LSTM1.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "import numpy as np \n",
    "import pandas as pd \n",
    "from keras.preprocessing.text import Tokenizer\n",
    "from keras.preprocessing.sequence import pad_sequences\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense, Embedding, LSTM, SpatialDropout1D\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.preprocessing import LabelEncoder\n",
    "from keras.utils.np_utils import to_categorical\n",
    "import re\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Imbalanced \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "df= pd.read_csv('test17.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df\n",
    "df = df.dropna()\n",
    "df.isna().sum()\n",
    "df = df[[\"text\",\"Sentiment\"]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = df.text\n",
    "Y = df.Sentiment\n",
    "label_encoder = LabelEncoder()\n",
    "Y = label_encoder.fit_transform(Y)\n",
    "Y = to_categorical(Y)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "max_words = len(set(\" \".join(X).split()))\n",
    "max_len = X.apply(lambda x: len(x)).max()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(16434, 162)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "max_words, max_len"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "tokenizer = Tokenizer(num_words=max_words)\n",
    "tokenizer.fit_on_texts(X)\n",
    "X = tokenizer.texts_to_sequences(X)\n",
    "X = pad_sequences(X)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, Y_train, Y_test = train_test_split(X, Y, test_size=0.3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_1\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "embedding_1 (Embedding)      (None, 24, 100)           1643400   \n",
      "_________________________________________________________________\n",
      "spatial_dropout1d_1 (Spatial (None, 24, 100)           0         \n",
      "_________________________________________________________________\n",
      "lstm_1 (LSTM)                (None, 200)               240800    \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 3)                 603       \n",
      "=================================================================\n",
      "Total params: 1,884,803\n",
      "Trainable params: 1,884,803\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "None\n"
     ]
    }
   ],
   "source": [
    "embed_dim = 128\n",
    "lstm_out = 200\n",
    "\n",
    "model = Sequential()\n",
    "model.add(Embedding(max_words, 100, input_length=X.shape[1]))\n",
    "model.add(SpatialDropout1D(0.4))\n",
    "model.add(LSTM(lstm_out, dropout=0.3, recurrent_dropout=0.2))\n",
    "model.add(Dense(3,activation='softmax'))\n",
    "model.compile(loss = 'categorical_crossentropy', optimizer='adam',metrics = ['accuracy'])\n",
    "print(model.summary())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\ab\\anaconda3\\envs\\tf\\lib\\site-packages\\tensorflow_core\\python\\framework\\indexed_slices.py:424: UserWarning: Converting sparse IndexedSlices to a dense Tensor of unknown shape. This may consume a large amount of memory.\n",
      "  \"Converting sparse IndexedSlices to a dense Tensor of unknown shape. \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 4827 samples, validate on 1207 samples\n",
      "Epoch 1/10\n",
      "4827/4827 [==============================] - 15s 3ms/step - loss: 0.7803 - accuracy: 0.7464 - val_loss: 0.6965 - val_accuracy: 0.7548\n",
      "Epoch 2/10\n",
      "4827/4827 [==============================] - 12s 2ms/step - loss: 0.6159 - accuracy: 0.7661 - val_loss: 0.5396 - val_accuracy: 0.7788\n",
      "Epoch 3/10\n",
      "4827/4827 [==============================] - 12s 2ms/step - loss: 0.4299 - accuracy: 0.8245 - val_loss: 0.5241 - val_accuracy: 0.7945\n",
      "Epoch 4/10\n",
      "4827/4827 [==============================] - 11s 2ms/step - loss: 0.3165 - accuracy: 0.8720 - val_loss: 0.5411 - val_accuracy: 0.7920\n",
      "Epoch 5/10\n",
      "4827/4827 [==============================] - 11s 2ms/step - loss: 0.2279 - accuracy: 0.9159 - val_loss: 0.5523 - val_accuracy: 0.8028\n",
      "Epoch 6/10\n",
      "4827/4827 [==============================] - 11s 2ms/step - loss: 0.1468 - accuracy: 0.9540 - val_loss: 0.5777 - val_accuracy: 0.8061\n",
      "Epoch 7/10\n",
      "4827/4827 [==============================] - 12s 2ms/step - loss: 0.1014 - accuracy: 0.9695 - val_loss: 0.6318 - val_accuracy: 0.8094\n",
      "Epoch 8/10\n",
      "4827/4827 [==============================] - 11s 2ms/step - loss: 0.0743 - accuracy: 0.9778 - val_loss: 0.6536 - val_accuracy: 0.7970\n",
      "Epoch 9/10\n",
      "4827/4827 [==============================] - 11s 2ms/step - loss: 0.0586 - accuracy: 0.9847 - val_loss: 0.7651 - val_accuracy: 0.8020\n",
      "Epoch 10/10\n",
      "4827/4827 [==============================] - 11s 2ms/step - loss: 0.0512 - accuracy: 0.9869 - val_loss: 0.7634 - val_accuracy: 0.7970\n"
     ]
    }
   ],
   "source": [
    "verbose = 1\n",
    "epochs = 10\n",
    "batch_size = 120\n",
    "validation_split = 0.2\n",
    "history1 = model.fit(X_train,Y_train,batch_size=batch_size,epochs=epochs,verbose=verbose,validation_split=validation_split)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_performance(history=None, figure_directory=None, ylim_pad=[0, 0]):\n",
    "    xlabel = 'Epoch'\n",
    "    legends = ['Training', 'Validation']\n",
    "\n",
    "    plt.figure(figsize=(20, 5))\n",
    "\n",
    "    y1 = history.history['accuracy']\n",
    "    y2 = history.history['val_accuracy']\n",
    "\n",
    "    min_y = min(min(y1), min(y2))-ylim_pad[0]\n",
    "    max_y = max(max(y1), max(y2))+ylim_pad[0]\n",
    "\n",
    "\n",
    "    plt.subplot(121)\n",
    "\n",
    "    plt.plot(y1)\n",
    "    plt.plot(y2)\n",
    "\n",
    "    plt.title('Model Accuracy', fontsize=17)\n",
    "    plt.xlabel(xlabel, fontsize=15)\n",
    "    plt.ylabel('Accuracy', fontsize=15)\n",
    "    plt.ylim(min_y, max_y)\n",
    "    plt.legend(legends, loc='upper left')\n",
    "    plt.grid()\n",
    "\n",
    "    y1 = history.history['loss']\n",
    "    y2 = history.history['val_loss']\n",
    "\n",
    "    min_y = min(min(y1), min(y2))-ylim_pad[1]\n",
    "    max_y = max(max(y1), max(y2))+ylim_pad[1]\n",
    "\n",
    "\n",
    "    plt.subplot(122)\n",
    "\n",
    "    plt.plot(y1)\n",
    "    plt.plot(y2)\n",
    "\n",
    "    plt.title('Model Loss',fontsize=17)\n",
    "    plt.xlabel(xlabel, fontsize=15)\n",
    "    plt.ylabel('Loss', fontsize=15)\n",
    "    plt.ylim(min_y, max_y)\n",
    "    plt.legend(legends, loc='upper left')\n",
    "    plt.grid()\n",
    "    if figure_directory:\n",
    "        plt.savefig(figure_directory+\"/history\")\n",
    "\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x360 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_performance(history=history1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2587/2587 [==============================] - 2s 788us/step\n"
     ]
    }
   ],
   "source": [
    "accuracy = model.evaluate(X_test, Y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Model Performance of CNN (Test Accuracy):\n",
      "v: 81.02%\n",
      "Loss: 0.704\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(\"\\nModel Performance of CNN (Test Accuracy):\")\n",
    "print('v: {:0.2f}%\\nLoss: {:0.3f}\\n'.format(accuracy[1]*100, accuracy[0]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Balanced"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "df= pd.read_csv('test17.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df\n",
    "df = df.dropna()\n",
    "df.isna().sum()\n",
    "df = df[[\"text\",\"Sentiment\"]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = df.text\n",
