teach_16_CNNs_part1.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "aad6ab19-5164-4969-b6eb-11531c45abb5",
   "metadata": {},
   "source": [
    "# Introduction to Convolutional Neural Networks, part 1\n",
    "\n",
    "We will start looking at image classification, the task of training a model to classify an image into one of a pre-set number of categories. The image below shows common computer vision tasks and is from [Shivani Kolungade](https://medium.com/@kolungade.s/object-detection-image-classification-and-semantic-segmentation-using-aws-sagemaker-e1f768c8f57d)\n",
    "\n",
    "![Image classification, object detection and segmentation by Shivani Kolungade](images/classification_detection_segmentation.png)\n",
    "\n",
    "\n",
    "## Image classification of an American Sign Language Dataset\n",
    "\n",
    "For this exercise, we will explore image classification, starting with a \"simple\" neural network and then adding convolutional layers, data augmentation and more to build better models.\n",
    "\n",
    "This notebook is inspired by the Nvidia [Fundamentals of Deep Learning](https://www.nvidia.com/en-us/training/instructor-led-workshops/fundamentals-of-deep-learning/) (instructor led)/ [Getting Started with Deep Learning](https://courses.nvidia.com/courses/course-v1:DLI+S-FX-01+V1/about) (online, asynchronous) exercises. \n",
    "\n",
    "## American Sign Language Dataset\n",
    "\n",
    "The American Sign Language [manual alphabet](https://en.wikipedia.org/wiki/American_manual_alphabet) has signs for each letter. We will skip J and Z because those signs require motion. \n",
    "\n",
    "The dataset we'll use is from [Kaggle](https://www.kaggle.com/arjaiswal/sign-mnist-using-cnn/data). There should be 24 classes...but for some reason, there are 25...It turns out they skipped class 9, the J class.\n",
    "\n",
    "We'll start with this dataset because it helps with the transition from the tabular data we've been working with to image data. Images are a grid of pixels, where each pixel has a brightness. For greyscale images, there is one grid, for color images, there are typically three grids, for the red, green and blue colors (RGB), often called color channels.\n",
    "\n",
    "Let's start by loading the data and examining some images.\n",
    "\n",
    "## Load the data and examine some images"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "e07f7283-df18-44a7-a7c8-d302ea4bf742",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "import matplotlib.image as mpimg\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "\n",
    "import tensorflow as tf\n",
    "from tensorflow import keras\n",
    "from tensorflow.keras.preprocessing import image_dataset_from_directory\n",
    "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
    "from tensorflow.keras.preprocessing import image as image_utils\n",
    "\n",
    "from tensorflow.keras.models import Sequential\n",
    "from tensorflow.keras import layers\n",
    "from tensorflow.keras.layers import Dense\n",
    "\n",
    "from helpers_plot_history import plot_history # A function to plot training history. \n",
    "                            # We 1st used the code in 15_neural_networks.ipynb. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "0d970832-2721-4a1d-a7c5-38463c740210",
   "metadata": {},
   "outputs": [],
   "source": [
    "sign_train = pd.read_csv(\"data/sign_mnist/sign_mnist_train.csv\")\n",
    "sign_test = pd.read_csv(\"data/sign_mnist/sign_mnist_test.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "c611958a-b1c8-4985-a700-e6796e9e72f7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>label</th>\n",
       "      <th>pixel1</th>\n",
       "      <th>pixel2</th>\n",
       "      <th>pixel3</th>\n",
       "      <th>pixel4</th>\n",
       "      <th>pixel5</th>\n",
       "      <th>pixel6</th>\n",
       "      <th>pixel7</th>\n",
       "      <th>pixel8</th>\n",
       "      <th>pixel9</th>\n",
       "      <th>...</th>\n",
       "      <th>pixel775</th>\n",
       "      <th>pixel776</th>\n",
       "      <th>pixel777</th>\n",
       "      <th>pixel778</th>\n",
       "      <th>pixel779</th>\n",
       "      <th>pixel780</th>\n",
       "      <th>pixel781</th>\n",
       "      <th>pixel782</th>\n",
       "      <th>pixel783</th>\n",
       "      <th>pixel784</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>3</td>\n",
       "      <td>107</td>\n",
       "      <td>118</td>\n",
       "      <td>127</td>\n",
       "      <td>134</td>\n",
       "      <td>139</td>\n",
       "      <td>143</td>\n",
       "      <td>146</td>\n",
       "      <td>150</td>\n",
       "      <td>153</td>\n",
       "      <td>...</td>\n",
       "      <td>207</td>\n",
       "      <td>207</td>\n",
       "      <td>207</td>\n",
       "      <td>207</td>\n",
       "      <td>206</td>\n",
       "      <td>206</td>\n",
       "      <td>206</td>\n",
       "      <td>204</td>\n",
       "      <td>203</td>\n",
       "      <td>202</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>6</td>\n",
       "      <td>155</td>\n",
       "      <td>157</td>\n",
       "      <td>156</td>\n",
       "      <td>156</td>\n",
       "      <td>156</td>\n",
       "      <td>157</td>\n",
       "      <td>156</td>\n",
       "      <td>158</td>\n",
       "      <td>158</td>\n",
       "      <td>...</td>\n",
       "      <td>69</td>\n",
       "      <td>149</td>\n",
       "      <td>128</td>\n",
       "      <td>87</td>\n",
       "      <td>94</td>\n",
       "      <td>163</td>\n",
       "      <td>175</td>\n",
       "      <td>103</td>\n",
       "      <td>135</td>\n",
       "      <td>149</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>187</td>\n",
       "      <td>188</td>\n",
       "      <td>188</td>\n",
       "      <td>187</td>\n",
       "      <td>187</td>\n",
       "      <td>186</td>\n",
       "      <td>187</td>\n",
       "      <td>188</td>\n",
       "      <td>187</td>\n",
       "      <td>...</td>\n",
       "      <td>202</td>\n",
       "      <td>201</td>\n",
       "      <td>200</td>\n",
       "      <td>199</td>\n",
       "      <td>198</td>\n",
       "      <td>199</td>\n",
       "      <td>198</td>\n",
       "      <td>195</td>\n",
       "      <td>194</td>\n",
       "      <td>195</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2</td>\n",
       "      <td>211</td>\n",
       "      <td>211</td>\n",
       "      <td>212</td>\n",
       "      <td>212</td>\n",
       "      <td>211</td>\n",
       "      <td>210</td>\n",
       "      <td>211</td>\n",
       "      <td>210</td>\n",
       "      <td>210</td>\n",
       "      <td>...</td>\n",
       "      <td>235</td>\n",
       "      <td>234</td>\n",
       "      <td>233</td>\n",
       "      <td>231</td>\n",
       "      <td>230</td>\n",
       "      <td>226</td>\n",
       "      <td>225</td>\n",
       "      <td>222</td>\n",
       "      <td>229</td>\n",
       "      <td>163</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>13</td>\n",
       "      <td>164</td>\n",
       "      <td>167</td>\n",
       "      <td>170</td>\n",
       "      <td>172</td>\n",
       "      <td>176</td>\n",
       "      <td>179</td>\n",
       "      <td>180</td>\n",
       "      <td>184</td>\n",
       "      <td>185</td>\n",
       "      <td>...</td>\n",
       "      <td>92</td>\n",
       "      <td>105</td>\n",
       "      <td>105</td>\n",
       "      <td>108</td>\n",
       "      <td>133</td>\n",
       "      <td>163</td>\n",
       "      <td>157</td>\n",
       "      <td>163</td>\n",
       "      <td>164</td>\n",
       "      <td>179</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 785 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   label  pixel1  pixel2  pixel3  pixel4  pixel5  pixel6  pixel7  pixel8  \\\n",
       "0      3     107     118     127     134     139     143     146     150   \n",
       "1      6     155     157     156     156     156     157     156     158   \n",
       "2      2     187     188     188     187     187     186     187     188   \n",
       "3      2     211     211     212     212     211     210     211     210   \n",
       "4     13     164     167     170     172     176     179     180     184   \n",
       "\n",
       "   pixel9  ...  pixel775  pixel776  pixel777  pixel778  pixel779  pixel780  \\\n",
       "0     153  ...       207       207       207       207       206       206   \n",
       "1     158  ...        69       149       128        87        94       163   \n",
       "2     187  ...       202       201       200       199       198       199   \n",
       "3     210  ...       235       234       233       231       230       226   \n",
       "4     185  ...        92       105       105       108       133       163   \n",
       "\n",
       "   pixel781  pixel782  pixel783  pixel784  \n",
       "0       206       204       203       202  \n",
       "1       175       103       135       149  \n",
       "2       198       195       194       195  \n",
       "3       225       222       229       163  \n",
