ML_DM1_SupervisedLearning_NeuralNetwork Dataset2.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "34ed6713",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.96774194 0.98924731 0.96774194 1.         0.98924731 0.97826087\n",
      " 1.        ]\n",
      "Accuracy score = 0.9846056234555534\n",
      "time for training and predicting the test sets = 2.7421942400033004 s\n",
      "Test score = 0.9846153846153847\n",
      "time for the learning curve = 11.585565645997121 s\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7fcb19e64340>"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x396 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Dataset 2 : Wines\n",
    "# Output : Type of wine : red or white\n",
    "# MLPClassifier() --> hidden_layer_sizes=(100,), activation='relu', solver='adam',max_iter= 1000\n",
    "\n",
    "import numpy as np\n",
    "import time\n",
    "import pandas as pd\n",
    "import io\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# split your datas between training and test\n",
    "from sklearn.model_selection import train_test_split\n",
    "# comparing datas\n",
    "from sklearn.metrics import accuracy_score\n",
    "# model \n",
    "from sklearn.neural_network import MLPClassifier\n",
    "# cross validation\n",
    "from sklearn.model_selection import cross_val_score\n",
    "# validation curve\n",
    "from yellowbrick.model_selection import validation_curve\n",
    "# learning curve\n",
    "from sklearn.model_selection import learning_curve\n",
    "# will be used to convert string into float\n",
    "from sklearn.preprocessing import LabelEncoder\n",
    "\n",
    "labelencoder_X = LabelEncoder()\n",
    "\n",
    "#dataset\n",
    "a = io.open('wine-quality-white-and-red.csv', encoding='latin-1')\n",
    "df = pd.read_csv(a)\n",
    "df = df.take([10*i for i in range(649)], axis = 0)\n",
    "df = df.apply(LabelEncoder().fit_transform)\n",
    "\n",
    "#input\n",
    "X = df.drop(columns=['type'])\n",
    "\n",
    "#output\n",
    "y = df['type']\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X,y,test_size=0.2)\n",
    "\n",
    "time1 = time.perf_counter()\n",
    "\n",
    "model = MLPClassifier(hidden_layer_sizes=(100,),\n",
    "                      activation='relu',\n",
    "                      solver='adam',\n",
    "                      max_iter= 1000)\n",
    "\n",
    "model.fit(X_train,y_train)\n",
    "\n",
    "validation = cross_val_score(model, X, y, cv=7)\n",
    "print(validation)\n",
    "print('Accuracy score =', validation.mean())\n",
    "\n",
    "#Testing the test set\n",
    "predictions = model.predict(X_test)\n",
    "score = accuracy_score(y_test,predictions)\n",
    "\n",
    "#time after training and predicting the test sets\n",
    "time2 = time.perf_counter()\n",
    "timeA = time2 - time1\n",
    "print('time for training and predicting the test sets =',timeA,'s')\n",
    "\n",
    "print('Test score =', score)\n",
    "\n",
    "#Plot the learning curve\n",
    "train_sizes, train_scores, valid_scores = learning_curve(MLPClassifier(hidden_layer_sizes=(100,),activation='relu',solver='adam',max_iter= 1000) , \n",
    "                                                         X , y , \n",
    "                                                         train_sizes=np.linspace(0.01,1,50), \n",
    "                                                         cv=7, \n",
    "                                                         scoring = 'accuracy', \n",
    "                                                         n_jobs= -1)\n",
    "\n",
    "#time after doing the learning curve\n",
    "time3 = time.perf_counter()\n",
    "timeB = time3 - time2\n",
    "print('time for the learning curve =',timeB,'s')\n",
    "\n",
    "train_std = np.std(train_scores, axis = 1)\n",
    "valid_std = np.std(valid_scores, axis = 1)\n",
    "\n",
    "train_mean = np.mean(train_scores, axis = 1)\n",
    "valid_mean = np.mean(valid_scores, axis = 1)\n",
    "\n",
    "plt.plot(train_sizes,train_mean, label = 'Training score')\n",
    "plt.plot(train_sizes,valid_mean, label = 'Cross-validation score')\n",
    "\n",
    "plt.fill_between(train_sizes, train_mean - train_std, train_mean + train_std, color = '#DDDDDD')\n",
    "plt.fill_between(train_sizes, valid_mean - valid_std, valid_mean + valid_std, color = '#DDDDDD')\n",
    "\n",
    "\n",
    "plt.title('Learning curve')\n",
    "plt.xlabel('Training size')\n",
    "plt.ylabel('Accuracy score')\n",
    "plt.legend(loc ='best')\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "b106d8cd",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.98924731 0.97849462 0.97849462 0.98924731 1.         0.97826087\n",
      " 1.        ]\n",
      "Accuracy score = 0.9876778200761371\n",
      "time for training and predicting the test sets = 0.5779653849995157 s\n",
      "Test score = 0.9923076923076923\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/ye/opt/anaconda3/lib/python3.9/site-packages/sklearn/neural_network/_multilayer_perceptron.py:559: ConvergenceWarning: lbfgs failed to converge (status=2):\n",
      "ABNORMAL_TERMINATION_IN_LNSRCH.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "  self.n_iter_ = _check_optimize_result(\"lbfgs\", opt_res, self.max_iter)\n",
      "/Users/ye/opt/anaconda3/lib/python3.9/site-packages/sklearn/neural_network/_multilayer_perceptron.py:559: ConvergenceWarning: lbfgs failed to converge (status=2):\n",
      "ABNORMAL_TERMINATION_IN_LNSRCH.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "  self.n_iter_ = _check_optimize_result(\"lbfgs\", opt_res, self.max_iter)\n",
      "/Users/ye/opt/anaconda3/lib/python3.9/site-packages/sklearn/neural_network/_multilayer_perceptron.py:559: ConvergenceWarning: lbfgs failed to converge (status=2):\n",
      "ABNORMAL_TERMINATION_IN_LNSRCH.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "  self.n_iter_ = _check_optimize_result(\"lbfgs\", opt_res, self.max_iter)\n",
      "/Users/ye/opt/anaconda3/lib/python3.9/site-packages/sklearn/neural_network/_multilayer_perceptron.py:559: ConvergenceWarning: lbfgs failed to converge (status=2):\n",
      "ABNORMAL_TERMINATION_IN_LNSRCH.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "  self.n_iter_ = _check_optimize_result(\"lbfgs\", opt_res, self.max_iter)\n",
      "/Users/ye/opt/anaconda3/lib/python3.9/site-packages/sklearn/neural_network/_multilayer_perceptron.py:559: ConvergenceWarning: lbfgs failed to converge (status=2):\n",
      "ABNORMAL_TERMINATION_IN_LNSRCH.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "  self.n_iter_ = _check_optimize_result(\"lbfgs\", opt_res, self.max_iter)\n",
      "/Users/ye/opt/anaconda3/lib/python3.9/site-packages/sklearn/neural_network/_multilayer_perceptron.py:559: ConvergenceWarning: lbfgs failed to converge (status=2):\n",
      "ABNORMAL_TERMINATION_IN_LNSRCH.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "  self.n_iter_ = _check_optimize_result(\"lbfgs\", opt_res, self.max_iter)\n",
      "/Users/ye/opt/anaconda3/lib/python3.9/site-packages/sklearn/neural_network/_multilayer_perceptron.py:559: ConvergenceWarning: lbfgs failed to converge (status=2):\n",
      "ABNORMAL_TERMINATION_IN_LNSRCH.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "  self.n_iter_ = _check_optimize_result(\"lbfgs\", opt_res, self.max_iter)\n",
      "/Users/ye/opt/anaconda3/lib/python3.9/site-packages/sklearn/neural_network/_multilayer_perceptron.py:559: ConvergenceWarning: lbfgs failed to converge (status=2):\n",
      "ABNORMAL_TERMINATION_IN_LNSRCH.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "  self.n_iter_ = _check_optimize_result(\"lbfgs\", opt_res, self.max_iter)\n",
      "/Users/ye/opt/anaconda3/lib/python3.9/site-packages/sklearn/neural_network/_multilayer_perceptron.py:559: ConvergenceWarning: lbfgs failed to converge (status=2):\n",
      "ABNORMAL_TERMINATION_IN_LNSRCH.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "  self.n_iter_ = _check_optimize_result(\"lbfgs\", opt_res, self.max_iter)\n",
      "/Users/ye/opt/anaconda3/lib/python3.9/site-packages/sklearn/neural_network/_multilayer_perceptron.py:559: ConvergenceWarning: lbfgs failed to converge (status=2):\n",
      "ABNORMAL_TERMINATION_IN_LNSRCH.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "  self.n_iter_ = _check_optimize_result(\"lbfgs\", opt_res, self.max_iter)\n",
      "/Users/ye/opt/anaconda3/lib/python3.9/site-packages/sklearn/neural_network/_multilayer_perceptron.py:559: ConvergenceWarning: lbfgs failed to converge (status=2):\n",
      "ABNORMAL_TERMINATION_IN_LNSRCH.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "  self.n_iter_ = _check_optimize_result(\"lbfgs\", opt_res, self.max_iter)\n",
      "/Users/ye/opt/anaconda3/lib/python3.9/site-packages/sklearn/neural_network/_multilayer_perceptron.py:559: ConvergenceWarning: lbfgs failed to converge (status=2):\n",
      "ABNORMAL_TERMINATION_IN_LNSRCH.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "  self.n_iter_ = _check_optimize_result(\"lbfgs\", opt_res, self.max_iter)\n",
      "/Users/ye/opt/anaconda3/lib/python3.9/site-packages/sklearn/neural_network/_multilayer_perceptron.py:559: ConvergenceWarning: lbfgs failed to converge (status=2):\n",
