4_TrainingOnSVHN.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Training Classification Model\n",
    "\n",
    "This script reads the files generated by \"3_Preprocess_SVHN.ipynb\" and trains the Classification model in charge of recognizing number strings in the input images."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from __future__ import absolute_import\n",
    "from __future__ import print_function\n",
    "from __future__ import division\n",
    "\n",
    "import tensorflow as tf\n",
    "import numpy as np\n",
    "import os\n",
    "from six.moves import cPickle as pickle"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Read training data from binary files\n",
    "\n",
    "Randomly crops a 54x54 section of the training images (containing the number), converts the image to gray-scale color and applies random brightness/contrast transformations."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# This function reads one record from the binary files\n",
    "def read_binary(filename_queue):\n",
    "    label_bytes = 6     # Labels contain: Number of digits + 5 digits\n",
    "    img_height = 64\n",
    "    img_width = 64\n",
    "    img_channels = 3\n",
    "    image_bytes = img_height * img_width * img_channels\n",
    "    \n",
    "    # Every record consists of a label followed by the image, with a fixed number of bytes for each.\n",
    "    record_bytes = label_bytes + image_bytes\n",
    "    reader = tf.FixedLengthRecordReader(record_bytes=record_bytes)\n",
    "    key, value = reader.read(filename_queue)\n",
    "    record_bytes = tf.decode_raw(value, tf.int8)\n",
    "\n",
    "    label = tf.cast(tf.slice(record_bytes, [0], [label_bytes]),tf.int32)\n",
    "    image = tf.cast(tf.reshape(tf.slice(record_bytes, [label_bytes], [image_bytes]),[img_height, img_width, img_channels]), tf.float32)\n",
    "    \n",
    "    # This is the size of the crops \n",
    "    height = 54\n",
    "    width = 54\n",
    "\n",
    "    # Randomly crop a [height, width] section of the image.\n",
    "    image = tf.random_crop(image, [height, width, 3])\n",
    "    \n",
    "    # To gray scale\n",
    "    image = tf.image.rgb_to_grayscale(image)\n",
    "    \n",
    "    # Apply random brightness/contrast transformations\n",
    "    if (np.random.random_sample()>0.5):\n",
    "        image = tf.image.random_brightness(image, max_delta=63)\n",
    "        image = tf.image.random_contrast(image,lower=0.2, upper=1.8)\n",
    "    else:\n",
    "        image = tf.image.random_contrast(image,lower=0.2, upper=1.8)\n",
    "        image = tf.image.random_brightness(image, max_delta=63)\n",
    "    \n",
    "    # Subtract mean\n",
    "    image = tf.sub(image, tf.reduce_mean(image))\n",
    "\n",
    "    return image, label   \n",
    "\n",
    "# This function create training batches running processes in parallel\n",
    "def input_pipeline(filenames, batch_size, num_epochs=None):\n",
    "    num_preprocess_threads = 2\n",
    "    \n",
    "    filename_queue = tf.train.string_input_producer(filenames, num_epochs=num_epochs, shuffle=True)\n",
    "    example_list = [read_binary(filename_queue) for _ in range(num_preprocess_threads)]\n",
    "\n",
    "    label, image = read_binary(filename_queue)\n",
    "    min_after_dequeue = 2000\n",
    "    capacity = min_after_dequeue + 3 * batch_size\n",
    "\n",
    "    image_batch, label_batch = tf.train.shuffle_batch_join(\n",
    "        example_list, batch_size=batch_size, capacity=capacity,\n",
    "        min_after_dequeue=min_after_dequeue)\n",
    "\n",
    "    return image_batch, label_batch"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Reading validation and test sets from Pickle files\n",
    "\n",
    "Read data from picke files and return a tensor with the corresponding transformations"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "def load_valid_test(num_valid, num_test, filename):\n",
    "    with open(filename, 'rb') as f:\n",
    "        save = pickle.load(f)\n",
    "        valid_dataset = save['valid_dataset']\n",
    "        valid_labels = save['valid_labels']\n",
    "        test_dataset = save['test_dataset']\n",
    "        test_labels = save['test_labels']\n",
    "        del save\n",
    "    \n",
    "    height = 54\n",
    "    width = 54\n",
    "    for i in range(num_valid):\n",
    "        temp = tf.cast(tf.constant(valid_dataset[i,:,:,:]),tf.float32)\n",
    "        # Randomly crop a [height, width] section of the image.\n",
    "        temp = tf.random_crop(temp, [height, width, 3])    \n",
    "        # To gray scale\n",
    "        temp = tf.image.rgb_to_grayscale(temp)    \n",
    "        # Subtract mean\n",
    "        temp = tf.sub(temp, tf.reduce_mean(temp))\n",
    "        if i==0:\n",
    "            result = tf.expand_dims(temp, 0)\n",
    "        else:\n",
    "            result = tf.concat(0, [result, tf.expand_dims(temp, 0)])\n",
    "\n",
    "    valid_dataset = result\n",
    "    valid_labels = tf.cast(valid_labels[:num_valid],tf.int32)\n",
    "    \n",
    "    for i in range(num_test):\n",
    "        temp = tf.cast(tf.constant(test_dataset[i,:,:,:]),tf.float32)\n",
    "        # Randomly crop a [height, width] section of the image.\n",
    "        temp = tf.random_crop(temp, [height, width, 3])    \n",
    "        # To gray scale\n",
    "        temp = tf.image.rgb_to_grayscale(temp)    \n",
    "        # Subtract mean\n",
    "        temp = tf.sub(temp, tf.reduce_mean(temp))\n",
    "        if i==0:\n",
    "            result = tf.expand_dims(temp, 0)\n",
    "        else:\n",
    "            result = tf.concat(0, [result, tf.expand_dims(temp, 0)])\n",
    "\n",
    "    test_dataset = result\n",
    "    test_labels = tf.cast(test_labels[:num_test],tf.float32)\n",
    "    \n",
    "    return valid_dataset, valid_labels, test_dataset, test_labels"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Compute classification accuracy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def accuracy(predictions, labels):\n",
    "    return (100.0 * np.sum(np.argmax(predictions, 2).T == labels) / predictions.shape[1] / predictions.shape[0])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Define network architecture"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Ready\n"
     ]
    }
   ],
   "source": [
    "image_size = 54\n",
    "num_labels = 11  # 0-9, + blank \n",
    "num_channels = 1 # grayscale\n",
    "num_digits = 5\n",
    "\n",
    "batch_size = 64\n",
    "patch_size = 3\n",
    "depth1 = 32\n",
    "depth2 = 48\n",
    "depth3 = 64\n",
    "depth4 = 96\n",
    "flat_size = 1536\n",
    "fc_size = 3136\n",
    "\n",
    "# Binary files where the training data was stored\n",
    "file_names = [os.path.join('SVHN_data/','Train%d_64.bin' % i) for i in range(1,4)]\n",
    "# Pickle file where validation and testing were stored\n",
    "valid_test_file_name = 'SVHN_data/SVHN64.pickle'\n",
    "# Number of Valid/Testing samples to load into memory\n",
    "num_valid = 1000\n",
    "num_test = 2000\n",
    "\n",
    "graph = tf.Graph()\n",
    "with graph.as_default():\n",
    "    # Input data.   \n",
    "    tf_train_dataset, tf_train_labels = input_pipeline(file_names, batch_size, num_epochs=None)   \n",
    "    tf_valid_dataset, tf_valid_labels, tf_test_dataset, test_labels = load_valid_test(num_valid, num_test, valid_test_file_name)\n",
    "        \n",
    "    # Variables.\n",
    "    conv1_1weights = tf.get_variable(\"W1_1\", shape=[patch_size, patch_size, num_channels, depth1],\\\n",
    "                                    initializer=tf.contrib.layers.xavier_initializer_conv2d())\n",
    "    conv1_1biases = tf.Variable(tf.constant(1.0, shape=[depth1]), name='B1_1')\n",
    "    conv1_2weights = tf.get_variable(\"W1_2\", shape=[patch_size, patch_size, depth1, depth1],\\\n",
    "                                    initializer=tf.contrib.layers.xavier_initializer_conv2d())\n",
    "    conv1_2biases = tf.Variable(tf.constant(1.0, shape=[depth1]), name='B1_2')\n",
    "    \n",
    "    conv2_1weights = tf.get_variable(\"W2_1\", shape=[patch_size, patch_size, depth1, depth2],\\\n",
    "                                    initializer=tf.contrib.layers.xavier_initializer_conv2d())\n",
    "    conv2_1biases = tf.Variable(tf.constant(1.0, shape=[depth2]), name='B2_1')\n",
    "    conv2_2weights = tf.get_variable(\"W2_2\", shape=[patch_size, patch_size, depth2, depth2],\\\n",
    "                                    initializer=tf.contrib.layers.xavier_initializer_conv2d())\n",
    "    conv2_2biases = tf.Variable(tf.constant(1.0, shape=[depth2]), name='B2_2')\n",
    "    \n",
    "    conv3_1weights = tf.get_variable(\"W3_1\", shape=[patch_size, patch_size, depth2, depth3],\\\n",
    "                                    initializer=tf.contrib.layers.xavier_initializer_conv2d())\n",
    "    conv3_1biases = tf.Variable(tf.constant(1.0, shape=[depth3]), name='B3_1')\n",
    "    conv3_2weights = tf.get_variable(\"W3_2\", shape=[patch_size, patch_size, depth3, depth3],\\\n",
    "                                     initializer=tf.contrib.layers.xavier_initializer_conv2d())\n",
    "    conv3_2biases = tf.Variable(tf.constant(1.0, shape=[depth3]), name='B3_2')\n",
    "    \n",
    "    conv4_1weights = tf.get_variable(\"W4_1\", shape=[patch_size, patch_size, depth3, depth4],\\\n",
    "                                    initializer=tf.contrib.layers.xavier_initializer_conv2d())\n",
    "    conv4_1biases = tf.Variable(tf.constant(1.0, shape=[depth4]), name='B4_1')\n",
    "    conv4_2weights = tf.get_variable(\"W4_2\", shape=[patch_size, patch_size, depth4, depth4],\\\n",
    "                                     initializer=tf.contrib.layers.xavier_initializer_conv2d())\n",
    "    conv4_2biases = tf.Variable(tf.constant(1.0, shape=[depth4]), name='B4_2')\n",
    "\n",
    "    full_weights = tf.get_variable(\"FW1\", shape=[flat_size, fc_size],\\\n",
    "                                    initializer=tf.truncated_normal_initializer(stddev=0.05, dtype=tf.float32 ))\n",
    "    full_biases = tf.Variable(tf.constant(1.0, shape=[fc_size]), name='FB1')\n",
    "    \n",
    "    s1_w = tf.get_variable(\"WS1\", shape=[fc_size, num_labels],\\\n",
    "                           initializer=tf.contrib.layers.xavier_initializer())\n",
    "    s1_b = tf.Variable(tf.constant(1.0, shape=[num_labels]), name='BS1')\n",
    "    s2_w = tf.get_variable(\"WS2\", shape=[fc_size, num_labels],\\\n",
    "                           initializer=tf.contrib.layers.xavier_initializer())\n",
    "    s2_b = tf.Variable(tf.constant(1.0, shape=[num_labels]), name='BS2')\n",
    "    s3_w = tf.get_variable(\"WS3\", shape=[fc_size, num_labels],\\\n",
    "                           initializer=tf.contrib.layers.xavier_initializer())\n",
    "    s3_b = tf.Variable(tf.constant(1.0, shape=[num_labels]), name='BS3')\n",
    "    s4_w = tf.get_variable(\"WS4\", shape=[fc_size, num_labels],\\\n",
    "                           initializer=tf.contrib.layers.xavier_initializer())\n",
    "    s4_b = tf.Variable(tf.constant(1.0, shape=[num_labels]), name='BS4')\n",
    "    s5_w = tf.get_variable(\"WS5\", shape=[fc_size, num_labels],\\\n",
    "                           initializer=tf.contrib.layers.xavier_initializer())\n",
    "    s5_b = tf.Variable(tf.constant(1.0, shape=[num_labels]), name='BS5')\n",
    "        \n",
    "    # Model.\n",
    "    def model(data, keep_prob):\n",
    "        # First Conv stage\n",
    "        conv1_1 = tf.nn.conv2d(data, conv1_1weights, [1,1,1,1], 'SAME', name='C1_1')\n",
    "        hidden1_1 = tf.nn.relu(conv1_1 + conv1_1biases)  \n",
    "        conv1_2 = tf.nn.conv2d(hidden1_1, conv1_2weights, [1,1,1,1], 'SAME', name='C1_2')\n",
    "        hidden1_2 = tf.nn.relu(conv1_2 + conv1_2biases)\n",
    "            \n",
    "        lrn_1 = tf.nn.local_response_normalization(hidden1_2)\n",
    "        # Pooling\n",
    "        pool_1 = tf.nn.max_pool(lrn_1, [1,2,2,1], [1,2,2,1], 'SAME', name='Pool1')\n",
    "        \n",
    "        # Second Conv stage\n",
    "        conv2_1 = tf.nn.conv2d(pool_1, conv2_1weights, [1,1,1,1], padding='SAME', name='C2_1')\n",
    "        hidden2_1 = tf.nn.relu(conv2_1 + conv2_1biases)\n",
    "        conv2_2 = tf.nn.conv2d(hidden2_1, conv2_2weights, [1,1,1,1], 'SAME', name='C2_2')\n",
    "        hidden2_2 = tf.nn.relu(conv2_2 + conv2_2biases)\n",
    "        \n",
    "        lrn_2 = tf.nn.local_response_normalization(hidden2_2)\n",
    "        # Pooling\n",
    "        pool_2 = tf.nn.max_pool(lrn_2, [1,2,2,1], [1,2,2,1], 'SAME', name='Pool2')\n",
    "        \n",
    "        conv3_1 = tf.nn.conv2d(pool_2, conv3_1weights, [1,1,1,1], padding='SAME', name='C3_1')\n",
    "        hidden3_1 = tf.nn.relu(conv3_1 + conv3_1biases)\n",
    "        conv3_2 = tf.nn.conv2d(hidden3_1, conv3_2weights, [1,1,1,1], padding='SAME', name='C3_2')\n",
    "        hidden3_2 = tf.nn.relu(conv3_2 + conv3_2biases)\n",
    "        \n",
    "        lrn_3 = tf.nn.local_response_normalization(hidden3_2)\n",
    "        # Pooling\n",
    "        pool_3 = tf.nn.max_pool(lrn_3, [1,2,2,1], [1,2,2,1], 'SAME', name='Pool3')\n",
    "        \n",
    "        conv4_1 = tf.nn.conv2d(pool_3, conv4_1weights, [1,1,1,1], padding='SAME', name='C4_1')\n",
    "        hidden4_1 = tf.nn.relu(conv4_1 + conv4_1biases)\n",
    "        conv4_2 = tf.nn.conv2d(hidden4_1, conv4_2weights, [1,1,1,1], padding='SAME', name='C4_2')\n",
    "        hidden4_2 = tf.nn.relu(conv4_2 + conv4_2biases)\n",
    "        \n",
    "        lrn_4 = tf.nn.local_response_normalization(hidden4_2)\n",
    "        # Pooling\n",
    "        pool_4 = tf.nn.max_pool(lrn_4, [1,2,2,1], [1,2,2,1], 'SAME', name='Pool4')\n",
    "\n",
    "        shape = int(np.prod(pool_4.get_shape()[1:]))\n",
    "        flat_layer = tf.reshape(pool_4,[-1,shape])\n",
    "        hidden_fc = tf.nn.relu(tf.matmul(flat_layer, full_weights) + full_biases)\n",
    "        \n",
    "        # Dropout\n",
    "        hidden_dout = tf.nn.dropout(hidden_fc, keep_prob)\n",
    "        \n",
    "        logits1 = tf.matmul(hidden_dout, s1_w) + s1_b\n",
    "        logits2 = tf.matmul(hidden_dout, s2_w) + s2_b\n",
    "        logits3 = tf.matmul(hidden_dout, s3_w) + s3_b\n",
    "        logits4 = tf.matmul(hidden_dout, s4_w) + s4_b\n",
    "        logits5 = tf.matmul(hidden_dout, s5_w) + s5_b\n",
    "            \n",
    "        return [logits1, logits2, logits3, logits4, logits5]\n",
    "    \n",
    "    # Get output for the current batch using dropout = 0.9375\n",
    "    [logits1, logits2, logits3, logits4, logits5] = model(tf_train_dataset, 0.9375)\n",
    "    \n",
    "    # Define Loss\n",
    "    loss = tf.reduce_mean(tf.nn.sparse_softmax_cross_entropy_with_logits(logits1, tf_train_labels[:,1])) +\\\n",
    "        tf.reduce_mean(tf.nn.sparse_softmax_cross_entropy_with_logits(logits2, tf_train_labels[:,2])) +\\\n",
    "        tf.reduce_mean(tf.nn.sparse_softmax_cross_entropy_with_logits(logits3, tf_train_labels[:,3])) +\\\n",
    "        tf.reduce_mean(tf.nn.sparse_softmax_cross_entropy_with_logits(logits4, tf_train_labels[:,4])) +\\\n",
    "        tf.reduce_mean(tf.nn.sparse_softmax_cross_entropy_with_logits(logits5, tf_train_labels[:,5]))\n",
    "    \n",
    "    # Optimizer.\n",
    "    global_step = tf.Variable(0, trainable = False)\n",
    "    learning_rate = tf.train.exponential_decay(0.002, global_step, 5000, 0.95)\n",
    "    optimizer = tf.train.AdagradOptimizer(learning_rate).minimize(loss, global_step=global_step)       \n",