       "4       157       163       164       179  \n",
       "\n",
       "[5 rows x 785 columns]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sign_train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "ec5299a3-8f71-41db-95f0-b887f2332cd4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(7172, 784)"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_train = sign_train['label']\n",
    "X_train = sign_train.drop(columns='label').values\n",
    "\n",
    "y_test = sign_test['label']\n",
    "X_test = sign_test.drop(columns='label').values\n",
    "\n",
    "X_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "15551a19-09a4-43f9-98ae-55d1337e2516",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAACMYAAACJCAYAAADNVqahAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/MnkTPAAAACXBIWXMAAAsTAAALEwEAmpwYAACa8ElEQVR4nO29ebhlVXnu+057kbboKSiqCoqeoqiiLSk6QQHBEFCJKDYEzVFzc425xuTofWJO4vEm53hOTiI2x2jkEMVggwiCIK0UUHRVNEVfdEXfK6jYz/vH2jXqHS97fDX2qrn22nvz/p5nP3yr5lhzjTmab3xjrMX3Nm3bwhhjjDHGGGOMMcYYY4wxxhhjjDHGmKnGy4ZdAWOMMcYYY4wxxhhjjDHGGGOMMcYYYwaBfxhjjDHGGGOMMcYYY4wxxhhjjDHGGGOmJP5hjDHGGGOMMcYYY4wxxhhjjDHGGGOMmZL4hzHGGGOMMcYYY4wxxhhjjDHGGGOMMWZK4h/GGGOMMcYYY4wxxhhjjDHGGGOMMcaYKYl/GGOMMcYYY4wxxhhjjDHGGGOMMcYYY6Yk/mGMMcYYY4wxxhhjjDHGGGOMMcYYY4yZkrxi2BVYG5qmaQAsBLD3yN/OADYFsAmAFsAzAG4B8AMA/9627U+GU1NTS9M0rwfwDgAHA9gKwGsBPAHgQQA/BnBe27aLh1ZB8yKaptkAwJsAHAJgPoDtAawP4GcAVgK4EsC/tW173dAqaYq4/yYuTdO8HL11bS8AC0b+uwd6fhEA/rZt209V3Gc+gH3RWyd3x+p18hUAngVwG4AfAfha27aPdfsUL2266sNR7jsXwLsAHA5ga/Tm7JMAHkZvzv6obdvz17b+Zmw0TTMPwB8DOADALADrYrUvXQLg/ziGGV+6WuOapnktgMMAHIqeL90BwIYAfonevFuC3l7j4oE8iCni/eDkx304sek6lmmaZnsA70dvvz8HwHoAXgDwCIDrAJzRtu15HVX/Jc8gYtGmaY4A8D4A+wHYHMBzAO4G8G0A/7tt2593UnkDwPuJyY5j0cnPIM7M7EfHD/vQyU8Xfdg0zXsB/FsfH39527YH9/E+M4LPtic/7sPJzYD28x8C8AYA2wJ4DYDHAFwP4PS2bc/urPKDpG3bSfuHXqO3lX9PAPiDYdfZf8W+3ATAtyr68cZh19V/Wb/9JXob8Zo5eDqAdYZdZ/+5/ybLH4DvrKFPPlV5n8cq+/h5AB8Y9nNPpb+u+pDutw6ALwL43Rru+5NhP/tL6Q+9DIz/C8DvK+bZGQBeM+w6vxT+ulrjALxzxD/W3Od8AJsO+9lfSn/wfnDS/7kPJ/Zfl7EMgL8C8OuKfr4YwLRhP/tU+Ou4/14N4BtruN8KAHOH/dxT6a/LPhy5n/cT49d3jkUn+V9XfUj3sx8d/z60D53kf130IYD3Vs5j/fvqsJ9/sv91NQfhs233of+G3X8NgP8M4LdruN95ANYf9nOv6W9SZ4whHgZwDYDb0ZtgT6AXbO4E4G3o/Z9ImwL4TtM0R7Zt+6NhVdS8mKZpNkfv8GvXkX96AL1fyd8K4BcApqP3f14fMZQKmogd0JtrAHAvgIsA3AjgKQAboffLweMBvBy9X9JvNjIHfz/+VTWj4P6b2LxcXj8D4Gn01rSx8hR6/xfZrQAeRW+tfDl6/7fTsQD2RC+7xZeapvlt27Zf7bPOJqezPmyaZl0A5wI4aOSfnkAvuF2K3v9htiWAGej9H0vb9Flf0x//A8Cf0etzAFyG3v/9vhmA/dGLR18O4I9G/vv28a3iS5Ku1rhV2X+Anv/8EXpZDZ4A8DoAi9DLdvga9GLVi5qm2b9t218M5rFMAe8HJz/uw4lJJ7FM0zQfAfAZ+qcfo5cF6EH0fPKeAE5Cr88PBfCDpmkOaNv2d/1V24zQ5X7iNAAnjNhPA/jf6GVz2gS9dXQfANsB+GHTNPu2bftgXzU2ivcTkxfHopOfrs/M7EfHH/vQyU8XfXgJgD+sKPcyAP+O1ZkU/m0Mn2FGx2fbkx/34eSmq/77m5E/oPcDmO8CuADATwHMBvBu9DLTHAng7KZp3ti27W/6rfTAGfYvc9by104vA7DLGsq8HMDnsfoXS7cPu97+y/qnAXA59c8/AHhVUH6bYdfZf1l/fBmrNwVNocwi5P93y/uGXW//uf8mwx96v8L9DIC3Apg18m/vxdh/0btrqX+pzF/TfZ8F8OphP/9U+OuqD0fe9+/0vv8DYL2g7NbDfvaXyh+AmVj9f4v9FsAbC+Xmiy+dN+y6T/W/rtY4AJ8EsBjA0QBeXrjPLuj9EGrVff522M//UvnzfnDy/7kPJ/ZfF7EMev93NfvakwvlZgF4iModO+znn+x/He4n/oDe8wCAGXL9ZQC+SmW+Nexnnyp/3k9M3j/HopP/r6s+HClnPzqcPrQPneR/XfZhxWcdQfe9e9jPPhX+OoxFfbbtPvTfkPoPwG5Yffb9GwBHj1LmlehlSV913z8f9rOHzzTsCoxT578SvV+jreqU2cOuk/9S3/wn6pcvDLs+/htz/21UWe5PqZ8vH3a9/ef+m6x/g9r8jdz7Jrr3ocN+1qn61+fmgTfn561pI+G/ce3PU6hvzlxD2f9OZf+vYdd9qv91tcaN4T5H030eGPbz++9F/eP94CT/cx9OnL8+DtIOo/LXrqEs++TPDvtZp+Jfn7HoMnrPUYUyr0Xvy95V5XYb9rNO1T/vJybHn2PRyf/X5ZmZ/ejE+bMPnfx//fRh5X3PpPv+52E/51T9G1T/jdzbZ9vuQ/913H/oSQiuKv/fg3Kvxeofaj+NCfzDppfhJUDbS9lzN/3TFsOqi1lN0zQNgL8YefkzAB8fYnVMH7Rt+2xl0W+Rvfsg6mLGjvvPCLeR7XVyYvGxkf/+Hr0fVLTDrIzJ2Izsu4uletxF9usGUBdDdLXGjeE+5wP4+Yg9o2ma9SvfZ8YB7wcnP+7DSY3XyklM0zRzAMwbeXl327bnjVaubdsX0MuusArLRk4svJ8YZxyLTn666kP70SmBfegUp2maaQDeMvLyd+hJn5nJh8+2Jz/uw4nHoWSfXio0EsecOfJyGoA3DbJSa8NL4ocxTdO8DL1U96t4bEhVMTmL0NOPA3r/l/Vzw6yMGSjPk/3aYikzUXH/vTTYjmyvkxOEpmm2BXDIyMvL2ra9Z5j1MS/icbLXpM/K128fQF1Mf3SyxrVt+zsAv+jiXqZ7vB+c/LgPJzVeKyc3fKB5wRrK/pDsIwZQF9MH3k9MeByLTn7W1If2o5MY+9CXDO8E8OoR+8K2bR8eZmVM3/hse/LjPpx4bE32nWsoy9ePGkBdOmHK/zBmJCvJ32P1r8tubNv23iFWyazmQLIvaZrmFU3T/EnTNIubpnm6aZoXmqa5v2marzdNc9jQamm6YDeyHxhaLUy/uP+mOE3T/CcAe4+8fBzAlUOsjslZBKAZsS8BgKZp/qhpmgubpnm8aZpfNk3zUNM0322a5q0jcY8ZP84H8OsR+7imaQ4frVDTNPMB/MnIy7vRS79sJgadrHFN02wGYNORl78A8OTaVMp0h/eDkx/34aTnSvRksABg76Zp3jdaoaZpZqKnKw/0Ui//++CrZirgdfKGNZS9Eb3/yxoAdnFcOmHwfmJi41h08rOmPrQfndzYh7404Pj0q0Orhekbn21PftyHUwJeAyes8sQrhl2BLmma5ggArxl5uQ562UiOA7DHyL89DeCPh1A1Mzp7kf0UgKuw2vGtYtuRvxObpvkPAO8bSclkJhcfIPsHQ6uF6Rf33xShaZoD0UtlB/T+T4iZ6GmRHzDyby+g52d/Nf61MwV4rby/aZrzABwpZaYD+MORv8uapjm+bdtnxquCL2Xatn2kaZq/BPBPAF4O4MKmac4BcCl6uqqbAVgI4G0j128DcOyIJIiZGHS1xvF9fti27e/X4l6mT7wfnPy4D6cebdv+cuSQ85vonUF9tWma9wI4F8CDADYCMB/ASejFpw8DOK5t26eHU2Mj7ED2/VHBtm1/2zTNwwBmoCeFNR3AQ4OrmqnE+4mJjWPRyc+a+tB+dHJjHzrFaZpmDwB7jrx8CsD3h1gdswZ8tj35cR9OOh4DMGvE3gHAzUFZzgC748BqtJZMqR/GAPgagM1H+fdfo7eg/WXbtveNa41MBGvEfQ69Q8/nAPwrer+gfwV6v8p+D4BXAjgBwKvQOxg1k4SmaRZi9a+uf4nel4dmkuD+m3L8I4B9R/n33wG4CMBft227bHyrZNYAr5WfQm+t/BV6esdXotd3ewE4BcC6AA4GcH7TNIvatv01zMBp2/Z/NU3zOIB/QO8A85iRP+ZJAJ8A8PW2bX8BMyHoao1rmmY2Vmc5AID/b+1qZtaCr8H7wcnO1+A+nHK0bfudkR89/QuAndHLHnugFPs5emvlv/nLpAnFhmQ/VSpEPI1ePLTqvf5Cd/h4PzFBcSw6+answw3Jth+dfNiHTn1OJvvr7rcJj8+2Jz/uw8nFYqz+YcxJAD42WqGmaV4D4O30TxsOtlr9M+WllEa4A70J9cSwK2IyNiR7e/R+NT+3bdu/aNv2G23b/p+2bd8P4PXo/WAGAP6waZq3w0wKmqbZAsCZWO1r/t+2bR8cYpXMGHD/vaR4AMCFAFYOuyLmRWxI9vYAngWwf9u2fzKyTn69bds/BzAPvf/DGgD2AfCR8aykwXcA/DlW94GyKYC/RO9HvmYC0NUa1zTN6wCchV5mCwA4tW3b67qppekQ7wcnP+7Dyc+lAP4Mvexpo/E6AB8FcIplCCYU65L9y4rynOF3vY7rYvpjQ7K9n5ggOBad/IyhD+1HJzcbkm0fOsVomuZVAN5J/2QZpcmLz7YnP+7DicmXyf5I0zSaNQ1N07wCvYQXW9E/v7xpmtcOunL9MKV+GNO27RZt2zboPdcG6P2g4gsAdgXwRQDXNE2z3RCraHJ0/H2gbdsXabGObOg+Qf/0fw+0VqYTRjbnZ6OXThLopRP97PBqZMaC+29q0rbtfm3bNiNr5boAFqD3f5NthV7/3tg0zYJh1tG8CF0rPz7ar+bbtr0HwH+if/qzgdbKJEZiyxvR+3HMrwG8G8CW6GW523Lk9X3oHaJ9tWmazwynpmYVXa1xTdO8HMA3AMwd+aelAP6fLupo+sP7wcmP+3Bq0jTNpgAuA/Aj9H4s+mH0JJNfNfL6eAA3oReT/gOA05ummVLnVVOEdtgVMH3h/cQEw7Ho5Gct+tB+dPJhHzq1eQuAjUfsG9q2jSRCzATAZ9uTH/fh5KJt2yuw+keDrwBwbtM0ZzZNc0rTNG9rmubj6O3l34le1rvn6e0TUtpzSh40tD2ea9v2qrZtPwTgzeilYdoVwI9GglczfHiCPNC27Y+Csv8G4Dcj9r5N06wblDVDZiRt1vfR+4U80EsteULbtt4ATgLcfy8N2rb9edu2S9u2/WsAC9HzyVsDuKhpmq3id5txhNfKXwI4PSj7A6z+P5SmN02z88BqZQAAI3NlCYBdAKwAsFfbtqe3bftY27a/Gfnv6eilVr5n5G1/1TTNm4dU5Zc8Xa1xI1/Yfg29gzQAuBPAkW3b1vxfoGbAeD84+XEfTh1G+moxejLJTwPYt23bz7dtu3JkrXyqbdvvAtgPwNUjb3sngA8Op8ZG+BnZNf/HH5d5vljKjCfeT0wgHItOfvroQ/vRyY196NSGZZScLWaS4bPtyY/7cNLwQfQywgC935W8Db1MMmei96OmXdDL7vsWAK8cKfdC27a/Gud6VjElfxijtG17AXqbBaCnhfXu4dXGED8he2lUsG3bn6O3yQOAlwOYOZgqmbVlJAXhdwEcOvJP1wI4aqQPzQTH/ffSZOT/dvnHkZcbwpm5JhI/Ifv26KBz5CCO/88l/x/1g+eTADZZZbdt+8xohUb+/ZP0T//XoCtmXkxXa9yIxMeXALxr5J/uAfCGtm0t8TJB8X5w8uM+nNR8GMAOI/Z/a9v2vtEKjcQ4H6V/8lo5MfgJ2RuXChXK/KRUyIwrPyHb+4kh4lh08tNnH/6EbPvRycdPyLYPnUKMfOn+xpGXv0QvA5eZpPhse/LjPpy4tG3767Zt3w9gf/SSWKwA8IuRv1sBfBq9/4lpBYDXjLztsSFUtYqXxA9jRvgh2QcPqxIm406yf1pRnsts0HFdTAc0TfNKAN8CsEpnbhmAI9q2fW54tTK1uP9e8nidnJh4rZzYcOaXi9ZQlq/vUyxlBkLHa9znAJwyYj8A4NC2bR8OypuJgde5yY/7cHIylrXyGqz+P+t3bJrGsczwuYvsmVHBEW35VbIiP8fq/2veDBfvJyYAjkUnP2vRh/ajkxv70KnLe9D7H7AB4Ky2bX8yxLqYbvB+cfLjPpzAtG27pG3bk9u2ndO27etG/nZr2/aTbds+BWBvKn79sOq5Jl5KP4zhtHcbDqsSJoM1G9evKM/BZE0gasaRkc3bGVidxvUWAIe3bfvs8GplanH/GXidnKh4rZzYcFrPNR2Icn9YAmQc6XKNa5rmnwB8aOTlQ+h9EbGyi3qageN1bvLjPpycVK+VI/+nNZdZZyA1MmNhOdl7raHsPKz+guk2S/FOGLyfGDKORSc/a9mH9qOTG/vQqct7ybaM0tTA+8XJj/twcnME2VcMrRZr4KX0w5jtyX5qaLUwzPlkz48KjuiS7zjy8jcARk2/bIZD0zQvR09j9fiRf7oNwGFt2z49vFqZWtx/ZgSvkxOTy9H7P8UAYJcRPfNRGUmpvSf9012lsqYz+Mu7bdZQdluy7V/HiS7XuKZp/hGrU7k+it4XEfd2UlEzHnidm/y4Dycn1Wtl0zSvBbAp/dOoEoVmXLmA7DetoSwfhP6wWMqMN95PDBHHopOfDvrQfnRyYx86BWma5gCslvp8AMDFQ6yO6Q7vFyc/7sNJStM06wJ4x8jLCS1P95L4YUzTNC8D8Mf0T1cNqy5mNW3bPgDg6pGXM5umOTwo/j4ArxyxF49Vf9cMjpH59VUAfzTyT3fB2saTBvefIT5AttfJCULbti8A+P7Iy9cAOCko/masTrt8X9u2PoQZPPx///1RsdSLr0/YdJJTiS7XuKZp/h7Ax0ZePo7eFxF3d1JRM3C8H5z8uA8nNWNZK4/H6n3/LW3b/mowVTK1jKx1y0Zezmma5sjRyo18Ufh++qczB103U4f3E8PDsejkp4s+tB+d3NiHTlneR/bXnJ1pyuCz7cmP+3Dy8vcANh6x/20i/0/3k/qHMU3TfKRpmv3WUGY9AP+O1b/WfQbANwddN1PNJ8n+303TbKsFmqbZC8Cn6Z/+28BrZaoY+SX8lwC8e+SfVgA4pG3bx4ZXK1OL+2/q0zTNe5umeeNIX5fKvKppms9idUrgXwP413GpoKnlbwH8dsT+h6Zp5mmBpmlmA/gC/ZPXyvHhDLL/36Zp3jBaoZF//wT90+kDrZXpdI1rmuaTWN1/T6J3GH5HJxU1a4X3g5Mf9+FLAl4r/7hpmneNVqhpmj0A/BP9k9fKicPfkv2Fpmlm8MWRL45PBbDq37/dti3/IMoMH+8nxhnHopOfjs/M7EcnN/ahU4gRZYK3j7xsAXxteLUxa8Jn25Mf9+Hkp2ma+SMZYUa79vKmaf4Gq7MaPgTgr8etcn3wimFXYC05GMD/bJrmbvTSnS1HLzX979BLvzsfwB8CmDZS/rcATpnIv1R6qdG27SVN03wBwAcBzARwc9M0/4re/0n9CgCLALwHwKtG3vLltm3PH+1eZih8GsApI/ZvAPwzgH2CNW4VF7Zt+4tBVsxU4f6bwDRNMwv5/xkNAHPJPnRE55r5Ttu2y+j1PAD/BuChpmkuRE8b+Un0gstpI/f7QwBb0Xv+n7Zt71z7JzAd9SHatr2zaZpPAPgHABsBWNI0zdcAXIlezLMXev9n2aoA9QL0DvDM4PkKgJMB7I3e/z12YdM03wNwIXox6cYA3gjgWKz+QfoPAXx7vCv6EqSTNa5pmvcD+Du6/jn0/k/POWu4z+K2bZ32dfAcDO8HJzsHw304Yekilmnb9ocja+Ox6K2FpzdNcxKAc9GTAlkfwEEATgDw6pG33YSevzVrQYex6NlN0/wHen20LYAbmqb5EnrzdWP0vjTeZ6T4owA+2tEjvOTxfmJS41h08tPZmZn96HCwD538dNWHwtuwuq8uadv2/rWrpSnhs+3Jj/twctOhDz0ZwHuapvkheiowD6G3d98BvR8arpKmexrAW9q2/WkX9R8YbdtO2j8A30PvV501f/egp/859Hr770X9+DL0Nhe/X0Mf/guAVwy7vv7L+u6yMcxB/ps57Lr7z/030f/Q+6JorH3zXrnHP43hvY8D+KNhP/dU+uuiD+V+f43epiF6/7cArDPsZ38p/aF3kPnDyv49E8C6w67zS+GvqzUOvf97rJ/7HDzsNngp/MH7wUn/5z6c2H9dxTIAXoteBpia918CYIthP/tU+Ouq/0bu9Wr0sv9E710BYI9hP/dU+uuyD0fu5/3E+PXdZX30nWPRCfTXVR/S/exHx78P7UMn+V/XfThyzx9T2ROH/YxT+a+L/oPPtt2H/htq