judge-books.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Judging books by their cover using Flux.jl"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Hi! Today I'll be showing you how you can use Flux to help classify books into different categories just from their covers.\n",
    "\n",
    "The first thing we'll be doing is importing all the packages we need.<br/>\n",
    "`Flux` will take care of our ML infrastructure, `Images` will allow us to process the image, `CSV` will allow us to extrapolate the data from csv files, and finally `Statistics` will be used to help calculate the accuracy using the `mean()` function."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "using Flux, Images, CSV, Statistics"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The following shows how we can read the csv file and checks out the first item to show you the structure of it.<br>\n",
    "The following output represents the information we have of the first book.<br>\n",
    "Column1 = Amazon Index<br>\n",
    "Column2 = File Name<br>\n",
    "Column3 = Image URL <br>\n",
    "Column4 = Book Name<br>\n",
    "Column5 = Author<br>\n",
    "Column6 = Category Number<br>\n",
    "Column7 = Category Name\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "CSV.Row(1): (Column1 = \"1588345297\", Column2 = \"1588345297.jpg\", Column3 = \"http://ecx.images-amazon.com/images/I/51l6XIoa3rL.jpg\", Column4 = \"With Schwarzkopf: Life Lessons of The Bear\", Column5 = \"Gus Lee\", Column6 = 1, Column7 = \"Biographies & Memoirs\")"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = CSV.File(\"judge-books/books.csv\"; header=false)\n",
    "first(data)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Next, we'll be checking out the length of the dataset for our reference."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "51300"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "length(data)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's try to get the file name so that we can load it."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\"1588345297.jpg\""
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data[1][2]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Interpolating the string containing the file name into the path of the image to load it."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAOAAAADgCAIAAACVT/22AAAAAXNSR0IArs4c6QAAAERlWElmTU0AKgAAAAgAAYdpAAQAAAABAAAAGgAAAAAAA6ABAAMAAAABAAEAAKACAAQAAAABAAAA4KADAAQAAAABAAAA4AAAAACiWSQ0AABAAElEQVR4AczdB5hmR3Un/M55cs5ROQICJA0SIJEsksjBXoxh17J3DbY/g8EBP8Y2zwPGJthkFgS28WITjEkSQRJKgJBQQDnNSJM0eaZnpns69/er+++5anqmZXsfhy21auqeOvmcqltV977v23jT9V9qaGhoampSK2NVGR8fHxwcBGypSnoBFTg1cnNVhoeHBwYG1BWDqVVjYzMqXHU0uqgKDvDT1qUNJ7Iax4sOo6OjwUdScyziJ5UgoKp56kQ7MjKCPBJr2skNPCmgDj7MFJAUyOHpcnC4v62tTSNKtra2aoc/nP7+/hkzZgAODQ21t7dzGn20dWESKzR0kUhKINpKTGkcb8EQTnoprw1SIwRNbxoQ9QYBZoBqEjFXtPXGmYBNLUWTlIquyNIgKEzUIKgiYnh4NJAanl6s4KSdhrbS0lT4h1aD2giVjo6OwGNXdCvt8ZIJxxZOPhYIUrxD3uP0VQuET+mUEpyoxTDG1xAiQdTH5Q6Inxp+ijZkpYSoGhXYKiDBZLB2uB0lmsikYKrrXg3ZgA891fDDKgjHrUmBRoQagkZKkPFJLz5BYKwul/jXmBo9PT1qbsWQB+SoRpAxQQI/ygBCCGdotVD84ci3cA4JNHwQqoOprWgrSI42i0o1EJyIQHBT8FE0APWmBB+E5mkftw5J3RW29DnKpvyrFzwlmAEGrhaXo/0TqQ8IZ3xsQu1Q/Yt1C5fF72HHJDTa3FoMPXoJJ47T0AseCJEaTyBGL5yUcK45uATPZc0h3IjQFYTUUS91jaxBKyRKpezE8A3byWh1G4e6rZHL4Kvr3hiLZ3NrC+Yw3U5KOlVtzuntPaSuEpQ3GtvbS6YqIOPjQ/Fqc3Nrzd8E3NqKQ5mHFF5E5ZILdCEMPFZHQ5BoqAGeEszAQWgYOMU0wHOpDkPXNTAQmGEI/q8vSCCHuUYu60ZE15dEMMpl3KVRa6V93FIzrFkFrcyFKdxtzIHmkiO0kSnEKCDq5IFaO73QXAY/TCfXuIUP/MBRaWTmi39dVgO+VA3VCJvQ6WjGRFAtLnxSI4mIXNaEk3Emt4Nfk1AgbKkHmDZgDTczxdcyDzBoGNLfzd3aprOzUzpmEt27d+/+/ftPOeUUCBxSzKkKKm3MM89NdoUs1ZVeJHCiRmTBBMGjVrhWsgbCBHTHg5kYpRHRLS1lzgaphUJWQh7mQQgw8Mnt9EoPjbRrKmhjoxMTUHrV0VmdRpApyUa9NecaPw1o6Qp+jdbC+0CBQsVIKYIn5VMQghOEuq6ZJopTpLpkWKXqhLrww18XhqAgSghBGpsaG8aZUU3kFTjkgeuKGgW/6jXDhZV6rGIp5HrGJ1gWxMmF+FqcRpFIGmbjJfkGBwbV3d3d0VLmQTB3qjOAW0qzuHLGjFlz585/4IEH2ts7zYDbtz82Z86cjRsf6evrW7ZsBSYaM2fOVK9atUrWzp49W05bpBIl1xNvrACbjy4TqVLcSCFzdaVUMRLUuryILaoWq6O2Wrup1Czm1ZKdo+NZ5DWbmxlabvLFrhR+oLlECXCyW9LWqytok3vB6YyJARZWIErhU/QpnndZVGywxix1kd7YVO4U1YxDyYJTJdhkzpPbhbrScDKwRBenGgSJnZEdYMjqNp1AaKnWVnS5rDlMaQShJoksdWIPOfzDcDJcl8v0auOjXZdaSo2TLvAnUEZvjaaNNnw0FJE7dOiQFMlYP3jwoDQSGI4dHRsR+2YL5KaWsfHRocHhtvbW4aGRjRs39h44KLu3bN46b/7crVu29czo/uktt/b1H16/7oQlS5bccfvPDh8+vGnjI0968tlH+gcGBo/g1tHZ3trShs/I8Ki2adj8R+7u3bvNx4sXL86ornWL2mqmuWFo0LauWQRep45LvmJCRlTsjXuR6Kps/bmIg9cl8Qr/AHHQmAyZLD3tSJlch0/EgdfkdaOWOFlK2nBqkoktZE1Wy6jpQbQhRGR8AaIBUgNr/CmNmpzL0lYbA/UUEtGpdXF0LQ6wLiEnVIEAU9HIZYRCDlwdxQKfXIdqMgTV5EuTnIxRNMqcNzDU1tYhAQYlZVtLT89M7UOH+trbW7eN7pC1R46YFMdk8r69+02rw8OD4+2NCxcsmj9/IdqdO3dX82Xbo49sNvuMjAwZCK2t5rCW0dFhkBUrl1oYUIA4mvAMSzXYpWaFS73ULpaXMuEfvUq69OYOBgJf0QDETSl4Fas0wkpbY7Lh2hXC1OwPZjX3F3TM1TVtel2mFIyKj8s0IFSwCXjdntKo8afAy5kIL1SalWpCTsU9NCGo23FS0MyCekF4YQrf+jKESRftmlwjTAjVq2igSoLWkACDE1ahmsxfbwpgjVMjTGnArCFH6SYgVLI3FwC3Ywm6b98+jb7DA+74SEAcncyfP5+IAwcOmGtNsaeddgZ8Oi9ZMvzII48sWLDArHnfffe5ucuYkZH2c88914yIVWTVdsVp/GbXhEoSQ5CpCMlySRxkzBVdhMZj/f2Ha1Yw4430Bl8NQRcF0otDZlNdwSQXEI4Cog5JGuFf+ialWvIkfCCED5VCGEGTWU0ZGOGmRlK3j9vApFYGQuON1/yf8J3MHQanRw81pjSDoMFIRcMloIbCoVU8ColLfOFERVNFci4emSxFOyKiaBBENe7TawAQIS3whBk0vZxVVD96HqlBHPyIrjHhhFBX9NeVmQZ+OBAaVcnKHOZuSwTCdI2PFrnRLVJcYqINqIGnEhxSgkNijYyVSwg4a4NziEY8PDQyoIt03MitC1YKQmjxAymoBoeO4BBvczsEJOB2bPQPz0qjUhUNB8uRAjTcogDvgeAAAl8BjyAQROGJnCC6KXDgEweiQFAgOwZmGAgdQDQgV52lqkgfT/2JywqfDpEelfAnbjJ+fTmxNUvf5Bolg1mCGK/oB4HN3KFRNK0GqzpurS9xj9eQw8cBfgyOJWo82QDI7IxObHEAVGsn6dXaUReVLqzga0/xly6ykFM14kJVG5/L1DDhQMZcrWTGwgRzOEq6mkvApLX7ODMdFTlyj9NlmBhEOwxMKpQyM3G0FCy3cm23fjiCpRdjacl5+FQuZG8Z4agK/dH7Ox1cZrRQ3iUcRsU6foMPmW+1Y2kQwiQKQYCvbXgHX2/8HEHxXhCIUMIK2/Cp/ZPeSNcFrmBCJTW/BaKmT/hoE6oObajIKjiVVjAxjC0aQdaLoaJRl2kTlDF0xdSyXZs9uNQeCT1IvEADqsDHnd6odGmDJz80lOiROuriowEH/9CqQQwPlmOudodFgj+4XoXTwxYEW3X0gZPLGhOEzoDBT5jjEV2AujCMv8BBsIosDRY5hretsY3m0dEizXxFZINN0ujIWNlVly3ruNtqeQSmj5Mbmmyn7F5truHbDKEBYavDieHBwbIpp3LptPGa2FNH8yhcmxOH14pRFSUronbhW830QdNGHj5ItOEHM5Zi6xJcb5ggSWRdwgF378TWJdrgI1G046iICMMwSZcaoRJCjWDCwZYgCCDg2mFVsCs0lwFqKDCVyJ02QfGCSntDmf0SJTcCZBGMNQTiZ82aZUEGByRi1HBAlMOH+xHWmoGQHQcBKpCDr12GeDW3JVGCCa6hjuoUoFVsRhg1IChBCH7axLmk5JEjR7QlHM4aevFR46OBNnzCFrwOmzjLKjknz7CqMrVlrLEMKjy7urpk58hwuUsaofSp+FeqwqVSU0NzQ1lEouVJo72re2JGhAwuHTToQCsFSfSnicuMUm38aYi5EQEe5ZnDNEAjFoTO6vgEsl6XdAfBQXGp0KG3t9clWS6Djwm59EEoClEJh6ABQqYD0UXnamCj1U4NLfBAIKOFHMJcwiwOrHJDg86s04CDCk5K2pBzWWJ/3MJaAnRt27Ztz549dOLf6I04PsLaCd8999zDpLgpCGgR2nDcdddddhLBJxgHltBJzVoQWqYGQYV8eGx0z/59TZ7fNDZ0z5zR2NI8a+4cE5Q/Xf6O2FgPu8ePjJrZWpo1Gpqb2jo72rs6YXb2dDe3tba0t8HXwOdg3+GR8bFNmx/FDeTAoYOYINRu7WiHoBcyJhq6MMcNWzhITHv+zCmGyc6dj2lUQXFTHr3jjts6Oug88vDDD8JxW6+mnsEya46N3H77rSAagKk3bnzI3h+5Pw0xHRoaONTft2vvnr6BI2RRng75Y74G5SHQsxw3NjfRqq2jowz9traunp72zs6ly5cfPHz4jjvvnJigqi5wCDK9SKqKcPAzz8uJ2267bfv27TyvJIl5XpcdYUIgrBrJMLVLtQDBd9Z74403injiiEqw9vUe4LT4kJ6JlAZnxgqaA+ZP3CmDlYQhJVqpU3TRV60E0vzmN76yupyoQNOnZszdd99tx2pnajksC2Wq2j2XirivWbPmhz/84UMPPfS0pz3Nad/mzZuhyUu9+CxduvTKK688cmTAzpRUoxZ5dZq9Mebh/PDDD9MVbcg5Yttj280KGzdu5C/5DRNOpsCdO3c691bmzp0LSBYlbZDpaQo3VLZu3RoNicMcTxAFghN1CpOFs2lvx44dHGTXzMW7du0iF6vHHnvM3YAVrOZHat9yyy0eHsPRa6BSYOHChRQgS/vRRx998pOf/OCDD95xxx24GV3mVFrF79ddd51s4Qenm9TTRSieNvtcwY3Y4jAyNo4hhbUpAB8HCis8AKgRE8BxuOGG60844QQMIdsYEX3zzTdnGUZnOOCcxjRASo5UUSOLVj/72c+owTPY0oSxIkWZU089Vcree++9qIaGhnOYAIKcnmqsSKQGZOJWrlwZQwgyNsxBRPOhXtoKFg4cIjRCAEIrQGh81XeonEIIE4biu3r1al0gKfxcl0CmvcXrJvWqq65605vexKcrVqy49tprZQZ5zk3uv/9+5tFeWbRokcBff/31EtelQxa1xyexEx/JIXimTz6iN0UNyic96Uk/+tGPNDj05JNP/slPfkLchRdeyEg4vPPUpz4VK+me6IrWj3/8Y1448cQT165di4r99MmczUheEDCRAKGJtDjttNM4VzqSOG/ePFGhEon0gSY28TtDDBto9JF5OBMafYgb6if/0FlnnbV58yOo7r33bjiYoFqxYpnU3b59q3n04MEDPT1dAkD0/Plz9+/fax597LFtdld8deaZZ8rCqn0vNfjHmJcoBJn8cZPKjKUGQwSPk/mEY3N0RQHZwwT6cP5NN91kUGHCQClizDCZvVu2bOE9CM6tfvCDH9DQHDF35izD7OKLL2aghEACWZFneJILjSeNSYZTwOggCCvhMGxiURIOxAjh5HXr1lFMlyDKJzrLXTi04iIIjF2+fDkg9TiZOCrJBGw71qxlIJxkP6u5gm5wiFNjqIHPv5ygFJUxWKAhjLWSjP2I+c4lX0sLg4nL5KgcIvj2228/++yz+ZePcMhalMcxoS63GsRPecpT1JKYrhQ1jERLtvEXhmLMBcJ2+umnI0SybNkycBM20WwAhMxfAsYMVOCsyjSmQRa/nHTSSdjqhcllsRyJggMTuEbwNIi79dZb5b1gm5POP/98+PKYYovnz2EIuxgIc9OmTZkgJY0zThIpJuqkoMXEuMrMh5baivBfdNFFX//614njQJGjs9TEn5e6Z8396U9/SkNMpD5jCeIoOY0PiOmNdRSI9xYtWkAKtsTRRFyITsGZM5nPgVSFL3W4EabjW+arFYOfYxFiq1BYLKAxpxrAZU/HRfDV5DJE7kpNytNTMuEME3OxoCQmIKjIJZ1ocIKMGUzAoxKTCcWE69xywTmc/5O7fAJZSfhSlxy94eq/T8eUGgZPqTHSBZMX6CHhRJeW/E4zbQlqJjcQeZMwAaOKzGMA7cfKg/GJrTpWMQa3jHUQGqsVXWhndnUn/4jTMB7MQBhyIj6Zg0uKtbZGKGtlCem0lY7adKM2btyUBmRtscdBb+ZIs4X7gy7z6zOf+UwcJKUhbpp56UtfmsyLXJwpxhZGGRh0EB5dJgxKigrmOJvgze5z587WhRxE0aAJ0VwncpgbloDf/va3sUV+9tlPJlosjSi63XDDDREnm40TFmX651uucDnaVNYGPCNHraCwev7zn6+XtzMO8aGnQOCDtnms7CU4GXOE/Mz5nEkftlNMl7YSX/UNlo0Li9Qwa7aizBw4ungDNzWJgNCQpwFBIQUaBbRx0KXWBmm1y6u2YtqAikvIaqUWoUEBpflNv/wKF8cWfVRX68KFkbSp9C5jxaU4USt8YyeIrIXPO9oKuBcEaIYwhUORgxiahjIRwokPOAQihgYGud5K0exLihWhwFj/6QUxQ6jdXzhIimOicDRyXfhQlQJEKMi1oyE4CHNgqkOCp6HsViDVBABPcCs8qY8tQtmPCW1FVKEeoahAhAeyfMWWXJhmGlPCsA1cdSqnJku6aCSQ5g+TmWGAg1TWZSazRseNehJd2rHOSDCjGz/SkfnuGxoMMa50bd+5jceIA+Go8EdOFnIqaWOCJIOqtamNRNzYiBycmSRSm2JsZKAaBENMOrq61GIXnhq0jbvAkxWAeCbD0CZ8GCJJwqhdwtFFokttciG467nEgVwNcjUUDawUDRIVDWXaBNWHPp5FRhIIG+Q0w0jiUDXWeKmJZIZkEoPvfe97BrGbFITD/X2ykLoVbXX+Uk1jRdf2do6zlJFw7gh8TURrcwuTBJ678RQM4siSH5iYbFzKIbWUgmNOikNpm7sePooUj/60QqhQQGELzyJU40ZPppnV3MI40WxNE9kJbvqBgBWS4sKWFrOazKAMWqxA3KaNRgpbo0OTvr29ByQizeFA4JwkjTb1kvEwmW9gILS3Nh7YwkBA4tAqnAYuj+WlqReJtGPpkuWLKc80xWWtjIbZAQ4rrBmYQHniPAnDDbleE79lAHLKc0gJfzXFuExM6QOKLdEUiA5ooZmkkROKkKoZn+Csw5CZLMcEK2xdsigjBDfKMI1uCGUlhtCE2GoQW151iQTz0GorYTVtgsJmWMwQPGLIw669vdxJ3ZLwMmkR6ZJmONp20JXGVt8yT+POO++cOXtWspyKOEBjvDspQqElxbEF5lZavCknDvUetOo1N4g9HEykmjQljsa8jFxSumQecdAoSRw7aUUlNTSbaNEljlxTNafQkyB+NIqMCoGHAIghoEFio/CMZzxDL5W++93vunWadcJQqHCAyQNEoxUtbTrzu0vTg0sKO6y0WRYSCQoTGhEMhwOZqpQB51sS8TSDUpj+VrfMZGz8BmgNLdsQmkclrgZ99uzfLRzs4jpwrJDwG3FqgcCT7evXryeXlK72boT0dxkrUMExDimAnM6GpXShszWP1+ToL7mpyr2OaNiFBH/MhZh63MuBNo7IacLGLMNcMsTYZilXcx0m/M+9pGjgk3suHMrYqMlj41ZkYaophomSNhunTVAKkc2qr3zlKxwnAN/85jcZfPhwHzEyjDv4lBLaHGcEc5w21fGFbznFmNnV1gcre2SJxeN4omU2WrmI1p5JAJwS6N38yKM8hQkDXFrYyQCuZAxxVOcOwZPcJipJyXjuYBiIxMITQwimtHPOOQcOIHFUkn/UEy0FDra8Y3uHHJVEl68XXHCBQWXzx0bz9NVXX82VmAgJD1rRSzXF2ICcBb6YCacIYa69evUq9/ovf/nLascxGuJqiDKEjbHXbKpBCp5bthTmfEIrExI+kBnr7m+UOsUDxwG+zEb4w5t+1Nzccu+9Lp1j7DcB33DDjcuXr7jtttsPHTr8wAMPCm53t9vO8NVXX3Peeec7nP3Wt76FJwOtFr7zne+wCEOxwI3t4mLgWe+KOF/t7y13ZE4WFF61zM1+1xbKRM5w63UZzDoOpL9EZzWI17T52WIMjmTQoLBaHIkmUVDY0t3ZaRuqi8kKtslFCa1Rl5Ka7GxpmTZBKZeCl5uR8PMjjw8MlDcSGJPbn2QSZukobAacrWjmXdxB7BBv/9kdAqNNOVL5iJ2ZaHkNHK1gaJgJJJzzbrMpY9x2dRFqmWjw8Ygsp4lhYKlnOVVNV6PcxAVYoeVWykQBTLgPHzGAT1s+MlgpLNHFnqP1uhvizKGcdemll0oaaef0hL0Uxo22SCCIhEFFigZW5loZSUMjhFZ8hZsRdcYZp3OrkFMAmrw0OYkxAzds2CD/IFNDwrnbwmxpaXXJTBJdUN4kxMNcR3+jxfCwWGKaIWHUdXR30IftdCOUaWKBv1QwHdCHr1hHNFbm0T279hJNBBK2iAuPEcReVFwHbi4UGjrgvHb9OmzdRvCHxhDSeZ5oBlIMnCxWqLmFOM7HVu3OiQ+gZKW8uCPkLlKK5h0dRHMlN/KJXhBGURWaknyTo2mQju0T7eKxIB4XFhJMA5IsSHCkBC7osYMGbrJh4cte9jKpQAPIqGA+trtMrsjJC7IoZqBQV3QVNpgajT+QA3v2ihAgPvKY3iJHFg6AZmXcsH32s58tgThUGxpx4gEZ89yIRY5E+gPGVAjQ6CzYuOHP3RqEYstlko8gns2UzHBAVgCCoCUIPgUAwxlPmtPBTYb+MmPHju1wRBGypwkyHhNjQESJk0NY4UwNKSJH9+/vRZtgY8UzyGHiT203a+bzsFSjhvwbaSzH4Eiw1YgVkgwhKgiATKYku+jQu+dgUoRd0EwQ9KQD/iAwaUIuQyJxeLz4kJ6UhKAXXFuIoYGTogGowScgISSLDnyInH/o7JIgrhAROvAS5MMH9hNKN8x5AIdYDchwNqIC0XBJ+rQJCgMezYKd9CKAA9kJLh2jfaIYBxmjxhlk4slmg2MLXQpaSjAmCsUvlIamV1cSvaWhPCCguuRjHkfDRxvnms4hcxYERrKcIMaT6J5lJhYevoZD/0pbzcePEaAlLSiPrfzgTbc8WkkgmJzC0QhJN8z0IqGDoMKR9Fk4crrRQgFqQ2O1S2iUcRqPA1uwQs4WrChPK2HDTQM3IiQZX82cORsTl1l1xCL3B/qb6XGGrxdntBzS1D6x6WGpeREHcg1paqgB6ZYcIo70juYyjfEzNUAStfg84WO1BhwGEuR1I2ggSEgskKogYY5LmHXswkEX9bTJjarRluZJngA5gVEdLSXjQUILEkHYaiukJUOSoNPe4smAERq1S4Vn+/vLaYsUIYC/2IBjLAky7jTQ61Lq2BXGxULoFmCRIDB0pZYsV8dm+plFyvJoYABCTBVpbQU3IecgaMTxNW7gVBIP4miSQUln6sGHJthmWapGBCD8TAbkCqdoURKmNlUpIP/c3SyqRN1sJOoEmaXiU1ThL4SGJc5uwTBpQnM6yH4+owkEfpBAPENnCuCgEekuNeDgbLFIEM6YcIvUpDYIV+vVwCGGs8ulicudwCdGyifem6R7+84d1tx7FsxfaPltAdbR0ek9f5npXPW+++7fv3efxVJCTj08hYMfLLKpx5MUS4DgsM5pn16CaAsOwvMCodfswEv0pDnbM0KSdskBOEzTxRWsYALNFb4iDhPjzbElztwOWUGoTWISxmUgiQvMaRMUQfSLWwm+5pprBGzuXP4vUxcWX/rSl9x6olDcJ8Yyhg1UJMnaubW9PO3gBSrqzWSDJyo3REsxWkK2VuM7wfY2GrkQSOEdOYSWODWVxFXuco1tDXwQTneJrRjr4k1JCU1cecTyXMiFwfyELQRy1bip8Y9DZaGVu9W9jTPva4uQrKV5/IC52Fj+ioe7mHmLXPytKdetW4fbFVdcEbbr169jJrY4SHRucZcXKj6hG6BL+DS0AMDKfYxXMWQpDbkCmrmT2oQSRx/4lBHpMkHOnkkNJHzi7k8NceFqizlPmDhfsAllr1uKVfKTzy5vC7g/SC8hw9Cgwhwa/vjQDX+XfK722auSIw0NMOMcT32pRDeYgmWWERq7ew4hQpYnfTk8IUDIHMgGLQVYSpxTEboJREdbG1oMMU8WqlMid0o97bN4eGIpY2hGA227WnvAOXPmRUVMmURFeSZOXM9OicJNGgYN4yk0Y85srmEG5/IppWlPV3tq05VUyG4GiRcvuPWqn3xXKr/qVa/KeOB3DhIJOijSRVZxU7JBmlKDhnJCL6e4FAwISVkMqUC0INFT7loOBo3CXMwul8Isgfjd1sSDTTnBIkCaUM/hC56SiWexkrUCqUErHpBwfCU8FCPdpphWWBGNJ6C3h8477zwGstqGWgqa0mzFeIMn1607AYIltQ272ws+gopQw8qYFYC2iTSRWIQuW7WcP6UgHZgsrRVLf2OMYggNdYpJaBBMsOUrSvKSCLLFQkXUuEiRPXTmgcwIqK698QZOsAX88Ic/jJsdEj54soiv6Ka2NOIfSea0wWNtfLCVcARRlRocQkmj3Vi66KKLOIQfGOvSLT5BQZ5RgVzvlLysL6ftgBEvGBZxvbARTz/THm0IcCgjioxkPAg04WewsW56U5gqIxGKCmcZSXwhMFTkAuraP9rfgZiHKIoJR+AmdRAai7lEy3dUoow2IF/jzyNobbolOr87V6KPE5PMUvAlJX04QqJDwCQ4NOR66mXqsuvU4HeqalCJFG4VIVmVmxTdDA+iIeg1AuWH7BEzI01DIGEalvzGD7xhyAk8JVnKaRpqTsBf79Of/nQ+NI/iyV5udOZCYVkooW2xGWhg8wOrpRS5OCCRtaRTBq0xQxx9qBG/EY0QkPK04mp3OYohBOcNVmPIJ1xHPXINHmorvAcNhAN5g3SDhAf4irgIlQMRxL28Cg7BRMYu+DFEKBMm4qidewWGMJkj1pCRKzDhqHN5bD3tJkk8yGAASzTU4oHX4sXlZkFFLsYXnIr0k3lsA5SL4CDaLN93sLydBMeMhU8yTObhZuMvmSjNy/Y3JOrt6Zg4OpAfJj/8w5NJvKytoNXFGBAIoh5HS0E+4g46iArpVNLWEDbeh8/7SNTakHW5NLOKlnyiA7lk0Zky0KSdNky2u/OaSilv0pLTcsJcJZ8MWlZIBetX5zZcIb2cmgFq8xX+ONM5nom26iobusxG+EDG32xqqtYwc0NgkbnfVMQKDlSODJf5gs4YUkyRiERQyTiR+jQkFE/68+rhA72YU55dRgIz4UsaDjHViYi0Y5GRyZ+Fnfc5q5UPVWX53/zN34gRzzAThA85ljm8p5dcaoRQg8IgosMuyLyHyiUR8hIh6Yf272NFEhEOWahSBzilnjZBcaetxNIQAFwoJJy2oeA4atMpdXrZCZ8Sal4wIgthU3m0yB5Ky1oaCzZ3hycOWLEBjkI5CQqoN4piInKkMEZUANW4YQWIFRe7v2OIj+FrkkALJ8oHPwqHA90Egx/pw1NYwRE/bQgu8RSGeF+DMiBEIIGAuRu08NMKhA4wMdQQfpnhZoWVtGaOWlu6U4+NRCex1DxGATnU3Fy287nrkeVWzgQk8ttME5dGCteR2NhaThXA8edqfOC7+7uNkC7zJCh8CYEJzX0SgBVuHWZxqsoV6VUNpLKVYRGhcRcn4NbcXjYAPOZSjRYTmgPSNjmgzS5+Yzi02i4hoFVo9dKHaXCMTOMnOTO7pxxQakPDR/jooAAiObZMu0lCM1l1wjgFpGJesl4bOwKYp8F3cDiCSRynl05uRr7oQltvyBMn5FQHTFJiAoedGiYufPCkcRJXLRsiJUzQ8pSk4SZSEBIEh84QFJyL0VUhCyvkrkhxSUnk2nwkivGRqMsnhLqkKcVgaqMyn3GoLiIEvsbXJWnw0cCfMlRyy4FpuiKF5lhpQ8BQcgeZXA08q9COSRSTHKeJpQa5MYFuWTuSyCeokHiSy9saCTM+NDcyTbSST9KTyD9yQmLJe5/qxJAObAE04AnFX75Co5VspjygNol9R47QFknUwxAhzRWYLtUwE3EKU0Di6oUWzwSTtuCM4jRCeQNDKo2NlLlGGxq5GGI1XXbqmjZBCaAlDMW97J7q5EXbvKPGkaLEqF0SyS+Gsrs/f4G4ffCXkDOLKrTUZgNHIGEV5VwKoQZ84tR6Z82Y4Vb1jW98w+3JQOdHNujlC+IgoyVdZthnuHTjgACNtgrXi5ZGvFzBShtzVHIIAgXkBC9zHydihT8l4YiZJIv7hFzeq23XLAoFG4Jp0pDAikV2PDwj6tq65JBeW1X4ZjIDlYYyww2aXKoq0DCnlUZiY0SAky6tFdZx42c/+1n+dH/PhEfc9ddfT3P7xZNPPYWXFJfgBoyjEgbyvEu0soEVFHDJvQ51NBiokEgf+otmJhFS6MkVNOQKl74tBxX+UTI+xwGOmt9g8hi5urQ50yU425G4JAgHKsGnibhr0EpiwPSoU4AsqS3wLGawrWXxybHF/IdViR95UPEq1wViVihvu/D+tdder7F581a9UIhniTFRoZXc0qs2+o14o4T2MJ/ylCeJkK8rogS9paNlskcjEGKk1bp7JXWFVoSQYMLFLMEKZnYM4sdyvazlTUI1jHVU1kbbdu5gtjQ644wzZDZZzg1AaCKlbMJsY63MkL/oRS/yuB8ccwpc98MbvY790AP3Q+MpRgkeR69ZsyYSBXLz9m2tjU2YyDmX9ge6LBPV1DAqQGzX7CTQIrSFn7do4fB4w4LFSxxaGTyU3LqjvK5ABHw1QvfxV7ziFW61FKb5RRddhND0YzxI9zImZ/ScdtaZHT3dl19+OYdI0x17ds+eP8+fxJJeIFJZBkM2wWfYcIuFAV/ZL5NrHey+bxRJDo4VR4sTw8/qtn9ocMGSxQ8/+ohYOHVhxd3336eLpU5XuIs3BNEEYXSZaxS0xgChuuScJ/W29mKaHJX6cT7P01A0bYItNrzXRhPugunIjxrzZ8/iAdMNu2ioN6GXfsdmZ0lLkatLMFxq4KJgLRdFnQBZSD+jHPy5z32uAcc7nmhr4H7xxRc7hCKSs4T5kksu4amMZnnJNonITi7AjT2CYW9EEC9gKClBNJCz0BIeFYZYOchgA19LEQ1HB1wpSGIJBxNTiAGAlv08TkMQogGxspPlO5ozxHYqAwAfjradwor+VHWpi1wmEwROYdzw1CWtZYNcNKJEjmgm6DXNm6v4gf7SjrtxgM/GGEK69MXEpkqvaQ8rQcVcgL2qIi85x3gTMwrLVzVtcXZ2SJa8p56EiHXGPAOFw7pTzYGUIZE4cINNCCiJymAwnrGVcBwlrNAwZ5p40UShfPA5h1ZMMO8YyQYANNmJOe+RInYyjAdALI5jDoleMEAFwewugl4uwcfunnrGCU9m4sOQYw0SvgKkiXQSXFFj7HTZiW3jtd/7AiSFAWjUuSx91ZRmuJgeWMuVOC5fvpTN4sFr7GePy7iMN4UNmvAQqcgDKwUM2SachjgHgauhEcFHDIZfYZZFBTNkFUwuloUix/W8w0IRYp5pVc3p5Epuj1Kph4MiuiIBX+Yxmwg1f8EkS0M8pKORTRDRAowVM4mjJHwNaSRjpJE5gJnbHt1s8mCaXpz1JqGplPCYCLUlIik07J5V3uw0EoRQxnMRr5KuwVKGE0E6v7GUP9V6dWnrwo1QQOIkTeXFJuZgJfmoFE04gXMkCiB8HuAKCrDXpYyhm9uLV2FIx5yNvCo/wLWlCCBDFNO8ZOK96IAcULxIdCSHvwQ1OAHpg5x19MzYSBxZrVcEceBGasS9lCTIJR1En4GkNI+XJ6VIYJKVNg+wVH1safzBd/8OKo70VuC5VJjqUqaTgQuH0onXzKq4M1XNjzSmHCo4MQAcN5LQVtwKW2iMAcQZVSAuwdUgyCux49JIm8flOqGGh2J48JTJiSxTo0SRQ9yE53hzeXuDthyHGzdJFylFDQUcPkL4IiQ1+Qs+nnQjVzEGoMk8+lNMYNynRMWagSHzZpUHDRgiAacJErSkUIM+aKkaW5jsm3LoQFvKyFoNVtCEtuRSWJuq8KEZdRqGjS7caIhhAhl7ucIRL1rOIZRzxJgaarRCQ2ErFllLHBzKY1uZVfKe7aRDEy+x02WSw9NocYlJcpoJGiC40dA4N8ErgLRNFHThHwT4mGvHw7xEIraA2kowMQTHkH8A0+ugniHacJAHqIasPrY0v/mNrwqUvLqApI0XbTgIPRk4NlafKQkCSKLlkvF8JMYEKxrRAydoCj4g8CG7jDsCyaXYaPBm/Mg7/EgNkZAc8N10hDBsNdxuxJLPQCCb84SBLzJXYUIHXhZ+EeVl6gkYW3TJpJgASISMUROhgQn9TUiSBubIUPnSB+LIkh96Ncy7eiUr7yOU4kkshOKAp2yjUniKHD50IDEzEyoW4QaHUDUcQArT0CUOUs3iUg5pKBAkIimKmQIaSDTEigIsstJlL/O1+UGd+ww1IKupwYdSX5pShkTSMRQaEmU5M2d0dc/o6ZE+vmrS5/s9FHZG2t7a5pMsgFVylQ/xlPzynSnmtfHybTa+HsJi0RemoAIpf1Kw6sUBAjjaAq++AzWpWaVTcXtiSqVjS/Nl//11ummvaKTkMjUIFvQJAgEuGZxUA4fA8phKAIgSHCRsV8OfYF1lPIRwAKzbbkz8zvsmM4HhYoQcx9FSRIZJQQF2X+NQseFfOngAjxVkq3uzmlC5m4iZNEIiO+U3ZfABxBYrjcQPH+ERM2jkoqUMQmqgEmm54u7lBmexTxlDggj6WF1xLrkSgtqYEGHikZe795YvuTDSiMBBQlOSUIszImxirCaRuEXgBiF5b9hYN2NivkSomMIZmPsJiYp7upsm5rhJVoT0l1KZCLFCm7UmJTmWDuKCCQTiKClfrT24ERrlk+V0JosCRh0TllWvBYpj4osPrfDRQIKnLq7QSzp3MQoQjjYc6gUBjoYuVOABuqyGsKvCJL0aSgEdrzT/6ptfi3VF+fjjpggDRKJGTwZVCCMOhELphRByjoATSXorzJI31RZ84qCHj8BrHIQp8EXCy+fmJHdhgmwFJA2nC4AdrnWSbLNUgu95oDT1xFzAEN74ox/yFEGcy+OY00SycreJ56qrrhJUaSTDMrtb5JmZKG+DYlYznYCQDqhNHD5GiC7ioDkG1OVVc7tAUrwfwwSGyBJhg2+XRk+0Am/ede77/e9/n2IUkChY2c6bnJhAbQtTCzVaGWZqJKQzmTh7KZlEcyZIR3DrCiRyywhxY2EvbgTBMd4sPKxNYWaYObjRkPrZlpkjoRl7MgkabpjTn2PlEG5Uwg0TTtOrQZD0XbN6NTQ+FFkBZQifo1VTTB1IcglCYprLZCeJCrRwoJXL0HKdXNPWFQgOSaQAw2dyPbEOq3j+XEUYbRRMcalYP57B1MIRMJI0EMcevkthVZRQ64KAG8srruWRDM0Ao5m5ymQjpf7pn/5JDDhLMSVQw1TBWvimH6lmi2o2AjEVfe1rX7MJdW4iThyNQxKLu1FJcQmkS2rKNvtiARAtnMWAXDX3McHcKVpiLHK0khn0RGWmkZpSxycFTHjShVwZoAsaWcYJQcyXjvJszZo1ELB1yMdFq1ev9vFimxVzNjXw1CtBKcBe/lEwMYdJ8TgNc/5hrEIlKY6EXc625DQP6CWIixDigzMN6c9R4GrHczClPkJSeB4To9RhCIa05S7ijDoTASYGKq+aRFlEEEIuIoU+OPCPmxVfiRRHwdHFXhC9gBCSKvEnnKSBhgLoUh1CdQhRhbnLYE5XN/74ui/XxBqT8UJMgEKnNKx6axy61oWucFIQgoNUCpUtm9hEOUyQQ0hmhBVkvSn0hhObwxyEs8QbQwVJ9NGLTx6lopUBMCmgAVmA4wK0qKIDWgguJZAo2gZRTBcIbgjlFgT8MYQM2Ozb1ivzKYwtZPAoH1ok4ZCbnUeRJlckkkY2GDa589CnZk5tEuEgpB4pxGnHLWoFBH+0rIAAn1wKRDoIKdEQH/jSSG94xka0qBxzwmSaN57oQw29BoYxhrPRogsfkHDz7B6cVmyMK9RK7VWiISjwKUmixrElXTEkyofKjv9Y5CeANN7646+jr+MEtbinymuqKNr0C3eNyQkKOTjq+C7OSl2x4ejqFYRqWQZS2zMZp2aiwa28Q5/ileo8IgrE+xAQgqu5gOaDo+Xb6mgSj2jD4X04tXraQZYE2oJkprRmkDFwkl5JFJc4g4SnS19PHz5JSnxyGcxojg+F5YEwGzBiqW3ZYALDB2c1nmhjnXacA5O4IuWow6M2zCCo0bpEApM4JifzNKISz0DQVkOAr8ZT0XArp4nUtDQHybiNh2FiFXI4yV2faCColoUnqqgBQSGXktET/2hehP18gRZAlIGGJyqm/Dziv3BVjFewiwcxwkWNr6ILA43ANcLP5RTGtUng2owvRpTl2kQwAFGlFNZVAawZ6gotoFJf0oFuWCVIcGRDerW5WDbo0kDFg/C52+2MhHhTPCI3DayoFw4QNFxqxNjMJURgXkJ4dALGLepFDaykVxQzHiJIRnp7S+rbc+BDB3kDzWUEUS/khGJIYTV91CAa0EhRAtFAogsViWpw9gZN7VKNj1qZTOiSRe7a2FImUlgUHKwgxExK0hwrwMaWsvWJK1KTrgDWpYZjpUt93AINiS7qKTBDeFzk6YATHyeIMeEVPbRxxLcGxqRklC6lZqrNSITRyWWl0kSSFdSjyZdGTagRG2p4MqzWgWJJJgiAdHCp1uZxUhx/aAASzd2KhhIFolVNiAl4eKIiK4IStjA0o6DCByb1AFHJfkKxhe9SEVHDIDmKPGjyYPnM8lXcZmjTKrhkNYchDBWek5UHBFFTCWftmn/UU8OHENFBox5M7ZqnBlVTk06x2BhZulxihUpNhEsMqQchA1sDGqDsD2YUwzOXNataVhBwQ3hsCRqG0FLgTId8LHkNaX79a16Ei2tcMFVwUdKmXPSDwCpawtIb+gpxoh20Cr1U4YMz3PCfLKJuh486rNRoOQ55ou5SPNRiz/UwXapdWj9JpuHR8vheg3rq4EsvaOYtfFCZe4RB2+lMpVJ5DOEgCT4pDMdNwsGJqtQAT1Cdg6ZEK0yiA7b2FuGGMJrjs3d/ea5h3sr6ARP7GHI16Ik2aZ0jTEAcrFbB+ZZWShYMBGFiAOAGB+fkEBJdgNQOCdEB4m9TyFhM0IJjixYmBBw4RP4h16YGE/Qq4JB5A353e2ehN+ScaLrGQ58TT0NInlS/f0R22eOIVHMLOLRj/yRTRZtfayrqOkDHqpyR/ltK8y//0ssYUJeiUjURMqMGagNyBzuJqvmDT267TAkTyAqV0IIc7ZwYAJiD6EpvzYdzAhQDTlBAbC2dj/Ajn0oX23Nv9wiG9vYdj9nhmqjcTzdu3Oh4BYkgSQu7b9t5k5lUxlPD3shjYpG4pnrhaOvWrVLHuZL9O1Y42NXawzrtcoAg7URxz65dHhgyxJbZpp6e4W9H/8///M9eNzFBfu5zn/OqALiXBwaHyyNBUaeMzTKhipyjgGfrzqEcEVDYSSp8NmLrKbY1IhKHRM6SaMVSgnxlBs05yhEB8+UfKmeZqBQQB0kOMfRu2rQJPj0xoTP9vQ7GltXVlwkIHM6QuVGCUsZLMzwDGVXtTILIbXOmXhVdSmKhRohPSNJV99aXkxuC67LOh3SViJe0/TeUkm1RguyIpx77tUUlSanh8lhhaANEEj7QapPSVV9qTNbLZZADhBy5EkW0HKNIF0mmdkr30EMPmUskn6SUN0ILAaGY+TSCI27v7EA2pVEYB6xk3pYtW+CrOVdSykJRQYWPoMoMTHCAjDk0cqWvYPsEnAPLnPCj1WUzDtlQwZyqFDAFghgDpl4TkvqLX/wii17zmtc4kMJKAkkFZ0yeaJPCh7LEh3gyF8pRTAwk/ClDFuVpawRKLHygSXGXijyjtvxbUx1j6ZKCZfDs2SOlcHaKxATjwWDDFlzasZdWVTKXaIJwET8QxCIQJiCX8eBGTlwRicyEkBglDTjWZWKq1hs4xY5bgpzEgJxEAjwu8hMAy22CNvRWE4kFpoys1ZoMZ214RVGYKblMV5QAh1xPh7VmGin4xwUuEeIQErnoTE5ScqIICSr3CSQNBcC7c+YYhImuiF511VWXXXaZz/c4cTT3CAmg/NAQM0eDanDTmKlLav7t3/6tXqmDIc7yI5rQQWwkigcBMgah4NGNaAjhKU3pgApPhK985Sud/+ODm/BLGlY4VjSoCKKkhqNy0hklXZzVGxUmb9ngFgyZCDhsWVt9VwpWqIijibGECencAu5EzAEwNLrJYC6iKjR2cZFUU2jukjjDwNQusnwiBGTRWQMfmsOkPM78QJxaL2BspAxMdWJNmaSHecHAYKmijRyOkqAfW+OcsKox+b9O0Mbrr/68MGCRRFFzXOSlEUXBU3KpS0PRYICil7pU51DkLqkYDhq5DARQAYmUKTUOgUAIDil4CphLjZR04dPg8a8lUfW4ItkG3yUmtKp1gMmnmCcqSKIDw2HSnN+1cxkOEDT8kFdkcbEAI4SJlQGgJkhvvB9aXzwFaHpDK4GQ8K62fKJM9YsfxdUg4o3DwUPlzV8N4nCgMyoNEDVMcAjw0zaPSnRTqV65yBzn/GZQQxqaQkgcCEHbGFDTUB2GQaCkBuZBTk2KxRJg4DVhRPNS9ARXkARTbyA1sOapS5vcIFMvkPCva43oplEjaLhsvOmG/+MHBMYb/A6pEzCirGUtiR3/+RKU0q7h6fUjpYE3N7WWH5NuaB7xPHzYz6+MtrV2tLVbXLdqw4Gvd3hkUHtsvHzeGhwE3AeztNX4g0SiOvDgoAo+KXi22Lk2OqSbwAktnhzFrQwRA9Flp3btiHiKqQo0UQlCjVP7FASOS7Qw1SDILXr9xIelvd/V9O6EdnnxoallcGigrbVdu3yS3Nle9VM12qYow4AgtLhl6PqtD+kYBSq4LTMp9Gw5MtDLOj4ptoyXj6zx5/DQaFd3R+2f+FNE4PADb2u3NLe1d7SqteGjiudH/XpT5VVU8FvbvOA2EVOeDE84zTaiVdwTqdq3+MPRq8aBrMROHMlKFEBwSHxFvI5v9K+jWUc28GRO5AYnmVDbHvxomBrnFpqVTC3nKdJTkmqyx5v/ZWOfc38TdEm3URNYs3Y1KqACiqlLA73EcqzKXcwMV5AqbYpH4JOEthp17IfCa8aWtv/4Anlpqw1zKlRbvQoothXcL3eUXj+a4feHACt8OVStliemB8IIrfKsLJgUeaaA+yu0xb4y5gBljwJH3rBdGw6L2EHhdBWFmsrgHvOjH03N8g4We4dHHBcYdaM+KcJYVFKtJFyb5K2+aqb88Fe5jTAOh4onL4HlRIw480qZ4GWVdjGrcC7Dg8foo8az8nD5yWtS7FGpY9odHhocItaoYUuLGJl0PXOyPiaC95p9s0iJhmHk0L7YiENxmletmpuF2wm/WAgSM2EWo40vd3WJyf+09Asj1KYSbpCowg2ki6N5paLS9uNPJX/kv1twsa78YnShJovOlRWMKtEvdbnHtqJiauXn8vZTcdJI0ZmSLpTitBKI6sJv7pAuPqDVhFEeQiRUPtAmMHQyB1APnL8qIrfUkgRVmeCCnds7452HEuZVQ0yKUUUO/qSWzwq6RaACVOOWdlioj14WqqO0JQbaikPxWlhlbeXHknmmznK8UHNAQhZuYah2qWASW4LsMgwhpFFfBiEMS/ZMkONZfFWFp8Tb6NVTkScFaV5c768aDGUYVMYWK0AIlJRS5yiVSwuAsqynHaAvc9E2Qqu1BD+XZ0WVB/C3uije852kZuiZM8uLpMzxyoos9JW6eVgAKJh2cUyQM06HQSrX0qRkWzWcyo1Sb5UHRTq46aPkXwlOmftZxyviSGLi7vZTaVIyweMR3MuwLj4UIEKLiZ4YZJPCFrQxXxsunvALUjHTVcJdciCZECkQSl9VKkENLeMNfkCtiE6YCrS05bbxJONMCRMPMLBG6FZSMSljovigmoHQD/tVH2aPOUjPkSq2JRQQFA3qwq80KG6owswRGTEahSG1TdDC6u4SVdWlPWaeKw6K6iUdqoJzmXlMqCX2crT8cHphU53hlezAVshNHfLJzyGV0lz9uHrpxENdYEefr9AGEFswbAiq7hiY+Phu0WW4zJfa/jM4mqrJoPITfAwRllnPbzRZmxbvhxUpwgPBraJSpqRA6WpxRyrD3sakwi26m5hpemSgPIhnmYmN8u7mFNPLz/5GRifefwvh4NBoS2tHQ/UjcY3VyMeosr08msJZWx43lLmDfqUY1bGxmF9NQNAq6cX2kgTiVc2vmqWrcuxINRNXfvOz5AZiiYhLPqv+SjQVtFYXVdQMPP8WZ/oBKLbgqfBGmbWLkMxWJRZwKhX8+3hpKb/AVO5BhYze5Z+q4duO0YCoY2Rqp8r2IkX7Uc9/28TQDceljV1Zr/h56v5y0lY0KLuHsn3DhMDq7oN/xnSR97gWk1p4SqBySlwpo0daGPxuaiAZhTQp68Fq4nRTgoKb/LQWw7Xkd5MP5nt5wnTl/SnzqyelZeI3ZgyiSrIJQGaUdLRAH2/jwRJT6lWcEZJVMq9KV+xh+uqK8tRREWaE5jnqmeSIqBCqW2HzRHKD0FDNOdSWJxqiEBOEdqT8LBnbyiaPPabDMB8ZbmCHnxsYGizxHxwor3fogqZub+vGluhqSi6nGYou9fiYpajXZKhWSqEyoobKnUcEjKnGtqay1KrGUhlilpFysew3DO8Jf+IjlwpOQ5kj0TJBTUMNCuQyqk7phRO4yUoOVIdcxcO2VqioXe7uJekbLOiJwIxiSECsUtTHlsbtj9wFeqzswqkq6c2ltj2mkDADpIisHAGSrCWMVXrVIY947crmMtowiRm13MkNfNBW2hdfKEEOuUvIUSYua2lnJz8WxqIMvYqcbwb1SW3fdAKIyL3SCgTD8uNuau3qzmL80MdIs8CVqeRyhXMJK8jiPjjkYp1CtBKLNJJ/bIeTNs3Hxie+brKoXpWkYPSPaRVPGktf81KZ0nCToOqgMc1mK23OVLSDhiUmwc8IkQp2YBSISgIUnKAhdEn/Y90IPzz1Bkcj5gS5Rgiay7qAhIpuoVUrwSRUdupyCcgchPQBSTtok+tEM5BICbfGsSN73RSEp5raSpDKDsWtyBACr/anel0Ep1GYzfMGJ5xWnmoaG/DZhuZG7fApQ4HBnM4ELrBhtLSqOIQzeLlTYGK5M8F5gn8FbwAvj9kki50tbXGxVPeTh60FUmlS90rI6F9wYFZy7abtaBokX9G64lDpZmNQ5ErHsh2s8NNbXrkaagGH7/8q0Tm2mvDMXS0T+GW9UHb0Y8OmCiZUHCbJLfY2Dcl9GRE9C62sHylBqr2PY2mLsdEzMlr8E39Gf3uv4dGWLu+/FQ+MDpQNWbTlN/uwCYt4I/q4tqC0ug0kFlW+Gh/mrrHiZ94rmlTWoaqGZcEvvIpFkQVe/JB4xS69qKpBPuHn4Jf7u+mprD6TJ4l48qcsqfif9Pi/yhM/o0sKqom4V5Eqcp84Hwb79th8WS/8K+v6qKKsP6wuqhotD005JiiLlknHSVP413xqOA7aNZXbn+MJPGGWrWJ1BENifXgxgd9slTLiGzPLvaxx3Bdn2mKrrd6GBofBq1u63ypu7T/SB27nmhrE3XVk0KfYx2CqO9o73XaLcznN3Oa3NAeHW9tsbI0xhzu2AH6/sxWOuuyAy2KBPuXgSbv82qyB2myZUW5nDmWiYUmK6me3NTSN2sxq0Mq82FCOlo71vwMdHsCBh/mkbParw534rfYS2gkp/+oIHivrPwJSa5V41bkBXrcjN7bU8Z3iDXODaBipxXeTa9lXBWqitkhOr2VE5WhVmUTLuClDwU3TcHQ3MTNhaEiWrWzBKUPk5ziHj3jqFV0cjq5LTSswy421ItHhPlIWh5X0o3wKTpGONfiQI72mtvLD2I2tglWOEZpNh+4oBnf70Miofd7ImLNM47Y6WSyzW/ndWEvnxha4zVbOTbauFbyppX10qCwM/fop8dYyxUTjxBFgWeUW3SoPlC0XI8ueqCSRZHIIhJt9flM1s1hjjBpc+FgemtGdL7XZbvoiIE+GvDvnTSKro7KZdnUqzAAAQABJREFU8JHo4/inZGeRXRxovuWH4oqiJRtRZT4uPokfKp8ch89/FZxWMqEEtCRAsSKaAFU5M2GLLjdkFxXacfSvUmEijSouR9sVaxDsiqciQ/sofEJqddsqvUfhJBWSqki+qfjho7YqN7eXGaRM8dAKpj8bGn8l+wqfsoovWez8tZgBOsEzctWtzW3WcnDLLQC3cjt2Z7dftT80XQ03enBQbqVedGyvMJ3etQ8PFRyW+UfmwKyWCq3uuBLZHGlXUrK0WEaLcS9N+ZKjcllmrCZU7h3mdv9ZCFCP/x0jVqe0JlM3OsXN0VFfuT2ajTEassAw31spl60se8rgLJRHff54LCsfclFxa1lExjmYFEe5RFjWTyXmcPin4ByXz2Se/5ntapbhO0qWqE2qXSpRuNSV5mI3GefxduPg4d6KYGpVZUABmkCUNNRZepeO4qwIK+3gTG4U6BMW5KhSQww3fk8jdY1Qo4WlS6VQtflqK5/1kdlwvTzqbXwfrvC9EhIdc5rzQlmCFw8UT5UtERwHvdrgEoY306vtVqC2efLMWT3eYPZtzvMYOLJiyjM2y4+fv52V43fu9uxrYLC8zleNuupBXTnlKSed5ZjTvF7t8duajv8RCNbhw6Q4XEPBtsrL4oOYXzf0Fuj/M6VWj0ZPoFvdNRl/shGce/xSOyIsHmc0yRGPp2d1DIFRxEyua8IpYuBwP2AQXFZnigXoHuwyBQJNtAt5ldClUZUg7Nu/U4LOmeMn4cqXnfieeK+LjDkrHHH45d2AsvDwg9Ay42DvwYULFzsScu60a7cPtvtFlE6fT5bcs2fPdTq6d5+vivUzpN4Rcegz2tjU2dXZ4yvnrP8G+6xr262TsiL0oK+7p/PA4YOStbtrxsBg/5H+wdlzZrb76eKxwUOHD/kGxX0HfPdOedsaQvWAoMWrJjN6Znk4bmMxPDxg7jCzTuf/YnJldbzEtDLHVF6K+S4V3lIgT8dnwln/6f/QKjLrxhQVElM3CHDtOt+moLnFH79M4Rt2NWp61WmA1wh1o0Y+biMKIeflkIhlwhCegIGrJ8PDreos1cZN97/rXX80Z47fHvB7137Lwou6Yx/96MdsYX7lV97krSOHG+95z5+df+6G//aG/75vn4+w+SGiubt37/rjP373WWed+Rtv+VUriv5+r/Ga2HwHRseHP/xXR470/8Ef/KH27t17Zs3ukZp9/YdseizqHcfOmNmtBjGznnHmaX/4B3903/13ve+97z946MCsmXNGRssxkzenXvrSl33wgx+cN3eB95VMhSbj1avWvuUtv7lu3bpyMOrU0qLVzqasLI9TaidwS9pl1XD0YJLhaAJXK4Ech9F/EYhKkRzFavUCry/rxnRqTpugCGrWdQMwo1aDpDrrtZ2fBZgaRNGeruDJ3bgl+dQiYQFR8BG6rsp0x2OwcFD6B3v39e4abfAhtZ721qaDffvMf4f69z/66KY9+3eYWT28vuOuW5781LPafNK4ZbRrRiv8bTse6RvoPTJ0cLzJ08JD3nIxB9mWHD6y39sv6m07fE3f/M6e1uHR/pGxwYamEdukcsLaOto/4PdkrXVtUxruf/CuscbBoZG+R7b4rsYFw2P9noRbE9LF34GDe+3PnFNyRTmlbRhauXqZFa9bfGVX2Xe2Nxw/BIWkOmNmqQb8XPp9ENelr1oAVN4qgzznoC7/3ymiQxm6VYEq7UCSGOq6PIHOx/dOeE0RMHFZPRVMehIQN2n4TZCSWCCVK2vZ9Duu+CgNzcm8Fxy9jAjTIweXhVv1y2Xuht5TTOpPx+TMM0+/5JLneQHewhFPv3rn8/KrVi27/PJPL1/uC7m98OY37B7yHXUXX/zsz372s35ky6fhzz33aaeccoKV6KWXvugjH/mId3itFMFPO+0kP0DQ19e7cmX5qm+anHJq+dE073E6ebbcnzGzy2uaxpK3M9ne2dUxZ8bMU0875ZRTT4RPf18B7rcaNzzjTSAXXLDBS+xu9JazknLt2jU9XZ1GrUWFUUQiDtPNoAmBWubFRTn6plJcQQfn897re2IXHddv/wnAxDf1RI4mE2QtG1heDbA01bk5HKtYy1hTOeWfKNUz07St57nGTiKcLIhu/emtV111lS8NlDG8401sv6t+4TOf6d1dfrrv3nu/+tWv+n6Ez3zmMx4ii4Az36JH08RLHkdlPP6vg0N82rvbdj248xd/8ReF0K9+zp5VfqxEMOTWS1/64s9//vPuhx74lMeYJhy75ypC9v322k5F2zpmnnv+s6674Sd5muKNumFPEceb7ntwY4MX0rrK/fGnt9+xbcdOM1tHd9fYeLcF6Pkbnrti2ckm6raOO5pbu5taqi+0PnJk7vzl3Z0L+vof2b2nPH1A+5rXvWHjxo3vfe972Uv1TY9u/+3fead3+K+9/sfSsamls390dN7spZ3dFp19Di2bfZJ+tHHV8jXzZixYtGDpffc8XB4RdbQe2H/43Kedf2SgvLRmArVRM5BbPDefJkOJbu8o72SwvKO785Pv+9SnP/1pQD+xJOrCyV1+xenqq7/ry3Udz8r44llrhp8vTROHOD8P/fe7coAs/xQsqackKT0J803m3uwH9970dddd51VrMV2xatXznve8l7/85Z4JlKnHwQaPVRkL0xFJXaptxzT3F0hcgJ3CRy5f97rX3XDDDS7zLF7iSd8PfvDDGzac99rXvlZyvP/97xdIH7Twzrk5phYzXYMZuW3JZt/b4TM0H/vYx3zWYveu8tvoevFXU73MyuVI5fils61zzZo1mMh1Bb752Gvt0gIHc1VU9Z0iPg7hbdzhwfK50PXr1w8MD3SU7fzEb44h5FMuI89HJiQQQuMEN9Mnq8WAwqYrvVTRS2i+FEStjQOFOzs6Fy5YLJmcBlSrT+f/Jfujj2UrHL5FrtBEfVzDwJFQKUb9+fvf7ztv/+RP/uSmm26hCSbJCTVuxUv/RYUmKeTHA7mUnVu3bPne9773oQ996M4775s5s6xz3BIf3bLl8sv/9vzz/9oC3Wdj3AEQciyq41ow7S0+0aoeUje85CUvMTWaMIQHr9WrV/vIAb/Mm9fjRfe3v/3tgIJk5pMBXh0nLH4vkfAE6HgFgmxAKOnlt698eec737l169bvfucalohK/A5N43HtzfGCMWmecNJ+0tqTFi9Y7PMYScc9O/Zc+c0rx4ZHfDQB5/LO+ew5N//4pl2P7RrxGuVg05yZc8550jk+/dDV2u7pjFHr7MBNwQvYGh71Pu2ccz7/2csZ62W2OXO7brv5tnFf8+i/sfGh/qH+g/0vf8nLX/CcF8TGebPmtTe0jw1556JMBnZUvYOHfPRi+/bySSNO81omMxVuQMIWWWXwuNRwKzieeyorPVusPqWuNmZPPfW0D3zgg894xoU48AkOpq14hscmmOQeOMk/5ez2P7JYcDOIeoRUjaISu/bt3fetb337rW99KyfMn19+2VEsXvCCF3i53cdjfvjDnzznOc/xAZ61a9d2tk+8SHBcNY+/QAyqBPJpFR/iueKK78hXkwS/vO1tb/Opg4cfftjI4H0TDA1MSDSAIKhioIgHr1H0uFIBE7AgwDeSfKbHJ4G0Y+HkvAzycVl59czywxcPlWBXkxGdt2zZ4qARNx+ps0exCqSwlRwc05Kb9aoVq8quvOHxj/OT61iKmcNjIwt8j/dCX79YfofADI2hUWTMcDEEtwgp7l5h5VrmzpZ2OnAFBJpDQHjOOU97+9vfMW/eAlNmlYtegGojMfbyTCXu8dsiyJQCEyT4RavqWxfjVX4giDjvHQdtCu3ky+M67d8RSJPJ3BI7WrHaHGnOoqcPRcmlT37yk5/4xCcA3Wn/7M/+WDje/OY3d3WWXw2g8GQmk9vTJhBJihkR37lzyw1OaDnIbMoxAuOOY9Epq8SGByHrFTxtyNJLLQwevRz3T+ScJjpWFDa1V/Wc3fQeOKQBP9npnmmWoS7mquqvUj48q6bH477Dc8OGZ2zb5nO0pl6JYpHnB8O7TjrxlAsveNbuXXsxdMoDLgU1TjzxZGy7u7rdzY1292LwJLdsbvUBjPLkNlNUeZnPKakf9Dp40JGo32fy+5k+E2zacODARLcgmeT7i30zByofTBvZtu2x5z3vBS++5MUnnGAfVlYaBe/oLwVWjwbKy2YRMTkYU9rCVudxnAnCb8WflU/in8o5BXBc/1DvP/hvQuvJSUYly/QHHtgIyEwrJesi34QPno3vq1/9ar9kx6hHNz8qbaTvFNvry2kTFDF2Jms5Z/7gqXJZfQFxf/W5b3Oe7zD/+Mc/LgbWvzgiEWDzaDApR3YtaUoDGkiSHnOY7sc++Yo2mDEYT42S6NOUoZGhtpYO39iW2VeQvFGKxMLxtNPO8EfPMoeZjBwyjZTjMF9z5/mRSNuyQOPBSHE46dPIvnmMKQRK5aLYmC/qLj9MKqMkmTHgpkEKEYZ9ZkeriLLuKCtmFo12dnRb+fgZ2oO9vluP62A2zp4117rIQT1LC2XlLuSFzfFKJbGkNd1oGCVNNgh1KVN8cjweBTaN2/7dwMfKDWv3cZHkfCbQVnQc1FDd/AVywvoTXv/613+n/LJm+Zog1k2n0LQJWgVjzDpX0rhxk4SFnaP5GTsiTZyz58zxUdd3v/vd1mq6uF7R4DtuhTOdVHA4bFMrkJOv1qColPKqRPkgS7aEZa46ymrqPFFeqhzz818znvKkc7SP9A3YBvXu7120YNGJ60+c3TN7xbKV4H2H+g8f7Ovdf3DBvAVnnHYGHIyMtJ/dfqen4sM+GTfe6AHnT2++5c7ykwxDbY4Dmq1KW2/76W3XXnPdyNDITA+Bunrwufr7V2/bvt1rS20+5dLY7GdMb73l1jvvuGugf2BkqDzSdPLzwxt/fMcdJpE7M9EaGNYJvih0YDBveRYr+KoyatoZznzM8EyTamMmW4KSE44kypsGpVTOmd4/XP0f+Uex6FbX0ccNZ+nSJdrGuUFLTQiVOeUrTLhJRskr84UPq1Tj/GiEf/7f2rCfB+NUZRgW1vumaJfcgrWPt5p1+MWA8Plzn5R+05vf/Au/8PwkKDSZXRxYHFeYTOV79BqCpDR64Cet9VjzRRAOVJepT8AhnEyERO0/0HvqKadv2bxt4Ej58SgPMJcuXb5s2Qo4K1astFrr8+SnlMPz589ftHARzrQbHBj2XR3uDwzkI3dn39xo2jPejB455ED80KE+ZH7V3aS4Zcs2Orub04qhlXreWhp26yfIzIiJWzomW7ZslUyPPlqWwjKJgUQ8+ODDZRapPuXMwJRYcWxNKy6Fw1EEKZkXCNaGf5RBERrIsUz+cyC1JpMtcu5hqcME/qEev1nBu1FmsW4e5RNdSCgJbTpVp93Fc6jstCdPDmHh9q12T3/xi1/MTbnp8JPDFOtU3/2i4R5EDznHa2TTzOjRAKEicr3FmxasIyP23fMXLPDllX4XjX4gjgYF3rP1Su+SQOW1ORPAaINDQTwlPFYFuTqYEMW2AvfhhNaXvOTSZz/7YjY7AYNAbbhS7p3v/P1NmzZZK+/Ysd0KsrPdFyFRvjBh0T/+w5e9aYSz8RBXLlsirZt80ysOhw72nXnG2V/+0lct9nlWekEzRGfOmE0luQvoFOmZFz77m994Oru4pX5xxI3+qu9f09dXPiDAHLV7ESfIe3zmzplrBTBrpm9rmhjD5Ti5+jrINEwwphbGigL1FO24JVMRk8vnjKup1DpevJzsHDp4wM+KMU2RDey0u8KHXOSY0D+RDQ6t0tBFKyYIECmxpUSZ0OpI0bfYC5AI4pPbtF+SROuV2YwZQN6o/F+Oz9y7xTldqKxBbZU0zKeEwswURpDLtKPJlHraBIVXFK1+0dadN5Ywz3cK33DDDRs2bCADAuNLKjQ1vfCFL7zyyitdylHIVIlgrkFFD+nLePtrCC5PPOkk2emTT5hs37atLCSqXyVjJ9vQasg/avhVM8Oxr+/Qxo0bZ8+eqU2irrA1Aq3toMk2IjjFNGyZaGA89NBD/GX143WNA/v9YNRwW+vw4JED+EhKKb5y5QorbLQLF81HYpkoeHff49d99p/79PMf275z2bJlvu3LJkm4N2161FctPfWpT1+8aLGs8oKIfVhnR/uhw+VnSHfs2OWVFCcqvb37HRLs3r2z73D56IUuChiRdgPCsGvXblO47HzgwQf0mpu3bt7GIjrwJ89QWCDlilovIE86MbArhU9nNeSUtPlKA+E1V1+9du1q3pAfloB2CDywbt06IVBCgrOF2c4dO0TKBGHgCYQvAKcVr9bDHjmedRClLw70dwBCSd+EvW/vXisYkVq0qPxaHL+Zy/w2BZ5ykcLOJf26ltEl1ljBcZBinTNvwVwBIgtQF62wxZz+E1ZN+cc3Wj7+13eg//G/3rHRwf379vzvT3/Sa41Llyxau2bVyhXL5s+b87JLX+LDWdx1+FAvhL7Dkmz05p/8eEZP19lnnbHjsW0+qD3smPFI35H+w7oO7N97y803/e7bf2fVyuVzZs+cO2fW7FkzLrxgw6c++