    "    \n",
    "    # Predictions for the training, validation, and test data.\n",
    "    train_prediction = tf.pack([tf.nn.softmax(model(tf_train_dataset, 1.0)[0]),\\\n",
    "                                tf.nn.softmax(model(tf_train_dataset, 1.0)[1]),\\\n",
    "                                tf.nn.softmax(model(tf_train_dataset, 1.0)[2]),\\\n",
    "                                tf.nn.softmax(model(tf_train_dataset, 1.0)[3]),\\\n",
    "                                tf.nn.softmax(model(tf_train_dataset, 1.0)[4])])\n",
    "    \n",
    "    valid_prediction = tf.pack([tf.nn.softmax(model(tf_valid_dataset, 1.0)[0]),\\\n",
    "                                tf.nn.softmax(model(tf_valid_dataset, 1.0)[1]),\\\n",
    "                                tf.nn.softmax(model(tf_valid_dataset, 1.0)[2]),\\\n",
    "                                tf.nn.softmax(model(tf_valid_dataset, 1.0)[3]),\\\n",
    "                                tf.nn.softmax(model(tf_valid_dataset, 1.0)[4])])\n",
    "    \n",
    "    test_prediction = tf.pack([tf.nn.softmax(model(tf_test_dataset, 1.0)[0]),\\\n",
    "                                tf.nn.softmax(model(tf_test_dataset, 1.0)[1]),\\\n",
    "                                tf.nn.softmax(model(tf_test_dataset, 1.0)[2]),\\\n",
    "                                tf.nn.softmax(model(tf_test_dataset, 1.0)[3]),\\\n",
    "                                tf.nn.softmax(model(tf_test_dataset, 1.0)[4])])\n",
    "    \n",
    "    valid_loss = tf.reduce_mean(tf.nn.sparse_softmax_cross_entropy_with_logits(model(tf_valid_dataset, 1.0)[0], tf_valid_labels[:,1])) + \\\n",
    "        tf.reduce_mean(tf.nn.sparse_softmax_cross_entropy_with_logits(model(tf_valid_dataset, 1.0)[1], tf_valid_labels[:,2])) +\\\n",
    "        tf.reduce_mean(tf.nn.sparse_softmax_cross_entropy_with_logits(model(tf_valid_dataset, 1.0)[2], tf_valid_labels[:,3])) +\\\n",
    "        tf.reduce_mean(tf.nn.sparse_softmax_cross_entropy_with_logits(model(tf_valid_dataset, 1.0)[3], tf_valid_labels[:,4])) +\\\n",
    "        tf.reduce_mean(tf.nn.sparse_softmax_cross_entropy_with_logits(model(tf_valid_dataset, 1.0)[4], tf_valid_labels[:,5]))\n",
    "    \n",
    "    # Save logs\n",
    "    tf.scalar_summary('loss', loss)\n",
    "    tf.scalar_summary('valid_loss', valid_loss)\n",
    "    tf.scalar_summary('learning rate', learning_rate) \n",
    "    saver = tf.train.Saver(tf.all_variables())\n",
    "    summary_op = tf.merge_all_summaries()\n",
    "    init = tf.initialize_all_variables()\n",
    "    \n",
    "    print('Ready')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Train the network"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model restored.\n",
      "Initialized\n",
      "Minibatch loss at step 0: 1.204979\n",
      "Minibatch accuracy: 91.6%\n",
      "Validation accuracy: 89.5%\n",
      "Minibatch loss at step 500: 2.072063\n",
      "Minibatch accuracy: 89.1%\n",
      "Minibatch loss at step 1000: 2.023313\n",
      "Minibatch accuracy: 88.1%\n",
      "Validation accuracy: 89.8%\n",
      "Minibatch loss at step 1500: 1.977567\n",
      "Minibatch accuracy: 89.1%\n",
      "Minibatch loss at step 2000: 1.346531\n",
      "Minibatch accuracy: 93.8%\n",
      "Validation accuracy: 89.8%\n",
      "Minibatch loss at step 2500: 1.831203\n",
      "Minibatch accuracy: 89.4%\n",
      "Minibatch loss at step 3000: 1.722450\n",
      "Minibatch accuracy: 88.4%\n",
      "Validation accuracy: 89.5%\n",
      "Minibatch loss at step 3500: 1.986029\n",
      "Minibatch accuracy: 86.2%\n",
      "Minibatch loss at step 4000: 1.580312\n",
      "Minibatch accuracy: 91.6%\n",
      "Validation accuracy: 88.6%\n",
      "Minibatch loss at step 4500: 1.987525\n",
      "Minibatch accuracy: 89.1%\n",
      "Minibatch loss at step 5000: 1.971294\n",
      "Minibatch accuracy: 90.3%\n",
      "Validation accuracy: 89.7%\n",
      "Minibatch loss at step 5500: 2.010910\n",
      "Minibatch accuracy: 86.9%\n",
      "Minibatch loss at step 6000: 2.033176\n",
      "Minibatch accuracy: 87.2%\n",
      "Validation accuracy: 90.1%\n",
      "Minibatch loss at step 6500: 2.312181\n",
      "Minibatch accuracy: 85.6%\n",
      "Minibatch loss at step 7000: 1.979577\n",
      "Minibatch accuracy: 89.4%\n",
      "Validation accuracy: 89.9%\n",
      "Minibatch loss at step 7500: 1.580547\n",
      "Minibatch accuracy: 92.8%\n",
      "Minibatch loss at step 8000: 1.976210\n",
      "Minibatch accuracy: 87.2%\n",
      "Validation accuracy: 90.5%\n",
      "Minibatch loss at step 8500: 1.977649\n",
      "Minibatch accuracy: 88.4%\n",
      "Minibatch loss at step 9000: 1.671758\n",
      "Minibatch accuracy: 90.0%\n",
      "Validation accuracy: 90.1%\n",
      "Minibatch loss at step 9500: 1.694518\n",
      "Minibatch accuracy: 90.0%\n",
      "Validation accuracy: 89.8%\n",
      "Test accuracy: 89.2%\n",
      "Model saved in file: Models/ClassificationModel.ckpt\n",
      "Total number of training steps: 70002\n"
     ]
    }
   ],
   "source": [
    "num_steps = 10000                               # Number of training steps\n",
    "savename = \"Models/ClassificationModel.ckpt\"    # Where to save the final model\n",
    "with tf.Session(graph=graph) as session:\n",
    "    session.run(init)  \n",
    "    summary_writer = tf.train.SummaryWriter('TensorBoard/svhn_train',session.graph)\n",
    "    \n",
    "    # These are required to start the processes reading training batches\n",
    "    coord = tf.train.Coordinator()\n",
    "    threads = tf.train.start_queue_runners(sess=session, coord=coord)\n",
    "    \n",
    "    # Load existing models if available\n",
    "    if os.path.exists(savename):\n",
    "        reader = tf.train.NewCheckpointReader(savename)\n",
    "        reader.get_variable_to_shape_map()\n",
    "        saver.restore(session, savename)\n",
    "        print('Model restored.')\n",
    "\n",
    "    print('Initialized')\n",
    "    \n",
    "    for step in range(num_steps):\n",
    "        _, l, predictions, batch_labels = session.run([optimizer, loss, train_prediction, tf_train_labels])\n",
    "        \n",
    "        # Save logs\n",
    "        if step%100 == 50:\n",
    "            summary_str, gbl_step = session.run([summary_op, global_step])\n",
    "            summary_writer.add_summary(summary_str, gbl_step)\n",
    "        \n",
    "        # Prints minibatch accuracy\n",
    "        if (step % 500 == 0): \n",
    "            print('Minibatch loss at step %d: %f' % (step, l))\n",
    "            print('Minibatch accuracy: %.1f%%' % accuracy(predictions, batch_labels[:,1:]))\n",
    "           \n",
    "        # Prints validation accuracy and saves current model\n",
    "        if step%1000 == 0 or (step+1) == num_steps:\n",
    "            saver.save(session, savename)\n",
    "            val_prediction, valid_label = session.run([valid_prediction, tf_valid_labels])\n",
    "            print('Validation accuracy: %.1f%%' % accuracy(val_prediction, valid_label[:,1:]))    \n",
    "    \n",
    "    # Gets testing set and its labels\n",
    "    test_dataset, labels_test, prediction_test = session.run([tf_test_dataset, test_labels, test_prediction])\n",
    "    # Accuracy on testing set\n",
    "    print('Test accuracy: %.1f%%' % accuracy(prediction_test, labels_test[:,1:]))\n",
    "    # Save final model\n",
    "    save_path = saver.save(session, savename)\n",
    "    print(\"Model saved in file: %s\" % save_path) \n",
    "    \n",
    "    gbl_step = session.run(global_step)\n",
    "    print(\"Total number of training steps: %d\" % gbl_step)    \n",
    "    coord.request_stop()\n",
    "    coord.join(threads)\n",
    "\n",
    "    session.close()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Performance on Testing set\n",
    "\n",
    "Randomly selects 10 testing samples and plots them with their corresponding label."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABHsAAACJCAYAAAC8RDTvAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzsvXmQZdd5H/Y7b3+v92V6pqd7BrMAswEzADkECZAgIIEA\nCLJIU6QoyY5NxpGiJZEdlV0quxKV4yhO7LhKTiqSHUdKWVSpHEV0SIqLQgIQCIAgQYADDDBYBtMA\nBoOenq33vd/+3skfr7+L7339nXvf636vBwPcX9Wrt917zneWbz3fOddYaxEiRIgQIUKECBEiRIgQ\nIUKECBHi/YHI9SYgRIgQIUKECBEiRIgQIUKECBEiROsQBntChAgRIkSIECFChAgRIkSIECHeRwiD\nPSFChAgRIkSIECFChAgRIkSIEO8jhMGeECFChAgRIkSIECFChAgRIkSI9xHCYE+IECFChAgRIkSI\nECFChAgRIsT7CGGwJ0SIECFChAgRIkSIECFChAgR4n2EMNgTIkSIECFChAgRIkSIECFChAjxPsIH\nJthjjKkaY1aMMf+iTeWfN8YUjDF/3o7yQ4RjeKOh3ePVJC2/uk5L1Rhz4HrTc6Mj5MX3B8JxvPER\njuGNhVAvfvDQ7JiH43J9EcrUGx/hGNbjAxPsAWABnLDW/jP6wRjzx8aYMWNMxRjz1UYKMcbEjTHn\njDETdYVbezOAf9lakkMIbHkMjTEfNsb8aF0IXDPG/EOv8HAMW4228py45p71MV1ef5Gh9EUAsNb+\nqbW2a52mEFtH3dgaYwaMMT8xxswaYxaNMS8aY37BdbMx5neNMa+uj9XbxpjfrSs85MXtwgYeJRhj\nvrrOQ7/qutkY88+NMUXGc8vGmH1e4eE4bgckL95ijPm2MWZ6nR9/YIw55LrZGBM1xvzRuj6cNcZ8\nxxgz7BUejmGroenFe4wxp4wxS+tOxK/7FWCM+dfrYzVjjPlfAq516txQL24btDGPGGP+J2PMlXW5\nedoY0w2E4/IewJZs11AvviewVdvm54wxT6zbsxc2FH6DjeEHKdhj1l8cZwD8VwBON1HOPwEw1Sqi\nQjSFLY2hMWYAwA8A/HsAfQBuBvBYi2kM8S62jeestT+x1nZZa7uttd0APgdgBcAjCk0htg45tqsA\nfg3AkLW2F8DvA/hPxphOnzK+AqAXwGcA/ANjzC+3i9gQTmg8CmNML4D/FsBrDZTxl+t8R/w33mIa\nQ/hDjmEvgO8AOARgJ4Dn17+78NsA7gFwHMBuAIsA/qgtlIYAxHgZYyIAvgXgT6y1PQD+NoD/1Rhz\nXL3ZmN8E8LdQG68TAD5vjPkNn/oa0bmhXmwvNDn7PwK4C8DH1m2WrwDIK/eF2H60wnYN9eL1xVZt\nmzUA/wHA7wZcd0PggxTs2QBr7b+31j4JoNDI9caY/QD+MwD/qq2EhWgYTY7hPwbwiLX2L621ZWvt\nmrX2jTaTGIJhG3nu7wP4hrU21+R9ITYBa23BWvuGtbZqjDEAqgBmARQd1/+BtfaMtbZqrX0TNWf0\nE9tIcgh//CsA/zuAuetNSIjmYK193lr7NWvtorW2AuB/A3DYGNPnuOVWAI9aa2ettUUAX1//LcT2\nYCeAAQD/EQCstS8AOAfgmOP6rwL4N9baa9baawD+ADV9p6JZnRui/Vh3OH8HwK9bay8DgLX29XX+\nC/EeRMhH7xs0ZNus69H/G8A720JVm/GBDvZsAn+IWkRQRt9D3Bi4C8CCMeYZY8zUerr6nutNVAhf\nNM1zxpgMgF8E8GdtoimEA8aYl1Ebq68B+GITxusnAZxtG2EhGoYx5qMATlpr/88Gb/n8+paSV40x\nv9VO2kJsCvcBuGatXXD8/xiAzxhjhtdl598F8P1to+4DjvWAzSsAfnV9a8/dAPYC+InjllsBvMy+\nv4wwOHej4TiAEoBfWt8+OWaM+a+vN1EhWopQL77HsAnb5n2DMNjTINbP/ohYa797vWkJsWmMorYq\n9g8B7AEwDuD/uZ4EhXBjCzz3iwBmrLU/bgNZIXxgrb0dQBdq27i+aYzpCLrHGPP7qKXbfq3N5IUI\nwPqWkn+H2taeRvB1AEcB7ADwGwD+e2PMr7SJvBBNwhgzCuDfAvhHrmustd8E8BKAK6ht4ToC4Lof\nHvwBw2+gJjMLAH4E4PestVcc13YCWGLfl9d/C3HjYBS17Za3ALgJwC8B+B+MMZ+6rlSFaBVCvfge\nwyZsm/cVwmBPA1hf7frXAP4b+uk6khNi88gB+Ctr7YvrGQe/D+Djxpiu60xXCIEt8txXAdwQJ+S/\nH2GtLVpr/wi1M5N8jVdjzD8A8PcAfNZaW9oO+kL44rcBvGytfb6Ri621Y9baSVvDs6ilR3+5rRSG\naAjGmB0AHgXwb621/8nnuj8A0I3aOXYdAP4KG886C9EmGGN2A/hrAH/HWhtHLUvnnxpjPuO4ZRW1\n8SL0rP8W4sZBDrUDZH9/XV++CuAvAXz2+pIVohUI9eJ7Ek3ZNu83hMGexkDR9x8bY64B+CaA3caY\nq8aYvdeXtBBN4BVsfLpB+LSD9yY2xXPrK9k/hzDY815ADEDW9ef6kxD+CYD717cyhLj+uB/AF9e3\nFlwD8HEA/8YY84cN3m8RLoZcd6yfCfIogG9ba32f1gTgYQBfs9YurQdc/wjAR40x/e2mMwSAGo9d\nttY+DgDW2rcA/H+oHVyv4SyA29n3OxBugb3R8IryW2iLvn8R6sXrj63aNjc0PtDBHlN7pHMKNSZM\nGGOS64eLSryK2rafO1BTsv8lgMn1z5e2i94QG9HEGALr54gYY04YY+IA/hmAn1hrV7aL3g86toHn\nvgrgGWvt++JQtRsFxpiPGWM+QeNrjPmnAFIAnnNc/3cB/M8AHrTWXtxOWkP44j9HLf389vXXC6hl\nQP6edrEx5m+tBxZoP/zvAPj29pAaQsN6pupjqOk2ddwEXgHwVWNM97pe/G0AV6y18+2kM4SHs6gd\noP3zAGCMOYja0yRfdlz/5wD+sTFmtzFmBLUHTzi3wDZpI4XYBlhrLwD4MYDfM8YkjDFHUXsK2/eu\nL2UhXGiGj0K9+J5Es7aNMcYkASQARNbHO75dxLYaH+hgD2oGURbA3QD+eP3zJ+VF60+MmaYXgHkA\nVWvtjLU2jMZfXzQ0hgCwfpL+f4fa4ZOTAA6g9qSnENuHdvPc30N4MPP1QBK1/dCzACYA3AvgYWut\na3vBvwDQD+B5Y8yKMWbZGPN/bA+pIVyw1i4LvisAWPYJiP9tAOeNMcuo8d2/tNb+x20iN4SOLwI4\nCeC/WOct4q9Rx/X/CLWV57cBTKGW6fPF7SE1hLX2HGqPdP53xpglAE8C+H+ttf/Bcf0foxYUeBW1\ngNB3rbX/l08VDdtIIbYVfwfAPtSeCvQ91M5peup6EhTCF83wUagX32PYhG1zL2rbLf8atYXnLGrZ\nsjckzAclVmGMyaI2uH9orf3nbSh/DMBuAF+31v56q8sPEY7hjYZ2j1eTtPx91B5BnABwq7V2/HrS\nc6Mj5MX3B8JxvPERjuGNhVAvfvDQ7JiH43J9EcrUGx/hGNbjAxPsCREiRIgQIUKECBEiRIgQIUKE\n+CDgg76NK0SIECFChAgRIkSIECFChAgR4n2FWLsr+NM//VMLAHSOlbXWe3EYY2CMQSQSQSQSAb+H\nQPfQe7Va3XC/37lzVG+1Wt1QllaOLI/TrWVE8d9kG+X11tpAWpv5XdJPn3/91399ywfxXbhwwatU\na0czCGq3Btkm/s7HQ/us3dsIjVqZ/F2WJ+uhuRw0f/xopnsPHDjQksMU33nnHRuPx5FMJj0+KxaL\nKJVKKJfLqFQqHm9YaxGNRhGLxepeAFCpVOrKJTojkQhisVhdu7U+l212zSm/ueU3LpIuY4xT5vjR\n5SqL2upqo5wX9N7d3b3lcfz+979viT6aI9Fo1HvxtlarVe+aWCyGZDKJzs5OVCoVrKys4Ac/+AG+\n+93voq+vD5lMBuVyGaVSCcViESMjIzh69ChOnDiBgwcPYnl5Gfl8HtZaxGIxpFIprw/K5bL3Ipkc\ni8WQSCQwNDSEdDqN5557Ds888wyeeuopTE1NIRKJoLe3F7t27cJv/dZv4aGHHsL8/DyWlpawtraG\nQqGAYrGISqWCSqWi8pY29tRuQiKRQDKZRFdXF3p6etDf34/Tp0/jz/7sz7C0tIREIoGvfOUruOee\ne3Dt2jUsLS2hVCrV1WWMQaVSgbUWDz74YEt4MZPJOGVqK+Anwzo6OtDb2+t7/9LSElZWalvZ/Xg4\nHo+jt7cXyWTSWVYul8Pi4mLdOGpoRD5rvBX0n+zfXC7XkjF8/vnnbbFYRC6X8140b0mearKEZC/n\nXfqNftf0CfE2tYm/IpGIx/8uXSnrkH1Hdhfvt2q1imq16vEhfQ+S2dq80/Sln90m6+fv1WoVv/mb\nv9mScfyLv/gLlQFlG48dO4YTJ07U/XblyhW89NJLyOVyDfGxa842artK2oLsUBd2796Nffv2oa+v\nDx0dHRvqcMlWP/tNztFG2pHJZFoyhnfdddeGSjUdTbwm+5v4IhqN1tGqfW4G2hzmv3F+kvYQ5zNu\nT3I/SX4PukbKAoKf7Uo0cJ0rZcBf/dVftWQcu7u7rUsWNGq/b8avagWa9Ws4tkLXZuuVOnJxcbEl\nY7hr1y7L5yPxHLdT/fSdBhcvuuRfM/3ZSP816jsE3avR5pIN1to6fSf1MNkYP/dzP4dPfvKT2L9/\nPwYHB/HQQw+pDWp7sEczClyD4XLKZFmuMqSi8VNKW2GuRiZYo45qI/Reb2hGAP9vM+UBzTEMQQoG\nPyOTrieB4qpX1s8VmixTGy9OixSgkk5+L/3vuq7ViikajaJSqWBtbQ3pdBqZTGaDMUEOCvBu8CYe\njyMWiyEajaJYLGJ1dVVVvslkEj09PYjFYoHjRHA5Drx8bZ74veQ91MeyLk0Iu8rjZfH+agbd3d1N\nXa9B9qfWN1xpcDojkQji8TguXLiARx99FJcuXUJPTw86OzuRTCaxurqKZDKJwcFB3HPPPfjCF76A\nQqGAtbW1Dco4aAwqlQpKpRIKhQLS6TROnjyJaDSK06dP4+rVq4hGo1hYWMDq6ipyuRxSqRTi8bhn\nbGt87XJiJS/xcZYGNYHmNgWDEomEF6iU/U1GLg8ihdh+bCbQQ9/bYeRrAYugAIYrGOMK7mjtkfqI\n5qWUSVK/SOdR8parHjIwpSGqBdT87CCp5ySvafK0mb5tNd5r9lgr5/HU1BQWFxdx/PhxL9ij2SDt\n6IN28KNWHuk/qlMGN+R/5ITy+3m5jch/6cNwnuEBS+13jXc4b3MnmaAFbjcDlw3Gwe1h132tgJ9s\nea/x5GbQCh9K3r+ZfuH6oZXIZDJ1POV6aQGfIHr94gP8OolGypbXauPk0sl+Zbr8DIIM5riCO/SZ\ngjzRaBTlchnPP/88xsfH8ZWvfAU7d+500tP2YA+wMeLvMoL8Jq00JrRBlwPWyP9+k12jp1HG2AwT\nbUWYuRyhVoAyOTQHbLPwG+egcXMZpvKzy7DW6pSfXU6/Rgs3lukzKWS/uaX9L42FViIajXrGe7lc\nRi6XQzQaRTwe39Aeuj4SiXhZGxQsIudXG4N8Pu8brZdtSyaTiMfjdb8Vi0Xk8/k6mvj9QC2jIJlM\n1gU2pHFEoM98pT3IUXGNOZXXboPHBU2uyLa4HDIKcKysrODs2bMoFovo6upCMpmsmxulUgnGGGQy\nGU/R0P3cgOYBFeov6VBSlkNnZyd27dqFvXv3Yn5+HmtrayiVSqhUKnjxxRcxMDCAkZERxONxX0fX\n1Rd+1xFd3Kim9kSjUZRKJS9ryRUYbjVPtnr+aGXwNiSTScRiMVhrkUgkYK3F4cOHsW/fvrp7Ll++\njDfeeKMuU6dSqSCfz3tjT/MqmUwilUp5c0eiWq0in89793K4xssviCO/Nxr8aRf8gjRBxqtrfgfN\nY/mdv/xkH9HDeZjfq8lZ6ZDKejR6XfpSa2cz9k477ZtG656fn8fY2BistUgmkxsMa5eTIP9vtD6t\nLNlvm5UdxNM8205zeCQd8lo51/j/Whl+bdtu+MkSP7tNc960cuh/ro9lRo+W4aOV3cx4+y1CafaN\nRCM+GC+jHeMY5Js1U0477tlscCXo8/VAq+tPJpN1wUnaIeAX8AnSmdzWbASSb4L0qibPtirHXX6s\n1I9SBsjgDn+RH1Yul71+KxaLmJ6exo9+9CNcvHgRn/rUp1R62h7s0VIFXUp+M4Ee3qF820KQk031\nSVokjY0yglR0sk7Xf41MHL97Xf0YZGw2Ay3Ys1Uh71JIcvVAMxobbRsPuviND0Fz9qUDS/9Jo56P\nMQVDXPPc1TZZb6sRi8W8/i2XyygWi15WB+9f+kzCeHV1FaVSyRPYtJ1LzutqtYpsNuuVQ9AMRLqX\ntgTx//L5vJpNwsvo7OxER0fHBmGpCXe55UfrW95ul5Dm12rp1y60cixln/g5aNVqdUPqdiwWQy6X\nw8TEBDo6OjA0NOSNf6VSQbFYRCwWQz6fRy6XQ6lU8sZKbq+VCkt+p21hNHd6e3uxf/9+TE9PY3x8\n3Lv+1KlTWFtbwy/+4i9i3759DTt0jfSr5GOSZTSXI5GIl71E/UUOsSynlZk9WvmtgGZExmIxZDIZ\npNPpOrl7/PjxDYbBs88+i3feeQfVahWpVAoAvK2ePOvPGIN0Oo2Ojg7n6jL1qytw66e/XN+1//zm\nSzuDA0HbVYMMR5cuc5Ul7R9NX9EclvXyxQct2CPBeZoMTa0dLv0my+W6QvZLM+PdbvjZfNPT05ia\nmgIA9PT0IJPJ1PFxK3jYpfP8bEDXmDQqH7meaNZWdc3fRtrQyrEN6gM/O8xlY2jlSl2nZSZLeajZ\nCpqDp/GzbI8LNHY0H3kA3mVnagH4RuaO1POt1It+80/+FuTA+33fDF28Xo2ORuvwG5dW0MfRChuq\nWVCwh2f20DEQMttH7gJw2QSc35qBq0zN5/LDVmWWa8w5D3G7WQvy0EIs34pKZRSLRTz55JOoVCr4\nkz/5E5WGbc/sIUdJYxaX4eZSwH6/NToozQR0gq7jtAfR1midzU6udhhHWgCmGSHhcpZd17qMWbqX\n3l2GNjdwqYwgYSLrp3a76NDKdBkc8j8/xpd90GqQEKby8/k8SqWSt4WGOwH0nWf0kLCenp7G5OSk\n1/ZqtYrOzk4MDw97wn5ychKzs7N17e7t7cXu3bu97Tr5fB7lcrmOxlKp5Dy3i76XSiUsLy8jnU4j\nkUhsMKhksKdcLqNQKHjnsRA9QM0wonK0sabPUmbxfgwyKls1ltTf2moIKQue+snPIKBgTzKZRDqd\n9rKqjDFe/wwNDeGBBx7AkSNHkM/nUSwWYa1+Jghvm8vBowASZXjQGNB8isfjSCQSSCQS3nbBoPRe\nGl+XjOV8x50YUpojIyP4whe+gFdffRVvvvmm1y/0IkOZ7rHWesq2VdjMfGjUeeOfKcgTi8Wwc+dO\nfPjDH/a2E