/43c53MV77sWwK7Dfuaav8meMeY96GUUORi9X1VvCWAzAOsAeB7Ag+hpeH4fwLlt2/56ONU0EW3b/h7AnzVN8w30fr12MFb/OvBhAJcD+ELbtkuHU0NjjJm0/Gf0/m/cgwHsj57W8WYA1gPwc/T+T6QbAZyP3q+Bfz6UWpoq2rb9TNM0Z6Ont/pGAFsDeCWAx9D7P5W+2rbtJUOs4kuStpe14IimaQ4DcCKAfdHrm9ehN89Wovdr+tPatr1yaBU1Zmri/eDkx334EqBt2xcAnNQ0zb+g1+cL0csYux6AFwA8AuAa9L4w/GE7cvpmJg5t2/4KwDuapjkNvf9jcD/05urzAO5G74vA/+39xMTG+wljhof96OTHPnTy0zTN9ujtPQDgpwC+O8TqmDp8tj35cR9Ofv4FvSwxBwOYA2DzkX9/DD31l28BOGvku/4JT+PzBmOMMcYYY4wxxhhjjDHGGGOMMcYYMxV52bArYIwxxhhjjDHGGGOMMcYYY4wxxhhjzCDwD2OMMcYYY4wxxhhjjDHGGGOMMcYYY8yUxD+MMcYYY4wxxhhjjDHGGGOMMcYYY4wxUxL/MMYYY4wxxhhjjDHGGGOMMcYYY4wxxkxJ/MMYY4wxxhhjjDHGGGOMMcYYY4wxxhgzJXlFdPH9739/u8reaKONsmvrr79+sl/3utcl+zWveU1W7lWvelWyX/nKV47673rtFa/Iq/Xyl7981GtRuRdeeGHU+mk5tpWXvazud0NN0yS7bdtiud///vfFa/y+hQsXNsWCY+SMM85IN9ZnjerKlJ4vajt9Vm7L3/3ud6Pa0T20rvy6n3to/fj1b3/722Ldf/WrXyX7kksuycptvvnmyf7Xf/3XTvrwkksuSZWOxmPt2OK+1PvVtndEbTlt41VEfVn7OWeeeWb2+oILLkg2+zH2YQDw6le/OtmLFy/ubA4uX768rlECSn0YtQOXA/L+ZVvL8T2jeVYqp3WqnYN8f14PAOCRRx5J9mc+85li3fn1dddd10kfXnDBBanSuubwa/aHWn8ux/NOyzHqX/l9kS/gNmBb21vbbhU6N0vPqO+P5m6pvtEaMmPGjM7m4Be/+MXUhxqj1PYN+we+puX4mfg9Y/ksvhbFKyV/ruO01NfaL1F/MDyW9B78eosttuikDz/72c8WHV1pbEX1YkptMxrcPtF4L5WrbV+ln3UxinGieIH5i7/4i87m4Mknn5wqd9ppp2XXdtlll2TPmjUr2Tp/+PU666yT7HXXXTcrt+mmmyZ7vfXWy67x3Hjta1+bbO0bnpO8h7jxxhuzcjvttFOyN9hgg2TruIr2LqVy0diMfDGzYMGCzvrwtNNOS32ofVPyZ9E+LVrTate7iNI4r41TlVLcWhvz6Gvut3/5l3/Jyi1btizZTz75ZCd9uGzZsqoHj8YT90WtH4naIOqLkj+rfU+pv5QoNor6NvKp/L7DDz+8szn40Y9+tLgn5HOVUryi5SI/Uor9tGztPB7LWlui1Pe1+5ixwOvuBz7wgU76sGmaVJltttkmu8brR9RupbM09cl8P93v8prG79N+Lt1fz/BK+5/adVDHTWkfM9rrVUT9fMopp3Q2B0888cT0QTq3eH8xbdq0ZOsZKscv3JbartE8Lu0NdM0t7Sei+RjNn9JaP4j5zXPw0EMP7aQPFy1alD5kk002ya5tueWWyf7617+e7A033DArN3fu3GRzG0T7S/0sHh8ci2o8y322xRZbJPv+++/Pyj300EPJ3n///ZOtc6tU36ifozlYO4/32WefzubgxRdfXLUnrJ0XEdHzMVE8VIoV+EwZyMfBL3/5y+L9eF5En/vrX/862VE8FJ2B8+sTTjihkz485phjUv/p2CqdZei8YB/K41i/8ynFO/pZ0bkO14Ov1Z7T6bgp7fO0LWq/r4ruwbzlLW8ZyJ5e68Z9xWNaYxS+xv0UtVfUN9E5LLdLaR+q1H5n0u89+uGYY47ppA9XrFhR3EuUfH+0DvS7/tfuKUttF33fWFqnFPbDY+nnfsbAlltu2dkc/N73vlf8fqLkO3RelPxZFOPrfC+1bdQmUSzaz56t37nVz2dtt912nfThOeecU1wHS2uE1reLuLv2jLlEdP4Rzenaz609/6mNwd7xjncUG80ZY4wxxhhjjDHGGGOMMcYYY4wxxhhjzJTEP4wxxhhjjDHGGGOMMcYYY4wxxhhjjDFTklBKiVOe1abD7jetXJRqrDb1KqfSY/mUBQsWZOVmzJiB0YhS8NSmKqpNQT2Wz14b+klz2kWqt9qUf1qO24FTLEXprGtTTEUyUPxZOtZLKTBnz56dldtxxx3RNbVpqGtTuEUpqGrHZzRWuV05hSfbwItTtK0iavvo37nuM2fOzK5xulr2Eb/5zW+yclHa0q6olffQVKYl/zgWv1H6rCh9Ls+LLtIk1qZA03SbTzzxRLI55aGmBKxNnT8WovSBpWv9ykJEqQVL1yKfF30u3+/ZZ59N9h133JGV43mxcOHCZGvK6H7SYXaRArCG2pSN/aQMjdIA16YIjuKhiNJn6ZgoPX+/0j61sVxXcMrkyDdGzxNJtZXuVztXo/SqteUiamOyUvyk9aiVXOoS9iP6Gdy/tfOH129Ofw/kaZw13TxLH5XaBMhlKLjut956a1aO5TBYwknHKcu78vjT9OBdpCYeFP3IrnSRarxfalMv87M8//zzyeZ1EQCmT58+6j10rnHd9bO473ls3nPPPVk5TjE+HpT8aBSzRf62NpVuRO36VopTozHE5XR/Uiqnr2vTrA+K2tgjkkFixjLnSr4gilsjaturdi8U1anrdbdfSvtgJfKhEbX9Urvm1Prr2n1Hqa76vqhOtXJAXVIbo0QyLqXzVR0T/LpWZr5WFkFlXDhm4XK/+MUvsnIcy/DnqtxQJC3JlPZga3pfv/D8V4krbgOO+zhGBerPwOfMmZPsZ555JrtWknGJ5hZLoj355JNZuccff3zU+2l8wvd/6qmnkq2yo/zM0ThnasutLf18n1B7LlN7ThfVSdeYUkx41llnZeV4vOy3337J/tnPfpaV07lbIvILvIfq4vkHQe1ZS63sV7Qe1d6j3zWyRD+x2lgY1D4yWt9Ka5qubyXZXb1fJIdVqkckj8z0G0eX1q3I346Xf6yh9uysaymo6B5RW5VkV6K9109+8pNk81oHAFtvvXWy+exG94PRXrHWfwyKqA9Lvr/f+VP7XX/tWOJYS+cmxx7RWsefVfsdfr8SooOgi+/la/fEtc8Tnd3UjvFaibFSfaPPjaiVwY5wxhhjjDHGGGOMMcYYY4wxxhhjjDHGGDMl8Q9jjDHGGGOMMcYYY4wxxhhjjDHGGGPMlMQ/jDHGGGOMMcYYY4wxxhhjjDHGGGOMMVOSUFC6Vj+wpLEGlPW+x6InXKvBxfV44oknkv3Tn/40K1fSvupXOy9637C1rMeiGb4Kbe9I/7cWvketLn2kc89ag9zGkaZYpK8W6dSxDt5rXvOaZKseL5frin60bbvQxdN7lOa4ji/Wnr7pppuSff3112fl9tprr2RzX7LmLQAccMABxWsM12m77bbLrrGGNGtGsz3a666INH5LenglLVR9j2puRtqATDReIo2+EtFnla5F+qF6beXKlcnmOfjzn/88K9ePv1sT/WgL12ocR3rC0T0i38jtzXqorGkP5Hrwzz77bLJ57QSAK6+8Mtk8t2bOnFmsn/YD923k/2vXy7FS26799KGW4+eLtFxrdYfZF9TqT9eOnYh+1p5BEbVpiX59Qa0eagT3WbRu1Y73Ujn9d26baA2J1tza9h0rrPWs4/O1r31t1T24T/nZd95556wcrxd33313dm2XXXZJdvSsHDfcfPPNyda6b7bZZqPWSe+9fPnyZLO/Pf7447Ny3B+D6osuiHxM5G9L8WOtv+0X7puoThzXn3XWWVm5I488Mtnc76o7XvpcIH/mhx56KNkcOwO55nlXRPscvlYb80RE/cefXdKNj+4XXYv2q88880yyub232mqrrFy0l+tCE3xtiOJHHltdzK3oDKgfrfHSvYF8vYrWxdIep9/2rtWlHwT9+vfaNq09m2O0XDSOSveLzghKsajeu+uzqy6Jzg1LsarGY/x8fG3dddfNyvF+V/uX151f/OIXo/47kO/9otiUX3PbaTmO5din6rjk/eKcOXOya6U9rN5j0HNw/fXXz17z8/Bza79wvbjt58+fn5WbO3dusr/yla8UP3vjjTdOtrbBBhtskGweN88991zxfhxTa2zB91+8eHGyZ82alZVbuHBhsvV8rDZeGHT/jfYZpTO3aA2L5jS/7vdMn19zWz7//PNZufPOOy/ZjzzySLL5vAYA9ttvv1HvF6FxTckXRGtAV0RxWu1ZS+k9ted0UdlojxDFGqU1J1oH+3nGsTCodbC0hulrLveqV72qr3LRZ5W+j9T1MvIFTGlsqh9lv1z7HVc0/sbjLI3heuo+LIrN+iF6tto24DbWPVsJHitLly7Nrn3/+99PNq99W2yxRVaOz5emT5+eXYv2/kztPnesRG1XOpPXsV+KU/V+pbk6WtkSPGfuuOOOZF999dVZuX322SfZHEP97Gc/y8px+3OMpus0t3+/39kPYn7287uIKLaopfZM5le/+lVWrhSTbLLJJsX71X4vWXtmVNtH0dlbhDPGGGOMMcYYY4wxxhhjjDHGGGOMMcaYKYl/GGOMMcYYY4wxxhhjjDHGGGOMMcYYY6Yk1VJKUYrWKOVWbarCftIEaVocTifIqT6jdGqDTvNYm4JxUKnuOB3doFNE9ZNiMkovFskglfotklwq3RvIU0zquOI25HJ6j6eeemrUz1obIgmMWomafqSwNFUhp+Ll9K333XdfVm7ZsmXJvueee5J97733ZuVYruV1r3tdsjklLQDMnj072ZrergSnagPyVMSPPfZYsjWFqUrNdEVtuux+0tePRfajlIosGle1afpq/UeUnpnnnaZ8ZWkMTu2mddc0e11QStOp12pT3EfpzUvvicrquH366aeTffvttxfvz2mSSym0gXy+33rrrcV7cz123HHH7Nq2226b7H7SEq8ttdIekb8t9W8UG0Xpc7mdo5SVtbJn0TitlTkpvScqWyvvtDZoWl2mNJ6i90QSDFH9OZ7gVOVLliwp1ollAzUtZW1qx1oppVoiWZlBwX4kGu/RulKSgolSK3N8oXB/aqp8vgfHdprSfJ111kk2r1s6VzkFKcvw6PyplYxkxktyKfIxtf6xtA8ci7xuLZGPZdhX8DqmffHwww8ne8aMGcVyTBSPs1wSjwmgXl5sLERyh0y0RpeuDSLNcK1/K6UEV7/AqdAvu+yyZL/xjW/MyvFeQ9OnM9HaPKjYpjZ+ZGpj97GshbWSAVxfblfeiwG5f9x8883HXPexnKGU3qfPO+j4tFaeKvKNEdGepPZsjtuAfRTvq4E8LTr7RqU0fscy9mrXg0HFNtF9S2npo3mxww47JFslAji+ePDBB7NrHFOxn9KzKJYH5LVOn4OliVmCXtc33mfze3Tffs011yRbxx/LE/AcHNRZKMNjTfcJ/GxcF+4HIO9PXnN0Px7J8pXks7W9WdaT56Oel/E+m+sUPSOPIY1tmX5lbAbVn/2co9SeqUTfQXR9Vs5zEwAuvfTSZPOY07Ot3XffPdm8x1Fpj9oYOJLeHURs1+85RD/36Po7H55bKsXGczeSf4jGJTMWWahhUnvmGe3Vozbh96k/K5251UoaRWfgHIt++ctfzspxnHPQQQclW+dqraRlrZTIIFA/vbZ+u9+xWSvLx/GJfufDTJs2Ldn6HRI/4ze+8Y1kH3DAAVm5H/zgB8n++7//++war/fRd5aDOqMpfQehnxmNwVK52jMefR3twVlyZ5tttkm2fpd4zjnnJJv7UOGzV45l3vWud2XlSmfvei3qp0HEMv38fqKL7wejz+L5xPMMAG666aZkc3u/+93vzsqVnitqQ54/GgPXxnvR2XZt/02cVdUYY4wxxhhjjDHGGGOMMcYYY4wxxpgO8Q9jjDHGGGOMMcYYY4wxxhhjjDHGGGPMlKRaSqk2na2mIeI0SpHEQJS+qfRZmoKQ5TY4lZmm2epHAiSin7RGmhZtPFK+akqr2jTYteWiFFT9pNyuTX3FROnLOMVkJOMyc+bM7Nott9ySbE73peU4NXdX1KbsrU0RxffTFK+cElLTJLM8y3nnnZdsbhsAePLJJ5PNaV415SuPB5Y90zp96UtfSvZf/dVfJVvT1fL9VBJhzpw5yebUvppylFMmdkk0jkvp0WrTjUXp0KLUi7X+huXpOG2a1pH7TeXLeG6xPID2E6fH1PT1LGHAaWNVmmQQ6Qpr09HVptutlaeKnoWvqQQDzzVeB1esWJGVY5ky7medgzxP7r777mRz2mYgl1G76qqrsmsf+9jHks1+ZrxSwdamh6yVdIz6qTa9bzQ/eX3ieaF1KqWN7TeeiKQJattpEETpQpnaevUj3alwXMnrI5Cvi4ccckiyI1mS2hiM52fte5SoPQeVspfbSOMLXs+jccxrC9c7kkHUNaK0X9H1iH3ns88+m+wtt9wyK1dKE81ykUDul+fNm5fsaEwoXad0Hyu18h61Ekm1Pqt2XdfYoyQdpGscl+PxonKbfI3vF9U9ktNj2QyWSwBeHDN3QT/p5qN9Y7QHqf2sKNYtjXGdMyUpjuicgec3nx0AeWzDsZHeg8fbeO3pI2olE2vXcm6/KB7gsasSSbvttluyOVZlCRZ93/3335/so446Kiv36KOPjloHlczT/V2JKJYbtKxLvz68tEaPJRat9cN8jf3m1ltvnZW78cYbk83rpe6rS9IekY+vlfUcL2pTykdnqDxeOaW8xjIc/+v6NmvWrGTz2qTxKPvwjTfeONm6xrDkNfs9la7g/TmvkfqMfI8bbrghu8bPzP6ja0mH0eA21liUz6O4/7RfShJsWl9+n65HJR8VSR1wXKl+gKUZGfWNLMnFddI9Pc/3Wsnc8ZqrUSxckmSJypXuPdrrGiLpHJ5Peq7Gn8XzmyUMAOCBBx5INsv1RtJd0X6K3xedrw6TWrm9aAxG1yJJNO4/jmO22267rByfkXHfqk9+5plnks0+WfeXXI8oxh7GfjCSN+JrJWlwIG/X2nLah6V79CubzuOA1y39voPXLT7bVr9euz8sjUUtNwj6PceM7lFC+5afleMY3Xtx3956663Fz2V/yHJ00ZxmH6dnRvz9w2233ZZdYwktjh2GcbZdK5EUyZRF86yfc3SVBmU43thjjz2ya3wGfuqppyab95NAHkfy9w5nnXVWVu5973tfsrmfororg5ZSql3fonJcx7GcLXI9WFJT+5nnE/tGXs+AXLasdGak17h+OkY5Pun3PKM2jnHGGGOMMcYYY4wxxhhjjDHGGGOMMcYYMyXxD2OMMcYYY4wxxhhjjDHGGGOMMcYYY8yUxD+MMcYYY4wxxhhjjDHGGGOMMcYYY4wxU5JQtC7SE6vV0a3VpqvV7GX9rPXXXz8rx1q3Jb0spR/900inK9L0KmlpDZJIO7pWQ7BUbizPwM/OfRjpjbHuneotcrlIU5WvsW7gJptskpVjneHly5dn184///xkL1iwINmq5xuNs36pbeN+xpPqLbJG5lNPPZVdu+KKK5J95ZVXJpt1boGyDqbqyfMYiHRFWQuQx8BY9MhZb/KMM84Y9X7Ai7V+u6K2b/rR9FRKutdaD372SDeRy6mGLM8nnmc//elPs3Lsl1l7lccbAOy4447JVk1snlsPPfRQsrXPBqGFHGnRltYtLVfS7Yy0ciNYz1HnFs8n1vfUucr6rWyrViTXaeHChcW6subrXXfdlV1jTfO5c+cmW333oNbFaN2qXd9KmuZRvKJaw+utt16yWT/8kUceycrxeGcdY9avBoCf/exnyeb1SOcq67VGeqrRtVI7DUpDnolizEhDnuknFtNr3J/clzznAODRRx8d9X7aL6UxFfmx2nJRXMToPB6Unjz7FfXvtc9UWvd1Lef+4NhO4bGr7cVzi9c01UzWPl0F+zwg16Xne6ieda2O8zA05Wv3hJFWcK3Oez/7k2g/we2sbcfjkee49m2pTlE8qvEVj7ONN9442Tw+gDzm7opa38/UllNq/Ujt+hHFqXyP0l4TyGMlvqbrdO247Le+a0Ok112KX6I+jPqTn0HL7bDDDqO+58knn8xesy/m+m6wwQZZOdYk5/VT58Wzzz6b7KVLlyZ7u+22y8pxPKT9y/BzaZ8NOraJNNAjn9LPuUD0LNF44P7jvdy6666blePxsGzZsmRvvfXWWTn2f7XnU1F9mUHFLkpt3Xjti/aEL7zwQrJ5Lwbk+2SNm/iefJ6jc2bGjBnJ5n678847s3Ic86yzzjrJ/sUvfpGVK+39Fe4P9Qv8mue+3m8QfpTbQGN39i/cpnpeVoL7EojXEn7W0pmY3vPpp59OtsaOHK/wPbQN2b9y2+uc5rFYinOB+v11l0Q+q3Stdq8Xxay1qC/ie0Z+iucd7/11DeMzA97j6JlYF3UfBLXzOtoH1J5J8OsoHt9tt92SzbE5kLcj79/UN/Ic4nbUc9HHH3882bfddluy9Xst9t3bbrttdo3HR+0euktq1zf2HTq3SntK9bd8TX0Rv+5n76nwOsv9tmLFiqzc2972tlHvrW2h/pwp+c7x3t938d1trX/VdYvPO3mtu+yyy7Jyu+++e7K32WabZOv3VY899tio9+b3APncYh+qewn+vlDP9mrXmkHtB6Nxx+M/mqulMRidP9X6WI4pgbx/OfbQvuEzaz73Xrx4cVbuyCOPTDZ/t3D99ddn5U444YRk63xkf1l71tQV/ezRas9koj2knpny+K/9fj3at/SzHnXx3RgT/R4jwhljjDHGGGOMMcYYY4wxxhhjjDHGGGPMlMQ/jDHGGGOMMcYYY4wxxhhjjDHGGGOMMVOSMDdNKQUuUE7r02/K69rP4rSinGoJAFauXJlsTgukkh216etLzxil44lSF0UMKvVdlH4okl0p0W+6Pu5f7rcoXVKU/pg/m1NCafpDTkvJaZ9UuoJTGd5xxx3ZNU6vyHIvWr/58+cX69svtfMpGpNcjttKU1lz25977rnZtauuuirZnLZOU+Lxa277SDqBr2kK4KOOOirZLF2hlJ4RyFNqv/Wtb022ypLcc889xfuvDewTdM6sbWq2ft/PbaTtxWkp2UdwimEgT4PHaZFVimf27NnJ5hSHmvJ1yZIlydaUoZxelOur7TmodIWrGIv0TOl9kRwFE0lV3H///cm+4YYbsnKcOpJ9l66XLCvAqWG1Tbn/5syZk2xNwx2lCuZ0fMNIF9qPpKDWrSQVEqXMVMk+bkv+LE3dWUL9KMtjsZQVpxwF8pShXD9NkxhJFpXm1nhIKUWpePsZQ7VxWhQ/1cY/pXSmSq00S20Mpp9VmoNKrQTTWOGUqrrO18qUla5p+vpaeU1eVzS9Kqe65vurZCbX44knnki2xhd77rnnqJ+ldY3SLg9bSin6/NI4j/xo5G9qfRHLe6iP4tS6HLeeffbZWTmW2+R4MZLhiupXqiuQz1n2a+qzzzzzzOI9+6XrNMH9yJ1E94jaqrZ+3C+8B9F7sK17i9q98XinXQbqJST6kSBQuC01xT/Pk4svvjjZ7DcBYN999x31PSotwH6Q99LROsZSApdffnlWjlNsq0QI7911/8P0e55TS/RsEf1IYUX3qI17eT6qhMTrX//6ZLOUKku/AHlMHM3vKA5gIrmvYVC7T+Bxx8+uY4LXIJWX0DT1q9C09LzvKMmLA/kekfuXJa2BPG16tJ+I4tOSdFAkh9gVHH/q/flZeb8VyQ/UxqJ6Fs3XovWI+5nPYbS9eaxwnThGAnI5CI5LozOHaA2pjdnHi1L8Ge0d+1lXgf4kgTgu0TMC7kMuF0kKcv1UZobHt4710tyKJP7GA27Trs/0dL2fOXNmsg844IBkqzxLSUpN24b7jOeZroPsW/jMVD+Xz1b5ewogHzsszc22flaXRH67tA72K0dfkonTsnwt8lnReSVf4+98eKwA+djk+ukc5PrVSsL2Ix0yVmqlcUrvicpFflLbh+Nzvr9+R3XLLbckm2N/3TvzHGRJa53D/JrjHf1+kOMkHg/A8OPP6HyxVqarHxlsnYMlqSzuMwC4/fbbk33cccclW7+75f7l74OWL1+elbvyyiuTzec9+v3sww8/nOztt98+u8bfhQzi+/2IWhnKyIeWfMpYpIR4TnIMq+sRv+Zy+lmlfWP0nR3XQdfpftq+3zhm+LtIY4wxxhhjjDHGGGOMMcYYY4wxxhhjBoB/GGOMMcYYY4wxxhhjjDHGGGOMMcYYY6Yk1bm6ojRpbGvqpdpytWmHOZ2vSnGwtANLB2g67PFOVzYaUfreLuknnXUtY0ltVEoDpfIe/Jr7RlOQcposToGmqWE5fSFLJHGKQyBPXfrmN785u7b55psnm1PIah9qHbsgSrPFaaFq+5nTVmm6VpZnue+++7JrnD6L66Sp7rj/opTmf/zHf5xsnqsqU1CS4IrSEUYybZyy7y1veUtWLpLuWhtqJTJq04fWpmXUz+Lnq03/x6k7zzvvvOway+pwWnRNpV6SzdJUfPxZN954Y3atJN+l/pvnZ1dE6T1L6QmjFIRReuJSWwF52mROAanr20033ZRsTk+oaeRKsmeabpIlsxidL/xaxxT7iWGnnozSFXK/1cpEaOrsLbfcMtm1EkkRvM5oSmxuc5Zx+Y//+I+sHKcaZdkQltAC8jGhvqUkvzAeUkqRNAsTrZe1Y61WpoD7ReMJTgPK81jnTOn+tSkgo9SQUXrg2nt0CfvmSFoxolROn41f6/zk1+zr1LeXYnRNp8zPwuuUyhzwelmKc4G4LWrlJQY1J6NYph9Jutr76bzZeuutk82pjTVuYHkVlmfR2GPx4sXJ5vVO5SNL60PUFzqO+B58jeUuAOBd73pX8Z79Ups6u1aChYnWi9o69SNPp+/jdLwaG5XipkhSMKJWSrNLate4SOqL24/X/AULFmTluF00vTgzbdq0ZGvqZZZZ4lTXGleyH50+fXqyde/Iz8Xzh6W0gVzOZ//998+usT+J+noQa2MUy6ztOUytbF30WdGehFHJ3D322CPZnKpcJQU55uR+UJ9cK1MWlRu0tC4Qz8Ha9ZqvqcwA+yx9npLEo8qz8D1ZmkNjFJ5rvNeIJJL6SZOu9S2ticBg9oi8H9XP4/NEbo9onPGzaaxSkksCypJU6vP4GtdX42iW2+FrKnvM5zUcd0R7Bq17KY4br3Uw8qP9SCRF5WrlBiN4jLB0A+/bgbwPuA95Twm8WJpntM/R+0VnnJG0zHhTau/a807tE95v8ToF5PsCbh89B+M2Ybk1PXNkv8kxiX43wbKSOgZKqB9mKW3+rJtvvjkrN2/evGS/8Y1vrPqsGqI9UGk+6fgsSRDVfucIlOWTIlniCJ53HKfyvhPIvyeplQ/uR3ptUHBdIt9f6+NqZXE17uB1i8fATjvtlJXj9j7jjDOSfdVVV2XlTjjhhFHrdMEFF2Tlli5dmmx+Xv2ug2VCjz322Owaz/cojhmPMxmlNCZ1XpSk6mv3DEDu3/i7UN2X3Xvvvcn+xje+kex3vOMdxXvz2NHvWXl/wd/v6f6Ev1/abbfdsmscR4/3bwKi30UwtTFIP98L62dzfKhxBs/rXXbZJdkas3KMU9qzA7mv5fude+65WTk+r1HZLa5TqS2AMZwzV5UyxhhjjDHGGGOMMcYYY4wxxhhjjDFmkuEfxhhjjDHGGGOMMcYYY4wxxhhjjDHGmCmJfxhjjDHGGGOMMcYYY4wxxhhjjDHGGGOmJKGYVqQ1VtL7i/SJIz3uSIu3pI2mmlasO8V6nKwHqdeYSEevVn9dYT2tSN9qUPpz3If9ahv2o9OsWml8T9b71Huz9i/r3mkfPvbYY8lm/VzWBwZyPTru91133TUrt9566yVbdTy5vpEGXK0OY79ov5T0f3Wc8RjkOaMaqqwhr21QGgPRmN5xxx2T/fa3vz27xtqRrAMXfW7UvvyMWo418m6//fZks24kMBgta6V2nmtduB2i+cj3V+3bkg70ihUrsnI8t1iXkTUDAWDJkiXJjjSJL7vssmSzvrzW76mnnirej/0+aw1qWwxCHzKaW6U21X6u1eBmf/Xcc89l1/hZea1bf/31s3I8F1jfk/U3gVxH9dprr0029zmQ63FGn8sao6p9Xlpzo3bqktq1N+qnkk7mlltumZVjv9cvfH+eF6pF/eijjyY70me+/vrrk839u++++2blFi1aVKwTj79Ie34QRLFjac736wtq1wEe/zNnzsyuse5taf0Fymua6pGX6qTrZc171uZ9awOPH/bh0edrXXhd4LZT3zZjxoxkb7jhhtk11pHndnjyySezcg899FCy2RfzGgbk+sQcX6geOWvKs8Z2pIOuY7h2PzUoos8s7QOjukXXStrzQK4fzfEK9zuQr2sPPPBAslW7nGMRjld4DAD5PqF2X6RzvPQ+vjcQz5EuiPbgXMfasaXPFd2D5zFf07biucF+VDXqn3322WRvscUWyea5CeTa2Xw/1b3mekT+MNqf9LNvrqGfua51Y7/H/nH33XfPyvHeQNuI/SjHETvvvHNW7uGHH072xRdfnGzeVwP52sjxovpH3uOrzy59rlK7xg2iD9WXlT6vn36O9id6rbSWRPfgcaRnMtxPHA9dc801WTn2tdFeLlpDxmOvHlG7DpZsheM99fs67xjeL/L79D081/g8hM/bgDxG4frqXp3rG/nv6FyG5/t4xzI8VtUP8ZqxySabFOtVO3/4Hhr/8zrG53Ha3nz2wv2nn8WxKT+HxjG8Rk6bNi3Zuu+IzvbHew8YEfkDvtbv+W2/1xheZx9//PFkP//881m56dOnJ5vXQR0TvCfheCry83quFu0DS/foimjMlNbdfs+L+H66R+AxEZ1j8tkLxyD6nQP3Lc+7zTbbLCun53ur0PiA2z7qL567ek500UUXJfsjH/lI8R5jJZpb/Jr9be13iRr3RWtpP/viqByvq3xNz/aWLVuWbB4H0b5dKflYnQNRHNAvtd9PRp/N/cRjUL9f4jXnm9/8ZnbtK1/5SrJPPPHEZOv3dDw/d9lll2R//OMfz8rxmcAf/MEfJFvPZDj+4T0Mn5sDwNKlS5PNZwpA/j0Sn0tEcXSX1M7BaE9b8rfR2qZ+Ss/BS/fgz7rwwguTrd8v8TiI1m0eE3wOoHGdrrOlOkYxzyD2g1Ec1c+6G633jPYLxxc833me6Wue0xrb8jrI/lT9On8WjwE+Qwdy31vbLtqetT7UGWOMMcYYY4wxxhhjjDHGGGOMMcYYY8yUxD+MMcYYY4wxxhhjjDHGGGOMMcYYY4wxU5Iw1zynvInS80WpNEvpUMeS8pXvGck7HXTQQcnmtKKXX355Vm6bbbZJNqfPitLsRJ/L6YoiaY8oDdOgUsP2c9/a9FFRajBtS07ryXVSyQ1Od8VprzUtG6cK5PRLCxYsyMptvvnmo34up4EG6qUFSqn8B0U07mrhfuJ0npz2DQDuu+++ZHMqOuDFac5GuzeQ99NJJ52UbE2vzWkHS2nVgfrUZbXpzjnVrEoscFrELqlNoVcrOcC2pkrj9HHaJpzqjFMFcppAALjllluSzf2kfcGfzeU4bRqQy71w+m1NxcdzUtuCUzwfdthho74HAM4991x0TZSCsLQ2ReWitMtcjlPRAWVJtyg9LstH7LDDDlk5TpP/pS99Kdnz5s3LyrG8Dt9PU93xNZWW4ZSV4+E3lShtain2iGIZngtdSCfpOL7//vuTzWsdS7VonXh8aN1ZAovTd//oRz/KyrFfeOc735ldK6WzHI+09tEc7DrdZDRW2GdxjBnNQY6jVfqB07LyvdnHAeWU9Bo/8TqtdSql29Z+jdJvrg38mSoXUJvunNdsfg+nUAXyNUhjO457WKZMU69yyvr99ttv1HsDwM0335xsnrfHHXfcqM8A5GNCnzdKGV0rUTQe+4lILrU27XW0P+I08iqVxf327//+78mePXt2Vo7jhksuuSTZe+65Z1aO7z9nzpxk33HHHVk5nl88DlTap5ZSGmqgPs39WOhH7jEaS1HsXioH5L6uFEcCeT+z5IOmoedU8ewTDzzwwKwcv4/3mipjFUkT8HNyH9VK/K0ttfI4UTn2dZwOWffZ3C4ao7CU6lvf+tZkq2wm35PnE89HIF8beXyovBPHKLxPjdbgaI0fbyIpJa5n7fhhv6HprGslf6K9S600AccoLDOpbc9r9bbbbptsHV9dSJsPKn197XllrUQ8ryuRL9Kxw2Oc21VjFC7HfaPnOqUz30jWgolS9GvdOZYrSeYNCo4/NcU+n1VqOzKleEclb3jvxeMdyM9a2CerTBnXMZJn4X02x6L33HNPVo7XxVoZotoz+/GSPYvqxv4s8m2lZxiLFEoJ9XvcBzxG+PsIID/34XGleyaNlVYR7eeitYd9hD7vIGLRCPYb47lWR98RlOSTOEYF8viEy2nMyv462gfU7tUjvzkoWc9I+qj03UW0p43KRXJMpbUq8kV8D52rHLNyu86aNSsrx98tsnQn7yGBsvSgEo31QewnonFRkvWM9hL8nkg6TM/B+Oz4i1/8YrI/97nPZeV4b8GSaIcffnhWjs8ErrjiimSz7BEAfPSjH002nxU8+OCDWTl+ln/+538uXmMZPJXxHQ8ppWheRHv/UtynZ2J8Tc9kSmNc5wz7RI5zvvOd72Tl+MyNY1b9HoPPx/kcRv2o7m2ZqA2ZQfnR0eqxprowpbmqMQPvLSIpJe4jPWPme/I9+Pt6ID9nZd+osQR/P8hrKX93D+TnFLVnUkptLOGMMcYYY4wxxhhjjDHGGGOMMcYYY4wxZkriH8YYY4wxxhhjjDHGGGOMMcYYY4wxxpgpiX8YY4wxxhhjjDHGGGOMMcYYY4wxxhhjpiShaF2kSdaPPnGtDmmtzlakt/v1r3892apVdcMNNyR7/vz5yVZNONav3m233Yp15/pFGmSRfvSgNDxrde0izeaSLpeW4/Zj/Uwg16PjfmM9TiDXi2M9XtXcZT001hdkLTqtE2tFjkVHjtuG7xGV64ronqVrWi/WD+R+uPfee7NyrCeo+oGs7chtH2lUszad6p+WfIHWnecMX1MduaidWMt67733Tvall16alVMd5q4YlLYk8OL24nHNzw3k+ousY6z1Y51q1gLVPmStwdmzZyeb21jruOeeeyZbfSC3v9aJfcb222+fbNU1ZF3troj0Vde2b1X3kse1+ijWYZ02bVrxnqytyXOV9U+BF/vKVaguJWtgs39mbUgg12OfO3du8VpJD3OQRLFMaX3Tcjz+dW6tLY888kj2+q677ko2j+mSrjiQ11fHFftlnnesOQoAt956a7LPOOOM7NrJJ5+cbNbfVl9cq4s6FnRM1hDpiZZ0rdf0Pr7Gc0v1bDnu4Dmj6+XMmTOT/e1vfzvZm2yySVZu4cKFyb766quTrf5/r732KtaJY0zuP+5zINda7hIek9qf3K78TKpve8QRRyT7mmuuSbY+A9+D+wkAnn/++WRz32ibv+Utb0n2lltumWyNE26++eZkl9YpII+voj0Tj8dorxExKL/az9zWupS00rUd2K+wljuQr2Xsi5csWZKVe+9735vs7bbbLtn/43/8j6zcAQcckOyDDz442RoH897x2muvTfbuu++eldO9JMPrCD+jrun77rtv8R79Eq2D/cB9pvEh97PuNfg1+yzVNL/uuuuSzbGo6ozza/ZfrC8P5Osd113X89oYJdq7jFdsw/Rz3sL9pOOW24V9IACcddZZyeb94R577FH8rF133TXZ7HsB4KKLLkr2qaeemuwFCxagBK+n7NeBPOaJ/FZtnNAVvA5G52W19dh4442TrfEc+xddc7feeutkc9vzuQuQ+02e49y+QB5f8LhRH/rggw8me86cOaO+H4jHL7/mZxyvOcdzXfuM5xPHAzoGuRzH9doO/Eza5s8888yo9VB/xjEQx356Tldat/SslfuU21/9d7S+sc/g++vzR2eq/cJzgc959fO5nPYfzwWOHXU88HjXWJTPNe+7775ka9zL99x5552T/frXvz4rxz6V95DarzynS3Gpfm7t2b7eY1BzsvZ7h4jIx9S8R+FxzPMMABYvXjzqPbSvn3rqqWTzmZ365YcffjjZ0XcLfE3nls7rUrna9uyKfsYM11Hrz/tsXveA8pm++h1+zeeOGsfwuRjf+7nnnsvKcd+yL9Q4muMF9a/ROjQeRPsJHofsO9WPluK0aO8bzffoHgzPC41R+B58dqlrGM9d9rfz5s3LyrEvGMT5WL9wH0Vtxe2r533MjBkzkq1jlc8x9bN4Tfvud7+b7Ouvvz4rx+26zjrrJJu/AwTys3Jec88777ys3AknnJBsPrPWWJn7T7+L/Nu//dtk/8//+T+TzeMGeHF7dEXteI+ucfwZfQfBcSXvzQHgG9/4RrL57OLNb35zVo7bkuuke/VvfvObyf7gBz+YbJ2rt9xyS7J5PGsf6vsYfh/bg4g9lcg3ls7KdA7y/OS1RPeDvJZEZ9Y8Z3R94z0Hn+NobMv3i/plq622SjaPG/53II9Vou+Jo3Uw8l2MM8YYY4wxxhhjjDHGGGOMMcYYY4wxxpgpiX8YY4wxxhhjjDHGGGOMMcYYY4wxxhhjpiShlFKUqrCUlklTAZXSBEXlVAaAU/hyKhxO6QOU0+pqih++3913351sTs0FAHfeeWeyNT1wLaW0VJEESpfUpkeL3sOp/TjNI6f+BPIUWZq2ivuDbZWGYNkrLqftddBBByWbUyxpytBayR4mSiPJ7xuPlHjRZ9RKKbE0DreVpnbU1wzPjUjWglN3bbbZZsnW9JBMbXr3KMVbJLPEKe045TqnIgXy9LKDojZtdyQ9xSnQojTOmvJvxYoVo9rqR/mzue3Uj3JqQE5JeOCBB2bl2C/ze3ScchpGTUPLUhmc2k2lEziN3CCoTa1XK4ERpVvUMfD0008nm1MaaopG/mz2r5pqnuc0p6hkCREtx+ON5QWBPK0oS3XpZ5fkDAZJPxISUd36kfYB8tTlbHMaViCXa2G/rFI8PHf32WefUd8D5OlqeQ3QOc3jaunSpdk1jpuOPvroZA8qRShTKwvB9CtDE8GxAbejxofcdixJ9e53vzsrt//++yebpULOPPPMrNysWbOSzc/x+c9/PivH44PHg8Ljl+VhgFxetEt4DKokQKkPtRynn95xxx2TzXEpkKfHVx/LEnLsizRVOcv18Fqqqfd5D8FSgSpVx744SlEfyXzV+rFB+dWobqV0tFE9oznJ80tjtTvuuCPZHNOxrBUA3H777clmyRFdnzjFMqdU5jEG5LHHhRdemGyd/7XyurzGa1r7QcSj/ax9kYwJ+0NN+xvF/Nz+X/7yl5P9h3/4h1k5nkO8H1dprQ9/+MPJ5viffSqQp/PmcioTVJuaN2I80jD3I5Wsr9nWecbtoPEjc8455yRb5zT7xC222CLZOrc4LTPH+5qym2URIniNiyRCmPGQ4tE1rfT5Ufp63tOzXNx//Md/ZOU4rtSU2HzOxvst3VNxTMF9pnstPnvh++k6yHMykgCpPfuI5ucw5mCpbtHzsa/U86dSOnUgjylY8lbhvQZL+6h/5H7jcwZtx5L0USQJr6nReVxEkiCDllLSvRL3H+9pdZ/D7X3YYYclm6VOgVzyM5If5rbX/SXH/xzvaNvw+sbrpbY9Sz/xc6lviuK90nn+eO3poxi6VK5fWc/az+K4Rr/H4D0zjwmWWALyucprrs4tHi/sB9Qvq2QCU4pztF2ied0v/czryNdG92YJnGi/xW2l3yVwv/BaqnsJXi/Zr+t3HXx/3ptEUlgaU5ckP8fDh+rnRHvVWokktnU/FI1PLluygdw/3nbbbclm/woAhxxySLK5r/mMGsjlkvkcYCwyUKXvpcYjFq2V3qmV8mSfdNVVV2Xl+DtB3W+VzmRYmgjIz9L4vEbXS/4OgqWZeP8B5HuLN7zhDaPWFcj7TMcKx1MsL/v+978/K8fn910Sfa9ekuHUMzFuc4559H7s2x566KHsGj/f1772tWTr9wR8rsHvUTnqiy++ONnvete7kq3rKvcp+1TeIwEvljZnuG3GWwI5ii3Yz/Ec0f0g15m/r9H1gs9J9bsE/i6Ax4CeQ/HYOffcc5Otc5plPjlWVulI7c9VRLJHkZRUdO5UizPGGGOMMcYYY4wxxhhjjDHGGGOMMcaYKYl/GGOMMcYYY4wxxhhjjDHGGGOMMcYYY6YkoZQSoymValML1kopcXodTbXDcgGc/kdTU3H6ek4NpelCOaWSpkVnOMU8p9liiRhFU/yUUmprup9BpW+KUnqXUt1p3fgap1HSduA0hJp2lMcIp16LpE84LZamW+JUy5yiUNN2cYo1rlOUkjFKCT4e8kklatPe6Xzk9ua20tTKLBegEhs8rjmtpqawmzNnTrJLaWKB8riMUoRFYzlKfV6Sv2L5GODFqS67IkojyXA9NaUtzy2+B6cIBfJ0n5oimPuX0z6qDAXXI0rj/Na3vjXZLL+h0kylVJGanjWSWeK0eOz3WVoDABYtWlSs76CJUk+Wrqk/4XmiPpTbi/tM5xaPf74HS4MA+dzlFKHHHXdcVo7HEcvn6frLaX85LT6QS1occMABye4i7V0NUZsz/Uj2jAVuM5YDYVvLsbzBsmXLsnK77rrrqJ8Tpc6N4jD2+5tuuml27bLLLks2r81HHnlkVk7jty6oTenNROmwo/UiGgM8Bzkm0ZS9zPnnn5/shQsXZtdY7oj78jvf+U5WjlMAc5pXTZ//2c9+Ntm6vu23337JZrm0SCKjS9jfRHISvOboGsZrCz+fyiD9+Mc/Trau6+yb+LO0LUsSZrzGAvlcYNmQSGIsSkEdxQulsTlecWk/a1xt6myFfRjLGwH5Xo99p6bP5dTG3Ncqm/vII48k+6abbko2SyQA+VobxUa1UkqRXEo0R/qldt9eux/lOupegNcS3efxPoFjA01Lz/3J99A0zhzbct157wPkbcz7H5WTY4k77QceA+OddnlNlFKx69jicpyKWucFp0Nn+VUg79+f/vSnyeb1Dsj7cOedd062+tuddtop2SwDovJoHJfwHpPPfIB8nYxSVEcpmwdB9Nm1PpTbntc6TY/NbaLSujyHuO1VTov3ihz/q/9jGU6OX/XshmN+XmNVcqnWb0YMqm+j8xZul2itK8UAKn3Cn6UxBcfa3P4ao3D8yKn/dQ/HPpH31nrOUCv5xX2o54UlOSZtz0H0IfsNjTG5LyLJXL7GfaZSCxyzqkRD6bm1rVhints0kkPnmOaII47IyvGaxvNxLPKcpTh1IqyJpTindu/f72fx/NE9Pe9XojHNMQtLFWgsw/s2Xg90L8o+I5Lpqd0TD4Lou4mI0rjT+rIv02ulMzddL1euXJlsjlP1vJM/i+Mp/R6KxyL3n9aP/YzGZyWp9H7bc6yU1jqgvH+plVyqlepVOHZkvwnk8Sf7bI1ZeT/A8+nwww/PyvGZKkvo6bl0rZQSU1uuK2rvr2sir6Xs8/QMeJdddkm29gvHJ+wnWToJAL70pS8l++STT062noHz9wB8ZqpyPX/1V3+VbB4PkXScjmvel/KaqwzqjKZW3prHk0qCMbzP0/0E30O/I+Q+Xb58ebJVFv5Tn/pUsnlu8dwEgCuvvDLZp556arKPP/74rFzpTEbjaN7H6PfEHANFskqDmIPRuOC4m+N77RdeB3gvrn6yJLkE5M/N79O5UPodg8asfIbC66rGPqVYKNoHRNdqZXYjnDHGGGOMMcYYY4wxxhhjjDHGGGOMMcZMSfzDGGOMMcYYY4wxxhhjjDHGGGOMMcYYMyXxD2OMMcYYY4wxxhhjjDHGGGOMMcYYY8yUpCzUJ0R6d5HOYEnrUzVrWXtVNexYK4zvz3qe+pp1H1V/jnV1WZNVde9Yt+t73/tesj/wgQ9k5SKNN9bjYh2sQenNKbVa94yW43ZlnUzVKGO94xdeeKF4f9bk03toX61C9ey4La+44opkqwbc61//+mTXaulyOaCst6b3GESfRvXka9xnqjHLGpCPPvposlmTFMh1MbW9Vdu6BM/dqD1qNYlr9aXZt2j/cdvwZ2k7sYZfl9Q+a6Qzzmy55ZbFctyHqg3Ir/nZVaeatQa57T72sY9l5fbaa69ks6+MNGRLfaHX1Aett956oz7HnnvumZVTXeAuKGlpK6W1LioXaT4rPLeeeuqp4mex7i2vg/zvALDJJpske5tttkk2+wsgH1PcL6pPz3VSf1HSG1YGpaMb6U6WrkWxjPqYWrj9WAtZ9aeXLFmS7KVLl45aBwDYb7/9ks1r56tf/eqsHI8zLqfjjcupf2Td2M9//vPJ/ud//ues3CmnnIKu4b6oHUuRTjrb2gbRWOE2YT+k85jjDvZdl1xySVaOdVi32GKLZKteK8P11TiLP1dj7IcffjjZHKvpnFNt6K5gvxJphEfXuP3Z1+tax+uRxvXst+66665kazvw+9j33nfffVk57l9+TxRH8riKdIxrY8x+tXTXhtrYNIoHornGY5T3h0Aen7Jf4vUOyNchtnWN4/HD/bnVVltl5Ur9pj6JPyvqw5I2MhDPg36pjUuiclxnfm7WtgbyPuM2BfI24Lj7gQceyMrtsMMOyeY+17iX5zFr2fP7gXwe8zPeeuutWTl+/da3vjW7xn6V19JI97pLIt/B7Rqtcbzv5vdcd911WTkex9F+gv2tlrvpppuSzesdnwMAwNZbb53sBQsWJFv11BmOc3QO8riKfCw/v/rsQcSjGpuViPZK3CbcZ7w3BPIYX/0JzwV+38Ybb5yV43l80UUXJftb3/pWVq7kQzfddNOsHM/VFStWJJtjFyCf4zq3SmdZUT8Pito4s7ac7p2ieIDbmfdpDz74YFaOX3PcpLHe7bffnuzddttt1M/R+nIMyn5A2WCDDbLXpfVN+7bfvVYE1//ZZ5/NrvH8jPaDHEfyOqD77O233z7Z2t6XX355su+///5ka3vzfoDrpOd0epazitmzZ2evea6W1gy9Fu2HozhuULFp7dyKYtF+5mrEjjvumGxdt04//fRkH3jggcnebrvtsnIPPfRQsnluaWz0yCOPjFpO51UUB9TuiQcRi0bUfldRQvuL54WOAfYvPJ/0TIb7hX2t+jz+voSfY+XKlVk5PrvktU7jIvab0fhlv67tN6h1MPIPtd8R1p6vRte4jdh3XnjhhVm5t7/97aO+/5hjjsle817mscceS/bXv/71rBzPE36Pfjc5a9asZOt3XOMRo9QQ7SV4jkTxK++Xdc3h+cN+Uu950EEHJVvXtw9/+MPJ5jj1kEMOycrxOdHf/d3fJfuaa67JypXWBvW199xzT7L5HA3In+uAAw5I9nidydR+n6DfqzPsH3m8czwIAHvssUeyNc7hNmMf9o1vfCMr9573vCfZPFd1zeHPuuCCC5LN657Wnfca2v6nnXZaso866qjsGo8/rsd49GH0HX0pFtU9As9PPtvSPi99Fwfk/ouv8VoH5OsYf4esZyilmKT2bCnyR/3+lqEWZ4wxxhhjjDHGGGOMMcYYY4wxxhhjjDFTEv8wxhhjjDHGGGOMMcYYY4wxxhhjjDHGTEnC/HhRirPatOAMpzvT9IGcllLTN3FKO07do6lzZ86cmWyWddBUd5wKndNozp8/PyvHaZ44/aum9+KU31H65Nr0/V1SmwqK66OpkzjNH6cN03TW3L/6fJw+7vHHH092lKKVpQU0/V5JDkLTtTL9Sh1xOqYondOgUmevQsdIKbWWtim/j+eCyjBwetUopWokKcLpq2vlAqKU0aXxG6XXjT6L/Za2E4/LLqlNX899qGOLU8RxOU0DzNfYBwJ5ql6eu+uvv35WjuW23vzmNyeb0wQCuWxBNLd4XkRzJPKBnPqdU4Zy+nXgxe3RBbXyV9F7olTDJXhdAXKZDk7tqM/MdeI5rinueb3jdlRJBF4PSra+VukLvn+ttNgw4L7ROci+Q2OAWtjHcL9p2l5OuV3yX0CerpnnvsrHlNLs6pjgFKSa+veWW25JNvcvS1cMCo71audPVK6f9NpAOaW0pubkOcR+V1NbcorQefPmJVtTrrOP5vSaOgcPP/zwZKtU4mWXXZZsTvWrY+Xqq6/GICilBY3QOciSbzz2OT7XcuqLSnKs2uacMpdjBY17eWzy/InSlkfjr/SeqOx4+dGuU8ty/+rc4msckwCxvArDfcNpnlVygFMxz507N9na1yUZoSjVfK384qD3D1qXSFKw9h4cl91xxx1ZOfY/kTQc+wVNk8zzriSzC+TnBxwbcQpzIJ/TvFdRSV+WuGBZQyCXVorkXgY1J6N0wyXpD90vsc9nnxXFfjNmzMiu8V6S1zj2vUCeGp3HhKZk5zHCn8XnOkA+93ld1JiM03nX7iuVQci4RFKQ7PP4s6M4hG31IdzGejbC5yssuaRwam7eT2haepZZYsks3V/y8/O4WbRoUVau1m9G5dZWkqOGKOV3JDVRuhatg9q/vI7xfv/uu+/OynH8z/sOLVeSLNUYqnRmo+syz0n1sfycY5E27wKuv8ov8747kgApvUfnKs9B3atzWd6Lab/w3o7PgtSH8vks7zUimfBa6c5ofYvuMWwppX7KRZIUOhdYEvnGG29Mtu6ruBz7xG233TYrx+3HfaiyMDyWeB+jEsiRNEStRNIgYtMojin5g9qxpO/nWCWSqWdfoJK57F/Zr6n/YH917bXXFj+L5WQi2Vi+n9a91E7DmIP6mbx+ROc3JR8T+SKNobgsn0fxnANyORU+/9Q5s/feeyebz+z0fnwmxuh3mFwn3ccMQwZ5Ff2su+oL+DWvUxpT8hnnDTfckF1juUH2SdwPQC4J+aEPfSjZ3/72t7NyvO9jH6rnb+wX2J/qOs37Ij0n4j0rz+nx+o43iqNKcZp+18XxBscQOo45/lf5Ko5feH3j+wG5BOtJJ52UbJWQ4/0J/15A5WJ5brEvVl/J11TS/sgjj0z2eH8nUft9CO+dVTK8tMZH67uegXHfsn/VecznaOyT1Rfy9wc8L/Q7h35+o6CU7tFv3OKMMcYYY4wxxhhjjDHGGGOMMcYYY4wxZkriH8YYY4wxxhhjjDHGGGOMMcYYY4wxxpgpSV0evVEopf/R1D18jdM+ajoxTu108803Z9c4lfFtt92WbJVj4lSsnDpS00ZxiiVOiafSFZwWjFMLcWpZIE5/VJKMGa/0abVpmjilmKap4zRznMZKU61FUjCcRpXTrapcAxPJonA6M27LXXfdNSunaZtWUZumFyinfIzKdUWUvrWUKptTrQJ5OmxOB6lSCzwGIhkAvqbpq3lu1Kbp7CeFe5QyLkpVyOU0Da1KrnVFlB6S25LHKqeRUzglnD4r+zZOLw/kfcW2yrjwHD/++OOTre1TSuFXK5ulYyxKe80+iNPra901bXfX6HwsPXetfFaU8lTnD/tQnuPazzw+9tlnn2RrWspSOn3240DuJzj9no4Hfh+n2wOAjTfeGKOh7Tkeqc/1M2rXSL4WpeNltBynlWS/rPN4p512Sjan6lTpMF5LuT80jTOnyoyel/v30ksvza6xz+A5qBKUg6AkERERpcMey/sY/myWJtCUo+edd16yeT7qvTnFJKfpVRm8adOmJZvTdWuaUp5nKovB9eCxopKC+ixdwbFBrexIJOHKc0ZjRU6ZG0lI8P1UlorvyW2p7cV+j9fwSDIjkqntR+5moknS1abWj8px+6k/4/WP5XHUj5bGgcq/HX300aM8xYvnF8+bKD175KNKKaqj1OFdUStBEsnwcNzM8dbs2bOzcosXL052JFvK/VeKE7QeOt95rLDEi/bf9OnTR733Y489lr3mdtJ5XIpRhiFnVithqvtg7g/2h5zSHACWL1+ebE3fzDHjggULks3SSUC+J3/ggQeSre3IdeJ5q+3KzxJJM7GP0DE8zHMZrSdTGkNRXBytJfw8KnNz8MEHJ5vbVPdULGvFfaR7LZZSY4kkjnGAfB/DnxWlnp9okoK150eRFE3pWjSndc/FcQnPLfVn/D4+O1Pfpn66BNeX56qObX6tKdl5r8FSiZrinfcuJ554YlX91gTHbPp5PD4j6XRuA5Zu0PiZ5R9UsoPPsHlNVF/L84llc3Q/wfePJM9L1O5/9XV0bjHs2DTyI6W9Xq3kEpD3L0t6nHDCCVk5jo/4PRrb8nzicix1AORzmv2ojrHaNY3fp/HCIM62S7KBQHl91nJcr2gd57bSucXtzXOQ55neQ+WTmNJZuc4D9t28H9EzdX6t/pWfOVpror3o2lCSdIyo3Xfo/fiZ9Ds4/s4wint/9KMfJXuHHXZItspH8xnLcccdl+xDDz00K8ffR7KUiMpw1cp6jzeRbyyNmcgX8P5N+4/35ip9zfK3LC2nsjn/9b/+12T/zd/8TbLVN/Jen68deOCBWTk+O2e/oPtLnvt6Ps5jluOAWtmztaXkK4Hct/H5lu6zue+53ip/eeWVVyabv1sA8v7mvYFK8J511lnJPvbYY5Ot3xmw/+W9nI4/9gW8hun3Irz/ieLc8Z6PtTLcvF5E8RyvTSpBzm2l+zeeM9xWuh/R89TR6gfkcmk777xzsmt9oT5jrex4VK72TM0ZY4wxxhhjjDHGGGOMMcYYY4wxxhhjzJTEP4wxxhhjjDHGGGOMMcYYY4wxxhhjjDFTEv8wxhhjjDHGGGOMMcYYY4wxxhhjjDHGTElCUUDWe1JtppJOVK1Ouuq0sRb4nXfemV1j7WLWEmRNXQA4/PDDR72/ajayvtjrXve6ZEeahjNnzky2PhO/T9ulpAWm2leD0raONARZ8491+Fi3G8h1FFlrT3Xa+H6sCQcA06ZNSza3n2omsz4a10M1PVkT8pRTTkm26vpxO7NenpaLtMe4DyO901r9srEQjQtu7w033DDZqkPK+rOsA6f9zPNEtTn5Ndv6zBdffHGyeT6qnl1pXqj/4PZmjUTVgOTxodf4uVizkuc0AGy//faj1qlLoufbbLPNkq36nvzs/Dyqrcm6iqr5x+9jDUG+NwAccsghyeY2Ut3yWk3Z0hiONNf1uRhuJ/ULixcvTvZRRx1VVb81Uau73Y82pc6z6HN5XrN/5b4E8rHDmpuRRjrfg3WWgVynkseX+m5mr732yl6rVulodQAGo2UN1I9VRn0bv+b4Qscq63iydi6Qxzncn6qly2N8k002GfXeQL4GLFq0qFgn9gWsL6ux0Q9+8INkP/XUU9k1jpW4vqxjOyiidZD7lsdPpBNfe28dj/xZ3PY6B1mHlf2mak+zzjX3ra5NXMfly5cnm+MqfX355Zdn11jLnu//yCOPZOV47HWJxiUM+ya22X8B+fPxmFYdY/ZZrG8L5LEHt7nGDfyabZ1bPC8iuA9rddbVB5XeN6j9gxKtd6V9UPQMUZuU1jGFy2ncxJ/NPuuYY47JyvG4evDBB5N9++23Z+XYJ/I4jXyNrrtcJx6L+vyD6NPaGCXa+3M/33///clmTXAgj1HYVwL5npLb9NBDD83K8fzk/YquW9y3XD8+OwDyPcnKlSuT/dxzz2XluO11TJXmQBTbd0mkoV37maW4UOcZxx4a7/EemmMP1Z7n/QXfX9dC7l/duzAcM/Pedv78+Vm5jTbaKNk6Xkr9FsXIXcHjKYr/I0pxso5B9i8c9wHA9OnTk83Pef3112flDj744GTznN56662zcrwGc+yhe39uY/YLeibA7VTrtwY15yKic8NofSutg/oM0V6Y9yEc8+hcZf/GY0djPZ53/cQ8Op455tQ48wtf+EKy+SxX9/T8LCeeeOKonztWuN30HIzXEu4jHZ8l36B7gYcffjjZ22yzTXaN17do7LL/euaZZ5KtfoDjGH6Pnnfy/oQ/N4pB9Hm5bHQGPh5EZ0ls1+4jI3+jsQy3JftKjnGA/EzkoosuSvaSJUuyctttt12yt91222TzOALytZTnlvqj6Fya26PftahfonFX2g8qpX7SecHnT3oWxftzbmM+6wJyP8Ttpu3N847Hhp6T7L///snmvuQzIiD/nkWfl30N31/brJ/zr7FSu0ZHczXad0TjmNuI+433ckA+xnkNYJ8KAKeddlqyjz766GTvtttuWbkDDzww2TfccEOydS8U+RbdX4wnX/3qV5OtY4THZHSO/IEPfCDZvJ7p94Mcf86ZMye7xvFJ6XtJALj77ruTzfsM3XNwzMm2ftdRij+j7yX1nIjj3q222irZ6oMG1c/Rd7yl71o1Jt9iiy2SzXNBY5l77713VBvIYw9uV90n8D2vuOKKZOuZ58KFC0d9jmXLlqEEl9P25/HH9x420VkZ+ygeu7qW8LPyWNU1ncenxrP8Ph4rGu/wHOT6aTk9e1mFPmNpfe93L8fP1e/3S84YY4wxxhhjjDHGGGOMMcYYY4wxxhhjpiT+YYwxxhhjjDHGGGOMMcYYY4wxxhhjjJmShPnVatMoRqm/GU5jo2lyODWQpnzllLj8uZqWklPYcXpmTcvLacE4naym7eIUm5xCLUoJ30+qamB80i4r/HycfkhTLHG6JE55pumJuZ017eMuu+ySbE7PrKmbua+5TVTWgdNKnXrqqcnWFOk77rhjsjnVl6be1zRQTG0KfE0d1zX62Zwmi59N06aW0sBp+nCen5qCi1Nrcco9TQ/H9//mN7+Z7I9+9KPF+/Hnapo6TrGr1xjuP02Dx6kyWepJ61QriTBWePzo3Fp33XWTzX5Fy3H6uZKtrzU9JL/mMcFzDgCuuuqqZLNsh6a6K6UWVbh/+Rm1nzhFm853fi6WvOL5DQBf+9rXivXol8g3l1K+RlJ5fE3XHJ53mi6PfSWPDx23XK6UOlTrweNB257HTeT/OR2jSiKUGA/5CP2cWmmsKO0exzkqb8REUhycYlLbkn0dy+3o/VgaYt68eclmSSQgTx/M8gPf/va3s3KcenannXYqXps9ezZKRCmS+6U2HXQUs/YztiL5Sfbdmhp27ty5yS7JZwHA0qVLkx2l7OXUvCxfopJl/Pr000/PrnEaTY7VNDXpkUceiUHAYzDqGx4zOrc4xmI/p2mA+bX6ogsuuCDZvAZpzMNxFNdJ/WgpjXzk22rT33YhBzYo+v3M0noazXG9xinKWa5BYb/KflT9F8sHclp6TWvOPpDnuKZn53UySufKz6XtMihZwVXonCn57WjfUZKFAvJ21P5jSRaOgTTtMssdMepHeQzwvNU9DteR66Bp87kttG9r15tBwe0fpeeO6sN+NJKF5DbS/Re3LccK+rncH7zuRGm6uZyOKx63bPN8BOqljSPZhq7il1pqJcpLvkGfmduYZT70npxanc9ngFz24+STT062ttW1116bbO7za665JivHY4r7VtffaGyX+kXrNKhztUhGr1bGpeQ7brzxxqzcHnvskWzdq/O6xfs0Pc/iecJSLer3br311mTzuQmf3wF53/AY09iI49YvfvGL2TX27VEcFp3N9Qv7OfVD7O+j+c9rEMepvC8A8jNm3nsB+VzgeadrM7cP95nKU5XaKjqP4Dmo/qM2nmXGy2fW7vVKUrtjIVpL+JyG1y2W9QaAK6+8Mtm8lqq8B/ti3tPzWSiQjzOdxyW0nUqS9l3snbuiH7ly7Wc+m1I4xmefFMmPlWJ4IB8PHGfpecoBBxyQ7FtuuSXZOh7Yz6hvZH870eTMGG6jSNapX2l6Xp903jEsk8dr2GGHHZaVO/vss5PNa+K73/3urNyee+6ZbJbzUQl17jf9zis6Cxg0HLNF+1TuM5ZsBHJJMG4DvR/PBZUv5PWT10SVCme5Kv4eoBb1C6XviXUfxHNS78E+9J/+6Z+Srf3MbajfPXVFNGe4Puqz+Jn4+wSeL0B+bsJnl0AuWc5rmMaEJYlsba8NN9ww2bNmzUr2nXfemZXj+7Ov1DXxpJNOSvYw1zMl6jP2Gyx7qv1X+i5Y91C8pmncy2dlfI33GEAugcxnANF3lrXfCUT7pdrYjcv1288TZ3QYY4wxxhhjjDHGGGOMMcYYY4wxxhjTIf5hjDHGGGOMMcYYY4wxxhhjjDHGGGOMmZKEUkr9pESMUtFyihtNi8MpfzR9OqfWYgkcTe173XXXJZulIVRyadttt002p73T1EKcoonT0EdpqPRaKVW2ph0aVNpsTt/FbQLkz8vpszSFIKfFiuBUWJqK7Lbbbhv1/poalvuXUzutWLEiK8dpl1l6S8cEtzOnstR0wVG6wlJqJk2ZqunhuoDHj6Yk43nC/af9XJJC0X5m9LM4RWuUnorHALe3SkNwynROt6vyPwynuNX6cb9rqj+uE6ezjOZxl/B40rTs/JrHu84ffs22pt2MJHG477n9dBzz/f/xH/8x2Z/97Gezctx+/IyaKpP9B6eDveKKK7Jy7B9VTo8lZPiZNT3psccei2HRT5rXSN6oVloiklxiWyUFeV3lsaI+lP0at71KYbEEi6auLaXSi+Qjukx3GKVorZU44de1afo1dTZLPnDaXr0fz0n205pa9D3veU+yuX+1b5hLL7101PcAeZp1lXfi1xxDKYNIpd2FRFLJv0f30/fwfOI1V/uZ/TrbGu+wfNznPve5ZKskJMdF3GfqP9inap14fvLarHHLEUccgUHAbReldo/g5+W5wCl2AWD33XdPtsavJVkkXXO5vlwuisM4jlTfVvIzOvdrpd74fYOW3RkN3cNw3TgGqN0T6v3Y72lMwXMvkvbkz+I5pPfjOcWpY9U/8jrJ6bvHIuERSS7W3qNf+Ll1nPUjN8j7KO0/3jOoXBv7wTvuuKN4D06nzKnnWT4XAM4777xks1/QeLuUxj2Si4qkEidayvpSH+p7OL7gdV0lqjj9ue6rWOaB4wGNPUpSzLrucOpr3buU4PTSKqvK4099bG08Oog5GO3p9exhFTpuS1Io2s+RJBif/3BbqSwkn93w/lLXS4Y/S+Vd2L+yH9dyfI9oramVuh4UtVKIWjcux+dbKqXEcqmRPDLPLZUy5/00rzkcfwJ5/7J0ssatPB55vuuegWVKWAJU4XEfyYV0Be9jdWzpfrpUjtcFXqdUEoDHsUoz7rvvvsnm9lEfyn6Zz8u0TpEsElPa40QxgVKSDBvLPdaGaE9Y+kwtx30V1ZPfp33D0g48HzVuYH/LY4zjGgB4xzveMern6vrGc5LXTj3XjeQ6SzG7zsFBxzm1+5dofeNn0XnBUhwqr8lnkmyrRGvp3E79BV/jtl+4cGFWjmMXlizheHi0+jKleFbbaRA+FIjnPdet9iyT92/Rvliv8fjnNUz3/ux/eW/H4wMATjnllGTz+svfO2k95s+fn2yVj+RYV/1CSc5M23MQsSjvA3TvzJ/Hz6n+haU2WUpJ4b0Ej339bL4/vwfI99ws5bN8+fKs3KJFi5LN8zaSaWRZQl5jgTjW5Xtwv+tn8brRpZRSaR1WONaOzpvZ70XfTyg8rnfddddkqxwyv2Z5a917cizJMZSWK/WNxmEqJTtRiNZW9g38nal+78f+hcejzlX2jbqX4HZkv8kSWUDu13m863f0PAd5jVR/V9rzRnFkJLMU/c6kNo5xxhhjjDHGGGOMMcYYY4wxxhhjjDHGGDMl8Q9jjDHGGGOMMcYYY4wxxhhjjDHGGGPMlMQ/jDHGGGOMMcYYY4wxxhhjjDHGGGOMMVOS0QUKR6jV7uVykS5gpDvPsBYaAGyxxRbJZi1r1bQq6b9vuummWTnWz2JtwmeffTYrx+9jDUhti0i3qqTJFrVTl7B+l2pVsq4f65KpfhnrE7PWm2q7sR6d6iiyBhrbqiV+3333jXqPm2++OSvH42errbZKtj4jv481JVWbM9IvK+nxsrYiALz+9a9H1/DnqX4z62Jye6gGLmvbctvr/bj/Ik1F1otTfdWSzzjttNOycqwtyFrW3Jd6D9Z9ZF1fIH+Wgw46KLs2d+7cUetXq2u7tnAbsZYjkOvtcb/pHORxzde4P4F8Pmn/st7gBhtsUKwTz3fW3Tz//POzcieddNKo9Xv00UezcuynWT9Xxw7rjt56663ZtWnTpiX7TW96U7JZMxEA9tprL0wEojUh8vWRLi/fk/11SaMxeg+Q6yk/+OCDydZ1kMcYj1FeCwBgwYIFyY7Wd76mepODWgdr7xvpGPO16H6PPfZYsrUt582bl+xLLrkk2TpXOfbgMf5Hf/RHWTnWUeV5rNrKrF/P/kO1ftlH6HjhZ2b/PWzYj0e680x0LdLN5vbme2g8we3K69vjjz+eleMYjP2u6j2ztvUb3vCGZB933HFZuSeffDLZRx55ZHZtww03TPb3v//9ZKuW+m677YZBUNJDB8pxg45B9h1RbMt6vDz29X0c86gvYv/G1zQ24rnK40P1fUv7pEgXWsdpqWytdm6X9OurS9rl2tf8WnWlS7r0vFcE8nHB5bQPS/tA/ncg35vutNNOyVb/zf2hayG32yB042uJ6sW2xsm1PpDXII1neU4+/PDDyVb/OH/+/GTfcMMNyVbteZ6DvMayDeRrM+8vI0159Qv9nIt0Se3nl+aZ3oN9oMYDHPvpOJg9e/ao99N4lNuydA4A5LGl+k6G68H9y3savV/tGU00JwaBjq3SnlTbtHSWxn0J5HssXQf5jOzGG28c9d+VpUuXJnvGjBnZNY5f+Jo+I/sCPptQHxr5yZJ/1XKD8q/6TEw/Y6YU3wH5s6qf4rVq5cqVyWZfCeSxH48lPX/j59pxxx2TzeefQL7H537j/gTyvYueM/CzcJvp2jyIOaj7MkbruQr1ITxXOVbR82Ze3/R8i8+mnnnmmWSrv+Y4hj9L11zuW453eM3WuvM99BlL5/daNjo/nUjU7vvGsnfkPp0+fXqyNZbheIjjlxNOOCErN2fOnGTffffdyY5izOg7GKZ2LkVn4F1Re0bEY1rXR34e9l26X+DvjdjXAvk5GLfdNttsk5XjOch+TvfqvEfgfta9CZfjccP9DwDLli0btQ5A7qs4ntV20rigK2rnSeQfSuM1WvP1+fh8mONSXhOBPEZZsmRJsvmcG8jnJK+d/P0UkJ9ns//W72B4DNeeg4wHPGei9ub669kij09uA/0el/uM21ThMa7tyHONYxKNd3hvzmufxjsca3BcyvtELadzic+JonOdnXfeGYMg8umlNU3jV/Yd7FPUj3LswWdsQN73kS/mvQH7RP0+iMcc9w1/7wnkZ54c122//fZZOa3vRCFar3kfHO05Snsq/l4HyNc6XWd43xfF4+xDeR7ruTF/T8d0EVtEZzJRe9Z+5ztxI1hjjDHGGGOMMcYYY4wxxhhjjDHGGGPWAv8wxhhjjDHGGGOMMcYYY4wxxhhjjDHGTElCKaUotWBt2kGGU1BpWkBOcROlfeTP1ZSXnJaJ0wRpSi9ONcR10nIf/OAHR70WpdTUVD2ltqktt7ZwijhNP1SS4NDUgNxXUYrpWhmQKN08p6rjz9I065z+n9NIabpgft/b3/72ZG+00UZZOW4LTfHMqVB//OMfJ/u2227LykUpiPuFU8lpCjt+Bm5HTT/HryNpJm4rbW9On7XJJpskWyWzuG+5TXV8c9o0HissdwXkqcDuuecelHj/+9+fbE3xzHO8JLuh9e0STg2ofchziMexpi8r+SxNXcyfpT6G5zWn/9MUdvw+Tj/3wx/+MCtXkqFQmSt+Lv5cTS140003JVvH5imnnDLq545HytAoVWHXKYRLKe6BfI6zHaWkZx+hfoHTRXIaRE03yXOSP3fPPffMym233XYoMeiU9Guitg8jonTZzOWXX55sTvEK5Kl12db7cfpJThUZpQJkX6zyOOxHeYxFfkbXcJYqOPjgg4v1GAS1Ekkcg4xF8rJ0D10TSmkaNWbiNZLXTh0PnPqTUzcfcsghWTlOC3rAAQeM+n4gX3817T7HVixVt8cee2TlVKakK2pTtnMba3zIz8t+Suus6wfD92fpEJ5nQD7XNB0sw/sObn/137VSbNHeqh9JuC6prXft5/N7dD/B/lbvx3EUzxttc5Zy4Gs6b3iulMYYkMuisc/WOCyiNNbV1wyiD/tZBzX1d2m/qz6Q+1PvwSl4OV7n9OZabvny5ck+6qijsnJHHHFEsjkGVv/Bewhub537fC2SRBt2XKOU4tEoTmbfGe0JdX7yvoHbS/dwPJ94Pqq8lu45S3Xn+i5cuDDZOvejlMol+R31t4OQ4omkf2opxYF6P95vsQwSkMu4sJSS+jyedxwvakpznuN8XhadtUTyWdx/0VrKz6zzcRjzs3ZPyPOJ98IsYQTkZ3icCh0AHnrooVGvqb/lPSKvVdrmPK5YjkDjH+4Pnrcqbcx7TK0T+4/IZ+u60gW8343kUqN6cD1LZ9RAHlvsu+++2TX2lVwnbSv2y9wXKmfM7RhJeDNRXBqd6ZbuMR4yPEoX309E62V0D/3eoPQeLsdnJzrfeR6zHJNKAPG6yvvN0vgFXtyHkaz1oGEfEu1zSrJKQF5ntvVMk+cJn+cDuSwcx/e6bnEbcz3U57F/jeYWjzH+LF6Xgdwv6J6e68FjRdtpUGfbtWOm6zhK78djnse4SvbwOQp/X6OSq3zGwnsDjY14fePvg3Q8s8+OJIvGO15h/64+j/uW66WxALfJ2WefnWyViON5ofOT+ymS0mafwfVQv3bNNdckm+eTSiTxuSjPM41B+LXukfgsgs9FDzrooKycSql1ReRjSuuynvuyD+N21ZiHz4A19uCxxGsV+1Qgn4Pc7/r9RGk/oWeoHEPxWDz++OMxGYjmP493Xqc0PuRy0W8aOP5keUEg34+X5oXWg9+j86J2bSj5oOh3Fgp/ViRjbiklY4wxxhhjjDHGGGOMMcYYY4wxxhjzksY/jDHGGGOMMcYYY4wxxhhjjDHGGGOMMVOSUEqplEoLKKf/0XJROtSazwXyFD2cdkdlSTidGqfg0vQ5nP6HU0+edNJJWTlO+cREaRX1s0opAceSJmht0JRZDKfX4/TEmiKLU/BzujiV8OBUfprajNOdc3onTgULlNOTairA+fPnJ/vYY49NNqf6AoCtttoq2fyMN998c1aOU41y6logT8vGY2KfffbJymmavS7gcaISQTyeOA2ZpiTjMcDXovkYpT7n8aFjhVPkaTo6hj975cqVydY03NyfnD5dZau4nD4Xt1MkezYo/vIv/zLZ+++/f3Zt0aJFyWZfpH3I7czXtL34mUqpyYHcV0a+iOexpj+88MILk80yLpq+jeWYOJ2izv3rrrsu2R/72MeK9Y1SsA6iT7kNIom4flC/xvNOU81zv/NYUVkIHu/cL3o/ThFaujeQr798vyOPPDIrp7JqNYxHmmUgTtHHRLEMxyWczlEl9e6+++5kn3jiidm1W265ZdR7aKprriOvnZGcHNdP5yD3Dc/ByEcoLAnEdRpvolTdUSra2jEQUYrhNBbl9uf3aD+XyqksGccdpfTJQO5TNe0yjz1OlfmJT3wiKzcIOTqg/3nH6JqxCpV14HVR1y32Zzz+NW0sX4tSQfPcKsnk6OtozEZp6UuyEcOQjOgiZX60XvM41/TwHMvzHNA+5Drx+qd15didy2ka2cMPP3zUOqmUUtQfpbHUr/Rbv0T357GqawTPJ25vTbHNcYTuE0opsXUesyTgMccck2yWQAPyPuO4RveoPKZ43Vafx+V4DwnUpwgehpxZSaZH68zPy32j7cX9q3tCvsbtr3sSnls8TzRu5c/mZ9S5tddeeyWb5Q10z8REMgal/eFor7ugn3vWppGOzp90Pfr2t7+d7P322y/ZOrcOPfTQZPN81LFSalONefg8glO4qyRFrXzmsOXMoj1h9Azs9/bee+9k6z6NYx6O24Bc5ornme4rS32jbcdzmvtan5Hfx/fWuc+xlp4H8RrOa0wkm9UV3FZRnMZtpbFFyZdpzMD9rp/Faem5HTVmZaJzZKZ0ZqefFY3RKE4tPVckudQltfERP6uuYaWxFbWDzi2OA7nNVR6Zx/Wuu+5arBNL/bCskq7hvN6xTMdYZKVKewgtp3XsgkgqlfusJJek5fh+uobxNT6fAXK/xPGErqXcjjyf9DsHXt94f6/+j30jjymVwea1YfHixdk1XQ9WoWtuJAW8NtSet9T8+5rK8evoc1mWavr06dk17hseE7q/5L7mflLfzn6f7R122CErx/0RSUYOQrozIoodSzK5uj6zz/v+97+f7AULFmTluC/0HjxPou/9SjKc2n8qOzfavYE85ozkkvi1ftYnP/nJZO++++7JVl+lY6craucTt7nG2jzu2Feo32dfp/EQj3Fe+1Qajn0n31/HBO8vOE7Sc02OOfksTuWWGe0bvgefH+g5Otd3ENJY6te4L3h8qj/n74a53SJfo7E6xx18pqbyY7fffnuyuV947QTyGDaSTK+NZyNKMedYJDGzcn3VwhhjjDHGGGOMMcYYY4wxxhhjjDHGmAmOfxhjjDHGGGOMMcYYY4wxxhhjjDHGGGOmJP5hjDHGGGOMMcYYY4wxxhhjjDHGGGOMmZKEwpGs26TaVyXt4lrdedb6A3INN73G99h4442TrXp8XA/W1lKN6qeffjrZrEe47777ZuVYQ62k0aufq1pXJa1MbRfV0usKfgbVhGYNMO5P1hoDgGXLliW7pC8G5M+qupusQ8h9zVpmQK4zxxpoJ5xwQlbuT//0T0e9n46dRx55JNmsLaoagjxGVq5cmV1jrToeL1dccUVW7tprr032n/zJn6ALtttuu2Sr3t9zzz2XbO5bbQOeC6xfrZp+PAZVW491AUvjBsjHAN9f5yDXnfVgjzjiiKzcnnvumWzWdlTtuC7mz6D0PVnPkXXdAeDiiy9O9o477phsni8K+w7WygbyuavznX0BP6uOAyZqVx4H7JfVB/L9eWzeeOONWbltttkm2XvssUd2jedrpFc4aPTzStrLkR43twdrYgLl+aP3YH3cSAeU769+ge/Bvpa1W7XcMccck2zWyV4TpThgEPrxo8HjXed56Zr6GJ4/rO2s2rFHH310sqdNm5Zd47LcN6oBz+0ye/bsZPN6BgDnnXdesg888MBindgXsK1apTyntZ24TpE+8yDg8RPphFZriFK5SIdU5xY/d+Qb+R7cPqpLy2vDeuutV/zc0hhVrWAeH6ohe8899ySb5/Fb3vKWrJyuG4NA531p7eXxCOSx2Zw5c5Kt8QrHF6q3zm3G84Q1woGy1rWOMZ7jvD5Ee6bS/mm099XcIyrXJaX9DFDuwyhG5PdoOW4X7d+SDrbOB56vHCvxHhAozxud47vttluyeT2N1hSlNs6s1Q4fC9zGOo65P7m9L7jggqzcLbfckuzjjz8+2bqWcJtqLMMxxqxZs5K9aNGirByvfaW+BPL+5DnNeuFaR66DrqsHH3xwsrnPgdw/8hjVfh1E//WL+hge16UYFshjnvXXX794D25LXXe4P7icxqN8P247rd/ee+896jUdf4z2TelsR/tsEH3Iz6n+pRRTlDTU9T3qk6O9Hbc/xx677757Vo7bgNdj9aHcxlxfjYF5DES+u1QHIPddg9q3R0TngUy0zm+55ZbJ5nbV8yf2TXfeeWd2jWO6aIzwOIj2Qptsskmyt99++2TzuRGQPxffQ/cdvG9n/63v4/ppX+v7uqC0DwNy38DPqX6I5w/3EZ+3AcDmm2+ebD13VB+4Cu1LnkO8pkXnHzxH9H6l/VTt/gkox6LRvnEYRM9UmjPR9xh6PsJ7DSbaF3N8peftfD/ek0Rr01ZbbVUsF1GKX6J1oys47otif56Pek7Caz7PhZ133jkrxz713nvvza7x2SXfX9uR7x/5K2bu3LnJ1rm+YsWKZPPz8/MC+djT/TCfmfJ3E+oXBnVOGvkOfh2tgzX3Bsrn10D+fHx/nYPTp09P9kMPPZTsaN3i77k05uG5ynEYjymth46D2vYYBNE5GK93UczA5bifP//5z2fl/u7v/i7Z+sz8PSC3sZ7JlL6/0j0H9x/PM54vQB738rqq6zSPgXe/+93ZNY6XOVbRfub4dtNNN8V4UFp79fs4PiPjcaDv53VL/RSPC46por3XE088kWyNm3hdZP+t6y3HXnyWqftB9vtXXXVVdo2/g+S4XMdptMfsl358o/rzkn+J9lTR2TF/V37TTTdl5ZYvX57sGTNmJFu/8y59h6l+ncdb5I9qiWL2WpwxxhhjjDHGGGOMMcYYY4wxxhhjjDHGTEn8wxhjjDHGGGOMMcYYY4wxxhhjjDHGGDMlCfPM9JOKMZIS4jRwmgooSkvP0hCcPkvTPHGKY75/JA3xnve8J9mcik6pTTEZpXyN6Dflz5rg+mhKW+4rTmXGkkBALhvB7a/pFjnVskopcZtzqrQo5d8nPvGJZL/5zW/Oyl133XXJZqknvR+nzuM6bbvttlk5TiOlMi5bb711ss8555xkL1myJCs3iJSh3H86jjn1GF/TVGOcZo5T00XpFTUNFqdQ43RiOgY4HR2na9O6s2zZ2972tmRrCkKuY5TiLUoRX5uSa1ApX/kZNG0v943OO4ZTpfG81VSD7Ec5TSBQ9jFRyle+n6bO4/txX2saQu57Ttl33HHHZeUOPfTQYj1KfaN9Pei0lNH9I5mLUv1VIoLHiqZN5bYrSbUA+dzlPlLZOu4XTmmoY4p95WGHHVasHzORZASAclpXfR31YWn+8Pqg79N0mvyafYH2IX8W+3P17Z/+9KeT/d3vfjfZDzzwQFaOJRAjKbwoRX1JvmQ8iOKoyPfXoM9SO3ZLUjtAvi6y39T5znOa/TWntAfyZ3zssceSff/992flOD0wp9kH8nShn/rUp4p1H5QP7ee+uuZwymNuE+0zTsOq45j7hu8X9Tv7ZR0vHPdG0kBM5GdqfdUwUjCzn9J24L7ia1HcwL5I0yGX5E6AfD/BfcNzA8hjEY5NtRx/Nsv0zJ8/PyvHaZM55tEU77VSSpF03yDS2UcSIKV6qR+5+uqrk82ySiq3weizcJ9xXKLrG7c37881xuS9P99D9zjcz1H/8bmASnnwvIt8xqAkXvjZo3MU3h9GciQcB2o6a36f7l14PvE9NEUz35NtXQu5f/nMQSWcuN94/EVSVrXXIok/llReG6L9M39eae87Wj1XUStlrvXgNOMqZ8Z7cj670TnI84LPWnSd4DU3kj3gtSFK8z8MOObX8cnweFcZTU7ZznGcxu633nprsq+55prsGvvcSGK4tM5oOa4j97uOUx47/Iy6hpekgtZ0jRmEzDz7LvVrTOTDeXzeddddyZ45c2ZWjuUb1A/za7bVh3I/lyTn9DXvJzQG4/7kcaPzqvY7gGETxdA1/76mcvy6VqYmOh9R38lw3/C6qrKQDJ/RRN9BRPJApfeM9r4u4PEZ7WW4jhqflGSs1F8tXbq0eI3heRytuZG0DPftvHnzkq1+ge/P66/KhEb9zm3I8ZT2azTe1oZaSeCof0s+Nrqf9g0/H/swjTHZ13Mso99PsP/lswTtm5JUoK71UbxSiuVq5abXhsi/l8a71ovXI25vlX1cvHhxsvX7PP7+jWOhaD/Cn6vjW/f3q9DvRNgXRGfg/Mwnn3zyqPcG8nmnc1DHWFfUfm/FMYCuTVzXSH6HY131oxyj8HjXmJDnIO/lVM6Y42DeD6o/5Fj8iCOOSPZnP/vZrByff6rUXklOV2XYdV53Qe13Dhwv63jnscv9p23F7a37DJZP4u/Ub7vttqwcj6PPfe5zyd51112zciV/pb679jwlit1K76uNC1/0WVWljDHGGGOMMcYYY4wxxhhjjDHGGGOMmWT4hzHGGGOMMcYYY4wxxhhjjDHGGGOMMWZK4h/GGGOMMcYYY4wxxhhjjDHGGGOMMcaYKUm1UG+kYxZp/7EeFWsIq74s65XpPVhLjjWyVH+OdeVYQ431jgHgwx/+cLIXLlw46nMAsTYnw1pakTYna2kNSn9cYe04rRvrsj/00EPJZs13IG9/7ifV5uU2534Ccp16rtOmm26alfuzP/uzZB944IHJ/vGPf5yV4zpuvvnmo9YVyDUPWUd+xowZxfqpphrrOt9xxx3J3mSTTbJyg9BhZV1C1Sxk7T7uP9bE1GslrWIl0iDkecdjSD97hx12SPY73vGOrBxfY7QNa+dJpNPHRBpzg5qTkaYtPy/XTTUEV6xYkWzWHeQxrfdQ+Pm47yPdTa6H9g2PA77HbrvtlpXbZ599kr3nnnsmW+cPrwk6Nvm5In87CKL+47HGdYnqxXNJn5M1VFXzleckr6vaf+yXWW9StUNZf5Kv6bp62GGHJZs1NiOtSH3+Wh3U2nJjpdRP0WfWxjzah9w36rN5jHN76Rjjcrzm3H777Vm5Z555JtnLly9Ptj4jj4moLZh+9T4HAX/eoPXkI23sUp9FaymPB4X7mWMhHTc8P3k9V11eXic4pgGAr3zlK8nm+Ec1gCM97EFRWr+1/bldNPYooXsNngu8N+D9CZDHXmzr2sz9Wzt2Il8ZjfVaBuVHlyxZUvwMHjds6/O9+tWvTjb3zaxZs7Jyu+++e7JZYxoozzf1WdOmTUv2c889l+xHH300K8fPwuNKNerPPvvsZPNaPRY/yvFL7Z5h7ty5VeXGgsa7HF9wHH/XXXdl5XjO8Hxk3wPkeuQ6t9h3PvXUU8m+6aabsnLf+973ks1xzkYbbZSV47iE+0z7hevLY+rjH/94Vo4/S/1Mqa/73buMFW6jaNzxtaeffjq7tmjRomRzHK6+TWNBhudQae4Dua/j9te9C/uCkrY6AHznO98Z9X4R2jf97NU/8pGPjPk9o7HLLrskW/0Yzwv2L1rf0hqtaw7fT+N/vgf3u7Y3+1A+P1D4fhzLcIwK5M/Mc/W+++7Lyl1wwQXJrt1DKzwH99tvv2K5sbLxxhuP+hlA3s5cNz6nAvL5w7H7LbfckpVbunRpsnUecyzDfa9rM7c51zeKE3j88ZoNAOuuu26yebyor4w+i9umdE6q17qCYwEe36N9/ip0rvJ6yTG5xjHcdnq2ym3Cc0vLcX25Hlonfq1rJFNam/Tfo3i2n3JdUhtzdbFX5efTdl25cmWyua/1bJtjqmjd4r0Gz3cdE/xcc+bMSXZ0dhbtNXi+jwfR55X6T30Brwtc7txzz83K8fcR+r0FtxfPQZ0L/Nl8LSq34YYbFj935syZo5Zjfw/k/f7ss89m13gs8tjT9XI8vm+K/DSP96gutefX6mN43eF66LrF9WD/rXEvn4nw94oaQ/G44jbffvvtR38IxM/IfmY8zmG4ztov/Pm15wl8Pz2Xvvrqq5N9zDHHFO/B8zE6Q+FyUbxTmo8AcOuttyab+1bn2cEHH5xsjn30/txmGlfwPndQ1H73pWOQ/R4/j65TvI/XvSGPH553OmdKfkrPeHhd5Xuob+O9/0UXXZRsPePh81D9TcDs2bOTve222yZb420+jzj22GPRBf18b6L+iuFYkX0XkH/no+c1vAd54IEHkq3tzd/L77rrrmusq9JvPFa7b2F0/FZ/J1BfLWOMMcYYY4wxxhhjjDHGGGOMMcYYYyYP/mGMMcYYY4wxxhhjjDHGGGOMMcYYY4yZkoS5uqKUiKVrmjae00pyuklNc8SpnDSNFb+P01FFqfg4Reuf//mfZ9f22GOPZNfKJdSm4NF0csNOu8xoSituy3vvvTfZmj6qlK5QUwNyOU1hxOmuOC3ZKaeckpXjNFtcP07tBACXXnppsg844IBkqzzLE088kWxO7c1jAADOP//8ZF9xxRUoEaW3i9Jb9QvfU1MSc+oxTpGlUkpcjtNsqdxVJPnA6T153ml6T5bK+dM//dNka/rRSDqthmi+9JuqcRApe4GyhJG+5s9XKStO/cYppzXVPz+fjtWSxAuPHSBPN8gp5nj+ALlv59Ts2223XVaO1wpOu8xjSlG/zK/HS4ZuFZyeTyUB+DXbOpe4ztw+2n88x3V+lvyrjiluY053qOsqrwf8nkimgH2yji9+rfFBKY1jNB91DowHtZJspfcAed9rGkJ+HUmH8Ti48847k61rM6+Xm222WbK1b0qp1HWd5vppGlr2OyznoGvpMKkdW4z2H49d9VF8T24f7Reea+zXdU7z+9ifqmQWr7m8nmv6UU6HefTRR2fXOPUn12+8pJOiuVXbb/w+bksdqyzToSl92a+wr+MYGCjL0EVp0EupfruiX3mwrmBZtygtewSX4/VkwYIFxfdEUhq87m655ZbZNe43jotVLoTjYp4PvPcE8vS+UXrwKB1xafzoPOQ5qnvYfoniKN7PsX/nPRSQ9x+3ve697r///mSrdC3XQ2OqEtyXmpaa5W95rOh44LX5yCOPTPYhhxySlePU7BrLlOLPSEa5S7i91I/y2OJn1brxfpqlTzUe5XVHzw/4GsueqS/meVeSCgJyf86+U+9XkrGM/HIkERLtzwYhC6J7ZoafIZIDYV9RkqQB8vMDnWd8Dz5/Uykfjit5DLDPBPLxxuNI9x08Bvg9uv5yqu8oxXi0HxzUHIxki/iZeOzqGRY/3w033JBsjenY1+l45M/msROtOZHsDftsrsdWW22VleMzgpJUGpC3f7SfGlQ/leC4fvr06dm1SLqF4flTkt0E8vGp19gHcvuob+S5y/GrluP9PpeLxk0Ue3dxPj7esrsRtTK8CvehSmlw3MO+TmNM9sV33313snV/wtdYCkGlcXm/uM022yRb1+mIYfZNaQ0D6uWRS3tXjQ+jMV4bf5aIJGX5zFTLcbzN79l///2zciwBqPsRHh/RuYVKh3QFr3V65smvS/4LyNdILqdxHz+fflZJkk7hOcljLvrOhOWq1d/yPbjuKqcX7V9LEiHjvSbq+Kw9Zy+d62g/czuq5CLLzrH/UrlVntcsd6V117GzCpX/4e+hOXZRefE/+IM/GPV+QB7P89qgY1RjqPGg1Ic6tkrfXeh+iPsjklKq3edxO+sZKpfj8cJSR0Du29g/suQPAFx22WXJ/vrXv55de+c735ls/s6L5VwBYNmyZcnuSkqpH3T+sG9k/6flOAbRuVX6jkDnMc9B9qcax9TK55X8XHSeMug938SJWI0xxhhjjDHGGGOMMcYYY4wxxhhjjOkQ/zDGGGOMMcYYY4wxxhhjjDHGGGOMMcZMSapzqNemzd56662z15zWhtNlcRpRIJcLiCR/ND0hw+l13vSmNyV73rx5WTlOO1Sbtr30OWuilPK137RlY4VTKWl6QU5dyzIJmu6fUyyxPW3atKzcLrvskmxOzwwA++yzT7I59dXFF1+cleNxwal6tb04/SmnyGJZJQDYe++9k83pX08++eSsHD+zjhd+zuuvvz7ZOhbnzJmDruH+47SJQJ42jFOSaapmToVeShkP5HNBU8lxv3M7cgoyIJdv4HR2mkqwlCa6kzRYlXJJ/aZVHSv8mdoOJbkF/XdO0cdp7zRNHfepprpjOG3dhz70oewazw1OExul4i5JxGgd2c9pet+ob2olbgbhR1nWpN90+ZxWbvPNNy+WYz+k853nbpQul9OYc+o8ndPcZ2yrJNx3v/vdZHNaPU0Vyf2pqSw5pSyn8NN7cJrESFqjS7hPa+WTmNp000DeDtwf2p/c9zx/NCUut2WULpjvwb5F5yqPU52f3FeDkBiopd80iiX/rveLUqrusMMOyea5qrEVtyuX4zhL68t9e+ONN2blODUyr+ecgh/Ix8OnP/3p7BqPAfbd4yUBEvVb7bzjMchtMnfu3Kwc+zNdBzk+4jppKtBSqmutK89dLqeyArXPGElo9OOfuoTHaL8+gJ9v5syZyd55552zclFMy5JGXA+Vd+X5xvJ0KgPC8RBLh6h8JNed1zgd2yW5EL0H+wZNJazpvbtGxxaPV5YmivxjBMcRGqdyn/E1jY9LqfL13/ke/Ll33XVX8X2nnnpqsvWZuH61chraTpE07drAdYs+k+umaZNZroHRfuJYQeP/0lxQ6ZxS3KqxB8937g8dp6U9hPqjSFqmtOfUcTUIfxtJYZXOiCJZPn4WbQP1h6X3MT/84Q+z13ymwn2pn8VyHjyOIik27hf1tdE+l9upNsV2l0Rzm/uNfdE111yTleNr7Pv1WXme6PgstUNESX4JyGWVeRx88IMfzMpxrBSdEURy9yX0OQYxByPJWIbbNzofKsmSKRqLcjtyPfQskf0mn/vpHOT2LsX7QHmvGMnKRudltbJKXRL5xJLUdLSWcBvpvSO5Wf5eg+NUlukA8nNklhVR2Mfy+a9Knv+X//Jfkt2vhGtJhk/n4KBlXSK/1sV5Xxd7z9L9tL35/C3aI/AY5bUgOlfTfS6fTXIfbb/99lk5/U6nK3ifoDEmr+fcDpHsMdsq68nXdJ3hNuL3qS/mczXeX2n78PeR3L8aT3G/cZurvDj3tT4/9/cw9/e180zrX5IzVV/L5zUq8cK+kf2p+gVu70gejPf6vObqeCide+v40nnH8DPzc+l5kkrwdUUU/3P79bOWq8+KpM5K+3iNAXl9437SMcH7SJZE/uhHP5qVY9mzc845J9ksewTkZ02LFi3Krn3xi19MNreF+rRBxDYlmVP9PO4LjSNLMkj6XQ7HHXoWrfv2Vegc5Hvwube2Fdejn32KEt2j9p615ZwxxhhjjDHGGGOMMcYYY4wxxhhjjDHGTEn8wxhjjDHGGGOMMcYYY4wxxhhjjDHGGDMl8Q9jjDHGGGOMMcYYY4wxxhhjjDHGGGPMlKQsnDkGWNNPddRY0431sg488MCsHGtzrlixIrtWq/fN92ddZNVd09djJdKFju5dq83eJaxzytq0QK7Nxvreqre4xRZbJPu9731vsvfcc8+s3I477phs1ZVmXTnWE7z//vuzcqyVxnqQhx56aFaO9Qp57Jx//vlZuV122SXZrMepbcH6aKo/t9FGGyX71ltvTbbqDrLWclewVh/rDAN533JbrVy5MivHbc8agTqXuE1VV+7II49M9hvf+MZkT58+PSsXaa8yJY33Ws1YrXukSVv6rKhOXVLSXozqo/9e0jjW9qp9Ph5LO+20U1Zuhx12SDZrQGp/8nNFWtwlzcpIS1yfq6S1+JrXvCYrd8kllySb/dHaoJ9Rola/cMMNNyy+h59N/TDr0LN2pNaP78kakKpLyf3H71F9VfU7q9AxOWvWrGRr25fqOwxN3X41q0vvi+aqaomzjiprC6uuNK9HvK7oHOR4I5pbqgdb+neOUbQtuN+22mqrUe83KCKd+FK/aLna+fnCCy8ke+ONN86ucXzLGq0cP+ln8zWtE2vncvtqXFTSmtfx9ZnPfCbZs2fPzq6x/1jbGHi8qI2nOaYEcp+lutIM943OLe4PtnUN5zia577q+5aI/FGtfxxU7BIR7WeieIDnF8fnunfkcqp/zPrwrPuubc7jgO+n85XbmbXLo/UginX5mo4/jtt5/uoz6jjrAq6Xrjm8T+A9uMYDpXGtcQL3e9Q+JX3sNV1jeO/Cn6ttuPPOOyd7s802S3Y0lqPxy75YfRX37Zw5c4r3Hyu1eyR+Jq0bj7XIx3L76fzkvQGPHR3v/JrHjs5B1ZhfhfrA0joe7fuUkn/qV/98LNTucyLfz33WbwzNz8o+9IYbbsjK8ZzZZpttks3zAMhjSfYZuk9gv8N1UL8SjUt+H49znceD2l/w55T25mO5B+/1tF2ZKPatHRO8D4zOaC688MJk8/kPkJ+JsR/V/URpTivR+c2g0fWNxxbXJfInPAbUr/G+mPd1QB5XRv3OdeT765rLc4v9rvq1Wj/D79PnH2afAXndxsNvr0LnDPd99F0Ff0/Cvk59Fp+3X3/99cnefvvts3LHH398sqM9ZtROtfHVIOB6RnMr6tvS/IzOFmvR9ijdQ/uP96LsD7VcyfdG528az3JZXpvXX3/9rNyg5ue1115bvFZaI6M1jP1XFDcovGf7xCc+kWyNRTle5zmodeI1kmMjvV/pe63TTz89K8d+QfuX78+2luNnPOSQQ9AFtbEvzwU+N9b3RbEttwG3G5DHE3w+Hp1Z8/243YA8nuJ+1s/l+rIPnTFjRlaOx6LWif0676t0Pdc6dgU/k9aN/U90TlL6Du6BBx7IyvH80fHCz8t14vYBcr/HZyEaK/Kenuu+ZMmSrBx/B8lr55133pmVW7BgQbLnz5+fXTvjjDOS/dWvfjXZH/7wh7NyRx99NIZFND+5P3lerLfeelk5jj91/WFfwL629mxE92ul96mvjfYgNZ8LlONUrUNtnOiMMcYYY4wxxhhjjDHGGGOMMcYYY4wxZkriH8YYY4wxxhhjjDHGGGOMMcYYY4wxxpgpSTPe6fOMMcYYY4wxxhhjjDHGGGOMMcYYY4wZD5wxxhhjjDHGGGOMMcYYY4wxxhhjjDHGTEn8wxhjjDHGGGOMMcYYY4wxxhhjjDHGGDMl8Q9jjDHGGGOMMcYYY4wxxhhjjDHGGGPMlMQ/jDHGGGOMMcYYY4wxxhhjjDHGGGOMMVMS/zDGGGOMMcYYY4wxxhhjjDHGGGOMMcZMSfzDGGOMMcYYY4wxxhhjjDHGGGOMMcYYMyX5/wGN47P5bJmtkgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 2880x2880 with 20 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(40,40))\n",
    "\n",
    "num_images = 20\n",
    "for i in range(num_images):\n",
    "    row = X_train[i]\n",
    "    label = y_train[i]\n",
    "    \n",
    "    image = row.reshape(28,28)  # Note that we reshape the row into a 28X28 pixel image\n",
    "    plt.subplot(1, num_images, i+1)\n",
    "    plt.title(label, fontdict={'fontsize': 30})\n",
    "    plt.axis('off')\n",
    "    plt.imshow(image, cmap='gray')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e1a431d0-ad72-4e4e-a4ce-782d9844eedb",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Just to show that there are no samples in class 9\n",
    "sign_train.groupby('label').size()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d03dd1da-5bd6-4f04-8124-d8f4e0fb3def",
   "metadata": {},
   "source": [
    "## Normalize the data\n",
    "\n",
    "As we mentioned in previous classes, networks train better with standardized or normalized data. Normalization tends to work best with images, so can look at the min and max of our dataset to get those and divide by the max value"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "5062a132-60e3-401a-9514-63eaf9519027",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0 255\n"
     ]
    }
   ],
   "source": [
    "print(X_train.min(), X_train.max())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "905ce62a-c8ea-46d9-8023-b27e18e5aaba",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Normalize our data (get values between 0-1)\n",
    "\n",
