Skip to content

Instantly share code, notes, and snippets.

@MohammedNagdy
Last active February 3, 2019 23:08
Show Gist options
  • Save MohammedNagdy/3b5829588dd7fa6a377fb1cc0d592c5f to your computer and use it in GitHub Desktop.
Save MohammedNagdy/3b5829588dd7fa6a377fb1cc0d592c5f to your computer and use it in GitHub Desktop.
Display the source blob
Display the rendered blob
Raw
{
"nbformat_minor": 1,
"cells": [
{
"execution_count": 15,
"cell_type": "code",
"metadata": {},
"outputs": [],
"source": "import numpy as np\nfrom sklearn.model_selection import train_test_split, StratifiedKFold, cross_val_score, GridSearchCV\nfrom sklearn.linear_model import LogisticRegression\nimport sklearn\nimport sklearn.manifold\nimport pandas as pd\nimport matplotlib.pyplot as plt\nimport urllib\nfrom sklearn.metrics import accuracy_score, confusion_matrix, recall_score, roc_auc_score, precision_score\nimport sys\n%matplotlib inline"
},
{
"source": "<h4>So, because we know that the data is already clean and the columns have equal number of samples. Therefore, we won't do much on the cleanning side.\nBut we'll encode categorical varibales since that data has a lot of them.</h4>",
"cell_type": "markdown",
"metadata": {}
},
{
"execution_count": 16,
"cell_type": "code",
"metadata": {},
"outputs": [],
"source": "#reading the data downloading from the UCI library\nwith urllib.request.urlopen('https://archive.ics.uci.edu/ml/machine-learning-databases/statlog/german/german.data') as response:\n df = response.read().decode()\n\ndf = \" \".join(df.split())\ndf = df.split(\" \")\ndf = pd.DataFrame(df)\ndf=df.values.reshape(1000, 21)\ndf = pd.DataFrame(df)"
},
{
"execution_count": 17,
"cell_type": "code",
"metadata": {
"scrolled": true
},
"outputs": [
{
"output_type": "stream",
"name": "stdout",
"text": "Status of existing checking account object\nDuration in Months float64\nCredit history object\nPurpose object\nCredit amount float64\nSavings account/bonds object\nPresent employment since object\nInstallment rate in percentage of disposable income float64\nPersonal status and sex object\nOther debtors / guarantors object\nPresent residence since float64\nProperty object\nAge in years float64\nOther installment plans object\nHousing object\nNumber of existing credits at this bank float64\nJob object\nNumber of people being liable to provide maintenance for float64\nTelephone object\nforeign worker object\nLoan Acceptance float64\ndtype: object\n"
}
],
"source": "# label the columns\ndf.columns = [\n \"Status of existing checking account\",\n \"Duration in Months\",\n \"Credit history\",\n \"Purpose\",\n \"Credit amount\",\n \"Savings account/bonds\",\n \"Present employment since\",\n \"Installment rate in percentage of disposable income\",\n \"Personal status and sex\",\n \"Other debtors / guarantors\",\n \"Present residence since\",\n \"Property\",\n \"Age in years\",\n \"Other installment plans\",\n \"Housing\",\n \"Number of existing credits at this bank\",\n \"Job\",\n \"Number of people being liable to provide maintenance for\",\n \"Telephone\",\n \"foreign worker\",\n \"Loan Acceptance\"\n]\n\n# making numeric cols numeric\ndf[\"Duration in Months\"] = df[\"Duration in Months\"].astype(float)\ndf[\"Credit amount\"] = df[\"Credit amount\"].astype(float)\ndf[\"Installment rate in percentage of disposable income\"] = df[\"Installment rate in percentage of disposable income\"].astype(float)\ndf[\"Present residence since\"] = df[\"Present residence since\"].astype(float)\ndf[\"Age in years\"] = df[\"Age in years\"].astype(float)\ndf[\"Number of existing credits at this bank\"] = df[\"Number of existing credits at this bank\"].astype(float)\ndf[\"Number of people being liable to provide maintenance for\"] = df[\"Number of people being liable to provide maintenance for\"].astype(float)\ndf[\"Loan Acceptance\"] = df[\"Loan Acceptance\"].astype(float)\nprint(df.dtypes)"
},
{
"source": "<h4>Because our model can't read letters. We have to transform words into numbers. There are a lot of methods to do this. But in my opinion, the most accurate one and the one\nI prefer to use is binary encoding.</h4>",
"cell_type": "markdown",
"metadata": {}
},
{
"execution_count": 18,
"cell_type": "code",
"metadata": {},
"outputs": [],
"source": "# get the cols that contain string or categorical variables\ndf_obj = df.select_dtypes(include=\"object\").copy()\n\n# binary encoding\ndef bin_encode(data):\n newData = []\n newData = pd.DataFrame(newData) \n for col in data.columns:\n data[col] = np.array(data[col].values) # tranforming from data frame to numpy array to ge unique cat-values\n cat = np.unique(data[col]) # every unique categorical variable\n for childcat in cat:\n newData[childcat] = [1 if category == childcat else 0 for category in data[col]] # transform every unique variable to binary\n return newData\n \nencoded_cats = bin_encode(df_obj)"
},
{
"execution_count": 19,
"cell_type": "code",
"metadata": {},
"outputs": [
{
"output_type": "stream",
"name": "stdout",
"text": " A11 A12 A13 A14 A30 A31 A32 A33 A34 A40 ... A201 \\\n0 1 0 0 0 0 0 0 0 1 0 ... 1 \n1 0 1 0 0 0 0 1 0 0 0 ... 1 \n2 0 0 0 1 0 0 0 0 1 0 ... 1 \n3 1 0 0 0 0 0 1 0 0 0 ... 1 \n4 1 0 0 0 0 0 0 1 0 1 ... 1 \n\n A202 Duration in Months Credit amount \\\n0 0 6.0 1169.0 \n1 0 48.0 5951.0 \n2 0 12.0 2096.0 \n3 0 42.0 7882.0 \n4 0 24.0 4870.0 \n\n Installment rate in percentage of disposable income \\\n0 4.0 \n1 2.0 \n2 2.0 \n3 2.0 \n4 3.0 \n\n Present residence since Age in years \\\n0 4.0 67.0 \n1 2.0 22.0 \n2 3.0 49.0 \n3 4.0 45.0 \n4 4.0 53.0 \n\n Number of existing credits at this bank \\\n0 2.0 \n1 1.0 \n2 1.0 \n3 1.0 \n4 2.0 \n\n Number of people being liable to provide maintenance for Loan Acceptance \n0 1.0 1.0 \n1 1.0 2.0 \n2 2.0 1.0 \n3 2.0 1.0 \n4 2.0 2.0 \n\n[5 rows x 62 columns]\n"
}
],
"source": "# concatenate binary vars with numerical vars\ndf_num = df.select_dtypes(include=\"float64\").copy()\nn_df = pd.concat([encoded_cats, df_num], axis=1)\nprint(n_df.head())"
},
{
"source": "<h4>Here we split the data into a training set to train the model and minimize the error, and a testing set to see how the model will perform on unseen data.</h4>",
"cell_type": "markdown",
"metadata": {}
},
{
"execution_count": 20,
"cell_type": "code",
"metadata": {},
"outputs": [],
"source": "# split the data for the model \nX = np.array(n_df.drop([\"Loan Acceptance\"],axis=1)) # the features\ny = np.array(n_df[\"Loan Acceptance\"]) # the labels\n\nX_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)\n"
},
{
"source": "<h4>This is the where we visualize all of our dataset in a compressed form using t-SNE model. This helps us to see all the dimensions in 2D space.\nNotice the green dots are the accpeted clinets and the red ones are denied ones.</h4>",
"cell_type": "markdown",
"metadata": {}
},
{
"execution_count": 21,
"cell_type": "code",
"metadata": {},
"outputs": [],
"source": "# graph the data using t-SNE after reducing it to dimensions\n#first reduce dimensions bt t-SNE\nmodel = sklearn.manifold.TSNE(learning_rate=10, n_components=3)\ntransformed = model.fit_transform(X)\nxs = transformed[:,0]\nys = transformed[:,1]\nzs = transformed[:,2]\nxs=pd.DataFrame(xs)\nys=pd.DataFrame(ys)"
},
{
"execution_count": 22,
"cell_type": "code",
"metadata": {},
"outputs": [
{
"output_type": "stream",
"name": "stdout",
"text": "[[10.104637145996094 -0.1401769518852234 'accept']\n [-9.141085624694824 -6.318431377410889 'recject']\n [-0.6053515672683716 0.11326328665018082 'accept']\n [-1.8682620525360107 -10.056123733520508 'accept']\n [-12.729700088500977 -1.413717269897461 'recject']]\n(1000, 3)\n"
},
{
"output_type": "display_data",
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAABEQAAAJrCAYAAADpr0b6AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4wLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvpW3flQAAIABJREFUeJzs3XmcjXX/x/HXdbY558zYx74vCVEI2SIlJYSkZKtUlKhf2ve6b+13e0lCaKFSKClliSyVJYzsJIx9m/3s1++PyzYzx1Ixh5n38/HwMOd7fa/rfC5398yZz/X9fj6GaZqIiIiIiIiIiBQktlgHICIiIiIiIiKS15QQEREREREREZECRwkRERERERERESlwlBARERERERERkQJHCRERERERERERKXCUEBERERERERGRAkcJEREREREREREpcJQQEREREREREZECRwkRERERERERESlwHLEO4HRITEw0q1SpEuswRERERERERCTGlixZstc0zZInm5cvEiJVqlRh8eLFsQ5DRERERERERGLMMIy/TmWetsyIiIiIiIiISIGjhIiIiIiIiIiIFDhKiIiIiIiIiIhIgaOEiIiIiIiIiIgUOEqIiIiIiIiIiEiBo4SIiIiIiIiIiBQ4SoiIiIiIiIiISIGjhIiIiIiIiIiIFDhKiIiIiIiIiIhIgaOEiIiIiIiIiIgUOEqIiIiIiIiIiEiBo4SIiIiIiIiIiBQ4SoiIiIiIiIiISIGjhIiIiIiIiIiIFDhKiIiIiIiIiIhIgaOEiIiIiIiIiIgUOEqIiIiIiIiIiEiBo4SIiIiIiIiIiBQ4SoiIiIiIiIiISIGjhIiIiIiIiIiIFDhKiIiIiMgZtXbvWuZsnkOKLyXWoYiIiIgc4Yh1ACIiIpJPbNwIM2ZA4cLQqRN7jCw6vd+apANrcYYi+J0GD5e9gR7dn6FSkUp4nd5YRywiIiIFmBIiIiIi8rdkBDJI9adSOqE0NuPQYtOuXWHKFDBNsNvB4aDbLSZLSwUIOjj0icPklW2f8fMjX/JLNRcPtXiIp1o9hWEYMbwbERERKaiUEBERESng/CE/k9dM5q+Uv2hUrhFtqrTBSE+HiRNh92649FK2XVCRu74dyLT104gQASDBmcDozqPp/sF8mDz56AXDYbbGh1mUCEF79vfKdEEmITKDIYbOHcqrC14lEA5QK1SURge83BLfnBZPjcRwu/PwX0BEREQKIsM0zVjH8K81atTIXLx4cazDEBEROWekB9L5ctWXfLHqC6ZvnE7EjGCaJh6HhwsTqjPzv3/h9YXB7+dgYRc17wqxxxnIdR1XxEbWCzZswVC28eWl4dJ+kBaX+71r7YG1JcC0gREBVwSCNjABbwj6LIfrUytQfPhYGtS5/Az9C4iIiEh+ZRjGEtM0G51snoqqioiIFDA/bf6JUq+U4pYpt/Dt+m8JRUJWQgSTzFAmv+9N4oULUyEjA0IhRtfM5IAtdzIEoHh6BDMcyjVeey8YUZ65uELQdCuYh3bJ2EzwOyBisxIkGS4YdxF4/9zG+RdewfK3Hz+dty4iIiJyhBIiIiIiBYgv5KPz+M5khbKOO8fvgLH1j77+pQKE7NHn7vdCOEoJEFcY3pkG3oC1CgTAHYTiWbCyFHDoHGck97lZDphcC7xhOP/+F2DnzlO7OREREZG/QQkRERGRc0EwCKmpVtHSf2HGphmEIrlXdOR0bJLjgj1gD0efF3DArMsqgy33R4o+K2DGOOi2Gi7ZBvcthO4rYXH5o3NsUW7HboLnUIhhTPyfjz9pvCIiIiJ/lxIiIiIiZ6OMDBg+HHr1gqZNrVa2JUpA9erw/ff/+LKBcACTkyRVTOiZdPRl/yXgPk5CxMCg5sffQ//+4HKBwwFuN7RsCV4vzfbE8cVEg18+iuOabW7WljY4tqlMKMonEWcEbjr0/vYIOIPHeXMRERGRf0FFVUVERGIpEiEwbw7DN0xgVNZ8TJuNW2pcz939RxK3Z7+VGMnJ64WffoLGjf/22x30HaTs/8riC/uOO6esqwRrXvVTOCsCmZmQkMCyy+twY9sDrNu//si8kt6SfN3ja5pWbGoNZGbCgQNQurSVGPH5rNeFC8OsWZCSApdfTsvvb2D+1vkAuP1g2KzEB1gJktemw12HfqwHXHZcSaugZs2/fa8iIiJSMJ1qUVUlRERERPJKWho4ndYKCoDVqzHbXclVbXcyv3yYTKc17InYabw1wuwPzehLOQ0DOneGSZP+URhjfh9D/6n9CUaCuY61rdqWb3t9iys1Az77zKrf0aoVtGkDhkG6P539vv0UiStCEXeRf/T+87fMp91H7cgMZQJQOAOG/AYVUqHDOiiVYXWcCcXZsT/wEI6hz+e6hmmafL/heyb8MQG3w80tF91Cs4rN/lE8IiIikr8oISIiIhJjpmkybvk4Zn7+Eo+N3cR5u4LYbHaMa6+F99+H+vWZa9/GNb2s7irHSvDDj+OgafJxLl67Nqxa9Y9jW7t3LW/++iaLkheR4EqgdeXW3NbwNioWqfiPr/l3LNi6gEdmPELSriQqFanEM5c9Q9faXcn46UfSPhjGXxsW8U7VPWwuHcfAOn3pccdbGIfqlJimSc8ve/LNum/ICGZgYOBxenioxUM83frpPIlfREREzl5KiIiIiMTCvn3w1lswfTpL7Lt5ucp2Rk70U+iYrrWmy4VRrRps28aL9dN5sk2ULi4mPD8DHp0f5T3sdqu2yNixZ/JOYiKQeoCmT5RhTeEAWYdWzMQHoHekLsNfsAqLzP5zNp3GdyIjmH07kdvhZmG/hRTzFKNSkUoYRpT2NyIiIpLvnWpCxJEXwYiIiORHK3atYM7mOSR6E+lcqzPegxlw0UVW3QyfjwbARwsh5+/lRiAAmzeD3U7ZNHCHID1HQsQbhBLHK/Ph8cDjj5+BO4q9iSOHsL7Q0WQIWKtnxgZX8uCqeVSv05Ipa6eQGczMdW4gHKDxiEY4TYPikThGNXyWq7rcn4fRi4iIyLlEXWZERESOIxgOMmbZGK766Cq6TujKd+u/wzRNImaEPpP60HRkUx6a8RD9p/an/KvlWfK/IdYKEZ+VybABLtPqmpKL0wmBAN1WgyMCORu/OOwObtzosQqoxsVZRUrj4+Gqq2D+/HxbZHT65hmku3KP2034ecEEABJcCdhtOZfUQCQSIUSYLCNEsj2Dq5c9QOWnCrFy98ozHbaIiIicg7RCREREzj2ZmbBhA5QrB4mJZ+QtwpEw7T5qx6Lti45szfhx048MbDyQ+mXqM2n1JLJCWdnO6cwEtgRDuZ42mECuzRuhEDzwAAlvvsnssZlcfwPsKAQYNkqVqMgXN0ykSPc4WLoUKle2Cpva8v9zjAqe0jhD2wjm+IRiN6F0ySoA9LmwD68tfI1QJJR9Us5/ZAO22NKp9149pvWcRvvz2p+xuEVEROTcoxoiIiJy9vD5rBUWh9u25mSa8MIL8NxzVh2NYNDqtjJmzNHOLafJpNWT6DupL+nB9Gzjboeb+mXq88u2X3KdkxCyM2dUmIY7coQNhA1wHPqRG3G7sV19tdUlZsECeO89zH172XDtpZjXduK8snULbP2LPxfPoO7kK4903AEwIlDab2frc1k47NaB0b+P5u5pd+O0OQlFQrmSU9F80OkDbmtwW4H9txURESkoTrWGSP5/1CQiIme/UAjuuw+KF4fzzoOSJWHYsNzzxo+H55+3VoikpVkJlClTYODAv/V2B7IOsCNtByd6KDB1/dRcyRAAu2Fnb+beqOcYLhdBb1y2saAdFpWHLy+0c9Bt4CtVHNtjj8Hnn1sTmjeHjz7CmPYd5935GDXL1SvQv7BXbdSWibWepniWQSG/VVD1vMw4fur145FkCEC/Bv3YPmQ7I68dyTvXvEMcubfQ5HTv1EE8O/OpMxm+iIiInEO0QkRERGLvgQfgvfesRMdhXq+18qN796Nj9erByij1IOLiYP9+65wT2Jm+k15f9WLelnkYGFQsUpFxXcbRrGKzXHMfmfEIry18jWAkmG28kKsQvev1ZuyKsbkKe5bwlGBn6P9w/Pc5K6ZgEOrVI/mT4aQWjuP8xPOxGXoWcSpCAR8rfpmCp3AJal10xUmTRHdPvJWRy8YQcOY+VjYV3psKHdYf2lXToAH2iV9C1apnJHYRERGJLbXdFRGRc0MgAMWKZU+GHFa3LiQlHX1dqhTs2ZN7ntsNmzZB2bLHfZt9mfuo9149dqXvIsLRKqcJrgRW372aCoUrZJu/bt866g+vn2srRnFPcTbfu5nOEzqzaPsi0gPpxNnjsNvsTOkxhbbV2kJKCixbBmXKwPnnn9q/g/wrpmkycsSd9N8xwho4lD9xhmD921Ax5eiyWBOIGHDg94UkXtQ0FuGKiIjIGaQtMyIicm5ISYFwOPqxbduyv27ZMncPW4AiRay6I8cxfuV4yr1Wjh3pO7IlQ8DqJDNiyYhc59QsUZMPO39IvDOewnGFKeQqRJmEMvzY50cKxRViRt8ZfH795wxpNoT/tvkvGwZvsJIhh+Np3VrJkDxkGAZ3DHifRT1mYT/mv5Fr10LJjOwfeAzAZsKets2Yv2V+nscqIiIiZwetEBERkdiKRKyVFNFWflxxBcyYcfT1mjXQpIm1muRwEuXYrTXp6bBpE+FyZbGVSMQwDLanbafGWzVOWHTzxgtuZML1E6IeywpmMX/rfDwOD00rNI3a7lXOLqFIiMHTBvP+kvd5ZK7Jc7OidPkBQgZUeNzNpqf24XWeeLuViIiInDu0QkRERM4NNhv873+56394vfDii9nHatWy2tD26WMVX73qKvj+e7j+enjiCZLqJNL83YY43ymF51kH/SbdwicrPsHk+Mn/eGc8rSu3Pu5xj9ND22ptaVGphZIh5wiHzcF7Hd9jco/J7KpS4rj/64cN8IcDDJ07FH/In6cxioiISOxphYiIiJxYJAK//mptbWnWzNoOcqp27rRqgFSpYiUwTuTbb+HZZ+Gvv6BBA6u17sUXn9r7DB/OjqeGUOu2LFLjOLIcIM60U7ZYRZJTk3MVRwWwGTYqFK7AHwP/IMGVcOr3JecMMxgkUKIorrTMbKtElpWGlv0g41BTIAODB5o/wMtXvhyTOEVEROT00QoRERH599auhWrVoF07uPFGa2vLO++c/LxIBO66y0qEdO8OF10El18OqanHP6dDB/jtN9i1y1r1cQrJkK0pW0nalYT50ku8VzcLv51seyP8Rpid6TuP29nlhjo3sPiOxUqG5GOG04l96e+kxnFkpUi6A5rfDhmuo/NMTF5Z8Apjl42NSZwiIiKS95QQERGR6CIRKxGyZYtVmyM1FXw+ePhhWLDgxOe++y6MGwd+v7WyJCvLOueOO05LaDvTd9JidAtqvlOT5qObk75jM8tKgz9Ky1WXzUn789rjdXqxGTYMDLxOL0OaDWH89eMpGV/ytMQkZy9HjZqkbt3AbT3jeaUZdO0BWQ6iFhZ5fNbjeR6fiIiIxEZMEyKGYYw2DGO3YRgrjxkrbhjGj4ZhrD/0d7FYxigiUmD9+iscOAA5t1ZmZcGwYSc+9803c7fR9fthypRc42n+NL5b/x1z/5pLOHKcbjOHhUKYH3/M+qY1efCVhVy+yke6P52F5aHxdnDn3hVDMBLi+cufZ/bNsxncZDCDmgxiRp8ZvNru1RO/l+QrFUtW5/1xByj81nA2NqoavcoqVntmERERKRgcMX7/McA7wLhjxh4BZpqm+aJhGI8cev1wDGITESnYDhyI3uLWNKN3hDlWSsrxj2VkHCmgOvr30QyaNgin3YlpmnidXr7r9R0NyjbIfV4kAp06EZk7h0szrY4xbTfCyIbwYDuYPAFebwoBO0QOpfvdhovWVVpTu2RtAJqUb3LS25b8y2l3MqDRAFpXbk3tYbWjzqmZWDOPoxIREZFYiekKEdM05wL7cwx3Bg5v4B0LdMnToERECrjtadtZuXslwUsaQyCQe4LXC926nfgibdta3WNyKlcOEhMB+HbdtwyYOoCsUBap/lTSAmnsythFu4/bEQxHWerxww8wbx72zKPtcxOCMGAJZDihfS94fXFx2u+IJy5io5ijEIOa3cOkGyf9nduXAqBWyVq0rNgy17jNsDHsmkOrnyIRq6bNrFm5VzuJiIhIvnA21hApbZrmDoBDf5eKNskwjP6GYSw2DGPxnpM9qRQRkZPak7GHy8ZcRvU3q9NsVDNKjazJ0ntvsBIgh1eKeL1Qsyb07Xvii73wAhQtCnGHWnjYbOBwQM+eEAzyw8Yf6DKhC6FIKNep+zL3MWxRlC0506ZZtUxyMIG2m2BLWQ+7/vMwU0ek43s2zP7HU3nlyldwO9x/819CCoK5t87lnib34HF4sBk2qhatysw+M2lRqQWTvnyO5BIuUi+9hNQObQklFrdq4oiIiEi+EvO2u4ZhVAGmmqZZ99Drg6ZpFj3m+AHTNE9YR0Rtd0VE/h3TNKnxVg02HdyUbdzr9LK4zpvU/mK2tU2mWzcrGeLxnPyiO3fCo49av0gaBoTDkJBAuHIlyt28h92Zx09mu+1utg7ZSqI38ejgf/5jteLNsWol1QV3dbHzc9NyrLhrBUXdRRH5p0b/NoJrrhhAqfTsT41CcU4cvy2GCy+MWWwiIiJyas7ltru7DMMoC3Do790xjkdEJN+757t7ciVDADKDmfwnNBM++cTasjJgwKklQwBKl7Y6y0QiVjIEID2dNQfXk5lx8KSnf7zi4+wDffqA3Z5rnmGzUa3Pvfw+4HclQ+Rfmz7qUbyBKB+QAkHo1cv671lERETyhbMxIfI1cPOhr28GpsQwFhGRfC89kM4HSz847vEtKVtOfIEtW6zEx8EcSY5t26xjOcRnBAmHc2+VOZYv7GPzwc3ZB6tWhY8+goQEKFwYChWCxEQKzZ7Pfzu+SglviRPHKXISoUiIyIH9HLt2NrkQPHoFdOwJ/aqt5KYHq/HeovfICmYd9zoiIiJybohplxnDMMYDlwGJhmFsA54GXgQ+NwzjNmAL0D12EYqI5H9r967FZXfhD/uPjDnC8PRPcPciKOJfCCMugQcfhNWrrS4z110HVarADTfA7NlWrRC/H4YMgaFDrS0yTmfulr1AlYNQK9XF8hJBImb0p+0JrgQurXRp7gPdusE118D8+eByQfPmVm0SkdPAYXOwrnYpXF9Zi1OTSkGL28Bvh4ADbBGIGH8xYdpAnpr9FEsHLKVikYoxjlpERET+qZjXEDkdVENEROSf25m+k6pvVMUX9h0Z+3ASdP8D4nMu5Di8ZcXlgmrVYMMGKxFyWHw8vPMO3HKL9bpxY1i6NPs2A4+HzU/fy2Vx49mftZ/MYCZhM3zksNvhpmaJmiy+YzFOu/P03qzISYxaOoodD97J//0c4pre8HMlIEr3aYBralzDt72+zdP4RERE5OROtYaIEiIiIgVROAxTpsAXX0B8PI9W3sAbtl/xhXyUSofNb4DnBLtashywqiSUyoCKqTkOXnABrFxpfb1pE1x6KaSlQTBodZtp1QqmTCHidDBn8xy2pGxh7d61TFoziWAkSM96PXmoxUMkuBLO2O2LnMjY5WOZ/tytTKhjYh4nGXLYtvu2Ub5w+bwJTERERE6JEiIiIhJdJALXXgs//QQZGWAYmB4Pn3c5j5trr6H5FpNJHwUp4ov+8+HdxvDwlWCPQMAOzbbBxM+h+OGSCqVLWx1mDgsGrZa527ZBkybWqhGRs92cOcT/eBmZJ1mkVCq+FH8M/CN7R6RD9mbuZcHWBRRzF6NFpRbYjLOxdJuIiEj+o4SIiIhEN3Uq9OhhJUOOFRdHxtqVpAbTKVO3KcaxW2EOSS4ENQdDpuvomCsELbbArHFYK0B69LC60oic4waN68H7Gz8nZDv+ZyW33c3DLR/mmcueyTb+4rwXeeanZ4hzxGGaJkXdRfmxz4+cn3j+GY5aREREzuW2uyIichwHfQd5+MeHqfpmVWq/U5s3fnmDUMTa27I7Yzcb92/MXqj0wAF47jlrm0rv3rB4MUyalDsZAuB0Ej93IWVr1Mfo0SNqe90n22RPhoBVbHJhRdhW3AFFisB//3s6b1kkZl6+aTRtarQ94coOX9jHR8s/4uX5LzNy6UhW7l7J8MXDeXzW4/jDflL9qaQF0tiWuo2rP7ma/PAgSkREJL/QChERkXOEL+TjouEX8dfBv450hPE6vbSu3JqsYCbFfpxH3yVhXGEoY3pp2PtBGDkS9u0Dn89aveF2Q+vW8MMPVh2RYxUqBGPHQteu1jaXZ56BYcMgNRUqV4bkZOreEeSPkrl/bhQK2vgp60Ya3vcyVKiQB/8aInlnzd41jFoyitd/eZ0w4ePOMzBydWw6VoIrgVl9Z9G4vLVtbE/GHpJ2J1GpSCVqFK9xRmIXEREpiLRlRkQknxm3fBwDpg7AF/JlGzcweGaWyZCFkBC0xtKdELZDEb+Ru/VtoUIQCkFWVvbxwoWt2h9RVoYAkJzM/R/34R3fXAI5filMcCWw+4HdeJzHOVckH2g2qhlLti8hGAked06jZLhvIVRKgenV4d0mcMBrHSvsKsxXN37F5VUvZ8j0IQxfMpw4exyBcICmFZoy6cZJFHEXyaO7ERERyb+0ZUZEJJ+ZuGpirmQIgBExScw8mgwB62t3kNzJkMPuv99aLVKokJUIKVrUKnx6vGQIQPnyPHj3pxT1lsBlO7pvxuv08nLbl5UMkXxv6k1TaVOlDXH2uKjHe6yAn8bAjSuh5VZ4bB4kvQeJh3aoBSIBLqlwCSOXjmTE0hH4Qj5S/ClkhbJYsHUB/ab0y7ubERERESVERETOKqYJgUDUQ6v3rI467ohYT6NzjR9vZX8wCP37w/btMGYMTJhgrQxp0eKk4ZVJKMPyu5YzqMkg6pasS7tq7ZjSYwp3Nb7rpOeKnOtKeEswvc90tt63Facte/sZRxiGTYP4INgPjXlCkJgJD823Xt9Y50YSXAm89strZAYzs53vD/uZun4qaf60PLgTERERASVERETODqEQPPKItWLD44GaNa06H8fI+QvUYc4wtPor93jEiDLZ4YCLL4aKFaFYMbjuOmjfHuKiP/GOpkxCGV696lWSBiYxvc902lZre8rniuQHJeNL0uG8DtiO+RhVa6/VijqnuDB0XGd9/cXqL1iyfQkHsg5YAzkWcNkMG2kBJURERETyihIiIiJng0GD4O23re4vkQisX28VN/3ttyNT6pepH/VUt2nDHmVnjMMEGja0tsYUKQJeL9SrB19+eYZuQqTgeKv9W5SML4nb7gbggBucURIiAHsP7SbzhXy89strDN1UlW2vQvhZ2PgGdPvDOp7oTaRsQtk8iF5ERERACRERkdhLSbG6u2TmWAGSlZWthe0zlz2D1+nNNsXr9NK/7cP8/Nb9pLkN0lyQ5bBWhxg9e1oJleRkmDgRfv0Vli6F0qXz4q5E8rWKRSqy4Z4NvFmuHzcn2Wi4y2BnfK5FH6Q74bXm1tcRM8Lm1b/Qb+xyyqdZH8KqHYSxk+GG9S5GdByBYURb2iUiIiJngrrMiIjE2sqV0Lw5pEVZKl+9OmzYcOTlnM1zuG/6fSTtTqKEpwQPtXiI+5reZ/0S5ffD3LlWO93LLrNWhojImZOZCaVKWSu7ojCBPxKh3iDrtdvu5qGFNp6dlnv7m++8qrjXbTqDwYqIiBQcp9plxpEXwYiIyAlUqWLVEMnJZrO2vByjdZXWLB2wNPp14uLgyitPf3wiEt2cOWC3H/ewAVQ/CBUPws7iTorEFWbwnD1R57q37jhDQYqIiMjxaMuMiEisJSTA4MFWjY9jud3w5JOxiUlETs528o9RIaedNv4y3NnoTpbduYxEb2L0idWqWX9nZcEvv8C6dacxUBEREYlGCRERkb8jGMxd6+N0ePFFeP55KF/eSoS0aAGzZ1tFUEXk7NS69UmnFDKdjH1oIW+1f4syhcrC0KGYOZOfXi+88AKMGmVtwbnqKmjQwFohlpx8hoIXERERJURERE5Fair07m2t5ihcGOrXh0WLTt/1DQPuvRe2bbOeEM+bB02anL7ri8jp53bDF19YCQ2PJ/dxjweuvtraFgf4Q37uLLeU/lcF2FIEwgakVS4LH38MiYlwzz2Qnm59v8nMhBUrrLbY+aDem4iIyNlINURERE5Fx45Wx5ZAwHq9fDlcfrlVELVy5djGJiKx064dbNlitbNetQpmzbK+L3i9cPvt8NJLR6b2/6Y/X6z6gqyLQoy8yBrzOlP4rmEJWj30ppUMPVY4THjVSua/PIjdndrQtlpbirqL5uHNiYiI5G/qMiMicjJJSdC0ae6tMi6Xtarj5ZdjE5eInJ1M01r1dYz9Wfsp92o5/GF/runtqrVj+ohMa2VYzksBESDggG1FDPbeezvNnng/1/VFRETkqFPtMqMtMyIiJ7NxIziiLKgLBKwnwSIix4qSrNiRtgOX3RV1+sYDG6FjR0JxuY8bgB3whOC8fSYX/vcDDjx+/2kOWEREpGBSQkRE5GTq1Tu6VeZYbjc0a5b38YjIOadqsaqEzXCucbthp2mFpnDnnaQV8xI6ycKP+CDEv/7OmSnuLCIiUsAoISIicjLVq0OHDtmLJtpsVo2AO++MXVwics7wOr081vIxvM6jHWYMDDxOD0+2ehKKFGHGxJf5vbxB5CTXimDCjTfC9ddbtUvCuRMtIiIicnJKiIiInIrx4+Hhh6FMGShUCLp0gcWLoWTJWEcmIueIxy59jPc6vEftxNoU9xTnmvOuYeFtCzk/8XwAOl3Sh8e7lyDLeeLruPwhmDrVSobcdJPVySZysjSKiIiI5KSiqiKSb6zcvZLHZj7Gr8m/UqFwBZ5s9SRdanWJdVgiIqdsS8oWpt59Jbd+to4I4A1ZdUQO89vAFbHG0lwwoS6sLAUNAsW54YWv8TZpEaPIRUREzh6nWlRVCREROXdt2waPPw5JSeyvU5XGFb7jT48PE+v7mtfp5X9X/o+7Gt8V40BFRP6ewM5kbDNnsWPBDyRM+IqEVB82w4YtHMGIRPizKDS9HTJc1p94PxQNGPzWYyblGrWJdfgiIiIxpYSIiORbk9dM5vlvHmL7zvVc+hf8ZzbU2A9hAy5bGEfJAAAgAElEQVS/GX6ucnRukbgi7HlwD077Sdagi4icrUwTDhyA776DW26BUIh2vWFmNYgcs/nZEYbrMiry2atbYhaqiIjI2UBtd0UkX3rz1zfp9VUvFmWuJ7kwfH4BNOoPG4uDw4SvPgOOyfMGI0GS05JjFq+IyL9mGFC8uFW7CCv5OytHMgQgZIdvvNvg3XeheXPo3BlWrIhBwCIiIucGJURE5KyUvH0NHz/fg6ld6hC48XpYsABfyMcTs54gM3i03WTEZi0X/28r63VRH1Tff/Q6ETNCojcxj6MXETkD4uPhpZcwANtxFvg6wiYMGgQLF8LXX8NFF8Fzz+VpmCIiIucKJURE5Kzz1HcPUeO9OtyV/hk9a6+mbJUvWdr7CjaNeAkjW3lBS9gGP1e2vjZMCNitrz0OD30v7EuCKyEPoxcROYPuuw/b4HvovAacoeyHXCG4KSnKOU88AbVrwwMPwK5deRKmiIjIuUAJERE5q8zcNJPXfnsDn8MkPQ7S3LDfC+27+Sj59EsEwoGo51U+aO2USS5qsC/RayVDLurL29e8nbc3ICJyJhkGvPkmwx6ZS7WAh0J+iAtCQtCgzn4br/x4nPPWrIG33yZUry7LV/x43O+lIiIiBYkj1gGIiBxrxJIRZJhBci4EyXLAmpIGXRJbMmXffHwh35Fj3gA89jMYbjflFi5hbbnCFPcUx+v05nH0IiJ5o2TDS1nVIJ0f13/Pul2rqFu+IZe16InhP8EKkECA0P69zBjcgdbtvXzQ6QO6X9A974IWERE5y2iFiIjkndWr4eqrrX3wZcrA0KEQyr7mOy2QlisZAtZQhi3Mh5e+ynW1ryPOHke800sx3Ly762KuHPw6pKbiqFWHCoUrKBkiIvmezbBxVc1rGHzpA7SpdjnGrf1Oeo47DJevDZLiT6HPpD48NvMxfv7rZ/JD10EREZG/S213ReSMygpm8cTsJ/hx9ih+fj2FQv5jMrFeL3TvDmPGHJk/bvk4Bn49gIyIL9t1vAHYPb0e8Yutjgmp/lT2Ze6jYpGKOGxa7CYiQjgMNWvCpk3HnRIBPqkLfa+3XtsNOx6nhzqJdZh580zVXBIRkXxBbXdF5KzQcXxHhi0axi0/peAO5vimk5kJn30G27cfGbqp7k00qngJ8bgAcITBE4T3llUgfsp3R+YVjitM1WJVlQwRETnMbof162H0aGjVymrV68j+PTLLCfe1P/o6bIZJD6SzfNdyHp/5eB4HLCIiEltKiIjIGfP7jt/5Zdsv+EI+LkmGuEiUSXFx1laaQ5x2JzP6zmBs90/pW7sHgyt2Y3H7SfT9diuUL593wYuInItsNrj1VpgzB9atgxYtMN1xpMZBqgvuuRr2xec+zR/283HSx3kfr4iISAzp0aqI/HN+Pzz8MIwaZa32KFcOhg+HDh0ASNqdhM2w8q5JpaDJNnDm3KXn90P16tmGHDYH3ep0o1udbnlxFyIi+VOJEvDTTxhbtzJ26lM8s/tz9pN53OnBcDAPgxMREYk9rRARkX/mtdes4qhvvgnp6RCJwLZt0KkTTJ0KQM0SNY8U6nutGfhzpmDdbmjbFqpUydvYRUQKkooVGXTnaEbe8DGXVb6MeGc8Ro7q1Q6bgy61usQoQBERkdhQUVUROTWrVsErr8Aff0CxYjB3Lvh80edWqABbt2KaJo0/aEzS7iQC4QDNtsDwqXDBHrDFxWHcfAu8/jp4PHl6KyIiBdm6fetoPqo5WaEsMoOZJLgSKOYuxm93/EaZhDLZ5u5K38Ubv7zB7M2zqVq0Kg80f4CLy10co8hFREROzakWVVVCREROaPPBzXw4+Rl2TvmYq9eZdFoTwRGtFsixbDar2wFw0HeQgd8OZOKqiYTNME3LN+WDdu9Qp9yFVgFAERHJcym+FD5e8TGr966mUblG3HjBjXic2ZPT21K30eD9BqT50/CH/diwEYeNEXE30KPnczgqVYlN8CIiIiehhIiI/GPJqcks2r6Ijfs38sSsJ4gEfATskOCHC3fBrLEQFz7BBQoVgtTUbEPhSJiwGcZld53Z4EVE5LS4/evbGbtsLCEzlP2ACfFBGFi6I88PmqRuXyIictY51YSIfoKJiMU0MZOSuGfZC4zcMgmXPY7UwKGkxqGFHOlxsKwMjG4Ad50oB/nAA7mG7DY7drQiRETkXPHDxh9yJ0MADMhwwbu7p+KbcjdvdX0/74MTERE5DVRUVaQgM01IToY33oCyZfn49sZ8uHYCvrD/aDIkh0wXfHLhCa7Zsyc89dSZiVdERPJMCW+JEx7PdMLIpA9J9aWybt86dqXvyqPIRERETg+tEBEpqMaMgXvugbS0I0NvXWs99TsZTxCrQ0y1apCSAqVKQbduMHgwFC58xkIWEZG8c3+z+7lz6p1kBDOOP8mEqm9WxR/2E4qEaFGxBROun0DJ+JJ5F6iIiMg/pISISAESMSPM/nM2xsSJXPbUaGz+QLbjKe6TXyM+AANWOKFvXxg2TIVRRUTyqV71erFqzypeX/AqoWCAkA1ydOvFZwbJ8u0/8vrnLT/Ta/iV/ND+U6hRA1yqGyUiImcvbZkRKSC2p22n5ts16fpZV8q/8n6uZAhA19Xgirpd3MDj8OC2x9Gn8rV0+34LvP++kiEiIvmYYRg8f8XzJD+wg/cC7fDk+PngMm2YxyRICvtg0kdBvn54ORkN65JexMPsJ/sQikT5wSIiInIWUJcZkQKizdg2/PzXz4TNMOnPWR0CctrvgYv7w+54q1aIAzsuRxz3NbuPKkWr0Lpya84rcV7eBy8iIjG3aN5nPDT7MRZHkilTuBw2h5N1+9cdOf7dR3DZZnAf04Us0wFvto7jivvfpUn72/I+aBERKZDUdldEjjiQdYAyr5YhELZWhSx+Hy7eEX1umgs+bOzgh4ZFqHpVDwY2uZvaJWvnYbQiInIuGDp3KM/9/By+kI9yqbDhLXKtIgEIGhBwQLDBhRSdMc9qzS4iInIGnWpCRFtmRAqAQDiAcczG74fbWk/toilUuiL39HyTqf/bztsd3lEyREREohrYeCDF3MVw2pyUTofAcXZROk1rVaJ3aRIMGpS3QYqIiJyAEiIi57iMQAafrfyMUUtHsSVlS9Q5pRNKU7VY1SOvZ1aHa2+CJeUMAk4bxMdDy5Ywdy5s2QIDB6oQnoiInFBxT3F+H/A7dzW6C1+Nyjgjxgnnu0ImfPYZhMMnnCciIpJXtGVG5BxgmibJacm4HW4SvYlHxudsnkPH8R0xMIiYEUKREI+2fJSnL3s61zUWJS/iinFXEAwH8YV9xDvjKRVfikV3LKKEt0Re3o6IiORDGS8/B088EbVG1RE2G2RlKekuIiJnlGqIiOQTC7YuoO+XvUk+sIVIJMwlyTBsXzOqvzGWsp81JsWfkm2+1+lleu/ptKzUMte1dqbvZPTvo1m3bx2tKreiR90eeJ3evLoVERHJ51Z/+D/2P/swF22P4A1GWYrcuDH89hsAKb4UPv/jc3ak76BZhWZcUe0KbIYWL4uIyL+nhIhIPpCcmsz575xPRjDjyJg9DFUOwn/nORnQ3U1aIC3bOQYGt9a/lVGdR+V1uCIiIkQiEYaOuZVBg8bhCVqFVkNOBw63B+bMgQYN+H3H77QZ24ZQJERmMJN4Vzz1y9RnRp8ZxDniYn0LIiJyjlNRVZFzhC/k45MVn/DErCcYnzQef8h/5NgHSz8gFApkmx+2w+4EWFksSCTgy3U9E5PMYOYZj1tERCQam83GU/3GUnTzTuxPPInZoQOOBx6EVaugQQNM06T7F91J8aeQEczANE1a/pHOPa8tYHvzuvD++xCJxPo2RESkADhOnwkRyQvJqclcMvISUvwppAfSSXAl8MjMR/j19l8pk1CGDfs34Ddzb8Y2gYqpEAyHIEdV/3hnPD3q9sibGxARETkOW6nSuJ7+T67xTQc2sSP9aO/397+BPsvBE44AG2DRnTB4MEyaBB065GHEIiJS0GiFiEgMDZw2kJ3pO0kPpAOQHkhne9p27v3uXgBaVW6F23DmOi9sg0bboGdqZTwODw6blduMd8bTrno7Op3fKe9uQkRE5G8wDIPDW7br74DeK8CTs/FMMAgdO8KsWXkfoIiIFBhaISISI6ZpMm39NMLm0U+B9XfAZZtDHFz+FVyTQe8Le/PCzy+QvG8zwUP/b/UGoP16qJQGvW99nUfK12bs8rGk+FLoUqsLbau1xTBO3PpQREQkVqoWrUqFwhVYv389V22AuNAJJt96K/z1V57FJiIiBYsSIiIxZDu0SMuIwKdfQqd1YI9A0B6C8uXxzpzJkgFLuOPzPvy0aholsmDAIqizz+Ctt3ry3ws7YxgGz1/xfIzvRERE5NQYhsHEGyZy2ZjLCHgzMI2AtRc0isjWLZQfWoy4hCIMvmQw/3fJ/2G32aNPFhER+ZvUZUYkhrpO6Mo3677hpmVh3psKCTnLhZQvD1u3gmGwN3MvE//4gjR/Glef1556pevFJGYREZHTIT2QzrfzPqTbVffhCOXcM2PJcEKhR8G0gdfhpUe9Hoy6Vl3URETkxNR2VySW1q+HpUuhcmW45BKIsoXl06RPuW3KbfjDfn4abdJqS5TrJCTAzz9D/fpnPmYREZFYmDwZunbNNZzpgA8uhv9rf3TMbXez4Z4NlC9cPg8DFBGRc82pJkS0ZUbkNNmbuZfVO5Jo+NDrxE/7EZxOME2oWpXfP3qFSfvn43V6uanuTQDc/vXt+MJW21zH8boLGgaETrS5WkRE5BzXpQvs3AmtW8O6dWCz4TciTK5l8tCV2ae67C4e/PFBNu7fSPnC5RnSbAgtK7WMTdwiInLO0woRkX8pYka4b/p9vL/4fYYshMd/9BN/zNaXkN1gdlWDq3uDw+bAbrPTqWYnJq+dTCAcAKD/Inj1hyhbZkqWhB07wK790iIiUgDs3w/r1vHwhuG8tukTQmbuhwJOm5NgxPqB6XF4eKf9O/Rr2C+vIxURkbPYqa4QUdtdkX9p2KJhjFw6En/YT7+F2ZMhAI6wSas/IyT4IgQiAbJCWXy15qsjyRCA0Q3h1/KQ5jpUV87thvh4+OwzJUNERKTgKF4cmjbltvaP4XK4sh2yG3YMjCPJEICsUBa3f3M7IxaPyOtIRUQkH1BCRORfen3h62QGMwHw5lzhcYgJeI455rK7iLPHHXkdssOVfaHnTS72D74dnnsONm2CNm3OYOQiIiJnp5olajKt5zRqFK+By+bCZXcR74zHjNKOxsRk8JQBvDy0vbVVVURE5BSphojI35DqT+Xpn55mQtIEDMOgz4V92J+1H4ASGfBnUSiVAY4cn8e2FoFdCUdf2w07l1S8hN+SfyMjmIGBgSfOQ61eAylx5St5eEciIiJnp9ZVWrNu0DoO+A7gcXjoOL4js/6cFXVuwAGPhL5n0rMVmfZwEsU8xfI4WhERORephojIKQpFQjR4vwHr9q07st0lzh6Hx+nh6t8OMvprcIXA4Ogfnx2CNri6DyyodPRaXqeXHUN28Gvyr3ya9Cl2m52+F/WlVeVWsbg1ERGRs963676l+xfdyQplHXeOKwReT2HGdfuYTud3ysPoRETkbKK2uyKn2ZQ1U7h5Yi/iD2SwzwN+pzVeLcvDylez8OSo+xYyYMTF8FJL2FIUbNhwO92Ypsln13+mD2oiIiJ/04vzXuTRmY+eeJIJHqeHkdeOpGe9nnkTmIiInFXUdlfkNDM/+IBNwzJwH0p8fNAQ7r8KOiT5sdnsQDjb/LABfxWFbYUhPuLg7kuHUL1Yda6rfR2J3sS8vwEREZFz3CMtH6G4pzj/N/lOsnLuTz1GViiLIdOH0OOCHthsKpknIiLRKSEiciomT6bDsB9x+o4O3f47mAbsKuHEkSMZAuCIQOl0uGG1jWeufp7z2z6YhwGLiIjkT/0v7k/xzbu4bclTpLqw9qge69DrXRm7OP/d8xnTeQwtKrXI6zBFROQcoC0zIqeiYUP4/fdcw5kOaH9vCX56NwPD58t1PGgDR5VqGH/8YbXSFRERkdNi67TxXDyvD+m2MFlOsEUgcriI1zHinfEk3ZVE1WJVYxKniIjkvVPdMqM1hCLHsWT7Em6efDNtx7XlteJrSXPlnmMzbIzr9w3Go4+C14tps2FitdkNu+NwDLgL47fflAwRERE5zSpecxOrn9hFhwuvx2bYiNjIvVoECIQDDFs0LM/jExGRs59WiIhE8fGKjxkwdQC+kI+IGcETtlEmJcLSEVD02IUgxYvD7t1gt8OyZfDppxCJwI03QuPGMYtfRESkIBm+aDgPzniQ9EB61OMdz+vINz2/yeOoREQkVlRUVeQf8of8DPx2IJnBzCNjWfYIOwrB203gybmHBr1eeOklKxkCUL++9UdERETy1J2N76Rt9bbUHVYXf9if7ZjH4eHSypeycOtCPlj6AWmBNG6ocwPX1b4Ou80eo4hFRORsoC0zIgCrVsG110KJEqxofT5GMJhris8JX11SCBITrZoin34Kt98eg2BFREQkpxrFa9Ctdje8Tu+RMYfhoFBcITICGbT9qC1jlo1h4qqJ3DrlVjp82oFwJHdRdBERKTi0QkRk/Xpo2hTS08E0KbZmP8E2gDP31BJ1m8CeGXkeooiIiJzc2K5jeeOXN3h30bukB9LpeF5H7rnkHpqPao4vfHTPa0Ywg3lb5vHNum/oUqtLDCMWEZFY0goRkaFDITMTDtXTqbEfau0Fe46HRvHOeO695N4YBCgiIiKnwmFz8EDzB/jz3j/Z8+AePuzyIav3rsZhz/0MMCOYwVerv4pBlCIicrZQQkRk4UIIZ89+fD0eah60EW/3UCSuCHH2OB5s/iCdzu8UoyBFRETknyjkKoQRpf2MPQJF5/wKu3bFICoRETkbaMuMFBjhSJjZ379Hyqcf4nV4OO+Oh6nRohPUqGFtmzlGhVT44wMnS5d+w25HgCblm1DCWyJGkYuIiMg/1a56u6jFU11huO2LjTCpCaxZAx5PDKITEZFYUkJECoRQJMSHPWrRe9JGHBFrLPjJtSy563oufuwxmDPH2jZzmMeD0bUrF9e+IjYBi4iIyGkR54jj+17f037slYQz0sCEgB1e+x4u2h4m3bWFzx+/hmuHfkGiNzHW4YqISB7SlhkpEKZ99xa9J23EEwJnxPrjDUGd9yaSXtQLH30E5cpBXBy43dCrF4waFeuwRURE5DS4pMIl7Ex4hs8nOflwCmx/Fe5cYh1LCEDVST/RZHgjsoJZsQ1URETylFaISL4VMSPM+nMWmw5sYstXb3OlmXuOPQJ/jX2bWq98CF27wp49ULiwlRQRERGRfMNVszZXJbshLZht3ATiQtBh9lZGNB/BvU1VQF1EpKBQQkTypV3pu2g1phXb07YTMSOEK/iY2wemfwye0NF5pgEOh8t6YRhQqlRsAhYREZEzq107KF2aSHo6tkOd5ZILQbs+sKUI2MwIGT8MYXvadl5s+yKGkbsQq4ic28KRMNM3TmdR8iIqFalE9wu6k+BKiHVYEkOGaUZ5bH6OadSokbl48eJYhyFnkY6fdmT6xumEIkezH+4g3PsrvDjj6LwsBzhXr8VRo2YMohQREZE8tXs3gfNr4DqYBkCTO2BpGQgfU3M13hnPh50/pPsF3U98rVAIvv4afvkFqlSBnj2haNEzF7uIZBcIwPTpsHs3XHop1Dzx5/mMQAatx7Rm7b61pAfSSXAm4HK4mHfrPGqXrJ1HQUteMQxjiWmajU42TzVEJN/JCmbxw8YfsiVDAHxO+LC+lQQ5/Gf3Uw8oGSIiIlJQlCqFo9sNhAzYXBRWlsqeDAHICGbw5q9vnvg6aWlw8cVw883wyivw4INWUmT58jMWukiBcKoP65OSoHx5q+7fPfdA/fpwxx3HP3/jRv5s24gZ9y7hjxfTuX8+ZPrTOZB1gF5f9co1Pc2fxjdrv2H6hun4Q/5/cUNyttOWGcl3wmYYk+jfDP3FCpF0f2c8djc1b3uIytXOy+PoREREJJZsQ4bA+E9Jjcs60nkup4O+gye+yPPPw7p14PNZrw93quvZE/744/QFK1IQRCJkDX0G+1tv49qfAnXrQt++YLfD+efDVVdZXx+ay3/+Y/3JmfwYPRri4+GNN7KPr1sH9epxQSCAART1wX9mQ509cFsXk1V7VrErfRelE0oD8MmKT7jt69uwGTbsNjsOw8GUm6bQqnKro9fMzITvv4esLLjySm27P4cpISL5ToIrgYZlGrJo+6JsiRGHzcF1F1xPk86jYxidiIiIxFSdOtimfE2Vu/qR6dya67DL5qJrra4nvsb48UeTIcfauBF27ICyZU9TsCL52x+7/6Df21ewxLYLYyA02wKffpVEqYcfxGl3YrjdUKYMzJ8PJUvCU0/Bq69GXwkSicDbb0PlynDffdaYaULr1nAoGXKYNwQ9k+DJy2FfMdidsZsPln7Agq0L+G7Dd7ku3eHTDuy4f4dVb+Snn+Daa49ePxSCF1+Ee1WQ+VykLTOSL43uPJoi7iJ4HB7A2g9cNqEsL1zxQowjExERkZhr25bXRvTDsOd+Nhg2w5SJL8Mbv7zBt+u+JRwJ5z7fbs89dirHRAoQ0zQZu2wsdd6tQ8mXS3LdZ9exdu/aI8f3Zu6lxejmLLLtImyDkB36roCSmeCKgBEMWtvT/vwT7rrLqhnyxhvRk5GHRSJW0uTwnAULrBoj0eIzoO5uqFK0Cs1HN2fo3KFRkyEAwXCQb9Z+Y60MufZaK660NEhPt97r0Udh2bKjMaSnn/r2H4kprRCRfOmCUhewYfAGxi0fx5q9a2hSvgk31bsJr9Mb69BERETy1sGDMHOm9QH9iiugQoWTnxMMwsSJMHkyJCZC//5w0UWnPbQUXwqT1kwizZ/GVTWuomaJvKvrNWb5GEJmKNd42Awz6PtBGBg4bA6KuIvw6+2/Uq1YtaOT+vWD556zlssfZrNBvXpaOi9yyH/m/odX5r9CRjADgMlrJjNj0wx+H/A71YtXZ8yyMQSCfsxjHtHf+AfE5cxBhkIwZQocOADhKAnKnAzDSqLUrg2rVxM2TKKlKZ1hOFi2GFmhLNID6Se8ZCAcYE/mHhaOeJpGkSDOnBP8fhgzxvp++corVuKkZEl4+WXo3fvkMUvMqMuMiIiISH713HPw9NNHf4kwDKv97KRJ4PFEPycQgDZtrAKhGRnWL/pxcfDOO1Yi4DSZuWkmnSd0BqxWmBhwd+O7+V+7/5229ziRyq9XZkvqlqjHSqbDNeshYsDUmpBZKI71g9dTsUhFa4LPB1dfDYsXW8mjuDjwemHePKhRI0/iFzmbpU//hlLzOpPlyP67pt2wc3P9mxl17Sj6TenHh8s+zHY8Y6i1nSUXu91KQJYubSVGTsTthuRkKF6cHdO/pHCn64kPZp9iAvsqFCdl+W/Ufa8uvtAJVp0c4nF46L3c5NUpPgoFokxo2BDWrDlaUwis7wuffgqdO5/0+nJ6qcuMiIiIyDlie9p2pm+Yzrp9607fRadNg2efzf5E1TThhx9OvNf9k0+OJkPAWv6dlQWDBlmrTE4DX8hH18+6khHMICOYgS/swxfyMXzxcGb9Oeu0vMfJ9KzXkzh7XK7xW5fCX2/A29Pg/9k77+ioqrcLP3dapiQh1NB7RzoISq+C9CIiiChVsGDDisBn+6mANBFEVJr0IiCCdEGadAgdQm8JkJBkkky93x+HlMnMQEJJgpxnrVlJ7tx750zNnH3ed+9JK+Hid9Bhv43BKwen7GQ0wsaNwlTx66+FmeO5c1IMkUgAtm3jxKBu6J3eC+8u1cX2sNUA1C5YG53Gs2FheTlwKJ7HuBUFmjcHvR6++EKIDP4wGKBzZ8iVC4AlIVc4mk/BlmrWqwI2LUwd1Z28W/ezebKNm1/D9p+g2Wnfp1VQSHAmsLxoom8zZosFjh71FENA/D18uP/xSrIcKYhIJBKJRCKRZBFu1U3/Ff0pNb4Uzy96nmpTqtFsRjNibbH3f/KxY0X1QlpUFWbO9H0dwIIFKWJIavR6UQHxANhwZgOKonhttzqsTN8//YHcxt34pOEnVMhbAZ2SMiErHgXf/wkmJwQ5xMXshF+Xwe69f3imzygK1K8vzBu7dhVVIhLJY4rdZWfDmQ2sD1+P+9NhFI5IxObLnEGFUkevwtSpBBoCvTx63m4FVwLBoUneXXxmHT4M58/D4MHw009QpIjvgWg08OOPyX8G6I20fcXIzKoQrxNiy+pSUPlNLcXPRBHc/SVqX1LJmQh1L8HyudD6RNINi4qWIH1Qctv9tSD4pClY9eBM+ggLDISmTUVrjy/Onk3HIyjJKqSHiCTboqoquy7vYufFnRQOLkybsm0waA1ZPSyJRCKRSLzZswfmzxeT5O7doXr1ux5yJfYKbee2Ze+VvQAkukTJ9tYLWxn4x0DmdJlzf2OKjPR/ndMpet71Xp3wEBzs+xi3G3bsgEmTRLtN376i/SZJ2IiMhFOnoFSpu/poON1+Jg6IiRXAxZiL/LrvVy7HXqZ5yeZ0KN9BrCYntXv7EFQyQqAhkN39d/Plli/5/O/PcapOuh0GrY/VXxXodBRKTSjFjr47KJO7DABx9jhuJtykYCzolq0Q1Tjt20Px4vc1Nokkq1FVlTh7HCa9yauKA+DCrQuM2jaKzec2E2gIZP/V/Wg1wqnj+K5Y8luh/TFYUQ4SU33MmB3w0d9uWP0Ji0eU4/UdKofzwZYi4NDB1SD4XwMY+xfoAeX2hcuXhfD4778i3nrdOuHZkdb+Qa8X+zRtCkDzf2/wpprAgA4wIFXXik6j0PXn7V4VHWYnfPcX2ILNGOs1onmpFqDAsPXDkvcZ+zRsKg4v74cnc1ai7pBRIno3f364cSN5vzMhMLsKxBUPpN35f6hXpJ5PIViStWRbDxFFUc4CsYALcN6p/0d6iPz3iE6MpvbU2pyOOo2KikFjIH9WuNIAACAASURBVKcpJ//0+YfSuWQ5qkQikUiyEcOGiWqMpFQDoxHefRc++8zvIVa7lfLfl+di7EWf1wdoA4j+MBqjznjv4xo+XHiIuH3M8HU6MXGoXFlEWNavn3Ldxo3Qtq136bfBII5Jqh6xWETyw//+BwMHij75gADxOLzwgljF1flee4u1xZJ/TH7iHZ63YdFbmNtlLkadkY7zO+Jyu7C5bAQaAmmmlGLx3/nRrl0n/ASeew4mTEgujb9nbtzg47+HM/boL3y8wclHG52ksT0gUQsfN4NxTys0Kt6IVT1XMXjlYOYcmoPWrWKMt/Pdej29w25bN37xhXgNSCSPGInORNaeXss3c18nOPwiZ/PqadqsH2NajiFAJ6qgwqPCqTm1Jla7FYfbu9Js7QxofgYSdDC4DcytDIoKORNEBVbnY4BOxw0TmBKcmJ0Qp4OC70GsEfZOgepXfQzOaISDB6FMGejUSZg+pyU4GGbMgI4d4ehRHNWrsKKEkxe7CLFTRaTZKBoN8Z+58SVPqIBiscCqVdCgAeeiz1F+UnkvnxGL3sKfPf+kYbGGYkP//jBtGgCzK8OA9qKKxKlTMOvNdKrQiZkdZ0pRJJP4r3iINFFVtVp67ojkv4NbdVNlchVORZ1CvV2vZnfbuWa9RreF3bJ4dBKJRCKRpOLIEfjuOyEeuN3iEh8Po0cLcz0/zA2bS1Sif2NAVVW9xIIM8/bbYsXSF06n8AX591945hnYuTPluiZN4MMPxeQjMFBUg1gsQoRI3UpjtQqj1bffhrlzhRBy65aoPJk3TwhFfggKCOKX9r9g0pkwaET1p0VvoV3ZdrQq3YoeS3oQ74jH5rKJA2LjmPrFAZQ1a0QVht0OCxdCo0a+BZ/EROHtUaECVKwong+bzXOfyEhR4VKoEF+9+CvHZ+ekcvMeYPCumlEVsdKtorLl3Bb6Le/HvLB52Fw24lU7N00wuKWDNYUSxW1/+qnwE5BIHhHORp+l6YymhHxuwdqlPWu/PM+cBW72TLDx7NuTGbjwpeR9P93wKTG2GJ9iCMDwJhCvF61nvy6Dm9/A0QmweB60TOXRERLvTjZQPZUHtCqggsWXYSmI91b58kIQKVtWfC6lxW4XnwsAv/yC4nDS+RhcGgM/rISJq+DQJOgUHsC1QjmYWRV+qgEXUxXGKSA+xzdtAqBYSDE+rP8hZr0ZjSKmzxa9hc4VOtOgaANx0N69QhQGbgUIMSRBL6peVFSsDitLjy5l9anV/p8ESZaQ3QURyWPI2tNruRjje8UsLCKMK7FXMnlEEolEIpH4Yfly314cTiesWOH3sF2XdiVHUfqiaEhRchpz3t/YcuYUfffvvw8FCoi/tT7CJ+PjvcWLTz8VgoLdLlpT7HbPiNkkNBphKJr2usRE+OYbqFIFtm3zObznn3ieI68dYUTjEQx9eiireq5iTpc5hEWEYXN6ihc9DoLZDhp3qtINu1305m/c6Hlit1vEC3/2mRCljh4V1TKtW6eU16uqEEM2bhRCSUICRY9doeMXi9DlyYcKuBRRpmzVwzf14FRucaiCwqIji0hwet7neAN8eXtuhMMhBBuJJBtjd9n5astXFB1blFLjS7Hp7CY+2eim/XEhZoTYxM/G4W7qjFtExMn9qC4Xq06twq36chYVbC8KHbsrhOUT76N4LeRJhGrXINBx257D6UTrEuewaaHZSxBtAhRYWFFUZfnE7RateRMnChPjJFFEUYTZ6tdfi886gKioZAPUnInQ6yC8sh/yx8ENnYOS/a289iy81QrKvAFjnkp1O0Yj5M6d/OeIRiNY/9J6BtYcSJ/qfVjy/BJmdJyRUu0xfnxyleC6kiLSNy1Wh5W5YXPv+JxIMp/s7CGiAmsURVGBH1VVnZrVA5JkDv9e+je5MiQtKqpfNVoikUgkkkxHrxciQ1ozPY3Gtz/HbSrkrUA+l4nOexKoewmO5IFfasB1i2iX+andTw+mrDokRAgT33wDFy6I1dW0rTAgytBTExYGH30kRAe7v+VaxH2/U/LMoUNCeNi7V6zopqF4SHE+bvCxxzaD1uA12aocISZSXrhcQvRo1ixl29q14v6kFmmSqmE2bxarx3v3wsmT3s+b3Q6XL7OsHJwNAadGlPvvLSiu1rugeZlmbLm4NaV6JRXnQm7/oqr+DRYfILG2WNaFr0OjaGhesjkWg48Vc4nED53nd2bDmQ0e4t6g3d6xt2YnvLzXzZEWdRhWR0t0xbtH1G4qo6NaaTeqy8WCBdDhOMmtaGk/2f4sk2KiCvBtfeh6BArF+nnfgxAfihaFoUNh0SIhggwcCHXqpOzTvj2uuXPQxnuKl1YDrC/uwoUKqbyQP20CLcKhyjXEZ9vzz3scV7dwXeoWrut7POfPJ1er+UyhQYipeq3//wuSrCE7V4jUU1W1BtAaeE1RlIapr1QUZYCiKLsVRdkdeSfTMMkjR5EcRdBrfH9Y5DblpkiwH1dpiUQikUgym+ee823uqSjCANAPL+d7hv3jExm1FnofgBF/w6kJUC8qiL0D99K4eOMHP9a8ef1flzYudtq0OwshSSgK1Khx531sNuFTkk4q5q1I/sD8KKmmTfvzQ5yvrwZarWiJSc327b5FGptNXAdCHPJVLeN0EqdX6dkF3m4NQ59JEUNQIU+Cws+WHh5jS0LjhqfP3/7DYLjj8++Fqoq2pZUrPUwZ78SiI4vIPyY/Ly97mV5LexE6OpSVJ1am/zYljzX7r+5n49mNXpVOlQfDey0gNk2OQYAT9ofY+aV8gt+Fy9QUMOdjxDoX0V8Lz5C0vjyQHOTCTRO4Us1KY4xQbRC82Rp21iks3k9eB6tCDO3ZE5YuFZVqqcUQgDZt0NZvgMMkVA83ouJreNdcGPXe0b12nTBBRaMRIkuqCpG78swzor0QIaq4ffxbMOlN9K7aO/3nlGQK2VYQUVX18u2fEcBS4Mk0109VVbWWqqq18t7pH7wkW7Pi+Aoq/VAJ05cmKk6qyLJjy3iu4nMEGgJ97r/wuYXSiEgikUgk2YeiReGHH0R5tcUiLkYjTJkChQsDwqTwvTXvkfPrnAR8EUCr2a3QfPgh+eM1yaufZicE2WDjllJUzFvxDjd4HxiN8Prroqw8NWaztwFsVJSovvCFySSMC3PnhtWrxf23WMQkwhdOp6g4SSeKorCs+zJyq0aCEkWrzNIKMOlJ+Ko+HAi9vaPBIBJtGjf2PEHBgr69BYxGKFRI/F6jhm/Bx2Ti36p5fK/wKlAmSqGAI4Avmn6RHMMJQgyxOGDEPxrx+LzzjmgXSg/h4cIToXlzkZ5RuLAwqr0DF2Mu8tLSl4h3xBNjiyHWHovVYaXbom5cj78OwKWYS8wPm8/68PVe0aZ3RVUhOtpnlcs/5//h6Z+fJvCrQMpNLMdvB3/L2Lkl90yiM5Flx5Yx68Asv+3ltxJv8cafb5Dn2zzk+TYPb/z5BrcSb3ntt+fyHp/C3tUg+L4O1O8j2l2SOBMCg9p5brsTI/9y8M4OCLJ7V4Qk4VZEa0yDSLOXgJCoh4V1Arnywzeo/j5bwsPh//7PO2kmCa0WVq5E/8t03O3bcavzs8QvXcBTQ0b7Hg+ifQejES76fnz98uqrkCcPGAyYHbBogfjsClT1GHVGjDojbz75ZooBqyTbkC1TZhRFsQAaVVVjb/++FvhMVVWfLjQyZebRZMnRJfRa2svDNM6sNzOj4wwq5q1I90XdOXb9GG7VTWhgKIueW8RTRZ66wxklEolEIskirl2DP/4Qv7dr5xE7++xvz7Lx7MbkhAIFhaivVXL4qjrX6cRE1NeE/kHgdsPnnwsjWKtVTL6/+w46d/bcb/Fi6N3b00QVhAjxyy9CdGjQICVF5tgxnJ9+gmbREq/VNtVgQHntNXE7GcA+djR//voxq4s6mF5dpFQ4NGBwQd+DGibk643y3VjIkcPzwFu3oFgx8TM1uXKJsvakx3bAAPjtt5QWIr0e8uZlz0+f0XhrP+LSLkqr0Om4wpIuC6BDB5acWsFXW77iUuwl6uWsyueXy1HBnVukX1SunL47qarC+PXkSU9zWItFrHq3aJGy7eZNGDcOVqxgTFUrn5Q8i0317Ccw682MaTmG01Gn+X7n98nl+UEBQWx4aQPl8pS7+5hmzhS+M1FRIjVoyBAx6dRo2H5hO81nNff67vZN8294/cnX03efJffErku7aDm7JS63CxUVp8vJB/U/YGTjkcn7uNwuqk6pysmbJ5Pjqw1aA2Vzl2X/wJRYXID14evpNL8TsfZYn7cXaIO5i6HlKfH3gHYw4+5p4uI2ncJI1XKHLvdELSwrDx+20vJhYCvCSgXx64UVyd5KZr2ZKqFVyG/Jz1OTlvHaTtX3+cxmYZo8aFD6BoeIPC85oaR3aowdVs2GBueBLl1ElUhGuH4dRo0S3lK5chH95gB+L+smzmGlVelWIinz1i3hqxQUBE8/7V9Iltw36U2Zya6CSElEVQgIn5M5qqp+6W9/KYg8YmzbBm++yXj9XhZUVNlWBA/puERICcKHhAMkG6gWCCqQBQOVSCQSieT+OHb9GDV+rOFVln51FIT68lTV6yE2VkxEHyZut2ghURTxpX/tWihSBPr1g+LFRWVAy5bCdyN1zO7QoTBihM9TDlk1hFqffE/nMHfyxMUN2M0GjMdOifNnhGvXiClfggKDE4hPI05Y9GZWvPAHTUo08X3svn3QrRtcuiT+LlZMmJw+8YTnY/DTTyK6NyYG2reH4cNR8+WjzLiShEefRU01VzHb4Y9FeppcMQph6M8/4cknuS8OHhSTorTCE4jo4yRj3uhoqFpVCG82GyMbweeNwJ1mLmXQGOhRuQcLjyz0Mu0NCQhhY++NVCtQzf94li8XkcmpfWbMZpEk9MUXNJ3RlI1nN3odFmIMIXJoJDpNdrYnfHRxup0UGF2A6wnXPbZb9Bb+6PFHcovdiuMr6LGkB3F2z5axQEMgczrPoV3ZtjB9Onz1Fe6Ia5Qd5CTclOi7BUaFEacKMXLeVXC5MA4DWzqf3gIxcHqCMGRNixuRvrK3ALTtIdpjWpZqyeqeq1l+fDk/7vmReEc8PSr3YNTWUZy9dRany8kbO2Hcaj/tDcWKCYPlDDBh5wQ+XDMUh8OOSxFj7XEIpq64PS0xGiEiQggXD4rJk0Uct14Pqoo7KAj36j/RVa764G5DkswjHburqmq4qqpVb18q3UkMkTxi/Pwz1KsHe/bwxg6Vv2bBgoWgpFoUORt9liShrkBQASmGSCQSieThEREBf/0lzD8fAkcij/g00ZtWHWz6NF/D9Hpo0+bhiyEgViVdLjGhf/VVURUwahRUqgRr1ojKj7/+Eq0/bdpA9+5icu5HDFFVlZ/3/cwrbd182QCuBIpe/dWloeFAQ8bFEIDQUNZOfk9EcaYh3pHA7IOz/R9bvTqcOCFadQ4fFkkzqcWQpMdg4EBx/YULMGkShIaiKAqre6+lWM7iBGqMBDt1GJ0wYhM0OeEQgtWNG9CqlXeUb0aJjvbtZQI4rkfw+p+vEzo6lEJjizCs7AUSbhu5PnsKjD4mm1qNlmPXj/lMMIq2RVP357p8ufkOX6uHD/c23Y2PFwkaDgcHrx30eViiM5Eb8enzPpFknM2nN2CP9Y7ptjqsDFwxkASHEFz3X92P1e793FvtVvHcff01vPEGnDqFJiaWLZMSqBAJqPDUeei1H6reDnMMNARScuhX8N57YLGQ+w4p4GmFsJgcAagB3r4fbuBgKNTpBw37CDHEpDPRpHgTFEWhQ/kO/NnzTza9vImiOYpyzXoNp9sJCvxU07cvByA+xzPIm3XeZFf/3QzdpeetHfDX7FRiCIj35dKldzpFxti1SzyWCQlCgI2NhcuXiaxXjYG/90t+DiWZT7YURCT/QVQVfv8d+vdP3qRBOEe3PimcpJMoEFRA+oRIJBKJ5OGyaxeULAmhoWLCX7s21KqV8sX6xAno0EG0YxQtKto93H6iA+5AudzlxBf6NHzb1MCFykXF6rvFAoGBIoXlp5/u956lnwkTRHxlUnWC3S4mvy++KMQSvV78/scfMHcuNPFTjXGbBGcCLi38ryEUfA8CP4E2L8K+nHdPpPBCVcWlStV7X6FVFPEclyiR4UNL5ypN+JBw1ry8gTnxrbjwHbyfNj3Y6RTi0f1Qq5ZPrxbVZGJU/tNM3TOVCGsElzVxjKmj8kwvYUT55CXoHiZK/JOw6C30r9Gf4zeO+705m8vGF1u+IDwq3PcO58/73u5wwK1bFA8p7vtql4PO8zvz9T9fJ/tV2Jw2lh5dypTdUzgccdjvmLI7J2+cZPbB2Ww4s+GOUbPp5q+/oEYNzhe00K1/CMGfmwkdHcrwjcOT21zSkrDyd7+ePqeiTtFlQRcASuUqhUVvJm+caHlJwmKwUMZUGL780qMaqUAc7JmicmAKrJkNk/6Erb/AmpkQgpGuFbsKP5uJExl6Oj/mNC0rBq2Bt+u+zVOFn8KsNxNkCMKsN/NU8Qbovvifl1+RI0DLkE4BHL7tA6TT6AgOCGZAzQFe9+vkjZMej0eiHs7n8NpNcDdjZz9UCo/lq3Uq362B+ufTeJ04HKJN7UExZUpyLG8SGsBigzMrZtJ9cfcHd1uSDCEFEclD51ZCNLualifhuU4+TY8CHdB7v/jdrDczstHIzB2gRCKRSB4vtm6F+vXhzBnxt8slVvr37xcxixcuiMqJFSvESt6FC/Dpp9CxI/TpA6+8AuvX+zfyS0WlfJWoU6gOAdo0VR8mI+ZNW2HLFvj+e5EucuiQMOW7X2JjRfqCrzaM1Myf7xlNm0RCgqiayACKoviNo2xYNAMmgtevi+cgIAD0elp8MAWXy9s4wKw382KVFzM0xoyiKApPFXmKwmEXyONjdVxV3eKxvh/MZiFMmc0pXgJmMzEFcjGhaiIOd8p9T7zdZrD9drHNtOWw+PcAehVtx8tVX2b5C8vpXKGzV7uE97hVlh9f7vtKf0awgYGQMyefNfnMw0w2CZfqYtvFbXz292dU/7E62y9sp8jYIvT+vTeLpwwhvmZlbAE61FKlUKdP59C1Q+y9sjfjZq8ZJT4evvpKVAfVqCEmpUnCgs0mvFtiYjwOsdqtfLz+Y4qMLYLlSwsVJlXg1T9epeO8jpQcX5Jz0ee8b2fvXiGgFism0ka2pVXPbrNyJXTqxM1j+6j1YjyLC94i1p1AhDWC0dtGJwsbydy8Cd270/DDyTj8rBW6VTebzm7i+PXjdLkRyoGxiZwfC9e/hRW/Qd54hUBDIB0sNXwmYhld8MQ1CLQLA1SLAxpeUNh/rZN4rhUFXnmFIUsvM6TJR5h0JoIMQRh1RgbUHMCoFqPY/MpmtvbZyk/tfmJnv52sfWkthrfegalToVw58fqpVw/9+k106zOGcrnLUSCwAK9Ue4W9A/eSy5TLa1xP5HvCq7rujWchXp+SToOiiPdOBlKsPJg40b95tKJA06b3dl5fXL/uU1RXFTDFO1hzeg2X9/4tWtaKFRPJX59+KqrIJA+VbOkhklGkh0j2xeV20e+9skz8Ptx/jjjwRxno0y8PI5v8H4NqDZIVIhKJRCLJOKoKmzYJ/4gSJYQHg14vvlCuWiWub91aXHbu9H2OgABhJvrrr2KF0B8WC7z0kkhY8UdiIkybhnX+LN6pdImZBSOwqy6eKvIUk9tMpnJoOo0304vLJTw+Jk8W99vpFOXx//ufb+O+p56CHTu8t5vNcOCAdxTvXdh/dT8Nfm2AzWnD4XZg0Bow6oxs7bOVJ/I9cfcTuFyiZSc8POWx12pZUjuQF9vahZGk24leo6dvjb5MaDXhoX9fiEqIYlDfUKYtcnh9j3EYdOjPXYD8+e//hnbtEi07V65A+/a8V/AwYw5O9trN6IBRa+H1fxHPaeXKQsi7zYiNI/hs82dex6XGpDMxuuVoBtce7H3ltm3CyDWth8h334n2ImBu2Fze/etdIqwRuFTvyaRRZ8SsM3Mz8SZ1L8C6GWBJVSQVb1AY2VzHlPpGAnQBzOsyj2Ylm9358bkXnE4Rw3r0aIrwZzZD69Y46z6JMvL/UACNyy0qoSZNwq3XUXdaXQ5FHPIy3ATQKBpqhlbn36KfiwjnwoWF506nTuI2kuZVZrMwJm7VyvMEFSvC0aN8Ww9GNoKENF0lJp2JzS9vJjI+EoveTL0Ob6A9egwcDmZUhT4dvH1jAIJdemY0/56Oz77jIYTatHCuUCCG/Qcprs0tKuIS01mxlScPl8MPsPncZnKZctG0RFN0Gh1Wu5Xzt85zNe4qn2/+nH1X91EoqBADaw2kU/lOFA4unL7z3wVVVan9U23CIsKw3W4V02l0tI/IxYKjT6A9eQqqVROGv9Xu4ItzJ5o1gw0bfF/XooXfCjBVVdl0dhOzD81GVVV6VO5BsxLN7vx5NH26SPlKI1Qn6KDwOzDgsJEvV9nRpBZNtFrxv2z//odntP0f5pE2Vc0oUhDJvqw8sZKT/Trx5haH33IkFbA3aYhh/SYphEgkEonk3rBaRWxpWJho/QgIEO0uH38sTOx0OjFZcbnEZNtHnCggJjKlSwuzy7thNsM//wi/irQ4HMIz6/Dh5MmlajGj9u2LZvyE+7ijd+Dzz4VHQNrJ7MiRQihJy8yZMHgwWK2czglnQ6ByBOQrVjHDFSJJnL91nvE7x7Pvyj5qFqzJkDpD0j9BWr1aGKGmrbqwWLg69nMWVNFitVtpU7YNVULTGWl7n2w8s5Euczsy7+cYnr4gqlpdiGqNme2KMmixj2qBB8DEnRP5cN2HxDs9S1OCEmHOEmh7AjG5XrcOChTwOO6DdR94mfimxqQzET4knPyBfoScrVvhgw9ExVKhQuL1062bxy6qqrI+fD1dFnYhxhbjdQoNGty42TAdmpz1vonoAMjzPri0otrn5BsnKRhU0O+Y74nFi+HllyHOs2LGpddhV50ehp9ukxFNn76sebs9XRZ0wXwzjve2QYvTcDEYRj8Nf5cQgtSmGQq1o8xorFbx/rLZfFcZlC4tKlBSc1uo7NwNlvpI1zZqjbhUFya9CdXpxBSTwKrZKjVu+3oMaAu/VBePW2pMDtgX3Z1yUxd7C7kWixCKa9VK8QvyVRmWBptJT8gnWvQaUaVh1ptZ99I6nsj3BLsu7aLxjMYeaUMAOkVHlfxVWPTcIkrkzHirWlpibDF8tO4jfjv0G063kw7lOzCm5Rj/r92MMmECfPSRt2+OXi8MjHPm9HnYW6vfYtreacQ74lFRsegt9KrSi8ltvUXMZOx2qF8f28F9BNicyeayIxrDkgpweBKYfBWrmEwiRWewDwFTckekICLJFny5+UsSRw5j2N8QkKZKTOV2r16JEuKLl8mUBSOUSCQSyX+CDz8Uxo+pVz+1WiGCZMT7I08e4Skye7b/UurU5//sMyG6pGX+fOjb17ttxWiEI0fuydfiruTKJaJS0xIaClevem93u4kd8DKdnLPZVljF4BIpEv0q9mJ89+lolEzurB4/XkzEbTZuBcCiihBhgYbn4Omub4mI3UzmaORRav1UC1tiPF2OQrfDEGOAn2tC4dbPM6/rvIdyu1EJUZScUJLoxJRyea0LCsfCqQmgCwwWJfh6z5aCSGskJcaX8GmqatabcatuprWfRs/KPe97jMeuH6Pm1Jpek2KAGlcgNBZmLIW8Pube8TooNQSuBt2Ohc1Vllh7LKVzlWZYw2HJqSn3xRtviHa0NLjx7Rmgmkx8/edHfL9yJHt/cBOSKL67Jk1c32wNoXEwbDOY/eipHiiKEEtSP0dFi8KFCwxvDN/WA5u337IXuePh8hgRN30pCCq+BjEBJBteGB3Q+KKOVSdqi6qVtAQHi1CDrl2FWPLBB/Djj0IUzpdPfP9OI9y4NRqWVFR4rqvnZ2CR4CKce+scLWe3ZF34Ot93G4XCwYU5M+SMR8xvtsRqFULRuXOeVUSffw7vvOPzkLCIMJ786Ukv0dGsN/PPK/9QvcAdsoltNm78NIG9Ez8iwuBiSm34pxh8uFPP53850bn9zMvbt4dly+7lHj7WPNIpM5L/DiVzluT3mmZcPl5pbq1GGLWFh0sxRCKRSCT3x8yZ3qXgLheq203fdpDzAyj0DvxfI3D6K0bU6WDaNFFNkZ6kF73efxnzqlW+PTx0OuEb8qBRVf+95jf8pH9oNPRrZeOfUnoS9HDLCIk6+OXMYqbsnvLgx3g3KlYEvZ6dhaDo2zCkNXzaBJ7pBe3zb/RpTvuwqZC3AhXzVOC1PRpGrRFG8KVvgsYQwNt1335ot5vTlJMtr2yhRr6q6F2gdwlh6J/pGnRGM/zyi5cYApDXkpcVL6wgtyk3QYYgggxB5Dbl5v2n32di64mcf+v8AxFDAMrnKU/Z3GXRKp6TXrMdvl4rUgT9pYK4Fbhx+6uf3WUnLDKMc7fOsf7MetrMacPCwwv93/C5c/Dcc6ICrEAB0TKRVBVx+rTwkxg3TvhWGI3pvj+q20UxXR4+3qZNFkPgtvGlA8auhpf3pVMMAWEGrEuTUztyJJjNDNwDBjekTrv1J0A6NLCupPi9UCxs+xkanwWtW5jq9t8DSxZrhemxr+/TdnuK6aheL9qfoqNFBcT586KSJkeOlMfKbOZWkI63WngLwlGJUey+vJt9V/b5vdsqKtGJ0X4Fk2yFxSLa1b78UlT0dewo5iZ+xBCAP0/+6fOzyOa0sfLkyjvfXkAAuV8fSuF/DvHbh63ZV9pCoaBCtCzZ0r94pNEIIU3y0JBh4ZKHSqcKnXgn/zsMaB/P1OXiQx1AoygYli5D27pN1g5QIpFIJP8N/Ph9ODSwtDxEm8Tlm3qwLz/8Pv/2DgaDWCmtUEGIKkkTh5UrYcAAYbyaFFGbtmJEUbxaCZLJn19MPtKOS1EejHFqWhRFCAq+Wl38GGVa7VaWHVuGze2ZbBHviGfcjnG+PSYeJs2a4S5RnM4twohJ8LazxAAAIABJREFUNY+1amGj8yQz9s+gb42+mTsmYNPJ+mjW7cN0+2FqcAE2TFfRDTLBg7FL8MkT+Z5gz6D9RN24iG7uAoL2boAXisJrrwmvFT80KdGEq+9dZdelXSiKQu2CtR/aSv2KF1bw7G/PEn79JLr4ROxa+GwjtLgdYqNzCY+E1O0pcXoYVxccfmYh8Y54hqweQteKXb1bqW/cECv6N2+Kyq+YGPjmG9HiVqeOEBxSV4SlqYR3Aw6tEJjSyg+OQDOdar/ExZNveFU1A6CAIb2F9WYzDBnibWLapw8kJlJo+HD+nnWL/u1U9oeqaLU6CgQW4Nwt7xYstwK3kvRZjYZKkW42zki1g9EofJGGDhVCWeqWQJNJeAVNmCAEml69RJqVXg8hIWKfypVFqta0aaLlsG5dXtIv4lKEt3CrUTTE2eMoFlKMGwn+Y5bdqpvLsZfT+WBlMYGB8Pbb4pIOLHoLOo3Ow/AYQK/RE2gITNc5KuStwJ89/0zZcOYMTKvgu5XTYBDveclDQ1aISB4Iqqqy7cI2lh1bRoQ1JQvcqDOyve92LrVvTOH3tfTrrOGbVytxI/wwAa3bZuGIJRKJRPKfIckXJA1u4EAoRKUq4kgwwJrScCwPYoLwyy/CsyIszDO6sXFjMUm4cUNcv3ChmOQEB4vjzGaYNcvDv8GDfv18ruBjMgmzvofB+PGeK8RJCQzjxvnc3VdbRRJRiT5abx42Gg2HFv9ATKD3TNnqiGfavmmZP6YdO7CMm4TJLmbIJ3LD4gpwJMQhyupTs3+/MNHMlUsIbLNmPZAh5MxdmKDX3xEr1z/8cEcxJAmdRsdTRZ6ibuG6D7VtoXBwYQ4OOsgOR2+Wzoero+HdVF0bOq2WrXUKciGHqMy6FSBEyeF3TnDmZsJN3xPuH38UniCpRY+EBCFgjhghfrfZUi4gxEmLBUwmbhbLR6/n9cTrwZVKq7DqwT7mW0wBFgqV9V1hb1K15Ogz2HcVRt684r2WVJXSr58Yjy8GD4aICKofvsHu7+3EfBxL7EexjGoxCoveu+LMYdDS+GawOH/nzqJ1onp18f42mcRtzZkjBI69e4UpdL58oi2vbFlhID1+vPAXqlZNGEanJV8+0fo3Zw68+SZtavfwnSbkdlG3cF1GNhrp8/ok3Krbb+rUo07Xil19blcUhW6V/Ajkd6NECSHsBQR4imhmMyxYAOXL39t5JelCVohI7pvwqHCaz2xOZHwkGkWD3WXn/aff5/+a/B8AxUOKs7H3RmE8pKpYDNIlWSKRSCRpcLlgzx7hy1G9uu9UFH/89ZdPvw+rAXp19t5d5xJCSflzqpgwmP1/sSc4WPzs1EmUmCelDrRoIYQRf5QuDb/9Jkwdk3xM8uQRk1pfQsmDoFkzYZ74f/8nfEoqVxaTspo1fe6e15yX/IH5vValNYqGFiUfkmhzF9SgIDGh9BEd+1B873buFLGsJ06ICoOPPxavCRATlBEjwOnEroVuz8GaUqK6wKlRqRG1gpW2GIIDgkVlTv36KW1SUVHCwPLKFXj//Qc/7mzGEznKwOWAFBHiNjqDkaYvjaC2OoUzl48QjQ1VI0xXAw0WjDojkfGRXufTKBqCDD7eX9u3+05JUVXfq+tarXhOW7QAgwF9wdz8O6UqDXNd5uP1Dp68BOdzaQh/vScv9e4PgPnj4aIlJ7XRpl6PvmEjcn71Hew5LD6rEhPFa9VggI0bxaT24kUoWFAII3dCo0n+bDFrxOdPpwqdmLRrErsv704WKy16C0PrDaXAyDTiSvv24v5qtZ4T6AIFRKUHCJGoe/eU16TTKS6vvSY+z5IqRFKx8cxGvt/1PZHWSPKY8nBdvU68Mx6toiVAG8CUtlMw6U20K9eOH579gaFrh3I9/jpqqt4fs95Mh3IdqJC3wp0fg0eU0MBQ5nSZQ88lPdFpdKiqisvtYkanGfdnDDxkiHhely4VQnz9+uJ163CI6sXDh0V8dNeu0mrgASNNVSX3RVRCFOW/L09EfITHdovewvyu82lTVrbESCQSieQubNokWk8SE8XEJigIfv8dnnwyfcePHStMVe2erR+fN4Avm2qxKZ5iicUGG2dAbU1h4UeQEfElo9jtYvJkMkHVqt4l9FnMmtNr6DS/EzanDZfqwqA1YNFb2DNgzwNJicgobtVN4e8KcyXuisd2s97M2GfGMqDmgAd3YytXitddUlyqViuep23bhKdCuXLJk+9hTeC7p4W5ZhIBboXnqvVkVqdZYpKyZIlXiwaBgRAZmSEvi0eSy5eFCJg2vcRigQsXsFoMfLv1W2Yfmo1W0dKneh/ervs2Cw4v4NWVr3oYs5p0JvpU78P3z3obovLhh+L9nua97pEilRqjEb79Vhis3iYqIYqxO8by+7HfyW3KzVt136JD+Q6ex40dC8OGifPa7VC3rvDayJVL3M7mzSKyulAhUbVxJ1E1AzhcDuaGzWX+4fkEG4IZUHMATUrcpZzGHy+95LtKKakyrqtnpcPobaMZsWlE8nNh0pnIbcpNncJ1KBRUiAE1B1Apn2d1klt1cy76HD/t+YlFRxdh0psYVGsQ/Wv0zx6Gqkki0EN4/8XaYlkbvhZVVWlRqoUQRh80ly4JoTY6WghbgYHis+nff4X4tWOH+L1oUWEGbjDc/ZyPETJlRvLQuRRziapTqvrtIXym1DOsfnF1Jo9KIpFIJI8UkZFiZTWtAWlwsFhtvVMVRhJ//SW+3KeJ17yU30yFQW5i1ZQVZb0LqkRq2LU4F8r6DaKK4jEnLCKM77Z/x/Hrx6lftD5v1X2LAkF+WoEyge0XtvPM7GdwqS4SHAmY9WbqF63PihdWoNc+oOoaVRWvu3NpPBsUBZ55RqzUvvtu8gQ/31CI9FHgatAasH5sRVeilDCoTEtgoGhjKFPmwYw7O7N8OfToIYSlJBYvFnHYflBVlW+2fsMXm79AURScbic9K/fkhzY/YND6mNydPy9ahlK/1w0G0aJ04oS3IGM0iu1FinhsTnQmMnnXZGYdnIVeq2dgzYH0rtrbcxJvtYpV+dBQKFYsI49E9qBfPyF8pJ3rBQcLoaR9++RNUTv/puCq5iQqnlU2Zr2Z0S1GM6j2oMwY8YMjLk6IYHPnCkHkiSdg6tT0i+zZhU6dYMUKT6FPq4W2bUUr586d4v4ZDOKzZuvWh5Ng9ogiBRHJQ6fX0l7MOTQHt+o7zvDJgk+ys//OTB6VRCKRSB4pxo8Xq75py+ADA0VkZu/edz+H2y2MQ0+eTFk51ukgf352/z2Xvmte42jkURSgrakqP5V9j1ytOz+81hXJfROVEMX8w/OJsEbQqFgjGhZr6G2weT9ERwvfBF9mvMHBIoljyJBkoS7wY9GClRaNosH6sRVji9ai0iktRqNotQp+CKvH2ZH4ePj7bzFpa9QofWlNQIIjgfO3zpM/MD85jDnuvPPu3dC/v/D90WhEhcaUKTBxomh/cjiEsKXVwqhR8PrrHoe73C4a/NqA/Vf3J0enmvVm2pRpw4LnFtzT3c6WbN0KLVt6tv6AEJmvXUtpu/jkE/5cPoYX2to8zIyTaFmyJX/1+uvhj/dB0qyZuP+pW7gCA+HQIShePMuGlSFUNcX0Oy1arbgutQCo0QjD4Z1y7pVEegUR6SEiyTixsbB4MSvPLcaNbzHEqDX6NR2SSCQSiSSZyEjfngA2G1y/nr5zaDQiyvadd4QBncsF7drB+PHUKliQA68eIMYWg0FrwKj7j7cu/EfIacrJq7VefXg3YDYL0cyXIJInj4jfTNVm0eI0LC8H7jTdVWa9WbymPv1UlK6nnnyaTNCz5+MjhoB4XFu3zvBhJr2JcnnKpW/nWrVg3z4hVun1KW0Cw4aJSrGlS8WEsWtXKFnS6/CVJ1dyKOJQshgCItlm5cmV7L+6n2r5q2V4/NmSevVE8sw334jPSK1WiMdLl6aIIWFhMHYsuXLbfEYkKyjkC8yXueNOD1arEMDmzROvucGDxXtNUeDYMeE1k8bPBptNCPBjx2bNmO8Ff+2cbrd3NZTbLYydIyOFya8k3ciUGUnG+Pdf3IULo77+OuaYBL+7FQsp9uiV10kkEokk82nSxLcBoV4vVpjTS86cIj3BahUCy8KFwtzwNsEBwVIMkaRgMAjD27TmhGazMEHNnVuY4ppMEBjIZxsgJBG0aSwqrHarqJRt2lS8/goUEFURJhP07QuTJmXaXcouXIm9woLDC1gXvg6X29vs+IFhsXh7JpQvDx99JJ5DH2IIwKazm4jzYdrrVt1sOecdNftIM3KkEAjGjhVVNFeuiOqJJH7/Hex26lyEvPGgpFnnNOlNvFY7m0W+2mzw9NPCPPrAASF+vPqqiEkHOHXKt5eGwyEEoEcFRYEuXbwrGfV6/6aqiuK7okRyR6QgIkk3B6/sJ6JlfTQxMShWKwN3gynNwoqCQtXQquwduDfdWdwSiUQieYxp2lSsZFpSGTRYLKJHutZdK10lkntn7FjRo280iioOkwneeitlYtWpkzA1/OEHpjQ0onWDK41PpEFrQOH20nq3bmL/ixdFyszEiY+dyeGnGz6lxPgS9Fvej87zO1NkbBGORh7N6mF5UDCoIEattziq1+jJH5g/C0b0kClWTLyme/Tw9mQyGECjQQHWzIKS0RBog+BEMKFnTMsx2S8+d+FCOH3as7LQaoXZs4UYUqmSd3UICKGybja7L35wq25ibDG4x48TniBBQeK5CgqCUqWE2OqrHa1UKf9R8BK/SEFEki5uxN/gtf81wJSQooB8+A88c0qIIkEOLRa9hWr5q7Gh94Y7ZpNLJBKJRJKMogjTuHHjhDDSsCFMnizM8CSSh0lAgKgCuXBB+F5cvQpffumZBJQzJ/TqhbZvf2JDPCfRAdoAelTu4eltoiii5Sad3hn/JVadXMXYHWOxuWzE2mOJtcdyNe4qrX9r/XAik++RXlV6+UxAMWgNtCvXLgtGlA7CwkQbYJ48YsI/Z06GT3Hyxkl6Lu5J8XHFafBLA1adXCXaim6b4Ja+CScnwKbpsGSpnojeYQ+3be1eWbvW24QbRAXI778LAaF9e88qCkURf7+Wzapd0qCqKuN3jifPt3nE5ZfyTPx1EOrcucIfZ/588Vr44guR6pRUXWk2i/SZ337L2jvwiCJNVSXpYsy2MSz+7RNWT7MRnCbp7FgeOFC/NCW+/43aBWs/WNMziUQikWQbTt88zVt/vcX68PWY9Wb61+jPyMYjCdA9fpM/yeNFvCOejtNb0fy37fTe48LgVNlWPS8N524lqNhjkCCTDtrNbccfJ/7w2h5oCGRT703ULFgzC0blm7/P/k33xd2Js8ehqir5A/Oz9PmlVA7NvNQpVVUJiwjjRsINahaoSVBAmuqNyEjhTbNwoag4Sj1nM5th+HD44IN03dbJGyepObUmVoc1OQzBrDczrtU4+u9yi8qoJL8Ktxt+/FHE9mZHPvlEmOX68v/JkUMkEZlMQkCYPFkkzjRrBqNHZ/u0pym7p/Dumnc9Yqj9Ro47HLBsmYgJL1ECXnxRCLiSZGTKjOSBseXcFt5d8y67L+7i8hjIn0aUdRgN6MdPTCkxlUgkEsl/A6dTRGn+/Tdx+XNRM2Ecp3QxyV+oTToTzUo2Y