+7fv39Dmfv27cMXv/hFvPTSS3j99dc9/u/q6kIul0M2m60b6/7+fpw8eRKDg4MbaJma\nmsLjjz+Oy5cv1/3O57BEUGBHfnbpgKD/WwHNSJV82ciYabpJ01WaUyEddFcfaXyrZUNqRqiWdaDV\nzWlr1JFpxInVXq3UjX40uP7L5XJ46623PEO7FWh0rmzlfnn9xMQECoUC9u7di66uLqct5XJwg2ww\nfr3Uka0cQ825lw4RzWmyV4Kyr7UFDMkPMpOV6uNl83v5lkj+fyOQMsZ17gmnQ/I27w9N/soFIg2y\nT1ptp0qZ0ory+PtWytD4YCu6XJu3rW47H/N26EGJRCLh2Q08qOPK8NHmr9YOenf1t9R9Qf6eq0xt\nTLQ6moHLh+T/ET/xYA8/D5MCPZR5z3mctnX5zcW2B3v8jBb53TVAcsJyaJO50ckg798Mgtql0RRE\np18dfr+3y7DlwZ7NCE5N8fsxoOvF7+fl8PJc/awZ0/JdYz6NBr9yZTv9+sSP+Vu5WiLp4TTzzAue\nsQPAc+55NN7aWoBobm4OExMTdcZ3f38/Ojs7PYNqdnYW4+PjWFlZQaVSQWdnJ0ZHR9HT0+Nt3Uok\nEnUBCYJLcFF9lUoFuVzOO1PI1ZdEW7Va9TIU5HWUXUTg++KDFIKcGy5F1SqDiMaD9wWvRxqkpHTp\nPKbFxUVks1kkk8kN27eMMRgcHMRdd92F/v5+rKysoFQq1W3jkvziahv9RhlkhUIB1WoVPT09XrDB\nGOONHSkrfj4UBYQ0XvCrm9NK76REi8UiBgYGsHPnTmSzWYyNjWFlZQWLi4ve1kAKPkql285gTyuc\nOG1+cv5IJBLYs2cPhoaGvEDl5OTkhoyd/fv34+233/b6wRiDRCKBcrlcJyer1SpisRh2796NPXv2\nbKAnkUggnU57ZQTpea0/XPrVz9n0k/GthMvB0vQM4K8HXf0jsxLpf5c+1MqWZXJ65Mq81Hdclkhs\nZquyrJ9eWj/QNY3YkFuBy9aU/3O9XywWcfXqVZXmzaAZ/eCyM2U5fnOPfpuZmUEul/PO7QHgBd+1\nfuFzUNJBZ7TR/bIuaaO1cgzl1kJJq7yW8xXRJOd4ULBHBkG5btbK8csGdvG07O9mXvx+Oa9JJ3LQ\n/42cA8R1QKtsG618F/zG189OayU9csw2U56f7dQKSBrbMVYSlK0js3rky7WNK8iG1nhH03V+ZwG5\n7Hftf4nNyC2XL6vJILI1eaCHtm/xhXB+PQ/8urBt27g0JS6hDZC8rtHDMjWlrYGYIYg21738s/wu\nFST/vVH6ZH1B97TDsJVKT3v3o8fPyJXgRqZkDs3QkELCNa+ClKLLCNVo4AaoNFR5GdR3rrmsCS9X\nm1sN2QfcOaY66bdoNOpliSwsLODSpUtYWlqqy/Sw1iKbzeLtt9/2hD09VenZZ5/FysoK7rvvPvT2\n9uLNN9/07rvpppswNDTkpDOIT+hMGD5vtHZSORTQorLp91wu553bkkql0NnZWXcdf5f3u+huB6SR\nqtXJx5MyZ1KpFFZWVvD000/jlVdeQTQa9TKZqA8HBwcxODjonedEZzrwYA+vQ8o1uZpprfUyhnK5\nHHp6evD5z38e6XQaY2NjKBaLMMZgZmYGHR0dXnZYLper20LiaqerH7QnQVSrVe9snkQigc7OTu/A\n8RdeeAHz8/O49957vXM4crkcgI2G+nsZmjNMmXsA8Oabb2J5edk7lBuobdFdXV31shOIX6amprC8\nvFxXJgV7aJyz2SwuXLiAb33rW0in0xv4NJfLYXp62gvm+slfoJ7XpXyXv8nyZPtd/dJqBLVpK+XJ\nzIStlCnh0nH8pc193tdct7nq0ept5nrtmq04WY3UpznCrvq3Orc0ndLIPJY8QGO2mYB0Pp/H+fPn\nMTExAQAYGRnBgQMHNtSp2bMcs7OzePvtt3HTTTdtCP5qfdbKMeTj1egDOSSk/bCZYI+rXM12ld/l\nwgZvg1y88HvJYI2mL2W2vuy3oIdPcPt2OwIIWv1B4xrk0HP4lSUdcm3+umRsI7QF1b8VaL5Xu8Gz\n5mhOyqAP7STgC8p+NLr8sSDbwLXV2uUvAPUxBpc/yeugzy6/TX52+ZT0nW9t07KfyK4mPSVfLmxL\nZg9/BzaeoSGv8xM01EDXii+HxkhyImlCw2WA+rVLM/I0h6gRwagJ6mYQZPw2C7/MniD6pGDk37Wx\noPpksIc75lROkJHNaWxEMUpaXUawpJ8zoTSCOV2aYefq4yDltBlMT08jmUyio6NjQ/u1qDD/j57i\n9fbbb+PUqVPI5/Mb+pzSCXm7s9ksXn/9daysrGBoaAjz8/NeO40x2LdvnxrsaYRXjDHo6Ojwzvxx\n9Rldt2vXLi/TgGjgQQmac5FIZENqvqbY/Xignc4IsDHow5UTKQdrrZetkslksLS0hNdeew0XLlzw\nVl8pQ6pcLqOrqwuxWAyFQsE7t4fGSa4caG2n67hhXC6Xkc/nEY/H0dfXh/379+PgwYMYGBjA0tIS\nIpEI8vk8FhYW8M4772DHjh1Ip9NIp9Ne6qpUpLKPpXznmWg0h2k7WS6X8w72pt+uXbuGcrmM4eFh\nGGPQ1dUFAB797eDHZDLprd5obWoUQTKen7cyPz+Pubk576wdklUU7NH0Bp/rxBtEI2WKLS4uOu/T\nHBT+u6s98lqXrHfJfsL1cERk/S4DVXMCg3S3ZiTy/4L6T5bNjV75Wb5zHSfh+q0Rne03frLsVts2\nfnXKtnKd1Kr60+n0hqc0uvTJysqKp3clDfw3PwdF02OVSgXz8/NIJpPo7u7G2toapqam0N3djUwm\nA6B2/tbS0hISiQS6u7uxsrKCbDZbR8Pk5CSuXbvmPd2Q10XtbNe4aQ4V/851CH12bV8kcOeJ84Zm\nm1I5HFK+cVuK08K3g7t8GqmD5RYuv0VyzW6XWzh5OZTty+1av36+3jJWQ6P+Cb9em5vN3N8u2m40\nSP+HL7zxp/vyLftBcsHlh1Ed9K7pGu1ICJr7Uqbz6yWPaHVpW0U1+SPboQWSNTpokZ23jTLeeb3R\naBSjo6N1GZoS23JmDw2oJhjkIEtFJg19F1NebwQZLJspj3C9hYKcwJry0KAZHlyJyN/pXUt75fdS\n2ZI5pQDQGI8rRZqTciVIa7eLif0M2iC4FGej/dssXnvtNezcuRM333xz3WOmuZLn4AZAoVDA+Pg4\nnnvuOXz729/2DlCm+2m8yJDhhgJt43r88cc3PHmLHnlNaJSHaNxuueUWjIyM+PaXMQZ79uzBZz7z\nGe9QZ0430U7CtVgsYmlpSa23EaXkQisevS7PCeIvvm2L6KTzeTo7O5FIJLC0tIS1tTX09/d7W3PW\n1taQy+VQqVSwtLSEbDaLSCSCXC7nzQs624crQt5ePt78P9oGRoqezkY6ePAgrly5gqWlJSSTSRQK\nBTz11FPI5XJ48MEH0dHRgWKx6AUh5Pi69Ai1nx/+Z631UmGB2pPcUqmUd15UIpFAoVDA008/jZmZ\nGTz88MPo7OxEqVRq2XkcEr29vcjn81heXt40j7t0p99noJahs7KyUmcs0Pi67vHLtJT3BDmj9C6d\nCXm9fEkFKLvDAAAgAElEQVTHhl+nQaO11brUxYsab/J7/OByOrSAjOt+l+FLL23BjOtcaehqYxQE\nSYOkJSiFXzOsSe63MsvOZYtsh505ODiIo0ePqnYuR7FYxLlz57C2ttbQtocgaIGBnp4eHD16FAsL\nCzhz5gyOHTuGvXv3AgBWVlZw7tw5DA4Ooru7G1euXMH4+Hhd3ZRBeO3aNczNzdW1YXR0FEePHvV0\n/XbZ8Jp+l/Vr8gpAXVYroDt/BJlNKhcaqC9IF0o+DnISudwL4h1ZjnZGkLStOf3kPHK7WKNP0vhe\nhUb3BxHbxXM8s4fqlVu6KLPHFezR9LamS6l8jQ9kAFTKOxmTkHzVCL9L+St1s6sdkq9ITlC/aQFg\nbstK3kwmk/j4xz+Oo0ePOsdlW7ZxNaqQpPMuhZdsfCNK3yXI5f+NQAoNP+N2q3AZ3kHCth20uOAn\nRGkiuhwSl0Ery9aYxc8p0e53KVBuALjokUyqXcPbKpWsS4gFfW6Hcvr+97+PQ4cOIRKJYOfOnejp\n6fGMZ79tXJFIBEtLS/jpT3+KZ599FleuXEGhUKgbS+JLnkYueX9hYcF3jAl+4ysFZDabxeXLlwMN\ndNpG5sdL/H4+dvx9dXUVCwsL6OzsRG9vL+bm5rCyslJXZldXF/r7+9vCg7KdXCnRixuc6XQa8Xgc\n58+fx+uvv461tTVP4VJmy+rqKiKRCI4fP44TJ054QSCuZGSKKH3Xtg1o7aYslnw+j4GBAdx77714\n7rnncOrUKVSrVRQKBYyNjaGnpwef+MQnkEqlkEwmkc/nvSCk64lArjo57/JtYUQLzSE6s6dUKmFt\nbQ2ZTAYHDhzAzp0761afWsmT8rBw3o5G2yd/93OU5bznPKRl3vAy6Do+z+h3SYtmXGk0ae+8LtdL\nc3Ikra52bofBr+kdl96QaFSXyXHwu9+vL/k9fnUE0arNHX6NrI+PpYsurQ2yzFahUZ3AbY+g64Lq\no36Nx+NeJqHW71QWHbCeTCbrHHOeWddsWyU9xWIR8/PzWFxcxMrKSl22Hy0KVCoVJJNJTE9Pe3pP\nllkoFFAoFOrop6xi0kuDg4N1W6VbCVf/SzmgyR1XGX68TGX5ySmum11Z642W34hPJe1Wydty24fM\n3KV7pdO5XTLUBW3MGqEp6Do//ar5fEF1EX1+ZQbp7iAEle+iq92Qc5U/gl0+cl0L9kg5K9si/X6q\nhz4TXH4dL5fLAqmjaO5rc0CzQfx8RNk3RJ+rXVQn+ahaX0n7pqenB8PDw2p7gW0M9lADOLSO1gQZ\nF5bSQdeyOiSameTata4B1Awev3IbceJdbW8GQbRsBlxpETZLlx99jfaPH8P5lSOZV6Ycumhptq2S\nDpfxsJ34+te/jjvvvBPd3d340Ic+hL6+vjrnyiVEI5EI5ufn8fjjj+PMmTOeEaApKMnTLuFE8OMJ\nTQlIw+PatWuYnJysawenh+6jTA4S4C5nRnM+uCJZW1vD+fPnsWfPHvT393uPESdUq1Xs27cPg4OD\n6jaVrYIrAy4zpcFoTO28Htq29rOf/czbfke/0bamlZUVDA4O4sEHH8Qdd9yBa9eu1QWwqExa6aT+\n4yufmqHMg4VEe6lUws6dO/Hwww9jbm4OTz75pHcm1OzsLHbt2oWVlRX09fV52wHof863Gj/xeSbp\n0jIryACZmprC3NycR8Pk5CTuv/9+fOlLX/Iyz1pt6HIHTcqdRg3MIAPRzzHg/CLnuasuY+oDPlr9\nzZQl6dHK0Rwo7pT4laU5U+2Cy2Hbar2ao+bXHmkAusZX67egrB5+rxx/+b989xvXILtA6qbN2ESN\nQuoxv+u0oEGj5UubFtD7n8utkZER7Nmzp85ZOnfuHObm5rZs7xljsLy8jLNnz/o6kXNzc5ifn/cd\nf9lWoLaFdGFhAUAt2/SOO+5oW7AniJZ216MFpDlfchuE4HLOJd0yIEP3anJB6iwZ6OFbfDW5yiEz\ngd4raHQeSrt/sz5hkH/od23Qdc3O0Ub8zc2Uu1XIxWO/QI82x4Po1YI/Wl/76X+Xze+iQ35upE7+\nm6uMoMxZOXe1V9DCDOG6nNmjXeNS+tKApw7iQR7JyH50NEOzK8gjy5OftfsbZTw56RoRZn5ltQJa\nQM3PoJUTNMioa0SBaM6ELE/OAz+D2O97I3S4yuVMyct2CRyNP4KUy2ZRqVRw8eJF/PVf/zWWl5cB\nAAMDA+jp6UE2m0UsFqt7Cgc9LvvatWsYHx/H2toarH33kew824GMGE63Nj6yL1wCSuN/3rdB/M7L\nl/vspVHtKicSiSCbzeJv/uZvcO7cORhTy+yZnZ1Fd3c3+vr6MD09jaWlJVhrsWvXLpw4cQKLi4ve\nIdQcu3btChihYFDQQzoNvN3UP7Q1p1AoYHJyEpOTk8hkMt5WumKxiGw2i+7ubgwPD3tnH2nZklpW\nD700WjRnvFKpoFQqeVvLjDHek8GIn2dmZvDEE0/gzjvvxPHjx71HsvNxkhk+nPfpdxpbbU6S8UHp\nxIlEAslkEgCQSqXqfisWi4jH4y3dNkK0SLpkm1zXu35zGST8M+cLTaZq5fnpQXm9rNeFIPkv57g0\nHuk/P6OLl9WIMbQZ+Nks8jNBBq20lTuNf7gc1YxVXgZfTdXGhGwoWTZ3AAmcp+UYaHpY0qX1SZBN\nwOvl95HOaUcgXdal/aY5bM3OKTkfuV7iAXX+mG5jjHrQOaBvm2iGLn59kJ2q8ZB2rSyH65VisYiL\nFy9iYWEBJ0+ebIjGIMit4Lx//BaciTaX3uPg80/Tc8C75xZKu1cri4+nRrt2r/xNs2kIcvuWbDPP\nMpL/SSe01XKzlWhkPnIboRmbOuhazT9z3ROkw5uh670M6mMZ3NG2bbnegcb8NHmfJkuDFnaJNpKj\nGu8GzbGgAIxWnvRjXG0PksWyXhe2LdjT6H+acUPgziUfGJcjrZW9WcdeK18zmDWDpxFFwq/nE0LS\ncD0EQiP9oN3jR6+f0vUzzKXhz8vSjHo/R6UR+hoRxq65J2lxzQE5d9o1zpTFMDk5CaC2R/+OO+5A\nd3c3stks4vE44vG4x1eFQgELCwt466238Oabb3pnuHCBLdvq4gcp5OR9Wp8QuFDUxli7j/6PRCLo\n6OhAT0+PJ9ClUc2FreTfXC6HRx55BN/5znc2tFu24+jRo9i9ezeMMd5jeTnuu+++JkdsI7hh72cY\n0jk7hUIB8/PzmJmZweLiIjo7OxGPx1EqlbxHPI6OjuLmm29GIpFAsVj0+kWOm1xxln0nlZYrs4ce\nsw7Ae6Qk7eGen5/Hj370I3R2duIjH/mIl93DDVlyhAh83mm8z5W6XMGsVquIx+PeQaSZTAaZTAap\nVAqpVMo7rJobzq1AI8aO6x6/33g5mnPuMngA/UBRYGM/uvqB97PLUdEQZMTxl2Yscn7Q9vNLudEq\naLrBZbvw67S5qAV8JF/z7DZtbpOBHYlEvEfbao+3BTYe8ugaL95fLholvbI8lw3jJ79cRnk7dKOL\nviC9xK9pZF5p93JZSsEd+k4H1BNNFKDmY89lMslR4gd+TpkxZsM5fZIeks98/lBdkUjEy3KUh8o3\nA6L5ypUruHLlSluCPQQZCPabN34yTc5Z4N0FD21M5Vk//Aw9Kcs0eSmDtH40uxxbqotnjWmggI/U\nn1rGruyXGw3N0NzotZvpB5euvxH7VAPNWy3Q45L3dJ+f7PWrj4N4kOxbF2/wjCPNptfK1qAFXLT6\ngtqn/Sd1sKRPBpT85MW2HNDsIlY6BxxyUshGyd/ky88o1erSPkvaeb2aodKIgSPLbJQu+i770s9A\naiVck8x1raQ3yEjaiiHuGm8517RrpNDxM+D571p/uNomaZF1uIzlVoPP37GxMRQKBXR3d+PgwYMb\nDHkKlBSLRTz77LN4/PHHsbCwUGeISOHkciTltfJ617yV80iW4Tf/SOh2dnbiy1/+Mh544AEMDQ0h\nn8+re9aB+kfNk2EslQM3IGVbpqam8Mgjj3gZMhK/8zu/0/BYNQvuDBH9yWQSZ86cwenTpzEzM4N0\nOu05C8ViER0dHRgaGsLP//zP4+TJkxgYGPAOVOb9wdsrM3o05eJScNVq7fDrfD7vnY2zf/9+79Bo\nYwxyuRzGx8cxNzeHVCqFTCaDbDbrlUlP5yJDVjrAmpzhY5pKpRCLxVAsFjE9PY3z588jmUx6mT0U\nXKLfKOhD/dHK8dJ43Y/vGzUOuX7yC85qDohWJl0rt9Jp9EkeCWoTle8qT9KoBXuko6TZGEEyYzPQ\nZLf8TWuPNIK1FHfOW7yt0hHjdfJsHh7soRcvi59JxueLH1xnLcj2+ukxlw7UrpX9xtvaykw7v3pd\nOp4jaE7xOSh5gzvr3DmRwR6g1v886Mz1k7W1rco33XSTdwbQ1atXcfnyZQC1p2HddNNN3kMCNNm9\nsLCAixcvoq+vD3v37kVfX59XdldXF44ePYrLly/XbVuW7febS3ycOe2tgMZrsq9d1zZbNvAuL3Bo\nDh/w7qHMWpCaywFNZvOFCQk5dzS7Rsp8Kkv6MEF99kGDZku0suztrnM74dJLmuzzsyeCypfl8MA3\nZZHzRU2ecccD2vyoAM0+ojr9xo0CShRgl5k9UldSmXQ/r8evP+l6KWustUin0+jp6XH227YEe4hA\n+R5k5EmDRwvmaI2W18iytfq0z1owxWVk8nulceYqQxtkl5Eor5fQJmer4Gcwu4IY8rcgQ0mOncs4\n18ZWjrP87jp7gK51jSP/X96j9UkjQQht/rgchFYLfm5gTU5OYm5uDg888ABWV1e9LS3yUdPlchkX\nLlzA2NiYl/nDDRPZD5ojogkq2Xdam6UDH+S40W+Dg4MYHh5GtVpFT08PHnroIXzyk5+sE/wy+k5y\nhoR+uVz2xmVoaAgjIyPeIb78aRqcrpWVFbz22mstHLHmQP1M2/HoIE16igtXaoVCAalUCv39/Thy\n5Ii3/WxxcbFuDLVMJs530uGgd02hkQKmR7EPDAzgwx/+MM6cOYOlpSUv42h5eRlXr17F+Pg4UqkU\n0un0hsAcB6dHczqJTnoMfaFQwPT0NC5duoTZ2VkMDQ0hk8l4bctmsygUCgDgBYhavY1LU/oEP74P\nkg/SiA8K9mg8qoECPX57zDWDxo/OoN9cY8ntAqpXguteaS+0CkSbFvCRtPuNCw/IcAOZrpWOG7eJ\npFMon3rCgz6SF/l3l13E//M7c8FvLkl7oZn+ddlvrdSNrrJc/bKZOSTvicViyGQySKfTnrPO9RPp\nXhnsocxOCniRDAfgbcclg58OVQaAzs5ODA8Po7e3t27eSJquXr2K7u5u7Nmzp47uVCqF4eFhVCoV\nrKyseLp+bW3NO2jeZV9pCJI3zcIlA3hww2UvEvz8Ee0z/87tH/5odSmDND71O7hWW2Di7+TEUsYs\n1S+3A3J6eVmurCO/cWsnWiWfNdvwvQZNvjRCZ6Nj0Ijv2w649FyjAZ8gmeviXZrrZL/T01QLhUId\nT1Cgh7K6ebakVo+0m2S9ZJtSoIcWTKUMcGU48TbLOeHnS0p7nJ/RqWHbgj1A41kcmgNMRo9sIHU2\nnzTciZNCUqurUUijTdIrP7vavxW4hIOfYdQKuIIa0iHXaNN+p36S5WhbRFxlAI2tysmypDOqKTjt\nfymQJM3SIQ0yEjQHppXZA0GwtnbA8dtvv409e/YgHo9jdXXV+09LxZSOA10LYEPbXf0qjSBN0Luc\nMzmm2vePfOQj+LVf+zWvDbfccgsSiURdHWQUcYeNnKNyuew9cjsej+Puu+9GqVTCk08+iampKdWJ\n0/q2HeDyTNvWQe1IJBJIp9OeQ5HL5bw20VOx4vE41tbWkM1mkc/nvSdEycwe3iZpNFOwR9vjz/uH\nG8IU8Dl48CB6e3sxPz+Pt956q26l5a233sLXvvY13Hffffjwhz/sKVLafsJTz/m7K/ODZztduHAB\nTzzxBMbGxpBMJr16K5UKVldXMTk5iSNHjmB1dRXGGCSTSS8I2ipoWRtyjDW4nA0OmvdBDoOs3yUD\ntTH1uz/IYZA0aN9l2S64rtFkdVBZm4GU343oE5dD58qc4RkCvF2uMihwT4Ys/cb7goLZnHdlG6ST\nyHWAFuyRc5rGQZPRQbrd79VOaHbUZq7R7gHebVt3dzcOHDiA7u5udRWaZ2xI/UpbraLRKIaGhtDX\n1+fJuFQq5QUb+BOv+H8yu4Pgsrt43w8MDOD2228HUDvT76233sLs7GxdGzV7qd2gzLUgkENIn/3A\n26HNc7qGynItCMjAK2UN07Y74lN68aA5XQfU8z+3PWn+cB3JX8TLpDe1ra5B/ouUa1IWvZexHXKj\nETRiM2+1XD/4HSPRTvjpRT//UcI1jvQ7f9JqqVTyngpYKBQ8viB6qE7iCU1P+dXN+5LqpHrJhuY8\nQrwsbbNGbCVev+vViM+8rcEelwPn5/hKI0IOCnd4pAAKMhCCDGzNSfUz2vwmM58c2m+asG3EWObK\nRtLbKiZu1PjSxlVjGpcD6VcWleeCVpfmjPO5w+mRc48iwHLMNYepEQQ5Ra0cLxe0Pj537hwGBgbw\n0EMPoaenp864pP3/rm0GvM/lO3ccePv5/y4HjN/D+1/OE82JqFQq6O/v9x4hDsCL4stUeb63ngwq\nojkWi3ltpMyeVCpVN1aNKqlWghuamrAn2qk9uVwOy8vL3iqGtRalUgm5XA579+7Frbfeiu7ubuTz\neRSLxbrx19rG5w85GKREtQA78RIvj4JpPT096O/vx+joKPr7+73xSCaTmJ2dxTPPPIMjR44gk8l4\nKzTRaLTuzAgpj2WGBO8TyuwplUq4du0aVldXvWAP0V8sFrGysoLx8XGcOnUK+/fvx86dO+tW2FsB\n7py45r+GIJ3F/2s22KM5MXS9y8APkmuyPo5GdLVWvtZWv8+azGkFtLZqekS2wy/Qo51zQA6b3J6h\n3UvOJB1szgM+XG4RD3FjW7ZD6j8/PeCS9X4Zuy6jmtPV6H2thKZ3+O+bKY9AtEejUXR3d3uZhrQK\nzfUSz/IB3pUZfBwSiQQ6Ojq8OviWnlgs5m3bonrL5XLdvAmyO2R7aVurMbXD9eUiimyv1gftgFwk\n8xunRuYQnwNau/i8BzY+7ZHTw/lSBmCJT+k8JLK55IvqkIFfmi+xWKwu4MPtHNr6TL6SHItGs1Zl\nX7STB7W65fdG6g/yKbTrt8MOp3dNFzeLRu9z2XWtHscgvajVK3lGQpP90i/QAj20iMmDPWTby61b\nWp28fI0WABsCPTzLjstu2hHAdwZsZoFfs/s1OiW25YBmaShKR48LTteKpHQsuZHCP3PBJeuke/l3\nSav87GJ+miw8ICX3xfMB4cafPGtC6y8XA0vh0IjzvFVIB1tTlq5JxsfHlY4adC99lu+aQJH9w1dH\nNYbltGgGrqs+ag+Nr5aRpLUrSPi1U9HIvqlWqzh9+jRWV1dx++2348CBA3VjG7T6zMdW60utjZxH\nJe9KaEqJC3Yp8LhxUyqVkE6nvUOHC4WCl90iD24D4BnOlUoFqVQKHR0dXlCCG30u44/T2M4xlLIF\neFfOkUyhp3DRuTQXL15Ef38/Ojs7UalUUCgUkM1msX//fnzhC1/wDiLmSkprl1S0NB94Vg/PuKFr\n+UoyAE8xGmOQyWQwMjKCAwcO4Nq1a1hbW0M8Hkc+n8fCwgJyuRwymQxyuVzdqqfsE5LHPDWeO9vc\n2E4kEp7TkkgkPD7mab7vvPMOvvWtb+GLX/wi9u3b5x1c3Sq0Itjjd62mU/l/fJ5yHSvLpGs1faU5\nPxqfuGS8JiddOlHWIz9rbaeyuZxpJVxt5npE9rvLxtGCNlLeasEeeS4Pz2AjR5IyBrjuktt1XX3N\nP8tgDw8WaPOHy2muG+WqJ11L79KY1V7thjYPm/nfBamr+HYDLbOHnzNB9VFmjxZg4/fRHAHqg3rS\n8ZDy2a/NEq7502peC4ImFzWelw6SC1LPaeVLuSnroH7VMu7olUgkvC3XxKuS72Wwh79orGlMaT7R\nIgw/oJt0rnQupQ3K398P0OSc6zp6lzpSXgNszs/S5Fwz8NP3fvdIWdtOG5Xq1BIgpCyj37itKOej\ny+fkIJlKNjvZ/PREYW7XVqtVjz/8FrCoXPku9ZMM9JCPwfUULTjTfTxWQLLCNSZa+7V+CZpP2xLs\nkZPL5QRLY0lblaT7paPnZwBwZ8RlLPsZjhqNfoEpTiePpksaNSfZb9D5dbJf28nA0lCTxhunyY9e\nubIgDQO/eUGfeXDN5XRLR4Z+cxmPmhOrjTtXwrLeRvvIpSzo93YasprAzefzWF1d9QQgp3dubg4T\nExPIZrMbVpy548BfvG+koOd0SKHVyJynz3zc+Dj29/fj0KFDOHbsWN3c4HW4jDP+Xi6XvSePlctl\nvPLKK/jZz36GlZWVusNOtfnabkVKTqQ2nygtnM7quXjxIi5fvlzn7JVKJcRiMQwPD2N4eBg7duxA\nPp9HLpfznA3qWy2LiH+X7y5Dhvc9OSF8b/Ott96KfD6Pxx9/HGtra55iNMbg7NmzePTRRzE6OorO\nzs4NfC9lMT+vhBxka21doKdYLOLq1atYXV31DmwG3nWUgNoKeS6XQy6Xqzu/p5XjSHA52Nr3Zv7j\nfeMXFAniVS43XPKLy1tZjuYYu3iHlyXnjx9cukfqyXbw5mYMdjmH+XhptHKeco1lUCBJlqcFibhM\npev42DYT7JFBaD9DXZMdmoxvJF291ZDzkf/eKB1URiKRwI4dOzA4OFjnnEsnQbaX1y+Dbrx8nrVK\nYy/lcSQS8YINsi1a37raGIvFMDIygmg0iunpaW+b8HaODaERB1HOq0bLdDlVZDPRtTKDRwZ5KLDD\ns+7oQQCUYcoXKjjvWmvrxpYHC3m2lpS/3JHk7fELnrpsYBkkorn1Xgcf856eHgwMDODw4cPYu3ev\n9z+9S12hyeJisYixsTFcvHgRs7OzKBQKvr6PtRb9/f0YGhrCkSNHsHv37g1+q6sfG7EN+Pv8/DzG\nxsZw9epV5/bKdsOPZg7NhvSDJpeoD2mRV2b00ItnnXO70GX/aDaPtH9l/ZInNb7iQWAqp1VjElTO\ndXn0umYMaoa7JsC4UqKO46vJEprB50ePX4fJrAZtjz2VwWmkySDbzSeDiyHlBJerMNshbGXQQgYz\nXHD1uTTqNQTNDz+nncDHidrBVzvkGHFjiNfBFSbfc82VIh8Pqoe/u/qQ091Iv2wFpJylkcqFDzcO\np6amcO7cOSwvLzuDmlQW9RHvQxdfcQHKeVcbT9mvXGBzYR2JRDA8PIxPf/rT+NCHPqSuXnHDSTpP\n3LAql8vIZrNIpVLIZrM4deoUnn76ae+AYx7saURJtxJc3vH5Qm2Ix+NIpVK4du0afvCDH2B6ehod\nHR1IJpOIRCLI5/NIJpPYtWsXduzYgVQqhXw+v2FVGagfG80RcP1G71K+cYeEr8DQVrIzZ85gYmLC\nC0xVq1W8/PLLWF1dxS//8i/j+PHjG7Ye8HHlxjYZzTRnuNFNT/yqVqsYHBz0gjikrKncaDSKYrGI\n1dVVRCKRlgZ75BkTLkNO+8/vPg4uo/i1mkErjR1Zvgwyyv/5dZL3eH2awSblkcZTQUaR9n/Q961C\ncyj57/w71R8kM6SdIwPbQY64Np6aHJb6lFY6uXzXnDt+PZfxriCVtsDl14eaQy0DH40661uFNmfl\nb804UMbUzv/av38/BgcHkc/nkc1m6wI+2pNcpBwtl8uePOOPZrfW1m394npS6k9pM8tFmkYcr3g8\njn379iGTyWBpackLim+HQxkEFz82GjDk81obBwAbbBDSGVwXky6iwA4Fe+g3nmHKM3u0YA/fLsK3\naxWLxQ22GbdDOW/y36W+p3ZzG0ezw3mfbgcfbhaafO7v78exY8fwpS99Cffdd98G3pB8LfWnMQbL\ny8v45je/iSeeeMLjXykrZf2Dg4O47bbb8OUvfxl33XVXnV8Y1Icu31DSaYzBW2+9hW984xs4deoU\nlpaWvAxOvzLahSB9Te9+9oEfrK3PkOTZPDzQUygU6vwS4idJq1/fSB+N0y6z7HhGJq9XW/xvFeRc\n1bCt27ikUyyJ5EaGXJGSAosrLfqNN1aWv5XO5TRyYewK9nAaaSIY8+6qNi9XDrwUqLItfoaiZpS0\nEi6DjNcv69WMef4f3as5N9Khk/NElktzgf+ujQtnPN4uDXys+QqNNHg5uACS58Pw+SBXUGWfbRfy\n+Tzeeecd7N27F7t27UIikYAxBq+++iq++93v4vLlyxvmu6RZGv0u45/f55ovHLyfqtUqRkZGcPvt\ntyMej28wvvr7+73HzkYiEW+llD9ymxtLfH+tXOWORCJ46qmn8NRTT3lPItNWwLV5086xczlwZKTR\nwcz5fB7j4+MwxqCjowOJRALVatV7MkEikUA2m/UyV+QjKqkPqQ6pqPi4cP6RxpPWH9Vq1VPO+Xze\nOzCaxsKY2hk76XQay8vLuHjxIgqFgmck8zlmra0bMxn4IaOV7qH66FHrqVQKkUjEazsZR7SFLJVK\nYXl5GXfeeSf27dvXsnHkMikoKOE3n4L+cxl50klxIcghCqpHKyfIwdfu0XSlpKPR31sFP/r92iMN\nXK3P+LVy0UAaddpiAsGYWjZfNpv1vvPyXXYMD2jTPUELcbx8ut7VNvmS9MvftGzZ7UazdqR2Pec5\naodrYUhC9o1cQLTW1m3FlfXIcrltDQBdXV04duwYOjo6NtSp2Snctpeyfqs2d6vA516jNh/9T/qF\n2s4DP9L2Jd+AB3j4eTwU7OHfKdjDt1zKrDn6XK1W6wKCPBuMZ+1xW0n6GxxasEGzwxvVL+9V0Ph1\ndHRgeHgYd911F+6//34cPXq07imfXL5IG4vLOTo38CMf+QiGh4exuLjoDPYA7/ZzV1cX+vv7sXv3\nbpRKJSSTSc8n9NN59K7JWI3O0dFRPPjggzh+/DiWlpbwyCOP4Mc//rEqq9vFn5IuzWaV9TdKi/T3\nOHNqcbEAACAASURBVF/wrB560fZY0mNku8s572cLaf449+0lz8kjIgDU8TEHH5NmxmUzfNf2YA8J\nSOkgap3tMiZkh0gFROVJJ3KzwkgztrnDr+2nl0EFLjx4EII/4UIzXOSgyzZKJcqvbcZJaAaNGuau\n+jjdLmNdRs9dAtc1F2iO8aBikHCUik3Sw517vrc6kUhgZWUF+Xy+LpWaQGOcyWS8J2DIwA+PuEtD\nql2GkmbodHV1obu7GwsLC5iamvLSywuFAt5880387Gc/qzvbRAa35FyUzgJvizQIJS2SRm7AxONx\ndHV14dChQ7j//vvR0dHh3VsqlTA/P49qteqdu5PL5Tz+6uvrQyaTUWUKp5u2QS0vL2N6eho//vGP\n8e1vf3vDI4xJpnGag4yiVkLOay6brLVYWlrC7OwsZmdn0dfXh97eXi9jKZ/PI5PJoKOjA9FotE4h\nulb6qI3S6JeGkuak0bXy3VrrBXtozIjP+DxaWFjwglQya4dok4+v5XzLr6lWq5iZmcHS0pJ3yCh/\n9DHxZDKZRDabxczMjJfxcODAARw4cKBlY6gFiWWfy89+17n+k7JPXiOvC4KrjCCDCfBPgw7SKy5n\nUsr6rbRtuyDbKfsCqOc7HgjQdCnXkfTOeZMOZOd90ujCFa9DBsP5dVpfN2Ls+8kROT+2K7NHGt3S\nxuLzzUWHqz+stcjlcshms3XnGfkFemRZ/B7epzTWlB2kyW4qmzJPeEA8nU57j23XnCrNfiM5nE6n\nvcWV6wGXPpb8ovGaZhcB9Zn