      "ABNORMAL_TERMINATION_IN_LNSRCH.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "  self.n_iter_ = _check_optimize_result(\"lbfgs\", opt_res, self.max_iter)\n",
      "/Users/ye/opt/anaconda3/lib/python3.9/site-packages/sklearn/neural_network/_multilayer_perceptron.py:559: ConvergenceWarning: lbfgs failed to converge (status=2):\n",
      "ABNORMAL_TERMINATION_IN_LNSRCH.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "  self.n_iter_ = _check_optimize_result(\"lbfgs\", opt_res, self.max_iter)\n",
      "/Users/ye/opt/anaconda3/lib/python3.9/site-packages/sklearn/neural_network/_multilayer_perceptron.py:559: ConvergenceWarning: lbfgs failed to converge (status=2):\n",
      "ABNORMAL_TERMINATION_IN_LNSRCH.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "  self.n_iter_ = _check_optimize_result(\"lbfgs\", opt_res, self.max_iter)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "time for the learning curve = 5.7281134320001 s\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7fcb1ef58250>"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x396 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Dataset 2 : Wines\n",
    "# Output : Type of wine : red or white\n",
    "# MLPClassifier() --> hidden_layer_sizes=(100,), activation='relu', solver='lbfgs',max_iter= 1000\n",
    "\n",
    "import numpy as np\n",
    "import time\n",
    "import pandas as pd\n",
    "import io\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# split your datas between training and test\n",
    "from sklearn.model_selection import train_test_split\n",
    "# comparing datas\n",
    "from sklearn.metrics import accuracy_score\n",
    "# model \n",
    "from sklearn.neural_network import MLPClassifier\n",
    "# cross validation\n",
    "from sklearn.model_selection import cross_val_score\n",
    "# validation curve\n",
    "from yellowbrick.model_selection import validation_curve\n",
    "# learning curve\n",
    "from sklearn.model_selection import learning_curve\n",
    "# will be used to convert string into float\n",
    "from sklearn.preprocessing import LabelEncoder\n",
    "\n",
    "labelencoder_X = LabelEncoder()\n",
    "\n",
    "#dataset\n",
    "a = io.open('wine-quality-white-and-red.csv', encoding='latin-1')\n",
    "df = pd.read_csv(a)\n",
    "df = df.take([10*i for i in range(649)], axis = 0)\n",
    "df = df.apply(LabelEncoder().fit_transform)\n",
    "\n",
    "#input\n",
    "X = df.drop(columns=['type'])\n",
    "\n",
    "#output\n",
    "y = df['type']\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X,y,test_size=0.2)\n",
    "\n",
    "time1 = time.perf_counter()\n",
    "\n",
    "model = MLPClassifier(hidden_layer_sizes=(100,),\n",
    "                      activation='relu',\n",
    "                      solver='lbfgs',\n",
    "                      max_iter= 1000)\n",
    "\n",
    "\n",
    "\n",
    "model.fit(X_train,y_train)\n",
    "\n",
    "validation = cross_val_score(model, X, y, cv=7)\n",
    "print(validation)\n",
    "print('Accuracy score =', validation.mean())\n",
    "\n",
    "#Testing the test set\n",
    "predictions = model.predict(X_test)\n",
    "score = accuracy_score(y_test,predictions)\n",
    "\n",
    "#time after training and predicting the test sets\n",
    "time2 = time.perf_counter()\n",
    "timeA = time2 - time1\n",
    "print('time for training and predicting the test sets =',timeA,'s')\n",
    "\n",
    "print('Test score =', score)\n",
    "\n",
    "#Plot the learning curve\n",
    "train_sizes, train_scores, valid_scores = learning_curve(MLPClassifier(hidden_layer_sizes=(100,),activation='relu',solver='lbfgs',max_iter= 1000) , \n",
    "                                                         X , y , \n",
    "                                                         train_sizes=np.linspace(0.01,1,50), \n",
    "                                                         cv=7, \n",
    "                                                         scoring = 'accuracy', \n",
    "                                                         n_jobs= -1)\n",
    "\n",
    "#time after doing the learning curve\n",
    "time3 = time.perf_counter()\n",
    "timeB = time3 - time2\n",
    "print('time for the learning curve =',timeB,'s')\n",
    "\n",
    "train_std = np.std(train_scores, axis = 1)\n",
    "valid_std = np.std(valid_scores, axis = 1)\n",
    "\n",
    "train_mean = np.mean(train_scores, axis = 1)\n",
    "valid_mean = np.mean(valid_scores, axis = 1)\n",
    "\n",
    "plt.plot(train_sizes,train_mean, label = 'Training score')\n",
    "plt.plot(train_sizes,valid_mean, label = 'Cross-validation score')\n",
    "\n",
    "plt.fill_between(train_sizes, train_mean - train_std, train_mean + train_std, color = '#DDDDDD')\n",
    "plt.fill_between(train_sizes, valid_mean - valid_std, valid_mean + valid_std, color = '#DDDDDD')\n",
    "\n",
    "\n",
    "plt.title('Learning curve')\n",
    "plt.xlabel('Training size')\n",
    "plt.ylabel('Accuracy score')\n",
    "plt.legend(loc ='best')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "752f6f4f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.97849462 0.95698925 1.         0.98924731 1.         0.97826087\n",
      " 1.        ]\n",
      "Accuracy score = 0.9861417217658452\n",
      "time for training and predicting the test sets = 2.474482977999287 s\n",
      "Test score = 0.9846153846153847\n",
      "time for the learning curve = 9.72245999200095 s\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7fcb1f4a67c0>"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x396 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Dataset 2 : Wines\n",
    "# Output : Type of wine : red or white\n",
    "# MLPClassifier() --> hidden_layer_sizes=(100,), activation='relu', solver='sgd', max_iter= 1000\n",
    "\n",
    "import numpy as np\n",
    "import time\n",
    "import pandas as pd\n",
    "import io\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# split your datas between training and test\n",
    "from sklearn.model_selection import train_test_split\n",
    "# comparing datas\n",
    "from sklearn.metrics import accuracy_score\n",
    "# model \n",
    "from sklearn.neural_network import MLPClassifier\n",
    "# cross validation\n",
    "from sklearn.model_selection import cross_val_score\n",
    "# validation curve\n",
    "from yellowbrick.model_selection import validation_curve\n",
    "# learning curve\n",
    "from sklearn.model_selection import learning_curve\n",
    "# will be used to convert string into float\n",
    "from sklearn.preprocessing import LabelEncoder\n",
    "\n",
    "labelencoder_X = LabelEncoder()\n",
    "\n",
    "#dataset\n",
    "a = io.open('wine-quality-white-and-red.csv', encoding='latin-1')\n",
    "df = pd.read_csv(a)\n",
    "df = df.take([10*i for i in range(649)], axis = 0)\n",
    "df = df.apply(LabelEncoder().fit_transform)\n",
    "\n",
    "#input\n",
    "X = df.drop(columns=['type'])\n",
    "\n",
    "#output\n",
    "y = df['type']\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X,y,test_size=0.2)\n",
    "\n",
    "time1 = time.perf_counter()\n",
    "\n",
    "model = MLPClassifier(hidden_layer_sizes=(100,),\n",
    "                      activation='relu',\n",
    "                      solver='sgd',\n",
    "                      max_iter= 1000)\n",
    "\n",
    "\n",
    "\n",
    "model.fit(X_train,y_train)\n",
    "\n",
    "validation = cross_val_score(model, X, y, cv=7)\n",
    "print(validation)\n",
    "print('Accuracy score =', validation.mean())\n",
    "\n",
    "#Testing the test set\n",
    "predictions = model.predict(X_test)\n",
    "score = accuracy_score(y_test,predictions)\n",
    "\n",
    "#time after training and predicting the test sets\n",
    "time2 = time.perf_counter()\n",
    "timeA = time2 - time1\n",
    "print('time for training and predicting the test sets =',timeA,'s')\n",
    "\n",
    "print('Test score =', score)\n",
    "\n",
    "#Plot the learning curve\n",
    "train_sizes, train_scores, valid_scores = learning_curve(MLPClassifier(hidden_layer_sizes=(100,),activation='relu',solver='sgd',max_iter= 1000) , \n",
    "                                                         X , y , \n",
    "                                                         train_sizes=np.linspace(0.01,1,50), \n",
    "                                                         cv=7, \n",
    "                                                         scoring = 'accuracy', \n",
    "                                                         n_jobs= -1)\n",
    "\n",
    "#time after doing the learning curve\n",
    "time3 = time.perf_counter()\n",
    "timeB = time3 - time2\n",
    "print('time for the learning curve =',timeB,'s')\n",
    "\n",
    "train_std = np.std(train_scores, axis = 1)\n",
    "valid_std = np.std(valid_scores, axis = 1)\n",
    "\n",
    "train_mean = np.mean(train_scores, axis = 1)\n",
    "valid_mean = np.mean(valid_scores, axis = 1)\n",
    "\n",
    "plt.plot(train_sizes,train_mean, label = 'Training score')\n",
    "plt.plot(train_sizes,valid_mean, label = 'Cross-validation score')\n",
    "\n",
