4.+Cifar10_l2_notebook.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "from __future__ import print_function\n",
    "import keras\n",
    "from keras.datasets import cifar10\n",
    "from keras.preprocessing.image import ImageDataGenerator\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense, Dropout, Activation, Flatten\n",
    "from keras.layers.normalization import BatchNormalization\n",
    "from keras.layers import Conv2D, MaxPooling2D\n",
    "from keras.regularizers import l2\n",
    "import numpy as np\n",
    "import os\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Defining the parameters\n",
    "batch_size = 32\n",
    "num_classes = 10\n",
    "epochs = 50"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "x_train shape: (50000, 32, 32, 3)\n",
      "50000 train samples\n",
      "10000 test samples\n"
     ]
    }
   ],
   "source": [
    "# Splitting the data between train and test\n",
    "(x_train, y_train), (x_test, y_test) = cifar10.load_data()\n",
    "print('x_train shape:', x_train.shape)\n",
    "print(x_train.shape[0], 'train samples')\n",
    "print(x_test.shape[0], 'test samples')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAdAAAADECAYAAAAvbXA5AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvOIA7rQAAIABJREFUeJzsvXe8JVd1JvqtqjrpnnNz59xSN40SygEkJCSRETbDA+eAbfwG2+M0Djz78cZ4jAfPe+OxzWPGfsY2/GyMB0wwwWCyLECShSRaqaWWOufum+/JoWq/P9batde593a4955Wt8T+fr/uOrd2hV1779q1V/oWGWPg4eHh4eHhsTgEF7oCHh4eHh4eL0T4D6iHh4eHh8cS4D+gHh4eHh4eS4D/gHp4eHh4eCwB/gPq4eHh4eGxBPgPqIeHh4eHxxJwQT+gRPTjRPSVZZz/DiL6di/r5NF7ENG9RPTO05RtIqIKEYVnO/bFBCI6QESvXmD/K4lo9yKv9REiel/vaufh0Tu8mMfnBf2AGmP+3hjz2gtZh+8XXKwfJmPMIWNMyRgTX+i6XAwwxnzLGLPjQtfDoxunW/B4fH/jolXhElF0oevg4XExwb8THh6Mi+VdeF4+oET0fxDRXiIqE9EuIvp3sr9LBUtEhoh+iYieA/Cc2vcrRLSPiMaJ6P8hogXrTUR/RkSHiWiWiB4holeqsvcS0SeI6G+lHk8R0Q2qfB0RfYqIxohoPxH9ynlrkGXgDG35XiL6qDpui7RdRER/COCVAD4o6tIPyjGvIKLvEtGMbF+hzr+XiN5HRPfLOZ8nolEi+ntp3+8S0RZ1/GmvJbiUiB6S8s8S0cjcep7meX+WiJ4moiki+jIRbe5RU14MuFH6cIqIPkxEeSJ6FREdsQeI5PNuInocQFX681oielTGwMcB5C/cI7zwQEQbiejT8q5PENEHiehSIvqG/D0u43xIjv87AJsAfF7ehd++sE9wceNM45OI7iGinUQ0LXPLy1TZaedgmd8+SUQfJaJZAO94Xh/qdDDGnPd/AN4OYB34g/3DAKoA1oIb4dvqOAPgqwBGABTUvm/Kvk0AngXwTimbe/5PABgFEAH4DQAnAOSl7L0AGgDeCCAE8H4AD0pZAOARAP8JQBbAJQD2AXjd89E+PWrL9wL4qDpui7RdJH/fa9tN/h4BMAXgJ6W9flT+HlXH7wFwKYBBALuk7V8tx/8tgA8v4lpHAVwJoAjgU7auZ6ongLdIHS6T674HwP0Xug961I8HADwJYKO033cAvA/AqwAcmXPcTjmuIOPzIIBfB5AB8DYAbQDvu9DP9EL4J+/+YwD+RMZiHsBtALYBeA2AHICVAO4D8Kdz+uHVF7r+F/u/M41PANcBOAXgZumHn5Z2zeEsczB4fmvLnBBAvg8X+t+FauSdAH4QC39A75pzrAHwevX3LwL4uvzuOn+B+0wBuFp1wNdU2eUA6vL7ZgCH5pz7O5APxMX8T7Xle7G4D+hPAnhozrUeAPAOdfz/qcr+GMCX1N9vBrBzEdf6ozlt35KX6LT1BPAlAD+nzgsA1ABsvtDt3oN+OwDgXervNwLYi4U/oD+r/r4dwDEApPbdD/8BPdd2fzmAMTveznDcWwB8b04/+A/o2dv3tOMTwJ8D+IM5x+8GcMfZ5mCZ3+670M8399/zokcmop8C8B/BkyUAlACsALCQ48jhs+w7CJbAFrrPbwB4p5QbAANyH4sT6ncNQF5Uh5sBrCOiaVUeAvjWwk904XCGtlws1oHbUuMggPXq75Pqd32Bv0uLuNbcPszg7PXeDODPiOiP1T6S68693wsR5zSu5xy3DsBRI7OKOtfj3LARwEFjTEfvJKJVAD4ANnX0gxdrU89/9V7wONP43Azgp4nol1VZVs6JcfY5eKFvwwXFebeBis3qQwD+A1ilNwRWXdFpTlkoPcxG9XsTeIUz9z6vBPBuAD8EYFjuM3OG+2gcBrDfGDOk/vUbY954Duc+bzhLW1YB9KnD18w5fW67HgMPaI1NYFXrYnEu15rbh20A42e57mEA/35OvxSMMfcvoY4XI846rgW6744DWE9Eelxv6nXFXsQ4DGDTAjb394Pb+WXGmAGwOUi3sU9bdW440/g8DOAP57zPfcaYf8C5zcEXXR88H05ERfCDjwEAEf0M2Ba2GPwWEQ0T0UYAvwrg4wsc0w+gI/eJiOg/gSXQc8FDAGbFWaNARCERXUlENy6ynucbZ2rLnQBuJ46rHASrPzROgu0KFl8E8BIi+jFxTPlhsGr1C0uo17lc6yeI6HIi6gPwnwF80pw9dOUvAPwOEV0BAEQ0SERvX0L9Llb8EhFtEIeq38XC43ouHgCP81+Rtn4rgJvOZyVfZHgIPMn/EREVxXHrVvD8UQEwTUTrAfzWnPPmvj8eC+NM4/NDAN5FRDcTo0hEbyKifrxw5uAunPcPqDFmF9h+9gB4EF4FdphYDD4LNjDvBPDPAP56gWO+DLaZPQtWGTRwjiK/TORvBnANgP1gyeivwM4zFw3O1JbGmK+CJ+DHwW0190P4ZwDeJh6fHzDGTAC4B+xsNQHgtwHcY4w5m1S4UL3O5Vp/B+AjEMcuAGf1cjbGfAbAfwXwv8Tz7kkAb1hs/S5ifAzAV8DOEvvAdqIzwhjTAvBWsP1/CuxI9unzV8UXF9S7vg3AIQBHwG34+2AnlxnwHDO3Td8P4D3iPfqbz1+NX1g40/g0xjwM4OcBfFDK9shxL5g5eC6oW1V98YGIDIDtxpg9F7ouHh4eHh4eFhctkYKHh4eHh8fFDP8B9fDw8PDwWAIuehWuh4eHh4fHxQgvgXp4eHh4eCwBiyJSGBkZMRs3bkB3eBT/7g776QZJ+M7SZd3FnnkuoZ+9uJY81xwp/vDhw5icnFx2JZgjFgBn+uq6J0wy7/go6l4PdWJVL6ljEAQL1lnuKBtXlpFrkpwXx+6+STy/X438ZR8+DF2dKCCpOh+TzbmyxF5XTsxmMqrqRu49v87GAM1mB512vOz2XjG6wmzatAkGqm2NtFfaNqoosX1h5pXZ57a7qKuV5l9rPuzxZwlFNHMvYub9Mgk/T6tRT8s67TZfXWilw6wLIQ6iLACkraCHmu2Lp3c9Om6MWXmmJzgXFAs5MzTQ17XP1imQ8QI1Vu08E4T8TlDg3o1EnjOx/aLmpFDGbxhGXdfhy9uxZ6+l2jBmvgWjjieRO5Iknnc8SX8lUudOp63uI++gPF+n03J1b8tve83ENbpJDMqtBPVOsuwxHgSBCcLuecK2RUbeuUw2mle20N9xzHVttbiNOh3HTTF3ftHnuf61faL6KbT3ds8fS1tkMlzWbrv72DrYPokiV/cw4v60wyFOVL+mQ8S+p7p/Cc16Fe1W46ztvagP6IYNG/C5z30emSiX7tv77D4AwKc/+RkAQKvhBoxtIDshd73r6cxC88qM/EGpgKwbf+4FFkBX3y1mzKmXyv6WuuiQxSDkG7zu9XcBAK677qq0LE46uOeeexZxzzNVh4BMFne96cfTXSdOcqz93p0cCXTpJseXsHbzFgDA0UPcJ/sPaB4DbsusDLBGveHqnH5c+e++Pte/w6PsRV6r8+Q7O1NNyyIZ0LHidInBf/QPJnK+4zlfuWItAGBqgslGNm5anZYdPXYcAJDL8qBfu3o4LZueqgEAZqabfP6Uq0Oj0UGz0RszxPr1m/DPn/s2EkWQFcpEHUZ8jycffyAt+8zHPwIASORlb7eaadlb3/rDAIC77uSom7aaEMOQJyqj10UWc9ZFwZzlCQCQnpzkA59IWZyozgj4Prsf/i4A4OF//vu0aGyMSbleuYNDG4slRxrVufI1/GOQiZECKqVlTan0bTcWesJ+NDTYj1/6qXsARQxkP1a5Aodx50L1MZGPDgU89jL5QloW5bielmAtlIUAANiPdGlwFABQGBhNy2ZOcbSbXWDkii58vFGZ4R+hu1YiA75ZPgUACAI3jWb7R/gY+dZNnHCRdEGO62paPEeeOnUoLWvVeUw3pycAAO2yGuO1Kj71cG/IpqJMiNUbVqDZdGM1m+Nnu+yyywAAff1uQRPaBTTNF5TqMoccPczv7tFDjgdkdrYMAMhls3IdtyC2fUfyHvSr+61dtwoAkC+44+OE22toZAgAMD4+mZZNT/N97LpjeGQkLVu1lonOEpn7yhU3xso1HlPG8FwXGHe/gAI8df+XcC7wKlwPDw8PD48lwH9APTw8PDw8loDFk8kTIVQqi6OHWHz/6Ic/BgCo1p2dJQKLx0a+06R03VZNkyqjutS7VreVkVsqO9oZbK3ptaF1Y4tQ4Wo1krEiPp+vVWO5PNd661ameLzueqfC7SUy2TxWbNyG17/9Z9J9odgn/sf7OCXhdS+7NC279Z4fAQB84P2/BwCIyKmPtl66DQAQx9w2+/Y5lZCJWS2WyRjZOhVupcxl0zMz9mh3HsSOplSKkbT3utWsEiw4DRu2bmD1ysAA1+HYEccbXZ2Rds5xux9vOzVNu8ltn82IOliZO4eGBlCvumOXAwNCx4QIlA3Y2mdaNX7+T33iH9Kyp3Y+DQDYvPU6AMDJKWfT+quPfAoAMLyS6W6vvPKKtCyR8d0Ru68e0iRlytLkylIVrqqzjNPUHEtOFVUR9daBXY8BANavdSrzlStYTRnLq3XwqOMpOXJsFgBQWM9j69Ir70zLogFN37t8EBGCTBaxs/wgEDOAfeuzRadutXbHuCVjtlBMy3JFNjck8v4XS04VW8qLCta+16oO/aM8Vo/ve5yvrWztfUNs5tX2yuoE51Ro1ipyzCpVBzY91Cs8tvNZ9y5FxX4AwHSN1eeRes9CSuQ+rM40eafWLLb6EWZOR5O8OIRRiP6hQWTUPG1tooHYDI1unTmm+IXsnH1F7oPhUac+lcfB0Ai3R6jaodHiwkqF1cixcfO7VfUW+53ZgOzF5LgodG2Ty1pTG28GBxxxUU7sqdY8aFwV0GmwWWh6egwA0OpoVlFCHDdwLvASqIeHh4eHxxKwhHRmBl1eZ7J8jsRgnwvc2jmMWPywRlwoSTJddtttoMuCrrIuD65z8GBM1LrALMqJSHlwyCo0FKeJWHnThdTqqqc5B6l4KSiWBnDz7a/Fmq2XpfsGxJD+hh/+KQDAuqK797orbgEA3HAHO4FUjj+Vlm0Vh50rrnsVAOBLX743LTu4h1feScyOC7Wac2Dok9X/1q0sZTcblbTs0OH9/EM5yAQl7vNb7vwJAMCGDc4Z6ODeB/kap3jlefSwc3IKM9y+JXH2KM+4lW6pxJJnqcjDtdbvHDrqnWb3inmZiE0wx+uUtzOz/Nxvfsu70rKVax4FANz35b8EAAyv2ZaWrV7F7XXffc8AAO795jfTsje94fUAgG0v4X6N1eLXOqiRFTMT7VUkznjKm9B6nHZkdR6ETgI9Oc4OKd++/14AwPEx57SyfjvnILj1VW8FAJSu3ZKWXV2ULHNBVq7ppLx2p7dx40TE3pVGOwrZvhfnttBplLJFlkySJksQfaV+dzGR6HI5ljyHVzgnYdNgibBdY8mrVXWZyoZWsiYpJn7ewweeSMuuuYX7Kopdu05WZ6TufHy+32Xls11TL7NWJMy4uhsRs1stroP1KgWAjrRrXiTPduzmsCYF3XPnMhCGEUZGRtFW85mVKknaOeiai6XuxmqnXDtYadE6A2lNoe3D/gHur9LwkLtmwP1UFQm00XDzTb6P54++opMyo4grUauxVJhVknvOKhZSR0itfRQvX/kmZVVf5EVyjQK+ZkyuPeIkWTDKYSF4CdTDw8PDw2MJWKQEasAO827JbAL5LbaiRK0AbOxgYmNF1crGxm+ZwLpJz5dAXexdl5Ho9NVLV0sLxcvYJzjTClqViQRgQ45IrbzWb2H3/v4hlq4SHVbQQ2G0E3cwPTmBZ596KN13+TWcGWhAQgxa5HT1Dz74bQBALG7/q9Y4G1AYsF3r6mtZSjo0PpOWTZw6AAC4YQuv2NvKxrD9zncAAK57BUu1R59zUu1nPvZhvl9rNt03WWXJ4K4fYLvtwb2PpGWPPvY3AIDqLK88jYp/WSvhOM02P8/4VDkti8WmVZAwgGbdrRZrrVYa/7dcGEOIO0GXxHPgINs5/+ZDHwQAXLLRSZmzEk5z/CTbUW64+UfTsgGRSp57klOXHj3pEtPsfJTDSt73n/8AALBh/Za0LGnZuEOpU6LHq427dc9rJc+OtRMpcbZPwjpuv4E1E0/scbbES257MwDgOpGwVFQDah1e1jeaXIl2y7V3ctYMdIuFgYkTZJWNLJHmD0TCSXRojthAcwUeo6GK+7Pvdp/YGnPKPtqJeVy2YW3abswmMT98oY/PO3byQFrWt/t7AIArLr8l3deWcVso8ftvba8AUJ7mfrY200iF0sRtljytJKTnKSNtno9Y2xKpGKdGs4VeTSxBQMjmssgXXHiZld5s7LV+n+zUZiXPYtG1aaMp0pv0ycCQmzf6MmxXthJlVkmUJLbJwSEuq9dd3+fFVh2rcWx/J4bbNIGOrZV9qc+Ks1XHidhAEXRdR/+2z26UDbTTbLnY4LPAS6AeHh4eHh5LgP+Aenh4eHh4LAFLcCICuhh7rLpVQlsSpaa1jCuWwo0WUO86hyFVdg5ORGdSxQZYwFlJjk+SM5ynVLEtEfH71rEq7o7X3p2WXXslq/EGR1gNQtoFuocq3Eatimd2PoSpslP/7XuaVUr7n9gFACgW3A2f3v24nMcODIMZ5/ATHDsCAPji5/8RAPD4rlNp2Yi4n994Ezu1DK/ZnJbNDrGq+NsPsCpyOHQqkssuewkAwMRO/zewktWEHeI67Dv8WFqWFUehgVE+fnXWqbfyGb7ugXGus27S2TKripI29+vsrKtDVIgAuL+XA2OAZifEc89+L9336U9z2Ep//zUAgK9/44tp2aiwNA2sZtX3vv33p2WhqFlPjbGzymXX3JaWlayrfYYdK9pquNpIqmAOwxDgTAXaQtFKVX3iAKJe6VKJ+/Xma9lhaPuWTWlZ4Spm0WqIb78OT4jbYo6RPmgr+sZ6Q+l6ewADwJDp8pXK5kQ9m+X3q5t6jg+0LEOa8rFW4bFDomYOlUOVEbVhXtS6DaXCbdbYwWjNWjGLJE7d+MDD7Ci2Xr0Tw6u5HaMct10S6/HH9bNq3bjt1I1tcc7LyPNo9TOB1ceJOCZNjI+lZZ1Oq2eOckQBcrlcF+WdZSVqWiatwN0rZRLKdIeEAG48RqKSLhWcyci+EyTPWml3m10AICPsXrlB7TAkLF16Lhazhv0OFApO5WtV/6F8WzSrkVVF27HdUnaKujiTtZtcl1bNhfWgEztvsLPAS6AeHh4eHh5LwKIlUMIc3lorOVp+3EgF2tt9dkWjpFNjpVFL5qskUDNH8gy0e3Raj9OvEEwX+focMuQFSNStEVwTn9/4ihsBAHe++QcBALWOW71869++BQAovfwqeSxFit5LJ4skRlybxsEnH0x37X2MHVBy4hI/OuwCjk8d4DAF+4TNAecoMDPDUtyRY1/hYyfd81xzw8sBAI8e4mNWVE6kZQ88+F8AAE89y9e+TLhTAaBfgp2bHSc+XF64GgDw0Pe+BgBowzk5bdhxPQBg1TpxVpp17uuTRzg4PZxlJ6Viyzk52ZXj2CRLDZpTO4xbKan9cpEAqMcJyh03frZf/iau3wmWSsfGHeHA8WPsmGKd6h47+mxaFmXYQeKGl78NAHDrK1+TlmUDdpA6svdJ3tHakJb19w/LNYXQXTkzRCKVxqaW7rOanLo4VpUrbiXdJ9JZJGEXrdpEWnbwIeaurle5fw5MOI1EW4Ls7aiutZxDV6XltBq9QJTJYmTNJlSnHRlGIg44WctDrDRXUdg9N2gS/Ei0GE3R2CQqdMIIn6rltLZSLgA0K6wlWLGWx/almy9PyybqHKo1OfZMum/7FaKNkvli8pQjOcjmhKM3x/1YLStnJcOSbmDntcBJyP1DLLHOVPidMCrMpL84MCc8Y3kgoi4J1CIvkp1REmgUWS5oPr6l6pX+Fqcq/TzD4uRVrslYVXMEJTbhhDhqag+2xJKLKC1nYpN2cB2aTUUCIePfOh91OZWJdGn966DmiVgc4+ri9NiouHdqpK+Y9tHZ4CVQDw8PDw+PJWDZNtD0G2wZ9iNVZhn4A1kCaBuopSybawtVv1Pb6UJBxGfIxgVN5Wel2TQEwJ1odeNbLmHKste+4a607JIdvO+pXc8BAI4cPpKWbd+4BQAwOir2vi5iiQXqtUREUYhVIwOYqbgVrrUf2MwCDbUai2GzHAhVnKLIyopLdyRBzNmMs9s8s4sloWeeYelq3Uq3ci/l+IHyWUs/51ag1TavHBM4W+ZTj+0FAGQOMUnC+k3OdnTz7Uw1uPcgr+pPVB3VYEeCoze+lO9df8JJ3Y3acam7kFrotESdDhYeDItHbfYUdn7zf6KUd+QPwRRL3o9+4yNclwlnm0qom/QgCFw7bFvHlHcbDPfdzq98LC278fqbAQDtyQMAgCcedjbXXB+/M6faXIeypi20qehmnbRWkSB0G/YTqlRPK4usgbj5WtYwmMhpR576FmebCKQPp1tuPIQSvhJIlptZRft2sN1bGyhAMEEWA6MqM5qEeyTWDqy0Oi25f0ZsX0HGtXlphG2YjUnORjR93GkESgOStUjmpEjJDnXJfNJscBtee+31adlRoZtsKo3IrJAkbLpkB99n7Lireqps4+s3NLWp2HQLYjttquw9aSo22Tey1lEmzs5Md/mP9AKhCtUK5xAomMS1dyDqxki2rbbKKiQ/I7EvF1Vf9NkQJNmXaBpXsWlWqmKHbM9P+aYJG6y0bGKb1nC+ls+eV1e2zLnaRi11Z+aE7HTUe5PEMbBgusf58BKoh4eHh4fHEuA/oB4eHh4eHkvAolW4BgakmIgC0ZtGkgUiDJwoHIjYbkRtqJmIrDrX0HwVrgt7sWK0KrNhLKl6cn6IS9DFhct17RhWjeT63CO/4pZbAQB33i0u/R0n/n/5Xna2WbOCVZDXX3NDWmadP2anWb1DOj1Gb0hxuM6dDsbHx7FhneP7pBF+xkceZ5WyDveIQtum/HfcVAZ/UbPEwvTTbKuk0R1xhmnzdqzlnHsyG1j1VZ1htVVl0rVtvsUqqVPTzrGk1mLnmXYgHJjPuRCcjeu2AAD2fZsdjL75rXtdHUS9cvOtdwAArrjh1rTsXz/zcX4+aec+xSFKcYBKpzeNHpfHMfPND6GQc45ZG8VB4grJVHNYqeS2rmW148tfySw1173ChapsLLEq+ugzHFoUZh2Dy9ZRVpcOFFnlOJ5x/VRvsNPJyWOiOhanMQCot8WZp+7G21hDnC6kW/JKVdYv7bVRsqrs2LbVPWydx8rxMp8/XXFtmBUnrUSSDp+sunf6UI+JiIwxSDotdIL5U5ENBcrlXNiCNcHYTCakeZiLPPZKa1m1Gtec2rXd4XHfL5lTOk3nNNJnxGwww+/zwEqXpL48wW13dGx/uq+zittjRpJfVyrqfZEE322rZlQ2nYKEttgE1h1y7029wv3ev2JU6uvaY2Z6Cr0yUyRJglqt1hUmYlWieanXUL8bqwUbxmKdddQcXhEHt4zM13mVNNtOTNUyP2NTZQlycz5vdJiSVbNmVIibDW2RRE1oNZxjYkfqbkNxtHo3IsuFy2O9pRi1GjYcS8ZPqDjcg0WY4bwE6uHh4eHhsQQsXgIlQItZVtqzLsx6IRlY429gCRVUqEq6klmASMH6JdFCqy5xMMLpJVBtALYhKus38mr/3739B9KybTu2AwAe2bkTAFBVWUhuuZmliSMH2WnkEx//VFp2bD+HWmzb8A6pp6t7AsXbuUyYJEarVUGp3zn1TAn/qruj4iWW3yktqgr36NgzQj5m5WonZVmq0US4LYeH3OpvsszPP7yCr3X5lWvTsr5+lkDHZ5xUtucoH1cGO8Gs3eaC9xvg4/IlvneoJQtZzT66kyW22+54Q1q2bguHC02fYKms1O9Ws42ZaRD1ps3L9Tbue+IE7tni8ju2pD+HxFnrlsudU9QP/dRPAwCuuYbzgWZyLmyoPcbST2nHSwEAzdCNyQF5fojzyooV29MyEqehNcO8LzGu7//h65zRJVJL5ETCJrKR1bS4+5wSx6D909yvNw84LtyaOEo9uJcdYDJZVXfJNhTINc1Kt84O7L2dQLYsEAzCuIEg6/hkY2GWiMWph9R8E+W57SLhjNXan0jmlCjP5CdRSeURbYq0Z0PWVGaOKOJrigIGtZkpVcYSZRiMqH2SFagijlsZzSsrGjLRXJT6HbmA1aw1xGnGKPkl21eULWubyidOpmVBsLicUmeDMQZZJeFZR5qCSKD92jFL2ikgkSiVs9mUZJUJsjYbi3svE+mzWQlJ6uiwFCupShfYfL+Ac3LsemJxGhySUJVAkSVMi/RvnSsjxalsnaGq4phVVSFe9brMdSW+VnHYafmGin04ctyFfJ0JXgL18PDw8PBYAhZPpGAMtNRnM60kIoUZTbsVymrAipTKdTqYQ6SgV2PpYmUBKj+nu6fuY8HZSwAgk3U7X3kr29Luuvt2AMDUjLOL3HffAwCA9RtYShoZWJ2WffULvNrf+T22d9Zmp9OykuWHsBnPz9M6hAgIwwRHjrog98PHedUb2/ZWthHbutamEKrGiSUUIJBV2eiQsquKhsAGl/eXnHR66DCvxocGWSrLDrqyulD4aSoza3dLMnyfQuT6fM8uFlsiwyv2laNOmrOEAW2RmnOjTtL90d/4Pf4h9Gt797gML1/48P93rqxbZ0USEKp9WTyrqAIPz/Lzj0pg+E/+8rvTspteweEoxx7hkJuZg3vTskyZ+ykzyJIVrXHPs2/fAQDA6hGWavokywwAjI2xbbs9y5Lh9nUu1+QrruUA/2fKjsyhTCIFkB37aiyK1FwvscQT5p1ta/OVbOfrazMlZE29FyYnNrE+kQTbSlrrbTpQmLiDTnWyi0YzlehEnWUUzWBTpL6+Ph6HpDIHhfJOZGTMRUqSCoXCj2JuE5O4Pg5siJvY7SbHXJ7amalxqZN7l1oSOmRE1ZPPO6mnJZKQJULQJCstuX6jIfdWIWEFeeeadaGWqzsChlw+3+3dRwGSAAAgAElEQVQ/sgwkSYJqtYqhIafZGBxgKbkipA9NJZ0PFHhs5qRNZ8tOirPTbFHKcmp+b4lirCPhIkbpzBKZg1oNvlZHhbE0DO8j16Rp5p2CXCvOO4nfciNEee6LQIW/WG1DLIQtnQWyNq1fw/PaigFFK5oLkX3s3MKGvATq4eHh4eGxBPgPqIeHh4eHxxKwJCairhS/1oEmZBFYJwC2GRCsCrY7G4tl95/vDGRS1WMwt8glExaVn+YKXbeBHYXuft0r032r17AjwYEjBwEAOcWBuVEYhZ54nHkuH33o4bRs8hSrTa079XXXXJGW3XwDO4ZskCTQsVLFhD2kIkqMQaPZRlRx6oUArL4wklQ2pxLv5qV968Rl2m3bRjeMrmCVxfSUWzsVS6zCMZE4hJFT7/av42dMiFU5+447p6X+Ab7W9stelu572ctY3RYOshPMsZNO/TZ+jB0jNr+Ejx/Z+tK0bO0m5oPNicNUYb1T7+ZE9dgWh4EBcmqkTdu3Yv9z+9ALmMCgWWjgkbZzk5/JiWp0B9fv0ssdTyrq7HUyWOI+aeU00wk/x+AOHjeDW5yj0MBhVhGGogJsKh5Yq6prCu+nVvO9XMZ3adqpW5/KstrYJiInpUZLZoWtyrLoKBf/FaLSbIuqsaOSFA/0SXaNfmGfckVoNXurwyUihFE25a8FgGyRVduBzBE6rCARdV+zxk5BWaU+zfXZkATLr+3Os2YNS7KTdLHZ8DM1JLSlppwJjZihBlc600WzyqrOiqj3Nfd2Q8aEdcrKl9y7VBde2IaoLvNKFUmSOaY6y057pJLNZ0uDC7OxLQHGGCRJgkrFhdBYhyI77yaRCgMU9Wwg+7LKS3RAVL0jomLNZ1xfnJji69c63LYhuUEUiCo1Iw5XGWXCiOV4nR2lKvrgTp+9vnIctexJoubXzE+2s63DkGYbsqFRA6K+7i+5ORamc87t7SVQDw8PDw+PJWAJEmi3hJVmTrGuyWqBSiKBWrIF0jEudtVGdqMkV8yRQLvye/IqIlvga1119ZVp0U03cwaVRGUAGZ/iVeLAIAe9Hz/quEzvv+8bAIAjwtsaqhCUrVvZ6ePOuzmw/85XuSD5/gLXvTzBUq3RGVjo3IzP54IoymDFyhWoVt3z5GTVlxF+4VV9zv2/0McruWN1fuZO5FaEm7Zyponb73wd15ncqq8d8TWG1rDDSkU5BVl+XcvjGapA/QFxKBpVLuBocV1DqdcWlbpnepIlrZERvk+16Z4rzAs3pSxUKxVXh0qF+yyQlWgp5+73Q2//Efzln/8FeoEgC5Q2oisTQ1zme27azuPHHH0uLTsp0tDwSpaYVm/a4i5mQ3VWstSYFJwEk18h++osGQZTzkmsIE5AdWnv8rQL+A87XLY9dtlbyrO80t+f5zGcc35CyGX5nRockvcnceO7KKvzdWt45R31uXes2eCy6Qne1qpufC/gh7EsUBAiKvSjXnMSUUOyx9iwINJ5M0XisDVKlKTWEZ5c60yoA/utFJtIG7SqTooP7dwjjkwlFe5TlG6LlRhebrKE2pbxq3OkTo5zX/YP8xjXjlspYYE4QPX1u3fXStTlKR7rGcUpHmT7euZElMlEWLlyZZf0a+tlpbKayo4yXZWsN0JSYZ23ACAj2U7soJhVUu1Ymds3Sd9rRbAjxCdZcV4s5J1WcDZtUzc32D4LRSrUuWutdFmR8xodHWJp52Ie4wN9TsosSvv2Sw5TUprMWiM+53HuJVAPDw8PD48lYJESKGHeNzewVGLUteXfc6RSo1aSJPbR1H1dS3H8O7EnBm6FkhG36suvZlqyS7e7QP2paV79rV3tVpCNGq9CvihhKc/scvayQFYdm8W29PJbnC3vztfcBABYIZnV9+91mUP+8ctMRXfLy7YBAK69/CVpWU/zgQJAEKLRdtLYymFeMa1fvQUAEJZceERpLbfF7du5bTKjjhAgP8hSW6bIAfRR1mUcyQrFWEbsSWuzrp+sjSlK97mlmaVH616t2b7iFXuG3EpycJBX49Wa5G1UeTebM+JqXuHVb6Q0ElHCK9tTz3F2jUBJHft3705XocuHQUJxFzFDQ1a2tYMs1e+nh9Kygthk8sR20eKAa9NoNdtwSezLHZXFhCKRkESCp9YKVcbbQMI1pqacpFQUO9nwsDt+cC+HXtEBllxGnPkf5Szfc3QVj4MuYhIhKbCvX63syqzEWatyx2pzYSbTOxs/X7uD6uw4OmqM26wb6dSgQhOCyPpa8N9aMssIxWPqa6HmIksVGQrVXmPaPVRTJMqcEC9kCo78oF3jdipPuH7IWxICK7iqFyAnZXXpP2snBZzkWRzkcWKp/QDANn9phOebrCLlYC1Qb9qdKEA2m+0Kr7ESaJoRRtndY8mc0mrwvkFlc7bEFVWRRKtKck2ZVu35SgINRSuVF5+BQuQkw5aNeFRzUJ/0Z9tIJqq608pUpV6tjoS4JCo3s0isBbnW0ICr+2ohgaFY7NIqlKZpgnMmTvQSqIeHh4eHxxLgP6AeHh4eHh5LwBITajvQHEYgzRqUah0WkIfTzCmiZukotdzmTayWtNlY9h51LEC5AVZxWKaLfOAu/vY3vAYAsP+QS3D76U/8JQDgwBHet2qdU+/eehurae+QbBorBx3f5aH9zAjzj5/8KADgke+6pMfVCeZJvGb7Fn4GzYXbQxVup9PBxPg4NLnTBnH0ufV1PwgAGL3xNWlZJExKk4cPAACOTDg+x0I/q/FGxdElq5yPcjm+QUZUhDoUqW2TZqeJZ52aJisORaSYS1qiioslFEDzC9sEwm3hxySVSiYvxv9U2x84lUpljJ21jj1zPz/DoFOVtpvVLhXachCEBsX+GLEK6Vg7yGqfap3Vx3uecuri9aPcF4moALe/6UfSssJ6VqPH0jSRchRqHhmXa7JqemjEqQw7FZvViPtk9Qqnri0JX2qG3PNeRqwqnphm9/1TDx5My8xmVqcPbeEx3xo/kZZVLK9wka81VXZ90ZFMPf2DNoxFh2n0jusZkDCWMIdm073jYdTtRKidRkhCmEIJR6PQsYeROEZZVa7yX0uzEIWi+80qHuaGMFwVxXErCJ1KsTHJoVekVHw2Iba1XSTqWgPC8DMxzn08M+nCc/LSf7OS9WXP7t2uglL3FSt4fhtZtS4tminPIgh745xos7HohNU2obZleStktYMRt/0pUWGfUo5oK0f5We08oKi305C6lmSEqiv1ruUxt6rbWGWGyku9SopFKiezQqXM6vBE8fHm5RNWyLBqPquyHlkfMutMF6lMYfUGz42nxrh/2upTmC0NpAxHZ4OXQD08PDw8PJaAZUugqcTQ7Qm+wB/ojmxeSGIVrBennvXrWdIYn/lOWtaWEI0dWzhA/aZrrk7Lnn2GOULve+DxdN/KVezyf8lL+fitIjUCwEoJPzhwiFfjn3voX9KyJ77H5AqTwoFrYrfqLwRuhQYApisDS+8CzTNhiLUD/Vizyklcb77z1QCAHa9/MwBges2OtOy57/JzP3nflwEAE1Un/a3M3wkAuOpmzhwSlNwzjEnex5lJdnxoJUqKFueLrKy1VIw12kLUUG449/VWjVea1tGnrjhHQzk+siQYHVdmw5iywqGbbbmAaLtW/t/ufgUAYF2/W/HXqi08/ZgjwFgWDECx6SIjsKEQ0aDkpgzdyn3duvUAgGqV6zpzyHHhTkqOVCvd9A+6OhdkKNXF+aKgclM2p/n4AcnKEQ+5sVUQp4ucemeuXM2ZavoGWRr4l284CXl1k1fll/WxFNuQcQ4AJyT/5HiG7x2oqSAWsdmIR0e97qQHG+jeK1AQIlfoRybrnG2sU4slP2gp6S8WDQhJmMRQy7VPXhy+InH8sQ5DfCL3Uc2OVaX9iER6sfNVs+7em8opltqzM8fctfpYYxBK3libxQQAoiz3Q1mkzHrDjePpCZZ29u05AAA4dOKAu2bI1xoXAperLt+WFm3Zug0dJXUtDwbGGJdVB07yzIrUR0quqsj7PD0rTlRK+xFJaEsxI8erOnaE09c64bXVnBKJI1Ii2sNKRfH+FkQLoHL+2sw2JH2/uuicSiPRFiTiPBTHrr2bMoc0pQ4zHRWeI89Vkdy6RcXxPTg4eM4Sv5dAPTw8PDw8loBlS6BLhTOP8gpA6+Q7Esi/biPb7e66zUmZgxLZfOXLmBhg504nbX7pK5wVo1xzq50NWzi0I5TF0aOPPpOWTY+zHvzkUV5d1lTQupWMQwnFSTRFoeS1RBruoLjOepi5b7BUxJtuvQWFAbcauuIyfp49z3JO0v17jqRlp3axnW4k5BXXmrUu48IQeMXVnOHVXlsZLE4c4mvkpS9G165Py6akKRsS6F1VFFtNkaBCZaQdltCOlqzsyhW36osksNmSb+QVKUMkUlgiAe77nt2VlmXE1fyu17Ct+srNa9IyJAkGSm5FuhwQCESh8sEHGhLS0cjx83RIrYyl/huuZqn+K/f9a1r29PceBQBcu30jAGD1xo1p2bYr2fY+IhlamjOO3CNbYhuOJR+ZUVk5LMFDqd+tlgdG+B1pi9R209XXpGWbV7IUOzDFBBZP7XFZXHbOsDRU65c+cN2EmWkJX2mL7UhJA0hDYdQJywARIcrmsXL9Jem+Rp0lyLbYzAOVb7NW5b5pyDhsN5VkNjQo5/F4iZW9LiP5WG3AfFOFQmSE7CArAf1Joqj8JD9rZczZMnMB+1NEa1hKDAadndpKzzYjTKwo4SZPsj11uMSS28brXV89LtqzzgDX89RJR9jRrtdS+r/lwhiDxLS7wj2i0NL18b1nyy5kZ0LGTk3aa4Wyydt8w5ZqL1QkBjYLE0n+4oGcm8MKWbFXRnx8ZdppHyoVyeakNHlty64i4TWhmrtCG9YkmqIZZWuNhFowJ6QRmjLQSpzFIcn+lFc2YapBh+udCV4C9fDw8PDwWAL8B9TDw8PDw2MJWKQK1+D0TjI2nEUZXy1TvplfZtnuA2EgimOnUoHh36OjLIJfdfmr3CVFfP/eTk50/fQup+pbMcrG/dHVjit1Qphcnn6Kj+uoDCVWGxVI/cLQNYcNR7GqhESJ9FbtnLZEF3N/75ws8tkMXrpxNcaqJ9N9lZok+BV37dop5xjSmWbV0miJn2NAOa60JYxi1xOs+h0LXRtlaqxC2baCVRwtxcRTEeeGepNVuEaxSQ2K6rQz40I0vvVPXwAAZIWU9arbX5WWWf+DuqjokopT3bQkXODxB9lh7Gtf/XJadsM1zPRjXsmq/MqMOy+AmZNZYxkgVpPmVFhOPWGVULHEjlylsnv+poQJPZmw+u0fvvTVtGxdkVVLV77iVQCAlTucus76yE8dZDVdR/GyDm9hVWZrhq9dGXfq3XCAVfKZQReK1aza5M58zbxicOlMcr88/SSzaN171LFpndrIfVzMcz3LNeXsJQ4mTVHJzU659shGvV1zG2PQiVuIsm6sZjpcp4aothMVVtIRruW6ZN2YOOlCcwZH2SkwEj5hnanJZulI66+caOpVHo9ZYSAK1DxAovY72XL71ue5PcJpCYFQoVq1SNTAEnqT01y40p6rVnKoyqaXOOazSy9ntf74SeY0LledyjYIM/jqd55AT0Bg05Oai+3YaUuoT0cxEWVyrNocyHDb5HT4j00eLk6UfVl3zTX9wrZkM+NolWhkw6NEna4cNJviMNRRXMBmTlhS0nJzbFOc2joyBzRU7NKIzE9FcTqyWWdsMwBALOrhdqecltVqVcTxuTlteQnUw8PDw8NjCViCE5FZ8E9HqKCdaNJYFdnqrCoSfCsXuOLKy9KiH/jBewAA1Qav7P7pn7+dljWbvEJpy+olhjMM90kew4NHncQ2MyNu+lLPjOKAtHW2gfg6ID8Nz5krbUJH58zNGsNP1CsUCjlcftV25AbcSpVKvMqeOi659FSePSuZNcQtv2RUBgTDq8vqLEs2jx91kuuWPl5lilCC6SOOLzh8yQ18H3E0yKrMGPEpDtq/7xMfTvfte5wJJ06M84puzbqVadlayUWKk3z9saMu6P/pRzgU5dTBAwCAHeucs8L29XzegEhLZLRDVy9lfgNDnS6+14I8byIEEbOTjpziG3vZ+eqBUyw1n9C8tX3sIFQfEieim+9KyxrH9wMAHv+nvwMA1E44Cf7KO8QhQ0JWOlW3MobwQE9PO6KQU+McLvH0Ya7Ld594Ni0LxBmkLUJQ7TKV1WeY+zyfEycKlbknCu1LzQMi6bj2yAa9a20AgDFIOi2ERecYBZuPVpyHIvXOFgfZUSgmkZxVCMS0kBdghOsYZN3cQPJu2wwqeRWob6V2++ZqSWV4NYfUtde6DDjJFLd/RRzkGk1Xh4aEHxm5fk45QAUlfocKkvM2ViM3IyE3/cJfHRWcRq7VanVx1y4HBIMgTBBGbp5y6Zf5R6S0DCVxarPEHqEK77CSqp0TSoq/NytOoaFkJWqpHLuW27spIUikpMakaR2GXB2yItW35N6UVdmSRJsTi5NnRoUU2aoa4dDtKE2VbU/7DC1FENNqt9LQnrPBS6AeHh4eHh5LgP+Aenh4eHh4LAHLjwOlOVuzQKFlKVIpyyS8CpGI/5Rx6pYvfYVTRu3afQAAMDWleDJtzKGcH8fzuTljxTJiVcqp0G8WWDPYfUarXxejquqxWktQKBZx1cuv74qD23OE1bRPPs2qUjPsuEARshpuUhxChlS9qM1OCfUaq/XWrHDsRmPPHgAAdAJWf4wOur6ot1iNXiO+diHjrvncwxx3G5ado8ttL+PUbl+/9wEAwIGdD6ZlK65iJ6BDD3G85IP3u7K1a9gx5jd/5WcAADsu2ezaQdJA9RdYLdTRzDRE56xuORuICNkoRKDiU2fGJUkzser2WOxemWdnuZ1PiSNDtujarTzFar2OxNsFWp0u/Xnjj/0CAGB671PufoeZH3VFyE4bpZy733HhOM6W3H32H2FzxT/cx+PhSNY5tKzeyP25QsJ6rXMNAIRlbsOaaNYayjEjW+Dfw6PcDn1FxSJjepcwHuA3pxUbZLV6TdTJ+RKrOjtN51BDMnFYJ4+ZCedEdFycU9ri3NI35JytbCLonKh1NdNNJmtT9okaW8X6RnLNUKVba0sbjDe5zs1ExbDXWPUajkpsacHFKOckLjVjE1fPTqZlE2Osxp+R9Gc6TrVRr6PV7E0caJIkaNSrIHLtbdWyWeHApUD1t7wLWVFz59R4bItjVSiHx4qx6siMsJqJ81ZHJSSPxUEzFnYiUsSzsThfBYlyOM1w22cHuA+zOdfeFERyLa5no6XVtHxeu8N1aLXd/B5bs106d7iyMAwXZMhbCF4C9fDw8PDwWAJ6wERkJTxxyNElZiEnG0bqOi0n7HzSOT9Y1h+bFSEINEuRXWEIs4hmhjH2PHW/1EHIqNp21127DF1M4EwVGbRVtoLyDEuQ9TF2GikVnDSS7+PfY2O8Kl814BxQQllJjhGHo2QD57hSG2eniMGb7ubr5N39ZiwDirjzV44fTctWr2cp0UR3p/vae5kZaptk1Jne93Radv8hDvd47llmg7rrthvTsh9722sBAOtXsdQQdvENi9Qhq9NYMZ4YY3rWbUkCVOsGnapiQWnySnhCVrZPKp7UKUkm3xGHjFBF09woIRXDbXHkEcchALBJIfqH2OGk/6UvTcv2PPwtAMBuYdgaVGEDE5Ms1R48pXhjbZiAhC5teqkb+0PDXHfrKDQx5hw57GqcZFw0m8o5RMZbX1HeOyUN1Ou9c5KzMEGIRM0RlvvUZgBqK+mvJRNGIk4f5SmdsJqlvqjAkopR4S8kv0NxSMlGbny1JIyiWeb2jRXrWEekqmpDSWWSeHv1JcyIpRM8V8TpKyfOTppTNRJpyWaeaqv7VKt8bxuWpJPE1+sVJElvQrWMSdBsNZDJ6pBC3rZFS2VUGFsoPLeRSO76PEtXPSMMbu2qa4dQQltIwsAC0pIhP7edk0hpfKSJECpHvlyJ65Af4XtHOdduiYShtJriVKq1iOJgGYbzpclYnFFtuGJGOUeGYcZLoB4eHh4eHucTS5BA536ZbZjHAlIcdUunXXwDqXmUurYAEIobdZDq4rVNU+6XLjRUfeaFlyxQ3zNKKzTvlw2ZILWymb820Xt6KMUaA+q0ETed/WBIsqhs28zhIUdnXchOp842lY6QHhw8qvT6QmxQWsuS4eykk4j2TPIq/hvPcljJdVdfmpY1MpJ3UVbik8cc9+6W7Xzc8KVOgnqqwnXIjvNx3/76N9OyyzdzSMev/8JPAwDuvuOWtKxPcvbFYt/UGSFs60YijdVV4HUYdCf5WQ46HYOJ6Ri5nHKTF4mnWuPV757Y2eOjYbERSsD2lsDZo1eJ7at+gO2bjZyzD4Ui/WRFgj14xGX6uPeBRwAA39rFfbFi0PEZ5yTMYDpx0kn/KLfTyDZ+lVt9TluTiL1welwkHmVrqjT43o0an1eddmW5gmRhER7gfN5JHZWpHpFWWFCAMMqio/hNjUiXtTJLc9ZuDwAGLElmJB9oruikJRvSUJlme3VWcZ+GeZZKEznGqNAJy6kQC0lDW5GtNMWW1+539tQVlzLZxaC8SzXFHWszi5Sr/A42VaYdIzbQlr2+IgsIRcIzEYt1sXF9HCNArzi2iQIJ05mfjSUUKUznA7YkJTWxZTYb7ryJkyItSy7aRPESZ0MeXwV5lzKRu6aVCAOb4UnZo632MKMIQbLCoxuKLTQxrn86tq7ycSkUnWYhSfl+rb3TtUM+sNlb7LFYErwE6uHh4eHhsQT4D6iHh4eHh8cSsEgVLmG+StSqROaLyS5sxe5UhutUh7sAg5G9pqh1uq45lwhpIY2poQV+x+4RVA3lRlKmGG7mqG67whDQrcY6X65HBIMQMVT0AVp1Vg0dO8a8pvuPOxVuUVg/RtawA0NLqaJa4oATVljtQi3nYLF2kOMcDu3la91ypWM+ik+ySurQnt1yIed6PzkkajTVHBvXs5p2/EkOUXndHbemZb/0k28FAFyxfSuAbs4mq5W1aZaM4uO0zDC2T3SqIwqCnrV/kgD1iknVtnIHAECtIom+1cNausyGYZVj/4hj07l3N7PifH3vZwEA//41N6Vl2yWFWFFSMf3rbscK9end7KR1qCyu/mWXRquvj+s1NOoGxI4NrLIV/xm0lMNZfYaPq00lUj/3/oWhTWou71iX1UNCAmyYhiJDajV6pC9Pb0UIoiwaNZVKUPo+7lgHEaf+tvUs9PMY71NhbLGkMavIeSZQia77WBUeCAfs2CkX/mIdSKzZoKnCMVqiRh5a79LRFYSdqCPXaite2YSsw5uk+lOmH5vGL3UiUu9NSxzjwkjCbLJqnNUbC89zSwARIZvJIKP4nq1DZ1PSx2nziR0LobBTZZSD4YCMw3ye269RcSrc2Smep2YmxCSjmIWClNWI/+7rU46QwiRU7KhvhZjyEjF5JIHr847MIomEIhqV8DsjE6dN19Zl7Wt379KhcPVmo4tH+UzwEqiHh4eHh8cS0MOE2jasRO1Jv+qxPqT7d/oJV27sc5yBFh0oTwvcKL2GLku6tt2S7hzJWlmZk/R4e0JvV+UpiEBBhFyf66Y9+4V/9jtMNnFszK2krcPTGuGc3bx+bVq2ehVzea7pcMjKU8ed1FOQjBFDa5l/tm2cyHHsOQ4vylR4Xxi4gO5nH2ayhKtuuC3dd8kAX6tfuI1vv9U5Cq0bZQN/EssKX4UntcStvNHgpWFOJVhvSWD8QnygJlhAK7IsECqzSvrNC2ezDGEVUZHWZ1rIFr7w9O60zPr5TEo4wtXrXSL3HTez09XXnuG+/PP7XKjP/kl7b5vFwtVFqKExpcbiqaPcbus2iRSl1BXHhWpYqEAR5d3KPZF3JBYSiEQ5ZrXb1olofngO9VjfYkAwQYRGUyexlgB4G86WdY4hNgzFJqweKLkyE1iuYHYiGj9+IC3L2Ww6Irnmsi5LSkfCUKzzUKzaIhFSB6MytNTl+EyWx6iV3ACgI31TEG7f6rTT2DSFDMFmuWmSq3ssElRVJPF6ww20mdlKGvi/bEjGIVKcxp1EuGIlpIMUWYbltI1EMgwUkYr4ZSFf5DYqDbs2yg/yNU4eYqe7mQlFRFHldracy7HmwpXn1JEnoZ3PW5KkXA3IjlzDyHQRKA2RJeIoSEJtPXLr0s4t4ePVJDpxp9P9CTkDvATq4eHh4eGxBPRMAnUSW9de+d+SGCjX6ZSAoftYhv2uL5Thpft+p6mNOs5Ki3ZlMp+uzx6TLCiBdm+7niet1/mxghIFiAr9XTXedOkOAMC6TdsBAG1lV7bu9Hv2sx1t9z4XcnLZNl4Jbx1nibV/pcuScsWlLJ3mRnhVvjJ0EuhIh21wew9z2Etfn+uLIckMUz7gSDBO5nlld9P1LGUNjriVflMMDySSZ7ulw5MkU4W4rBttO7K2T1mdxiqoPO4YFUK1PBCIg91DtRoVeyjJqrc04CTjRo2Pq05ynacnnbRoo0m2jDAl35oBF47SCHjF/p0TLHXtPOayedihb1fgRFozY+S+rg+ee1JsU2IzG1R2zulT8hzWjp9x17LhLwHZDBzuuYzY++2Qt+EwAJBLadRcmMXyQABlEKhA9rbYCLNFDh3JK7tWS4g9Mhmhlhxx49iSrMQSBmNJCQBgXDQuA8N8/MoRFR5kM4eI2NHX5yTDhthTq033vKHk/4zaGXkC1z5V0dS0ZaxS1lH5NcX/4OQppr6MI1eWL/I4qYlNeuyEC22qVsuIe5XzVqCnVBs6kpV4Hp0dJZLwExtSmOiQQmuTtKFBSmzLF1hiXbWa2zsi136VGQl7EZIFbW/sdOaE+sBlh+mIXTTRNISSmSWTF0lZ9UVQl/CcYI4tFEAQcNv3ZbrnFgCI8wYRndun0UugHh4eHh4eS4D/gHp4eHh4eCwBy1bhplz2Vr2mVA1k2fZtxEpXeIlsrN5Ue07bQ+wxC/zuqIsAACAASURBVDAEnbFOXeo/UT1YA7lSiNoQFVt3E2uF8hwVrtLvmnkW5vPjRBQbg3KziYGCC4+45iXM/vMb7/gBAMDOp5wDymOSveY7Dz8JADg26XhCdx3msJfHn90FAOgrODXVxqH7AQCr1rDT0TeUmubgIVYH7z3Gqt/houNmvef1rwMADCreyqFLWe226VLmyQ0ipzbJiJNHqyEZF0KnprGsJB1R8yYLqOiNjK2wSxUT9Kz1DQw6nVYX65RNImGZUZSmE/nUTd6GQTknnZqNUQhYNffK1746LavUWc33rae/zKcph44R6RcbdlTpqPApsk51rn6WGWbfM7zdcaVy5ChIxhXRyOfy7iWzz2U1V9pEEQlDTknCbepTuj16O9aTJEatWu7qbxvuYaeLTuLaIBLHnSBjt248BoZVg30S0zM07JKyzwiTTnl2Ro5xoRMR2MElFmc1MsopSCqhHYUqwt6UE7apRGUHopDr1ajyfWZnnXre8tkGwgp29KhLjG6I1bpWZT9Tc8/crLVSfvBlwxgYkyAI3DuUk0EdSmhLoto7nYtFbdpRma7sXJ+GG6r5M+WrpvmJuG1GLWtVM5qFSto7VuPBcoFb3y5S828U2Gw5wiyk2JDqbVaZJzUJQVOJ0lOHRLlfYObMKZ4L18PDw8PD4/xhiRKoOi2VJEVyUBKoifmrvkD0gTuPLGGBKhKnAUMLhC2cqVqWFlEHwdosLNbxpMsZyNbdOjupuhubEUJWWcqGbykW3aXOjxNRo9nEc3v3Yvumrem+Acm+csc1nHfzupe6ssefYylz++ZNAIB7H9iZlu09xGWVPs4A0lRR3AdneNX23BiHYbQ7itNSHE/uup3DUTatWZWWbVnPDgI3XXdFuu+yHcwTanMLBkq6SmT1andllTiXzMnPt1DISjLnWEAytPQsyNwgWzDIqMHYB5F0JCi/03bOELZtigP8PszOKmcXO96sN9CM40t9ejdnpTk6wS7+eRXU/hPXsHPYgDhofP6ZA2nZk1PifKEJSaSu0+O8b+9uJykNDNrcuVyHbNbVL1/i46tl3pcJtbOSOHmJFGHgrtlu98p5iBF3OpiamEBfqV/t4/ta55FIOeJY6YUCy6HtnslyJSdtHnv5tpLiZDszxU5xA0o6zct5DYkTMrEL1bJSz3TZET3EE+yQV+znOkeBk1Y6kqLE8eq6d6nR7g7VqtddmdWQNSREptly92PikF4NcoCiBGGUdO3jOotDjZmvbQtiS6igcnEaHvcZyREa6axZ8nOqxoQKjZZr05rsS4SoI5912oCOSJQtlXaWAgltk4kjo0KKMqKJsPNNpys/tEjUC0i6oZyXxAu1a4JzbW8vgXp4eHh4eCwBS5NAdbZwCQC2QfyB0oNbA4t1O+6SDNOVo+jDlVRh3cLDeaEugAs96f4b6kjTFcZibZjJvKPdMXJvFR5BEn1O8zKXA7C0fiQrSHI2ECwgNS8VcRxjZmYG+w4eSPe95BK2gRaEBq6gVpIvv4LLrhcp8HrJGgEAO59hqefAKV4979rniBROlXnlXZnlsjUrB9OyV4vk+fY33AkAGB1ykkJWVoSRkhbbEphv4vntZvvArhrbynZkJU5b1lG2v7mhRNrOHoS9o/LjmwAZZypBKJJgriiaCCX9tcRNfniFhLgMqpyC4sZfPsJS5he+8C9pWVvGvM06pFJN4gu7OVzo5y9hCeldN7g+fP+j3Icnq+6E/iLfp1JhiSduK4q2RCTjacmkM+bqt20l33toSEJwxt01q+VuEofBYXfNKXVcL5AkCWrVGsKMIksQDVSUY9t/VmXmsOPE+Ta43s+KpBoVZC5S9IDWzjc5xbb8E0edX8HQiEij6TTg3qlGg9u1oaTZyUkmaoDkxi31u2tlJbNIRWytLSWB2tCWpvgA1FX9MmmwP89hbZVjNInjLh+M5SAICPlCDrm8a1P7PiWxpcNTErW1c8ocGYbO5twSVpGAJEOOkiTtXNpX5PPyxUpaVBOzsA0D0+FsKc2fyt5iKfmstiHQLAtW+pc651R+ZEv5aXOyan+MXE4yIsnLHqv5Jm63F9SALQQvgXp4eHh4eCwB/gPq4eHh4eGxBCxJhduVuMGqXkUEJq3hafPlrWaUtGE4zVjNGx3h4soWUFukoS0LKmNlo9WG1jC+0OGWhWK++jkQ9ZextP3KfdtI2AIlVn2tnap6p1AkIoRRhOmKc4Xfe+gAAODSDcwelFFOOg1JPGwJZ+64yTkY3X4rMwNVpM57DjsX+gMH+XfSYXXVtq0b0rLNGzm0xSbLTWLnRGLEAE9K/WZVt80at1smqzPw8PGteH7Ik0UyR80LOFVvGiqlTAHtTtIzJiKuU4KWCtXoHxAzgjCW1KpKtZTh+q/ZKMcrC0VGOHQn+oUTtKY4USetSoq3a9e78XNimlV3j01wCNKrRl077NjB7bxOJecmcYyYnpLQk6JTsdfq3G6rN7Jaa3SVq2BBVNKBMK7knZ8OsgW+fnnWyFapt7SPRg+QJDGq5dk0JAQASiVJLi1/t9qalUZCLsTZpNvhRUIaRJUYqjFEwklsx/jR/S78a3KSneEGhjkEK1Lv1NQkOx3Fam4oz3Jc0PTEKQBAsejqXixyiIods7MzLgF7Q94hy7XbUariYhrmYc1QTrapVctdjnPLhuk2kThHJ24vHUJiWb/sVptd7LxpL1WtuGvWKzyOs9JPOil7oY/VwE3hnNZOi9QS9b1S2+cS68hnMw85djPLc2vDZCjQjkIytkWtm8s59bMNU7HtYFRfmEX4bHkJ1MPDw8PDYwlYtARqEACKmzIQRxqC5JlUX+5AWP5tcKyV2ACAAskeEXSvxgEnjSaL/bxbTgYlbtrfgXVE6XIwkrpYrtCOImyw4StkpR/NAWmdm2TV09WMvVyiG3acUQbt8RkhRxAniq0bnbQYyUrYcoGGKiDaCO/kcJ5XYbdefWladuvLOCSmJsHmrY52rxc+yZAdJZoqUNm2pBYkrX3ftnOouSllldiR9m6356+qrQOZdpqYK5W21Q3jpLP4bD2nBSEMQtQbitNWskrUZ60zhXueokiXNj9sy/lJIKmL81Get2W14i8In+7wCr7Wmk3umis38FgaO8kr+OcOOTIMjPA1hkaUs5IMjQHxg6nOqjCTHB8/uFp4hguarUTCJppWslDB6SK4xdKHhX53niVZOHpQOc4tA8YYtOJ2t7NZ6iwmhBvaMc8+ehoIr6QlkSKsdBqqLC6GKl3HN6quXWeFQ7oqXLWBCuk5eYz5pBvKqccilTKnJueVhSm3r+vbGcktaiU3rWWx86CdZxoNRdxQbyLukRNRHCcoz9bQV3L3djzIMid3ze+8TbUByinK8le3RbrvqDHUEQejtvSF5aoFgNHVnBmnILlPx0+5vrBhWZFqm0Ifzz1kHYtU/5CEgIXiDBSqsqI4d+VFSl2ISKEyW5HnUnNK25zznOIlUA8PDw8PjyVg0RIogZAkbvW546VM2fb7v/8bAIBOrKXMlOiPN11kCZauqzvzisaZv+4LhbjMu036