stacked_model.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# The stacked model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import os\n",
    "import re\n",
    "\n",
    "import matplotlib\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')\n",
    "\n",
    "from sklearn import preprocessing\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "from sklearn.model_selection import KFold, GroupKFold, train_test_split\n",
    "from sklearn.model_selection import RandomizedSearchCV\n",
    "from sklearn.utils import resample\n",
    "from sklearn.pipeline import make_pipeline\n",
    "from sklearn.metrics import mean_squared_error\n",
    "\n",
    "from math import sqrt\n",
    "\n",
    "from sklearn.ensemble import AdaBoostRegressor, RandomForestRegressor\n",
    "from sklearn.tree import DecisionTreeRegressor\n",
    "from sklearn.preprocessing import PolynomialFeatures\n",
    "from sklearn.linear_model import LinearRegression\n",
    "from lightgbm import LGBMRegressor\n",
    "from catboost import CatBoostRegressor\n",
    "import xgboost as xgb\n",
    "\n",
    "import pickle\n",
    "\n",
    "matplotlib.rcParams['axes.labelsize'] = 14\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "__Notice:__ In `sklearn` Polynomial Regression is defined indirectly.\n",
    "First, you have to obtain polynomial features to linearaze data and then fit with vanilla `LinearRegression`\n",
    "```python\n",
    "poly_reg = PolynomialFeatures(degree=3)\n",
    "X_poly = poly_reg.fit_transform(X_train)\n",
    "pol_reg = LinearRegression()\n",
    "pol_reg.fit(X_poly, y_train)\n",
    "```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Import the datasets\n",
    "data1 = pd.read_csv('./datasets/Data_training.csv', index_col=0)\n",
    "data1 = data1.apply(lambda x: x.str.replace('\\\\[|\\\\]', '', regex = True))\n",
    "data2 = pd.read_csv('./datasets/Data_training_2.csv', index_col=0)\n",
    "data3 = pd.read_excel('./datasets/data_2.xlsx', index_col=0)\n",
    "\n",
    "data = pd.concat([data1, data2, data3])\n",
    "data.columns = ['Tg', 'Tk', 'n', 'damage']\n",
    "data = data.astype('float16')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 307,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<seaborn.axisgrid.PairGrid at 0x7f520c4b71d0>"
      ]
     },
     "execution_count": 307,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 20 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Basic EDA\n",
    "sns.pairplot(data)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Since most of the `damage` values are near 0 (usual working conditions), but we want a good accuracy for extreme values too, lets split the dataset into 2 groups: \n",
    "__working__ and __extreme conditions__, and will sample them equally with `resample` (bootstraping by groups)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Define split variable\n",
    "split_val = 0.1\n",
    "data['groups'] = [1 if f > split_val else 0 for f in data.damage]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Bootstraping by groups (equal sampling)\n",
    "n_samples = data.shape[0] * 3\n",
    "data_resampled = resample(data, n_samples = n_samples, stratify = data.groups)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 329,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Normalization\n",
    "scaler = MinMaxScaler()\n",
    "scaler.fit(data_resampled)\n",
    "\n",
    "pickle.dump(scaler, open('scaler.pkl', 'wb')) # save fitted scaler\n",
    "# scaler = pickle.load(open('scaler.pkl', 'rb')) # load it back\n",
    "\n",
    "data_scaled = scaler.transform(data_resampled)\n",
    "data_scaled = pd.DataFrame(data_scaled, columns = data.columns)\n",
    "data_scaled['groups'] = data_scaled['groups'].astype('int')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 330,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<seaborn.axisgrid.PairGrid at 0x7f520c23e0f0>"
      ]
     },
     "execution_count": 330,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA3UAAAN8CAYAAAD/AYPoAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvnQurowAAIABJREFUeJzs3Xt8ldWd6P/Pep59zQUSQvBCUFFRGym3KKKdU5lyprUV9XUGvBIraA2g1ZnW60xL2ymnZ1T02GorF9uCCioWZ35a1OpvqLYzY6kaEKqxgIiWqIUQEpLs7Gc/t3X+2Nmb7GQHIZed2/f9euWl2dkkKzvfvZ61nrXW96u01gghhBBCCCGEGJyM/m6AEEIIIYQQQojuk0mdEEIIIYQQQgxiMqkTQgghhBBCiEFMJnVCCCGEEEIIMYjJpE4IIYQQQgghBjGZ1AkhhBBCCCHEICaTOiGEEEIIIYQYxGRSJ4QQQgghhBCDmEzqhBBCCCGEEGIQG/STuosuukgD8iEfR/PR7yRe5eMYPvqdxKt8HMNHv5N4lY9j+Oh3Eq/ycQwfR2XQT+oOHDjQ300Q4qhJvIrBROJVDCYSr2IwkXgVvW3QT+qEEEIIIYQQYjiTSZ0QQgghhBBCDGIyqRNCCCGEEEKIQUwmdUIIIYQQQggxiMmkTgghhBBCCCEGsUCufpBS6pfAbGC/1npilq8r4CfA14BWYL7Wekuu2ieEEH0tkXA50Grj+pqAoRidFyIczlk3LMRRk1gVg4nEqxhM+ipecxnxa4CfAo938fWvAhPaPs4Dlrf9d8A75e4XjvnffHjPxX3QksOOtU193R4hhrtEwmXngRiL11ZT2xCnrDjK8soKzhidL4MPMaBIrIrBROJVDCZ9Ga85i3at9e+VUqcc4SmXAY9rrTWwWSlVpJQ6QWv9aU4aKIQQfcSyXOrjNiMiAZ6umsHru+q489/eYfHaatZXzWCsDDzEABL3XIrzgqz7xnmYhuL1XXUSq6LHUv1ganWiJBoiEDDY35LA8XyCpsGYgjCBwLGdDDrQaqcHyAC1DXGJVzFgHWi1+bCuiaerZuD5GtNQbP2onlF5wR7H60CK9rHA3naf17Y91mlSp5SqAqoATjrppJw0Tojukngd3izLZVd957ty9/39RO78t3dwfd3fTcwg8Tq8WZbL3oZEp3i99W9PG3CxChKv/SHb1jHTzJycjYoEO03gsvWDx40IccXKP6QfW1FZwVnHFR7TxM71dXpCl1LbEJd4FQNSXsjglNIRXLVqc8Z7IS/U8zQnAylRisryWNZ3pNZ6ldb6HK31OaWlpX3cLCF6RuJ1+Erdmc52F/mCCaWUFUcJGNm6vv4j8Tp8fVa8DrRYBYnXvpBIuHzc0MpH9TH2HYrz10NxPqqPUdvQSmOrxc4DMa5ctZkLl73Glas2s/NAjE+a4lyx8g9cuOw1XnnnE3bVZz6nq7iyXZ3x2KK11exvSRxTewOGoqw4mvHYQOxbQeJVQKvtZ30vtNp+j7/3QFqpqwXGtfu8DPikn9oiPoOc2RPiyBIJl8aE0+VdZM/XLK+sYHReqJ9aKMRhluVyIG7jHSFejy8I91PrRK60P+9TWhDmzovO5I4N2zNW0h7etLPTgHTNgunpx75UfgLXPLo54zlH6gc7PuZ6xza4HZ0XYnllRadVQOlbxUDUlyvLA2lS9zzwTaXU0yQTpByS83RCiMEq5rp4bVuPyoqjGZ146i6yHOQXA4HjeBy0nM+MV4nVwcVxPPa3JFAKtAZPa0ylGFMQJhg0s/6b9ufTlswuT0/o4PBK2pLZ5bxSsz/9b2ob4rRfFPN150Gr2UVcmR1W08qKowTMY9tEFg4HOGN0PuurZkj2SzHgHamP7fH37vF3OEpKqaeAmcBopVQt8H0gCKC1XgG8SLKcwfskSxosyFXbBoPuZNgcSN9fiOHCslwOxm0Sns+HB1pxPS/rXeSCiCGDDtHvLMtlXyxBXXOC+phNNEDWeC2JyqpHf0lNzlITlqCpcDxNaX6IUCh7H+I4Hn/e38LDm3Zy3QXjuevZ7Rl/z7PGFGSd2LVfRSiKBrOuKJTkZ8ZCWXGU9osMhuo8aN36UX3WuAoFDj83tRI4phsrwuFwQJKiiEFhZNTI+l4YGe35ibhcZr+8+jO+roGbc9QcIYTodZblsq81ge9rDKU4fUwBhoK9B2PpTFcBQxEJGURNGYCI/mVZLg0JB0MpSgvDlBSESLge+w/FM1Y9iqIGkYjEay64rp+RcKQkGuQvh+LUHowzuiBENBTA9RVaw/5YAqPVYXResNPkbn9LMtnNktnl6QkddMgMWZzX6ee3X0VojDtZVxRKC8MZE7GOk7Pf1nzaadB6SukIJpRkrqalsl8+s/B8XM8n0M3sl0IMJq02nFgUyngvREKKVhsKIj373tJLCyFEL7Asl1bPpTnusqjdYGbZ3EmUFIS55cmt/PiqKWgg3wx0eYddiFywLJfapjh1zYmMM1PL501jZF6Y2oY4x4+MUBAxiIbkLF1fSSWnab9F0nY1q363m9c/qOfZRedzoDnBU298xHUXjGfxui3pv9W9cybx2Ot7uGXWGZxZmp/Rp6RW3Lpabevq/E7782krXtvNsrmTMuJj2dxJhAKq01ZH08ycnI2KBDOek1pVHJvl5sCJRdEsLRFiaAqY8HGj3WmlbmxRz/tZGVUIIUQPWZbL3qY40aCZntBBcvB0x4btLL1sIrfOmkDAOPJ5FiFyIZUUZe/BOEueeydzFWfdFp66cQbv728hYCgKw2GMAZhFcChIlTvZ+HYtX5s0lpufPDxhe2TeNAAcX3PHhu1ZV9zuejb5eHrlrd2kLrXi1tVqW1fndzqeT4sEDDYsOh/b9TEMRTRkUBzNHhMdJ2fZJnBCDHfxLrJfrq+aAfk9+97yjhNCiB6w7eSd9mjQxHb9rHfF80ImY0ZEKImGZEIn+l3cSybxyQuZWePV15qyUVHGFMiErids26UudrhWWzRkUBA8vEpfH7d5eNNO7v7q5/j6L9/IGOTdtG4LT954XjobaVcrbqnHO668jSkIs7wtU+W9cyZ1OlN3pHNrcj5NiL4zXLJfCiHEoGLbLjvqkum/n7rxPAKmkfWueKvtEQnIuSTR/yzLpdX2MZSi1fa6XMU5cUSe3IDogfZ9Q21DnIX/4xSuvWA8zVaCgGFTEg3h+po5FeM4GLO7SPcP4cCRV9xSj3dceQsGTc4aU8D3LzkbpWB91Qx8nTzrK7sFhOg/QyL7pRBCDCWu69OUcCkIm6z7xnmAojnu8OMrp/CP69/OOINSWhimJC/Y300Ww5xluRy0bGzPx/fhtDH5/Oyaqdz85NaM7IOyonx0XNfnYKvdltQkmZEytSoXDhjpCd0VFWVcPHksV63anLFaNq44TEl+iPqYnXWQ9+GBGJ8fO4IVlRU8lGXFLXWmbnllBaX5nbOTBoNm1mQoQoj+kx82WLPgXPYeTO7iabU9xo2Kkh8eRNkvhRBiqLAsl7jv8temREZSlEfmTaOkIMiPr5xCcV6IcNDAVKBBEqOIfpWa0NW32BnJNpZXVrBh0fkkXD+dkVBWlLtmWS4tjovt+qCgIebw/Nu1XDx5LDe1e11XVFZwZUUZD/zHLm784qksWPNm1jM0pYVhVv1+d6cJ28+umcYPnn+XH181hTNK87OuuBkKvn/J2UcsbSCEGFg8HyzHT59nTvUXnt/z7y15Y4UQ4hhYlkvcc2lN+J2Soty0bgtaKxzPZ9b//R2ur/n+8++ie75VXohusyyX3fUxdvy1JT2hg8OTi9RZDpnQHVmqBERj3GHHvhZabY/9zQmunH5yekIHh4t0Xzp1LFPHFWEaqsszNMflh/mHWWfw2Ot7WDK7nA2Lzufx66ezbvNH1LUkCBgqmTWyOI8Ti/IYW5zHuFH5jC3O44S2z2VCJ8TgkXA6jx0Wra0m4fR8Vic9QT+RYt9CDE5xzyXhamwve1KUgzGb40dEKCuOEjQUt8w6Qwo3i36TKi6+cG01D1w+uYuzW8nJhUzoOmtfcqAh5uBpTSzhZtxlf+KG6Vlf17rmBLfOmoDn6y7P0EQiASaU5vMvl07Ecn0+PBDjRy+8R11LQoq+CzEEOX2YKEVW6oQQ4ihZlkuT5eF6mg8PtFJWnJnCu6w4Sn3MxtewvLKC/LDBhJJ8GSyLfpGa0NU1JzKKSbdXVhwlaEoSn2xSJQeuXLWZP33cxMK11RTnhdJ12yA5GDtSX3BqaT4j8wIsr6xIPye17TU1YQuFApxQFOWEgjBnHFfAj6+awvqqGdJ3CDEEmW2JUtorK472SqZhmdQJIcRRSN2xB0DBS3/6lJ9dMy1joHbvnEk8W72XaNCgrChMUV5EBmWiX1iWS8xxMZSiJD+UPLfxWvLsVvuYXVFZwaiIJPFJcV2fmGXxcUMr+2IJiqJBrqwo48SREZbMLsfXne+yP7RpFys7TNpSfYFSirjtM6EkWfvtd3fM7HLCFokkt1meXJLcXil9hxBDT8BQLJub2Q8vmztJsl8KIURfsyyXhHb5y8FEOptdqhP+/Y79PH79dA7GbOpjNo+9vod/mHUGr7z7KX939gkU97CQqBDdYVkuTY7DviabxWurKS0Is2zuJO7YsJ37X97B0ssmcsrofMIBg1GRoEwe2qTOy9Y2Zr7Xl1dWMKYwyOJ1W3j8+umdtlLWtSQ4bkSYdd84j7rmRLovuGXWGfy2JtkXRCIBKcYthCBoKkoKQiy9bGI6+2VJQYigKSt1QgjRZyzL5WDCpsXy04M8gNKCMJbj89VJJ7KvycL1NRPGFHD935yKaUDF+NFyFkb0C8tyafVcHFeTcDyWzC4H4L7fJCdz918+ObklMGrKhK4dy3LZ32rTand+rx9oTmA5mnv+/vO8sO2TTqtyj8ybxqO/301eyOSEkRHOOr6Q784+m+o9B6QvEEJkSDg+z761l7LiKKWFYcqKozz71l5JlCKEEH0ltd0ydXg5NcibOq6I279yZqf04+vf+IhLppQxKi/EiLAMlkXuJRIuHi6fNiYTo7SvZ3b/yztYsOZNfnfHTEZETKJmQGKUzPe542mCpjrie/3eOZN49+NG1t5wHgdakqtyP/3tLm6ZdQabd9fxxTNKaUn4uL7m784+QTKKCiEyKfjimcely5yk+hV6vlAnK3VCCNFR+wQJFy57DcfT6Tvzi2aelh7kQXKyd/OTW6g8fzxjCkIU58kgTuReIuHiapcDMS89oYNkfN717HYWzTwtnXExPygTOuj8Pp+/+o2MJAbZ3ut3Pbudkflh/s+LNRTlhTjr+EK+d8nZfFjXxCmlI8gLBuVcnBCiS1qTtV/pjdJHMqkTQogO6uN2egvW1HFFNFs2K9q2XBVFg9nTwmvNiHCQYNDsp1aL4czVLo1xH6+LdNkl+aF0xkWpa5aUep+XFoRZeW0FD1w+mYOxz36vl+SHWPCF8Wg0BWETz9dMO7mEs8YUyPtfCHFEXpZkS6kxRE9Jzy6EEO1Ylovray44tYRFM09DKTjQbAMeT9wwnaBpZK05FZa08KKfWJZLfauH7WrCgezxecLIiGwFbMeyXIKm4vHrp6ff4wnXw3QVJxZFWHvDeelVu2yvZcBQjGzbZl1S0I+/iBBiUDFV9n7FVJIoRQgheo1luRxoKzR885dO556X3uPb67fh+j63PPU2f3v/7/jhr99l+bzMRAkrKysYnS/JEETuWZZLU8KhKe4yf/Ub3PrU1k7psle2rdDJhC4pkUhmB407PgdjNn89ZAGau//tT8xd8QeuWLmZJsvh4U27OpWAWHltBWMKwhw3MiqvpxDimOWFjKx1K/NCPZ+S5bRHUkpdBPwEMIGfa63v6fD1k4DHgKK259yttX4xl20UQgxPqfM1izskmNBas/q/97BkdjlF0SCNcYeN22pZPf9cDsUdSgvDHJcflm1XIucsy2V3fYxW26Ml4fLA5ZNpjDv8+5aPWXrZRE4tTdbUOC4/LBOQdlocl7omO3328FcLz894/Va8tpub1m1JZgttVwIiGjQoyQvJe10I0W1x22fj28kxhGkoPF+z4a2/cN0F43tcBilnvbxSygR+BvwdUAu8qZR6Xmtd0+5p3wWe0VovV0qVAy8Cp+SqjUKI4avJdjnQnMgY2N317HaevPE8rrtgfKcMePlhk7yQSWFUtl2K3LMslybbpdX2iIZMvvXM252yXf74qimMzjclPjvwfNjf9l73tSbSxet3amk+D1wxmVDAYLSsdAoheoOCL33u+Izslw9cPnnQZb+cDryvtf5Aa20DTwOXdXiOBka0/f9I4JMctk8IMUxZlsv+5gRLnnuHK1dtZunGGm7/ypmUFoRRqKyZqnwNvtaE5GiyyDHLcrF1MmZbEi6LsmS7vHXWBAKGwpT4zNDxvW45mXXp2r9+pqEIy4ROCNGbNNz2q20Zfc5tv9qWnAH1UC4ndWOBve0+r217rL0fAJVKqVqSq3S35KZpQojhKJFwiVkW9XG7y4Gx7fnZM1X5GkMpGeyJnLIsl71NcZotn0Vrq8kLmVnj85TR+eSFZBW5o47v9SO9fkFDyVlEIUSvcrrIUJyqidsTuZzUZVtY7PgbXA2s0VqXAV8DnlBKdWqjUqpKKfWWUuqturq6PmiqEL1H4nVgsiyXRsthT32iy0721NL8dDbB9lL1vkqiQy85isTrwGVZLgfjNgXhQLp0QWPcyRqf0aBB1Bz6k5FjiddUZtv27/WuXr9w4HB2SyF6i/Svon0tzJSy4iiGMbiyX9YC49p9Xkbn7ZU3AM8AaK3/AESA0R2/kdZ6ldb6HK31OaWlpX3UXCF6h8TrwJPKcmm5PnXNCQIdOtmp44pYPf9cnLZVutXzz8nIVLViCGcTlHgdmCzLpbYpzl8a4ig0hqHYsOh88kMmP71maqcMjUXDZEJyLPHaYDnpAdXUcUXJTJaFYR6/fjpfLh8DHH5/j4oMzfe36F/Sv4qQafCza6ayev65rK+awer55/Kza6YSMgdX9ss3gQlKqfHAx8BVwDUdnvMXYBawRin1OZKTOrmVIYToNZblsi+WoK45QX3M5tnqvdx50VmsqKxgUVsh4jsvOpM7NhxOjPLgFZN5dtH5WK6PrzUjwgEZ8ImcsnyXoGlQGAlwoMVJbyFMHbJ/8IopjC4MEw0aFEeChMMSnx2FgorWhObpqvM4GHO4ad2W9Gv4yLxp/MulZ2N7moKQJJcRQvQN0wDTMFjy3Dvp/mdFZQW9MKfL3Uqd1toFvgm8DLxHMsvlu0qpHyqlLm172m3AjUqpbcBTwHyte6HEuhBCALadTAE/7+d/ZO6KP7B0Yw3XXTCe+37zZ0IBxdM3zuChq6emJ3SQ3Ib5rWe24XiaeT//Iy0Jj0IZMIscsiyX2oYE837+R2ob4p3Of972q20U5QWJBBVFYZnQZWNZLp82Jrj60c28+0lzekIHydfwpnVb8DQ0W668v4UQfcZy/E59+KK11ViO3+PvndOeq63m3IsdHvteu/+vAb6QyzYJIYYHx/Goix2uTQWHE6IsmV2OqQw+rI9x/Mho9sQoWrO+agal+SFCIRn0idywLDcjuUdRNJg1Pg/FHfJlBblLR/Ma+r5mTGFY3t9CiD7T8VwvHE6+1lPScwkhhrxEwqX2UBxDqXRnOnVcEYtmnkZRNMiYEWHywwGioXx8nTxX077TLSuOYrYlRpEBn8gVy3I5aNm4vubHV06htDCMoVTW+CwtDDMqEuzH1g5srq8pLQhzz99/nhOLollfw6BpMELe30KIPpQ619ux/xlsiVKEECLnHMcj4btYjs+H9a3pJAm3f+VMlm6s4cpVm7n2F2+wvyl5zu6Hv36Xe+dM6pR4QmpViVyyLJcPDsa4YuVmLlz2Gv+4/m0+aYzz8KZdLJvbIT4rKzguPyzxeQTRgME/f+0s7v63P3HrU1s7v4bXVlAcGR7JZYQQ/acwbLC8siKj/1leWUFheHAlShFCiJzyfU2T7dBieSxsS4Ly02umUhAOMn/1G+k7ZaUFYeqaE4wblcecinE8t/VjlswupyQ/RFFeiMKwJE4QuZPacln1RDJml8wupygapNX2+F/TxnLfb3aw9LKJnFqaT8BQjM4LyTm6z+Bq+NYz2ygtCLNo5mkUhAOsWTAdQ8FH9a2MzpebNkKIvteS8Nn4di2r55+LaSg8X7Phrb/w9QvGMzKvZ99berAsTrn7hf5ughCiFzTGE1i2j+drlswuZ1PNPhKOj+vZGdswb//Kmdz17OFsl/fOmcT9L+9g695GXr19JsVR2dYmciORSK7QOV5yu2DH2Fw+bxoAC9a8ye/umElJVCZ0R8P1/C5fz4c27eInV03p7yYKIYYBpeCLZx7HgjVvZow5VM93X8r2SyHE0GNZLvXNFp82Jrhy1Wa+9MDvWLqxhnkzTuYX//UB9TE7vfVh0czT0oM8OJw8ZdHM05JFiE0l5+hEzjRYDj/+j50URgLcOmtCp9hcvG5LOjYDhpLVpaMUMFSXr+etsyZg9sJ5FiGE+Cxak3XM0Ru5/mVSJ4QYUlJFmg9ZbqdMlzc/uYU5FeNY8dru9Lm5rjLhleSHkkWIo6H++DXEMOO6Pn89FCfh+sypGMc9L73HSSV5Xcbm8soKSiQ2j4pluRiG6vL1PGV0vryWQoic8Pow+6VM6oQQQ4ZtuzTZLkbbPoauBsRb9zZy/8s7WDK7nDEjwulVu5Sy4ignjIxwekm+rISIPue6Pjv3NzN3xR/481+bKckP8UrNfj5tjHcZmxMkNo9afdwmbntdvp4hU1Y8hRC5YbRlv2xPsl8KIUQHju9S15Lg6798g137W7J2nKPyQ5QVR9m6t5GlG2toijs8cPnkjExUKyorGJ0niRNEbjTEbfY1JXjg8snkh0xGFyRvNDzwys5OmVhXtK3QSWwePaUgFDB4/A8fZn09QwEZCgkhciNkqE7Zd5fNnUSoFyZ1clUQQgwJluXSGPeJBk1Wzz+XrR8dZHllBYvbtmAmkyJUYBrwdNUM6ltsmi0HgF/81wfpbJelhWGOyw9L8gnR5xzHI+Y47G9KsOS5d9Jxunr+OSyfN43F67Zw/8vJTJcnl+QRNA1K5WbDMTOVotl2WfCF8az+7z3p9/rogjCRoMFIea8LIXJFweiCEEsvm0heyKTV9hhdEIJeONYrPZkQYtBLJFx21ccyJnDrvnEeP3qhJp0OvjHu8PBvdzKnYhxLN9Zw75xJPPDKTgBunTWBk0blEQkalOZLgXHR9xzHo8GycT06nf1csOYt1lfNYH3VDGzPJ2QafNJocdKoqEzousHTmvmr30yXMyiKBmm2XE4YGcEAeb8LIXLG9WHZyzuYUzGOPExsz2fZyzv4waUTe/y9pScTQgxqluVyMG6nJ3SQHBjXNSd4pWY/r9Tsz3j+4pmnp7NNLZldzsInqlmw5k1evT2ZHl4GeKKvWZbLQcvB1z6en/3sp+X4hIMG316/jQevmkIkZDIyLKU1usPxkokJahviLHyiOv34pm9fCEHZeimEyB3H87OOTb57sd/j7y29mRBi0EoOjm0Snt9pYNy+bEFKWXGUkdEgU8cVUdsQp6it/lxZcZRI0JBVENHnLCu5qvyD59/ho/pkzGaL04CpqGtOUNeSIGgoSYzSA4EuEhMETNUrGeeEEOJoddUf9UZZFZnUCSEGJcty2RdL4HiaDw+0pjvJqeOKWHvDdD53QiErKysyDiPfO2cS97z0XrrOV2Pcoaw4ysprk4lRhOhrqVXlBV8Yj2kogqbikXnTOh2aVwoaW5106QKZ0HVfwFA8eEVmMqQHr5iMUhA0ZRgkhMidrvqjgCRKEUIMR5bl0mQ7mErhac0po/P4+XUV/Ht1LVdNP5n6mM17nzaz5cN6Hr9+OofiDvubE9z/8g627m2k6ounsbKyguL8IBsWnU9JXohg0OzvX0sMcZblYhqKp6tm4PmahpjNzeu2UloYyojT+36zgwevmkL5CYWMDAdlQtdDvtacUBRJv+7JO+IaQykKIlJ0XAiROz7Z+yOfnu8akCuFEGJQsSyX/a0JGlsdblq3JSNj4KVTyrj2l2+kH0utzM2pGJc+S5Oq8xUwlAyYRc5YlsvepjgHmhPcsWF7OkZ/fOUUfvTCe3z9l2+kz3gma6cZEp+9JBw0+LgxkZkJt7KCsUVhmuI+hZH+bqEQYrgIB7ruj3pK9h0IIQYNy3Kpj9vYrk5P6IC2JAhWpyyCdz27nTkV4yjJT26tTNWl8rSWAbPICcty+bihlfq4Te3BeHpCB8kY/cf1b7No5mnpM56pGB0VkfjsLXHb75RIafHaauK2T0lUtl0LIXLnSP1RT8kVQ+TEKXe/cEzP//Cei/uoJWKwsiyX9+tjeL6mMBLolBglL2RmzSJYkh+iKC/Ef3z7i+w9GKe0ICQTOpETqaQoi9dWs/Yb53UZo6nJ3JgRYdYsmM6IsCnx2YtcX2d93V1fy+sshMipI/VHPZXTlTql1EVKqR1KqfeVUnd38ZwrlFI1Sql3lVJP5rJ9QoiBJ5FwiVkWB+I2luNRnB/C150zBrbaXtaMUqPyQ7QkHO77zZ8pLQwzSoo3ixypj9s8vGknS2aXEzRUlzHaanssmzuJprgDaIolaU+v6jL7ZS8kJhBCiGPRl/1RziZ1SikT+BnwVaAcuFopVd7hOROAfwK+oLU+G/jHXLVPCDHwWJZLg+Wwv8Vl174WfvTCe/zvje8SNOmUMXBscYQfXzkl47EVlRVEQwYl+SGWzD6bCaPzJSGK6HOO49FsWSgF/3xxOWcdXwhKM7Y4wrK5kzJidOW1FZx1QiGnjM7nhKIoJxfnSYz2ssKIwfIOmXCXV1ZQGJETKEKI3OrL/iiXt6unA+9rrT8AUEo9DVwG1LR7zo3Az7TWDQBa6/2dvosQYliwLJfdB2MsfKI6I/HJc1s/prHVIWAYPHXjDHytcTzNqt/tZtf+FpbMLqckP8TxIyIYBiQcTTRoUJIfJhCQQZzoW47j0WQ7fNqYYFG7g/DL5k7irT0HuXTqWJ66cQae1oRNA9DEbY+RkQAlkrGjT1i2piB8uL8wlML1PSxbMyL62f9eCCF6S1/2R7mc1I0F9rYBHeQlAAAgAElEQVT7vBY4r8NzzgBQSv03YAI/0Fr/JjfNE0IMJA0JJz2hg8OJT1bPP5cFa95MP76+agZXrtqc/nepLJfrq2ZgGopw0OT4gnyZ0ImcsD0Hy/bTEzpIxu4dG7azZHY5Fy57Lf3cX9/yNyQcj3DQ5ESZ0PUZ29dc+4s3M86xlBVHebpqRj+2SggxHPVlf5TLSV22zaIdTwUGgAnATKAM+E+l1EStdWPGN1KqCqgCOOmkk3q/pUL0IonXY2dZLpbjZT1MHAoYPHD5ZBrjDite250uIN6xg2y1PU4fU0A0aMgZumMg8dp9luXyYX2CvHDXCVFSyoqjHGhOcMZxBVJcvAeOJl69LhITeL2QmECIYyH9q+jL/iiXV5FaYFy7z8uAT7I8Z7PW2gH2KKV2kJzkvdn+SVrrVcAqgHPOOUd6ZTGgSbwePctyOZRwcH3NhwdaMyZrU8cVceusCTiepjHu8Gz1Xm7/ypk8t/Vjls2dlFH765F508gLmSRcj1JJOnFMJF67x7JcDlo2AEHTYMOi86mP2ax4bTdb9zambzQA6e2YpYVhmdD10NHEa8g0+HL5GOZUjKMoGkz3HyFTVu9Fbkn/KvqyP8rlleRNYIJSajzwMXAVcE2H5/x/wNXAGqXUaJLbMT/IYRuFEP3Eslxqm+LUNSc4sSjKQ5t28dBVU7n16a2UFoS586IzMyZu986ZxGOv7+Hq6ScTDZnc8/efJ2gaNMYdRheEiAYNomZABsyiz1mWS5PtAMntJ1et2twpTm/50gRKCkL8+pa/oTgvSChgMCIo8ZkLoYDili9NYHFbbcuy4ijL500jFJDsl2Jgcl2f/S0JHM8naBqMKZAz4UNFX/ZHOYsQrbULfBN4GXgPeEZr/a5S6odKqUvbnvYyUK+UqgFeBe7QWtfnqo1CiP5zMG6zt604s2koSgtD5IdNll42kf97xeRORZtThcVPGpXHvzxfQ+Uv3uDKVZtZurEGUBTnR2TALPqcZbm0OC51zTY7/trS6SzdXc9u546vnEVRXpCb121l8dpqDGBMocRnrsQdPz2AgrZiv+u2EHd6XuxXiN7muj6HEjZaJxfytNYcSti4rsTrUNCX/VFOryha6xeBFzs89r12/6+Bb7d9CCGGCcfxCJiK08cU8NSN5xEJGHzn4nLm/fyP1DbEWV81o8vC4kpBXUsCOFzGYFQkmO3HCNGrUqvLQdNg4dpqHrh8ctY4PRR3+OCATV1LguWVFYyKypbgXOrqDIsvZ+rEANTq2F0+PiIgCZUGu77sj+Q2oRCi31iWS33cxtMareFHL9Sw4AvjKSvOw9eHO76ukqGUFoaJ2y5PV83A8zVB02BUJCgrIKLPpbZchgNmOlaPFKcnFkVYXzVDztD1