    "plt.fill_between(train_sizes, train_mean - train_std, train_mean + train_std, color = '#DDDDDD')\n",
    "plt.fill_between(train_sizes, valid_mean - valid_std, valid_mean + valid_std, color = '#DDDDDD')\n",
    "\n",
    "\n",
    "plt.title('Learning curve')\n",
    "plt.xlabel('Training size')\n",
    "plt.ylabel('Accuracy score')\n",
    "plt.legend(loc ='best')\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "12a62601",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfsAAAFlCAYAAADs50HhAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/MnkTPAAAACXBIWXMAAAsTAAALEwEAmpwYAACE8klEQVR4nO3dd5xU1d348c+9d3rb3mgLLE1BRFDBgi0So8aoKIImJKBJnhg10RjFEpUIKmpigoklJuZngiZqjKKkxxYUEWWVqrCw1IVle5te7vn9MezIsrtsm9mZ2T3vPM9Ldsq5Z86U772nfI8ihBBIkiRJkjRgqcmugCRJkiRJiSWDvSRJkiQNcDLYS5IkSdIAJ4O9JEmSJA1wMthLkiRJ0gAng70kSZIkDXAy2EuDwvz583nmmWfa3f773/+e66+/vtPn/epXv+L+++8H4Dvf+Q47d+5s95h//etfzJ8/v8s6/PrXv+bNN98EYPny5axcubKbte9aIBDgl7/8JZdddhmXXnopl1xyCc888wzJWFn7/vvvc+6553LllVfi9/t7Vcarr77K+PHjefzxx9vcLoTgS1/6El/96ldjj/u///u/DsuYP38+5513HpdeeimXXXYZF198MYsWLcLn88Ue88477zB//nwuvfRSLr74Ym6++WYqKyu7LLu3qqqqmDdvHgBut5t58+Zx8cUXs2rVqtjtkpQIMthLg8I111zDX//613a3v/zyy3z961/vVhm//e1vGTNmTK/rsG7dOsLhMAA//OEPueyyy3pd1pGEEHz/+9/H7Xbz0ksv8frrr/PHP/6RN998k+XLl8flGD3x97//nTlz5vDKK69gsVh6Xc6QIUN444032ty2fv36Hp1A3H777bz++uusXLmSv/3tb/h8vtgJxKpVq3j00UdZunQpr7/+On/729847rjj+OY3v0kwGOx1vY+loKCAF198EYDPP/+curo6/v73v3PJJZfEbpekRDAkuwKS1B9mzZrFgw8+yPr16zn55JMB+OijjxBCcMYZZ/D000/z1ltv4ff78fl8LFq0iFmzZrUp47zzzmP58uWccMIJLF++nFWrVpGZmUlxcXHsMbt37+b+++/H4/FQU1PDhAkT+OUvf8krr7zCli1beOSRR9A0jbfeeouxY8dy3XXXsX79eh555BF8Ph9Go5Gbb76Zs846i1dffZX//ve/qKrK3r17sVgsPPzww5SUlLSp18cff8yuXbt45pln0DQNgKysLB555BEOHDgARK9yv/71r/OVr3yl3d+TJk3iS1/6Etu2bePKK6+ktLSUp59+GoDy8nIWLFjAu+++y549e3jggQdobGwkEokwf/58rrzyyjZ1+d3vfsdbb72F2WympaWFH/3oRyxbtoy1a9eiaRqTJ0/mzjvvxOFwcN555zF58mS2b9/Oj370o3btPW7cOCorK/nkk0+YOnUqAK+99hpf+9rXeO+993r8GVAUhenTp7N69WoAfvGLX7BkyZLY+6coCt/97ncpKipqF+w3bNjAo48+SjAYpKamhtNPP50HH3yQcDjMkiVL+OSTTzAajQwbNoyHHnoIs9nc4e0NDQ1ccskl/PWvf+Wuu+6iqqqKSy+9lMcee4wrr7ySTz/9FICnnnqK//znP+i6ztChQ7nvvvsoKChg/vz5ZGRksGvXLq6++upu9ShJEsgre2mQMBgMXHXVVbzyyiux21566SWuueYaDh48yAcffMCKFStYtWoVt9xyS7vu4yO9+eab/Oc//2HlypW8+OKLuN3u2H0vv/wyl112GS+//DL/+c9/qKio4N133+XrX/86kyZN4vbbb28T1BoaGvjBD37A3XffzapVq3j44Ye57bbb2L9/PxAN5Pfccw9/+9vfOPHEEzscitiyZQuTJ0+OBfpWI0eO5IwzzuiybUKhEOeeey7//ve/ufrqqyktLaWmpgaIdmXPnj0bIQQ/+MEPuPXWW3n11Vd5/vnn+f3vf8+GDRvalPXtb3+b8847jwULFrBo0SKeeuopqquref3113n99dfRdZ1HHnkk9vixY8fyz3/+s12gb3XZZZfx+uuvA+Dz+SgtLWXmzJldvqaONDU18c9//pPp06fT0NDAgQMHYicRrRRF4Wtf+xoOh6PN7X/84x/5wQ9+wF/+8hf+/ve/8/bbb7NlyxY2bNjARx99xBtvvMGrr77K8OHD2b59e6e3txo9ejRLly5lxIgRvP7665jN5th9K1eupKysjL/85S+8/vrrnH322fzkJz+J3e9yufjHP/4hA73UI/LKXho0rrrqKi6++GLcbjfhcJj333+fxYsX43Q6eeSRR1i1ahV79+5l48aNeDyeTstZu3Yts2bNigWEK664ghUrVgBw2223sWbNGn7729+yZ88eqqur8Xq9nZa1adMmRowYwYknnghEg9/UqVP56KOPUBSFiRMnUlhYCMDxxx/Pf//733ZlqKra57H51t4Oh8PBrFmzeOONN1iwYAGrVq3ihRdeYM+ePezbt4+77ror9hy/389nn33GlClTOi139erV3HLLLRiNRiDao3DDDTe0O25nLrnkEi699FLuvvtu/vvf/3Leeee1O6k5lkceeYSnnnoq1j7nnnsu3/zmN2Pvr67r3Spn2bJlrF69mqeffppdu3YRCATwer1MmDABTdOYM2cOZ555JhdccAGTJ0+mubm5w9srKiq6PNY777zD5s2bueKKK2J1PHKeQVdtJkkdkcFeGjQKCgo4/fTT+cc//oHX6+WCCy7A6XSydetWvv/977NgwQLOOOMMTjnlFH76058es6wjg+uRwedHP/oRkUiECy+8kHPOOYfKyspjBuJIJIKiKO3KDofDGI3GNmPeiqJ0WNaJJ57IH/7wByKRSJu6bNq0iRUrVvDoo4+2q3MoFGpThs1mi/37qquu4p577qGkpISSkpLYVanT6YxdZQPU1tbidDo7fW0QDVRHvj5d19sc+8jjdiQvL4/jjz+e1atXs3LlSu644w4aGhqO+Zwj3X777bGhiyNlZGQwcuRINm7cyOmnn97mvh/+8IftJm1+4xvfYPz48cycOZMLL7yQjRs3IoTA5XLx+uuv88knn/Dhhx9y8803c9111/H1r3+9w9vPPvvsLuus6zrf/va3ueaaawAIBoM0NTXF7u+qzSSpI7IbXxpUvv71r7Nq1SpWrlwZm5j38ccfM2nSJBYuXMipp57KW2+9RSQS6bSMs846i3/96180Nzej63qbAPj+++9zww03cNFFFwGwcePGWFmapsUm6LWaMmUKu3btYtOmTQDs2LGDjz/+mFNPPbXbr+mkk05i9OjRPPTQQwQCASAaiJcuXcqwYcMAyM7OZsuWLQDs3LmzTZfy0Vqv1J944gnmzJkDwKhRo7BYLLHXWllZyVe/+tVYmZ2ZOXMmf/7znwmFQui6zgsvvNCtoYUjXXbZZfy///f/aGlpYdy4cT167rHceOONPPDAA+zduxeInng9+eSTbNu2jdGjR8ce19zczObNm/nxj3/Ml7/8ZQ4dOsS+ffvQdZ133nmHBQsWcNJJJ3HTTTdx2WWXsWXLlk5v744zzzyTV155JTY8tHz5cm6//fa4vW5pcJJX9tKgMn36dJYuXUpGRgbjx48H4Ktf/Sr/+c9/uPDCC9F1nXPPPZempqY2Y/FHOvvss9m+fTtXXHEFLpeLCRMmxK42b7nlFm644QZsNhsOh4NTTjmFffv2AdEJfo899libK9vs7GyWL1/OkiVL8Pv9KIrCQw89xKhRo2KTtbrj8ccf5xe/+AWzZ89G0zR0Xeeyyy7juuuuA+D666/njjvu4H//+x+jR4/usit4zpw5PPnkk5x//vkAmEwmnnzySR544AF+97vfEQ6H+eEPf8i0adOOWc7111/Pww8/zGWXXUY4HGby5Mncc8893X5dAOeffz733Xcft9xyS4f3v/fee5x00kmxv51OZ2wS3rFccsklCCH40Y9+RDgcJhAIMHHiRP7whz9gMplij3O5XHz3u9/l8ssvx2azUVBQwNSpU9m7dy9z5sxh9erVfPWrX8Vms5GRkcGSJUsoKirq8PbumDNnDlVVVVx11VUoikJRURHLli3r1nMlqTOK3OJWkiRJkgY22Y0vSZIkSQOcDPaSJEmSNMDJYC9JkiRJA5wM9pIkSZI0wCV0Nv7GjRv52c9+Fks40urtt9/miSeewGAwcMUVV3DVVVeh6zqLFy9m+/btmEwmli5dSnFxMVu3buV73/seI0eOBODqq6+OLWvqiK7reDwejEZju/XLkiRJkjQQCSEIhULY7XZUtf11fMKC/W9/+1veeOMNrFZrm9tDoRAPPfQQr7zyClarlauvvppzzz2XTz/9lGAwyEsvvcSGDRtYtmwZTz31FJ999hkLFy7k2muv7dZxPR4PZWVliXhJkiRJkpTSxo0b12Gyq4QF+xEjRvCrX/2qXTKI8vJyRowYQUZGBgDTpk1j/fr1bNiwIZbzesqUKbEEFFu2bGH37t289dZbFBcXc9ddd7XLW32k1rSc48aNa7NWti+2bNnCpEmT4lLWYCbbse9kG/adbMO+k20YH/Fsx2AwSFlZWSwGHi1hwf6CCy7oMA+02+1uc9Zht9txu9243e42Qbw129jkyZOZM2cOkyZN4qmnnuKJJ55g0aJFnR63tes+3lf33c1+JR2bbMe+k23Yd7IN+062YXzEux07G77u9wx6DoejzSYjHo8Hp9PZ7nZd1zEYDMyaNQuXywVEtyntbhaqSZMmtdlJqi9KS0u7zBQmdU22Y9/JNuw72YZ9J9swPuLZjoFA4JgnDv0e7EtKSti7dy+NjY3YbDbWr1/Pddddh6IovPPOO1x00UVs2LAhlgP7uuuu45577mHy5MmsXbuWiRMn9neVJUmS+lU4HO72jnzJEgwGk12FAaGn7aiqKgZDz0N3vwX7VatW4fV6mTt3LnfccQfXXXcdQgiuuOIKCgoKmDVrFmvWrGHevHkIIXjwwQcBWLx4MUuWLMFoNJKbm9vtK3tJkqR01NLSgqZpvfpB7y8lJSXJrsKA0Jt2DAaD+Hy+LnecPFpCP03Dhg3j5ZdfBqKbTrQ677zzOO+889o8VlVV7r///nZlTJw4kRdffDGR1ZQkSUoJ4XAYTdNSfhvbUCgUtwnQg1lv2tFkMuH1egmHwz06IZRJdSRJklJE61ylnlq1dT9/+6z9hGhpYGrd2bInUrefSJIkSeqSPxThlpXrURQ4f2wRFqOW7CpJCdabhHEy2EuSJKWxh9/ewu56NwCPvL2Fey84sU/lLVu2jK1bt1JTU4Pf72f48OFkZWXx+OOPd/ncZ555hhkzZjB58uQO73/ggQdYuHAhQ4YM6VXddF3n4YcfpqysDFVVMRqN3H333QwfPrxX5Q0mMthLkiSlqV11LTzy9tbY3w+/vZX5J49mVE7PJm8d6Y477gDg1VdfZdeuXfz4xz/u9nO/+93vHvP+u+++u9f1Anjvvfeorq7m//2//wfAm2++yYMPPshTTz3Vp3IHAxnsJUmSUtTtq0p5ZePeTu+vdvvxhyOxv/3hCCc8uop8h6XT51x5YjGPXNLztd133HEHjY2NNDY28vOf/5wHH3yQQ4cO0dDQwFlnncXNN9/MHXfcwUUXXURtbS3/+9//8Pv97Nu3j+985zvMnj2b+fPns3jxYv7xj39QUVFBXV0dBw8e5M4772TmzJm88847PP744zgcDjIyMhg/fjw33XRTrA6FhYVs2bKFf/zjH8yYMYMvfelLnHXWWQC88847/PrXvwbg+OOP56c//Slr167ll7/8JWazmczMTB588EE+//xzfvazn2E0GrnqqqsYMmQIv/jFL9A0jeHDh3P//fd3moUunckJepIkSVK3zJgxgxdffBGv18uUKVN49tln+fOf/8yf//zndo91u9385je/4amnnuKZZ55pd7/JZOJ3v/sdd999N8899xyRSISlS5fy29/+lhUrVnSYFG38+PEsWbKEN998k69+9atcccUVbNiwgXA4zJIlS3jmmWf461//SkFBAZWVldxzzz38+te/5vnnn+eUU06J9QAEAgH+9Kc/cemll7Z5TEFBAa+99lr8Gy4FyCv7JApHQhi0gXcGKUlSfDxyybRjXoXvqmvhhEdWxa7uLQaNzbdd0qdu/GMZNWoUAC6Xi82bN/Phhx/icDg6TAwzYcIEAIqKijq8/7jjjgOiV+vBYJD6+nocDge5ubkAnHzyydTW1rZ5zrZt2xg1ahSPPfYYQgjWrFnDzTffzMqVK3G5XOTk5ABw4403xsorKCgA4JRTTuGxxx7jnHPOib2O+vp6qqurufnmmwHw+/2cccYZfW2mlCSv7JNACJ0GzyGqmnYRDPuSXR1JktLU6Bwnt5/3RVbRRedNTFighy9mga9atQqn08nPf/5zrr32Wvx+P0KIDh/bVVmtcnJy8Hg81NfXA9Et0o+2du1aHnvsMSKRCIqiMHbsWKxWK7m5uTQ3N9PY2AjA0qVL2b9/P263m+rqagA++uij2FbprVvAZmVlUVhYyJNPPsmKFSv43ve+x/Tp03vWKGlCXtn3s0DIS4OnCl1EUFUDjd5q8l3Fya6WJElpatF5k1ixfheKAref1z870Z166qncfffdlJaWYrVaKS4ujgXV3lJVlXvuuYfvfOc7OJ1OdF2nuLjtb+P8+fN5+OGHueyyy3A4HKiqyiOPPIKqqtx333383//9H6qqcvzxxzN58mSWLl3KTTfdhKIoZGRk8NBDD7Fjx442x7z77rv57ne/ixACu93OI4880qfXkaoUcfTpWJpr3Qwg1TbCEUKnyVeDJ9CMqnzRoSIQuCw5OCxZfa1mypObZ/SdbMO+S+U2bO3u7mlWtVVb96MoCl89flgiqtWOx+PBbrfHvdzf/OY3LFy4EJPJxI9//GPOPPNMLrvssrgfJ1X0th07+px0FfvklX0/CIR8NHoPEdEjbQI9gIJCi78Om9mFqshkGJIk9dwlEwfGOnO73c5VV12FxWJh6NChXHTRRcmu0oAhg30CCSFo8tXgDTSjKEqnY1hCQKOnmmxHUT/XUJIkKXV84xvf4Bvf+EayqzEgyWCfIIGwj0ZPFRE93K2JKr6wm0DIi9mY2htgSJIkSelHzsaPMyEEjd5q6loq0EWk2zmMVVQavdXtZrRKkiRJUl/JYB9HwbCP6ua9eANNKErPmzash3D7GxJQM0mSBrL9dZ+xv/7zZFdDSmGyGz8OhBA0+2tx+xtRFbVXgR5AVVRa/HVYTU6ZbEeSpG4J6yHW7fobigJFmWMwqPK3Q2pPXtn3UTAcoLplL57Dgb6vFEWlyde39aqSJA0em/e/iztQT4u/ni0V/4tLmTt27OC73/0u8+fP54orruDxxx9P6BDj7bffziuvvNLmtueee45f/OIXHT5+/vz5lJeX8+qrr/LWW2+1u7+rLHj//e9/qaqqoqamhsWLF/e63gB79+7lu9/9Ltdddx3f+ta3ePTRR3u813x/kMG+l4QQNPvqqGnZj65Hen0135FAyIsv6I5beZIkDUwt/jq2HPgiwG+ueJcWf32fymxubuZHP/oRd911FytWrODll1+mrKyMF198sa/V7dRVV13F66+/3ua21157jTlz5hzzebNnz+ZLX/pSj4/3xz/+EbfbTV5eXp+D/WOPPcY3vvENnn32WZ577jn27NnT4QlIsslu/F4IhgM0eg8RigTjcjV/NEVRafLWYDHauz3BT5Kk9BIM+/EGmoEvrphDoTB2cyZqJPq9L93zL/bVb+2kBPCH3ET0cOzviB5m5SePYTE6On3OyNzJnDKq8/Xrb731FtOnT4+lltU0jYcffhij0ci6dev42c9+hqqqXH311eTl5bXbVS4cDnPzzTcjhCAUCvHTn/6UkSNH8sMf/hC3243f7+e2225rk5b25JNPpr6+ngMHDjB06FA2bdpEbm4umZmZ/PCHP6SlpYWGhgbmzJnDNddcE3ver371K3Jzc7nqqqu455572LlzJ8OHD48lnSkrK2PZsmXouk5zczM/+clPaG5u5vPPP2fRokU8+uijLFq0iJdffpk1a9Z0uEPeb3/7W4xGIxUVFVx00UVcf/31bdpryJAhvPbaa9jtdiZPnswvf/lLDAYDuq6zdOlSNm3aRCgU4qabbuL8889n2bJllJaWAvDlL3+Z73znO212FPzNb37D7373Oz7++GOEECxYsIALL7yw0/eru2Sw7wEhBG5/Ay2BOhTUhAT6VrqI0OyrJcOWl7BjSJKUPI3eqjaBGiAc0rGanOiiddtaEU3E0ZmO7upjb3t1dTXDh7dN0nNklrdAIMCf//xnbDYbX/rSl/jzn/9MQUEBf/jDH3jqqaeYPn16LG/+zp07cbvd7Nu3j9raWp577jnq6urYs2dPu+NeeeWVvPHGG1x//fW8+uqrzJs3j71793LxxRfz5S9/maqqKubPn98m2LdavXo1gUCAl19+mYMHD/Lvf/8bgJ07d7Jo0SLGjx/PqlWrePXVV1m6dCnHHXccixcvjm1lK4TgnnvuafdazjnnHA4ePMgbb7xBMBhk5syZ7YL9Lbfcwp/+9Ccee+wxysrKOPvss7n33ntZt24dDQ0NvPLKK9TU1PD888+jaRoVFRW8/PLLhMNh5s2bF9uid8aMGSxYsID//e9/VFRU8OKLLxIIBLjqqqs444wzcLlcfXpfZbDvpnAkQL3nEGE9hNIPox+KouAJNmI3Z2DQepY6U5Kk1OYNNBOOBLsc/jup+MucVPzlTu93++v5+8YniYjoSYOmGLj4xBvIsOWhqb37eR8yZAifffZZm9v279/PoUOHgC92vmtoaOhwV7nbbruNPXv28P3vfx+DwcD111/P2LFj+frXv86PfvQjwuEw8+fPb3fcSy+9lAULFnDttdfy0Ucf8ZOf/IS6ujr+8Ic/8J///AeHw0E4HG73PIjOMZg8eXKs/kVF0QRl+fn5PPnkk1gsFjweDw5Hxz0enb2Wc845h3HjxmEwGDAYDFgslnbP/fDDD1mwYAELFizA4/Hw8MMP8+STT5Kdnc2UKVMAyMvL45ZbbuF3v/sdJ598MoqiYDQaOeGEEygvL2/TrmVlZWzdujXWRuFwmIMHD/Y52Msx+26qaamIJsih/7rVFVQavFX9djxJkhKvdb5PPOb5OCzZHDfki8loxw89E4cli4ge7vWEunPPPZf33nuPffv2ARAKhVi2bBllZWVA2x3jOtpVbt26deTn5/P73/+e66+/nscee4zt27fj8Xh45plnWLZsGUuWLGl33OzsbEpKSnjyySeZNWsWBoOB3//+90yZMoWf/exnfOUrX+n0NY0ePZoNGzYghM6Bg/upqor+bj7wwAP84Ac/4OGHH2bcuHGx5yuK0qaszl5L62OP5dFHH2XNmjVAtAdk1KhRmEwmRo8ezebNmwFoaWnhuuuuo6SkJNaFHwqF2LhxY2yzn9bjjB49munTp7NixQr+8Ic/cOGFFzJsWN/3PJBX9t0moB8DfatQ2I830ILNnLhtKyVJ6j9ufwO6CMdtUu/EoWeyu3YjCgrHHxH4dRFBU3r+E+9wOFi2bBk/+clPEELg8Xg499xzueaaa/joo49ij1MUpcNd5RRF4ZZbbuEPf/gDqqpyww03MHLkSJ544glWrlyJ0WjkBz/4QYfHvuqqq/jOd77Dv/71LyB64rF48WJWrVpFZmYmmqbFxuOPdP7551NaWsqcOVdRNKSIzKxMInqYr33ta3z/+98nJyeHwsJCGhqieUxOOukkbr/99thJR2ev5cgd8jrzy1/+kqVLl/Lzn/8ck8nEsGHDWLx4MXa7nbVr13L11VcTiUS44YYbOPvss/noo4+YO3cuoVCI888/n4kTJ7Yp77zzzuOjjz7immuuwev1cv7553faI9ETcte7bigtLWVISQbJCPYAKAqFrpFxnfGfDKm821i6kG3Yd8lsQ13oVDXtorPfknBIJ8te0ONd7yoatqMAQ7PGx25TUDBopoRM8k3Urnd9oesRwnqozW2qoqGphpSd6Cx3vZPaEHp0e9xMW0GyqyJJUh80+2oQAuIde4YdEeRbCUSvr+7TjRCi3WRHiPZuCF3HoCbmpCedpPel4iChKAqeQDPBcCDZVZEkqZfCkSDeYHO/Bh1djwyK/TZ0EUF0sgxBCEE4EkQXqZfopj/JYJ8mVEWlUU7Wk6S01eSr7ZeVPEdqvbofyDq7qm/zGKIBv6vHpYvenMDJYJ9GwnoQd6Ap2dVIWeFIgHp3JXXug7IXREopgbAPf9jT5eNUlU6Xl/XWQL+670kAj+hhwpFQ0ttDFzqhSKDT3oiuRCKR2KqI7hr4gzkDiIJCi68Gm8mBqmjJrk7KCIUDNPvr8Ic8sURHNS17MRtsOK05mA3WJNdQGuyavDWo3bi2UjUVt68RAIMhfj/PYTXS63X3HQmFQh3Oiu9vutAJR3peD0VRkjKOrwv98LLI6JBCKBTpUTsKIYhEIkQikR5/PmSwTzNCQKO3mmx7Ub8dUxc6/pAHq9GeUisCgmE/Lf46AiEvitI2o6GqaIQiAWpbKjAZLLgsOZiNtiTWtn+EIyFAyERMKcQbbCYcCXT7u6OZBZ5QHXocO6dURSXPNbzrB3ZTeXk5J5xwQtzK6616d2Wfuuaz7AWYDO0T5cRbOBKgxd9AIOxFIXqhJkSEyr31nHTiyd0uR1EUTCZTr04EZbDvhvcqWsiJ6Jw/LjtuZf63rB4Felymoij4gm4CJh9mY9sr1lVb96MoCl89vu8JGCB6FunxN/LXzdG1pl8el43ZYMduycBs6Fvg7EtdA2EfLb46AmFfuy2Fj25XVVEJR4LUuSswaBaclmyspr6vWe2r+L9XOk2+WrzBZoyamTxn/H7YB7u+vFdCCFo6SKDT1fdf1VTUXnTedVauEDph4cdm7nkWts5ef0+XB3anzJ7wBppADWPQOv/+d6U5WIVLy8dhzkhIXcORIE2+2livY2t6XoB/b6un/pCH6af0z4m5DPZd8Ici/Lz0EEaTypmjM7EY+n5l6w/r3Pfv3SgKvSpTVVSavNXkuUbEuqH8oQi3rFyPosD5Y4uwGPvWze8Lumn21eIOBlj87z0oCpw1Ogvw4mtpwaAasRgd2M2ZGLSe7Z/d27oGwl6afXUEw35Upf3eBMdqV0XRiOgh6j0HMfrNOC05SQv68X6vPIEmmn21CCFQFIVQ2I8v6E6Jk5p019f3yu1viGbePOKz2tfvf6d1PebnX8Xtb+hxsI/3ZzVeZbae3Pa1XRVUmr3VhMJ+Mm357br1e1vXcCREs78WX9Dd6W/V4v/uJRKK8P2LI3Fp164kNNhv3LiRn/3sZ6xYsaLN7W+//TZPPPEEBoOBK664gquuugpd11m8eDHbt2/HZDKxdOlSiouL2bt3L3fccQeKojB27Fjuu+++Hk9M6IuH397CQU8IPPDkmgP86Oy+XzE9saaCfY3RPrrelhnSg7j9DTit2bF67q6Pbov7yNtbuPeCE3tVt3AkQKO3hmDYh6KoPPXBwXZ1VRUNXeh4Ak24Aw2YNCtWkxO72dWtrsqe1jUQ8tLsryUU9qMoWqcbEHWnXVVFI6KHafBU0uIzYbdkYe/gB1DXBb5QmEZ/kL9tPYAuBF8ePwS7yUCm1YTZoPZ6vC9e71Ug7KPJW00g5CcQEXiCYd4sa0AHZo7047TGp9fg3fIqFODskvjleUhEma3lqorCVyYMwWk24rIYMRl6/0Pal/dKFzotgfp234l4fP870lW5YT2IN9iCzdT9bJzx+qzGu8ymI05uW/W2XRVFxRdsJhTxk2MfiqZ9ERZ7WtdIJBwN8qGWY26W9sSaCvYfrmu82rUrCQv2v/3tb3njjTewWtt2NYdCIR566CFeeeUVrFYrV199Neeeey6ffvopwWCQl156iQ0bNrBs2TKeeuopHnroIW6++WamT5/Ovffey1tvvcWsWbMSVe02dtW18MjbX2wv+cQHB7hych4jsno/xrO3wc+THxzsc5mqotLir8dmcrG30demng+/vZX5J49mVE73v9S6iNDkrcEbbIl1jXdVV0VRUNAI60GafTU0+2uwGBzYzRmdjo8f3abHqqsv6KbFXx8b71SOMSmxp+2qKCohPURl0wG8wQOoqoOIsBEI6/hCOsFIGF0oCKGz5D+bQIGSHAeaqhLWdTRFwWzQsBg1rEYNs0HDZtLItJiwGg2oascnAj15/UcKRXRa/CEOugOYD9bS7KvBE2whrCuEIjoKoAv4+f8OoCgwLsdIS7AGo6Fvm2cEwxEeenMLKDAm19mnwJnIMo8utyTHgaooRAQYVAWLQcV8+L2qaAkwtNnX5Ulbb9+rVs2+mnZZtuP1/T9ad8qNXt3XdzvY9/X1J6rMaL6CpjYnUX1tV0VRiehhqlv2km0fgtlo7VFdW3cpjW5ZrBzzoufousajXbsjYely//3vfzN+/Hhuv/12Xn755djt27Zt49FHH+XZZ58F4MEHH+Skk05iw4YNTJ48mYsvvhiAmTNn8t577zFz5kxWr16Noii8+eabrFmzhvvuu6/T48YzXe7Xnn2bv392oM1tTrNGcR+DfUug7brXvpTZGpSb/W3TRLosRkq69eERRPRIh2txe1/X6C9ctPvqi1SVXq+XSl+ky7rqIkJEjwA63U1R3FFdR2ZZ+NbJhWRaNZxmFZdFw2FWMSgQjOiEIgIFgUFTEUJHUTQMmguD5ozV+Q8fl/PH9bsA+NbJo/nmKSWd1kEXglBEoCoCo6oR0iO4g2GafWGa/UFaAmF+/9HO2JVCqxGZdr51ymhyHRZcZuPhK1IDdpOBiC4IRPRYQN+zexPDR+R02C4vfFLDnzfUAXDNSblcc1IuFtOwPs047snrT2aZPSm3rKyMkjFjYydtRk3FF4rgCUbfpyZ/iOZAiD98VM6ehrbL5br7vRJCENYDHP0+xfv73/NyBZpq7NZqnvK6lk6/q16vF5ut5/N2jlVmd4UjwXZL1uLdrqqqsafe12VdBQJdD/coYU9Hdb34+KG8cd15vaprq6Sly73ggguoqKhod7vb7cbp/KKx7HY7brcbt9vdJtm/pmmEw+E2XTV2u52WlpZuHX/Lli19fAXQ1Nh+TbsnGKG8tuv1sp0JtI+pCSnTHQixraqhG8/u/FwvvnVVDpfZ/nht69q7c8+O6rqnwc9P/7un3e1mDVxmhQyTgsuskhH7t0KGGVwmFZfRij9s5s+fHIo970+f7GaUwY9JU2gKRmgORNr+NxihKRD9b3MgQribL2Vfo4cl/93c4X12g4rLrOEyKWRadFwmhYx9zYfrquAyRevvDuq8sskXe95fNtZyvN1DjqUGRO/G7mt8If78SWWb1z/G5CfX2rM5Gokus7NyizUfRk05/J7osfemKRiheUN17L1rCUXQu/ledf97BR19lhPx/e9duV2fAHb5XW3p+XKBrr//3dE/7dq939We/151VG5TY1NsN7xE6fcJeg6HA4/nizfA4/HgdDrb3a7rOgaDoc34vMfj6faevvG4sv9/I8dxwiOr8Iej747ZoPLO96b0uRv/vKc3EAjrcStzf2OQc5/6NFZPi0Fjy+2XdNotFAwHaPJVEwz7jnmGH8+6CqFTtmMHrsLxzHziQ/yHy7QYND66+RxybUEieqjXS/uOrqtJU7j7S8PxhyLUekM0+SLU+8I0+MI0eKP/3dUUOcYZua/dLSFdsPjDgx089gsGVSHbZqbEZSfHZibbZiLbZibbZsZpNmA2qNz/n80EI62vX+V3c0+nxR9kT4OHem+QRl+Qem+Aem+QOo+fem+ASs+xs6AptP3ZCemw7KMAxVmgKuFeJWPfU+8mdEQUDOmCR0qrGZXd+4l/uzspc2QfyuysrvevqzzGM6Kf52ybmeHZZrJsptj7lWUz4TIbsRoN/OSfG454r479vWoVCPuobanocLw2Ed//npariwg5jqFYjMfegGVXXUub378jX39vNxM6VpldEUJQ3bynw+9sItp1T72X83+zqc1v1ebbvkquPYQ70AhHzRnorqPrajFoPLfg/D5347de2Xem34N9SUkJe/fupbGxEZvNxvr167nuuutQFIV33nmHiy66iA0bNjBu3DgAjj/+eNatW8f06dNZvXo1M2bM6Le6js5xcvt5E7n/P5sAuOH0oX3+UhZnWfj+6UP4xeqKuJU5LMPAD2eO4uF3dgKw6LyJnY4rNXpr8AdbDq9LP3ZXXjzrqigqCgpFLrj+tCJ+8V50eOSGM4aSZfWii2OPc3Xl6LreeMYwrpsenaDm9oeo8wZo8ocIRXS0w+PpEV3Q7D98EuANx/7b4AtT74uw6aAHd7DtD4vLbODk4bltgkO23Uy21Uy23YTTbEQ94gcgGNZxmA3k2MwMy7RhNxsJRkTsM7XovElcPXUUEJ0UeKjFR7XbT02Lh3CkAYEPBZVgWKfRH2HbnoMYndlf1NMbpsEXYeshL55Q27q6gzqfVfno6MSlOzq6ZmkJhNlU2dir8jrTEgizpY9ldlRXl8XA9BF5ZNvM5NhMZNnM5NhNNFVVMm3ieGxGLfZjLYQgpAsyrUby7BaGZtgwGTTcwfAR71XH36ujNftqOp2YlYjvf0/LVRWNFl99l8H+6N+/7r7+RJXZ4q9vt7KhVSLadWS2jeuPKPPms0ZjNdTg9h8O8r0cHju6rvFo1+7ot2C/atUqvF4vc+fO5Y477uC6665DCMEVV1xBQUEBs2bNYs2aNcybNw8hBA8++CAAixYt4p577uGxxx5j9OjRXHDBBf1V5ejxz5vE79Z8jtGk8v0zhsalzBvOGMZfN9WgKMSlTEVR+faMHF7aWImqKNx+3qQ29wshaPHX4w40gKBHQTXedQW44czh/HVzLYoC158+BCVOWwd3VleHxYjDEu0mdgdC1HmigT+CIMtmIMtmgJz25R1qDvL913YTPNz1aNIUfn7JMIpcVhTFiKoa0RQLqmpp06bBsI7TbCDHbmFYhg2bue3XbNF5k1ixfheKQpv3SlUVilxWHCYfuVY/DV6NRr+ZZn8Ig6aQ7zASzNQYNqz9D0NHdX1y9igKnSYEOmZjEZras/W8B5u9XPfi2tiVrUlT+f280yly9T4jYSLK1PUIe+v38f1Xy496r4YzImsIBq3t2HKZuxa7yYAQgmBEjwX4YZl2jFrb70Zn71VnvMGWw0tDOz+RTsR3qqflhiJ+AiFvl4mmevr6u6M3ZUYiYdz+hmP+diXkt+pwmQDXnZoFEJesezecMYxXNtUQCUXi1q5dSWiwHzZsWGxy3iWXXBK7/bzzzuO889pORlBVlfvvv79dGaNGjeL5559PZDWPyWLUuHVaITlF9rith7UYVH56wSiUw/+OB6tB4/4LRpNhy22zZtMbbKHZW4suDp8R9/Bzmoi6JqLM7pbrMBtxmNsG/mZ/iOARV/ytCl0mrjghOzbp7crJORS5LIBAiCCRSJCwiM4hiegqDrOVbLuDEVnZZNocnf4wWYwav7jsZBRF6fS90lSVXIeFXIcFIaDBF6TBG0AIgS5Em96Dzupa6IwGdwWVcKQBTe3ZMrchLhvzThoZm/R29Ukj+xSUE1GmEDrB8CEKHFr798ppJhiuBXJjAT86cU7HbjKQ5zAzNKN9gD9SZ+9Vx3WJJtDpqscsmZ//Voqi0uyvJc844thl9uD1d7uevSizyVfTZZBN9G+V1Ri/cGkxqCyeVUz9oeZ+WWMPCZyNnyzxnI3fqrS0lCElGfQ4UvYzXUTIshdhMzkJhn00+WoIhbufpjPRysrKYsMzqcYTDFHrjgb+QFjHoEXf62BY5/pXo4k6nrx8FKYjfkTCusBm1MiwmshzmDFp0S+trkdAUTBoJoyqCYNmwmK0Y9TMHf5gdXcOBcD27dvJGz6KBm+A5kAIXRexk5Rj1RV0TIZ8NK1ngTUYjnDti2tBgd/PPS1uS+/iUaYQOoFQJboeQVGUTl+/LnRUNYdcRyZ5DjNV5duZfkr3U5R2l9vXQJO/ttMu/FQjRIQcx7BepZHu7Zh9bwRCPmo9Fd3aWyCd6HqEivJapp9yelzKS9psfKn/qYpGs7cGf9CNP+SJroNPkx+eZLObjNizo1f83mCYWrefpkAIVVH47ox8FBRMBrXTAH8kVW0N+mECephA2EuLvw6FwycAmhmjZsZksOEJNMRyG3RnOZSiKIcn/UWv2luv+Jv9ITS1bV2PqhWhSEOPg73JoPH9M8ejHP53PMSjzGigPxQL9NFy1djrNxpUIrrAYTGQabGSaxPkuqxYjHbqdsf/pF0XOs2B+rQJ9BDNKtnsryMvxfeMaPJVD7hAnwwy2A8wAhHdbKGfd3MaSGwmAyMOzw73BsPkOy34QmEcZmOnAb4rrYE8ooeJ6GF8QTeCmmNm2eqOLKuJLGs08Df6gmTbzfiCHc/cF+ioeDEaejYZ6Nwxhb2uXyLKFELHF6jGoAiUozYEmTkqE4MGGRYTuXZLm6GZes9Bsu1Den3cY2n21UZ3qUqz710w7CcQ9qXszpBufwNhPRS3eT2DmQz2knQMNpOBEab4f01asw/GU6bVRKb12JPwFKAgIydte3x0oVPbsp9wxNHjE1oFlXrPQUK6P651CkdCeANNadmmqqLS4qvD7IxPauV40oVOs79eBvo4Sb9PpyRJvaYLnRZ/TxKYpI4vAn2o1z1XCip+0Yw/1LckNkeKTh5L35/SQNhHMBzfE6B4aPLWRHtLpLhI30+oJEk9pigK7kBDh+mRU5mIQ6BvpQD1nsq4BPxg2BfXE4dkUBWVZl9dsqvRRjAcwBtslsORcSSDvSQNNuLwVVOaEEKnJk6BvpWCEpeA3+RLn9n3xxIIewiGe57+NlGafNUDol1TiWxNSRpkFEXBG2whHEmdH/fOJCLQt+prwPcF3QTDvctOmGpURaPFX5vsagDgDbQQSsFhhXQng70kDUKqotLoTY0f985EA31FQgJ9KwWFBk8lgVDPgrYQgmZfbbeWS6YLf8ib9BNAIXSa03wORKqSLSpJg1Qw7CUQ8ia7Gh36ItAH+2HcVqHec6BHAd8TaCQiwgmsU/9TFZWmJI/dt/jqe7RdrNR9MthL0iClKCpNvtQbu+/fQN9Koc7dvYAvhE7LAF0S5g+5CUeCSTl2OBLCHWiQk/ISRAZ7SRrEwpEg3kBzsqsRI4SgtuVAPwf6KEXpXsBv8tUywLKMx6iKlrSZ+U2+atl9n0CyZSVpEFMOL7tKheDVGuhDkUDSru5iAb+TiXeRSPhwAp2Be/XpC7kJR0L9ekx/yJOyQ0oDhQz2kjTI6SK6fWgyCSGocx8gFPEnPZC2BvyOZtqnewKd7lAP74jXX4QQNHkHfrsmm2xdSRrkFEXFHUjexKjWQB8MJz/Qt1JQqD0q4AfDAfwhdxJr1X/8/Xh135r/XkosGewlSUIkKdFONNAfJBj2pUygb3V0wB9MY8oKKi3+xI/d6yKCO812C0xXsoUlSUJRFHyh5n6diS2EoN5zkGDYm7JBVEGhzn2QFl/9oEv04gu2EIkkdnlho7dGpr/vJ6n5DZMkqd8pqDT5+mes1h/yUNOyj0AodQP9kZr9tWlRz/hSaE7A1b0udHzBFpq8NfiD7pTr0Rmo5Ba3kiTF+MOehO5vHgh7afbVEgoHUBQ1bQLoQMqU112KouALNpNhy+316xdCEIz4CRzOzheMBIlEgqAoqIoqA30/ksFekqQYFZUmbw35rhFxLTcY9tPsqyUQ9h3+kU+PIC8pNPtqybQVdOvR4UgIf8hDKOInpAcIh4MIRJuTBVUdfCdOqUAGe0mS2ghHAniDLdhMzriU1eSrxR/yoiqqnIiVZhRFwRtoxmVtf3WviwiBkJdg2E8o4icYCSL0MIqixa7YFUUdgHkG05MM9pIktaEoKi2+WqxGR6+7WcOREM2+WnwhtwzyaU+h2VeH1eQkEPISCgfwRGqobCiPdcdHHwWKKkNKqpLvjCRJ7UT0aKIdpzW7Z8+LhGn21+ILtaAgg/xAEL26b8QTaIxd3QuE7I5PMzLYS5LUjqKotATqsVsyuxWw9cNbk