V7qimB/14v5M9A6xj9J8ycPSYF11xSVypEoS2UtaP4N5K6FQgohnJNPEsbHxtGyT2EUjsUkmKqegkdV80hTpL1Yramt/ETuv1hSkwUZkNQXKxV1c+7toDqXdInEZ11kVUnf5BTLcWJuEXYlryrCUTkzuk1WB1CBzzrRbZ0OSGNRqBhmVGaRZtaFYVhJ3ZZ2WEBTICt6SEgDORX+l5OlsF51EgUlut/UbxZ6Xd+flZEVdl3dn10l32uExyWajum5wWKRESS4SR8pWL5cNIh4Hrbbr17qY5hLL6afsvpkMazI6koEkqzLwmMJ5WHMncRd9XDPNfMLPlsvOt13Z/Jhd2Wpk27HjUQX9tzs2VIPPK/Y5uyVZmr9J9gUoV5wUPzY2NqdOzpbfVyhKHdyj2DGdBvSr8dKWtk5DJ5RBeUrsqZZEol5RYR+NVs9soESETCaDgYGBdF+/SGqWQGNiUkvnLDVbzZOuRyRjNSW8yCn6vdD6v3C79w+6+w2X+HfmEj7v4H6XNapS4b4o9rnQoOFhHtyTk9xGlbrqC0ukIDb0KKeketGWGMkkY/OeAmrOEqnWKKm7E3fO2avCS6AeHh4eHh5LgP+Aenh4eHh4LAFLUOGargSoq9awC/g997yWyxfUps0XiJXy7vT3WqrdfH7Sl3M9/LRYSPNr1TRJVzhG70IqDJirMVQMGpF1vpJKTEw4dYtNArtx3XoAQE6pQayjFKUsSlqfLk46wiqltPBoBZYtR8JSVFJmmwkhVurWclW4QJP5Q8seZ7Vu2pXeGvitii5Q2RsiOaFtHSuUNitIljFO5sIQ4jhETlW9OiN1zgmjStbdrCYUQlXp84ZyisqXWO0YWVVW4FTZpVHet1Ic1NpK9RWLit0M8DPvOuE4dAeENUirzK0GdlY0+UGfu49JnYEYSU2p2JqSiFjCBYZHXXt3WtwvBcne0lZE0HGP41hMYtBptdFUTjrGcNvF8nCaJcayWQUSohB2JQLnrX0fqxV3zUnhr+2IE0wYOocS66TSminLMe48ktbTyZg74jxUl+TeBaUO7hfVqGXwabV1e0l2G1EZ6zCuSsU6s7TlfsoURtQzy5AxQKsdo6kc5cKQ36uWsC5Vppy5oSwsVrUa11kz9BSL0gdC/pNRL44NL7Eq37wKdbMsQEUJL1mv5szpaW6HNJsLXFJz6/DWbLp2s+xTiXA7d5QjaFXUwXXxbNSmHjsXWYczyw4FAI36uavMvQTq4eHh4eGxBNBiQgCIaAzAwfNXnTNiC4AWgGMX6P6LwWZjzMrlXuR5bO8cgEtlexTAqefhnr3Exd7eVwE4AKB8Hq59oXCxt/m5YB14zO8/TfkVAA7h4ui3F0N7LwfXA3gSUKmBzi/Oqb0X9QG9kCCijwA4Yox5z4Wuy4sNRPTXAGaNMb9+oevyYgQRHQDwTmPM1y50XTwciOi9ALYZY37iQtflxYDzOc6JA0S3G2P29Pray4FX4XoAwGYATy1UQKQJGD0uFIhoibl7PTwuPF6s4/ei/YAS0bVE9CgRlYno4wDyquzniWgPEU0S0eeIaJ0qey0R7SaiGSL6n0T0r0T0zgvyEC8AENE3ANwJ4INEVCGijxHRnxPRF4moCuBOIhokor8lojEiOkhE7yEJwCOikIj+mIjGiWg/Ef0HIjIv1hdmGbiGiB6XcflxIsoDZx3Lhoh+iYieA/AcMf6EiE7JdR4noivl2BwR/TciOkREJ4noL4iocJq6fN+BiN5NREdlPtlNRHdLUVbGdpmIniKiG9Q5B4jo1fL7vUT0Sem7ssxNV1+Qh7kIQUR/B2ATgM/LPPLbMn5/jogOAfgGEb2KiI7MOU+3cUhEv0tEe6WNHyGijQvc6zYiOkxEdz4vD3cmGGMuun8AsmA9/a+DGY7fBqAN4H0A7gIwDuA6sP3i/wVwn5y3AsAsgLeCPYx/Vc5754V+pov5H4B7bRsB+AiAGQC3ghdYeQB/C+CzAPrBtuhnAfycHP8uALsAbAAwDOBrYAfi6EI/18XyD2z/fAhscxsB8LS022nHspxnAHxVzikAeB2ARwAMgf0yLwOwVo79UwCfk2P7AXwewPsv9LNfDP8A7ABwGMA6+XsL2Ob/XgANAG8EEAJ4P4AH5/Tbq+X3e2UueZvMSb8Jtp1mLvTzXSz/5rTXFhm/fwugKOP3VWAz3OnO+S0AT0h/EYCrAYxKmQGwTd6BwwBuutDPa4y5aCXQW8CD9E+NMW1jzCcBfFfKfhzA3xhjHjXGNAH8DoCXE9EW8IvwlDHm08aYDoAPADjxvNf+hY/PGmO+YziLbhvADwP4HWNM2RhzAMAfA/hJOfaHAPyZMeaIMWYKwB9dkBpf/PiAMeaYMWYS/HG7BmceyxbvN8ZMGmPq4L7oB/BSsP/C08aY48T++T8P4Nfl2DKA/wLgR563p7u4EYMXKJcTUcYYc8AYs1fKvm2M+aLhDN1/B560T4dHjDGfNJyi6b+DF5e3nNeav/DxXmNMVcbv2fBOAO8xxuw2jMeMMROq/O0A/hLAG40xD52X2i4SF+sHdB2Ao0aWHoKDqiz1IjPGVABMAFgvZYdVmQHQpTLwOCccVr9XwGkELA6C2xuY0+Zzfns46IVcDUAJZx7LFno8fwPABwH8DwAniegviWgAwEoAfQAeIaJpIpoG8C+y//sehh1Pfg0sRZ4iov+lVOVz+yV/BvOD7osEPLesO82xHozFzAcbAew9Q/mvAfiEMeaJ5VWpd7hYP6DHAawn6qJl2CTbY2CnFwAAERUBjILDL46DVYm2jPTfHucMvXAZB0s+m9W+TeD2Bua0Ofgl8Dg3nGksW3S5yRtjPmCMuR4cYvESsNprHEAdwBXGmCH5N2iMKcEDAGCM+Zgx5jZwexsA/3UJl0nHtvgAbMALI6zu+cKCWZfV7yp4oQcgdVDUi7zDYNX66fB2AG8hol9bTiV7iYv1A/oAOC/YrxBRRERvBXCTlH0MwM8Q0TVElAOrqv5NVIv/DOAqInqLrCJ/CcCa57/6Lx6IausTAP6QiPqJaDOA/wjgo3LIJwD8KhGtJ6IhAO++QFV9IeJMY3keiOhGIrqZiDLgyagBIBZp6EMA/oSIVsmx64nodc/LU1zkIKIdRHSXtHEDvNiYn8397LieiN4qc8uvgWMSH+xhVV/oOAngkjOUPwuW8N8kY/g9YNW6xV8B+AMi2i4Ocy8jolFVfgzA3eDvwi/2uvJLwUX5ATXGtMCOQO8AMAW2wX1ayr4O4P8C8Cmw9HMpxNZjjBkHr1L+b7Aq7HIAD+P5C759seKXwRP2PgDfBk/8fyNlHwLwFQCPA/gegC+CFz9LmaC+r3CmsXwaDIDbewqs+p0A8N+k7N0A9gB4kIhmwc5cO85PzV9wyIFt8+Ngle0qAL+7hOt8FjwXTYF9AN4q9lAPxvsBvEdMCG+bW2iMmQHwi+AP5VHwnKJNbP8dvCD/CtgZ9K/Bzkf6GofAH9F300UQXfGCIVJYCkTNcgTAjxtjvnmh6/P9ACJ6A4C/MMZsPuvBHh4vEJAnXfBYABelBLocENHriGhI1DW/C3aH9mqW8wQiKhDRG0XVvh7A7wH4zIWul4eHh8f5xovuAwrg5WBPrnEAbwbwlnN0ofZYGgjA74PVWt8Dxzj+pwtaIw8PD4/nAS9qFa6Hh4eHh8f5wotRAvXw8PDw8DjvWBRfaTaXN4W+EmqS/JXBEmwgIZs6cNMmJTWwCU3nnoU0C+6cYDcAwIrVHCJULBXTonKZMwuVZyXxauyc4Oz1u5JfB7zTxLw3iXUi6a7bddWv0MfOX8MjIwCAjkoaXW+05N72mi4ZKxHQqFfRbjWWnQJ3sL9oVq8Y6koobPPs0gKZy8m2vi1Th9j6T89wuzWazjG5L8+e5JmMTdbt+OOTJJbzJXn4/OaD1mIkcoDNu5sruMTFgVTe1qWjEgpnJfE0jPRX4srs9RtNrkO92eoqixMgMWbZ7R1EgYkyQZrwWz+PfVrd7vZnIHXPZNyz2kI7NoxKSm0zgJPdqvtl5FpRxFudwNr2gelKR8/3sYmBF85oH8izuLJOR967xNZFJXCWnzZ5tX2H9P3KU9Vx04P0WlH/kMmOzuEiMF2bBZG+62rsmfQ8u2+hd+SMV11c0RmVd6evw7lg7qU7k8cRV6aXPcZXrFhhNm3Z0vUez7unKqN0u9A8LRs7hy+gzaQ5WwColyVJvOE5yM47ABDI3JME7l3K5Dh01L57+i7pb/utMQtU/lz6oOujARw+fAiTkxNnPXFRH9B8vojrb3o9du/8brovDCR7vbzweTWJlKtVAEBLMopn+5xHciwParOy12ouC3yc8DXf9b//AgDgjrvvSst2PrYbAPDxj38OALBv7+NpWS4vDUzqpQqlYeV7VptuqPvY4wKpk8tCfuW1zOj1M+/8WQDAiZNTadmD//b/t/deQXalR5pYHnO9Ke9Q8K4BtCHaG7bh9gzdDGNGs8tZN7urWUkhhV4UI+lNodCrQg8KRShiZWKfpBAVu0uN0dCJpptskk2ym23QDt2wVUChUIUy99b15jg95PefzNsAu4FC7Tz9+VLA/e895j//OSfzyy+/ZCEM8w5KYnmJ57MOvf7Tv6K9sPFKkf7LP/8DirqN9LO5ae54X8jzXEbqxW7eWrniNBER7VsQTYOsz+f/3R8wGfnNd99Lx/Yf4u+tr6Em3JHzWVxk/YqNGhyXtrRGzGf5ms9OyHN0cpL3XSzxsczPj6Vj5z/8mIiIVtdYnWtzu5aOPXiSj+H4Ma68aPXkGHI5XlPv4tr/+Oey/obhkGq9u+se/3nmZxyaPl6gUimt9aZcnvedoCpnOBTHw/P5HKemFoiIaG5BiMcenIVuH9fOlXUXhDyHjeYtbEeOYbzK1zeD37uefunxnASBdhp5vFLh35VL1XQsiTJ8DB0+dt+tqG1hrM3nM1BOyWDAxzoIhpiDtI8D5bPsbP3oW6/tSU/J3NQ+OvPffetTL0LzbDDOw+3X16yJfiC/6+Hr5vaP1Tad1PH8/JSVcwdfzNEPYfPPdPt3+1L+3c9js4X4Di8Cx3Ho+v/wH37GPu7eFg8dpu/88g2iUObUw/PPxdx4jh5jyyT8nSS6vTptc2uTiIj6gVBN9i2wUzRs8Ttgc/WtdKxde5u/32K9EEdd3/FSCccia9UvzfK+M3h/uBJQublxHoND6Kk5zhcniIioUOTnU5KfTMfMsgkD3vdQPUeb/T792T/8w9vO805mIVxr1qxZs2ZtF2ZfoNasWbNmzdou7J4g3DAMqF67Sb567boG/jTQg2q/7OZ487kC/52YkBC60WAYq9ls8oH4ciilHIfoQwPhuArO8BGqMwJFOVfgpajbv+0YCsifRh6H6Jm8hOoZwGO9AFi8KyeWz5vj4WPoDzrpWA5j1XGWGi2pPF8S9Ucw/fuxbq9P596/QO2mQJ3HjjHUOT/PCoWVvMAZE9MMdbjVOT7mvoINBzzPJ/YzpFopPJqO3ajx98anGa7N5wUGqRT5XMZKfK6HFs6kY4+c5X8XALEQEbW6PJfBkGHGza2NdKwTXCQioks3toiIqFZvpmMzc7yNxyb52PdPigKjmc7+gC/sG29/mI5ttUK6E8S3G8sXCnT69EMj28tmeaG1O3ys2ZzK6SI3WCjy3PjqxjDQa3WMx+JE1kR/wGuwvsPnE+mcLiCoGOBZqTCejgXAncJA0h0mZ9xu8b47LSW6Bbit1WRoLZ+T85oYZ1jL5G1bLeE1BEGIn/P+Bn05PoruOw03Yo7jUNbzKFS5XpN3drw75BHxT5M/jxOZc888FPAs8hS0mgCWNOjsSB4NcKnZi6fiijSHd6fcsns7hJt8itOhskK37Ufnq9P9ALpPVJIyoWS3qdQ7mucQxSNQNnLdwLkzKh+eNd/DnyAaOSEiIvJxTwRDuZ8DpJ0uffQrIiI6/+5307HxKYZip6Y59VHKy7M1zu/wMai8u+e2cXycPnBVfpRivq9CrMtuV5QB44RVBXNlhnKz6r0QxbymezHfL72+zEd3mJgl9LlmI1Br1qxZs2ZtF3ZPoZLrERUrDm164pHGcI2MF5JkxXMYr3K0c/DQYSIiOoS/RERXry4REdHSEv89fFjGanUm7BjnUjmZVKpyxHn6oeNERDQzLiSVS+c/IiKibkdIN+NFjtA6LnvtUSRuyMnTp4mIqDrDHsrUzHQ6ZrzX9bXr/LuhnPOJoxwdlUDkqY5JwjuKQvpxXnlI92FBlNCt1pCGrkTu0Tifd1DhYy5NKQLPDEdvq3Xu0HTzxkfp2HSWI5n9B7nZwUEl+dx69XUiIqoN2RN88szpdKzkcXS6usoJ/40Pr6dj2+N8ns//oyfTz7wSH9cQwe/Kkny/6yBin+RI9/2Pz6VjnYQ9x5k5Hls8IB29QkRCV64x6WZuTq5TJu9TayAEr/uxTCZHi/uO0Py8RNTTU6xl/atf/xLHoghjYC97JkRW6zSBVx+CdOT64tLmC7yGZ2bMfiT6mpri9ZzEiE5DRcpr8La6bVlvccw7HQz4uPqKjOeCDFYs8DrN+oLWbGywp2+QAk0i8sBqKoPQ0e0LOaSuUIO9MIeIPCcZYfomYGKGaTQuFht2c2JY/3I/u/ii2ZZhlxMRDQZ8DqE7GlERKS6Q2U50e/jh3iHWuBMhKUmjyjtQ+z9FTtJs6pREZCLYT217rwJQl4gKjkeB2n6UVlB86tiJyAEqdyfulZmlGGSz7euX07H6eSYk/vxNJiv+7B15DsSI/kolRqSKGTm7aoUJfONVIfIV8DwtAgWrVIWMOjHJn41V+Z6YmphLx6amT/K25g7zeSpmf9RgFKyKbnblgqz/gj9Lvnt3Ut42ArVmzZo1a9Z2YfcWgboulSoFCnzxmDMZ5B2Qrwgi5clG/H4+epTp/S++9GI6duIUR1IXL7IXcuaM5NbOvceRydQke+OG0k9EVC2yF/3gqcNERPRB/3aPu5zPyEGj/u7Bhx8iIqIvPCQN51944QUiIlo8wlGPk5PpaNSYmt24xZ7UjRvSonF1nccyzgDnKR5UMZcnd4/cxXwhTydOnaapWSlHOXyUI7Nyjr3rqTEpW4hRmrBybZnHVL7u9FN8rjmUs/zipz9Kxxrr7IWdWmTv7cO3RDp4ZZPzr83IRFSSK9sCFf6xP/zj9LOxCl+zXJU9w6OPPJKOOcjXPv0Cd6brtSW3e/0i99GdgkfZVChCcwfXGJ7rqaNSNxgMJ2l1W3Ki92NhENLmrW1ydBKdeE4d5No7LYnGQuQ1vSF8cVcil0KRf5fN8e/yBVmTAcLzCtayrjEe9Hit93q8ttpNWd9JwtfTdWRbPu6/fJZ/V8xLdGpqaUtl9uZLRfHqTdmY6/N+8kXdOnQ0EskrVMkp6rm5f3McIt8drS+OEdsYdCvRqreywwAAIABJREFUhbKmptVEnio08hFdmtRpXh2qQbHiOyS3nE/Fd+4dbuCRjz5d867L5tLA0x35P6lfSN3kbUPkOaZmV45zr9XiYmc0GPYwT565CGowxL7NFfAUv6PX5jW0tsqR5zUVgV67iRzoEjdbqczIcyqD/X3y3iUiIgqG8j4x5WJeRvMJUBuNAy2osqoi7qFqhZ8tc/PybDh5gvkXp07xMWSzErlurPPzvIlSsrFxWf+TUwENh3fXZMdGoNasWbNmzdouzL5ArVmzZs2atV3YPUG4nudRtVSlSlFgIiPPFiPk9VWIb6jf29usPJPLCvT03DMM4x09zPDp1JQQZU6fOsHbdDicT5TCRbXAYXwjw2F/PhFI8dA+JpfkSkIsevzZZ4iI6E+++adERHTk4MF0LA8swSicaZW/DGQEe4u8ze3Dkpw+D7hxc5tJN0YhhojJR94euSWTY0X6p994jLpDgR6mDzEUkgU6m4nX07GdBsMZXzjDxzzYFmhk0Oe5vAxI+tV3PpYd9UHxB0Hk6q3NdMgDNJIDAcUNhZgRgQzTBCGFiKg6hgbyZd63kxXoZnKW/725wbDJoRMC2x85doTPocaEoMaOkFWmp3mb4xO8v7IgOFSv1eiHv71Ce2G+79Hk+DgFoaQhzn98noiIAsChjlLa6g7MdeeFUynJdXIJSjko6+k0paSoi8+GA1PWoEpjDKHC4c+iUCmrGFJMRhZqFlBXESpfriNlTYYLY+CoVldUpAxE6KL0RssqGqgshkpLpSTw1tTEBO2tJeQ6CTm6/A0Qom9K5NS3A4yZUpNIqeYQFMw8pG2y6peGlGNg4TtLHhq7XT1opIwllRGMb//2pzbr3IlElErSJbcNyVdGN7RXJKKEiMI4IU9t35TveSi9ipVKkSHerG7zPb66vJSOhV2+Rzd2mJDz5ppcxGEHBDuX115bKcAdOLif/x4/gYOSd4bv8zsiUrC1WZtGZchVx97voRyyg7KUG1I2d32TYeTXobqmFbwGQ/6dC7KekYYlInr62b9HzbaU1nyW2QjUmjVr1qxZ24XdUwQahxF1Gh1amFlIP+v1ODr04fiWlBc+O8s0/elpjog6HXmrz0xNjPzNKcJLwZSeYNv9nngHtQ2OoK5+yF6FG4ln89QzXE7x8Nmn0s+e+yJ/Nllhj2jYEA8qBk2fMohYHYmuwpAT5L3tNSIiirZvpGOPHeKIaLnM53rtuhCMev3B7bz4XVoUBdRortONZYmIruHf1QX2pnodiUBbG+w5HkOpCsUSqr31yx8TEdH5JZYwvX5NpEwncc0KxNs8dEii9KsgJF3eZC8zCBRJDOKjr/3wx+lnvx+DIANN29KslJyUKxzFL13+hIiILpx/Jx07BZLX/FGmns+TKt6HF3wIOrwH9onIwvLFK1T6N7L/+zHHIfLzzoiGahFkoATF2+2OrMUcyDzVMq+fUkGivzaaHhCIXRlX1lbe538nKI1KlMh7AsF4D4Ih/b4II3