XG9Y+XzFfw+evddP/v933tHT9eMZUuWWzgumLfwf7z5V3ds3/nhD37ovKef297W0uMhz/y5yF/0wudf8e2vl6AM9e/ds+sv3v8+oil5wvq1s2Z2LVk8f/GieZ0dvnug4e+/8Ln3vfdPNXw0cvbsjoULSJ5JPsi5T3/KA/fffdqpzootSxqWLnVmWXbCPd3t5517jqmISo9ueuSrX/nihvOfNmtm5/Jli+bO6cH/mRee91u/+b8eevCeQwc99x8aHDg8NHj405/6WFenV1AbZs3u6uhsmTO3p7unHbtPfuqju3Zv/9XL3jRzVpe/VauX/dl7/pjmf/2RD3Z1F3mrVqxct2atb0To7uya0d1DARrO7JnhDyMQlw/e/0D/4T7vD97yk5u5ZdGCxfyzdPEyi+AfXH3t0MCg1fOXv/QPaNtam9esXjlv7sxVK5cypKO9+R3veDtDvKVlhG/Z8ujHP/7Rs88+c8WKZd4c8JbCKaec9KQnnfXnf/7ezZsf2bVrB0yPar0O9vKXX5o8cYBDm6VLF9t3gLgLXXPNVc997sVr1qxCbqwtnL/g9a99HcUO7NtPw+9ccSWdfeJryaLFSxcvOXH9Cf4YuHL5irlGxqzZFz7z/NvvuOVw34FDh/cPDB7uP3KQGweOHOrv6/U3cPjxPy+ClndBH89Omfp4dh7oO3xgdGRgaPDIpo0PLVwwT1pIUMbLsM6Otss/+7+lpl7Jd7B3P8P8yU55vGvnY+AedMrRwYF+5J/4+EflVndXh/hpdLS3zprZo83iiy961iObHkabVH75y146a8bshfMXrVqx2t/Tn3ruBRsuhOavq7N9+bIlhgcF5B/yX7vszYaQw/ONDz38Vx/6sNzFdtnSxXCkkQhdecU3pMLoyJG3/c5b29saFy+au3LF4sUWoA0Nv/5rv9p7YJ/X6rdtfeRVr3ypHF20aNbyZQtXLF/8J+9+F312Prbjz9/7Pt6H3Nba2NXp7MkL8g04L12yYEZPB/5//4XP79yxdWjQL+8M79u78ztXfvNZz9ywaPE8b5msWbtCaBcumvu/fuOylauWuky+smTGzM7PfPaTf/mB95nFn/TkMwQP59NPPe2Fv3DJS1704te95rVG4/Kly1avXEW6YJ92yqls9OeLIO+842ec4y8J2t7aUSdof9+hb33z62znqGVLF7a3Nam//70r3JaZ40cT7rrrZ2eddYatHxfMmuV4siwtpIxktXBasmTRVVd9b9u2LVaG/pDceOP1F130LHN3T0+XBJXQp512yiWXvMBlchcE1aIFC5ctcYtokKMCvn/vvk0Pb3zORRcDGmPykoFsYRGc33vHO4eG+w8e2qeWmhN/fb0G+b85Qf2wNzKKmuF++Q2/tGTxQn+yRo7y0Eknrhdg+ccvGjJSba41k5nSkppy1Dz0gb98f2tLk9SRWKje9Ye/f8/dd9526y2mPRlP6d/+rbfKcn9k/cob39Da3LFg3uJTTjp91Yq1c2fPk6YXPuOZ1/3g+tt+eivzTj7pBKwyj8r1T3/qEwme+h1v/12+4BdxNRg+/KEPYLhzx3bZw/g/+P3fnTO7Z+GCOTO6Z7780lccPNCrd8/ux/Re9f0rTH4z/HhaW+PvvfNtRuZ999591hlnzp456/RTz/ifv37ZF/7ub+66844HH7jPLHXeuU+j84knrKOGbPjc5Z/Zs9uPGY/yAKu3bH5ELN34Vq9e6d66yvOAVWWj9trXvvrzn7/8T//03a95zatMRcuXL/2VX/lleSxXFsybb7J5ZOOm2pDf/s3f8nDlhHXrxd6jog994IN7du0uvaNjVYIuWTh/yYplq5cuXnE0Qb3ubztXkph6AiQKfP7TW36CRrb5drA777xjnnE7a4aEk1Mnn3zij3504wMP3PfkJ59tLqSG3hkzur/znSsOHz5orYJq587HbrnFrxsuYMKyZUvMtSecsI4t1P7KV750zz13vfGNb7Dicgcw5ufNmWsgGdV08NEDmfqGX/pvYuFPjkLwd/lnPqt3cKhPXqaWqSZRaXawd+//TYIic7N2H7/2B1eLirujW7zYsN9k9v/99m8W+0eGTH5lNhofFde//qsPbdu6WeLKToQ/uelHCOWTmdUc/GuX/Q9AJJIeZ3PnCy95gWnpS//4ReT4vO61r5agy5euOnH9KcLgXvaqV7za2ZDDnUO9B00hr3zFyxYvWkC6mRLbF7/oEudKwjN4ZEgSu4PwlJC7OUodU7u8qW7Ww7ffdjNNFi2ca+F+80234EYFo2v/vl3/+A9/p8vkal68+67bjR/3pmdd+Oyvf+0b/Yc95Cg3B+Mnc7y1yvp1aww2w9UweP7znqOXLf7I2vzoJg/VJKi1moZnbCIqL8NELfBf//rX/F4c+MqVy9evX0vVP/i93xe5HdsfG+g/YlFxykknu92vWLbcDMSiB+673/29Yj9w+623Gb3eQUmCHr3FlwRF+7GPfJRK1kJucaKzb+9uf4S6a7/gBc+TWFJQwrllS0R3fLl4xx23ST5AU6kp37gy1yZBIXz/+981nBDSU6/3u0899eT77runmDwyZElQ7kcLFpog1XzontO7/wBb/H3jn78OooshgrJ29ZrtW7dZA0hNN/fMnRLUvd4t6AkSdNpjoGxxLJNtJzds2PD6179Wg4MVGxprbW+p3XD99bYyNivW8hbIJ518stPXvI1mn27T85a3vIVTrMRR2cF4fgDT0tii2/ZQIH0HrGey73rXuwSMrMWLrXXKeTh8l9bO73nPe0BsNSy3CT3v3A39feWplTU1HOtuLzR5z933zJ1+yunz58x36VNyHa0dH/rLD82Y4TcO24YHBh30XPP9a5ja1d41q3vWe9/z3p6eGYPlq1y7QO7+2V26tm3ZcdbpZ5ywdt3QkYE/ffe7r7nm6he/+EU4H9zX60PD0LAd6h+c2T1z6MiQZ/qecGpf/4Prb7rxxz4Z5xOp3grwgTu7UgdMDv9sF+npcOMP//CPKOtb73btKt8F7gzEoccZZ5xmm+JlPHZ54sd8bnQSc/XVV993/322LCA2FhdccIGZWJctBbfwG7i2ekqx23Mg7fvbnDCI1Bve8AbbLBsXmz8P/K688rtEC5x4efzt9Dou9SjYV+DiTA2hcaT4F3/xF9xu14s/BcD5n7fR2ph6tm7vRXMIei+99FI7OZxpOH/efE/F7Z/o57DNfqi9rbxAo/CDzZPaMUdVlXM3/LPH10Y+xZz6ctoEpS694WHKqssuu8wBKFWwYxvjlV//9V/HWkIBEuxEQy7CcbrhiPS6a6/1XZ4slGq4ecncpKKXxiAH/KxR5a8vfOELv/EbvyG/XTprDIJIaHjCfsIJ67mbF/BPpvpAhS460I3c4MuPOfNme0TEayX7ncG0tj5w772cNcdvTh469P8zdyfgflbVofDPPGdOIAQICYEwKPMkMwFBcGr92lr9hIJDr4q2tvZW2+rV2+G2j8916AVt1aq1dnAWFYqAzCBWGWQQFCUJQ0YynOQkZ56+39rrnDf/nJyD5qm9z7fzzz77XXvttddee+21h3fv/eKEbSFoUuakLQs0sdp63333dXW2zZ0TX/6UyQknn2pl13Hkvl39Yq0aqDZV4upBCOanZqByEcUpi+kqNiycyDfbMOLZDrUoW728ZkvOGVSYqF1++eUUSCzF1YaXH364u3rnL1iwc8eOz3/+86ZXyBL+lq1bLr744qrsWdPkwMmCy0AK30YnU2+Mbd229a//+q9dwKv6zXdM/z/96U+nmUh82/PwiRoKlsxe9rKXWbAUxrMCugo4VzbUEUdZ0RdLAnw1wjdn56trqxxKzaVOW+157NFHQ1ZFOHDkgg3FhxxVVjJVNA6cr/o4YcjTOlfOWemOfXjxqxblyg0ikX48Pl7LFp5z7rnv/L13eK+MaQzJUmGU5A/+4A/kCgErgPKAoANQi1/+8pfx7ViPtXSYFlMOPXQJFF2rG2VaWn0h0xddfT227Xd/983y0k9ZO3T4scGNjmP9/HnzZ8UdA+ND5sXWF73YcTOULDIjBSPT0fq+kbreuuaRXb3bX3LOqZu2rm9srW/paHr08Ud37o6t7wP9gx1ds9daM3dgraFp3qJZP3zovk2b17d1xSS6f2jwZ6vXDAyPtHV2rTzm2CbLRrt2xOubRofhbWEathLhA/QNLQ58jvF3Wj0aHdnVL9Px+AKOVVi36+jS3FPa1jowMmznwfDgQNxm4J6poeGu+PRlh8OdRjVGPoY3+sotz28yHrCCeMSKpW+/+q2+NDZ7jiWeoS1bNz/40P0uXjQj2rrtedOR173+tXErkw0xDfZdDbkryydI7XEwdLYPzBBQd2xh6/HHHytbJuJK21NOPuWKy69saWnXAq2P/vzJp376xJOmnl6nkXDPzl0HHbhkyEfFBkfsxTQiP+5Fx9N3EOdjF85f9PSaZ7wVE2XRzaVg2GT0mCU/rxNNcBXBCMeAe1dP9yknn5A1rvro5I6dO77//e+nttEHFcQXlRZBg4wmxZZZoFUxegPrsna9mPC6b1A3Z7+vs1WTv/KKqmw7mFZzATUd1AVYUDldffXV7LwGnQ2LBaWjf//3n7rnnns0fVlbIaO7lIZNFXApOs4kRIGv1QZ/ucWzZClc/kZsBZepEnqUS6bN5MJEkEn4cGCmpiYRppqR1td7hCDWHtBYkS9fJbT+iiWpPOpuvHUUBROTOkHt2+4421IZcos88tIslRoRTkItTZ/FNvzpn/4pX17JSfKWpjRZAlcxVSwEAkk0WXP6mUUHHCALojUi0psnAmF+5jOfMeyBjKAWfs4552Qs5OwcUc684OCKySTVAxcvfs973gOftLt3dNtbjWcIgdzSYhcszkGwF64+PpnsJSMEYpI1w+ZbCPFY5NazqyffrlUQZcmkCUmZA6Zx9RhUJ6uVJZJvYv5K/BlNa0qZaASyeAaIH/3oR5XOaFLlqSTFMzF829veRjWJHpekJgmN03dYJQbBJYhi5HJ0Pk7LeiJXCB4rSBWoEpbIWAHmCAiXfBqW85IEas3wMaaSKKiawDxOmIP7778/SNXXU1Bl0TN6uuCCC8QqLeKoSagOVDzVEdD5nnbaaXbgEkiVS6Ex0cyC4KTDwGQw/iY/FIh8ZAHiRnAbHLR5PISBKf63v/1t7R+CHK3823ZOtzI5EpmQjz08ED6nRIZSN918EzTKfcXlV7j825s5LS3V1Nimf2DPfjaFUiLIIdJCUz1m8TEpVpTWK5ZIMy/wQC5OWNaCAmIxkDjyguMFhHv7daeJ/CvxZ1RQ+clbxskQkTGZF1100fnnn69IDAyIWlRzGs3//J+xvY+wFBVQlZezOPFSy6PC4DVHjcL5OIX7kEhBE+DEFsS9qnlKEo8wockimKyvX7ZsmTGiMLZx4kMFvogIzpSqJPUniop0dLTTD4aRsL0FxWQxiu1hQcv3YfnKwikj+pTSezL7qd0Gb3qeRm5fZpL/hGcRMpxwuWT1s9zen61atep973sfNAYVAn3SVDQkBjL5se5D2pQJz1xqWwpTGI4CGgKyAuaRixZadW8zujUSxbOiKQJkQqQxpi8pKwnRp1UpXGQxoJgqEQ8eoTGx6s5l6Wk2K3htiWAizqWCSpVostYdeciC/0r8GRWUCJIJ2WBI3hwp6EH0pDlMxh+IsHmiV6AGGUqvGuALoxCyKNzziabiWHmq8H8+QHZcDHxnzVq+fLl80Sd0+qR60P/Zz37Gx4A+EaZysSs6emh0gsrqE0444QQFTJ6z4MplfmPlwZzm1ltvFfuRj3xESRUtqmE/nXw1UQLxqp3aXXXVVeadevBoMJ3e+bR99rOf1cmQp9wNbxhCk6ccLFEsIsUtIvysmiyjT4soJouAPZqKsoJjja8uNE7alglRkJwQaCRI0kxISizRjHZYnCycJALglSZ4lMRjUEtomd/ASSIkiZNM/ivxZ1RQWWqFuCEyeSejBEemH/zgB62hKIZCEjoH8x3veIdqs76AdZyrCWZJoOIyiCBaVDP9KiqJyysDGesxIbVoU8IwKzT5Whnw2RfvggFlp39kOyVxApMPgh/FkUqPT1klYYEyivlUc2KjaosqM2lGgQ4JKrXBtw9x+8QMCvp33cUUTjzWcltbwIRTRFLiBMgt27ZHYWntN7DAJErzpm2Av/Vbv2VqX0ks6SdZPgdHE/JCnE7QKtK2SEJBZUdFMlONVhk5+CACfMVM3aqKoFCJnwjKqx5lnbxBy3pEpGIj8CcVV2yFkJJhjSvi/8nAHgWaQkiW2MWi2lJnHoVBzCEuu+yyq69+G6EQJSXQ5UWZ6+stoWUZIJNRKigIB0L0KKSWoDMlO4/QEpjCKukmIFVUlSpjK8zklhr5Hg0lA5eR6nc6RS/PTHqxatPGhRdeiDHFEcWsGihzKhVvDkWlHVJ59hPecMMNiqmqKIHSvf/97ze2sSUKRCwKOKlyr7iqAhlV+4ifTII9xTfWlB00pEwxjT41b5ykTmDGgNhQKikorAJCpsECgCAIgiBF12Ea21x77bWAohDhUzI+BD5MTkKPCpsBEFHCaCZZYVkggrfEAYeTDkSAL2HwubdBEQUZvCIF8z/vZlRQxcjSVs1RZjhTbJVkpY3h1HZxQ9Z01ADIelv2mKSQXXyKgK9WKLoyCFdMZ2mTbBUl05QpuVfIAhxOMgkfDsysCWEu6DQ0rFy5UkIIslNNtFA70VN7PPzww51XoQfZ3vTy1vkxr3ZtnLJwiKBukUI7ZW/5LK2pLCxlW6CmAcKQJSec5CSZTxHhISWWlVoVk7iEVTzGOHWPDuJIeZQjgnpqYXDcGusvW7ZMd5/9exYtKaRuCWcBFUpefFlIePMtNzvOyoYhjhMSs/YjOwjJjFQpQ5VYJBoKmi7xhWWHH5yobMS5BPIRyUdhAXqAMkxF4wN6RAfbiZyZ/if9GRU0Jj2lMvSG73rXu1gaWRMiWWBO+HOf+5yOniLiBpCZIT69v0f7GpUfstKm8whHKoWGIFDLN8iUR6kSmEKS3CMnAJOfAZCqIYET1dLDDtNZqy1RXtytXbv2a1/7Gk480l1dNiHi2fK13p8iyoiNXLz4QN+DC67GY+u7VzuMsWrGMw02BQSMOqurIxYtU60nh8hyyRKfZPhcVeqKT8RRSzUVhiDMQTZjs7JDXyXnnnpqzbvf/W48py0AiYoviyRJWY4ooIygrsAjapLbiP3e974X25Ikt/BZ5dRREI72G9tAEMUlJjqIiJUR7SQQW0sRBQHHSTYqohaQKgslrdz5si7EohEKqwX6IWEC05ckU9UCf5nwjAoqcU4wFTjnQOoGN2RBOkqiRt/xjrcrsGKkgMjCmugPf/hDLNKAqgAC0pICtJnYrbiHA98j/EAu5VRajqSq5HDkmzWdQI+AwgweHRJAinrprLPOzO0MH1mmFLQh5t133+1RKnBJ0KfZDzzwgE6cpUEBGwqifyB3ecHERpY380q2JUy2k4cMJwIiYgv7E60rk4MjyBk3+4YnpVG7kiB4xBGHO2IlSUoAJ+CUhcwpBOKZVqwoMlc6qTiUDUmtOWiBkKFxCgszKfMlwU+uAYuVi4RZWI+IywLBjBIrgElJSg5hLtLEyEubQTkaWRnpJgPECC7Vr8rNqKAKhj8ca1IWsa+77jr1hD9w5RGFRa8ozUg8QiMRzHnd7F2ZWDUqIdbxjVeQbOsV3+CoeZRWbFX+CqEKiBVOOglMSIbBUeBQ41vB8SrPeEMsyQoYhlJTBl4nbtycWwVQwLm3lJhXIjO/7FLVDa1V0iojcI9wEsKXi0yTh8pP9jxWgYyqUlVwyQGTAp/QTL9kAZ/Ds/n7oUuXJmOkiiXNhkoRODYgSC7Wo1itUS1AYPYInL5+6EMfMk+YsGHj48uXL88WIi0RQZNcvaKDQsoNtWTPIxtJGtpDcgiCq6ypxIm6nCxmlg6dRECTwUrgr8qfUUGTs+RJD8M0xmpi2QzPl71imzZaKyFipdKY8G3d/t577zX0IQ7wiSIVZqWCIJglrwqQZcuoDGcUSAKrxypJLVlAaElTQHWecsopmgehi0qJg7ORZh7ahF4eMlHy4WhCWD3rrLPS5Gj9a9asUZx8FAtTxSOVdiJ5kKTiQaAKZ2zl1/Kf8hRVVXbGemXwyCOPCctRFrk+31sGTrLm1q9fb2TZvX07Ch5hyg6mBm88feWVV771bW/zBoHmiSLku++5+7Zbb/UWHpMgtq/bvUCHZC259WBphSue4WTHDShs3561XnmRFYiE8DlRkqAgpbAoDlxxPKa4+OScUb8qf0YFzQwwpOTmByyQvSp0TrHxlOKmlBQUQilCjJQ1Pgtv5KtnpCXQ0MmCkV0qjceEoyOcJReAmfD9KpvkUmVybGCYIqoVXMnOI5XFM2CefKK+mzdvkQW4WAltgjFsrYRu5idWVBZTWHJ+xTxMsSD75bKkkghkwQUw/I//+I82vCUpWdiduWrVqtzbICOdNe20oSSZgZ+qln0uICFv2rjRHE4rygnfQYsPslwfjarI2WK+FQwqKLssiFpLIh4VJAxH3A8w4XBiNllVkMeUMB9GJCwG1SO5YzgxM2rb9m3qXSeWxfmV+C+koClHzUIPorG6PinLpvAChOXVYc48YCoqZ3FUyzaX1wo5YUXKolsYz7pPsllsZZAKRAAaYFWqTFX7WIVr0YRT7gIpdwN8CkfuOJRELIaN6jI5q2+PhSjC5XOsflYJicNJiVdsg9AGPvr45ARqWfXIJfEpPrQKAscjsgV9ggKB3HHHHaSESfwQteswSDUnyDoiWbuNY82atdHLT7bnwkW0baxKaHhgVKMimA9RSqQT82aBEdUhGHkb1VRKiR+ZwhHQUARMf2lwsoqgZYUjjzwyhYBaqdXwINQWMzMiZJiIiMI/OocccoiWn5Ksyv6fCcyooDhIpskoOWMa7WbATaoUhvBhenvRRRdt3bo94eRIG0IR6+tJLVlHCgUqbkSowwhhTNZchvOxtvy1Rdo3tkoeUp4c40pOWPloAieMTwqhbjZs3GDjowqDr9kYt2GMWigC3kAskUJGAQk2KaWfPEiejCXxLEuVb2bHT5zaQEYlEX4mT4SST+zkMF4nExKmmpgR9v4dn9imeY7t33XXXXfffe9xx71YvojwM5AMKALtxLk26bU+deEgWK/9xCc+kR23tqf3yJJK5T4VexotswTDZR5moSO5gqOw0hqsyz15rkqdxfEILkePZKibSkg2PGmXL1+uXAmE+Z93MyqoPJQBxwLmFtlRGoAnK7RQeYhAdTJX9lqnOIwEYAIqq840+cv6oLX6IKWqgODKWTmPiJNvZp12iwLCTxzMVGnhpP6lsCCIEuZ7AcOcRE2WSpXjvLnz1FMK3XBFr4dJFCDQDBAu8wDMBSZ08IOgKldSj3jjZ0Z8mInDxxhSyVKS5XOiMjbxk2BSAKeI3/nOd7KOMSmWJG1pZUE9Mm+W+f7qr/6KgRdFGhhImhUFxxJINcXC9OolsOFRXdBs78CouG2m9EZ9SYsOItYoCl/RZowljHFXrFghLFMab0sQhWa84UCunEfJuYoHyCnwBBKvkZXWnsR/VX6DK9n8yqZQvmvR1HT85C1LX8JwaNMRAgcJhJ18uOqqq0iBvPhZbEpJIyGrSOVR96eceiqmbbalFu5Q8MW3WV3z7FZ77NGf9ff5ZFFX/XhHnc2frXN7d49s3tT9/7zmdc89u1F+1itQRkdyimJUoJyqSpXktJQQXa7jUQAD/NERW5lcjmA3pru6290ialPpJRdfOm/uAvfGqyoqSCNVXtTBcBzQOeecs8rmdqlZx9FDDlniHlpnNlzSZm/lKaecVF7ZY8E02eamYcdUxDqE5PQIORCIw5N2Rjo26LyKb21hCdCuxviwh0+OuUth1GfBGtxj7yvWSiTrbNIkZKbR1s5I22O15vbb7yI6UWbf27fv8B7BQDn7dwX/0he/+L3vfV+npLxZZDyhjzKyQ8Ou02+zQ7RUmjnN+J/+2Xu3d2/16JMP7R2t11z7twIer7zqip5dO0UNDvVj6bl1z+zcuV3COMJfN/r0M2u2bd/i0cZTCP/jA+/bsXO7VD093ZD9ZC1fs59t27qN7AxSXF1hwXTuvIUPPPAjCW1g7R/oVeqjj1npLKq0bruKj4SPDMc+1zHfmmrs69+NQ1GhYLEJOTTNzz5C/qBPoE5CpgR+gQW1iYH4tHLaQBGZ0uuu+5b7FtWtDfD41ll4W0OTDIBUA2047rjj6EH2WUb3gAamCmnw7lUTeFxDW9zgQFwGwWxIfujSQ1SJR6f9aDl8vhwBtVlWIRc+AF37IXdO7hC046w8pg6E07Q0EqyKkjtfP0jdlUUUuBzzQkAJWSBvR+Ott7oqW+JtXCJF1GRBe/Bm+YJFIQTLNxbVUaD3EMpdFbuhsS1MkYaki0jbjA2s0i0MCPOhZZsHl5fe3M1EhkZ4A7HhQ2O48MILzS+hgTh4bmnPIqxGqwpIgO4iqPXiBAKylsDwjzhLKd9XvepVDuek7YB/22233XffffJ2qObcc89RFuIFtzPByxc1wCJoih6RQpagyJP5NNPFIYuD5zTwBCUAaEVcdqo+erbx8e9973vQJEcNJ44PwExJGjUJSIUfYlEoUSScAsG/xwQKIAJ5WjejgkrMaFELecsGFTlRLwcC3VurFr/y5S8bgDqWpN9nADCRuRoMKZjaUp2vec1rjjxyhZ5Wt6OOHnjwh5/4xLVef6gM1vn667/lQJbj1ddc87cqxlsRMtQiZM3JkTRJQZGYaMyQYOhZ7OWINVfVRlEWLJyn7WnHTAsDwCpoqa3tLRe9dFVLaxNLw96ceNLxHV3qpt8jybKmqhOT6DMJ9pfg3yMnLx3ipZe+VGEVR7WR3b/927+tfuoplffwj370lre85ZlnnmNOzK9pFc0nBDR/9NBD/+13fzfJ8iVPuWfzTmqlVYanSKh5gyC5WqE66liH47pqcCVVXpsyHdiS3Bk3POQOI0MRpJSdg0nmEMgKPgo4MZ2n+rJG09zIiyUVAdlWHovZeLZOaQHY/bqEKWsHy+ArqbILuFLTDElxcBh8lmbGz9p3tu6aaz7uZaEbu6S98847DQ8UVhg1hunSSy9lsJgn9YKaKsIhGcoUMzRYrcuLkBHkJES88gWmcXncc1/fseORYatmuzasf8ZVCLNntS877GC3IejdDj1kiTONyC+YP9chV6dvnSfmz57V+dnPfLqcooxjkHnW0XUMzp46r3zowYdIZRPYJRdf9LrX/rYDogvnL0DkU3//SXKw/valL/4riLOjTmw62axLfdGxR99y83e2btmMlPsg7r3nrpeccZpz3zLNo5XOHvzeO9/mhggnA/XFSlG64ziK+dnPfBJx2u5Awac++fFyNrpXR8xYOO/vAHtnR0tzU52zyM89u8b5+v6+HkScOeb/5IlH587pdEZ5/rxZLptQzHPOPvPSl11sqKP4F5x/rgOliuyoKv/kk0445eQTHbX87df+5t9+7CPK6PwqDpVCEST8/n334geTjnw4KVEOe/TfdustRgtOg0B2UNYR6je/6SqYN1z/ra9/7SvfveUmCQEVExo5oOBYqesIJCFquRARZvLyC2RDhuV8qZPHZOg4s4QwneTOKMdcCUQdqQuUV11w3hmnn3rkEYc7oAxNidwVQM5OdOAQq/x77r5T2RVEZfkd9+JjhRGhAK/9rd845uiV2EMNJ8h+7rP/IJWfkqqs0049WZJjjznK4XIIRPcH7/q9tWueIn+idtS48gX89tXAhLjhLW50mPZHQRHatPE513K4ueDK3/l/3U3gvPn5552j1vFEBNilMXyq842vf9VpG+IgL0d+aYMCo+xwOjWC40ey7j4QcLMA6Vz3ja/AX/3UTx04BkkEla3aDPJIx0/NrXvumb/+X39JIUgWDmlmmRX7qJXLTz7puMcefZCS+WC27Ab6e6jgvffcvnCBK0w6KOKdd9xSzvgPaXUuIyGjd//h7+X9C24NoZTSgksroYDz8rffdrP7S+i37HCipMrrUO/7/uxPfvbkTyiKShXVxNCXWqepeWQenF5itVIRgbe8+Y1ajipXeTmipQ2KoJh8zQwRRZORQPUjK4Naj6qZbN1OgAcCgUkCMvKji7/26lc6Kk2MDnz/xZ9/EASOKD40+of5f/7C51UHX3PKo8m4QiczPeTgg77wT/+Yesy++KlBdXfH7bfiQVmomgBVY0Te+Y63J1cYI3/ElYL2G5GjIO3HPvphQFF+cCBr5NqV2qXZH7/2Y6mdpaa8Q+gWYBT40/5m7PtLNxtnD5jrP/uzP/X+zX4zlvySl1325je/2SDGJNEQRKyTfmJPO/104zJ2O322WsekBzVSYOo/+MEP/NEf/dGtt37XpYQ7dnS/+MXHHX30Ua9+9a8RP0GsXevlTd+b3vTGJv1xQ1O+StaP6xQM9Yz51qxZIywXwyOMGYrpO4yW5N7X50aX52yet+Cn15OdTmRwcEBv+J73/LGNAcuWLXPKVic2NupO2ninB+H1r389IvpE4zZZoIlhURhWZP2jd0tez3h/9tBDD5v2wvEKyuZlhbKMcuTKlS9/+cvNwXXEMkUfk1Z23vKWN2WfKwlqWMK2ftAYSaGMTzIjK24WwkD0j/LS52Jy3bp1iHs0slQ6/aCBRxlFzLMu9tBDD0F75StfiQjKOfbQCyOO5urVq83EddNwVNNyaz3NzcLoK6we1gv62Hl9xRWWsWw2sKStb1UoubzhDW9wItKWP10zlvCAoDG3Xl51Z3dMOIiTmH78t1/3OhNlS62yxqfFxFWrVhl4eCMMTXmffvppG2cVBCllQUF5JccSCrbF2FymdsRm/w4Bk9JO6+qp7bQRgEjwSUG1oZKix7Rxhumx4iVPz2/ebKSv2qybQOPUGQlyROkeTNe2m8c1NFC4bvd4+Yqms4WmhzYhbNmyyTluw0fvnK1yWFMvFmQvjlJwpIxyLn9U0aKotEfWj1jxRlLQ8NzW5uruHb74ZhCb/AMSUGtLhwmN8a5dbhLi2WiOj1sSTEkhZdyGDnxaUkoRcwIq/Owzz4Aor1gjquccAlm61BxFcouRFWP7BrY8/zwB0hgJBbDkIdEIUzhY8kYgLHJdvurMV1/JavKZ+LXhPRTKJteKph3QC8pcJ+clagcE2/gkB4/WCpSLPFHDFTjpibIRCBwdWnP77beb95CJhJRMtTqpy1DJ1OsiyCSpe3RqD5OAFEa5qEdylSWSnHYmY5TemNiUlKgllxFpUFO+x0w1xZ/RgiKaCp5Vrs5AFCNncyjSTjtxQA448EA1BIIVSeBnUbELaPyunQ35pmvZzW6aMmwl1LSmsWnjxnVFs8MiSqtsjjPV18WrVGSRkqmS4Ji6IOLRAbMsTMaCWF1NNNZUgYXhC/T390nFlKLMJR1aPtA/rG0hooEhEm1pKF79URoUZJ1sM64KggK+169bZ1Yk7FI4S30qQJtkMk22zVWFUwK0Ch0UODwgzgmnZLKBCQPyIWSrwLBqZsCYHJnihxFK3Xp67Vr5ag9exMtdksxXLpmXYnLgqfHZDhFHUykUM+0WsuQMQvt8c0kNyiVbIyFILqGyQ5BQcjSxKBeKKza59Q7TlF8pQEySDinrnZYaUNbA4CASDayuTlmQkrUSqUeVy44CCiMoKjUqZUUakmQpBPZ1L2RB98UukOlf682ALPOJlfkpCMnZFGA8js/YZqZBhh6LefvjfMZjOjeTjIoCTJNgRv6nwf2/AZqJ//3nM+TDrDA6BjDWYTxSKb0/7fzCF75gJUtebFCqMiM4bfFm5mda9BmBe3blTEGZOYPpGZqSfM/jDDWcZmYPWhWaQYGq+CmB/WwuNHr/FDSawHQu7cq+MTPJbSb4vhReGDJTvtl17Jt2fxWU0Wdc2U/nDY1imT32D1nGT5UphRGkfgMO68sS6/n2zRRkpvIWizltiumB9eaw08bMnMH+Kuj0+DMpaNxI81/r9pf+r0ZBZyzvftbYf7WCmhjIgjo6vvu+971/wYL5ODfMsK5pVKODdpHWVW98Y45gyqBi+jOcM+vP/tXuxExq30QzCdQgal/kF4TsH/7+tvgiCDonl1/Sn57ZmfJlNaZPMAN0poqZAX1ipXqm2F8ePhP/M8Fnokw7Vb3LOGwMWrgw3uN7kUYR4evin3jiJ15VeJ324Y98xByIHbUTeCZS08L3sz3W1VsjnZbQTF3G6Nj+MRRX7kznZmwAM3yZbjoaATOpstJgKL8//jTEZqrIyVthpiaZyZLtr4LORGdqfpPP+0t/pnJN0tvn73jsm/b62rsf407sMZx6eVM3RpQpFbaIltvqYjbcOP0cYCY+3XC1T5YvBJixi59ZgaY36TNmMoOCzlQAew5mJDVdRH3dxEGl6SKnhU0/5JipImdS0Jnwp83yBYD7q6Az1ctM/MwEn4mlpsbYiad2TIzoHxvpMef4CRdVBTA/OhbH6PZ1M9Xv/irojJOkfbNMyEwZz4RfDNs0kTPRcShrGuwXAI1NP+mZMYVNPNO5mfgptnmaBDPhq7xpsGeeNMykcNMSAZwp35ngM/EzE31KqfNELbt11hQFEJoKIkAp8QxehaclNSM/0d3th7PitZ9ddv30Jn3GPGeYNZP0DEn2k/7YflrQGZa9ZuZnJos+Pf9WOactlw2D08L//weMKXw2G4bTCquVYMNNEBpJWUNfrXDnQYlY8d3PHrV8V+iXL3XTJ/7hQzKm71pGNhfhbCKoZDjZhcAlRBKxCUk0JUnuK6BHUdnOBDwmHT4i1i+SFF9sOjg6kAxnXkkETpV1Qjwmnf6+ICsXboJUbJtwuWTJcW9zBmGkMbb0V0Qy7FFy+eZjhQDe5XvNPkJX537QxnF3KNOzBtc4j/k8nWvOycz1yha4G2JPamP9eEP/iDEx3tAatYPWO8LY9Fg/3t4atsclnTyfdpeFtwrh+f73ZL6yC5yymC8m4VMDo3uOnAdnSBdMzUI4Hjkbuoq0SWZsOOYY4FUuApyowJx8a5DCnEi+TycQdGuqKSmkb8uXKHQqhAwk/VpM4YyCvC9cqcVavS8CiHcNbDYiTQ5li5BGHAXlPIpI1CSUlZeFST3ge0yXGXttkI8p38xPFIL8SdwJztDPjOTFJQU4Amq3opNRGZs8FISJ4xMJHx+LV3Ni0QTh2zgsEAzEu8kom2e6QG1hqF+xsigxe8ISJjBj+ela6mz5s9XWHtsmS4SjUtC7Bje8+noiVWM/GkLhRut9PwzV4fHYyu3eXUricnL6TLTenHnXOyHS8olHd8Oig1BTW7BducTBSVVefIqNchXnEtr4CHcUNxgmLK+lQuZxYCHzZN8AAEAASURBVH0CJ67TLQUsSHvGiIBIpZ816FHapM+vZCI8xSUbFUIV0OYyCn4FFOCmUPiFj5mEBHDIZXHiU8kZgVFxXC2hKHkpFWxhmPRa2TJJRsEXZYUMUCAfk0iFICAqY6