0bz7CtyTH6AURRNdsenORnsJUnqmODw5Kz8zh8idJp9dXiCjSBkkJekVCWDvSRJHVIUBW+wGYc5C4NmbHNfdJvXBtyBRhAiGuTlpZ8kpSwZ7CVJ6pSCQpOvhhzHECC6btrjb6QlUI8Q+uEgL6O8JKU6GewlSTomfyiaaCccCdLiqyciwnKtvCSlGRnsJUk6JlVRqWupAEVBQZEz7CUpDclgL0lSl+RVvCSlN/kNliRJkqQBTgZ7SZIkSRrgEtaNr+s6ixcvZvv27ZhMJpYuXUpxcXHs/pUrV/Lss8/idDq5/PLLmTNnDsFgkDvvvJP9+/fjcDi49957GTlyJFu3buV73/seI0eOBODqq6/moosuSlTVJUmSJGlASViwf/PNNwkGg7z00kts2LCBZcuW8dRTTwFQX1/P8uXLee2113C5XCxYsIDTTjuNd999F5vNxssvv8yuXbtYsmQJzz77LJ999hkLFy7k2muvTVR1JUmSJGnASliwLy0tZebMmQBMmTKFLVu2xO6rqKhgwoQJZGZmAnDCCSewceNGdu7cyVlnnQXA6NGjKS8vB2DLli3s3r2bt956i+LiYu666y4cDkeiqi5JkiRJA0rCgr3b7W4TkDVNIxwOYzAYKC4uZufOndTW1mK321m7di0jR47kuOOO45133uH8889n48aNVFVVEYlEmDx5MnPmzGHSpEk89dRTPPHEEyxatOiYxz/y5CIeysrKkCnC+i7ajlJfyDbsO9mGfSfbsG+E0LGqWZSWlvbL8RIW7B0OBx6PJ/a3rusYDNHDZWRkcOedd3LTTTdRWFjIxIkTycrK4pxzzqG8vJxvfvObTJ06lYkTJ6JpGrNmzcLlcgEwa9YslixZ0uXxJ02ahNlsjstrKS0tZdy4cchg3zdlZWWH21HqLdmGfSfbsO9kG/adrkeoKK9l2rRpcSkvEAgc8yI3YbPxp06dyurVqwHYsGFDmw9GOBxm48aNvPDCCzz88MPs2rWLqVOnsnnzZqZNm8aKFSs4//zzGT58OADXXXcdmzZtAmDt2rVMnDgxUdWWJEmSpAEnYVf2s2bNYs2aNcybNw8hBA8++CCrVq3C6/Uyd+5cjEYjs2fPxmw2s3DhQrKzo/tmL1++nN///vc4nU4eeOABABYvXsySJUswGo3k5uZ268pekiRJkqSohAV7VVW5//7729xWUlIS+/eNN97IjTfe2Ob+7OxsnnvuuXZlTZw4kRdffDEh9UyW6ua9AOS7irt4pCRJkiT1jUyXmwQRPcy2yrUA5DiGoqnybZAkSZISR2bQS4LdNRvxhVrwhVrYXbsx2dWRJEmSBjgZ7PuZN9jcJsDvrtmIN9icxBpJkiRJA50M9v1sW+VadBGJ/a2LSKxLX5IkSZISQQb7fhaOBJNdBUmSJGmQkTPD+kkg5GVndSkN3kPt7htfeFoSaiRJkiQNFjLYJ1hED7O3djO7ajcS0UPYzZk4LTkcaiqPPabZX4Pd7EpiLSVJkqSBTAb7BBFCUNlUzo6qj/CHPJg0C+OKTmVY9gSE0GnyViPQ8Ye8lB1aR76zWC7BkyRJkhJCRpcEaPAcYtuhD2n21aAoKqNyT2RU3hSMmin6AEVlQlG0677RW83u2g3srtnImIL45EiWJEmSpCPJYB9H3kAzZVUfUdW8G4DCjNGMLTgFm6l9F31r5rxsexEHG8vYXbuRoVnjsJqc/VpnSZIkaeCTwT4OQpEAu6o/ZW/9VoTQybDmM6FoBpm2gi6fa9BMjCs4lc0H3qXs0EecOOJL/VBjSZIkaTCRwb4bmiMHMTQ3tctjrwud/fWfU15dSigSwGJ0MK7wVApdo1GU7m+HW5Q5hn31n3GoeRfDPceTbS+K90uQJEmSBjG5zr4LYT3EwdAGtlWuJaKHgejku+rmvXyw4xW2VX6ALnTGFpzCmWPnUJRR0qNAD6AoSmwMf1vlBwihx/11SJIkDXTVzXtjm4xJbckr+y5s3v8uIeEhFILdtRvJd45k+6EPqfccBBSGZx9HSf40zAZrn46TactnSOY4DjaWUdGwneHZx8Wl/pIkSYOB3GDs2GRrHEOLv44tB/4X+7u8+lPKqz8BINcxnPGFp+KwZMfteOMKTqGqeTc7qj6mMGM0Rs0ct7IlSZIGstYNxiB6YTYmX65uOpLsxj+GdeWrYl33UQJNNTKt+EKmjfxKXAM9gNlooyTvJEKRADsPn1RIkiRJxyY3GOuaDPY9lGUrJNc5LGHlF+dMwmZysb9uK25/Q8KOI0mSNFDIDca6JoP9MUwvuaTNuI+qaBw35PSEHlNVNcYXzkAg2Fa5FiFEQo8nSZIkDXwy2B+D05LDpKFnx/4elXdihwly4i3POYIcxzDqPAeoadmX8ONJkiSls3EFp7b5W+GLLKVSlAz2XThh+DkYFTtWo5NRuSf2yzEVRWFC4QwUFLYd+hBdj3T9JEmSpEGq2V/b5m+z0dovF2bpRAb7LhhUI0OMU5hQdFq/LuVwWLIYkTMRX7CZPXWb++24kiRJ6UQIwd66LQBYjHZURcMf8uALupNcs9Qig303uLQh5LtG9PtxS/KnYtIs7Kr5FH/I0+/HlyRJSnWN3iqafbXku0ZyXNEZDM0aD9BmG3FJBvuUZtTMjCk4hYgeZkfVx8mujiRJUsppvaovzplEvquYMfnTUFColMG+DRnsU9ywrHE4LTkcbNxBo7c62dWRJElKGb6gm+rmPTgtOWTZCgEwGSzkOofR4q/DHWhMbgVTiAz2KU5RVI5rkzd/YCzFS6cc1ulUV0kaTPbVb0UgKM6Z1GZPksKMMQAcapRX961ksE8DWfYiCjNKaPLVcLBxR7Kr02etOayP3FwoVaVTXSVpMAnrISrqt2HSrBRmjG5zX76zGFXRqGzaOWAukPpKBvs0Ma7gVFRFo6zqI8KRYLKr0yetOax9oZY2KS5TUTrVVZIGk8rGHYT1IMOzj2u3UsqgGcl3FeMNNrdbljdYyWCfJqwmB6PyphAM+9hVsyHZ1em1dMphnU51laTBJLrcbiuKona6Q2hRRgkAlbIrH5DBPq2Myp2MxehgT91mPIGmZFenV9Iph3U61VWSBpM6dwWeQCNFGaMxG20dPibXMRyDauJQ0y7ZlY8M9mlFUw2ML5yOEDrbD61LdnV6LCS8NKX5igJPoJGwHkp2NSRpUGtdbjciZ1Knj1FVjQLXSAJhDw3eQ/1VtZQlg32aKXCNIstWRE3LXmpbKpJdnW4JRQJsq1xLRWg9wYi/3f3DsiYkoVZdG5N/crvbvMFm3i97mQMN2xFCT0KtJGlwcwcaqXVXkGkrIMOad8zHFmVGZ+VXNu7sj6qlNBns04yiKIc3eFDYdmgtegoHHCF09td/xntlL7O3bgsGzEwZfj6j805q87g9tZtSspvt6JUPo3KnMDrvJEKRAFsOrGZt+Urq3AeSVDtJGpz2HZFEpyvZ9iJMBitVzbsH/R4jCUv2rus6ixcvZvv27ZhMJpYuXUpxcXHs/pUrV/Lss8/idDq5/PLLmTNnDsFgkDvvvJP9+/fjcDi49957GTlyJHv37uWOO+5AURTGjh3Lfffdh6oO3vMUlzWH4dkT2F//OfvrPqM4t+sPfX+rcx9gW+WHuAP1aKqRsQWnEKy3UJAxilzn8NiZtsOcRY17H1XNu9stn0kmT6CRfXVbsBgcRJfvKpTkn4SmGhiePYEdVes52LiD9Xv+QZ6zmPGFp2I3Zya51pI0sIUiAQ427MBidJDvGtnl4xVFpTBjNPvqtlLnOUCes//TnqeKhEXMN998k2AwyEsvvcStt97KsmXLYvfV19ezfPlyVqxYwfPPP8+qVauoqKjg5Zdfxmaz8fLLL/OTn/yEJUuWAPDQQw9x880386c//QkhBG+99Vaiqp02xuSfjEE1sbO6lGDYl+zqxHgDzXy69z+s3/MP3IF6hmaOY+bYqxidNwVViX7cNNXAhKLTmFB0GuOLZqAoKtsPrUupdezbKj9EIJgwZAYTik5vsxGSxejghGHnMKPkMrJshdS07GXNjlf4vPIDguH2wxSSJMVHRf02IiLMiOzjY78nXZGz8qMSFuxLS0uZOXMmAFOmTGHLli2x+yoqKpgwYQKZmZmoqsoJJ5zAxo0b2blzJ2eddRYAo0ePprw8+uZs3bqVU0+N7ld81lln8cEHHySq2mnDZLAwpmAaYT3Ijqr1ya4O4UiQ7YfW8f7Ov1DdspdMWwEzSi5j0rCzO5wtm+8qJt9VjN2cQXHOJPwhN3tqNyWh5u3VtOyj1r2fbPsQ8p0jY3U9WoY1j1NGfZUpI87HYnKyr24r75W9xJ7aTYO+y1CS4k0XOvvqP0NTDAzL7v48nwxrPlajk+qWPSl1QdHfEtaN73a7cTgcsb81TSMcDmMwGCguLmbnzp3U1tZit9tZu3YtI0eO5LjjjuOdd97h/PPPZ+PGjVRVVRGJRBBCxFIh2u12Wlpaujz+kScX8VBWVgYoXT6uPwlhwKjYqGjYht5ix6w6un5S3OsgcOuHaIjsJkIIDTPZ2mjsoTyq9jdQRUObx0fb8egynGgYKa/+lECDEYNi7q/qd1AXnQOh6MmTNTCEHTu6l7GwQJxIs3aQxsheth9aR/mhjWQbRmNTctuk8YyHjtpQ6hnZhn3X323o0Wvwh9041SHsLu9Z+mpTJAuf3sLG7R/i0PITVMOeEULHqmZRWlraL8dLWLB3OBx4PF9sy6rrOgZD9HAZGRnceeed3HTTTRQWFjJx4kSysrI455xzKC8v55vf/CZTp05l4sSJaJrWZnze4/Hgcrm6PP6kSZMwm+MTNEpLSxk3bhypFuwBct1O1u/5Bz7TAfJzT0BB6fAqtC9a88IfXW6Dp5LPK9fSEqpDUwyMyZvGyNzJ7bJZtSorKzvcju05GjS2HlhN2FbL8cPPjWv9e2J37SZCh3yMyD6e44ac1PUT2phAMOynvOZT9tdtpTr8GVm2QsYXzWg3a7izNu3KsdqwL+UOJl21YTrrr/c/GW24btc2CMPkkjNw9HB+TIs/lw927kOxeRlXnBrvva5HqCivZdq0aXEpLxAIHPMiN2HBfurUqbzzzjtcdNFFbNiwoc0HIxwOs3HjRl544QXC4TALFy7klltuYfPmzUybNo277rqLzZs3s2/fPgCOP/541q1bx/Tp01m9ejUzZsxIVLXTTo5jKPnOYqpb9uI50IimGslxDO004PZUa2741mNpqgFfsIXth9ZR1bwbgCGZYxhbcCoWo73XxxmaOY79dZ9R2bSTETnHk2kriEv9eyIQ9lJe/QlGzUxJfu++gCaDheOKTmNE9nGUHfqI6pa9fFi+kqKMMYwtOAWrydFhm8ZDosqV0sNAfv+bfDU0eqvIdQzrcaAHcFqyD08G3k8oEsCoJa/3MFkS9mmYNWsWa9asYd68eQghePDBB1m1ahVer5e5c+diNBqZPXs2ZrOZhQsXkp2dDcDy5cv5/e9/j9Pp5IEHHgBg0aJF3HPPPTz22GOMHj2aCy64IFHVTkvjC2dQ3bI3uoY94md37UbG9DJYHa01NzxAefWnKArsqd2MLiJkWPOZUHQamba+d4u1Lin8aPcqtlWuZfroS+Pe/d2VHYc+JqKHGFd0BiaDpU9l2c2ZnFT8ZerdB9l26EMqm3ZS1bybkbmTEUKPtWmi3qt4liulh4H8/vdkuV1nijJK2FG9nqrmPQzLGh+vqqWNhAV7VVW5//7729xWUlIS+/eNN97IjTfe2Ob+7OxsnnvuuXZljRo1iueffz4h9RwQFFBQEETXqu+u2ciQzLHYTF0PdxxLu9zwtRsAMBvsjCs8haKMMXENyFn2QgozSjjUVM7Bxh0Mzeq/7rYmXw0HGstwmLN7NPmnK9mOIZxWcjkHG3ewo+pjdtV82ub+hL1XcSpXSg8D+f0PhLxUNu3Cbs4kxzGs1+UUZkaD/aHGchnspfS0rXJtLNBDNIf7h+Wv47LkACAO/y/6EBF7bGsim+jf4vD/Rf8tBPhCLW1ywwPYTC5OGzMbg2pMyGsZV3Aq1c172FH1MQWuURi0xBznSEIIth2MrvCYUHRat5f0dJeiKAzNGkdBxijW7ny1zWY6uojwwc6/YjF2b3JlMBikekf73ff8IXeHefynFstesMGgs30cBsL7v7/+c4TQKc6Z2KeLC5vJRYY1nzrPQQIhb6c59QcqGewHqFDET52ns+xuSnSq4eEvjnL4f4dvpPXWjpaP2c2ZCQv0cHh3v9wTKa/5hN21GxhbcErCjtWqsqmcRl81Ba5R5DiGJOw4BtWI3ZzZbuc8XdcJdXN9foQIoXD7rIm6nrqZFKXE6ygDZbpvhQ3ReQj76z/HoJooyhzb5/KKMkto8lVzqHlXn4YE0pEM9gPAhKLTqHMfiJ3Zq4rGaSWXR7vwjgjoQI/OjL3BZtbseKVNudFUvYk1Ku9EDjRsZ0/tZoZmjU9oV2Q4EqLs0DpURWN84fSEHadVR+/VGWOv7PZr7GwW9NHvlYLSL++VlBoc5ixq3fvb3NbgPcRnB9cwJn9an+egJMuhpl0EIz5G5k6Oy0VGoWs02yo/5FBj+aAL9oM35+wAYjO5GJV7YuzvUXkn4rBkoaoaqqKiKiqKovS4C6yjcvtjDFBTDYwrPBVdRCg79FFCj7W7dgOBsJeRuZOxmpwJPRYkrk2PLhdFQVPkufxgUNlYzp66TWjKF8GwMKMEmynj8N4U6ZnoKbpn/RYUFEZkT4xLmWajjWx7EY2+6nY9bAOdDPYDxKi8E7EanViNzrY/+ilablcKM0rItBVQ1bybevfBhBzDG2xmT+1mzAY7o/L677Ul+r0yahaE0NlR9XHcypZSU72nks0H3sWgGjll1MWxz9WkoWdxxtgrmVB4GigK2w+tY83OV6hq3pOSm051pMF7iBZ/HfmukVhN8UsY1po+91DTrriVmQ5ksB8gjsw3H8/1tYkqtytf7O4Hnx9am5DtZLcfWocuIowrPDWh8xCOluj3auLQmTjM2RxoLKPJVxO38qXU4vY38One/wAwZcQsMmx5bT5XqqJSnDuJmWOvYkTORHzBFjbs+y/r9/ydZl9dkmvftb1xWG7XkYKMUSiKOuhy5ctgP4B0lsM9VcvtSoY1j6GZ43D766lo2B7XsuvcB6hu3kOmrSB2pt+fEvleFbhGMqEomnhqW+XatLmSk7rPH/JQuvdfhPUgk4acRY5jKNDx5yqa6Ol0Th97JXnO4dR7Kllb/ipbDqwmEPImo/pd8gabqW7eg8uaG/cEW0bNTJ5jOO5APW5/fVzLTmUy2EspbWzBKWiqkR1V6wlFAnEpUxd6LNPYhKLT+j15T3/IcQwl3zWSRm8Vh5oG1xXMQBeOBPlk77/xh9yMzT+ZIVndm6XuMGcytfgrTBt5IQ5zFgcatvPejpfZVbMh5TaI2Vf3GRC9qk/E97Mw8/BOeIPouyGDvZTSzEYbo/OmEIr4Ka/+JC5lVtR/jjvQwNCs8e1y1g8k4wunoygqZYc+Srkfc6l3dKGzYf+btPjrGJY1gVF5U3pcRq5jGKeNmc3xQ85AVTR2VH3M+zv+QmVTeYe9QNXNe2M59/tDOBLkQMM2TAYrha7RCTlGvrMYTTVQ2djxax6IZLCXUt7InBOwGqNbyLoDjX0qKxj2s6O6FINq7Jc1/MlkM7kYmTMZf9jD7pr2iXik9CKEYOuB96hzHyDPOZzjhpzR66teVVEZnn08M8fNZWTuZAJhL5v2v81Hu1fR5P1inkdrvv1tlWv77YTxQOMOwnooume9qiXkGJpqIN9ZjC/UMmjmtchgL6U8VdUYXzQDgWB75Yd9KmtndSnhSICS/KmYDdY41TB1jc47EZPByu7ajfiC7mRXR+qD8upPONhYhsuay+ThX4pLpkejZmJ84XTOGHNlbNjnw10r2bT/Hfwhdyzfvi/U0iYdb6IIIdhXtwVV0RiWfVxCj1WYMbi68rv1aamoqODdd98lEomwf//+rp8gSXGW7ywm2z6EWvd+alp69xls8dezv/5z7KaMuK3bTXUGzcS4gsM5C6oSm7NASpyK+m2U13yC1ehkavEFcV89YjdncNKIWZwy8mKclhwqm3ayevtLbfZy2F2zMeFr02vd+/EGmynKKEn4yXiuYxhGzcyhpvKErPZJNV0G+3/84x9cf/31LF26lMbGRubNm8frr7/eH3WTpJgvluIpbK9ci97DL6cQgm2VHwCC8UWnJax7MBUNyRyLy5rHoaZyGjyHkl0dqYdqWvbz2cH3MWpmpo38CmZD4nK6RzduuoxJQ88ChXZ7brRObE2UvbXR5XYj+iG7napqFLhGEQz7qPdUJvx4ydZlsP/tb3/Ln//8ZxwOBzk5Obz22ms888wz/VE3SWrDaclmePZxeIJN7Kvb2qPnVjfvod5TSZ5zOHnO4QmqYWpSFIXjDucskEvx0kuzr5aN+99CUVROKr4Aey/2cu8pRVEZmjWeHHv7fSLitSKmI25/PXWeA2TZi3BZcxJ2nCMVDaKu/C6DvaqqOBxfZC/Kz89HVeVQv5QcY/KnYVBNlFd/QjDs69ZzInqY7Yc+RFFUxhfOSHANU1M0n8AYmv21HGgsS3Z1pG7wBVso3fsvInqIycPOJSvO6827ctzh2fpHavRWsWn/24RF/IP+3sMn8P2Zsz7LXojZYKOqaXfapRPuqS6j9tixY3n++ecJh8N8/vnn3HPPPUyYEL/9viWpJ0wGC2MKphHWg+ysLu3Wc/bUbsYXclOcM6lfroxS1bjCU9AUAzuqPh4QO6INZMGwn9K9/yIY9jGh8DQKMkb1ex2O3m9haOY4XJZcKpvKqQh9xM7qT+I2Qz8Y9nOwcQdWo5N854i4lNkdiqJSmDGasB6k1l3Rb8dNhi6D/b333ktVVRVms5m77roLh8PBfffd1x91k6QODc8+Hrs5k/3122jxHzvtZ3RG8QZMmpWSvJP6qYapyWJ0MCrvRIJhH+VHTLySUktED7Nh33/xBBopzjmB4tzk7c525D4Oxw05gxmHx/NVDJRXl/L+jpfjsla9omEbuogwImciShxWGfREUeYYYOB35XeZmHvJkiU89NBD3Hrrrf1RH0nqkqqoTCicQenef7Gtci0nj7y40/XGZYc+IiLCTCg8HYNm6ueapp6RuZOpaNjO3rotDMuagN2ckewqSUcQQrCl4n80eA9R4BrVL9suH0vrfgut/wYYmjWeluoIaoabPXWb2VTxNvvqtzKh6LReJanShc6+us/