RBtshkZVv9DqLLJshAqnTDaEqbSClSUWb0KU3TgdL4NYFRDmhSTlWg6yYEe2EOOeR7rhLtlwPwhnz/O+p6DBGtRB6fZ6CikQhRZgHHW8nJcTsQoQji24UKXG+UGBXdiQX4GSL9n9at/fQZ3v7v2wUizCbM2nM+HdZ+ZsR89+YQP/R15ASgT0pnVFDvoknBb956l4iI/ubb/3c69shRvrdf+vJXiIjodCAI47l3+PncBfK1pVCqMu6X4vg4/j+Vjnnm3nDlmhhdad81zS5U9IyI1QiCuOp35roYkZ7JSUEm2y1GRYt5Jta1WvJuevnZZ+i73/4W3Y3ZCNSaNWvWrFnbhd2jlF9IW5tbIzTvZoNxZlNUrMdMofhTT3FEeOKYVO+XkbNJy2CUF2dyMJPo8uFNiIcyi4LZWXSf+OpX/yAdy5Wn8TuVp+lxPq+5zKUOYV9KJ+qIKiYPPc4fFKSA3kXJRL/J5/fez19LxxYXOALad4ZLNIIpibovrXZHchv3Y+1Oh15/8w16502Jfl0U01eqiEADicDHZpjCffQfP8jHrqT8ajt83sMOe15RR3KMkws8X1WU8XT64o0Z7++ho+wt9gdqm3WOxv7tX/0g/ez821wC8+izfM2f/MofyX7mODI+dpARjJvLF9OxGx9xh4ZDx1nE4eaGlLGYYKmCnGutI7ndsZkF8nxVtnQfFsURtXotGqgoKwvPtt/maCibkwivUua1WCrw3yMHT6ZjQ0SLN1cZITh06Eg6dvAgl3XlEEnWarImL13mdXr5CneeadXUtUBkZaJOIqK2EU5AHitfkONLEnRTAbpTrapOLbjf4shEpxJ2GJQhRC6425WxnDr/PTGHyHMc0kGgiTDOLPJ67jQkerllugJhbWcLck59lIBkUc6SVfnRihE3ASIwUMIopm2b9KeTZ5jZwp2k/Ix0oG43KLGlSWbq8xqNWHWpinluSmSsSoecTxfa7N4cgkxeNLp93jduNFWWY5CBHfBYjAgIEVEb1+Cdtzk6XduRtbqzzajczbU69quQAqy1o/s5FzqlxGCWVpgf4SpZVXMRQhyf5tkYqc08/hYKUqpluiodgUjPiWPH5Pgg1rO5jb91ed6srK3RMLBlLNasWbNmzdq/N7MvUGvWrFmzZm0Xdk8Q7mAwoMuXLlKYSHgbopn04cMcHn/569KI9Otf4eTywX1MHjFNnYmIXAO3mLIAhVG4pjMLlF0SBbcEAcOGM4usyJNRJSuEcoCgeTP9pL7GhJW4zTDQB++8kY612gxBnOJKF8pM7ZffodnzB7/5NRERvf2aNHEeh2rLH/0pwxmlBSGPdFoNiuO701H8PBv0hnT5/evU39pKP5sqMVyYQZnD/KKQdE48xpCyA+3Si+dFazafZwj2pacfIyKiyarMWzbP55PDfG+cv5SO1ZoguqDLh9YsbQJKbHgC+ay/y2okl1YYOu/sqGbgCwzhV2ZBHCvI8ttoM1lrc4Uhzw8+FlWT5VW+FrNTfA511Wz7yP791O6o7iT3YWEUU63RpqHShfXRCHthltfwWEUgQ6MxGw75/HsKWj19iq/FkQMMp/ueQJ8ZD9cO2zxxTMp5nnmcF+PNVYbtX33t1XTs44tcerS2uZZ+ZuDWEjREsznxicOQcVEDGepm6OYcDXx3J4We4RBNxPtyXtFed2MhoowTjxC3xkBqOzbHqZumgqXzgO82Vq4SEVEmJ/D9ypDvw8PQEz5YlG0O8vzZaoNhxiYgSSKiCHFEA2VJWSW+nUeZ0FCRrgYuH59phB6Qfh7yPGZjKFglGg7GGA6rVJTUT73P1ybEriMtAL6HcU4UR9Tu7VBFQd8pco3dZFQT980aP29Nt56D+6Vp/PmP+Pny41dYS7tUkc5BZ58AkXMflydeBjRLRBSD1FYAHLwwKyWMWR/vA0UGilFfWN9huLWo5i2P6290m3M5gZh9U6YY8Nxu3lxJx0xqagzKWu22nPO5985Rr3d3zxQbgVqzZs2aNWu7sHuKQJOEKApdChTtfhI9KP/pP/sXRET0p//gT9KxGei0evA48qpUJQvvPaWQ38GxTYzepSuHmcuxp+3Ao09iGUtAqOnUpESjVed/1zfY8/zZa+dlbIc9x4+W+XeZcUVIQInGTo0j16sb4r2HXY6u4r/8DhERPfc1KZtZvnRztH/ifVi1UqEvf+kl2q4J4acyfZiIiKbGef4OzCrCFKbiF69y1HLlipTXZIm7ITR3uNC4mBUxDB+lJxE24BdVVO9zBLldh4ccKPq/6cmqREdjl6PZW02ev7feEZ3aqfFlIiJKMnwNE1U9Pw9izTgELDKqt+Zkhb1KE2k++9TT6diZk0fof/rXQq2/H4uimJqNLoWqVMfoxxrCyPaWdH7Zijky3r+Pj33/onjnYxWOnrIZ9ow91cNwY4OvQbfN+wkGQkJpI2KfHGdk4U///p+lY7968xdERPTtv/k36Wch9JirE7wfV0UPvTYiSAgpxCqKSntu4uvjYxI9GCpMD8QpTapqNfYm2jfmEJHviJgJEVEEotryNV6/bVXKU0U52nMnmGDUVw+ODVR0HZjgtT0/lBKvGtblMXQQag/kd5Upvu8/vLGJ/cn1Pw3t6aVViaDWQRDLptqxQqxLu5dEhtAk0XPg8O8m0MXpoCrx+nCN7/FeiOsSy/qPE+c2vd7d2jAY0vW16/TICdGojnAepgfnTkOedX/53Z8QEVEHwgLrq0Jo3N7k+ZIyICV+gBKkrGdKSVTpFSL9QY+3mVV1M1NjfH3yOYkyCc/6DMRzfLXG0/IfrO1EISkRnlW1GiN4TUVyMtzVMvod++odkyR3P9c2ArVmzZo1a9Z2YfcUgVLC0mKZjODMj0K0YBolK826yMBVXfYcs1n+fqIo0AbjTlJdLLWflDIOb071/DPlAS7yjMWs8tSG7B03N5fTz26u8L/Pvc0R2IXLcnyNFoqrqxwtlJSW39JVzjP9+X/8HxER0clHJO/0219zl/XtBudaNzfFY1u9vpn2WLxfK5VK9OTTz1FjR4r354+z4EAO0dvKe6+nY00ITmQxb/OTMt/1FnvJ776P3KInEdG+aY44++gKsrIl0omTRUjFISfTHIp31himXQzTzzzXdMvgz9ZU/0hTkL+6xfmntupd+ZX97Ok//xyvp08uiDzf5DgfVw4iAaaLCRHRRqtNwR36N+7GPNehStGjbFak64xHu4qOE0Eg+xoOTOcI9pZPqvXzEboMvfMWi0VEgUR/VXSs+f3f+z0iIjp4UHLvK20+748vcyH63IKMPXSW89e5MaHqv/Y6i0g0O+xlD4YSDUWmV2ifjytU6zIPFzxryl8iuY/2g18wRCS+sSH3jBZD2QtzCNJsCo0w0m1XGm2ch3zfw/0+B+SlOCVoyc9v8dzVG3y8QUfu2Qs1jpz2j6F/pCqbOVBi1KNT4TmpF2QusgEjAvNlJZaAiLiYFHG8cj3qkPXcMh12FD/AdPkxPU13lCiHdJfhP7qMw4npjgjdbiwIhnRzfZUeP/lw+lkGxzPEWj9/Q6Ltt99i7sd/8PWXiYio8Lx0s/pf/9X/QkREl3f4ftalQZvbvB6HOMfKhKB705P87wo4EHEszxuT0y+XVDkWypKiCCIjKj9qyn9cj49doyXmez7y2E5GFlKcQGTElMiEClnzM3S3E24jUGvWrFmzZm0XZl+g1qxZs2bN2i7s3khElFCShFQqSzi+tsbQyNUlhk+ef/J0OtZvc2jfRyI9XxHCS5LqSNLIXyICZkGUIMx2ldJHExqG5SofeqIqRkx3gF5dKOrrKyABfcQU5pkZgeeOPcqNqr/4lW8QEdHyDSGI/PBn/wcREf3sdVbIOXJMoLQJwEfj+5m4EfkC4Ww1ehTqti73YY1mi37041fJjwUa+eZTf4+IiAoVhsx7F99Lxw4/zLq/m1CHam1cTceM4tPiIpNFtBTq5hbP0RAQSb8jUMc09ERnK3y9uomUDVyp8/dbAznfPMooEtOAeyA7arf4uALAmYOh7KcJ4oLRTX68LCSnzTU+vrEyYB2lY3r91s4oxncfli/4dPr0LE1NSVPenRrDSzduMCGlXpc1EmO+2h3+bH1DOkF02gwlfXKRy6hcBQkdO8KqRMO0y5Ac/4FDDJ+Sx3O0vi7waQgY/SRgfCIiP8v3wY9f+VsiIup2BQ4zijouYE/duMSUovkO/74fyu8GAZptAwLrDwW2jeI91sJ1iLK+R8NY6db6/O8uHk+Oyu8MQUApQX1qqiipiMksyD0ef2dMlUfkQ342TEzxvXpwQp4D+2bRcLzKcx+15HnjYi5K80L4aeEeCB1TciFxyDbWxK9W+Rl0Y1NB3mBKdQN046kp6NI1nbh527pbSpIke6ZElMtm6ci+AzRUimL9gLf+7hUmXH7wkRAt+z2+93yoE5VVM+tBwOcapiQ4OdclrPsIkOzzL0up1hNn+d9l3ONlVYroGlUx1US9B53m+Rn+nqPiPvOoNU3Xu6rEzcDnvint0jcA5reH9aReMbRvYZaymbtTN7MRqDVr1qxZs7YLu6cI1HGIPI81cY01W+yhjUFn04nEQ23UOImfKxmtQ/EyI7gORn9RORxpF4nI9APtiNff2WLiTr4A4ksgRJRByF6Vl5PoJQOP7uA0f1bdJ7qLM184S0REbYgzXL4q2qzz8xxl3lxj8lEmL+e8cIC1cGcmOCIKlR/i5XMjWpr3ZXFESb9OZ1/+/fSj6gx3StnEcR059Vg6lqtydOlDEzjxVG/IBP3vcGx5RdoozrI37sBDu7UjXtxOx5Qy8PWaLIkXd2SCt7VWlyizg39OVKG/qqRTt1vsjWZRDO9mZN7ePvcBERH98Mc/JSKiEycFyWhAC7VQ4oiiUpLzOn5winLZe+PC/S7LZn1aPDhB/a6cjyGtOYTCeLWGS2hMOjbOa0t3Qkki/t4CdJNDRbAolvh3LYh7bGwJ2aUHHeJCke+n6Wnx+LvoeRj2JCI7dYSj0dXjHD2sXJMygxAF/g4IFrmSRGs+1oED4kgnkvtoaZ23ZTpV6J6bw3hvCHLGXMehvOeR68k1NJ1isuYYQ3mmFNFfcnYa4h+e3JeHS+jMgefGYdX7d/IAz+MiShTmi4qIgohwBjqqNV/10e3wWBwJUTAt70Gpm/o6TU0xMnQwx9doW3UV6iNiNfGko87ZCFq4IH75as4TJxmJwu/HHMelfL5Av3z/o/Sz69f5mbq0zGun1xbxkxJKbt55l/Vujx8VJO7QPK/R9jbP23ZDoR8g242jxM1XKEsOwhgl3Aeur8v+eL6jSJXGgeCTg8COnokyCK1GnCSp6h6rPKe5LErR1DPPiN10QawbdFTP6Wox7XP9eWYjUGvWrFmzZm0Xdo+uu0MuOSPF9EHAb/Ub1zlH1GgINXsK0k4TM5xb0HJmxiOPTc9CEjcuRB4hQW6hX1OlJzc48qqOsycUOOJVx0iIOkXJte47zF6oi84S1f3SFaN0kKXlPrjIPUKPHJBi8gPzTxARUQ6CD1OzErmWyhwBme4WFy4vp2OdYf+Osmi7sbGJCn3tmy/T8edEnCIyMlgR53RiVY6y/CGXTGyv8/nU2hJVmCAtj/Pp9mW+O130ksS2fYX/x/DmV3fQizIvS+bQAnt4AUnEVkI+ZXaCv1dryjHUENnNoYZA6WpQHf343nqXPeOQJPJq1gwCwdfwgQfkGg6CmOI98s6jMKZGvUsbt6QAfxMCGh2IHpAqsq5UOCI+gQ4yVSVldv0a54AMrX7Yl3m4dYsFAj46z3NU35H97bT4uh44cJiIiBb3SQeJjMfHMOhJRLYwx2jDi8+8SEREVy8LinLu43M4ZL6uusFFxgiZ4FIHqh+oiYw8yKKNFQTR8VoSie2FOUSU91xytS9v8m1AlCo5mfOH9vH5xg6vlySUsYch/Rcj0mhsiATmCsq4Ljf5+eR2pAPOyYOc8+52+TvXN+V5c+oEr7VWS1CwtU3Ob26t8XWcU0X/DxxjRGACkeSsmvS12HSs4s9cxUPImrI+PDvM84qIKN6zDChRq92hn/7i13RjVc6nvsnrz/SN1fm/KZScPPoYl7089gVBhh6Y5LW68QRHeO9+LNyT85/wM+jlx/h6Vadkf+WY0SYn5GdzNxBEKcS8Xbokcqw7eIZMjfP1zajuS5UyIxFZrNkwUUgPcqytiB80Oy3dk5TPa7vG0XY3lPsz7jUoGKpo+jPMRqDWrFmzZs3aLsy+QK1Zs2bNmrVd2L2RiIi1HjUUG0PB5Poqwx61rsBpBw+jcXCVw/ih6lISQ7c2Bp0+CARCMhRroyTRbwksHELZorG6zL9rCXV6cpohNScv9PWx/QyB+WjaWpyREoVMiSGBR3IMA2zfEgWOENCyAWN91fXBRccANw/oISfJ6epcnrzM3vglURDQzo1VuvCbn8j25xlC6eK8r18T+O9dqNLcuPQWERHVNwSWyJkuLtD5VHwSGh/jMVNKklcM7qP7GZ6aKIEmrqDfD5bNHMk1XxiHWhAS/leaCioG+aWB7R84OJ+OjYU8022c11A1A/exDi5eWiYiokV09+FjnxhtvnsfFgQJ3boZUnVc1k8DDdWNKo/vSflDHnrCbZzj0tW307FLF1m5qAGVllJBJrxQ4PPZ3ub1Nj4hZVC9PsNiN29xqqJcFop/3ud9r6+JxrELHdaxMYZbX3r2i+lYHR1Hlq4tE9EoGagDSDkDbD+TlYvuO3wNI3RzGZJWaZFj3QtziCjrSJkEEZGPko4CYLSSImBF0GSuFfneS9R6rBDPlQ/oeeOWkLOufMIdhoYgpIyr9X8SDZcX5/l54CoCy5FZLnXzFyVtcHOCr9vki8/yfhOZ1yo6tTwMvefqqjyfvn+Br9sQzzpf6SP7BrqNUIKjbtD2MNozXmIYBLSxvk5hR0GdKE0ZmrSaKgtzsS6aTT6Pm2sC0xKUz/YV+JjLx+V18sTcIrbN8+0VhHxW9fm6VKoM6763Imt8dYtJWJHqhhJjrS5f5O8vzMv9PwdY11T9ZFy1PrE0Oi2G6zduCsFuDUS1PFTyyhWBkR86E9L/Vbi7tJCNQK1Zs2bNmrVd2D3z/13XpVi1TkhQVFzbgmeoqMyRx2/3riEoRBJVhNDsDCP2eoZKw3M4YI8jj2RxpLpjhCg+31hhskRuv/gApTGOHNzitPo+CmXRd9SrypiDY8+gcLtYFu8lFlFK/r/yAA25KYj4WBxFw56eK6Tai/drjY1t+v6/+j+p64t3tP8sRxh//C/+UyIiarckav7tmxwB9UDkevBR6bjw4CMniYjoV69yNNtSpJZxdDcoog9jpyveYhzxEinDI8tlZMnU0CN0QpUEGCRhFb1WbzXVtsBYcUtMtjn90Nl0rIo+m3WURWWV3qUDSvnaRhPn+W469vU/+FpaArAXFscetRqyFt2Eo7GyiSAjiQw66Eyy3OJos666PbjocGP0PisVQW2mUcQ/Ns5/u20hNwy66F24wV56W5HyZqY5GgqUAMV1aD0vLrDnPjUh6/vLL36ViIhe+dkrRER0/vIn6ZiTgQ40FExcdY8NUSCfKwAxcnX5marZ2APjMhaHwkDmvAgRibjNJKB1oE1ERJ0qr6ETx1mbNVRMnIprOjXxdx44IQSsKUQtLiLDoidzWEY3EQ+9TisnjqZjO22e/5tKD7jW5Gu0P+Y59xQBxUf5RgZ6ysFAyHB5lADlcrwmqkpTvIB5LZZ4bHFeUJDL63Xy9igEjeOIup02BYFsL4B+ryG8aVEbs+6vXmVRlquXRaP6hcN8zfIQysn48ux3IN6RrpdEnhtDND31gJ5kFYGtW+exQkEQkRzul3PXWajk4CHpenTyFJO2DAHMkCSJpNzFLN+BWmPvvfc+ERHtm+N1UduU9Z8/skZOcnflWjYCtWbNmjVr1nZh9yik4JDn+SP9OQ2t30Qmk+NC5c8aejfEFRKdA0WOsYeecJ2u0OMTULhj1DlkffGWmsiBZlz08OyLh74F2bOxWclzeui0YqTl6lsSJRQy7DmZ7imlimDxAejxAbDyoVL57/Z43wPIRm3dFM9mYzWkcLg3ZRWDKKGleky1ntDxvTnG8020feaYeGNf/cOvERHR9RWmgGd98bLrq1wcX86zzxRmJdq+gTkpo1tOTuUUN3fYq8xk+ZwmyhKBzWT5OnW6KiLa4s/6A/4bKR+tUub10EfXnEvXJJf3wtOm1IndxYLKOTuIkI2HvLYmFHdy92auidhbbrW3adAX79NFBw0XQgqh6t1p+obWtlbxf5mHw8irPXCKo+xCQZVwYX37GRSZ+xLhjaFbyCTW5DCUsTyEG2YqEmWaov4JyA8ePXQgHXvi7OP82VGOxP76//vbdOyjSyzX1sW1SCJVzI5oZ4gcZKxCkiTZu/nGFslNIhrX5VGQymy6vC73jcmaW5zlqG97g++5lsqPukO+/00Uvt6Ue70z4OfLAGhGxpVrNVHgOY97/NmQ5F7vIDffUtfWq/I6DhP+nVsQadMdn/fp4dHaD2S97J/iezY2ZSLqvsnjGTeFR+ak6rEbDXqf0jrdvUVhTM16i0KFYpjrawRyAnXMhQwf19YGP2e3t0SucibmZ8hDC3yseSUMEeOeNZ1e9OE7kCaNIelZ8mS+Z/BY8lXnLg+cm8mneL6LJcltby39NX8f76RA5dJNF5cyyhofmJZ3xeSjhtPB6NGwI8/wXNQjVwlgfJbZCNSaNWvWrFnbhdkXqDVr1qxZs7YLuzcSkeOQ53sp/EpEFCcMAT35NDdaPfHA8XTMRaeHEMnzJNYEBKPKAW3WnEBcQ6jftNFdpeQpin2ZIRwHxIvNy9I5IDPDUEBZlSEEOEPHZ2ykrZRUDDXbgWZopBoOZ02pCv5qFpFRTTEdAFxHILVi7iS57pu0F9YLiT6oxRSF4ue8jE4c4YDnJlDdUZ59htWTnn+ZSUevvyGdWn72/e8TEdE0VJQckoR/C4SCm+jKcnhWSEumYukWOkdEKrdeAuHh8oY0zY6Ghh4OLVwFv0eA22qA4XXnEBcEo9WbTIrqqfKEAhSfDEFpMBR4a9DtUxLvDbzFKQqPQqWvHILgMI5OQpUJgfnLKA06cghNqbNCYDhz5kEiIpoCbOcpbc0mutLkkKLQupumfGqnzWvz4uWVdGwHJRzZrECGpvFvo8nXbjCcU0N8fF945FEiIprZJ2VD3/r2t4iI6Je//gURESVK0co1DZZB7AuCvdW/1eYSpwzySonMhR7wzCyng5KezGs+x2vTQKuJKrHJoNG6b0g6Y6KgNB7zZ1nMecaXOS/nGDfMO/z7SJWltO8A4TZBWGlu8bx0lM5ri3jtDNF0u9cQeLKFZ10Pqay4JfPqoOyrAPLQhbpqYt6NKNojdbM4jmnQ7VMYyvYMZJvC8+p26qLpfQ9/QwX1v/E+w7nFkJ9/jxySEq86Uh0DpAEWF+UauuiaY7oZqcY4NFmJJXaYAAAgAElEQVTiNR5EQqwLBzxvJZ/nrVeXtbJ8y3TuwvkpVbQQ6cVsBl2zErnmfRBVp8f4fq635RnW7gwoUunGzzIbgVqzZs2aNWu7sHsWUnDJozCUhKuDbiflCntvA0W26RjNR3gFiWpCGSNhbWjOifL6TETRqrNX0Ffah4VZJpsMUADevClCApPwBAdDiSAMvXmAwvxAdZIZ9kGQQcJb6x826kzcMf3mBqFOuiOqgseazYj3fvTQEcplVZX2fViUxNQYDmh2TiKH/Uc4Av/Z975DRESPfOnldOz7f/k3RETUx7x9+WvSxeXp//a/IiKipQsXiIjor/+fv07HsohiCiAu5PJy/PNjHIIasYSMIpAFZh5U/9OZCm+jhiXSV/M2gFeaL/L13L8o52Vo69soel5aF7KCjwgt6UPoQfUK/fCDi9TryXq8H0sSXnueL+foYw2XoLlZLQtJbv8+Juw8cobLhRbmhLw2McGRhFl3po8lEYuREBF5IEzpCK8Dz/ji0jLvryLEh06Pvf+uIjwUgNyI/rJcCxOYR7gG+2YW0rGHjnNPxnffYtGNRlehCHCrA6AJkbqGripp2QvLeB7NV8s0npVnwwxEOyYrvCbiUMYiRJJVlE7Me4JwOCD/hMFoSQgRkeswghCBFBapSKoDItZGm6O+9Y7MYRsoyXZNiHw19BveqvN1qO1I9xIjhDGAEEBTzWvX4W2FMchZqu9uAhGI+cefJyKihYNCDhwkHoV7E4BSkiQ0GAwo48s9LugIz2W7LWuuiH6rGZBE+115fl7vgkR4LMJ35Dm9hfK1FsRTihVZs+fO8Xx96SFcX1UGd2kZ3YEU6maQTNNMRT/DfSBXiUEIlViI6S5ltMsdhbKQwxFx4jPJtFRV6jH9G7pF72eajUCtWbNmzZq1XZh9gVqzZs2aNWu7sHsjESVEFEWUKCZJBGWgW9CdrNWkTZCLmtAclHkcleCNAefGwJlCBe+ausw2lF22GkL8WdzPEG4BtXDFWRXqo9ZuMFT1XyANtaCxGivCSYw6vp5pqaYa90ao1esBeusrmC1AO7fr1xjKuXBBiB7DoUMdBYHcj2VzeTp49AT9+b/8J+lny5+wCs8bv2JVmTPPv5COvfcBtwJb+YQ1H43+LRHRl7/+JSIiygC6ikkgi5qZX2AkNzZl/lodnpsS6vRMg2kioqs3+XsZRRQykM9yAwStWPlogC4d1D1GSollc4fnrNlgyOviktR6FrH9Ko7B6Qn89rc/eIV2mgKT3Y9FcUSNZiOFhYiIStBcnZxkCHAeKQQioiLqnIdGSUWtLQP95gA3hzmNCaF1FZgPA0fOp9ECzIc65Fml3dyAupOBfokENjMwnFGVISJKUV0DbykizOMPM7FoeYkVZn78i1fSsSZqqz3TPFzdm0m0R1gimUMKaLO7Qa22rLnxSW5V2MNa1SSxHrRsO6hj/PX7y+nYhRUmnvQBmw/0PQsMtIs0Qrsjz5RBl893gJSO0U4lIoqQrhp2pB2XeeaFpsm2Vq3Bs8QFlO5EspaM0pHj4PuxqnWcOoG/DLO3PKnTDqKE4j1S2wqCgG6t3xpJO7mAwdN6UHW9i3nTrnKZiIiGA1V3m+V74/xlnpvHFuSYo6Fp2M7nulOXObq0zNf6qSN8T02U5Po+dIx/11ePDUNcSvDoCfoyF308s4e4X7RKnoF644iPL4rlWnRwrXeWkAJUnKGTs9EIOe2zzEag1qxZs2bN2i7sHrVwE3KdgBKVxO3BK774CWvTrq6KukwOkUZg9FPv4Kk5UIvQTVwDeNht6KKee+eDdKy+zd7iA1DgqY5Lo+sYyi4xyX5CdHbpoJFuT2l+evC4Yng40UCRUeD1NuEZRyO+Bp/PyjKf69a6REvZTPFT5Tq7t7nZWfqLv/gvaCov3vK//d9fJSKiiWkuF2rtCLkhQbT8x//8XxIR0e/90TfSsclxdKoY4ya2Y6/+Nh0bBKxv2UGEst2R+VtBQ+k5aLkWshL97LT5e5Njcu3aRgsUJT5F1eWjg0itXGFCTmlMyn/qiECNDm2govgIyMIOrlMpIyVPH39ygfp7RCKihCgKicJErt/kOEeZvmfKH+R8xsbZg84a0pgiwhniS77Iv3cSXTYAxRtcL92g3njQZn+lsuwvg65BbdUo3XUMIYnXZK8v67uLSCzBNcgoPdXZaSajffNPvsnbVKjNb97hMqw+Shd09LN3qsNs9VaPvv3a+zRsCBnwzFGOQCvYme+osiVEkn3M3S/e+jAdW6/LORARJSNRmyGZmL8qwjAlC0CdvECiTYowB6FsO8FnWTS1N8o6d7IkkrWaS9DEHOjPMCMR2/Qx1qr2iyjZUiUXIdEetYxnpKJSLtNOQ4hPhiBU3+bIO6MUhbo7PBd1EKfGJwTVyuT4qK5sQ/+6pshtUDDqI+rOqTLFCi5sBs+SSOkZv/MR33sbqquX+Z6bqo7J88kQDA2XzFXXNQMUKAPkylWNz426WRklT14iBNI46t/1hNsI1Jo1a9asWduF3aMWLpGfcSlUb2cP3uHyVc4D/uwnr6djyYv8xZkp9pwd5TmYUo8iPHRX9cYzpSfXVrmU4cpVifDW19nri+C1P/PFR9Ox6RmOCHJ5yS30ESFHEXtQ25sSsZk8VREF2H3VhWR7nb83wH7yKp9ovNn6Fntxg67qJEMBRdHdFeF+niVxSFG3Rq/+6IeyfdDIS+jy0axJwfVTz7xERET/8D/5z4iIaGJmKh3z0cdxgOjH6LASEU1NcslFsgO9UCUoYXpBriKnqTULhvjPoCnnm4dHaHp4DlQ0nkEufN8C5xHzBdUHc5PztiFQiq5COfptlEEhepgvyLW4td1I9Yrv11zXo0JhLM3PExEluNZGDGRtQ3oKRui4ks/Cy1blL2aaJvEPT1Hog8gUmaPUQYl7rG9j3aE0KKvKNEwAWVLzVkEXG1Ne0lICFPkudKYH0NxNZFsFNH0tVxmZ+OKzz6ZjK2tLRES0tMz50UxOfhffZYH53VqvM6Bzb14mj+Qeuvg+6zZHiDiSkQjY5BH5/7pJSdYfzS3rPrFGrCJEaYOjhAkcrOMI3I5YIWWx0fFOVOmexxGni9R3ooQ9YlNyAaQiUTrHhGOIjM52Sa5jCR1gEtRPhAopix1vz0LQcrlML774PH33uz9IPxsHV6WMdVXblJ6fOawTM5UthQz1apxzLkC39l//ZCkd2zdZxDZRSpKTe+PsSZSQFPmkNpuCqJw8wp99oax0daGG4yInr3RHKMH7Z0BmbpVmccxRbwsdpbbrslgMimXQyiQQ8Zh3lsaoO5T7/LPMRqDWrFmzZs3aLsy+QK1Zs2bNmrVd2D1BuKVymZ565hn67veE8m4aY+8gnP/lz99Ix5p1hjgfeZjhiePHpdXSGPRNDQFjqFoPXbzIEM7lq6Y0RhK8GyC1lABdPfHFx+T4RmBWfFbhz3J5hhSuXrmejhlK9v4DTFoIlIrSTsNoP2I7JAnoHijQTShxJL7SeXRdImdv/JIwiqher1O7J7DE/gNc1vD0i9xYO6Ng8Sde/BIREU3P83e02gpBa9QQBsYmRE91GiUaBg/rdARS6UONZghItato7Abq7Claec/wwQGDJUrSY7yEpuYgKTQU/NyF9qfrGcxTyDOT0wwzTk7xcVaronW8VeuksN79mue5ND5VoawiShGg21qTUxSm2TsR0U6Loau19Ut87JFAcmMVPkbTSmxqSshuLtSczDwbJRwiolobECyaqCeKTJJF2U9lXOAmA906jiljEQh3B824DbSsekhTEPC1iKGmdeCAKN/sX9yP8+J7ZaDWWBTtFZ0FloTkBpvkqbWaQ9lSaLA6hdO6KZaJVIQqk3Kj0XXgKMg6SSHx2yHcEGUpBPWgrCrVGslXydaIiCgzMKUqar0YghsasQ9UH6+MY0iOfFxeSVStVjY4XeVso/xLrX/H9SkYyBq5H8tkfJqeHaNQNUbf3OR7bxoa4rpB/dhYFceA8qGOPKdNCzEHa/SDa3I///YiP7vzIBO9+ra8aiYneO1NzfL9Mj0jZMKFaR47OCukI0qJcriGKqVg2h96UMRrRzJPE+N87FPQP67OynXK+ljTIA/VVBrq/aVZiugtuhuzEag1a9asWbO2C7unCDRJYgqCIUW6mDox+rMcvW1vSwL6jd9wqYRpTjo3K57GGMgLhpRw8cKVdOwnP/olERGtXGePpl6TxLVR9zeau7r8xeiBBiMF1PzvXoeP4dplSQ43UYDfrLFXVR1XDbVxXH0UbG+1b6Vj7Q7vOwINPZuVdgKe75Hr3WN10O8w3/NoqlqhsRnpsPHSn/wZERE99SIThn79V/9bOnb4kcNEROQ4hkyhIwf+W6lyFLdv/5F0LETpg5MxBA25vj66SkQQSNWcCBP9DBXpJgu6epLwZ6HSMfWBNhjSzHUU8RMRDZp8rTsoXZqbEjThOEqWZhDFzc7MpmPn3pYyhvs1x/Upl58kTzXpTmLoq4Kc4CoN0UFgyEC85r1Y1mm9xZ9tNni9jalG84U8IsiE14mZWyKiKGGv2TDuDWJARDSAt7zTkDKLDjrbFAr8g3JJotNWCx1IEM1klY5tB3rHgwCdSCqyn4lJPtZiKYfjU1FHZo8LWXyH4skcDTu3E9eqIR9TPpJj84xAASJQT5F7THmJbghtLA2qHJTPqXIHowPiI+jJ9OV+dhGhO4kQEx1DePIh4KFC+xgiHKYMpUuyXjLQ73VApiuW5Hm4DTGTxDGiJrImYnIpUgIZ92NJElMUBiPryjwHzZSEiig2DIxmOeZbPdtMZYqZST+nomYgk4bTs96SOeqBGOcCSckEgtwcyPEaP31UC+vw3Df7fH37gaxHc6s6xNesti1dXFZu8DaCiK/B5rasiwCIxKkHH+b/J3LNn3jiGfr2vxOS1WeZjUCtWbNmzZq1Xdg9hUrdbo/ef//9kd6GpkuHn/bXE8+ug64RS1fZC38leS0de+ZZLj958CEuIP75L6T85Ve/epuIuKidaNSj9H3e/tw8l2hksnIKpovAYHh7nq7bQqG+6pxgeoNuwzPP5cTLNJ1GGqBYd/tKhhB+Rxb0bZ0zcB1Xubv3Z67rUr5Yoq/8o3+efvbYF79GRETDHqKdsuSVC2X25EyPVi1H5YBeX0SeeP/ho+nY6nWO/nN5dimrVYli8lnQw+GVeioETdtfqnySyR8WQNF3PcllmBnMGu9XdfXx4f4WCrzvQkU8wto2R1ynEYkemZOxp86eouvbeyPllyQuDQZ5SmLx9r20cwp6dyrZPdOVaIj8lKcKwk0lVT9mb3m4I7lTI+UXI8JyXfHAcznedxZyhZ4v2+z3b+8MYuTnzPUZG7s90i0h/19WheRlU7oV8t9yINdpMOS1FRkRARJuQEVdl72wXLFEhx9/hkjlHR0jQuHwnEeBuvdwngaBMmIrREQOmRKq0W43RHJLDgPI/PUkimnj+rkmXz+h8o0J0JlY5sCUtjjE3YQSnSbFs8REXp7K0RZN/hjozDAvEVuCPK+LiNpTZX2xSxQ5e5N7bjRa9L3vvUKhypWbf9cgmjABgRAiIh+Sfz3wPXQ3ni7KpEwnrbySB8xU+Tlz9qEHiIhockpF9bj9jx5hFKyguCs/fY3fEa2edGqanWAE7soyS5WaLlpERGNl9AqtGGRBnl1Z3LMzE3xPTM2orj7ojDM/z/PtKGnTmekK+Rpq+wyzEag1a9asWbO2C7MvUGvWrFmzZm0Xdo8kIlbm0SF3bJL5yMBHijTi4v28vMQlALfWRSf36We4/GRtlTUw33tHyCAeEvdGa3F8XEouiiUOtTtoVPvB+6KTu7DA5JJ8XulPZk3Taz7VWCnjOIlpBAs/QkGvnSYUSAIeyyvySATMxs/d7n84jkOuuzcQbqFUoYefeYnGVSPkFF5M+Phe+MY/SMeyUHUKQwOrC2ziADZ0wZg4eOxEOvabXzJsUijKPBvrA5pPoA/re6rcAPtLSOCOWp3hv3DIx5AryNj4GB9P0GP4JKfg0DyuceTwtRuqjj+TUFR68fkvERHRqTOn07HqviP06rnLtx33bixJYgrjHjla9iUlKaD0QEHSWRBGslkDwQrMF4B4M+jz31CdT7+HuXQZbqqoxs9eBt0lAJnVL62lY13Aju22pCFC6LfGWMsbm6rDB0gr0+NMVnnguFzzKtIwPRAzWrcETl9fZ0WZIfRfFTJHsYIy98KKnktnx/O0OCtdZ0yXmxLITGtbQkxsQ/d4G6pZVzeENNJJRhvZxyq1YP4dg4ToDOUZ5gCKjJFkGKjm3i7Wv6u0jI3SkenuEanyGUNuipB/8hVUOhg08Xue80gTIM36QJnYCFnLS9J93q/FcULDYTiyPT1PRERDRVjycYyGpLazI/CpUXoyHMJyRqUb8GrJ4x4/NCnX5sBDXyAiouoYkzaLRUk7bK7z8/zwpJBKZyY4BXh0GqpyiYL0QbYaGB1z1Y1lCDWnPlItPdXi5eYtvofe/4BLtXxfyszmr72Zdu/6PLMRqDVr1qxZs7YLu7cI1PEocKvkl1SBshFHxZ9MXt7JedOrEGUR4+MSER3ZzwXmK1cvEBGRq1jaY0WUk4DunFXJdh+6oxsb7CG88Zv307GHH2JizIEDErGVEVXV6/D+fNUVoMDHOjPFBcTdQLzrTpePOZdjTzVWpR1Bn73f+A7+R05Fv/drjuNQxsuSq3VUEXmbnqyVinhvxhtzTIGzOj5DvjCe56GTZ9Oxs09yT9F33+DyoUxO9pcFG8aIMsRDiQZ8RGWx0hz1fFNczft58ISQnFauczQ1bLL35xclOjUlRC2UG5WLMo9nn2SPde4Qk4j8CSnr+cKLX6Zi+X+mvTDPdbg7g3L+zXylGqwqEDAENtMBIlb9BjOJiU54/jodITqZIu4KSF9lpftpSmgq6DK0tHQtHdveYvJQbyDXYDDk7ZpowElk3sp5Xtcz44zMVMdEgCLxeD002lw+tHZrOR27ubqMc+drkS/K8bl7RGYx1ui06Xuvv05jCv0o5FAeVuW/TdVtpwtkI4GoxFDdggPDOkyjKxUZYn7SaEt13HGAHMRp+ZuQukzFyohYQlrKAe1cRTozwhsJniUhyfVIPEYcDLcpo+5PJ0ZnEjz7ElfXi/lpr977tiShMAwor/TCu6brDuYmUlFzE/fqcGjIajJWgUiN+d22Qhi3dnhd9tr8rFy6LkjKoXVE4Ahdc1mZ71aNEckPQxFlMPVFA0SQuYIcuwvN4SCAoIgiX/kghRpkLaf6DxOQJKfKx9BSpLLG0nXqD393hx1tNgK1Zs2aNWvWdmH32I3FSWnNxtLOB57JMSqJLHirhqJdLosHMF3l7Sz1OYdx/IjQlp0M52wuXOHcqelrSERkmrl3h+yhNLsKq0Yus1CWSLcCwQbjncYqmsuhz+SRRc6/1COJQG+h60aIzupaWi5C/si9A10+2SMZPyKiJI4oHLQpVuUeMTzbHHppao8rbdxhhBQ+ldvQY5m8zNGTL/weERFtbbNYxOqS5BRNAXUWJQU7juScevDSiiUpj5ie4Yj44Qe4X2lJ5YkvfsR57mIB9PJ5EUSYW2TUYB/o5DNKdOOx514kIqLJww8SEVGgvPOCl90z4QqihDw3HJE5M9fWrOuB7pJBplOFk/4+/R1QgHwBkpXK4w9wPXM5lE/lVTcPdP+ZmuS5yXiSq+tC+q/V2lCHbBAgRKAqH11E+cr8HK/vakXKE4YBRxYmwlhZWUnHXDwWqiX+vqtygtWqOZ5LtBcWe1nqTh6igeJOUMz7D1rI5WekNCHBXEWO6TGsIgWUEXnmHlS5MiP9l2B6kkDdG+iX6uCvr3OpJr+p1lwM1M3kMn0Fn5lewA4ifEfxISKgEoQSnERFlQ7y6THOYYSr4Tq0V+1YEkooDMORTjUmkjTlLHpP7TZHkn3k8nU3HnNPFIu8zm6qkh2jtVPv8rl2bor4x9bwYyISmcCpWeka5eX5+dHvqtIu8E+SAs+zowQbPOSoDTVB50eHuI+bWxz9BgrJ6OMadNFX1vTmJSLyfZ/6PV129rvNRqDWrFmzZs3aLsy+QK1Zs2bNmrVd2D2WsSQUx3GqPEEkcIbhFsRqLDLNYaGLmM0I+aMMss3iAkO3/aF0cbkFQkMLye226g4So9TEAVml1hT90VsbTLLYNyv7efG5J3kbUIKJFHmg0QVcgqj/7IMPpGO5PMMLyysMWd7YkNIBw5SODU6hCEauE1IU7w3cEgYBbd68SeOzAnW66Pbg+kYFScHHiYGW8H8FxrgG1jLdO0igjrnFw0RE9BWUxHzw25+nY506w4Xb2wyZZxyBTw4eY4KQgSmJiBYPcalEPub5vnlLtIf/8X/+F7wNwC45