g1Sx345/F1OKj4sBqiXh4bgaQ4HjaVytKY5K/cqhoC/7o6PafqmU+mIXH/9DKXWuUmpUj1sihBhWUnXorly1mS/e9xrzfv5H7rzoLEZEg1z96GZ27mtJF+dc8dpu7p3ToWhzZQWhgOIbj1ejgJNL8jmxKCoDZtHnLMsl7rnsa7a5+tHN/PmvzclV4i7iNBxQaA2l+TKh6w9B0+DBKyZn/F0evGIyQUmUIgYg24WmVpf5q9/gSw/8jvmr36Cp1cV2+7tlojcEDJW1Pwr0wqTuaK8ur3G4/EDqp7b/3FdKPQ9cq7WO9bhVQoghzbJc9rfanerQ7T0YZ8lz71DbEE8PkO96djtb9zby2Ot7ePz66RyKOxRFg4SDBgdjDssrKyiR7WwiR1JbLsMBk8Vt5+fax+r9L+9g6WUTOakkj6ChCJgK14fR0RChkEzo+oPWmlDAYOllE8kLmbTaHqGAkT6zJMRAknB9Fq+r7nDmqpr1UldxSPC0JtihPwoGDLxe6I+O9jbVxSSTm1QCp7d9VALvAnPaPqYA9/S4RUKIIS215TJgwJLZ5ayvmsHKayuYOq6IvNDh+l5b9zZy/8s7WDK7nN/ediFzKsZx2zPbuOWprXxY34qvoSgaZEJJvqx+iJxIxW5qhae0IAxkxuqPr5qC7fnc+9J77K6L4flgKiUx2o8s12fd5r+kt8KWFUdZt/kvWJJ4QgxArq8pLQiz8tqK9PWxtCCMK2dAhwYNj7z6PraX7H9sz+eRV9/vXLm7G472KvO/gX/QWm9q99gHSqk64F6tdYVSygMeBm7pebOEEEORZbnsqo+x8e1aLp1SxtKNNRmp3wvCAVbPP5e8kJkuLr50Yw1LL5vIwieq0/W98kImIUMxMhIkHJbBsuh7qdhdvLaa0oIwt86awP1XTObTxjgPvLKTrXsbWbqxhtXzz2Xpxpp0HTrTRBL39LNIwGDejJOobYin74zPm3ESEUkRLwagSMDgB5eWczCWLJMSMpOfS7wODeGAkbWkQbgX/r5HOxoqJ1lbrqOP274G8Cfg+B63SAgxJCXTNDscaE5Qef54rn40Wctr6rgiFs08jaK8IGjS2y9TE7jRhWEKwiabvn0hhoIDLTb54QAjwjKhE7lhWW46dpfPm4av4eYnt2TckHjs9T3cMusMCiImT9wwnXDAoDBiEFIBidN+ppTCdv2MvuXBKyajlCRKEQOPxOvQZns+kaDBmgXTMRT4GrT2sf2e7xw42mlhDfAdpVQ49UDb//9z29cAxgF/7XGLhBBDjm271McSNCeSJ70VmiWzy/n3my7ggSsm82z1Xmob4izsUOPrjg3bOdTq8P3n3sXXmvqYzYlFEcpGSEIU0fcSCZfGVouDcZtWxwOgJeGmJ3RwuBbd9y85m8KISWvCJ2QajIqEKIxEZEI3ADiez6P/+UF6u/eS2eU8+p8f4Hiy/VIMPBKvQ5uvYdnLO9hd10Jdc4LddS0se3kHvTCnO+qVupuAXwMfK6XeIbnz8/OAD8xue86pwCM9b5IQYiixLJcDcRvP1xhKsXl3HaMLQhlbLx+4fHLGebqU2oY4nq95pWY/VV88jUjQlJTwIicSCZdDtkNd8+GEPqltMqUF4YxYrW2Ic7DV4ZKH/4tXb58pMTrAKAXXXTCeu57dnrG6KhUNxEAk8Tq0GUYXf99e2F17VN9Ca/1HYDzwHWAL8DbJVbrxWus32p7zuNZ6Wc+bJIQYKizLZW9TnF37WvjrIYuP6lupPH98ei85JAfEt/1qG3nhQDrFb0pZcTRd+6soL8SYwrAMlkVOxD0Xx9WdMrQuXreFW2dNyHhuWXGUTxqTF+dwwJAYHWC0Jj2AgsOrq5J3QgxEEq9Dm+/DY6/vyViJfez1PX2/UqeU+iXJBCnNbaUKVvb8RwohhgPLcmlsO4fU/mzAum+cl3VFTkE6LXzHs0rL5k5CoymOSsIJ0fcsy6XV9nF9nTVWTxmdly4e2z5Ol1dWSFKUAcjr4u/oyShZDEASr0PbmIIwt8w6I10Sp6w4yvLKCsYUhD/7H3+Gz7qdeB1wN9Dc458khBg2bNvlYNzGB+7YcPiOYyoFfGpAnFJWHEVBuhZdwFAYhsJQ8L1LziZoKEblhQgGzX74bcRwkipbYBoK0Flj1VAqHaemoVAKvn/J2bLtcoAyDZX172jKfjYxAEm8Dm3BoMlZYwpYXzUD19cEDMWYgnCvjG8+a/ulRJAQ4phYlsunzQk+OWThepl3HBfNPI0fvVDDvXMmpbdalhVHeWTeNCIhM12LDpJ3Jk2lKCvO47iRUZnQiT6X2i68c18Lrq+7jNVQwEjHaV1zAlMpxhbnyYRugAoYimVzM/+Oy+ZOIiCDZDEASbwOfcGgydjiPE4uyWdscV6vjW+O5gok671CiKNiWS6762PpLJZPV83IuONYFA3ySs1+6pptlswupygapDHuMCo/yIOv7OSZ6lrKiqNYjkdBOMDIsGxlE7nRvmzBkufe4bEF07uI1RAPvrKTupYEluMRMA2J0wEu4frc95sdGX/H+36zg59cPaW/myZEJxKvoruOZlL318+qjaG1llvoQgxzlpXccrmwrTjzktnl5IdMHpk3jZvaEqO02h5lxVG27m1k4RPVQPIu5E+vnpqe0K26toKivBAjQgFZ+RA5YVkuda02vtZYjk9pQZhPD8Wzxuoj86bx+gf1EqeDSMBQ1LUk0n9HSP4tA1L3SwxAEq+iu47mSlQFNPZ1Q4QQg5fva2KOi+NrSgvC3P6VM9MJT75cPoZ13ziPxtbkitxPrprCPzz9dvqA8MprKyjJD/H7O2ZiGopwwGB0YaS/fyUxTFiWy+6DsYyyBffOmcRzWz/mgcsnc9uvtmXE6qi8IOurZkicDiIBU7F83rR01t1UaYqAKYNkMfBIvIruOppJ3a+11vv7vCVCiEHJslyaEg5O24HfW2dNSE/opo4rYk7FOOqaE4yMBlm6sYYfXHo266tm8Okhi/qYzU/+YycLvjCeEZEA4aDJ6LxQf/9KYpiwLJeDlsPCJw6vLhdFg7TaHv9r2ljueenPLL1sIqeW5vNBXQy05mDMljgdZBxPs3Hbx6yefy6mofB8zYa3/sLXLxjf300TohOJV9FdnzWpk/N0QogupRJLHGhOsPq/93Dbl8/k5JK89ISu/YpdagWk2XLTH587vpAls88mZCoUMDISJByWrWyi71mWS21THNNQnVaX00XGC5O1ES3H44zjCggYEqeDkaHgi2cex4I1b0oxZzHgSbyK7spp9kul1EVKqR1KqfeVUncf4XlzlVJaKXVOb/58IUTvsW2XFsclEjB57c/7+O7ss7Ecn3DAoKw4yqKZp2UtoBo0k91OaWGYzbsPcPWjm7FcXwbKImcsy8XyXaJBk6BpZKwuw+Ei43de9DmK8oI4nmbLR/USp4OUL8WcxSAi8Sq664hXJq31Z036jppSygR+BvwdUAu8qZR6Xmtd0+F5hcCtwB9762cLIXqXZbnEfZeY7REKKC6ePJZrHt2cPkP3y/nnYCgjawFVSB76/vBACw+9upuVlRUclx+WgbLICctyiXkunzQmWLy2mgtOLeGmL52eNVZNI1kzynYcTi0dIXE6SPk6ezFnX8soWQw8Eq+iu3J5dZoOvK+1/gBAKfU0cBlQ0+F5S4H7gNtz2DYhxFFyHA8bl1jCx/M1vq94YdvH6fNIjufjuD77m62sBVQ/qm/lrOMLObE4j2cWns+YgjCBQK/dPxLiiGztYtk++SGTx66fjuN57KmLZY3VoGkQNBXHF+VLnA5iyXqXWYo5SzZBMQBJvIruyuUVaiywt93ntW2PpSmlpgLjtNYbc9guIcQxaLEdPqpPcNWqzXzpgd/xL79+l4snj2XpxhquXLWZu//tTzRZLi/96dNOhZvvnTOJhzbtwvZ8xuSFObEoKgNlkROW5dJsWXx0MMGVbbF73S/fwHY179Q2Zo1VU4Ht+YyKBCVOB7FIyGB5ZUXG33d5ZQWRkPxNxcAj8Sq6K5crddluMaTXkpVSBvAgMP8zv5FSVSRLLXDSSSf1UvOE6BtDJV4TCZcDrTYAi9uKiwPMqRiXrkMHyW0id2zYzpLZ5dz/8uECqmMKw3z7mW3UtSQIGEpqew1QQyVe20ttF25N+BmxW9sQ56Z1W1g9/1zubIvZVKz+60vv8f1LzqYkEpJYHcCOJl7jts/Dm3ZmFHN+eNNOvn/J2ZCfy9aK4U7iVfSlXE77a4Fx7T4vAz5p93khMBF4TSn1ITADeD5bshSt9Sqt9Tla63NKS0v7sMlC9NxQiNdEwmXngRj//7uf4vo6nd1y5bUVTBhTwJLZ5UwdV5R+fm1DnKJoMF24+bZfbWPn/hbqWhIsr6ygJCrp4AeqoRCvHTm4NFseWsMDl09m5bUV6XhNnp1TGbH6YX0rt8w6g5KoTOgGuqOJV9fX1DXbGY/VNdu4knlC5JjEq+hLubxavQlMUEqNBz4GrgKuSX1Ra30IGJ36XCn1GnC71vqtHLZRCJFFzHUZGQ1w7vjRGErx5fIxXHfB+E7lCu5/eQdb9zZSVhyl1faA5NaRFZUVlOQnizbLQFnkkmW5xG1NY6uTXlFuH691LYl0VrlUrI4pDDEiFJQ4HSIiAYM7LzqTOzYc7q+WzZ1ERLbUigFI4lV0V84iRGvtAt8EXgbeA57RWr+rlPqhUurSXLVDCHH0fF8Tsyw+aUywe3+MhWuriSUc/vlr5VlTLi+aeRplxVEevGIyZ51QyG9vu5DV88/l12/XAoqxxXkyUBY5Y1ku+2IJbFd32iJ817PbuXXWBJZXVjAyGuC3t13ImgXTOWFEmKKITOiGEk+THiDD4S3inix8iAFI4lV0V06vWlrrF4EXOzz2vS6eOzMXbRJCZOf7mn3NcTwfLMdj3KhkNq6PDsY547iC9BbMRTNPS+/7//zYEaxZMB3QfP+5d3ilZj9lxVFWXltBcSTY37+SGEYsy+VA3MZQ4GlNaUE444zKitd2c1ppPgETqh6vpq4lwcprK8gPRgiFZEI3lDien/Xv73h+fzdNiE4kXkV3yZVLCJFVQ2uCWMJj78E4eSETSG673FSzj7NPHJF1C+aKtlWPAy02V08/me9eXE7QNCiWlQ+RQ6kVusZWh5HRINFg9u1M4YBBwvP5yVVTJE6HMNnOJgYTiVfRXRIhQogutdoeT73xEY1xh8ZWm+9cXM51F5zCf7z7Kd+5uPMWzEVrqwHF8td2U1oYZkTU5ISiqAyURc5YlkttU5wP6mJYjseeAzEcX2ffzuRrbNeXOB3i/C62s0neCTEQSbyK7pIrmBAig2W5NCUc4q7PT3+7q9Nq3MrKCr488QRcT2cUR4XkxcfXmu9cXE40ZFAQkiyXIncsy6XJdmmI2Sx57p10zD5xw/SssepqzYhoQOJ0iHM8P+vf3/FlO5sYeCReRXfJSp0QIs2yXOrjNnHXx/E0cyrGdVqNW7i2mj993MTuuli6OGpKWXGUgKGIBk1GRcNSsFnkjGW5NCYcHM/nW89sy4jZDw+0Zo9VpSROhwHTUFn//qbKVj5XiP4l8Sq6S65kQgggOSjeVR/jylWbuXDZa+w92EpJfijrHcOiaJCHNu1i2dxJ6YtPahWvJBpizIiIDJRFzliWy77WBPubEzhZVpAf2rSL5fOmdYrVUdGQxOkwoBTcOyezr7p3ziQMGSOLAUjiVXSXbL8UQqRX6BavrU4PiB/atIsHrphMWXE0Y5BcVhylMe6wdW8j9/1mB09cPx2j7WpzXH5YziWJnLLt5HZhz0uWLXjyxhmdYrauJUFLwpVYHaZ8DY+9vicjm+Bjr+/he5ec3d9NE6ITiVfRXXKLUohhzrJcDsZtXF+zZHY5U8cVpb/WbDmsrKzodMdwxWu7geRg+cP6VgKGkkGyyLnkzYjk+U/TUCybO4mQqXikw6rcvXMmsfq/90isDlNBQ3Hz355OyEwOeUKmwc1/ezpBWfoQA5DEq+guuaoJMYxZlsvugzEWPlGdTipx75xJPLf1Yy6bOpabn9zKlRVlPHXjDA60JBgZDXLPS++xdW9juoRBJGgwpiBMMGj2968jhhHLcvngYIyqdrH7yLxp+Fqz9g8f8fj10zkYs6mP2Tz2+h5unXWGxOowpRQYSmUkz1k+bxpyREkMRBKvortkpU6IYSpVnDk1oYPkebm7nt3OjV88lbue3U5pQZhpp4zC9X1ueWortz2zjTkV41hfNYOll02ktCBE2YioDJJFTlmWS4PlpCd0kIzdm9ZtwfXh9Q/que2ZbdTHbIqiQa6efrLE6jDmeJrF67ZkxMridVtwPMkRLwYeiVfRXTKpE2IYsiyXhoSN52cvSxAKGMm6czNP465nt7P81d08Mm8adS0JFj5RzW2/2sbowjAjw1KsWeRWaruw3UXa70NxR2JVZPB09n7O0zJIFgOPxKvoLrnCCTHMWJZLzHU50OJwoDmRNRGKUsmUykXRILUNcZ6prgVg9fxzMQ1FOGBQEg3JIFnklGW5vF8fY9HaapbMLs8au580xtlUs4/V888lFDAIGEpidZgz2/qzjrEiKeLFQCTxKrpLVuqEGEYcx6PBcrAcn8Vrq3mntpEVHRKhPDJvGuv+sId750yi1fbSX3umupa/e/D3fP2XbwDIIFnkXH3c5tdv17J6/rmUn1DYKXZTSXyeqa5lwZo3CRqKMQWSFGW4CxiKn1w1JSNWfnLVFAKSeEIMQBKvorvkSifEMGJ5DiFT0WL7XFlRxt9+7jge2rSTJbPLKckPUVoY5tX3/srK//yQNz5s5M6LzmT5vGnp/f1lxVGWt9WiEyKXLMslEjS4ZEoZC9a8SW1DnC+Xj2HdN84DwPc1/9ohic+oaEjO0AmUghGRAEsvm0heyKTV9hgRCaDktrYYgCReRXcNi0ndKXe/0N9NEKJfOY5H3HVotTWOpwkHDC6dOpZ5P/8jtQ1xXqnZDyTvCK6efy7wZ7bubeTqR//I1HFFPF01A8/XspVN9AvLcmlMOCRcn0Xtaim+UrOfmk+befLGGXz7mbdZNPM0vnNxucSpyOB6mgVr3uq0ne2Zqhn92CohspN4Fd0lVzwhhjjLcrFxabL8dEpky/HRaEoLwhkXjtqGOGaHLR51LQkAji8IEw5LlyFyK5Wl1fM1vs4es57ns3VvI0s31vB01QxG54UkVkWa4yfjpn0x5xWv7cbxJfGEGHgkXkV3yVVPiCEslSnQ8TWer7E9j288driu17K5k7jvNzvYurcRSN4NDJpG+pB2WXGUlddWMDoqg2SRe6kJne36eL5mw1t/4c6LzuwUs66vJVZFl8IBg3vmTMQ0TAwFJQVh7pkzkXBA9rOJgUfiVXSXXPmEGKIsy2VXfYzFazOLM19wagnPVNdS2xDnjg3bWXrZRBaseTN9DqnVdtJn7E4YGZFtbKJfdBW/L2z7mFtnTciIWaVgfdUMiVWRVdBQJFzNTeveyIiloCSeEAOQxKvoLpn2CzFE1cft9IAYDhdnvvGLp6afU9sQZ9yoKBsWnc/aG87j1ff20djqsnRjDdGQKYNk0S9SK3TZ4nfuOSdlxOyIaICE60usii5Zrs9NHYo537RuC5br93PLhOhM4lV0V04ndUqpi5RSO5RS7yul7s7y9W8rpWqUUtuVUpuUUifnsn1CDBWW5eJ2UVi8/Zm5VO2b+pjNt9a/zfrqWsYWR/np1VMplVTwoh+ki4u72YuLJ+skmtTHbP7PizWYSkmsiiPqqi905YySGIAkXkV35WxSp5QygZ8BXwXKgauVUuUdnrYVOEdrPQnYANyXq/YJMVRYlkttU3Lwm6pzk1JWHCV1XUht6fjXl95j4RPV1LUkWFFZwROv78E0DUaEZJAscsuyXPa3JrDbzoBmi99QwODhTbtYurGGW2edgau1xKo4okAXfaHU/RIDkcSr6K5crtRNB97XWn+gtbaBp4HL2j9Ba/2q1rq17dPNQFkO2yfEoGZZLp82xjlo2bTaHodaHR6ZNy2jgOnyedMoigZ47uYvsGR2OWv/8BFzKsaxYdH5rK+awdjiMF+/YDwTSvJl5UPklGW5HEo4NMVd9tTF2PDWXzrHb2UFjudxxbnjWF81g3HFYY7Lk1U6cWSRgMHyLH1hRBJPiAFI4lV0Vy6vhGOBve0+rwXOO8LzbwBeyvYFpVQVUAVw0kkn9Vb7hOgTuYhXy3J5vz7GorXVLJldztKNNZQWhPnXORNZs2A6hgJfQyigcH1NNGSwdGMNtQ1xXv+gnmVzJxE0DYrzIhTn9UkTxSDRH/2rZbl8cDDGvqYES557h9KCMLd/5Uxe2PYxq+efi2koQgGDf6+uZX11LSsrK+QMnQCOLl49rQkFVEZf6PkenpbtbCK3JF5FX8rlFTHbunHWCFVKVQLnABdm+7rWehWwCuCcc86RKB+CjrVg/If3XNxHLem5vo7XRCJ5BilVlLkoGqS2IU5tQ5x/evYdFs08jaJokBOLotTHEjzy6vvc8qUJPHjFFHytabU9SgvDsoVNALnvXy3LpcFyqHqimgcun5yO3ftf3sGimadxMGZz3IgIpqE4//TRXDz5RIryAjKhE8DRxWvc8bm7XV+Yqvv1k6um5LStQki8ir6Uy6tiLTCu3edlwCcdn6SU+p/Ad4ALtdaJHLVNiEGr2XZJeIeTSjTGnXSdua17G1n4RDVlxVHWfeM8vvnkVpbMLmfxui2snn8uh+IOp5bmc1y+bGETuWdZLvtiCRTJRABdxe7TVTNosRyaLZeivCAFQYlVcfRMQ1HXkmDhE9Xpx8qKoxhyRkkMQBKvortyuUH3TWCCUmq8UioEXAU83/4JSqmpwErgUq31/hy2TYhBK+H67GtKpPffr3htN/fOmZSxH3/Z3EkAGSt5pqEoLQzLhE70C8fx2F0f40cv1KST+mSL3RWVFZgKDsVdTi3Np2xElJCsKotjEDQUy+Z27hOl7pcYiCReRXfl7MqotXaVUt8EXgZM4Jda63eVUj8E3tJaPw8sAwqAXymlAP6itb40V20UYrBJlS6IBg1WVFawaG01W/c28tjre3j8+ukcijvsb05w32928L1LyikrjqZXQ8IBQ84liX5hWS71cZufbNrJdy4uRym4d84k7np2O/e/vIOll03kpJI8IgGDxlYHX8PYoijFkaDEqzhmvtZEQyZLL5tIXsik1faIhkx8OaMkBiCJV9FdOb06aq1fBF7s8Nj32v3//8xle4QYzCzLZVd9DMv2iAQNfv12LWtvOA+l4IO6GLc9s42texuB5J2+xlaHe+dM4rHX97BCEk2IfpJK6qMUXHfBePY3JRhdGOKx1/ewZHZ5+gzJvS+9x3dnlxMOGuSFDPICAcJhiVfRHYpHXn2fORXjyMPE9nweefV9/uXSif3dMCGykHgV3SNXSCEGqfq4zeK11dzz95/nW8+8TW1DnIZWl2/OOp2SghB1LckjqalU8KUFIeqaE/zT1z5HXtCUCZ3oF/VtSX1Wzz+Xx17fwy1fmsC/vvgeN/zNqdz2q23UNsQpK46y8toKIkGTsGlQnB/p78+/k+gAACAASURBVGaLQcw04NZZZ6STSaW39UqGeDEASbyK7pJRnRCDUGrbZW1DnJHRYHqFY1R+iGse/SOlBeH0Y622h2V7NFsui9dtYd03zmNkONjfv4IYhizLxVCwZHY5AUNxx1fOYsGaN6ltiFPXbLNkdjkl+SGK8kJEggrL8RmTF+rvZotBzvE0D23ambES/NCmnXz/krP7u2lCdCLxKrpLJnVCDEItjkvQUCz8H6dQGAlS4vrUx2wKIoF0Svj2mbNevf1Cnv7jR6ysrJDEKKLfeLjYnqYkP8TO/S3pTJdAOtslJOPV1zAmT7YIi57ztKau2c54rK7ZljNKYkCSeBXdJVdLIQYZy3KxXJ+gqbhkShmVv/hjeovGz66ZxpfLx/BKzeHksamkKF+/YDyl+SHJHChyzrZdGuMOdS02C9ttKeoqXiMBk5K8oMSq6BVh0+DOi87kjg3b07G3bO4kQrKfTQxAEq+iuyRChBhEUslRrlq1mbf3HkrvuYdkuYKbn9zCP331c51Swo+KhBhbnCeDZJFztu2ysy7Gu582pyd0cOR4LY7IhE70HtfX6QEyJGPvjg3bcX1Z+RADj8Sr6C65agoxSCQSLgfiNg+37bWfMKaAJbPLWfHa7nSWy9qGOCh4pmoGtq8JGIr/x969x0dZ3vn/f133nDMJJISAh2A9FK3RRSUesHuoXXdrD3T5/RasB2grbQ3YrW67Su0eaHeX734fRXTbaiuI3YoFW7Cw+9W1/aq7dm13V201UlmlBUStiaecyWnumbnnvr5/DDNkyAQhkMkkvJ+PRx6QyRCumXzyue/rvq/r85muKpcyjjoGUtz5xG7+8iNnc8eV59GTSOdjNhev3//MxQQcQyhgmBZVvMqxlfFtwT7jXAxmdJIsZUjxKqOlI6fIBDHoZffRHVwVa/XCOdz+2C62t/RQXxPjtY5BzjqhipCD2hbIuItHHP78j87kk9/75bCYbe9P8lrHILFwgFNqYprQyZiIBIsvZ4sEtVhJyo/iVUZLESIyAbiuR2tPkp1v9Q1bcnnrth0sv+yM/MnynU/swcv4mtDJuHNdj32uz7KNw2P2pstn5+N15pQo0xSvMkZ8KLqczR/fYYkUpXiV0dKkTmQCyPWkqwgH8ok+p7U7wewZldx33UX5ux8Bx+gEWcZNMunRNeDSmUiR8vyiMXva9Hg+XqNBR/EqY2akGEx7Ok2W8qN4ldHSpE6kzCWTB3rS9STS+aISOfU1Mfa09dM1kKK9P8naJY1MV28vGSfJpMfujgEGkj6ebwkYisastZb2/mS2kE9MfRNl7AQcUzQGHceM04hERqZ4ldHSpE6kzPW4aYL7k/y6J/fynWvnFlQLXL1wDtuaWzi5JsaWpnmcOT1OJKK7HjI+OgZTNL/agW+zhXo6+lOsWTSnIGbXLJpDNBxgc9M83lureJWxFXJM0RgM6SRZypDiVUZLR1KRMua6Hh0DKU6ujvDA5y6hvS+Jb302ffYSOvqTdA6kuP+pV/nzy89UlUspC9GQwxW/cxJexpKxlplToyRSaVYtOJeKcIDBVIa6qggOMK0irAmdjLlIyHBSTYwNSy/GMeBbCAUNkZBOkqX8KF5ltHQ0FSlTruvxzkCSh7e38icX1OeLTdTXxLjnk41MiWV/ff/6Yw3MjEc0oZNx57oevoXO/lRBhdZ7PtnI2SdWkfR8go5RER8pKWOg3/WG5dDqWGC8hyYyjOJVRkvLL0XKUDqdoSORoj/psfjS04ZVD1y2sZmKUJATp0Y1oZOykE5n6E15pDx/WIXWZRub8fb3TdSETkqtv0gF1mUbm+l3VXhCyo/iVUZLR1aRMtQ1mMr2pLHQ1usWr4TlW06s1IROxp/reuxLpmnvT+GmM8XjNaN4lfGR2V9oaqjW7gQZq2bOUn4UrzJaulMnUmZc1yOVsQykMrT1JXHTmaKVsCIBtS2Q8ZdbJtyXzNB+qHhV2wIZJyNVEwwY7VGS8qN4ldHSpE6kjLiux8udA1xz7zNctuZJVj70IpXRIHdceV5BJay1SxqZFlPbAhlfruuxt3OAxd/9BX/0jz87dLxG1bZAxkcoYLh7cWHV4LsXzyUU0EmylB/Fq4xWSS+bGmM+DHwLCADftdZ+/aCvR4DvA41AJ3CVtfa1Uo5RZLyk0xk6E6lh+5G+8IPtrFk0h5XzG6iNhzlxalT7kmTc5eJ1meJVylw6Y/n2T/ewcn4D1bEQPYk03/7pHr728XPGe2giwyheZbRKdpQ1xgSA7wB/DLQCzxpjHrbW7hzytM8C3dba9xpjrgZWA1eVaowi46k7kc43GR+qtTuBYwyrHtnJuiWNOkGWsnA48bpW8SplIONbHt/ZxuM72woe/5uPNYzTiERGpniV0Srl8suLgZetta9Ya1PAZmDBQc9ZANy//+9bgcuN0SJimfxc16OtL8kr7QNF19LPmJLtU/fe2rhOkGXcvVu8nlQd48FllzJb8SplIBRwisZpKKAdKFJ+FK8yWqWMkJOBliGft+5/rOhzrLUesA+oLcnoRMZRbtnlnU/sYfXCOQVr6dctaSQeDqp1gZSNQ8XrPUsaqasIc1J1TPEqZSHgGL519fkFcfqtq88n4OiasZQfxauMVimPuMWi8eD6rIfzHIwxTUATwCmnnHL0IxMZQ4cTr7llbK3dCW5/bFd+Lf3JNTGma/malJDiVSaSw4nXRDrD/3rk1wV7lP7XI7/mW1efX8qhiiheZUyV8sjbCswa8nk98OYIz2k1xgSBqUDXwd/IWrseWA9w4YUXqnGHlLXDidfg/hLGrd0Jtrf0sGxjM/U1MR5smqcTZCkpxatMJIcbr+39SZZtbM4/Vl8T050PKTnFq4ylUi6/fBaYbYw5zRgTBq4GHj7oOQ8Dn97/90XAT61Vt0WZ/GpjYdYuaVTbApkQFK8ykYwUr7WKVylDilcZrZJdUrXWesaYLwCPkW1p8D1r7UvGmL8HnrPWPgz8E7DRGPMy2Tt0V5dqfCLjKRoNMrs2zpameXi+JegYVQ2UsqV4lYlE8SoTieJVRqukEWKt/Qnwk4Me++qQv7vAlaUck0i5iEaDnKykLROE4lUmEsWrTCSKVxkN1UcVERERERGZwDSpExERERERmcA0qRMREREREZnANKkTERERERGZwDSpExERERERmcDMRG8DZ4xpB377Lk+bDnSUYDjl5Hh7zYfzejustR8uxWBGongt6nh7vaB4ncj0eotTvJYnvd7iFK/lSa+3uMOK1wk/qTscxpjnrLUXjvc4Sul4e82T6fVOptdyOI631wuT6zVPptdyOPR6J7bJ9nrejV7vxDbZXs+70es9Olp+KSIiIiIiMoFpUiciIiIiIjKBHS+TuvXjPYBxcLy95sn0eifTazkcx9vrhcn1mifTazkcer0T22R7Pe9Gr3dim2yv593o9R6F42JPnYiIiIiIyGR1vNypExERERERmZQ0qRMREREREZnANKkTERERERGZwDSpExERERERmcA0qRMREREREZnANKkTERERERGZwDSpExERERERmcA0qRMREREREZnANKkTERERERGZwDSpExERERERmcA0qRMREREREZnANKkTERERERGZwDSpExERERERmcA0qRMREREREZnAJvyk7sMf/rAF9KGPw/kYd4pXfRzBx7hTvOrjCD7GneJVH0fwMe4Ur/o4go/DMuEndR0dHeM9BJHDpniViUTxKhOJ4lUmEsWrHGsTflInIiIiIiJyPNOkTkREREREZALTpE5ERERERGQC06RORERERERkAtOkTkREREREZAILluo/MsZ8D5gPtFlrzy3ydQN8C/goMAhcZ619vlTjk4kpmfToGEzh+ZagY5heESYSKVlYi4hMSsqtMpEoXmUiGat4LWXEbwC+DXx/hK9/BJi9/+MSYO3+P0WKSiY9dncMcMOmZlq7E9TXxFi7pJEzp8eVzA/h1K/8+Iie/9rXPzZGIxGRcqTcKhOJ4lUmkrGM15JFu7X258aYUw/xlAXA9621FnjGGFNtjDnRWvtWSQYoE4rrenQmUkyJBtncNI+n9rTz5X9+kRs2NbOlaR4nK5HLEJrIihyeVMojkfGoqQjxwOcuIeAYntrTrtwqRySV8mgfOHAnoi4exvehM3HgsdpYmGDQoa0/STrjEwo4zKiMEAwe2c6gjsFU/gQZoLU7oXiVstUxmOK19l42N80j41sCjmH7bzuZVhE66ngtp2g/GWgZ8nnr/seGTeqMMU1AE8App5xSksFJ+XBdjz2dw69y3Pan5/Llf34Rz7fjPcQCileZSBSvx69UyqMv7fFmT3JYfr3pg2eUXW4Fxet4KLZ0LBA4MDmbEg3QelAM/fPnL+Wd3tSwuJo5Jcwn7nk6/9i6JY28b2bVEU3sPN/mJ3Q5rd0JxauUpYqww6l1U7h6/TMFvwsV4aMvc1JOkzpT5LGiv5HW2vXAeoALL7yw/H5rZczk7tAVuyq3uWke9TUxgk6xUBo/itdD01208qJ4PX61D6QARsyv5ZVZsxSvx97QSVs06GCBpOcTcAyVEYeW7uGT/qmxINfe+wtauxP855c/OCyGUp4dMa6GPrZ8UzMPLruUk6pjhz3eoGOor4kVTOzK8VwAFK8Cgyl/xDvLNfGj+97lVP2yFZg15PN64M1xGouUmXQ6w9v7ErwzkBzxqlzGt6xd0sj0ivA4jVJEZOLJ5VfPt4fMr8qtE0s6neGN7kHe7Bnkje5BXu8a4I3uQdLpzIj/Jrff56r1z/DFzb/ilY4BFq17mg+seXL/nYUkdz2xe9gJaco7EDcZOzyGMoeIq4Mf8zL+Eb3O6RVh1i5ppL4mOxHMTTQVr1KOxvLOcjlN6h4GPmWy5gH7tJ9OIHtg2t3enz+wBPZflRsqd1VOG6NFRA7fkeRX5daJI53O8Ju2fv7uX1/i1Y5Brlr/DH9w25Nctf4ZftPWP+LEbuj+tOWXncGKrTuG3Ulb2Dir4N+0dicYelPMMcNjaKS4Chx0N62+JkYwcGSnppFIkDOnx9nSNI+frbiMLU3zdC4gZSt4iBx7tEo2qTPG/BB4GjjLGNNqjPmsMWa5MWb5/qf8BHgFeBm4F/h8qcYm5SuV8ugaTFERDrBh6UVs+uzF/PrNfUWvytXGVMJYRORwHWl+lfHluh5vdA/y287sHbe39mXvwKVS3rDntvVnl0gubJzFrdt2DLuz1tafLPp/DL2LUB0LFb2jUBsvjIX6mhhDbzI89HzrsBgKB03RuAoHTcFj65Y0MqMycsTvTSQS5OSaCt5TG+fkmgqdC0jZikcc1h30u7BuSSPxyATaU2etveZdvm6BPyvRcKTMua7HvmSavqRHe18yf7WwvibGnVdfwNs9g2xpmpffqF0dc4hGlcRFRN6N8mt5yu0Zz73vFWGHqkgYz/PZl0zT3p9i+ZC9bKsXzuH+p17lxsvP5Ky6OOHwgZ9RbnI20sRspKVeQ/en9STSRfeq1VVF8o8fPDlr7U6wpbmVRRfVF8ZQJMSU2lDBY7nqlw8uuxQv4xMcZfVLkYkk5VkiQYdVC86lIhxgMJUhEnRIeUe//FJZWspOrrrlQNIjkcqw8qEXC64y3rR5O9++5gJ2v9PP7JmVhIKGWPjIr+yJiBxvcvk1k/HpHkwrv46z3ATbkp2IpTyfN3sSfP/p17jx8jN5Ty38tjNJR19y2M/q1m07WDm/4UD5/iGTutzkbKSJ2UhLvXL7027Y1My6J/eyZtGcgkn/mkVzCAdNweQsV/1y6OSsLl58cnZykYsDR1IURWSiS2csSzc8O+x3ckvTvKP+3prUSdlxfY9pFSGmRIO8vc8tepWxJh4mHg1SFXWoikRwyrDKlYhIucnlV8+3JD1f+bVEDr4Ll1vO2tKbYDDpMZjKFEye7rjyPO56Yjdf+/g53LCpmTuuPK/ozyp3J+7gO28zKiOsXdLIXU/sZvXCOfklmLk7ayMtcRy6Py1X/XLr8ktJeT6OY4iFHWpixWNCkzORdzeWhVI0qZOy4boeru/lyyWvnN9AOOAUvcoYcAynTqkgFAqM44ilmCNtUVBu319ksslNKCrCTr5/mPLr2Bg6eauKBkikfEIBwzt9w3u0zaqJ0NqVfe8Pvgt3849eYOX8hvwJ4Eh33HKPH3znLRQK8L4ZlXzt4+dgDGxpmodvLY4xzKiMHPJnG4kE1bRbZIyMZQsOLVyWsuC6Ht3JFINJn1gowMbPXMyMqjAnVkdYs2hOwYbSb3ziPEKO0QmHiMi7yC23/LeX3iKRUn4dK67r0d7n0rF/QhdyDF0Daa5a/wz7El7RvlQDSZ+KcICKcGDEgiS5E8B1T+5l9cLCn9XqhXPY1tzC2iWN1MWHF7IJhQKcXFPBSdUVnFxTwaxp2SIi+tmKjJ9I0OG+6y7kvusuYkvTPO677iLuu+5CIsdgL6kuxci4c12P3pRHV3+abz2xm4WNs6iNh5kWD5Pxswe9H14/j6SXoaM/RU08zNRIaLyHLSJS1tLpDL3JNCEHLjptOp9Y/wx1lRFuunw2p9RWEHQMJo7y6yjk7sgZAwFjSGZ8vIylayBF50CKbc0tXP/7p1NXGSHgmBGXWw2msq0FRipIUhVz8nvcbn9sF6sWnMtpdXHCjsE48LWPn0NdPFxQJEVEypdvLZiD7soZk338KCkLyLjJTeba+pL0Jz2+91+v8On3n1aw9v/uxXNJeRY3nWFfIs2JU6PMjEdUiU1E5BBc16M3maatP0V/0uOWH71AXWWEW644a1iOrQhbegaVX99NbiJXGXHoS/r5xtndbpqMtXz+gecLKlPe+5+vsPyyM8j4dsTlVvXTYgwmvWEFSe75ZCM1FQEqgmFm1waHVY3Uz0hkYrKWgsJHuQJEU47BkmdlBRkXruvxzkCSjG/pT3rMnBIp2k