cnn.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# TensorFlow Convolutional Neural Network for Image Classification"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Tulika\\Anaconda3\\lib\\site-packages\\h5py\\__init__.py:36: FutureWarning: Conversion of the second argument of issubdtype from `float` to `np.floating` is deprecated. In future, it will be treated as `np.float64 == np.dtype(float).type`.\n",
      "  from ._conv import register_converters as _register_converters\n"
     ]
    }
   ],
   "source": [
    "import time\n",
    "import math\n",
    "import random\n",
    "\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import tensorflow as tf\n",
    "import dataset\n",
    "import cv2\n",
    "\n",
    "from sklearn.metrics import confusion_matrix\n",
    "from datetime import timedelta\n",
    "\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Configuration and Hyperparameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Convolutional Layer 1.\n",
    "filter_size1 = 3 \n",
    "num_filters1 = 32\n",
    "\n",
    "# Convolutional Layer 2.\n",
    "filter_size2 = 3\n",
    "num_filters2 = 32\n",
    "\n",
    "# Convolutional Layer 3.\n",
    "filter_size3 = 3\n",
    "num_filters3 = 64\n",
    "\n",
    "# Fully-connected layer.\n",
    "fc_size = 128             # Number of neurons in fully-connected layer.\n",
    "\n",
    "# Number of color channels for the images: 1 channel for gray-scale.\n",
    "num_channels = 3\n",
    "\n",
    "# image dimensions (only squares for now)\n",
    "img_size = 128\n",
    "\n",
    "# Size of image when flattened to a single dimension\n",
    "img_size_flat = img_size * img_size * num_channels\n",
    "\n",
    "# Tuple with height and width of images used to reshape arrays.\n",
    "img_shape = (img_size, img_size)\n",
    "\n",
    "# class info\n",
    "#classes=[\"dogs\",\"cats\"]\n",
    "#classes=[\"s1\",\"s2\",\"s3\",\"s4\",\"s6\",\"s7\",\"s8\",\"s10\",\"s11\",\"s12\",\"s14\",\"s15\",\"s17\",\"s18\",\"s19\",\"s20\",\"s21\"]\n",
    "classes = [\"001.Black_footed_Albatross\",\"002.Laysan_Albatross\"]\n",
    "#classes = [\"001.Black_footed_Albatross\",\"002.Laysan_Albatross\",\"003.Sooty_Albatross\",\n",
    "      #     \"004.Groove_billed_Ani\",\"005.Crested_Auklet\",\"006.Least_Auklet\",\n",
    "       #    \"007.Parakeet_Auklet\",\"008.Rhinoceros_Auklet\",\"009.Brewer_Blackbird\",\"010.Red_winged_Blackbird\"]\n",
    "#classes=[\"001.Black_footed_Albatross\",\"002.Laysan_Albatross\",\"003.Sooty_Albatross\",\"004.Groove_billed_Ani\",\"005.Crested_Auklet\",\"006.Least_Auklet\",\"007.Parakeet_Auklet\",\"008.Rhinoceros_Auklet\",\"009.Brewer_Blackbird\",\"010.Red_winged_Blackbird\"]\n",
    "num_classes = len(classes)\n",
    "\n",
    "# batch size\n",
    "batch_size = 32\n",
    "\n",
    "# validation split\n",
    "validation_size = .54\n",
    "\n",
    "# how long to wait after validation loss stops improving before terminating training\n",
    "early_stopping = None  # use None if you don't want to implement early stoping\n",
    "\n",
    "train_path = 'B:/New folder/tensorflow-image-classification-master/train3'\n",
    "test_path = 'B:/New folder/tensorflow-image-classification-master/test'\n",
    "checkpoint_dir = \"models/\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Reading training images\n",
      "Loading 001.Black_footed_Albatross files (Index: 0)\n",
      "Loading 002.Laysan_Albatross files (Index: 1)\n",
      "Reading test images\n"
     ]
    }
   ],
   "source": [
    "data = dataset.read_train_sets(train_path, img_size, classes, validation_size=validation_size)\n",
    "test_images, test_ids = dataset.read_test_set(test_path, img_size)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Size of:\n",
      "- Training-set:\t\t56\n",
      "- Test-set:\t\t84\n",
      "- Validation-set:\t64\n"
     ]
    }
   ],
   "source": [
    "print(\"Size of:\")\n",
    "print(\"- Training-set:\\t\\t{}\".format(len(data.train.labels)))\n",
    "print(\"- Test-set:\\t\\t{}\".format(len(test_images)))\n",
    "print(\"- Validation-set:\\t{}\".format(len(data.valid.labels)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Helper-function for plotting images"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Function used to plot 9 images in a 3x3 grid (or fewer, depending on how many images are passed), and writing the true and predicted classes below each image."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_images(images, cls_true, cls_pred=None):\n",
    "    \n",
    "    if len(images) == 0:\n",
    "        print(\"no images to show\")\n",
    "        return \n",
    "    else:\n",
    "        random_indices = random.sample(range(len(images)), min(len(images), 9))\n",
    "        \n",
    "        \n",
    "    images, cls_true  = zip(*[(images[i], cls_true[i]) for i in random_indices])\n",
    "    \n",
    "    # Create figure with 3x3 sub-plots.\n",
    "    fig, axes = plt.subplots(3, 3)\n",
    "    fig.subplots_adjust(hspace=0.3, wspace=0.3)\n",
    "\n",
    "    for i, ax in enumerate(axes.flat):\n",
    "        # Plot image.\n",
    "        ax.imshow(images[i].reshape(img_size, img_size, num_channels))\n",
    "\n",
    "        # Show true and predicted classes.\n",
    "        if cls_pred is None:\n",
    "            xlabel = \"True: {0}\".format(cls_true[i])\n",
    "        else:\n",
    "            xlabel = \"True: {0}, Pred: {1}\".format(cls_true[i], cls_pred[i])\n",
    "\n",
    "        # Show the classes as the label on the x-axis.\n",
    "        ax.set_xlabel(xlabel)\n",
    "        \n",
    "        # Remove ticks from the plot.\n",
    "        ax.set_xticks([])\n",
    "        ax.set_yticks([])\n",
    "    \n",
    "    # Ensure the plot is shown correctly with multiple plots\n",
    "    # in a single Notebook cell.\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Plot a few images to see if data is correct"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZgAAAD6CAYAAACccFGwAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvNQv5yAAAIABJREFUeJzsvXm4ZedV3vn7hr33Ge88T3VrnjWU5NKAJlu2LGyMsTFgTIIJUyAdGswD6Tx0JyEkPEkMCTQECDFJ3GZqsGxjYwsPsoStkixZUkmqUqnm+c7DueeeeQ/f9/Uf+9xbMs2T7rbbuhH3vv/cOtPep9beZ797rfWu9xPOObawhS1sYQtb+P8bcqO/wBa2sIUtbOHvJrYIZgtb2MIWtvBtwRbBbGELW9jCFr4t2CKYLWxhC1vYwrcFWwSzhS1sYQtb+LZgi2C2sIUtbGEL3xZsEcwWtrCFLWzh24ItgtnCFrawhS18W7BFMFvYwha2sIVvC/RGf4HXG53dPc7PFSkU8lTjBj4eNgyRniZxFo3AOkEcNnGAHwQknsNJDy9yKE9ho5DEOqSUkNXEYQyewIYW1YrJ93TRSFpgHHEzxPggmoaWaiESQUc+wDqHcOC0wrOOUnUVhCRQmkKhwGpllSRJUIUseS9AODAOKo06WU8jjCPQHiGG+moV6SnymRyNsIWNDN09nTTjiFatjq2ZJedc/0bH/vVCoVBwfb29AAghAPjbHSvE3/Kcwzm3/n7nIN2EaG/rxvNr773x+DVbXXtrexd/c/fib9n1N75nbbvpFxB/8xX3mteB2bm5TXWMfa1c1l+7fL0m0Ov/fm0wXxO9GweU1x7LG8f2/35gnHPfcB699r3p49fuw/2N7d3Y7d/8Wt/4uf8eHA6oNuM33DHedATj9Xby9n/9y/RcsUQTGTgzxY69+3ju2JPc+eBDhNUQkbNcvzZFn87wnFliqCpYnOhh+PwcQwcOUu4KOfOV42QOF0BrOmYjmhND7FlNyOJz4cpV5s0i4dgA+yZ7efKZl7nrrW/ikac+QXj8Gm+56y6uXqpwbvpF7r//LQwPdLN4+jxfuHSOI7tv4sE77uD3fv/3ufXu28BpHtx1mD879xIT24YpL5fxp2e4WltlNpfl/Ufv4InHv8reiUn6OorUMz4nn3+ZegFEoZP4C09hn12+utFxfz3R093NL/zch/D9AOfs+o9daz+9KcBhrcVZh1QS50BKQZIkCCGRUuKcwzgHzmKNwVqHMQYArTXK8xDWYp3DGkcUtzDGkjgDJr0IWWva+xPpxcS59vVFYJ1FCokQIIRsP5vuU7yGzIyzSCTWWqw1gMCYJN2GNel+gH/+L395Ux3j3UOd/NefvA8h0uOmlcbgEM6hlUeSJBhrEdJC7JBa44TDCYlSEuEciXPgHFJIrDNo7afnCwKLQUkPaw1JkiBlegyiKAIEWkicszhAaQ/tKdLNWYyxSCnax8wipUQqSRxGJCbB8wIkjsRaTGKRDqxMuccYg0LijEEqSTMKEUZghePdv3PsDXeMNx3BhJFhj9fPdW+KUyLCXZtiaGiEo3e9metXr6PyPtXpy1QHCoyYAjumDVYrRkpVMn1dlK9O87aBnTx/fZHagW0MFxXxIY+eL18nefc9vPK5Y9yy6wjyyjl278nQTBrccedNfPoLn6IQd/CT73wXjxdyjAWL3H/0++ka7GP6zEVO11bZs3M3zZUFTl48TeW+u3iscoEfHz7KH3z8j3jfP/gRPnXyJd4zNMFv/8kn6d87TvLsGf6i2sDbM0Lnjt1cf/pZDh88wqfDGbpVjpFai54jBzj37JMbHfbXFVJKgiCz/mNfu2FVyuEkCNu+qKsb2YFzIIVEKgU4pFQowFpHgkAIi9QKTylSCnCgFDaxCGnw/WD9LtfadL/O2fXvtPaaEAIpJUqp9QvQGmE40gy6/Qmcs0h7I6NSSrX/+gA46aP+X90B/92Ds5YwTBDWYJzDSoNtky6mBUoihCNJ0vhKKVFSoX2PJHTEcUwSRgTFLBKBFJJ6WEcKhZAOi0ALSEyEEOD7fjszkWAsEUmbPAQmtsSJQAJRYnDO4en0+JokQWtNM4zWz8NWKwYc0kGUxHhaoz2fKIpRUmCcQyqNlpJ8Jkez2QTsfyca/+Ni0xFMgMTVIlYOjrHj5BUm77+X3kyBT9ZOkalW2HbHm6k0M2yr+5TOL1Dt76I0JOnpzrF6bJo3HzrEl55/hW2TkwSLJbLNIVb0Cov3DPDq5ZMkdp6FjOI7b1P89l8+wg9NPszl6mU+8NDDfPpTf0W1ssLOjOapiyepju7glae+gIkTHt53lKn6PF+/eAGRyzDQoVn+4gUuPdDLW9/2dv7wq1+kulDno+UFJt73ECoIGOjpYbbV4uq5S7Tyg7xQELzwsf+Cvljj7n/0TjKx5fMXL250yDcEWitA4fk+uPSC4pzDJRaDJfACrEiJwsQmvcgriTMWk8QopdGeRiLxlAR946eyVqJSSiK1BmtxQuCMwRiDlOkFS2nFWl3EOYdt3zGvcYKUgrQNqkgSk5ZN2+WQtaKYFIBQKKWQQqQZk00zqrWM6G8rt/1dh7WWKAxR0uGMIHQJzjnC2BCGIQaDRhNkAgCcaZHJ5alVa2itMDFIrQjrMUIYhPSwLkb7Pi5KkNrHqHa+KSRhGNJqhWQDnzCKUL6H5+n114QQCAtOuPQvYGKD9DROCKx1aK0IpIfVBqQC63BKkBGaKMhSy4/yshUUvQKRdshWhGeb9JgGOfXGPMibjmCMsjzbucRePcjJ/gajgcAM5LileyePV56mQIm9S3AsmWfHrSP4j51nuOUzOBcxtWeAr586iRUZBh64l1dOv4jf55O7HjO+Y4j8E18lijLkXlnl1Jtu5ubJ3djxLvquXOR3fus3ePjWI1yeK/HksafJbBtFJpbGUonm9QZ/9oU/4i0//jCqbok8n2Slwc7vOEj3rt1cKM/TeX6Fn/vZH+OJY1/loUNv4v/4+guMjW3DXL/MvsnDPDV3lYO9/Ty7b4Lv/aHv4NgXPs8Pv/07uS02PM7xjQ776440cxEoJbGSdkaTlsa0VCTWIpzDCocgzSycMQgpCTJpNiKEAClwToNLLxKOBCkkSImUCuEcBrBJjBAaKVMykcpijEFrjQO00u1MxKblN5OWV3zfXy/z3Hh9LQsyrGVTUkocIJ3DGJDCYZ1tZ0GbzxHdOsdqpU4+4+Pa194gk0HjcGiE9sl4AdYYtBfgdBaMw1c+QgrQjnwu274hUNQqVaTW+L5Pq2UwSUpUuXy2fR5p/IwljBL8jCaMTJtgBEIKAj+g2WriaS+98RBpf0YKgZCSQiGPMQlOgpYBxsZIL8DPd/Cs6eRkUGSmqbFC0EoE9QQyokCsHKOdmny7PPtGw6YjGO0cO+oZgplljhSGqGcds08+RqVUI7OrFxF7PBdfoyuSLF27xPItknilTM/QKPlCkctmhXctFzjhzeCZDCuVVSYTj0vxHNFIB2P9Q1zNNBgJ5rhlcg+ffuxPeceBWxkf28ZTl65z5223w/HnGSLGtFrYlQhfJKixLqYbVZKs4tWnnmPk3js4MTfDB/fupz6d58QtlupKmcPbRvnTz/8VHpJs7wRvGtvLs7VlhsNebs50c1bAsU9+giPv+0EiL+HY1JmNDvnrDmssq6urWJtmFK8tTynVzgakQnsevuchvbVMJ62bG2dQLr3or5WyhBCgJNIpkGkN3xiz/ppsZxhx3M5WzI1MRSkfJdI6umiXu7RO/0ZRTKvVAtKsS3s+WglAIwVUa3WkAmcFQUYTxhEKjdUOF7czLffGvLv9ViClpLe7G0fa5xJCYJzF9/12KTEtUUqVlj2NiRFC4vkBSqp1kUYSJ0jl8IL0c1EUoZWHEBLtawQSS5pnSqEx2QJSGrKeQNgIaxMyuSzWWjJBBmMTjLFYkZ5v1lkUEgQE2RxRHGGxKOUxm+3l87ab42ioxhRWm3i+IEYyXmuS+AEL3Tni0FG2b8xjvOkIph42oBCQyWe5ODXFeN825O7d5MuL7GyVyCZL7KrDgu/T2DGJ96UvoFwXuhVTyq3yZrudBVYJP32cvrccYOi5RaZvLdI/NM6Vc0tcXlxk4vA+5NwcX3zmCe7bcZAXaiXu3radmtA889wzhLFFVA1Hdk9wtjfPfC1icvc4GeNhE4uMImazCT//0Lv4lT/7z7xn7+3MxQ1OPfkVmhfP0LnrNv7Bw2/h9/7T71HIB/Tesofvfsc7+MUf/Bn+ya/+LB/+8Ed45rm/4qVSwnfdfiuf5NWNDvvrCvcadZfneevPCylRUuJ7AUqr9kXJ4aKEuBWtE5GSksjFJLEjjCrEcZwSFQohDUp5CCkQiPXmrxCCbC633rRXnkJIhWpvM45jlJZYlzZ3vTbRZTIKz9MYY4njhCgKCfwArSVxnJDL5khMghWWJDH42iOxDuUcQqVKRmPemPX5bwXGQhNNMZfFxSFCa7RIm+xKahKToJQkSQyxMShPEzZb1BsW3097YOmGHEIKwjDG8yye56U3Dlis87gYFcj4krBW4mrnTi7ZDEPO0uNroqRJIW6xX9bokC1aUQNfewglkEIQxXF6o+FASIjDFg3nUZIF5mWGR8JhdGkVTzVx1RZeYvBdliELtUIOJzU9LYfwffbtuYnHNzbk3xQ2HcF42RyjFs42rsEQCBuy3CwhdvQywggvzl2hW/cy2NFLYXGWZ+7cSzVsIOIVJnPdvNSc4cDhoxwpam4ev5t5eY5Xytfoe6nErtGdXL9cRufzjOVXsDdP8tmLLxLNVqlLj/l6g6WVZcbGtnP1xKt8qr7Mnh07SKZeJsDjykwZWYtw4330ninzsfg0b9m9nytNx+yp66zs7iSfH0fS4tG/fJQDhw5xXcV06IAPve1HueeH3wE9XRTun+CDd97Hf/z4p3n03MmNDvnrDiklhUIBRHqHuna3Cil5SCWRsl0CsxbafQ1jDEliSKxBa4/A98jm8+gowhlwLiExBiUlQsr2nTNAWsKyJm3wJkmCafdjlFIEfgalJa1WQpLYtMwmFb4X4AWpUikKI8IoSj9vDZ7vo6UmikOcA98LsNYRC0OgfSKboJxO+z1SbWC0NwaLTcsfnqjy8G6PoWKenPOYqsUMFxQ2CdsKQEVoDGXVS0ZZyhkoeg20SFDSIZzFeQrP83BIlNZopcn6mrDRYtkV+ZwbwJYamKCT2Zol21plyctRz/v4ZUel2sQfG+RO0WQnNaTnsS0ukxUJGtA6wAlHEhsqKsNL/gTPqx4qYQvVDEkwqNlVko4Osp0dJC7Ca0T0RAG5ri5yfpZC9yB37jzI72x00L8JbDqCaUVNFq8t04vDz/cxMNZFo1gluVblIyt/zUPbDvGqWSGf1Tycm+Dm+TK5ge1EzYgeP0vnzFWWzn6JZGyEv5p+mf0u4OjkER5/5in2jI0zsK2f0vl56jvuoVNcYF8Ssv2e7TzxiT9nNayTDy3LL19h7JadaN9w/Cun0lp7Z0z11CKq2MH77r+Xjx8/RcfUOa6uFunuH2Pvjp089uqzVJSHt6KZW15hxBvG7dxGVFpCD2XYfWSSpJawODXFRx75JJ3FLo4M9PM5Tmx02F93GGPWVVxKyfUyCqSzDFGcgAC11mtBkwk0LkjVW2uy5DhqIaXEDzRaZ9ry43YfxVqcYU2WhtIKicAPgpTUnME5QRxHtFoh1lo8HeD5ubYIQSAFCHwyeY9cvoAxSVsuDZ7WhCYVIMRJjNYKry3B1VJihF0nsc0GEWRY7ernLyuGQsmyI5nlRKOTooh4c+8KiRDMFrbzfP8YxuToiUIWlULZkPG4zl7l2J6LyTXnUim6s0SNOol1BNks1+ngYpTHWk3ey3GlWsY4gezqYjGJsbUmSIE/0EOjZfiyynJS5yB09HrdTLo6D8hFYunjNUvMqW6mvBEuNzxacYMgmyEzM8+8MIzkB8ga6BmeZHn2GuO9QwxNTnJt8RLNuXM0VIuvnKhvdMi/KWw6goladf549THet+PNTPSMsOi38Dr7KbqYf9T9Nk6VZzk8tpuw2aTSmaNmppCVHiwBV8Mmg5PD7N29k8Vmkw4vg5UJFxenOHBgN7IFJhNTDLOMI/nYH/06FaE5WX+C7Yd2Uezt4vLVOTqHHeXLM/TuGsD1Z7jpvndw7vHHyGYlciWBwgDv3G+ZXW4iI8NAZ5Fa6Ro600XUWuGtRw/wZx//DCdOLHK0t4dqFHPop7+TqGH4b49/lA4vj7pumL1Pk5Wbz6zBWkuz2Ww3WdP6t9YKrTXrF3UhscYSW4uQAufAJLatykqzAutSibNsl8PWiCWVEBtSpbIAIbHWUa/W2nMStBv9Ct/zyWZyaK3bmY0lDFs0m+1ZHGdxFhwWECgl0Non8IN2/ygVLERhhLExQinyuRwChxcESOfW97mZEGtFbXSISrlM0l3kalTAq6+wVEvoyQ4xOuDxgu3ktCkwVl6lW2XY3tVNs9jBofFJhIl55MLLhJHmfrXCsK1S0AElDE/4B2G1yrWaxQtXqPZ0IapNtJY0KjVaElASJTVxaJGBT9iosdhZxPqKpXrMmaZkJTeCH+TZUZjkgt+JjhzdExMMSEVrYRF96zYO1moM9A2zbed2YiQr23bSme+g1axT61S81CoxKBKeufDCRof8m8KmIxhf56gQ8pXpywyqUUqrdXbleyjnHH7TMWK7KdQtk307uXjxPJ2ZbnoCxXKzQcYDEUrONGYY8rpYWl1gRje5c/QAZ+aukcvm6JRdDBYSVD7LB95yP7/5B59gz4MH8YIMN+1/gPLKx+F6maoA3Yroz+bRixVWZhsM7xsm1jFf+I0/Yd87buHUF5/hre9/J2dmr2GX6wwHRfSOYf7kLz5Hx/ZB5CvXWXr1LOe/fJKbP/ggY3e+idXcIGZ1jtU3D3Bz0snAzr3Apzc67K8rrHU0mw0cAiVluxYv1gca1wYWlRRo30ej8bTC2FRJliqA11RdkCQxITeEAjiHEwKhUmmqlAKtPIrFDoRgvUTmnGsP2aVyZM/zyGRUe/gSbGKITUwURalM2SQIoREqJTdn1uZfJEqDSwTWxKyWy7h2ec85Ry6f36BIbxy0VHTEivHRfZysLxP1FFjN9TGwsMyryyWenl1AxfM8ePcEwvmYZpndY+O8dPI49XyOyW07MZUq1YrPs/kJTrWqjPd3cYmY6swS9zYtrquTVRGSTwydgwMsFvJoNEVhyFebFHIFGtIjKwXe0B4aSZM51yDfUSQqz3EldmAsiyLLO2+9l4GBYfLZHMYaBLBUWaUxf5nxPbemsncHi8Usr167zKde/iodlTqJg7OlGbat1ji10UH/JrDpCKYr28lP7/4As60SPcOD1C5c5Lyb5dC2A7jYkHgLLMctaotXGBoaolxeZbpSobMbyrWEBT9meHCM0vwihY5ubu4f5/rSMn7Go7FaZXL3MKVSjfLVea7P17j3PQ9Qu3CJkV0HOPb4x0ku1zj63d/PmdPPUa4ucP3pi8ytzjKxbz/NczMM7+ml8/btaCfp2z7O185dolcaJvbu4bFHv8bDtx1i9ouv4nblqVccV65VIQ7o9jW//uHfQK4aRr7/Lq4S8s6j9zN37fRGh3wD4NrT+LRLTm59wHGtVCZE2iiOG41vnLZfn5CX7SxEoOSNn0m6nZRkhJXt9ztaSYtmq9lWLmkC38fzvPVJe2scWEeShOuGJGtSZGtNOr2/Ni3eahELcLY9O4NDaw/fD0iSGGsjYhOjSBvJ1bnqBsR4Y+Gs48D4XnaNbWe4scKXp84w6RcYEgXOL5UYHBjDJQm33HkPrzz3ONI2+dzjj9OR87l24SSz01Pcc+R+6pUSly6dY8jPc251BWsV44UhMlno7uzAVFfIB530D/fymZ/4AP/Tb/8u5WqDrp4iOT/H4PgOpDOcrS+w2HBElTorKwvk8gWWXcxe18M9N99BaalErdlieGAALRWZTJ7Pf+rjTHQJgo5Bent7iJ3j+ZPP8djZlxmVPksKgljT1dHLYN+2jQ75N4VNRzCGmJmMY/baFH7eJ1coUlte4HjjJAP5IqVmzN2HDvLy5QuUmiHZzgLl5WVyopeh4UFMq4zGY6Sri2q5gWv2oBNJf8cwSdZSWl6kFlZ54s8+wpPxq/zCT/46Xz73T3niwisgPUYPjvLZRz6C0ZqdQxPM9Xoc3H8b164tsZJUCXI7uLJ6iY9+3y/z44//c4biHLon4fiXTlAsNTn5Xz9PXK1TPz+Ld9tu/BenKVWqnGu2EBcqtN62jcv/8XMc/tD7mV2p8NW5pY0O+esO15YQO1y7P3HDy0utzZO0swgpPaQQGJuWqJyz63Mn1hqscZgkvJExAFp6baGAXScJh8O1bUGUspg4xrbnWlJSS5v6QqZkl9KGw/P8NMOREkXar4miOLUs0ZLA81HaQziBEwm+88kmWWx7eDRrEkRn54bEeSMhPU080Mnx6jQEPmHUoJHJYCZH6cw4RnQHg10DHNy7h4HuLmq1MreLDJ60vPTMF5laWOS9t9wGDm46cifTSwsUih0o5VHIZnEmppUkSKUQOuLtu/qJo5Bf/9EfxTbK3P+//wYHukY5duYpGs6Sz+bxWhG+5xNFVRKdILD0jA9RyOUYHRtDIDCJQTiLiVt897u+h4yvsUikdSyWSuzuGaHztgHmV5d47659jPX2YTF4SvGRf/NLGx32/8/YdASjC/30hx7zA8Ms12JkV41Dtx4kWogZGOqhfPwi12cXyWVyNMs1wsRw2227uXx1mun6BV44f4bhbJ46ionBmwnmr5IN4OKZ69x1cBcvfuYvkEcf4MgDu/nqU1corywRBl1UXrxIz/Y+Xq2s8KbbbmJxocxyskLfod0cP3YKL2zSe3Qv8xcvono9PvTL/5y7e8ep9Wl6tr+JD/zwEf7FT/8MzVpCZ6GLpdNl7tIjPDl9mdt/6B288l/+GkeGzlcbGN0PX7+KnBhle7GHzTbLv9ZDAUccxa/pm9j1LEZrvZ7NOCHXVWTWOmJj0j6NVEgpUEJikjgdtHSWmBgrHIp0AC9xDi0FOFJJcRytT3PrthxZCENskvYcjL0xPyNTmbLW3jdYwmQyOZyztJoRzUY97cc4gRNpNhMEAUEug7OOsD1Hs5nQikM+98KX8aQE7aOsRZZXuVSrUmo1uSwW+eED+3nlxItM7tpDd28/UkqarTp7Dt/NXcOjzKwuU2816csVGOjrpTNfJEkipBQ4HZD1PCSCN43kwICvBL/6a7/G+3/ge9kVC6KpKQY9xdWOLC0Ts6uji0rGY2/TYvNdWByXl+bZV+hjdaVEf28vfYViepyxZDKpaEQ6h5KCwd4uZnxHtDDHA7ffRhLF1FsVrizMMdTTs9Eh/6aw6QjG1JbJdGa4yRsjEjVWk5j56xXOzp/HvwCHdxwgaVbpHByiRoVD4/s59sxLHN13Oyu1EvdMdJDtzGGjVFpKRvPqi8+y5/DNPD99mtVD2+nxalT7drE9OMbXTnyJW+44ynypRG1mjnu+80EuXr5GKypRvWbZ97ZBamcvI3VA46szFAcCMnKA4oFB/CP7eKgKT3gR/+6nfoVCd558PsPK0hL3vPtdvPSJL9M7MI4/2E3v2F7mL12icb2K0BGH3vsgxb4cevrKRof8dYcQkiBIp/Hzudx6435tkt/aVF0mZeowaKwhEikZaJUKAgwO6dLmvlASKRTWpYquNSIwxqQzMg6slhhrUy8zJ9BSpuok2/bHkopUDyAQSISQxHFa7qo22iaayiPwfQRQq9egPTOTypQ1SqdluyhJ52VAILUm62c2LNYbBSEFndkMVkiSxCJyOZo5QUtrKiYi30z43aceZU8F3uO9l207dqY3FkrTymf4+JnnmV26zlSlRFdHN5P4/Mhb3s25l19g7/6byORyIKFUKsPQAHc+/N0st5r8wg/9PZZLZf6Xhx/kJ/7i0/iZAi6vkVZxsTTNYkazs7eX5eUp/v5d76Yjm2HnwAiNZoO55QUWa3VGu7sBUAKcFLgk4sz0eT576STnl6+jZZbp5gpTV86yEDfxRUDvGzRL3XwEYwyN0irWwNO1izx86+2sRDVu8XewbXIX9blVdFeWelJnMDtAqVxmf/84F69dxMsEFLJduCjm7NVXyQedDA0N8crFs5ROn2Pvux9kvGuMelilWWtQCHqYK1f51KMvIFcjbrnpHi6/ehVdKXHk9iM8fuqLrLx8gR0Hd7Ew08LaMgcm9lHvC6guTjPXGuEvjh1nqH+C7/qxf8xXvvAZmtNXSBLD0597FNc09Hb188xvfYpDt91JubOIiSw7j+5htbrK5z7/xzxw070bHfINwpr/l2xbryTrmYy1a/MqN6TMUghoS4etEKg1ky+Rlr7WhAGQ9lSUkDiVZiCB0lgMIjHg0l6OTQzGWayN1sltrdcDaz5kaT/Hk6mJJdYQR+nci/Y8tJJo0l5NEkXETuCkwFNtmxqZWmQ2W29MCeu3AmEhERls2MIIh79cJpAevbLAfQPjDO6b5Ey1xAM7j+JwnDh9lnKzynyzwmOzJykaj/6pWY4WBrjUbNAvasyee4XK8hy/+58+yzu+64OYjOaPXnmc2yZ+iYPbd/PmO24n3z+AiUOuLMwTduepWUO8ukpH4qhrD19rKkszqOUqT51/iT1DI+wfGwHrMdLfw4Xrlzm7eImjE3tAKwLt8ezFl/nYU4/SJzP0RBY/iPjaqa9TcwbpoJ6UuVSd3+iQf1PYdAQjkUyVprFVyT237qG1kKO+MIvf0c309UssLdfoKHRjhKFZbxFkJZnODIussKewl69deJkd2WGCTBcujDg9d5G3fufDdDZqnKhcpTNfg2ada5U682aBwSNvY/DwYV76zP9JRU+jMgLTtDz7ykmG3rSfpa9cZXGpTkeUIbEhp8NF6tMJd++8g+hKi8EHfpwP7vb4l3/wCCvnz5MLsnheH1lbp3PXJKWFJUZ3THL1+gy9E/uoTF1lZVrROTeD7/nk89mNDvnrDufSqXjteSitUVKlsyRxOmNirWs38gHW7GEcJkqt2p2U6xP4UglwAqRG4tLBxvU5GIfAYm2STvJn01hb6zBYPGNJjFm3cbdtIYGUIi3PDg0HAAAgAElEQVR3OUuUGCRgXUpOQHs6PyESbYFBan2GcKk5Z2IsGJOSIun/YbPBCYHG4QcaJxQ1pbCFHJlMF/NSMdhdwE8qPHr1Od657WaOv/QMjf5OFlt1Htp2K8OZDo581z5yxRzLS4s8/+SXWCxVsTqHGhjmN559hNAajvSM8tcvHmdp8A7QjvLqKh2dBTw/wEqBSCzKJpSNRcUNAhuzmoTIjGO2NMXZpfOcmj3D5YUZmmEzHaIF/vTrHtlMlrDeIDItpPRY0Q5PCjSGZismCRtYYXDW4effUMvArGPTEQzKUUlaXCmdRSwpxrXCdCdE1QZnr83Q0znAxOgoFxZPoDNFfBxJrMgRsFSeJpJL1Lw8Y/kO+oeHOHXlAvNz16BHMpHbznxrjrAeMZBkyA4donJtCeF7bJ/cTr4zT3lmmsj3WL2yyNt/4AE+88xVhvr7CefrvOntb+HM2XPY87M898rnmNi+h9sDx698+gV+8ed+np/6Z5dx2iEVmMoqtVoDFUfo7DBu/gLxfAXXarB4bhYZzxFrQ/aBwkZH/HVHKlNuYuu1dfUYIrXvSNd7EetZxZo/2Y3PWrAG0850pGlbymARNrVxl22JstIKJfT6+jFrn5dSgpUo3yMQct2U0hhLYmJwawtXSXydritiEoc1MVEUp2yCQ7cnywUQ2gRpHJGJkW1vS9G+A9Z6E/6MkwTfOKq1MrJYoBob/EqNl6s1mibmhYXLCOljwyauucxiIWFp4SLDXb183533UC9V6O3Kc+XSOS5ePMOdd9/LR//8D1mMKqx25qj7llLG5+e/4/0888op9nU0mJmvceTOo0xOTlAqFYmeexqEwxqDchAbQ1RZQSmJbyWzS1NID14uzeGEgmweaxIcAtts0YpignwOTwSEzRYoSwuBEgpV7CJo+rSSkHwrIqPfmMO0m+7MjOOI79h7GxMdE0yOTXD+2hX6OnuJalUmR3eR9wPOvHoJ+iynL16gONBBZyihpVBjvYy1JrGuTpAf4urla6zMXsU1WkxrgQ1nsIWE69Uqeb/OwcmjzMSzuPoC/Z07aczO0hF3Mde6jowtz588R9fEIFPPnuLen3kf8XKFnm07CDp9mlc0F8+dpVqaQWXy/Np/+Pc88pFf53t/8CdIait0j4yxdOkKiXIk85p8Vx/VuWm6xscYy+VpeVkqV6/z6X/95xsd8tcdzjnCMCSxFpvE7Sa7wPM8pKdRRqy7Ja8RjBBpz0T7wXqfZc2GRUuFE6nkeW3Y0iUWIyxOKhIbp+UspdcVaJCW4WIbo9rPSSnwPI01hjhZ21Zaosv4Pi7jk0lSsUEURVhjUt8xJfGQGGVRKDCWOI4RxoCxJPaN6bT7rcAKqIQVGmGVWFk8v0AtXMXTGYqFHJVKGWXq+H6Bk4tLeL6g5hIWowa/e+yz/P0j97NcqTC7MIWVgqhRoZ6FUj5DNVDUophH3/YP+ezjj1MTHlGyRNbT3HrLTVjjWJqeIpCKDiFpiIhG0kQHWTJ1h7MRJolJRAI5jVM5jARhDEmjhpYSJT1EkCOOEhKlcMIRxo6gkMdGjlajCtkC0gQ0vBZKvzEdszcdwfgq4KlTT9LfKrAsoLG0Qk7NMNx9C6vNKjPV6wwMjDFfirnv1tsp11eZqSxxx1uPcvapE8hmnbnWVVyuwmWxzEjvGHrbBOHsJfbu2MflxVcYHsoz7I+SCYr89aMf5b4de2lpTS5XpBxP4/f1UURx18SbmJ5c5NBDd4Mo8FL9EoWwwGD3XVwsPcH9dx3kxakZXLXG3HKF3/nd32JHbwdnqxWWZqbI9xdBd3LHzbv42tdexHmClcU5yiqHiBq4XA/Ge2OemN8KlJIUi4XUW8wkmPZFe839OIF1R2Up5fqKhTiH8CyeVIj2Y6DdaE8tWVLzTLHek4lMjHKpa64xCUrpNmms9XogNhHGpU1dISRIQRCk65SkooN0wh8n0rKe5xHkUnv5OIlx7UXHPAEio3Bt/7S1YU42oReZdY5WtYLUmkxQJEpadIh0UbB6rUZeaUzgg9KEcYyQml7fJ4wjLsxd5+c/+fsUC12EzVV8qdnVKjFr6kSADuE/HDjC//pr/xuHb7mTL9XO8veKHXzoF/8ZSkhaUYuV1WXCygoq04EJQ3xjyWhBGIY4CRERQXcniZII62ErK8iOIoHOYnwPKwSZxWVaUhAUu6g3Kni9A4S1GmE5xCt46cqYYYhp1Ulyb8xKxKYjGM/THLjjKNFUlRmWGZncxitXv06tf4WTX3yaod5h+nos/Z2DJPkQW7J06CzHHn8ckWkykh8iqBfIJnnu27+bv/r0F1iqzzMx1M/16wFjg1mOXz7HyHbL186/wL5bv4OZS9fYd/N+/vRH/hViraHrYqxTIAWf+Mxn+fCTj/DQvd/HyZPPsrvYQzxc4KkXL1IsZJDd/djSIseefZmBgQ4e/eQf8IGf/BAryyVEINi/b5wvfuFLaN8jblQRoowUmu6eHirzcxsd8tcdrm2fIoRASYWfDdaXsF0bqFxr+AscCL2+dgdC4JzAmrRXs9ancRJsHK8vLrXW29FCIX1wbs3DzJDE6eJhSqVmlK7dO7FtojLWErr2cshS4SmdzsNIibEGl6TLLjvADwIkac/BxEk6m4NIfdR8D8zaktCbDFLiOorYZo3yzDn6dECoNAQahSRJwMvmiXKCfKCwsWDVxUTNkB4rEFmfVtJAxIIWIS++coqkA4IgS9PCx/78EVbuPsKj1YtYWuyePNyeoXJUKhU8L4OOLPXyEhlkupplxhF2BOiuLgLniK1BmYTexFDtGSabCZh3VQoZH1upksvm8J1HXFshMzCA1DnyMsRt603P36RFHEZ0DI2QtJq8Edv8m86oyjqozzSIurP02x5yfQHb9+/g+LEn+Z4fewg1bNCjuwm7IvoOvpUF7xo24+gpdpBJslysXsD2OupM8Orzp3n4oTsY6xkkXvRYsPP84ZlnuHvndq60HPMty/6uHirlRc5dOod1dSLb4tLFF/mRH3w3h3aOc9/hAwzli+wf3cbM9HXGO7vR3jKdE3uYHB3Boenu6QfpMK5Fveb48K/+G8JylVy+gPIkn//MV5FSs//IUbpGR+kbP0zPYC/VlWX87OaTsAJo7SFlas+i1Y3ZF6VU+69uZzESz9M3Vh5sD2UqpdPnpcSKtCSWWNdWpDlik6TL7jpDGMfpY2PXFxJLCWjNIUDccGAWCiUVnuen1u5CECcJrVaTMAyRIl3wzA9yqRtAe50ZTymy2SzZXI58LksmmyHnZchksgT+JhRy4PCFwXmSTGcRL/EQmRwu24EvJFKDwWETw+LSHIlNiIwl0JIVkeCQ1CtVQs+nGtaxfR3EwsdahZpe4NRwL0vz8xgB0svwjve8D5skhGGI53mMD/YjvIBqvcLC9AVWWyFJEpPUK4S1VWy5QlclJE4SdL1GYbmEXVgi72v86Vm6mxHWWPIyoTOXoysxdCUtOowlKleIKnVcYtBCEDabmDfoksmbjmCMc0wvXULZBvldA2Q7Ezp0B29///dyaanEbGmey89/hWRuiTNP/AE3j+9lGZ+F0iJL5Smmzl9nZ18X5avPsLxQ5vEnjhOWGhzcO0HzWhnbSDh1/iorS9MYnXDs1a+zc/d+xrrHaDXrVKqLLC8ucPT2o3z/B95L58QQoRpD6kE6TR7iDk5MXSJC09QeHV15rk6d4abb7ieRGp0NePy507z7PQ8TFBSea3Fh6jKT+/ZQWl3FEBFWFklamt6efu788XdvdMhfd0gp8X0fT6XDi4kxJEmbEJKEKEwvBiZJ7fdbrRATJThjcCYhdjdmXaTU+MojkwnI+D6el5KTdKTN+9hgE0MUhSRRSBSFxHFMq9mk2WrSilrtBcY0vufj+d76ImZOSDylyWYyBEEGKdNJ/majSdiqE0VRWqVrL7mbJAkCgR9k8LSPzvoUCnmyuU14E2EM1fkSzalphHWUe7I47SHiJvUoZLXRII5C7MwMvpehWauTLC0TC4HVmuZqFXNxilatgkwMorZK0KqRxA26QkPf4hIdNkaELUQ+ACRhM70JwIEodEHTYEVCFMaUTIVkfh5aIflaiJxdZjVqolUBL1bYuTmKniLXiujq6qE+NUuzWWW5PMfiSoVKEhMrybKJyUUhOd9jIsjj5YtIEdF03v9jSP5HxKYrkRkbc7p0hpsOHuDytRmqjRJXL5/jwNxhFmt17rz9Now3TM5NU0Nz6uQUGVHHBYpizwSDO7dzfaZKq1mBmmNwoI+6EJwOr3D0wCj9y3n6B/o4OzVDnxOMjk6yWGoykBmnXFllpVnD9xyj46PkCzmG+0Y4P3WKn3/zu/i1T/0lQUeTUVFgfnGFex98OyePH8PL5hk9vIuXn/4yDUrsOXyYSzPXuefwbXRODHHfzTfxU//zP6VrpAfP95DGI1EeqtjDwuXZjQ756461jAHtkKQT/eAQQrXXuE+zCpOkpaw1a38p0mHJdH4m7bEIseZonLZkpFRIJRCodSPMxAmklevCANsuqwnS5+I4wpgE0Xbg9T2NCIK0HGZFaqxpHa69Zs3ajAxSkJjUGFGuOT47hw3DVNUWO2LCtVbR5oIQiCTCk4rKtWlksYdgdIjS/DxSagKroFoh29dHUXkUQ2hKj4XZBcLKCt1CY0dGKJcWsEojq3U6uwdZLlcICnmijMSPHTbfTVxPMLYtrJASqTxyfkB+8RplKcjfcit+bx9UI4LWKrZcRQaKYTmBjevIXTvJJhLd3UuttMhg9yjuziJSKJrlFUwuIKiHVJt1iiZLIFvscIqZSjl1jGi16AsbzGx0zL8JbDqCkUjunriXUlShJ5djoOjo8G9htlLmph0HGPaLXFq5yvGL57j1jruYSl6k0nsvPdUXcfWIcCVESUlP1xBzQZmo0YB6wrb+Tl64NEN2eJiZqWX6cl20zs/z2O//MTvHxpE/sJ1ypQTSI9/RweDECFEcc640j5i+zKfmrvKjRw/x357+CpVaSJhAM2zQOTDASDDApTOnGL/vbsb1AMvNOsP9Oc4+/yw78h6/9Pu/iVEeYWwo9A6yWl2ko7PI97ztPv78k5tPRSaEIJPNpEvlJslr3I35BpsWRHo+CA1CpD8F5RyvteVfN8lsv9+2J+td27LS93082s+3rWReq0xLJdFtGfOaA4BLFUVCCLSnAHVjnRlrU4FBG0pqhEz3Zk0qJHA4xJpUmTST2mwQiUE4B1qTzxZxXb0UC93ojh5Wr1zA9zVSCFZNTE+2QNNFdMgc3VpR15oojlNHhHKVuHMAr3+AuLJIl1JE1iPJFQjI0buSZ2X1BLVqFSElwvfQAhIT8/nf/D32/4t/ghfk6LsyTxBkKE0vIOtV3OQIiV8n8VrMnn6BTj+DWSzRpTWivMTw0ABz1SVGCh1crtXplYrtMg+dOeLrs3h9HpXEMa6yFLN58p59Q67qtOkIRkgPz8+RLFc4M3WO3r4hDk3splI7weULp1ntGmB12x52OEW9tMIte49QXTzB8eoyuzrHibozlJZW6MgVyDYlIxM7yRQsolJm/NYDnLxynM4rLZKkSn9nHz3veC/N1ip6agobh9hMBi089h84zO6de1n90pM8cmaFib0389jLL5C9+gq9ux+i1ThFZByLq9C326d5zjLYM0rQn2dvGLDSTAi6i7z41NM4VeQtP/QgVy8sUfWbfOTD/5l//MEf5qOf+BOy+X7gykaH/XVGOsgYtcK2w/HaejAexjg8rddnYFLotFfjpTMnCNle5dKuk82NSX6QQpJYQxJHJNYgkannmK9ASbSQSOTa3CRww8I/JQOHMannmDHJa4gIpFZIIdfNMI1L0Ci00jjpoO2yrKXAGkuSxJtyRUsnJX7fMEEuQ1chz2rsqNarmDikf3I7SmpqLUMurLNcWkYiwCX42QyhCMiZhARD/+5JwiSm2dnFSlZRbOWQtRaZmo9f9HGtFp4/mtr6OEdWKYyI0VKzsDzPTwzu4frcIrMLc5w5c4aB7m62v+1e5nE0G/M06pZMoYBqWPq8bla7HLNzcwStPE0/h1Ye+foicU+epF6l5iy+dGyzAtcy9BUDFrPQUU82OuTfFDYfwdiETDKD6h6iy+9nT8cAK8s1Bvxhdu4b4aunj3Gk0sEL09MoIbH9Pos1jW8lLy1domc5w/jBYZbOLdFMElr1ClfG3oG79IccsDmaj59j+8FDRM2YxAgKxQ727N7FW+66je07dlNPQIR1ckGGamS558h+PvbCR5k5EWI7clyXEwRJg4ndO5m5eprxvkFOPv802WwvdHjcfWAfJ168yrnjX+bWPb0cGBnklUqBhdle/t3P/yj//t/+W/7hz/4c3/XeH+OlmUt49ZD5l17c6LC/vnAOqRW5wKNQLKyrwVL/MItSilarmUqU20sg+9prZygOZ826fDmVJYO1cl3mHJsYKSVBkMF3aWNftEtvzlqcJG3KtklKtM00waGVRmuNQGFsasTp+/46mVmbmmKm3zWVLydJ8hqjToXv+/jKS8tlbWXcZoPVkmZg8XBMNWsQtXDaJ1so4hJLU8e0whAlEoLhAXzrKC3P0tvdR2lpgQHPQ+Q7qJHgTIQNQzpyebyGwkYeQ0PbmGtdwgsUXbqHMAzTlUplOmgbGUPUMixX6qy0QoaGxxgYGCGfzfHy0y+TyeWpTV0geOhuGqUVdOhRLMSoTAdz3UViCzZaJSl0MdpVoNyMmS1kubmjm6lyGa+7gwezOebqFexKA0+8Mdvlb8xv/S0gshGlepHF81eomAatUDIfL/J/kfeeQZLd5b3/5+Q+nXt6enLa2d3ZnJNylhASQRICkxFgovy/2CaYe218/Tc2XBtsbMDGJAM2GRkQQRIKgHal1a5Wm8NsmJw7zHTuk8+5L3q09769UKUt1Xyrpqameqqm5+nu85zn9zzP53t66jiLRg2n6nJuskAsrTJw38d54dJhcvU6ciREtFYjkkrx64eeJrn3rSiBSUNssPjsP9CbCVM8d4YgcLl44SLt/asom1VEUaSQW+DbP3yY0yfP0KI33Q0lLUIq00Y8meHa3T144XlaboiRdU12bljFyRcOEsQjNKolJkZHCUcjtIYVfnLgCUaHj+EGMjOGQ//gBu657Sp6S2f5i0//M0cujXPzG+7mO1/7Cl68wgrcwcP3fYqLi+TzefL5ArVaFcMwml+W1bSoptnTEIUmeLJhNqjV6tTrNUyz+btCwOULuOPYl7f1X1yaJPCbfZwXkf3+i2j+5u/IqoaqhpaNxpqNfElqJirHtQgCH1mWl5/P/wFovugRw/JRm6Kpy2PMEr7vUTcMKpUKjm0hSRKq+vJsAP9ecj1E06Q6MUVtKY/ZqOLZDpVKmZmTx7EWi6iBj6DoVF2T/GIRRwhTKC2BLDAezzCGiadHkMQwCTlMwlVJoRJJJqgbdSKRKGo6guksEovFkFUVSWzCTEVBYGYhS9UwyOeyLCzMslSpMzIxgWU2GB+/SLlh4B45i1qq0yYrhDpilMcv4S+WqFRypG0fzAaNXJlAFtmqxfCyS3hYOKbJ0uIigWGwJZrAbBhXOuK/k1ZcBeN4Pt3tbbhGDC+XZ3jqJDffcT1WZgtTkwdpS0ao1LMsTFZQf/VFerr7aVRrlComQihBWLbZdfMezj/77+iqRDTZymYd5kfGGHnqJO3JFEulOZ5/5mk82aGYz9La2oKqhvjN/mMM9PfT2d3TdE30PSKREPlhh1TbOiJOwL4da3juxCW2X3U15UqZnDXNNffexfbe7YxfOMSq9R0cGj9BOpNGDqUYm7zI6cMvYHsq0UyGTroYPjhK/7rNhIoBOC/P0vr3UbNj0bzQs+yb4rrNi7gkihhBAIGPbTuXj6gkSURRmhdySZKQZKVp9uX7NBq15tKbAKL0f6FhlnsqEDSnvDwHmnzM5eM3gQAf3wvw8ZdHlAUkWUFVmh89zw9wHedy8rpsaPail43QHJv2vBcp0Mt/l2Xb5+Xm80pTYNkIs3lSbe1U21PgytRPn8RP6yQ37sCYHcFEJJLqQsRD9ExQoji2jSopBCUXPSpi130ET0GZsvEFjYpfIKYlCakhTFFCsQX++O5riUajeKKALOuAhy0IZDLtpJNxFnI6gSRRWMo3raxDEcKOQ7lokZuaZV1/D0u9aeZGztKRyjB+6QytosR0pcHQjl0sRiR0KeD4yCWiIiTbO8gv5Ei0tEA8gleuE0+/PHH9K66CCckqB6dPcPDcKSLJCNWGzakz50jkh6k3KoyPz+K5DlV7jkMHfgWNgMGeIfYNbCCe2chS2SNkyuwbup7K0xc589mv8ugnvszCI6fJZ6cZnRnHaFhMzoyQn88xMzXC6PmzdHRkyHSnOfLs8xBIBIQRJRUvECn6An0bOvnOVx9nplSkJRHlwMHHCPkhUnqajvBqxh75AZvXXsXccwvcuOdG0t1tOOUiZ86PsKenHdETsMtlVvVdRX5qini6g6CqobW/PCF5v68CYdnvnhddIZsXfGG5wtBCIcJhnWg0SkjVCIV0ZEXG930sy6JaKVOrlKnX68iyTCgUIhyJIssyjuNQrxvUGwbOiwME+KhSc/GxuazZPJIL/GbSUGUFebnx73sehmlSrzcwGo3L/RlBEJadK1V0Pdz0mvG8y4lQFAVUVUMLhVFUFUFoUgs0beWNKYuigOAHlB0Do1ikfOhZIu1t6KZPdWkK33NRU0liiSgYCoInEly6AD4I4SSCpxBqxPCmDHRbRvViJCMtCIGMJELZqiEUq9hLZW7ZsaeZXBTlstWDoGkYtRojE2PU6nVUJbSMGPJomDXKlSKS3Ly8XnzmEF4ySWRwDYYJyUgCOZ5gzboN+J5Di6sQlKtELJMZHabKRSbDIuML0yxYNWKaSuC9PAc5VlwFIyPRGXSS1aYRFJt0LEGpVKGxNEMsmSKTVtjQ3wfnVPLxMUKKgiMbiI5IyssTikeZLcwzVZ3j9gffzhOf+hJSYYmF3AKS4GM1GoQVHdfzESSZ197zBkRfwq5UyI8ucdpu0LfuFFs2bgBFJBaOcPN1d3Dq9Fm2b7oJjAKFpQrXDN3IpWMHybs+SWOErltex3xtgbvf9oc8/IVPk9m2hz0dEc6XOjk/P8c9123kkhhFCExuX7WdwtkJpstJskf2X+mQXxGpqra83ChcrmI8v9kUD4KmSVdI01E1FU0TMCwbCdBCESTRB0QaloUQQEiRcf2ARqPRRPVLkIokkFQV13HZe92tlBsNNMknHNLwfZ9f/vRHRKOJ5lSZL+D6Hr7vLFdLctNnRhAuT50JgojtO/i2t9z0Fy/7lwRC0whNFJsJSpJFRDRkSW5WR8EKPAfVQniqhhRLkhIVip2dNCZGCa1fh2RYyO1dBMU8BlFaKxXyRoHYwGqCkXGkQR2lI0rj/Cix8BBOtURLxEa2ZwhpPq6xSEyTsQ0TQUtRNxqEQmFsx13uqTUtsNdvXMff/flHect7/xuTtQp+EFCv1wgpMmE9gm0ZSJKIZdmU/vXb9L/lNXgtMtGuLbi2i+j5qHWTeCaMZOtkZZEeFyqBw+LMAlo0iXH4OItrhrBL5Ssd8d9JK66CKbk1BtekiMdkFqcXaO/oIdaSZKBjEEd0COki8zlQ4wobbr6Rjr4B9JpPTBYJK1HKvsdSzWLL/X/NoScPMjM/T0trO0vVJRwvwA0cFhYm2LbjGt77jvcxNzHHrw/s58TYCI8cPMChQ8/zs4cfJ7eYp1YuoahR5seOYVo+bbEYs0sm+/ZswBEF/FSUj7zqGoz1WzhfXKTRUMlls4SGttETCpOtiRhinPH2Bvlyjnu2rOHAd3/ApcfPU5gPcOeGEa2Vd0QmCAKe52GYBo1GA9OyEEQRWVYIh8O0tLSQSCQuVyyeHyAhIUsSntdcbnRsn1rd4Jvf/S9M08L3veaypCxD0LRY9l0X33EYn55kcnoK27axHRtRkLjzNfc3rZhdD8sx8TwX1/eWR5hF1FAIVVVQVQ1ZbvZoEuEkejiKruuoIQ1FUZCWx5ybY88Q0LQisIwGhmFg2cbl47WVJFmWSW7ZRDqVoVwuIbRE0LZtwfUEPFfCq9ZQ23qQ6nlsw6atZ4jG/BQdg+txJQu5WCWSSqGHJQTHoFZbpFErU3MNlEBAkGUaRZNaZR5kBde2CTwPw2hQKpUIHBtJVQhFYnz1i/9Ao1alWi3jeA6WbaNpKoHffF3i8TieJlE6dQYRD1V1MQWbelhF0HVKizlaZIUNbb3IagTFBi0SwpI8Ml3dGIbBnPTyvIlYcRWMRIgXjp6mb3MvHXqE2WqDseMX6G3dQlxrwa875NvnSUspeuQkP37sJ+xdv5u6VkZra6M71YWQVJj64T8ghsBLigirUyiTIe665W7yhQJ9a4aQVZVHH32EcGuSwHcpLVWoL2UJpzI8sv9ZHvntk+zdsIV//MI/8eDb3sQ//+x50q0daOkoY8NnSbeGqebSPHRolDoKjUoFr7OXZw8cYN2G7eSdGgmrhujU2ZCV+enFZznwwjHe+c4/4HsPPYJyTS/G47Pkbl4L35y60mF/SdU8mgpQVA1V01Bl+bJHjKQoeJ53GXTp2BauH+A6Nl7NbfbGAp8vfeM72LaJKEp85l++zCf+7E+wjCYbyvN9DL9xmcZ8+sRxPEkhpqq4yTjppIrlBsiKhKzozYTkuvgCTfy+0CQ0m26AKAdN90vAD5zl7yDJ0nKFIy4PDojLE2bNxBYEAbKiENZ1HMe+kuG+InJ9j3JxhnT/RtqTG8nOz4Iss3btVvJzkzSMOr4QkNl9DbWZSczJORKr1lDNlQnVXEy/AzUkseTMoBLHaNQQk+3obgWxPUVhIUconsAQbIpLBcLJVmRFoVarNScF3eYNRzKdpiWV4vvf+zZvfONbMet1qpZJqbxISNObNytOA0mWKZwdYeL5k0iiiO8HzIkia6/eQyKTpD47j5RKYLfJ/zIAACAASURBVLoumqDieTJ6Smd+fp6WVAdZxbrSIf+dtOISjOeZ2IpCY6zAmaDIYGcPu4ZupKTPY9Vtugd6aTd85pbmGddyXLt3NwvzeRxHQTYnYUbi1r2b6dwe40tfe4jYll309g6w1k1TKM6ix9McP36I3Tv3YTfqjL8wiiD5SFqURHsHjmVSWzKwnRpjiRn2//Yxrr3+emZzeWbmCpRHhqkU5xh9coINW3fgCho7d+6gXqozNnKetB6jVM/Smk6RaG3j0nSOi9lpBvfdQUiN8/ThQ6wb7GPy5BRVRUR6/CwrrQUsSTKJZAJRANf1sSyzib73PLxG/f+yTJYQJQlNVUilUvi+j2mZ/P3n/4XAa7oJNtH/Jn/zvz5Pf187b7zvXhRVwfN9JFFARAHNoWbZeARIImTzZQ4dPc/orMPufnBc77KJGEqAKPq4gYkoSU1zMm/5cd/HXyYka4LWJCsLKp7rLMM6m+PIQdDczXEdh6pjI0or7mNMKBzFalhcOnsITdXxPJuWkI7pdqOsXoMzPEy4r5f52jyBGtC2Zzt+tYwbjWLn6ggFCCwPPaqjBSHsoIEcGDTEBg1DQFVFBFdB00VK1TKSFqFmNDBNs7mjFARUTYPw8kh8dyrB4Ud+yi1veDOIEnPzMyiKsjwZ2KxknCAgHItSqhURRQ2h3mD0N8/iqRJb33ovhhYiUjNxHBcNActwcWMhKktLOKWX52HTintnCm7AvsF1nDjzAoqsMLfkICdjRJ12OiMaPXo7udAUudElNsYSLEzmyS3Osm7rBqYms2QyKv/0L1/khltuQsyIvHbv7aRPLPJk6SiGJdDVk6Svo5NT588yOTsKXoAXSESlEPVSifaONs6dP09vdyc33nYDviBj1mqU58YZ2rKNUlEh2bmBV+67lvb2Nhq1BjIShitzzXXXcebkWcJtGTQny6nZCus721m3uo9Yfyu1XIC2aQMnf/2fvPu+V/G135xEkWVOzH3vSof9JdWLCcT1PLRQCDVQlz1amo/ZrovvOtiWjeO6VKtVqtUaFdPmBw89hCAEuEEAUnPh0nUdDE/g4sgUn/6nf+Z973kfkugiuvDs86ewvQb9fQOsHRzCciyOnbqEH6g06iaSFEUL6UiShOv7SAhYroMYNFH/nu2CKKIpCrKiICoKjtfAkyQUoFIp4XguIgJaSEdTVexmE+Dy/7kSZRp19HgLqaUF/FCSYjFHy+Aqjh8/QXdvG0Z2FFeVwDVIayrl7BSBHmNy7Bzt2/eitC7RWDRIda2i4Di0j4Ux4iDnbQI9TnVkHG1VD74i8ejhQ+xev5VINIauhfACD3d58dIO2fi+T1jVsByDb3/lS7z2LW8nnkxRLS4ha/IyyaG5bIugoqGRWttP9w17oeHj5srkpxZpy0jYnQmEXJGSV6WrpYVQIKGGJdTq4pUO+e+kFZdgEqkIeVGk76qrOLl0hkTdwwg8Dsz8lldv3ElFClgYM9nUPUi2lkVMy0QT3eQWc3S0tzB+5iwPvOtB9u//OYEjYu+/yFIAG7ftRZMVRscvIgUiuhwhkWonPzuNqClU6yVcxyObnSMei1EslvnQ+z9IpV5hZn6arXf0kZ8Zpz3TwsL0EiVBZT67SH9/B8PV57BmMvRt3ojkO2SiAXEpw1K5wWylTG1iikR+kGRII1uqMJ83+fX+33L1tg0oQ22c+PmVjvpLqyDwKZfLqKpC4AeIUnOsuOn7Ii3vrTSRK7IsIWsxPvfFL+N6IgQNBEG+7OXSpCprzY15X8Z14Mtf+yZGo8yff+wjnLs0T0eLw7wokEq3cObsMAcOj7B12xYWFgrUtnSgmeWmmZmsICgKqiJRrdQQRJFotOnzoSgKtmVRbzTwCDg7OslVW3oJh8NNrxhBRlVEGobR7P00l3hQJAlpBVYwAQHpSIxiooMW1UdRZIZPvsAt9z7A4WO/RVOitMbbWBw7Qz0UwlMkjJlJ+ldvorwwh+r7xNuSlOuLtHZ2EgQuclwia1VImA5iOo4S1ql7DX4+OkVCjZJMpmltTSMEAZF4DNt1adc0arUaBg2imszC/FzTltu2m6PuToCHhyhJBJKIF/hoqszAlq1U81VaWuLUTQGjoGCv76N25hCJNT2kCGN5Debnc6zq7IBtHfDElY76/7tW3DvT9XxqQYp4dpyrxBSBLjFWK7CvZzuKq3BxYoazE+e58aZXUDk1RtGo4oRb6BBCSK5L18AgT/zjvyNVbSqlGvrtg3T3dBARILdY58hSiUsXzxO4LtFomkRrG9Mz06TiMeQgIBxP0NbRzv6nHiVw6jiexfPDI3SFOxkeKSAaWdLpfkRZQ/UtdEmlx9kIq1V+ffAQd73iVg4+9zx3Xb2ZieFL9Ny8mbO/avDML37MwOAg1aUKpmtwbLTIUL1GqbH6Sof8JZcoioTDOq7nYdnNs+vgxa13HwIhaDb89SiHTg3zzDMHkBEIAmt5ZNjEFwRkJYzruAQ4xGOt1CpFBEFGcCCsJ/jcF7+M5/us2/5qtq3pRdZkXjg1y8LMDKbnsW3rEAcOnuGmHR0Ikgiuuwy+9FFVdXnJ0gffx3YcPM/nNwePcO7ceRBkzp1fzVvuuw5FFDFtA8EPoYc0fF8FmkwzkRd3blaYHIdypUhXR4iqFmb1tr24isjImSNEFotkenrJRENMij6Ba9GTGSRItuHm5+iIxYlGk5Tnpon3rcHVVQqKS1pPovW10lgsQSyEpXm09AyRPXOKQt3C8AtUDJNkNIIU0TENh9n8EkldpzOTxvVFBrq7+LuPfZgP/fXfIkjNPaVAlppIIVHCqDcYWjOItVBAaE3QMFyc7jZUr8jS/CXU/k6KtRLRaAtGNIKSTJEtVwnHXp6X6pfns/49FAgCQn2KOVvG82rsaVmNYc/h58Y4YkcYWNNNpKuTk2eeZ2BDL+roIjMlmXRvlGhaYfKXRzCTrdx+z+088qOvsDA/weH9T2F5LqbZoG5a+IKFW7cwrDqO63LNTTdjFAvUDIdbXnEjn/non1Av5Vgolzl+fpSD5xeYGZtlwytWU5qOcP2Nezn829NcKM4iVFKU8nM48x67Nm3g0ugIiiJw4KmHad8zhB5qcOLpn5HpWs3MhYts+2//naf/5sMkOwcYm59nX3sbKwwUQxAE1Gr15SqkiYkRBAlJEhEVGRER2/f54UOPosd0tuzYyaFnn8T3HLygOdYsSBK+5ywvOspUqjlARBF0bNck8D0ESUWLJXjskYc53bWN+vf3Uyxk6Vy7hsX8DEdrJmZ5iWPxGHuGdGy7WalEIhEsy8I0jebP0Rg/+OkjTI7PULdqyGj4vsPiwjmi+u3Ytk1E17FtG8PwEGSFkCIjCSK2YxN4K2+KzAt8DLvG4shZauE49VqFRqVMduQ8yqo+ytUqeWuSdM8Q+akLZB2DtlQXmVSC+VKJuZnxZtIvZbEWfQJVxDAb6KqG0xbHc3UatQbtegiprYdHnCJ3mS04QYNSvc7ofJZ4JIyuKgitrdiz86zq7aZQKNE2MAiBh+MFiEKwzJWD3Vu20DY4xAtHx9C9CK7hEVQLyFYPCa2O192F0r6K+tQIniQQswWEXBmvK0OlvHSlQ/47acUlmJios33frRReeBxN28BPR86xrbMTUzDYFOvmyMUT1F2bmKAzd2qaWWuJjKNiNLp5MjdMW6fD++94Iw994stMT06ydHoSKaTjWTae4GMbBgJNZ0KxUUMMR5kZGyGxpo1/eO+HODc5wtzCNHOFHMdGCxwdHWZGFikLRdyGgFmeY3JqGj2aJJXuIhpNMHDVdUiyj2vLZMdGaGvvxxbrXBqvUnnyMDe84V089/DP6bp6iI7etchaDCGaJtq/gUvZ01c65C+5mlRi7zIJOQgCEJpHZPVyhWg4wpFjx4kIBufPDGPb1vK8vkDgOc0FxmW6susH+J6FgIwXBHhBBQUJIaSh6zr1ah1ZilMoziMKASFZ4NKxx+kZuBpbsJGjCQJF4xsPnySixygujlMvz+NZPlpUxfcFouEonlfDcUXSrZ0UsnkCUcC3Gnz6H/+Ne+//AxJhj5AoooV1QrKK/yLaX2rSllecBIG3vP79/PD7X2F9WztzxTJD6zZTbuul5NapZmeRWuNUc3N0dXZh2zYXfv4jCnt2o9gGZbNOJhGjsyXFqXNnaF+7gYWx87Su3opXMgh8iYaa4fTT+0nv3UW8JrFUson6TfsHVZYpVg2skEcQLLKqt4eabSNqKk6t3uzheR6arvGXH3wHNUfg6FKdMaOGGJOwUx7RTBq7LhI3MgwXzyCcLTJQaxBfPUij7FIo1WnLpJB8B6H15bnJv+ISTD43z09+8jmMc+P82/v/lKVyO6PFMcxaHX0wzuq2Li4WJ+jojtPT0Yd1/jh2qpOl0igpy+e+ex7gqeePwdU72birB3s6x/DjzxDRotiSjOcHdLR3UsjOYHo+zlKBv/mLP+WWvTfx4P/4HwzevYO/+M+vose3kF9qsKp/K25+nl2bewhmi0wuGGxVs+RcD6lWxdOqzE2No2oRJuZrRLSAHf0p/u3QDOsH+gi39/P0Y//FXR+8lQCRWOUQm268gflLE5hLo/T3ruI4z1/psL+kkiSJRLIFSRKaZBWaeHXf8wmHwxAE3HDjdYQ0FcOy+NpXv06lWsVyXERpGZ0PEHh4jtU8P/cdZElClXVcT8B3AvBBEX2QVOz6IoIgokZbuePVD3Ls8AFauntZmimQXayBrLOYO029WsB3ZZBdXMvHsm0i4TSBLBG4BrlcjnAkSqOURw4p2HaD//rh92jvWMe73nQjICJKAoZhICE0YZgrcExZkUOMTI3iygKzREgkRcp1A0f0sQyXwPZZquTo7OphXTTOucIim267Az0sMzY3Q6x/A2ojz1SljBJRsItVEmqSam6Otp4uSqbFmoGrsHyD/KULiB27OKZOsdvyEQnwBZBEH9vzsWyH0ckJCsU46XiC8mKewHW58+Ybuev6qynXDapOwNkTdYSkgU0WoRjBKJvYjTL1UJFQexrNEchZGgOnpvAki0ipSsG2SSWjRNs6rnTIfye9PGfffg9pyThzlQLVNQk+dugXJAcVOqYWefjPPkukxWWGLK6usHHteg6f+C1dA7u5pjtFZ2sv12+/igNH9hOy4PbbtqEEdV71h3ejb1jH0H038Ad/+j5SfRnmZ6YYvOE21ly9j3V33ErV0znwq4e55pqr+Ojd7+XD7/oIuuAx0BmlWCxRyee5dlOGN93zCoIli9qUxMGfPMNtt9zDz556CNPVOHN6jFV9HQytG2SmOInWkmGhUWSqXGPPHfs4+PAB2rp6UGI+21+5AzWWJj8xzrGTF650yF9y+b5PuVyiXC5TrVWaAEvDxPVcJEkipOuEVBXPbY4iv+uBB/jQH/0Rb7jvfsKqRuD5CL7L3p27wPcJ3CYRmaDZC3F9g0gyjWXbBLJLNB7GcVw838Ku5Dh97BhC4FEqFPEEj6VsFgmBSHwVicR6bKuG4IPjeHiOS91o4Do+nucgywFGrYAoKxCEME0Ty6wwP3eOz37h6+TKNSzDRpVkVC1EKKSi6eErHfKXXEokwtOf/wJeIBAszRLuXA2KiGk2yJ4/DV0DCDNZpkaGmc5OExg1SkaD0dlxkrqOLqo0TBtZ1qlp7SzOzFKRPBbxKeQXkS2fytknqfR2kEhmqPlV1LDEVNjF9j08xyHwRYJAwFvmzCmKgmnZxFIpPvbgB9i7ezdFN2C+bvOj8RM0GrNYuUVCShJThFp2Ds+p4oaqmEtFXF0itGkDxXKDaMGmGAmItiWpLywx+cLL8yZRWGlbwD0bVgXf/NV/8Hf/+hPe8I67mJsYIZWO8elHf0yX51DMVemuihyrZdm1fiPEdBZiMqploygxrurvIpLqopZd5Me//BEfePu7qRQNzjxznHvvu5UTZ0c4dPAsputx8+7rEGMh1Nl5Th5/lu9+92E++M730PAsUqk4a7u7Gdy6le7+XpxKGSFw8AWBzz/0KJs7O6iUGvRdextn5/eztv1azp49SadkkA1c8sdPsuvOa5iZgscefZj7Xv9qTp46xare9azqaOPk+bO0yTrP7H+K2eHTR4Mg2H2lY/9Sqa+3L/jIH//JZazHiw1+27abvh5+E8cCy9RjWUaSJDQthCCKeK6L57nYtoPv+8zMLfDzX/6Scr2GgI8kKmh6ikZjiUikhYH1V3PhxFMEgks43IMV0tF8H1UJI4eSCDhY9RpmbRFFU2iUZ0kkk5TKJRzHRJajJFo6cVwb3yphWSZaOIFtVnCdOl4QIEsqsqwSjibp7WplYmwKNaRjOya33Hgr3/nOV1fUa6x0tgVdt99MaO0AlaUsvV19TAU2HR7Mz89gL2XxU+1kOldjLM2g+CKWJrGtZx3HzzyHGYlDOU88FKIs6hileXRNwRueof89/53G9BH8uXlCmTQ4DoIo4NbKqIjsS3QTFEwkSUNTFcKahKZpCKKAKgbUTJP6YpGF+XnS3Z385MJB2tQk0lKGSq2EFrdQu1vJj51C7+2ASJKwAAteQNisIzkCOgrt+3YzcfQoWm8GfJ9LX/j2y+41XnEVjO+6fPyLn+e9b7uNv//cJ/nN6Ul+9ewxEqZH4HrceNWdXHCLbLOjjF0cY9ZziJsOOzoG6Mm0MjVfplYp0ggC7rj6ToJIhHB7krpU4rcHnqbaELn+2uuR2gUW6g3mp2fZvLELeWgH7/3rB/nUZz6JMNiFMNCG1dXF+akp/u7L3+LPP/W3/NfDv+Sb3/wGOzs6qck687PjLJ46RPZ0iYmxWVojSe559avZq0q8/S0PIAQi1eoCD7zr3UyenWRwxyoCUeah73+f1tQAzzz9FJ2r117pkF8BNduqoiJdZnoJQtN9MhaLkUymSKfTdHR0kEwmURQF13WpVMqUi0uUK2XKpRJGo45pGrS0JHj3A2/jEx/+MLfd/EpSiQSWWUQQBEyjjm9ZKJqOa1nYRg5JUNAlFatSpFEtIgoaptfADQTqZhk/CCiVlwg8C1lScZ0qqVQCo5FD1eKEohG8Zb+awHPB9/FtE9sw6Oxso25JVBsVdD2CIqsk0/ErHfCXXr5H0N1B2ZNYmplA1XWUxSVyc9M0CIgmUnj1OuXZSRbmp+ndcC21XJbT588gKAqrenrZvWEnZURCtklbWwepRAvt192AeOZRVFcAXyE/MYGMSM0OiHT2IcY6eM6rEdM1XM/BMA0My8ULPDRRQETC9yXUsEq8tZNHzj9PQg1TzRWw6g6uncYLLEqT40TXDBHaej8uIql99yIqAnT0Iq9bjWEZ1GsFtK4elkp1GpXqlY7476QVl2A8zyOs6Xzwbz7Oa978Vq5el6QlE0HwLC42bHatybBz/U5e9/m/4t1/9THKToGdGzZy4txpglAIkhrt8QyrMm0oSkDDMokIAgs1n3TfTnozLRCJ0LN2DUK9SsU2SUpR6maOoqrzZ1/4Al/6yF8yP19golbn2fGLhPvb6N19A6+84xZOnhxBkVz+4h3388ZX301ItHnna27nf779ldx/zWqe/PkvkFr6OTw8RX4GBtdu5MT4c6zbtxXJ7SKkx9m+YS2j42fY/t59qB0rj6YsCAKKojTRLra97L/i4rhes4rxPbygWdEgNCe7wpFlBpgWIqzrxOJxwuEwoVAIVZbxPY9SvcKmdf28481v5ON/8iE+9qE/JsBhZuY8r7nnnSCC49aQrBKe6+AKNq5TZyl3BsnziUbTSIKCH1gEgdfcl/BMBFlmbmaEro6NuL7RxJCoMp7XBHMKfoAggigHjAyfpbAwjihKLC0tUK+XycRW3hGZqoVIyAJafp7VPauw0IhF4qxasx7ZaqCEW5BViQ3dbfQPDjF8+glikoBtLSGGW8jOTHFh6hLtyTaqRp1KqYzt+HiCjRBPUi8tUVA0YuEUpfkFQukUS/kyiqRhV4qUJIGIqlGpVMgX8kg+aIqMIPh4gctEbpGnXjiHOVFCyCloWhpXzSEol2jkswhmBbGwgPXr/8A/cQln9jwdvkSyFhAuVjDLJUqOhVPKIyMjBy9P19IV1+R3PYkXzh6lJ93KN7/zPZK+TG5khHve/yZuCcewZJ07Xn8P/vwMP3/mt8SlTo4fPMqaZJrc+XHmSjlKh2eYn7iE2zBpufZaFtwqoS6RfD1LvhLCsgsMdq8nssOhkJ3j6YkZenYNMXMxixHxeeDTf8knXv8mvvfMeXYPbmW4dJKd1+9iPlfnHffdzX8+8Rscz6RwYZKrX3snwyfP8Jlv/TubhnqpzQp48V7qPkxMzXLvVeuYH+2gLZEk0x3wyLceZ/vNu+joMNDqEi1blzh4pYP+Est1XXK53GVbZFXTkAQBRVEvu0fiNxEfftA0CyOAUCi0zI6yl5OPiCQJSJKMpgEElz1YggA0ReR//tnHEAKRf/rqt4joCRzHQVbSLBVOEY+14CshIok4lVKWwIXAriDLGq5j8aJJoYKCooQoFkZxUcD3UZZ9Z2Q5RODa2J6HBHh4OFUXRZLwXAfHcoiEV16CCRApGpUmriUcpbGUh8UC+SAgmu6lUK/TkYjixeLolRJ1LU7d9allF9ByBaRUFNFwcFSFdWs3MZ7Pora2IFYWUZUYhVOnWfuK2zA9kXBuAtHxiEgeVauMoCXZLzd4kxglL3hcOHWSi2dP8p63voVMZwf5s8OMr19Nsl6h1WrBKOaQpRhiukal2EAWVLyuDNrgIEZpAbc6Qe5XTxPqbyXe145VztO6ZoBqqYosy2iWQSz+8rxRXHEVjCSI7LjjTmqGyfpMinq+gZTUOfj0s9Tyi3zzm/9ObuwCR8dmiJYMrhropG4bPHNumAEhwp1XX03L6hamTkzyof/2CSQnx8dfdTOv2baXDXs28duFX5NubUGteXiGwzWr11KIm1CyaWnV2Ta0ky2bb+Yn40Xe/9preNW9N3D/He9mc9/1LIoajYbLtrW9PPrkfs7NjPPUsUscy07Rl+rg1u4N7F6VpNyYRAZ27t7CxGyBt73iajrS7RQvLbJzVQ/zhUuIjkF+ukS5lLvSIX/JJYoioVCIUCiEoqgQBHhBgG1bVKsVarVqc/vasgABWZYRJWl5ydIlEJZ9V3SdSDiy7EapEQqF0XUdTdPQtBCKqizbKwd85IPvRA13Ydk1quUJNmy+HVEyCaw8luFQWhjDMm18bFzHxHdNAPxAYse++4hGkqCEUCRQVRnTNPjAO9/Jn//ZR3nwA+9FcBwEsTnhhucgqyEGuwb4q098DEFccR9jzOIiEccgogS4pk0tvp6Sr1Kcvohse8Q1mTohzp48jqtqWNlpdt/6OloDDSQVXwmTWDuEa7rk/QDZ97ELOcx4K4s1k4SoMTY1SnVhhkXDQFMCHKtOhybQun49/XGV4YvnkCwDTZbpaEujCB6OZ1OqVChMXyKeihHa1EnnbbvwPR1d7KWRt/FI4awdxO/tR0Ciq3sTZshi1c03Ujhyjqpfw3BAODuOM5tDiGhkc2NXOuS/k1ZcBWM4BqWJC7T1pAnFUmT6DLp7N5OIyuzauJV8NkubGRAfWs2qTIpWVWImN8u61f3EevpQnIBOrZN3v+f/wwiV2bFhN88+eY5ayMAfzfOBV9+H7EM8k2EqN0mkJLN6zxZmLlzk3FNjXJge5Ya+zTiqxueen+cOJUou1IVrXmK2VONTn/xL3vueB6hXFnjtGx7gfW99Ez/47ndIZdq563Vv5PCx09yiGjxxcpRaPc9guo1vPHcYLW7gFERaNvcw/MwBjh4X2bxtI1bZvNIhvyLyvQDTMptNfgFikSiapiFLEoEoEvgepmljOTa6rqMIEq5pUa1XcT2PwPMRJQlFkZsN3GVzrxcHAgzDwPMEwMcnwPdcAhfS6RZmp8fZuuHVjF4IEDWZsKyy76rdDI/OI4thbr/9Njas6eOf/vUrKJE0A4NhTr4wQiDIRLQIH//wh3AsFyQB17XIZDr45F9/kmg0xre+8U3uu+916BEd0zTxfbCslydp9/eRpIWIdfQTUUMsXjxBpOzhCA7VaIJieYaBntUUalUCx8dUMtjJIgvzIwjbt6HYMJCJMnX8EFI8Q3H4CJ4t4ZUKDNzcTlf/EAXHR/NshMAhoqUIKWEmTQczW8AMtRNzVK7t6eXEyCTB+i6G5BgnzpxAMDwqvoocNVnUIxSrWcScQ8+OtQTDSyRaNxDel6SSXkdp9AKS3oZ/cZxEXzdnf/Y4iYiCHfg4UQXN9fG3rUOLd9BwTl7pkP9OWnEJBjzW9w/R057mpz/6JXXPwbRgy5pe1q0d4k//8EGe6HiSofvfxK39vTx65Fnuv/H1TGYv0p6KMTk6Rqa3kzMn5jnxrQn23diLE+6jUDtLixLn6f/6NWMLF7n3Pe9i6sAC+WSOySPjtLS1cO3WHeTbXKyZPHOT57ntxlcxL7VzZ0eKz/zw12xZexVGcZFZW+NTn/k8Zw79lrqW4pbXvpPq/AVEARbGL1ASJXSvzqWCwY9/8VMe+upf8djwGHJfgccOXyC9eQ1vGfoDIi1X8b3P3XWlA/6S6/9wx5rulQJNszDDMJoe94rM5cpFEGlU68vWyc3jMEEQ8YRmJeO6LgQCgiggy82jMcOoo2khvCDAsZuIFz8IePe77+aRXzzK5vVbaGlNEQQeomPguB56Zj2cGUXSFK7duw3LdtCiGcJhlc39afb85V/QaDTwXI9CIYdA00FRkkQMy0IQBGq1Oq97/f0omkapVEKUJKLRKKqy8hwt8X2yC9P0DWzE1mKkq0eZ8y0GeweZLZYYmZ2it6uTUGeScFcXLd0Z1HKWYiFLS0hmcb6CpKl0d6QpaCIFx6f76utxF7NML85TWhijrXeAhhRn9OIFUr5NMhanOjuLdHo/XkcvuZLP9LYBRAJ+MD6BqsKrQt0c6lFwKiJ+opWIVabmVlFEhZ7eKEqnS86sICoyuqQSOAaOriPPQrUd/gAAIABJREFU1Ej2dFJamkLTIti+h7huLfVCjcpTDxPqbrvSEf+dtOJq63g0juk4TI5Ncu/9b+Tu+1/PnrtuwpBUHnr+eV734Q+y5eqrefP6ddy1eT3nHtuPUBhnTWc/lWKJYyPnsNQwc40cNdkkF7M5M3YEq+6ipEz2vuImdlx9GxeeG+aVG/qIrepmx66ttLa00zoUQVpYRI63UlqS+N4Pv41YqfKNZ/fz4Ktupmedxk8f+yVb9mzi8aNP8ORvjlEvzTFXGGPWNHjXW9/EzOQlLl46ydjYJdS0Ss/N6/jz7/6M9991JxcuFNixdgtywePipd9w8jefZfcdN13pkF8BCUSiEWKxKLFYjHAkSnT553AojKpqqKraPD6jaU7WxLAYNBp1LMvEdZerGEHCD5rIGdcLsC0LXQ8jiiKqLKGHdHRdX/ZlCdi9ZydaSGXLpkEkScJD4nVvfhtWpcZtd93FQNdqAh9kVWX71l28/pWvxPU8GnUDWVHQw2FSLWmSqRTRaATbcfE8H0FWSCaTzYRTbzQtg0UR13URhJVHVRYjOqoFZ557llX79lEJQNYUzFwWy7PoHNjEkulR90RmHv4+8fIii5UymbYBZmYnqRZKiJluZvM56uUKawcHCeolKkuLZC+dI5AVgniayuwFwok4gqiiywJeJkV8zSYSsTCe51At5bBqZcKhMEc//iluvuEmxEBHWMghWzZLSzUi6QSppIqSijDY1oUv6mi5WUQxwNTSRLUEem83RlsYrWcNDRfUS3nq8/MIYxPIEnjmy3OZdsVVMNncAu3JGNVcwPDp8/h2A0EOOHlhjKuu2YkpdpOdm6RSKvGJL/2C1f2ruOaaaxmenuQnh57nlpvvokeNccMtO1i1McPI7BT7rl6N7TY4eXqMx39znFv2bKM91cqBY0e5IJn4DZOJQpY1XRlcT0c2Rpkom7z+7qupuBWmS3McvOBy4oUpxLVxDM9n9HyB//z6l/joxz/KZ//xa7zhhmt5359+kK8fPc7sU09wyz1vRvYSSJVLaFoLmzbfwonnH+N//ceXSdRl5qrT7LnlNcwuvDwXtH4fBYGPYRiIYhMVI4kSwWUeZIDvNBcug8BFkppgzFBERwzA8wMEmpNoruNhORau4xAELs8feY67XnMP1aUlfvKzh7nphhuYm8tRqReZmZxh1fpNEEis37oH1xZ51X3v4MlfPcFzB/azbu0m2tMiu3as4wtf+ToPPvAW7rxpK5LQxAoFAjjLezeS1DSkEiWRaCQCgOM3HTF9z1/+v2RkLYTg+/j+ytplA3BNCzceYc3gIAunzxHRZMy6h9DSgj5xkUSqHT+aoDg9Qc+urSwaVbKLRerl87TH4kTX3YRfusBMbo5EspPpo4cIRJFwVz9Cfga3XkdTowwN7WDi6BGW6gVIZ/Ajcax4kvLFYbabPo5cI9Q5gCDDzv/4Cq+oelTdCnFdaR6/Fj2q1jyHpSIdmW5WVx1CS1VkF+qzJdriPSRaEjhaCF83mLkwSSaZxBVcxFVrCHW0UaosoNgBPHmlo/7/rhWXYNLJJPPZElXPIozG9ldcT48WZe+2GaamJjnwi18h+BbfvniKsNKBWS9x75vuI9HWhyQrzIxfZP01V9EjhUlEWvjp97/O1qGNlMuTjB08wVve+l6ePPQoH3nXB/h+bZaFo5Ncs/s1pP0zrN91E8cmD7Mmuop43mbD6jdz+tRPECIxhusufT0RwrrOuaLE2s19/NszJ+jdsJMP/fG7+PX5i3zx7/+a6YmjxDSViakLtKYd1m3tIypoFF93Nz985nHuu/EO/sMRuSCNI1RBXtSudMhfcomiSFjX8YMA3/eQNAXfcTEdC9swsT0XWZYJPB/HdZEkGd9t2rJVq1Xq9Tq6HkZWRIJAxvUtNCXMbbe9gqOHD3H40CEqlSrf/u53ue6aazh7bpjBgUHWdnfzmyOH6e5dg6Lr7Nu+ls2btjKVs9i6sZX2sMBb3/l2WhJpjrxwlAsXh3nbW9+BHwRNx0vfR5ZUbM8Fz0cWVALBJfDBcx1sP0CRJQRJBiHAsy2a1P6VR1OWJIWK0cCYn6VH1RkZPUmqb4gAl/BAPxXLJxKWKPhQtR0WRi5gaSpbol2MNvKIlw5Tc4t0Z3pwlOZOS3VhBgmfdPcQjcBnIXuBkCDjdPSRiCvU5qcQtTSNkfOs6hngkjFF25qtjJw9TrSjg5Dn8KS2Bk06RSMUpt22cHOLVK8aZPvqjUyePY4Yy9AjDGBbDvHOFCklyrRbINaZIbwwSqrqs5SsoXa3o47Pkhu+SF97D9WweqVD/jtpxW3yp/ragnd89pNE+sL86G8/T8u6VuTubq5t6cWsVjlw/jwbt6+nNDFLoqubNX19/PyhH6PGoyRWreX44aep5HJs3bud9Rs2k8l0kI6HOXUqz/y5CRJd7XTGXFzf5cjkFEZxhmhSR8ynyVZzdO8cYu/gEI6noHsyDSGHHaiYZQsl0c7UqRn+/zdcz1ePXEKXXBBsVmXWcPLCcd5352uQ9RAPPvheVvevRt6RIJszuW7fDSyOTaJGW8k5Bq9ub+PD3/g8r3zNrUzMTbL/kz962W0A/z7q7ekJPvTgH1Fr1Jmbn0eVpMuI/Fgs1hxVDnwkWcVzbFTtf7N33mGWHOW9fquq00mTZ3ZmZzZHbdYq7EoCJaIQIFuykQgmCYzxBWxssMHm2tdcHIBrGS7BXDIIEFgIhCSQhLRa5bAriV2FzXlmdmZ28smnu6vq/tFnpUGAbIK0Xs38nuc8U13dU91dX1V9udrH2sQcFkUhju8hAN9PoaOQKIox2lAoFpAy2cF4967dTBbzVEpFiqVy8tVDKfGCgM6uHl70kotobcjg+z4/ueUWGtI+Y8MjOOkUuWwb82c1Mjo+QVNzIzbWaK0ZHhtl+fLlHDl4iFPXr6dWraGNplqp0Nvby4JFi8lPTNDY2EAUhsRG4ygXrWP+8RMfn1Y09rpm2cVvvwx31nIO3/htus49n6BQZCzbQGXwCJlMjoznMVAssLyrk1G/kXxpnA2TPoOtDo9PDJARMT2pNIdVF5WDD3NmWzcvTbVQqlbI1yK+71Rpb00xOjCOGxbIBS66uZuxwijx+CRu90LidCNdw0cwtkox1cTIlifIzW5EEVMYH6Ohcy41P6B52SXUJnZT2vsA5zYvRguJcT36s4IFYYYiMXHlGLv3H6KxpxvpCzqqjRx48E6qGR9VVvRv3XTS0XjaaTBKCK6++3ouv+A85p+2msLN23nVlefymR9/j5WrVnHBBRtYNHc+jWefw7/duI29X7yGYyaiyTGs7+li5FgPpdZGJkdrDI9FeLk8hUKRw7V9dK/pIlWUrG1ewlCL5NBojbWnXsCeoZ1U22DDwvU8uO0JIlLsHn6CXN7j4OOP0rl4LS3NGfYXHuKicy/knoe3c2S0yMtWreb2xx9m5JGddJ66gkPVPtZ2nMXpc5aTTwka27txjWbn9n3MaWvHoujfUeInpW18+HVv4+rttzFvdtuJ7vLnHdZaKpUKIyMjpDwf5SoCP0UQBMmGlK6DMSbJ8vcyCCEQSlIulvA8H2zyzfc4DCkUCyjlUKtVSaWCJLsey6JFi8iXCjRkskzmCxwbGWbvgf3EUcTE2Bhb7t7M4OAg5XKJXC7NUSExcUQuk2XHtp+x6PLXk8llyKZzBCmfqBbS3NzCD6+9DutJrv/Jj+no6ODSSy5hciJPEARMTI7zwN33smr9OpqbmyhM5GlpacbzTk7p9reBxTI50I8zOYpctJjhPXtobmkk5wZMOClSyks+a20ER44NodJ5TK3ME+MeNDeyPNeONzrGPhFD727esfgcFs5qoWlWC3sOjnDRuhXc/u43s/aSN7GjOsKozfF788/ncztvJCcc3jZrId89fAizaB7prmaOjYwzv1KBlfMJR8aZX0mR1xkKu48Q+gH20U9y6vmXkFn5apyJkPHSICYnGTp0hHBWB/GYxKuWmTtrMbUIxoIazYUa3oI25lQbGdowC7ZuOtHd/mtj2jn5hQA72s+dd2ymu7WVvJSMjUxy4fnn88htd9HV3o31JB/50L+ycm6Kt37pXzm2cz8t6Tau/tTnWbl0LS9+2UV098xm4IwPs+/Jw2SUS61So62pg/WnLuPh/p3kiwKZqbF04UKO9A2hxwocGa9SPDLG1q1bSActDFTKXHDJS8i1pnh480Pkhhu4e+tdPDq2iz88dQ3joc+ZPV3c1zfEa087ja/fdAtfvfX/8f6//xCHf/YIbTpFWPE4/ezFHBg7zJN92xmv7GLO6nn89PDjnDH/FPKV6RfCmnyvPiaXayCXy5FKZUgHAX7gk81mEEoRRzHSUSilkEKijcFxHKI4TL5eiSCOI3zPJ5tJk05nMMYSRTGBH5BryJFNpfF9n3nz5rF6xUqampqSTzNLRblSoVypMH/hKnrmr6Wjs4daGFKLIhoaG3nkZ4/gOR79Q0cZHR1jaHSYifwkb3nLW3jDH17OlW/6I4aPHeNLX/ky2598goHBEdJBmjWnrqN7Vie+8mht70Aph2A6bnYZxeQWrWFVdi5BLaKjvQNwKObHUSamNjpCXCiQSqcJQ42qhvhlzZ+95AJyBwaQE8c4raEHvzTGX6xZTcfsuSxbfwa7+idpbs7wwLadvPr9f8Z5G89mdcNqTmtcwWmLZ7FoRx+jP/gJS50sQX8/3o2bsNc+wKJtI/zdKy9jwa5B2g4O068necmFL2dWtpnueaey4bQXsTQjEUPDzF/eysL5XRQGxnjzaS9iZKBMj/aY7QQsTadoiMqIB3ZwrDpMpWMWuZZWZg9EJ7rLfyNMOxNZ0JiyH7zu42y67wHOOeNM7v/ijaxbs5am03v42t03MHrdFq78xAcQXgppLRNeE733PcaytT0UaiWEiNGRpWfJfEwYMNi7l/69vSxcsQqn6JPOFwje/hG2fOBdlNJ52k9ZSDx8DNdbBO4AG5as4o57HyYk5sjIEOVdQ1z2tteyb+tRFqw8hwULW+k/+ihntrZzw8gkdnCIqCHF3v6fcc68c+jqamDzt7/P56/6Ou//7j+wYs46frbtIPnSXl6y+qUcMv20BoLtveP4sSY0ZR795A9OOtX6t0FTY6M964wzcZRCKoHr+vi+T61Wpaunh4ZMFmMNUggMgoZslmqtQhxpyuUKDQ050pkcAoO1CVOx1lIul7A22VqmVCpRMxHlySKOKzkyMMSRw4coFApIN4WuaTQhqXQj7Z2LWDK/jS1b78dEEZmGBl7zylcyOZbnhh/fwNpTT2XRgvns3XuQ+7bcR2tHG45wiOOYc846i+u/fx1eKsWs9nZmd3XR1dlJU2szzU0t5BqymEjz1x/5m2lF46Czw7oXrGdBth3V0Ul1dJQgk0ZnGxgrFMj6AkeD1SFuYxt6aJQ5bU10VRuZv2QOVglkTXDs4CH8phZcX/DosR10+s3sfmSEU89cwpolywh8ybbePoojk2xc1M1nN1/PqmWLOHL7DpasX0drRyP39z7J6zdczNnzZ7GlVOIvPvB2rnj1FTRMCnpevp5tO55geccqWjs6+fr1/8Ell72EowcPE2UaePjgTszAOK9Ycw4tDlgt2Hl0kMWNHvvDiLt33s85szaQ1xNc/5mTzww67UxkqcYcXvMiLrt8BXd//ysM7D7AFS//A/bduYc/Cs7Gu3wj3aMt5Jo62fbkdtriMTprKToOQH8V4lSMbGtmgdPMzif3wro3MEv8iOU9yzhUGKKfSRZu/hbLXraMxx7qY+hQP2fMX8awKLNo9nJ29B6gvaWZ2GuEyMLLFrJ44ToKUZH2llY8NLd8bxPp89YTVnPkdYGVc5oY39NC88AxJsiSMg5vvfx1XPm+P+Zr+zZRGBhl1cqVtM/pwvbn+cHnrsZbsZT3vvdKbrnr5hPd5c87UqkUq1atAiVBJ0zC9/0ktNh3kU6yuWXaDzBYdBST8tJUqdHY4uMIxeTkGEoqCoUi2YYsANnGBuJahON7KM9hdHiESGvS2SyZIIPjBlTCMVyraGhspFgoU6uG2LjCZKGIsIpiOY9SLgcOHuDgoV5CazhwsB+tNTUd8YZLL2NkdIT9vQN0trYwPjLExjNOZ8260wGNsQanvutAqZAnqlXJZnMntsNPAKzW9MzqwUtlKR46TPv8+QxP5GmN87Qqh9nZxVSiEoeP7GGurxiVMSsauok7m/jeNd/mtCsuQDw+yT53nI2pJnqPDLA028Pu/AH2dU7wyvaN7Nx7kEw6S2Uiz9nr1/G1A1t46Ytfy4End9Jx0YX0BzHL58zhnMaAb95zPTdsUjSnGrnogksZHg+4t/Q4r90yhwU9KwiqsOXBLUwc3EZm/JX4UcCG9tkcuOchHDfN7qMHWL9oBbHQHHxiK3LhKvYN9PNXb/tLHt/2AP19Iye6y38jTDsGozyf6//s73jr69/OmdEiHg7vJ18epX+on97ewzS1N5Hrb2BJ90KWr1nB7h076FzQTRTVkCMFxITkZetO5SfX3MjKFWs5svkHVCOfBx+7l8WrexistXBweC8iHTC3u5ude0foPTrOrMVz2fLYNub3LOG6r17Di159Kb2FEm8+6/fZtvcQg7054upO7h87wGWXvYpbHr+b95z7cn64q8ChoyVyixo5MF7jXWeu5vF77iZkgChfYPTQIKeuX81EucyN224nWzKc/tbXoQoTfO+6O2hqbz3RXf68QylFW1tbfWt+U0+4lERRhDXJ4pTyfErlErUwRAmF4zo4noMf+AgjyGYzaG3IZnM4jiLUMa5UmCjCxjFKgOe6LFgwH0cpiuUynu/S3NSIHzSxfNlSHtryEFI6NDVmyaXStLS1I5VDVCsxdGyYwwcP0DG7i8BPMTA0TDqd5oGHH2F0cpI/vOS15DJZ8vnJen5OHiEELa1tSEcSVaoUSyW01thpGKacTWdZ7DbgO1l2ZceIhwfY2LmMyWqJTCpNICN6Zs1iYXMnd++5mVycYv/+AwRellXnnMX3f3or6/zZNDZ1cM+9W+iZs5TTl67m1m/dyRUbz+cbgw+Re+AQ7tIsuqS5Z/dmipNFhoNmLvvQ+zhyzQ3MXtbFd66+mgbTyqLl63AbDN/9wj/Rc/65qH0hZ19+FseOTlLycmz99NXMmzufSy5+J9f+6DoKxw6TetnFtK9sI5PvZGLPJryeJVTLQ6xctJT2+Quxjs+em2+gr2+A/oPjJ7rLfyNMOx9Mq5Pm1We9irs3beKu++4k25zhW5t/yuNzJW/6xPuY85ozWHjJWjb1b+ej3/0c91T7+Ox3vsbV132P7YcPop2Im27+CeTSTFTK9A710t+7h9oCj4IWDI/vYjwfc/TgHtJNDis3nMJQvhfpQF+jgwnh7e95I6e1tvKa8y8A3c/Q/hHa5xr29N5DoxswnnFRqTRhMaSoj9EauGxYvYqdYwO896r/ydymFO96x3u54967+Pjr/oJ8QZOWGZoaGqiFk2S1Q/9gES2O0dE8/RzAWmtKpdJTzCW2BmM0QV2LqVZrlCplMuksqSCgUCpSKhQYHBhg15M72blnB3fcdSdP7trBwOAAIyMjVMplIm1IBxlSqTTFUoXm5jYqhZBCPk8U1fCCLLNmzSad9jnSe5jA85k3dw6jYxMc6e0jl0rT1tzIiuXLmdfdwzkbz0YYqJWLGANtjTkGh4aZP7uH+x58gPu3PkRDUyNNzc04jgtKMj46wsjoGEEqzezZs8k1NCDk9AtTFmHEY4O7ObjjCXLHLJeseAXp0KMznaK7uZmeTDe1QszWe+/k3DUXsHvHQYLlXZxz3nmoCU32nh0syqZ4ZGAHpcYJtk7s5UvXfYpDnVUGm5vQjzyO05Dh0o2X09W1nFxfhUvOPJ8/vfydhA/vwAtaaY1m0dWzDL9rDtn2CjfffjN/8OG/ZPXZpzKvvZElRwfp3fM4UfkIPQvn0bXmNKKaoU/30bB6DdueeJhVqfXcu2MzQz09bNr2MAMFy7e/8R2KQ/3oSp6ykwZr+Mhn33Wiu/w3wrTzwSxcuMguX7qYI+Eor/j4+zk2PsIT99/P4sUb2TAnIJdp5Jv/8V2a582jcW4XulRENXShi5NQKdORyYDjUCtU+eoDm/jzi99N9chOxtyQUlWwd9sTFHSIrUa85GUXQjVDJepl0ZzFCCO5+Z47uPjM86HYxtUPfZVVXeexasU8jhYP8uBPb+WVF1/Ej2+5gxdtPA2xq5eB2a2olixRb5m2eWkOHd5Ld/csBg6P8Iqu1Wx96CH++dOf5Zt3/Ign+h7n7FWn8+2f3MzE4EE2rj6P4ajIfR//6klnu/1t0NPdbf/8Pe/DmCT8V2uNnwrAQBiGjAwfo1Kp4Po+vuclznnXAW3I5/N0dHQgpaJWqzI2NkZfXx/phhyVQonYJJ9VTqVSLJy7CNdTRFpzdHCILY88Qn6iiHAksVZ0d3bSnPPIZXN4KY+Dh3qZHBtm3dp1FEtFHnhoC6tWrWHhgh6K5YjZHS0IYZEiMSxIJXEdB2OSfJ5jx4499W0bIQSOq2huakEKwUc++g/Tisap1ma78KIXMyfTyY57HmPtGS9iweJ5HBsc4pZDm1jSupx5LQ3occ2+2n4yopGWVJqO5m7OX7qQf92xicM/e5TTggVsmezj5ae/iPGFzUwcOcTKlgWcu/xMbt65nT6vyNrGxXSJACoFKsWY3Yf30dHYyc7oAJNDQ/QVJml00rz89LP4/pHHeJXTysODe1AjHotPWcPu6l6CoyPMv/RSDv/sCbKjPq++/Cy+s+km8oNHWSzmkFrSwpOPbmXduedRfizP6hev5qWnreBTk/u41Mlxz+HDfPs9Hz7paDztTGRj4yMMnHcGG06/iNHho3Q2tzLZPY+16RJfvPYmCrrG4vZObr/xNv7yQx9k73iB1PgI1ZExpCfYOz7CnZ++GrP+FF6+5iKu++l1eKZEd9d8gsDjvI3r2H/0KAsWdDKcn6BcGuaMOS/mieI+mqs13vGaN9Mvy9zznTvZuGIjYVzl7ke30NzUyKV/9MeYss+8dUM8umMb5yxeyfj+/dhJy5HHjrB21jnk3Abu+cEPOPviS7l502Z62tp4/fv/iK/8y5eoPjDJ1l37GX9kL+e//qWkOzoYeXz0RHf5847jDnkhJLr+meRqqUy1ViMMQzLZLH7g09jUhDWWUqEEMvnU8tDQMfbv38/4ZJ7mXAOnrFxBZ2cXs9rbGRgaBAS79+xh/tp5jE9OUIvKKOVQLpYIqxFz5s5nspCnp7ub/v5DZLLz6ZuIaKrC/iOHqBbLRGY7rmPpmj2b7Y8/Rq1WZenSxXiOg7WKwWP9dHV2US6XmQxDHMehubmZjo6Op7a1KZVK5CfLxLGhrXX6mUFNVCM9WuPs9QvoXthJY8Wj49TZVPdUWCtXc+TAYXzXoXXxbDJHCuw+tJO5c5dy8MB2Rlpj8sMjvPsP303/8FHe5J/PRL6K3dnPSN8RKis66a/k8SYDNmyYzejjg+xpzXB291J2Pf4Yux98lCPd6zjtlDW0nVKiUhOMTOzmrLWrODQ+wZ7eMZqciFe98XIKvcdgbyez3/ga9m/bwrIxRaExw6GfjfDJP/lbevcPcP+j91KcgHf9j39k6w03MnvZLAYHBvnbb9zHqgvPYrzJo6//6Inu8t8I006DCTqb7Bv//S+59dH7WT9/Fbd8+odc8sHfY7xWYfDoAMuaOyDl8cAtW1lwykp27NzBueevYnBwlJeuXMXGxS9m2+Et1GLN1299ED8q4I5VGG8ok/9pH6vedCHrlrycvt776e9sYJHJEOTamRzppWlJD+XxUQqTBZYvms/YEU0qyvMfX7mRD37mw4wd2Mf8WZ38+LYtrL/4dK7+h4/z+Y9exQc3XUOPlcxZPJd0pZVN267nwuUbObxtD7ZqOFwdZv7CZXzsIx/jyn/+33Qv7sSTEXGtyJLulXz6je876SSf3wbds7vtn1x5JdoYqtXqU05+3/epVCpMTkyQyWaxxqCNxQ28JHPeSSK3KpUKwyPD1Ko1co0NuI4iFaSZ3TWbMIrIpDMcPTpAqiHN8MAgBoEUgr2HeglrBoFLKu3QEATs6x+hrSmLEwj2794PImbe3G7Gx0ZpaW2jra2V2Z1d9B4ZYMG8bkKryaXSaCxpPyAMQ4rFItZaPM8jnU5jrUVrTaFYQMcxjuNy1Wf+77SisRv4dt6pq+np7mJ2+xz6Uoagb4R8LeRQ3yFmz1tAWCkzOZZH6yqLly6mHEbkB0dZ1jGXXX2HaZ7bDkWHanWEeXMWUi3mGSRkvLefUxYvY9/RQ2RSzXQ0d3B0YAeqpwcrBdWipjIxwtz2Nvr6j7FyTg8HhwZIt3cwt6WHA2MHsKOTjIwNc8F5F7Pr8EFmOR6zm9p4snqUwclRNvSsIkpJBgYLNEyUCRqy9NcmOGX5UiqDJaQPdz18D0sXLmJuVyd60vD9r//7SUfjacdghBDDwOET/RzPM+ZZa0/OLxb9Bpih8QsfMzQ+OTDtGMwMZjCDGczg+cG0iyKbwQxmMIMZPD+YYTAzmMEMZjCD5wQzDGYGM5jBDGbwnOA5YzBCiFYhxLb6b1AI0T/l+DnL/hNCvEoIsVsIsU8I8cEp9YuEEFvq9d8RQrj1+r8SQuwUQmwXQtwmhJjzK9rtE0I0PVfP/buAEMITQowJIf73M+rvFUKsE0I4QoiJX7PNS4UQy39J/Qx9n2c8n/Stn5uh8fOM55vGzzWeMwZjrR211q6z1q4DvgD82/Fja20IIBL8zp6hPuA+C7wcWAm8WQixtH76k8AnrLWLgTLw1nr9w8B6a+1a4AbgX35Xz3MC8EpgB3D577DNS4FfGJzW2lHg9Bn6Pq943uhbx+TMHH7e8bzSWAjx3OZCWmuf8x/wv4AP1MuLgSdIBuzPgHnAxJRrrwC+XC/PAn5AMoC2ABv/k/u8GPjxlOP/CXwQUMAIoH55anVkAAAgAElEQVTZdVOuPwO461e03Qc0PaNuI/BA/T3uA5bU6+8HVk257iGSyXIhsB3YBjwKZIAG4I768WPAq5/RT18BngRuBoL/5P2vIRlM9wBnTKm/F1hHklg7Afxb/X63Aa31a/4E2Fp/vmuBVL2fxoCD9WeeX2/rH4G7gT8HFtTPH62396L6c98GVEgmS1S/x2pgPzBaf9czZuh7UtB3MzAE7AF66s/eW//fSr0/Jur03Qocqp9byMwcPplo/Fi9vZ4pa/ET9fY2T5nDW+ttPQYsfNb3ebaTv6sfv8hgzPHOO95hU66dymC+d3xA1jvmiXp5A/CFX3KfK6bWA28DPgV0Arum1C8Atv2S//8C8KFfY3A2ThnwrwS+Vy9fCfyfenkF8FC9fDOwoV7OkkwaF8jV6zqAvVP6KQJW149/AFzxLH2cIVnkA+BPgat+xeC0wOX1+o8Cn6qXW6dc/y/Au+vlbwG/94y2PjPl+GbgOuADwB/Xj019QM+q33Oyfu2/A58Gvgz4JJNghr7//en7RpI5fC3w/fqzW+AVx9smWfT+nUTyvgL4av1ZZ+bwSULjevmPge/XyzuBWfVy05Q5fPzePv8JwzxRW8Xst9Zu/S9c91JgmXj6m+PNQoiUtfYhEonimfhlu/7ZZ6l/+h+FeAsJd37ff+G5jqMJ+KYQYtEz6r8LbBNCfAh4O/C1ev19wKeEEN8BrrPWFoUQCvi4EOJFJAvzHCHE8c9Q7rPWPl4vP0IyQX8VXgvcZq2tCiGuBR4WQnzAWmuecV1MslBAMvC+Uy+vEUJ8tP5OOeCmZ7nXd6eUNwCfqZe/CfwTiZZyV/34uinX3k8iOQ0Cc0ikwSUz9AX+e9P31SSaxCPA++v1eeAv6v6OH9Xr7gc+QiLZuvVnnZnDJw+NIZmzx/0/95H0zbUkzBHqNBZCzAN+YK3d9yz3OGEMpjSlbPj5wRNMKQvgTFu39/4X0EeycB1HD4lEcAxoE0Ioa62eUp/cRIhXAn8FnPdr3AuSxfJWa+3nhRCLgVsArLUlIcSdJAPmMhLJA2vtx4QQNwAXA1uFEOcD55FIUeuttbEQom9KH0z9HKXm2en1emCDEOJQ/bgDOBe48xnX2V9x/E3gImvtE0KId5CYDn4VSs9y7vj5d/L0wM0JIZqttVcLIbqA3ydRxV1m6HscJxN9IdFQ/56Evo+SJG1fLYR4APgH4CIhxLnMzOGTmcZT5/B2IcSaKTS+GLhNCPEWa+3dv6qBEx6mXOfO40KIJXVn4e9POX078D+OHwgh1v0nzT0IrBBCzBNC+MDrgBvqA/KeKW2/hbrUJYQ4Hfgc8Fpr7a/7VZ9GoL9efuszzn2ZxFl5v7V2sn6vRdbax6y1/0xi811Wb+NYfWC+DOj+NZ8BIUQzyUDosdbOt9bOJ5HiXv9LLndJbLwAbyBRlyFRzwfrTtY3TLm+QCIN/So8SGKbBngTiZ0dEtvsgySSbwycVZcSzyGxR/8Y2McMff9T/Deg7+vq5fUkdnsAbwp9x4GiEOJC4ACJ6egIsIaZOfxfwn8jGr+Jp2m88Bk07hZCLLTW7rPWfppkDq951hd7NvvZ7+rHL/pgtj3j/OUkZpXNJAPluA+mncTm+xiJs/hz9lnst/VzryFxRu4H/npK/WIS59Q+EvXQq9ffSWKy2Vb//bBer4CHp/x/H8lA7Kv/PkHi0N5Dokp+jEQdnvos+4CXTjn+dxKn2WMkqq1HIqU8ROIE/SI/70jdNuV/PwR85Fe88zuAbz2jrp1E6nP5RQfhP5FInbfztIPwPSSLw50kk+o4Dc4lscVOdRCum3Kfhfyik38bSTTP4/X3valOjwMkDsdR4CfAohn6nhT03cwvOvknp9D3KpI5PEoiGQ+TBAE0MTOHTyYaP9PJP3UOX1Wv+wiJgLiNZA43/bL3Of6b2YvsOULdNn0bcIqd6eQXHGbo+8LHDI1/e5xwE9kLEUKIt5E4w/5mZmC+8DBD3xc+Zmj8u8GMBnOSQQjxBX7ReXeVtfabJ+J5ZvC7xQx9X/iYTjSeYTAzmMEMZjCD5wQzJrIZzGAGM5jBc4ITlQdzwpD1hW1JTwkkn5LCJUS9KBLOKwB7/NzURizJBfbpa4QQIARC+gjpIqQPUoKQYE0SrWEtYJI/1mBNjLUaa0MwMRaBlArq+wiKeoqQNRGYGKPBCoGxFgtYK7DWojVI9fS72ONPXL/uWIERe5J9Ce+3QWMuYztaGjFGAyCsnUJj+fRPPV22JPQ5vq1WQq8kv81Sp2/9r7UWa47nvj1dFkKClMmYsAZjDFIqrDVoHWN1jFQOUjr1+5EMOiwi+a96OwrpOCBAIIjCmNHJEsrzkUpSKhYJy0U0hmUL5mGtZef+Q9OKxulUyjY2NgL1aVWfjAKJkEm/CfH073hfMqWfj8M+3cjPQQiR0Omp/0nmFD+3JoinxsYvtlr/P/uL9/z5x7BTnuNXv3NvX99JR+Npx2BaM/DXL1fY2CJEQntbX7jBojBIKfFcUMriCgcjDVIYhBVIKbBCI+ocRioBfoDbsALptUDQgeu3IFINSDeHUpa4VkNHReLaKLpSxtYm0GEeXTpGVBvCxCC8FE4wGydoR/otyFoeW+snrh2kVpRU8orYtnDldw7wj3/QwiXv/Qb3fvsT9A5u5/fe/i06li3hxi99me13fZV/vHEnn/3TpcRWEGnBZzebafVp2e6uTr72ib8liqqAIK5V0LFGuQ5ekMJxj+fAWYRQxHEEcY0ojvH9NMpxMTrE4mCFRcchSghsrNHCYrVGGFNnIhpjYrACP5VBOgFSSsJKkWq1QCqdw1pJHFWo1UqI2JBtnoVULsbECEfhemmM0URhFaEjhEqRymXx3IB/+X8/onXeEryGZvb87AFuv/4a0lmXm77yUVyvBY3FmJjz3/TeaUXjxsYG/ugNV1DnHCipEFIghEEJF4SL44DjKpR0kEqgpMJx3ITRyGThl1IikBgbYy3J8XEpE5BKYKVAaJEIkj/HBRJpTkhQSoEVxDoijmOstUiRtC0diZR13lQXUBAWaw3WCKwxdSZlMUYka4oBg8FqQ6w11hj+4kMfOuloPO0YDIBjDVaAlC5CaLQSWG0wGBzhYWyMMaAcS0yE1BLhC4QFKS3GguMopLKo1DxEaiGyoRvpNyGDRpTjYISDEgpsFWtqRMUJ4loeW8ljo2PEtTGiqBdrwUk1g+pG+c0ovx3HARMViHQfuqwwMcQBXHjl5/k/r2vhlHPeSba5m7WvfA/i0dtomNPO4O4t7Nr2Zc649B184sqViCzImn1WieiFDOkHZFIpkIo4jJCAFAIV+CjpoE2iTQgsxhjCcomoMEZ5cgQpHCId4gQZXC+F8jxMWCWKK4SVMsJoorAC1qKkAuUjlSIKa0htkVKh4xpoQ1iropSL66WRroONNVFUQ1fLGBPjZxpwXItyHayRFPMlUmnJl79+J8d0itWnvZh7N93Eljt+xIoNi9n03c9RqRVx3BROKoU0JmFw0wyJgploE650cF2FcjykUEDCKCwGow0mCjFYhBBIIXFdD9dTieZqLVIalHIR0iBQSZ3jJEIlCq011aiM1Rop3brWCULUE/NjS4hBCoW2MTo2CASx1YDEhjFCKASgpAvK4kgXYxMh1xy3aKCxGqIoJo4tmBCtE4FXyV+2U85/f0xLBqMUoCRKxmDBkRajLDJOBohSEkRitpBSIqVGaIVQFuGA585F+Y1Y5UK6G5nqQGbaEY4PeJg4xlAhMgVEFGNqeUxlGF3oR4fjaFNGx4NYI3HcNmSqG+W1obwAx0+j4go6LkBssAaUlLhE3PHFK8j4loMPf5H9D36RmjGo9Cl862OvII4F2Szs3/pldFWCEASOgyY60d19QiClgw5rxCbE8T2ElHUTRGJ4BInnByCSBURKmUigQRVtDC6SVCqDE/iAJKpW8OI0fqYJYTQmjtA6RHkpBAJrNUK6aKMhjvH8NNJxsdZi4giBwfVz4EmsCRFhDWNdlBsglcJxPYJMhls272bfsb20LFhDSynm/37knfzzVR/iY3/6GqTvYrQl8DyUUhhjMHENHT9zq6oXPqSUBIGHtQIhLNoabKzxHIlyJVaCEh6J5TiRsmKjsUagTYzQFuV4yfwXCsdRGANhVAUrEXGMlAJsRBTHRFENV/qkM35d01EYoxHYxJGtBMIaDA7WWKwRSGkJtUHiE2tDbCKsMZhQE1EDbRNztxWJOVw6YI+vSQLPD7BKoSRgZhjMSQNjBK4yWARKJtKGkBKhDBIJMvGTKCGRCKwROL5A+Ao3ux6VnoVx0ijrg59G+o3YTDvWBJiwhA2L2No4No4wcRkTl4nzh6gVdxBHMdKRySDyc8igHRm04bjtOIGPoIaOSxCF2JrGIolrTSx+2Tv42Q9/yDu+eSdfe9erWf/md7H1R5t42bveRGmsQLalje33PoTjVXjkvs189Ju38r+uOA1T2cu0U2OsxYY1arUyUro4Tl2bxKJ1TFwqElarRNksQSaHFAJdC8EaHD9Fyk+hHBfluUjhYIwBR6HIgRTYKFlwrNY4fgprYkykcT0v8Y0JizAQR1VqlRJhuUBYLSOUh59pSBhNwpYwGryUx94n+7h+yw6Un8Np7GHz9d/m6MFdbLr1WirFPLUwxLEGYyyu7yNdDxNHxIn4e2L7+0RAQCbdgBEx1iQ+LwsYY7ARKEdilUU6qq5wWFwUQqp6A8d9KgYrBVobarUaOg7xXRfHTfygWsf4riCTagAkYRTV66t1Ch7XjBL/TGwS31uyA9ZxH1BiTk/MayYx46Gwom4qM7qukcVICwZLbCQmihBxjBCgT1IldfoxGAuR4Sn1WCiJlAZpQOBhhQZhEv+8sCjHRXoxnjMH0bIcmenGSTUmtlPhsFusZnC0Sm0gZMGiuURRkVj00NYhaNz7HWw4iC4XCCu7iGOJlA7CxDiZFqS/BNdvwsl0oRwHqyuYWglbGITaAYxxMTrCuBMcueOTpHKCq9+9kGx6Iw9c+ydQs1z78Z8QiwoyNigcCiZGa8kHL8sR4DNdw9CN0UjlJFqKFVSrZbCasFJiYmSAuFrBS2XJtHTguh7l0gQ2CgkaWtBRjHQDnMjFT6ewJjGPukEKIQRaaawQxFojpMRxM1gnIqxW0CbGxDoxnZEIMyBxXA/H8xBCYrTGT6XBGsaGJ7nqmptwMw7LzjiXbfffw+abvsGW234IsaZSLROVihgdEulkUbQmQyAUQZBCIQnD6gnt6xMBayz5Yj4xdzkuUrqJH0QkvpWnPoFmSZwkGKQj8Z0UwpEIazHHgzisJapVAIvnBgglpmx1kuzGG0YxJo6J4phMpgHlOIlJrM5QEiYmcQ0gE/+KFA5CktynPg2NMaA1cRSjbZyYbaWDwWK1TXx8NsaaGjZK/EtSPB1IcLJh2jEYi0BrS1VbfM+hFhmEI5AYfBHV7bBJhJEUMY6nUcE6ZPM8nMYFyFQz0sA/XbOdOatOpbNTUDU+zbMaae9sYKBP0NPi4rgumQv/jtGHbsMM/g2hFEhrEU6Mnz4VEWRQTgbhNeOkG7CRxkZjyPIQleo2ojLYWKEtGG2JhUUYgTFVKtXNOLEgFpDzysSRJpSgTUhGCqyIiQ0YwkSymoYSrnIUyglwvYCoWqZWLSHimHJpnDCqoRyF63nElTxx6JHKZBO6C0FcKyN1iCRLNY4Sd6+q28FF3RfneMi4DMZgFGitwRiEtUnQgGPxvQAbBHiBn0SFGUMYhjhSMjkwzhduuJcx47PmjNPZsX0XX/77d3Drt79I8YrTOXZkL3GkQQqkSpzFjpdEJjpequ4ktkgvQGp9orv7eYdSimw2izYxcRQCBiGdRFtVCoxEusfDQpPoQGsE1bCCi4/nKkxoCeMKDgKERGCw0iS+WcdLeJWJweokoFA5OEKhjcVzJXFkiWNNbDXSCqQ8HvVpMEIgbQRYpLBoLUAKlJB4nsJzPeJYE0UVojBKYhWERWqLVS4SiVUOEpswwxPZ2b8Fph2DMdYS2kS8iUODK0FqgcWlLGukpIMQGheLchQiSGNyrfgNc1HpVrTx+crdx5i99iwyjbMQGYdGN40X+BwbKGBNmUzDPCwWJS1zXvRq9i86Ff2jV+Fm0gh/ETbdiOu3It0MTtAATgply0BEFI4RVwRWJ2ZXbWUiJVmLsTqRgAFPgKvAaI10FUprIiOIDWgLjq2H3qpn7Y4XJLSOqVWqBBkHazSO55FSTUhjSTW10opF100P1hg8P/GzaBOjayHCrTuAoxArBLWwmoSDp9IYY5HKwQ18QCT+G1vFGpP4UpwgMZsJcL0AUauiowgpJNWwRlyt4gcB/3rDdlat24jYf5CvfPKvueSKS/jelz5FMdaoII0rFNoWqZWLKOngpwI8LwBH4QUeVhtq5TJCSOL419md/oUDx3FwcLAOWKEx2mK0RYsocfJrB6nqtCQGqxDSIgUYJFIZfJkCQCmJtgKJRTouxhpMFGOsftoKoMBzfaRQScqAjUBYpAWkwVhRj0hLTOymnr4QGlDC1iOTTZJWIATSVQROBtfRGAyxrlEtV5FxhFQurqNQjsAYg66dnL7UacdgLKCNQIokfyGyIK3EVSGOFRgLrlAImajUbvpMTG4RMt2K63n88/cP09DQSCbdinRdujraQTo0pF2aMh5tjWkmSyG+srS1NrD/UD9Le1oYuvwOKps+jJdrw0k34PktCKWSwAATEccldHkcXdubhDIKg9FJqKO1BmHBkRJrk1wZbTWm/iWdlLDEKkm7UbEkNhrtCGxsp6HuQuKD0ZpqqYgtGPxMFjdIYW2MdByUFyAl6FotMVkIi5ASqZMooziKiGshVkn8dIbAcRM7uxKE1RqO4yClRBuQrofQMVZHaJssQK5SdVNZTCk/SRxFCCGIwgquknzqW/eyZNlKHrr9BnY/ei8Pb91MaWSYcqWKEqA8H2sMOdNBFFaIajVqYQ2jJFZrolqItYaoWktSLE5S88lvAzHVbFSP/hLC4ngSjUHa4z6RCFcm0WXKcfC9ACMMEgtKYU1iQQujCKN14pOp1Z5uWxiwSQCIRRDFMZYQKQUClTCyp8zQSeSaFAorkugvqcCXiakMK7HGEsc1wria+G9s4j+q1RIhxnMclKMI/KAexBETh1WMPTm11GnHYKBupsUiZRJBohyLJxXStXiOQvkxjtuACHowfhprQobHLN944CDNHV3MXrCUfLGCIyyjhQpdjR6TpQhlI3IZHx3VQCpMXKVrVo7WXJrAVeza8FcEfT9NcmSswUYl4uIQ0kaoaBxb3YWNEtNHEshmEBJcZbFWYpEIqbGxwDgWY1USFYNBWYFvBbFjiTVoa8CVVKehE1hKhXJdjEnyW8oTFaQTEMe1JMIr04ifzSQhq56HReAGAZ6fJoqqmDhG1PMahBA4bhINZmONn0nMGY7ykI6D47hP5T3oOEJKRRxWiIpFrEkc/9nGZjKpLENHJvjB7iFSzbO46St/zi2bH6A0Okp+dBzl+HgpgQ5jpOshVRIFp40lcFw838fxfLTWaGOwOsZNCaIwRKppqKaSmMmUI/CcVKJp1JnFcTzNgASudLEkvhApE2ZkrEBIi7aWlO9DPUcliiLCOK6byARSTUmWFBBHMVEYohwH13WQUtXbFNTDhBDCwXEFSrpYI4h1jUqlgEWDdfFcHwBrNZVKwtCklEQ2yYEpV6sIHBzXxc/4+PbkFCKmHYMRgO9YlCNQgO+B6wuEq3CUi1EKpeYgvACrUkQx2HCc64dXkOuYTybbQKkaU5gcJx10IuOYvoESuYYM2VnNOK5gVnuOJLLQIS5NcmSyQGtLExeedxYHdjcx8cSN2CjGVAvo4hBhdQhqewhLJhnkRlCPOkBYgVIWIQxCSLQRWGVxlMBGYGQMVmIwGKGQscVNdHZ0bKflZkBaR9TKRfyUTyRVkjRhNGgNVhOHRURRI4DYDVDSAZvkTAhjMVLiIjBGE8cxulZJIpR04pS1QqBFhPICwkqVWEf4foByHMqFItX8GIXRAZRSeOksjutwx08fo9CxnOFDh7nz+qv48Q+v5+jOx+thtpDK5qhVK+gwQiSrIFGlRLVcQiqF1QbfAG4SuGCsxsYaJ0jhKO8E9/jzDyHEU/4WQ8TxhEtr6n4XaxMrgLFYNJoYa5LIUYRB1ZmCdJLcFG0SgdMIg+M4icYZJ5qnNjbxiWCwpu7PsVApl6gIhee5pIIU0lE40sHz3afC3sESC017excTkyMIQBuLiS1aa8IownH9ukYEvrBI4SQJ3TaJbou15mR1wkw7BqMUNDQItNX4vocUBsfzQLZjtYdwNEakEqlWuNSqJfYdm0tLdxPVWBAbQzgxgo1jGhpzSKFJ5xrp6sgQhVUyfhPVaolM2qdai3E9l86uTgJfUcxPML+7jUceHMfEeSiOEubvx4QWE5nEtCUdpAGpPbARSFvfbiSZVEqIJFxeSLRjEmlJCuIwCbl2lMCgEEqA0jgnafz8b4X6ljxxrUYcx0ggFgIhE5+JDUMUiW27VsoTRxon8HGUADeDE2QgriWRaEIRVvJUa5Uke186uH6A5/go169n4NfwM1nSuYbEJyYEfiabOP4F3HD7dgqZDu7/yr9w7NATxPhMHu2lUs1TrVRw3RQ6CtFxhEBStjp5tkoJE+tkSyEMtWIR6bqkG5pwA79uJnLriV3TDRZdz5rHGIRS9Sx8gYljLBZrwFGJadN1JKCQUqAN6DgEY4iqEVWjUEoQpDL1BE2LUAoHEi34uPdEx4nQJpIwaAiIdUxYN7Vm0hlix+JaN0nI1pLff/0V9YhCi7aJVeLeO+5gaHAo2dJGBIBNdhIxFoup7xKS+FfjOEzG0UlqhJh+DEZK3LTAt13guomjXfgYFSG0ITIpDAVElKKmx4itIb/sLZjJCo71KORHkcLBqoDuthTbd/eTy5RZ3NNENufjumAjS61SpSGbZaw6ieNItmx9hAvOPYf86Bim0I+OxohL27E1QxQCSlGt6STmHYMmRh7XXozGFR7KaqzQBG4SdukFAoxPbGp4gcJKksVVS5AaIwXEJ+nI/C2gHBc/3ZBoECYiDmsIx8Uag+MFiR8jDtFxkisTlwswbvBSQZJVrVIEQYByA9xUjjiuEVeKYGOqlTJSKhwnRaqhFT+VwXEEJo6olctIpZBKkW1qR7iKz37jLloXreb+a/6OoaFxMpkOrv38JykVxxFC4UiBUw+bdTw/ESZItj7J5JooFSZAJAKGsQbpiiQsWViUH+D5buJHmmZIkt9NIhwogSMljkqWZ0f5KJkIFKAxRqKUTBzrSuAasF7i3I9DhzAMqYYhUjpJxB8qCRM29inrwPFthRD17HoFXtrDjZM8KW1jKtUirhNgNSgX3vC2dybPaAzIJKcO4NwLL+SLX/gsnkoR68Q85io3UXgMaBuhdZy0ezw0/SRNN5h2DAaVwk0vQykHKVxiLTDCYK3GCIOggo4NUtWQMsN9k6/l8YNbWbFqPQN9/eioQjbTQMr3ufXWe8lmG3DEbMI4xtGKsbEJHGIsimwg8JpbsHGVszecRlgp46YVurALGw4gMIm6bKFaibBIohhiY4mPJ3hKieMIqlrjuhZHKAQax5WJQ1jFuCmfWHQiHYXRAssYMiwSOwZRAzhJs7R+Qwgp8bNZpFDUSnlwPVLZJpTj8P/Ze9Noy86y3vf3NrNb3W5q76pdfZ+OJIQkEENoD6BgQATlMLzqafTqBXtRj8cORdGjHMReUVAOjeLFBukhBAgkQEISklQqlbaqUn3Vrt2sfs053+5+eBc59/NhjGTUKJ9PGRm7xtjjXWvP553P8////rYqp1BKgSkrRHB4W2OcRfr4uaks+odUmqHSnLzdprMh7ldcXTEcD8DUSJ2Rt1okU++Ec3E3gndkeZN3vOfT7Lr6Jj71nl9kvLZCbwy3fuoDrK+excuEtNFBJlM3edYgazYghPjWIuNMPikaYF1cSONRWqMQUy9XVMMFe2EugL+dkkKQpholk/i3MOWBee/5FuQ1WDfFxRJZY0JAUARvMV5QmxJnwXtL3shJtP6W/TVOVH1ERhFA6wRkwDiJ1vENQ6ORU1EBiKmcuMQYw40vfGkcsfnYIGRC/BnvsULwY2/6Cd7/3vfiPVhTU/oa62zcCwWLD+CDi03pAh5CXHQNRgiFSGbiH7Kcvsp6hxQjnDEgmqhsFp1ohMo4u1JTdftMBiO6Z48RTI1vtjlz5gxFu0XrWc+j09asrw3YutBi03wH4UFOSQF1MLSTBolWjIYlrWZKNTqL1ga8jOowEcg01BbEt7T0VhCCpw5gahfRL7XAawhBkyIQOHwKWiagOpAvgC6QtkboMYnvEpIzwNozfexPa3nnqAZdkqyIjmoXl/BCSOraIAVkRZNGZwYlwDqHsxZbVxGGqBTO1kiVkDVbkVrt6jilaLdJWi18bfE+7mSMqTCDCVIpkrygaDb4vffcwdLu53D7h36ZoBJ6NvCZz3+e9dUzGGdpzc5Fz42pwVmSPCNYS12XEcqosqeozZH8ML3JOk9pDbauIykC4pjoYqsAiUpQIr691cYRjJnu0xzOg3EGNeUNel9PdyLRVe+8QUiFEhKlNLYOZIkC4XE2IJCEoFAynruYPuXTNCPqyZj+TJhy7iRCgRQdUI6tu3YzHsexau0swgeKoiBPU6SUTMqaq665hnvvvJva1FjjokzeVfGSGGfi4InIGnFhLlMvugaDEKhsPoLrRFyqK+HxrkmaSIKUoFKQgZ7ZymgwYn5pE8snDzPqr3L+xGE6cxspJxM2zM2zZ9sMl+6cxbp4owzeTA1firKqWJyfZzKZREVSiGgQXGQWKR0iNwlHpiTeBoIispB0dOR7wAeJVAEfJLWLahlHiIayWuGkI2mOEGEO6QUiyQkiR1iFEAUXW4Oxdcm544+jsiZKShrNNkIN44IWmPTWGayeJclQV38AACAASURBVCBJ0mTaVDRCRpNcBB9Gx/2o10NKibUGreP+Rel06r4W4ByummDMhDxpo1sZf/Wxh9h82W7u/8SfUjODMas02gvo8TqD3joy0QiZxhGOD1H+WlUM184x6J5D6wa60UYET9GeI80zbG0gBOp6QjkcxH8XpuM/efH9GTvv6A2H0Qk/ddULmca/Z5WhAB0c1lnAInyMzgg+StITnYHwJFJHz4uH0ag/BZPGzyaVKd6biHEJgSA9iZR8SzkjptEMBIFQnjRJkSLhpTe/irIqWV1d4777H+Jj//JPDHtdynrC6tlz/O37PsD2XdvZs/9S7rv7XsR0n+SCxdqoLCPISBQREeEs5b83mAuihJAkWYGzJYQa4RMcFUEGrKtJdYbxNTIoPnQw0Dt1mi2X7mBtfR1rPeVwzLVXb+axI8c4cfwxzrzvfo4feYTCW278rtfy5p/8cZ71rP0sL3fJU8lat0sjz6NSKM/JixQTMhJXTvNCiAy0IEgywDmCmzqFbTRphSkBwyvQAaZOMco6IHVNw2hceRpnBTpdANnC+kAiBDUXn8IIAsEMGA7Oo2SCnTSQawXNDUs0WnMRYFlDr7uCG1TgPVKnFEkKKt4w8+YcRavA1DVapzhncXWFQJDmBTJNUSrKl9OiQbPZYjQK/OlHD6KaObe9/xdp0eDU2TVGNuVzn/kouBJdNDDW4IPD1Y563I/8KR8VbmmS4O0EP478Mz9J6E3GFFLjfB1HcETybmkq3KiHJXmmD/xpLykVzWYbHwJKSqSICU5hakgGMcXlx5e/NM3ItEIIjfEGKfXU4xJHYmrKDfPBIYOj9tF7howKMoSaRntEVZcQ/ikFIMh4KQkeISrKSRSXrK2tsbZylquuuBRnHV/48ldYXNrIW37xLfz6r/0ma5M87vq0ptOJ/piyrrC1n0aIOIQCLXOUujDnZBddgwne4ash3hu8d+gwnP5/E19nTY8gHaYMVF1JY65N9/Rxlrbv58abbuDlz/sZtu7Yzic++H6sr/jkpz9Pe9fVLLUFd33lcyid8JKXv4JtWzZx5WU7yZIEJaPP4h2/9mvI1ROkK4qt2wNSarR0SKVwIaCEQCUSFeIfi1QRjAcggiRoH6MDpnA94UEFR11bVGVR+XGqyZOgFFLtoFaCINvP3GE/QyWVptXaSJ6bKC0lELylKrskUiPSjKI9S5o38b6Kt1Njqaoxk3EfX1fUdYkxs5FlpjXBGaTO8MFjTUUqwJp469SkfOGLj/Plk2fZtn0H93zyT/Ajxen+KkEoLrnmucy2CqBJ3mwTrAEhMfUY32xFMrO16DQlTLEvUid4ZxkO+ojJECt1NAkHN40JEBRpE9vwmAt0Afzt1reyv6yrwau4R/URga+UQieCLNUEoRFBUDmLVhKlBFqAsQ6daAgq8sKkxE33X8GAD4IsSSMpW0RuslSCNMRdj5ISZGz3GkA6Wo1ZalNT1xV1VbN5yyLVyNFqa2687lpOLy9zfvk8D919iOZ1r+Oq63fx+Nc/hseClLSyApcS37aCn2L8Jfy70fLCqOANk8FJlJB4ez76CainOQyrSHwkmTpYG22kLVIkgkzWXL5nC1nRoqprXvmGN7A43+FH3/QTXPWiH+bk2VOUaolhOebRQwe54vKd1JWl0cjRWnPrv32Uyfg8xw6f4PpLrsP425EejJAob0lVZCFJIdHa4ZEoD9IIgvSgwBnQOkxv3AIVbcDIoAnSYScGlWik9kh3BCUlTi0900f+tFcIgcm4C0LiXI1OCoSS+ImlX56Njn6VUXTmaMxsQAmJc46is4G2LSOMcPqgct5jqipSdYsWmZTUpmJYjRG25t7Hz/Pg+YxKSjbNbuKW//XbbFraSq97BOMEb3jTj/Cj3/1yysmENM1QUqEaGc4YlGhRiRLvLEpFXwZSEbFX0YuR5wWJUkilCS7F+xamGsVlsa3RcoZOa+aZPvKnvbz3jAZrSJ1GQ6NIyVKB83GUpITAW0GQAZUopBAUKsMFT6Ki8ktohfUePc2QUUIQpCIQaOQ5zntsMGj09E0lIvqtizBL6xzI6FVDKxJpuf4FN9GfjBkMhhw5fBTrPKtrxxiZWeYXtzGpa3bvu4wz3RM0z5RsygJJIRDkOFcRfIRfmrpmXBsUCYkGxIX5qL4wf+tvo5yZMFy+D2cdBB1DfgIIJamrqOhCOEa+g6ssuu05t7rK3kv2c35tSFH08QK2LM2zNqiYa+V86C9/mTe96ad4/StfzOHlMSeOHePu+w7x0hdczwYVo3t37LuUEF5P1vgmjc07WbnvTpKkRoWoiNEqkEpIE0EiJcJLssQhE7BOxLwJMW02PkYkCxEBjFJJgnQkUuBEQKiURDSnCZyTZ/rIn/ZSOmV+6350luNdhCEKlZLmDYK1VKN1qrqmrkaIAWiVUZUD1HQ0JtOMNMTdiJYCOTVxHzsx4CO3HSTNEnSSMVxbJlvazZ5LLqd3+hi3ffIPCcbyyIMPs+vK63jfH70da0qsMdSjPuP1KgZZ6YRhdxWlErIsKtWSZoG1Ju4GAWcM1nuUzlFZjqmqqcxaIZMO1lrUZMykXsf015/pI3/aS0pBkTfRWhKERMmpM18EtEhBOpwQKClj4JxKqKyNSj9ZIpEokYKAGoPUMb1U6RSPwHoHQkX6ugMcCBkvdEpJ8GCcJ7gIszTW4kxgNB7ircfUhizP8OWQhaXtbN24iXE5YfP2RfK0ybsf28UrVMXZ9RUuvfb1nPrGJ0hUDiKKjpTS5Dr+HqWd+m8uwLroGoz3MBkLXBAI6RBTOCQh4KK4kSAF58I+1s6fQSUpOzcvYU2Nw1LWJc7PYmrLymqXZmMzl195Fb/zh38GteTmbZsQ1rBn73YmvXUmwxHNomB9bY3z/SGv/oH/i6OHT6OqX2f94d8k4NACEhEnwb52iESR6ShLBU+iFQgHmSQ4gVACmcQcG50A6QxSLCD1PEp3puiQMc5M8LYGVp7JI3/aS0iJzlvoNEHg0b4gy1sIHeNzdZ6ROUeSpEgl0UoTwkaAmA1Sm7i8Dy7O95Wiuz7hk3dHVJAXYLzishc9n1NHH+Ker9zKmQc+R9Fs8OST56is4/3veiumnuBC/E455zB1TaE1IoSoXqsqtFIgDSkFWqeINI5HnXdI5xDBIZSm0exgq/FUMRbQPiAaMzRn53D+wnz4fHsV9x9VXaF0SnAST0yQdCLic6SQGOuYlCVV1UXLiNzBRwmyUYEizaIazDmsJ6Lyw7egmS6GggF6mq8uhcI6P138x3G1Cw6tNAiBC4HS1AgpmJmdwa8HhoOS+x56lEbmWNqyi9Fohfn8KsbD81y6czv9lR6XXPk8jtz/ZQZVTe0NrdYcIvNIk6GDJk0vTDPtRddgAIyLIUVSyel8PhC0iKhsqQDPN9aWaLfWKKsR1nWwkxJXBQYjSzmpeOLJPq1mQX8wJssTrrvqUrTSOKHp9VZRwdJsFDGZLs04eO9dXP3cF7C4cxtWCGweOHVvoFkwzZ+QWBdvYErFzHflBSpTEGJmtxaSkLtpMLiAxpWg5kkbDULSAZ2gdY4LGuljgwmTwTN93M9ICSEwVT0de0R3t68qfFGQFA2KJNKQhbOYKYwyLQqESqhVhXcVuU7Jmi2OHHiCTx84xsz8PIm3lOWQqr+GMzuZXdrB3V+5hVZLUY/OsnMp5eO3fpXuqSO42tCYnUegooCk0USnGQKBSlLMZAxSopI0qp28RyiNTjMSqbHOUE7G1NUEEWIMc2SjpRExohN0mj7ly7iYSghI05QkJNTWUttolNSJmoovYtRBOS5RQtFszSKnyl8hp6BMGS9uQkgUOr49IJBIhA5ID16GCInxHjmlI3siSFPIqGYTEqyr0CrBlDUiCLwPaCXptDocfvwwWQhAxpmTJ2mQMj7zIdZ3v5wzTz7E9n17WRfz7H/eK7j39k+QEikNziQ4X0bvjrowzbQXX4MREXApkhCtBVISxP/+75ilLeg0c06dFuzeuY2ZuQ5XPPsynjx1gmvn2xjvMLVhIhSTomQ07DM722F1dYWFxQUSqfC1pWhmFHlCIiX9QcW9X3sEWcywact2jCxY+75/hVteCypBYKNxckpyDkIgJRA8Mg0kSRuVbkGomfjFlhrV2oFIOjghSZIMpCIISSIVpU3Ruo3UnWf4wJ/+klKSFw2MmWCraJYrbZ9Rd5Ukb9CYW4wLc2MRUjAZdFEiOrqFgCTLyfIGSZLw53/+DxiVs3Hv1Zw/cwTV6JDOLZJZwdn7bsF0dvIDb3wjD37985jhJu47dhLhKhqdGcx4EsdgeQ6yFbNjfPxsfZpj88a3fmG8tZTjIb6qcLXFecOk32Xt3En6y8cxtqZodGi15smaHXSqIW2BUMiLzEgL4H1gUpkpI87gHCjlAB+hpl6iEk2z2Y7kbOtx1kYztYjjMBks+ATUNGRQKJRIcbJCBRmNm8JPs2AEQYSnSM2e+MyIaEuHkilCgZAKb2oajYJESrqDAddcezWPHDmBGY9x9ZiRL2mOVzl54gRKzHNy+UGef+VejirNdS/7Ie767PtxQ4d3HpVkqJR/d/JfKCWARipwBIRUEWmvpxgOBMhA4iTrkwl5q2B1fZ26mnDgPsn+S/ZjnWR9vYsMhjTRDAZDvK8ZDPoURYvjJ0+zuDBPkimUlPS7A/78L/6W+dYih548y77VPu1Oj42b2jxb7eXzn9zFtuLYFGipkBISAVp5kkKQZpuQ+VbINyCTdgxMs2MECU4WSKmRKsPJMA0Xg9o5pEzjH8dFGAgTfKCqS1xl8MHiTY2txnhbImQLrTTCGJROSJMEyTwqieh25y15q8OjB5/kM984xFjPIgicOLtMO8u55MqrMcbyyGTEE/c9wmt+9NV89qMfZfN8Cy0NV24P/NZ//x/85x98I/MbMky1gm+2o2Of2Py0imZOoQqC85HELKLZ01sTA8umzSjPEvTmHYggSRJNEDGm11tPbbu4eswFC6r6dkpMwyqFQOvIFPQhMJkYvPdYb/EuqsqkEKgEGnkbIUDrOOIyXuAZk6Y5iY4KPstoSjaeGuhFHEnj4kNeKgkuEIQmeDM1Q0rQARkSHrj7Tp570wsYV4ZhXaOygg6CDZ0WvQDHl0+jXOCa3Uvc+Y0PsqZeydzuPfzt+/4Xr/m+7+Hwga+zefsuquWTIAPeVpjKYf+dpnxhlJQgdXzs+mAgUwg7xXAo8CJgBUhajEeriGAZq+ikH/Z7HLMWLeGqKy6hHNfYcsKe/fto5BntVoFUknI85vbPfYqrr76ccVXTX5tw1bP2cuLcKvfddze91ZNs2XMV23Zt46of/0eGH34BUjkSEUikQKWCVELSvgLZ3osq5nAyR+mEuqoQssCbimqyhhjZKMMkIEQatfoiIia8kKiLUKYciDN0oRTae1ySkeYZSbYZqbKpIVaS6IRcJzTSBuu9EQ8+8Ci33P0gtdM4IZjftodQxzCxa666hIX5eZ574/WcO7dOd/Uctyyf4fF776Qqe7iwwNZdl9A9dIi+Kfndv/wgP/DKF3Dpnhn6K2cIDoTW0T2eZmRpgUgkSmrSvEArTZKmeK3xxsagOCWReNw0+Cw4j1YpJnjscJVyNJxCMsfP9JE/7RVCYDgcxqwcW1NXFus9qY6plnIKNzXO42yFqSwurIL0dBozqESRqIwT6xPW149FVZ5zzM0vcMM1e5mYEblK0TqJURmpBOvwLkqWJQEno/IzZtHE0VyaK4bDESpNaDYKlHGIpMnsxo1MbMUNN9yEmPQpTcmrX/Uq6rLmwOED7Lr8Wk6cOsrZEyc5vnKC51zzXFYO3cXs7AJKJ1FCegHWRddgBJDq6f7FTrMelIVERs25D3ilGNYVjURhLCjj6HbXaTcb6CwjzVLW1laYne0wszCLqWuskvSHgK8JAV72mteS6YSv3fZlTg8Nx1bg1a99FX/61l/mHffdzd/87cfwIeFZl2zmk2cdm7cGtJAIGchygcwvIzR3oTo7IW3E13IZUF5hhSMEh6wH1FVNCPVT7nPrK+pqQlWOkNJiQ/ZMH/nTXiF4klSjZFyoJ5nils/cy3Ou3M2J1TXuPfAIle5wvj+kKJoYLOOqZMe2bTS27Gdrp4VUCXffey9Cp1S9M7zkhS/giUMHOXDbF/m1d/0mf/DWL/D2t72df/3055mfmWNutoPwgXarQ3c0YN+lu/nM1+/hyn0vo7++TDnq463D2RofPI2ZJZK8gZSCpOgws2GRPMsJSkOaE6QgqSo0cX9UO4d3FgHkKsUqhcoKjInImIutnHWcX12jrisgNhznLfW39jBoQnAY52KGjnF4VSFtyrBc5+z5EcP+kFAPmN+yg26/y7DX5cypszz40AO02g0WNm7hpVfuJmm3USEhiECSxuhylCBEJjc+ePQUcuqdwyEItcGFQGd+llG3y97t2wkh8Kwd23jgnrtRQeJEIMklV+1IsfVZDh4XbGgqvvzl25j0z7JtbiPnzp0lSRXtmQ3P8In/n9XF12AkyDzKHEUzRZMThCOICVp2sN4gnWSuNc/hc2fQmaS/ssrCwgKVsxw5fISrnnUFWqd0uwParTbj8RghQRtDVQ6ZnZ2l2+3T7hRsvWIX41DQ6LTpjRImcoa9Oy7ledfu5ZHDXc4u93nJHxzlyF/sRSYC1XDI9stQzTaivYjIWiA1wfsp9FCgRQsnMrSQkJQIQCqFsSMYB5ATZCKxpkdZls/0kT/tFZxjPOgxO7fIH/3xhzljFO2ZBT534iG2bNxE1yg2z82QeMUll1/Opz/2Ye78+q0s5pLaeQolGFeenhN0csW1N93Mpz71Sb7/Fc/nUGjyg9/5PVz/4pv4wq2fZ7jWZ2HLAlLAoL/GlqWNdPxeGhsWUCuneed7P86PvfFF5B6CLfEhR0pN1mggvKcq65h6OeqRFy2SZoeiM4NWKUqn1LIkbbZIpUTJBJQE56nrmmBnpmOgC9OE9+2U957xeMTUkkxQnlQmeCGw1mC8RSFQSkT8k5YR2a89D9x/jC2X7kTUYBsZh594kDxt41zF3quuprvaw9aGjVt3c8vdD3Pp/i1cvm8TshQQNGme4n2g3+tFE6YMCDRSjhHeIaVACRkl6NaxuLjA+dV1tizO0h2OmFQ9jHEEL1FC0mk1SNKUjYuaz331K+y/bD/1cMzG3Qtx9A0XbGrpRddgpCrIZy5BqAIlc5zrInyIAWMhR/sK6SvOLx9n3/69HDl6lKzIUbqg2+sz32oyLivuvfc+FjduYOOGFsvlELGq2b1rK0WjwWg0irG5Nmfz3BK/8tP/kTOnKsYTz9v+6I9oNwNr3fPs2rGB0+sS3ZbsfPMRznxgL7p1PXJuG6gGqAYm1AhjpwNnB0IiVYFIaoy3kQbsYp5EknXQKkdPEmxi8GaeLOkCB57pY39aSycZ7/nHOzixOiBttJjdso3te/YxqcfMNpt886uf4+P/+DeoIPnnAF45dJAsj+P5fu/rns9rb7yKNXIWtOcvP3Ock8ePIouXsnnnTta2LbJj524e/cZtnDm5zob5BrLVRPqayaiPbhX4coBXCeW4z1995A5+4T/fzHi0TngKy59gTY3Kp/Rm71g78wQoTXNmMzLNKfIC5xxVkqCm0bxqSlgWgHcxfTH/lljgYioBnVYrfvfxCBnpFiEIgk+onCfNEjKVsnXPPs4sr/HQ4SGnTj2BDTVveM2r+fM/+xN02mRudok9O3dxz70P8MA372RhYSvBjXj4nju45urruOfe+xn1Le3NC+ycLZDjkqRR0Gg1mAz7SCGZ6cwgpnugSEDXUemX59PRpmBhbgPDwQhBMqU2O6w3UHlMNabVLHjlC6/n0GNP8JF/+AIv+Y7rMHXATxMxL8S66BoMQqOyrfjg4m6OzSDHBJHERNRUEGyLvNOk1ytpFm0QnuPHH+eqq69hZCzzC4v0eqsM+n1OnDjJzOwctXEszM+hlKJZJDTyDFNOSGXBZz/yUZ7/+h+gU444dnzMhg0KiaDdCpTjMWxoIhWQ34hqLyEaiwipCcEjvZgm8UXfi/AGJwXY6LZM0zYuhIj4dg4hDSJJ0MES0gQnL76Hz5nVASu6w5bLdzMzv4HVJ5/kL37nJ2hoQd4ueNV3v5jf+6U/5oodu/DW8/CJx7HdU6wOPU0/Zjhco989TasxR2kSDh34Gldc81z+28/+NO/60Gc46C1f//Q/cGr5NMGPcXWXs8d6LG5oUijByE4YnHiM0w8f5F3vfBs/9pO/w8+87a/4i7e+CfutFF+VkKQ53kZulVKaPG9QlhOUUpEAHMDYCZRj0JJ6MsCUFVlaoBptFIJxf41BcvEJOQSghCTJBEmSoKXGEWJIl6958MARbnj+i3jiyGEeP30Pz7n2aq6/epYrL5vn5pe/hL9+9/uw1YSjh59g67adHDn+BC/+Dy+kNhPuufNrOBeYnZMcO/0kWZbzxTs+zw+98f/mlgPLbNq2h6s6m0gGD9GeXWQyGtHvr5OlOUWWPCXESHTM6imNYW5ujrW1dWbnZkh1jlIGYyzGVngfowT6wxJBYPembfzaf/v5SPVIHaiAtRfmqPuiazCBgHATcBVoCc4hqHChQsoNhDoghWOmvcT5s31m5mdYPddl/97L8C5jYaFBfzjk5JkzXL7nEuqQMRrVbFyY44F77mLrzh1smJnBlRMazYIsVVAk/PV7v8Rrbn4+W+bgzIrAyoSqd5zRRHI+VRSNnO1vfB+Hb/kz5s0EHzRCSZzwBCeQ0xjkEDyYcXR92z4+SLzwJLqNVAIZGqAdw+Eao+FBgm0+00f+tNd42Oeej/8daZHwsf/3Pbhyll/4j++mHJeUk3WCKZHDFR554Hz0uzTapHNbWZjVIMFYE5e4AfJGm+fs/TonTp/FuD7HH7yNr95zJ+tnz6GE59IrLufcyRMsbmhTNGaYDCzHH/0yL3z5q6l1ix/+yd/nln/8S7prK6QN+Kkf/3Xe9WdvJ88KnHXUdYmtSgbdZYKU5I0G1gYkKkbwhgwvHGneomjNAlBPxkz66wwmQwKBVFx8l4g0zdi2fQv4AEJhasNab52d27djPNx598N86IPvZWnjVqyx3Hf/HbSSGYRyfOJf/hmJoTeRZHmTc6eP0yhm+MbgK6yvrTA7t5lh7zSNosXp40fZuX0PjUab97/nnbzsDT/C6MQD3H++jSzmEeY0+7emSOsxtqKyjk6ziRCCqq5wztPIMmxZIqVASsFLvus7uf2LX0KImhhtXqKUjuZOZ7A46kkdyeoEvJ/Gfl+AddE1GOFrfH2a4BzCFjjRQ1gbM0BEQMgGnopTT9S05pYoR2PmNm5gpb/ClnaLlbUup8+ss2//5XRXV9m3fxe9Xo8lHdCJZuXkKR686+ts3LTAfKFQ2nL8kYdZ3P3DmNowcglpMkDJFraxjSt2eA6fGlIHAVnOhv/wc1R3v5NERMEBhKhU8eIpwyVaoskQyQLOC5S36ERNUyqGeOOQKjad0l58qBhraz77ib9lNBwQhkMcYKsJo/4qSmuyrCBtzKCTAutKhMoheOpyTJq2yfP4EDeTPmIi+ZWf/yl+6Kd/FVNKlo88ykOPPUmqBJdubzPunWV+YTNZ0QJnWD5/mg1Luxn0RzSEJM/hu3/wzTxx9Cjf+PSH+Z9/8bu8+OYfpiEsfSNQWU4IcOm2Tdx8w3V812tfAdUYvMdOxkgl0HkeVXBpgatLtE6QEBfLSYpUFx9NOdoNkpgmGiRfffBhvnnwALMLmyCdZWgU6A6Hjz7OZZdcwUuueTGX7N3HRz7yL0xGE9KGRNgRMjjacwuMywGrq2PKyYjXfM/389Xb76DRyDl58kl66+uYumZmx9V87fbbecGLXkGS5zQ1bNp+NVsXZjh39iyTc4+zfn4YvTAEsizHGUNlDFJKNi8s4lz0t3jr4merJULmeBeQCdGIK6MKkuAw1qGiHvuZPvL/o7roGkwINa56DO9SvGwjVIWwE4TcBCrB+TGiHpA3N3Di9Cnm5+ZY2NhhXI3prq1yySX7UY0OnSJh68ZNPPzIURbnGxw4eIjUVhw8eJCxNdi7JuQtxWjUYzKCRnIvp2avYzRs89yr5hkOYTRyPBY0N1yxwP2PDxisDVncusBoyysxpz6HDxalJcETjWHC4bx4Sp8vVUaQJnL8kYSqZFJ2oRpTlyO88xC6z/CJP/21f+9OkBmNpgKV0MkbmKpE6pQ0z4GoOlJJhqYZ23JvFXxAJwkgkGlA2UhQds5w6Z4lHn74cf7pX/6ehhRkOqE/GNFoztFu5GhhOHD/fVx+1XXU6RInV3t88uMfZ9/enRhSvvHpvwepSSQ8eM+tPP87vpPn7N3GqbNdJnXFqXMr/MlH/pVtiynXv/K1eGOpTRURRt5hbYzRHfbWKQcDJsM1pK+x3mPcxbfkB6ic42Ofup3WwiZuunoXhx47wdryGVpLLRySjVu2cejko5w4/hjr6+cRKuWn3vxf+PXf+l1C3zM3v4nxqMuwv06atZj4Ac1Wh89++t9obNjJWr+HIMS3zyw2clN3uf/OrzK38zJefNP1HD9yhLI3y3gyYef2K7j59VfjiQ5/5zzWOoppQupkMkElCXmeUdqSyXhIVdaYOqJtYpaNIpuGkhEiqbku6yk26sKri67B4KEuBUrWJGKNeuyRmUKEGaTeSCJqbLJAMbuJVmnolSOytXUSaWkvbqaqKsrRCm62xerKChuXFpiYarp4b9Lr9Wi3ckyhGQ3HLD95gP/x62/hXR96hAfu6XPdtS/mc6sDbnruxgjL62bc9diIq/Ys8NjRVVbPeRqb9tF98nbmigq8Q0mFCx4XXCTuAjJJQGZIEgQKXEUQFa4cMJms4m2Jc8vU9sK8+Xw7JaVGpQW+riJDLok4j7zRRicpxhiciR4mrRKklLhyjJ/iZaw3tDvzZEWBUgpT1/z+b76V//RjP8lyt8emRUWz0GiZksiKYSjZgwxJ3wAAIABJREFUuGGJublNTGTG/YeO8sbXvIBr9mzibb/3R7Tm5rHlhHG5jtYKMx5C3mGlN2b/JZdw/PgyWafBDbu3cMMrvxfrDChJJnLKsiTYmvGoR6hqhsNVfJB05jcgQqC7cpZMXJjjk2+nhIAi1bz+NS/mox/+Gq96+6/w7g/8K0eOHoFHDyGswMjA5qWd1HbEaNzn6JFjfPHWLxBkQARJb7hMkTXx1lOZMZ12h0aj4PxKD9MfgjfkjTbKWNbW16geuoOitZHLr34J50/dz2NPbCWTliEdtu3bxXj1GCbUZCKNGTR5QVlOSFKNsYHgLG5KcrjmudfzmX/7KAABSXA1zgsCFhn+d47N9AcuWCvtxddgJKSNbchkLtJysxFCZMhiByLJkVmTIsnYkA5Zz2dpK0uwE2YWFqgnE4b9Abv37UEmDbIMiizHmoyNM02On3qSLdv2cOrkE0CNr2v+8FffgqkD//0Nexj5Fu/44Ae55nmv4ot3jHjOlYv0uz3CIGOt1SJRY+pSYCddZm78fvp3foCZROBwBOdROsF6H7X3tUNpF8c7UxqsSBKK2W0okVGXy5SlxbgL9av5bZQQEfujNMHW1GW8OeqiiQfSJEVUCgigFFhD2p5DKR3JvMbjnEPrFOMsadFCKMmmbfvYsHGVte46iZQ0Uti1ZRelMJw9dRbV3IQpWuzY1KDdmeVTX/wSN73idVSDFbyQFI1mVIWVNXsvv5oH73uAlZV7ueyKZ5M15vh/fvKHWV89hzM1Qmc0Gi20VoSQ0UgbyDlF0y4hnCdtNHEE0tY8Slx8S37nHGvrffJGh5t/8IV885v3oouMK698Dvc/cBcqTUmkovYTnIPReMyD37ybECNccAG0hwljEhkZYmOR40NJ8Ib+uYdobdhDpzNDqjXzMwusdFfZtmsfxw7dw+JShyMHbmPjtl24usdkZZGbbriMPMlweJAa5ywgmIxL8qygDoFUKcqyZPPmrQgRc2hCCIigpsyNFB8CnqmhUwTwesofvPDqomswQrZQ889DJRkEE9F2WXMak5wi0iZSZfzCD83y8+/4Z0alpDnTJCCZmZtBpwknT51mYdMmMj1Hd2QYd8+zY9sS41oys30v5049Rm9YobA4axDOYvE0lOftP/IsfvuDn2br7mt5rJFz2f4ZTjx6lC+vlLz0pbv5+w9/iM0Lu7n5dTcy/s4307/t3QgvENEHimIafKRFBPrZCUIneA8ybaCEInQUMm+ykDRIBueBh5/pY39aSwhB0ergnSOpCyb9LrauydMMnWToJImMJyXwQWDrEpXk6CzDVhXBFcgkRSlFIgQiydBaMzGGgWkytmMS10OgOdOtyWcKWjObULPbee13vgRVD7j94BGsD7hqzDt+65dIqNBpPs2L9wyWz7B562Y8i9z/0AM89znfgVQqfi+FJNGaJE2QaUEmRIzulZpgKlxVR9SRD7Q685j6wpSwflslJEWrgzOWcT/w2c9+EIEHavbvvwypMh557AGyZBFLjTUlpa8o8g51NWZmfiPj/jrVoEQ0BEWjIFGBsnRkRcFknLJpfgZXG9IkYWQrijyjVwcSJgwGHmowzSW6q8fpDUYcPrmJ/XstzUaD0XiMc5ZGUaCkxDhHmqYY55BSMjE1RZZgfeSkhSCRMkquY8BZjlSSLM1R8sKdQlx0DSZITT63NZqzBMisiXXTHHYhCFqhVBO8JyQNtAyoRDIuK7JswGAo2L//CkLt6PXX2X/JpagdG+mPR2xc2sLKWpfzfUdDK+x4QMBCsHFpR2Se/eoP7OGJM8v8861HqctXsWXbEg/f+yVoKK541nPYurXNsRPLbN+zj+OX/yCdgx+a+iDibUcKiQ+R6CqFABvFACrJUI34YBy7HkYJ8ubFpyIL3uNqgyeASkgbTSAgdIw5Vjp+1vEFRpFmOcYYlFRInYK3EaJoDSIQLwne48oh/d4YM17jS1+5nYOHDvHOv/oQt37tmzz7+hbf94I9rA1rtsy2WFub8MSRo7zshufRyCTl0FCXJVJKVJIAhqUtu5hUFXmzxcGHH2Bp207W1nt4bzBljdKxAYoQYoQAFlObmMJIwJg68uvUxUhTFpi6jtJ9abnpxpv4rsUZ3vkHf4wUkoWlBd73N+/l537uZ/EBrHMURYHXkna+SJYVLOzdxNEjD+O9xfkUPc1VUkJTFA1Wz5+i3WrhvEcqwcbtVzB00BuMcOWYzduWWD12iDBZpWi0efCBb3Ljs/fSahRkiUZohSQ2l8FgCATSJAGd4IwgzVokOLTOAD8diQm8j6SOGAfg0SohTS7M6POL7psppMJahxWCoHKQGSpvREy/TBBoAh5vxkhdYKoJzUaTDQsbSXWDpgqcPHGYtW6P0gSOn13m3NkuVSVRQjE/M8vL3/gjWOsIsgF+aoT0luAcwVQE4dm5tcPv/ehO7vz833Ho4GGefcNL+NK//gHjusFH/v6fOXHyJEePnsZXFUkq8FP4oZCAjA0lUnjrqKNXAZEolEwJIVCN1vHlGF9ffBgRISTOWeoyyrllkpPkbZTWmLpk3Ftj1F9j1F1h2F1hMlzHliNqU02pxmPGvS7D7hrlcIArx9STIX/z7r9mPFjlji9+iXI8ZvfunSzMzfKLv/QzbF1aYuXsMrbs8dCjR3nFS68j15I7vnkf3sc4XghxWW8M+XxOplPmZ+d4y396A7d+9AOsrqyT5BlZ3iRJEryNCH+0wkvBaNSnt3qa7vlTjNZWGfXWWT9/hnH/4hNy+OnnOxr1sdbRahV89ZZ72bVrA6PxgNFkmdlWY5qoJEiTlGo8wdWWqqxYPX+aY4cfw7oEY2NWj3UliQ4oJXAipddfY3XlNP3hOuP+gG5/wOqJAwQ7xmvN2soqWzctkacp6yvLhKBYX10lOEe7GaXjo3JMXdfUxgAC4wOBAL5mMuniXID/3wJfYqawTUWzUVAUBaYuGQ0vzM/4omswAfDeIpzFi2jMwjiCABE83hl8NcSaGjseMN9u011dYTKpqOsap3KSRGHKmlYjwVUBYx1zc7PoJCXRCc204FzfxXwRH2OOVaJjPHMIWGNQdkBvNOC3/+uzOfClv+euL93Oda94M3/3zp/n1Hqf9XMDThw/jqpqymFAIuNNNQhAIALYuorBVNPRihCaqlxn5cQ9dM/dhnMlOrv44nQRIiZHPtWA4zxbiDgOq8sSX1bUVUV/5TzdlfOsnj3NuLvKZDTAVhOcMQTnmVQjxsMug94qfrKKJ6euJkzGQ/CgXM3Be+7lMx/7Jz7/mVv58tfuZm0w4YMf/BfyjTvYvPMypJSkRZvO4mZmF7dQdGb5uz/+n5AG3vJfXs933HgtOtFkjQIhBGU5oTYlwVicrfDGkeiMVnuO5swGvPNUVRlJwkog1EU3iCAgUDplptOhyDTSGl500362LlzDYGXA8SdPs3HjIu961zuRSYO0yJldXEAmjUhCNp5yMqBQHiUjt6yqa4q8TVXVtAqFSpqs9ycYn2Od4fzpRzGjmizR2LrPymjMgTs/z7nlcygpGXRPc/z0SaqqwhiL1FFAUteGPEtpNApSrRDeo7UiSTJCsAxHQ5zz+OAwXiCVotFoYqbiDiUV4gIdk110DUYgUDLBi4TgPd4bECEmW04TLb0NEAwyLahUQaPVotVqotIGiQpkjQXGdkJvMKYz36TIc44dP45xDq9TvFC87od+ChXENKAoPgCUUggZ89W9iCMuj+X3f/altKsDHLjtE7zuR36ZXe4cDxz8Bmvra9x/6HEO9DeACIRgYpIekiBiYp/SGTLNUTLDVUN6Zx6jGh2lmH0OrZldJBfh+CR+hjaGQgWPCA7rDM5alIzLLKk1SZZTNJq02jMsbt5Gc3aezoZFOoubmd+8jdlNm2m157HVhFCOqcZDtu+/hPXTx1g5cww7GXLq9EkO3f91tmzZzPd/7yu5646vs7q2zpdu/TiPPXQ/E2Pw3k0vBxGMWFc1ZjKhlUl2bt/ApBxPl/8ltqoJxsQgPCnQUoOz0TPRbNLZsMim7bvZuG03G5a2smHzVmbmL0wQ4rdTzjlWVtc5+MijPPDQwzxx/CSHnniSZ129g9/5rd8i9TndtfPs27aDt//GW1k9d4Zhv8/M7AKD/jlKU5GmDcrSY5xHaklZGga9FYSQCKFZXNiASjMGy8coWm2CS9i8ZXOM1NAJiWrhg+b8So/l5bOsrU6YjAUra+v0BgOMqZFSUhQ5s50O1nsmVU1pIly105knb7TI8wzn7XT8LWAK0ER4krQgSXOy4sKkol98Vx+poOg8FRTENKFOOI8QAedrIivdI2RFnjTAwaC3TGdmDq074MfkjRZKCFbPLbN5aQNFYxYnHGVlqMYVYyviQ0UqtBQxlS4IUDFcKoTwVJ5FQPC933UdnWaD3/jrf+PKG28ke+R+Tj9ymGKmg9XR+yJE1McHwHmPSlK0LnC2phosU/fOMe4eQaLIdYqwYzAXIcrdT8OlYjemrCq0ieFUKk0pxyPsZECvv0ZetJhf2k7emSNvtoCAKSsQkBYRNNo0NSEE0naLWz72Ic4cexKxcg6ZJJxfPsX5lXX2XrGTTfMZSlQcuPurlNWEs1XJ1972EwyrklRqglQ4W2OqMVIofuMt/5VurxdhiUJjQ42UgmoSMfT1ZIgIEdCaZC2GkxEiBIosR0iFMzWmHOG4+HwwSkl0AvMzc9SuRoRApgtWlk+TtxVvecuP8gu/8Tbe/zfv4bnXXcGO7fsoqwnrZ59kZn6JuhxQjkbk7QbCgKs9RdHAmIpm0WA0GaOMYH6mSRc48fhhFrfvwhoTm4ETCHp4HFlWkKQZk/Ead9z1ZS6/dC8bFxOMd1jryRKNcdPkTKWoqgqCxwaDQpE3O1TV/8fee0dbdtR3vp8KO518bup7O7dyACFEENgEgwnGxpEBAw7YGDM2+GE/P5jHvGcPxgkwA87A8uCxjTHJCTO2GUzOCCRLKGe11Ln7xpN2rKr3R+0rXZpWS2oBcr++37XuuvvU2WdX7fpV1a/qFzO/GZSaosgYDQZYV6GDiCCUpKdp0NozjsEIIVBB5OXiVvjQK5X1eYNchUBjKcE5Gt05jh0+QpJ0aPT6aBlS2ZxYdwilJgwURhgOLY5IoowoCFlZHSIbMVZanv3S1yCDq5EOKuvt252tfKIzU1EKfC4JJxBYhqMJ/9dLL+KvPnkbK0PD3V95J1se/YPMRSns3IXFYZ1DKOGtmlYOkGWrOOcXLlst0ezME4SJd8K0iqS7DfjqI93t31FYa0gHy+goIWq2ibRG1GZ4Qkrk3DzFuEmjN0MYxQSNBtnaMquH9oEOkKZAqZCg2SJudWlNz2CqCqFCBocOoRBMb99Fo9XinkOLXP70H+Bdb3kty/fcxTt/65f5b2/7S77yoXeQG8Pdd9yIEhIVNwgiLwKLW22CpEmgA8KwQZnllFmKrSribtcbJwjpfZ6kYLh4EOsksZKUVcV4uIap9UUmH5OfgWmxBdBrNr14UOp64+XIJz4IbGVzvutxl+OcJEoS/unDH2SSlrzw517Nyv79GDMgabQpMp/VNKyNBnSoGWcFlXFYJwiCJnNbOoxCzdLRAwTzW1BKc2hpla1bF7BOMTW9k+WD99BqSG6/4zB33nUHF1+0m2YQMhqNAOGjJ+NPKBLQUYyW2m96lCROYhwOU1q01piyIglalEXFYLCEDpJHtsNPEWccg8H5UN84H6HYL/g+86NzAvAeuErAJY/7Lj770Y+yfddWEBqnQ5wLSIsJu3edS5i0WF1bIYwU1gl0s8FUEFLYilFh+OpVV/Dkp3XJ0xG2Kr1S3jic9CFftFQ4NBh/HBbOoJXg555zEW949yeZ33UR5WiVQ0tLCHkWUvgER7aaUGQ5+fAApswJky5BMksQbfGGCrqBDDTW+pitZxqU0jT7MxRpxmRlkbwoCEJN0uqitIbKkY1WOXjnjZiqIEy6KGkZrx0hnRQEcZNGsw0ypDU1i0DSbHdodnqIvMHr3vBbvP2tbyGXiuu/9nmMdX5xShq0q5K3/ddXIFVAp9VAx4k3LMlTijQlT72orTSWZqdH0u6jtSZqeRGIDiLaUQME3hm0qmi3+xjnvDw/HZJl3oHUVgW4Wf//DINDIHSLUTqk2fSRLLzOURIQMRhlXHLJozi8dIg9u3b7KA4OyuExmknAsIipqgCt/XqgtKayE4xpAiVSBKRZhrMKJwRxo0WB5NCB/VSmZPfZF1HkE2zYY7z3ZlpTHSbDlDgp+fA//RPPfMZ30WwHNFpN1taG5HmOsZ55BHV07FgpSgWVLcAKcBIpPSOKmjHCKeJAUwwmRNHpaUV2xjEYn4u9wLkSIXwmOmsqrDA4VxKgEa7CVZY0yzi8fIyddjdgmGq3Obq8QnN2ClMUpOUSU9PzDCdryMqSjycgJCa3VNmEzvQciAlUziv3K+NPFhiE8iG7BQIRWFzl07GKICG3KX/1W/+Z1/3hP1AMVth78BAyiglFiKkK79VPilANwniWRmMbJgCF9QwmamCKMTYfUI4WH+Ee/85DCEncbBMmTYo0IS4yRmsrHLvndsqypNWdweYFofZJo4bLe8kmy6igjZCKshoxGhviQLB8cJVsvEa70yNpdkm6M1xx9a3ka4sMjpXIJCaKW1jrSNpdn30Qgy2Nt/oTEhVGVGWGtZak3cEWJY1OTBCEXkdXVSjtdXPOVlTGgHCYsqQqK4TE3+MChI5JWj4hmXZQFAXCnnmbCCHAmoxGotHSSwacFURRRFVVLMxs5+rrr0YkDaZ7fWbn5nFRzKte/Vr+5B1/QmAsDkOgA4rCkKUZgVTIqO57WxBoRVllmEmB1pK5qSkOVRVyMmRl8TCNdh9BTpAkTIYjet0GgZacfdbFrK2sEUQxYRAQRt6ysyhLr/fFgpMsLh+j05v1EZfT3G8ejcNJ0CrA2pwwimi2usjT1Jn2jGMwOAtl5n1IlMLaHOEMTlRoITG2REuJcTl7b7+Rpz/laayMBighmGQTts7PMlgdsFT59Kr5aBkVhjTjhJWVMa1GQEFFWpVEDnQQUugC5aS3ZJLghAahEcJ58ZiIsEoipAQLERH79t1JfugGxsMJS8sraBV4axdbUTqLkI6kt5MgaYMMCFWEqwqqfJFiaT8mG2ImhxmP73ike/w7DoejyDOqqkQISdTuEXdmGA+XmQyWkUhUNIVzlqqYEDXbDAc9smyEE5rKGMIwotHqEyUNMAuESROpNCpqgHR8/0tezvv+6HcgjRlVR6mqkqTbxxQFUZIQRDFIiXGOUAfE7SlkFJMPh2RFhrKGlDGmOgpAGMZEzQ7W+MgDzlofm0oqUMpbxMnKi/qkwlYFRgZIBeZMtONAsLK2TFEYpqenGQ5GFGWBEDAzNcU9S8vM9WdQsaAwjgpBu9XkZT/xwzznWd/Nj73ghYwmq9hMoMIIbMTbfvO/ccudt/OlL1zFrXfdiAoDwrDHR/7+b3jW859LOrbsWNiOwHDLrTfSbPYweYkLLHk6xlYTpqe28PjHPZovfflKnv/DzyMv/OkyiiJ0oCmLEofFVJanP/u5fOkzn0UqgbMl1jpAEQYRzhmiyIfoN+ubjtMQZx6DwWFNjpIhxuRIYRHOoWydy14IhBOYvKTbmmYtS5FFRbvfIwh98qBWt+PT1BYVqYBQKEZmRKcVU7mKqhKEoSIUTWx2hCpfobKSOG6hVQOjQIkQ63x6XwFI5xBC4qRDKu8E+Guv+Ul+5Q1/xL/95ZsYDNfI0zWy8TIIjVRNZNDw4U3QlOkK1fAYxXgvZnyQKj/mrZFO0zDfDwfOOcZrSxSTFBVHuM4UzXZMs9MjbrSI4hhrLXGrTTYZYsuC1myGKVPK8ZiV1aOM1pYYrB6l0WzTa/dRYUxnap4vXnkDReXQss8P/Mxr+fVffQXPffazEVJQmdq/yloczhthhLHX+ykFBtLhEFOVICRCQBAEyCBES4WxFWU6IUsnYLzHfxAnhGGEDBNkECG1QsqAAEdZ5ARRk+g03d0+XIRhk/n5DnlumJ7q4JAo6a01pVYMBmvsu+Em5uImcTskDheIw5Bev8Mf/96bePmrf4H3/+372bl1gQ994O8568JzufS7Hs9/etELOHJwkV94zS/xsX/5ELfdfif/9r/+kR944Ut5y5veAM7yylf9MivHDhF1pgnwjruBUpx/4TmUtiJJukjhXQucc5RV5cP+WEdVVcShojW/lVYzIUtLoijAmQpbKgSGKs/Ic28qr7FYcXou1adnqx8OnANrsWTghE/WhZdvO+uQgfZyNCfZfd7Z7LvtTsbOMDM9RVaVJE0flj1pNomSiMI5TJbR6k2xNh6zsGUGU63SbffYt28/R+/+Ci5fw1lBGTYIGtPIsE0QdxFhQumALEMIiwhitI6xEpRTVCbjwLGUxUO3YQpDni1SVhVx0kckfQgibDkmXV7EFgcpVm4lL48h8gqLIpAKIatHuse/47BVickLwjgmbveIkwRTFhhjvKy9KnHGEbY66CjGmYosn+CqkrKR0lzYyWTpiI9DFsbEcRMdx+hGk7f96TuJwy6D8RL93lbe8Kfv51nP+B6E1iAkKorJxyOqwYAojhHO+sRxUmJtRW9uAaG1D0lTFX5hSppEYUxRVYRhQKM3jZI+O6PfTAgqWxFIjXM+rq7WAc5YrDVk49PTCe/hoKpKRumAOAtQQlNUlsqULA9WmelN0UxaTE1JMBF7D+7jnuUjzM9u47GXPIYgDNl9/oX8/lv/lO3z8witeOaznl4n9ZOoIKTd7fFfX/daVkcj4iQmaU/x4b/5G44tLrJjz1n860f/mZ986ctYOXYYm1vO3bmDl/zET7Hr3HNQEvq9KW+JQO2eIMS90SLCWDFOM0bpgCNLa6gAppM+aT4kakZoYVAqIdQhk/GEMAp95I7TEGccg3E4nMkRhGAy708iBSDRQYhUEdZVyLjLQjLL5z73Oebmt7I2mRBGMc5amlGIiCOOHj1Kq9VnMBmB1JQIVodD8jTHhJqiGnPgwBIzyZ2sLo4II4cOBDKMCaKd6HArItCYLCXNVghDTdiYI2j0MVWBdo4/e/33k0+Gvo1CEcVNVJighCVbuY1ieIx88EWqgQBXYaz00Qi0oagEpROcvrFYTw3GGCaTAdMLO4mSBGsdxtQBQ5XElIa8yGh3e5TW+ThRSRupFFGeEiYtelMLIARlmVGmGabIaIctFscVrV5AGGuyfMx5517CT/zsL/K+v/ozwmbLpwFQiqLMCYIG4MhHQ9LhKkJJGt0pkqCJVopsZP2pZTAkUxNw5j5mFMaEUUKRZ5jJGGclubBUeeYda8MIW1rS8RprR/Y/0l3+nYcQhCqmKgxGOpIkJh8MGA2H9FotolAyGefIMOEP3/E/+KHnPYfDB46yOi64+LyzmZ3q89SnPh5TeWOfdrdHURReDyoVjVbJE598OUEc0+r600iSNFB1CJhekvDu//EOirJgx46dKKWoqorl1RFlljI93fWbh8BntVxnNk44xlmBw+d8kRJCrTmytEiz2UCXBZWSOKsoTEEQ+I3EZrj+0wUOnBDkRYqyJUJJKqcIZEhlLWGkEU7zvk/dzmXPfSYXXHQReeaoXEZLK8JAsrI6YHp62vsjOEer0yFqxFTjFGkF/XabY8eOcslllzC+4rNM2RjnBhSZosgsuBzLbejg1trlRmAt5IEhHPVp9B5FEHUoXQlBBLZAyohQR5T5Gun4AG48Ip1ciUkleerjFqFUHXTV4IxDSssZGCmGIIqJGx2MsZRFidIKrTWVLUgnE+IkIZIJzjmiJEbIWlkOCB14vU2j6SMClAFSamg2Kdst4riFcIK1wZBue5a15WNI1aE1NU1Z+TAkDh/8NsvGmKrEmop8OCBsNcnHY0xRgApRQUSsm1S1uTFCE4YhItBoFXgTVq1xJvKOomlBPpmAFLjx2KcNVoqphR2PbIc/AtBS0W23KauSosiRUpBEDc4/+3yqqmCcFbQaU/zGW9+Kc44LLtjGs7/nB0FpGo0EhyDPMqTWFHlBGIZUpiTPchAOpb1T9GQ8oiEkQgkfpFsoirJEKUV/bhZTlJRVRWktYRDQ6rTIlEApSRhFVLbeMNRWgHleYKyhzAvGozVUIJEiohkKJmsjBq5kZmoaJSqQirIqvQi82NTBnDaQOLQKEUoipUaowA8o1cQJ7XOAtHZy/Q03s7Q6ZrbfZzQqSK2CNKU71ScvHVGziXUB3VYI0tKbmSXNCqo0JZ6eZTJJiSpLZZZQWhCEfgeNMVQ4RJIQqnMJwhAZddA6AKURBFgZIJzCFCUiX2KcjyjSm6myAVXhMKWkqCTOGSyCAOXD31if2dJKRZFDaU/Pnc/Dgfd1CjGm8voQp7DWYZ03E1VKo7WkyFOyWh9SFRkyjP2iUOTk2cSLopT3R9Faox0kzTaCkLm57YQCAgTtqS4OSbo2QAcaFcZkpvTBE53X8bVntxC1ulRlSTkZE0QKJwVKRQRhQK4FUiiSOMHUp5iqqpBWo6OEULdwLUOcdRFKooTwznk6wp2BSn4hBUEoQQUELqCsCnQoiTVMrKIsUo6NU3rtLt2pLpc99imkRUkQwCTLUWGA1JqsKChNhcbhHIzGQ6RQOKDb7SKtJR2PKLVGK0Wr3Wa8NsCVJcZYoijygVBrCzFTWXQU1hkoveSgqh2rrbVIKTCZj0toi4JO3GCYThgOJ8xN9cnKDGcr0sIiIhBoiixFB9Ej2t+nijOTwQQxGgUIb7mlAqRUWGmRzmCdIm7PE7R6zBcpvakZWu0uy0fuxoQ90iyjlSR0ei2GaUY2GRIlHaSy9NsJR9I15gNB6hzV9CWUq18liEMarZ1E3V04kSCURugAHXVRMkIIBWYM1QhbTLDFIsV4kXSyl3w8Is0kRW4RQmGkxYCPPqCkj/yMwUpF6QzWSIR1GAHlGWjCioPu9CxFUaCCwEcvtpbJaMBkdYVtUT3JAAAgAElEQVQ0GzE1s4VWf4alA/f4kD3O4mxOf34PjXYXHQkf7l+HTMYDynTM2//wLQgnafc7CDqs3nMz7dkuf/am17N29BDOVDijsZVBoRB1OBpZR+qOkgQdxpRFTjYek49WGUxGhFHEeG3V+7m0egRxBEIhnCPPMqIoJAwbPqOiTr1YRzhkaTDCEajT00fi4cA5R2UL+lPTHDuySlmWREHA4uoatoK7Dh3kE5/4BGtrE85NQpbXVryRjqj1rkWBFbC4vEqRZfS6Xao8Q8iAdDJGCMnR7DA6CDEGwiggUJLhaECQJFQjL3IdpSlaK0KlEUoThwG6PgWXRYlTAmMMxhiKsqSqCoqiIMsy3vnOv+B7nvJYQh3R78Btd97Nju1b0FpRFIJYxUyqEiE1pjg9I3KccQzGmYz06E0IGRMEPh+DVIHPCOlKjJVI5Thy9EK6lWF1MkY2x5TZhN6WrWSDFUqhGQ7GaCEZDteIoibWDjh0JGfHzm202zNkpGhg1XV57PZnIKQkiKcJW1tQUYIVyuuDqgyXjTCjA9jhXZjsZopJQZlbSqvIK8ukFFTG4qREOEFeCawB6/A+NvgjuLmXmTgvPkF4r/AzDMb4iMhR0gZnvLOplMwsbMfOLaC0IptMEBK2nn0BVVUSBBFFnjJaW+bI3UdQKvR5Y8IIqhLrDI+7+CI+e/UdhEHI1FSPc3Z0efNvvJ4ym1CmY4wpGSwfpcwmJO0pwiShLHJcnfLYVSUqjDBlxszCHtozW1AyIE2HREVOEEYICelogMQxGa6BkBRBiJ6MyPOUcjhEBRFBEiGkQquATJyBjpYOlgeLxGGTUMFXv3wT+47uZ1KOwMVkoyEz3Vm2bVMYK3njb/4OT3vaE7hr7xHCKKI7vYWnP/nJpGMfJXl6bo4inyCFwBgB5YRJVtHvtknaUyhRIIQk1AHHllZBOoQRGOEYro0474JzsdYxGo6IwoD+1BStJKZC4KwPsFsUBWVZMh6P+Ou//iCz22ZwlaawEw4dPUCz06IylqPHVmm2O0wmQ59UTgU4ET/SXX5KOOMYjDUT8rU7UTqhChKohjgVI2lQmRWsKQijGZqd70HbCe2kR1FUJHGElJq1YcpsECHiEKslvf4UTgWsDce0p6bJJiPa7TaryxPWFo/QimKaQUxhHQJDUUxQ+FzrlDkuW8KOjkB2O2W+RFWAcZIKSWXAVqClP75XPqoNxnhrlxKLq/PMCCtA4q1ghA8B7sQ6qznDmIwQpOMhxlaoIPAhOpT3QyqKgtCFaKVROqAsc28mHipC3WKm1fZKVevj1JVFgatypAp47o8+n+e/6IWESUhVGl71qv8DaywqSpBKw2RAb24BW1a1jH+CUgFxo+3l++MhZZFi8pQDd95Ib7yTZn8OJSGIm+R5SjEekY3WqMoUW2Sk4wEGjbUFRVXS6czS6swgtfL5YuR9uYbOLDjWjhlMvsjO7XM857lP4h//8ZM8/7u/n26vwfU3XM3hlSFZUaFwLC4f5cDth7GqzYUXnM+NN97E7/zuWwjDkCdf/nhWVwaUZU5eKbLhErPbFmhETQ4uLtEc5ijt6LY6rJQjEI5iPGHLtlkOLa6QZxnXXXc98wtbKIuKvJBUtmIcxZQCpHWEYcSH/u5DOCvI0gwp4Nyt27n2tts5a/s8zWCKmdk+ZVnSSAKMkUxMyXSrhxKOY0srj3SHnxLOOAbjrCMf300Qh8jCYI0gCGdwQRNlBYFqIW2JUpIy7pEOFuk0Z1leXmbb1jmSRgcdxkRxgjGatCrp9iIuPG+W4WDE2uoaB/btp9VpMR4ZfvxF38/kb/87TmqE2o4LIqRo4NwEaYfkxZ2IsURKyHJJYQxFYSkrn8cCIb3YC++0VxqfWsAJEBVIFAiLkY5QC4TwzMjiTZ05Q3UwvZktjAdLlFWBcBalI/LUe9PnqQ+9IbUijL2y3+Q5UimKssI5S5amYC1lMcEWBVGcoJtdKjsmnawRN5q84x1/RFmWlGWKEwJrLDII0ElIlWc+plRVMR4s44QnRVlZwvYsnSQhDGNskfskd1qTqDZBGNOd3oK1hrLIsGVFPhmSj9cAQaM3i1CSMk2xVYGOZlDBmaeEEcD2bVPMTk/zu297J698xQt48ndfRpwo9t51D7fecwhjHaHW3LF3Pwtz0yS9Fru2zHHd9V+j21tg1/Y5knafT3zyk3Q7c9hqkRtv2Meec3YyGCyTlxYlBd12FxU7Bitjdu/YRdxoMFxdZTgZc/TwEYR0SB3yjO99OqCQlJSV4MiBg3SmenQaLdr9WVaPDRmlazSbDZI4YWU85FFn76KoKmanuxRlgVQCYyDQjoYIyYoRhVH3Ol2ebjjjGAz4UCI2d9gwRusQEW7DUVLaFSItMUWTJzxqls9dd5jhcsraYB/9boter4Mwhqg7QyhBCcd4pSKIGqwsrhImETJoUFYTuu0u2y/vsPfuY/TTFOEccBOlc7jK4ZxEOYuQEcKUFELgjKSsLBVelyKAzFQ4Kymt17tY57zrtrBICcL5MDShVrXDps+aqWS96Jye5vMPC6b0lkXTCzuZDAeMVhYp7ZD+1CwlOdlwQLpaUJUFzU6PIOly9K4baC/soNHsEccxjQSQClMmrB7ez4Hbb2AyWSVudQjjBnHYoD0zT7M7gzN4IwEEyoEOY4xzJEnDpz8OQoLYByu0ZUlaZLjcR3euygxhNKrZRgeaRncKpRRrS0co85yk0abR7WLMVp+rqChxUhCGsT8Fa4E6Ax0tjXVU0rL/6EF+6AefSaCb3HHoNj7+6c8TiJjzH72dbCxpdGN2bNvOymDAXbfu58Mf/Fee8NRLOXpgxPxCj2azwQXnnctoMuTuvSUXP/oCtm/dyVXXXsl8p02alvRmpul3Wky6E7QOKMsCJwVJlLB9x3YG4zFKCq695jqKrCKMAtqdFru276DZ7tKf6hOGMTMz04wHqzSaLRYXj4EWTNIhnVaXokyxpUNrTZanCBdRVetGAikuOD2X6tOz1Q8DUoXErUcjkhApQ5SO0FGTKl31UZZllzLewme/cC3xzAL9+S24smD79u0cPXqMWCeIsiCzjiiRbO3HKFlgdcA4LRgXGd1ezOrKgCWT09Il7cKbOBonsJWjtF4uK4QEUfqEZC7CyYIwEsjKolF+EuGVkqEAY0AEkqoC6+VhOGmR1juJOkAohbDUEQnwqQLONBGZsxy58ybWejN0p+fozS5Q5BOfSdApivGEKEmQUuOqisnKER+B2hps7S+T5ylxo0UQaTqzc0wtbKPKU9aOHGS0doSVwZC15WO0ZxaIohiJRAeaweoxH4MMCLSmSMeUxjC7+zx0EJE0WnR7s5gqJ08zTBXgak9/GShsVfn8M5VPkmaFz9luBN45sywRgUTVSmWT58jgDKMv3jz7Xz/8aZ73vKdSZYJ999zD1vk5nve9T2Nxecg4mzAaLbFj96PI05JmR3HB2Rdw/qPOotudQquISTngtltv54ovfpknPPFydu/ZyerSgMXFw5x19rlccP6FzM1sJ0kktnJoDV+94vMgFVmakqYFRZlSVIZ20qZ0hkYCURTSaDbptDrEzQa33HITO7bvYsvWrVx37Q089nGX0m63CJFM9bssLS8ThgFIgy2UDyljnU8lUhniIGJuy+wj3eWnhDOOwQjVIN5yFgLt813rAOscSgQoMQNhB7nnR6juvII8M0xN9RCmYlTktOI2WENhDIqCJOqzMsjRLme60WRoC6rSMjMzhauGHNg3Ym7XDFkJWvqFwgnpTWelwNWm7RWCSqRoJFoByiGlQDmLMmCkoMRhKn+SERoqfDhx5RxIr8yXzi+YUkuUMD69sjjzjjBCSKwtyZePsVaWDGoHxdb0DP0tO2m3mkwmY3QYoaSiLAriIqXMcx/puEixZclkdcWLprRmZvseWtNzBI0W7XInkfZKV4egzAvS4TIqCJla2MFoZZkyy5iMBwxWjpCuHMWkI1TSpNXfQmdqlqTd81ksE/8cpfwpxFoBzvjYdGGMVIrxcI10NKRMJzib+1TQeQo6wqZDwlb/EevrRw6OF77geRCUzE7PceTYMovjJb581bWEScVlFz2JI8eWGI8KetNN5uYu4P1/8z6oGmzbMwPS0mg1GC2nPOZRT6Db6YCATrvDcG2ADhR33XkrB/feRthoEErFoy99Ijv3nE+jGdPpTPP5z38clUu6nR5IRV6kBCogTnyqYxEojhw5wsUXX8IdN9/CcDhgzzm7KU3BZDAm6Pcoy7JO8e0zVGmtcFpQ5d7rX4Q+GsTi4qYO5rSAUBqZzBDqGGsynM3rDX5AEDUxaLLS8KjHXMjRg4eQUqKiJmsrS0RTUzipcZM1er0uy0trLMzNMCgKrBTMt7qILKeYTDgyzNizZys7tyYcvGoPs8XdOFEBzkdxxmFwSOUZnbWS0jmsC5CiROkKDRghEEpRGkuFI0SSGYtygso6rFQoJ8BYjDCgQQiLRNTyske2vx8RCGhPLfC1f7+K8y7sI50hsiXj1SUGhw/T3baLTn8KoRRXXXM9l1/6GHS3z+pgSBRqtNbkkwHWWhrtDlWRM1w8xnhlmXwyxJmS7vQCQatDGIZIpZB6hmI8BKHpzMyTTUb01BZmt53FeLBCMRkQxC1MnrK6/y6K2QWiRgtrSoQOqfISoSVJo4nWmla7g2h1GA/rSM6tDjgDQpEOlknXVnBVxUhqquzMywcDECaSlTXDHftu41Of+RKNXonLu3T6fbJqlWc+9XIOD1b50heuZ/fOlP7cAmWecdPXr6fZ63D+xecxOz3NYy89h6KCNMswdpleuwvSEkYhWknKoiKIQg7dcwd7999Fq9snkHcRRhHUWWt1IEmSPnEcEbiAldVFlIXzzj+X5eUhj37co4l1E6klk3REFCbMt2MwJXESooWkrBRSOmwpCFXojVSExBpzn8j7NMMZx2AAVNTAypDK4XcJxqFCDVGCDrZw5HDJwX17CZIuBw8f4ezzzmcaRyNQWCnJXJNBZljo9xnn3omqBFxW0EjaLMx3Gd18DzNTLdaGJcWeH6G87ffRQqK1j9gspfSJjhCYyuGUxVU+6VmgJLb0Zsk+qq5DYFHSO5hFTmKkQ5iQoqoojAPjkFqghEM4SaAEWvsEZWciBuOUp37XU6iQ3H34GDNBzJeuuo4wbvD0bpuJNdx8zefZv5oR2oJdOxaoZESku9x8w22ce/F5YEFQMTh2jCAISVpN2r0Zb3lmKkariygBOEEQhjjhwFZkdfi3oixRWtOe3UogtyOloiwz0sGQ8doKZVFgyoKo0SRsdjB5ymBp7B1/pSbPRkQy8Dva0npGphxB3KbRnsJay5S1fozwx49kd3/HYa3lU5+8gsc89mL2HznKU552GTt2zDFaLYgbGtEK+eJnv8r8jh0kBBzceyu7ztvBkSMr7L7gAg4f3MdNN95CIgMWFnagwgpbGTpJnyooIXAkcYs0naACrxuxwtBKOjQbLRAw3LuPQlrm5zr0p+ZBSIaDCUE7Zlt3G3Mz8xSmYnZulsP79zOYZMzPbWFuZpqgpRkYaFiLxXgjIGmw1m9A6xWFPM9J4jpB4mmIM4/BCIkzEkGFXN/ea43UERLJbe4cvnDbLeAiRqMx2+e3kA5HpJMxeRKydaZFt91llBvSoqKsShqRJLQ5KozptdrklWVuNmE8TMlWF7l7RTBbe/YaW9VKep/drrQGpwzSCdDah5o3FhQE0vrUzs6HjcBKJBalFVSWwlQIJFqADBxaOySSUEAY+GunT8+B+XAghOCsc8/3Hu9BTBSHFEXFc549x+rRIxzcv5d0uMT8novZdckcRWG5++brGC/eQ9KZI2q3uOnLd1GWJVP9LVRFCtbR33E27akttFvTTMZDkiBgtLqMKXOqMqTRnaLIc7TWJP1ZyjylqiqyyZhBnXUyDhsILem153BGkDOiyjOipE2r0/ERB/BJ08IwIJtMCKMIHUmqylBlI6yxFMYQJBEyilCcnrvbhwMpJI973GPQWvHEyy4iqyqMkURNmKQl6doy5563m6XDazzvh5/F3fsPYG2Js4prv34jZ+3exe6zFrjhhrtpdTRREGMRVJUliCW2dKyuHGOq16M0PkZhVRXEcYKwhiLP2XH2HowzHNp/lM996p+5+OKL2XPebtLBGqLTpshSSifo9xu0zz0fFSpc5SjLjCiIOba4SL/bZDBIMcL7amntJQ+OEKUlaTEiX1tDbp5gTg9IpWh0GtjKIqMYpzQ2iJHCYoMOd920yM8+/1K+dMuYRpDRChW37F+h2ejR7rQxNiNBsLXfZJAbOs023U7oEyDZionJsZWm2+0TuYx+9yzSvEKuTPvQIFKArEBLFBLjLNrHc0Y6hxMCKQVCOh9R13qz4wDnY5ZhMUZgQogRPsaR9ScVpQXSCQItUAof/iYWwD2PdLd/R2GqkqP77qSzZQtNIenPbMVUJWWZ05+aR+pLscaydPgAxpR0O21mH/9UTJFhC0NepJTjISoMmawdYbh0gGZvhmN7b+Ke275OZ3YbURDT6k3T6c94GgYhzlSEQUA2HpGPBpRFTtzt0+r2kNMzlGWBlpqyLIiTps+uOdCkgxVWj+2nysZkaUajO0V7ao4wDAnrjJjOer+mqNVGKcUkzSjzklAqcnPmOVpa5+PMIQRFWRCHATiJDQQ333YnBJrt8wtcff0XuX3vnTz5qc8gqBQf+dxHsFnF7fv2sW1+C7u3bOELn/gqT3r6YzFFQVE5GnGDVi+i3dmKcwotLUWaU45LhMyIkxmarRarK2tkeU670+CFP/UCbrzmBoZra+zadRZRqAljST7IUSrk8//2b+w+Zzfbd+8mabRwVcl8K6GqSqyUqEASqISympCPCxqtFs4awkBRGIk9TVWp4kzLFyKEOAbc/Ui34zuMXc6509MM5RSwSeP//2OTxqcHzjgGs4lNbGITm/jO4PQU7G1iE5vYxCb+w2OTwWxiE5vYxCa+LdhkMJvYxCY2sYlvCzYZzCY2sYlNbOLbgpMyGCHEtBDimvrvsBDiwIbP37YsR0KI7xdC3CKEuF0I8boN5WcLIb5al79PCBHU5c8QQlwthKiEED9ykufuF0JcV7f/OiHED9blWgixeoptfe/J6jzu3u8RQtxQt/Uh9Z8Q4iwhxIu/1e2raXy9EMIJIdaOo/F+IURPCHGOEOKah1jvy4UQ8yf5fpPG3/zbbwuN63u21H33c8eV3y+NH2j+PxCNTxWnMDb+ixDiJiHE14UQHxdCnDCH9Pq7fqvb+61E3a/LQojfOq78C0KIS09lHAshfkwIccG3tqUPDidlMM65Jefcpc65S4F3Ab+//tk5VwAIj2/ZSageNH8CPAe4GPhpIcR59ddvBX7POXcOMAF+pi7fC/w08KEHUcVT6/d5MfAH36p2P0j8JPBm59xj1/vvIeAsfJu/pXDOLQF/BnwBWGIDjR/mo18OnHDxEULEbNL4RPi20LjGjwNfBl7yYH/wIOb/y4H5E81/IcQp+did4vy/ErjMOfcY4CPAm0+l7v8g+D7gRjy9vlX4MeCEDOZU6fSg4epw7w/0B/wG8Nr6+hzgevyguxrYBaxuuPfFwLvr6y3AP+AHwVeBJz1APU8F/mXD518HXgcoYBFQJ7qvLnsv8CMnefZ+oFdfPxm4sr7W6+0HOsCngH8HrgWev+H3P1uXfR34i+PrBN4E/DkgT1D3LwDLwF3Ae/DM/e11P14H/Kf6vvsrvxJYA64BXlO3+e11n14LvGLD79+BH6T/C/jfJ+uT+jdfBh4HrABv3EDjEvifwA3ArUBRt+lDwE8B7wbeWI+BNeBo3Z7/FxgBt9TtDeu+/3Xgi/X/zwNX1G2/oS5T9e9urPv4Y8C/AM+sP19T9+GPb9L4lGl8JzB/fH/V9L4B+OsNNE7qe94IHAAO4ef8OXjv3RLI6t+tbqDxLcAngcuAq+p3WqmvnwT8nxto/N66jnUa31bf33yo878ufwLw2QcaGxvKnlT3zdX4sXluXf4l4FEb7rsCz/A2jsV/B5r3N564b53887qPPgrED0Cn9+MZwueBJ2wo/wJwaT0mVoHfr+v7ODC9Yfx9rW7f3wJJ3U/rY/IaYHf9rN8BPgf8CrAH+HTd9o8D2zes49fXz/t0Xfbouo5r6vvPOun7PAwGY9c7gA2T9wQM5oPUTKV+uevr68uBd52gnhdvLMdP+D/A74Zv3lC+B7jmFBjMdTWxJ8D3nWDxCYB2fT0H3FZfPwa4GZiqP09trBO/EPwptW/R/dS/caH6cfzCoOp321fXd3/lzwI+vOFZrwJeX19H+AmyE3gRfiBLYDsweIA+2b3er/gJ9uENNHbAMzZcD+vv3gP8FZ7BTK3TGD85XlYPyi/gd/Pv2tD3v7qBxsvAU+rP/4xffOaBCgjr8kvwA/mjwOV12QeAH9uk8SnT+PeA1xy/6G6g8fpcfQ/wK+v9QD3/axq/HD//r+EbF739wK/W9H03nol8sh4bv4s/KV+PZ1TfjWdW65uBj+LXhPXfrjOSBz3/6/J3rffZ/YyN4xlMd0Nd3wd8sL7+OeC/19cXAVdsbGd93appeH/jaX2T9uj68z8ALz4JnZrAQSCuaf/2Dd9tZDCOepMF/CbwB/X19Ib73wz84onWxfpZf7zh80eBn6ivXwn8XX19E7Clvl6n0zs31B3xAAzz4RyP7nDOfe1B3Pcs4HxxX9j4vhAicc5dgd8VHI8TBUVwJyl/qHiqc261PnZ/TAhxMX5nvrH+twghnoKfRDuEEDP4ncsHnXPLAOv/a7wR+JJz7hcfQjueArzPOWeAw0KILwCPP0n58eKW5wAXbpDZd4FzgacB73fOWWC/EOIzD9COl+AZBPjJ/9wN31X4RW0GvwOaqsvfix/Y1wPfi99x/RB+8D+tvmeA30y8d8Pz1utpA9o594X68xeBX8b3fQ68VwjxT3Xdrv7+D4QQ78NP5gcKsLZJ42/ERhp/AM8k/+gE993lnPtKff1e/GLzB3ga/zz3zcG9wB340+bx+CB+1xzhF8p18VqEZ46r+IXrl4F/wi/AUNMYf/qRdd+s40HNfyHEy/A77Nec4N77Qw94jxDi7OPKPwBcI4R4PZ6h/sXGdtZj8e+dcyMhhOLE4wngdufcdfX1VXhmf3/4IeDjzrlMCPG3wJVCiNfWdN6ICn9CAU+n99XXlwghfrN+pzZ+43Z/+MCG68uB59fX7wHW9T9fxPfN3+KZI/iT3a8JIXYB/+Ccu/0kdTwsK7LxhmvLNw6AeMO1AJ7o7pPdbnPOpSd57n5go5JuO56rHwVmamJuLD8lOOduxe+ij5dN/jR+Il/mvOx5sX6fk2Xu+irweCHEQ0nMcX/RhR5s1CEBvGpDv+5xzn2y/u6hMN6XAK8QQuzF76y3CyH2nOA5x09yid+9/Ql+oekCbwHe6ZzbxomZwPqYOcQ3xsHbgl9cjwIpfqf7ROBfgYPOud8G/jN+x/g8YOHBvNgmje/FRhr/A/C4DTTeiOOf6YQQDTyNP4A/wf1PPLO4v/k/5gTzH38CvKEeG8/GM50n4hdRtYHGOfAiIcS59e8f1PwXQnwf8F+AH3YPTff1O8DHnHOPqtsYAzjnxsBn8Iv+C/AnN44bi1+r23l/44n6fdZhOHn8x5cA31fT6Wv409DTTnDfN9Gp/v8e/Knl0cBv8410OB7jk3y3jp8H3oBnil8XQvSdc38N/Cj+vT4uhDhR++7Ft0Q5X3PYFSHEubXC70c3fP0J4NXrH4QQD6Q8/gpwkRBilxAiwosDPlLvaD6/4dkvw++ATgm19ctOvjkSZBc46pyrhBDPBrZteI8XCyGm6t9PbfjNvwBvA/5ZCNF6kE34XP08JYTYghcZXHmS8iF+V7KOjwGvWlfSCSHOF0IkG34vhRDbgKefpA8uwosHtjnnduN3kJ/hxIrmPcC4nlAvwct3NX5x+SReNPKC+rmXnqC9G/ExQAshfrSm8UuBf6u/uxK/A3sdXizyv4UQZzvnrnXOvQnPMLad6KEneL9NGh9H45rOb+V+aCyEeEJ9/RK8KCXB03iCZywv2HD/EC8asngdi8YzlPUFKMWfQl+N19l9VghxGV7G/yk8jWeBxjqNgV+q63vKg53/QojH409lP+ScW7y/vrgfdPH6JbjPaGAd78Yz1y8559bqujaOxauB87n/8fSgUW9cLsf3zTqdXsOJjTICvNQA/NxZlwQ08SfioC5fx8nmIvg190X19U/ixxd4/cpX8HqwFWCbEOIs59ztzrk/xM+JS076YieTnx0nH/wNvlEHc7z+48fxx+ZP44m9roOZBf4OrxC6EfjTuvyEOpj6ux/EK5XvAP7vDeXn4Dn77fgd1bqs/sn4k88Yv3u4ti5X1Epe943y+WvwMvqXuW+Wz8/hRXdX4nfSt3Kf0uvleLHQNcCfHy/fxHP8T3I/csnj7n2oCuAQv/h/HT/wFF7Oen399yn8IJJ4OemNwD/ilcAnlM/jdzm/fRyNf7+u9xz8qWJdPn8dfgHO8bvJd+En4JvxiuMj+MX/aE3/F+GDES5yn5K/t6GuX8YvQHnd/i5+AbsKv5jl3Gcg8M56LBR48cAmjU+RxnXZZcB1G/prI43/DD9XNyr534y3MLwV+Mu67dfUNF435HhpTZvP1/R/N/cp+Vfxivub63Z/Ea/MXwZe5+6T7V9f1/1pHtr8/wxwuG7HNcA/nmRsHKj/78fro55S1/XFuq9uP66vbgeeteHzxna+t6bZCccTx62TwOuBX7sfOr2C2uBhQ9ksfj4FfLOS/3fxSv5PcJ+S/5fwc/EzeMa4vgY/DS+W3Kjkv3RDPWdxYiX/R+oxcT21Pgj4Nfy8ugYvYeid6H3W/zaDXW5iE5vYxAlQ+9N8HLjQbS6Up4RNT/5NbGITmzgOQoifxSu0/59N5nLq2DzBfBsghPgIXva/Ea91zn3ikWgP/AhpAdMAACAASURBVMds0+mM/4j9+R+xTZv4Zggh3oU33d6Itzvn3vNItOfbiU0Gs4lNbGITm/i24IxLmdxqNl2/3wfnMMZgncFah/fT8f+d88YPYoPVqlIKhEBIAQisMRhrUUIgpERK5VMdC/89gHMWax1SSpxzWGv9cwVIqVh3DXIOhABvgOcwxgIO57g3F7e15t7PUkqEWK/rmw011suFEAgE9+zft+hOs0x4Dwf9ft8tbN1oySzu7WuxwULY1u4FUkpfeq+v1n39dzyOt2Pe+HnjM9Z/+Q3fbYADT/jjH8463e67Pv42vyl0G54D11577RlF404zcXNT7Q194+7tf+ccDuHt6h04HEr6ueVRU805bJ2uHAHO+nknhcC6dbt8gRMgnU9LLqS812DfWIuo56TjvnmHq8eOqGlkAecQon6iEBhbK9hF/Tgh6/FQj9V1BbyQSOGft/fAkdOOxmccg+l0Orz6F14JlWU0HmCdIdABeZ5TFN58XmlFWZYYa9FakYQB1kHlNMUwY5gu4lxBq9mm0erS605jrENIgVKKTqdHGIYIIVhdXWUymRDHMVIKisLXU1UVSim01iilKIoSpfx1s9lEK40OPHnKssQ5h5QKZz2jKsryXibjPEdESkkcxxhj7q0/DEN+5pU/f0allp1fWOA97/1rjDH3MmOpQEmFkgpRT3ILGGNA2HsXByUVQmqEVDjAGoOUAikVSRihlLqX6W9k8uuf79sQWKwz61Y6fjwZg0CilCIMw3ufVVUVUkomkzEOh9aaJPJ1bdwwOOfIsoy8SLHGfsN3Z5913hlF4363xetf+QL6rQRrwQoBtqQRhsiqQChBURSkRmBliBSGtgaHRkjrGYupyNG4ypDmFVbHFKWjKlPazQilQyoncc4inSDPxyRJgrUFSmqcMTincFLi8IymLDKk1GAsVkicChhnKZM0R0kIpEIIQWYqGklCpDX/H3vvGSR5et/3fZ7wf/6h0+TZcHsJd4dwOOQgEmACKREiGESTLNIWbBJk2ZINiqLJKoeyraLKgZaL4gu6yrJKlFSiyGKVIAaJJCCCASDP5DEAh3TAxb3bcJtmZmd6uvufn+AXT08f6BeiC1BhtRz9qrZ2amant+fpnueXviEzGmst/oRa6i1KCZQICJPTtg1N0/Ij//P/fde9xqcuwQghUFJh8pQ0S3DOUdcNSiWkxtK0LU1Xk2iNFAJCYNZL3vSu93D2vtfSiYS19QlP/sFv89InHwc8B/sHFMMho/EIpRRlWeGcI0kSkiQhTVOSJMEkGmMSmrqh7Vr6vqeuK7SOF5rWGiEE1lqcc1R1hRASY+JjxItR4pxDaU1dVYzXJiil0cvLyHu/vPigrhvK8v8Pn+ovVnjvODw6QErJqBggCVivIWFVaSqp0MsEH5yPSbvr0blBAFrGZCESjVDx18QvH1vzSudDkHjnCIALHmctbVOvEstJAaGNQSUJepmAetvR25NOOSadPMuR4iT5CbwLON/ivcc5FxOREKTmleSjTHIHTvjOh5QKj2bhNMq1KN/RB4/tQWmJdB3eSYJWJEKSGyh0Sh8czgmUEHgZyKSgCw6bpvwPP/Nz/J2/9ddJs4wEhUoNouuRtqH3gswYhLcYpXBeAAlSBXrn6b1HJhqVFlhrsT5QdTVCt2Q6Z3d7jLcNwTakZgDSETw4rxAolJLgOqquY1HVqCRhYzxG1j3eWgqT3ukj/5Li1CUYCLR1idUGYxKcj52Bcw6TGtIsYxyGeAKNDbz7Xe/gwuveyXPXFjR6jZvHcx5/8jPMujXqwetIF39Kkk3QStF17bIa1eRFhhCCjY0N6rqm6zq6vkMnmsFwSNIZrLO4ZfUKcRTivacsy1WySdMUHwJtGwnBw+EIkxq890gpSHRUhD+pZpXWUW+l6zDGrLqy0xRSSgaDQUy2gFIaryS97SkXJW0bX6eiKBgOh6Q6FgInyeCk61ntJ5fjipNRZ9O0SBmTgpLL5BNiYk8STZZOCCGszl4QOyXvPUIqfPDLsU1McFolr4zplv+n9x5CWHWxJ4VH27ZUbRO7ISHQrY7/9tRFoDApvbWEoGmDIPQWFzyhc2RZzmw2ZzjJ0SJQlZLbboFJU/DggiV4h1KgZcrTL9wg1Zr/6e//U37qxz8QO96+RgtPnim8SKjaDik9EtASrPD0QWDwiNTglcf2DVoKyCBPM2QAXA1tjUXgg6Jpa+LcTeNCjw2Buq5YNA1Ga4bDIUIp7MnIzmQIfXde1Xfns/4ywlnH/t4eQWrSNEVrxWr34h2EQJblmCzjrW99E+96z7fwmctzPvG5T7GwCdrAhd11bhzN+EJrefi+t6JmVzmezVFKxQsrz3C2RwpJ39vYwXzRZe9DIJUKZS1OWcJyH2RtlGXKsgyl4ygnhEBdVYQQGI5GIALWWbq2xTlH3/V47/DBr/ZEXdcRACUlbdPcucO+QyGlpCiKeOFLhQiBnkCmMgbFAGstTdNgjEHrOM4AUEoud2FfNOoKAX+SHIRAa02WZVgbP9e1dUxKSiFU7DBPkr0xBudcHMMRH9ekZpWcnHP0fU+/HHcmy+IAwNn4vmiWna6zr0hz2eCX712NWRYipy0EAeEbdAh4ZVDeIVKJryvSbIwWgbWBJuAgSERwIAzOdei+i7sNqQm2Q2nNhz7yOK0HleT8X7/wK/yt7/8uFh2I4GmFR2iFCwKsRksQiUA4DwGG4wJhWxwBJFjbY9H0NoDwkCT0AYRM8Rasc3jlyFWCsSXK9xgdWBsbvCqWhY3AC0cfAgFH29ydheKpSzCJSTh7/jw+CPxyx6KURGuNBLquQycJt6dHvOfrv4FPP3+Lx598mr3ZnMO548w9D3Lx2h6ut6Sp4fGLFY/pOULEijiEwOx4hkkN4/EaSrWrLkRrTQieZNkxSSWRy6+FZRUaQlhdPkoprLWr6loIQdO0lFVF17bkWYYS8XIjsKzmAn3fk5gECUwmkzt63ncinHNUZYMNlmFRoIQkLO/guPdK2FgvkEqukjssE4sShOCQIokJRcT3CTJKYJ3scZRZdi5mmUCERGmNUnKVnJQIEERc8kpWnZG1lul0uhqFCq34/8IF4usNMoCSCUlmSIxejsYEXdvRNi23jm4wr07jGDQglWCgNUp6rPU4BGkq0H4WO8sEROJw3sUCQAiQGYlIsU1NL0CGBIRjZ2Obm9ObhCB5+dBS1h2FkSQKemmorcI6SZpqeu+pWkfTBpSCXjpSIVEKikST5zm9cyw6gXWArcmCQ4iAVQKRBAge38wIQeGSDI3EB4f2LV6nCCFjIZoIbHAocXd2qacuwQghGYwm2OXi3JgE72Ml6R0IFG3X8dAjj9AHjdKGx3/jF/EqJ5g1wtHLeK0Yrp3FB4t2M+zWAzw0bnBe0DQNeZKQZRlt21GWJUop0jzO10MIjEajVSIwOiHPstVlI6XEe0+SJK8sqGWsjE+SX5amDPKc2we3OTg4IC9y8rRAKUlRDCiKAmMMbdPQ9/2ffyh/4SKCLdI0jYvessQuz7YoChKT0DQx45yMmk7OWmlNnmUIHT8+AU988cjMObdC7PW2W468QPRL0MXqLnBoZVBKY5KYHE6AAaPRaPl6R9ShWI7IxPL1P4mT90wInq5vV8tgpSR5nnPGnGX7SxIVv7vDB6gZUnUdOIsSis56Zh0MTMZGnjGQPaNiiEg0Ps6f8daRGYPPMuZ9CxjSBPYPXwZpSIxANI6f+D9/mf/lx78TETSZkaxngtZC2TS4AJkU5EbgFNi2JgBaSTohUNRoJciNIc0SrBvETtZ3JD4ghcY5T59kOBKCCAQpsD5BY9G+RwqLFhbnIwjp352l41c2Tl2C8d7T1PXql71tO6SUJEmKSgWtDaj6iO97/w/w+B9/hn/6zz/E00/8FqO1TYrRiAfe/nbOveYv8dRz1yBR7O0dkLqKexJD3Qt2d7ax3nE8O+Jwfw/nHMPhCLmIOoARgugYDEaIENFFWi/n6ELgQ6AYDKjrepVsqnrBbDajXJQ4F0c1w6Jga2uLtfVxHKPVLUImBDx1XXJwsE9TLTg+Pr7TR/4VDyEg4FmUFakxjNc3IHjmiwV7Bwd0XYf3ntFgxMbmJsPhAIjd63Q25/KVqygpGE8mnDlzhjSJI7HgQxw9ftFESiBJlAIp0Mv9zck+5eTfO+tolqNKrTVSSowxpGm6Sl7ee7yLsFetNELFz/e2x1u/3M+B1oZimOOdwzrHrGk4Ov6SnKDv6hAEdDejF5Kq7VFmQJYnbOYjdCKp24bWexZ9RZpIsC2htwCYPENrhXUeK3puHy2YrG8xnS8iAGhssMc1v/7h3+cD3/XN+N7RqIDvPblRECxKxde97ixBWpwXZCZSDxwS4VmNVQdFBBP1XtO1J6MuiQsJwTmkiKBqIz0iBJSSBC3RSIzQOBtiRr0L49QlGGCVXL4YcgrQW0eWCKZdwc1be0yrkqqpWXSOg8tX+Ojv/i5Wal64MWNvf046HlPt7fH8pSu85f5HWV+XzGZzymrB8fHRMpEMyPOc0XiyQnlZa6mqisFgwHQ6pe/7mEyWFXLbLpFD1lKWJU1T4b2nKOJj2d5SlguuXr0KyySU5/Frx8dTyrKMAIFEs7NzV8Hm/52EEIIiz8iIo8bgA3VTkyQJ9993H3o5ejyazjk6mlKVc8bjMVmec2bnDOfPnkcQeU5t09C1cY+mVBxTJssxmNaak9mb87EDJYToY6Ak2pi4j8kjl0KsupFA1/f0XYfzfjmq1SRKIxDYYOm9peu61XvhZL+X6ISmjm4XSimGRcFo9G8Tyv2LGUoEBsqitCRPFGBBgO1qZOdJsoIgsyW6LMF2lkEiaTzY1qG8RUvBxnDAWpYwO57ilUTKBCU1Z7YnPHFxn/d7UIlCBgg6wXmLJL7uPZLOCWqbEbxFGY3vaoZFjhcSs0SilmWJMSm5yRkXYwCsaynbnq73CDpEEDR9QHqH8z2hVzgtyU2CMbHDvRvj1CUYrdUKYXSyCPfeR66JdxzOeuZhwgsvXGRjY52v/dqv5qH7dvnqd72LYn2L28c1WTFE+B7hO9pqysbmJqUckVc3uT0tKYxha2MDIWPF2vcWay3j8RilFG3X0jY9x8fHDAYDkiSJI7rl/kUvESN916GUYjQacDKjL8sS5x15mnFUH6GWMOi2bePXXExqa2trDEej1WOdpjgZddmmR0lJshxVhBBQAYQPGKVZW1tjc3OTQZosF++aftkhiqDQSqMSSbcsCk4IuH3wqwSTpXn8WOm4N4HVWO1k/BZcoPd2tV9bLfSTCDE+6VQlcgVTziSr5BiI33fys518XwgBGeLS/7SF0ppsUOCBrUST6LhzrKqYaKqmorWGMkgK4UBmTD1gLUJCYhOGhaSpagKBgdHMrMdozebGhBDgTWsZP/1PPsR/80PfhQ2Kqm3RUuBDj5AK+g7vNQ6JC5qjsiPVBlcHGteTKIdRmkQJaFus7SnSHKTAAip4Eix17xBSo7AYJbHa4NB4Kdmft+zNKsq2urMH/iXGqbt9nPMroqNzkVcgpMCTUpc9yT1vYf7Ss/zLX/11/su/8V/wzjc/hn/b2+hszwuXruHEgMtXrmAGOQe397l2/RqboxHPvFjwutEtUCMW5ZwsNWiToJRGKkXbNty8WaJ1vMz6Pla8x8dTpJSkyz1M3AulS45FHLEEH0hTg+0taZqR5wXOWgaj0WqxnyRp5O0Qf/kigc9xGhGszlqOj48xSUbTtrTTKc0S5m0SteIVSa1BJPRekigJeJzrsM5xPJ3S9T0eSJd8JpNksQBIJMELfB/fQ3oJcxZCEAhIfcK+lnRdQ9u29H1D07Z0S1j6xmQNrdWyMzIIAR6PDwEpFNIFFCC0WilA+GX3E+f5DtdbympB3fzb/Pv+YoZ1gb1ZQ7AdidHkWpEXivFoghCC8TiCKWazGYkRLFxHgiTPJJ0HXMX+IkNqRYKkFSmSniAEqZbsbBQgFE8+8zIv3bjN6x+6l7VhRt00BKGpGhc5SUlgoqDtHUhF78BaQCgaJ+i6BqECk1SSSUHrHd4r+r6jtRYCdE5RlQ1CKVKtsAQqq5jNj/AuKgsMdH6nj/xLilOXYCCslr3OOareklrLcXWD3Uf/Iz78q/+Mm7cPefOb3sJP/u//Gz/4Nz9IED3D4RqbmwlP/OnnmZZHXH/5Gkop1scjbGdprOTcubPMa0FvO/I8BwRVXdE2Dd7ZJTQydlBZlpEkyYpf4axDLcl93odI9FQqPo6HqqqihIyHokgxJqWu6wipNia+EZfzf+ccWZZiraOs7s7K58sJpTXj8QRvI2chUZLJ2lrkorgeqUSseLWm7Vq6pl51JGmWRqTh2bMEoHd+mXxACY02Bq1fWcIomcTk4CNXqbc9bRnHWjLExzTLDmo8mdBZS13X3Dg8wFmL94EiTf8MiEApRZaky5/Bxr3NEoKulKK3DiEFJs8xRcrkFFYRIXiMq/AICpEggKb2BLcgTfRq1xW0puwCaT6kq+e0KLSR9J1iO1NUTYP3gfe841H+zR8/hVSC7Y0xRZGjk4TJJOfv/exH+Qd/5/0IEdAmicRNHUiLOJFwtsFaR2JSBDAvF/R9T2sdISR4IVHa0yOZzloCkOuAFoEgNEIGNkcjtOzRUtI7x0hW5KNxhDQHtyoe77Y4hQkG6uWSP4SAERIXFtz7xu/kaFGxe+4Cb/mab+PWtZf40R/9UUgyHJrj2ZxnnrvI3t4Nyq4hCMHB4VFMDEKglKauK7r+le4hhIAUkjzLyTKDkCqiR/qerutomoYkSRgtR1ld1y2TjlhxKU52RGmaMhgUIKBpmqUkzRip5GpxGBFKkqJIl9/vGZ/C+byzlv29AzKjY2cRoK7muBCWiVjT64BWHVmWMVxbf6UDWULGg/NIIcjUUlpGRhKllCCX/CQpJMFbECLuT4whkRK9RPw1fUPftFT1gs5FNJ8gdlDb6xsrHkzdVjFphMidMcYg04yyqeOYTilMmmGUjt/T1gQX6LqG2rnIAj9l0fWWz1+5jdKK0WCI0bA90Cg5REsRL2Zn0SLqhHVNiVGaXoSootFYdKhZ39jA+8DTz/4eAUtnE/I8Q0rJfFaxu77GdHGLawvLQCrSriFLNImWuPIogjxMilaKpiwp24Z5B4WRtK0lZGtoJVG6I9eS9UFKCD2zsmHaemrnsb1H2IZcw9bagNaBE4rMWDqn6TpHuEuv6rvzWX8ZIaVa7TzatqXzns2Hv5nLh3MOXr7EaOf1PPPsiywOXsKYb8LrhJev3uCZZy4yL2tUIqCVZPkQXTaMioLFbM7L128g7t+gt5Ed3PctUkKe52itcLZfkh4Fg8FgxZlpmmbF0j8h3OkkxWi94s/Uywq76zqcdwiIcFUfhTPVstM5IfzZvo9aV1LetcvBLyestRG6bZKVXI8XkTRpO4tdkhalgiRJosrCElmWJEnUgjOvnFtYcpggJvETmHLv+4jychHllWU5eZ4zGA6RUtLadvU9J4KIMR9FMmxELyZUbZQWMtpEjo1a/hGCfrnHUVKCj/ua9fXN1fPqcaeSyZ9ozbmNMT705JlifZiR5XkUn12iMSOoxjEaDFBJAgj2D/bo2h4nFTLNuXh9j4vXprzvm7+Rf/gvfhMlDYNigBeWpz73NLOjGhMsv/prH+Nvf+A7WDSOeQjYeYsWHaNBSltZEinIiyGbgxHd3i3q+Zw6JCTdTYI23PKQJ5K18ZA0MaR5zjA05N7TKElTCVovuHHUkGcJmrjnS6UjCIf7c0/k3884dQnmZAwxGErU6BH06EEuXnma81ub1KXl05/+Jda23oNKz/D4b/8ePss4Oppy8cUXGe9skw03OaxCJFQNxrz8/BcQvWW4IbnwwAPcLzXH85qmaVas77IsOT4+JssGJEnEwJdlvRx5aNq2jZeY0gwGI9IsA6Bdss0na2urrgcXxRGD9yilV5W3tS6i0JYjFYTAeo/R6k4e9x0JnSTUVU1dLjBJEpFaBAbDIePBiPEojjY8brm/SuLlv2T0d11H3dtVgl7M5xSDAUWer0ZpJ8t6OFGzfoVLE0KgqWta16/eb0VaoBONkks0mYwXCC6QKINRRE0062LCokclGrlU2RUS5JKA692JyKWg7zoWi/mdOuo7FkZLHrqwE0eHgAsBZ6Fzlr5v6PpAUIrdjRE6G1HkOXU1o0gNqZY0Tcf/88QnWcicrY0NLl97AedauhbQEiM173zTm3n2uef53MWKJ5+/ivKOYCu0rXni08+xu7lJJi1bW2dZ39pkNq04uH2ZqnVcOH+Bc7qlrmsSKZh3nqb33Ji2aFkzyTVrkwlt52hnMxySqq0pEqgWLTrT4CRGB9LEYO7SOvHUJRjvHcfHx2ydvZdp0/HUb/0y9z5wjp//xd/hDY8+DOlr8PUVUvsSBzfO8OQXvsDe0Yy3v+s9VE3CF57/NKNiyH7lGa9t4Z0jNwYZPNevX6cYjBkOJxRFQZZlhBBYX18nz3MODw9ZLEpGwxF+iQxK05S1tbUVAukE0TY9nrJ38xZ5nrOzc4a19XXaNqoCnMigrPgTyyXwiQTJyeeSJKGqTt8CWGvNPRfOYQMoKdBfJMevpMbayCGpu1gI9F0b1bKdW+mTaaXobOwo0+EALwRt7zBGkaiEsKySIWqdBe9xIcRxKXHHkyxf/5M/znukiCrN8os6okHwOGUI3nA4a5lWFdkoY12m+G5KZ6NqRJpGaZvgBe7E+iEo1sabd+ys71RYD7dqR7Adzlq0FqTAcJiR5cM4eh6OuXXjGldvzei8IM1zbNVifeD60RFdWjCQGmcbRonizNoGTV3ju55kWGD7QwKCnY0J+wdznvj0Z3nV2pBrTeDMZIRUgpIc3UN/OEOGmPxv3z7k3PYGi7YjkZLOOYZZgnGBtmmw1jNlSGlLtFEobdiaBDw5TW8jotRaehuYdwGJw6R3Z6F46hKMUprhcMBvffyPYFBQHh/zc//sd3j/f/J9fOalOXvP/C4f/G//My5d8mil+IavehuX96d8+KO/zf0PvomdnTMMZMdwOKBxcWHfLUp8PeOlF2YM188yKI6idthwyHA5Ljk6OqLrOra3d3DOkuV5rFRFRAW1bUtdVSv4tPeOPM/QWnN0dLTkweRRQsJ7jDHLxf+JlEwShfyWCgFCiCgZk5w+td0QAotygfWBNElITRQE9d7Tdy4SW53DEhNKLQTzxQJtTCQvzmYUeY5WKiYFpQjO0XvHvCqRSJq2+TN8qhO4uQTCUrvMer/qalh2PnIpORNOjFwATM5vfPwFLk1rpuWCd33913FBKny9wBioFxVp35G0sWNVSq86I+FfEUI9TWESxX1bY7SO72+TpiSmoO8a2qqiD5Z6MSfPUnZpcNbiwjGVb+jSggsbE7qRITgoTIoOjld/6zo/+S/+IBJeEYzHI6azF+itIATPP/zl3+fv/uBfIzcd06OWvmrwQpIqRV96jBH4rqVZTJkeHvCaV91H23ckclk8CoEfOLqmpnEVrk9YNJIeiTEpeMvQJEwKg1KB4BWzco6zgYC9swf+JcapSzAnsuy3Dy4RbgSSyQbve9938luPf4pveft5Hnr3t/L8xYs0ZcPa5hrDrOA1ox0++ZlnuX7t87z7a/9z9q9f58Wnn4Jsg67v0YnG2pZ3vvPdpKNzUa15CZE8OjqiKAo2NjeQIo631te3EErR1DVlWTKv5thllZqohLwokFKyvbUdSZhCE5ZGZFpHpFi7FLsUCIoi8npmx8csypJBUax4E6cx6qbh5es3yIeDaHNQV8jAElAxWXV5tq5wHowxnD1zJvrynHjvhOgCF6SgLBckOonAkERSdw1Be3rbEYSinJWEw70VWONkmiGX6tivJKKomu29R+qMgQ6QpfzsLzzJS1cXkDVMNjf4wlOfZ7FR8I1veZjGG9bPTvDlAoHFeY/tquXjSgJyNSY6TSGlIsvyla2CkppydpvZfEbTWbog6a1FJkOU2aSpD0nwyOGIoYCmmeORNM2CFkOvIel66BO6vgchOJ4uOHtmm6svX2VtnHI8DfzWk0/hQvRjyrKUzfUNDqcznPcsyjljack1fObpZ3j++i3e8djrGQ8U1ewI2zlGkyFZMcQ2LS5YsuBJcCSuRQRoakHdSBCSzcmIc+vrVH2LTrI7feRfUpy6BOO8Z2N7m8Pr+zit2DQ5k3TGT/1338vBwQwI5IO1CPftWo6mh1TtMYPhiKbt+Ll/9DM8+s6voWp6FjefQojAZH0L4yWf/+zn0fllyqrmzJkzrK2tsbG+TV03HOzvoZPIIJ8vplSNZTQakWrNhXvuWS77+5WopRDRNVMKhdRiaSkQSYDHx8f0fb8ibt64cQMpI9dmfX19udyM3Bl5Cpf8Wms2N7cZ5MVSwFJiXVRLmM/nq45jY30jKlYvV6hSSrI8j0rG1tI6x/R4yvH0GCUEWZoSEk2yVIBQSmO7nvXhiM72uLbFdz1IgVTRh+TE+O0kqZ10qz0dVab4xCdu8vlPPcn8+Dbb976O2+0BG2vbbLz+Hi7PSnYKw+15ibaWxNVL/xi/BKpYeutP5ZK/aTtefHmP7fUBmVYrO4TRaEzSdXRtjdOK0rZ0ZYW1PXXjsHZOoiQ6m6ASyISE4OlkQSsVPtSUZcPmuuPcuR22ttZ47JF7+cTnnuHp8grn77mHRdkiRfTnIcQOd2M85sJYkyYSgeCMV7jBmFsHU24dKoa5IgTNrSu3yBLFuTM7jAdZNBOzirr3UQ5GpRglsE6wP60I44wskVh7d6qin7oEYxKNwINwHB7s86bHHmVnXPD0Zz7D9rkLaKVI0wzrHK2TTMYjplePcL1H6JTz5zdo6pats/cyGk042r9CWZYMTE5dd9TzaiUb9LnPfY719XUuXLg3ku7a6LQHcM/5syipODo85NKllxiNxqyvb0amf9su8fWvSPm3bYsQgkXTLZFpcUwyn88jX73hzgAAIABJREFUz8IYhGSlYWatZbGYn8oEk+iE9bUN5NKuGljCxCWjkX7FkdJ7AgKpJWUZu84Vb8g5eu9pbY82CSZJCEsXydTkpGkaz13LFcFSSUVvLf0S6cfSfTQsn0eSJBiToFRC5gQLYfnQL36YpnkWUd/miU/8NufuvQ93/X5y834ee90F3r21hZxMaNqKGwcR9CHwqzGpSdUrvjWnKEyiOLc5xrqeumoRUi5HoYFBmjMZRfM/h2BRlrR1FRWM3QglYgFQliW9yWM327e0LkF0JX/01PM8+OC9K4fauu954PwO0+OG8+fPU1UV0nuCdVhbIzyofo4pUuZVBzohySck2RhDiCRb6xFmBJOcp557jj987iWUCKTBceHseTbXN3ju0kWCzHnsta9G4wlaU3cW72GYmzt95F9SnLoE09Q1n/3UJ5msj/je7/lO2nKBCB3Xbk755Kc/wXy+QGvF5uYm48kmg2FBajTb21scXd5jPJrQBUldL2jbiN5JkoThcIgxDcVkTJpljEYjzp8/v4K/6kTjvCD4wLVr1/jkn/wpg+GA0WTCaDRifX2drutIjeHw8JDBYMDh4W1GwxG965eAAdja2lp1OLPZjLNnz1JVVRz/jMdYa+M8WuuVWsFpC2stV69cpipnICWDyZjcpDjrljDypRukVrRNSwgOrRTFYMBkvB5Jqs6SJFGQ0tkuCqMutcGSJLoLBuc47OuotCAVeZqhhGCSDqIS8xcld8krKsxSSkhTLn7uGsibDAdD7nv9Y/yV//g8733Lm/ngB36IZz/7Kdof+CDf8trv4XNPfZ7XvPoRtkZrsYNJJFLIpUWvPZU7GO8cdT1nc3ObLCuiiKwPlFVJU9c085p6UdK5gLWe3rlozWEypIvGf8J7EqMIQVIMCwSC7//e9/HPf/33mc5uk2dRBVkKgdEK50pS1zPIJThL31uESTBaYpSm7lq8GVC3HWniKY9naBOFLou8wNclqdI8cOE8efEwoW+omp50uMaimnPmnlfTeMefPPUsk/UNEtvwzje9BqkMs/LuBOucugQTgCzNeedb38zx9JC6Ltm/+Cx5UZDlKVmeEjwEL9nfv8XRkaZIR8zmDUmiOVg0NM2U7e1teu8pshFOadLRiL494GDhyIzk5s2bjCfrDIdD9m5cY3f3TFzM1xWvevAhzp87z3R6TEDQ1B3PPPMsDz/8CG3Tsr29vZQg0Vy+coXRqOD2wT7D0Qhr+5jMkoTxeISWmjRJSYyO8ijG0LctSgictRhzd1Y+X05IKdne3ia/cA6k5PbxlKaqo7VxEEtjN0GuIncoIsR6ysPbXL91kzzPWR+vk6ZpRO3lKV3XohNDURQMivGqC9oJFq2jL3tnLXjQRNa9FCJK/YSAFBFhhoiqy1ZIPvvpp5mXB+xun6cpGy5dusUvln/Ce//GD3Pz2mXe8fbH+Kmf/mne+ta3sXt8jFmO5touQtZtbxEnopunLJIkYXv7PN7W7N+6vgTHuJWgrPcBUwzJRU+SjdEykKgUH15J8k3bYHtL17d0dUNiNGnfsLk2QPpX5H/qpsHjefj8DpMiwRiFFgW9jbJTwqSMipx127JRLuhDGvXJnMMrhVMZBIswCg/0PmNRNaT0rI1H9InBuoLQ96wZyeYjD9FWNbb3vHT5BmfO30OR3p2v8d35rL+M0Doicfq64/bREWW5IMsz6iqOlvI8x5iEs2fuxYco61EvGoSSrG9v0FrB2oakXkTF4iTN8DrBOsFDr32YmhFpImnqhr392ywWCx544EFu374dTcSk4vbt2wwGA1796lejteHo6GipnWXI0hzvHV3XMxgMuXAho6rjboglV6YsS8oTXxkVZ/zOW9bWJ9y+fXvpbBnNrm7dunWnj/wrHiEErHcsSgcEdFAMB6N48RAVkmfHx9y8dW1FehVCkKcGozTlomRRl0v1hAHntnYZZBll01AvgRlSSeqqJk1TJpMJWZYDgbptWJQLgg9R5mUpxLixsY4WCodnWpbMDm/xyWdfZKjHCGVY9JbtRDHOMrZ2dnjdY69n68wai7Ji9577eOLJp9jZWmNrc4DRhq6PXVW3HKOetggIetuS6JT1nQEBjW9qop6cY17OottrUBwf3MIBO7vnGI/HUYC26zFZQddUtJ2m1RrhHWd3d3ngwhQXHCIErLVLjx+PI+HJ56/y3ne+lt4HRonmcF6R5TkEh7dQDIdY5xECDqcd+cBEN0vf4+ZzqrZDeU8+GKHGZ1F5QZgdkMuWkA9o6ymEQD4Yo+QYrw3z1rGY3p22G6cuwdR1xfT4NkUxZDKZMBgMkEpFQp5z1HWNc5aXr15lfXMNYwzF1gC3PeawNhzcvk2ap9SLkr7vyYZDbIBivMFTn/9jrCqYjMfs7Oxw73330vc9h4eH3HfffWitqapqZS729NNPk6Ypm5ub7OxsYW1Pnq9jvQMEa2trVFVFulT7zfKcsiypqoq+b9nf38dZT5oaxpMxQgjOnz/H0dGUa9eu0bU9586du9NH/hWPQKCqKrx1K5tjqZd20n2P0Irx+hqIuMPquo6NjQ289wyHQ86eORs920Ogrmv29/epsgyT5wwGA5x32N4yGAyo+44Xr1yO8v/HR1R1Sect3juUB6UVRVGQmwLpI0w+z3OG44Jrt/ZQWU4XPITArRs3MMvR5nBQsLU24cEHH+RjH/sYZ87dw4ULF7D9HK01xaCI49C+P5VL/rbreP7ii7RqjBqM6euGlIZ149DFgEQIhuMxx9Mp6WBCLw3zzjLfO2A0yEmTlL4umc5u4zq/dLx05FlC1/VYH7UDQwgkKqHVCq3gX/3mE3zXN7yd3lnKtmVrsrZMWCXoOBp1zlGWDU61JFowSA2+k4REkpYVrSxQa2fovadranxT4myDkjWD4QZdSLDB4XzAEMB5ps3duUs9dQlGCMHR9JC6btnc3ERJhfcBZz1JYpjsrlFWM/ouMJvNolNhgOFD5+iKdSpruXH1MrkgsvSrhrXdHa5cu8lXv/211I1ltijZ29tjPInjrN3dXS5evMj29jbD4XD1XKSUVHXJ7KVj2q4lzwqmW8c89NAjJEnCfD5fOtpFb/drL1+j7drYuiMYDoc0TYOUksPDA6y1XLr0IpPJBK01eV5wPJvducO+kxEcLtiIEnOeRCRIKZBL2LBHMMgHZCZDDRI0sDYcopSmKluU6BisDygGu9S1RacgaUFXqK5Aa8DVTPJ1xsOo9/awfAilJFW1IBCid8uy63FLTkxm4lguKIs0I7zyDEzB8fyQNBVcffEyui25/JkneelT9/LUn/wBb3jDGzFKc3g0596ddY5mB0iiNI2Sp4/nBERzPrOOci1Fs8egyMA5yrrGd5adnR0OKs/MD6lqR9tWMdEPBtyeBnBzUrtgczQkGSW0TcVifgwELl66wtro1UtF7ID1lmADZd+TpAlXb9xCa4nrG0pnSdKlFYSUCJHQOQfWsznKcULStD1NWUY/oQCpb+n2LyJNjvMKLxOcStBG0Fez6HIqEoKUCFK8tf9hRHY3hVkaQR0fH7N7Zpeu7bDOIpVgUc4ZDIbkmznBR2HJ6XyBrUvMYIKwHaO8QIeOsmzZ3d2hrEuEaNibBrIiYzyOpkJHR4c0dUlqUiajETeuX40qsMbg+h4hJCY3ONcgCdi+5fq1yxxOD3jb294Wx29tS2chuI4sNSAlNkh06KmrmskkyshMp9MVL+bm9evMZjNcCLzxjW+8w6f9lQ/vXJTVWcq2+KVx2Il0j1IKrRRr6+vREiE02A4EKb1TjNbW2dufcnit4Q8//jHKRHP27C4bkzH37pzn6c89wfvf/900dUmWtginUTqj7TsQMNrdXcFm19bWlsTImNhkkJGprT3To2M2z25HT5Odbbzvme3fYHHrJq0v+cAPfg//43/9N3nve/8qDzzwMDIIwtvfwP3337si0mZLGPRpi0RJdgqH63usF1RO0nSGFk1ihtxsIhBDpQJDS14U9C4Wks5blDRUyTpt7cg6WDMJOzvnqduKedlFsuVyb2a0YG4t22ubvGT2sNUxlkAvNEZKhDAYEztk4SC0zdKOwyOUBBHQ0iD7lmKQ03uPrS22PCZJUlLvaKXB+RQbMnovcAJS25MlFV7F++JujFOXYJTSpCZDCIVSkuPpMZPJBGljC5rnOW3b0nYdRVYwGo2ZbGwxDQOeuXVAVcbZ/OJ4gZCay5evsLG1y7nts6ikozm+TYUkyzKKoqBpGpq6IS8KhsMRt27dZDqdrWDGi2vx49FoRJYp+r6nn8/4vY9/nK/9mveQj4e0TU1VK/ouKiVHaf+C8XhM8FFpeXvn9VRVhZKSmzdvcOHe+0AIFovyDp/4nQnnHDqJ5EhjTLQ/kK84URpjojWxFMzLjNorLr90yK9+9HHSLCdBMptdop5OGSaKg4tDblYtj37dt3HjiWf5mZ95H4ien/3H/4A3v+FBFBZlBW3fMS8juvBErucE9SdltMCNn5OcO3ee2rZ4Db1zNFXFZG2DHEdajHngvodQwvORD/8aH/ud3+cnfuJ/5f4HfwItJa966CGU1OTZ6QNxQFTMbuqatu1oVEGPRaUFXgZ675BLpWqTZjjvqOoSfKBVCi0TQhInCF0I9D00vSKbl6wPBIuqpel6us4ipca7juF4RDmrOH/uDFePel51bgx6TJ8UtN6TS41wLbI9xqQGqTRlSHHZJlKCxlM2LVYEXD9HyBSZa7Tw+E6gTYbKd0iCRdmYTLq6Igw3kN6T3KVIwVOXYIQQjEdr+BBJi8YYfAgYY6Iu1ZLJnWXZii0vdYqaTFZw4LqsUNrQdS3bZ87hHOwfTFEPvArVH1O1lqOjI7SOvhQBT1XN2dk+w/33PRi5LUtJF7s0P2u7Fmuj86UOCi88H/6Nf8VrX/da7r//ATZ3zqAXDQRLMz+KjG5rUTLK/Hd9i9aatu/ZPXtuRdQ8c+bsnT7yr3gkiWFzcweJJMsMJjUEKcjSjL6Nr3GaJ9yeBj79hUvMOk/XweHtKYNwjO4WjDfW2dl8FdpkyOE2rpoxmFX87od+gTf/pa/hB77lW2n29vjXv/cn/Njf/T/473/sv+KvfPXrGCoQ+MiFEdnKlrswJwTapU23MJBA0QVsPcPWoEJP3TssirLuGecpeFAIvumbvoFHX/8o3/Le7+B93/qX2dnZ4YM//MPIwEpx5jSFDYLjUBB0HCFJV0F1SCo8Ih2iKZBeIqUlcT2JFgSV4ZeHJSVY1yEQSKD3DmehPlowLKKKefA2/i0SEh3o/YwH793hH/3a7/KTH/xujJ0jXE/voFt0BCkJ0lB3AZUkFFi66irSC5JEYq1G6GguZzHIvKDqOvLCEazD14eENHpFhZO9mgsEJH1S3LnD/jLi1CWYaDY2wAcXu4umjXL6UpKmKdZG9FZd16ilm2CqDa2Part1VUEIjMYj5nM4Opqytb3LcGOdp6/t88adgtaW0Xe96+j7Hq0V88Ux5aLl0de9nr7vGRQ5R0dHZJmkqio2Nzbi3D5J6Op4CeZ5ygvPP8vNg2NeuLrP0aKjq2Z84K9/N0ZZnHS0TZSaD3gWS8tklSw9RXyg704fD0ZKyZnds4jE4oMHPIM+JU9zDoUiLRLSNOelvQNICg72LvHyCy/z9nc+xhve8G3cuHGD42nJ7z/xae5/5A1cfvIjfN17v42b12ve/c3fzmJ2i356RLa+yeTcLj/4usc40vfykU9c46vecB+bkxEZHW2QK86NltB1Pd5bmqbjeiOwfY9IPNXsmFQPMJnBB0emE8zSu/3k50HA7u4mf/jEx/hr3/4ddF3H3/6RH0GIaKl82iKRsJs7jEmxVtJ24HvBvLFIpQnVLOq9hQLKaTSGS1NckBA0wWRIElxw6NQAGpUlYAu0lqQmo+0sKkmQIoCCg8NjXveaHbYmW2gpcElG4jxaODCSJB9i+xbXVHRqnQUZWnq6ADdnPZKOtbFBp9FEMJGWIALWjAmyIxGB4GoCjr6zCGfxocUMttDj8Z93JP9exqlMMH3fkxcZk8mEvu8hgJBiyYy3zI+rJWM6quJWixk23WI0zFDGcO3iiyQyesJsbd/DwcE+TXOdnQdehe+us3v2PEZqynrO4eEhVVUxHEzw3vGFZyK7v75Wc889F8iLTQbjjuPZMR5HYVIG6QgbgCLn2nMHvHjpOgbHhXO7+HCWxy96jm9c5pvevkvbN4TpjFI5us6zNhqgCMjgyXKD1qdvhNLUNb/ySx/hM9dm7N57P2e3dnn5uae59OyneOFgyn/6/T9EdXTEcLTGL/36R8hlx32753n8ox/lC598kq3zZ1k/9xCvftUjvPUdb2ZtuMZRrXnX138VN16+RGENly9+gcmF18BMsHH2PNy+xHQw4uYnrrA7UpxZH/LqnRG4GX0zp+0teZrTuR4fGn71o8+jbEvXVwyLCc73YDSpyuiqit62XDmYsnnfDs47gvNRNVhJfvrv/xQ//mM/Bt4tK+3ThyJzPlA1nuNqjlAC2bXMFxUuGWGKTYLtKHxN2zbIbIzXCb53YDtEovB9B1qSKRnRht7T+wXCB77nG9/Nk89fYTzOyZXCCc/+jSOu3jjgwrldzmxvMO0kWgicjYrOMhlCKHAiJSQFSMX+wR697YE4EXFWYivPqEhIXYdzDcPg8OWMxnqSpZV6ryRKZxQ46MHPGlz6HzqYuyK88/Rdz3wx+zO+HmE5Jgsh4LFIneC9wwVou4YiTXHOQWfZOX+OUZpycHib4bqi9Y6dyRY394/ZeuTVyPppjtoGnGA0GjEajei6jsWiJM+jVbMxKfu3riJR7OzsYFxDPhox2N3g2StzPvKbH2eUSg729qNHzNou2w/ez41Lz9Ld+j2GG+e5Yi8wHHvOqD0Wey+SC8f+4RxFhOeORgOy7O58Y345kWUZ7/nat/E92w/y9/7JzyPP3kexscYjf/n7eEeR8aGf/1ne+ejDfPz5lyn3Z7zq7W/hYH7I7NIlhqOUjcGI9YHi0uc/QXn4EvViwfl7HuRjf/ArPPeJP8L1islWgXMfZbSxyaNf/e20AbZ211DGcJAkvDwa8tlnbjDaGpGPc959zy5FqBAOkj7l80+/TNW1mMTQdmXcDdUdQrtIyPWKf/07f8xjH/iry+KHCF11HQ+9/mF+6d/8ClX4f9l78y/LrurO83POueOb34s5MjIyclKOUipTEhIgiUlgMHS5bWzsoheU3e52G1yu8liuoXtV2W2v6uVyDcZDFV1tU2VsMAUWYIQNCGE0IKFZqUzlGJkZGRlzvBfxpvvufE7/8BLV77CWcuXK+v4FsfaNd/fde3+HkCzNbk2acm64FkpMblGt1hBujNQ2jjQ4liIWHkHuoHMb0gEELZQAIQ2kfZSShO0QWWygSiM41Rq50eSZZqY6wp999Qn27BrDIHj2pYu0NltoNXSyPrR7jE/86Vf5lZ/9ewjLB7cOyiKLBoihSRBSC8YbYxgpyfIcZUlsZSGVRZprUlEkyzLCPMO2h4QCoxNy3UVJgbAK9CyHtfV1iipisnpzTqm3XIMRQqAs9bqK+3site+lS9q2jZSwsrI0TCE0BtcpcPXsWSrjs+SOYaAzOhubTE/PsNpqUms0UK7H/olxetspJbuCpbtobdPrdonieHjgdYZ0Y8/ziPEI1QzkmrHGbgLZwZQrPP+dk/TDLXrr82xEgne/70foRcP0yovnXmNx4RLTVUXWX+PJqy9T8yrk47v4wP1vphRfxu1noLPrKYy3pspbG0PRr/NTP/phzl66iuv8EeWRKQ7cew+TM/txsoSHP/0pyvVRhFvgwmsvsbW2Mkyj9MssLl/mwpVLOFaRQbBGPuhx7exJ9hy4gx/9+V+iMTqN60CtUKcfD1i5Mk+cw+hoDZ0bTJYT9ns4lSrri2vsmBrn6xttjuyeJO9vMjuxm3a/TalSIo0jMp1jY0jIyYymXK9SwSHNFS+/dnKoncozpAW9Xp/lzZWh83MUYSuXubm5G13yNxyWHNqzaJGS9tchDCj4HpZfwM075NstlBDYyiYnQVhgOQ4JkhyF0Qav0UD5JfI0IexsUaiNYkjIMk2WBkRpgkeBt7/5dp558TwrqxtDFwDHZ3KygXZrxEmMyjIwoPzScJrMYqSALIuRwqFcLBBF8TA+oFDEtp2hhVSeDu2DLIeCX0Q4CqcwRhJsI/MINxwwXfZoxxBY1Rtd8u8Lt9zb53sH/O9NLa7r0u12cRwb13WH2pTBgEZjlGazSRwnuI5PsL1I5k6SJgOCbgfH8xFS4HhlpI7o9wJS3SKKIkaP7kdsBCADMgFO7KC8ItL2ya061QNvRjmS7a0uI+Uapy48g5OlXFq4yiCOefabX2PXsTdzz/s+RLDZQnXbnFu9Qr66xOzsbTSmx9leOA2iyErzGuWgyaNhm/e86+3s32vY3O4TRRlxv0uib0564w8CKQWZE/Hnn/k3PP3cZT75Xz/Pc9/9W3Yd3M1Xv/4I7c1Fdt5+gmqlQb3W4PlnnmDvwaN0+wG+tPG9BrZoMjo1i1uvM//SY5T9EtdWLpK/IpCVWVyvjOsVmZiZojWQmNgibWWMTlRxvByhLFLhYZyc+cUVgm6b9VaHo4fn+MLnHqVYcGltrOM7DmkSoxyXKEtQjoPIDcaKaQUp2AK74DDohaytr5FlCbiSOI1JyZBCsbi8eKNL/oZDYFBpF0s5SFXB1CsM+m3s7W0ouCjHIUuzYbSChmwQouwE5di4jg0GwigdmpmWxygVKiRRQJYMiLot7r3zIC+evsK77z+Osm2O336Q0VqVtWaHfXMFxuoVoiSn7FnEYUgWR4ikP/QdlBKdRTiWjU4jMgFS2SjPI0wTZJ4PTU+lyyAd0Ot32dxYxy8WqFTKOIUKeVrAK/jIXFPobBNFyY0u+feFW67BfE8z8D1b/DAMh6FhSUIQBK9PNFEUMT4+DkAYDBBC0hgZIclLaKEROVy4cJG3PvAAV5bWII5RwjA3N8fmepcd9cME62ep1EuMHNzHZifgzIVFfvj972BjbZXm1UX27dvPhbPPUE5jLly9CJnGcwu87x/8BqFRzL/yEqtXLlGp+8zUqrR37ENnAa2FeUqjOyjonKNHHuLa4iJxlvHyyVdo1hQjjRqTMxPklTLZLWh2mec5QT8gcmxOvGkXv8KP84cFlye++XVU1kN4FkW/jFNoEGUWtx26g2ZrkzSOGGxHVKdmqY/sYquXMFrMrlvjp/iFIqvLyxyoHeDK/CVmdu2g0y0yGAwYnSyjZERzKSIa9BCOh/KrNMoepeoYSIt6tcyzz77K0uVN2p021WqNzlaLkldES0muLMrVKiLLKBcdBv0BzVaTJE5IshTP9VHFIlEckykH5UuUVEhxy/2MQVnoQh1pNGESIQcptmURRaAjg1PycQuSimPoddoIu04mbTIgMlCqj2KTI9tbZN0mubLQaKRTpDBW5KF37OQvvvgVBII81/ieYs/uKTrtDtu9kNnZKT7xn/6Ef/rzPwFJMozTUD5epYJluQgBUX8LK4sh7iDzEGUVwamQaINODcoSFDyXQsGn57hYQmJyiHU2JG4MApQQFIslsvxGF/z7wy33n/m9bA7LGnp6fa+xCCHwPI9+v0+WZVSrVZIkGeolbJtGvY5j24RJMAwtu+4EcPHCOSyvQtGTLF9bAGlhMo1wC8jiLHkU0V3vEnSb/PD7f4jl1TXiRFOdmOH8lVU6cc7G/FVM3KTbTygUSlxbWWfvwQPUXIuR/TP0e5ssnLvA//Szv4wtDN998gl8O2d1s83py2vs3HWIsR272Vq9wktrGxwUCSvXXqQ+OsHo6K0XpzvUOhVwveGNbc9dc3xg/XbCjQ0uvvYKRrQZnZrk3JmzuKVRBp01iDpMjY3QjiRF36fbD/EdycryKaT0MFlAoG12zN3HVq9HoVqjH2ra587hlGqYJngsUK95bKx3yCwPpyJROWysbWBJi37rLN1Yksmh2WLQ7+NaDr2wg+f5uI6FEppBOEBbglo15s4DdwxV4koRRRFJkhDq7HW7fmmgWCze6JK/8cgS1GALqWxc4ZBKQ5CBtsp00gR3kGOlfXzHolSoUjIxSoG0FBJBHG6QG4OxFQQ9+ssdlF9CO0WU50MSc/naMvfcsR9pWVhK4TiK1a02S8ubvO3eO2mMjDHhWXSzAVE6jG6Ok4DMKaEcF2H7pFqSJzkmM6hoG8fq4tsFYi3AK6HzDN3v4mQhjueSdjMEEdIuYbwSidEYZVOu35yWT7dcgzEMI2aTJMF1XWzbvm7L0ieOQyzLolgokOU5GOj1elhSkaQ5cZKQBCEyy6lWqzSbTcrVGklvg37QploZwbZs/JpDJ1hH5AUGaczk5E6STsKTjz+FEgaERXtjnvXVJt3MYWTuMOLKMiaN6LZCMmGTlXZTHFxj8fLVYVriyD4e/twfozs9xmencYTD8bvfy9G7jvDamYtE/SZXL7zC6OzttLXP0cNzeEmH+flLN7rkbzjyfCjCW1vfGlK4jWF8Zoy33X8Yk/d4+tknePSLn2VkfI6R6V10l0OMzogzg1AeveYKuSjSjpcollzK1R24joVdrNBcukBmWzTGJri6sMjs1DRpew2jwBQ91jZjGpMNVC4Jsj7tzR5hOhTtprpLpF2iwQDSDN/zEUpgREqaDvfxV69e5eDBgyTGMDY+QaFcIs9ztrfbxMlQl2UEw9CyNCWNYjY2Nm50yd9wCCmpupI8TxE6Is9SMswws0eCJ3NSkZAEEcFAoRojpIOAGAvLr1EtOpiojw5DhKUoei5G5NgqIu+3CRHsHRnjwvklDh/ZBa6DMXBgz27iKGf+8iLvvv8enp5f4823jVLBJkli8jxD6m2CfkoYpthimDqqkxAjJbk2CAIKliLrd4fJlnJIk47DPkMTKIHrpBRVRJpCHA+IN27OFdnN6aD2A8AYQ7FUwKBptjbZbK6ztr5MlkfrUUXxAAAgAElEQVQYY4ZOvNdzO7JM43sF3GKR8bKPVBalskupUsMvFfFKFSbGp1AIfGnIBz3WL3yXfhBz7dJVTBRSbIwSBoqNhQsYqejrOmvnnyNPckbqNXbUS2ydf47a4bdh/DJCCorVMvPPf5WLly8RZT6D6l6OvOXdjNjwtoceoiIkA3LOvfjXPPPYl3nz/Xdx+dIyH/nfPsZ77z2CZ+Uot86B+z7AkbvecqNL/oZDa02706ZerzM+Ps7U1BSJidh35xyzO6c4euJNlK2cvLVAvLXIW/7+T9BOUlrLSwTbG7QHOWkYE8fb1MtVgkTQTV1anT7d5iL9lUUqWUxD5KRhQNTe5Mq5UzQ3WrTbq2ytbrEdGXwZIYnwHZskSWi2WqRJQhgMIMvodDp0+tfXskJQKBZezxDqdrt0uh1WVle5trREL+gTZ8ObQZoO0zkbjQazs7Ps27fvRpf8DYcBAm0TWxUid4TIrSDsAsp1sYTA5ClKQrlYxnYcbGkoj01QLTgIk9KnRDv3sFyPUqWEkgy1cYM+QTgg6XTAdzi4fweOY2PyIbuv6Dvce/wQE+N1pCU4u9yha00PU1ElWI4Ntk1x+iB3//CHcAo2lq0QSpBmOYMkIQhjoihCKYtypUzZU3h2gm87GGlIs5QozuglEGYJpZKPL2/OHdmtN8EYw7Vr14arL8vCsjzSNKXT6Q3t9237dbpyrVrEAEmaUS1X2Aq79Poh3VaTsZE6UqdcvHCOeqVIM/UZKwuEv5Pm4jx7ds4Sa8M99x5hdQP0/oMEsWHHkaOcXXocnQYkUUaUacI45vzTX8K1BfWxnUSDjNv2HSQOm1w4fRKhDa9+8U+ZPnqUk88+Rclx8MojBMmA1tJVHv2L/8zYSIUvfPFreDpi/8wop15+lRfPnWZ9Y/tGl/wNx3Z7m1QHuN4Yju3Sam1x9PBRwjBk8uem+Mc/9y84ePQEL730AlsvPcPS5cuM+CWMnWAXPQaDPhOzU4xY42x1+3S6Z7nt0FvIc8NmMMDKeqxsjtFqbrF3fIIgMDiOh+85KFVA+XXazS0o2UBEc62FaMywZ+cBli6eo1ZyCToDdJpg4gjhWAjAr9RQQtBcWUUUXDKdUCiVca6zEId2NwrbtoeuDXFM0XFvycAxW8J4SZFnITrL0TJFG4Pl2KQSbHfoRRZqhcSilYHpGspulbqVYHSbUMQMwoggSdFCYZRE5wK7OoUIAza3ThJnGb42OK7PYJDgF10KnsfhgwcIwog7D5XZc/QuepTY54eMTM8QWyVyDCNegd13Pcjf/Mf/m7HRBvEgGMadCw/tuPTzDHKDuu6IrfOhi4ilJFo45JlBqAI97eHepHK2W7LBNBqNYeBQFBH0A5JkmG9fLBWoVCtstbbIsqFNBECt2uDaylXcyRmyzMar+bQ2VqjWRyn6BZRlMVmoILIeWThAD5q0u3XIM/7wX/9fvOfHf51zl65QdTOee/Ir/MzH/wlf+dS/pVQqc/ubH6J69gWC1irGCKKwxyBUJGnG6mab2dmdzM2M8/DD36A4PTbcww8yam6Z6ekprl5bQmiD9JbpL64TlUpcLRrq9QkG1xY4cfAIj97gmr/RMDrnzz/9KV545SSf+P0/YHJyGmkkSZJQr9V57PFHOHb0Pmbn9nHlyhkcHeHXJyjWDjJz9E485TH/wpPExpClKY36NJ1Ol6KX4xUq1CybPEvJs4TV5SWKvociI2xv0trepjq+i2qpSH2kytaWx86DEywvr7J0sUsWd4njkEKhTJJESHvoeu04Nq3VNUbHx2lMTSKkoFwsMVYfI46Gcb9SSrQepjN2u12EELSum5zeehiKoIWARGuEcCj6QzKG9ApESDJjg2ujoxBlctAW/VQRxIKiY1FyDIWCg84hS2OyZPiiz8ItSBI+/Pa72VIW/X7A5bNXWFrb4J/88s/juh7CUujMMDo1zeyeOfxCDWVLkiTChAntXp8/+9wf8Zefe5hPfuLTtF77NML1qYzaSCGGZrdYxFGCX6mRZ4YkTYmiBKUE0rbIjUWmM7KgT3qTLptuuQZj2zaFQgHD8Dga+jFSWsRxBEJw5fIClUoFpSySNEUKQbvdRtkOszumWb16lYvzVyiUxwi6berjO1jbbLFvbppuWKZ9+TRTM7tYDiNmq0UiT/Hoo1/nwx/6Ef7qv/we90y7fOvRpxidmmV9dZ31UNHbWiEzDpbRiKxHuTJGwbeY2X0b8y8/xdWrC/z4P/6nXH7iL2mMThGnNsKzee3CJaZ3TrJzz504RQtPngaRsTY/jzcjMI7Nk08/fqNL/sZDCJ577gXuOnE3v/Grv0YURbz1gTez59AB7jp2N2My4U8/+x/5lV/858zMHqTXWWdmzx30RYkw9hmbGGF81wGiXh8HweLiOUxkcHfOUR8vMuitoLsdXLtAnqTEBLhZSDMZwzIuDTejt36B+U4Dz3UxsUfFc9GuJsoFMleE7RaWknjFAtL1UL7LZLFMlmQobZgcq9ENOixuBdjJFrXqCM1m87pJ5/DrNk0TLIbO0LcaDCAk6NQAkiRPCbYjEnKMVUQ5LgaNGPSHfl5SYrTBURJpeQw0BJmNFQZUPYPnuLiFMgIQOidD8Md/8zWmxkYIwy4//TM/TblcRCFBgpA29dEqUb9P0AuwHJ80G97G2v0+SRQxt2snv/gPP85zKy3uPv5R1l78c1xLDcXe1z3kvHIJLAfbSdEDja0FWWqQWqCTLsoYcqdw05oB3XIN5ntfgHmmMa6gWC6jlEVBD6nKu/eOIK+HVBljGAQBcRxSruwiCFNcAROz+7FLEywvnCOINSO1KotnXqQ2c5ji2B56iSRP4ezJbzKz9xg7Jw+gfIeqI1jtSEx2BTk+TS42CJpL2KowtBMxGSDJB5ucfGaB0vRhGhWbkSNv4+uf/QRTlTqyGCL9EbY3LrHv6P1YNoRGEDU3SZIBadgnzUJefOFbuL4z/Iq7xWApxU/+5E/ypS9+mUqlwvnz51lcucbo2AgT42O8dvoc97/1AVrtVaYnd1Op7mGkUUdFAuKQKxfnqe2YQ5d6GGOYKQrwKsSZg1OycEtTlCoe3UGL3so6tu3Raq5QsWp0koSFS+dQOsEeK1HyfJIwxK1VEXlO2S6j/DrumMXq6jWk0FhKkvZ7UK8QtDvsnpik1+thLEWn20WF2zQ3t6nVKhQKBTBDt26u3wuz7Najoutc09zcAjSuZ+M6Ho5lY+caYyRp3EY6JbAVSRIipE2WRxDlWFKR5wm2XYLCKH0pGASruDKnUHCwhOHqap+PfOhH2Tk3ScGvDn0JjUEjMBhcxxkG24UDtrZbSN9Fp5pup0tshsJZtKGzucCZ85ssL1TYO32MYu8iDLawlEAZMFmGMnpocpoZHK+CW8ixpMS1SyTpcP2p05vzGd9yDUYpRa3aQF4PGnPsIfvDcRxs2/7vmd65xmCoNxpEg5jzy6vMHN/DhQttNCVkmlEqlYl1SHtpEVGok5PTDSWqd5L6yG6ymTu59x3vZ+/YOP/1K1+i2+tBKsm1ZO+R4+goIO1uUz/8ZsRrTxANesTJkC3SqFfZ3lygWLXZbl5l50idRJYY9LoUgnMQp/Q2LrK+2WF2/0GWFy5g6wipY5I4YWpyHN+rsry0dIMr/sbDdTw+/9mHKUyNcNvhw/zGP/sN/vRP/wvvf/8HmJ3ewQsvvMAf/vEnCaI+eC7ViV1EUcjV8+dReUaxWkVHm2hhI60C1Ur9uuO2IAq2CTKDm9SoeVWqh8apVUscvv9dXH7tNDJNWX/tDL4rKIUxecMia2oqU+NQLlPyKoiCA1mE4ymSNMF3Mwq+z/aZszjFArnUWMLgOpJr8xe5/fAM9VKdLMuG6n3bJk6y61HZw7vhrQZlKaq1Kmme4doCIS0ELkUlhjHHxsZSNsaAKHvESQTGoIwmSCJcBXEeQpySa0lijxAqxYm3PsTS6iUmqwE6TdCpRssBUWbjWpI0BamgE/TJdYYAlhYWaW5tIhjGP9iux+TEFPmVZxgVFu8Y7YDYwG/N8/DzZ/iJe28nNooojdDawRhBnttoJVgbxPR7AcWCR8GxcVybquPg+zenYPqWazBCSAwCS9nEWQxCoK5PKzCccKSURHGMYdhslONQLjosn36RUq1GuTHO8ivfIU0VdmEUq1jGr4ywuniByX13sfXSaaQzw+xUlbndO3j67BWqBYee7bH/8DHOLW2wOH+WuLdFlKRk2CytX6Wk7NcdBuzcpVZvYERGlGQ4WqEthckTMmNjqYj2wmUOHD7E6sZVsjQj6GzRqPgUfI9OZ5uVpSuo2s3Jn/9BIJTk//zNf8XKuXM889LzPPZ3f8cv/MIvcP78BV544SX+1b/8bXKhKXglNpcW6HcCxsbHGCm5CO2Q5DHN9SVst4RfrpEWakjLIuxu41hQcBSKjKDbxy2X6SR98qzO7P5DuErA7YeoFIpsXL3E6lYLKROUJTFpQqD71CplPGnj2hZpEpImITqPmbptH1vb23TDABewLYHlOkRxzGp/DakUruvSWV/H8zxq1SpKDhlqtxyMIU9jwhziXJLrhCzOkCam6A21a7bl4tgWURKjhCIzmhQQRiEluGgsk3JtdZXlviCTkjwOEJZFluc4toVtuwwSF1vaJI5HEoUASCUQxpDmOSkhTuyj8xyjNUYYFp77BqlwsSuTZMoij0IGg23ec/wOOtrDCIXOM2wybNtC2Bppcmzfp1b00WmGTYIxFp3+gL5yb2y9v0/cgg1GYFsW2uSkWYK4/tt0HAd1fZedpime5xFF0TBcrNcj1i5hbiiXRjj3zLdRymCrIp0kYaSg2Fx4jbE7H2Dxib/hrnc9xIH9h/nc5x/h7BNfYfzE/ay88jQWGd1UUq9Ae20Bzy2QRRFrl09y9K63c/G5b1EulRjEBse1yBkQ5wbZW6cvXLIkQ2QaQ8T6xjr+yDTzF86B7RIMUrxihcWFM+zaux/HNfR7mqR567HIyDUf+8jP8Pd+5sN84Cd/gotnz/CpT32Kubnd1Ip1ZnfO0e33yLOMbnOVzfUN8sE03V4PqWxGJyfpDVLGp3fR2Y6xPYWQkrDTphMFFEbH6UR9sijDLziYTNNdaePt3kW3uYUcGyHsp3ilCvvH5miuXsRkKUmc4pQKxHFIf3udRqNBp71FZgnyTLDdahJ2eox4Psq3KbgufrFInGX02z20MeTfE1jaFlvdDkXn1jMzhaEOxrJtVByQpBFCKDzbRSkXBChLoHVCfxCDUhgERipMnoNSxLj85h98hqLv8J6338fMrmlWri3x1JPPMzFZozFSx5IWju9S8MtYjkOat4dNJU2QysJSw4/VODdEySYSSZ5njDoJEz7E0ga9RW4k0oJQZiSyiMkFQoEsjWGEIMqzIYMs3MaxDI5VInMstJaIoIfludyc88st2GCMuZ65kiTUajWiKCKOYzY3N5menrye2aFRSlIul4YpiM4EC0GXu+89wQvPPk6pUkMJjSV9du+Y4tuPfomDJ97Oye8+yV0PPkiz0+f5//ffcfD4O7l8ZoHmqScoegX6qc3ClctYriYJ+/S6XSqT01jdHq+dPI3l1lla73D87uOEmcPa4gJFleBYPuVilW6QMjIxzsKVVzh07G42tjoMoqGxZR4HIBNKtTLzly4y1pgmiyV/8J/+DT/2gXff6LK/oej1+3zgx36EJ7/xLV597kUeeughjuw9wLMvvsBjf/ctkihF5wapJFbBZs/UNIVCgTDtEATbrC538YpFOuspYyPTLJ15mcmpafrtreEHSlzAt122t9ewLU0uJQUE260ylgSz1URnglykGNHBaTWh4DAwFqbfR7kuju/R295icuc0/fUWyrbIsoza2AhBnlIRDv0g5PCeGfJBn5FaY7iu0Zo8z7FtC2MgSsJbkkVmdE4YBOgsxbd9HNcmDEOQFtpArzt8xqFxUUUXlE+qDZbROCS89OpZjh3cxbHbD+E6Ds31TbxGjfefOEq316bgeayvrxMPAoIwwnMLRFmOKyS9fhcpBZZjkWlBmmSkmcFIuHeuhglDNpsJubKQyhnmvigL6VdwsgFKKLS20P0umXAxjoO0PbQcI81S8jjCUj46TxFeFUtmKHNzEjluuQajtSYMQ1zXff3eMjY2RqPR4PLleQ4dOkSz2eTUqVN4vsvjjz/OvsMn+Llf+y0efvhhjBG8490/xFe/9AUKvsN3nvwG9eptvHb+Mrt3KLbbms2r8+y+7TALly9hbEHZnyZMM3bt38+lM+d50z0P8sp3vkQYhKSDLkm/i1euc/vRuzh9/hyLi4u4pTqO1cWYhG5UgbSHkoqt5gqOVMRRzMLFU4zvvY/ZfXu5dvopfCtDFSdIYvid3/lt6tUaG51bT+UdJzEXL14kNwkn7jrG1PQ4QSfk1CunyJOEv/iLz7C93eZjH/sY1UqVVmuTVsvgOC6FkkcSJwziPnEasrJ+mYnxGS5f3qBWLiOlRIUFVpa3qI2MkMddtBRI26Pf3qTkFkjSDE2G57t0W1uEMoYgItcWuWujQijWygihsS1JfXyMTrMJcUBqEpRSBGTU63Vc1yWKQjS8HvecGk2SDnUfmJuVX/SDwQDFchXLUfS7ffphSJqk6AwMGg9QnseIo0iTkGTQQ+cZyq8RG8XydsLO2Tm63ZBiQdDp9dg/NcL81UXSJMGxLKRSaAFSKQZBwCCKSWybIIoJ+n0kgkbV4/D0GLYtcQCjwK1WKNckeZ4z6IX0c402BqEcjLLJkwBXZAhhsPM+OvPQaYyyXbRlI9FgcqRTJTMpykiMuDm1Trdcg3Fdh9HRUYKgT6/Xpd3uEMcx+/btZf7SPC+/8hITExPs37+fV0+d5PbbD/OWt72dR7/2CDrpcvytP8xjj36dRqNOpTJFR/s0F1/k9iPHuPLSeaaPDxgkAdvtEqVylWbzKrPH3sHa8jwJNu961zvALpPFIbaSmHBAmiWMTo1zaeEKKo6xLEnUvEpB2QSJQ2N6kna7hSQmT/uUbJf1lUV2jFUJNl5lI7vG9maTZrhNsTHHlx/+ErZnWF1Z5/y5Cze65G84bNvmgQce4JFvfo29t93GkTuOMVEf5dVXX+FE4Rh//bUv8PSjT6G1od3vUKlUhwf0LMdRHsVqjU6nizaacqlCp7uF63msb24BhmtXL6Ism7BbBWmQSiEmZxH9TeyRcaLMYNsWcWTR295COTZaR1jKIdqWON7Qbdd1HLL1FNe1GBkpstlqYdJhyNTE1Dhp2mHx6mWkEAglKXg+cRzTDQLCaOg8UfI9oujmfPn8IBBCosnRWlCqlCiUfPJ8GIOuECh7GI3R6/XwS0VqjkUURfzxl07TaV7GcSu021ukucESGdNTDZ5/8SQCg+sqlLJIkxzbdVASMqFQWcaP3H8nZc8j0ilaQxhl5GlIGBsyWxG3AwoWSCGwbYtqvUghzYijkCRuEkURlrRINcNVvQBbGNABQbcD0kLaNsIqIZXAGJ84i8i1d6NL/n3hlmswzWaTv/rCXzIyWicKQ8rlCs8/t0DB/wDvf//7+Ju/fYQ0TTl79iyNxiRRp8lmXw89gfD49je/wYkTJ7g2fwp3bI4DpS2qRcGl+TPsvfs4/e1Vjhw5ztlTz1Ou1bnn7vdz/vTT7D94jKWrV/lv33qEid1H6Q6g4EkcR1L3d2CCjLJfJFOwvX4V2/XphjGlSpX1a/NYjk2YpNjSYbvXBpWh8ox9J95Jc+USlWLAPfe/h9/4l7/O1vYW4WqP5fVVgu3WjS75Gw6lJO9973t50z0ncF2Xl77zNNeuLLB/1yxPPv5dFtdXmNw9xaxt0el0EFIQX2cS5mb4tWlbNpZnIS1Q6TDAzfNrZFmGbSeUimWSeDilBEHA0uoVLMtiYeEMaZrgeR5JnGE5DhpQmcGxPUqjo3ipR5ZF2PU6g94221qzc9cspWKZarXGCy+8wNTkKK5r6HTaALiui85z8jwnjmOM1sM0Vt+nVrs5s0J+EGitIVf00gTHVq8LUU2Wk2Q5YbePMYZqtYI2MBgMcB2PsWKKlU+SmxyT9rnv2EE81+ItR2fR2GRZRprnw1uNSXEcHw3kWU6apXS3t9k2OZYz9DEsFosYt0SpaNB5TupIkAoLELZEG4G0BF7RxnITipUKSskhAzBLyZIcbTSptHFTTaxzZA6YLlnSRSgHS3kYfXOuQW+5BgOQZRmrK2vYtkWlUuPQoUOcOnWKCxfn2bFzipkdOxkbG+Pll19mpNZA+RVmG7vovnaa6Zk6r3z3CXYeOkGvtULz2lliI6hU67iuxWZ7hU6Q49kSqeEbj/4ts5MVnv3aFxmf3c1DP/VzrC2exwkb1AuKxVaLejkjDAdEwTZ+sUCvG1CqWpQKHt32FlJ6xEGAY0N1dCfv/LGPcP7Vp3n16ce4cv4kXmWM1lafj/7s/8KFixdwPYetzXVOn3yFvTtmbnS533DU63U2Njb4nd/7f/joRz6CcB0O3XM7URTxof/9Jzh18jWeePTb7Nu/hxMnTvDNRx9nfn6ePMuRtsYIQbFUIoj63H7kCBcvXsJ1XeqNIrZt0+4F9Lp9LM+lHfSHERC5wXNdUkvj+R7GaCwFXsFjdm6O+ddeJdYxcbODySWV+jSWzCmXy3T7GWEwwPFK9Hp9du/ew4XXXmZmZpxarc7a2ur1FFQHrTXl66s6x3GGTLT0Zj0Bf//QKLbDCGEkQaIx0gFilNF4rk+pWkGKofWKwmD7JbIs5YPvuZs0yYcfDpZLlgwdqlfaGZYCS+aIPEIpCUj6g5AEmzRJ0Jph8JtTIos1MjZsBT2KjsGzFY5t4fkllDRESYLRmlxrhDYIIRDWMD1XCIXRGm2GLEGtNY4UFFyHOIzRhmHCpZQkaUIcDbD4HzeYmwK2bbNjxw4sZbNz507Onj1Dtzv0IXvrW97O2NgISRrTbneZmZlmY2Obu/bt5aXnX2bHWIOB1nSjjCROOHJgji+ffII3Pfg+Tr/4JJsrV9lotZnc0WB0bj+ri9e44/hRgmvn0ZZh99HjbPVD9h+5g6bn0OoMKMYp440S5xa7VD3D6uoa0rLYMTPLpfnL9AcJRvfwiw0a0zPsOXoHZ68ssfvgXZx94Sm6rWWSaMBPfuijZCTYtsPK6hIXz51hcnSUJAxvdMnfcMRRQjQY8I/+0a/x6jPP8md/+Vne+bYHcRyXxx57bPi8peTy4jLf+sYTCCEolYooS3DvfXfw3SdfYGpihkHQ48LpC9x9/5088La3MNIYY319Ha0158+f4633v4VUKBauLHD04CH++ot/zcLCNbrdgCAIcFyfJEq4fOECQtgoBcfvPU6apZx79RTz8+sI6VAp19hcu4JTqXH3PXdy/sLz/K//4KPs3LmTKI6ZndvNVquF7br4vo8tHCw1jCKIkpiNbudGl/wNh9EZsVYI10HmkGcJuAXCKKIdhvgFCykdXAZk2sYTISnDyUFKSUFJfCkwRQ/jewT9PnEYMtA5BokQGinBty0skaEsyDKNsnykFCgjMGgS49DPIcBBDyKUleI4ClvZ+CrDsxRpkg0nXyURUpBfn7aU+u/yiCzPyDJD7jgY6RDGKWgQtsLyLER2c35E3HINJs81rc02s7t3Yzs+U1M7mZ6aotVq4RYtPvmff58P//2PoqShl/S4trpCa32dUsFje2ub2uRuPvhTH+WlZ54izgIqtQbt7S08xyZOEnbtOch2cwOlPISRmHabub1HODK5A39iHz9y1yGefvYFLlsNjj7wdl74/HnS4hT3vO+H6C98l3w1RNk2qfKYu+sBCo4PyQaD3FCuTNELM6Z2TBH0tobUy0ywd2YnD77jzTS3tsiTnIuXLjPaGGW0XEHcgkdgISWO57G+uc6b7rubO996DyOVEmmSc+L4MX73d3+PXj/A94r86q/9CpM7Rrh48SK3334H/X6Xu4/fi0kEp0+dolgscvL8Kzzz7NPUaw1azRY7Z3Zz993H+f3/8IeMTYzysY9/HJ1Irl5eY2lpjXK1xO49u7nzzjsYHRshCHqgBI7j8Cf/359g2RYlt4CyNZnJEXZMGITk/ZTnnnqc43fcST7IOHHkLtJkaHQ53hghy/LrqatDFmQcRdj28Bh9y0FItMlxUGDZmCxBZSm2ELiuhDyAKMLYNlLE1/VvEAuLJNa0ejmIlHrJoeooyuUypbIADEkUkWUZkRHkBmSeITGUfIc40+g8QJjrwYXESBxkHpHmKWR9dBCTKYumURjlYfIEx+QokVEolXBcjzzPh1qafLju01mONoowSZACtNFg+cNpBhutbk4x7S3XYBzH5i0PvIn+YAAqY/+BOYJ+wJlzr9Lt9Nm7ex/d7R5jY+O4BZ9LC9usLi3T3GqTZjl3zU3y+c/9GcfuuY/Lr54kDgOSsEsURhgEjm2zY2aW2T376QUpxYpNWKxipQIKZb7+1HN0t5rcfnAPr778NLuP3kfPGeH4XSf4xOd/l2P3f5ix2Vm67Q797RY9Ayf23sEzTz9KsLbN9IGDKJ3SD7pk0uPTX3qEze0rLFy+zMjoGMJ3cYsFJmt1bNse7qpvNRgoFkrctqtInMTYQhOlEVEUMTJT5zd/91+RJjl7Z3exsrLCzPgUJDnd1jZawcj4KHEQ8+C7HqRarTK9a5owHLBr/ywPf+GvOXbvcdY3NvCqVaQq8Ku/8i8YBAGf/fNP02xtghrqqn75l/45lmVh2za5TpjbO8feA/spFSusLyyjdY4QkjzO0EaRa+iHEXsPHebet7yZpaWrjI6MDdmOuUYJRRqn2AosBFGuEQimGrdgqJwwNFwF+fB36QtJkgRYWGhjyABpOWhlk+UxOs/xHIUvE2LpoNUAqQVJmNDqS5SVYUkby5Y4lkWxXKEoNHGUkMUQGU2WJOgcpO2gdIYSFpmyMMpGI7BUTm5yoIAtNXXPwUiBzl3AIAXkaUoaDwkaKAvHtnEsC2HbYFs0LPqQtZUAAB5PSURBVIc4N5g0JgwDlJAIE5L/D7PLmwN5nvP8c88RBAGDwQDbtul2u0xPDxXvQghW1pbYs2+OTi9kefUas/0O4aCPXyjx2Nce4f43neDl114d+hHlKZYES4AGMDl+oUx7ELB7/x2Mq4SXLlxiLUxJlzeJBjFzkyVaK4tUK1WEp3CNRVUZSn6NreYV7IJNuxPx7re9le++9AKrcYoz2kAEMdMjVa6uruEXCnzwp36aj338Y/z6P/uH7Nqzn4XFRfyiy4jjkKQx9XqdtbW1G1rvGwHHsanX65y7OM/ISJ0wHGBkil8o0B+E1GplXM/lb771DWZ2zJAv52w2N0FAEscU/SqVYpmlpSWyPGbvgd0Ycuojo9x337184b99ESkEUir6dod33/9OevGASmOE8clpTp16lT/65CeJ+l0eeeSrtFotyrUy/8fHf567TtzB6VNn6A86Q3six2ZqcpyrK8vEYQfbcqiXPb792JO8/33ve92+SEmJMPL1ZEvP8ygWCkOx8C04pUopKZc8lCyS5xlaCEyWk+uhN5sUgJAYC/LMRuc5WqfEsUFJTdGzUEiM1mQCBrEgzTVZlpNYhjhNcWyJ7xVwLIuKZRGnQ+cPoyHPh69OS2sGaUhqHJTtkmuN0RnkQC/CCDBCUy76ANiuhxYGjCA3OTmCPNVoA55UBGEPIW20AOH6WMJGkN20L+qb9e/+vuE4DtVKg10zc5w5c4YHHnyQ1bU1Op0OmOEDrjZ8Pv/wXxAEml6/i19w8COPSqVCJAXNzSa636fpu4yPjRAO+li2QEoL17EIBwNEqcLJxWtUwnXCOMcTiiToMzE+g+lvYIREYBHGGaMjFWwr5hd/7dd58vFXERhCKVleX2OqVqdeHyFb8WgGLS6efQ23XGF9vc38xauMjVZIugkr/XWq9TrX5i+QLV9Dlnxcx+Oll1660SV/w2GMwXEclpeW6ff7VKsVxkYmuHp1kTAJaW6t0wvazO3eQ65z/v3vf4L777+fY8eOMTJSp1yssbXZZNfcHOvrG/SigI3Ndb77nedYXl7jwXc+yPjkBHv27uPLX/oyi1tLtLd7fPSnfwaF4L3veS+/9Zu/jTA5WZ4wPjGKEIqiX8B2HMJBOFTjK0UYRSwuLpJjqJSLJEnMv/8Pv8dv/ea/xi/Z9Ls9ut0u5XKZYqGIbRdwXYcsS1DKQgpBsXArqvkF3SAmv54Y6ToWthBDxp5jkWmDGrqD4fkuSlmAIQwjAHJjiHKNySBHYtsuwhvSnC3AUgJhhky9TOeoDHKdYxhSi/sRJJlG2hZK2GACRJoidIbJDWkyYHx6EuU4JEl2PVkVNBmpNmSJIROQZin9MMEgyfSAim8zWrXJoogwNVhWjpYWVn5z2gHdcg0mSRLqjQqtVotKvcw3vvl13vve97FjZpoXX3yR22YOcOnSJUqFKuMjZeq9Po9+5WEO3fEmBv0AZQlWNyKqY1NEmy2aSRcj+oxUyuS5ZpDkNMam2d7aRhYUs3tnefmVU1jK4Hklep0mfr1CtLbMdiw5cuggcZoR90J6iWY7aOPGRWpTO8l0xmijyJPf/Rbve+idPPLZzyKSkGwrQqQZ0xNTtJsJf/fUM9x+x0HWVi+zff4snutRTiWf/6vPcvyOO250yd9wKKVQQrB7325eeeUV9KLm9MlX2bd/H5cuXWJtdZ1BOKDd2qJWrfLe//kDfOYzn+E7TzxLmIQoy2J6fJLla6soNZwgwjDkyNEDfPCDP87O6WnqoyP0w5Df+PVfotuO+cbffouvPfoNjt15O3me8e1vf5tr164xPj7OzMwMzzz7DNM7dnD2ymWm53ZRtCyEZXHuwgV8yyFNU0yqkFpRLhX5t//uE2yuNfnQB3+URqP+en5RFIcIbfA8DykFOYr/v70zj7arqvP853fGe+97981J3ntJIBNESMA0YRSCVUIpZYldDkvAtrUKre4uq5bYDrRdrXRhUYpWidQoy0LtxiwKpUQKsZBGBIUwCfogARIIEDK95A33TXc64+4/9r7m8njvZYAQqJzvWnfdffbZ5+x9fsP+7f377bPPnuHhI03y1xyCwiu45GyXJE2p16pMJOhPIOCCpHrHjlqdehCSpgmW4+I5tv7QoCi8VJG4KShBqZBUWcRRRIIiifSmma7j4LgucRpgW5Z+N0WEzrzScZM0IE1rlIMILPBsm3zex+looTRVox7XCBOLSr2OUyhiqYi87SBpRJQmqKiO7+ZxLQ/Xtsg7FkEQ49g+HiEqjShKSqTqR5rkh4SjzsBYIrz44g4syyKOFSeeuIqHHnqIJUuWkM/n6e7uYXR0lNHRYUZGhli8ZBlBuofNTzzCiavWsPCY49n6/DZaWjzGShaWq0chwyO7cN0WXL+H1LJYufJN1LGpRQFeLk99cphyNaAWWVhxmdach1TKRFGJvaWAUrSA3S9uh/IEFS+hEE4hdZ/IcqFe486f/Bs987qpxTG7du1iyaKlRI6ivQClkb14rKC9tcjad1zAyN4RRkdGcByXMHhjrj55JQjCgMc3bmS0PMbKE46jo6ODM05bSz6fZ3xinP6+fvbs3cPdd9zFhg0b2LJlM6eeeipPPfUUgkAU8+zmF3jzmtV0zevgsYcG6OnsolYJGNy1B1tZ3PyDH1IotrJ58xYc22X1qtV8+UtX0tpeYGh4F6VSid5FHRSLHpVghLe85WS+evU1WDmf91/0fo5fupQXXnyRE9ecwInHnUR/fz9PP7GJloL+bHKuWKTF89m+cydxHFNsbaWnuxvf95FUYds25XKZ1HJoaWk50iQ/MkhTQhWTpimO59PtuthAFIV6x456jQRwPI+c45BEIUFYp16zsWwzEHGEJElRArEIrutR8PSms7brkiQJjuPs21DUcbAQXNsmwWzdE0a0dbjEcUSlUqEchXgqhUTRkvNQ9Zi21lYc10JFCisNSOI6BcDPtzBVr2DHVbBtUsvDcmzqUYoSn3oQUk8sCrmOI0joQ8dRZ2Bs26Gnez7FYpEkThgeGdLfSx8ZIYoi9uwZpKuzh3J5ilq9yvYd28j7Pv0rFrJj6yaee/ZJdu8d4qx1Z5FTPlPVSYIoprt7AbbbQoQwMjVFENQRFRFVJvHyHdSmJrByRbq7eqlVx9kztAev0MLkVJmgWmXDXT8iHJtibGwvq9auY8XyhezZvp1aEiNhQGuxg+HBQRKxKBaLDI6O4VcmSOvjVCvjjO89lsSyeOSFFxgdHqG9vZ2TT1pDefzoW8IaBiGWZXPC8cexfXiIjvY2JmtlKlGNjo42quVJujva+dBHP8iF7/89dm/Zzc5dO3nnuW+jra8DpVIWzO/7zaanH/2DS7FsG9sSLIHv3/xDyuUqzz77Au+88G2cf/55bN7yFINDWwl3BkBKPagjFoSlSXK5HIouPn/l/2TZohO49Zbb2VLdxo9+chsrVi5n81ObWLt2LW9ddx7vvOD36O3r56b1N9JpdhhQSqGUYmx8HNdxaG1pw4pTLMfDscHyjzo1RgGubePYeg83sR39TlCS4uUL1Go1akFInAhpPQSpg7go5ZJzhZzngxWjkpQoiVGphe0kYAmVMCDnOtQqKcpKsYOAVBQ5xyUJQ5QIpDYpgoie6ZTrMamC1HYp5DyUcsn7Mb7vk8+lgOivavoutbBOvWpTSyLtOrMcYr/VfKvKxrX0QNhKA4qeRZoGJPXsRcs3BBzHYdnS5XqWMl6ipaWI63pYloVIjVqtjmunRGHMgt4+avU6tWqV0vAQPV1FhkZLzG/LsfXJASbGKhRaCsRxSjBVxnZ9lCN4Hb209fahopDqRAkbi3y+lSBIiGo1rFjvbxRUy+zauRcnjSkP7yAKQnCF1rzLVGkvniQ89/zzRFFArVymrb3Int278H2fjs75JNUpIpVQHxtm4FcPs+70MzjmmEWctHoVYag7WX/psiNN8tceAnFaZ8f2QezJCl/77tdwXBgdHeXUs9/CGWedRf+8HsbHx8kVO1iy1qf/pIXk/Bw5r0Acx1SigDBJuH/DvZQGR9i9ezdPP7sdz3ZI45jzzl/Hhe87D88u8Py250hJdYyuXgcl5HMxlm0Z46DlqVqusXtoKxe861y2bN7B6MgoV1z5Bb765a/w4APX8e1/Ws+8eQuoVepcc83X+dRll9HV3YXn+VgKJIVCPo+y0fWg3804Gje7TFLFyFhAKhXEa8FxFK1pRN7TLrDWQp6C51KamCCybKLUJUwUjuNQj6ESJ/iuRxLWUdik2LhpiigdpB+zLb2bsShI6ojt4EqC57mESpHEISl6URCAJwmOKHK+h+85SKpIsPVLsKnCchwsR7tb865DPpfi1QPiFOw0JlQBiVUgShOiOIVUf4hMiXaDWtYbcyn6UWdgRIQk0W9QAyxevJjx8XEGBwfxfY9KpUJv3wJyeZ+h0jAiNn29iwhqdUqjo7S1FSkWWxkbn6KjzaJQyFOv13EJcW2LkdESMjrE6PZniNKYvnnzqUwOkcu14uSLVNUUKgmRNCSoTtLSmmd8aC8Fz6IWlim4BXZu3cj8BT0sP3Yxv7h3Czk7YXy8zujYJFGiXxZLX9xGqhTzOwvM611ILUqYimK8WsBjj95DrVrj9NNPI5/PH2GKv/ZQKmW0NIxvK2q+zblvXUtrRx+FQoF6ZYpfbdjAbTt2Mj5Rph4ElKdCtm9/kSRJiOKE1rY2OluL+gNVjk3/4iX09Czk0598G6vedAJPP7+FF7ftpjwR4HkRcRrjODaFrm7yuRaKre3GwEMYRSRxgl/IY1kWURgxMVHmpJNWM7p3mKuuuoqolpDzWihPVeju7iFNhCuuuIJ6tUqlXGG4OsynL7+cY5Yfz5YtW/jrL15Bb28vPT09eguZo3Apum0Jvi+gPOJoAqklRAJRoYDt5HCtBJUKxUIe8VzSSLuygqBOPVVUU4s0raIsC1vplWUqtfQsBIUkIbboWUmCj2O5VJOYaiXGcvRCElKFjYV4NrXYBQUTiYVTC7HSlLybkLcsIktwlU+1WsaxXcJIYVkO+ZYinuvq2G5UJybEsx2SIMVxLVoch1iEKIlJoyzI/4aAiFCtlUmShEWL+xke2YvjuCxbvoSJiQn6+/splUra162EvJ9nz969BGGIY9tE9QClUrq6uoiiiNbWVhzHoTRSwraE/t4FBJUK1WqA+MLk8A5IU8ajCKUcVGsrdhwTRoqpyXEmBrdhW4qabaOSlNHSENTHGdz2JI89oKfPu0cqpCqlvb2TYlsnw8PD2K4LIiTEuLkc/b197Bka5ZSTT2H5sctJVUqh4DN8FAaAPc+nu6OLsfIURS+P09tJmiYEQYVUFCtWHk97dxeJROwa3M3o0B76lrTQ2lqko60H27aYmCqzZMkSFi9ezMjwKLWgjusX2DM6hi05Vh6/kkq1TDmo4eZcVGpRrgR0dXTi+x56YCv4fp40TVEpWLZFXQlTtSrDL2ziltt/wC/u+TnXf/ObrFq9mvN/6+20tbVz00038fyObTz04IP82acvRynF7bfeShRGOI7NcGmEDfc/QEdbFz3zu47KTyaLCPPaW4nTlFpNSC2LSi0gqYPrBlTDCNdKSR0Puxxgu5ZeOWa7FAp5Oq2Y1LZwU72ay06VHmBE+gOASawI4hhPEmIVIpHgYWFbCXEMmI0qIxSWUnhRFd/1sJTeCSBfyGHZPomktDoecRhiKSGIEoq5PJYkOBbU6lW9E7PXApZ+s1/5FraArRxsUTgxyBtzAnP0GZggDCiNlQiCOkkaUy6XyeVyiNXJ2NgY5XIZ19XbyHR09bB9+3YqtTqt7e0MDg7i2hZBEFKpBoRhyODgICJCsaUFVVU4nkdHZxvHHnsME9Up/XpUmjI5PsHi3n52jY2hgpB6HLGsbwlhPSZRCUESEgcBpVKZaqWG61n4uW4SJXT2FJmcnKQWJqjJSR2ctBSOYxNM1rGTFEeFzO/p4eknBzjhxBNwbIf77ruPycnJI03y1xxJHBMEAZ7rMjU5RS6Xw/X0RoYqhVxLC6tX9YOjWLr0OAo5HSifmJikkG/B93Mk5lMOlmXRu6BfGwlJqVT0DgC+rzc7LER6iXAu14Ln+YRhxPj4OEE9oLevj7a2NlzXxVIO1WqZepyQpil7xnaze/cuzjzzDH7+s3sYHR6hs7+Nvr4FvOM95/HowINgh/zjd66jr7eXbc+/wLoz1lEsFvFyLmeffTaem8NxLaamjj4ex0nKrtIkYtu4lmClirZiAUul+K5HnESIcgnCKpZvkSKQJLSqlKA2SU0cUBaOLXiOgFjk/Ry5QgtxmoKKCer60wmWpbfeV0oRxzGeaxMmIJZ+C98RC7wCSoQw9ajFQhh7+LHehSFNtCusra2DNE2J45A4iSnXAxQWVqxnX0lqIVZK3vWxHbDSENu2SBPzjt0bEKKOspe0RGQYePFIt+M1xrFKqXlHuhGvFTIe//tHxuM3Bo46A5MhQ4YMGV4bvDE3uMmQIUOGDK97ZAYmQ4YMGTIcFmQGJkOGDBkyHBbMaWBEpFtEBsxvj4jsajr2DlejROSdIrJFRLaKyGeb8peLyCMm/0YRcU3+b4vIr0UkFpHfn+O+O0Vko2n/RhG50OQ7IjJ+iG1dP1ed08r+log8adp6UPQTkWUicvGr3T7D400iokRkYhqPd4pIh4isEJGBg6z3UhHpneN8xuOXX3tYeGzKLDC0++i0/Fl5vD/93x+PDxWHIBuXi8jTIvK4iNwlIotnue9OEXld77li6FoSkb+Yln+/iKw5FDkWkfeKyJte3ZYeGOY0MEqpUaXUGqXUGuA64OuNY6VUCCAar9pMyAjN3wNvB1YBHxaR483pvwK+qpRaAVSBPzD524APA98/gCrWmee5GLj21Wr3AeJDwNVKqf/QoN9BYBm6za8qlFKjwDeB+4FRmnj8Cm99KTBj5yMiOTIez4TDwmODi4AHgUsO9IID0P9Lgd6Z9F9EDukViEPU/0eBU5RSbwZuA64+lLpfJ7gAeArNr1cL7wVmNDCHyqcDRmOfo/39gD8HPmPSK4BNaKH7NXAsMN5U9mLgepNeANyCFoJHgDP3U8864MdNx18APgvYwAhgz1TO5K0Hfn+Oe+8EOkz6LOBRk3Ya7QfagJ8BvwKeAN7VdP0fmrzHge9MrxP4MvAtwJqh7v8GlIAXgBvQxv0aQ8eNwPtNudnyHwUmgAHgE6bN1xiaPgF8rOn6f0QL6Y+An8xFE3PNg8BaYAy4sonHEfBt4EngGSA0bfo+8J+B64ErjQxMAEOmPf8LKANbTHs9Q/svABvM/33Aw6btT5o821z3lKHxncCPgbeZ4wFDw4syHh8yj58HeqfTy/D7SeC7TTzOmzJXAruAQbTOrwC2G/mom+vGm3i8BbgbOAV4zDzTmEmfCfz3Jh6vN3U0ePysKd9ysPpv8k8Dfr4/2WjKO9PQ5tdo2TzO5D8ArG4q9zDa4DXL4q+AltnkiX395LcMje4Acvvh0z+jDcJ9wGlN+fcDa4xMjANfN/XdBXQ3yd8vTftuBvKGTg2ZHACWmHv9JfAL4JPAUuAe0/a7gEVN/fgmc797TN5Jpo4BU37ZnM/zCgxM2iAATco7g4H5HsaomIfbZNJnANfNUM/Fzflohb8WPRre3JS/FBg4BAOz0TC7ClwwQ+fjAkWTng88a9JvBjYDXea4q7lOdEfwD5il37PU39xRXYTuGGzzbDtMfbPlnw/c2nSvjwOfM2kfrSDHAB9AC7IFLAIm90OTJQ26ohXs1iYeK+C3m9JT5twNwP9FG5iuBo/RyvERI5T3o0fz1zXR/lNNPC4B55jj29GdTy8QA57JPxktyHcAZ5i8m4D3Zjw+ZB5/FfjE9E63iccNXb0B+GSDDhj9Nzy+FK3/A7y009sJfMrw93q0EbnbyMaX0DPlTWhDdTbaWDUGA3eg+4TGtQ1DcsD6b/Kva9BsFtmYbmDam+q6APieSX8U+GuTPhF4uLmdJt1qeDibPDUGaSeZ41uAi+fgUwuwG8gZ3l/TdK7ZwCjMIAv4InCtSXc3lb8a+OOZ+kVzr79rOr4D+E8m/V+AfzHpp4EFJt3g0zea6vbZj8F8JdOj55RSvzyAcucDK8VsCgd0ikheKfUwelQwHTJDnpoj/2CxTik1bqbdd4rIKvTIvLn+r4jIOWglWiwiPeiRy/eUUiWAxr/BlcADSqk/Poh2nAPcqJRKgD0icj9w6hz5090tbwdOaPLZtwPHAecC/6yUSoGdInLvftpxCdpAgFb+dzSdi9GdWg96BNRl8tejBXsTcB56xPVutPCfa8pMogcT65vu16inCDhKqfvN8QbgMjTtA2C9iPyrqVuZ89eKyI1oZd7fi80Zj1+KZh7fhDaSfztDuReUUg+Z9Hp0Z3Mtmsd/xD4d3AY8h55tTsf30KNmH91RNtxrPto4jqM7rsuAf0V3wGB4jJ79WIY2DRyQ/ovIR9Aj7E/MUHY2dAA3iMjyafk3AQMi8jm0Qf1OczuNLP5AKVUWEZuZ5Qlgq1Jqo0k/hjb2s+HdwF1KqbqI3Aw8KiKfMXxuRoyeoYDm040mfbKIfNE8UxE9cJsNNzWlzwDeZdI3AI34zwY0bW5GG0fQM7vPi8ixwC1Kqa1z1PGKVpFVmtJ6P+p9yDWlBThd7fPdLlRK1ea4706gOUi3CG3Vh4Aew8zm/EOCUuoZ9Ch6um/yw2hFPkVp3/OIeR5hdoP2CHCqiHQeRBNmUpi58mcq9/Emui5VSt1tzh2M4b0E+JiIbEOPrBeJyNIZ7jNdyS306O3v0R1NO/AV4BtKqYXMbAQaMjPIS7cpWoDuXIeAGnqkezrwb8BupdRVwH9Fjxh/F+g7kAfLePwbNPP4FmBtE4+bMf2eSkQKaB7fhJ7BfRttLGbT/woz6D96BvikkY3fQRud09GdqN3E4wD4gIgcZ64/IP0XkQuAy4H/qA4u9vWXwJ1KqdWmjTkApVQFuBfd6b8PPXNjmiz+0rRzNnnCPE8DCXNvz3UJcIHh0y/Rs6FzZyj3Mj6Z/xvQs5aTgKt4KR+mozLHuQb+CPjfaKP4uIh0KqW+C7wH/Vx3ichM7fsNXpXgvLGwYyJynAn4vafp9E+BP2kciMj+gscPASeKyLEi4qPdAbeZEc19Tff+CHoEdEgwq1+OQfuSm9EODCmlYhH5HWBh03NcLCJd5vqupmt+DHwNuF1EWg+wCb8w97NFZAHaZfDoHPlT6FFJA3cCH28E6URkpYjkm663RGQh8NY5aHAi2j2wUCm1BD2CvJeZA81LgYpRqEvQ/l0H3bncjXaNvM/cd80M7W3GnYAjIu8xPP4g8P/MuUfRI7DPot0iPxGR5UqpJ5RSX0YbjIUz3XSG58t4PI3Hhs9/xSw8FpHTTPoStCslj+ZxFW1Y3tdUfgrtGkrRMRYHbVAaHVANPQv9E3TM7ucicgrax/8zNI/nAYUGj4E/NfWdc6D6LyKnomdl71ZKjcxGi1nQjo4vwb5FAw1cjzauDyilJkxdzbL4a2Als8vTAcMMXM5A06bBp08w86IMF+01AK07DU9AC3pG7Jr8BubSRdB97gdM+kNo+QIdX3kIHQcbAxaKyDKl1Fal1N+gdWLuT+bO5T+b5h/8c14ag5ke/7gIPW2+B83sRgxmHvAv6IDQU8A/mPwZYzDm3IXooPJzwP9oyl+Btuxb0SOqhq/+LPTMp4IePTxh8m1MkFe91D8/gPbRf0S93D8/H+26exQ9kn6GfUGvS9FuoQHgW9P9m2iLfzez+CWnlT3YALCH7vwfRwuejfazbjK/n6GFyEL7SZ8CfogOAs/on0ePcq6axuOvm3pXoGcVDf/8RnQHHKBHk9ehFfBqdOB4L7rzHzL8/wB6r6gR9gX5O5rqugzdAQWm/e3oDuwxdGcWsG+BwDeMLIRo90DG40Pksck7BdjYRK9mHn8TravNQf6r0SsMnwH+j2n7gOFxYyHHBw1v7jP8v559Qf5xdOB+s2n3BnQwvwR8Vu3z7W8ydd/Dwen/vcAe044B4IdzyMYu878THY86x9S1wdBq6zRabQXObzpubud6w7MZ5Ylp/STwOeDzs/DpY5gFD01589D65PLyIP+X0EH+n7IvyP+naF28F20YG33wuWi3ZHOQf01TPcuYOch/m5GJTZh4EPB5tF4NoD0MHTM9T+OX7UWWIUOGDDPAvE9zF3CCyjrKQ0L2Jn+GDBkyTIOI/CE6oP1nmXE5dGQzmMMAEbkN7ftvxmeUUj89Eu2B12eb3sh4PdLz9dimDC+HiFyHXrrdjGuUUjccifYcTmQGJkOGDBkyHBZkLrIMGTJkyHBYkBmYDBkyZMhwWJAZmAwZMmTIcFiQGZgMGTJkyHBYkBmYDBkyZMhwWPD/AcN7O8Hgjvt1AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1c34c0edf98>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Get some random images and their labels from the train set.\n",
    "\n",
    "images, cls_true  = data.train.images, data.train.cls\n",
    "\n",
    "# Plot the images and labels using our helper-function above.\n",
    "plot_images(images=images, cls_true=cls_true)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## TensorFlow Graph\n",
    "\n",
    "The entire purpose of TensorFlow is to have a so-called computational graph that can be executed much more efficiently than if the same calculations were to be performed directly in Python. TensorFlow can be more efficient than NumPy because TensorFlow knows the entire computation graph that must be executed, while NumPy only knows the computation of a single mathematical operation at a time.\n",
    "\n",
    "TensorFlow can also automatically calculate the gradients that are needed to optimize the variables of the graph so as to make the model perform better. This is because the graph is a combination of simple mathematical expressions so the gradient of the entire graph can be calculated using the chain-rule for derivatives.\n",
    "\n",
    "A TensorFlow graph consists of the following parts which will be detailed below:\n",
    "\n",
    "* Placeholder variables used for inputting data to the graph.\n",
    "* Variables that are going to be optimized so as to make the convolutional network perform better.\n",
    "* The mathematical formulas for the convolutional network.\n",
    "* A cost measure that can be used to guide the optimization of the variables.\n",
    "* An optimization method which updates the variables."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Helper-functions for creating new variables"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Functions for creating new TensorFlow variables in the given shape and initializing them with random values. Note that the initialization is not actually done at this point, it is merely being defined in the TensorFlow graph."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {},
   "outputs": [],
   "source": [
    "def new_weights(shape):\n",
    "    return tf.Variable(tf.truncated_normal(shape, stddev=0.05))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "metadata": {},
   "outputs": [],
   "source": [
    "def new_biases(length):\n",
    "    return tf.Variable(tf.constant(0.05, shape=[length]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Helper-function for creating a new Convolutional Layer"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This function creates a new convolutional layer in the computational graph for TensorFlow. Nothing is actually calculated here, we are just adding the mathematical formulas to the TensorFlow graph.\n",
    "\n",
    "It is assumed that the input is a 4-dim tensor with the following dimensions:\n",
    "\n",
    "1. Image number.\n",
    "2. Y-axis of each image.\n",
    "3. X-axis of each image.\n",
    "4. Channels of each image.\n",
    "\n",
    "Note that the input channels may either be colour-channels, or it may be filter-channels if the input is produced from a previous convolutional layer.\n",
    "\n",
    "The output is another 4-dim tensor with the following dimensions:\n",
    "\n",
    "1. Image number, same as input.\n",
    "2. Y-axis of each image. If 2x2 pooling is used, then the height and width of the input images is divided by 2.\n",
    "3. X-axis of each image. Ditto.\n",
    "4. Channels produced by the convolutional filters."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "metadata": {},
   "outputs": [],
   "source": [
    "def new_conv_layer(input,              # The previous layer.\n",
    "                   num_input_channels, # Num. channels in prev. layer.\n",
    "                   filter_size,        # Width and height of each filter.\n",
    "                   num_filters,        # Number of filters.\n",
    "                   use_pooling=True):  # Use 2x2 max-pooling.\n",
    "\n",
    "    # Shape of the filter-weights for the convolution.\n",
    "    # This format is determined by the TensorFlow API.\n",
    "    shape = [filter_size, filter_size, num_input_channels, num_filters]\n",
    "\n",
    "    # Create new weights aka. filters with the given shape.\n",
    "    weights = new_weights(shape=shape)\n",
    "\n",
    "    # Create new biases, one for each filter.\n",
    "    biases = new_biases(length=num_filters)\n",
    "\n",
    "    # Create the TensorFlow operation for convolution.\n",
    "    # Note the strides are set to 1 in all dimensions.\n",
    "    # The first and last stride must always be 1,\n",
    "    # because the first is for the image-number and\n",
    "    # the last is for the input-channel.\n",
    "    # But e.g. strides=[1, 2, 2, 1] would mean that the filter\n",
    "    # is moved 2 pixels across the x- and y-axis of the image.\n",