T5WqTPM6KJSZPDM+LZD+9D4+PhdlGbGC7ZjUsjdlQ0lCNAiJoqHMm+nUYUhEsZrTAUgnKWkruZ3oYbrL5JQKkCEb9fr8+5yi12ewlvrQJy4WVToChLFGSzSbyxWGuhPAZ4kiu+SMWKm5D1V5glGUGt6BLFEpo0k30haBKsFxo8xt378/c8l1SpTPKt1ms8Ywr67SdDa6tYYwN1QlJEZT2Ogwu7G+L/HXpF9UKidOG3DjGisIP8FaDRXBkDLmHoKWcaI0YNPtmybw+kzRZDtmeNLztb4uf98PakREFChlJj6fPSJvJUSUJCPpHQPnZjImJaHgapRJddIuPEpDOG0gb9a4vE6MIlm6G/Ve2KkxnJvOs7qfC3gvdLryzB9iLkyzbEd1KApBbnTNe0Edu9H0NefqkMypIQkN0FEnVGpXxXx+pBzzs8xGoNasWbNmzdou7J7ZF2EYjmi/mogzY8pYdERkolO8p1tKD/Rvv8M9RRN0alnflrGPrzKZJW+0abVHlOqIwltUhbOmd+XyRen5CS0GegS9SB88eTgd+9XPuQTmt+fO8+9JCElj6Dk5VebIYXl5XfYDbyx2Rz1X/rd3W2Hybs33PJqcqJDjiDeUAbHDzOmdCqLT66OuRRTzPDe3mGq+eUPKI5otJgqE6Kij6d5ddMJYv86e8dlHn0zHXv7aV3m/GUn4l8aZENDrMjJwTPUPDIYB/nJUVyjJfopjHHkWKvzXRHV8ruZaIxJREb+zhz5gFBG160QZVQqSdRA5Yk4HKhobmkgHJAdX306m/MVETAN9X7CnPxjgvFTJjlPkz2o1jvzn56QMaHKKI8N+IBG/C2ELs+RCFVAapMh4+oEioxEQgsThz4ojErcgoUXmvJQgh7vHPrfjUOj7FChEJEZ0HIMZ4usSDpyoKfPwlRhBJgY6AwKOp6IlE5VG+GwknkM5hUG3ErWmPOzHzarSCddEVyhjUcGKIXEluN8c1dmHQHyKTY/RjF4TQGVAoMyVdAlaQl1vlLx5P+YmzkjHFYkkcSxatAFogI/9p4IFJBrlRoxjckqQkYV9jAJevrQ08l0iIXm1G/zcGWrCDnY9GMjxedA/L0B8O+vL87aMHsMN6K53+oIUmMdEiH276ply6sw+bJPP78aKdNuKI6KNuxTDsRGoNWvWrFmztgu7xzIWlzK5cprDIiJy09Qaeg9mdVmFiRwgkaU8gFd/y1HfDjoANNri9ZQRjWRQQuBkNIUcXiLw8HxOvHeTK7lyWaKrH3yf83ljFfZeilqUALnZj68uExHRzR3pclFEmUdMEBJQ1HYnzZkg36FycN4edWIhIgrjhOrNgGZGCsIxF2nlhIrGTB4A/w+G4tl1djiCXl9m4YrtmzfTMR8yhwl6FzZuSN7y/DnOTZ/7kMsWXnru0XRs7SpH+rEq8ekP3iUi6REaK3p4NOSi6pmDXOIyf/jZdCxTKI+cj0PyOwfH5biIPhLxMvmY9yY/5LkRjZVbNK2K+idQomNyIjs92XcfiX9T5pHxNPIAlCJNSOvOOOb+4Sulgija2OD7Yeka55wPHV5Mxw4d4eh+fGZ/+pnrmx6YPF/rqyLY0Nw0OXr+o2X42l1eDxPTfO1n9omUnomaPJQu6Sjf2aN8c2quS06+IKUnJLKgrsf3nhtLWO3gWWJy/476nRPis8R8R62LVEDhdrGA6FNojj5DU2riqcg7NIIDCJI8HR0iYovT3KzsJwZy5aEUcKCQKpPXCwbo9KTKBRNKKNnDeXdodN4+bXdC0EroM6tRjF5vdO31VCQ5MWFKphyMiQSgyb/nIZDTbt0e1SYqeZyHHGgpzxOeVWVfC4t8r7au8L7X16RcxpyhQUmPKIGdb/w+5AQrORyDkijNufTf/4//L92N2QjUmjVr1qxZ24XZF6g1a9asWbO2C7tnCDebK6YKKkRE+TygBqPD6mi4ERRwwCW+0lgl/Dtb4e9UfIHs8iCbpDCLgnBNUt9HiK8VNQx6XPAFgnjltbeIiGgAuHZtQ0L8fInD/xK6hOSLQoZJUJJgYJesKt8w+4w9cwyKrBNFUjJyn9bY2aHvfu879F+f/W/UpwbGHCVo8X8MXAgYNFLlHqCKHzzJGsQHDh6UY0bJxCa0LDcuiNbq9Ws3Rrb9xq+l3Oj4GYZkDzwg5T/BwEA1/P2yKkeZ2fc4ERFNzjEsmSuIyk4K2dxRaxUkAEMm2kOYXFul4tOLL41TVTVWz6cKW+j6EAu8G2LqY89A+prcNKrHqhVp3BR+5M8CpYrz01cYgl1ZQReXQOCt6jjDTeUxmTcPSk9tNIyvbyldVsd0ywCcruD+bpevnV/g3xVyCtKMTYeXJo5PjiFUZJC9MMdxKONnyY912YapiUMHHFUeZJplB4A8E/UIS3xT9mLOV7F7sAlzGbJKstpAlmm5g1qDLo2Swfh7KI/Avejq5tfQtDUdpcJQdS8BqTIDQpJGSmNzjUCEVNU55Hsu7dWSd1yHvFyGoiC4bcysE20GbiWcV06lzNrQuU0VjBTsurJyA99pj3yHSOYbPx+Bk1NtdXW+rRZ/f+NW7bbje/9DVhnyTPmQPjM/y/MAAADKSURBVAXcjwcPcOrjz/6JpIwWF6bMDomIaKwq93UcxbahtjVr1qxZs/bv05x76TPnOM4mEV373C9aO5Qkycznf+2zzc73XZud7797s3P+d2t2vv9u7a7m+55eoNasWbNmzZo1NgvhWrNmzZo1a7sw+wK1Zs2aNWvWdmH2BWrNmjVr1qztwuwL1Jo1a9asWduF2ReoNWvWrFmztguzL1Br1qxZs2ZtF2ZfoNasWbNmzdouzL5ArVmzZs2atV2YfYFas2bNmjVru7D/H0f0cDnxbuy8AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 576x216 with 10 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plotting some random 10 images\n",
    "\n",
    "class_names = ['airplane','automobile','bird','cat','deer',\n",
    "               'dog','frog','horse','ship','truck']\n",
    "\n",
    "fig = plt.figure(figsize=(8,3))\n",
    "for i in range(num_classes):\n",
    "    ax = fig.add_subplot(2, 5, 1 + i, xticks=[], yticks=[])\n",
    "    idx = np.where(y_train[:]==i)[0]\n",
    "    features_idx = x_train[idx,::]\n",
    "    img_num = np.random.randint(features_idx.shape[0])\n",
    "    im = (features_idx[img_num,::])\n",
    "    ax.set_title(class_names[i])\n",
    "    plt.imshow(im)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Convert class vectors to binary class matrices.\n",
    "y_train = keras.utils.to_categorical(y_train, num_classes)\n",
    "y_test = keras.utils.to_categorical(y_test, num_classes)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[0. 0. 0. 0. 0. 0. 1. 0. 0. 0.]\n",
      " [0. 0. 0. 0. 0. 0. 0. 0. 0. 1.]\n",
      " [0. 0. 0. 0. 0. 0. 0. 0. 0. 1.]\n",
      " [0. 0. 0. 0. 1. 0. 0. 0. 0. 0.]\n",
      " [0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      " [0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      " [0. 0. 1. 0. 0. 0. 0. 0. 0. 0.]\n",
      " [0. 0. 0. 0. 0. 0. 0. 1. 0. 0.]\n",
      " [0. 0. 0. 0. 0. 0. 0. 0. 1. 0.]\n",
      " [0. 0. 0. 1. 0. 0. 0. 0. 0. 0.]]\n"
     ]
    }
   ],
   "source": [
    "# Printing sample data\n",
    "print(y_train[:10])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = Sequential()\n",
    "model.add(Conv2D(32, (3, 3), padding='same',\n",
    "                 input_shape=x_train.shape[1:]))\n",
    "model.add(Activation('relu'))\n",
    "model.add(BatchNormalization())\n",
    "model.add(Conv2D(32, (3, 3)))\n",
    "model.add(Activation('relu'))\n",
    "model.add(BatchNormalization())\n",
    "model.add(MaxPooling2D(pool_size=(2, 2)))\n",
    "\n",
    "\n",
    "model.add(Conv2D(64, (3, 3), padding='same'))\n",
    "model.add(Activation('relu'))\n",
    "model.add(BatchNormalization())\n",
    "model.add(Conv2D(64, (3, 3)))\n",
    "model.add(Activation('relu'))\n",
    "model.add(BatchNormalization())\n",
    "model.add(MaxPooling2D(pool_size=(2, 2)))\n",
    "\n",
    "\n",
    "model.add(Flatten())\n",
    "model.add(Dense(512,kernel_regularizer=l2(0.01)))\n",
    "model.add(Activation('relu'))\n",
    "model.add(Dropout(0.5))\n",
    "model.add(Dense(num_classes))\n",
    "model.add(Activation('softmax'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "conv2d_5 (Conv2D)            (None, 32, 32, 32)        896       \n",
      "_________________________________________________________________\n",
      "activation_7 (Activation)    (None, 32, 32, 32)        0         \n",
      "_________________________________________________________________\n",
      "batch_normalization_5 (Batch (None, 32, 32, 32)        128       \n",
      "_________________________________________________________________\n",
      "conv2d_6 (Conv2D)            (None, 30, 30, 32)        9248      \n",
      "_________________________________________________________________\n",
      "activation_8 (Activation)    (None, 30, 30, 32)        0         \n",
      "_________________________________________________________________\n",
      "batch_normalization_6 (Batch (None, 30, 30, 32)        128       \n",
      "_________________________________________________________________\n",
      "max_pooling2d_3 (MaxPooling2 (None, 15, 15, 32)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_7 (Conv2D)            (None, 15, 15, 64)        18496     \n",
      "_________________________________________________________________\n",
      "activation_9 (Activation)    (None, 15, 15, 64)        0         \n",
      "_________________________________________________________________\n",
      "batch_normalization_7 (Batch (None, 15, 15, 64)        256       \n",
      "_________________________________________________________________\n",
      "conv2d_8 (Conv2D)            (None, 13, 13, 64)        36928     \n",
      "_________________________________________________________________\n",
      "activation_10 (Activation)   (None, 13, 13, 64)        0         \n",
      "_________________________________________________________________\n",
      "batch_normalization_8 (Batch (None, 13, 13, 64)        256       \n",
      "_________________________________________________________________\n",
      "max_pooling2d_4 (MaxPooling2 (None, 6, 6, 64)          0         \n",
      "_________________________________________________________________\n",
      "flatten_2 (Flatten)          (None, 2304)              0         \n",
      "_________________________________________________________________\n",
      "dense_3 (Dense)              (None, 512)               1180160   \n",
      "_________________________________________________________________\n",
      "activation_11 (Activation)   (None, 512)               0         \n",
      "_________________________________________________________________\n",
      "dropout_2 (Dropout)          (None, 512)               0         \n",
      "_________________________________________________________________\n",
      "dense_4 (Dense)              (None, 10)                5130      \n",
      "_________________________________________________________________\n",
      "activation_12 (Activation)   (None, 10)                0         \n",
      "=================================================================\n",
      "Total params: 1,251,626\n",
      "Trainable params: 1,251,242\n",
      "Non-trainable params: 384\n",
      "_________________________________________________________________\n",
      "None\n"
     ]
    }
   ],
   "source": [
    "# summary of the model\n",
    "print(model.summary())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "# compile\n",