8ULK7J4oBJJJtC8Oe6t/6BJFOX4qk6HEhoqPBp8+eA8ZjT6tRGbz2/22h4cEMzy7stpVDwDnkCZgFt1ExYRhl6jp3ye8pm6oHfh1gWenfMsZ6LOoNPosLvsfNXsK96q+5bYwWoVlSCXL/ue+INoLYyM9O8nkYoei3sw//B8r2TIkIAQIoZGoI+4Lir1FEVUV4SG3u9dfHicOSNiln20xahmM8q332b7ShB/FBxTkCtxV7y2FwoqxMV3LmbBiB5tZMqM5L65lXiL8pPKczXuqtiggUqvwbGJYHGA0QnxBnA9WY0cr7yStYOVSCQSyQNh35V9zDk0B22ijY8+XkXwuasocXEEGHTsxknLXrCjiNg3wZnA+vD1HL9+PP0JERLJo8iePbB9e7IYAqA4nSK+d/ZsYer4mNOtUjd2X9ntsboNIub2yUJPZtGo/KNRNFQJrZIlt91mThuORh7FpaaYzn6y4ROqhlalSYkmoiXm+nX/YgiIio6zZ4U4cBe2nN/iJYYAONwOzt86T6kCpR6dhc0SJWDAANSJE0krYSnx8STs3oGJR1MQSZ5zpcGXSCJ5cEgPEYlPVLebjp+UIirK84150wxF34b3W8D/GmsZ8XY1cmza4e2ALJFIJJJHiqtxV6n8Q2VqTK3B6O2j0YyfiOH4KZQ4kcagtzsJssO8RUCq4lK9Rk9YxCPk3C+R3AsHDnh6iyRhtcLOnZk/nmxI3xp9qZCnAha9MEjWKlpMOhNT2k7BpL97FcPjwuGIw4RHhXuIISDassbvHC/+2LrVt09Gaux2yJ8+E9hCQYV8bne6neQx50nXObIVzz2H3eg994jTwy/sy4IBPRhK5Srlc3vpXKUzeSSPF1IQkfhkQ8eq7A64gc2HzpFogKXNC2Ib/jFffrHd9xcEiUQikTwyqKpKy1ktCYtMETZ6HQSTj/S+PPFQ5kbK307VSZnc2bsvWyK5b0qWTP6+49Sk0gRNpnSt0D8OGHVGtvbZyuQ2k+laoSuDag1iZ7+dvFjlxaweWrbiZsJNn6auANes18Qv5cuL9CN/mEzQvXu6PSM+afCJV+CBUWfkuYrPkcOYI13nyFbUr8/x3Cq2VA+jU4F4PQwveJwER0LWjS2JhAQIDxc/08moFqO8nieTzsSoFqMe9OgkqZCCiMSL40e2UG9VGPF3KPq49M4lPmvyGUbdHT6sJRKJRJK9sNuFC33v3vDxx6IXG9h3dR+nbp7y3NX393UUNeW6AG0AtQrUyrKyc4kk02jUiPU1c1HhdTB8Cjk+gk+agtOgA9k2nEyALoBeVXuxsNtCJj47MVNNSh8VahSogdPtrTabdCY6lOsg/njlFdDdwdmgXTthfJpO2pVrx+gWowkOCMaitxCgDaBzhc5MbTc1o8PPHigK7foHMrsKxOvAoYHVpeHJ/hBrVHC479Bq9LBRVRg2TCQCVakifg4b5mmK64eO5Tuy8LmFVAutRqAhkGqh1VjcbTHty7XPhIE/vkgPEYknR44Q0qUnAU6odwG2FAWPBj2VR1NJlkgkkscdq1XE2p46JX7X62H8eFi0iMulXGgUzzWSqTXg23XCMyoJVVG4ks/IhVw2AjR6uj/RnYmtJ2byHZFIMp9dl3fTvkUk8bfnsbEBMO5phajerfkhb96sHZzkkcJisDCq+SiGrhtKgiMBFRWTzkSBoAIMrj1Y7BQaChMmQL9+IvUlNTqduD6D0c6Dag+ib42+nIs+R15LXkKMIQ/oHmUNTat25FXNbPp18BSXqud7guCA4CwaFSLNZuxYiE/lpTN2LISEwHvvpWzbtw/+/FOY4z7/PBQoAMCzZZ7l2TLP3tcQnG4nuy/vRkGhVsFafiuSJAKZMiNJ4dw5qFIFd0wMGuBQPqjXF2xasOvA4ASNCrN7LKJLxS5ZPVqJRCKR3IEzUWfYfnE7oZZQGhdvjHb0GBg50rt8N3durp4+QPGJpbC5UgwjtS5YtABahEOARo/OYBRl2ps34yhdEq1G6yWiSCT/VdrPbc+KE95pSgatgYj3IuRikSTDbDm3hfE7x3M17irtyrZjUO1BnhP5xYuhTx+IifE+uH17Ebn6GHMt7ho1p9YkOjEaq8OKUWfEoDWw+eXNVM1fNesGljevMMT1tT0iQlSKDB4MM2eKpBy9XrTjzZoFXe5tfuVW3aw8sZLfj/9OrC2WNafX4HQ7sbvsuFU39YvW55vm31CncJ37vHOPFjJ2V5IxEhKgTh04dMhj84VgmFgH9haAyleh1RvjeKbtkCwapEQikUjuhqqqDF45mOkHpqPT6FBQyGnMyYnpgQQcPOJ9QGAgbN7MOxGzmLx7MonORI+rPzC14CtTOzSFCkHbtmAwZNI9kUiyD6GjQomIj/DarlE0HHj1AE/keyILRiX5T3PxIpQpA4men8mYzTBqlJhUP+bE2eOYfXA2Oy7uoHye8vSp3od8lnxZOyiNxnd7jEYDLhesWwcdO3qb5ppM8O23Iv44JAQGDYLGje96cy63i3Zz27Hl/Bbi7HF+9zPpTMzsNJOuFbtm8A49usjYXUnG6NYNwrxTAorEwDdr4bpF4fS4EdSVYohEIpFka+YcmsOsg7M8hA2rw8rhOCM1fB3gdoPZzJiWY6hZoCbfbf+OS7GXqJi3IiMajaBR8UaZNnaJJLvicDuEk2oaH3m36vYZZyqR3DeFC4uWmV9/TZk8BwSI1orevbN2bNmEQEMgr9Z6lVdrZVLstdMpUqXcbqhb13fKZqVKPudUVKwofv72m+8EIZtNtNTYbldq/vGH8B756KM7DmnJ0SVsPrcZq+POqUQJzgQGrxxMp/KdZAtNGqQgIhGmeuvW+TX7cRv0BO8/RN3S5TJ5YBKJRCLJKJN2TfL6YuRW3Uyo6eCXqyY08alaZhQFihaFsmVRFIWeVXrSs0rPTB6xRJL9aX4pgJVBEJ+qQMpshyJWLVpFTi4kD4kJE+Cpp8TPmBjRUvHuu8J3QpK5/PMPdOokBAtFERUfCxZAixae+40bJ1qaUnuImM1iO/hP53S7U8QQEMd/9hn07y+MWf2w4PCCu4ohSVgdVs7fOk+JnCXStf/jgmz+lQiDPX/GTIqCtlVrAqQYIpFIJI8E/kpmF1czENWptYhyDAyEoCDIn1/0ocv4dInkjkxa7mT5XChxEzRuCE6EYX/DH7NVKuSVsbuSh4SiQI8esGMHHDkCn38u2ikkmcutW9C6tfAGiY0V4lR0tGh9iUjTStesGaxdC82bQ8GC4ufatWI7wIsvpl/Q0uth8+Y77hJoCERJW7rmB5fb9cib6T4MpCAigQoVPBXJ1FSuDPPmZe54JBKJRHLPdKvUzWckulFvImTmAuEV9f33YmXr/HkoWzYLRimRPFrExd6gRxeICAS3BlwamFAXDA43mnRORiQSSeZz+uZpftj1A7MOzCLG5sOgNj0sXuy7kt7thrlzvbc//bQQQS5dEj+ffjrluiZNRKyyySQED5MJtFpRceKLuwhgfWv0xaQ33fUuBGgDeLbMs+Q05bzrvo8bsmVGInoUO3eGpUs90weCgkT/munubzKJRCKRZA+G1BnC3LC5nIs+h9VhRa/Ro9fqmdlxpugbLl1aXCQSSboZ3NHAdbMN9+05i9UAiToY0tHAUllhJZFkSz5Y+wET/p2AgoJW0TJo5SCWv7CcpiWaZuxEN2/6XjxOTPSdKHMnFAUmToQBA0TsbmAglCsHHTp4ttmAqCRp5O3jdTXuKvPC5hGdGE3LUi35oN4H/O+f/yUbqTvdTnIE5CAiPoIAbQAqKg2LNmR6x+kZG+tjghREJILp08Uq4aRJohSsQQPR61akSFaPTCKRSCQZICggiD0D9jAvbB6rT62maI6iDKg5gNK5pAgikdwLqqqytqiDtNapLg2sKuYQK8dSFJFIshWbzm5i0q5JXslpHed1JGJohM9KSr80aSISYnxRqdK9DbByZXFJYuxYeOstUTWiqpAjB6xeLapHUrHyxEq6LeyGGzc2p43R20bTtmxbTrx+go1nNxJkCKJ1mdYYdUZuJd7iSOQRCgYVpFhIsXsb52OAjN19XNm+HWbMALsdnn8eWraU/8wlEolEIpFI0qC63RhHarH78E612CHumb+hYcPMH5hEIvFL7997M/PATK/twQHBzOk8hzZl26T/ZP/+C/XqiZSZ1CiKMLkdNeo+R3ub2FjYtk1U6det69VGk+hMJN+ofMTaYz22W/QWZnaaSecKnX2eNsGRwKyDs1h9ajVFgovwaq1XHwvvIxm7K/HP8OEwZoxoj1FV0UfeubMQSKQoIpFIJJmD1SoMTrUyoUIiyc4oW7bQ9ajCwvIqjlTfnA1O6H4I0K6TgohEks2wO+2+r1Bvx2hnhHPnRFJMTBoPElWFEyfubYC+CAqCZ57xe/Xmc5tRfMzVrA4rMw/4FkRibbHUmVaH87fOY3VY0Sk6pu2bxpzOc+hQvsODG/sjjDRVfdwIDxcqZnx8ijmQ1QpLlsDWrVk7NolEInkcWLNGeHiEhEBwMLz9tqjWk0gk2ZPTp5m4Rke5GxBoA6ND/KwUCWPWayB37qweoUQiScMLlV/AovdOc3G4HTQr0SxjJ6tZ0/f/abMZGje+twHeAxrF/9Td33UT/53ImegzydG8TtVJvCOeV5a9gsOVQWHoP4qsEHncmD1bOCKnJT5eRC/Wr5/5Y5JIJJL/MqoqYvO2bxcGbN9+m2Jg7XTCjz9CVJTwcpJIJNmPqlXJYddwYApsKg7H8kDFSGh0FhSDDrp3z+oRSiSSNLQt25Y2Zduw8sRK4h3x6LV6tIqWae2nERQQlLGTlSwJXbqIAIok41O9Xixs9Onz4AfvhwZFG/iM2LXoLbxc7WWfxyw4vMDLRwXA6XZyKOIQNQrUeNDDfOSQgsjjgtMpcq+XLAGHDzVQpxMuxxKJRCJ5cNjt0KqV6D+22YQ4ktaYLSEB5s8XrYxypVkiyXYsMZ8jZ0EHT5+FpmfERQXRZrxwIYSGZvEIJRJJWjSKhnld5rH53GaWn1hOjoAcvFjlRUrmLJmyU0QEXLwogiXuNg+aMQNq1xYBFHFx0L49jBwpzE8ziQBdAIu7LabDvA6oqDhcDvRaPd2f6E67su18HhMcEOxzu0t1EWiQcz+QpqqPD59/Li6+xBAQ0boHD8ooRolEInmQfPcdDBvmGWnui+Bg2LQJqlfPlGFJJJL04VbdPPt2Pn6edoM8CaC/XWS7Lz9M/bQVPw5elbUDlEgkGSchAV56CVasgIAAMT/68EP49NNs56foVt0cv34ck95E8ZDiANxMuMniI4uTY3er5q8KwK3EW3yz9Rv06H2XAAAgAElEQVQWHF6AWW9mUK1B5DLlou/yvsktMyDEoop5K3Jo0KGsuEuZRnpNVaUg8jhw44ZYvfAXFxUQAFOmwMsvZ+qwJBKJ5D+D3Q7R0aLCI7VJ6hNPwOHDdz/eaIRr14QwIpFIsg1Xrp3GVLw0IWkqzq16qPt+Tg59cTNrBiaRSO6d3r1FqERiqje22SxaWF98MevGlYb14et5cemLxNpicatuyuUux5Lnl1AiZwmvfRMcCVT7sRrnos9hc9kAMOvNtC/bngJBBZi8azJ6rR6A3ObcbHhpg8/z/JdIryAiTVUfB0aO9C+GaLVw9KgUQyQSieQeuBR9gWtv9UfNlQuKFYO8eWHy5JQd0rPoYDbDG29IMUQiyYbkWvsPWh/Waxo3vBymz/wBSSSS+8NqFW2qiWlUzvh4+PrrrBmTD85Gn6X9vPZcjbuK1WElwZnAwYiDNJreCJfbe143L2wel2IuJYshAPGOeJYdX0br0q154f/Zu+/oKMougMO/2ZJtKfQmvQpIR0AQRECKIoqAUsQCIgo2FPRTRFERC4oKFkSwoIBIlSZIFUWKYKEjvUMSCCHJJlvn++MFks1uEAWy2XCfc3IgM7OTO5Rk9s59763Vg9Kxpbm18q2svH9lvk+G/BvSQ+RqMHNmzvsqVoQK8h9CCCEuli8hnq3DB5K6ZD6W1AxqJIDmPbszIwMGD4aCBVWjxfvvV0np7EtmrFYwnJ1OMWQIPPZYbl+GEOIiWJJT0XQjEPgGxOaDDtGyxE2IiJOcnPOymPj43I3lAj7b+Blevzdgm1/3czrjNMv3LeeWSrcE7Fu5f2XAspisbp96O37dj8fvYU/SHn7Y8wNr+q6hRtEaVyz+SCIVIleDqKic9334Ye7FIYQQkczjYfOqmZyqWJLK42fQdFcG9Y+BzZvtOKdTJUEAnngC6tfPbNZmt6sGbKtXq6dUBw+q6pA8tmZZCHFWy5aYTcH3US5bFNV7PhGGgIQQl6RECTUdJjuDAW66KffjycHBMwdx+4JH/eq6zrHUY0HbyxcoT5QxxPcqnwuXz4XHr/pIun1uUlwpPLXoqcsfdISShMjV4KGHVNPUrDQNmjSBtm3DE5MQQkSClBQ4fhx9wKMM6GJj3ZNdKeD0Yz+bBMkxjXHkiPrValUjd2fOVEmS999XSZD6MuZOiIhQsyZa9+7gcJzfpNvtWJo2R2vfPoyBCSH+E4NBPRC22zO3nZu2+frrl/3L+fw+5u2cR//5/Xlh2QvsOrnrol7XpkIbHGZH0Hav7uWG0jcEbX+o/kOYDYHL+IyaEb8evOZPR2fVgVUXeQX5nyRErgbPPqsynna7+oiJgfLlYcaMcEcmhBB509Gj0KaNWtJSqhRzV4xjUi0f7fdkTpm4oFq1Mn9vMKjk88svQ79+0itEiEjTvr36f2wwgKahVa2qehAY5DZaiIjUpQv8+CN07AjXXquWt/7xhxq/exl5/V46TO5Az1k9Gb9xPKN+HUWdcXX4dsu3//jae667h7JxZbGarOe3OcwO7ql5D1UKVwk6vkxcGRb2WkiZ2DLYzXasJisNSjbAYrSEPL+M3M0kU2auJhs3qo9y5dSNftZJCEIIIRSfT90UHThwviH17T1gfjXY8hHUTPiH19vtsHgx3HjjlY9VCHFlrV0LrVurpXDnWCzqQdPixeGLSwiR503eNJn+8/sH9fZwmB3ED4nHbrbn8EolxZXC6DWj+XbrtzjMDgZcP4AH6j6AQcs5GavrOnuT9mIz2ygVU4qBCwby+Z+fk+HNbCJrM9l45oZneK3Va5d2gXmcjN0VQggh/i2vF5YuhbvvVstlzmrbG5ZUgr4b4f1FEO3JfInHAE4z+A0ajgY3EPXWKGjaNAzBCyEuu7vugjlzgidG2WywYweULRueuIQQed5tU25j4a6FQdtjLbF81/U72lVud8VjSPek0216N5btW4bFaMHldXHHtXfwdeevz4/hza8uNiEiU2aEEEJc9XZMH0fBIcModjARPSoKgzdwosS9m+DXMjCxPtQ9Dg/9DhkmtXxmSzH4emR3XrzrPQpGlwjTFQghroh9+0KPz46KUr2CJCEihMiB3RS6AkTX9YClMFeSzWxjfs/57E3ay66Tu6hetDpl4+T7Vlay+FEIIcRVS9d13hh7D2V6PUrxA4loOhhcbjx6YEKk52ZodhCi3fD4bVDtMejdzcAnY++nzv50PrxvKiUkGSJE/nPzzaGn9bndULNm7scjhIgY/Rr0C9kY1Wqy0qxss1yNpWLBirSr3C44GaLr8NVXUK2amoLXurVqsXAVkYRIfjNrlirVrlIFnnwSjh8Pd0RCCJFn/XLwF6qNn4nVE7jd7Iesz4RNGPjhWyPT59t44ng5+jUdyIdj9zK4/5e59pRHCBEGgwer6RNZ+67Z7fDcc9IgWQhxQW0rtWXA9QOwmqw4zA5iomIoYCnA/J7zMRnyyEKNUaNgwAD4+284cwaWL4cWLWDTpnBHlmukh0h+8uqr8PbbkHa2cY/ZDIUKwebNULRoeGMTQoi8YtkyGDQItm0jOdaC0+2kZFrwYW6LiajYguqTLl3gtdegSJHcjVUIEX4HD6qx2UuWQLFiMGQI3HMPaDkO3hZCiPP2n97PpL8msenEJioUqECPWj2oX7L+xZ/A6YRJk1Qj57JlVQKjWrVLD8zlUvc1qamB2zUNOnVS/ZMimDRVvdqcPo1esgQnTC6i3aqsG1Cd0AcPhhEjwhqeEELkCatXqxG4WSZGeDQw6JB97pbXYsZ0PB4KFMjdGIUQ4XfsGLz0Ehw+DHfcAf37SwJECPGfjPx5JCNWjcDtc6OjYzFaGNRkEK+3fv2fX5ycDNdfr3oWOZ1gMqmH3t99p8YGX4rdu6FeveCECEDp0nDo0KWdP8wuNiEiS2byicVfD6fcoy7KPwVFnoW7u8IZCyrzt3RpuMMTQoi8YdiwwPGZgFkP/mGYZgbfQ30kGSLE1ejbb9FLlUKfMAF90SL0Rx9VlSHZvncIIcQ/2XNqD6+teo10bzo+3Ydf95PuTee9te+xJX7LP59g9GhVpXbu+4/XC+np8MAD4PNd8KX/qEQJdb5QKlW6tHNHEEmI5AObV8/irhMfcCgOXCb1Mfda6HI36mmGdEAXQghl69aQmzNMsKEUuIxwPBr2PX4vljEf53JwQohw83rdeHv1RIOADz0xER58MLzBCSEiztydc/Hr/qDtHp+HOTsuYknKjBnqAXd2Lhds23ZpwUVHQ58+qi9SVnY7vPzypZ07guSRbi7iUrwzoS8Z2XIeLhOsLgt7SkRR6ZlnwhOYEEKES0oKfP01/P471KoF990HBQtC9eoQHx90uCXKxrpvR/C7PZrO13bmOof0XRLiajR3zEA6+4OXk2sA8+blejxCiMhmNpoxaME1CJrXi3n8BIi/VvUpy2lJXk7Nm30+ldC4VB98AFYrjBunqkWKFFHbbr750s8dIaSHSKTbsoUm79diXZngXXEZMCu2H62eH5/7cQkhRLgcPKjGYZ5bE6tp6qZhwwY1eat9e1Vueo7dDs88oxpTCyGuag/3v4ZnZh7lmXaworzqyfboBhi6CoxRFgzpGeEOUQgRQY6mHKXSmEpkeAO/d9g8sOVjqOh2wBNPwMiRoU8wZQo8/HDm0AxQU6/q1Lm843E9HvU14uLyTb8k6SFylfj1nadofgCiQiz/chmh1v3P5n5QQggRJon7t3G8YXX0rA3CdF1VjPTqpUbJzZqlurNrmnoS8vLL8Mor4QtaCJFnLCzvpslDsLAyOKMgPhrebgb33gUZbVqGOzwhRIQpFVOKT2/7VI3exYzdDVYPjF4EFZNQSYj33gtZvQpAjx5qWYvFAjEx6gFP+fIwe/blDdRsVn3T8kky5N+QCpEIdtJ5kgkditPrdx91BsBpK/jPprjsbnjoeEk+mHg0vEEKIURu0HVczz5NA+f7bP74bHl7KF6verJy9jVX4w9+IUQOliyh8fR2bCipn7+fOsfihe0Pb6JCmVrhiU0IEdFOpJ5g7iM3o+/YTse/oVRKlp2xsTB1Ktx6a84nOHQI1q6FkiWhWbPLdv/i8/tYtm8ZB04foGGphtQrWe+ynDcvuNgKEekhEsFWH1rNgtpWHluTxobx8GIrWFIJCqbD4+tg4LTV4Q5RCCFyxe9fvclTSWPQLf9wYNYbCEmGCCHO0XV4+GGsLYOTIQAmm4Nt6QepgCREhBD/XvHo4vSz3AB/7AR/tiarKSlqKmj79mDIYQFHmTLq4zI6fOYwLb5oQaIzEZ+uJtbcVO4mbqtyG2//+jbxafHULl6b0W1H06xss8v6tfMSWTITwQpaC/JnKQNjG0GxNPhqNhx+F/4YBwXqNEarUCHcIQohxBX37ZZvab73RX4u7edAHHhD5Dl0UE9VcrrREEJc3U6ehKNHqXcMzCGWIft1P1UKV8n9uIQQ+cfjj6sGptnpOnz6KQwcmKvh9JrVi4PJB0lxp+D0OHF6nCzdu5SnFj/FweSDZHgzWH9kPW2/actvR37L1dhyk9wZRrBmZZsRZ43jhVs0mvaFES1geEto8piVasPHhjs8IYS44nYk7qDP931wGv2gQZoFPr4efNmTIpoGkyaFJUYhRB6XnAwrVoCu8+Q6sPgCd1u9Gs3KNKNq4arhiU8IkT/UrQtffqn6gWTndKp9OfUSuRipqWpM79SpKsl7AafST7H28NrzlSHnePwevP7ArLDT42TYimH/Pa48ThIiEcygGVjSewll4sqwt2w077WP5f1WNgbe+z7XX3N9uMMTQogrRtd1eszoQc2Pa5LuTQ/YN6g9vHQTJEeBHzhYyIh/8SJo0yY8wQoh8q7PPlPVYw89BLpOhdOwbBLUOwoGv+od0svRhDnd54Q7UiFEftCtG9SuHXqfxQLbt/+38y5aBCVKqAasDz8MpUur72858Pg8aDl3XAuy6cSm/xZXBJCmqvmAruusP7KeZFcyN5S+gRhLTLhDEkKIy87n97Fw10Lm75rP1q0/sca5M+Ra/3M0HdpE1+bL/j9QKqZU7gUqhIgMmzZBkyaBY7jPiY7GpXsw9XkI4/tjZLmdEOLyeeAB+OYb8GUvR7PCjh1QrhygluqtP7KepPQkbihzAwWsBUKfLzkZSpVSVSZZ2Wzwxx9qsl4INT6qwfbEi0vAtCjXgp8e+Omijs0rpKnqVUTTNBqXbhzuMIQQ4orx+r3cOvlW1hz6la8mpbGzCfgv0CbJ7jcysdFIut8mo8eFEDmYOJF0v4ujBdXEB1vWKnG3G8vs2Ree+iCEEP/FkCEwfXpgAsNqhdatzydDdp3cRdtv2pLoTMSoGXH5XIxsPZJBTQYFn2/u3NBJW49HJV5eey1kGJM6T6LVV63w+D1keDOINkdjNBhxezJI97vOH2cz2Xil5SuXdMl5maS7I9Cx5CPM+2QQq17szcmVP6hGPEIIkY9N2zKNXw/9Spf1aXT6mwtWhpgNZp6+6XnuuXVI7gUohIgoft3P89oyCj/jp84jUORZGHbz2QbMAG63etMihBCXW82aMH8+VKkCJpNaKtOjB3z3HaCq/9t9044Dpw+Q6k4l2ZVMhjeDF5e/yM8Hfg4+n9MZPLkGVAVKWlqOYTQs1ZDdT+xm+E3D6VuvL2ML9eLwiDSeXe4iNkMtGyxvKcHULlNpWb7lZbr4vEeWzESYr5e+S8uugyl5BjRU48BTtSpTYs0mVRYlhBD5UMcpHVmwawFbPoRrT8L4BjC4LTijAo8zakYODjooS2SEEBf05sxBvPbnGJymzDcRdjeMWA6D1p7dYDJBYiLExYUnSCFE/qbrauSu1QpRmTc064+sp/Wk1qS6UwMO19DoVrMb07pOCzzPgQNw7bWQkRG43eGAhQuhRYt/jmXLFqhT53xiRQfcRrDoBjh8WPVaijAXu2RGKkQiyIHTB6jfczClk8Gkg1GHKD8U+2s3Z55+LNzhCSHEFWMzq4TvtqJg0KHPH1D3OESfreg0e8Hkg1daviLJECHEha1fzzu/vR+QDAGVYH3rxiwbDIbQ0yCEEOJy0DSIjQ1IhgCczjiNQQt+m66jk+hMDD5PuXIwdCjY7ZlLZ8xm1Vg1KSm4V0koL78cUGWicXbilt+v9uVjkhCJIAt+/YoaCQT1AzYA1q++CUdIQgiRK/rW6wvAC62h2BAo+D+IdsOIZdB3Izy5FkYU687QFkPDHKkQIq/Tn3yCU9bQ+xLtZ38TFQV33qme3AohRC5qUroJHp8naLvdbKdL9S6hX/Tii2p8eMeOqroNYOdOuPdeuOkmcLlCv+6cPXty3rd580VGHpkkIRJBimzdl+M+kyv4P40QQkQyv+7nw/UfUm1sNXrN6gXA7sKQ6IB0MyypCC+1gud/hqfiK/DcY1PDHLEQIhJoGzZSPSH0vusSDWC346lfl/jRI3I3MCGEAGItsYy6ZRR2s/38aFy72U7lQpV5sO6DOb+wfn1YvRq8XtVQFSA1VU2amTDhwl+0bduc97Vu/S+vILJID5EIsu+n7yl7850Ys/2V6YCndCmiDh0JS1xCCHElPDjnQb7b9h1Oz9ku7DpBJXJmL9RMgP4PfMgjTQbmeoy4XLB+verhVL9+jqM51x5ey9h1YzmeepxO1TrRt35foqOiczlYIQQAxYqxJCaBO7uD04z6vqKrKTOTyg1i9pGlzDDtREOjauGqTL5rMrWK1wp31EKIq8yaQ2v46LePSEhLoHP1ztxf5/7zS4hD2rABWrVSfUmya9QI1q3L+bVnzkDx4sF9SKxW1UOkcOH/dhFhdLE9RCQhEkl0ncQyRSh85FTAewId0L7/Hjp1CldkQghxWe0/vZ/qH1Unw6t+MBt86s1KWojl/OWT4LdiL1Lk2Zczy0Rzw8yZ8OCDag2w3w8FC6qu8bVrBxw2fuN4Bi0eRLonHR0dm8lGmbgybOi3gRhLTO7FK4RQ3ngDRozg18JOhrdUvYlqnjQw7JqedCu0lPi0ePx65lr6AtYC7H9yP3FWaa4qhMjDNm2CZs1UVUh2zZvDqlUXfv3Ro9C5s0qs6Lp60DNtGlSqdGXivcKkqWp+pGkUWfsXGVUr4jVqeI0G/GYT2gcfSDJECBHR1h5ey40Tb8Qx0oFlhIU2k9oENBTruhV8IX5imXzQbjcUfnWUWu+fW3btgt691VOYM2fUzcehQ6qs1O0+f1iaO41Biwfh9DjRzw70TPemcyj5EJ9u/DT34hVCZHruOejTh6YJVn78PpbDH1lYbO7DiXs6kuZOC0iGALh9bqZsnhKmYIUQEcHvhzlzoHt39bDk5xDjca+0WrWgSJHg7Q4H9O+f48t0XWdL/BbW64fx/PqLasLq96vESIQmQ/4NSYhEmtKlse3cg2nzVkw/rcJwOhmeeCLcUQkhxH/26cZPafFFC1YfXo3T48Ttc7MnaU/mUhng8d+g1T6wZmuXZPHB4DWguVzoCxbAyJG5E/TEiWqNbnZuN/z44/lPNx7biMkQXLWS7k1n1vZZVzJCIURODAYYO1Y9DV2xQv362WccSDuCyxfceNDpcbL/9P7cj1MIkWd5/V4W7lrIhN8nsOXEZujSRTUwnTYNvvoK2rfP/eksmgbffw+FCkFMjFrOa7Op2Hr0CPmSnYk7qfZhNZpMaEKbSW0o9k4x5u6cm7txh1ku1haLS+L1wgcfwKefQnq6+of90ktqvJIQQkSoVHcqT/3wFB5/6MbQGho6Om4jzPgOBrWDr+qC2wi1T8AnC6DyqXPHAsOHq4q56667soHHx2c2LMvK54OTJ89/WsBaAJ8/9Li7ovaiVyo6IcTFKFhQfZzVsFRDzAYzbp874LDoqGgal26c29EJIfKovUl7afFFC864zuDTfeg+Hx2tXqY6fRhBLTdxOuHtt6FvXyhbNveCq10bjhxRS3jj49WEmZo1Qx7q9Xu5+aubOZ56/HwVK0CPGT3485E/qVK4Sm5FHVZSIRIp7r4b37AXVZn24cP4P/4YGjZU/9mEECJCrT+yHpMx59y80WDEYrTwZSMzfoPGuAWQNhLSX4c/PoUmhwOP93k9JL6fC1UiHTqoEtTsfD518wF4fB62xm8NWPpzjt1s5/HGj1/pKIUQ/0Lzss1pUKoBNlNm00KL0UKFAhXoVE2WJgshlK7fdeVY6jFS3Ck4PU7S/S4WVPQxvkG2Aw2GgKrRXGO1QteuMGBAjskQgGV7l5HqTg1IhgB4/B4++/2zKx1lniEJkUiwdSvuBfMwpmd2/TV4PHiOH4WpMmZSCBG54ixxXKi5d82iNTk++DhvfLoHe/feYLPhB8whii6OR8Ozt0C92KmUeLMIA+c+yumM01cm8M6d1VOYrFV6VqsqSS1fHrfPTcuvWtJvXj9S3Jnd3q1GK1aTleE3DadNxTZXJjYhRIC/T/7Nb0d+w+UNXg6TlaZpLL53Mc/f+DzlC5SnTGwZnmryFKv7rA659E0IcfU5lHyI7Ynbg3oNOaNgXPb2nUajWrqSRyU4E4KSIaASIkfOXD3TSyUhEgF2LZqCWw9eq25Od5GxLAxZRyGEuEzql6xPyeiSIfdZjBZevflVClgLcE2BMmhffQXr1rG3enF0CPgRfsIBdR6BsY2gegKsfOckH94xDgoWJKFhDbWO9/DhkF/nPzGZVO+Bd9+FihXVUyBNU0nqpk35ZvUn/Hn8T9I8aQEv09HZ9fguhjQbcvliCUXXVbNXX+jlOueke9LZlrCNpPSkKxuPELnN52PR+Oco/b8oaoxR6+MLvlWQyZsnX/BlVpOVYTcNY9+T+zg46CBvtnlTpkEJIc5z+VwhKz8B0kPlTTt2vLIBXYIby96I1xf8HtNhdtChSocwRBQekhCJAOOOzcOvBW9PN8H2giGa+gkhRITQNI1F9y6ifFx5zAbz+e0xUTFMvGNicJl6rVoUHj2ODLNGkhXSzr5kVFM4bYUGx2D2NLj2pOopUiADov/ajvf113i7d0XKv1mCwm8VpteoGzh4azNo3Fh1ULfZ1K+TJl188CdOqN5Oe/eqbuzp6ZCRARs2MG3G8ICmsOdYTBa2JWz7D39S/8KUKXDNNVC4sOqPMHy4ii+bt1a/RdFRRWkyoQkl3y1J71m9z485FiKiJSWRUr0iT25+m+NRHnwG8OMn3ZvO/bPvZ+GuheGOUAgRoSoVrERhW+Gg7VbNTI8dJlUREhsLcXGqj0f25bW6DmvXwoIFkJh4RWL8++TfDF85nP8t/R9rD6/N8bjyBcrTr0E/HObMGG0mG1ULV+XumndfkdjyIqn/iwCbTm0j3gF2D5iyPBL1arC3c0vqhS80IYS4ZJUKVWLvk3tZf2Q9ic5EahevTenY0mhaiEwwULjtHbge6geTP6f7rV76/AmzqoPbBC+vBFu2Xqc2L/TqrDOnugen6wQA0/xr+fE62P4RFDmXt9i7Fx59VI3SHTgQgN2ndvPW6rfYeHQj1xW7jueaPUfNYjVh+3aoUyd0Y1WPh9iEMxB8v4Su6wE3HqBGek7dPJXZO2ZTJKoAj5S8nYa12kF09L/5Y1QWLoR+/TL7S6WkwKhRKs7XXz9/2NTNU3n1p1cDkjYzt8/EYrIwodOEzPO5XKrqZe5cKF4cHnlEXbcQedngwWxJP8jRGPAZA3f5dB+3T7mdsR3GMqDRgPDEJ4SIWJqmMfmuyXSY3AGv34vL58JhdlC+QHkGT1kEv/wGFgu0bq1+zWrfPrjlFvVAxWBQk+mGDoUXX7xs8X268VMGLRqEx+/B5/cxdv1YetfuzSe3fRLyvuqD9h/QolwLPtnwCSmuFLpf151HGj5ClDHqssWU12kXWrsdKRo2bKhv2LAh3GFcMUsrajzSET7/HhofAb8Gh2Oh3+0w+8MECkaHmDcthBD53R9/8PBXXfkmei/pZytFDoyGsmcCD9tfAKoPhAxz4HarB4b+DC+uynbeuDg4eZI/EzbT/IvmZHgy8OpejJoRi8nC4nsXc2P3Z2HNmhxDW1bTzh3d/aT5AisuSsWU4tCgQ+fLbV1eF82/aM62hG2kedIw+MHqg9FLTfRvMlAtyTEGvqPbm7SXJXuWEGOJoVO1TkRHZUmcNGwIGzcGB+RwwKlTEKVucGp/UpvN8ZuDDrOarJx89iR2s11VvDRpArt3qwSL0ahu7j75BO67L8dr/ze2J2znvbXvsfPkTm4qdxOPNXqMYo5il+Xc4ioWHc2CUmn07AJnrKEPsZqsbBuwjQoFK+RubEKIfOFoylG++OMLDiQf4KZyN9G1RlcsJkvOL9B1qFED/v47sGrT4YCZM6Fdu0uOKT4tnnLvlwuq9nSYHfzQ6weal2se/CK/H375BZKSoFkzKJJ/3ldqmrZR1/XsnV2CSIVIHpfuSafJYbhnK7TrDQXSweGBozEqQVLQEhfuEIUQIjzq1ePDGpsxPVKeiaUTcBthU3EofSZwPehfxSHKF5wQyTDDL2VCnPfMGUhK4qlFT5HqTj2/2af7cHqcDFwwgL/WbrlgaK33+Hn2+kG8vfZdem41cOdmD06bibovvx6w9vjrTV+zLX4raV5VqeE3gNMAg9p46fHReGJjY+HVV88f//zS53l/3ftoaBgNRvrP78/8HvO5qbyabMO+faED8vnwJMbzY+pfJDgTLtgsLTkjGfvxk9CqlUqGZDkHTqfqWt+16yWPfV+6dyl3fHsHLq8Ln+5j3eF1fPLbJ2zsv5GycSFGFB44AF98AUePQps2qrGt2Rx8nLjq+Pw+nlv6HJM3TSbdm07r2508t0qN587O7AOPEfy6n5nbZzK46eDcD1gIEfFKxZRiaIuhF/+CLVvg0KHgJaxpaTBmzGVJiPyw64eQDaCdHifTtk4LTojs3Kl+niYnqz5objcMGwYvvHDJsUQSSYjkceM2jKNkVRi2Em79G34pB9FuuHM7nCgZIzeDQoirWpTFzscfH+DdD99n1G8f8GrLE9y8XyWOzyl3GjHNYBUAACAASURBVLwhOmaZvVA91PJdXYekpBzX3W4+sRndYkXLuEC/DV3npd4TeK5oBbS9+4jK8KFrfrSeA+HFY9CoEWzcyIwDr5NWLLjXiNkHvxZJp+3772F45RXQNFbuX8nY9WODnvzc8e0dnBh8Qj2ZqlNHNXvNHo7Hw8iHqjGukYFksw+XL/S0jZioGIpbCkGzKurGLRSDQSVFfvxRVZF07Ahvvqn6lixcCO+8o8qBO3SAIUPg2DFVVXLsmDq2d290q5WH5j4UsGTH5XPhzfDy4vIXmdQ5Wy+XxYvhrrvA61U3bFOnqqVAq1ap/i/iqrUjcQeNP2vMGXdmadicqrC8HAz4DT5tCGlnK7+tHnC44aRDLV/z+S/cdFgIIS6b5OSgis/zTp26LF/CbDSjEbwsRkMLXgKj6+rn9JEj6vfnjByp+qu1bn1ZYooE0lQ1j5uzYw6PdIQSQ6DN/fDqTbC7IJj98GHfWuEOTwghws9mwzbkeV767jgrP0xj5SfPcaxqKfwGDR2oe8pMnQQDUdne+0T54fH1Ic5nNkNGBnE5VODZoxxqyUhONzagem8kJBC1bSdRGW4ANF1XFRZDh0KnTvC//1HgxBm04H6n6BrEuoCUVF5d9hIAn//xechGrTo6K/afTYKMHBmyckPz+RiyxMnyj1LRnOlB4wLPH6dpxE//Ek5fYFxxSgp89ZVKcJw+rZITDRuqHiXduqmEzLZtMHYsVK0KTZvCxIkwbx48/TRcfz3x8fs4nno86NQ+3cfiPYuZvX02DcY3oOS7Jbnr285se+wedKcTv0f9WZKaClu3qkSLuGr5/D5aftkyIBkCZyutzGDUYeoMaL0H6h2DIb9AhbMDlUwGE3dee2cYohZCXJUaNAg9+c1mU1WXl8FtVW4Lmei1mq3cW/vewI0bNkBCQmAyBFTFykcfXZZ4IoUkRPK446nHSbZBsk2Vdzuj4MPGUHYQVL7htnCHJ4QQeYo9ys5tfd+k5M4jGHx+tG3bYNw4Fnaaxu017iLKGIVZM1LtlIHF35qoGGrabGws1KjBE42fwG4KTC7YTDb61+uH1qEDlCt34aQIhHhOg7r5cDpB1xmwQTV9DdwPcRlQMx7mV4W31o1m9vbZuLwudIL7fum6jtt3NlHQpImqprj++uA/Gy+UPQ19fs853kRnIp/PGobudl/wugL4fGqZ0fDhmc1cQVVynDmjqkjO3QSmpcHevcR9NS3ktZy7nntn38vvx37neOpxZu+cQ+0eyViGQdQwaNMbdhdCnXfKlIuPU+Q7y/ct54zrTMh9bhOsLqOqa3/8GlZNhOJp8GcZEzaTjaHNh1KtSLVcjlgIcdWy2eDDD9VDC8PZt+B2O5QvD/37X5YvEWeNY3KXydhMNhxmBzaTDavJygs3vkD9kvUDDz5zJjOO7JJC3RzlX9JUNY8r/k5x4tPiQ+478cwJikVL8zkhhLhY6Z50MrwZFDQ6YNkyNTUlMTGwaeh338Ft6inLgAUDmLRpEhajBZfXRc+yt/HZqJ0Y9u9XVSBms7qh6dxZ9bfwBmY3dHJIimTxVjN4uSVYfOp4hwc67oRv6qilPl4jFLMX471273Hv7HuDEglGzUjSc0nEWGIyNy5aBN27qxLdbJZUgLb35xxP28NWFk7RMDrT/yHybEymoOvPUZMm3PVUSRbsWpCZzEElnHT04PG/Wf4gDX4omAG7x0BsVAwGZ7pKMjVtCjNnctjiIik9iWuLXIvZKMtK87Opm6fSb14/0jxpQfs0NPpWuIuxW8txOvkE8e1b8H3BE+jodK3RlRpFa4QhYiHEVW/DBlWBcfy4Wkb64IOX3JMru1Ppp5izYw4Z3gxuq3Ib5QqUCz4oNVVNj3Nmqzy129WS1AGRP4XrYpuqSkIkj4t7My7k0w+jZuTE4BMUtoeY6yiEEOLiOJ3wzTeqqqJsWTV2t2rVgEMS0hLYdWoXFQtWpMTg4SrxkbWCwmRSPUH++ktVQGQRKiESaluiHabVgDMWNUnstZvAleW9vIZG64qtWbFvBT49sBzWbDCz6dFNXFvk2syNv/2mmqKmpgYc6wOm1Ib77sr5jyQ2KoZDM8oSu20PnOuTomnBZbVZRUWpRnEXmxDp0IHTs6bQaWonNh7biNlgxuVzcde1d/H9zu9DvsHNyu6GkcvgiXWZf5YnHNClh4GN5aMwG8yY/DAu5SbujqoHvXpBNakGyG8OJR+iytgqIXviRBmj+LP/n1QvWj0MkQkhRASYOBGeeEL9rPf71cSbKlXg11/zRX8uSYjkEzd/eTMrD6wM2l7UXpTjg48HTCsQQghxhRUoELLqApNJPelZsuR8UiTVDHsLQuWT4DGBpkO6Gf4uDM0PBp/CbTFR7GkfyZbQP5eNmjEoGQJqnN6YDmPoU69P5kZdV4mdvXsDOtqnmaHV/bC+dM6X6DA7iB+4H/ub76peIV6vShZt3hyYCMoqJgaqV4c//gBPlo62mpYZz/kv4FBVOLfeCqimmPtP76dO8To4ohwUH1WcDN8FGtae9eDv8PnczM/rPwybi6uKmnPsblj7mWqeu6O0hcMPd+eWIeMwWnKYxSoizuAfB/PJb5/g9GY+5YwyRjGvxzzaVmobxsiEECICbNyo+nHFx8Odd0LPnmDNHz8jLzYhIu+m87g327yJ3RxYRmU32xl1yyhJhgghRG7LPi4vq6lT4eOPoUULtl9bhMdvhfr9oeQQ6HUXdOoBFYaY2FcmOrM81mJRyZQyZYjq9wif3JpzIzMdHbMheAmIQTNQMrpk4EZNU1NgqlYFhwNfTDRpZniyfWYyxGq0Uia2zPmfMQbNgN1sZ2yHsdjjisAbb6gRt/HxMHu2qgLJTtOgbl1YuRIWLFDLVqxWlSCJiYERI1QyJSZG9WaxWuGZZ84nQwCuLXIt7Su3p2RMSWItsfSo1QOb6cJPpuxuaHAs8/PNxeDvIoHJEIAH/oCqp8Ckw3WHXNz46lf8VsXBdyN6kH40RFZKRJxRt4xiSpcp3FLxFhpd04iRrUZy6tlTkgwRQoiL0aABTJgAc+dCnz75Jhnyb0iFSAR4atFTjF03Fj9+NDRqFK3BTw/8JMtlhBAitz3wgGrkmbUKwmCAm2+GpUvPb5q2ZRp95vYJmgpjNVnZft96ys/7GbZsgXr11NMYh+P8MU0mNGHdkXVBX7p0bGlOOU8FPAk3aAZKRJfgwFMHMBlMwfHquqrsSEriYJXivPnHGFYfWk2VQlV4rtlz1CxWky///JJ5O+dRIroEAxsNpGGpHB6mbNgA/fqpuA0GNSln9GgoUybwuEOHVF+W6tXVjZXfD6tXw8mTcMMNas3yBbi8LgYuHMjkzZNBB48nA03PTHYY/FA4HXaNgbizKyV+rAR3d4PkLPdxcelw7N3gprWpZrVEqWQamB/qj+HNtyAu9EQhIYQQQkQmWTKTTyzctZBu07sF3FRHGaNoVqYZy+9fHsbIhBDiKpSYCI0bq6qJ1FSIjlbJjLVrVaf4szw+D80+b8bWhK3nv387zA761uvLBx0+uOCX+Ov4X9z4xY2ke9Lx6T40NGxmG9O7TcdmstFrVi/OuM7g1/1UKVSFWffMolKhSlfyqgOlpalmsqEqRi7nl3GncSr9FNHHT/Ls2235tngCHiM0OgzpdjPrPvagoXqInLTBNc+AK0tO6PYd8PXszKRJVilmiPGc7ecSFQUPPQRjxvzj1CAhhBBCRAZJiOQTrb5qxYr9K4K2W01Wdj++m2tirwlDVEIIcRVzu9USkr/+Uo06u3UL2SE+w5vBxN8nMmXLFBxmB482fJQ7r70TTfunuTOwM3Enr616jQ1HN1C1cFWGNh9K49KNAfDrfnYm7sRmtlG+QPnLfXV51skT+5myaTK/p+3m+muu54FDRbD3uv9849dhbYyMbm7CqasMSNtdMGM6xGRre3Ju0VPAolO7HQYPxvfyS3y39TumbJ6C1WSlb/2+tKvU7qL+zoQQQoggbjesWaOWmN5wg3qgIHKFJETyiRof1WB74naivHDb31AiFX4uB/tKO/ilzy/ULVE33CEKIYQQ4aHrqpGr349evz6zdsxm9NrRnDz0N3euPsnrS3WM2W5zchqFrMfF0fHjZmzZuhLHaSd7CoHZ5uDhBg8zut3o3LgaIYQQ+cnSpeqhid+vfl6ZTDBzplpmK644SYjkE48ueJRVC8ex4kuwesDsVzdzc6rDHeuTcdhiwx2iEEIIkfccPMjBt4ZSaOJk/GfvdSw+iPLlkBABJjbQuOa0TsUkKJkKz94CX91gZcujW4KWJW1L2MbPB36mmKMYt1W9jSjjlV1CdDXRdZ1VB1axfN9yijqK0v267hSxFwl3WEIIcfESEqBChfOT585zOODgQShUKDxxXUUuNiESogNb3qBpWnvgA8AITNB1/c0whxQW10SXYvZUKOIMLO/tuhW2j36ROkPHhC02IYQQIs8qW5ayH32N/t5E/pg6mj+/fIvuv5zOsUJkXyGNx9vrWLyqF0m3bfDKMlhSxU2jzxpRKqYUjzV6jPaV29N1elf+Ov4XRoMRs8GM1WRlxf0rqFmsZm5fZb7j9Xu589s7Wbl/JWmeNGwmG/9b+j/m95xPy/Itwx2eEEJcnGnTcp5MN2MGPPxw7sYjcpQn57ZqmmYEPgI6ADWAHpqm1QhvVGGycyflkoP/oqJ0qPjBpLCEJIQQQkQKLSqK+vf/jz4rkkhZ/gMZ7duo5qnn+oJoGq4oI4/cqpNhhmQbZJhhenXoeye8stTP5M9O0XThFp6d9zgVPqjAhqMb8Pg9ZHgzSHGnkOBM4M5v7yQ/VN2G2zebvjmfDAFI96aT5kmj2/Ru+Py+MEcnhBAXKSnpfI+rAC4XnDqV+/GIHOXJhAjQCNit6/peXdfdwLfAHWGOKSzqFaqJOYfkYvTJFLUeTQghhBD/qHjz9th/WAK//w5dukDlyuh33MHNfQ0sqRx47AN/wbyp0GMLtN8D7/4Iqz71YHWH/rl7LPUY2xO358JV5G9f/vnl+WRIVi6viw1H8+fyaCFEPtS6deiG60adl0w/s2Jf8NAMER55NSFyDXAoy+eHz247T9O0hzVN26Bp2oaEhIRcDS43ten0VMjSXkCVYfnkaYkQQgjxr9SuDdOnw65deGd8x9ri3oDdMRnw7mJweNS6XYBoD1Q+CQ/8GfqUBs2Ayxtixq+4oJ8P/EzD8Q2xjLBQenRpjqQcyfFYmfYjhIgYN9wAt96qeoaclRoFM6r6GJH2A7dPvZ3/Lf1fGAMU5+TVhEhO/c4yP9H18bquN9R1vWHRokVzKazcZ4my4auiHlulm2D4TdCxBzR/EAbdU4DT3tQwRyiEEEJELrPRTO3itQO2NT4CbmPwsQ4vdMmhCMRqsgadR1zYb0d+o/3k9mw8thG3z82RlCMcTD6I2RA8ltJuttOgZIMwRCmEEP+BpsG338Lnn3PypkYsqmrkgTvgvs6go5PmSeODdR+w59SeXA8tIS2BmdtmsmTPErx+7z+/IJ/Lq01VDwNlsnxeGjgapljCzvTFl+i3tGFk4wyeWQOaDiYd9hU8TZc36/PDsB3S3V4IIYT4j8Z1HEebSW1wudLwGiDVDMYQy1X9wElb8HarycqULlMwGkJkUUQQPTER9ysvMcw5EWcZd8BjMLfPjVEzYjfbcfvcWIwWDJqBWffMkj9fIURkMRjg7rsZWWAdo9esD9qtobF4z2IGFBqQayGNWj2KYSuGYTFaAPXz68feP1KnRJ1ciyGvyasVIr8BVTRNq6BpWhTQHZgb5pjCp1kzNr73LM//AgVcEOdWZbzXJsCEsfuZs312uCMUQgghIlaT0k34o/8f9K3WgxuOGOi/AazebKWpQLoZPmqkbmINGDBqRpqXbc6ux3fRpmKbsMQeab5YN45rRpXAXugTlpR2h6wJtpltfNP5G15v9TpjOozh0KBDNC3TNPeDFUKIyyDOEhey8s1oMBJric21OH45+AvDfxqOy+fijPsMZ9xniE+Lp93nN1/VTavzZIWIruteTdMeAxajlu9+ruv61jCHFVZRi5cSle3fqUmHIik6CT/OgevuCU9gQgghRD5QpXAVxvWcQsKnf+PYuhHz2WzIuaSIywAvtYQdNYsxotGToEGnap24rth14Qo5oui6zvPLnued1aPw2XPoFn+Wz++jXeV2dDZ3zqXohBDiyrm39r28+cubePyegO26rnNHtdybG/Lphk9J9zgDN2rgTE3i55mjadltSK7Fkpfk1QoRdF1fqOt6VV3XK+m6/nq44wm3okluTCEa20d74MHXF8CiRbkflBBCCJGPZBw5QMwvG7FneQChZfnN5/WhfeX2vNDiBV5o/oIkQy6SX/czYMEA3l79Nj4unAyxm+0MbDQQuzl4OoMQQkSiigUr8lmnz7Cb7MRGxRJrUR/zeswjxhJzwdfuOrmL9t+0x/Kahdg3Ynly0ZOke9L/UxynM5KCKh9BtWNI/XzcfzpnfpAnK0REsGJd78e5egP2wMQiGmA/lQKdO8NPP0GjRmGJTwghhIh0G+eNp64Bsr9n1wCzD/r/BseahCOyyOT1e3lx+YuMWTeGdO8FbuDP3qHHmRw83exZXmzxYu4EKIQQuaRXrV50qtqJ5fuWYzaaaVWhFVaT9YKvSUhLoPGExpzOOI2OjtvtZvzG8WyL38aS+5aQnJHMpL8m8fux36lTog7317mfgraCOZ6vW+U7WLF1AWnZWk96jNB87bHLcZkRSRIiEcL4YB/MH7xHxr59WL0hcnsZGdCtG+zfr7oaCyGEEOJfOeLw0TCHAgYj0Gm3xpnreuRqTBHL6+WxkU352v0b6cFL5wOUOQN751fCuH0nmlEapwoh8qcYSwx3XHvxS2TGbxxPujcdPUtdR4Y3g18P/8qiXYu4b859pHnScHqc2LfaGbFqBGsfWkvlQpVDnq97vd58NuFR/ijiI80CBr/ql/XOjxBXotwlX1+kyrNLZkQ2djvmjX9iffQxdEMOf22HDsGUKbkblxBCCJFPlG15B4figpupAqSZwG6NpV2ldrkeV0RJSYFHH+V0AStfuX/D+Q/JELMP3jhSHdPKVZIMEUKILDYe20iGNyNou1EzMnT5UE6mn8R5tieI0+skKSOJAQtynlgTZbayvOpIPvvRQpdt8NDvsOoLeHSbHUaMuGLXkddJQiSSxMTA66+jmXO4u9B1+Pjj3I1JCCGEyCcalW5Ml3vgmCNw1YzbAIk2iPedIf7p/rB5c9hizNN0Hdq0gc8/57DDh/kfhhZoaIxtM5pe07ZBqVK5E6MQQkSIuiXqnh+Pm5VP97EpfhN+PbCk0a/7Wb5vOboeKq2vmJ8ZQo+eI5mxtBCfLjTQgJIwbhx06XLZ448UkhCJNDEx0LNnzvudzpz3CSGEECKk/Un7qPqsjd2FoM6j8HJLlQQ5Gg0bSsI1qdBqr06RMROgcWN4/arv9x7s119h2zZwuyl/GryhCj50MBlMOMwOZtw9g/4tBuV6mEIIkeclJfHUJgfPrtZocDRzs8VooWHJhiETJaC+v16QpsHTT0NiIqSlwZEj0Lv3ZQw88khCJBKNHw8OR/B2m+3CyRIhhBBChNT6/XoctLlxRkFiNIxoCeUHwc9loeExNerepIPRr0N6uiov3r073GHnLVu3gl89sYx2w1Nrwe4OPMSCkffbvc+JwSe4q/pdYQhSCCHyuJ9+grJliX3+ZV5Z6uHnLw18PQtsBgv31bmPBb0W0Kt2r6CkiMVooft13dEupp+kpoHVKr0nkYRIZDKZYOZMsNshSrUJ1qMd/NmoDD90qExCWkKYAxRCCCEix96kvRzUk/Fkq2hIi4JEOxhCVR/rOsydmyvxRYxq1SBLn7PXl8EbS6F0Mtg80PyQgZ87fMfARgNxRIV4sCOEEFc7rxfuugtSUyEtDc3rw+b202uPg7Rqkxh/+3iio6J555Z3qFuiLg6z4/xHa1NVxlZ5Uv18yk7X4b33oEQJMBqhVi1YujT3ry8P0i60xihSNGzYUN+wYUO4w8h9+/fDxIkcP7aLdpXWsEc/iclgIsObwZONn+TNNm9eXIZQCCGEuEptjd9K60mtOZF6Qs3XzeaNJfD0GojKPn3GaoW33oInnsiVOPOElBT4/XcoUgRq1gzer+tQty7s2AHubKUhNWrAhAlwww25E6sQQkSin3+Gjh3hzJngfR06wMKF5z/VdZ21h9dy6PcV3Pb8F9j3HUYzGNT36ClToFmzzNe+8gq8/XZgewWbDZYsCTwuH9E0baOu6w3/6TipEIlEKSkwYwasWQODBtG16WG2eY6S5kkj2ZWMy+fio98+Yvq26eGOVAghhMiz1h9eT+1xtTmRFjoZYvZCugn8Od0tXU1N6D74AIoXh06doFEjqFNHrT3PStNg5Uro2lVVsBqN0KoVbNqkltNIMkQIIS7sQsUK/sDMvKZp3FCiIXc/+iGOHXvQMjJUwuPgQbjxRhgwAFwuyMiAUaOCe02mp8NLL12Bi4gs/9B1ReQ5CxfC3XefL0k9bHWzcaAfL96Aw9I8aby39j3urnl3OKIUQggh8jRd17lz2p1BXfrPMfihcDp07T0Sa/eCMGhQ5nIQvx8++QSuuSYXI859To+TbzZ9w7G5U3j+rdVEubzqBhpUguO22+DPPwNfVLAgTJ4M33yjbuwN8uxNCCEuWpMmob9vOhzwwAPB2xctUstrQiVSPvtMJUc++ijnRMvWrZcUbn4gCZG85MAB9VGjhip1yi4pCbp1C8junbaByQWEaDS8NX4rx1OPUyK6xJWLWQghhIhAS/Yu4WT6yRz3l/BEsbTbLKo3vk1t6NQJ5s1TVRCdOql12PlUiiuFsevGMuLnEXj8HmZ861X3Gln5fLBrF2zfDtWrB59E06RZnxBC/FtRUfDdd3DnnSqJkZGh+ka2a6feB56TkgLDhqmliGlpoc/l9cLy5Wr5TU7fj2vUuPzXEGEkIRJO6ekqaxcTA/36qX+wFosqberXD95/PzBDOGdO0D/maxMhyhf69CnuFGp+VJNDTx/CbrZfwQsRQgghIsO+pH3c8e0d/H3yb9w+d8hjNDSGdHorMxkCUKoU9O+fS1GGT3JGMg3GN2D/6f34dHWDUTwthzXWJpMa3SiEEOLyueUW2LcPpk2DU6egTRto2jTzfaDfDzffDFu2qPeNF2I2q4loTz8N774buGzGbodXX71y1xEhpI4xHHQd3ngDihaFhg2hTBlYvFhlAJOT1a8TJ6py3KyczqC1YyY/jJ8HFj1ba/yzTmWc4oVlL1ypKxFCCCEihl/30/TzpmyO34zLl/NNZEFrQfrW65uLkYVXQloC6w6v46TzJGPXj+VIypHzyRCAudXAGeoRmtcL9evnXqBCCHG1KFYMHn8cXn5ZNT3N+lB8xQrYufOfkyEAHo+aAPbKK+qjSBF1rurVYfZs1WvkKicJkXCYNAlef12VN6WmqiSHL1uZh9MJo0cHbuvQIeT6ry4H7DxdvmeOX+6bTd9cjqiFEEKIiDZo0SD8x49zy26oFmJCvYZGxYIV2fHYDmIsMbkfYC7znDjGA89UouybxWj3UROuebsYH6x5jwxvRsBxHzWC49GZSREd1JPFUaPUunYhhBC556+/gid5hWK1qmbWNWqoJMjgwZCQoN53btsGbdte+VgjgCREwuGNN3Je65VVUlLg5xUrwnPPqZuQc0tpHA645x4aNe6c42lS3CmXEKwQQggR2RKdiXSd1oWKw8ew/z34bjpsHA+rPofYsz1CaxWrxaoHV7H78d0UdRQNb8C5wePhxcdr8J11LxkmSLaCS/Nzynkq6NAzVqj3CLx2E2yq6EDr3Fk1eR8wIAyBCyHEVa5SJdVmITurVe0zGNTv77sP5s4NPk76OwWQHiLhcOLEPx9jMKhRddkNHw7t28PXX6vMYPfu0KoVHXUfBs2guuXr0GsTPLsaiqbBLxV97Gq7iCqN2l/2SxFCCCHyMp/fR7PPm9Hix7/p+wfYfOoDoNFhmDQbOvfUeKLxE9xY9uopHdbnzuXjKqdJjwrcntOI4VSbgemdKjDg/hUQV+bKByiEECK0W2+FQoXUioJzqwwMBoiOVpO/bDb1uSQ+LopUiIRDo0YX3m82q0arb7wRen+TJmp80mefQevWoGmYDCZeu/k1jJqRl1bCuPlQOx5KpkHnLT6K3dSBE1vXX/ZLEUIIIfKyH3b/wMmkozzzK0R7AvdZ/NB+DxRyG+lVq1d4AgwTfcsm0syh9xnQsJqsxFnisJvtlIopxbQu0/j78b8pI8kQIYQIL7MZVq9WzVfNZtXgulkzWLNGJUWMRkmG/Av/qkJE0zQDEK3r+pkrFM/V4a231D/i9PTMJqlWK1x/vWqo2rQpPPOMarb6Lzx/4/MYUtN44rWR2L2Z20062Nyw/flHKT5342W8ECGEECKPOn2a06eOsmzSKzy3IpVSOawe9WkwuFZ/bGZb7sZ3ha3ct5KhK4ayI3EHVQpVYUSrEbSp2Ob8fkPVa6n9i4G/ivmDXnuDrSrfP7aa9UfWU8xRjPol66PJzbUQQuQd11wDP/yg3jvquqoKEf+Jpodo0hlwgKZNAR4BfMBGIA4Yrev6qCsf3sVp2LChvmHDhnCH8e9s3aqWv2zYoHqDDBsGLVte8mlXTh9Fvd7PEhei6fC+0g4qHEq95K8hhBBC5FknT8J997F094/c2dWLpoPXAOPmQc/NYM5223M0zkCJk24MxtDT2iKFX/czf+d8hv80nC3xW/D4A8thbCYb07pO4/Zqt6sNLhe/NC1Nu/aJZJjUUhmjD6x+jZX9VtOw3A1huAohhBDi8tA0baOu6w3/6biLWTJT42xFyJ3AQqAs0PsS4xM1a8L06WrG9LJllyUZAlC6emOifMHb/cCmaCfTt06/LF9HCCGEyJM6dCB95RLu6uIlLQpSLZBhhqGt4bQNXGfvi5MU2AAAIABJREFUfLwauKKMxE6ali+SIV2+60LX6V354/gfQckQgHRvOk//+HTmBouFG7//g3U7m9NjK9Q6AfeeuoaNPVdKMkQIIcRV42KWzJg1TTOjEiIf6rru0TTtwmUlImwqX9eCNbWKUndTwvmmcQDpZnizmc5fc+6nSekmsgZYCCFEvrH28FqGrxyO6/ffmP/nKVZUgOx3KkfioOaj8Nk86LgbPLe0xvbO+1iuuy48QV9Gi3cv5sfdP4ZMhGS1+9Ru/Lofg3Y2K1S6NNfNWMU355ryRXhiSAghhPi3LqZC5FNgP+AAVmmaVg6QHiJ5WN3l21jTsBgZRnCa4LgDeneGtWXAp/uYumVquEMUQgghLotle5fR6qtWLN6zGMfxU3gM4Mnh7iYhBr67DowWG7bpsyHSkiGpqWq5bdWqqtL0/ffB42HOjjk4vc5/fHkRe5HMZEhWRqMkQ4QQIj84fRrWrYNXXoGePdUgjpQcmmgJ4CIqRHRdHwOMybLpgKZpN1+5kMSlshUowl/v/Y+7F/4Pa5qbozGgn73/8fg8nHFJPksIIUT+8MQPT2BLTsdrhT9LgMULrfeFToo43NDzb4u6UYyJyf1g/yNd15m88Uuu7zyQcsczsHrPlr8MHQpLlxL9RDUMGPAT3CD1HLvZzvM3Pp9LEQshhMhVTif06QOzZoEnS7Xg99/DyJGqb2XJkuGLLw/LsUJE07R7z/76dPYP4Ilci1D8J+0rtyfNauBIXGYyBNQNUceqHcMXmBBCCHGZ6Fu3MuXVbfTarKbFHImDSXXB4IcJc8HmAbMX0MHugU5Jxbh15HQYMiTcoV+Qz+/j+x3f89jCx/g/e/cdHlW19XH8e6YlMwkBCT303iwgINJUmqgoCBYUFb33CmLHju3aQEUseC2vShFRFBVRAUUFERWlSu+995oyM5ly3j9OJIQMEJDMpPw+zzMPzD77zKwjSGbW2Xut52c8z23f3MaU1/pTabc3OxkC1gfg6dPpbzYnzhF33NdzO9wMbDOQAS0HRCF6ERGJuj59rORH4JitkxkZsHu3lUCXiE60QiQh69fCcwtFjmhQtgF3NLuDD+Z/QHogHYAEZwI9G/bkgpQLYhydiIjIP5SWRrhtG84+AH4HhLO6wva/AhaWh3vnwM8fwottYFpN6Nrkej655tOC1z522zb49lu8BHBe2Z1wxQp0+KgDC3cuJC0zDZfNRWY4kzfWQ4nMCOeHw9Res5fXLn2N+6fcDyZkhjMxMalRqgYvdXiJqxtcjdPujPqliYhIFOzZAxMngj9Cm1GAYNBKlkhEJ227G/Ekw3CZphnpx3JMFMq2u1Fgmia/bPyFDxd9SCgcovfZvelSu0vB+zAoIiJyqkaNwntnX9y+ID/Wgh7XQ7orwrww2O12djy4g7IJZaMe5vGYpsm0R6/DNe5L7roCVpQFRxguNmrwa9xOvEFvjvkD/oAXfgZP8JgXKlECRo+Gq69mv3c/U9dPxe1w06lWJ+Id8dG7IBERiY2lS6FVqxPXCklJga1boxdTAZDXtrsnTYgYhvELcKtpmhuznjcHhpumee4ZiPOMUEJERESkGFm+nPTbb8X9x1xsWAXEB7eBtcnwU0044AETa+vMWe5STLrpe1pWaRmTUH2/TMX+xJM4Fy+DUqWgdm1o1YqRZbfSctBHXHC71Rr4b3FZW3z8xyzoKJMO64ZB0lG3o8IG2MqVh82bwRUpGyQiIkWd6fWSWbokcb7jdBpzu+Gxx+Dpp6MbWIzlNSGSl7a7LwJTDMN4E0gBLgNu+4fxSYxlhjKZuGoimw9tpnlKc1pXaa2VIyIiUvCNHk3grjv433k+uiXDPVfAjGrWoQu2wu8j4av6UP+wg473vEbJ2++JXmzLl8OCBVC9Ostql+TfY69n3sHleNrBghVQc2saxtatmL/9xk1miBFNrO0+R/M7wBHK/dJ7E6DjLfDZl1AhLasIXP36xH/1rZIhIiLF1Lr96/h06accusTkmamQkJUT+XvJgxEfD+3bWwkRiShPW2YMw7gY+AnYCzQxTXNnPsd1SrRC5NSs27+ONqPakJ6Zjj/kx2lz0jylOVN6TzlhUTYREZGYSk0lrUp5Wt/gZVkZKO2D/W4IZXWMtYWhTAYsfRuSXhlG3F1RqgEfCMB118EPP4Ddzl63Se3bvRxyhsGAO2fBkKmQcMx2l8MuSH4Ugsd0vI0L5k6UuOwuDAwchp0G6W5e6vIqHdr1yd/rEhGRAmn1vtX0GNeD9QfW4wv6MDFpvQle/QEqpcKysjC9jp3Hn59ByeatYx1uTJyxFSKGYTwFXAe0A84BfjEM40HTNCf/8zAlFm4YfwO703cTNq32fJmhTGZvnc0rf7zCk+2ejHF0IiIix/HbbwxpGWJ1MoQcOZMhAGEbpDthUl2oOXYID7s+pnJSZe44/w461ep0RldCLtm1hGV7llEvuR5NRn1vJUO8Vt2Pt8+HVAeQ9XYjmkHVVHh0Zs7XcIWg+Vb4s1rO8aDDRpzdid1mx27YcdqdTLtlGmU9ZUnNTKVO6TrYbcdkUUREpFgIhAJc9OFF7ErbhUn24oaZ1aBl3+x5Hkccd9apTMkYxFiY5GXLTBmghWmaXuBPwzCmAMMBJUQKoT3pe1i8a/GRZMjfvEEvoxaMUkJEREQKLrebsfUD+LLqa4Sz2sqX8sKtC+C8XbCgAvxaDaa4tjF3+zbmbp/LN6u+oVPNTky+cfJJEwm+oI/xy8ez7sA6zil/Dl3rdsWRGYQJE2D9erznNqTb/neYufUP7IadkBmiydYg34cyj7Tl+6BpdmxgrfZ47iKotw+6r8weN8iqAxLOnh9v2rn67Ot4qt1T/LrpV8p4ynBF3StUIFVERACYsnYK6ZnpOZIhkSS4EqhSskqUoiq8TpoQMU3zvmOebwI65VtEkq9CZoSNyVmC5rGl60VERAqQtm1xfm0DrJ9lJlBzP8z+ANwBa0vK1cthYj24+3KwhyFkQJgwP6z7gWGzh/FAywEcmjqJ/5v4X6a4t1G5ciPu6/EyzVKas+ngJlqOaEmqP/VIy3obBn2WxzHsJzslDnl5/HI7vzUJ4rNnfxCdlwz3dYGR38IeD+xKzB16hgteaZWdEDENg/SKyYRK+aizL40NpSExPon7Wj/I420fx2Fz0KBsg3z+DyoiIoXNzrSdJ/xO97fD/sP8uO5HutTuEoWoCq+8bJkpCzwKNASO3J4wTbN9PsYl+aRCYgVql67Nsj3LcozH2eO4ofENMYpKREQkDxwObm91N08uHobXBRjwziQ4ywd2Ew7HQdvbYP1ZkJZ1/OgbaM//+jy3jl9H0/C77E4y8TrBtnc3X71/IR9d/CbpbwxhzNKdLC4Hb7WADaUhjMno+j6WlIQ5H8CoxmF8xywy8TtgVFOYVhPu/9OqAXJsXRCAHSWwCqDGx2O43ZT+6Rfm1q+PaZoqbC4iInnSqkor8lIH1B/yc8uEW9jx4A5tszwB28mn8AmwAqgBPAtsBObmY0ySzz7p8Qkl40ricXoASHQlUie5Dk+0fSLGkYmIiJzYPde8Qp2UxtjDQBg6bLCSIQDPtYNVZbLa2P6dXzgqKVJu60FeXfIeOz1WMgSsrSoVDoZof/ldXPvdJjquh7vnwOJ34YIt2XNWloHfq4LvBLeSNpeCJzoc9d5HsWHQvmlPeO