/n8w88zw+vnwdY4uEo03QwExE5JNf1aBtM4Q/Jr63dCVbObyiaY7c0zeOkqVGm647PMOl0hrb+JL61GJMtFPL6/n3fuROyBz53CYu/+4uilSmrYyHu/fkr3L14bsGkL9f7rzZGvvrl5qZ5+L4lFHCYHg/nl0kGg07RqpEiMvGkfZu/gAPZfLFi6w42q/qlTDSplEd3Yni/ne9/5mJq4+GiS1R8a+l1PWbXxjWhK0KFQ0Qkx3U99nYOsOyg/FpfEztkz7KZ8YgmdEOkUh49ieydzqHHqnX7K0oOPSFr70uOuCeuz/V4sLmVmoogm5vmkSlyt03HNZHjR2aE6pf+Mah+qUIpUjKu67GrfYCdb/XlD5KQDeav/99fM70ykt8EnlNfEyPkGE3oRETeRW4FxLIi+fU7185lMJUpmmODjlF+3c91PXpdl93tA7xU5Fi1fFMzCxtnFfybzoFU0fe1rirCOSdW8bMVl/Gp959GRdjhxKoIJ9dU6P0WOU6F9hc/GkrVL2VCye1FuGFTc9FKX4/vbCOdyfCda+cWVPdat6SRioijA6CIyCG4rkdHIlX0rtHjO9twDERDDmsXF+bY3DJAyb6HAxmP/oTP8hGOVbk7cENta27h7oPe13uWNDIzHmHG1Bjvqc1WnZwWj+puqMjxzsAdV55XkC/uuPK8YbVTRkPZRcbc0L49h+q3Y4zDA8+8yqoF53Lq9ArCAYd4xKEyrBMOEZFD2ZdM42Vs/q7Rwfk1Fg7y1Ydeoq4qzAOfuwSAgGOYrn3KQOFxKhZyuO+6i4iGAiNWpcw9Xl8T46bLz6SuKsyWpnlkrCUSCFBXpabtIjKctfBP//VKfs9tTyLNP/3XK3zt4+cc9fdWJpcxNfRAGXQMP//yZTgwbNP46oVzuO3RX2cPjpVhUhmfjLVEnSDBY9C7Q0ZPe/ZEypvremT2V8k+56QpbFh6Edfd9+yw/Lrq/zuXusownm8xBqZFNaGD7PvXNeQ41et6LN3wLHWVkWFVKVcvnMOOlq6CapTRkEPGh5lVER2vROSQYmGHFR9+H61d2YtF4UD281hYfeqkTKVSHn1Jj7d6kwWbzFcvnMP9T73Kn33wvWxumsfb+1w6B1Lc/tgutrf0sPOtPjY3zcMxKtssInIoqZRHIuPR0lWYZ7/xifPYuvxSWrsTRfOrQRO6HNf1eLlzoOD9W7NoDnWVEba39HDbo9necKfXxQk5hoy1vGd/j7+gY5gaCel9FJHD5mWgZyBV0NLgG584j6nRo+8PqktKcsy5rsfu9gF6k96wTea3btvBwsZZ/NkPtuOmfRate5plG5vZ3tKTf07Gt5rQiYgcgut6vN6ToM/1h+XZLz34Aol0Rvn1XeRWkhz8/q3YuoPll50BwPaWHpZueBaA+596lX0Jj+mxMDOnxpg5Nab3UUSOSCrj86UHXxiWs1MZ/6i/t7KRHFPZhuJpqvYf6IptMs+V1Q4YRmzIqgOljKcjXXL62tc/NkYjERnOdT1c3yMWCuT3Kg+Vm7wpv47MdT263PSI71917MBV89z79qn3n6YJsYgclZFaGmTU0kDKSW5fwkAqwyvtA7zZkyhatjVXKKUv6bFuSaMqsYmIHCbX9UjhsS+R4a19LinP50MNMwqeU18To6M/xT3Kr0W5rkdrb4Jdb/eRzvjcd91FXDCrOv/1+poYg6lM/u+5902tCETkaIUCDh9qmME9n2xkS9M87vlkIx9qmEEooD11UgZSKY99SY93Dto/d8eV5/Htay/gCz/YPmxP3ZpFc6iNh/m3l94q2HCuq6AiIsW5rkdv2qPtoFy7dvFcINu6ILcnLBZymF4ZZsPSiwkFjPLrEL0pj/a+ZMGeljWL5nDbo7to70/yjU+cx8k1MX6+4jICet9E5BiKhgw3Xn4mNwzN4UsaiYYmWJ86Y8yHjTG7jDEvG2O+UuTrpxhj/sMYs90Ys8MY89FSjk+OnOt6vNWXpL/I/rmbf/QC/a7HyvkN/GzFZfzw+nnMmhbjU5eeym2P7iKV8fnbR36D51v6k54OnCIiI8g1Fk+mh++hu+GB57n1I2fzsxWXsemz2XYFf/vwTgZSGf7oH38GoPy6Xyrlkc74+YqWcGAf3V3XXsDmpnmcMq0CxxhmxNUoXESOrcGUn5/Qwf4cvqmZwdTR76kr2aTOGBMAvgN8BGgArjHGNBz0tL8BHrTWXgBcDdxdqvHJkXNdj72dAyz+7i9o6x3e8La1O8EJU6PUxsOkM5abfrida+/9BcYY6qrCeJnsno+QYzizrlIHThGRIobm2jd7EkVzLRbaepO80ZPI33HK7avTProD+tMeqYxf9D1MeT4Z35KxlmpVtRSRMeCPsKfOn2B76i4GXrbWvmKtTQGbgQUHPccCU/b/fSrwZgnHJ0eoN+XR1pfkjivPY1o8XHT/XEtXgkXrnua6+37JLVecRV1lhFu37eCvP9bA1udeZ92SRqbFwoRCgXF6FSIi5SuZzFZoTKQzrJzfQDrjF821r3cN4lvL/U+9yk2Xz+buxXPZ+tzr2kc3hOt6uGmfjE/R9zASdAgHjdo9iMiYCTimaP4JOBNr+eXJQMuQz1v3PzbU3wJLjDGtwE+AG0szNDlSyWR2QrfyoRe5av0zrHnsN9y9eG7Bpvw1i+Zw5xN7gAPtDJZfdgat3QmMgU+9/zTeWxvXwVNEpIhk0mN3xwBXrX+GReueZtUjO6mMBrnjyvMKcu3qhdlce/OPXmBh4yzOqItzwpQIn3r/acxWjs3rS3t09qe47dFfs3rhnIL3cN2SRhwHpkQcvV8iMmZCATPsfPnuxXMJBY5+UlfKzFVstAffa7wG2GCtvcMYcymw0RhzrrW2YKGpMaYJaAI45ZRTxmSwMrJUyqNjsLC3z+M72wDym/IDjuHGH2zP90eCA2Wi62tix1VzccWrTCSK1/LROZgatvfiCz/YzppFc/INsX/zdl++uThAbTyMMQYw1MVDhMOTO8ceSbwmPZ8bHnie1u4E7X0pVs5voDYeproiTDRk8H0IOZP7/ZLxpfwq6Yzl2z/dw8r5DVTHQvQk0nz7p3v42sfPOervXco7da3ArCGf1zN8eeVngQcBrLVPA1Fg+sHfyFq73lp7obX2wrq6ujEarhSTSnns6RjgrX3usDXBj+9swzHQn/TwLbT3Jwu+nisTvWbRHCLB4+dqqOJVJhLFa3lwXY/kCHu/QgGHaMihrTdZ0Fy8viZGXVWEYMAwJRyc9BM6OLJ4HdofantLD8s2NrNo3dMEHHCMIR5xjov3TMaP8qtkfMvjO9tYtrGZq9Y/w7KNzTy+s23C9al7FphtjDnNGBMmWwjl4YOe8zpwOYAx5myyk7r2Eo5RDsF1PdoHUizb2EznQKromuBw0GEwmSHpZYb1oFu3pJETp0aYURVhig6cIiIj6k6meac3WTTPVleEqauKUBMPDcuxsbBDjYp8DOO6HqGAU/T9DAUcqqIOsYDeMxEZW4fKQ0erZBnMWusZY74APAYEgO9Za18yxvw98Jy19mHgZuBeY8yXyC7NvM5ae/RTVzlquXLauSud657cyx1XnsfNP3qhoM+Pm85w73/u5fGdbXyoYQYPfC5bXtsYQzzskM5YaqIhIhEdPEVEDpZOZ+gcTGGtJRpyWLNoTr78fn1NjHuWNNLe53Lff7/Kl/74TDY3zSPj23yOjQeDmtAVkcFjMOUNez/XLp5LOJhdehmp0PsmImMr4MDaxXPzS8FzeShwDOoFljSDWWt/QrYAytDHvjrk7zuB3y3lmOTdpdMZ9nYNsGxjMyvnN+SvMJwwNcqqBedSEQ7Qk0jny2jfd91FtPeleHxnGzvf6mPl/AZWPbKTB5ddyknVsXf530REjk/pdIY97f1889938zfzz+ELP9hOXWUkv/diMJVhWmWY1p7snrDrv5/Nycs2NlNfE+PBZZdSW6mJSTG9ruW6+54d9n66aZ+0Z5kSCo33EEXkOJDOWO46aE/dXcdoT52yv7yrzsHsksvcHbpvX3sBiVSG9r4kSzc8BnRxJAAAIABJREFUO+z5+xJpbrnirPzm/dp4mLVLGpkW1UFTRGQknYMpvvnvu/n0+0+jrTe7b7m1O8Gyjc3552xpmseqR3ayeuEcbn9sV7741D3KsYeU9v2i7+fPVlyG46C7myJSEr7N7qnLFRjMWTn/4NbdR66Ue+pkAnJdD99aVs5vYEvTPJZfdgZexrJi644R99V1DqTy7Qvqa2KcODWqstoiIofguh6eb/nLj5zNrdtGzq89iXS+RcxNl8/mpOoYW5rmcYZy7CEFTPHeUEHHUFhfW0Rk7Dgj5CLHTKw+dTLBuK7HvmSa7oE0qx7ZyVXrn2HVIzupiYfzd+0O7vWzeuEc1j25l9buRP4O3fHSukBEZDRc12Nv5wBXr3+Gtr7ku+ZXyFbBPHV6HItVjj0MxlD0/TQGpqk5u4iUyKFy0dHSUUCKcl2PtsEkAWNIpDOsnN/Auif3sr2lh9c7B6mvibG9pYfbH9uV7/UzNRbiy1t3sL2lJ3+HTicbIiIjc12PnmSatr4kd1x5HtPiYT7UMIPHd7YV5NfqijArfvRCQfuCaNChJqpKl4cjYAzhgMPGz1xMxlre3udy/1Ov8rWPn6P3T0RKxlr4+a53uO+6iwg4hoxv2frc65xae9pRf29lMhkmnc6wL5mmN+GxfFMzdZURbrp8Nrd/4jze6knw8K/e5L7rLqS126UiHGAwlWFKLMiax36Tn9Dd88lG6uJh9fwRETmElPXoT3r5z1u7E9xyxVmcVlvB3FNrqY2HqY2HSWYy+d6fuRxbWzH5m4sfC67r0TfkPQZDfU0Ff/WxswkFjsHlcRGRwxQNOVx76amkvWxxf2MM1156KtHQBGppIBNDMpltLt7el2TlQy9SVxnhlivO4tZtB0pAf+faC3DTPisfejH/2DevOp+lv3saTX9wBtMrI8QjAZ1siIiMwPN89rkpkp6fz7dDc+zHz69n+abm/GN3XHkeaxbNIRRw8qsglGMPT1/aYyDp4fk+n/zeL/Pv6boljcSCx6COuIjIYbIW9g2m+fyQlgZ3L55LPHT0uUh76iTP83zaBlLc+cRuZk2L0dqdYPllZ+QndJC9itw1kM7318g99sUtv8JN+0RDATK+T7WqsImIFOV5Pq91DZDKWDI++b5pcCDH5iZ0ucdu/tEL+Rw7NeZoyeARSHo+XQPpYe/z8k3NeL5a4YpI6SQ9Pz+hg2wu+vwDz5P0jr5ik44KAhwoigKWT7//NFq6slcPqmOhfODlVIQDwx5r7U5wyrQKKiMBqqMhQsfgioOIyGTUnUgxkPTyk7QjzbHRoAp7HImMb0d8TzOa1IlICXm+LZqLjsUFJt2pE1zXY0/nAIm0T9Kz3P/Uq4QChvs/czG1lZFhpVcHU5nipaEDhqmREJGIrhWIiBTjutnJXCTo0N6XJOgML2/9bjk2GNSh+0iEHDPiexrWeykiJVQs5+faqxz19z7q7yATXmcixQ2bmtn02UsIGMtXPnI2XQMpXm7r5/nXOlm7pJEbhuztqK+Jcs+SRpYNeeyeJY3MjEe0JEhEZAS5FREBByL7VzPc/R8vs2bRnPzSwFyOXbeksWBPnXLs6AWDhtPrKvj+Zy6mayBF50CKbc0t/PkfnUl1RFsFRKR0omGH+5ZeRGtXIl9ssH5ajGhYhVLkKLmuhzHZTvbRkENHv1dwIrF64Rwe+VUrW5rm4fkWYwwPPd/K/7y5jw1LL8YxEA06TFPrAhGREbmuR286TedAiqaNhTn2X55/g1ULzuWUaRVEQg7/0qwceyxZH9p6k3zpwRcKChPMrNIkWURKy/MsvYl0QXGsb119PlOOwSo3rTs4juWuGg+mMoQDDknPH7Y5/9ZtO5h7ai0d/Sk+sOZJrr33GT7wvhlUx8KEAoaMtTrZEBE5BNf1GPA80p7NT+jgQI69vGEmSzc8Cwb+/l9fUo49xtK+zU/o4EBhAvcYFCYQETkSad/y55t/VZCP/nzzr0hrT52Mlut6vDOQJJnxGUxlWPnQi7y9zy26ebM2HubtXjf/+ecfeJ6mD5xByDHMmhLTyYaIyAhc16PLTfHWviRvjZBjq2Mh6mtivNWT4PGdbcqxx1hmhMIEvoqkiEiJjWU+0qTuOOS6Hns7B1j83V/w67f68qVVexLpops3p8XDrHtyb/6x1u4EoYChIqKy2iIiI8nl2l1v97N8UzOdA6miOXYwlWH1wjnc8fhuQDn2WAuNYWECEZEjMZb5SJO641CXm84XORnasmDdk3v55lXn54Mttzl/yy9/y/aWnvy/zwVfPKQN5iIiI+nen2tz5fR3v9XLuiWNBTl23ZJGKiNBbn9sVz7PKsceYwbuvPqCgvf9zqsvAM3pRKTUxjAf6RLgcSaV8khnfFq7E1wwq5pp8TD1NdlG47NnVDIlFmTVgnPzFXnCQcP/31jPj198J7+hc+2SRmpjYfWiExE5FGP5+p/+DidVx1j2+6fygffN4M4ndrNyfgO18TB1VRF+sbeDk6fFae9PAijHjoFIwKEyGig4tlVGA0QCuq4tIqU1lvlIk7rjiOt6dCZSBBzDhxpm8On3n8aax37D6oVzuHXbDq7/g9NZuuHZgrW+9TUxNn72YlYtOJdTp8eJBB1qY+pFJyIyklxhlK6BNF/55/+hrjLCnddcwDX3PkNrd3bfHGTz6/c/czE3P/iCcuwYSno+n9nw3LBj24NN88ZxVCJyPBrLfFTSy1TGmA8bY3YZY142xnxlhOd8whiz0xjzkjHmB6Uc32SWazD+d//6El7G568+1kDK82nvS3H7Y7tYOb+BcNApunnTMYbT6+LEwwFmVEZ0siEiMgLX9WjtTTCY8kmkMqyc3wBAxhbfHB9wDF/9eAOnTq9Qjh0j6REKExyLanMiIkdiLPNRySZ1xpgA8B3gI0ADcI0xpuGg58wG/hL4XWvtOcAXSzW+ySyZzN6hi4Ucbrr8TJb80y+5bM2TrHzoRW654iwAlm1sxjHFN28GTHah75RwkGBQy1VEREaSyHi4aZ9r7n2GReueZtUjO7nlirMIjrA5PmAMN/5wO44xyrFjZKT3XoVSRKTUxjIflfJy4MXAy9baVwCMMZuBBcDOIc+5HviOtbYbwFrbVsLxTUq5JZeeb4mFAtz26G9YOb+B6liInkSa+596leWXncGyjc30JtLcvXhuvhpmrkFrOGiYElafJBGRkQxd3t7el+SOK8+jJ5Fm3ZN7uXXbDh5cNq9oft2XSLNuSSMz42qEPVYiQYfvXXchb3S7+T0sJ9dEiWgCLSIlNpb5qJRHkJOBliGftwKXHPScMwGMMf8NBIC/tdY+WprhTT65JZc3bGqmrjLCt645n0+//zRu3bYjf1KxeuEcpkSD1NfESPs+lZEAG5ZejGPAt+yf0IV0siEiMoJcrm1+tYMLT5vOyodeLMixtz+2i/a+FFMrgsPyayTgMDWiHDuW0r5Pb8Ir+Ll886rzqdJ7LiIlNpb5qJSXqYrdVzx4AWkQmA1cBlwDfNcYUz3sGxnTZIx5zhjzXHt7+zEf6GTRmUjlJ3Rf+5MGHGPyEzrIruG9ddsOqivCrJzfwN89vJMVP/ofWroGsUA05DCjQlePj5biVSYSxeuR60ykeK29lz8+50SW728XAwdy7E2Xz6YiHODun+6lpWsQ31r2tvcTdAwzp6q5+NE4nHj1ffjill8V/Fy+uOVX+H4pRyqi/Cpjm49KOalrBWYN+bweeLPIcx6y1qatta8Cu8hO8gpYa9dbay+01l5YV1c3ZgOeyFzXw9u/GfPLHz4LA/lWBkO1difwMgciaXtLD0s3PEv3QAoDOtk4BhSvMpEoXo+M63qEAoZT66bw1j63aI49vS7OvT9/hQebW1m64Vk6+1OsemQnVnU6jtrhxGtmhMIEGf0ApMSUX2Us81EpJ3XPArONMacZY8LA1cDDBz3n/wAfBDDGTCe7HPOVEo5xUnBdj5c7B3ilfYD6mhgnTInSNZDmtY7Bopszd7f15zfzXzCrmvqaGHVVEWpj4XF6BSIi5c91PfZ2DpDOWG7Y1EznQKpojn2lfYAFF5ycz6+DqQzr9veik7E3YmECo0IpIlJaY5mPSjaps9Z6wBeAx4BfAw9aa18yxvy9MeZP9j/tMaDTGLMT+A9ghbW2s1RjnAzS6Qxdbor2viTTK8N859q5WKAiHODOJ/aweuGcgi72qxfOYd2TewuWCd2zpJHp8YDu0omIjCCXa4H8qoh1T+4tmmPvfGJPPr+uW9LIOSdN4b21ceXYEnEcw5pFhT+XNYvm4Kj6pZSpVMrjje5Bfts5wBvdg6RS3ngPSY6RscxHJT2iWGt/AvzkoMe+OuTvFviL/R8yCl2DKTr7U/kNmB9qmMGqBefiV0Zo70/me9LNnlHJnrZ+bn9sF9tbeoDs7d8z6uJMiwcIB0Lj/EpERMpX12CKdMZnSiybK+trYmxv6eH2x3ax8TMX09aXpCeRLsixufwaCYQIhQLjOfzjStLzue3RXQWVn297dBffuvr88R6ayDCplEdLT4LXuxL56ojutBizqmOEw7oQNNGNZT5SdEwi6XSGtG+566d7WDm/gRlVEabGQvQlPW579DesXjiHW7ftYNnGZu677iJWPbJzWEf7UMDRCYeIyCG4rkfGZitYvtLez//9n7dYs2gOK7buYHtLD691DuYvrOUov46fcMChrqpwqWtdVZhQQC0NpPzsS3q46UzBY246w76kR50mdRNeOODQ3p9k2cbm/GO548PROqzoMMacYa3dO8LXLrfWPnHUI5Gj4roee7sGmBINDmtbsGbRHNr7Uvm7dLXxMDOnRPjmVefnK/DU18RYt6SRmqhOOERERpLbs5yrcplbYvkvz7/BqgXncnpdHIA7r76AmzZvV34tA6GA4cY/nM0NQ3oErl08l1BAyy+l/Pi+pdctLHm/ZtEcpleqsM9k4Bi465oLuPGHB44Pd11zAcdiNfjhTvkfN8b8rrX27aEPGmP+CPgXoOrohyKj5boe7YMplm1sZvP184a1LVixdQcr5zewbGNz/srA1uWX8g8//jWrFpzLqdPjREMONeqVJCIyolyD8WJtC1bOb2Dphmf56c0f4FPf+yV1lRHl1zLhen5+QgfZn9kNDzzP5qZ54zwykeHSvmXF1uHncYrXySEWNlRGg6xacG5+eW1lNEgsXLo9dY8C/2aM+T1r7T4AY8wfk53Q3XzUo5Cj0pvyCBjY9NlLyNjipVKrYwf2yNXXxDhxapRvXn0+jjHEIw7xYJBIRCccIiLF5CZ03gjlqKtjIeprYkSDDlua5uH5Np9fp8Wj4zRqgZFLiPu+7nxI+Rmx5L3idVIYTFmmxoJUzKzE8y1BxxAKGgZTlimxd//3h3K4Czi/ALwI/MQYE9s/ofs/wF9Ya+85uiHIaCWTHh19Lu39ST6x/hm+tOVXpDO2aKnUwVQm//e7F8/FGFj83V/Q2j1I2rOa0ImIjMB1PVp6E+x+px/fHznH3rOkEePA767+j4L8KuMrEnSK/szCQe2pk/IzUrxGFK+TggHe6nHZ/U4/b+/L/vlWj8uxWAx+WBGyvyrlJ4Fesq0G/gX4orV2/TEYg4xCMumxu2OAnoTHso3N1FVGuOWKs7jt0V8PK6n9ravPJxpy2Lr8Uh743CVsevq3pDOWVQvOpa4qwtSIKl2KiBSTTHrsS6bp6Euy8qEX+YsHXxhWjnrt4rlURoL41mItbGmap/xaRoKOYe3iucN+ZkG1NJAypHid3KZGQlQcdCOlIhI8JseKEW/PGGPmFnn4fwMPAN8HmnPPsdY+f9QjkSPSMZjirid28zcfa6C1O8Fd15xPz6DHZ3/vdHxr+fqf/g7RUIDqijD9yWy51O0tPdTXxFi14FyCjuGsmZVMi4W1z0NEZASe9fB8SyjgsHJ+A+ue3Mttj+7KF0Vp6Rrkqw+9lM+vW5rmccLUKCHHKL+WiaTn88gLb3DfdRcRcAwZ37L1udf51PtPG++hiQyjeJ3cotEgs4hREQrkl1/WHqNjxaHu1D0HPLv/z9zHk0A9sHzI15496lHIEXMcuPEPZxNwDMt+/1Sqogdm+G7apzIa5B9+/Gs6+5N84QfbWX7ZGUB2Xfap0+PEIw4n1VTohENEZASu6/Fmb5rd7/ST8S3hgMPX/qQBgKUbnqW9L8lAKlPQ69O3lpOmRJVfy4gx8JE5J9HanaC9L0lrd4KPzDnpmFSbEznWFK+TXzQa5OSaCt5TG+fkY3isONR30SWBMuR5Ph0DSRwM/UmPmVMiLP2902nfvzRoaPnbv/ro2XQNpgoKpeTWZSfT2uchIjIS1/XoT3skUplhufXLHz6LFVt30DmQGlaEyjFGbQvKTMAY0p5f8HP8xifOwzE6S5byo3iV0RrxTp219rfA14Aua+1vD/VRuuEe3zzPZ1dbH4vWPU0qY7nvv1+lvS/Fr9/qG1auecXWHZwwNcq6J/dSXxOjJ5GmviaW3chvrPZ5iIiMwHU9BjyPgVSGz4+QW1cvnMO25paCIlT3LGmkNhY+1LeWceD5li89+ELBz/FLD76Ap2qCUoYUrzJa71Yo5dPAURbYlGPBdT3a9negv+mDZ+Bby8LGWSzb1ExFOFC0/G0q49Pen+Q712Y33D7wuUs4pTZCTUT7PEREinFdjwwe/clMfunTUK3dCQyG+596let///SCIlRn1MaVW8vQSG0oVCJeypHiVUbr3SZ1utdbBlzXY0/nAIOpDLcv+h0aTq7mlfYBauNhWrsT+btwQ+X6JW1YejHT4iGmxkJMiQWIBTWhExEpJjeh6xjI0DOYpnMgVTS3hgKGz/ze6ZwwNYpjDCdMjTIzHlFuLVMBxxT9OTrapCRlSPEqo3U4LQ10aWCcub7HtIoQ4YBh1rQ4iVSG6VURTpwapb4mxron9w5rY3Dn1RfQ56b5o3/8Ge/0JjFAVThMUH1ORESGcV2PbjdN50AmW42sMsy25pZhuXXt4rn0uWmuXv8Mb/a43PyjFzCgCV0ZCzpmWBuKNYvmqES8lCXFq4zW4RyF3jbvsjnTWqtd4WPEdT1aupPc9cRuFjbOojYeZnpVmGQ6g28taxfP5YYHnuf2x7Iltt9TW4FjDEnPo2sgewevripCbUwTOhGRYlzXo7U3QUtXgopwgMFUhjNPiHPj5Wdy1xO7WTm/gdp4mLqqCDtaugiHQvmG42u1j67sxUKGk2pibFh6MY4B30IoaIiFdJIs5UfxKqN1OJO6JqBnrAcixXUmsv3oPv3+07h12458JaS7F8/lL3/4K/7qo2ezasG5VIQD9CTS3PzgC7T3J1m14Fxi4QD3LGnUsiARkUPoTaaHVRC+e/FcfvzCGyxsnEV1LETnQIr1P9/LyvnnsOqRl1i3pJG6yjBTIyHl1zLnpi1t+9x88YlcNcFooIKp4z04kYMoXmW0DudI9K/W2rYxH4kUyC0FMgZWXPE+lm54tqAS0ucfeJ6V8xvwrWXphuGtAivCAe3zEBE5BNf1GMx4pH3Liq07iubYZRubC/7NX3+sga9+/Bymq7H4hJEeoZrg5qZ54zwykeEUrzJa77YeT/vpxkGuMMqG/36FroE0+xLpopWQzppZxQn799UNlVsWFHSMTjpERIpwXY+2wSRvdLu8tc8tmmNr44XLKutrYqQzFgftoZtI/BGqCfqqJihlSPEqo6Xql2WoM5Hihk3NLLrwFJZvah6xAturHQN0D6b41tXnD9tQWz8tpn0eIiIj6HbTpDzLDQ88P2KOrauKFOTWuxfPJRw0TFNunVCCI1QTVOEJKUeKVxmtQ07qrLXOsVx6aYz5sDFmlzHmZWPMVw7xvEXGGGuMufBY/d8TScZmr9IEHENrd6JodcvVC+dw5xN7+MIPtpPyfG6/8jyevOUyNjfN47TpcWZNielKsohIEa7rkfZ9HMOIOfbuxXN5ePsbbFh6MT9bkc2tJ0+NMqNCS9onHEPRY+i71IATGR+KVxmlkh2ZjDEB4DvAHwOtwLPGmIettTsPel4VcBPwi1KNrVx4nk9/KkXAZK/SZHxLfU2M7S093P7YrnwFtqmxEF/euoPtLdn6NY4xXLX+GbYuv5STpkap1sZ9EZGiXNejO5km6BgMZsQc66Yz3PHve7jj3/fw8xWXaQ/dBOZbuP+pV1k5v4HqWIieRJr7n3qVr378nPEemsgwilcZrVLWuL8YeNla+4q1NgVsBhYUed4q4DbALeHYxp3n+XQlkrR2J/m7f32J7366kXDQYeNnL+a+6y4CYNUjO0ln/IIJXX1NLN98vK4qwvS4TjxERIrJ7aPL+Jak5/NGT4J/+vSF+Yndqkd2kvR81jz2G97clz0EZZuNO8qrE1g44PBnH3wv4YBT9HORcqJ4ldEq5VHqZKBlyOetwCVDn2CMuQCYZa19xBhzy0jfyBjTRLbVAqeccsoYDLX0epMpUp4l41v+6qMNpH2fxd/9Rb6c7drFc/F8H2MM7f1J4MAt+fufejXfuiAc1olHuZmM8SqT12SN11xF4Yxv2dsxwJ1P7KG9P8k3rzqfH15/Ce/0JukcSHH/U6+y9HdP47ZHd1FfE2Pdkkamx7WHrlwdTrwGnGxF6K6BdP6xinAAnSNLqSleZSyVcgZQbDVwvpSPMcYBvgFc927fyFq7HlgPcOGFF074ckCu6/FWT5Jlm5rzk7g1i+ZQVxmhtTtBa3eCG/aX11735F5Wzm/gfSdU4fkWN53hrz/WwPR4QFeSy9Rki1eZ3CZjvLquxytdAzRtPJBjVy+cw+2P7eKLW37F1//0dxhIZXjfCVX81Ucb6HXT/ONV5xMJZIuihEKB8X4JMoLDidd0xtLRnyroQ7hm0RyqoqGSjlVE8SpjqZSzgFZg1pDP64E3h3xeBZwLPGmyu0FPAB42xvyJtfa5ko2yxFzXozORyk/oILtxf8XWHdx33UX5pZat3QmqY6H8EqEtTfPwMpbKSJDp8QBBowndZHHqV3483kMQmVR6kun8hA6yOfbWbTvyfehCAYdVj/wPW5rm4dtsXo2HHaaEtT95MvCK9CFcsXWH+n5JWVK8ymiV8mbus8BsY8xpxpgwcDXwcO6L1tp91trp1tpTrbWnAs8Ak35C93LnAB39qaI9SfYl0txyxVlcMKt6f38kP7sUc0kjGWupiASo3T+hi0R04iEicjDX9Uhl/KI5dsb+lgWDqUw+rxoDvrWa0E0i6vslE4niVUarZEcsa61njPkC8BgQAL5nrX3JGPP3wHPW2ocP/R0mn4G0R1U0SGB/T5Khv8T1NTE6B1KsemQnqxacSzTkUFcVYdvySzGOob0vSU0sRGU0doj/QUTk+OW6Hh2JAxWFD86xVdEQa5c0ctLUCBkL7X1JqmMh6tUSZlIZ6RgbUN8vKUOKVxmtkm67tNb+xFp7prX2DGvtP+x/7KvFJnTW2ssm61061/Xo6HN5szfJ4u/+ght/sJ27F88d1pNk3ZN7ae1OMGtajNse3cV19z1Lr+vRl/CojATVAFdEZASu67Gnc4Cr1z8zYt+ncNBw1xO76RpI5/NqrVoXTDpGfb9kAlG8ymjpyFViuRON6liQjr4kd1x5Hj2JND/7TRubm+bx9j6XzoEUtz+2i+0tPdTXxNjbPpBvYRBwDKGAQ12FTjxERIpxXY99yXQ+x9oR+j4tbJzF4zvb+MpHzlZencTU90smEsWrjJaOXiXWmUjxWnsvp9dNKahsdMeV5/HUnnbOPrmaVY/sHFahDbJXa8JBh5qYeiaJiIwkg1dQPe5DDTO48fIzuWHT8OqXubxarbw6aUWDDjf+4WxueOD5gjZB0aBqxEv5UbzKaOkIVkKu6xEOOsyZNY3F3/0FdZWR/JWYwVSG9500lZlTImxYejEBBwyG//2Tnfk7dmuXNFITc4gEtexSRORgruuRsh79SZ+2ISsh1j25l7ue2M0Pr5+H5/u81jHI7Y/tor0/ydoljUyNOYRUQXjS8nzLIy+8wX3XXUTAMWR8y9bnXmfp750+3kMTGUbxKqOlo1iJ5FoXQHYzfl1lhFuuOItbt+3IX4lZt6QRN52hsz/JzT96gbrKCDddPpu//GgDkYBh6v4JXVBXa0RECuRybNAxdB7U4yl3V+6dXpd/+PGvueny2fk+dLGwQ1gVhCc131r+4KyZLN3wbEFMWKtqglJ+FK8yWpodlEBuH91V658h6fl0DqS46fLZ+QkdZMvVLt/UDBj+6b9eYeX8Bv76Y2fzntoKHGOJhBzi0agmdCIiB8m1h7lq/TOkfZtftgQHetLddPlsaivDw/JqTTyqCd0kZy3Djre3btuBKsRLOVK8ymhphlACnYlUfi9Hxrdsa27hlNqKon1IvIzPF/5wNu87oYrplREqwgGMMVSGdNIhInIw1/XoctMsH5Jji+XWU6fHCTqG2nhYefU4M1JMZHSWLGVI8SqjpUndGEsmPbwhv6D3/vwVbvzD2bT3JfPlanPqa2L8tnOQ3kSaxd/9Ba92DJDxLTMqIrqSLCJykGTSozuZwhvSXNxA0dwaDhh2vtXHJ7/3S+XV40ww4BSNiWBAp0BSfhSvMlqKkDHkuh4dg6l8I0mAB5tb2fj0b6mvibJuSeOwPiR3PrGHE6ZG81eW1TNJRKS4/rSH71OQY/cl0nzzqvMLcus3rzqfjLXUxsOsW9KovHrcsUVjwqA7H1KOFK8yOjqqjZHcHo/lm5qpq4ywdvHc/D6Pp17pZEHnyVSEHVYtOJeKcICeRDpfjS1gsicosZBKbIuIFOO6Hm/3Jlm2sZkfXH8JqxfO4dZtO3hzn8u25paCHk/rf76Xv/zo2ZwwNcp0TeiOO9bC+p/vHRYTX1PfLylDilcZLR3ZxkCuCltuj0drd4L+pDdsAldXFebPPjibP/vBgV4kaxbNoaM/xfpPNjItFhrvlyIiUnZyOXbZxmyOfb1zkO8//Ror5zdw0tRo0bwaCTpMi2ruM0klAAAgAElEQVRCdzwKBRyu//3T+dKDL+Rj4hufOI+QlrNJGVK8ymjp6HaM5e7QuelMwUbX2x7dxS1XnMXNPzrwS3rLFWfxwDO/ZeX8BmrjYaorwoBlaizE1EiQcFg/HhGRoVzXY2/XAPFwMJ9j73h8N3+34Bw+v381xIcaZvD9z1xMn+sRjwQBqwndccxaSzh4YGXMYCpDOOioRLyUJcWrjJaOcMeQ63q8M5Bk+aZmVs5voL4mlj/p2N7Sw/1PvZpvJrmnrZ/bH9vF9pYeHmxuBeCnN3+AKdEQ06ui4/kyROQInfqVHx/R81/7+sfGaCSTm+t69CTTfOvfd/O1j5+Tz7HbW3owULBc6eYHX2B7Sw9P/MUHmBoLaUJ3HHM9n799eCfLLzuDCgKkMtnPv3X1+eM9NJFhFK8yWjrKHUP9aY+AY7jjyvOorgjyw+vnkc74ZHzL1ude5yNzTuKdXpcTpkYJH3Qbvb4mRsAxTItHxmn0IiLlzSObY7/28XPwfMuPll/KO/tckp5PLBxkW3MLj+9syz8/WzFOefV4F3QMC847gbNPnIJvLSdWx1hw3gkEHDPeQxMZRvEqo6VJ3THiuh7v9CZZtqmZ959ey5JL38Nn738mv9TyvqUXkUz7fOWf/6dgn8dtj2aLo9zzyUZmxiM4+qUVERnGdT26BjPsG0xzwwPPU1cZ4csfPosVW3fkc+raxXMBeHxnG/U1MeVVASAecWg8bTrX3nvgmLx2SSPxiPYoSflRvMpoaVJ3DLiuR1cixbL9hVGWX3YGn/reL/NLL1u7E7R2JVj50IsFj63YuoPNTfMIGENNVMuDREQO5roevSmPdMbHWvJVhFfOb8hP6CCbU2944Hk2N83jbz7WQCjgKK8KAANJnxv2H59hf6xsamZL0zyqK8Z5cCIHUbzKaOlod5Rc16O1N4ExhtbuBBfMqsYYCoqkAFSEA8Mea+1OYK1lemWEUChQymGLiJS9XH5t70syJRYqyKPVsVDRnJrOWKbGAlSFQ8qrAoDn26Kx4vkqPCHlR/Eqo6V7uUepK5GipStbUru+Jsbyy87gtY7BfNPInMFUZthj9TUxHGN04iEiUkTn/vy6YusOwgGnILf2JNJFc2ok6DAtHlVelbzgkOb0OfU1MYJalitlSPEqo1XSSZ0x5sPGmF3GmJeNMV8p8vW/MMbsNMbsMMY8YYx5TynHd6Rc1yOZ8akIB7jziT18+9oLOHNmJdUVIb7/mYv5UMMMIPvLOGtajG9edX7+F7W+JsY9SxqpjYXH8yWIiJQt31qmV4ZZOb+BoGMIBQzfvvYC6mtirHtyL2sWzRmWU6dF1d9TCoUChrWL5xbEytrFcwkFdZIs5UfxKqNVsuWXxpgA8B3gj4FW4FljzMPW2p1DnrYduNBaO2iMuQG4DbiqVGM8Ern2Ba91DAJQVxUmmfb55A9+md/Y+p1r53LjH86moz9Fn5tm/c/38sDnLsFAtiKb+iaJiBTl/j/27j1OqurO+/1n7br3jb7QiKExKgJJa0Bpo+jMZEyYSZwR9ZwBr3RGjREwXjIZNc6cPCQZeeUZr0+iiYLk4gUw3njO0SEm+gyJmdeMEk1DcLQNIqLpFrXv9K127dpV6/zRdNlNVys0dFVfvu/Xq1+hqzfNKvOrtdfaa63fz/WJhQLsMz6rN9dn+tW7LpjPHUvn4RiDYwyPLV+YqeN0VGFEfapkYXAMg+p+OQawGiTLWKR4lZHJ5d3vNOBNa+1bAMaYR4Hzgcykzlr7mwHXbwVqc9i+g+a6Pq1xj1TaEgoYZpRF+dY51Sz7ye8GHWy95pFtrD7/JKYVRygrDPGdc08knkxREA5oQicikoXr+rS5SfxUGscxrDwgYcANT+xg1eJqVm+u5/7aGqJhBwtMjalPlexSacuKDdsGnVOqKovxxIoz8tgqkewUrzJSudx+OQNoGPB94/7XhnMl8MtRbdEIuK7P7tYeLlq3lS/c9Vv+6X//N11uip6En/Vg6/GVhUyfEuEH/2cX7+1zSSTTVGjwISIyRH//euH9L/K5O56ntdvL2q9+anoxjy1fyPTSCO93JCiOOOpTZVjJdHqYxBPpPLVIZHiKVxmpXN4Fs60bZ03lY4ypBU4F/nKYny8HlgMcc8wxR6p9H8t1fVriHndveYNVi6spjYXoiCf5xY53ufzPj+eJFWcwtShMylre3+fy8Itv81ZzD3OOKuKFt1r5+l/NprIwTDiswcdkkq94FRmJfPavHYkkTV0J7rpg/v6iu1H+/R//Esf0Pb3+8X+8xQtvtZJMWQKOIe6lmV4aoTCss8mT1cHEa8AYVvzFsSw99RgCjiGVtjz5+z/hGG1nk9xSvMpoyuXsohGYOeD7KmDvgRcZY/4K+Bbwl9baRLZfZK1dB6wDOPXUU3OS49XzfHa39QBw2ZnHcfOmvvpIX6yexo1fmktbdwI/nebLP/vwTN2aZQtY/+I7HF95Amv3J0XRhG7yyUe8ioxUvvrXxs44cS/FqqdepbIowq1LTuKDfW6mLl1VWYz7li3gK39xLHHPJxwIUVYYIBYMEwwqkfNkdTDxWhB2WHxyFVc8+PKgYs4FYcWN5JbiVUZTLiPkZWC2MeY4Y0wYuBh4euAFxphTgPuB86y1TTls28fal/Bp6kxQXhjG89P88JJT+O1NZ/Htc0/k3XaXpi4vayHcv/nM0QQdwwkVhdoeJCKSxb6ETzjo0OOleOSq0/nhpacQcAKZCR309alf27iNcCBAS7dHKOBQHI1qQicfq9fLXsy519N2Nhl7FK8yUjm7G1prfeBa4FngdeBxa+1rxphbjDHn7b/sDqAIeMIY8wdjzNPD/Lqccl2fXi/FtJIIrd0eq556lVv+rZ69HXHe3+dSEA4MW1z82KmFFEV13kNEJBvX9Umm0nS5Pj/7z7doaOvrVx1D1j414BhmlscoU+kCOUgq5izjieJVRiqnjzittc9Ya+dYa2dZa7+3/7VvW2uf3v/nv7LWHmWtPXn/13kf/RtHVzKZorXb5c3Wnkxmy/5MbCvPmsVNT75Ca49Hr5catrh4LORQENTgQ0TkQP2JUf7UFmfF+jqW1Mzk5k19/WraMmwB3qqSmB6UyUFTMWcZTxSvMlLatzKMZDLFH5u6ae/1MxO50lgo8/RkWnGExvY4a5/fTVlhiPLC0NBCuF+uoTQSIhQK5POtiIiMOclkio5EkhUb6phaFKayKMKsysJMv+qnU9x3YAHe/WeTNaGTQxENOayprRkSS9GQhkAy9iheZaR0ZxxGU3eCqzfUseHK0zMTuY54kqqyGJVFEabEQlSVxdje0MG/PF3PN8+ey8zyGI9etZC0tYSDDhUFYU3oRESy6E0m8fy+1N3RoMM3z55LQ1s806/+86ZX+e551fx8f58adIwmdDIiiWSauj0tPHLVQqy1GGP4df17VJx4dL6bJjKE4lVGStP+LDzPJxpyeHT5QkKBD5fB1z6/m3svXcD1i2Zz6y9f57Yl8zIDkJuefIW2niS3bH6NgGOYPiWmCZ2ISBZ95+j6atpUlcXwUpabnnyFe7bsGtSvXvPIdvZ2xPneL+oBNKGTEQkGHOYePYVLf7yVv7zjeS798VbmHj2FYEBDIBl7FK8yUrpDHsB1fdrdJK6f4oPOBMeU95UmuHrjNrY3dPDMK+9y6cJjea6+ieYub1C9uoBjuG7RHKYWqGaSiEg2/fU+U2mLMfCDi06my/VpbI/T2B7nzmd3ZvrVGaUxbtn8mvpVOSxuMsXtv9o56H59+692cvclJ+e7aSJDKF5lpDSpGyCR8HmztYd7trzBkpqZVBSGcZNpnv9jE6vPP4ljKgoAaO/xMk+SV6yvA/qeNj+2fCEVsTCRiP6ziogcyHV9PuhJ0NyVoLXHY1NdAzd+aS6xUJCqshiN7fFMv9rfp37n3BOZWqB+VUYu4BiauxOZ+zX03bMDKuYsY5DiVUZKa7kDtPR63LPlDS478zhWb65n6doX+fufvcTn5k7jni27uOxnL9HU6WIMWZOi6LyHiEh2/Zkul/3kdyxd+yKrN9dz2ZnHceezOwk4ZE2KEg07elAmhy0Wzp54IqZizjIGKV5lpHSn3M/zfCzwz3/zab78s5cGFX285pFtrFpczYr1dYQCDhYojAQzB/jDQYeySEgTOhGRLFzXpzXuseKAgro3b3qFVYurcZNpfrHjXR5dvpBUui8pSmHEIeoE1a/KYev10vxwyxuDtrP9cMsbfPvcEykvzHfrRAZTvMpI6W7Jh0+QV2yoY82yBYM+SGuf3832hg5K92e7PKokisXiGEM44BAKQCyggYeISDau69PY2Vc0/K4L5g/qVxvb41QUhokEHf7+zOOIhR2sBTeZ1oROjph02tLc5Q16rbnLI61izjIGKV5lpCb9HbP/jMeKDXVUFkVIW1i9uZ7G9r7U2rctmcdDL+yh10tx25J5XP/z7TR3J3h8xRmUxVSDTkRkOK7r0+Z6NHcluOnJVwb1q3c+u5Pm7gSVxREefmEPl/3Z8RSH1afKkddfMmNgDN6xdB7RoLazydijeJWRmtQR0r8lqLkrQWN7nJVnzeKaR7YN2R70rXOqKYoEM4OQNbU1lEc1+BARGU5//+qnbWZwAh/2q9cvms3a2hp+8/r7LD65Sn2qjJpklhi86clXSGrlQ8YgxauM1KRdqetfoQOYEgvxxepplMZCmQ9Rv8b2OJ2uz9GlUe6++GQcxzBVCVFERIbluj7NvR4Wi7Vk7VdnVRZiDHzxxKMpV58qoyiVtlljMKVBsoxBilcZqUl5F3XdvtIFK/cf2u/PLNST8DNptftVlcVo6UpQVhCiIBygKKRzHiIiwzmwf33g8s9m7VcDjiGeTFFVElOfKqMqGHCyxqCKOctYpHiVkZqUEdIa9zIDDuh7AnL1hjqOnhJlzQFptdfW1lBVHmNqLMzU4qgGHyIiH+HA/vWeLbuGloDZn577mFJN6GT0OVnKEN2xdB6Oyn7JGKR4lZGadHdT1/Xxh1nabu5K8L1fvJ7JftnrpagsCjNF5QpERD5Wtv51e0MHt/9qJ+u/chpNXQl6vRRTi8IUBoOEw+pXZfQl/DS3/2rnoMzWt/9qJ3dffHK+myYyhOJVRmpS3VH7twU1dyWyLm1XFkdo7k6wYn1dZkumJnQiIh/NdX06PZ+mrkTW/rW5O8EbTd2s3lyf6VdVUFxyJRxwMvf2flVlMULaziZjkOJVRmrSREh/YpSVG+q4Z8sublsydGn7d7tb+PlVC/ntTWfx2PKFzK4o1IROROQj9D8se21v57D969raGj4zo0T9quRFOGj6jlIcEJPhoPazydijeJWRmjR31k7PJ21tpvjtU9vfzSxtzyiLYa3lqJIojoHKgogGHSIiB6HT82nuSvDJigJWLa5m7fO7ufPZvq1D04ojTCuOZLJgVhaGteVScq47keI3r3/AI1ctxFqLMYantjVy/ikzqCjKd+tEBlO8ykjl9O5qjDkbuBsIAD+x1t56wM8jwMNADdAKXGStfftw/13X7dsWNDDb5cDit6sWV7N6cz3319bwCSVDERE5KP1966qnXh3St/ZvY199/klMK4kwq7xQEzrJi0jQ4dTjyrn0x1sHFXMOq5izjEGKVxmpnEWIMSYA3Av8DVANXGKMqT7gsiuBdmvtCcD3gdsO5990XZ+97b20ZMl22V/89v79S9wbv3o6s7QtSETkoPQXF8/Wt648a1bfueRlC5g7vYhZ5epbJX/SwxRzTqvul4xBilcZqVxO+08D3rTWvmWt9YBHgfMPuOZ84KH9f34SWGSMGdEmYtf12dXaw4XrtvJuezxrtsvjKwuZVhKhNBbiqEJtuRQRORj95+je2+dm7Vs/Nb2Yx5cv5KiSKOVRFRaX/Bou47WvQbKMQYpXGalcTupmAA0Dvm/c/1rWa6y1PrAPqBjJP9Ya97h6/xPkjngyc+C0X1VZDMcYikNBZpQVaNAhInKQ+lfoWnu8YfvWgGMoiyl7sORf0DFZ4zSowl8yBileZaRyOanLFo0HPnY4mGswxiw3xvzeGPP75ubmrP/YwCcda5/fnTUbW2WBniDL6DuYeBUZKw6lf/2ovvWoKTFCoUAumy6T0MHEq3HgrgvmD4rTuy6Yj9ERJckxxauMplzOaBqBmQO+rwL2DnNNozEmCEwB2g78RdbadcA6gFNPPTXrenT/k47G9jjbGzq489mdrD7/JI6vLCToGCpimtBJbhxMvIqMFYfSv/b3rasWV1NRGOboKVH1rZJTBxOvNg0//c+3BhVz/ul/vsV3zz0xp20VUbzKaMrlvP9lYLYx5jhjTBi4GHj6gGueBi7b/+elwK+ttSMaBFfEwqwZUOejuTtBZXGEWCjAtCKdnxMRGamB/ev2hg5Wb64nGgpoQidjUnkszPWL5rB6cz0XrdvK6s31XL9oDuWxcL6bJjKE4lVGKmd3X2utb4y5FniWvpIGP7PWvmaMuQX4vbX2aeCnwHpjzJv0rdBdPNJ/LxoNMruikMeWL8RPW63OiYgcIepfZTyJRoOcoHiVcULxKiOV0wix1j4DPHPAa98e8GcXuOBI/XvRaJAZ+hCIiBxx6l9lPFG8yniieJWR0LFLERERERGRcUyTOhERERERkXFMkzoREREREZFxTJM6ERERERGRcUyTOhERERERkXHMjLAM3JhhjGkG3vmYy6YCLTlozlgy2d7zwbzfFmvt2blozHAUr1lNtvcLitfxTO83O8Xr2KT3m53idWzS+83uoOJ13E/qDoYx5vfW2lPz3Y5cmmzveSK934n0Xg7GZHu/MLHe80R6LwdD73d8m2jv5+Po/Y5vE+39fBy938Oj7ZciIiIiIiLjmCZ1IiIiIiIi49hkmdSty3cD8mCyveeJ9H4n0ns5GJPt/cLEes8T6b0cDL3f8W2ivZ+Po/c7vk209/Nx9H4Pw6Q4UyciIiIiIjJRTZaVOhERERERkQlJkzoREREREZFxTJM6ERERERGRcUyTOhERERERkXFMkzoREREREZFxTJM6ERERERGRcUyTOhERERERkXFMkzoREREREZFxTJM6ERERERGRcUyTOhERERERkXFMkzoREREREZFxTJM6ERERERGRcUyTOhERERERkXFMkzoREREREZFxbNxP6s4++2wL6EtfB/OVd4pXfR3CV94pXvV1CF95p3jV1yF85Z3iVV+H8HVQxv2krqWlJd9NEDloilcZTxSvMp4oXmU8UbzKkTbuJ3UiIiIiIiKTmSZ1IiIiIiIi45gmdSIiIiIiIuOYJnUiIiIiIiLjmCZ1IiIiIiIi41gwV/+QMeZnwGKgyVp7UpafG+Bu4G+BXuBya+22XLVPxqdEwqel18NPW4KOYWpBmEgkZ2EtckgUryIiR576VhlPRitecxnxDwI/Ah4e5ud/A8ze/3U6sGb//4pklUj4vNHSw9Ub6mhsj1NVFmNNbQ1zphaqM5cxR/EqB+PYf/rFIV3/9q3njFJLRMYH9a0ynoxmvOZs+6W19j+Ato+45HzgYdtnK1BqjDk6N62T8cZ1+55ylESDPLp8Ibf/3Uk0tse5ekMdLb1evpsnMojiVUQmO9f1ebe9l3dae3i3vRfX9fH9NHs74rzT2sPejji+nz7k39vS62UGyID6VhnTWno93m7u5NHlC/ntTWfx6PKFvN3ceUTidSw9wpgBNAz4vnH/a+8deKExZjmwHOCYY47JSeNk7HBdn12tQ59y3P53J/HN//0qftrmu4mDKF4nN8WryOhRvOZetq1jgYBDU3eCZCpNKOBQHg3RGv/wmopYOGs/eFRJmAvvfzHz2traGj51VDHB4MGvOfhpm5nQ9Wtsj4+5vhUUrwIFYYdjK0u4eN3WQZ+FgvDhr7ONpUQpJstrWT+R1tp11tpTrbWnVlZWjnKzZCxxXZ/WePancmfOrqSqLEbQyRZK+aN4nbwUryKjS/F65CUSH66ofbAvzvv7+lbSGtt76eh1eaOlh4vWbeUv73iei9Zt5Y2WHvZ2xrnw/hf5yzue57lX97KrdfA1w/WDnm8HvbZyQx1N3YlDam/QMVSVxQa9Nhb7VlC8CvR66ayfhV7v0FepDzSWJnWNwMwB31cBe/PUFhmDXNdnd2vPsE/lUmnLmtoaphaE89RCkQ8lEj4diaTiVUTGjf7zPhet28o/PPoH3mrpYenavsla38pCgh9ueeMjJ2dfqD56yKD1o/rBA1/zU4c2uJ1aEGZNbU1mYte/8qG+Vcai0VxZHkvbL58GrjXGPEpfgpR91tohWy9lckomU7S5SQojwcxTuYEfiv6ncjoYLWNFj++T2r/1SPEqIrmWTKZo6k5gDFgLKWsJGMO0ogihUCDr3xl4Pm3V4mpuevKVIStpqxZX81x9U+bvNLbHGbgolrZDB62BYfrBwAGraVVlMYKBQ1tviESCzJlayGPLFyr7pYx5HzUmOFw5W6kzxvwceBGYa4xpNMZcaYxZaYxZuf+SZ4C3gDeBHwNfy1XbZGxzXZ+9nS57O+Ls/KCL3U2dWZ/KFUUddeKSd67rs7e9l31xnzc+6KZ+7z7Fq4iMWDKZGpRgpKkzzrvtvXie/5F/549N3fzLv73GnpZeLlq3lc/d3rcV8o9N3SSTqax/b+AqQmkslHVFoaJw8ApYVVmMgYsMjhm6HXL7O61Z+8Fw0Ax6bW1tDdOKIgf3H2aASCTIjLICPllRyIyyAvWtMmZNiTlZPwtTYoc/JctZ1FtrL/mYn1vgmhw1R8YJ1/XZl0gScAyVxREqiyPEkz5N++KDnspFww6xgDpxyS/X9en0klj6BjYnTCvCMdDQ1sOjyxdmVu4UryIyUP+qWv89LRQw+GmoiIX40744jW1xphaFiYWD+GmDtdDUk8DpTTK1IEQ4PLg/aepOcPX+VbWbN70yZLvkY8sXMqOsYEg7Bq4idMSTWVcUKosjmdcPnJw1tsf5df17rKmtGZQU5djKEmZXDF5Nq4iFCQYdHl9xBn4qTTDgMK0ockhJUkTGm14Pqsoigz4LBRGHXg+Koof3uzWqkDHLdX0aO+M0dyUyW0AyTzQKIjS2x5k+pe8TUBQMDrmpieSS6/o09SbY15vk6o3bMvF6x9J5VBRFuO6R7fzg4pOxQGFA8SoiHz64bO72WDlgEnTfsgX8Yse7fOXPj6elK8HPX3qHy848blDfctuSeTz0wh6uWzSHuZWFg/qU/hW34Vbbhju/038+7eoNdax9fjd3LJ036P57x9J5hINmyFbHQGDw5Kw8Ghp0TWVhmHA4yIzo0H7vE6WxLC0RmZjCQWhoTwzJBDuz7NBXqA+kUYWMSf1ZAxva4qx66tUhTxl/ftVC3mzqJugYjp4SwxmDWa5kcvA8n+YeDwt4vs0MuqAvXm968hVWn38S1y+aTdD56PMsIjJx9PcNB05s+vWXO+lJ+Nz4xI5B/cbXNm7jkatOJ5m23PTkK1lX3G7e1Pd6ZuVtwO/uX3EbbrVtuPM7B55PiwYdnlx5Bp6fxnEMsbBDWSyS9Z574OQs2wROZLLrSWTPfvnY8oWUDl08PyRa45Yxx3V9GjrjJNOWgnAg61PGtLVUlceYVpT95iKSC57ns7O5L1Ncc1cCx5A1XgvCAY6dWkhFLKwJncgkMLBv6E/rv7O5B9f98Bxca9zjh1veoLI4MkxmSEh9zIpb/+sHrrxNK4qwpraGTXUN3LZk3pDzOx91bm3g+bSjpsSYPiXGMRWFVJUVUFEY1T1X5DBMluyXIriuT5ubJBxwCDomc+A621PGT5QUaIAsedWZ8ImFAjz8ldOIBB2cYbJa9XopokGHqJ5cH7Jj/+kXh3T927eeM0otGZlDbb+ML/33LLB9GSbTlkgwANhBT+PPPL6CWCjABz0JgnGPilgYP21ZUjOTP7X2Zu033m7pYe704o9ccet//cCVt1AowKemFfGdc0/EGHhs+ULS1uJ8TPZLERldQcfwxeppLKmZSWksREc8yaa6hvGV/VLk47iuz5utPZkCphet20osHODeS08Zkh1LKx6Sb67r815ngssfeIkv3PVbLlq3lZauBPd/eXBWqzuWzmNmeYyKglCeWywiR0oymaKp0+XN1h6++/SrNLTF+zJM3vE8S9a+QEu3R+X+1bALa6qoPeOTXP7AS5l7267WHqbEAlQUhrlny66hq2nLFnDPll0EHVg7zIrbbUvmsamugTW1NVQWDq3JFgoFmFFWwCdKC5hRVsDM8r7MkLp3iuRPSczhukVzWL25novWbWX15nquWzSHkvGU/VLko/Q/7Vx5wD7jlRvq2PjV0zOZA0OOoTwW1oqH5JXr+rTEvaH74vefg/nBRSdTVhAmEnJwDASMUWIUkXHO99O09PRts44n06StZeWGOm79u89wwwFn4lZsqONHl5zC/3XfC1z1ueO54sGXs56hqSyO0Nyd4M5nd7JqcTWlsRC9XoruhE9zd4JkyjKnsjDriptj4DvnnjjkrJ6IjF1d8exn6h5fvpBiZb+U8c51fdy0T8JPZd1nnEylae5KEgsHmFVeqAmd5FXfAwgvc9ZloMb2OH7KkkylWfS/fstvbjwLP51mZomyu+XKaG/X1HbKycn30+xs6uLuf3+DG780l3fbXT5ZUcCqxdUcXRrL2hdMKQjzxeppBBwz7Bmaowoj3F9bw4oNdaxYX5dZgXvgv/awZv+ulHA4OCgJioiMXzpTJxNaCp998RRvt2Q/VxBwDEdPiVKhFToZA9rdJHZ/35stXt9p7eXYigKqymJEAoaji2KKW5FxynV92uIeaWDF+jp+eMnJtHZ7mazMVWUxNn719GHPxH3n3BMz32c7Gx6NBpldWThoBc7sX4HTPU9k4gkMc/Y+oDN1Mt65rk9HPE0qbYc9VxAJOrq5Sd65rs+77b14qTQAj2x9mzXLFgw543LPll2kLayprdFWYZFxqP+z/k5rD2+Jg7YAACAASURBVK1xj+dee4932+M0tseZWhTN1G2Dvifs3/tFPWtra7L2BX7aUhhxWHPAz/tX4YC+lbgBZ976z8Cp7xCZeIwh6/nYI5FUVj2G5E1/jZ6rN9SxanF11nMFFUVhyqMaGEt+DYzVgYV//XSa1eefREE4QEc8yZ3P7qS5O0Es5FBeGFTcjgPaTin9kskU3V6Sxo6hhYEri0I8cPlnSduhW6eeq2/iX84/KWtfYIxhXzzF7IrCQcW49aBSZHKyFh56YU9mrNsRT/LQC3syq/qHQz2K5EV/cfFEMsWqxdVsqf+A25bM4+ZNr2TOFdxfW0NJVGngJb/6Y3XgwebKogien2ZmeQzHOFzzyLbMAPD+2hqee+09/vrEoykrzHPjReSguK5PY2ecaCgw5LPe0pWgJBokFg7gmOxbKbvcJNOKI6w4YDL46/q+viAaDaoYt4gQdAxX/cXxfOPxHZm+4vsXzj8iJQ3Uw0jOua7P7taeQTe/25bM46nt77JqcTWzpxURcAxlhQGCClHJo/6BXijgZAZxp8ws5cYvzeXmTX1bsL5YPY2Hv3IaxsAHnQkCDtQcNzWztUpExrZkMkV30scxZlACpAM/61VlMX5w0ck8eMVnufyBlzOv3XXBfO58diff+78/k1mNM8bw6/r31BeIyCApawkFnczKfq+XIhR0SFslSpFxpn/VY8UB6Vxv3vQKqxZXs3pzPY8tX0hJzCGEtq9J/vSX2QgFHIIDDjavPGtWZpAHfVuv6t/r4s4L5lMUCVJeEKYkElLsiowD/Z/zZCpNOOhgYNjPemN7nH947A/ct2wBG648nZbuBK09Hj/9z7e4btEctu5u5nNzKulOpPHTlr8+8WhtsxSRQayFax/ZPmS1/7HlCw/7dytRiuSM6/p80JMg4aezpnOtKAyzpraGoqhDNKBBseSP6/q82drDhfe/yD88+gdae7xMIoTSWChr/E6fEmVaUZiyAg3iRMaD/s/5d59+lT++38X7+1yAj/2sx70U//OZekoLwnxqejHfPvdE3m7u5NjKEgpCIWaUFfDJikIlOxGRIbKdy21sj2ulTsYH1/VJpH26En1ZLoMBJ+uZhKOnRAkFDRETJBQK5LHFMpklEn2ryf/2h8bMtsqWLo9gANZfeRqhYeLXMYaSSEixKzJOdCd9CiMB/vlvP01Ll4ebTPHePpdPlEbZ+NXTcUz21OO9Xoor/uw4LJaiSJDuRIoFn6xgWlFEn38R+UiBYfqVgNGZOhnjXNenI5EknuyrQ3fPll1UFvetyA3MLnZ/bQ3RkENhUFsuJX9c16fT8wkFDOfMn8Hf/+wlKosifPe8atp6kni+xQIPXP5ZrnjwwzM1a2trmKYVOpFxw3V93uvsy3J55vEVrDxrFm09Hq09Huv+YzfXfH42G7e+k0ngNfCzPrUojAGm7N9mXVGU73cjIuNFNOzwwBWfpbEtnjlTV1UeIxo+/M2TGoHIqEkkfHa39bBi/eCEKHc+u5MfbnmDR65aiJ9KEwk6xJMpTegkr1zXpz3hkUr1HWRu7faoLIrwzbPn0uulBhUb/v6F89n41dN5f59Lr5diWpEmdCLjhev67EskaelKsGbZAowx/P3PXhp0n7r3N7tYUjMzU2anojDM0VOiOiMnIocl6Vs6erwhY4qSyOH3Kzk9U2eMOdsYs9MY86Yx5p+y/PwYY8xvjDHbjTGvGGP+NpftkyOn//B5/4QOPkyIsvKsWTxX34SfShNwDCUxh6qSmG6Uklee9fF8y/udLn98v4ufv/QON35pLkdPiQ0pNvyNx3eQTltueGIHlcURSmOhPLdeRA6Wh4/rpymOBomGAqzMkrhrSc1MKgrDbG/oYPXmeiKhgCZ0InLYkmmbKWcAH44pkulxdKbOGBMA7gX+GmgEXjbGPG2trR9w2f8AHrfWrjHGVAPPAMfmqo1yZLiuT0/Kx02msh4GLY2FqCqLEQ46lMYcIgGdQ5L8cl2fP7UlMoO7/qf1D72wh//nb6uzxrHjGB5bvpCKWJhwWAM9kfHAdX3+1PrhZ/3JlWcMm7jr6ClRfnvTWQQdQ2WhPucicvjS6WESpYynSR1wGvCmtfYtAGPMo8D5wMBJnQVK9v95CrA3h+2Tw5RI9K3O+ak0jjG0dHnDHjJfW1tDUcQhiJKiSP4kEj6+9emIpwc9re8vLv6tc6oJOoYvVk/jufqmzN+rKosRcgzlenIvMq60xj1WbqijsijCqsXVlBaEeODyz3LPll1sb+gA+j7flcURCiIOMUfHAkTkyAnsH1MsqZlJaSxERzzJproGAuOs+PgMoGHA943A6Qdc813gOWPMdUAh8Fe5aZocrkSiLzX0wPNzd10wnx9dekqmHkf/IfO4l6KyKExUN0vJI9f18fHZ05qgOBr8yILDa2prgL6adP1xrAmdyPjiuj5+2lJZFBnyGb9j6Txu/9VOmrsT3F9bw9TCAGHtIhGRIywUMFz7hdl8beO2TP9z37IFhALja1KXrbUHrjVeAjxorb3LGHMGsN4Yc5K1Nj3oFxmzHFgOcMwxx4xKY+XgeZ5Pa6835PzcDU/s4Na/+0zf09BYiF4vRdxLURAJZLKGTQaK17HHdfvKFvhpS0tXgrKC0EcWHL56Qx2PLV/IP/9tNaGAoTjqTNj4VbzKeHIo8doa9wg4husXzR7yGb/pyVd4dPlCHKAw6lAYjY5202USUv8qyZTNTOigr//52sZt4674eCMwc8D3VQzdXnkl8DiAtfZFIApMPfAXWWvXWWtPtdaeWllZOUrNlYPhuj7vdSVIpLIXFI+GAqxYX5dJKDGzPMas8sIJOyDORvE6triuzwc9Cd7b92FClC43ydraGr5YPY1ZlYVZY9lPW95u6cFNpogFJm78Kl5lPDmUeA0FDWlrOX6Yz3gqbUlZO6E/35Jf6l8lNcyZutQROFOXy0ndy8BsY8xxxpgwcDHw9AHX/AlYBGCM+TR9k7rmHLZRDoHr+uxu7WHZT37HGx90U1UWG/TzqrIYU2Ihnlx5Bo8tX0hVWYTSSbRCJ2OP530Ys0vXvsjqzfVcduZxfP//vEFFUYivL5pDQ1s8aywHHcPc6cUcUxpTwgSRccbzfJo7PTa++DbWMuxnvLIgos+3iIyaYMDJ3v8EDn9KlrNJnbXWB64FngVepy/L5WvGmFuMMeftv+wG4CpjzA7g58Dl1trDn7rKqOj0fJq6Etx1wXxKokHuumB+JlCrymLce+kC7nj2j0RDARwHnaGTvEomUzT3eKzIkr78ms+fQCoN8WSKaMgZEstramuoiIX5hCZ0IuNSa2+Su7e8Qe0Zx9K8vz7dwM/42toaSmITd1u1iIwNQQPfv3DwGOP7F84nePhH6nJbfNxa+wx9ZQoGvvbtAX+uB/4sl22SQ+d5Ph1u34RuYPHEn152Kj+46GSmFkfY+X4X5YUhLjntk0wrClOiFTrJI9f16Ugk8QdsezhlZikrz5rFJ6ZEcYzh4nVbM7H8o0tP4da/+wyhgMMnSmMURjTYExnPvFSaaz5/AvviPt94/A9UFkVYff5JHFNRwHsdcSqKQoRyOyQSkUnIt5ZoKMDq80+iIByg10sRDQXwj8AaVk6Lj8v453k+u1p6eG1v55CCrVc+9HvKC8MEDazeXM+bTT1MLY5oQid55bo+Tb0JmroSvNXcQ1VZLJPhcvXmevbuc4es3l37yHZ6vBQ3PLGDtLUUhRS/IuNZNOhQWhDO3Le2N3RwxYMvc9nPXiKZslhrdJ8SkdFn4eqN27jiwZe5aN1WrnjwZa7euG1o6sgRUA8mB60/Y+CK9XX88JJTsh70bOvxqCgKs2bZAqYVRyZVlksZe1zXpy3ukbbQ2u3xy/9+j9uWzMPz05nsd6Wx0LDFh++vreGoQp2xERnvUmlLc1ci62f92KmFlEdDeWqZiEwmyWESpfjjrPi4jGOu67OrtYdEMkVlUYTSWChrYfHWHo/K4gix8OQqWyBjT3/MXr3hw9qJty2Zx7a32zj35BmZ2O2IJ7PG8vQpUaaqFp3IhOCnLa09XtbPeiSoVToRyY2AY7L2Q844Kz4u41h30qcoEqCyKMz/umg+//rM69x76QKueWTboAHzQy/s4dvnnkgsqKQokj/9K3RXZ0mK8vBXTmNPS0+mU137/G5uWzJvUCHi+2trNKEbJcf+0y/y3QSZhAKOYVNdQ9bPetAcgQwFIiIHIeQY7lg6j5ue/LAfumPpPEKa1EkuuK7P+50J7tnyBlf++fFUFEV4rr6J0liYh79yGm09Hq09Hg+9sIfrFs0h4EBZQTjfzZZJynV9GjrjOMYMu0X4ni27MoO77Q0dPPTCHh7+yml0uT6lBSGOKoxoQicyQbiujzFw2ZnH8dALe1i1uJqKwjDlhWGKIgFKwtp6KSK5kbaWWHhwopRYOEBaiVIkFzoTSboTPt86p5pPlMaIhvpqbDxe18gNj++gtcejojDMd849kU+WRyiLhAmFAvlutkxCruuzL5EkFgrgmOy1qFp7PLY3dHDnsztZtbiax5YvZNXiE4mEHCqLwprQiUwwrXEPz7c89MIeltTMpDQWorXH49Zfvo6Xsvq8i0jOpCzc95s38VJpoC8z732/eZOUEqXIaHJdn4T16Uz4xL0U7+9z6fVSfOrooszS8faGDlZvru9bOg44hI22XUp+uK5PY2ec9h6Pbzy+g8qiCHdffDJff/QPmS0Oa5Yt4Ie/3gXA9oYOVqyvo6osxqrF1cw5qpijizShE5loHAMFEYdrvzCbr23cNmjLUzigrZcikjvhgMNVf3E833h8R6Yv+v6F8wkfgeLjGr1IVq7r057wcDDEvfSgenR3X3wy06dEBy0dTyuOUBLWhE7ypyfpEwkGqCiK8MDln+VX//0eRZHgoDiNhBz+8YtzqH+va8hZ0FvOP0nxKzIBOcbQ3OXxo1/vymy9rCgK4/kpjsCOJxGRg+YYKC0MDxqblBaGOQJH6jSpk6GSyRSdiSSt3UmaBxQYh74zSV9/9A88ctXpeKk0x08pJBJ0KFWmS8mjRMJnb2diUKbLjV89nWU/+d2QDFPrv3Iadyydh2MM04oj/OsvX+cf/moOFToHKuPQoSaeefvWc0apJWNX2sKK9X19w3P1TUBfX3DfsgW4fjrPrRORycRLpbnigZeHjE0eX7HwsH+3ztTJIImEzzvtvXQmUqzcUEdBOJA12UQ63VdgPOgYTegkbxIJn70dcVp7B2e6rCyKkExlrwXT2uPR6frc8MQOQkGHfznvJGZXFukcqMgE5aXSWfuCkliI4JF4PC4icpCGG5v4R+BQnSZ1kuH7adrcJD0Jn4ADd10wn/LCcNZkEwD3f7mGCqV9lzxJJHzebO3hu0+/SmLAoO3CmiruunA+AWf4RCkVhWHW1taQspayaEgTOpEJrL8u1EBVZTGCjiEW1jBIRHJnuP5IderkiOnPGhhwDMXREK3dHh3xJL9+/X3uW7ZgyOHyWMhhelGhJnSSN/sSPk2dCb51TjVvNffVnTvz+Aq+9vkTaOlO4CZT3HvpKVzzyPastRSjQYeikM6Bikx00aDDzy4/lXfb3cwZlhllUaJBh4KAPv8ikjsf1R8dLvVmguv67GrtobwgREe3x4oB55JuWzKPX+x4lwevOA1joLkrQUVRmCnacil55Lo+TV0Jfv7SO9z8N5/mni27+NGlp+AYQ+1Pf5eJ37sumM/3LzyZ8sIwbzZ389ALe/j6ojkqLC4yiaTSln29yUEJv75/4XymRHUfE5Hc+qj+6HBp34HQmUjS0pXAQmZCB317fG/e9AoLjq0gFDC4yRRVZTHKYroRSn7tSyQxwLfPPZGgY7jrwvk4xmRWlKEvfm94YgdtvR4AJ1QW8a1zqplVoRVmkckkmbaZ9OHQ1zd84/EdJNNKfSkiuTWa/ZFGNpOc6/bVoSsrCJFOZz+8ObMsxnv7XFJpy5RYiKKQwkbyo3+bcCKVJmUtF6/bmnnStf7K07LGb0VhmGDAEAs5lIT1QEJksknb7Pc2q3oGIpJjo9kfaaVuEusfIO+LJ7n259vBZD+8OSUWwlrLQy/swTFGg2LJC9ftS4xS/14Xnm+HrMq93dKbNX4riyPEwprQiUxWgWHubY5R5ksRya3R7I80qZukXNfng54EXsry9Uf/QGN7HIPltiXzMsHWf6bOGLjpyVe4btEcyo/Anl+RQ+W6Pi1xDzeZYmZ5DMcw5EnXPVt2sba2ZlD8rq2tYUosQElIEzqRycoYhr23iYjk0mj2Rzkd5RhjzgbuBgLAT6y1t2a55kLgu4AFdlhrL81lGycD1/XZ3drD3Vve4OazP50ZHLvJNA+9sIdVi6spjYXoiCd56IU9fOfcE3l8+ULKlVxC8qB/hW7l/vOeD1z+WWaWF1BVFhs0sWvuTlAcDWbid0ZZjOJo33Mrxa3I5GUtw97bRERyaTT7o5yNdIwxAeBe4K+BRuBlY8zT1tr6AdfMBv4Z+DNrbbsxZlqu2jdZeJ5Pm5vk7i1v8LXPn4AFnlx5Bq09Hr/67/e49guzB5UvWFOrWnSSP67r07p/hW7V4mrWPr+bX/73e3ztC7OGlNq4b9kC7v31mzxe10hVWYyfX7WQXi9NWSSc77chInkUDTlc94XZXD3w3rZsAdGQNiuJSG6NZn+Uy5H6acCb1tq3AIwxjwLnA/UDrrkKuNda2w5grW3KYfsmvP4zdGlr+R+Lq+noTXL5Ay8NKl+w7e1WHr1qISlrCTpGEzrJmwNX6PpLFEyfEqWjN0llcZhHly8klbY4xrDhxT2ZCd3a2hoKwg6FQdWhE5ns0hZKYkEevOI0HNP3fSjQ978iIrk0mv1RLkc7M4CGAd83AqcfcM0cAGPMf9G3RfO71tpf5aZ5E5vr+jR0xmnpSnDTk6+wanE1qzfXDylf8POrFnLL5te45LRPMveoIg2IJW/aE8nMhA4+LFGw+vyTuOLBlzMPIu58dicA91xyCl/49HR6vRTlhSFN6EQEAM9Ps+wnLw3arl1VFuOx5Qvz2CoRmYxGsz/K5d6DbEcAD5yXBoHZwFnAJcBPjDGlQ36RMcuNMb83xvy+ubn5iDd0okkk+rawhQMOD/xX3z7e2dOKWLW4mlNmfvift7E9juenueLPjmNmeYzymLatHQmK10Pnuj5uMjWo0ztlZimrFlfzyYoC7v9yDZVFEW7e9Aorz5pFc3eCd1p7uOGJHUwrjlAe1QrzSCleZTw5mHj105bKogj3f7mGx5YvzPQfKS3VSY6pf5XR7I9yOeppBGYO+L4K2Jvlmq3W2iSwxxizk75J3ssDL7LWrgPWAZx66qnqlT9CIuHzRksPbzd3UnNsBZedeRw3b3pl0JbLO5/dyfaGDqrKYsRCDsdPLWRKRNkCjxTF68Hr3yLsp22mREFje5xTZpZy45fmZo3disIw93+5hmjQYeNXT+eowohi9zAoXmU8OZh4jQQdvnn2XG568sP+446l8wgHdaZOckv9q4xmf5TLHu1lYLYx5jhjTBi4GHj6gGv+P+DzAMaYqfRtx3wrh22ccHp9n2lFYT79iVLSlsygGD7ccrnyrFlUlcX4/oXzCThGEzrJC9f1aeyM81ZLD8m05Z4tu1izbAFVZTFWnjUra+xev2g2R0+JMr0kQjQU0IRORIZIp21mAAV9/cdNT75CWit1Mkb5fpq9HXHeae1hb0cc30/nu0lyhIxmf5SzSZ211geuBZ4FXgcet9a+Zoy5xRhz3v7LngVajTH1wG+Am6y1rblq40Tjuj6NHQn+1BYnkUyRTKWzVrGfM62I1eefRGlhWBM6yZu2uEdDW1/nFjCGyuIw0ZDD6vNPYva0oqyxe+zUQlLWUhgMMqOsQLErIkOkrM3af6SsJnUy9vh+mr2dcXq9FKm0pddLsbdTE7uJYjT7o5zuPbDWPmOtnWOtnWWt/d7+175trX16/5+ttfYfrbXV1trPWGsfzWX7JpL+VPBXb6hjWkmEpi4vs51toKqyGKGgw9yjiphZEtOgWPIimUyRTFsKwgEa2+NYLN86p5orHvw9Vzz4MruaurPGbiRgmFag1TkRGZ5jTNb+w1H1cRmD9rkenp+moa2X5q4EDW29eH6afa6X76bJETCa/ZE2lE9A/cXF39vn0tgeJ+gYZpbHuGfLriFV7NcsW0DAMSosLnnjuj5tvR4Bx9DrpVjxF8fiGEN6wNOstc/vHhK799fWKG5F5GMZw5D+47Yl89CcTsaiVBoqikLMOaqI6VOizDmqiIqiECkt1E0Io9kfaTQ0wSSTKdpcj6auBHOnF/GL6/6cjt4kTV0JmrsT3PnszkwV+14vRUVRmFjIaGAsedG/ouyn++oinnJMCdOKI1y8biurFldnEqVsb+jgzmd3svr8kzimvIBI0CEWdhS3IvLxLDz0wp7Mva8jnuShF/bw3XNPzHfLRIaIhaGhLcHy9R/WaF335RqOKY/ku2lyBNhh+qPvHIH+SCOiCSSZTNHW69GT8CkrCOH5lpJYiNWbX6O5y+O2JfO4edMrrFhflynQ7KctUUdhILnnuj67Wnu4ekBx8cdXLKSpK8FdF8wnmUrzo0tP4dpHttPYHqe5O0E05JC2adIYCoOKWxH5eMGAwzWfP4G2niQA4f3fBwParCRjT0/C8oN/f2PQoP8H//4Gt5x/EkXRfLdODlc05HDdF2Zz9cZtmbHPmmULiIYOvz/SqGiCcF2f3W09lMZCtHR7g1Kl9qd+H7hKN6M0RjTkUBRSgWbJPdf1adl/5rN/i2VlUYTWbo9VT72aid27LpjPHUvn8YnSvv3mwQA4KEOriBw8P53GTaaH9C2ptPazydiTtjZr+SmrxD4Tgp+yFEQCPHjFaTgG0haMsfipcZYoRUaH6/o093qsWF+HhSGpUvvLFmxv6GDF+jpueGIHFjShk7zo33LpGFi1uDpTfPOGL87JPLmCvti94YkduMk0jjEk/BRt3UlN6ETkkFgLNzyxY0jfoooGMhbZYcpPKV4nhlDQkErbQYlwUmlLKKhEKQJ0ej4BA48tX0gqnT1V6vSSvjX7/m2XsZDOI0nu9W+5fPiFPbT3JFm9uZ5bf/lHwgGHmeUFw5Yt2PDiHqKhALMqChW3InJIVNJAxhPF68RWHA4RCwWYWV5AZXGEmeUFxEIBisOhw/7dmtSNc67r09Kd4Lv/9hpvt/byVnNP1lSpZYVhnlx5Bhu/ejrFsSBFIQ2MJbd8P82+RJKWrgS1ZxzHig11VBZFuPFLc1n11Kv88f2urLG7r9fjvJOrmKpMlyIyAkHHydq3BB0NgWTsUbxObKFQgOnFUWIhh4BjiIUcphdHCYUCh/27D2mEZIy5CFgETOOACaG19rysf0lGTSLh0+YmWb6+jp/8fQ3BQICAAxuuPJ3/+Uw9z9U3ZfZiBx36kqIEHW1fk5zzPJ+2eJKW/WfmNl51Og9c/llCAQc/bTnz+IpM2YKB5wjur61hWkmEKZEg4bBiVkQOnWPg/toFNHV5FIQD9HopphWHcVTSQMYgxevEFwoFmFFWcMR/70GPkowxdwD/APwG2AtoHTiP+iZ0fYUo7/9yDX4avvrwS5nB8NraGv7xr+cQCQb411++zjfP/jSFkaAmdJJz/UlRUmlLd8LnyZULaelOsnJA1sv7li1gw4vvZJL5fGp6MQHHaHVORA6bY8BiBiVKWVtbo0GyjEmKVxmpQ1nL/XvgEmvtF621l1trrxj4NVoNlOziKZ+eRIr2niRNnYnMABn69l6v3FBHMBDg7dZerl80h7KCILN1HklyzHV9Gjrj7Pqgm/f3uQSMIeHbIfH6tY3buOpzx7O9oYPVm+sJakInIkeInxra56zcUHdEss2JHGmKVxmpQ5nUOcAfRqshcnB8P837++L0JNI0tMVZsaGOgnAg66HaUMBQFAkyrSRCYVCZLiX3+s/QrXrqVS5at5XuhE9zVyJrvAYc01evpbaGCk3oROQISQ6TQCypdIIyBileZaQOZVK3DqgdrYbIx/P9NDubunh5Twt+2mYmcx3x5DCHag2xcIASlS6QPHBdn2TaDiqxURAO0NrjZY3XSNDhseULOb4iongVkSOm/4HRQFVlMQLazyZjkOJVRupQJnWlwNeNMf9ljFljjLln4NdoNVD6uK5PU3eChJfkuMoS9sWTzCiN9e213p9gor8T6F/tiIUdTtCWS8mxZDLFex1xPuhJDCmx0RFPsqmuYUi8rq2toSDsUBhxCBrFq4gcOSHHcMfSwX3OHUvnEdIgWcYgxauM1KGMnqr5cPvlpw74mdaER5Hr+rQnPIKOYXppAd0Jn6QPt2x+jXsvXcA1j2zjzmd3svr8kzh2al82ncJIkIKAVugkt5LJFG80dbNi/3mAR5cvpKoslpnYrX1+N988ey4P/NceVi2upqIwTGVxhNKCAL1emrJImEhEMSsiR47jGEqiQVaff1Imm2BJNIijQbKMQYpXGamDHj1Zaz8/mg2R4aXwcTB4KUsqbSkMB/mXZ1/jufomSmNhHv7KabT1eLT2ePzrM69z/aI5TIlpQie519ydyEzoAP53XSPrrzyN1u6++NxU10BlcYRvn3siftoSdgwBx5DwLGURnaMTkSPPpi3RcF+xX8dA2oIxFqszSjIGKV5lpDSCGuNc12dvZ1+yif6zSVVlMdYsW8AVf3Yct/9qJzc8voOVZ81i9rQiVi0+EddPURDQ/7WSW67r4+/fbnnKzFK+efZciiJBvvzTlwaVLgg4Bmst+3o9jiqJ0tLjMatc24RFZJQYiHvpQWVU1tbWUBzJd8NEslC8ygipPP0Y1xr3aGyLD0o20dge5+qN23CTaW780lwAVm+u553WXhwDM0tiGiBLTqXTlp6kj+MYvlg9jRu/NBc3mebqjduGlC54u6UXxxgKI0FiYaMJnYiMKqWIl/FE8SojpZHUGOW6Pu1uEmNgZnksDU6jmgAAIABJREFUa3rbgnCAG57YwerzT6KiKEwsHFAqeMk51/XpTCTx0paAA986p5rv/aKeb579qWHjNmUtVXr4ICI5oBTxMp4oXmWkcrpSZ4w52xiz0xjzpjHmnz7iuqXGGGuMOTWX7RsrXNdnd1sPD/7XW7T1JGloi2dNb9sRT9LYHuf4ykI+MSWqFTrJOdf1aXM9PuhKcMu/vUaX6xMKGC4787hh47bXSxFyHMWqiOREcJgU8UElnpAxSPEqI5WzSZ0xJgDcC/wNfZk0LzHGVGe5rhi4Hvhdrto21rS7SVasr2PpqcewckMd92zZNSQF/G1L5rH2+d2Z2iVFqkUnOea6Ps29Hqk0bN7xLt8+90QCjoO1cPOmV7LG7R1L5zGzPEZZNJTn1ovIZBEcJkW8BskyFileZaRyOQs4DXjTWvsWgDHmUeB8oP6A61YDtwM35rBtY4Lvp+n2PLxUmjOPryAWDnDXBfPpiCd5avu7rFpcTWksxLSSCP/42A6auxPcX1tDWUyrHpJbruuzu7WHFRvqeODyU1l66kx2fdBNQTjAUSVRKosibG/o4M5nd2bidkZZjJBjmBIJKV5FJGcSqTT/77Z3eeDyzxJwDKm05cf/8RbXLjoh300TGULxKiOVy5HVDKBhwPeNwOkDLzDGnALMtNZuNsYMO6kzxiwHlgMcc8wxo9DU3PP9NM09Li3dSRyg9oxPcvG6rZnMR/deuoCNW9/hhbdaefCK0/jBxSf3DZBjDpFgON/Nl48w0eLVdX26kz6RUICHv3IaheEAf2rrZdVTr2bi9Y6l87j9VzvZ3tDBivV1VJXFeHz5Qk3oxoGJFq8ysR1MvEYCDhecWsUVD76c6aO+f+F8IgHlipPcUrzKaMplhGRbN86c+jTGOMD3gRs+7hdZa9dZa0+11p5aWVl5BJuYP03dCVJp+OGWNyiJhfjaAVkDr3lkGyvPmsWPLj2Fm57YgYHMhC4Y1Ad9LJto8dqd9HmvM8HlD7zEF+76LfXvdfGNx3cMitebnnyF6xfNBsikYy5XEp9xYaLFq0xsBxOvKWuH9FHfeHwHKavEE5JbilcZTbkcYTUCMwd8XwXsHfB9MXAS8LwxBmA68LQx5jxr7e9z1so8CQbA9+F/nFONY2DNsgXs3eey9vndbG/ooLE9TluPh5tM0dydIBRwKIxG891smUQSCZ+WXg+AqzfUUVkU2b/vv4AHr/gs7+9zueu5NzLxOquykN/edBZBxygrq4jkTTJlOfP4Cq763PGDtrMpRbyMRYpXGalcjrJeBmYbY44D3gUuBi7t/6G1dh8wtf97Y8zzwI0TfULnuj77Eklauj1WDCg0eduSeWyqa+DGL83lzmd30tydoLXHo6IwzNraGiWakJzqP0P38p4WPv/p6VQWRfjOedXEvRSX/PjDbcJ3XTCfW3/5R5q7EziOIeAYyrTlUkTyqCAcoPaMTw7aznbfsgXEwoF8N01kCMWrjFTO9u1Za33gWuBZ4HXgcWvta8aYW4wx5+WqHWOJ6/q82drDntbezIQO+pbab970CktqZnLzpr5tbP2TvKOnRDmhQsWaJXc8z6fTS1IcC/LZ46biGMP1i2bT3pPkpidfGRS3Nzyxg+sXzWZtbQ1pazWhE5G88/z0kCMNX9u4Dc9P57llIkMpXmWkcjrastY+AzxzwGvfHubas3LRpnzqSfoUR4MUR4NZC02WxkI0tseZWR7j9l/9kesWzdE2NsmpdNrSnfRp6vJo7kqw6qlX2XDlaRw7tYCmzkTWuD2+spBYOIABxaqI5J0/TDFnX8WcZQxSvMpIKcNGniSTKfZ2Jlj2k9/xx/e7hi0uXlUWIxIM8J1zT2S2Vugkh9JpywddcXq9NG4yxczyGGceX8H+M6/0eqmscRtwDG4yRVlMWVlFJP9UzFnGE8WrjJQmdXnguj773CSxUID1XzmNeVVTWFtbM6S4+Ka6Bu5btgCL1Qqd5Nz/z969x8dZ13n/f32vOWWStE2appxSBLGohbscUqCw+1tRlEUp9HffrYAQFxHsAQEPiLg/rLLW3RssrgJCW2QFpSAg9XHLIgr3D0XvXcCFUKhaKFBAG0GSpmmbw1xzzTXX9/5jJkMmmdAcOqfm/Xw85mFmMpl8R9759Pud63voTSaxFlLpgNnTY8TCIT7/kbmkA0s0ZJgzMz7igNR1ba1MjzscPD2uXVlFpCLU1zisHfZv7Nq2VuprVKOk8iivMlEaJZSY6/okrc9f9yS56bGXWNI6h6a6KAc31LC+rZW6mjARx2AMfG3RUUTDhulRrUuS0nJdn67eFF29SR5/8S3OPfFduCmf2miY637xAo9u6eT0ebO59uyjuHfZQoLAEg45zKxRVkWksvhpmFUf4cefWUhgLY4xmR2n0+VumchIyqtMlHpfJTR4aHO/l+amx17iwlMO5+qNm3O7G61ra+Vffr6Fry06ip39Hg21UWqjRp1kKRnX9dnppkilA8Bwx3++xmc/OJd/+MF/5e3M2tXr8eiWTra82cvqxUcz94B6DehEpCKl05bXdwzkNnZqaczMMjhiVl25myYygvIqE6VruSXiuj7dCY+kH9DVm8ztbDl0d6MVG9pZ0jqHVDpg5d3PEg4ZasJalySl4bo+23b2c876J/nAmsfZNeCxpHUOn73n2RE7s6449Yjc/cNm1TFL04NFpEJ5gR2xU+9VD2zG08YTUoGUV5koDepKwHV9Xu7u59zbniLpB7nz5grtbtRUFyVtYf0nW2mMRbQuSUqmx02x/K63j9ZwU+lRc9oQz5yT2NIYpybiaEAnIhUrPcpugml1kqUCKa8yUeqJlUB3wmNl9hy6dGDZ2L6dr3z0/bQ0xvP+cFsa4zRPixGPOEyP1qijLCXjuj5eOsjl8bg5DdTXhGmsjRbM6eDOrOvbMh8+yNR12Fd+Xu4miLyjSMgpWMciIX1oKpVHeZWJ0qihyFzXz505ctycBnYnUvzjx97P7oEU/3bhAi7+4TO5OdPr21qZVhOiLhwmFtN/GikN1/XZmfAAuONTJ3DTYy+z4tQjuOyeTdzxqQXccv7xuSmYLY1xbjn/eMIhw92XnMQBdTF9+CAiFS0WNtx50Qls35mgNhpiwMsc0RILa4t4qTzKq0yUemNF5Lo+OxIeIcdw+rzZXHjK4Xz+vufyBnEPrDiZnf0edbEwGKgNaUAnpeO6Pq9097MieyV5cEF2fSxMR0+CV3cMsLF9O6sWzePghjgz4hEMmd24dMyGiFSDVNrS6/qs+tkfcnXu5k8cR73+rZUKpLzKROlabpEMdpbPu+0pLr9nE9ecOW/ExijLN7STSlvqYmEa4iEOqNVVDymdwc17Bgd08PaC7MFpl+se38aFpxzO6oe2cNbN/8H533+K3QlfAzoRqRqBhct/vCmvzl3+401oiZJUIuVVJkq9siIY3lnu6Emwx/ULLny11hKPOsScsDrJUjKDHzq4qXTBXHb1Jfn2x4/hyp88zw2PbGX14qM5bFYdsZBhpgZ0IlJF/CHrhQd19CTw00GZWiQyOuVVJko9s31scFv4WNjhjk+dQNgxhEIOYAsufA05hukRne8lpZNMZs6ie+a1HZx+9MEFc9nZm2Td49tYtWgeDfEIhzTGqY06zKyrKWPLRUTGz3FMwTrnOFqjJJVHeZWJ0vTLfWjwCl0s7JBIBVx059N88Nu/4fzvP4WbSnPjecfS0hgHyK2p0zQ2KaVUKk1fyicWNnxw3oH0JVOsWTo/L5c3nncs6x7fxqbtu1h+VztX/uR5HAN1YeVURKpPfcxhbVtrXp1b29ZKfUxdIKk8yqtMlHpp+4jr+mzr7mf5hnZ+9OkTc0cYQOay+afvfIa7LzmJ1YuP5tCZtYRDhqa6kAZ0UlJ7kik83xJYSxBYPn3nMzTXx3JX5Aa8NAc3xOnqSwLZDx8+2crMmqg28BGRqtSfDLj5sZdydW5XIsXNj73E1886iobacrdOJJ/yKhOlXto+4Lo+b/UnWb6hneb6GMYw6sGRF935NL+96lTqakKE9X+/lEgqlWZPMkXPQIo+12f29BhAbs3n8rvac8/9zVWnct+yhaTSlqjW0IlIlfMDy6NbOnl0S2fe4189c16ZWiQyOuVVJko9tX1gj+djraW5PsYN5xxDOnj7vK9N23cBmSse6SCzri7sGOpC2hhFSsN1fZKBj5sKqI2GiIVDpNKWdACnz5ud9w/H4DrPHz3xGouObWFuU51yKiJVLZw9VmhJ65zclY+N7dsJaY2SVCDlVSZKvbVJcl2fPW6KmojDN//fo/hz9wC10RAA1549j2sf3EJXX5J1ba088MyfWdfWqisfUjKu67MrmSKRSvPWniQ1EYfL7tmUO/vm1guOB+DRLZ2ZIwzaWqmNOvzDKYdrvaeI7BciYcOX/v69/KXHBSAacvjS37+XiA5zlgqkvMpElbTHZow5A7gRCAG3W2uvG/b9LwKXAD7QBXzaWvunUrZxPAanXe7o85hZFyFtyTsscs3S+dz8iWN5ubOf5vqoOspSUqlUmm07+1l+19sHi3/nnGNoro/lpl1eevez3PGpE7j4b9/NgJemqT6KtSinIrLfsAHsHkjl/fv8nXOOYUZNpNxNExlBeZWJKtlWOsaYEHAL8FFgHvAJY8zwCcKbgAXW2vnAA8C3StW+8Roc0HX1JnlXU5xYOMSldz874hDnVGBpnpZZZ6eOspTSjn4vN6CDTCa/cP/zXHn6kbnndPQk2NnvceVPnmf2tBiNsQhN9TXKqYjsN/zA8oX7nx9RC32d5iwVSHmViSrl/qgnAq9Ya1+11nrAvcDioU+w1v7aWjuQvfsU0FLC9o3Z4NEF6cBy0IwaPN/S1ZssuDmKMYbm+ijTozqLTkonmfTxRjnA9MAZb58119IYZ/a0GKsXH82sen3oICL7Hz+wo25eJlJplFeZqFIO6g4Btg+535F9bDQXA78oaosmwHV9Xunu59zbnuLK+58nkQro6k3S3e/lzhQZ1NIYxzGGGTEN6KR0kkmfP+9OEDKmYCYHF1u3NMa5fsl8/ucvXmDWtBgzYpraISL7n5BTuBbqMGepRMqrTFQpRxqF0ljwYwdjTBuwAPjAKN9fBiwDOPTQQ/dV+8akx01x02Mvccv5xxFyHHYNeHT3e2xs3871S+Zz9cbNuTnQ69tamV2rqx9TXSnz6ro+u5MpdvV7xBtqWLN0Plc9sDlvnWc8EuJnn/0bZtVHscDXzjqKWZoaLFnlrK9SfId95efjev7r151ZpJbsG2PJa9gxBWthWJ1kKTHlVYqplL24DmDOkPstwBvDn2SM+TBwDfABa22y0AtZa28DbgNYsGBBya5Hu65PyIH/72PzCIcM5932FKsWzWNj+3YuPOVwfvjEa6xaNI+muijN02IcUBdTR1lKllfX9dnpegQWvnD/86xZOp/aaIjVi4+mNhpiwEtTGw3x5u4E//3WJ/nNVacScowGdJKnXPVVZCLGklcDBWuhushSasqrFFMpp18+Dcw1xhxujIkC5wEPDn2CMeY4YD1wtrW2s8BrlM3g1vBdvR7/8vAWkn5mvdK6x7flBnRLWufkBnQN8ZA6ylIyruuzbWc/56x/ir9kd7b81i+3FrwU7qaC3DRMDehEZH8XYEfUQpt9XKTSKK8yUSUb1FlrfeAy4BHgBeB+a+0fjTHfMMacnX3aGqAe+Ikx5jljzIOjvFxJDe50mUilWb6hnSWtc/hz9wAtjXE2bd/FDY9szQ3oDpheQ100RE1IHWUpjWTSZ2fi7Z0udyVSuWz+04Nb8NIBIccw94B6rn1wC5GQw/q2Vg3oRGRKCAK49dev4KUDALx0wK2/foUgKHPDRApQXmWiStqjs9Y+DDw87LGvDfn6w6Vsz96kUmn2JFP8dXeS5Rva+fbHj6GjJ0FDPMJ1v3gxt4Zu0/ZdrH5oC+vaWrFY6iNhYjF1lqU0ej2f5JCdLtc9vm1ENq9fMp/AWrr6khw4o4a6qKMBnYhMCcbAhaccnrfm/fol8zGazyYVSHmViVKvbhSpVJoXO/uIR0Is35B/BWRXIkVXX5IbHtnKqkXzaIhHGPDSNNdHtdOllFzSD9g9kMlmR08id/V49eKjmTMzzraufn74xGt8+m/fnVtsHdeVZBGZIqwlt+a9IR5hVyLFD594ja+fdVS5myYygvIqE6We3Sg6+5Ks3NDOPZeclHcF5Jbzj+eWX7+cuxKy/K72zE6Xn2zVgE5KznV9HAPxaIjvnHNM7sDSrr4kNRGHq36yma6+JOs/2crMuij9SZ8G5VREppBo2OHyD81l5d3P5q58rL3geKLhUm4rIDI2yqtMlHp2BXieT03E4acrTyaVtnlXQO5+6k985aPvJ5FKc8enTmDAS9NQG9FOl1Jyruvzcnc/M+Jh3tjlMqs+yr3LFuL5AZ29Say1XHPm+5lZF6U+FiIcMjTF40SjyqmITB02sETDhjsvOhHHQGAhHaSxOsxZKpDyKhOlYf8wruuzy/VJeAED2YPFv3f+cbmDIJ94tZtUOshNuWyqj2pAJ2XRnfB46LkOAObMrKU+Fmb3QIp0EJDw0jjG0Ov6BNYyPRphZl2NBnQiMvUYiA+rffFouPDpuSLlprzKBKmHN0Qy6dOxJ0FXbzLv0Mfvnnssa5bOJxJymBGPsOaRF/nEie9i9rQY02u04YSUnuv6OA6cecwhnP/93+UdUDqzLsJhs2pxjCEWdjTdUkSmNMcYdg+kRkxnq51RU+6miYygvMpE6Updluf57BjwCDkmN6AD6OhJ8Pn7nmOP67N03ZPsTqS4/ENzef9B0ziwIUY8HClzy2WqGTwzMQjg0mzRh0xWr3pgM2/sShJ2HBriIWqjRgM6EZnSkn6Q6yBDplauvPtZkr72iJfKo7zKRKm3R2ZA91JXPys2tLP2guPzdhxa9/g2Nm3fRUM8QktjnIbaKGCpjRri4QiRSKjczZcpxHV93hpIkk5n5tavWjQvl1HIFP/aaAhrLWkLddrlUkSmOD+wNNfHRvzbntYaJalAyqtMlHp8wM5EihUb2mmujxFYWP3QlryzQX74xGsMeGnWtbVSFwsRcQzT47oMLqXluj67kyn6En7umI3BjN7wyFY2bd9FS2OcAS9NJORQG9J5iSIiNWGHL5/x3rxlFWuWziem3QSlAimvMlFTPiGu65NIpenoSbDi1CP47D35l7yv3riZa86cx7yDpnHg9Bg7+jzqI+ooS2kN7nS55c3e3IAO3s7oilOPyBX+OTPjNNZoHZ2ICGSufAxfVnHVA5vxdeVDKpDyKhM1pQd1ruvTnfB4fcdAZmplPJL7IxrU0ZNg10CKVGAJh2BuU506y1Jy3QmPlRvaqY2GCmZ07ux67lu2kCOa65kzI66Miohk+YEtWDc1nU0qkfIqEzVlB3Wu6/NWfxI/sERChjsvypw5N3h0waDMOroIs+JRGmpr1FmWknNdH2My6+dmT4sVzGjIMTgOTI9qyqWIyFBhx4xaN0UqjfIqEzUlB3Wu6/NKdz8X3P47PrDmcb7y09+T9ANmT4ux9oLjc39MLY1x1re16hw6KasUPj39KVY/tIUv3v88a5bOz8vorRccT100RGMsqpyKiAwTCzsj/m1fe8HxWqMkFUl5lYmakj3A7oTHimHrkpbf1c7dl5zEP/98C6sWzaOpLkrztJgGdFJWruuzOxHk1tF19CT41i+3snrx0by7uY4+1ycacaiPhJVTEZEC/MBy869ezttN8OZfvcw3Fh9d7qaJjKC8ykRNuV6g6/qjzlc2Br5+1lH4gSXsGJriuvIh5eO6Pjtdj/SwvG7avouL7nya3375VBpqI8qpiMg7SPoBj27p5NEtnXmPf/VMnfsllUd5lYmaUtdyXddnW3c/r3b1F5yv/MKbvZx721P0uj7NdeooS/m4rk9/KjPtcrS8WosGdCIiexENOwVraETT2aQCKa8yUVMmIa7rsyPhkUilqYk4fPvjx4xYO3f0wdO5b9lC3tNURzSqjrKUh+v6dA149HppOnuT/OL3b3L9kvkj8tpYF9KATkRkL0LGcON5x+bV0BvPO5aQ0cYTUnmUV5moKdEjHLxCN/TA5u+dfxzX/Y//RiTkcHBDnA1PvsbZx7VwxEwdWSDlM7iJz4phh4v/bNNfcvPrD2mME3KgxlFORUT2xg8CQo5h9eKjqY2GGPDShBxDOtB0Nqk8yqtMVEl7hcaYM4AbgRBwu7X2umHfjwE/AlqBbuBca+3rk/293QlvxIHNl92ziVWL5rH6od/z488spO3kw2mu1VQ2Ka9Cm/hcvXEzqxbNY/ld7bQ0xrl/2UIaY1FdTRYRGQNr4bJ7NuWtTW5pjHPfsoVlbJVIYcqrTFTJpl8aY0LALcBHgXnAJ4wx84Y97WKgx1r7HuA7wPWT+Z2u6/NGz8CoG6M01UVZ39ZKLOxoQCdl906b+DTEI7Q0xlnX1spMraMTERmz4ZtNgQ5zlsqlvMpElXJN3YnAK9baV621HnAvsHjYcxYDP8x+/QBwmjETm0Tsuj4vd/dzzm1P8eJfewsuOj1wRg1HNNUxe7oOFZfyGpx2OdqmKLOnx7j7kpN4T5OmB4uIjEckNMrGE6Eps62AVBHlVSaqlAk5BNg+5H5H9rGCz7HW+sBuoGkiv6w74bEyO41t3ePbCm40MUtXPKRCDE67vOmxl0dkdV1bK3XRsM5MFBGZgJBD4Y0n1EeWCqS8ykSVsodY6Irb8GvJY3kOxphlwDKAQw89tOAvGzqNbdP2XdzwyFZWLZrH+w6chmMMdTFHHWQpifHktaMnkcvq4KYotVGHmXU1pWyyTGFjyatIpRhLXhOpgG8+9ELeYc7ffOgFvnvesaVsqojyKkVVylFNBzBnyP0W4I1RntNhjAkDM4Cdw1/IWnsbcBvAggULCk4yDjuGlsZ43sBu9UNbWL34aGZNi3HgtPrJvh+RMRlvXjdt35XbFOXuS05iWjRS0vbK1DaWvIpUirHkNeIYuvqSLL+rPfdYS2OciKMt4qW0lFcpplJezH0amGuMOdwYEwXOAx4c9pwHgQuzXy8FfmWtnVCnoikeZW1b64hpbEcdPJ33za4nEglN7F2IFMFoeT2gLqasiohMwsx4lHUF6uvMeLTMLRMZSXmViSrZlTprrW+MuQx4hMyRBj+w1v7RGPMN4Blr7YPAvwF3GWNeIXOF7ryJ/r6amjBzm+q4b9lC/MASdgxNWkMnFUp5FREpjpqaMO9RfZUqobzKRJU0Idbah4GHhz32tSFfu8DH99Xvq6kJc4j+CKRKKK8iIsWh+irVRHmVidBeOiIiIiIiIlVMgzoREREREZEqpkGdiIiIiIhIFdOgTkREREREpIppUCciIiIiIlLFzASPgasYxpgu4E97edosYEcJmlNJptp7Hsv73WGtPaMUjRmN8lrQVHu/oLxWM73fwpTXyqT3W5jyWpn0fgsbU16rflA3FsaYZ6y1C8rdjlKaau95f3q/+9N7GYup9n5h/3rP+9N7GQu93+q2v72fvdH7rW772/vZG73fydH0SxERERERkSqmQZ2IiIiIiEgVmyqDutvK3YAymGrveX96v/vTexmLqfZ+Yf96z/vTexkLvd/qtr+9n73R+61u+9v72Ru930mYEmvqRERERERE9ldT5UqdiIiIiIjIfkmDOhERERERkSqmQZ2IiIiIiEgV06BORERERESkimlQJyIiIiIiUsU0qBMREREREaliGtSJiIiIiIhUMQ3qREREREREqpgGdSIiIiIiIlVMgzoREREREZEqpkGdiIiIiIhIFdOgTkREREREpIppUCciIiIiIlLFNKgTERERERGpYhrUiYiIiIiIVLGqH9SdccYZFtBNt7Hcyk551W0ct7JTXnUbx63slFfdxnErO+VVt3HcxqTqB3U7duwodxNExkx5lWqivEo1UV6lmiivsq9V/aBORERERERkKtOgTkREREREpIppUCciIiIiIlLFNKgTERERERGpYuFS/SJjzA+ARUCntfboAt83wI3Ax4AB4FPW2mdL1T6pTsmkz44BDz+whB3DrNoosVjJYi0yLsqrVAtlVUSkOIpVX0tZoe8Evgf8aJTvfxSYm72dBKzN/q9IQcmkz0s7+lm5oZ2OngQtjXHWtrVy5Kw6dT6k4iivUi3256we9pWfj+v5r193ZpFaIiJTUTHra8mmX1prfwvsfIenLAZ+ZDOeAhqMMQeVpnVSjXYMeLk/CoCOngQrN7SzY8Arc8tERlJepVooqyIixVHM+lpJa+oOAbYPud+RfWwEY8wyY8wzxphnurq6StI4qTx+YHN/FIM6ehL4wZjPaSwJ5VVAeZXqUS1ZBeVVqovyKsWsr5U0qDMFHiv4Dq21t1lrF1hrFzQ3Nxe5WVKpwo6hpTGe91hLY5ywUyhK5aO8CiivUj2qJaugvEp1UV6lmPW1kgZ1HcCcIfdbgDfK1BapArNqo6xta839cQzOS55VGy1zy0RGUl6lWiirIiLFUcz6Wkkrnh8ELjPG3Etmg5Td1to3y9wmqWCxWJgjZ9Vx37KF2qFNKp7yKtVCWRURKY5i1tdSHmnwY+BUYJYxpgP4OhABsNauAx4mc5zBK2SONLioVG2T6hWLhTlEHQ2pEsqrVAtlVUSkOIpVX0tWsa21n9jL9y3w2RI1R0REREREZL9QSWvqREREREREZJw0qBMREREREaliGtSJiIiIiIhUMQ3qREREREREqpgGdSIiIiIiIlVMgzoREREREZEqpkGdiIiIiIhIFdOgTkREREREpIppUCciIiIiIlLFNKgTERERERGpYhrUiYiIiIiIVDEN6kRERERERKpYuNwNEJkI3w/o7EuSSgdEQg6z62OEw/qMQiqT8irVRHkVESmeYtVYDeqk6vh+wItv9bJiQzsdPQlaGuOsa2vlfQdMU8dDKo7yKtVEeRURKZ5i1lgCa+QqAAAgAElEQVRVaKk6nX3J3B8DQEdPghUb2unsS5a5ZSIjKa9STZRXEZHiKWaN1aBOqk4qHeT+GAZ19CTw00GZWiQyOuVVqonyKiJSPMWssRrUSdWJhBxaGuN5j7U0xgmHFGepPMqrVBPlVUSkeIpZY0u6ps4YcwZwIxACbrfWXjfs+4cCPwQass/5irX24VK2USrf7PoYP730ZDzfkg4sIccQDRtmxmPlbprICMqrVBPlVUSkeIpZY0s2qDPGhIBbgI8AHcDTxpgHrbVbhjztq8D91tq1xph5wMPAYaVqo1QH3w94a4/HyiGLTNe2tTI9EtFCfqk4yqtUE+VVRKR4illjS1mhTwResda+aq31gHuBxcOeY4Hp2a9nAG+UsH1SJboTb/8xQGYu8soN7XQnvDK3TGQk5VWqifIqIlI8xayxpZx+eQiwfcj9DuCkYc+5FnjUGHM5UAd8uDRNk2riB7bwItPAlqlFIqNTXqWaKK8iIsVTzBpbyit1psBjw9/BJ4A7rbUtwMeAu4wxI9pojFlmjHnGGPNMV1dXEZoqlSzsmMKLTJ1CESs/5XVqU16lmiivIsWjvEoxa2wpB3UdwJwh91sYOb3yYuB+AGvtk0ANMGv4C1lrb7PWLrDWLmhubi5Sc6VSNcWjrG1rzf1RDM5HbopHy9yywpTXqU15lWqivIoUj/IqxayxpZx++TQw1xhzOPAX4Dzg/GHP+TNwGnCnMeb9ZAZ1+ihD8jgOHDA9yr3LFubtHORoDb9UIOVVqoXn+bzc3c/rXXtyeQ07hqZ4lJqakm6WLSKyXypmn6BkVdpa6xtjLgMeIXNcwQ+stX80xnwDeMZa+yBwJfB9Y8wXyEzN/JS1VhP5JU9Xv8e5tz2VNye5pTHOfcsWckhUHQ+pLMqrVIuu/vwF/DAkqxrUiYhMWjH7BCWt0tkz5x4e9tjXhny9BfibUrZJqo8W8ks1UV6lWiirIiLFtb9slCKyT1TbQn6Z2pRXqRbKqohIce0vG6WI7BOjLTJtrqvMhfwytSmvUi2a65RVEZFiKmad1SR5qSq+H/BKdz/tr+3gns8sxNq3F/JHtT5JKozyKtXEcRxm1Uf48WcWEliLYwzhUOZxERGZvGLWWfUqpKp09iVZkV3If+1DLwKZTznuX34yB2shv1QY5VWqSWdfknPWj1zAf//ykzm4If4OPykiImNRzDqrj9+kqqTSQeEFpumgTC0SGZ3yKtVEeRURKa5i1lkN6qSqREJO4QWmIUVZKo/yKtVEeRURKa5i1lnN/5GqMrs+xj2fOQnPtzgGAgvRsGF2fazcTRMZQXmVaqK8iogUVzHrrAZ1UlXS6YDdCT93QO7grkEH1AWEw/o0WSqL8irVRHkVESmuYtbZcf20MabGGLPUGHO1MaYh+9gRxpiZk2qFyBjtGPByfwiQmYe8ckM7Owa8MrdMZCTlVaqJ8ioiUlzFrLNjvlJnjHkP8L+BaUAD8BNgF7Aye/+SSbdGZC/8wBZeYBrYMrVIZHTKq1QT5VVEpLiKWWfHc6Xuu2QGdQcAQ1vzIPDBSbdEZAzCjim8wNQxZWqRyOiUV6kmyquISHEVs86OZ03dKcBCa23amLxf/Gfg4Em3RGQMmuJR7r7kJLp6k3T3e2xs387lpx1JUzxa7qaJjNAUj3LHRSfQsTNBbTTEgJemZWZceZWKpPoqIlJcxewXjHejlEiBxw4Fdk+6JSJ74fsBr3T35w5zbmmMs66tlWde28GB02uo0WHOUmH60z57EilW/ewPuczeeN6x9NdFqNE+VVJBVF9FRIqvmP2C8Uy/fBT44pD71hgzHfgn4OeTaoXIGHT2JXMdDsjMQV6xoZ2DGuvw/HSZWycyUsIL+Ny9z+Vl9nP3PkfC02HOUllUX0VEiq+Y/YLxDAm/CPzaGLMVqAHuA94DvAWcM+mWiOxFKh0UXFzaVBclGg6VqVUio0uPsiA60MYTUmFUX0VEiq+Y/YIxX6mz1r4BHAtcD6wHngG+DBxvre2adEtE9iIScgouLm2eFqOpTms+pPLEwoUzGwnpzC+pLKqvIiLFFx2l1u6LfsG4XsFam7DW/sBae5m19lJr7e3W2sTef1Jk8mbWRFjb1pr7Y3j7wMYYjnZnkwqTTPr0Jn3WLJ2fl9n1n2yluT5W5taJ5FN9FREpvsZRam1jTaFtS8ZnPOfU/cMo37KAC7xird20l9c4A7gRCAG3W2uvK/Ccc4Brs6/7vLX2/LG2UfZv3QmPmx97iVWL5tEQj7ArkeLmx17i62cdxSFaxC8VZseAx0V3PE1zfSyX2QEvzay6KOGwrtRJZVF9FREpvmLW2vH89C1AlMwOmIOr+Rwglf06YozZBJxRaDqmMSaUfY2PAB3A08aYB621W4Y8Zy7wj8DfWGt7jDGzx/uGZP/lB5ZHt3Ty6JbOvMevOXNemVokMrrBA0Y7ehIsv6s99/hvrjq1fI0SGYXqq4hI8RWz1o7n4+JzgE3A35DZKKUm+3U78N+B4wAD/OsoP38imat5r1prPeBeYPGw53wGuMVa2wNgre1EJEsH40o1UV6lmiivIiLFV8xaO55B3b8Cn7PWPmmt9bO3J8nsivlta+3zwJXAB0f5+UOA7UPud2QfG+pI4EhjzH8aY57KTtcUAWBWbbTgPORZtVrEL5VHeZVqoryKiBRfMWvteKZfHgYMFHh8IPs9gNeAxlF+vtAQdPj+nWFgLnAq0AL8H2PM0dbaXXkvZMwyYBnAoYceuveWy34hFgtz5Kw67lu2ED+whB3DrNoosVhlr/dQXqcm5VWqifIqUnzKqxSz1o7nFf4L+FdjzCettX8FMMYcCNwA/C77nLlkrsAV0gHMGXK/BXijwHOestamgNeyZ+LNBZ4e+iRr7W3AbQALFizQgU9TSCwW5pAK72QMp7xOXcqrVBPlVaS4lFeB4tXa8Uy/vAQ4GPizMeZ1Y8xrwJ+zj12SfU4d8M1Rfv5pYK4x5nBjTBQ4D3hw2HP+F9npm8aYWWSmY746jjaKiIiIiIhMKWMeJlprXzbGHA2cDryXzHTKF4D/ba212ef8r3f4ed8YcxnwCJkjDX5grf2jMeYbwDPW2gez3zvdGLMFSANXWWu7J/jeRERERERE9nvjuvaXHbw9kr2Nm7X2YeDhYY99bdjrfzF7ExERERERkb0Y16DOGDMTOAM4lMyZdTnW2m/sw3aJiIiIiIjIGIx5UGeMWQj8HEgCzcBfgIOy918HNKgTEREREREpsfFcqVsD3A18DtgDfAjoB34M/Nu+b5rI21KpNJ19ydz2r7PrY0QioXI3S2RUyqxUC2VVRKR0ilVzxzOomw9cbK21xpg0ELPWvmqMuRq4h8yAT2SfS6XSvNjZx8oN7XT0JHIHNb5vdr06HlKRlFmpFsqqiEjpFLPmjudIA2/I128B78p+3UfmWAORoujsS+bCD9DRk2DlhnY6+5JlbplIYcqsVAtlVUSkdIpZc8dzpe5Z4ATgJeBx4JvGmAOANmDzpFsiMgo/sLnwD+roSRBYndsplWm0zPqBMiuVRfVVRKR0itk/GM+VumuAN7JffxXoAm4GGoFlk26JyCjCjqGlMZ73WEtjnMBmLmOLVJpoyCmY2WhoPCVXpPhUX0VESqeY/YMxv4K19hlr7a+zX3dZaz9qrZ1urV1grf39pFsiMorZ9THWt7Xm/ghaGuNcv2Q+//zzLZoiJBXJMbBm6fy8zK5ZOh/HlLlhIsOovoqIlE4x+wfjOqdOpBwikRCNdRFWLZpHQzzCrkSKGx7Zyqbtu7jmzHnlbp7ICK4f8K1fbs3L7Ld+uZXvnndsuZsmkkf1VUSkdIrZPxjPOXWNwLXAB4HZDLvKZ62dPenWiIzCWlj90Ja8ecgtjXHCuvQhFSjsGLr6kiy/qz33mPIqlUr1VUSkNIrZPxjPBM4fAWcBG8kM7lYNu4kUzez6GGuHTRFa29bK7PpYmVsmMpLyKtVEeRURKY1i1tvxTL88FfiAtfbZSf9WkXGKREK8b3Y99y1bqANypeIpr1JNlFcRkdIoZr0dz6BuG+O7sieyT0UiIQ5prC13M0TGRHmVaqK8ioiURrHq7XgGaZ8D/qcx5hhjjD6+ExERERERqQDjuVL3ChAncwg5xuQv6LPWaqAnIiIiIiJSYuMZ1P0YmAFcAbwFTP7ocxEREREREZmU8QzqFgAnWmv/UKzGiAzneT5d/V5uMWlzXZRoVMcrSmVTbqUaKKciIqVXrNo7nlfYAkyf9G8UGSPP89na1c/KDe109CRoaYyzvq2VI5rqqKlRx0Mqk+v6bOvuZ/mQ3K5ta+W9zXXqMEvFUH0VESk9z/N5qaufFUNq77q2Vo7cB32E8fz0V4F/NcZ8Ffg9kBr6TWvtzr29gDHmDOBGIATcbq29bpTnLQV+ApxgrX1mHG2U/UhXv8fKDe0018dYtWgeDfEInb1JGusizHJQB1kqjuf57HQ9OnuTfPvjx7ArkWLd49tYuaGd+5Yt5BBlViqE6quISOntGEhx02Mv5erurkTm/rVnH83BJRzUPZz930fJX09nsvffcaOU7I6ZtwAfATqAp40xD1prtwx73jQy6/Z+N462yX7IDyzN9TG+9Pfv5eqNm/M+0Qg7DrPV6ZAKs8v12dmfYtXP/pDL6/VL5nPDI1vxAy1Dlsqh+ioiUg6WC085PK/uXr9kPmYfbFUyniMNPjjk9qEht8H7e3Mi8Iq19lVrrQfcCywu8LzVwLcAdxxtk/1Q2DFccdrcXPABOnoSrNjQjh8EZW6dyEh+EOSmVEAmr1dv3MwVp80l7Ji9/LRI6ai+ioiUnrWMqLtXb9zMvvjcd8wfxVlrfzPJ33UIsH3I/Q7gpKFPMMYcB8yx1j5kjPnSaC9kjFkGLAM49NBDJ9ksqVTNdVG8WbW54A/q6Engp6vnqofyOnWk0rZgXg+bVUdzXbRMrRof5XVqUH0VKT3lVdJB4X5C2pb2Sh0AxpiDjTELjTF/N/Q2lh8t8FjuHRhjHOA7wJV7eyFr7W3W2gXW2gXNzc1jb7xUlWg0TDTk0NIYz3u8pTGOU0VXPZTXqSPkmIJ5jYZM1axRUl6nBtVXkdJTXsUZpZ/gmMnX3TEP6rKDucfJXGH7T+Bx4NdDbnvTAcwZcr8FeGPI/WnA0cDjxpjXgYXAg8aYBWNto+x/Qo5hzdL5uT+AlsY4a5bO11Q2qUjhUfIaUl6lAqm+ioiU1mj9hH1Rd8fz0fF3gTQwD3gaOAM4APgG8IUx/PzTwFxjzOHAX4DzgPMHv2mt3Q3MGryfHUB+SbtfTm0GqI2GWL34aGqjIQa8NLXRUMHLviLlprxKNVFeRURKq5h1dzzTLz8AXG2tfZHMtMkua+1PgavJbG7yjqy1PnAZ8AjwAnC/tfaPxphvGGPOHn/TZSpoqIlQG8v/7KE2FmZmbXWsT5KpRXmVajKzNqq8ioiUUDHr7niu1MWBHdmvdwKzgZfIHEo+fywvYK19mLePRhh87GujPPfUcbRN9kO+H/Dyjn7+/bkOli44lJBjiIYdmmujRCLveIKGSMkpr1JtjDHEwg5zZtbiGAgsRMMGsw/WdoiIyEjFrLvjuVL3IvC+7NfPASuMMe8CPktmOqXIPuO6Pm/1usysi/APpxxONOyQDix3PfEa3YnU3l9ApMQ6+5L8+3MdXHDyYUTDDtGwQ8gxDPh+uZsmksd1ff7SM0Bnn0vIGCIhQzqw3PabbZz//d/R2ZcsdxNFRPZLnX1JXu/qJR7J9BHiEYfXu3r3Sd0dz5W6G4EDs19/A/glmTVxSeAfJt0SkSzX9Xm5u5+HnuvgY/MP4bP3PJs7oPHWC44nPO49W0WKL+zA2ce1cP73f5fL63fOOYaGuijxUJiamurY/VL2b3urrwB+WufUiYgUQ13UMLM+zrm3PZWrvevaWqmLlvBKnbX2bmvtndmvnwUOAxaQOVfuJ5NuiUhWd8Lj5sde4twT35XrcEDmHI9L736WVBWdoSRTRzqA5XflHzz+hfufp7vPozvhlbl1Ihl7q6/LPnAE4ZA+ORMRKYb+ZMCKDfl9hRUb2ulPTv7DtHf86NgY84OxvIgxBmvtpyfdGhHADyxLWuews98rfDBuoEGdVJYgsKTSQcG8zqqPKrNSMfZWXyMhw+z6WJlaJyKyf0uNcvh4ah/0E/Y2H2j4yYh/BwTA77P3jyZzte+3k26JSFbYMTTVRenu92hpjOeFP3OQsz5Flsqyx01ijCmY15BjdO6XVIywYzhweg1/3eMWzGvYMYQ1x11EpChCzuh9hcl6x8ptrT1r8AY8QeY4ghZr7d9Za/+OzGHivwR+N+mWiGQ1xaM0T4uxsX071y/JP6Dx1guOR/1jqTT9yYC/7nYLHigaCzs0xbVFvFSGpniUpvpowfq6vq2VmAZ0IiJFExnl8PFIiQ8fvwI4zVrbP/iAtbbfGLMaeAz450m3RgSoqQlzAHDFaUdy02MvsWrRPJrqosysi7Lu8W189kPvKXcTRfL4geVfHn6Br589L+9A0eZpMWbWRLVJilSMmpowIdfjwlMO54dPvJZXX+tjIayOHhcRKZq0tcSHHT4ej4ZI28lPvxzPR3L1wMEFHj8IqJ10S0SGqKkJ01wf5RMnvouGeITufo8r73+eJ17t1lQ2qThhx9DVl+SfHtyCN2TnwGkx7XoplSew8MMnXmNJ65xcfb3uFy+QSlua6nRVWUSkWKyFW3/9Sq6v4KUDbv31K+yDMd24rtRtBO4wxlwFPJV9bCFwPfDTyTdFJMN1fboTHmlreXdzHf/88y08uqWTlsY4a9taNZVNKk5TPModF51Ax85E7pO3lplxZsQi5W6aSB7X9XGM4R8/9n5e3zHAdb94ka6+ZK62OvrQTESkaJriUa4643107MysqYuGHK464337pG87nkHdSuDbwJ3AYE/FB/4N+NKkWyJCpsPRsSfB9iGd4y+f8T6uPfsorM38MejKh1Si4V1hdY2l0gyvrwBrPj6fabEwM2IR1VYRkRIoVn9hzBXcWpsALs1eqTsi24ZXhq6xE5msPZ5PV2+SVT/7Q+5QxjVL53NEcz2zp9eUu3kiBY2W2+k16ihL5VBORUTKq5h1eNzbXFlr+621m621z2tAJ/ualw646oHNeYcyXvXA5rx1SiKVRrmVaqCcioiUVzHrsPYuloqSHuVQxiCweJ5fplaJjM51/VFzm9ah41JBVF9FRMqrmP0FDeqkokSyhzIO1dIYxxhDX0qdDqksnufzcnc/b+52C+Z2X5w7I7KvqL6KiJRXJOQU7i+EJj8k06BOKsrMeJT1ba15hzJ+++PHsPqhP5LwNEVIKktXv8fKDe1c/4sX+fbHj8nL7bq2VmZqp1apIKqvIiLlFXYM3zknv7/wnXOO2SfHdWlltFSUmpowM+sirFo0j4Z4hF2JFNf94kU2bd/FNWfOK3fzRPL42WkUHT0JrvvFi7ncHtIYpzbqaPMJqSiqryIi5ZVIpfmXh1/Mq8P/8vCLfPe8Yyf92upxSMUYPJ/ODyzRkJPrbEDmkwwdOi6VJuwYTp83O3eI865Ein/7j1f5+llHUR9ReZXK4Lo+ibTPgBeovoqIlFHYMTRPy5/F0zwtWn1X6owxZwA3AiHgdmvtdcO+/0XgEjLn33UBn7bW/qmUbZTycN3M2qSVG9pzW7zeesHxbHjyTzzxajdr21pprtNUNqksTfEoV5x2JCuG5HZdWyu1MYdAs9mkAriuT0/SY0dfSvVVRKTMRus3lPrw8UkxxoSAW4CPAB3A08aYB621W4Y8bROwwFo7YIxZCXwLOLdUbZTy6U5k1iY118dyl6S7+zw+/5G5XGmOpDEeIRrVlQ+pLHuSKbp6k3z748ewK5Fi3ePbWLGhnXuXLWQAj0M0/VLKrDvhEVhyA7rj5jSw4tQj8PyAKz48ly+H38v0WFj1VUSkBHYmPG567KW86Zc3PfYS1551FAdPss9Qyip+IpnDyl8FMMbcCywGcoM6a+2vhzz/KaCthO2TMvIDS3N9jC/9/Xu5euPm3KcXay84ntnTYupwSMUJAktnn5d3gOj1S+ZzwyNbdZSBVAw/sFjIDeiG19j1ba3MPChW7maKiEwNBi485fC8Onz9kvmwD2bAl3L3y0OA7UPud2QfG83FwC+K2iKpGGHHcMVpc3Mhh0wnZOXdz5JSB1kqUFdvMjd9AjJ5vXrjZq44bS4hx2iNklSEsGMImcxRBitOPWJEjV2+oZ2u3mSZWykiMjVYy4g6fPXGzdh90NUt5aCuUA+n4FswxrQBC4A1o3x/mTHmGWPMM11dXfuwiVIOqVSahrjDYbNqRz0Yt5opr/sn108XzOu7mmoJO2afzI8vB+V1/5FKpamPOYRDZNZs1EULZjbpp8vUwslTXqWaKK+yvxw+3gHMGXK/BXhj+JOMMR8GrgHOttYW/PjQWnubtXaBtXZBc3NzURorpZFKpelxPXoSAbFw4QMZQ1V+xUN53f/4fkBolIOcIyGHhlikao8zUF73D6lUmhc7+/jtS13YAGZPj3LgjJqCmXWquMYqr1JNlFcZre+wL/q6pRzUPQ3MNcYcboyJAucBDw59gjHmOGA9mQFdZwnbJmUykErx1h6Pu554DccY1l7QWpQDGUX2pZ0DHtGww60XHJ+X11svOF7n00lF6OxL8nrXHg5rns61//5H/tydIGQyV+yGZnbN0vlEVGNFREoiEnK48bxj8+rwjecdSyQ0+SFZyXoe1lrfGHMZ8AiZIw1+YK39ozHmG8Az1toHyUy3rAd+YowB+LO19uxStVFKry8ZcPNjL3HZh+aydN2TNNfHWL34aA5tqqWrN0ljXRRTyo8eRMYg5MAbu1y+96uXWbVoHk11UWbWRQk5kE5X93Rh2T/4geW4dzXxjX//Ixf/7bv5/H3P0dGT4PR5s/nRp09kdyLFroEUzdNizIhFyt1cEZEpIRyC6fEIqxcfTW00xICXZno8Qji0D1578i8xdtbah4GHhz32tSFff7iU7ZHycl0fP7AsaZ3DpXc/S0dPgo6eBBfd+TQtjXHuvOhE6mIhqni5h+yHgsDipgJWZjP76JbMpIKWxjj3LVtIwtcBdVJ+9bEQfck0S1rncOVPns+t4Xh0Sydb3uzlrotPpD4Wpi4W0pVlEZESSXgBF93xdN66usH+A3WTe21VcimLwcPG+5P+qIv3u/uSREM1HDwjPsqriJTeHjeJP8pCZz+wmi4sZee6PjsHUhjMqPW1c0+SkGOoi6kbICJSKu/Uf5gsTWyTsuhOeNz82EvURByap8UKLhod8NKEHUM4rJhK5UimMgO30RY6V+uul7L/2J1MsaM3ybd++QKz6kevr83TYjTXKa8iIqUyWv9hX3wgrI/opCyMgav+/n3sTqTYnUhx50Un8Kns5ejBxfvT4xGMLnpIhUkFlk1/6ubuS06iqzdJd7/HxvbtXH7akUyLaZMUKT8LuKmAi//23exxC9fXpvoo8YhDJLIPFnKIiMiYxKMOd1x0Ah07E7k1dS0z48SjVbRRisigVCpNT3+K5dmDm1sa43z33GP58WdOIrCZ7V67el1u/P9f4utnHVXu5orkeJ6PY2BmfZwLbv9dLr/r2lo5qCFGXzJgRm25WylTWSqVZkefx6qf/aFgfXWMYUefyw2PbFV9FREpMTcV0Jvw82r0TecdR3108h+waV6blFxnXzI3oIPMXOLP3/ccW97s5YLbf8cf39jD5T9+jstPO1JT2aSi7E76BAFcvXFzXn5XbGgn4QXKq5RdZ1+SFQXq6yud/bza1c//861fq76KiJSJtXDFvZvyavQV925iHyyp05U6KS3P80ddJHrg9BpuOf94wiHDfcsW0hSPaiqbVAzX9QFI28L5tdYqr1J2wSj5PGxWHX/dneCBFSdz0Iwa1VcRkTJIj9IHTmujFKkmnueztSvzaXGhRaKz6qMkvBRn3vQf+IE6yFJZfHw8PyA06iYpKqdSXq7rE1gK5jMaMvS5KZaue1L1VUSkTCIhp2CN3heHj6sXIiXT1e+xckM7Nz32MtcvmZ8LdUtjnOuXzOef/v2P1NVEuXbR+4hoW3ipIK7r81p3knNve4rL79nEmqX5+V3f1qpdBKWsPM9nR8Ljn3++ZdT6emBDreqriEgZ1UQMa9ta82r02rZWaiLa/VKqxOBB44MHjN/wyFbu+cxCOve4dPd73PDIVjZt38WWN3u5b9lCZmqth1QI1/XpTmQ+kBjM77d+uZXVi4/m8Fl1bH2rl8a6iHYRlLLq6vfw/IBHt3TS1eupvoqIVKABL+Dmx15i1aJ5NMQj7EqkuPmxzMaAjTp8XCqd6/rsTqYIO4YHVpxMd7/Huse34acDlq57Mu+5gwcwamqQVALX9dmZ9EasA920fRcX3fk0v7ryA6x+aAv3L1tYxlbKVJdKpYmEDBYnV2PTgeqriEil8QPLo1s6eXRLZ97j15w5b9KvrcouReW6Pjtdj6Qf8PqOAW567GW6+pLccv7xubVJQzvL++oARpF9IWl9Io4DoYA7PnUCtdEQuxIp1j2+ja6+JIGFdW2tzNLUSykjL50ikQryzk285sx5qq8iIhUmEnI4fd5slrTOyV2p29i+fZ+sqdOgToommfR5pbs/t7324NqOGx7ZymfveZb7ly9kXVtr3vfXtrVqm22pCK7rk0xZ/MDyl55E3pkya5bOp3lajMbaMJFwmGhUpVTKw3V9Xu9O5p37ef2S+dzz1Ousb2vNe1z1VUSkvGoihss+NJdL7342V5tvveB4ramTyuW6PjsT3ojzkq7euJlVi+ax/K523tjlctCMGu5bthA/sIQdQ23U0dQgKTvX9elJeqSDzJkyX7j/+bwcX/XAZu5btpABL01zOFbm1spUNbjec/i5n4N1dno8zL3LFpJWfRURqQgDXpAb0EGmZl9697Pcu2whMye5pk67X8o+Nw2eGMYAACAASURBVNjRSI1yFkdDPEJLYxw3lebN3S4WuOD232GB+mikLG0WgUx2/7o7kRvQdfd5o54p4wcWxxh1kqUsBuvsaOd+NtVFsRbOu+0pUmmLBeIhZVVEpJxG61ME++CcOlV4mTTP8+nq93AcsAEk05n1cw214YJrOga8NN855xjA0N3v0Twtxtq2VmbFo9pBUMrGdX3e6k+yayDFzLoIf9nlYoBpNWFOnzc7b1FzS2OciGO0i6CUlOf57EykiIQMfck0uwZSNNZFC+ZzZl2USNihoydBJGSYpcPGRUTKLqo1dVKpXDdzNpK1lp4+P3993AXHc8v5x/HZezblrelwvTR+EHDgjBjf/z/buPaso5jbVKcOh5SF6/q4gc+OvhQ9/SkOmhFjV8LnSz95Pi/LAI9u6aSlMc66tlZmqpMsJTRYa8HS3efnrZW7tUA+1z2+jZUfPCK3OYqyKiJSfrGI4fLTjswdkzTYN45pTZ2Uw+C0n5qoQ38ywPMDYmFnxPq5lXc/y3X/47/lzuIY8NL0uSk+8f3f8cCKkwG44rQj1TmWkvI8n+5ECmst1kLaWmpCDtGww8z6KK5vC2b5x59ZyDVnziMSMtTFtDZJiiuZ9On3fVK+JRVY0oHlzd0ufjrgKz/9/Yj1GHd86gSW/d0RzIhHcFNpnni1m2UfOEKbo4iIVBA3ZXMDOsj2MTa0c//ykyf92iXtlRhjzgBuBELA7dba64Z9Pwb8CGgFuoFzrbWvl7KN8vagzZjMJhGOA0GQmQcccgyRkCEaNvx1d5Lld2WC+cCKkwvOEY6EHJbf9V+5x+5btpCWxjjN02LUR0McVF+jzrHsU57nkwp8+pKWaNgw4AW5jXhqwg79qTRv7U5SE3FYefezNNfH+PIZ7+WqBza/Y5YDa4mFM9PeZtWqkyyTl0ql6UulSKch6b+d03jUwVroGUjR1ZvMZbOlMc6PPn1iwXzu7Pc497aneGDFyfS6PmvbWmmIh2mZrhorIlIp/CAYZZ1+MOnXLlmlN8aEgFuAjwAdwNPGmAettVuGPO1ioMda+x5jzHnA9cC5E/2dQxeSxyMh0oEllQ5yA5NUOvPpZyTk5O5bLFhIZQcwUceAA37209JYODPn1drMJgmpwOIYCJnMZdPBT1QjjiESdhjw0kQcQzhksDZz6GDaWsIm8/2kH5AOLPGIgx9AOghwjMkNqIyBwEIs7JDyA3xrCRmDk308EjL4gc08L8i8fsQxhB2D62feazzqMOAFOCYzl3ew8xAJOYQNuecZA/GIk9cJTltLR7fLlUOmot150QnZwdrbnzR093sF18/tSqTy7g94ada3tTI9HiLuhNXZGGJwx1CTXZs4+N/AcQxBNm/J7FXRwII/mOXsf+vo4ONBQE04hJ8OiEUc3FSQt7vogBcQWEvYcYiGMhn2s7mNhhycbCbiEYfsn0Pu+4Pt8fyAmkgIa21e5snmdlqNQ68b5HKcDixOtq01EUPSt3jpt38uGnbo99LUhB0smQ5uKNveRDaPNdn356VHfi/iGGJhh2Q6gOzfWThk2NHvc9EdT+dNU5seD5O2lpXZ3adWLZqX6zTD6Fke/P/90Ia4jjBgWF6tIZUOCGdrT4AlCCCwmToTDZvcf6uwk6lfaUu2Tma+V5PN6tAPkyKOkzfYqQk7YCCRytbNsEM6m4mh2W2oDdHvBm/X8ZBDTRR6E2+/1mDNr4k6JL0gl93BnIdDg8/JvP5gVsMhQ2Az9XdovQ4ZQ9rC4B+Nn/3dg+8nGs60z0+//TcTDTmEHOjs9fKmrt9x0QmEHcP2nW8fpQGZf/j/1D0waq0d/MBsTmOcaNhQGw4TiymrItVmaP817BiaNKNpvxEypuCausFxxGSUMiEnAq9Ya18FMMbcCywGhg7qFgPXZr9+APieMcZYa8e9JYzr+rzc3c/KDe0018f46qL387l7n6OjJ8Hp82aPOCNi7QXH89Dzf+FD7z8wbwCzZul8ZtVHWfPIVrp6Pb666P2s/802Lv3ge0h4ae74z9e49IPvIeIY9rh+3ieqay84nrue/BNPvNrNHZ9awB7Xz7VhsIP5vV+9TFevl3elYPCcoR8+8RoXnnI4v936Fmcec0hee69fMp/fbn2LRcccws2/epkLTzmcqzduzmv3t365la6+JGvbWml/bQcL3zMLz7d5rzP0eXddfALd/akR83xfeGNXXqdi+87/2969x8lR1vke//z6NtOZhGRyQ8wkEBDUqAnJjIA3RFAP6wWOJwGjyR5ABAnrZVFZXHePevS4LzHLIiAkwAFBkxUwWdcssOJZEPEuSSBRQTAEJANorkCSmUnfnvNHVXWqbzPdyUzP1Mz3/XrNa7qrnq56quvbT9fTVf10LxNaEyUHFSsffIorF84tqcPKpZ1ce/+TAMX7rUmvU6wOXam+Pu83/a69/0kufOuxJRm8/sPzMTMurXJWKdiH39/4HB9YMIPL12wulnnihZfonD21Yn/e/Wg3N/70GTra09x6fhcv9WSLQ/Z3tKe5ZvGJ3PmbbSw5ZRaJmNGTyZfMD9a35JRZ9GTyJXW56px53PKzrXzijBO4+9FuTn31kSWZuPrceRw5sZXn/J8FCL9WHvzDdrpmTy6dvrST6+5/suprJJgXfI/o1vO72NeX55N3PFJS12njW+je01u8TO22C05i6vhUMb+T0skBs3zTX3cyqSWpzPrCeb341OP42zsfLclrX7ZQzPC750yv+P5A0L5d9LZjaW9L8bMnt9M5eyrX3f8k5715Nrf/4mk+/e4TyOYOdr7DnfJ/uufxqplYvmguhmNvX6qkk3TLeZ1k9lJShxuWLOCeTc/xvhM72PD0To5/xcSK9uuIdKIiqyuXdmI4PrZqY0nuW5Mxrv/xlorXb7Ct//De15LJFdi5L1OR/ese+GNJG9vtt7HjUvGKT3Svvf+PrFiyoOR5Cdaxcmkn6aT3Ic/ElhZ16EQiKHz8Gn6/09gDo0MyXv136pLxw+/UNfMnDWYA20L3u/1pVcs453LAS8CUQ1nZrt5M8QVxyWnHFTtTAAs7Z1b8RsSy1RtZ1DWr+GYcTPfefPtY2DmzuJyFnTPZsz/L5Ws2F29v35sp+bQ/WOZFpx7rH1D2ldQhOMAMllv+2CvWesu+Yu1mFnXNqqhvMH2Zv4zgYCRc70tOO654re7pc44iEYtXLCdcLhGLV73O9/Q5R5U8t+NS8eLZjMAj217k9l88zW0XnMQDn3k7Xzn79WTzeT500tE8dPlp3HnxKcyY1EI6GafjiLQapjK7/N/0W9g5syKDu/dni/utWlYuX7OZi049tjg9KHP6nKOq7s9FXbOK95/b01fxG2yfuuNRLjr1WHb7ua72G23B/PK6fOZ7m1jYObO4nvJcXnbXJjI5V/W1cvaCjsrp/nNSbbuDeeFtCTp04bpectpxxee5e09v8cx6kN/gDEcgnOUff/Y01lzyJo7Tm2mJcF6DDh0czGs4w0EeqrVvl921iW27e4tZDdqyhZ0zvfaoym/55PLUzMTlazYzc3JbxXci41Xatkv9Nj9o48qzesmqDVWzesmqDWzfm6nI/e792aqv32B7snno3tNXNftBjgNBG9uTyZdkE2DHvgP0ZQus/ujJPPhZr22dPXUcX3z/6+hobwFiTB/XoryKRFT4+BUOvt/t6s0Mc81kMGTzrurv1GXzh/+TBs3s1FXrgpZvQT1lMLOLzWy9ma3fsWNH1ZWFf7un/JP48vsQvPFb1enjUnEmpZPFx01KJ4ufoAa3q32iGiwTqDk/vNxq8/qrVzC9v8cHt73LRem3XN5V/+2M8hOlPZk8azds46pz5hUPODra01zwltkcyOX52n8+TioRY8WDTzF9QgutqTjJeIx0PMGM9nFj7mCjkbxW25fh7NST3aBMocb+DDJZvuzyMgPlur9M95fbWjl0Neo70GtkoG0Jl+loT1Nw8OeX+7hy4VzvzMuDT7F80dySLH/yjBM4ojXBUeNbeMXE9Jg64zGYeYXamQ2mj0vFi1kNt7G1chIz+s1Etd9tq7WsIKO1Xiu1HjcuFa86rb9tjVntjE5pK/2OZtDGtrclK7J549JOjpzYQl82R65Q8H5KxuFdnhpLMH2MfYeunryKjBSNHr8GgrZNoq/W79TlG78osUIzO3XdQPjjyA7g+VplzCwBTAR2ly/IOXeTc67LOdc1bdq0qitLxGp/El9+H7w3y3zBVZ3ek8nzYm+2+LgXe7PFT1CD29U+UQ2WCdScH15utXn91SuY3t/jg9tm3vdA+isXPnsRnh8PPZcd7WmmT0jxyTNO4JafbeVr/+MN3P/pt/Pdi07h6CnjmJhO8qX3v45jp47jS+9/HcdMaeGI5Ng70AhrJK/V9mU4O/VkNygTq7E/86E3hv5yO1Cu+8t0f7mtlUOrUd+BXiMDbUtPJl+87V3iAN/6+dP8831P8JWzX89V587j2Klt3HnxKfzk8tO462Nv4lVT2sZsZgczr1A7s8H0nky+mNVwG1srJwVHv5kIt/2BWssKMlrrtVLrcUGmyqf1t60FVzujk9tSJW1sx+Q0H3v7cdzw4y0AfOfCk/jJ5aex+qMnM3FckrgZ7ekURx6RpDURY0o6xZETx+YVEPXkVWSkaPT4NRC0bRJ98Rr7dzC+U9fMTt3DwPFmNtvMUsBiYF1ZmXXAef7tRcADh/J9OoAp6RQrlnYWP4m/ZvGJxSdx7YZt3LBkQcmb6IolC1iz/tmKs0/ep6StrN2wrbic8Ceowe3pE1IVn6iuWLKAmx/a6r1Jt7eW1CE4wAyWW/7YKxd6y75y4VzWrH+2or7B9BX+MoKzDuF6r3zwqeK12A889gK5Qr5iOeFyuUK++JwVt2FpJ6mEsfqjJ/P9S9/MV85+PS3JOEdNauGL738dMyePozUZoyVhGFb8QeajJo3jle3jaGsdmwfGjZqSTrFyaWfVM6CT25LF/VYtK8sXzeXmh7YWpwdlHnjshar7c836Z4v3Z7S3cvW5peu7ZvGJ3PzQVib7uS6fH6xvcpWzCFedM4+1G7YV11Oey6vPnUcqYVVfKz/Y2F053X9Oqm13MC+8Ldcunl+xLa85agI/8S8BfuXEVtLJOF98/+u45kMncvyR4xnfEsfM+yL60VPaeOWksXlw3IhwXr/xwRMr8hrOcJCHau3b1efOY+bkdDGrQVu2dsM2rz0qa69uWLKARJyamVi+aC7bdu9nZdn68lXathv8Nj9oH8uzutJv+8rXsXJpJ9MnpCpyP7ktWfX1G2xPMg4d7a0Vy/vGB0/kzt/8idUfPZmfXH4a3/7ISYxviTOzPc0X3v86ZkxK0xL3BjVKJWJMaPE6cdMnppnQ2kp7m9pYkdEkfPwKB9/v9NMko8NQ7l87xD7Toa3M7D3AN/B+0uBW59xXzezLwHrn3DozawW+A8zHO0O3OBhYpZauri63fv36qvOC0YPyBUdrnaNfgiuOUhkrG/0yV3Ckqox+GTeIDTT6ZcKKIxqWj35ZKDhay0a/DEa3DEaJq3f0y+LIbVVGv4wbJPsZ/TJm0Fo2+mUwtHafP9pc3B9lsCUJ6USKRKKZnwsctmH/mGugvNYa/dI5h5k36mRqgNEv84UCLYcw+mXBz0TMvNEnWxsY/bLgzwtGaz2U0S97Mnla/NEvM7kCsRqjX2bzlfOqjX4Z818HMaAlZbQllddG1Z1XZ8U89jf6ZdCGjMTRL4PsDtrol2Vtdfnol8HrDVzxuSg4VxztOIJt7IjOa+CYz93T0DKf+dp7D6dKMnKN6Lxq9MvR7RD2b115bWpCnHP3AveWTftC6HYfcM5gra+1NcEMvQgAmNxWf9n2BsrK4GltTfDKJuS1Wfv3iPTAZerSX32V1WHTcF6bvB8nVsnfEa3NW7+IyKHS8evoNlT7NzIfAYqIiIiIiEglfQwgIiIiUoUu1xSRqNCZOhERERERkQjTmToRERGRQaAzeyIyXNSpExERERkGjXYCG9Vop3GoO6Xq9IoMnab+pMFQMLMdwJ8GKDYV2NmE6owkY22b69nenc65M5tRmVqU16rG2vaC8hpl2t7qlNeRSdtbnfI6Mml7q6srr5Hv1NXDzNY757qGux7NNNa2eTRt72jalnqMte2F0bXNo2lb6qHtjbbRtj0D0fZG22jbnoFoew+PBkoRERERERGJMHXqREREREREImysdOpuGu4KDIOxts2jaXtH07bUY6xtL4yubR5N21IPbW+0jbbtGYi2N9pG2/YMRNt7GMbEd+pERERERERGq7Fypk5ERERERGRUUqdOREREREQkwtSpExERERERiTB16kRERERERCJMnToREREREZEIU6dOREREREQkwtSpExERERERiTB16kRERERERCJMnToREREREZEIU6dOREREREQkwtSpExERERERiTB16kRERERERCJMnToREREREZEIU6dOREREREQkwtSpExERERERibDId+rOPPNMB+hPf/X8DTvlVX8N/A075VV/DfwNO+VVfw38DTvlVX8N/NUl8p26nTt3DncVROqmvEqUKK8SJcqrRInyKoMt8p06ERERERGRsUydOhERERERkQhTp05ERERERCTC1KkTERERERGJsESzVmRmtwLvA7Y7515fZb4B1wDvAXqA851zG5tVv9Ekm82zpzdLJl8gZuActKZi5HKOWMzI5ArkCo5kPEbCoC9XIB4zzCCdjNGT8eYnYkY6FSNXcGRzrjitNRnDOcgXvGm5giMeM1oTMbJ5RyZfIBkzEnHDOcgWHPmCIxkzMMCvT1+mQN45YuatO27m1T9UPp2Kkc1DJl8oTmtNxYjHYF/fwXqOS3n1njouRUtL02Itg6yvL8dLB7I4YFzKyOTggJ/X1kSMgvOyEPf3ea+f1WTMaEnEOJAvgINcwct6uEzCL2MGfdkC5r82Cs6RinvLzvrLTsaNbN7LYTxmtLXE6DlQIOvfTydiZPIOhwM/4/GYFXN/IHuwbCputCRNeR2Fstk8ezNZsjlHIm7FXAUZDNpV5yCViBEzyOVdsY1LxWM4HHEz8gVvesLPWybnZRO8ZRX8efhtpQMyucq8xkJtcbZQIB5qX3MFV6xPvuBIxGO0JIzebP/ta1tLjHQ8Eems9vXl2N2bKe6X8HOWjMeIx4zebL7YtvRlHdl8odju5AqF4ntVa7K07UnEDGKQy7mDr/uYkW4xXu71yo1vidOXPficTkmnaG1NlLR5OT8XCX+ZqYQxoSVFLldgV2+mpM2KmRGLQaEA09pSpFKV+yabzbN934GD+9x5WZs+voVkMl7zuTpwIMfOnkyx3XV47XDcf09uiyfY5T+XiZgxdVyKeDzG9n0HyOYLJOMxJrcmS8pMa0tRKFAybUo6RSJR+rjp41tIJBo/3xCuc1CnKOdV5FA0M/G3Ad8Evl1j/l8Bx/t/JwMr/P/SgGw2z5/29LBj7wG+9fOnOe/Ns/njn1/iLSdMp1AocCDnuHT1Rrr39NLRnmb5orl8/YdPsGPfAb5z4RvZtT/LslUbivNXLO1kcluCD9706+K02y54IwB79me47K5NdO/p5d1zpvPx048vWfbKpZ3kCwX+5l8fKU676px5PPD4n3nfiR0l6/nmh+eTjBkv9+W4fM3mkmUYjo+t2liy/r5sgUvK6vnMjpc5ZtoRnDC1TY15BPX15dj2ci89B3IcNamVPb15Xu7Jsmz1RqaNb+Hvznx1STZWLO3kuvuf5EePbaejPc2t53exry/PJ+94pGaZG5YsYOK4JP/6y2c49dVHcsXazRXLrpblasu5Z9NznP7aV/CZ720qeT1NndDC8h/+oVj2+g/PJx6LKa+jTDab5897+4odgZ4Dea574I9c+NZjSzJx5cK53P6Lp/mH976WTK7Azn2Zkhxfu3g+8Rgl7eS3Lngj2XyBbK5ATyZfmvslC+jL5ottb5DHbz7wx5J8hu9/88PzyeYK3PzTrZz35tlcsbZ0edeFytZqX488wjERIpnVvr4cW3btL9mm8uco/F4YzNuxN1PR7ly7eD4T0gku+NbDxWk3Ll2Aw0qWv3zRXKZNaOHrP/wDk9Iplr7p6Io25fgpbcU2r3w/X33uPCa1pYjH4U+7DnDd/U9W7LsgW5844wRePa2tpGOXzeb5w/Z9VR+3Ymknr5k+vmrH7sCBHE/u3M+yVRuqtrvXLD6RI9LJku1fsbSTiekEH77ZO0740vteQ+fsqSXv8f926Zv4y8uZiuOLI49Ice6Nvyx5z3/NkRMa6tiF6xxe9mhoW4/53D0NlX/ma+8doppIFDTt8kvn3EPA7n6KnA1823l+BUwys6OaU7vRY/u+A2zb3cvlazazsHMmV6zdzOlzjqJ7dy/xWLz4pgLQvccrd8lpx9G9p5dELF5sFIP5y1ZtoFCwkmnbdnt/wUEFwMLOmRXLvmTVBnbvz5ZM+8z3NrGoa1bFevbsz7J978GDnfAytu/NVKz/kir1nH/0FJat2sDOnkwznmoZZLt6M3Tv7mX3fu/MRzbnWOZn6pLTjqvIxrJVG1jYObN4/7k9fcUOXa0yl67eSDbnWNQ1q3iQU77salmutpxFXbOKB+/B9MvXbKZ7d29J2d37s8rrKLR93wEO5BzP7enDOWPZ6o0s7JxZkYkr1nptcTYP3Xv6KnL8yTseqWgnu3f38peXDrB7f7ai/M59mZK2N8hjeT7D9/fsz3LZXZuK7wklWSwrW6t9zeRcZLO6qzdTsU3lz1H4vTCYV63d+eQdj9C9u7dk2va9lcu/fM1mtvltwUWnHlu1TQm3eeXrueyuTXTv7mVvb6HY/pTvuyBby1ZtYMf+0n2zfd+Bmo9btmoD2/cdqPpc7ew52PGqtv2fuuPRiu0P8hFMO33OURXv8Zmcq3p8EX5c8T2/Rt1qCdc5vOyo5lXkUI2kjzBmANtC97v9aS+UFzSzi4GLAWbNmtWUykVFruAYl4rTvaeXSekk3Xt6KThvWswoNnqBoBx4l2ZUmx9cAhQYl4oX5wWCdZU/NigbnhaPWUXZasustYxg+8rL5Qte/XOFun+nsSmU1/oE2Q1uh/NaK19BdqF2LsJluvf0EjPArOay61lXrRwHeR3Hwcwqr6NTkNFw29pfdoKy9bZx4fnl8+rJZ7XXxuG8jvIjLKeBevKaK1R/b6v1nJXfLn9cve9JQVtQq60It3m1Hh/Uvb99V60dGehxtdqd8HPVyPt6zA7eL1Q5lsjX2Afluere00suX6hat1pq7d+R1raC2lcZWiNpoBSrMq3qK9I5d5Nzrss51zVt2rQhrla0JGJGTyZPR3uaF3uzdLSniZk3reCgoz1dUj4oB953LqrNj1nprunJ5IvrCATrKn9sTyZfMS1fcBVlqy2z1jJqlYvHvPonYtWiNHyU1/oE2e3J5EnErCSvtfIVZBdq5yJcpqM9TcH/PlGtZdezrlo5DvJaT72U12gLMhpuW/vLTlC23jauVptYb86rZfBwXkfx4LtjI0w9eU3Eqr+31XrOgtv1vq/1t19f7M3WbCvCbV6txwd172/fVWtHBnpcrX0Zfq4aeV8P959iVY4l4jX2QbysHh3taRLxxg5Na+3fqOZV5FCNpE5dNzAzdL8DeH6Y6hJZ08e3MHOydz3/2g3buHLhXB547AU6JqfJF/LcsGRBsfELrvtf+eBTdLSnyRXyrFjaWTJ/xdJOYjFXMm3mZO/v6nPnFaev3bCtYtkrl3YyuS1ZMu2qc+axZv2zFetpb0syfUKK5YvmVixj+oRUxfpXVqnnI3/axYqlnUwdl2rGUy2DbEo6RcfkNJPbkiQTRjJhrPAztfLBpyqysWJpJ2s3bCven9HeyrWL5/db5oYlC0gmjDXrn+XKhXOrLrtalqstZ836Z7nqnHkVr6eOyemSspPbksrrKDR9fAstCWNGeytmjhVLFrB2w7aKTFy50GuLk3HoaG+tyPG1i+dXtJMdk9McObGFyW3JivJTx6dK2t4gj+X5DN9vb0ty9bnziu8JJVksK1urfU0lLLJZnZJOVWxT+XMUfi8M5lVrd65dPJ+OyemSadMnVC5/+aK5zPTbgpsf2lq1TQm3eeXrufrceXRMTjMhHSu2P+X7LsjWiqWdTGsr3TfTx7fUfNyKpZ1MH99S9bmaOi5VfH+utv3XLD6xYvuDfATTHnjshYr3+FTCqh5fhB9XfM+vUbdawnUOLzuqeRU5VOZc805Pm9kxwN01Rr98L/BxvNEvTwaudc6dNNAyu7q63Pr16we5ptEWHv0yblCoc/TLmD+qV63RL4ORAMtHv6wYcS1f8EbvCo1+GR65rdrolzGjeEYwXH6QR78c9o/tlNf+1TP6ZTZfqBjZcrhGvwSHC60vPPplMO0wRr9UXke4fke/9EcajPltcPnolwW/DQavDRzK0S+D9jW4ZDQ4Wz3Io1+O6LzWM/plXzZfbDdqjX4ZM2gZwtEvC/7jh2L0y+BxhzL6ZSbntbsDjX6ZyxdIHMLol8Hjmjj65YjOa0ADpYivrrw28ycNvgucBkw1s27gi0ASwDm3ErgXr0O3Be8nDS5oVt1Gm2QyzvR+GuyBtLcNYmX6c5jrmVh6tUXz6i1DprU1QWtrA81SPft8kHIxadzhPV55HX2SyTiTD6OtHcnK8xp1ra0JXtlI2zKQOl+/E1r7n19Pm5dIxJhxCHVPJuPMaG+84WppSTBjgFEjq9XnlZPSA5ap53GHop46i4x2TXsFOOc+NMB8B/xNk6ojIiIiIiIyKoyk79SJiIiIiIhIg9SpExERERERiTB16kRERERERCJMnToREREREZEIU6dOREREREQkwtSpExERERERiTB16kRERERERCJMnToREREREZEIU6dOREREREQkwtSpExERERERiTB16kRERERERCJMnToREREREZEIU6dOREREREQkwtSpExERERERiTB16kRERERERCJMA125xgAAIABJREFUnToREREREZEIU6dOREREREQkwpraqTOzM83sCTPbYmafqzJ/lpn92MweMbPNZvaeZtZPREREREQkaprWqTOzOHA98FfAHOBDZjanrNg/Anc55+YDi4EbmlU/ERERERGRKGrmmbqTgC3Oua3OuQxwB3B2WRkHHOHfngg838T6iYiIiIiIRE6iieuaAWwL3e8GTi4r8yXgR2b2CaANeGdzqiYiIiIiIhJNzezUWZVpruz+h4DbnHNXmdmbgO+Y2eudc4WSBZldDFwMMGvWrCGprMhgUV4lSpRXiRLlVaJEeR3djvncPQ2Vf+Zr7x3U9Tfz8stuYGbofgeVl1deCNwF4Jz7JdAKTC1fkHPuJudcl3Oua9q0aUNUXZHBobxKlCivEiXKq0SJ8ipDqZmduoeB481stpml8AZCWVdW5lngDAAzey1ep25HE+soIiIiIiISKU3r1DnncsDHgfuAx/FGufy9mX3ZzM7yi30GuMjMNgHfBc53zpVfoikiIiIiIiK+Zn6nDufcvcC9ZdO+ELr9GPCWZtZJREREREQkypr64+MiIiIiIiIyuNSpExERERERiTB16kRERERERCJMnToREREREZEIU6dOREREREQkwtSpExERERERiTB16kRERERERCJMnToREREREZEIU6dOREREREQkwtSpExERERERiTB16kRERERERCJMnToREREREZEIU6dOREREREQkwuru1JnZHDN7dej+u8xslZn9vZnFh6Z6IiIiIiIi0p9GztTdAswHMLMO4AfAZOBvgP8z+FUTERERERGRgTTSqXstsNG/fQ7wa+fce4C/Bj402BUTERERERGRgTXSqYsDGf/2GcC9/u2ngCMHs1IiIiIiIiJSn0Y6db8DlpnZ2/A6dT/0p88AdtazADM708yeMLMtZva5GmXONbPHzOz3ZvavDdRPRERERERkzEk0UPYK4N+BzwK3O+d+608/C/jNQA/2B1O5HngX0A08bGbrnHOPhcocD/w98Bbn3B4zm95A/URERERERMacujt1zrmHzGwacIRzbk9o1o1ATx2LOAnY4pzbCmBmdwBnA4+FylwEXB8s3zm3vd76iYiIiIiIjEUN/U6dcy4P9JnZ6/2/tHPumTo7XzOAbaH73f60sBOAE8zs52b2KzM7s5H6iYiIiIiIjDWN/E5di5l9A9gNbAI2A7vN7Boza61nEVWmubL7CeB44DS8ETX/r5lNqlKXi81svZmt37FjR72bIDIslFeJEuVVokR5lShRXmUoNXKmbgWwCPgoXsfrVf7tDwA31PH4bmBm6H4H8HyVMj9wzmWdc08DT/jrKuGcu8k51+Wc65o2bVoDmyDSfMqrRInyKlGivEqUKK8ylBrp1J0DXOCcW+2c2+r/rQYuxOvsDeRh4Hgzm21mKWAxsK6szL8D7wAws6l4l2NubaCOIiIiIiIiY0ojnbr9wHNVpj8H9A70YOdcDvg4cB/wOHCXc+73ZvZlMzvLL3YfsMvMHgN+DFzunNvVQB1FRERERETGlEZ+0uA64Itmdr5zrhfAzNLA//LnDcg5dy8Hf7Q8mPaF0G0HfNr/ExERERERkQE00qk7BXg78JyZbfanvcFfRpuZFS+ldM6dVeXxIiIiIiIiMsga6dTtBNaWTXt6EOsiIiIiIiIiDWrkx8cvGMqKiIiIiIiISOMa+vFxERERERERGVnqPlNnZr+l8sfCi5xzcwelRiIiIiIiIlK3Rr5Tt6bsfhI4EXgLcP2g1UhERERERETq1sh36v53telmdjlw9KDVSEREREREROo2GN+p+zdgySAsR0RERERERBo0GJ26U4GeQViOiIiIiIiINKiRgVLWlU8CjgLmA1UvzRQREREREZGh1chAKbvK7heA3wOfd879aPCqJCIiIiIiIvXSj4+LiIiIiIhEWCNn6gAws2OBOXi/Wfe4c27roNdKRERERERE6tLId+qOAG4BFuJdeulPtrXAhc65vUNQPxEREREREelHI6NfXgPMBd4BpP2/M/xp3xj8qomIiIiIiMhAGunUnQV81Dn3E+dc1v97ELgY+O9DUjsRERERERHpVyOdujSVI2AC7AZaB6c6IiIiIiIi0ohGOnU/B75iZuOCCWbWhvcbdb8Y7IqJiIiIiIjIwBrp1H0aOBl4zsx+YmYPAt3+tL+tZwFmdqaZPWFmW8zsc/2UW2Rmzsy6GqifiIiIiIjImNPI79T91syOB5YCrwEMWAWsds71DvR4M4sD1wPvwusMPmxm65xzj5WVmwB8Evh13VshIiIiIiIyRtXVqTOzJF4H7vPOuZsPcV0nAVuC37UzszuAs4HHysp9Bfg68NlDXI+IiIiIiMiYUdfll865LPBuvB8cP1QzgG2h+93+tCIzmw/MdM7d3d+CzOxiM1tvZut37NhxGFUSGXrKq0SJ8ipRorxKlCivMpQa+U7dvwH/4zDWZVWmFTuJZhYDrgY+M9CCnHM3Oee6nHNd06ZNO4wqiQw95VWiRHmVKFFeJUqUVxlKdX+nDngW+EczexuwHtgfnumc+5cBHt8NzAzd7wCeD92fALweeNDMAF4BrDOzs5xz6xuop4iIiIiIyJjRSKfufGAPMNf/C3PAQJ26h4HjzWw28BywGPhwcQHOvQRMDe77o2t+Vh06ERERERGR2hoZ/XL24azIOZczs48D9wFx4Fbn3O/N7MvAeufcusNZvoiIiIiIyFhUd6fOzG6tMcsBfcAW4E7n3PM1yuGcuxe4t2zaF2qUPa3euomIiIiIiIxVjVx+OQ14G1AAfudPez3eACgb8AZR+bKZvc059+ig1lJERERERESqamT0y58D/wl0OOdOdc6dijfYyb3Aj4CjgXuAqwa9liIiIiIiIlJVI526TwFfds71BBP8218FLnPOZYArgRMHt4oiIiIiIiJSSyOduvHAUVWmv8KfB/AyjV3SKSIiIiIiIoehkU7d94FbzOwcMzvGzI42s3OAW/B+mBzgJODJwa6kiIiIiIiIVNfIWbVL8H6LblXocTngVuCz/v3HgYsGrXYiIiIiIiLSr0Z+p64HuMTMPgMchzfq5Rbn3P5QGY16KSIiIiIi0kQNf//N78RtHoK6iIiIiIiISIMa+U6diIiIiIiIjDDq1ImIiIiIiESYOnUiIiIiIiIRpk6diIiIiIhIhKlTJyIiIiIiEmHq1ImIiIiIiESYOnUiIiIiIiIRpk6diIiIiIhIhKlTJyIiIiIiEmFN7dSZ2Zlm9oSZbTGzz1WZ/2kze8zMNpvZ/WZ2dDPrJyIiIiIiEjVN69SZWRy4HvgrYA7wITObU1bsEaDLOTcXWAN8vVn1ExERERERiaJmnqk7CdjinNvqnMsAdwBnhws4537snOvx7/4K6Ghi/URERERERCKnmZ26GcC20P1uf1otFwL/OaQ1EhERERERibhmduqsyjRXtaDZUqALWF5j/sVmtt7M1u/YsWMQqygy+JRXiRLlVaJEeZUoUV5lKDWzU9cNzAzd7wCeLy9kZu8E/gE4yzl3oNqCnHM3Oee6nHNd06ZNG5LKigwW5VWiRHmVKFFeJUqUVxlKzezUPQwcb2azzSwFLAbWhQuY2XzgRrwO3fYm1k1ERERERCSSmtapc87lgI8D9wGPA3c5535vZl82s7P8YsuB8cD3zOxRM1tXY3EiIiIiIiICJJq5MufcvcC9ZdO+ELr9zmbWR0REREREJOqa+uPjIiIiIiIiMrjUqRMREREREYkwdepEREREREQiTJ06ERERERGRCFOnTkREREREJMLUqRMREREREYkwdepEREREREQiTJ06ERERERGRCFOnTkREREREJMLUqRMREREREYkwdepEREREREQiTJ06ERERERGRCFOnTkREREREJMLUqRMREREREYkwdepEREREREQiTJ06ERERERGRCFOnTkREREREJMKa2qkzszPN7Akz22Jmn6syv8XM7vTn/9rMjmlm/URERERERKIm0awVmVkcuB54F9ANPGxm65xzj4WKXQjscc69yswWA1cCHzzUdfb15djVmyFXcKSTcfIFRzZfIB4zknEjm3fkC45kPFa873DgIFtwxGNGKmYQg1zOkS04WhJeP9g5R8yMbMERM4ibAd7j8gVHMmYkEzF6MnmSMSMRN5yDXMGRd46EefMP5ArkC450MkauAPlCgZgZZuAcmEHBQUsiRjZXIOcccTNi/vRk3MgVnFeu4C0/GTMSMaMv521rOhWjJ1OgUHAkYsbEdIy8g3HJFImETtaOFH19OXb3ZrAYOH9fJmJGLGYU/LwdyBVoScQoOMgFWfb3dSqYXijQmoiTyxdoScboyxaKyxoXZME5ErEYqbiX4Zyf21Q8RsygL1cgnYzhvxyK84P6ZHIFWpNxnHMlmcfP7YTWGHv7CsUc5wuOmF/X1qRxIOfI5A8+LpWIsT+TpzURwwEHcge3LRYzJrYkSKWa1lxJHUry6oxsvkDCb3sKOAoFKDivfU0ljN7MwRzGDPIOv5305rX6WQ0yE4tBMua1kcHjWhMxMOjN+u1mwmvLMvlCSXYnjYuzv69wsB2Px2hNwd7eg8sK2vzWVIwDmUIxu0HOE/GgjLf8IKuJuFFwXvtb3l5PSsfYn4GJLQlaWpTXkSR8PJCIGVPSKVpbtY9EZPRoZot2ErDFObcVwMzuAM4Gwp26s4Ev+bfXAN80M3POuUZX1teX44+79rNs1QamjW/hH9/3Wj51x6N07+nl3XOm8/HTj+fS1Rvp3tNLR3uaFUsWcPem5zj9ta/gM9/bVJy+fNFcpo5Psfy+J9ixN8M/vu+13PiTp7j0Ha+iN5PnWz9/mkvf8SqSMePlvhyXr9lcsszv/PJP/GLrLr51fhcv9+WKdehoT3PDkgV884E/smNvhr8789Ulj71y4Vxu/8XTnPfm2Tz0xF9477wZJfW9cuFcHnriL7xv3gyue+CPnPfm2VyxdnNJvb/+wyfYse8AK5Z2cvej3dz402foaE+zcmknx0xp4aUDGSaijt1I0NeXY8uu/Vx7/5Nc+NZjSzJ4/YfnY2Zcunoj08a3VGRl+aK5fH/jc3xgwQwuX7O5WOaJF16ic/ZUlq3acDCTZVm49fwuXurJctldB9d3zeITufM321hyyiwSMaMnky+ZH6xvySmz6MnkS+py1TnzuOVnW/nEGSdw96PdnPrqI0tyefW58zhyYivP7emteK08+IftdM2eXLFtR7QmeDkZZ+aktDp2I0Q4rxefehx/e+ejJXntyxaKGX73nOl84owTSnIYtG8Xve1Y2ttS/OzJ7XTOnsp19z/JeW+eze2/eJpPv/sEsjnHslC7d8OSBRyRTvBP9zxetd1cvmguhmNvX4pLQuu75bxOMnspqcMNSxZwz6bneN+JHWx4eifHv2JiSVZXLu3kiHSiIqsrl3ZiOD62amPF9nzijBM4dkoL217qZebEtDp2I0T4eCDcFh4/pU0dOxEZNZp5ND8D2Ba63+1Pq1rGOZcDXgKmHMrKdvVmig34JacdV+xMASzsnFnsIAF07+ll2eqNLOqaVTwQCaZ7b+Z9LOycWVzOws6Z7Nmf5fI1m4u3t+/NFN/4w8u86NRj6d7TS/eevpI6dO/p5dLVG4vLLX/sFWu9ZV+xdjOLumZV1DeYvsxfRnAwEq73Jacd59Vj1QYWdc0qzrtk1QZe6i3Qlymwfd+BQ3l6ZZDt6s1wyaoNLOycWZHB3fuzxf1fLSuXr9nMRaceW5welDl9zlHF10BQtjwLz+3pK3bYgmmfuuNRLjr1WHb7uS6fH6xvt/8aCM/7zPc2sbBzZnE95bm87K5NZHKu6mvl7AUdVbdt+94Mz+7uZcf+TDN2hdQhnNegQwcH8xrOcJCHau3bZXdtYtvu3mJWg7ZsYedMErF4sUMXPO7S1RvJ5anZbl6+ZjMzJ7cVO3TB9HgsXlGHS/02f9mqDZw+56iKrF6yakPVrF6yagPb92aqbs+yVRt4sbfAtt297OxRXkeK8PEAHGwLd/VqH4nI6NHMj6isyrTyM3D1lMHMLgYuBpg1a1bVleUKrtiAT0oni7er3Yfgjd+qTh+XijOOePH+pHSy4nZwv9oyAcal4lXnh5dVbV5/9Qqm19qe8LKDegT3cwXvabXKp1cGWSN5rbYvw9mpJ7tBmYJz/WayfNnlZcalDma+1vxDyW3Mqi/T1ahvUI8gszK0BjOvUDuzwfRxqXgxq8G0SelkzZzEjH7bzXDbH6i1rCCjtV4rtR4XZLLa9uQKjnGpuPLaJI0eDwTC74MizVJPXkUOVTPP1HUDM0P3O4Dna5UxswQwEdhdviDn3E3OuS7nXNe0adOqriwRMzra0wC82Jst3q52H6CjPU2+4KpO78nkebE3W3zci71ZejL5ktvB/WrLBGrODy+32rz+6hVM7+/x5fUI7gfffUnEdenlUGskr9X2ZTg79WQ3KBMz6zeT5csuLzNQrvvLdH+5LTiqTrca9Q3qkYhV+8xHBttg5hVqZzaY3pPJF7MabmNr5aTg6LfdDLf9gVrLCjJa67VS63E9mXzN7QkuWVZem6PR44FAsK9EmqmevIocqmYe0T8MHG9ms80sBSwG1pWVWQec599eBDxwKN+nA5iSTrFiaaf3HYgHn+KaxScWG/W1G7Zxw5IFxfvBd3rWrH+Wq86ZVzJ9+aK5dLS3snbDtuJy1m7YRntbkuWL5hZvT5+Q8suWLvPmh7bS0Z6mo721pA7BdzqC5ZY/9sqF3rKvXDiXNeufrahvMH2Fv4wrF86tqPfKB58qfndgzfpni/NWLu1kYjpGayrG9PEth/L0yiCbkk6xcmknazdsq8jg5LZkcf9Xy8ryRXO5+aGtxelBmQcee6H4GgjKlmdhRnsrV59bur5rFp/IzQ9tZbKf6/L5wfom+6+B8LyrzpnH2g3biuspz+XV584jlbCqr5UfbOyuum3TJ6SYNTnNtLZUM3aF1CGc12988MSKvIYzHOShWvt29bnzmDk5Xcxq0Jat3bCNXCHPirJ274YlC0jEqdluLl80l22797OybH35Qr6iDjf4bf6KpZ088NgLFVldubSzalZXLu1k+oRU1e1ZsbSTSekYMyenmTpOeR0pwscDcLAtnJLWPhKR0cMOsc90aCszew/wDSAO3Oqc+6qZfRlY75xbZ2atwHeA+Xhn6BYHA6vU0tXV5davX191XjDaVb7gaK1z9EtwxVEqY2WjX+YKjlSV0S/jBrGBRr9MWHFEw/LRLwsFR2vZ6JfB6JbBKHH1jn5ZHLlNo19WM+wfyw6U11qjXzrnMPNGnUwNMPplvlCg5RBGvyz4r4WYeaNPtjYw+mWQrWC01kMZ/bInk6dFo1+GRSevzop57G/0y7w/GuVIHP0yyK5GvzxkIz6vGv1SQkZ0XgPHfO6ehpb5zNfeezhVksM0hPurrrw2tUVzzt0L3Fs27Quh233AOYO1vtbWBDPUaAMwuW24ayADaW1N8Mom5LW9SVk4Ij1wGYmuhvPaX+6GIJMTq+TviNahX39brXXIsNLxgIiMdmPqNI2IiIiIiMhoo06diIiIiIhIhKlTJyIiIiIiEmHq1ImIiIiIiESYOnUiIiIiIiIR1tSfNBgKZrYD+NMAxaYCO5tQnZFkrG1zPdu70zl3ZjMqU4vyWtVY215QXqNM21ud8joyaXurU15HJm1vdXXlNfKdunqY2XrnXNdw16OZxto2j6btHU3bUo+xtr0wurZ5NG1LPbS90Tbatmcg2t5oG23bMxBt7+HR5ZciIiIiIiIRpk6diIiIiIhIhI2VTt1Nw12BYTDWtnk0be9o2pZ6jLXthdG1zaNpW+qh7Y220bY9A9H2Rtto256BaHsPw5j4Tp2IiIiIiMhoNVbO1ImIiIiIiIxKo6pTZ2ZnmtkTZrbFzD5XZX6Lmd3pz/+1mR3T/FoOnjq293wz22Fmj/p/Hx2Oeg4WM7vVzLab2e9qzDczu9Z/Pjab2YJm17ERymvFfOV1BFNeK+YrryOY8loxX3kdwZTXivmjJq9NzapzblT8AXHgKeBYIAVsAuaUlbkUWOnfXgzcOdz1HuLtPR/45nDXdRC3+VRgAfC7GvPfA/wnYMApwK+Hu86Huf+U1wj/Ka/Ka5T+lFflNUp/yqvyGpW/ZmZ1NJ2pOwnY4pzb6pzLAHcAZ5eVORu43b+9BjjDzKyJdRxM9WzvqOKcewjY3U+Rs4FvO8+vgElmdlRzatcw5VV5VV5HLuW1kvI6cimvlZTXkWtM5bWZWR1NnboZwLbQ/W5/WtUyzrkc8BIwpSm1G3z1bC/AQv907hozm9mcqg2bep+TkUB5VV6V15FLea2kvI5cymsl5XXkUl5LDVpWR1OnrtonFuVDe9ZTJirq2Zb/AI5xzs0F/ouDn/KMVlHav8qr8hql/au8Kq9R2r/Kq/Iapf2rvI7tvA7avh1NnbpuINyT7wCer1XGzBLARPo/JTqSDbi9zrldzrkD/t2bgc4m1W241JOBkUJ5VV6V15FLea2kvI5cymsl5XXkUl5LDVpWR1On7mHgeDObbWYpvC+Srisrsw44z7+9CHjA+d9SjKABt7fsmtyzgMebWL/hsA74n/5IQqcALznnXhjuStWgvCqvyuvIpbxWUl5HLuW1kvI6cimvpQYtq4nBrdfwcc7lzOzjwH14I+vc6pz7vZl9GVjvnFsH3AJ8x8y24H3CsXj4anx46tzeT5rZWUAOb3vPH7YKDwIz+y5wGjDVzLqBLwJJAOfcSuBevFGEtgA9wAXDU9OBKa/KK8rriKW8Kq9Rorwqr1Ey1vLazKxadDv6IiIiIiIiMpouvxQRERERERlz1KkTERERERGJMHXqREREREREIkydOhERERERkQhTp05ERERERCTC1KkTkWFnZseYmTOzruGui4hIs5nZ3WZ223DXQ0SiS526CPIPfvv7u2246ygSUF5FREREhtao+fHxMeao0O33ATeXTettbnVE+lVPXtubWiMRERFpCjNLOecyw12P0U5n6iLIOffn4A94sXyac+4lADM72cw2mlmfmT1iZu/xz4ycNpz1l7Gl3ryGmVnMzK43s6fN7PimV1okxMweNLMbzOyfzGynmW03s382M72HSsPMbJyZ3WZm+8zsL2b2+bL5S83sYTPb62fte2Y2IzT/NP+9/K/MbIOZ9ZrZT82sw8zebmab/GXfbWZTQo97o5n9yM/wy2b2MzN7U9m6TzCzn/jHDU/4xw37zOz8UJkZZnaHme3x/+5ROz22mFmbmX07lOG/D19CbGbPmNmXzOxWM3sRWO1Pf4OZ/Zef2d3+62BiaLm3mdndZev6kpn9rryMmf2jv+59ZvYtM0uHypxqZr/y571kZr82s9cP9fMy3PSGNEqZ2XjgbuAPQCfwd8DyYa2USB3MLIn3BnAa8Fbn3B+Ht0YiACwBcsCbgY8Dfwt8cFhrJFH1z8C7gIXAGcB84NTQ/BTwRWAe3tUNU4HvVlnO/8bL4cl4VzvcCXwBuBiv/Xwd8KVQ+QnAd4C3AScBjwL3mtlU8D5MA76Pl/NTgPP9erQECzCzccCPgT7g7cCbgBeA//LnydhwFd7+/wBwOl5W31ZW5tN4x6BdwOf9fPwQ2IeXvw/gtae3HsL63+6v8wy819G7gSsBzCwB/AD4mV/mZOAaIH8I64kUXX45ei0B4sCFzrle4Pdm9lX8T0tERqhxwH8AE4G3Oed2D3N9RAKPOee+4N9+0swuwjugqHawLVKV/4HrhcBHnHP3+dMuALqDMs658EHuVjNbBjxuZh3Oue7QvP/lnPupv4yVwHVAp3Nuoz/tdmBRaLkPlNXlE3gHxGcCq/A6mq8G3u2ce84vcxnw89DDFgMGXOCcc36ZjwHb8Tqgdx3K8yLR4Wf4I8D/dM79P3/ahYQy7PuJc+7rocddBIwH/to5t9efdjHwYzN7lXNuSwPVyONlcB/wOzO7ArjFzP4e70OIScB/OOee8sv/oeENjSCdqRu9XgP8zu/QBX49XJURqdNqYDLwTnXoZITZXHb/eWD6cFREIu04vDNxvwwm+Aemvw3um9kCM/uBmf3JzPYC6/1Zs8qWFc7kX/z/vy2bVsyomU03sxvN7EkzewnY688Plvsa4PmgQ+d7GCiE7ncCs4G9/qVt+4CX8M4UHjfg1stocByQBH4TTHDO7Qd+V1Zufdn91wKbgw6d7xd4+ZrTYB02+6+bwC/xXlfH+ccOtwH3+ZcGf9rMZja4/EhSp270MsANdyVEGnQP8HrgLcNdEZEy2bL7Dr2HSuOs35lmbcB9QA/w18Ab8c6kgXfQGhbOpANwzpVPC2f0dn95l+Fd9nYi3tmVYLn1HDfE8C7bPLHs7wTgxgEeK6NDkOGBsrK/yuNqPSaYXqDyNZKsv2r+wpy7AO+yy4eAs/CurvhvjS4navSGNHo9Drwh/MVRvGuYRUay/4v3HZF/N7N3D3dlREQG2Ra8ztgpwQS/IxcM4vAavO/Qfd4595Bz7g8M3hnhtwLXOefucc79Hu9MXXgk4seBGWb2ytC0LkqPFTcCrwJ2Oue2lP3p6oqxIchw8ZjS/77cQAORPAbMM7MJoWlvxsvX4/79HZRmErwPDcq9wX/dBE4BMkBwuSXOuU3OuSudc6cBDwLnDVC/yFOnbvRajXfN8c1mNsfM3gkEI2zpDJ6MWM65m/A+Sf53M3vXcNdHRGSw+JeM3QJcaWbvMrPX4Q0UEfeLPAscAD5uZsea2XuBrwzS6p8ElvrHBG8E7sA7EA78P+AJ4HYzm2dmpwD/gjdwSnDcsBrvss4f+CNtzvZHGrxKI2CODX6Gb8XL8BlmNgfvA9kY/R9frsY7e/dtfxTMU/HO7v5b6Pt0DwDzzewjZvYqM/s7ql+5kwBuNbPX+ccJXwNuds7t9zP5NTN7s5kdbWbvAObidSpHNXXqRin/Rfd+vNFy6ZgPAAABvElEQVSvHsEb+fJL/uy+YaqWSF2cczfijZyljp2IjDafxRtB8vv+/9/hXSaGc24H3hmF/453EPpFvLZwMHwEb6CKDXgduluBZ4KZzrkC3oiELXjfl7od+CregXqfX6YHb6TOrcD38AaguB3vO3V7BqmeMvJ9FvgpsA4vw5vxvkNX8/jSz85/A47Ay9cP8L4L95FQmfvwRnX9Kl5OjwFuqLK4nwC/5+Dr6AG8Ud7Bu3T5BLx8PomXz9X4o2OOZuYPXiRjgJmdjRf+6c65ncNdHxERERm5zGwe3nfoupxzG4a7PjIymVkL8CdguXPuqiFe123AVOfc+4ZyPVGknzQYxczsPLxP07bhXev8DbwhXtWhExERkRJm9gG8S+T+iHeW5F+ATXjfpRMBwMzm441m+Ru83z+8wv9/53DWa6xTp250OxLvNPZRwJ/xRha8YlhrJCIiIiPVBLzL1GbiXU75IHCZ02VdUunTeL9rmMM7m3tq2e8oSpPp8ksREREREZEI00ApIiIiIiIiEaZOnYiIiIiISISpUyciIiIiIhJh6tSJiIiIiIhEmDp1IiIiIiIiEaZOnYiIiIiISIT9f2tkCQyIy+qWAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 900x900 with 30 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.pairplot(data_scaled)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 331,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(data_scaled.loc[:, ['Tg', 'Tk', 'n']], data_scaled['damage'], test_size=0.2, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 332,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Define function for CV for \"weak\" models\n",
    "def Stacking(model, train, y, test, n_fold, with_eval_set = False):\n",
    "    folds = KFold(n_splits = n_fold, random_state = 42)\n",
    "    test_pred = np.empty((0, 1), float)\n",
    "    train_pred = np.empty((0, 1), float)\n",
    "    for train_indices, val_indices in folds.split(train, y.values):\n",
    "        x_train, x_val = train.iloc[train_indices], train.iloc[val_indices]\n",
    "        y_train, y_val = y.iloc[train_indices], y.iloc[val_indices]\n",
    "\n",
    "        if with_eval_set == False:\n",
    "            model.fit(x_train, y_train)\n",