    "Y = df.Sentiment\n",
    "label_encoder = LabelEncoder()\n",
    "Y = label_encoder.fit_transform(Y)\n",
    "Y = to_categorical(Y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "max_words = len(set(\" \".join(X).split()))\n",
    "max_len = X.apply(lambda x: len(x)).max()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "tokenizer = Tokenizer(num_words=max_words)\n",
    "tokenizer.fit_on_texts(X)\n",
    "X_train_seq = tokenizer.texts_to_sequences(X)\n",
    "X_train_seq = pad_sequences(X_train_seq)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "from imblearn.over_sampling import RandomOverSampler\n",
    "os=RandomOverSampler()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\ab\\anaconda3\\envs\\tf\\lib\\site-packages\\sklearn\\utils\\validation.py:71: FutureWarning: Pass classes=[0 1 2] as keyword args. From version 0.25 passing these as positional arguments will result in an error\n",
      "  FutureWarning)\n"
     ]
    }
   ],
   "source": [
    "X,Y= os.fit_sample(X_train_seq, Y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, Y_train, Y_test = train_test_split(X, Y, test_size=0.3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_2\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "embedding_2 (Embedding)      (None, 24, 100)           1643400   \n",
      "_________________________________________________________________\n",
      "spatial_dropout1d_2 (Spatial (None, 24, 100)           0         \n",
      "_________________________________________________________________\n",
      "lstm_2 (LSTM)                (None, 200)               240800    \n",
      "_________________________________________________________________\n",
      "dense_2 (Dense)              (None, 3)                 603       \n",
      "=================================================================\n",
      "Total params: 1,884,803\n",
      "Trainable params: 1,884,803\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "None\n"
     ]
    }
   ],
   "source": [
    "embed_dim = 128\n",
    "lstm_out = 200\n",
    "\n",
    "model = Sequential()\n",
    "model.add(Embedding(max_words, 100, input_length=X.shape[1]))\n",
    "model.add(SpatialDropout1D(0.4))\n",
    "model.add(LSTM(lstm_out, dropout=0.3, recurrent_dropout=0.2))\n",
    "model.add(Dense(3,activation='softmax'))\n",
    "model.compile(loss = 'categorical_crossentropy', optimizer='adam',metrics = ['accuracy'])\n",
    "print(model.summary())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\ab\\anaconda3\\envs\\tf\\lib\\site-packages\\tensorflow_core\\python\\framework\\indexed_slices.py:424: UserWarning: Converting sparse IndexedSlices to a dense Tensor of unknown shape. This may consume a large amount of memory.\n",
      "  \"Converting sparse IndexedSlices to a dense Tensor of unknown shape. \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 10884 samples, validate on 2721 samples\n",
      "Epoch 1/10\n",
      "10884/10884 [==============================] - 27s 2ms/step - loss: 0.8664 - accuracy: 0.5958 - val_loss: 0.4730 - val_accuracy: 0.8350\n",
      "Epoch 2/10\n",
      "10884/10884 [==============================] - 24s 2ms/step - loss: 0.3191 - accuracy: 0.8837 - val_loss: 0.2707 - val_accuracy: 0.9067\n",
      "Epoch 3/10\n",
      "10884/10884 [==============================] - 26s 2ms/step - loss: 0.1489 - accuracy: 0.9532 - val_loss: 0.2195 - val_accuracy: 0.9302\n",
      "Epoch 4/10\n",
      "10884/10884 [==============================] - 27s 2ms/step - loss: 0.0915 - accuracy: 0.9739 - val_loss: 0.1833 - val_accuracy: 0.9434\n",
      "Epoch 5/10\n",
      "10884/10884 [==============================] - 25s 2ms/step - loss: 0.0595 - accuracy: 0.9833 - val_loss: 0.2137 - val_accuracy: 0.9316\n",
      "Epoch 6/10\n",
      "10884/10884 [==============================] - 26s 2ms/step - loss: 0.0419 - accuracy: 0.9884 - val_loss: 0.2096 - val_accuracy: 0.9408\n",
      "Epoch 7/10\n",
      "10884/10884 [==============================] - 25s 2ms/step - loss: 0.0341 - accuracy: 0.9910 - val_loss: 0.1901 - val_accuracy: 0.9511\n",
      "Epoch 8/10\n",
      "10884/10884 [==============================] - 24s 2ms/step - loss: 0.0296 - accuracy: 0.9901 - val_loss: 0.2249 - val_accuracy: 0.9452\n",
      "Epoch 9/10\n",
      "10884/10884 [==============================] - 17s 2ms/step - loss: 0.0294 - accuracy: 0.9926 - val_loss: 0.2072 - val_accuracy: 0.9438\n",
      "Epoch 10/10\n",
      "10884/10884 [==============================] - 17s 2ms/step - loss: 0.0276 - accuracy: 0.9915 - val_loss: 0.1840 - val_accuracy: 0.9500\n"
     ]
    }
   ],
   "source": [
    "verbose = 1\n",
    "epochs = 10\n",
    "batch_size = 120\n",
    "validation_split = 0.2\n",
    "history1 = model.fit(X_train,Y_train,batch_size=batch_size,epochs=epochs,verbose=verbose,validation_split=validation_split)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_performance(history=None, figure_directory=None, ylim_pad=[0, 0]):\n",
    "    xlabel = 'Epoch'\n",
    "    legends = ['Training', 'Validation']\n",
    "\n",
    "    plt.figure(figsize=(20, 5))\n",
    "\n",
    "    y1 = history.history['accuracy']\n",
    "    y2 = history.history['val_accuracy']\n",
    "\n",
    "    min_y = min(min(y1), min(y2))-ylim_pad[0]\n",
    "    max_y = max(max(y1), max(y2))+ylim_pad[0]\n",
    "\n",
    "\n",
    "    plt.subplot(121)\n",
    "\n",
    "    plt.plot(y1)\n",
    "    plt.plot(y2)\n",
    "\n",
    "    plt.title('Model Accuracy', fontsize=17)\n",
    "    plt.xlabel(xlabel, fontsize=15)\n",
    "    plt.ylabel('Accuracy', fontsize=15)\n",
    "    plt.ylim(min_y, max_y)\n",
    "    plt.legend(legends, loc='upper left')\n",
    "    plt.grid()\n",
    "\n",
    "    y1 = history.history['loss']\n",
    "    y2 = history.history['val_loss']\n",
    "\n",
    "    min_y = min(min(y1), min(y2))-ylim_pad[1]\n",
    "    max_y = max(max(y1), max(y2))+ylim_pad[1]\n",
    "\n",
    "\n",
    "    plt.subplot(122)\n",
    "\n",
    "    plt.plot(y1)\n",
    "    plt.plot(y2)\n",
    "\n",
    "    plt.title('Model Loss',fontsize=17)\n",
    "    plt.xlabel(xlabel, fontsize=15)\n",
    "    plt.ylabel('Loss', fontsize=15)\n",
    "    plt.ylim(min_y, max_y)\n",
    "    plt.legend(legends, loc='upper left')\n",
    "    plt.grid()\n",
    "    if figure_directory:\n",
    "        plt.savefig(figure_directory+\"/history\")\n",
    "\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABJoAAAFWCAYAAAAhe/fuAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAgAElEQVR4nOzde3xcdZ3/8ddnJvfm2lx6v98vlLaUAqVAYuWOgIhCpUBRYHEVV13dRX8uILsqq67iKrsuyM0KdFmQi1hAwYY7FIpA6f1CS0NLm6RNmnsyme/vjzNJJmmSps1kZpK8n4/HcWbO+Z7v+Zw5ln77me/FnHOIiIiIiIiIiIj0li/WAYiIiIiIiIiIyMCgRJOIiIiIiIiIiESEEk0iIiIiIiIiIhIRSjSJiIiIiIiIiEhEKNEkIiIiIiIiIiIRoUSTiIiIiIiIiIhEhBJNItJrZnarmblonysiIiIikaM2nYhEghJNIgOEmS03Mxfazu2izOOh44FoxxdJZva50H2UmVlirOMRERERiZSB3qYLu7/FsY5FRPqGEk0iA089cGXHnWaWA5wXOt7fLQM+BHKBThtgIiIiIv3cYGjTicgApESTyMDzNHCRmaV32H9Z6PXPUY4nosxsKF7j6j+AtXTSAIsnZjYk1jGIiIhIvzSg23QiMnAp0SQy8DwMJAGXdNh/JfAn4GBnJ5nZl8zsPTOrN7NSM1thZqM7KXd+WLktZvblrgIxs8+b2RtmVmtmh8zsT2Z2XC/uDbzGlQ94BHgI+IyZZXdx/SIz+7OZVZhZjZmtM7PvdigzKXSve82swcw+NLO7zSwjdLyle/f4DueND+1fHrbvfjMLmNkYM/uDmVUAr4SOHWdm95rZNjOrM7MDZvakmc3sJO5EM7vJzNaHPY+/mNlpoeNvmNn6Lu75GTP7yMz033cREZH+baC36bplZilm9iMz22lmjaHXH5lZcodyc83saTPbF7qXj83ssfB77kkZEYkc/UNEZODZD/yFsJ4+ZjYRWASs6OwEM7sJuAc4BPwTcC9wKfBqqAdRS7lPAU8CqcC/hOr7EXBxJ3V+Gy8ZtAf4DvBDYFaozqm9uL8rgeedc6XASiAR+Hwn1/8i8DwwAfgF8I/AC8BFYWVmAG+F9j0I3Bi6p5PxhuUdCwOeAxqBm4C7Q/vPAmYCvwe+DvwKOAl42cyGh8XkA54Afgxsw3set+M9m9NDxe4HZprZCR3ueRhwJvB751zwGOMXERGR+DDQ23RdMjMD/gB8F3gZ+Abej3ffBR4NK5eP196bBvwc+CrwG2AYMKqnZUQkwpxz2rRpGwAbsBxwwGLgi0AzMDJ07GbgAN6vYvcDgbDz8vDG+L8MJITtPz9U30/C9q0FyoHcsH0zgID3n5PWfWOAJuCHHWIcForjwbB9t4afe4R7nBSKaVnYvheAlzqUywAqgHVAeodjFvb+r0AdML2Ta1mH73V8h+PjQ/uXh+27P7TvPzupL62TfZND3/3/C9t3VaiOW7uJKTsU9y87HP9W6NzD7kebNm3atGnT1j+2gd6mC7+/bspcECrz4w77fxraf17o80Whzyd2U9cRy2jTpi2ym3o0iQxMTwC1wNLQ52XA/znnGjsp+2kgGfiFc6515RLn3J+ADXh/0RPqdTMfr7dMeVi5jXg9eMJ9DkgAHjazvJYNr6H0OvCpY7yvK0P39UTYvoeAxR2Gtp0FZOE1TqrDK3DOudD95AGFwArn3KaOF2opd4z+q5P6alvem9kQM8vFS4ZtAcJ7Jn0eqMLrxdRpTM65CrxfIZeaWUJYkSuBtzq7HxEREemXBmqb7kguCL3+rMP+n3Q4XhF6vbDjkLowPSkjIhGkRJPIABRKavwBWGZmJwNT6KKLNV7PHIDOkhMb8IaehZfb3Em5jvtaulGvA0o7bOcBBd3eQNeW4XWbHm5mk81sMvAeEAwdazE57PpdmYQ3zK27MsdqR8cdZpZlZnea2X6gGijD+z6Ow+uh1GIysNU5d6SVZO4D8oGzQ/XPBuYCD/Q+fBEREYkHA7hNdyTjgdLwRBiA86ZOKKPtXl7Cm8vq+0C5mT1rZjeGftDjKMqISAQlHLmIiPRTv8dbjeRHwE7g1WOow/C6Gre8J+xzx3LhWpLYFwANx3Ddwy9gtggvOTQJ2NpJkSuBf+sQT3e9knpSprvj/i72N3fxK+NKvB5UPwfeweu1FATuoH3SP/w7785fgI/xhtr9KfTaFLqOiIiIDBwDqk0XAa33Eurt/UUz+xneEMFP483NebOZFTnnPuhJmZjchcgApkSTyMD1At6kjUV44+q7Sl7sDL1Ox/u1K9z0sOMfhu3rqONEkNtCr7udc+/3MN4juRKvJ9A1nRw7Hvi+mS10zq2hLRE1B+iq8bAtrEx3WlZ06biy3fgjnNfKvFXxzsGbd+kHHY7l4P0y12IrcIaZpXTXq8k5FzSzFcA3QnV8EfhTx1/+REREpN8baG26ntgJnG1mueFtm9CwvVza7gUA59w7eD/k/auZzcGbg+ofCWs39qSMiESGhs6JDFDOW3Xsa8APaFv5rDN/wfuF6h/C5/sxs3PxVhT5Y6i+T4C/4XXdzg0rN4PQ8K0wj+FNJvmD0Cpq7YRW/+gxM0sCvgA865x7tOMG/HvoHlpWZfkzUAl818zSO9RlofspA1YDV5nZYQ2tlnK0NbCKOhT56lHcQnPotd13YWbLgJEdyv4f3mTmN3UTU4v7gRTgv/FWTfndUcQkIiIi/cBAatMdhT+GXr/VYf93wo+bWU4n7aONeIumZPe0jIhElno0iQxgzrnHgcePUKbczG4Ffgy8YGaP4iUtvg58hJfEafHPwLPA62Z2F96SuF/D6zV0fFidH5rZP+ENE1tjZo/hrWwyFq9nzwd4K4701PnAUOCpLu6h2sz+ClxuZt9yzlWZ2Y148xX9LdTz5xO8X+kWhTaAG/G6n68J3c9mYARwCd7yvjudcxvM7BXgh6FlgfcBFwI5PQ0+FM9q4J/MLAXYDizAW26443xOv8frnXSLmc3F+xXTH4r5Xbxu8y31bjaz14HL8L7fP/U0JhEREek/BlCbLtxVZlbYyf7fA6tC8X3PzEYDbwIn4/2o+LRz7plQ2auBG83scbwfBxOAy/F+tHv4KMqISAQp0SQiOOduD01S/Q281T2q8CaevMk5dyCs3F/M7GLgh6FtF/A9vKVvj+9Q5y/MbAtel+Tv4v33Zg/eZN6/OcoQr8TrFdRdIuVJ4Fy8Rs8fnXMrzOyT0LW/g9ebaAde46UlxvVmthDvF8Kr8Roce/B+EQwfzrYsFPM/4q388jBeL6L1R3EPX8RrpH0ZrxfSGuBM4D/CC4WGxF0E/FPouucCh/C6er/YSb33A6cAK7uYG0pEREQGiX7Qpgt3XRf733DO7TSzS4Cb8dpQS4G9eEm028LKvoj3493ngOF47bT1wMXOuSePooyIRJD1bgVvERGJJTO7BrgXOCk0P5WIiIiIiEjMKNEkItKPhYbOZTrnZsU6FhEREREREQ2dExHpZ8xsCPAZ4FS8+Qq+HNuIREREREREPOrRJCLSz5jZeLyliSvxJjz/ZmhFGhERERERkZhSoklERERERERERCLCF+sARERERERERERkYBjQczTl5eW58ePH91n9NTU1DBkypM/ql2Oj5xKf9FwixwHOOYKu7TXoHO6w127K0PY5EAiQkOD9dWCh/zEMM9r2ARbaYRa2zyscVqatbFuZtjpbyrYv3+FaLWVDnzst0xpH++sD7b8HDv8ejvR9tLz2tHxvWOi78ln7VwNcMIjf7ws9b++548C7etg+vBhDh1tjCt8XD0Zmp5I7JKlP6l67dm2Zcy6/TyqXY9JXbbADNY18XFHH1GEZJCfo99J4o7/r45OeS/zSs4lPei490137a0AnmsaPH8/bb7/dZ/UXFxdTWFjYZ/XLsdFziU8D9bk0NQepb2qmIeC91je1fG6moSlIfaBtX31TkIbwz6Ey7fZ1LNdST9g1AsGjTx0Y4AeS/D6SE32kJPpJTvBe62prSElNI+igOehoDjqCzrW+bw69D4beB4O07uuKI34SHN1p970k+Ehu/V58JCf4u39N9JMSdk7455R2dXX+mpzgI8Hf9T+UI/lnJhj2HJuDjkDoeQZCz7rlc3P4vua2Y80dN+doDgZpDtLFvvBXr45Fk/KYOTIzIvfTkZnt6pOK5Zj1VRus5GAti/99Nf9w/gyuPW1ixOuX3hmof9f3d3ou8UvPJj7pufRMd+2vAZ1oEhE5kqbmIIfqmqioa6Kitin0vpHK2o77mqiobaSyronKugB1jQHqA8Fuky1HEp70SUn0kZLgb01CpCUlMHRIS9LC75ULJThayieHfe6YBPHKHJ4g8fvssDiO9S/Tlp4+7RJTLiyB0S5JRbtER2fnHJbYcq41kdExGdJ6Ttj1Wp5Fy713l+QJf01K8HX6vQwkPp/hw0j0xzoSkd4ZnZPGqHRj9eb9SjSJiIjEKSWaRKTfc85R09jclggKJYkqQ4miirpGL1lU27avMnS8uiHQbd0ZyQlkpSWSlZpIdloiw7NSyEpNJC0p4fDETigh1F2yJzyZ5OvnyQ0zw28M+CSNiMSXOfkJPP/hAaobAqQnqykrIiISb/S3s4jEjcZAsDUBVFnX2GliqCWZ1JJIqgwd6244WaLfyEpNIjstkezUREZkpTB9RIaXPArtz0pNbEsopSaSnZZEZkpCt0ObREQk+ubk+XnmwyZe3VbG2bOGxzocERER6WDQJZqampooKSmhvr6+13VlZWWxcePGCETV/6WkpDB69GgSExNjHYrEkfqmZkoO1lFysJbi3U1sKN7Wmhhq6WlUWRegMpQ8qmls7ra+jJSE1qRQdmoSI7NTWxNDLT2OslKTwt57r6mJ/taJrEVEJDYi1QY7eeJQ7hnpSK3fx8aNByMUXf+lNpiIiMSbQZdoKikpISMjg/Hjx/f6H55VVVVkZGREKLL+yzlHeXk5JSUlTJgwIdbhSBTVNzXzcUVdazLJe61j9wHvfVl1Q/sT1m8mye8jK62l11Aio7JTmDkis11iKKs1cZTUmkjKUO8iEZF+LVJtsKqqKg40+qhtbGb68IxB/UOC2mAiIhKPBl2iqb6+PiJJJmljZuTm5lJaWhrrUCTC6pua2RNKJO0OSyS1JJVKq9onkhL9xsjsVEbnpLJkegGjc1IZPTSV0TlpfLTxXc5bcgYpiT79+RMRGYQi2QbLSEmksq6J+qYgqUmDd5Z7tcFERCQeDbpEE6B/5PYBfaf9U3giqX2vJO91f4dEUoKvLZFUNC2f0TlpjAklkkbnpFKQkdLlxNA1O32D+h8DIiISufZCRorXhK1qaBr0f7eoDSYiIvFmUCaaYqm8vJwlS5YA8Mknn+D3+8nPzwdgzZo1JCUlHbGOa665hptuuolp06Z1WebOO+8kOzubK664IjKBS7/UEGhmT0X9YQmklvf7DnWdSCoMJZJG57QlkoZldp1IEhERiZZEv4/URD9VdQEKejiLgdpgIiIi0aFEU5Tl5uby7rvvAnDrrbeSnp7Ot7/97XZlnHM45/D5Op+P5r777jvidb761a/2PliJew2BZvZW1LdLIu0OSyrtr2rAhS3G5vcZI7NTGJ2dxulTwhNJqYwZmqZEkoiI9BsZKYmUVjUQaA72aA4/tcFERESiQ4mmOLFt2zYuvvhiFi9ezJtvvsnTTz/ND37wA9555x3q6uq47LLLuPnmmwFYvHgxv/71r5k9ezZ5eXnccMMNPPPMM6SlpfHkk09SUFDA97//ffLy8vjGN77B4sWLWbx4MX/961+prKzkvvvuY9GiRdTU1HDVVVexbds2Zs6cydatW/ntb3/L3LlzY/xtSIvGQJC9lXXsPnD4sLaSg3Xsq6rvcSJp9NA0hmUka0JtEREZEDJSEthfVU91Q4DstCP3RuqK2mAiIiKRpURTHNmwYQP33Xcfv/nNbwC4/fbbGTp0KIFAgKKiIi699FJmzpzZ7pzKykrOOOMMbr/9dr71rW9x7733ctNNNx1Wt3OONWvW8NRTT3Hbbbfx7LPP8qtf/Yrhw4fz2GOP8d577zF//vyo3Kd0rTEQZO2ug7y0tZSXtpSyYe+hwxJJI7JSGJ2TyuIpee2GtY3OSWV4ZooSSSIiMiikJfnx+4yq+t4lmkBtMBERkUga1ImmH/xxPRv2HDrm85ubm/H7209AOXNkJrd8ZtYx1Tdp0iROPPHE1s8PP/ww99xzD4FAgD179rBhw4bDGjmpqamce+65AJxwwgm8/PLLndZ9ySWXtJbZuXMnAK+88gr//M//DMDxxx/PrFnHFrccO+ccO8treWmLl1h6fUc5tY3NJPiM+WNz+FrRZMblDlEiSUREBpTetMHC218NgSDNQUdakl9tMBERkTgxqBNN8WbIkCGt77du3covf/lL1qxZQ3Z2NsuWLaO+vv6wc8InrvT7/QQCgU7rTk5OPqyMC+8qI1FzqL6J17aV89LWUl7eWsruA3UAjB2axiXzR3H6lHxOmZRLRkpijCMVERGJb36fEWgOEuxlm0ZtMBERkcgZ1ImmY/3Vq0VVVRUZGT1c6uQoHTp0iIyMDDIzM9m7dy/PPfcc55xzTkSvsXjxYh555BFOO+001q1bx4YNGyJav3iag451H1fy8pZSXtpayjsfVdAcdAxJ8nPKpDyuP20ip0/NZ1zukCNXJiIiMgD0pg0W3v4KNAfZsPcQwzJTGJaZEpHY1AYTERHpnUGdaIpn8+fPZ+bMmcyePZuJEydy6qmnRvwaN954I1dddRVz5sxh/vz5zJ49m6ysrIhfZzD6pLK+dZ6lV7aVUVHbBMBxo7K44YyJnD4ln3ljc0hK0DA4ERGRY5Xg95GWlEBVfYBhmZGpU20wERGR3rGB3HV3wYIF7u233263b+PGjcyYMSMi9fdlj6ZoCAQCBAIBUlJS2Lp1K2eddRZbt24lIeHY8o+R/G57o7i4mMLCwqhes76pmTUfHvDmWtpaypZ91QDkZyRz+pR8Tp+ax+LJeeSmJ0c1rngSi+ciPaNnE5/0XHrOzNY65xbEOg5p05dtsI7tr32H6tl3qJ6ZIzL7zTyGaoNJNOm5xC89m/ik59Iz3bW/1KNpEKuurmbJkiUEAgGcc/zP//zPMTdwBhvnHFv3V/PSllJe3FLKmg8P0BAIkpTgY+H4oVx6wmhOm5LP9OEZmFmswxUREemSmZ0D/BLwA791zt3e4fhY4AEgO1TmJufcqqgH2oWMlAT2HYKqhgA5vVx9LlrUBhMRkYFMf6MNYtnZ2axduzbWYfQbB2saeWVbGS9tKeXlrWV8csibGHRyQTpXnDSO06bmcfKEXFKT/EeoSUREJD6YmR+4EzgTKAHeMrOnnHPhkwZ9H3jEOfffZjYTWAWMj3qwXUhN9JPg81FV138STWqDiYjIQKZEk0gXmpqDvLu7IjQcroz3SypwDrJSE1k8OY/TpuRx2tR8RmWnxjpUERGRY7UQ2Oac2wFgZiuBi4DwRJMDWmZAygL2RDXCIzAzMlISOFTfhHNOPYlFRERiTIkmkTC7D9Ty4hZvEu/Xt5dT1RDAZzBvbA7/sGQKp0/N5/jR2fh9asSKiMiAMArYHfa5BDipQ5lbgT+b2Y3AEODT0Qmt5zJSEjhY20htYzNDktW8FRERiSX9TSyDWk1DgNe3l/PyVq/X0odlNQCMyk7lguNHcPqUfBZNziMrNTHGkYqIiPSJzn456bhSzFLgfufcf5jZKcAKM5vtnAu2q8jseuB6gGHDhlFcXNyukqysLKqqqnodcHNz8+H1OIcB5YdqCKb0jwnBI6m+vv6w7zsWqqur4yIOaU/PJX7p2cQnPZfeU6JJBpVg0LFh7yFe2ur1Wlq76yBNzY7URD8nTxzKVaeM4/Sp+UzMG6Ku9yIiMhiUAGPCPo/m8KFxXwbOAXDOvW5mKUAesD+8kHPuLuAu8Fad67hiz8aNGyOyWm9Xq/6W11fTEHT9ekXgY5WSksK8efNiHYZWaopTei7xS88mPum59N7g+8knDhQWFvLcc8+123fHHXfw93//912ek56eDsCePXu49NJLu6y341LCHd1xxx3U1ta2fj7vvPOoqKjoaej9UmlVA394p4RvrPwbC3/0PBf86hV+8uxmKusCfGnxBB669iTeveVM7rtmIdecOoFJ+elKMomIyGDxFjDFzCaYWRJwOfBUhzIfAUsAzGwGkAKURjXKHshITaCuqZmm5mCnx9X+EhERiY6o92jqwRK644B7gXzgALDMOVcSOtYMrAsV/cg5d2HUAo+gpUuXsnLlSs4+++zWfStXruSnP/3pEc8dOXIkjz766DFf+4477mDZsmWkpaUBsGpV3KxOHDFNQcdr28p4cWspL20pY+PeQwDkDknitCl5nD41n8VT8ijISIlxpCIiIrHlnAuY2deA5/DaZvc659ab2W3A2865p4B/BO42s2/iDatb7pzrOLwu5jKSE/mEeqrqAwwdcvjqc2p/iYiIREdUezSFLaF7LjATWBpaJjfcz4DfOefmALcBPw47Vuecmxva+mWSCeDSSy/l6aefpqGhAYCdO3eyZ88e5s6dy5IlS5g/fz7HHXccTz755GHn7ty5k9mzZwNQV1fH5Zdfzpw5c7jsssuoq6trLfeVr3yFBQsWMGvWLG655RYA/vM//5M9e/ZQVFREUVERAOPHj6esrAyAn//858yePZvZs2dzxx13tF5vxowZXHfddcyaNYuzzjqr3XXiSaA5yC1PfsBXX6jli799k3tf+ZCs1AT+6ZxpPH3jYt76f5/mjsvnccn80UoyiYiIhDjnVjnnpjrnJjnnfhjad3MoyYRzboNz7lTn3PGhNtifYxtx51ISfST6fVTVN3V6XO0vERGR6Ih2j6aeLKE7E/hm6P1q4ImoRhgFubm5LFy4kGeffZaLLrqIlStXctlll5Gamsrjjz9OZmYmZWVlnHzyyVx44YVdDuP67//+b9LS0nj//fd5//33mT9/fuuxH/7whwwdOpTm5maWLFnC+++/z9e//nV+/vOfs3r1avLy8trVtXbtWu677z7efPNNnHOcdNJJnHHGGeTk5LB161Yefvhh7r77br7whS/w2GOPsWzZsj79jo5WfVMzX3vobzy/cR+njUpg+afncvLEXK08I9KVxho4uBMO7iSv9D3YmQhpud6WmgN+/dkRkf7FzMhISaCytomgc/g6tJ/U/hIREYmOaP9LoidL6L4HfA5veN1ngQwzy3XOlQMpZvY2EABud871Lgn1zE3wybojl+tCanPg8H+MDT8Ozr298xPCtHTfbmno3HvvvTjn+N73vsdLL72Ez+fj448/Zt++fQwfPrzTOl566SW+/vWvAzBnzhzmzJnTeuyRRx7hrrvuIhAIsHfvXjZs2NDueEevvPIKn/3sZxkyZAgAl1xyCS+//DIXXnghEyZMYO7cuQCccMIJ7Ny584j3F01V9U1c97u3eWPHAW67aBZjG3ZSOGNYrMMSiS3noO4gHPgQDn7ovR7Y0fa++pPWorMB1v97+/NTstsST2m5MCS3/eeOW0oWaG6zwa05ALXlUFMKtWVQE9pqy7x9sy+FCafFOkqJF71og3Xa/gIYfhwZhf/KgZpGahubSe/kxya1v0RERPpetBNNPVlC99vAr81sOfAS8DFeYglgrHNuj5lNBP5qZuucc9vbXeAoltZNbmrE1xzgmDkIdDg/2NRIQw+W7l2yZAnf/OY3efnll6mpqWHKlCncc8897N27l+LiYhITE5k9ezZlZWWtjY+qqiqqq6sJBoNUVVURCASoq6trvZ9gMEhNTQ3r1q3jJz/5CcXFxeTk5HDDDTdQUVFBVVUVzjmqq6tJTk72biH0ua6ujoaGhta6GhoaqK+vp7q6msTExNb9gUCAmpqaTpcnjsXSulWNjp+/Xc+uqiDXz0lmbMNOLUcZp/Rc+oALktxwgJT6T0it20tqXfjrJyQ017Qr3pA0lLrU4dSlz6Q+b4n3PnU41fVNZCUGSGqsJLGpisSmQ21b9SESD24Kfa7E5zr/b6bDR1NiZtiW0fq+MSmr3eeWLehLVnKqG7H+M2PB5tbnnthUGfr/x6Gw997n1veB6k7r8f6/kcGO6kw+2dUc5buQwSY9OQEzo6q+qdNE08UXX8y3vvUt3nnnHerq6pg/fz73338/paWlrF27lsTERMaPH099fX231+mst9OHH37Iz372M9566y1ycnJYvnz5EevpbqqrlrYagN/v19A5ERHpN6KdaDriErrOuT3AJQBmlg58zjlXGXYM59wOMysG5gHbO5zf86V1L/x5r26mq+V1D59+8nAZGRkUFRVx4403csUVV5CRkUFDQwMjR45k6NChrF69mo8++oj09PTWa2RkZJCeno7P5yMjI4NPfepTPP7445x//vl88MEHfPDBBwwZMoRgMEhGRgajR4+mtLSU559/njPPPJOMjAwyMzNxrm3pXzMjPT2ds846i+XLl3PLLbfgnGPVqlWsWLGi3fXAa/Q0NTV1et/RXlp3b2Udy377Jntq4e6rFrAk1ItJy1HGJz2XYxRohMrdXm+kjr2TKnZBIOwfMeaH7LEwbALkFMLQiTB0AuRMgJzxJCelkQxkd7hEcXExc3vybJyDxmqv10ptOdQeaH1vteUk1ZSRFL6/civUHQDX+QpQJKR000tqaFhvqrzQkL6hkNCT/8IODBH/M9PS46i1t1FpqAdSWee9kOoOdl6P+ULPJg9y8mDINO/9kPxQz7eW996rpWST5PMxHZgeubuR/q4Hvb+7UtdF+wu82cyHJPmpqg8wIuvw4+np6RQWFvKlL32JpUuXAlBZWUlBQQGJiYmsXr2aXbt2dXv9008/nQcffJCioiI++OAD3n//fQAOHTrEkCFDyMrKYt++fTzzzDOtf4YzMjKoqqo6bOjc6aefzvLly7nppptwzvH444+zYsWKo/tCRERE4ky0E02tS+ji9VS6HPhieAEzywMOOOeCwHfxVqDDzHKAWudcQ6jMqcBPohl8pC1dupRLLrmElStXAnDFFVfwmc98hgULFjB37lymT+++Sf6Vr3yFa34eJKEAACAASURBVK65hjlz5jB37lwWLlwIwPHHH8+8efOYNWsWEydO5NRTT2095/rrr+fcc89lxIgRrF69unX//PnzWb58eWsd1157LfPmzYvbbto7Squ58p41HKpr4ndfWshJE3NjHZLIsWuobksgHfywfVKpsqR9oiYxzUsc5U2BKWe2JZKGToSsMX07t5IZJGd4W874np0TDEJ9RSj5VBaWpDo8WcXBnd7nhsqu60vObJ+ESstr/zklE3yJ4E8EX4L36k8K7Us4/Fhnn33+/tHTqjngJfJqSsOGqHUYrlYTNpStu8RR6tDWxBDDZrUlitJy2yWNSMuD1GzvOxKJUxkpieytrKMxECQp4fB1b9T+EhER6VsW7dVpzew84A7altD9YfgSumZ2Kd5Kcw5v6NxXQ8mlRcD/AEG81fLucM7d0921FixY4N5+++12+zZu3MiMGTMici9d9WgarCL53XZn/Z5Krr53Dc7BA19ayOxR7X+yHDA9Z4JBr7dKUlqsI4mIAfNcjoVzXiKl4zxJLUmlmtL25VOHhiWQQkmklvfpwyKeBIm7Z9Pc1D4B1e0WKtdUG9kYWhNQ3SWoepi4OqZyiWzYuJGZ44d3PedRXQWHjz4HMC/51pIYGpIb9j60hfc6Ss3p94kjM1vrnFsQ6zikTV+2wY7U/qpvambLvipGZaeSm57cZbmBJFptsCOJu79PBNBziWd6NvFJz6Vnumt/RX1ZIefcKmBVh303h71/FHi0k/NeA47r8wAlrr218wBfuu8tMlISWHHtSUzKT491SJFRUw7718O+DaHX9bB/EzTVQOYor/dK3rTQ61RvyxjeP3pdDBbBZji0p/NE0oGd0NhhXrPMUV7yaOo57ZNKORO8HiODmT8RMoZ5W0811noJp4ZDXqIqGAi9NnXyORC2v+Pn7s47wjlNtUe4Vtj+4JHnB5wJsBFaE0ctiaKCGZB2Wte9jgZA4kikN5ITfCT5fVTVBwZNoklERCSeaP1q6TdWb9rPVx5cy8jsVH7/5ZMYmZ0a65COXlM9lG32Ekr7PoD9G7z3YSuAkTrUG7oy/0rvH5Dl26BsC7z7oDc/TovkzLDEU0siaqqXrPAnRv/eBoOGKqjYHZozqUPvpIpd0NzYVtaXCDnjvMTR2FPa907KHgeJKbG7j4EoKa1/9f5z7ojJrzVr3mDhGed6SSYljkR6zMzISEnkYG0jQefw6UcZERGRqFKiSfqFJ9/9mH985D2mj8jggWsWxv8vlMEgVH7k9UwK76VUvh1caNUlfzLkT4NJRV5iqWCm99rV0CjnoGovlG6Gsq1e8qlsC+wohvcebivnS/CSGvnhPaCmQd5kbwl66Zxz3lxCFbuh4iMvmVTxUdtWufvwOW6S0r3vumA6TD+vfa+krNFKDkjXzELzR3WdFK4dsgfS86MYlMjAkZGSQHlNAzUNATJS9OOLiIhINA3KRJNzrtNlaeXY9eVcXyve2MXNT37AieOHcs/VC+KvwVh7oK1nUksvpf0b2/c+yh4Hw2bDzIvaEkpDJx3dxM1mkDnS2yYVtT9WfwjKt3oJqNLNoSTUVtjybPshOunDveRT/rSwnlBTvWFcA/3PRMs8SRW7OkkmhXopNRxqf05imjfBdvZYGL3Ae235nDPBG6o00L83EZEIilYbLD05ATOjqn7gJ5qiPd+qiIjIkQy6RFNKSgrl5eXk5uYq2RQhzjnKy8tJSYnsUCDnHP9VvJ2fPreZJdMLuPOK+aQkxrCHSKDBS+Dsa5lDaYP3WrW3rUxqDhTMgrlfDPVSmuX1dknu40njUzJh1AneFq65CQ7u8obrtSSfyrbA+//XfmWvxCHt53/KD70OnQgJcd57rEUwCNX7Ou+J1JJI6jhhdHKmlzTKHgvjF0P2mLBk0jhvyJL+OyEiEhHRbIP5fEZ6cgJV9UeeD60/66s2mIiISG8MukTT6NGjKSkpobS09MiFj6C+vl5/sYekpKQwevToiNXnnOPHz2zirpd28Nl5o/jJpXNI9B++RHGfcM5LULQkklqSSmVbw4a9JXnD0SacAcNmegmlYbPib4Juf6I3ZC5vMnB+237noHp/2/C7lu2j12HdI23lzOctYx8+EXnLkLzUnOjeS3PAS+qFJ48qdrUllipL2s+RBN58V9ljvcTZ5E+HkkphyaTBPum2iEgURaoN1tP2V3VDgIraJhrLkkmIVhsiBiLdBhMREemtQZdoSkxMZMKECRGpq7i4mHnz5kWkLmkTaA7yvcfX8cjbJVx9yjhu+cwsfL4+St7UVRyeUNq3of0KYdljvUTS9PPbeinlTurfE26bta3qNeG09scaa0I9n7a2T0Jtf6F9ImdIflsPqNZtipfA8R1Dgz7QCIdK2noftQxpq/jIm++q8uO2RF+L9GHe9UbMhRmfaeuJlD3Ge588QFYlFBEZACLVButp+2tnWQ2f/1kxt35mJstPjUzbT0RERI5s0CWaJL41BJr5h4ff5dn1n/D1JVP45qenRKZ7faDRS5Z0TCod+ritTEqWN4/S8Ze39VIqmOENSxtMkobAyLneFi7Y7PUg6jgP1IYn2k+SnZAa6kXVPgmV0FTtlW83pC0smVS1FwifZyI0J1X2WBhzMhzXYVhb1mit3CYiIl0anzeEiXlDWL25VIkmERGRKFKiSeJGTUOAv1uxlle2lfEvF8zky4uPoVHoHMn1pbDlOW9i7n0bQsPetrRNiu1L9IZ/jTvVSygNm+1N0J05Mr6GvcUbn9+bs2noRJh6dvtjNWVhvZ9CPaFK3oYP/kBL8mgxwKvh9SV4k5Bnj4WJhYcPa8scBQlJUbk1EREZmAqnFfD7N3dR19hMapJWAhUREYkGJZokLlTUNrL8vrdY93ElP/v88Vx6wlHMNdDcBDtfhk1/gk2rOKVqD7wROpY1xhvuNvWc0LC3md7wrv487C0eDcnztnGL2u9vqoPy7VC2he3vvMik4xe1JZMyRnjJKxERkT5SND2fe1/9kNd3lPGp6cNiHY6IiMigoESTxNy+Q/Vcdc8aPiyr4b+umM/Zs4Yf+aSGam/OoI1Pw9bnoL7SW4p+8hK2BD7D1NMuCQ17y+r7G5CuJabC8NkwfDa7y4Yy6fjCWEckIiKDyMIJQ0lN9LN6U6kSTSIiIlGiRJPE1K7yGpbd8yYHqhu5/5oTWTQ5r+vCNWWw+Rmv59KO1RCo91YVm36Bt00qgsRU9hQXM3XsydG7CREREYlLyQl+Tp2cx+rN+3HORWbeRxEREemWEk0SM5s+OcSV96wh0BzkoetO5vgxnSw1f3BXaEjc0/DR6+CC3nC4E67xVoEbewr49X9jERER6VzR9Hye37iP7aXVTC7IiHU4IiIiA57+hS4xsXbXQa65bw1pSQk89HenMGVYqOHnnLci3Kanve2Tdd7+gllw2rdhxgUwfI4m7RYREZEeKZxWAMDqTaVKNImIiESBEk0SdS9tKeXvVqxlWGYyK758EmOyk2HXa209lw7uBAzGnARn/RtMOw9yJ8U6bBEREemHRmWnMm1YBqs37+e60yfGOhwREZEBT4kmiapV6/byDyv/xoz8ZFYU1ZD1yj978y7VlII/yVvmfvE3veRSekGswxUREZEBoHB6Pve+8iFV9U1kpGjlWRERkb6kRJNEzR9eW8+Lf3qIB9Lf45Tad7DHayApA6ae5U3mPfnTkJIZ6zBFRERkgCmaVsD/vLiDV7eVc87sHqxuKyIiIsdMiSbpW1WfwKY/sfv1/+OC8jVcktiMSxiGTf+Cl1yacBokJMc6ShERkUHLzM4Bfgn4gd86527vcPwXQFHoYxpQ4JzrZAWP+HXCuBwykhMo3rxfiSYREZE+pkSTRF7ZtrbJvEveAiAQHMaLuZdSeNGXSBy7EHy+GAcpIiIiZuYH7gTOBEqAt8zsKefchpYyzrlvhpW/EZgX9UB7KdHv47SpeazevB/nHKZFRURERPqMEk3Se87Bnr+Fkkt/gtJN3u4Rc/nL8Ov46a4pnHjiKfzrxcfh96lhJyIiEkcWAtucczsAzGwlcBGwoYvyS4FbohRbRBVOK2DVuk/YuLeKmSM1VF9ERKSvKNEkx6a5KbRSXCi5dOhjMD+MPxUWfInGyefwrefKefr9vfx94SS+c/Y0/XooIiISf0YBu8M+lwAndVbQzMYBE4C/RiGuiCucmg9A8Zb9SjSJiIj0ISWapOcaa2H7C15iafMzUF8BCakweQl86l9g6tmQNpS6xmZu+P1aXtxSynfPnc7fnTEp1pGLiIhI5zr7Fch1UfZy4FHnXHOnFZldD1wPMGzYMIqLiyMSYEfV1dXHXPe4TB9PvLmVmZRENigBevdspO/oucQvPZv4pOfSe0o0SfdqD8CWZ2Hj07D9rxCog9QcmHYeTD8fJn0KktJai1fWNfHl+9/inY8Ocvslx3H5wrExDF5ERESOoAQYE/Z5NLCni7KXA1/tqiLn3F3AXQALFixwhYWFEQqxveLiYo617gsbN/NfxduZt/BUstISIxuY9OrZSN/Rc4lfejbxSc+l95RoksNV7PZ6LW162hse55ohczTMvwpmXABjF4H/8P/rlFY1cNW9a9i2v4pff3E+5x03IgbBi4iIyFF4C5hiZhOAj/GSSV/sWMjMpgE5wOvRDS+yCqcV8Ku/buPlbaVcMGdkrMMREREZkJRoEm8y7/0bQ8mlP8Le97z9+TNg8Te95NKIudDNHEu7D9Ry5T1vsu9QA/cuP5HTpuRHKXgRERE5Vs65gJl9DXgO8AP3OufWm9ltwNvOuadCRZcCK51zXQ2r6xfmjskmOy2R1ZuUaBIREekrSjQNdhW7YeVS+GQdYDD6RDjzNph+AeT2bG6lrfuquPKeNdQ2Bvj9tSdxwricvo1ZREREIsY5twpY1WHfzR0+3xrNmPqK32ecPiWfF7fsJxh0+LQaroiISMQp0TSYHdgBD1wI9Yfg/P/wkksZw4+qivd2V3D1fWtI9Pt45IZTmD5cq7iIiIhI/Cqans9T7+3hgz2VzBmdHetwREREBhxftC9oZueY2WYz22ZmN3VyfJyZvWBm75tZsZmNDjt2tZltDW1XRzfyAWb/Jrj3XGisgaufghOvPeok02vby/ji3W+QkZLAYzcsUpJJRERE4t7pU/Ixg9WbSmMdioiIyIAU1USTmfmBO4FzgZnAUjOb2aHYz4DfOefmALcBPw6dOxS4BTgJWAjcYmYao3Us9r4P958HOLhmFYyce9RV/Hn9Jyy/7y1G56Tx6A2LGJubduSTRERERGIsNz2Z40dns3rz/liHIiIiMiBFu0fTQmCbc26Hc64RWAlc1KHMTOCF0PvVYcfPBv7inDvgnDsI/AU4JwoxDywlb8MDF0BCKlzzDBTMOOoqHl1bwlcefIeZIzL53787mWGZKX0QqIiIiEjfKJpWwHslFZRXN8Q6FBERkQEn2nM0jQJ2h30uweuhFO494HPAL4HPAhlmltvFuaM6XsDMrgeuBxg2bBjFxcWRiv0w1dXVfVp/pGVVrOe4dbfRlJjNuzNvoWHdbtp/pUf2551NPLSpkVm5Pm6Y1si7a17rm2B7ob89l8FCzyV+6dnEJz0Xkb5TND2fXzy/hZe2lvLZeaOPfIKIiIj0WLQTTZ0t7dFxmdxvA782s+XAS8DHQKCH5+Kcuwu4C2DBggWusLCwF+F2r7i4mL6sP6K2vQCv/CvkjCPhqic5JXPEUZ3unOMXz2/loU1bOXf2cO64fC7JCf4+CrZ3+tVzGUT0XOKXnk180nMR6TuzR2aRl57E6k1KNImIiERatBNNJcCYsM+jgT3hBZxze4BLAMwsHficc67SzEqAwg7nFvdlsAPGplXwf1dD3jS46gkYkndUpweDjh/8cT0PvL6LLywYzY8+exwJ/qjPIy8iIiISET6fccbUAp7fuI/moMPv6+z3TBERETkW0c4WvAVMMbMJZpYEXA48FV7AzPLMrCWu7wL3ht4/B5xlZjmhScDPCu2T7nzwGDxyJQw/Dpb/8aiTTE3NQb71yLs88PourjttAv/+uTlKMomIiEi/VzQ9n8q6Jt7dfTDWoYiIiAwoUc0YOOcCwNfwEkQbgUecc+vN7DYzuzBUrBDYbGZbgGHAD0PnHgD+FS9Z9RZwW2ifdOVvD8Jj18LohXDlE5B6dIv01Tc1c8OKtTzx7h6+c/Y0vnfeDMz0i5+IiIj0f6dNzsfvM1ZvKo11KCIiIgNKtIfO4ZxbBazqsO/msPePAo92ce69tPVwku6suRtWfRsmFsHlD0FS2lGdXlXfxLUPvM2anQf4t4tns+zkcX0UqIiIiEj0ZaUlcsLYHFZv3s+3z54W63BEREQGDI2BGohe/U8vyTTtPFi68qiTTOXVDSy9+w3W7jrILy+fpySTiIiIDEiF0/NZv+cQ+w7VxzoUERGRAUOJpoHEOSj+d/jLv8CsS+ALv4PElKOq4uOKOj7/P6+zbX81d1+1gAuPH9lHwYqIiIjEVtG0AgBe3KzhcyIiIpGiRNNA4Rw8fwsU/wjmXgGf+y34E4+qiu2l1Xz+v1+jtKqBFV8+iaLpBX0UrIiIiEjsTR+ewfDMFFZv3h/rUERERAYMJZoGgmAQnvknePWXcOK1cOGvwec/qio++LiSL/zmdRqbg6y8/mROHD+0j4IVERERiQ9mRtH0fF7eWkZTczDW4YiIiAwISjT1d8Fm+OONsOYuWHQjnPcz8B3dY31zRzlL73qDlEQ/j/zdKcwamdVHwYqIiIjEl8JpBVQ3BHh758FYhyIiIjIgKNHUnzU3wR+ug7/9Hs64Cc78VzA7qirqm5q59oG3KchM5tGvnMLE/PQ+ClZEREQk/pw6OY9Ev1Gs4XMiIiIRoURTfxVogEeuhg8egzNvg6LvHnWSCeDtnQepagjw/QtmMiIrtQ8CFREREYlf6ckJLJwwVPM0iYiIRIgSTf1RYy08vBQ2/8kbKnfqPxxzVa9tLyPBZyzUnEwiIiIySBVNK2DLvmpKDtbGOhQREZF+T4mm/qahCh78PGz/qzfp98LrelXdq9vLmTsmmyHJCREKUERERKR/KZzmrbRbvLk0xpGIiIj0f0o09Sd1FbDis/DR6/C538L8K3tV3aH6JtaVVLBocl6EAhQRERHpfyblD2HM0FTN0yQiIhIBSjT1FzVl8MAFsPc9+MLv4LhLe13lmzsOEHSwaFJuBAIUERER6Z/MjKJpBby6rZz6puZYhyMiItKvKdHUH1R9AvefD2VbYenDMOOCiFT72vYyUhJ9zBubHZH6RERERPqromkF1DU189bOA7EORUREpF9ToineVeyG+86FyhJY9hhM/nTEqn5tWzknjh9KcoI/YnWKiIiI9EcnT8wlOcHH6k2ap0lERKQ3lGiKZ+XbvSRTbTlc+QSMXxyxqkurGti8r4pFkzQ/k4iIiEhqkp9TJuVqniYREZFeUqIpXu3fBPedB021cPUfYcyJEa3+9R3lgOZnEhERGezM7Bwz22xm28zspi7KfMHMNpjZejN7KNoxRkvRtAJ2lNWws6wm1qGIiIj0W0o0xaO978P95wEOlq+CEcdH/BKvby8jIyWB2aOyIl63iIiI9A9m5gfuBM4FZgJLzWxmhzJTgO8CpzrnZgHfiHqgUVI0rQBAvZpERER6QYmmeFPytre6XEIqXPMMFEzvk8u8uq2ckyfm4vdZn9QvIiIi/cJCYJtzbodzrhFYCVzUocx1wJ3OuYMAzrkBm4UZm5vGxPwhrN6seZpERESOVUKsA5AwO1+Bhy6DIflw9VOQPbZPLrP7QC0fHajlS6eO75P6RUREpN8YBewO+1wCnNShzFQAM3sV8AO3Ouee7ViRmV0PXA8wbNgwiouL+yJeqqur+6xugClpDbywrYbnXlhNsl8/yB2Nvn42cmz0XOKXnk180nPpPSWa4sW252HlMi+5dNWTkDmizy71+vbQ/EyTNRG4iIjIINdZJsV1+JwATAEKgdHAy2Y22zlX0e4k5+4C7gJYsGCBKywsjHiwAMXFxfRV3QAJo8p47p43SRw1k8Lpw/rsOgNRXz8bOTZ6LvFLzyY+6bn0nobOxYNNf4KHl0LeZLhmVZ8mmQBe3V5GXnoyUwrS+/Q6IiIiEvdKgDFhn0cDezop86Rzrsk59yGwGS/xNCCdOCGHtCQ/qzdp+JyIiMixUKIp1j54DP73Shg+x1tdbkjf9jJyzvHa9nIWTcrFTN3BRUREBrm3gClmNsHMkoDLgac6lHkCKAIwszy8oXQ7ohplFCUn+Fk0KY/Vm/fjXMfOXSIiInIkSjTF0t8ehMeuhbEnw1VPQGpOn19y2/5qSqsaOHVybp9fS0REROKbcy4AfA14DtgIPOKcW29mt5nZhaFizwHlZrYBWA18xzlXHpuIo6Noej4lB+vYXlod61BERET6Hc3RFCtr7oZV34ZJn4LLHoSktKhc9rWW+ZkmaX4mERERAefcKmBVh303h713wLdC26BQOK0AgNWbSplckBHjaERERPoX9WiKhVd/6SWZpp0PS1dGLckE8Oq2MsYMTWXM0OhdU0RERKQ/GZWdyrRhGazevD/WoYiIiPQ7SjRFk3NQfDv85WaY/Tn4wgOQkBy1yzcHHW/sKGfRRPVmEhEREelO4fR83tp5gKr6pliHIiIi0q9EPdFkZueY2WYz22ZmN3VyfKyZrTazv5nZ+2Z2Xmj/eDOrM7N3Q9tvoh17rzjnJZiKfwxzl8Eld4M/MaohrN9TyaH6AIs0P5OIiIhIt4qmFdDU7Hh124CejkpERCTioppoMjM/cCdwLjATWGpmMzsU+z7eRJTz8FY++a+wY9udc3ND2w1RCToSgkFY9R147T/hxOvgwl+Bzx/1MFrmZzplkhJNIiIiIt05YVwOGckJFGv4nIiIyFGJdo+mhcA259wO51wjsBK4qEMZB2SG3mcBe6IYX+QFm+GpG+Gtu2HR1+G8n4IvNiMWX91WxtRh6RRkpMTk+iIiIiL9RaLfx2lT81i9eT/efOgiIiLSE9HOeIwCdod9LgntC3crsMzMSvBWQLkx7NiE0JC6F83stD6NNBKam+AP18G7v4fC78KZt4FZTEJpDAR5a+cBrTYnIiIi0kOF0wrYd6iBjXurYh2KiIhIv5EQ5et1lmXp+BPRUuB+59x/mNkpwAozmw3sBcY658rN7ATgCTOb5Zw71O4CZtcD1wMMGzaM4uLiiN9Ei+rq6i7rt2ATs9b/lLzyN9k+8Wp2czK8+GKfxXIkmw80U98UJLNuL8XFpTGLIxq6ey4SO3ou8UvPJj7puYjEXuHUfABWb97PzJGZRygtIiIiEP1EUwkwJuzzaA4fGvdl4BwA59zrZpYC5Dnn9gMNof1rzWw7MBV4O/xk59xdwF0ACxYscIWFhX1wG57i4mI6rb+xFv73Cih/E877GZMWXsekPouiZ/72ly34bCtfvvAMslKjOwl5tHX5XCSm9Fzil55NfNJzEYm9gswUZo/KpHjzfr5aNDnW4YiIiPQL0R469xYwxcwmmFkS3mTfT3Uo8xGwBMDMZgApQKmZ5YcmE8fMJgJTgB1Ri7ynGqrgwUthRzFcdCcsvC7WEQHw2vYyjhuVNeCTTCIiIiKRVDStgLW7DlJZ2xTrUERERPqFqCaanHMB4GvAc8BGvNXl1pvZbWZ2YajYPwLXmdl7wMPAcufNwHg68H5o/6PADc65A9GM/4jqDsLvLoaP3oBL7oZ5y2IdEQC1jQH+9lEFp2h+JhEREZGjUjitgKCDl7YO7KkHREREIiXaQ+dwzq3Cm+Q7fN/NYe83AKd2ct5jwGN9HuCxqimDFRdD6Wa4bAVMPz/WEbVa8+EBAkHHqZNzYx2KiIiISL8yd0w22WmJrN68n88cPzLW4YiIiMS9qCeaBqRDe70k08FdsPRhmPzpWEfUzuvby0ny+1gwbmisQxERERHpV/w+44yp+by4uZRg0OHzxWYFYRERkf6iR0PnzOwCM4v2fE79Q8VHcN+5UFkCyx6LuyQTwKvby5g3NpvUJH+sQxERERHpd4qmFVBe08i6jytjHYqIiEjc62ny6EngYzP799AE3QKk1u6Fe8+FugNw1ZMw/rARfzFXUdvI+j2HWKT5mURERESOyelT8zGD1Zv3xzoUERGRuNfTRNMk4C7gC8AHZva6mV1nZpl9F1qc27+Jue9+FwJ1cPXTMHpBrCPq1Bs7ynEOzc8kIiIicoyGDkli7phsijdrQnAREZEj6VGiyTm30zl3i3NuAnAmsA34BbDXzFaYWVFfBhl3PlkH95/nvV++CkbMiW083XhtezlpSX6OH5Md61BERERE+q2iaQW8V1JBeXVDrEMRERGJa0c975Jz7q/OuSuBqcBa4ArgeTP70My+aWYDf4Lx5EzIm8q7c38MBdNjHU23Xt1WxsIJQ0n0a4otERERkWNVNK0A5+ClrerVJCIi0p2jzj6Y2Rlmdj+wGZgN3AmcBfwf8APgd5EMMC7ljINrnqEubUSsI+nWvkP1bC+t4VTNzyQiIiLSK7NGZpKXnszqTUo0iYiIdKdHvY/MbBxwdWgbDxQD1wN/cM619B9+wcxeB34f+TDjkMX/0ravbS8D4JRJmp9JREREpDd8PqNwWj5/2bCP5qDD74v/tqCIiEgs9LRH0w7gOuAhYLJzbolz7uGwJFOL9cCaSAYox+61beVkpyUyc8TgnbNdREREJFKKphVQWdfEu7sPxjoUERGRuNXT+ZQ+AzzrnAt2V8g5twUYXBODxynnHK9tL+eUibn49IubiIiISK8tnpKH32es3lTKCeOGxjocERGRuNTTHk0vA8M6O2BmI8wsPXIhSSR8dKCWjyvqWDRZ8zOJiIiIREJWaiInjMth9eb9sQ5FREQkbvU00XQPcFsXx24FfhuRaCRiXt1WDsAizc8kIiIiEjFF0wpYv+cQ+/4/e3ceX1V17n/882SeQ2YgCVPAPXvDdgAAIABJREFUME8GEKgaqnWorVZtHXq11Wpte9Xetj9vr3bweu1we+1gJzvQarWj1VYtDtVaJWoJICgKAgIJICSMCZAQMifr98c+wCEGOJgzJfm+X6/9OmfvtYfnsBIgT9Z6VmNrpEMRERGJSoEmms4Cnj5O2zO+dokiS6rrGJqRxJjc1EiHIiIiIlHMzC4wsw1mVmVmt/fSfp2Z7TWzN3zbjZGIM1osGJ8HwEsbtPqciIhIbwJNNGUCzcdpawWyghOOBEN3t2NZdT3zSnKwfrA6noiIiESGmcUC9wEXAhOBq81sYi+n/tk5N923DeqR7KUF6QzLTNL0ORERkeMINNG0CbjoOG0fBKqDE44Ew4bdB6k/1K76TCIiInIys4Eq59xm51w78DBwSYRjimpmRnlpHq9sqqOj64Tr5IiIiAxKgSaafgLcYmbfNbNJZpbte70HuBn4UehClFNVWa36TCIiIhKQQmC7336N71hPl5vZajP7i5kVhye06FVemk9TWycrt+6PdCgiIiJRJy6Qk5xzvzKzAuAO4Et+Ta3A15xzvwpFcPLeVFbVMTo3leFDkiMdioiIiES33ubYux77TwJ/cs61mdlngYeA97/rRmY3ATcBFBQUUFFREeRQPU1NTSG7d6C6Ox2xBr/952u0bU+IaCzRJBr6Rt5N/RK91DfRSf3SdwElmgCcc980s58Ac4EcoB5Y6pxrCFVwcuo6u7pZvmUfF08fHulQREREJPrVAP4jlIqAHf4nOOfq/XZ/Bfxfbzdyzi0EFgKUlZW58vLyoAZ6WEVFBaG696k4Y+syqg+2UV5+dqRDiRrR0jdyLPVL9FLfRCf1S98FOnUOAOdcg3PuWefcH3yvSjJFmdW1DTS1dTK/RPWZRERE5KRWAOPMbLSZJQBXAYv8TzCzYX67FwPrwxhf1FpQms/G3U3U7D/eejkiIiKDU8Ajmsxbvmw+cBqQ1LPdOfezIMYl79FSX32mM8ZkRzgSERERiXbOuU4zuwV4DogFHnDOrTWzu4GVzrlFwOfN7GKgE9gHXBexgKNIeWk+33x6PRUb9nLNGSMjHY6IiEjUCCjR5KvP9ALesreOo/P5/efwK9EUBZZU1TFhWAY5aYmRDkVERERCzMzygVTn3BbfvgGfxvs/2wvOuSdPdg/n3DPAMz2O3en3/g68Op3ipyQvleLsZCo27FGiSURExE+gU+e+DzTgzeE3YA4wCvg6sAlvlJNEWGtHFyvf2a/V5kRERAaPB4Ev+u3/D94v/y4AHjez6yIQ06BgZiwozWdJVT2tHV2RDkdERCRqBJpoOhsv2bTTt2/OuW3OuW8Dv0ejmaLC6+/sp72zm/ljlWgSEREZJGYCLwKYWQzwOeArzrnxwLeAL0QwtgFvQWk+LR1dvLplX6RDERERiRqBJpqGAHudc91AI5Dv11YJzAt2YHLqKqvriY0xZo1SfSYREZFBIhNvJWCA04Fs4A++/ReBsZEIKmRaG8mufy3SURxxxpgcEuNiWLxhT6RDERERiRqBJpq2AIdXHFkL/Jtf24fxCkNKhC2prmNaUSbpSfGRDkVERETCowavHhPARcDbzrla334m0BqRqELlhbuZ/Na3YNvySEcCQHJCLHNLcqjYsDfSoYiIiESNQBNNzwDn+d5/E7jczGrMbAvweeAngT7QzC4wsw1mVmVmt/fSPsLMFpvZKjNbbWYf9Gu7w3fdBjM7P9BnDgYHWztYXdPAvJLcSIciIiIi4fMAcI+ZPQp8GVjo13YGsD4iUYXK+79KW2IePPpJaIqOUUQLSvPZUneILXWHIh2KiIhIVAgo0eScu905d6Pv/d/xpso9BDwOfMg5971A7mNmscB9wIV4v3272swm9jjta8AjzrkZwFX46j/5zrsKmIRX4PJnvvsJ8OqWfXR1O+apPpOIiMig4Zz7X+BWYJfv9cd+zdnAryMRV8gkZ/HW5Nuh5QA8ej10dUY6IhaUehUlKjR9TkREBAgg0WRmiWb2VTObdviYc26lc+6rzrkv+RJPgZoNVDnnNjvn2oGHgUt6nOOADN/7TGCH7/0lwMPOuTbfEr5VvvsJXn2mxLgYZo7IinQoIiIiEkbOud865251zt3vnHN+xz/rnHsokrGFwqG00fDhH8E7/4IX7op0OIzISWFMXiqLNX1OREQECCDR5JxrA76KVxC8rwqB7X77Nb5j/u4CrjGzGrwpe7eewrWD1pKqOspGZZEUr0FeIiIig4WZTTCzM/z2U8zs22b2hJndeqJr+7VpV8KsT0PlT2DtE5GOhgWl+SzbXE9ze+RHWImIiERaXIDnLcdbyeSlPj7PejnmeuxfDTzonPu+mc0FfmdmkwO8FjO7CbgJoKCggIqKir5FfAJNTU0hvX+gGtsdb+9q5vJx8VERT6RFS7/IsdQv0Ut9E53ULxKgn+GtALzMt/9d4DrgFeD/zCzJOffdCMUWWud/G3a+AX+7GfInQF5pxEJZUJrP/f/awtLqes6ZUBCxOERERKJBoImmLwN/NLN2vFFGu+mR5HHONQdwnxqg2G+/iKNT4w67Aa8GE865pWaWBOQGeC3OuYX4CmGWlZW58vLyAMJ6byoqKgjl/QP11OodwCqu+cAsZmjqXNT0ixxL/RK91DfRSf0iAZoMfB/AzOKBa4AvOOd+ZWZfAD6Dl3waeOIS4GMPwcKz4c/XwKdfhMT0iIQya3QWKQmxVGzYq0STiIgMeoGuOrccKMErMLkJaAQO9tgCsQIYZ2ajzSwBr7j3oh7nbAPOAW84OJAE7PWdd5WvZtRoYBzwaoDPHdAqq+tJT4xjSmFmpEMRERGR8ErF+38ZeKvMpQKP+fZfB0ZGIqiwySyEj/4G6qvhiX8H967B7mGRGBfL/LG5LN6wBxehGERERKJFoCOaPkUv09ROlXOu08xuAZ4DYoEHnHNrzexuYKVzbhHw/4BfmdkXfc+8zlfYcq2ZPQKsAzqBm51zXX2NaSCorKpjzphs4mIDzRuKiIjIALEZL8H0MnApsMo5V+9ryyXwXwb2X6PPhHPvgue/7tVsmv/5iISxoDSf59ftpnpvE2PzIzOySkREJBoElGhyzj0YrAc6557Bm37nf+xOv/frgPnHufZbwLeCFctAUHugha31zVw7d1SkQxEREZHwuxf4uZl9DJgBXO/XVg6sjkRQYTfvVqhdCf/8bxg+HUafFfYQykvzAFj89l4lmkREZFDTEJh+rrKqDoD5Y3MiHImIiIiEm3PufuBc4GHgfOfc7/ya9wE/jEhg4WYGl9wHOWPh0euhoTbsIQwfksz4oeks3rAn7M8WERGJJgElmsxsr5ntOdEW6kCld0ur68lJTeA0/eZMRERkUHLOveyc+75z7oUex+9yzj0dqbjCLjEdrvw9dLbCo5+Ezvawh1Bems+Krfs42NoR9meLiIhEi0BHNN3Xy/YnvNXnWoCfhiQ6OSHnHEuq65hbkkNMjEU6HBEREYkAMxtiZv9lZk+a2RLf65fNbEikYwu7vFJvZFPNCnjuK2F//ILSPDq6HEuq6k9+soiIyAAVaI2mu3o7bmYGPIJXnFvCrHrvIXY3tjF/bG6kQxEREZEIMLMS4CUgD1iCt3pvAXA3cIuZLXDOVUcwxPCb9BGovdUrDF5UBtOuCtujZ47MIj0pjooNe7hg8tCwPVdERCSa9KlGk281uF8DtwQnHDkVS6u9+kzzSlSfSUREZJC6F9gPjHHOvd85d7Vz7v1ACXAA+EFEo4uUc+6CUWfCk1+AXWvC9tj42BjOGpfH4g178P6bLCIiMvgEoxj4GCAhCPeRU7Skqp7CIcmMyE6JdCgiIiISGeXAnc65Y6pf+/b/B1gQiaAiLjYOPvoAJA+BP18DLfvD9ujy0jx2N7axfufBsD1TREQkmgQ0dc7M/r2XwwnABODfgEeDGZScXHe3Y+nmes6bWIA3g1FEREQGIQfEHqctxtc+OKXlwxW/hd98EB77DFz9MMSEfsHls0vzAFi8YQ8Th2eE/HkiIiLRJqBEE70X+24DaoCf4f3GTMJo3c5GGlo6VJ9JRERkcFsMfMPMVjjn3jl80MxG4tVpeuG4Vw4GxbPhgv+FZ26DV74HZ3855I/MT09iSmEmFRv2cPOCsSF/noiISLQJtBh46H/9I6ek0lefaa7qM4mIiAxmXwBeBDaZ2et4KwLnA6cD24EvRTC26DDrRqhZCYu/DcNnwrhzQ/7IBaV5/HRxFQ3NHWSmxIf8eSIiItFECaR+aklVPWPz0yjISIp0KCIiIhIhzrmtwHjg88BaIB5Yh7dQy1xgRMSCixZm8KF7oWAS/PUG2L815I88uzSfbgcvb9ob8meJiIhEm4ASTWb2LTP75XHafmFm3whuWHIi7Z3drNi6T6vNiYiICM65dufcL5xzNzjnPuh7XQiciTe17oTM7AIz22BmVWZ2+wnO+6iZOTMrC2b8YZGQAlf+DpyDRz4BHS0hfdz04iEMSYln8YY9IX2OiIhINAp0RNPVwCvHaXsF+HhwwpFAvFlzgOb2LuaVqD6TiIiIvHdmFgvcB1wITASuNrOJvZyXjjdqanl4Iwyi7DFw2ULY+SY8fZuXdAqR2Bjj7NPyeGnDXrq7B289dhERGZwCTTQNB2qP07bD1y5hUllVjxmcMSY70qGIiIhI/zYbqHLObXbOtQMPA5f0ct43gHuA1nAGF3SlF8BZX4Y3fg+vPRjSRy0ozaf+UDtrahtC+hwREZFoE2iiaRcw8zhtMwFNQA+jJdV1TB6eyZCUhEiHIiIiIv1bIV7R8MNqfMeOMLMZQLFz7qlwBhYy5bdDyTnw9y9DzWshe8xZp+VhhqbPiYjIoBPQqnPAI8CdZva2c+7pwwfN7IPA14GFoQhO3q2lvYtV2/bzqfmjIx2KiIiI9H/Wy7Ejc73MLAa4F7jupDcyuwm4CaCgoICKiorgRNhDU1NTn+8dN/Q6Tq9dg/3uCl47/Qd0JGQGJ7gexmTEsGhFNdPjdoTk/tEmGH0jwad+iV7qm+ikfum7QBNNdwLTgSfNrB7YCQwDsoF/4CWbJAxWbN1HR5dj3ljVZxIRERmMzGwvfsmgE0gM4JwaoNhvvwivLMJh6cBkoMLMAIYCi8zsYufcSv8b+QqQLwQoKytz5eXlATz+1FVUVBCUe08aAfefx/yd98O1j0NMbN/v2cPqrk3c+8+NTC6bS25aIN3RvwWtbySo1C/RS30TndQvfRdQosk51wqcZ2bnAwuAHKAeeME593wI45MeKqvriY81Zo3KinQoIiIiEhn3EViiKRArgHFmNhqvHudV+C3y4pxrAI78dsvMKoDbeiaZ+qXh0+FDP4C/3QwvfhPO/e+gP2JBaT4/eH4jL2/cy2Uzi4J+fxERkWgU6IgmAJxzzwHPhSgWCUBldR0zirNISTilrhMREZEBwjl3VxDv1Wlmt+D9/y4WeMA5t9bM7gZWOucWBetZUWnGNVCzAv71Ayg8HSZ8KKi3nzQ8g9y0RBZvUKJJREQGj4CyFWZ2FV4RyO/20nYbsM0590iwg5NjNTR38FZtA7e+f1ykQxEREZEBwjn3DPBMj2N3Hufc8nDEFFYX3gM7V8MTn4O88ZA7Nmi3jokxykvzeH7dbjq7uomLDXQdHhERkf4r0H/tbuf4y9k2A3cEJxw5kWVb6ul2MF/1mURERESCIy4RrvgtxMTBn6+Btqag3n5BaT4NLR28sf1AUO8rIiISrQJNNI0D3jpO23pfu4TY0up6kuNjmV48JNKhiIiIiAwcQ4rhow9A3QZYdCu4YJXAgveNyyU2xli8YU/Q7ikiIhLNAk00NeOtQtKbYqAtOOHIiSypqmPW6GwS4jTsWkRERCSoShbA+78Gax+D5b8I2m0zk+M5fWQWi9/eG7R7ioiIRLNAMxb/BL5uZvn+B80sD/gq8I9gBybH2nOwlU17mphXkhPpUEREREQGpvlfhNKL4B9fg3cqg3bbBaX5rNvZyO7G41WiEBERGTgCTTT9F5AGVJvZo2b2YzN7FKgGUoAvhypA8SytrgdgfonqM4mIiIiEREwMXPpzGDISHr0ODu4Kym0XjM8D4KUNGtUkIiIDX0CJJufcNmAa8FO8qXIX+l5/AkwHgvOvsBxXZVU9GUlxTByeEelQRERERAaupEy48vfQdtBLNnV19PmWpQXpDMtMUp0mEREZFAIu9uOc2+ucu8M5d4ZzbhwwD3gB+A6nkGgyswvMbIOZVZnZ7b2032tmb/i2jWZ2wK+ty69tUaDPHAiWVNcxtySH2BiLdCgiIiIiA1vBRLj4J7BtKTx/Z59vZ2aUl+bzyqY6Orq6gxCgiIhI9DrlqtJmNsfMfgjUAM8DHwEeDvDaWOA+vBFRE4GrzWyi/znOuS8656Y756bjjZh6zK+55XCbc+7iU429v9q+r5ma/S3M07Q5ERERkfCY8lGY8zlY9jNY85c+325BaR5NbZ2s3Lo/CMGJiIhEr4ASTWY22cy+ZWbVQCXwGaAA+BIwzDl3c4DPmw1UOec2O+fa8RJUl5zg/KuBPwV47wFrSVUdAPPHqhC4iIiISNic9w0oPgMW3Qq71/XpVvPH5hIfa1Ro+pyIiAxwx000mdkYM/uKma0B3gRuA9YDnwDGAQascs51nsLzCoHtfvs1vmO9PX8kMBp40e9wkpmtNLNlZvaRU3huv1ZZXU9+eiIleWmRDkVERERk8IiNhysegsR0+PM10Nrwnm+VmhjHnNE5qtMkIiIDXtwJ2qoAByzHG8H0V+fcfgAzy3yPz+utwJA7zrlXAX9xznX5HRvhnNthZmOAF81sjXOu+pgHmN0E3ARQUFBARUXFewz15JqamkJ6fwDnHBXrW5iUE8NLL70U0mcNFOHoFzl16pfopb6JTuoXkSiRPhQ+9iA8+CF4/HNeofCYU64+AUB5aR7ffHo9NfubKcpKCW6cIiIiUeJEiaZ3gJHAZKAc2Glmz53iCKaeavBWqzusCNhxnHOvAo6Zkuec2+F73WxmFcAMoLrHOQuBhQBlZWWuvLy8D+GeWEVFBaG8P8CGXQdpfO5lLp03ifJZxSe/QMLSL3Lq1C/RS30TndQvIlFk5Dw475vw3B2w5Idw5pfe020WjM/nm0+vp2LDXq45Y2SQgxQREYkOx/11jHNuNDAfeAg4B3gS2G1mv/LtH28k0omsAMaZ2WgzS8BLJr1r9TgzKwWygKV+x7LMLNH3PtcXW98my/cDldVefaZ5qs8kIiIiEjlnfA4mXw4vfgOqF7+nW4zJTWVEdorqNImIyIB2wnG/zrmlzrlb8eoonQ/8DbgcOLz0xqfNrCzQh/lGQ90CPIdX7+kR59xaM7vbzPxXkbsaeNg555/MmgCsNLM3gcXAd5xzAz7RtKSqnhHZKRpeLSIiIhJJZvDhH0NuKfz1Bjiw/eTXvOsWxoLSPJZU1dPa0XXyC0RERPqhgCaYO+e6nXPPO+c+BQwFLgMeBS4FlpvZ+kAf6Jx7xjl3mnOuxDn3Ld+xO51zi/zOucs5d3uP6yqdc1Occ9N8r/cH+sz+qrOrm+Wb67XanIiIiEg0SEzzajR1tsMjn4DOtlO+Rfn4fFo6unh1y74QBCgiIhJ5p1zJ0DnX7px7wjl3FVCAtwpdVdAjE97a0cjBtk7mluRGOhQRERERAcgdC5f+HHa8Dn//8ilfPndMDolxMVp9TkREBqz3tmSGj3PukHPuD865DwcrIDnqSH2mEo1oEhEREYkaEz4M7/sivPYgvP67U7o0KT6WeSU5VGzYG5rYREREIqxPiSYJrcqqesYPTSc3LTHSoYiIiIiIvwVfg9Fnw9P/D3a8cWqXjs9nS90httQdClFwIiIikaNEU5Rq6+xixdZ9zNVoJhEREZHoExsHH30AUnPhkWuhOfCaS+Wn5QNo9TkRERmQlGiKUq+/c4C2zm7mqz6TiIiISHRKzYUrfgcHd8Ffb4TuwFaSG5GTQkleKos1fU5ERAYgJZqi1NLqOmIMZo/JjnQoIiIiInI8RafDhfdA9Qvw0v8FfFl5aT7LNtfT3N4ZwuBERETCT4mmKLWkup6pRUPISIqPdCgiIiIiciKnXwfTr/ESTRueDeiSCyYPpb2zm689/hbd3S608YmIiISREk1RqKmtkze3H9BqcyIiIiL9gRlc9D0YOhUevwn2bT7pJbNGZXPbeafx2Kpa7n5qHc4p2SQiIgODEk1RaMWWfXR2O+aPVX0mERERkX4hPhmu/B1g8OdPQHvzSS+5ecFYbnzfaB6s3MqPXtgU+hhFRETCQImmKFRZXUdCXAynj8yKdCgiIiIywJnZBWa2wcyqzOz2Xto/a2ZrzOwNM/uXmU2MRJz9QtYouPzXsPsteOoLcJJRSmbGVy+awEdPL+KH/9zEg0u2hCdOERGREFKiKQotqarn9BFZJMXHRjoUERERGcDMLBa4D7gQmAhc3Usi6Y/OuSnOuenAPcAPwhxm/zLuA1B+B6z+M6z49UlPNzO+c9kUzptYwF1PruPxVTVhCFJERCR0lGiKMvsPtbNuZ6PqM4mIiEg4zAaqnHObnXPtwMPAJf4nOOca/XZTARUTOpmz/hPGnQ/P3gHbXz3p6XGxMfz46hnMK8nhtkdX8891u8MQpIiISGgo0RRllm6uB2Ce6jOJiIhI6BUC2/32a3zHjmFmN5tZNd6Ips+HKbb+KyYGLvslZBbCI5+Apj0nvSQpPpaFnyhj8vAM/v2Pr7O0uj4MgYqIiARfXKQDkGNVVteRmhDL1KLMSIciIiIiA5/1cuxdI5acc/cB95nZx4GvAZ98143MbgJuAigoKKCioiK4kfo0NTWF7N7BllryBWa+/mUaf30pq6fejYs5eVmET5/m+PYBx/UPLOP22UmMyuw/pRT6U98MJuqX6KW+iU7ql75ToinKVFbVM2dMDvGxGmwmIiIiIVcDFPvtFwE7TnD+w8DPe2twzi0EFgKUlZW58vLyIIV4rIqKCkJ17+Arh6Iksp74LGd3VsB53wjoqrI5rVz+80p+vLqLRz4zh7H5aSGNMlj6V98MHuqX6KW+iU7ql75TNiOK7GxoYXPdIdVnEhERkXBZAYwzs9FmlgBcBSzyP8HMxvntXgRsCmN8/d/0q2HWjVD5Y1j3t4AuGZqZxO9vnEOMwSfuX07tgZYQBykiIhI8SjRFkcoqX32mEtVnEhERkdBzznUCtwDPAeuBR5xza83sbjO72HfaLWa21szeAL5EL9Pm5CTO/18omgVP/Dvs3RDQJaNzU3noU7M52NbJtb9eTl1TW4iDFBERCQ4lmqJIZXU92akJjB+aHulQREREZJBwzj3jnDvNOVfinPuW79idzrlFvvf/4Zyb5Jyb7pxb4JxbG9mI+6G4BPjYQxCXBH++BtoOBnTZpOGZPHDdLHY0tPDJB16lsbUjxIGKiIj0nRJNUcI5R2V1HXPH5BAT01tdThERERHptzIL4WO/gfoq+NvN4N5Vc71Xs0Zl8/NrTmfDroPc+NBKWju6QhyoiIhI3yjRFCW21jezs6GVuarPJCIiIjIwjT4Lzr3Lq9W09KcBX7agNJ8fXDmdFVv3cfMfXqejqztkIYqIiPSVEk1RYklVHQDzx6o+k4iIiMiANe/zMOFieP6/YcsrAV928bThfOOSybzw9h7+89E36e4ObESUiIhIuCnRFCWWVtczLDOJUTkpkQ5FRERERELFDD7yM8gpgUevg2W/gIbagC695oyR/Of5pTzxxg7ufmodLsDpdyIiIuGkRFMU6O726jPNK8nFTPWZRERERAa0xHS48g+QPhSe/S+4dyL86hxY8iPYt/mEl/57eQmfPnM0D1Zu5Yf/3BSmgEVERAIXF+kABNbvamR/cwfzVJ9JREREZHDIOw0+twTqNsH6RbBuETx/p7cVTIGJF3tT7PLHH3OZmfGVD06goaWDH72wiczkeD71vtER+hAiIiLvpkRTFFhaXQ/AvLFKNImIiIgMKrnj4Mz/523734H1T3qJp8Xf8rbc07yE08SLYehUMMPM+PalU2hs6eTup9aRmRzP5acXRfqTiIiIAEo0RYUlVXWMyU1lWGZypEMRERERkUjJGgnzbvG2xp3w9lPeCnX/+gG88j3IGgUTPgwTLiGu8HR+dPV0PvXgCr7819WkJ8Vx3qShkf4EIiIi4a/RZGYXmNkGM6sys9t7ab/XzN7wbRvN7IBf2yfNbJNv+2R4Iw+Njq5uXt2yT6OZREREROSojGEw+9Nw3VNw2ya4+CeQM84rHn7/uXDvJBL/cQe/PquVqcPTuOVPq6isrot01CIiIuEd0WRmscB9wAeAGmCFmS1yzq07fI5z7ot+598KzPC9zwb+GygDHPCa79r9YfwIQbe65gCH2ruYV5Ib6VBEREREJBql5sLMT3hbywHY+Jw3ve71h0h+9Zf8NSWXvyfP5MGH3iTtxhuYOiIv0hGLiMggFu6pc7OBKufcZgAzexi4BFh3nPOvxksuAZwPPO+c2+e79nngAuBPIY04xCqrvPpMc8doRJOIiIiInETyEJh2pbe1NUHV88SsW8SFG5/joph/0PjAj2ksvZCMGZdDyfshPinSEYuIyCAT7kRTIbDdb78GmNPbiWY2EhgNvHiCawt7ue4m4CaAgoICKioq+hz08TQ1NfX5/k+/1sKI9BjeXFEZnKAkKP0iwad+iV7qm+ikfhGRk0pMg0mXwqRLieloZfcbz/DqMw9x9oa/w4a/QEIajDvPKyQ+9gPe+SIiIiEW7kST9XLMHefcq4C/OOe6TuVa59xCYCFAWVmZKy8vfw9hBqaiooK+3L+1o4vNz/+DT84bSXn5xOAFNsj1tV8kNNQv0Ut9E53ULyJySuKTKJh1GWOLzqX8l69wTtJG7j6tmuSqv8PaxyAuCcae661gd9r53sgoERGREAh3oqkGKPbbLwJ2HOfcq4Cbe1xb3uPaiiDGFnYrt+6nvatb9ZnjfSObAAAgAElEQVREREREJCgmDMtg4XVzueb+GNZuL+PhW75Hxp6VsG4RrH/SW8kuJh7GlHsr2I2/yKsBJSIiEiThXnVuBTDOzEabWQJeMmlRz5PMrBTIApb6HX4OOM/MsswsCzjPd6zfqqyuIy7GmDU6O9KhiIiIiMgAUTYqm19cczqb9hzkxt+uomX4XPjgPfDFtXDDP+GMz0LdRnjy8/C9cfDgh+DVX0HjzkiHLiIiA0BYE03OuU7gFrwE0XrgEefcWjO728wu9jv1auBh55zzu3Yf8A28ZNUK4O7DhcH7qyXV9UwrHkJaYrgHlomIiIjIQFZems8PrpjOinf2cfMfX6ejqxtiYqB4Fpz3TfiPN+EzL8P7vgRNu+GZ2+AHE+D+86Dyp7D/nUh/BBER6afCnuFwzj0DPNPj2J099u86zrUPAA+ELLgwamztYE3NAW5ZMDbSoYiIiIjIAPThacNpbO3gq4+/xW2Pvsm9V0wnJsZX9tQMhk3ztnO+Dns3+KbX/Q3+8VVvGzbdKyQ+4RLI1f9ZRUQkMBpKEyHLN++j28G8sZoTLyIiIiKh8W9zRtLQ0sE9z24gMzme/7l4Ema9rLGTVwpn/6e37dvs1XNatwheuNvb8ib4kk4XQ8EkL1ElIiLSCyWaIqSyuo6k+BhmjNCKHyIiIiISOp87u4QDzR0sfHkzQ5Lj+dJ5pSe+IHsMzP8Pb2uogfVPeYmnl78LL/2f1z7hYi/xNHymkk4ig41zXkJ6+6tQswJqV0J3F2QMh4xCyCyEjCJvP9P3Gp8c6agljJRoipDKqnpmjcomMS420qGIiIiIyABmZtxx4Xgamjv48YtVZCTHc+OZYwK7OLPIKx5+xmehaQ+8/TSsXwRLfwpLfgiZxd7qdRMuxrq7QvtBRCQyWhuh9jWoWQk1vuRSy36vLSEdCmdCfAo01nrnNde/+x4pOb5EVJEvEVXol5Qq9NriEsP7uSRklGiKgL0H29iw+yCXzBge6VBEREREZBAwM7592RQaWzv45tPryUyO52Nlxad2k7R8KLve25r3wcZnvel1K+6HZT/jfTFJsH02FJ8BI+ZA0SxIygzNBxKR0Oju9lalPJxQqlkJe9YDvnW68sbD+IugaLb3PZ5XCjE9Bk90tEDjDi/x1FALjTW+1x3QsB22LYXWA+9+dmqeL/lU5Dc6quhoIipjOMTGh/yPQPpOiaYIWLrZy/DOL1F9JhEREREJj9gY44dXTefggyv5r7+uJiM5nvMnDX1vN0vJhukf97a2g1D1T3ZWPkpRaw288j1w3YBB/kQv6VTs27JGaapduDgH+7eQeWAtbE/1fkCPS4TYBG+LS/SOxSZ673smC2RwaN7nG620wre9Bm0NXltSppdMmvgRKCqDwtMhOYDSL/HJkFPibcfTfsiXeKrxS0j5tvpq2PIytDX2uMggrcA3Cuo4o6PShkKs0hyRph6IgKXVdaQnxTG5UL/hEREREZHwSYyL5ZfXns6//Xo5t/5xFQ9eP6vvi9MkpsOkS6nam0VReTm0NXk1W7a/CtuWwZq/wErfwtFpBVDsG/VUPMdb9S4uoc+fa9Dr6vRGoex8E3athp2rYdcaaGtgBsAbAdzDYrykU2yC1yexiSdJTvmOHfe8k1xz0uvjj8YTExPiP8BBorsL9qw7OlJp+6tQv8lrsxgvMTz5Mi+5VDwbsktC92efkAq547zteFobfSOj/EZEHX6/dwNUvQgdh469xmIgfVjvI6IOv0/LV2I1xJRoioAlVfWcMSaH2Bj9NkdEREQiy8wuAH4ExAK/ds59p0f7l4AbgU5gL/Ap59w7YQ9UgiY1MY4Hr5/FFb9cyqd/u5I/fvoMphUHcYGaxDQYU+5t4P1wu/dtL+m0fbm3rX/Sa4tL8gqKF8+GEb7kU0p28GIZiDpaYPc62PWml1Da+aaXPOhs9drjkr2VAad8FIZN5c13DjBtymToaoOuduhs7/Het3W29Xjf4Z3n/76rA9qboWu/9/5413R3Bvczx8R5Saf4ZG/UypARkDnCex0yAoYUe6+aqnmsQ3VeUulI0e7XjyZmUnK86W/TrvK+/4bP8JLG0SQpw9vyx/fe7hy0Nrx7RNTh6Xq734KNz0Fny7HXxcT5klGFvY6Oim9v8O6t0ZfvmRJNYbZ9XzPb9jVz/fxRkQ5FREREBjkziwXuAz4A1AArzGyRc26d32mrgDLnXLOZfQ64B7gy/NFKMA1JSeB3N8zh8p9Xct1vXuWRz8xlXEGIfsiMifUSHwWTYNYN3rGDu7wffrcv9xJQS+/ziosD5IzzTbfzJZ5yxw3eH/haG7yRSTt9SaVdq72RHM5XeD0pE4ZOhVk3eqPDhk6FnLHHTB3af7ACxpWHN+7ubl8Cqs0vmdV2bGKr1+RWe48EWI9r2pt8o1k2wqZ/vjuBkJTZewIq0/eanDVwv5a6OrzEyvYVR6fB7d/itcXEQcFkb6pr8WxvGlzW6P7/Z2HmTeVLHuL9/dIb57zC5Q01PUZH+UZI1b7urazZ1XbkkvkAr2Uenf6XM9Yb3XV4XwnNk1KiKcyWVvvqM/V1iLKIiIhI380GqpxzmwHM7GHgEuBIosk5t9jv/GXANWGNUEKmICOJP9w4h4/+YinX3v8qj352LsXZKeF5ePpQmHixt4E3SmfHKt+op1fh7Wdg1e+9tuRsX40n36in4TMG5lLpB3f7pr29eXQK3P6tR9vTh3mJpPEXea/DpnnJk2hMFsTEQEwSxCeF7hnOeaubHXgHDmyHA9u8rWG7l2DZ8pKXmPKXkH40+eSfgDq8peRE559nbw4naw9Pg9ux6mjiLW0oFM/yCvcXzYJh0yEhTN/b0cbMGyWZkg3DpvZ+jnPe6C/fiKiqFS8wNtugvgq2Lfem/x4uhg5e0fJsXwIqZ4zvdSxkjxmYfze9B0o0hVlldR25aYmMy0+LdCgiIiIihcB2v/0aYM4Jzr8B+HtII5KwGpmTym8/NZsrf7mUa+9fzqOfnUdeegSWGI9PhpHzvA28H/zqNvmm2i3zftjb6PvSi4n3kiwjzjha7ym9IPwxv1fOecmRw9PeDtdUatp19Jys0d5nnPkJGDrN+wE5LT9yMUcjM0jN9bbC09/dfngki38C6sC2o0mpd5YeLXp9WHxKj+TT4fcjveNp+ZFJRHW2eV8jNX6jlRp8f3XHJnhfK2XXeyOVimZ7tYj6S8IsGphBWp63DZ9Oza5UxpaXH23vaPWSl/VVXqHy+irYtxmqnoc3dh97r4yio8mnI8moEu9raBDVo1OiKYyccyyprmdeSQ6mb3wRERGJvN7+Q+J6OYaZXQOUAWcfp/0m4CaAgoICKioqghTisZqamkJ278Hs1mlxfHdlM5f9+EVun51Eavyp/181NH1T7P2AP+VjxLc3ktH4NpkN68lofJv05QuJXfpTAFqShtKQOZ7GjPE0ZE7gUGoxWOSL/Vp3F8kttaQf3Exa02bSmqpJa9pCfKdXJ8cRw6HUYprSJtBUcBEH08fQlDaarrhU7wZdQC1Quw6/gYanRN8zABnAJEieBMnAMO9oXEcTiW17SWrdc+y2azNJW5cR33nwmLt0xSTQlphHa1IerUkFvtf8I1t7QpZXjDpAvfaNcyS21ZHRuIGMxrfJaNxI+sFqYpxX96o1MY/GjNNoLDmPxoxSDqaPwcXEe9fWAXXVQPV7+UMSn+N/z6QD0yFzOmQCoyG2s5nklp0kt+wgpXmH91q3g+TtrxPfeXREnSOG1qR8mlMKaUkeRnPKcFqSva01KTcq/r4KJiWawqh6bxN7D7YxryQn0qGIiIiIgDeCqdhvvwjY0fMkMzsX+CpwtnOurWc7gHNuIbAQoKyszJX7/zY4iCoqKgjVvQezcuC0SXu58aEVPFidyG8/NYfkhFP7wSc8fXPx0bed7d6IoO3LSd6+jORtyxm6u8JrS8zwrZw1x6v3VFjmFSkPpY5Wryi3/yil3WuPTmeKS/LqyJRc4Y1AGTYVy59IWnwyoYxM3zN90Hbw6Aiohu3EHniHlAPbSDmwHQ68Bjvrjj0/NsEbTZTpNxLKf6pe+rBjVjurqKigfP4c2PEG1PhNgzu40zshLsmbKjr5fO/ruWgWSRnDSAI0vi10gvY907zvyAgo21dNcn0VyfVVsGfxsavlxSZ40+7860AdHhGVPrRfjk5ToimMllSpPpOIiIhElRXAODMbjTdu4irg4/4nmNkM4JfABc65PeEPUcLl7NPyuPfK6dz6p1V87g+vsfDaMhLionhZ+bgErw5N8SzgFm+q1P6tR1e227YcKv4XcN4ok4LJR1e2K57jJQDeq9ZGr0j34YTSzjehbsPR1dYSM73pbmWf8l6HTfOKnMfqx69+JTEdCiZ6W2/aD3lFpg9sO7ZWVMN22PQPaOoxrSomzlvpbMgISB/KzK2r4OWtR79uskbBqDN9SdJZ3tdsbHwoP6GE0uHaUMWzjj3unFdja1+133S8am+/6nmv8P1hCWleEuqYouS+6XhRvEKn/qYLoyVVdRRlJYevyKKIiIjICTjnOs3sFuA5IBZ4wDm31szuBlY65xYB3wXSgEd9U/+3OecuPu5NpV/70NThNLZ08pXH1/D/Hn2TH145ndiYfvLbdDPIHu1t067yjrUcgNqVXtJp+zJY9Qd4daHXllF4NOk0Yg4UTOk9EdS0F3a9eezKb/s2H21PK/CKc5de6CWVhk71Egb9cBSCnKKEVMgr9bbedLT6ElHvvLtO1LbldMVmwrzPe7XGCsu8GkEy8JlBxjBvG/W+Y9u6u7yvmcN1oA4nona8AesWHV1xErxVFP0TTzklR0dFJYZoFdEAKdEUJl3djmWb67lw8rBIhyIiIiJyhHPuGeCZHsfu9Ht/btiDkoj6+JwRNLR08H/Pvk1mchzfuGRy/60vmjwExp7rbQBdnb4l4P1GPa19zGuLT/GKSo84A7CjK8AdnsYEXgJp6FRvmfhh0733/akQuYRXfBLkjvW2XrypaY3SU0wsZI30Ns45tq2z3UtaHilI7nvd+i9Y/fCx56YVHF0Jr/wOyCwM20cAJZrCZu2OBhpbO5k3VvWZRERERCS6fa68hAMt7fzypc0MSU7gtvOPM2Kjv4mNg+HTvW3OZ7xjDTVHk07bl8ErPwAc5JbC6LOPjlIaOsVLXImIREJcAuSO87ae2pt7rIznm4q38Vl4/9fCH2rYnzhIVVZ79ZnmqhC4iIiIiPQDt18wnobmDn66uIohKfHceOaYSIcUGplF3jb5cm+//ZBX0yk+ObJxiYgEKiHFW2ygYFKkIwGUaAqbJVV1jMtPIz89KdKhiIiIiIiclJnxrUun0NjawTefXk9GcjxXlPWhgHZ/kZAa6QhERPq1KF5GYuBo7+xmxdZ9Wm1ORERERPqV2Bjj3iunc+a4XG7/62qefWtXpEMSEZEop0RTGKzatp/Wjm5NmxMRERGRficxLpZfXHM604qH8Pk/rWJJVV2kQxIRkSimRFMYVFbXE2NwxhglmkRERESk/0lNjOM3181idG4qn/7tSlZt2x/pkEREJEop0RQGldV1TC7MJDM5PtKhiIiIiIi8J0NSEvjdDbPJTUvk+gdXsHH3wUiHJCIiUUiJphBrbu9k1bYDzCtRfSYRERER6d/yM5L4/Q1zSIiN4dr7l7N9X3OkQxIRkSijRFOIvbplH53djnmqzyQiIiIiA8CInBR+e8NsWtq7uPb+5ew52BrpkEREJIoo0RRiS6vriY81Zo3KjnQoIiIiIiJBMX5oBr+5fja7G9v4xP2v0tDSEemQREQkSoQ90WRmF5jZBjOrMrPbj3POFWa2zszWmtkf/Y53mdkbvm1R+KJ+75ZU1zFjRBbJCbGRDkVEREREJGhOH5nFL689neq9TXzqwRU0d7hIhyQiIlEgrIkmM4sF7gMuBCYCV5vZxB7njAPuAOY75yYBX/BrbnHOTfdtF4cr7vfqQHM7a3c0Ml/1mURERERkADrrtDx+dNUMVm3bzxcWN3Prn1ax+O09dHZ1Rzo0ERGJkLgwP282UOWc2wxgZg8DlwDr/M75NHCfc24/gHNuT5hjDJplm+txDuaNVX0mERERERmYPjhlGE/cPJ97/7aclzfu5ck3d5CblsDF0wq5bGYhk4ZnYGaRDlNERMIk3ImmQmC7334NMKfHOacBmNkSIBa4yzn3rK8tycxWAp3Ad5xzT4Q43j6prK4nJSGWaUVDIh2KiIiIiEjITC0awicnJfKL953J4rf38viqGn63bCsPLNnCuPw0Lp1ZyEemFzJ8SHKkQxURkRALd6Kpt19l9JzMHQeMA8qBIuAVM5vsnDsAjHDO7TCzMcCLZrbGOVd9zAPMbgJuAigoKKCioiLIH+GopqamE97/+dXNlGTEUPmvl0MWg7zbyfpFIkP9Er3UN9FJ/SIi/VFiXCwXTB7KBZOHcqC5nadW7+TxVbXc8+wGvvvcBuaMzuayGUVcOGUo6UnxkQ5XRERCINyJphqg2G+/CNjRyznLnHMdwBYz24CXeFrhnNsB4JzbbGYVwAzgmESTc24hsBCgrKzMlZeXh+BjeCoqKjje/Xc3trLz2Re4/uxxlJ9VErIY5N1O1C8SOeqX6KW+iU7qFxHp74akJHDNGSO55oyRvFN/iCdW7eDxVTV8+a+r+frf3uIDEwu4bGYhZ47LIz5Wi2GLiAwU4U40rQDGmdlooBa4Cvh4j3OeAK4GHjSzXLypdJvNLAtods61+Y7PB+4JX+inprK6DoB5KgQuIiIiIoPcyJxU/uPccXz+nLGs2n6Ax1+v5cnVO3hq9U5yUhP48LThXDazkCmFmarnJCLSz4U10eSc6zSzW4Dn8OovPeCcW2tmdwMrnXOLfG3nmdk6oAv4T+dcvZnNA35pZt14q+V9xzm37jiPirjKqnqGpMQzcVhGpEMREREREYkKZsbMEVnMHJHF1z80kYoNe3h8VS1/XL6NByu3UpKXymUzi7hk+nCKslIiHa6IiLwH4R7RhHPuGeCZHsfu9HvvgC/5Nv9zKoEp4Yixr5xzVFbXM3dMDjEx+o2MiIiIiEhPCXExnDdpKOdNGkpDcwdPr9nJ46tq+O5zfvWcZhZy4ZRhZKiek4hIvxH2RNNgsG1fM7UHWvjs2WMiHYqIiIiISNTLTInn43NG8PE5I9i+r5nHV9Xy+Kpa/uuva7jzb2s5d2IBl80o5KzTVM9JRCTaKdEUAkuq6gGYN1b1mURERERETkVxdgqfP2cct75/LG9sP8ATq2p5cvVOnvar53TpjEKmFqmek4hINFKiKQQqq+soyEhkTG5qpEMREREREemXzIwZI7KYMSKLr31oIi9t2OvVc3rVq+c0Ji+Vy2YUcsn0QoqzVc9JRCRaKNEUZN3djqXV9Zx9Wp5+wyIiIiIiEgTxsTGcO7GAcycW0NDSwd/X7OSxVbV87x8b+d4/NjJ7dDaXzfDqOWUmq56TiEgkKdEUZBv3HKT+UDtzS3IiHYqIiIiIyICTmRzPVbNHcNVsr57T396o5bFVtdz+2BruXLSWD0wo4NIZhZxdqnpOIiKRoERTkKk+k4iIiPQnZnYB8CMgFvi1c+47PdrPAn4ITAWucs79JfxRivSuODuFW94/jpsXjGV1TQOPr6pl0Zs7eHrNTrJTE/jw1GFcOrOIaarnJCISNko0BVllVR2jclIoHJIc6VBERERETsjMYoH7gA8ANcAKM1vknFvnd9o24DrgtvBHKBIYM2Na8RCmFQ/hqxdN4OWNe3lsVS0Pr9jOQ0vfYUxuKpfOKOQjM1TPSUQk1JRoCqLOrm6Wb9nHxdOHRzoUERERkUDMBqqcc5sBzOxh4BLgSKLJObfV19YdiQBFTlV8bAznTCjgnAkFNLZ28OyaXTy2qobvP7+R7z+/kVmjsrh0RhEXTRlGZorqOYmIBJsSTUG0uraBprZO5qk+k4iIiPQPhcB2v/0aYM57uZGZ3QTcBFBQUEBFRUWfg+tNU1NTyO4tfROtfZMPfPY0+FhxMkt3dFK54wBfeXw/dz6xhun5scwbHsfUvFjiYgbm1Lpo7RdR30Qr9UvfKdEUREurvfpMc8co0SQiIiL9Qm8/Wbv3ciPn3EJgIUBZWZkrLy/vQ1jHV1FRQajuLX3TH/rmcsA5x1u1jTy2qoYn39zBylVtZKXE86Gpw7l0ZiEziocMqHpO/aFfBiv1TXRSv/SdEk1BtKSqjvFD08lJS4x0KCIiIiKBqAGK/faLgB0RikUkLMyMKUWZTCnK5CsfnMC/NtXx2KpaHlm5nd8te4fRuamcMz6fqcVDmFqYyciclAGVeBIRCTUlmoKktaOLle/s59ozRkY6FBEREZFArQDGmdlooBa4Cvh4ZEMSCZ/42BgWjM9nwfh8DrZ28Pe3dvHEqlp+u+wd2v+1BYCMpDgmF3qJqamFQ5hSmElxdrKSTyIix6FEU5C8/s5+2ju7VZ9JRERE+g3nXKeZ3QI8B8QCDzjn1prZ3cBK59wiM5sFPA5kAR82s/9xzk2KYNgiIZGeFM8VZcVcUVZMR1c3G3cf5K3aBlbXNLCmtoHf/Gsr7V1eTfzM5HimFmUypdC3FWVSOETJJxERUKIpaCqr64mNMWaPzo50KCIiIiIBc849AzzT49idfu9X4E2pExk04mNjmDQ8k0nDM7lylnesvdNLPnmJpwOsqW1g4cub6ez2ypplpyYck3iaWpTJ0IwkJZ9EZNBRoilIllTXMbUok/QkLZEqIiIiIjLQJMTFMLkwk8mFmcAIwCufsWHXQVbXNvBWTQOraxv4+UvVdPmST7lpvuRTkVfvaUpRJgUZSRH8FCIioadEUxAcbO1gdU0Dnzu7JNKhiIiIiIhImCTFxzKteAjTioccOdba0cW6nY1Hp93VNPDSxk34ck/kpycytchLWHnT74aQl67FhERk4FCiKQhe3bKPrm6n+kwiIiIiIoNcUnwsM0dkMXNE1pFjze2drN/ZeCTxtLq2gRfe3oPzJZ+GZSZ5iSffqKcphZlayVpE+i0lmoKgsrqehLgYZo7MOvnJIiIiIiIyqKQkxHH6yGxOH3m0nmtTWyfrdjSyusar97SmtoHn1+0+0l44JPlIvafDtZ+yUhMiEb6IyClRoikIllTVUTYyi6T42EiHIiIiIiIi/UBaYhyzR2cfs5hQY2sHa2t90+5qG1hTc4Bn1+460l6cnczUwiFHpt1NHp5JZopqxIpIdFGiqY/qm9p4e9dB/vP80kiHIiIiIiIi/VhGUjxzS3KY61eSo6Glg7VHEk8NrK49wNNrdh5pH5WTcky9p8mFGVqgSEQiSommPlq6uR7gmH8MREREREREgiEzOZ55Y3OZNzb3yLH9h9p5a8fRYuOrth3gqdVHk09jclOPTLlrr+9ixN4mhmUmk5ygGRgiEnpKNPVRZXU9aYlxTC3MjHQoIiIiIiIyCGSlJnDmuDzOHJd35Fh9U5tX66nGq/f06pZ9/O2NHQDcs+Il77qUeIZmJjM8M4mhmUkMH5LMsMPvM5MZmpmkciAi0mdKNPVRZVUdc0ZnExcbE+lQRERERERkkMpJS6S8NJ/y0vwjx/YebOMv//gXBaNL2dnQyo4DLexqaGVHQyuvb9vP/uaOd90nOzWBYZlJvi3Zl5Dy3h9OSiXGKRklIsenRFMf1Ld0s7W+hWvnjop0KCIiIiIiIsfIS09kQk4s5TOLem1vae9iZ8PR5NPOAy3sbPRea/a3sGLrfhpa3p2MyklNYJhf8unoqzdKKj8jUckokUFMiaY+WFffBcA81WcSEREREZF+JjkhljF5aYzJSzvuOc3tnexsaGXngVZ2NrR47xu899vqm1m+uZ7G1s53XZeblnhM8mmo3yipYZlJFGQkkRCnWSEiA1HYE01mdgHwIyAW+LVz7ju9nHMFcBfggDedcx/3Hf8k8DXfad90zj0UlqCPY/2+bnJSEygtSI9kGCIiIiIiIiGRkhBHSV4aJSdIRh1q6zySfPISUkeTUlvrD7G0up6Dbccmo8y8ZNThelHDMpP5/+3dbYxc51XA8f+ZmX3xOnaa1rEhtsEucVpCVJRiQqghclIoaSlUBYTaKnwARPlQaIIqVW0llFBIeEdFwo0aSov6ogQpTUMURWksikVb0pC0rlTiJK1x4tiNk9huHb/Fu57Zw4d7d3dmveuuvbN7b+z/TxrdeZ557r1n9GjGx2efe+eSVw333ENq1fJhBrxFifSKs6iFpohoAluAXwb2Ao9GxH2ZuaNrzAbgw8CmzPxBRKws+18N3AxspChAfaPc9weL+R4mZCY7DnbY9LqLaTSiihAkSZIkqXJLh1pcuvICLl05ezHqyImTPN+1Guq5QyfKS/ZeZtf+Y3xt50GOzlCMWrlsaLL4tGr5MMuGW4wMtrhgqMnIYIulQy2WDjWL7WD5vOx3xZRUjcVe0XQVsDMzdwFExF3AO4AdXWP+ANgyUUDKzBfL/l8Btmbm98t9twLXA3cuUuw9dh04xqHRZNNPrPjhgyVJkiTpPLZseIBlwwNsOM3VIIfLYlT3Tcv3HXqZ5w+f4DsvHOGr3z3A0bE2mXM750AzegpQRYGqxchgs9j2FKhaLB1sThWuJotYU/1LBpouMpDmYLELTauBPV3tvcDPTRtzGUBEfI3i8rpbMvPBWfZdvXChnt5/7zwAeH8mSZIkSeqH5cMDLB8e4LLTFKMykxMnxzk62ub4WLvcdortaIdjo22OjbXLbdme1n/g6GjPfmPt8TnFFwEjA01GhqYKVj0FqlNWWBXbyQLXULOn0DXaSTrjSdPilc4xi11omukTNL0e3QI2AJuBNcBXIuKKOe5LRLwXeC/AqlWr2LZt2zzCnd2/bz/BRUPJ09/+H54Jvxjq5OjRows27zp7zkt9OTf15LxIknSqiGDJYJMlg01gqC/HPNkZL4pUpxSoJopTnZ7+osDV4fhoUeg6cHSM3a+z6E8AAAlISURBVAePT409g1VXbH2AViMYbDUYbDUYKreDzQZDrWZP/9Dk8yaDzcap+7Sm9hlqNhgaaEyOm36syXMMNBhqTr1m0Uv9sNiFpr3A2q72GuC5GcZ8PTNPAk9HxFMUhae9FMWn7n23TT9BZt4B3AGwcePG3Lx58/Qh8zY+ntz0X1u5YsUA1157bd+Pr/nZtm0bCzHvmh/npb6cm3pyXiRJWhwDzQYXjjS4cGSgL8fLTF4+2Zm2kmqqkHV8tFhJ9cR3vsuaH1vPaLvDWHucsc44oyeL7Vh7nNF2h9F28fzIiTYHJ8ZMjG9PjBunPT7XytbpTRS9uotXcyl6NZu9BaruVve6jOh6pbe/V8yymGMuxzrt+bsaPbt0NfY8O8ajo0/SjKDRiMltI4Jmg3JbPCafT4wtX5/onxrLDGODRjDD2OnHZbJvtv7u49bBYheaHgU2RMR64HvAu4D3TBtzL/Bu4F8jYgXFpXS7gP8DbouIi8pxb6G4afii27HvMIeOn+TyS/tTQZckSZIknRsigpHB4pK5i5fN/n/Gbe3dbN68oS/n7IwnJ8tC1WinM1mA6i5GFcWszmQxa3TamIni1tkUvTo9Vx9OFb26V3Z1l8Ky64XpJbKefWYbN8txT7dPbywzxwjQ6YyTu3fR6VPxbrE1orcA9cD7f5F1K5YuagyRc17T16cTRrwN+BjF/Zc+lZm3RsRHgccy874oSox/T3Gj7w5wa2beVe77e8BHykPdmpmf/iHn2g/sXqC3ArACOLCAx9fZcV7qyXmpL+emnpyXufvxzLy46iA0ZYFzMD8b9eXc1JPzUl/OTT05L3Mza/616IWmc0lEPJaZG6uOQ72cl3pyXurLuakn50WamZ+N+nJu6sl5qS/npp6cl/lrVB2AJEmSJEmSzg0WmiRJkiRJktQXFprm546qA9CMnJd6cl7qy7mpJ+dFmpmfjfpyburJeakv56aenJd58h5NkiRJkiRJ6gtXNEmSJEmSJKkvLDSdhYi4PiKeioidEfGhquNRISLWRsR/RsQTEfF4RNxYdUyaEhHNiNgeEfdXHYumRMSrIuLuiHiy/Oz8fNUxCSLiT8rvsf+NiDsjYrjqmKQ6MAerH/Ov+jMHqx/zr/oyB+sPC01nKCKawBbgrcDlwLsj4vJqo1KpDXwgM38SuBp4n3NTKzcCT1QdhE7xj8CDmfl64KdxjioXEauB9wMbM/MKoAm8q9qopOqZg9WW+Vf9mYPVj/lXDZmD9Y+FpjN3FbAzM3dl5hhwF/COimMSkJn7MvOb5fMjFF/Yq6uNSgARsQb4VeCTVceiKRGxHLgG+BeAzBzLzEPVRqVSC1gSES1gBHiu4nikOjAHqyHzr3ozB6sf86/aMwfrAwtNZ241sKervRf/Ma2diFgHXAk8Um0kKn0M+CAwXnUg6vFaYD/w6XJJ/ScjYmnVQZ3vMvN7wN8BzwL7gJcy86Fqo5JqwRys5sy/askcrH7Mv2rKHKx/LDSduZihz5/uq5GIuAD4AnBTZh6uOp7zXUS8HXgxM79RdSw6RQt4I3B7Zl4JHAO850nFIuIiilUa64FLgKURcUO1UUm1YA5WY+Zf9WMOVlvmXzVlDtY/FprO3F5gbVd7DS6nq42IGKBIcj6fmfdUHY8A2AT8ekQ8Q3GZw3UR8blqQ1JpL7A3Myf+8nw3ReKjav0S8HRm7s/Mk8A9wJsqjkmqA3OwmjL/qi1zsHoy/6ovc7A+sdB05h4FNkTE+ogYpLg52H0VxyQgIoLiWucnMvMfqo5Hhcz8cGauycx1FJ+XL2emfxmogcx8HtgTEa8ru94M7KgwJBWeBa6OiJHye+3NeJNQCczBasn8q77MwerJ/KvWzMH6pFV1AK80mdmOiD8CvkRxF/pPZebjFYelwibgd4BvR8S3yr6PZOYDFcYk1d0fA58v/9O2C/jdiuM572XmIxFxN/BNil9z2g7cUW1UUvXMwWrL/Es6c+ZfNWQO1j+R6aXtkiRJkiRJmj8vnZMkSZIkSVJfWGiSJEmSJElSX1hokiRJkiRJUl9YaJIkSZIkSVJfWGiSJEmSJElSX1hoklSJiLglInKWxw0VxJPlz2ZLkiSds8zBJC20VtUBSDqvvQRcP0P/zsUORJIk6TxiDiZpwVhoklSldmZ+veogJEmSzjPmYJIWjJfOSaqliFhXLqV+T0R8NiKORMSLEXHzDGOvi4hHIuJERLwQER+PiAumjXlNRHwiIvaV456KiJumHaoZEbdFxP7yXFsiYmhB36gkSVKNmINJmi9XNEmqVESc8j2Ume2u5t8C9wO/BVwD3BwRBzJzS7n/5cCDwFbgN4G1wF8Br6VcEh4RS4BtwErgz4AngUvLR7cPAF8GbgDeAPwlsBv4m/m/U0mSpPowB5O0UCIzq45B0nkoIm4BTvnLWGl9uX0a2JqZb+na75+BtwFrM3M8Iu4CfgZ4fWZ2yjG/Dfwb8KbMfDgi/hC4HXhjZn5rlngS+EpmXtPVdy/wI5l59TzeqiRJUm2Yg0laaF46J6lKLwE/O8Pjua4xX5y2zz3AJcCasn0V8MWJBKf0BaAN/ELZvg7YPluC0+Whae0dXeeRJEk6V5iDSVowXjonqUrtzHxsphciYuLpi9Nemmj/KPBsuX2he0BmdiLiIPDqsus1wL45xHNoWnsMGJ7DfpIkSa8k5mCSFowrmiTV3cpZ2vu6tj1jIqJJkdh8v+w6SJEMSZIkaW7MwSSdFQtNkurundPav0GR2Owt248A7ywTm+4xLeCrZfs/gCsj4g0LGagkSdI5xBxM0lnx0jlJVWpFxEw3edzT9fynIuITFNf8XwP8PnBjZo6Xr/8FsB24NyJup7ie/6+BL2Xmw+WYzwDvAx4qb4D5FMXNLi/LzA/1+T1JkiTVnTmYpAVjoUlSlS4EHp6h/0+Bz5XPPwi8nSLJOQH8OfBPEwMz8/GIeCtwG8VNKg8Dd5b7TYw5ERHXUfzk7keB5cAzwMf7+3YkSZJeEczBJC2YyMyqY5CkU0TEOoqf1v21zLy/2mgkSZLOD+ZgkubLezRJkiRJkiSpLyw0SZIkSZIkqS+8dE6SJEmSJEl94YomSZIkSZIk9YWFJkmSJEmSJPWFhSZJkiRJkiT1hYUmSZIkSZIk9YWFJkmSJEmSJPWFhSZJkiRJkiT1xf8DDqKaudXbAxAAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 1440x360 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_performance(history=history1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "5832/5832 [==============================] - 3s 579us/step\n"
     ]
    }
   ],
   "source": [
    "accuracy = model.evaluate(X_test, Y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Model Performance of CNN (Test Accuracy):\n",
      "v: 95.16%\n",
      "Loss: 0.183\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(\"\\nModel Performance of CNN (Test Accuracy):\")\n",
    "print('v: {:0.2f}%\\nLoss: {:0.3f}\\n'.format(accuracy[1]*100, accuracy[0]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}