F5TJVCVjipZULVWdERlbH8rDCBinLERTEm3jxJxSVBF5jIyyNk/6VJPypn72WjxAEOzMLYRKoSBmwaH7SL3dDex9tHfdfQbvB4d6K3GaKCE1aISEfGbaiXqW/iiSWnUM3o/ggwJN7UPNGkyxBJY5pQUB9NzhzllfLhw8/yJlf8lDk4R0EZ7eSWRVZlWVhl5zJJkoqEdbEzNRN6FEg/MdMH4TLJlOJn2kSoRU5IYTXqiwNJX0AuFSRz9JiBTFg9VgFw9PMxlRA+F69fk2KKoBIEbNmn/RPwSDUzJb+WV0STCDrCoriknhDALDxg5pVoCZcWPJPwwyBNPiZ+EtEAPBaJ7FHoYK+8agtMIghdtFFc5Y3bwRy6EYopnzxY64kO7SU+xJN+cpi5VEApEQjPfhZdfFgmZY/eJo4q4JV6sZrx+rt+VHev/48PeEuDIzUXLBRqwUOpxNDLiA85RdNibKFGdDmemqIOeFE1QE44H8PXQMJ+lqTjoyU3CB751FoAwTLGiCD6BVTCSGUxy1OUi8twYJZaSwWogBmQcAokH0u9TxBJPhMzydYSr6XzAnA8pAT4HMyQUUU0daWCgHOpjpkg00vmsTbLLBU4hKi8Iu6kg0Jipi+Kq9JmjnwEExMFsYh4TMyE85OHKm3C651+kTYOhgQ+FWRQEMm0qJC8SL79SKGtk8teiYCaQBIXMDLGDGAyBl6YUF5m0JauumArOpkoVOlIYpeTAB7Y2eCtPvb1RIebOloEFYkmOAkM7AXDOc8oepL5BJFSfJwIB7XixHrEUrAahtkIWF5gpewx1Jj4mLCr5OFIFKTK0Ro7kzyikC7DlV8JSmzmCJIBOJVLmtXj3oEgvCfTouUgcGaiMxNckqz3JJgVEQqKempVgqakr0SWTGSRKoYkSYhU6GbaDFRhyImT9LN4wsmHWAGPSYFq5WPCIXtMHwI+k58KYtpDPkUBVAwElkANNjU0s8TqCrD8heNfPIZDM8nW5gKCeOWL4gbLhIadRmU0vugQXMCJIY23soE/5uszIPazl+ShWHr6yMX5k1I6zSaeiwMZjQoNFXY2UTUqlBgEamWSCVMmgT3JmLEG5GgwoyPxNWah3C5jlDGS2jnR1O0KC6taH5NXFDL3lB5I1Tcm/JfxpZqChiwHziWH+IEDOAXzFz6iIK2ElY8gFzviMi5VBGhfMQUXk1nm8phHqfip8pLgQNrkQ1QGEM+xKUw4fDhiuYqPJOVREuHkJ8kmJl8qLpNPyYhJ8QtP6cIwmZs0UKpCbEItoqMP/lVonSlMZsRP+gLp0K8CmZ1HH2mnR/iS2vmBYrRK2fFr1u6j7dStTr8x0tgQKjEyHhsuZSwNNQ4KYWQVn2Y1Kkl8fKNe44wBSSio6VQGI/OwoAqKtdCuIkU+iB/5Bf3SxhQ1zHZwZct3lAOK9AQQVEpNR6uIjy9MdCZwCkuBLJCT1Hgo2YBwVTio7O3QAYAArTZQTZKqqCpdYlaPVWBauOSZBYXJFqsUXJwV9iezNMqsskGFkwZEbD5WgURLokk3KSQCOJr8DFQIYjOvilpGQQMp6EXAJbtaglJBINPEFK4Ysz0PsGzS4xfZ5cu0WBWacBQgpvTFSetvJJmsyAynX8VWj/V1DrSgO9pkLNFgaBtqIbWjIZQXLawZbZYSpOrHUBWd8CcMZ1DNsMTRnEKZgidDVWxkWSDVuooBZYeQDgLWhWVAF6l2nU4dFaw1pxykSzboZZnqT9agVOIyiyKJibrIjFImYqvHxKn8Cr5vYEraYHJSwlXyXxhIxUjjCNljZhQbS7GOKBDrKEyFK2sHlTqL4jIBPwNVUVPfpUomKuRMkptYk46E6CPOz1SFWFRS0qzQMlPAzIWvISV9Udn65VgyjQYHIjYpJ00rjpLYlYMNaD5ta6UWWukVo0ECYiPbJ/ogMJFKGQlk1i2N80fHBtW5L9NZZRqKr3C30Uj0WbQ4MUt0Dcyjabz2Q1vLakYqbNhUQ9PoZ1wCRBRRTGVylr4s5yn5yFichUiB8DOMzxRIJC89T5Y9gNEq4jWP40A+neRznYNjgz5a7w2QeSF+grrcae1I9P0jMQTa45IOHygrSDFlUZtLhfPLBFJ0fASTbKZK+L4UZJSYFXKVUKlVRxWbaWN7c4ZEpCNNEKLxmGVIeEUx8af4cBKSAfwli6ghxVGIKioJ5mP6mVaSfFQMSYSTAX5Sg5YBfjqfzooJ1Vgptn5Tb6+9aVdDvkzVrApdtDc00O88Jh0VnhywxSAMb2o6iecahbSZhayFseGgYnNLvXOAba1Ng0N9jrWNDg60tnQ2tXXEIafhusFCxqS+qTmUnjEzZwnTGpdfuA+2XS5eyZjkN/hnOcriKBM6FuvbhK/51CoHZrKeMJYFTDZSIMIGHK5EcDezVmlbvBbns1QOTytXvCiIpsem+BNmFj9Z5SnV9JHNMgokJNL833IKKKsq3ypQcTUF4lBKnKmQjCNKPgwOnC9Kyqw2j8IQsiz5mDhJIeGVn1miCYJCRcSjVOkqapATX2yG+RmLAuRMnqQ8CqTrmh3HsiyYcwWzXO8Be1QR4l6N/v5eXx5xq0scyLPO3jDxxgtK1QwEwgKVJiSMvtwhBI4lG0arYXywzw0rPpPHmjbUjw61tswaHW7SHOJdRFwo4jt0cqwf7BsO7UTCgGC0YSSOk1iAsp6nxzUI1RXQPPZtuLm+sbWhyZy/DC414OjH5CipsJ8Ap5hFoybqtTkO/pPMcEtTIzOsTGyOs6e6BtZV8hi+RgsNuyhlSm9SWiHbKlwo7/VYGzUlnJxMARYK+8KmgVTJqxqHVAELHawFM1GEGjfRL4PA5qSHwQ/0kiDhgBmYKQOCqJJU9EGSVG0qwFomkM2MMrmofERkWkiyWjHM0iTbasuJbnNpB23VriO8WpwpScOYbnm3A+2jg3XuuesbioUzSVIRhZGKuiytVACryRJfbHuzQ3PDo+P94yO9FplaGg36rBA0DQ+4aaOtsb5VL+5TkkMOPQ/01w2MtzV0RemoYLFPDKopffnaH032UWQH2H13xiDBx+x2jNS1jjXN0bSypHjIsss6HWaSn4wS6zCfcYTK6N1t0NLW0NxqZNbS2DxsjSHSqGM6PSlDc6fiImJvhyxAwhH/r3OyqIgLy7d6nBJIBhOYvAnHK0dOTYQkJhOLrhLXJquA+wbkndnzxaafaMKyoEmcQEYlWXXJ1WZBRTzCSX6EQdKcC0/JF07/YJ8jX94Ix+jPGfA2Nx85vNvsqoSG8YGRwZ6Gur4rL//NDmPRoZ3jwz16VdS4KiNlx1XykJlWuWCjfrSvXFnSPdS35XW//YoDFnU0NvSNje7qaPXt5MjLj560tLoyaby+vLE3PXPTidWfMGfjrsqub24aM4dBp6l+aMHcllNPPPKs04+ZN7t+bCi2qTsQhyv2mygy6yKVCQ+fopLnsPHDPQrS1jT8ht9+tS939u3aquWMjhh71OkwfChUOwhzHu9fx4ZG94wTKpoIclneFOm+gq0k8KsKkCSHWgbSryVewROYFcGPSVKCMn3yrTCqreK7whaYQnTKYyapshdAqvYRPiKAAolcm0QU0SV+5pVihSPA58RKm75As8sjAmjaEpNjM4S68SFvyH3+YGRowGeMzz/vrJeuOvP5TWv+477v9fZuH2+MT9GhprI5VYWUR+VNsgKZlwDMocEdPq3d1DJw4vFHvOT0Y8dGur/21eucyx/rsDtkZMw6aT3VG21pa2pr6WxsGu/rjtGfZXhWU7PRrTOWLoZoYkhHXAM23NXefuEFp1Omhx64o6EuPkaTosaJfDGjOBhLZjAgVh3xOY/9A1u9nT7lxJMvPP8lWzY/d8/3ftDjSrO2TtN4GwZ07BpxiClWWcdM49BJUhnIcOZS0PZq8wmvMGsDkGsfa8J7DGQNcCKIoBA/A8K1dGrDiQOSmEoqwJ/o4itUgayeKrNMwCcgvtiMquD5WFGo4FWgIpgQuWarkCRTgXNJv3RSQTJjE8EjIkkHJiA/cfqH+2K9OmYe6oMwbDIaZju62ppG6wZ37+x5yWnHb1q/5uQTjn74/nuGG0e6yyVYqQHoZK1nLhjzKBd9LpMmTHvmtDYP+gBy98arrrps+7Z1Jxx/xB23tQ0P7tzds1XHXt80m7X0wt7MxDgQJy7rjBopLdkA1lJkXFsy2OvCt7HR/s6O5hXLFy+a397X23P0UYc++eRPNvTsUBYGMk6ljUc78dhoUbXIWZgzhnWLH4fDzgYK3XjSicds3rD21JOO/c6N17t2wEYUrNaZQRkI8yPNaOneWcqJ+gqZFidSebPIk7Bf/Hcm/FIPvzg5DPnyKzoZqB4VTTgfExOEC/OT0MwkH/mpDQkktYRnygTu61d0KjSBDKPGCcuSn3qAQiZJOGBhKXQxzYkAnIzNXimJSAUz01oDZDhaGppaW0yZm9sspo8P1ZvoDvYumNd59MrlSxYvYvBamutf8fJLvvnNbx666NBMi5Qqd0jc9Ci5kpdc6K7bO/S86Du42Nyzratz0THzFx9z9JFNLdaadp9/3tm33HLvgS2L6xt9pnuOyysHRwd2D/X0D+8YGhloqZuDGrPJern3q6utdXRkcKDXRR2D9e0uA2w85+wz7H0yJFh1/lkb1q0+bNFhCijTMqOPqyKoGghpYCYHJFn2FMXIpnUve9lLXTrhWp7OWQsvvuj8n615pmd3/7wFCxua2xoaW13+tyNuYjNQZfQmDAqClbSViwOpAqJ+oYM8Lc4vo6CZeyav6GTuFbDiB2Qv/Nf/ziWqisuU4rK2iEz9sSUSJC1RRYcmjFmVJClWtT6FDuEmE0lWLMoJBEmXzCXBigGY0LiKYMW6VFV4YGx3i9GfbXSjY/Nmt15y3kvOPuvY8aGtLXU7Gka760e2N44PNIz2x25N3W1T0/bOi/oHvMzp/NEja26/+8E1T28bbfDtylbfVldrdeM+2zw0f17LyScedc55p6xYfmjH6m8p8PhgXLM4GnuWqMGs8aaO8ebO/pH28Y4Dfrp6y42337/heZdC1TU2tfYN9lv30b3WDQ/Mam8+/qjDLj73xKNXHNjXvbqtvhc/9UM9Y8N9cQmlu8saGnfPPnZwyAuuBb0DrV/82i0/X7ujd9gcrJNqNXptNNqzZEHH2acffeYpK8dHemZ1NrftfMBtQUy9YYr7ukcb2kbr54z4Nc5tbF18130/vunW7z+/Y6ChucPQWI/fOh7NL5if7ABTqvmYfiVqVZCVta/PxgNWdaGa0klSEcka9JjZZUaSwMQDHwVjq32JT0sZUJL6N1x1KRJJOkFJS1xtqapsqI1wrcv8sJXZZFSSyiiQDKAsLC+sV7kIVGiZNssjXJUQQoWW5cwk/NHG4fIKMF4z1I0O7Ny27sQXLT/tpMMvOPtFQ30b2xpdJNLdNDbixXl7a5vF0a2zL3r0sSdu+u69a555vqX9gIHhlvrm2WPeSRJc/YgLa13G3dhgpXPEQviRK5a95bT4boTBJV8LjYtW3dJT3yrVui27v33LfT9+cmPL7IN39TeMNrZbeYwF2RFblR2VGjOjaWscaqnrO/TA9ne8+TcHdm2c1TrUMNo3Nrjbyio9VuNPNx68c+fAffc//sjjz9U1L+wdbh+p6xivi2tjm6ymDO5qbxyoG962cG7TpReedeaZpzRtuNMc3S67YWtrGGudPTja2jV/2S13PvTDB5968LFnOucsGW9ob26fVd/sIp3G5tG4RkXtpABT/ulX8NCDSR1QzEq2VUBsil0gXRUlICpjM0BtUIZWVV/CJ5NM3wbgJ6lagsKNJ568MrVzSsbZLGCAQ6jBmbC1otJlQkwI1LJSC08ifEnSr+hPkom/kisVP9Ey00So8JNs4ahwFexYGGRpm90cOqujc1fPTvcYuunv2GOOVZcNsTWzbYyNbOxsapl9zedv/e537x70wrzRl67bmppdduz2OvcCx1A2lip1/cNDvbt2u4que1vP9jUPLz748Lb2OXUNHWN1rSMOPda1Do82/9O/fuOGG+/Y0j1Y3zJneJQqMMPe8yNom5/lV2/ZG72NHOztH+w3xOj5ype/evCSA90519Lc0djkc64u1em0yv4fT2y+7vqbN23uHRxuHq3rbGuf29DUFdOd2IRllTV2hbgn2TqvS7+eenLNmSetGNQoxxuH670YaB8eb93ZN/a5f/rav990z/BYa9esA5pa3DDaotH7MVx2o6QwQ0yTE00QYX6KOhFqYxOSftaXcLqs4sqvqkOAXuZqdNUYqrTwE1P9T1Ka8W9ipr+nuSSv6UsalT85qZwkHRTjZWGNE5X5pGLhrIrMVFVj8ph8w08cgVrKCawgVSDxKSgIlhJehaMirTXGbox6l/752mZvz1B75wENrfOHxjrmtC+c09U8NjJk/DvYNxB7RRvndLQtGhhxncuspvq2uoYuW9TBzZcsao6MuLtpmK1t7XQf9q6d3QNP7d51UdOC8ZZFfaNDs2Z1tLuTbEdPe/vso4457b4Hnm3pamxt7nSB38golWKmKbcNe5TUoRAr6C3tnfPHh5r6+7aO1bctXX5SfWtznzvS6sa9l+ofGOyY1TlvwZxt2+5obG1qbOkab2wbHrGkykwbAjSMjhgTt9BRXXTfoFXPRf2DrY0dS7xUaGppnNXeZhfFeEPb1t7nx+rbO7oWuKKmuaW9ubFjCBsYiPlZbERMOWdtpjD57FyGU5JR2UW25Jz4/HQQBLJ+1RFu0gF6ZHE5aQGravWYdOAIgyeRQvAXKyi0Cr/xhJOOzJySepVH8p1WOlnnFy4nEkOodbjMRxSSYOVLmA6kcsl0LQVRHmWRAUkqggpfCa5KMkFK2c2irTKZtPcP9ZePSVKRV7/q105ddcnA7oFv//ut377h9u5do11zDpl/wIrOeUc/8uhPh0YdNbK62NXU2qESjSxjWaY4F0sL2edrQj8+0vDqSy8654JLm9vnPfDwT2+57Z77H3y8c+7ipUcc5wbpBx9+sqF5znh9B9vse7neXLq6fmh0ICiYnDDH9c3ClhjMzE879YyXv/LXRsaafr5m/bf//baHHl296MDDFxx1StfsJT9/akPP7pHRujZ6NmrNtLUz3lzFqMYwNW5g9RKqqaFFA3zTm9666KAFGzbvuvsHD33927c8t6n7xNPP6x9qXHDAskcfXzs63jY00jg4Ei8J5EtoVEinlhJTOCKdENqkacioFHX6WS/QRFV+phKVFaEuKnUsuYSmZhS/NruqElEQ5op2JL29fBQ8Z6a1EdHF5zOMRKr8pCg2IZPJJnp8sclKEqVYGaj1JfGYdAT2pjNJb5KtxKwU1COXuUuYJd+TZjIU7Vb9Z11EgsbW9naXC7zy1b92y43f+di1f79pS8+2HYPPbdjx8I/X7h5oXHnU6evXb+7ZHYs1Nsr5EYylmJEweOMUqrXJKyh3Zw401DXP7przO1e9/pHHfv53n/n87Xd9v6dvZEt3/89Wr3viiTXnnHfJUNG2UTvCYn3JPN1+J9MwU0uTG9oa2unmPGbMdYe/c9VVq59+5l/+9cv3fP/+bTsHt2wfePCRpx74j8dWHHPy7LkHPvzYk17N4r2+scVouHxrgehi0Btv9t3Y2NK+cuXRZ7zk3G9e9/Ubb7n76Q1bN23b9ezGrV+97sbZ85acec7F//GDx5w2GbLIFsodRiv2lcSqcFQ8l8Kswmn2VE1GBX5xIfTiErPy8yVLpZpQsloTwk+HRiZBX1VyMEESmU+2Fc1fJlB/5VteCW+Cu/Ink1GIzMCjvMUk39UsrDaJcBY1SfElB8wAOmIlzySAwrqY6jEhiaNgmTCzE5Y7alyipS92IhUFG2N2fMOqSd/IWrn52aud5zc/2zWrbXdvN0zTEatJjQ3NXT4D0bHINeY9u3Z7MzhonxKb4yVPfaOLwWXR6MbMhngvPzI07Lp873i6OvvdoeUDC1aOmr0f8gJej1bvpWn77LkLtmzrGXKus655rN4wlGbX1bf4woGNTfYltcSqwCiFGWtpGps9u+X5Tc+hMDDYZ6Art5ZGnyhv7e3vP/Sww57fsg0zLZ2zGll0vfN4kU+8JrUjxEui2HY6f+4cFx3u6l0/OOSaSCdKh33oI0TZaEQya9bshYMDY739+nanSghtOOQSYgrJk1WKqwqolJRz1kKGxYLzEzglQD4VmspKl1OiqnbklTgZm1nXQko1Jhd7+RVlgQyn33jyqUfvhVi0J+lWeZQkoaAgvAwIZ8IklCyCFJzEDF8sfE5UYiZOKr1UmTBx+CkgOAnJVBWOx3TJEt9ZoMKJXGJbL122U9NtylZirHDGVW2tnfVNHWYeRoGDI6Yb47t7+8p3T8hbcx7VIY/VUVOzZhS874kvpjXZCYXh0fHe4QEb61jGlvauRvN3qlLfzETaRN/vS2BWk2SpE0alPlayfNnZFyKMaTWrGBm7Pp1Wj49t3dY9Yh/SuBbQ4Zs59Y0dhsHD47FhfJDtjk15ca+sF7bDRsyN9UPDfU2xx3mcxwa7+Flbovq9hgLtc+taOsatIDW0sLuORA+PNgwO1ffaZxAnkGycMqfy8RMmRh3t6dZT/inYFClIPuK7siMgtUIWhmazYhpdsZDTCVupTWRhFJIICPpScQJpYvJRPok/rZ846UMQCAFlqBaaiTMPYdxgTgATIfcSSPzKvOFMLJdpxQokcX4OyRMBNRAJJcnVtXyskKFl+ZNUhnNFNiG1/phTl+Oux/X5PXzaNRGjSZbHp2h8UWg8hqb2MDWxDI5CMiduHI/d53DYN5vkYz/+MJkNWxuyfdPl13RFshFGaHzI29Iwq9SvadB4gDU0R44lTFrnDUEIwgFLBTVfZ32clNerqxSdrCS6W00lBGVFU07e8cRJZLpI+21bjow63MyvhFTJmyjTPXOXOB0w2NaqbcthfNhuVPnSeHWtyhq7tBgHP1REHOnUrGwxHKfQePF9Tk1UdvkymY5oFDE3xYJMUsNS2in5UoKosqw4fiKDSJIVkbFZOyBqLWPTz3pBOUpZUvHllXSmEEkcCLUus0ZWkipVIgTPopNKgjLMF5eQTF/BPSaEDwhHIJGrDGoTyjjTJhBmpC9p4SewQkg6YhP+y/i2AscZ3zr6Y8xFpjbSS05bcaZmyUyITKMfj6lxOY0ZkozzEhjwK6XQUdd7X6pIGKY9Mi/njQMSRD3psWPNld5FqoIRI1g9KfWStQDc0KqY2KQAY5hbkGVOqwMYGh7Zl8Git1CUguGOk8TOddDjoGkuHfOb1JFQFp0+Y4/F0ZZIh6hRr6yj1UnnxaxU2BbGcKgsNIKMW0CKI9t0lcwFsiLSr4VnuPIj/8m6zlSScMJpeipMwMRMPQYvSfG2J1CLXIUr+pisiAhM2MUKLwOZ9xTgJIlIX6EJo5hMkEMa1IzNnEAylg8ZTiZJIkkzfUVKUpJnbMVAIlSPtQFdbigZwsxGMWpCpFSqWVS40EU45QBumRQlMHWsYMQSUVRs1HxIMhSNIikncxcMB0dqOwpuiZNOKEhZ8KQdpK9/hhMKYTkhUEuNUEj40U9T+ZiOx5Y4QoqMi86HHTZqBoldEfE+wIBADpw0Jb+i1ZQbHosa4Og1BWlnTIiiZ+ZaWXtD25KzwT36wb+/aKImkaiskWCvVMREyYvqgHiEVgETUj1WUZF40lWxAmCJkwE8AYKkixLt40RVsAwn4QQmZM86KKjoKsGUQEULinCFWVEEVH4uxZHJxeZkqJJR4kPWKQgLcBVyQmqzrkWohVfhMCG0w6o4+Ud7wwKCqoREUhuL4oUHEkYy4rmAlIoMnZCMF2oXqlB0zEAQuLkc4Yjhabiq4EhED8+aRjJCiYkN0ewlw4BHHqHtlorQtqW+NAKKimCwraslIv21koZuGxUI0cTIMQQdKhsyhyCpZNHgS9XbnhI9RLxymnCWt2BGisJI5C8m+I4ke3gDTI1JuKgqUHgGmECeAhdbIYiqHhPNYxWo0IJWqRW+2ERIv8RMZJ2k+LUMQ9tLQSVIuiKqDJLWlMcqA/CMyiSlNUefrvycQI5OMpz0M0nVwkSxsslrhVBxkvAquwqtCpi+RC3E50bD7BW7EYpYjJe/bI4OnjER8LOUXmpVetOPqOho6OGkCLtF5SQPtUuY0SaiajxUAaWoaDIddYYzAtESiu7QvDIIoA4xEKBXdTF0LEqVaUtjCCqFRFHcqNAYBNPLwPfz4JMnBruhviqBDk/gYzu029m9kqOLd2hxbDKMqVycKQ02YzgTQwMcRsUXQ1laW1EdEK6UCx9TXUYlNOs0wwnP+qogtciA8CvKGVCtFalaajmHEZVoFZ0o8GRXXBEEnBjbJq0Ku0qfqBk7iRN/azETTslS9xHlJpFDHFNYyahEq6JSg/mZtqKQyC/gmy/oLmlnTA+rk5z0UIHpj0qKuioNgC7GrMJzsKf68ww9vfJo3BZGEWbERz1HpuiUbbGhdaEpMU0KZNOcPBITE3Q6SmlCPdRJc14HGXmVMWA0ALHRrYe2kniQQsQfz5ODIinl4dVA9ANGok3WBkIIpVFlR5AyaXT7E9LmSJFbcBun4+KAaFy5UwpS2gz9wHq8I5ioF1GpNOikKwXJ1iI44dSI2MmnPcAKUlVZQiAnpDYgqqrKiuAk2b1UoqJWa9RTkZL+xCy+yiwD/CrX2qjCxARDFWkI4CUKLIpXpdXscgya+pexBTe6/lo+MkcILG7mWPmAXPU4JWBeXfKPZchiQPTVYSntuAtrWhkgBKJGVWUOGr1upCW6XI9F50KH1YyFGYqowvXz+lBdfCiU4YtKo1i64CDiZ9wYUioaEtmU9hDKGMOMYshdeReUgxmT69jcTrGiz41GFdQiMTMIFuZy3PY/dK1VGf3EOdHQ+zh1T9diWI2vkNjIYHCCD4PvouLBqY4iMMPYklWs6CosFZdVjlcmK0Vsih1PwVtxKd6suCo8GRkEuaqXq02YOJlQuAoIyyVjM7mwQMIFajETLRHkIgr7FZN7aUOSyARJJZMltrDEUb+lB8+oAomi0sWEo+5REpCMTTifSwjf8gQcThh+rdsXUhs7Jaw6I7eYyFtGD+tjJRGneW9ldo+lKy/pip7H+C7qOLUtJaAmHUYDVtnGcJJTFiX1vrIIOgwni0RJYrApjS0JKBaZQ45SFPUo1SAmyiBJttWYU8dYIIbJoY7UNOxxMYwkUJJboYr1S5utbba3fOvj2HC8HoiEQY0FLnk4xipJ3NAUyw0yMjcseVs3I/rIkQsVNcSItPGI1cAplIJKcflY+RVOkdRULxV0MqmChFg81hKvTQMhowCDmcID30wwU6UPwlXhRM60fG6PBYWUWWY2HisIEql2AtIEyeJgFhFFwtT9MuKMpQfx3sfY8Dtn1lLfDXe00rsNnTClNa2E4PuIUiGblJOZyLreMd+24aExn+Aq4544qmsn8ogzb96rsENOeTT7pG4c+bXbrHvTVrep2+ShNoeG44hw9KdRVyOYiU/XWaDBXOiT1jlmB2XkEvpnxTFqLt2oWAY1BsRlqW+8wecBrEJvdz9DmKJ4EerskLD1TfC2PfdChAkTWxpmVBodHB+3GmrBNUxj0MeTlVDLn4UZgsWZEaTvjGUzxgd9ioya6imgFfZxC6cxrgxaTAr6LD1yvtmWA/aoHrtTMOydkrXSojExQmF+DLfihZKNWg5L2SRAj836tdtIGn2G9hyTNs2yqFoo8lBdg5dkDbt74jSB7HKB00iGC8bK6YOsZTyjLAzHkRvI4bQlLvqSKLKGVeQga7tewbINq5iUf+By0iEukDqTGSFeqj5i97KgJcleXmYGJBA8TOfAI/+iuDjOPJybWbp0qe/orH8uPv/oXJtjYT5v70Nn3i9S3GeeWZupJEQVkWRRFajaWB5pbPR1OTvNbXug380tDfkhbiuGdlSpHGc4O+fOvvzyy32l7/ktG5T0Jz95nJ2JN5HWrFtbMWPDvPkTibe20q2pi3a1pZGdeY8k3mEuXry4o71LsRzD2LwpzrW1tEZ9tLQ0+mrb1m2b8B8dbxGLvNR5CRfRl+rxqFwRNdkIvTUFycoA5FKeyEbWxQFC4LCRx6CjiZW9QpA5cFxKAielB5iKMgGJ4WsTtqtUpErAXj4UCxyKrhqNTyAED/Q0GkOMoaORlE/vJZ9ZL3JJfqCBcAhG8cs7JIUaGIxv1XnkBLjIoEDK0wt5MBHnc5kk5ZB5ieJmVNAqv9ocAJNWAmvDyTqKgMFsueSc7E499TSfQI0vX9eNrV7785tu+s6WrRsTuSozail0PkUc6B9sb2/yJVZfeqUu37np+iee+DELCp9Y+31McWiI/voK6qtf9Zr/c81H165d+64/eMeRRx7x4IP3v/9//ElXVwftxAAlCYM9+TICa7VlmRKWtWGbLGhzlmLBgoU+wLpg/mKPvX07fb1948Z1SmfbBL3HZylleNpAhtFk0TIMkwPhKy/KwrVRkdBe1HLfCZpySTR+FLN8yDApeAQUzuSpCSV10KRgkTbWDYJ/Bk8R4ANyZQE/WJzUn+AnFKN0/8nRpB+lcH60REcD4NBMOknQI1fBKSv77TEaY5GHtGlJYQXqHpF7jB8XOEUsAhz6fBBZcIrAT7ioGRU0qBeX3CSJhAhXrhaCXORf2lx3d7dPO7I3G9b1/vqv/zq03b27fQf3q1/96kFLFlV8CMAXK1AsqZbtOYzEW9/61nm297a0HHvs0b/+mlfOm7/not2+vgHfBj/uuBNk8fDDD61Zs4atZXKcJRLlUyk4cbiH/XbvMyvhtHOcKZuQQBFcii+DpQxkbbccR/niYFD/0IYNm3w5eNH8JSbCCN58882Dg87Ck3nsFzGESM6jvDGJUpBSc0abE539HtUUy/hVQqsSogMIv4rKgOLjnC+KDx8FvAlQ3BRX5FsyEkVKYQwLmkdEpOLDLPXtb0yWMptQJd25PQQxvw89NsCOxKFiezoZzyhkFtlCkh9hjQpFAvedmklbkHo/0YTEvoBLyhCysJlLli6ByXzy/0IKGjwX7ZEsA5km807qGU66GctXKn5UeRnNSBvz4nI7s05BbBYyaaZfpBy1pYsfjuFjyz//8z//3u+9yzejHnvsEd9UNWlNmjCXLVvma9hmC727+2wOWnzQAZ/61KcWLz5w3bpnjUeLuYl1GI17aNCdCLGSwJeRtLWNumJeAGMsKEx6LCNOpjJiEMhOp6k+OHpjCIzO0HAoSmIqsUCKSg7wRckOMHDMpwOP4sYfacsoNRihKP19MQx14C9riHCoMtF1tNsH7Skp21IQK1KoKVSoS5mqwoSAoLT4FOBkASh3aKype0ziJcJ46CN4YSx4kyZ1GiRdsAg1ppBcsB8GcjQ2tTU72B+724K8iGyARGonTo5QM1N0MmUkLmWfpP2L/0oCSSqZZloQ1GZU0EwwhXCVcZKoZQKtzAAQmlIlJL5vTiDR28Y3rqksp1SlCBOqLGHhJnRocCi6coqyevXaK664wlhWdZBOa4tuLhRIfZu/hj61Nm/f1g1ZPe3a1dPbuzu7I5kj2NYWnwTfuXOXR61cpvjBW0yhwhXjWWNBI9NgJ6of0vDQ6I4dWzZu3Lji8DlFa+uoJvZpqn1GKCdaJAgdCskSL/6RFkZNeDI24PBLvqG+HlWDWBBmni8sF74orCqdxgAzU4lKfJQxQGFIr5jI8IWjInNMWbQzKWd2JFH6WbVD6WKlhSYrfnOLW/smjD38SFKUqihccIQktVeZkSQarS9GGzgZB4/qY8Db2lpmzZrT378LQsGJDAUUYTLr6f9WyDDTpaoktthgpth+sTMqKAzRlZ8Z89OBC1S+QBKdVIKIkpyIfZO0r79P3Yoi9LLvNW5Np3aTxAKZS4Jq2XZOUoCpswaHTNscilMJg4NxAZP6a6pvoiYMreK0tcW1dWraVGxW1xwzfQqk5lgWe8SQFWvjL9FHLc4gPThUn8OnLPgHH3zwokWLwtqVD1YDanTQ8CPcErOukgDnkU7CPZoX1V2aILgATaOQNBBZ6yx8kZkOpkfOnLhM/iATTothCSyxcmFS+fjnCqYqxEYYMPM2qiNWEqWGKTthvstcyhwxBohNzmVFHxLzcQJhTaMi3B9QDLBVWmK0RVB9ITw4YGOfi9Zam22kbfTKwDJMzPaLH/ptYlC2D4z1Org1GN8wl5ATyFLLRLhIZaJyiSF/E7VcxJLCgRkslcliUkgifC6pFFL772VmfE5RERDgy6xyvjOO9QRS0MyVH0pTjEBmK6FHQDqUUiNB29r7+vp37+4DSSICMKGFPhRLWdr0sJpTOXRUVCz6jNXv8NXyxhafd58/byHtNCb2Mb+SMDOc6mcR0E+ucCiwYcMGCkpqmbuxaU5go1uocYlQoeENnaKRMbTQCSxadKCE1q24Iie5h+TFykIB2fgFCxY5CMA6Gtht3ryl6GLcLNnvcp++2ATNdhqU9/UO9ezcvaN79+5dAcQYIgJwDMRxqIyoadvwsdRv7yv9jpdmLCL9ZpKwV0e3enslwtUA0Sl+8j80OIIBWcRg13tX9Osa0aTkMGWKbSJVraSqh0FCtSoveRRmwuJ6nCrf6Z4lkWn6kiQPFIBLgYNMWFBIKMDmZzh9GFzmJyCWyCTmIOPbY5WBQGaTCBmrSjJhR1sHOLl7hGnlh1wQSTqlkLH2YfHozDPPnjd30UEHHbx8+eE+C7lly/Nf+MLnm1vGD1x8wKpVq4wW5s45YHff7q6OeZYjX/vbv0nWFNGlDFSQvEzFLr74ZZdeeol8tm3bwlKsXLkS/Lrrvv7kk08uWDgHA65UwHlsDZ5Uvij+ZI+RHGKSIoIrFFZLG4iSiVU0jgYgCxIDxEnNZl3w07Nrx8knnfqKV17G6gjTj9e97nWbNj5//wM/uOfu7x1w4MKWZsuTjVu3bH/Ri477b7/7NjjPPP3ciiOWb9ywefWanx+xYuUt371pw/pNFoAGB4Z/67W/ceZLzn7k0R8dsGixQvlmppU12nz99d/auWOrVTAKuvn5zS960YvOPvtMqkNByY0wDznkkDVPbb7jjtseePAHxx1/DHO4edPzF1988dKlhzuDsGzZcoUybe3u3nra6SfNmdthM+wxK093guCZZ5658cYbzUGVy1icWtNmDeDd7363oQ4568TU7HPPPbfk4IXf+973Hn/8cWJJ4RCprEmMANOJSsUQEAUBPAUbYi+PBjYJEQWNyyR7uvhMA2lfl2kSjonEREIA9YrWvglFQeAySkCSTKWCBWhJRqlsFQafPv3kJz/p73v0nHPOP++885cvW/HoYz+CrLvRpr///e9v27rjmKNPePllrzEvpdbXXXcdw8lyaNDQurpm//mf/yU9vummm6677pvsDEdBP/CBD1zzf6655957/uIvP6BqkaqaikzxgBltWIBkQDjUsKdLAxROFSRfwFKmmJiXVKx71IfWB6evN+4F+aN3//Fll1z22c9/9otf/CIFwufxxx/vQ9YveclLKO6//du/qV2G6rzzznvjG9/453/+5xoVrqj7SSed9PrXv/6oI476+te/Lq/e3sE//uM/ppHve9/7GDC16MuuK1as+OD7P3jdt7+xbVu3xTgVQp/QsVqyZs1Tn/zkJ+kNfigWO/qH7/rAy1522V13f/eaaz6mcSrQj3/8xA033PTf/+hPTCjnzpn/nve85/rrv/nAg/+xfsPTO3Zu72o/6L3vfS/GTj/99D/5kz+howSIW8xcc801P/jBD0xeNQBAxF/zmtdY7Ra+4447cKLgqVV8ksmaTV9UxpJkqkER9YTkIVe6kQkTn7/HZFYJQCuiFVAgUyIkjAN+1hN4kgPhPKbLMH8KwYQkEAX1jenU+6zm7dt36PjoXJubbBomrtiNYdng8I7unXRLf23tGQ8SDgz0WcxCjRBRVkmnnnqqUew3vnHdY48+PnfuPNpz++133HnnXe4VOf74E9kABkCmHB2VJF3FpEdcIT5RuskuHocgGZvIfE5hkYLPCWhgb3/7Oy666OI1zzx9ww03Ugick9a///t3vvvd2xbMXfSqV/za+eevevbZdXPmzHv969/w8MOP0hi9PLuzbt2G66771ic/+enb7rxDQo3ijDPOPOusc376059t3bq9vb3TEHzb1u6vfuXrH/+7v5szZ25np8+4dz/++E/e+c7ff8XLX9XR2vWpT/0DIvPmzfOuAT/PPvvsY4/+2NG/VRdcdNVVbzIEMN4gtOzBGQXiJcwf/vCHa9euzTWjLVu2btiwce6seYi/9KUXt7cb2XsROPCqV72aJG+++ZZdu3Zv2sQ8N1vEu+aaazdu3Pz881uXLl2G4VQKASL0Y3EykD5R+RERBlQ0J+CRDEk1dSZlziftdLEpM2uo1s80tZAMJ7JYhRdGAi15iAXM+hNIuEc4MCtqmZCfTioIuISJXcUDp1uKoQchPryPjLi6Kz4KDRNaUhYrrRoFKVEWJsxamg5ffsRv/MZvxEmPunrHfE84/iR9NLkbxy5ZskSSebPnWdr80pe+5E1N8oB5RLgsQiEbPMtOOB3MqkQFM2Id3aysKeaxZGKm8bz0opdfctElaXflBYeBWb58+aZNm4466igE3ST2hje84frrr2c1Na1LL7706KOPfuSRR/TaF1544VNPPeVT2Kwgy/TYY4/RMNaXEfWy40c/+pFG6FrQBx988Oabv+sTgcg++bOfXHbZyxFhNHb378Yebhla5eJj7Oc/X71r967WtrY3vO7y++6795ln1xIvDYZgjI6f9evXP/rooy6R9NKyrb1lfE68I3Bvo36JNjOWCst4W++78vIrP/zhD5v1+oI8hTYSkN3f/M3fsMTZ6WenBJiSQXyKGEFS7ALpijwn0CrkxEl/rzGoNKAVXj4mIUBRwiqPSzTAzCAxMyxKgBNbuYpIDWRCd5UKQaK3ZmHMtHHjehMDpgIdr4Eay+I5pjL3pEzuqrm5qaOM4mPfCTmyuObdLMqi+QdaLGAtVMOmTc/PnTsbnJG46647161bN29+J/Gxuzhx2RIn92gMmlkUXROP66jQzJUEYZlGmYuDZgtblpePFDBV0AAYfh3imWeelbuMDj9sxYD3gP1xed3GjZtM2r74xS9l8ZFeuHAR4/S///eHP/axjy07dPmhhy6lFmHS6pueWvvUv/zLv3znO985+eST9SS6gquuuOqEF5+4YsURXe2WNUKr7r737o9//JotW7ZRMn2x+ybjNFZDnIIiT/STbYxZAImVDa+XB/vPPvtcg0WnW5l5JhDn0CyoBT+L5vqwNva2b4vJpROnogygn31mHel58fvE4z994KGHTj351M6OWSuPPNplLG7XZ4A/94W/v/vuu2UkbQqKLy0HkoEMZyysBIboJ/tqELEpnNpYHE5jQWuTJXb6mV+WCjkOHDI4AwNeC8moBCYcJNIUPviSUEpp1TEBgZjoeLf50Y9+OG2qyWPJN5pj1RcX3kAsGIWRkLBAoq9PJVs0f5FU8nVRDJrMJzUFsFayYMF8s876hoWaexlIW7+emGyigxODz/AnGwPKhXiUsXASZGVqO1/2TeiDg+zc1XP11VffeONNa9asyZGGXTFPPvXkDTfcsHBhTHgZP9aOycQkJVYiI06DtieeeEJCs5YzzzzTkhZbC/mI5Ud84H98QJWLhW8Sw2KRTCIsX3aUQe1555xnYP37v/9Oc3yzq1XntbjiWRMV1bBJm8FqzOVxaH5G8ArY3tquDbD0OakSpQpExTqgHQs6Me81hkblmIXlmw9hQ6mFGfJ3vvOdpnq6BXyuWLFiyUFLKP2bfudNRx555Ec/+lFZyLHQjCE7SUrF54ogJ/rxhIlCNimLJUNVnMiiuAzjas8kKSN+oS8lJvgwkc488jHTZnWC45ITkD1TI1bAYzqPUgkrOQER7hVXXCnM8pnhcmIdqiQcREpeADGLguORcaVIkdwh8bLeJEwDRHE4POSQpbpL+sEFL6MmJbuESdMkLC1o7DWabOWSs6CF+YlpX5LiZ4n4HDS6TnAMQSmfbVntpj6673/6p3+Wr67wxBNPnNUxyxxFgDpqJHRLT3rQQQfp0/XdtI3aMefXXnvthz70oS9/+ct6fJzoUinu4Ycf/u53vRvOV77ylbe//e0UURd/++23G1ai4KMgOtw//MM/RPzII4/atu0+tnlgeKCtvIsixuiLZtvHpAHHXhkjV8tbjvhTv3vvvZdgjfoMK80pdU9KJ0cDIVWjUDaOkVQEyssvpOi0IiOLT72QLt6sVNHQMUnCzxVXvvaMM87Q9ZmZwZS2kpuAQqUjOgF+paAeIYQ8i1Ykcgh2b7cXuVrSM4WRUA1sgID8knrJOCoynVzTidWwoiKpyITGxhMgKacf8h0YOPTQQ3VVVXdPsoCyaGuJxTyPWTxJSE3HZ5AHon3LgsraKgXfKtI/fO4fund264BIkHAhaO7ku3PHLkTUtykUm4EHCZURw2mhMSxcsS2QnKuqLIXHDPDxwOconwWdSy65BA8UjmS+8Y1vGNKhnBRkjaxHymfcBu39738/X0L+kYcf+dKXvtQkXcEpClaltStl9dOrtSJJRK06b9WFF15ICIh7NJZF5y/+4i9SIB0dXV/72tdu/PebZDFvzjxlUTTs4TaVFWWvfKzmMnIKrj8BkTV8vm0G0AQ4EDRJSVrMK10WgWAVQene/ta3a4fGph6xunr16m8XZ5qlEmUtOSIcVjPAB+Twn0XIx/TFIiUjPGMg2c6qkXVSaDz+xCMqWrWBzEN6rsKGgI8E8j1mxvLzmL4AIP2gHJrawQcf8tJVF+zu27Gzp9sF8g88+MMFC+Z0dXVoKgsWLDS+sYh97jkX/Pc/es+czvnXffNb5qRLlx6mfs8996xjj15pINQ32PPzn//UfMzIadmy5Zs3bTv33HPOOOM0K8jDo7333HPn4Ycf6bswAwPeX8+6//4HrJWsPGr5mWeftGnz0z9f/djAYE9rW4tVnt99y9XWt6/91Mda2tu8JYkzGvFRmjG3iAwMD4Kw9t6ueC++6MBFxLP8iBUXXXxRa7yGtgja9PDD5ijDrNjixQd7G3DgAYfM6pq/6oJL3nH1u+bNW7R9W88dt9/d2Tm7o7Ph/vvv29W74/gTjjv++BetXfuMiqRqxhX68SuvvNKI7bbbbiXeefPmXLDqvCNXHnb4iuVPr11rAZU8jXOuvvrtwyN9X/zy5/sHdlhlvOC8cw5eerABevf2XUY0u3ZvefkrLnn71W+++ZYb7rrr1o7OlkUHzrvp5hs2b9l45llnnHDCizdsWGed1LZGdXLhhave9ObLV6994tOfufaRRx8wwLZaN3t2u9WLl112gfslZ3e2Do7uevDB+xcuXGC0MKtrblfX3FUXXnDIIUtsR92w+emnVv909py21WuevPCi8w4/YulJJx/PRGzZsolSGGdddtmlr3v9b37iEx/v7t6uQpviigD1H+rtUTh/INWP/lAPJYWUKkQUXOpPKrHH1CW6W3/5Gy+Dva+rtC2peJSYn7QSmH6mFeaoPx8T+iPDFP6a1Rt0Rps2bdDyumZ12AZaVDfG1FoqQ7l48RJRdg/p/h577PH777/fItxZZ73EJJFOZweNJkP1ox89ctFFF7k9SfO1fuRdnLbrDZ95wNNrn73//ocM/82FTUW7utpPPuX4F7/4WJcbWJmaO3dh7+7Bnzzx1J133n3AoV3Z1r0XJT79dRYt5OVEZWxwaT3ttDOWHXY4Ungb6Y8Jh36cCZGQnikUUbBDWjzNk0Rr1LM/9NBDSPX2bePv2NFzxBErFeToo17MWD733DrSMBJl/Bgb1UmMc+bM0mwefeyhU0457ZCDl7I7xotz5s5av37dl778L9u7Ny9ZEn36QQcdvGH9ltNOPfOgxUu7u3cetmwxQ3jXXXcYZOtkycdivsGi+b7w6aefesQRR6TNs3dABT/6yE/vvPNOZAFtqMXqypVHWDfo2bVTv2x0a8hhA4rROfbYlAtXvUIZ0xTOndeldD/72U+tG+jHjVKMCs468xzSMDj0Fkq+n/3c36kd9s8AVHKyUnw+oECtSz1BkPQ4UWTuMbujVEpAgSoWwowKqmyZTdItNIOoQEJqYzMsViC0vgR0Fh5ndS0gUHqgw1UlyuZFGsmKynKuX79RwGYPnKl7804JbcZXeMtDhEXQGNUDauJS7dzRayaxdeu2Hd09ar1n90YK6rTj1q3dhx5yGBkZSIUJGdi5YME8X2lvbWnbtm0n23nAooPmz1/YPxLTannpW6J7GYslWBB3eOaIXG15x21coBlE/9jUpmoF2Da6pS0ZVuoZMCkhTnIV1jBOJeHZ3vsiccajQaNi2iFbMtQ1w1Ef5MMe8y1F0O/1G54x4Tt4yWE6JK+gfIept3e3W6Xmzu3a2bMDV+j07OyfM3uBC5iEmVVEWEcCURfoaPZ8+iELlYNmNh5GCpMNrpjU7zU3EDKpwjQc0q2LIlX8YxtByp0lGhuN9s8EKpd39EbP+NQyWRM1SEHN4kUdcMBiuqCACxfhKjpSOAKqnkM5FRTcY/oZSzgZRXrgMEEQFwZPnMT3KPaFFBRGuuSgZB2T7qTC5yCkH7SKQ12BwQVE2RSHFT1LiQSe0PvY1h0zKKkiqqzu2lAQuJC2b99K4rbSEejixQewlOqDWPV01j7sUZItTiw4t3XE2viihQeRs/URL1dUTLsLZRqpwvCu3T2zZ3m32dje5j7NLmrXNbcdfWxgKRuwh7g8wUUy7h6ZfG9ZZE58de0209e7v24w7GscOIn950y1jWnmGU4OxeagOESqS6xzB11f/44gMmaEF2NWB1TIwaBdGbNuNCG2RsFVPEzDHu3B1SAGcshGW929w6mSOXM6MQ9Hs2xp7rSXY1ePQblcYuHCzXkaA9vMIi46YAHrbm5e1ChmnLSNkikjnMGBsGT4xIwyZu0wpVqd3NUUTrKmaAmIpq1F9fTs5GsGbIT+FjLmmRXocT1kp5er9jSxOHP6+ru1BwlDG4o+IJjhUuN7vATykSptLKYQqU6SAPI5pcOPZIkf2jatk2WVIOqz2EXJqsQViSp58iKVtJJwkLXdMgiOuaECh2+dJu4pcHJoT3cgSUjQzUVjTbqS008/beHC+aqcoH/4w/9gFYiPuGmtPbUk4j0KY7Z48eiSQ+Z6vzx7lirpIzuvOinowkXze3ZtO+qoFZuf3+S+WWq9dcsObcHe596BXgyrezkODtn8QUuiZ2+PLynFzMC7/iyR4hj3GwmUJmTfhFtj6VnszFAR2j/t0SrYb2GyDiGVCT4iqetqdO6chWqRguqCkVXTtKerK04L6RlYIBs7N2/asnbtk11dsxYtPFAjLMP3JkPJOXNnL1u2jKI3NrT5KEJ32+45s+c3t4ZV27VrJ57ZP773+OYulFVaIvr/mrvzML+Ps0Dwfd+HultSSy3JOq1bvmQ7duRDvq8ckDiHCYRhQ9jlnJ0nZNgwMAvsLOzMMzs8w0BgMjw8IYRMJgEmkASStR0nxGfsxIdsy9Z9tc5Wq+/72M9b1f3Vzy0r8Ac7z5Za9atv1VtVb7311ltVb11yQXkZ2W8AK62U7RMBMTEH3PzpqZBRKM5QUjg3NjWYEtx4wyqDq4MHD4BcunRxx+I23KwgxCd4Yww7CkhQ9aLdQVVBGJmiXqZb5pzszsgkXgjLZ85LdmLx8cmR7QIgx+UZAJcbg+b4EspsWuQHyyLR7Jkxy2BFEEc2UeNhQr5qwbhTcQD7RNyo7ykMHWmqVtmVzVZ3dldtAABAAElEQVT/8i//sh10n/3sZ5cs6fjZn/ufVRXtxr59+4zYLKyZwn/wAx+2lxlwTVXtt5/825Pdp0iBnnP9t95yu232+Iaoe/KpJyzKYC96pbvvvreutgkXmYGd6Tl/5+27IXSs+8hv/ta/PtdzJjVjUhA1Up8Ck3SMQb2GwNBK0rhFwQFk6mfiwN8nGRAx53sMbMpTWbRNTai9banaVQsf/ehHly9dLil6eOdewBjwaG9endu2dYeK/+7fP/3UU88Y/9hK4i2R6ZlRm2NWrVrhYOD69Ztuv+1uD9Y1NjR9++9D22qWKV9cIoXlXZ0Y1HJObWXtk8/+Pc2/NpAY0R32WlSgjb3gLMgohTqTdCSqVYoisAEgZnNL05133nn65IAG8JEP/xg9ib7ilT2v/M1Xv7J69Sowixa1W5yTI2yPHDlq/PrYY4+1tcceoFS5waDc7Fy/HNmIW5iUY7ApejKwAix3AGxuDp4ActzLStAcDI4RwScHW/xsp5Co1fwJM+TIGWRPbkFxbVDAzKWDQbMc5cnARCyhkMMNWvO//3f/fv3a9b/yq5/8+te//tGP/jiF4uYNm9euXYviOnpNFmVNX/KKCz3i3jdfGB2ZkMvoyOSyzpXXX3cTHRPK/tcv/Lc9r764Zs0V0vnc5z5/7z0P/uzP/Pyp093//a++fsN1NzQ01HUu7jxz8gxi6q7zaUd7vFzaqIhuniVH6fGdejQKVRAmk4+NGozyJMwvVgYf/rlFKzGEscX5nn6iy+cHPvCBkfERRcZeVFHUaj09Z6VgpLF509Zf+IV//gs/9wsPP/zBP/uzP336mSf1/1PTY62j488///1XXt5z7bU3vOfB91tDMnI0Av793//9tWtXW7ui6A3ur4jDdDYhfPynPo44RDWbKB0dDf2RHbFCNRXScWn7Mn/WgchIIyVZG4KREdhYTRkbfO5znxvqL6N5/ciHPzIyNqzpvvTSSxSfxleG4BYOli9bIaOf/Mmf2rZ5O7Xgtdfs/N3/+NuyU3CMKxFEyLaSln5y589MOrZPzBpMkIxPROYMuHnjM+De1hTRgrMS6bEgsyCJIi5aZ8hs8+dgg+dQr0zhkz9zAypg9FAWKnDn2MQYdZ3Vv+PHu3/jt37rn3/iXxCfuSdio/W69WuGxwa/8IXPm3gi3Oo1ZqJXkArYV2rU+2yz387OmAJLCiEs6rz86stXX331iq6VaVRK4edmxVBWQxCSHKgGJWWEJ5/ElnN8ySeVfi5IKOBcZNWvOD4B8OEAzEgqA/ABzxDJfFKPH6toBJU+l+cTTzwRJ/LOnFq2dNknPvFJan99qFBaT0yzbduOzqXL+wYGkKWxoRFWUkAEhcLlponSNB+3beqLf/FFmd5www1QyjM5iaMqaphr/uqnfm3S+zfjI7/0S79kewd8NEbFgok0M6oYEUqmqu5SV+fVFbUQIB3QViK37LqN/Sd/8tlvfesJp7QtKVtVsklAwflLRKk5UulTf5RKnT8zfUDiZlgBg6QGrAll6omewC8yaP68uMq8ILjISQZSZ3JO0kWajI1SMXIVlH04cj1ljOUNXiGxjlgKDyFoiQUyZ5FTForuKG52gle07zRV79GPaO54SAopSqw8YUfC49FHHzVIMizMpZVO1Ee8bVxnKqPsWFPW+j75mur+0R992hIBzZHqr6nSfU9ZpDadZWM2nzjNzs5qFzD7i+MN1qTnB6aJCEgE+dxUlJdDAXNBOBiJ8IdPRgmAKADgJpQn4gAg9jLRFMrcq6trpSXy5Z3LB4a8LFpx//33p41LoROor7OZaIIwxpGIZj9gpoNqlqDEDx8+TDRiZbbtI4oGDLBRuw5H1kaQZCr5LUdR7HhatWr1rl232EJjroJH5ZWbKCTTcHxGdvU19Q1ePikrQy7aEvQ0vlc6KZMIFGrWcnt6e7ysctddd0FDoRjpM5kI0GCkKZQJiiTDDSsNSRHQIUtfTCJ6BuPwmX2CdPMRF/7mzHK0yCEZQHJlCy1NReGF8wefE4IZB39g+APP4TMaJbcfq3u0RfcY4Zl4VMbChnhGVC3Ni8SClmowR4aeUInoPtiJgpV6Z7sSH33sm07Z40j5GtrKOWMIvbyTCAnUffYUkUE4exnZNkHKxXyfdEEjVaWeYCVKtA1vbLgmPnb5xaogeOwLJe6EW0iLJDQNmDhDCCVaRVbcKJCNIJmKyL8gN8/MtRwZTJUrGrQJk9CXQaznAhbhYMBIXEZRgLSii0oanimRpoIsNikDzhIOxraYICacZSqGuCD1xcYDf/zHf1xX3eCdsfqahve85z2NjU3UOxCGm4wAy41NiCq4TqFvsG9g2EmvYZ6aimTVoFUVPrASA8VEfPzxx2UBJlcQh6z5s4GxSw3/wqBMNqIwOTqfBQBzcjEHXBqcoXNCOVRaHAWDcoDhCY/sYGfM8qCkqbHFLNWGBvNrZEJKhB4IPZGl9Bj3KDCKOGngcILtOTEBLg/5Z20X6TGN7HIFy0JGePe11/Y8/fTT6omKQG6l6KX19Lg1Kbc9UTAcABxAujz77NPOAp09c8ZCy5ZNmwn/0eGR4G3icGx8aGD46OFjPWfP0zh5513CdTX1WJYiCT4aDJ7G3BbMCRV7l+yH18rseAdjvw8+tlsZQbJBh0x01IB/1EHa/pKplz1xkuZx6NDhB+5/yGYOYHvf3PuHf/iHq1evQYQgUTLiMqmY+RIKV9I342wK5ltvvZXs1KELxUOIhnERDbzcJchYwTpx4oRd0m/sfwOY6jLXWX3FWjSBQIy6k1DItvG9dKorqxc1L2ppbMH0moSJKUiVRdKvXbv2oYceAtza3PrYtx576qmnICOpXC75CrqUNZWDPzshwBmGOwqWUJUChIvQwjHX1eYIpTboItFEpcggm+KzwKNIDkCOxQdpSIszvf36bl3M1VfveOnlH5ilbt26FUGtTzz33PNyUfdX7bjOSAu5d+683i4QPZ2BlAksHcf3v//C0WMHZaTWpZxlg8QN54qM1Dydp6z5M+GI47Mh31h8xEJog4HB84OUMp2ddjzNHj5yUCd41VXb77//PtTpWrFMvYJ85eVXn3zySbVCbZ43p6Iiheu6dcssVOKb/fvfvP7jN+q+9ux57etf/6pp2+DAQFkz+gLUC0e3nioiWIQ+ChoZecv63GgCpZRXHDiR9ac+9SkKin0H9j3xxHd0DgpqyNjUrNOI8ZIcgelkMZZ0YUg51dfnYdJaPGpQ5BgJCae8ucpyTft0eEYVWMazImWaSOB95St//b998lcHhobAv/e9P/prv/arVpjMooL5XQWYDq6oLLSCKjwRWS7ECgNzS4HMu971bidSHn380Tff3G+2t3xFSy5FrhpgPuHJoZiZENntMxufBQCfws2RDYAMoxSXncWLCQhENmKmWHMxc05QyQkVMiP7o5RWyBCBP/ETH/3xR37829/99r/5N78NfGh4wMbyu+6665FHPvKe97yPLolwtaV8YGDYmvU1V++8YecS5NNSX375xVBnDvSjVMIEg8YffKSfs478kzKVj4rhmRBQl6E5xyWZWIgghcyjZCf69F7oVbuf/vSnbT995pmnTcIIciuTn/gXn1i5YvXNN7/zd37n31IdojYWHxwc+Pmf/8X3vff9/+nTv/f4498y+fniF7/8wQ9+4Bd/8ZdsE/nLv/wrrWhwUK3roCP/oNFcY+EOn4wtX8oH7dPcgvKLqshIcfPmrXB/4lvf+Yu/+KsjR44pxYYN62jvz/eeViiy2XbZHF3B4bNz506Nh4rKBPCeO++R5v+5dy/ksQ4WAa0ueALOY/27745N0M6EGLiTdic+cmJF1yrs/o4bbr722p0nuo8kqoJP9esFe31CMjI1IN6yZesdd9y1ZcsWfmvXrt22edvj3/7Wxz/+cSJmYGCIUm96NhYmsIcci2IWjnlqzJMjiDHHkZDMJvkF0QruLPV5iwQt5cK5Ri/beZOTy9hIIgqQGCInzR8g+gpCKdEx1s/8zM/ce/eP0up+8xuPdnc7irBI/0hS/u5/+I/bt2+n//uVf/mpT3/6j5599rkffP8l42UDdqkZAyAuKTIamo7Y+qkrHxmxQDxnch2AjO+EVsaEnb7mLIEwMWblSBSMZXdsKtguZkLpqae/Q1FlbXzdunUAnn76mYceenPTxi0dbUvg5vwT1dXkzNjv/Pa/7Vy2ZN+BA1/60peNkuVOo/7Zz/7pvfc8sHJV10d/4p+RxG/s3Wd7smXAeMwzsgtyg6SuYpAl42TsYNpncZz6Rjj5RJmwePGSf/3rv/Hg/e/pPtUt6DOf+SNbQNo7mpPWNXo9faCNAgY4Sktx8eyzzzY32xwzhpIPPvigWaAeHP0zWRSZkSNmJfmMiAxMfWJrc6k/+7M//+Qnf4XyEU8//PDDv/zJ/7Wrq9NgRN1JAeqwtW2lvrZxcnrcvnp5ff7znzdh1R6oma666mq7sb70xS8989xz//k//xcFb2oJzlZYETMPcDORWvKHP4fQ7GAzAHJotn3y5GbndDIMrLLI8TkXDCibzJaRVe6uUvZyDdAEnMHyp0SZXA3gc9u94447HrrvIYvhRpYoqO9ob+/Q0VjXcYTlxPGThp56jQ9/+BEElZ0ztdQfkiUTpYbcgV8yAGTkk80jydHYJMqHnXCcs0o+lSgKlROJSkvNxk5Q61W400DC2Q+taNeuXcQPMUamYl+5o6d+1mTDat5HfuwnTHuXLemC4/ZtV23Zss0AbuPGzR/64CMOFbU0LmpftHj37Xc6kGxORvzAQ6YM/PWwMuWDblGOOKaT8c9zr/ikOdcU1frho4dXLF9hLPSZz3wGQxiUp0FNxM3FBMxI1ogQd2KOP/3TP8U0WpfpiyCQ8lIEWYnic/fu3XaBQIZqk/Z+x/arzfr1VE6+Nzc2E8Y33nCTmRlgRgoJ8aAnh08jUTRJFdduBG+/yKFDh+xetU/ZmMSWP+otmCgm4Fz70MvR2YVDasUnSCb75HyLIJ4wz01L64I/yMt28aAZEOCyIxJOXJLTzXbQfb5xZOTYwERR/r6hPo8GYURzI/5aLWARkx0zpLaW9jWr1jg/ZKeSXkMzFZG0AIB3bVaQu/HnhOsY080tGRk2AGRMtsY3d42ioiKWbjE7JAVgHueQZz7DlFf0DfSCIcCIH/5ikTSqVmMUKSEQYwZdoV4M+/LENOvWbUA4dSZIhdl0DGe5rFmzBiOKLmhkLN+0MTfujIXdhEYUJG0tnUMhERZu+vpz53pee3WvpNQ3mmy6cpO2/bWv/bU0xc3EhGRsALDps61NCsBoK7GLPdq4U0TFKTISRY9sGOCMqE1P9FaqIB6xrag5dSquS9l05WY8ozoc4frff+PXFi9pcbODFMj93IxTvvYzxm0X2SACAczfztrYnE/Zt/IK+2v/1a//Uq4OmUJS5WaEOQAz+TO7Sz8zwnxAZjdIRZNaUD6xkKBY+ykilzrA+cwxM0xhZ89M6CKDnChPYAzJJAUcUFdded5dFePDM7OTjXWN/NFOCiSHI2xykIKb6zx3ZIfH5NQQBZS5AZ7xh6ekANe6eFEzztBQ7qKCPUo1NnTOGkWMpa2GiuAQ27DbXNLBGteHRFHjAlGbFitt9rFyOJEW0sfqGxtsAG1salE9bvqob2gZGp6ormkM6ufni6yYT4zX1FWdOtONKe0fcsvBxMzohf6Tf/3VLxofQwkm0lcKnM1+be8LTS2GvGNDI+egjaLKbk+ICXLQAmPOxK0kCquBNzbWjY4NNjbVenHGHpTqiVab5Ds7V1DWOKO6qKXpwsA52+G+/vWve1Gsory+yoU/U/GCJ0K5mMwuLSVtbukaGR0wwPjGN/8W6xBj1AhoawTzyV/+l7/yK5+C3s/97Mee+NYzX/zyf6FI0Ryi1ZdT3pX93d99deXKrg3rNkDy5ne8gxx5+ZXn3c3mjJT9jXU1S6e9F1JmvjiZVHyx5SDdvWpoO7NkiWXYUTePxP6VsqqVXesJZvmqlBLOVmNzHQWcs4EPB3+1hmLZBFyI3hAiUa3JAYzeEEAi8vzGiLlkLvMjfmmIzBhJwCwjp6ipJqIpCPJJINlE6NlyPSbS4wZjIKyp3RutKySZ2tvfOz41fuLUCaoNNS0LWOaMMIECS0rVMhIHYOtg1Pa8QNIkokeexw28fOeFaZmOSS+h5uSVUZK7btECIC09coloz5RYwCTV3h6LKAjY29erSgyH7aKStT5UGWtiEcCZ9GbJMpDHFnzYxOd1113HnWUwHFLXFAIJthk97mDl1BpRg0wCAw5iLk+xl0qasvZpP4Kmu3//fhEDbbOqVG0prkfooj1ISiIc0EYNDthClWg0fVFMjW39+vU8qe5REljCKoqZN4/++Z//uYiwcBCUfJWCT/hz2EGWPme8nJxJnXei5YyQK1WEquERa5WyQ2G5iyUXWcCco6gppchEYGdHxExGlMLwyO65sMTN4FM+hd8PdaT0wwrCJfkKV5WXURE1Awhl+D/BfPcJ7WHLpk0rViy3AcfYG3+88cbr3Js2XamQXm793d/9v8+cOaUaiENjf+OOMXcKjQ6ZGOE9icN7AV5FSeRoi53sctYXLpwPNAxhyypsiUI4fGOShI3wFnFow6hkXcbolUsKL+I91FVpuNhz/uz58+eEtptANTVxX+g7r0kohIo/32+P+qh8lYuRnYpRfMz0sY99zPE3LKURAlBDDBYscE70CAbPPmDE5eZga7nIYlvMunVrQCp+d/eJRx/FWFEuAJgAcWKfjQ43bl4JnYa8JIJi/f0XNG+fe/fupZuzawTC2PeRRx5xegTm9XVxjaP6grO6gJtP8/qXXnlJQ3VM9PZbbjWgNN8ytKXT1QOAdc8fsni5WakVE2EhEK+MVpQ5sEBFODxC3k/YGy7lgv65QebyygWeUaIEwJ0/c6nzJ9tnNrmis5t/rlDR/+ExqDipbHNpzhd1Lm++kkZEDskB9qktKpVPWqShobG77rzn1tt2oeP53nMXqEUXNd9558M0lXjo03/z+7R6dvSQQ0eOHjK0MndpaWrBTKJT6KjX6MfLQv7LgZiRl8InLgk3VjBZISmNJnU3tEhYFuQVV6yE2vBIrIVUVcOtzDVXao4mwYq4A0o2DS1dukSb18pcidvREduK9VRioZtzZ52dS/lA8jd/8zcNtt558zs/8YlP/Mmf/AlleBalkLTxiubyD/7gD2St94ezSRFyJUJ75c2bx3UmxfKFKpZS/cDghkp4BZiNyRgLPzU1NOw7uO8rX/nvL774faHKhUfZZ4fPWXbEDWNTZFtol1esWCGRBGBNDvPFWsbP/8LP2uSF8mh4xx13oAYlGgwbG21PjrVltDOSR08jUZc70OJpEs1NDdRK9n+QEW5htgHg8H67Extlh0qOMhvIUhvDXDtHcxNBSytf+MLnf/qnf2bzxm0a+bp162yUhgA6A4A5ImSuYPMvPjkSZeZaZnJHE020Cos7G0EcOeJlt9tlIHClMbkxaPZhZwMnBnzmXbQIcqS77PCNrVnbtu246647VqxYRelonmsXh5kryfHFL34JN2gkN9ywc+XKK5TLyu/IcHC2LfFSk6wbDeKy8JgFReeOcIytcUJ9qh5ds/QpFy/0ujCo0ba6xka7MNsPHNwHJdR87fWXXXIr8Zvfac3vxikDzpo6Ihy+xM++fQd09JZz7rnnLjUCNxjKC3scOnRA1f7t3/2NqY+JtjUb01gVrLYIG/M5Penjjz9OGlFnEtJKjQuNphJZ6BziasUtm7etS6M9HS7uyQ2s++RRrEN4k09Ll6zOE51jx44eO3YEGOkuBQXktnFEWays2jpjjGTSVlE1qs1gUMRREPwqX4xr8oTa2oOV97Vr10qB+JTL33ztvw0ODiOFm59Vektz+z333Gu+ZbipCNrA+d4zNK9ahWG6glRXWMwc1ITOnqOINWF1GnZAk/iLv/yynbVScDSqoaG5taXNmVJnY5Yuq9d7yIv6jKyBtqyhhF81IUZt8lFZcAg0Yl16zvDPQb4zC3FkMLE44vNy+0Fz8FxKJWyaMZBuTki6aMSAFMSTDS02H62ZsDlw4JDKoAG1+QETCDT7TJNmV6nDpIrKGssK1bJdDKbaJItYEiFybO5MdRDKAcWLvMqDBLnkAPAl2ulyiVK7TNTHwMAgEUIA2yqhHwcbm4udjZscHxulTqwiIAkDq/nmMRhCrFh7mi4ztLBcqS9zJ9TgUD9hDDIkRzI4SR9qtwSZYe6cJ3BQzQWHHkiTKAVPZDF0Nr+ZxbuJ6WMDnnYCH80sRUm9wVScbTKojZqL94lDX4OVBwYowOMIkUbiRjF0k2xbW/vQcByrZ8RCOk1UvoIUn9vRLszEgWNQANii9lC12gQjtWm7VOOOb4c64mbk7u4TKK/4mo1tN844GE+WTcf0sbGp3m150QPUeJ02jlZrmQrlol2dpTvuzDVwuTH94qWNAZ96D/kCy8XXjFWQYjIcDH/GJIkN58ya0sy0TYFhCXqL+UcyaGn87Jau1LPhw8HmiSgFltAy466rJRXGbSu2A/dCbx9CdC1fYU+7u+idt8yhJAMYdIyZ5rw+iEBSZ9BNhYm+Q/GSZ6g/lFyO6ikAZiwU2d0dV2uoCf5CxTLkxd+aR7QdlzM5WzMhQbKZ5hWMqNEGcn+i/nRkemSxsJTaskWXrJVULqascQw0NJLUX4feShAACUFbrdfUxmV3iRR8aK1i3uCTCbCkEJM7lM3iYTg2EppLXBKBYTJsjKPRR470r/7jKtKRPTM7ghvwurigMQeAzMe6EXypPdgkYM9HhnExI1oZxKgWlImmGJ1nhfZvAI1BceGRI4f041NxQelUU8OSRNI4TmhwhXQwCTERQ+G4aJytXNS0CA7X8YlBBYcAfBhlCYyT4Q8g8wNbaKrZYMFcyBQjuIgjKjEJ2uz2CUl25VXXXjlHkrf+FKm81TsYEaFF5lAS2HDnOhCFJ3IwYvmUmfVOF9jq7Azv2DJVi+bWAHR6IAFwm6VKhA/dkAT5p7qcG1FgCOIw8UGgo4AZhyhAQqY67hpJ/DoZcwjvCOiYYvIal7LGwj3PKUmjV4avDvZNCMb5HkYTEwqYPpLcTaaShMiNDQD0mEwWn6KDYWMINjoQM7g2sR2KI1S0JfDkiBIpdeQSJDEGCM5W5LC97V0t2SgpGzfAkyKWmwQOsmCr6vRwBwpYUkqnSnT3kJGmHEHhtpaW5uPHYzUkxvQN9VTxEldBjjdxJB6IigttgLchK53JpEyosjNBh26KlmozdBc6GqRAYd0OSU++oA/GSqSuMPxXmETM0H6kpGJfgZTBKJu8hEoNPB+GIwiXGJQ/8gPgyXCIku0ckZspQsN9OQmaIOegc5zsk1MsUpclJHgyPGXD5AwyDIqXRgSWP4NYyfjMcTlEoYRjuIkZKXOAVLwk8/Les+AkkPxB4gSfWnbY6VK1+E0H2cIRkKgWjuCZ1LoCGCsAnrf50PUQw2DiOhxr90ojVhIYqhmGspNpTgHXGofM4xnJMikv6cSyitVaoUQJ2hhoEDC5LDqEzKAZn4hSNidfM3dG3IStwYZSIGpgFRUf7TM6h/JoZsDY0EskRNIQxjyTCTyzf3jGsz7xYF24A6u0Y4Fcr44o0M74RMSUb8wSk+EfOQR554qWvG0qKUjNQ13PdaSZDXglHIIp0Y0nh+JjYnmlFjvHACm1i1bBwQmlgJEO+7Kz+BxcJKAwOU6ulRyayhz45KKGq8TkuIKYwl0kmCusNEjU/BkR5scPHKKwc7WBKRDgLlIrcQQ1VXsma/KfSzb0TzlK2jsSXJgqgDd3OseX8Eypqq0ALo8hBFRlyvBBPp8hmRK/pvTnLKFgQuwFW8x5qmYrqzkF3Jl8tcz0m4HSdqfEJVFMJk4OJpOKEc8vph2FOY78S9hljlekRmIVcgFklCGnlmkkvZRlQIZjnoODAPMG/CVgIW4XmtR6UzoKieeC9zNX5RTmkQ8CZqJJISMDjGBemGD6zgCcJThEwGUZVIQU8aKVfXB64ZVRYQsqUMkZFPlBvoAv0szw/AufnEhKUL7xF1WbVhRQQY+hCYKXS5Fazmi+QWfvi3nN1UQkJChVj0sZC8ZJGRSocuSKLxLPjgQgzWBKts95zDmZjH8IIW4MDLuEevBKUbVSjtOeaWRcJJsQyZjN9STAsEOy53okH7ILtpl/eAQas/FGnoKEmFSIlFGUDmoJh0J8RlypwSphkvKKNplJpCAZFz/zfBns7i8E51tNpnnk8lZzkdoJmagdjmwnvH2FESsjw35rChe/clCG585G8EVuuwhb4gKXvwqHMkvFZ05LaHawi3hF6gEWj6IuRKsUuIiV0zFii0pKhk92sHO+GaZAif/bJqUaSoOKuCLyZ+dYBUypIwNkG2TV/Mo+5hBL+1QB7Ky4ACbxnH6OklKGFW5N95vF09xRbYnFpYd6IV81Ww7fMEp2oJTpZPyZRyMceeaJsjgySdlZOt1sZCdKMsFA2VUECZ03c/AAilBcSKT4BDMXnH4WfF4MwrsXq/eiN1fumksjZiGik0EWAEW+2V36WZqQFIoicOS47MsyaE7o0uQW+BRgpSiWZhxDvfk6KHHQv7ylpaJe/iur1CcSOWzSgtDT5UU5dbPyUniGQ3bZvAWfuSqP/HFO5owASwyvUxY3l1y95IiFgxohAQIFEsZnJJR2UYjoMyPAxm1MkXV2FHaO6FOOeVhc5KuMEZpyz2DySRGVaK4649Nb9nJJwsxnMLs5YxgUyJwSuOUCJrqJEv55qIAxEv4SBBYRoyDgE08Gqxk1zKt7IuDiUCG+Lhr+c1E1qTnXxdDkUjS/Cbc5S9Z8goLzsiwHZMgF0YvPi0imyi38fxiDJsIFZKlDrnxyljmVHJpnrKXAhRtAkX2OwlbBhRtAYZAh1b7+PUR1TiRn57PI99IEi9QWOEBmnBEop5CT5V84RCl4KMOLwnCTlKZE5ARgMAz0FJYj+4gLjO2TCfaLzyQP4h0ZPTKPGAIl8RlZ5eEvYB/KmePx923NLCUS8BCMKOkUIW5PFIvRcM4xZ8pO0efyzWinKgIVKGWw7M5Y5egRnAyYeedcQYrP5IjsShxR+6VG+m9rwOSUoQTzjGTCZ2EKOTWJcLBzrJwm92UZVB3kPIqkU+qhysqRc7pFaGbQ4jM7wMwTK6iQcIh4CY/QCmVTGsuJNJH8qYisoAEDPtc4yCIKR/GZPeft3IRyA4hcIvpFo+LnIhYOgfBEnNK/4CR+ZTZ5YFVz/Hhv2OaNYC8rrtas3VKaztNBKmgSV5F5TyjnnsYk6SJ9NUQaZnXVPCYJXDHjFUV9dszS5ngpnjYOTBLawW9coatIw9N0kDB4bt5EO8kUxseZFVIFBUUFJfpE2TOhsp3IGwinz6APh0+OS0xQIHlyzFE+fV5kstRsLsabz4LuLKbtjJQzTJ5WzlfgxSiFKxeq+JQUc1kGpWADCiKXM9s+MWJOKBcp0kjGp19gTM7DJ0cGKxzFJ5mUPXP0HDd85kbxkXXSLc9xmOOgQnOa2ZEjLiCQoB9iiugZ21LITEr+TPbPwDTvHFkzr2hUuVmjafETpKBcXtFzrGq7KKPQMcjLowXKJmC5vJgjZZCYOJqiQyDZcIo41/gzMkjtjpPUp4din3o0DXxjKJKNOCDF5zB84ekzEWSuFvwQ4Rkg2XNFK6KXsgtP6bydmavQtwua8yvKkL8lJWuY5IwK4gRal08sA2c0igTR7eJ1IMJyAFuLz6A5WhaokT1FazQ4lRBfQe9UN6l9JCKn0SSy5uKmxhltKBO9SJ9D9MzDmY0zqwHkXREPmse4Kq/KcKd8pBH1IS9o5xSmK92WUxFyZrYyXkUNsSfvmTJHecqpUocXt9mwc8vRg2/sP/D6xPhoTd1Wq5Rti9tUdixzVTq1ONrU2Do1OmbNxCUd5TNDTptOT/VPzQxu2bxhYGp9dZWbm8aWdiw6vP9o37kBS15r1nStuW1d3Fn8zA+Gx0zQO5au6FrUUX+s++DMkfEli1sWdzZWN1bAwZWTfYMTr75y0MJkTWVjtSlW2diyJTXXX7d+eOT4oYOvv9K/4lz3mc3LVvef3r95feXOa7a4pPzZFw8MzVS3ddkm2vTsk09N9s+saW/dsGp2xZKGr33viKMH1fUNV1193V9/9e8mpuMVTk+hoIsNmvmIpuahYjCnlzWJ36inTOI5DgyfoD5hEIte/i5KxAhKZp6VEmBU9lxkAInI6fct1kUAVZcaRmj78VSuu0IhpfqykUUOYudMs12kyj8g+GavwsGTiMaj7GDhtKEwuH++tCDzp5wygChFIhksp1akWeSa4xafCxw52WyDzMDsokjcouBUDM3gRdSIPtCiaNA9vTOXbHMt/ufOnL3KWY2rd/T39dV7QmRmsrGhdnZyfHrStj7LrYNlHuYccWPCTKUuvHw6mHx2kp78xx750H333nPi2NHDBw+89IPv19dUP3D//TPTk8O2tw4MfuSRD5/qPjlit7MVdrqf8Yn+vt7+C94OLb9x53XmsI998xvPPv20JrJm1UpbkeKtUSyknU5NHDx40EXPD9z3wKEDh0ZHBpYv8xaeTcHTb+x99Zqrd2y8cv3B/W8cPrTve88+SRps3775jdf3HDy07+4779i6bbPbqfYd2P/qq68792xnLVJkOlAtFLTKjkzDBeT9H/wJk8K8bdZFKMcCABzFzHXxOdh3KVACCCsEVxpMBEBaqspgpYkCyJ/s7FgAE3FT6/zhhBOaoy+w+Yse2KR0soPtSWD+/ldqbKEjBBY8aslR96b1mkkPD/aPDg2ODw81Lu/sPX3GwoYzyDY+Wcn00EB7c8vJ7rMdHUsnx0amJkZt1a+vq7l794N37N71W7/xa+PDrsuzUbX2mae/+1L1yzarhFiasYY+MktNOznhro54N9itn7qO6anFHa2bNq5vbq20A79v4Pzj3/yGSyGWLV1rowEWnZqYnBmfrKuq7jl9prbOBTM2wA3NVEz3D4yv6eq4cKamt+fU5FRLQ32FW27K6qq/8fW/ItdXr+ysrpjsOXe6uX7SCZq62kYv63zn2991roB4iTXx+SWlTJxM+f+f2OoRVuxsCgx9ZgzVbAbIQQvsi2PQIkKOpl1ygM4RcigbK2QfoZgyw2TbJzlXhPLM/oXNIYXMoDluQLzVFGiUpsOdTYYtYHymkcRMGqiqpqxgjJYQx97MoGMqMzs+Mrpi2fKNG67UIdyz+1ob0jzx/fAHP7R23bovfOGL+tl3v+t9r7+2b9oDtShZNkWsbt28wSqJvtIt4fWNtY0NXlm1X2gU22N96+TDg0MdrYtsoTNPsupObVZbZY9x+eGD+5767hMPvPve667/5FT5ZP/AwN987dGnn3xx2bJ1rjoMcldVNNY322U1PeEiz2W9o8NLli91evTM6aMtdW5KGlvU1llXW3b73befPH/uW08esLOttaNpavC8UyKjQ26OXX7TO285eerMt7/9960dS+y3Mkqxbdu75EmU6rRDfWr+pfeOTmWhYMok/P/cLq0jNQ6jghEFFaEcajZ5XEQ080au8bdI0Iy1AA7JFYUokuMzYSaZjFT8Zpsjw2c82Dlj8Fn8AihS++GODLkA3ieT88qFySjxnBvq5kRDlRg8qqxapd04sw42OZvk8PDUjMOl5jcffnj7jddse+XFZ/ftfen++3bfdOM1Bw/sa6yrnBofMvwtL6Nudd42bjo2flrU2lJzOs6TkWnuj129alXvuYGh/iEStL11kd139hOau5RXOWnUPGlH9OhoR0uTTcfffvLRNeuvmCybuO/+Bx/54CMnj54fHJpwDYqVd6TyNJHRx9RE+ejQZOuiulMnDo73DjRWjk8O91jntGGv9/zpEycO9dm1Pj5aWW4zytjklKvOptyf33thhLzv7etf1Nqh24hmODsZrxUbu8fZ9qT9CCkQ4/tEnYsV/8Mp/08bWsozUs7cqb64cw0W2fnM/oVPqWMhgxagMbybN6WJ1lTm65BiCIhjsh2MkiQ2R/bJUUX0zZMpEik+S/Eo3EVojlLYhX9OObeHCNU0DPaTDtq2kEgn1UgGINGNGBxXd/Hi8ROn7RjHuPbsLO9c8uyT33W44/3v+6AUvvXEt9y+mYW7nVKuE3v+ey9u27a5ssoZlaEq17yV1/2rX/11yqJ/93/9B+zVPzA8Mj7hYvkjx2x2Hm5tW2yTnsOnYyPjS69Yu/ue3V/+qy/ufXMf/ujv/8sf+/E220VGzvY21tnwgQ4Vw2MTM7NVTS2Lndsra545efLEtjUbbrpu6ze/+mczZVXUWh6JeeKJxxctXdzZuayluXlwLLb8DY9NtrUuokT59re/27F4sVOvy1esennPHhc2O4pSEVui5nQFMeRQmESHf7RkKGrgn8wBhwVp5borbKG5frOjqGsA3Dn6xS4eUPYqjZNj5hS5Gb1J/sxtomAUniJiWYabfwaj1M4Rsw93BhOa/S+1c1IZEjCT3RlyQUS71WLUMReWf1TNDElvYlXmwi3znqq6lkWLr71+ly2e3lT4f775qNnu+nUbX37p1XfctHvxkuWuYVSw2nR3l4OjDi1949G/X7ys84Ybdx85/oKDpSuvWHrg0KFnnnpJUjP6q4rqp59+4Zrrbzpyoq++YbSyut22p3E74csqPbx6y+47j57q3vP6S2vjbsSavW/sP3myp66+KSlS4zb0Cc9vN7Q1t1S1d64+Njp+9VU777pt1/TIYG1T+2xly+hEWX1TG8WC00A/+q732//055/9XNwn3bqsf9hLKQ3V6T6Bm2/epeDe1RwZi2NDcfQzTSUv0plEjQacafI/2s7VtKCyIHERvZJqzVVcimIRvfyjH3vIRzbBC8lgMk22NELh9hw7N6jEikmrl6JIIXuyc2pS4NDxsaGVjbg+2ZdLX6gUAEg/JRyWz8JIR1w2/9QY5hen5yCgp6cWxTyJnmmyobG+raXZJvlYEKqoHDt7Ro2OxsHi6r6BoV233Hbw0PHzF4Zq65ooqoh/naOzG2Pjg/0DPZ1LOtqXbK+unKmpHXUCpLq8aXzMXMct95Nneg5u2rKhuqZldLR8cqrOqaaJmb6eC2c6m9qNELzf0Ll8mcs4dccjoxNeksC9bkIwOauuKquv9aSBQzznibrJls6RocENa684uPe12oqpjrbWmjonj4emK2sMIptbWg4dPNjaWNfu1vDZUQP8/qm4YcC+T8OVfQcOooOncxTdjmJkKfQtUTuxbkDX9vb1WNDzUocqyKYIWlAFhf8PcUhBaLY5ohZTPUI4x8o+2c4ABXAGmEsBg+bvBXYOLuILzT7j6eRrzizb2R/V3pJuISCng9czWpSpiaXyInUkyJ/NFAkWBQgSp0kYH8bCAVtSDHh5ZdymKuLKwlB8YvpgzBiOUWrX1dvM61WKSqNJ+Yqk8kyiW8vjmERtQ62rms0/GptbDx0+MTzqbnLPgyBe7AiOCyMqbBYYqTRFmmqanBiurHHIYaamoskpe2fc7X6enLlgDp82W9dXV7W5mrmiZsxmzYq0ndz41XpQ7Ad1s2mt+3wMZHXEZU44O6JXESP5UYfNTbd63WOuSVTaFTHr7DqmoiSIs5Uz5bU1DbYux25rB6mq4jovG6mHxmNjl4N+BTVwqgLa6sm2axRxcDbScbB1DYXJFGPn0LetrxxURMkOkEX95liFDSC7ixrMn4V/dhSfRcQF8PItgkodb+niSwOyu0i3KFtOV2j2AVD4XBqdTwbI2QejJY5ko3JOXHSh2ebQWy3wT5HmIuag0oxop+PMRtzVHYcT1IyxsYNPQ4534hKzWNcPxL4i04dKu/ZHR+P4xOj4zMOPPEzM/tcvfXkSo9F+eqajvIaUi10jZdjDyScVPx1HqIjMCgfSzULMRtw7Aouqcg/U4CjO2QaMV5E62NB0OXWCRyMRklzD4DAxqrSrVBmITM+IOXbillmH1LTp+uom6oKy8nFbpKYnZmE+U2b1n4yeKnOiuqxSU6iuKSeh3V8Rh/mqpROvQUf/kKfG6YgwNoHKHMHnK1uOb5lEzhPuUjLOh0R9cReVu8C/+CwcQYCSLq6IngHeNqNL4YvULnX8wwyaMSgYKKfOLhyXolKKFhqB1P54kmdBsnmTI6aU5iw+AjMMtwpgMlgWnEX0nDtgANx5oc9hCdyBDynhQ4SRIulILs2hO70IGCdUaitrpmLRfPb3Pv2HDpY3ty7ySkFVLbkYtSJ/8y2zIhPimelxqEy7ULzcMJpQdty0RtKhKMqTMXMoHfGkXe58rNuMxa6jSpLYPEc1E2exK8+IWKWHbEQEWCmgO3VikSxYYWbSRdzyTHnrDeImcgqjca2IMtfZEOkEz1Q48maO5Qy9M27iMsadIfSr62vwiJNLsTAR1A6OYapq4viek/yZgHyy4x9jZ+BIZT4W+ueIcl+QAp8CLAclNKLxXhplAeSCpC79vCyDSqjIQDSfiVvQd655leYEshS4NBu4omXmMzAZdZ/YBVhprEgiFTXzsdCcY8ZExAzALiIGDrOTRLEuqEJlWiaaHG1uMlKcJDJB4km/Vv4cWorLsBx/G3ajzuzETGiRnJrUmxORkzO4MFScZiTgzccjWWwUe4hx5ojzvQ4L1VCX26HoSp+6yqGxfuI+dotgqVAfEa7uoDPDygwRW/LIv/RHtaqwxKV1KOlXkmrYC/IMIT82MVxTZ88zzWgcAKJXhTLFAY3ZzIRWXT5tVMAdjzvUTs/0Z/KCJMeDT+NYZmVbe6ujrRplbDdVjnTrTiZahmfLLrsznQv/UkcBU0BmB/9M+QyQ3UUQz8JfUGlo4c6OghOKTAuAwqfUUXn1dRtLv0vdYhaRMwZhL2w/EaMUsjQFbmwAJ3TMYOyC7STGzc5RciIxIZ1vD2IxGaYYEhQoZUdcbRFDt7GZSX/Dt+16x1XbN73+yovbt21sbqgdGxnyYq9sSBciIDRJIQIhM1Vbb7tN3KRAxWgmZ50T/+BCI8CpSSfQxytmxstn3NZeff+9u2emBly08a4H3n3i6NGmxqbmptr+vtOzU2OVxJv7uWcrLUFVlPmzRmqFadpIUDkxH2YMZrHEXzZZOevFI7ugPM8sW7cYg5mqmJ3addPOG27a/sbrr89M2hcBB3dPTLnxJN6zx8bk/ky/UUf5rJv37KDyQhzJGdtH4iBgDE8rl3UucR2AGbIz2iGbg6AhDPQbiJPpWdgRWGIyGQuq5urwyeBvJsuL0moSW2iRhqBssk+KOhdagBWOBXFLEyncpY65KVWpV+HOuRb8xD9nkzEotXNQ4VOkwCE6WzmFSrBgOEP7vNZfBAnNwNmR82VnNCSSHcVnTrZqdtLk11/17HjF9Gh91dSyxc31VdMt9ZV33/7O0YHzDVVlNeXTDdVlQ73n66orFrXU11bPtDTXbtqwyoNvVeXYxVU7PVXl43VVk3U1khqtq5yUTpODbmN9g72nO9vd9DVVWTbe2d54/mx3JemLvys9oTS9qLmutbF2cnTA+LG6bKLOvKd8tsmSf3WFE5/OpTW450y3PjECvaa62fqaWcfiCeKG2rK2lrrrr9nRXDMzMXJ+6aK6sqlB/rXmSlOeMDnXWD3bUDPb2mCoO11bPr5qufl9vbmUFlZVU+76MQ7c56KQuvrq++67x+0j4yM2BozSoaKMt2VwJ1ohu0+GOztyNWX329oZIHhz3oQQKDGlqWXvSyurNJcCh+xZGlSk+raY8PwHGLSUP2AL0QV5lH4WGXMU+ZGIkBAxx+UvlJtEZDgkm4FzIQtFSfbMmfLMoZLKKOV0JFXlNryyycSmM6rz/Knjk8MXXJ5gbejo/tfOnTqyfdP6rqVtV65b9Z6H7nX7nbdfVq1aUlE+esuunVu3rqOuuWrbuvf9yL2L2+qu2rb2iq5FLQ0VNRXju3ddt7i19oF7bmmISctwS718hssmR+sqZtqb6lZ3La2cmXjo3jtWdy1Z1tF63523VM9OvOO6qxa3NHzkAz+6dcPqRr0xlp2dXreqq6Ol8QM/8q725vqNa1ftuvHq8eHeu257x8rO9obq8jtuuWlmvH9y5HxLw6ysH7r/jmuv2lRTOXXHbe8om+pf2la/fnVnc13Zhz/07q0bV61Y1mYuZi3Uqeg777zDOWPX6m7ctMEBa7PDXm/p9fdOTk3kLc/oE4R9q44J9TJVf4gtVlGPwDIDlbLBgri5RnLtFHVUmlHBDNlRpO8zJ55TWJBs8XlZBs155MjZnfEuzS+7s3+R4gIHkSk6mMyL2cENjH9hcqycIM/SAuegApKjNAtdrDGaQVfl7NTIUH/n0vbFba0Dfb2v73np/ns8p7375puuX9LearPIvjdeXdzeumP75nfefIPBal2tFxT6aqvLBwfOnTx2aGJicGlH8/XXXbX5ytU7r91+3dXb3nxjT/+Fs4uaGwb7zttdMjVOX9nX4hqp1SuXtrcN9PVs3LB+fHTowvlzO7ZuIvdWLV86ONA7MjzY7fmH4SGj7IqZ6WVLFtPrL1nc3tnhTrK6rs7FRw4c3rL5ymWdi8+dPdnUSGLODA/0lk2PdR8/bFfVyq6l9vjdc/dtxsGNDVV37751fHTw6m2b97zywplTJ8bHho0v/d19953Hjh8xUPjgBx92dwNy8dTgteRsUDIJ0KA8U0ox7kt9CgAMlHmIT6Z5DLPmdYjZswAuYHKVXa6OgBU5BjZvxSfHKk2z1F15zc5N+RtckZAkis+McWHHyhBIPJdmqgRgmiiYb8ZaMM5CDC035g0+rW7XNcQdGh4i0tWbHZdR1wmuHncRVQRQ7VkklExtRZkZcflUhUsrJswdbHlwYbZB13QofaZDqxOek65xrarVvVlkty1ggv5ylN6vutFDgaPTDeu23Dgy3bTvWH/XumtOXZh+7sU3u88NjU5V9A4M0ym6sPTs8f59b3aXl7fUN6wcHms+ea68+8zsqQsVda1rzvXPHj01dKpn5Fyvrt3QpOXU2bHBqv6dt93y5rEzfWPlm66+5uX9bwxOjixd3XXiXG9tc9swVGvLq1rLDp3eP1tX3rlm1fnB83sP7h2dHpqtnq5pqbzh1hu6e091rFhysPvkIPVmTW3f2PiGbdveOHx4xuXTLU3dvTNVTZ2Lu7Y+/+LBZStXNC1qfvp73x8cLZuYbRiZrvrenj0K3Lp49dETtgCMmphXzNR7vn50yKpu4+L2K156af/4ROWKlRsOHOym7TWCxajTE1MmvwjdUFUZbzJWVI2PGYIbsmANl4c1TIy7DwLZMTNVGt1Duj3UrhMv75TXxAZrG1CoO1KVxeaBqbgs2LDBnE/nl3iMaFPVcfuV5sFHWrlLLPgvO7KduRBkwY6ZwQpeLGIVPtlR/pM//a4Mmu0i0dL4RWSOPPsuQvnkBqcrz2Kv4HIwgZDbfdJIVCtUgJhlxm7zaACmzLHzrabG6jZQEXmOTtFfhtYmjVdD+iKJl17nG3FkkvEBFviMxBIRMDXi0+VEVllo9Wtqq1waaBDhdh19onvC9H0uhCkbVofVo2OTtY1NptT9AyN0P9Tp7pNx8mjWtJk2cnK81luxEzZqVEzXjh44cMgTdVJ2zb5LzmwQgZKLG1x+dPLkCW6X8rorb9euXa+9tqdjaadLi/bseVXrcSeZG0XtDTVRcUcV0rkN1ODbNeSyQxXCr6G+VdFc+6PI7mNKl6+EVtyLREqqIC5dam5qYeOzoG2Zx2Wmmpoa+/r6PdvgosqennMbN26iYuOP5HS32JDcsJuX2xar8dEJU9u0iTlubiBJaqstU9GkKHcoJVyKY4LoIikcZ6EVJjJiT1mzSI5cp4hMpaG8Qfa0cMXDaohPwAxHNhmeOzvYqYLm7IupRYpzJgMLWmAqr05X34AqAgrQnFmpDSZ/ZoQKO3sKnc9uLjX+Tu1SJaYunZZOj+M0z7iFn+iavbua1DBmo+7/qfVkd5l1lTbPXVdV2FhRW13pfmE7fRtcFkbxUmUxZqq8tqrB29jsuOyyoq6maspuUMp0i0iU6lERbia2xW50JKrB9k1kCq4m+N02NG4DnqVIZylGXMDknhwZ0x/F09xUSxPTk55NmqTdoSiPHRgVJA3eW917jmK1pbaqtXymoa6mbWIUMs1l0w3ti1Y01LUfP9qzbvXW0eGZvt7xY4d7hvrtvqufHq9hV5Y1EngtDZ2zU7RINAeLysOzYWayZnaqTmpTk8OTE5MuXoJ8nbftQ3k50+ilQ+1jOFrI+Og4/9GREYTQKalzXY+BBG0a6lGKuUIRk43J3i3U1L+xJ8vm69iQANjlaLRcVCmC1AKyk8LYGK+5rkYDRAD8Zse0W7DVlIuho++jXoguMm6VMFEw8tXTSdGgAuur91zRiSmDlVV0wRsFM2QfdjalPDbvd/E3p3Dxe951Uc2U4xepl2YJOPuz5yXZXAI5FlvDYgvFC6XwZVSHs3EFK63iJP0ixbSLT+s8GFdTiwmrcICH20YxGDdlSmVZC80MbqMtTyTCxs6Wh1C1UGnmL8gd6W78dlFRaGBmhtSSg5aqihTRmOEAE+MxrdxlYOSrjHgS1hZgpsaGly3zTlLZ2OSoJiJflePSI1WIt8tnrVlO0Gm6M6ttUZMLF8ji06fPLVmyrL9vyF3atj6RKR4OkBSR5eYz9/J1da0cHRvzVonrIyl6hofGhga9phUPzetDXU/XEpedhy7eZVYIRRgn8R+b5Soq4/IzxxfyAFJBiHw0JHexopkQ4+1n5eKjQ2hudvucp7CJfBfQeYczDi2lTgnZw62Y7BCPsX1ktqmxsu/COXei8bPKgNpaMHYkKEIcVNtjPabU3E2N6FNNhSw5MtRWLyu5Uo7hngEtoZJqFlYpx2CAJKHmWCO+S/hEFeSA7J9tnhkmfwIoQudSueQnGFS0HHMB9NumNY9lATsnNdFdUGaOHDHnRTOi9vXGdvVaIBka7L/ttl1rVq/ou9BzxaquJUvauo8fiaZcrk2HDCN1pmieXQU4NkKIcKMm4ap7NBGxf57s4kNtiZNCHli4iUvkoEGiev7QwEolBCaYoL6udtvWbadPn9x9++3bt2892d19y03XrnK3bfmMq76RnIpe9dtTglPJksDEzuTxoX/2kx/p6Fj08osv3Hf/A6tWdZ46ddjejpaWuqGhXg/AmqIbjLjabWTELZttH/rQu/ftf4Xb3dz33X/nuvUrX33thY4OL4fYWjrY3tHU3+fORIk7hElcDjQ2VY2N91dWKXWsI6S7IXQmsQ6lHFFp0Q1M2A29zKSvfdGkN+vL4qa+1hb7A99/4UIvnC3H+6OsswtVXHewhczU2Ko8UuVmO9pfOt6J62/Yetedt+/d+6oL8E2cyFEDCj1N6iK0Z6NLDW2kpaXhoYfuVSn9vYO0s62tjZs3rj/RfSzobMlL14e1iNTo3p1biPUw9CJpCe9c0QW3FJ+ZRSJC1M5beCx7ZoBSsFKf7A5ZfalvkdwCh09Jp3Yzp7/Nn2zcyRTwRZpGfnFIJO2lHR8du/mmG7dtdbDm5MDghSOH9h85tG+RnWW2rDfqQ4yQPLRaMTPpYroJ9uzscPuimsZGS0Bu+C1rb3WfN9YanpzsH+o/DZJuyJCOChBKKobWGg4GmkZ4qM64Bc5LP87x7Nix49U9r7no+vDhg+3tbW+8+bpBqttZCVjjTsKSoK7SB+gBZh2Rq/WGWCA5MFBVM7qsq/Ho8VcXd3qwZaqhaapv4ETf4LHq2vHGZov459esWzw6ce7Yidc2b11+8vTelrbyc70HF7VXHTn26oWBo4vaK0fHzzS1zlTXjbe0VTQvml26rKFjSbUbKWfK++qbPJTdZHxpj7VbalWFNzQ09WhaDXUGIXffc5fnInrOn0uX0VV6XZzW8/TpU0ePed/jcHWNWvAO5ZwucwAADyhJREFUBBI1LGprdjdgC61sbY1BlEHqkqUeha+jNjuw7/VjR/YRrPr/2uqK5kZJVnR4YCW9djit+yqfFrS8s+Pcme6y6fG66rKxoYGDb+5tqq9paqhvrK9pbW5qjOemsMrcBMBcLJuioue+59mx8F/AEj6LoMJRxL3UEcPhbBYkxDP7lNo8L00iRyexMoPmz2yLaxNxWLPaXlVLU+tN73gnyBee+96EeQpG6+93TMwoakXXsqVLOpa0k0b3neh+7dZbd9xw48by8v53v+e2665bNzN7Ydu2FS2ts42N0w9/4L4LFw4/+NCuVat0dzM7dmxbuXLlLbfcYm70Uz/1Uz09Z90x66kGeGahjndxp8nHnj173E5a39icJhA1D7//g+7UBXnrrlsWd7Tdsfs2+nWxCB8Pq5m+7N9/4Lbdd5SVD/7ef/o/3v/w3bfcusOmu//pY++vbxjfcdWqW2/b0bmsdu269itWtz7/wmMrVrYs72p64MFblq2oO3nmjfd94O6tO1becvtVW7avqG0Yv/eBm6Zme+66d6fP9iVVd9934+TMuVtvv+rKTUu3bdvu7mbNDFMajbzvfe9zn6MidHV1rVu3Tn/thB2U4jrV8dGtWze74Z++k9hav2HtHXfcvryrc8uWTavXrNq4ccOy5Uttfbr5ne9wb/emzVcC9pbkFatWvvD88+9+17uuuerqTVdu3LpliwHue979EGmL57Zu2bTz2usa6uo3bdxw5PAhQvdn/5ePj44M3njDtdu3baKbe/Deu65YtWJl13J0NnbXevHoXEXHYCfmc2q5YAyOgmFKOaHghwUOn2h+KWThEwyaMyi88mdpNtmdIbkzfgVA4DtveBbpZHj9QzyAEuOWas9y1tY29p+PO6Rbmts8g37n7ru8LvXmGweN3q67+jr2qVOHdlx95YYNK/Yf3LN9+4blXe2f//wf6xB/5L33f/8HT29Yv/Jczwk3r8+Wj/X0nKitL3NFvAomGlXwV77yleXLV3iXw8FiuSt5dPqzs1esWrN//8HEAHU33Xzb333jMa/PGFD1XRj8sQ995GT3aTuGXnzhJUNDJz1he83V17vn9AcvYujBm5wA2rV708Zt33vO1fGzFy6QqkOPfPgnjh497lQQ/aBbj5Xx/vsefOGFH7jw9o039u3fd1DVbd605R03vvPgwcPVVfXLOrs8Tmd2/8LzP7AP//jx7lWrVl9//Y3Hjp2gDXj++ReMTb1stmvXrYsXL9m7982dO29oamzeuHGj56M049zY3MK8bftWL82ZqXDjPwBnz5529E+3cM8997hBHFtTWUjzve9999NPP42/e3rOGwq/8+Zbenp6vTb94g9e9hCZW6TPnT1/9kwPImzffhV7/borX3vt9fb2juGBfiOYVStXDA0OLGptXrJ48VNPfddTf5u3bIrrb1LtG6EWFR1eb2XQXO+ZDXJQqc8Cd+bOH8Kjldfu3ETGgFhgS4gnk3MqbFkW/pkttSGGuzSIOxvv98QKsgO95a76dh34QU+vUtZ4Qalz2bKDBw6fOnnae6xOMrz4/ZdtiTQNc/e7mQG667/IDy8QwM2gShamJu4F1r9s2LCBXORz8mT/8ePH6XHc3I4vt2/fsWrVqiOHjwXR4rqOCp2dC95PHD+BA0SH1ZEj3Z1LV4yOTeO2V1/d296x5NjRbtdaO4DmQvgli5ceP3by2JFjW7dt9+DMiy/u6+7udcVxTdWixobOqcnasdGKU9399ilfsWpzfV3HwQOnZmdqz5we7GhfKai/b3LNFVt6e0YPHTy95+WDnUvWtrV2nTh+fmKMDqu+r9ep0daxkbKes0bYZkvlvT3DunQl7ehY7Fr4kye79ePHjx978cUfPPDAg889+xxFUkzVm1tWrFjp6u6TJ8+sXLkKH3uuaO/eNxSQp1tnT5zoXrrUo4blZ8+6gLfuxRdfAtPY2DQSbwCFpgxtMaVOw2jowP4jyrtp09ZXXnl1ZHhEZ2MHIJgVK67oOXOWHm3lqisee+yxO++8q79/wEt4mtPX/vbr8IyBVFwKbGiRdF5YIabEc6yCtty53smLgk8K5sk+7MKncGT2KwAuOlxgK8XMXqAFFJxaRC51AM6ReeaIuDN7Zp8MzCc7Guq8OcnEbjYtLxKvsALugethQz3zULdLW7omJ8CbaJrphp6/qpI2iB4OZjHOik24U41OV4yPtTa37L7z9hPHjj/z3LO33XLruQsThw8eqayuaqhrGBufpPOra6gPnZ87ZSo9wj5Cu9LQVB+b6J2DsyHTVCtdaVkfg7+4PaWiOtSoDDWt++xhyNPGXx0un5rq5jMnT3VRfw55ZKzcW2THuk90dXYpyagnXFubYtHI81xazmBfY1390PhIMw3r1LRT8lMzsx1ti+2MnrT2YG9/XYM3TPSnJsXwaWppdPk0fSR1WK33XGxjnhqjPxoY6jNBvHbnNbpWbBEzkVQWmwYvnO91xBPFQuNgw2hsYrX7yYlTuvnYWsXH4oh90oaLaIIC1VUOk5qNxelkPRhorZQ0pd9gUAC4XpvcpfSIYUYV3bsjohOu4n/Pe3/0O9/5TvfpM2fO9rR1LI5DsyZisQErpi5RV57ofevlNImQYRlfCWW4Czt5vL2VyX5pWOWOazbgzmxyWgXzXgqdfUoBCjeE5HGpQRkCQMpGfsHbNjjgVs1OaflRPlugiGvPpS0FqxSNyDU1Q31qCtBQRuNZ3WyHbllFXUVZfVW1Sp2pqWsd7J+oqGp0qDiOPlpUTIxFESAXbOTGmNimhuHTiMkBD5e02y+Z5suVNNmmzp7VtKAl97TdJ8ZT5qU2A2kXwehjFEwcIllxse7ltbzZ0bGZC70WTiitLNMYrVKwU0kC83rBzBBBNj7tpPvoCCVrbPSjEPNK29DI5MiQKVBF7/kBrWF4aHx41A7rqp7efnZvv8fEsJAJt7OkVBaxK74hmuLkq3v2OAyQ7nHEvtUuEHFZs/kzqhEJKK+RaEJZYLD5Z7eyhwALwUy14vU9q6OUHdz211cjhI2sCm4zKToouJPUU7pA+gPnU6mdxjyA1NS5rIuu/5nnXtA7eolCTaEuZpNzKPmB+5l2ACYWllReZh4OnzDJrFLqmX0uZ+dELg2NM0nRQ6c+WrAiKVsUNTH+pRGyj+SKFEEypZ8ZhicHxZzCZGiy0Ng8ZoIUyJY1k5mh0Yzr0GNLOlM24yWQUFuKTplH+ecz3PTacZG25cLp8+fPu1RDECE3VTnswanYB1+B9DSbIaVJUAQPuStrywChe4eiAYT1uhhvQEko/ITKVKmzCE+ZInsgnElRW9VMF0VdaWGB0tXGYO3I0VZ7MUOxSPVCo0XXHe64yGTKYi2VZ2wQrSapLUQE64iDV1zwXl2R1m9mST5a4dDRVI3DDYbUwxaLSe5q+6eryX5NJFaAcECccPcXB/Atg6X+xLAReaUM0dhVTVsUZSmEFXcOTeNFQGnTX6oQZSQm0nQnbvdEBhpTyhZuZKEXzrxn3jY4bPgRSkLI5denxGTQNKR6Ms7OZKKxU56RF5M/A/qt5nL+orwVcO5rbqCQ0gwLa2YbH7xthEs9c9wiY58FrhxxLELlpsY7VWbTWiJSdOYmT+rO5YJR/hCoiWMwqp4xmnUstaFDtfMYMjXdmZzyhkoDkbd0WaOXVai3h0amqhocjiNF6Nulp4cjmGVn7ZjWiTAtp973viDWdPQjZYGLcF9VqPptNY6K55Ps3CYVwDArMMNnFRMzA8C139geDD6Uf+X2OzsrCi3Lg+gV77KVW9J2GkRhQ+GrXWo01sKcLYpLvEmmtHGbBAjgUNbQuNLjGtNV+yTESDUkwe5EuE7W2Q9XR5Fz9F7KgpDWkh1eMpwIDY9VDTv0UzeLvdAOBaLByTeKHfsiYqSIqS1j5sWLGCzmOXEg5EYnRpcUpQjK294Qh5mgERxPBT3hIFVjVGBVtT0P0UiiDchbDQS94ltctC4ZIqq18P2nMyGrch45TZllc7ksouaSAVbAFJ58sv/FULI4evRZ/S9uCHESRXNMiIoiBFWIFlRIBYtYhi7x0sWUarDjt8oekRDPwUASJ6XiWCS6qC1a0+jDsRdtspihzAoyRi8YsVA25DR5Om4xKrahyU+UaISipTM9ISHi8KebiyUSI13LqlCOYVZkh+Wt2MMy0o6fUFkTwAQkxONwCTSiRcboIhbSKqtt2Ii6t9Mt7LiGJJZh4O+4Xhz5qIwFblcuKL/COiMFTIFivOu5HP/jtAHMgyX071ovVkUTjQY609pGrOhKEb8Gal4PkUIIFEVXmdBLmh9CWsbR9PmH2AlhoKQxPAARjTDoTohbxI9CxT7qkAjEgsFPwAvVZMDHsSroQFyOMbiM0GRQ2G+EJs/syEHZZ4F7AUARejn/kOpFQlGIlPHloIvkOEphuHPEnFTxyWEgF1WdTOKMGFNh1BCO2AuRYgKkxYfMtNZUVU9yo7WbCFBT7x+kQSVk0gfZZjFpwBknx8iJspp6g1fJq36Hy6L7C0SCrjHkx4QxATL48vBw6lm1kNgPpaIs8lhuZlIFBALB6CoyWpGyqNDIV2RIakshJNSsoHjYSnZz7WIKB2OBaFEyRRVNyLgVtljG2Dq64ugSo22m5mQyHbQAlj/rm1rE9e6Sg57jI+mIdtKB43/FCjFomuN2lJjBVGsMBhHRnQdDhDhDEwwOAz0HHjWUzAJHAYFp9lZrNUou5bG1H51lJJW438BG/xi+uMoZjjCCV3AueRI8yU9bU+BoWpboUr/GN5nMNolEUfVgeBc8wD+7C0cRlKP/4+3Ul80nLRUmZ3a5JABktNgFTIo316QyTjk0+Udh0RwwivrHRJ3hSw0zpRaiKBVJ1g63BWRUihlFzGEMzvgY8asbm8mDznbZ4ZA4Ej5qX4RVb6nFfCuSDhYzoCQYfenugq+Cub2AhqpzZ6CjXViAh0wWf3GmN/LFbpIlgAWlOYF6IXGDL5UWl5MfiT6RZlRSWqG2WuiTLJnzTHhkH/CpZnFsRFTwCAypJM2Ikl880uQY6UszRqzaT9wMmnoPYhYBcX9wqw0GWouUkkhUbGHBs9HYIEShKzvMnVgyJCw04txJ/Bj5YNooYBQKBhWxKUk28caDBIGkYVVVdTz3CJ/UV1ipDl2yV0JTsUAF/pFNMokzgzWzyWHcAfxPYf5fyub6pDH8sY0AAAAASUVORK5CYII=",
      "text/plain": [
       "224×224 Array{RGB4{N0f8},2} with eltype RGB4{Normed{UInt8,8}}:\n",
       " RGB4{N0f8}(0.788,0.765,0.671)  …  RGB4{N0f8}(0.769,0.745,0.659)\n",
       " RGB4{N0f8}(0.792,0.769,0.675)     RGB4{N0f8}(0.769,0.745,0.659)\n",
       " RGB4{N0f8}(0.812,0.788,0.694)     RGB4{N0f8}(0.769,0.745,0.659)\n",
       " RGB4{N0f8}(0.816,0.792,0.698)     RGB4{N0f8}(0.769,0.745,0.659)\n",
       " RGB4{N0f8}(0.796,0.773,0.678)     RGB4{N0f8}(0.769,0.745,0.659)\n",
       " RGB4{N0f8}(0.796,0.773,0.678)  …  RGB4{N0f8}(0.769,0.745,0.659)\n",
       " RGB4{N0f8}(0.8,0.776,0.682)       RGB4{N0f8}(0.769,0.745,0.659)\n",
       " RGB4{N0f8}(0.776,0.753,0.659)     RGB4{N0f8}(0.769,0.745,0.659)\n",
       " RGB4{N0f8}(0.773,0.749,0.655)     RGB4{N0f8}(0.757,0.733,0.647)\n",
       " RGB4{N0f8}(0.773,0.749,0.655)     RGB4{N0f8}(0.757,0.733,0.647)\n",
       " RGB4{N0f8}(0.773,0.749,0.655)  …  RGB4{N0f8}(0.753,0.729,0.643)\n",
       " RGB4{N0f8}(0.773,0.749,0.655)     RGB4{N0f8}(0.753,0.729,0.643)\n",
       " RGB4{N0f8}(0.773,0.749,0.655)     RGB4{N0f8}(0.757,0.733,0.647)\n",
       " ⋮                              ⋱                               \n",
       " RGB4{N0f8}(0.369,0.388,0.294)     RGB4{N0f8}(0.365,0.384,0.29) \n",
       " RGB4{N0f8}(0.365,0.384,0.29)      RGB4{N0f8}(0.369,0.388,0.294)\n",
       " RGB4{N0f8}(0.365,0.384,0.29)      RGB4{N0f8}(0.369,0.388,0.294)\n",
       " RGB4{N0f8}(0.365,0.384,0.29)   …  RGB4{N0f8}(0.369,0.388,0.294)\n",
       " RGB4{N0f8}(0.373,0.392,0.298)     RGB4{N0f8}(0.376,0.396,0.302)\n",
       " RGB4{N0f8}(0.361,0.38,0.286)      RGB4{N0f8}(0.376,0.396,0.302)\n",
       " RGB4{N0f8}(0.357,0.376,0.282)     RGB4{N0f8}(0.376,0.396,0.302)\n",
       " RGB4{N0f8}(0.369,0.388,0.294)     RGB4{N0f8}(0.373,0.392,0.298)\n",
       " RGB4{N0f8}(0.373,0.392,0.298)  …  RGB4{N0f8}(0.373,0.392,0.298)\n",
       " RGB4{N0f8}(0.361,0.38,0.286)      RGB4{N0f8}(0.369,0.388,0.294)\n",
       " RGB4{N0f8}(0.353,0.373,0.278)     RGB4{N0f8}(0.369,0.388,0.294)\n",
       " RGB4{N0f8}(0.357,0.376,0.282)     RGB4{N0f8}(0.369,0.388,0.294)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "load(\"judge-books/images/$(data[1][2])\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now, we are going to create a preprocessing function that will convert the image into grayscale and a floating vector of that (containing all the information) to be passed into the neural network."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "preprocess (generic function with 1 method)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "preprocess(img) = vec(Float64.(Gray.(load(\"judge-books/images/$(img[2])\"))))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Creating a function that allows us to create a batch so that we can easily group our data together:<br>\n",
    "Each image will have its floating vector and an array with its onehot label associated to itself."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "create_batch (generic function with 1 method)"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "function create_batch(r)\n",
    "    xs = [preprocess(img) for img in data[r]]\n",
    "    ys = [Flux.onehot(img[6]+1, 1:30) for img in data[r]]\n",
    "    return (Flux.batch(xs), Flux.batch(ys))\n",
    "end"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Using the function above to create a batch of the first 4000 images in the dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "([0.7607843137254902 0.7568627450980392 … 0.0 0.4; 0.7647058823529411 0.6862745098039216 … 0.0 0.4; … ; 0.37254901960784315 0.8 … 0.0 0.6823529411764706; 0.37254901960784315 0.9607843137254902 … 0.0 0.6823529411764706], Bool[0 0 … 0 0; 1 0 … 0 0; … ; 0 0 … 0 0; 0 0 … 0 0])"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "trainbatch = create_batch(1:4000)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Checking out the structure of the floating vectors in the batch of 4000 images.<br>\n",
    "As you can see, the array has a size of 50176 items for each of the 4000 images. <br>\n",
    "I'll be keeping a note of that number as that will determine the number of inputs we pass in to our neural network later."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "50176×4000 Array{Float64,2}:\n",
       " 0.760784  0.756863  0.615686  …  0.0313725  0.807843  0.0  0.4     \n",
       " 0.764706  0.686275  0.462745     0.0313725  0.780392  0.0  0.4     \n",
       " 0.784314  0.709804  0.498039     0.0313725  0.792157  0.0  0.4     \n",
       " 0.788235  0.701961  0.537255     0.0313725  0.807843  0.0  0.4     \n",
       " 0.768627  0.694118  0.513725     0.0313725  0.803922  0.0  0.4     \n",
       " 0.768627  0.709804  0.529412  …  0.0313725  0.803922  0.0  0.4     \n",
       " 0.772549  0.686275  0.521569     0.0313725  0.811765  0.0  0.4     \n",
       " 0.74902   0.705882  0.517647     0.0313725  0.803922  0.0  0.4     \n",
       " 0.745098  0.694118  0.541176     0.0313725  0.8       0.0  0.403922\n",
       " 0.745098  0.694118  0.533333     0.0313725  0.823529  0.0  0.403922\n",
       " 0.745098  0.694118  0.529412  …  0.0313725  0.843137  0.0  0.407843\n",
       " 0.745098  0.694118  0.521569     0.0313725  0.541176  0.0  0.407843\n",
       " 0.745098  0.694118  0.521569     0.0313725  0.145098  0.0  0.407843\n",
       " ⋮                             ⋱                                    \n",
       " 0.368627  0.866667  0.231373     0.27451    0.584314  0.0  0.705882\n",
       " 0.372549  0.862745  0.227451  …  0.192157   0.592157  0.0  0.690196\n",
       " 0.372549  0.858824  0.227451     0.180392   0.584314  0.0  0.670588\n",
       " 0.372549  0.862745  0.231373     0.172549   0.603922  0.0  0.694118\n",
       " 0.380392  0.85098   0.219608     0.141176   0.592157  0.0  0.686275\n",
       " 0.380392  0.854902  0.227451     0.156863   0.592157  0.0  0.694118\n",
       " 0.380392  0.815686  0.219608  …  0.2        0.592157  0.0  0.690196\n",
       " 0.376471  0.847059  0.223529     0.27451    0.596078  0.0  0.709804\n",
       " 0.376471  0.835294  0.223529     0.176471   0.596078  0.0  0.713725\n",
       " 0.372549  0.847059  0.207843     0.0901961  0.6       0.0  0.717647\n",
       " 0.372549  0.8       0.223529     0.0823529  0.6       0.0  0.682353\n",
       " 0.372549  0.960784  0.278431  …  0.113725   0.6       0.0  0.682353"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "trainbatch[1]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Checking out the structure of the array of onehot labels in the batch of 4000 images. <br>\n",
    "Each image has 30 items associated to it, with 29 of them being 0 and 1 of them being 1. <br>\n",
    "The index of the one and only 1 represents the category of the book, except that number needs to be subtracted by 1 because the categories range from 0-29, not 1-30."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "30×4000 Flux.OneHotMatrix{Array{Flux.OneHotVector,1}}:\n",
       " 0  0  0  0  0  0  0  0  0  0  0  0  0  …  1  0  0  0  0  0  0  0  0  0  0  0\n",
       " 1  0  0  0  0  0  0  0  0  0  0  0  0     0  0  0  0  0  0  0  0  0  0  0  0\n",
       " 0  0  0  0  0  0  0  0  0  0  0  1  0     0  0  0  1  0  0  0  0  0  0  0  0\n",
       " 0  0  0  0  0  0  0  0  0  0  0  0  0     0  0  0  0  0  0  0  0  0  0  0  1\n",
       " 0  1  0  0  0  0  0  0  0  0  0  0  0     0  0  0  0  0  0  0  0  0  0  0  0\n",
       " 0  0  0  0  0  0  0  0  0  0  0  0  0  …  0  0  0  0  0  0  0  0  0  0  0  0\n",
       " 0  0  0  0  0  0  0  0  0  0  0  0  0     0  0  0  0  0  0  0  0  0  0  0  0\n",
       " 0  0  0  0  0  0  0  0  0  0  0  0  1     0  0  0  0  0  0  1  0  0  0  0  0\n",
       " 0  0  0  0  0  0  0  0  0  0  0  0  0     0  0  0  0  1  0  0  0  0  0  0  0\n",
       " 0  0  0  1  0  0  0  0  0  0  0  0  0     0  0  0  0  0  0  0  0  0  1  0  0\n",
       " 0  0  1  0  0  0  0  0  0  0  0  0  0  …  0  0  0  0  0  0  0  0  0  0  0  0\n",
       " 0  0  0  0  0  1  0  0  0  0  0  0  0     0  0  0  0  0  1  0  0  0  0  0  0\n",
       " 0  0  0  0  0  0  0  0  0  0  0  0  0     0  0  0  0  0  0  0  0  0  0  0  0\n",
       " ⋮              ⋮              ⋮        ⋱        ⋮              ⋮            \n",
       " 0  0  0  0  0  0  0  0  0  0  0  0  0     0  0  0  0  0  0  0  0  0  0  0  0\n",
       " 0  0  0  0  0  0  0  0  0  0  0  0  0     0  0  0  0  0  0  0  0  0  0  0  0\n",
       " 0  0  0  0  0  0  0  0  0  0  0  0  0  …  0  0  0  0  0  0  0  0  0  0  0  0\n",
       " 0  0  0  0  0  0  0  0  0  0  0  0  0     0  0  0  0  0  0  0  0  0  0  0  0\n",
       " 0  0  0  0  0  0  0  0  0  0  0  0  0     0  1  0  0  0  0  0  0  0  0  1  0\n",
       " 0  0  0  0  0  0  0  1  0  0  0  0  0     0  0  0  0  0  0  0  0  0  0  0  0\n",
       " 0  0  0  0  0  0  0  0  0  0  0  0  0     0  0  0  0  0  0  0  0  0  0  0  0\n",
       " 0  0  0  0  0  0  0  0  0  0  0  0  0  …  0  0  0  0  0  0  0  0  0  0  0  0\n",
       " 0  0  0  0  1  0  0  0  0  0  0  0  0     0  0  1  0  0  0  0  0  0  0  0  0\n",
       " 0  0  0  0  0  0  0  0  0  0  0  0  0     0  0  0  0  0  0  0  0  0  0  0  0\n",
       " 0  0  0  0  0  0  0  0  0  0  0  0  0     0  0  0  0  0  0  0  0  0  0  0  0\n",
       " 0  0  0  0  0  0  0  0  0  1  1  0  0     0  0  0  0  0  0  0  1  0  0  0  0"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "trainbatch[2]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now, we will be creating the model containing the neural network. As we have 50176 items associated with each image, that number will be the input. As we would like to know which category the book most likely belongs to, and the fact that there are 30 categories from 0-29, 30 is the number of outputs we'll be needing today. Now onto the architecture of the model, here we will be using the reLU activation function for the first Dense layer and an identify function for the third, before ending with the softmax output function."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Chain(Dense(50176, 512, relu), Dense(512, 64), Dense(64, 30), softmax)"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model = Chain(Dense(50176, 512, relu), Dense(512, 64), Dense(64, 30, identity), softmax)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We are using the cross-entropy loss function."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "loss (generic function with 1 method)"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "loss(x,y) = Flux.crossentropy(model(x), y)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We are using the Adam optimiser with a learning rate of 0.0001"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "ADAM(0.0001, (0.9, 0.999), IdDict{Any,Any}())"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "opt = ADAM(0.0001)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Setting up the accuracy similar to that of [this](https://github.com/FluxML/model-zoo/blob/master/vision/mnist/mlp.jl#L23)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "accuracy (generic function with 1 method)"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "accuracy(x,y) = mean(Flux.onecold(model(x)).== Flux.onecold(y))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Setting up the callback function that will be used when we are training the function to show the present loss and accuracy."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "cb (generic function with 1 method)"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cb() = @show(loss(trainbatch...), accuracy(trainbatch...))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "loss(trainbatch...) = 10.965167f0\n",
      "accuracy(trainbatch...) = 0.08525\n",
      " 51.497023 seconds (1.04 G allocations: 22.411 GiB, 7.80% gc time)\n"
     ]
    }
   ],
   "source": [
    "@time Flux.train!(loss, params(model), Iterators.repeated(trainbatch, 1), opt; cb = Flux.throttle(cb, 1))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now we can train the model. Here, we'll be using the `Iterators.repeated` function to repeat the training process 360 times as well as using `Flux.throttle` to print the loss and accuracy every 60 seconds."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "loss(trainbatch...) = 17.33493f0\n",
      "accuracy(trainbatch...) = 0.058\n",
      "loss(trainbatch...) = 9.069943f0\n",
      "accuracy(trainbatch...) = 0.0485\n",
      "loss(trainbatch...) = 5.563376f0\n",
      "accuracy(trainbatch...) = 0.0355\n",
      "loss(trainbatch...) = 4.8754497f0\n",
      "accuracy(trainbatch...) = 0.0385\n",
      "loss(trainbatch...) = 4.1969633f0\n",
      "accuracy(trainbatch...) = 0.05425\n",
      "loss(trainbatch...) = 3.6376808f0\n",
      "accuracy(trainbatch...) = 0.0425\n",
      "loss(trainbatch...) = 3.5965037f0\n",
      "accuracy(trainbatch...) = 0.0345\n",
      "loss(trainbatch...) = 3.3209286f0\n",
      "accuracy(trainbatch...) = 0.0955\n",
      "loss(trainbatch...) = 3.3053942f0\n",