    "        else:\n",
    "            model.fit(X = x_train, y = y_train, eval_set = [(x_val, y_val)], verbose = 0)\n",
    "            \n",
    "        train_pred = np.append(train_pred, model.predict(x_val))\n",
    "        \n",
    "    test_pred = np.append(test_pred, model.predict(test))\n",
    "    return pd.DataFrame(test_pred), pd.DataFrame(train_pred), model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 333,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Define lightGBM model\n",
    "model0 = LGBMRegressor(boosting_type = 'gbdt',\n",
    "                       metric = 'rmse',\n",
    "                       n_estimators = 1000,\n",
    "                       num_boost_round = 1000,\n",
    "                       early_stopping_rounds = 100,\n",
    "                       # device = 'gpu',\n",
    "                       n_jobs = -1,\n",
    "                       random_state = 42,\n",
    "                       verbose = -1,\n",
    "                       verbose_eval = False)\n",
    "\n",
    "# Create the random grid\n",
    "model0_param_grid = {'learning_rate': [x for x in np.linspace(0.001, 0.05, num = 5)],\n",
    "                     'max_depth': [int(x) for x in np.linspace(2, 8, num = 4)]}\n",
    "\n",
    "model0_rs = RandomizedSearchCV(estimator = model0,\n",
    "                               param_distributions = model0_param_grid,\n",
    "                               n_iter=10, cv=3, verbose=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 334,
   "metadata": {},
   "outputs": [],
   "source": [
    "# # Define RandomForest model\n",
    "# model0 = RandomForestRegressor(n_estimators = 100,\n",
    "#                                bootstrap = True,\n",
    "#                                oob_score = True,\n",
    "#                                n_jobs = -1,\n",
    "#                                random_state = 42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 335,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Define AdaBoost model\n",
    "model1 = AdaBoostRegressor(base_estimator = RandomForestRegressor(),\n",
    "                           loss = 'square',\n",
    "                           n_estimators = 1000,\n",
    "                           random_state = 42)\n",
    "\n",
    "# Create the random grid\n",
    "model1_param_grid = {'learning_rate': [x for x in np.linspace(0.001, 0.05, num = 5)],\n",
    "                     'base_estimator__max_depth': [int(x) for x in np.linspace(2, 8, num = 4)]}\n",
    "\n",
    "model1_rs = RandomizedSearchCV(estimator = model1,\n",
    "                               param_distributions = model1_param_grid,\n",
    "                               n_iter=10, cv=3, verbose=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 336,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Define polynomial regression model\n",
    "model2 = make_pipeline(PolynomialFeatures(),\n",
    "                       LinearRegression(n_jobs = -1))\n",
    "\n",
    "model2_param_grid = {'polynomialfeatures__degree': [2, 3, 4, 5]}\n",
    "\n",
    "model2_rs = RandomizedSearchCV(estimator = model2,\n",
    "                               param_distributions = model2_param_grid,\n",
    "                               n_iter=10, cv=3, verbose=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 337,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Define CatBoost model\n",
    "model3 = CatBoostRegressor(loss_function='RMSE',\n",
    "                           random_seed = 42,\n",
    "                           task_type = 'GPU',\n",
    "                           # boosting_type = 'Plain', # set if it exceeds RAM\n",
    "                           # params for early stopping:\n",
    "                           iterations = 1000,\n",
    "                           od_type = 'Iter',\n",
    "                           od_wait = 100,\n",
    "                           silent = True)\n",
    "\n",
    "# Create the random grid\n",
    "model3_param_grid = {'learning_rate': [x for x in np.linspace(0.001, 0.05, num = 5)],\n",
    "                     'depth': [int(x) for x in np.linspace(2, 8, num = 4)]}\n",
    "\n",
    "model3_rs = RandomizedSearchCV(estimator = model3,\n",
    "                               param_distributions = model3_param_grid,\n",
    "                               n_iter=10, cv=3, verbose=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 338,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n",
      "Warning: less than 75% gpu memory available for training. Free: 6126.1875 Total: 11178.125\n"
     ]
    }
   ],
   "source": [
    "# Train \"weak\" models\n",
    "kwargs = {'n_fold': 10, 'train': X_train, 'test': X_test, 'y': y_train}\n",
    "\n",
    "test_pred0, train_pred0, lightgbm_model = Stacking(model = model0, with_eval_set = True, **kwargs)\n",
    "# pickle.dump(lightgbm_model, open('lightgbm_model.pkl', 'wb'))\n",
    "\n",
    "test_pred1, train_pred1, adaboost_model = Stacking(model = model1_rs, **kwargs)\n",
    "# pickle.dump(adaboost_model, open('adaboost_model.pkl', 'wb'))\n",
    "\n",
    "test_pred2, train_pred2, polynom_reg_model = Stacking(model = model2_rs, **kwargs)\n",
    "# pickle.dump(polynom_reg_model, open('polynom_reg_model.pkl', 'wb'))\n",
    "\n",
    "test_pred3, train_pred3, catboost_model = Stacking(model = model3_rs, with_eval_set = True, **kwargs)\n",
    "# pickle.dump(catboost_model, open('catboost_model.pkl', 'wb'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 339,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE score for stacked model: 0.0009052603756906132\n"
     ]
    }
   ],
   "source": [
    "# Create \"strong\" model over \"weak\" predictions\n",
    "train_preds = pd.concat([train_pred0, train_pred1, train_pred2, train_pred3], axis = 1)\n",
    "test_preds = pd.concat([test_pred0, test_pred1, test_pred2, test_pred3], axis = 1)\n",
    "train_preds.columns = ['lightgbm_model', 'adaboost', 'polynom_reg', 'catboost']\n",
    "test_preds.columns = ['lightgbm_model', 'adaboost', 'polynom_reg', 'catboost']\n",
    "\n",
    "strong_model = xgb.XGBRegressor(eval_metric = \"rmse\",\n",
    "                                # min_child_weight=1.5,                                                                  \n",
    "                                # reg_alpha=0.75,\n",
    "                                # reg_lambda=0.45,\n",
    "                                objective='reg:squarederror',\n",
    "                                n_estimators = 1000,\n",
    "                                num_boost_round = 1000,\n",
    "                                early_stopping_rounds = 100,\n",
    "                                # max_depth = 2,\n",
    "                                # num_leaves = 4,\n",
    "                                eval_set = [(X_test, y_test)],\n",
    "                                kvargs = {'tree_method':'gpu_hist'}, # enable GPU\n",
    "                                seed = 42)\n",
    "\n",
    "# Create the random grid\n",
    "strong_model_param_grid = {'colsample_bytree': [0.4, 0.6, 0.8],\n",
    "                           'reg_lambda': [0.01, 0.5, 1],\n",
    "                           'learning_rate': [x for x in np.linspace(0.001, 0.05, num = 5)],\n",
    "                           'max_depth': [int(x) for x in np.linspace(2, 8, num = 4)],\n",
    "                           'num_leaves': [20, 40, 60, 80]}\n",
    "\n",
    "\n",
    "strong_model = RandomizedSearchCV(estimator = strong_model,\n",
    "                                  param_distributions = strong_model_param_grid,\n",
    "                                  n_iter=100, cv=3, verbose=0)\n",
    "\n",
    "strong_model.fit(train_preds, y_train)\n",
    "# pickle.dump(strong_model, open('xgbr_strong_model.pkl', 'wb')) # save the fitted model\n",
    "\n",
    "y_pred_test = strong_model.predict(test_preds)\n",
    "print(f\"RMSE score for stacked model: {sqrt(mean_squared_error(y_pred_test, y_test))}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 340,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE score for stacked model: 0.0009052603756906132\n",
      "RMSE score for LightGBM model: 0.012030017903305844\n",
      "RMSE score for AdaBoost model: 0.0006192848001356577\n",
      "RMSE score for polynimial regression: 0.006991856066514724\n",
      "RMSE score for CatBoost model: 0.012473316848840706\n"
     ]
    }
   ],
   "source": [
    "y_pred_test = strong_model.predict(test_preds)\n",
    "print(f\"RMSE score for stacked model: {sqrt(mean_squared_error(y_pred_test, y_test))}\")\n",
    "print(f\"RMSE score for LightGBM model: {sqrt(mean_squared_error(test_pred0, y_test))}\")\n",
    "print(f\"RMSE score for AdaBoost model: {sqrt(mean_squared_error(test_pred1, y_test))}\")\n",
    "print(f\"RMSE score for polynimial regression: {sqrt(mean_squared_error(test_pred2, y_test))}\")\n",
    "print(f\"RMSE score for CatBoost model: {sqrt(mean_squared_error(test_pred3, y_test))}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 321,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE score for stacked model: 0.0006769867078193512\n",
      "RMSE score for LightGBM model: 0.0088917921494916\n",
      "RMSE score for AdaBoost model: 0.0008210242015147392\n",
      "RMSE score for polynimial regression: 0.005867663762123875\n",
      "RMSE score for CatBoost model: 0.011301488462533735\n"
     ]
    }
   ],
   "source": [
    "y_pred_test = strong_model.predict(test_preds)\n",
    "print(f\"RMSE score for stacked model: {sqrt(mean_squared_error(y_pred_test, y_test))}\")\n",
    "print(f\"RMSE score for LightGBM model: {sqrt(mean_squared_error(test_pred0, y_test))}\")\n",
    "print(f\"RMSE score for AdaBoost model: {sqrt(mean_squared_error(test_pred1, y_test))}\")\n",
    "print(f\"RMSE score for polynimial regression: {sqrt(mean_squared_error(test_pred2, y_test))}\")\n",
    "print(f\"RMSE score for CatBoost model: {sqrt(mean_squared_error(test_pred3, y_test))}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Visualize the result metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 322,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_true_vs_pred(y_pred, y_test, title_text, ax=None, **kwargs):\n",
    "    ax = ax or plt.gca()\n",
    "\n",
    "    ax.set_title(title_text, y=1.01, fontsize=19, fontweight='bold')\n",
    "    sns.scatterplot(x = y_pred, y = y_test, ax=ax, color='red', s=100)\n",
    "    ax.plot(np.arange(0, 1.1, 0.1), np.arange(0, 1.1, 0.1), color='black', linestyle='dashed')\n",
    "    \n",
    "    rmse = sqrt(mean_squared_error(y_test, y_pred))\n",
    "    ax.text(0.17, 0.8, f'$RMSE={round(rmse, 4)}$', fontsize=12, fontweight='bold')\n",
    "    ax.set_ylabel('$Damage_{true}$', fontsize = 15)\n",
    "    ax.set_xlabel('$Damage_{pred}$', fontsize = 15)\n",
    "    ax.grid()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 323,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_true_minus_pred(y_pred, y_test, title_text, ax=None, **kwargs):\n",
    "    ax = ax or plt.gca()\n",
    "    \n",
    "    ax.set_title(title_text, y=1.01, fontsize = 20, fontweight='bold')\n",
    "    sns.distplot((y_test - y_pred), kde=True, norm_hist=True, color='red')\n",
    "\n",
    "    rmse = sqrt(mean_squared_error(y_test, y_pred))\n",
    "    ax.text(-0.015, 1350, f'$RMSE={round(rmse, 4)}$', fontsize=12, fontweight='bold')\n",
    "\n",
    "    ax.set_ylabel('$Occurence$', fontsize = 15)\n",
    "    ax.set_xlabel('$Damage_{true}$ - $Damage_{pred}$', fontsize = 15)\n",
    "    ax.grid()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 324,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize = (5, 5))\n",
    "plot_true_vs_pred(y_pred_test, y_test, \"Stacked model\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 325,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize = (5, 5))\n",
    "plot_true_minus_pred(y_pred_test, y_test, 'Stacked model')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 341,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(2, 2, figsize=(10, 10))\n",
    "\n",
    "axes = axes.flatten()\n",
    "\n",
    "# plot_true_vs_pred(y_pred_test, y_test, 'Stacked model', ax=axes[0])\n",
    "\n",
    "plot_true_vs_pred(test_pred0.values.flatten(), y_test, 'LightGBM model', ax=axes[0])\n",
    "plot_true_vs_pred(test_pred1.values.flatten(), y_test, 'AdaBoost model', ax=axes[1])\n",
    "plot_true_vs_pred(test_pred2.values.flatten(), y_test, 'Polynom_reg model', ax=axes[2])\n",
    "plot_true_vs_pred(test_pred3.values.flatten(), y_test, 'CatBoost model', ax=axes[3])\n",
    "    \n",
    "fig.tight_layout()\n",
    "fig.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## How to re-use the pre-trained model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Import the new dataset\n",
    "data = pd.read_csv('Data_training.csv', index_col = 0)\n",
    "data.columns = ['Tg', 'Tk', 'n', 'damage']\n",
    "\n",
    "# Define group variable\n",
    "split_val = 0.1\n",
    "data['groups'] = [1 if f > split_val else 0 for f in data.damage]\n",
    "\n",
    "# Normalize the data\n",
    "scaler = pickle.load(open('scaler.pkl', 'rb'))\n",
    "data_scaled = scaler.transform(data)\n",
    "data_scaled = pd.DataFrame(data_scaled, columns = data.columns)\n",
    "\n",
    "# Make inference\n",
    "x = data_scaled.loc[:, ['Tg', 'Tk', 'n']]\n",
    "y_true = data['damage']\n",
    "\n",
    "model1 = pickle.load(open('lightgbm_model.pkl', 'rb'))\n",
    "model2 = pickle.load(open('polynom_reg_model.pkl', 'rb'))\n",
    "model3 = pickle.load(open('catboost_model.pkl', 'rb'))\n",
    "strong_model = pickle.load(open('xgbr_strong_model.pkl', 'rb'))\n",
    "\n",
    "model1_preds = model1.predict(x)\n",
    "model2_preds = model2.predict(x)\n",
    "model3_preds = model3.predict(x)\n",
    "\n",
    "models_preds = pd.concat([model1_preds, model2_preds, model3_preds], axis = 1)\n",
    "\n",
    "y_pred = strong_model.predict(models_preds)\n",
    "print(f\"RMSE score for stacked model: {sqrt(mean_squared_error(y_pred, y_true))}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For example"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Import the new data_interset\n",
    "data_inter = pd.read_csv('/home/mraevsky/Skoltech/Surrogate_modeling/ciam_ml_model/Data_training_2.csv', index_col = 0)\n",
    "data_inter.columns = ['Tg', 'Tk', 'n', 'damage']\n",
    "\n",
    "# Define pseudo_group column\n",
    "split_val = 0.1\n",
    "data_inter['groups'] = [1 if f > split_val else 0 for f in data_inter.damage]\n",
    "\n",
    "# Normalize the data_inter\n",
    "scaler = pickle.load(open('../ciam_ml_model/scaler.pkl', 'rb'))\n",
    "data_inter_scaled = scaler.transform(data_inter)\n",
    "data_inter_scaled = pd.DataFrame(data_inter_scaled, columns = data_inter.columns)\n",
    "\n",
    "# Make inference\n",
    "x = data_inter_scaled.loc[:, ['Tg', 'Tk', 'n']]\n",
    "\n",
    "model1 = pickle.load(open('../ciam_ml_model/lightgbm_model.pkl', 'rb'))\n",
    "model2 = pickle.load(open('../ciam_ml_model/adaboost_model.pkl', 'rb'))\n",
    "model3 = pickle.load(open('../ciam_ml_model/polynom_reg_model.pkl', 'rb'))\n",
    "model4 = pickle.load(open('../ciam_ml_model/catboost_model.pkl', 'rb'))\n",
    "strong_model = pickle.load(open('../ciam_ml_model/xgbr_strong_model.pkl', 'rb'))\n",
    "\n",
    "model1_preds = model1.predict(x)\n",
    "model2_preds = model2.predict(x)\n",
    "model3_preds = model3.predict(x)\n",
    "model4_preds = model4.predict(x)\n",
    "\n",
    "models_preds = pd.concat([pd.DataFrame(model1_preds), \n",
    "                          pd.DataFrame(model2_preds), \n",
    "                          pd.DataFrame(model3_preds), \n",
    "                          pd.DataFrame(model4_preds)], axis = 1)\n",
    "models_preds.columns = ['lightgbm_model', 'adaboost', 'polynom_reg', 'catboost']\n",
    "\n",
    "y_pred = strong_model.predict(models_preds)\n",
    "\n",
    "# Rescale data back to original scale\n",
    "data_inter_pred = data_inter_scaled.loc[:, ['Tg', 'Tk', 'n']]\n",
    "data_inter_pred['damage_pred'] = y_pred\n",
    "data_inter_pred['groups'] = 0 # pseudo values\n",
    "data_inter_pred = scaler.inverse_transform(data_inter_pred)\n",
    "data_inter_pred"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}