89+PRT2LIF6te3QlQyRERE8qhRuUZ0r9+dBGfCSed6g16W71kehagKr7wkRJJN0xwBBEzTnGGa5r+Alvkcl+Sjcyucy8b7NzK001AebvUwo7uPZkG/BZSIK3Hyk0VERGLIaXey6M4lPND6YTxxHjKPuuk19pzc3VmOdtk6g29rh/E7c44P/RGS/ODO2jnqCkNiAN6blD0nZIMl5aHDeqvmx/F4XXCWswQehzs7ZpuTs9yleab7G3DrrXD55eB0Hv9FRERETmDM1WN4o8sbVEyseMJ5oXCIBNfJEyfFWV4SIlndjNlhGMYVhmE0ASrnY0wSBaXiS9G/eX+GdBpCjwY9cNjyUl9XRESkYBjSeQipA1NJv+aq7C0sJ1lBXKvKuST7cq/G6LA+e5XJ0RrtBk+m9XtHGOrthTe/h1J+q/jp8ex3hfj+pil0rdOVs8udzZ3N72Rx/8VUTtLHJxER+efsNjv/afofejXuddw5BgZ1Steh5lk1oxhZ4ZOXhMgLhmGUBB4EHsLqMDMgX6MSEREROQmbYaPse2PYXrMMaU64binEBY6ZZAIGuB1ubnz6CwbMc5CQmXNKmivy6wdtkGkHRwiqpBpcshFqBRJZ9VN9nmr9BC5b5BPrlq5Lu2rtmHjjRBb3X8wbXd6gUolK//RyRUREcris9mUkOiNU8saqHTmh14QoR1T4nDQhYprmJNM0D5mmudQ0zUtM0zzfNM1voxGcxFg4DH/9BbNnQ+DYT5giIiIFQFIS1VfuZMbwJ0koXQ530EpgGCY4jyrCP/668Xy0eSKu2/vz4FwnriDEB6zH2OZxhN0529r6HPB5I8DhoEejnvza7G1szz4HY8dSZs5SHu/0LMO6DDtSj+tvboebFzu+GIULFxGR4q5DzQ5cVP2iHPVE4uxxdK/fnS0Dtmh1SB4YealQe2SyYfxlmmbTfIzntDRr1sycN29erMMoWhYuhCuvhIMHwTDA4YCxY6GL2jaJiEjBlZaZxtVjujJ962+EMalfpj5NKzTlq5VfETbD2A07vswMa7JpFUy1h2HS92dx6VIvhstl3QRo2ZLQhK8wkpKwGce/f/Thwg955pdn2Ja6jTql6/BKp1e4ou4VUbpaEREprkzT5JMln/DqH6+y9fBWElwJNCrbiLtb3E2X2l2KfcFuwzDmm6bZ7KTzTjEhssA0zSb/KLJ8oITIGebzQaVKcOBAznGPB1atgsraAy0iIoXDtPXT6PZZN9ID6SecZzfsfN/6XTp5K0KtWtCgQZQiFBEROXWP/PQI78x958jPN6fNSWl3aZbeuZQynjIxji728poQOemWGcMw7jYMo1TW08n/ODIp+CZOhGAw93goBKNHRz8eERGR0zR60eiTJkMAQmaISZlLoWtXJUNERKRA25O+hzdnv5nj51sgHOCQ7xBvz3k7hpEVPnkpqloBmGcYxufA70ZxX3tTHOzdGzkh4vfDrl3Rj0dEROQ0hcKhk0/KcrL2hSIiIgXBom3zKRGwUekwOI/62uYL+Zi2YVrsAiuE8lJU9UmgDjACuBVYYxjGYMMwauVzbBIrF10UeTwxkfXtzmbEXyOYsGICvqAvunGJiIicohvPvjFHsbnjsRk2/tX0X1GISERE5B/IyOCcS/uw4UUva96EfUPg8V8B0/pZVuOsGrGOsFDJywoRTKvQyM6sRxA4C/jSMIwh+RibxErDhtCrFyRkf4A0PW7uvy6JRqvu5b4p99Hn6z5UerUSf+34K4aBioiInNjldS6nR4MeJDgTMDBw2py55hgYfNjtQ8ollItBhCIiIqfghhsot3E3iQHwBKFEJgz8De6eDfH2eAa0HBDrCAuVkxZVNQzjXqAPsBcYDnxtmmbAMAwbsMY0zZivFFFR1XxgmvD55/D+++D3M/Hac7gh/aNc+7ArJlZk6wNbT1iBX0REJJZM02T2ttlMWj2JEq4SXFXvKqZtmMbU9VOpeVZNHmvzmJIhIiJSODidEcsb7EqA2eNe5aorHohBUAXPGesyYxjGc8AI0zQ3RTjWwDTNFacf5pmhhEj+6zq2K5PX5K6pW8JVgh9v/pGWlVvGICoREREREZHiIRgKYHe4iFTU0wSMpUuhUaNoh1Ug5TUh4jjZBNM0nz7BsZgnQyQ6vEFvxHEDiJ8xEzZPh5QU6Nkzx1YbERERERER+ef6TurHI8lQf1/uY4bLZZU+kFOifQ6SJ73P7o3H6ckxFheA79/L4Nx+/8V86imC/e/An1KBHbOmxihKERERERGRomfNvjWMWjiKAV0g45hlDRkO4J13QA1hT5kSIpInN59zM80rNSfRlQiA0+bkkTkOWuy0Y6SnY4RCODK8OA+lsb9bZ+6afBcn244lIiIiIiIiJzf0z6EATKkDl/eG36rCbg/MqApX3GKHf/87xhEWTifdMiMC4LQ7mXbLNCatnsR3a7+jrKcsT44ei8O/Icc8G1Bzv8mPv4zg9dK1eOBCFfURERERERH5JxbtXHTk9zNqQLujuus2KFM3BhEVDVohInlmt9npVr8b73V9jxfav4DLFjmfZgL+oJ+HfnyIlsNbsjt9d3QDFRERERERKUKaVGiC3bBHPPZyx5etLqEZGdavkmdKiMjp69OHcHxcjqEwsKEUbCkFJibzd8yn+2fdYxOfiIiIiIhIEfBgqwdxO905xuyGnS61unDl77uhQgVISoLkZHj9dSVG8kgJETl9DzyAcd55pLmsREiaEw7FQ69rs6cEw0EW7FjAvG3zCIZz98sWERERERGRE6tdujYzbp3BhZUvxGbYSIpL4qFWDzGRG+Hee2H3bgiF4MABePJJePPNWIdcKCghIqfP7caY+QfL3nmWQe2d3HcZVB0AS8tnTynpBbvXR5v/a07pp108P7gLZlCJERERERERkVOxbM8ytqdux8CgXEI5WqS0wPHMs9ZWmaNlZMALL2iVSB4YRaETSLNmzcx58+bFOozCLzUVJk2C9HTo3BmqVs3zqWv2raH3V72Zv30+YcIAePxgGuB1Zc9LyISnDzfhkf/9daajFxERERERKZJGLRjF3d/fTUYgO/nhcXo4/Gwm9swIN5xtNvB6weXKfawYMAxjvmmazU42TytExPLzz1CxIvTrB/fdB/XqwbPP5vn0Osl1+K73d1QoUYE4u1VXxBnOmQwBSHfBc0kLmPrHx4TCoTN5BSIiIiIiIkXSEz8/kSMZApARyGBt8nG+0lesWGyTIadCCZFiaunupVz68aUkDE6g1suV8F95ubUyJDXVWmLl88GQIfDnn3l+zTKeMiy+YzGPtH6E8yueT8Zx/v9Ld0KPH/5F1SEVWLVn5Rm6IhERERERkaInGA6yI21HxGMPdQiBx5Nz0OOBl16KQmSFnxIixdCGAxtoNaIVP637iYxABo0W7sAf8uee6PXCqFGn9NrJnmSeu+Q55vWdRyOzTORJBqTaAuzw7uWKV5tg+nyncRUiIiIiIiJFn8PmoHxC+YjHVrWoCV99BeeeC2431K8PH30EN90U5SgLJyVEiqFX/ngFb9CLiVU/Ji4ERColY5pWUuQ0Db30NdyB4x83bbDT7mPx0IdO+z1ERERERESKCtM0mbp+Knd/dzePTX2MFbuXg9fLcxc/i8eZcyWIx+lhUPtBcOmlsHChtdJ/xQro2TNG0Rc+SogUQ3O2zcnRAndqTaveRy6JiXD99af9Ph3a3Mz3F7xJqz3x2I9TLsQehvSJX532e4iIiIiIiBQFpmnSa3wvun/Wnbfnvs3et4ZQulYjwokJ9L30cabtv5JKiRUxMKhWshofXPkB1za6NtZhF2pKiBRDjco1wm7Yjzw/6Ia7L4MMB5gOhzWYkACXXQaXX/6P3uuiK+9h5v8yeLXpY3gycx83DTh/axgefRTuvBN+/FHtoUREREREpNj5fu33TF49mfRAOr0Ww7DvTMqngS1swv79tHx3IttcAwk9HWLj/Ru58ewbYx1yoae2u8XQst3LaDG8RY4qxfGOeG5PaMebu8+3Cqt27w7t24NhnJH3zAhk0PrhZNYk+Eh3gSMEzhB8ONHGdcsNMpwwoU6IzWWdtKh8Ae1H/oJht5/8hUVERERERIqAW7++ldGLRgOwZhjUPhBhUpkysGdPdAMrhPLadtcRjWCkYGlUrhHf9/6eOybdwep9q3HZXdx63q0MufQ1cMTny3t6nB5m3TyDcXdfzKRqfioeCtNvuZuGW3ysTA7T9jbwOcDrCOAO/c7Zrzdm2n1/4Xa68yUeERERERGRgmLOtjlMXT/1yPMqh48zcd8+CATA6YxOYEWcVogUc96AF5fdhd0WpdUY+/fDxx/Dhg3W8+HDado7jYXlrSKrf3OH7TzW/mmevujp6MQlIiIiIiISA7O2zqLDRx1yrOBf8jY0jrQQJCUFtm6NXnCFVF5XiKiGSDHndrqjlwwBKF0a7r0XXn8d2rdnVyIsL5szGQLgtYX4cOGHBEIBMkMRio+IiIiIiIgUAQ//9HCOZAjAI52sGo85eDzw8svRC6wYUEJEYqdTp4jdfv+2K30XnkEe3C+4qf9Wff7Y8kfUQhMREREREYmGxbsW5xr7vi5c18tGZuMG4HZD/frw0UfQu3cMIiy6lBCR2ImPp8KnE6l7wIZxTNtfAwNvwEvQDBImzKp9q2g9sjWDfxscm1hFRERERETOtEWLGPSri8d+hVr7ch6a0dCDbdFiyMiAFSugZ8/YxFiEKSEisXXxxXz68CxKORJIMOIAcDvcGBiYR60fSTkEr/wAF/d+gr29r4aVK2MVsYiIiIiIyD/36KPQqhV3/rCfZ3+Bxe9C/znWIY/Twz0t7sFhUx+U/KSiqlIgpGWm8fmyz9lyaAu703fzwfwPCJgBAOruhdkfgDsIcSEI2Qzs8W6YMgXato1x5CIiIiIiIqdo/nxo185a/XEUnwMaPRjPNR3uZXCHwdGt91iEqO2uFCqJrkT+1eRfAMzcPJMRC0ZAyDr2yo9Qwg9//1NgD5vWPxz9+sHy5bEJWERERERE5HR9+SX4fLmGXS43K6sOwdnp7hgEVfxoy4wUOK2qtKJxucZHnl+8MTsZksOaNZCeDkAoHIpKbCIiIiIiIv+YzQaGkXvYsOF0xccgoOJJCREpcAzD4Jdbf6Fd1XYAHI6LPM+023l57huUfrk0jucdVH6tMmMWjYlipCIiIiIiIqehVy9wuXKPh0Jw1VXRj6eYUkJECqREVyIzbpvBgn4LmH9NK/xxx6wRiY9nfvv6PDtzEAd8BwDYlrqNW76+hYs+vAhfMPfyMxERERERkVgJhUPZ31POPhuefhri4yEuzmqtGx8PH3wA5crFNtBiREVVpeALhaBvX/jkE+sfCb+f0CUXU77FDPYZ3lzTbdjo37w/b13+VgyCFRERERERyZbhPcx9w3syZt90goZJwzIN+L+r3qdVlVawfj1MnAhOJ/ToARUqxDrcIiGvRVWVEJHCY8cOq4hqzZpsLm2nwdsNyAhkRJzqdrhJHZiqqswiIiIiIhI7a9dy+aCGTE8J4HNmDyc43Cy4YxF1kuvELrYiLK8JEW2ZkcKjYkXo0AFq1KBCYgVsxvH/+maGMvGH/BzyHWLMojG8P/99Nh/aHMVgRURERESkWFu9mvVtGuVKhgD4M7289udrsYlLjlBCRAoll93FY60fO25SpG5yXX7b9BuVXqvEnd/dyYAfBlDvrXoM+nVQlCMVEREREZHiZtH2v3jg+TY82zKTj7+ETa/BI7+BPas5ZtAGS3csjG2QgiPWAYicrsfbPk5GIIMXf38RE2vrl82wEe+I59XOr9Lz8565ttQM/n0wnWt1pnlK81iELCIiIiIiRdywsfcycPn/8NeEsAHjG8IVa2DE13DOLrjpGnAFoWVFfSeJNa0QkULLMAwGdRjE2nvXcsf5d9CkQhNubHwjs/8zm/RAesTVI76gj9GLRscgWhERERERKep2b1rOYyv+h9cJYRtgQHocTK4Dv1eDHiuh+n5wm3bua/dIrMMt9rRCRAq9mmfV5N2u7+YYW7xrccS5YTOMP+iPRlgiIiIiIlLM/Pj5izhC5Pqmne6CBzvD/02CWxcbXP/6FConVY5JjJJNCREpkjrX6kwgHMg1nuBM4LpG11lPdu+Gd9+FP/+Exo3h7ruhevXoBioiIiIiIoXfL7/AmDG4t83AaJL7sM2EuBB0vwG2dZ1G/HmXRD1EyU0JESmSynjKMKzLMO6fcj+BcIBQOITH6aFHgx50rNnR6vfdvDlkZHDY9PHpgZ9YvWIYza6+kx59XibOGR/rSxARERERkcLgoYesG61eL5c5TMymuafEhcDvgIDLzuTk/fSMfpQSgRIiUmT1Pb8v7aq14+PFH5MRyKBbvW60q9YOwzDg4Yfh4EFWnRWm9b/B6wiT4QqTuP5NnnrsHWb3/J7kVh1jfQkiIiIiIlKApS2ex7IJb1HB6adaBngC8M1n0K0XGCaYBgTscO1S+KgJeBwu9mbsjXXYksUwTTPWMfxjzZo1M+fNmxfrMKQwSUqC1FQu/DfMTgHzqPqrziD8a5mT/xu1B0qWjF2MIiIiIiJSYA2ZOYRnpj6J0xcg0w4XboXx4+AsH8xKgek1IN0Ja0rDl42sIqtuh5u/+v1F/TL1Yx1+kWYYxnzTNJudbJ66zEjxlJBAmgvmVcqZDAEIOOCLugEYNy42sYmIiIiISIH2zcpveG7Gc3gJcDgefE6YWQVuuMY63ny3ne86V+f1jh4+P9tKhiQ4E+h9dm8lQwoQbZmR4ql/f2xDXwR8EQ87QsCOHVENSURERERECodX/niF9EB6jrFMB/xSHXYmQoX0MNNu+omR+6by8eKPiXPEccf5d3BNw2tiE7BEpISIFE8DB+JZupRLNn3Jz9VMQvbsQ3EBuGWFEwa0iV18IiIiIiJSYARCAfZk7CHZnUycI45d6bsiznOFYJ8bKjhL4apRmztq1OaOZndEOVrJKyVEpHhyOuHzzxm1ZCatP76E/XZr358zBI32GjyTcQG0bx/rKEVEREREJIZM0+T1Wa/z7IxnCYQCGIbB/Rfcz6W1LmXTwU0EwoEc820m1NkPXHtpbAKWU6IaIlKspZzdmrXPHeKT5Nt5eXklJs2uyZ8NXiXh+6lgGLEOT0REREREYujDeR/w1NQnOOw/jDfoJSOQwRuz3yDeEU+p+FK4yFpqboInE4ZNAZenBDzzTEzjlrxRlxmR0zF1Kjz9NKxdC40bw6BBcOGFsY5KRERERETOlHHjqDXrBtaXyv2dOSkuiRU0ZNeUAAAgAElEQVR3reCNWW/w04LxVN14gIfmOmlbuz08+yzUrRuDgOVvee0yU+ASIoZhPAPcDuzJGnrcNM3vTnSOEiISVRMmwE03QUZG9pjbDd9/DxddFLu4RERERETkzFi+HJo1w/OAF68z92G7YefwwMN4nJ7oxyYnVdjb7r5umuZ5WY8TJkNEoso0YcCAnMkQAK8XHnooNjGJiIiIiMiZ9d57kJnJ2ZFrp5JSIgW3wx3dmOSMK6gJEZGCKTMTtmyJfGzJkujGIiIiIiIi+WPnTgiFGPqjVRvkaB7DxWuXvoahmoOFXkFNiNxtGMZiwzBGGoZxVqQJhmH0NQxjnmEY8/bs2RNpisiZ53JBiRKRj1WoEN1YREREREQkf1xxBSQk0HYzTP0ILtkAZdKh5TaDCZeOomfDnrGOUM6AmCREDMOYahjG0giPbsC7QC3gPGAH8Gqk1zBN833TNJuZptmsbNmyUYxeijXDsLbGeI7ZK+jxwJNPxiYmERERERE5s66/HurVA4+HC7fCz6NhzzsJ/Jn8KJ0vuDHW0ckZ4ojFm5qm2TEv8wzD+ACYlM/hiJyaxx8Hnw/eeAPCYXA64amn4N//jnVkIiIiIiJyJsTFwe+/w4gR8NlnkJQEd94JXbvGOjI5gwpil5mKpmnuyPr9AOAC0zR7negcdZmRmPD7Ye9eKFfOSoqIiIiIiIhIzOW1y0xMVoicxBDDMM4DTGAj0C+24YgcR1wcpKTEOgoRERERETmezEyYOdPqFtm6tfUZXiRLgUuImKZ5c6xjEMlP/qCfV/98lVELRxEOh+l9Tm8ebf0oCa6EWIcmIiIiIlJ0TJsGPXtayRCw6gF+/jl07hzbuKTAKHBbZk6HtsxIYWGaJpeMvoQ52+bgDXoBiLfH07BcQ+b8Zw52mz3GEYqIiIiIFAH79kG1apCennPc44GNG0GNOYq0vG6ZKahtd0WKpN82/8a87fOOJEMAfCEfq/eu4v+ev4q5d1zFqgnDs7PYIiIiIiJy6r74IvJnatOEceOiH48USEqIiETRnG1zyAxl5hpPC6SzceZ3NH1vIinX386f55cnM9Mb4RVERERERCSXzZth8WIIBq3nBw9a9UOO5ffDgQPRjU0KLCVERKKoclJl4hy5Czl5MqH2frADiQE4e8ke/uh3BUycCDt3Rj9QEREREZHCYPt2aNkS6tWDNm2sDpBffgmdOoHLlXu+260aInKEEiIiUdS9fnfcDjcGRvagCc4w9FqaPZQYhLajp8NNN0H16vDAA9pGIyIiIiKS5bD/MCP/GsEr9zRl3ra54PNBaqq1+qNPH3A4rIKqCUc1LkhIgCuvhBYtYhe4FCgFrsuMSFEW74jn93/9zvVfXM+KvSvANKm+O5NxX0BJf865dhM4fNh68v77cO65cMUVULo02JTLFBEREZHi6c8tf3Lpx5cSDgXJbOTFWR+6roZPx4PNxNoW8+abMHo0XH01jBpl3Vy89VbruWGc7C2kmFCXGZEY2ZG6g1AwE3e1WiSnhk44N2CAaUDYgINumHxnJ257aQo2Q4kRERERESk+QuEQKa+lsCt9V47xhEx4byL0XpI10LEj/PRT9AOUAkFdZkQKuIolKlL5rGrsG/MBqS5Ic0IAODZFGQKCNnCFIT4EFdKg1+s/8eCj58UgahERERGR2Jm/Yz4ZgYxc4+kuGN4064nbDZdeGt3ApFBSQkQkxup2uw3/+jXMHNCTiTc1J5DozjnBAPcxC0gSAtD98yUs37McQiHIyP1DQURERESkqAmb4eMfM4C4OKuwat++0QtKCi0lREQKgDIptbn05S/pMWYOrk8+A48H7HYAAsf5v7TGAfht8B0EPW5CJRLYVrkkCz8bFsWoRURERESiq1mlZrjsubvHJARt3LarotWM4K+/ICkpBtFJYaOEiEhBc9VVMGcO/PvfmB07WpnuY4SABRWh7I+/48gMYA9DyrbD1Lnlfu59pQN/bPmDolAfSERERESKkWDQapnbpw88/DCsWpVrisPm4MvrviTBmYDbYa2sTnAmcFGDLtw0aTMMHmw1IRDJAxVVFSngpt91Bc0/+I7EgPU8DGQ4oeuN8NMYq2Xv34IGjDkH7rrWTeNyjZl6y1SS4pQdFxEREZECLjMT2reHRYsgLc1qm+tyWR1irrsu1/Q96Xv4bOln7MnYQ4caHWhXrR2GusdIlrwWVVVCRKQQGDuwK+cNn0yFNJiTAsO6lGLwN2k02RrMNXduRWjRD1x2FzefczPDrxoeg4hFRESkyElLgz/+gIQEuPBCsGmxuZxBH3wA99+fuzZeYiLs3m0VShXJI3WZESlCbnxxEvV3h9i6cRE1Z6/iu6dW0nhX7mRmpg1mVc76fSiTsUvGRjlSERERKZJGjYLy5eHaa+Gyy6BKFVi8ONZRSWEVCsHkyfDggzB0KOzcCZ9+GrlRgM0Gs2ZFP0YpFhyxDkBE8sZm2Din/DlHnh++oQfxn35BwlFbaXwOeLVV9jmZoUxM09TyQRERETl9ixfDXXeB15s9lpoKnTrBtm3W1gaRPFq1bTFGu3ZU23SIuBAEnQ7szzyDcc45kU8Ih62GAyL5QCtERAqp5JGfsqz/NWxLgjQn/FgLWv4HNp1lHbcZNjrW7KhkiIiIiPwz778PPl/uca8Xfv45+vFIobXhwAaWXNaUOuutZAiAIxDESE+HFSsiJz5KloTmzaMbqBQbSoiIFFZ2Oy2GfUHS7sOMmzWc6/+TxMZK1t5Kj9PDWfFn8c4V7xz39L0Ze3ll5ivcPvF2Ri4YiTfgPe5cERERKaZ8Pvj6azhe3cEDB6IbjxRq737zJD2WhIh0u87M9EPPnhAfb9UNSUqCMmWsrTWqVyP5REVVRYqIg76DjF44mgU7F9CkQhP6nNeHUvGlIs5duHMhF314EYGgnxo7/Nw7C2odgJZnNSZx8k9QoUKUoxcREZEC6c47YfhwCARyH4uPh40brdoiRYlpQjjM3oy9ZK5YSsVSVTDq1j211/jmG3jqKdi0CRo2hJdegosuyp94C5Hne1Xi8XE7sEc4FnY4sC1YYCVCfvnFap3bubPVaUbkFOW1qKo2/IkUEaXiS3Ffy/vyNPfWr2/lsP8w3VbAJ1+BKwhOE/ybl0LFitCvHzz/PJQtm89Ri4iISIEVCsGHH0ZOhgAMHFhokiG+oI+FOxdSYfUOqh8IQ5MmULNmzkmHD0P79oT/mo9hwt5kePBSWFPGYMz8alww8gfIS2Lk44+tz1J/FwidNcsqRDt5MlxyyZm/uEKkvKccQdsO7OGc4yZgejzQqBEYBtxyS0zik+JHK0REipkD3gOUH1qecCDArlcgOcKWYMBKjCxebC1VFBERkeLH57Na7IbDuY/FxUWuKxJlWw5tYcvhLcTZ4/hl3VTs8xfQ7actfFrlIP+ruoNDpo8apWpwaMcGvhntp/6uMKbdRgJOjO5Xw5gx2UVhq1TB3Lo1x3aOdCd0uAWWlYM148pRYdVJisiaJqSkwI4duY81bw5z5pzR6y9sFsydSP1WV+EO5hw3AePLL60tMyJngFaIiEhEDpv1v32jPeCM8PnmiP37Ydgwa6XI/v0QDEK5ctEJUkRERGIvPh4aNIBly3KOGwZcfHH+v38wCAcPwllngd1OMBzk+RnP89bctzjoPUhSfBIZgQwMDPwhP7YwOMLwUDOwhbMXtizfu5zx46HxdogLA4Ew4Mf89luM116DRx6xVnEckwwBcAfg6RnQ83oYWeMgj//4I1x++fFjzsiAPXsiHzv2v2NR4PPBxImwbx+0a2dtDzqBJs2vZPHjd1B38P9hhMFuAnYb5qOP4lQyRGJA1WlEipkScSW4uPrFZLisDw3H5fdb+1/btLFWi1StCo0bw8KFUYtVREREYuzdd63OH/asqg9Op1Xw8tVX8+89TRNefhmSk6FyZWu16uuv029iP4b+MZT93v2ECXPQd5DMUCb+kB+AsA0yHRCyQeCo277uTLhiTVYy5ChGRga8k1WAftasiKHYsG4i+ZywtmQYdu06cexut7WqJpIqVfJw8YXIhx9afxeuuw7694dzz4U+fSKvKDrKOc++S9zq9QReeAZz0PM4lq3A+cLgqIQsciwlRESKodHdR2NWqcKqZAieqCvv2rXWB4TMTCtBsmyZVRBs376oxSoiIiIx1LYtzJtn1XRo3tyqjbF4sVXrIb+8+SY895xV08Pvh4MH2T34CcYu/IiMYMYpv1x88AQH09KsX1u1ing4DCwuD4l+aLfFgNatT/xmNpu14uTY9rEeD/z3v3mOucBbtAj+9S+rzszfgkEYOxY+++ykpxs1apA48L84Bz6Zt7osIvlECRGRYqhiiYqsfnAjaff3J90FoUhJkbg4a0ns0T/owFp/OmZMVOIUERGRAqBBAxg50qp/8b//QfXq+ft+gwZlFyTNsi7eS5w/dJwTTuyAB9ZHaLxn2u3Z219atIBq1XLN8TphUFson2FwfePr8/blfeBA61GihNUhJWuFCzfccFrxx5Rpwty5MH16zj+TRx6J3Io5GIS3345efCL/kBIiIsWUzbDR9rF3KJkexP78C1YBMJfLanVWogRcc03kJY9er7VyRERERORMM82INThqHQCf/RSaQRwz9d/dIM0J/qxvP4E4B0ZyMgw+aqvGsmXQpg2mYRAyYEl5g1uusdM6WIG557yFe/jovL23YcCTT8KBA7Bzp7XNpm/fvMdeUCxfbiW/2reH7t2tWnIff2wdW7Pm+OcdOhSV8ETOBBVVFSnu7HZ44gnrsXmztR2mYUOrVsjXX+een5h43GWlIiIiUoTt2WOtIE1Kyr/3MAyoVQvWrcsxXC4dem0pyed1MvEGvcc93RkEO9Biu425lcL4HTbqlalHo6at6VXzT/r8kU47bznKX9oDbr8dSpfOPjkhAX77DQPrNc4Gxv+Ta7HbrYKwhVEwSObF7XDs2ZfzDnrfvlatkAsugA0bcp0WsAG9b8QZrThF/iElREQkW9Wq1gOspaMXXggzZ1qrQsBaQVKpklqiiYiIFAPbDm5h8VtPcc5n06mwYQ/2zIBVI+OSS+Cjj/Kv+9zQodC7d84tGm43H1w7hkrxf/L23LdJ9adSL7kelUpUYr9vP43LNiYlKYUyjiR6plWhRnJtwi2aEzBDxDnirNe4Mn/CLWp2pO7g9y9e5aL0fRz7Jxz2+7G9/z489RTmhAn4g1ZB2/gQpLpgaTl4Mvk7fjIfw2ZoM4IUfIYZae9XIdOsWTNz3rx5sQ5DpOjx++GVV2DECKuwaq9e8NRTUCrCRtyjTZtmLRVdvdraj3vvvVYxNpt+MIqIiBRIGzfC/PlWJ5TmzflkySfE9e7DZSvDJBxblNTptGppLFlirejIDz/8YH2WWLMG6te36op06HDksGmaGPn13kXBjBnw1lvWqp6rr4b//Of43W+yHPId4vovr2fGphmEM/04gyYv/wR3zT1m4rXXwuef88cPw7n1u75cu8SkQhpMrQWT6oInPpHx142nc63O+Xd9IidhGMZ80zSbnXSeEiIickZNnmy1XzumGBqlS8OECVaPehEREYkt07SSGeGw9WX500+tlaChEMHqVbm2yVrGjAuQGDjO+YmJMGXKybuuSPS98Ya1Ffrvz2JuN9SoYRXFzciASZOs8a5doWzZI6dd9vFl/LzxZzJDmUfGPJkw/nPoklU+Ls0J8R+MxNHnNl6Z+QpP/PwEgXDuvyRPtn2S59s/n2+XKHIyeU2I6HatiJxZDz6YOxkCsH+/Vcl9x47oxyQiIiIs372MT+5ozb6SLkybjWCtmtC/P4wbBz6f1eY2PR1j5SpenRzEfbxkyN82boxG2HIqDh2Cxx/P+VnM67X+rPr1s7ZG33OP9ahaFUZbhWJ3pu1k+sbpOZIhABkuGJKV80p3wupyNuy9bgSgQmKF7O1IR/E4PFQsUTFfLk/kTFNCRETOHNO0tskcTzB45AfvlkNb+GHtD6w/sD5KwYmIiBQvmaFMDvkOYZomU9dPZfStTeg+8g+SDwcwAMf6DZgffJDrRoY9FKbqIRO//QQvHgrB+efna/xyGmbPtlb6HCsjA8aOtRJf6enWw+ezEmJbtrA7fTcue4TzgFXJML06DLzMyZSRT2DEWUmQHg164LDlLklpt9np1bjXmbwqkXyjoqoicuYYhlVgbdeuyMf9fkKbN3HL+N58tfIr4uxx+EN+OtTowBfXfoHb6Y5uvCIiIkVNZibp33zBnateZ1x4MWEDqpeqTpr3EEunB0g4ZtWHcZzt8wFbrs612dxuuOwyq7aHxMzS3UtZvW81jcs1pm5yXQD2xodJ9GcQH+kEw7BuXh0tHIYvv6TePXdiHudPfHdJB1f8x8k9Le7hsY7PHBlPcCUwvc90eo7ryc70nRgYlHaX5vNrP6e0u3TE1xIpaJQQEZEza+BA6+GN0BIvMZHx5fcxYeUkfEEfvqAPgGkbpvHADw/wbtd3oxysiIhI4RcMB9l2eBul96ZR4uJLuabDTqZXCeF3Aias2b+Gkl5yJUNOJFQmmeu6HOaTzwLYwxAXBFcYjAoVYMAAeOCBfLseObFUfypXfnolc7fPxWFzEAgF6FSzE29d/hbNZt/CTE+AGn6wH53fcDqt5MexQiHw+4lzxDGk4xAe+ukhMgLWiiGnzUmJuBJMvnEy55Y/N+KNq/MqnMfae9eyat8qwmaYBmUaqNitFCoqqioiZ5ZpwksvwdNPW1tk/hYfDw0aUPG6zez078t1mtvhJv3xdLxBLz+u+xF/0E+nWp10h0FEROR41qxh1OeP82BwMn47hDN9XL7KZHIdrGTIUYww7B0CpX0RXsfptB4ZGdm/nziRzefXZvyizyi7aA2tqram5uW9rWMSU7d9fRufLv0Uf8h/ZCzeEc955c/jr51/kbInkyljoHKqtcrHBtgeepi4YW/lvmHldsO8edCwIQA/rvuRITOHsC11G51qduLR1o+SkpQSvYsTOUPUZUZEYsvns9q9jR1rtey96Sa4917crycfWRlyNAODKb2ncPvonrzwnY+rlwYxge2dL6TehxMhOTn61yAiIlJQDR3KlJGP0/PqABlHlX5wBa0vwYEI68Dv+xMG/ZxzpUggzolz7GewaRNMnw61a8Ndd0GtWvl+CXLqwmYYzyBPjmTI3+yGnZAZIuUQ/DwaUlIBE+zA4e6XU67ueVYHGl/W57D4eLj/fqulsUgRo4SIiBRIncd0Zur6qbn2qZ5b/lw27FnD7GEZ1DgAcVmrOjNtQLWquFat1V0pEREpdoLhIN9Pep11347mnDWHuKTaxRg33wJXXUXbG3z8Xi3CSSZwzK4Fl91FOXdZrvptF09MD1Eu1WRvpVKUe3c0tiuvisalyBkQCAWIHxRP2My9/cVm2AibYX4fAS22gvOoj1qmw4FRsaLVfrdiRahSBa6/Hpqd9PuiSKGU14SIaoiISFS90eUNLhxxIb6gj8xQJk6bkzh7HNc0vIYlwwdTKTU7GQLWfmXfzh0wcSL06BG7wEVERKJsR+oOWr/VhL2pu/BXBFc5qLfvY6b3GEeJEGwqFfk8RxgcIfBlrRyxmQZuh5tZt88mrU8aSw9t4pzy51AhsUL0LkbOCKfdyfkVz2fu9rk5xm0YdDxcho7z9nP+tmCOZAiAEQzCli3Ww+OxVopkJUNM02TN/jXEO+KpWrJqtC5FpEBQ210RiaqGZRuytP9S7m1xL22rtqXf+f1YeMdCKiZWpOGOIJ4IBd/ivAFYsiR7YNUqa3nnCy/AihXRC15ERCSKbp94O1t8u0iNg0wHpMXB0nLwZJsABIO03gy2CHUyPUF45heoesigZKaNq2pdztzb55KSlEK9MvXoXKuzkiEFzKq9q5i8ejKbDm466dz3ur5HCVeJI21y44MGJb0mb4/azYA/ct5YiigjAx58EHw+ft30K1XfqErT95pS7616nPd/57Fu/7ozcEUihYO2zIhIgbD18FYe7VuTdycESMrMeSyY4MYxfCT06gVDh1oFWwMBq4CrYRC+qB2BlwYT1+yC2AQvIiJyunbsgOHDrWR/69Zw882QmEhmKJPEwQkEwsFcp5TOgH1DYFUyNO8L6U4IZ93m9NjjGVKiB3ftqgbnnw9XXaUtpzG2aOcixi4ZSzAc5NpG19Kycssjx9Iz0+n+WXdmbpmJy+7CH/