RyQ97u7nJT1vZs+o0nje0urGAQBoRs7XLHO7ayAjJcwEbuRWn3+hyqiyzGxY\nlcsydpZ3FAAAFNd31TAAlK/VXrW7j5nZ+K0+hyQ9OH6rT0lb3X1T9bWfNLPtki5I+pC7H2xx1wEA\naEmuI8s0loFMlNVzbuBWny7p7uoDAICu6/bocT00loGM5NqrBgCg3XLNgQPTWI5miEb3oY9m2EZV\nKa666qpkPKqGsW7dumT8kksuScajP6Bolq4U37c+msEbvRfz589PxqOKIVFlkMOHDyfjBw4cSMb3\n7t2bjEvSoUOHkvEBqIYBYMBF+aDe+aPdVRKi9Q8NDSXj0TFE8ahKxoIFC8J9iio5RbkxylFRzowq\nNkXxotUwmvnMej2H5Lr/A9NYBnKX80xgAADaKeccSGMZyEiuvWoAANot1xxIYxnISK69agAA2i3X\nHEhjGchIrr1qAADaLdccSGMZyETO12sBANBOOefAgWksRzNpZ82alYwPDw8n41Glh6uvvjoZv+KK\nK5LxJUuWJONRNYf9+/cn4yMjI8m4FFfKOH78eDI+Y8aMZDyqABId29y5cwutJ3pPd+zYkYxL0osv\nvpiMnzlzJvydXpBrrxpA50XVEKIGRb3qCdFr0TmnaCWGostH243iUX5atGhRuI3ly5cn41FuPH36\ndDJ+7NixZDzKpdF6yqzWVPRzKxKPqnN0Qq45cGAay0AvyLVXDQBAu+WaA2ksA5nI+e5FAAC0U845\nkMYykJFce9UAALRbrjmQxjKQkVx71QAAtFuuOTA96w1Ax43PBK73aISZrTezp81sh5ndk3j9XWZ2\nwMy+U328p/SDAQCggLJyYDv01chyvdm406ZNS8bnz5+fjK9YsSIZv/LKK5PxtWvXJuNRVY2o9/TS\nSy8l408//XQy/uyzzybjkrRr165k/OTJk8l4VBnkNa95TbiNlKgySPReRzOWo/dOkmbOnJmMR7OW\ne0WrvWozG5L0gKSbJY1I2mJmm9x9+4RF/8TdN7S0MQClKKsiRZmK7lPRc1fU8IniUX6KcrUkLVu2\nLBk/cOBAMh5Vtzh48GCh5c+fP5+MF32PmskH7a5i0m49O7JsZg+a2X4z+25NbLGZfd3Mnq3+G9du\nAdCwEnrVN0ja4e473f2cpIcl3drWnQb6GDkQ6JxOfLtaXeadZrbdzLaZ2RcnW2cjl2E8JGn9hNg9\nkh5193WSHq0+B9Ci8dnA0aMBKyXVfqUwUo1N9A4ze8LMvmJmq8vYd6BPPSRyINARrebAmm9X3ybp\nGkm3m9k1E5ZZJ+nDkt7i7tdK+tXJ1jtpY9ndH5N0aEL4VklfqP78BUlvn2w9AOpr8HqtYTPbWvO4\nq4lNfVXSGnd/vaSv61//LwOYgBwIdEZJ1yw38u3qeyU94O6Hq9tN3/WtRrPXLC91973Vn1+StDRa\nsJrMm0nowMBpoOc86u7X13l9t6TakeJV1VjtNmovwPucpE8U2UcA5ECgHUq4Zjn17eqNE5a5SpLM\n7O8lDUm6z93/ut5KW57g5+5uZuHRuftGSRurO5bnldtABsZ71S3aImmdma1VpZF8m6Q7ahcws+U1\nif4WSU+1ulFgUJEDgXI0mAOHzWxrzfON1f9jRUyVtE7STaoMKD1mZj/g7kfq/UIz9o0nXDNbLmnS\nIexOqDerM7qv/JIlS5Lxyy+/PBmPKkNEFR2iKhzRvem3bduWjD/xxBPJ+HPPPZeMS9L+/emP5dy5\nc8n43Llzk/HofY1mGkezkxcsWJCMR1UyFi2K58xEn2eva7VX7e5jZrZB0iOq9JgfdPdtZna/pK3u\nvknSr5jZLZLGVPl6+V2t7TUwcLqaAztRMaCsyg1R/igaX7hwYTK+Zs2acJ+iHHL69OlkPKpGdfjw\n4WT81KlTyXhZ1U2a+Zyj34niU6bkVUG4E9+uqjLa/Li7n5f0vJk9o0rjeUu00mbfpU2S7qz+fKek\nv2xyPQBqlDET2N03u/tV7n6Fu/+nauzeakNZ7v5hd7/W3d/g7j/m7t9r4yEB/YgcCLRBCTnw5W9X\nzWy6Kt+ubpqwzF+oMqosMxtW5bKMnfVW2kjpuC9J+gdJV5vZiJm9W9LHJN1sZs9Kemv1OYAWlVAN\nA0CJyIFA57SaA919TNL4t6tPSfry+Ler1W9UVX3toJltl/QNSR+aMJfnVSa9DMPdbw9e+olJ9xpA\nw0q6ZhlAiciBQGeUlQPdfbOkzRNi99b87JLurj4a0ld38AN6HaPHAIBBlWsOpLEMZISRZQDAoMo1\nB/ZVY7neTNOo0kNUxSKaYbtyZepmaNKcOXOS8RMnTiTjO3bsSMajqhfbt29PxnfvnjjJ818Vnak7\nNjaWjB89ejQZP3IkXWUl2u68efOS8enTpyfjs2fPTsaluBpG9DeQa2+1FtclA6hV9HxQLweWdW4p\nq0pGJKogFeXqqHKVFFd+Gh0dTcajfBrluvPnzyfjZVWYqPd5Fq0mEn0OOTVOc86BfdVYBnpdTicu\nAAA6KdccSGMZyEiuvWoAANot1xxIYxnIBNUwAACDKuccSGMZyEiuvWoAANot1xxIYxnISK69agAA\n2i3XHNhXjeWhoaHwtei+8lF1i8suuywZX7JkSTIefcDR7NqnnnqqUPzFF19Mxo8fP56MS3F1i2im\n7syZM5PxCxcuFIpH243eoyhebyZwvc+6l+XaqwbQedE5sJnzX1nnlmifovN4tN1oPbNmzUrG161b\nl4wvWrQoGZfivBnFoyoZUdWLohUmotxb5nm/6Pudm1xzYF81loFelvP1WgAAtFPOOZDGMpCRXHvV\nAAC0W645sJzK2QBKcfHixbqPRpjZejN72sx2mNk9dZZ7h5m5mV1f2gEAANCkMnJgOzCyDGSijLsX\nmdmQpAck3SxpRNIWM9vk7tsnLDdP0gckPd7SBgEAKEHOd/BjZBnISAm96hsk7XD3ne5+TtLDkm5N\nLPdbkj4u6Ux5ew8AQPMYWe6AqVPjw4mqWETVMJYtW5aMz549OxmPZtHu3LkzGX/mmWeS8T179iTj\nUdWLc+fOJeNS8Vmx0fsXVcmIZi1H64mqZ5w9e7ZQvJ810KseNrOtNc83uvvGmucrJe2qeT4i6cba\nFZjZmyWtdve/MrMPtbK/AIpJ/R+PzslFKxiU2Zgouu2o0kMUnzZtWjK+fPnyZHzNmjWF1iNJhw8f\nTsZfeOGFZPzEiRPJeDMVm1KK5uRmKlgU3dfcRnJz259xfdVYBnpZgzOBR9296WuMzWyKpE9Jelez\n6wAAoGxUwwDQkBJ61bslra55vqoaGzdP0nWS/q460rBM0iYzu8Xda0esAQDoKEaWAUyqhF71Fknr\nzGytKo3k2yTdMf6iux+VNDz+3Mz+TtK/p6EMAOg2RpYBTKrVXrW7j5nZBkmPSBqS9KC7bzOz+yVt\ndfdNJewmAAClY2QZQF1lXa/l7pslbZ4QuzdY9qaWNwgAQIu4ZhlAQ3LtVQMA0G655sBJG8tm9qCk\nn5a0392vq8buk/ReSQeqi32kOprVVTNmzAhfGx4eTsaXLl2ajC9YsCAZj3o9Bw4cSMaff/75ZHz3\n7t3J+LFjx5LxqERcvV5YVLJn+vTpyfi8efOS8ajs3uLFi5PxqNRcdAxRWbyojI/Uv2Xlcu1VA4Oq\n7BxYpBxYtGw3zxNFy5xFeWj+/PnJ+Nq1a5PxKA/VywVR/t27d28yXjTPFm3YRe9FM6Xjyipn10x5\nunbKNQc2clOShyStT8Q/7e5vrD663lAGet343YvqPQB03EMiBwJtl3MOnHRk2d0fM7M17d8VALn2\nqoFBRQ4EOifXHNjK7a43mNkTZvagmS0qbY+AAZZrrxrAq5ADgZLlmgObbSz/rqQrJL1R0l5Jn4wW\nNLO7zGzrhFv0AphgfCZwvQeALJADgZLlnAObaiy7+z53v+DuFyX9vqQb6iy70d2vb+UWvcCgyLVX\nDeBfkQOB9igjB5rZejN72sx2mNk9dZZ7h5m5mU36f7Op0nFmttzdx6eT/qyk7zaznrLVq4axaFH6\nW7IoHlWMiGbe7t+/v1A8qnpx/vz5ZDz6I6k3kzU6hqjSx4oVK5Lxyy+/PBlfvnx5Mh59DkePHk3G\nR0dHk/FDhw4l45J05syZ8LVexugxkL9O5cDovN9MJYTotWgbRXNOVOlh1qxZyfiyZcuS8VWrViXj\nQ0NDyXi9PPHcc88l41GlpaLn3+iYiy7fzEBItK6ifxsXLlwovO12ajUHmtmQpAck3SxpRNIWM9vk\n7tsnLDdP0gckPd7IehspHfclSTdJGjazEUkflXSTmb1Rkkt6QdL7Gj4SACFGj4G8kAOBzikhB94g\naYe775QkM3tY0q2Stk9Y7rckfVzShxpZaSPVMG5PhD/fyMoBNC7nuxcBg4ocCHRGgzlweML1/xvd\nfWPN85WSdtU8H5F0Y+0KzOzNkla7+1+ZWTmNZQCdw8gyAGBQNZADR1u5/t/Mpkj6lKR3Ffk9GstA\nRhhZBgAMqhJy4G5Jq2uer6rGxs2TdJ2kv6tex71M0iYzu8Xdw4o1rdRZBlCisu5eNNlMYDP7RTN7\n0sy+Y2bfNLNrSj8YAAAKKCkHbpG0zszWmtl0SbdJ2lSzjaPuPuzua9x9jaRvSarbUJb6bGQ5qv4g\nSbNnz07GZ86cmYxHM2+jWbQnT55Mxk+dOpWMRzNQoxmr06ZNKxSXpIULFybja9asScavuSbdZrrq\nqquS8UsuuSQZj3qGBw4cSMZHRkaS8X379iXjknT69OlkvNcvY+jQTOAvuvtnq8vfospXUqnb+QIo\nWeocVbSyQZF1j4tyS70KGinRvka5KMpDUTWlJUuWJOPnzp1LxqNqSpK0Z8+eZLxo/o2OuawKI5Ey\nv2nMrepFpNVjdvcxM9sg6RFJQ5IedPdtZna/pK3uvqn+GtL6qrEM9LpOzAR299q6hXNUmdEPAEBX\nlTHg5e6bJW2eELs3WPamRtZJYxnIREnVMCadCSxJZvZ+SXdLmi7px1vdKAAArci5IhTXLAMZaeB6\nreHxW+dWH3c1uZ0H3P0KSb8u6TfLPAYAAJqR611sGVkGMtJAr3qysjmTzQSe6GFJv9vY3gEA0D6M\nLAOYVLtnAkuSma2refpTkp4t7QAAAGgSI8sdUPQ+7VLxi8mjKhlRJY65c+cm4wsWLCi0P9F2582b\nl4xL0urVq5PxqOrFtddem4yvXbs2GZ8xY0Yyvnfv3mT8hRdeSMZ37dqVjNeb5Xz27NnwtV5VxvVa\nDc4E3mBmb5V0XtJhSXe2uOsAGpSqoFA0DxWtwlBP9DtRPo1y0Zw5c5LxRYsWJeNRNaUolx45ciQZ\nj/KHJB09ejQZj96/qVPTTaKokkS7G2/1KpVE2y5a3SQnOV+z3FeNZaDXlXGimGwmsLt/oOWNAABQ\nMhrLACbV63WiAQBoVq45kMYykImcv4ICAKCdcs6BNJaBjOTaqwYAoN1yzYE0loGM5NqrBgCg3XLN\ngX3VWK73Jp88eTIZP3XqVDIezX6NqltcdtllyfjVV1+djEczVo8fP56Mz5w5Mxm/9NJLk3EprmJx\nxRVXJOOrVq1KxqdNm5aM79u3LxnfuXNnMv7cc88l49Fs5mj2sySdP38+fK1Xdbs0DoC8FK1sUO/8\nUVbVi6gKUlSZacmSJcl4VBEqsn///mT8pZdeCn/n3LlzyXhU9SI65qg9UJYyq5uUua5OyzkH9lVj\nGeh1ufaqAQBot1xzII1lICO59qoBAGi3XHMgjWUgEznPBAYAoJ1yzoE0loGM5NqrBgCg3XLNgTSW\ngYzk2qsGAKDdcs2BfdVYPnv2bPja6OhoMn7gwIFkPKqSsWLFimT82muvTcbnz5+fjEdVMk6cOJGM\nT58+PRlfvHhxMi5JS5cuTcajWctRj27Pnj3J+NNPP52Mb9u2LRl/9tlnk/FoNnP0GUj5/odqRc4z\ngQG0T9H/91HFg6LVM6S4GkZUBWn27NnJeFTdYuHChcl4lNNOnz6djEfVl+pVTRobGyu07eiYo/WU\nVSWjzHwW/S31QpWMnHNg+n9JDTNbbWbfMLPtZrbNzD5QjS82s6+b2bPVfxe1f3eB/nbx4sW6DwCd\nRQ4EOifXHDhpY1nSmKQPuvs1kn5I0vvN7BpJ90h61N3XSXq0+hxAC8Z71tEDQMeRA4EOyTUHTtpY\ndve97v5P1Z+PS3pK0kpJt0r6QnWxL0h6e7t2EhgE4zOBW+1Vm9l6M3vazHaY2asSuJndXR0le8LM\nHjWzy0s/GKBPkAOBzigrB7ZDIyPLLzOzNZLeJOlxSUvdfW/1pZckJS+QNbO7zGyrmW1tYT+BgdBq\nr9rMhiQ9IOltkq6RdHt1FKzWP0u63t1fL+krkj5R8mEAfYkcCLRXz44sjzOzuZL+VNKvuvux2te8\ncgTJo3D3je5+vbtf39KeAgOghF71DZJ2uPtOdz8n6WFVRsBe5u7fcPfx2ZPfkpS+zzmAl5EDgfbL\ndWS5oWoYZjZNlZPEH7v7n1XD+8xsubvvNbPlktI3bu+getUTRkZGkvGoQsOyZcuS8aj6xMqVK5Px\n1atXJ+PRvp45cyYZj/5IopnM9X7n0KFDyfj3v//9ZHz79u3J+JNPPpmMf+9730vGo8/g6NGjyXg0\nA1nKawZvmRo4ruEJI1Qb3X1jzfOVknbVPB+RdGOd9b1b0tcK7SQwYMrMgUUqVhStYFBmNYypU9PN\ng6gaRhSfOXNmMh5VkojywfHjx5PxqIKUFL9P0bFF70XR97UT+amZzzolt1ya2/6Mm7SxbJVP5POS\nnnL3T9W06RnkAAAgAElEQVS8tEnSnZI+Vv33L9uyh8CAaPDuRaNljVCZ2c9Lul7SvyljfUA/IgcC\nndHrd/B7i6R/J+lJM/tONfYRVU4QXzazd0v6vqR3tmcXgcFRQq96t6TarzNWVWOvYGZvlfQbkv6N\nu8cFygGQA4EO6dmRZXf/pqRovP8nyt0dYLCV0KveImmdma1VpZF8m6Q7ahcwszdJ+j1J692965dP\nATkjBwKdk+vIcqFqGADaZ7JZwI30uN19TNIGSY+oUuLqy+6+zczuN7Nbqov9jqS5kv6HmX3HzDa1\n65gAAGhEGTlQak/51L663TXQ68roVbv7ZkmbJ8Turfn5rS1vBACAkrWaA2vKp96sygT3LWa2yd1r\nKxWMl089ZWa/pEr51J+rt96+aixHlSQkac+ePcl4VNEhund8JJrZG1XDWLhwYTIe9ZyiYzt48GC4\nT7t27UrGn3766WR827ZtyfhTTz2VjD///PPJ+L59+5LxaNby+fPnk/Fcv45pp1yv1wJQjm79H48q\nPUTxadOmFYpHFSai8/jJkyeT8SiXRtUf6uX9qKJStK5uxZvJddHvlFUlo1tK+P/xcvlUSTKz8fKp\nLzeW3f0bNct/S9LPT7bSvmosA70s55nAAAC0U4M5sCvlU2ksAxlhZBkAMKgayIFdKZ9KYxnICCPL\nAIBBVUIObEv5VBrLQEYYWQYADKoScmBbyqfSWAYywTXLAIBBVUYOdPcxMxsvnzok6cHx8qmStrr7\nJr2yfKokvejut4QrVZ81lqOZr5J06NChZHz79u3JeHQf+t27XzWaX3c9r3nNa5Lx4eHhZDyaUXz0\n6NFC+yNJO3fuTMajKhYvvvhiMr5/f7rjdezYsWT87Nn0NxrRLGdGU/8V7wWAcdH5IIo3UwmhaIWG\nqHrRqVOnkvHDhw8n41EVi6GhoWQ8Eu2PFFeLiipxRG2IKHdF71H0+XRzMCT6nHPLOWXsTzvKp/ZV\nYxnodYwsAwAGVa45kMYykIkidygCAKCf5JwDaSwDGcm1Vw0AQLvlmgNpLAMZybVXDQBAu+WaA2ks\nA5mgGgYAYFDlnANpLAMZybVXDQBAu+WaA/uqsVzvTY7KmY2OjibjUam2HTt2JOOPPfZYMj579uxk\nfObMmcl4VDYnKmlz+vTpZFyKS/lEvxO9R1FpnqJlczC5MnrVZrZe0mdUqTH5OXf/2ITXf1TSf5H0\nekm3uftXWt4ogI6JyoDVKx0XnZej83uUc86dO5eMR2XaonKoU6ZMScYjzRxzlNOiHBi9R9F7UTTX\nNVPar6heKREXyXVkudhfK4C2Gp8NHD0mY2ZDkh6Q9DZJ10i63cyumbDYi5LeJemLJe8+AABNazUH\ntktfjSwDvayk67VukLTD3XdKkpk9LOlWSS/fNcfdX6i+lmcXHgAwcLhmGUBDGug5D5vZ1prnG919\nY83zlZJ21TwfkXRjSbsHAEDb5Hq5CI1lICMN9KpH3f36TuwLAACdxMgygLpKuiZrt6TVNc9XVWMA\nAGSr29cl1zNpY9nMVkv6Q0lLJbkqX/t+xszuk/ReSQeqi37E3Te3a0fbJerFRLNoo3hUPQMoooRe\n9RZJ68xsrSqN5Nsk3dHqSoFBVXYOTFUr6EQDIdrGhQsXkvGoqkJ0jioaj6phFK16Ua/CRFTFoizR\ne1q0IkUzFSw6UVmjG3p5ZHlM0gfd/Z/MbJ6kb5vZ16uvfdrd/5/27R4wWFpNmu4+ZmYbJD2iSum4\nB919m5ndL2mru28ysx+U9OeSFkn6GTP7v9392lb3HehT5ECgQ3p2ZNnd90raW/35uJk9pcokIgAl\nKmsmcHV0a/OE2L01P29R5fIMAJMgBwKdkXM1jEJ1ls1sjaQ3SXq8GtpgZk+Y2YNmtij4nbvMbOuE\nGfwAEnKtMQmAHAi0W645sOHGspnNlfSnkn7V3Y9J+l1JV0h6oyq97k+mfs/dN7r79czgByZ38eLF\nug8A3UEOBNov1xzYUDUMM5umyknij939zyTJ3ffVvP77kv5nW/YQGCCMHgP5IQcCnZFrDmykGoZJ\n+rykp9z9UzXx5dVruSTpZyV9tz27CAyGnK/XAgZV2TmwnY2BZs4fRatMFN3G0NBQMh69D0XX30zF\niKJVKYquv+h2y/ybyLWx2Yicc2AjI8tvkfTvJD1pZt+pxj4i6XYze6MqpXRekPS+tuwhMEB6+UQH\n9ClyINAhuebARqphfFNSqpvUczWVgdzl2qsGBhU5EOicXHMgd/ADMtHt2b4AAHRLzjmQxjKQkVx7\n1QAAtFuuOZDGMpCRXHvVAAC0W645kMYykImcZwIDGCzdarRMmZK+/UMz+1O0WkXR9RTdp6LbbWZd\nuTY2G5FzDix0Bz8A7ZXr3YsAAGi3MnKgma03s6fNbIeZ3ZN4fYaZ/Un19cerd+asi8YykJFc714E\nAEC7tZoDzWxI0gOS3ibpGlVKPF4zYbF3Szrs7ldK+rSkj0+2XhrLQEZy7VUDANBuJeTAGyTtcPed\n7n5O0sOSbp2wzK2SvlD9+SuSfsImuUaGxjKQifHrtXLsVQMA0E5l5EBJKyXtqnk+Uo0ll3H3MUlH\nJS2pt1Iay0BGcu1VAwDQbg3kwGEz21rzuKsT+9Xpahijkr5f/Xm4+nyQDNox9+rxXt6l7T5y8eLF\n4UmWmWlmW2ueb3T3jTXPU73qGyes4xW9ajMb71X34mcF9JLRixcvkgM7JINJ0b36GeecA0fdfX2d\n13dLWl3zfFU1llpmxMymSlog6WC9jXa0sezul4z/bGZb3f36Tm6/2wbtmAfteFs1yQkAQI8jBw7W\nMQ/a8baqpBy4RdI6M1urSqP4Nkl3TFhmk6Q7Jf2DpP9N0t/6JD0rLsMA+kuRXrUa7VUDAJC76jXI\nGyQ9IukpSV92921mdr+Z3VJd7POSlpjZDkl3S3rVRPiJuCkJ0F/a0qsGAKAXuPtmSZsnxO6t+fmM\npP+9yDq72VjeOPkifWfQjnnQjrfrqtcgj/eqhyQ9ON6rlrTV3Tep0qv+o2qv+pAqDWoAnTWI58dB\nO+ZBO96+ZQwoAQAAAGlcswwAAAAEaCwDAAAAgY43lie7FW+/MLMHzWy/mX23JrbYzL5uZs9W/13U\nzX0sk5mtNrNvmNl2M9tmZh+oxvv2mAGgKHJgf+YDcmB/62hjucFb8faLhyRNrBl4j6RH3X2dpEfV\nQLmSHjIm6YPufo2kH5L0/upn28/HDAANIwf2dT4gB/axTo8sN3Ir3r7g7o+pUmmgVu1thr8g6e0d\n3ak2cve97v5P1Z+Pq1LfcKX6+JgBoCByYJ/mA3Jgf+t0Yzl1K96VHd6Hblrq7nurP78kaWk3d6Zd\nzGyNpDdJelwDcswA0ABy4ADkA3Jg/2GCX5dUbwLRd3X7zGyupD+V9Kvufqz2tX49ZgBAMf2aD8iB\n/anTjeVGbsXbz/aZ2XJJqv67v8v7Uyozm6bKSeKP3f3PquG+PmYAKIAc2Mf5gBzYvzrdWH75Vrxm\nNl2VO4dt6vA+dNP4bYZV/fcvu7gvpTIzU+XOcE+5+6dqXurbYwaAgsiBfZoPyIH9reN38DOzfyvp\nv+hfb8X7nzq6Ax1iZl+SdJOkYUn7JH1U0l9I+rKkyyR9X9I73X3iBIieZGY/LOn/k/SkpIvV8EdU\nuWarL48ZAIoiB/ZnPiAH9jdudw0AAAAEmOAHAAAABGgsAwAAAAEaywAAAECAxjIAAAAQoLEMAAAA\nBGgsAwAAAAEaywAAAECAxjIAAAAQoLEMAAAABGgsAwAAAAEaywAAAECAxjIAAAAQoLGcGTP7rJn9\nx7KXBQAgd+RA5Mjcvdv7MDDM7AVJSyWNSbogabukP5S00d0vtrjumyT9d3dfVeB3PiTpTkmXSxqV\n9N/c/Xda2Q8AAFIyzIG/JumXJQ1LOiHpTyR9yN3HWtkX9B9GljvvZ9x9nioN1I9J+nVJn+/Svpik\n/1PSIknrJW0ws9u6tC8AgP6XUw7cJOnN7j5f0nWS3iDpV7q0L8gYjeUucfej7r5J0s9JutPMrpMk\nM3vIzH57fDkz+w9mttfM9pjZe8zMzezK2mXNbI6kr0laYWYnqo8VDezDJ9z9n9x9zN2flvSXkt7S\njuMFAGBcJjnwOXc/Mr4pSRclXVnyoaIP0FjuMnf/R0kjkn5k4mtmtl7S3ZLeqsp/4JuCdZyU9DZJ\ne9x9bvWxx8x+2MyOpH4nsS2r7sO2pg4EAICCup0DzewOMzumyqWIb5D0e60cD/oTjeU87JG0OBF/\np6Q/cPdt7n5K0n1FVuru33T3hQ0ufp8qfw9/UGQbAAC0qGs50N2/WL0M4ypJn5W0r8g2MBhoLOdh\npaRDifgKSbtqnu9KLNMyM9ugyrXLP+XuZ9uxDQAAAl3NgZLk7s+q8s3qf2vXNtC7pnZ7Bwadmf2g\nKieKbyZe3iupdmbv6jqraqqsiZn9X5LukfSj7j7SzDoAAGhGt3PgBFMlXVHCetBnGFnuEjObb2Y/\nLelhVcrdPJlY7MuSfsHMXmdmsyXVqye5T9ISM1tQYB/+D0n/WdLN7r6zwO4DANC0THLge8zs0urP\n10j6sKRHGz4IDAway533VTM7rsrXSb8h6VOSfiG1oLt/TdJ/lfQNSTskfav60qsulXD370n6kqSd\nZnbEzFaY2Y+Y2Yk6+/LbkpZI2lIzg/izzR4YAACTyCkHvkXSk2Z2UtLm6uMjzR0W+hk3JekhZvY6\nSd+VNIOi6QCAQUIORLcwspw5M/tZM5thZoskfVzSVzlJAAAGATkQOaCxnL/3Sdov6TlVbg/6S93d\nHQAAOoYciK7jMgwAAAAgwMgyAAAAEGipznL1VpSfkTQk6XPu/rFJlmcYGz3B3a3T21y/fr2Pjo7W\nXebb3/72I+6+vkO7BKAOciD6FTnwlZpuLJvZkKQHJN2syn3dt5jZJnffPsnvNbtJoCO6dWnS6Oio\ntmzZUneZKVOmDHdodwDU0WwOnDLl1V/oRnmRyyTzULTd0uuf28WLF7uy3ZxzYCuXYdwgaYe773T3\nc6oUFr+1nN0CBpO7130AyAY5EChZrjmwlcswVuqV92kfkXTjxIXM7C5Jd7WwHWAguHvXevQACiMH\nAiXKOQe2dM1yI9x9o6SNEtdrAZNh9BjoL+RAoHG55sBWGsu7Ja2ueb6qGkMDunmNWtFt5/jH26/X\nvufaqwbwKqXlwBzPsUVFx9Duc3W99Zf1vpaVG8t8L/r1Ovdcc2Ar1yxvkbTOzNaa2XRJt0naVM5u\nAYMp1+u1ALwKORAoWa45sOmRZXcfM7MNkh5RpWzOg+6+rbQ9AwZMztdrAXglciBQrpxzYEvXLLv7\nZkmbS9oXYOAxegz0DnIgUK5cc2DbJ/gBaFyuvWoAANot1xxIYxnIRLevyQIAoFtyzoE0lgtK3X1J\nimemlhUvc6bxhQsXCm0jihftAXbi2Hpdrr1qAOVIndfKqqpQbz1l3YWurJxWdLudaESVtY2yKlU0\nkwN7vUpGrjmQxjKQkV45oQEAULZccyCNZSATOc8EBgCgnXLOgTSWgYzk2qsGAKDdcs2BrdyUBEDJ\nLl68WPcBAEC/KiMHmtl6M3vazHaY2T2J1z9tZt+pPp4xsyOTrZORZSATOc8EBgCgncrIgWY2JOkB\nSTdLGpG0xcw2ufv2mu38Ws3yvyzpTZOtt+ON5dQb0YmqB0Vn6kZVL6ZNm5aMz5w5MxmfNWtWMj53\n7txCy0frnzo1/gijP7qzZ88m4ydOnEjGjx8/noyfPHkyGT9z5kwyPjY2loxH1TkGEaPHQH8r0hgo\nq5JEM9soq7pFlEuHhoYKrafMdkLRCk9RvGjljrIqjNRbV6RXqmSUkANvkLTD3XdKkpk9LOlWSduD\n5W+X9NHJVsrIMpCR3E5cAAB0SgM5cNjMttY83+juG2uer5S0q+b5iKQbUysys8slrZX0t5NtlMYy\nkImcZwIDANBODebAUXe/vqRN3ibpK+4+6dfbTPADMjJ+zVb0aMRkkxuqy7zTzLab2TYz+2KpBwEA\nQBNKyIG7Ja2ueb6qGku5TdKXGlkpI8tARlodWW5kcoOZrZP0YUlvcffDZnZpSxsFAKAEJXy7ukXS\nOjNbq0oj+TZJd0xcyMxeK2mRpH9oZKWMLAMZKaFX/fLkBnc/J2l8ckOt90p6wN0PV7e5v9SDAACg\nCa3mQHcfk7RB0iOSnpL0ZXffZmb3m9ktNYveJulhbzCxdnxkuYwZrUVnlErxjNyomkRUlWLBggXJ\n+KWXpgfnVqxYkYxfdtllyfiqVasKrX/OnDnJuBT30I4ePZqM79q1KxnfuXNnMv79738/Gd+zZ08y\nfvDgwWT81KlTyXhUPUMqbyJcJyqxNKrB67XKmNxwlSSZ2d9LGpJ0n7v/dXN7DaBdilYwqHc+i6pP\nFK38FOXGKBdFOTOKR+uJKkLVE+WQqMLTsWPHkvEoZ0YVpKKcFlWiaqZSVPS5RaLc0oM5sJH1bJa0\neULs3gnP7yuyTi7DADLSQCegjMkNUyWtk3STKtdzPWZmP+DukxZmBwCgXXKtCEVjGchICb3qRiY3\njEh63N3PS3rezJ5RpfG8pdWNAwDQrFwrQnHNMpCJya7VarDH/fLkBjObrsp1WZsmLPMXqowqy8yG\nVbksI32tDQAAHVBSDmwLRpaBjLTaq3b3MTMbn9wwJOnB8ckNkra6+6bqaz9pZtslXZD0IXdPX1AO\nAECH5DqyTGMZyEgZPefJJjdUZ//eXX0AAJAFrlkGUBd38AMADKqcc2BLjWUze0HScVW+yh0r8RaE\nk223UFyKy6xEZXCWLFmSjK9evToZX7NmTTJ+1VVXJePr1q0rtJ6lS5cm43Pnzk3G64nK47z44ovJ\n+JNPPpmMz5s3LxmPyvFF/wmi8jj1yuYU7X0WLcHULbntD4BYu3NgdN6KzqVReThJmjFjRjIelWpb\nuHBhMr5s2bJk/PLLL0/G165dW2j5qNzq4sWLk/F6JdROnz6djO/bty8Zf+6555LxZ599Nhl//vnn\nk/GRkZFkfHR0NBmPStnVK59aVE4l4urJNQeWMbL8Y+6e/gsAUEiuvWoAIXIgUJJccyCXYQAZybVX\nDQBAu+WaA1stHeeS/sbMvm1md5WxQ8CgGr9eq94DQFbIgUBJcs6BrY4s/7C77zazSyV93cy+5+6P\n1S5QPYFwEgEakGuvGkASORAoUa45sKWRZXffXf13v6Q/l3RDYpmN7n59pyb/Ab0s1141gFcjBwLl\nyjUHNj2ybGZzJE1x9+PVn39S0v2l7VkT6s2KnTZtWjIezQQeHh5OxqOZwJdddlkyHs3sjdYfzViO\nZsWeOXMmGZfiqhTRNi655JJkPKrQEVXVOHLkSKH4yZMnk/F6x1b0P03UW03NEO5Wz7bbdygC0Lgy\nc2BUqSA670fn9unTp4fbmD17djIeVX5atWpVMh5Vcrr66quT8agaxsqVKwvtT1T5qd4xR06cOFFo\nn6J8HbUforZIVOHp3LlzhZaXiufAXqiGUVYONLP1kj6jyo25PufuH0ss805J96lyKdW/uPsd9dbZ\nymUYSyX9efUDmCrpi+7+1y2sDxh4jB4DPYMcCJSs1RxoZkOSHpB0s6QRSVvMbJO7b69ZZp2kD0t6\ni7sfrl5GVVfTjWV33ynpDc3+PoBXY2QZ6A3kQKB8JeTAGyTtqP7/lJk9LOlWSdtrlnmvpAfc/XB1\nm/snWyml44BM5Hz3IgAA2qnBHDhsZltrnm909401z1dK2lXzfETSjRPWcZUkmdnfq3Kpxn2TfStE\nYxnICCPLAIBB1UAOHC1hsuxUSesk3SRplaTHzOwH3D09sUo0loGsMLIMABhUJeTA3ZJW1zxfVY3V\nGpH0uLufl/S8mT2jSuN5S7TSvmos15vtGc0enjlzZjI+a9asQstHvaGo0sNLL72UjB86dCgZj2bX\nDg0NJeNSXPUimsEbVQyJZhsvXLgwGV+0aFEyvmDBgmT84MGDyXi96iaR6G+gV0Zse2U/AZSnaKWC\n6NwYncOlOHdFVTKiPBHlg6hyR1QFKVp+9+6J7ZqK6Nii45Kk+fPnF/qdaBtR7oqqZEQVPfbvT18a\nG7VP6uXAKFcUjedWJaOEHLhF0jozW6tKI/k2SRMrXfyFpNsl/YGZDatyWcbOeivtq8Yy0Mu4ZhkA\nMKjKyIHuPmZmGyQ9osr1yA+6+zYzu1/SVnffVH3tJ81su6QLkj7k7ulRuyoay0BGOlFj0szeJel3\n9K9fTf2/7v65ljcMAEALysiB7r5Z0uYJsXtrfnZJd1cfDaGxDGSkEzUmq/7E3Te0tDEAAEqU67er\nNJaBTJR096JGakwCAJCVnO9iW3wGFYC2uXjxYt1HA1I1JlP3cH2HmT1hZl8xs9WJ1wEA6KgScmBb\nZDGy3IkKBtE2omoS0T3Zo+oW0SzXM2fOJOPRjN/ojyGaORzNopXimcDLli1LxpcvX15oG9Es6ige\nzaKOPoNmZun2yozfSAN/85MVZG/EVyV9yd3Pmtn7JH1B0o8XXAeAkpR13qq3fLSNc+fOJeNRFYuR\nkZFk/NixY8l4dN6PRDkwqgwRVWWSpFWrVhWKRxVAIlH1jLKqm9TLB9H7FG27j3JgV2TRWAbQ8Ezg\nyQqyT1pjcsKs389J+kSR/QQAoGw5V4TiMgwgI+PXbEWPBrxcY9LMpqtSY3JT7QJmVvsVwi2Snirt\nAAAAaFIJObAtGFkGMtKhGpO/Yma3SBqTdEjSu1rbawAAWpfryDKNZSAjHaox+WFJH255QwAAlIhr\nlgHUlfP1WgAAtFPOOTCLxnJZPYl664leO3v2bDIezeyNRMvPmDGj0P5EM5OjahjRbFwprnoRzbxd\ntGhRMj5v3rxC2y5a3aKbFSxy68Xmtj8Auic6BxatmiTFlZmOHj1aaF2HDx9OxutVZioiWs/cuXOT\n8dWr48qXixcvTsaj82yUG6P3+/Tp08n4qVOnkvHoM4iqbzWjV6peRHLNgVk0lgFU5NqrBgCg3XLN\ngTSWgUx0e7YvAADdknMOpLEMZCTXXjUAAO2Waw6kzjKQkVxrTAIA0G5l5EAzW29mT5vZDjO7J/H6\nu8zsgJl9p/p4z2TrZGQZyETOM4EBAGinMnKgmQ1JekDSzZJGJG0xs03uvn3Con/i7hsaXW9fNZbr\n9Tqimb3R7NRoxm80y7XoPeKbmc2cMn/+/MKvRe9TVMUiikeimb3nz58vtHw9vT7jN8LoMTB4yqoI\nFJ1j620jyjknT55MxqN8EFWSiHLj7Nmzk/GogsUll1ySjF955ZXJuCRdddVVyfiKFSuS8ej9i96L\n/fv3J+MHDx5Mxk+cOJGMR1WwBjEflHDMN0ja4e47JcnMHpZ0q6SJjeVCJr0Mw8weNLP9Zvbdmthi\nM/u6mT1b/TddcwxAIRcvXqz7ANBZ5ECgcxrIgcNmtrXmcdeEVayUtKvm+Ug1NtE7zOwJM/uKmcX1\nB6sauWb5IUnrJ8TukfSou6+T9Gj1OYAWcc0ykJ2HRA4EOqKBHDjq7tfXPDY2sZmvSlrj7q+X9HVJ\nX5jsFyZtLLv7Y5IOTQjfWrPyL0h6e7H9BDDR+PVajCwD+SAHAp1RUg7cLal2pHhVNVa7nYPuPn5H\nus9J+l8mW2mz1TCWuvve6s8vSVoaLWhmd40Plze5LWBgMLIM9ARyINAGJeTALZLWmdlaM5su6TZJ\nm2oXMLPlNU9vkfTUZCtteYKfu7uZhUdQHSLfWN1Bsj0QoBoG0HvIgUA5ysiB7j5mZhskPSJpSNKD\n7r7NzO6XtNXdN0n6FTO7RdKYKt8avWuy9TbbWN5nZsvdfW+1hZ6eEtqiojOEm6mGEc1yPXv2bDJe\nVsWIaF+j+KxZs5LxetUwot+ZM2dOMj5jxoxC+xTNHI4qjESVRMqcCVy0+khuemU/gQFXag4smtOi\neL3KQkWrYUT7VLS6RZSj1qxZk4y/7nWvS8avu+66QstL0rJly5Lx6Dz7wgsvJONR1YuXXnopGR8d\nHU3Go/ZG0SpYUv9WhCojB7r7ZkmbJ8Turfn5w5I+XGSdzV6GsUnSndWf75T0l02uB0ANLsMAegI5\nEGiDXHPgpCPLZvYlSTepUq5jRNJHJX1M0pfN7N2Svi/pne3cSWAQcBkGkB9yINAZOefASRvL7n57\n8NJPlLwvwMAro+dsZuslfUaV67U+5+4fC5Z7h6SvSPpBd2fyEZBADgQ6J9dvUPvqDn5Ar+vUrT7N\nbJ6kD0h6vKUNAgBQklxHlpu9ZhlAG5RwvdbLt/p093OSxm/1OdFvSfq4pPRsTAAAOqxnr1nuJc1U\nw4hElR6mTk2/ZVE1jClT0v2RaPmoIkU0o3jp0rC8p5YvX56ML168OBmfPn16Mh5Vqzh+/HgyfvTo\n0WS86Ezgej3M6H0ts4JKpzV4vdbwhHqtGyfcwSh1q88ba1dgZm+WtNrd/8rMPtTKPgPovOg8UWaF\nhCjXRTkqqjxx7bXXJuPXX399Mv7mN785Gb/66quT8UsvvTQZl+L348CBA8l4lPej3HX69OlkPMqZ\nZY6aFs2BvaCnr1kG0DkNnOhG3T2dZRpgZlMkfUoN1JUEAKCTcm3s01gGMlJCr3qyW33Ok3SdpL+r\njrosk7TJzG5hkh8AoJsYWQZQV0nXZL18q09VGsm3SbqjZhtHJQ2PPzezv5P072koAwC6qdvXJddD\nYxnISIdu9QkAQHYYWQYwqTJ61ZPd6nNC/KaWNwgAQAkYWW5C0Zm99d7k6LUoHlWrmDZtWjIezRCe\nOXNmMj537txkfOHChcl4VNli7dq1yXi914aHh5Px6JhPnDiRjI+Ojibjhw8fTsZPnTqVjF+4cCEZ\nj2b71lN0XanPv1s925xnAgMoRyqvFa3iE6mXA6NzY3Tej+JRZaaVK1cm46997WuT8euuuy4ZX7du\nXTIe5a1671GUc86