      "accuracy(trainbatch...) = 0.093\n",
      "loss(trainbatch...) = 3.2756648f0\n",
      "accuracy(trainbatch...) = 0.09375\n",
      "loss(trainbatch...) = 3.2393825f0\n",
      "accuracy(trainbatch...) = 0.09825\n",
      "loss(trainbatch...) = 3.2274425f0\n",
      "accuracy(trainbatch...) = 0.106\n",
      "loss(trainbatch...) = 3.2144039f0\n",
      "accuracy(trainbatch...) = 0.10725\n",
      "loss(trainbatch...) = 3.2027128f0\n",
      "accuracy(trainbatch...) = 0.10075\n",
      "loss(trainbatch...) = 3.188115f0\n",
      "accuracy(trainbatch...) = 0.11125\n",
      "loss(trainbatch...) = 3.1736515f0\n",
      "accuracy(trainbatch...) = 0.1135\n",
      "loss(trainbatch...) = 3.161421f0\n",
      "accuracy(trainbatch...) = 0.11875\n",
      "loss(trainbatch...) = 3.147987f0\n",
      "accuracy(trainbatch...) = 0.11825\n",
      "loss(trainbatch...) = 3.1349843f0\n",
      "accuracy(trainbatch...) = 0.12\n",
      "loss(trainbatch...) = 3.120817f0\n",
      "accuracy(trainbatch...) = 0.13075\n",
      "loss(trainbatch...) = 3.1071382f0\n",
      "accuracy(trainbatch...) = 0.1305\n",
      "loss(trainbatch...) = 3.093386f0\n",
      "accuracy(trainbatch...) = 0.14325\n",
      "loss(trainbatch...) = 3.0794728f0\n",
      "accuracy(trainbatch...) = 0.14175\n",
      "loss(trainbatch...) = 3.065045f0\n",
      "accuracy(trainbatch...) = 0.14825\n",
      "loss(trainbatch...) = 3.052664f0\n",
      "accuracy(trainbatch...) = 0.155\n",
      "loss(trainbatch...) = 3.0374594f0\n",
      "accuracy(trainbatch...) = 0.159\n",
      "loss(trainbatch...) = 3.0237405f0\n",
      "accuracy(trainbatch...) = 0.1665\n",
      "loss(trainbatch...) = 3.0095434f0\n",
      "accuracy(trainbatch...) = 0.1685\n",
      "loss(trainbatch...) = 2.9965284f0\n",
      "accuracy(trainbatch...) = 0.1745\n",
      "loss(trainbatch...) = 2.9828465f0\n",
      "accuracy(trainbatch...) = 0.1795\n",
      "loss(trainbatch...) = 2.9681509f0\n",
      "accuracy(trainbatch...) = 0.18675\n",
      "loss(trainbatch...) = 2.9550605f0\n",
      "accuracy(trainbatch...) = 0.19775\n",
      "loss(trainbatch...) = 2.9409826f0\n",
      "accuracy(trainbatch...) = 0.18775\n",
      "loss(trainbatch...) = 2.929104f0\n",
      "accuracy(trainbatch...) = 0.19425\n",
      "loss(trainbatch...) = 2.9131331f0\n",
      "accuracy(trainbatch...) = 0.2025\n",
      "loss(trainbatch...) = 2.9007783f0\n",
      "accuracy(trainbatch...) = 0.205\n",
      "loss(trainbatch...) = 2.8866603f0\n",
      "accuracy(trainbatch...) = 0.22425\n",
      "loss(trainbatch...) = 2.8728306f0\n",
      "accuracy(trainbatch...) = 0.22475\n",
      "loss(trainbatch...) = 2.8597357f0\n",
      "accuracy(trainbatch...) = 0.2215\n",
      "loss(trainbatch...) = 2.8446953f0\n",
      "accuracy(trainbatch...) = 0.235\n",
      "loss(trainbatch...) = 2.8307788f0\n",
      "accuracy(trainbatch...) = 0.23725\n",
      "loss(trainbatch...) = 2.819147f0\n",
      "accuracy(trainbatch...) = 0.23925\n",
      "loss(trainbatch...) = 2.8027794f0\n",
      "accuracy(trainbatch...) = 0.24925\n",
      "loss(trainbatch...) = 2.7878103f0\n",
      "accuracy(trainbatch...) = 0.25425\n",
      "loss(trainbatch...) = 2.7745626f0\n",
      "accuracy(trainbatch...) = 0.26225\n"
     ]
    }
   ],
   "source": [
    "Flux.train!(loss, params(model), Iterators.repeated(trainbatch, 180), opt; cb = Flux.throttle(cb, 60))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "loss(trainbatch...) = 2.7598846f0\n",
      "accuracy(trainbatch...) = 0.26825\n",
      "loss(trainbatch...) = 2.7465343f0\n",
      "accuracy(trainbatch...) = 0.27075\n",
      "loss(trainbatch...) = 2.7333057f0\n",
      "accuracy(trainbatch...) = 0.282\n",
      "loss(trainbatch...) = 2.7183459f0\n",
      "accuracy(trainbatch...) = 0.28525\n",
      "loss(trainbatch...) = 2.7092032f0\n",
      "accuracy(trainbatch...) = 0.27625\n",
      "loss(trainbatch...) = 2.6927774f0\n",
      "accuracy(trainbatch...) = 0.299\n",
      "loss(trainbatch...) = 2.6804216f0\n",
      "accuracy(trainbatch...) = 0.29175\n",
      "loss(trainbatch...) = 2.6631036f0\n",
      "accuracy(trainbatch...) = 0.3035\n",
      "loss(trainbatch...) = 2.6506827f0\n",
      "accuracy(trainbatch...) = 0.314\n",
      "loss(trainbatch...) = 2.6370416f0\n",
      "accuracy(trainbatch...) = 0.31275\n",
      "loss(trainbatch...) = 2.6238384f0\n",
      "accuracy(trainbatch...) = 0.31825\n",
      "loss(trainbatch...) = 2.6093378f0\n",
      "accuracy(trainbatch...) = 0.33425\n",
      "loss(trainbatch...) = 2.5947564f0\n",
      "accuracy(trainbatch...) = 0.334\n",
      "loss(trainbatch...) = 2.5829868f0\n",
      "accuracy(trainbatch...) = 0.334\n",
      "loss(trainbatch...) = 2.5783112f0\n",
      "accuracy(trainbatch...) = 0.33075\n",
      "loss(trainbatch...) = 2.5564036f0\n",
      "accuracy(trainbatch...) = 0.3585\n",
      "loss(trainbatch...) = 2.5427842f0\n",
      "accuracy(trainbatch...) = 0.356\n",
      "loss(trainbatch...) = 2.5326293f0\n",
      "accuracy(trainbatch...) = 0.34475\n",
      "loss(trainbatch...) = 2.5163062f0\n",
      "accuracy(trainbatch...) = 0.37225\n",
      "loss(trainbatch...) = 2.5043914f0\n",
      "accuracy(trainbatch...) = 0.3685\n",
      "loss(trainbatch...) = 2.4927158f0\n",
      "accuracy(trainbatch...) = 0.362\n",
      "loss(trainbatch...) = 2.4762936f0\n",
      "accuracy(trainbatch...) = 0.38925\n",
      "loss(trainbatch...) = 2.465825f0\n",
      "accuracy(trainbatch...) = 0.39925\n",
      "loss(trainbatch...) = 2.4528575f0\n",
      "accuracy(trainbatch...) = 0.39025\n",
      "loss(trainbatch...) = 2.439254f0\n",
      "accuracy(trainbatch...) = 0.3955\n",
      "loss(trainbatch...) = 2.4253373f0\n",
      "accuracy(trainbatch...) = 0.4105\n",
      "loss(trainbatch...) = 2.4137037f0\n",
      "accuracy(trainbatch...) = 0.41775\n",
      "loss(trainbatch...) = 2.3996987f0\n",
      "accuracy(trainbatch...) = 0.4235\n",
      "loss(trainbatch...) = 2.3855271f0\n",
      "accuracy(trainbatch...) = 0.4225\n",
      "loss(trainbatch...) = 2.3732436f0\n",
      "accuracy(trainbatch...) = 0.43025\n",
      "loss(trainbatch...) = 2.3715036f0\n",
      "accuracy(trainbatch...) = 0.42625\n",
      "loss(trainbatch...) = 2.3489606f0\n",
      "accuracy(trainbatch...) = 0.4325\n",
      "loss(trainbatch...) = 2.343701f0\n",
      "accuracy(trainbatch...) = 0.412\n",
      "loss(trainbatch...) = 2.3256476f0\n",
      "accuracy(trainbatch...) = 0.45525\n",
      "loss(trainbatch...) = 2.3111808f0\n",
      "accuracy(trainbatch...) = 0.45075\n",
      "loss(trainbatch...) = 2.300154f0\n",
      "accuracy(trainbatch...) = 0.44\n",
      "loss(trainbatch...) = 2.2846942f0\n",
      "accuracy(trainbatch...) = 0.46225\n",
      "loss(trainbatch...) = 2.2743857f0\n",
      "accuracy(trainbatch...) = 0.466\n",
      "loss(trainbatch...) = 2.2647107f0\n",
      "accuracy(trainbatch...) = 0.4595\n",
      "loss(trainbatch...) = 2.2502787f0\n",
      "accuracy(trainbatch...) = 0.45825\n",
      "loss(trainbatch...) = 2.23634f0\n",
      "accuracy(trainbatch...) = 0.47525\n",
      "loss(trainbatch...) = 2.226215f0\n",
      "accuracy(trainbatch...) = 0.4755\n",
      "loss(trainbatch...) = 2.2119472f0\n",
      "accuracy(trainbatch...) = 0.4815\n",
      "loss(trainbatch...) = 2.199222f0\n",
      "accuracy(trainbatch...) = 0.48125\n",
      "loss(trainbatch...) = 2.1923325f0\n",
      "accuracy(trainbatch...) = 0.472\n"
     ]
    }
   ],
   "source": [
    "Flux.train!(loss, params(model), Iterators.repeated(trainbatch, 180), opt; cb = Flux.throttle(cb, 60))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As you can see below, our loss is now 2.18 and the accuracy is 50.3%."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 154,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2.1843717f0"
      ]
     },
     "execution_count": 154,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "loss(trainbatch...)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 155,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.50325"
      ]
     },
     "execution_count": 155,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "accuracy(trainbatch...)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now that we have the model trained, we can use the other books in the dataset to test our model and give a prediction. I'll be testing it out on the 5000th image in the dataset, which belongs to category 2, \"Business & Money\". Let's load the image to see what it looks like."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 156,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "224×224 Array{RGB4{N0f8},2} with eltype RGB4{Normed{UInt8,8}}:\n",
       " RGB4{N0f8}(0.447,0.0,0.039)    …  RGB4{N0f8}(0.471,0.0,0.004)  \n",
       " RGB4{N0f8}(0.506,0.004,0.071)     RGB4{N0f8}(0.533,0.0,0.055)  \n",
       " RGB4{N0f8}(0.561,0.0,0.055)       RGB4{N0f8}(0.569,0.0,0.055)  \n",
       " RGB4{N0f8}(0.58,0.0,0.031)        RGB4{N0f8}(0.553,0.0,0.035)  \n",
       " RGB4{N0f8}(0.592,0.0,0.031)       RGB4{N0f8}(0.541,0.0,0.027)  \n",
       " RGB4{N0f8}(0.553,0.0,0.02)     …  RGB4{N0f8}(0.533,0.004,0.031)\n",
       " RGB4{N0f8}(0.486,0.016,0.02)      RGB4{N0f8}(0.486,0.016,0.02) \n",
       " RGB4{N0f8}(0.451,0.047,0.031)     RGB4{N0f8}(0.471,0.02,0.016) \n",
       " RGB4{N0f8}(0.42,0.063,0.059)      RGB4{N0f8}(0.482,0.02,0.012) \n",
       " RGB4{N0f8}(0.416,0.055,0.059)     RGB4{N0f8}(0.506,0.012,0.016)\n",
       " RGB4{N0f8}(0.447,0.047,0.071)  …  RGB4{N0f8}(0.549,0.0,0.008)  \n",
       " RGB4{N0f8}(0.478,0.024,0.067)     RGB4{N0f8}(0.573,0.0,0.008)  \n",
       " RGB4{N0f8}(0.506,0.012,0.059)     RGB4{N0f8}(0.576,0.0,0.008)  \n",
       " ⋮                              ⋱                               \n",
       " RGB4{N0f8}(0.42,0.388,0.0)        RGB4{N0f8}(0.4,0.357,0.0)    \n",
       " RGB4{N0f8}(0.408,0.365,0.0)       RGB4{N0f8}(0.4,0.357,0.0)    \n",
       " RGB4{N0f8}(0.388,0.325,0.0)       RGB4{N0f8}(0.392,0.361,0.0)  \n",
       " RGB4{N0f8}(0.38,0.329,0.0)     …  RGB4{N0f8}(0.392,0.365,0.0)  \n",
       " RGB4{N0f8}(0.404,0.369,0.0)       RGB4{N0f8}(0.38,0.365,0.0)   \n",
       " RGB4{N0f8}(0.384,0.345,0.0)       RGB4{N0f8}(0.369,0.353,0.0)  \n",
       " RGB4{N0f8}(0.431,0.357,0.059)     RGB4{N0f8}(0.38,0.365,0.0)   \n",
       " RGB4{N0f8}(0.412,0.325,0.09)      RGB4{N0f8}(0.424,0.4,0.0)    \n",
       " RGB4{N0f8}(0.431,0.361,0.031)  …  RGB4{N0f8}(0.373,0.341,0.0)  \n",
       " RGB4{N0f8}(0.4,0.341,0.0)         RGB4{N0f8}(0.396,0.365,0.0)  \n",
       " RGB4{N0f8}(0.463,0.42,0.0)        RGB4{N0f8}(0.439,0.42,0.0)   \n",
       " RGB4{N0f8}(0.286,0.243,0.0)       RGB4{N0f8}(0.267,0.251,0.0)  "
      ]
     },
     "execution_count": 156,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "load(\"judge-books/images/$(data[5000][2])\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The array below returns the percentage certainty the model has for each of the 30 categories, in order. The first one corresponds to the 0th category, the second one to the 1st category, and so on..."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 157,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "30-element Array{Float32,1}:\n",
       " 0.06904125   \n",
       " 0.0013687051 \n",
       " 0.30564067   \n",
       " 0.0019003467 \n",
       " 0.0122752255 \n",
       " 0.00063050323\n",
       " 0.014400939  \n",
       " 0.00066149054\n",
       " 0.044060275  \n",
       " 0.0011265653 \n",
       " 0.00024397735\n",
       " 0.005299801  \n",
       " 0.003381016  \n",
       " ⋮            \n",
       " 0.00074706454\n",
       " 0.0004789304 \n",
       " 0.008010443  \n",
       " 0.019767081  \n",
       " 0.0015127813 \n",
       " 0.42793602   \n",
       " 0.0020457897 \n",
       " 0.0016720602 \n",
       " 0.0029231526 \n",
       " 0.010123464  \n",
       " 0.03700547   \n",
       " 0.017807292  "
      ]
     },
     "execution_count": 157,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "testprediction = model(preprocess(data[5000]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can sort the previous array to see what the model predicts the category to be and its percentage, from highest to lowest."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 158,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "30-element Array{Float32,1}:\n",
       " 0.42793602   \n",
       " 0.30564067   \n",
       " 0.06904125   \n",
       " 0.044060275  \n",
       " 0.03700547   \n",
       " 0.019767081  \n",
       " 0.017807292  \n",
       " 0.014400939  \n",
       " 0.0122752255 \n",
       " 0.010123464  \n",
       " 0.008010443  \n",
       " 0.005299801  \n",
       " 0.0048549897 \n",
       " ⋮            \n",
       " 0.0016720602 \n",
       " 0.0015869794 \n",
       " 0.0015127813 \n",
       " 0.0013687051 \n",
       " 0.0011265653 \n",
       " 0.0008684021 \n",
       " 0.00074706454\n",
       " 0.00066149054\n",
       " 0.00063050323\n",
       " 0.0004938163 \n",
       " 0.0004789304 \n",
       " 0.00024397735"
      ]
     },
     "execution_count": 158,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sorted = sort(testprediction, rev=true)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's get the highest value and search for its index in the prediction, but remember to subtract 1 from it because the categories are from 0-29 whereas the indices are from 1-30."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 159,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.42793602f0"
      ]
     },
     "execution_count": 159,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sorted[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 160,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "23"
      ]
     },
     "execution_count": 160,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "findall(x -> x==sorted[1], testprediction)[1]-1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now, the model has predicted the book to be in category 23, which is \"Science & Math\", with a 43% certainty. Unfortunately that's not the category of the book, but I have to say that it is sort of related to Math, so that's quite close. But let's look at the second highest prediction and percentage and see whether that's correct."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 161,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.30564067f0"
      ]
     },
     "execution_count": 161,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sorted[2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 162,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2"
      ]
     },
     "execution_count": 162,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "findall(x -> x==sorted[2], testprediction)[1]-1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Yeah! The model has predicted it second in category 2, \"Business & Money\", which is exactly the category the book belongs to, with a certainty of 31%. Not bad at all!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### thanks for reading :)\n",
    "### written by kfung"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Julia 1.3.0",
   "language": "julia",
   "name": "julia-1.3"
  },
  "language_info": {
   "file_extension": ".jl",
   "mimetype": "application/julia",
   "name": "julia",
   "version": "1.3.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}