8mvzhNgMFccg+5Fk8MuhDASG+jYsWMjQ+JBuR62kK8vD/+aKV5tzz\nNkp+AvRtWUE6arM+lgtBZTVblzG1bLqdO3fi9ttvx91334377rsPqVQK8XhcXYR1+R2RSASlUgnx\neByHDh3C4cOH63xBSR/nU36OlrXvLlBx+5/fR2W55KeLzh07duDOO+/02jI1NYUXXnihLttvO/Sh\nnw2j6Xn5v2vO8pfM6OGBHlrEJJ4BUMcrklaNT+g/SafkIRkslMEeqtMlQ11j4tdH8n7g3UVUF7Yl\n2APoBg6wMbNHrtDyKDe9qBzeSbz8Ro1PQiP/k7EjnT7N8AGwYfABbIi08wwTroSlMScNAI152g2X\nI8dp2ooQ4QwuDVepvLSMA81Q4kKBvstrZKCG6uGHXJJCplWYaDSKp59+Gi+++KJnMHOUy2Wk02k8\n8MADOHLkyIZsEu1JKLxf2+mkcD6JRCI4efIkTp48iaGhIe+ahYUFXLx4EfPz805nQPIx0Ut8ws8D\n0DKYNF6m/6XDY4xBf38/PvrRj+JjH/sYTp48iXQ6DWNqKy0TExN47LHHsLKygnvvvReLi4sYGxvz\ngkQHDx70AiEy4BaJRLwAAxlkL7/8Mv78z/8cFy9e9DI/pGLVeG47nUrefySP4vE4pqam8NJLL+Hs\n2bPo6OhAOp1GLBZDpVJBPp/HysoKRkdH8fDDD+PgwYOeMnQdVkd1ab9LJ1SbC34OXrFY9DJ4uru7\nsWfPHkxPT2N5ebkuw4eMsnK5jHg8XrciZ63dcLClNJqj0SiSySTm5ubw7LPP4rXXXvN4O5VKIZfL\neYGwWCyGPXv2YH5+HisrK54cANBSR4Zv42pU/wRBu8bvt1bJGWmE8nfNcOOfXQ6+vN7P4fajpxXt\nc0GzXehFMkwz5KWzTtfJbVR0vewjKYc045OXGY1GkU6nN5TNDyTnK4/c8aF2yva5+Fr+RvdqfcDb\nx1PfXXNCc9CvB6Q8dMlHDcViERcuXMDi4iKGh4ed857KlfPeJUvpnWwMbvPQO9erJFN5FhBd43LO\npJyn6zo6OnDkyBFcvXoV4+Pj12VstGMSCNIH4Z+1FXnqa764QPYD6STSHfzMnXQ67b3oP5kRLBck\n+GHNrmMigHpdW61WvfLK5bIaYJDvfgFFGUBwYTvtm1bAWotEIoHR0VGcPHkSDz/8MI4dO4auri5Y\nazcswMoAquaH0llMtH2LQ+MbKpcCsOl0GslkEtbauqwTl87TMj+kf8RpBWr8WCqVkM/nMTQ0hMOH\nD+Py5ctYWlraQG87eNUlu/3ki999MoBCL5J1PJOHAj2UKU7+lt/imuYXbgbagoVsF4c2rrIPXf3C\n76dr1tbWsLi46KRv2w5o1gjlBPNJLY0nGfCRglrL+qAO4EqyUQeNGy2aIccFOKfNFeyh6+SZEPTZ\nZRBudfK1ipGDJq2Eq08budZlPMszAqSRTHXIsZeZURqz8K0eFOghJ4/2VGezWVy+fNmLHgcFezKZ\nDAYHB2FtLfOks7MTO3fu9AILmqPMhWGQQ7NZ8P7hfWBMLd1/YWEBy8vLmJmZQS6X2xDkCXKkqP/4\nNgBuoGtt4jQQXXRPNBrF4OAgDh8+jHvuuQe33XYbhoaGkEqlEI1GMTMzg3w+j4mJCUxNTWHnzp2Y\nmpry+jUWi2FiYgIDAwMwxqCrqwuDg4Peyg5vGxl0V69exTPPPINoNOplgPD2+WX2tNsg0uQnN0Ry\nuRzOnz+PmZkZL7MsFot5j/WlA0KPHj2K7u7uuqeTBa0qUzv5fJUOmrxXk8d0Ha3KkIyk4NTCwoIX\n2KE5d/nyZVy4cAE7d+70suWI7ynYI51SPhej0Siy2Sxef/11XLx40dvuRn1mrfV4vaenB9Za7Nix\nA8YYzM3N4Z133kEymcS+fftaMo5cZmj8oKGRoFAj0OarlAuu+7Q6+ZzU+JvrD3rn+ttlzHBaNTkj\n2yLrbnfAx09XcX7QMgj4nn4KyLhkLL+X6qW6paNK21bJGObZBZw/uFyWq9dEv3QkpB7m4yh1LaeV\n30PZc7xP5KGWQfOhnZicnMTU1BSAWqbL6OgoMpmM1xZNxjWjpyuVChYWFgAAO3bsUAN2vGwaK/4b\nUB8kk9/psFJXP1KWJ5f7fDw5NMdFLsbEYjEMDAygUqkgm81iZWXF2za4XZBOOrffOa2uuUeONz00\nAEDdIcq0bZzmcjKZREdHBzKZTN2L/0b3aT4OfZeLx9IPIrop85U7u5weze6VLxnQon7hMtblfGoy\n6XqgGf63trag2dPTg5GRERw6dAjDw8PeQxq0jHs+j6hvqZ+Bd5/qCbybwe+qm/qV/EMaQ5KD/CEV\nLt2qzRsuk7k9Wq1WvcUxKrOjowMDAwOYmZmpK5dodPnFW4HU7fxd+krafVo5coxIxtFTdmXAh+xa\nflByUN2af6jJQ7/7JP/xclxjzN81uvzq4vdS4MuFtgd7pODRFJcUKjyayoUhD6jwTpUr9TJzQhol\nfkJLGsLc+ddeUtjKsjgT0uST18i+4vfKa7Q6tgOtMri0e0koaw4s72stmMcjxdp3yWRcmPD5REKZ\nH6hHKzSpVArnz5/Hd77zHZw/fx6XL1/G8vKylwkgx5+eePSDH/wAr7zyCgqFAm6//Xb8wi/8AlKp\nFIB3D6nlkCuqfsJpM+BBCvp8+vRpzMzM4NOf/jR6e3tx/vz5DWe2+Dmj3FiQDgHvf5nWKI0K7Tvx\nzcmTJ/GJT3wCd955J/r6+rynfiQSCVy5cgUXLlxAoVDA/Pw8fvjDH3qBVeJ3Sp+ORCK49dZb8dBD\nD+Hmm2/GyMhIHY8D8M50Id7mW4GINi2zyU+mtNJI4vVJfiHDNJ1OI5PJeG3nBsvu3buxe/dudHZ2\netu4uJPVDL38nkbkOr1zuS0P1qOXte8+FrhcLuP555+HtRaf/exnMTQ0tCHYwxcDuNFM/FytVrG6\nuoorV65gfn4eAwMD3qHVZOANDw8jmUyiWq1651dcunQJr7zyCqy1GBsbw2c/+9mtDF8d5LyXfdUs\nNHmhGRnNGHiaDOXlkLySCx2u1XS5ysyDHjJzgNfDaaZrZMCM8+V2OCZ8vpGs4CuQ3OCXfUB9R3NP\nrtTy8vln14sceJJltN2Ab8/k20ek3OYBHT7m/He5fZLzIAWXZP/wxRDpZNB9JAfI8dFsRdkH7cKZ\nM2fwyCOPAAD27NmDL33pS9i/f/+myuJ0umw5F29xG1i7TtqKQP05LFym87lIPCJXymk++tnHmqPB\nP/f29uL48eN46623MD4+vpku2zQow4XmHLfLuVzgi2lAbQ7LJ/iQTKKMnUql4mVjUGZ3MplEZ2cn\nuru70d3djc7OTu9FQR8K9mgOG4ANvKfpS/6dggUu/pQ8JeeDzGKRtLgCC3503QiQPqgWOJCBT5KP\n1LdAvfzhfCN9Ev5KJBLekQPEkxSIoOCin37VMnhcMpvsPFfA3KXzWx3wcdUdVAf9zwPY0lbkCwPE\ns/ypqcTDUpf42Uaabib4+fWuF5e7fF5wNKPH/PqumXFre7BHGndcmXBBrAlAV1CFzl2w1iKdTmNg\nYMAzbOSqEx8APqn9FBqng9MgT9fnB4Zqzl8ul8P09LS38kFGGF3HGVROEhedmpKV/7UbWt/5GWJB\ndMr5QFkK1M/0WcvuIiWvMZe20sjrlwKUp9Tyg27feOMNPPvss3jhhRdw+fJlLCws1GVryTbTPKRH\nHxYKBe8xj65VVG6IuRTtVsH7jvpmZWUFExMTOH36NK5evVo37+hgZul0aGPGf6Ox4RF4EtR0nVzJ\nkjSScWOMwZ49e3DLLbegu7sbkUjEC6YZY7zoPvHR4uLiBnlDoP6la0dHR+voob7RVtpc/al91r63\nCtIYlzxjrcX8/DzW1ta8x7laW8suS6fTOHHiBG699VbPOaSUf2n08TZozqfmvGh0aeXJsii4dvjw\nYe8sgvn5eS/jplKpYHx8HF1dXbjvvvsQjUaRyWS88aVgDwefl7lcDhMTEzh79ixWV1c9hwBA3Xkm\nJ06cwPDwMKrVKqampjAxMYFyuYzFxUWcP38e09PTWxo7DpdTr/VlM2UCG8dEjpXm0LkcEj99w/le\nq5PLaV6GNLSlkSTB5RavV/vfpZvaAZLdtJAjsyD5/Pdz+IB6J1QGsbSXnD8k8/hWLDlm/Lt0DDRw\n2SK3cPHypIFMkH0gr+P3ay/XfOP91mosLy/j8uXLAIBkMoliseg02P3gx8+83XJRhI+vFnzjZfAg\nDYANzjy3J6VtKbM5JZ9q8kEbGz6f6MBh0kPXMzAg+1H2bSNyQvIbD3qSjZhKpbytWxTkoaCPdi4M\nD5ByyIwGeud2Lj29kvQ2nXlHLwpa0XX8ARTavAjqO94nLj3+Xoa1tWzdW2+9FUePHvXOBnQFenhm\nD/fpCCQzI5Ha1n8A6iO2iS95RiUPZPDjHIrFotOZl4EdbjdrPEt1S9/XBS7XWjmmrsCHpvfkfbIM\nmczBDyan4A49aIR+o2xxWVcjPhWXw822V9Kt+b0u+8bVF5oskDI9aJwJ25bZI4M9nKHkJKBJLYMs\npOwnJydx/vx5AMDw8LC3LYMLcamcNIOJ08g/cxq5cOfBB37gGrWDr7QYU3tE34ULF9DR0YGhoaG6\nwBAXvDzgI1dbtGyWZoyOVsFP0NNvmhKVfesqm88J19MMeJYFN5boOw+8BI0/DyiSEctT3inYMzs7\ni8ceeww/+clPvHNgKCOAG+Z+TKfVKwM9vI8aEUqbAS+P5pwxBqurq3j66ac31C/T9rXyNKXIeZ6n\nSNPKL9/DLjNnqFwyYqvV2uOxd+3aBWst8vl8XeYRDwaSA88Du9baurG6du0aJiYmMDQ0hPvvv9+j\nlzsyMrDoN6cbddTbybM0VolEAqVSCZcuXUIul8PevXu98SkWi+jv78e9996Lm2++GdZaFAoFdQuX\n5lTJsdbGna7j80fKWW6s0P/Vam0v+913342dO3fCWouzZ89ifHzcG7vJyUn09PRgZWUFxhh0dnZ6\n84zGjurguiSfz2NpaQkvvPACXnzxRZRKJWQyGS8Lo1QqYW1tDR0dHfjoRz+KkydPolqt4plnnsHb\nb7/trSJdu3at5cEe2We8v7V+9YOfjpAy0Q8u44L/J9vhmtsu402u1GnGoNY2Psb0O3d6eR9pzkqr\nwQMhNJ9kkJg7mpou5P9RG2QfSDuAB074dx7AkXTKABnfyuXnHGh6k/qSryjzwISkWS4GULs1I9lF\ny43iYAKN8bBmvEtdxftUG1c+98n+0TLmeB3atZr9IcFtePrOx/562aUEzb7nL2lXcbnA+ZcCAtZa\nz/bkD+fgtoZcCNYCQPF4vM5JlUEdVyaGzESRW814sIcO/KYXObz8KZt+QdRm+/hGQmdnJz75yU/i\n4x//+Ab9wwMJMthDdgXwLk/wLV40tnS+IJfjtABA5yWSDcG3bvEX9xsJXOdxmUB0cfnA75HjrPnX\nEprO2QqCAhOavSxtDDlOvO9kFjgFe/gZPdx/5rpT6kLNRtmM7nHdK9vViJzU/CkpE7TxDSp7W87s\n4R2sBTGkEywdYnrNzMxgcnIS8/PzdROEXycH1++lgeiUDMWDPrQncmJiArlczhnsofSycrmM119/\nHbt27cKOHTtQKpXq9vbLaLJkCL/J5meItwpcWWr/0e8yKCAFoDaJZTt5lgJ/WsHU1BTOnTvnPZXn\n+PHj2Lt3b135PDtDRoSJPr41h9fDFXq1WsXS0hK+973v4fnnn8frr7+O6elpLzhEUX3eN5rDIgWG\nNOTkPOfXap+3CumY8e+uYKxrNVU6UbzNJIz4IxE7Oztx2223YW1tDWfPnkWpVKobb14Wjd3hw4dx\n7Ngx3HTTTV7mHtFJe66ff/55PPHEE5idnd1QDtHEg6ykOHg2S6VS8dJA5+bmMDMzs8FJkeOgOUX0\nu7zexTubhRwjWlGl8wKSyaSXXUbjQNu16IDA1dXVun5wyQ/NEXA55BxcDkuaubzg8yuTyXhPxyJj\nlrKTEokE1tbW8Oyzz8Jai+PHj3uGFNVDRhbdE4lEMDY2hlOnTuHs2bO4du0ajKmtuFlrvUdSZ7NZ\nRKNRL00+kUjg6NGjqFQq6OvrQywW87KlWgVXsMeFRq+Rn4MMF80Y4fdp8ozPA+Ip3h6ewcHL4AY2\nPzNEk51+bdd4Txq0nMZ2yVN+dgO9KFBcqVQ2yA9Jp9afLtnhsps4L/FMES6fZRYQsDGTIMhY5ffy\nOuWZMvxa0rGyra76gsbGT/5sBXy+cszPz+Oxxx7D6dOnvfppW/HBgwc9eqRedc1Va2sBhJGREfT3\n9yOVSnlymsswoolnvlG50o4gvuIZtDQP+JzTnCfKAKF5HIlE6rJvqWz+tBu5KEAPYuB9sGvXLm9h\nRR4Ky/unleNIbXDZV0Qj7zuy8ckWp4wkmtP8wRz04ocv86dr8adx8Ze2QKkFcviii8wC4sEeHvCh\ncSE/g170O3eQ5VZtPhZan2mOMvVhK2Vou2GMqQu8kJwE3NmsQeVpfdNon2hy0FW/lHWusZPj14y/\ny8tqFTQ/z0U3/1/2hbYwRHOayyR6cT6SclIuXEsZremgoH7TdKWk3VWGyy7T5ocrsEO/33bbbThy\n5Ii3+8GFbXv0ulRUfhNVChsyzKenpzE2NuYZEVQuOYGNMKA20TRjUAaaZLAnl8vhypUrmJmZcQZ7\n6HOhUMDy8jKA2r5m4N3sBk63NJ634iS22iBqpD9pzDgN0lCRjC3L4QGZRCLhbTe5dOkSfvjDH2J5\nedkbg76+Pm+lRZ64Tv1IAoKUODeqqQ7++MtEIuEdyPrNb34TTz755IYVHTIKpLDmgYtoNIrOzk70\n9PSgVCqho6NDjTBrq668/1oJ1/hRX2lK3U9pcONfCiJufCaTSYyMjOC+++7DzMwMLl265K2I0DYj\n3l4at0OHDuFTn/oURkdHYYzxAjyRSMRLZ37llVfwzDPPIJPJ1D0BjM87yr7jgVXOr0DtrJ6ZmRm8\n+eabuHLlijqvpbPm6ltXP7USRAPPSKPf+PYDMtbp6QTkjEoDUyo7P6cvCHJVmsqV857/DsAL7srH\n0NIZCblcDqdOnUI0GsXQ0BC6urrqeJ6y7uhchWg0ivPnz+Opp57C3NwcKpUKBgcHvWB9oVBANpv1\n5hU5OolEAvv37/eelpPNZvHqq6+2NNgjHedGjLFG5pEmZ13GhgsuZ9xl3AD15/TwQAc3SvgKHb1o\nS56mG4juQqFQ1/fSSOIGvLZgwstrtYNJekXaCfwMEW3sGnEuXPpTK5cHQnjbXU6bFqCTRqRGp1av\n3EIJ1OuPoMCSZvRqfdEOOUplSxhjsLi4iKeffrru/2QyiYGBARw8eHCD/gsq19pasGd0dBSDg4Pe\nE324fpHZyHwBg2ef0pjKbAEqj2csA/VnKvHgEAV7SM7yrFmynXhAgYJPXL5z2qytHWzf3d3trcBz\n8MW3VvKiFjzX9Lf8jQf55Xx0BXp4sIdnn8vgD9mMZA9xPuTykBbEpE7mB8zyYA8P+ND2LXk4LQXm\n+OKWDPZIe08L9Ei5wfvyRkIsFkNnZ6dnD7n8D+037aXJLZcM0OSc9n8QXLS4/AqSJzJ7qN1j59dm\nPo/87BLiD7loT3KLBzuJH0jucd9Z+vFyPrt0kLTfNVnuaq+r7UE2Fy+L9wF/lzraWovDhw/jwQcf\nRCaTqUtEkNi2zB6ZEicnp5y43GFYWFjAxMSEd1YKH4iVlRW8/vrrGB4exujoqDPCyVcepPDSJiQv\nRwZ6KNWTb+MCNj4indpLE3Z6ehr5fB7Dw8Po7e310r7lvmzNYG2kn1upPJuBn1ELuLOPXEzG+3x5\neRlnzpzBT3/6Uzz33HNYW1uDMbXzZH7yk5/gV37lV3DLLbf4brsBoI4lBXpIeVtb2yb02GOP4Xvf\n+x7eeOONugOYuYLWVmj6+/uxe/duVCoVpNNpfP7zn8fx48dRLpfR39+PRCJRl4LH5xn/3qjw3wpk\nH8k5J3lECkf+u3SqeJAtkUjgc5/7HO69914cOXLEy6544YUX8PLLL2+YG9ZazwDq7e3F3r17EYvF\nsLa2hmw2C2MMksnkBh6VQTMqj6/g8f9oHGlsKXPw0UcfxdjYWF0fuRS95qRIR7Vd4EZ2KpVCsVjE\nU089hZdeegmlUskL/hSLxbo0bzIG+TlKckypfL92yvbSb1LBcsclEqk/A423w9paQH9gYAAPPvgg\nBgYGkE6nsbS0hNXVVaRSKVhrMTk5iRdffBEAcPLkSdxxxx11QRo6fBaojT1tx+zq6vL4kgx8SgHu\n7e3Fvn37MDAwgEQi4TkjqVQKt99+O+LxOGZnZzExMdHSMXTNEe33RuaTZqC49BuwceyCjBHNwOFB\nA5nRY4ypy9zhTia9br31VvzSL/0S+vr66oICnK6nn34a3/jGN7z6ZRakZqRTnZJnW8mXNL9lsEfa\nDlJeNiLbOe9xmmWQhv6n36nN0lDm37kjQGVzY5oHxYF3s4Y0uS8XKvg9HNq2ou3Qc41iO5wgCa0P\nOa/wMeJ6SjrccvWbriFbh9uUJCt5MEHOETkneIYuDwbxxS8JCpjv3Lmz7vdLly7hypUrLR93TXZw\nPuTOI5/HlNFD3/lY8Eep05MtKfvU9Yh1GjP+pEl+jo527g4P4PDx0II9dA93eOUTNaVdKtv9foa0\nTzS7tdFyXD4pAC8TnHhL6t6goAbRxN+D2qP5zVzPaLpnenoab7zxhndW4XbBr78ln3F9TffKcvxs\nCDrkemRkBH19fbDWYmlpCZOTk14AiAe/XTuAOD3ai/7bSl80Mg+14I6UX9p/QWVvS2YP8K6Rwg0w\nbfJqEzmfz+PKlSveQZ68UwqFAq5evYpkMok9e/Y4MxKA4Kgi/19braMXrTRaa32DPdT5pCBXV1ex\ntLSE7u5u9Pf3q9F0rmylUa4NpDSuXddtFXJyufo0yFHRjDxpRPL+Xl1dxalTp/Dcc8/hwoUL3pN6\nLly4gIsXL+LQoUPo6upCZ2enc8WBxocEoNy2RU8smpqawoULF/DUU0/hiSeeqDtTho8lCR0A3mn7\nvb292LNnDw4cOABrLTKZDO6++24cP35cVd68b3igRQt+tApyHDQniY+Dazy5EymFJIEMlEwmgzvv\nvBOf/vSn0dXVhVgshsHBQe9xtrxM6l8ybDKZDHbs2OHtS8/n83Vzg9crFTOnhc+Ljo4O9PT0oL+/\nvy5AGI1Gsba2hnPnzuHKlSt1Z4BJ+cQFr18ftwO8z/k2xFwuhzNnzuCVV17xHNByuez1HbWRH8qs\nKQiXouVtk/OI0yXlN4Fn6fCtkFKB0VOw1tbWMD4+7m2xo1Xs1dVVvPPOO8hmsxgcHMRHP/pRjxdp\nzGgeES20XaJcLntZeTwIduzYMZw4ccKTybS1Ih6PY/fu3TDGYNeuXejo6GjZOLr4ptnfXNdwHm3E\nmNQCQ36ymn92yXKgfnsoT8WORmtPhvvYxz6Gz33ucxgaGtogF+k9Ho/jzTffxNWrVzE/P68GGPn1\nPMuHt03SulVowR5pu7jsENkG2R5Xn7u+UxlkQ2hOn6SLg48NjRkfQ75YxumURju9S5tIOl2a7PGb\n3/K/djquLjr6+vowPDyM7u7/n7sv/Y3zus5/Zoac4XDfN0mkJMpabWuxLG+y1dhxHDtxY6BJ0yAo\n0K8t2g9F0X7rv9AiBfKlQFCgDVq4SF03qZ3Fie3Uq+RNtmXtC7WQlEhx52zkLL8PxLl63sNz3xlK\nQ1r+HWAw27vc5dxznvPcc+/bvOL+3Ge6TlzPQqGA+fn5wL5/rBvat3Bbso3Vv3NfM57kPX80QWQR\nAhxwMWbhFwCHi7QflHrGYjF0dnauaEveALuaYvliqbO0kzWxwZlTTJRFIhFH8jDBw58564fJHm67\nYrG4guiRLB4h0PT+OvziLB8d4Mrx/JmX/nGfrpbosL6HYcJqSTWufTuBuRUz+uLTSCTiiDZtT9kG\nyFi5nXppf6Fxle8lejw7O4uRkRFcvXoVN27cCGT5raf4YlYfvrF8G+uyzyaVSstPPpMsYV6eahE9\nPl329ROX7U7GgNWvVn19RI7+rv8LkzUne9iYlmuksEEmTksPLOkE7kQr4GMQ5etkvqcmeJgdvHTp\nEkZHRwMpsABMAMPOUwCTDwzqct5tYoEzC8zy/6u5npwjs0a1tbXIZrM4deqU26xVZviLxSJStHed\nkAAAIABJREFUqRRee+015HI5l8ZmAVnuUyZ6xFFLgHf06FH85Cc/wbVr15zz5gCf048la6W7uxt7\n9+7FkSNHMDg4iI6ODgeUOzs7HTnF66/FGfv0Xc5fCx2wAg6rD+TdN2Z1cMjHSf2krmyEh4eH8fLL\nL+PatWsr2pdnsCT9u7a2FqlUyi1DkmPEHnAwKXrDZZDyibMYGBjAoUOHcPDgQbS0tLglPLwhOD/9\nTewAp9Bz3VYTsFRTBKgKaZLJZHDjxg3cvHkTjY2NiESWNydOpVLIZDJob2/Hhg0b3NNBdFAuYhE8\nFniyglEOELmcekmm9JX8L2SNXmImM5Zi3yOR5f12MpkMLly4gMnJSXdtuY6k84quFArLj83N5XKu\n7xYXF5FOpzE/P49UKoU9e/bgqaeeQnt7uwPk0ejyMgYB4V1dXdi2bVvV+k+L9lXW59Vey0cSVAIM\ntFjnaPvhC+qtvXp6e3vxwx/+EE899RT6+/udXec6yLUOHz6Mzs5O/Mu//AteeeWVQB2Z3JD+LUdS\nVUt8M6rWxBUH79xOXF+utw/4hvkFCyjqoE8CWg5ExJ/pANNqY8vWcRDB9QZu4R9NCNwO0bNettUn\n999/P775zW+iq6vrtq+Ry+Vw4cIFzMzMuCd0iuh20XZYRONcSyoJSKz7cT/pzB/GLVbf6bG/nv2l\nyR65P2fCcf24/WQci97LdyZ65LM8bUsv6ZLsHrlmsVgM+CImejhDR75Lxi0vy2PCiJfFc2aDtfmv\nRb6tRgRbsP1iW8AE7/8P4os5ffacj7PiN2B57MzPz7vMYt895bNvPJUrk/Y/krH8xRdf4PXXX8fZ\ns2dDY7S1FIuQ4N+1vQg7n7N6eB8qjjWuXbvm9tuUe/LKDF8bWGW0pFy8K8fIu3XfsNjfV3+rDfQk\nKd/bJ+uS2aOzebRoxS+n3HIMAzzfrJXV2FY59D2tR64LAZDNZjE3NxfYCJTJHA1+maQq92KxFM8H\n5K3jqgVwV3Mdqw7lAko5j/toaWkJx48fx3vvvYfLly8jnU4jkUi480XRR0dHMTo6umLPHn1d7lMh\nfIQYuH79Oo4fP4433ngDn332GQCsWPvIxAIA7Ny5E9u2bcOmTZuwe/duPPjgg24ZCJdPUj0t4G2R\nLFz+2wnKKhEOODTppMtSiX7q34WQ2bFjB/bs2eP23AGWs9suX76M+fn5wD5GnNHT0tKCoaEhdHd3\nu+vpGWL5ndsauBV4cIDCZWpra8MDDzyAzZs3uxT3YrGIubk5zM7OIhK5lfbpc/h8f6tt10OknLW1\ntRgZGcHZs2fdZpiSvZLJZJBKpbC4uIiDBw/i0KFDaGpqCrSlvLTd0nXReski5+rZf22XhWydn5/H\ntWvXXMCZyWQQj8exc+dOtLW1BfaKELArfSKbT6fTaQwPD+PDDz/Eli1b0N3d7Y6NxWIYGxvDhQsX\ncObMGUxOTrrxXyrd2pg5EomgpaUFGzZswKZNmwKzsKKTxWIRdXV12L1794pstDvtv0q/345OaX/q\nIwp0f1Xy4ntokCuffSClVFreUPDgwYM4cuQIdu/ejWQy6c0ciUSWs6qSySQ+/fRTDA8PY3R0FAsL\nCyvO0VIpCLpdsUgdKzDg+nD7aLFssEWkcd1YfCSPRfjw8UzEMZhk38lLzcPKy23AAbe2M+UIn/UM\nSrToQKSzsxNbtmzB/v37HeEbNo70b/wejUYdORCG0XzBkfbFYcGCFus4S2fkXRMJTKr6JnErKcta\n9K3oNBBc6sjjMAxzSmDIvoYJHv4sLyaDBE8K7uBJKCZ09OPR+RHSvLxLE696fPJ33Ue+Ooa1ezm8\np/v3yxyfdyo61tDjLCxzRo+DsHbVxJjVvjpm1P/pFSZ6tQl/LxQKmJqawvnz5/H+++/jnXfewejo\nqDf5YS1iCxFNoLBdYU6Aj7dsno/oEcJUMrgjkYgbQxzrlfPF+n6VxNW6j3T7Wp99sYRVFl0e7bst\nwofL55N127NnNcbBalBJmdZgSoy8lZ7lM17WANVMKe/Rw4SPfvy63DcMGGsApAFiWKf72mc1x1VD\nfOCGgVGl9/O1E7dJOp3Gz372M/z617/G1NSUa28Bj8ViMZCBoQecdU29X09NTQ0WFxfxxRdf4Mc/\n/jEuXLgQ0AHL4MhSwmeeeQbf/e533SyPBJLpdDoAzqz28oFdXxutpWijX2786Dpxe5dKpYBRfvLJ\nJ/Gnf/qnaG9vd0sBgOA+OvIbpyoPDAzg6aefxtDQUGCPAJ2KabHx0t9yTek3CfKTySS2bt2K9vZ2\nRxCWSiVcu3bNERC8fM+yRXJtS9fWS6LRW4+dP3HiBN59912kUik0NTUhkUg4xydB8WOPPYann37a\nLevi65TTR663DmBEfOfLudJezc3NuH79Ot566y0sLCwAWCYAm5ub0dTUhGQyGcjwEnAlAYZs4pzL\n5XDmzBm8+OKLeOGFF9Da2orFxUV3z3PnzuFnP/sZRkZGMDMzg40bN6KlpcVdN5VKob6+Ht3d3Whv\nb0dtba3brJnJnng8jpaWFjz88MPYv39/9TpwHYX11LJJ8lmnBGvCwDfW9HU5sNIgJRaL4dvf/ja+\n973voaenB8lk0tl0LpMOxEqlEvbu3YtsNotXX30VZ86cMXWNP1vBSjUlzD6Wm7mzwLbGJ2G+gf/n\n/pWxIlhJA20mwK3+scge3nvL8m1cX411uD5WPXxtutZ+r5zw/Tdv3ow/+ZM/cZMPYWUrV+5EIoGh\noSF0dnYin88H8ILvej7/bIkPR/oCHa0D/D9ni/AsOvtxH75eb1laWnKfGVdw/OHTWx2fiF8NI3jk\nu2T4CA4F4LJ6BA/xU7N4E2Urs0dn6zDZw/2kf5PAV0SPPW4D3zFhevVVFwvHWZPzot++bE15zHe5\n9orFlh/QwhOKuhxiV7VN1TFLGNHDWDWTyeDKlSv4zW9+g2PHjuHkyZMO47J9Xg+x8AO/Sz1ZLPzB\n19D7XLHO8ybzHCOELdvyxew6JmXR+hPGOcjxuiwWAcVl4vbQRBf7cF/SgE/WhezR4gsUrO/WANVM\nvQAbq7GtjrOAIN/DWr5VU1OD6elpTE9PY2ZmxmTz5F2Xy1I2n5HQhse69u20+Z1IJYNA399yKvy5\nXHBYKpUwMTGBsbExxGIxlyYLBNeb37hxA9evX0c+nzf7Va6liTo5991338Xbb7+Nq1evIpfLBQg8\nvf45n8/j/vvvx8MPP4xHHnkEvb29DljzTI3WV90u5YIp/m+tQO9qdCkMDHAfaqNcV1eH1tZW93Qa\ni0Dhtm1pacGWLVvw8MMP4/Dhw4ElcXKuODlp3z179rgnfAlwlnJL2xcKBTQ1NWHXrl3Ys2eP2yhb\nwNLk5CTeffddHDt2zO0Pw/2m+0SPR8sh8LnVFAYBQlqOj4/j0qVLbi8BmYFKp9Oor69HW1sbWlpa\nHAnEelVOz6wxxcGmtqn6JWP2/PnzGB8fRzQaxfj4OMbGxtwGlpOTk25JnQCgjRs34utf/zreeust\nHD9+3I1LHsP5fB6zs7PucbOc3VcsFl1mEPeJPMEklUph+/btOHLkCHp7ewPp9byJqZwnAH8tZC3s\nte/aFuFjAV4LUGg90ZMXWqzgJBKJOJ0U3eBMEy43E6ul0vIGjPfddx/eeecds65ynuWT1yKI8el8\nOSBYSZl9ookeS5jwsQJ5iwTiLAE9c619lFVu+Z2BrYB6rR9a33RblmvztfSJltTW1qKlpQXJZHJF\nua1yWeXjcSY2jJeFW2WxMLHuh7C2qAQ3at8telIq3XpQgrzCyATrmpZvXCvhZYk+skfKwv/JOweI\n4lfr6+vR2Njo9oQUsieZTKKxsdERP3V1dY7oscaTZPPw49E5u0dn9miyhzGohR05IAwj5vldyiq/\nsb1djaxX/96uSN1kvz9ejm+RPYD9JGYdd5bDxHxvsYNcJsZxlg0UndRlYZJH3mUfsE8//RTHjh3D\nRx99hCtXrrhYhAkPuY+ILls1hLc34Xqyr+AVOlwW0V+ZlOPxIWNGlvZLO2pyh+/Nus331y8fjtVS\nTd+jx6jU3UfsWmNcXydsPK452aMLoBvLang+zgekOHD2ET3W/Xxl4wFm7d4diy0/refkyZMu2NDX\n0U6VZz1XC+juJtEgv5xihYEd/mz1O19XshBkY19ZdiXKv7i46DIyZFNVvh+3uSZ7FhcXMTIyghdf\nfBFHjx51a6+ZkNAzoIVCAYcOHcLf/M3fuGPFcfMabDY6fD3dPpbT1gQQA6z1Ft+Y0gBf2omNEwC3\nttxyavLO9e3u7sZTTz2FRx99FAcOHEA6nXaZKWLANdmzf/9+1NbW4he/+AUuXboUIH5Z2tra8MQT\nT+C+++5DJBJx5VpaWsLExATeeustfPTRRy5TSy9bsAKXSoOTago7NdHlmZkZjI2NYcuWLUgkEq7d\n0+k0uru73e+8LAoIEj06s0KLdojAraVynLmjXxLUnDp1CkePHsXMzAxqamrQ2trqQNj09HRAVwBg\n69at2LVrF0ZHR3Hs2DG3p4voAG+IyfsZiH4kEglHfsnvpVLJkT3ZbBabN2/Gs88+i0gk4rJ6uH24\nTWT2ttpSiQ5V4sgruQ9fzwoAtT32ET5yPekPn1iABUAAtOjZKQbcfJ9SqYSOjg7kcrnAcjprLFqf\nK/W9qxGNO8qRPFxmX4Cvy2+J9pVWP3LgbpWJCR85Vi/jkvMYYPJ9rQCa8ZJcWwNtqx5W/X3/r4X4\nrs0bm2qcwuLDPPq7hXN9+Eefx2OwkoxMvlZYUMrX5//0RsAW2aOvvZ4ED4texuXbR4rtFpdT/Kns\nAZdIJNz+PEL4SOap3rhZNtuWDBvOhhJ/IzgxnU67pdW8ObPsSWhl9vAeeyysE2xPLSlnU3yEfTmp\nxH99WcK6XygsP+ihrq7OPcyF8Quf48uqYT3yZWVUEm8ySeGzqboMnISgyZ5sNoubN2/i3XffxZtv\nvolTp04FJi19iQlrFWfmcrkVeNBqL0266kw1yXiTl4wX3jvM2lOXl29ZoolTEe3DRfSYKtfPleA5\nyw9o8tYicn2ETyWyLmSPdlZW0OebPZL/LcXhQEwrtNw3TPj+FoPKCuQbOGHXtcoepvzcLpV0YqXH\n3alYxEwl9w4DQD6QYgFnJk8YTOg+kfP19fh8CZBLpZL7LBvz6qdvsVFsbGzEpk2b0N/fj/r6ercR\nGhshcdxcVgtYhLWtb2ZmLcQKiFgHrX4IcxQCOrZs2YIdO3Zg7969aGxsXGGgRGTMNzY24oEHHsCD\nDz6IRx55BBs2bEA2m3XZWtJfnGmh2W4dQHKQKMtxNm7ciI0bN6Kpqck5es4a02OdbRJfT4A2gMBn\nboew73ciYp9kbyh5QllDQ4N7Kog4ynQ6jb179+Jb3/oW+vr6AmSPpXNaKgFy2p5pQCSk+NTUFKam\nptDc3Izm5ma0trYCgCOlFhcX8c477yCVSmHnzp3uiTctLS3YvHkzMpmMI2008JE2lr6Rzddl1lU2\nZy4UljdcTiQS2Lx5M7q7uwNL3jhtXkT0TIKfr5JUShhoG2SNKQYjcm0OOK17WfYsjOTg61g+81e/\n+hV+97vfuU0nfWSL/o+vV03x2UjrpQNi9oMiHHhUigUqCa71WGcy0wKU1qSDLpN+iUj50+k0isWi\nd4PMsNeXIVY7dnZ24tFHH8XevXtRV1dX8XX02NJikZzAymBPjmW90Rk4Fm4Oa0dLD+V67MskO0UT\nPhwI+frtyyJ8bkcikcgKbJhIJNzmy0Ls8BIu+Z0fux6JRFbsG6nHkCa4rUDOetKQzz9zX/mOWWtZ\nD5x6p3I7ZazEpvp+SyaTbrNubVMtcoHHrUUoaX8m/y0tLSGVSjkdunbtGq5cueI2Bg8jetZSfGSP\nHhs6u4cxh8RSnNXDmT3ArfiPt+UQHC9+k8cagBXjz+IjwuIcHb+uRqxYjycU9e96by7fBFmlsuZk\njy+YBYJpbmGzfCJ6MOhgOozoKef8NCmgX2GgjckBi2xYDcBZLejhe6+l6H7kelvl0ef5riG/62uw\ngdD7u8g9fOsx9WerX2UpydLS0gpiD7gFskRn29rasHfvXgwODronAulHaspLk1Kcwqjr6AuI1sKB\nVgIA+bsVyIhYRlDGb29vLw4fPoytW7cimUw6hySAUR+fSCTwwAMP4IknnsDu3btRU1PjggXuO0nf\nFBIIgDP0FvCV73J8R0cHOjo6nBNeWlrC3Nwcbty44fSAZ1KkDzXho9tTt6vVf9XqT559TKfTmJyc\nRD6fD2Qk8Yz6Pffcg8OHD2N6ehoLCwtmZg/XT+pTDiBY+qJ1RdpTQEk6nUZfXx86OjocmbO0tISp\nqSncvHkTX3zxBWKxGAYHB9HU1IR8Po+enh7s2rULZ86cwcTEhBvveszK/fP5PKampjA3N+cCGNls\nfWlpCZlMBg0NDRgcHER/f7/TBdlIXeuQAPhCobAmmT3SD7ptLT3ToMQ6RovuJ+57FisIEZBRU1OD\nhoaGQGBTKpUwNTWFVCq1wjZzvcJsGuupr9ysmx999BFeeuklNwbKkSs6KK22WDpvlcF3f5+dsIKC\nMAnTC/ax0tbaRvqIHsZYlv2ysIqcI0s0BXhb7eBrl0p+Wwv/yO0oDwl4/PHHsWXLFhPH6LKV89uS\nLcJBoMbCbEt1Pbm/5Bzdb7p9K6kzX5e/a6JHnoLJWTO63OuBRSsRbleN0fX/jDEkq0cIHf0kLsnk\nEaKHH/Qh/oLLoO2qxno6mCtH9vj0iycotFg4Xf8f1n7yXbdlufveLSJlFVynf1+NiK744pdoNOr0\nwtpbybIdFu5iO+CLGQuFAtLpNNLpNG7cuIFr1665pfJMpFg4bi3sp4jsV2URVHx/XlIH3CJ7eJWE\nED28p5XYP8GYMmZ5L10eW3I/i1jRes7lZd22Ylqfvff9xi+JL3WZdAaPtazzdgmfdSd7uNGlQX1B\nr5wPhKdp82+VOBrfoNKEgH4EswVkNYOp617u3tbsH18rrDPXcsD67lfJPdkpWEDIB0b14LNe8l8Y\n883l0P0qwfD//u//4uc//znGxsZWnGsFPhs3bsTzzz+PPXv2hK6N1lk8rCNSJvld2skH8Bn0VVs0\nQLfGV9jMNNdPl1n3l8xCLCwsOOKHA8pYLIZ77rkHQ0NDKJVKbhM8ds5iCMXYCwjr7+9HOp12G/vK\nS5whL9sRZyFkXKFQwCuvvILXXnsNo6OjK5ZvaqMsDkzrtjXuLedQDZHlRHV1dThx4gQ++OADXL9+\n3WX2AMv7F9TX12PLli1IJpNYWFhAOp1GNptdsQeAb1ZY656v/JFIJAAsOOU4kUigWCwim82ipqbG\n7Xugl0RJ4C7tKrOmpVIJu3fvRmNjI8bHx3H58uUA8SuOXgi/UqmEK1eu4P3338eJEydw9uxZd4zo\nwsLCAjZt2oSnn34aAwMD7tH0sr4dQEA/RZaWlgJBTrUkrF2rIVbgyIG8/C76wI/2lc/t7e14/PHH\nsXPnTmzatMmRsT/96U/x3nvvmaR7WIAh4ymVSiEej7sAWNtR2XeKg8ywl54YYFu8FgGo2Ai9r4PP\nbmpfyC99XDV9u88WWcGn9tOWz/b5XPmNgb6vPcoFOaxD2q9X055aEo/H8cQTT+Chhx5ymzKHiY/g\n0GXcuHEjNmzYgLq6uhUzziJMsOnrcAAj97SyETUG0uUI00NN9jLZo8lCuZYP03xZonE5f9dBqJRT\n/Aln9gjZw6QPP2qdM4KlzTSOtAJ+q6wWyWMt4/LFFPxu/V/O/lljTOrgu0dYwHs3CJdN+lUT2Jbu\ni1i2Ssg9WYon9wHgssF4XOr+9+H9cuUPk+PHj+Odd97B2NhYaF/z+FxLkXHAmYCCEaUtZBk+Ywcm\nNySTRz+9Ti/hEuynyR4hqzXnIO0gUol/9pFE/H9YX/FxohNyX8aYnNUkdlcnEvD+aZpELCdfCtmj\nDYlvAFpiBfR6QFYqfB6DRn4KFwd/zc3N6O/vx/z8PLLZrDu3WLy1RliXIay8YYSFFUDK+Xerca1E\nyoFKEd2vzAxrhx0GHsXQCOEDwD3J54MPPnAZEXxtbfyLxSJaWlpw3333oaenZ4Xx5jLrFMAwZ8ni\nC7q/DLHa1QJIGrjIcTMzMzh16hQGBwexY8cOAMttIYZKB5cA0NraipaWFiwsLDjjzkBKjhXGXwiC\nhoYGtLS0uOMY2InzECMpZZRjFhcXcebMGXz44YcBPQgLPDg4K9eGUu9q9qNk79TU1GBsbAxHjx5F\nJBJBIpFwWTS5XA7d3d3YsWOHI8NkXynLQbGOWmPJB+J9442/i/MCsMK5sx7V1tZi06ZNGBwcdICp\nWCyiq6sLdXV16O7udgSQBB25XA4zMzMoFovuv8XFRZw7dw7Dw8OYnZ1Fc3Mz6urqHIiIx+Po6urC\nzp07UVdXF1jCxVl9ojvSTpIl9GUI2wT+HnaM1Sc+3bWCvI6ODrS1tSGZTGJgYACPPfYYdu3ahYGB\nAQd2h4eHUSwWkclkMD09jRs3brh20sBJhHVL9FFS3vk4IXtyuRxGRkZw6dIljI6Ororwkbpzu1RT\ndAZPGKlh3d/yezwxIGIFYdZ/Yb/xWGd7ZmUcyPGWvmk/Zvk1aQ+5l24jn321cFK5eq2FSHbhzp07\nAfgnN1gqwWRNTU3o7Ox0AUwY0c51FfvM41MwEJMBt+NrNN4Rm60zTniCQJdZxNKFXC6Hubk55+cX\nFhbWBL+Ww90sOujVkwdM5gi5I4G8LJXWT+QtFAorNsKVe8n9NB7Vk8ucPcpxSTmyx6q/D1/73rVw\nxqU1WcnH3e0i7S8TQnopu57cFbEm8qWveEm3XIcfHMH4wZfZY9lVXWa+vi8uvn79Os6ePYv5+flQ\nf8P3DIux71QYS3C2jvY5PA4YrwuGZZJHCA7pE156GY/H3YtjBr6Xz0+WS7IoN3b4Pj4OQ/+mJ6Xl\nOCa7NNljZVm2traira0NHR0dFfXnuu7Zw43C/1XKTJUDBtx5cm3fNbiT9Wwxr9/lfTz6+/vR3t6O\nL774AlevXg0ESlYg4AMrFjjUrzBAr6+3VoO22mIppK+P+DO3C58XNlj5XCbtZCmHBHfWDCwQNFjC\nUDc2NiIej7sUdYvssRy9Jnz0Z1/bhBntOxU9Bn1OyCoPjyF9fjQaxYULF3D+/HkMDg7iscceC70O\nkzjyEifJy7Y44yCbzboNhwUMcVAubaaBqjgE6VN5MhUDL6kTAyyxEVxX7lNfYFSuPW9HZA8kWa50\n6dIl9PT0oLOz0xErmUwGQ0ND+LM/+zOXWcPprxZot4JyljBnYgVt0m9A0IlJP0pfCXkXj8fx2GOP\n4cCBA26fnaWlJbcpZl9fHzZs2IDJyUmXZTM9PY3JyUmkUimX1SQzsJIdJNkXkrnT29uLvr4+t8mv\ntAvXj8c9BzlfFtlzu6LtIwCnxzxONAgpFArYtWsXnnjiCWzZsgV9fX1obGx0GQn5fB61tbV44YUX\n8PDDD+PChQt4//338etf/xqZTAYAAraB33XZCoUC5ubmVpSds6g+/PBD/PjHP8bU1FRgrFoZNVaG\nzVqJhSUqJX1W6wfvVCwQagFV61hfsOEjoTQZYBE+/D3M7vz/KD6CTQfUrLvcV2LHwwJJljB8BNjB\niF5OIPbCh1P0Z3lfWFjAqVOnXCCay+XurPE8Um6JrdUGrI+c1SOEj+AF/qwngXlPKk108z14mZjP\nFko2rM4IYbzBn1l8cZVe9qGDUvlu2TLBc9qGfhUIHhGrruzndNswrtNkD7eJYAw9diyiVH7nPmQJ\ns3lh2DEs9ijXJmspli2RrRT4d+3DmehggkNnLwoeZ6JHHpijM+14KZeIFcdYMaDVZpY/1Hpl+VnL\n1zLZo3WHYyJN8shxQ0NDePrpp9HU1BSKz0XWJbOHP2vwFxZA8Lv+/XZEX5M7mll2K6snGo06ln/D\nhg2Ix+OIxWLIZrOYnJwMBAPaKep3CwD6QKE+TwODtR64IhzcAyufcMASBvAthyvH8oApx5Tqe1mE\noQaXEvhJ+XWwYLUv15efAMQ6I86SryPn6DLqttDtIt99jqFa4qsnS7n/+H8GDa2trejs7ERDQ4Nb\ntmXVlR2oBjFWdoClCxqkyPFi7IVEOHToENrb28ueL+XRIIiFZwqs/vHZvGpIJBLB1atXcezYMXz6\n6aeBMgqRKY8Y7e7uxvT0tJtZ1c7FagffmLbq5rPNDG5kJjSfzyOTybhMmrq6OkdMScp8e3s76uvr\nUSzeehqDpE3v2bMHmUwGb7/9NiYnJx3Jp58Uk8lkMDEx4ZYIybLNbDaL5uZm7N+/H7t27UIkEnGO\nlG2HBigMCKtJHKx23K3W74X1n7a1nAE3ODiIffv24f7778eePXuwadMmtLa2mmOiu7vbPammWCxi\nfn4e586dcxMhHMjzOJqensb169fR3NyMRCKxYozk83nMzs7i+vXruHTpEt58801cunRpxSSMDris\n/llLIiHMd5fL2rXKIm3LZbXGmGW7w+yrJT6bqm1VmA+2jmciQpdbE3J6f7EwO1ntvvPJ1q1b8eST\nT6Kvr29FXcPKpttG+we2KxzMc19KwM94QSY8RNhPSvYUBxKWHlhBB5eN62jZB2tSS2Mdvgb/Vltb\ni7a2NkfGRyLLpMjExAQWFhYq7JXyop+Oy+XhcvvGqs620RkDnO3Dy4d5qbnVJj6cKNeQhwnI5Ea5\nZT8+YpazVaTPfEvCZIzqJWIWUc1tJ+9flUmPMBun25PfgVv19J0nWamyX6RgHp44CotXRKxYSAuT\nTty/NTU1qK+vx44dO3Dz5k3cuHED4+PjAftQDp+uh7Dt88V7QvZwRgsTHGwnGQfwGOUYXmfa6Xb2\nTQxVEnPocahtJY9dS4+YVNTtwxk8/EQ/6yl9iUTCPWSEtyjxybpl9oR990m5QaCvGeaUZaD4QJlO\nq2RAqQ3fwMAABgYGEI1GcePGDczPzwc2lLXKZTF+PoDG5b0bxDJIPlCp+8Bn3GQA6ACG2lA3AAAg\nAElEQVRdzypp1l0Hq/pcn3Nnssd6XJ/0sWbzrXLLtQRgyAy1bCLM7aAHuAZ3VhAg7VIOZFZDyo0b\nPo4ljAS555578PTTT2NwcBAzMzOoq6vzBj7WNaWfmDy19M5awiHHif40Njbiu9/9Lp5++mlHIjDg\n8hEbmpgrl6mgQe5q2nY1UiwWceLECfzoRz9CPp9He3u724BYnsCVSqXcXjQCInnPBV+5fOOZ6+MT\nfY44JMmwyefzbpNmeSqYPIq2ubkZXV1diEajyGazDkzJUr6Ghgbs378fdXV1OH78OG7cuOEcnAQ8\nssHy9PQ0hoeHMTU15TbjLhaXlxp1d3fj0Ucfxfbt21EoFFzGkybvmADhTYqruWePz6dVI6jVAZi2\nmWyTOM16cXER9957L/72b//WzUA3NTW5vZf4egCQSqVQU1ODgYEB1NbWoqmpCa+++iquXr0aSlqM\njY3hwoUL2Llzp5sd5bIvLi5iZGQEb731Fv77v/8bExMT5tIJneGjbYqlr2tBGviCSLbx1uyh5eO5\nfVerI5a/siRMJ8oFRTow0nZCB00MeHVArbNSrPaz7Opayr59+7B3795AoOTTI91X2qZycMbBoN7P\ng/0LZ6iIzdaBQVif+PyNPg6ws0E0rtJkt8ZAWs84WC6Vlp9iKsvh5PdcLodPPvkE8/PzlXZLWdFk\nj64r3983Nnm/Ob0fCC/p0hn/+roifF0AASJJ+5hyT9yx+lj7KtYvDhhzuRyi0Whg6ZHG2FIHHqds\nn8Ls+VdBLD2Xz7q9ZUxqvCRSLC7vN1csFtHa2rpi6Z2IxqxafBiYyyll4Hd51dbWIplM4sEHH0Rz\nczM++ugjnDx5ckUM6iPn1iq28Nlx0TeuG8d1orec4aNtIGfIMQEbj8cDuuvzJ9IePFGkxzLrBo8z\nbjdffMcxLLeF9ptMsmocFkb0WBk+su9huVhjXcieOzEMVmeJaOfkO9YCDWzsdfAPANeuXXMz5HJe\nV1cXent7A2VoamrC9u3bMT4+jhs3bpgBlbx6enrQ3d2N9vZ2s23uJDBcS2Drm5W0+kMzixr8SLtL\nXXVwrme+rDRla5CVC1ylr99991289tpr+PzzzwNBA++5ZJV5aWkJ8/PzqK+vd0QPO0eLpBLHK+Vk\nhppBgryLwWFns1YGmevH71xnbmMul2+McXvofTisgIf7RZZvSSYHp41LW3F7xGIx5HI5TE5OYmRk\nBEtLSw6EsUGVGRiujwbyWpe4PlxG1g3L9qxHcPlP//RPGB4eRiQScU8HkQBZ2rCnpwctLS2BJUpS\nZnn3EaOWaHCkRcYr97EECvJ0uqeffho9PT04deqUs61CTD366KM4cOAAenp6EIvFAsskBayWSiW3\n9E4yg+RekvkBADdu3EA2m3WbQMt12tvb0d3d7ZZy6swdaQN+120gDrwaEqYXYX2xmmuz3st3C5jk\n83n09/fj8ccfx5EjR9DZ2enavlAoIJPJBPpdxoI8aalQKKC5uRk7d+7Ehx9+uCJA4HKVSiV88skn\nqKmpQWdnJzo7O93/pVIJr732Go4ePYrp6Wlcu3bNLf9goocnYTTRw22nfYz8Vi3x+Rg9gWSRPJX0\nr4/o8LVvuWtbwbyFOSx74fO9Vpmi0ajbd0smQnQgyTZVfItV/vUIKnWAXuk9uT/D9KqjowO9vb1o\nb29fQfj4/KOvnzhw0MEf+0opl/bjPlvu0wPrd63nfLz1mUlywTnV7lcmeziICpvp5jHL+3XqrRys\nDALWYbkXX5d1QxNCjH81vuU+sgJ//S72W8/884MpampqHOEjWEvqoCcgLfHZobtVrLaz2s168VI8\nJiI4PpBXoVBAKpVyY0x0hTEQ42Yd95Sz1SKiW2JHtR+JRCKoq6vDxo0bsWnTJoyPj7tJTSbSLVmL\nftV75/myZrSu6/1qdCaanK9jdsunWPWUcahJ67CYls/XdlDbXh2PcqwR5mt9vtW3BFPrsb6WT9ac\n7KmWMulMEBFfI+p7W4GpBiAyWIHl2ceRkZEA27dr1y50d3cHnGVdXR22bt2KSCSC69evmw5UXh0d\nHdi9e7fbeKpcEHW3CLeTDnqBlY7I6iPL0eljtILLoNZ9rGc8rM9cdi7/8ePH8c///M/ucZpsLOTa\nFiGRTqdx9epVxONxdHR0BIINDirF+YrR8g1ObhO9NGwtg5NygYIWDex9swRyvjbouu912woLPzk5\niZs3b6K1tdUBLrkOl0WCTJmpGhkZwfnz55HNZh27LyDaemShZdQtY2sFcBZw1mUr18Z3Kv/6r/+K\nRCLhnmwlZE8ul8P8/DwSiQQGBwfR3d29guyRcrLj0O0RZl99dpYDDdFhAbKSnfP4449j48aNGB0d\nxaVLl1wmSbFYxMDAAPbv3w/g1qa90l+yvEj+43EmZMPRo0fx+uuvIxJZfpJXa2srGhoaUFNT4/Zl\n2rBhg9tcWD/NQNrIAubcZtXsz7BrVUoGVHJdbR8Z1DIh3dPTgz/6oz/C/fffj/r6euef+AkYImKX\nhZhNp9NoaGhwGwbyPbkN5f3ixYuora3Fs88+G9C7YrGITz75BD/72c/cBrZM8PC7BSR1n621aB/h\nIzSsMvLLKqsVsOj/dVl8x+hxa41n6zvbOwuMcvnYnkQiyxvGy/lWUK2Xcvnw2d0sYUEUS3t7O3bu\n3OmIaw5iuM0AOJsGILDXhPZNEkT6+tJX3jsdFz4d1veR99Xg9TsRzohinKjJDNZdrY880+8jfDiY\nZ/ziw1MaR5RKpRVl5bbg6+jA1Qr6eAmuXgKSTqcDWUg+fFaJ+Oz53SqWXbNsmCZ6mFgoN7aKxSLS\n6bT7X/b0ZAyksaT+Tfevrw4ifA5j0kQiga1bt+LatWtIp9OYnZ0NnZzlOsrnagnHij7Ch22+tIe1\nCTFPxrGP1Uu6w7CAbjdfLFIOk0lZ5d3SIR0v+XSPxXdN/VnrrrSJPtcn65LZsxrRDa8/83s551bu\nHmzs2cBHo9GAEunO040P3DLEPIB8BsYCX7oeYXXyAcS1Fh9ItZTX5/y04+Vz+FVTU4NNmzZh69at\ngXTfsEEs3/mzNjy6v63MHikPp9yePn0a//iP/4jnn38eL7zwglueYqXdSiCrg0W9n4GlE2sBhHwS\nZgxFfAaGHQkfq9tCxAJWsVgM8Xgcc3Nz+M///E+cOXMGjz76KLZu3Yr+/n53zWg06vb24LXv+Xwe\nX3zxBX73u99hZmYmMF65LMlkEo2NjW6JkOz3otfda5ujHYjOCJJjtfPWul1NR9rV1YV4PI7Gxkbn\n7GQWb35+Hu3t7fja176GXbt2BcCLiNU/HKyGiQ445Fy+JjufYrHoZhETiQR6e3vxgx/8ACdPnsTn\nn3+OTZs2Yffu3di6dSuKxeKKx/yWSiW3kXp9fT1aWlqwefNmZDIZZDIZRKNRxONxt9QoFouhrq7O\ngS4OJA8cOIBDhw6hq6vLAQvdFmGZO9Uek5UAi0qOX025wsAukz/Wy/JNMkHS0NAQyKLzgWTRmwMH\nDuCxxx5DV1fXCnDy3HPPobOzEy+99BLOnDnjnYHV7RFGQoe1352KFdCFvXiJqg+YSTuFAUM+jn8P\nC06kzzVAtO5R7ncfrmHhiQsJjPVLZwRaPsiq75ctt4M55bxyvl5jK9Z7vj/7Iwsv6/JauJptvw6G\nypE6XN4wTLYewviNfZF8t9qLA1FeEhqW1cMZ3VJPJo+sttCbR1ttz+SR9Ttg4yve0FYediCf5elh\n6XQamUwG8Xg8kO3Dk2Y+PZT7yvdyAeXdJGHxhcbdOsBmMpVFln5L5oyItv+sEyLSbj6sbcV1HF/K\nu1xbPicSCfT09ODIkSNoaGhALBbDmTNnMDIy4vUXjFurLVZs7cOXgrskbmJdtiZpAXu/1bCsHi6X\nnF+pbbPEit8t0sc6Nqxc+rdy5YrH42htbXXL7C19ZflSyB4d9PsavhLQqysXdj0RC4CJ8eNNOyXL\nx9exDGAl40P2pJBjS6WSC87q6+tXDc4rlbV0rmHtWYnB0AZNzit3jqyTTqVSK+5TTje08dXEjkX6\nAMHZFO7jiYkJjIyMoLm5GRs3bgw8nUYMkmy219raisbGxoCh14QPg3kdxFhkYrXE14bWuLMMVhhQ\nlbrU1ta6zVelTUulEtLpNLLZbGDMRSLLS5A+/vhjTExMIJ/PY2JiAoODg+jr63PkRiQScbNqQgxk\nMhlcunQJp0+fRmNjIxKJBKLR6Ird99nhCNkzPj6OixcvYm5ubgXh4wO9uh0ractqS2dnJ+LxOOrr\n652uCEECLD/C/t5770V/f/+KJ035HJBVRx0s+t75eO1sBIzKvkGNjY04cOAAmpubkc/ncc899+Dg\nwYOBPWN4k0qeuYxEljf+vu+++7C0tISTJ086GyF2mtdyA7eWl0UiEfT392NoaAjF4vL+PdbsiQYD\nGkBUcyyuxmb6fqvU9lqBo3yORCLo6OhwS9z0Mlq2WXINua/sZZRIJNzscnNzMzZt2oS5ubnA8i85\nLxqNYteuXXj44YfR0tISuHYkEsGePXvQ3t7usmqXlpawsLCA2dnZFf5H94cV/K6lnD171mU0+Ygd\nC5BapIYeNz7x/acJIguQ8jGW6P/CiB09+cW/cdtzEO4j/HnjTm1/fYRTNSWsTcrJagImX6DJ5bA+\nh9lvadewdvH9ricydB9YeC8sCPEFc/r/avcfsDKA0/eXfrLqaMUBmvAR0sfan4WvrX2HtcSUx4Jv\nMlJnDwH2U310gCzLnOXhBnrpK38WTCQZtnxdjUGlzFzfr4L4xoUPC4Xh7kjk1kMw9HiRNhMcJv2n\nz9djV+sQl836bMUJQkRu374d2WwWp0+fxtjYGK5du2a2iZRjrfrQsvFMvOq6yGfBf7w5s+AQ9u2W\nMJ4BEMAvWo+tuIsnr+VVzk75YqIw/Slng3TcatkNtgucCVrON35pmT2Wg/EZbPlfi69xdeNY//vI\nntHRUVy/ft3tNQGszNphZQGWFauzsxOtra04f/48hoeHA+Vsb2/H0NAQGhoayrYJ16tSuROwshrx\nKbB2pJZUMoC4bZeWlnDlyhVcuHAB8XjcLVfR54cNIA0wrXWevIRPyqANNevIRx99hKtXrzonzjuv\n9/T0YPPmzXj22Wdx+PDhwOCVa1uPIJTrcBuvBagNE1//cJ9YzLkcz8a1ra0Nu3fvDmTh5HI5DA8P\nY2xsDKVSKbApmrTl7Ows3nzzTbz//vuora3FD3/4Q7zwwguB/pSNdScnJzE2Nob5+fkVgEoD2Fwu\nh0wmE9ic+d1338VPf/pTTE9PuydGcbqzFmvci23xgfSw325Xurq6AgG2LHcplUpuT7GGhgZEo9GA\nblov3VZc1kodGNexVCoFZmJisZjTd3nyFrBMSP3BH/wBYrHlx6ezgxZyioNHAbTNzc342te+5mau\nIpHIik36RKey2SxSqZQDDnNzc1hYWAjMZLLz16RGJBJZ8WSc9RqPawHCtF2R+tbV1eGRRx7BI488\n4p6qFTZDJW3my7jZuXMnXnjhBfz+97/HyZMnA+cXi0XEYjH09PRgy5YtiMfjgU1DZfw2Nzfj+9//\nPh599FFcvnwZR48exZtvvrkCvAnxy0t9uf3Y/3MbVEt+8pOf4L777sNzzz3niGbGFNZny1ZoQL/W\nYFyLtvth97VsAvc/7yvBftQiwTTBbrXNemGb2xEdPOjffedYRE8l97Iwrs4c0AGkDqrkXIt81JNT\nlWK61dRhrXRa+36+JxBcIqNjAsaJGiPqZV28vFwH7XJvIVW4DX3tLeOAgzcfMWyNvbAJTK6/1Nuq\noyzVlSwhPUEk9+dsqa+K+AJ4HS/oc3zX0H0D3OoXwRq8PQSfr3GU1lPxqz68ZflijiE0VimH19ZK\ndHzFNl58tpU97CN6pB68PYDoKpNI3M6yZJb3/2EbsLS05Mqmx7o1GVOpWONG+7Uw28v2WibTZJsQ\nPc6XlpYwOzvrMtvL+cp1ffQ6sLpgyDLa+nsymURzczM6OjpWDGwfiPABD9nITGb7Gezr2W898CRl\ncmRkZEU96urq0NHR4TpIyuEro/xfCRgIa6O1EK3ElfSv/l8DI999UqkU5ufn0drausI4hxlpbZg5\nRdfnHMXhchDMfS+v6elpTExMBECbDM6Ojg6MjY2hra0NTU1N6O/vR1NTU0BnOLDVqZlrLZUYLg0s\n9H8Mati5cIZTTU0NWlpa3Gz/4uIiJiYmcPToUZw4cQKFQiEAcqRdJKtHjP7ExIR5/1KphOHhYbz3\n3nsYGxszSWO5fj6fx8cff+ychxB77777Lk6fPu32vbGeriHtIN/DnHVYG1fTsba0tLg6ZDIZt+Sp\ntrYWu3fvxp49e5BIJJxD5JkUq/zWWKqU5PGdzzMxpVIp4Jhl2UZ7e3tgDwt9L00S5PN5JJNJt8G9\n/F9TU4OGhgY0NTW5DDDOTKmrq0NDQwPi8XggU0iPNwZNPHPCurleGzRX8n+lx7Bookfq29XVhb6+\nPjQ2Nq7YI63cS5M9LS0tGBwcRGNj44o+7enpweDgIDZt2uTsovYHkcgygbdr1y5s2LABvb29LsPz\nzJkzuHr1qquL9A9wK5DR+sizfdWWN998E7W1tXjmmWe8viVslk73i9S/VCqtsENyXDmfx36P76M/\n6/MFhMt7OQyl8Y+ecWaCwSqbj+jx3UfudbeJj8DQdkz/5vuf7VI5H8wZH/KfLoMvg0j3g/bFleDT\n1eD2tRqDujw+PWJ/xL5cn2vhQv0CgpiN21OCM3lCpiZ22L/ooM96yXFyT427xAZyXVg3+Httba2z\npXV1dchms0gmky4ozuVyyGazbgLJl1Vh2aa7TcJ002cTw/Te12/ArXHK+IB9k9hzxmIsPO6lrS07\noDGohZkaGxuxe/duXL9+HadOnXJ4S7fJWo5La1Jd10X0Sz95irO7uW48fovFosOTABz+lfoI5tTE\nEfdPLpdz/SvllGM4NrHITZ9NtET7Pa1DevxK2cX+yGcLVwC39rhkv+3tl0o6705ltaTFaqSlpQV7\n9uxxjx7j63FDc0NoptFixfX+HxaA4rppJ6nrXc6w+H6/GwHOaiSs/GFBpLzrvhJhY6gdu+57q2+1\nI+eZkEjk1hIFfV3LyIvjT6fTuHjxottD5rvf/S727dvnjtUGjOuh63O3OFLWawBmPzAwsN7T6TQu\nX76MX//61zh9+rQDQ5pcYSML3FoCoIFMPB7HuXPn8NJLLwVSGbVRra2tRS6Xw8svv4xf/OIX7pq1\ntbUoFApuKZ6UxQKKGqBb4DtMqt2XkrXDWTPF4vK+RI888gj27duH2tpac2NdroMPyFv1ZfBhAWqf\n3WNQyhk+8i4ZSul0OrDHlWX35BrRaNQBBFnWJ3syyePCxZEvLS2hvb0dHR0daG9vR21tLbLZbKBd\nuO4MpnmcSxuvByFbTizfshrxkTTRaBQNDQ2oq6sL7JnkA5VhhI8PkG7fvh3PPfccNm/ejEgk4mZA\nLT8p4zcajaK/vx9PP/00fvSjH+HChQuuDHyugCYGaWHtVw05ffo09uzZ4+6ryR0f4WP1hy/4ZF+j\nxwWPOR28W0Gj9qdM7sg1NOEZhlXC+lrbGjnHF9Rye1j9ZZEjd5NYNsvqW/5PZ85JmwIrJ4B8mEb+\nA25NIIWNW74ejxWewLH8g+UrfL/xO9f3y8A0Pjxu6ZGFIaxAi8cpj0vBIpLVAwSXymmCU//vazMr\n8Od+4rLq/pQyyxYDem+fXC7nMp9TqZSrn0xKayKRbcXdglG1lMMn/Bsf5/usx5625Tz2GTswgSq/\n++yAfJbv5fw82w++d3d3N5555hnMzs7iN7/5jWl7wupcDdExlrYt8hKCR3SRN2jWesb4gh/aIXGV\n4Eupj960nMkeHdNIGSUukH1axb4yWcVt5vPlvmMt+81kD9sRHa9ynKVxhaXfZr+sphNvR7SjCVNe\n3zktLS3YsWMHJicnMT09HTguGo26gE0CBt2wchywcmaEB6YmA6LRWzPRFljS12NjoMvI9xSQpZk+\ny7H62ouNQZhDX0/xKbsF4MrVy2JBLeMbVkdu85GREVy4cAEnT540nTiDahEOhLRT1XXmuo+NjeGD\nDz7A0tISTp8+jQcffBAdHR2Bx0WL09WOQqdir7doQ2g5OSZB2dhq4yrZcfK77Nkj+xlZ/QrY7R6J\n3NroUNosk8mgvr7eTBVl0CUbGEvfLS0tuX1vZL+ecoBV64Xch6WcLlZDhNhinRXyorGxEU1NTW7d\nvk6D5bJY9iwsUNDncZ0sW6vrK0SNOC7J6GH90GOdAatcj2eDpP4CZmtra5HP51Eqldz+BTt37sS+\nffvQ2dm5YtbIsi1cV84U0cTZncrt6IPW0dWUh/tTL7np7u5Gf3+/2+uoXPAuopfvFAoFt4l2Y2Oj\n06eenh7s2LEDhw8fxr59+9DR0REImKwyAsv92tzcjObmZsRiMXzjG99ANBrF0aNHce3aNa+/Eb1Y\nzfi8HVlcXMTw8DBeeeUVHDhwADt27FjhUyohfLju/D2s3HocWkRPGPHD12V7zve38IgOQrReiR0R\n3KSzrXzXs+pltdFayZ3qhhXENTQ0oLu7Gz09Pd77cdvJd+kPaymD9sXymyYA5Fqcfcz3lhfrp3Ud\nq6/CfADjVp+er4VIsCdtx/XlpcFcDm1PdSwQNnb5Ptw3YfrNmW58X+tY9n/8m5Sd62FlGUkdpC/l\nuyydTSQSAeJHiHXRO8YPur/v5nFYyfV8WC/sc9h1fLrus2+6/WQc6wk1y6ZY8R4fF4vF3CRmWFta\nelgtsTJjtJ+Xp22J/uXzebS2tqKvrw/bt29HKpXCqVOnHC6cmZlBOp125xeLRTdhyPdjLGs9jReA\nm8wSPyXXZAwpe13pdgrD0j69YXxsYQO5Lq+MkM++jOFYLIZr167hN7/5DQ4dOoR7773XJBNZ1oXs\nCSN4fIrL/7W2tqKrqwvnzp1zZI913UoAhXaWVlYPZwsAt5Yl6IASwAonIL/JvfS7Bb6sl8+A+Nrx\nyxBfEFAJQcXf9f96gFhtHGaE+X/p4+HhYfzHf/wHzp496wXiWqx7apBlydzcHCYnJ/HJJ5/g2LFj\naG1tRXNzM+rr6wOkEgM6ACsc9lqK7i/feJJ33RdsnCQA58ylfD7vls/Mz8+bxF0Y2SPXyufzKzLt\nhCSQDRN9M+ry5AxJ8eT/hAzSIFfX29dW3EY+PdbnVkPYqUlbiUPjp2vwOmVf3Sw94+CjEtBvXYuD\nR9FpKUsymXSbHApBICSMkEFM4kQit4gtBglSX3GIQtox8ZhOpzE0NIQnn3wSqVTKPUnDSr/XNlcD\nlWou4bod8dm6cvpl6S6TNLFYDH19fdi4caNb6qYJQm3nuU/1Zs7t7e1obm52mxYDwMaNG/Hss89i\n//792L59u5k5ZOlpNBp1BHksFsMzzzyDbdu2YXJyEleuXHEpzD4/b7VNNUFtJBLBlStXXIbhrl27\nVpA9vleY//LZY7mn5SfLET1WxhOPUx/Z4yOn5Bh552VcQiLyxNZq