cSVfMjCqARBWkonw9e/bsZDyqOHX27NlkvN6x5dqobEXOOZA6y0BGcq0xCQBA\nux6c3ukAACAASURBVJWRA81svZk9bWY7zOyeOsu9w8zczCatMJX1yDIwaHLtVQMA0G653sWWkWUg\nI4wsAwAGVa53saWxDGRi/Hqteg8AAPpRgzlw2My21jzumrCa1F1sX3FBfe1dbBvdNy7DADLC6DEA\nYFDlehfbLBrLRRsI0QzRetUTopm9RatVRDOBo1mxS5YsScYvueSSZHzp0qXJeDTTOIpL0qWXXpqM\nz5kzJxmPql4cP348GT948GCh5cfGxpLx6POMPrN6vxPplVHZXtlPAM3pVoc4OmdG8agyRFQRKqr0\nEJ3Tjh49moy/8MILyfhLL72UjNd7P6OcU7QKVlSN6oorrkjGi1aEOnfuXDJ++vTpZFwqr4JKbnK9\ni20WjWUAed+9CACAdsr5LrY0loGMMLIMABhUud7FlsYykBFGlgEAgyrXu9jSWAYykfPdiwAAaKec\ncyCNZSAjjCwDAAZVrjmQxjKQkVx71QAAtFuuOXDSxrKZPSjppyXtd/frqrH7JL1X0oHqYh+pXiPS\nlKjUSdHSKPVKx82ePTsZj0q7RWXXohIyUQm3FStWFIpH61m2bFkyHpWsk+ISP1Fpt6iUz7Fjxwqt\n58KFC4X2Z8aMGcl4vRI4RUsC5fofcKJce9XAoOpmDiyqXg4sWqIzOhedOZO+4dn+/fuT8WeeeSYZ\nj/LKrFmzkvHo2Oodc1QaNsr7UenWefPmJeOrV69OxqMyd1G51ZMnTybjUUk5Kc5pvV5SLtcc2Mgd\n/B6StD4R/7S7v7H6aPokAaCCO/gBWXpI5ECg7XLOgZM2lt39MUmHOrAvwMAbrzMZPRphZuvN7Gkz\n22Fm9yRe/0Uze9LMvmNm3zSza0o/EKBPkAOBzikjB7ZDIyPLkQ1m9oSZPWhmi6KFzOyu8Xt4t7At\nYCC02qs2syFJD0h6m6RrJN2eaAx/0d1/wN3fKOkTqtz6E0Ax5ECgZD07shz4XUlXSHqjpL2SPhkt\n6O4b3f36Vu7lDQyCyXrUDfaqb5C0w913uvs5SQ9LunXCdmovFpwjKc+LxIB8kQOBkpWUA9uiqWoY\n7r5v/Gcz+31J/7O0PQIGWAM95+EJI1Qb3X1jzfOVknbVPB+RdOPElZjZ+yXdLWm6pB9vbm+BwUQO\nBNoj17k5TTWWzWy5u++tPv1ZSd8tb5desZ1C8WjmqyQtWpT+luzyyy9Pxl/zmtck42vWrEnGL7vs\nsmQ8qmJRNB7t/9Sp8UcYzVo+ffp0Mh5VmCha3WLu3LnJeHQM0Xqi/ZSkU6dOha+lRMeQ28zbBvZn\ntIwRKnd/QNIDZnaHpN+UdGer6wQGRa45sF5liOi1qBpGtI3ovLxv375kPKqa9OyzzybjRateTJ8+\nPRmX4mpRUR6PthFVr1qwYEEyvnjx4kLLHzhwIBk/ceJEMi4VrwjVK3Ld/0ZKx31J0k2qjGiNSPqo\npJvM7I2qfH37gqT3tXEfgYFQ0t2LdkuqrWe0qhqLPKzKV8oAEsiBQGf09B383P32RPjzbdgXYOCV\n0KveImmdma1VpZF8m6Q7ahcws3XuPj6s81OS0kM8AMiBQAf17MgygM5ptVft7mNmtkHSI5KGJD3o\n7tvM7H5JW919kyqz+N8q6bykw+ISDABABnp2ZBlA55TRq67eIGHzhNi9NT9/oOWNAABQsjJyoJmt\nl/QZVQaMPufuH5vw+i9Ker+kC5JOSLrL3bfXWyeNZSATOV+vBQBAO5WRA2vuNXCzKtWgtpjZpgmN\n4S+6+2ery9+iyr0GUnfpfFnWjeWi97KfM2dOuK6oykQ0K3bdunXJeFQ949JLL03GowoQUXzWrFnJ\neHTM9f6wzp49m4xH95uPenTRPi1ZsiQZP3/+fDI+Y8aMZPzQofTNsY4dO5aMS/HfRq/PEO6V/QTQ\nfkWrXtSrjhS9VrRKRnSOis7XUTWMaD1F8369ahhRXo6qUkRVlqK8UvRziPY1qghVr7pJJHr/Irnl\nnBL25+V7DUiSmY3fa+DlxnIz9xrIurEMDBpGlgEAgyrXew3QWAYy0e07FAEA0C0N5sCu3GuAxjKQ\nEUaWAQCDKtd7DRS/IAZA24z3rKMHAAD9qoQc+PK9Bsxsuir3GthUu4CZ1U5Ka+heA4wsA5mgGgYA\nYFCVkQPbda+BjjeWUz2DaPZm1IuIZprOmzcv3G40K3blypXJeHQv+OHh4WQ8ml07f/78ZLzeDN6U\nqIJFVHlCKj6zN6pWEVW9uHDhQqH1RNVKZs+enYwfOHAgGZfiY6hXQaMXMHoM9Lci1QqiZaPqCdG5\nV4rPs1GViagSQ9SYiXJRFI/yR7T+aD+j96Lea9F7EcWjNkdZ5+vova73t1LWtotWz2i3XO81wMgy\nkBFGlgEAgyrXHEhjGcgII8sAgEGVaw6ksQxkgmuWAQCDKuccSGMZyEiuvWoAANot1xxIYxnISK69\nagAA2i3XHNjxxnIZM4GjWbEzZ84M1xVVYoiqVSxatCgZj6pqRFUy5s6dm4wXrYZx9uzZZPz48ePh\n70SVIU6cOJGMR7OTZ82alYwvXrw43HZK9HlGlT6OHj0ariv6DxUdQzTbOFq+G6ilDPS/Iv/Hi1bD\niKo5SHHFpig3RtuIzr3ReTyqynTmzJlC6y9arUmSrrjiimT8yiuvTMaj/B69F1FePn36dDIeVQaJ\n3rtm8lP09xXlwJxyTs45kJFlICO59qoBAGi3XHMgd/ADMlLGHfzMbL2ZPW1mO8zsnsTrd5vZdjN7\nwsweNbPLSz8QAAAKyvUutjSWgUyMzwSu95iMmQ1JekDS2yRdI+l2M7tmwmL/LOl6d3+9pK9I+kTJ\nhwIAQCFl5MB2obEMZKSEXvUNkna4+053PyfpYUm3TtjGN9x9/ELCb0laVepBAADQhFxHlrlmGchI\nCT3nlZJ21TwfkXRjneXfLelrrW4UAIBW5XrNck82lpu5l3k0qzTqqUQzb6MqGVE1jKjqRbQ/UQWL\nAwcOJOP79+9PxiXpyJEjyXg0g7doJYloZm+0nmj5aObwyZMnk3Epnj08NjaWjEf/AXOaedtgz3nY\nzLbWPN/o7hub2Z6Z/byk6yX9m2Z+H0B7ReeDqCJUvSpLUeWnKHdF1TOiqlPRuTc6v0fxSFRZasWK\nFeHvRNUwovi8efOS8agy08GDBwvFo5wc5broPZWK565o+R7MgV0x6WUYZrbazL5RnRC0zcw+UI0v\nNrOvm9mz1X/TrUgADWvgeq1Rd7++5jGxobxb0uqa56uqsVcws7dK+g1Jt7h7uvcEgBwIdFAZ1yy3\nY5J7I9csj0n6oLtfI+mHJL2/OmHoHkmPuvs6SY9WnwNoQQnXa22RtM7M1prZdEm3SdpUu4CZvUnS\n76nSUI6/mgAgkQOBjmk1B7ZrkvukjWV33+vu/1T9+bikp1S5LvJWSV+oLvYFSW+f9CgAhMqYCezu\nY5I2SHpElf+rX3b3bWZ2v5ndUl3sdyTNlfQ/zOw7ZrYpWB0w8MiBQGeUVA2jLZPcC12zbGZrJL1J\n0uOSlrr73upLL0laGvzOXZLuKrIdYFCVcb2Wu2+WtHlC7N6an9/a8kaAAUQOBNqrhHk7bZnk3nBj\n2czmSvpTSb/q7sdqJ9m5u///7N17lFxXeef9789tXSxL6OK2ZNmSL4BIUCAxeR2TeSGJEyBLziQ2\nrCRE9prEZBHMZOGE22TGkIzj8VxeYCYQ5n29mCjgGDKA8RAIIjFxWEAWISEgQQhgCQchy7ZkWXJb\nd1mW3NLz/lHVTrnZu7pO16mq3VW/z1q1VPXU6XOpbp1nn1N7P1tS8gibB7GpuY4ye26bFaLUkcBm\no8450Kz3OsiBExFxRR3bqjLIvaPGsqR5NE4SH46ITzTD+yStjoi9klYDfev7mKu2kJuDHuDgwYPJ\neK6axL59+5LxFStWJOO50cm5ShJHjx6ttN1HHnkkGW9XDSNXWSNXlSK3r7nqI7k/6tyI31xFj717\n9ybjuRHFAMeOHUvGc8c2V5Q6EthslPU6B+bOsbl4Lgfm4u3WlauSsWZN+pvp1atXJ+OLFi1KxnO5\nMbevueWrVvOAfL7OVQ2ZmJhIxnfv3p2Mf+9730vGH3rooWQ8l69zbZd2v8+q5kpuqWE/qw5y/6lO\nBrl3Ug1DwAeA7RHx7pa3NgM3NJ/fAHxqpnWZWV7JsxeZjSrnQLP+qCkH9mSQeyd3ll8C/CrwLUnf\naMbeDrwDuFvSa4EHgVd3skEzy5srV/9mI8Q50KxPus2BETEpaWqQ+xhwx9Qgd2BrRGzmmYPcAR6K\niGuyK6WDxnJEfAnIzQLysgrHYGYz8N1js7I4B5r1Tx05sBeD3OfkDH5mw6jk2YvMzMx6qeQc6May\nWUF8Z9nMzEZVqTmw6MZybvRuruJBrgoDwMMPP5yM50bePvnkk8l4rnJDbtRt7hhy1Rxyo3FzlSTa\nHXNuhG3VK7eqo5lz2z18+HAynqvakVse4Pjx48n45ORkMp77D5j7/QxKqVfVZtZ/ufPBqVOnKsUh\nf848ceJEMp6rjrRs2bJk/JJL0jMGr1y5MhlfvHhxMr5gwYJk/Oyz082Vdo2rXNWpXHWLb3/728n4\nP/3TPyXj27dvT8YffPDBZDyXr3NtmnbHlvvbyOW0qssPSqk5sOjGstkomRoJbGZmNmpKzoFuLJsV\npNSrajMzs14rNQe6sWxWkFKvqs3MzHqt1BzoxrJZQUq9qjYzM+u1UnOgG8tmhSi5v5aZmVkvlZwD\n+95YTl01VB29mRs5mquq0O5nDh48mIzv3LkzGV+yZEkyvnDhwmS8akWPkyfTU5TnqnO0G/2cqwyR\nk/u8q8ZzVTJy+5P7LNod22xGD88FpV5Vm1nv5P7f585nufNfrsoS5Ktb5KpM5Kog5eT2NRcfHx9P\nxufPn5+M5/LBvn37svuUq0rxne98Jxn/7ne/m4zv2rUrGd+/Pz1Tcq4tksvv/chbpVW9yCk1B/rO\nsllB5npj38zMbLZKzYFuLJsVouTZi8zMzHqp5ByY/l7GzAbizJkzbR+dkLRB0v2Sdki6OfH+T0r6\nuqRJSb9U+0GYmZnNQh05sBfcWDYryNSVde4xE0ljwO3A1cB64DpJ66ct9hDwGuAjNe++mZnZrHWb\nA3vFjWWzQkyNBO7yqvpKYEdE7IyIU8BdwLXTtrMrIr4JlNk5zMzMRk5NObAn3666sWxWkA6uqscl\nbW153DhtFRcBD7e83t2MmZmZFa3Ub1f7PsCvSvmSqv1T2pVKy5XUOX78eDI+MTGRjOf2P1eWJ6eu\nMi51fi1R17rqWs9s+iflfg+lDhqYroNjnoiIK/qxL2Y2WFVLyrUrt3n48OFKP3PgwIFkfMeOHcn4\nV7/61WR8+fLlyfiiRYuS8arlVnPHBfnSsLn8niv5lms/5Eq65sqn5lQtzwr5z6lqKd7S1NAv+elv\nVwEkTX27um1qgYjY1Xyv4425GoZZQWo4oe0B1ra8XtOMmZmZFa2DHDguaWvL600Rsanlderb1Rd3\nu19uLJsVoqbZi7YA6yRdRqORvBG4vtuVmpmZ9VKHOXAg3666z7JZQbrtrxURk8BNwL3AduDuiLhP\n0m2SrgGQ9GOSdgO/DPyRpPt6eEhmZmYdqaEaRk++XfWdZbOC1FFHMiLuAe6ZFrul5fkWGicQMzOz\nYpT67arvLJsVYqYr6rkyQMPMzKyqOnJgr75dnfHOsqS1wIeAVUDQ6Ez9Xkm3Aq8DHmsu+vbmHa3a\n1FnZoOrP1DVTTNWqF7n9rKt6xmxU/ewGWeljkJ9THQY5Q5GZfb9B5sCc3HkiVzEC8tWiTpw4kYzn\nKkns3bs3Gc/l615Ximp3zsy9l4vnqljklu91u6LdZzGsN09K/Xa1k24Yk8BbI+LrkpYAX5P02eZ7\n74mI/1Flg2aWN6wnQLM5zDnQrE9KzYEzNpYjYi+wt/n8qKTteJIDs9rVVA3DzGrkHGjWHyXnwErf\nj0i6FHgR8JVm6CZJ35R0h6Rk5XFJN07NNtbVnpqNAPdZNiuXc6BZb5WaAztuLEtaDPwZ8KaIOAK8\nD3gOcDmNq+4/SP1cRGyKiCs865jZzM6cOdP2YWaD4Rxo1nul5sCOSsdJmkfjJPHhiPgEQETsa3n/\nj4G/6Mkemo2Ikr+CMhtlzoFmvVdyDuykGoaADwDbI+LdLfHVzb5cAK8Cvl33ztV5y73qfOl1VbGo\nqs5qDr2urNHryhNzvbLFbLirhVlZ5lIObLd8XeeWuhoz/aiy1OvzadX1V60M0k7VNs1cUer+d3Jn\n+SXArwLfkvSNZuztwHWSLqdRSmcX8Pqe7KHZCCn1qtpshDkHmvVJqTmwk2oYXwJSlzB9qSdpNkpK\nvao2G1XOgWb9U2oO9HTXZoUoub+WmZlZL5WcA91YNitIqVfVZmZmvVZqDnRj2awgpV5Vm5mZ9Vqp\nOXCoGsuzuSKZK5Uh+nFspV7RjYpBF103M5tJr89RdVZ5qCsv97qy1GzUVU2kpJxTcg6sr46JmXWt\n1ILsZmZmvVZHDpS0QdL9knZIujnx/gJJH2u+/5XmzJxtubFsVpA6pvrsxYnCzMys17rNgZLGgNuB\nq4H1NEo8rp+22GuBgxHxXOA9wDtnWq8by2aFmBoJ3M1Vda9OFGZmZr1URw4ErgR2RMTOiDgF3AVc\nO22Za4EPNp9/HHiZZuhT48ayWUFquLPckxOFmZlZr3WQA8clbW153DhtFRcBD7e83t2MJZeJiEng\nMHBeu/0aqgF+ZnNdB1fO45K2trzeFBGbWl6nThQvnraOZ5woJE2dKCZmtdNmZmY16CAHTkTEFf3Y\nl1b9bixPRMSDzefjFJCc+zzysq/H3Otj62D9RfyOZ+GSAW333ogYn2GZiYjY0Je9MbO6TZw5c6ao\nHNhno3bMc/V4i86BM7y/B1jb8npNM5ZaZreks4GlwOPtVtrXxnJEnD/1XNLWQVwdDNKoHfOoHW+3\namoE9+REYWbdcw4crWMetePtVk05cAuwTtJlNHLdRuD6actsBm4Avgz8EvD5mOHun/ssmw2Xp08U\nkubTOFFsnrbM1IkCOjxRmJmZla7ZB/km4F5gO3B3RNwn6TZJ1zQX+wBwnqQdwFuA76saNZ37LJsN\nkWYf5KkTxRhwx9SJAtgaEZtpnCj+tHmiOECjQW1mZjbnRcQ9wD3TYre0PH8S+OUq6xxkY3nTzIsM\nnVE75lE73iL04kRhZrUbxfPjqB3zqB3v0JK/fTUzMzMzS3OfZTMzMzOzDDeWzczMzMwy+t5YlrRB\n0v2SdkiacQTiXCXpDkn7JX27JbZC0mclfbf57/JB7mOdJK2V9AVJ2yTdJ+mNzfjQHrOZWVXOgcOZ\nD5wDh1tfG8uSxoDbgauB9cB1ktb3cx/66E5ges3Am4HPRcQ64HN0UK5kDpkE3hoR64EfB97Q/N0O\n8zGbmXXMOXCo84Fz4BDr953lK4EdEbEzIk4BdwHX9nkf+iIivkijLFera4EPNp9/EHhlX3eqhyJi\nb0R8vfn8KI36hhcxxMdsZlaRc+CQ5gPnwOHW78byRcDDLa93N2OjYlVE7G0+fxRYNcid6RVJlwIv\nAr7CiByzmVkHnANHIB84Bw4fD/AbkOaMaUNXt0/SYuDPgDdFxJHW94b1mM3MrJphzQfOgcOp343l\nPcDaltdrmrFRsU/SaoDmv/sHvD+1kjSPxkniwxHxiWZ4qI/ZzKwC58AhzgfOgcOr343lLcA6SZdJ\nmk9jmt3Nfd6HQdoM3NB8fgPwqQHuS60kicY0ytsj4t0tbw3tMZuZVeQcOKT5wDlwuPV9Bj9JPwf8\nITAG3BER/7WvO9Ankj4KXAWMA/uA3wf+HLgbuBh4EHh1REwfADEnSXop8LfAt4AzzfDbafTZGspj\nNjOryjlwOPOBc+Bw83TXZmZmZmYZHuBnZmZmZpbhxrKZmZmZWYYby2ZmZmZmGW4sm5mZmZlluLFs\nZmZmZpbhxrKZmZmZWYYby2ZmZmZmGW4sm5mZmZlluLFsZmZmZpbhxrKZmZmZWYYby2ZmZmZmGW4s\nm5mZmZlluLFcGEn/S9J/rHtZMzOz0jkHWokUEYPeh5EhaRewCpgETgPbgA8BmyLiTJfrvgr43xGx\nZhY/Ox/4J2DJbH7ezMxsJqXlQEm3Ar8LnGwJ/3BE7OxmX2z4+M5y//1CRCwBLgHeAfwH4AOD3SV+\nB3hswPtgZmbDr7Qc+LGIWNzycEPZvo8bywMSEYcjYjPwK8ANkl4AIOlOSf9lajlJ/17SXkmPSPoN\nSSHpua3LSjoX+AxwoaRjzceFneyHpMuAfwP8P3Ufo5mZWUopOdCsE24sD1hEfBXYDfzE9PckbQDe\nArwceC5wVWYdx4GrgUdaro4fkfRSSYdm2IX/F3g7cGL2R2FmZlZdATnwFyQdkHSfpN/s5lhseLmx\nXIZHgBWJ+KuBP4mI+yLiCeDWKiuNiC9FxLLc+5JeBYxFxCerrNfMzKxGA8mBwN3A84HzgdcBt0i6\nrso2bDS4sVyGi4ADifiFwMMtrx9OLDMrza+t3gX8dl3rNDMzm4W+50CAiNgWEY9ExOmI+HvgvcAv\n1bkNGw5nD3oHRp2kH6NxovhS4u29QOvI3rVtVlW1rMk64FLgbyUBzAeWSnoU+PGI2FVxfWZmZpUM\nMAfm1qEa1mNDxneWB0TSsyT9PHAXjXI330osdjfw65KeL2kR0K6e5D7gPElLO9yFb9M48VzefPxG\ncx2XU/PVu5mZWasCciCSrpW0XA1X0vim9VMVDsNGhBvL/fdpSUdpNEh/F3g38OupBSPiM8D/BL4A\n7AD+ofnWycSy3wE+CuyUdEjShZJ+QtKxzLonI+LRqQeNr8DONF+f7vIYzczMUorIgU0bm+s9SqPe\n8zsj4oOzOywbZp6UZA6R9Hwad4QXRMTkoPfHzMysX5wDbVB8Z7lwkl4laYGk5cA7gU/7JGFmZqPA\nOdBK4MZy+V4P7Ae+R2N6UNeBNDOzUeEcaAPnbhhmZmZmZhm+s2xmZmZmltFVneXmVJTvBcaA90fE\nO2ZY3rexbU6IiL7X2tywYUNMTEy0XeZrX/vavRGxoU+7ZGZtOAfasHIOfKZZN5YljQG3A6+gMa/7\nFkmbI2JbXTvXnCzj+8ylriNz6Rjq2teq65lLn1EvTUxMsGXLlrbLnHXWWeN92h0za2O2OTB3vrPB\nq5qjhtWgcm/JObCbbhhXAjsiYmdEnKJRWPzaenbLbDRFRNuHmRXDOdCsZqXmwG66YVzEM2d62w28\nePpCkm4EbuxiO2YjISI4c+bMoHfDzDrjHGhWo5JzYFd9ljsREZuATeD+WmYz8d1js+HiHGjWuVJz\nYDeN5T3A2pbXa5qx2tTVV3Y266pLqb/4lLr2tep6BvkZpf5mBrk/pV5Vm9n36XkOtP6q2je56vJz\nqT0wKKXmwG76LG8B1km6TNJ8GnOsb65nt8xGUx39tSRtkHS/pB2Sbk68f7GkL0j6R0nflPRztR+I\n2fBzDjSr2dD1WY6ISUk3AffSKJtzR0TcV9uemY2YOvprdThC//eAuyPifZLWA/cAl3a1YbMR4xxo\nVq+h7bMcEffQSLRmVoMarpyfHqEPIGlqhH5rYzmAZzWfLwUe6XajZqPIOdCsXqV2Ven5AD8z61wH\nV9Xjkra2vN7UHEA0pZMR+rcCfy3pt4BzgZfPbm/NzMzqM5R3ls2sPh32yZqIiCu63NR1wJ0R8QeS\n/hXwp5JeEBFlnqXMzGzoDbpfcjtD1Vgu9UO2cpX2N1PDVXUnI/RfC2wAiIgvS1oIjAP7u924mY2m\nXs/0Ohu9rm6RM8jPYq7PQljqneVuqmGYWc1qGAncyQj9h4CXAUh6PrAQeKzGwzAzM6ts6KphmFm9\n6hgJnBuhL+k2YGtEbAbeCvyxpDfTGOz3mijtFruZmY2Uoa2GYWb1qqPNmhqhHxG3tDzfBryk6w2Z\nmZnVqNT7Nm4smxWk1KtqMzOzXis1B7qxbFaIQffJMjMzG5SSc2DfG8upEZmlfji9VNfo10GOcB3U\nyOE611Xa32OpV9Vm1jtz6byfU3Wfcsd81lnV6g60225d1TDqyte55avG2+n1MfRaqTnQ1TDMClLq\nSGAzM7NeqyMHStog6X5JOyTdnFnm1ZK2SbpP0kdmWqe7YZgVouSRwGZmZr1URw6UNAbcDryCxgy2\nWyRtbg5sn1pmHfA24CURcVDSypnW68ayWUF899jMzEZVDTnwSmBHROwEkHQXcC2wrWWZ1wG3R8TB\n5jZnnJDLjWWzgvjOspmZjaoOcuC4pK0trzdFxKaW1xcBD7e83g28eNo6ngcg6e9ozEdwa0T8VbuN\nurFsVgh3wzAzs1HVYQ6ciIgrutzU2cA64CpgDfBFSS+MiEPtfqCvUrfYq47SLLECRK/jdX49nxtt\nnIuPjY3Vsp7cf4LTp08n45OTk8l4u/dy26hztHEvlbY/ZtZ7/Tjv15VPc8vn8sS8efNqiZ99drq5\nsmDBgmR8NuvKyX12Tz31VDKey0+5XHfq1KlK6wc4efJkpZ8ZoRy4B1jb8npNM9ZqN/CViHgKeEDS\nP9NoPG/JrdTVMMwKcubMmbYPMzOzYVVDDtwCrJN0maT5wEZg87Rl/pzGXWUkjdPolrGz3UrdDcOs\nIKVd5ZuZmfVLtzkwIiYl3QTcS6M/8h0RcZ+k24CtEbG5+d7PStoGnAZ+JyIeb7deN5bNClFXn2VJ\nG4D30jhRvD8i3jHt/fcAP918uQhYGRHLut6wmZnZLNWVAyPiHuCeabFbWp4H8JbmoyNuLJsVpNur\n6k5qTEbEm1uW/y3gRV1t1MzMrAalfrvqxrJZQWq4qu6kxmSr64Df73ajZmZm3Sp1bI4by2aF6HA6\nzzpqTAIg6RLgMuDzs9hdMzOz2lSZ0rrfumosS9oFHKXRQXqyk9p3qbIz/fhwcuVuqpY/y5WcmT9/\nfqXlc2V2cvuZKzkzm6uwXKmdRYsWVYpXLRGXK3Xz5JNPJuPHjx9Pxtu9l9tGqf8Bp+tTjckp/zyK\nEQAAIABJREFUG4GPR0T6j8vM2ppNDqy4/oGtq2quW7x4cTK+dOnSZHzFihXJ+PLlyystv2TJkmQc\nYOHChcl47hhyn1GuFFwudz3xxBPJ+NGjR5PxQ4fS5X0nJiaS8XbvHTx4MBnP5cbSDPOd5Z+OiPxv\n1Mw61qcak1M2Am/odoNmI8450Kwmpd7YcjcMs0LUNBL46RqTNBrJG4Hrpy8k6QeB5cCXu92gmZlZ\nt0qexbbbSUkC+GtJX5N0Y2oBSTdK2jqtn6WZJUz12co9Ovj5SWCqxuR24O6pGpOSrmlZdCNwV5R6\nGW82NzgHmtWo2xzYK93eWX5pROyRtBL4rKTvRMQXWxdoDj7aBCDJidmsjX7UmGy+vrXrDZmZc6BZ\njYbyznJE7Gn+ux/4JI2yVWY2S6VeVZvZ93MONKtXqTlw1neWJZ0LnBURR5vPfxa4baafq3KwVUfv\ntls+V7lh3rx5leLnnHNOMp4bkXvuuecm47mRxrnPp+qo23Zy+zQ+Pp6M50Yn5yp6VB0JfOTIkWS8\n3d9KbmTvqVOnkvHc30ZJDdCS+2uZ2TPNNgem5M5DVSsOtcuBufdy5/FcJYlcrjv//POT8bVr1ybj\nl156aTJ+8cUXJ+OrV69Oxpcty08+WrXqRU4u3xw7diwZP3z4cDL+2GOPJeP79+9Pxvfu3Zvdp1wb\n4qmnnkrGcxU9Un9Lg8qLJefAbrphrAI+2fyjOxv4SET8VS17ZTaiSmq8m1lbzoFmNSs1B866sdyc\nIexHatwXs5FX6lW1mT2Tc6BZ/UrNgS4dZ1aIQffJMjMzG5SSc6Aby2YFKfWq2szMrNdKzYFuLJsV\npNSrajMzs14rNQf2vbGcGoVa14fTboRrbsRvbkTpokWLkvHcyNtcJYmlS5cm47lRurlqDrnRtbnR\n0pAf/Zobzbxy5cpkfNWqVZW2nZvnPvd7PnHiRDLeTqn/obpR8khgM+udqtUZctqdF3Pn61wOzFV+\nylVHuvDCC5PxdevWJeM/8AM/kIznqmGsWLEiGV+wYEEyDvnPI3eezeXfXPvh9OnTyXiuelXuM821\nN3K5ut17uc+jXVuhFCXnwPI/PbMRUmqNSTMzs16rIwdK2iDpfkk7JN2ceP81kh6T9I3m4zdmWqe7\nYZgVpNSrajMzs17rNgdKGgNuB14B7Aa2SNocEdumLfqxiLip0/W6sWxWEN89NjOzUVVDDrwS2NEs\n7Yiku4BrgemN5UrcDcOsEFP9tdo9OjHTV1DNZV4taZuk+yR9pNYDMTMzq6jDHDguaWvL48Zpq7kI\neLjl9e5mbLpflPRNSR+XlJ5msoXvLJsVpNur6k6+gpK0Dngb8JKIOCgpPbLTzMysjzrIgRMRcUWX\nm/k08NGIOCnp9cAHgZ9p9wN9byz38mvmdiOKq44EXrhwYTKeq4ZxwQUXJOO5Ebzz5s1LxnNzzefu\nKj7xxBPJeDu5EbnnnXdeMp6rhpH7Xebmps/9fnLL50YUQ77SR26fcvFeVmeZjRr6LHfyFdTrgNsj\n4iBAROzvdqNmVr+q56J2OTD3XtVqGLkKT7k8sXr16mQ8lzNzVR5y58Zc9SXIV1rKxXP5NBfP5evj\nx48n40ePHk3Gc8eQWw/k82bOXOniV0MO3AO03ile04w9LSIeb3n5fuBdM63U3TDMCjHTKODmya6O\nr6CeBzxP0t9J+gdJG3p3VGZmZjPrMAfOZAuwTtJlkuYDG4HNrQtIar2CuwbYPtNK3Q3DrCAdXFXX\n8RXU2cA64CoaV91flPTCiMjfojEzM+uxbu8sR8SkpJuAe4Ex4I6IuE/SbcDWiNgM/Laka4BJ4ADw\nmpnW68ayWUFq+Kpsxq+gaNxt/kpEPAU8IOmfaTSet3S7cTMzs9mqo7tIRNwD3DMtdkvL87fRGLfT\nMXfDMCtETdUwZvwKCvhzGneVkTROo1vGzvqOxMzMrJq6KkL1gu8smxWk26vqDr+Cuhf4WUnbgNPA\n70wb8GBmZtZ3pQ5E7Htjud1o3UFsNzcS+Nxzz03Gly9fnoznRvyOj48n41X/IPbvTxcsOHnyZPZn\ncvPW5yp95PY1N2o5N0I4t6+nTp1KxnMjfuushpFT2n/MOq6cO/gKKoC3NB9mVqgqVXxmUjUH5vJE\nrhrG+eefn4yvXJmuTJnLpTmPP56+ns/lG4CJiYlk/MiRI8l4LhdVrZJRNafl4rmqHZA/hty6cu2B\nYcyBveA7y2YFKe3EZWZm1i+l5kA3ls0KMdVfy8zMbNSUnAPdWDYrSKlX1WZmZr1Wag50Y9msIKVe\nVZuZmfVaqTnQjWWzQlSYocjMzGyolJwD52RjOTeqt90I4bPOSpeUnj9/fjKem59+1apVyfiFF16Y\njOdG/OZGsubmez98+HAynptrHuCcc85JxnPHVrWix6OPPpqM50bj5vb12LFjyXi7Sh+lXn12a1iP\ny8x6L5fn2r2Xq4aRyx/Lli1LxnP546KLLkrGc7kxl+sOHDiQjD/00EPJOMDevXsrrSu37apVMnI5\nMFcRKpf3c1Wf2v1MroJGrhFaWuO01Bw446Qkku6QtF/St1tiKyR9VtJ3m/9WqwFjZklTV9a5h5n1\nl3OgWf+UmgM7mcHvTmDDtNjNwOciYh3wueZrM+tCybMXmY2wO3EONOu5knPgjI3liPgiMP07i2uB\nDzaffxB4Zc37ZTaSSr2qNhtVzoFm/VNqDpxtn+VVETHVGehRIN2RF5B0I3DjLLdjNlJ899hsTnAO\nNOuBUnNg1wP8IiIkZZv7EbEJ2ATQbjmzUTfoK2czq8450KweJefA2TaW90laHRF7Ja0G8pOzdyD3\n4bSrblHV2NhYMp4b8ZsbqVt1xG+u2sbBgweT8cceeywZ378//RG3mzv+Wc96VjJ+wQUXJOO5Y8hV\nz9i9e3cynqv0Uddc9lDeCN661HFVLWkD8F5gDHh/RLxj2vuvAf47sKcZ+v8i4v1db9hsdNSaA/sh\nl09z1TAWLlyYjOeqYeTyx8UXX5yM5/JTrgJErppHuzyRqxiRi+eqWxw6dKhSPLee3HZzxzCbPFd1\nXam/i0Hm11LvLHcywC9lM3BD8/kNwKfq2R2z0dZtfy1JY8DtwNXAeuA6SesTi34sIi5vPtxQNqvG\nOdCsB+rosyxpg6T7Je2QlB18K+kXJYWkK2ZaZyel4z4KfBn4AUm7Jb0WeAfwCknfBV7efG1mXahp\nJPCVwI6I2BkRp4C7aAxGMrNZcA406486cmCnN4wkLQHeCHylk32bsRtGRFyXeetlnWzAzDrXwZXz\nuKStLa83NftETrkIeLjl9W7gxYn1/KKknwT+GXhzRDycWMZs5DkHmvVPDV1Anr5hBCBp6obRtmnL\n/WfgncDvdLLSOTmDn9mw6uDKeSIiZvzKaAafBj4aESclvZ5G6auf6XKdZmZmXamhz/KMN4wk/Siw\nNiL+UpIby2ZzTQ1X1XuAtS2v1/AvA/mmtvF4y8v3A+/qdqNmZmbdquHb1bYknQW8G3hNlf0qorFc\nV9WLduvJjfhdvHhxMp6rGHHppZcm47kqGUePHk3Gc5UhclUyciOEc9U8IH8Mz33uc5PxNWvWJOO5\nkb3Hjx9PxnPHfOrUqWQ8dyXZ7veZGxld6kjaTkz11+rSFmCdpMtoNJI3Ate3LjA1ir/58hpge7cb\nNbP61VkRKreuqpWiVqxYkYzn8s3atWuT8dz6c3li5cqVyfixY8eScahe6WPevHnZdaWcPHkyGa9a\n4SlXJaNdPqh6Y6XOv6Ve6TAHzvTt6kw3jJYALwD+pvmZXABslnRNRLQ2wp+hiMaymTV0e2c5IiYl\n3QTcS6N03B0RcZ+k24CtEbEZ+G1J1wCTNGYme013e21mZta9Gr5dbXvDKCIOA+NTryX9DfDv2jWU\nwY1ls6LUcWc8Iu4B7pkWu6Xl+duAt3W9ITMzsxp1mwM7vGFUmRvLZoUoefYiMzOzXqorB850w2ha\n/KpO1unGsllB5nKfazMzs26UmgPdWDYriO8sm5nZqCo1B/a9sZz6IPpRDWP+/PnJ+PLly5PxqvPc\nj4+PJ+O5UbG5ag5LlixJxnOVKs4999xkHOBHfuRHkvEXvOAFyfiqVauS8V27diXjuRG8Obljzo1A\nzlUAabft3N9Aqf8BW9VUDcPMhkTuvDWb81zuZ6rm39x5PJdjcxUpcnJVMnI5tl2eyK0rVwVr6dKl\nyfiiRYuS8dyx5X4PufN71Xg7c6HqRU7JOdB3ls0KMhca9WZmZr1Qag50Y9msIKVeVZuZmfVaqTnQ\njWWzgpR6VW1mZtZrpeZAN5bNClFyfy0zM7NeKjkHurFsVpBSr6rNzMx6rdQc2PfGci9HaubmuIf8\nqNjzzjsvGc9Vhjj//POT8WXLliXjJ06cSMaf85znJOO56ha56g+50bsA69evT8Zz1TByn9GCBQsq\nLZ8baZyL5z6j06dPJ+Pt3iv1qrRTc33/zay9VA6ss+pFTu7cksstufPygQMHkvGHHnooGc/ltFxe\nOX78eDJ+5MiRZLzdOXPhwoXJeC5v5ip95H4Puc8uVwXr1KlTyXiuoke7Y2uXH1Oq/o0NSqk50HeW\nzQrhGfzMzGxUlZwD05dRZjYQZ86cafvohKQNku6XtEPSzW2W+0VJIemK2g7AzMxslurIgb3gO8tm\nBen2qlrSGHA78ApgN7BF0uaI2DZtuSXAG4GvdLVBMzOzmvjOspm1NTUSuMur6iuBHRGxMyJOAXcB\n1yaW+8/AO4F05zozM7M+qikH9oQby2YFmeqzlXsA45K2tjxunLaKi4CHW17vbsaeJulHgbUR8Zc9\nPRgzM7MKOsiBA+FuGGYF6eDKeSIiZt3HWNJZwLuB18x2HWZmZr0wZ6thSLoD+Hlgf0S8oBm7FXgd\n8FhzsbdHxD2z3YmqJU2qxgHmz5+fjOfK1+TK0OV+kbljWL58eTL+Qz/0Q8n4s5/97GQ8d2zPetaz\nknGACy+8MBnPlb/Llbupesy58ju5zzQXz62nnTrK4wzy6rWGbe8B1ra8XtOMTVkCvAD4m+ZncgGw\nWdI1EbG1242bDZu6c2Dq/3g/ynrltpErf3bo0KFkfOfOncn4vHnzkvFcSblc7s2VRMvF2+WJXG45\n++x00ye3TytWrEjGjx07lowfPHgwGT98+HAynsu9uZJyUD0v55TWR7i0/ZnSSWvkTmBDIv6eiLi8\n+Zh1Q9nMGmrqr7UFWCfpMknzgY3A5pZtHI6I8Yi4NCIuBf4BcEPZLO9OnAPNeq6uPsszVYSS9G8l\nfUvSNyR9SVJ6UooWMzaWI+KLQLoKuZnVqtv+WhExCdwE3AtsB+6OiPsk3Sbpmh7vvtnQcQ40659u\nc2BLRairgfXAdYnG8Eci4oURcTnwLhpdE9vqps/yTZJ+DdgKvDUikt87NAcgTR+EZGYJdfTXat7l\numda7JbMsld1vUGz0eQcaFazGnLg0xWhACRNVYR6unxqRLROB3kuMGMrfLbVMN4HPAe4HNgL/EFu\nwYjYFBFXdDMoyWwUzHRFXWpfLrMR5BxoVrMOc2DXFaEAJL1B0vdo3Fn+7Zn2bVZ3liNiX8sG/xj4\ni9msx8yeqdSRwGb2L5wDzXqj1xWhpkTE7cDtkq4Hfg+4od3ys2osS1odEXubL18FfHs266lbuztv\nuVGlx48fT8YfeeSRZPz+++9Pxh9//PFkPDdCuOpdwtwo3VyVj3bvnTx5Mhnft29fMr5nz55kPPcZ\n5UYCnzhxotL+5EZpw9wf8ZszV/bTbJTVnQOrVr2YTUWoqtvI5cZcdYujR48m49u3b0/Gc7kxV6li\n4cKFyfjSpUuTcchXo1q2bFkyfu655ybjufy7ePHiSvHcMeTWn6uSAfk2Ta5qyFzRh4pQ091F45ui\ntjopHfdR4Coat753A78PXCXpchr9PHYBr59pPWbW3tRIYDMrh3OgWX/UlAOfrghFo5G8Ebi+dQFJ\n6yLiu82X/xr4LjOYsbEcEdclwh+YcXfNrDLfWTYri3OgWf90mwMjYlLSVEWoMeCOqYpQwNaI2Exj\ncO7LgaeAg8zQBQM8g59ZUXxn2czMRlU/KkJFxBurrtONZbOC+M6ymZmNqlJzoBvLZoVwn2UzMxtV\nJefAvjeWU1cNVUfp5q482n3ITzzxRDK+e/fuStvYu3dvMp4b/ZqrSJEb8ZsbIZybm/7Zz352Mg75\nahJjY2PJ+K5du5Lxb33rW8n49773vWS8apWM3O/m1KlTyThUH/Gb+xsr7Sq2tP0xs/Lkzme5czvA\nokWLkvFcNYmzzkpPw5CrXvToo49mt11l/bncuGTJkmT8ggsuqLyNXNWL3PK5zzuX33N5PFcNI/d7\nq9o2grmfQ0rdf99ZNitIqVfVZmZmvVZqDnRj2awQnqXPzMxGVck5cLbTXZtZD5w5c6btoxOSNki6\nX9IOSTcn3v+3kr4l6RuSviRpfe0HYmZmVlEdObAX3Fg2K8jUlXXuMRNJY8DtwNXAeuC6RGP4IxHx\nwoi4HHgX8O66j8PMzKyqbnNgr7gbhlkhahoJfCWwIyJ2Aki6C7gW2NaynSMty59LYxYyMzOzgXE1\njD7JzZUO+Xnr9+xJTxl+6NChZDw3mjU3gjc3yjU3F3xuLvuLL744GW83WjY3ajn3x/jAAw8k4zt2\n7EjGc1Uvcp/d8