    "model.compile(loss='categorical_crossentropy',\n",
    "              optimizer='sgd',\n",
    "              metrics=['accuracy'])\n",
    "\n",
    "x_train = x_train.astype('float32')\n",
    "x_test = x_test.astype('float32')\n",
    "\n",
    "# Normalizing the input image\n",
    "x_train /= 255\n",
    "x_test /= 255\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 50000 samples, validate on 10000 samples\n",
      "Epoch 1/50\n",
      "50000/50000 [==============================] - 23s 466us/step - loss: 7.7600 - acc: 0.4627 - val_loss: 5.6869 - val_acc: 0.5794\n",
      "Epoch 2/50\n",
      "50000/50000 [==============================] - 21s 424us/step - loss: 4.4442 - acc: 0.6208 - val_loss: 3.4234 - val_acc: 0.6570\n",
      "Epoch 3/50\n",
      "50000/50000 [==============================] - 21s 424us/step - loss: 2.7485 - acc: 0.6848 - val_loss: 2.2938 - val_acc: 0.6800\n",
      "Epoch 4/50\n",
      "50000/50000 [==============================] - 21s 429us/step - loss: 1.8381 - acc: 0.7260 - val_loss: 1.5850 - val_acc: 0.7266\n",
      "Epoch 5/50\n",
      "50000/50000 [==============================] - 22s 430us/step - loss: 1.3440 - acc: 0.7552 - val_loss: 1.2163 - val_acc: 0.7493\n",
      "Epoch 6/50\n",
      "50000/50000 [==============================] - 21s 424us/step - loss: 1.0646 - acc: 0.7745 - val_loss: 1.1516 - val_acc: 0.7192\n",
      "Epoch 7/50\n",
      "50000/50000 [==============================] - 21s 424us/step - loss: 0.9034 - acc: 0.7933 - val_loss: 1.0193 - val_acc: 0.7460\n",
      "Epoch 8/50\n",
      "50000/50000 [==============================] - 21s 424us/step - loss: 0.8083 - acc: 0.8107 - val_loss: 0.9186 - val_acc: 0.7662\n",
      "Epoch 9/50\n",
      "50000/50000 [==============================] - 21s 424us/step - loss: 0.7471 - acc: 0.8230 - val_loss: 0.8786 - val_acc: 0.7778\n",
      "Epoch 10/50\n",
      "50000/50000 [==============================] - 21s 424us/step - loss: 0.7010 - acc: 0.8390 - val_loss: 0.9381 - val_acc: 0.7551\n",
      "Epoch 11/50\n",
      "50000/50000 [==============================] - 21s 424us/step - loss: 0.6732 - acc: 0.8489 - val_loss: 0.9158 - val_acc: 0.7692\n",
      "Epoch 12/50\n",
      "50000/50000 [==============================] - 21s 423us/step - loss: 0.6520 - acc: 0.8607 - val_loss: 0.9022 - val_acc: 0.7790\n",
      "Epoch 13/50\n",
      "50000/50000 [==============================] - 21s 424us/step - loss: 0.6356 - acc: 0.8705 - val_loss: 0.9355 - val_acc: 0.7746\n",
      "Epoch 14/50\n",
      "50000/50000 [==============================] - 21s 424us/step - loss: 0.6199 - acc: 0.8814 - val_loss: 0.9078 - val_acc: 0.7893\n",
      "Epoch 15/50\n",
      "50000/50000 [==============================] - 21s 423us/step - loss: 0.6065 - acc: 0.8870 - val_loss: 0.9459 - val_acc: 0.7840\n",
      "Epoch 16/50\n",
      "50000/50000 [==============================] - 21s 423us/step - loss: 0.6044 - acc: 0.8931 - val_loss: 0.9563 - val_acc: 0.7812\n",
      "Epoch 17/50\n",
      "50000/50000 [==============================] - 21s 423us/step - loss: 0.5898 - acc: 0.9001 - val_loss: 0.9900 - val_acc: 0.7763\n",
      "Epoch 18/50\n",
      "50000/50000 [==============================] - 21s 423us/step - loss: 0.5878 - acc: 0.9041 - val_loss: 1.0765 - val_acc: 0.7570\n",
      "Epoch 19/50\n",
      "50000/50000 [==============================] - 22s 434us/step - loss: 0.5870 - acc: 0.9050 - val_loss: 1.0502 - val_acc: 0.7738\n",
      "Epoch 20/50\n",
      "50000/50000 [==============================] - 21s 423us/step - loss: 0.5797 - acc: 0.9112 - val_loss: 1.0858 - val_acc: 0.7630\n",
      "Epoch 21/50\n",
      "50000/50000 [==============================] - 21s 423us/step - loss: 0.5738 - acc: 0.9148 - val_loss: 1.0647 - val_acc: 0.7783\n",
      "Epoch 22/50\n",
      "50000/50000 [==============================] - 21s 423us/step - loss: 0.5727 - acc: 0.9185 - val_loss: 1.0699 - val_acc: 0.7796\n",
      "Epoch 23/50\n",
      "50000/50000 [==============================] - 21s 423us/step - loss: 0.5743 - acc: 0.9193 - val_loss: 1.0504 - val_acc: 0.7801\n",
      "Epoch 24/50\n",
      "50000/50000 [==============================] - 21s 423us/step - loss: 0.5714 - acc: 0.9210 - val_loss: 1.1176 - val_acc: 0.7750\n",
      "Epoch 25/50\n",
      "50000/50000 [==============================] - 21s 423us/step - loss: 0.5776 - acc: 0.9221 - val_loss: 1.1691 - val_acc: 0.7661\n",
      "Epoch 26/50\n",
      "50000/50000 [==============================] - 21s 423us/step - loss: 0.5720 - acc: 0.9241 - val_loss: 1.1083 - val_acc: 0.7768\n",
      "Epoch 27/50\n",
      "50000/50000 [==============================] - 21s 423us/step - loss: 0.5690 - acc: 0.9270 - val_loss: 1.1939 - val_acc: 0.7672\n",
      "Epoch 28/50\n",
      "50000/50000 [==============================] - 21s 423us/step - loss: 0.5698 - acc: 0.9280 - val_loss: 1.2362 - val_acc: 0.7547\n",
      "Epoch 29/50\n",
      "50000/50000 [==============================] - 21s 423us/step - loss: 0.5782 - acc: 0.9268 - val_loss: 1.1818 - val_acc: 0.7715\n",
      "Epoch 30/50\n",
      "50000/50000 [==============================] - 21s 423us/step - loss: 0.5737 - acc: 0.9299 - val_loss: 1.1457 - val_acc: 0.7812\n",
      "Epoch 31/50\n",
      "50000/50000 [==============================] - 21s 423us/step - loss: 0.5683 - acc: 0.9311 - val_loss: 1.1112 - val_acc: 0.7876\n",
      "Epoch 32/50\n",
      "50000/50000 [==============================] - 21s 423us/step - loss: 0.5653 - acc: 0.9339 - val_loss: 1.2151 - val_acc: 0.7661\n",
      "Epoch 33/50\n",
      "50000/50000 [==============================] - 21s 424us/step - loss: 0.5613 - acc: 0.9337 - val_loss: 1.1499 - val_acc: 0.7893\n",
      "Epoch 34/50\n",
      "50000/50000 [==============================] - 22s 433us/step - loss: 0.5692 - acc: 0.9331 - val_loss: 1.2178 - val_acc: 0.7733\n",
      "Epoch 35/50\n",
      "50000/50000 [==============================] - 21s 424us/step - loss: 0.5650 - acc: 0.9350 - val_loss: 1.2198 - val_acc: 0.7720\n",
      "Epoch 36/50\n",
      "50000/50000 [==============================] - 21s 423us/step - loss: 0.5548 - acc: 0.9387 - val_loss: 1.2664 - val_acc: 0.7671\n",
      "Epoch 37/50\n",
      "50000/50000 [==============================] - 21s 423us/step - loss: 0.5500 - acc: 0.9408 - val_loss: 1.2149 - val_acc: 0.7777\n",
      "Epoch 38/50\n",
      "50000/50000 [==============================] - 21s 423us/step - loss: 0.5553 - acc: 0.9374 - val_loss: 1.2757 - val_acc: 0.7742\n",
      "Epoch 39/50\n",
      "50000/50000 [==============================] - 21s 423us/step - loss: 0.5517 - acc: 0.9408 - val_loss: 1.2132 - val_acc: 0.7757\n",
      "Epoch 40/50\n",
      "50000/50000 [==============================] - 21s 423us/step - loss: 0.5524 - acc: 0.9396 - val_loss: 1.2232 - val_acc: 0.7748\n",
      "Epoch 41/50\n",
      "50000/50000 [==============================] - 21s 423us/step - loss: 0.5692 - acc: 0.9365 - val_loss: 1.2630 - val_acc: 0.7789\n",
      "Epoch 42/50\n",
      "50000/50000 [==============================] - 21s 423us/step - loss: 0.5710 - acc: 0.9382 - val_loss: 1.3511 - val_acc: 0.7668\n",
      "Epoch 43/50\n",
      "50000/50000 [==============================] - 21s 423us/step - loss: 0.5642 - acc: 0.9408 - val_loss: 1.2863 - val_acc: 0.7722\n",
      "Epoch 44/50\n",
      "50000/50000 [==============================] - 21s 423us/step - loss: 0.5556 - acc: 0.9431 - val_loss: 1.2528 - val_acc: 0.7752\n",
      "Epoch 45/50\n",
      "50000/50000 [==============================] - 21s 423us/step - loss: 0.5387 - acc: 0.9451 - val_loss: 1.3636 - val_acc: 0.7569\n",
      "Epoch 46/50\n",
      "50000/50000 [==============================] - 21s 423us/step - loss: 0.5372 - acc: 0.9460 - val_loss: 1.2695 - val_acc: 0.7792\n",
      "Epoch 47/50\n",
      "50000/50000 [==============================] - 21s 423us/step - loss: 0.5387 - acc: 0.9446 - val_loss: 1.2671 - val_acc: 0.7794\n",
      "Epoch 48/50\n",
      "50000/50000 [==============================] - 22s 432us/step - loss: 0.5372 - acc: 0.9457 - val_loss: 1.3002 - val_acc: 0.7692\n",
      "Epoch 49/50\n",
      "50000/50000 [==============================] - 21s 425us/step - loss: 0.5419 - acc: 0.9465 - val_loss: 1.2621 - val_acc: 0.7788\n",
      "Epoch 50/50\n",
      "50000/50000 [==============================] - 21s 423us/step - loss: 0.5338 - acc: 0.9478 - val_loss: 1.2990 - val_acc: 0.7775\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x7f81bc367e48>"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Training the model\n",
    "model.fit(x_train, y_train,\n",
    "              batch_size=batch_size,\n",
    "              epochs=epochs,\n",
    "              validation_data=(x_test, y_test),\n",
    "              shuffle=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}