    "X_train = X_train/255\n",
    "X_test = X_test/255"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a634d683-d002-48ea-8ca0-bd2b3def5a53",
   "metadata": {},
   "source": [
    "## Labels to categorical\n",
    "\n",
    "We want out labels to be categorical. As they are now, the labels are 0, 1,...,24 (to be honest, I am not sure why there are 25 classes. Should only be 24...). That would imply label 1 is more similar to label 2 than to label 23 and that label 3 is less than label 4. Converting to categorical labels converts these to named categories, removing the order/associations among the names.  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "b46986cb-a71f-4f30-85dd-73171afc13ca",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Convert our classes to categorical\n",
    "\n",
    "num_classes = 25 # Not entirely sure what the 25th category is...\n",
    "\n",
    "y_train = keras.utils.to_categorical(y_train, num_classes)\n",
    "y_test = keras.utils.to_categorical(y_test, num_classes)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e7b934eb-ed00-4d42-89cb-9deba47e35a2",
   "metadata": {},
   "source": [
    "## Make our model\n",
    "\n",
    "I've kept the code format here as it was in the Nvidia exercise. This is another common method of making a model. So far, I've used the format below to keep things similar to `sklearn Pipelines`, but I want to expose you to this `model.add` format too. The code in the block below is the same as this:\n",
    "\n",
    "    model = Sequential([\n",
    "        layers.Dense(units = 512, activation='relu', input_shape=(784,))),\n",
    "        layers.Dense(units = 512, activation='relu')),\n",
    "        layers.Dense(units = num_classes, activation='softmax'))\n",
    "        ])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "c6175207-a26f-4c9e-a1b3-08c37b5322a9",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2022-03-01 15:19:23.295292: I tensorflow/core/platform/cpu_feature_guard.cc:142] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  SSE4.1 SSE4.2 AVX AVX2 FMA\n",
      "To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.\n",
      "2022-03-01 15:19:23.780042: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1510] Created device /job:localhost/replica:0/task:0/device:GPU:0 with 79125 MB memory:  -> device: 0, name: NVIDIA A100-SXM4-80GB, pci bus id: 0000:bd:00.0, compute capability: 8.0\n"
     ]
    }
   ],
   "source": [
    "model = Sequential()\n",
    "model.add(Dense(units = 512, activation='relu', input_shape=(784,)))\n",
    "model.add(Dense(units = 512, activation='relu'))\n",
    "model.add(Dense(units = num_classes, activation='softmax'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "d76321fd-ffda-47f9-9391-85b73e0e3246",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense (Dense)                (None, 512)               401920    \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 512)               262656    \n",
      "_________________________________________________________________\n",
      "dense_2 (Dense)              (None, 25)                12825     \n",
      "=================================================================\n",
      "Total params: 677,401\n",
      "Trainable params: 677,401\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e1f64f65-90ed-4182-afe4-26d94a544a37",
   "metadata": {},
   "source": [
    "## Compile and fit our model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "7740dac1-a3eb-4e44-9703-05df91c24715",
   "metadata": {},
   "outputs": [],
   "source": [
    "model.compile(loss='categorical_crossentropy', metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "f5a10ffd-40dd-487e-9dfa-8b9cfa642ecf",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2022-03-01 15:19:25.339673: I tensorflow/compiler/mlir/mlir_graph_optimization_pass.cc:185] None of the MLIR Optimization Passes are enabled (registered 2)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/20\n",
      " 65/858 [=>............................] - ETA: 1s - loss: 3.1917 - accuracy: 0.0803"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2022-03-01 15:19:26.566877: I tensorflow/stream_executor/cuda/cuda_blas.cc:1760] TensorFloat-32 will be used for the matrix multiplication. This will only be logged once.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "858/858 [==============================] - 3s 2ms/step - loss: 1.9244 - accuracy: 0.3883 - val_loss: 1.6048 - val_accuracy: 0.4670\n",
      "Epoch 2/20\n",
      "858/858 [==============================] - 2s 2ms/step - loss: 0.9624 - accuracy: 0.6758 - val_loss: 0.9109 - val_accuracy: 0.6864\n",
      "Epoch 3/20\n",
      "858/858 [==============================] - 2s 2ms/step - loss: 0.5857 - accuracy: 0.8045 - val_loss: 0.8255 - val_accuracy: 0.7358\n",
      "Epoch 4/20\n",
      "858/858 [==============================] - 2s 2ms/step - loss: 0.3967 - accuracy: 0.8731 - val_loss: 1.0551 - val_accuracy: 0.7316\n",
      "Epoch 5/20\n",
      "858/858 [==============================] - 2s 2ms/step - loss: 0.2942 - accuracy: 0.9098 - val_loss: 0.7375 - val_accuracy: 0.8055\n",
      "Epoch 6/20\n",
      "858/858 [==============================] - 3s 3ms/step - loss: 0.2515 - accuracy: 0.9328 - val_loss: 1.2855 - val_accuracy: 0.7479\n",
      "Epoch 7/20\n",
      "858/858 [==============================] - 2s 2ms/step - loss: 0.2183 - accuracy: 0.9453 - val_loss: 0.7729 - val_accuracy: 0.8469\n",
      "Epoch 8/20\n",
      "858/858 [==============================] - 2s 2ms/step - loss: 0.2163 - accuracy: 0.9513 - val_loss: 1.6917 - val_accuracy: 0.7266\n",
      "Epoch 9/20\n",
      "858/858 [==============================] - 2s 2ms/step - loss: 0.1802 - accuracy: 0.9601 - val_loss: 4.0188 - val_accuracy: 0.5721\n",
      "Epoch 10/20\n",
      "858/858 [==============================] - 2s 2ms/step - loss: 0.1756 - accuracy: 0.9666 - val_loss: 1.1778 - val_accuracy: 0.8231\n",
      "Epoch 11/20\n",
      "858/858 [==============================] - 2s 2ms/step - loss: 0.1764 - accuracy: 0.9663 - val_loss: 1.1678 - val_accuracy: 0.8003\n",
      "Epoch 12/20\n",
      "858/858 [==============================] - 2s 2ms/step - loss: 0.1559 - accuracy: 0.9685 - val_loss: 1.2439 - val_accuracy: 0.8133\n",
      "Epoch 13/20\n",
      "858/858 [==============================] - 2s 2ms/step - loss: 0.1620 - accuracy: 0.9702 - val_loss: 1.3891 - val_accuracy: 0.8186\n",
      "Epoch 14/20\n",
      "858/858 [==============================] - 2s 2ms/step - loss: 0.1565 - accuracy: 0.9720 - val_loss: 1.3335 - val_accuracy: 0.8277\n",
      "Epoch 15/20\n",
      "858/858 [==============================] - 2s 2ms/step - loss: 0.1492 - accuracy: 0.9741 - val_loss: 1.2542 - val_accuracy: 0.8431\n",
      "Epoch 16/20\n",
      "858/858 [==============================] - 2s 2ms/step - loss: 0.1379 - accuracy: 0.9742 - val_loss: 2.1318 - val_accuracy: 0.7906\n",
      "Epoch 17/20\n",
      "858/858 [==============================] - 2s 2ms/step - loss: 0.1442 - accuracy: 0.9732 - val_loss: 1.3413 - val_accuracy: 0.8479\n",
      "Epoch 18/20\n",
      "858/858 [==============================] - 2s 2ms/step - loss: 0.1341 - accuracy: 0.9785 - val_loss: 1.7921 - val_accuracy: 0.8086\n",
      "Epoch 19/20\n",
      "858/858 [==============================] - 2s 2ms/step - loss: 0.1367 - accuracy: 0.9783 - val_loss: 1.4696 - val_accuracy: 0.8484\n",
      "Epoch 20/20\n",
      "858/858 [==============================] - 2s 2ms/step - loss: 0.1207 - accuracy: 0.9805 - val_loss: 1.4348 - val_accuracy: 0.8417\n"
     ]
    }
   ],
   "source": [
    "history = model.fit(X_train, y_train, epochs=20, verbose=1, validation_data=(X_test, y_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "bc9cc2eb-7737-4710-8112-01be4d9517e9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Call the plot_history function we made and imported from helpers_plot_history.py\n",
    "\n",
    "plot_history(history)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2013534a-5b93-4246-9e6e-ba024b3a0089",
   "metadata": {},
   "source": [
    "## How can we improve our model?\n",
    "\n",
    "Honestly, that's not too bad with the data we have. About 80% accuracy for greyscale images 28X28 pixels seems remarkable to me! But, notice that the test accuracy is well below the training accuracy and test loss is going up or staying the same. We are overfitting our data.\n",
    "\n",
    "We can do a number of things to improve on these results! Let's take a look.\n",
    "\n",
    "Let's look at some slides that cover convolutional kernels, pooling, dropout and data augmentation for image classification in neural networks: [Lect_06_CNNs slides](https://docs.google.com/presentation/d/1uSk7xHWZ9H6YihUP4OdHpIVws_2py_HBfbby7GpZDCA/edit?usp=sharing)\n",
    "\n",
    "Then, we can move to [part 2](17_CNNs_part2.ipynb) of this section."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9cffb72b-2b1f-46e1-a27d-d3d94874845c",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Tensorflow-2.6.0",
   "language": "python",
   "name": "tensorflow-2.6.0"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}