QVCNDs8bHtf7d4bLkYjO5qGneS1gP9es+Gf2py1OosrIyPB5Pf9RFkrot9/BEu3pPJdXNezp8TIPnEJVN5bisuQzNbL8H/GCkqQbGF56KEDplh9YluzrSUcoOreNQ8y4ybQWcMOyclEjl3FG+fVWJbkN95tg5X6zPL1/J5or/9TjfflXTbgJhD0Mzx2FMwgm5oigUZpQQEWFq+jFTYH/r8speVVXOPfdcRo0ahdlsjt3+xz/+MaEVk6SujC+cQW1LBdsPrSPXObzNDnLRpXbRZULHFZ2eEj+aqaLANZos22dUt+ylzn0gtomKlFx767awp24zdnMmJ434cr/uMtlbNpOLk0bMos59kG2H1nKwsYyq5l2Mzp1Cce4J3XoNrbvbjchJXu6LoowSdtV8SmVTeaxbf6Dp8p247bbb+qMektRjdnMmI3ImsbduM3vrtjD6iDzhBxrLaPbXUpQxJu67ZqW71pwFa8tfY1vlWk4bMzsu2dik3qtq3s22yrWYDFamFX8Fk8GS7Cr1SI5jCKeXXE5Fw3Z2Vq1nR/V6Khq2Ma5wOgWuUZ2ebDd6q2nyVZPnHJHUISWHJQunJZtadwXBsD/t2r87uvyGn3rqqfh8Pt555x3++9//0tzczKmnntofdZOkLpXkn4RJs7Cr5lP8IQ/VzXupbCxnx6GP0Q53NUrtuay5DM0ajzvQQEX958muTtz096YtvXVkPRu8VWza/w6aamBa8VewmpxJrl3vKIrK8OzjOHPcXEbmnIA/7GXj/rf4ePffafa1nUjb+voTtWd9bxRljEEInapOhgXjraZlHy2R/kty1a2kOr/+9a8pKipi2LBhPP300zz11FP9UTdJ6pJRMzOm4BQiepiyQ+vYVrmWzw6+TzDiY1TeFCxGe7KrmLLGFpyMphrZWV1KMOxPdnX6LBmbtvTGkfVs8dXx6d5/I4TOicPPx2XNTXb1+syomRhfNIMzxlxJnnMEDd5K1pa/ytYD7xEI+2Kv//PKDzjUVI7DnEV2Bwl8+ltrrvxDTTsTfqzW3B9V4S2E9VDCjwfd6MZ/4403+Mtf/oLFEu3WuOqqq5g9ezbXX399wisnSd0xLGsc++u3tlk6Y1BNjMw9IYm1Sn1mg42S/KmUHVpHefUnHDfk9GRXqU9aN20B2F27kTH505Jco44dWc+Pdv+NsB5k4pCZAy6zo92cwdTiC6h1V7C98kMqGrZxqKkclzU/9vohcXvW95TV5CDTVkC9pxJ/yJPQC4XoZyC6MdWWiv8xZcT5CTtWqy6v7IUQsUAPYDabMRhSf+KINHgoisrII5J/AEREmEC4Z7OCB6Pi7InYTC7213+G29+Q7Or0mjfY3GZXtv7YtKU3jq5nWA8yPPt4hmUP3ERluY5hnDZmNscVnQ4o1HsOtLk/lebUFMWu7ncl7BhHfwY2V7xLi78+Ycdr1WWwnzFjBjfddBNvv/02b7/9Nj/84Q+ZPj25az8l6WiHjkqIIYSe8E07BgJV1RhfGN0+OJ3z5n9+8AN08UW60/7YtKU3jq4nRNPiDnSqojIiZ2KH6/BTaTfGgozRKCgJTbCzaf87bT4DET3MuvI3Ena8Vl1eot999938+c9/ZuXKlQghmDFjBnPnzk14xSRJ6h95zhHk2IdS5zlATcu+dmuqU11tSwUN3vYTnbyBJiJ6OGWWsDX5amj0VrW7PRW6sPtLqu82aTZYyXYMoc59AE+gKa4rBLzBZrYfWkeTrzpuZfZEl1f2Xq8XIQSPP/44P/nJT6itrSUU6p8JBZLUXROKTmuzaYeqaLHMX9KxRZfizUBBYfuhD9NmQ5AWfz3r9/yT0r3/JKKHgLZB0xNs4v0df6Gyqe/pXPsiEPax5cBqPixfSVgPcmQ9B9vnNB2+p0UZ0XX2R/cW9lY4EmJH1ces2fEK1c17cFpy2qQE1lQD00u+FpdjHUuXwf7WW2+lujp6JmK329F1ndtvvz3hFZOknjh6045ReSdiM7mSWKP04rBkMzz7eLzBZvbW92z74P4WCHnZcmA1H+x8lTp3Bdn2IZxWMrvN3gejck9kZO5kAmEvm/a/zbpdb3R4VZ1Iuh5hT+0m3i97iQMN23GYszh55EVt6znIPqfp8D0tcI1EVbQ+nyQKITjYuJP3d7zMrpoNmDQLk4edy2kllzM6d0rscScMOwenJTsONT+2Lvu3Dh48yNNPPw2Aw+Hglltu4dJLL014xSSpp0blncjBxh3Rfx81YU/qWkn+VCqbdrKr+hOGZI7BbLAlu0pthPUQe2o3sadmExERxm7OZHzhDHIdw1AUBbs5I/b+l+RPRVMNDM86jrKqj6hq3s26XW9QmDEag+h5/vaeqmnZz/bKtXiCTRg0M8cVnc6w7ONQFZVMW8Gg/pym+vfUoJnIdQ6nunkPLf56XNacHpfR5Kth28EPaPRVoyoao/NOYlTeibG8+9E2KCMc0pk07Ox4v4QOdRnsFUVh+/btjB8f3aCgvLxczsaXUlJHm3ZI3WcyWBiTP43PKz9gR9V6Jg09K9lVAqKTLQ807mBn1XoCYS8mg5Xx+TMYmjW+Tea/jt5/m9nFlBHn0+A5xLZDH3KoaRcKuzEe8jMqb0rcc7F7Ao1sq/yQWvd+QGF49vGMyZ/WJiPbYP+cpsPrL8ooobp5z+Glgt0P9oGQlx1VH3OgsQwgtpnR0YmSontUzKC2srnfNt7psqUXLVrEtddeS0FBAYqiUF9fz6OPPtofdZOkHku3yWWpZlj2ceyv/5wDDdsZkX180pO8tO594A7Uf3GFlDu50x0MO3v/s+yFzBh9KZVN5XxW8QG7azdS0bCdMfnTGJY9oc/pgkORILuqP2Fv/VaE0Mm2D2FC0Wmdds8O9s9pqr/+POcINNVIZVM5YwtO6XISpa5H2Fu/lfLqT4joIRzmbCYUnUaOo/NkQXnOEQSqa+Nd9U51GewdDgcLFixg6tSpLF++nIqKCrzertcv67rO4sWL2b59OyaTiaVLl1Jc/MUbvHLlSp599lmcTieXX345c+bMIRQKcccdd3DgwAFUVWXJkiWUlJSwdetWvve97zFy5EgArr76ai666KLev2pJkjqkKioTik5j/Z5/sK1yLaeM+mpSZou3+OvZfmgdde4KAIZmjmNMwcl9SnSiKApDMsfQXBXClB1gV80GPq9cw776rW2GA3pCCMGBxjJ2HPqYYMSH1ehgfOEM8l0jB9Us+4FGUw0UuEZysHEHjb5qsjrJBSCEoNa9n22Va/EGmzFqZsYVnRGXE8h46zLYL126lB/84AccPHgQh8PBypUrufHGGznrrGN38b355psEg0FeeuklNmzYwLJly2Jpduvr61m+fDmvvfYaLpeLBQsWcNppp7Ft2zbC4TAvvvgia9as4Ze//CW/+tWv+Oyzz1i4cCHXXnttfF61JEmdynEMJc9ZTE3LXqqad6Eq0Z+JeF6NteaFP7rMQMjLjur1HGgoAwTZ9iGML5zRq3HTzkR7CKYwNGscO6tKqWjYzid7/0WOYyjjC2e0uxrvrK4N3iq2HfyAZn8tmmJgTP7JjOzmTm9S6ivMKOFg4w52VX/K8Ozj2r3/7kAj2yvXUuuuQEFhRPZESvKnpuwmOl1+KnVd58wzz+TWW2/ly1/+MkVFRUQiXS/NKS0tZebMmQBMmTKFLVu2xO6rqKhgwoQJZGZmAnDCCSewceNGJkyYQCQSQdd13G53bG7Ali1b2L17N2+99RbFxcXcddddOByO3rxeSZK6YXzh9MNXLOtQFQVQyHEMjUsga82NDsTK7GryXSKYDTYmDp3JiJyJbK/8kDr3AT7Y+SrDssYzJn8aZqOtw7r6Q27KDn0US7xSlDGGcYWnyn0YBpgcx1AMqpladwWeQGPs/Q9FApRXf8K+uq2IwyekxxqySRVdfnOtViu///3vWbduHffeey9//OMfsdu7/lC73e42AVnTNMLhMAaDgeLiYnbu3EltbS12u521a9cycuRIbDYbBw4c4MILL6ShoSG2CmDy5MnMmTOHSZMm8dRTT/HEE0+waNGiYx7/yJOLeCgrK+PodbxSz0XbUeqL/mpDlzKUpvD+2N/rt79Nljayz+U2hPfg06NZ49ZvexuDYqYhsocIQTSM5GjjcOqFNFT6aGBHn4/XkaPb0ClKMBiyqI/soqJhGwcadpCpDUcXkVhdP972Foqi0hTZh0DHpDjI0cZg8WWwb/eBjg4zoA2G77KqG4EAvlAL67e9jaaYaYjsRieEAQvZhhJswRwq99VSSc/G34XQsapZlJaWJqbyR+ky2P/sZz/jL3/5C48//jgZGRlUVVXx85//vMuCHQ4HHo8n9reu67Er9YyMDO68805uuukmCgsLmThxIllZWTz33HOxXoTKykq+9a1vsWrVKmbNmoXLFV2LOWvWLJYsWdLl8SdNmoTZbO7ycd1RWlrKuHHjkMG+b8rKyg63o9Rb/dmGLb4cPij/Itg36xWcUDKj1+uihRB4gk3s3fl+7LZGPdpFrioao3OPPfkuXo7VhrqYzoGGbeysKqUhsqfNfU36PgBMmpWxhacwNHPcoB2XHwzfZW+wmT07vHB4qX3rZ1VTDYzNO4XinEl96unS9QgV5bVMmxafDZsCgcAxL3K7rGlBQQE33nhj7O/bbrutWweeOnUq77zzDhdddBEbNmxo88EIh8Ns3LiRF154gXA4zMKFC7nlllvYvn07RmN0GUJGRgbhcJhIJMJ1113HPffcw+TJk1m7di0TJ07sVh0kSeq9HdXr2/ytiwgf7HwVi9GOEAKBAKEjELG/hdCPuO+I2+g8OYnFYGd6yaUp0Q2uKirDs4+nMGMMa3e+hi/UdjMdq9HJaWNmx33JnpR6ontF6G1uMxtszCi5LCU+qz2VsJkks2bNYs2aNcybNw8hBA8++CCrVq3C6/Uyd+5cjEYjs2fPxmw2s3DhQrKzs1mwYAF33XUX11xzDaFQiFtuuQWbzcbixYtZsmQJRqOR3Nzcbl3ZS5IUf7oeJhj2oaCiKEr0/1FQVC36X0X54j6OuF9RQVFw++sJRQJtynRac1Lux9OomXBYMtsFe4clSwb6QcxlzU25z2p3JSzYq6rK/fff3+a2kpKS2L9vvPHGNj0GEE3Hu3z58nZlTZw4kRdffDExFZUkqUMTik6jzn0gtkOXqmicMfbKPqU39QabWbPjlTZlplpu9FYdvf5UrasUfwPt/U+thYCSJKWMROQxT4fc6K3Sqa5S/A20918Ge0mSOjUq70SsRidWozNuecwTUWaipFNdpfgbSO+/zP6QRLqItNnuUZJSTSLymKdDbvRW6VRXKf4G0vuf3rVPQ0IIUARWoxOzwUajt6rN3saSlGoSkcc81XOjHymd6irF30B5/2Ww7ye60NEUDbslE4clK5Y32RNsIhwJJrl2kiRJ0kAmg32C6SKCUTPjMmdiM7naJeHIsOZT07Iv5TZNkCRJkgYOGewTRBcRzAYbTks2ZqOt08eZDGasJgeBUNc7CUqSJElSb8hgH0dCCEBgMTpwWrIxGrqXrjfDkkdVcLccu5ckSZISQgb7OBBCB0XFZnbhsub0eIa9phmwmzPxBJoGba5tSZIkKXFksO8DIXQ01YDdnI3dktGnK3OXNQdvsLnrB0qSJElSD8lg3wu60DFpFuyWTGwmZ1zKVBQVpyWHZn8titxdT5IkSYojGey7SQBCRLAaHdgtWZgN1rgfw2HJxBNojOViliRJkqR4kMG+m+ymDOzmTAyaMaHHcVlzafBUysl6kiRJUtzIiNJNGba8hAd6AKvJgdFgSfhxJEmSBhohBGFdJinriAz2KSjDmhed4S9JkiT1gI7dnJHsSqQkGexTkMlgwWJ0JLsakiRJacVksGHSrIdznkhHksE+Rcmre0mSpO4TQsdidGAx2tGRk5yPJoN9itI0AzZzhjxDlSRJ6iab2YVBM8mtwzsgg30Ky7Dmyox6kiRJ3WA22GIbihk1U5Jrk3pksE9h0UQ72eiyO1+SJKlTutCxHJHgzKB2b1+SwUQG+xRnN2diUBO/5C8ZZPIgSZLiQVUUbKYvJjUbNbMcAj2KDPYpTlEUXNbcARcYBQJDGn0hdRFJm7pK0mBjNtjbJCKzGG1ygvNRZLBPA1aTA9MAS7RjUI3kOYchSP0vpBA6mbYCzEaL/AGRpBQjEFiP2qPEoJlQVBnejiRbI01kWPMGzNi9QCfbXoiqaBi0dBhbi/6YZNuHYjE5EMgrfElKHQoWo73drXKSXlsy2KcJk8Ha4Qc63ehCx2HOjgV5UxoEe4NmRlVUFEUh216EzeQacMMqkpSuLAZbh6uWjHKSXhsy2KeRgZBox6iZcFqyY3+bDKk/tmbS2g6hZNrycVlyU77ekjTQCaFjM7k6vE9O0mtLBvs0YtCMaZ5oR5BpK2xzFm412VO6W1wXOmajrd3tTms2Llu+DPiSlESqomExddzjaTbaEbIHLkYG+zTjsuZCGibaEULHqNgwGdp2raXDuH1nwycOcwaZ9sK0mGQoSQNRRyfirQyaEUWVu7i3ksE+zaiKisucfol2DKoJk9Lx5j5Hd5OnEqPB1GZJz9FsJifZ9iFp3NsiSelJFxFs5sxjPkZO0vuCDPZpyG7JxJBGZ6xC6GTa8ztN/WsyWlO2O9ykWrt8jMVoJ9c5LIUHIyRp4NEUI+YuliTLSXpfkME+DSmKgjNNEu0IIbCZMzAZOg+aVqMjJa+MOxuv74jJYCHfORyF9BtikaR0ZDF1/d00aqaU/G1JBhns05TN5MSYwt3frTTVQIY175iPURUVgyH1zsAVopm4usugmchzjkBVNPkDI0kJpOs6dlNml4+zmBxpN+SZKAkL9rquc++99zJ37lzmz5/P3r1729y/cuVKLrnkEq655hr+8pe/ABAKhbj11luZN28e11xzDeXl5QDs3buXq6++mmuuuYb77rsPXZdvHkCGLbUT7QghyLQVdGvnvlQctzcazMccr++IphnIc43AaDDKgC9JCWLQjBi7cYGgqQZUVW53CwkM9m+++SbBYJCXXnqJW2+9lWXLlsXuq6+vZ/ny5axYsYLnn3+eVatWUVFRwf/+9z/C4TAvvvgiN9xwA7/85S8BeOihh7j55pv505/+hBCCt956K1HVTivmFE60E+2+d2I2dj3mDdFZtal24tLbExBVUcl1jMCcwnMRJCmdWY0dT/btiJykF5WwYF9aWsrMmTMBmDJlClu2bIndV1FRwYQJE8jMzERVVU444QQ2btzIqFGjiEQi6LqO2+3GYIhOQtu6dSunnnoqAGeddRYffPBBoqqddjKseSm59EtVtC67749kMdpRUmiKm65HMBt6fyIVzbY3BKvJIQO+JMVRRESwmTO6/Xg5SS8qYVO63W43DscXZ1+aphEOhzEYDBQXF7Nz505qa2ux2+2sXbuWkSNHYrPZOHDgABdeeCENDQ08/fTTQPQqsbUr2G6309LS0uXxjzy5iIfS0tK4lhdPfr2ZkO7rVnd5fxBCYFGzOKDWt7vvWO3ojdShkxqTDgWCSrU5Lm0afX+8PR4S6ExZWVlcyhnMZBv2XbLaUEGjSus6BrQK6T78eny+y/EkhI5Vzeq32JKwYO9wOPB4PLG/dV2PXalnZGRw5513ctNNN1FYWMjEiRPJysriueee48wzz+TWW2+lsrKSb33rW6xatQr1iN2LPB4PLlfH6RGPNGnSJMzm+JzRlZaWMm3atLiUlQi60DnUtCslZoILIbCaHGTZC9vd11U7Nnqq8IXciaxetxk0I3nOEXErr8VXT4u/rs8Bv6ysjHHjxsWpVoOTbMO+S1YbCiFwWLJwWXO6/ZyIHqaycRdaio3d63qEivLauMWWQCBwzIvchHXjT506ldWrVwOwYcOGNh+McDjMxo0beeGFF3j44YfZtWsXU6dOxeVy4XRGtyrMyMggHA4TiUQ4/vjjWbduHQCrV6/m5JNPTlS105KqqDjN2SnRXawqKpm2/F49N5XG7U1a9+YadJfTmk2GTK8rSX0i0HF0kUjnaJpqSLlAnwwJu7KfNWsWa9asYd68eQghePDBB1m1ahVer5e5c+diNBqZPXs2ZrOZhQsXkp2dzYIFC7jrrru45pprCIVC3HLLLdhsNhYtWsQ999zDY489xujRo7ngggsSVe205bBk4Qk0JXX8XgidLMeQXl+9mo12SIFxeyH0hEx8tJszUBSVRu8hFLnqVZJ6zKRZezW73qiZCEWCCahR+khYsFdVlfvvv7/NbSUlJbF/33jjjdx4441t7rfb7SxfvrxdWaNGjeL5559PTEUHCEVRcFlzaPRWxW1suCeEEFhMjj4FSVVRMRrMRCLhONasFxSOmQSoL2wmJ5qiUec5mBLDLpKULoQQPZqFfySjZh70wV5eXgwgNrMLQ5KWmSiKQqatoM/ldCc9baKZNEtCJ/OYjTZyHcNToA9DktKJjs3S9XytjhgMlkE/hCaD/QATHS8X/ZrQRQidDFs+ahx6FMxGa9LH7eM9Xt/hMQxm8p0jMBmS/3olKR2YDDZUpXdj71aDHT0Flyj3JxnsBxiTwUphxmjs5ox+OZMVQmAx2rGZnHEpL9lJgnQRwdzJ/tjxZtCM5DiGkOsYiqYaEPJaX5I6FJ1H07sufABV1dBUYxxrlH5ksB+AFEUlw5ZHgWsUZoMtsUFfUci09737/oviVIyG5GW8UhS131P3mo028l3FuA4nIZJpdiWpPbu5d134rYzq4M6kJ4P9AKZpBrIdReQ6R2DQTHEP+kLoZFrze9211plkjtsnerz+WBzmDAozRmEzu1IyK6IkJYvZYO/zxGOjNrgz6clgPwiYDGbynMPJshdFd2SLU3ex2WjHZo5P933bcpO33t7Uxf7YiaYczlOQ7yzGqFkQabCNsTTwRUTyVsjoQsdi7n0XfiuTwTKo58fIYD+IWE2OaHexJZp9qm/dxaLXyXO6YjHakrIoTReRPo0LxpNBM5HrHEq2YyiaIsfzpeRRUMhzDCdZOTBURcEWh++l2WgjFfJ4JIsM9oOMoig4LFkUZozq9SQ+IXQyrHloamLSNCiH19v3N0XRkn5lfzSL0U6eawQuSw6in1dZSINbdE8SlTznCMxGGxm2gqQsX4tHFz5EN+cazJP0ZLAfpPoyic9ksPVo16neSMb+9qYUHdP74gRtNEZ1cHdFSv1DCIHRYCTfVYymRU/qbSZndD5JPy/rtcZppQ+AQQZ7abDq6SQ+gSArDslzumI22NH1