03_model.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "38cac848",
   "metadata": {},
   "source": [
    "<div class=\"alert alert-success\">\n",
    "\n",
    "\n",
    "\n",
    "# Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "efa6adfd",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>table {align:left;display:block}</style>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "from scipy import stats\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "# Increases Display Resolution for Graphs \n",
    "# Only works inside notebook\n",
    "%matplotlib inline \n",
    "%config InlineBackend.figure_format = 'retina'\n",
    "\n",
    "# Decision Tree and Model Evaluation Imports\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.feature_selection import RFE\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.tree import plot_tree, export_text\n",
    "\n",
    "# import sklearn.metrics\n",
    "from sklearn.metrics import ConfusionMatrixDisplay\n",
    "from sklearn.metrics import confusion_matrix \n",
    "from sklearn.metrics import plot_confusion_matrix \n",
    "from sklearn.metrics import classification_report \n",
    "from sklearn.metrics import mean_squared_error\n",
    "\n",
    "# import sklearn.preprocessing\n",
    "import sklearn.preprocessing\n",
    "from sklearn.preprocessing import MinMaxScaler \n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.preprocessing import RobustScaler\n",
    "from sklearn.preprocessing import QuantileTransformer\n",
    "\n",
    "# import Environment and Helper Files\n",
    "from env import user, password, host\n",
    "import QMCBT_00_quicktips as qt\n",
    "import QMCBT_01_acquire as acquire\n",
    "import QMCBT_02_prepare as prepare\n",
    "#import QMCBT_03_explore as explore\n",
    "#import QMCBT_04_modeling as model\n",
    "#import QMCBT_05_evaluate as evaluate\n",
    "import QMCBT_explore_evaluate as ee\n",
    "import QMCBT_wrangle as w\n",
    "\n",
    "# set constants\n",
    "&alpha; = 0.05\n",
    "alpha = 0.05\n",
    "np.random.seed(1992)\n",
    "#random_state=1992\n",
    "\n",
    "# Turns off warnings\n",
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "\n",
    "# assists with processor speed\n",
    "#import matplotlib as mpl\n",
    "#mpl.rcParams['agg.path.chunksize'] = 10000\n",
    "\n",
    "# Left Align Tables in Jupyter Notebook\n",
    "from IPython.core.display import HTML\n",
    "table_css = 'table {align:left;display:block}'\n",
    "HTML('<style>{}</style>'.format(table_css))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "9556ef13",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<module 'QMCBT_wrangle' from '/Users/qmcbt/codeup-data-science/project-2_zillow/QMCBT_wrangle.py'>"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# allows import reload without needing to clear kernel and rerun\n",
    "# reload(packagename) \n",
    "from importlib import reload\n",
    "\n",
    "# This code refreshes all of my helper files (w/o the need to stop the kernel)\n",
    "reload(qt)\n",
    "reload(acquire)\n",
    "reload(prepare)\n",
    "#reload(explore)\n",
    "#reload(model)\n",
    "#reload(evaluate)\n",
    "reload(ee)\n",
    "reload(w)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "1b8dafdc",
   "metadata": {},
   "outputs": [],
   "source": [
    "# use custom wrangle function to automate Acquire and Prepare\n",
    "df = w.wrangle_zillow()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "aa1d44ce",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "________________________________________________________________\n",
      "|                              DF                              |\n",
      "|--------------------:--------------------:--------------------|\n",
      "|        Train       |      Validate      |        Test        |\n",
      ":--------------------------------------------------------------:\n",
      "\n",
      "\n",
      "Prepared df: (47603, 18)\n",
      "\n",
      "      Train: (28561, 18) - 60.0%\n",
      "   Validate: (9521, 18) - 20.0%\n",
      "       Test: (9521, 18) - 20.0%\n"
     ]
    }
   ],
   "source": [
    "# use predefined function to split data for train, validate, and test\n",
    "train, validate, test = w.split(df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "ada30c05",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_______________________________________________________________\n",
      "|                              DF                             |\n",
      "|-------------------:-------------------:---------------------|\n",
      "|       Train       |       Validate    |          Test       |\n",
      "|-------------------:-------------------:---------------------|\n",
      "| x_train | y_train |   x_val  |  y_val |   x_test  |  y_test |\n",
      ":-------------------------------------------------------------:\n",
      "\n",
      "\n",
      "   X_train: (28561, 3)   Index(['bedroomcnt', 'bathroomcnt', 'calculatedfinishedsquarefeet'], dtype='object')\n",
      "   y_train: (28561,)     Index(taxvaluedollarcnt)\n",
      "\n",
      "X_validate: (9521, 3)   Index(['bedroomcnt', 'bathroomcnt', 'calculatedfinishedsquarefeet'], dtype='object')\n",
      "y_validate: (9521,)     Index(taxvaluedollarcnt)\n",
      "\n",
      "    X_test: (9521, 3)   Index(['bedroomcnt', 'bathroomcnt', 'calculatedfinishedsquarefeet'], dtype='object')\n",
      "    y_test: (9521,)     Index(taxvaluedollarcnt)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "(       bedroomcnt  bathroomcnt  calculatedfinishedsquarefeet\n",
       " 0               2          1.0                           926\n",
       " 1               3          3.0                          2425\n",
       " 2               5          4.5                          4483\n",
       " 3               4          3.0                          2362\n",
       " 4               2          2.0                          1522\n",
       " ...           ...          ...                           ...\n",
       " 28556           3          1.0                          1176\n",
       " 28557           3          1.0                          1102\n",
       " 28558           3          2.0                          1240\n",
       " 28559           4          4.0                          3272\n",
       " 28560           3          3.0                          2860\n",
       " \n",
       " [28561 rows x 3 columns],\n",
       " 0         115097\n",
       " 1         776889\n",
       " 2        1133856\n",
       " 3         676411\n",
       " 4         860000\n",
       "           ...   \n",
       " 28556     425000\n",
       " 28557     362332\n",
       " 28558      49237\n",
       " 28559    1397970\n",
       " 28560     428949\n",
       " Name: taxvaluedollarcnt, Length: 28561, dtype: Int64,\n",
       "       bedroomcnt  bathroomcnt  calculatedfinishedsquarefeet\n",
       " 0              5          4.0                          4024\n",
       " 1              3          2.0                          1254\n",
       " 2              4          2.0                          1292\n",
       " 3              3          2.0                          1378\n",
       " 4              4          3.0                          2150\n",
       " ...          ...          ...                           ...\n",
       " 9516           3          1.0                          1798\n",
       " 9517           5          7.0                          6641\n",
       " 9518           3          4.0                          3435\n",
       " 9519           2          1.0                          1278\n",
       " 9520           5          4.5                          4211\n",
       " \n",
       " [9521 rows x 3 columns],\n",
       " 0       1663334\n",
       " 1         45302\n",
       " 2        388528\n",
       " 3        130715\n",
       " 4        407342\n",
       "          ...   \n",
       " 9516     258220\n",
       " 9517    1637431\n",
       " 9518    1015473\n",
       " 9519     228983\n",
       " 9520    1267425\n",
       " Name: taxvaluedollarcnt, Length: 9521, dtype: Int64,\n",
       "       bedroomcnt  bathroomcnt  calculatedfinishedsquarefeet\n",
       " 0              4          3.0                          2655\n",
       " 1              4          2.0                          1462\n",
       " 2              3          3.0                          2273\n",
       " 3              3          3.0                          2165\n",
       " 4              4          3.0                          2484\n",
       " ...          ...          ...                           ...\n",
       " 9516           3          2.0                          2208\n",
       " 9517           3          2.0                          1899\n",
       " 9518           2          1.0                           840\n",
       " 9519           3          2.0                          2079\n",
       " 9520           5          3.0                          3602\n",
       " \n",
       " [9521 rows x 3 columns],\n",
       " 0        320000\n",
       " 1        440098\n",
       " 2        171043\n",
       " 3       1812185\n",
       " 4        147620\n",
       "          ...   \n",
       " 9516     466957\n",
       " 9517     859492\n",
       " 9518     485000\n",
       " 9519     291628\n",
       " 9520    1270451\n",
       " Name: taxvaluedollarcnt, Length: 9521, dtype: Int64)"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# split data and target for evaluation of models\n",
    "target = 'taxvaluedollarcnt'\n",
    "feature_cols = ['bedroomcnt', 'bathroomcnt', 'calculatedfinishedsquarefeet']\n",
    "\n",
    "w.Xy_split(feature_cols, target, train, validate, test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "24c19660",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>10</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>bedroomcnt</th>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>bathroomcnt</th>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4.5</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.5</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fullbathcnt</th>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>bed_bath_ratio</th>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.11</td>\n",
       "      <td>1.33</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.5</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.33</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fireplacecnt</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>age</th>\n",
       "      <td>66</td>\n",
       "      <td>68</td>\n",
       "      <td>15</td>\n",
       "      <td>56</td>\n",
       "      <td>63</td>\n",
       "      <td>69</td>\n",
       "      <td>20</td>\n",
       "      <td>31</td>\n",
       "      <td>40</td>\n",
       "      <td>29</td>\n",
       "      <td>60</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>calculatedfinishedsquarefeet</th>\n",
       "      <td>926</td>\n",
       "      <td>2425</td>\n",
       "      <td>4483</td>\n",
       "      <td>2362</td>\n",
       "      <td>1522</td>\n",
       "      <td>1785</td>\n",
       "      <td>1065</td>\n",
       "      <td>1620</td>\n",
       "      <td>1204</td>\n",
       "      <td>2184</td>\n",
       "      <td>1213</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>lotsizesquarefeet</th>\n",
       "      <td>7869</td>\n",
       "      <td>11580</td>\n",
       "      <td>10936</td>\n",
       "      <td>24277</td>\n",
       "      <td>10630</td>\n",
       "      <td>7215</td>\n",
       "      <td>7466</td>\n",
       "      <td>5700</td>\n",
       "      <td>902</td>\n",
       "      <td>10952</td>\n",
       "      <td>8549</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>has_taxdelinquency</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>has_hottuborspa</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>has_basement</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>has_pool</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fips</th>\n",
       "      <td>6037</td>\n",
       "      <td>6037</td>\n",
       "      <td>6059</td>\n",
       "      <td>6037</td>\n",
       "      <td>6037</td>\n",
       "      <td>6037</td>\n",
       "      <td>6037</td>\n",
       "      <td>6059</td>\n",
       "      <td>6059</td>\n",
       "      <td>6037</td>\n",
       "      <td>6037</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>logerror</th>\n",
       "      <td>-0.075689</td>\n",
       "      <td>0.029119</td>\n",
       "      <td>0.056621</td>\n",
       "      <td>-0.05261</td>\n",
       "      <td>-0.186242</td>\n",
       "      <td>-0.034413</td>\n",
       "      <td>0.090559</td>\n",
       "      <td>0.048001</td>\n",
       "      <td>-0.037959</td>\n",
       "      <td>-0.006433</td>\n",
       "      <td>-0.041282</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>taxvaluedollarcnt</th>\n",
       "      <td>115097</td>\n",
       "      <td>776889</td>\n",
       "      <td>1133856</td>\n",
       "      <td>676411</td>\n",
       "      <td>860000</td>\n",
       "      <td>1087043</td>\n",
       "      <td>98129</td>\n",
       "      <td>556000</td>\n",
       "      <td>268300</td>\n",
       "      <td>471970</td>\n",
       "      <td>241620</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>county_Los Angeles County</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>county_Orange County</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>county_Ventura County</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                    0         1         2        3         4   \\\n",
       "bedroomcnt                           2         3         5        4         2   \n",
       "bathroomcnt                        1.0       3.0       4.5      3.0       2.0   \n",
       "fullbathcnt                          1         3         4        3         2   \n",
       "bed_bath_ratio                     2.0       1.0      1.11     1.33       1.0   \n",
       "fireplacecnt                         0         0         0        0         0   \n",
       "age                                 66        68        15       56        63   \n",
       "calculatedfinishedsquarefeet       926      2425      4483     2362      1522   \n",
       "lotsizesquarefeet                 7869     11580     10936    24277     10630   \n",
       "has_taxdelinquency                   1         0         0        0         0   \n",
       "has_hottuborspa                      0         0         1        0         0   \n",
       "has_basement                         0         0         0        0         0   \n",
       "has_pool                             0         1         1        1         0   \n",
       "fips                              6037      6037      6059     6037      6037   \n",
       "logerror                     -0.075689  0.029119  0.056621 -0.05261 -0.186242   \n",
       "taxvaluedollarcnt               115097    776889   1133856   676411    860000   \n",
       "county_Los Angeles County            1         1         0        1         1   \n",
       "county_Orange County                 0         0         1        0         0   \n",
       "county_Ventura County                0         0         0        0         0   \n",
       "\n",
       "                                    5         6         7         8   \\\n",
       "bedroomcnt                           2         3         3         2   \n",
       "bathroomcnt                        2.0       2.0       3.0       1.5   \n",
       "fullbathcnt                          2         2         3         1   \n",
       "bed_bath_ratio                     1.0       1.5       1.0      1.33   \n",
       "fireplacecnt                         0         0         1         0   \n",
       "age                                 69        20        31        40   \n",
       "calculatedfinishedsquarefeet      1785      1065      1620      1204   \n",
       "lotsizesquarefeet                 7215      7466      5700       902   \n",
       "has_taxdelinquency                   0         0         0         0   \n",
       "has_hottuborspa                      0         0         0         0   \n",
       "has_basement                         0         0         0         0   \n",
       "has_pool                             0         0         0         0   \n",
       "fips                              6037      6037      6059      6059   \n",
       "logerror                     -0.034413  0.090559  0.048001 -0.037959   \n",
       "taxvaluedollarcnt              1087043     98129    556000    268300   \n",
       "county_Los Angeles County            1         1         0         0   \n",
       "county_Orange County                 0         0         1         1   \n",
       "county_Ventura County                0         0         0         0   \n",
       "\n",
       "                                    9         10  \n",
       "bedroomcnt                           3         3  \n",
       "bathroomcnt                        3.0       2.0  \n",
       "fullbathcnt                          3         2  \n",
       "bed_bath_ratio                     1.0       1.5  \n",
       "fireplacecnt                         0         0  \n",
       "age                                 29        60  \n",
       "calculatedfinishedsquarefeet      2184      1213  \n",
       "lotsizesquarefeet                10952      8549  \n",
       "has_taxdelinquency                   0         0  \n",
       "has_hottuborspa                      0         0  \n",
       "has_basement                         0         0  \n",
       "has_pool                             0         0  \n",
       "fips                              6037      6037  \n",
       "logerror                     -0.006433 -0.041282  \n",
       "taxvaluedollarcnt               471970    241620  \n",
       "county_Los Angeles County            1         1  \n",
       "county_Orange County                 0         0  \n",
       "county_Ventura County                0         0  "
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.head(11).T"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "048b01a0",
   "metadata": {},
   "source": [
    "## Predictions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "07294fa3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>actual</th>\n",
       "      <th>logerror</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1663334</td>\n",
       "      <td>0.04901</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>45302</td>\n",
       "      <td>-0.003143</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>388528</td>\n",
       "      <td>0.002243</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>130715</td>\n",
       "      <td>-0.02305</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>407342</td>\n",
       "      <td>-0.023093</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    actual  logerror\n",
       "0  1663334   0.04901\n",
       "1    45302 -0.003143\n",
       "2   388528  0.002243\n",
       "3   130715  -0.02305\n",
       "4   407342 -0.023093"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "predictions = pd.DataFrame({'actual': validate.taxvaluedollarcnt,\n",
    "                            'logerror': validate.logerror                           \n",
    "                           })\n",
    "predictions.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b2fc9135",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "<div class=\"alert alert-warning\">\n",
    "\n",
    "\n",
    "\n",
    "## Simple Model\n",
    "\n",
    "$$ y = mx + b $$\n",
    "\n",
    "$$y = f(x)$$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "id": "311af606",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.linear_model import LinearRegression\n",
    "\n",
    "# X must be 2-d\n",
    "X_train = train[['calculatedfinishedsquarefeet']]\n",
    "# y can be 1-d\n",
    "y_train = train.taxvaluedollarcnt\n",
    "\n",
    "# 1. make the thing\n",
    "lm = LinearRegression()\n",
    "# 2. fit the thing\n",
    "lm.fit(X_train, y_train)\n",
    "# 3. use the thing (make predictions)\n",
    "X_validate = validate[['calculatedfinishedsquarefeet']]\n",
    "predictions['simple_lm'] = lm.predict(X_validate)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "id": "fbba28a4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>actual</th>\n",
       "      <th>logerror</th>\n",
       "      <th>simple_lm</th>\n",
       "      <th>multiple_rfe</th>\n",
       "      <th>polynomial degree 2</th>\n",
       "      <th>polynomial only interaction</th>\n",
       "      <th>lasso_lars</th>\n",
       "      <th>glm</th>\n",
       "      <th>baseline</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1663334</td>\n",
       "      <td>0.04901</td>\n",
       "      <td>983390.124531</td>\n",
       "      <td>410135.180989</td>\n",
       "      <td>931010.714799</td>\n",
       "      <td>935951.199147</td>\n",
       "      <td>952006.761725</td>\n",
       "      <td>449783.851861</td>\n",
       "      <td>449783.851861</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>45302</td>\n",
       "      <td>-0.003143</td>\n",
       "      <td>294426.378284</td>\n",
       "      <td>410135.180989</td>\n",
       "      <td>293207.372986</td>\n",
       "      <td>302881.887929</td>\n",
       "      <td>304346.928579</td>\n",
       "      <td>449783.851861</td>\n",
       "      <td>449783.851861</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>388528</td>\n",
       "      <td>0.002243</td>\n",
       "      <td>303877.866500</td>\n",
       "      <td>410135.180989</td>\n",
       "      <td>242519.921994</td>\n",
       "      <td>258440.924719</td>\n",
       "      <td>243292.550466</td>\n",
       "      <td>449783.851861</td>\n",
       "      <td>449783.851861</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>130715</td>\n",
       "      <td>-0.02305</td>\n",
       "      <td>325268.076672</td>\n",
       "      <td>410135.180989</td>\n",
       "      <td>326422.853321</td>\n",
       "      <td>333055.259481</td>\n",
       "      <td>334969.014207</td>\n",
       "      <td>449783.851861</td>\n",
       "      <td>449783.851861</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>407342</td>\n",
       "      <td>-0.023093</td>\n",
       "      <td>517282.521475</td>\n",
       "      <td>410135.180989</td>\n",
       "      <td>499389.755074</td>\n",
       "      <td>504174.009760</td>\n",
       "      <td>507417.168764</td>\n",
       "      <td>449783.851861</td>\n",
       "      <td>449783.851861</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    actual  logerror      simple_lm   multiple_rfe  polynomial degree 2  \\\n",
       "0  1663334   0.04901  983390.124531  410135.180989        931010.714799   \n",
       "1    45302 -0.003143  294426.378284  410135.180989        293207.372986   \n",
       "2   388528  0.002243  303877.866500  410135.180989        242519.921994   \n",
       "3   130715  -0.02305  325268.076672  410135.180989        326422.853321   \n",
       "4   407342 -0.023093  517282.521475  410135.180989        499389.755074   \n",
       "\n",
       "   polynomial only interaction     lasso_lars            glm       baseline  \n",
       "0                935951.199147  952006.761725  449783.851861  449783.851861  \n",
       "1                302881.887929  304346.928579  449783.851861  449783.851861  \n",
       "2                258440.924719  243292.550466  449783.851861  449783.851861  \n",
       "3                333055.259481  334969.014207  449783.851861  449783.851861  \n",
       "4                504174.009760  507417.168764  449783.851861  449783.851861  "
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "predictions.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "18ab479c",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### Examine The Model\n",
    "\n",
    "- **intercept**: Value we predict if x is 0. Depending on our x variables, this may not be valuable\n",
    "- **coefficients**: define how our prediction changes as x changes. A one-unit increase in x causes our model to predict a `coef` unit change in y, holding all other independent variables constant"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "id": "173b63bc",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([248.7233741]), -17472.732831977657)"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lm.coef_, lm.intercept_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "id": "92abba76",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "taxvaluedollarcnt (y) = 248.723 * calculatedfinishedsquarefeet (X) + -17472.733\n"
     ]
    }
   ],
   "source": [
    "print('taxvaluedollarcnt (y) = {:.3f} * calculatedfinishedsquarefeet (X) + {:.3f}'.format(lm.coef_[0], lm.intercept_))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "id": "09ad5115",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x13d7dd0a0>"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAB18AAARACAYAAAChuJorAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/YYfK9AAAACXBIWXMAABYlAAAWJQFJUiTwAAEAAElEQVR4nOzdeXxV1b3///cOKINEIEJAGQIIFMQRiqjgDZbWqeXbirfSeyu2ltvambZWpYM/9dpvFf1eLdrb2lbFlg6XKuotbR0poKaKlFgZBAMSgqCEIQwRBMXs3x/xxAxn2HvttYdzzuv5ePR6Sc6wzj57yuezPp/luK4rAAAAAAAAAAAAAEAwJXEPAAAAAAAAAAAAAAAKAclXAAAAAAAAAAAAALCA5CsAAAAAAAAAAAAAWEDyFQAAAAAAAAAAAAAsIPkKAAAAAAAAAAAAABaQfAUAAAAAAAAAAAAAC0i+AgAAAAAAAAAAAIAFJF8BAAAAAAAAAAAAwAKSrwAAAAAAAAAAAABgAclXAAAAAAAAAAAAALCA5CsAAAAAAAAAAAAAWEDyFQAAAAAAAAAAAAAsIPkKAAAAAAAAAAAAABaQfAUAAAAAAAAAAAAAC0i+RsBxnH91HOdux3GedRxnv+M4ruM4vw3hfU5xHOc3juO87jjOYcdxdjiOs8xxnCtsvxcAAAAAAAAAAACAtjrHPYAi8UNJp0l6S9JWSaNsv4HjOJ+XdK+kg5L+LGmzpF6STpZ0saTf2H5PAAAAAAAAAAAAAB8g+RqNb6s56bpRUqWkJTZf3HGcs9SceF0j6ULXdbe3+/1RNt8PAAAAAAAAAAAAQEe0HY6A67pLXNfd4Lqu6/U5juP8m+M4SxzH2eM4ziHHcdY5jvNDx3G6pHn4bZI6Sbq8feL1/fd/N8DwAQAAAAAAAAAAAHhA5WsCOY5zn6QvqLla9mFJeyWdJelmSVMcx/mY67pH3n/sQEnnSvqHpLWO45wnaZwkV9I/JS1xXbcp6s8AAAAAAAAAAAAAFBuSrwnz/tqtX5D0iKTPuq77dqvf3SjpBklfkzT3/R+Pf/+/GyT9TdLkdi+52nGcaa7rbgxv1AAAAAAAAAAAAABoO5w8syQdkfSF1onX990sabekz7b6Wfn7/71M0mhJ0yT1lDRc0nxJp0j6i+M4R4c5aAAAAAAAAAAAAKDYUfmaII7jdJd0mqRdkr7lOE66hx1Wc5I1pVOr//6H67p/fv/f+x3H+dz7j/2wpEsl/SGMcQMAAAAAAAAAAAAg+Zo0vSU5kvqqub2wF3ve/+9hSX9t/QvXdV3Hcf5XzcnXM0XyFQAAAAAAAAAAAAgNbYeTZd/7/33JdV0n2/9aPefV9//b6LpuU5rXTCVnu4U2agAAAAAAAAAAAAAkX5PEdd23JK2VNMZxnDKPT1ul5jbFfRzH6Zfm9ye//9/NwUcIAAAAAAAAAAAAIBOSr8lzh6SjJd3vOE6v9r90HKe34zhjU/92XfeIpF+8/8/bHMcpafXYUyR9XtIRSQ+FOGYAAAAAAAAAAACg6Dmu68Y9hoLnOM6nJH3q/X/2l3SBpE2Snn3/Z7tc1/1uq8f/t6SvSmqQ9ISkLZLKJA2V9C+S5rmu++VWj+8uabGksyS9JGmpmteNvVTN7Yavdl33jlA+HAAAAAAAAAAAAABJJF8j4TjOjZJuyPKQOtd1h7R7zickfVnSmZJ6qTkRu0XSk5J+67ru+naP7y7pWkmfUXOS9pCkFZL+y3Xdx2x8DgAAAAAAAAAAAACZkXwFAAAAAAAAAAAAAAtY8xUAAAAAAAAAAAAALCD5CgAAAAAAAAAAAAAWkHwFAAAAAAAAAAAAAAtIvgIAAAAAAAAAAACABZ3jHkChchynVtKxkjbHPBQAAAAAAAAAAAAA3g2RtN913aF+n0jyNTzHduvWrWz06NFlcQ8EAAAAAAAAAAAAgDfr1q3T22+/bfRckq/h2Tx69OiylStXxj0OAAAAAAAAAAAAAB6NGzdO1dXVm02ey5qvAAAAAAAAAAAAAGAByVcAAAAAAAAAAAAAsIDkKwAAAAAAAAAAAABYQPIVAAAAAAAAAAAAACwg+QoAAAAAAAAAAAAAFpB8BQAAAAAAAAAAAAALSL4CAAAAAAAAAAAAgAWd4x4ApKamJjU0NKixsVGHDx+W67pxDwkA8p7jOOrSpYtKS0tVVlamkhLmGwEAAAAAAAAAwkXyNWZNTU16/fXXdfDgwbiHAgAFxXVdHTp0SIcOHdKBAwc0aNAgErAAAAAAAAAAgFCRfI1ZQ0ODDh48qM6dO6t///465phjSA4AgAVNTU06cOCAtm/froMHD6qhoUF9+vSJe1gAAAAAAAAAgAJGli9mjY2NkqT+/furtLSUxCsAWFJSUqLS0lL1799f0gfnWwAAAAAAAAAAwkKmL2aHDx+WJB1zzDExjwQAClPq/Jo63wIAAAAAAAAAEBaSrzFzXVeSqHgFgJA4jiPpg/MtAAAAAAAAAABhIeMHAChoqeQrAAAAAAAAAABhI/kKAAAAAAAAAAAAABaQfAUAAAAAAAAAAAAAC0i+AhGYPHkyrU8BAAAAAAAAAAAKHMlXFDwSnwAAAAAAAAAAAIgCyVcAAAAAAAAAAAAAsKBz3ANAdGrqG1W1cZfeOnREPbp21sThfTSyX2ncwwIAAAAAAAAAAAAKApWvRaBq4y5d9ovndf6dz+imRa/ov56q0U2LXtH5dz6jy37xvKo27op7iJKkBx54QJdeeqmGDRumbt266dhjj9XEiRP129/+Nu3jGxoa9IMf/EAnn3yyunfvrp49e+q0007T7NmzdeDAAW3evFmO42jZsmWSJMdxWv43efLkltdp/+/WPv/5z8txHG3evDnQWAEAAAAAAAAAAFD4qHwtcAtWbNH3Hl6tJjf971+sbdCM+5br1mmn6rLxg6IdXDtf+cpXdNJJJ+lf/uVfdPzxx2v37t3661//qhkzZujVV1/VzTff3PLY2tpanXfeeaqrq9O4ceP0la98RU1NTaqpqdGdd96pL3/5y+rVq5duuOEGPfDAA6qrq9MNN9zQ8vwhQ4ZENlYAAAAAAAAAAAAUB5KvBaxq466sideUJlea/fAqDejdTROH94lmcGmsWbNGJ554YpufvfPOO7rooot066236stf/rIGDBggSbr88stVV1enH//4x/re977X5jm7du1Sjx491LVrV914441aunSp6urqdOONN8YyVgAAAAAAAAAAABQH2g4XsLmLN+RMvKY0udJdizeEO6Ac2iczJenoo4/W1772NR05ckSLFy+WJK1cuVJ///vfdfrpp+u6667r8Jw+ffqoa9euiRgrAAAAAAAAAAAAigeVrwWqpr5RL9Y2+HrO8toG1dQ3amS/0pBGld2WLVs0Z84cLV68WFu2bNHbb7/d5vfbtm2TJL3wwguSpAsuuEAlJfHMH/A6VgAAAAAAAAAAABQPkq8FqmrjLuPnxZF83bRpk84880zt2bNH5557rs4//3z17NlTnTp10ubNm/XrX/9ahw8fliTt3btXkmJr6+tnrAAAAAAAAAAAACgeJF8L1FuHjkT6vKDuuOMO7d69W/PmzdPnP//5Nr/7wx/+oF//+tct/+7Vq5cke9WljuPoyJH0nzuV6DUdKwAAAAAAAAAAAIoHa74WqB5dzfLqps8LauPGjZKkSy+9tMPvli1b1ubfZ511liTpiSeeUFNTU87X7tSpkyTpvffeS/v73r176/XXX+/w8/fee0///Oc/A40VAAAAAAAAAAAAxYPka4GaOLxPpM8LasiQIZKkpUuXtvn5E088oXvvvbfNz8aNG6dzzjlH//znPzVnzpwOr7V7924dOnSo5d/HHXecpOZ1WtM588wztWXLFj355JNtfv6jH/1IdXV1gcYKAAAAAAAAAACA4kHb4QI1sl+pzhxaphdrGzw/Z8LQsljWe5Wkr371q5o3b54+/elP69JLL9WAAQO0Zs0aPf7447rsssu0YMGCNo//7W9/q8mTJ+v73/++Fi5cqMmTJ8t1XW3YsEFPPvmk1q9f35IknTJlih588EFNmzZNF198sbp166aKigrNmDFDkvTd735XTzzxhD75yU9q+vTpKisr09///nfV1tZq8uTJHZKsfscKAAAAAAAAAACA4kDlawGbNWWEShxvjy1xpG9OGRHugLI49dRTtWTJEp1zzjn661//qp///Ofav3+/Hn74YX35y1/u8PihQ4equrpa1157rRobG/XTn/5U9913n7Zs2aKrr75a5eXlLY/9j//4D33ve9/Tvn37dNttt+n666/Xfffd1/L7KVOm6NFHH9WYMWP0P//zP/r1r3+tIUOG6MUXX1RFRUXgsQIAAAAAAAAAAKA4OK7rxj2GguQ4zsqxY8eOXblyZdbHrVu3TpI0evToUMaxYMUWfe/h1WrK8jWXONKt007VZeMHhTIGAIhb2OdaAAAAAAAAAEDhGDdunKqrq6td1x3n97m0HS5w08cP1sDe3XXX4g1anqYF8YShZfrmlBGxrfUKAAAAAAAAAAAAFAqSr0Vg4vA+mji8j2rqG1W1cZfeOnREPbp21sThfWJb4xUAAAAAAAAAAAAoNCRfi8jIfqUkWwEAAAAAAAAAAICQlMQ9AAAAAAAAAAAAAAAoBFS+AgAAAAAAAAAAAIVgxzpp0zLpcKPUpVQaVimVj457VEWF5CsAAAAAAAAAAACQzzYtlZbdJtVVdfxdxUSp8lpp2OSoR1WUaDsMAAAAAAAAAAAA5Kvq30jzL0mfeJWafz7/Eql6frTjKlIkXwEAAAAAAAAAAIB8tGmptGiW5DZlf5zbJC36ZvPjESqSrwAAAAAAAAAAAEA+WnZb7sRritskLbs93PGA5CsAAAAAAAAAAACQd3asy9xqOJO655qfh9CQfAUAAAAAAAAAAADyzaZl0T4PnpB8BQAAAAAAAAAAAPLN4cZonwdPSL4CAAAAAAAAAAAA+aZLabTPgyckX5FXJk+eLMdx4h5GB0OGDNGQIUMK9v0AAAAAAAAAAEDCDKuM9nnwhOQrAAAAAAAAAAAAkG/KR0sVE/09p2JS8/MQGpKvyCu/+c1vtG7duriHAQAAAAAAAAAAEL/KayXHY7rPKZEqrwl3PFDnuAeACO1YJ21a1ryQcpfS5rLyPJvdMHjw4LiHAAAAAAAAAAAAkAzDJktT50qLZkluU+bHOSXS1LuaH49QUflaDDYtleZdLP3sLOnx66QlP2r+78/Oav75pqVxj1CS9Kc//UlTpkzR8ccfry5duuiEE05QZWWlfvazn7U8Jt2ar0uXLpXjOLrxxhv1j3/8QxdeeKF69uyp3r1769JLL9Xrr78uSdq0aZM+85nPqG/fvurWrZvOO+88vfzyyx3G8fnPf16O42jTpk264447NGrUKHXt2lUDBw7Ut7/9be3fv9/X5/rDH/6g8847T71791bXrl01evRo/ehHP9Lhw4cNtlJmDzzwgBzH0QMPPKCnnnpK5557rnr06KG+ffvqyiuv1N69eyVJL730kj7xiU+od+/e6tGjh/7P//k/2rx5s9WxAAAAAAAAAACAiIy9QprxSHNL4XQqJjX/fuyMaMdVpEi+Frrq30jzL5HqqtL/vq6q+ffV86MdVzu//OUv9clPflKvvPKKpk6dqquvvloXX3yx3n77bc2bN8/Ta6xYsULnnnuuJOmLX/yizjzzTD388MOaMmWK1q9frzPPPFNbt27VFVdcoY9//ONatmyZPvaxj+mtt95K+3rf/va3dfPNN6uyslKzZs1Snz599JOf/EQf+chHdOjQIU9jmjlzpv793/9dGzdu1LRp0/S1r31NZWVluv7663XhhRfqyJEj3jaQD3/605/08Y9/XH379tWXv/xljRgxQg888IA+9alP6YUXXtCkSZN05MgRzZw5UxMnTtSiRYv08Y9/XE1NWWbEAAAAAAAAAACA5Bo2WbryL9JXX5AunCOd98Pm/371heafU/EaGdoOF7JNS3OXmUvNv1/0TanXoNgOvl/84hc6+uij9fLLL6u8vLzN73bt2uXpNf7617/qt7/9rT772c+2/GzmzJm6//77dc455+jqq6/WD37wg5bf3Xzzzfr//r//T/fdd59mzZrV4fWqqqr0z3/+UxUVFZKkW265RZ/+9Kf18MMP6/bbb9f111+fdTwPPPCA7r//fl1yySX63e9+p27durX87sYbb9RNN92k//7v/0773kH86U9/0uLFi1VZWSlJampq0gUXXKCnn35aF198sX75y1+m3UaLFi3SJz/5SatjAQAAAAAAAAAAESofnXdLThYaKl8L2bLbcideU9wmadnt4Y4nh86dO+uoo47q8PM+ffp4ev6kSZPaJBUl6XOf+5wkqWfPnpo9e3ab311xxRWSpH/+859pX2/WrFktiVdJKikp0e23366SkhLdf//9Occzd+5cde7cWffff3+bxKskXX/99TruuOP0u9/9Lufr+PVv//ZvLYnX1LhnzGhuJXDyySd32Ea5tgMAAAAAAAAAAAC8ofK1UO1Yl7nVcCZ1zzU/L4YZEZ/97Gd19dVXa8yYMZo+fboqKys1ceJE9e3b1/NrfPjDH+7wsxNOOEGSdPrpp6tTp05tfjdgwABJ0tatW9O+XusEZsqwYcM0aNAgbd68WXv37lWvXr3SPvfgwYN6+eWXW1oVp9OlSxetW7cu08cxlm07jBs3rsPvcm0HAAAAAAAAAAAAeEPytVBtWmb+vBiSr9/5znfUp08f/exnP9Ndd92ln/zkJ3IcR5WVlbr99tvTJhTb69mzZ4efde7cOefv3n333bSv169fv7Q/79+/v+rq6rRv376Mydc9e/bIdV3t3LlTN910U86x22R7OwAAAAAAAAAAAMAb2g4XqsON0T7PgiuuuEIvvPCCdu/erb/85S+aOXOmnnnmGV1wwQXasWNH5OOpr69P+/Pt27dLSp/ITEn97owzzpDruln/BwAAAAAAAAAAgMJA8rVQdSmN9nkW9erVSxdffLF+9atf6fOf/7waGhr07LPPRj6OZcs6Vg9v2rRJr7/+uoYMGZKx6lWSevTooTFjxmjt2rVqaGgIcZQAAAAAAAAAAABICpKvhWpYx/VKQ31eQI8//riOHDnS4eepitfu3btHPSTNnTtXdXV1Lf9uamrSNddco6amJl155ZU5n/+d73xH77zzjr7whS9o7969HX6/Z88eVVdX2xwyAAAAAAAAAAAAYsSar4WqfLRUMVGqq/L+nIpJsaz3Kkmf+cxn1LVrV02aNElDhgyR67p69tlntWLFCo0bN04f/ehHIx/TxIkTdfrpp2v69Onq2bOnnnjiCb388ssaN26crr322pzP/8IXvqCVK1fqZz/7mU488URdcMEFGjx4sBoaGlRbW6tnnnlGV155pe65554IPg0AAAAAAAAAAADCRvK1kFVeK82/RHKbcj/WKZEqrwl/TBnceuuteuKJJ1RdXa2//vWv6tq1qyoqKjRnzhx95Stf0VFHHRX5mO6880498sgj+tWvfqXNmzfruOOO06xZs/Sf//mf6tq1q6fX+O///m9ddNFFuueee/T0009r7969Kisr0+DBg3XNNdfo8ssvD/lTAAAAAAAAAAAAICqO67pxj6EgOY6zcuzYsWNXrlyZ9XHr1q2TJI0eHVLFafVvpEWzsidgnRJp6l3S2BnhjCHPfP7zn9evf/1r1dbWasiQIXEPB4AFoZ9rAQAAAAAAAAAFY9y4caqurq52XXec3+ey5muhG3uFNOOR5pbC6VRMav49iVcAAAAAAAAAAAAgENoOF4Nhk5v/t2OdtGmZdLhR6lIqDauMbY1XAAAAAAAAAAAAoNCQfC0m5aNJtibUjTfe6Olxn/rUp3T66aeHOhYAAAAAAAAAAACYIfkKtPPAAw/ogQceiPQ9b7rpJk+PGzJkCMlXAAAAAAAAAACAhCL5CiSA67pxDwEAAAAAAAAAAAABlcQ9AAAAAAAAAAAAAAAoBCRfAQAAAAAAAAAAAMACkq8AAAAAAAAAAAAAYAHJVwAAAAAAAAAAAACwgOQrAAAAAAAAAAAAAFhA8hUAAAAAAAAAAAAALCD5CgAAAAAAAAAAAAAWkHwFAAAAAAAAAAAAAAtIvgIAAAAAAAAAAACABSRfkQibN2+W4zj6/Oc/3+bnn//85+U4jjZv3hzLuCBNnjxZjuMEfp0hQ4ZoyJAhwQeUEOm2y9KlS+U4jm688cbQ3jeJ29FxHE2ePLnNz2688UY5jqOlS5fGMiYAAAAAAAAAAOJA8hUAEsRWshsAAAAAAAAAAESvc9wDALK55ZZbNHv2bA0YMCDuoQA5nXnmmVq3bp369OkT2nssXrw4tNe26etf/7o+85nPaPDgwXEPBQAAAAAAAACAyJB8RaIdf/zxOv744+MeBuBJ9+7dNWrUqFDf48QTTwz19W3p06dPqEloAAAAAAAAAACSiLbDSLR0a762Xh928+bN+sxnPqM+ffqoa9eu+vCHP6w///nPGV/vD3/4g8477zz17t1bXbt21ejRo/WjH/1Ihw8f7vDYRx99VJdffrlGjhypY445Rj169NC4ceN01113qampKeNYN23apLvvvlunnnqqunXr1mEtzHRS63i+9dZb+va3v61BgwapW7duOv300/Xoo49Kko4cOaIf//jHGjFihLp27aoTTzxRP/3pT9O+XlNTk+655x6NHz9ePXr00DHHHKPx48fr5z//edqxS9L//M//aNy4cerWrZvKy8s1Y8YMvfHGG1nH/cQTT+jiiy9Wnz591KVLF5144om65pprtHfv3pyfOZfUOqJvvPGGZsyYofLycnXr1k3jxo3T73//+w6Pb73e6osvvqiPf/zjKisr67D/+NkHJH/bJduarw0NDfrBD36gk08+Wd27d1fPnj112mmnafbs2Tpw4EDLfr1s2bKWz5/6X+t9KNOar4cPH9att96qU089Vd27d9exxx6rc889V3/84x87PDbIMeRVpjVfU59n165d+tKXvqTjjz9eXbp00ZgxYzRv3ryMrxfmvgYAAAAAAAAAgC1UviJv1dXV6cwzz9SwYcM0Y8YMNTQ0aMGCBfrkJz+pp59+Wuedd16bx8+cOVP333+/Bg4cqGnTpqlXr1564YUXdP3112vx4sV66qmn1LnzB4fE7NmzVVJSogkTJmjAgAHat2+f/va3v2nWrFlasWKF5s+fn3Zcs2bN0rPPPquPf/zjuvjii9WpUydPn+fdd9/Vxz72MTU0NOiTn/yk3nnnHf3hD3/QpZdeqieffFI/+9nPtHz5cl100UXq0qWLHnzwQX3jG99Q3759NX369DavNWPGDP3+97/XoEGD9B//8R9yHEePPPKIvvrVr+q5557T7373uzaPv/POO/Wd73xHvXr10hVXXKFevXrpiSee0DnnnKOePXumHe9//ud/6oYbblBZWZk+8YlPqLy8XKtWrdL/+3//T3/961/1/PPP69hjj/X02TPZs2ePzjnnHPXq1UtXXnml9u7dqz/+8Y/67Gc/q23btumaa67p8Jznn39et9xyiyZNmqQvfOEL2rVrl44++mhJ/vcBk+2STm1trc477zzV1dVp3Lhx+spXvqKmpibV1NTozjvv1Je//GX16tVLN9xwgx544AHV1dXphhtuaHl+umRra++8844uuOACLVu2TKNGjdLXvvY1HTx4UA899JCmT5+uf/7zn/rxj3/c4Xl+jyFb9u7dq4kTJ+roo4/Wv/7rv+rQoUN66KGH9IUvfEElJSX63Oc+1+bxUexrAAAAAAAAAABY4bou/wvhf5JWjh071s3llVdecV955ZWcjyt0tbW1riT3c5/7XJuff+5zn3MlubW1tR0eK8m98cYb2zz+8ccfdyW5F110UZufz5s3z5XkXnLJJe7Bgwfb/O6GG25wJbk/+clP2vx848aNHcb53nvvuVdccYUryX3hhRfSjvWEE05wN23a5PWju67ruhUVFa4k9xOf+IR76NChlp8/88wzriS3d+/e7oc//GF3z549Lb977bXX3KOOOso9/fTT27zW73//e1eSe8YZZ7iNjY0tP3/rrbfccePGuZLc3/3udy0/r62tdY8++mi3d+/ebbbze++9506bNq1lW7f2t7/9zZXknn322W3G5LofbOtvfetbHT5jRUWF522Set9Pf/rT7nvvvdfy802bNrm9e/d2jzrqKPe1115r+fmSJUtannPPPfd0eD2/+4DJdkmN4YYbbmjz83POOceV5P74xz/uMK6dO3e6b7/9dsu/KysrO7xua+m2449//OOW/f7dd99t+Xl9fX3LvlVVVdXms/k9hrKR5FZWVrb5WWqbLlmypMNjJbkzZ850jxw50vLztWvXup06dXJHjx7d5vEm+1o6nGsBAAAAAAAAAF6NHTvWlbTSNcgR0nY4H9x4o+Q43v73pS91fP6XvuT9+WnapWrqVP/PiUBFRYV++MMftvnZBRdcoMGDB+vFF19s8/O5c+eqc+fOuv/++9WtW7c2v7v++ut13HHHdagGTbe2ZklJiWbNmiWpuQ1qOtdee62GDh3q+/NI0k9+8hN16dKl5d/nnnuuhg4dqj179mjOnDnq1atXy++GDRumiRMnavXq1Xrvvfdafn7//fdLkm699Vb16NGj5efHHHOM5syZI0m69957W37+u9/9Tu+8846+8Y1vtKmwLCkp0e23366Sko6nibvuukuS9Ktf/arNmKTm9sunn356h+1polOnTpozZ06bMQwdOlTf/OY39e6776atPj799NN11VVXdfi5333AZLuks3LlSv3973/X6aefruuuu67D71PtfoO4//775TiO7rjjjjaVu+Xl5br++usltf3OU/wcQzZ1795dd9xxR5uq8JNOOkkTJ07UunXr1NjY2PLzqPY1AAAAAAAAABHbsU564R5p2e3N/92xLu4RAVbQdhh56/TTT0/b0nfQoEF6/vnnW/598OBBvfzyy+rTp49+8pOfpH2tLl26aN26tif23bt36/bbb9df//pXbdq0SQcOHGjz+23btqV9rTPPPNPnJ2nWq1evtAnfE044QbW1tRo3blyH3w0YMEDvvfeetm/frgEDBkiSqqurVVJSknat2crKSnXq1EkvvfRSy8+qq6tbftfesGHDNGjQINXV1bX5+fPPP6+jjjpKDz74oB588MEOz3vnnXe0c+dO7d69W8cdd1z2D57F4MGD0yayJ0+erJtuuqnN50hJt/1N9gGT7ZLOCy+8IKk5qek1YetHY2OjNm7cqAEDBmjUqFEdfv+Rj3xEktJuK6/HkG0jRoxI2yZ40KBBkprbEpeWlkqKbl8DAAAAAAAAEJFNS6Vlt0l1VR1/VzFRqrxWGjY56lEB1pB8Rd5qXwWX0rlzZzU1NbX8e8+ePXJdVzt37tRNN93k6bX37t2r8ePHq7a2VmeeeaauuOIKlZWVqXPnztq7d6/mzp2rw4cPp31u//79fX8WSRnXEE1VMqb7fep37777bsvP9u3bp7KyspZ1Tts/vk+fPtqxY0ebx0tSv3790r5///79OyQZd+/erSNHjuTcnm+99VaghFi2MUkfjD3d71oz2QdMtks6e/fulaSW5LhtqXEef/zxaX+f+nlqHK15PYZsy/a+ktpUcke1rwEAAAAAAACIQPVvpEWzJDdD/LGuSpp/iTT1LmnsjGjHBlhC8jUf3HhjsNa+v/xl8/9MLVpk/twESCUtzzjjjJZqxlzuvfde1dbW6oYbbtCN7bb9888/r7lz52Z8ruM4xmO1oWfPnmpoaNC7776ro446qs3vjhw5ol27drWpOkxtn/r6eo0ZM6bD623fvj3tezQ1NamhocHy6Nuqr69P+/PUmNIlpNNtf5N9wGS7pJNKNGaqlA4qNc5M43nzzTfbPC7fRLWvAQAAAAAAAAjZpqXZE68pbpO06JtSr0FUwCIvseYrCl6PHj00ZswYrV271nMCZ+PGjZKkSy+9tMPvli1bZnV8tp1xxhlqamrSM8880+F3zzzzjN577z2NHTu25Wep/z/d59q0aZNef/31Dj8/66yztGfPHq1du9biyDvasmWLNm/e3OHnS5culdT8Wb0w2QdMtks6Z511lqTmNYK9VJOm2gC3rv7MprS0VCeeeKK2bdumDRs2dPj9kiVLJKnNd55PotrXAAAAAAAAAIRs2W25E68pblPzWrBAHiL5iqLwne98R++8846+8IUvpG2/umfPnjYVkUOGDJH0QZIv5aWXXtItt9wS4kiD+8IXviBJ+t73vqeDBw+2/PzgwYOaPXu2JGnmzJktP//sZz+ro446SnfffXebRGdTU5OuueaatAnDb3/725KkL37xi3rjjTc6/P7AgQMta50G8d577+m6665rM4ba2lrddddd6ty5sy6//HLPr+V3HzDZLumMGzdO55xzjv75z39qzpw5HX6/e/duHTp0qOXfqda5W7Zs8fjJmr9z13V1zTXXtEna7tq1SzfffHPLY/JRVPtaEtXUN2peVa3uXrxB86pqVVPfGPeQAAAAAAAAADM71qVf4zWbuueanwfkmbxqO+w4znGSLpH0cUmnSBog6R1JqyXNkzTPdb1Om5Acxxko6T8lXSjpOElvSnpU0k2u6+6xOnjE6gtf+IJWrlypn/3sZzrxxBN1wQUXaPDgwWpoaFBtba2eeeYZXXnllbrnnnskSVdccYVuv/12fetb39KSJUs0YsQIbdiwQX/+8581bdo0LViwIOZPlNm///u/63//93/1xz/+UWPGjNGnPvUpOY6jRx99VLW1tbrsssv02c9+tuXxQ4YM0a233qqrr75aZ5xxhqZPn66ePXvqiSee0N69e3Xqqadq1apVbd5jypQpuvXWW/W9731PI0aM0MUXX6yhQ4fqrbfeUl1dnZYtW6ZJkybp8ccfD/RZTj31VC1fvlzjxo3T+eefr3379mnBggXau3evbrvtNp144omeX8vvPmCyXTL57W9/q8mTJ+v73/++Fi5cqMmTJ8t1XW3YsEFPPvmk1q9f35LwnzJlih588EFNmzZNF198sbp166aKigrNmJF5fYPvfve7euyxx/S///u/Ou2003TxxRfr4MGDevDBB7Vjxw5de+21mjRpkudtlSRR7WtJUrVxl+Yu3qAXaztWaZ85tEyzpozQxOF9YhgZAAAAAAAAYGiTYUfJTcuk8tF2xwKELK+Sr5I+Lennak6SLpG0RVI/SdMk3SvpIsdxPu26rpvrhRzHOVHS3yWVS/pfSeslnSlplqQLHceZ6Lru7lA+BWLx3//937rooot0zz336Omnn9bevXtVVlamwYMH65prrmlTRXnCCSfo2Wef1ezZs/Xcc8/piSee0KhRo/Szn/1MH/3oRxOdfJWkP/zhD6qsrNT999+vX/ziF5Kk0aNH6+qrr9ZXvvKVDo//zne+o+OPP1633367HnjgAZWWluqCCy7Qbbfdpn//939P+x7XXXedJk6cqLvuukvPPfec/vd//1c9e/bUgAED9KUvfSnj8/zo3bu3HnvsMV177bWaN2+e9u/fr5NOOknf/e53jV7fzz4gmW2XdIYOHarq6mrddtttevTRR/XTn/5UXbt21ZAhQ3T11VervLy85bH/8R//obq6Ov3P//yPbrvtNh05ckSVlZVZk69HH320nnrqKd1xxx36/e9/r7vvvludO3fWaaedpp/85Cf6t3/7N9/bKkmi2NeSYsGKLfrew6vVlOEq9mJtg2bct1y3TjtVl40fFO3gAAAAClxNfaOqNu7SW4eOqEfXzpo4vI9G9iuNe1gAAACF4bBhVzfT5wExcjzkKRPDcZyPSDpG0l9aV7g6jtNf0ouSBkn6V9d1F3p4rScknS/pm67r3t3q53dI+rakX7iu++UAY105duzYsStXrsz6uHXrmkvmR49m5gbQmuM4qqys7ND6GTCRD+faqo27NOO+5RkTr62VONL8mROogAUAALCAziMAAAAReOEe6fHr/D/vwjnSWcapGsDYuHHjVF1dXe267ji/z82rylfXdf+W4efbHce5R9L/lTRZUtbkq+M4w9SceN0s6b/b/foGSV+SNMNxnKtd1z0QcNgAAOQ0d/EGT4lXSWpypbsWbyAICAAAYlMoVaJ0HkmWQtmvAABAGsMqo30eEKO8Sr7m8O77/z3i4bEfef+/T7ZfI9Z13UbHcarUnJw9S9Jie0MEAKCjmvrGtJUW2SyvbVBNfSPBKFhDsBMA4EUhVYlWbdyVNfGa0uRKsx9epQG9u+XNZ8s3hbRfAQCADMpHSxUTpboq78+pmMR6r8hLBZF8dRyns6Qr3v/n4x6e8qH3/1uT4fcb1Jx8HakcyVfHcTL1FR7lYRwAAKhq4y7j55EcQ1AEOwEAXhValSidR5Kh0PYrAACQReW10vxLpLY1cek5JVLlNeGPCQhBSdwDsORWSSdL+qvruk94eHzP9/+7L8PvUz/vFXBcAAy5rst6rygabx3y0rTB3vOAlAUrtmjGfcszVl6ngp1/XPF6xCMDUFPfqHlVtbp78QbNq6pVTX1j3ENCkfNbJWo6uSwqQTqPwJ5C268AAEAOwyZLU+c2J1azcUqkqXc1Px7IQ3lf+eo4zjclXS1pvaQZtl72/f/mnAObaaHd9ytix1oaDwCggPXoanY5Nn0eINFqEfmlmNpiU41uXzHtP2EqtCrRpHceKZb9ttD2q2JULPsqAMCisVdIvQZLy26X6p7r+PuKSc0VryRekcfyOmrrOM7XJM2V9IqkKa7rep22mqps7Znh98e2exwAAKExDSAReMpfSQhSEexEPiikRKSX457Wm3YV0v4Tt0Jcnz6pnUeKab8txP2qmBTTvgoACMGwyc3/27FO2rRMOtwodSmVhlWyxisKQt4mXx3H+ZakOyWtUXPidYePp7/6/n9HZvj9iPf/m2lNWABAnnBdj9mlGI3sV6ozh5b5Cj5NGFpG0EmZkxlJSG6mk5QgFcFO5INCSUR6Pe6pRrerUPafpEh6laiJJHYeKbb9thD3K7+Ses+aS7HtqwCAEJWPJtmKgpSXyVfHca5T8zqv/5T0Mdd1/d6xL3n/v+c7jlPiuh+s7uw4TqmkiZLelvSCheFm5TiOXNdVU1OTSkoKZQleAEiOVPLVcZwcj4zXrCkjNOO+5Z4qEUsc6ZtTRuR+YAHLlswo7dpZjWmqUuKegZ+kIBXBTiRdoSQi/Rz3D1VvpRrdkkLZf5IkqVWiQSSt80gx7reFuF95lZQJeSaKcV8FgKJCJSpgRd5l+xzHuV7NideVaq54zRg9dBznKMdxRjmOc2Lrn7uu+5qkJyUNkfS1dk+7SdIxkn7juu4Bm2NPp0uXLpKkAwdCfysAKEqp82vqfJtUE4f30S3TTlFJjhxxiSPdOu3Uog5gLFixRTPuW56xcjNd4lX6IMnxxxWvhzm8tPwGqUyTo14Vc7AT+cGkLXZYauobNa+qVncv3qB5VbWqqW/09Dw/x/11D68yrkZHR0nafwpFEqtEg0p1HvEjzM4jxbjfFuJ+5UWue9k471m9KMZ9FQCKwqal0ryLpZ+dJT1+nbTkR83//dlZzT/ftDTuEQJ5Ja/uWB3H+Zyk/5T0nqRnJX0zTSXTZtd1H3j//x8gaZ2kOjUnWlv7qqS/S7rLcZwp7z9ugqTz1Nxu+Af2P0FHpaWlOnTokLZv3y5JOuaYY+Q4TuIrtAAgyVzXleu6OnDgQMv5tbQ0+dV608cP1sDe3XXX4g1aniYYM2Fomb6Z4FnwUfCazMgkrhn4SVtftViDncgPSWmLHbQqyc9xb9ohn2r0jpKy/xSapFWJ2pKUziPFut8W6n6VTb5XjRbrvgoABa/6N9KiWdIHDULbqquS5l8iTb1LGjsj2rEBeSrfImhD3/9vJ0nfyvCYZZIeyPVCruu+5jjOh9WczL1Q0sWS3pR0l6SbXNf1dzdpqKysTAcOHNDBgwe1devWKN4SACLX5LpqanLlSnIklZQ4Kolwkkn37t1VVuavsiEuE4f30cThffJ2/aew+UlmZBJ1u84kBqmKMdhZiAr1PJGEtthe2gVffu9yfftjI9MmY0yOexNUo3eUhP2nEBXq+vSpziO5kmFhdx4p1v22UPerbJI2Ic+vYt1XAaCgbVqaPfGa4jZJi74p9RokDZscxciAvJZXyVfXdW+UdKOPx29Wc5w/0+9fl3Rl0HEFUVJSokGDBqmhoUGNjY06fPhwy/qEAJDv3jp8RPX7D+nA4fc6/O6YLp3U79iu6tElnEuR4zjq0qWLSktLVVZWlnfrao/sV0qAoh2byYwoZ+AnMUhVjMHOQpLP68R5EXdbbK9VSa6kO56q0V9Wv6n/7xMntdnmYbcOT6EavaO4959ClpQqUduS0HmkmPfbQt2v0knihDy/inlfBYCCtey23InXFLdJWnY7yVfAA/5aT4CSkhL16dNHffrkb5AMANrLVTUkfVBFcNn4QdENDHnLdjIjqhn4SQ1SFVOwMw5hVaV6qciccd9y43NrEqpp426L7bfC/tXtjR22eVRB5lQyKAnfW1LEvf8UMhtVokndV+PuPFLM+21Sqo9bC2s/SOKEPL+KeV9F8UnqNQuwase65pbCftQ91/y88tHhjAkoENz9AACsy/e1jJBMtpMZUSVHkhqkSmKwsxCEWZUa5rk1SdW0cbbFNq2wb7/NowgyD+zdTfOf36x/vr5Pq7ft6/D7QqiCNkFb9XCZVokm6RyTTVydR4p9v01C9bEU/n6a1Al5fhT7vorikC/XLMCKTcvMn0fyFciK5CsAwLp8X8sI0fI6o9h2MiOqGfhJDlIlJdhZKMKuSg3r3Br2uP2Ksy12kAr71tvc9Jhx1NzO2Iute97W/Be2ZPx91N9bUtBWPXx+q0STdo5JIvbb+KuPo9hPkzohzw/2VRQ6rlkoOocbo30eUESScwcHACgIhbCWEaLhd0ax7QRgVAnFpAep4g52tpaEMZgKu+I/rHNrUjsVxNUWO2h1UettbnLcTxs7wNP34VWxdpigrXo0vFSJJvUck0T5vN/avH7HUX0c1X6a5Al5fuTzvgpkwzULRamL4TXX9HlAESH5CgCwqhDWMkL4TGYUmyQzMol6Bn4+BKniarUoFUZrr7Ar/sM6tya1U0FcbbFtVBeltrnJcT9xeJ+s1egmirHDBG3VkyOp55gkmji8j2ZOGqpfPVub9XFJ2m8L4fotRbefJn1CnlecY1GouGahKA2rjPZ5QBEpiXsAAIDCkoS1jGrqGzWvqlZ3L96geVW1qqmnHUqS+J1R3DrpNGvKCJU4wd4/juRmKkiVa+zFGKRasGKLZty3PGMgMpWI/+OK1yMemXdBqlK9CuPcGsW4g5g+frDmz5ygCUPL0v5+wtAyzZ85wVPLN6/XBRvH3jM1O1VT32h83E8c3kcLrjpbT377X3TD1JM0oFe3wGOK8ntLCpv7TzEI494p6eeYpFmwYovuey574lWSZk4amoj9thCu31L0+6mfe9kkV41yjkWh4ZqFolU+WqqY6O85FZNY7xXwgMpXAIBVca5lVCiz7wtdkBnFXmfaZ9I6yRF1i1vWV+2oUFp7RVHxH8a5NR86FQRti+33umCjwn7Jqzu15NWdLa8/f+YEo+M+9fluWvSK8VhaK8YOE0lqq55UYd475cM5Jim8Xg8l6b7najX5Q+WxXg+TdP0OenxHvZ8WUtUo51gUEq5ZKGqV10rzL5HcptyPdUqkymvCHxNQAEi+AgCsimstI5M2toiejXUrcyUxMxnVv1TXf+IkSdJlv3g+liQ9Qaq2CqW1VxQV/2GcW5PQqcArk7bYptcFP+2Cs2n9+guuOtvouDcNBKYTx/eWFHG2VU+ysO+d8ukcE7d8ux4mYby2Jg7EsZ8W2oQ8zrEoBFyzUNSGTZamzpUWzcqegHVKpKl3NT8eQE4kXwEAVsWxllGSZt8jO1szitsnMZ9/bbeeeqVe2XaBmvpGPfLSVj1cvS2UQLOfxApBKjuJ+KSIouLf5Nx66sCeqtq4S0+s2Z52n4yzU0HYglwXglbYZ3t9v/uuzYBePnxviE4U906FfI6xKd+uh0kYr82JA3Htp0zIA5KFaxaK3tgrpF6DpWW3S3XPdfx9xaTmilcSr4BnXCEAANb5qRqysZZREmbfwxvTRMIzNTvTBqNG9ivVzsbDuvnPr2RNvErN3/1DK7flfC+/gWbaXZsppNZeUVX8+63IXLV1n1Zt3dfmZ633ybg6FXgVJCAd9LqQqkr60Z9f0brtwdbyCnLdsRnQ4zyE1qK4d0r6OSYpor4e5lub3nSvY3PiQNz7KRPygGSI+1wAJMKwyc3/27FO2rRMOtwodSmVhlWyxitggOQrAMC6KNcySsLse3hnmkhov4Zi633GTwDZK6+BZtpdmyuk1l5RVfzbqMhsv09G3anAi6ATGmxdFyYO76PHvvUvmru4Rj95akPOCR5+X98LWwG9KL435I+o7p3i6IaSj6K6HsbdpnftG/uNntee7YkD7KcAJM4FQBvlo0m2AhaUxD0AAEBhmj5+sObPnKAJQ8vS/n7C0DLNnzkhcEIqyOx7RC9oIiGVOPrjitclmQWQvUoFmjPxW3nBPtdWobX2mjVlhEocb48NUvGf69zqRet9Mqpxe7VgxRbNuG95xuO6/TkgHdvXhVlTRuq3/xFsm2d7/WxSgcAgovjekF+ivHdK2jkmiaK4Hto4t5q8b2srLNyvBZk4kA37KdCspr5R86pqdffiDZpXVZvz2Ck0nAsAADYlM3oFACgIUaxlVEjVc8XAZEZxe63byIUdEMjWIo9218EUWmsv04p/k/NjpnPrwuqtWrPNW2VRap9ccNXZkXUqyMVWK8kwrgutt/ktf12nJa/utPr62fhtN91aFN8b8k+U905RdkPJV2FfD5PSpreu4WDgzjNhtTxmP0WxYxmVZpwLAAA2kXwFAIQuzLWMCq16rhgESSSkpBJHk0L+gzdToJl218EVYmuv1Dqhdy3eoOVpPteEoWX65vvBKxtBrtbn1pr6Rt206BVf403tk37G7ZVJUtnWhIYwrwsj+5XqX0b2NUq+mo7LtN20yfdWzMKcKJYUqc9YvWWP0fNN9+EwzjEmkvodh309DKNNb0VZd9U1HPT2oq0EXfc1zIkDSdlPgaixjEpbnAvMJfU6CwBxIfIMAMhrhVY9VwxsrFspNSeOxlb0tjewNDIFmsOqvCg2fhLx+dLay0vFfxhBrqD7pK1OBaZJZZsTGsK+LsRx3ckVCDxlwLE6fVBvlZd2IdjlUzFU+2T7jH4E2Q5RdEPJJB++47Cuh2FNFhs/pMwo+Rq080zYky7j3E9hH99jbrYr4wsF5wJ/8uE6CwBxIPkKAMhrhVg9VwxyJRK8C5C99SDTH4m0u7ajkFt7Zar4DyvIZWufDNKpIEhS2eaEhpH9SvWh/qV6dbv3tuR+rgtxXXcIBNpXDNU+uT6jV7buncLshpJOvnzHYV0Pw5osNmbAsXqo2v/rBu08E9Xkl6j3U9hFIsg7llHJjnNBbvlynQWAOJB8BYCYEUQNrhCr54qBjTUU17/ZqIG9umnr3retjy9boJl21/ZE1dorKefasIJcYeyTfrZZ0KSyzQkNC1Zs8bUetMl1Ic7rDoFAO6Ku9omr4tNG4jVf753yraIrjOthGJPFauobtaPxsNHrBt2+TLpELiSCvGMZFQSVb9dZAIgaEUAAiAkzcu0p5Oq5YjCyX6kG9u5m9FyThK0XuQLNtLu2y0tFn2niomrjLt3y2Dqt2ba/w++iPteGGeSyuU+aXJ+CJpVtJY9TQSDX41gcmV0XuO7kv6iqfeK83/PzGTPJ5304Hyu6bFe425yYE7R9ta0kKJMukQmJIH9YRgVB5eN1FgCiRPIVAGLAjFz7oqqeg30LVmzRb5dvCf19Shxp2tgBerh6W+BkCZUXwWQKKqer6AuSuPi/f3lFv3q2NuM4oj7X2g5ytd+OJw84Nm2SOZN0+6TJ9WnBii2Bk8q2ksd+k02j+pcaf/dcd/JXVNU+cd7vmXzG9vJ5H05iRZefhKqtCndb59ag7attJkG9Tn5xJH10dD/V1Deqb2kX7sGKAIkgf1hGBUEk8ToLAElD8hUAIsaM3PAkZT28uN8/n/itUjPVOoB8yRkDrSRLqLzwz28iNUjiIlfiNSXKc62tIFfVxl269bH1Wr1tn/FY0u2TJtenrXsOavbC1UZjaJ1UNp3QIEnzqmr11qEjOvjuEd9BoHXbG3MGgbKd05Ny3YE/UVT7xH2/Z/oZz/tQX/3LyL55vw8nqaIrzupnG5PFgravDqN6OtfkF0lyJT35Sr2efKVeEp2FCh2JIP9YRgVBJOk6CwBJxRUTACLGjNzwxbUeHq2k/bPREjGTgb26aea5QzsEkG0lS2g76o/fRGqQxEXVxl2eEq+tXyOKc62NIJfXpHI2mfZJv9enm//8imrqG2V6CLdPKvuZ0OBI2vf2uzr/zmcM3/0DmYJAfs7prMOaX6Ko9on7fs/0M75zpMnaGOKUlIquJHS7CTpZLMi9WpjV0+3v555/bbeeeqU+4zWJzkKFjUSQfyyjgiCScp0FgCQj+QoAEWJGbuFKQnAt39hoiZjN1r1vh97Sj7aj3pgkUoMkLm59bL3vMUZxrg0a5LKReM20T5ocj+u3NwYaS/tktJ9WkjbePyVdEIhzurl8qAIOu9onCfd7pp+x6rXdqnptt6T8njiWhIquuKufU4JMFjO9V/vK5GG65IyBkRz7I/uVamfjYd3851dyTgais1DhIhHkH8uoIIgkXGcBIOk44wFAhJiRW5iSElzLN6bHg9/3CPvYoe1obn4TqXMeX69VW/211E0lLiQZt+P1s7+YfN8mQS5J2tl4WDsbD/tOvJ46sKcuOWOApzFGcTy2l+48mGtCw9Djumtzw0GrrcrbB4E4p5vJp+4PYVT7tD4nrH0j/HNQLja2dT5PMkhCRVfc1c+tmU4WM702lJd2jfQeKEnbGvEgEWSGZVRgKgnXWQBIuuK+ywCAiDEjtzAR8DETxX4d5bFD29H0TKpm/CZeU4ImEL3sL0ETTLOmjNBn713ua1x3Ld6gA+/435dXbd2n//fp0zztl1FfZ7JVTrSuYG6fSK/dfdD6WIK2X+acnn+VwjarfbKdE/yyeRyaTvZoL65JBkEnNMVd0ZWE6uf2TCaL5cPfLknc1ogeiSAzLKMCU3FfZwEgH5B8BYAIMSO38MQZ8Mn3Ssso9uuD776neVW1ebuNCkGUFZVBg7259kkbCaa+pV18jytdlZJXXivporzO5KqcyLWdbWofBCKI75+fSuHrFq7Si5sbNOaEY2M/H9uo9rG9r9o+Dv18xmyinGRgs4I6zoquJHe78TNZLB/+dknytkZ0SASZYxkVmCqGyul8j7kAiBfRfACIEDNyC08cAZ98au2YTRRj/PnS1zr8LJ+2USGIsvolaLA32z5hqxVt1O19vW7/qI4HR9krJ7xuZxvSBYEI4vvnp1LYlfTQyq16aGXzv+M8Hwet9gljX7W9Hbx+Ri+imGRgu4I6zoqufKgY9SIf/nYplG2N4IohERSWicP7qG9pFz3y0latf7N5GY9Rxx+rS84YULT3N8itkCunCyXmAiBeJF8BIELMyC08UQd8omjtGNXsTlstEf2KahsxS7bZjsbDkb1XkD+ATxlwbNbvx1Yr2rXb9psO0YjXhHRUx+Otl56S9bjzs52DyBQEIojvj0mlcGtxtyMOUu1je18N634v12f0I8xJBmGttRxXRVc+VIx6EfXfLib3ToWyrRFcISeCwpQpybTk1Z1aWbeHJBOyKsTK6XxbTgNAcnG3CQARY0ZuYYky4BNWYLL160c9u9PP8eA4ktzmyqmgwtxGkqxsx0JJ3r70+p5I3qd1sNckiTj7otEZf2erFe2CFVu0sHqrr9cJys/+7ff6NLJfqdZvb/T8+hOGlmn6+MEZfx80kednHJmCQATx/bFRyR3XmqIpJutg2t5Xw77fa/8Zn36lXlWv7fb9OmFOMghzrWWT7ziofKgY9SqKv12C3IMW0rZGcIWYCAoTSSbYEMd1Nixhx1wAFJfijBIAQIyYkVtYogz4hBmYjOsPb7/HgyvXWpvHMLbR5fcul5Q5QexlO4adBLf1R7HX6t81EVR6tg/2+l3n8IvnDs3ZcthE6yqx1B/yERR1tjh1YE9f363f43FA725WA/JhtWT+yuQT1f2oTp72d4L4/thKxkW5pmgmftbBtLmvRnm/1/ozmiRfw5pkENVay36+46AKqdtN2H+7LFixRbMXZr4+5rp3KqRtDTsKKREUJpJMsC3K62xYwoy5ACg+JF8BIAbMyLUrzj+sowj41NQ36pGXtoYWmIz7D2+/x4Ot9onSB9tIUtZ9yOs28vJ3WrbtGGYS3FZSN9vrDOzVTeeN6qsZZw/RyH6lkaxvmi7Y62edwy+eO1Q/+PhJWR9joxVtkBalE4aWyZV8nwOuu3CU7/fyezzaDMiHUVU3YWiZr+1QiEH8MK+RNpNxUawpaoutfTWu+72kTTIo1LWWC6nbTVh/u/xi2Wu65bH1OR+X6x60kLY17CmERFCYSDIBbaWbDDbC2aqJJWvUQ2/rLXVTVdPJ2uAObPl9Pt2/AogeyVcAiAkzcoOLo01uOmEFfLJ9Pq+8BCaT8Ie3n+Mh02N3NB7Sz5du8v3eX/z1P1TXcLDDz1vvQ7bX9ku3HcNMgttK6uZ6na1739b8F7Zo/gtbdObQMg0u6+5pfKayBXtzBYpPHdhT1104ytM2DNqKNkiL0tbnDJvVvNn4OR5tBuRtV9WZBtgLJYgfxTXS9rUg6cm0FNN99cIx/TTmhJ6x3+8lbZJBoa61XGjdbmz/7bJgxRZPideUbPegYW9r/l5DoYmq4wCQT1pPBjunZI1mdX5YE0o6XqeWN43S3CPT9Pemk1uex3EBIB2SrwAQM2bkmknS+jRhBHxyfT6vcgUmk/aHt5/jof1j7168weg90yVepQ/2oW99dGQo61C2345hJcFtJXW9vk7Ki7UNWhHS+p1fmXyiLjljQM59JRUofnLtdj1cvVWNh46otGtnTRs7UOeP6e/5/YJWiZlWdTlqe86wWc3rhdfj0VZA3mbiIUgyoxASJlFdI02SeNmku2YlMelh+p1/5/wPxT72lKRMMqjauEsPrTRbCzsf1louxG43Nv52qdq4S7MfXu37ectrG3TDn9ZoyHHHdDgXeNnW08YO0IF3jujuxRs8nU+SMtETsK1QOw4AQaTuQy/rtES3dL5XnRxXris5zgePcV1pQsl6zT/qFs0+8kU9+N7kxE8GAxCf5P+1AgBAO3G3yU3HZnDNb5Irm1yByUL6wzuMIGyTK935VI31101Jbccwk+C2krom1b9hrW9aXtrF0/6XKWj6+Nr6Di2SswlaJWb6B/ml4wa2SYzlOs+cMuBYzb5odGyB4KABeVuJPBvJjHxOmAS5RpokOv2usZxN6/N4kpMeSascNWEyycB2IjzoRLMkHn/p0O2mo7mLN8g1/N5//fe6lv+//bkg07bufnQnLazepusWdkz4ZjqfJGmiJ2BboXYcAILo0bWzzilZ05J4ldomXlv/u5Pj6tbOv9I2t496dA0+6RVAYSL5CgDIO0lok5uOreCazRa3Ya2xmMQ/vMP6jsNKIEofbMewkuC2krpB2uaGwcv+57dFcq5ETpAqMdOJAWNOOLbDzwotiN/+c1w6doD+sbnB83a+ZdopOvjOe6Fsh3zd1ibXyNTzTBKdftZYziX1HvmQ9EhK5WgQXicZSNJlv3jeaiI86ESzpCWzvaDbTTOb9xSZzgWtt7XJ+SSJEz0Bm4IuqQEUoonD++ikzg+3JF5z6eS4+mbnR1Q2/KshjwxAvuKqCQDIK0lrk5tOkOCazYCUl8BkIf3hbbv9ZRRS2zGsJLitpK7p64TFS0W33xbJuRI5QVrRBm1bnE6+B/GzVTUO7XOMNu86kHXiQ2o7R5F4y6dtbXqNvPze5Rm3t5fjI1cSz4vR/Zu3c74kPQqhPbWUe5LBghVbNHvh6kD7RzpBJpqFnczOtwkX+cb2PUW2c4Hp+SSpEz0BW8K4NwXy3Uhnq1SyvkOr4UxcVzqrZJ3kbJU0OvTxAcg/yYucAgAkEfjJpJDa5KZjKyDlNTBZaH9422x/GYVUwHvtG/uMnp8rCWkrqZu0Sudc+59JUN9LIse0FW0htCi1KVcVUu2uA3Icaehxx6h214EOv09yy9+4mV5Dch0uXo6P9km8n/5to3YfeMf3GPwmPa5buEozJw2N5T4pn9tTt5duksEvlr2mWx5bn/O5fhPhQSaahZnMTnKr60ISxj1FpgSoSRK1b2mXxE/0BILi3hRIY9MySd4Sr20et2mZVE7yFUBHJF8BIGEI/GRXSG1y07ExTj+ByUL7w9tm+8uwje5fqh8+uiZQpW6u79hWZXOSKp1z7X9BgvpeqldMW9GG1aI03ybqeK1Ccl2pbvcBzbk0vLbCcQj7+wrzWue1uiv1eW5a9Iqv11+/vVFPrt3u+/jduuftlveK4z4pX9tT57JgxRZPidcUP9V/ppMEKsq668fTTsn6HqbfQz60ui4UYd1TtE+AmnYCeOSlbUbvny8TPYGUQmifD1h1uDHa5wEoeMmJpAEACPx4UEhtctMJOk6TKptZU0ZkbTnZmqPk/+GdqxKpoqy76hoO+n5dR/bWf3XUnGgI8npekuC2KpuTMuHDS+AnaPW41+oVv61obbcojXqijq3Ekt8qpIert2nBVWf7fp+kier7Cvta5/X4MD0OH67eavS8lDjvk/KpPXUuVRt3afbDq30/z+v+YTpJ4F/HDcx4nAQ5xvKl1XWhCHPbtU6Amp6H1r+53+h5+TLRE0gplPb5gDVdDO/jTJ8HoODlRyQaAIoAgR9vCq1Nbnum4/zK5BN1yRkDCibwG1S2SiRJOv/OZ3y/5rc+NkJzn86eOEp1Hsp2GDvvZ3GDJF69zj63VdmchPV0vQZ+bAQ/w6pe8TIxYPzQMh1450jWBEaUE3VsJg3zYc3uMET5fUVxrfNyfJgeh40Wjt9iv0+yYe7iDXINL1Je9g/TSQIH331P86pq065LG+QYS9L6nmFUUCetKjvMe4rW556ok6H5MtETaK2Q2ucDgQ2rjPZ5AAoed4cAkBBJCvwkWaG1yW3P9PNdd+Eo4/ecu3iD50SgK7v7XtgBwUyVSCbbeNaUkfpwRVnO4ISkrI/Z+/a7enW7eWsiv7PPbbQUq6lv1If69dCK2gZr1b9++An82Ah+hhmwbT8xYO22/VqxuUF1DQdb/vfQyubHpktuRjlRx3bSsNDX7E4n6olVUUyU8HJ8mB6HpZaSF8V8nxRUkNbtkrf9w/R7+fnS1zr8bFT/Ur3qoZNEpmMsKZNCwqiOT/JSJn7uTfxofe4xPQ+NOr5US17d6ft5nG+Qrwq1fT7gW/loqWKiVFfl/TkVk1jvFUBGJF8BIAGSEvjJF4W+Pk2Uny+ufS/ugKDpNvYanLBddZtiMvs8SEuxbN+TV6btmj93doWG9DnGd+AnH1q3Ss1Jspe27NHDL231ldyMaqJOGEnDQl+zO504JlaFldRI8XJ8mH6GaWMH6vG19UbPbS/TtYrgcnZBW7d72T9sThJY72MyU7pjLAmTQsKojk/6UiZe701MXjfd/+/HJWcM1Mq6vQU70RPIpJDa5wPGKq+V5l8iuU25H+uUSJXXhD8mAHmL5CsAJEASAj/5pNDXp/H7+fqWdknbhs+LOPa9KAKCXpKjpvuQ18B9ugDGvKpao89z4Zh++s75HzLe5jnb3R7XXeMr2ra7zfU9eZHafg9Vb/UdxLzpkycbvefIfqXG6/qmDOjVzfi5XtTUN2r+85v12xe2eKrWuu7hVVq9ba9KHCeyyRJhJA0Lfc3u9uKa3OL1/OY4Mmot6+WaatrF4fwx/a1W7rZf/zHbpJ9Lxw7QwXfeK/qkbNDJDmF0ZbCp/TEW96SQMCa65MtSJrnuTfxqnwAN0i3H5kRIJnwAQB4ZNlmaOldaNCt7AtYpkabe1fx4AMggPyMZAFBg4g785KNCX5/Gy+f7yKhyPVS9VdcuXNXh916rR6Pe98IOCPqpqPW7D9mo1jXdbv2O7aqqjbv0xJrtxoG7Du1u39ivF2sbtKXhoOp2N//voermx36of6lq6huN1/yT2m6/Ab27RVqt/tXzTtR1C1cbP/9L81eGUoFtWknsutL8F7YEet9s+0v7wPCAXt1CSRoW+prd7cU5scrr+c3v/uinuss0eWEzKZc653qZ9JPE1qxxCDLZwc/+EVbloxetj7G4J4WEMdEln5YyaX9v8rf1O/TsBv/nzkz3DkE6nQSd6Bl3lxcAgKGxV0i9BkvLbpfqnuv4+4pJzRWvJF4B5EDyFQASIO7AT74q9PVpsn2+l7bssVI9GvW+F2ZA0KSi1us+ZKta13S7/fr5ug4/Mw3ctbS7rc7c7tbvmrSnDuypS84YEEqlsYnp4wfrnmWbVLvrgPFr2G7JaKOS2FSmpL+NttLtXy/bubfQ1+xufx7ZvNts/7M1scrr+S2siRGmx73NpFyPrp09T/pJJ+h5IB/vT0zPv478T5yxXfnoVetjLM5JIWFUx+frUiapbiF3PFXj+7nZ7h2C3H8EmeiZ9LbPAIAchk1u/t+OddKmZdLhRqlLqTSskjVeAXhW3FF7AEiIYqsGsq3Q16dp//lsVo9Gue+FGRAMuk2y7UNJ2N7pmAbuqjbu0uyFq43WYc1k1dZ9+n+fPi3r9xR1tfqPPnWyLr93eaDPaaslY5Dkjw3pkv5hJIO9JA0Lcc1u20ls2xOrsp3fwp4YYXrc20rKTRzeRz98dE2g/dzkPJDPFW+m67HOudRs4kymSQI7Gg/p50s3+X49L1ofY3FOCgmjOj6flzJ5cu12NRpMPpk1ZUTW+6Ag9x8mEz3zpe0zAMCD8tEkWwEYI/kKAO3EUaVQ6NVA+SqpFSv51E6utTADgmFuE5uvbRrUzvZ+fgN3P3x0jdXEa4qX7ynKavWJw/vo1kuDV8/ZOIb87ENhaD/2sJLBXpKGhbZmdxhJ7Kg/c9gTI0yP+/bPu/fZWm3b+7bn950wtEySrJxv050HMn2eQqh489v6+XsXjQr8WdpPErh78YZAr5dN+305rkkhYSz9kM9LmTxcvdXoeeve3J/zMUHvP/xM9MzX+3QA4UlqXAEAEC6SrwDwvrirFAqxGihfxb0vZGO7ejTKComwAoJhVtSG8do21zOU/AXuFqzYEqgVbzZ+vt+oqtVtVc8Faclosg/ZlG6iTljJYK/nxUJZszuMJHZcE6uimBhhetynnjeyX6nv+yTTa1w6qfPAzsbDGe8RRvUv1avbG3NOcEl6xZvXSRKOI80JKYkc1tIa6Y6xuCaFhLH0Qz4vZWJS9er3eWHff+Rr22cA4UhyXAEAEL7477ABIAGSUKVQaNVA+SoJ+0I2tpOlUVZIhBUQNN0mdzz5qsac0DNrgiGM5LTN9QxTvAbufrbkNTtvmEYSArfptE8qPVOzU0te3en7dUxbMtpM/viVbqJOWMlgv0nDQliz23YSO/V9xblNktzG3+Q+qbpuj9Ux/PRvG/TnVW9mfP/1PtbLjqrizXR/inuSRBivm23yYhyfN4ylH/J5KZNSw/sI0+eFIZ/bPgOwK+lxBQBA+JJzlwoAMUnSujxxB7qKXZL2hUxsJ0ujrJAIKyBouk0eX1uvx9fWt/w73ezjtdtyt7IzGZOtiszWcgXuauobVddw0Mp7pTOgVzfNq6pNbBItlVR669ARo+RrvrVyzDRRJ4xkcJBuDElO9mVjO4ld4kgzJw2NvDoi35Lffu+TbE8KWfTym1bbtodZ8Waj2iaKSRKZXtt2m34vkxejnhQSxrIj+byUybSxA9vcl/l5XlLk270CgHDkQ1wBABA+kq8Ail7S1uUphGqgfJW0fSEd28nSKCskwgoI2gqut599vGDFFi00XH/M69qX6Y71zbsP6Nd/r/P9nrkCd2FWYJZ27awvzV/Z4edJbKcVdUvGOCqCs03UsR3gLdZuDDaPpwlDy3TqwJ6677nayKoj8rkNnp/7JNufIa71sv2yXW0TxiQJL/ugnzb9jiN9qF9p2upjv5MXo5wUEsayI/m6lMn5Y/qrtGtnX22ES7t21vlj+oc4Kn/y6V4BQHjyIa4AAAgfd3gAilqS1+XJ12qgfJXkfaE128lS08qSnY2HjT53GAFBm3+opmYf7zn4juY8vt440O5nTO2P9XlVtUbvuWHHW1l/H2ZVRaZA6Yu1Dbr8vuWaeurxGlFemoiJJFG3ZIw6kPK5cyp00/85OePvbQZ4syU0Cn0Ckenx9LmzKzSkzzFttsvOxsOezou2qiMKpQ2el/sk29WTYbB9bs6Hahs/+6CfVtOXjR+Ud+eeMJYdyeelTL5+3nDd8th6X49Pkny5VwAQnnyJKwAAwkfyFUBRY10epOTLvhBG9eisKSP02XuX+xqH6ezcMAKCtoPrTa700yUbjddyDNq+zzQAt+jlNzR9/KCMz4+rqsJ1pT+9/KakN1t+FmdlXdQtGaNO/gw57pisvzfd5r+cMU7b9r6dM6GRzxWV7WVL4pgeT0P6HKMrJw5t87MfPromsuqIfEjM2eZn0k8cbJ+bk15t43cfnD9zgubPnOC51XQ+Tl4MY9mRfF3K5KrKE1VT36iF1dtyPvbT4wbqqsoTIxiVd/nc9tmGfJv8AIQhX+IKAIDwkXwFUNRYlwcp+bQv2K4e7VvaxfcYgszODSMgaDu47qflXWs22veZJutcZQ+imwZYhx3XXZt2d1wr1m9rwNbirqyLuiWjr9aZkqaMLtfT63YYvVeu79k0MOylrWOhVFSGmUBu/7yoqyOSnpgLQ2rSz+yFq427GTgKp+WwZLfiLR+qbUz2wQVXnV3QS3LU1Deqpr5Rk4b30diKXpIcdT+qU+DPmK9LmfzXZadrZL9S/XTJxrT3GaVdO+vr5w1PXOI1JV/bPgdRSBOvgKDyKa4Qqx3rpE3LpMONUpdSaVilVD467lEBgFUkXwEUNdblQUo+7Qu2q0fjmJ1rOyDodZuEyZG99n2zpozQ5fctl+vzs2QLopsk3Yb1OUZ/++7kDt/TgF7d0q7x6oeXyrqwAsZRt2T0+36XjR+kuYtrdOdTG3y9j9fqmTACw4VSUekngWyjuinK828+JObCMn38YK3Ztk/zX9ji+7mOpKmnHf9+Bb9dtivekl5tE3QfzMeq1my8JKxsfN583G5XVZ6oqypP1JNrt+vh6q1qPHREpV07a9rYgYla4zWdfG77bKJQJl4BtuRTXCEWm5ZKy26T6qo6/q5iolR5rTRsctSjAoBQFMmZHQDSY10epOTbvmCzejTO2bk2A4K5tknYLh030FpQaeLwPvrEqcdrkUGwP1sQ3W8F5sThx+nuxRs6JD5N16VtL1NlXRQVFLn2l1MH9tRpA3uqfv8hzauqDZz49XrMStJlv3jed4LCT/WMSWA4VyK8ECoq/SaQr7twlP6xuSFQEjvK82/SE3Nhm3H2EKPk662XnqIzBve2nnwNo+It6dU2xb4PtkbCypvzx/QPLdkaZkVwvrZ99qtQJl6FLd+qzxFMvsUVIlX9G2nRLMltSv/7uipp/iXS1LuksTOiHRsAhIDkK4CiVuzr8uAD+bgv2KoeLaTZuem2ydo39unxtfWhv/eYE461+nojy0vVep1Ur7IF0f1UCLtSh0RFKvFpM1DfvrIuyoB0uv1lR+NhvbBpt1Zt3adVW/e1ebyNlrPZjtlcnz0Tk+oZG8ng1PboW9qlICoq/SaQ/7Z+R+DqJtPz6No39vnefklPzIXN9Do/ffxgSbK6dnNYFW9Jv54X+z6YQsIqXlG1yM3Xts9+FMLEqzDRjrk45WNcIRKblmZPvKa4TdKib0q9BlEBCyDvJS9qCgARK8Z1eWwqpIBCvu4LQatHkzw713T/ar1NauobI0m+5ksQPUiFcCrx+YlTTzAaWyapqqa4AtKp/eXqP/5TC6u3ZXycrcRvumPW62dvL0j1TNBkcND9IUnVbKbtUG/+1MmaP3OCcXWT6f77+Np6Pb623lfwNumJuShMGVXu+Xtuf523tbb4KQOO1eyLRodyDU3y9VxiH0whYRWfOCqO87HtsxfF3MreC6rbi1u+xhVCtey23InXFLdJWnY7yVcAea+w/ooBAAPFti6PLYU4k7dY94Ukzs61uX+ZfL7Srp3V6KPSJoztEWYQPVPF529fqFOuGEGTKy16+Q2jsWWSqmqKKyBdtXGXrv7jy9q+/5Cn9w2jEsnPZ5ekiuO661dXfDi09QD9JMJN94ckVbMFaYd65cShxtVNJuen1vwEb5OemAvbghVbNOfx9Z4em+46b2tt8dXb9mvu4g0tr2lTEq/nKVUbd+nhLJNbsimUfVAiYRUnKo7too14ZuxrKNa4QkY71qVf4zWbuuean1c+OpwxAUAESuIeAAAkwfTxgzV/5gRNGFqW9vcThpZp/swJzEp934IVWzTjvuUZg0epYPAfV7we8ciCK8Z9oWrjLu17+13Pjw97dm4Y+9esKSNU4nh7bIkjff284b4eH8b2SAXR/fAbRB/Zr1RXThyqb0wZoVfrG3MmXlNcNSeobenRtXOggLQfNfWNmldVq7sXb9C8qlrNXVyjy+9d7inxmpJK/Npi8tnrdh+09v7p+EkGm+ahklTNZqMdauvj6cqJQz0fi37OT+mkgre5AuFRnFOSym9l+XUXjkp7nc91j+BVmPdJfq93tq5f7c+trc/Nqev66m37srxCeoWyD6YESVghGJMJXsiMNuKZsa9BKs64gnask164p7lq9YV7mv8tSZuWmb2e6fMAICGSE/EAgJgVw7o8NhTDTN5i2hf8rjEZ9uzcsPYvv7OPLxs/SL26HxX7bOWwWla137cH9OrmO/nXeOiIHEdyA7bglJq/n7ArKLJVU5uwWYkUdfVIrnObSTLYRJKuDXG2Q7VRUdnkSnMeX69LzhiQ9ZpVrG3w/FaW/239Dl1VeWLa36XuERas2KLZC1cbTz4I6z7JtNrG9J4nV6eKKaPKNefx9Ub7drZ9MF/v0UhYxYOKY/toI54e+xpaK5q4wqalzW2F01W3VkyUeg8xe93D/ibZAkDSFPZdDwAYKNR1eWwppnWqCn1f8FsJNLp/qX74iZNC/T7D3L9yrXXafm1Gv48Pg+2WVbYTkJ849Xj9ZdWbgVpwpqqanliz3ej5XgLSficZeGWrdV5UwXiv7byjqLBKWjVbHC152wfibpl2ih6u3uZ7LeaUVVv3adXWtpWF7du051MbPFuByrAC4dPHD5akwEnzMO6T/Fy/grT597Kmoen1JtM+mO/LXpCwigctcu0r9lb2mbCvIZ2CjitU/0ZaNCvzeq51VVLd381eu0uBbjMARYM7eACAZ8zkLSx+K4GO7XZUqAGTKPYvv7OPkzBb2VYSOIwE5MjyUn1m5uCMY8uldVVTWAFpv5MM/LBViRRFMN5LkiS1dmjYFVZJrKiMcq3MXMmjOZeeoiXrd+jxtfW+X7u9dGvCJmFiSTa2k2thBsJzbUsv/F7HbF6//JwX2rdFDPPcesqAYzX7otEdvucg400KElb2+Lk3o+LYviSvMR0n9jUUlU1LsydeWxjeLAyrNHse8k7BV4ejaJF8BQB4xkzewpHERHqU+5ff2cdxz1YOmgQOK0ieGkP7sW3Y0ahFL7+Z9c/s9lVNYQWk/U4y8MNWJVLYwXi/7bw/O2Gw0Xi8SEJFZSZRtOT1kjz6x+YGfWx0P9+vnUmTK123cJVcuS3VmkmYWJJOGMm1sAPhrbflHU++apQ093IdM01KZ7p+BW3zH+a5ddrYgWkrXgth2Yu4ElZJO9aDMDkWqDgOR7G2ss+GfQ0Fbce65nVYDzc2V6Wu+h8PiVdDFZOk8tHhvDYSI987mgC5cHUHAHjGTN7CEVWiM6lVCfkahEwF0VPjf2LNdlVt3JVz/GEFyVv/IdQ+wD99/C5flXVhBKTDXrvU1h+CYQfj/bbzfrld61pb4q6ozCXslrx+kkdPvhK86rU1V9J1C1drYfW2NkGMuCeWtBZWci2MQHima8iYE3oaJV9zXcfCSEoHafMf9rn16ff3/9bXtkJa9iLKhFWhBTVNj4V8qDjOx3vTfGplH5V82NeATDKehzYtlZ6+UXrjpWgG4pRIlddE816ITSF0NAFyIfkKAPCMmbyFI+xEZ1KrEvI9CGky/rCC5LmSfyaVdbYD0mGuXXrqwJ5Wg6JhBeNNvv9VW/fp5AHHas22/b6e15rjSJdPGKzy0q55E0SWwm3J6yd5FFJBYaKDGGEl12wGwnOdgz/Ur4fRe2W7joWRlA7a/SLsdaGrXtutqtd2S2rerpeOHZC4bh1BRJWwKrSgZpBjIcktcvP93jTpreyjluR9Dcgk23noruMe0tQDD8ux9m6Ost7pOiXS1LukYZOtvSOSp1A6mgC5EA0HAHjGTN7CEWaiM6lVCfkehDQdfxhBcj/JPz+VdbYD0mFW3V934SirrxdWMN70+z9jUG+98sZ+44pp15Vq6t/SzZ86xewFYhRGS96wKwX9CDuIYbLdwmyFbysQ7uUcvGKz2Xec7XsIIykdtPtFlB1NXqxt0ArDYyfJy16EnbAqxKBm0GMhiS1y8/3eNCWprezjYnNfY5sibNnOQz896if6+FsvyrGXeZVO+3dpb51U91zH31VMaq54JfFa8AqpowmQDclXAIBnzOQtHGElOuOqSsgVmMj3IGSQ8dsOkofdOs5mQDqsqvsvnjs0lM8fRjDe9PsvL+3iKRmcTZIrz7zwM3HAyznIRI7aAGNhBDGCVG+F3Qo/aCDc6znYNfiyst0nhZWUDtr9IuqOJqbHQNKXvQgzYVVoQU0bx0LSWuTm+71pOklqZR8nG/tavldEIz9kOw99v/Nv9fESy4lXSTr+VOmSn3VcP3ZYJWu8FokwJ10CSUPyFQDgSxJnjUeh0GYdh5VIj7oq4SOjynXZL57PGZjwO65bH1unRd8419sTPAi6/wTZrjaD5FG1jrMVkA5jnF88d6h+8PGTrL9uiu1gfJAq91zJYC+SXHlmg9fgqGkS6GMn9dNT6+qNknq52AxiBK3eCrsVftBAeFjrZue6TworKR20+0W+BPzzZdkL2wmrQgxq2joWktQi12aCvND+TikEQfa1QqmIRvK1Pw+NcLZqYskaneRs1qc7PWM/8So1J1ml5kQrydaiFPakSyBJ8uOvEQBAYiRt1njYCnnWse1EetRVCdPGDtCcx9fnDEx866MjfY9r9bb9mnr3c5p90ahA36+N/SfodjUd/y9njNO2vW/HFsizEUg0mWSQycBe3TTnX6M7p9kKxgetck8lg2/40xr9+u91vl8n6ZVnQfgJjpomgd7Y+3YoideUVBAjyPFmo3orijW/TQPhpi2jc1Utt79PSvcdhJWUDnpesHluDVO+3p8FVYhBTZvHQhJa5NpKkBfy3ymFwGRfK8SKaCRT6/PQOSVrNKvzw5pQsj7cN62YRMIVoU+6BJKE5CsAwLckzRoPU6HPOradSI+yKuEjo8qzJl5TmlzpzqdqjMa1etu+QN/vghVbNHvh6ozBd6/7T9DtalrlfP6Y/kbvG1S2QOLgsu762nknavr4wZ5fz88kg3Qqyrrrqz7fM0lsVbk3NZm9f75UnvnlNzh6yzSztW/XvLHf6HlerX1jv6fuAdnYqN6Kak1500C4icvPqlBNfWPG69i0sQN04J0j+u6DL+vF2gZtaTjY4XGDy7obvXeu487GeSHouTVsxbzsRSEGNcOYoBFni1wb98xx/p1Cpa0/fva1QmsZjuRKnYcu67REt3S+V50cV66rcKpdJckpaV7PFUUvikmXQFKw1wIAjCRh1niYimXWsc1EepRVCZf94nnPgYkgcWHT7/cXy17TLY/lnjns5fVtbNd8aReeK5C4peGgrlu4Wvcs26QffepkT99JapLB7IdXe64gdCRdftZgzTh7SEGcz2x8/y+9vsfovfPxvOiF3+Dow9XbElkpuHDl1kATRGxVb0W9pryfQHiQdZNv/tTJHa5j3Y/upIXV23TdwtU5XyNdQtYLL8dd0POC1wlccSikZS9MmAYnN+xotDwSe6KaoBGVoPd2cf2dQqVtuAqxZTiS661DR3ROyZqWxKsUcuJ16l3SsMkhvQHySaFd04FsSL4CAAKJc9Z4mIpp1rGtRHpUVQmmLSBN+f1+F6zY4inx6vX1bWzXfGgX7jWQKEm1uw7o8vuWa47Hao7UJIPrH12jTbsOZH1sahvkYzV7JkG//5r6Rq3Z5r8C89SBPQvy+mAaHJ1z6Sn6x+aGRCWqcg0lV+DeZnvTpE4SCXoObn0dyzXBxAavSWkb1wUb60L75bedczEy/eyLVr2p6eN3JXLbRT1BI2xBzytx/J1S6B2BkqAQW4YjuXp07axZnR9uSbyGpmJSc8UriVe8r9Cu6UA2JXEPAACApAky6zifjexXqisnDtU3pozQlROH+r65jWoGo2lgIgiv32/Vxl2a/XDuaiY/r29ru04fP1jzZ07QhKFlaR8/YWiZ5s+cEFvAzE8gUZLc95NCXveHicP76G/fnaw5l56iigytPOPeBmEK8v2bHnOnDexp9LykM90eB995z7j9cJxSgft0bHc8uGXaKSrJUXURdXLN1jnYzwQTU36T0jauCxOH99GCq87W586u8D1eE7deekpir2NJkQpq+uVmOdaTYNaUETnPDylJr34Ocl6J4+8Uv5W2cdyrF4JCbBmO5DqvbLcmlKz33BnIk76jpAvnSOf9sPm/X31BuvIvJF7RQSFd04FsqHwFAKAdZh2biWoGY1wBBi/f79zFG4z/gM30+ja3q0mVcxStxU2rmU2qOaaPH6zp4wcXbMv0bEyr3E2Pua173i7IVnhBgqNJrCjzIlNbQ9sdD5K4prytc7DfCSZ+mSalbXW/GNLnGL9D9m3C0LKiPof7MWvKCF1+73LfSy8kuYVpPnTx8CrIeWVeVa3Rewb5O6WYOgLFiXUQEaUh+/4hyXKr4YvmkGiFJ4V0TQey4QoNAEA7zDo2F1bbyNZB1rVv7As4SjO5vt+g7ZA3Z2mHa3u7emkXHta6XukC5kEqJEwDxYXaMt0Lv5/dNKi35NWdWvLqzoJbBy5IcDTsaqD/c9rx+vOqN0NJ8qUL3IfR8SCJa8oHPQfbapdfcVx31e3uuA6sjaR00HNi2Md3++1azOdwLyYO76Oppx2vP738pu/nJnkyYRInaJgyPa9E/XcK65BGh3UQEanDlrt2nf0NEq/wpZCu6UAmJF8B4H1JCvIhXsw6Nmd7BmO2BGDUcn2/QZMqv36+Tuu2N6ZNUkU9MzSMdb2yfZcDe3ULMtxEB4q9SPr1J+j+VGjrwAUJjj6xZrvl0bT19Y+M0JgTeuqnSzaq0fKEoHSB+zA7HiQpuRb0HGwr6f6vYwfqgpP7J/J8YbIveEXVg5kR5aWS/Cdfkz6ZMIkTNEyYnlei/juFjkDRYR1ERKqLxf3mpE9JF/zI3uuhaBTKNR3IhCgxgKIXVnUX8hezjoOxNYMxVwIwarnGayNYmS1JFdXMUL/reg3o3S3wd7l179uGo22W9EBxJvly/bGRVPGzvyRdkOBomJWvE4aW6aUtezTn8fWhnDczBe7D6niQNEHOwbbOUT26dk5UUro9P/tCe6VdO6edMEDVg7lCn0yY5GPBK5PzStR/p9ARKFrFck1FAgyrtPM6Z3+DxCsCK4RrOpBOftxVA0BIwqjuKkaFNkuNWcfBBZ3B6DUBGBUv36+tYGW2JFUUM0Ntr+sVxXeZL4Hi1vLt+mO6fmBrhbQOnGlwNKzPXuJIHxlVHuqxlmnsxbRmk+k52NY5Kunbzuu+4EiaetoJGlHeo832K7T7ybgxmTA/+D2vRP13SqEn8ZOmmK6piFn5aKliolRXZfb8E86QPnojrYYBIAvuhgAUrTCqu4pNvlRtmWDWsR2mMxj9JADD5vX7tbmv50pShTUzNIx1vaL4LvPtPFPM159CWQfONDgaRmvW1Hs8VL01tGMtV+C+2NZs8nsOtvG582WSV5B9gaoHu5hMmF/87P9R/p1CEj96xXZNRYwqr5XmXyK5TR4e7Ein/7vU/9Tmqtny0aEPDwDyHclXAEXLdnVXscm3qi2/mHXsn62KFZMEYFj8fL+2kypxJKlsr+sVxXeZj4HifLz+zF28IVDVa2uFsg6caXA0SGvWTO/Rt7SLrl24KvgLpuFnAgprNqUX9PqQb5O88mVfSPr4bGAyYWGK8u8UkvjxyJfzKPLcsMnS1LnSolnZE7BOiTT1LmnsjMiGBgCFgOQrgKIURnVXMfFbteXK1cF33su7PxqZdeyN7QroMNdE9MPk+7WZVJGiT1LZXtcr7O8yHwPF+Xj9sZ1Ez/d14NoHQm/+1MmS1OZnA3p107a9b6u6bo9q6hvbXPe8Bs0zaW7Zery+/pERLa85r6rW1sdrwyRwT/VieqbXh3ye5JXUfaGQO7e0x2TCwhXl3ykk8eOT1PMoCsjYK6Reg6Vlt0t1z3X8fcUkqfIa2gsDgAGSrwCKku3qrmLjt2rruoWrO/w8X4JbzDrOLowK6KgTM44j3TrtFCsTBIImVdqLelvYXtcrzPEnOVCc7XyRj9cf20n0fF0HzkvCRpKnpE6uoHlp185qzHD8uJK27z+snY2HW/aJMI61ocd1148uOSWRx1g+Mrk+MMnLvkLv3JIOkwkLV1R/p5DEBwrcsMnN/9uxTtq0TDrcKHUppb0wAASUn5EPAAjIdnVXMbFVAZVvwS1mHXcU1rqVUSZmUkEim/tgriCnH1EnqWyv6xXW+JMaKPaSnMvH64/t907a9+bFghVbNHvh6oytl1+sbdBn710uR8r6mNbXvb6lXXThyf016vhS1e87pH7HdtWQPsdo4vA+emnLnpzv1/q1wjjWyo/tmpffVZLluj5UHNdd4yvKNGbAsUzyCkHc623HOZGPyYSFLYq/U0jiA0WgfDTJVgCwiOQrgKJku7qrmNisgAoruIVohLVupem+kC3pkU6YQaL2Qc7Nuw7o18/XGb1OlGyv6xXG+H85Y5zOH9Pf+usG5bWa6hOnHm/0+nFef2y+dz6uA1e1cVfWRGhruR7T3A1ile6vqtX67Y0dfn/m0DK9c6RJcx5f7+m1wryGxt3uulCRBItPXOttJ6nNMZMJEQTnLwAAAO/IIgAoSraru4qJ7QqoJle69bF1WvSNc62+LrwxDZ6EtW5lajwDe3XT1r1ve37tVCI102z8UwYcq9MH9VZ5aZdQgkSZtmPrIOe67f62WVxJKj/rejmSpo0dkPH3JsncXLb52C+i4qeaatGqN43eI87rj633ztd14P7zz6/4mtiRiyulTbxKzUl6P8dLKkG04KqzrR9rEssthIkkWLTiWm+7GNsco/Bx/gIAAMiN5CuAomS7uquYhFF9tXrbfk29+znNvmgUCe6IBK3CsL1uZbbx5JJK6MQxG9/PdvST1IwzSeVnXUJXzWs6L6zelnGf8fO5vUhi+3c/1VSum309z3Tivv7YSKLn6zpwNfWNejVDojQpUgki28ealMzjDTARx3rbcbc5BgAAABCfkrgHAABxmTVlhEocb4/N12qdMIQVFFq9bZ9m3Ldcf1zxeiivjw8sWLFFM+5bnjGRkqrCyPZd2Fy3Mtd4skmX0BnZr1RXThyqb0wZoSsnDg0taeV3O6aSmrnOO0lIUk0fP1jzZ07QhKFlnh6fbZ9JfW6Pp9ucktb+3aSaqvHQEc/bIynXHz/XzPYmDC3T/JkT8rKq65GXtsY9BE+qNu7yfI7xI2nHG2AqjvW2TdocAwAAACgM/DUNoGh5re5KQiIkScJoI5rCzP/w2arCsLVustfxpBPmmq25mG7H6eMHa2Dv7hnbI8f5mdqbOLyP+pZ20d1/26BFL+dulZttn5k+frDWbNuv+S/4X/c23bhssFUhbVpNNfW04/XnVW/mzfXH6zXTkTT1tBM0orxHXq0Dl2l/WP9msqteU1IJolznGL+SsO8BNti6b/EqrjbHAAAAgGQv5gFzJF8BFLV8SoS0FvcFNIzWhimpmf9J2+aFwqQKI913YWvdZD/jkaSBvbpp5rlDY79pDLId42iP3Pq9Dr57RJKj7kd1yvi+pm2gs+0z5aVdgnwESc3teoNuo6Att9szrYoaUV6q+TMHW7/+hLlf5es1M5tc+8Phd9+LYVT+tU4QtT/HPP1Kvape2+37NQf27sYf5ygYtu5bvIqjzTEAAABgO+YBcyRfARS9OBIhppJyAfWzJqQJZv6Hw2YVho11k03Gs3Xv27Efm7a248h+paF/Dq9J1NbnjwUrtgQ6tjPtMzbalzYeOhLo3JDrs6XaJ9867VTPLXKDVFPZvP5EdX3Ip2tmLl72h3yR7rttfY4xSb6e96G+gccFJIWN+xY/4mhzDABIqB3rpE3LpMONUpdSaVilVD467lEBKEBhxDxgjuQrALwvikRIEEm7gNpubdgeM//ts12F4acCOt26lflaFRJk3Kn/RpGw8pNETZ0/Zk4aqvueqw08qSLdd2RrYojp92+r5XZ7Nqqpgl5/4rg+JP2amUuQludJkytBZLqPzjh7iOGIkqsQJg3AXND7Fj+ibnMMAEigTUulZbdJdVUdf1cxUaq8Vho2OepRAShQYcU8YI47ewDIA0m9gGaqgOp+dKfAQW1m/ttnuwoj6LrJ+VoVYvr+9z1bq5sWvdLh52FUrZsklppc6VfP1lp5/3TbyNZ60Zm2f66kiq2W2+1FXU3VXlKvD0nnt+V5UnlNEA0u664tDQc9v67NfTQJktI5BPEKet/i972ifB4AIGGqfyMtmiW5Tel/X1clzb9EmnqXNHZGtGMDUJDCinnAHMlXAMgDSb+AZqqACpKAZea/fWFUYQRZAzJfq0JM33/r3rfT/jyMqsS4E0uZttGsKSN0+b3LFWRo7V/bS1Klb2kXay2304mymqq9pF8fksikdXgS5UoQma7dbHsfTSfKCtSkdQ5BvKJau9rWxByqtQEgD21amj3xmuI2SYu+KfUaRAUsgEBsLjMGe4hsA0DC5esFNBXcuvWxdVq9bb/v5xd7ciAMYVVhmK4Bma9VIWG8v82qxCQkljJ9honD++hbHxuhO5/aYOW1vSZV/mWk2dqVXlscm1RT2Qio5+v1IW6mrcOjVOJI1104Sn9bv8MoQWS6drONir9soq5ApTIc6US1dnWQiTlUaycXCXEAOS27LXfiNcVtkpbdTvIVQCD5uqxXoSP5CgAJl88X0InD+2jRN87V1Luf0+pt+zw/r9DaHSZF2O1R/a4BaTKeiuO664k121W1cVdswS5b7XPbs1WVGHdiKdc+M2vKSD22ervWb28M9Np+kipLX93p+70kfy2mvVZTSdJlv3jeSkA9n68PcTJtHf6vYwfqiVe2qzHk1uepBOhl4wfpqsoTfQf6TdezHdirm84bVa7TB/cK9gEysFmB6nWbJK0ynKRNsoS9drVpm2OqtZOJhDgAT3asS7/GazZ1zzU/r3x0OGMCUPDydVmvQkfyFQASrhAuoLMvGhVbS060FWd71KDjkaS63Qf1X0/VtPw7rmCX33F7ZaMqMc5j3+s+c/0nTvK9/dq/dhStlf22mM5VTWU7oF4I14c4mLYOHzPgWF0ydkDg1tnZpKto9ZsgMj02tu59W/NfqNP8F+qsn1ttVaD6SX4kqTKcpE3x8tvmmGrtZCIhDsCzTcvMn0fyFYChfF3Wq9CVxD0AAEB2hXABTc38L3GyPy7sdodI3nfhdTyZpIJdf1zxut2B5RB03NkErVyN69j3s8/43X7tXzuq1sqm+//IfqW6cuJQXXByf0nSE2u26/pH12j2Qu8BdS/7QSFcH+IQpOX5xOF9dOulufddx5GG9jkm7e8mDC3T7/5jgp789r/ohqkn6eqPjdQNU0/Sk9/+Fy246uxA511bx4btc6tJBWp7C1Zs0Yz7lmf8fO3HHKQy3Ca/40bhmTi8jxZcdbanY97GsQK7/CbE4+5AAiBmh/139wn0PABQ/i7rVeiKO/ICAHmgUC6gfmf+IzxJ+y5yjSeXuKo/co17YK9u2rr3bd+vG7QqMY5jKLXP9C3tonlVtZ5aanr93tPtj1EENoO0P89W5eaF1/anhXJ9iFrQFux+zqG52syGUWFpi61zq40KVJNqwCRUhlPFGEyhtWnOVcWepGptfCBp7csBJFwXw/Ox6fMAQOEvMwYzJF8BIOGScgG1EQDL1ZIT0Unad5FuPA+u3KotDQc9Pb/Jlb7/yGr969iBkX6ObNuxauMu3bToFd+v2aNr50DfS1hr0rbnSLr8rArNOLtCOxsPG7XUbL/9Nu8+oPp9h9Tv2K4a0ueYjJ877La5QVpu52pN6JWXgHpSrg/5KGgLdq/n0LDXlGxv8+4DVl/PRiLBxtrEJsmPC9+vPPfLZmU4SRszcbVpjvueiHW8k4eEOADfhlVG+zwAeF/SlhkDyVcAyAtxXkDDCIBFHYxGenEHGdNJ7Rs19Y1t1nb1Is71YG3u0w9Xb0ubtE33eTJ9h6Zr0pY40sxJQ3Xfc7VZn5tqA3zZ+EFW1kHzu/3CbJsbpOW21yo3P6+Xa7vwB5aZVOvrXN9Xrv0hKdezoNXW2QRNJAStQDVNfsycNNTofW1dM0jamIljbc2krMmbhGpttEVCHIBv5aOliolSXZX351RMYr1XAIHZ+hsX9pB8BYA8ENcFNI4AGMKXlCBjNjZaZ8a9f760ZY/R81Zv25f2560/z4De3XJ+h17OGa21TqhO/lC5p5aqcbXUDGv/DNpy20+VmxdeAurF9geWzUkjqfbBcx5fr1VbOx53+dIO31a1dTZBEglB1yY2vR5s2/u2cWW4jf2MpI1/cVxTknSvyzreyUNCHChSO9ZJm5ZJDa9J+7dLxx4vlQ1rrk71kiStvFaaf4nkNuV+rFMiVV4TfMwAoOQtM1bsuEsHgDwR9QWUdcqSx0YwOEiQMcpKWVtBq7j2z9TxY1uTK123cJXkSK6H73D+zAme1tJtf/7w2lI1rpaaYbRWdiRdOnag8fhMqtxy8RpQL4Y/sMKYNJJ6zXSJ15MHHJsX28x2tXUmQc7JQdcmNn3vtdv267SBPT0flyWO9JFR5brsF89b2c9sJG2S2KEiTFFfU5J2r8s63slDQhwoMpuWSstuy161WjGxObk6bHLmxwybLE2dKy2alT0B65RIU+/K/loA4FPSlhkrZtwRAkAeifICyjplyWEr6WAaZDR9/yD7qc2gVRz7p+0KyNbclv+TWeo7nD9zghZcdXab7+Lgu0ckOep+VKec30u2lqpRttRMty/NmjJCl9+7PNem8MxVsOC6jWrt9vyMo5D/wAqjMi3Xa67Ztj8vOjuEea5pLcg5OejaxKbv/VD1Vjk+Hj/xxD6a8/h6a/tZkKRNPnSosC2ONs1Ju9dlHe/kISEOFJHq3+ROlkrNidn5lzQnTcfOyPy4sVdIvQZLy26X6p7r+PuKSc0VryReAYQkKUvkFDOSrwCQh8K+gLJOWXLYTDqYBBm37jno+/1tBI1tB62i3D/DqIA00TpQHMY5I4qWmrn2pbOGlen5Tfa2dZDguu0WgxOGlkmS5lXV+kqkFtofWGFUpiWt2s1UlOeaoJ8/yNrEQd7bT176WQ/nND/7hOm49x96N+u2CrsNblwTOKJu05zUe13W8U4WEuJAkdi01FviNcVtkhZ9U+o1KHcF7LDJH7QxPtwodSn13r4YAJDXSL4CADpgnbJksJkgMA0yvljbkDN43fr9TZK16YTRVjaq/TOMCkhTYQaKw14HzcvEA8dPWZtHptvMZrW240h7335X59/5TIffFWrVWyZhVKYlrdrNVFTnGhuJhCBrE4dxPQjC6z5hMu5R/Us19+nc+2cYEwPirraNem3NpN7rFts63vmAhDhQBJbd5j3xmuI2NVe1eqlcLR9NshUAilBJ3AMAACRP1AEwpGeSIMjENMjotWqoyZVu/vMrvpLFucY0a8oIlVhMrkW1fybtOAgrQRPmOmheJx5kWvc2KJNtNqBXNyvv7UiSK726vTHt71MTGP644nUr7yc1T86YV1Wruxdv0LyqWtXUp3/vqAWpTIvyNeMSxbnGZiJh+vjBmj9zQktVd3sThpZp/swJaSfm2L4eBOV1n/Az7tTjbF33/ViwYotm3Lc847ERxnmnvajX1kzyvW6QYwX2pRLiuY5lEuJAntqxLvsar9nUPdf8fAAA0qDyFQDQQdQBMHRkux1eFMHC9RmSRek0udL3H16tX33uwxkrSLxWf3gVdP/02ooxacdBWN99mOugRbWOZSZr39jv+bHZqsX8Gt2/VOu3N/qqNg8S5I270i2XMCrTklrtZiLsc00YiQTTtYlT14PZC1dbW+c5KC/7hJ8qxm99dKTueKrG1xhsdDdIShvuqNfWTPq9biGv452Ppo8frIG9u+uuxRu0PM01c8LQMn2ziLpSAAVl07Lgz6eqFQWG+w/AjmRFBwEAiRB1AAwd2U4QJC0hKEl1DQd1/p3PZE3y5Ap2+WG6f/pNUCXtOAjruw9rHTTTdSwd+VvfMZuFK7fqzCFlOauKcrVGzsZxpMsnDFZ5adeWP2h/+OgaX9XmQdrh2lxPOixhVKYludrNrzDPNTYTCZmCN34DONPHD9aKzXv00Mqtgcdkg9d9wmvSxrS6OujEgKS04Y56bc18udcttHW88xkJcaBAHQ7Y3STo84EESfrkXCDfJC8SCwCIXdQBMHRkO0GQ5BvkXEmedMGuB1du1ZaGg57fw3T/NElQmRw/pV07qzFhFapehLEOmunEg8vPqlBNfWPgJL3UnMS9buEquXI1ffzgtI/xWi2WTqqisPX+brvaPZukVLrlEkZlWtKr3fwIYy3U8z7UV9+7eLSV63kYwZsxJxyrh1YGHpoVfvYJL0mb6ro9RuMIMjEgyvOOF7OmjNDl9y73NAnFUbCW2NzrIsVvMpWEOFBgugQ8noM+H0iIfJicC+Sb5EURAACRSxd0CCOpEqV8n5VuO0EQRpDeJi9JntbBrrEVvUPfP4MkqPweP18/b7jmPL7eeqvdsAPFXltqOmpuqdn+u013nJomEspLu+jmT53c4TUfrt6m1dv2+X695gTsai2s3pY2SWTaGjlTRWGU7XCTUumWi+l77mg8rLsXb0h77rdd7WbjWhPkNfyca7z4l5F9rZwzwgreJGkikclYsiVt4pgYUEhtuE3k+70ugqG6B4AkaVhlvM8HEiBfJucC+YbkKwAUsVxBh5mThuq+52pzrlNme024IAolkBJGOzw/QUbHkdyIF9bzk+Txs46e6f4ZJEHld3yXjR+kXt2Psra+beq1owgUe2kN7Uq646kaPbdxl2a9P6ZMx+ngsu5G40glINonN0b2Kw2UnEqXJDJtjfzLGeN0/pj+aX8XVTtcm5VuYU9yMZ008vOlr7X5d+tzv61qNxvXGhuvYXttbBvXxzCDN0mZSBTGxJY42uCanneeqdkZynE/d/EGz63XXQWfGBLFvQSSKYnVPfk+cRTIW+WjpYqJUl2V/+dWTGK9VxSEfJmcC+Qbkq8AUKS8BB3+sblBMycN1aqt+7KuU5aUm64kBlJMmSYINu86oDuefFWNh46otGtnTRs7sCXZ4zfIOHfxBm3b+3bQj+KLn3aGXtfRM9k/bSSo/I7P5vq2UQeKUy015y6u0U+eyhw8f7G2QZffu1xS5vVZ/bSTbj+GTD+/5IwBWli9zeh1pY5JItNqsWzHU1RVb0Er3WrqGzX/+Tr9bf2OtJ/H9iQXG5Wd7c/9QavdbFxrbF6vbJ07bCUUww7e+JpI9P7/yTaZyHGks4aW6YVNDZ4Sf2FNbImjDa7peWfJqzu15NWdLf+2cdzH1QI5zHsJ2GcjQZm06p5CmTgK5LXKa6X5l0huk/fnOCVS5TXhjQmISNKWoQAKCclXAChCfoIO9z1Xq/kzJ6hvaZdEz8ZOWiDFBr8B5lVb9+pL89suhvf42nqVdu2sr583XFdVnugryLh62z7Nf6HO0qfxzk87Qy/r6JmOwfR57Vuc+hnfxOF91Le0ix55aaseX7Ndtbv8JyIH9u6mOZdGX6FTtXGX5j6du2opjILqbAmIqo279MhL5onXlNZJojCqVKOqejMd+9pt+3XZL57P+Ye57Ukutio725/7Tavd/Fxrrnt4lVZv26vy0q5tjvswrlfpzjU7Gg/pt8u3eOpiYCuhGEXwxu/3N6B3t4zXvNR6289v8jbmsCe2RN0G19bnsHHch90COdt1OKx7CdhjM0GZpOqeQpo4CuS1YZOlqXOlRbO8JWCdEmnqXc3PA/JcsS9DAYSJ5CsAFCGToMOCq85O9I1VkgIptvhJOriS3n43/R+KjYeO6JbH1mvjjrd0+6dP8xxknHF2RSzJV5PkULZ19KIaQ7bneRlftsCiHzMnDY1l3zZdAzWoXAkIm+NKJYnCqFKNqurNdOwLq7d6TpzbnuRiq7Kz9bnftNrt1se8r83sutL8F7a0+dmZQ8u0/+13I7leDevbQ9/66AjNfTr7MWAzoTj/+c1Gz/MbvPH7/bW/5m3Y0ahFL7+pRh/n+igqIKNug2uzjXPQ4z6s1ut+knZe7yVI0mYWxraxmaBMUnVPIU4cBfLa2CukXoOlZbdLdc9lflzFpOaKVxKvKBBRLX8DFCOSrwBQZJIUdLClED9TSq4A8+CybtrS4K018IMrt2p4eQ9dVXmipNxBRpOg7NA+x6hu94FAyS7T5JBNpmNY+8Y+3b14g++AY67Aoh9xBOZM10ANKlcCIoxxVW3cFVqVqmnVm59gt+nY/e6afpOGuT5DukkjB989op8v3eRrXK3P/X6q3ao27tItj63Tmm37fb1feyb7o5frVbYE04f6l6pE0rrtjR1+ZzuhuGT9ztwPSsMkeOO3WjF1zavauEt3Pl3jaZ92JF1+VoVmnF0R2f1C1G1wbbT2TgkyWSCMSS22qwppD5tZWNvGdoIySdU9tieOMikAsGDY5Ob/7VgnbVomNbwm7d8uHXu8VDZMGlbJGq8oOFEtfwMUI44SACgySQo62FKIn6m1bAHmS3/+d1+v9dMlG1uSr174TQb96FMnS1KgCrWggUsbwSfTMTy+tl6Pr61v+beXgKPXwKIXttZs9Mv0GPSiV7ejtPftdzv83EsCIoxxvXXoSGhVqn6r3iRlbAWcad+zWemWS9CkYbrP0HrSyLyqWqNxtT/355qIYnNyhKls16tc43t1e6NKHOk7Hxup0q6dQwvM19Q3aqvhOuFBgjd+Ox/4SXi4UiwTtaJsg2urtXeK6eQ225NabCftaA+bWZjbxnaCMinVPTYnjjIpAAhB+WiSrCgaUS1/AxQjkq8AUGSSEnSwqRA/UzrtA8wPVNX6apkoNbcgfnLtdp0/pr+nx3sNyjqSvvXRkR1aPH7x1/9QXYP3dUuDJA9tBp9sJai8BBxttcW1tWajiTCPpdaJ1wG9uukjo8o9V6GFMa5UkiistRm9Vr1t3XMw6/tn2/dsVrrlEiRpmOv4ieLcb3NyRBCZxuwnwfSTp2s0f+aE0AIlQSY7RBW8eXLt9rzqlGG7pX4mtlp7p5hMbrM9qcVm0o72sJmFuW3C6GyTlOoeWxNHmRSAopKqRj3cKHUppQoVsCSq5W+AYkTyFQCKTFKCDjbl02eyUcUSdG3Qh6u3ek6+St6Csq6kO56q0XMbd7UkOUf2K9WPp50SSoKqvTCCT7YSVNkCjrba4tpcs9FEVMfStr1v63fL63TKgJ6ejpswxtV6gkFYazPmqnoLGuz2M6kiaM7RRtIw0/ETxbk/rrWM28s05iStd26aDB/Yq1vowZug18186ZQRRLrzTvWWPVryqv9W0qb7gq1JLbaTdkk6zqLk5Z41zG0TRmebpFT32Jg8xKQAFI1NS6Vlt0l1VR1/VzFRqryW9VeBgMKaWAwUu+RG0gEAoUhK0MGmfPhMtqoybbS/9FstK32wra556GW9sfdQxse1T3LaTlClCwTubDwcSvDJZivGTAFHG21xba3/F2RiQJTHkp/v0fa42s/wzTUxYWCvbjpvVLlOH9zL6P0yVb3ZCHZ7qbAdVNZdD63cajT2lAdXbtXYit4d3t/GZwj73B/XWsbppBtz0tY7N02Gnzeq3PJI2rJx3cy3ThlBtG/tbZJ8Nd0XbN0z2EzaJe04i4LXe9awt00Y3Q2SUt1jY/JQsU4KQJGp/o20aJbkNqX/fV2VNP8Saepd0tgZ0Y4NKCBhTiwGihnJVwCISBTrdnmRlKCDTUn/TLaqMm21vyw1CPgsWLFFsxeu9lQF1z455rWNaq61OzMFAku7dg4t+GSzFWO6gKNpYHHiicfpoyf1s3IesTExYGS/Ug3o1U3bDNd79Mvr92hzfdNMM3xbV4vNf36zlqzf2bLu5da9b2v+C3Wa/0KdtXXXbAW7a+obVVPfqEnD+2hsRW9Jrrof1bnN9WleVa0eWhlouNrScLDD+c30M9zwpzUactwxLeML49zf+lq99o19vsYYlkxjjmu980z3M6b79oyzK4zHkout62ZSun9EfS8Zx+Q2G/cMNpN2cR1ncfFzz3rgHbPt7HXbhNXdIAnVPUGPrWKcFIAitGlp9sRritskLfqm1GsQFbBAADbuwQC0lYy/IgGggNlch9KWJAQdbEvqZ7LZEsxW+8tpYwf6enzqM/h56/bJsVxtVLPJFQj0W8nrN/jUNrlWp7+t32GcZGwfcDQNLH70pH66cuJQo+e2ZrNd80dGlWv+C3WBx+SV1+/RRvtoLzN8X9qyR79bviX0ddeCJgK8XJNS29TWtan9+c30M/z67x/sX6mx2jr3B21LG5ZsY456vXMv+07SJkLZum6GcZ/m53oY171kXJPbgtwzSPaSdlUbd+m+Z2uNXisfq6X93rNOO8Pf/WSK120TVvI/CdU9QY+tYpsUgCK17LbcidcUt0ladjvJVyCgoPdgANoi+QoAIQpjHUobkhB0sC2pn8lWSzBb7S9Lu3b2td6rZB68Tpccy9RGNRNbVUvpXtfvHw/NybU6q60r42yZbXutsBlnV0SafJW8fY9B20d7rcyOat21IAm3OK9Jrc9vNpISrcca9Nxvoy1tGHJdr6Jc79zrvjNz0lD9Y3NDIiZC2bpu2k4Q+02kxn0vGefkNr/3DCk2rq1Bzws7Gg+bPbGVqAOffu9ZV9SZHV9ez0FhJv+TUN0T5NiKevINELkd69Kv8ZpN3XPNzysfHc6YgCJieg8GoC2SrwAQkiiD8SaSEHSwLWmfyWZLMBtrg0rS188b7uvxQYPXQWfY26paas9v8Cms1pVxtsy2vVaYzRa/Xnn9Hr2eG/qWdjEKdEe57pppwm1H4yHd+XSNr2tSTX2j0Xtlkjq/2Wrhmhrr/JkTNH/mBKNzf1gTPLwY1b9UPbsdZXy9Mt2Huh/dydfj/dzP3PdcrWZOGqr7nquNfSKUjeum7WSi30RqEu4lkzq5LRsbVYVBzwv/fH2P8XPjqHQ2ud+r233Q6L38jD3M5H/c1T1Bjq0oJ98Asdi0zPx5JF8BAAnBnRcAhCTKYLypuIMOYUjSZ7LZEszGTPVPjxuoqypP9D2WIIKM21bVUjp+g0+2W1e23j8/1K9UK2obPLV1tpUICGutMD9BUseRbp12ig6+856eqdmpJa/u9DUeyd/36PXc4Pc8EfW6a6bXiafX7fB9TZoUwjWpauMuq9e61FgXXHW20bk/rAkeuTiSrpw4RNPHDza+XplOeHi4epumjx/s+fF+72dWbd1nnAy3Keh103Yy0SSR6nfbf/+R1frXsQOt3/ckcXJbrmMmSNLOxnlh9bb9Ruf5uCqdTe/3Bpd115YG70lYvxPIokj+h1Xd42U/NT224uycAkTisOEEQNPnAQAQApKvABCCqIPxQRViS5HUZ0oFPp5Ys70l6G/rs+YKqqzdtt/oddMFjIPMVC/t2llfP2+478RrprH4EWTctqp90/ETfLLZunJn42H98NE1Rq9nMxHw079tMHperkpmPy1+XVdaWL2tpYLHJPmahCCr6X76xV//Qz+edorvz2CacHtz3yFfj19e26CxFb18PceLtw4dsV4l3fr66ef7tT3Bo8SRp6pPSXIlXbdwdcsxYLqG86VjB4R6v2F6P9O3tIsWXHV2rBOhglx/wkgm+k2kznl8vVZt3efrPep2H9R/PVXT8m+bVZJJmNzmpyLUNGln87zgt/tHnJXOpvd7Zw4t09Y9B0NtS5205H8ufiuXTY6tODunAMZ2rGuuTD3cKHUplYZVZq5S7WK4r5o+DwCAEJB8BYAQ2Kx4hJkwW7Z5ee2tew5qYfVWo9dPFzA2HeuNU0/S5w2D+pnG4keQQFhY61L5DT7ZSAI7jnTqwJ6eq3DasxlYrNq4S4teftPouV6+k1xB0tZaV/DkaxDRdD+tazhoXL3kp6IrGMf6K6bOKbY/g8n10+YEj1QC57LxgzT5Q+We9n8peBXbwXfeMxmu5+1luo1++rcNuuvfxlqd7OA36Wd6vvzljHG+10bPxSSh5zfxmk4YVZJxTdgzqQg1SdrZPC/4vT7E2TXH9H5vzAnHavyQ8NtSJyH570WQymW/x1ac6zEDvlT/Rnr2DmlPbcffVUyUKq+Vhk1u+/NhlWbvZfo8AABCQPIVAEJgGowPK9lUbMJs2ebltS+/d7kkeWojm066wJTpDPcgiddMY/Hz/kECYmGsS2USfLJxXLqu9Ktn0wQc0nAc6fIJg1Ve2tUosJgrMDl38QbjfdPrd5Laby6/d3nO90pV8Fx34Sj9Y3OD58DzqQN7entgyILsp6bVS34qjIPoflQn6+v4+q1K88rkOLV1zW2fwEklCRas2KLZC1d7PgZMqtjCvt8wff1FL7+p6ePttJg2nUxlet20nXiVwu3kkEuY68FGJUhFqN+knc178WzXh/bjGdCrW6xdc4IkRUf2K42sMjXJ3XpsVC77SS7n43rMKDKblkp/uVravTHzY+qqpPmXSFPvksbO+ODn5aObE7N1Vd7fr2IS670CABKF5CsAhMA0GB9GsqnYhNmyzWsgPUgeIVvCMo4Z7qbtQW28v+0gkWnwKerj0nWlmvq3dPOnTvH1PC8Jir6lXQIl0vxsOz9J3iZX+tv6HZo5aajnJPV9z9Vq8ofKYw8mBn1/03Ua/VQYm9rReNhqhWr785vNz2BynJoe2xeO6acxJ/TM+X0trN7m6xgwqWIL+37D9PVd2anKCzqZKimVYXFPrrNdJRk1GxWhXpN2Nq/56bZ3tmu1CVtdc4K2sU2X5D747nuSXHU/qrNq6hvVt7RLYhOnNgTZT00nmeRbS2YUkaVzpKU/9vZYt0la9E2p16C2FbCV1zYnZt2m3K/hlEiV1xgNNZ2kV9kDAPIDUX4ACEGQ2eMIJoyWbbYDZZnkCvzGNcPdb/LFsfT+JoHA0q6d1ZgmyB0k+BTHcem3msVrguITpx5vPCY/lcym60QeeMd7giLfkwmtma7T2DrY/chLW/XzpZusjuu3L9TplAE9rVSoZjq/tQ/Yb951QL9+vs7360d5bH/n/A/lPBaiWvs97PuNIMdX0Ko8G5Opgqz7aTPomoTJdTarJKMU1bGUYuuaku6ametabcJmYt/GZIWR/Uq1s/FwaEt/JFWQ/fSlLXsCTTLJl5bMKBKblkpPfF+qX+vveW6TtOz2tsnXYZOlqXOlRbOyJ2CdkubK2fatiw2EuXQRAKD4xP9XIAAUoKCzx2EmjABdGIGydBx5S1jGMcPdT3vQYX2O0c2fOjmWxG+JI91z+Tj1Le2SMfhkEpgyrf4Nys+ajF4TFKZrvTryVxFm2mJzzbb9vh6fhGRCGO1E/bZGH9mvVOWlXa2Pw1VzUmv+zAmaP3OCcYWqlwkhravS1m33dy43vX6Gea2Oau33sO83gp7/glTlmUymSnf+93PdDCvompRAra0qyShFdSyl2Ljmp0tMer1W+2W7UjfoJL8wl/5IMtP9dP7zm/W75VusdOxJcktmFInq3+ROlGZT95y0Y13b1sFjr5B6DW5OzNY91/E5FZOaK14tJF6L9fwFAAgPyVcACElSWt0lVRizs20H6MIKlKVz6biBnv+Ii2OGe67gdUVZd331vBM1ffxgq+/rNxCYKfEaNKBus/WqV16rWfwkKEyHP/W0430lD6JssRl3MiGsz+q3NXqY47hr8QYtuOrsNuedtW/s14u1DdrScDDr800mhER5/QzrvaJc+z3s7TVryghP6zenY7odTCdTnX/nMx1+njrHL7jq7KzXzTCDrnFN4mkvjvbHQe9VojyWUoJc8zMlJv1cq/1I1+0jiCCT/MJc+iPpTPe3Jet3Wu/YA8Ri09JgideW11nWcd3WYZOb/7djXfPvDzdKXUqlYZXW1ngt5vMXACA8JF8BICRxtYhNOptVJe0Dept3HzAaU6aASViBsnTGnHCs7+dEPcM9rrZmXgKBHxlVroeqt+rahas6/H5on2O0edeBjIkDLwF1P9W/tnipZjFJUJj4+kf8JWuibLEZ91qKYX5WP4HWMMfRusJ4ZL9SvbRljx6u3pr1WHAkfftjI40So1FeP8N6ryjXfg97e00c3kdTTztefzKonPd6Hmt/TbFZUd7+HB9kslWQoGuQJLajYOvJp0R5brZ1vxflsZRies3PlJgM81r9k6drNK6it/WOJyb3e2Es/ZEvTPe3rXvf9vX4JHT8ANJadlvwxKvUnFjNpHy0tWRre8V8/vKK1uYA4B/JVwAIURwtYpPMVlWJ7TVY0wVMokpqpeTTPhB20jfdH3bZAoG51sqq3ZU7Ke8loJ7reLbNyz4RRsvb9kxauka5P8e9lmLYn9VroDXscaQqjL0mqVwFSwpEef0M4732H3rXaCzdj+5k9Lywt9fXPzLCKPma7f2yXcsH9Orm+72yyXWOjyLoOnF4H33itOON2r9fftZg1dS/FfjaE9W52WYVcdjrGmeS65g6dWBPnTawp8pLu+YMQod5rQ4zCeDnfi/qtXmTJsr7nrg7fgAd7Fgn1VXZea0u0e/bxX7+yoV1cAHAHMlXAAhZXNWCSWOrqiSMNVgztU6LCuv9NvP6h13rbWWzNbSXAGa64/mF2t2q2rg7+ABaqSjr7ulxplWfXquoTFu6mrTYPHVgT63aus/3e8X9x34U7US9BFrDHkdqX4uyMiDK66fN96rauEtzn95gNI6Hq7cZt28Pc3vZXls217V8m89qMC8y7Y9RBl2/8ZERRsnXGWcP0ch+pW2+2wdXbs3Z9ru1qO41bFcRh72ucTa2jqmwOzQkIQkQ9dq8SWOynw7s1c135asUf8cPoINNy+y91rBKe6/lUbGfv7JhHVwACIbkKwBEJOoWsUljI2AfxhqsmQJ0UQU2inG933RM/7Cz3RraawBzZL9S7Ww8bLUCu7W6hoM6/85ncs4mNq36nHra8frzqjdDbenqdx3K6y4c5Xt7JmXiQthrAns9H4U5jh5dO8dWGRDl9dPGewU5LyV5e9laWzbK9dTbS7d95z+/2ei1TIKuQROJrb/bsRW9I1sb2Y8wJmhEuQ50OkGPqSg6NMSdBIhjbd6k8bufnjeqXPNfqPP9PnF3/AA6yNYq2I+KSaG1Fc6G81d6rIMLAMGVxD0AAEDhCxKwb812oi1bgC5oYMOR5Di53z/u9X5r6hs1r6pWdy/eoHlVtR22eRT8/mGXmp0cVmtoL7OfF6zYohn3LQ+9NXUq6fzHFa+n/b3pvjO8vFTzZ07QhKFlaX8/YWiZ5s+cEGgGc2rNvBIfx8GsKSNyPr7185IycSH1WT0O3bdc56PUcVxdt0cfP/X4nOceE0HW4Yyyk0DcbJyXkrq9TI7pdKJcTz2d1tt3wYot+u0LW4xex3SdeVvnOVvfh0227vfa8/pZHUkfHd1PNfWNsdzPZBLFto87CRDH2rxJ4/eYnHF2hfH7AIlio1WwUyJVXhP8dQxw/krPZDIVAKCtwr5SAAASwUYrH9uJtlzByCCBjdRrD+jdLbHr/VZt3KVbHlunNdv2d/hd1Gu3mFbJhJWgyBXAjLpqK9tsYtNWsz95ukbzZ07QgqvODrWlq991KFOBy1zbNwkTF9qbPn6w1mzbb1TFkkumz5mtVXdp185qTLMvTxhapr1vv6tXt3tPTKQq755Ys937oFuJOykQhdRx9EzNzsCvleTtFXRt2ajXU08ntX1T53LTU/mv/16ndW82+r5e2jzPRbk2shdhtm70su66K+nJV+r15Cv1kpKzFl0U7enjTgLEtTZv0vg9JuNqqQ1YFbhVsCNNvUsaNtnGaHzj/NUR6+ACgB0kXwEgJsW0BqyNVj42E21egpGmgbJ0yaSkfdf/9y+v6FfP1mb8fZRrtwT5wy6sBMXBd7O/bpCqrVQw/fTBvfTFX/9DdR7X68vWmtGk1Wzr1wu7pavfNfOSlkzwo7y0i/XXzBRozdWqu/HQETmSpp52gkaU92izzas27jJq4UllQEfZEuCmkr69gqyDmYSq3tT2tVGBa3q9tHmei3Jt5FzCbt3Y/rM+/9puPfVKfcYEepLWops1ZYQuv2+53JAmbsV9TYxzbd6k8XNMxt1SG2hjxzpp1QKpfm3zv8vHSKdNz90KuHy0VDFRqqvy/57HDZc+/l+xJV4lzl/psA4uANiR7L/sAaAAZQvUJmWGvm02AvamAb3PnV2hIX2OaRP4kJq/h+q6PVkDlH4CIo6kWy89RdPHD+7wuySt95sr8ZoS1dotQf6wCytB8fOlm7Sybm/aYzFI1VbrYHpNfaPnxGtKptnEE4f30ayPjtCdT/lr9RT17GQ/x0GSkgkpXsZie5/MFGj1Wn3tSvrzqjc0f+aENvuyaeUdlQFt5UqAmxrQq5vdFwyJybXN9Fo+sHc3bd3zttFz20udW2wlzE2vl7bPc0m414hqgkZq3fWb//xKzsplv99PWNedicP76NZpp+i6hasDv1Z7SUkCkEhsy8sxmc8dP1BANi2V/nK1tHtj259veFKqurM5CXvhj7MnSCuvleZfIrlN3t6zx/HSR74vjb3CdNRWcf5qi3VwAcAOkq8AEKFcgdokzdC3yUbA3jSgN6TPMbpy4lBJzQmLHz66xnPi229AJOnfWdXGXZ4SrynZqi1tCfKH3QUn97c8mg9kOhZNk8VfmTxM1134waxx27OJj+16lNXXS4okJBP8TJixeaxkC7SatupuzaTyjsqAD4TZfnzbXjtJxiQyvZbPnDS0pd1866RYpmt6Jqn9cV6V92uhF0Gul0k4z9kS5QQNG+fB1qKYHDl9/GDtPfiubnlsfaDXaS1JSQASiWbyueMHCsATP5D7/E+VdaniHWul33xK+j93S2NnpH/MsMnS1LnSolk5ErCOdN73m5O1CcL5qy263QCAHZwVASAiXgO1UVUcRslGwD5oQG/Bii2a/fDqjO3eMiXbCikgcqtBsC/s6sggf9iFvYZaumPRNFnc/ai2n9PmbOKa+kbjdSaTMDs5aJVRmNWxfifM2Nons51XbK7BNHF4H/Ut7aJHXtqm9W82r/886vhSXXLGwIzbkMqAZjZa1maShOMyLN2P7mT0vEwt0k33x+df2200jmxY6yy6CRq216KzOTky1zXpqsoT1av7UZq90Hy94RRHyUsCFNJ9c5SS2PEDReCRL8t9+Q/ZE68tXOlP35B6DcpcATv2CqnXYGnZ7VLdcx1/XzFJqrwm1hbD2XD++gDdbgDADpKvABAR2zP0803QgL1pUmNn42EtWb/DU5VBpsR3IQREauobtXrbPqPnhlkdGfQPO5P1Tv1ofyzamgVs43VsrDUZ5+zkbOOvKOuu8UPKNGbAsRmPs7CrlEwnzPg9190y7RQdfOc9z+cVW1XTmbbfkld3Zmy7LVEZIAVrP+5FIVUNtL9uPly9zfdrZEvOed0fpbbVs0+9Uu97HF4kvZtAFKKYoGGze4StyZF+rkm5Avx+DOidvDblhXDfHJdCqoRHgm1aKj3xfal+rcfEa4rbnFjNljwdNrn5fzvWSZuWSYcbpS6l0rDK3GvHJgDnr2Z0uwEAOwrnL3sASDDbM/TzkY2A/awpI/TZe5f7et/rH12jTbsOeH58tsR3PgdETAOVUrhVWEH/sPO6XzmONOS4Y1TrY19IaX0s2poFHOR1auob9dO/bdCil98MXDVjIzFmEpzIVWVU13BQdQ0H9VB187/bB66jaOFuOmEm7HblNqqmg26/VOvMny7ZqMY04ynt2llfP2944luxmwpyPvWiEBLWNiaHpEwbOyDr76ePH6yNO97K2Vb/vudqNaK8VA9Vbw187sykkKuWvYpigobN7hE2JkeanFMzBfgfrt7mebKcK7uTNW0nG/L5vhkoWNW/8dAaOIu655oTq7kSqeWj8yLZmgnnL7rdAIANJF8BIAK213fMV0Fb+fQt7eL7Pf0kXlMKLfEtBWux6KUKK0jALugfdn72qzmPr9fPl77maVytpY5FW7OATV5nVP9S3+sb+h1Xa7m+U9PKU5O1MlsHrgf07hZ6C/egE2bCbFsWtGraRpXXghVbNOfx9Rlfo/HQEc15fL16dz+6IBOwYSbYCqFqIFciyq+D77yX9fdVG3fpvudyr+Ha5ErXPbwq4/IDNhRS1XIQYbdutNWFwsbkyKDn1NYB/pr6Rt206JVA4zERxXq3ABJg09JgideW11mW14lVeEO3GwAIjr8OASACNmfo57sgrXzCrjZq/175HgBPCdpiMdsfUjYCdjb+sPO6X3U/ymy9wdbHoq1ZwH5ex5H0an2jtaRBtnF5+U637jmYdb26F2sbdPm9yzXn0o6Vk6ZrZaYC1yP7lYbewt3GhJmw2pYFrb4OWuVVKOuXB/lewkqwFULVgMnkilxy3Qv52afDTLxKhVG1bEuYrRttdaGwca63uaxIHJM1o+gkASAhlt0WPPEqNbcSRlFgHVwACIbkKwBEwNYM/UJi0sonymR0ISW+5y7eYNxi8ZQBx2b8nmwG7Gz9YZdrv7JxLNqaBexnrULJXtIg27i8fKeX37fc01hcSdcuXKXV2/aqvLSrenTtrAG9ugWq3G1ypfXb/QV8TKqCbE6Ysd22bGfjYZV27Zy23W8mqWpKG1Ve+b5+ua0JI7YVStWA6eSKbLKdt8Nef9ePQqhaDkMYrRttdaEIeq63vaxI1JM1C2UyDQAPdqyT6qrsvFYXrnXFhHVwAcBc4Ub1ASBBbM3QL3ZRJqMLJfEdNDA9+6L0LaXCCNhF8YedrWPRVrI41+uk2MplZBuX1+/UbxJ4/gtb/D0hBH6rgpI6YcaknaujD9bMDFpVFff65UHPDaYTRtK9r9/ETzaFUjUQViI0V/eFJCiEquV8Y6MLhY027iYyXZNsX3tynTPzfTINgHZ2rGtuCXy4sTlBOqzyg/bAm5bZe59hlfZeC3mDdXABwL/CiCwDQMLZmqFf7KIM+BRKcClIYPqL5w7NuB3CDNiF+YedzWPRVrI49TpzF9foJ0+ZVylnc96H+up7F4/OOq4wKtaSwm9VUBInzJi2c3UlXbdwtRZWb9Pg3t2N3ju1/eJav9xvtWq6Y3Jn42HPE0auW9g8YURSxvcd1b9UjoJNjPjK5GG65IyBBXOtDyMRmuteKAldKgqlajlOJtdRW0sWmEg9z3alqs12yrnOmX1Lu8Q6mQaARdW/kZ69Q9qTZv3ziolS5bXWWgUf6nuaurLeKwAAnpB8BYCI2FonspiZJM5MFFLi2zQwOK6il37w8ZPS/i7u6regbB+L6ZLFfgPJVRt3ae7T4SReJelfRvbN+v5Jat0ZBr/VREmcMBM0Of5ibYNWGH7Hqe0Xx/rlfqpVB/TuljHhUNq1s/d1QSV99bcr1Xj4SMbnrN/eKEcKlIAtL+0ayzkxrO4CthOhXs6/UXSpGNW/VD27HcVaZyEI2gY8aBeKoOd625WqNq49Xs+Znzj1BF9jTgkymYaWlYC5tMfPgZXSX66Wdm/M/MS6Kmn+JdKYaYHH4ErqetF/Bn4dAACKBclXAIiIrXUii52fxJkJJ+GJb7+BK9PAYLagXFzVb7aEeSyaBpLDrjrN9RmS0rozLCbn0yRNmLGVHDfdxVLbL+p2zH7am1+3cJWkzJ/Rzxq5krTPw+NdNSdfO5c4OmJwAEddtWljvdtsbCZCvZ5/TcfrNWle4kjXf+Ik1joLga1144N2oQhyrrfdJaGmvlEf6leqFbUNnvfP1uPxc85c9PIbPkb8AZPzVtjnHqCQpTt+RjhbVdH5dxre6WWVeHkRt0laszDQOFxX2nbSTA0cNjnQ6wAAUExIvgJAhGytE1mMWgfVPn7q8frzqjezrj1Z4kgVxx2j2l0HfL3PnACJb7+BPz+PNw1chdE+NY7qN9vCOBaDrCcZZtXphKFlkqR5VbVt9jVJLftf9ZY9ob1/3EwrUpM0YSYpyfGo2zH7mZQQV8dsVzJKvEr2kpVeriW2El3Z2DoG/Jx/TSsFp40d4PvYZq0ze8JYN970+wlyrrfVJSHbPZ6f8URxzvR73ori3AMUqvbHz6c7LdHVnR5U/5K9Bq/mSl2OlQ7v9/9MV3p12Oc0avodBu8LAEDxIvkKABGztU5kscgWkCrt2jltNVMqcCvJV5Xs9y4aZRT48ZsY9fv4IIGrMNqnRl39Fhabx2KQQHKYiTXHkfa+/a7Ov/OZ0N4jaqP6l6qmvjGSitSkTJiJe+JCqmo9ynbMhd4KWwqerPR6LQkj0ZWOyf5xyoBjNW3swEDnX5PKxYnD+yTi2C5WYa4bbyLIuT5ol4Rc93jppBtPVOdMv91Aojj3AIWo9fFzTskazen8Sw0qCfg3w+H9klPSXAnr0bZOA7Rn8i06+dxPBntvAACKULKioABQRKigyC1XQKrx0BE5kqaedoJGlPdIG7j1Us3gSJpzqdmMe7+JUb+PtxG4st0+Nerqt7DZOBaDBJLDSqw5kuRKr25vDOX145BqAbp1z8HIKlKTMGEm7okLm1t1ELh07ADjlph+JKXaNyxB1wj2cy15qHprZIkuv9eb2ReNtnKMmlQuJuHYLka214239f0F2R+8dFpJd03yeo8nNV/TLz9rsGacPSTteKI4Z/o9byUtyQ7kk9Txc1vnn+vTnZ6V41h64TGXSGsfSZuATS2n4EraetwkvffRmzRk9Ic1wNJbAwBQbEi+AgASyWtAypX051VvaP7MCWkDNmFWrvlNjO45+I7mPL7eVyLVRuDKdvvUKKvf8kHQQHIYibXR/Uu1fntjbK1Yw9B+/4y6ai3OCTNxB6Pr9x/y3RYzaPI77mrfMAVdW9zvWrh+zwPZEl25xNWuO8i1nslw0bK1bnxY64j62R+8Vq0O63OMbv7UyR3G47dNcE39WxnHZnrO9LP+sZ/zlu0ke7FjkkhxqalvVOe6Z/T3o3+uE0osLw3Sd7Q04wpp2e1S3XNtfuVIUsUkOZXXaBBruwIAEBjJVwBAItmcLR9WdYvfMf7XkzW+Hj/n8fVatXWfrzFlClzZTkLbrqbNZ0EDybaSD81V4Mfr6x8ZoR8+uiZvEq83Tj1JrqS1b+zXitoG1TUc7PCY9vtnTX2jauobNWl4H42t6CXJUfejOhVsQNJkwoNN+w8d8dXC3UbyO+5q3zBNPfX4QNsminUd2ye6/IirXTeVrPnBxrrxSVhH1E/V6ubdHatibScnTc+ZnzjteP1l1ZvWJ0vYSrIXu7AmGSDZdiz7lX571C0qcUK4m+9SKg2b3Py/HeukTcukw43v/7xSKh9t/z0BAChShRvVAACEJuzAZliz5W1Wt5iM8Z33vK+vI8l34jUlU+DKZmA67OqmfAqeBw0k20istU5m5NtamTcueqUlgPj/Pn1a1u/eSxAyqftJUH4mPNj2/Gu7PSXxHEm3XnqKpo8fHPg9CzmYPKLcfB+N6vgOWnkcZyKUStZkC7pufFLWEQ06SdB2ctL0M37jIyP0mfGDrU+WsJFkL3ZJmGSAGGxaqnNeuTmcxKvUnGBNKR9NshUAgBCRfAUAeBbV7Ot8mC2f5PUIcwWubAWmw6huyscZ/kEDyZK/xFrzum8VKi/tkjaZkeR9M5P2AcR0+2exByG9TngIg9e3cyU9XL3NSvI17mrfMAWp6o3q+LZVeUwiFO0FXTc+CeuI2pgkaDs5GWRJiFQXDpuTJWzcGxWzpEwyQAyW3aYS+Zuw61nFJJKtAABEiDtbAIAnUSY+8mG2fJJn5kcZuDKtbkr3+Je27MnL5FrQQHLq//dTSZzt80e9b546sKcuOWNAy3e5Zts+Laze5vt1sgUQCUI2yzXhIQlsrtkXZ7VvmILsm1Ed30k7fvKpGwKyC5IkTMo6ojYmCYaRnAy6JITNyRI27o2KWRImGSACO9ZJqxZI9Wub/92jn1RXJVfvr79qk1MiVV5j+1UBAEAWJF8BADlFnfjIh9nySZ6ZH0fwxWvALltlqxdJTK4FCSS3ZquSOMp9s8SRrrtwVJsxXfaL541fL1MAkSDkB9pPeHimZqeWvLozlPdyZLZmqK0uBF4nJTjvDzQfcrTpjn0/oji+g47RpnzshlAowkx4myYJk9IZxcYkwTCSk2EvCeGHrXujYpSUSQYI0aal0hPf/yDp2o71xKscaepdzeu8AgCAyCQ3cgwASIyoEx/5MFs+qcHeMAJXtgKwuaqnvUpici1otUmKjXUSo9oujqSPju6nmvpG9S3tYlyV1F7rAGJNfaMeeWkrQcg0UhMe3jp0JJTka4nT/P0++Uq97+farM70Oilh656DsbRk9sNR5mPfK9Pj23EkN+D5KWrF3mo8LlEkvE2ThKbnlmdqdlpNItuYJBhWcjKMJSFM2bo3KjZJmWRQCBLZNaH6N9KfvqnIpowdN1z6+H+ReAUAIAYkXwEAWcUx+zofZssncT1C24ErmwFYr9XTXiUtuWa72iRI6z9b++ao/qXq2e2ojO1tXUlPvlLfkpw7c2iZPtSvR6D3TLnhf9foPVeBPkOxBCHDqIRMBedr6huNkq+2x5RuUsLBd9+T5Kr7UZ1VU9+oicP7aP7MCVZaMpd27azGhLaWNzm+R/Uv1ZUThySiGs4rWo23FVUCIcqEt0mS0PTcsuTVth0CgiaRbT0vrOSkjYlc6fh9vSRV4uaTfFh+JekS1zVhxzpp0zLpzVXSy79XJInXHsdLH/m+NPaK8N8LAACkRfIVANJI5CzZmMQx+7qmvlEf6tdDK2obPP1pGtds+SStR2g7cGU7AHvLY+usb6ekJdfytdoknRJHuv4TJ7UJ3j7/2m499Up9xmPyxdoGrbA0GeH5TcFfp1iCkPsPvRvo+RNPPE6nDuql7kd16nC961vaxew1Q9rHR/Yr1c7GwzkDqn1Lu+iRl7bp50tf8/0eX508TD9busnGcDtwZadq3+/x/Wp9oxw5WZPTUZ6fvKDVeLMoEwhxJLzbJwk37z6g+n2H1O/YrhrS55gO5yBbnzVoEtnWJMGwk5O21nANsh8m6d4oX+TD8itJlqiuCUvnSMt/Ib29O9z3ae+8H0iV10b7ngAAoAPuzgCglcTNkk2AKGdfm6wHGudsea9Bs7DZDlzZDMBWbdylWx9brzXb9lsZW2tJTK6FVW1iMg7TfbP9MZVKeN3851dyToZIwDyEFsUQhKzauEtzn94Q7DVe262q13a3XONa76dJ60LgJ6Babpg4DivxmmKjaj91fM9euNrTMee+f66eP3OCFlx1duznp1xY77BZ1AmEOBPeOxsP67E123Pef9vsOhI0iWyrajXpyUkb+2FS7o3yRT4sv5JUiemaUDVXWvJj6cgh+6+dTb+TpQv+Ly2GAQBIiMKPSgGAR4maJZsgUc2+NlkPNO6AlJQ7aBa2r0w+UdddOMrqa9oKwNpa4zWTJCfXbFWbBGGyb2Y6psKoXA5bMQQh/RyruWS6xiVlzT6/AdVpZwwMZRw22Kjanz5+sOZVbdb67Y2eHt/6XJ2E81M2rHcYfQIhzoS33/tvm11HgiSRbVatJjU5aXs/TMK5J2nbOJ2kTXzKJ4nomvDIV95vLRyhAeOkT/63VD462vcFAABZJTdqCQARSsws2QSKYva1n/VAHUmXn1WhGWdXJCbI0D5odt+ztdq6923Pzx/Yq5tG9OvRZj0yr7of1cn3c7KxFYC1vcZrOsVyDAaRLaArKWcAMszK5TAVehCypr5Rj7y01fqa0+mucUlZs89vQHVFXXLW427PRtV+TX2j58RrSr5Uh7LeYfQJhLgS3qb33za7jgQ5LmxXrSYhOdlaIhJZluRbd6OkTHzKJ7F3TdixTlr8n9Krfw3+Wl71HCR98qdUugIAkFAkXwFAhRVcsC2K2dd+tr8rJTZ4nAqajexX6itgMudfT1VNfaNR8tV29aetAKzNirx0Cj25ZlumgG66n3ld4zWp4gxChl1RY9Ka3a9017i422KaBFTrdh8MZSw22DhvF3J1aDGud9j63HHw3fciTyD8/+ydfXwV5Zn3f5MEEgInQAgJmEgILzG8qihFGqpYlNJVu4hdabdoH5dtbfu0i267Ylu7rWtbpS9aarfWFdZVbHfpGvQRrVREEEwxIqEQXkIIhMQESAgJcCC8Z54/wsSTk3m5X2fmJNf389mtnJwz95yZe+6Zc/2u33UFJXiLPn+rrjoic12E1bUqS+BClkISsbpRWBKfEonA7osHNgDv/gyoLRXfBi+DRwGfehCYcq9/YxIEQRAEwU3i/kIlCIJQRE8KLuhCZ/Z1Tzz+IgGToYL9CVUHW1QEYEXOKQ+U4a8HP4Q93RhwD0LqCo774ajRXcY7Frs11hIY3tp1BKvK6xE9exGRtBTMm5KH2ROGad0f0YDqiMx01LWET4Q92Hwaz5fWSM2/nuwO7U39DlWuuzICQhCCt+zzn53oWV7XKpTIpuK6CJtr1UL0vtdTEjwSubpR0IlPiYbodbyx6qj4c2H5i8DqRYDZLjQ2MzMeBPr0R9P5Pihtn4D6lJEYcCYFxSH+PUwQBEEQBImvBEEQPSa4oBOd2dc99fjzBkzC0t9JRQBW9JyywDPH/HShJLrjxU9hTycLbhhh61rRKY764ajxo4y33ZjxpcTtjuGaXY3aSzaKBlQ/UZCJ+ta20M3rFzbXdv636LELgztU17oXlvuhblSvuyLXiXUODx47LTRmGEodx4qez5fWCImv/7u1HlPyB/coEUv2vpcICR4sa1CiVzfqqc5qHYje39bvPdpl3WC+Lx/Y4I/wmj8DpSO/GXM9nwdQ1fnn2P2leUIQBEEQ4YLEV4Igej2JEFwIA7qyr3vy8ecNmIShv5MKx5GucxM7x9yOqZ99vXSM5XfgJAhhTxf3TB/Z7TWd4qhfjhrdZbztiL2Ogy7ZKBpQnXBFBqaOVNcbUgeixy5Id6gfa2wY7oc60bHu8lwnKhy3MoJ3VWMUG6v4RVLA/RlDdN7VtbSFruysDCrWbB0JHqqeb1jXoJ5UXUels7qnCnSqnu2Z78vv/ky/8AoD64d9GQtd7ocf1LRgwfIyjBzSHzXN3RNpwtjTmCAIgiB6CyS+EgTR6wmDeyRR0JF93RuOP2vAJAz9nVQ4jlSdGwPAXVPyMCE3o3OOlVY34+5nNzsG3K7OG4jl79W4ByiWleGOq4fjm58eKxVs4g1uel03forGsQQh7OnATgjgFUdNmGg7f4l5bfPDUaO7jLcT1nUchpKNMkJjYU5EaW9IHYgcu6DcoX4J8WG4H+pEx7rLegxUOG55BO/Ye19VYxQf1rbg8IlzwmO7PWOIXBcWYSw7K4KqNVtFgod17ncdOokPalpsy8DzPt/wrEGnz4slA4a9uo4oQT1n+oXM9R+P53rQtMeHHq8G9t3wOBa+29/zejZN2AqvQDh7GhMEQRBEbyFxItcEQRCa6E29xVShMvuajn9XwtDfSdZxpGrfTAAftbbhF3dfDYAt4MYScDEBvLb9MF7bflg42MQT3FxcsgP/WVqDyiPRbn+3xq9vbQvEWRiUsKcaJyGAVxxdXFLR7XWnOeKXo0ZnGW83rO8bhpKNskKjU+JQ7qB++OqKrUr3VRSRY8ezVhuQd4f6LcSH4X6oAx3rLquwrsJxyyp46+oj7jUuz3URTxjLzvKias2WWXd5zj3P8w3vGjRvSh7zvseSCNV1eAm6goVfyFz/8biuBwfelR/AjeyJwJyf4PtrU9Fu+iAmEwRBEAShBRJfCYLo9fSW3mJhhY5/d1gdxrrKhsk6jlRmnlti1dHoOS2lQ0WDTTzBTROwFV6t8RcsLwPMjve54RU4EZkPqoU9A97fw41pBZmY7OFejsdpHqoSOKw58sAthYikpXQe36aomHOL11ETRBDYWmPDVLJRRRlau8QhVWuVNQ+vGTEI//jCh7YOLy94j521Vj9cUsF03TW0ngEgfu8IQojvif0OVa+7PC5UWccti+Bd1RjF0+/sw+vbD0vdD5zG9zrvrM8wToS17CwLqtdskXVXxFnNKgzxrkFbBNf2RKquw0IYKlj4hez1H4/j9XHO/rleDAPInQKkDwGyJwBXzweyx12+njcqG6UnJJcQBEEQRKLRs54qCYIgBOnpvcXCDh1/e5wcxn6UDZN1HKnMPC+tbsabO49oK43LG2xS7VoyOQOU8YETmfmgUtiblJuBG0YNwXObarg/+/WZo3Hntbmd833mVdlMZWLd5qFKgaPdBJ5cW6VkW7zH3O8gcOwaK3oMdZRs5Amo5g/pb/u6nYCnYq2K70ctIrxa8B67vMHpTFkPJthc+E5rRdBCvMqKG0Gjct3lKbsseu/68vR8jMzq7yl463K6WvA8/1nPMN97pQK1x/ivx0QtO6t6zeZNxgMgLHp5CUMi87dWcC3uaeJUGCpY+InXbxhebK+PVFXrgwF87mlgyj2246omkZNLCIIgCCIRIfGVIAgCPb+3WNih48+On2XDZBxHKjPPDzaf1l4alyfYFFQZWIvYwInsfFAl7F0xMA0LbsjHd1d1L9vrRSQtBYvnFHV5zW7utV24BMBEep8UpnkY1rKBvMdcdL2bcEUEuw7xOzMeuKWwc0zRY6jr2FsB1Ude3enY2wzo6HsWO++9EhQWzijwdFvHOlu9ejfLwHvslq7bx5zA4eXCd1srwiTEJzqq1l3essui53BkVn/cV1zg+h4VfWTdEHn+Kx6Thc9PycMvBRJnwnr/8ELHms2TjHf3s5ul5oCbMCQ6f/Mz07lE2J5WXSfoxJmgsHuOLK9rxfq9Rx0/M9aoR3HSTgzAGZxCP5S2T8Q+M8/++hh1k/xO9s8G7noOGDXT9s+61iG6LxMEQRCEf5D4ShAEcZme2lssUaDj701QZcNEHUeqMs8bT54V/iwPrMGmMARlS6ubmUsxu80HVdfT6OwBwoH36NmLWLJmj62oKuN2C2vZwNxB/bjeL1rGW0R4BTrEcAvRY6j72NcecxZeLax5v68p6iqsflDTgg8PtmDhjALsqD/BdP/RKfrvazrF/F7VLny3tSJsQnwiI7rufn3mKObkEzt0nUMVfWTdkHn+C+sapgtd35clGU/VeuQkDInO36kjM/FRa1uoq+voLKve2xNnYp8jny+t6Sa+jjXqsSB5LT6dtA1XJnU/VmXtRWg9/SCAuHmRPQ7ILwZqS/l3KjkV+MRXgM/8xPVtutYhui8TBEEQhH8k5q8KgiAITfTE3mKJBB1/d/woG6b62Mee01e2NeCZDfu5t5EzME14fF5Ygk1hCMqeOnuRez489vpurHngxi6vq+rPW3uMLbDpxDMbDnT5t4ry2WFN1Ni0rxmzJwzj+ozKMt5exAblRI+hjmNvrU3LNrH3A243wVQGu90Elr9XgxULp2FoJFVqDZRdH1ZvP4T5U68MzIXvdO/oiSJWUM8aor3uF88ZJzWurnMo20fWja/PHN2tMgIPYVrD/ED393VLiFK1HjkJQ6Lzd0JuBqYW+FNdh3dN8aONCCXOfEzssfxk0k4sSlmFaUmVju83TWBaUiXMbV8H8i50Lw1800PAijsBs51tB66cDhR/Cyi6jXt/VRLm+zJBEARB9DTorksQBGFDT+otlojQ8e+O7rJhugNAhTkRLJ5ThK21rdwB55EO/Rt1wBJsCkNQtu3CRe75UHkkil+v29fN1aFC2JPpcWmHivLZqoRl1ayvbOL+jMoy3l7EBuVERSLZ9Ts2gN0UPYu/fnQCFQ0npLbphSU6rrx/unTCiQwm2BNndAXH7e4dPUnE8kPw8CKIXvc6zqFq93U86X2SpT6vew0LW7JgUGs2oG49chKGZOZvYU5Ea3UdkTXFrzYiPTFxRhTr+vibul/i3pS1SDI6BFbDsH+/9bphtgOr/wkYdGXXEsGjZgJ3LAVWL3IXYI0k4I5f2/Z1Zdlf1WtsGO/LBEEQBNFT6XlPVARBEATRA9FZNkxnACg+MHnXlFx8eLCFK+A8NJLKNaYMbRcu4fnSGtdAajhEPYdIkQdPra3CdfmDuwReZIW9/CHpqD2mVnwF1JTP9tMxykr98TNCvdRUlfH2IvZYVzVGcVVOBFtqWsByCGVFIrcAth+o6HOnYn1g3Q+dwfH4e0eQoo5K/Oyb7kYQve51nEPdPdBVzHEdQrdKAV+1gBuEsA+oW4+cjpvs/NVVXUdkTfGzjUhPSpyRpnQp/rvpcST3OdP5kpPw2g2zHXj35937s065Fxg0ouNvte91/1z+DOCmf3Hs6+qF6ufYMN6XCYIgCKInQ+IrQRAEQSQAQfdq4w0AuQUmC7L64+Cx0zA5As5+iZ12ZZEn5mbgu58d1+V7BynqTSvIFHYDObnqRIW9JAOYmp+pRXwFxMtnW/jpGOVBtJeaFTx+a9cRfHXFVuX7NW5YR9UBERFUViTyCmD7hYo+dyrWB5b90Bkct7t3BCXqqCKovulOBNHrXvU51F2aVMV3Vy10qxLwdTmwgxD2rXFl8RKGVMxfldV1RNcUP9qIWPSUxBlpXvk6sP0PkPLS174HNO3p6Pcay6iZHf/XtAc48C5wLgqkRoBRN3V/Lycqn2PDeF8mCIIgiJ5OUtA7QBAEQRCEN2Ho1WYFgLxYuaUO9ywvcwz01DSfBkxgVJZ9OeFpBZlYsXBal6DlolljkSRm9pRmZ8NJfGlZGX7yxu7O16xgiN/7ZAVOZBwmlqsunuIxWVh5/3S89eCNuOPq4Z7eWitwOyE3Q3hfWHDaX1bmTx2BFQunYVpBpsK9kkNWsGg4fsb7TQJUHoniO3/c7nr92mF3zfLAGsD2AxVikor1gWU/rKC6DuzWGNbvpVrUUYWO+50ssevuD+8Yj2/fWogf3jEebz14I1beP135MVR9DnW6ryfnDURpdTOeXrcPz5fWaL0PsK5hvGKbkzPY6znJEnD/uOUj94EcUPV9eZBdj1iEobCtQSJrikwbEVF4nqF7nEBX+Qbw2xuA7X9Qs70D7zr/LXsccMPXOpyuN3xNWni18LqeR2X1Z35mD9t9mSAIgiB6OuR8JQiCIEJN2Ppp6cTtu4r+WG6KnnUsX6mjjyxrYNIEcPDYaSy5axLazl/yPL9hcDA+t6kGAPD928YD8HYtFQ2LYG9j1NXha2EFTdzeGhs4kS3F7OaqK8yJ4OkvTsEXpjYzObJkAoKsyLoR48sN7mo4iS0HW1CruFctK/GCBe86p8ttZgJ4ubye+f23jMvGQ3OKpNdkngC2bnYdOoGn1+2Tvt9Y68P3XqkQcoazilq6XPhO95wg3Joq0N03XRYRN57o85HKc6jzPO+oP4Ed9V17Pcs6QmXLzqpwLPrlwNZVZtcN0fWIRxgKyxokuqa8sq1BaDyZ56Cg3NCB0bQHWPcoUP0OcOmc2m2f0//Ma4fX9VxazfbMThAEQRCEv5D4ShAEQYQSXeXYwgjrdxUpvfvMhgN4ZsMB22Omo48sb2ByVXkDVt4/nen9VsDtx6/vxp4jwQQ/nttUg5lXZXceR69gCEspVSvY1dp2Hr9ZX42ojbAWSUvBN28e0+lSke0rySLesQZu/eiBq0psLMyJYFtdK1Ztqw9U7Cspr+88hiLrnE63GQ9v72nC7PHDpAL5IgFsnazZ1Yg1uxo7/y0r9jx37/WY/dRGoc+yvk91YopXyckgRB1ZdPZN9xsVz0eqzqHfPdBV9OQVLTurSsD3s+QsoLbMrhei69HCGQVc5zMMa5DomlJ5+KTQ52Sfg8IiWmvlwAbg3Z8BtaX6xkgN9n7gdD2H4ZogCIIgCKI74YjcEARBEEQMqvppieLnD1ee7yrjcLI7Zqr7yPrhLCoek4VIvz5cY6hmyZpK/OLvru42R+4rLuj2XtZgV31rG5asqXQ8t9GzF/HEmkoMTu/bef4WzRqLBcvKXN2yTvCIdyyBW909cFWJjWEpb7uz4SQWLCsD4Ox2dlvnwhQcXSzZG1M0gO0XsvcbP/rtqXTh85Sc9FPUkUVX33S/Uf18pOIc+t0D3a+evPGoEPDD7sBWgUii3PL3uia2sRLkGuT32qDiOahHC3QblgAbHod7DRkFjLpJ7/YlSaT7MkEQBEH0Bkh8JQiCIEKFX+XYnMb2023L+11XLJwm5XCKP2aq+8iKBia/8sKH+Om8SUzHNgwuuR31J2ydbE5zhKVUGFOpZhN4qGQHWtvO4/6bRqN4TBYeuHUsnlrL35dQVx/Bh0sqtIS9VO0vb3nbgf364MSZC0rGjodlN5zWOb/dZm6Yks4sXQHsJKPDTbX8vRppUUj2fsMjTon221Ppwg+TuC+LdTzK61qFPh8Gl3lnufRDJ1Gytd5z7fBbnGR1OxoAbrpqKMYNz0B6n2S0XbgIwEB6n2QMSEtBSXk9djawuQJVOEJ5USHg9yQHthu8iXJBnE9ZRNeGouERrN97lPtzKo9NjxLoDmwA1nwPaNqlf6z8Gcr6uMbSI8VwgiAIgiAAkPhKEARBhAy/y7FZBOG2FfmuK++f7upwYt2OFagXwelzooHJ2pY25mMbZpec1xxxCnbxCoKPv/mxA3bRrEK8WXEElRxlmHlddazMnzoCOxtOYsX7tUq3q2p/RYR7XcIrD07rnN9uMzdknFk6xK3Y8oljsgcocTvL3G/87LfntM6IlpxM1KCwWzIVD3Zl4FlQcdxkvoPfYpZsSdOqxigeXb2ba0y/HaEqEtZ6igPbi0R1+PJct6LX1p3X5mFr7XGt1RB6DeUvAqsXAWa7/rGMJOCmf1G6yd7UYocgCIIgeiskvhIEQRChIahgTRBuW5nvGutwemVbA57ZsF9oO6rLYcqIKKzHNuwBSN45IurkXVzy8Rg/uH28dlcdK/dMz1cqvqrc3zAL917YrXM6en3KIOrMUhVYzBvcDwtnFHQLllui0GOv7+ZKUrBD5n4Thn57PCUnEzkozOLyZeVXb1fhuvzBzN9V1XFT8R38FrNkSpomgiNURcKa6oojYSURzmf8uLzXrcwztB/VEHo8Bzb4K7ze8Wtg1Exlmwy6xQ5BEARBEP6QWE/xBEEQRI8mqGBNEG5bFd+1MCeC7Eiq1HZUBoBkjwnLsa1qkhNP/IBnjojOAxNdHcx+ueq8KMyJYGJuBnPpSDdU72/YhXsv7NY5L0HPTzZWHRVy96kooZxkAEvu6jpX4gWglGRDePuxyNxvwtJvz6vkZCIHhVX3deZZz1UdN5XfIQgxS6SkaSI4QlUkrKmuOBJWEuF8Wshct6LP0GF6bktY3v2ZP8Jr/owOx6tC4TXIFjsEQRAEQfgLia8EQRBEaBANuvyhrBYHm09jZFZ/7iB2UG5bVYEp2e2oDACpEFHcjm1pdTNe33FYeNt+wjpHZAKNsWPodtXxiEXf/ew4fGlZmdA4FnmD+3UT02QRdQ7lZ6ajtqVN2X6I4jRXujrh6/HMhgM+71kH6/cexfq9R4VckTIllOPXJlXlZp1QIQ6Eud9eogeFecu4s8CynvMct8UlO2DCxPypI2zfo/I7JErSSaI4QmUT1lRXHAkriXI+Zdc7mWfoMFRDSFia9gC1pfq2nzYImPR3wNSFWnq8BtVihyAIgiAI/yHxlSAIIoag3Si9HdGgy76m09jXdLrz3zzBf1VuW965oyowpWI7KgNAKvpQOjl1lq7bBzME5VVZYXEcyQYaY8fQ4apzE7JGZKbjEwWZmHBFRpcxisdk4SufKsBzm2qEv5dVPlYlotv7xs2jQ1Ha12uuFOZEsHjOOO5ecqrhdUVWNUZR1RjFLeNysHZ3I3gOc/zapLLcrBOJVv6TF9VBYT+fq0TLuLPgtZ7zHDcTwOKSCpSUN3R7VlH9HRJlvgbtCGWdpyoS1npDydmgzycrKtY7mWfosFRDSDgOvKtv21fdBnzxD9o2n6j9kAmCIAiCECMxfo0RBEFoJpF7m/UkVB1jnuC/rHNUdO6oCkyp3I6KAJCKPpS7bErW6gyq64JlbsnOebsxVLnqvISsupY21LW04eWtHf+One/fv208AAgLsDrWW1HHkeVOY5nTn5+Sh5fL62V205Gm6Dk8vW6f53Up6yKdd638d2BxRfI6VCflZuCaKwcjO5JqewxUl5t1Ir1vst4BAkRlUNjv56rS6mZ8b1WFsu3F47aei96f7J5VVPemTpRn16AcoSLzVDZhrTeUnE0Eh6/K9U72GdruuY0EWRfOaWhBkpoB3PgdoHiR+m3HkGj9kAmCIAiCkIPEV4Igej2J3Nusp6GibK0Fa0lEGeeozNxRFZhSHeBSIdxZgcl/+K8tOHeRvx/TltqWbkGnpug5qX0KApa5JTvndbmaRISs+Pn+/dvGY+ZV2fjKix+i7fwl5u3oDMCKOo68gu05GamYPT4H90wfibrWNi2JAs9s2N/l307CgGgCRKxgMDZnAB5/s1Jqf91ckSwOVQPAreNzMH30EKags45ys3Z8d1UFDBg98nlAVVDY7+eqoB3PMoJp/LOKyjLBiVau1m9HqMw8lRXbekPJ2bA7fHWIYCqeoSkhmIFUheta9njg048ARbep26YLidQPmSAIgiAIeUh8JQiiV5Povc16IotmjcWCZWVcpSedYCmJKHo+0/smS88dVYGpMAa4hkZShYRXAKg91obZT21Uti8G4DqfDAB3XH0F+qYk4S/VzTh04qySceNLSToFaBfNGosFy8uESirrWo9Ehaz4+V48JgvP3Xs98zVtQG5+egXCZRxHscH2FZtr8U5lExqOnwEANJ48hxXv12HF+3UoGhbxnHMqcBMGvAL7k/MG4uq8gciOpNkep/tvGo1B6X3w8KoKqVLfdi4h1vuuCeCt3Y1oOH7Gs6e3n854v58HVLuf3LanIijs93OVX45nt32UDYrHPquoSqhJxHK1fjpCVc1TGbGtp5ecDbvDN4wiWK9OCK58A/jr/wDnTgCpA4FrvuAsiI66Sc2Y13wJmPtbNdtiJFH6IRMEQRAEoYaEu4MbhvF5ADcBuAbA1QAiAH5vmuYCzu0cBJDv8OdG0zSHSewmQRAJgureZoQcz767H79ZX61UtPDqkyPqHC0pb5CeO6oCU6oDXCoCgapLJ8qw4IZ8/PWjVlTYlDOOpKUgevYiXtt+qMvr+ZnpmDoyExNyO/qZPvLqTiF3MauD4Yl5k7C4hK9kpi5Xk6yQFT/fi8dk4Ym7JuHhkgqma7uh9YztPrnNSR6niKzjaFtdK35fVut4rVUeicIAm+gPj/d44SYM6HZmsbJicy0emzux89+8wv6uQyex69DH166d88fv9caP5wHV7ieW7akICvv9XOWH49lrrVURFLeeVVTMqUQuV+uXIzRMz/+qWgWEkTA7fMMmgvXahODSpcC7PwPOn+r6euVrzqWAs8cB+cVAbanYmD6VGLYjUfohEwRBEAShhoQTXwE8gg7R9RSAegBFEts6AeBXNq+fsnmNIIgehspeP4Q83/7jX1FS3qBl2159cnido/Om5HILZU5zR1VgSsV2VAb7w1Qea31lE+qPdxf0ACDqsJ+1LW34qLUNnyiYjMKciJC7eOWWOlfnYKyDYf7UETjedoG51KtOV5MKISt+vucNTvdWIy//OTaoyDIn61vbuJ0iosIkj2vTMICinAgqj3TvTWZdjwCkxU0vYUCVM+vxP+3B+r1Hubfx0vu1mJQ7EHdPvVKJQ9XufAax3uh8HmBxPy1YVoYld7G5n1jdVA/cUii0v9bc8/u5yg/HM8taqyooXlrdjPuKC6RK0feEcrW6HaH0/O8vYXX4hk0EC1NCgG+8eCdw4B3nv587Caz9V+Do3u4O1ZseAlbcCZiMVXb69AeGTwY++S3fSgzbkQj9kAmCIAiCUEciiq8PokN0rUaHA3a9xLaOm6b5IxU7RRBE4qGj1w8hxrPv7tcmvALegXle5+jp82KBfqe5oyowJbMd1mD/g7cWMol+YSqP5SS8ehHvLuCZIzsbTjAJqbFjdJZ69XCI6nY1qRKyYuf70nX7mMvXWkFFFlF1wbIyAN7uUSenCK8wyRMcNU1gYL8+eOvBGz1LIcdes20XLuKZDQeY9wnQLwwU5kRwY+FQIfE1VlCvauwuRIsQfz6DWm90PA/wCPwPlexAa9t53H/TaOnttZvAr96uwlXDIthrkzDgRGxQ2O/nKt2OZ9a1VlW/emvt5Un2MQDcNSWvs0pDT3o+1eUIpef/YAibwzdMIlivTAj4473uwmssf/09MPSqrk7VUTOBO5YCqxfZCrAmOtbHdhhomfptZN32AxV7rYQwtoshCIIgCEIP4YmMMmKaZqfYahiG21sJgiBcCWOvn97Kb9ZXa90+S2Cexzn69Lp9QvvhNXdUBaZ4t8MT7H9ybRXeqDiMf719vJZeumEj1l3AOkfqW9uYHayiY+g8vqqELGu+iwYVP6hp8RRVeaqNyjpFRL8HANxXXOD6vthr9vnSGqH90y0MyMw569jPUDhvY89nUOuNjucB3jK6j79ZicHpfR0dsLxuqiR0BHtFgsJ+P1fJHP8kA1g4owA76k8oWWt5gulOWGsvb7JPj+v9qBl6/icswiKC9bqEgAMbgN3/j+8zG3/RvUzwlHuBQSOAd38O1L7X5U8GgPfbx+HXF+/EXzZNxCfqN3OX69dF2PshEwRBEAShjoQTXxWTahjGAgAjAJwGsAPARtM0LwW7WwRB+EHYev30Vt7adcSx9KsqWH+0sjpH2y6I3SZEP8eDiOuVN9i/90i0W9nPeFQ5gcJAWU0L3tp1BA3Hz+DU2YuYM3EYFs4o6Px37HEurW7Gw5wlqa0xLAdD0CX6VAV5rLVSNKioo42jjFPEr+BoWIUB2Wu6rKYFU/IHK92n2PPJu2+T8wbizmtzcersRew6dAJrdjVyj6/6eUC0jO5ih/5/ItvbcySKB28di6Vvu98X7ILCfjxXxa6Luw6dEBovPzMdP503qUu5ZNm1ljWY7rUNizAk4ogStvKy8dDzP2ERFhEsrPd9JTTtAQ68C5yLAqkRYNRNwJ8W82/n3Emg8o3uJYNHzQRGzcSb77yDD955Bf3NMziFfihtn4h9Zl7n2+xaFgRJIq/xBEEQBEGw09t/QQwDsCLutRrDMO4zTfNdlg0YhrHV4U8yvWgJgvCBsPX66a2sKq/Xun2REmHezlFRWUiHnNSBaL9W0WC/UxnXWFQ4gcLCV1d0v93bHdel6/YJn+V4kS6oEn2qhHPruIQtOPibd6rx6y9e6/m+eAHh4LHTQuPxfv8wCwPy17T6xcC6bngdTIvnFHUR30TEV9XPA8KJCg6ubtHtZaT1wYqF07iDwrqeq6oao1ix+SDWVx4VLiMfy3Nfvh6FOZFu1/hnJg6TWnPnTx0BANw94YGOZAC7fXps7kQACLWYaaGyb7xO6PmfiCUMIliY7/vCHNgAvPszoLZU3Tb/+j+2/VpLq5vxf9eeQbs5x/XjLL9d/CToZEuCIAiCIPQT4qc17TwPYBOAXQCiAEYB+CaArwJ40zCM6aZpbg9w/wiC0EyYev30ZnS6XnWVCEvvI3b7FP2cFyz9Wp2yvWV65nmVcbUcBV49TBOV+OMqKmRbhEmklBXZYtfKsAUHV28/hPlTr3Sct24Cggi83z/MwkDxmCwsnFGA5zaJlUZO75Oi3BFvXTcyDqZtda3c46p4HlAl8AP2rm4ZN5VIUFj1cVR9LVrjHY2ewyOv7tQiEradF6twkR1Jxd3Pbg69cOmEzHOI39DzPxFP0CJYmO/7QpS/6NiLVYpz9tUOeMvry7SgUA0JrwRBEATRswlXNMxHTNN8NO6lnQC+ZhjGKQDfBvAjAHcybOc6u9cvO2KnSO4mQRCaCUuvn95MRJMwo7NEWJgy1Fn7tTple8sKfl5lXOdPHYEtB1vx8lb1DudJuRmoaDjZ7fW8Qf2UuKNYiD2uVY1RqW2FSaSUKaEZv1aGJcBlYcI58OYlIIjA+/3DLAys3FKH5e+JCa9AxxxX7Yhvu/DxGibiYLLWUB5knwd0iIrWdmPngYp7FasDX/Vx1HEtJhkdDlO3+ScrEoreU9ftaXJMUgqTcGmH7HNIECyaNRYLlpUxJYYZoOf/3kIiVRwJbUJA+YvAa/8ELdV+Ugd2e0kk8VGmBYUqEqVSAEEQBEEQcoQnyhcefocO8fXGoHeEIAj9hKXXT29m3pQ8oXKPbuQPScdP75yk7XyFKUNdNttbheDn1dNywhUZeNmpSL8gk3IzsPpbn7LNGC+tbsajq3erHdAF67jOkDy/YVtfvIQsO+zWyjD2/7ULvLEKCDy4BUfd3A5hTAxScXys7yjjno3nmQ0HsLX2eGegktfBxNvzGuB/Hojdl31NUazeflhLNYB44c/Pe5XK46jjWkwygIUzCrD8vRqtIqHoPdXrq4ZJuIwnkV1nBBEWwnjf56L8RWDTk0Crmnu7Ldd8odtLohV8vH676CSRKgUQBEEQBCEHia/dabr8v/0D3QuCIHwjDL1+ejOzJwxDJC1Fafnhz0/J03q+wpKhriLbW8Vx8nL66DgXD392HIDgXArxlNW0YEr+IOHPh9XBEC9k7Wo4iS21Lag91tbtvW5rJU9Q0TA6eljqJj7wJiIeueEUHGV1O4QtMUj2+MTO8e319qUDRbELVLKsDaKlwq8ZMYjpfbocrk7EC39+3atUH0fV16K1NvkhEuq8FsMoXCaq64ynP7tbtYTeBpVI1UcY7/ueNO0BtiwDdq4Czmi+z/UdYNvvVaa8fhAkYqUAgiAIgiDEIfG1O9Mv/++BQPeCIAhfCbrXT2/nmzePweNvVirbnh/lW8OQoa4i2/to9Jz0fngdb9XOx698qsA1EBGc09IQ+1QYHQxxxAtZb+06glXl9YievYhIWgrmTcnD7AnDHD/PG1R8ubxe+/mLDbzJ9uuNxyk4yuN2CFNikOzxiV0DVR9rC5FApU7HjI6yuV7IJj6I3qtUHkcV82POhBxMuGJgl+c4v0RC3fefMAiXsSSi6yxRBeMgoRKp/hCm+74rBzYAb/8IOLTNvzFvesj25TC1gmGBKgUQBEEQRO+iR4uvhmH0ATAawAXTNPfHvD4BwGHTNFvi3p8P4DeX//mSbztKEERoCIuLrrdx/02jUdUYRUl5g5Lt5Q7qh+dLa7SJ6FWNUVQ1RnHLuBys3d3o6p7QmaGuItt76bp90vvB8t1U9Xn8yqcK8P3bxvs2Hg/pfZKFgu5LwuJgYMApALtmV6NnAJYnqJg7uJ/28xcbeBMVEOxwCo6Kuh3CkBgkc3zi10CVxzoenkBlVWMUG6uOCo3jtfbqKJvrhZNj1Q83lUrnkYr5MeGKgfhWnIjsp0io+/4TpHAZT6K5zoDEFIyDhEqk+ovofd+354Q/fx/Y/Bvv96lk9CygeJHtn8LUCsYLSvwgCIIgiN5HwomvhmHMBTD38j8ti8V0wzD+6/J/N5um+Z3L/50LYA+AWgAjYzbzdwAeNgxjPYAaAFF0iLS3AUgD8CcAv9DyBQiCIHopXkGBX959DQpzIvjN+mqpEsSRtBR8dUX3BqMimfnx+5zeNxkl5Q3MP5x1Z6jLZnurcqAdjZ7zDAqwCgBOTM4biMVzipiPpex4IgxIS+EOun/3s0UJE6xUEYBlDSqynj8DwD3T8/Hi5lru7xM7l0SFgNnjszF9dBZTsFPG7RB0YpDo8ckb1A9LPj9ZybFmxStQqaIUcJNHxQDVZXO98HKs6nZTqXQeqZgfKrcr8jme9UtkmgQpXMaTaK4zIDEF46CgEqnBwXrf99WVHITwOn4ucPcLjn8OSysYFijxgyAIgiB6HwknvgK4BsCX414bdfn/gA6h9TtwZz2AqwBci44yw/0BHAfwHoAVAFaYph/dxgiCIHo+PEGB+28ajftvGt2lpOm5i+34sLaVeTwn4ZYnM180OG8AmD0+BzeMHuIqwqjKTpfN9lblQGN1mrEKAEMjqUqOj9d4qrH2kzXovuSuxHGJqAjA8s57lvNnAnhxcy133+j4wJuoELC38RT+496pnu9LdLeD6PFZaFMi3A/RxSlQqaoU8Evv12JS7kDb61dXWWUnDAO4ffJwlNe2oqox6nhdFY/JwtBIKl7ZVo/Kw1EAQNHwDNx5ba70HJO9F8WuDbsOyfcDttsfv0VClvtdYc4ArHi/zrd90kEiuc4sElEwDoreXCI16IoTLKzcUoeHSyockzg+qGnBgmVlap43D2zwV3jNKgL+ZgkwaqbnW8PQCoYFSvwgCIIgiN5Hwv2CME3zRwB+xPjeg7BpwGaa5rsA3lW5XwRBEER3RJ1ysycM69I/UlXAnCUzX2YsE8DaPY2495MjbQM0qrPTZbO9Vf2Y5xGJWJ2PqgJcduOV17Vi/V6xcqNOxB7XhOnZxYFMAFZm3lvnb+m6Kvxq7T7HACOP8GoXeBM9F7XH2pjmfqK7HVQKLH7Me6dytqqc8Cac7yU6yyrb7osJvLb9MIDDna9Z15WVyLKr4SS2HGxBbUtbl8+u33sUW2tbpZ1Roveio9FzeOTVnUrFarfyyyLIHBev+11VY1RIfA3TvSORXGcWiSgYB0GiJw2Jkij9bUurm12FVwsTwOISQVdy0x7gwLvAuSiw9XnRXXUnfwZw2y8+Hic1Aoy6Ccgex7wJP8rrq4ASPwiCIAii90F3cYIgCEILKkuVeYlZPK43t8x8FcF5p+3r6pklk+2t8sc8r0jkdxnV2PGeL61RKr46iXlh6NWpApkA7La6Vul5X1rdjKVvOwuvPDgF3gpzIhiRmY66OHGKBZa5n+huB5UCy7Y69koGotitbapLATut9WE4Zx/UtOBLy8qY36uiXyPvvWhy3kDlfVHdHE2icxiAdA95p/tdIgqXdvCW25+cN1DvDl3G6f7bU467bhI9aUiEROpv+2+v72Z+LjIBPPb6bqx54Ea2DxzYALz7M6C2VHDvGDGSgJv+pUNo5RBb7UiExEdK/CAIgiCI3geJrwRBEIQWVJcqcxKzcgf1s+3x6oZTZr6q4Hz89nX2zJLJ9lb5Yz4MggMrKr+3VxZ90L06VSAagP3Zmkq8U9kkPe95r0unZAyvwNsnCjKFxFeWud8UPcu9XSBcbgcVZf2stVA38eVsDx47raUUsN29JEznjBUV/Rp57kULZxRg+Xs1yoXXB24pRFVjFOW1rbZCKc8cNgCcOHMBs5/a2O1vKt1viVIu043iMVlYOKMAz22qYXr/8vdqMPOq7C7HT2WiEotzsSccd90ketIQL4nU37aqMYq9R6Jcn6k8EvV2JTftATb+HNi5CmIdqTkwkoA7ft1ZVljFGhD2xEdK/CAIgiCI3kfiRQcIgiCI0KOzVFm8mPV8KVuwL574zHzVffpit6+7Z5ZotndhTgT5mendSlGK4KfgIBtUEQl+2KEriz5sQSPRQOrbe5qY3+s070Wuy+jZi/iPe65Dw/EzXMdwwhUZeJkvjwMA0BQ95/mev34k1scyTG4HFWX9RBJcCrL6o6b5NPP7dZWzdeLxP+3BjYVDO+dYmM4ZDyr6NbLei1S7kMcNi6AdwJNrq7r9LVYoZZ3DVs+aSgdxQ6X7LVHKZXqxvZ59jYuda6pLvPI4F3vCcddJbyuRmkj9bV/ZVi/8ucVzbBymfjldLfJndDheR83UUuY5zImPlPhBJDph+51KEAQRdhLzyZggCIIINX6WKlOVma+6T5+1fb96Zolme08dmalEfPUjAKUyQMMb/Hh83iS0nb+k9YdmWPuM+RVItZv3otdlw/EzuK+4gOszosf2pfdrMSl3oKMIU9UYRUUDv/g6KTcjdMEMmbJ+ogkuX7tpFHM5eF3lbN1Yv/doZxlz6zpVkdwRBCr6NbL0OVVxbPIG9cPNRdnIivTF0redRZN4odRrDhcNi2BvYxSmj+63RCiX6Yboc87SdVVc584LXufiioXTsGLhtIQ97rpJpBKpsmJAovW3rTzM53p1/Vz5i8DqRYDZLrlXLBjA534NTLkXQGKVeVZFT0m4IXofYf2dShAEEXZIfCUIgiCU42epMlWZ+arLpFnb97tnFm+294TcDLxczj1MF/woiaU6QMMb/NAd9AlzAMrPH9IrNtfisbkTO//t51oi6og24S7CiK4B11w5WOhzumFN9Ij/O4tD2I6285cCLWfLg3WdLpxRgA8PtgS2HzLY3XtEhA2ne5Ho9TBnQg4mXDGwy/il1c1MQnu8UOo2hx95daen8Bq7XVXut7CXy3RD9Jz+aq13L28ekVvEubjy/ukJe9x1kwglUt3EgBGZ6fhEQSYmXJHheT57XX/bpj3AgXeBwzuA7X+A9hLDADqE16eBKfcASKwyz6pJ9IQbovcR5t+pBEEQYYfEV4IgCEI5fpQq6+zlx1GOMpb4H7Sq3X1N0XN4et0+lNe1Cn3er55Zsj/s/SiJpSpAYxdcDYPrJewBKFVlmlmId5H6XfaQxxEdi5sII3otZ0dSXf9uN58B+CYguIlrTsFwEU6dvYj7igsCKWcrQrvZ0dMyaCFYlNj5qsPlIHo9TLhiIL4Vd6+RLRMaP4fD4H4Lc7lMJ0TPKeulwSJyy567RDzufhDmEqleYkBdSxvqWto62wm4rVmJ1t+2aHhGZ8UFHj6Tvhd49iHg8DYNe+XCkDHAbb/s7O8KJFaZZx0kcsIN0bsI++9UgiCIsEPiK0EQBKEcnaXKVAT27TLzVf9IeGbDfqnP+1XqVUZY86skFm+A5nurKvD56/I6gxhHo+c8BYShkdTAgh+JEIASFSV5iXeRBlH2MH8IX49RCycRRrWAzLsGypQC4w0KegXDRbCOg1/lbFXQbgI76k/g8XmTsLikIujd4cI63rpcDqquB9VCaWl1M763SuxcJaz7TRF+PK94idy9zrnoE2EtkcoqBsTitmYlWn/ba68cxP2Z76W8hPl7/qR+Z9wYPAr41IOdZYYtwpDoEhYo8YMIO4nwO5UgCCLMkPhKEARBKEdXqTIVgX2nzHw/3X0s+PmjRURYU+EKZRF2RAI0tS1t+OXaKqb3xgbjeHuEqmDllrqECEAVj8nColvG4qm1+7SPFfvD3c+yhyrWF7tAvkoBWWQfRUQyEcejSDCchYPHTuP50prO+aC6nK0uympaMCU/nKWj3Sgek6XV5aDqelAptsle+0G538KCn4Kb09qeaM7FRCKMJVJFqxy0m8BimzUrkfrbAh197Xn4XspL+Eryn2Bo2p9ujPwU8Dc/B7LH2f6ZkiUIIjGgRAmCIAh5SHwlCIIgtKC6VJmKwL5XZr5f7j4v/O6ZxepsMADccfVwfPPTY6X2j0fY8UNQCapM0sotdXhY0BWnMgDlJoJXNUaxYnMt3qls4g72yRD7w92PsoeqhMP4QL51bPMG9UM9x/GzWwNk9rHdBBaX7IAJE/OnjnB9r6jjUVfJ3xf+Utv53zrKRuqk8vDJoHeBC2vePfLqTm0uB1UJFarENhXXflDuN53wuN79Sl5zO+eJ5lxMNMJUIlW2yoFpAl97aSt+t+C6znUrEfrbxsKz/n0yaWeH8OqH8jr8WuDWH3UpL2wHJUsQRGJAiRIEQRDy0K8NgiAIohOVQRXVpcpkA/ssmfms+6wTv3tmWXg5G4COkrBHTp7D0eg54XnBK+z4FWjxu0ySFfAXnWYqjoubCF40LAITwN4jUelxRLF+uPtR9lCVcGgF8mXKozutAbL7aAJYXFKBkvIGRwGTx/EYK+b6VfJXR9lIogNr3vnhclCRUKFKbFNx7fek8nqifX55zqkB9n6vsbid80RzLiYqYSiRqiIpL3r2IhYsK8OSuz6+l4S5v208btfCWKMexUk7MQBncAr98MXkdfqE18GjgGv+HkiNAKNucnS6xkPJEgSRGFCiBEEQhDz09EIQBEEIB9u8UFWqTDSw/+VP5mPkkP5cIvL8qSNQ3XQKz22qcX2fAeDW8TkYnd0fz2w4wL1vdvjdMyue4jFZqG9twwc1LY6BUdE+f0DHPHt4VQVMjlKWfgZa/CyTJBvwlz0uXiJ4ZYCiq0XsD3edZQ9VCofFY7I6HM0M89yJhTMKun0Plfvodg3zzMtYMfeqnAFK9o0FJ6d6GEWUouERrN97VHo7X585Gul9krHr0Ams2dWoYM+6Envveb7U/d7nBI/LQUVChQqxTcV1FaT7TTUyfX55zukDtxTiScbWAPFjOJFozkVCHFVBfbse82Hsb2uH3djLUpbg5uQdSDZ8yh41koA7nvJ0udpByRIEkRhQogRBEIQ8tCISBEH0cmSCbSzIlipbuaUOT7xZyT0uAIwc0p+7j2dpdTOWv+cdfDYBvL2nETkZ7iU8WQmiZ1Y8rG5M0TK9j7y6k1mQspyoj82dyLx9FfhRJklFwF9mnujqzWlHkgF8uigbb+9p4v5s/A93XWUPVZW2nlaQiW11rVgsWEraYvl7NRibHemy3qouv213DYvOyw9qWrDF517Zdk71sPXtBoA7r83D1trjUvs0rSATi+cUAeg4R6rF1/h7j18uB9mECpHzXTSsq2tP9rricb+FoVyrGyr6/Frn9Ik396CioXvJ7dhz+l51s3KhNJGci4Q4KoP68feSMPa3tSN2/Svp8wimJB3wp6ywhZEE3PFrIeEVoGQJgkgUKFGCIAhCHhJfCYIgejEqgm2s8JYqK61uxiOv7kRN82mh8QCxIDKP86vdhLCr6earhmLKiMGhCsLyfneeMr0rt9Rxn0sr8MXbL1MGP8okyQb8ZQNQunpzxmO5Q64ZMUhIfHUTXVReLyrOuQHgxJkL0sIrYL/e6piX8dewzLwUnU5fnzkK2ZE0HGw+jRc213p/IAY7p3pY+nYDH1+nMvsULxCpFJgNAE/cNalbD2A/XQ6yCRW8x3ZvYxR/3PJRZ2KDzHXF6n7TVVlENSru/9Z3tRNeJ+UO7CJa6RBKE8m52JuRTURQfd7i7yVh6m/rxqJZY3HVi9dgiBH1V3jNnwHc9C/CwqsFJUsQRPihRAmCIAh5SHwlCILoxegU22SQLdtpwRsMFnF+1beKiYI3Fg7lduXqRHefv1+9vU9ov0qrm3HtiIG+ia9+lEmSDfjLBKD86s0ZSUvB7xZc17leiPxwB4DnS2u0Bz5lz7kV81RZqrndBL63qgLPffl6FOZEtM3L2Gs4iP5Md16bh8KciLJSt2Ho2w10vU5F98lJIFIlMJsAfrthP64dMbjbMRRB5tlANKHCOrYPl7D1zzbjEhtEr6v8zHT8dN4kz++su7KIKlTc/72+a0XDiS7fVZdQmijOxd6IqkQEHVUO7KqehKG/rRvFf7oVZlIU/uiuBjDxLuDG7zD3dfWCkiV6F2FPZiCcoUQJgiAIOUh8JQiC6KXoFttE6Sx9qyBw7vZD3e5HoOrSnqL7FgSi392rTG9pdTMef3MPDp84K7T9U2cvIi+zv9BnRfDjvLRduCT82dsnX4Hy2lZUNUaFAhd+zfHo2YsYGknt/DfPD3fDAI6fuYDZT23s9jcdTjGZbRUNi2BvY1TJehVPbUsbZj+1EZ8oyMRdU3LVD3AZ6xr2uz9TbGa8ylK3XuKLbuwCxbz75CYQqRSYa499PMes6yrRXA7zp47A86UHmZMfYhPJRK99KykintjniqboWbxUVsfV4zyo5wLR+8Jv3qnGr794rXAVFV1CaaI4F3sTqhMRVFc5CCL5SJjSpcDGXwDnTvojvCpyutpByRI9n0Sp/kA4Q4kSBEEQcpD4ShAE0UvRJbbJoqokqlMw2O1HYN6gfkJj5Q7qhwZOZ+bR6LlQBQFlxQ+7IOe2ulZpgcBPQUi3gOA291h5bfuhLv/mDVz4GWCMXStYf7gbAGACex2EFB1OMRGxKT8zHc99+XquPsaifFDTgg8PtqAgq79UGXYnrDnhZ7AkPjNedalbO/Gl7cJFPLPhgNA48UTSUhC1uZasQPHQSGo317aTIGTdP3gEItUCc+x1lTu4H+qOsc+zoF0OVY1Rbtd5bCKZCqG5tLoZT7xZiYqGE1z7YeFnZRE7RO8Lq7cfwvypV0pVUdEplIbdudhb0NHiRHWVA7+Tj4Ro2gO8+g3gULn+sa75EjBsMjDqJmVOVycoWaLnkijVHwhvKFGCIAhCnAR4yiQIgiB0oNJppApVJVGdgsFePwJFS9t+uigbK97n61cYZKDVDtHAU1P0LO5+drO2UrbpfZO1bDce3QKC19wThTdw4WeA8dTZi92CaY/Pm4RV5Q22P9zHDYug8kjUs3yoDqcYb0mtn86bBAC+lHAGOr7zwebTMAwoF3uboucAqCvlaMC9B6xdZryuUrfx4st7+44JC2QWSQbwuwXXYWgktVug+Gj0HJPDQ0VAOTZg/fQ7+7B6+2Gp7bWbwOKSHdz9exfOKAj0XiabSCZbTu8nb+zGc5vEymbH4kdlESdE7wsmgCfe3GPb49UN67sCILGlF6CrxYnKJJQwPY93o/xFYNOTQKv8OsPE9G8Bn/mxP2PFQMkSPQsdSRdEsFCiBEEQhBgkvhIEQfRSVDuNVKCiJKpTyRvWH4EifGpsFrf4qirQquoHkOgPXpayijKsKm/AY3Mn6hsAHWVudZZJ0jn3AH63iF+8vLUev1xb1e31TxRkYsldk9B2/lKXefvIqzuZhR/VTjGRklqifUpFMQEUDOmP2mOnlc6ll96vxaTcgbh76pVKSjkuuCEfVY1R5sx4yxHOi4hT/ZorB0qJr/H3l9jxg3J4HI2ewxs75IRXC5HTvqNeTsyWRTaRTKacnirh1UJ3ZREncgWrfgDgFl4tvvLCh6htaev2OpWh7FnobnESLwbsajiJ1TsO4dzFdubxgiyb7kr5i8DbjwJt/rVECUp4JXoeupIuiOChRAmCIAg+SHwlCILopehyGskg66odnpGGX9x9te0+qipnHM+0gkxs2ud/CWfVPXREXW+6S65aAo4KR54TC6aN0FruStfci4U1cKHK3ciCXWAd+LiM7hPzJuO+4gIA4ehBzVtSK4gecTXNp7HkLmf3sAgmuor3sqUc75mej8KcCFNiiKgjXNSpnh1J4/6MhVtJtSAdHn6sL24E6dgE1CSSiZTTK61uViq8Av6vKSpK4Yvidn+gMpQ9B79anMSKAXdOycWCZWVMySRBl0235Y3vAFtfANrP+zfmFdcCt/xIS19XovcRhmd6giAIgggLJL4SBEH0UlT1OlOJrKv28MmznQ6q2CCpqnLG8SQZwOS8gVgmGIAVDbTqclipcL3poLS6Weu+yQgyXuiae3awBi7CcJ7jRaiw9KDmKakVVI+4tvOXsPL+6Y59lh8uqeB2MLabwOKXd2DhpzpKyK5YOE2olGPeoH6d59IrM17UEe5U3YAF0XP29ZmjsXhOkePfg3J4+Lm+uKHyOuSt5qAqkYy3nN7jb+4RGtcNP9eUpeuq8Ku1+4TczrrpbWUoe3IJxyBanBSPycITd4m52QPltX/qcLvqviqHjgOu+z/AuSiQGvGlryuhnjCvG2F5picIgiCIMEDiK0EQRC9GtteZalQEQOyERxXljONJMjp63S1/r0Y4TCISaNXpsFLhetPBqbMXte5bGEtpF48egtL9x4TG8wpchOU8x4pQYetBzVJSK6iArfWd7fbR+vfDqyq4Xen1x8/g0dW7AXzsnq9vbeMSc+224XScRBybbu5TFkQ/d+e1uY5/C9LhoePeJoKK65C1moNdwFllIhnLtV/VGMVOwXK7bvixppRWN+Ox13ej8khU+1gy9IYylKormISRoFqciLjZA2XJaOCMT+v5Z58gh2sCkwjrRtie6QmCIAgiSEh8JQiC6MXI9DrTQWFOBAVZ/VHTfFpqO/HCo+iPubxB/VB//Ey3162gjWy5R5HjqdthZQWswhSctYJwXsG0vMH9UN/a/Xx5EeZS2rrG8zqW44ZFYALa54AlQokGWpui5xTvETt+lnCOxetYzZ86Au9VN2P1dvE+oLFJLCwOIq9tXDNiUBfBLHdQP6Hj9tjciVKCpY6KD0E6PMISqPzfrfWYkj9YeC1lqeawYFkZRjo8HxQNi8AAm19MRSKZDtFbpLIIr/NJtMy3F5Ny5XopO9GTy1AG1SNaJSzzL8gWJ7xu9kAoXQqs/SG0u10tpn+LhNcEJlHWjaCSLgiCIAgijNDdjSAIopcTpuzw0upmHJQUXi1ihUfRH3NWCU67oI1suUfRQKsfDqviMVnI6NeHaxydxM69oZFUzJk4DEXDI2g8cRY5GWkYmdUf6X2T8d1VFdzbDmsp7YgPgQuWwGT839L7Jjv2G83PTHfs4+dGaXWz8Pry149ahT4ngt1xmlWUzX1NFgxJR80x/uNkkd432fFvKns4WkksKxZOEy5D3G4CD5XskN4XCxWCJc85YxHqgnR4hCVQWdfSJhzwZa3mYAKOiVmVR6IwAE8BVlUimWrRm1cQFnE+iZb59sJ6RtRVyr4nlqEMske0CnjmXxhanLC42QPhla8D2//g33jTvwV85sf+jUcoJZHWjSCTLgiCIAgibITjFztBEAQRKGHJDl+6Tm3/MUt4lPkR6BS0kXG+iDpvZB1WrOc3LH0EgY+DcF7BvpNnLgj1jgxrKe15U/KwZlejL+O5BSbt/jZ/6gjbufTnnUfwy7VV3OOfOnsRhTkRIfdURcNJ7c4olYImAAzNSENqn2RhV/F3V1XAgNFF5KpqjOI37+zD6u2Hla6hVhLLyvund7lHLN9UY1sVQDeyotfKLXVYsqaS6b2sQl2QDo8wBSpFA76yFSQsTACGARTlRGyvLZZEsth1re3CJQAm0vukdLtfqhS9eQVhUeeTquMci3UP1VnKXrXQHfRzLhBcj2gViMy/sLU4CZSmPcCBd4F9bwH71/kzZko/4O//hxyvCU4irRthSLogCIIgiLBA4itBEATRSZDZ4boEv9LqZtxXXKD8R6BoQNCAuPNGdMxdh07i7mc3M7kUSqub8T0BB6kODABXDk7Ht/67HK/vOOzYw1J03gwf2E985xgRCUCMyuqP2ROGhTpwYbdWiCYHWELGNVeKla7U6YzSUabzg5oWJpeeE7EiFwClwrAdse556zhbfV39Zl+TeBlsXuff4jlFTC7OIB0eQZW+doI34Kv6vm+awMB+ffDWgzdyCWysCRaq++nxVhYRdT7peL6KF43nTx0BAFhcovb5QZXQLdsnUZVoG2SPaFlE51/YWpwEQvmLwKYngdYa/8bMyAMWvAxkj/NvTEILibhuUNIFQRAEQXRA4itBEAQRCnT0UAM+FixV/wgUDQguuGGEcB8e0TFLttY7ijyxLgUTphbniigmgJfL67Vtv+H4GV96Iy2aNRYLlpc5isfx1DSfRml1c8IFLmRFqOxImtDndfW91FWmE7js0kOHU491XsTSbgI/eHUnDh477cv1Gitw61qrWVi94zDmT/24TDWPIMLr/Hunsgn33zTa831BOzx41gk3jMvZALLTiSfgq2MuWaWx7ysuYHo/T4JF7P1SRPSePioTnxydJSzeiTqfVB9nJ9G47fwlpeMAapIUZPokyoq28QTZI1oWGeddmFqc+Er5i8DbjwJtPt83r/kSMPe3/o5JaCMR1w1KuiAIgiCIDkh8JQiCsCEMZcl6G7oElF2HTnSWHlb5I1D0R+I900dyvb97GUR+vGJl7SawuGQHICgEJTJ+9EYqHpOFkUP6O/YrjMfEx6VeEylwIStCiSYXNEXPCn3OCx1lOmMx0VEidWC/Pty9VAHggKL+2CzErs+61moWzMsBfcDZ8WsniOh2jQSZKKGi5KslfNS3tilJOGAN+OpMnGAVf3m/r3XPWDynCB8ebOH6bFlNC775aTGBSXQO//D/7USdQC9uO+ZMyME/z77K8diqPp8qkhRk+iTKiLZOBNkjWgYVa2hYWpz4woENwBvfBo5V+ztuagZw43eA4kX+jktoJVHXjV6bdEEQBEEQMZD4ShAEEYPqDPfehkxARWUPtVjW7GrEml2NnedvxcJpSn4E6nY7rdxSh39fv19Z0NQLs/P/+Y8BYGJuBioaTgYyPmupTNH5XdUYZRZeLaygpR+BC5WBUBkRSvQ7bK/nL1XshV99jyuPRPHWgzfilW0NeGbDfu3jiRK7Putaq1kpq2nBgmVlTG5+SxBR4Rpxu06Cdnh4rRN5g/vh5quGdib/uF3vbtthhTXgq2susY4vmmDRbnY4ox+fN4mrzK5MHz7ROfzC5lqhz9nhJrwC6nvhqkhSkHELi4q2bgTZI1oGlc67IFucaKfyDeDdJcDh7f6OO/xq4KbFQNFt/o5L+EKirhtAL0u6IAiCIAgbgr8bEwRBhAQdGe69BRWitW5RO/b8rbx/upIfgTxCk8EYSCytbsYjr+7kFusSGRMITHi1cHO5yc5v2aClrsCF2/cakZmO/3vz6M4+fqzIiFCFORFMzM3ATs65sKP+hPK+Vn6W1i2tbkZ6n2TfxhMh9jyFIQGJxc0fK4jIuEZYr/+gHR5DI6mYM3EYioZF0HjyLHIGpmHkkP6264TbtRK73jz51l6s2dXIvS+sAV9dx4JlfNkEi7KaFiycwVbaOP5zIuvVwWPBPhOwJI+pOp+qkhRk3JoyJXbdCLJHtAyJ6rzzjdKlwMZfAOd8fpYdMga47ZfAqJn+jkv4SqKuG7H06KQLgiAIgnCBxFeCIAjIlSXr7agSrUWcpLzEnz/ZH4GW0PRwSYW3adQEGlrPuL5l5ZY6PLyqoteV/g0Ldg4NFfNbVdBSZeDC63vVtbRhcUkFfvfuAfx47kSu9U5GhLr2ysHc4iugvq+VnwFjS0wPK3mD+3U5ttvqWgPcG3ZiBRHR47uvKYqn3q5ivv6DcHh4icOi60ZhTgT/PPsqIfGVdb3Qdd9nGV9FgsUqwZ7kIutV4wk95dVZYE0eU3E+i4ZF8IPbxyt5xhY9x69sq9dWpjzoHtGiJLLzTjv//UVg75/8HTM9C7jlh8CUe/0dlwiERF03dEAOWoIgCCLR6AVPwwRBEN7oynDv6agWrXmcpKKoPn95g9OZ3mfC/RhYx5KE1+CIF9xUze+wBS15+hzWNJ/GguVlWMLp+LcTodouXARgIL1PMqoaoxgaSe0WMMmOpHJ+mw7cxFKRQI1of2URBqSlIL1veJ2v9a1nUFrd3KUUZ6JgCSKi6/3q7Ye5XbaAfw4P3RU7/Aj4qr7vs46vIsHiL/uPCX1OZOycjDShsVSwhMOFKns+B/bro+z5TPQcVx6OCn2OVVQPske0KD3BeaeUpj3AluVAxf8CZ4/7N66RDNzxKxJdNRFmYS8R1w2VUGsogiAIIlEh8ZUgiF6PTFmysPwgCwrVojVryVJZVJ6/pev2MbdKdTsGor3nCHXEi52883txyQ4snFHQLVgTtqAl71wzJRz/hTkRHI2eYw6YqBSqRQI1bp/RxcmzF/Bvq3f7Np4I1rqViOtUaXUz7isu4BYRI2kpiDKKN0EkZflVsUN3wFflfZ9nfBXJLScFxT2RsUdm9RcaS5bvfraIO/FG5nyqfD7z23XJKvYG3SNaBHLeXebABuDdnwG1pf6PPfBK4MGdnf8Ms1CYaCSCsJeI64YqqDUUQRAEkcgkBb0DBEEQQSPTj7E3IyNauzF/6gisWDgN0woyZXbPExXnT9UxkO09R3xM0bAIJucNFPqs5dR8vrQGP/x/O7nPSX3rGTy6ejdmP7URdz+7uXOOWUFLHgqy+msJoonONUtc4mXlljrcs7zMcUwrYPLHLR8BUCdU847L8hkdFAxJx6/WsidwBEVZTQve2nUkIdcpSxCZVZTN/BnDALPwasFyf1OJSPKTCFbAN8lwf59MwNfrvl+Q1R+G4vGDDEyLjO33/hoAfnbXZNx/02juz1rnMz+TrTJIPKqer0WPWdHwDKHP8Yi9XnN+WkEmViycFiohYdGssZ7rgEVPdN5hwxLgxbkBCK9JwK3/1im8llY34+5nN2P2Uxvx6Ord+OXaKttnT4INkefFoEjEdUMW3kQzmv8EQRBE2CDnK0EQvR5V/Rh7GzKitZeo5NQ3ryl6Fs9sOCA0bjwqzp+qY5CIPxQn5w3Endfm4uCx03jhL7VB7w6AjmCf1Svu9qc3cfUOHTcsgkde5RdcnYjPwl40aywWLCtjFtkOHjvdWepVJTJzjdeRJOrMk3XX8Iy7uGQHKhpO4MyFSyjZWu+7CNradoF7zCQDyB/SHzXNp7XskxOi/S3dSDKAhTMKsPy9Gm2O2gFpKVi5pQ5L1lQy79Ptk4fjte2HucdS3XvYCdUVO7wcVDJ9nFmoaoyiqjGKGWOyMCV/MAAT6X1SuuxLaXWz0vH96DNvh6gbUMf+DuzXByfOXOj2uuz5BDqe4z5/XR5+ubaK+7Oqnq9F3Zp3XpuLZzbs5x6P93gF0SNahl7rvDuwAfjz94DGXf6Om5QCTPo74M7fdb5EDkC1+FVBQiWJtm7IQq2hCIIgiESHxFeCIHo9YevHmCj4IVrH9817WtC5Y4eK86fqGIRByDcu/z+WnrNJBrB4ThGKx2Th+dIa3bvGjFXyFwC++9lxzKUyDQCVR6LKhbf4YM3ILHbBzNQUQJCda6ziUlVjFN/jKDsZGzDhKXNqAJg3JbfLazyBGhPAiveDSx44biN8ePH4vEnIG5yuvT92PLxOUC8iaSn43YLrUDwmC2OyB2grN5/eN5lr24vnFOH8xXYA/OIrz/UlEzhVmfjDWmpRR8CXZXxr2zrG96PPfCyybkDV+/vALWM7eznrCOCH4fma95h9uijb9xK7fvWIVoHuRIxQ0bQH2PhzYOcqwK/UrGGTgUEjgWu+ABTd1uVPiSgUhp1EFvYSad0QhVpDEQRBED2B3q0cEARBIHz9GBOFIIJqKgNyKs6fqmMQtJBvuRRMmNyOhjAIxxY76k90/jerQ8MwAJj6wmpWsGZoJJXbqagjgCA717zOt0zPVOv78vQMNAEsLqlASXkDFs0ai6GR1IQsjcvDwWOnMX/qCF/6Y8cSUbxORc9exNBIKgDvgD5P/9VYBvbrg5/+qZLrGL1T2YQ5E4dxjwUAuw6dwNPr9rmKWCp6y4muu8s21aAwJ4LiMVlCDiqVwqeog0tlwNmvPvOAGjeg6v21zp+uIHUYnq95j9mSNZUYnN5Xe6/jRKbHO+8C6etqAJ97Gphyj+M7ElkoDCMk7IUfnVW2CIIgCMIvSHwlCKLX43eGe08hiKCaqiCCqvOn6hgExbuN4AABAABJREFUGRzJz0zHN24e3Rnc5nU0BC0cxxIfFGFxaBw/cwF7j+jt01hW04JXtjUIfVY2gBAfHM0d1E94W4D7+fYSU1iwvq/XuYvHEmpun3yF+OAJQuXhjvk6f+oIAB3isx/Mm5KHNbsalW4zdn7bBfSrmqJ4fcdhYdetXUlVL8pqWrBwRoHQeGt2NXY5RvFiqqqSkaLrbsPxM7hneRlzqWfLQdXadh7rKpukBONYwuTg4l1r3Mgb3A/1rWe6va7SDahqf/14jg3L8/X8qSNwvO0CHn/Tu/S4NecWzyliKu/e40rsctAjnXcblgAbHodvTlcAGDIGuO2XwKiZjm8hoVA9JOyFH2oNRRAEQfQEwhMxJQiCCBDKcOcniKCaip5nKs+fqmMQVO85AKhtaeviHPRyNFQ1RvF8aY0yMU818UERt+8DALOf2ujLflUeZu8/G4toAMHNXSfqIgScEwVYxRQvYr+vde5WbqnDwyUVnqHQdhNYvf2Q3A4kGG3nLwl9LjuSiqboOeb3TyvIRP9U9T8b7Oa3FdAvrW7GU29XMZVCV03D8TNK1uRYMTV3cD9lgqOM0NNuAs9tYi8X327CVbSK/Y7XjBjE5IQLm4Mr/j6xseoo1u89yr0dq/S9bjdg/P7uajiJjfuOMl/TBvx7jg3L8/W6yibm93rNeYuCrP748dyJvVJ47XGUvwi88xPg1BH/xkzPAm75ITDlXs+3klCoHhL2wk8YStcTBEEQhCx0VyIIggB7WbIkA3jglkJUNUZRXtvas8psCRBEUE2m5xmLQ4G3jJqqY+B377l44h1X8Y6G0upmPPLqTlsxIi0lCWcvtvu5u444BUXsHBph6lfrhEgAwctdJyq8uiVP8Igpbth935LyBmYPSkCXj68UDc/o/G/RIOCNhUOxqryea91aqrDntoXb/FY1p0Q4dfaisjXZElMLcyLKBMcgE3bsaDeBh0p22P4t3hkr4+ACoFXYtO4TxWOyhMRX3WV84ynMiWBbXStWbWO7loOA5/lal4NUZM6xUHuMr5UAEULKXwTefhRoExM3hRg8CvjUg0yiqwUJheohYS/8hKF0PUEQBEHIQk8OBEEQl/Eq5VY0rCOY9uTaqm5/Eym71xMIIqgm2vPMq+yfaB8+3mMwNJLaxTlqBWv97D3nhJPjauWWOjy8qsLRgRYW4RXgC4r4GZQqGh4RDubzoMqBGo9b4oDKwHb899UVNE9k7rw2t/O/RYOAE67IwNSRfOuWjvPgNL+DPu/W2qxqTW43gUrO8uZeJSODTthhJT6xR9TB9Y8vfIi6lrZur6t8/opNvhqRmW47nhNBtKMQXe9N+NMPMvZ4fmlaPv76USsqGrpXgVBZltkO0TnnBfXVTGAObADe+DZwrNq/MXOvA/7234HscdwfJaFQPSTshZ+wlK4nCIIgCBnoaYwgCCIGpxKlJ89ewNK3nV04rH3aeiIsfTVVB9W8xpycNxBX5w1EdiSNyR0j24eP5Rh8uigbL5fX27qDrOCxyt5zosQHE0urm5lKvqrEAHDLuBys3cPfX5JnnvkZlLrz2jxsrT2uPYDA6xZMTUnCOQ/x3Ct5QlVg2+776gqaJyo5kdRuZbVFsNZE1rVbh0vcbX4Hfd6t4xr0muxWMjIMCTusxCb2iCa9OAmhKp6/3JKvWIhPTuGtoCGKjDtcZz9It+M5KXcgrrmS/flMBToTraivZgLyxneALc/5N16fdGDmw0DxIuFNkFCoHhL2EoOwlK4nCIIgCFFIfCUIgrAhtmxcaXUz00M/S5+2nopXn9Awj8nqHPE6v277s62ulUvctbazYnMt3qlsQsPxM8zfRwWxwcTHXt/teylXE8DJsxeEgiIAuvWkbTh+xnZ++HWdWsEa3QEEEbfguYvtePCWsVi1rQG1x7qLGyzJEyoC207fl0rmdeVTY4d2+bds8JB1HVV9Hrzmd5DnPT64aneMdh06gTW7+JNDePE6DpY4vLhkB+pb/b1P8GIl9syZOEzLtr2ev5zmuFfylRexySmiFTREUOEO19EP0ut4VjScwK5DJ6STFXme/XQnWlFfzQShdCmw4QngArujXZpPfBX4m59Lb0a01Pwjr+7slZWZWCFhL/yEoXQ9QRAEQchA4itBEIQHPM6C3l6CzM9eZ6rGVH1+7Xqlioi72+pa8fuy2sAcTU++tRfzpuRxl8pURVlNC5bcNQkfHmxhOgaGARw/cwGzn9ro+d7YILjuvomxwRrdAQRRt2BGvz54919uFk5kkA1su31fKpnXlQm5Gd1eUxE89FpHVZ4Hlvkd1Hl3C67GHqOn1+3zRXxlOQ7FY7KwcEYBHl29W/v+yFJW04KFMwq0bNvp/uwmiBYNi2DvkahwglFscopsBQ1eVLjDVSc5qEpm8xqDV+DW/UxOSULqUZrM2bQHePUbwKFytTvpRv4M4KZ/AUbNVLZJkVLzvbkyEwsk7CUGQVTZIgiCIAhVUESLIAjCBRFnAZUgSxz8OL8i4i6AwEtJrtnV6Iu44Ebb+UtMQREDAExgL6NQHBuMWjRrLBYsK9Pi7jXQPVijM4AgGgC2PieayCAT7PD6vmEKpEzKzbDtV+gnKvpOixxTVeeBdX6n901WMh4PBsfx8UscZj3uYbpOvGg4fkZb0kv8/dlLEOVNLsofko7PT8nrJgb5ITrGo0LwUz2PdScrigrcoq5BVihJSB1K3eMblgBlvwPO+FQyvn82MP5zwNR/FOrr6oVoqfneXJmJBRL2EoMgqmwRBEEQhArolwJBEIQLos4CKkGWGOg+v6Li7uNv7gl9Dz8/OHX2Iu4rLnANiowbFkGlgHPJCkatWDgNt47PwVu7NQjNhv3LugIIogFg2cCxSGA7PzMdz335es/vW5gTwaTcgahoOCG1jzLcc8MI3DN9JEqrm1HREJy70K3XmO7gocg5npSbgXlT8mznt9fcLylvENpPGRZMG8HsDPIjCBu5fF3GllF3WiN0i0sq2XXopJCDixXr/swqiPJQe6wNn5k4rNs5CKJCigrBT+U81p3MJitwL5o1FguWl8HUMOdIlFGDMvd46VLgnR8Dl87r2dFuGMDN3wNuekj7SKJ9yHt7ZSYvSNhLHIKoskUQBEEQMpD4ShAE4YKsk4wIN7rPr6i4uzNgd11YsILLbkGRR17dKexaje1BqEN8NT3cBqoDCKJBNRXBON7Stz+dN4n5u19zZbDia1XjKRTmRPDnnUcC2weWXmO6g4e85/jhz47jKgFruZqGRlIDERHPXGhnfq8fYmf07EXbMupO7i+dgqZKttS04Bd/d7WQg4sF6/7MI4jyEJ98FVSFFNl12y2ZQwTdyWy8Avfil3dg4acKUDwmC0ej57B03T4twqvq49hbUeYeX3YrUP+Bnp20Q0N5YS+Kx2RhaCSVqc1GLFSZyRsS9giCIAiCUA2JrwRBEC4E5SQj/EH3+SURXo74wFp8UEQk6B2Pzh6EgL9uAxFBSFXgWGfp2+xImvT+yWAFLIPsQ8pzzHQFD2XOcVVjFE+/sw+vbz/smCxhuZpunzxc7Y4zsuXyeWYVroMSO53cX6znxwC0lFlnpbalDVWNUU8HV/6QdNQea+Pe/oC0FCX3Bifi7+uvbBNzactWSJFJAGBJ5uBFZzKbyPmsP36Guw/ytIJMfLooG0vWVEr10Cb4kXaPv/Ed4MP/BMxLenbQjs89DUy517/xYqDKTARBEARBEIkBqQMEQRAuBOkkI/Sj+/ySCC8Oiyi4YvNBJWPp7EEI+Os24BGEDKgNHOsqfRuG66i0ujmQdT1svca8zvGk3Axcc+VgNJ48i+dLa5DeNxkl5Q3M11a7Cazeflj1bjNR29LG5TQV7b+nAif3F8s1WJgzACver/Nzd7thCQBubm0A3M4uoOO8iAoTLFjrkZuLmwUVyVkiCQAyvZ/dEF2ndx064Xl/1Hk+Laz7YfGYLAxK78N0XS+cURCatTmRkXKP//ke4MB6TXvmhHFZeL3H53E/hiozEQRBEARBJAbBR7MIgiBCTJBOMsIdFaU1C3MiKMjqj5rm08yf4Tm/QYo1QyOpXY7PybMXsPRtPWUYVWOg49y4BWRXbqnDS4oEhFNnL2p3selwG8RfA7mD+qHh+BncMi4Ha3c3errbTABPvFmJa64ciOxImpIStTpK34YhuH3q7EXf+mrmDeqHm4uG4p7pI0N5L7E7x03Rs/jrRydQ0XACFZJl08O2RLn1GRTtv6cCJ1e91zVY1RgNXHyNFwCc3Nqiz186S4QXj8ny7E3JgqqerTwJADqTOUS3uWZXI9bsanRMcgD8EYxMfHw9zZ86AtVNp/DcphrXzyx/rwZjsyPMvaKd6O19JkXE9a8kr8aYZxYAYC8Xr4QAygzbQZWZCIIgCIIgEgN6+iIIgvCAt88dlSDTC0vPQNYgYGl1Mw4eYxdeeZ2CImLN5LyB2FHP39/y6zNH485rc7sE7OKDd9fnZwYiEvBiAljxfi1WvF9re06t3mCqRBor2KnTxaYyeCzruIqlQyzrOt94ryM7VJa+9Uv0dMMKWPpRarb++Bn8vqwOk3IHhToAb53jlVvq8NTbVQmR2CGKW5/BWLHzlW31eGbDAd/2y81V73QNhuF6Kq9rxfOlNZ4ik+jzly6BYVpBJo5Gzym5T6gSQL0SADqSObJxz/R8reuJ7LxyS3LwSzCyrqej0XNY/p678Aow9B/1QOXzbCLD83w01qjHr1N+jaKkehga96kL/TKBifOAqf8IZI/za1RXqDITQRAEQRBEYpAU9A4QBEGEHUuUSfL4la+rlBvxMSu31OGe5WWOwT0rePfHLR8xbW/pun0wOQKoBVn9uc/volljPeeORZIBLJ5ThE8UZHKNMa0gE4vnFHkGVovHZGHl/dPx1oM3Ys6EHK4xvDAA/Oyuyfj6zFFKt2t3Tnl6g7FgndP5U0dgxcJpmOZw/McNiyA7kio0hqrgsdc1oALe68gPeK4jHVhzhPV+IIsV1Pej3KYMViJETxZeLSynqROFOREsnjOOe/2WRWSOBH09rd97FI+u3o3ZT23E3c9udvwOos9fOp7DLHFXxf1HdYWU2Hv7D+8Yj2/fWogf3jEebz14I957+NN4bO5E38rey8wrp3XPz+dqSxDl7T/Ki+rn2USG5fnok0k78WbfxVib+hDGJdfD8GP9Ss/qKC+8uAa47ZehEV6Bj5MdeKDKTARBEARBEP5D4itBEAQDXqLMtIJMrFg4Tbr0GOEMa5CfVbQQ6TF1oPk0qhqjXJ8RCR7zCra8buvCnAj+efZVXJ9xY1pBJl76x475n95HvUMl9pyKnDc34oNRbkHsNx+4ES/94zShcVQEj/0UusIm/vkletoRP0e87geqEA3q+4nqRIhYAtQGHbGccW74LWyKuOqDvJ7i8RKZRJ6/RIQJNwx03J+HRlKl7z86K6QU5kRwX3EBvjVrLO4rLvBdaFExr+zWPdXn042DzaeF+o++tesIni+twdPr9uH50hrXdUL182yi4/V89KOU5/H7Pj/FuCSfhOh+mR2i60P7gSn3+jOmALp/KxAEQRAEQRDyUNlhgiAIRnT0MiTYEXEiuAV0RINZIr07vcoCxvdhYy2BK+O2Fu1n/Njcia7zX1d5QOuczpk4TNk23YJRdiU7LUcML6rcBjqFLjtYriM/CaK/ptMcib8fbKw6ivV7jyof362sbNCoToSI546rh+P1HYdD56r1ugfoLmEej+ia63U9TcrNQN+UJGytPS65h954lW8Vef6aVZStbH4uuGEE7p56JZ4v9S5F60ZvqJCiYp22W/f8KPkOAI0nzwp97qsrtnZ7zal0sOrn2UTH7nl0rFGPH6Y8jxuSKpFi+HgTuOo24It/8G88Cfz4rUAQBGFBMTCCIAgxSHwlCILgRGUvQ4INkSC/l2gh2oNT9HNd+wI2oPLwSQBA0fAI7rw2r9t+8gq2Ioj003Ob/6XVzVhV3iC8P16U1bSgaLiaa89yMrEev5Vb6oTEFFVuA91ClxNhE/9ir6Mn39qLNbsatY3FErC0rodTZy9qEV8BsYQPP9Dtxvrmp8di/tQRoetTzXIP8Fq/U1OScO5iu5L9kbkHuImaR6PnsGBZmZJ9ZIFFZGJ9/lq5pQ5L1lQq27d7po8EINe7W8U9O1FQsU7Hr3t+JTXkDExTti3L1f3ALYWIpKXg1NmLaLtwUfnzbE/Aeh69wdiJJSnP4sqkY/7uQGoGcON3gOJF/o4riR+/FQiC6N1Qf3KCIAg5SHwlCIIgQo8Ol6qoW0jG2Wn342X93qN4ZsMB2x8vut3WKrPmRcVJXhpPiLlSusFRFlG03K9Kt0GQZQfDKP5ZpbN1ia+TcjPw8GfHMZ87XY5vQE7wsYhfQ3IH9UPD8TOXxYBLAEyk90nhWl9U7JcTllu8MCfim9jOCuu5ttbvlVvq8O/r96Oupa3zb5bwmp+ZjqkjMzEhNwO5g/rZuue8ONh8Wvr+YCdqPvLqTvhtOlYhMqkuzx5buUD0Ov/6zFFYPCc8/SL9ojAngglXDBS6bu3WF93VD6YVZGLkkP5Kt9luAk+urZLeThjvwyopHpOF93OWYGjrdn/6uVoMvxq4aTFQdJvwJoJ2g1FlJoIgdOH1+95KMnpi3mRqv0UQBOEAia8EQRBE6NHhUhUVxEQ/J/PjRafbWkXWvJ+9SHMy1LhSTI8yl7GIlPtV7TbQKXSFeWw3REpnszA5byBe++YMrs/ozPhWnfDBAksmuy7B2c4tLiPiqKYpehZPr9vHFNz2WvdrW9rwUWsbPlEwGQ3HzwjtD0+5U1ZknPaT8wbizmtzhUtxy4pMKsuzx89F0eN557V5anYoAVGd6BYvNC3fVIN6wWsnFutcD42kSm9LB2G9DyuhdCnw9qPINi/51+w7bRBw9wvAqJnCmwibG4wqMxEEoRLe/uQsv6kJgiB6IyS+EgRBEKFHh0tVtOepU2DDLeM87D9eZLPm/exFOjKrvzLBjaXMpagI8djciUJBMKdzoNNZ6UWQY3uhug9gkgEsnlPE/TldQjCgL+HDDZZMdh1rlJtbXHQeGoBSB+czGw50+bdTkJ133Z83RZ1AJ+tEEHXaG+i4forHZAmX4j547DSeL60Rcm+pLM9uV5pe9XNDb0BXopslNBXmRLBgWZnUNR6/7uhay2UI833YDc/nyle+Dmz3ucfq6E8D97witQlygxEE0dOh/uQEQRBqSMyneIIgCCLUqC59pSt4J9LzNB6WzPdE+fHCmzXf0b+23tcgpTWXVAluZTUteGvXkc4SrHbCuQi87i2veXTXlFyh/VBB7qB+gY3tBWvpbMtM4zZlZMtEL5o1FguWl8FUqPSJCjcq3OheySCqBedRWf3x2NyJjsdf9Lw8cOtYLH1bX4KIU5Cdd91fX9mkdL9kknlEXXa3js/pHEtULHrhL7XdXmN1kaksz/7EXZNsRRMVzw29Cd2CdfGYLIzM6o+a5tNC+2dXpUJ1Uo8KEi2g7fZMMy/vJL45sgGjWkuB/e/4t1OR4cCdv5NyuwLhT6gkgoNKPxM9BZFktt7Qn5wgCEIEEl8JgiAIZegqwaUreCfb85Ql811EjAn7jxfRUqayxPaBZDlvrLiV7dRR8jqepeuq8Ku1+xyFwQ9qWvDhwRYUSASYZRAth+oXrKWzAbi+Z96UXJw+f5G5pGw89a1tSi2WMsKNKje6VzKISpHi4DH3uS16H1g0qxDX52c6nvuCrP442Hxa6tS1m8Diko+D7CJBq2Onz0vsgfN+iSTziAqn00cP6fxvlWIDq4tMVWnWaQWZmD91hO3fVPZK7y3oFKyrGqNC98WvzxyFO6/Ns13jWc+xXySae9rp2fiTSTvx05TlGNncCPjZxj45FbjtF8CUe5VsLlESKgn/cPtdNCl3IK65ciCyI2kkyBIJg1/JxwRBEL0BEl8JgiAIJeguwaUreCfa85Q1813UBRf/4yUs2dQypUxliD+nXudNFmu+3j55uNDnWcSL0upmPPb6blQeiXq+t90EappPwzDE55QoidBrjrV0tt170vsmo6S8AYtLKrptl8dx991VFcwCntd5lBFuVJZeBdyTQVSKFF5B6qrGKK7KiWBLTQvTcY5dM9zmx7a6Vjxsc+55MQF89cUPMW9KLupbw5OwIJLMo6LahGpnNIuLTEVpVpbnBxW90nsTOgVr0QBxdiTN9ZrQ/YzBSqK5p52ejf+nz6OYlrQXhl89XS3G/y1w94vKNkduMCIer99FFQ0nUNFwostrQfQEJgge/Eg+JgiC6C2Q+EoQBEFI40cJLp3BO5Gep7r7nFo/XnS5iUVQUcpUBKdzGn/eNlYdFeox6ES7CazecVjos17nRFTEHtK/L1pOn/f1HCRSrzmW0tmx71GVNMK7HlyVE8HAfn20CDcqS6/GbtPpuKoUKeyC1CJOe6c1I35+8IrmXpw+fwkr3q9TtDV1WHOC9f4mKpz+te54l22qLt/qJdDL3gt5nh9ke6X3NnQJ1joDxE7nOHdQP9uKGapJRPd0/L3w28n/g28kv4YkAz4KrwZw1WeBL/638i2TG4yIRfR3EfUEJsKO6G+/RPrNSBAE4Re0MhIEQRDS+FWCS7fbhLXnqWpnmR0D0lK0u4l50Sk45w3uZ+sWYzmn1nkrHpOlVHwFOtyJkbQURDkCvJNyM1znkYyI3XzqPL7yqQLsqD/hmxsnkQK/PKhKGhFZDyqPRPHWgzd27odK4UZH1rnXNu1Eil2HTmDNrkbusWKFws37j2Ht7kYucZTnPqA7iSYsLN9Ug0dX7+72ulsCz6yibO55HX+d6Cjf6uYik3Hbij4/8PZK783oEKz9CBDbnWOVrm47Yudjogj8sffCZX2WYFbSdv+drp/4KvA3P9e2eXKDEbHIPENQT2AizKiogEIQBEF0QOIrQRAEIYXfJbjC4DbR4SyLJ71vsnY3MQ+6BeeFMwpQPCaL65zazQEdAdHo2YswwN7Os6LhJO5+drOjqCEr+Ly1qxHvPnSzrRun4fgZbN5/DG/t5he97Ei0XnM8qEoaEV0PvvLCh/jpvEm4r7hA6PNO6Mg6Z91mrEjx9Lp9QuKrk1DoRXYkFd+eXejYqzMeP5JowkK9Q99muwQea115cXMt9zh214mO8q1uLjIet60BYMEN+fjU2Cw0HD+D8tpWvFd9FICB9D7JoRa6EhmVgnVQAWLVru6vzxzdbc6VVjfj7mc3h6LyCQul1c0Ya9Tjf/v8CAONNn+F1+yJwJyfAKNmah2G3GCEhYpnCOoJTIQVkWS2nvybkSAIQgZ6CiQIgiCkCKoEV5BuE90Z7NMKMlFS3uCLm5gV3YKzFWxkOaduZUiLhkW4hFJW7rh6OF7fcZj5nDi5klUEa2pb2jqTF+KPl9VHVgWJ1muOB5VJI6LrQW1Lmxbnuo51QGSbosFmJ6HQi6boOXx3VQUMGEzH048kmkTASuBpbTuPdZVN0uuT3XXilDR1sPk0XhAQed2uOZ4WBQtnFGB7/QmseN99H8IodPVkeJLrggoQq3R1TyvIxOI5RV1eC1vlE08ObMDcTd/Efakf+Ttu9njg8/8JZI/zZThygxEWqp4hqCcwEVZ4kox68m9GgiAIWUh8JQiCIKTojSW4dGawGwDmTcnF4pIKrs/p/vGu83w5BULtArDb6lpdA5KVR6LQYbYYmx3BioUjuNxbdq5kVcEap+QFVWVUE7HXHA8qk0Zk1gMdznWZ0qt2iAoVQcyddhNYXLIDJkxPB2wi34NU024Cj79ZqWx7TutTfMLI86U1Qtv3uuZYWhRMzhuI5e/VMK2XoRO6eiiiPe6DChCrcHXb7Y+qkvi+8efvA5t/g8F+jpnUB5j1A6B4kZ+jkhuM6ETlMwT1BCbCCE8yW0/+zUgQBCELia8EQRCEFL2xBJfOHxd3XD0cbecvCX1W5Mc7q8NE1/lyCjw6BWBZ0NHCcdehE/jMxGFYef903PH0e6hoOMH0uXhXsqpgjd12VJVRle2dbEfYetapTBqRPU46nOuqSmIa6Ag2P71uH/d5Uy0Cs2ICWFxSgZLyBle3YiLfg8IO6/Wl00Xm1qLgaPQc9/URGqFLEWFbk2WcnkEGiJ3m2cmzF7D0bfdkKKf9UVUSXztlzwLvLgHajvk77lW3AV/8g79jxkBuMAJQ+wxByWBEWGFJZlP9m5EgCKKnQVEHgiAIQgoqwaWWb356LP6884jQZ3l+vPM6THScL7vAo1cANijW7GrEml2NmJibgZ0NJ7k+W1bTgrd2HcHsCcOUBWvstiPrqr35qqH47t+MUxqAF3Uy6SA2OL7rEJt4Ho/dcVchMqp2rqsqiWkC3UqyjshMx/+9eTRTb1XVfRF58HIrJuo9yCqXy+raDAKeHsEi187R6DmuJID49z7y6k6hY9cT+vOFaU2O3SdZp2fQAWK7eXZ9fib3/qgsia+N1/4J2LYCMNv9Gc8i4wpg7jPa+7p6QW6wxEdF8onK80rJYESYcUtmI8c2QRCEN3SXJwiC6OXIPkj3hBJcvMdAV69A67iIbr/tApv4KuIwUe1iKxoWwQ9uH98leMEagA0SXuHV4qsrtuITBZm4a0qukv2wC/rIZs7fWDhU6XWpu2cd63Ur66SOxSnYtmjWWHxpWZnUtlWVnYs9Ll+alo+/ftSKCsF5a0ddSxsWl1Tgd+8ewNduGoW285ccz4HKvogiuIk1QTlzZbCC+XdPvRIzr8p2FHbyBvUT7p2rAregdOz8bLtwEcMyUrm3LyOAylYISOT+fGHtI6rK6Rm2ALHI/qgsia+FJ8cBJw/pH8fCSAKGXwPM/a1vfV1ZCFrsJ8RQmXyi8hmC5gmRCNglGREEQRDeKBdfDcPoD6AQwADTNDep3j5BEAShBpU/QBfNGosFy8qYyr2GqQSX6DHQUR4q9riI/gh/ZsMBbK097nruZBwmKl1sA/v1kSq1l4h8UNOCDw+2oCCrP2qaTwtvxyl5QTZz3m7OeAWNnf6us2cdz3Wr0kntljQyNMIvIMUju664HZdJuQNxzZUDkR1Jw4C0FOQO6oeG42cuC2CXAJhI75OCpuhZvFRWB5PheNU0n7btTR1/DryC1LqFQjexJkhnLi/xwXw3Yae0uhmPrt4d2H7qToKQEUBVJE8lYn++sPYR1eH0DFuAmGd/VJbEV0rpUuDtH/njdh02GRg0ErjmC0DRbfrHEyRsYj/hjo7kExXPEGFLSCYIgiAIQi3KxFfDMPIALAVwB4BkdFQqS7n8txkA/gPAN0zT3KBqTIIgCEIMlT9ArYAqq/AalhJcMsdAdXmo+OMik03tde5kHCaWi+3hkgrpvqrxwVNV/UrjMaCnB6wo7SZwsPk0DANMAlc8bskLMtdVfPDHS+CcVZSNdZVNjn8/ceaClp51PNdt7uB+yoRXr6QRFYKOzLridVwqGk5g16ETeOCWQgDA3iNRDEhLwWcmDuty3u9+drPQvIzFbg0KWih0EmuCduayMGdCDv559lUJITRZ14lI/0teRAVQFSJVIvbnC2sf0dA7PX1G9D6grWxp0x7g1f8LHNqqZ/vxjPgk8A9v+jOWIsK0BhP26Eo+kX2GCFNCMkEQBEEQelDylG4YxnAAZQByALwGIBvA9Ji3lF1+bT6ADSrGJAiCIMRQ+QOUx1UWphJcssdA5XdwOi48buJ4nPZbhcNk/tQR2NlwslsfSBGs4GlpdTO+t6q7g04F00Zl4v0D4SoragIoGNIftcdOcwVrvJIXREX7+OAPi8DpNoaIiF5W04Ilaypx57W5ruUYea7bwpyIMuHVK2lEhRhj1/uPtawy63F5cm1Vt9ctp+rQSKqyBAinNSjIILWTWOPlzA0aN+HViSBKKlu9aFW5W70QveZUiFSJ1p8vzH1EQ+v0DAjR50vlz9YHNgDv/gyoLVW7XTcGjkg44ZVIDHQmn4g+Q4QpIZkgCIIgCH2o+uX4Q3SIq7eYprnBMIwfIkZ8NU3zgmEYmwAUKxqPIAiCEETVD1Ce/pzGZXEnLD8wZY+BaJ/bx+ZOZC5NVjwmCyMlytPa7bcqh8k90/OViK+nzl5UWhbWdoxz4QzQ1jSfxpK7JmFVeQNTsIY1eYG3BJqBrsGfIPvuPrNhP57ZsN+x5DfvdVt5JCq9T27HPVYc3XXohPRYf6073pmMwFMOXbZct+VUvX3ycPGN2MAawCzMieCqYRHsVXC+3HATayxn7sotdfjpnypx4swFrfvCikw5wkWzxmLB8jJpJzML+UPSMXt8Dpa/V+Pb2iEigFY1RtEUPSs9dlieZVgJs7s0dE7PgBF9vlR2nirfADYsAY5sV7M9VkYUA//wJ3/HJHoFfiSf2FX3aIqew18/akVFw8lu7w9TQjJBEARBEHpR9avlbwC85lFSuA7ApxSNRxAEQQig8gcoT8Df9LGEnReqjgGPyGU5C3lcX1WNUam+oED3/VblMFHlqtpY1YQPa49rKws8OW8gdtTLi2K6aDt/CSvvn97N4Rjbj5O3fxhPCbSCrP748dyJSoU8FdiVrdVVljqe6/MH46bCoZ5OUx3uvodX7cC+pqirgBV/bFQdl3YTeG37YentxMMawDSUj9wdL7FGdyIILzLlCDvbAfj0XSbnDsSyTTW+lnjneZ5Qec0mYn8+0Xv/LhvRQDWhcXqGCJHnS2n+/H1gyzLgonxyAhf9hgB/95/AqJn+jkv0GvxMPrH7nUc9gQmCIAiid6NKfM0BsM/jPRcA9Fc0HkEQBCGAqh+gYS5h54WqY8AqcomWlVLRQxIAXtlWj8VzxgFQ6zDhdVjasaX2uPiHPUgygKtDLr5aAXHVpVi9SqDlD0nHN2aOxvypI7q87pfAyUJ82VpV14MXR06cxbdcAtk6Bbp2E3huUw3T+6xjU9Wo1y2qAq8AZlVjVIlL2Qu3NThIx7cdMuUIgxCRV+9QL9y7wSOArtxSp6RPOaCnP58fgXnRe/+WWv33g8CdniGkeEwWFt0yFr9au8913iopW3pgA1DyFeB0k/g2REjqC9z+S2DKvf6OS/Q6gi5tTj2BCYIgCKJ3o0p8bQFwpcd7CgEcUTQeQRAEIYCqH6BhLmHnhaiTw+7YeYlcMmWlVP3of2bDAWytPW5bxpUVu89ZwcGn1nrlXvmPAeCBWwp9cdPJoLNsYvGYLAyNpOKVbfWoPNwhbBUNz/DsqRomYsvW+tXfr/74GcckkdLqZmUijizWsZmRAO4vr3Pnx7zzEmsef3NPaIRXmftG2ERkHfAIoCqv2fgS7bLwlheXQXQ7tcfaXJPmVAnHgTg9QwqrS1u6bGnTHmDjz4GdJWKfl+Hm7wM3PeT/uESvhEqbEwRBEAQRJKqeKEoBfM4wjGGmaXYTWA3DGAtgDoCXFI1HEARBcFLVGBXuSxj/AzToLGJRVm6pw8vl9UKfbbtgv+92fX5UuFdU/ui3SpUunFHA/Vk30SIjrY/srmnBBPDk2irkZ6YHvSuu6Cqb6BS8Xb/3KLbWtjoG9YO+Pu2wHPN+BsGckkQeeXVnKIRXi7KaFkzJHxT0bnjide50zzs3saa0uhlPvFmJnT6UV/WiaNgA/PqLU7pVmeC5r4ShbLhOeJ1+//b6bmXX7F1T8jrLoMvi5U62K70uQ2FOBCMy01HX0sb9Wbv1ULVwrLuSSFjwup5ZXOsGgAdvLRQXoMtfBN75MXCqUezzMuRMBD7zEyoxTPgKlTYnCIIgCCJIVEWyfg7gbwG8axjGAwDSAcAwjP4AbgTwFIB2AL9UNB5BEATBiIpeZ/E/QBMxi9hyBInj7qNUXVZK9Y9+1pKm8bgF+IIU667PH4xPFWa5luWrFQg0+4XqsolWUHfz/mNYu7vR8Zi4BfXDmuVfWt3saxDMbl6v3FIn3YNZD2H3d3uvZTrnnZtYo6I8r2EAI4f0VzI39h45haPRcyjMiQiJW2EqG64DXqdfVWMUexWWs56Qm6FkO6zu5PjS63bwiPOfKMgUEl/j10MW4XjBsjLccfUV+OanxzDf53RWEgkalusZANO8MAH86u0qXJc/mO9YlL8I/Olf/O/pmpoBjL0VuPFfgOxx/o7dS6Eeo12h0uYEQRAEQQSJkmiHaZplhmF8FcDvALwe8ycrjfwigH8wTXOXivEIgiAINlQEl+1+gCZiFrGsIyi9T7Lr31UHO0SCBToYGkl1/FuQYt2Hta34sLZV+XYHpffB8bYLyrcbi8qyiSLJFU5B/dxB/ZTsk2pOnb3o6/VgN69/u36/9nFFSO+THIp1wgmWAKau+4KbWKOiPG/e4H5YctfkLpUP3t7diNL9x4S2Z6KjlHR9a5uQKzJsZcNV8TlOEc/ilW1iVS6cUDVPeZ5FYkuvxyIizk+4IgMvb+Xf39j1kPW6MQG8tv0QXtt+iMsJq6uSSJCwupzzh/SXnhe2HNgA/O//Ac6of15ypX828OXXSHD1ET9LmScaVNqcIAiCIIigUBY1NU3zecMw3gPwDQA3ABgC4ASA9wH8xjTNvarGIgiCILxREVx2+gGaaFnEKhxBTkKjzmAHT7BAF259entaECfJAP7976dgZ8MJPP5mpbZxHrilEFWNUZTXtkoFlmWSK2KDtyrc8Tqxrj2/rod4EbqqMRpaF/WAtJRQrBNOsAQwZYR1S2AdGkn1vTzvwhkFnWtgbOUDUfEV6Cgl/cHBFpgCrsgwlg1Xwes7DmH+1Cu510ir17UKVD2/iDyLWKXXrfFFSxarSJoTuW5i9+eaEYOYrlMVlUTCIODyuJx53fPx88KWP94L7P5/XNtVwohi4B/+5P+4vRi/S5knGr2ltDlBEARBEOFDqWXFNM19AB5UuU2CIAhCDNngstcP0ETKIlbhCBIpWykb7LCCBQ+XVATWa9ItoB8Wd64KYud78ZgsDErvo+24P7m2qttrvEK9iuSKspoWLF1XhaVv6+8T+fWZowGYeGbDAe7PWseENXgmy1dXbO1yPsLsKLQEBT+OiwhuzvlYeO4nBoAFN4zAPdNHdhEeWIUVVeV57a5VFQFbL+HVIt79JlqJoHjMEJRWiwvGuuFy+WlAdaUC0c9Z5agfXlUhJM7LJs3JXDftJvBQyQ7bv6l25IXJ/ae7B7NbchyevQk4/Fd9g9vRdwDwhd9TT1efUVnKvCfTk0ubEwRBEAQRXsLZ3IsgCIKQQja4zPIDNJGyiA8ek+vHZ+d68SvYkTc4vUNtCEhU2dcUdXWQhNl1F8/wgak4fOJct9ft5jtLkCaSloK39zQp2TdeoV5VUNetZ65K0vsk41uzxmJr7XGutWlSbgZKq5vx551HOufeioXT8LWXtiKq0eln9S2cPT4H5y+1axtHhrzB/TqvQ6/5GhSu4kAMvPcTGeeOCjF9WkEmAOD50prOdTG9bzJKyhukt81DrPtN9B77wzsm4JFXd4Y6iYbJ5RdH0fAMrN97VGpc1c8vou7kU2cvorS6GV97aauwOA/IJc3pSkJR6cgLk/vPjx7M3eZT0x7gwLvA5t8AJz7SOnYXUtKAqf8IfOYn/o1JdKKilHlvoSeWNicIgiAIItwoEV8NwxjB+l7TNOtUjEkQBEE4IxokmzMhB/88+yrmH6BhzyJWUU7VyfXiV7Bj6bp9zMFWHby2/TBe23642+uxDpLH501icuMEjZ3wOil3oOMc9QrSPL1unzLxFWAX6lUGdf06ZaKlgysaTqKiYXeX1yblDtQqvFqYAP68u1H7OKLcfNXQLv92mq8nz17wxd1sB4/Y5Nf9RLY8rwHgxJkLmP3URqntqMISuI9GzyGSlsJ1bViJRYmQRLNicy0emzuR+f13XpuLZzaI92qelJuBhz87TmmgXtSdvK8piifXVnGv1/GitUzSnM6y1ioceWFz//lRMaFzPh3YALz9I+DQNu1jdmHYNcDMh4Ci2/wdl+hERSnz3oiK0uYEQRAEQRAsqHK+HgRb/M5UOCZBEAThgGiQbMIVA7l/jIY1i1imH6aF4eB68SvY4YdzQpRYB4klmvzg1Z04wNm3LGgqGk5gwbIyLLnL2QnjFKQRDaS7wSLUh7kMrhMqSwdXNJxQuGf6sMrjNp48h7c0iLj3TB9p+7rdfL0+PzMQV+yuQye41j0/7icy161x+X8rj6jrJyrLrkMnhe53sYlFfpX0lmF9JV+iS2FOBP36JOPMhUtC41U0nMRjr++GCWCvzfkWKWErKvat3n5YOFEm3n1+7YjB+NK0EVi/9yjqW890e79TkoOO+10sso68sLn//OjBXDwmC/jz9zucrn6SPwO46V+ovHAIkC1lThAEQRAEQehF1a+oF2Evvg4CcA2AfAAbANQqGo8gCIJwQTRIJhNcC1MWsYp+mJG0FPxuwXW2wTm/gh1hF9niHSTvfGcmfvLGbry4uRbnLoazVKsdJjr60bW2ncf9N41m/pyuwK2XUO9HUFcl8WW7w1oiVzUmgA8PtmJvo3qhzq4Uuht2omZT9CxeKqvT6lhfs6sRa3Y1IndQP3y6KBv3TM9n2m+d9xPR67ZgSDoOtrSFzuG/seooVpXXcwuv8YlFYb8u64+f4RLyqxqjwsKrhZvILlLCVqTvKq+bOR7rfuFWCYTl+vSjgomoIy+M7j/dYvW0gkwU/vUJf4XX3OuAv/13IHucf2MSrsiUMicIgiAIgiD0o+RXgWma/8fpb4ZhJAH4AYCvAfiyivEIgiAId0SDZGHsASTigJLth2kAjsIr4F+wIxGCI7EOkpVb6rD8vZrQOqe8ePzNSgxO76s1kM6Km1CvO6irEqey3U4Ox1XlDQnjbGVhjwaHpNMxZSFe1JyYO9AXt2PD8TNY8X4tVrxfK+QYVInIdTs5byDS+iSj5libxj0ToynavZy6G26JRfHX5a/f3ofWMxdU7ao0PAlMfiQviZSw5SnxbBiQLrE+IC3F0xndcPwMfl9Wi0m5ztVPdN7vYhFx5IXR/adrfRtr1GNG8k58rd9pYHOJljG6kZIG3Pw9oHiRP+MRzASRbEsQBEEQBEGwk6R7ANM0203TfBQdpYmf0D0eQRAE8XGQjAdeJ5VuSqubcfezmzH7qY14dPVu/HJtFR5dvRuzn9qIu5/d7Bhsky3Vm2QAS+7qXmo4Fr+CHYkSHCmracHSt6vwcEl4S1aysnjVDq5A7qJZY5FkeL+PFzfhXUVQV8UuGx4bsXPXxVOYE8F9xQX41mUxricJrzowADxwS6GywP78qSOwYuE0FGT1V7I9FizH4B+3fOTbmPHwXLdJBvClaSO0ik4alhBHomcvYmgk1fU9hTkRpPdNDpXwCvAlJPmVvGQlILFilXj2mn9JBnDH5OGSewek901m7oW6uGQHVm6pc3yPrvtdLCLnLYzuP5Hn8IKs/o7H9++S12ND3wewNvUh/DDlReQc8El4vfqLwCONJLyGlJ6UbEsQBEEQBNET0S6+xvAXALN9HI8gCKJXwxtcFnVS6WDlljrcs7zMMdjtFLyvaoziybf2Co87rSATKxZO83Q++hXsSKTgyFNv7xPuSRdL1oC+CrYijqkpkM6Lm/AuEtSNJckAHry1UPjzFgumjcA0h/2Iv5aqGqN4vrQGT6/bh+dLa1BlU4o37GW2/SI/M93xbyaAJ9dWuSagiFB7zN9ezZZjMKhzziOAPTFvMtrOy5WvdWPcMP+TnliO+2/X7/dhT/jgSUjyM3nJKmHLipX04LV+js2WmxvTCjJRUt7AnBRlAlhcUuG4vui638UyIC2F6X4R/xkR/ndrvdY1iPc5/MdzJ3abF3+XvB67+34ZP+/zHEYm8fU9lsMAPvcb4M7f+TgmwUtPSLYlCIIgCILoyfhpqckE4F9aPUEQRC/HCpJ5OR5Y3Gl+wtqvNbbcHwDHXmasfHl6Ph7924lM7xUpv2cFO3jKKIuMMyk3AzPGDsUzG8IXOGeh+dR5fOVTBfifLR/ZlluMpKXgmzePwbrKJm1ONN5ecDp6JXpdjzylK2OZVpCJf7rsMn2vulnqGGZH0vDY3Emuc9qtz2B+ZjqmjszEhNwMFI/JSogy237w+evy0A4Tv1rrnNAg0m/SCdky7aLEliwPAq/rNvZaeZojIYOFm68aihsLh6J4TBYeeXWnksQVHryutarGKGpbwldi2W2uxK5DbRcuoeU0XzlmWXhL2DqVXrfWz6rGKHYdEq8EkGQA86bkYnFJBfdn3dYX3b2BV5U34NHVu7u97lauXHQNqWtpU7aO2iH6HF48JgsfbX0Tw1//MlJMf+cxAGDIGOC2XwKjZvo/NsENz/Ng2JJtCYIgCIIgejq+iK+GYdwCYD6AnX6MRxAEQXTAE1wOCzxCQLsJPPLqTtQeOy0tHozkLLvJG+z4dFE27n52s60I5RZU5B3n4c+O43LghJGzFy6h4kefwVu7jmBVeT2iZy8ikpaCeVPyMHvCMAAdvSpFxEdWVAXScwf1w4/f2IM6DiGDxZXAGtQ1AMwen4MbRg/pJvSLCrgWltsovo+ohVefwdqWNtS2tOHl8o5/uzk+exNN0XP4fVmtpyAn0m8yHtky7bLwJjqohkUAe760BuV1rUrH/e7fjOvcfhDH38spGEYXutO66Jbg4SeiySPx62dpdTMeeXWn1PcxLot5p8+LJ7S4rS/FY7IwNJKKV7bV4509TdjbeEp4nHicSs+7CcIy/WhVrKNuCD2H/9ftuPLgJuX74klSX+D2XwJT7vV/bEKYRE22JQiCIAiC6A0oEV8Nw3jHZftXAhhx+d//pmI8giAIgh2v4HKYEAlE1zSrKZcpUhKYNdgxb0oulqypdHyfW1BRJKhSXqtWKPCb9XuPAgBmTxjWKbbGw3pcRJENpIsKAjyuBNnkCusYPryqAqbAMXS7Zlgd7LGE0WkXBNs+auVKQFn88g4s/FSB0JoeBpGNN9GBBd77nQ4BzIlYETHIsstu6HKhf33maJw6ewEvvV/H5fZ1Whe9Ejz8JFbQFn3eUvF9Imkp+N2C65Q4tu3c6UGK3W5CqUwykW4XPvNz+Gv/BJS/oGUf3DGA6f8X+MxPAhibUEEiJtsSBEEQBEH0BlQ5X2c6vG4CaAXwZwC/ME3TSaQlCIIgNOPkTgsTQQWiRfsfsQQ7Pl2U7Sq8WrgFFXmDKm0X9PUn9IP61jNd3HBOAUuv4xJJS7EtXcyCTL9A0QC6qCvBaRiW4a1j+LWXtnIdK69rJqhStolOQVZ/7Gw4yfWZ+uNnOst0urno7QhDqWeV++AmDLEeG52CXryIKPvdDQBXDYug8gh7tQOW+52ufql3XpuLwpwIJuYOZD7GTuuiSIKHTorHZEnNPxXfxwA6hVdAzXmMdaezXBvx1Ra21bUqPU9OQqlsQpYKF76XuOr6HP7kOODkIeGxxUgCZj4MzFzs87j+kAhJpypJpGRbgiAIgiCI3oKSX9amaSap2A5BEATRuwlCCJDtf+QV7Lj72c1cLrZfr9uHoZHUbttiCaqEpfyiCkqrm3E0eg5PvFlpW4YwNpBtHZdXttWj8nCHCFE0PAMjh6Tj4ZIKoX6Kou4A0QD65LyBWDynqMu4XgE0r0A4a1/Q4jFZ+N2C67BgWRnTsbLc3M+X1tju21u7jvSIORgEsk5+3l6wukQ2HlTtA+/1YHd9HY2e0yq8xouIMt/dALDkrsnIHdyP2fFnALgyMx3Pl9a4BuR1uKNiRV/WvqFubq0wJXhMK8j0FBm9rk3Z72M3v1SdR+t+zHJtmADW7mnEvZ8c2Sk2up3rSbkZqOBMOIkXSmOv5S9NG4F1e5pw6MRZrm0C4i58YdG9aQ+w7lFg75vcY0pzzd8Dc5/xf1wfUJGEk8gkQrItQRAEQRBEbyH4iAtBEARBXMZvIUBl/yO7YIdIGeWymhbMfmpjt9djA0Yi/TUTjdd3HMbW2t2Of48NZOcO7tct0GaVLi7I6s8taIk6oQHxAPqO+hNY/PIO3Fw0FBNzB6KkvME1cAiA6Xyz9rMrHpOFJ+5i6yGbP6Q/FpdUdPtb0bAITAB7OVx4iUaSAVw7YhC21h4Pelcc4elhePLsBZ/2yhkV6y9r0kO7CSxetQPLS2ts52kkLUXLGpqTkYoRg/th1+ETeL70Yuc6LvrdC7L648dzJ3Z+ntXxZwJ4eWs9Xt7a8W8nIUKmh6YddklO8QlFB4+dRuOJs8jJSMPIrP6u4rCKXrkzC4diQ9VRqW0AH/d0l1mPZb/PtIJMzJuSi9PnL+Lpdfu6JMRMyh3o2EeVlVNnL3Ld2+LdqW7JY6XVzahocL7XO2EJwioTzkQSAIWSoA5sAN79GVBbKr6zgjQOGI+ceY8Do2b6PrZOrLm1ef8xrN3d6JhIxpugRBAEQRAEQRAykPhKEARBhAY/M9F19j+ygkAbFQR2LT6oacGC5WVYYhMwClv5RRVsZehb224CD5XscH0Pr/Aq44SWDaDXHz+DFe/Xub7HChzmD+kvHAh3wsuNVpDVHwePnXY8pjylTxORaQWZmJw3EMs21QS9K56wnPPS6mYsfVuuJ6Qs+UPSlWyHRxgyTecEAdEy5RaWCLaz4STeqWxCw/EzAIDGk+fQePIctsSI9pbwyStyTsrNwOpvfarLa6xO0njchIhFs8ZiwfIyoX7QsXglOYm4tFS0KEjrI1+4yfpuL5fXS63Hot9nzoQc3FyUjZLyBtuEmE8UZCInIxUVDUKb76TtwkWhRLL4Mr525/rPO48I7dPLH9Zjz5GTSp97eBMAeZI+Hl61A2OMjzBlx6NA3fsSe8mHaQJb2gvxp/YbUNo+EX9/62zcN6rAt/F1I1LxhSdBiSAIgiAIgiBkEBJfDcO4V3RA0zRfFP0sQRAE0bNR7baxY86EHPzz7Ku0lOTSXfbXvCw2tradx/03je58PUzlFxMZQ9IJ7VfP4naTX1Rm7Wfn5FBK75uM766qkBZiEhEDwIO3FuK6/MG4ZzlbaeYwUFbTgiVrKpHeJ9m2bHUY1o3aY22Y/dRGqVKQKlyQMtx81VDcWDi08/iu3FKH35fVeh5bS/hcOKMAHx5sYe5/+vBnx9n+Lf7a3dVwEiXl9Z7z1UmIKB6ThSfmTRIu3Q7IJzk5lV5X0aJAVmjPG9wPS+6ajKGRVM8koHji12PR79NuuldA+KCmBYbQluMR2wpLGV/Riie7DvOVKmaBd56yrqF3J63Hw33+gMzX5crK87K9fRSWXPwC/tI+sfO1niQ2ylR8YU1KIxIH6nVLEARBEEQYEXW+/hfA/TvcuPwZEl8JgiAIRxbNGsvcv04EXcKrn2V/H3+zEoPT+3b2LaT+mmq4KiciVYYuiJ7FPPD0s4t3KPH0Lu5pmAB+9XYVCnMiCXcMntmwv8u/LZFzaCRV+bphGB3XkIgD2hIiH7ilEJG0FK7gqV9JD05MGTEY9xUXdO4Lz32g3QSWv1eDhTMKsPy9GtfPsZbJt67du5/dzPxjzUmIsBy1P359N/Y4nNeiYRH84Pbxtr3KRe+1Xj0br1JwD49ItjlYOKMAxWOy8HypmBM+dj0WFSDdyqtayC5Z0woykd4nWeizLPfEsIhfvO0GWJ69Ppm0E0tSnsWVScdkd4+JdhN4+dKN2G2ORGn7ROwz87r8XaalQthQUfGFNSmNCDe9vccvQRAEQRDhRvRX531K94IgCCIkUNZs8BSPyWLqX5dkdPSe5HEA6go8BVH2d3FJh1OpqrFnl3r1k8ojUalAnN89i3kRFYdJ4O8IaveEssqWyHn75OFKt2sJg1ZCyIrNB7F+71HUt55h3ka7CTy5tqrb617B04OcLnDVxF73Im7idrOj5/OKhdMcywbzOkhF+43brX/FY7Lw5gM3oqoxile2NaDysuOwaHgEd16b162srCwsPTS3KFiP5k3Jw5pdjcKft86F6Loa+zlRYUD3I4dVhl/0OYPlnuhHxRMvRNoNeCV9fC/lJXwl+U8w1FiPPblkAg9f/Cr+99JM27/LtFQII6oqN/AkpRHhQ6jnMkEQBEEQhI8IRQlN03xB9Y4QBEEECWXNhguv/nVWIBoAs0tWZ+ApiPKdJjqcSjNoXipFJhAX9jWiKXpW6HNBOwt7K1bJGNW0m8Dq7YeVbS9eGCzMieCxuZMAdIiA//jCh6hraRPevlPw1EqW+sv+YOen9b1lkhTKalrwXnUzHpvbUR5UNglM9Jp1W/8KcyJYPKdIaLs847MkMqlwc86eMExY9JucNxCl1c34884j2HXohNA+xAqTYRAg44l1Wg+NpAptg/WeuGjWWCxYFkxJd1ZHeTxuovtLKY+hOHmPb8Lrqfa++OrF73QpLxyL6HcMKyoTwsJesYRwhrfnMvX4JQiCIAgiCMJt0SAIgvABypoNJ069J+MD0awuWV2BpyBdgWU1LZiSPyiQsXsqMoG4MAbQY3mprA6Tcgdxr2MUnAyGJ+6ahFXlDbYJKLKIihzzp+ahaFgGlzAoI7xaxAZPAWjtrc1DbDUF2SSFZzbsxzMb9itJ+FLhxgwCPxKZYhOxRNsc7Kg/gR31YqKrRfz55dkXXYkZFlYp6diECt57G0+lkeIxWbh1fA7e2i3uRBZBpidxvKv33uQ1+JfklRhgnPNNdAUApKRj+xcqcGndPkCBaz4RUJkQFvaKJYQzPPcL6vFLEARBEERQ0NMmQRC9GsqaDT/xvSfjYXXJ6jpvwbsC9Ub5ciKpmD0hByver9M6TliQDcTp7lksg8mwjtklO1Bw0n/GDYtg/tQRmD91ROc5Wb6pBvXH2cv46mDhjFFcLkyV62O7Cfzg1Z04eOx0KK6v+GoKqoRLFQlfotes9bkgWjCIJjIZRsfaxkJ8IhZrmwPV2AmTPC0Xbh2Xgz8LCJVeoq0B4MFbC22rhPDc20QqjUwfPUSr+Pr1maOQHUlTNqetOfSTlOfwheT1SPZTcLWIXAF8ew+KL+9Pb2mdojJJhH7TJSYqS+sTBEEQBEHoRFk0zTCM/gC+AeAzAHIB2NUnMk3THK1qTIIgCFkoa7ZnwOqS1UHQTqH0Psla3ZYr/nFax//2EvFV9voOKpjPitM65lZ6fWJuhl+7R1xmz5Eo/rjlI9w99crOBJTCnEigwr5Iz2zV6+OBgPu7WthVU1CZpCCa8GXdA0X74Kb3Tcbdz24OpAWDqFC/YNoIVDWe8nSIOyVizZ86AgDw2w37UXtM3qXthYGOpLKn1+3r9pzCmkxW1RgVEl8X3JCPqsaoUKIajzgsUmlE97N1fH9iWQpzIniv/0PIvVjvr9MVAJAE3PojoHhRt33qDcKSqrVW5J5GhAMdpfUJgiAIgiB0oOTJ1TCMQQDeAzAewEkAGQBOAOgLoN/ltx0CcEHFeARBECrQkTXbW7LOw0oQgae2C8GKrwPSUrS5LWMDU7wCb/6QdJimiboWPqdeQVZ/1AQksLAG4ryuc68AetDEr2Nepdd3Npz0ce96HqJzenHJDlQ0nMA90/NRmBMJVNgX7ZmdyK7pSFoKojbisZNIpVo84kn4ckueYKUgq3+gLRhEhfrsSBoemzupy7rcduESABPpfVJcn8Xcjtvg9D64fmQmPjl6CErK65WtgyaAFe/XdnktVthmSSYT7cFqrSWiz6o6K43oLtv/yKs71SUPvPI1YPt/Iw/QXXykK0YScNPDwMzFPg4aPlScQ9F7GhEOErW0PkEQBEEQvQ9VEZFH0CG8LgTwXwAuAXgKwGMApgH4DYDT6HDFEgRBhAKVWbNuATzdbhFWSBjWRRC15j7GOo86RJnYwBSvwCviIPruZ4swKL0PFpdUcH9WFpZA3Motdfj39ftte1jGX+exAfRXtjXgL9XN2C7ZI1Al1jrGWno9kchIS8HJgANsBoAFN4zAPdNHAgBmP7WRexuWSLPi/drO+RWEsC/TMzvo+54oSQbwuwXXYWgklfm+qUM8YimT6JU8wYIB4OCx057le3W2YJAtlcybfOV13FrbLmDdnkZMyh2oPQHFTth2+z6yPVhlEtV0VhrRWbZfSfJA6VJg7Y8AtKvcNQYM4HO/Bqbc6/O44UR2rZW5pxHhQPZ+QRAEQRAE4Reqnj4+B2CjaZrPA4BxufaOaZomgPcNw/gbABUAvg/gB4rGJAiCkEJV1qxXAE+3W8SLRBCGE5n9TacCGzs2mKpDlIl111iuu4dLKlx7xolgGMCSy9fHt/67XPHW2bACcXYB5aPRc3jk1Z2u7kW761yFG00X1jrGU3o9UYievYi7puRiVXmDe3/Dyw0Qlc9nAEvu+ngePF9aI73ND2pasGBZWWc/xvi52nbhIp7ZcEB6nHhke2brdrPpID4wzyMo6RCPrEQJp7VJVnhNMoD8IezubF0tGE6eESuQJLIfrEkn7Sbw1Noqof2aMyEHE64YiKboWbxUVictbMef/7um5OLDgy3aerB6oaPSCKu73wBwy/hsrNvTxDX3RZIHqhqj2PXX9zFz24MYdPYj/9PtUjOA737k96ihR3Stlb2nEeFA9PzReScIgiAIwm9Uia9XAng95t/tiOn5appmk2EYbwL4Akh8JQgiJKjImuUJ4Olyi7gRdmE40SmtbsZagb5rKrALpjo5Upqi5/DMhv3cY9j2RrosWKkiNhBWWt2M17cfVrdxDkyYjr0OWWk3O0rFfnCwBecvXsLq7YeVC3uqGJCWIlR63Ys+SQYuBKzmmgDqW8/g4c8W4Tfrq21Lx0bSUvDNm8fg7MVL+NXafWrPkwHkDu7X+U9VZe5MAE+urcIbFYfxr7eP7+Y0e2/fMVQ0yLmr8zPT8fnr8pRWR9DpZhPBMICrciKoPBLt9je7wDyPw88Sj1S693c1nHRcmyJpKVLHdVpBJuZNyeXeXxZHLiul1c2eyS1OjBvGLgDGnsf/3VrPfNxED++EKwbiMxOH4SsvfOgpvFq0m8D3XqnAc/de3/m93JJ4CrL6ezqWE83lx5JIZgJYu7sJRcM6jpHdtewEa/JAaXUz1v3pj/ha8+O4Mymg0vtjZgML/jeYsUMOj1B/6/gcTB89hCr+9CBk3f8EQRAEQRB+oUp8bUNHqWGLEwCGxb2nEUCuovEIgiCkUZE1y+Ma0+UWcSLswnBPYOk6xaINI17B1HhHytPr9gmNEysaWfOJNYjMyl1T8jq/R1DHE4AyR68J4OWt9Qq2pBdL7FaNqPDK677zoqymBR/UtDie0+jZi3j8zUolY8Vjxq31qsvc7T0StXVZN548K73t2pY2fGbiMKUBSh2CpCjW2nn31Cs9RVXRqhHXjhisdJ9Lyutd5zEvX/5kPkYO6d/5fUWd2bbJOZys3FIntfayfC6oCgQvb63HLwVcs7XH2jD7qY34REEmrs4biOXv1Tg+x9U0n4YBYFRWfxywWTsTzeUXe03OmTgMhcMG4KXNdY7nufJIFIaAFdUreWDlljoMe+3v8YOkChhJ/NsXxkgC+ucA13wRuOWHPg6cmOjsQUyEH57ELurxSxAEQRBEUKiKBn2EDverxW4ANxqGkWyapiXKzgBwRNF4BEEQ0shmzYq4xlS6RbwIszDcE5BxDU4ryMRjcyd2CfznDuqHhuNncOrsRTRFz2J7/QnssOkRKhJMEhV/mqIfizm6ytM+VNIh/A+NpAZamjRIU96k3AzMm5KHzfuPYe3uRu37Yq1jv3+/VvNIbORnpuOn8yYBgFKHZJDnNHat17GuxibN1Le2Ke3bq0JUi0e1IClC/NrpVjZVpmqE6qQG1fN45JD+uK+4oPPfqlow8NKZ0COxjcojUU8RLaie1rU2vcF5+OByAokXJjr69S65axLazl9S2oPVL2QEctGEMKd1rrbkXzFv+6/RJ9m/SWMCMMbPBe5+wbcxewo6exAT4YbV/Zxo7n+CIAiCIHoWqsTXdwHcbRiGcbnP60oAvwbwhmEYqwHMBHADgGcUjUcQBKGERbPGYsGyMqbgn4GuWbOiAVYdge14wi4M9wREz781j1j6pakKJokGHLZfFn91lKeN5dfr9mHOxPiCGb2HioaTWHBDMt7eo194jc3+rz4aXL/iWD5/3cfuZ5ZAWqJgrfW6+p62m8Bjr+9GVWNU6fFSVSY5Fh0ua1aKxwzBD++YwLx2ylaN0HH8VBK/f6LJOfsk+52rSuix5paOfriJQrsJrCpvwMr7pwe9K9wsXVelvvQ7A92u0//+IrD3T8gHAB/drnvb81Bz/Q8w53Nf8G/QHoiOHsRE+CH3M0EQBEEQYUeV+PoCgL4A8tDhgv0dgE8DmAtg9uX3lAJ4RNF4BEEQgSMaYN1YdVR7ZnaYheGewub9x4Q+d+v4HOYggKpgUmFOBBNzM7Czga9v2Y76E50CsE7Kalo6e7f1Vn67fr92kcBA1+z/Cxfb9Q7ISKz4w9LvL1HYdejj623RrLH40rIy5WPw9DpkhVWM40kOkREk8zPTpVyEPMIrIF81QnWZadXE759oUHr19kOYP/VKoc+rTOhZtqkGj67e3e112X648Shud66cREugK61uxmOv79ayhrHQeR2svAfY85qvY582++LlSzfhpUu3Yp+Zh2/3L8QcX/eAIHoO5H4mCIIgCCLMCEcHDMN4EsB/mKZZaZpmOYCvW38zTfMigHmGYVwHYAyAgwC2mKYZjigfQRDEZXh6TJpQ08dv/d6jWL/3aOe/3XrHieJXGcHe+kO3tLoZa3c3Cn12+ughiveGjWuvHMwtvgId3/XgMTV9ON1Q0a8ykZEtUcnCXdfldSmR2j81HCJR/NoXH0g72HwaL2wOR4lkHrbEiEtDI6kB7gkfTvci63zsajiJLQdbbOes0/1M9H759Zmjcee1uZj91Eahz48bxpfAoqJqhIp7uU6hL37/RJ3Z8c9EPKhM6Gk4fsb2dZF+uG48cOtYLH1bT/l9VSRKAl2Q5aAtPnthLfDoYsDH8IRpAjvb83HHhce7vB72hA2CSATI/UwQBEEQRBiRedJ/AMAiwzBKAfwHgJdN0+wSOTVNcyuArRJjEARBaEM2yKpKLHXrHSeKaCCH9XNu/bl0iMlhg0e0j6cpei4Qd0q2oPizbFONY3BbJUckxFdZZ1xvYcIVGV0SJs6HwPkaSUvB0eg52+shNpC254je0tc6qG1p67zWVySIeFxkI1by9GN0up+J3g/uvDZXWBw0ADxy+3iuz6ioGiFbZnpaQSauzEzHy1vrhT7vtW27a23RrLFYsLyMu3+mqNsy7KWZ45lWkIlFswpxfX6moys/DPehRDiurGW9dXF30nr8MPX36L/B33PVbgL/felmfP/iV7r9rSc/LxMEQRAEQRBEb0ZGfP0BgH8AMANAMYBfG4axAsBzpmnuVLFzBEEQOpENsqrs4+fUO04U0W2wfM7LsaBDTA4TsuUSn9mwH89s2O+7SC0qyPshvALAXz86IfzZ5758PYCuPf9Onr3g6VIKexlJ1ZSU19uW5wyS6NmLTOvFolljcc/yslC7zuyw7hfvVDYFvStM7G2M4o9bPuo8FyIONbv7mcj9Mm9Qv877NO/5NwAsuWsy9/qqqmoEz/4a6HClT7gio7N6xPOlNXhZcfpqbL/neIrHZOH2ycOxevth7u2KuC0TyekXe9zcylv+eecR/HJtVaD7GtbjGnu8Xt5aH8g6/nfJ6/FYyvNIMy76euNvN4G1l6bg/ovfsf27U0IEQRAEQRAEQRCJj/AvNNM0fwLgJ4ZhzALwFQB/C+BbAL5pGMb7AJ4DsNI0TX+itgRBEJyoCLKqFATseseJIhLoZgkAsToWVIvJYUJVuUS/Reqedh4sYp168fPXzaU0rSATnxw9BE+9vc+X/XTDL8eUSNlpP2g3gcUlO1DRcBz3TB9puw4Vj8nC4/MmBV6qkpeDzadR1Rj1LYlBFjNm7QYgfLzt7me898v642c6kwXyM9Nx7ZWDUF533FM3GZXVH4/NnSi05qmsGpE/pD9qmt1LticZsL0HqF6vkwzggVsKUdUYRXltq22bgMLsCAB+8VXkWSpR7kfW+bEr1WznEA+asB1XHte8CuwSqj6ZtBO/7bMUgwz97RNiMU3gvy7OxqOX/o/je9wSIojEpLe2YyEIgiAIgiDskU6PNU1zHYB1hmFkAvgygIUApgO4AcBThmH8HsAy0zT/KjsWQRCESlQEWVULAqIl/OzgCXSzBoCWrmPvd6ZSTA4TKsv6+SlSq3Rqh4ljp87h+dIa2wCXk0spvW8ySsobQiG8JhnAN24ejcUlFUHvSqCYAFa8X4cV79c5usLnTx2BvMHpjoJ6GHlt+6GEEouBjnVpyZpKnGi7ILXvdr1QRe+XtS1tTAkKBRLCq7WPsp9jdQsXZPXHjx32VWS9jqSl2PY5HTcsgnYAT9q4MmOvNd3tCmJJhPvRtIJM/BNHdYqgn3XC5qD0u6+rdb5i7w8/T3kGn0/eBMPwZx8sLiSn474zD+C99omO73ES9onEpLe3YyEIgiAIgiDsUVabyDTNFgBPoUNwnY4ON+zfAfgGgK8bhrEVwLOmaS5XNSZBEIQMqoKzqgUBkRJ+drAGulkDQLI9csMKb5a66rJ+forUiVq61Y2jp853uuOcAlyxLiW/A8JuxLreSsobQi1E+ImbK9wS1JesqcQzG/YHtIfstLZdwIr3E6Pfayw76sXLgMcSfz/TLaDXNJ+2nTtO67zd6zJVI3j6WdYec3fi8SZQ/W7BdRgaSeUqvx57rfktWIvej/qmJGnpV50/JB2fn5In7FYLUlAOm4PS776u1ve37g9H3/g3ZG75FZJxyZ8dsDCSgBu+gT6f+Qm+Xt2MCy6VN3iEfSLcBNWOhVy2BEEQBEEQ4UdLYxjTNDcD2GwYxj8B+Ht0CLDXA5gCgMRXgiBCgcrSvHYOu/K6Vqzfe5R7v1Q6K70C3TwBINkeuWFDNEtdR7CMR6SWCbYkaulWVrwCXH4HhN2Iv/Z0COOT8wYqE9H8xssVnt4nOYC9Inixu5/F3y+Xb6pBvcKyzLFzB4DjOu/kFL1qWASG0VE21It40Uu0OoTdui6aQBUrBLOsKdbxWrFwmpDb9mj0nNA93vp+D6+qYDrWQEdZ2fnX52HF+3Xc43nx0zsnSd/fg0hwCqODkuc6kKXL9y9dCqx7DEPbL/gzuEVqBnDn74Ci2zpfcusPHMZnYkKMINqxkMuWIAiCIAgicdAivsYwBsBkAPmX/+1z0R+CIAh3VJfmjXXYPV9aIyS+qnZWqgoAqeiRGxZkstR1uVu8RGpVwRYvQT5vcD/UtyZGf0o73AJcfgaE4/nc1cMxNjvieO2pFMYNAE/cNQlt5y8lrPgKuLvCVa+ThB7czpN1DVjOdZW0m8Ajr+5E7bHTjteTnfAKAHuPRGHAvn9kLPGil2h1iNuf3mTbj9la11csnCaUQCUiBPOKh9GzF6UcXdb96Aev7sQBxv6414wYpEV8HRpJld4Gj2A+79o8rNpW7/m+hTMKsKP+RMI4KEWuA1E6v3/GUeDp64FjAbQSGD8XuPsFxz/b9Qcmeg5+t2MJymVLEARBEARBiKE8cmUYRgQdbtd/RIfT1QBwBsCLAP5D9XgEQRAyqC7NG79t0X3SgWwAyM9+cDoRzVKPFa+vyhmALQdbmN06LLiJ1DqCLU67fulSCGyhktgFuPwMCMdzzw35eGyuc+83C0uI+N4rFag95t3f0gkTwKryBswIUUBeFCdXeJjEhrCTn5nO1C9VB+l9k20Tf4COtXhjFX+CEis1HmKeGyY6fsCMGxbBniPRbn+3E71Eq0PYCa9Ax7q+YFkZltw1GSvvn86VQCUqBA+NpHIngcg6uorHZOGd78zEyi11+O36/bZzNf54B5EAxQpPxZE7p+QyC+uJ4qAUvQ7cSDKAB24pRCQtpev3P70VePdeoLZU+ZiepA8FPr8MGDXT/7ETiESZtyL43Y4lCJctQRAEQRAEIYeyiHhcn9d0dMQsdgB4DsBLpmkmrvWCIIgejcrSvLGoLGusA96ASNjEZFF4s9R//PpuRPr10S7cOYnUPMGWxQzBFi8h9/DJs6y7HGriA1w6AsKs3DM93/tNlykek4Xn7r0es5/aKDVmWU0LpuQPktqGDOl9k9F2Xk2/PTtRJMj+ionCnAk5+OfZVwGA9HwSZXFJRSDjqsAEkNGvD9568Eame6WOKg8mgIdKPl7X3e7Rsff0XYfEfnaVVjfjvuIC5A1Ox9de2uroDo7HzdHF+qwxf+oIzJ86gun9Osr7qjx/rBVHeCqTJIqDUvV1UDQsgh/cPr773HrjO8CW55SOxcSA4cC9rwDZ4/wfO4HoDaVx/W7H4rfLliAIgiAIgpBHSnw1DCMTwL3ocLmOQ4fgehrAfwH4D9M0y2R3kCAIwg909WZSXdZYBTwBkfjjMSl3ICoa2IO6forJLIhkqdu5nnTgFCDhCbaYJvC1l7bidwuus91emHqe+kFsgGvz/mOB7IPINaBOWPSv28PNVw3FlBGDMSAtBbmD+uGrK7Yq2/YuB2dgEP0VE4kJVwzsnHskVIthJWTdV1zg+V6dVR7+8YUPsezL1zuu6073dF4OHutwCw+NpDILrxZ2CS8i4guLyKijf7mO88cqmCaKsMqCquNoAHjw1sLuz8WvfA3Y/t9KxuAiqQ8w6wdA8SL/x04wektpXD/bsfjtsiUIgiAIgiDUIPzryDCM/wYwF0BfdPw+2oYOl+vvTdP0J1JNEAShGNUBMJ1ljUVgDYgsnFGA7fUnpIK5fonJPOhyPxoAFtyQj+xIKgakpaCkvN6xlKQdTgKdSLAlevZiZ6nK+KBWkD1Pg8AKcJVWN2Pt7kbfxzcAzJuSK/RZFcJi23n/+i3fWDi0U6B6vrRG6ba31NpfAzoEmJ5EU/RjFzsJ1eJY9w2vxCyd9+8zFy5hwbKybmKU1z2dlxf+Uos9h6O4SvA5yEp48UN8mT91BI63XcBv1ldzC8V2kENMDSqOo23FmT9/H9j873DvxKwWE4CJFCR94UWg6Dbfxk1kZErjJlqJYj/bsfjtsiUIgiAIgiDUIJOaOh/AKVzu5WqapjqLA0EQRA9CV1ljXngCIs9tkhNP/BKTY2EJ2ugoCwl0BOiqGqOdfT0LcyLcjme7/RcNtpiwD2r1NuebFeBaum6fj+HajzHRUXa1pLyBu8SeCmGx8YR/JaTT+yZ3/rfq66z2WJuje8NaX3/8+m7fXOqJwvb6j6sUFI/JwsIZBdJre29k+aYaPLp6d7fX492bukthmwCeXFuFNyoO419vHw8AWhIPPqhpwRbB73Dq7EXf+hKu3FKHJWsqlXz/sFXpSGREroP8zHR8/ro8+2e38heBNd8Fzp/SsLfOXDKB/7l0M75/8Sv4xLuZWJTSTAI9AyKlca3PJVqJYj/bsfjpsiUIgiAIgiDUISO+3g/gD6Zpnla1MwRBED0VXWWNefDL9cgrJsseE57ShjrLQsaW9+JxPC+cUeC4/xGJ/Y3v9xRkz1NRRmX1x4Fm8ccM65oLWnQWdXlZwuKSNZXYUc/fwzEnI437M6KsKm/A/KkjAABN0XPKt+/m3igek4U3H7gRS9dV4VdrgxHaw8iO+hOda9LKLXVY/h4JryLUHz9j+7rddb1o1lh8aZnerit7j0Rxz/Iy5A/pr+2eLrrZAWkpvvQlVFlCP4xVOhKdRbPGYsGyMqZ5lGQAP503qfscOLAB+N9/AM742zLANIFXL03Hgxe/1flaTymTC+h1l4qWxnWbK2E+9iKJBqKJHn66bAmCIAiCIAh1CD+Nmab5nModIQiC6A0E1ddLlwA1KTcD86bkCQVxRPvBxcJb2tAPd7H1/Vkcz5PzBmL5ezWO+y9bSjFWEE6k7PdIWkpn31orULh8U42jCGJHsgGsr2xC35QkjXvKjqjLq3hMFl775gzc/vQm7lLWI7P6i+yqELFzbV2l+hLPLPN30axCXJ+f6XjN9UZKq5txNHqOSjNrIv66HhpJ9W3cGonEFF3kDurnS19CVclkQVTp6OlYz3aswqvt8f/V1cDxgzp2z5WzZgr+4cJD+Ev7xG5/k3VqB42KZ26WMUTwmithPvY8Jf1ZEz3sBHI/XbYEQRAEQRCEOigVjiAIQgNh61uky/VY0XASv7yb/7up6AcnWtpQZ1nIeIHIzfF8NHrOlx6MliC8rykxSrIaBjqFV+DjhAWeUs5AR8nAx9+sxIQrMjTuLR+iLi8A+O5nx3EH+PwSgixKq5uxra4Vh46rL3fM6t6Iv+b+UFaLfU3hE6n84tTZi72u13OSAeQP6e+bOBl7XSdihQFVTCvIxKZ9+vsSqkom86vlQ2+Cpwex7fF/cS5wYL22/XPCNIE3Ln0C37z4gOv7ZO7hQeJHD2ZAb4nbsB57nko3XokeXgJ50bAIKjnaK1A5dYIgCIIgiOAh8ZUgCEIhfmSW82AJEG/vVu9Es+AJmlrvV9EPTrS0IU+WOi9tF7oHnpyE+Ede3emLIGL14Fu9/bDQ5yfnDbQteRtJS5F25sbjFpwS7YG66xC7W9QPLJcXAO4EDRZBKfYYLl1XpWy/WTh19iL+q/Sglm3zrpuWYH+w+XSvFl/bLlwKvOy2n1iCDgBfklssrOs6kSoMqMRK+Hjo5R1Cn+c5bqIC95wJOZhwxcBQJMT1RHhKQRsGMG9KLqoaoyivbcVn9v0IY4+8DkP/bnbhUp8BeLRtHl68NIf5MyJO7SDxqwczoL/EbViPPUulG69EDxaB3ABggK0sPJVTJwiCIAiCCAckvhIEQSjCr8xyFtxEYNXwBptV9IMT7StV1RgVFvFYeGbDAWytPY5FlwMeTudgUu5AVDTw9/AUwerBJ/JVpxVkYuX90x0F5NjX9zVFsXr7YeFegSzBKSvAde/yMlxKYCff/Gc3o7XtQrfXnRI0WN1EBVn98eO5HeUSP/OrjdjL4ZBQQduFS6htaVO+3fzMdOHP+ll6OZwk8IXCyM1XDUXR8AwAJtL7pGhf550orW7ukf31vIL9SQbwwC2FeK/6KBo4SsPHwnPcRAXuAal98C0SI5gQqd7C82xnmsDikgp8N+UlfCP5T0j2W3VNGwTc/QJePJyPF1fv5v44b9JhkPjRg9nCj+TSsB57t0o3sfvrVAmH5V5lomM9NoyOa8gJKqdOEARBEAQRHnpehIAgCCIA/Mws94Kn7JsKeIKmMqJpbPBC1PliBW28stRl+KCmBQuWlQFwDlj7JbwCYj34gI7gjpU179SrOP71+VObmY+pAWD2+BzcMHoIcgf1Q8PxMyivbe0UTpyCa6fPXUxo4RWArfAK2Cdo8LiJao+dxoa9Ta59hHWy46PjWrZb29KG2U9tFKoeEKbgo4EOMdivcrjTCjKR3qfnP+o3nzqPZzbs7/b6JwoysXhOEd6pbPKlB/CpsxfxmYnDtI/jNwtuGIGqxlO2x3DcsAjaATy5Vs5lz3OdigrcJeX1+ERBpvbkNxUE1bpi5ZY6/Hb9ftskGrf1l/fZ7pvJJXggeRVSkgK4UV3998CdzwAATtXsE9pEojjcVT1zs1KYE9Ha2gMI/7F3el52S4qNpKWwJy4AKMqJYGC/PsIuW4IgCIIgCMI/en5EhiAIwgf8zCx3g0eoUQXP95AVTS1Egy+xn3PKUj/YfBovbK4V2r5FWLTBaQWZwk6kOyYP556jscd0xeaDWL/3KOpbu48fWxqUt0z3qvJ6rn1KNOITNHjXluc21ejdQRdK9x/Tun2R6gEiweB+fZJx5sIl0d10xAR8E14BoPHkWbxbddS38YLCKZnlg5oWfHiwBU/Mm4x5U3LxcEmF1rV5QFqKL+KDKE4l5L24Z/rIbpUOBqSl4OTZC1j6tnw/4bzB/biEHtFnJxP6k98sRMXToFpXlFY345FXd7quT27r7yvbGpjG+Uryajyc8j9INgJ4SkrPBj7/HDBqZudLokJ+ojjcVT1z86CztQeQOMc+Fq+kWN4WHpVHonjrwRsB8LevIAiCIAiCIPxFydOrYRj3AviraZqOjX4Mw5gIYIppmi+qGJMgCCIsiGaWL1mzB+l9UpT+YOYRalQwrSCTa79ViKaA2oBZfJb686XBiVcqsfo9lde2Cn1+bLb4fCzMieCxuZNQ1RjFK9vqUXm4o/xtdiQNmQP6IL1PClZuqXMtU+wU6BXtMztmaH8caD4diCOUFytBY2gkNZQiTpCIVA9YNGssFiwrYxLeDANahNcgOHisDQePqS8DrRPjco1bVZepNV/G5kS0J8VY81G3+CBCkgEsnlPE3Y4g9h4fe68srW5W9h1vvmoo1/tlBG6dyW+AnHgaVOuKlVvqmBMT4tdfnhYX/6/v9zHZqOm4xv0kuS9w2y+BKfd2+1OYy+uqQNUzNw+sJd+9Sue6bT+R0JUUW1rdjPuKC0hsJQiCIAiCCDmqUgf/C8CPADiKrwD+FsC/ASDxlSCIHoVoZvkzGw50+Xd8YI7XOSEiAsuQFFOWlhVVoqnOgFmiBXbsiO33VNUo1vdTxl2gqudwuwksLtmBDw62YMIVGSgek4WI4H6NyR6AR/92opZS0zooq2nBb94RK4nY0+EVUD6sbWETXtHh+H5t+2Gp/SPESO+bjEWzxmJQeh+lwep2E9r7H8eKlH72my3I6o/aY+5JJfH9/1hFU7t7vPVcsnyTutLm90wfyf0ZGYFbpqyqGzLiqerWFazPj9a4PIfRWn/rW9uY9vne5DV4JHkF+him/8LruM8B81c4/llEyOdNOgySoJy9Xq09rOonMskgiYKupNiwl18mCIIgCIIgOvCzbksywlOJkSAIQhmqfgBbgbmFMwqwvf4Et3NCVASOZVpBJibnDfTsGRkfzGVFlWiqM2AW5rKRLEzKzcDDnx3Xecz8dnao7jlsAnh5az1e3trx77HZA4S2M29KXrdS0xurjmL93vCWZSUR0BkWAaW0uhmPvb4blYzCmwnnfryEftrOX8KSNZV4Yt5krFg4zTFwPyk3AxUNJwPYQ2c+XZTd5d8s4gPLvdaL1JQk17628f3/WIXh+Hu8qoQau/0TEVOs7yFaSlqmrKrT9mTEU1WtK3idt6LCUFlNCz442OLqXPxNyq/wN8kfIMlvwRUArvob4Iv/zfRWHiFfJOkwSIJ09jq19ohPBOipxx7QmxSbiOWXCYIgCIIgeiN+PrUVAhCrfUgQBBFiVP4A9urZ6OacEBWBi0cPwS3jc7oERGZele2ZsS4SnFEpmuoMmIWxbCQLk/MG4rVvzujymp/ODj96Du9rOsX9mUhaCmZPGNb5b6t85kEf+28S6nETUESTADbtk09iIcSxxKkVC6fhsbkTu5QtLxqegTuvzUVpdTMqGnYHvKddeeLNSgxO79t5X65qjKKqMYoZY7IwJX8QAAPpfZK7iQ9u91oWKo9EOwXrx+ZOZHI7srrSrHu86oSaWPYfPYXS6mah54n5U0dgy8FWvLyVvw+4ateYqHhqleYXaV0Rn3zC67yVFYachNe7k9bjp32eQ0ogouttwBf/wPUR0YSERCAMzt741h6x9ORjD6hJinUi0Y4FoQfR/uIEQRAEQfiHsGJgGMZ/xr001zCMkTZvTQYwAsCnALwhOh5BEERY8fsHsJNzQlQEvmV8Du4rLujyGmvGuggqRFMrsH3LuBys3d3o6XxZOKOA6zz5WTZSFVZPPzv8cHZUNUZDe7zmX2/fH6/x5Fmf94RQiZOA4kcSQG/n1vEda68O2k3gay9t7dbfef3eo9ha24oRmelaxpXBBLB41Q60tp3HusomV+dh7D009l77m3f2CbndYwXr+Hu5E6z3eN3XUvOp8/jSsjL87C6xPqYTrsjorIrAg8qkORERs6ymBXc8/R4qGk4IjxubfFJa3YzFJRWen4l9fhRtSeDEJ5N2YnnKEvRLCqBn9vBrgVt/BIyaKfRx3oSERCLszt6efOx1lQZOxPLLhFpk+osTBEEQBOEvMr88/0/Mf5sArrn8f3aYAMoAPCgxHkEQRCgJokytXdk5HeXF3DLWRZHJdBctfbhsUw3GZke4grteASGVyJbR9HIFsB5zA8Ct43JQ1RjF0EgqCnMiTMF5HeUoVbLsvRq8V92MH9w+vssxyhmYFuBeEbI4CSi6eqwRHzMsI1Xr9uOFV4sPalpCu9aYJvD4m5WOf/+gpgULlpVh9vgc3DB6SJe19Gj0HFZLlBnn7YNs4XWP9+taeqjEu4+pHUGWVbVYsfmg0OdkhFfgY2GntLoZC/9rC/PnrLkyQ+Ex+N8+/4rrk6r97+k68fPAjd8BssdJb0pn0mGQJIK7tKceex2lgROx/DKhFpn+4gRBEARB+I/ME6GVWm0AOADgVwCW2rzvEoBW0zSpth9BED2WIMrUxpedC0N5MVZEMt1lSh+aABYLBnd1n9JpBZlYef90VDVGsWLzQazfexT1rWe6vS+SlmIrSLC6AljEZBPAn3c34s+XHW1OY1pZ1fWtbQnjMKw8EsWCZWVYEuOwGjmkf8B7RchgzfnYgG3bhYuhFed6En/9SE446q3Er7HWWvpvr++Wvtew9EHmQWe/QjtExeOgn3vWVwbTN3xAWgpWbqkT6ntbVtNyuRy2HD9LfgafT9nkf1/X/jnAXf8h7HR1Q0fSYdAkiru0px171cczUcsvE+qQ7S9OEARBEIT/CIuvpmnWWv9tGMajANbHvkYQBNGbCKpMbXzPw7CXF4uFJ9NdRelDE8Bjr+/GmgduZHq/zj53FrHnoDAngsfmTgLg3MNH1hUQf8w37z/mWrbZzX22YHkZYOoXp1VilQa1ghEqAxIGgMHpfdDSdkHZNnlIMjrKay9/ryYhxHBZphVk4mj0HB55dac2gchAYs1vv+hw6pP4qgJrLXXqn8mLXR9k0fuGzn6FdoiKx34999gdRwCoP949WcoP0vsm47ur+IXXjxFXTL+X/BK+kvInX52uJgAjOQ24f4MSp6tqwu7a7Knu0jAjkhwim2hJ9GxE+4sTBEEQBBEcSmqhmKb5qIrtEARBJDJ+lqm1iO8nlAjlxeJhyXRXVfqw8kiUKbjrR89It3PgdExUuQIKcyI4Gj2HxyTcVqrEAr8xY4IRKkuGmwCyM9Jw/MwFbfOmIKs/apq7FxKZnDcQV+cNREZaH9w2ebhU+dJEIMno+M66qw3cNSUPH7W2+baeJwoDUvsEvQs9CpVr6caqo51rm2xPOF39Ct2wE4+90P3c43Ycg+o/PK0gEyXlDVLr3/6mU9yf+WTSTvxXyhPom9QuPrAApgnU5NyKUd942ddxWUi03os9zV0adniTQ3634DoMjaSSQE50Q7S/uMqKGARBEARB8KO0EYVhGMkArgIwGECy3XtM09yockyCIIgwYZdZ3nbhEp7ZsF/LeG0XugdHE6W8GCuqSx++sq0ei+e4uyZUiL2fdxFuwnAOenNfzLKaFry16whmTximtGR45ZEoHrx1LJa+refY1h47jSV3TULb+Us4dfYimqLnsO2jVuyoP4Ed9b3Dieinw3dCbgbunJKr1JnYE3j/wLGgd4FwYP3eo1i/9ygKsvrjYPNpx+Qalp5wTdGz+nbUAVHBV9dzj1cFjLqWNpHdlSLJAOZNycXikgqp7ay9XPqahbFGPZak/A7XJh3wva/rKTMVX73wbfzorm/4OzAD1HuR8EI0OYTEMiIe0WoUIklNBEEQBEGoQ5n4ahjGDwA8CGCgx1ttRVmCIIieRHxm+dbaVk2lMe2jYD2pvJjq0oeVh6Ouf1cl9p46dwGPzZ0IAKE7B3738gsjX12xtdOVorJkeEZaH6xYOE2LA77dBFaVN2Dl/dOxcksdnnq7qtcJ6I/PmyTt+GKleEwWHnl1JwmvcdDhCD92Dvl4vHrCvX/A/3vEgDTxn6aqn3v8qIDBiyXQnD4v70pm+VqfTNqJH6f8J0YlHZEejwfTBFrb+2P+xR9in5mnvE+vCqj3IsFKT0uKJYJBNDkpiCoWBEEQBEF8jBLx1TCMhwA8CuAEgBUAPgJAd3mCIIjLqHTXxZLexz2fpSeUF/P7R6MqsXfNrkas2dWIibkZuPbKwciOpCrZrgr87uUXVmJdKaoE041VR3Hq7EXMmTgMC2cUoOH4mU4HPGDiaPQ8Xt5aL7z9spoWX/oRh5WDx9p8Sxz48n9+gMMn/Hf/EYRf2PWEK61uxhNvVmKfQFlaWVSID6qee8JWHSJWoHl63T7t4/085Rl8PnmT707XSybwxMUv4rlLdwCQ69OrE+q9SPDQk5JiiWAQTU6SSWoiCIIgCEIeVXfirwBoADDFNM2jirZJEATRY2AtO8VLb/hBpfo7Fg3PcP27arF3Z8NJ7Gw42eU1kT5gKgM2YcyCHp6Risz+qdh1+KT3mxViuVJWLJzW6Sh9uKRC2N1nlf20mJibge9+dtz/Z+/e46Oq7/zxv84kQy4wXEIgQGJgIAl3KFiINNqqVGpb7Sp04VsK7bpobbtW2m5/xbZ2W2ursN+2irZrrbBsi7VfrFBXbEWR4qopRiQuVyEJDAkJEAjhMhAuSebz+yOcOJnM5Vw+5zIzr+fjsVtJMnNOMsnMmc/r836/ezzWM0YNMnWMX2896KpQwE77bfz9YPCangblenHWwtnNbhM+E87JjR1aqhvtCi7c1B3iazeOwZ3TCnt8n12beawx37MVj3hXIVOx95cgJID1HTfg/+v8WvfHjM7ptRpnL5JRqbAplpxh9HnQbc+fRERE6UbWivY1AJ5h8EpEFFuitlNGWPGGym27smV/j3dOK4z7eTsCbS1zwNTHYe/Rc3g30Bp1tlyiEDfWY+nG0P7Yucs4du6yI8cOCeD7f96Nz08vQr/sTHzzllI8tllOZdGepnP44qoq3HODHz/47AQAHz4XLP1/76Pl/BXd9+nEnEGidHG6rR0r5k3Ghuom6a3D3aqyrgUng5cdraiPV91YWdeClVtqowZeRjYzJeKW7hDl/jwsu3VclM8Yf5AG5Xpxuq2918e/kbEe38xcjwybK107oODx9rn4Vee8Hh93cytWzl4kIruVFfgw05+na+OHG1u2ExERpRtZq6/NEu+LiChlxWs79eCLexx9Q2X34qZWRt5sxjJ+WPwd55V1LVhfbbwlrB6x5oDFexwixQpxEz2W86bHD6DTUf2pNvxic41l9//MWwEA6A5gK0ry8S83leChjfssO2YqGjfc16OyOJUV5+Ugw6Mg06Ng+sg8TC8eiB++uBdXOkNOn5p0IwZk48sfG4UVm/a7ouK07Uon1t07CzXNQazdVo9X9hyLuVFi/DAfhviy8GatOwI7I85f6nC0ze7npo7AkBit+RNV42rZzKSXFd0hFAB33+DH6rcDmn7O8drt5nqNv+2+bcoIrH2nvvvfz2b+FBUZ+2xvLwwAmPAPyJz/e3yuOYjBLtr0l4jR34+9TfZ29iCi1KJnjJFbW7YTERGlG1mB6fMA7lQUJUsI4UzZChFREonWdsrJN1ROLG7qIWNmrgLgwdsmxPy8E+0WI+eAGTmHyBBXy2P53uFW+PP7ItByQcJ3QVo981YAN44d2v145/aJP7OZeir35+HOaUV46o1DTp+KLSYM749vzxlraTvYkYNz8YPPjMdX1u6Qe8c6fWFmMe79xBgMzPW6Yp6xGq6UFfjw8B2T8PAdk1DTHMSf32/E/mNBXLjcAW+mByVD+2HU4L44Ebyc1OFr4xn7ZilH89LOo3hp59FeG70q61o0/T7E2sxklCXdIRQgv18WRg5O/NqbqN2umfObVNg1euFF7/cx1XPYmdB11MeBf9rY/c9ka8Vq9Oe/vroRM/15jlxHE1Hy0zrGyK0t24mIiNKRrHeW/wbgOgAvKIpyvxAiIOl+iYjShlNvqJxa3NRDxszcu2/wxzxvrT8DK6hzwMy0fFRDXACaH8vDLRegKIBwQZVZOlmxaT9euu96AMD66iaHzyZ5eBTg5nFD8eCLe5w+Fdts2tuMTXubMXaYDzXNQUv+VutPteGD4/qrscr9eRCAtMBODTOsaM9v5nzClRX4cH3JEOyoP4N3D58GAFTWnbLtnLq6XfTD2ncapN/3uu32dHxIJHKjl55q3MjNTGZYcY0jBPDoK/sTfp0/vy9+esekuOdg5vyuvPk4Atmr4UTmiqz+wIK1wOgbnTi6NEZ//gLOXUcTUWpIdJ3k5pbtRERE6UhW+LoXgBfACACfURTlLIAzUb5OCCHGSDomEVHKceINlVOLm3qZXZRf/XYApUN9USsOnGy3CHSFv6/sOW7qHKoCrVj+ivaWnQKAf3Bf1J+64HiVWTrZ1XgWNc1BAPKCq1TnUYC504pc05LWbgeOBy29/8d1zjcO77xgtiOBKvw1JVZ7/tw+GXjyb3VoPH3R/AF1nI/Kie4IRQNzsOTqxqGygq4Q3orw1U3UjV4CQvdzpLqZyWwVpcxxB3rVn0rckaKswIdJhf2xR2cb2z95f4SPnqu1v9o1Nx/4p5eBoeNtPrA1zPx+OHkdTUSpId4Yo2TqIkBERJQOZIWvHgAdAMJXA6K9rXNkky0RUTKx8w1VTXNQ6uKm1ecc7WdzIngZz75Tj0Tr4bEqd438DGQ7fOqClHPY3XRW19cHWi5gxbzJ2FDdFDXQ9mVnImjB7Lt0V1mXvO1J7Vbuz8PN44ambfBqB70/1vDOC2Y7EgDAlKIBqKxrwat7jvd43QhvRWo0+ByYk4k+mRk4EdQ+FSXaPHWnuiM0nrnY63W0OC8XDa1t9p6IzUIC+I83Dhq6bWVdi+HrjvBri7EFPmwPtOr++zBLazg37ZpBmsPXxzOexOcytsHjkXGGOigZwOINSV/pGo2ZcRiyNgkQUXpLtpbtRERE6UhK+CqEGCXjfoiI6EN2vKEyGgJFLm5W1rVg5ZbaqAFi5Aw3GcJ/NvOf3qZ5cTTaoqYbgrDms5ccO3bblU6su3dWzOA8/ONt7R1pM2/TSudtDLQV6A/X3EABsHzeZCyYUYz5T2/TvcA9cnAuLlzuQMv5K5acXzr7SPHA7v8uGpSL0gKfqercXY1nsaux58aR8NcNvcHn1KIB+FhJPu6cVoiyAh8q61pMz1N3sjtCZV0LTgYvx3yNTVX1p4wFzIcNzDKPdw3jBC3h3FBfVsL7WeVdgdmenbZXugoAyoR/AOb/3t4D20gdh/HA+t2GXmPNbBIgIiIiIqLkIKvylYiIkpDRECj8dokqkiJnuMkko3LXyEKtbAX9sx07tvpYxgr7wz++ppIj3WWINk/SCh4FuGV8AV7d12zL8WRRANw+dTimFQ8yXJluNLihxNTQwMo2vOGvGy9UN+o6RrY3A8tuHdf9b7Pz1J3ujrDt4Ck8/PI+Vn5r1HxO32YmJ9pJa5EonIv3OlKqNOK/vd9HjtJha/AqAFzOHoLsB+rsO6iDFswoxvbDp/HCDv0zk+3chEVERERERM6wZPVPUZRBAPoJIY5Ycf9ERGReTXMQe4/qa1OrUhf9tFYkxWr5a9babYcN3S58UVPvQq1s44f5MCq/r2PH33v0LJ7cUqupTbRbFgvnTCjA6x80u26xXCs7Zr31z87EU4uuRU1z0LHwVVGAb32yDBfbO7H/WFd7zHHDfRg1uG/MVtdA1wL+SzuP4aWdx1Ccl2vjGZMWh09dsKUNb0gAyzbsgtB5jGhVg2bmqTvdHWHzvmbdlXWTC/tjt855oKmiYID2zUx6fo8VBVhUXoyhvmz0y85E4cAcfGXtDhNnGl+i19tov6s/zliDxRmbkWF3e2EA7dmD4Z3/n8hOwRbD8Uwc0R8vGPg1sGsTFhEREREROUfaVb+iKP0APATgiwCGoGvtLPPq58oB/AjAg0KIalnHJCKiD2mdtyqjvZ666KenFaPWOWZardvegGffaUj8hVGEL2rqWai1wvnLHbaEcbFs2tuMTXs/DOfitYl2y2LhrDGDMXv8UFdWKyUycnBu99/lTH+eZRV1IwbmYIgvC0M0tKa0wvAB2eif48UvN9f0+PjWAye7f8eG+LLwq7/VYuPOYzHDpVSfbZmMms9e0t2Gd2COFwX9s3GgWV97Yr3Bq+rRv36Aj5cN6fE6aHSeutObTvT8CEYOzsUzX/ooKutasLtpn2Xn5GajBmvfzKTn91gIoKb5PB6+Y3L3x6x8Dk/0eltW4Os+/s8yn8EXMrbCY3N7YQDAwJHAwnXwDh3vwMGdZ/T6zcnrPiIiIiIisoeUVVRFUQYAeBvARAD/C6AFQPg7sN0AbgDwBQAMX4mIJEo0b3Xe9EK0XenE+UsdqD0RjBt0aFHuz+ueBWq25a9RarWK0e8jfFFTz0KtFY6cvohH/voBBuV6cbqt3dFzAeK3iXbLYqEalsSrZLPagBwvzl408HiJrvbNFSX5WDq7VPMsSr32Hw9izmNvYqY/D2OHmZvJqdct44diy/4TOBZjlrH6O7bkej9e3mXu+YiMKx6Ug4bTF3Xf7lJHp+7n/jMX23HGyN+LQVsPnMTWAycB9N5Qoneeuls2nWihttt2OjB2kpbXqZrmIP78fqPpaxgrn8O1fB9LZ5di6u/Ho69yxfa5roAC3PIQULHU7gO7SngIrpV6HU1ERERERKlN1mrCD9AVvP6TEOL3iqL8CMC/qZ8UQrQpivI/AGZLOh4REUHbvFWZVRkeBbh/dikA460YE80x00Jv1VWk8EVNNwSKb1wNCdwiVptoI4uMsoUvWkZWsr1Z82HgYjVDwSuA+tY2PLSxqyJtpj8PS673Y/XbAcsqeN8NtEJRuuao2hVyvv7BiYTHCgngmbc4Q9hJE0b0NxS+vnHA2Ta8epmdO+7Ua4TRv9nKupakCoxlShRqyej8EX4No3WesF6awrnKlajY/G+AAy2GMfom4EsvOnBgd9ITwodfRxMRERERUWqT9XZtLoBXhRC/j/M19QAKJR2PiCjt2TFzL5xHAZbPndK9EG20smbbwVOmzsNIxW248cN6Vj2pgaIe5f483bdJNmqb6EhLZ5c609oQsRctywp8uKvCj+nFgxw4K+PeDbRi9dsBLLnej3ILf5/U1q12PWysZE0O44f3d/oUbKNuKIm1aaimOYg1lQE8uaUWayoDqAlrj2zkNcIsj9I129qI85ecbWXvlESh1rrtDVi8usr05qHIa58FM4qxdkl5zOfwcn8evvfpcZpfNxOGc/v/Ajw0GNj8b7G/xipFM4Efn3VN8Brv79ZOagif6DGOvI4mIiIiIqLUJmtbdBGA9Qm+5jyAAZKOR0SU9sxWf+pR7s/D/RFzQI1W1mze14zKuhbDi09GK25V+48H8fz2Iz0qoIxWLSxaVZXSQVO0NtFmK30+f20hJo4Y0GPe4it7juHxzbVxf5ZaFi2TsdorJIDVbwewdkk5hviysGhVFU4EL0s/jkDXxoP+OV5H2jRTdOOH+bD/eNCR55EN1U3I79cHLeevOHB0+0WbO56obb/artjK9rKR1NfbmuYgXt3XnPgGEfplZ7qiS4GdFADf/GQZapqDqK4/3WuWr8zNanuPno36uphonvDAXG/Cc4j7Ole5Evjbz4BO+a8PCWUPBB6ot/+4MWj9u7VKtMd5wYziuKMQol1HExERERFRapO1ShkEMDTB1/jRNQuWiIhMMlv9qdWtEwvw7Tljo7a/M7qAJNB7AVwPs7PsBHq31NUaKIYvjK7b3mDqPJJFtDbRiRYZ49lQ3YSZowbjrgp/98fKCnwY1j8b//HGwe6ZheESLVqqC6GHWy7oOhe3UEOhdffOwmMLPmJZqP/B8SB+u/haTB85EPuPBXHw5AU0tPb+eZM9vnbjGCy7dVzC9vFA13PP5KIB2HnkrLTj16fhYx++oURL2/7wdsVWtJdV3TR2CD5eNqRHWDfEl2XovtTnSTsD40hFg3Kw+LqRWL5pf3flvdbbNRpohT24Xx/8cnNNr4+rQZzMzWqb9jZj097mqCFfvHnCpsK5394EHK2W8w3okZENfHEdMPpG+48dg96/W5m0hL7r7p0VN4QnIiIiIqL0ISt83Q7gNkVRfEKIXv1+FEUZDuAzAF6WdDwiorRmtvpTq4kjBsRcMDJTWROtolIrGdWN0Sqg9CyMqlU0qVz1qooVdkdW+uw9eg4v7GhMeH+R82Qr61qw/JX92N3UO1QqzsvFv9w0BgtmFEe9Lxnz+9xC/ZuoKMnH7VOH46Wdxyw5zlfW7rDkfkm/O6d1TePQ8twzOr8v/rj9iN2nmJIq61pwMnhZU5Aa/nxlZtNJItOLB/XYkAIAJw1UwPuyM3EyeBllBb7uTUUPrLf/tWrJ9X7k9snQ3IP8c1NH4L6bS1BZ19I9E1uPWNXb7wZasWh1la4AWCsjIZ+WCtkefv8PwKE35J64Fpl9ga9sAYaOt//YcWitYI41t94MvaEvw1ZnMPgmIiIiIjeRFb6uBPAKgL8qivKV8E8oijIewDMAsgE8Iel4RERpzWz1p1axgk51caM4L9dw8BWtolILWQtpsVrqalkYtbPls9MShd1qpc9NP39D832q4fcbB07gmbcCMb+uobUN39uwGwqUXovbWqoFk436N3HfzaWWha/kDuX+PM3PPe83nMay9bsdPNvUcv5Sh67n8PDNOtEep71Hz2LTXv3tgcNFe55dGWXmdiLBSx09AqCiQbmmzsuo3D4ZmjcoKQAWWBhWWRG8qoyGfPEqZAEAf/wCcOCvEs7QgFt+AlQsdebYCRj9uzXLydCXtHG6FTVRsuKGBSIiImtJCV+FEK8qivJjAD8GsAdAOwAoitICYBC63lcvE0L8XcbxiIjSnV2zLSMXKmRWGhoNkGXOsosVAMdbGLWr5bNbaFmsWre9AQGdLX+rAq2aqseiLWbKnN/nJurfRLrNa0w34XOjI4U/99Q0B7F2Wz3WvuOeWYupoK29Q/fflrpZB0CPRcpPTRqGT00aZjp8jXyeNfM6E/6cuXJL/FnaVigamIP11U2an5/DRxEk43OfzJAPz38J2Pff5u9HJyGAk77xGPqdd2w/tlZG/ibMdFkJ51ToS9o42YqaKFlxwwIREZE9pK3eCyF+oijKWwDuB3AdgMHoej/9VwCPCSH+JutYRETp7tyldsuPEVmZJbvS0EyALGuWnd4AuKY5iF++dsDcQZNI5O9ALL/eetDS84hczEzVyuPwvwkn5zWSdRR8ODc6llRqp+1OiqFb3f2796LOSJ7pz8O4YT7sP95r8oom44f13uxjdrRASAA/fmkPak/YPwd7+siBuiv3w0OyZHzuMx3yvbECeOMRuSelgRDA+52jMbfjp/jXj5fhG7afgXZG/yaMdllRORn6WiHVqtxYlUykHzcsEBER2Udq6ZQQYiuArTLvk4iIeqqsa8HK1/W3I9RDQc/KLCsqDc0sfqiz7Myek9YAOB3DkHjVeeFqmoNRAwnZqgKtWLHpA7Rd6UzZxyEYthlA1u84ucui64rjLmSlYjttNyn35yHXm2HotrGe58w+H11q78SaykCPMETGaAEnglcAKByUY+h2akiWrM99hkK+P38V2PlHa04ogZ2h0VjR8X/w99AkAD2vh9wY0Bn9mzD7t+RU6CtbMla5afk9ZFWys9z4XEHxccMCERGRvezpW0lERNLYUfV3+9QRPd5oyT6m1orKeBbMKEbRoFzctWY7rnSGDN1HrDeT4YsJtSeC2LjzmO2tG52kKImr81RmK7T0eOqNQ7YdywmPv16Da0cO6v65q7/jT2ypjdqiedwwHw40By2daUhyLZ41KubnUrWdtluom4rU9sFuETjVhoc27uvxseI8Z2a1ytB2udPQ7cJDskTPfUUDc9B45qLhc7SCrpDv1R8A235l3cnEIARwVuTgH9sfQq0o6vG5ipJ8Vwd0RrulmB3T4VToK1OyVblp/T1MtarkZOLm5wqKjxsWiIiI7CU9fFUURQEwDIA32ueFEA2yj0lElC7smjd6380llh1Ta0WlFkN8WYaD15F5uVFbPaZbhWuk0fl98fAdkzS/0XbTAmOyi7bIUVGS373IGK26gJWSySPRppNUbaftFgJdP+N50wudPpWE7OgmYJXmc5cM3a6tvedrSbznvj+/34Sn3rC23b1emkK+Q28Av58LwFhAbUZIAI92fAHPdN7e63Pl/jy833Da1QGd0cV/s6GBU6GvLMlW5aYnKL5wxdj1p9uqkpNNsoX59CFuWCAiIrKftHcFiqL8I4AHAEwGEKufl5B5TCKidGNHlWFkQCDzmB4dFZVamDm3r980pse/rQ6xPjd1uKsraEfm5eLrN43BghnFmr5eXRCvbjht8Zkll+H9s3HqwmVc6TT2SMda5Cgr6D0XEkhcIUbucfO4oTE/Z9fGmnT3bqAV7x1uhT+/LwItzrTlTXUFA7IN3e6pNw5hR/2ZXhVT0Z/73PdKmvC65qfDgQ77Q/X2kIKVnXPxq855UT/vUbqem9we0JUV+DDTn6freVJGlxWnQl9ZkqnKTW9QPHdaUfwvjIGbBo1LtjCfekqVNupERETJREoQqijKvwB4AkAHgLcBNF39byIikmhv0zlL7z9aVaqsRYpyfx7ul9yGyui5Dcr19ggZ7Wj3WTrUh+Xz8l1XpTgyLxfPfPmjmt9Uszo4vta2K4aDV5XeRQ61QmzFpg9SvjVzMlv+yn7sPXoW991cGrXqnuwREsDhUxegwI0RXvLL7WP87WVkxVSsiv9cr7FjfH56EY6cbpO+UWXcsOibYwAAv5wInGuUerxEhADahQefbV/eq71wOHVD3AvVjboDuiG+LNtnPS6dXYrFq6s0nausLitOhb4yrNvekFRVbnqD4u31xv6O3VKVnIysCvM5O9YeqdBGnYiIKNnIuvL8FoATAD4mhAhIuk8iIgqzbnsD1lcbW8CbOKI/9h09F3ehOVZVqtFFilsnFmDiiAGWvok2em5jhvTrsbhkR7vPftmZCasUB+V6cbqt3doTCeNRgEfmTtb82KRzi1utQc3lDmNtsMMZXeS4c1oRw1cXEwBe2nkML+081msmGhe27CVEV4v1w6cupOXzmbXM/UBDAli2fhf+szKA/cd7z+ed6c/DWIPXExML++Pn86d2L/TvPXoO2wOtqDfZ5lmJ9sF1i4EPXjJ1v0YIAWwJTcXd7cvift3IvFw8Mncyhviy8N31u3QdoyrQijmPvdnr41bPeqwoycejcycnvA6R3WXFidDXrHXbG/DA+t2GbutElZuR7g/1p4z93bIS0xgrWtZydqy9kr2NOhERUTKS9SpaCOAZBq9ERNZQKzONLmk+tuAjOBm8HDP0i1eVavRN74QRA/ANixegjJ7be/WnMeexNzHTn4d50wttqeJUzzXeHLtX9xzHLzbXWH4ugP7FSTuqg60yb3oh/vx+k6lzn3dtEY60Rq+YumZQDo6cvmjiDHtqazc2D9BIhU6kAdmZOHepgxWBFous8OPClv0OtVzAinmTsaG6Kerf9cjBuYYX92Vz07nEMzIvF22Xzc8zFUDU4BXo+tvZbvA5Tn29i2xlHP563NbeoXsTywfHgx8GDIfeAH7/D4bOzwwhgLrQcNzS/gtNX692vFhTKe/tux2zHhNtYrOiy4pToa9RZt8zOLEZyGj3h+K8XF0zst1SlZwswp8b9x49a+g+YoX5nB1rv2Rvo05ERJSMZK30HAGQJem+iIjShtY2S2YqM9WFhrICX8zQL95ChNFA5/HXa3DtyEGWvmEzGzaZWcTVI9piT7Q5dna1HjWyOCmjOtipNp+TCgcgr28f/H5bveHK1Ikj+uPn/zg16t/PvKf+LvV8dx05gye31BqqGje7meDspQ58/tpCbKg2F1ZTYpyJ5ry2K514+I5J+PP7jdh/rCvsGze8P+6cVggAUav7nPD56UX41KRheOSvH+CNAyedPp2Y6lvb8Ltt9ZYfx8hTU7zQJfz12GgYWVnXgrJnyhyZ69oW8uLujv8Pfw9N0vT14T8L2UGbHc9r8TaxWRWsORH6GmX2es2JzUBGfw9n+vPQeLotqaqSk4HMESPRHlvOjnVGMrdRJyIiSlayrqz/C8BXFUXxCSGib1MmIqJuetosGWnzpIq20BAt9EtET8s1lZ5ZP2YYObdwVudLehZ7rP5ZzRw1CD+9U3ubYZWZ30HVuGE+/PC2CY7Mi135ei3OXDTXzjlWxdRre48jKHnxuvLgKVQePNX9bz2t19qumK86W1/dhAduHYe/7D6GXY3GKh1IG/V5ct29s0xXLZN+q94K4KGN+3p8bOuBk9hRfxpLZ5e65jE5EbyMB1/cY9m5KABunzoCt00ZjqYzF3uEWe83nE7argcqPa/DRkKgVd4VmL15p+7byXDymluxOPgvMSuFI0X+LKwI2uy6/jNyPWuGE6GvXjKu15wIuYz+Hk4c0R8zRiVPVXIykD1iJNpja9XsWEosGduoExERJTOPpPtZAWA7gNcVRfmEoijuePdBRKRBTXMQayoDeHJLLdZUBlDTbO0eknXbG7B4dVXMxRG1zdLz248AMF4NqUDeQkNFST6WflL/my911o+V1HZwnqhD15yld7FH3ZFslXcPn8aDL+5J+DsV+Tfx5/ebTB+7pjmIptMXsdSBN/Fmg9d4u743GJzDrEfkc0I8MqqYhAB+ubmGwatN1OfJpbNLXfk8lsqazkRvF67+zU0tGhB9nqfNnn2nPm6gogAoHJhj+P675hEfxVfW7sAre45j+shBuKvCj7ICHxbMKMbaJeUot/C1yQwlwQOkKMBtU4ajuv60pms8PSFQqdKI/X0WY7Znp/2/J0UzgR+fxZAl67Dpmx/Ht24pTXgO0a5JrAo07Lj+c0pZgQ93Vfjxjdml3X8nbmG2g4pTVW5m2qEmeo4q9+dh7ZJytq7VwIoRI5GPrZnZsWSe1vfN3LBAREQkh5StrkKITkVRfg3gTwD+BgBK9HfCQgjBoVZE5Ap6qk9lHlNvmyWjYcq8a4ukLjT0z/Yaut2v/laLJ74wXdp5RJOoHZwTjLagWzq7FF9cVWXRWcWfoSSzzVgk9Xd62a3jMLhvH5y6cEX6MayQaNd33YnztpxHvNZrMmZyRTLanpmMqaxrwV0Vfk1zBckeIdFVGesGiX4dBGIHyXpFe40Ir/i753fvoV7HjEWrLSovRk3z+aiv/b7sTAQvdeClnccAHOv+eLxrPC2v2fM9W/ET738iWzHfaUC3opnA3Zt7fXjp7DJ8dGSe7ra4MmaFxxJr1iNZx8wGLCer3My2Q02GquRkIGPESLhoYb7RDQJ8PpEnmdqoExERJTspQaiiKP8A4AUAGQACAI4CkNuDj4hIokQtleIFVEaoiwGr3wrobrN066Rhho45cUR/Q7eLxeiCzsadx7BgRovliyLhCy+/fO0ANu1tlnK/eigAFl03EotnjURZga+7glTP9zvEZ/0I9WhBnuw2Y7GO++gr+607gGSJdn1X1rWg7uQF284nsvWalWE52Wtv0zkA7txIks7SNQMPCWBZjM0ebgpeAWCoLxsP3zG5x/VF7YkgNu48FrMlfLxrvHgh0Mc8e/DrzMcxyOPAz2DsZ4Av/DHulxgNoMyOb4hF9jxZSsxo+16Z3XKMktEO1e5W1KlERsvqcLEeI6PPC3w+kYsbFoiIiOwhqwr1xwDaAHxWCPG2pPskIrKEkepTo4sRZsORqkArllzvN3Rb2QsoRhd0BICfvrwPvhyvLVXGZQU+TBwxwJHwVaBr8eJknPl8ib5fsy3jtAoP8qxoM5bstOz6Xrml1sYz6qK2XkuFOYz0ofXVjZjpz8P8Gdd0L4i9tvc4vrJ2h9OnRg5R4Gz4KwTw1Wd34DeLru2x4cNt1GsTNXSprGvBY6/XJPzZxbvGiwyBvpSxCd/NWIe+yuWEbY6lyx8L3PeurpvoDaDUNpSyX1OsmCdL8Rm9ll4+b7LjbXm1/h6yHao1ZD6/x3uMjD4v8PnEGtywQEREZC1ZM1/HAvgjg1ciSgZ6WiqpAZURiWa7atV05qLuOaBWzGwys8jxwfHYu6n1zLPUysk36FWBVizSMdM3kp07u9UgT3absWT35Y+NxLp7Z8X9nZddIaDH2m31DF5TjEBXEBS++Dln4jBLZ0CTu00YLrd7hRHBSx1YtOrD1yujr09f/thILL6u2JK5qJHP0zKu8dQQ6JHMZ3AoayF+4v09+nlsDl49XuBL/607eDXKitm+DMfsp1Zu61Huz8OCGcUWnZE+nN/qHFnvPxI9Rmbm+xIRERElG1mr0y0AkmN4GxGlNSOBiRpQ6QkzZVYSnr/UIaUVVzxaWg5ZORcsVgWK0VZITr9BFyaqqtva7W2rde/a9xBocVcbSbP6ZWXg/GXjs/gaTrVhTWUg7u+bkxVgf9t/gsFrCopsKw1Y1w6U3G/vsXNOnwKADzcGFA7KMbyxadTgvrirwo9JhQOkbhyJ3Ggm8xpvwd9uhMg8ZUlgnNCEfwDm/972w8ZqQ1k4MAc/+8sHulpOFw3K6X6dZFWVvax+z2A1tkN1htHn91snFmDiiAGaHyOz832JiIiIkoms8HU9gFsVRfEKIdol3ScRkXRGA5PKuhZdb/pkVhKqb2ataMUVry1ytPa486YXWlbtF9kGV895RbIyKJYlJIBl63dhyfX+iMUKe5d6Uy14/fz0IqyvbjR1H1sPnMTWAycB9P59UxcDX99nf1trVdOZi44dm6wVGQRZ1Q6USA/19fnhOyYZur36/JlopvG4YT4caA4m3MAERA+NpFzj/fmrwM6u2aq2B69FM4G7N9t91F7CWzgbHZ3RePoiHtq4D4D88RJmpEOglyrte9kO1V5Gfw++PWes7scp2TcIEBEREWklK3x9EEA5gD8pivJNIcRhSfdLRCTV4VMXLL+d7HakWhcttcyoDLdue0PchRm1Pe7yuVNQOCjH1OxaraoCrVi5pQYrX48dXoef10eKB8ZcREuGirFoi5O53gyHzyp5KQqw5+hZqXMS1d+3Jdf7sbPxrKsDfUoNkZt91Of+n768Dx8cDzp4ZpTO1OsOvRubJhf27/H7nKiqLdG1CdD1XK+GRuH3U91w2tD3dvjUBeDVHwDbfmXo9qaN/QzwhT9advdGAkctj4NW4ddtTrWLNbupL9mobYT/Y+vBqBXLet8zUOqzsyI1VTYIEBERESWiCC1bixPdiaIcAuAFMOLqh84AOBvlS4UQYozpAyYBRVF2TJ8+ffqOHTucPhUiCvPVte9h0179FWu3TizAbxZ/VNPXrqkMdAdqZpX787Du3lm9Pm52535lXYvmYFK5+v8kvFxoogCmwrPwRTQti4ceBVhyvR+r3w44HtQqAG6fOgIv7Tzq7Im4RP/sTPTJ8KDlAicbqAbmeHHmIpuMpLJ/vaUM34hR5bFySw0e31wrdYMBkVY/un0Cygp8ujc26Q23KutaYm40U40d5oMCYL/JDQn/mLEVKzKfgceJ/sKjbwK+9KIld13THMTabYexdf9JNEbplhDvMdFzjaiHRwHWLim3PUjRei1oJhx2U0VtvKB5ZF4uvn7TGNfMeCV30fO3L+PvOd5zPTcIEBERkVtce+21qK6urhZCXKv3trIqXz0AOgA0hH0s2ltYR8bmEBGpCvpnW36785fkzOyM12bJbCsuPW2RRff/s4fZQ4VXWGitFh7iy8Lxc5fw8s5jjoYaAmDwGuacpL+lVDJycC7ONEbb30Z2GDk4F5+fXoR+2ZnYUN2E3U3yH4t4c5+Xzi5DdmYGHn1lv/TjpoNMj4K7Kka5YrNNMtp79BzuqvDrboWtt/KxoiQfjafb8G6gNeZr8gGToevHPHvwrPcRZ0LXnMHAskOW3LXWVsHvBlqxaFUVbplQgFljBvcIC2WOzggXba611SrrWjT9robEh7ON9Zyf2ypqEwXN9a1t+N6G3VCgOFaFTO5ld0Uq5/sSERFRqpMSvgohRsm4HyIiq43K72v57fplm39qtbLNkuy2yG4UuYgW6439yeBlW9opE8nysZJ87GT46hgFwPSRgzDElyWtw0Gkp944hB31Z7oX7SOfu/7f9iOWHDcddIQEnnkrAH9+Xxw+dcG2jg6A+a4ObrD96mtloo1N0egJt9TAzKqf19vef0Gh5zQUJ4LXW34CVCy15K71tgoWAF7b14zXrs4wn+nPw7zphZZeE0XOtbaaniBZbzisZ3yHHUGn1UEzpQfZY2604HxfIiIiSlWyKl+JiJKC0TeKem5n9s2o1W2WKutaLLlft4lcRIt8Yy9znhmQGgvr5H53TivEjvrT3DDgkMOn2rBoVRUmjuhv6XHeDbRi0eoqjBrcF4EWY7PKKTb1Z1o0KAeNp3u3ZLXCJ8bm438OtCT160R9a1t3cKZubLr9ybc1V4BrDbesqrx8y/svKHIqdJ36BeDO31h291qDt3jeDbR2B+xWipxrbRUjmw21hsNuDDqtDJopvbAilYiIiEgOS8NXRVG8ACYBaBNCHLDyWEREWpQV+DDTn6drMabcn6frjaaRYxQNzMGSG/y2vKmV1RY5kXJ/Hs5cbDfdGtCMWItoMhYpw6mBOatoyUrqc9HS2aWWzOMjbQSAPUfPWX8cAVcHr1OKBmBXkldhN56+iFGDc3FD6RB4PMDv/l5v2bHeOJAaG5/Cg7Oa5qDu1tuJwi0runP8xftdTPA0OhK6npzxr3h54OKu8KIyYNl1nqzA2o6XFbuuQ41uNtQSDrst6LQyaKb0xYpUIiIiInM8Mu5EUZT5iqI8ryhKXtjHxgDYC+A9APsURdmgKAorbYnIcUtnl2qe8RVv7qrMY6z4/BTcVeG35Q2ujLbI8dw0dghe+9bHse7eWfi32yZo/jlYtSYabfFNZlWNAnRXKut53Cl1VYwZjFsnFki9T/X3TDVysLEW6kQyeBRg2a3jMNOfl/iLXe7wqTb8oaoeE4cPSInvx2rhwZmZcMvI5/T6mGcPDmYtxMQM+4PXM0NmYP7wTZjx1rV4aOM+/GJzDR7auA9zHnsT85/eJvX7TLZxElZfh6oOnzK2eSXR7cwEnVax4m+RiIiIiIjMkRK+AvhnAOOEEOHvQn4BoATAVgC7APwDgLskHY+IyLCKknw8OndywpDMzNxVO45hlNXH+njZkO4QWc/P4Vu3lFlyPpEVFrIXKQW6KhqAD79f5q/p7ZMTCjBxxACp93n71BGoKMnHuu0NWLy6ypKKSP/gXNw0dghuGjsEt08Z5kxrTnK98NetVNlworYGnT1uaEp8P1YKD86MVjDGu52sqsh3vV/BH7yPIMPux3PUDVj32d2Y3vitmNca6izQ5yXNb062AM2ua97ms5csuZ0bg04r/haJiIiIiMgcWdtOJwDYrP5DUZT+AD4D4HkhxP+52n74f9EVvj4j6ZhERIYtmFGMokG5eGJLLaqiLI7JmLtqxzH0CJ/bU5yXi4bWNkuOE/n96Pk5vF3XIr16I7LCworFr6pAK1Zs+gB3Tivq+gAHwLqeosCyeZoVJfnSf89umzIcP3xxN559p8GyX63AqTYETlnzvECpIfJ1S91wIrONu1NCAvh/7zakzPdjlfDXeKMVjPFuV3PCXHXg37xL4fectH/zSMFk4Gtvd4010NAWXuYs0GQK0PSO8jCjoH+2JbdzY9Bpxd8iERERERGZI+tqewiAY2H/nnX1vv8fAAgh2hVF2QzgC5KOR0RkWkVJPipK8nuEkv2yM6XO47LjGIlU1rXYOov0wRf3YGlEqKz152DFLMvIRU2rFr+eeuMQnnrjkCX3TfIJAWRleizJye//4/u4q2KU1Pv8ytodUu+PyIiH75jU67Ur0QabZBI41YYnttRh2a3j8Lf9J5L++5EtMjgzGhrGul1lXQte3nUs6ucScSx0BYBbfgJULAVg3SzQeNdPyRKgGR3lYdSofGPt+RPdzo1Bp+y/RSIiIiIiMk/WO4AggPD+ep9A11rm22EfuwTAnqSBiEiHsgKf5UGoHceIZt32BlMVPOoapp6bq+30ls+dgvkzrunxuUQ/B9lVVOOH9T5esixSkvX2Hw9CgfxC5f3Hg1i2fjcG9+2DUxeuSLxnImdV1rVEfQ6PtsHmTzsaLeuwYKWmMxex/JX9WDFvCh6+YxIq61qwt+kc1lc3OtLQYOTgXNS7oCI9WnBWVuDD5MIB2N10VvP9xKt8XLmlFkLnD/lF7/cx1XPYmdB16heAO3/T/U8zs0Bj/UzibaCb6c/D0tmlKByYo++8NZL52phMYzYS3c6NQWdZgQ8z/Xm6fv/srEImIiIiIkpHsma+1gL4tKIoWYqi9AHwjwB2CSHCe+6NBHBC0vGIiCiByroWUyFmuT8Pz95djuXzEs9sjaS20zPSenXBjGKsXVKOcn+e7ttG+uB4sHseK9D1M9lQ3WT6fil1CABQgHHD5C9AMniNT0HXBglKHrE6B9Q0B7GmMoBX9xwHAHxq0rCknn8t0PUadjJ4GRUl+dh+uNWR4PVjowfjM5OHO3DknhT0Ds4q61ow/+ltuoLXeJWPeoPL32T+HIeyFuIjGQ4ErzO/Avz4bI/gFZA/C1Sd8Z1oduzLu44aOm4iy+dNlnItNm6YD2uXlPfakGc1NZDUQ0sgadX9mqVnBrfdVchEREREROlIVvnPbwGsQVcI2w5gFIBvRnxNOYC9ko5HREQJ6Gl9B3RV13x+elHMtsj/sfUg6nVUMant9Ib4snS3XFYXeBetrtJdBRPpl5tr8Nfdx3BDaT5Wvx3gHD/qRQhgQI4XJUP7ou6E/Bmw1JtaBSUg8MCG3ab/zskeaucAtcJ1b9M5bD/cGvW1YaY/D3ff4MczbwXsPk0pQgL46rM7EHRwnubfD53C3w+dcuz4qkXXjewRnBnpqhGt8jG8UnrvUW0h7n9lPoJPZOxxptI1dzDw+f8ERt8Y9dMyZ4Fq3UAXEsDGncZaNcdT7s/DghnFWDCjuFfL4w3VTbpC9wE5Xsfa2+oZZ6EnkLTqfs3Q2j3GiSpkIiIiIqJ0JCV8FUL8TlGUsQC+cvVDv7r6fwAARVFuRlcg+x8yjkdElKpkzYY10vqu/lQbPjVpWI/jmZ0XWxVoxZzH3uz1cbVVXryFHyPtB2PZfzyI/ceDcu6MUhJnO9qn3J/XvRC9WMIGC7JPbp8MzH96m6bXhHcDrXjvcCtmjc7DtkPJ+fflZPDqJkN9Wd3/baSrhvo3r77mG7m2uCdjI76f+UdnQlcA+NyTwPQvxf0SmbNA9Wygk/0UGhkWho+MqGkO4qGN+3TdX6LWynrovU62KpDUer8KgG9+ssy2oDPRDO7Iv0UiIiIiIrKOtMF3QojvA/h+jE+/DWAQAJaTEFHKMhOcapnppWehxEzrO/Wczc6LjefdQCsWrarC7VNH4L6bS3r9nIyEx0TJICvTg8sdIadPwzFTigZg3b2zAADzn97GSvQkkt+vj+7XhJCAtOB1ZF6uru4LJE94OKi3q8bkwv7df/OA/muLUqURf/V+F5kKnAleNYSuKlmzQJ28BkoUQsq4vjR6e6PXyVYFkonuF+gKxn+5uQZv17XovpY3KtoMbjMbOomIiIiIyBhp4Ws8QogrAKQMPlMU5fMAPgHgIwCmAvAB+IMQYpGB+yoC8BMAtwIYDOAYgBcBPCSEOC3jfIko9ZkNThMtRKozvZbPnaJ5XpbZ1ndm58VqIQC8tPMoXtp5tNfPyejiHpHbpXPwCgC7Gs+iprmrCp0bLJJLy3n7ZxjfNHYIPl42pPu1IVonBbKe+vM3EgrubjrXXfWo59riY549+F3mI/B6jJyxOQKAMus+4FM/03U7dRaonp9R5CxQ9WfkBC0hpMzWylrJuE62KpBU73fllho8vrk2ZhWykWt5s8IrlomIiIiIyH5SwldFUUYBmADgf4QQF65+LBPADwHcga6K1/8rhPizhMM9iK7Q9TyARgDjjNyJoihjAPwdwFAA/w1gP4CZAJYCuFVRlAohhPNDlojI1cwuCOmZ6fXAhl0oHJSjade82dZ3eitbzIr8OZlZpCOinhTIb0tpBjdXkFbTiwfhrgp/97/1BltknhoOVta14PsGQ0G16lHrtcVO7z+jv+eS7ZWuQgBv5MzGTQ9sMHwfZmaBmu04ovW53qMAj86djLYrnbpDSJmtlbWQfZ1sRSBZWdeCla/HDl71niMREREREaUGWXuJfwRgLYDLYR97EF3h62QA1wF4XlGU6yQc61sAygD0B/A1E/fzH+gKXu8XQtwhhHhACHEzgMcAjAWgb6szEaUdvQtC0QIHPSFnSABPbKnV9LVmWt851e4u/OdkdJGOiHob1LeP06fQw6q3Ath79JzTp0FJIPy1oKY5iLEFPjg18jMdqeHguu0NWLy6ynDb5/OXOjRdW2zO/FcEshZiQIa9wasQQFPnQIy58hz6fP63pu5LnQXqSXD+ke19ZXQcuX3qcM3HXTCjGHdV+PGN2aW4q8KvOZCU1VpZK6uuk2VKhnMkIiIiIiL7yQpfZwHYIoToAABFUTwAvo6uatJidFWUXkBXcGqKEGKrEKJWCGH4ramiKKMBzAFwGMCvIz79I3Sd62JFUfoaPlEiSnlmF1uMhJxVgdbulp3xqK3v9AivbnGK+nNiRYD7TSkagMXXFaNoUI7Tp0IJtF6wv11sPE1nLmL9jkanT4OSQEVJPirrWjD/6W2Y89ibWPtOvauquFOZGtIBMB0K9svOjHttsc37NQSyFqI085jtoWsw1Af+y8/hho7/iDvrVI8FM4qxdkk5ymNch5X787B2SXmPjigyOo7cd3Op7uPqZeb6Ui8rr5NlSYZzJCIiIiIiZ8gqLSoAUB/2748AyEfX7NRGAI2Kovw3gBskHc+sm6/+72tCiB7D14QQQUVRKtEVzl4HYIvdJ0dE7mdmsUVdgDIacqrt+xIx2vrO6Za/VVd/rmwv2WVkXq7haiOreBRg2a3jeswBVGeonQhewrNVDTC+RYrSAX89KJFyfx7ebzht+fxx6m3cMB9+eNsEVJTkY/7T20z//CtK8vHqnuO9Pv5H70O4znPA9vbCANAhgBUdX8AznbdrmnWql54ZozI6jqgBZ1mBz5LZpuHMtFbWw+rrZBmS4RyJiIiIiMgZssJXL3quo1Vc/fffwj7WCGC4pOOZNfbq/9bE+HwtusLXMiQIXxVF2RHjU4Zm0RJRcpCx2GI05NR6O7X1XaKF68jWd25o+VtZ16Jrcc9K2ZkeXOoIJf5Ci3z9pjGuCh8if1+AnjPU5j+9jcFrHL7sTAQ509gSN5Tk4+26Fga7KUBRgJvHDXXVc186GZDjlTaGIFpXjY959uBZ7yMJW+RaIQTgb1Mfwwf9b8CI7Ey8JjGUjEbLjFGzHUeUKAFn+HHVIPbVPcelBLFGry/1svo6WQY7ztHKIJ2IiIiIiKwja4W9EcCUsH9/BkCLEOKDsI8NBeCWAV8Drv7v2RifVz8+0PpTIaJkJGOxpfaEsZZj0cLRWAszC2YUo2hQLp7YUttdURouWrWHG1r+nr/UoXlxz2qXOkL47eJr0XTmYvfPN7dPBjZUN0X9mcpU7s/DghnFqDtxHs+8FbD0WFrPJ151kFPzgpOFogD33VSCFZv2M1CywFt1Lcj2enCp3bnNEmSeAmDF3Cn4z8oA/04conbqMBsKhlc9qq8b7/e5GwOVNkeqXVEyB55Ff8InAXzSgcPHYjYsHDW4b9TX5cq6FqzcUhv1dXmmPw9LTVT7Grm+1MvoZsBEt5MZZlp1joC1jx8REREREVlPVvj6MoBvKYrycwCXANwCYE3E14xDz9bEbqYuByRc8hFCXBv1DroqYqfLPCkicg+ziy2VdS3YuPOYofsIX2jRujCjpwVdWYEPxXm5aDDQ6vZrN47BjvrTpgM49eeUaHHPLi/vOoonvtDzKX1a8SDMeexNy44Zvmi9szHWXiF7fO3GMbhzWqHl1Tt2y8r04LKNVc1CAFv2n8CS6/1Y/bZ1wZIbW1XHo0BeG2IGr8ltdH5fPHzHJLxX34r9xzkT0Unq9YJRkVWPZX/7CgLZf4XdmasAoIy6Afinl20+snZmO44EWi70GGsBAOu2N8TdvPZuoBWLV1dh+dwphmfA6r2+NHL/Mm9nRZgp+xxVdjx+RERERERkLVnh678DuAPAt6/+uwnAj9RPKooyEsDHADwm6XhmqavYA2J8vn/E1xER9WB2sWXlllpDYYPavg/QvzCjpfWdaqY/z1D4muvNwOxxQ02Hr4UDc7r/O3Jx73DLBby08yhOt7Xruk+PAky7ZhB2NJzWfT4bdx7DghktvYJvq6iL1kN8WVix6QNHq0nL/XlYdqu2TvpOzwvWS2bwes2gHPTNykwYGL0baMV7h1ux5Ho/djWelbqp4Gs3jsad04rw6p7j+MXmWJMV3Gfe9CIcOd3m6AYLct7kwv7Y+I0bUFnXgsc31zp9OmlPDdKMGJmXi0fmTkZFST7anrwOOac+gALYGrwKAZwI9cf9Rc9j3T/NsvHI+smoXgwfa1FZ16Kpa0hIAA9s2IXCQTlxzyFRuKrn+lKPsgIfZvrzdF0DhV8nh7MqzJR5jirZjx8RERERETlDSvgqhDihKMpkALOvfuh/hBDhq4/90BXMvirjeBIcuPq/ZTE+rw7NSZ6VSyKylZnFFqOtWcNnelm9MDNxRH+8EGuidRz9sjPxyp7j+m8Y4Strd/SqQlAX92qag/jdNv2NFB6dOxltVzoNha8CwBNbanv8DK0KGscN8+HOaYV4oboR312/y5JjaOWJMkcuHjfMC9br9qnD8fLOY6YrL4+cvqi5jWZIAKvfDmDtknIM8WXhz+834qk3Dpk8A2CoL7vXbMVkMLGwP34+fyrWbW/AA+t3c2ZrGlIAPPDp8QCMb06i6IxW+PfLzuyxEUqPZ778UQxbdytE6x7kGroHc9pCmbi747v4e2gScLWFsptnZBq5powUfk2yckut5s4KIdH7+kblhra3S2eXYvHqKk3fT6xrFquvmWWcYzhZjx8RkZM4r5qIiAjwyLojIcRFIcTLV/8vGPG5vUKIlUKI/bKOZ9LWq/87R1GUHj8DRVF8ACoAXATwjt0nRkTJY+nsUs1VHAp6BqdG3D5leI/KWb0LM3oYXcQpHJgjrUpTrUJ4fvuRHh83+vNru9JpanFKnYGnsipoPHA8iOWv7Hd8dmpky0gtknHx7/ylDsybXoSh/bNifk1WprbLJaEjMVL/LssKfBjqy9Z+wzjUxfc9TcnVuCO3TwaArjbjy+dNhseJWZDkuKbTFzk3WjIFwE/+YaKh266vbsRX1urfhbWu/xMofaoI/Vv32N5i+HLIg+9euQcTrvy+K3i9yg0bUmqag1ix6QPcteZd3LXmXazYtL/HNcXS2aWmnvvUaxIjf0OR1zdAV6Xo4tVVMe8r1jWabBUl+Xh0buLXhXjXLHZcM5s9R5Wsx4+IyCmVdS2Y//Q2zHnsTTy0cR9+sbkGD23chzmPvYn5T29zxWsyERGRXaSFrypFUcYpinKnoiiLZd+3gXPxXj2fMeEfF0IcBPAagFEA/iXiZg8B6Avg90KIC7acKBGlFaMVk6VDu3aK2rEwo1Zh6FHuz0PTmYu6bpOIWoWgvkmraQ7izZqThu5L/bkX5xmvwwl/s2hV0Cggb/6lUeX+PKxdUq57jpiR3xunbT1wEi9UN+LEucsxv8ajtaRVJ/XvUlYVdb/sTFTWtWB9dZOU+7PLhrDzXTCjGGuXlKM8yX6PjJpSNAAr5k22PaRyGwFg2fpd+P6G3U6fSkr51i1lWDCj2NDz8p6mc7q+/meZzyCQtRDlV96x/fe5UwDfvXIPxl55Fs+Hbur1eSdb4lfWteBTj7+JOY+9iafeOIStB05i64GTeOqNg5jz2Ju49fE3UVnX0h3gGX25UTexGF3UDr+d3kpRqxfSE70uxLtmsSvMNHOO4WQ8fkRETnHLxh0iIiK3kFa2oyjKRwCsAjAt7MNrr37uEwBeAbBACLHR5HHuQNd8WQAYdvV/ZymK8l9X/7tFCPGdq/9dCOADAPXoClrDfR3A3wE8oSjK7KtfVw7gJnS1G/6BmfMkotSnpzVieNtaoxWT6u3MLMzoafVjpI1adb3+lr6JhATw8Mv70D/Ha6oi6oUdjabnYIYv4L5voH2xmykAFl1XjMWzRvX4PVFbRr3fcBo1zeeR7fVgWP9szJ1ehDkTh/W6n6WzS7FoVZXjAbJMF9s7LbvvyroWaVXUuX0ysPwVtzQZ0a4qoi1o+JznX752AJv2Njt8htZQFGDZreNQUZKP7YdP44UdjU6fkqMEgPcseA1JV+OH+XDrpGFYUxlA8aBcbA+0Wva8vK/Pl5CjdBgODo0SAvht52fwaMeiuF/nVEt8La3U9x8PYtGqKqyYNwULZhSjaFAuvvrsDgR1Bsbf27AbChTDQbMVbYtlCn9d0NPG0q5rZjPnGE7G40dE5ATOqyYiIupNyjtRRVHKALwBIAPASnTNUv102Je8CaAVwOcBmApfAXwEwJcjPjb66v8BXUHrd5CAEOKgoigfBfATALcC+AyAYwCeAPCQEII9z4goJjM76Y2+yVBvZ9fCjFqFkehNVHgbNavanu0/bv5+61vbTN9HeAD+vRSq0FIUYMXcKT0qMtZtb8Cvtx5EQ9Sf21ls2tsMX3Ym7rupBPd+4sMGExUl+fjmLaV4bLO+tn3p6vylDnxqUu8Q24hn36nHbp3Vam6hLnRHLlqPH9E/ZcPX8FbyQ3x9HD6b1DJumE/K60ayUhQgBGDOY29aepxnMx9GRcYHjoSuf+q8Ad/t+Jqmrz8RvIQnt9TaOnOusq4Fy9Zru04QAJaFLUT/ZtG1ujcxqYvZXywfaeh8ZbQtlvVzjRdelhX4dB3HiTAz1jlqCWXNbtAkInKKGzfuEBEROU3WVfqPAPQBcK0Q4gNFUX6EsPBVCCEURdkGYIbZAwkhfgzgxxq/9jAQu/OVEOIIgLvMnhMRpR8zO+nvqvBjpj9P1+JWuT+ve4HG6oWZyMWhR+dOxobqJlRFOd9yfx7un13a/cYp1d9AGZm5mwwWlRd3B6+VdS148MU9CLQk7rwfvNSBR1/Zj7oT5/F//3Fq98eXzi7DX3cfx4E0Dj+0+tOORkwfOUj3c0I0yRq8AsDW/Sfwyp7jaTXvU20l3yXdGw/L86kJBXj6Sx/t9Vq2//g5rNue+tXFCgAI6H7+nVI0ALsatc2L/qP3IVznOeBI6HpBeDHpyu903e6pNw71+PdMfx6Whl27JGKk2vKrz+qblyvCFqIrSvKxfF7izW+RQgL43yPGqsfVn4WdlaLR7mPlltqorwN6HzOVG8JMPd+X2Q2aREROcMPGHSIiIjeS9a5iNoANQogP4nxNA4BbJB2PiMhRZnfSG2npqzK6wJKohV2ixaEV8yaj7Upn3MVHdeZnKgYo44b5uivzUu37G+rLBnC1ReKG3RA6g+U/7WhEydB+PSpg/+22CZp/x9NZQ2sbFq+uwpLr/XjvcGva/rzerE2/eXXhlWb7jyVvcO42140ZDKB39dmKTcnXkluv8VcrfrU8jSgA5l1bhIkj+qOiJB+VdS0Jw9f7Mtbj25nr4XFgr0BIAI92fAHPdN5u+r7UmXPLIzo+RDISBhp9HQV6LkQvmFEMAJqrZ1W7m85hcuEA7G7SFqQDPTf4OdX2dt32hrhhs9bHLJLTYabe78vIdXT440dE5AQnN+4QERG5mUfS/QwEkGgruQdd1bFEREnP7E56taVvogXM8Ja+KnVhRq/HX6+J+cZo3fYGLF5dFXOx591AK763YTf69snEN2aX4q4Kf8w3SktnlzqyMGu1s23tAIy/uXSzvUfP4jt/2oll640tGANdFTtrKgN48ur/DvFlYeknS1nPp0FIAKveCmDJ9f6U/Nuh6HL7ZGD+09sw57E3sfXASadPJ2XECk1yvRk2n4k9bho7BD+6fQJe+9bH4cvx6ppFf6S1rfv1PF549u+ZTyGQtRDf8dofvAoB/Lz98xh9+bnu4FXGKahtes1cFy1eXYXntx/p/pg6ksDo66h6H6q2K8bmjX/kmgGaH6fIDX5OVIrqnROo5zrMyDWzrDDT6Pel5zo68vEjInIC51UTERFFJyt8PQGgJMHXTARwJMHXEBElBRk76RfMKMbaJeUoj7EoVO7Pw9ol5VF3+BsJONXZKpFkL3ppDZaTzbFzl7BuewP2JnFr11g27W3GCzvMteO8cKUTD23ch19srsFDG/dhzmNv4rHNtbpm1lklGX4VBYD/t/0IWwemCX9+X3xvw+6Uq6J3gwdf3BP1dUpWG9EpRQMwa7T+DVBW+d5nxuOuCj8AGG75B0T/+dyTsRGHshZifuZb9rcYBvByx0z4Lz+HX3XO7f54uT8Pz95djj/cHfv6SauQAJa/0rtxk9HrIhkjCcIXoo0uSg/1ZRve4OdEpaiROYF6OBVmGv2+zGzQJCJyghtavBMREbmRrFe6vwH4gqIoY4UQByI/qSjKDHS1Jv61pOMRETlKVlswdbaX3nliFSX5WPrJUjy2Wd8CVLTZKkYWhxIt9EwrHoQvlhfjlT3H0XL+iq5zdLN/33QApy6kzveTLtwQAGsRvNSBt9Kw/W66UQAcPnXBVHUcxfZuoBWLVlXhW7eUSWnZH65wYI7muah2CL+uMNvyL/LnU+NdCK8HtoeuAICimVDu3oyy5iB+FOfaKPL6qa29E0+9cVDXoXY3ncPtT76NBz49ztBcd/W6aIgvS8pmivCFaDOL2QtmFKNoUC6e2FKLqijnVe7Pw/1R2ibb3fbWjjmBapiZKFCXGWaa/b6MPn5ERE5wusU7ERGRW8kKXx8F8I8A3lQU5ccARgCAoigTAXwcwI8ABAH8XNLxiIgcZ2Zua6TI2XRa9M/26vp6VfhsFdmLXvHmo6WCZAleFSRP4EiUbkbl90Wg5YLTp5HSBIBfbq7Bs+/U4+NlQ7rnmpqdSd505qK8kzQp8rrCbMs/NXR7rPH/YITnjDOh66f/HSi/FwA0bUqL/Jq2dmM/g91NZ7vnbn6keKCh66I/v99k6NiRZFahGt3gJ/P6NhG75gTaHWbK+L6MPn5ERHbjvGoiIqLopISvQogDiqLMA/BHAL+6+mEFwK6r/3sGwFwhRIOM4xERuYETO+nDyZitInPRa932Bk1t+sh6o/L74nDLBQawRC7E4NU+J4KX8cKORrywo+vf44alxiKfcvW6YogvC2sqAzh/qQOv7j1u6L66qyt/fR3WnfwAihOjcUdcC3zlbwDib+Ka6c/D0qtBn+yNXmoL4S+WFxu6/f5jckYSnAxe7r6+krWYrXeDn53Xt3bOCbQzzJT5fRnZoEnkFtw8kD7s3LhDRESULKQ12BdCbFIUxQ/gywCuAzAYwFkA7wBYI4RIzTIoIkpriXbSTykagKlFA9B87hLWVAakvuGUMVtF1uKQ1vloZI9AywUoALIzPbjUEXL6dIiIXGH/8WBKdAYoGpiDF6ob8d31u0zf16I3bwQ2nwFg73xsAUCBAvz4TPfHEm3iUltKq7eXLSSArQdOWnDP2kWOdnBqMduuSlEn5gTaEWZy/iGlOy0badhyNrU4vTGdiIjIjaRe3QshzgBYefX/iIjSQrSd9CeCl/H+kdPY1Xi213w4WW84jd7+RPASntxSe7VFX6eh+4hcHNIzH43sIQAGr0Q6TCkagDunFeL8pQ68vOsoDjSfd/qUyAKp8FJ15PRFHDltrgXyKu8KzM7YCeWypJPSSZn+ZeBzT3T/W+smLqsfv0aDP9dxw31SgtvI0Q5OLmbbUSmaqnMCU/X7ItJCy0YatdX7/BnX2HtyZCnOqyYiIurJ8q2ViqIMAnBFCMEea0SU0tSd9Ou2N+Cx12ssf8NppB0dADz1xiHDx1SFv2EyMjeWiMhtdjWexc//cSrKCnxoa+9g+Eopq67PQmQo9la6diuaCdy9udeHk30T153TirCj/oyU66HI0Q5OL2ZbWSmaqnMCU/X7IkpE60YatdV74aAcBnEphvOqiYiIPiQlfFUUZTaATwF4VAhx+urHhgL4E4DrAXQoivJrIcS3ZRyPiMit7H7DqacdnSyRi0Nrtx227+BERBZau60eB7ihhFLU371fw3DPWSg2p64CgNK/CPj23qifd9smrsKBOWg6o70CVr0uknVNFm0kRCovZqfqnMBU/b6I4tGzkSYkerdaJ+fIfn3hvGoiIiJ5la/fADBJCPHdsI/9HMANAGoB+AAsVRTlHSHE85KOSUTkOna/4dTajk6WaItDW/c7Ox+NUodydRhjEhc/UZJb+06906eQNHK8HlxsZ2vzZPBH70O4znPA/tBVABdCmdj55QNxr3Uq61psPKvEbh43FH+oqtcdmsm6Jos39zMVF7NTdU5gqn5fRLEY2UgT2Wqd7Mf5vERERNbxSLqfqQDeVv+hKEoOgM8D2CyEGAtgLIAjAL4q6XhERK5j5g2nGQtmFGPtknKU+/NM3U8i0RaHapqDaNRRHeIWChxqt0gxKQBWzJ2Cu2/wO30qRKQBg1f3W5W5AoGshZiVYX/w2hkCvtj+fUxu/z2aEsxRjVbp6aTFs0bi0bmT4UnwM4t2XSTjmiwdF7kT/dzK/XlYu6TckvmQNc1BrKkM4MkttVhTGTB9XR7Oye+LyG5GN9K4bQNOOlm3vQGLV1fFXMNQxyU9v/2IzWdGRESUGmRVvg4FcDTs3+UAsgH8FwAIIYKKorwM4E5JxyMich0zbzjN7vaN1o6urb0TT71x0NT9qmLNE0vWN8ufnjQMB1su4MBxeQtsZM5tU4bj7bqT2LjzmNOngoL+WWg+d9np0yAii10ttk85670PYrrnkO2BK9DV1WNVx2fwSOei7o8tWx9/zEK8Sk+7qS2Eywp8hueshl+T3f2799DQ2qb7+OnI7tbKdlV7pXLLaKJwRjfSuG0DTrrgfF4iIiLryXqnexlATti/b0DXWsabYR87B8DasiwiIge54Q1neDu6NZUB0/c3Mi8Xz3z5owC63qBV15/usWCUrG+W/7rnuNOnQBE27nI+dFWdOHcZ37qlFB8cPYfgpQ74sjPRx+vBS//rnnMkIvO+eUspVr6ufVyAGR4Flh+nVGnEJu934VFge/AaEsAvO+bhV53zen1OAHjwxT3Y+p0bo97WLYu5kaMVzIZmZQU+PDp3sua5nwo49xOwp7Xyuu0NcUMHtdpr+dwp0qpSU7FlNFE4oxtp3LQBJ51wPi8REZH1ZF3lBADcHPbveQBqhRBNYR+7BkBylkgREWngtjecMoLR+tY2fPv5/8WepnO9PjfTn4exXESiFCQAPLa51unTICILlfvzsHR2GT46Mg8Pv7wP+y3uhGBl8Poxzx6s9T6CDAcqXYUA9oWK8Nn2f4/7dYGWC/jZX/bhB5+d0OtzZQU+zPTn6R7dIFO8uZtmQjM9c2AFuhbD1duRNdKx2otVt2QHo38nyf73lYw4n5eIiMgeslb8fwfgcUVRqgBcATAZwEMRXzMdwAFJxyMich23veGUFepGC16BrqqA7Q4ulBIRERkRXmFYUZKPTd/8OFZuqcHjm2td24a43J+Hsxfbe4XEB7yL0McTcqTS9X87R2Nux0813+aZtwIoHerrVUlYWdeCsxfbZZ9iLyMH56L+VO8WwPFaCBsRLehau6Q8ZgvjcFZUXFJP6VTtZVdrZSLA2EaadG617iQnxyURERGlE1nh61MArgOwAF3rGRsBrFA/qSjKTADjAfxR0vGIiFzHbW847VhMsWKROlVn8BERkTvVNAfRP9uLedcWobLuJI6d7T3zWQ3o3q5rkTZPPZGJw/vj8x8t6q5Sq6xr6W5h+773nzHQc8n20FUI4FCoALPbHzN0+8j5r4nav0Yyc43w+elF+NSkYZZVACYKumaPG4p3D7dCpFHFpdukU7WXE62ViZbOLtXcaj2y1TvZxw3jkoiIiNKBlPBVCNEOYKGiKF/t+qeI7Nt1CMA0AIdlHI+IyK3c9IbTDW38jIj80Q31ZeFk8DIDWSIikkIA+OnL+/DgbRNihmXhJhcO6K6MHOLLsi18/fxHi3BXhb/73xUl+dgx+N8wMFjnSOh6JpSLae2rzN0PPqwk1Nr+VTV+mA8P3jYBm/Ycx9p36nUfu192pmVzN7UEXXqux5K94tKt0qXaKx1bK5M7aG21Hq/VO1nPbeOSiIiIUpVH5p0JIc5FCV4hhGgRQuwUQpyVeTwiIjeoaQ5iTWUAT26pRU1zEEs/WQpPgkVRu95wLp2d+FxksWoh+ESwdwUSERGRGR8cD2LRqipNgdjuprNYvLoKz28/0r2xyQ6v7D6GFZs+QE1zEKj+PfDjARh03v7gtV0o+GL79zGtfRVuGjsEX/7YSFP3p1YS6mn/CgD9c7yoKMnH4lnGjl84MKf7em1NZaDr5yqB3hBZK/XnRPKkS7WXkdbKRLIsmFGMtUvKUR7jtbLcn4e1S8pZce0gt41LIiIiSlVSti0pijIIwHAAB4UQl8M+fheAOwBcAPC4EOJdGccjInKDeO3lxg7zwYOuxd1IsmeLxaN197EMHy0eiJYL7Qi0XJB+36x6JSIi2fS8toRXiOnpcmHGu4dPI7PhLfx/f39E8pZZbYQAlrXfg+dDN3V/7ONlQ3BXhR9b959EQ2vv+ala/fn9JlPtX/V29vBlZ+Ira3f0+riMuZd6Q2Q9kq3i0u3SodornVork3tVlOSjoiQ/6gxs/p45z23jkoiIiFKVrHcRjwBYBGCo+gFFUb4B4HF0jeYBgDsURfmoEGKfpGMSETkmUXu5A8eD8CjAt28pgy87s8cbTqBrMa26/rQtb0IXzChG0aBcPLGlFlUWtiDeXn/GsvsmdxgxIBtHz15y+jSIoir35+HhOyZ1L/K1tXfgnUOteL/hjNOnRilCrRBbd+8sPDp3Mpat323p8fZ5FyPH0+lIi+G/dM7EfR3f7PU59TrmX24aY+r7/7vJ9q96A/BgjMpFs3MvjQRdeiRbxaXbpUO1V7q0VqbkYFWrdzLPTeOSiIiIUpWs8LUCwBYhxMWwj30HQBOAhQCGAfg9gG8DuFvSMYmIHKFnjtLjr9dg7ZLy7tlmD764J+oinYzKi3jCdx//6m+12LjzGKtJSTcGr+RWCroWhSIX+Wqag5jz2JvOnRilHLVCbFrxIMuO8YF3MbIdCl1PhHwob3866ufDq14WzCjGY5trcfycsdeFnY3GptGoYaTa2eOB9btNX8+YmXtpNOjSKpkqLpNBOlR7pUtrZSIyh/N5iYiIrCergVUhgID6D0VRJgC4BsCTQoi3hRAvANgI4OOSjkdE5Bgjc5TWbW/A4tWxZ8uplRfPbz8i8Ux7Kyvw4YkvTMfdN/gtPQ4RkZ1unzo86qKQnfM5KX1U1rVYErq96P0+AlkLkZNhb/AqBHC6Mxf+y8/FDF6jVb0MyPXacXo9hIeRRYNypd2v0bmXVgdWXOyWb+nsUng0/n0lY7VXOrRWJiI5OJ+XiIjIWrKusHMAhG97rkDXGKXXwz52EMBtko5HROQIo3OU3j3cCqGhUtZo5YUelXUtWPVWIPEXpqAh/frg5PkrTp9G2lDAeb1kj/tujr04btd8TkofsgO332T8HHMyqzUHQjKFBPBoxxfwTOftMb8mWtVLTXMQB6LMtbda+Dms3FIr9TXGyNxLKwOrZKu4TBapXu2VDq2ViUgezuclIiKyjqx3i00AxoX9+1MAzgHYGfaxQQDC2xITESUdo5UuiYJXlVp5YeUCiOzFymTC4NVe86YX4YXqRqdPg1LcoBxvr8WhyAWkpZ8sxcrXtXctIIpHVuD2s8xnsDBjq+3thYGu65JDoQLMbn8MADB+mA8fRAlTy/15uD/KWASr2+1GEx5GWjVrVe/cS6uu15Kx4tIpRgKDBTOKUTQoF09sqUVVlN+jWL/3ySAdWisTmcGQMTrO5yUiIpJPVvi6FcCXFUW5D10VsJ8DsF4IEQr7mhIA1vbTJCKymB3zkIxUXmhl1WIlUTR7jp7FpML+2NN0zulToRR2+mI71m1vwIIZxaisa8HKLbVRn+eG9c82PJ+SKJyMQOZQn4VQFDgy1/VyyINx7c92f+xTEwrw9Jc+itf2HseG6kYEL3XAl52JudOLMGfisKj3Y/d8SCUijLQq/NX7fZUV+FA0MAeNZ+TtMXZLxaUbA4rwczoRvIz3j5yOeo0x05+HpQnC01Su9tLT8YFBP6WLeNeIWp4ziIiIiPSSFb4+CmAegJXo6jJ4HsCP1U8qijIUwCcAPCPpeEREjrBrHtLabfV4+I5J0u/XiUoVSl/7HWhJSelp2frd+M3/HMLhlgsxK/sZvDonK1PB5Y7UKDsOrxDTW10GALXehcj0OBO6ngrl4qPtq3p9bmj/bMx/eluv72XT3uaYC9Iyr4e0tKgXoqtzB9AVmlkV/lY3nMaayoCuAO6mcUOw9p0GKcd3Q8WlGwOKeOcUzbuBVixeXYXlc6cknFWYitVeRlsrp2IQTQQA67Y3xP170POcQURERKSVlHfNQoiAoigTAXz+6odeEkKEvwMdCeDXAP4o43hERE6xa7Hp2XfqMblwgPQ3f3ZXqhAR2SXQcsHpU6AYUiV4jawQ01Nd9r53CQZ6LjoSul4SGRh/ZW3Mr3n2nfqY4WesBWmZ10OLritGTfP5qO1fY52LVZvhth44ia0HTgLQHjQunjXKUPj628XXounMRVcFXW4MKBKdUywhATywYRcKB+XEfQxTNXDU01rZjYE7kSyVdS2ankO0PmcQERERaSXtXasQ4jiAX8X43HYA2xVF4RUMESU1I3OUjBCw5s2fXZW7REREbqZmoFrznGitYNXqsgfW7455P5u9/4oSzzHH5ro+13kTftBxT/yvS3A/0RakZbbbXTxrFE4GL2PRqipN57Js/S4snzfZ9HET0Ro0Gp2xGd7SWQ0AX91z3LEA0I0BhdZziiUkgCe21EY9z2QMHPUGxVpaK7sxcCeSaeWWWs3PIfGeM4iIiIj0smUVXlGU2QC+gq5ZsDl2HJOIyCp6Kl20tNKLxYo3f3wjSURE6U6t+Go83Yb/eOMg6k+1xf36kYNz8fUbx0QNHooG5UZ9nX/T+w1c4znlSOgaEsDmjum4t/M7Uu8z8ppERrtdtY3zgy/u0Xy9JAD89C8fYNwwn+Xt7bUGjUZnbLopAHRjQKHnnGKpCrSipjnYI6RMtsDR7O9JrNbKdgfuqVplTO5V0xzUvWk62nMGERERkRGWha9X57z+M4C7AfjRlUGctep4RER20TNHae60IrxQ3Wj4WGbe/MVa4LCjcpeIiMhNigblYMn1flSU5ONk8HLMICO3TwYyMxScu/hhm/76U21Ytn43fr31IP7lpjFYMKO4+3PLX9nf4/YbvcswyXPEsUrX/wlNwj+1f9+S+4+8JjHabjfc3OmFhhbHg5c6sP940NQmN620BI1GZmzaHQDGC77cGFAYOadYKutaus/TjRW+8Vj5e2JX4O6mTQaUXirrWgzfjuErERERmSU9fFUUZQ66qlxvv3r/CoC3AawC8CfZxyMislKshSotc5RG5/fFH7cfMX0Oet/8xVvgKBqYg3HDfbYsVhIREbmBRwG+cXMJ2q504uevHsDmfc0xXwPbrnTGvJ+G1q4Q9jf/cwg/vWMShviysLupa2/pKu8KzPbsdCx0PRAagVvbf275scKvSWRs6Gq70ml4cVylKF0/g5ifB+DP74tDJuZCawkaF8woxpm2dvxqax2Clzp6fd6XnYn7birB/BnX2BoAJgq+Zo8bij9UGQvRrQwozP5ehDsf9ni4scI3Fit/T+wK3JOtyphSy/koz8VW3o6IiIgonJTwVVGU4eiqcl0CYCS63uMeBzAMwBohxBIZxyEisovWHdqx5ii933Aay9bvlnIuet78JVrgaDxzUcp8tnj6ZWXg/OXYi9dERER2ysvtI+01GQACLRewaHUVbp8yAgBQ0+eL8CrC9uBVCOBKyIOx7c/adszIaxI97Xa13J9eAsC4Ah8G5Hhjboa7/+o1m3q99mbNSWw9cFL3sRIFjeu2N2DFpv0xfxbBSx1YsWk/BuX2wQvVjbYEgFqCLzPhuZUBhcz77pfdtezhxgrfeKwMiu2oCEy2KmNKPerfvl23IyIiIgpn+IpCURQFwKfRVeX6mav3dQXAnwH8F4BNV//NFXgiSip6d2hHzlGqrGvBAxIXebW++dO6wGGFz31kOKZdMwgVJfl46KW9qDx4yv6TsMjnpo5A6dB+2Hv0LDbtbXb6dIiISKeWC1ek36cQwIP7bsfKrKAjoevFUCYmtP/e3gOj9zWJ2m73gfW7DXXUkLHAvf94EK996+MAEHeepHq9dv5Sh6HwNV4YqCdkWrZ+l+6flZEA0I7rQisDCpn3rQZ6a7cdNnR7J1qQWh0U21ERmExVxpSajP4+8feQiIiIZDDzjuYwgCJ0VblWoytwfU4I0f0OQXGi7xYRkQkydmiv3FIrtaWv1jd/ehY4ZCr35+GJ/zO9+98+CxfiJhf2x9zpRb2qjK1cXNy48yievbscABi+EhER/uL9LiZ4Gh0JXUMCGHPlOXsPHCbaNYk6A9dIdbGsBe612w7j4TsmawqdrKiE0nMNZvRyRW8AaMd14eGWC1hTGegVdMsg63ej3J+HsgIf1m1vwLMGZxQ70YLU6spUqysCk63KmFKTkfb46nMGERERkVkeE7e9Bl3vHR8FUCGE+FV48EpElIyM7NAOZ2ShIR6tb/5kH1crjwLcP7u0x8fmTi+y7Hi7m86hoiQf35hdirsq/N3zd5dc77fsmAJdjzPbTxERpbf7MtbjUJ+FmJjhTPBaGxruaPA6Mi835jXJghnFmOnP03V/6jWOujhuhp5KVtmVUHZdg+kJAO06p99tq8dDG/dhzmNvYv7T26TOaZXxe6Fep6qbK41m0U5cA1pdmWp1RaCZ8JhIpqWzS+HR+Jod7b0tERERkVFmwtct6Kp6fQDAcUVRnlIUpVzOaRER2c/MDm2VzAWDyDd/Nc1BrKkM4MkttVhTGbDsuFopAJbPndJrEWbOxGGWVr/+8rUDvb731W8HLDse0PU4Fw7MsfQYRETkTvdkbEQgayG+410Pj5l3TwYIAezrLIL/8nOY0/4Lew8e4es3jYn7eTML3Etnl8JMnt14+mKPa4N410xGQr14m+Hsugbbe/Rsj+8jHieuC9WxHM9vPyLtPvX8TkXyKB9ep5qtAnaiBanVlamy/w4i2dHWmEgLtT1+oueS8OcMIiIiIhkMr44LIW5RFMWPrpmvXwZwL4CvKIpyAMAaAM/KOUUiInvIaO8la8Eg/M1fZV0LVm6pjRoMz/TnYensUkcWKuZNL8L8GddE/dx9N5Xg0Vf2W3LcTXubsWlvc/f3/pOX99nSbvnXW2sTfxEREaWUg30WwqPAkUrXg6ECfLL9MXsPHMPo/L7d7YVjURe4E40CiFzgVq9zzL6UV9a14GTwcsJrpoqSfCydXYrFq6s0XT8kqoSy6xos8vonXkDgVIAVbyyHEVp/pyKV+/Nw/9WfkdkqYKdakNoxq1Lm30Ekq8NjIj0WzChG0aBcPLGlFlVRng/CnzOIiIiIZDF1ZSuECAD4nqIoDwL4B3QFsZ8EsBzAI+jq1ligKEofIcQVsydLRGQlGTu0ZSwYhL/5W7e9Ie6Ck1plcNuUEaaPq9fEwv7d/13THERlXUv3LNabxg1FTXMQ66ubLDv+u4FWfHFVlWX3H2ln4znbjkVERM563fstjPE0OxK6doSA0nbr2gt7FGDJ9X6sfjugKXRRFODhOyZpum+9C9yJrnP02HbwFB6OsyFLvWZaPncK5s+4xlBQHI3dYVHk9+GGcwqnjuWQFWIk+p2aUjQAU4sGYKgvG/2yM3vNnzVTBexkC1I7ZlUa3TCh9b6NYPhFVqkoye/ekBH+vtWKmdVEREREgMnwVSWE6ASwAcAGRVGKAdwD4J8AFAK4DcBRRVHWAlgthNgj45hERLLJ2KFtdMFAAbDoumIsnjWq+82fOp8q0YJkSAAbdx41dFwztFTlfmHGNXh59zEEowTbfftk4MKVTjtOlYiISJMdmXcjL6PNkdBVCGC0xTNdw8PPkqH9LAtdIhe429o7AQjkejNR0xzEEF8WTgYvSwteAWDzvuaE1bPhlZmyKqHMXPsZ/dYTVZg6HWCpYzlkBRpmQhOjmytjjdewk5WVqSqrKgLtCI+JjFBnjRMRERFZTfqWWCFEA4AfKoryIwC3oyuI/RSApQC+YcUxiYhkMLtDWw0ijVgxr3f1gp75VAKALzszashphXJ/Ht5vOJ2wKve9w61YPncKBuZ6saG6EcFLHfBlZ2Lu9CLMmTgM85/eZqoVHBERkQw13i/C6xG2h65AV+haHRqNee0/tfxYD98xqXvR2eo2jGUFvrhtgH3ZmVLHBmi9q/DKTBmVUEZCJgAYld8X9acuGP4ZxKswNXpOMoWP5ZDFSGhidHPloutGxqwstouVlamRx7GiItCO8JiIiIiIyK0sC0KFECEA/w3gvxVFKURXCHuXVccjIjLLyELVlKIBKCvwmWqb971Pj+u1uGNkPlXwUgcUpWsR10oKgJvHDdVclfvAhl1Yu6Qcv1n80e5FnQPHg2g6cxHzphfivcOttsxsJSIiirTKuwKzPTsdC13PhHIwrX21bceMDMSsbMOY6NrIrg1j0URWZpqthJo3vVD3dVug5QJWzJuMDdVNUcNvLeJVmOoJvqzg1NzZSEYDycWzRko+E2PsnFUpuyLQrvCYiIiIiMiNbKlCFUI0AfixoigP2XE8IiKj9C5U7Wo8i089/iZqmoOGQk8FwKDcPr0+bnQ+1W1ThuMvu45ZttCmoKtK94XqRs3HCAng4Zf3oX+ON+rCpD+/Lw63XDDces+Ie27w45m3AjYekYiI3OZgn4XwKHCkxXAoBIyxcK5rLLECMdmhi9bRCTIYbd+7dlu95lm2ibQZHKPQdqUT6+6dhZrmIH752gFs2tus+z5iVZhqDb6s4uTc2XCp0P42mWdV2hkeExERERG5ia3viISwuh6LiMgcIwtVB44HDR9PIPrMLqPVAmVDffg/S4pjLnCYMW6YDz+8bQKG+LLw3fW7dN12f5yfUaDlAuxc955c2B8jBuZgbEE/HGg+b+ORiYjIDeq8C5HhcSZ07QgBpQ6Eriq7AjE9oxPM8CjALeML8Oo+/aHls+/UY3LhgO7uI2aCLaPXba9fPe+Kknz0y/Yauo94xy4alIuyAl/c67BERuf3xaGWC7pv56YwTc/mSsXF7W+TdVZlMofHRERERERGuWM7KhGRiyTaoS1btJldRhdH1YWMIb4szHnsTUP3oQC4ferw7naA44b7cOe0ou7FkTWV8itG7dyZs7vpHHY37bPxiERE5AY7vf+M/p5LjoSuV4QHY688a++Bo7AjEDMyOsGIkYNzMWNkHs5fbjd0e3UD3Om2K9iy/0TUc57pz8NSDVV5Rq/bKg+eQuXBU4Zum+jYZkZiqDxK15zgWHN7Y3Fj5ajWzZVCdG0eUG9H8iRreExEREREZITH6RMgInKjipJ8rLt3FiYV9rfleOrMrvDjG6HezmjbYqBrMXJn41l87zPj8b3PjMdQXzZe3XMcayoDqGkOumaGFxERkRY7vXchkLUQAzKcCV6/2P59VwSvI/NybQk+zFyDJDJnQgGK83IBAPWn2vBCdaOhVr2qkAAefWV/zGDx3UArFq+uwvPbj8S9HydDumjHltH2OXwO59LZpfBo/NvxuLRydMGMYqxdUo5yf17Cr9X6uBMREREREcXCylciohhqmoPY03TOtuOFz+w6GbwMX3Zmd/WpFuFVBmYD0vpTbTErZ4cPyDZ130RE6ahfVgbOXzY2F5KM+Zt3Kfyek7YHrkBX6Prbjs/g0c5F9h88hq/fNMbwbfW0C7Vqk9bo/L54/YNm2+eXhkT0ERHhjMwVlWFyYX9U1rXg1T3HezwuZts+R87h1Fo5Gh7YupF6XotWVSXsuqLlcU9nbCFMRERERBQfw1ciohisrNyIRl2sNNImLrLKwMqZbsfOXrLsvomIUhWDV/ts8n4HYz1HHQtdf9kxD092zrP/4HGMzu+LBTOKdd+usq4lZsvZWC15rbgGUZSuGfE2567doo2IiKRnrqgs0UYpTCrsb2jz4JdnjcSo/L4xQ7REYzkiA1u3WrmlVvPvkZbHPd0YeU4gIiIiIkpHDF+JiGKwu71uv+xMQ23iIqsMKutasKG6yaKzJCIicqefZT6DhRlbHQtd60P5uLH9CfsPrsHDd0zSfZtEm8HU1qzL507B/BnXdH/ciuBl1OC+CLRckH6/eqgjImJV9+mZK2olo11bRuX3xV0V/rhfU1GSj4qS/KStejQyjzjR455OjD4nEBERERGlIynhq6IoZUKIGg1f9w9CiP+WcUwiIqtZWT0aTUVJPh58cY+uBTtfdiZ+s+ja7oVOI1WzREREya6uz0JkKHBkpuvpUA6mt6+298A6FA7M1h2Iat0MFq01qxUteJ0OXlXhIyKiSVQd6mZ6Nh2WFfiSMow02tUm0eOeDsw8JxARERERpSNZycJ7iqJ8XQjxbLRPKoriBfALAP8CIEPSMYmILGXngkG5Pw8AdC9UBi91YIgvC0DXosgDG3ZDMHglIqI0Uev9IjI9wpHQ9XLIg3HtUd/+uErTmUu6K/f0zAyN1prViRa8dtASUEZWh76+rxmVB0/ZcHbm2L3pUBY75hHb3Q3Hjcw+JxARERERpRtZ77AuAvidoiizAXxdCHFR/YSiKGMArAMwHcC7ko5HRGQ5Kyo3olHntZrZjX8yeBlffXYHg1ciIkoLG73LMMlzxJHQNSSAMVees/fAJump3JPRmtUtLXhl0xNQhleHJkP4mmxBmZ3ziJM1mJaF7ZqJjEnWFu1EREQkh6x3EVMB/AHAlwHMVBRlgRBij6IoXwDwGwD9APwcwPclHY+IyBZ6KjcUAOOG+fDB8aDm+w+f11pdf9rQOW47eAo/2bgPKbS2SUREFNUq7wrM9ux0bK7rc5034Qcd99h/cJP0VO7Jas2azC14YzESUCZDqFnuz0uqQMDuecTJ8Bhaie2aifQxsjmEiIiIUo+U8FUIcVxRlE8C+DcADwKoUhRlK4BPA2gBMF8I8aqMYxER2Ulr5YYaos6fcU2PHa4ngpews/EsdjWe7XWbcn8e7g9742V0V/3mfc0MXomIKOUd6rMQikNzXQ+ERuDW9p/be2CJ9FxjyGzNGtmC9/CpC/jd3+sN3b/TjAaUdnVSMUrtwJIs7J5HnGzBdCJGKvHYrplIO6ObQ4iIiCj1SOufI4QQAB5SFOUEgF/jw+B1uhCiSdZxiIjslqhyIzJEDW8zp9Ky0GF09yuDVyIiSmUHvQvh8dgfuoYE8FLHLHyz8xv2HtgCeq4xrGjNGn5t9MExfS1My/15EICj4aXZgNKtM3DDO7AkCzvnESdbMB2PmUo8tmsm0sbM5hAiIiJKPVKvhhVFuQdd7YU7ATQA8ANYqyjKF4UQx2Qei4jITpGVG3rntkQLZKN9jZsrI4iIiOy0z/sl5Hg6HKl03Rcqwmfb/93eA1tEb+We1a1ZjYZfToWXMgJKN87Ajdw8mAzsnEecjMF0LGYr8diumUgbs5tDiIiIKLVICV8VRfEBeAbAPwI4CmAhgHcBPAHgHgD/qyjKPwsh/iLjeERETtESopqhd8asS9bviIiIuikAJo7oD2+GgveP9G67n0iV96sY6jnnSOgaEsCYK8/Ze2ALGancs7o1q9Hwy4nwUmZAqXZSWbFpf9RxFHb52o1jcOe0wqRspWvXPOJkDKZjkVGJx3bNRInJ2BxCREREqUVW5ev7AEYD+AuAfxJCnLr68XsVRXkdXcHsS4qirBRCfFvSMYmILGO0wtUsPQuSt4wvwKv7mi0/p1Rz49gh+ETZEOxtOoft9a2oP9Xm9CkREaUUAWDP0XO6b7ff+0VkeYTtoSvQFbw+0vEFPNN5u/0Ht4iZyj2rW7MaCb8S3UaWm8YOwcfLhlhy7VdRko+X7rsetz35FvY06f8bGT/MhwPNQVMB9FBfVtIu9Fs5j9jua367yKrES9d2zURaydocQkRERKlDVvh6DYDvCCF+GfkJIcSfFEXZAeD/AVgKgOErEbmWmXlIsmhdkKxpDqZt+OrLzkTQ4ALc/xw4iXtuGI27KvwAgBWbPsBTbxySeXpERKTDe957MNhzwbHQ9WhoICra/8P+g1vIbOWeHa1ZjYRf0W6z6u0Azl5s1338WL73mfGWL4R/79PjdbdR9ijAg7dNQOPpNlMVwEYDTDeweh5xqpFZiZeO7ZqJ9JC5OYSIiIhSg6zw9QYhxLuxPimEOKQoyscALJd0PCIi6czOQ5JJy4LkEF+WpefgJiMH5+Lz04tQcyKIl3cdMxy8Al0VWeG7+u+cVsTwlYjIAfdkbMT3Mv8Ij0Oha3OoP65r/03crxuQ45Ua7MnmUYBvfrIMvuxM6ZV76maw5a98gN1RqjRltWY1En6pt6lpDuIXm2tMHT+cXa1S9c6AjQy0igbl4vt/3m2oe4fRANMNOHtUH9mVeOnUrplILys2hxAREVFyk/IqHy94DfuaDgDfkXE8IiLZZMxDkikydP3UpGG9FkGMzF9KVvWn2rCj4TT+58BJKXNuw3f1p9PPkYjILQ72WQiPAkfmuu4JXYPb21do+vrpxQOx9cBJi8/KmERBR6wNXNE+DqDXx04GL2PlltqowevkwgGaQharW7oaDZeisbtVqtY2ytEe54qSfDzzpY9izmNv6j5uMgdjnD2qj9GKur1xWmKnS7tmIr24OYSIiIgicYsVERHkzUMyS2/bYz3zl5LdG5IXv9UF28q6FnSGQlLvm4iIojvoXQiPx5nQNRQCxrQ/p+t244b7XBW+Fg3KwZLr/SgcmIOmMxdRXX8aNc3BHsFHvGsJM237w+1uOhu3G4ie6xkzIY6sdo1aW6XKDpwig6zDLRfQfO4SCgZkY9TgvnHvP12DSM4e1c5oRd366kbM9OfF7fQTr2KdwSylo3R9TiYiIqLYpIaviqLMAPApAIUAovXDFEKIJTKPSURkltl5SLIWGPS2Pa5pDqKmOYhPji/A5n3NUipC08mqtwJ4aOM+p0+DiCgtfOBdjGxPpzOhqwDGXNEXuqrunFaEHfVnXNEdwaMAi68biVf2HI8Zak4tGoDVbwdiXkvICF5VIQEs27ALu5vOYKgvu/sa6P2G05quZ5Zc78fOxrOaN5xFI6Ndo5ZWqXo3x+lldO5oOgaRnD2qndHvXcBYpx+r/06I3C4dn5OJiIgoNkUI88v1iqIoAP4LwCIACrqu18OXVtR/CyFEhukDJgFFUXZMnz59+o4dO5w+FSJKYE2lsRBu8XUjcSBGcFucl4uZ/jxMHNFfUxhbWdei+Y2aAmDsMB/2Hw/qPmciIiI7fS/zWXwl46+2h65AV/D6SMcX8Ezn7YZur84bPxG8hGerGiDhbZNhHgWYO70QG6qb0qLbBfBheBav+q6mOWio9e6XPzYyYWWpKtHmOKDr2uz2qSNw380ltlcxaTk/LT/LZFNZ1+LI7NFkq+qc//Q2w5tHyv15WHfvLE1fm66/h0SR+LdARESUWq699lpUV1dXCyGu1XtbWeHrNwCsBPB7AE8AeA/A4wCeB3AjgAcA/BXA94QQ9aYPmAQYvhIljye31OIXm2t0307daaJFrN3e6gLOqrcCaDpzUfc5EBERudGPMtbgnzI3Oxa6vhmahC+3f9/+g1ug3J+Hm8cNxYpN+9MmeFV5FGDtkvK4IZqRcOkPd8e/T5WezXEqrRV+MkM8p4JIN7A6DFXvf2/TOWw/3Ir61rZeX+Pmqk4jv8PhXvvWx6VuItXyN02U7NL5OZmIiCjVuCF8fQ9AXyHE+Kv/DgH4sRDiJ1f/PQ3AOwC+KoRYY/qASYDhK1HyMFr5qpeiALdPGY7SoT6cCF7C/x45i91NZy0/LhERkV1+lvkMFmZsdSx0PR/KwuT21Hm7sWDGNVgxb4qp6rVkl6j6rrKuBV9cVSX1PlVGf+7xqpqsbM2abFWZbhbvcYrGzZVs67Y34IH1uw2NKPnR7RNwV4U/7tfo/TvRU1FLlMzsek7mcz8REZF1zISvsma+jkVX1WvU+xZCvK8oyssAvg4gdVZDiCgl2LXrVAjgpZ3HAByz5XhERER2OtRnIRQFjsx1vRJSMLb9D/Ye2AZ5ffsYmk2fSqoCrahpDsZcSB7iy5J+nwBM/dxDIvrMzETtKNV5uEZDPKOzY6knLW1DI8V6zN1gwYxibD98Gi/saNR92/MJ5kQb+TvR8vdHlAqsfk7mnGUiIiJ380i6HwVAePnWBQB5EV9TC2CcpOMREUlTVuDDTH/kUxYRERFpsdn7rwhkLYTHY2/wKgTQ2Qn4Lz+XksErAOR6M/Dk32qdPg3HVda1GPqc0fs0c7+qkACe2PLhY1dZ16Ip0FNDPLPHJ2O0Pk7RRD7mbnKlo9PQ7fplx9+vb9XfHxHFt257Axavroq5+UHdzPP89iM2nxkRERGpZFW+NgEoDPv3IQCRZbil6ApliYhcZ+nsUlPzkFLBoFwvTre1O30aRESUJHZ5/wk+zxVHKl1DIWBM+3P2HtgBOxvP4PUPTjh9Go6LV32XqDIvllf2HMObNScBAOOG+3DntKIeFUpG7zdceIXfyi21mq8z1RCPFUv20/M4RePGqs7KuhZs3Gms806i30Gjfycy/r6I0pXezTxurMgnIiJKB4YrXxVF+ZKiKFOu/vNd9AxbXwEwU1GUHyqKMlFRlH8B8A/omvtKROQ6FSX5eHTuZHgSLCA7MMLONjNGDXL6FIiIKAlUe+9GIGsh+mfYH7yGBPDF9u+nRfAKgMHrVfGq7xJV5sXybuA0th44ia0HTuKpNw5hzmNv4lOPv9ldkXcieNnQ/UaqrGsx1ZqV7COrxbfbqjpXbqk1NO+13J+XMEQ2+vdn9HZEpG+TiJsr8omIiFKdmbbD/wXgjqv/vR5AhqIo/qv//ncA9QAeArALwJMAzgB4wMTxiIgstWBGMdYuKUd5jBbE5f48LLqu2Oazss+YofLaL39qQgFuHDtEyn0REZE7fMyzB4eyFiIvo82RatcLnV6Mvvwc/h6aZO/ByXGFA3Nifk5mNc+B40EsWlWF7zy/E8++Uy/lPs9f6mBr1iQh6+ftpqpOo4GyogD3zy5N+HVG//5YhUdkDDfzEBERJQ8p2w2FEC8CeDHs362KokwDcA+AMQAOA/i9EMJYrxsiIptUlOSjoiQfNc1BVNa14PylDvTLzkRFST7KCnyoaQ5i7TsNTp+mJZ564yBGDc6Vcl+v7mu2fWGeiIisc9Br/0xXoCt0bQ8BZWlS6UrRfWXtDsz052Hp7NJeoU1ZQdfmMRkViwAgALxQ3SjlvoCuCj+2Zk0Osn7ebqrqNBoo3z5luKaA1Mjfn5aKWiKKzsxmHv7dkZvEWnMjIkollr0rEEKcBfBzq+6fiMhKZQW+qBd+shf43ObwqTZp9yXSeH4uEVGqcDJ0PRvKxkfa/9PeA5NrvRtoxeLVVVg+dwrmz7imx+eWzi7F4tVVpmZ1WqWiJN/wYrmbQrx0IOvn7aaqTqOBculQ7Qvgev7+PBoraokoOm7moWRXWdeClVtqo66pxdpoR0SUrMy0HSYiSktLZ5cmnA1LRESUzA56FyKQtRAZGfYGr0IAlzo98F9+jsGrCaPz+zp9CjGV+/Nwzw1+Q9dSIQE8sGFXrzCzoiQfj86d7LrrM7XCz+gi4uGWC1hTGWC7SJvIWOw1UtVZ0xzEmsoAntxSK/3xtmMmq9a/P48CLJ87hYvqRCZwzjIls3XbG7B4dVXMYgZ1o93z24/YfGZERNYw++o7UFEUXQMQhRCp2a+TiNKGusDwvQ27XVlhQUREpJUvKxPByx9WQ3zgXYRsT8iRSlchgNFX2F5Yhk4h8NvF12JDdSOClzrgy87EdaMH48cb99lyfDVk+UjxwJgt5W4cOxRPbKlFlc5uIiEBPLGltleAs2BGMYoG5Rq6TyuEV/gZ7Zzyu20fzp1lNYj1zHa40VvVaUf1j10zWRP9/ZX783A/f3+JTOOcZUpWlXUtmtbQ1I12hYNy+HtLRElPEQb7QiqKEkLXSBw9hBAiLbZbKYqyY/r06dN37Njh9KkQkUUq61pcs8BHRERkxvvef8ZAzyVH5nULATzS8QU803m7/QdPM77sTAQtbj2oN2R5be9xfGWt/vdMr33r4zErDMPniFU3nMbWAyd1378Zavgc3h65sq7FdGvkaPdLchl9nPQ+Nuu2NyRchJb1eM9/epvumazr7p1l+Hic40dkLbv/polk4O8tESWra6+9FtXV1dVCiGv13tZsEHoOwBmT90FElJQqSvJRUZLfvcCwt+kctte3ol7i3NR0NWt0Hj41cZhtFTpEROlqvfdBTPcccix0PRfKxlS2F7aNVcGrAuD2qcNx382lukOWpjMXDR2zsq4l5rHKCnzdn1tTGbA1fI0VPsvonMJqEOsZeZz0bjiwu/rH7pms4X9/RCQf5yxTsqlpDuruKlEVaEVNc5CvJ0SU1MyGr48JIX4i5UyIiJJU5AJD+G7v2hNBvLTzmINnl5yqAq247+ZSU63fiIgovkN9FkJR7J3pCnSFrh0hoLTdvS2Gf7v4Wrx/5DSeeuOQ06eSFASAl3cdw4IZxboXyc4bDIS13s6OkPLLHxuJUYP7Jqzwk9EaOVbbZZIn0eM0Mi8XM/x5mDiiv6GqzpVbajUHuzIeb62BMmeyEiUH/k1TsqmsazF8O4avRJTM0qIFMBGRnSLD2AE5e7D2nfo4t6BI6kKTnl29RESkzUHvQng8zoSuoRAwxsWhq+pMWzuG+rKdPo2kYjQk6pdt7C2p1tuZneOZSLk/Dw99bpLmr4/snHL41AX87u/6rhPdXA2SKi1nIx8nWd+PU9U/nMlKlFr4N03JxOqNdkREbsXwlYjIYkN9WU6fQlKqCrRiiC9L065eBcDUogHY2XhW9zByIqJ04WToKgQw+or7Q1fVAxt2Ye70IqdPI+kYCYmMLgzrud3S2aVYtLoKQvJFgpl2jupmvTWVAUO3N1INYmUwWlnXgpVbaqMGizP9eViapCGA7Ba6Tlb/WBUoE5Ez+DdNycLqjXZERG7FZzEiIovxgtG4h17aix99biLWLinHik37savxbK+vUXf1AsBPXt6HA8eDdp8mEZGrven9Bq7xnHJsrmttaDjmtP/C/oObEBLA5n3NTp9GUtIbEhmpTC335+kOeJfPnYwH1u9OuElLATBvehE2vN9oSztHO6pBrA5G121viLtR7t1AKxavrsLyuVMwf8Y1ho+TCow+3nuPnpN2DpzJSpRa+DdNbmfHRjsiIjdiIkBEZDFeMBpXefAU5jz2JnK8GbjY3tnr85MK++P+2aVoPN2WsDqWiCjdVHm/gqGe846Frp0hoCQJWgzHcvZiu9OnkJSMhEt6KlONVpuqLRp/+vI+fBBjo9a4YT788LYJqCjJx53TC21p52h1NYjVwWhlXYuma7CQ6KooLxyUkxLXxkYrzYw+3ut3NGLmqLy0D6+JiCj52LHRjojIjQyHr0IIj8wTISJKVVbPGksH0YJXANjTdA6LVlUBANsNExFd9a73XgzxBB0LXdtDQFkSh65kjpFwqfF0m6YXcgXmqk0rSvLxyjc/jprmIP78fhP2H+uqJhw33Ic7pxX1WOSzq52jldUgdgSjK7fUat78ZnQusJuYrSI2+r0LpFZ4TURE6WXp7FIsXl2l6ZrBzFgHIiI3YYBKRGSDpbNL4XFgETwdCDB4JSJSHeqz0JHgVQjgUqcC/+XnGLymufXVjbrmWqoBodbX8sJBOcZOLExZgQ/Lbh2HNXfNxPc+Mx5Dfdl4dc9xrKkMoKY52Otr76rw4xuzS3FXhV96FYa6SU8PrdUgRoJRPWqag7o3F6pzgZPRuu0NWLy6Kub3rFYRP7/9SMz7MPJ4q4w8RkRERG5QUZKPR+dOTrguJmusAxGRG7DtMBGRDdQLzUTVBx4F+OYny+DLzsT5Sx3Ye/QsNu1158y5QblenG5jS0YiIjc46F0IjweOhK7tIQVl7X+w98DkWnuazulqY6snIBSQVzlp9RxUPayoBjETjGoNmPWE7JG3S7ZWgjKriPU83pH0PkZERERuoY6AsGOsAxGRGzB8JSKSLFZ7OiMXmjXNQdeGr3l9+zB8JSJyWJ13ITIcCl0vhTIwvn2tvQempKC1ja0dAWE0Vs9B1UvPJj2t1SB2BKNG5vuauZ2TZLZXVh/vB9Zrr/gOl4zhNREREWDfWAciIjdg+EpEJInWCopoF5qFA3PQdOYiqutPo6Y52P3v85c6UJyXi4bWNge+o/iOnrno9CkQEaWtvd4vIdfT4UjoKgQw+gpbC1N8IQGs2LQfL913fcyvcaJy0o45qEYk2qQ3ubA/Hvj0eM3nYkcwamS+r5nbOcWKTQILZhRje+A0Xqhu1H0+yRheExERhSsr8DFsJaKUl1zveoiIXEpvBYV6oRkvsHW7i+0hp0+BiCjt7PN+CTkOhK5AV+h6MZSJCe2/t//glJR2NZ7FbU++he/FCA2dqJyUWcEom3qc5a/sx+6msz0+t7vpHFZenfep5XzsCEaN/lySrZ2gVZsEJhb2xwvV+u832cJrIiIiIqJ05HH6BIiIkp3eCgp1AWfd9gYsXl2VlMErERHZ71CfhY4Er0IAlzs98F9+jsEr6abOgH1++5Fen7O7ctJMBaMd1GvDyOBVpW7mi/azjGRHMFpW4MNMf56u+y/35yVdpYtVmwTSJbwmIiIiIkpHDF+JiEwyUkGhNbAlIiI66F2IQNZCeGye7SoEcKozF/7Lz2Fs+7P2HZh6KPfnYeyw5AqrIkVuQFPZHT6ZqWC0mtHNfLHYFYwunV0Kj8bnJY8C3D+7VNf9u4FVmwTSJbwmIiIiIkpHDF+JiEwwWkGx/JX9DF6JiCiuGu8XEchaiIwM+0PXUAjwX34O17avsu/AFNXDd0zCv902wenTME3dgBbO7vDJiTbHWhnZzJeIHcFoRUk+Hp07OeFxPAqwfO6UpKzatHKTQDqE10RERERE6YjhKxGRCUYrIWK1kyMiItrr/TICWQvRJ0M40mJ4S2gqRl95zt4DU0yVdS2oKMlHdmbyv3WL1sLXzvDJ7jbHWlnVDtmuYHTBjGKsXVKO8hhBerk/D2uXlGP+jGsM3b/TrNwkkA7hNRERERFROrL2XSQRUQqrrGvBqrcCTp8GhRmZl4v61janT4OIyJBN3u9grOeo7YEr0BW6Hg8NwKz2p+w/OMV1/lIHXtt7HJc6Qk6fihSVdS09Qik1fErUcldG+OTWGZtm2iEnCvgWzChG0aBcPLGlFlVRAt5yfx7un11q+nusKMlHRUk+apqDqKxrwflLHeiXnYmKkvyUaJO7dHYpFq+u0lSdrHeTgF2PERERERER2YfhKxGRAeu2N3BmqwuNHeZj+EpESedX3sfxWc+7joWul0MKxrX/wf6Dkyb9sjOxobrR6dOQJloLX7vCJ7WCUU+V6ci8XLy653h3BbIVQaLV7ZDtDEbLCnwpEbZGsnqTQKqH10RERERE6YbhKxGRTpV1LQxeXeq1fc1OnwIRkWYf8+zBH7yPOBa6doaAkna2F3a7ipJ8vJ5Cr2+xWvjaFT7pqWAEgPrWNvxic033v2f687BUchWiXe2QUzUYtYsdmwT4GBERERERpQaGr0REOq3cUmsqeJ1c2B+7m87JOyEiIko6B/sshEeBIzNdQyFgTBKErr7sTAQNVgSmCnVupM/imaN2yu2TEffzVodPWisYY3k30IrFq6uwfO4UaTNM3doOmXpjhSoREREREWnhcfoEiIiSSU1zUFerukgeBXjg0+NRNDBH4lk5SwFwzw1+OFC4RUSUdA56FyKQtRAZHnuDVzV09V9+LimCVwBpH7yGz42cO73I4bPpzaMAhQauZzZUN1lwNvosmFGMtUvKUe7PM3T7kAAe2LDL8KzWSGo7ZD3UYJ6cUVbgw10VfnxjdinuqvDzsSAiIiIioh5SZws1EZFBenaum1lkC58BddO4IVj7ToPh+3KT26cOxw8+OwGn29rxwo7UmUlHRCTTQe9CeGwOXFVCACdDPsxsf9r+gztkStEA7Go86/RpRDWsfzaaz11CvKLLyLmRcyYOc1UlsNpetfF0G5at363rtlWBVtQ0Bx0Pq6JVML6wo1Hz7PiQAJ7YUiut+lRPO+TwYJ6IiIiIiIjch+ErEaWtyroWrNxSG7WSNdY8r/MGFz2LBuVgxbwPF1EXzxqVMuHrbVNGYE1lAKfOX3b6VIiIXGeT9zsY6znqWOh6JpSDae2r7T+4gzwKsOzWcXjwxT0ItFyw5ZgKgEXXjcRQXxba2jsBCLRd7kTdyfNo7xDom5WBccN9uHNaEcoKfKisa9E9N/K+m0rw6Cv7bfl+4lGA7vNbUxkwdB+VdS2Oh68qtc1xTXOwx2xXLWQGyVrbIUcG80REREREROQ+DF+JKC2t294Qd3Er1jyvfgZnri253t9jkUxtL2emhbEb+LIz8ZW1O5w+DSIi1/lZ5jNYmLHVsdC1MwSUJEl7YZnUYAoADp+yJ3j1KMA3P1kGX3Ymzl/qwFBflqb5j7HytVgfv/cTY1DTHMR6h9v2CnxY8Wl0U5rR21nJaHcTmUHyghnFKBqUqzuYJyIiIiIiIndh+EpEaaeyriVhVQHw4TyvwkE53YtcRhe7ot1u6exSLFpVFbftoNu5pf0hEZGbHOqzEIpif4thda5rssx0lW3EgGz833+ciiG+LNzzu/cgbHiB9ef3RVamB7+MUjEZq4uG0Q1gAPCL+R/B2YvteP2DE9K+ByPUik+jm9KM3s5KRgPhN2tOahpdoVW0dsiy7puIiIiIiIjs4b53vUREFlu5pVbTPC2g9zwvIxWr5f68XotlastjLafhUboqZ3c1no1aBUFERO7g1FzXdA9dVUP6Z8UcJ2CFwoHZqD91QVeIamYDmMrjRDl1FJV1LYY3pR0+dQFrKgOuChSNBsJbD5zE1gMnu/8dK3TXS22HTERERERERMmH4SsRpZWa5qDuRdnIeV5LZ5di8eoqTQGuR+maixYuUcVLuMj2cjXNQfzytQPYtLdZ1/dARETWqfMuRIZDoasQwOgr6R26qnYeOWvbsRQAR89eSlhdGxmi6t0AtmLTftw5rbC7+jG3TwY273PHNcD5Sx2Gxyj87u/13f8tK6w0S9bx41UuExERERERUXrwOH0CRER2MjPPS1VRko9H506GJ8Eiuzp3LnwxT2vFi6qswIchvqwe/544YoCucyciImvs8X4JgayFyMxwJnjdEprK4NUBHgUYld9Xc1tjtYuGkQ1guxrP4qGN+/CLzTV4aOM+LFu/2zXjCtRK0aWzS2Hm118NK5/ffkTOiRmkBskyqKG70etOIiIiIiIiSm4MX4korRid5xV5uwUzirF2STnKYyzSlfvzsHZJea+KBz0VLwCw9p16zHnsTcx/elv3Ap4b56QREaWT17z/ikDWQvTL6HAkdL3Q6YX/8nO4u32ZvQcnlPvz8OjcyQi0XNB1u6pAK/78fpNFZ+UMmZWqbgkrl84uTbi5Tis1dCciIiIiIqL0wxV8IkorRoPLaLerKMlHRUk+apqDqKxr6W4JGGt+mZGKF1V4Czun2/IREaWzQ30WQlGcqXTtDAElaT7X1QlFg3Jw09ghWDxrFMoKfFhTGTB0P/uPnZN8ZvJ4MxS0d2rfHebNULqvdZa/sl9KNa4aVjp5naN2N9HTpSSeyNEVRERERERElB4YvhJRWjG6oBfvdmUFPk2LamarOdSqkLVLyg3NVyMiIuMOehfC49Bc11AIGMPQ1RGTRvTHEF8W+mV7uz9mtIuGm+kJXtWvX7zqHZy+2I49TfJCZTeElQtmFKNoUC6e2FKLKgnXWpV1LQxfiYiIiIiI0gzDVyJKK+o8L73B5YMv7sHS2aWmqjFkLNaqVSFLZ5di8eoqKVUZREQUG0PX9LbnaFewuPXASTz1xkEM9WWhfLSxuaDjhvuw9cBJmafnqLfqTllyv3aGlbG6l0TrblLdcNrQ45eKYT2RVlo7BBERERERpRqGr0SUdowEl+FtfyPnuGola1ZrVaAVQ3xZUtviERFRT06GrkIAo6+kZ+h6y4QCvL6vWUobWyucCF7Gxp3HDN32zmlF2FF/hp0rErAjrKysa8HKLbVRH4uZ/rzuDXfh3U3WVAYMha+yrv+IkonWvzEiIiIiolTlcfoEiIjsps7z8uhcUFfb/hptH1w4MMfQ7aKprGvBghnFWLukHOV+YxU4RETU2z7vlxDIWoiMDGeC15MhX9oGrwBw4HjQtcGrGeX+PJQV+LB0dqntv1eJjMzLdfoUerA6rFy3vQGLV1fFDMHVDXfPbz/S4+OyRlfUNAexpjKAJ7fUYk1lADXNQUP3S4nxZ+0Mo39jRERERESphNtwiSgtGZ3npbb9NbIA13Tmou7bxKJWhYS3xfv+ht14r/60tGMQEaWTD7yLkO0JORKMCQGcCeVgWvtq+w/uMg2tbU6fgnQKgPtnlwLoet2+bcpww9WzsnkU4Os3jcGy9budPpVuVlbDVda1aOoaom64KxyU030+RkZXqKG7emxWAtqDP2vnmPkbIyIiIiJKJax8JaK0VVGSj4fvmKT7dlWBVkM752W20YusCikr8OGRuZOl3T8RUbp437sEgayFyMmwP3gVAujsBPyXn2PwmsL8+X17hAtZmRkOns2HFAW4bcpwnDh3GYNyvU6fDoCeYaUVVm6p1TyuQd1wF27p7FLNnVM8yoehOysB7cOftbPM/o0REREREaUKVr4SUVoz2kL4l68dwMQRA9AvO7N7JlikmuYgKutacP5SB/plZ6Km+ZzZ0+0WbYe4kYoMIqJ0dqjPQiiKM+2FQyFgTHv6thdOJ4daLqCmOdh9reCG12lfdiaClzrw0s5jANxThauGlVaoaQ7q/tmrG+7Ux04dXZGoss+jAMvnTkFFST4rAW3En7WzZPyNERERERGlCoavRJTWjFajbtrbjE17m7v/Hd7CLF6rMxniVYUsnV2KxaurNO84JyJKRwe9C+HxOBO6CoG0numaru7+3Xt4dO5kXLjc4Uhr5a/dOAa53gzUngjipZ3HEJTYjUOG8LDSKkY33FXWtfS47ko0uqLcn4f7r14TqmMh9FYCMhA0xkjVJX/W8sj6GyMiIiIiSgUMX4korUW27zVKbWE2d1oRNrzfaFn4magqRK3IcNPsNiIit3AqdAW6QtfnOm/CDzrusf/g5LiG1jZ8cVWVI8cu9+dh2a3jUFnXgl9urnHkHPz5fZGd6cEHx3uPbQgPK61kdMNdtNtVlOR3h6vhXU7UbiiVdS2Y//Q2QxvxIisBYx0jktavS1WsunSezL8xIiIiIqJkx/CViNKazIW+kABeqG6Udn+RtFaFqBUZX1pdhU5WwBIRoda7EJkOhq7nQ30wuf2/7D84EYCbxw0FAPzk5X1w4rJgcmF/bPzGDQCcDQiNbriLd7uyAl+v81+3vUFT69t4KutacDJ4OWYnFa0dV8K/zgpuCnxZdek8K/7GiIiIiIiSFa9yiSitJcucVL1VIRUl+fjurePw6Cv7LT4zIiL32u39J/TzXHEsdOVcV3KDv+0/gZvGDcWBKFWnVvMowAOfHt/972hhpVlaA0CjAaSe22mdOZrItoOn8PDL+2LeT3fHlemF2FDdlPDrlooMhTcAAJaRSURBVM+dgvkzrjF3UmEq61rw6CsfYE/TuV6fszrwjYVVl86z42+MiIiIiChZMHwlorTn5jmpt04swLfnjDW0UDmpcIAFZ0RElBwO9VkIRXFmritDV3KTqkArVr99yPbjWj3HVW/Fp5ENd+X+PF3XYHpmjsazeV9zwirlkABe2NGU8L5CAnhgwy4UDsqR8lj87C/78MxbgZiftyrwTSSdqy7dUoFsx98YEREREVGySP53GkREJqlzUmVUKshmNHgFuhYAiYjSjVNzXRm6kptV15+25H6LBuWg8fTFXh+3eo5rota+sQJAPRvuPApw/+xSzedkZOZoLLIvR0MCeGJLrenHI1HwGn48mYGvFulYdelky+lYrPwbIyIiIiJKJgxfiYjw4ZzUJ7bUosolLYhj7QTXsrtd5gIgEVEyqPMuRIZDoasQwOgrDF3JvTos2l225Hp/99xRvVV34dczbe2dAARyvZkJ70Nra99oAaDWDXdGqnaNzhy1S1WgFTXNQcOb+irrWjQFrypZga9W6VZ1aXQDgtWs/BsjIiIiIkomDF+JiK6qKMlHRUk+Vmzaj6feOOjouUTbCa5nd7vbFwCJiGSp9S5EpgOhK9AVuj7XeRN+0HGP/Qcn0mHEwBwEWtqk368akuoJsCrrWrD8lf3Y3XQ27tfFqtzT09o3WgCYaMOd0apdGbNDFciveg1XWddiOGxc/sp+3bcxG/jqlS5Vl2Y2INjBqr8xIiIiIqJkwvCViChCrjfD0eNH2wmud3e7jAVAIiI3czp07QgBpWwxTEniy7NGobLulNT7NFI1qLVtLRC9cs9IZ49oAaC64U7mrEyzs0M9CvDJ8QV4bV+zqfuJx+j1YU1zMGFYHouZwFevdKm6NLsBwQ5W/I0RERERESUThq9ERBHMLp6ZEW0nuJHd7U5+D0REVlrlXYHZnp2Oha6c60pW++3ia3G67QoeWL9bShVkuT8PcyYOw7hhPuw/HpRwj7GrBuMFLXqCV1Vk5Z7Rzh6xAkC9VbvxmAm31Ou/muagpeGr0etDMx1V7N4QmOpVl7I2INhF5t8YEREREVEy4eo8EVEEuxZjbp86HNOLByXcCW5kd/vDd0ySfLZERM471GchFMWZua4MXckOalCqMhvAhoekP7xtAhatqjId6EarGkw0GmH2uKG6g1dVeOWe0SDPjgDQyMzRkXm5eObLH+2+/hviy7Lq9AD0vsbVWpVo5ufnxIbAVK66lL0BgYiIiIiIrMHwlYgoQlmBD5MK+2NP0zlLj/ONm0sTLoIY3d3+fsNp3QuARERuddC7EB4HWgwL0fV/o68wdCXrKYhSTWpiCGhkSFpRko/l8ybjgQ27IeLcpwJgVH5fBFou9PpctKpBLaMRzF6PqJV7RoM8uwJAvTNHH5k7uce1oJEAV6vwNtGJwvLIWbtmfn5OVpimYtWlmzcgEBERERHRhxi+EhGFURej9AavOd4MXGzv1Pz1RQNzuneux1sUMrq7/YH1u3H3DX68d7hVc9UsEZHbOBW6Al2h65bQVNzdvsz+g1NaumVCQXdQpY4ciBeSxhOrtaraknXFpv3Y1dh7hufIwbn4+o1jsGBGsaaqQa2jEWSorGsxHOTZFQDKmDmqJ8DVKrwCWktYHjlr1+jPb3Jh/5QLP53m9g0IRERERETUhVfgRERXJVqMikYB8K1bynDtyEG6Fsoaz1zEQxv3AYheYaB6/8hp7ScTRgBY/XYAd04rxPrqJkP3QUTklD3eL6Ovp92x0PVcKAtT29fYf3BKawX9s7v/W8/IAdXnpxdhYmH/uK1V1U1m0YJXAKg/1YZl63djfXUTls4uxV0V/rjHNHKeRp2/1GGoMjS84tMOZmeO6glw504vxIbqJs1Br9awPHzW7hBfFirrWlA0MAeNZy7Gv2GEBz49XtfXU2Ju34BARERERERdGL4SEcFY5Ubk4pmWhbJoolUYqGqOn9d3Z2FCAniv3lh4S0TkhE3e72Cs56hjoWtnCCjhXFdyyLPv1GNy4QB8pHigobazNSeC+Pn8qTE/r2eT2buBVixaVYVPjB0C9c9x3HAf7pxW1B1kGhmNYIZauae3tW+vVs42MDtzVE+Ae+e0Is1Br56wPCSArz67A0GD7WrvucHPwM8CybABgYiIiIiIGL4SEQHQX7kxpWgA1t07q8fHEi2UxRNeYRC+UJXt9ei6n0j1p9pM3Z6IyA4/y3wGCzO2Oha6hkLAGIau5DCBrmuBL5YXG7r9rsazqGkORg1ZjGwyEwDeOHCy+99bD5zEU28cwthhPvzbbRNQ0xw0dJ5Ghc+u1dPad4gvC2sqA7oDUBnMzBzVGuBq/TojYbmZ4PUHn51g6LaUWDJsQCAiIiIiSncMX4ko7RlZjIq1wKkugK3cUoPHN9dCTxFsSABPbKntEb4O658NIHprQCKiVHCoz0Ioiv1zXRm6khuFRFfIaVRlXUvUsE9me+ADx4NYtKoKcyYUyLlDDSIr97RUht48biheqG7Ed9fv6vX5eCMf3EZrgJvo6yrrWmSeVlRTigZg2a3jkuLnmsxkzBYmIiIiIiJrMXwlorRndDEq1gJnZV0LVr6uL3hVVQVasW57AxbM6Kp6mTu9CJv2Nhs6PyIiNzvoXQiPx5nQVQhg9BWGrm7XJ9ODKx0hp0/Ddo2n9c3VDHc+olKxpjmIP7/fJL09sADw6j57rk9iVe7Fq/h8v+F03GAq3sgHKxltQyxD5O+GLEUDc7Dkaothtra1j9nZwkREREREZC2Gr0SU9owuRsW6ndnqkgfW74YCBfNnXIM5E4chx5uBi+2dxu+QiMhFnApdga7Q9bnOm/CDjnvsPzjplo7Bq1nqXNTKuhas3FJr60xWK2ip3Ius+NTaYjnWyAcrxHs87KrCVX83ZGs8c1FK8OpkMJ2szM4WJiIiIiIi6zB8JaK0Z3QxKtrtjLQwjqTOfFMXA/P79cERE1UwRERu4HToeiQ0GB9vf9L+gxMZMNSXhRPBy7pvV1GSj3XbG3TPd3Ujo5V7ejbBRRv5oIeW0CvR42FXFa6V4W6sbjBab+t0MJ3szMwWJiIiIiIiazB8JaK0Z3RBJ9rtZM3TUhcDh/iyGLwSUVJ713svhniCjoWul0KZGN/+e/sPTmTCpyYOw9p36nXdptyfh5PBy64NXhUAD3x6HP62/0TUNqnDB2SjYkw+Jhb2N1y5Z2QTXFWgFTXNQV3H0xoY6qnCXbZ+F3Y3ncXiWSMtCdLKCnyY6c+zpBraaBcZtwTTREREREREsjF8JaK0Z2QxqtyfF3VhTOY8rapAK/78fpO0+yMistuhPguhKM7MdQ2FgDHtnOtKyWnxrJHI9nrwzFsBTV+vzkU1O/rASrdPHYF7PzEG935ijGVtUo1ugtNTuaknMHyhulHz4yEArH2nHmvfqbes4nPp7FIsXl0l/XfESBcZN7aHJiIiIiIikoXhKxER9C1GKQpQVtAPT26p7V4wBLoWkaobTks9r/3Hzkm9PyIiOzjVYpihqzUG5Xpxuq3d6dNIG5ML+6OswIcffHYCACQMYNW5qEN8Wa6e8Vo6tF/3f1vVJtXoJjitt9NbyWo047Sq4rOiJB+Pzp0svTraSChqZ3toIiIiIiIiuzF8JSKCvsUoIYC17zTYc2JEREnEydBVCGD0FYauVrh/dimOnrmouQrTTTwKXFsJGstHrhnU/d8/+OwE3Dh2KFZs2o9djWd7fW34XNQ1le5+fIxUR9p1DK230xMYmv21k1XxGa3KeO2ScjyxpTZq+2dfdiaCOkLsWN1gEp2THe2hiYiIiIiInMLwlYjoqgUzilE0KDfmYpQTxg33YeuBk06fBhFRXE6FrkBX6FobGo457b+w/+BpQg1+jp+7hI07jzl8NvokW/AKAEN9WT3+XVGSj5fuuz5hq16Zow+sYEfVotFjaLmdkcDQLDMVn1rm0g7xZfX6nToZvKy5G4za7trIuRmhpz00ERERERGRkxi+EhGFqSjJR0VJPl7bexwbqhsRvNSByx0hvFcvt52wVndOK8KO+jOubiNIROnL6dCVLYatN3pwLh58cQ9fh2wUXoWpZzaqHZWlRvmyM3EyeNny4KyswIeZ/jxdv69aKzeNBoZmGan41DOX9q4Kf4/PlRX4NHWDUdtdGwmGrW4PTURERERE5DT3vkMnInJAvCoBJzz44h7MHjcU2wOtptvXERHJ8pb3PhR5Whm6poFDp9pw6FSb06eRVipK8jVVLUaGXnbPw1SgvbVu8FKHJTNMo1k6u9SSyk0ngz89FZ965tLGamucqBtMeLtrI6xuD0366NnkQcmNjzURERGRffjuhYjoqkRVAk54N9DK4JWIXOVQn4VQFGfmujJ0lWPW6DxUBVpd9XpHH/rtmwexobpJU9VieJBppOrTiHHDfPjhbRPQeLpN13VTeNgXrd2tjACgpjmImuYgPjm+AJv3Nce9ftJbuelk8Kcn+NUzlzZeW2O1G4wVYY2V7aFJOyObPCg58bEmIiIish/DVyIiaK8ScIILT4mI0pBTLYYZusr3TqAVd1/vx67Gs66ZcU4femFHU8KviVW1OLVogJTw1aMA3/xkGS62d2L/sXMAuubQ3zmtqEfwVjQoF199dgeCGsPBkEDMrzcTAOjtXGKkctNoMKEoXc9jZmgNfo3MpU3U1riswCe9Ms7K9tCkjZ7W1FZXq5O1+FgTEREROYPhKxER9FUJEBGlEydDVyGA0VcYusomBLD67QDWLinH8lf2Y3fTWadPiQyIrFqsrGvB6rcDpu9XTzA5xJelOXhVxfp6owGAls4lCoA5Ewpw3ZjBcSs341V5Gg0M504vNL3BT2vwa3QurZ62xrJY1R6aEpPRmpqSAx9rIiIiIucwfCWitGekSoCIKNXt9y5Clifk2FzXkyEfZrY/bf/B00RIACs2MXhNduFVizI2kn3txtFYdut4zV//5/cbzR0wgt4AQGuwIABs/qAZX/rYqKgho9aWnEYCw4qS/LjzUxPRU/FpdC6tE/NsK0ry8ejcyQkfP73toSkxWa2pyf34WBMRERE5h+ErEaU9o1UCRESpaIv3WxjtaXYsdO0IAaVsMWyLXY1yg1cFbJXvhF++dgBzpxdJ2Ug2anBfTV+nt82vHpEBQLxqVBnBgt6WnEYCw/D5qWu3Hcaz7zRo+lvRW/FpdC5ttNtZMes10oIZxXGDaSPtoSk+K1pTkzvxsSYiIiJyFsNXIkp7Tuz2JyJym3syNuL7mX90LHTlXNfkt+i6YtQ0n+ccWZtt2tuMTXubpdzX9zbshgIlbttfLW1+zaoKtGLd9gasr26KWY06b3qh6WDBSEtOM4FhWYEPD98xGZMKB1hS8Wk0pAy/ndYqYFnCg2mrw15KrtbUZA4fayIiIiJnMXwlorRntEqAiChVHOqzEIrizFxXhq6pY/GsUQC6WtHuOHwa7x4+7ewJkW6J2v5qDStleGD97pjVoe8GWrHdYMgfHiwYrZw1GxhaVfFpdC6tes56q4BlKivwMfCxQTK1piZz+FgTEREROYuJAxGlPbYyI6J0ddC7EB6PM6GrEMDoKwxdU0WBLwsPvriHM9RTQLy5fzLmymqV6DBGT0MNFmS05DQTGFpV8WlkLi1grAqY19DJR2ZranI3PtZEREREzvI4fQJERE5TqwSIiNLFQe9CBLIWIiPDmeD1TCiHwatLTCrsL+V+moOXGbwCGDcsNSr31JAxnJGw0o3UYMFMS06Zygp8uKvCj2/MLsVdFX7T1Z8VJfl4dO5keBI8t0e2NTZSBUzJR0ZrakoOfKyJiIiInMXwlYgIXVUCiRapiIiSndOha2cn4L/8HKa1r7b34BTTrNGD+foniUcB7pxWiFT5ca7dVt/j37JDR6eowUIqt+RcMKMYa5eUozzG5sJyfx7WLinvbh1spgrYbjXNQaypDODJLbVYUxlw5BySmZFNp+GtqSl58LEmIiIichb7iRAR4cMqAbvmmBER2Wm990FM9xyyPXAFONfV7Va/HcCS6/1Y9XYAgq9/hqlVhC9UNxpuh+s2a9+px+TCAd0BnZnQ0ZediaALQsvwYCHVW3LqaWtspgrYrqCmsq4FK7fURg2JZ/rzsNTAjNx0ZbQ1NSUfPtZEREREzkmOd45ERDZYMKMYRYNy8cSWWlSlQFs9IiIAONRnIRTFmUpXhq7uFxLArsazKCvw4cBxVpDFMm6YDwNyvFGvD8r9ebh/dimG+LLw3fW7pB2zaFAOGk9flHZ/Rixb/+FsT6Oh49duHIPrS/I1BwBWiQwW0qUlZ7S5tJGB7OGWC4bu264q4HXbG+JukHw30IrFq6uwfO6U7s0CFJvWTaeRrakp+fCxJiIiInIOw1ciojDhVQJrt9Xj2XfqU6aChYjSy0HvQng8DF0pMW44is+jAD+8bULCKsI1lQGpx11yvR8VJfn45WsHsGlvs9T71kqga7anen1kRI43w/IOIwoQ93otWrCgtuTU02432VtyxqseNcKOKuDKuhZNvzchATyw4cPNAhRfok2n6qYS/iyTHx9rIiIiImcwfCUiChO+qDp6SF/MnV6I9dVNTp8WEZFmToauQgCjrzB0pdQRGdpFqyJUya4CVIPdb88Z61j4Cnw429NIWAkAj79eg2tHDrK0w8jyeZOxobpJd7CQTi05E1WPGmFHWLNyS63mcw6JDzcLUGJ6WlNTcuNjTURERGQ/hq9ERJBfCUBEZDenQlegK3StDQ3HnPZf2H9wIovoqQaqaQ5i79GzUo+tLogbDT1lUmd76gkrVeGBmPp/P3ppD37393op51buz8OCGcVYMKNYd7CQLi05tVaP6mFHFXBNc1D37334ZgHSJt6mEkotfKyJiIiI7MPwlYjSnhWVAEREdnE6dO0MASVsMUwpYKgvCx8vHYKJhf01VwNZsXlLQe8KSyOhp0xqVW9FST6WfrIUj22u1XX7yEBs1OC+Us4rshrVSLCQDi059VSPamFXFXBlXYvh2zFgIiIiIiIiJzF8JaK0ZkUlABGRHV7z/itKPcccC10515VSzYngZWx4vxEz/VM0BTdWbd4S6ArLgA/bulo9MzWR8Nme/bO9hu4jPBCTEWTKrEZN5ZacRqpH47GzCthoK2/ZLcCJiIiIiIj0YvhKRGlNdiUAEZEdDvVZCEVxZq4rQ1dKZSEBPLBhFwoH5XSHS9ECuZPBy5YGoe8GWrF4dRWWz52C+TOuAZC4QtNK4UGbjEDMbCtlq6pRU7Elp9Hq0WjsrgIOD/3tuB0REdH/397dx0d113n/f38DgRAYbkJKoKHAQJJiW9BSIdaovUGru+pupdruYlm3P1zrulp0vX7beq3+1MvdLd1dt9K6uq6y1cX1ErdQV7vb2hpbrbFNEbDQG5qkBGhSSBvCTbi/yff3x2ToMJmbc86cmTMz5/V8PPIImZlzzjeT5JDM+3w+HwAA/MJfJQBCy+9KAADIt6BaDBO6IkziM0olpW0pHKkanfeLt1IFwYkVmuue3K11T/kzNzWT5NmeXoOtVwdPnPexm1bKRtINi2a6agmNGK9h+UfeOltzpo4PtArYa8hbyi2iAQAAAJQHwlcAoeVnJQAA5FOQoau10txThK7Irn5ylXoPnsj+wBLQ3j2gm9e2y6YJBgcL1NY0HgQnh0lNdRF95frLdOj4Kf3kmb15O36q+bNeg61neg6N2I+TVsrxNrfxCmC44zUsnzN1vG5pifq8Gne8VEgnXywAAAAAAEGoCHoBABCUcp0HVTWaUztQLl6qXK7uscs1alQwwetrQxGCVzgyf3pEew+VR/Aaly54LbT27gF19A2mvO+T1zamvN0PRtJdN5w/27Otq1+f//Gznva3refQiM/jpsWztG5ls5qjNSm3aY7WaN3KZoLXHJR69eiqpY2qcPj/X4UZebEAAAAAAASBylcAoVWu86BOnR0KegkAcrS1cqUmVxwveOAq0WIYMXNrx6u7/6ic5I9G0sDRU8xQz6O2rv6U1XxeKwNvW9qov3nweb2wL3WoO396RF943yXnBXDrN+3Jec5tqs8jsZVy8mxdKhhzV+rVo24rpIslNAYAAAAQbuWZPACAA+X64gwvfgOl629Hf1vLRz1G6IrA7ew/qjfPmaLf7jqQ9bFW0quDJ/O/qBDL1K3DzezUeGVgS0OtHvr0O9TRN6gHtvZqx97DkqT5MyL6wOUzRwRvbV39OQevUubPo6kuUjSBX7nx8j2SThAh+U2LZ2nmlGrd09qp9hQhcvyCgnL93R4AAABA6SF8BRBaXioBACBfdo5ZLmOCaS9M6IpUnASvKIxM3TpyqQxsqovo9vfMz3r8Na2dvlzcVa5dR4qdH9WjbV39WtPamfL35iXRGq3Kc/hJhTSQX/xsAQAA+Iu/fgGE2qqljbr5O+2O2ioCQD68VLlcFRXBhK7WipmuQAnIFmrlszKwo2/QtwvVqEwMTi7fI9laTj/dPaAVa9u1etnCvM/npUK6OBDUlY+gL6wAAAAoV4SvAEKt58CxoJcAIKSCCl2lWOjaOvRGffT07YU/OABXnM7fzFdlYFtXv+dtExXTHNGw8vI94rTl9JCV7ti4TfVTxmUMagjtShtBXXkppgsrAAAAyg3hK4DQir+YRNUrgEIKOnQ9MyQ10mIYKAnZ5m+m4ndlYKY5rU55+TyQP26+R9y0nB6y0j2tnUXZthi5I6grL35fWAEAAIDzEb4CCC2/5pcBgBOdlcs1OsDQlbmuQGnJNH+zkHKd05rp86AKsrh5aTnd3j2gjr7B876OhHalj6Cu/Ph1YQUAAABSI3wFEEp+zi8DgGx2jlkuY4KZ60roCqQXGTtKgyfPBr2MEXKZ0epVuiA01zWsfFt0RKBGFWRp8Npyuq2r/1z4Wo6hXRgvGiCoKy9+XVgBAACA9AhfAYSSX/PLACCToFoME7oCzrz7shm6f3NPQY5lpIyjDoyk97/xQn3y2oaCvrjtJAhdEq3xfNHa2l936+qLp50LYqiCLB1eW04nbldOoV1YLxogqCs/flxYAQAAgMwqgl4AAATBj/llAJDOS5XL1T12uUaNKmzwGg9doyd/QPAKZPFnb4/q0gsnFux4NyyaqeZoTcr7mqM1+v5Hm3XPH19e0Be212/aoxVr29MGK/Eg9I0zJ6nC47ksHqhJ7qsguVguWF5bTse3yyW0KzZOf1Z+tOnlAq8s/3IJ6lCc/LiwAgAAAJlR+QoglHKdXwYAqQRV6SrFgtfWoTfqo6dvL/zBgRLzZ2+P6q/fe0lBQ55L6yfqH298Y8FalmY7jpsg9DtPdKt5bo3auwdkHVYxJooHal6qIC+IjA1di9di4bWKM75duVTXlWPrZDcI6spPrhdWAAAAIDt+cwIQSuX0ggiA4AUdutJiGHDug4tm6q/fe4kkqakuovnTI9qxL/8hbPx3j6a6SF6DJaetUd0EoVbSUztj+xs7ukInzwy5XtcDW3s9VUFed/evRtxezi1ei0lTXcR1y+nmaM257+9yCe3KqXWyFwR15SfXCysAAACQHW2HAYRS/MUkAMhFZ0DthaVY6Hr2LC2GAbcurX+91fD6TXv0YgGC18RAKp+ctka9J00464SX4FWSduw97Gm7VMq5xWuxWbW00XHL6Qoj3ba08dzH5RDalVPrZK8I6sqPl7+FC/X/GAAAQLkgfAUQWm5eTAKARFsrV6p77HJVEroCJSexJernNm6Xhy66riQHUvnipjXq3Y925H09+cZc2MJoaajVncsWZP2ducJIq5ctPC9wK4fQjnmnBHXlKpcLKwAAAJAd4SuA0HL6YhIAJNo5ZrkmVxwndAVKUGIg4KaVqFepAql8cdtGuNDmz/A/iIm3eEV+3bR4ltatbFZzmgCuOVqjdSubdePii867vRxCu3JpnZwrgrryk8uFFQAAAMiuePr5AEAAblo8SzOnVOue1k61e2y/ByAcgprram3sbe4pAlcgF4mBgJdWom7Nnx7RF953SUFesC7E55OL5miNPnD5TH3z8Z2+7zve4rWYArty1NJQq5aGWnX0Daqtq19HTpzRhKrRammozfjcr1raqBVr2x1dGFCMoV05tE72Qzyoy1ZdT1BXWrL9LdwcrdFtzNcGAADwpLz+IgAAj4KoAAFQGoIKXaVY6PraUERLTn+r8AdH0bm4LqIXy2iOYCElBwL5bAlqJH3mXU0FDZGKvcXptfOnqakuoplTxqnnwHHf99/W1U/4WiBNdRFXz3Wph3bl0DrZLwR15cnrhRUAAADIjPAVQKit37TH0Xw0AOHTUblclQGGrkNDor0wzkPwmt5dNyzQ95/are29h0fclyoQyFdL0KDCh2JvcfqLHa/q1qvmaf+RU3nZf7F//mFXyqFdvHWym8ryYmud7CeCuvLl9sIKAAAAZEb4CiC02rr6CV4BjLC1cmUgM10lQlfAi+jUav3zYy9pz8CxEfctqJ+UMtTxqyXon189T9WVowIPH7x+PkaF6f7R3j2g77Z16/jps3nZf6m1eA0quAoyMCvl0K7UWyfnQ76CulL8/gAAAABSKa2/UgHAR2taOwleAZxn55jlMiaYua6EroA33ftHhq5x23sPacXadq1etlA3Lr7o3O1+VdhNi4zVLS1RX/aVC6+fz6ff1ag1Py/M70M/3PRy3vZdjBWTqbR19WtNa2fKKsol0RqtylP1Z1DHTaUUq+tKvXVyKSim71EAAADADxVBLwAAgtDRN+iqfRiA8vZS5XJ1jy38bNd46Bo9+QOCVyBPhqx0x8Zt581FbaqLqH7yuJz3XSztbuOtUd1ojtZo1dImrVvZrGaX23px7FT+Wj2XQpi3ftMerVjbnvb3z6e7B7Ribbt+5HNIHdRxy81Ni2dl/FlpjtZo3crm8y7ygDN8jwIAAKAcUfkKIJQSX4AFEF4vVRY+cI2zVjo5VKH5p79f+IMDITNkpXtaO8+rnLp2/jSte2p3Tvstpna3XlujJraDXffkbn3/qd15aUVcPcb/56pUWrw6HXURv1Cgfso4X6r8gjpuuSrl1snFiu9RAAAAlCsqXwGEUrFUqgAIRrzSddSoYFoMnzobq3YleAWc+8iVs/XF91+iBfWTPG3f3j2gjr7Bcx+vuHJ2zmsqphAg3hq1Iss5LV1r1Ka6iL5y/WVafUP2fXjxRz5XBJZSi1c3oy7iFwoEcdzbN2zTfW3d5/2cYKSmuohuaYnqU0sbdUtLlOA1B0H9bAAAAAD5RvgKIJSKqVIFQOEEHbqeHQ5dm2gxDLgSqRqt6y6drpaGWm3vPeR5P8mth2fXVHveV7R2fNGFLn60Rs22Dy+aozX605aoIj79/lVKLV69jLpIvlCgUMftOXBcX/7p87ru7l/pxm89SacY5FVQPxsAAABAIZA+AAilUqiSAOCvnWOWy5hgQtehITHTFcjB4IkzWrG2Xe9bOCOn/SR3vvjENfN0+4btnva1a/9RtXX1F93vFH60Rk21j1cHT3pqSZzYGviT1zTozod2uNzD64yk1Tcs0E2LZ3neR6F5DTDbuvpzCvdzDU7jczZXL1tYEiE3Sk9QPxsoPrTyBgAA5YjwFUAoNdVFtCRa4/pqawClJ6i5roSugL+GrPTTZ/bmtI/kzhczp3ivfLUp5sgmC/IF5aa6SM7HStzHjd960lPwmtga+Nar5qmjb1AbtvS6Xkt8X6UWBHoddZHriAw/RmwwZxP5FNTPBopHW1e/1rR2pvybfEm0RquWNnLuAQAAJYvwFUBorVraqBVr2x3PGQJQWoIMXa2V5p4idAX8lut/2a8OntC9rZ3ngtA1Oc4PjLfATA45y+UF5Xh4vGv/UU8XrN25bMGIsPSrN75JTXURff2xLg06DFGaozW6rUSes2ReR13kOiLDrxEbQw4uMgC8COpnA8Vh/aY9+tzG7Wn/Fqf6HgAAlDp+awUQWi0Ntbpz2YKMf/QBKD1Bha5SLHTtGarR209/vfAHB5DVNx/f6fs+k1tg+vWCcpBVs5nCYzeOnTqb8vZbr5qnW6+ap0ee26eNW3o0eOKMIlWjtWzRTM2pHa8HtvZqx97DkqT5MyL6wOUzS7YFpdfQMtew08+wNN1FBkAugvrZQPDauvod/Q1O9T0AAChlhK8AQi0+M8zrvDcAxSPo0JUWw0A4JbbAbOvq1x0btmet0M30gnLQVbPZwmM3srUHve7S6bru0unnPm7r6tfnf/zseZ/7Yy++pm8+vrOkKoYTeRl10Ryt8aVltJ8jNpizCb8F9bOB4K1p7XT8fwzV9wAAoFRVBL0AAAhauqoMAKXhhcqb1T12uUaNCqbF8NmzUvTkDwhegZCKt8Bs6+rXx7+/2XFr5PgLyonWb9qjFWvb04YR8arZH216OZclp+W0GskpN+1Bg/7c82nV0kZVOPz/qcJIty1tLPhxs2HOJvIhqJ8NBKejb9D1RSHx6nsAAIBSQvgKIPR4MQkoTVsrV6p77HKNGzVE6AogMC0NtVq/aY9uXtvueIZpXOILym7bMLZ19Ts6RkffoO5r69a9rZ26r6074wvYbqqRnHBaqZSvz71YxEddZAuZKoy0etlC3yq8nB7XCeZsIh+C+tlAcLyev0vtvA8AAFCSf0EZY2ZK+j+S3iNpqqS9kn4s6cvW2gMO97FL0uw0d/dZa6enuQ9AmTl2mspXoNTsHLNcxgRT6Up7YQBxzdEavTZ4Up/buF3WY2gZb+fqdxtGt+2LvVQjZeKmPWgYWlDetHiWZk6p1j2tnWpP8Tw3R2t0Wx7aKmc7rlP5fL6DnG+M4AX1s4FgeL3wmQumAQBAqSm58NUYM0/SbyRNk/RfknZIWiJplaT3GGNarLX7He7ukKSvpbj9iA9LBVDkMr0oCaA4BTXX1drY29xThK4AYuItMHOtFj1y4kxObRhThVTZ5rbGW/iuXrZQNy6+SJK/VUVGsXmO97Z2Zg3T/P7ci1lLQ61aGmoLHjYmH/c7T3Sr9+Bxx9vna85m0PONUTyC+tlA4Xmtoqf6HgAAlJpS/O3lG4oFr7dZa++N32iM+SdJn5H0t5I+7nBfB621X/J9hQCKXrYXJQEUl6BCVykWuh4ZGqMFp79b+IMDKErxFpgXRMbmfBHXhKrRObVhTA4m3LbwrZ8yTi0Ntb5WFVlJ657afd5t6cI0Pz/3UtFUFwlk7fHjNtVFtGJtu6Pfg/M1Z9PLBQIof0H9bBSLMITPXi+o4EIMAABQakoqfDXGzJV0naRdkv456e4vSvqYpBXGmM9aa48WeHkASoTTFyUBBC/o0JUWwwCSJbbAvK+tO+f9tTTU6mfP7vO0barA1GsL33xXFaUL02hBWXjxOZvZfh/O15xNrxcIAOUqTFXgTXURLYnWuLpwKV/V9wAAAPlUEfQCXLp2+P0j1tqhxDustYOS2iRVS3qLw/2NNcbcbIz538aYVcaYa4wxo3xcL4AilGt7QAD593Tlreoeu1yjRgXTYvjsWSl68gcErwDOaWmYqo+8dbbec9l0XRAZKyn3ADD+gnLnq4Oetk8OTHNp4ev1hf2PXDlbK94yy9G5Oh6mJVa70oIyGDctnqV1K5vVHK1JeX9ztEbrVjbnperUywUCQLlav2mPVqxtT3vujl+48qNNLxd4ZfmzammjKhz+fp+v6nsAAIB8K7W/WC8eft+R5v5OxSpjmyS1OtjfdEnrkm7rNsbcYq39pZMFGWM2p7lrvpPtARSWlxclARTWzjHLZUwwoSuVrsDrjKQbFs3Uxq09XLQkqa1rv9q69p/7eEm0RhfXTfC8v/gLym1d/frpM3s97cPPFr63tEQ9VSN9+Q8v043felLWQ7WtFI4WlMXaSjSIOZthmvELZBPWKvCgq+8BAAAKodTC10nD7w+luT9++2QH+7pP0hOSnpM0KGmupE8q1rr4IWPMldbaZ7wvFUAx8vqiJID8C6rFsLWxt7mnCF2BRFbSxq09uv098/Xf2/dqW8/IX8GbozWaOWWcNm7pVdjy2ae7B7Rpl/cLum5YNFMtDbWx4NLD9qnaMObawnfV0kbXs0BzDdPKuQVlqbQSLeSczTDO+AXS8domvhzctHiWZk6p1j2tnWpPcY5MbPEPAABQikotfM0m/nJt1l9frbVfTrrpWUkfN8YckfRZSV+S9AEH+7ki5UJiFbGLsm0PoLCYDwYUn6Dnum4ZmqsbTv9N4Q8OlIAhK33t5x06fnpoxH2X1U8898LoskUzddfDO1IGtOXMabVnKvdv7tHMmnGeOnKYNG0Yc23h66Uayevc23965EVdeuEkTagarRsW1eu3uwZchb7Fbv2mPRmfx3QzcMsdM36BGKrAg6m+BwAAKJRSC1/jr+ZMSnP/xKTHefEvioWv78hhHwCKFPPBgOIRdOh6fGiULjmdPH0AQLJUwaskPdt7+Lzw6CeffJved+8Terb3cIFXWJqspLsf9TbL8v0LZ6SsBvKjha/baiSvodjDz/Xp4ef6zn0crR2vXf1HM15FWyotKMPaStQJZvwCMVSBv66Q1fcAAACFUmp/wbw4/L4pzf3xS6DTzYR14tXh9+Nz2AeAIhWWF7aAYhZ06MpcV8A/yeHR537vDY7b1sK7xmmpX6T2q4Wvm2okv0Kx7v6jMkaKTh2v7v6jKddZKi0ow9xKNJswzPgFnKAKHAAAoLyVWvj62PD764wxFdbac5fhG2MiklokHZf0VA7HuHL4/c4c9gGgSDXVRTRj0ljtPXQy6KUAodNVuVyjCF2BspMYHjltW1uOmuomqKPvSEGOlSnw9DK3NR0n1Uh+hmLWSrv3H9VdNyzQsVNnS7IFJa1EMyvnGb+AG1SBAwAAlLeKoBfghrX2JUmPSJoj6S+S7v6yYtWq/26tPSpJxphKY8x8Y8y8xAcaYy41xtQk798YM1vS14c//L7PywdQJKZFqoJeAhA6O8cEE7xaK509K0VP/oDgFcijeHgkxdrWrlvZrOboiF+38+6aiy/Q7Jrqgh9XkvYdOlGwY2UKPOMBeEWW861fLXzjYZpfhqy0cUuvbmmJ6lNLG3VLS7SkgrcHtvZ62s5rC9JStGppY9bvz7hSmfELuOX13LthS0+ozhcAAAClqqTC12GfUKw18D3GmB8bY+40xvxC0mcUazf81wmPrZf0gqTWpH18SNIrxpiHjDHfMMbcZYy5X9IOSQ2S/kfSP+b7EwEQjJrxY4JeAhAaL1UuV/fYwrcZjle6EroChZP4YnBLQ63W33qlHvnMO7TiLbM0c/K4gqzhj5fM0u6BYwU5VrLDBWoF6aQKMFsAvnDmJK1b2awbF1/ky5rchGlOJIb5paKtq183futJffPxlzxtH6ZWooW+QAAoRl4vXInPW//RppfzsCoAAAD4peTC1+Hq1zdL+q6kZkmflTRP0j2SrrTW7newm8ckPSApKmm5pL+UdJWkX0v6iKT3WWtP+b54AEVh/ozSqZ4ASlU8dB01Kphq14ND4zT3FKErUEhPvjTy1/Ctew7oP9r3qOfg8bwff+HMSeotwHGC5KYKsKWhVrctbdRl9RNH3Let55DWtHb6Vj3lNExzo5Qqu9Zv2qMVa9tdtxtO9OpguEZiZLtAoDla4+sFAkAx8nrhSnzeeimdJwEAAMKmJIdFWGtflnSLg8ftkjTiV1lr7S8l/dL/lQEoBXOmjg96CUDZeqmy8FWuccx1BYL16PN9auvqP1el1tbVX9DZr9t6Dmn/kWCvn5w5eVzegma3VYDrN+3J+Pw/3T2gFWvbtXrZQl8CrpsWz5Ik3b5he877kkqnEtSv7/PvP7VbC+onhSpsjM+J7ugbVFtXf0nO+AVykcuc9MR56wAAACg+JRm+AkAuNmzxNosLQHqErgCszn8heE1rZ8GC17igK1+vmX+B/qN9j++fd3O0RrctbXT8IrvTQDBePVU/ZZwvL+AfO3U2533ETagqjT9V/fo+t/L3a1FKmuoihK0IrZsWz9LMKdW66+Ed2tZzyNW28Rbt/PwAAAAUn9L4ixYAfNLRN5hTSzgA5yN0BZAocVZnGP+/XXHlHF1WPymnSsh/XXGFeg8ez6kK0E0g6Gf1lJ/VqqUQQPr9eyWVbEA4tTTU6gOX17sOX6XYxTaErwAAAMWH8BVAqDAXB/DPzjHLZUwwM10JXYHiFdb/a5ujNecq+GZOqdY9rZ1qdxnMVY2u0HWXTs9pHV4CQb+qp/yqVo0/l8UuH9/rVLIB4eT14pVSadEOAAAQNhVBLwAACok/ToHcvVS5XN1jC1/xGg9doyd/QPAKFLFd/UdD9/9thZFuW9p47uOWhlp95frLXO/nxJkhrd+0J6e1eA0E/QgS/ajYTH4ui1m+vs/DegEDEGZeL14plRbtAAAAYcNvaQBChT9OAe9oMQzAib7DJzSndnzQyyiYCiOtXrZwRPDoNUBb/dAOXT5riufKxyCrp5rqIloSrfHcitco9XNZrPL1e2X8a9HRN6i2rv6cWlADKA1ez3ulcr4EAAAIG1IIAKHCH6eAe4SuANyom1QVmv9vm6M1um1pY8rP12uYeeDYaV1396+0JFqjVWn2nUnQ1VOrljbq5u+0y8vI25vfMls3Lr7Il3UUQv3kcXnZ76uDJ3Tjt55MGWJ7/b4AUNy8XLxSKi3aAQAAwoi2wwBCJf5HLYDsOofbC48aFcxc17NnaTEMlKI5U8erqS6iy+onBr2UvHjjzEn64vsv0SOfeYfW33pl2hAs1zDz6e4BrVjbrh9tetnVdkFXT7U01Opdl9R52nZaZKwvayiU3oPH87Lf77fvSRvAeP2+AFD8Vi1tVIXD37lLqUU7AABAGBG+AgidVUsbFUABH1BSdo5ZrtEBVLsSugKlLx7ife733hDwSvLjHz70Rt3SElVTXUQdfYO6r61b97Z26r62bnX0DZ57nB9h5pCV7ti4zVULYy8XmvlVPRV/Pk6fHfK0famNh8jXzFebpWzYy/cFgOLX0lCrO5ctyBrApmt3DwAAgOJRWn/dAoAPWhpq9el3NeruRzuDXgpQdIJqMWxt7G3uKQJXoJQlhnhHT+YnmApS/PNr6+rXmtbOrG1hZ9VUa8/AsZyOOWSle1o7Xb3Ivmppo1asbdeQg96/flRPZXo+3Ci1ICHIsNjL9wWA4nfT4lmaOaVa97R2qj3FOTVTu3sAAAAUD8JXAKG0ammTfrz1FXX3Hw16KUBRCHqu69khqYFKV6CkJYd4G7f0BLgaZ+KnPCfzSeOf3/pNe/S5jdvTBpvxtrCrly3UX1wzT7dv2J7zOtu7B/TIc/vUe/C4jpw4owlVo9XSUJu2WjVePZVpnfHPKdfqqWzPh1OlOLsw6PCjvXtAHX2DJfe8AcispaFWLQ216ugbVFtXv6PzPgAAAIoL4SuA0Pqb6y/Th7/THvQygEAFHboODYn2wgiFBfUTtb33cNDLcMWY7O1Pzz1WI0O8wTy1ZPVLvHqo58AxxyGlJEdBY7wt7LqVzYrWjvflYq+Prds84rbEKttkhaieauvq9yV4NZKWLaofcXuxBw/xFs+5Vvzmoq2rv6ieEwD+aaqL8PMNAABQoghfAYRWS0OtGqdNUOerR4JeClBwhK5A/iycOUkfuLz+vMCoratf23ufD3ppjjVHa7RsUb2jYK1+cpX+/oNvHBHiRfLYkjVSNdpTuGskvf+NM/TJaxvPe0HbaUh547eedBw0xtvC/s31l+nmte2Og2w3Eqtsb1x80Yj78109taa1M+fgVYpVHt++Ybs2bOnVquHqaSdtnYuBmxbP+eB17myxB9sAAAAAUMoIXwGE2skzQ0EvASiorZUrNbniOKErkEfbeg7pHz/0xvOCjJ89uy/AFbkTb6/b0lCbMZSM6z14QmtaY3PUEwOxZYtm6uHn+vKyxsETZ2SM9P6FM9Q4LXIuPHpt8KSnSk8nIWVH36DrCsf27gFdEBmr1Q5aAHsVr7KtnzIubSCZj+opL89HNk93D+jm4a4k6Z6qbIFzoTlt8ZwvbufOOp1XDAAAAADwjvAVQGh19A1qz8CxoJcBFMzOMctlTOGrXQldEUaJrUDbuvr1n5uLf/6pNHIGaDyUXNPaobsf7Uy73dPdA7p5bbvuSgjErrt0uucKVSeslR7ctlfrVs46t96mukhOlZ6ZQsq2rn5P62zr6tctLVFHQbZX8SrbQoZmXp+PbJzkl04C50LK1uI5n9x8/m7mFRdDsA0AAAAApYrwFUBo5etFQ6DYBNVimNAVYRZvBZot7Cgm1ZWjtOqdjSNCl7aufn0tQ/AaZ610+4bzA7FPXtOgOx/akZf1SulDx3xUenpt7xrfLh5kP/LcPt2+YZsOHDvt5/LU3j2gjr7BgrWO9fp8+CWIwDmTdNXTG7b06Nk8zXtujtY4/no7nc9bbME2AAAAAJQiwlcAofXkS/uDXgKQV0GGrtZKc08RuiK8JlSNVltXv+7YsN1RJV8xOHb6rFY/tENTqsecF8B+5cHnHX8OVtLfPPi8Hvr0OyRJt141Tx19g9qwpTfrtlfMnqyGaRO0fpO7KuFChY5u27smb5ep3atfEiuu883r8+GnQgfOTiQH/011kbzMhDWKtQd3ys183mILtsOKubwAAABA6Qr+L2YACEBbV78eeT4/c+iAoAUVukpUuwJx1WNG6ePf31wywWuclXR7QtVbR9+gduwbdLWPF/YNnheIffXGN+n4qbP6nyxzb7fsPqijJ71VUxYidPQaRLU01BasArqQ1ajFEswVMnD2Il8zYaO14x1/DbzOKy62YDssmMsLAAAAlL6KoBcAAEH4/I+fDXoJgO+6Kpere+xyjRoVTLXr2bNS9OQPCF4BSbdv2J63Waf5Zq1018M7dF9btz7/gLf/Lx/Y+nqla1tXvx5+LnPwKsWC3x37jng6XiFCx6a6iJZEa1xt0xyt0WuDJwvWerqQ1aheno98CLr9sRM3LZ6ldSub1Zzm+VpQP9H1Pnf2H1VH38gLIzr6BnVfW7fube3UfW3d56onvWBER+Gt37RHK9a2pw3L43N5f7Tp5QKvDAAAAIAbVL4CCJ31m/aou/9o0MsAfLO1cqUmVxyn0hWAb7b1HNK2nkOet9+x9/UZl27anXrlJnTMpZXnqqWNjlvIVphYW1i3n//smmr99XvfoI+t2+x8o2GFroZz83zkSzG0P3Yi3UzYloZatXX1a3vv8673mVj1m6lasn7yOE9rLoVgu5wwlxcAAAAoH6XxlyoA+Ogbj70U9BIA3+wcs1zGBFPpSugKIBsv7U69cBJA+NHK02kL2QojrV62UBdExrr+/HcPHNOc2vFaEq1xtW1ztKbgLWJbGmq16p2NuvvRzoIeN3kNpSR5Jqwk/SxLS+504uFotrbWvQePe9p/PoNt5pmOxFxeAAAAoHwQvgIIlY6+Qe0eOBb0MoCcBTXX1drY29xThK4A0ps/4/VqvHxzEjpmC6firTxXL1uoGxdflHFfNy2epZlTqnVPa6faU4SjzdEa3TYc5N7X1u3480jU1tXvqco2CBOrKgM5rhRM4JwPXkPOCVWjHVdLepGPYI95pqkxlxcAAAAoL4SvAEKF2VUodUGFrhLVrgCc+8DlMyXlv22pk9AxH608M7WQTQxCvH7+R06ccV1lG1RgFVRr2kIGzvmu0vT6tWtpqNXnf/xsXoLXfATbfl4EUW5ymctL+AoAAAAUH8JXAKHC7CqUKkJXAKVi/vTX26rms22p09Axn608U7WQTZRLRaOUvcp2dk21PnHNvECDKr+/xvH/5jJ9yQoVOBeqSrOpLuKpzbSkvLT1zkewzTzTzHK5UAMAAABA8SF8BRAq+XwRGMgHQlcApcRI+sL7Ljn3cb7Ck8TWvpkE3cozl4rGxH+3NNRq/aY9+sZjL503PmH3wDHdvmG7NmzpDaxdq5/HjH9dJTlq65xPha7S9NJmOh8dXdwE224qgplnmlmuF2oAAAAAKC78pg4gVML0Ig5K384xy2VMMHNdCV0BuGUk3XXD+aGNl4q+dK65+AK9o+kCVy1fvYZTX/9Fl+7548s9bZvIa0Vj8udXzO1am+oimj89oh37Bh1v0xyt0VeuvyxjcOekrXO+BFGl6aXN9JbdBzwda+aUceo5cHzE7U6DbbcVwUFfBFEK/LhQAwAAAEDxIHwFECp+vggM5EtQ1a7Wxt7mniJ0BeBOptBm1dJG3fyd9oxtZJ343O+/wXUQ47Ul50+eeUWXXjhRt141z9P2ibxUNCYq9nat6zft0Ysugtf455itZbOUva1zvgRVpZmtzXTyz5nXqseVb4uqpaHWU7Dt5UIA5plm59eFGgAAAACKA+ErgNCZVzue8BVFKegWwweHxuny02sLf3AA5/zx4ov0w00v5xxUFsJHrpytObXjs4Y2j7/4as6fj9eQIZeWnHc+tENTqsfkXEnqpaIxUTG3a40Hw06/vqZAs1pzEXSVZrzNtJOq31yqJb0E214vBGCeqTOrljbq5rXtsh4v1AAAAABQPAhfAYTOg9v3Br0E4DxBh67Hhyp0yenvF/7gAM7z9zfEKsXeNGuy7tjgPNAKyu6Bo6qdMPZcVVuqIOdv//t5ffuJ7pyOk0vIkGvId/sGfypJ3VY0xgUVBDpt9+smGJaki+sinsPsQrUgLpYqTaeVwYWslvR6IQDzTLOLt3J2GrwW+0UMAAAAQNiF568ZAJD0yHP7NBiyq+hRvIIOXZnrChSXzldjrVvjQd0XfvysdvYfDXhV6T3+Yr8ef/H1oCp51mNbV78vwavbkKGjb1APbO3Vjr2HJUkzJlVp76ETno5v5V8lqZuKRun1KkMvvAaBbmZ5egmGd+wbdB0Mu50vmqtSq9LMta21U7lcCMA808yytXJO5HQuLwAAAIBgEb4CCJWNW3qCXgJA6AogpW8/0a2rL5527kX1XfuLN3hNJXnW450PvZDT/tyGDG1d/frKg89rh4v5o0742VJWclbR6CaMScVLEOh2lmchKkS9zBfNValVaXpta+22kjiXr/ctLVHmmabhtJWzJBmJ4BUAAAAoEYSvAEKFqlcEbeeY5TKm8MEroStQGu56eId+8sm3uW7nWizisx6trJ7tPexpH1OqK7X+1ivPBS9OQqL1m/bktVWz3y1lsx0rl+BVch8Eepnlme8KUa/zRXNVSlWaiT8bH26epWd6Dmlbz6ERj0u8kMFrJXGuX+9CVeiWGjfnej8r8QEAAADkF+ErgFCJhGh2FIpLUNWuhK5AadnWc0iPPLfPdXvPYjJkpW889pLn7Q8cOy3JebvZtq5+3bExvzNyC9lS1o/gPVs4kxxob9jS43qW53sum+5pbU6DYa/zRXNV6DmqXmT62bisfqIuv2iKpkXGjrhYIZdK4lwrgr1W6JazoGY6AwAAAMg/UggAobJs0Uw9/Fxf0MtAiAQZulorzT1F6AqUmnJokb974FhO23/9F516cNteRyHR/Vt6ZPNcJVyolrJewphkmYLATKGdG+3dA1r5tqinbZ2EakGHUsVcpZktQH2297Cef+XwiAA110piPyqC4/Os72ntVHuKr2/Y5pkWonU3AAAAgGAQvgIIlesuna5I1WjaDyPvmOsKwCv+j5J++szerJWsQ1a6fcO2vFa8xhUqDPIaxsRlCgJznSObbOvLB/zZUZKOvkH90yMvetrWr1CqWKs0cwlQc60k9qsiuKWhVi0Nta5nzpajfLfuBgAAABAcwlcAofPJaxp050M7gl4GyhShK4Bc0SJfjgPVQgSvhWwpm0uoYjIEgX7MkU22Y++gp+3SBaR+VOX6GUplq9KcXVOtxdEaHT11pmBtYL0GqH5VEvtZEdxUFwld2Jos11bOAAAAAIoXv7UDCJ1r5k8jfIXvuiqXaxShKwAf0CK/eHhpKZtLRV8uoYq10v1beka0i5X8mSPrl1QBqV9VuX6HUslVms/1HtamXQPaPXDs3Nv9m2OPTZxDnA+5BKh+tbct1orgUuVHK2cAAAAAxYnwFUDo5NrSD0i2c8xyGRPMXFdCV6C8LKifqOsune66vSf85zZAylS56TSYyzVUSZyFG5/36ccc2VTmz5iox158zfV2yQGpn1W5+Qqlmuoi2rrngDZu7XE0hzhx1qpfcglQ/Wxvy9xW//jVyhkAAABA8SF8BRA6zEmCX4JqMWxt7G3uKUJXoNxs7z2sG7/1pJbOn6bf7hoommrFsMkWICVXtx4+cVprfp6+uvTp7gHdvLZd7184Q43TImkrYr2EMcmS533m46Kz5miNPnB5vb75+Euut81XVW4+Q6lcZq36JZcA1e/2tsxt9Y+frZwBAAAAFA/CVwCh80zPwaCXgBLHXFcA+fR094B+u2tAyxbVa8Pm3oLMNQ2791xap0svnJQ1QMplLqm10k+e2Stp77nbUlXEuglj0kmc9+n3RWfxAMiPqj2/qnLzHUp5nbXqp1wC1Hy1t2Vua+5o5QwAAACUp4qgFwAAhfabl/YHvQSUqJcql6t77HKNGhVMtevZs1L05A8IXoEQGLLShs29il4wPuilSJJa5k3VH7xxRtDLyJtLLpykTy1t1C0t0bRh0vpNe7RibbuvLXzjrWp/tOnlc7fFw5iKHP+fic/79HMOqpH0voUX6oLIWEmxoNjpOlMFpPf+ojP3NeU5lMpl1qqfcglQ40G5G7S3LZybFs/SupXNak7zNWqO1mjdyua8tLMGAAAAkB9UvgIIlUee26djp84GvQyUGCpdAQTBStr52tGglyFJeucldWqqi+in2/bKlmEp7td+3qErZk9JG3D5OZc0WapWtdnmajrV1tXvayhpJf3kmVf0k2deOVe167Vqr62rXw8+szf9Rg7d3Dwrr6FULrNW/Qwvc600pr1tcaOVMwAAAFBeqHwFECobt/QEvQSUmJ1jgpvrSqUrgGLR0lCrNa2dZRm8Sq+3ik3nzodeyOv83VTHb2mo1fpbr9Qjn3mHrrn4Ak/7PXLijKeqx6a6CbrswokZHxOv2jUynqr21rR2+tJSe1qkyoe9pJfLrFW/5VJp7LSimva2wWqqi+iWlmjWSnwAAAAAxY3KVwChMpiHF8JQnoKqdqXSFUCxiYdqfrbbLUbxVrGJYUdbV79WP7RDz/YeDuT4UiyMmT9joh578TXX+4y3HHY7R7aj74ijx8WrdtetbNb6W690XLXn16xXyfss1HzvPx/rynU+aLaK6uZojW5LmkEMAAAAAHCP8BVAqETy/AIdSl+Qoau10txThK4Aike8es5r69VSk9gqdv2mPXlrNezk+PGP17R2eg4q4yGa09DOi3jVbjxodVKp5+f306uDJ3Rva2feWrTmMms1H3INUGlvCwAAAAD5RwoBIFSWLZqph5/rC3oZKELMdQWAkVa+LaqWhlpt2X0g6KUURLxVbD5nvDo5vpR7+Js471Pyb45sKumqdtPxsyXvNx/fed7H8Vm0foWfuc5azQc/AlSnQTkAAAAAwD3CVwChct2l0zVmdIVOnRkKeikoEoSuAJDe2l93q3FaJO+tXYtF/PNc09pZ8OA18fi5hr/J8z7j0oV2G7b05NxaOblqN5N8fj/FZ9GuXrZwxJxZr9y0bU733OdDsQSoVNECAAAAwPnC8SoKACS47MKJ2rLnYNDLQMC2V/6pJlScInQFgAziMz3vXLbAl/0ZSQFkmo7Fg8mg5ttWjxklKbfwN928z0SJoV1H36C+/NPnvR0sQWI1a7YwLt8zRePft/VTxvlyrFxnrZarTG2x/a5ABgAAAIBSQvgKIHT6Dp8MegkI2M4xy2VMMHNdCV0BlJohK23c0uu69WqyubXj9ZXrL8tphmk+xVvF3tfWHdgaNm7p1eWzpnh+frLN+0zFr/mrE6pGOw7jmuoiilSN1qCP7YeTJc6i9UOus1bLTba22PmoQAYAAACAUkH4CiBUOvoG1XvweNDLQECCajFM6Aqg1LV3D+iuGxbot7sGXFdkzq6p1ieumaebFs86d5vTFq6Fktgq1s95pG61dw/oga09nrb986vn6vb3vMH1dn59vodPnM74dU0M4940a3Jeg9c4t7Nos/Fj1mo5cNoW2+8KZAAAAAAoFYSvAELlga29QS8BAQgydLVWmnuK0BVA6du1/5jeu3CGHnxmb8bWwUbSuy6p05XzpqYMpZy2cC0UI+nT72w6Fw4FPd92x95BT9tVV3pbtx+f7/zpEa35efZWyfEw7sPNszI/0EduZtE6VSyzVoPipi223xXIAAAAAFAKCF8BhMqOvYeDXgIKKKjQVYqFrieHKjT/9PcLf3AAyINvPv5S1sc4bb2arYVrIVlJ//Roh3606WVdM/8Cvb3xgkDX45XXEDXXUKxi+P9YN2HcYy++ltMx3QiykrkceZmJ7HcFMgAAAAAUO8JXAEDZCTp0pcUwgLC5cm6NvvyHl40IVzK1Z03XwvV3ew7qv555xdf1VRjpA5fXa+OW3rRVuz0Hj2vdU3u07qk9eZ9Hmsn8GRM9hZNeQ9Smuojneb4VJlY1/E+PdrjarudA4UZABF3JXG68zgjORwUyAAAAABQr/hIFECrzZ0QKWm2BwtpauVKTK44TugJAgT25c0Bfe7RD37j5CkmxoGVNa2fKQG9JtEarEqpjk1u4fnznb31dW4WRVr4tqrW/7s7YLjlRrsFrc7RGyxbV6/YN211v+4HL67V59wFXYWhztCanYGvV0kbXc3jjVc4dfd7aJBcK7W795bWSmApkAAAAAGFSEfQCAKCQPnD5zKCXgDzZOWZ5IMGrtdLZs1L05A8IXgGE2v88u09/+9/Pa/2mPVqxtj1tePh094BWrG3Xjza9POK+tq5+bd1z0Lc1NUdrtG5ls57pOVSw+bJ/+a4mrb/1St20eJaWRGtcbRsPUVctbTzXzjebCiPdtrTRw0pfF5/D6+SYl104Uf+64gqtv/VKtTTUeg7V6ieP87SdG7mG0hjJayUxFcgAAAAAwoS/gACESlNdRJUVRqcL9Qos8i6oFsPWxt7mniJwBYC4bz/RLWNi58dMhqx0x8Ztqp8y7lxl4vpNe/S5jds9h6R/fvVcTYtUjWhx7GVGZS7u/nmHrpg9RS0Nta4qShND1HgYmu35qDDS6mULs1Z3Zmr/HOd0Du+zrxzWx9ZtPlfB7DVUu3b+NP1H++68heJ+hNIYyWslMRXIAAAAAMKE8BVAqHT0DRK8lgnmugJAccoWvMYNWeme1k61NNSqras/p+BVinW3SFXleO8vOr3v1AOb8HnlEqJmC0MX1E/Umy6aor7DJ3RfW3fKQNVN+2fp/Dm89/6iUw8+szdtq+Z4BfOn39mU+QlJY8WVs3VZ/UTdsWG743bQbnz6nU0lGfg5CcqD5GVGMBXIAAAAAMKG8BVAqDywtSfoJSBHhK4AUD7auwfU0TeoNa2dOQWv86dHUoY7bV39evCZvTms0Jv459VUF8kaosZnp6YKChPD0Hgg9+rgCf3u5UPa3ntI23sPn/f4xEA1WyVxPDxdvWyhblx80Xn3vTZ4Uv+9LX3wGjdkpa/9vEPzp0e0Y5/z2a/xMK6pLqJnew9r3VO7HW/rVKTE2ty6DcrzwWnw67WiGwAAAADCorT+IgWAHO3Y6/yFQRQXQlcAKE8PbO3NuS1wuv8a1rR25qWq0om2rv5zwVVLQ60uiIzVA1t7tWNvLDCdPyOStlo3WTyoXL9pj+7+eUfWQHXl26Ja++vurOFYqvbPklyF4fHHVRh5CuOmRcY6O5BLXmfRBsFNUP6mWZN9r4z1UiHtZ1tsAAAAACg3hK8AgKK3c8xyGRPMXFdCVwDlaswoo1Nng2/F/5uu/pz38cK+wXOVpnGFnvWaLB7+pQu2HnvxNf30d3t1zfwLtOLKOSPWnhywvTZ40lGL3iEbm73rVGL75/ix3T5vO/YN6jPvatSan2cObVOFcV5nxmaTr/36zWnL7SEr/dWGbSnvy6Uy1muFtJOK7mWL6nX01Bnd29pZlC2UAQAAACBfSuMvUgDwyfwZE/XYi68FvQw4FFS1K6ErgDAohuBVkp7pOeTLftY9uVtfuf6ycx+3+RDq5mJC1Witae3Q1x5NX33bc/C41j21R+ue2qMl0RotnT9NrTteTRl+Vo2uyFsVb2KbZK/P28SqSq1b2ey6vXK+qiJLpdoy15bbUuYW0pm4CX5TVUinaos9oWq0qseM0oYtvbp9w/YR+ypUC2UAAAAACBLhK4BQufyiyUEvAQ4EGbpaK809RegKAKXmsR2vnvdx0G1nv/ebXdq1/5jjxz/dPZCx4vTEmSE/lpVWvE2y1+ftyIkzacO4TBWPTXURLYnW+FqlHJ8pW+z8rM5OF5Bm4ra9dGKFdKJ4W2zJeyUtAAAAAJQTwlcAodJ78HjQS0AGQc91PTI0RgtOf7fwBwcA5Kzn4PHzWg8H3XbWTfBaDOKhq9fnLXG7xDDOiVVLG7VibXvOFaDSyJmyxSQ5lH518KSv+88UkKZai9vgN7FCOpVcK2kBAAAAoFwQvgIIlaCrYJBa0KErLYYBoDzEqzel0mk7Wyzi4anX5y2X57uloVZ3LlvgKLjLJNVM2WKQbu5vPmQLSBPX5EXiz1gyvyppAQAAAKDUEb4CCJWgq2Bwvo7KD6uywhK6AgB8kXiRVT7a2ZazeAjm5Xnzo83vTYtnaeaU6owzY6+dP02/2PGqq5myQcvWhjcfMgWkcbm0l04lH5W0AAAAAFCqSCEAhEqxvSAXZjvHLJcxwcx1JXQFgPKUfJGVn+1s42ZPrdbuFC2FF86cpG09h/w7UAElh6dunjc/2/w6mRl761XzXM2UDZLTNrx+cxKs+tFeOlE+KmkBAAAAoFQRvgIIFapgghdUi2FCVwAof8kXWfnVzjbRBxfN1Lsvmz4i/Gvr6i/J8DVVeOr0ectXm99sM2PdzpQNips2vH5yEqz63V7a70paAAAAAChlhK8AQueGRfWErwEIMnS1Vpp7itAVAMpZuta32drZujWhanTK8O9nz+7Led+Flik8ddIGuBjb/BYLL214/eLka+J3e2m/K2kBAAAAoJTxlw6A0Nm1/2jQSwiVoEJXiWpXAAiLbK1vU7WzPXb6rL75+Euuj5Uu2Cq1EMlJeOqkDbBfSqWVsFNe2/Dmys38XT/bS/tdSQsAAAAApay0XiEAAB/s2DsY9BJCgdAVAFAIblrfJlesbt59wLfKv1ILkb5y/WWOQ7p8tvlt6+rXmtbOlF+HJdEarSrR6tog2um6nb/rZ3tpvytpAQAAAKCUEb4CAHxF6AoA8FvthDHqP3JqxO25tr71s/Kv1ObKr3tyt75y/WWBrmH9pj0Zg7+nuwe0Ym27Vi9bqBsXX1TYxeXIayX0n189V9MiVedVAG/dcyBv83f9bC/t588TAAAAAJQywlcAoTN/xkQ99uJrQS+jLO0cs1zGBDPXldAVAMrXwNFTuuuGBTp26qyvbWndVv5dEBmr+9q6065h6fxpJRO+Prbj1UCP39bVn/V5l6QhK92xcZvqp4wrqQpYr2v9wOUzR3xfN9VF8jp/16/20n5W0gIAAABAKSN8BRA6H7i83tOMN6QXVLWrtbG3uacIXQGgnA1ZaeOWXq2/9Urf9+2k8u/a+dN0/5Ye/dWGbSPuj7fGvSAyVt/9zS7f15cvPQePq6Nv0Le2r26DuzWtnY4qJKXY1/+e1s6SCuv8bsNbiPm7frSX9rOSFgAAAABKFeErAMAzWgwDAAqlvXvA17AwUaZgK1vL16e7B/Th77T7vqZC+LPv/VZ/t2xBTkGYl5mtHX2DriuE8/n1T+RnuJmPNrz5nL/rl0IExQAAAABQzAhfAYROW1d/0EsoeYSuAIAgtHX15zW8SQ62nLbGLVW7B47lNFPV68xWr7+L5fPr7yVEzibsbXhLISgGAAAAgHwgfAUQOs+9cjjoJZQsQlcAQJB+/nyfJBWsgs5Na9xS5XWmai4zW4+cOONprV63y8ZriOwEbXgBAPC3swQAAKWA8BVA6LR1vRb0EkoOoSsAoBi0vbRfbS/tl+StGtHNC39eWuOWKi8zVXOZ2TqhytufoV63yySXENkp2vACAMIqH50lAAAoBYSvAEKlo29Qew+dDHoZJWXnmOUypvDBK6ErACATN9WImV74u6x+oi6/aIqmRcaeF4iFbUyBm5mquc5s9foiaz5enM0lRHaLNrwAgDDJZ2cJAACKHeErgFAJ2wupuQiq2tXa2NvcU4SuAIDMnFQjZnvh79new3q29/yRBEuiNZo1pdrv5RY9pzNVc53Z2lQX0ZJojasAtzla43twmWuIDAAAUitEZwkAAIpZRdALAIBC2rX/aNBLKHovVS5X99jlGjUquGpXglcAKC8VRvqzt0fVHK3xfd/xasRUnL7wl+zp7gFt2NLjw+rS8/u/WD/253SmqtfZq8+98nrIvWppoyocLrrCSLctbfR0zExyCZEBAEB6XjpLAABQTqh8BRAqfYdOBL2EosVcVwAoHxdNGaeXDxzPeT/jKkfp+OmzOe1jVs043bls4blqhvjcy139R/Xb3QfOC+S8au8e0CPP7VPvwePnzdN088JfMo+bOVJhpNXLFurbT+xU56tHct5fc7RGF9VU6/7NuQXGTmeqep29umFzj5bMqdGNiy9SS0Ot7ly2IGs4Hn+u8lEN4zVE9rodAABhQGcJAAAIXwGETN3EqqCXUHQIXQHAX7NrqjW5ulLP9BwKbA09B47LKPcAMdfgVZIOHDt9XnD22uBJPfTsPtcvymXzsXWbfd1fviycOUnvXTBDa9u6PQev77m0TpdeOOm8GbX3tXXr/hyfAqcBp9cg1Or81oI3LZ6lmVOqdU9rp9pTfD80R2t029LGvLUh9Boie90OAIAwyHU8AQAA5YC/GgGEypza8UEvoajsHLNcxgTXXpjQFUA52j1wTHv8zRVds4q1oTUmds4N0uCJM3rkuX267tLpWeevlrsF9RP14eZZumPD9pyC8UsvnKRPJbXhzTWgdDNT1cvM1rh4a8H4elsaatXSUHuuIjqxcjnfL8B6fc6YSQcAQHp0lgAAgPAVQMjwYllMUNWuhK4AwqIYskUraX5dRJPGVaasKiyk2zdsU3f/Ua1+eEfgYXBQjKT3Lbww5+BVSl15mUsg6mWm6qqljVqxtt1TkJ6qtWBTXaTg1S5enjM3ITUAAGFEZwkAAKSKoBcAAIXUVBfR/OnhfcHspcrl6h67XKNGFTZ4jYeu0ZM/IHgFgALasW9QyxbV65HPvEPvubQusHUcOHZadz4U7uD1rhsWauPWXl+C+XQXk61a2qgKl/+/e52pGp/Z6vXXCa8tCf3m5jnzElIDABA2dJYAAIDwFUAIHTkZvlY2QYWu0uvB69xThK4AEIQ7NmzX7/Yc1KUXTgp6KaE0fWKVvv/RZu09fFwv7hvMeX+ZKi/jgajTMLE5WqN1K5t14+KLPK3lpsWzdMOimZ62LZbWgk6fM68hNQAAYRPvLOEGnSUAAOWGfg4AQqWjb1A9B44HvYyCCaq9sESLYQAoFlbSHRu36cPNs4JeSii9OnhCz/Ye0tce7cx5X04qL29aPEszp1TrntbOlO2mZ04ep2vmT9OKK2f78iLnpfUTdf8W99sVU2vBbM9Zc7RGty1tJHgFAMAhN+MJ6CwBAChHxfMXLwAUwANbe4JeQkF0VS7XKEJXAMCwISs903Mo6GWE0pCVvv5YV87tht1UXrY01KqloVYdfYNq6+rXkRNnNKFqtFoaan2vKimX1oKFfM4AACh38c4Sn9u4PWMAS2cJAEC5InwFECo79ube7q/Y7RyzXMYE116Y0BUAitO2nkO6rH6inu09HPRSQmcwxxa7s2uq9XfLFqR9YTJdYBh/y6d4a8GnU1SMplPMrQUL8ZwBABAGdJYAAIQZ4SsAlImgWgxbG3tjpisAFL/LL5qi51857KgFXLG55uILJEmPvfhawCspvG9/5M0pA8G2rn6tae1MGXwuidZoVYFe0KS1IAAASIXOEgCAsCJ8BRAq0yJVQS/Bd0HPdT0+NFqXnP73wh8cAODaL3a8qpVvi2rtr7tLLoB9R9MFammoDV34mq5KdP2mPRlb+T3dPaAVa9u1etlC3bj4oryukdaCAAAgEzpLAADChvAVQKjUTKgMegm+CTp0pcUwAJSe3oPHtfbX3Vr5tqi29RxK2QKuWLU01Oq1wZMaV1mh46eHgl5OQZg0VaJtXf1Zg04pNm/2jo3bVD9lXN4Dz1xaC1INAwAAAAAoJ4SvAEKlurL0T3uErgCAXAxZae2vu7VuZbMuiIzVuid367Edr6rn4PGgl5ZWc7RGW/cccBQ4ehGvyLSyWY9hTOz/w3wzku5KUyW6prXT8fMwZKV7WjsLUm3qtrWg17bJhLUAAAAAgGJW+ikEALgwoap0T3uErgAAv8QDufW3XqmvXH+ZJGlNa4fufrQz4JWNVGGka+dPy1vwmlyR6aR6M11g6JfqylH69kfenDZ4dHvs9u4BdfQNFiygdNJa0Evb5GKZcQsAAAAAQCalm0IAgAf1k8cFvQRPdo5ZLmMKH7wSugJA+UoM5Nq6+rXm58UXvBrFKlLv39LjW/BqJF13SZ3eMm9qyopJp9WbN3+nXfkogDVS2uBVigWQXrR19RdNdaiXtsk9B44VzYxbAAAAAAAyIXwFECpbXz4Q9BJcCaraldAVAMIhHsi5aWNbUEbad/iEb1WmmeaOJstWvTll/BgNHD3ly7ri4u2PM63vyIkznvbtdbt8cNs2+SsPPq+OvsGimnFbTmjjDAAAAAD+InwFECpPdHqrFim0IENXa6W5pwhdASAMftXxmnbtP5rXFrq5sFa6+9GOnPfznkvr9JfXXexLoJStXa5XToNhryMUchm94Gc456Vt8o59g44fW8gZt6WONs4AAAAAkB+ErwBCpffA8aCXkBFzXQEAhfTYi68FvYSs/Mg4/QpenbbLdcNtMOw1DPOyXT7COa9tk90o9IzbUuRl5i4AAAAAwBnCVwChMmSLsaei9Hzln2hcxRlCVwAAfDZz8jjfQrh8tGd2Gww31UW0JFrjqnq0OVrj+jnIVzhXqPbHxTTjttja+nqZuUsFLAAAAAA4R/gKIFQmjavUoePFM/NMknaOWS5jmOsKAEA+XDN/mi/78dIuN5to7XhPIdyqpY1asbbdURBcYaTblja62n8+w7lc2h+7UQwzbou1ra/bmbu0cQYAAAAAdyqCXgAAFNKV84rnhaOXKpere2zh2wxbK509K0VP/oDgFQBQ9lZcOduX/eSjXW53/1F965cvOXpsR9+g7mvr1r2tneroG9TSNzgLlW9YNNN1cOYlnHOqUCFeoULedNZv2qMVa9vTBvbxyuEfbXq5oOvychFBvI0zAAAAAMAZKl8BhMrS+dO0vsAvciULaq6rtbG3uacIXAGc74pZk7V5z8Ggl4ESY+TPPNZ88tJuNy65Veyu/qM+ry7mzod2aEr1mLStezNVTzqxYUuPrr+83nHomUs45+S59tI22YsgKzWLua2v14sIiqmNMwAAAAAUO8JXAKHSe/B4YMcOKnSVYqHrmSGpkUpXACn0Hz0V9BJQYoyR7lq2UP/W1q0d+4qzIs5Lu10p97DTi9s3pA7gss1ddcJt29hChHNu2iZ7CflzCd39UMxtfb22Yy6GNs4AAAAAUCpoOwwgVH4XQGVXvL3wqFHBVLvGWwwTvAJIZ/f+Y0EvASVk6vgxumHRTD3R9VpRB6+rly10HWhlaxWbL1YjW/c6rZ50wk3b2EKEcy0Ntbpz2QJVZPm9qMJIc2rHu16Ll9DdL8Xe1tdrO+ag2zgDAAAAQCnhLygAobKt92DBjrW1cqUmVxwPrNJ1aEjMdAUA+G7/0VO6f3NP0MtIa8akKn36nY1p2/im42fY6UV794AeeW6feg8e15ETZ3T/5h5f1+K0MrVQ4dxNi2dp5pRq3dPaqfYUYWVztEbLFtXr9g3bXa/lgshY19v4pdjb+nqtsA2yjTMAAAAAlBrCVwChcurMUEGOs3PMchkTTKUroSsAIMz2Hjqh2zds14YtvVq1tNFxaOSmVWy+fGzd5rzt22llaiHDuZaGWrU01I6Yr9vSUKumuojua+v2tJYg55MWe1tfLzN3g27jDAAAAAClhvAVQKhUjxmV1/0HNdfV2tjb3FOErgAASNLT3QNasbZdq5ctzFoF66VVbKlxWpkaRDjXVBdJuX2xB5mplEJbXzczd73OTgYAAACAMGPmK4BQWTR7Sl72G/Rc16EhglcAAJINWemOjdvU1tWvjr5B3dfWrXtbO3VfW/d5Mza9tootJW4qU1ctbcw6jzUun+FcKQSZyUqhra+bmbteZicDAAAAQNhR+QogVFa+ba7Wb/JvTl1Qla4SLYYBAHBiyEof//5mDaaohlwSrdGqpY167pXDAayscNxWpsbDuWwzcPMdzpVCkJmsVNr6Opm5e5uLtt0AAAAAgNcRvgIIlaa6iKpGV+hEjrNfX6i8WVUVQ4SuAACUgFTBqxRrTXzzd9oV8KjXvPJamVoM4VypBJnJSqWtb7aZuwAAAAAAbwhfAYTOuy6Zpp9u2+dp20cqP6vGir2ErgAAlIlyD15zqUwthnCuVILMRMVSOexUupm7AAAAAABvCF8BhM7731jvKXzdOWa5jAlupiuhKwCgnEweN1oHj6euSEXu/KxMDTKcK7UgM64YKocBAAAAAMEgfAUQOr0Hj7t6fFBzXQldAQDljODVHxVG+vQ7mxSpGl22bWP9CjILXcFbDJXDAAAAAIDCI3wFEDpH0sx9SxZk6GqtNPcUoSsAAEgvTNWTuQSZbV39WtPamXJ27JJojVYVYHYtYSsAAAAAhAfhK4DQmVCV+dQXVOgqUe0KAAiHmVPGqeeAu04UkGbXVOuDV8wMdfWk2yBz/aY9GVsWP909oBVr27V62ULduPgin1YJAAAAAAgzwlcAoVM/eVzK279TeZeWVjxD6AoAQB5VGOmaiy/Quqf2BL2UklJhpL9btiAUVa5+aevqzzorVpKGrHTHxm2qnzKO5xcAAAAAkDPCVwChs/XlAyNu2zlmuYxhrisAAPlUYaTVyxbqTbMmE766EH/eCAbdWdPamTV4jRuy0j2tnTzHAAAAAICcEb4CCJ0dewfP/TvIua6ErgCAsGmsi6h+yjg11UW0JFqTcgZnMYtUjdagw9nxbs2uqdbugWMjbg/TXFc/dfQNuv7+au8eUEffYCjbOQMAAAAA/EP4CiC0gqh2tTb2NvcUoSsAIHxe3Dd4br7mqqWNWrG23XFlYtAqjPTJaxp018M78rLmD14xU+++bLrauvp15MSZUM919UNbV7/n7XjOAQAAAAC5qAh6AQBQaPNnTAwseP3B2WsIXgEARe/9b5yhfP0XGZ+vKUl3LlugigBmrScyw2+ZxNv+3nrVvLyteULVaDXVRXRLS1SfWtqoW1qihIA5OOKxQtnrdgAAAAAAxFH5CiB0/tdTzQUNXq2VXhy6UO85/Y+FOSAAADl6e8MFapg2QXc/2pmX/cfna66/9UrNnFKte1o71Z7HFsSzp1Zr9/70LX0lpV1DctvfmxbPysuaaSvsrwlV3v7U9bodAAAAAABx/GUJIHRGSdlLXHxgrXRmSGpkrisAoMTcsXGbPtw8K6/HiM/XbGmoVUtDrTr6BtXW1a+1T3Sr5+BxX4/1wUXZW/omriFb29/kNSc+fuueA7p9w3ZX62uO1lDl6jOvYTYhOAAAAAAgV4SvAMLlS5PyfghrpaEhaR6hKwCgRA1Z6bEXX8v7cRLna8bff/mnz/t+nHhL3/gxj5w4c24maGLo2VQXcRWCpnp8U11EB4+d1p0P7XC0jwqjc9W38E9TXURLojV62kV1MiE4AAAAAMAPhK8A4BNCVwBAOek54G/1aSq/6njtvKrReCDqt+oxo3Tjt55MGcQtidZoVUJbYT/cetU8Ta6u1B0bt8va9I+Lz5Kl2jI/Vi1t1Iq17RrK8DWIIwQPjtOKcwAAAAAoFcZmejUAnhljNi9atGjR5s2bg14KgARDX5qkCp/3aW3sbe4pQlcAAHJRP3mcen1uOTy3drx27T+aMYCLh6A3Lr4op2Mlh0jVY0Zp45ZeR7NkkR/rN+3R5zZuL8jXH+60dfVrTWtnwS6KAAAAAAA3rrjiCm3ZsmWLtfYKt9tS+QoAObBWOjg0TpefXhv0UgAA8F1dZKz6Bk8W7Hh+B6+StLP/aNbHDNnYnNv6KeM8hT3ZQqS7bligY6fOUtkXgJsWz9LMKdW6p7WTELyIZAvFn+4e0Iq17YTiAAAAAEoS4SuAUDkrqUKx0NQY7/uxVjo7JDXQYhgAUMYKGbwGbchK97R2ug7hnIRIv901oNXLFuqWlqgPK4VbLQ21ammopb1tkWjr6s9ajSzlflEEzsf3PwAAAFA4hK8AQuX77/qd/vSRN3kOXq2VTg0ZXXz6P/xdGAAACFx794A6+gYdBxKESKWlqS5C2FQE1rR2OprDK3m/KAKvo70zAAAAUHiErwBCpaWhVkMPSxUV7ipfrZWGhqR5VLoCAFDW2rr6MwZ0idVj92/uIUQCXOjoG0wZAmbi9qIIvI72zgAAAEAwCF8BhEpTXURXVP6nNp35kKRYAJupBTGhKwCgHFQYaeXbolr7627HYWEqC+onatmimTpy4oyee+WQHn6uz79FFokjJ86kvD1T9ZhThEili5at/mjr6ve8Hc+3O1TmAwAAAMEhfAUQOtMnj9PcV36glyqXp62AJXQFAJSL5miNbhtuK9kwbYKjF+PT2d57WF+9MRY63dvamffwdVxlhY6fHsrrMZJNqBr5J1K26jE3CJFKCy1b/ZXu4oZ8bRdmtHcGAAAAgkP4CiB0xo+JnfriwWrnqOUaNer1+8+elRrPEroCAErPH7xxhi6fNSVtdd5Ni2dp5pRq3dPaqXaPFZzx8DBVSOm3WTXVerHviKdt490t3EoOH9q6+nXHhu3yIXeVRIhUSmjZ6j+v541CnG/KCe2dAQAAgGDxFwyA0Bk/dtR5Hzee/YF0NqDFAADgo58+s1dva7hAt7RE0z6mpaFWLQ21+uJ/PavvPbnb9THi4WEhKqQOewwqK4y0etlCfeW/n9egi31EqkafFzy0dfXr49/f7FvwKhEilQpatuaH1+eI59Yd2jsDAAAAwaoIegEAUGjzZ/CCAgAgf9KMES8Iq1gQ5OSF9zm14z0dIx4eNtVFtCRa42kfTu09dML1NtGp1Vq3sllvmjXZVfAqSYMnzqijb1BSrOrx5rXtrveRDSFSafDSshXZeTlvNEdrCARdor0zAAAAECzCVwChM2eqtxebAQBwwkoaFWAC6zQI8qMCbdXSRlXk6XOdWl3pabs/eesctTTU5lT5Fa969NK2OBNCpNKQS8tWZOfmvFFhpNuWNuZ3QWWI9s4AAABAsAhfAYTOhi29QS8BAFDmzvoc2rnV3j2gux7ekTEM8qMCraWhVncuW5CXAHbm1GpP28Urt3Kp/HJT9egUIVLpyCW4R3ZOzxvx9uFUi7tHe2cAAAAgWISvAELFSyUDAACl6JuPv6Tr7v6VbvzWk2lDIT8q0G5aPEsr35Z+xqxXNdVjPG0Xr9zyWsF17PQZ339XIEQqLbRszb+bFs/SupXNak5zAUhztEbrVjbrxsUXFXhl5YH2zgAAAECw6CkDIFSoSAAAhM3T3QNasbZdq5ctHBFkxCvQPrdxe8ZKz0zhYVtXv9b+utvvZWv+jIl67MXXXG8XX6P3oNPfMt7maI1uW9pI8FpCaNlaGC0NtWppqFVH36Dauvp15MQZTagarZaGWkJAH6xa2qgVa9sdVfFTmQ8AAAD4i78OAYQKFQkAgDAastIdG7epfsq4ESHgTYtnaeaUat3T2qn2FBWf2cLDfLTolaQPXF6vzbsPuKpCTa7cml1Trd0Dx1xt/8qB467WmWx2TbU+eMVMQqQSRsvWwmqqi/Bzkgd+XFwDAAAAwBvCVwChcuw04SsAIJyGrHRPa2fKF9i9VqDlq51/PET1UrnV1tWvNa2drtcV3/4rDz7vcdWxffzdsgWEGCUu3rI1l+AfKAa5XlwDAAAAwBvCVwAh428rQQAA/PCG6RH1HDyuwTx3aGjvHlBH32DakMhtBVo+2vkntr90W7nVc+BY1sdm2r6loVbjKis8r53qsfJBy1aUC9o7AwAAAIVH+AogVKorRwW9BAAARnhbY62+/YT/c1NTaevqz+kF98QX8LfsOeDjylK3v8xWuTVzyjhdc/EFsrKegtfkyq+6iVWSDrle+5tnTxkxUxeli5atKDe0dwYAAAAKh/AVQKgcO3026CUAADDCT373SsGO5XX+udd2vk5lan+ZWLm17snd+sWOV9V7MDabtefAca17ao/r482uqda3P/LmEWHEskUz9fBzfa7397F3zHW9DYobLVsBAAAAAF4QvgIIGZflMAAAFEDf4MmCHWtClfs/AdZv2uOpqtQNJ7veuueA/qN9ty/r2D1wLOXt1106XZGq0a5aQEeqRuu6S6fnvigUHVq2AgAAAADcInwFECrVlZz2AADhllil5yRQauvqz3vwKklPdw9oxdp2rV62MGX73nysI10L5k9e06A7H9rheD+fvKbBv0WhKNGyFQAAAADgFCkEgFDxUu0DAEC5aI7WqKkukrGF8JJojVYltFJd09qZ9+A1bshKd2zcpvop40a0cs3HOtK1YL71qnnq6BvUhi29WffxoStm6tar5vm7MAAAAAAAULJIIQCECjO5AAD5ZlScTe6NkW5b2pi1hXBiBeqbZk3O24zXdIasdE9r54gK3XysI9NFWV+98U1qqovo6491pWxBHKkarU9e01CQ4JWWtwAAAAAAlA7CVwCh0lQX0fzpEe3YNxj0UgAAZag5WqNli+p1+4btQS9lJCs9/uKrWvvr7qwVpPEK1A83zyrM2pK0dw+oo2/wXMDY1tWfl+Nkuyjr1qvm6dar5umR5/Zp45YeDZ44o0jVaC1bNLMgM17dVCgDAAAAAIDiQPgKIHSKsRoJAFC6jKT3v3GGPnlt47mwcMOW3oJXjGZjJX37iW7Hjx+y0mMvvpa/BWWROI81XXvgXMRbMDtx3aXTCxK2JnJToZxqRi4AAAAAAAhGRdALAIBC6ugb1ItUvQIAfGQlPbhtr14bPHnutlVLG2VMcGvyS8+B44EdOzFw9Xtme8VwC+Zi1dbVnzF4jYtXKOerMhi56+gb1H1t3bq3tVP3tXWro4/fQwEAAACg3FH5CiBUeHESAJAPyXNKWxpqNWfqeHX3Hw14ZaUrMXD1s7VuhZFWL1tY1O1617R2Zg1e41LNyEXwaBkNAAAAAOFF5SuAUMlH20IAAKTX55RKsWq3QgSvU8ePyfsxJldXunr8/OkRNUdrcj5uYjDVVBfREh/22Ryt0bqVzUXdprejb9B1y+rE7z0Eb/2mPVqxtj3t1zHeMvpHm14u8MoAAAAAAIVA5SuAUDl2mvAVAJA/657cra9cf1neOy1MrBqtv37vG/Tq4ZP66qMdeT3WwWOnHT/WGOkL77tELQ216ugb1ANbe/TNx3e6PuaC+onnzWNt6+rX4ePO11FhpDuXLdCxU2d15MQZTagarZaGWsczXoPk9XsncUYuguO2ZXT9lHFUwAIAAABAmSF8BRAyZTCADwBQtB7b8aqk/Hda+Px7L9GNiy/SfW3deT2OW3Omjj8XJDXVRTQtUuVpP2+6aMq5f6/ftMdRmBUXbytczNWtmXj93qG7R3GgZTQAAAAAgLbDAEKlunJU0EsAAJSxnoPH9chz+86bV5oPd2zcprau/qILbbr7j57X/tZrIDgtMlaS8yrCuPnTI0XfVjgbr987+f6eQ3a0jAYAAAAASISvAEKGFyYBAPn2sXWbtXFLb16PEa+Y82sWaipee0Ukts3NNUh0U0UoSZPGVRZdIO2W1/WX+uddDnJpGQ0AAAAAKB+ErwBChRcmAQCFsL33UN6PEa+YW7W0URUOk1Ij6YZF9Vkfb4z0rkvqPK0rsdo1lyAxrFWEXgL15mgN816LAC2jAQAAAAAS4SuAkGmqiyg6dXzQywAAwBfx1sN3LlvgKIC1kjZs6VVjXURvmJ4+rLNWenGftxAzsdo1lyAxzFWEbgL1CiPdtrQxvwuCI7SMBgAAAABIhK8AQujoKaoLAADlIV4xd9PiWVq3slnNDoPOF/cNakeWcHX3wDFPa0qudvUaJIa5itBpoF5hpNXLFtLZo0jQMro0dfQN6r62bt3b2qn72rpLvnoeAAAAQPC4xBZAqHT0DerVwZNBLwMAEDL1k8ep9+Bx3/ebWDHX0lCrloZard+0R3ds2K5so1JdjFJ1LFX723iQ+LmN2zPOb00OEsNeRXjT4lmaOaVa97R2qj1F++XmaI1uW9pIcFdE4pXebtpl0zI6OG1d/VrT2pny67UkWqNV/HwBAAAA8Kg8XpkAAIfKoRUhAKD0LKifqEsvjGjTrgM6cOz0iPsXzpykbT3u58SmCgY2bOnNS7CaTab2t16CRKoIXw/UO/oG1dbVryMnzmhC1Wi1NNQS2BWpVUsbtWJte8YLDeJoGR2c9Zv2ZLwg5OnuAa1Y267VyxbqxsUXFXZxAAAAAEoe4SuAUCmHVoQAAO+MvFV8et0u7uHn+s77eFZNtZZEa3TphRPPBWk3fuvJnCvmOvoGXe3DL07a37oNEqkifF1TXaQsP69y5LXSG4XT1tWf9esjSUNWumPjNtVPGcfXCQAAAIArzHwFECrl0ooQAOBec7RGq29Y4Gnb1TcscDxP1Yk9A8e0cUuPxo8ZfS5U8zobNVE+OzzMrqlOeXtztEbrVjY7rg5rqovolpaoPrW0Ube0RDOGin48J0ChZZvB7PZnBv5a09rpqDJZigWw97R25ndBAAAAAMoOKQSAUOGqdQAIn/dcWqe/vO7icyHfhi29rqspb1o8SzctnqVHntunj63b7Mu6kquq/KiYy2eHhw9eMVPvvmx6QdvfUkWIUkXL6OLkpTtAe/eAOvoG+boBAAAAcIzwFQAAAGXt0gsnnfeieS4zGa+7dLrrVriZxKuq4qGhl9moifLZ4WFC1ei8tr9NF1Ll+pwAQaJldHHx2h2graufryMAAAAAxwhfAYRKPtsxAgCKU3IgmWs1pZvw1onkqqpcKubyGUDma99tXf1a09qZMtBeEq3RquFglSpCALny2h0gn10FAAAAAJQfwlcAobKr/2jQSwAAFFiq0DCXakqn4a0bqaqqvFTMNdVFfK3MjWuO1uQl5Fy/aU/G5/Hp7gGtWNuu1csW6sbFF1FFCCAnXrsD5LOrAAAAAIDyw18QAEKl7/CJoJcAACigTKFhLtWU2cJbt/ysqvK7Mje59bJf2rr6HQXYybNxAcArr+cQzj0AAAAA3CB8BRAqdZOqgl4CACAHb5ge0Yt9g57mtaYLWb1WU6YKb5975ZAefq7P9b78rKpyWplrht9neirTtV72w5rWTscBcfJsXADwwkt3gHxV/gMAAAAoX4SvAEJlztTxQS8BAOBRhZE+/75L1HPgmKt5rU5ninqVGN529A16Cl/9DhWdtlWW5Kn1cq46+gZdt0ZOno0LAF646Q6Qr8p/AAAAAOWN8BVAqNRPHhf0EgCgZDRHa7RsUb1u37A96KWMqMB0Oq/V7UzRXBVTVZXTtspeWy/noq2r3/N2hK8AcuG0O0A+K/8BAAAAlDfCVwCh0nvweNBLAICiNXtqtT64aOaI8G3Dll7XVYp+SlWB6SRYDGqmaLFVVTlpq+y19bJXXmfc+jkbF0B4Oe0OQPAKAAAAwAvCVwChwou2AJDe7v3H9O7Lpo8I4dyEiU5EqkZrMMX5OF5pe+zUWccVmJlCw6BmilJVlZ3XGbd+zsYFEG5OuwMAAAAAgFu8egEgVI6dPhv0EgCEkJG0bFG9Htja61uAmS+p2ro6DROdqDDSv9x8hS6IjM3ri91BzxSlqiozr593WJ8vAPlT6Mp/AAAAAOWP8BVAqHS/djToJQAocm+dN1W/eWm/b/uLVzfeuPgiLVs0M20YVyzSdQjIFiY6kVzpWe4zRamqSq+YZuMCAAAAAAD4ifAVQKg8v/dQ0EsAUOTedUmdzgxZV6FQ1egKnTgzNOL25OrGxDDunx55UQ8/1+fbupNFqkZp8IT7av9MbV1ThYnHTp+RZFRdOUqvDp7U714+oO29h0dsW+hKz2KaKUpVVWrFNhsXAAAAAADAD4SvAEKFma8AsolXJd78nXY57bB78uyQ7rphgeNZpU11EV164aS8hq9eglfJWVvXbGFiMVR6MlO0+DEbFwAAAAAAlCNeXQIQKhOqRmvg2OmglwHAoTdMj2jHvkHHIahfWhpqNad2vLr7nbUqt1bauKVX62+90vExvIZ8b5lbo+deOazBPFxM4ldb12Ko9GSmaGlgNi4AAAAAACg3hK8AQuXKubXaM/By0MsA4ECFkT7/vkvUc+CY7tiwvWABbHxWqNPgNa69e0AdfYOOQ0evYdL/+cPL1FQX0ZrWDt39aKenfaRSbm1dmSlaOpiNCwAAAAAAygnhK4BQWfn2qNb/lvAVKHYmqc1opso4vx05ceZcAOtWW1e/47Ao13CwrWu/pzWmUq5tXZkpWlqKoWIaAAAAAAAgVxVBLwAAACDR1PFj9P2Vzbpx8UXnbmtpqNX6W6/UI595h774/kv0wUUzZfJ0/AlVoz3Ph3a73aqljapw+IkkhoMdfYOuQttMmqM1Wpf0fJeL+EzRbM9xuYbPAAAAAAAAKDwqXwGEitdqNgCF86dvnZM2BEusjFscnaLPbdzuqKrRjZaGWs/nCrdzXOPhYLbPIzkc9Lq+P796rqZFqkLV1pWZogAAAAAAACgkwlcAoeK1mg1A4TgNMJ2EaoeOn9aOfYOOj53rzE8vAZ6XcNDruay6crRuaYl62raUMVMUAAAAAAAAhUL4CiBU3FalAWF21w0LCjprNc5NgJktVGvr6vc08zPXeaxuuQ0HvZ7Lwn4OZKYoAAAAAAAA8i3cr8ABCB3aSgLZxVvcxmeApgsFJZ277dXBE3qm55C29RzK6dheA8x0oZrXtr5SbB6rl+A2F07DQa/nMs6BAAAAAAAAQH4RvgIIFS/VbECYpJt/mS4UTL4tOaTtPXBc3/l1t6Nj+xVgJvM68zOX4DbfCl2ZCwAAAAAAAMAZwlcAobNqaaNuXtsu66CaDSgVtRPGqP/IqRG3R6pGazDDfNBxlRW6/KIpeteldb7Mv0wV0jbWTdAdG7Yr049cvgNMrzM/vQa3hRBEZS4AAAAAAACAzIwlfcgLY8zmRYsWLdq8eXPQSwGQwvpNe7KGQUAxGVVhdDZFyhapGq1PXtOgW6+alzZYjN++a/9R9R06obqJVZpTO96XsNWptq7+ogww3XAb3BbC+k17HFfmxttIAwAAAAAAAMjsiiuu0JYtW7ZYa69wuy3ha54QvgLFr62rX3/z4PN6Yd9g0EspOxVGjqrx/PAHb5yhU2eGNHjijCJVo/WWuVNlJT35Ur827TqgA8dOp1+npKEUt8+tHa/u/qOewvkLJ1WpYdoE7d5/TLsHjo24vzlao2vnT9OD217R9t7DWfeXGEw+8tw+bdzSc+5zXbZopq67dLqHVQanGAPMUlcOwTYAAAAAAABQTAhfixDhK1A6OvoG9cDWXu3Ye1gDx06p98Ax9R9JH9hJ0sKZkzR7arX6Dp3QgWOnVGEqtP/oiZTbja4wOpMliawcZTR9YpXGjq7Q/qOndPLMWVkrVVQYWWtVXTlKtZEq/dHii/Tq4Al97ze7dfTUWUef3+gKozm11bp0xiTtOXBULw8c07FTQ6ocZTS3drz+8E316hs8oc27Dqj34HEdOHZax1Lsu2p0haZPqtK0yFhVjq5Qw7QJmjN1vIykp3buHxEIJj6vkjR/RkQfuHymXhs8qdUPvZAxeFw4c5Jqxldqy+4DGjxxNmUI6jRUiod9v9tzUC/2DWpcZYXqJladt85UYWCmQGvm5HG6Zv40vb2xVr0Hj6cNErMFjcn3108el3F/QCYE2wAAAAAAAIA/CF+LEOErUNoSQ4xjp89IMqquHJU10MjW9jW2v7OSrKorR+cckDzy3D5978ldeuXgcY2uMJo3LaJo7XhHa3X6+ecrxDnXCrf/qPoOn1DdpCrNmZq6FW6QoRKBFgAAAAAAAACEC+FrESJ8BQAAAAAAAAAAAEpPLuFrRT4WBAAAAAAAAAAAAABhU5LhqzFmpjHm34wxrxhjThpjdhljvmaMmRLEfgAAAAAAAAAAAABgdNALcMsYM0/SbyRNk/RfknZIWiJplaT3GGNarLX7C7UflChrY28VJXn9AQAAAAAAAAAAAIpQKSZP31AsML3NWnu9tfYOa+21ku6WdLGkvy3wflBqNm2SrrlG+u53g14JAAAAAAAAAAAAykhJha/GmLmSrpO0S9I/J939RUlHJa0wxowvxH5QYnbulP74j6UlS6Rf/lL6wheko0eDXhUAAAAAAAAAAADKREmFr5KuHX7/iLV2KPEOa+2gpDZJ1ZLeUqD9oBTs3y995jPS/PnSD3/4+u2vvio98URw6wIAAAAAAAAAAEBZKbWZrxcPv+9Ic3+nYhWtTZJaC7AfGWM2p7lrfqbt4Mzjjz+uX/7yl5Kkq666SldfffV59//sZz/TU089JUl617vepbe+9a2v33n8uH761a9qy9mz0uTJet/Chbpi8/CX64YbpL/7O+2fOlVTC/GJAAAAAAAAAAAAoOyVWuXrpOH3h9LcH799coH2g2I0NCStWyddfLH09NPn3/fWt0ptbdL990tNTcGsDwAAAAAAAAAAAGWp1CpfszHD722h9mOtvSLlDmIVsYtyXAe8+OUvpT/5k9i/3/Sm12//+MellSslY1JuBgAAAAAAAAAAAOSi1MLXeEXqpDT3T0x6XL73gzy7+uqrR7QaTvTud79b7373u5M3kq69VvrFL/T+9na9//d+T/roR6XKyhHbT51K02EAAAAAAAAAAAD4o9TC1xeH36frF9s4/D7dLFe/94Ogvfyy1NMjXXnl67cZI/3DP0g//rH0//6/UiQS2PIAAAAAAAAAAAAQHqUWvj42/P46Y0yFtXYofocxJiKpRdJxSU8VaD8IyqFD0p13SmvWSDNmSC+8II0d+/r9ixbF3gAAAAAAAAAAAIACqQh6AW5Ya1+S9IikOZL+IunuL0saL+nfrbVHJckYU2mMmW+MmZfLflBETp2KBa7z5kl33SWdOCF1d0vf/GbQKwMAAAAAAAAAAEDIlVrlqyR9QtJvJN1jjFkq6QVJzZKuUaxN8F8nPLZ++P7digWtXveDoFkr/ed/Sp/7nLRz5/n3XXEFVa4AAAAAAAAAAAAIXElVvkrnqlbfLOm7ioWln5U0T9I9kq601u4v5H5QAL/6lfSWt0g33XR+8DpnjvR//6/09NPSO94R2PIAAAAAAAAAAAAAqTQrX2WtfVnSLQ4et0uSyXU/CMiePdKnPiX95Cfn3z5livSFL0if+MT5c14BAAAAAAAAAACAAJVk+IqQqKqSHnvs9Y/HjpVWrZLuuCMWwAIAAAAAAAAAAABFpOTaDiNEpk2Tbr9dMkZasUJ68UXprrsIXgEAAAAAAAAAAFCUqHxFcfvMZ6Tf/33p8suDXgkAAAAAAAAAAACQEZWvKG7V1QSvAAAAAAAAAAAAKAmErwAAAAAAAAAAAADgA8JXAAAAAAAAAAAAAPAB4SsAAAAAAAAAAAAA+IDwFQAAAAAAAAAAAAB8QPgKAAAAAAAAAAAAAD4gfAUAAAAAAAAAAAAAHxC+AgAAAAAAAAAAAIAPCF8BAAAAAAAAAAAAwAeErwAAAAAAAAAAAADgA8JXAAAAAAAAAAAAAPAB4SsAAAAAAAAAAAAA+IDwFQAAAAAAAAAAAAB8QPgKAAAAAAAAAAAAAD4gfAUAAAAAAAAAAAAAHxC+AgAAAAAAAAAAAIAPCF8BAAAAAAAAAAAAwAeErwAAAAAAAAAAAADgA8JXAAAAAAAAAAAAAPAB4SsAAAAAAAAAAAAA+IDwFQAAAAAAAAAAAAB8QPgKAAAAAAAAAAAAAD4gfAUAAAAAAAAAAAAAHxC+AgAAAAAAAAAAAIAPCF8BAAAAAAAAAAAAwAeErwAAAAAAAAAAAADgA8JXAAAAAAAAAAAAAPAB4SsAAAAAAAAAAAAA+IDwFQAAAAAAAAAAAAB8QPgKAAAAAAAAAAAAAD4gfAUAAAAAAAAAAAAAHxC+AgAAAAAAAAAAAIAPCF8BAAAAAAAAAAAAwAeErwAAAAAAAAAAAADgA8JXAAAAAAAAAAAAAPAB4SsAAAAAAAAAAAAA+IDwFQAAAAAAAAAAAAB8QPgKAAAAAAAAAAAAAD4gfAUAAAAAAAAAAAAAHxC+AgAAAAAAAAAAAIAPCF8BAAAAAAAAAAAAwAeErwAAAAAAAAAAAADgA2OtDXoNZckYs3/cuHE1b3jDG4JeCgAAAAAAAAAAAACHXnjhBR0/fnzAWjvV7baEr3lijOmWNFHSroCXUm7mD7/fEegqAAB+4/wOAOWJ8zsAlB/O7QBQnji/A+ebI+mwtTbqdkPCV5QUY8xmSbLWXhH0WgAA/uH8DgDlifM7AJQfzu0AUJ44vwP+YeYrAAAAAAAAAAAAAPiA8BUAAAAAAAAAAAAAfED4CgAAAAAAAAAAAAA+IHwFAAAAAAAAAAAAAB8QvgIAAAAAAAAAAACAD4y1Nug1AAAAAAAAAAAAAEDJo/IVAAAAAAAAAAAAAHxA+AoAAAAAAAAAAAAAPiB8BQAAAAAAAAAAAAAfEL4CAAAAAAAAAAAAgA8IXwEAAAAAAAAAAADAB4SvAAAAAAAAAAAAAOADwlcAAAAAAAAAAAAA8AHhK0qCMWamMebfjDGvGGNOGmN2GWO+ZoyZEvTaACAsjDFTjTEfNcY8YIzpMsYcN8YcMsb82hiz0hiT8vcKY8xbjTH/Y4wZMMYcM8ZsM8Z82hgzKsOxPmKMedoYc2T4GI8bY96X4fHjjDFfNsa8aIw5YYx51RjzI2PMG/z43AEgbIwxK4wxdvjto2kew/kdAEqAMebtxpgNxpi9w6+p7DXGPGKM+f0Uj+XcDgAlwBjz3uFzec/w6zM7jTH/aYy5Ms3jOb8DBWSstUGvAcjIGDNP0m8kTZP0X5J2SFoi6RpJL0pqsdbuD26FABAOxpiPS/qmpL2SHpO0R1KdpGWSJknaIOlDNuGXC2PMHw7ffkLSekkDkt4v6WJJ91trP5TiOP8o6bOSeiTdL2mMpD+SVCPpU9baryc9fqykVkktkn4r6ReSLpL0IUmnJF1rrW335UkAgBAwxlwkabukUZImSPoza+13kh7D+R0ASoAx5vOSviKpX9KDiv0uXyvpckmPWWv/KuGxnNsBoAQYY+6S9FeS9kv6sWLn+AZJfyBptKQ/sdZ+P+HxnN+BAiN8RdEzxvxM0nWSbrPW3ptw+z9J+oykb1lrPx7U+gAgLIwx10oaL+m/rbVDCbdPl/S0Yr9Uf9Bau2H49omSuhQLZlustb8dvr1KsV/Cr5T0x9baHybs662S2iS9JGmxtfbA8O1zJG0ePv58a+2uhG0+J+nvFPtj4Kb42ob/uPixpOclLUhcMwAgNWOMkfSopKikjZL+l5LCV87vAFAajDEfkvQjST+XtMxaO5h0f6W19vTwvzm3A0AJGH4NplfSa5IWWmtfTbjvGsXO2d3W2rnDt3F+BwJA22EUNWPMXMWC112S/jnp7i9KOipphTFmfIGXBgChY639hbX2p8m/KFtr90n6l+EPr06464OSLpD0w/gv98OPPyHp88Mf/nnSYeIX0/xt/Jf74W12Kfb/wFhJt8RvHw4J4tv8VeLarLX/JekJSZdIusrxJwoA4XabpGsVO9ceTfMYzu8AUORMbCTIXZKOSVqeHLxKUjx4Hca5HQBKw2zFcp32xOBVkqy1j0kaVOx8Hsf5HQgA4SuK3bXD7x9J8WL/oGJX4FRLekuhFwYAOE/8hZszCbfFz+EPp3j8rxR7Ieitw61pnGzzUNJjJGmepFmSOqy13Q63AQCkMDyLabWkNdbaX2V4KOd3ACh+b1Wsi8H/SDowPBvwdmPMqjTzADm3A0Bp6FSsje8SY0xt4h3GmHdIiijW8SCO8zsQAMJXFLuLh993pLm/c/h9UwHWAgBIwRgzWtKfDH+Y+It52nO4tfaMpG7FZpHEW+GMl1Qv6Yi1dm+KQ6U65/P/BAD4YPhcvk6xed7/O8vDOb8DQPFbPPy+T9IWxea9rpb0NUm/Mcb80hiTWBnFuR0ASoC1dkDS7ZLqJD1vjPlXY8ydxpgfSXpEsREityZswvkdCMDooBcAZDFp+P2hNPfHb5+c/6UAANJYLekySf9jrf1Zwu1uz+Fezvn8PwEA/vj/JF0u6W3W2uNZHsv5HQCK37Th9x9X7IX1d0pqV6xd5VclvVvSf+r1sSGc2wGgRFhrv2aM2SXp3yT9WcJdXZK+m9SOmPM7EAAqX1HqzPB7G+gqACCkjDG3SfqspB2SVrjdfPi923O4m8fz/wQAZGGMWaJYtetXrbVP+rHL4fec3wEgOKOG3xtJH7TWtlprj1hrn5P0AUk9kq5K04I4Fc7tAFAkjDF/Jel+Sd9VrOXveElXSNop6T+MMX/vZnfD7zm/Az4ifEWxi18VMynN/ROTHgcAKBBjzF9IWiPpeUnXDLe+SeT2HJ7t8amupOT/CQDIQUK74Q5JX3C4Ged3ACh+B4bf77TWPpN4x3CHg3jHmiXD7zm3A0AJMMZcLekuST+x1v6ltXantfaYtXaLYhfX9Er6rDFm7vAmnN+BABC+oti9OPw+XT/4xuH36frJAwDywBjzaUlfl/SsYsHrvhQPS3sOH36xPyrpjGJXZspae1SxPxImGGNmpNhfqnM+/08AQG4mKHYOfYOkE8YYG3+T9MXhx3x7+LavDX/M+R0Ail/8PHowzf3xcHZc0uM5twNAcXvf8PvHku+w1h6T9LRiuc/lwzdzfgcCQPiKYhf/T+Q6Y8x536/GmIikFknHJT1V6IUBQFgZY26XdLek3ykWvL6a5qG/GH7/nhT3vUNStaTfWGtPOtzm95IeI0kvSdojqckYE3W4DQDgdSclrU3ztnX4Mb8e/jjekpjzOwAUv18p9mJ6ozFmTIr7Lxt+v2v4Ped2ACgNY4ffX5Dm/vjtp4bfc34HAkD4iqJmrX1J0iOS5kj6i6S7v6xYP/t/H74iBwCQZ8aYL0haLWmzpKXW2v4MD79fUr+kPzLGvDlhH1WS/mb4w28mbfMvw+//2hgzJWGbOYr9P3BS0n3x2621NmGbv0+8UMcY84eS3q5YW+RfOvwUASBUrLXHrbUfTfUm6SfDD/ve8G3rhz/m/A4ARW749/T1irWA/P8S7zPGvEvSuxVr//jw8M2c2wGgNDwx/P5jxpj6xDuMMb+nWLHSCUm/Gb6Z8zsQABP7uQCKlzFmnmL/WUyT9F+SXpDULOkaxVoVvNVauz+4FQJAOBhjPiLpu5LOSrpXqWd17LLWfjdhm+sV+0X/hKQfShqQ9AeSLh6+/Uab9MuIMearkv5SUs/wY8ZIuknSVEmfstZ+PenxYxW7evKtkn4rqVXSLEkfUuxKz2utte2eP3EACCljzJcUaz38Z9ba7yTdd704vwNAUTPGTJPUJqlBsRfrn5Y0W7GZgFbScmvtfyY8/npxbgeAojYcbP5M0jslDUp6QNI+xcaIvE+SkfRpa+2ahG2uF+d3oKAIX1ESjDEXSfo/irU6mCppr6QfS/qytXYgwKUBQGgkvAifyS+ttVcnbdci6a8lXSmpSlKXpH+TdI+19myaY31E0iclXSJpSNIWSf9grX0wzePHSbpD0nLFfrk/LOlxSV+01j6f/bMDACTLFL4O38/5HQCKnDGmRtLnFQtc6xV7of7Xku601o4Y4cS5HQCKnzGmUrEK1D9S7NxbrVig+rRi5+tHUmzD+R0oIMJXAAAAAAAAAAAAAPABM18BAAAAAAAAAAAAwAeErwAAAAAAAAAAAADgA8JXAAAAAAAAAAAAAPAB4SsAAAAAAAAAAAAA+IDwFQAAAAAAAAAAAAB8QPgKAAAAAAAAAAAAAD4gfAUAAAAAAAAAAAAAHxC+AgAAAAAAAAAAAIAPCF8BAAAAAAAAAAAAwAeErwAAAAAAAAAAAADgA8JXAAAAAAAAAAAAAPAB4SsAAAAAAAAAAAAA+IDwFQAAAACQkTHGJr2dNcb0G2N+YYz5cNDrKwRjzC5jzK4CHWuxMeY/jDG7jTEnjTGHjTEvGWN+aoz5K2PM+DTbNRlj/tkYs8MYc8QYc3T43183xsxLs82cFF/f5LerHTwm+e1P8/okAQAAAECRMtbaoNcAAAAAAChixpj4H45fHn5fKeliSddLGiXpbmvtXwawtIKJB6/W2jl5Ps7Nkr4nyUj6haRnJZ2VFJX0ZkkXSWq01nYlbXebpH9S7CLrX0n6rSQr6QpJV0s6I+kvrLXfTtpujqRuSYckfS3Nsr4r6U9T3P5pSZMkrZF0MOm+H1trf5f+MwUAAACA8kT4CgAAAADIKB6+WmtN0u1LJT06/OFca+2uAi+tYAoRvhpjqiXtlTRe0rutta0pHvNWSc9baw8m3PYnigW2A5I+YK39VdI2b5f0Y0k1km6w1m5MuG+OYuHrbref2/BzMltStJy/9gAAAADgBm2HAQAAAACeDIeDOxSr0lwsScaYLyW0ql1ujGkfboG7K76dMWbGcHvcXcaYU8aY14wxG40xVyQfwxjzp/E2tsaY9xpjfjPcTveAMeZ+Y0xjqrUZY6qNMZ8zxvxu+PFHjDFPGmP+OMVj4211v2SMWWKM+W9jzEDCca1iIePspNa63zXGTDHGHBtuC2xGrkQyxjw4/PgRn1+SyyRNlPRsquB1+Dn/TVLwGtHrFavLk4PX4W2ekBRvD/0NY0xVlnUAAAAAADwifAUAAAAA5CIeOCa3VfqspH+TtEfS1yU9JEnGmKhiLXE/IeklSV+V9DNJ75X0G2PM+9IcZ5li1Zs9irW5fVLSDZKeMsZcfN6CjJks6deS/k6xlr3/plhl6AWSfmCM+Zs0x7hS0hOSqhK26VCs3fKh4bcvJ7z92Fp7QNIPJc2V9M4RT44xMyW9R9Jma+3mNMeN2z/8/sJ0c11T+KCkKZKettb+LN2DrLUPS9okqU6xdtEAAAAAgDwYHfQCAAAAAAClyRjzTsVmv1rFgr1E10q60lq7Nen2f5F0oaTPW2v/NmFf31BsVun3jDGzrbVHkrZ7v6T3W2sfTNhmlWJVn9+QtDThsV+TdLmk2621f5/w+CrFAtz/bYy5P8VM0uskfdxa+62k239jjPlTSbLWfkkjfUPSLZJu1ettmOM+qthc3OR9prJTsedxsaQ2Y8y3Jf1G0nPW2lNptnnb8PufO9j/o8P7fodigXGiycaYL6XY5nfW2h872DcAAAAAQISvAAAAAACHEsK5SsVC1+sVq3y921q7O+nh/5ocvA5XgV6nWDXs3yfeZ639jTHm/0q6WbEq139P2t8vEoPXYV+X9ClJ1w4HtruNMVOH9/HbxOB1+BgnjDG3S3q3pOWSfpe0v9+lCF6zstb+1hjzW0l/aIyZbq3dN/z5jpK0UtKgpP/rYD/WGPNBxSpurx7+/CTptDFmq6SNkr5prT2csNmM4fcvO1hq/DEzU9w3SdIXU9z+PcUCawAAAACAA4SvAAAAAACn4uGclXRQsRa9a62130/x2KdT3Hb58PsnrLWnU9z/C8WC08s1Mnz9ZfKDrbVnjTG/ljRveJvdilV2jpJk01RyVg6/f4PDNTv1DcVaFf8/irU7lqTfVyzo/GaKSt6UrLV7JF1jjHmDpHdJerOkJQlvnzDGXG2t7R7eJF3b51Tij00183W3tXaOkzUCAAAAANIjfAUAAAAAOGKtNdkfdc6+FLdNGn6/N8028dsnp7ivL8tx4vueOvx+8fBbOhMy7MuLHyo2v/bPjDGrrbVDirUhlpy1HD6PtfYFSS/EPzbGzFcs3L1S0t16fW5r/Dmb5WC38YrX19yuBwAAAADgTEXQCwAAAAAAlKVUlZiHht9PT7PNjKTHJapLs018X4eS3t9trTUZ3q5xuGZHrLXHJX1X0hxJ1w23WH6PpHZr7TNe95uw/x2SVgx/eG3CXb8efv9OB7uJP2ZzrusBAAAAAKRG+AoAAAAAKJT4DNi3GWNSdWKKB6JbUtx3VfINwzNV35a076clDUl6ew7rTOWsYu2MM/mmYgHurZI+Ovx411WvGQwOv0+sQP5PSQckLTHGvCvdhsP3LZF0engbAAAAAEAeEL4CAAAAAArCWtsj6VHFqkM/nXifMaZZ0nLFgsQHUmx+rTHmfUm3fVKxea+PWWt3Dx/jVUn/IenNxpgvpAp5jTHzjDFRl8vfL+kCY8y4dA+w1nZKapX0PkkfV2wu7nqnBzDGRI0xtxljJqW4z0j66+EPf5VwzEFJnxn+8AfGmJYU275V0g+GP/x7a+3LTtcEAAAAAHCHma8AAAAAgEL6uKQ2Sf9gjLlO0m8lXSTpQ4pVrN4yHCgm+6mkB4wxD0jqkvRGSb8vaUDSJ5Ie+0lJjZL+j6QVxphfKzYz9kJJb1BsFuwfS+p2se7W4e0eNsb8StJJSc9Ya3+a9LhvKNbet07SvdbaYy6OMUnSGsWemzZJzypW7TpNsVbDcyW9KumziRtZa79njJms2MzZJ4wxjyvWWthKukKximKj2FzaL7hYDwAAAADAJcJXAAAAAEDBWGt3GmPeLOnzioWnV0s6LOlhSX9rrd2UZtONkv5VserP9yrWPnejpM9ZazuSjnHYGHOVpI8pVk17g6QqxQLYTsUqRR91ufS/kTRZ0vsltSjWUvh7ioXCiX4iqV9Srdy3HH5B0gckXSfpLZJuklQj6ZhigfPfSfqatfa15A2ttWuMMT+TdJukpcPbx6t0j0j6M2vtD12uBwAAAADgkrHWBr0GAAAAAABSMsb8qaT7FKuI/W6wq8nOGDNXsaC0zVrr99xZt2sZLel/JL1L0qestV8Pcj0AAAAAEAbMfAUAAAAAwD//S7EWv4EHndbaM4q1c35O0j3GmJUBLwkAAAAAyh5thwEAAAAAyIExZpZi7Y0bJd0i6RlJ/xnoooZZaw8ZY96r2LouNMZMstYeCnpdAAAAAFCuCF8BAAAAAMjNXEl3Kjab9VFJf26tHQp2Sa+z1u6W9KWg1wEAAAAAYcDMVwAAAAAAAAAAAADwATNfAQAAAAAAAAAAAMAHhK8AAAAAAAAAAAAA4APCVwAAAAAAAAAAAADwAeErAAAAAAAAAAAAAPiA8BUAAAAAAAAAAAAAfED4CgAAAAAAAAAAAAA+IHwFAAAAAAAAAAAAAB8QvgIAAAAAAAAAAACADwhfAQAAAAAAAAAAAMAHhK8AAAAAAAAAAAAA4APCVwAAAAAAAAAAAADwAeErAAAAAAAAAAAAAPiA8BUAAAAAAAAAAAAAfPD/A3wKcZ1LyhVDAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 1152x648 with 1 Axes>"
      ]
     },
     "metadata": {
      "image/png": {
       "height": 544,
       "width": 943
      },
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "x_pred = np.array([-25, 100]).reshape(-1, 1)\n",
    "y_pred = lm.predict(x_pred)\n",
    "\n",
    "plt.figure(figsize=(16, 9))\n",
    "plt.scatter(train.calculatedfinishedsquarefeet, train.taxvaluedollarcnt, label='actual')\n",
    "plt.scatter(train.calculatedfinishedsquarefeet, lm.predict(train[['calculatedfinishedsquarefeet']]), label='simple_lm')\n",
    "plt.plot(x_pred, y_pred, ls='--', c='red', label='linear model prediction line')\n",
    "plt.hlines(0, -25, 100, ls=':', color='grey')\n",
    "plt.vlines(0, -10, 100, ls=':', color='grey')\n",
    "\n",
    "plt.xlabel('Property SQFT')\n",
    "plt.ylabel('Tax Assessment Value')\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a9608bce",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### Model Interpretation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "ddad9ad7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "For every 1 foot increase in Property SQFT, we predict a 248.72 dollar increase in Tax Assessment Value\n"
     ]
    }
   ],
   "source": [
    "# taxvaluedollarcnt (y) = 248.723 * calculatedfinishedsquarefeet (X) + -17472.733\n",
    "\n",
    "print(f'For every 1 foot increase in Property SQFT, we predict a {lm.coef_[0]:.2f} dollar increase in Tax Assessment Value')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "c8429583",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>SQFT</th>\n",
       "      <th>predicted_taxvaluedollarcnt</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1000</td>\n",
       "      <td>231250.641264</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1001</td>\n",
       "      <td>231499.364638</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1002</td>\n",
       "      <td>231748.088012</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1003</td>\n",
       "      <td>231996.811386</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1004</td>\n",
       "      <td>232245.534760</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1495</th>\n",
       "      <td>2495</td>\n",
       "      <td>603092.085538</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1496</th>\n",
       "      <td>2496</td>\n",
       "      <td>603340.808912</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1497</th>\n",
       "      <td>2497</td>\n",
       "      <td>603589.532286</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1498</th>\n",
       "      <td>2498</td>\n",
       "      <td>603838.255660</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1499</th>\n",
       "      <td>2499</td>\n",
       "      <td>604086.979034</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1500 rows &times; 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      SQFT  predicted_taxvaluedollarcnt\n",
       "0     1000                231250.641264\n",
       "1     1001                231499.364638\n",
       "2     1002                231748.088012\n",
       "3     1003                231996.811386\n",
       "4     1004                232245.534760\n",
       "...    ...                          ...\n",
       "1495  2495                603092.085538\n",
       "1496  2496                603340.808912\n",
       "1497  2497                603589.532286\n",
       "1498  2498                603838.255660\n",
       "1499  2499                604086.979034\n",
       "\n",
       "[1500 rows x 2 columns]"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "SQFT = np.arange(1000, 2500)\n",
    "taxvaluedollarcnt = lm.predict(SQFT.reshape(-1, 1))\n",
    "pd.DataFrame(dict(SQFT=SQFT, predicted_taxvaluedollarcnt=taxvaluedollarcnt))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c324e22a",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "<div class=\"alert alert-warning\">\n",
    "\n",
    "\n",
    "## Multiple Regression + RFE\n",
    "\n",
    "$$ y = m_1x_1 + m_2x_2 + b $$\n",
    "\n",
    "$$ y = f(x1, x2, ...) $$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "e9b56cc7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "selected top 2 features: Index(['county_Orange County', 'county_Ventura County'], dtype='object')\n"
     ]
    }
   ],
   "source": [
    "X_train = train.drop(columns='taxvaluedollarcnt')\n",
    "y_train = train.taxvaluedollarcnt\n",
    "X_validate = validate.drop(columns='taxvaluedollarcnt')\n",
    "\n",
    "lm = LinearRegression()\n",
    "k = 2\n",
    "\n",
    "# 1. Transform our X\n",
    "rfe = RFE(lm, n_features_to_select=2)\n",
    "rfe.fit(X_train, y_train)\n",
    "print('selected top 2 features:', X_train.columns[rfe.support_])\n",
    "X_train_rfe = rfe.transform(X_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "290d44cc",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(28561, 17)"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "c9ad594f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(28561, 2)"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train_rfe.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "554f94d8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>actual</th>\n",
       "      <th>logerror</th>\n",
       "      <th>simple_lm</th>\n",
       "      <th>multiple_rfe</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1663334</td>\n",
       "      <td>0.04901</td>\n",
       "      <td>983390.124531</td>\n",
       "      <td>410135.180989</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>45302</td>\n",
       "      <td>-0.003143</td>\n",
       "      <td>294426.378284</td>\n",
       "      <td>410135.180989</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>388528</td>\n",
       "      <td>0.002243</td>\n",
       "      <td>303877.866500</td>\n",
       "      <td>410135.180989</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>130715</td>\n",
       "      <td>-0.02305</td>\n",
       "      <td>325268.076672</td>\n",
       "      <td>410135.180989</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>407342</td>\n",
       "      <td>-0.023093</td>\n",
       "      <td>517282.521475</td>\n",
       "      <td>410135.180989</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    actual  logerror      simple_lm   multiple_rfe\n",
       "0  1663334   0.04901  983390.124531  410135.180989\n",
       "1    45302 -0.003143  294426.378284  410135.180989\n",
       "2   388528  0.002243  303877.866500  410135.180989\n",
       "3   130715  -0.02305  325268.076672  410135.180989\n",
       "4   407342 -0.023093  517282.521475  410135.180989"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 2. Use the transformed x in our model\n",
    "lm.fit(X_train_rfe, y_train)\n",
    "\n",
    "# 3. Make predictions\n",
    "X_validate_rfe = rfe.transform(X_validate)\n",
    "predictions['multiple_rfe'] = lm.predict(X_validate_rfe)\n",
    "\n",
    "predictions.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6ef3d35e",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### Examine the Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "fd59071c",
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['county_Orange County', 'county_Ventura County'], dtype='object')"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.columns[rfe.support_]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "239907d1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([118759.06708793,  80470.5058521 ]), 410135.18098900304)"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lm.coef_, lm.intercept_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "3be50427",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Tax Assessed Value = 118759.07 x Orange County + 80470.51 x Ventura County + 410135.18\n"
     ]
    }
   ],
   "source": [
    "print(f'Tax Assessed Value = {lm.coef_[0]:.2f} x Orange County + {lm.coef_[1]:.2f} x Ventura County + {lm.intercept_:.2f}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "1e38ebb9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Orange County</th>\n",
       "      <th>Ventura County</th>\n",
       "      <th>predicted_taxvaluedollarcnt</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>490605.686841</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>528894.248077</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>410135.180989</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Orange County  Ventura County  predicted_taxvaluedollarcnt\n",
       "0              0               1                490605.686841\n",
       "1              1               0                528894.248077\n",
       "2              0               0                410135.180989"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X = pd.DataFrame({\n",
    "    'Orange County': [0, 1, 0],\n",
    "    'Ventura County': [1, 0, 0]\n",
    "})\n",
    "\n",
    "X['predicted_taxvaluedollarcnt'] = lm.predict(X)\n",
    "X"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fff2d044",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "<div class=\"alert alert-warning\">\n",
    "\n",
    "\n",
    "\n",
    "## Polynomial Features\n",
    "\n",
    "First off...what the heck is a polynomial?\n",
    "\n",
    "$$ y = a_0x^0 + a_1x^1 + a_2x^2 + ... + a_nx^n $$\n",
    "\n",
    "or a simpler form:\n",
    "\n",
    "$$ y = a_0 + a_1x + a_2x^2 + ... + a_nx^n $$\n",
    "\n",
    "`degree = order = n`\n",
    "\n",