LTvV53Prr6I5w2B0ybBh99ZH3muekm6NwZDIOth7fy9py3WfLLOC74YxP95oQpl34KQSUlsXviOGr+fg3pgewTbYaNiokV2Xj/Rhw2bSaQwks1RESk0Hlx2rP07vksFQ+bR5Z6hmxgq5iCsW4dbNsGjRplFwPLYgI+B0xoW4Z6Y77n/BTthxURkUJg3jy45BIrye/3W1sZkpNh3jx8pZNIHJxIyAzlOq2kDw4OLwNpaaw8K8RT7UL8UTlMpVJVePKa/9GtfrcYXIxEMvi3wQz6dRC+kA/TNHE73fRt2pfXu7wOwL+++Rdjl4zNUSTV4/DwaJtHefrLPTBqFKRnJSwSEuDGG+H997PfoGpVaxtMJA5Hzo5/R4uPZ9joO3l09du5CrSWcJVg3DXjuKzOZad93SKxlteEiLbMiEiBMbDDf9k79VtWNKtKyGYQttuwdbkMY9YsiIuDr7+27pAcwwDcQbjq970MfLYtnyz+hE8Wf8LGgxujfg0iIiJ51qcPpKVZyRCwtjLs3AlPPUW8I57WVVpjM3NWR3UGocdqu7V1dPly6j/0El9cMJRtd29g7pOblQwpQDYc2MDzvz5PRjCDsBnGxCQjkMH7f73P/O3zCYaDuZIhABnBDH7+dhiMGJGdDAHr9598An/9lT0WFxf5zZ1O6N8f2rTJvULI6YTmzVkSfyhit5qQGWJ76vbTvWyRQkUJEREpUJo268o5czZhzwxg82diTP4OKle2DtpO/E/WthIwt7SPvl/ewh2f3USDoTW4q19lzOXLoxC5iIhINn/Qz4yNM5i1dRahcO5VHuzbB2vX5h4PBGDCBABGdBtBaU9pEsLW1oXETKicZvByg3vg9tutjiEPPAADBkD16vl4NXI6Jq+ZHHHcG/Ayfvl4guEgwQhbogDaLEuNvLrD54Pvvst+3rcvuN0559jt0KIFvPkm/PYbvPSSNScpyfq1eXP46isuqX4Jic4EHMf89TRNkwurXHgqlypSaGljmIgUTHZ77rEePWDgwIjTTaD7DXAoDkxbGLJOH11mG+1ubcb136yz2syJiIjkE9+hfWTOm81c33p6LH4CDOvLZYIrgYk3TKRZpaNWb7tcEVc9Ake+4NYuXZv1921g3LJxrNzyF03tVejZ+nbizioThauRf8ppc2KLcP/ZxOTFmS8yee1kapSqwdoDORNjBgYpleqDY62VIDuaywWJidnP778ffv8dpk61ntvt1rarTz/NnvPAA1biZPFiKF8eatWCYJDrX11Ct9e9ePywOhnuvhz+rO/hyrpX0rBswzP1n0GkQFMNEREpXN57D+67D9Pv5+hFxCvLQNO+4HXlPqXtZoNfKz0Bzz/Pij0r2Jm2kyYVm1Aq/jj9CkVERPIoGA7iC/gYf28Hrhk1h4ANnGFYXRq69obtSda8UvGl2P7AdtzOo+7md+liFc08eiWA2w1PPGE9pPBauZKMVwazYNon/J4SZlhL2JGUe1q8Ix4Dg2A4SCAcIM4eh9vpZn73KdRscgl4vTlPcLthXYSbPIsWwdy5UKUKdOwY+cbS0e68E0aPtrZpZfG6DL5//xG63TwIu+0k54sUcCqqKiJF15Yt8N//EvzoQwKGiS0M8ypBpz7gjVBIv8l2mLKjPVd0Pczyvctx2pz4Q36eaPsET7Z7Mvrxi4hIoTd762z6T+7Pwp0LuWgTTPrYJOGom/kBA5aVgyb9reclXCUY2W0k1zS8JnvSrl1w8cWwdav1PBSyvsyOH6/OMIVVRgbcd59V/yPre5bPButLQ+O74JiSMNgNO93rd6dyUmWW7FpCy8otubvF3VQsUREmTbI67P2d3AgGYcwYa8XsP3HoEFSokKtIPYYBXbvCt9/+s9cXKQDymhDRlhkRKXyqVIGRIwm/8CzTXu7LvHW/823tIGFnEMi539YdgF4r7fS4eDULd+3MsVf3pd9f4uxyZ6sAnYiInFh6OmzfDikp4PGwdv9aOnzU4Ui70nv+tH7eHM1pQp390HA3LC8HoXCIfRn7ck4qXx6WL7fqPGzcCE2bQuPG0bkmOfOCQWjXzip6etRN5/gw7EmExJCNVEc4xykhM8Smg5v48rovc79e165W0mzqVOv1OnbMuV3mdG3ZYiXcjk2ImCYsW/bPX1+kEFFCREQKLVelKnQd9j1dgWeASasncf3H3QiYYQIOSMiEmvvhqk3x/Ne5h2AoZ7IkPZDO67Nep3xieYbMHELIDDGg5QAurn5xDK5GREQKnHAYHnsM3nrLuksfCsH99zP0wv34g9ndOSqkRe5UELBBctaOhDBh2tdon3uSYVhfotu1y59rkOiZNAlWroxYG6b+HsgknGvcYXPQPKX58V8zIQG6neEbN9WrRy7YarNZSTmRYkRdZkSkyOhatyuL+i3kfv953LDczjvfGcxd0Qb/x6Nx2iMUFwEW7FjAhSMuZMLKCXy76lsuGX0JbUe1ZeKqiWw7vC3KVyAiIjH3+edw3nnW6o2zz4b//c+q45CWZv06bBiLF04haGZ/oZxYFzIi3GZ0hWBBRUhwJnDbebdRJ7lOFC9Eomne9nkMnf06H9ZJ53CETrjl0+GmtQl4nJ4c4/GOeB5q9VCUosySmAj33AOenLEQHw9P/T979x0dVdX1cfx7p2YyCaEjHURpNnoRUKqiIoKIFfuLooIFC/rYwIr6KKiPWLAhKqBSlaKIKE2kIyBFpPeaQMpk2n3/OEDaBFBJaL/PWrMyc1vOgUkys+fsvZ8u3LGIHGeqISIip03o+FkAACAASURBVC7bBssiGAlS+rXSpGSm5Njtslw5XtBmVzoYx+3zQ5xzVlNuen0yDm9cYYxYREQK0eZ9m0kPpVOteDUclgNeew369s1RaDKWB6728W4dUwQToEgAFr4HZVPBF4YopqbVB12q8GOH2txV/y461uiIZVmHva6c+A6+dzr4fxm1o9ww8ga+W/UdoVAm3swIDhu+/xyabMp2osdD5IXn6d84xNtz3iY5kMyFFS9kYPuBnF/m/OMxERg40Dzn9+yBunXN48aNC38sIgVARVVFRLIZ+vtQenzXg4xQBjY2cS4T4AiEA3kPtqHVGpgy9MBDh4Xjy2Fw3XWFOGIRESkoG1M20vXrrizatginw0mxuGJ8dvlgWjfoauqFHMH64g7Oe8TP/uD+Q9tKZrq4b65Fh5VR7NKlKPvUq1TodHNBTkMKWEoghSd/epLxq8bjc/vwOr0s2bEEy7K4svqVvHP5O0xZO4Ue3/U4VE/moDP2w+Y3wHHwrdZtt8GHHx65+4uIHBMKiIiI5PLbpt8YMHsAG1M20v6s9szdMpdvV32b90AbWq6FqZ/l2u73m+XSzZrBO++YpdQiInLCC0VCTPhzAhtSNlC/XH1uHn0z65PXE7Ejh46Jd/pY+p6DqpuPHBChbl2WTvqMByY9wMwNM0n0JnJfw/t46qKncDlUou9kti55He/OfZfZm2YzY+MMonbeuh9gan+UTyxP5aTKTNswLc/+xEz4aZiHBnZZ+Owz1YgRKWQKiIiIHMHsTbNp+lHTvDtsGDscOq6MfZ4NhPxxPDfoOkrXqMfN599MMV+xAh2riIj8Tb//Dl98QUrKTm5zjWNKuUyC0RAWFqFoKEcwBMDtcNN7RoT+P8R+A3xIfDxMmAAXX1yAg5fjYcaGGbT/vD3BSPBQOtThJHgSqFCkAit2rcizL9Hl58dLPqdRw04FMVQROQK13RUROYImFZrQpVYXRi4fmbXxQLpMfsEQAAuIBgIUf3cI8ysNY3X4cXo+OY7q9doW+JhFRCQfgQCMGgXLl8PateZ+MEhCNMJQF3xxHvS4EvNLPIZQNMTa8yrB9J1mNeBBXi+cfz5s22ZWBvbtCw0P0xVETh4Hao2Zuza3j7k9T+rL4aQF06hZoiYbUzbmOc/r8VG//pXHdLgicuwpICIip7Vvrv2GKWum8N9Z/2VH+g7qj1/E++OO8OkgEBeB+3+D9AVBHDa4v70EHn8KnnuuEEYtIiLJGcm8MfsN/tj5Bxe7z+beXp/hTNlnusFk4wQSQnDTEhh6AcysHPt6frefNtc8Bnvmw7Bh5o1yQgIMGAA33VTwE5LCM3Uq3H8/LFsGSUnw4INsf/D/2Lhv49+6TIIngRvPu5FAOMCMjTNIDabic/lwWA6+7vo1TofqhYic6JQyIyKSzcpNi6lUvSFxGaH8PkQETNpM7v12fDzWpEnQokUBjlBERAbPH0yP8T0O1XcYNRw6rAT3YV7WRoA3m8DD7c1jCwsbc4LX6aVCkQr8fs/vpi1qWhokJ8MZZ6gI5qlm3jyT7pS9k1B8PJm33UyRsp8QjASP6jIuh4tKRSqxoucKXA4XP6/7mZ/W/kQpfyluOPcGSvlLFdAERORoKGVGROQfqFHhAti+G3r2hG+/BY8HihSBjRvNcmxMO8WYwZKMDPj44xwBkbV717IjbQfnlTnPvMgWEZF/ZemOpdw34b6sYpc2XLHq8MEQgKgFgQOvfD0OD62qtGJb2jbSgml0qd2FPs36ZP2e9vvNTU5uGzfCmDEmNaZTJ6hUyazkzJ4SBZCejvfjIXR46yK+3fZLvvVDXJYLp8OJZVl0rtmZN9u/idvpBqBV1Va0qtqqoGckIseYAiIiIrklJsKQIVmPAwH4z3/gww+JpqWyvCRU2muTmOv1kmXbh15k7UrfRecRnZm/ZT5up5twNMz9je7n/sb3UzaxbCFORkTkJDRtGrz/PqSkQNeucMMNJkANfLjgwzxvWO3DLek7IOSEL843aQ6Vkirx1bVfUcRbpCBGLyeC996Dhx7KetynD7z6qkmTibVC3u1mcJ1naDW/J3/t+QuH5SAjnEG8K54y/jJcf971PHXRU3icnsKbg4gUOKXMiIj8DRmhDK7+oA2jHvwVXzjXzoQEGDoUOnXi4k8v5teNv8b8lKmotyhjbxjLRZXVgk9EJGpH2Zuxl6S4JNOy9uWX4YUXTIDZts1Kjbp14aefwO3mltG3MPT3oTmuMfwr6LwCPLlLQLlc4PViR8L80uNyxlxaiQsrXkinmp30xvYUs2LXCtYnr+f8MudTdk8QatY8tLLzkLg4s4rzxx/zBkV8Pti+HTshgblb5rIueR31y9anWvFqhTcJETlmlDIjIlIAfG4fE+6dyerAU1R58jWckShWOILl90ObNtCxIxtSNjBn85x8l9wmZybTakgr1j6wlkpJlQp5BiIiJ47357/Pk1OeJDWYisfpofd5d/HMc//DEcjMOigtDRYuhJEj4frr6VijI18t+4rMSNYxvS6HeluhbLqFP+LE8vmgVCl45BHwerEuu4yWZcvSsvCnKAUsJZDClcOuZN6WeXicHgLhAEPX1eMa286b3hqNQq1aMHNmnhoi3HMPJCZiAY3KN6JR+UaFOAsROV4UEBER+Zssy+Lsh1+EDrfAZ5+ZJd1XXQVt24JlsTNtJx6HhwCBfK8RtaM8O/VZPun0CQCZ4Uz2ZOyhtL+0qtKLyGlhxNIR9P6+N+kh88Y0M5LJawv+h9XC5tnJuQ5OSzNtdK+/nk41O9G0QlOmb5hOxI4AsDMB6j7oZWnlV0jYmgG1a8Pll5sVInJKu2PsHfy2+TeCkSAZYZO2umDLPDpH7bxvdGwbypaF77+HBx+ExYuheHHo3RsefbTQxy4ix59SZkREjrGMUAal/1ua1GDqYY+7oMwFzL9rPn1+7MO7c98lakdxWA4almtI7wt707FGx0IasYhI4av5v5qs3L0yz/bETNjbH5zZX6I6nXDXXTBoEAChSIgRy0bwwfwPSAul0eHsDjzR4gniXHGFNHo5EaQGUynxaok8nWHO3APLBlnEhWOkxSxcCDVqFOIoReR4ONqUGQVEREQKwHvz3uPhHx4+9MlnLD0a9CDBk8CguYNiHlfGX4a/7v8Lv0edDkTk1JP4UiKpobyBY3cEdr8Cidnf48bHmzSHOnUKb4Bywtueup3KAyvnSJ86qO8cP89OjULoQPqq2w1PPWWKpIvIKU81REREjqMeDXpQo0QNXpr+ElPWTsEmZ/DZ7XDTu0lv6rxfJ9+gyfa07XQc3pEpt0wpjCGLiBwV27aZvGYyW/Zt4ZK0MpSbthCSkuDaa03djqN0bplzmb1pdp7tJeJLkFDMCekZYFnmDe2AAQqGnA5sG374wbSwz8yEm26Cq682K4RiKO0vTdmEsqxLWZdju9Ny8tf/XQ1vPAnffGOu26WLqR8iIpKNVoiIiBSw9cnr6TSiE4u3LcbGpkaJGgy/ZjjlE8tTcUDFmJ9sHeS0nISfyd3ORkTk+JixfgbthrYjEAmADRZw3TL4cnwcluUwdT4uvfSorjV9/XQu/fzSQ3UfAOLd8bx3xXvcfO6NZkVIaio0bw5F1B73VBaJRnBYDqzevWHwYFMzBkyHobZtYfRoExyLYcqaKXQc3pHMcCYRO4LX6SXBk8CCuxeocLnIaUwpMyIiJyDbtrEOvKiL2lHKvFaGXRm78j3ewiL6bO4+ksDu3fDzz6bVb+vWZimwiMgxtiNtB9PWT6NoXFEalW9EiVdLEI7mCtLaMGg83DMPE7jYsQO83qO6/owNM3j8x8dZtmMZlYpW4rmWz3FVzauO/UTkhPTbpt+4Z/w9LNq2iDinlzt/C/HapAhx2Z9iCQkwdqz5W5eP5TuXM3D2QFbsXsHFlS+mZ6OelPaXLvgJiMgJSwEREZGTwKcLP+W+ifflmzZzVrGz+PP+P3NufOst6NPHBEEsy3ydOBEaNiyEEYvIKW//fvj0U5aP+oBx0eV82iSOzUUd2Nj5FouukAIbB2ACIt98A+3aFe6Y5eQQicDevZCUxJ/71lH3/bqkhdIO7Y4LwWWrYdSIXOf17g2vv164YxWRk5pqiIiInARuq3sbxeOL88DEB/LkQDssB19f+3XOE+bNgyeegEDA3A5q3x5761YW7l7KuuR11D2jLlWLVS34CYjIqWXHDqhfn8juXdTKCFDVCffOTuPSbvDrYbIPUj2FN0Q5Sb33Hjz5pEmHcbtZ2aE6wRoBk3d1QMANE8+CDUlQKeXARo8HihU7LkMWkVOf43gPQETkdNexRkfWPriW+d3nc3Gli6lYpCJdanVhw4MbqHNGriKCH36YMxBywG5XkIZvnstFn1zE7WNvp/ag2nQb1Y1INFJIsxCRk4Zts3XQq0xrWJrvL/Dz/kMXsXLbMrPvmWdg+3acGeb3TFzEdHv5dAyQ36JiGy5al+3xRRcV4ODlpDR8ODz8MOzZY4qlpqbSZtQi+kzL+zfKG4Y/i2fb4HRCt26FN1YROa1ohYiIyAmiXrl6/Hz7z4c/KDkZonlritx2aQa/p68hRNaLy9HLR1O3bF0ebvrwMR6piJy0IhG2NK5N0cWruDACLhtSl09n7rjz2P3jz1w4dmxWm9JsKqVA6TTYleAgSrbfQTa4o/D+j16Id8JXXx11/RA5jfTtC+k5U0N9wSiPzoKXWkA020e0ATfUzEyAIg6TYvPpp1ClSmGOVkROI1ohIiJyMunSxVTdzybVAz9UjuQIhgCkh9MZNGdQYY5ORE4QwUiQIYuG0OHLDnQb1Y3p66ebHUOGUGLRKuLDJhgCkBCChhttxj7fDeLjY17PAgIu8Dg99GjQgyLeIvhcPtqVbMQ6Tx/O6Pc6rFt31B1m5DSzaVPMzXEhiM8Wf/O5fFxb50bKDxltVpXs2AHXXFNIgxSR05FWiIiInEw6d4b334fZs00etsNBRqIXyxUE8i493h/cn+PxuJXjuGPsHezO2I3DctCmShvG3TiOOFdcIU1ARAqEbR9qSxqKhGg1pBWLty0mLZSGhcXoFaN55uJn6DPoawJO8Ob6dZEQgkazN5LZ42W8fZ/P8Wl+0AG/VIFwQjxvXTaQ7vW68+4V7xbe3OTkd9555u9WLlbJkjSqfh4zN80i0ZvIfQ3v46mLngKH3qKISOHQbxsRkZOJywWTJsHo0fD111C0KCXvvJOKc7qxes/qnIdaLjpU73Do8cwNM7lqeFY7y6gdZfLaydR+pzZrHlhTaFMQkWNj5oaZvDfiEfYuX8i1CzK5PrkCnhde5usLHIeCIQA2NumhdPr+3JeHHOdkr2F5SARI9zpwP9gb5i6A777DdjkJR8JsK+7hx76dmdHmAeqWrVuoc5RTxKuvmtVDGRlZ2+Ljcb8+gCmqDyIix5Ha7oqInAJmbJhB+8/bE4wECUVD+Fw+iniLsODuBZRLLAdAnffqsHj74pjn/3rHrzSp2KQwhywif0dqKum/zWTYlu/5KDiblEAKf+1aRTAaxnaAPwgXbIOpX/m4qe/5fJP6W55LJHoSmRR/F3UefT1HmgJAmhsGPd+RR/uMNRtWrTJdrapUgaZND60+EfnHZs6Exx+HpUuhcmV47jno2PF4j0pETlFquysichppXqk5S+5ZwqC5g1ixawUtKrege73uFPNltSrMvYIku6nrp8YOiGzeDCtWwNlnQ6XD9NwUkYLz7rvYDz9MxA5wfcSmYTG44ibITOJQNbg0Dyw+A76umkGJ+X/gqOkgauctwJx2xSV4flpJ6sTviFqmNogrAt9cVpmevYdnHVi9urmJHCvNmsH06cd7FCIiOWiFiIjIaaLm/2qycvfKmPu+7/Y9l1S7JGtDKAS33QYjR0JcnGmT2L49DBtmHovIsZeRAZ98Yn7uiheHe+81aXLt2+eo6RG24I9ScME9kDv/5eo/4MlZTpr38JARzsixr7S/NJt7b8blcMGCBex+579sZT9JPR+mYt2WBT8/ERGRQqIVIiIiksPA9gO57IvL8mwv7itOuzPb5dz4/POmTklmprkBfP89PPYYvPVWIYxW5DQTCJhP0FeuzAp+TJgAVavmaVfqsuHMvXDuDlhaJmu7IwrFMqCe70wGtH+YhyY9hNvpxrZt/B4/k26aZIIhAPXqUeKjLylRSNMTERE5EWmFiIjIaeS9ee/x4KQHyYyYIEf14tWZdvs0yiSUyXlgyZKwe3feC/h8pruNZUE0CpMnw7RpULYs3HADlNDbK5F/5OOP4f77zc9XdpZlOsjkkuyFztfDz1WztsUH4efhXhq+8jlccw0pgRRmbJhBojeRZhWb4XQ4C3gScsJLTYW9e6FcOXA6zeMhQ2DqVDjzTOjRw3wVETnJHe0KEQVEREROQ8kZyXhdXnxuX+wDvF4IBvNutyyTThMOQ7t2sHAhpKYScjnItKJ0ucGieO0GDOo1iWLxxQt2EiKnko4d4dtv8273ek1AJNfPY7oLyjwK6V6LhCAELZvXF5Tk3jvehWuuKaRBy0lj8WK4807z1ek0qY99+8Kbb8LOnSYQ53ab27ffQuvWx3vEIiL/ilJmREQkX0V9RQ9/QPPm8NNPebfXq2deTL/9Nsyff2gpvzscxQ18OMamUpW5jH2lNFuf2E1SXNKxH7zICShqR5m1cRZLti9h1PJRzNw4kwRPAj0a9ODpi57G7XQf/gKlSoHDYVZeZed2Q5EisGcPBAJEgQw39GvnpmL5s3j24meJd8dzUeWL9PN2qpkxA/77X9iwAdq2hYcfhjLZVvMFgzBmDMydC9WqmVV6STGeA337mjTIg8+tcNikQj70UM7nXChkbrfear6nOguJyGlAK0RERCSvpUvhwgtNXYNQyBR29HphyhRo3Bjq1oVFi/KcluqGxt1Nwcfba1zHxzcMj3FxkVOEbcPgwQRfeI7I1i38XtbioXZRfs3WkMnn8tGhege+6vrV4a81bx5cfHHOeiGWBeXLw++/w/vvw9ixhEuV4I8b2uJo05ZzSp2DpTetp6ahQ036ysHng8djgh2LFpl0l717TTvkzZtN2ovfb46ZMQNq12Z/5n4+WfQJ634aTf++0/EEI0f/vePjzd+AqlWPfKyIyAnqaFeIOApjMCIicnKYtXEWo5ePJrPm2bBkCdxzjwmMdO9uXog3bmwOdMauRWABEcvcmbDm+6wdtg0ffGBqk1iW+dS7Xj2YObPA5yRyLGSGM/O2se3fH/uhh/Bs3IwvbNN4Y5TJQ6H+5qxDMsIZfLvqW9Ylrzv8N2jQAAYONG9GixSBhATT6nryZChWDB5/HH79Fde47zj/hgc5t/S5CoacKpKTTZrK1Klm9UYoZOrJZA+OBYPmuJdfNo+ffRbWrjXBEDApL8nJcOut7EzbybnvnssTU56gzHc/4/g7wRCASMQ8D0VETgNKmREREeZsnkPrIa1JC5mCjhYWfZr14eU334x9wp13wvLlOV6wR4FtCbCyJGBDkrdI1vEDBkCfPubFPkA4THThQrZe1pxW9/kpXqYKj3ta06nhzeaNod7oyYngr7/Y8OyD7Jz1IzNLB3jvong6XfoAz7V6DlcoAi+9hJWrA4wvBM9NhSu6ZW3zOr0s27GMKkWrHP77de9u0h5mzzZBkYYN9bNwqnv3Xejd26zuAFPbY9CgrN+V2YVCMGmSuf/VV3nrPNk2LF7M6xOeZuv+rYSiIQ7Ep4+e02med2XKHPlYEZFTgFaIiIic5iLRCM0/bn4oGAJgY9N/Zn/Grhgb+6Tu3aF1a2yfj0wH7PNAchxcfR2HXn0/1e4FcyccNjnsuV7gO4CkADT/I43fUpZx0+a3GfBIM1O/5OCnniKFIBwNs3zncrbu35q1cd48IuefxxnDvqP+2gA95sLsN9OZMnYAD0x8ALZujdn9xQHU2ZZzWzAS5OwSZx/dYBISTL2IRo0UDDnVzZ9v6oIEArBvn7nt2AF3322CH7GULm2+OvJ/CT9u9XeEoub8EedC4Ajla3A6ITHRPPeqVYMRI/7BZERETk4KiIiInOaGLB5y6MVzbv+Z8p/YJ7lcMG4c1tSprLnhEnpeDhUfgt/LADZ0q34NN19wszn2QDHIWBJC0H61uZ/ugaebh8hYPA8eecRs3LcP3noLunaFp59m9+olpAXTYl5L5GilBFKYMXcUOx+5l69vvIAy/RJp9EEDznzrTFoNacXOtJ1w77040zPwHMiS8UQhIQivjAvw8aKP2ZcUl7cA6gGrsnWf9jq9tKjcguolqhfCzOSkMniwKW6aWzAI55yTtWrkoPh4ePRRc//WW81qkuycTmjaFKtIVmHVBeXgzcamK1EEyBHCc7tNGuO0aWYsEyfCihWmRomIyGlCKTMiIqe5P3f/me++Hek78j/RsqBxY2o1/p5Pw2FaTn2DndFUbrnwHsomls06rlgxE0CJ8Yln0AEbszVFcNiw2h/kvM8/N6tK6teH5GRmF0/n9jNgzWcvgMtN++qX8e7l77IjfQcl40tSoUiFfzBzOa0cqGPT/9s+DKmcwsyPYXkJuO1WSHcABxYwzdwwkyu+vII5MYq1O4ALN4Lb4WZTeA+1773XpDxkS5tJc8PzLS3Axuv0cvP5NzOw/cBCmaIcR5EIDBliaiUFg9CtG9x7b96gRXZ79uQbVKNXL3O92bOz2qA//TR06mT2P/00/PKLqfUUDJpjkpJg6FDu2/4dj05+lPSQeV4+2RZGXOCk5/ZKdD/jCli1yrTabdPGrFA544xj/I8hInLyUEBEROQ017V2V/rP7B9zX/OKzY/qGg6XizvaPRZ7p9ttCkI++2yeXSEnvF8/63HQCWXSMJ+aPvkk7NjBen+YdjdDqvfAQXaI8avGU3FVRfxuP6FoiMblGzPy2pGUiC+R53vIaWbvXvjhB/Np+aWXmlQAgJdeYvzw53m+YyZDR0FSJgxqDIFcr4RC0RDLdi7jj4px1N6QkefyqR6TYlM5qTK88orp7jFgAHZaGrtK+HivWw1u6HYPo8+5lgRPAg5Li3FPCzfdZAqjHgyOrVgBX39tur7kU4Sazp1hwgRTEDW7YBCuuAJuuw3WrTPpWeeea57Ls2fDY4/BwoVQtqwpvlqqFFSpYs5xu+lRsQdzNs9hxLIRuB1ubGzCtStxZf8pkKDgh4hIdmq7KyIi1H+/Pgu2LcixzWW5WPfgOsoXKf/vv4Ftw3/+Y95A2jZhC1K8cFtn+K6GOcQbgstXw6ivHdC+PcydCzt30qctDGwCwcOE8N0ON40rNGb67dNNrZKRI2H0aChe3NQ7qVv3389BTnyffWbqL7gPFE2IREzxybZtoWRJ2nZOZUo1SHkJigShxe0wo3LeyyR5kxi5qzWt3pmAI5CV0pDugv81c7HvmT680PqFrBOi0axP6VX34/SzeLHpxpWrwC4JCfD553DVVbHPC4XMKo0FC0xQxLLA5zOr4w6mxmQXqzVzfDy8+CI8+GCew9fuXcu8LfOomFSRxuUbqyuRiJxW1HZXRESO2tzuc+nVqBcJngQ8Dg9NKzRlZa+VxyYYAuaF/ssvmzeO06bheqwP3/2nCzNq+/EHwRuGK1fB0Ek+k2Lz1luH2j6uKHn4YAiYT/Xnb5nPmh0roXVr0wVnxAizfL15c/M1+/GRECmBFE6FDwVOJ9tTt/PAxAc4662zaDS4EcOWDMv6P1y7Fnr0MPVq9u83t/R0U39m+XKwbXb6zaFpB0oztFsDcTHK52RGMqnz1P9wXH4FUa+H/T4nGS744Vwf/hdf4/lWz+c8weEwqRF6w3l6mjEjdupLair89FP+57ndMGUKvP8+dOxo0mx++IHMh+5nyKIhXP/N9TzywyOs2r3KHP/UU3mDLunpZvVdjJTEqsWq0vWcrjSp0ETBEBGRfChlRkREcDgcvHXZW7x12VsF/81atIAWLbgVuDESYt36RZQcOZFinlXwdF244w4TFOnRA557juYbMphcDTKO0CnB7XSzbewXnHnw01YwKwTS082np9dfT8jv47HJj/HBnEGEIiHKpFu8tashnR/7BGrVKvCpy1EKhUxQbNAgky4QDhOuVJG+jXfxUc0MQtEQf+39i+7fdmfJjiW81OYl+PJL8/+dm8MBv/4K0ShXrjJtod9pCE9Mh/vmwHsNYJcFoQOviPxuPw82eZASRcvByJE4NmwgcdUqqFmTThVUq0ZiKFPGBDdyF4+Oi4PyRwgqu90m3eammwBIC6bRdHBD1uxdQ1ooDZfDxbvz3mVYl2F0XLQo9jXCYdi+HfT8FBH525QyIyIiJ6ZQCG64gb1TxlOreya742zC+aTiA/hcPnbMbUnC2Il5dxYpAl98wd32t3y+4BPSyfo0NT4I346OY/oL3Rn451D2Z+4n0ZPIXfXv4umLnybBk1AAkxMyMiAlxbQRzd5CdN8+Uy9h48Y8p6S54bG2pvbHQXGuODY+tJGSL7wB/fvnbYUbF2dStTZtYvdH/6PurRns9cKH30LHlbApEd6+0MHEczyUOPNcejd/lK61u+oT9dNUJBph+NLhDFk8BIfl4M66d9KldpfD14IJBKBiRdi1K+d2vx9Wr/5bRUtfm/kaz/78LBnhnPVrisUVY/u46rhn/Zb3JJ8Pdu82X0VEBDj6lJnjskLEsqyuQF+gFtDItu152fY9AdyJ6Q52v23b3x+PMYqIyHHmdsM331BsxQrmz5nCU5kT+Hbvb/hcPpIzkwmGgwSjQQDi3fE81/I5EpYvNGkLud8U2zb74iw+mz2EADmXlqd74PorAuxb9D8ynea85MxkXp31KqNWjGJxj8UEwgE+WvgRe3/5gVZbvTSq3Y6k+hdCzZpZRTvlyJKTzbL/jz82ARGXKytF6vrrzTH9+sUMhgD4Q/D8VHi/AUQOBMe8Ti+Lti2ibceO5jq5C1Ralik2WbUqJfx+Fr/zOu/U2s9bF3uZ0fk87qvSlbeaXGHanMpp7As5iQAAIABJREFUzbZtunzVhR/X/EhayDyPZmyYwdiVY/n86s/zPzEuDqZOJdrpKhzbtpvnnN8Pw4f/7Q4uI5aNyBMMAVPId/HD3WiwaEneGiL33KNgiIjIP3S8UmaWAlcD72ffaFlWbeB64BygHPCjZVnVbduOsQZWREROCzVrUr5mTT7hvkObtqdu59WZrzJx9UTK+MvwyIWPcEX1KyAy3RRTzZ1n7/Ox9fyquH6LvcRklx9sK++KyXXJ63hj1hv8b9ZAPhyyl1ZronjD4LTHYwOWxwMPPADPPw/jx5tPg5OSTFpQrVqnfk2JcNis7ohEYOhQ0zFj3z7TJejMM03aU+MDyzmCQWja1LT8PFhvIRw27T/vvNOkHbRqBV98cdhvGReGEhmw48DCnVA0RLnEclC1lgmqDB+es0DlQw9BtWrm4Gefpdgzz/BUKMRTHk8B/aPIyWr6huk5giEAaaE0Rq8YzYKtC6hXtl7M82ZumEnPmT1ZfPMaEl1+7jvzOvpd8w5utzfm8Yfj9/hjbo/YEfwt2sBHH5lWuTt3gsdjusw8/3zMc0RE5MiOS0DEtu3lQKzlqFcBw23bzgTWWpa1GmgE/Fq4IxQRkRNZmYQyvH7p67x+6es5d7RoYVYg9Otn3ixYlvk6aRKVilUlVpKoFQWnTcx0nHA0zNtz3+ba6btptcasUMghGIS33zZFEYPBrBoCDgdUr26CJGeeabZFIvD99yZoUqUKFC9OZvkz+NMfoErRKidEao5t2+wN7MXv9uN1eQluXMekHwaxOx4ubtedM0uezaZ9mxj46T10HjiJJutMQMSyLBzhXJ9dTJtmOry8+ircdx+MGWNWfsQqPpmebjpltGplVo0cRtSCvXHmvtvh5pxS51C7VG2zYfDgrKCI2w0332y6f2R38DkhksuUNVNyBEMOCkVCTFkzJWZAZOmOpVzy+SWkh0wQdl84jYFrvmTbhCgfX/Vx1oGZmfDHH1CypEmvyUfPhj2Zv2V+jnFYWFRKqkTNkjXh+lpw3XUm8Oj3H/HnRUREDu9E+y1aHpid7fGmA9tERESOzhNPmBUHP/9sVmu0bg1uNz7giRb/4aUpz5LuOPDm3Ya4CNjEDogA7E7fzf/NjxEMOSgQyFtMMRqFFStMu9fVq82nuc2amcKH6enY0SiZLpPZ81c1aNIFuv4BH0xy4a5RG/r3Z1vzuvT5sQ/T/5zMGftt+qY3ocFVPfjGv57kQDLtzmxH3bK52glv3gzffAO//25SeRo2ZGObhry68G1+3fgrtUrW4tFmj1I+sTwvj32E0Uu/JnF/Jj23VKTOpbfSNf1TtuzbggW88Uclnq20mkynCUJE/niNbtW78vOan5j9ym6KZoATIBIlZOXTti493bQP7dbNtFHOnc6S3Zo15uvdd5uuGTFqnIU9Lj5s5sbjc+CIhmlWqRkjrhmRdYBlmX/ztm3z/z4i+SgRX4I4VxyBcM6fZ4/TQ3Ff8ZjnvDz95TzHZ4QzGLZkGK+0fYVS/lImRezBB83zMxiERo1Ma/CSJfNc75ra1zBjwww+WPABboepJJ3kTWLc9eOyPki0LPO7TURE/rUCK6pqWdaPQKzEySdt2x574JifgUcO1hCxLOsd4Ffbtj8/8PgjYIJt2yNjXP8u4C6ASpUq1V+/fn2BzENERE4dtm3z6dzBvDTuEXZE99Noi4NX5hTh/ltLMzOwKs/xca44XJaL3wakUntXjAseSUICTJxoVklMnGhSRHLJcMGIc+DeK+C2xTBoPER9cXTpHGLM2RGw4ODSFlcUPA4XQad5k3bdOdfxUcePzBuloUOhe3fzhuvA3/aox81OT5gW3Z38mRTGYTnwOr0kOeLZk777UDvj+CBctxQcNnxUH65cAYvKwsYimO9/gDsC98910O/HKP5cU4lwIECSW5EiMGwYbNgAvXub2iG5ORxmZccXX5jgUqtW8NtvOYMiLhc88QSRZ55mdfIaisYVpUxCmaP8jxA5sh1pO6j6ZtVDqz0OSvAksOmhTSTFHQhCbNkCv/wCSUmct+4xlu5cludaca44GpdvTFO7Aj3+M5LK27IFTdxuaNgQZs7MdywbUjYwY8MMyvjL0LJKS5yOw1SUFhGRPI62qOpx7TITIyDyBIBt2y8fePw90Ne27cOmzKjLjIiI/G1pabB3L5Qty75wGreMvoVxK8dhY+OwHMQ54/iw44fM2TyHxNff5vGfI8TnjWccXpEiJo3jpptiBkMOCjih6OMm9rD7VYgPwfISULvX4S/vd/v54uovuKp4U6hcOe9KFSBswYxK0Or2rG3eMGTmWiMaF4LZg6H+3TBwIjzRDlJjlED4cAzcGaP7Z9gCV6yXFImJprbIeedB1arm3zy3hASYNw9q1DCPbdu8WRw50hSs7NbNFLB16k2hFKzJf03m2m+uJWpHsW0br9PLqOtG0aJyC3NAv37w8suHUvJuuSzAl7XDROwYqWCAJ2rhDtt8PxSaZa8V7PPBkiVZ9W1EROSYOqG7zBzGOOBLy7LewBRVPRuYc3yHJCIipyS/39yAIs4ijLl+DMFwkCU7lhCKhqh7Rl28Li+danaiy8aFLF8+jVrbbeLDZsGGBVlFU/P7cCEz0xQVPcKHDxYmJSfggt0+ExA5e8+Rp5AWSuOTRZ9wVdqufIMFLtu8EYsPmo46kDcYAuCJwJIyUGGfue/IZ8jzy8F1yyAhVwpRvuVjixQxdTwcDpgxw9T1WLzY1FRxuaBlSxg4MCsYAubftXlzcxMpRO2qtWPHIzuYvWk2DstB4wqNcTkO/MBMnWpWe2VmmhvwxBQYdZZFmjv29YIOm6AHbusEq97O9nPidpsUOgVERESOq8M0VS84lmV1tixrE9AUGH9gJQi2bS8DvgL+ACYB96nDjIiIFBaPy0P9cvVpUqEJXpdZHuFz+5jQ/Wecs+cw5/WH2NX5Ujj3XFMs9brrYNQok8+fu1B4fLzpcFK5sqkf4sj/T+6WRNjjM4GIsqlm2+aj7OZrY5uAy99Y8emM8Zc1akHFfbArHpaVjn1eXAi+quch1WNWhByU6YwRQHG7oVw5U0j24Nxr14b582HHDtOCNxSCyZPV8lZOKG6nmxaVW9CsUrOsYAiY4snp6ezww1/FzM9MrV0wdYSPpkXOweVwYeUTGtyYBNuz100OheD88wt2IiIickTHq8vMaGB0PvteBF4s3BGJiIgcXp0KDeD+BnB/jJ3r18Onn5oUj82bTRDk/vvhqqvM/o8+Mi1n09IgI4Mo5hOJiGVWhfToYFaFvDLZ1AkJeBz0bRl7CX52frefWy+4FYpdCL1i59eELZhZMWt1iMfpwbbDRMi6vjMC5ffBn8UhzQsf1of+U6BPW3N+yAUJQTgvqTqvXvcR3eLupMeXf3LZnza43Tiv6oTlS8xKAWje3ASNmjePvXKleOwClSInsp3pO7n+FphZCZxRKJIJH42Dy7e7mFXzNbjsMqq9VY01e9fkOdfGtIwGzMq0fv1MqpiIiBxXx7WGyLGiGiIiInLC278fvvzSFAvduJHtG5YzzbuNly6MsCse3pwEV690QLFipD39OLXtt9mQsuFQQVXLBncUHC43QSL43D46VO/Al12+xGE5TK2SXr0OLeUHsD0e9vldNLo9zPZSPjIjmVxy5iXcWe9Oenx1MymZ+4hY0HCrxQeZ7XnmUhdTNs+gmK8Yj9brRbs/Mvl48SdsL+KgfavudGrb69An5imBFLwuL3GuuOPxrylSoHam7WT40uHsSt9FmzPb0KJSC+r3r8rSjPWEssX44oMwZ2gc5/yxExISeH3W6zwz9RnSw1mFWd2Wi9ahCkwa7YcyZeDhh+Hyy4/DrERETh8nRVHVY0UBEREROelFo6ZNrd8PloVt20xeM5lxi7+i2uYM7kpsSWrrZozY+iPJgWQuqXYJjcs3zmrFCbB2LXz1FSxdCsWKQZMmcPXV7IqmsmLXCionVaZiUkXz7ewoa3asJGHXPs6oVNsUPxURflr7Ex2HdSRqRwmEA8S742lYriFzNv9GejhnlyRnBO5Masn7D08FIBwN021UN8auHIvb4cbGpkrRKky5ZQql/fnkoomIyDGngIiIiIiIyN8QjoYp/Vpp9gZydkOKc8ZhWRYZ4bxto9tWbcvkWybn2Pbn7j+Zv3U+lZMq06RCk5yBSxERKXAna5cZEREREZGjY9smTczrzVvY+G8KRUJ0G90tTzAEIBAJxCyY6nP5aHtm2zzbzy5xNmeXOPtfjUdERArecekyIyIiIiLyr7z6KhQtajo6lSwJb775t7ot5dZrYi/GrBiT7/4KRSrgd/sPPXY73BSNK8pd9e/6x99TRESOL60QEREREZGTgm3bRKMRnFd1gvHjs3bs2QOPP27uP/DA377uvsx9DFk8hGAkGHO/3+3nlbavkBnJ5M3Zb5IcSObKGlfyZIsnKeYr9k+mIiIiJwAFRERERETkhBaJRnh+2vMMnD2QlMwU1s6AKrkPCgTg+edNy+u/mT6zdf/WQx2UcrOwuLL6lVx37nU4LAe31bntn0xBREROQAqIiIiIiMgJ7aHvH+KjhR+RHjLtbM9IzefAPXsgGDQ1Rf6GSkmVyK/RQMvKLRl2zbC/dT0RETk5qIaIiIiIiBSoSDTC+/Pe54J3L+Dst8/mySlPkhJIOapz92XuY/CCwYeCIQCrSuRzcJEi4PH87fH53D6eaP4E8e74HNv9bj9vXvbm376eiIicHBQQEREREZECdeuYW+n9Q29+3/E7q/es5pWZr1D29bLc9e1drNq96rDnbtq3CbfDnWNbn3aQHmudc2YmbN/+j8b4nxb/4e3L3uasYmeR6EmkdZXWTLt9GueVOe8fXU9ERE58Vn7LA08mDRo0sOfNm3e8hyEiIiIiuazctZI679chEA7k2efAQZw7jok3TeSiyhfB3r0wa5bpHtO0KTgcpAZTKf1aaTLCGTnOvWwlDJoAVbIvNPF6oU8f6NevgGclIiInMsuy5tu23eBIx2mFiIiIiIj8O7YNU6aY1rcTJkAkcmjXb5t/y7dgaZQo6aF0/m/c/2G//jqUKwc33giXXQZVq8LKlSR4EujVqFeedJZfzoS9vlwXzMyE+fOP9exEROQUpaKqIiIiIvLP7dsHrVrBqlUQCpkaHqVLw8yZUKYM5RLLYXH4ri8VF62F4c+YTjGBAytJUlPh0kth7Vpebvsypf2leW3Wa+zJ2EOdpBoMeHsVdbflapPr9UK9egU0UREROdVohYiIiIiI/HNPPAHLlpkARmYm7N8PGzbAXXcB0KpKK4r7iuO0nPle4u45EcjImRKDbcPu3TBnDg7LwcMXPsy2R7YRfDrInPuX0KzqxRAXl3W8ZZnH99xTELMUEZFTkAIiIiIiIvLPffGFCYRkFwqZ1JlwGKfDyS+3/ULdsnVjps54nB5qWqWxYtW1czggJZ9uNGPGwJ13gt8PTie0bGnqj5Qt++/nJCIipwUVVRURERGRfy4x0awOyc3phPT0HG1wNyRv4KEfHuK7ld/hc/sIRUM0LNeQifuvwvf4U+b47Hw+2LbNtNMVERE5SkdbVFU1RERERETkn+vUCYYPh3A4a5vDQbhFc/475w2+WvYVRbxF6NmoJ11qdWHktSPZtG8TS3cspUrRKtQsWdOky3w0BP780wRFLMsEQ157TcEQEREpMFohIiIiIiL/3Pbt0LixqfeRmgp+P7bPx2U9i/KLa9Ohdrt+t5+76t/FG5e+Efs6gQAMHQqjRkGpUnDvvdCkSSFOREREThVHu0JEARERERER+XcCAfj6a1i4EGrV4rOaQe6d3oe0UFqOw7xOL3/2+pON+zayIWUDDco14KziZx2nQYuIyKlKKTMiIiIi8u+sWQPjx5s6IJ07m3a6wIKtC3hs8mPM2zKPsollearFU9x0881w880AjP2qS55gCEBmJJPqb1fH6XDisByEoiG61OrCkE5DcDry70IjIiJSEBQQEREREZG8XnwRXnjB3Hc44MEH4bPPWNyiOi0+aUF6MJ2z9oAdTeGu5O5sT9tO76a9AaiQWAGX5SJsh/NcNhAJQCTr8egVo2k8tzG9GvcqjFmJiIgcopQZEREREclp4UJo1swUO83O5+OmNy9mycJJjBwB5faDDeyOh+uuAUfTpvz3kv9SLK4YDT5oQHo4Peblc6tRogYreq449vMQEZHT0tGmzDgKYzAiIiIichIZNgwyM/NudzopO/lXfv4Equ0BfwgSQlA5BX4YCitW/Uq7oe1IDabyWefPjvrbpQXzpteIiIgUNAVERERERCSnUAhirSK2bVpsj8MVzfsi0hmFG5dAeiidJ396ki61u9C4fOMjfiu3w02nmp2OzbhFRET+BgVERERERCSnBg1MIdXcIhHOrdMObyTvLn8Yyu8z9xdvXwzAex3eI8GTgNvhBjj01eUwZezi3fGckXAGz1z8zLGfg4iIyBGoqKqIiIjIqWbzZnj7bViwAOrVg549oUKFI5+3bx906ADz5+dcIeJ2g8sFAwZQ7YILCA/+GjJyptTs98C0Kub+WcVMK906Z9Rh6T1LefO3N1m0bRH1y9anc63OjFs5jtV7VtOySktuveBWEr2Jx2jiIiIiR09FVUVEREROJUuXmoKogQAEg2alh9cLs2bBuefC7t2wZAlUrAjVquU898YbYeRIc95BLhe0bAnvvgtnnWUCJVdeiT11Kla6KZqa5oKFZeGi28HnjWfMdWNoV61d4c1ZREQkm6MtqqoVIiIiIiKnkp49zUqPg4JBc7vvPmjSBN56ywRIgkFo3BhGj4aiRU0R1dzBEIBwGJYvN8EQAMuCMWOwPvgAe/Bgtqds4e2aybxZP0L5omUZcOkABUNEROSkoBUiIiIiIqcSlwsiMYp8WBbEx0Nato4uHg+0bw9jx8L+/VCihCmomltSEiQn5/sto3aU9FA6frcfy7KOwSRERET+ObXdFRERETmFLN62mF4TenHDNzcwfOlwQpEYgQswQY9YLCtnMAQgGCQ6aSKR3bvgjz+gbNm85zkccMklhx2bw3KQ4ElQMERERE4qCoiIiIiInOA+XvgxTT9qyqB5gxi+bDj/N+7/aP1Z69hBkTvugLi4nNvi4sDni3ntgB2i8VNlmNOtFezaZTY6HFnnFSsGr756DGcjIiJyYlBAREREROQEtj9zPz0n9CQjnEHUjgKQFkpj4daFDFs6LO8J/ftDmzYmAFKkiAlqtG4NXbqYdJpcUrywoHSUS67JID1kiqTicMAFF0CfPqZ+SJUqBThDERGR40MBEREREZET2IwNM3A73Xm2p4XSGL50eN4T4uLgu+9g8WL4/HP4/XcYPx5eeAGKFiXsdgIQsSDNDXd3ANsBUQu+rX7gGuEw+P3Qty+UKlVwkxMRETmO1GVGRERE5ATm9/jJrwh+kjcp/xPPPtvcDqpYEZYuZWqvK0j4dT5/FYM3msLCcmZ32II92bNqDlNEVURE5FSggIiIiIjICaxZxWbEu+PZH9yfY3u8O54eDXr8vYuVKUP4xee55JtrSQ2m5txnQeu1B+57vSbFRkRE5BSmlBkRERGRE1EgABMn4hw/gTdbvIhFzg4uZ/jPoGnFpvmenhHKoPf3vSn2SjHiX4yn8/DOrEtex6VnXUrTCk3xu/2HjvUH4dbFUGM3pktN+fLQu3dBzUxEROSEYOW3BPNk0qBBA3vevHnHexgiIiIix8aUKXD11QDYQEbGfm7pZDPynKxD4t3x9G/Tn16Ne8W8RJvP2jBrwywCkQBgWuMW9xVnVc9VJHoT+eL3L/j898/xurz8X4l2XDVmOdaGjXDppaZTTUJCQc9SRESkQFiWNd+27QZHPE4BEREREZETSHIyVKgAaWk5Nqe7oFZP2FA0a1udM+qw8O6FeS6xcOtCmn/SPKtrzAE+l48XWr9A76Za/SEiIqeuow2IKGVGRERE5EQyZgxYVp7NDhtu/D3nttxpNAct27kMh5X3ZV5GOIM5m+cck2GKiIic7BQQERERETmR7Ntn2t7m4olAUmbW43h3PHfUvSPmJaqXqB6zM02cK47zy5x/zIYqIiJyMlNAREREROQEMH/LfNp/3p4LN/UjEMnMsz/dDeOrm/tuh5vmlZpzd/27Y16rYbmG1C5VG4/Tc2ibhYXX6aV7ve4FMn4REZGTjQIiIiIiIsfZ3I2/8eITzbj87e9pO3cPQ8+xSXND9MD+/W74tjrMqGQeu51uxl0/DrfTHfN6lmUx+ebJXHvOtXicHhyWg2YVmzHrzlmU8pcqnEmJiIic4FRUVURERORY2rwZduyAWrUgLu7Ix4fDzKtTmhqr9pIYgkwnRCzo3xyq7wZ3FL48zwRE7AMfZcW741nZcyUVilQ44uWjdpSoHcXlcP3LiYmIiJwcjraoqv4yioiIiBwLe/dC164wcyZ4PBCNwquvwj33HP68L76g1qq9+EPmoTdivj48C8o8BpkxXq1ZWJSKP7qVHg7LEbPAqoiIyOlOARERERGRY6FrV5g+HYJBCATMtocfhjVr4Pzz4YoroHjxvOd98cWhYEgOFly8xc30qi4yIhmHNse743moyUN4Xd6CmYeIiMhpQh8XiIiIiPxbmzaZlSHBYM7tGRkwYADcey9UqADDh+c91xs7sGEBHc+7hmHXDKNq0apYWBSNK8qTLZ6kX6t+x34OIiIipxnVEBERERH5txYsgJYtYf/+wx/n88HatVCmTNa2cePgxhshLS3HofuLxhO/Yy9Ot+kUE4qEcDlcWJZ1jAcvIiJyajnaGiJaISIiIiLyL+zP3M+7gRn0aJPBuw1gv+cIJ4wcmfPxlVfCbbeBz4ft8xFNSMBOSiLx+6mHgiFgOssoGCIiInLsqIaIiIiIyD+0Pnk9jT5sREoghcw6YeJrQ9+WMGcwVE6JcUI0mlVf5CDLgv/9D+6/H+unn7CKF4cOHSA+vjCmICIictpSQEREREQkhvXJ6+k/oz/TN0ynWrFqPN78cZpWbJrjmK5fd2VH2o5Dj9M9EHDBPdfEMeHTEEQiOS/qcJgVIbFUr25uIiIiUiiUMiMiIiKSy+o9q7ngvQv4cOGHLNu5jG9XfUvboW0Z+UdWusuavWuYu2VunnOjDphYMQD3329WeTgc5hYfD48+CmefXZhTERERkXxohYiIiIhILk/99BT7g/uJ2lHK7gNvxGZd0XR6TuxJ51qdcVgOBs8ffNhrbOrbmwpdu8KIESYgcuON0OCI9d1ERESkkCggIiIiIpLLz+t+puKeKCNHwAXbwWlD1IJva+1g27UrKVexFhv3bcz3fAcO0kPp0LSpuYmIiMgJRykzIiIiIrlUdBVnzmCouw1cNliYoMhVf0Qp06YjRCJcUu0SnDhjnl88vjhnFT+rcActIiIif4sCIiIiIiLZTZjAr4//Ran0vC+ULMC5fgNMmMB151xH9ZJ5i6A6LSfDuwzHYellloiIyIlMf6lFRETktLdp3ybe+PUNBo55gug1XXBlBrHyOzgYhMWL8bq8zOk+h2cueobyieUp4StBh+odWNVzFW3ObFOYwxcREZF/QDVERERE5PS1cyfDB93LHZFRRB0W9/wWJTNk4zvcOV4vVKsGQIIngX6t+tGvVb9CGa6IiIgcO1ohIiIiIqenNWtIPr86d4S+IcMZJdOKkJBp4wkf4byiRaFz50IZooiIiBQcBURERETk1LdqFYwcCb//nrXtkUf4vmQKrmjWpslnQsB9mOs0bw6zZ0NcXIENVURERAqHAiIiIiJy6goG4aqroE4duOMO0wK3RQvYvx9+/BHLtsHOOvy3CjC6JqRmD4r4/XDLLZCeDtOnQ5UqhT0LERERKQCqISIiIiKnrn79YPJkyMgwN4C5c+G++yA+nktX7yeS/eMhC27pDF3/8vDOriaU9JeE226DDh3AyrfMqoiIiJyEtEJERERETl0ffJAVCDkoMxO++gruvpskh4/PRoMvZG7uMMThpMItPSk56ReTZnPllQqGiIiInIK0QkREREROXenpsbeHQvDoo7BsGV3Gj6fZRy6+qZZJepXyXPHccM45s3HhjlNEREQKnQIiIiIicupq2xa++w6i0Zzb69WDhAT45hv46y/OWLKEnmedBeeee3zGKSIiIoVOARERERE5dQ0YADNmmJUigQB4POD1mlSag6pVMzcRERE5rSggIiIi/9/e/QdZVR5mHP8+u/xYQCtQEEmMipG2YkmwErWJmrajwSS1qVg7mJroaAO2tU5NMo4ZY8TEjBrbOom1iba1SacaTSqoDU0tJEUyMWmDIyUimvgrFTABtQ4Tfi/79o97maxwUQT2nsue72fmzt59zznLw8zhzOXZ97xHGryOPhqeeAJuu63xuNypUxsLqh5+eNXJJElSxSxEJEnS4DZ+PHzyk1WnkCRJHcZCRJIkHRjWr2+s+fHTn8Ipp8Cpp/r0F0mStNcsRCRJUmfbvh1uuAHmzoVSGq8RI+Cd74QFC2Do0KoTSpKkA1BX1QEkSZJ2qxSYOROuvhp6exvlSF8fbNgA3/1uY20QSZKkvWAhIkmSOtd3vgMLFzaKkZ1t3Ah33NH+TJIkaVCwEJEkSZ1r0SLYtGn321sVJZIkSXvAQkSSJHWGVatg9epXj40dC8OHt95/2DC46KKBzyVJkgYlCxFJklSt5cthyhSYPBmOOQbe9jZYubKx7bzzoLu79XEnnwxz5rQvpyRJGlQsRCRJUnXWr4d3v7tRgGze3Hg99hicdlpjjZAJE+C++2D0aDj4YOjpaTxh5vOfh8WLG7NEJEmS9oKP3ZUkSe3V18fab9wDy5Zx6HPrYNu2V28vpVGMzJsH558PZ5wBa9fC974HXV2NmSFD/AgjSZL2jZ8mJElS26z89j380bzzeXxMLwDHDYO7euBXN+y046ZNjTVFdhg6tDFrRJIkaT/xlhlJkjTwtmxhw4dmcersAIJKAAAJu0lEQVSDs1j2y71sGQJbhsCjE+CUi2Dj0J32HzGiMRNEkiRpgFiISJKkAbXoc3/C7D/o4V2j7uHnw6D0+/RRumDzELj32H4HjBgBxx/fWFtEkiRpgHjLjCRJGhi9vVx5/Rl8YfNitvwG9HUBZdfdNg6F50cHJh3VWBvkwgvhox+FpL15JUlSrViISJKk/e+hh9gw8/e4dc56Ng3vN96i4xi5DaaPPAae+VHb4kmSJFmISJKk/er/rv0ED995A89Pgq4WM0L669kGU17u5vSb729POEmSpCYLEUmStN/cdP3v8qltCxh2DvQ21wdppWcbTNgAH956LFfeuISuQ8a1N6gkSao9CxFJkrTvtm5l4ft/jbknPsvmYbB5x1NjCqTv1QupjtoKX31gOGct+BEccUQlcSVJkixEJEnSPtm4djW3n3MUi9/cu+vjcwOURgnS3dd41O4HX5zIWUufheHDW/04SZKktrAQkSRJe6+vjwv+4khmr97OA2+l5aKpB22FaxfDIWPfxNu/OJ8Tjjix3SklSZJ20fX6u0iSJLWwYgUvnnkq/3r0doYUmLkSRm7ddbfebvjI9NlcdP9PLEMkSVLHsBCRJElv3BVXwNSpjFn0MN//e1h2GPzhD+GtL/+iFElf4/1fT7mcg265DYY4MVWSJHUOP5lIkqQ3Zt48uOkmALqBaT+DyS/Bj8fCon+Cecc2XodugD8/82pO+uCnq80rSZLUgjNEJEnSG3P55bsMjeqFX9oKH5oJG4bCyavgs++5npNmW4ZIkqTO5AwRSZL0xqxe3XJ40ivw+DgYt7mLa8+5hSNn/Wmbg0mSJO25SmaIJDk3yYokfUmm9xs/KsmmJMuary9VkU+SJL2GiRNbDgd4/u3/yJ1f384xliGSJKnDVXXLzGPATGBJi21Pl1KmNV+XtDmXJEl6PddcAz09u47PmAEXXtj2OJIkSXujkkKklLKylPJkFX+2JEnaRxdfDDfeCGPGNJ4cM3IkXHUVfPObVSeTJEnaY524qOqkJI8meSjJqbvbKcnsJEuTLF23bl0780mSVG8JXHYZvPgirF0L69fDddc1xiVJkg4QA7aoapJFwGEtNl1VSrl/N4e9ABxRSnkpyQnAfUmOK6Ws33nHUsrtwO0A06dPL/srtyRJ2kNdXY1ZIpIkSQegAStESimn78UxW4AtzfePJHka+BVg6X6OJ0mSJEmSaqyjbplJMj5Jd/P90cBk4JlqU0mSJEmSpMGmqsfunp1kFfCbwIIkDzY3nQYsT/I/wL8Al5RSXq4ioyRJkiRJGrwG7JaZ11JKmQ/MbzF+L3Bv+xNJkiRJkqQ66ahbZiRJkiRJktrBQkSSJEmSJNWOhYgkSZIkSaodCxFJkiRJklQ7FiKSJEmSJKl2LEQkSZIkSVLtWIhIkiRJkqTasRCRJEmSJEm1YyEiSZIkSZJqx0JEkiRJkiTVjoWIJEmSJEmqHQsRSZIkSZJUOxYikiRJkiSpdixEJEmSJElS7ViISJIkSZKk2rEQkSRJkiRJtWMhIkmSJEmSaiellKoz7LMk64CfVJ1D2gfjgBerDiFVwHNfdeR5rzryvFddee5X48hSyvjX22lQFCLSgS7J0lLK9KpzSO3mua868rxXHXneq6489zubt8xIkiRJkqTasRCRJEmSJEm1YyEidYbbqw4gVcRzX3Xkea868rxXXXnudzDXEJEkSZIkSbXjDBFJkiRJklQ7FiJShZKcm2RFkr4k03fa9okkTyV5MsmMqjJKAynJ3CSrkyxrvt5XdSZpoCQ5s3lNfyrJlVXnkdolyXNJfti8zi+tOo80UJLckWRtksf6jY1NsjDJj5tfx1SZUa9mISJV6zFgJrCk/2CSKcAs4DjgTOBvk3S3P57UFjeXUqY1X/9WdRhpIDSv4bcC7wWmAOc1r/VSXfx28zrv40c1mH2Zxmf3/q4EvlVKmQx8q/m9OoSFiFShUsrKUsqTLTZ9ALi7lLKllPIs8BRwYnvTSZL2oxOBp0opz5RStgJ307jWS5IGiVLKEuDlnYY/AHyl+f4rwO+3NZRek4WI1JneDDzf7/tVzTFpMLo0yfLmNFOnkWqw8rquOivAfyR5JMnsqsNIbTahlPICQPProRXnUT9Dqg4gDXZJFgGHtdh0VSnl/t0d1mLMR0LpgPRa/waALwKfoXF+fwb4K+Ci9qWT2sbruursXaWUNUkOBRYmeaL5m3RJqpSFiDTASimn78Vhq4C39Pv+cGDN/kkktdee/htI8nfANwY4jlQVr+uqrVLKmubXtUnm07iFzEJEdfGzJBNLKS8kmQisrTqQfsFbZqTO9AAwK8nwJJOAycB/V5xJ2u+aHwx2OJvGQsPSYPQDYHKSSUmG0Vg4+4GKM0kDLsmoJAfveA+8B6/1qpcHgAua7y8AdjdDXBVwhohUoSRnA7cA44EFSZaVUmaUUlYk+RrwONAL/FkpZXuVWaUB8rkk02jcOvAcMKfaONLAKKX0JrkUeBDoBu4opayoOJbUDhOA+Umg8X+Pu0op/15tJGlgJPkq8FvAuCSrgGuAG4CvJbkY+F/g3OoSamcpxdtXJUmSJElSvXjLjCRJkiRJqh0LEUmSJEmSVDsWIpIkSZIkqXYsRCRJkiRJUu1YiEiSJEmSpNqxEJEkSYNekkuSfLjqHJIkqXP42F1JkiRJklQ7zhCRJEkdJck7kixP0pNkVJIVSX59p33OSvJfSR5NsijJhOb4F5J8qvl+RpIlSbqSzE3y8eb4ZUkeb/4Zd7f/byhJkjqBM0QkSVLHSXId0AOMAFaVUq7fafsY4JVSSknyx8CxpZSPJRkJ/AC4FPgS8L5SytNJ5gI/L6X8ZZI1wKRSypYko0spr7Tz7yZJkjrDkKoDSJIktfBpGsXGZuCyFtsPB+5JMhEYBjwLUErZmOQjwBLg8lLK0y2OXQ7cmeQ+4L6BCC9Jkjqft8xIkqRONBY4CDgY6Eny2STLkixrbr8F+JtSylRgDo3ZJDtMBV4C3rSbn/1+4FbgBOCRJP6CSJKkGrIQkSRJneh24GrgTuDGUspVpZRppZRpze2HAKub7y/YcVCSI4GPAccD701yUv8fmqQLeEsp5T+BK4DRNIoXSZJUM/5GRJIkdZTm43F7Syl3JekGHk7yO6WUb/fbbS7w9SSrge8Dk5IE+Afg46WUNUkuBr6c5B39jusG/jnJIUCAm11DRJKkenJRVUmSJEmSVDveMiNJkiRJkmrHQkSSJEmSJNWOhYgkSZIkSaodCxFJkiRJklQ7FiKSJEmSJKl2LEQkSZIkSVLtWIhIkiRJkqTasRCRJEmSJEm18//wKEHFyBKmegAAAABJRU5ErkJggg==\n",
"text/plain": "<matplotlib.figure.Figure at 0x7fa39caffef0>"
},
"metadata": {}
}
],
"source": "# y values for graph only\ny_cat = pd.DataFrame([\"accept\" if i == 1 else \"recject\" for i in y ])\ndata_tsned = pd.concat([xs,ys, y_cat],axis=1)\ndata_tsned = np.array(data_tsned)\nprint(data_tsned[:5,:])\nprint(data_tsned.shape)\n\n#secondly we graph scatter-plot using matplotlib \nplt.scatter(x=xs, y=ys, c=['g','r'])\nplt.xlabel(\"x-axis\")\nplt.ylabel(\"y-axis\")\nfig = plt.gcf().set_size_inches(18.5, 10.5)\n\nplt.show()"
},
{
"execution_count": 23,
"cell_type": "code",
"metadata": {},
"outputs": [
{
"output_type": "stream",
"name": "stdout",
"text": "logistic regression accuracy: 0.805\n"
},
{
"output_type": "stream",
"name": "stderr",
"text": "/opt/conda/envs/DSX-Python35/lib/python3.5/site-packages/sklearn/linear_model/logistic.py:1228: UserWarning: 'n_jobs' > 1 does not have any effect when 'solver' is set to 'liblinear'. Got 'n_jobs' = -1.\n \" = {}.\".format(self.n_jobs))\n"
}
],
"source": "# train and test the data\n\nlr = LogisticRegression(n_jobs=-1, random_state=20)\nlr.fit(X_train, y_train)\naccuracy = lr.score(X_test, y_test)\nprint(\"logistic regression accuracy: {}\".format(accuracy))"
},
{
"source": "<h4>As a bank, we want to minimize the downside risk. So we won't mind if our model got couple of our good lenders as bad lenders. But it would be a disaster if it got\ncouple of bad lenders as good ones. Therefore, the normal accuracy measure wouldn't be accurate in this case so we use precision. However, because the precision\non its own might be optimistic we have to combine it with recall to get the full picture using the weighted F score as we see. The more weight that it's put to beta \nthe more importance of the precision is. And that's what we want.</h4>",
"cell_type": "markdown",
"metadata": {}
},
{
"execution_count": 24,
"cell_type": "code",
"metadata": {},
"outputs": [
{
"output_type": "stream",
"name": "stdout",
"text": "The precision: 0.8311688311688312, F beta score: 0.8454425363276089\n"
}
],
"source": "# evaluation process\nlr_pred = lr.predict(X_test)\nprecision = precision_score(y_test, lr_pred)\nrecall = recall_score(y_test, lr_pred)\n\n# calculate the f beta score\ndef f_beta(precision, recall, beta):\n return 1 / (beta *(1/ precision) + (1-beta)*(1/recall))\n\n\nf_score_beta = f_beta(precision, recall, .8)\nprint(\"The precision: {}, F beta score: {}\".format(precision, f_score_beta))\n"
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3.5",
"name": "python3",
"language": "python"
},
"language_info": {
"mimetype": "text/x-python",
"nbconvert_exporter": "python",
"version": "3.5.5",
"name": "python",
"file_extension": ".py",
"pygments_lexer": "ipython3",
"codemirror_mode": {
"version": 3,
"name": "ipython"
}
}
},
"nbformat": 4
}
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment