demo.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "initial_id",
   "metadata": {
    "collapsed": true,
    "ExecuteTime": {
     "end_time": "2023-12-04T17:15:22.510918Z",
     "start_time": "2023-12-04T17:15:22.508384Z"
    }
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "\n",
    "from matplotlib import pyplot as plt\n",
    "\n",
    "from sklearn.datasets import load_iris"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "outputs": [],
   "source": [
    "data = load_iris()\n",
    "\n",
    "X, y = data.data, data.target"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2023-12-04T17:15:22.770472Z",
     "start_time": "2023-12-04T17:15:22.767353Z"
    }
   },
   "id": "839171829f376684"
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "outputs": [
    {
     "data": {
      "text/plain": "       sepal length (cm)  sepal width (cm)  petal length (cm)  \\\ncount         150.000000        150.000000         150.000000   \nmean            5.843333          3.057333           3.758000   \nstd             0.828066          0.435866           1.765298   \nmin             4.300000          2.000000           1.000000   \n25%             5.100000          2.800000           1.600000   \n50%             5.800000          3.000000           4.350000   \n75%             6.400000          3.300000           5.100000   \nmax             7.900000          4.400000           6.900000   \n\n       petal width (cm)  \ncount        150.000000  \nmean           1.199333  \nstd            0.762238  \nmin            0.100000  \n25%            0.300000  \n50%            1.300000  \n75%            1.800000  \nmax            2.500000  ",
      "text/html": "<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>sepal length (cm)</th>\n      <th>sepal width (cm)</th>\n      <th>petal length (cm)</th>\n      <th>petal width (cm)</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>count</th>\n      <td>150.000000</td>\n      <td>150.000000</td>\n      <td>150.000000</td>\n      <td>150.000000</td>\n    </tr>\n    <tr>\n      <th>mean</th>\n      <td>5.843333</td>\n      <td>3.057333</td>\n      <td>3.758000</td>\n      <td>1.199333</td>\n    </tr>\n    <tr>\n      <th>std</th>\n      <td>0.828066</td>\n      <td>0.435866</td>\n      <td>1.765298</td>\n      <td>0.762238</td>\n    </tr>\n    <tr>\n      <th>min</th>\n      <td>4.300000</td>\n      <td>2.000000</td>\n      <td>1.000000</td>\n      <td>0.100000</td>\n    </tr>\n    <tr>\n      <th>25%</th>\n      <td>5.100000</td>\n      <td>2.800000</td>\n      <td>1.600000</td>\n      <td>0.300000</td>\n    </tr>\n    <tr>\n      <th>50%</th>\n      <td>5.800000</td>\n      <td>3.000000</td>\n      <td>4.350000</td>\n      <td>1.300000</td>\n    </tr>\n    <tr>\n      <th>75%</th>\n      <td>6.400000</td>\n      <td>3.300000</td>\n      <td>5.100000</td>\n      <td>1.800000</td>\n    </tr>\n    <tr>\n      <th>max</th>\n      <td>7.900000</td>\n      <td>4.400000</td>\n      <td>6.900000</td>\n      <td>2.500000</td>\n    </tr>\n  </tbody>\n</table>\n</div>"
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dataframe = pd.DataFrame(X, columns=data.feature_names)\n",
    "\n",
    "dataframe.describe()"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2023-12-04T17:15:23.050256Z",
     "start_time": "2023-12-04T17:15:23.047010Z"
    }
   },
   "id": "5dfc0fb3fa8c7887"
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import StandardScaler\n",
    "\n",
    "normalizer = StandardScaler().fit(X)\n",
    "\n",
    "X_normalized = normalizer.transform(X)"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2023-12-04T17:15:23.755951Z",
     "start_time": "2023-12-04T17:15:23.752767Z"
    }
   },
   "id": "91e447cfe9d4ebe2"
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X_normalized, y, test_size=0.2, random_state=42)"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2023-12-04T17:15:24.681536Z",
     "start_time": "2023-12-04T17:15:24.678928Z"
    }
   },
   "id": "7a697c889ac1f311"
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(120, 4) (30, 4)\n"
     ]
    }
   ],
   "source": [
    "print(X_train.shape, X_test.shape)"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2023-12-04T17:15:25.467666Z",
     "start_time": "2023-12-04T17:15:25.465333Z"
    }
   },
   "id": "7e6ee44145d11239"
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "outputs": [],
   "source": [
    "from src.models import Sequential\n",
    "from src.layers import Dense, ReLU\n",
    "from src.optimizers import SGD\n",
    "from src.losses import CrossEntropyLoss\n",
    "\n",
    "model = Sequential([\n",
    "    Dense(20, activation=ReLU()),\n",
    "    Dense(20, activation=ReLU()),\n",
    "    Dense(20, activation=ReLU()),\n",
    "    Dense(3)\n",
    "])\n",
    "\n",
    "model.compile(\n",
    "    loss=CrossEntropyLoss(),\n",
    "    optimizer=SGD(learning_rate=5e-2),\n",
    ")"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2023-12-04T17:14:47.992547Z",
     "start_time": "2023-12-04T17:14:47.989214Z"
    }
   },
   "id": "2f079980bdb342da"
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100 - loss: 2.2007 - val_accuracy: 0.6000\n",
      "Epoch 2/100 - loss: 2.4106 - val_accuracy: 0.6000\n",
      "Epoch 3/100 - loss: 2.3361 - val_accuracy: 0.6000\n",
      "Epoch 4/100 - loss: 2.6509 - val_accuracy: 0.6000\n",
      "Epoch 5/100 - loss: 2.3421 - val_accuracy: 0.6000\n",
      "Epoch 6/100 - loss: 2.5338 - val_accuracy: 0.6000\n",
      "Epoch 7/100 - loss: 1.8720 - val_accuracy: 0.6000\n",
      "Epoch 8/100 - loss: 2.8300 - val_accuracy: 0.6000\n",
      "Epoch 9/100 - loss: 1.7221 - val_accuracy: 0.6000\n",
      "Epoch 10/100 - loss: 1.6912 - val_accuracy: 0.6000\n",
      "Epoch 11/100 - loss: 1.2941 - val_accuracy: 0.6000\n",
      "Epoch 12/100 - loss: 1.4829 - val_accuracy: 0.6000\n",
      "Epoch 13/100 - loss: 1.3326 - val_accuracy: 0.6000\n",
      "Epoch 14/100 - loss: 1.8453 - val_accuracy: 0.6000\n",
      "Epoch 15/100 - loss: 1.2443 - val_accuracy: 0.5667\n",
      "Epoch 16/100 - loss: 1.4698 - val_accuracy: 0.5667\n",
      "Epoch 17/100 - loss: 1.6354 - val_accuracy: 0.5667\n",
      "Epoch 18/100 - loss: 1.3037 - val_accuracy: 0.5667\n",
      "Epoch 19/100 - loss: 0.9367 - val_accuracy: 0.5667\n",
      "Epoch 20/100 - loss: 1.3629 - val_accuracy: 0.5667\n",
      "Epoch 21/100 - loss: 1.2551 - val_accuracy: 0.5667\n",
      "Epoch 22/100 - loss: 1.1261 - val_accuracy: 0.5667\n",
      "Epoch 23/100 - loss: 1.1145 - val_accuracy: 0.6000\n",
      "Epoch 24/100 - loss: 1.5202 - val_accuracy: 0.6000\n",
      "Epoch 25/100 - loss: 1.0001 - val_accuracy: 0.6000\n",
      "Epoch 26/100 - loss: 1.0104 - val_accuracy: 0.6000\n",
      "Epoch 27/100 - loss: 1.4745 - val_accuracy: 0.6000\n",
      "Epoch 28/100 - loss: 1.1182 - val_accuracy: 0.6000\n",
      "Epoch 29/100 - loss: 0.5855 - val_accuracy: 0.6000\n",
      "Epoch 30/100 - loss: 0.8500 - val_accuracy: 0.6000\n",
      "Epoch 31/100 - loss: 1.1645 - val_accuracy: 0.6000\n",
      "Epoch 32/100 - loss: 1.0896 - val_accuracy: 0.6000\n",
      "Epoch 33/100 - loss: 0.7418 - val_accuracy: 0.6333\n",
      "Epoch 34/100 - loss: 0.8767 - val_accuracy: 0.6667\n",
      "Epoch 35/100 - loss: 1.0977 - val_accuracy: 0.6667\n",
      "Epoch 36/100 - loss: 1.2710 - val_accuracy: 0.7000\n",
      "Epoch 37/100 - loss: 1.0897 - val_accuracy: 0.7000\n",
      "Epoch 38/100 - loss: 0.8739 - val_accuracy: 0.7333\n",
      "Epoch 39/100 - loss: 0.7318 - val_accuracy: 0.7333\n",
      "Epoch 40/100 - loss: 0.9580 - val_accuracy: 0.7667\n",
      "Epoch 41/100 - loss: 0.7597 - val_accuracy: 0.7667\n",
      "Epoch 42/100 - loss: 0.8366 - val_accuracy: 0.7667\n",
      "Epoch 43/100 - loss: 0.9970 - val_accuracy: 0.7667\n",
      "Epoch 44/100 - loss: 0.8277 - val_accuracy: 0.7667\n",
      "Epoch 45/100 - loss: 0.8171 - val_accuracy: 0.7667\n",
      "Epoch 46/100 - loss: 0.7908 - val_accuracy: 0.7667\n",
      "Epoch 47/100 - loss: 1.1235 - val_accuracy: 0.7667\n",
      "Epoch 48/100 - loss: 0.6357 - val_accuracy: 0.7667\n",
      "Epoch 49/100 - loss: 0.9595 - val_accuracy: 0.8000\n",
      "Epoch 50/100 - loss: 0.7891 - val_accuracy: 0.8333\n",
      "Epoch 51/100 - loss: 0.5968 - val_accuracy: 0.8333\n",
      "Epoch 52/100 - loss: 0.7493 - val_accuracy: 0.8333\n",
      "Epoch 53/100 - loss: 0.6519 - val_accuracy: 0.8333\n",
      "Epoch 54/100 - loss: 0.8032 - val_accuracy: 0.8333\n",
      "Epoch 55/100 - loss: 0.5480 - val_accuracy: 0.8333\n",
      "Epoch 56/100 - loss: 0.5350 - val_accuracy: 0.8333\n",
      "Epoch 57/100 - loss: 0.9370 - val_accuracy: 0.8333\n",
      "Epoch 58/100 - loss: 0.7066 - val_accuracy: 0.8333\n",
      "Epoch 59/100 - loss: 0.7979 - val_accuracy: 0.8333\n",
      "Epoch 60/100 - loss: 0.5893 - val_accuracy: 0.8333\n",
      "Epoch 61/100 - loss: 0.6809 - val_accuracy: 0.8333\n",
      "Epoch 62/100 - loss: 0.8409 - val_accuracy: 0.8333\n",
      "Epoch 63/100 - loss: 0.8870 - val_accuracy: 0.8333\n",
      "Epoch 64/100 - loss: 0.7955 - val_accuracy: 0.8333\n",
      "Epoch 65/100 - loss: 0.5312 - val_accuracy: 0.8333\n",
      "Epoch 66/100 - loss: 0.6544 - val_accuracy: 0.8333\n",
      "Epoch 67/100 - loss: 0.6395 - val_accuracy: 0.8333\n",
      "Epoch 68/100 - loss: 0.4915 - val_accuracy: 0.8333\n",
      "Epoch 69/100 - loss: 0.7072 - val_accuracy: 0.8333\n",
      "Epoch 70/100 - loss: 0.6717 - val_accuracy: 0.8333\n",
      "Epoch 71/100 - loss: 0.7840 - val_accuracy: 0.8333\n",
      "Epoch 72/100 - loss: 0.6726 - val_accuracy: 0.8333\n",
      "Epoch 73/100 - loss: 0.6247 - val_accuracy: 0.8333\n",
      "Epoch 74/100 - loss: 0.4007 - val_accuracy: 0.8333\n",
      "Epoch 75/100 - loss: 0.7984 - val_accuracy: 0.8333\n",
      "Epoch 76/100 - loss: 0.7331 - val_accuracy: 0.8333\n",
      "Epoch 77/100 - loss: 0.6307 - val_accuracy: 0.8333\n",
      "Epoch 78/100 - loss: 0.4829 - val_accuracy: 0.8333\n",
      "Epoch 79/100 - loss: 0.5465 - val_accuracy: 0.8333\n",
      "Epoch 80/100 - loss: 0.4223 - val_accuracy: 0.8333\n",
      "Epoch 81/100 - loss: 0.6288 - val_accuracy: 0.8333\n",
      "Epoch 82/100 - loss: 0.3993 - val_accuracy: 0.8333\n",
      "Epoch 83/100 - loss: 0.3759 - val_accuracy: 0.8333\n",
      "Epoch 84/100 - loss: 0.7292 - val_accuracy: 0.8333\n",
      "Epoch 85/100 - loss: 0.4943 - val_accuracy: 0.8333\n",
      "Epoch 86/100 - loss: 0.6708 - val_accuracy: 0.8333\n",
      "Epoch 87/100 - loss: 0.5903 - val_accuracy: 0.8333\n",
      "Epoch 88/100 - loss: 0.5845 - val_accuracy: 0.8333\n",
      "Epoch 89/100 - loss: 0.4077 - val_accuracy: 0.8333\n",
      "Epoch 90/100 - loss: 0.6042 - val_accuracy: 0.8333\n",
      "Epoch 91/100 - loss: 0.5188 - val_accuracy: 0.8333\n",
      "Epoch 92/100 - loss: 0.4007 - val_accuracy: 0.8333\n",
      "Epoch 93/100 - loss: 0.4599 - val_accuracy: 0.8333\n",
      "Epoch 94/100 - loss: 0.5474 - val_accuracy: 0.8333\n",
      "Epoch 95/100 - loss: 0.5226 - val_accuracy: 0.8333\n",
      "Epoch 96/100 - loss: 0.4097 - val_accuracy: 0.8333\n",
      "Epoch 97/100 - loss: 0.5797 - val_accuracy: 0.8333\n",
      "Epoch 98/100 - loss: 0.5970 - val_accuracy: 0.8333\n",
      "Epoch 99/100 - loss: 0.5925 - val_accuracy: 0.8333\n",
      "Epoch 100/100 - loss: 0.4229 - val_accuracy: 0.8667\n"
     ]
    }
   ],
   "source": [
    "history = model.fit(X_train, y_train, epochs=100, batch_size=32, x_val=X_test, y_val=y_test)"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2023-12-04T17:14:48.441258Z",
     "start_time": "2023-12-04T17:14:48.401925Z"
    }
   },
   "id": "d22d9d73bd89e7fb"
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "outputs": [
    {
     "data": {
      "text/plain": "<Figure size 640x480 with 1 Axes>",
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAiMAAAGdCAYAAADAAnMpAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8g+/7EAAAACXBIWXMAAA9hAAAPYQGoP6dpAACH0ElEQVR4nO3dd7xT9f0/8NfJvvcmuXsvLnvvIWAFBUXcWq2DKlqltYLjSy1qrbsWf7Xu1aoV6sCJqMWJIIiI7D0u697Lhbt37sr8/P44+Zyc5Ca5Se5ILnk/H4/7KNysc6M17/teH4ExxkAIIYQQEiaKcF8AIYQQQqIbBSOEEEIICSsKRgghhBASVhSMEEIIISSsKBghhBBCSFhRMEIIIYSQsKJghBBCCCFhRcEIIYQQQsJKFe4LCITD4UBZWRkMBgMEQQj35RBCCCEkAIwxmEwmZGVlQaHwnf/oE8FIWVkZcnNzw30ZhBBCCAlBaWkpcnJyfN7eJ4IRg8EAQPxhjEZjmK+GEEIIIYFoampCbm6u9DnuS58IRnhpxmg0UjBCCCGE9DGdtVhQAyshhBBCwoqCEUIIIYSEFQUjhBBCCAmrPtEzQgghfRljDDabDXa7PdyXQki3UiqVUKlUXV67QcEIIYT0IIvFgvLycrS2tob7UgjpEbGxscjMzIRGown5OSgYIYSQHuJwOFBUVASlUomsrCxoNBpa3EjOGIwxWCwWVFdXo6ioCIMGDfK72MwfCkYIIaSHWCwWOBwO5ObmIjY2NtyXQ0i3i4mJgVqtRklJCSwWC3Q6XUjPQw2shBDSw0L9bZGQvqA7/v2m/4cQQgghJKwoGCGEEEIi0Pr16yEIAhoaGsJ9KT2OghFCCCFRJ5o+6PsCCkYIIYR0idVqDfcl9BiLxRLuS4gKFIyE0ZYTtVix5WS4L4MQQtw4HA784x//wMCBA6HVapGXl4cnn3wSAFBcXAxBEPDhhx9ixowZ0Ol0eO+99+BwOPD4448jJycHWq0WY8eOxTfffCM9p8ViwaJFi5CZmQmdTof8/HwsXboUgDgi+uijjyIvLw9arRZZWVm46667/F7j559/jvHjx0On06F///547LHHYLPZpNsFQcCbb76JK6+8ErGxsRg0aBC++OIL6Wc499xzAQCJiYkQBAE333wzAGDmzJlYtGgR7rnnHqSkpGDOnDkAgA0bNmDy5MnQarXIzMzE/fff7/Z6/HGLFi1CfHw8UlJS8NBDD4ExBgB4/PHHMXLkyA4/x9ixY/HQQw8F/M9m5cqVGDFiBLRaLfr164dnnnnG7fZXX30VgwYNgk6nQ3p6Oq6++mrptk8++QSjRo1CTEwMkpOTMXv2bLS0tAT82j2K9QGNjY0MAGtsbAz3pXSrGf9Yx/LvW82OVZnCfSmEkB7Q1tbGDh48yNra2hhjjDkcDtZitobly+FwBHzdS5YsYYmJiWz58uXs2LFjbOPGjeyNN95gjDFWVFTEALB+/fqxlStXshMnTrCysjL27LPPMqPRyN5//312+PBhtmTJEqZWq9mRI0cYY4w9/fTTLDc3l/3444+suLiYbdy4ka1YsYIxxtjHH3/MjEYj++qrr1hJSQnbsmULe/31131e348//siMRiNbvnw5O378OPvuu+9Yv3792KOPPirdBwDLyclhK1asYEePHmV33XUX0+v1rLa2ltlsNrZy5UoGgBUWFrLy8nLW0NDAGGNsxowZTK/Xsz//+c/s8OHD7PDhw+zUqVMsNjaW3XHHHezQoUNs1apVLCUlhT3yyCPS6/HH3X333ezw4cPs3XffZbGxsdLPUVpayhQKBdu6dav0mJ07dzJBENjx48e9/pw//PADA8Dq6+sZY4xt376dKRQK9vjjj7PCwkK2bNkyFhMTw5YtW8YYY2zbtm1MqVSyFStWsOLiYrZz5072wgsvMMYYKysrYyqVij377LOsqKiI7d27l73yyivMZOr654/nv+dygX5+C4w5w7YI1tTUhPj4eDQ2NsJoNIb7crrNmMe+Q2ObFSv/OA0T8hPDfTmEkG7W3t6OoqIiFBQUQKfTodViw/CHvw3LtRx8fA5iNZ2vljKZTEhNTcXLL7+M2267rcPtxcXFKCgowPPPP4+7775b+n52djYWLlyIv/zlL9L3Jk+ejEmTJuGVV17BXXfdhQMHDuD777/vsPjt2Wefxb///W/s378farW602ucPXs2Zs2ahQceeED63rvvvoslS5agrKwMgJgZ+etf/4onnngCANDS0gK9Xo+vv/4aF154IdavX49zzz0X9fX1SEhIkJ5n5syZaGpqws6dO6XvPfjgg1i5ciUOHTokXfurr76K++67D42NjVAoFJg5cyaqqqpw4MAB6T73338/vvjiCxw8eBAAcNFFF6Ffv3549dVXAQB33XUX9u3bhx9++MHrz+l5jfPmzUN1dTW+++476T5LlizBl19+iQMHDuDTTz/FLbfcglOnTsFgMLg9186dOzFhwgQUFxcjPz+/0/c4GJ7/nssF+vlNZZowareK51RYbI4wXwkhhIgOHToEs9mMWbNm+b3fxIkTpT83NTWhrKwM06dPd7vP9OnTcejQIQDAzTffjN27d2PIkCG466673D5Qr7nmGrS1taF///5YsGABVq1a5VYC8bRnzx48/vjj0Ov10teCBQs6rN0fPXq09Oe4uDgYjUZUVVV1+h5MmDDB7e+HDh3C1KlT3YKo6dOno7m5GadOnZK+d9ZZZ7ndZ+rUqTh69Kh0JtGCBQvw/vvvo729HRaLBStWrMDvfve7Tq9Hfh3e3mP+Gueffz7y8/PRv39/3HjjjXjvvfek92PMmDGYNWsWRo0ahWuuuQZvvPEG6uvrA37tnkYbWMOEMQazMwgx2+jwLEKiQYxaiYOPzwnbawd0v5iYgO4XFxcX1OuPHz8eRUVF+Prrr/H999/jN7/5DWbPno1PPvkEubm5KCwsxPfff481a9bgjjvuwNNPP40NGzZ4zZQ0Nzfjsccew1VXXdXhNvlv5p6PFQQBDkfnv/wF+7MF6tJLL4VWq8WqVaug0WhgtVrdejq6ymAwYOfOnVi/fj2+++47PPzww3j00Uexbds2JCQkYM2aNfj555/x3Xff4aWXXsKDDz6ILVu2oKCgoNuuIVSUGQkTsywbYqbMCCFRQRAExGpUYfkK9EycQYMGISYmBmvXrg345zIajcjKysKmTZvcvr9p0yYMHz7c7X7XXnst3njjDXz44YdYuXIl6urqAIhB0KWXXooXX3wR69evx+bNm7Fv3z6vrzd+/HgUFhZi4MCBHb4C3QbKD3UL5CTlYcOGYfPmzZB3NWzatAkGgwE5OTnS97Zs2eL2uF9++QWDBg2CUikGgiqVCvPnz8eyZcuwbNkyXHfddQEHf/w6vL3HgwcPdnuN2bNn4x//+Af27t2L4uJirFu3DoD479/06dPx2GOPYdeuXdBoNFi1alXAr9+TKDMSJrxEA1AwQgiJHDqdDvfddx+WLFkCjUaD6dOno7q6GgcOHMCtt97q83F//vOf8cgjj2DAgAEYO3Ysli1bht27d+O9994DIPaFZGZmYty4cVAoFPj444+RkZGBhIQELF++HHa7HVOmTEFsbCzeffddxMTE+OxtePjhh3HJJZcgLy8PV199NRQKBfbs2YP9+/fjb3/7W0A/Z35+PgRBwOrVq3HRRRchJiYGer3e633vuOMOPP/887jzzjuxaNEiFBYW4pFHHsHixYvdgp+TJ09i8eLF+MMf/oCdO3fipZde6jDtctttt2HYsGEA0CGw6Myf/vQnTJo0CU888QSuvfZabN68GS+//LLUg7J69WqcOHEC55xzDhITE/HVV1/B4XBgyJAh2LJlC9auXYsLLrgAaWlp2LJlC6qrq6VrCbsut9H2gjNxmqa8oY3l37ea5d+3mn28vTTcl0MI6QH+pgwimd1uZ3/7299Yfn4+U6vVLC8vj/39739njLmmaXbt2tXhMY8++ijLzs5marWajRkzhn399dfS7a+//jobO3Ysi4uLY0ajkc2aNYvt3LmTMcbYqlWr2JQpU5jRaGRxcXHsrLPOYt9//73fa/zmm2/YtGnTWExMDDMajWzy5MluEzgA2KpVq9weEx8fL02eMMbY448/zjIyMpggCGz+/PmMMXEq5u677+7weuvXr2eTJk1iGo2GZWRksPvuu49ZrVbp9hkzZrA77riD3X777cxoNLLExET2l7/8xesU069+9Ss2YsQIvz8fYx2naRhj7JNPPmHDhw+X/rk8/fTT0m0bN25kM2bMYImJiSwmJoaNHj2affjhh4wxxg4ePMjmzJnDUlNTmVarZYMHD2YvvfRSp9cQCJqm6cOKa1ow85/rAQBPXjkS86Z0b3czIST8/E0ZkDPLzJkzMXbsWDz//PN+78cYw6BBg3DHHXdg8eLFvXNxPaw7pmmoTBMm7bKmVbOVyjSEEHKmq66uxgcffICKigrccsst4b6ciELBSJjIAxCLnYIRQgg506WlpSElJQWvv/46EhNpt5QcBSNh4tbASpkRQgjp09avX9/pffpAV0TY0GhvmLS7jfbSnhFCCCHRi4KRMJFnRmgDKyGEkGhGwUiY0J4RQgghRETBSJjI+0SoTEMIISSaUTASJvLRXirTEEIIiWYUjIQJlWkIIYQQEQUjYdJupYPyCCFnrn79+rltIxUEAZ999pnP+xcXF0MQBOzevbtLr9tdz0N6F+0ZCRMzlWkIIVGkvLy82xd93XzzzWhoaHALcnJzc1FeXo6UlJRufS3SsygYCZN2amAlhESRjIyMXnkdpVLZa68VaaxWK9RqdbgvIyRUpgkT6hkhhESi119/HVlZWXA43P+7dPnll+N3v/sdAOD48eO4/PLLkZ6eDr1ej0mTJuH777/3+7yeZZqtW7di3Lhx0Ol0mDhxInbt2uV2f7vdjltvvRUFBQWIiYnBkCFD8MILL0i3P/roo/jvf/+Lzz//HIIgQBAErF+/3muZZsOGDZg8eTK0Wi0yMzNx//33w2azSbfPnDkTd911F5YsWYKkpCRkZGTg0Ucf9fvzbNu2Deeffz5SUlIQHx+PGTNmYOfOnW73aWhowB/+8Aekp6dDp9Nh5MiRWL16tXT7pk2bMHPmTMTGxiIxMRFz5sxBfX09gI5lLgAYO3as23UJgoDXXnsNl112GeLi4vDkk092+r5xb731FkaMGCG9J4sWLQIA/O53v8Mll1zidl+r1Yq0tDT85z//8fuedAVlRsLELTNC6+AJiQ6MAdbW8Ly2OhYQhE7vds011+DOO+/EDz/8gFmzZgEA6urq8M033+Crr74CADQ3N+Oiiy7Ck08+Ca1Wi7fffhuXXnopCgsLkZeX1+lrNDc345JLLsH555+Pd999F0VFRbj77rvd7uNwOJCTk4OPP/4YycnJ+Pnnn/H73/8emZmZ+M1vfoN7770Xhw4dQlNTE5YtWwYASEpKQllZmdvznD59GhdddBFuvvlmvP322zh8+DAWLFgAnU7n9sH+3//+F4sXL8aWLVuwefNm3HzzzZg+fTrOP/98rz+DyWTC/Pnz8dJLL4ExhmeeeQYXXXQRjh49CoPBAIfDgblz58JkMuHdd9/FgAEDcPDgQSiVSgDA7t27MWvWLPzud7/DCy+8AJVKhR9++AF2e3CZ8kcffRRPPfUUnn/+eahUqk7fNwB47bXXsHjxYjz11FOYO3cuGhsbsWnTJgDAbbfdhnPOOQfl5eXIzMwEAKxevRqtra249tprg7q2YFAwEiZuo710UB4h0cHaCvw9Kzyv/ZcyQBPX6d0SExMxd+5crFixQgpGPvnkE6SkpODcc88FAIwZMwZjxoyRHvPEE09g1apV+OKLL6TfsP1ZsWIFHA4H/vOf/0Cn02HEiBE4deoU/vjHP0r3UavVeOyxx6S/FxQUYPPmzfjoo4/wm9/8Bnq9HjExMTCbzX7LMq+++ipyc3Px8ssvQxAEDB06FGVlZbjvvvvw8MMPQ6EQCwSjR4/GI488AgAYNGgQXn75Zaxdu9ZnMHLeeee5/f31119HQkICNmzYgEsuuQTff/89tm7dikOHDmHw4MEAgP79+0v3/8c//oGJEyfi1Vdflb43YsSITt87TzfccEOHE4D9vW8A8Le//Q1/+tOf3ALASZMmAQCmTZuGIUOG4J133sGSJUsAAMuWLcM111wDvV4f9PUFiso0IXA4WJcPPDK7HZRHPSOEkMgxb948rFy5EmazGQDw3nvv4brrrpM+uJubm3Hvvfdi2LBhSEhIgF6vx6FDh3Dy5MmAnv/QoUMYPXo0dDqd9L2pU6d2uN8rr7yCCRMmIDU1FXq9Hq+//nrAryF/ralTp0KQZYWmT5+O5uZmnDp1Svre6NGj3R6XmZmJqqoqn89bWVmJBQsWYNCgQYiPj4fRaERzc7N0fbt370ZOTo4UiHjimZGumjhxYofv+XvfqqqqUFZW5ve1b7vtNinbVFlZia+//loq0fUUyowEyWJz4JKXNiIjPgZv/25yyM9Do72ERCF1rJihCNdrB+jSSy8FYwxffvklJk2ahI0bN+K5556Tbr/33nuxZs0a/POf/8TAgQMRExODq6++GhaLpdsu94MPPsC9996LZ555BlOnToXBYMDTTz+NLVu2dNtryHk2fgqC0KFvRm7+/Pmora3FCy+8gPz8fGi1WkydOlV6D2JiYvy+Xme3KxSKDr/0Wq3WDveLi3PPdnX2vnX2ugBw00034f7778fmzZvx888/o6CgAL/61a86fVxXUDASpKKaFhypbMaRyma0W+3QqZUhPQ8dlEdIFBKEgEol4abT6XDVVVfhvffew7FjxzBkyBCMHz9eun3Tpk24+eabceWVVwIQMyXFxcUBP/+wYcPwzjvvoL29XcqO/PLLL2732bRpE6ZNm4Y77rhD+t7x48fd7qPRaDrtsRg2bBhWrlwJxpiUHdm0aRMMBgNycnICvmZPmzZtwquvvoqLLroIAFBaWoqamhrp9tGjR+PUqVM4cuSI1+zI6NGjsXbtWreSilxqairKy8ulvzc1NaGoqCig6/L3vhkMBvTr1w9r166Vym6ekpOTccUVV2DZsmXYvHlzhzJQT6AyTZAqmtpdf25s93NP/+Q9I5QZIYREmnnz5uHLL7/EW2+9hXnz5rndNmjQIHz66afYvXs39uzZgxtuuMFvFsHTDTfcAEEQsGDBAhw8eBBfffUV/vnPf3Z4je3bt+Pbb7/FkSNH8NBDD2Hbtm1u9+nXrx/27t2LwsJC1NTUeM0c3HHHHSgtLcWdd96Jw4cP4/PPP8cjjzyCxYsXS2WnUAwaNAjvvPMODh06hC1btmDevHluWYcZM2bgnHPOwa9//WusWbMGRUVF+Prrr/HNN98AAB544AFs27YNd9xxB/bu3YvDhw/jtddekwKa8847D++88w42btyIffv2Yf78+VLza2fX1dn79uijj+KZZ57Biy++iKNHj2Lnzp146aWX3O5z22234b///S8OHTqE+fPnh/w+BYqCkSBVNLZJfy7vQjAin6Cx2B1wOLrWg0IIId3pvPPOQ1JSEgoLC3HDDTe43fbss88iMTER06ZNw6WXXoo5c+a4ZU46o9fr8b///Q/79u3DuHHj8OCDD+L//b//53afP/zhD7jqqqtw7bXXYsqUKaitrXX7bR8AFixYgCFDhmDixIlITU2VJkLksrOz8dVXX2Hr1q0YM2YMbr/9dtx6663461//GsS70dF//vMf1NfXY/z48bjxxhtx1113IS0tze0+K1euxKRJk3D99ddj+PDhWLJkiZTJGTx4ML777jvs2bMHkydPxtSpU/H5559DpRILFg888ABmzJiBSy65BBdffDGuuOIKDBgwoNPrCuR9mz9/Pp5//nm8+uqrGDFiBC655BIcPXrU7T6zZ89GZmYm5syZg6ysnm+6FlhXOzF7QVNTE+Lj49HY2Aij0RjWa3nh+6N47vsjAIDnrx2LK8Zlh/Q8v/rHOpTWuQKbw09cGHLJhxASmdrb21FUVISCggK3Zk1CIl1zczOys7OxbNkyXHXVVX7v6+/f80A/v6lnJEjyMk1XMiPtHrtFzDYHBSOEEELCyuFwoKamBs888wwSEhJw2WWX9crrUjASpEpZMCL/c7DaPcZ5xZXwfXONLyGEkDPDyZMnUVBQgJycHCxfvlwqG/U0CkaCJG9aLZf1jwTLc+sqTdQQQggJt379+nV5j1YoqIE1SJXdME1jd7AOW1dpooYQQki0omAkCGabHbUtrqU+FSGWaeSn9MZpxD4ROp+GEEJItKJgJAhVTWa3v1ebzLCFcK6MvHnVGCP2icgDFELImaUPDC0SErLu+PebgpEg8BJNdkIMVAoBDgZUN5s7eVRHvHlVrRQQ48yMUM8IIWcevmK8tTVMJ/US0gv4v9+eK/WDQQ2sQeBlmawEcY76dEMbyhvbkRnf+a5/Od4folMpoVUp3b5HCDlzKJVKJCQkSAeuxcbGuh3YRkhfxhhDa2srqqqqkJCQENCGWF8oGAkCb1hNN+pgdzCcbmhDZQhNrDwzolUroVGJySkKRgg5M/Hj7f2dAEtIX5aQkCD9ex4qCkaCwMs0GUYdxBJZQ0iLz3gwolMroHUGI1SmIeTMJAgCMjMzkZaW5vXsFEL6MrVa3aWMCEfBSBAqnA2sGfE68KNk/C0+k58SKccbWHVqpRSMUAMrIWc2pVLZLf/RJuRMRMFIECplZRrePOwtM8IYw/xl29DYasGnd0yHUuEekPATe8XMCPWMEEIIiW5BTdMsXboUkyZNgsFgQFpaGq644goUFhb6fczy5cshCILbV189MIo3sGbE65ARL/4M3hafVZnM+PFINfacakSVqePtZl6mUSmpTEMIISTqBRWMbNiwAQsXLsQvv/yCNWvWwGq14oILLkBLS4vfxxmNRpSXl0tfJSUlXbrocGCMuYIRoywY8VKmKawwSX9uMds63M7LNFpZzwiVaQghhESroMo033zzjdvfly9fjrS0NOzYsQPnnHOOz8cJgtDlTttwa2i1StmLNKNW+n5FU3uH3hD3YKRjkNEuz4yoncEIbWAlhBASpbq09KyxsREAkJSU5Pd+zc3NyM/PR25uLi6//HIcOHCgKy8bFjwDkhSngValRLpRzIxYbA7Ut7p3yBdWdpYZ4T0jSmiUzjJNCJtcCSGEkDNByMGIw+HAPffcg+nTp2PkyJE+7zdkyBC89dZb+Pzzz/Huu+/C4XBg2rRpOHXqlM/HmM1mNDU1uX2FGw9GeBCiUSmQotcA6Hh67xFZMNLsJRjhzapatQJaNTWwEkIIiW4hByMLFy7E/v378cEHH/i939SpU3HTTTdh7NixmDFjBj799FOkpqbi3//+t8/HLF26FPHx8dJXbm5uqJfZbfgkTYasRMMDE/l4r93B3IKRFovvnhG30V4r9YwQQgiJTiEFI4sWLcLq1avxww8/ICcnJ6jHqtVqjBs3DseOHfN5nwceeACNjY3SV2lpaSiX2a3kkzRcpvPP8vHe0rpWt4Pwmr31jNhcPSNUpiGEEBLtgmpgZYzhzjvvxKpVq7B+/XoUFBQE/YJ2ux379u3DRRdd5PM+Wq0WWq3W5+3hUOlRpgFcgYl8JfxhWfMq0FnPiIIaWAkhhES9oIKRhQsXYsWKFfj8889hMBhQUVEBAIiPj0dMjHhY3E033YTs7GwsXboUAPD444/jrLPOwsCBA9HQ0ICnn34aJSUluO2227r5R+lZFY2usV6O/1meGZGXaAD/o71imYZ6RgghhES3oIKR1157DQAwc+ZMt+8vW7YMN998MwDg5MmTUChc1Z/6+nosWLAAFRUVSExMxIQJE/Dzzz9j+PDhXbvyXsZXwafHyzMjMc7bXMEIH+vVqhQw2xzeG1hlmREN7RkhhBAS5YIu03Rm/fr1bn9/7rnn8NxzzwV1UZFIfkgex/8s38LKx3rH5CRga3EdWv30jGhV8rNpKDNCCCEkOnVpz0i0MNvsqGuxAPAIRjxWwpttdhTViNtox+UnAACa/U7T0Nk0hBBCCB2UF4AqZ4lGo1IgIVYtfZ8HIyazDc1mG07WtsLuYDDqVBiQqgfgvWfEbHMtPaPMCCGEkGhHmZEAyM+kka9912tVMGjFeK6isV1qXh2aYYTe+X2/Z9OolFLPCB2URwghJFpRMBIAb5M0nDTe29QujfUOztAjzhmMeN0zIh/tpQZWQgghUY6CEQ8WmwNWjwVk0o6ReN/BSLksMzIkw4g4jdgL0tnZNNI6eNozQgghJEpRMCJTZWrHhCfW4Pdvb3ebHKrwsgqec03UtEljvUPSDVJmpLM9I7SBlRBCSLSjYERmW1E9TGYbfiisxo6Seun75V62r3I8M3KsqhmnG8QD84akG1w9I16maVwNrPINrFSmIYQQEp0oGJE5Xt0s/fnNjUXSn6VD8vyUaTYerRH/btQhPlYtZUbarQ7YPLIeUmaE9owQQgghFIzIHatyBSPfHqxASa24M4RP02R6C0ac2ZJa5x6SwRkGAECcVindp8XinvXgPSNa2QZWi90R0FI5Qggh5ExDwYgMz4wYdSowBizbVAzGmLRnxF+ZhhvqDEa0KiXUSnEMWN43YrM7YHOIQYeYGRGDFsYAq52CEUIIIdGHghEnh4NJwcif5wwBAHy0vRTFta1Sc2maoWMwkuk8n4YbnG6Q/uytiVVejpEvPRNvo74RQggh0YeCEafTDW1otzqgUSpw3eQ8DM0woNVix3NrjgAAUvQaqaQilxirdvs+z4wAQJyG7xpxBSPtskZVrUohTdMAtPiMEEJIdKJgxIlnRfqlxEKtVODWswsAAF/sKQPgvUQDAIIgSH0jggAMTNNLt/G+kRbZ4rN2Z8ChUSmgUAhQKAQpIKEmVkIIIdGIghEn3rzKz5S5bGwWUvSuvSLetq963tYvOQ46tatx1bWFtWNmRCfLptBEDSGEkGhGwYjT8WpxcoZnNrQqJeZPzZdu97Z9leNNrENk/SIApF0jrRYvwYgsaNHQSnhCCCFRjIIRp+MemREAmHdWPnTOpWT+MiNDM8UgZGK/RLfv856RFrfMiGv7Kqelw/IIIYREMVW4LyBSHHP2jMh7PpLiNFjwq/54ad0xTB2Q7POxC37VH1MKkjA6J8Ht+94OyzPLDsnjpPNpKBghhBAShSgYAVDXYkGdc2lZ/9Q4t9sWnz8Yf5w5ALEa32+VWqnAhPykDt/XazseltfuLMXw/SLin/lKeApGCCGERB8q08A1SZOdENMh6BAEwW8g4o/3BlZepnG99a4trNQzQgghJPpQMAJZv4isRNMdvC8969jASpkRQggh0YyCEcjHeuM6uWdw4jTOMo2lY2bEvUxDPSOEEEKiFwUj8N682h1cmRHZ0jMvDawamqYhhBASxSgYgatnRD7W2x30Xso0/kZ7ac8IIYSQaBT1wUi71Y5T9W0Aei4z4nUDq5o2sBJCCCEABSM4Ud0CxoCEWDWS4zTd+txSmcbScbRXp/K2gZWCEUIIIdEn6oORY7ISjSAI3frcei89I2avZRpqYCWEEBK9oj4Y4WO9A7u5XwQAYp3TNN7KNFqvB+VRzwghhJDoE/XBiJQZSevesV7AlRmx2Byw2sWsh9+D8mjPCCGEkCgU9cGIlBnp5uZVwNUzArgmangpxr2BVQxMLHYKRgghhESfqA5G7A6GEzUtALp/rBcQMx4apfgWt1jEjIhUppH3jKgpM0IIISR6RXUwcqq+FRabAxqVAjmJsT3yGnEeh+V52zPCAxbqGSGEEBKNojoY4cvO+qfEQano3kkaznPXiGu0V1amUdMGVkIIIdErqoORYz10QJ6c5xZW7xtYabSXEEJI9KJgBD0z1st5ntxr9jJNQ6O9hBBCollUByPHq53Nqz2YGXHtGvFoYFXRQXmEEEIIAKg6v8uZa9qAZMRqlBiWYeix1+hQprH5OyiPghFCCCHRJ6qDkT9dMKTHX8PzfBrvB+VRzwghhJDoFdVlmt7gmRkxe8mMUJmGEEJINKNgpIe59ozYYbU7YHcwAO6n9lIDKyGEkGhGwUgPk+8Z4SUawLVbBHCVbKhMQwghJBpRMNLD5GUavmNEEDymaZTOs2koGCGEEBKFKBjpYbGajpkRrUoBQXBtfNVSZoQQQkgUo2Ckh+llZ9PwnhB58yrgypLYHQw2OrmXEEJIlKFgpIe5NrDapTKNvEQDuKZpAMqOEEIIiT4UjPQw+Z6Rdi+r4AHXqb0A9Y0QQgiJPhSM9DBvDazysV4AUCkVUDlPDabMCCGEkGhDwUgPk5dpXD0jHd922jVCCCEkWlEw0sP0zmkai90BU7u4hVXrUaYBaAsrIYSQ6EXBSA+L1boCj9oWC4COPSMAnU9DCCEkelEw0sPUSoWU9ahtNgMAdCovZRo1lWkIIYREJwpGegFvYq1t9p0Z4RM1lBkhhBASbSgY6QX8sLzaFmdmxFsDK21hJYQQEqUoGOkFcc4mVt4zolX56RmxUjBCCCEkulAw0gs6lmk6vu28TGOhdfCEEEKiDAUjvSBOCkZ4mcZLZoSXaazUwEoIISS6UDDSC6QtrBbv6+AB+dIzyowQQgiJLhSM9IJYjfdTeuU0tGeEEEJIlKJgpBfwMg3nLzNCG1gJIYREm6CCkaVLl2LSpEkwGAxIS0vDFVdcgcLCwk4f9/HHH2Po0KHQ6XQYNWoUvvrqq5AvuC/SBxGM0NIzQggh0SaoYGTDhg1YuHAhfvnlF6xZswZWqxUXXHABWlpafD7m559/xvXXX49bb70Vu3btwhVXXIErrrgC+/fv7/LF9xUdMyPeyjTUM0IIISQ6qTq/i8s333zj9vfly5cjLS0NO3bswDnnnOP1MS+88AIuvPBC/PnPfwYAPPHEE1izZg1efvll/Otf/wrxsvsWvdY9E6Lzs2eEyjSEEEKiTZd6RhobGwEASUlJPu+zefNmzJ492+17c+bMwebNm30+xmw2o6mpye2rL/PMjGi9bWClMg0hhJAoFXIw4nA4cM8992D69OkYOXKkz/tVVFQgPT3d7Xvp6emoqKjw+ZilS5ciPj5e+srNzQ31MiNCQA2s0p4RyowQQgiJLiEHIwsXLsT+/fvxwQcfdOf1AAAeeOABNDY2Sl+lpaXd/hq9qUMDq5cyDW1gJYQQEq2C6hnhFi1ahNWrV+PHH39ETk6O3/tmZGSgsrLS7XuVlZXIyMjw+RitVgutVhvKpUUkzz0j3g/Ko7NpCCGERKegMiOMMSxatAirVq3CunXrUFBQ0Oljpk6dirVr17p9b82aNZg6dWpwV9qH0WgvIYQQ4ltQmZGFCxdixYoV+Pzzz2EwGKS+j/j4eMTExAAAbrrpJmRnZ2Pp0qUAgLvvvhszZszAM888g4svvhgffPABtm/fjtdff72bf5TIFUwDK5VpCCGERJugMiOvvfYaGhsbMXPmTGRmZkpfH374oXSfkydPory8XPr7tGnTsGLFCrz++usYM2YMPvnkE3z22Wd+m17PNMFsYKUyDSGEkGgTVGaEMdbpfdavX9/he9dccw2uueaaYF7qjBLn2TPiZ88ILT0jhBASbehsml6gUiqkplWFAKiVQof7aOhsGkIIIVGKgpFewptYtSolBKFjMEINrIQQQqIVBSO9hPeNeBvrBahMQwghJHpRMNJLYjU8GOnYLwLQQXmEEEKiFwUjvYQflucrGNFSzwghhJAoRcFIL4mTekZ8lGnU1DNCCCEkOlEw0ktcPSM+yjTOs2msdgaHo/MRakIIIeRMQcFIL9FrOmlglQUptIWVEEJINKFgpJd0lhmRl29oCyshhJBoQsFIL5EaWL1sXwUAlUKAwrl+hPpGCCGERBMKRnqJMUYNoOM5NZwgCDTeSwghJCoFdTYNCd1lY7NQVNOCeVPyfd5Hq1Ki3eqgYIQQQkhUoWCkl6QZdHjyylF+70Mr4QkhhEQjKtNEkJ48LK+wwoR7PtiFopqWbn9uQgghpCsoGIkg2h7sGXl/60l8trsMH28v7fbnJoQQQrqCgpEI0pOH5dW1WAAATe3Wbn9uQgghpCsoGIkg0jSNtft7RngQYmq3dftzE0IIIV1BwUgEkQ7L64ENrI1tYjDSTMEIIYSQCEPBSAThK+F7YgMrD0ZMZgpGCCGERBYKRiIIPyyvJ3pGmtrEIIQyI4QQQiINBSMRRKvmo73d2zPCGEMTL9NQZoQQQkiEoWAkgvTUaK/Z5pD6UCgYIYQQEmkoGIkgPRWM8H4RADC1W8EY69bnJ4QQQrqCgpEIwveMdPcGVnkwYrUzOvuGEEJIRKFgJIL01Nk0TW3ui86oVEMIISSSUDASQXqjTAPQRA0hhJDIQsFIBOmpg/I8V8BTZoQQQkgkoWAkgvTU2TSNre7BCK2EJ4QQEkkoGIkgfM9Id/eMNLa5Bx8mOiyPEEJIBKFgJILwDaxUpiGEEBJNKBiJIDwzUtti6dZdIB0aWCkYIYQQEkEoGIkgg9IMAIBdJxvw4Gf7YXd0T0DiOdpLPSOEEEIiCQUjEWRkdjz+fuUoCAKwYstJ3P3Brm4p2fDMSKpBC4AyI4QQQiILBSMR5oYpeXjxunFQKwWs3luO37+zHW2WrjW08mAkOyEGAO0ZIYQQElkoGIlAl47Jwhs3TYROrcD6wmrc+J8tXcqQ8LJMdqIzGKHMCCGEkAhCwUiEmjkkDe/dNgUGrQrbS+rx07HqkJ+LZ0ZynJkRGu0lhBASSSgYiWAT8pMwuSAJAFDZZA7pOWx2h5QJ4ZkRamAlhBASSSgYiXDJeg0AoLY5tGBEHnhkxlOZhhBCSOShYCTCJcWJEzC1LZaQHs9LNHEaJRJj1QAoGCGEEBJZKBiJcClSZiS0YIRvXzXGqGHQOYMRKtMQQgiJIBSMRLikODEYqetiZiQ+Rg29TgUAMFFmhBBCSAShYCTCJeu7VqZpch6SZ4xRQ68VgxGLzRHUYXzduZqeEEII8UTBSIRLjutaAyvPjBh1rmAECLxUc+/HezDrmQ1ooWwKIYSQHkLBSITj0zR1IR6eJy/TKBUCYjVKAIE1sbZb7Vi16zRO1LTgYHlT0K9NCCGEBIKCkQjHe0ZsDiaVXILBG1jjY8TmVZ4dCWTXyMHyJumwvsqm9qBfmxBCCAkEBSMRTqtSwuAMIGpbgi/VSGWaGPE5DM4m1kAyI/tPN0p/DnXpGiGEENIZCkb6gCQ+3htCE2tTm0dmJIjx3r2nXMFIlYkyI4QQQnoGBSN9QFJc6LtG5A2sAKQsi8nc+fk08sxIFWVGCCGE9BAKRvqAZGkLa/ABQYfMiDMY6Swz0max42hVs/R36hkhhBDSUygY6QP4eG9dCJmRJmfQER/LyzSBLT6TN68CQJWJMiOEEEJ6BgUjfUByF3pGPMs0gWZGeIkmLykWAGVGCCGE9BwKRvoAqWckyGCEMdahTGMMcJpmnzMYmTUsDYA4CtxmCXxrKyGEEBIoCkb6gBTnSvi6IHtGWi122JylFj7ay8s0nWVG9jknaaYNSEGMWlyURhM1hBBCegIFI31AqNM0vESjVgpSQKHXihmSJj/BiNi8agIAjM6JR5pRDIZo1wghhJCeQMFIHxBqz4h8+6ogCABkmRE/o70HyxvhYECaQYt0ow7pBh0A6hshhBDSMygY6QP4aG9diwUOR+Dn0zS2ujevAq49I/56RniJZlR2PABImRGaqCGEENITKBjpA3iZxu5gUrYjELwUY4xxBSOB9IzsOy0eijeSByPOzEgVZUYIIYT0gKCDkR9//BGXXnopsrKyIAgCPvvsM7/3X79+PQRB6PBVUVER6jVHHY1KIZ0pE0ypptFjkgYI7GyafacbAIj9IgCQLvWM9H4w0thmRX0II82EEEL6jqCDkZaWFowZMwavvPJKUI8rLCxEeXm59JWWlhbsS0e15BCaWF2H5MkyI52c2ttqseGYc/MqL9OkG52ZkV4u0zgcDHOf/xGzn90Ai83Rq69NCCGk96iCfcDcuXMxd+7coF8oLS0NCQkJQT+OiJL1WhTXtgY13uvaMeL6x2xwTtOYbQ5YbA5oVO7x6MGyJjiYmA1JcwYhaYbwZEbqWy0oaxRfs7bFjMz4mF59fUIIIb2j13pGxo4di8zMTJx//vnYtGlTb73sGYP3jdSEkhmRNbDGaZXSn72VaviyM54VASAFJb19WF6drDzT1Nb5KcOEEEL6ph4PRjIzM/Gvf/0LK1euxMqVK5Gbm4uZM2di586dPh9jNpvR1NTk9hXtUpzjvXVB9E/IR3s5lVIh7Rzx1sTKg5GRsmCE94yYzDa0WnovKJD3x/DAihBCyJkn6DJNsIYMGYIhQ4ZIf582bRqOHz+O5557Du+8847XxyxduhSPPfZYT19an+JafNYxO8EYg83BoFa6x5aeq+A5g06FNqsdJi+7RvhYL29eBcQ+kxi1Em1WO6qazOiX0uP/2gDwzIxQMEIIIWeqsIz2Tp48GceOHfN5+wMPPIDGxkbpq7S0tBevLjLxXSPepmleXX8cQx/6BjtK6ty+762BFfA93ttituF4tdi8Ks+MCIIQlokayowQQkh0CEswsnv3bmRmZvq8XavVwmg0un1FO2kLq5eekW/2V8DuYPjfnnK37/M+iw6ZER+Lzw6Wi82rGUadtFuESwvDRE2d7GcNZr/KydpWnPOPH7B8U1FPXBYhhJBuFnS+vbm52S2rUVRUhN27dyMpKQl5eXl44IEHcPr0abz99tsAgOeffx4FBQUYMWIE2tvb8eabb2LdunX47rvvuu+niALyLaxydgfDkUrxHJldJ+vdbvPWwAq4MiOe470HvPSLcOGYqJFPDgXTwPrz8RqcrGvFF3vKcPP0gp64NEIIId0o6GBk+/btOPfcc6W/L168GAAwf/58LF++HOXl5Th58qR0u8ViwZ/+9CecPn0asbGxGD16NL7//nu35yCdk3pGPIKRktoWmJ07OA6UNaHdaofO2aDqrYEVkO0a8ciMHK9uAQAMTtd3eP1w7BoJtUzT4LxvtZf+GkIIIZEn6GBk5syZYMz3+SjLly93+/uSJUuwZMmSoC+MuOPTNPWt4vk0CoV48F1hhUm6j83BsP90Iyb2S4LV7kCrxQ7AWzAi/t2zZ6SoRgxGClLiOrw+7xnpzZXwbg2sQZRp6lvFx1WbzGCMSYcEEkIIiUx0Nk0fkSg7n0aeJTgsC0YAYNfJBgDumQReluEMPk7u5cFI/9SOwUiadHJvL/aMhJgZ4QcEtlsdftfeE0IIiQwUjPQRaqUCRul8GldAwDMjGc4yyk5n3wgfhTXoVFAq3DMDBi/TNO1WO8oa2wAABSkdyzT85N5KU3imaYIZ7W1odd23mk4aJoSQiEfBSB+SoneO98qmTAqdzau/mZgDQAxGGGM+m1cB7z0jJbWtYAww6lRIjO34GN4zUt1LmRHGmNsBeU1+Thn2xMs0AAUjhBDSF1Aw0od4NrG2WmworhVLK9dMzIVSIaCyyYzyxnbpw9uzXwTwPk1TVCPuFylI1XvtseDBiMlsQ0svlD6a2mywOZjs70GUaWT3pSZWQgiJfBSM9CHSrhFnMHK0shmMiSf65ibFYlimAYCYHWn0sX0VcGVG5GWaE7xfxEvzKn9MrEac0umNiZpajwMBQy3T1FBmhBBCIh4FI31IEt814izT8H6RIRliEDIuNxEAsLOkQbZ9tePAlKuBVZYZqfY9ScNJ4729MFHDm1flJSW7w/cUl5xbmYYyI4QQEvEoGOlDUqTMiPgBe9gjGBmfnwAA2FVa7/NcGgAwOPtI5MEIL/f4C0akxWe9khkRA4p+KbHS90wBjPe2W+3S3hWAekYIIaQvoGCkD/HsGSmsFE8zHpYhrssfnydmRg6cbpI+hP2Vadx7RgIIRsKQGUk36KRThgPZwiov0QAUjBBCSF9AwUgfkixN04gfsJ5lmrykWCTFaWCxO7D5eC0A/9M0fM9IY5sVNc08E+GnTNOLK+F5MJIUp5ECqkB2jchLNACVaQghpC+gYKQPSXZmRupaLKhpNqOm2QJBAAani8GIIAgYn5cAwDXyG+9lTJf3jLRbHbDaHSh2ZkXSDFopUPGmN1fC8/HlJL1G6nsJZAsrZUYIIaTvoWCkD5Gf3MuzIvlJsYhxTrkAwDhnqYbzlhmJkwUcze22gEo0gGzxWa9kRsQgIjlOI/0MgUzUNDgzI9kJMQCAmmZxfT4hhJDIRcFIH8J7RupbLThYJvaL8BINN86ZGeG89YyolQro1OI/+mZzEMGIgfeM+M82OBwMz3xXiB8OV/m9nz+1UplGG1SZhh+SNzBN3CJrd7AOpRtCCCGRhYKRPiQpVgxGHAzYUiT2hAxxNq9yY3ISIN/+bvQSjACuiRpTEJkR6bC8TkofO07W46V1x3DPh7thszv83tcX3jOSHKeRfoZgyjQpeq0UvFHfCCGERDYKRvoQlVKBBGcPyJYTdQCAoR6ZkTitCkNlAUq8lz0jAGDQunaNBF6m0UmP8XcAHc+cNLZZsa243u9z+hJqAysv0yTEqqVRaOobIYSQyEbBSB/Df9vn58p4lmkA91KNr8yIayW81e9pvW6P0aoQx7ew+ukbqZOVRdYcrPT7nN4wxmRlGo10QGAwo72JsWqkOqd/aigzQgghEY2CkT4mxbmFFQC0KgX6JXcMIMbLmli9NbACrvHeopoWNJttUAhAblKs1/vKBTJRIz/gbs2hCjAWXANpi8UOi3NxWbI+yDJNm/ja8bEapDpHoSkzQgghkY2CkT6GZ0YAYFC6HkpFx0PtJvVLAiA2r+rUyg63A65gZN/pRgBAdmIMtCrv95VLDWDXSJ0sGCmta5PGjAPF193r1ArEalRSMBLYnhHxPgkxrswIBSOEEBLZfC+VIBGJj/cCwJB0o9f75CXH4sXrxyHRy44Rjjew7jslBiMFKfqAXt91Po2fzIjH9MqaA5VufSydqZXGesVgIpjR3kapTKOhYIQQQvoIyoz0McmyzAg/pdeby8Zk4VeDUn3ezhefdXZaryfXRE3nmZHJzgzNmkPB9Y3Im1cByJaeBdAz0uZqYJWCEeoZIYSQiEbBSB/DV8ID3ptXA+W5abWzSRqO7xqpDCAzcvXEHAgCsPdUI8ob2wK+tlqPYCTQaRrGmKtME6tGql68VsqMEEJIZKNgpI+R94x0KRjRuQcj/s6kkQtkC2t9ixgQDE43YFxuAgDg+yCmauplO0aAwMs07VaH1PiaQGUaQgjpMygY6WNSnJmRpDjXtEgoPDMjgZZp+Gv6G5eVyiyxGpw/PAMA8F0QwYhnmYafr2O2OdButft8HC/RqBQC4jRKKRipb7VKQQohhJDIQ8FIHzM+PwFzR2Zg8fmDIQgdJ2kCZZBlRjRKBbKcZ7l0JkXa3eF9xXqbxY42Z8CQGKfG+cPTAQC/nKgNaDQXkJVpnM26eo0K/Ef19xw8I5MQq4EgCEiIUUPlnDbiTbGEEEIiDwUjfYxWpcRrv52A356V36XnkQcj+cmxXkeEveGZkcY279kG3i+iVgrQa1UYmKZH/5Q4WO0MGwqrA3qNOo8yjUIhSBtj/S0+kzev8sfx6aMaU+SdT8MYC3oHCyGEnIkoGIlSeq1r7DfQfhFAbCb1l23ggUSiMzsBAOePELMjgW5jlR+SJ71ubOdNrI2yHSOca6Km508aDsah8iYM+es3eO77o+G+FEIICTsKRqKUvGck0H4RoPNsA8+MyBttL3CWan4orII1gIPz6pxBjvw5pCZWf2WaVleZhovULazfHqiAxe7A6r1l4b4UQggJOwpGopS8TBPoWC/Hm2i9ZRvkmRFubG4iUvQamNpt0gF//vANrMneghE/mRHPMg2AiJ2oOVDWBAAormlBm8V3Uy4hhEQDCkailDwzEmow4jUz0tIxM6JUCDh3SBoAYONR/30j7VY7WpwfzkmybbN814jfYMRfmSbCgpGDzmDEwRD0unxCCDnTUDASpfQ6FdRKsaejf2pgq+A5f5tN6/g69jj3VfTTB6YAAH4+Xuv3uXlmRa10Na0CgW1hbXCWiBLjvJRpImgLa32LBacbXEvgDpU3hfFqCCEk/OhsmiilVirw9NVj0G61S8FFoFL87Bqpl+0YkZs6IBkAsL+sEY2tVqkh1ZO3BlggsC2sPDMS75YZibwtrAc9gg8KRggh0Y4yI1HsinHZuG5yXtCPS+ENrF52jdR5yU4A4gF7A1LjwBjwS5Hv7IjnKnguoJ4R2Sp4LhLLNAfKxMMJNSrx/34UjBBCoh0FIyRo/AO+xssHvLeeEW7aALFUs9lPqYZP0shPJwYAY0zn0zS8gVXePBuZwYgYfMwZIW6nPVxuon0jhJCoRsEICVqKnz4Mb9M03DRnqebn4zU+n7u2ueOOEaArZRrxeVosdrRaOj/1tzfsPy1mRi4bkwW1UoDJbMOp+sAPEiSEkDMNBSMkaFJmxFvPiJc9I9xZ/cVg5Ehls89Mhef2VU5qYPWxgZUx5rVME6dRQqcW/zWPhC2srRYbTtS0AADG5MZjYJp42KFnHwkhhEQTCkZI0HhmpKHV6rbEjDEmnQ/j2TPCvzc80whAPKvGG89D8rjOMiNtVjsszmuRZ2UEQYioLayHyk1gTAzo0gw6DMs0OL9PwQghJHpRMEKClhCjls6yqZU1sbZYXAGB5zQN5yrVeA9GOm1g9dEzwrMiaqWAWI3S7bZI2sJ60Nm8OiJLDMp4cEbBCCEkmlEwQoKmUAhSGUVequHNqzq1AjEeAQE3baAYjGz20Tfiu0zjmqbx1uzJy0MJHiPBQGQ1sfLmVR6MDJOCEVp8RgiJXhSMkJB4+4Cv87FjRG5SvyQoFQKKa1vdFn91eA4fZRoHA5rNHftGvB2S5+9aw8UVjMQDcAUjJ+taYfIzKUQIIWcyCkZISLxN1PjaMSJn0KkxKlv8IPY24lvb7H20V6tSQKMU/3X1toW1oa1j8yqXqtd1uNZAfXegAjtKOj9PJxBWuwOFFWIGhGdGkuI0SDeK7yW/jRBCog0FIyQk3raw+tsxIudrxNdqd0iBhudoryAI0kQNz4LIycs0nkLNjJQ1tOEP7+7AvDe3SBmbrjha2QyL3QGDToW8pFjp+8Oob4QQEuUoGCEhSTE4e0Zk47L+dozIyZefyfs/eDCjELyXW/wtPvN2SB4XajBSUtsKxoB2qwPv/VIS1GO94ZtXh2ca3fpaeDBykPpGCCFRioIREpJUb5kRPztG5CbkJ0KjVKC8sR3Fta3S92tlwYxCIXR4nL+V8I3+yjQhBiOVTa5R4Ld/KYHZZg/q8Z48+0U4yowQQqIdBSMkJN4Wn9XxHSOdZEZiNEqMy0sA4F6q8dW8yvnbNcKzKt7KNPKzdIJZu14hC0aqTWb8b095wI/15qDHJA033LlrpLDCBLuD1sITQqIPBSMkJClednfwgCAxzvuJvHK8VPPzMVcTq68dI5yrTBNcAyu/Vovd4XODqzcVjWIwwoOgNzeeCPkMGYeDSVtWR2S7ByP9kuOgVSnQZrWjpLYlpOcnhJC+jIIREhJvDazSNE0nmRHAtW/ky33luOKVTXjrpyIcrRR7JjwnaTijjq+E99Yz4syMxHR8rE6tlB4bzBZWXqa5ZXo/xKiVOFxh8rmsrTMn61rRbLZBo1JgQKre7TaVUoEhGXwTa9/qG2lqt+KKVzbhXxuOh/tSCCF9GAUjJCS89FEvWwkf6DQNAIzPS8TckRlQCMDu0gY8vvogXlp3zO/j/ZVpeANropfMCOAqK1U1Bd43wss0Q9IN+M3EHADAf34qCvjxcrxfZGiGAWplx//bDcvom30jW07UYXdpAz7cVhruSyGE9GEUjJCQJMZqpJXwvNejPojMiFIh4LXfTsAvf5mFRy4dLvWQAMBAj8wB53eaxhmgxPsIRniA4+/UX0+VzjJNRrwOt0wvgCAA6w5X4VhVc8DPwe33WAPvqa+eUcOzR8G8r4QQ4kkV7gsgfRNfCV9lMqPaZEaqXot6Z3YikMwIl2YQP+hvmV6A0rpWHKk04exBKV7v62uaRjyx13cDq9tjA9xy6nAwVDn7YTLidciMj8HsYelYc7ASb20qwt+vHBXQ83C+Jmm4vjpRw9+jRueafs9V/IQQEgjKjJCQybewmtpt0iSItybSQOQmxWLWsHRoVd7PtYmXzqdxb0JttdhhtYuv7atMY/TxWF9qWsywORgUgmuM+bazCwAAK3ecCnoJGg8yhvvIjAx1BiNlje1SYNUXVDkzI3YH87qmnxBCAkHBCAlZCh/vNZml5tU4jRI6tfdgoqv4BlbP7AYv0WiUCsT4eG2p+TXAzEhlo/gbf4peC5Wzx2NyQRJGZcfDbHNg1a7TAV93Y5tVmjoalOa9BBUfo0Z2QgwA4HAfWgtfJZumolINISRUFIyQkMn3d0jbV4Mo0QTLVwMrb5yNj1X7LBPIT/0NBG9ezYjXSd8TBAEzBqcCQFAjuEU14n1TDVoYdL6zRoPTxUDleHXwPSnhIl8M1+BlTT8hhASCghESMvkW1mAmaULlq2eEBye+SjRuj/Wyo8SbikbxROF0o87t+6FM5RTViMFF/5Q4v/fjI7/Hq/rOrpFK2fsQaKBHCCGeKBghIZNvYQ1mx0ioeHajxWKXxokB+bk0vl9bKvEEmxnxCEbSDB1PK+7MiWoxuOjvY0qI47efqOkbmRGb3YHaFtf70EDBCCEkRBSMkJDJt7A2BHguTVfwvg8AMMkyHK4TewPJjAQYjDS6JmnkpMyIKfDlaSecZZrOMiP9U8Xb+0qZRlyv7/o79YwQQkJFwQgJmXwLa6Dn0nSFSqlAnEZsUJVnOPwdkscFO03DeyE8yzRpBvHv1SZzwKvhXZmRwMo0p+rb0G7t2qF8vcEzIKOeEUJIqCgYISFLMbgaWF09I6GN9QbK6KWJtbMdI0AImREfZRqeGWm3OmAKYJTV4WAodmZGCjrJjKToNTDoVGAMKO4DZ9RUevTNUGaEEBIqCkZIyHhmpL7VIvVQ9OQ0DSDbNSILKviyNf+ZkeB6RlzbV7Vu34/RKGHQOs+5MXXeN1LR1I42qx0qhYDcpFi/9xUEQcqO8GxKJJNP0gBAY1vf2Y9CCIksFIyQkPGV8IxBWpGe1INlGkA+UePKSgTSwMpHalssdthkza/etJhtUtYjIz6mw+08OxJIMMLHevOSYr2eSeOJl3JO9IG+Eb5jhB8LQJkRQkiogg5GfvzxR1x66aXIysqCIAj47LPPOn3M+vXrMX78eGi1WgwcOBDLly8P4VJJpFEqBKlhtbS+FUDPZ0a8lWn4b+T+MiMGH82v3vASjV6rgl7b8cQEVxNr58EIDyo66xfhpPHePpAZ4dtX85PFjA/1jBBCQhV0MNLS0oIxY8bglVdeCej+RUVFuPjii3Huuedi9+7duOeee3Dbbbfh22+/DfpiSeThpRrey9mT0zRAxy2s7Va7lKHwF4yolQrE8ubXTvpGeIkm3aj1enswmZETAfaLcAP6YGZkcJp4yB9lRgghoQr6oLy5c+di7ty5Ad//X//6FwoKCvDMM88AAIYNG4affvoJzz33HObMmRPsy5MIw7ewcj05TQO4yjQ/Ha3BntIGbDhSjVaL3Xkt3oMH+WNbLfZOJ2q8bV+V4xM1gYz3BrpjhJNnRiL94DneMzI4w4BvDlRQMEIICVmP94xs3rwZs2fPdvvenDlzsHnzZp+PMZvNaGpqcvsikSnVIwAI9ZC8QPEG1p+O1eDr/RVotdiRGa/DwnMH+Dz3hfN1to2nCh9jvVxwmRExwxFoZiQvORYKAWg22wJ6/lA1m214ce1RVDQGvi/FE5+m4WvsG6lMQwgJUdCZkWBVVFQgPT3d7Xvp6eloampCW1sbYmI6NgguXboUjz32WE9fGukG/IMZEJeSBdKk2RXnDE7BmxtPIDcpFhcMT8f5wzMwMtsYUAbB1zp5T9IkjY9gJC3AYMRss+NUvbhWPtCeEa1KibykWBTXtuJYdTPSfFxDV73x4wm8sPYoTta14p/XjAn68fLtq4OcZRqT2Qab3SEdLEgIIYHq8WAkFA888AAWL14s/b2pqQm5ublhvCLii7w00tP9IgAwIT8J+x+bE1L5wuhlLNib8kb/ZZpAMyMlta1gTGyE9cwg+dM/VY/i2lacqG7BtAEpAT8uGDtP1gMA9pQ2hPR4vn1VqRDcsj5N7bZe+feAEHJm6fFgJCMjA5WVlW7fq6yshNFo9JoVAQCtVgutNvD/eJPw4YvPgJ6fpOFC7aPg6+Q76xnxtX2VSzMGNk0j37wazDX3T4nDOvTcWnjGGPaeagScr9FqsSFWE9x/Cni/TKpeC41KAb1WhWazDY1tVgpGCCFB6/F86tSpU7F27Vq3761ZswZTp07t6ZcmvcAtM9LDzatdFWhmxNf2VY5nOepaLG4H9nkKtl+EG5DWs4vPSmpbpWZTBwMOlZuCfg7eL8InjngvD9+GSwghwQg6GGlubsbu3buxe/duAOLo7u7du3Hy5EkAYonlpptuku5/++2348SJE1iyZAkOHz6MV199FR999BH+7//+r3t+AhJW8mCktzIjoQqkZ8Rmd0jlF19lmsRYDVTORV81fk7vLeKZkZTAJmk4fqBeoJmR8sa2oCZZ9pxqcPv7/tONAT+W49mjVOdkUbyX/S9y6w5XhvQ6hJDoEHQwsn37dowbNw7jxo0DACxevBjjxo3Dww8/DAAoLy+XAhMAKCgowJdffok1a9ZgzJgxeOaZZ/Dmm2/SWO8ZQt7AGunpedc0je8yTU2zBQ5nL4SvUWGF7DZ/fSPSjpEAm1c5nhk53dD5gXl1LRac+8/1uOZfPwf8/LxEwzenhhIk8BIVz4zwKSpvwUhZQxt+t3w7bl62LeDDBQkh0SXonpGZM2f6/Q+Kt+2qM2fOxK5du4J9KdIHJMZqoBDEdH9P7xjpqkAyI7xEk6rXSh/W3qQZtahoakdVk5/MSA3PjAQXjCTHaWDUqdDUbkNxbQuGZhh93vdQeRParQ4cqWxGfYsloOzUXmdmZPawNHx7oBL7y4Ifna/y6Kvxlxnhh/7VNJtRWteGvGT/Z/QQQqIPzeCRLhFXwou/Hff0ib1dFUjPSEUnkzQc7xup9lGmaWi1oM55knGwPSOCIEjZkeNV/vtGePYFAI5Udt77YbM7sP+0GHzcMCUfAHC00tRpBsYTL9PwMWeeGfG2El4esB0oi4xSTUOrBT8eqYbDQZkaQiIBBSOky3iqPjkusiegvB2y56myk+ZVTjqfxkdmhAcJGUYd4rycb9MZ3mfS2Vr4YnkwUtV5j8mx6ma0We2I0yhx9sAUJMVpYHMwFFYE18TqKtOI75O3M4M4+em++yMkGHn0iwO46a2tWHe4KtyXQggBBSOkG9wzezCunpCDswf1zE6M7hLIBtbOVsFz0uKzZu8bTOVjvaEYkBZYE2uRLBg5GkBmZG+pGAyMyomHUiFgRJZYAgo2SODTNDwo4ycme8uMVMiDkdORsU2Zv2/7qKmWkIgQkUvPSN9y/vB0nD88vfM7hlkgPSOuQ/ICy4z4amAtCnGsl5MyIzX+yzRFQZZp+CTNmJwEAMDI7HhsPFoTVJAg374aSM+IPHu0/3RjRJy5U9MsltB4PwshJLwoM0KiBi8ltFjssPnYD+LKjPgvOaVKh+X5KNMEeUCep4E8M1LV7LNh3Gp3oLSuVfr70crOyzQ8GBntDEZGZccDCK6XQ759NdnZMOsKRjruGZGXaWpbLFJWJZx4P09xJ8EeIaR3UDBCooZB50oEmnyM91Z0W2YktEkaLi8pDkqFgBaL3WfAc6q+DTYHg0Yl/t+4tsWCWj97T9qtdhx2LjgbnSMGISOzxP89XG6CxeZ7gZtcpWziSOGcOPI32lvp3NbalVHi7tRmsaPN2bBbVNNC48aERAAKRkjUUCsViNUoAXjvG2GMdbp9leM9I1Umc4cPM4eDuYKREHtGNCoF8pLEEdjjPhpTeSmof0occhLFoxWO+MmOHCpvgs3BkBSnke6fmxQDg04Fi92Bo1WBNbF67hgBfJdpGGNSJmRSv0QA4W9i5SUmQNw5U0+nDRMSdhSMkKjib6LGZLah1SL+xtzpaK8zGLHYHB2WqJ1uaIPZ5oBaKSA7wfv5S4GQNrH6KCUU1Yglmv6pcRicLp6c6y+g4MvORufESz0bgiBI2ZEDAfaNSGO9soDNtQ7e/YO9odUqZVzOHZIGIPxNrLxEwxX1cKmmzWLHsk1FKKH+FEJ8omCERBV/EzW8edWgU3V6cJxOrZQO3qs2uU/U8A+3/OQ4qJSh/1+MZ1U6y4wUpMRhULrYm+KvidWzX4Qb5SzZBDpZwjMjabLtu/HOMo3Z5nDbWcJLNElxGozPFzMj4d41UusRjPR038iDn+3DY/87iD9/srdHX4eQvoyCERJV/E3UBFqi4VIN3k/v5eO4oU7ScANS/U/U8KCnX3IcBqeJmRF/ZRqeGRnjDD64YMd7PbevAoBBq5J6QuSlGt6Dk2bQYlimEYIAlDe2+z3Tp6fVNnsEIz2Ysfh892l8uvM0AGBrUZ1bwzEhEYMxwNoW1kugYIREFX9bWAPdvsqlOSdqPJtYd55sAAAMz/S9xj0QPNtxsKzJa5NlsbcyTaXJ632bzTYpSPLMjIx0TtQcKm/yOWUk57l9FRDLPTxTJA9G+FhvRrwOeq1KCtAOhLCCvrvUtbj/8+qpMk1pXSv+umo/ACBGLfYqfbbrdI+8FiEhYwz44e/AWxcCbfVhuwwKRkhU4R+Y3npGAt2+ynmbqGGMYcuJWgDAlP5JXbrWEVnx0CgVqGk246THb9TtVjtON4i/yRSk6DEwTQ9BAOpbrdIODbl9pxrBGJCdEON2uCEAFCTHIU6jRLvV0eleE6Dj9lUuIbbj4jP+nqY7AzfenxLOiRpepuENwj2RGbHaHbjrg10wmW2YmJ+Ixy4bAQBYtes0Te+QyMEDkR//AZTvBo58F7ZLoWCERBW/mZEAt69yaV7KNMW1ragymaFRKjA+L7FL16pTK6UR3G3F7r+x8A9Qo06FxFg1YjRK5CaKH67eNrG6+kXiO9ymUAgY4QwS9p3qPEjg0zFpRvegxttK+AqppCPed2S2mC0KZ99InTNYm+DsYSmuae32AOHFtUex62QDDDoVnr9uLC4anQmdWoETNS3YXdrQra9FSEgYA9YvFQMRAJjzd2DMtWG7HApGSFTx2zMS4I4RzltmhGdFxuYmQOdMzXfFBOc47PbiOrfv86bLglS9NBkz2FnWOeql4XWvj+ZVbkR2YH0j8u2rvEzFJUgTNa7MDA9c0uM9MyPhLNOI1zc2NwGCIJawvGWTQvXLiVq8/MMxAMDfrxyFnMRY6LUqzBmRAUDMjhASduuXAhv+n/jnOX8Hpi4M6+XQOngSVVzTNB3LNKfqxbJHdmJg47g8M1Alm6bZUiQGDV0t0XCT8pPwb5zANo9ghJdTCpJjpe8NSjfg+0NVXidq9pR6b17lAh3v9bZ9lfO2a4S/N7xMwzMwJ+taYDp1CAZl7+/4MDQewnDBhEEsFjMMFagytaPyyDakZnWtx4f7z8fbMQxtOH94Oi5NqwHKawAAN/ZrxNE9xSjcfQrWCXaoleFdiU+i2MHPgI3PiH++4MmwByIABSMkyvjKjDDGcNoZjOQGGIyk6t0bWN36RQqSu+V6eSnheHULapvNSNaLAVCRc918QYpr3byUGfGYqDlc0YTTDW1QKgSM9BGM8PHeA2WNMLVbYXC+T568bV/lvAUjlR6TN/GxavRL1GBx8zMwvLnZ78/eU54HAC2ANcA0/uf/dd/zv8Gf8wSAf7u+PxHAV1oADMCb3fd6hITsgr8B0xaF+yoAUDBCooyvnpHGNitMZjFbkpMY2+Fx3vDMCA9GSuvaUNbYDpVCwPj8hG653sQ4DQal6XG0qhk7SupxgTPVz3tGCmQbXgfx8d4qk9thdMs3FQMA5oxIl4IxT/1T4hCjVqLFYseEJ77H9IHJuHBkBmYPS5cCIMD79lXOcyW8ze6Q3pt0ftaP3YZnlC9hgnIzHIISCn1a8G9KF1U2tYMBSNFr0WoRF93FaVRuxwWEqt1qR0ObFSqFgBR9x/eoqd2KVosdOpVCavglpNcpNcC0O4HJC8J9JRIKRkhU8bWBlZdoUvTagHs9Up0fNvXOLaO/FIlZkdE58Z0uTQvGxH5JOFrVjO2yYKRIKtO4gpGBaXooBHGapbrZjDSDDnUtFqlH4ZbpBT5fQ6VU4PHLR+C1DcdxoroFPxRW44fCaiiEffjtWfl48OJh0KqUXrevcp5bWGtbLHBIJR0tYLcBn96GCc0bYGFKLM9+Ar9f0Lvp4XarHVMe+gYAsOf+C/Dx9lL87ctDuGhwBl6dN6HLz//iN4fx6vrjuG5SLp769egOt5eWNeLiF3+Cxq7AtntnS+8ZIdGOGlhJVPG1gfVUvTg6mxNgiQYQMwG87l/TbMaWE7xfpHtKNBw/04X3jTS1u8Z3+6W4sjg6tVIaV+Wlmg+2nYTZ5sDIbCMm5vuf7rlmYi7W/Wkmvl98Du69YDBGZhvhYMDbm0tw3eu/oKKx3ev2Vc6zTCMv6SiZHfj0NuDAKjgUavzReg8+bBoR2hvSBbx5Va0U96LwvSd8tX5XuVbuJ3i9fXimEUPSDbDYHPhqX3m3vCYhZwIKRkhU8dUzwjMjwQQjgiBI2ZEqkxlbini/SPc0r3KT+onPt/90I9osdmmSJtWg7dDbMSidb2I1wWp34J3NJQCAm6cVSGWbzgxMM2DReYOw+s5fYfktkxAfo8aukw245KWN2Hi0GoD3iSMpM+J8b/l0UqZRDaz6PXBgFaBQw3TZW1jrmIATNS1oMXs/PTlQhyuaMOax7/DGjycCuj8PRhJjNRAEAf2cwUhJbcfTey02B7YW1cHhCGzslzEmm1ry3psjCAKuHJ8NAFi1s+9P1dQ2m7HpWA3tTiFdRsEIiSq8Z6TFYnfbNuoKRgLrF+FSnR/Ke0obcKpebBKd2K97g5GcxBikG7Ww2hn2nGrwWqLhBktn1DTjuwOVKG9sR4peg0vHZIb02jOHpOF/i87GsEwjapot2OXcLuu9Z0TsgeCBXqXJDCXseKDtWWD/SkChBn7zNuLHXoYMow6MicFEV3yxuwyNbVZ8GWCWga+hT3JOAuUmxkIhAK0We4dNun//6hB+8+/N+HhHaUDPXVzbiqZ2G7QqBYZkGHze7/KxWRAEYGtxnZSR66se/d9BzHtzCz7YFth7RIgvFIyQqCJvUjTJxnv5h0JuUnCn7PLMyOq9ZQDE1ep6bfe2YgmCK8DZXlznCka8nH0jXwu/bFMRAOCGKfnQqkLfeZKXHItP/zgNV4zNkr7nLzPCyzTVjc14Tv0qJrf84AxE/gsMvQiAa/lZIEvW/OHBUaAf6jwzwptLNSqFFIDK18K3W+1YufMUAGB7cWArsvc4l5mNyDJC7eeAxMz4GIzKDu++lSOVpg67a0LBD3F8ce1RtwMSCQkWBSMkqqiVCsRqxA9med9IaV1omRE+UcM3pJ7VzSUabpKz32N7Sb0rGEntGIzwiZo9pxqwvaQeaqWA307J6/Lrx2iUeO7asfh/vx6FeVPyMHVAx74Y+TQNs1tx7sGHcZlyM+yCyhmIXCzdd5jz3J5CP6cMd8buYNJm2ZpmC1otnZd8eDCSJNuRwks18rXwaw9VScHqsWrfhw/K+ToV2Rve2xOOzIjF5sB1r/+Ca1//BSVdXIXP38/yxnZ8sPVkd1weiVI0TUOijlGnRqvFLk3UMMZCamAFXJkRrruWnXnimZEdJfXIdy4668fLNA47YBE/MPsbHIgXWuGwAzoAFw/PQJrGDLR3/ZRcAcC1o+Jx7ah4wNYMeHz2xwsOGNAKweGAbeXvMa7xe1iZElsnPovpskAEcGVwCitCD0aOVJrQanH9Nl5a1+a3PAK4zqWRByMFybH4Ee5NrKt2nZL+fKyq2W1U2hfpVORc7/0icjzo5eXB3rStuE4KItYcrMRtv+of0vMwxqRtvADwyvrjuG5yXrdsHibRh4IREnWMMSpUNLkyIw2tVrQ4P9SyE4ILRuTnsygEdHu/CDc0w4A4jRKmdpuU2u+fGgdU7Afevx5oFH8r1QHYI4+PjgJ4qkcuqQMdgH28enMQsEGJO6x348bBF3W479AM3mgb2Ae9NztPupdPSutaOw1G+Lk0yd4yI86MU22zGesLq6XbTe02VJvMXseZOavdIZ23E0hmhAe94QhG1h2ukv689lBVyMGIyWyD1S42rmbG61De2I53fykJ+flIdKMyDYk6nhM1/AMh1RD4jhFOnhkZnmX0uVSsq1RKBcbLRnMFAci3FgFvXyYFIpHEpk3An4XFWOOY6LW/pF9KHNRKAc1mm3T6cLB4vwjnebKxN1JmRO+7TPO/PWWwORhGZcdLWahjXs77kTtSaUK71QGDTuW1sdiTKxjp/TKNPBjZVlzntjE3GLXOwC5Oo8Q9swcBAF5bf7zLE1IkOlFmhEQdzy2sUvNqkCUawH35V3etgPdlYn4SNh4Vzzk521AF7Yo7gdZaIGscMO8TQCv2Yaw9VIm/fXUQT101usevydPcF37EseoWvP6byVj19k4AQIaXYEStVGBAqh6HK0w4Umny2qtT02zGS2uP4pbpBVLAILfLmRkZmKbHsapmlAbwwc7LCslxriCSBw/FtS1wOJi0JO6q8dn46WgNSmpbcby6GdMGpvh8Xtd+kfgOa/K9kZdpQs0MheJEdTOKalqgVgrIjI/BybpW/HikGpeOyer8wR7q+Hup1+Kq8Tl4df1xlNS24u3NJfjjzAHdfenkDEeZERJ1jM6JGt4zEupYL+A6uRcAzurmZWee+PKzIcJJvGJ7xBWI3LgKiEsBVBpApcGsUbn44b45mDIoU/peb33p42JhhQpHqsXAQKtSSIvmPLn6RrxnHd748QT+u7kEj68+2OG2xlYrjjvP57nM+UHKm5D94b0SybLMSE5iDFQKAe1WBzafqMWeU41QKgRcOiYLA9PEUenOMiOdnYrsiWdGms22kDMToeBZkSkFyZg7Stzmu/ZQZUjPxTMjSXEaqJUK3HWemB3594/HYWrvvZ+JnBkoGCFRxzMzUhpi8yoglmniNEpoVQpM7u5+EYcd2P0+sP4pYP1TmFjyOharP8EKzZMwOhqBzLFiIBLjf7Nqb+LjvXxKJt2o8/lbP+/vKPSxa4T3hGw8Wo3GVvcPt12l4m39kmMxJjcBgNgz0pm65o4NrCqlArnO6ZZn1xwBAMwYnIoUvRYDeDDSyURNZ6cie9KplVIg25t9I2sPicHIeUPTMGtoOgBg/ZFqt507gaqVxqTF9/LysVnonxqHhlardB4SIYGiYIREHV89I6FkRjQqBd5bcBbe//1ZiI/txn4Rhx347A7gs9uB9UuB9Uuh2fj/cJfyUyQLJtQYhgM3fRZRgQgAxMeIH0xHnMGItxINN4RnRio7ftBb7Q6p9GG1M3x7oMLtdt4vMi4vUSqvlda3+t0EarbZpcMQ5Q2sgBjUAOK0EgBcOU7ckhpIZqTNYpeCLx4YBaK3+0Ya26zSkQKzhqVhfF4CEmLVaGi1YqdH/00gPMekVUoF7p4lZkeW/VwcUoDTE975pQQXPLch5N4k0jsoGCFRx3U+DS/ThJ4ZAYCxuQkYn9eNQQEPRPZ+AAhKYNxvgYm3AhNvRfmgefgy6Saob/k84gIRwJUZ4WfjpHnZ1MrxzMjxquYOH1yHy00w21zf+59zqRy3y7lgbFxeArITYyA4t6jyD0hv6lvE4FOpEDo0GufLmk4NWhXOHy5mDQakisFIZZO5w3lG3MHyRtgdDKkGrd/gyxMPfgMpL3WHjUerYXMwDEiNQ35yHFRKBWYOTgUArD0cfKnGtc3W9c/44lGZSIxVo67Fgi1FXV+q1h1W7jiFI5XN+EHWuEsiDwUjJOrIMyPijpHgz6XpMQ478PlCVyBy9VvA5a8AlzwLXPIsMue9iovvegnxSWnhvlKv+OIzHkh4m6ThshNiEKtRwmJ3uC0cA1xlmAHOxW4/H69FrfPDz+Fg2O0s4YzPS4RWpZSCAH8TNfJV8J5NpvJttheNypSmquJj1FI55biP7Ii8RBNMI2pvZ0Z4v8isYenS9/ifefkmGHUeZRpAzI5cMFzsRfl6f2QcBMjX/Hd1wRvpWRSMkKgj7xmpb7VKi7Oywx2M8EBkz/uuQGTEFeG9piDxzAjnL1OgUAg+m1h5GeaS0VkYnRMPu4Ph6/1iqeZETTOa2m3QqV1nwOTyLIOf/gupedWjRAPAbVqHH2THDXRmR3jDrKdgm1e53B5YfMYYw8L3duIP72x3G7G1O5i0O+W8oa5A9pzBqVAqBByrag76w7rWS/8NAFzobIz99kBlwIcM9hTGmBSMdNfJzKRnUDBCoo4rM2KTfitNN2q7dH5L0Eo2Ay9PAp7Kk33lywKR//S5QARwZUY4f2UaQN434r6JdbesDHPJaPGQv//tEUs1vL9hdHaCdAZMjvNMIX9NrN5WwXMjs4zQa1UYmmHo0IjcWd/IHmnzaoLP1/amJxaflda14ct95fj2QCX+8M4O6byY3aUNqGuxwKhTYYJsX018jFqa0go2O1IrTSa5/zOePiAFBp0K1SYzdpwM7FyfntLUZoPFWQKkzEhko2CERB1Xz4g15DNpuqRkM/Dur4GaI0B7o+vLYgKUWmcgcmXvXU83MnpkRvyVaQBgMN/EKlsLX99ikc7fGZubgItHi6O7W4vrUNnU7mpezU+QHsPPevEXjHhbBc8l67X44d6Z+Oj2qR1KOP6CkcY2q3Sto7MDm6ThcgJsvA3G8RrXNf50rAZ3vb8LNrsD65w9ITOGpHU4xG+2s1SzLsieCmnPiMf7qVEppOf8el9Fh8f1purmdunPJXWtYc/UEN8oGCFRR94z0tXm1aDxQMTaAvSfCSzcCiza7vq6t7DPBiIAkBBkMMLXwsszIzwr0j81DgmxGmQnxGBCfiIYA77cWy4tOxuX6/oN31Wm8ZcZ8f7hyaUatF436PJg5LiX8V5+6nBeUiwSfTyvL1nOowdaLXbUt3bPXg7e1zIgNQ4alQLfHazEkk/2ykZ6Uzs8hpdtthTVBrwfhDHmdWcLd+FIXqqp6LZAKxRVJtfZORabA+VN7X7uTcKJghESdfhv7y0WO4prezEYKdkMvHe1KxC57n0gdQiQMsj1FYETMsHw7BlJ76RMw3tGimtbpJKCt2CDl2o+2l4qjQ2Py0uQbud7Qvw1sLrKNP6vyROfqCmpbYHZZne7jY/Kjg5wv4icTq2U3p/uamI94czSzB2ZiVduGA+lQsCnu07jcIUJCgGYMbhj43P/VD36p8TBamfSht/ONLW7zqXxlmmaMTgVsRolTje0SSPa4VAtC0YAoKSGSjWRioIREnUMOtdG0EPl4sKt3O4s09itQOHX4sIy/rX1DTEQsTQDBTPEQETTi6WhXiIPRgw6FWI1/k+cSNFrkBSnAWOuMggf2x0rCzYuHpUJQQAOV5jgYOIkjjzrwss0ZQ3tPvdb1DT7/k3en3SjFnqtCg4GFMuaIBljUh/LuUNCm27q7tN7eWakf2oczh+ejmeuGQM+4DM+L9Fr4AC4siMrd5wKKJPBJ5v0WpXXXiudWolznc/JG4/DwTMY4b98ePpmfwXOf3YD9p8OX+AU7SgYIVFHrVQgViP+B/Swc/tnt/WM2MzABzcA718nLizjX1/d6wpErv/gjAxEAPdgpLMSDQAIgoDB6WLm4XCFyTm22wAAGCdrCE0z6jClwNVYKs+KAECaQQuNSgG7g6G80Xsq3t80TWfXOMBLqWbPqUacqGlBjFoplSWCJfWNBLA9NhA8M8KzOVeMy8aTV4yCQafCjVPzfT7umom5UCkErD1chc92n+70dfyVaLi5I10jvuEq1VQ3e2RGfDSxvrelBEermqUNvKT3UTBCohLvDWi3ir9Fd0uZxmYGPvwtcPQ7QBUDDJgFDJzt+jpr4RkdiADingm9VsyGBLoAbGiGeMDfkUoTTtQ0w2QWx3Z5PwknP8xtnMeSOYVCQE6C/w92f9M0neHjvfIm1lU7TwEALhiRjjhtaGeO+puo2XmyHg9/vh8/H6sJ6MO8qd0qZQL6p7pGlW+Ykoe9j1yAy8dm+3oohmQYpJN3H/7sQKfbSv01A3PnDkmDVqVASW0rDpWbfN6vJ/H3IzNe/HexyEuZhjGGg2XiLyU/FFZ1W2BIgkPBCIlK8sPbBAHITAh8c6ZXnoHIDR8AN34K/Hal6+vCv5/RgQjHsyOdjfVyrl0jJtfYbk4CVB5TH3NHZkLpnHTxzIwArr4RX02svLQQbJkG6DhRY7U78L+94lIvvjo+FK4yTcdrfvjz/Xh7cwlueHMLLn7xJ3y68xQsNt8r1k8496CkGbQweDTiBrKM7fYZAzAuLwEmsw1/+mi338kTvmMk2U//TZxWhXOcG16/CdMCNB6MTHKOa5d4KdNUNpml4IoxYMXWk713gURCwQiJSvKpiXSDrms7Rmxm4MMbnYGITgxE+s/s+kX2UTwYCaRMAwBDMsQP+iOVJtmZMwkd7pcUp8ETl4/EHTMHuJVwuFznrhFvTawWm0Na/x9sAyvg2gTLg5ENhdWoa7EgRa/F2QNTgn4+ztfis9pmM/afFn9bj1ErcbC8CYs/2oOz/986fOOjB0PeLxIKlVKB534zFjFqJX45UYe3NhX5vG9nk0mcq1QTnr4RKRhxlvhK6lo6BFkHy8U+EZUz0P1wW2mHRmXS8ygYIVFJvg+Df4iFxGYGProJOPqtMxD5MKoDEcC1+CzQMs0gZ2akvLEdPx4Rt4TKJ2nkbpiShyUXDvX6m36un7Ne6lvF33wVQsfx40DwzMiJmmY4HAyrdol9FZePzeqQwQmGvEwjL8VsOl4LQBx93vzAeVhy4RCkG7WoMplx78d7vDbpnqjhY736kK+nX0ocHrpkOADgH98WorDCe3mFNwMndZJlmjUsHWqlgKNVzX4PGwxUtckcVP8JD0bG5SZApRDQbnW4jfsCkEo0F47MQIZRh7oWS9j3o0QjCkZIVDLKJmpCbl7lgciRbygQkbl0TBYKUuKkFH1njDo1sp39HrxXwVtmpDN5fso08tXlnkvNAn1ujVKBdqsDhyqasOaQuESsKyUaQCwPCgLQZrVLpQIA+OmoGJSdPTAFCbEa3DFzIDYuOQ8GnQrNZhsOewkSjleJZZr+XQhGAOD6ybk4b2gaLDYH7vlwN6xeAp9Am4HjY9SYNkDMHH1/KPjD+OR+OlqDSU9+j6e+ORzQ/a12B+qcQWhGvE4K/Dz7Rg46J+pG58Tjhil5AMSTfknvomCERCV5ZiSk5lWbGfhovisQuT66SzNy10/Oww/3znQ7fK4zfKIG6Di2G6hcP1tYu9K8CogljH4p4vO/tPYYLDYHBqfrMSLLGNLzcVqVEukG8WflpRrGGH5y7vs4e5CrBKRRKTDWWZ7a5WXNuiszElqZhhMEAU/9ehTiY9Q4VN6EbV5O3w1kmobjZaztxV1bDf/zcfE92XIisNOA61osYEw8pTkxViOdzOw5UcMzI8Mz43HdJHGqaEdJvfR90jsoGCFRSd4zEnQwIgUiX7sCkQHndvMVRpfBssmZsSFkRQBXmaam2YJWi83tttoW14m9oeKlmm8OiCn8K8flBHVKry+8TMibWE/UtKCssR0apQJTCpLd7svPldlR4v7BbncwaQdKV8o0XJpBJzV9ets8y09A9tfAyo13XvOuk/VdGvHlZR5vEzHe8BJNcpwGSoWAfsnivx/yXSOmdqv09+FZRqQZdZgzQuxzeXcLZUd6U2jzaGeK8r2AOTwjZ2GhNQAZo4Bu+A9oj2AMKN8NWHp+tG5w+2lMFsSu+eEWASguDfzBm1+mQKSb8QPzAHhtTg1EfKwaRp0KTe02nKpvk6Z0AHlZIfjmVW6g7ENeEIArxmX5uXfgchJjsa24XsqM8KzIxH6JiNG4N1aPd44086kj7lR9Kyx2B7QqhbRmvqv6p8YBh1y7S+SCyTSNzDZCo1SgtsWCktpWtxOSg8GDkcY2K+pbLJ2u3+fBSKpB/GfOX7dY9vPwcldmvE76WX57Vj6+3FeOz3adxv1zh3o9IsCXqqZ23P/pPtwyvR9+NSiwMiURRXcw8uWfgFNbw30VvWv0tcAVrwGKXjyhNhA2C/DJLcDh1b3ycpcBuIx/Lq0J4QlUOuD69ykQ6SZDZJkRzx0iwchNisWBsiacrG31Gox0JTPCF58BwNT+yciM754Pfc/FZxt5v8igjlM6Y/MSIAjixFC1ySx90PKx3oKUOGn8uat4mc0zEyE/lyZF33lwp1UpMTLbiJ0nG7DzZH1IwYjF5kCJrPxWVNsSfDDiLNMUy8o0rhKNq9x2Vv8kDEzT41hVM1btPI350/oFfJ0f7ziFdYerUFTTgnV/mtEtmbOuOFJpwsvrjqGkrhV5SbHIT4pFXnIs+qfEYVxeYrf9u9IdojsYic8B2sJ7xHWvqjsB7P1Q/HMkBSTyQESpARJ8b4rsLs1mGyqdh2b1T9UjqP9L6ozArEeA/jN65Nqi0YBUvXROS1f6MHITxWDEs4k11FXwntfIdbVxVU4+UWO1O/CLsyfiVwM7/mZt1KkxOM2AwkoTdp6sl0oKvJTSHSUazlcw0tRmg805HpsYF1jWYEJ+InaebMCOknpcNT4n6Gsprm2BXTaSW1zTImWJfOHbV1OdAVO+s0xTUiuekiwIgisYkf07JwgCbjwrH498cQDPrjmCLUW1GJJuxJAMA0ZkGaXeJG/4uUlFNS3YWlSHKf2Tfd63J5XUtuD574/is92nwStje5zHLHC/P6c//nLRsN6/OB+iOxi5Zlm4r6B3Hfwc+OR3kRWQ2K2yQEQLXLcCGDS7x192z7EazHtzC7Lidfj5zlk9/nrEP51aif/deTbAxD+HKi/Z+4F5ge7F8Gdgmh7JcRowIOT1797IF5/tLm1As9mGxFi1z6BsfH6iGIyUyIMRPknTteZVuf7OYKS0rhUWmwMaldhiWON8Lw0+zqXxes15iQCKOvS6BMpzLLg4gL4Rz8xITmIsFM7JpWqTGWlGnTRJI8+MAMCV47Px4tqjqG2x4Kt9FfhKNur71FWjcN3kPK+vKR+F/nBbaa8HI20WO5748iA+2lYqBYxzR2bg4tGZKGtoQ0ltKw6UNWF3aUPAhyL2lugORqLN8MuBq98CPr7FGZAIwBWvhi8gsVuBj2/u9UAEEH9Tmz0sDTNDPOCMdL80Qxe34ALIlUoe7rtGQj2xV06nVmL1XWcDQIcNp10hX3y20blnZdrAFJ8jyOPzEvD+1pPYKZuo6YnMSKpBiziNEi0WO07WtUoNvMFM0kjX7GxiPVJpgqndGvT7x4MRQRBby4p8HHgnV2USM59pzmBEo1IgJzEWJ+taUVQjlnl48DAiy/3UZaNOjXX3zsTu0gYUVjThcIUY/BXXtmLt4SqvwYjN7pDKZQDw5b5yPHLZiA4nWfekf/94HCu2iL1wMwan4t4LhmCUx4nSpxvaMP2pdThaaUK71d6l4L87UTASbYZfLmaEPr4F2PsBUHccMHTfb3lBaTwFlO3q9UAEED9Y3pw/qddej/SOnCTv69UDOUslEN3VJyKXEa+DQgDMNgc+d54CfI6XfhGOT9TsOdUoZSxO9EBmRBAEFKTGYf/pJhTVtEjBiHxnS6DSjTpkJ8TgdEMb9pQ2eu2H8eeoMxgZn5eIHSX1QWZGXEFufrIYjJTUtiI+Vg2L3QGDVuV1oi4+Ro0Zg1Mxw7kv55cTtbju9V98jvyW1IlNxDFqJXKTYnCkshlf7CnDjWf1fNmZ4+PTD8wdij/MGOD1PlnOZt26FgsKK0wYE2LDeHejYCQayQOSU9vCey1hCETImYsvPjtZ5+oLAEL7bb63aFQKZBh1KGtsl85OOdvPJEZBShwSY9Wob7XiQFkj+qfqpVHbri486/haemcw0gwgHYBrTDo5gOZVuQn5iTjd0IadJ+uDDkZ4ZmT2sHQpGJH/8/XGs0wDiE2sG4/WoLi2BSql+NhhmcaAFuENc5ZyTje0obHVivhY94zHEWeWZVC6HpePzcYTqw/iw20ney0YYYxhf5m42n66nyMKBEHAiCwjNh6twf6yRgpGSJgNvxz4w49A6ZbwXke/s4HUIeG9BnLG4JtcWy121LVYkKzXwmp3oKHVCqBrPSM9KScxFmWNzobqlDjp5/BGEASMz0vE2sNVbiO+6UatdGJyd/HWxFrXHNj2VU8T8hPxxZ6yoPtG7A6GE84y1HlD0/CPbw/DZLahptniFmh48haMyJtY+aGDwwNsmI6PUSMnMQan6ttwsLwJUwe494McqRSvcVCaAVeOy8b/+/ow9p9uwv7TjRiZHe/tKbvV6YY2NLRaoVYKGJTuPygdmR0vBiOnI2exGwUj0SxjpPhFyBlCp1Yi3ahFZZMZ//fRHvx2Sp5UMxcEICE2QoORpBhsLRb/HEjWYHy+MxgpqZd6ErqzX4TjTazyXohQS16uHSn1cDiYWzZiy4laPPLFAfz14uEdfv7T9W0wO8tRA9P0yIoXyz3FtS0+g5EWsw0tFvGwO8/MCCAGV/y8Is/mVX+GZxpxqr4NB8oaOwYjVWJmZEiGHklxGlwwIh2r95bjw22lQQcjtc1m/Obfm5Gs1+KfV4+RGrP94YHF4HRDp43FI509MgecmZRIQBtYCSFnlMvHimO3Px6pxu/f2YELnvsRAJAYq4movQpy8vORAjkFWP7BzrMG3dkvwnnLjNRKJa/gyjRDMw2IUSthare5bXW1Oxge/Gw/DleY8J+fTnR43LFq8UO+v3OHCv85/W1i5WWrGLUScbLFcXzHSUlti2uSJohRcn5f/li5o5W8TCPut7l2Ui4A4LPdp9FuDe4U4E92nMLxanE8+OKXNuLrfeWdPoYHFiOzOg98RjmDo8PlJq9nD4UDBSOEkDPKXy4ahu8Xz8Dvz+mPFL0GpnZxNXxqkB+evYk3UCoVAs4a0Pk46JjceCgVAsob27HpmDii2ROZEf7hXWUyo9ksvo+hjkmrlQqMyRU/BOWlms92nZZ6QrYU1UnlE+6os/zBl85Jy8v8BCPyEo28ryQ3KQaCALRY7GhotUKl6LykIcenbjybWC021yQNX7Y3fUAKshNiYGq34ev9nQcTHGMMH20vla7f1G7DH9/biYc/3+83qNl/2hmMZHceXOUmxcCgU8Fid0i7UcItpGDklVdeQb9+/aDT6TBlyhRs3ep7i+ny5cshCILbl07X9RE+QgjxZWCaHn+5aBg2PzAL//rtBFwzIQdLLozc3qTxeQlQKgScOyQtoPXjsRoVhmWKH3p7TokfQt3dvAqIfRIpzqZf/uFf24UFcvKMDiB+iD/3/RHp9laLvcMhgDxQGcSDkZSOm1Q9eesXAcRtsFmyiaiBafqAd6UArszIsapmmG2uwKC4tgU2B4Neq0JWvPj5plAIUnbkg62BHzexq7QBx6tboFMrsOb/zsHtzqmYtzeX4Nev/Sw1Y8sxxrDPWaYJpCQkCIKrVBMhfSNBByMffvghFi9ejEceeQQ7d+7EmDFjMGfOHFRVVfl8jNFoRHl5ufRVUkIHEBFCep5aqcCFIzPw9DVjMGtYergvx6eBaQZs+PNMvHDd2IAf47mBtKun9frCSzX8jJqujEl7HvT34baTOFXfhlSDFucPF//5/HTMfRnXMWdJh48WFzhPTy6q8b1rxHP7qhw/fRkIrkQDiGOx8TFq2BxMytgArs2rg9L1bpmYqyfkQCGIGR/P04J9+Xj7KQDARSMzkRCrwf1zh2LZLZOQFKfBgbImr6WsKpMZNc1mKBWCNPXTGZ5B2R8hfSNBByPPPvssFixYgFtuuQXDhw/Hv/71L8TGxuKtt97y+RhBEJCRkSF9padH7n8UCCEkHHISYxEXxDQM/2AHAJ1a4fYbf3eS+kaqW+BwBHcujSd+7tDx6haUN7bhxXXHAAB3njfQazDCGJMyIwM9yjQltS0+TwHmmZE0Y8drzE92BW3BNK8C4mcZf4y8VMPHegenGdzun5UQg7OcW1jXHfb9CzvXZrHjf85dM1dPdK3NP3dIGh66ZJjzeao7PI6XaAam6gNeYsYzKPyx4RZUMGKxWLBjxw7Mnu3aCaFQKDB79mxs3rzZ5+Oam5uRn5+P3NxcXH755Thw4IDf1zGbzWhqanL7IoQQ4iLPjBSk6APalRGKghQxCCiqaUZTu1U6IyYxhMmkpDiNNKGz+MM9qDaZkZMYg+sm5UmNu3tKG9DYJo5iV5vMMLXboBBcQVFuUiyUCgGtFjuqnEGHp6om35mRAnkwEsI5SN6aWPlY7+AMQ4f7zxwi7ozZcKRjEOHpmwPlaDbbkJsUg7MK3HuHZgxOg0IADpU3oazBfcMwn6QZEUC/CMeDkYPlTW7n/oRLUMFITU0N7HZ7h8xGeno6KioqvD5myJAheOutt/D555/j3XffhcPhwLRp03Dq1Cmfr7N06VLEx8dLX7m5ucFcJiGEnPFyEmOknoiemKTh5BM1vERj0Kmks2qCxVfDbz5RCwC4Z/ZgaFQKZCXEoH9qHBwM2HxcvI1vXs1PjpN6O9RKhdTw62uiRirTeBn9zZeNyY7IDH7/h9fMiHOsd7CXZtgZg8UjJ345UdvpVM1H28TPxavH53YILpPiNFJm6YdC9yzL/iAmabiC5DjEaZRotzrcppvCpcenaaZOnYqbbroJY8eOxYwZM/Dpp58iNTUV//73v30+5oEHHkBjY6P0VVoaePMPIYREA0EQMMH54cSbO3sCD3ROVLdIzauhlGg4eUZnYJre7QRknh3hE0K8ROM5KSTfF+KNrwZWABiRLU4iDc0wdNiiGgiefThY3gSHg6Hdapc25/JJGrnB6XpkGHVotzqwtajO5/OW1rVi84laCALw6wneT4U+b6gY2Kw75B6MHJAmaQIPRhQKQcryREKpJqhgJCUlBUqlEpWVlW7fr6ysREZGYOebqNVqjBs3DseOHfN5H61WC6PR6PZFCCHE3Z8vHIIbz8rH/Kn9euw18pJiIQiAyWxDobNRsytn/Mh7Xf50/mC33S88GPnJIxgZ6BFs8WyNr/Fef8FIdkIMVt95Nt6+dXJI1z8gVQ+NUoFmsw2n6ttworoFdgdDfIxaOpRPThAE6Xwbf6WaT3aIWZHpA1Lc9s7I8WBk0/EaKctS22yWtvcGW3bio8qRsIk1qGBEo9FgwoQJWLt2rfQ9h8OBtWvXYurUqQE9h91ux759+5CZmRnclRJCCHEzIFWPJ64YicQeXHOvUyul9fTbi8Xf7LuyVn9Qmh7XTMjBdZNyceFI919izxqQDKVCQFFNC07Vt3YajHjLjDgcTFp65mtD67BMY8inRKuVCgzOEK/nQFkjjspKNL7OypnRSd+Iw8GkYOQaWeOqp6EZBmTGi1kWXso64CwX9U+JC/o4AL78LBImaoIu0yxevBhvvPEG/vvf/+LQoUP44x//iJaWFtxyyy0AgJtuugkPPPCAdP/HH38c3333HU6cOIGdO3fit7/9LUpKSnDbbbd1309BCCGkx/APf34qbFcOHFQoBDx9zRg89evRHT68jTo1xjjX9286ViON9XqWofztGmlos8LmbMhMjuuZRXdS30h5k2yst2OJhps+MAVKhYBjVc0dTpQGxP6Z0w1tMOhUmDPCd5VBEARXqcY5ncMDiREhnH8jNbGWiSWncAo6GLn22mvxz3/+Ew8//DDGjh2L3bt345tvvpGaWk+ePInycte2ufr6eixYsADDhg3DRRddhKamJvz8888YPnx49/0UhBBCegyfgDntnOLoqQ95wFWqWb23XCq3DPDMjEjjva0dPkT5YxJj1SE32XZG3sRaWCEGTEP8BCPxMWqMc56O6y07smLrSQDAZWOyOh3NlQcjjDHX5tUQJoMGpMZBqxJLTv6WyPWGkP5JLVq0CCUlJTCbzdiyZQumTJki3bZ+/XosX75c+vtzzz0n3beiogJffvklxo0b1+ULJ4QQ0jt4ZoTrSs9IZ84eJJY0Nh4V+0Yy43Udyg9ZCTqolQLMNgfKm9rdbvPXL9JdRsjGYnmZprO18tKIb6F7MLL3VAO+3Cv+An/DlLxOX3vagBRoVQqcbmjDkcpmqd8jlJOBVUqFtCRtf1l4+0bobBpCCCF+FXhMs3SlTNOZsbkJiJUdbufZLwKIH6K5SWKTp2cTa3WzGJyE2hMSiKHOfSLlje1+J2nk+Ijvz8drpfN3GGP42+pDAICrxmVLDaX+xGiUmOY8v2jVrtM4WSe+/ogQMiOAaxPrgTBP1FAwQgghxK/+HpmRnizTaFQKaWsp4D0YAVylGs8mVmnhWQ9mRgw6tdu+kuQ4TafjziOyjEiO06DZbJPO5vn2QAW2FtdBq1Lg3jmBn510nvNog3c2FwMQD75LCGEJHRA5TawUjBBCCPErKyEGGqXr46InyzSAq28E8B2M9PMxUdMbZRrAfZV8ICf/KhQCzpGN+FpsDiz9+jAA4Pfn9EdWQuDr/HnfSItFHO8NZtmZJ/l4r6/1+r2BghFCCCF+KRWCWyYgpQfLNABw9iBZMOLjNOJ+PnaN+DskrzvJg5HOSjSctG+ksBpvby5GSW0rUg1a6WTeQGUnxEilIiC0fhFucLoBaqWAxjYrTtW3df6AHkLBCCGEkE7Jm1h7cq8JII7yjs6JR5pB63NkVSrT1IYpM5IVfDDyq0EpEASx8fWF748CAO69YHBQByRy5zqzI0Do/SKAWBYb4gxs+F6XcAj+HSCEEBJ1+qfqAVTCqFNBrezZ32MFQcBHf5gKxsSGTW8KnGvqS+taYbM7oHJeU28FI/Jm00CDkWS9FqOy47H3VCNMZhuGZhhw9YTQzl6bNTQNr60/3uFaQvHideOQYtDCqAt+PX53oWCEEEJIp3gTa1fOpQlGZ/s2Mo06aFUKmG0OHCo3YZRzWZq/Q/K6U7pRizE58ahsMgeVmZg5OBV7T4nNon+9eLjbOvxgjMtLxFXjsmGMUXf5Z+3voxTWm6hMQwghpFNT+idBq1JgckFSuC8FgNgQOta5SOy61zfj4+2lMNvsaGi1Auj5nhFBEPDhH6bih3tnBlVmuWxsFjQqBS4bk+XWGxMspULAs9eOxaOXjQj5OSKJwMLZPhugpqYmxMfHo7GxkQ7NI4SQMGm12BCjVvo8g6W3nW5ow/99sBtbnWfmTB+YjE3HaqFWCjjyt7kRc52e2q12aJQKKELMivQlgX5+U2aEEEJIQGI1qoj6gM9OiMH7vz8LSy4cApVCwKZj4uFxqXptRF2nJ51aGRWBSDAoGCGEENJnKRUC7pg5EJ8tnI4BzqbWnMTYTh5FIg01sBJCCOnzRmbHY/Wdv8InO0oxuSC58weQiELBCCGEkDNCjEaJG6f2C/dlkBBQmYYQQgghYUXBCCGEEELCioIRQgghhIQVBSOEEEIICSsKRgghhBASVhSMEEIIISSsKBghhBBCSFhRMEIIIYSQsKJghBBCCCFhRcEIIYQQQsKKghFCCCGEhBUFI4QQQggJKwpGCCGEEBJWfeLUXsYYAKCpqSnMV0IIIYSQQPHPbf457kufCEZMJhMAIDc3N8xXQgghhJBgmUwmxMfH+7xdYJ2FKxHA4XCgrKwMBoMBgiB02/M2NTUhNzcXpaWlMBqN3fa8pCN6r3sPvde9i97v3kPvde/prveaMQaTyYSsrCwoFL47Q/pEZkShUCAnJ6fHnt9oNNK/2L2E3uveQ+9176L3u/fQe917uuO99pcR4aiBlRBCCCFhRcEIIYQQQsIqqoMRrVaLRx55BFqtNtyXcsaj97r30Hvdu+j97j30Xvee3n6v+0QDKyGEEELOXFGdGSGEEEJI+FEwQgghhJCwomCEEEIIIWFFwQghhBBCwiqqg5FXXnkF/fr1g06nw5QpU7B169ZwX1Kft3TpUkyaNAkGgwFpaWm44oorUFhY6Haf9vZ2LFy4EMnJydDr9fj1r3+NysrKMF3xmeGpp56CIAi45557pO/R+9y9Tp8+jd/+9rdITk5GTEwMRo0ahe3bt0u3M8bw8MMPIzMzEzExMZg9ezaOHj0axivum+x2Ox566CEUFBQgJiYGAwYMwBNPPOF2tgm916H58ccfcemllyIrKwuCIOCzzz5zuz2Q97Wurg7z5s2D0WhEQkICbr31VjQ3N3f94liU+uCDD5hGo2FvvfUWO3DgAFuwYAFLSEhglZWV4b60Pm3OnDls2bJlbP/+/Wz37t3soosuYnl5eay5uVm6z+23385yc3PZ2rVr2fbt29lZZ53Fpk2bFsar7tu2bt3K+vXrx0aPHs3uvvtu6fv0Pnefuro6lp+fz26++Wa2ZcsWduLECfbtt9+yY8eOSfd56qmnWHx8PPvss8/Ynj172GWXXcYKCgpYW1tbGK+873nyySdZcnIyW716NSsqKmIff/wx0+v17IUXXpDuQ+91aL766iv24IMPsk8//ZQBYKtWrXK7PZD39cILL2Rjxoxhv/zyC9u4cSMbOHAgu/7667t8bVEbjEyePJktXLhQ+rvdbmdZWVls6dKlYbyqM09VVRUDwDZs2MAYY6yhoYGp1Wr28ccfS/c5dOgQA8A2b94crsvss0wmExs0aBBbs2YNmzFjhhSM0Pvcve677z529tln+7zd4XCwjIwM9vTTT0vfa2hoYFqtlr3//vu9cYlnjIsvvpj97ne/c/veVVddxebNm8cYo/e6u3gGI4G8rwcPHmQA2LZt26T7fP3110wQBHb69OkuXU9UlmksFgt27NiB2bNnS99TKBSYPXs2Nm/eHMYrO/M0NjYCAJKSkgAAO3bsgNVqdXvvhw4diry8PHrvQ7Bw4UJcfPHFbu8nQO9zd/viiy8wceJEXHPNNUhLS8O4cePwxhtvSLcXFRWhoqLC7f2Oj4/HlClT6P0O0rRp07B27VocOXIEALBnzx789NNPmDt3LgB6r3tKIO/r5s2bkZCQgIkTJ0r3mT17NhQKBbZs2dKl1+8TB+V1t5qaGtjtdqSnp7t9Pz09HYcPHw7TVZ15HA4H7rnnHkyfPh0jR44EAFRUVECj0SAhIcHtvunp6aioqAjDVfZdH3zwAXbu3Ilt27Z1uI3e5+514sQJvPbaa1i8eDH+8pe/YNu2bbjrrrug0Wgwf/586T319t8Uer+Dc//996OpqQlDhw6FUqmE3W7Hk08+iXnz5gEAvdc9JJD3taKiAmlpaW63q1QqJCUldfm9j8pghPSOhQsXYv/+/fjpp5/CfSlnnNLSUtx9991Ys2YNdDpduC/njOdwODBx4kT8/e9/BwCMGzcO+/fvx7/+9S/Mnz8/zFd3Zvnoo4/w3nvvYcWKFRgxYgR2796Ne+65B1lZWfRen8GiskyTkpICpVLZYbKgsrISGRkZYbqqM8uiRYuwevVq/PDDD8jJyZG+n5GRAYvFgoaGBrf703sfnB07dqCqqgrjx4+HSqWCSqXChg0b8OKLL0KlUiE9PZ3e526UmZmJ4cOHu31v2LBhOHnyJABI7yn9N6Xr/vznP+P+++/Hddddh1GjRuHGG2/E//3f/2Hp0qUA6L3uKYG8rxkZGaiqqnK73Wazoa6ursvvfVQGIxqNBhMmTMDatWul7zkcDqxduxZTp04N45X1fYwxLFq0CKtWrcK6detQUFDgdvuECROgVqvd3vvCwkKcPHmS3vsgzJo1C/v27cPu3bulr4kTJ2LevHnSn+l97j7Tp0/vMKJ+5MgR5OfnAwAKCgqQkZHh9n43NTVhy5Yt9H4HqbW1FQqF+0eTUqmEw+EAQO91TwnkfZ06dSoaGhqwY8cO6T7r1q2Dw+HAlClTunYBXWp/7cM++OADptVq2fLly9nBgwfZ73//e5aQkMAqKirCfWl92h//+EcWHx/P1q9fz8rLy6Wv1tZW6T633347y8vLY+vWrWPbt29nU6dOZVOnTg3jVZ8Z5NM0jNH73J22bt3KVCoVe/LJJ9nRo0fZe++9x2JjY9m7774r3eepp55iCQkJ7PPPP2d79+5ll19+OY2bhmD+/PksOztbGu399NNPWUpKCluyZIl0H3qvQ2MymdiuXbvYrl27GAD27LPPsl27drGSkhLGWGDv64UXXsjGjRvHtmzZwn766Sc2aNAgGu3tqpdeeonl5eUxjUbDJk+ezH755ZdwX1KfB8Dr17Jly6T7tLW1sTvuuIMlJiay2NhYduWVV7Ly8vLwXfQZwjMYofe5e/3vf/9jI0eOZFqtlg0dOpS9/vrrbrc7HA720EMPsfT0dKbVatmsWbNYYWFhmK6272pqamJ33303y8vLYzqdjvXv3589+OCDzGw2S/eh9zo0P/zwg9f/Ps+fP58xFtj7Wltby66//nqm1+uZ0Whkt9xyCzOZTF2+NoEx2Vo7QgghhJBeFpU9I4QQQgiJHBSMEEIIISSsKBghhBBCSFhRMEIIIYSQsKJghBBCCCFhRcEIIYQQQsKKghFCCCGEhBUFI4QQQggJKwpGCCGEEBJWFIwQQgghJKwoGCGEEEJIWFEwQgghhJCw+v+iauIPTYIi2QAAAABJRU5ErkJggg=="
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(history['loss'], label='cross entropy loss')\n",
    "plt.plot(history['val_accuracy'], label='validation accuracy')\n",
    "plt.legend()\n",
    "plt.show()"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2023-12-04T17:15:18.272847Z",
     "start_time": "2023-12-04T17:15:18.222875Z"
    }
   },
   "id": "18da4456a0c63ffc"
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}