    "What does a polynomial look like when graphed on a coordinate plane? Well it depends on\n",
    "- The intercept\n",
    "- The coefficients\n",
    "\n",
    "Consider this third degree polynomial:\n",
    "\n",
    "$$ y = 2x + x^3 $$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "id": "800c7803",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import PolynomialFeatures\n",
    "\n",
    "X_train = train[['bedroomcnt', 'bathroomcnt', 'calculatedfinishedsquarefeet']]\n",
    "X_validate = validate[['bedroomcnt', 'bathroomcnt', 'calculatedfinishedsquarefeet']]\n",
    "y_train = train.taxvaluedollarcnt\n",
    "y_validate = validate.taxvaluedollarcnt"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e3ff7911",
   "metadata": {},
   "source": [
    "<div class=\"alert alert-warning\">\n",
    "\n",
    "\n",
    "## Polynomial Degree 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "a86efc91",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>bedroomcnt</th>\n",
       "      <td>2.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>5.00</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>bathroomcnt</th>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4.50</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>calculatedfinishedsquarefeet</th>\n",
       "      <td>926.0</td>\n",
       "      <td>2425.0</td>\n",
       "      <td>4483.00</td>\n",
       "      <td>2362.0</td>\n",
       "      <td>1522.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>bedroomcnt^2</th>\n",
       "      <td>4.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>25.00</td>\n",
       "      <td>16.0</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>bedroomcnt bathroomcnt</th>\n",
       "      <td>2.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>22.50</td>\n",
       "      <td>12.0</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>bedroomcnt calculatedfinishedsquarefeet</th>\n",
       "      <td>1852.0</td>\n",
       "      <td>7275.0</td>\n",
       "      <td>22415.00</td>\n",
       "      <td>9448.0</td>\n",
       "      <td>3044.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>bathroomcnt^2</th>\n",
       "      <td>1.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>20.25</td>\n",
       "      <td>9.0</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>bathroomcnt calculatedfinishedsquarefeet</th>\n",
       "      <td>926.0</td>\n",
       "      <td>7275.0</td>\n",
       "      <td>20173.50</td>\n",
       "      <td>7086.0</td>\n",
       "      <td>3044.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>calculatedfinishedsquarefeet^2</th>\n",
       "      <td>857476.0</td>\n",
       "      <td>5880625.0</td>\n",
       "      <td>20097289.00</td>\n",
       "      <td>5579044.0</td>\n",
       "      <td>2316484.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                 0          1            2  \\\n",
       "bedroomcnt                                     2.0        3.0         5.00   \n",
       "bathroomcnt                                    1.0        3.0         4.50   \n",
       "calculatedfinishedsquarefeet                 926.0     2425.0      4483.00   \n",
       "bedroomcnt^2                                   4.0        9.0        25.00   \n",
       "bedroomcnt bathroomcnt                         2.0        9.0        22.50   \n",
       "bedroomcnt calculatedfinishedsquarefeet     1852.0     7275.0     22415.00   \n",
       "bathroomcnt^2                                  1.0        9.0        20.25   \n",
       "bathroomcnt calculatedfinishedsquarefeet     926.0     7275.0     20173.50   \n",
       "calculatedfinishedsquarefeet^2            857476.0  5880625.0  20097289.00   \n",
       "\n",
       "                                                  3          4  \n",
       "bedroomcnt                                      4.0        2.0  \n",
       "bathroomcnt                                     3.0        2.0  \n",
       "calculatedfinishedsquarefeet                 2362.0     1522.0  \n",
       "bedroomcnt^2                                   16.0        4.0  \n",
       "bedroomcnt bathroomcnt                         12.0        4.0  \n",
       "bedroomcnt calculatedfinishedsquarefeet      9448.0     3044.0  \n",
       "bathroomcnt^2                                   9.0        4.0  \n",
       "bathroomcnt calculatedfinishedsquarefeet     7086.0     3044.0  \n",
       "calculatedfinishedsquarefeet^2            5579044.0  2316484.0  "
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 1. Generate Polynomial Features\n",
    "poly = PolynomialFeatures(degree=2, include_bias=False, interaction_only=False)\n",
    "poly.fit(X_train)\n",
    "X_train_poly = pd.DataFrame(\n",
    "    poly.transform(X_train),\n",
    "    columns=poly.get_feature_names(X_train.columns),\n",
    "    index=train.index,\n",
    ")\n",
    "X_train_poly.head().T"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "7aa04bfe",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "outputs": [],
   "source": [
    "# 2. Use the features\n",
    "lm = LinearRegression()\n",
    "lm.fit(X_train_poly, y_train)\n",
    "\n",
    "X_validate_poly = poly.transform(X_validate)\n",
    "predictions['polynomial degree 2'] = lm.predict(X_validate_poly)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "c6971115",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>actual</th>\n",
       "      <th>logerror</th>\n",
       "      <th>simple_lm</th>\n",
       "      <th>multiple_rfe</th>\n",
       "      <th>polynomial degree 2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1663334</td>\n",
       "      <td>0.04901</td>\n",
       "      <td>983390.124531</td>\n",
       "      <td>410135.180989</td>\n",
       "      <td>931010.714799</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>45302</td>\n",
       "      <td>-0.003143</td>\n",
       "      <td>294426.378284</td>\n",
       "      <td>410135.180989</td>\n",
       "      <td>293207.372986</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>388528</td>\n",
       "      <td>0.002243</td>\n",
       "      <td>303877.866500</td>\n",
       "      <td>410135.180989</td>\n",
       "      <td>242519.921994</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>130715</td>\n",
       "      <td>-0.02305</td>\n",
       "      <td>325268.076672</td>\n",
       "      <td>410135.180989</td>\n",
       "      <td>326422.853321</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>407342</td>\n",
       "      <td>-0.023093</td>\n",
       "      <td>517282.521475</td>\n",
       "      <td>410135.180989</td>\n",
       "      <td>499389.755074</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    actual  logerror      simple_lm   multiple_rfe  polynomial degree 2\n",
       "0  1663334   0.04901  983390.124531  410135.180989        931010.714799\n",
       "1    45302 -0.003143  294426.378284  410135.180989        293207.372986\n",
       "2   388528  0.002243  303877.866500  410135.180989        242519.921994\n",
       "3   130715  -0.02305  325268.076672  410135.180989        326422.853321\n",
       "4   407342 -0.023093  517282.521475  410135.180989        499389.755074"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "predictions.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7fcb0960",
   "metadata": {},
   "source": [
    "#### Examine the coefficients of the resulting model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "df65062b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "bedroomcnt                                 -36158.519991\n",
       "bathroomcnt                                -21317.566226\n",
       "bedroomcnt bathroomcnt                     -16535.164574\n",
       "bedroomcnt calculatedfinishedsquarefeet        -1.840363\n",
       "calculatedfinishedsquarefeet^2                 -0.017293\n",
       "bathroomcnt calculatedfinishedsquarefeet       17.993813\n",
       "calculatedfinishedsquarefeet                  282.915370\n",
       "bedroomcnt^2                                 1526.242288\n",
       "bathroomcnt^2                               17881.794780\n",
       "dtype: float64"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "feature_names = poly.get_feature_names(X_train.columns)\n",
    "pd.Series(lm.coef_, index=feature_names).sort_values()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9b6f151c",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "<div class=\"alert alert-warning\">\n",
    "\n",
    "\n",
    "## Polynomial Only Interaction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "id": "9539624e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "bedroomcnt                                 -41139.133625\n",
       "bathroomcnt                                -18835.097095\n",
       "bedroomcnt calculatedfinishedsquarefeet       -24.877100\n",
       "bathroomcnt calculatedfinishedsquarefeet       17.649795\n",
       "calculatedfinishedsquarefeet                  282.665351\n",
       "bedroomcnt bathroomcnt                       9796.352567\n",
       "dtype: float64"
      ]
     },
     "execution_count": 73,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "poly = PolynomialFeatures(degree=2, include_bias=False, interaction_only=True)\n",
    "poly.fit(X_train)\n",
    "X_train_poly = pd.DataFrame(\n",
    "    poly.transform(X_train),\n",
    "    columns=poly.get_feature_names(X_train.columns),\n",
    "    index=train.index,\n",
    ")\n",
    "lm = LinearRegression()\n",
    "lm.fit(X_train_poly, y_train)\n",
    "\n",
    "X_validate_poly = poly.transform(X_validate)\n",
    "predictions['polynomial only interaction'] = lm.predict(X_validate_poly)\n",
    "\n",
    "pd.Series(lm.coef_, index=poly.get_feature_names(X_train.columns)).sort_values()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "e4530034",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>actual</th>\n",
       "      <th>logerror</th>\n",
       "      <th>simple_lm</th>\n",
       "      <th>multiple_rfe</th>\n",
       "      <th>polynomial degree 2</th>\n",
       "      <th>polynomial only interaction</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1663334</td>\n",
       "      <td>0.04901</td>\n",
       "      <td>983390.124531</td>\n",
       "      <td>410135.180989</td>\n",
       "      <td>931010.714799</td>\n",
       "      <td>935951.199147</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>45302</td>\n",
       "      <td>-0.003143</td>\n",
       "      <td>294426.378284</td>\n",
       "      <td>410135.180989</td>\n",
       "      <td>293207.372986</td>\n",
       "      <td>302881.887929</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>388528</td>\n",
       "      <td>0.002243</td>\n",
       "      <td>303877.866500</td>\n",
       "      <td>410135.180989</td>\n",
       "      <td>242519.921994</td>\n",
       "      <td>258440.924719</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>130715</td>\n",
       "      <td>-0.02305</td>\n",
       "      <td>325268.076672</td>\n",
       "      <td>410135.180989</td>\n",
       "      <td>326422.853321</td>\n",
       "      <td>333055.259481</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>407342</td>\n",
       "      <td>-0.023093</td>\n",
       "      <td>517282.521475</td>\n",
       "      <td>410135.180989</td>\n",
       "      <td>499389.755074</td>\n",
       "      <td>504174.009760</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    actual  logerror      simple_lm   multiple_rfe  polynomial degree 2  \\\n",
       "0  1663334   0.04901  983390.124531  410135.180989        931010.714799   \n",
       "1    45302 -0.003143  294426.378284  410135.180989        293207.372986   \n",
       "2   388528  0.002243  303877.866500  410135.180989        242519.921994   \n",
       "3   130715  -0.02305  325268.076672  410135.180989        326422.853321   \n",
       "4   407342 -0.023093  517282.521475  410135.180989        499389.755074   \n",
       "\n",
       "   polynomial only interaction  \n",
       "0                935951.199147  \n",
       "1                302881.887929  \n",
       "2                258440.924719  \n",
       "3                333055.259481  \n",
       "4                504174.009760  "
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "predictions.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f375ed14",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "Takeaway: interaction terms can be useful if you suspect there's a multiplicative relationship between independent variables. Higher order polynomial features should be used with caution (usually when you have domain knowledge that says this is the right thing to do).\n",
    "\n",
    "Example: Telco churn, using monthly charges and tenure to predict total charges. monthly charges times tenure will give us a good number for total charges."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "65fe8f14",
   "metadata": {},
   "source": [
    "<div class=\"alert alert-warning\">\n",
    "\n",
    "\n",
    "\n",
    "## Lasso-Lars\n",
    "\n",
    "Performs intrinsic feature selection and noise reduction to avoid overfitting.\n",
    "\n",
    "The most important hyperparameter is `alpha`. \n",
    "- `alpha` determines the degree of regularization applied to the model. \n",
    "- The lowest `alpha` is zero, and this represents no regularization (identical to OLS). \n",
    "- There is no limit to `alpha`, but as it rises, it will reach a point where performance no longer changes (often because you've regularized all of your coefficients to zero. \n",
    "\n",
    "Generally speaking, the higher the alpha, the lower your risk for overfitting. \n",
    "\n",
    "`alpha` = 1.0 is a good place to start."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "d99463ca",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.linear_model import LassoLars"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "784fdb3c",
   "metadata": {},
   "source": [
    "`alpha = 0`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "6c99d444",
   "metadata": {},
   "outputs": [],
   "source": [
    "# create the model object\n",
    "lars = LassoLars(alpha=0)\n",
    "\n",
    "# fit the model to our training data\n",
    "lars.fit(X_train, y_train)\n",
    "\n",
    "# predict validate\n",
    "X_validate_pred_lars = lars.predict(X_validate)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "7ed1307f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "bedroomcnt                     -70982.107595\n",
       "calculatedfinishedsquarefeet      247.261933\n",
       "bathroomcnt                     52514.524033\n",
       "dtype: float64"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.Series(lars.coef_, index=X_train.columns).sort_values()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c42f339e",
   "metadata": {},
   "source": [
    "`alpha = 0.5`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "5c865e44",
   "metadata": {},
   "outputs": [],
   "source": [
    "# create the model object\n",
    "lars = LassoLars(alpha=0.5)\n",
    "\n",
    "# fit the model to our training data= \n",
    "lars.fit(X_train, y_train)\n",
    "\n",
    "# predict validate\n",
    "X_validate_pred_lars = lars.predict(X_validate)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "5be20684",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "bedroomcnt                     -70710.336619\n",
       "calculatedfinishedsquarefeet      247.107118\n",
       "bathroomcnt                     52377.032985\n",
       "dtype: float64"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.Series(lars.coef_, index=X_train.columns).sort_values()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cb876151",
   "metadata": {},
   "source": [
    "`alpha = 1.0`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "e40c62cd",
   "metadata": {},
   "outputs": [],
   "source": [
    "# create the model object\n",
    "lars = LassoLars(alpha=1)\n",
    "\n",
    "# fit the model to our training data\n",
    "lars.fit(X_train, y_train)\n",
    "\n",
    "# predict validate\n",
    "X_validate_pred_lars = lars.predict(X_validate)\n",
    "\n",
    "# Add lassolars predictions to our predictions DataFrame\n",
    "predictions['lasso_lars'] = X_validate_pred_lars"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "614c7027",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "bedroomcnt                     -70438.565644\n",
       "calculatedfinishedsquarefeet      246.952303\n",
       "bathroomcnt                     52239.541937\n",
       "dtype: float64"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.Series(lars.coef_, index=X_train.columns).sort_values()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "2277015b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "101505.35310822102"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lars.intercept_"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1ff8ebae",
   "metadata": {},
   "source": [
    "`alpha = 2`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "1f7fff12",
   "metadata": {},
   "outputs": [],
   "source": [
    "# create the model object\n",
    "lars_high_alpha = LassoLars(alpha=2)\n",
    "\n",
    "# fit the model to our training data\n",
    "lars_high_alpha.fit(X_train, y_train)\n",
    "\n",
    "# predict validate\n",
    "X_validate_pred_lars_high_alpha = lars.predict(X_validate)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "2e54f667",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "bedroomcnt                     -69895.023692\n",
       "calculatedfinishedsquarefeet      246.642674\n",
       "bathroomcnt                     51964.559842\n",
       "dtype: float64"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.Series(lars_high_alpha.coef_, index=X_train.columns).sort_values()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "5a579583",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "100903.18357780174"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lars_high_alpha.intercept_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "8d9d3d1d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>actual</th>\n",
       "      <th>logerror</th>\n",
       "      <th>simple_lm</th>\n",
       "      <th>multiple_rfe</th>\n",
       "      <th>polynomial degree 2</th>\n",
       "      <th>polynomial only interaction</th>\n",
       "      <th>lasso_lars</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1663334</td>\n",
       "      <td>0.04901</td>\n",
       "      <td>983390.124531</td>\n",
       "      <td>410135.180989</td>\n",
       "      <td>931010.714799</td>\n",
       "      <td>935951.199147</td>\n",
       "      <td>952006.761725</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>45302</td>\n",
       "      <td>-0.003143</td>\n",
       "      <td>294426.378284</td>\n",
       "      <td>410135.180989</td>\n",
       "      <td>293207.372986</td>\n",
       "      <td>302881.887929</td>\n",
       "      <td>304346.928579</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>388528</td>\n",
       "      <td>0.002243</td>\n",
       "      <td>303877.866500</td>\n",
       "      <td>410135.180989</td>\n",
       "      <td>242519.921994</td>\n",
       "      <td>258440.924719</td>\n",
       "      <td>243292.550466</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>130715</td>\n",
       "      <td>-0.02305</td>\n",
       "      <td>325268.076672</td>\n",
       "      <td>410135.180989</td>\n",
       "      <td>326422.853321</td>\n",
       "      <td>333055.259481</td>\n",
       "      <td>334969.014207</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>407342</td>\n",
       "      <td>-0.023093</td>\n",
       "      <td>517282.521475</td>\n",
       "      <td>410135.180989</td>\n",
       "      <td>499389.755074</td>\n",
       "      <td>504174.009760</td>\n",
       "      <td>507417.168764</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    actual  logerror      simple_lm   multiple_rfe  polynomial degree 2  \\\n",
       "0  1663334   0.04901  983390.124531  410135.180989        931010.714799   \n",
       "1    45302 -0.003143  294426.378284  410135.180989        293207.372986   \n",
       "2   388528  0.002243  303877.866500  410135.180989        242519.921994   \n",
       "3   130715  -0.02305  325268.076672  410135.180989        326422.853321   \n",
       "4   407342 -0.023093  517282.521475  410135.180989        499389.755074   \n",
       "\n",
       "   polynomial only interaction     lasso_lars  \n",
       "0                935951.199147  952006.761725  \n",
       "1                302881.887929  304346.928579  \n",
       "2                258440.924719  243292.550466  \n",
       "3                333055.259481  334969.014207  \n",
       "4                504174.009760  507417.168764  "
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "predictions.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bae16a9c",
   "metadata": {},
   "source": [
    "<div class=\"alert alert-warning\">\n",
    "\n",
    "\n",
    "\n",
    "## Generalized Linear Model\n",
    "\n",
    "The most flexible regression algorithm, the GLM allows for different distributions, beyond just the Normal Distribution for OLS (and other models based on OLS, like LASSO). We will use the `TweedieRegressor` method to create a GLM using sklearn.\n",
    "\n",
    "> `sklearn.linear_model.TweedieRegressor(power=0, alpha=1.0)`\n",
    "\n",
    "* power = 0: Normal Distribution\n",
    "* power = 1: Poisson Distribution\n",
    "* power = (1,2): Compound Distribution\n",
    "* power = 2: Gamma Distribution\n",
    "* power = 3: Inverse Gaussian Distribution"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "a12ba914",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.linear_model import TweedieRegressor\n",
    "\n",
    "# create the model object\n",
    "glm = TweedieRegressor(power=1, alpha=0)\n",
    "\n",
    "# fit the model to our training data\n",
    "glm.fit(X_train, y_train)\n",
    "\n",
    "# predict validate\n",
    "X_validate_predict_glm = glm.predict(X_validate)\n",
    "\n",
    "# Add general linear model predictions to our predictions DataFrame\n",
    "predictions['glm'] = X_validate_predict_glm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "e15ac011",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "bedroomcnt                      0.0\n",
       "bathroomcnt                     0.0\n",
       "calculatedfinishedsquarefeet    0.0\n",
       "dtype: float64"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.Series(glm.coef_, index=X_train.columns).sort_values()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "d76ae789",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "13.016522417153546"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "glm.intercept_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "e8a77996",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>actual</th>\n",
       "      <th>logerror</th>\n",
       "      <th>simple_lm</th>\n",
       "      <th>multiple_rfe</th>\n",
       "      <th>polynomial degree 2</th>\n",
       "      <th>polynomial only interaction</th>\n",
       "      <th>lasso_lars</th>\n",
       "      <th>glm</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1663334</td>\n",
       "      <td>0.04901</td>\n",
       "      <td>983390.124531</td>\n",
       "      <td>410135.180989</td>\n",
       "      <td>931010.714799</td>\n",
       "      <td>935951.199147</td>\n",
       "      <td>952006.761725</td>\n",
       "      <td>449783.851861</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>45302</td>\n",
       "      <td>-0.003143</td>\n",
       "      <td>294426.378284</td>\n",
       "      <td>410135.180989</td>\n",
       "      <td>293207.372986</td>\n",
       "      <td>302881.887929</td>\n",
       "      <td>304346.928579</td>\n",
       "      <td>449783.851861</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>388528</td>\n",
       "      <td>0.002243</td>\n",
       "      <td>303877.866500</td>\n",
       "      <td>410135.180989</td>\n",
       "      <td>242519.921994</td>\n",
       "      <td>258440.924719</td>\n",
       "      <td>243292.550466</td>\n",
       "      <td>449783.851861</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>130715</td>\n",
       "      <td>-0.02305</td>\n",
       "      <td>325268.076672</td>\n",
       "      <td>410135.180989</td>\n",
       "      <td>326422.853321</td>\n",
       "      <td>333055.259481</td>\n",
       "      <td>334969.014207</td>\n",
       "      <td>449783.851861</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>407342</td>\n",
       "      <td>-0.023093</td>\n",
       "      <td>517282.521475</td>\n",
       "      <td>410135.180989</td>\n",
       "      <td>499389.755074</td>\n",
       "      <td>504174.009760</td>\n",
       "      <td>507417.168764</td>\n",
       "      <td>449783.851861</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    actual  logerror      simple_lm   multiple_rfe  polynomial degree 2  \\\n",
       "0  1663334   0.04901  983390.124531  410135.180989        931010.714799   \n",
       "1    45302 -0.003143  294426.378284  410135.180989        293207.372986   \n",
       "2   388528  0.002243  303877.866500  410135.180989        242519.921994   \n",
       "3   130715  -0.02305  325268.076672  410135.180989        326422.853321   \n",
       "4   407342 -0.023093  517282.521475  410135.180989        499389.755074   \n",
       "\n",
       "   polynomial only interaction     lasso_lars            glm  \n",
       "0                935951.199147  952006.761725  449783.851861  \n",
       "1                302881.887929  304346.928579  449783.851861  \n",
       "2                258440.924719  243292.550466  449783.851861  \n",
       "3                333055.259481  334969.014207  449783.851861  \n",
       "4                504174.009760  507417.168764  449783.851861  "
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "predictions.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bc83ac41",
   "metadata": {},
   "source": [
    "<div class=\"alert alert-danger\">\n",
    "\n",
    "\n",
    "\n",
    "### TL;DR\n",
    "\n",
    "First of all, it always helps to know the distribution of your target variable before modeling.\n",
    "\n",
    "> `plt.hist(y_train)` to see the shape of your target (`y`).\n",
    "\n",
    "Secondly, you could try them all...doesn't hurt usually.\n",
    "\n",
    "Third, if you have a good feel for your data:\n",
    "- For a normally distributed y and a linear relationship: OLS, LassoLars, GLM(power=0) will work best.\n",
    "- For polynomial relationships, polynomial regression is best.\n",
    "- For poisson, gamma or inverse gaussian distributions, use the Generalized Linear Model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "ab127cf0",
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['bedroomcnt', 'bathroomcnt', 'fullbathcnt', 'bed_bath_ratio',\n",
       "       'fireplacecnt', 'age', 'calculatedfinishedsquarefeet',\n",
       "       'lotsizesquarefeet', 'has_taxdelinquency', 'has_hottuborspa',\n",
       "       'has_basement', 'has_pool', 'fips', 'logerror', 'taxvaluedollarcnt',\n",
       "       'county_Los Angeles County', 'county_Orange County',\n",
       "       'county_Ventura County'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.columns"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "152a66b8",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "<div class=\"alert alert-success\">\n",
    "\n",
    "\n",
    "\n",
    "# Evaluation"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2dec8f5d",
   "metadata": {},
   "source": [
    "<div class=\"alert alert-warning\">\n",
    "\n",
    "\n",
    "## Baseline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "a764a5b6",
   "metadata": {},
   "outputs": [],
   "source": [
    "# add a baseline model\n",
    "predictions['baseline'] = train.taxvaluedollarcnt.mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "fc6b333d",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>actual</th>\n",
       "      <th>logerror</th>\n",
       "      <th>simple_lm</th>\n",
       "      <th>multiple_rfe</th>\n",
       "      <th>polynomial degree 2</th>\n",
       "      <th>polynomial only interaction</th>\n",
       "      <th>lasso_lars</th>\n",
       "      <th>glm</th>\n",
       "      <th>baseline</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1663334</td>\n",
       "      <td>0.04901</td>\n",
       "      <td>983390.124531</td>\n",
       "      <td>410135.180989</td>\n",
       "      <td>931010.714799</td>\n",
       "      <td>935951.199147</td>\n",
       "      <td>952006.761725</td>\n",
       "      <td>449783.851861</td>\n",
       "      <td>449783.851861</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>45302</td>\n",
       "      <td>-0.003143</td>\n",
       "      <td>294426.378284</td>\n",
       "      <td>410135.180989</td>\n",
       "      <td>293207.372986</td>\n",
       "      <td>302881.887929</td>\n",
       "      <td>304346.928579</td>\n",
       "      <td>449783.851861</td>\n",
       "      <td>449783.851861</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>388528</td>\n",
       "      <td>0.002243</td>\n",
       "      <td>303877.866500</td>\n",
       "      <td>410135.180989</td>\n",
       "      <td>242519.921994</td>\n",
       "      <td>258440.924719</td>\n",
       "      <td>243292.550466</td>\n",
       "      <td>449783.851861</td>\n",
       "      <td>449783.851861</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>130715</td>\n",
       "      <td>-0.02305</td>\n",
       "      <td>325268.076672</td>\n",
       "      <td>410135.180989</td>\n",
       "      <td>326422.853321</td>\n",
       "      <td>333055.259481</td>\n",
       "      <td>334969.014207</td>\n",
       "      <td>449783.851861</td>\n",
       "      <td>449783.851861</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>407342</td>\n",
       "      <td>-0.023093</td>\n",
       "      <td>517282.521475</td>\n",
       "      <td>410135.180989</td>\n",
       "      <td>499389.755074</td>\n",
       "      <td>504174.009760</td>\n",
       "      <td>507417.168764</td>\n",
       "      <td>449783.851861</td>\n",
       "      <td>449783.851861</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    actual  logerror      simple_lm   multiple_rfe  polynomial degree 2  \\\n",
       "0  1663334   0.04901  983390.124531  410135.180989        931010.714799   \n",
       "1    45302 -0.003143  294426.378284  410135.180989        293207.372986   \n",
       "2   388528  0.002243  303877.866500  410135.180989        242519.921994   \n",
       "3   130715  -0.02305  325268.076672  410135.180989        326422.853321   \n",
       "4   407342 -0.023093  517282.521475  410135.180989        499389.755074   \n",
       "\n",
       "   polynomial only interaction     lasso_lars            glm       baseline  \n",
       "0                935951.199147  952006.761725  449783.851861  449783.851861  \n",
       "1                302881.887929  304346.928579  449783.851861  449783.851861  \n",
       "2                258440.924719  243292.550466  449783.851861  449783.851861  \n",
       "3                333055.259481  334969.014207  449783.851861  449783.851861  \n",
       "4                504174.009760  507417.168764  449783.851861  449783.851861  "
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "predictions.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8091d9b2",
   "metadata": {},
   "source": [
    "<div class=\"alert alert-warning\">\n",
    "\n",
    "\n",
    "## Calculate mean squared error for every column in the dataframe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "id": "851c4cc5",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "actual                         0.000000e+00\n",
       "polynomial degree 2            8.190023e+10\n",
       "polynomial only interaction    8.207193e+10\n",
       "lasso_lars                     8.228066e+10\n",
       "simple_lm                      8.479923e+10\n",
       "multiple_rfe                   1.221896e+11\n",
       "glm                            1.251606e+11\n",
       "baseline                       1.251606e+11\n",
       "logerror                       3.302344e+11\n",
       "dtype: float64"
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def calculate_mse(y_predicted):\n",
    "    return mean_squared_error(predictions.actual, y_predicted)\n",
    "\n",
    "predictions.apply(calculate_mse).sort_values()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "70365336",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE using Mean\n",
      "Train/In-Sample:  354521.86 \n",
      "Validate/Out-of-Sample:  353780.47\n",
      "RMSE using Median\n",
      "Train/In-Sample:  363920.04 \n",
      "Validate/Out-of-Sample:  363893.41\n"
     ]
    }
   ],
   "source": [
    "# We need y_train and y_validate to be dataframes to append the new columns with predicted values. \n",
    "y_train = pd.DataFrame(y_train)\n",
    "y_validate = pd.DataFrame(y_validate)\n",
    "\n",
    "# 1. Predict target_pred_mean\n",
    "target_pred_mean = y_train['taxvaluedollarcnt'].mean()\n",
    "y_train['target_pred_mean'] = target_pred_mean\n",
    "y_validate['target_pred_mean'] = target_pred_mean\n",
    "\n",
    "# 2. compute target_pred_median\n",
    "target_pred_median = y_train['taxvaluedollarcnt'].median()\n",
    "y_train['target_pred_median'] = target_pred_median\n",
    "y_validate['target_pred_median'] = target_pred_median\n",
    "\n",
    "# 3. RMSE of target_pred_mean\n",
    "rmse_train = mean_squared_error(y_train.taxvaluedollarcnt, y_train.target_pred_mean)**(1/2)\n",
    "rmse_validate = mean_squared_error(y_validate.taxvaluedollarcnt, y_validate.target_pred_mean)**(1/2)\n",
    "\n",
    "print(\"RMSE using Mean\\nTrain/In-Sample: \", round(rmse_train, 2), \n",
    "      \"\\nValidate/Out-of-Sample: \", round(rmse_validate, 2))\n",
    "\n",
    "# 4. RMSE of G3_pred_median\n",
    "rmse_train = mean_squared_error(y_train.taxvaluedollarcnt, y_train.target_pred_median)**(1/2)\n",
    "rmse_validate = mean_squared_error(y_validate.taxvaluedollarcnt, y_validate.target_pred_median)**(1/2)\n",
    "\n",
    "print(\"RMSE using Median\\nTrain/In-Sample: \", round(rmse_train, 2), \n",
    "      \"\\nValidate/Out-of-Sample: \", round(rmse_validate, 2))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "033a56c3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "image/png": {
       "height": 261,
       "width": 401
      },
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot to visualize actual vs predicted. \n",
    "plt.hist(y_train.taxvaluedollarcnt, color='blue', alpha=.5, label=\"Actual Tax Assessment Value\")\n",
    "plt.hist(y_train.target_pred_mean, bins=1, color='red', alpha=.5, rwidth=100, label=\"Predicted Tax Assessment Value - Mean\")\n",
    "plt.hist(y_train.target_pred_median, bins=1, color='orange', alpha=.5, rwidth=100, label=\"Predicted Tax Assessment Value - Median\")\n",
    "plt.xlabel(\"Tax Assessment Value (target)\")\n",
    "plt.ylabel(\"Number of Homes\")\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0f2b2bb4",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "<div class=\"alert alert-danger\">\n",
    "\n",
    "\n",
    "\n",
    "## Evaluate on Test: Best Model (2nd Degree Polynomial)\n",
    "\n",
    "Notice this gets a bit more complicated as we have to reproduce any transformations to the data we did when training. Polynomial transformations need to be applied to the test dataframe to evaluate the polynomial model. "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a56b8387",
   "metadata": {},
   "source": [
    "# re-create our polynomial model\n",
    "X_train = train[['bedroomcnt', 'bathroomcnt', 'calculatedfinishedsquarefeet']]\n",
    "y_train = train.taxvaluedollarcnt\n",
    "\n",
    "poly = PolynomialFeatures(degree=2, include_bias=False, interaction_only=False)\n",
    "poly.fit(X_train)\n",
    "X_train_poly = pd.DataFrame(\n",
    "    poly.transform(X_train),\n",
    "    columns=poly.get_feature_names(X_train.columns),\n",
    "    index=train.index,\n",
    ")\n",
    "\n",
    "lm = LinearRegression()\n",
    "lm.fit(X_train_poly, y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e7408d59",
   "metadata": {},
   "source": [
    "# any transformations applied to your training data must be applied to the test as well\n",
    "\n",
    "X_test = test[['bedroomcnt', 'bathroomcnt', 'calculatedfinishedsquarefeet']]\n",
    "X_test_poly = poly.transform(X_test)\n",
    "\n",
    "test_predictions = lm.predict(X_test_poly)\n",
    "test_actual = test.taxvaluedollarcnt"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2e9ca92c",
   "metadata": {},
   "source": [
    "mean_squared_error(test_actual, test_predictions)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}