/h2vTlTWvHhRFRWLmkGBawRGzTzok4RIiaELHbPRSq5jeLvcGRnW/Fjw7xeKGtfluIN5kp4M9hLQvUl8Qui4rHn98mU3G639Ois+IsJYk7zGv7sMmplc57Av3isZ9KU4EULHZnKQbe94jwtFUcixF/XbHBKL0RbX34HBPEmvH0/RpHRgPZzf3hNopMVff3jcLvplMxosOBLcfd+qddw+HAn1y/E0xYAxxcbru9L6Xrn9DbgD9QhByu3ZLaUPIXQc5ixcttxjPs6gmbEoToTQE7oPhy4i2E3x/b0ZzJP05JW91M6Rk/gc5kyE0BFCkGVrv0d9IvVH2tovjpVegb6Voig4rdkUZIzCYnLIq3ypV3R0XNbcLgN9K6Nqw2KyJ3T8XlMMh4Nz/AzmSXoy2EudUhQVly2XAtcosh1FGLT+/ZKYDbZ+GbePTkZKz2DfSlU0su2F5DpHJLsqUprRhSDLWoDDktWj52XZChO6T3y8A32rwTpJTwZ7qUuaZkhKznpznMfrOhMdp4x/cqBkMBnMOA8v05OkrgghyHEUYetFKlpFUcmyFyYk26MuItjMmXEvFwbvJD0Z7KWUpShKv6y3VzVD0rYGTgSHJXPQXr1IPZPrHNqnE3mTwYrLkhv34SNNMWJOUG/bYN1pUgZ7KaX1y3azanp34Xck05Yvx++lTiko5DmHx2W5qcOShdlojev4vcWUmC58ODw8MAiTU8lgL6U0i9Ge0KAlhMCc4uvre8NksGI1OWXGPakNIQSqqpF3OK9GvGTZh8RtyE3XdeymzLiU1RFVUdH6ef5RKpDBXkppJoOVRCbKF0LHYkqNfPjxlmHLB7kUTzosOhHVRJ5zRNxzZaiKSra9KC4n5gbNmPDhu8G43a0M9lJKUxQFYwKXxWmasd9XGfQXVVFxWeM/niqlH4HAbLR1mBUvXsxGGw5LVp8nh1r7YTOqwThJTwZ7KeUlcg18qubDjxeHOWNATT6Ueq51tUmOI35d7Z1xWXMx9uHzFhGRhO+7AYNzkp4M9lLKsxjt6AnY110IkfL58OMhM0m7lUnJJ4TAYckmsx/2s2iVbRtCb7PUmTRzv/S0mYzWQTdJTwZ7KeWZDNaEdD3qxHdHrVRlMljkZL1BSAiBy5aLy5rTr8fVNEOvtsMVQmDphy58aJ2kN7h6vGSwl1JedNw+/t3tBtWIpg6O7SHkZL3BRaCTaSvAkaDENF2Jboeb0aOAL9D7tb6DbZKeDPZSWkjEevt0zYffG3Ky3uAhEGTbh2AzJ7fXKsOa16P5IibNiprA9LtHMxpksJeklGM2xXfcfrCM1x8pOllvYE9IHOyE0MmyFSYlvfXRFEU5vByv6+EjIUS/zMI/kslg7Ze9N1KFDPZSWjBplrhupykYOPnweyLTVjDoZiEPJppqxJpCeSMMmolMW143epR0bJae5+fvC3OCc3ikGhnspbSgKEpcl8kZNFO/dhmmCpPBjM3kkpP1BiAhdOz9sGytp2zmjMPbL3f+mTMZbKhK/34fFUUdVMtSZbCX0kY8u90HYj787sqw5fXLboJSP1PUlAz2AFm2gk4nwwqh99ss/KMZFBnsJSnlmI2OuIzbR8frB2+wl5P1Bh4hBDaTM65DXfEU3Q63qNPtcO292GI3HgbTJL3U/GRIUgdMmjkuP2YCHWuSZyonm11O1htgdJzm7GRX4phMBnOH2+GaDfaknaSYDbZBM0lPBnspbUTH7ft+RW7QzP0+PpiK5GS9gcNidMZ9c5tEiG6Ha4uN3+tCx2JO3oRCk8EyaCbpJezToes6ixcvZvv27ZhMJpYuXUpxcXHs/pUrV/Lss8/idDq5/PLLmTNnDq+++iqvvfYaAIFAgM8//5w1a9awf/9+vve97zFy5EgArr76ai666KJEVV1KYSaDhVAk0Lcy5BUt8MVkPV+wRY7hpzFd6DgsWcmuRrdl2YuobtqDQKAoCjZj8nrZWifpDYar+4QF+zfffJNgMMhLL73Ehg0bWLZsGU899RQA9fX1LF++nNdeew2Xy8WCBQs47bTTmD17NrNnzwbgpz/9KVdccQUul4vPPvuMhQsXcu211yaqulKasBgdtPjre31lLoSOyWCLc63SV4YtD3/InexqSH1g0ixpNQdFPTx+X+euwGJ0JP1E06iaCOi+pNahPySsG7+0tJSZM2cCMGXKFLZs2RK7r6KiggkTJpCZmYmqqpxwwgls3Lgxdv/mzZvZuXMnc+fOBWDLli28++67fP3rX+euu+7C7ZY/ToOVyWBB6UMXvBACqyn5CUdShaqoZMjJemlLpNlVfSuz0YrDkp0Se1MMlrkrCbuyd7vdOBxfjMVomkY4HMZgMFBcXMzOnTupra3Fbrezdu3aWBc9wG9+8xtuuOGG2N+TJ09mzpw5TJo0iaeeeoonnniCRYsWHfP4R55cxENpaWlcyxus4tGO3kgDOqFePVdB5ZDW0uc6JFMiPoveSB06A78rs1VZWVmyqxAnKg6tOSlHHii/iWE9iE9v6PceBiF0rGpWv7VjwoK9w+HA4/HE/tZ1HYMheriMjAzuvPNObrrpJgoLC5k4cSJZWdGz0+bmZnbt2sWMGTNiz501axYulyv27yVLlnR5/EmTJmE2x+eMrbS0lGnTpsWlrMEsXu3Y7KvDE2js1XMtRgdZ9v7b7jPeEvVZDIYD1LTsS8jugqmmrKyMcePGJbsafSaEwGnJxmnt/1n4A+k3UQidg43l/f7Z1/UIFeW1cWvHQCBwzIvchL26qVOnsnr1agA2bNjQ5ssVDofZuHEjL7zwAg8//DC7du1i6tSpAHz88cecfvrpbcq67rrr2LRpEwBr165l4sSJiaq2lAasvdzfXhd6dB9rqR2TwYzdLDPrpRNFAYclM9nVSHuDJZNewq7sZ82axZo1a5g3bx5CCB588EFWrVqF1+tl7ty5GI1GZs+ejdlsZuHChWRnR89Od+/ezbBhw9qUtXjxYpYsWYLRaCQ3N7dbV/bSwGXs5bi9Qv9vtpFOMqx5+IJyPky6sKZwEp10E52kF052NRIqYcFeVVXuv//+NreVlJTE/n3jjTdy4403tnvet7/97Xa3TZw4kRdffDH+lZTSlkkz93gJXnR9vfxx7IyiqGRY82j0VskgkuJ0oeNI8SQ66cSgmQiEvcmuRkLJb7SUlnqTJ38w7V/fWzazC2MaLeMarKxGOwbNmOxqDBhmoy2uW2inIhnspbRkNTqIiO53u+lCx2yU6+u7I9OWLzPrpTAhItjTcLldKjNpFpQBnkpPBnspLRkNZjSlZ6NQFqNcX98dRs2M3ZwhJ+ulKKNmie7FLsWNoihoA3ySngz2UtrqSbe80WCS49A9kGHNTXpmM6k9gcBmzkx2NQYkoyqDvSSlJKPB0u2rT7Mmr4R6IjpZL19m1ksxKiq2FMg6NxAN9OV3MthLactqdHYrGOlCx2yQXfg9ZTM75WS9FCKEwGbOkD0uCWLpZf6OdCGDvZS2jAYTaje29VSI5uKWei5LTtZLISIt8+CnC6NmHtCT9GSwl9KaSe36ytNoMMvx+l4yyMl6KcNqcso8EQk00CfpyU+OlNaMBnOXgcgkx+v7JDpZT5UBP4l0oeO05CS7GgPeQJ6kJ4O9lNZspmOP2wuhY5b71/eJoqgUZIzEZnYh0GXQTwKLwSaT6PQDg2YasJ9vGeyltGbQuh63l8l0+k5VVDJt+RRmlGAzuwAxYH8UU40uInKsvp9YjXYEA3OOigz2UtozH2O9fXTTnIE76aa/tQb9gozR2M0ZyKCfeEbNLE9Y+4lhAE/Sk8FeSnsmzdppwJH58BNDVVQybHky6CeYLnTspoxkV2PQUBRlwK63l8FeSnsWk6PDcXshdJkiN8GODPoOSyZCBv24UlUNm1kG+/5kVM3JrkJCyGAvpT2DZkTraPKS0rvd8aSeUxUVlzWXIhn040YIgd3oksNQ/cyodb3CJx3JYC8NCCat/dm4SZPj9f1NaRP0sxAwIH84+4fAYZUT8/qb2WgdkJP0ZLCXBoSOxu3l+vrkiQb9HIoyRsmg30sWkwNV0ZJdjUHHoJkH5EVCz/YIlaQUZTE5aPTVoB3+cdRFBLNJjtcnW2vQd1qyaPE34Ak0IoTo9Y+pEOJw/nIFVVFQFRVVNaCqGioaqqKhKAoREUYXESJ6hIgeQtd1UARqmlzf6CKCy5Kd7GoMSq2T9CKRcLKrElcy2EsDgkEzYtCMsYl6iqLKmfgppOOgr8fSGMeCuBLd2a1NEFc0NEVDVaJ/GzUTmmqMBfbuEEInHAkT1gMYlUpsJmf0REBETwR0EY7tAaCgJv3KzmywYehgaErqH0bVLIO9JKUqk2YhEPbG/p3sH2ypvSODvtvfiC7CRwRxMwbViKrGv+taUVSMBhNGTJhVOxm2/HaP0YVOOBIiFPET0cPoepiIiJ4IhCJB6EOPRE/oIoJd7lmfVEbNjC/oHlC/ITLYSwOGyWDBH/KgKAomuTVrSlMUFac1tbqpVUXFZDBjMrS/ohZCp8lXizfQlPBNlQyqCavJkdBjSMdmNtjRqUYbQCEyPQawJKkbrCYnOjq6iGAxyh9LKX6Uw5kDc53DUBXtmPsx9IUQ+uEkRVIyGQ2mATc5UgZ7acDQVAMG1YiiaPLKXkoIk8FKvqsYhyU7McuzFFV24acI4wDLpCeDvTSgmDRLh2vuJSleFEXBZc0h3zkCTTUiiM+SQiEENpNzQI0TpzPDAMukJ4O9NKCYDFaZNU/qFwbNTL5rBC5LTpxyCOi45J71KWOgZdKTwV4aUGxmJzaTK9nVkAYRhyWLfFcxRs3Sp7F8i9GZkJUIUu9YjLbDOR3iTxc6DksWmtJ/QwUy2EsDiqpoGDrKky9JCWTQjOQ6h5Jpy+9VtkBd1+We9SnGoJnifvIlEGiqgTznCDJsef06ZDNw1hVIkiQlmc2cEc3m6Kk6vAy0e9dTZqNFTipNQUbNRDgSiktZQug4LTk4LFlJmZchg70kSVIcqYpGtmMI/pCHRk8VutCP+eMeXW4nr+pTkUE19znYC6FjNFjIshVgSOIMf9mNL0mSlAAWo52CjJFYTc5jdutrqkEm0UlRfZ+kJ3DZ8slzDk9qoAcZ7CVJkhJGUVSy7AXkOIZ2mIwnutxOJtFJVRajvVeTLoXQozkZnCNxpEiSpIR14+u6zuLFi9m+fTsmk4mlS5dSXFwcu3/lypU8++yzOJ1OLr/8cubMmcOrr77Ka6+9BkAgEODzzz9nzZo1NDQ0cMcdd6AoCmPHjuW+++5DVeV5iiRJ6cFsjCbjafHV0RJoQD08lq8o4LBkJrdyUqcMmhGlB7GmdUfHTHshNpMzgTXruYRFzDfffJNgMMhLL73ErbfeyrJly2L31dfXs3z5clasWMHzzz/PqlWrqKioYPbs2axYsYIVK1YwceJEfvKTn+ByuXjooYe4+eab+dOf/oQQgrfeeitR1ZYkSUoIRVFw2XLJc45AUw0IBFaTM+G59qW+6W4mPV3oWEwOCjJGpVyghwQG+9LSUmbOnAnAlClT2LJlS+y+iooKJkyYQGZmJqqqcsIJJ7Bx48bY/Zs3b2bnzp3MnTsXgK1bt3LqqacCcNZZZ/HBBx8kqtqSJEkJZTKYyXcV4zRn4zCn1mZAUnvGLjLpCSFQUMlxDCHbXhjrtUk1CauV2+3G4fhi0ommaYTD0f2Bi4uL2blzJ7W1tfh8PtauXYvX64099je/+Q033HBD7G9xxNaSdrudlpaWRFVbkiSpXzit2TInRBo41iQ9IXRsZhcFGcVYjPZ+rlnPJGzM3uFw4PF4Yn/ruo7BED1cRkYGd955JzfddBOFhYVMnDiRrKzo0pPm5mZ27drFjBkzYs89cnze4/HgcnWdIe3InoR4KC0tjWt5g5Vsx76Tbdh3sg37brC0oS4ieCLVKEfsgieEQFU0LIoLTW0G9ve6/P5qx4QF+6lTp/LOO+9w0UUXsWHDBsaNGxe7LxwOs3HjRl544QXC4TALFy7klltuAeDjjz/m9NNPb1PW8ccfz7p165g+fTqrV69ucyLQmUmTJmE2x2cjg9LSUqZNmxaXsgYz2Y59J9uw72Qb9t1ga8ODjeW0ZkoQ6DjMWTgtOX1OjhPPdgwEAse8yE1YsJ81axZr1qxh3rx5CCF48MEHWbVqFV6vl7lz52I0Gpk9ezZms5mFCxeSnR0du9q9ezfDhg1rU9aiRYu45557eOyxxxg9ejQXXHBBoqotSZIkSW0YNROhSACDaiLbXoAhDXfWTFiwV1WV+++/v81tJSUlsX/feOON3Hjjje2e9+1vf7vdbaNGjeL555+PfyUlSZIkqQsWgx2r0ZHW+xfIdLmSJEmSdAxOa/qvmkjNNQKSJEmSJMWNDPaSJEmSNMDJYC9JkiRJA5wM9pIkSZI0wMlgL0mSJEkDnAz2kiRJkjTAyWAvSZIkSQOcDPaSJEmSNMDJYC9JkiRJA5wM9pIkSZI0wMlgL0mSJEkD3IDLjS+EACAYDMa13EAgENfyBivZjn0n27DvZBv2nWzD+IhXO7bGvNYYeDRFdHZPmmppaaGsrCzZ1ZAkSZKkfjdu3DicTme72wdcsNd1HY/Hg9FoRFGUZFdHkiRJkhJOCEEoFMJut6Oq7UfoB1ywlyRJkiSpLTlBT5IkSZIGOBnsJUmSJGmAk8FekiRJkgY4GewlSZIkaYAb9MFe13Xuvfde5s6dy/z589m7d2+b+99++22uuOIK5s6dy8svv9yt5ww2vWnDUCjEbbfdxjXXXMOVV17JW2+9lYyqp4zetGGruro6zj77bMrLy/uzyimpt+34m9/8hrlz5zJ79mz+8pe/9He1U0pvv8+33nor8+bN45prrhn0n8XuxAifz8e8efNibZXwuCIGuX//+99i0aJFQgghPv30U/G9730vdl8wGBTnn3++aGxsFIFAQMyePVtUV1cf8zmDUW/a8JVXXhFLly4VQghRX18vzj777GRUPWX0pg1b7/v+978vvvzlL4udO3cmpe6ppDft+OGHH4r/+7//E5FIRLjdbvH4448nq/opoTdt+N///lf84Ac/EEII8f7774sbb7wxKXVPFV3FiE2bNonLL79cnH766bHvbaLjyqC/si8tLWXmzJkATJkyhS1btsTuKy8vZ8SIEWRkZGAymZg2bRrr168/5nMGo9604Ve+8hV++MMfxh6naVq/1zuV9KYNAR5++GHmzZtHfn5+UuqdanrTju+//z7jxo3jhhtu4Hvf+x7nnHNOkmqfGnrThqNGjSISiaDrOm63G4NhwCVn7ZGuYkQwGOSJJ55g9OjR3X5OXw3udwRwu904HI7Y35qmEQ6HMRgMuN3uNpmI7HY7brf7mM8ZjHrThna7PfbcH/zgB9x88839Xe2U0ps2fPXVV8nOzmbmzJk888wzyah2yulNOzY0NHDw4EGefvppKioquP766/nXv/41aJNy9aYNbTYbBw4c4MILL6ShoYGnn346GVVPGV3FiGnTpvX4OX016K/sHQ4HHo8n9reu67HGPfo+j8eD0+k85nMGo960IUBlZSXf/OY3ufTSS7nkkkv6t9Ippjdt+Ne//pUPPviA+fPn8/nnn7No0SJqamr6ve6ppDftmJmZyZlnnonJZGL06NGYzWbq6+v7ve6pojdt+Nxzz3HmmWfy73//m9dff5077rhjUOfO702MSHRcGfTBfurUqaxevRqADRs2MG7cuNh9JSUl7N27l8bGRoLBIOvXr+ekk0465nMGo960YW1tLddeey233XYbV155ZbKqnjJ604YvvPACzz//PCtWrOC4447j4YcfJi8vL1kvISX0ph2nTZvGe++9hxCCqqoqfD4fmZmZSXoFydebNnS5XLGT+IyMDMLhMJFIJCn1TwW9iRGJjiuDPl2urussXryYsrIyhBA8+OCDfPbZZ3i9XubOncvbb7/NE088gRCCK664gq9//esdPqekpCTZLyVpetOGS5cu5Z///GebMavf/va3WCyWJL6S5OlNGx5p/vz5LF68eFB/DqH37fjII4+wbt06hBDccsstsbHTwag3bejxeLjrrruoqakhFArxzW9+c1D31nXVhq2O/N4mOq4M+mAvSZIkSQPdoO/GlyRJkqSBTgZ7SZIkSRrgZLCXJEmSpAFOBntJkiRJGuBksJckSZKkAU4Ge0mSeqSiooLzzjvvmI/51a9+xa9+9at+qpEkSV2RwV6SJEmSBrjBm+NVkqQuhcNhFi9ezI4dO6itrWX8+PHceuutsfvvuOMOzGYzmzdvxuPxcP3113PZZZcBsGnTJubNm0dVVRWzZ8/mpptuwu12c9ddd1FVVUV1dTWnnXYaDzzwwKDNQy9J/UUGe0mSOvXpp59iNBp56aWX0HWdb33rW/zvf/9r85j9+/fz0ksvUVdXx+zZsznjjDMAqKur48UXX8TtdnPeeeexcOFC3n33XY477jgef/xxgsEgF198MVu3bmXSpEnJeHmSNGjIYC9JUqdOOeUUMjMzeeGFF9i1axd79uzB6/W2eczs2bMxGo0UFhYydepUSktLAZg5cyYmk4ns7GyysrJoamriq1/9Kps2beK5555j165dNDY2titPkqT4k2P2kiR16q233uLHP/4xFouF2bNnc8oppzBkyJA2j9E0LfbvI3fqOnLHLkVREEKwYsUKHnnkEbKzs/nGN75BSUkJMmO3JCWeDPaSJHVq7dq1XHjhhVxxxRW4XC7WrVvXbjezf/7znwghOHDgAJs2bepwr+5Wa9asYe7cuXzta18jEAiwbds2dF1P9MuQpEFPduNLktSpOXPm8OMf/5i///3vGI1Gpk6dyrp169o8xu/3c8UVVxAMBrn//vvJysrqtLxvfetbLF68mGeeeQaHw8FJJ51ERUVFol+GJA16ctc7SZJ67Y477uDUU09l9uzZya6KJEnHILvxJUmSJGmAk1f2kiRJkjTAySt7SZIkSRrgZLCXJEmSpAFOBntJkiRJGuBksJckSZKkAU4Ge0mSJEka4GSwlyRJkqQB7v8Db6FSqQxsADsAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 576x396 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ValidationCurve(ax=<AxesSubplot:title={'center':'Validation Curve for MLPClassifier'}, xlabel='alpha', ylabel='score'>,\n",