ePHk/Fc1Yt2FS9yV5il9nfqVAf7Py5pa8vrTRGxqeX1RcDDLa93Ay+evhJJbwDe\nAswHfmZ2e2tmJclVYYB8NYmVK1cm47lKD7nqRU8++WSleLtqRylVKyBB9aoUubycqyxVV8Mul8dn\nUw1jris1hw9VY9lsruvg5DsREVd0u52IuB24XdL1wO/RwXSfZmZmveQ7y2Y2oxquqvcAa1ter2nG\ncu4C3tftRs3MzLrlO8tm1lZN/bW2AOskXUajkbwRuL51AUnrIuK7zZf/GvguZmZmA+Q+y2bWkW5P\nFBExKekm4F5gDLgjIu6TdBuwNSI2AzdJejnwFHAQd8EwM7MCuLFsZjOq4yuoiLgHuGda7JaW52/s\neiNmZmY1czeMNqp+OLkRou2uSHKVIXIVNHLVM3Jzx+eqXuRG4+ZGJufWf+LEiWT84MGDyXg7uVHI\njz76aDL++OOPJ+O5UdFVq1vkfm/9+E+T+lsa1H/Wkr+CMrP+q3ouyuUPgMWLFyfjq1atSsZzlZly\n5/HDhw8n40eOHEnGcxUmcrk0lzPPP//8ZLzde7ljO+ecc5Lx3L7m2g+5zyi3fJ05MNc+KrUR2qrk\nHFhEY9nMGubCCc3MzKwXSs2BnsHPrBBTV9UlTvVpZmbWS3XlQEkbJN0vaYekmxPvv0XSNknflPQ5\nSZfMtE43ls0KUupUn2ZmZr3WbQ6UNAbcDlwNrAeuk7R+2mL/CFwRET8MfBx410zrdWPZrCC+s2xm\nZqOqhhx4JbAjInZGxCkacwlc27pARHwhIqYGXf0DjfkI2nKfZbOC+O6xmZmNqg5y4LikrS2vN0XE\nppbXFwEPt7zeDby4zfpeC3xmpo3O2FiWtBb4ELAKiOaOvVfSCuBjwKXALuDVEVG9NAPV5z+fTYOi\n6qjV3D5VrYaRWz63/lz1jJzjx49n38tdheWqVeRGMz/55JPJeO6zy/1+BtkQnAsjhEseCWw2qvqR\nA9tsu1K83fksd75etGhRMr527dpkfNmyZcl4LsceO3YsGc9VeKq6n+Pj48k45Ct95CqD5I4hV9Ej\nlxtzlaJyy1ettgH1VRGrY9116TAHTkTEFXVsT9K/Aa4AfmqmZTvphjEJvDUi1gM/Dryh2f/jZuBz\nEbEO+FzztZl1wX2WzYrjHGjWJzXkwD1A65XemmbsGZoTc/0ucE1EpGsLt5ixsRwReyPi683nR4Ht\nNG5zXwt8sLnYB4FXzrQuM2vPfZbNyuIcaNY/NeTALcA6SZdJmg9sBDa3LiDpRcAf0Wgo7+9kpZX6\nLEu6FHgR8BVgVUTsbb71KI2vqMxslnz32KxszoFmvVNHDoyISUk3AfcCY8AdEXGfpNuArRGxGfjv\nwGLg/zS7pzwUEde0W2/HjWVJi4E/A94UEUda+79EREhKHqGkG4EbO92O2Sjz3WOzMjkHmvVeHTkw\nIu4B7pkWu6Xl+currrOj0nGS5tE4SXw4Ij7RDO+TtLr5/mogeSs7IjZFxBV1dcg2G2Z19FnuRUF2\ns1HmHGjWH6WO2+mkGoaADwDbI+LdLW9tBm4A3tH891M92cOaVK3QUFc1jPnz5yfjuZG9CxcuTMZz\nV1u5kcaQH2F78mS6L3tuBG/ViiG9jkP10bpzoXtDHdUwWgqyv4JGyZwtkjZHxLaWxaYKsj8h6Tdp\nFGT/la42bDak6s6BVc5FVStF5c7VkD+/5yo95HLdypUrk/Hzzz8/Gc/lxty5LncMZ5+dbq6cc845\nyTjAggULkvFcDpyYmEjGDx06lIw//vjjyfjBg+miKHVVBoHqbZdRyYG90smd5ZcAvwr8jKRvNB8/\nR+ME8QpJ3wVe3nxtZl2o4aq6JwXZzUaYc6BZn8zZO8sR8SUgd3n7snp3x2y0dXBVPZCC7GajyjnQ\nrH9KvbPsGfzMCtLBlfNACrKbmZn1WqndRdxYNitETf21qhZk/6lOCrKbmZn1Usl9lt1YNitIDVfV\nTxdkp9FI3ghc37pAS0H2DZ0WZDczM+s131luox8fTtVRxVWrYcybNy8Zz43GzVXJyMmN3m13XLlq\nGKdOnaq0jdz89HVVErF/0e1Vda8KsptZParkoqq5sd35I1eJYffu3cn48uXLk/GlS5cm4+Pj48l4\nrnrGkiVLkvFcLs3J5SeAo0ePJuMHDhxIxh988MFk/IEHHkjGH3rooWR83759yfjhw4eT8SeffDIZ\nb/f7LLVR2S3fWTaztuoa7duLguxmZma9NOiKF+24sWxWkFKvqs3MzHqt1BzoxrJZQUq9qjYzM+u1\nUnOgG8tmhSh5JLCZmVkvlZwD3Vg2K0ipV9VmZma9VmoOdGM5IzdiuWo8J/cHkRsVm5OrbAEwNjaW\njOeqZBw/fjwZr6tKRl3x2ejlCPQ6lXpVbWb9VzXftDt35XJLrnLD2WenmwenT59OxnOVJ5797Gcn\n46tXr07GFy9enIznzo1HjhxJxgH27Pm+EvMA7Ny5Mxm///77k/Fdu3Yl4/v3pytv5vbpxIkTyXgu\nJ1dtVwyDUnOgG8tmBSn1qtrMzKzXSs2BLnxrVoip/lrtHmZmZsOorhwoaYOk+yXtkHRz4v2flPR1\nSZOSfqmTdbqxbFaQqTqTuYeZmdmw6jYHShoDbgeuBtYD10laP22xh4DXAB/pdL/cDcOsIL57bGZm\no6qGHHglsCMidgJIugu4Ftg2tUBE7Gq+1/HG3Fg2K4TvHpuZ2aiqKQdeBDzc8no38OJuV+rGsllB\nfGfZzMxGVQc5cFzS1pbXmyJiUw93CZijjeXZlM2pKreu3C8yV/olVyrm4MGDlZafP39+Mp4r7wNw\n1lnpLum5cnO5+BNPPFEpXrU8Tu4znc3vsx9/G71Ux35K2gC8FxgD3h8R75j2/k8Cfwj8MLAxIj7e\n9UbNbOBy5TwhX/ItF8/lg0OHDiXjDzzwQDJ+3nnnJePLli1Lxs8555xkvGrJOoCJiYlK8dyx5cqq\n5srx1ZXr5kreqlMHxzwREVe0eX8PsLbl9ZpmrCtzsrFsNozqmL2oZXDDK2h8/bRF0uaI2Nay2NTg\nhn/X1cbMzMxqUtMMfluAdZIuo9FI3ghc3+1KXQ3DrCA1VMN4enBDRJwCpgY3tG5jV0R8E3CfDzMz\nK0a3OTAiJoGbgHuB7cDdEXGfpNskXQMg6cck7QZ+GfgjSffNtF7fWTYrSA39tXoyuMHMzKzX6hi3\nExH3APdMi93S8nwLje4ZHXNj2awgNfTXMjMzm5NK7aftxrJZIWrqr9WTwQ1mZma9VFMO7IkZG8uS\n1gIfAlYBQeNr3/dKuhV4HfBYc9G3N29991zuyiNXCaHdz+RUrXpRdUTxkSNHkvFcBYuqcag+wjY3\nkjr3WVQdXZ3TjyvJdn8b0w3yyraGbfdkcIPZqJpLObDd+SP3Xi6n5XLUsWPHkvG9e/cm42NjY5Xi\nuZxWdf+heu6qWgWrrioWVfLTTEq9M9upUve/kzvLk8BbI+LrkpYAX5P02eZ774mI/9G73TMbLd1e\nVUfEpKSpwQ1jwB1TgxuArRGxWdKPAZ8ElgO/IOk/RcQPdbvvZkPKOdCsT+bsneWI2AvsbT4/Kmk7\njUFEZlajumbw68XgBrNR5Rxo1h8lz2JbqXScpEuBFwFfaYZukvRNSXdIWp75mRslbZ02gt/MEs6c\nOdP2YWaD4xxo1lul5sCOG8uSFgN/BrwpIo4A7wOeA1xO46r7D1I/FxGbIuIKj+A3m1kNdZbNrAec\nA816r9Qc2FE1DEnzaJwkPhwRnwCIiH0t7/8x8Bc92UOzEVHySGCzUeYcaNZ7JefATqphCPgAsD0i\n3t0SX93sywXwKuDbnWwwNepzNtUtUuq86qg6yjX3C85VmMjJHbPvKnZmrn9+c2U/zUZF3Tmwl2ZT\nESqXu6o2WtpVpUjJ7WvVc/hszpl1VZ/w+bp+pX6mndxZfgnwq8C3JH2jGXs7cJ2ky2mU0tkFvL4n\ne2g2Qkq9qjYbYc6BZn1Sag7spBrGl4DUZVhf6kmajZJSr6rNRpVzoFn/lJoDPYOfWSFK7q9lZmbW\nSyXnQDeWzQpS6lW1mZlZr5WaA91YNitIqVfVZmZmvVZqDiy6sdyP+dVLu4opbX8GaTYju+uorDKo\n38Gg60ia2fCqq7pU1WoVva5u0e64qja86jqGnLqqcAyrknNg0Y1ls1FT6lW1mZlZr5WaAytNd21m\nvVXH7EWSNki6X9IOSTcn3l8g6WPN97/SnMLXzMxsoErNgW4smxViaiRwu8dMJI0BtwNXA+tp1IJd\nP22x1wIHI+K5wHuAd9Z8KGZmZpWUnAPdWDYrSA1X1VcCOyJiZ0ScAu4Crp22zLXAB5t/8clUAAAg\nAElEQVTPPw68TO5MZ2ZmA1ZqDnRj2awgHVxVj0va2vK4cdoqLgIebnm9uxlLLhMRk8Bh4LzeHJGZ\nmVlnSs2B/R7gNxERDzafjwMTda681FGULWo/5sJ1dbx1/j4rruuS2jZczb0RMT7DMhMRsaEve2Nm\ndetpDuyHquflacsXkfe7PIYqnnG8c6CNMsU5cJq+NpYj4vyp55K2RsQV/dz+oI3aMY/a8XarphPA\nHmBty+s1zVhqmd2SzgaWAo/XsG0za8M5cLSOedSOt1sl50B3wzAbLluAdZIukzQf2AhsnrbMZuCG\n5vNfAj4fc+iWh5mZWUZPcqDrLJsNkYiYlHQTcC8wBtwREfdJug3YGhGbgQ8AfyppB3CAxsnEzMxs\nTutVDtSgbihJujEiNg1k4wMyasc8asdrZtapUTw/jtoxj9rxDrOBNZbNzMzMzErnPstmZmZmZhlu\nLJuZmZmZZfS9sTzTnN3DQtIdkvZL+nZLbIWkz0r6bvPf5YPcxzpJWivpC5K2SbpP0hub8aE9ZjOz\nqpwDhzMfOAcOt742ljucs3tY3AlMrxl4M/C5iFgHfK75elhMAm+NiPXAjwNvaP5uh/mYzcw65hw4\n1PnAOXCI9fvOcidzdg+FiPgijZIkrVrnI/8g8Mq+7lQPRcTeiPh68/lRYDuNKSWH9pjNzCpyDhzS\nfOAcONz63VjuZM7uYbYqIvY2nz8KrBrkzvSKpEuBFwFfYUSO2cysA86BI5APnAOHjwf4DUhztpih\nq9snaTHwZ8CbIuJI63vDesxmZlbNsOYD58Dh1O/Gcidzdg+zfZJWAzT/3T/g/amVpHk0ThIfjohP\nNMNDfcxmZhU4Bw5xPnAOHF79bix3Mmf3MGudj/wG4FMD3JdaSRKNKSS3R8S7W94a2mM2M6vIOXBI\n84Fz4HDr+wx+kn4O+EP+Zc7u/9rXHegTSR8FrgLGgX3A7wN/DtwNXAw8CLw6IqYPgJiTJL0U+Fvg\nW8CZZvjtNPpsDeUxm5lV5Rw4nPnAOXC4ebprMzMzM7MMD/AzMzMzM8twY9nMzMzMLMONZTMzMzOz\nDDeWzczMzMwy3Fg2MzMzM8twY9nMzMzMLMONZTMzMzOzDDeWzczMzMwy3Fg2MzMzM8twY9nMzMzM\nLMONZTMzMzOzDDeWzczMzMwy3FgujKT/Jek/1r2smZlZ6ZwDrUSKiEHvw8iQtAtYBUwCp4FtwIeA\nTRFxpst1XwX874hYU/HnfhT4Q+BHgePAf4uI93azL2ZmZtOVlgMlfQb4iZbQfOD+iHhhN/tiw8d3\nlvvvFyJiCXAJ8A7gPwAfGMSOSBoH/gr4I+A84LnAXw9iX8zMbCQUkwMj4uqIWDz1AP4e+D+D2Bcr\nmxvLAxIRhyNiM/ArwA2SXgAg6U5J/2VqOUn/XtJeSY9I+g1JIem5rctKOhf4DHChpGPNx4Ud7MZb\ngHsj4sMRcTIijkbE9vqP1szM7F8UkgOfJulSGneZP1TPEdowcWN5wCLiq8BunvlVEACSNtBo0L6c\nxl3fqzLrOA5cDTzScpX8iKSXSjrUZvM/DhyQ9PeS9kv6tKSLuzwkMzOzjgw4B7b6NeBvI2JX9aOw\nYefGchkeAVYk4q8G/iQi7ouIJ4Bbq6w0Ir4UEcvaLLIGuAF4I3Ax8ADw0SrbMDMz69KgcmCrXwPu\nrLJ+Gx1uLJfhIuBAIn4h8HDL64cTy3TjBPDJiNgSEU8C/wn4vyUtrXk7ZmZmOYPKgQBIeilwAfDx\nXqzf5j43lgdM0o/ROFF8KfH2Xhp3f6esbbOq2ZQ1+ea0n3NpFDMz65sB58ApNwCfiIhjXazDhpgb\nywMi6VmSfh64i0a5m28lFrsb+HVJz5e0CGhXT3IfcF7Fu8J/ArxK0uWS5jXX/6WIOFxhHWZmZpUU\nkgORdA6N7h53Vvk5Gy1uLPffpyUdpfF10u8C7wZ+PbVgRHwG+J/AF4AdwD803zqZWPY7NPob75R0\nSNKFkn5CUvZKOSI+D7wd+EtgP40BFNfP9sDMzMxmUEwObHolcKi5DbMkT0oyh0h6PvBtYEFETA56\nf8zMzPrFOdAGxXeWCyfpVZIWSFoOvBP4tE8SZmY2CpwDrQRuLJfv9TS6SHyPxvSgvznY3TEzM+sb\n50AbOHfDMDMzMzPL8J1lMzMzM7OMs7v54eZUlO8FxoD3R8Q7Zljet7FtTogI9XubGzZsiImJibbL\nfO1rX7s3Ijb0aZfMrA3nQBtWzoHPNOvGsqQx4HbgFTTmdd8iaXNEbKtr58xGycTEBFu2bGm7zFln\nnTXep90xszacA83qVXIO7ObO8pXAjojYCSDpLuBawCcKs1k6c+bMoHfBzDrjHGhWs1JzYDeN5Yt4\n5jztu4EXT19I0o3AjV1sx2wkRAQecGs2ZzgHmtWo5BzY8wF+EbEpIq6IiCt6vS2zue7MmTNtH52Q\ntEHS/ZJ2SLo58f7Fkr4g6R8lfVPSz9V+IGYGOAeaVVFHDuyFbu4s7wHWtrxe04xZCyndRz4Xr1Pu\nCq3qtqte6ZV6ZTgXdPvZddiP8veAuyPifZLWA/cAl3a1YbPR4xxoVrNS2w/d3FneAqyTdJmk+cBG\nYHM9u2U2eiKijqvqp/tRRsQpYKof5TM2BTyr+Xwp8EhtB2E2OpwDzWpUUw7siVnfWY6ISUk3AffS\nKJtzR0TcV9uemY2gDq6qxyVtbXm9KSI2tbzupB/lrcBfS/ot4Fzg5bPbW7PR5RxoVr9S7yx3VWc5\nIu6h8RWumdWggyvniRr6Pl4H3BkRfyDpXwF/KukFEVHmMGSzQjkHmtVrGKthmFnNariq7qQf5WuB\nDc3tfVnSQmAc2N/txs3MzGar1DvLnu7arBA19dfqpB/lQ8DLACQ9H1gIPFbjoZiZmVUylH2Wh13V\nKhZjY2OV4vPmzUvG58+fn4wvXLgwGV+wYEEyDnD22elf71lnpa+RJicnk/GnnnoqGX/yySeT8RMn\nTiTjJ0+eTMZPnz5dKQ7lXn12q9vjyvWjlHQbsDUiNgNvBf5Y0ptpDPZ7TQzrB2pmZnNGqanIjWWz\nQkxdVdewnu/rRxkRt7Q83wa8pOsNmZmZ1aSuHChpA/BeGjeM3h8R75j2/nuAn26+XASsjIhl7dbp\nxrJZQUq9qjYzM+u1fsw1EBFvbln+t4AXzbReN5bNClLqSGAzM7NeqyEHPj3XAICkqbkGtmWWvw74\n/ZlW6sayWUF8Z9nMzEZVn+YaAEDSJcBlwOdn2qgby2aFqKu/lpmZ2VzTYQ6sY66BKRuBj0dEvppA\n08g3lnPVLXIVI3LxXBWLXLWKRYsWJeOLFy9OxpctS/c9Hx8fT8YBlixZkozn9jVXDePo0aPJ+MTE\nRDK+b9++ZPzQoUPJ+BNPPJGMt7vCzP2Hmut3Zuf6/puZmc1Wn+YamLIReEMnKx35xrJZSXxn2czM\nRlUNOfDpuQZoNJI3AtdPX0jSDwLLgS93slI3ls0KERG+s2xmZiOpjhzY4VwD0GhE39XpHANuLJsV\nxHeWzcxsVPVjroHm61urrNONZbOC+M6ymZmNqlJzoBvLZoVwNQwzMxtVJedAN5YzclUyzj47/ZHl\nKkxUrXqRq26xevXqZHzVqlXJOMDSpUuT8dwxHD9+vNLyuSoZ8+bNS8bHxsaS8dxnnYsPs1Kvqs3M\nzHqt1ByYroNmZgNx5syZto9OSNog6X5JOyTdnHj/PZK+0Xz8s6R0TT8zM7M+qiMH9oLvLJsVpNur\nakljwO3AK2jMXLRF0uaIeHqqz4h4c8vyvwW8qKuNmpmZ1aDUO8tuLJsVoqb+WlcCOyJiJ4Cku4Br\ngW2Z5a8Dfr/bjZqZmXXDfZbNrCM1XFVfBDzc8no38OLUgpIuAS4DPt/tRs3MzLrlO8tmNqMOrqrH\nJW1teb0pIjbNcnMbgY9HxOlZ/ryZmVltfGfZzNrqcPaiiYi4os37e4C1La/XNGMpG4E3dL6HZmZm\nvVHyLLZdNZYl7QKOAqeByRmS+EBVLU9WV+m4hQsXJuO5sm65UnC50nHnn39+Mt5u208++WQyfvLk\nyWT88OHDyfiRI0eS8VwJutz6T59O39gs9T9NL9VwVb0FWCfpMhqN5I3A9dMXkvSDwHLgy91u0GxU\nzaUcWFWvS3rWtZ4680TVdY1ijuq1Yb6z/NMRMVHDesxGXrcn34iYlHQTcC8wBtwREfdJug3YGhGb\nm4tuBO4Kn+3NuuUcaFaTUlOSu2GYFaKukcARcQ9wz7TYLdNe39r1hszMzGpScjWMbiclCeCvJX1N\n0o2pBSTdKGnrtEFJZpYw1Wcr9zCzojgHmtWo1BzY7Z3ll0bEHkkrgc9K+k5EfLF1geZI/U0Akpzt\nzdoo9arazJKcA81qVEcOlLQBeC+Nrojvj4h3JJZ5NXArjQvef4qI7xvb06qrO8sRsaf5737gkzQm\nRDCzWSr1qtrMvp9zoFm9us2BLbPYXg2sB66TtH7aMuuAtwEviYgfAt4003pnfWdZ0rnAWRFxtPn8\nZ4HbZru+QcmNyB0bG0vGc9UwcpUnFi9enIyPj48n4ytXrqy0fG79AKdOnUrGDxw4kIzv3r27Unz/\n/v3J+NGjR5NxV8Nor+T+Wmb2TCXmwHYVJqrmtKqVn6rGzzorfa8ut9065XJLLhflcmkunst1ueUn\nJyeT8Xb5YBjzYx9nsX0dcHtEHGxuN92YadHNX+Uq4JPN/5xnAx+JiL/qYn1mI28YT4BmQ8o50Kxm\nHeTAmSbm6mQW2+cBSPo7Gl01bp3p/+6sG8vNVvuPzPbnzez7+c6y2dzgHGhWvw5y4EwTc3XibGAd\ncBWNibu+KOmFEXGo3Q+YWQHcL9nMzEZVTTmwk1lsdwNfiYingAck/TONxvOW3Eq7LR1nZjU6c+ZM\n24eZmdmwqiEHPj2LraT5NCbg2jxtmT+ncVcZSeM0umXsbLdS31k2K4jvLJuZ2ajq0yy29wI/K2kb\njanqfyciHm+33pFvLFethpEb2XvOOeck48uWLUvGc1UvcvGlS5cm4+0cOpTufpOrbvHQQw8l43v3\n7k3Gc1UvnnrqqWS86ojfdv9phrFR6WoYZtYql59ylSRy+Qng3HPPTcZzuWX58uXJeK4y03nnnVdp\nPUuWLEnGFy1alIwvWLAgGc/l6nbv5T7XXO46fPhwMv7oo48m4w8++GAynsuxjz32WDJ+/PjxZBzy\nFTfmcm7s1yy20fiQ3tJ8dGTkG8tmJZnLJzozM7NulJoD3Vg2K4jvLJuZ2agqNQd6gJ9ZQeqYwU/S\nBkn3S9oh6ebMMq+WtE3SfZI+UutBmJmZzUKps9j6zrJZIeror9Uy1ecraJTH2SJpc0Rsa1mmdarP\ng5LSHeXNzMz6pORxO24smxWkhivnnkz1aWZm1mvus1yo3KjY3Pz0CxcuTMYXL16cjOdGDueqXqxY\nsaLSdg8cOJCMAzzyyCPJeNWqF0eOHEnGcyOHS70ynAs6+OwGMtWnmfVfrppDrjJELg8BnH/++cn4\nhRdemIxfdtllyfizn/3sZPzSSy9NxtesWZOM56pn5Kp25I45l6sh//nlqomcOnUqGc9Vq9i+fXsy\n/tWvfrXSdnO5NLc/7X7m9OnT2Z+ZC0ptP4x8Y9msFB32yRrIVJ9mZma9NOh+ye24sWxWkBquqnsy\n1aeZmVmvlXpn2dUwzApSw0jgnkz1aWZm1muuhmFmbdUxErhXU32amZn1kqthmFlH6rhy7sVUn2Zm\nZr3mPsuFyo1OnTdvXjJ+zjnnJOPLli1LxnMjkHPxpUuXJuMnTpxIxvft25eMQ35++tx89rl56HN/\nvLlRyLnlc1eMk5OTlZZvt425rtSrajPrnVxVplx+mj9/fjKeqyQB+dySq8yUq2JxySWXJOO56hkX\nXHBBMp6r3JH7LGZTMSJXDSP3OeWqUeXyfi4v53Jv7phz7Y3cZzHMSs2BI99YNivJsF4EmJmZzaTU\nHOgBfmaFmOqv1e5hZmY2jOrKgZI2SLpf0g5JNyfef42kxyR9o/n4jZnW6TvLZgUp9arazMys17rN\ngZLGgNuBV9Aok7pF0uaI2DZt0Y9FxE2drteNZbOC+O6xmZmNqhpy4JXAjojYCSDpLuBaYHpjuRJ3\nwzArxEz1JX3X2czMhlWHOXBc0taWx43TVnMR8HDL693N2HS/KOmbkj4uaW3i/WeY8c6ypDuAnwf2\nR8QLmrEVwMeAS4FdwKsj4uBM6xqk3KjSXEWH3Dz0udGsuXnucyOBx8fHk/HcKOe9e/cm47lRtwD7\n9+9Pxk+ePJmML1y4MBnPjcjOxXNXhrlRy7n9ycUh/1XNXG9Q+s6yWVlKzIFVKw4BnD59OhnPVZl4\n4oknkvHDhw8n47kcdeTIkWQ8lw9y66+6HshXschV9Hje856XjOcqieTaFVUrP+V+B7nl221jruvg\nuCYi4oouN/Np4KMRcVLS64EPAj/T7gc6ubN8J7BhWuxm4HMRsQ74XPO1mXXJd5bNinMnzoFmfVFD\nDtwDtN4pXtOMtW7j8YiYuhv3fuD/mmmlMzaWI+KLwIFp4WtptMRp/vvKmdZjZu25GoZZeZwDzfqj\nphy4BVgn6TJJ84GNwObWBSStbnl5DbB9ppXOdoDfqoiY+s7lUWDVLNdjZi1899hsTnAONOuBbnNg\nRExKugm4Fxjj/2/v7qPluuv73n++Pnp+tOwjy7IeLGELguwQuxE2XaTEDbhLpokNKymRnaSmpTHJ\nRQ0E0luH9Bqu265L0saEe5dLo4BrJ5dgKJCgpErdxCGL0gWuBPgCkmP7IEvWs3z0/Cwd6Xv/mDnq\n+Pj3nZl9Zu85vzPzfq01SzPf2ee3956js7+/3579+27pMXffamYPS9ri7hsl/ZqZ3S1pRLWB8Htb\ntdtxNQx3dzML965+8fXYC7ABJHD2GJhcyIFAecrIge6+SdKmMbGHGp7/pqTfLNLmeKthHBg9jV3/\nNz2TrLZRG9x9TQkXZAM9raxqGFUUZAfwKuRAoGQ5V4Qa75nljZLul/SJ+r9fLW2LKhLNWo3uHR9V\nw5g7d24yHlW9iOLR7NqjR48m4/v370/GDx06lIxL8QgtqsQxa9asZDyq0BH9xz179mwyfvLkyWQ8\nmuUc/c4k6cyZM8l4s9nDk0Gno+qqCrIDeJVSc2DR6j7Rca5ZBaHoODs8PJyMR8f9qKpGVH0pqvQQ\n5a4o10Xb3+yYv3DhwmT8tttuS8avuuqqZDyqFHXixIlkPKroEeX3qPJIs33r1Uv2cv12teWZZTP7\nvKRvSnqDme02s/epdoC408xelPSO+msAHSphVH25ILu7n5c0WpAdwDiQA4HumbRnlt393uCtt5e8\nLUBfG50J3MKgmW1peL3B3Tc0vE4VZL890c7PmtnbJL0g6dfdfVdiGaDvkQOB7mgzB04IbncNZCTX\nguwAAFQt184yt7sGMpJrQXYAAKo2aS/DANAdJX0Fdbkgu2qd5HWS7mtcwMwWN9SIbasgOwAAVeIy\njAxElRWmTEl/BNE95aPZstdcc00yHs3GjWYaR7NoDx8eewOpmqiahyStWLGi0DZFFTqizyiaeR3N\n+I1mS0ezn5v90UQzrKOZ2pNl5nCuBdkBdF90PIuOf1FVBSk+jkfxK65If/F8/vz5QstHVSyi435U\nnSOqgBRVrpKKH0+jilDRPkd5ed++fcl4tG/R7y36/feyXHN133SWgckg14LsAABUjTPLAFrKdVQN\nAEDVcs2BdJaBTOR8vRYAAFXKOQfSWQYykuuoGgCAquWaA+ksAxnJdVQNAEDVcs2BfdNZLqsaxuDg\nYDIeVcOIKkycPHkyGY9mDkczjW+88cZkXJKWLVuWjC9evDgZnzFjRjIe3Z8+qtxx4MCBZLxoVY1o\n9nOz96JZy7mOVhtNdB1JAHmJjgdFq2RI8TEzqqgU5ZyonaiTE1VHiipDnDp1KhmPPosor0jSggUL\nkvHrr78+GY+qXUW5bs+ePcn43r17k/Eov0d5q5moTzOZc0hZOdDM1kr6lGoVoT7j7snb0ZvZz0r6\nkqQ3u/uW1DKj+qazDEwGuY6qAQCoWqc50MwGJD0q6U5JuyVtNrON7r5tzHJzJX1Q0jPttMsd/ICM\n5Hr3IgAAqlZCDrxN0pC7b3f385KelHRPYrl/Lem3JZ1tp1E6y0AmRmcCN3sAANCL2syBg2a2peHx\nwJhmlkja1fB6dz12mZn9HUnL3P2/tLttXIYBZISzxwCAftVGDhx29zXjbd/MrpD0iAreuZbOMpAR\nzh4DAPpVCTlwj6TG6gZL67FRcyXdLOlv6pMkr5W00czubjbJr286y9HM0alTpybjc+fOTcaj2bIL\nFy5MxqdNm5aMnz2bvkwm+o8SVbCIKl5I0vLly5PxOXPmFNqmaHZyNFs62ofTp08n44cPHy4Ul+IZ\n0NHvebLgzDKAUdHxIDrGRpWLpLjiQlTdomg1jGjdUf6I8k20z1GujnKyJP3Ij/xIofisWbOS8Zde\neikZ37VrVzIeVfqI9jky2fPZeJSQAzdLWmVmK1XrJK+TdF9D+8ckXS5rZmZ/I+k3WlXD4JplIBNl\nXbNsZmvN7HkzGzKzB5ss97Nm5mY27q+0AAAoQxk50N1HJK2X9JSk5yR90d23mtnDZnb3eLetb84s\nA5NBp6PqqsrmAABQtTK+XXX3TZI2jYk9FCx7RzttcmYZyEgJZ5YrKZsDAEDVcq0IRWcZyESr+pJt\njrgrKZsDAECVSsqBleAyDCAjbYycB82scSLCBnff0G774y2bAwBA1XKtCEVnGchICTUmKymbAwBA\n1XKtCNWys2xmj0n6aUkH3f3meuzjkn5Z0iv1xT5av6A6W1EJlqgEWVRCZt68ecl4VGouaj8aPc2e\nPTsZv+aaa5LxZqXjZsyYkYwfOnQoGT927FgyfvHixWQ8KosXfXZRybqonah0kRT/PidzqZ3RmcAd\nqqRsDtCvcsyBRUvKSdKFCxeS8eg4G61jYGCg0PJRSbkoN0b5IMofb3jDG5JxSVqzJn1eIcqbUXm9\nqETc3r17k/ETJ04k41EuRU1JObAS7Vyz/LiktYn4J939lvoj644yMFl0er1WVWVzgD72uMiBQFdM\n2muW3f3rZrai+k0BUMaouoqyOUC/IgcC3TOZzyxH1pvZ98zsMTNbEC1kZg+Y2ZYxk5IAJOQ6qgbw\nGuRAoGS55sDxdpY/LekGSbdI2ifpd6MF3X2Du69pMSkJ6Htl3cEPQOXIgUDJcs6B46qG4e4HRp+b\n2R9I+vPStgjoY5w9BvJHDgSqkWsOHFdn2cwWu/u++st3S/pBeZtUjaLVE6IZuTNnziy0fDRzePr0\n6cn4/PnzC633zJkzybgUz9Tdt29fMn7u3LlkfHBwMBlfuHBhMh59FtGoMIpHs6ib/cxk16v7BfSS\nic6B46mGER1PoxwYtRXltKiqRhSP8kRUxWn58uXJ+C233JKMS9Ib3/jGQut+8cUXk/EXXnghGd+z\nZ08yfvLkyWQ8qoYR/T5z7ThWKdcc2E7puM9LukO1myHslvQxSXeY2S2SXNIOSe+vcBuBvjDR12QB\neC1yINAdOefAdqph3JsIf7aCbQH6Xq6jaqBfkQOB7sk1B3ZSDQNAyXKdCQwAQNXKyIFmttbMnjez\nITN7MPH+r5jZ983sWTP7hpmtbtUmt7sGMpHz3YsAAKhSGTnQzAYkPSrpTkm7JW02s43uvq1hsT92\n9/9YX/5uSY8ofeOhy+gsAxnh7DEAoF+VkANvkzTk7tslycyelHSPpMudZXc/3rD8bNXmHjRFZzlQ\n9BcWzRCOqlhceeWVyfjZs2eT8ahSxZEjR8JtiqpenDp1KhlfsCBdVz+qehEtf/DgwWQ8qtxx/Pjx\nZLxZpY/z588n45O9s8mZZQDj1ez4Fx1bLly4UGj5KNdNmZLuTkTxqCLFnDlzkvGVK1cm4zfddFMy\nLklXXXVVMn7o0KFkfNu2bcl4VCVj//79yXiUY6O8VbRKRi8rIQcukbSr4fVuSbePXcjMPiDpw5Km\nSfqpVo1yzTKQEa5ZBgD0qzZy4ODoHTHrjwfGuZ5H3f0GSf9S0r9qtTxnloFMcM0yAKBftZkDh1vc\nDXOPpGUNr5fWY5EnVbsjZ1OcWQYykutMYAAAqlZCDtwsaZWZrTSzaZLWSdrYuICZrWp4+Q8lpa+z\nacCZZSAjuc4EBgCgap3mQHcfMbP1kp6SNCDpMXffamYPS9ri7hslrTezd0i6IOmIpPtbtUtnGchE\nSdclVzITGACAKpU1N8fdN0naNCb2UMPzDxZts286y9EvYGRkJBmPqlJEs1yj2azz589PxqMZwlE7\n0axbM0vGpXgm8PLly5PxpUuXJuNRNYxjx44l41GFjgMHDiTjhw8fTsajz1oqPoN7smhj+wfNbEvD\n6w3uvqHhdSUzgQHkbzzVMKKfKSse5aioqsbg4GAyfuONNybjixcvTsalOE/88Ic/TMajahg7d+5M\nxqMcGPUfou0ZTzWMXp3wnWsO75vOMjAZtHEAbDW5od31PCrpUTO7T7WZwC2/hgIAoEq5DgLoLAOZ\nKKkaRiUzgQEAqFLOFaGohgFkJNeZwAAAVC3Xew1wZhnISK4zgQEAqFquZ5bpLAMZyXUmMAAAVeOa\n5QkWjVbOnDmTjEcVHQ4ePJiMRxUdbrjhhmR8yZIlyfjs2bOT8aiyxenTp5NxSZo2bVoyPm/evELL\nDw8PJ+Mvv/xyMv7ii+lv9V966aVk/JVXXknGm1XDiKqY5PqH1o6cr9cCMLkVPTYWPRZFVS+uuCJ9\ntefcuXOT8WXLliXjUbWmGTNmhNsUVZF67rnnkvGhoaFkPMpRUf4tq+rFZM5n45FzDuybzjIwGfTb\nwREAgFG55kA6y0Amch5VAwBQpZxzIJ1lICO5jqoBAKharjmQzjKQkVxH1QAAVC3XHEhnGchIrqNq\nAACqlmsObNlZNrNlkv5Q0iJJLmmDu3/KzK6S9AVJKyTtkPQed0+XkMhANAs1qkiciY4AACAASURB\nVIZx4MCBZPyFF15IxqP701977bXJ+OrVq5PxaCZwVD0jqgohSefPn0/GT548mYzv3LkzGX/22WeT\n8WeeeSYZ37ZtWzK+a9euZDyqPBJtv5Tv6LMTOV+vBfSrXsmBkbI6J1OmpLsTc+bMScajnLl8+fJk\nfMGCBcl4s4pQe/akb166ffv2ZDyqdhXlzChHUfVifHLOge3cwW9E0kfcfbWkt0j6gJmtlvSgpKfd\nfZWkp+uvAXQg17sXAX2MHAh0SRk50MzWmtnzZjZkZq/5uzSzD5vZNjP7npk9bWbXt2qzZWfZ3fe5\n+3fqz09Iek7SEkn3SHqivtgTkt7V1l4ACF26dKnpA0B3kQOB7uk0B5rZgKRHJd0labWke+uD20bf\nlbTG3d8k6UuSfqdVu+2cWW7ciBWSbpX0jKRF7r6v/tZ+1b6iSv3MA2a2xcy2FFkX0G9ajag5swxM\nLHIgUJ2ScuBtkobcfbu7n5f0pGoD28b1fM3dR6/f+Zak9B1vGrQ9wc/M5kj6sqQPufvxxrv1uLub\nWXIv3H2DpA31Nsj2QBOcPQbyRA4EqldCDlwiqXGC1G5JtzdZ/n2S/qJVo22dWTazqaodJD7n7l+p\nhw+Y2eL6+4slpa+MB9C2XK/XAvoZORDojjZy4ODoNzX1xwPjXZeZ/aKkNZL+Xatl26mGYZI+K+k5\nd3+k4a2Nku6X9In6v18d19Z2STRaiWazHjp0KBmPqmFEopm60T3rX/e61yXjV155ZTLerGJEtI6h\noaFk/Ac/+EGh+EsvvZSMRzOKo88i2od+O8taxkzghuu17lRtRL3ZzDa6e2OJktHrtU6b2a+qdr3W\nz3e0YqBH9UoOjDSeIW90xRXpc2lTp05NxqOqF9dcc00yvnRp+pvvaPlIlG8k6eWXX07Go2pXJ06c\nSMbLqnqB5trMgcPuvqbJ+3skNZYVW1qPvYqZvUPSb0n6SXc/12ql7ZxZfqukX5L0U2b2bP3xTtUO\nEHea2YuS3lF/DaADuV6vBfQxciDQJSXkwM2SVpnZSjObJmmdagPby8zsVkm/L+lud2/rG6GWZ5bd\n/RuS0kNP6e3trARAe9oYVQ+OmSi0oX5N5KhKrtcC+hU5EOieTr9ddfcRM1sv6SlJA5Iec/etZvaw\npC3uvlG1yy7mSPrP9W9WXnb3u5u1yx38gIy0MXJu9RVU2xqu1/rJMtoDAKATZVzC4u6bJG0aE3uo\n4fk7irZJZxnIREl3L6rkei0AAKqU8x386CwDGSlhVH35ei3VOsnrJN3XuEDD9Vpr271eCwCAquU6\nObLvO8sjIyPJ+KlTp5Lx3bt3J+NHjhxJxl988cVk/K/+6q+S8auvvjoZnzVrVjIezcaV4pm9w8PD\nyXhUAeTYsWPJ+JkzZ5LxCxcuJOPRiDHXP46JkOv1WgAmt7KqXsyePTsZv+qqq5Lxa6+9NhlfuHBh\nMj5jxoxkPKqmdPz48WRciqteHD16tNA6on4Cuat8nFkG0FRZd+mr4notAACqlPOdauksAxnJdVQN\nAEDVcs2BdJaBjOQ6qgYAoGq55kA6y0Amcp4JDABAlXLOgXSWgYzkOqoGAKBqueZAOsuBaHRz7ly6\nJG107/ho1u2OHTuS8WhmchRvJvpPF+1b0WoVReNoLddRNYDJLaqGMTAwkIxHVSnmzJmTjC9YsCAZ\nj6pkRO1ElSeiXBpVcZKkw4cPJ+NRpaho3VHVKXJd+XLNgXSWgYxw8AUA9KtccyCdZSATOV+vBQBA\nlXLOgXSWgYzkOqoGAKBqueZAOstARnIdVQMAULVccyCdZSATOd+9CACAKuWcA4uXWABQmUuXLjV9\nAADQq8rIgWa21syeN7MhM3sw8f7bzOw7ZjZiZj/XTpucWS5J0TJqdHyQUsao2szWSvqUpAFJn3H3\nT4x5/22Sfk/SmyStc/cvdbxSABMuKg8nxeVHo9Jx06ZNS8ZnzZpVKD59+vRkPCrHFpV1u3DhQjI+\nPDycjEvS8ePHk/GoBGy0DvJ193SaA81sQNKjku6UtFvSZjPb6O7bGhZ7WdJ7Jf1Gu+3SWQYyUcZM\n4KoOFAAAVKmkahi3SRpy9+2SZGZPSrpH0uUc6O476u+1vTI6y0BGSjizXMmBAgCAqrWRAwfNbEvD\n6w3uvqHh9RJJuxpe75Z0e6fbRWcZyEgJo+pKDhQAAFStjRw47O5rurEtjegsAxkpYVQNAMCkVMK3\nq3skLWt4vbQe6widZSATbV6v1WpUXcmBAgCAKpV0zfJmSavMbKVquW+dpPs6bbRlZ9nMlkn6Q0mL\nJLlqZ7I+ZWYfl/TLkl6pL/pRd9/U6QYB/ayEUXUlBwqgX/VKDoyOLc0qaBQxMjKSjEfVLaZMSXc/\npk6dmoxHFSyOHj0abtPhw4eT8ZMnTxZaR9SBK1oFC611+tm5+4iZrZf0lGoVoR5z961m9rCkLe6+\n0czeLOlPJC2Q9DNm9n+6+03N2m3nzPKIpI+4+3fMbK6kb5vZX9bf+6S7//tx7xWAV+l0VF3VgQLo\nY+RAoEvKKNNXH7RuGhN7qOH5ZtW+dW1by86yu++TtK/+/ISZPafaJCIAJSrr7kVVHCiAfkUOBLqj\nZ+7gZ2YrJN0q6Zl6aL2Zfc/MHjOzBcHPPGBmW8ZMSgKQwB38gHyRA4Fq5ZoD2+4sm9kcSV+W9CF3\nPy7p05JukHSLaqPu3039nLtvcPc1E1HqA5hsRkfW0QPAxCAHAtXLNQe2VQ3DzKaqdpD4nLt/RZLc\n/UDD+38g6c8r2UKgT5Q0ExhAyciBQPVyzoHtVMMwSZ+V9Jy7P9IQX1y/lkuS3i3pB9VsItA/OHsM\n5GUy5cBmx4+oE3LhwoVk/PTp08n4oUOHkvGzZ88WWn7WrFnJ+MDAQDJ+8eLFQuuV4n2IqmGcOXMm\nGT9//nyhbaJKxvjl+hm1c2b5rZJ+SdL3zezZeuyjku41s1tUK6WzQ9L7K9lCoI/kOqoG+hg5EOiS\nXHNgO9UwviEpVYgx23qSwGSU81dQQL8iBwLdkXMO5A5+QEZy/QoKAICq5ZoD6SwDGcl1VA0AQNVy\nzYF0loGM5DqqBgCgarnmQDrLQCZyvl4LwOQWdUKiahgjIyPJ+KlTp5LxWtGQ17riivTtHKLly4pL\n8T5Hx9myls+1w5e7nHMgnWUgIxxkAQD9KtccSGcZyEiuo2oAAKqWaw5s+3bXAKrV6jaf7Y64zWyt\nmT1vZkNm9mDi/elm9oX6+8+Y2YqSdwUAgEJyzoF0loGMXLp0qemjFTMbkPSopLskrVbtxgmrxyz2\nPklH3P1GSZ+U9Nsl7wYAAIXlmgPpLAMZKWFUfZukIXff7u7nJT0p6Z4xy9wj6Yn68y9Jers1myUD\nAEAX5JoDu33N8rCknfXng/XX/aTf9nmy7u/1E7Tep9x9sMUyM8xsS8PrDe6+oeH1Ekm7Gl7vlnT7\nmDYuL+PuI2Z2TNLVmpy/K2AymTQ5sOhEqzYrSWS9zxWYrPtLDhyjq51ld184+tzMtrj7mm6uf6L1\n2z732/52yt3XTvQ2AKgOObC/9rnf9rdTOedALsMAesseScsaXi+tx5LLmNkUSfMlHerK1gEAUJ1K\nciCdZaC3bJa0ysxWmtk0SeskbRyzzEZJ99ef/5ykv/Zci1sCANC+SnLgRNZZ3tB6kZ7Tb/vcb/s7\n4erXX62X9JSkAUmPuftWM3tY0hZ33yjps5L+yMyGJB1W7WACoLv68fjYb/vcb/s74arKgcYJJQAA\nACCNyzAAAACAAJ1lAAAAIND1znKr2xD2CjN7zMwOmtkPGmJXmdlfmtmL9X8XTOQ2lsnMlpnZ18xs\nm5ltNbMP1uM9u88AUBQ5sDfzATmwt3W1s9zmbQh7xeOSxtYMfFDS0+6+StLT9de9YkTSR9x9taS3\nSPpA/Xfby/sMAG0jB/Z0PiAH9rBun1lu5zaEPcHdv67aLMtGjbdYfELSu7q6URVy933u/p368xOS\nnlPtLjk9u88AUBA5sEfzATmwt3W7s5y6DeGSLm/DRFrk7vvqz/dLWjSRG1MVM1sh6VZJz6hP9hkA\n2kAO7IN8QA7sPUzwmyD1Atg9V7fPzOZI+rKkD7n78cb3enWfAQDF9Go+IAf2pm53ltu5DWEvO2Bm\niyWp/u/BCd6eUpnZVNUOEp9z96/Uwz29zwBQADmwh/MBObB3dbuz3M5tCHtZ4y0W75f01QncllKZ\nmal2V5zn3P2Rhrd6dp8BoCByYI/mA3Jgb+v6HfzM7J2Sfk//6zaE/7arG9AlZvZ5SXdIGpR0QNLH\nJP2ppC9KWi5pp6T3uPvYCRCTkpn9hKT/Lun7ki7Vwx9V7ZqtntxnACiKHNib+YAc2Nu43TUAAAAQ\nYIIfAAAAEKCzDAAAAAToLAMAAAABOssAAABAgM4yAAAAEKCzDAAAAAToLAMAAAABOssAAABAgM4y\nAAAAEKCzDAAAAAToLAMAAAABOssAAABAgM5yZszsP5rZ/1H2sgAA5I4ciByZu0/0NvQNM9shaZGk\nEUkXJW2T9IeSNrj7pQ7bvkPS/+vuSwv8zHRJn5L0bklTJf0PSb/i7ns62RYAAMbKMAdeqVoOvKse\n+g/u/vFOtgO9iTPL3fcz7j5X0vWSPiHpX0r67ARtywcl/V1Jb5J0naQjkv6fCdoWAEDvyykHflLS\nLEkrJN0m6ZfM7J9M0LYgY3SWJ4i7H3P3jZJ+XtL9ZnazJJnZ42b2b0aXM7P/3cz2mdleM/tnZuZm\ndmPjsmY2W9JfSLrOzE7WH9e1sRkrJT3l7gfc/aykL0i6qex9BQCgUSY58Gck/Y67n3b3Hap12v9p\nybuKHkBneYK5+/+UtFvS3xv7npmtlfRhSe+QdKOkO4I2Tqn2NdJed59Tf+w1s58ws6NNVv9ZSW81\ns+vMbJakX1DtgAMAQOUmOAdKko15fnPxvUCvo7Och72SrkrE3yPpP7n7Vnc/LenjRRp192+4+5VN\nFnlR0i5JeyQdl/RGSQ8XWQcAAB2aqBz4XyU9aGZz62er/6lql2UAr0JnOQ9LJB1OxK9TrTM7aldi\nmU48Kmm6pKslzZb0FXFmGQDQXROVA39N0hnVThx9VdLnVTvLDbwKneUJZmZvVu1A8Y3E2/skNc7s\nXdakqfGUNblF0uPuftjdz6k2ue82MxscR1sAABQykTmwnvt+wd2vdfebVOsT/c+i7aD30VmeIGY2\nz8x+WtKTqpW7+X5isS9K+idm9sb6NcXN6kkekHS1mc0vsBmbJf1jM5tvZlMl/W+qXfM1XKANAAAK\nySEHmtkNZna1mQ2Y2V2SHpD0b1r9HPoPneXu+zMzO6Ha10m/JekRSclSNe7+F5L+b0lfkzQk6Vv1\nt84llv1b1b5C2m5mR+uT9v6emZ1ssi2/Iemsal9BvSLpnarVXAYAoAo55cAfl/R9SSck/V+SfsHd\nt45vt9DLuCnJJGJmb5T0A0nT3X1korcHAIBuIQdionBmOXNm9m4zm25mCyT9tqQ/4yABAOgH5EDk\ngM5y/t4v6aCkH6p2e9BfndjNAQCga8iBmHBchgEAAAAEOLMMAAAABKZ08sP1W1F+StKApM+4+yda\nLM9pbEwK7m6tlyrX2rVrfXi4edW+b3/720+5+9oubRKAJsiB6FXkwFcbd2fZzAZUuwPcnard8Waz\nmW10921lbRzQT4aHh7V58+amy1xxxRXcMAbIwHhz4MDAQNvriC6TNEv3Y8ZzWWXUVtFt6mVlft5l\nrLfMdafWcfHixVLaLirnHNjJZRi3SRpy9+3ufl61wuL3lLNZQH+6dOlS0weAbJADgZKVkQPNbK2Z\nPW9mQ2b2YOL95Wb2NTP7rpl9z8ze2arNTjrLS/Tq+7TvrsfGbtQDZrbFzLZ0sC6g57l7yweAbJAD\ngRKVkQMbvvG5S9JqSfea2eoxi/0rSV9091slrZP0H1q1W/kEP3ff4O5r3H1N1esCJrtcR9UAxocc\nCLSvhBzYzjc+Lmle/fl8SXtbNdrJBL89kpY1vF5ajwEYp07PHrd5HeXoqPrT9RH3JkkrOlox0H/G\nlQPL+IaoG9cmV20yXSs9Uetutt6yrqPO7RvLNrZncMy3NBvcfUPD69Q3PrePaePjkv6bmf1zSbMl\nvaPVSjvpLG+WtMrMVqp2gFgn6b4O2gP6mruXcV3y5VG1JJnZ6Ki6sbNceFQN4DXIgUCJ2syBwyV8\nS3OvpMfd/XfN7O9K+iMzu9ndw5WPu7Ps7iNmtl7SU6qVzXnM3beOtz0A+Y6qAbwaORAoXwlnutv5\nxud9ktbW1/dNM5shaVC1O0UmdVRn2d03qfYVLoAS5DqqBvBa5ECgXCV8u9rONz4vS3q7pMfN7I2S\nZkh6pVmjHXWWAZQr11E1AABV6zQHRt/4mNnDkra4+0ZJH5H0B2b266pdlvheb7FiOstAJkq6ZrmS\nUTUAAFUqKQcmv/Fx94canm+T9NYibdJZBjKS66gaQPeVeee4oncDLMtkqnpRVNFtLfOzLuv3mdvn\nndv2jKKzDGQi51E1AABVKisHVoHOMpCRXEfVAABULdccSGcZyEiuo2oAAKqWaw6kswxkJNdRNQAA\nVcs1B9JZBjKR8/VaAABUKeccSGe5h1U9y7msEeB4tjPX0WenenW/ANSkjnfR333RigdlVskoq52q\nq3CUecwsuk1lVZ6oOldP9LqLyDUH0lkGMpLrqBoAgKrlmgPpLAOZcPdsR9UAAFQp5xxIZxnISK6j\nagAAqpZrDqSzDGQk11E1AABVyzUHXjHRGwCgZnQmcLMHAAC9qKwcaGZrzex5MxsyswcT73/SzJ6t\nP14ws6Ot2uTMckHRzNEoPjAwkIxPmZL+6KdPn56Mz5w5MxmfNWtWMi5JU6dODd9LGRkZScbPnDmT\njJ87d65Q/MKFC4XWO57OYa6j0nZN9u0H0Fzqb7ys6hbNKhtMVNWLoqquSCFNbPWRMtY7nnVXvQ9l\n6XR7zGxA0qOS7pS0W9JmM9vo7tsa1vHrDcv/c0m3tmqXM8tARnIdVQMAULUScuBtkobcfbu7n5f0\npKR7mix/r6TPt2qUM8tAJsooyF7VqBoAgCq1mQMHzWxLw+sN7r6h4fUSSbsaXu+WdHuqITO7XtJK\nSX/daqV0loGMlPCV2OVRtSSZ2eioeluw/L2SPtbpSgEA6FQbOXDY3deUtLp1kr7k7hdbLUhnGchI\nrqNqAACqVsJE9j2SljW8XlqPpayT9IF2GqWzDGQk11E1AABVK+Hb1c2SVpnZStU6yesk3Td2ITP7\nEUkLJH2znUbpLAOZKOOaZVU0qgYAoEpl5EB3HzGz9ZKekjQg6TF332pmD0va4u4b64uuk/Skt9k7\n76izbGY7JJ2QdFHSSIlnvCZc0ZJvM2bMSMbnzJmTjC9atCgZv/7665PxFStWJOPLli1LxiXpqquu\nSsajfTt16lQyvnv37mR8x44dyfjLL7+cjO/fvz8ZP3LkSDJ++vTpZFySzp8/n4xfvJg+SVp1uaOy\n5DqqBvBaE5UDx3M8K1pirWg5s6LxK65IF+OKcmyUt6J2mq07yhNRPOrAFY0X/b2VmZ9yy3WRMrbT\n3TdJ2jQm9tCY1x8v0mYZZ5b/vrsPl9AO0PdyHVUDCJEDgZLkevMtLsMAMuHu2Y6qAQCoUlk5sAqd\n3pTEJf03M/u2mT2QWsDMHjCzLWNm8ANI4HbXwKRCDgRKlGsO7PTM8k+4+x4zu0bSX5rZ37r71xsX\nqJe12iBJZpbnkAHIRK6jagBJ5ECgRLnmwI7OLLv7nvq/ByX9iWo3RAAwDqMzgXMcVQN4LXIgUJ6c\nc+C4zyyb2WxJV7j7ifrzfyDp4dK2rEuiGbbTpk1LxufOnZuMDw4OJuPLly9Pxt/4xjcm4zfddFMy\nfsMNNxRarxRX4oj+w508eTIZ37dvXzK+ePHiZHzBggXJ+KxZs5LxqNrG8HA8Z+bEiRPJeFQlY2Rk\nJGwrJ7mOqgG8Wpk5sGjliWj5ZpUhinY0oraieJQzo+N+VK3puuuuS8ajSlFRHpLifH3hwoVk/ODB\ng8n4zp07k/GoItQrr7ySjEc5Nspb0XZKxaubVN1OWXLNgZ1chrFI0p/UP+gpkv7Y3f9rKVsF9CnO\nHgOTBjkQKFmuOXDcnWV33y7px0rcFqDv5TqqBvBq5ECgfLnmQErHAZko6Q5+AABMOjnnQDrLQEZy\nHVUDAFC1XHMgnWUgI7mOqgEAqFquObCnOsvNZnUWncEbzaJduHBhMr5ixYpkvGjVi2jGb1TZIqoK\nIUmHDx9OxqOZtxcvXkzGo/+88+bNS8avueaaZPz48ePJ+OnTp5Pxs2fPJuOSdO7cuWQ82rfJIOe7\nFwHovqKVCsZz/Ihy45Qp6e5BVN3i6quvTsajnPamN70pGV+zZk0yHuXMJUuWJOOSNH369GT82LFj\nyXhU3eK73/1uMr5lS/o+M9u2bUvG9+7dm4wfPXo0GW/WcezFXFFWDjSztZI+JWlA0mfc/ROJZd4j\n6eOq3Vjo/3P3+5q12VOdZWCyy3VUDQBA1TrNgWY2IOlRSXdK2i1ps5ltdPdtDcuskvSbkt7q7kfq\nNxVqqtPbXQMo0ejIOnq0w8zWmtnzZjZkZg8Gy7zHzLaZ2VYz++NSdwIAgHEoIQfeJmnI3be7+3lJ\nT0q6Z8wyvyzpUXc/Ul9nusB2A84sA5koYyZwVaNqAACqVFI1jCWSdjW83i3p9jHLvF6SzOx/qHap\nxsdb1UinswxkpITrtS6PqiXJzEZH1Y0X0RUeVQMAULU2cuCgmTVeLL7B3TcUXM0USask3SFpqaSv\nm9mPunv64nHRWQaykuuoGgCAqrWRA4fdPT0LtGaPpGUNr5fWY412S3rG3S9IesnMXlCt87w5anRS\ndpajGcLNZg4PDAwk49Fs2aj6RDTjN7qf/dKlS5Px+fPnJ+NRlYcDBw4k47t3707GpbgaRiSqADI4\nOJiMR5/djBkzkvFoFnUUj9qX4t9nNLM7+r+R24ziXEfVAMqR+huPjltFj0/NcmD03tSpU5PxmTNn\nJuOLFy9OxletWpWMv+Utb0nG3/zmNyfjUWWpaHuiyhaSdObMmWQ8qvwU/R4WLVqUjEeVOF555ZVk\n/OTJk8l4VBHqwoULybgU/9+I9q1oZZWJUkJO3ixplZmtVK2TvE7S2EoXfyrpXkn/ycwGVTuBtL1Z\no5Oyswz0ojav15qQUTUAAFUq45pldx8xs/WSnlLtm9PH3H2rmT0saYu7b6y/9w/MbJuki5L+hbsf\natYunWUgI7mOqgEAqFoZ3/a6+yZJm8bEHmp47pI+XH+0hc4ykJFcR9UAAFQt13sN0FkGMlHW3Yuq\nGFUDAFClnO9iS2cZyEiuo2oAAKqWaw7sm85yNMt1ypT0RxBVdJg3b14yvmDBgmQ8qvQQzXKNKlgM\nDQ0l4z/84Q+TcUk6fvx4Mh7tQ1S5I6qSEVWrKFqpIhpJNhth5jr67FSv7heAmug4WLXouBzluqjy\n0+te97pkfM2a9LzjH//xH0/GFy5cmIzv3bs3GX/++eeT8Z07dybjknT+/PlC646qW0Sf3ZVXXpmM\nR9Wupk2bVqj9ZqJcMVmqXkRyzYF901kGclfS3YsAAJh0cs6BdJaBjOQ6qgYAoGq55kA6y0BGch1V\nAwBQtVxzIJ1lICO5jqoBAKharjmQzjKQiZyv1wIAoEo558CWnWUze0zST0s66O4312NXSfqCpBWS\ndkh6j7sfqW4zXy0aeYxnlnH0M0WrZ0SzWaOZqdEs3aNHjybjBw4cSMaj6hlS/HVGNPt5cHAwGY9m\nRUe/h7NnzybjJ06cSMZPnjyZjJ8+fToZl6SRkZFkPNrnXEerY02W7QT6RTdyYNG/+yivNMuBUSWG\nqNrRtddem4yvWrUqGX/DG96QjM+ZMycZ37FjRzL+rW99Kxn/7ne/m4wPDw8n41JcreJHf/RHk/Fo\nn6PPKKp6EeXYon2U8VS2iH5msuSWXLeznd/c45LWjok9KOlpd18l6en6awAdunTpUtMHgK57XORA\noCtyzYEtO8vu/nVJY09f3iPpifrzJyS9q+TtAvrO6N2Lmj0AdBc5EOiOnHPgeKujL3L3ffXn+yUt\nKml7gL6W66gawKuQA4EKlJEDzWytmT1vZkNm9ppvfczsvWb2ipk9W3/8s1ZtdjzBz93dzMLuvpk9\nIOmBTtcD9APOHgOTCzkQKE+nOdDMBiQ9KulOSbslbTazje6+bcyiX3D39e22O94zywfMbHF9wxZL\nOhgt6O4b3H2Nu6fvhQlA0v+aCZzjqBrAq5ADgZKVlANvkzTk7tvd/bykJ1W7bKoj4z2zvFHS/ZI+\nUf/3q51uSBnGMyKJPvwofuHChWT84sWLyXhUtSEyffr0ZDya1RtV1ZCkmTNnJuM33nhjoXi0TXv3\n7k3G9+/fn4zv27cvGY9mM0dVMqR4vyf7pQq5jqoBvEqpOTA6bkXVl6Llm1VPiKphzJ49OxlfvHhx\nMn799dcn4wsWLEjGDx06lIx/+9vfTsa/+c1vJuNR9YzoM5KkhQsXFoovW7YsGY8qekTVq6LqWEX7\nA81EuaIPqmEMmtmWhtcb3H1Dw+slknY1vN4t6fZEOz9rZm+T9IKkX3f3XYllLmundNznJd1R38Dd\nkj6m2gHii2b2Pkk7Jb2nVTsAWiuhs395VC1JZjY6qh7bWQbQBnIg0D1t5MDhEr6l+TNJn3f3c2b2\nftUm6f5Usx9o2Vl293uDt95efPsARNqc7Tsho2qgX5EDge4oqeLFHkmNXxEsrcca19P4dcdnJP1O\nq0a5gx+QkVxH1QAAVK2Eb1c3S1plZitV6ySvk3Rf4wJmtrihms3dkp5r1SidZSAjuY6qAQCoWqc5\n0N1HzGy9pKckDUh6zN23mtnDkra4+0ZJv2Zmd0saUa2G+ntbtUtnGcjE6EzgDlUyqgYAoEol5UC5\n+yZJm8bEHmp4/puSfrNIm33TWS5a3eL06dPJ+IkTJ5LxaFbsqVOnkvGoVj7OIQAAGHxJREFUusV1\n112XjM+aNatQ+1I8Ozma8Tt16tRk/OWXX07GX3rppWT8hz/8YTK+Z8+eZPzw4bE3x6qJfgdSPKs4\n+j330EzgVj9fyagaQDlSf+NR9YSimrUTVY2Icsv8+fOT8blz5ybj0bErqo60ffv2ZPzIkSPJeFS1\n49prr03GJenmm29Oxn/sx34sGb/hhhuS8XPnziXjUf/h+PHjyfjZs2cLtdMsH0S57oor0hWBm1VK\nyUmuubpvOsvAZJDrqBoAgKrlWv6VzjKQkVxH1QAAVC3XHEhnGchEWddrAQAw2eScA+ksAxnJdVQN\nAEDVcs2BdJaBjOQ6qgYAoGq55sCe6iw3+5AvXryYjJ8/fz4Zj6pMRJUbDhw4kIxfffXVyfiSJUuS\n8ahSxetf//pkvJloNnM08/aFF15Ixp9//vlkfNu29B2UoyoZw8PDyXhU9SLaTin+fU5mJd29CEDG\nUlUJyvq7b1bxIKqSEFVBiuJRxY2oYkRUAWLmzJnJ+Ote97pkPKrCsXLlymRckm699dZk/JZbbknG\no8oaUQ48ePBgMl5WrmvWp4l+11F8MuSWnHNgT3WWgcku11E1AABVyzUH0lkGMpLrqBoAgKrlmgPp\nLAOZyHkmMAAAVco5B9JZBjKS66gaAICq5ZoD01f8A5gQly5davoAAKBXlZEDzWytmT1vZkNm9mCT\n5X7WzNzM1rRqs2/OLEcf8sjISDIezeA9duxYMv7KK68k41H1jGi9V155ZTK+fPnyZHz69OnJuBTP\nTv7bv/3bZDyqhhFVvdi5c2cyfuTIkWT8zJkzyXj0WTSreNGrHcdcR9UAylHkbzxadjwVD6JjZtFK\nUVFujNqJctpNN91UqJ158+Yl44sWLUrGJemGG24o9DNRxZAov+/YsSMZP3ToUDIe5cAy81nRHNKs\ngspE6DQHmtmApEcl3Slpt6TNZrbR3beNWW6upA9KeqaddjmzDGRi9HqtHEfVAABUqaQceJukIXff\n7u7nJT0p6Z7Ecv9a0m9LSo/+xqCzDGRktM5k9GilYVR9l6TVku41s9WJ5QqNqgEAqFobOXDQzLY0\nPB4Y08QSSbsaXu+uxy4zs78jaZm7/5d2t6tvLsMAJoMSvo67PKqWJDMbHVWPvZZmdFT9LzpdIQAA\nZWgjBw67+7i/DTWzKyQ9Ium9RX6OM8tAJlqNqCdyVA0AQJXazIGt7JHUeCvkpfXYqLmSbpb0N2a2\nQ9JbJG1sdTkiZ5aBjOQ6qgYAoGolfLu6WdIqM1upWid5naT7Rt9092OSBkdfm9nfSPoNd9/SrFHO\nLAMZyXVUDQBA1TrNge4+Imm9pKckPSfpi+6+1cweNrO7x7tdLc8sm9ljkn5a0kF3v7ke+7ikX5Y0\nWk/lo+6+abwbMZGKltO5cOFCMh6V2Ynaj0rUzJw5MxmPSsQ1+88Tla+JSr5FZXCisjlRGZzos4u2\ntWh5pFbvFVlHTkq6e1Elo2qgX01kDoyOBwMDA8l4s+NcdFw+ffp0Mh6VPT148GAyPnv27GR81qxZ\nyXiU6yLRMT/6LJr9TPS5njhxIhmPcuOePXuS8ePHjyfjUZnUorlRKl4+cDzlBrutrDv41f8WN42J\nPRQse0c7bbZzZvlxSWsT8U+6+y31x6TsKAO5yXVUDfSxx0UOBLqihG9XK9HyzLK7f93MVlS/KQBy\nHVUD/YocCHRPrjcc6+Sa5fVm9j0ze8zMFkQLmdkDozP3O1gX0BdyHVUDeA1yIFCyXHPgeDvLn5Z0\ng6RbJO2T9LvRgu6+wd3XdDKDH+gHo9drdXoHPwCVIwcCJcs5B46rdJy7Hxh9bmZ/IOnPS9sioI9x\n9hjIHzkQqEauOXBcnWUzW+zu++ov3y3pB+VtUjWKzhCNZtjOmDEjGZ8/f34yPjg4mIxfddVVyXi0\nnXv37k3Gjx49moxL0oEDB5Lxffv2JeNnz6ZvkR5V7pgyJf3fJ/rsos86ar8fz6T24z4Dk023cmDR\nCgbNjh9RJaeoAkSUJ+bOnZuMR5Ueohw4derUZDzazige5d5m646qVezevTsZ3759ezIeVYo6d+5c\nMl60UlQz46mgkVK0slTVcs2B7ZSO+7ykO1S7c9huSR+TdIeZ3SLJJe2Q9P4KtxHoCxN9TRaA1yIH\nAt2Rcw5spxrGvYnwZyvYFqDv5TqqBvoVORDonlxzILe7BjKS66gaAICq5ZoD6SwDmSjr7kUAAEw2\nOedAOstARnIdVQMAULVcc2DfdJaLVnSI7mcfza5duXJlMr5ixYpkfM6cOcl4VMFi//79yfiRI0eS\ncUk6c+ZMMh7NBI6qWESfRVQZZNq0acl4NPs5miHcbJbuZLjP/XjkOqoGUJ2yKhKMpxrGyZMnk/Eo\n50RVL6LKEFdeeWUyHuXeaB+ivPL6178+GZfiHHjw4MFkPKp6EVWjij676LMuq4JFM1FfJ0I1jPZ0\ncgc/ACXL9e5FAABUrYwcaGZrzex5MxsyswcT7/+KmX3fzJ41s2+Y2epWbdJZBjJR1t2LqjhQAABQ\npTJyoJkNSHpU0l2SVku6N5Hj/tjdf9Tdb5H0O5IeadUunWUgI52Oqqs6UAAAULUSzizfJmnI3be7\n+3lJT0q6Z8w6Gq9Fna1avfSm+uaaZWAyKOF6rcsHCkkys9EDxbbRBcZzoAAAoGpt5MBBM9vS8HqD\nu29oeL1E0q6G17sl3T62ETP7gKQPS5om6adarZTOMpCJNkfOE3KgAACgSm3mwGF3X1PCuh6V9KiZ\n3SfpX0m6v9nyfdNZjio9RBUdohm8S5YsScaXL1+ejF999dXJ+KlTp5LxF198MRnfsWNHMn7+/Plk\nXIqrWESi2ckzZ85MxqPPLqp6Ef0Ootm7uc3S7YY2RtUTcqAAUI7Uca2sybvN2omqWJw9e7ZQW9Hy\nw8PDyXjRqklRfOnSpcl4s32O8uyuXbuS8ZdeeikZP3z4cDIefaZFf5/jyXXROqK2JksFqRK+Xd0j\naVnD66X1WORJSZ9u1SjXLAMZKeF6rfEcKN7VwSYDAFCKEnLgZkmrzGylmU2TtE7SxsYFzGxVw8t/\nKCl9lrJB35xZBnJX0t2LLh8oVOskr5N0X+MCZrbK3UcPDm0dKAAAqFIZOdDdR8xsvaSnJA1Ieszd\nt5rZw5K2uPtGSevN7B2SLkg6oja+WaWzDGSk06/EqjpQAABQtTIuC3H3TZI2jYk91PD8g0XbpLMM\nZKSMuxdVcaAAAKBqud7Bj84ykJHcJlsAANAtuebAnuosN5tRWrQaxoIFC5Lxa6+9ttDy0Sjp5Zdf\nTsaHhoaS8X379iXj0cxhSbrmmmuS8ai6RVTFouis5aiqRplVL6I/qFz/0NpR0jXLADI2Uceo6Nhy\n4cKFQstH1TCi436UJ6LcOzg4WKidKLdL0rFjx5Lx06dPJ+NRnj158mQyXvR3WTRvNcuNRfPmeNbR\nbTnnwJ7qLAOTXa4HCgAAqpZrDqSzDGSiQGkcAAB6Ss45kM4ykJFcR9UAAFQt1xxIZxnISK6jagAA\nqpZrDqSzDGQi58kNAABUKecc2LKzbGbLJP2hpEWSXNIGd/+UmV0l6QuSVkjaIek97n6kuk3tTFSJ\nIZqRO2/evGR81qxZyXg0I/fMmTPJ+NGjRwstH7U/e/bsZFyKtzWaVRytO5rlXNYs2l6sbDFe/bjP\nQM66kQPLqmwwHlHnJIpHuTRaPtq3KD/NmTMnGZ8/f36h9iXp0KFDyXiUf6PqGVEFkIsXLybjRStP\nRPFmv+eoTxC1FW1rbnLNgen/9a82Iukj7r5a0lskfcDMVkt6UNLT7r5K0tP11wA6cOnSpaYPAF1H\nDgS6JNcc2LKz7O773P079ecnJD0naYmkeyQ9UV/sCUnvqmojgX4xOhs4egDoLnIg0D255sB2zixf\nZmYrJN0q6RlJi9x9tIL3ftW+okr9zANmtsXMtnSwnUDPG71eK8dRNQByIFClsnKgma01s+fNbMjM\nXvONj5l92My2mdn3zOxpM7u+VZttd5bNbI6kL0v6kLsfb3zPa939ZJff3Te4+xp3X9PuuoB+leuo\nGuh35ECgep3mQDMbkPSopLskrZZ0b/2yqUbflbTG3d8k6UuSfqdVu211ls1sqmoHic+5+1fq4QNm\ntrj+/mJJB9tpC0As11E10M/IgUB3lJADb5M05O7b3f28pCdVu2TqMnf/mruP3vP8W5KWtmq0nWoY\nJumzkp5z90ca3too6X5Jn6j/+9V29qJKzWbFRjN4o0oP0UzTqJ3I1KlTk/GFCxcm41F1jug/SbR8\ns/eiqhenTp1KxqPRXDS7dmRkJBm/cOFCoeWb/WH04lnWMs4eN4yq75S0W9JmM9vo7tsaFhsdVZ82\ns19VbVT98x2tGOhRZefAVJ7qheNZlOtmzpyZjEfVLQYHB5PxqErGuXPnwm0qWo3q5MmTyfh4clQR\nZVWWksqrbjIR2syBg2Muadrg7hsaXi+RtKvh9W5Jtzdp732S/qLVStups/xWSb8k6ftm9mw99lHV\nDhBfNLP3Sdop6T1ttAWgiRIOvpdH1ZJkZqOj6sudZXf/WsPy35L0i52uFOhh5ECgS9rIgcNlXdJk\nZr8oaY2kn2y1bMvOsrt/Q1I07Hl7sU0D0Eyuo2qgX5EDge4p4VuWPZKWNbxeWo+9ipm9Q9JvSfpJ\nd4+/nqjjDn5AJtq8e9GEjKoBAKhSSXfw2yxplZmtVK2TvE7SfY0LmNmtkn5f0lp3b2uuAZ1lICO5\njqoBAKhapznQ3UfMbL2kpyQNSHrM3bea2cOStrj7Rkn/TtIcSf+5fr34y+5+d7N26SwDGcl1VA0A\nQNXKmDTp7pskbRoTe6jh+TuKttk3neXoFxDNco3uBX/69OlkPKoMEc34LVoNIxptRRUmJOnIkSPJ\n+M6dO5PxaJ+jmcPHjh1LxqOqGufPn0/Go8+uF2aIF5XrqBpAOVJ/41E1hDKPgUXbKlpBKspd8+bN\nS8avvvrqZPzKK69MxqMKVVFOluKcFlW9iJaP+gmR6LMrWvViPFUyop8ps+JGlXLN+33TWQZyV9L1\nWpWMqgEAqFJZObAKdJaBjOQ6qgYAoGq55kA6y0BGch1VAwBQtVxzIJ1lIBNl3MEPAIDJKOccSGcZ\nyEiuo2oAAKqWaw7sqc5ysxFJVDUimhV76NChZHz//v3JeDSDN6qGsWDBgmQ8mlEcza49eDCu/BW9\n99JLLyXjL774YjIeVc945ZVXkvHoM6UaRmv9uM8AihlP9YyiVRKinDNt2rRkvGg1jCg3Ru1HOfzc\nubhMfFQpo2jVi+hzLfqZFv2sm/0+o/eKbmtucs2BPdVZBiaznGcCAwBQpZxzIJ1lICO5jqoBAKha\nrjmQzjKQkVxH1QAAVC3XHEhnGchIrqNqAACqlmsOTF9VDqDrRq/XavYAAKAXlZUDzWytmT1vZkNm\n9mDi/beZ2XfMbMTMfq6dNuksAxkZrTMZPQAA6FWd5kAzG5D0qKS7JK2WdK+ZrR6z2MuS3ivpj9vd\nrp66DKPZBxmVhDl+/HgyHpUzi0rRRGXUtm/fnoxfc801yficOXOS8WhEdfTo0WRcisvc7du3Lxk/\ncOBAMh59RlH5najET/SZRvqxc1jG2WMzWyvpU5IGJH3G3T8x5v23Sfo9SW+StM7dv9TxSgGMW1nH\nuvGUB4t+ZsqUdPdg6tSpyfj06dOT8agUXCQqBRfljyjfNGsrKmN65syZQuuOfm9RKbiySs01W/dk\nV0IOvE3SkLtvlyQze1LSPZK2jS7g7jvq77W9Ms4sA5loNaKeyFE1AABVKiMHSloiaVfD6931WEd6\n6swyMNnlOqoGAKBqbeTAQTPb0vB6g7tvqHCTJNFZBrLSxsi51YEiNaq+vaTNAwCgMm3kwGF3X9Pk\n/T2SljW8XlqPdYTOMpCJNu9e1OpAAQDApFPSHfw2S1plZitV6ySvk3Rfp422vGbZzJaZ2dfMbJuZ\nbTWzD9bjHzezPWb2bP3xzk43Buh3JVyvVcmoGuhX5ECgezrNge4+Imm9pKckPSfpi+6+1cweNrO7\nJcnM3mxmuyX9I0m/b2ZbW7XbzpnlEUkfcffvmNlcSd82s7+sv/dJd//3bbQx4aIPOZpJe+zYsWQ8\nqgyxc+fOZDyaUTye2a9FRSO0aGZv0Rm/ReNoLddRNdDHSs2BqWN80aoK4znGFs0t0TqiY1RUcSqq\nIHXkyJFkPKpgEX0W0fJSXPUi2reorZMnTxZqP+pXROuNcm+zfBC9V/T3XGafowxlVIRy902SNo2J\nPdTwfLNqJ5La1vLMsrvvc/fv1J+fUK2n3vHMQgCvleuoGuhX5ECge3K910Cha5bNbIWkWyU9I+mt\nktab2T+WtEW1kXd6qAigpZKu16pkVA2AHAhUqawcWIW26yyb2RxJX5b0IXc/LunTkm6QdIukfZJ+\nN/i5B8xsy5gZ/AASch1VA/2OHAhUL9cc2NaZZTObqtpB4nPu/hVJcvcDDe//gaQ/T/1svazVhvpy\nZHugiVxH1UA/IwcC3ZFrDmynGoZJ+qyk59z9kYb44obF3i3pB+VvHtA/Srp7EYASkQOB7sg5B7Zz\nZvmtkn5J0vfN7Nl67KOq3Ub3FkkuaYek91eyhRUrq6LDeO5bD4yV66ga6GOl5sAiuaXo8SCqGNFM\ntD1RdYszZ84k41FliKiyVFQpqmgFqWaifYgUrRRVtIJU0eWbKauKRW4nYXLNgS07y+7+DUmp38qm\nRAxAB3I7cAH9jhwIdE+uOZA7+AGZyHkmMAAAVco5B9JZBjKS66gaAICq5ZoD6SwDGcl1VA0AQNVy\nzYF0loGM5DqqBgCgarnmQDrLQCZyvl4LQHWKVjaIOhTNjh/ROqJ41FYUjypPnDt3LtymIqJ9HhgY\nKPwzZa276nZy7ThWJeccWLzODIDK5FpjEgCAqpWRA81srZk9b2ZDZvZg4v3pZvaF+vvP1G9j3xSd\nZSAjly5davpoRxUHCgAAqtZpDjSzAUmPSrpL0mrV6qGvHrPY+yQdcfcbJX1S0m+3apfOMpCJMu5e\nVNWBAgCAKpV0B7/bJA25+3Z3Py/pSUn3jFnmHklP1J9/SdLbrcW1UHSWgYyUcGa5kgMFAABVayMH\nDprZlobHA2OaWCJpV8Pr3fVYchl3H5F0TNLVzbaLCX5ARtoYOQ+a2ZaG1xvcfUPD69SB4vYxbbzq\nQGFmoweK4XFtNAAAJWgjBw67+5pubEujbneWhyXtrD8fVP8l537b58m6v9dP0HqfcvfBFssMu/va\nrmwNgLINX7p0iRxYoosXL5bZXNkm6+846xzY4v09kpY1vF5aj6WW2W1mUyTNl3SoWaNd7Sy7+8LR\n52a2ZSJGBxOp3/a53/a3UyV1gis5UADoHDmwv/a53/a3UyXlwM2SVpnZStVy3TpJ941ZZqOk+yV9\nU9LPSfprb3FKm2uWgd5y+UBhZtNUO1BsHLPM6IFCavNAAQBA7urXIK+X9JSk5yR90d23mtnDZnZ3\nfbHPSrrazIYkfVjSa6pGjcU1y0APqV+DPHqgGJD02OiBQtIWd9+o2oHij+oHisOqdagBAJj03H2T\npE1jYg81PD8r6R8VaXMiO8sbWi/Sc/ptn/ttf7NQxYECQOn68fjYb/vcb/vbs4xvXwEAAIA0rlkG\nAAAAAnSWAQAAgEDXO8tmttbMnjezITNrOQNxsjKzx8zsoJn9oCF2lZn9pZm9WP93wURuY5nMbJmZ\nfc3MtpnZVjP7YD3es/sMAEWRA3szH5ADe1tXO8tmNiDpUUl3SVot6V4zW93NbeiixyWNrRn4oKSn\n3X2VpKfVRrmSSWRE0kfcfbWkt0j6QP1328v7DABtIwf2dD4gB/awbp9Zvk3SkLtvd/fzkp6UdE+X\nt6Er3P3rqpXlanSPpCfqz5+Q9K6ublSF3H2fu3+n/vyEavUNl6iH9xkACiIH9mg+IAf2tm53lpdI\n2tXwenc91i8Wufu++vP9khZN5MZUxcxWSLpV0jPqk30GgDaQA/sgH5ADew8T/CZI/Y5pPVe3z8zm\nSPqypA+5+/HG93p1nwEAxfRqPiAH9qZud5b3SFrW8HppPdYvDpjZYkmq/3twgrenVGY2VbWDxOfc\n/Sv1cE/vMwAUQA7s4XxADuxd3e4sb5a0ysxWmtk01W6zu7HL2zCRNkq6v/78fklfncBtKZWZmWq3\nUX7O3R9peKtn9xkACiIH9mg+IAf2tq7fwc/M3inp9yQNSHrM3f9tVzegS8zs85LukDQo6YCkj0n6\nU0lflLRc0k5J73H3sRMgJiUz+wlJ/13S9yVdqoc/qto1Wz25zwBQFDmwN/MBObC3cbtrAAAAIMAE\nPwAAACBAZxkAAAAI0FkGAAAAAnSWAQAAgACdZQAAACBAZxkAAAAI0FkGAAAAAv8/Cob9C0Z3EzIA\nAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1eb804146d8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(15,15))\n",
    "for i in range(5):\n",
    "    plt.subplot(5, 2, 2 * i + 1)\n",
    "    plt.imshow(np.reshape(X_prototypes[2 * i], [28, 28]), cmap='gray', interpolation='none')\n",
    "    plt.title('Digit: {}'.format(2 * i))\n",
    "    plt.colorbar()\n",
    "    \n",
    "    plt.subplot(5, 2, 2 * i + 2)\n",
    "    plt.imshow(np.reshape(X_prototypes[2 * i + 1], [28, 28]), cmap='gray', interpolation='none')\n",
    "    plt.title('Digit: {}'.format(2 * i + 1))\n",
    "    plt.colorbar()\n",
    "\n",
    "plt.tight_layout()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}