+kK3DesE9/fdJjrzr6mpCTt2\n7EB9fX3gd+0ftJ2WelqblFoYlAkii+yxbLlPJ3k8aj3TZbf00BLtL9dCKiF7LB3U9sbav0MHkyy6\nLyvBVVrCxgGPV1876/sw7pU+l/pwkC1PPMrlcm4iTf6TDNi7IZu12mLhmEp+09dg8cU0vt/kHI03\nmFyTl295jhU/lytfuTJVQ3TWiY55OGYQsicSiaC3txePPvoovve972FiYgIvvviiezjPqVOn3FOZ\nxWZa8SHbMcYlTIiKXxKsKOWSSeB4PB7Qfe0Dua8tnWEsadlbKzNHrlmO4OHYdXh4GKdPn0ZbWxvu\nv//+sjHjupE9YeSMdRwfOzU1hbGxsQDRI+fre1nX1w5KM/g6m4eBUalUCjSwGEUO+GtqajA5OYnx\n8XFMTU05tk6UZXZ2FidOnEBfXx96e3sD5ZVgjR2ynoHgWR8NYrmcXxbpI2KVwerPMEPjGyDWICon\nfJ35+XmcO3cON2/e9A4iq7za8OqXjxASnZmcnMS//du/4cqVK3j22WcDT/Gylj+I+Az8WohFBGjj\nqYkZZsYFIAwODuLrX/86HnroIbcZeT6fd/umWOmQIrrPY7EY3njjDUxNTeH555/H/fffDwA4deoU\nXn31VRw/fhz19fVuY15NCDHQkfrpTD7+L8w26TbSwVo5fV6rccmOU2dW8eyFVR8NJOV6HHhbOmg5\nOHnXwFADX3HwANxyIWu5WSQSCYxJKafUp76+HkNDQ8hms85ZyzGygbfs9yUkj96nh8e46IMEVBaw\nr9TmVEsqvRf7T5+wjrPOWE/d0sSLDugZoOrU4qamJnfOwYMHXZ90dnairq5uxVjTts/SV84QEWLq\nBz/4AXp7e/HLX/4SMzMzgbbSNqYcMX8nUiqVsHHjRnz729/GfffdZ9owC2NYJKzeU0Cur8egttEW\niAzL6GE9tgg3H3jlc3hygvWIgTOTRcByvzBxEUZCcLvImNR6sZ5j8XZlbm4OX3zxBfr6+rBhwwYA\nwX4T38C6L/WUmWshp7X/4WBeB438buFqS0+lPFawqfVBXuX8Wqm0vO/ilStXUF9fj4GBgVAMcCci\nfkV0hW092y3WSY1tLKzj8/O6z3jT2Gw2i0wm4/pPj7FyhJA17nSfWThE+hJAAP/wk7ikjGInZDmM\nbNYve/hI1rlcR4/Zr8L4Y7HGR7kXYI8fHiPs04QokEkoi2Qs56d9cTBj2bA418It/L5e4tNhIIhb\nl5aW0NHRgf7+fjz11FM4dOgQmpubUVtbiz/8wz90bfjb3/4Wx48fX5HBY90vGo1iZmYGIyMj7ulo\nPO4E68lvUkb2ZXwsbxMh5ffF22F6ZtkYHXdaZFCYn4/FYsjlcpifn3cTnj5ZV7JHS5hj4mPm5+dx\n6dIltzSEB5llhPlcH3jRAKxYLCKVSjlwqZ2xHJPNZt1GZzIAY7EYUqkULly4ELe7f6EAACAASURB\nVOjAaDTq9iyZnZ1FNBp1G9fq5S5SPu2QrXbURka3x1pIGIljGSc+x2fgfL+HAQ35XKnIeZlMBiMj\nI8hkMmhpaTGBsFU/7dB5Ey9dTgs8zM/P4/XXX0ehUMDOnTuxdetWtLe3r0gvDBsnayVW/1TqBAGs\nADmdnZ345je/iW3btrmgRvQ7Ho+XnaViwFJTU4PPP/8c58+fR19fH5qamgAAH330Ef7nf/4H2WzW\nLcOSDB0eszIGff3DwsBQA0KrvTTQWm/hMa/7gDels2ypCJffAp86wJB3a1xKgMbny2e+brFYDABN\nKa9uf06j13UtlUpIJpPo6+vDxMTEisd75/P5ANkjm1iHLcPyjV3rmPWQSvXKslk+0QRLoVBAfX29\n28uIAyTd/xbQBW4RFDJmZU+ISCSC7du3o7Gx0VsurSM6+Jf6yfWLxSLi8Tja29vx9a9/HU1NTRge\nHnYzfr4+u12/UYkUi0V0dnbi4YcfxsDAAIDg3jlW1jCvv/e1sXy3CA4NHPVElCZ8+FjdBtKu3K8a\ne2ibyeXSPlFvtC5jmfVFzvH1tyZ75J1/t/r6TkTbwNWeq0Xqnk6nsbCwgNraWmzYsMHEMkCwHeV8\nsZU+Ipzto0XSyXX0WObraAzEZbtd/KX/W1xcxNTUlCPbORipJtbhLRwssge4RerL57BxZbWBNVa1\nD5asmEwm40gT38Qe40vdxxqXSr9bwSDbF8DOiBUbL8updb9Lhq1s4JzL5Zwec/2+itk+3IZs36Sd\nLYKTxRp/2s5FIssZ5MlkMjCJUgmmrxT3hx0jZVhaWnLYx6q/dc5a4Vif/+UxUygUsGHDBhw8eBCP\nPvoo9uzZ4x7G0dXV5YiZhYWFwJOyrLiTx8Tly5dx8+bNwH05i5z7holuJjOj0agbN1wXK8YtV3+2\nKdb402WxjvO9ZIuMpqYmxyVYsu5P4xLxKZcGPTK4tJG2SBE9O8L3YvAhLLcwsYlEAlNTU7h69WrA\naeTzeczPzyMSiaChoQHT09M4efIkhoaG0NvbG1AKWe+ngwI22FNTU/j8889doL+0tISmpiYMDg6u\nWJIiZJCINhz6s3ZIayEa6DFQZwDHx4tUYkwqITwsUOx76f99ToqP4T1Bstms2/9EUl7lxXrE+smO\nltvs+vXr+NWvfoWnnnoKR44cweLiotsQzEf2cL+ulVh9ZBkoC/hYYEfajo+Tzf/EIOtghIWdaTKZ\nBAC8+uqr+OCDDxCJRDA1NQUALqvHMpr6WhZ4svRCG2efTlnt59PZcv/fibDOsm4KKLPspPxmZeHo\n4EDO03poBQFsCyw7IC8pnwQzMvOiQayMLw2qEomE24Ca9+CR+zHZMzc3B2BZVwToWqLtNS/h5d+r\n2YfVHtdhAFUHFvK699578cgjj6Cvr29FAKIDER7n0t+ybE4mNawAnsvGYwu4tbafx58VmEodJMsj\nkUhg9+7d+Mu//Eu8+uqrePHFFwOTNHzftQSzGqdokOYjfNhfattiBarabsn9rGDPWtaqg0bdppw9\nozGVz25qXbLqI3ohfcLHWXjBsr8aXHO9v2zRbSnv2l5y/1qTBNKePAstNp0DDcYc8uIA3iImdGDE\nZdc+nn/3+f1KjuO6NTY2YseOHWVnne9U2I9o4trCUVZ9tFi6zkGhNQaE8MlkMg47Sj9apKj2lzJu\n9LjT45wJP9EDHi+8AoFjH0svZUJOJrElLhIcofXwqyi67OX01xqjpVLJtU86nXaEWCKRcA8b8JF6\nPv3ja4fFMZXUDwDGx8fx9ttv46OPPgrc98vqN11+K47Yu3cvvvOd76Crq8vhQx5nkUgE+/btw+bN\nmwO+RGMUsWexWAyff/45MpkMMpmMG0fZbBbDw8MOF0o5ZAKSyxaLxQLL2sPin7A6rkbHVoNVNNbw\nTVKyrPuePZVUhI/nSuighWcZ+V7lOkADr+npaYyPj+PGjRtuMMtaQtnYVdbxpdNptLa2uie/SBnT\n6fSKJV76s+x6z0HZ0tKS24Syrq4uUH8G3rp+YtQtAqPagSUDUcuZagNX7lo6ENFiGcYwYOV7cZlq\na2vR0dGBmZkZb7l8MzTZbBZtbW0YHBx0m3Zdv34dU1NT7jGV2qGy8SiVSpiZmcEHH3yAoaEhPPHE\nE85Jhy3h87VPNcUHAjUYtMQHdrRxFCZdAmlh6C3DJPeUgLtQKODSpUu4ePEiIpGI26tHyFUusx4f\nlkGUumiHLNfhMnB/yGd9L187hv13JyLlkVkK3txYB9rakVh2SQcHlnAbWmm0lq3V5AjbawGYQhiw\ncxUSXjb31ptEStAgQQ6v/S6VSo6glcfJst5xBguv+bbAApeT/Ua15HbAV9g5FqDUwbceq93d3di2\nbRva2trMR67ztbV9krYRvfAF/JHIcmbu9evXEYvF0NXVhbq6OiQSiUBWEPsViwTh+9fU1KC7uxuH\nDx/GpUuX3O+iF+XAVzVlbm4OZ8+eRTwex8DAQCBYs5YKi07Ld25DX2Bg1cUiemTM+NLAtUhbaaKO\nP/vGPGBnOPD/nPnF3y1iMAwHlMMD1RarLOV+s/RVJJVK4fr160gmk25CUNtP7UuZvJAXT1BKcC7Z\nQNyufH/uG8arPO6YBLCwlRaNDazxJe/xeBxdXV3IZDIYHx9HqbRM+Gcymco6o0JhjGrZjUrO9/UD\nv7SNkmPEBwluTKfTDjvyo58ZI+n76f7hcauJPh7z4ku5TzQG05iGSWcmjVjHeNJIt5Fu92r3452I\nZRt849Mar/p4rrdkkXPmtLQZEMzS85E9gi0s4uB26srXXlhYwLlz5zA6Omr6Qeuc9RJLHyV2np6e\ndvv8aRsSiUTQ39+Prq4ujI2NYWFhAcViEW1tbejp6QkQMjxu5ubmcO3aNdy8eRPRaNRNAkYiEaRS\nqRV4yPJPGk/oPmOxYin+zDFIuT7QfWX1nXXMl0r2sHCgxN8Be3BppdBggivJDs0CTFoRxMnNz8/j\nwoULSKfTgWvNzc0hm826687OzqKhoQENDQ0YHR11jkuuLWykDpD5GKkDlyGTyeDs2bPYuHEj7rnn\nnhWKx0BQjCy3oVxTy1oYYQZ0PmNolYWV0gIflvgARJi+WABSfmtubsaOHTtw6dIllx2i66eXwwjZ\nl06n8cADD+DP//zP0dvbi5qaGrz88st4/fXXceXKFaRSKQC3sgJEtwA4MJbJZHDp0iVcv37d6ZqQ\nPdLXX4YRtsabBnLlztP9KYExAw8GqDw7ZRlTHh/8eGUOxuX6ujysgwyUrICnHJC1ZrMskOAD/RqA\nVkPEPsgGxDLDJHXjeupxatVV15vrztfjmUStEwxs5Bx9f24fILgWWq7PYFP6mGd6LJvAbVEoFNwe\nPRJQF4tF95h22aBSO3ULZEtbaxLjqyKsnxYpK0FKqbScQVdfX+/a2hfY6iCH28uXqRKJRDAyMoJf\n/vKXGBsbQ2NjI7Zs2YKenh5ks1kUi0U0NjY6O2EFU1wGsdOxWMxl98m9daaQ3F/eq21fI5EIrl27\nhv/93/9FLpdDa2srksmkN7um3MsKRixwzN91Vo8mfSyyh/GS1gnGH6VSkLiywGpYYMXX5/61/Ibl\nR77sYMWyW4y9+D+fjS2VSpiYmMDMzAy2b9+ODRs2rJh4kOO0/vPkSE1NjSNJmQxnH6v7U67PPlH0\nRuuJxtBW8MPtb02ghMnk5CROnTrl7Es2m11lb4QL66W2GRwI6jHBeFJnyEqWsmDBTCazArOIf5GJ\nwXQ6HVhCrCd32cbKPTV20nYhErk1ySV9Jp95nDKZzIGfFQDq9uB38cHiK3U7cfvejX5RYzO2LcBK\nv659Dh+vJyGkbeUJ0IKbrfFbLh617JjlP33tW27cVdIva21LLWJCT1D87ne/w+nTp/FXf/VX+NrX\nvuZ8ofZ1U1NTePPNN3H27FkAwKFDh/Dcc8+544Fb+rlp0yZ85zvfwf/93//ht7/9LUqlZVJu8+bl\nJ4WeP3/e7VflK3OlYsWi8ruuu893Wv1uYQTffSvBN+ua2WM5Set3/X/YdX0AoZyiS8Pk83ksLCy4\n9apynswCy5pbMQzRaNQZfhYBmj6yR3e8lF3W2y0uLoY6o0okzDDcibBB0591EKAVls+3SCIueyXl\n8BnysP+EbOjs7MTExAQmJydXHAvcCqTZwbe0tODIkSM4cuQI9uzZg9bWVkQiETz22GOoqanB22+/\njYsXLyKdTrs20ANarpnNZnHhwgW8/fbbuOeee9Db27sqtrfaYo0VHVjIb9xWlp7JOTLzISAkl8vh\n6NGjeO+99zA/P79iuaIFkn1todtVg3FLpyxjGWYcLWOs7Qyfo22PpXu+QOB2hLPO5FH2zc3N6Ozs\ndI+hF8AnZWLQosemFl13BoVs41g4kOBr8D0tHdOb4rPuyFgUnQIQIIQkwycSibixJU/5kuWvPEsp\nG2UKQLOWkfBYZLvGafd3q/hsvw4qCoUCenp60N/fjz179qC/vx8NDQ2BseHTYzlf7meBW9axVCqF\nS5cu4cSJExgdHXUBSk9PDwYHB3HgwAFs2LDBPY2jWCyuWOokZZJ3TUICNinHer5WJEEksrys9JNP\nPsGOHTtceTljR++tYb3kWA3e+D5WPTQpbs38cxDH19M6wUE9jwErsA8LVPiz1plywZU+X7e11afV\nFguvWuUKK698TiaT6OzsxPz8PMbHx90TB7U+WP5J+lQIHnnp7C0hOaW9mcjV5ZFAS+wiXw+49VCI\nUin41EMmU+U6rBdhOiGytLSEubk5l8V+8+ZN77La2xFrvPi+6+OBYKavXpIuWFD2HdT2iCcIhfQR\n4kf8kRwj/lvO137IRwTX1NS4/pJJQlltIOfJZAYTPTyOuS34ZdkntifSPmyvLDtVDamWj41EImhr\na8OGDRvQ0tKCRCIBwLZB+je+hpA7PBlr6bomy7V908Ljhuvuw8PyXs72tba24oEHHsDU1BROnToV\nIGktPdD3qIawfWJdFLsm+pVIJFAsFpFOp3H58mX8/ve/d/t7WW02OzuLY8eOYXR0FJFIxBGrcmxb\nWxvq6+sD+H5iYsL1aSwWc5uQc4atXhrLRDqPIyuGELEwU5idZNH+0BqTmlNg/xqPx93+pWH9uG6Z\nPZbj9H33OdJywb7vmpX8rq8jT02anp5GLpdzx7Lyslgdy9fWSmIpgMXo83V85V4P0cZdGzN2ANxH\nFiD0GUGfsfP1tS8w0feSQd3Y2Bgg9Kzr6WVc27Ztw1/8xV9g//79zrHn83kcPHgQAwMDjvgbHh5e\nMXsmjpYB99mzZ/HSSy/h+9//Pvr7+9fF+FYqrIdhhIdVNgneE4kEYrHlpzzEYjFMT0/jv/7rv/Db\n3/4WjY2NLpVdjCk7Vy6DfObfrLKw6P7X44lBcZjjtJwwB/3ayWv94fLxezVEQCWniXd3d2NoaAjN\nzc3OeQHBjSk1iOG66XozUWO1oSYWdFvqczQJINcRIkaAqoBYeSx7JBJBfX2927uJg5S6uroAOSRk\nTy6Xc4A3Ho8jmUwimUy6vRMaGxsDhJI1ZqXMGrTd7aIDYg0K5LVlyxZ861vfws6dO7FlyxY0NjZ6\nbat81mA4Erm1cSe3J/uJ2dlZvPHGGzh//jwA4Ny5c7h48SKKxSKGhobw93//9+jv73f9XSqV0Nzc\n7LJ1RLTv0eXRvofbQN4rAcq3095zc3OYnZ3FxMSEOVZ8QT0HVjxORP/0+OLPvkBNg1UmfKz9Opjg\n0RmmQNB+WJiEReM7rXuiL74gy+f3rTZfa7Hqov8vh70KhQKampqwa9cujI2N4ebNm4E+Y1Kel6ty\noCP2q6GhAclk0k2UyHUk8BbShzNThNxmXRFdk71ZJDtBdIP1QYgO3uyf7aMPq2rfoKWvrw9DQ0M4\nfvy4y4j+skWPBZ3hzS8hvNgnMDHGGeFC/HBbynXkXB/Zw/0lfcaPhuZlZXKstURexyvs5+R+2hbp\nbEFeYaD7uNo2NayPKhUpT19fH7Zv346enh6HAZjctogejZNkXOiyML73+VnWEcsvcb+EEUO6bmFt\n3t3djWeeeQbT09N45ZVXAhMmGlevleilxNJObP9kOwyJG/L5PH75y1/il7/8ZWCPK43zpZ2i0Siu\nXLmC119/Hfl8HjU1Ndi9ezd6e3sDyyR92MDymYIt+Qm/2l9zn+t4xIpLrHjKh2NYP/ieNTU1K/ZB\n5fPq6urQ0tJSFqeu+549t3uufk8kEujr60N3d7dbL+kL+sOuLQ55cnISExMToWQCi2Xw5JXJZHDs\n2DGkUikcOnQI7e3t3sDVN/DCBrSvbj5wcqfC12UgqI2cpeh8DWkzS7ktUijs5ZsVtIxvoVBAX18f\nvvnNb2JpaQmnT582QSaTOclkEt/4xjdw+PBh9Pf3Bxy+gJ/GxkY8+eSTiMViePHFFzE7OxtoewHt\nDIrEAFmkiTYia+VILWdjBXZWm3MwIiBDxt+NGzfw7//+72hpaXH1zmazuHz5ciATQ9qZSQg9JrRR\ntQIeq16aiOX2lntpx8szltpBaAOqf+e20ut9C4UC4vE47rvvPvfo3TuV2dlZ5HI5pFIppFKpwFMz\ncrmcW9LFywgt0kmPTevFgFG3G/8uAYhlK5lkk3P4N9kQn/cH4CescUAzPT2Njz/+GJ999hmuXr2K\nXC6HhoYGd4zs0ROPx9HQ0IBisYgPPvgAhUIBDz30ELZs2eKAPJN0LBIIcXBabXuq278a5+p+ZR2U\nlwRuQuIlk0k0Nzeb5JYlum8toAIsBz8ff/wx3njjDUxMTABAACgVi0WMj4/jxRdfxNtvvx245wMP\nPIB7770XHR0daGxsXPH0TV2Wffv24e/+7u/w5ptv4uOPPw7tK8s+3IlIu5VKJbz33nv4h3/4B3zn\nO9/Bww8/HABqekkMBxpyjNW2Fvax6qdJJFnmoR8gIHtZAXB2W5aCyBjizX45i6scYcb+gf2pXEOT\nEb4sH00Ysu3RBNdaim5r/bmurg59fX0oFAoYHR115ArL/Pw8zp49i4WFhUC5JXON/Qdwa8JEfksk\nEkgmk6irq3Pktowjvl4kspzdqIMTbjOLpAmzIyKMW/g6vkDHZ9fa2tpw7733orOz08wOvVNh3bNe\ncgxPuorP5o2IBePlcrlAEKgfxMGibau0m24rHgPSx0yyWufymObNkyUwtcaSBKwy/kulUmACqFxg\nqYlizrSUAF6uVW0Js9+V+OEwIorHW5jtYXKax4+eXODr6phDx0Pc5my7ZGzpGMZXN/7M9YtGo2ho\naEA8Hkcul0NHRwd27tyJ0dFRTE1NVRRvVkuEQGZ7JPXT2WNMXMrYYPKb29CKi+SaAHDz5k2kUqnA\nQzuYHJL7CuZMJpOO9BYyXcYVj/lqxGHlSLqwc8KSP8rZXZYv7WlcqznfCiLi8TgGBwfR1dXlZnT5\n+Eqv29DQgK6uLsTjcZfype9nBZss+v9cLofPPvsMk5OT2LFjBzo6Om5bYSynKt99dap2cKKJAH0v\nH3Dhc9kRWyy2r5+1UvsCbet8OUaWLmzfvt2lNlr9zEFRc3MzvvWtb+HZZ5/F7Oys2xuFN8RtamrC\nQw89hMXFRfz85z/H9PT0ijpbg7UcuVUOWN+uaIPP77o9LHJEwBL3sRjaWCyGmzdv4uWXXwYQBCyx\nWAwNDQ0OcDDZpWedrDqvph00geVzsny8Bok6Bd6nU+zYJagRB1coLD/horW1FQcPHsTevXsrKn85\nmZubW0H2CPgQ3ZTgkftPgzyLFAuzexows+7oNuT76KwC1gsB1hJYlkolR7bIeTxjOjo6it///vc4\nffo0bt68iaamJrS0tDhwICRsbW2tI4E+++wznD59Gh0dHRgYGFgxW8T1E52Ue3M7VFuqfU3reqyb\nesZfUprr6+udDlkEuGV/xUZpYMt29MSJE3j33XcDAYPoXDQaxcLCAl599dUVevn8888jk8lg27Zt\n2LRpk/PNugzyfc+ePdiyZQvm5ubcE0jC2qja9lTq9Mknn+CLL77Atm3b8NBDDzk9YrJH2wtOJec2\n5IkADqgs/6iDeZ2WzjOsxWIR2WzWEap1dXWBVHuZwZexbAVO3IZhwJP1TsC2XhpktYsPv2igvxZj\nUouFQ+RzbW0t+vr6sLS0hPHx8cD+D3LswsICFhYW3O9MpkkQzu3L5KGQ1ryMS8aSHC+vUqm0YjZd\nL9tj/bAytKx25/OA8AyvMFwKAE1NTYHlE2sh+t4WlhT/xX7cIntqa2sd4SPv3K7cLkzayHjV2TLS\nT0zA89IuDlCtDCpN9iQSCZfFatkW0R+NJ/UY04QS65EmrPlYkbUgfCypNKaR+vI5VtwRRjIDNnnI\nS7n0ffQ1rPvpMWKNP9840vZX2+FoNOoI/sXFRTQ3N2NoaAjpdNrco1S3V6XtW4nIU3J1lhnrv0wy\n+CZBdKawbgsRHsOpVAozMzNuTPH1WHcFYwrRIy8mUtl/WvbOakctvnPKXcvq23IxYzn7C6zzBs23\nK1al5Hf9vz7Hdy3+rFleCfYikYhLcQcQePqB77rF4nJa1VNPPYVsNovOzk4vm2wFjZrICDOmlXby\nnYoP8GvDxgatkmtZYgWYYdcJO1+O0wbEIoj4WMshWy/9aEpdfw5wfbNpPj216lINYSPCht7SJats\n7OQENPFnBhd6dovbga9vOT0N7i0JCwr4s3bgPidbiT5YQYroQH9/Px588EE0NjZiaWkJ27Ztw7Zt\n27Bhwwa0tbWV75wKZGFhAblczu3XI7ZKZurr6uoQjUYdeJR684yUFazpvrZABR/HfSwkKJM64lDZ\nUcnnfD7vNuisr68HABcAaSdbKBSwsLCA48eP49NPP8WZM2ewsLCApqYmNDU1oa6uLjBOxbHL7JKM\n0U8//RTxeBz79+9He3t7gOzRYE42UC+VlicD9FMSvyyxbEc5Yf0UmyWb1be2tgZAkWXTrWCJP2uw\nLOXk8S7kg/wu5IZli0+ePImpqSkkk0ns378ff/zHf4yNGzeirq4ucE9dJh9G8LVdNcQiILLZLBYW\nFpzucyDFfaGfyiX10jOhLGF11AG9kJWiy5K5I8SZjFtpP8lgEL3n7MZK2lD7DCkP2xy2ndaTjSxs\nZtVxPaQclsxkMjh37hyKxaJJ9IhweUdGRpDNZjEwMIDm5uYVY0Q/mVNvysuEKV+f9UxnirGOSqDD\nezmFBZx8bQCB8/SkiuiRL2NoamoKly5dwoYNG9DV1VVZJ6yh6D7VdpL36hGyR9efJ7+k36QveYNj\nuTb7R74fZ4zzJJhueybt5SUi/ctllN8Zf+m+lTpwvfTyUw6W2T5xW37VxLLd0h4+YsWn27crFv69\nnbKHXVvGpSapWD/4vRoik3VaF4HwzbGt+FfbIwuPWONWZ5Ay9tDEqSZ79ESJJl98EoY9wvpK65jY\nEStJQO6j7xcWR4rcFcu4fMqrK8KKGwbsKi0TX5OVUJRTAigeKAw0LUcJLAc6O3fudL+HETk6mAwL\nPn1tuh6G11cuC2zrQWkJG55yA6mSOoUZZqsPrDa06hXGOus+0+XR5dJ18jH/1nWqIZroserNv+nP\n+hpAMDNG1zfsVa5Py+mG1V6VBCNarNke/V2TPJL51draGghe7rnnHjzxxBNob2/H4uIi9uzZg23b\ntjmnUw2Zn5/H4uIiMpmMc2oyo3Hjxg20traitbUVwK3sHV03aX/ddywWuPEFA/KffM5ms6ipqUEy\nmQzck0EyAJddACwv/ZmdnXVPzpJ7FwoFzMzM4JNPPsHx48cxNzeHRCKBlpYWR8SkUimX1aQD10Kh\ngEwm4/aNaWlpcXvEsNTU1LiNHGWDZ2D9Zi7LiQ8shI0j6SvRz3g8jra2NgwMDGBwcNAtdQvzo757\nsg+2QJkVgEqf+mxxsVjE2NgYrl275gKtb3zjG+jv7w9kv2jgU85fattTLdHBQalUwsWLF/Hpp59i\n586daGtrW0H08D4Y+mVlTfh8f1id+Xytvxp4M6bRwNLyE1ymsHJweXgsWz5V608lvtTK0rwT8bW3\nTxYXF3Hjxg3E43E0NTW5bEt9PZbp6Wlks1k0NjYCgCMAhLDWWEL0Qewhbz4vS1X1RApvqitjRcrD\ne1ToWXcRyx7oQMkifET0+BRZWlpySyzWQnz9Zv1u4RsdNEp2j95vR4JA9p9AkGzhJwTK2CoWiy7r\nViYh5HcmcTh7hK8vGXLadjKZLMGq3pjdWibKwX6pFFxeo8kf6yXnhrX9ly2VBNfc1mFEz534DtY1\nq80qiZV0GayY2XqaZjKZdA+Vkf0QecJGyIVq+kXZh9G39NGaJCoXV1o4VK4j44bHqc7ukbZiMlZs\noSznkt9kotDyh754ZLXxDZ/ja3+tK9puNTU1obe3122bEYbjgLsos4crqztWfw4DG1r0+T6FEqOr\n729lA1gDTXeEJhosBWYw7gPA5eq33lJJ2TgYCDMi2thqB2cRJBxQWNfgAEP3jcUgW8YkzNCXcwAW\nINX6bJF8FglVTQPM5dNiAXVL33QbMVnAZS7XVtUOunziswdWZoBFvurAhGfU9u7diyNHjgQIoIGB\nATzyyCMusyeRSGBubg6Li4soFApVye6Zn593T50SnRkZGcHCwgIymQzGx8fx8MMPo7W1dQXZY+m7\nDuiA4CaR2o7xOeI4xVE2NDQgk8ng5s2biEaj6OrqCix75OvK+nJ5Ssbk5CTef/99XLlyJfBUgXw+\nj3Q6jVOnTmFychItLS1oa2tDW1ubW4YiZI881UmCGXmiIrA8s3zixAnk83k0NTUFQFGpVEJ7ezue\neuopbNmyJRCwFAoFpNNpd2xLS8sd96El1RwP2n/Jq7e3F4cPH8YjjzyCtrY2RCKRQJ9KH4mOWLNM\nGhRb9/XZY53lZZE90vbFYjFAFAlglHvfLcL1KRQKeOWVVzA2Noa//uu/dvuTSF2E9JEsDr1cQvpJ\ngkltp+TdR0Jrm1VTU+Oe0BeNRt2yTwDuSU+8f4y+Fu9BogNFC7doApX1onPRdQAAIABJREFUCLil\ni3qfEu77sDaWl54tXgthuxgW1La2tmL79u24ceOGeySwdS15z+fzGB4exsjIiBuTQ0NDXp9bKpXc\nkq8rV65gfHwcpVIJfX192LFjh9M9Xr7H+EvuK+2mH7nOY1rjJD6fx2GlSxtYOjo6sG/fPrfZfrXF\nmkyxsJSFVVl/deaMflkZEdrPimibyjoPwNkCeUqbLqMuJ7e79Lne74RfQvqJ8Jjhcomf00u4ypE9\nUqa7ZULEktXgTt2nlQT55cTC1vo/nx/V45jroOtSLBbdBKCMhVKphJ6eHrS0tCASWX5gwsWLF90T\nrLj//1977/Xj5pXej3/I4bANp/fm0YyK1TyWbUlr++uy9jbYwQZYIAGym4vkIrlMkL8lSK6SvUkQ\nIAGCbfZerLErR9HCZVe2JatYktU1miJNZ+eQfH8X83uOHj58zst3RpyRI5wPQJB8y+nnKZ/TmqlT\nuc7XljJKee/nf9hseKmvuD6lPsX1DJUdJ8TlpsxyZjkvcznwaWsj2kf6tFp5SzvIb4IIyZB9+/bh\n9ddfx759++pktoYnTvZoTrb8zcE7QBAShD8rGxsvvGQyidHRUUMS0Ga7fX19qFQqWFlZ8WV+ZX5k\nGjXDiKZVd3d3o6urq07R2ggVvw4grzUDQTqjppyoDPg3LyP67VemfoaWdCrkNFTpqGofm8D1C186\nLUEUiFZufgaW9v7jQivTIHXJ35f1w51DGYdWl5qQ204+/dqFFr4tr37kDo22j4yMoLOz0xhQbW1t\nePHFF3H8+HEjL8rlMrq6utDX12fW+ZdKJbOhcrlcxuTk5JbyqCGTyaC1tRXd3d3I5XJYX183o5GX\nLl1CMpnECy+8UNcPtHKvVh/tXyDL1lZnvOxoJJqWrty6dQtLS0t48OABEokEbt26hampKYyNjdUs\n56C00Ugkhffw4UPcuXPHlCnJzWKxiNXVVVSrm0tk29vbzZ48nudhY2PDLGvjhgaNckajUaTTaaTT\naczPz5sTFgcHBzE6OgrP89DZ2WmMY2mk02h0s+SpRDONLdnnuHyJRqNmGQUd28t1Gu/LZMhTGNLQ\np7g0YlDKYv4JhR7N7OGGLJeH9Mzy8jJOnz6NcrmMl156ybRnLjO2YsA3G5o8nZmZQTwex9LSkiE5\nZRmQQUoGKpUTyR9pRHJnivqsNligyS6aFcIJAaB2SQ7JL24ca8v7CNLY5vpUlg93gOm3TLMWh1bG\nvL3J0eJmwKZ7uX0g9RXZb6HQ5hK+lZWVmn16ZB48zzMzgKrVqtEnWthUXjQzY3FxEaurqwA2Z0W2\nt7ebcOjoXXJstH7CZ4DI/Sj8dL90kjTCp1E/o6USa2trWFpaQi6Xa1gfjwOtzdp+c/tLI1AlmUr5\n5jMWuc4iyD6s9TFa0kmyQupaWzob2bO2dNBH+hJbIXokubVTxCtPp80X3Gp4/CNtH8oLJ9Tkf60c\nNT1gK2+bv6jlkYdvi5+nr1wuY3l5GTMzM7hz5w48z0MikTBL5SORCHK5HJaWlrC+vl7nzzYLmq4L\nQmbKb82vAB7NPtTIHupLpF852QOgjuyh7Qb4clnSmTw90qbidaT5iXRPk5ES0hfU/ETZv6vVqtn0\nPpVK1ZBCNuwa2ePX8G0N2eYYSseUf8v45Psa4VIul81oMTWWCxcuoFAo1GxypY1Qa86+1jA1Zq5S\nqaCzsxMHDx5EMpk0DdiP8LGVgyyDnXJMbM6z7VkpEAk2oRWkc0hII9CmlPwUpCRatHTxEditkj5a\n55WdXMtXs50UP4GzlbqVjogWbqP+HCRdjZ73Mwg0Q06Wuc3AI0MsmUzi2LFjOHr0KMrlMvr7+7Fv\n3z6kUinE43FznDzfgyKdTpuTb+gI1mZNX89kMmY0eG5uDmtra6ZNzs/PY2FhAeFw2CyPIiOU2ivP\nN5WRLNdGbZmeI5JpaWkJd+7cwZkzZzAzM2OOUC+VSvjbv/1bHDt2DMVi0SyzAh7t0UOOjHQeeBrD\n4bDZoDSZTJr11QDM5vyZTAaFQqGm/9MMIX7CFt/E8PDhw3jnnXdMGUSj0RoDiGOn5Onj9HGbTOXg\nbZz3AVs6JHkL1BrB3LDXDGMKS8pNTi7YdBy1Y5Irt27dwj//8z9jcXER+/btQ3t7O2KxmDp6zuUB\n/62RUc2GjNvzNvfKyefzSCQSNWVAzjs5eWSkEjnD9YvN8dN0mCR8yuWyeZ/Ci0QihligMGmGm00G\naoMQRLJKJ1jqEFnmmrz103+28g0in7aCZthNAwMD6OrqwsWLFw3Zozl28trKygrW19frwqN+WK1W\na/oNEdLpdBqXL182y34OHjyIgYEB03+ojni8fGYPyV7Ks01PU3lz8lzb8ydI/wqFQpibm8OVK1d2\nbDnXdqDZAn72IlDrNHK5ReCyp1qtGj1FJ2jRkhO5KTLvC9rgX1DYSHm6zsMm2Stne2kfPquukY34\npKHpBfqvDVwQuK1v8ys0+5Pq2k8vaz6vfFbzm/x8HMpHOBzGw4cP8Zvf/Ab379+vq5e2tjbs3bsX\n8Xgc6XTalxh4HPDy05Z9+rUXqefoGtet0k8mXSWJHjlgwdPGiR2+pIvvjcblgiYX6VvzZYFHtncQ\nH8hGgPl9otGoOaBE09MSu0L22MgQm6GmFch2BQqFpRUibyhytGNkZMQ4e62trdi7d6+p9IWFBayu\nrpoGwRuGzdjWCCZqzBRvqVTyJSR4PjTyx+/340Ajm/waLa9P+vbrLJoQC2qcc2eDj5LykT9ez6lU\nCuPj48YBpym0JKgpzI2NDXz22Wdob2/H1NQUksmkEQQbGxtIJBIolUr49NNP8cknnyCXy9UoD6m0\n6TMxMYHvfOc7GB0drSN8NIO52ZB1F8TQ1oSZZOqDhKW1DR6P9jz1Oc158GuDkuDRnBfpeHR3d+PA\ngQNm35auri4MDg7ixRdfxPj4ODY2NtDW1oa+vj7jKBGBQeHzo1v5Bo98M8XHwejoKDo7O2uWepBC\npRmIn376KUqlEp599tmae9zQ42Wg9UVbuWpG0Y0bN3D+/Hmzn04ymTSblp4/f96QMV1dXXjttdcw\nODgIALh9+zZu3ryJQ4cO1RyxzeuM0kMzgIjo4Qo9n88bsoePtPIjQOn0kvn5eXOkcTQaNaMiAAyB\nRHKYyoTyvpPYjvPqJ4c1Y5T+b2xsIJ1O18z60NLByQJu/BKkwSkNaT/DlJerplfC4c3T2nK5HDKZ\nTI2xxvN2//59nD17Frdv3w5kWO0UScDDX11dxXvvvYdCoYAf/OAH6OrqMjJCG/XkR83yNq8ZyLzc\nuNzZ2Niom6lBSzM40aulnRM3pBNp83dJyFFbAFAjO/kIqmxvBD4iK8k+DX42QjNn9vCRX2kwe56H\njo4ODA4O1hAohI6OjjqH3S8fUn56nldHfNBzvLw2NjZw8eJFzM7OAgDGx8fx3HPP1ZQFLVslYkiG\nydsbzbCh93p7exGJRGp0PDlF5FjIJWByxp5f36M95ebm5sxSwmbC5jRzOS5JD+lwcznUyA6X5Ul+\ngGbzyv4B1C91IXnHyXgeX0tLS92+IqTHkskk2trazEAIP8GNv8OXp0gHuFrd3Gic6tZWj9LhJBm0\nE2iWztUIKwKRpXyQgT9nm+mk6S9OnlA4/De3YbWw/PJP73DyhH6TPv/yyy/xySef4OzZs1haWqrz\nRwirq6t1ZdtM+0b2LT+CjP5TGvhvrvt4W5R9mudBKzM5OMRJb+q/tiWqQYge2bYoPi08nlbZhjRZ\nI/Wk1JmSePXDjpM90lHzKzQ/ASN/c6PBZnRpDYEqnxtKLS21G6a1tLRgbGzMVFIsFsPBgweNkNjY\n2DDLujjJ40f2cINKjmrZpotKQ0vmRZaHvL/TDgqHRvDIb1m/srNI4arVoRavzbmg57lB3NnZib17\n9+LOnTtYXFw0m4nxTkpkz+9//3sUi0X85Cc/MbO+aNO+eDyOhw8f4ne/+x3+93//F5lMRh3xkJ14\n//79+OEPf2iMANm5tXebjUZtipctfWtED/3X2qOfAgtK2PCP3zuyjWijx5Lpl2VeqVTQ29uLb3/7\n2+jp6cHGxgb279+PvXv3mmPjaR1wuVw2ZE42mzVOEW9n2sltzcCBAwfMb0p/KBQyymptbQ2nTp1C\nJBLB4cOHa/oTN3JIXvH6ClKn/Dc5ktevX8fnn3+O8fFxjI+PI5FImKVVX375JX7zm99gZWUFhw4d\nwr59+zAyMgIA+Prrr/HrX/8ayWQSBw4cqHGWuJIlsodkMRmtVBeFQgG5XA7FYtHklWbzEKkTj8dR\nKBQwMzODZDKJZ555pqadhMOPZkMR2SONkichTx8nXimDqZ3n83msrq6io6MDbW1tquFF8gl4pBOp\nDVG5SGOOjzDK6/IZAh+94x8pY6TcoXZ/69Yt/OIXv8C1a9fq2rDNnmgWeP/gZbi0tIT/+q//wsrK\nCk6ePImenh7Ttonw4afryHvS6JN9leLk7ZcGLegeET22pToyHxSWPNGE21eUHmoDcgBLI3tkPNKA\n1WCzHagN8D2cmgGbXUHo6urCoUOHzIlwfn2Ty88g0Ih1nm9yFIvFIj7++GN8+umnAIDXX3/dpIn3\nDyK4NbKHl9/y8jIuXboEz9vc6DOZTJq9Paif08yfUChUM/KtOcGyPLQ+ceHChZqTbpvZL2V7oHbG\nZZLm9GsySdps0l7n5ck3xObhcNnD2ztd404mla0ke3gZ0oAFET38JCEifOhD1zTihxMQkuzxO6lN\nlo1chtksNEPv8fdlfWmELScFNLJHazO2OIKkRYYXpC/w9MtB7XQ6jfv37+P06dM4ffo0Ll++jEwm\nU6P3+TcPcyd0o9TVXH7b0uEHqQ+0/up5Xt3MOBvZw3WIjdCTvIUtXTxvFI5G9mj9SfZzjejRSB6b\nf9kIu0L2+BE5fvc4KFP0TC6Xw7Vr17C6uorR0VHVQfEjQojRJcceqDVqNXaSnJvh4WG0tLRgZmYG\n6XS67lkZl6xImnI2MTGBvr4+40hyZ9KPyNLC302CRyO25H96jr6lAJDXpSDTiB5Zp/RMpVIxR1Lz\njswdez762dbWhlAoZIxaajuc7KlWq1hYWMAXX3yBRCKBl156Cc8995w5Aejs2bP49NNP8Yc//AGz\ns7NmhgWHZGd5eqQRwPOp5XenoRE7sv3b+rJNidj6O4dG5vDrfpD9W5a1tj5e7tB/7NgxHDp0CAMD\nA9izZw8OHz6MWCyGUqmEVCoFz9tcklGtVmv6aKFQMKej2JZC8I0Tm2UQ/dVf/ZUZef/jH/+IcDiM\ntbU15PN5szzp3r17WF5eNkte6IhlvikxGZW8H/H+SGUo65HXEzkE7e3t6OzsREdHB1KpVM2pIFRG\nnuchn8/j1KlTyOfzeOGFF3Do0CG0tbVhcnISiUQCr732Gg4dOlSj3C5duoSrV6+aWSg0+szrgDaj\npiPpM5kMcrkc4vF4zUZ8ZABTWLFYDB0dHcYBISOC1ri3tLSgUCigVCqZGZ7Nguxv2w1DI2o4uOO2\ntLSE3/3ud8YQnJiYMEs/ONkllwTxfs7rRhsVa2QgaWikxzQDDoAhN69cuYLl5WXVYAtqX2wXfnJa\n6j05oMHLjk/F5rN7OPkp4+W6jZcL8OikJ3ImGqWVy0yaoUjyQpI91Af5e372is2o5W2J7jciemwO\n+uPg6NGjNXHy357nIZVKIRyuny3DnTdK5/j4OKLRKO7du4d0Ol2Xn0aOobR5qN7j8Ti+/e1v49Ch\nQwCAkZERM9uxXC7jzp07yOVyGB0dNUsHeZiFQgF3795FLpcz/YbsHt4/+Kg3AGPTyGUO2izsxcVF\n3Lt3T9V16+vrpvx2oh8GhbRlJGz2mrTVqKxo9hP91+xZHq7m8PNBRE6GccgThEjHxmKxGsKHdByR\nO/SbZsRyncHbl7S9/6/D8zzMzs7i6tWrWFhYwNTUVM0sR23gi+weTdbIuuXxBLVVtbrn4Wnyh/+X\nPhIdXvH+++/jgw8+wK1btzA7O6tu+C3TsNOw+d8Sml0pw5H3JXFEckwj86TtqpXjduwDG9FDs/yA\n2r2LtLaj5VPO3NHqTNpGnJzyw66TPVQx/B791iANTHKUi8Ui7t27h1KphL6+PmOoau9LSNZeGi6y\nAdB/Unbd3d2IxWJm6UCpVKopdB4vFwbkYEUiEXR0dGB0dBRdXV1mSZHcwM1GBMiy8TOOmwHNmQ8K\nG3mgEQlBiB6ZNyJNqB5o9IKPkNIpCrQMZ3x8HLOzs3jw4AEqlYqpP4qLhP7q6qqZMUAb4w4NDSGR\nSOAPf/gDPvjgA9y5cwfZbNaMsPE085lboVAIiUTCGPWUbk0o+tX546CRQLDVkzS4/QSjn4Nlq1f6\nza/LMDk0YcjlhEbuhEIhpFIphEKhmpkcJ0+exBtvvIHJyUn09fWZ48L5EixamsVn6nCyhxMnvN9q\nRuLj4q233jIzigqFAm7dumX+02wX2iSUr0mWDhsf/dX6o3SwtfogQzeZTKK9vd1snExtPBqNmr16\nqBzPnTuHjo4OHDx4EGNjY5iYmDBkGCd6COl0Gjdu3DCzeSTZk8vlMDg4iD179pjjkK9du4ZKZXNj\nZzKE6Zh3IntaWjaXeSWTyZq9hAAYMoimSIdCITOK2izYyjTIswTuHMtnpSESiUSQyWRw7tw5AJvL\nT9bX1zE+Po62tjakUimUy2WzNxKBGyhSF1O43NihOuF6tZEhZ5Pv9HtjYwOZTMaMXHNyPp1OY2Zm\nxrQ3rkO2a8gFhZ8RRpCyUxI+fJYPv6bJDq2cNEKWrmuOCVA7I5Df4yQ1J7FlHkl+UFic8NHKQ37b\n8hak7BoZzdvBxMSEiY/bcDbHmP5TGfK09Pf3I5lMIp/P19kWHFrf1Z6hTzQaxfT0dJ19SWX/4MGD\nmmPdJbGQTqcxOzuL1dVVQwCSnaPJCprxzMkAObOHDwwUi0U8fPgQN2/erJmdIrHTfdFmB2sOIC9L\nbjtI+4E+/IQfWf/SNuKDjhQHbz+c7CXdQvFo9hPpK+0kIXo/Ho8jmUzWzPyhZ/geTZSORvJLkzeN\n7NVm4XHD9DzPbFa8urqKYrFoSDk5KMvtHRvZQ/9lf+dlYiNruK1PdaqtYuDvaXYzvUf31tfXMT8/\nj1OnTuG///u/a2YJcTKr2f5gI0h91SgNQWUC76+8n9E1Inp4OWiDllKPaPHb/E0tTVxu8me5DGjk\n2/Jy08pL2uM2H8kvrbu2Zw93IqTzIw1UglS0pJh4x9zqiLkU7Pw6hS8rhToPd5QikYgZkb5+/Try\n+XxNg9IaOqV1YmICY2NjSCaTNYrENrPHJoR3E7bGRb9tadIIBM0Yl9d4vLaO16gcqMxptITqjEaz\nC4UCFhcXzf4iss7pdIvZ2Vl8+OGHuHDhAp555hmMjo7ik08+wb1798x+T3zHeV7X1M7a29vR09OD\n/v5+o1D80q0pkN2ARtZopA1Pq3zfL9ztpIVD9iuNWNEMtP7+frz66qvmdL3BwUGMjY1hcnISw8PD\nZr+WXC5nyAnaZFmSPbSMi/bi4csXpFwKojC2Atr8k9LLNz4mhRKPxxEKhcyJJ/zkJaB2mQ039qnM\ntXLW2gIpOTI2+VRxz/PMhsrJZNIcsU79g8940ghuroyj0aghbCnNlUrFbH595MgRPP/88ygWi7hx\n4wYikQjm5uYMQVCtVmtGQ4ns4cYAGQS839F7nGD6vwRupHJD5O7du/jVr36FZDKJvr4+/L//9/+w\nd+9etLS0oK+vD+Pj4wA2651GhGXbJqNKGk5LS0u4cuUKVldXffu8zQjU+snS0hIuXbqESCRiZnER\n4VosFq2EwE45llo+bLClzTYLgD580Il0hTa7h9eJnIXVSJfKwSneF7i8oDzyZWYyfpK3QcqqkfEs\nP7LM+IyJZoCT3lweNjKgbfo5Ho9j//79aG9vx40bN1AoFOre08Ig+aPZw1wuyfRQ+8jn87h+/Xqd\n00Gyi2b18NFvWb5EJpD85jqDz+TjfSybzeL69etYXFysST+Pg6dlJ8DtAIJ0jng/4XknXUR2IO8f\nnFjkfgsnajkRQPfIZ+D6iuIlnUazTbndr+lo+paba8vf8ghp3k8oDZqtxGUp2Ts8TdqsaO6bNNsX\naVZ4sm1znc+ddJkXfk/TKVxu8nf5/j98QJf7rTxsniZqY1pf4fKIdEMul8Ply5fxP//zP7hx40bd\nLE6b3Jc6ZCf6o/TT/dqJJv+09GvcgNQRZJPw9+V7Nj+T15fUiVqa6D+vS8lr8Hty4EkLS7PD5T2Z\nLy6fGu2dtSsze/j3VjqyNGZ4RbS0tKCjo8Osh+cVY/sQpLFC16Sip8bHnQzP84yA7ezsNKOLi4uL\n5lQF3tB5nine9vZ29Pb2GqeSOzvaniIyrTbCY6cIIGk42KAZIrxR8sbJr9k6gExDUNKH1x3tQ0DG\nc3d3N4aHh3Hy5EkUCgX8/ve/N5seSuFLm8Dm83mk02ncuXMHMzMzGBoawv37983SGa5cpRIgQ4Jm\nhNHMEVu6tfzuFBopFpuA4WlsFK4Mk7/XSCZwIU7PaTJBji6TAdPb24uuri5sbGxgZGQEJ06cwNjY\nGABgaGgIo6OjhpzQNlXmBpAkY+m3jdjRjM5mIJvNmtHA7u5u7N27FysrK3j48CGATUKTTlu4ePEi\nRkdH0dPTU2OoNVIklGZJSmp9lDsJfHNHusedg2Qyacg1z/OMcc3lH5UV5ZFm1YXDj07XohHzbDZr\nNn7ev38/qtUq2tvbUSgU8Nlnn+GLL75AsVhEpVIxRnA8Hjej1HzwgUZU6T/FQ4pam7nwOGiGkdVI\nFpNxSM4f9adMJoPV1VVUKhWzZI2WohLZMzk5afY1onrkhADXi9yxyefzWFxcNJuwyjTa5Lj2W5OB\nlK/l5WWcO3cOly9fthrQmlHXTNj0sQRPCzf+JMFjI3xsxi59SyeWt2EpY7ls4vVnk6n8XU4oSTkc\npH/wtGh2giwrWW6SHGsW+MllPE6eNu6UcPC80P1IJIKenh4AMLO/gc2R+CBHjmvtXbsn71erVSwv\nL1vD0PoDybbV1VWjO2hWjxwgILnJSYx0Oo2FhQUsLCyYI+V5Wfj9biY4kcEh2xzJM430kA6q9AMo\nbC5XSY9w+0/qRN5G+ABGPB6vWe4tZ39IW5mHqe05Ikke2TZoUEgOsJRKJeTzeeTzeTNDXltpoPkn\n0sF90pD+Ji1rv3v3LiYnJ82Jm1Rn0meke+T481lcVDfchuLvU5jczpLlpOkqbWBC+0/xLSws4Msv\nv8THH3+MM2fOYHZ2to6YpHcakS071R85pJ5sZLfQs37lIa9Tn/aTj/Rf80X90uuXVt4vZXxaPfO6\n9tOBQbEV2bqrR6/bCs/mzJGzzD/0bCKRwL59+zA4OGiO+uWdSrLP0vmSHYCUl+aYknCX6fe8zfXc\nhw4dwq1bt3Dp0iWrIw88Yg/9BKfGmvP/vLwkGRS0vJsJ7oxzaMaGbeRVdjzeGRo5AbbOTe+TMqNr\n+Xwe8XgcJ06cQHd3N65cuYKbN28CgFnTzJUlTbGnTz6fx927d81JbdIQ5e1GmxYvZxFo8DOKHxcy\nXi4oefnJ335h2JwyvzTYwrOljfcrvqcI3xyZl/XGxgaeffZZvPLKK2Zj7ueeew6Dg4NIJpOmjsmw\n4SSPJHvkrB5a3iCNcN5XeV6aWYe5XM6QkFNTU0ilUmbpEvBoI8d79+7hl7/8Jd555x2MjIzUtD8t\nTbZ+pxm+WvuQ9SYdxHK5jL6+Prz99ts4ePCg6Yt0n5PtRC60tLRgZWUFd+/exfDwsNlDiYiedDqN\ntbU1s19XPB7H2NgYvve976FQKOCjjz4yS2RDoVAN2UNLfsgg48Y+H4GlzzfJoA0CXo7cyOTOMrXl\ns2fP4vz58zVG8N/8zd9gamrKEOVElvG+Jh0RKf/4N1A7eqbpZ0kg0O/u7m4cPnwYvb29Jiw6lv3L\nL79EtVq7hEumR2u7zYKWTylLeRmRU8BnBvBlXDSbwracXDpXVGaS6OG2jGZLAKgpL/6utozc8zwz\neilnHfNrfrAZuXSPvrVZDbJum0328PTZ5FwjvUig/LW3t+PQoUOmXK5cuYI7d+6oRr0tPE3Gamnh\nerKRUyedrlKphJs3byKTyeDZZ581Szl5nZJdxEkGem9+ft7MsON6UdPzO2XbcHkh42r0HvU1+s+J\nToIsWz7IR2XP/1N75WXNr0sfQM42kjYxf1dzHHnfkHnnRCzpOk7kENnTaHaPHFjbKbJnu76L1t4y\nmQzOnDmDlpYWJJNJdHd3o1qtoq2tDYlEos7O4WRgJpMxe3pSGVO49OFysK2tDbFYzFzn96V9w+V/\nEB1FYeXzeVy4cAH/+I//iGvXriGTydQsAZbvcztMDt5pZdZsSBsyKKQskff4b+67c13G73PI/hVk\nQMgmyxqRPbZ+S3nbbju36Sk/7BrZQ9AUvkysNCKSyST27NmDxcVFrKysmGdonaptRoxG+lA8Gtmj\ndTZO9sjnyABKJBLo7+/H1NQUHj58iPX1dfT396O9vR2e5yGdTuPhw4fo6upCb28v2tvbVcJHI3Rs\nhEfQsm4mGnU83oBlh7QJM+06rzcqI9pcmbcLbfmHjQSi92i6bnt7OwYGBjA4OIjOzk7k83lDCvF0\nScFQrVaNApCCWnNaaPZWV1cXjh8/jqmpKSQSCZN+zemW6W4WtPojwaMpMglb2fJ7jeL0ay/SCJVx\nUj1rU4vL5TI6OjowMTFhNjw/efIkXnnlFZRKJbS2tqKvr8/M5KEZW3yZlu3YdEnUEVnHy8xWT802\nhPj+Mm1tbWZJ6MDAgJl+39LSguXlZXzxxRd4/vnnjTFLGzjL+pBOuqxnm1PC+xM3EskpLJfLyGQy\nWF9fx/r6Ojo6OsySLjJSpMFIZMPc3BwWFhZw//59AKjZdL1QKJjZPj09PYa8I6d/cHAQw8PDGBwc\nRDgcNsdIEyFQKpUwNzdnNtenvNA3DSo0MsC+ySCdxctVjjLxKeF8ClqoAAAgAElEQVTc6ahWqzhz\n5gxCoc2Bib6+Prz88svo7e2tOa5dyr5CoYCbN2/i448/xoMHDwDU6nJ6lstrjfCRejwSiaCtrQ3V\nahW5XA7RaBSFQgH37t3D4uKi2S9NjqDtBtnDIe2L2dlZvPfee3j99dfxwgsv1Bn6krygNizriRvs\nsr3K+Lms1OSo5pxpDrjUxzIMru9sDjIPm5eLdt+mb6VRLgmqZoA72jwd8posE25H8nKi98LhMNra\n2sy9oaEhVKtVLC0tIZfLqc5MI32h2c0ctvbt1+6pX9EMJG7XUJ1qTovnbW66n8vlAjscQdO0VWhO\npc0hkjqMz1QkuUknY/HZL2QL8Blsnre5QTZfxsxJH+DRhvc833wWULVarSlzOSgaRJZx+0iWB+/3\nckCM8kZ7rdHhBrZBLttAdbOg2YN+zzYKo1gs4tq1a/A8D2traxgaGkJXVxeOHDmCPXv21PVfKWPX\n1tZw8+ZNPHjwwPievC1x/XXy5EmcOHHCEPj8viTg+cb8tC2AzXH3PA9nz57FlStXUCgUcPnyZVy+\nfBlLS0s1YfE2wt/l9r1m1+2GbdOI8NHkoM2X5NCIHv6s9EuknpE6x2YjaGmQPp/sl7zvyk3t6T2Z\nZ15WW4FWVhK7OrPH5tBp/3kHIZLkypUrWFpaMkJLGqga2aN9eFp4Z5CNn9/jDp7WoXp6etDT04Nz\n584hk8lg8v+fBg8At2/fxtLSEgYGBjA9PW1Ns0b47JYTuR3YBJP2zFYgCR0ieni90D2/Da2l4QbA\nEEd0QtHQ0BCGhoZw69atml3sNeXKw5T54h1XklTFYhGDg4N45513MDo6ing8jmw2a928cafB087L\nhvcDrY3z3371auvjW4XmAPLTYuSyqvHxcbz55pvo7u5GqVTCt771LUxPT9fsuUOzQvh/ufkyvyc3\nZaT08HLbDWVJ4ARhLBZDMpk0y26WlpaQz+cRDoexvr6O5eVlLCwsoFAoGKNAU5bSidEUEQd3BMgI\npmngZOySAbO2toalpSWsrq6ivb0d6XQa2WwWuVyu7pQyInoikQhu3LiBDz74wLwnSbp0Oo1UKoW+\nvj709vaaE9TK5TLa2trQ29uLvXv34uHDh8hms9jY2DAKN5PJYHZ2FouLi4bE58awNqV+N+u4GSAZ\nRtP6KV+cRNB0KF373//9X5w5cwblchnT09OGFOdHPXOZuLGxgWw2iy+++AI///nPzd5IQP0Sab8B\nGT+dSBuid3Z21hEAkuTRRr53qpw1GR4KhXD9+nX80z/9E9bX1/H8888bMkdLt1y6xQkfKXs5tP/S\nsJbOjCQqNAdZI5WkPtDqSiuboAa0H9FDDjInG5oFSjfPOye0uAyQo7KaQ8HD4tdHR0fR2dmJ8+fP\nm73XAPvsaFlGmhMcxCm2gdc/TwulXS7J1foYhRMEOylD/Zw6LW7eN/heU6TbaeYj2QFE5pBNwAcj\nNzY2zMAzLYOjfRxDoVDNQJLUeTytvL1L+aDVsyRzpLyUqyL47Gdu21Da6NAJ2gtPLumSewvxJU7f\nRFDZ3717F3fv3sWHH36IiYkJc6JdX19fXR/gNnA4HMby8jJOnTqFTz/9FJcvXzZ+oLSXPM/D3//9\n3+PQoUPo6OgwS865XcFlHCd78vk81tfXVVuDwvjP//xP/Pu//3sNwcAJQ6mT+fucxJLyfKf6pOY7\n+T2nXdf6swZNF8g6sqWL9zvN1+O6QYuTiFoeNtmUWn3z2eSafxWE4LSVayNdsONkj0ZeUOPjnUEz\nHLjx2dLSYsgSYHPJTVtbm5VxluwzF0y8EmR6NIOGC15iYv2YdspvLpfD/fv3zXF4tNZyeHgYXV1d\ndSRUkLKUZardl78fF5oxaFM8jcLhZc0NJ45qtYpEIoE///M/x8DAAH77299iYWGhTpF1dnbi+9//\nPl577TWMj48jFovVpIM6GW0wSp2QCBg6zjSVSuHUqVOYmZmpEbpcgPoRcbIeaeSnt7cXo6OjOHDg\nAE6cOGFmX2gbcGsEyW4oUW7cStjSxdsC70+8f2kCSbYRrfz4dfrN5QEneyqVCg4fPoyxsTFsbGxg\nYmICr7zyitm3paenx4xS2ZZlaSNXcuYQryc/BSbTvRP1x5UPET/PP/88PM/Db3/7W9y+fdsolmg0\nimvXruG3v/0tnn/+eQwMDNQY6FLJSQNRA88XlQ3N3kmn03V77QwODmLfvn3wPA+9vb1IpVLI5/NG\n6ZFsp9EPkrUUDk2Npg2eiaxbW1vDgQMH8NZbb2FkZMSQTpSn4eFhvPPOOzh16hQ+++wzs78W8Ghv\nrosXL+Lf/u3fzMlg3BgmnUFotmG0nbC2+o7sn9IYpGtEMHAigvcDuk5LF/joJ8VD18h5orKWMh94\nRP5w8oFfk7KRo1qtmtkEZJhr5I6ENK6aAU0P8jhKpRKWlpZw4cIFvPfee3juuecwNTVVQ7pRW7ct\n4+JkDzc+Ad35p+v8m/JP1zS5y2U4PS8NX3qHhyVtNV4O3Nnm6dWMVE3nSrJOzi5pFviePZwc0JwI\nXubckZLQbMJwOIxYLIaJiQmzJNFPrqysrJi9tOhZbi9pttN2dU4otLnk5erVq2YAjLcP2c/m5uYw\nPz9fc7y8LdytXN8uNCeW16k2O4bek+2Y3iE7kWyH1tZWs8cml2vlcrlmo+V4PI5qtWpm93AbQw5K\nynSSsyjJHqC27wYldeQWAnxAK8hAFxFU0laS8nsn4GcLB20/vI9WKpun1lWrm3vTra6u1sjCyclJ\nTE9P4/Lly7hy5QoePnyIO3fu4OrVq7h//74ZINZsWs/zcPr0aXieh3fffRcvvvgiYrGYIfiBeiK1\nUtncFmJ2dhZ37tzBhQsXcPv2bbVPf/LJJygUCnWyUaZDlo9mb8v3dkIv+l3j6ZA6iEPqKhs5IvNN\nz8mBEunDSh2jpVPze6QOkLpCs7t4ffG0834kuYugk1lkmdiwK2QPfXMBqY30SYJHCi46zYjC0xwz\nbZoh/eeKi779nFhupHqeV0P0cAOZv8fzkc1m8fXXX2NpaQkA8PDhQywuLiIajaKzs7OujDQ0Uuba\nu7tFEmjx2eKW5cnJARlepVJBIpHAD3/4Q0xMTODWrVtYWVmpceY8z0NXVxf+5E/+BG+++aZZzsdn\ny5CDQmQPxUFkAR293d3djZWVFeO4kgInYSENA54XW75aWlrMpsDvvvsupqamEIvFTNxcYcoy3SnC\nx09h8rZsi98mxHkZ8W9tpEET0jajQQpA+qaRuEQigenpabz88svY2NjA4OAgpqenEYvFzBKttbU1\nY6xxskczhKRQtTmdfoSPzFuzQYYDX8Z48OBBJJNJ/OEPfzAzWGizzRs3biAUCmF0dBRjY2M167q5\nMqIwyVlvJFeIqCHSjeQdOTS5XA4rKyvYt28fTp48ib6+PsTjcWQyGTMSSmngJ4pQOyTHN5lMwvM8\n04eLxSJyuRzy+TxGRkbwxhtv1OyzRGH29/djcHAQV65cwccff2zIXQA1RNj6+jp+/OMfY2xsrE73\n8D6qteXdhF97sz1P4MaU1jZJb/F8yrhoen8ikVBnIthGF7khI/U9twWkw8UdL4orm82apSYLCwsN\nB114Gvj3bqFSqSCbzeLKlSv49a9/jWQyiX379hkHgEgeIscoz3IZFx9FlLaG1J+2fivL0gZZjlJP\n8/dtjjKFI9uJZmtpcUojXCN9mjmzhy/j4raDTKfMP/2XS+b4ezydVLejo6M1z9nK5c6dO1heXjab\nzEuiR3u/kb7m+ZTPZrNZ3LhxA563SczzOpdkydzcnNnr0K/t7SZ43+B6TpKE3GGWg0hUzsCjQUFa\nBt7a2mpm9UjiRTtJl5YNSyJlY2OjzhmkcKRPQZD2pvRvtMMjKC6+FJ3niX+0gXKNIKIwCDtJ9jwu\nZB/0vM0tNdbX19Ha2op79+7VtPFXXnkFY2Nj+OKLL/DBBx/g6tWrePjwYZ0c4uB18sc//hFXrlxB\nZ2cnhoeHMTw8bOwcLk/ok8vlMDs7i+vXr+Orr77CL3/5S3z66ae+/baR3Wm7vhVS4HHhZ6docdsG\nYiShwfMudZL0QWy+Cr8vdYuWTi298llNz3E9Im1ujQuRXIXtP7U3W1n4YVeWcclK4/81sodnlJxu\nzYi0Mdra7B4uxIFaJ1VzcikO3hDL5XLNJpV8WjY3pmXFyGdkGWiNmn/LcrSV8U5jqw4HhyZwOJEm\nGU4iRQYHB/HXf/3XGBkZwc9//nO0tbWZ03yGh4cRi8WMg6t1CCJ8NOVM9/r7+/GjH/0IAwMDCIVC\nWFhYwOrqat3Iilav/NPT04PR0VFUKpsbmr755ps4ceIEhoaGapS+NrPHVma7AU1IagYkb79+abQJ\nWq3/+/UDTsKQ403ld/jwYRw/fhwnTpzA/v37DblBy4doRg/fdJA266W16tygkSNVvH5lPhops52s\nN15G3MjjU7BpVkVraytWVlZw+/ZtQ3bw0zwA1MhV6XxrxgXPG5XfoUOH4Hkerl+/jtnZWbS2tiKb\nzWJ5eRkvv/wy+vv74XmbRA0dGc/LnDsSRPLkcjncuHED3d3dZhS8VCqZvXpGRkbQ09ODSCRiRltJ\nnvCNqLu6ujAxMQEA5uhhz9skY+m4WnkUuzT+yUlrJtmzVWf1ceLmBgaVNekkSotfnc/OzuJf//Vf\ncfv2bfzZn/0ZotGo6uBKwkfqT1s747+5Duc2Qjgcxm9+8xt8+OGHCIU2T+NaWlqqa788PE3n7xZ4\nmlZWVmr2WJDGpu3DN37keobbLEEIH5uNIe0e7hT5Qcpx2zvSCLeVU5CPbWZGM8CdV5IPWnzSOaAy\n5HmRedIcGTkKzO/z9tzT04Pp6WncuXPHnBgqyzNIWXBbt9Fz4XAYi4uLOH/+vHqPwlhZWQkU3m6B\nZn4CqLHbKN3aEjSePi7PJNHseZ6xMaUTR+2FyBDa+yYejxu9ImcRU3uT7Yn3e0n48nRJGceJHT7w\nLZep8yVbnAiSJBelTc4Q4v4U4ZtM9gB20vXBgwfIZrM18mt5eRkXLlwwqzHoMAxpF3M5y3VrKLS5\nbOwXv/gFlpaW8OMf/xhTU1PwvEf7y1L8uVwOn332Gf7jP/4DMzMzWF1dxb179+rsgka2poyf35O/\nd9o25bD5CTKNUndLHcavB9EnMg5pe2jy3JZeGwmlhW2z5ySZpPlAcpBR9jk+8Ey2aEdHBxKJhG85\nSOzqzB7bt8w4LwBOrPDRDTIMZaH4sWHS2JEfui7TRddIgErFDzwS/rFYDJ2dnWZJQiqVMrOAKCwi\njGQla2TCN1GQcsc9CHidEeQohnS0qJxTqRRef/11VKtV3L17F6lUCnv27IHnbc7s6erqMu/Iuqd4\niZTjTnIoFDLKrq2tDdPT06hUKlhfX8fnn3+OQqFQo3ApH7yOIpEIurq6jPMzODiIPXv2AACSySRO\nnDiBgwcPIh6Pw/O8mpERbTrvTsOPmJGCS3PyZR/yi4fXtxSotj4v//NRqpaWFnN0fWtrK06ePIm3\n3noLY2Nj6O/vN32KZn3wted8JMsmQLV0BXV6dhOynKhsQqGQ2SeF+lRraysymQxWVlZqyE3epqWD\nR/0DQKClEuFwGJOTk0bxzMzMmE2aI5EIOjs7DflTrVaRTCaNUcQJCJKXGxsbWFlZwfz8PJaXl9He\n3o5oNGr6aiaTQXt7O/bv3282POVkIBmplK/R0VFMT0/j+vXrWFpaMks9acZfLBYz5VYoFGp0jyQC\nm9lHmzkzYStxckeHp8E2sw8A1tbWcObMGXR3d+Odd94B8Mg5omfX19cxPz+PhYWFumUHsvykfpMD\nPPw/ACwuLuLcuXP48MMP8atf/aqGCKETFKWs4gMJhJ2WsZpeDIU2R3Dn5uZw7949zMzMoKOjwxxv\n7Uf02MgeXvbcCKa8y2/tQ/1E2j6a3GskE2WcQQxkec1mi/kRP80C37Seb5LN9RbvM9L20/LB/3Py\nUyOAZPnT/2QyiVQqZQgE+SyPA9i0T7PZrFmWpsksmxPG72ezWWQymbo02hzRIHbFToP2SKF0auQO\n3+NEygsus0iv8t+cJAdgBkzoGg0m0ZJj0imk0+SMGyobKY+5bPMbbLSRPXKGj5zZw+0gPluH2gov\nO4pTm/lMkINhzYBGWmy1Hdn6JuUzk8kgnU7XxDE/P48vv/yypo3wWRkyHN6fuUz94osvkE6nMTo6\nimw2i+7ubmOjAkA6ncbVq1fx4Ycf4v3338fy8nJN2Frapd/Ky6eRvdrIlm22bcN1L7UPLW+aLONp\n4t/yPi8T7TtofjTfVCvPRra+1E9aunn6NJun0eQV+oTDYbM3GI+7UTp3bWYPfUuDQzNEbAy2DFMy\nYbJgtPB4emxp1dIFPFr+Iw1Vrlj6+/vR0dGBtrY2tLS0YO/evcaoorDi8bgRwHLHe61yvwmkj2aM\nyQYdxBDk4IJAi8vzPLPE48SJExgfH0dLS4shTyKRCFKpFKLRaN1+K7K98DRzg5mcYADYs2cP/uIv\n/gKJRKKO7OH1R3XS3d2Nt956C8888wwqlYpxPhKJBNra2tDR0WGW/vG08eUh3IgKSjI0E1KJ2BQO\nz7ufYLFd08LjfQio3ySXTx8eGBjAa6+9hsOHD+OZZ57B6OgohoeHTb0QySPJHr5sS8oGTvLIfPqV\nF+WnkeG8U+B1UalUEI/H0dHRgfHxcczOzmJ5eRnVatXMXInFYmbJVSj0aLM4AHWyjspHKlYtvxR+\nPB5Hb28vDh48iKWlJdy6dQvhcBhdXV1IpVI1SzDJQOb9k+IOh8OYn5/Hhx9+iMuXLyORSCCRSCAa\njZo9ebLZLCYmJvC9730PY2NjyGazKBaLdWQuHUFPexWtrq5iZmbG9DVaykUjbvwIXY2I5w5BM7BV\nh9jWtjQDSXuG7nHnR4av1THdozLlTg430r766iv89Kc/xc2bN+t0hHQIeN+X5U3fvB9++umnmJub\nw+zsbJ3zpjlumpG5U+AOhF/5h0IhXLx4Ee+//z7eeOMN7Nu3r87J4yQP/w2gpozIqaBwNXnW6EPv\nSj1JkLrIZgDbdJafwW2TLdLx5d/8mWbXabFYNGnhm2TLdGmzQqQTKB1DbTaS9pvbPPxdz/MwODiI\njo4Oq6NEv9fW1nDt2jWsra2p4XNobbWRzA8ia2Sadgs0A0ObHSNnaUmZQTY89SVuh0gb1fM8Q8aQ\nDK1UKobsIZKH7FZO9nASBoDarjSyh+9vJwe1iYzRjkqX++/Qb1oWyLcUoLYvj5SXZSLJnv8L0MgE\njSzg9xr1ay184NGSnrm5Ofz0pz/FuXPn8Pbbb+Pw4cOG7Ll58yb+5V/+Bb///e8NqdrI9rbpZd4u\nn7SPSOBLHWXf0+RcI5+DYJNF9Kz8bat3v7qUYTW6buMvuL2l2T9+xI5tlZLfoGMQebtrGzTbjARA\nZ6zlKJ/mHEiSx2+WjIyPOqUc2ZZCgBwfyfZzB5UrFBKUdI9GsrlTRUtN5LRKnhfNGdXKUftuNrTO\nItEobptBTPWodWoq52g0iq6uLvT19dXEVa1W65ScNtWUC0T6pvqkNJHTPD4+jhdffBHAo7X8kuyh\ne52dnTh+/LjZ3JXSQKyrRkLKmUcyjdx4bzYaGWuNhG4jpyYINGKBC0hJ9ITDYRw9ehRHjx7F8ePH\nsX//foyPj5tTgYjYIaKHb8iszeax9SVbedkEPN3XnpftrZng9cNlZCKRwNGjR5FOp/Hxxx9jY2PD\nLFHyPA+ff/45otEo9u7da/YS0GSjFp9U1lRffPQrGo2ivb0d3d3dZko9jUrTaWDa7BFuZIZCIayv\nr+Py5ctYWFhAKpWq2Zh5Y2MDiUQCAwMDZjYRX8JFfZ72QPG8zVMSY7EY+vv7kUwmTb3Q/gvLy8u4\ncuUKksmkWRZGz/B2uZuE+1bi8XuW91ebY87brPxNjgfpuNnZWXzwwQeYnJzE4OBgTR3evn0bX375\nJdLpdM0ou0yjjXzgbZEjFAphcXHRLNnSnG4p17h+32p5bgUUrxzJsz27vLyMe/fuIZfL1Tl2UqZI\nmSuJHnqWywLev4LoY172UpZJ2eBnx/G0NCIZZNnxtqbNPOMznOSnWaB9oEgWcQebp0Nrd1I++hFB\nWhloYci+SIS9n9MCbBIew8PDaG9vN7KUiB9O7vr1iyC2ngxH/m703k6Az8KyOc/SfpHtmeqZwpGg\na2Sf0ICg520uUeYkD18aLG1T2nOQ2ps2s0cSVhSvRvZQevigIr+mzfahsEqlUk3/5+nQ6lG2pWaj\n2bJaSz//rdm0NlnG35dtSsaRy+Vw9epVQ+zdunULIyMjADZnP3/00Ue4deuW2veDQOp2uub3vPab\nl0OzUCwWjS+syUsZt62varYKf8cGqZO09/3AbSAui2XYMp3SluFhcciBRLnnsMZp2CasbKXN7NrR\n60D9iB5d0xxAObqqGXRyVo8cvdcIHwIncTSFINNG78jRXhrx5KOltgZNYXEBTEy7VAiyUckwZaXv\nliNC0Ax5G/zSKA1NXv68/siB4+Uoy1J2EG5IAbUsK3f+KXz6/9xzz+HgwYNmM1luwFEeKpXNpYX9\n/f1mZhHFR2ni7VSOjPCOy+vY8x5tIN7sERRN0WnKws/A30ocGqjPaOXBBR+NQnV0dOBP//RP8d3v\nfhcAzAkHuVzO7FMjZ/JomyVKQnGn4GcENAOcnCaDjgzNkydPolwu4+zZs2bEiEiNn/3sZ7h8+TL+\n4R/+AZOTk1ayx2Ysa6Oh3Kik/y0tLejp6UE+n8fa2prZZFmOSpCDGwqFzFG1wOa+OgsLC8jlchgZ\nGTGkaS6XQ6WyuVRycHAQiUQCoVDIGKt8QMDzNpdMhsNhE/bAwAAGBwexuLho9l5Lp9NYXFxEsVjE\nvXv38O6772JiYqIurxRmM7HTbVFzzBvpRECvf9JtFy5cwLVr1/Dd734Xb7/9tnkegDk1SCu7IJCG\nHJdP0hFpZHxzW2InHRM/eaoZqfy/JHr4Nb4/FNdXWpz8tzaCqOkRjbCR/VPLi6298DLW6pHf0/7b\nZjaFw7XLTsmZJoe6WcjlcjVyiT58qSt3wKVjLvNmsyeB2v1gePlps08kGjmf8XjcnHwYCoVw7do1\nrKysGLsnaF+w2Zx+bfybAjkTB6itA9mPeH2QvtBmr1CZE4njeZv79OTzeVQqFdMmOTFJ9cln4Miw\nbeSh/HC7SbMp5QwAv8Fvyg/1L2obcq8jnneejqcJ1KZtcqqRMy3bGX8nHA5jZmYG7733Xg1BTcst\nNV0mZQeH9p+nXxtI4+/Z+myzbVU6NYzsQblBMY/TL+9aWqVtouVFe5Zf056VYci+qIXNr0k9qpEx\n/BneN7UTgSURxPvwdutqV2f2aIXFC4IyQtNo+bIpcq5J4NA1G+GjzZ5olE5Kj18joDTSOl7gEWEh\nyR6bMSiX9dicf27sSgPtSYGnRzP8/NImhRNdo28+Ms+dE17XHJwg0Ka8UZ3wTsadQgqDRq4pD62t\nrUgkEjWKkMKRRgMt05JLtHin10ZJKS5edlQOtnp/XNgEnZaeIGFp9S//a4JOM044yTM8PIy9e/ei\no6MDAwMDeOGFF9DX14d8Po9yuYxMJmP2MZDLtfioFS97Ld9B88lhU0DymWYqTxmXVBpESKZSKQwM\nDGBsbAwAaoy4XC5nDIxoNGpOAtKUkxxptClibiSHw2Gz5w31V76ngdb+KWwipL766itcuHDBzEqi\nDQ1pw+1YLIaTJ09ienraTI/XDFkqfyLSW1pacPjwYRSLRZw+fRrZbBaxWMzET+2Oln7RTCRen82W\nuc0mciU0R106CZoskjLT8zzThujY+3PnziGfz9fEd//+faMTGxnIMp0Eeoc7FTJdsl3yZ2S4zZaf\nEtII5GXGnSqSWfPz8/j666/xySefoFKpYN++fUilUsbp4rKQO3gUDtlG9F/e0+SQ1Ndau+N2jbzO\nwwH0k1OkUUzX+LcG6eDK5StyJg851JyIaQb4fjicTOKbs8rTnOTMLJ4P6RxTH5K2gFZn9K5NPmhO\nKP/maRkcHDTXS6USHjx4YPqt9n4jyDYUiUQwODiIcDiMhYUFs0z4SUDa77z8+G/NYaa64HvVSB1O\nv+WmzNVq1QwsyJloVMbSlpWElK1Opc7VBrEpfzxsTddqfZnkDrVPSTrKtMoykmneTTRy9Lf6rnbf\nRhIEeY/Lb9oImqdP02H846e/NHtWXgtit+4U0um0Sh5q+QPsM3ZsPorfrDP+20/GaeXG73EZrg1g\nSVsYqJ8Ny5+VtoEke2wrlOTJemQbbGfwesfJHm6AE6gg6bfmRNBURwqDb5Lm5zhKNtvPUA9SWFon\n54SP53kqc2kTDpy4kMa3Lc4gzqqtkzQDvGPwhivjD+pUy84uOwy/Rh9aSy/Txeudz+CQxqlsLxxE\nJNI7NPJPe/DI+g2FQjVhyRkpUsnaFK+fINoJB7ORArAJPp7erShAXr9A/Saj2jTkcrmM8fFxvPvu\nu9i3bx+eeeYZM4Wd+hwRFxrR08ipCQKtLwXtn/L6Vg3qRtAcyo2NDbOXVVdXF/bu3YtSqWT2OOGb\nEdOntbVV3duK0tpoajE3Aok4WV9fN/sB8NPR+FGzAOrCjkQiSKfTOHv2LM6dOwfP85BMJuuOW+/o\n6MCrr76K/fv31yzflOXCUSgUzBK3aDSKP/7xj5ifnzeEVyKRQDKZRCKRMMsvyZmU8qmZ/VEr92ZD\nthNJQnNHh4M7rhQOEd+VSgVff/01rl69au7RO9TWpDFmk4WaDrFBc0Zss4hsBlezIQcAeHlzO4VI\nx5mZGbO/VD6fR29vLzo6OozuoQ+1aZKLFAf/cKJH02laOUinV6sfDfK+lGt+deEHnh/uHHCyRS6N\n4bMmmoVsNmvSQDqI77nCnVzu1JPs4jPatFkZsj3IPiNn+vBvWxvWZLOU2/39/Wa55fr6utGZPGxu\nt2lxEjTnJRaLYc+ePYhEIlhbW6s5vGS3Iduz5ndwO1vaNFuVUa0AABzjSURBVPLQAj/dR+/T7GOC\nDEuTgfx3I/tLyjDyHeTgMKDrE963bOQC9SONPJfxy5k92+33uwnZtjmC2GZ+fUJ7Dqg97MCm7zQC\nRPqPPO028lf2Ye0e/78bSKfTde0OgOkbtvJu1N8oP376369MZVj8v5YW2W9svo+0s7T6pvvcZ+WT\nFfgkAcllSLKH66VGNhXHE9mzRys0XmDSOOREjxToJAD5+/zTyOCzKTqtEfF0EQFF31pntcUlZ6zI\nqZWaQxfE6dwpaIZhUId3K3EQeEOXI2pSGGr1Td/c2JLtS1OWFEYkEjEj/ZKdlqMevGNSZ+TtXTOQ\nZflJQRK08+4U/Iz/rRhzmqEi15BT/9nY2EBfXx+mp6dx/Phx7Nu3D93d3fC8zc2Xaco0X7pVKBRq\n9ryysd1PyvjcqXDlhy/n6unpwfe+9z20trbi7t27NU5US0uLKfN4PK7OhuF9TZ5II/PG+wHwaIkd\nX+IqHUyu7KgvUf3FYjGkUimzLM3zPFPf6XTa7AHENwCkcLXZPfQ+sLmZdDwex8jICNbW1ozTE41G\nzV4q6+vrNVPyKc+a4/BNgZ9jxuWc1I+ajLI5k3IWAy2Rle/anCRqJ379kOs9CldzLKSz7TeTqFGc\nzYA0yoO2keXlZZw6dQpHjhzBCy+8YE6JAzbTzWe2SLnGy5j6Ev3nJJCsd06Q+DlBfnmV/21GteYk\ncd2tkSLyw+WWtuFus0Cbx3Oyhwh0nhaN0JGb2dryIp0TKi+yUwjcobbZzbw/SUKBy3G6Fg6HEYvF\nsG/fPoyNjSEUCmFhYcEc565Ba8ue56G/v9/MHKXTFrPZbKD3tWeaCc2B4+XQ6F0bMaORIEHTEiRe\nP0ee/+a6U8vr40DKsN2QnbuJ7ebH7z1NhxK09m/Tr5pc5XHb2qYNUsbIsGzPNwvr6+s1toNmG/jl\nWwNPtx8BQ//9CFvpW9nKxI/ssYWn+X30LX1VOUlF+rCab0krH77zne/gyJEjAOwziiSe2MwereDo\nWZqmRBnWHG0KUxup1Jz5oA1Je9aWVi6obQ1PU5i2D4+LDDn+jpZuLW07IaS1NMo4eZ5teffrMDIe\nXv/cOeRlrY1W85FsLQ+29kIfYk/ltHIuAOR7fBmXlhdpMMiZbbZyeVIK16/vNFI8Wr1zUodvEtjW\n1obW1laUSiXs2bMHb775Jg4fPozR0VEAMCRPPp83J17wDZg5483rWyMpbGjUJrVrjYxZTVE3A5pQ\nJxKtVCqhvb0dL730Eu7evWscllAoZPboWVxcxPDwMNra2mr22+HtU+5PwdOvKUhyMOk9TurZ5B/V\nFZFDfJYkX0JJRA+FRfWtyX6N7KH0U1l0dnYilUqZJW2RSATr6+tYWFjA3NwcRkZGTPxBHYXtoJkG\nurymyVCpF/0cDN7GSdZxp5fu8Xe4bNScUFtfkKQEzxdPBzce5SwLzZCUadwJbNUQ5wNE6XQa58+f\nR7lcxtDQEAYGBpBKpUwf4B/a60Ea8WQjhEL1p2rxcpWELqVFK59GzoCfnmr0nOYAaISIjfiR15oF\nWn5E8dNeEzwuuVSLX49Go9b0c5JIW9bAZS1v97z+tDL2c5q4fULh0ObNvG5WV1etdqutbfT19WFq\naqrm+Uwmo5YrDyOXy2FtbS2wTb5VaDLDT85pDiB/j+sOP7vXlhbNZvfzIxrlx+Yv8GeCpEuDJsu1\n8thKmLuBrcYt217Qem0Uj2YTBQlXPrsV/dUo7TY9uxv+Iu2nyD+09cJ29bSsOz+yx3ZfhqfZHJp8\n1cKy+Th+Npkke+LxONra2mru8fRw2cW3fpmamsLLL7+M/v5+o/f9+ihhV49et8GmWKQRxQ0UP8e9\nWWneaiO03bM5PH6C+0kK0SAIkj6/8pOdSgtbU9aaI2ErR9l+uAKnGT8yLTxebkSTw2NTvjIcrc45\n+EgrxRXEYH4c2AwEnl4/Q0WWp5/wo2skwDjRQyQNABw9ehSHDh1CsVjE6OgovvWtb6GnpwctLS1I\np9NIp9PmRCd6V+5eH1TJavUi39mO4eTXfptdhzaDkYgbWqYVDofNDBkiLNfX13HmzBlUKhW8/vrr\niMViNdP6qV8A9WvKbY44xS/7LPUXWSZSidJJXuFwGA8ePMCDBw+QSqUM2UMze3p6ejA8PIxwOGz2\nKApSV9Vq1Zycks1mzVKGUqlk4qWj3ekUs1dffdUs55IDDM2CzQnZCvwMC25gaN9bcbp4XZGskoM4\n0ojj5Sbli3Si/PQkDzvIbBCbEbhb0PQZ5bVQKGBtbc3MBAaAe/fu4f3338fJkyfx8ssv1/UNuYEz\nlT0NfvBZPfzDjUqu76TO22q75nmThLpW5prxzPNimzEThAhqFkgXUZ5IvmjxyvTT8jKN5NGWfknC\nhw8o8YM+JNHM+41WjhzSnuDP0LWuri4cOnSozumyDUJROtra2uoGB/zaEt27fv06PvjgA7Psid77\nu7/7u8etPis0QtPPAQQaD/A0CkuzWfnzQa5zf0faZdvRG3KmO5e9QXyvRrrim+6rbBfNyJfUdfw6\nv8/7KT1PNg7Xt5Qurf81atvavZ3Qje+8806dLjp9+jQuXrxYk9dGaZb+ik0G8rzI8rTlsRFnYCtf\nLV1BwpO2V7VaxfT0NN58801T31p+aZsBPguovb0d8XgcpVKpZpZ1I+zKMq4g8KtwrbP4ET1BCBgZ\nt82BlXFvxRnU0q2lXYarXZNhBekgzerIfkY4/d+q82DrsJpzoMXRKG1+73Ahyp+RzqlGLtri5wLG\nry55XNIoaySgHheN2q4muCgtmmDV3ufvaVMRaUZOZ2cnBgcHcezYMbz00ksol8vo7u42Dn0mk0Em\nk8Hq6qp5R27ELZfu2OQEx+OUa9Dy28o7W4XsK/QtlxHGYjF0dHSYsgqHN4+pv3jxIvr7+/HKK6/U\nOBgUtnRwbPnkylcu66HrfGYQN1YofHJ6FhcXcffuXSwvL6NcLtdsIE3L9fbv34+jR4+io6NDVY5a\nGVEfrlQqKBaLAIChoSE8ePAACwsLJt2tra2oVCr46quv0NbWhm9961tIJpM1ZE+zsZ22FOQZqVt4\nH5HXpUEjHUwOWWfarFM/ua7BJit5nDJ+jQDg96TRLA21J4VQaHO/sXw+j5WVFXR1dSGZTJrjsVOp\nFDo7OwFsLo+hvXzksey8v/ElW1p83HGwtRVpwNpsFhmP5mTY9IMkG2SdacRIEEKvWeBOtY08k22K\nPrTHkI3s4cQOnyXMZ1xqs4i1PGrp8BvhlnXDr0WjUfT19Vmf097h7UnWMQc5qp2dnQiHw0in0zWy\nqFKpYHZ2Fqurq02tQ55/aUdupb1IHSsdTdvzjfwE3t+CwuZMbtWu0Gw6zY73S39QH+Rx4Cd/tgOb\nTvGri+34e/L5ID4QfQfRm3JwWqZf+x1E9+1EHfKTOmlG3/37983GzVxu0XMyHbKN8z5oe0eTi7Y8\n+vWlIGVi64+RSASpVAqVSgXpdNpqJ0UiERw4cADPPvtsnR3E+0AymTQEu7Tj+Kz8IHW9q0evbwVS\n4NoEcBCDvxH8Oqg0iLbS+W0V6Ce4beSALX27Ae68BzG0ZF3Je1qH1OILcr1RW5DX+MgGGSA8XTJc\naURQZ9MMrHD40VGWfFRVE1o8DJvw24361srBD5RW/q4Mh88iILKBZuWUSiUUCgUcO3YMb7/9Nqan\np7F3715zpCndz2azWFtbw/r6eg3Ro8Wp5Ynnw1bHPL9+bXY7xlUz35fQ2imVDS11a2trw4EDBzA3\nN4fV1VVDnszNzWF+fh75fF4lNKQzJtPvJ7M4yOGRe4jQPeoroVAI58+fx+eff45MJoP29naj3GgJ\n38bGBo4dO4a33nrLkD108peMX9YntcNisYhkMok33ngDkUgEly9frjlRr1KpYH5+HgsLC6DZPvzk\nH04qNht+8n8r7/Jrsv9py51lf7At7eEyipabaDrRzxCT4fHvoM/ZZlfI37Y0PCn9ydt+uVzGzMwM\nPM/D3r170d7eDgC4cuUK7t+/DwDo7u7Gm2++iZGREZXskXn025yV4tZGhjWZR/dszp4M35Zf/gwn\njmV9+e3J04jc2wmQLJX5tLWzcDhct6ek3yweP7JHy79t+ZiWHq3ty/Tb+nejvMoZR41slEgkgomJ\nCUSjUXz11VeYmprCX/7lX8LzPGSzWfzsZz/D559/3rR6k3Ya1Y1Mn8250+zRILJYyj9bedjC2o5M\n0nRws+xFKduflMzcKWwlT8222TQ9qT1HH232rJ+vKsNp9NxO1S3NUAWAO3fu4Ouvv8aJEycwNTWl\nyiyt7Qax83k7pf9Sfmn5lP3Hz8fW4pZ2FLetOjs7ceDAAeTzeVy9etUsEebh0EBse3u7OXyGxy//\n81NutXJpJN8JT5zs0RInG/R2DN/tdFSbIG7U6LYavp+zZGuEQcPeCWjpC2IQ8v82BSuVq197CJLG\nRs/yOD3PqzMaZV7lOkx6RzMObPlqlF4pmP0EVbOgCQ8OWz64AyHD4iN3ZDTTf9pTprOzE4cPH8ZL\nL72E6elpjIyMIJlMAkDNSVu5XA65XA7FYrFmM2e/fNB/mT56j9e1JtT9ENQB2sl+aAN3JGn21NDQ\nEL773e/i9OnTePjwoXEuyuUyZmdn8dFHH2H//v0YGRmpac+e59Uda6yVsZz1xr81hc6VEi19SCQS\naGlpwcLCAq5fv462tja0t7cjEomY2Vy0rnl4eBj9/f2oVjePupUjHdryJALtlRKLxdDb24vR0VEM\nDQ0hnU6bY3PpuQcPHuCjjz7CgQMH0N/fb0jIZhnTPE1aHcrfftf8nuVl4lc2QdLAf9MIo9xjhKDJ\ndPkt+wyHJgul86Y5Z1JmSodXi6uZCCI/CMViEfl8vqYP0RJD4NHSQ9vMHvrwWShafNxpaKRngxjX\nWn78vkOh2lOs+HW/WTDyN39Okg7NgKbLbGUujWvqX5Lskeml6xrZo5FddE/b/Fk7MEJzoCj9GhGq\n1ZWsX00GyHZRrVYRi8UwOjqK3t5eE04kEkFvby8ikQjGx8fNLKJQKIRsNouVlRV0d3c3rQ41skfL\nqwYqI760nkMrh0ayJOjAgJ/Nq/VJP7vNz8bW7HCb/t6pQY2dxlb8xKD2ud91m9y1Yav2oZ/vtJ1w\ndgupVMrE2d/fb/alKxaLNfJweXkZ8/Pz1nCq1So6OzvR09NjBt8IWtvnuoVkjLbMKR6Po6enx4TH\n+3SxWMTy8jIikQh6enpMWdv8QjmQlkwmMTo6ilKphO7u7pol2xRGa2srksmksbelbLbp9FKphKWl\nJeTz+Zq0aIS/hidO9gB2Acqvy9+NCBN5bSfSy9McxJDW0q59NyP9zezgtrRuJw6/PNmEGI+Xz8xp\nlD4Zti3NvMzlOulG4fHZD/ye1vlsylmWpwx7p2DLn2b4BTEG5UwCfqoa7dczPDyMd999F0eOHMHk\n5CQSiQSAzQ0c8/m8IXlo+Q5fthXUWdWMdw1+daPlVcohm7LfKblD0PJE5UMzqMbHxzE1NYWZmRl8\n/PHHxmlobW3F3Nwc3n//fXz/+9/HM888Y5wHyiNv03KdsRa/rBs/AjUUCplj2Wlz7kwmg/n5eUxO\nTpr9hiqVCgqFAtrb29HX14fOzk5EIhEz00cu5dPkKq8fIiBjsRh6enowOTmJ27dv4+HDh2Z0JRqN\n4sGDB/jZz36Gt99+Gz/60Y8Qi8UM4bgTMtXvN78WtO0TtkL0+MlcDi6XbAZzI/kqn7E5npoj6jci\nKMPcqsO3XfD+LkkCCZuOkPc5MeBH9tgMvFAoVLcciTtwNFNNS5+fM+lXBjYygZMT/FlJXGgzWzTy\nhzsLzUKQfDZy/uSSbJrVqM2OkQc+aHknOUmzhrT3eRj8vmwbfkQQl/d0TYKu8TzyNplMJnHgwAEA\ntUszKV66R+/SCYnHjh1rXDkBYbMHpS7yyx9Pu1+49KzNX3lcaH2wUd/UbB9Z//JZ6cjyZ/6vEj7f\nVPA+p9Udr4ud0FO7Cb7hfVdXF7q6uup0eDgcxsWLFzEzM6POKCbbhfYW6+rqMpsZ23xkLt/u37+P\nS5cumZMW+XMDAwM4cuSIORCBv7+ysoKLFy8imUziyJEjaG1trbN7KW3aVhIUjud56O3tNf/9EERW\nA5tH2mezWaTT6bq4JKGv4YmRPTZjiK5LgUTwGxHl/7dj4PK4NWfWT8j6OXkyT5rz5PcdBFr8zRYa\nNsOG35P/+W8p1Gzhy9/yGWl02BShnB0jFZ801Kk+eOfV6ptGgDShI5WnVh42gU9x0ygTT3sz4NdG\ng75v67dSEBJJQ7Ny6JSoY8eO4fDhwxgaGkIsFjP7WEiSp1Ao1JxspqHRvUZO3nbKQnOsbPHuBPyM\nMCr3UqmEZDJp9v0oFouIxWJ1J8vRyDFd5/3Gz+i1oVHeKU46wYaczmg0imQyaYgemtlQKBTw6quv\n4pVXXsHo6Gjd3k2S6JFp5ca+53mmbEqlklnuVigUatIUi8XQ2tqKaDSKRCKBcrmM1tZW40A3CzKs\nRuUd1PmW8sxPHzaSwRyN2j2/pjkWfuFpTqiU0zZntVH8tjiaASlPG+Wbns/n81hcXMTNmzcxNDSE\n/v5+0yfpOe7YS32ilYWsX2qvtuc0udXIobSVAU8b/y1n72gGv0b2+C3p0giKZoCXsVbmWl6pPvno\nMW+3tj355MwcWz75nmr8WVv58GuUTr+ZP3z/LXpG2jO8jWt7RMny4fpJy38otEmqT01Noa+vr6l1\nuF1oNiCHzC89q9m8Nh24E/Kn2Whkf8v0B5UR33TIeqTfQf29RnpHhsnD4T4Rj9PW5rZjw+92HdHM\ncCkvNzY2MDs7i2w2i1AohAcPHpg82cpueXkZX331FWKxGOLxuDkpVasbHt/y8jJWVlawsbGBUCiE\nRCKBWCwGYHNg+fbt22htba3TmXTab7Vaxc2bN2t8sWg0iqGhISSTSZXssaWpkf/B5XM+n8fCwoIh\nxPm7tK2Bn23khydC9mxF6DVy6DRjtpEQ8iMSKH02Q9hvumijSpXpluvoG6VPgxQEQdKzVdhIHr8R\nfHlNfjeKy/Y+UL+RsiQaCHxarjRA6RoPQ8ajtSlJ9NhYadmWNEdAxsXrjBthzcR2lAW91ygtVB58\n6ValUkEkEsHQ0BBef/11PP/88xgfH0c8HkdLSwvy+bw5IYmIHjpaXQpiTZDanBMuDDWFstUy4G1F\nloetD24nnq1C5onKHYAhVIBHMz3IISAiJBqNGsKHypDaN29/jQxYm9PL/4fDYbOEixN94XAYqVTK\nnI5VLpfNUpdnn30WP/jBD7C8vIxMJlM3q4fyzePhRCnva0T2EMEEAIVCwUwRjkajiMfjhuyJxWIo\nlUrGSGgmgg5a8P9B2pKUV0F0YRAZoxmf8r7NQLW9I9/lRox0ToPO5pDx8vB30tnS4tXuV6tV5PN5\ns+SyXC4jlUoZI5Y/z2f2bEdehUK1BCUPQ2t/mr7zi9fWz3mdyVkosj4bET1ytstuOM3cttHatNQt\nstw0nc/zre3xwwkeelaWG5WF3MiZzwKjGZxALakm+xDpBZL3PC/8Oc028rMzZR3ZZhoNDw+bo+B3\nEo1sCHnd5kDK53gd22xGGQ4nERuliaclqO0s0+sXpsyL9owW9nb10lawk307aPxB/BSbHgz6309m\nBC1TLQ1BdO5Ogp9uyPV5NpvFvXv3sLi4iFAoZJYt2+R5OBzGysoKlpeXUa1WEYlE0NXVhVgsZu1j\n9J3P57G6umr8Nb4klsgeWxih0OZyrkwmU9O+29ra0NbWhmg0amZtcvksdYHNduYrIOg+yUkqo9XV\nVauclQNDQfXhN2IZF0cQYSSfbbbACeK0ybRtJX5JSvjF802Bli5O/gQNQ+sU/L7mpHBnnTuh0iCR\n7UHOvuHGBv23OakyHTYDT3Y2fp+H20ihy3Ll6W0mGrUvTYhoRrw2tZGnn0iH1tZWHD9+3Mzq6e/v\nx8bGhqmfbDaLTCZTQ/IQSbQVgeZnpPFnHrc8vwmGiPxNCo3+07K5QqGA7u5uTE1NIZPJmJOustms\nOQUomUyaMqeZNqHQo+Od/Zw+qhO5zE4bXSbFSIRKMpnE5cuXcfbsWSwsLCCZTKKlpcXMBMtkMlhf\nXzdTV4vFojl9QFOmPH2cDKZyofZaqVQwMTGBn/zkJ/j5z3+Oa9eumeMtY7EYKpUKZmZmsLCwYNaY\nJxIJhEKhGiOmWZBluxU54Xedw6/NBm3PXP5q79p+B02Dn2PYqP9radJ0x07IUy0vMm2aLUOndczP\nzyMUCmF0dBQjIyPmeeovsr3LspEkGN2j/kvXqF9rG6ZTOm1Ej5/zJ3/zb78lR1Iu0Me2WbHMe7Pr\nkZcH1+2aA8b7As1glWHJNNrKR6tL+vjN3rGRPZzgtxFK9L9SqdTMMqI46R4fGNPyp/2meKWdYKs/\nuT/cdhFENjRqxzIMXv+aDec3A1aTA7Y4bOmWy5O36ufwdMt68POttDLjMoTf2wmy52mBrW79/AGg\n3k+ia/z9J22H+uHSpUt110KhkLHrtBmaWr/gPjLJkkwmg3w+72uPAo9mIfIZM9Vq1WwdYINNdnue\nh3K5jLt372J+fh7VahV9fX0YGxszadO2mtDytba2hrt379bM3qFyoD03JaEjv233/PCNI3sINsJF\nM+Z2On6bQLe946dUtiMctXQ8ic6uNeStpkkj6LR63ko52YxTv3rQ0sqvawpWI7f4MpFGQlxLt5YP\nzYjYTfjVYxDhQvmgjVwnJyfx3HPPYXx8HK2trTUb4xLBUyqVzEg3X74VxMC3Cdhm9JEgZbHd97eD\nRg4EUDvDJ5lMor+/3yxfCoU2Ry1WVlaQz+cRjUbNB4CZFUThaDPMbL9lndFvbfQ5Go1iaWkJFy9e\nxPr6upnVU61WzYls9CGiJ8jeTZpylUZAb28vDh06hLNnz6JUKtWcykVls7a2ho2NDTP7qdGSwu0g\niCPidy8o0cMdlseBZgTZ5Kjtfb9w/cgeIPjpOhSmJOl3yki22QkyHZIQLxaLWFtbQygUQkdHR817\nclkPzQrQyEF5j8/a0AgdTmbIe/I5W75k3vlvXl9+s3NkHuVHIwh2oh5lW+Fl0gjckZZhNnLstbLg\n7/FlXEHIHkovJ9a0MqVnKP2cHCKZz9uUlnY+G8jmjMi2x50arbwfB5KU2ooN6ZcWm02nhd0oLls4\nWtzchrLFp6VV+8/7sU1fN0pr0DJw2ISfDRq0TQZ9/kn4gn6gUyUlbDJcIy+4zcnbb7FY9I3bZhPT\njG4auPN7X0sb9cWlpSXzu7W1FePj4zXPBqmrUqmEBw8eIJPJmGuS9JKDpY3yGMgXcZ3VwcHBwcHB\nwcHBwcHBwcHB4enBzh734+Dg4ODg4ODg4ODg4ODg4OCwq3Bkj4ODg4ODg4ODg4ODg4ODg8NTBEf2\nODg4ODg4ODg4ODg4ODg4ODxFcGSPg4ODg4ODg4ODg4ODg4ODw1MER/Y4ODg4ODg4ODg4ODg4ODg4\nPEVwZI+Dg4ODg4ODg4ODg4ODg4PDUwRH9jg4ODg4ODg4ODg4ODg4ODg8RXBkj4ODg4ODg4ODg4OD\ng4ODg8NTBEf2ODg4ODg4ODg4ODg4ODg4ODxFcGSPg4ODg4ODg4ODg4ODg4ODw1MER/Y4ODg4ODg4\nODg4ODg4ODg4PEVwZI+Dg4ODg4ODg4ODg4ODg4PDUwRH9jg4ODg4ODg4ODg4ODg4ODg8RXBkj4OD\ng4ODg4ODg4ODg4ODg8NTBEf2ODg4ODg4ODg4ODg4ODg4ODxFcGSPg4ODg4ODg4ODg4ODg4ODw1ME\nR/Y4ODg4ODg4ODg4ODg4ODg4PEVwZI+Dg4ODg4ODg4ODg4ODg4PDUwRH9jg4ODg4ODg4ODg4ODg4\nODg8RXBkj4ODg4ODg4ODg4ODg4ODg8NThP8PXlCuPhsG2EwAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f07e4ddf350>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from PIL import Image\n",
    "import sys\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "\n",
    "plt.rcParams['figure.figsize'] = (20.0, 20.0)\n",
    "f, ax = plt.subplots(nrows=1, ncols=10)\n",
    "\n",
    "im_samples = []\n",
    "labels_test = labels_test.astype(int)  \n",
    "for i,j in enumerate(np.sort(np.random.randint(0, labels_test.shape[0], size = 10))):\n",
    "    im_samples.extend([j])\n",
    "    ax[i].axis('off')\n",
    "    label = labels_test[j,1:1+labels_test[j,0]]\n",
    "    ax[i].set_title(label, loc='center')\n",
    "    ax[i].imshow(test_dataset[j,:,:,0],cmap=\"gray\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Plotting the same samples but with the output given by the network"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA2UAAAByCAYAAADTTXntAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAACgQ0IZAABAh4QyAACADgllAAAAHfoHZkxu6ZZA13QAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f07e4dfa110>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "temp = np.transpose(np.argmax(prediction_test,axis=2))\n",
    "plt.rcParams['figure.figsize'] = (15.0, 15.0)\n",
    "f, ax = plt.subplots(nrows=1, ncols=10)\n",
    "\n",
    "for i, j in enumerate(im_samples):\n",
    "    ax[i].axis('off')\n",
    "    label = str(labels_test[j,0]) + \" -- \"\n",
    "    for digit in temp[j,:]:\n",
    "        if digit==10:\n",
    "            break\n",
    "        else:\n",
    "            label += str(digit)\n",
    "    ax[i].set_title(label, loc='center')\n",
    "    ax[i].imshow(test_dataset[j,:,:,0],cmap=\"gray\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}