      "                cv=7, estimator=MLPClassifier(max_iter=1000),\n",
      "                param_name='alpha',\n",
      "                param_range=array([0.0001    , 0.00535789, 0.01061579, 0.01587368, 0.02113158,\n",
      "       0.02638947, 0.03164737, 0.03690526, 0.04216316, 0.04742105,\n",
      "       0.05267895, 0.05793684, 0.06319474, 0.06845263, 0.07371053,\n",
      "       0.07896842, 0.08422632, 0.08948421, 0.09474211, 0.1       ]))\n",
      "time for doing the validation and learning curve = 45.7353050419988 s\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7fcb20c54c40>"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x396 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Hyperparameters : hidden_layer_sizes=(100,), activation='relu', solver='adam',max_iter= 1000\n",
    "# Vary : alpha\n",
    "\n",
    "time1 = time.perf_counter()\n",
    "\n",
    "#validation curve\n",
    "print(validation_curve(MLPClassifier(hidden_layer_sizes=(100,), activation='relu', solver='adam',max_iter= 1000) , \n",
    "                       X , y , \n",
    "                       param_name ='alpha', \n",
    "                       param_range=np.linspace(0.0001,0.1,20), \n",
    "                       cv=7, \n",
    "                       score='accuracy'))\n",
    "\n",
    "#learning curve\n",
    "train_sizes, train_scores, valid_scores = learning_curve(MLPClassifier(hidden_layer_sizes=(100,), activation='relu', solver='adam',max_iter= 1000) , \n",
    "                                                         X , y , \n",
    "                                                         train_sizes=np.linspace(0.01,1,50), \n",
    "                                                         cv=7, \n",
    "                                                         scoring = 'accuracy', \n",
    "                                                         n_jobs= -1)\n",
    "\n",
    "time2 = time.perf_counter()\n",
    "timeA = time2 - time1\n",
    "print('time for doing the validation and learning curve =',timeA,'s')\n",
    "\n",
    "#means\n",
    "train_mean = np.mean(train_scores, axis = 1)\n",
    "valid_mean = np.mean(valid_scores, axis = 1)\n",
    "\n",
    "#standard deviation\n",
    "train_std = np.std(train_scores, axis = 1)\n",
    "valid_std = np.std(valid_scores, axis = 1)\n",
    "\n",
    "# plot the figure\n",
    "plt.plot(train_sizes,train_mean, label = 'Training score')\n",
    "plt.plot(train_sizes,valid_mean, label = 'Cross-validation score')\n",
    "plt.fill_between(train_sizes, train_mean - train_std, train_mean + train_std, color = '#DDDDDD')\n",
    "plt.fill_between(train_sizes, valid_mean - valid_std, valid_mean + valid_std, color = '#DDDDDD')\n",
    "\n",
    "plt.title('Learning curve')\n",
    "plt.xlabel('Training size')\n",
    "plt.ylabel('Accuracy score')\n",
    "plt.legend(loc = 'best')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "bfed5ca8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x396 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ValidationCurve(ax=<AxesSubplot:title={'center':'Validation Curve for MLPClassifier'}, xlabel='learning_rate_init', ylabel='score'>,\n",
      "                cv=7, estimator=MLPClassifier(max_iter=1000),\n",
      "                param_name='learning_rate_init',\n",
      "                param_range=array([0.0001    , 0.00535789, 0.01061579, 0.01587368, 0.02113158,\n",
      "       0.02638947, 0.03164737, 0.03690526, 0.04216316, 0.04742105,\n",
      "       0.05267895, 0.05793684, 0.06319474, 0.06845263, 0.07371053,\n",
      "       0.07896842, 0.08422632, 0.08948421, 0.09474211, 0.1       ]))\n",
      "time for doing the validation and learning curve = 33.2218367330006 s\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7fcb1dfada00>"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x396 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Hyperparameters : hidden_layer_sizes=(100,), activation='relu', solver='adam',max_iter= 1000, alpha = default\n",
    "# Vary : learning_rate_init\n",
    "\n",
    "time1 = time.perf_counter()\n",
    "\n",
    "#validation curve\n",
    "print(validation_curve(MLPClassifier(hidden_layer_sizes=(100,), activation='relu', solver='adam', max_iter= 1000) , \n",
    "                       X , y , \n",
    "                       param_name ='learning_rate_init', \n",
    "                       param_range=np.linspace(0.0001,0.1,20), \n",
    "                       cv=7, \n",
    "                       score='accuracy'))\n",
    "\n",
    "#learning curve\n",
    "train_sizes, train_scores, valid_scores = learning_curve(MLPClassifier(hidden_layer_sizes=(100,), activation='relu', solver='adam', max_iter= 1000 ) , \n",
    "                                                         X , y , \n",
    "                                                         train_sizes=np.linspace(0.01,1,50), \n",
    "                                                         cv=7, \n",
    "                                                         scoring = 'accuracy', \n",
    "                                                         n_jobs= -1)\n",
    "\n",
    "time2 = time.perf_counter()\n",
    "timeA = time2 - time1\n",
    "print('time for doing the validation and learning curve =',timeA,'s')\n",
    "\n",
    "#means\n",
    "train_mean = np.mean(train_scores, axis = 1)\n",
    "valid_mean = np.mean(valid_scores, axis = 1)\n",
    "\n",
    "#standard deviation\n",
    "train_std = np.std(train_scores, axis = 1)\n",
    "valid_std = np.std(valid_scores, axis = 1)\n",
    "\n",
    "# plot the figure\n",
    "plt.plot(train_sizes,train_mean, label = 'Training score')\n",
    "plt.plot(train_sizes,valid_mean, label = 'Cross-validation score')\n",
    "plt.fill_between(train_sizes, train_mean - train_std, train_mean + train_std, color = '#DDDDDD')\n",
    "plt.fill_between(train_sizes, valid_mean - valid_std, valid_mean + valid_std, color = '#DDDDDD')\n",
    "\n",
    "plt.title('Learning curve')\n",
    "plt.xlabel('Training size')\n",
    "plt.ylabel('Accuracy score')\n",
    "plt.legend(loc = 'best')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "83424962",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.97849462 0.97849462 1.         1.         1.         0.95652174\n",
      " 1.        ]\n",
      "accuracy score for upgraded model=  0.9876444266346089\n",
      "Test score for upgraded model = 0.9923076923076923\n",
      "time for training the model = 0.9845322450128151 s\n",
      "----------\n",
      "[0.98924731 1.         1.         1.         0.97826087 1.        ]\n",
      "accuracy score default model =  0.9922861150070127\n",
      "Test score default model = 0.9692307692307692\n",
      "----------\n",
      "0.05898107357648248 0.05290443547566238\n",
      "log_loss for the model by default : 0.05290443547566238\n",
      "log_loss for the upgraded model : 0.016520107608501845\n"
     ]
    }
   ],
   "source": [
    "# Hyperparameters : hidden_layer_sizes=(100,), activation='relu', solver='adam', max_iter= 1000, alpha = default, learning_rate_init=0.01\n",
    "\n",
    "from sklearn.metrics import log_loss\n",
    "from sklearn.metrics import zero_one_loss\n",
    "\n",
    "time1 = time.perf_counter()\n",
    "\n",
    "model = MLPClassifier(hidden_layer_sizes=(100,), \n",
    "                      activation='relu', \n",
    "                      solver='adam', \n",
    "                      max_iter= 1000,\n",
    "                      learning_rate_init=0.01)\n",
    "\n",
    "model.fit(X_train,y_train)\n",
    "predictions = model.predict(X_test)\n",
    "\n",
    "validation = cross_val_score(model, X, y, cv=7)\n",
    "print(validation)\n",
    "print('accuracy score for upgraded model= ', validation.mean())\n",
    "\n",
    "#Testing the test set\n",
    "predictions = model.predict(X_test)\n",
    "score = accuracy_score(y_test,predictions)\n",
    "print('Test score for upgraded model =', score)\n",
    "\n",
    "#time after training the model\n",
    "time2 = time.perf_counter()\n",
    "timeA = time2 - time1\n",
    "print('time for training the model =',timeA,'s')\n",
    "\n",
    "print('----------')\n",
    "\n",
    "#default model\n",
    "model2 = MLPClassifier()\n",
    "model2.fit(X_train,y_train)\n",
    "predictions = model2.predict(X_test)\n",
    "validation2 = cross_val_score(model2, X, y, cv=7)\n",
    "predictions2 = model2.predict(X_test)\n",
    "score2 = accuracy_score(y_test,predictions2)\n",
    "print(validation2)\n",
    "print('accuracy score default model = ', validation2.mean())\n",
    "print('Test score default model =', score2)\n",
    "\n",
    "print('----------')\n",
    "\n",
    "pred = model.predict_proba(X_test)\n",
    "eval1 = log_loss(y_test,pred)\n",
    "pred2 = model2.predict_proba(X_test)\n",
    "eval2 = log_loss(y_test,pred2)\n",
    "print(eval,eval2)\n",
    "\n",
    "print('log_loss for the model by default :',eval2)\n",
    "print('log_loss for the upgraded model :', eval1 )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "cdce59b9",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}