    "    # The padding is set to 'SAME' which means the input image\n",
    "    # is padded with zeroes so the size of the output is the same.\n",
    "    layer = tf.nn.conv2d(input=input,\n",
    "                         filter=weights,\n",
    "                         strides=[1, 1, 1, 1],\n",
    "                         padding='SAME')\n",
    "\n",
    "    # Add the biases to the results of the convolution.\n",
    "    # A bias-value is added to each filter-channel.\n",
    "    layer += biases\n",
    "\n",
    "    # Use pooling to down-sample the image resolution?\n",
    "    if use_pooling:\n",
    "        # This is 2x2 max-pooling, which means that we\n",
    "        # consider 2x2 windows and select the largest value\n",
    "        # in each window. Then we move 2 pixels to the next window.\n",
    "        layer = tf.nn.max_pool(value=layer,\n",
    "                               ksize=[1, 2, 2, 1],\n",
    "                               strides=[1, 2, 2, 1],\n",
    "                               padding='SAME')\n",
    "\n",
    "    # Rectified Linear Unit (ReLU).\n",
    "    # It calculates max(x, 0) for each input pixel x.\n",
    "    # This adds some non-linearity to the formula and allows us\n",
    "    # to learn more complicated functions.\n",
    "    layer = tf.nn.relu(layer)\n",
    "\n",
    "    # Note that ReLU is normally executed before the pooling,\n",
    "    # but since relu(max_pool(x)) == max_pool(relu(x)) we can\n",
    "    # save 75% of the relu-operations by max-pooling first.\n",
    "\n",
    "    # We return both the resulting layer and the filter-weights\n",
    "    # because we will plot the weights later.\n",
    "    return layer, weights"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Helper-function for flattening a layer\n",
    "\n",
    "A convolutional layer produces an output tensor with 4 dimensions. We will add fully-connected layers after the convolution layers, so we need to reduce the 4-dim tensor to 2-dim which can be used as input to the fully-connected layer."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {},
   "outputs": [],
   "source": [
    "def flatten_layer(layer):\n",
    "    # Get the shape of the input layer.\n",
    "    layer_shape = layer.get_shape()\n",
    "\n",
    "    # The shape of the input layer is assumed to be:\n",
    "    # layer_shape == [num_images, img_height, img_width, num_channels]\n",
    "\n",
    "    # The number of features is: img_height * img_width * num_channels\n",
    "    # We can use a function from TensorFlow to calculate this.\n",
    "    num_features = layer_shape[1:4].num_elements()\n",
    "    \n",
    "    # Reshape the layer to [num_images, num_features].\n",
    "    # Note that we just set the size of the second dimension\n",
    "    # to num_features and the size of the first dimension to -1\n",
    "    # which means the size in that dimension is calculated\n",
    "    # so the total size of the tensor is unchanged from the reshaping.\n",
    "    layer_flat = tf.reshape(layer, [-1, num_features])\n",
    "\n",
    "    # The shape of the flattened layer is now:\n",
    "    # [num_images, img_height * img_width * num_channels]\n",
    "\n",
    "    # Return both the flattened layer and the number of features.\n",
    "    return layer_flat, num_features"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Helper-function for creating a new Fully-Connected Layer"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This function creates a new fully-connected layer in the computational graph for TensorFlow. Nothing is actually calculated here, we are just adding the mathematical formulas to the TensorFlow graph.\n",
    "\n",
    "It is assumed that the input is a 2-dim tensor of shape `[num_images, num_inputs]`. The output is a 2-dim tensor of shape `[num_images, num_outputs]`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "metadata": {},
   "outputs": [],
   "source": [
    "def new_fc_layer(input,          # The previous layer.\n",
    "                 num_inputs,     # Num. inputs from prev. layer.\n",
    "                 num_outputs,    # Num. outputs.\n",
    "                 use_relu=True): # Use Rectified Linear Unit (ReLU)?\n",
    "\n",
    "    # Create new weights and biases.\n",
    "    weights = new_weights(shape=[num_inputs, num_outputs])\n",
    "    biases = new_biases(length=num_outputs)\n",
    "\n",
    "    # Calculate the layer as the matrix multiplication of\n",
    "    # the input and weights, and then add the bias-values.\n",
    "    layer = tf.matmul(input, weights) + biases\n",
    "\n",
    "    # Use ReLU?\n",
    "    if use_relu:\n",
    "        layer = tf.nn.relu(layer)\n",
    "\n",
    "    return layer"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Placeholder variables"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Placeholder variables serve as the input to the TensorFlow computational graph that we may change each time we execute the graph. We call this feeding the placeholder variables and it is demonstrated further below.\n",
    "\n",
    "First we define the placeholder variable for the input images. This allows us to change the images that are input to the TensorFlow graph. This is a so-called tensor, which just means that it is a multi-dimensional vector or matrix. The data-type is set to `float32` and the shape is set to `[None, img_size_flat]`, where `None` means that the tensor may hold an arbitrary number of images with each image being a vector of length `img_size_flat`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "metadata": {},
   "outputs": [],
   "source": [
    "x = tf.placeholder(tf.float32, shape=[None, img_size_flat], name='x')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The convolutional layers expect `x` to be encoded as a 4-dim tensor so we have to reshape it so its shape is instead `[num_images, img_height, img_width, num_channels]`. Note that `img_height == img_width == img_size` and `num_images` can be inferred automatically by using -1 for the size of the first dimension. So the reshape operation is:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "metadata": {},
   "outputs": [],
   "source": [
    "x_image = tf.reshape(x, [-1, img_size, img_size, num_channels])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Next we have the placeholder variable for the true labels associated with the images that were input in the placeholder variable `x`. The shape of this placeholder variable is `[None, num_classes]` which means it may hold an arbitrary number of labels and each label is a vector of length `num_classes`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Tensor(\"y_true_2:0\", shape=(?, 2), dtype=float32)\n"
     ]
    }
   ],
   "source": [
    "y_true = tf.placeholder(tf.float32, shape=[None, num_classes], name='y_true')\n",
    "print(y_true)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We could also have a placeholder variable for the class-number, but we will instead calculate it using argmax. Note that this is a TensorFlow operator so nothing is calculated at this point."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Tensor(\"ArgMax_4:0\", shape=(?,), dtype=int64)\n"
     ]
    }
   ],
   "source": [
    "y_true_cls = tf.argmax(y_true, dimension=1)\n",
    "print(y_true_cls)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Convolutional Layer 1\n",
    "\n",
    "Create the first convolutional layer. It takes `x_image` as input and creates `num_filters1` different filters, each having width and height equal to `filter_size1`. Finally we wish to down-sample the image so it is half the size by using 2x2 max-pooling."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "metadata": {},
   "outputs": [],
   "source": [
    "layer_conv1, weights_conv1 = \\\n",
    "    new_conv_layer(input=x_image,\n",
    "                   num_input_channels=num_channels,\n",
    "                   filter_size=filter_size1,\n",
    "                   num_filters=num_filters1,\n",
    "                   use_pooling=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<tf.Tensor 'Relu_8:0' shape=(?, 64, 64, 32) dtype=float32>"
      ]
     },
     "execution_count": 124,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "layer_conv1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Convolutional Layers 2 and 3\n",
    "\n",
    "Create the second and third convolutional layers, which take as input the output from the first and second convolutional layer respectively. The number of input channels corresponds to the number of filters in the previous convolutional layer."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "metadata": {},
   "outputs": [],
   "source": [
    "layer_conv2, weights_conv2 = \\\n",
    "    new_conv_layer(input=layer_conv1,\n",
    "                   num_input_channels=num_filters1,\n",
    "                   filter_size=filter_size2,\n",
    "                   num_filters=num_filters2,\n",
    "                   use_pooling=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 126,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<tf.Tensor 'Relu_9:0' shape=(?, 32, 32, 32) dtype=float32>"
      ]
     },
     "execution_count": 126,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "layer_conv2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "metadata": {},
   "outputs": [],
   "source": [
    "layer_conv3, weights_conv3 = \\\n",
    "    new_conv_layer(input=layer_conv2,\n",
    "                   num_input_channels=num_filters2,\n",
    "                   filter_size=filter_size3,\n",
    "                   num_filters=num_filters3,\n",
    "                   use_pooling=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<tf.Tensor 'Relu_10:0' shape=(?, 16, 16, 64) dtype=float32>"
      ]
     },
     "execution_count": 128,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "layer_conv3"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Flatten Layer\n",
    "\n",
    "The convolutional layers output 4-dim tensors. We now wish to use these as input in a fully-connected network, which requires for the tensors to be reshaped or flattened to 2-dim tensors."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "metadata": {},
   "outputs": [],
   "source": [
    "layer_flat, num_features = flatten_layer(layer_conv3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 130,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<tf.Tensor 'Reshape_5:0' shape=(?, 16384) dtype=float32>"
      ]
     },
     "execution_count": 130,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "layer_flat"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 131,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "16384"
      ]
     },
     "execution_count": 131,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "num_features"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Fully-Connected Layer 1\n",
    "\n",
    "Add a fully-connected layer to the network. The input is the flattened layer from the previous convolution. The number of neurons or nodes in the fully-connected layer is `fc_size`. ReLU is used so we can learn non-linear relations."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 132,
   "metadata": {},
   "outputs": [],
   "source": [
    "layer_fc1 = new_fc_layer(input=layer_flat,\n",
    "                         num_inputs=num_features,\n",
    "                         num_outputs=fc_size,\n",
    "                         use_relu=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Check that the output of the fully-connected layer is a tensor with shape (?, 128) where the ? means there is an arbitrary number of images and `fc_size` == 128."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 133,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<tf.Tensor 'Relu_11:0' shape=(?, 128) dtype=float32>"
      ]
     },
     "execution_count": 133,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "layer_fc1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Fully-Connected Layer 2\n",
    "\n",
    "Add another fully-connected layer that outputs vectors of length num_classes for determining which of the classes the input image belongs to. Note that ReLU is not used in this layer."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 134,
   "metadata": {},
   "outputs": [],
   "source": [
    "layer_fc2 = new_fc_layer(input=layer_fc1,\n",
    "                         num_inputs=fc_size,\n",
    "                         num_outputs=num_classes,\n",
    "                         use_relu=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 135,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<tf.Tensor 'add_14:0' shape=(?, 2) dtype=float32>"
      ]
     },
     "execution_count": 135,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "layer_fc2"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Predicted Class"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The second fully-connected layer estimates how likely it is that the input image belongs to each of the 2 classes. However, these estimates are a bit rough and difficult to interpret because the numbers may be very small or large, so we want to normalize them so that each element is limited between zero and one and the all the elements sum to one. This is calculated using the so-called softmax function and the result is stored in `y_pred`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 136,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Tensor(\"Softmax_2:0\", shape=(?, 2), dtype=float32)\n"
     ]
    }
   ],
   "source": [
    "y_pred = tf.nn.softmax(layer_fc2)\n",
    "print(y_pred)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The class-number is the index of the largest element."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Tensor(\"ArgMax_5:0\", shape=(?,), dtype=int64)\n"
     ]
    }
   ],
   "source": [
    "y_pred_cls = tf.argmax(y_pred, dimension=1)\n",
    "print(y_pred_cls)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Cost-function to be optimized"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To make the model better at classifying the input images, we must somehow change the variables for all the network layers. To do this we first need to know how well the model currently performs by comparing the predicted output of the model `y_pred` to the desired output `y_true`.\n",
    "\n",
    "The cross-entropy is a performance measure used in classification. The cross-entropy is a continuous function that is always positive and if the predicted output of the model exactly matches the desired output then the cross-entropy equals zero. The goal of optimization is therefore to minimize the cross-entropy so it gets as close to zero as possible by changing the variables of the network layers.\n",
    "\n",
    "TensorFlow has a built-in function for calculating the cross-entropy. Note that the function calculates the softmax internally so we must use the output of `layer_fc2` directly rather than `y_pred` which has already had the softmax applied."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "metadata": {},
   "outputs": [],
   "source": [
    "cross_entropy = tf.nn.softmax_cross_entropy_with_logits(logits=layer_fc2,\n",
    "                                                        labels=y_true)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We have now calculated the cross-entropy for each of the image classifications so we have a measure of how well the model performs on each image individually. But in order to use the cross-entropy to guide the optimization of the model's variables we need a single scalar value, so we simply take the average of the cross-entropy for all the image classifications."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 139,
   "metadata": {},
   "outputs": [],
   "source": [
    "cost = tf.reduce_mean(cross_entropy)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Optimization Method"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now that we have a cost measure that must be minimized, we can then create an optimizer. In this case it is the `AdamOptimizer` which is an advanced form of Gradient Descent.\n",
    "\n",
    "Note that optimization is not performed at this point. In fact, nothing is calculated at all, we just add the optimizer-object to the TensorFlow graph for later execution."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 140,
   "metadata": {},
   "outputs": [],
   "source": [
    "optimizer = tf.train.AdamOptimizer(learning_rate=1e-4).minimize(cost)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Performance Measures"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We need a few more performance measures to display the progress to the user.\n",
    "\n",
    "This is a vector of booleans whether the predicted class equals the true class of each image."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 141,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Tensor(\"Equal_2:0\", shape=(?,), dtype=bool)\n"
     ]
    }
   ],
   "source": [
    "correct_prediction = tf.equal(y_pred_cls, y_true_cls)\n",
    "print(correct_prediction)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This calculates the classification accuracy by first type-casting the vector of booleans to floats, so that False becomes 0 and True becomes 1, and then calculating the average of these numbers."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 142,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Tensor(\"Mean_5:0\", shape=(), dtype=float32)\n"
     ]
    }
   ],
   "source": [
    "accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32))\n",
    "print(accuracy)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## TensorFlow Run"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Create TensorFlow session\n",
    "\n",
    "Once the TensorFlow graph has been created, we have to create a TensorFlow session which is used to execute the graph."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 143,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "config = tf.ConfigProto(\n",
    "        device_count = {'GPU': 0}\n",
    "    )\n",
    "session = tf.Session(config=config)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Initialize variables\n",
    "\n",
    "The variables for `weights` and `biases` must be initialized before we start optimizing them."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 144,
   "metadata": {},
   "outputs": [],
   "source": [
    "session.run(tf.initialize_all_variables())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Helper-function to perform optimization iterations"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "It takes a long time to calculate the gradient of the model using the entirety of a large dataset\n",
    ". We therefore only use a small batch of images in each iteration of the optimizer.\n",
    "\n",
    "If your computer crashes or becomes very slow because you run out of RAM, then you may try and lower this number, but you may then need to perform more optimization iterations."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 145,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_batch_size = batch_size"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 146,
   "metadata": {},
   "outputs": [],
   "source": [
    "def print_progress(epoch, feed_dict_train, feed_dict_validate, val_loss):\n",
    "    # Calculate the accuracy on the training-set.\n",
    "    acc = session.run(accuracy, feed_dict=feed_dict_train)\n",
    "    val_acc = session.run(accuracy, feed_dict=feed_dict_validate)\n",
    "    msg = \"Epoch {0} --- Training Accuracy: {1:>6.1%}, Validation Accuracy: {2:>6.1%}, Validation Loss: {3:.3f}\"\n",
    "    print(msg.format(epoch + 1, acc, val_acc, val_loss))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Function for performing a number of optimization iterations so as to gradually improve the variables of the network layers. In each iteration, a new batch of data is selected from the training-set and then TensorFlow executes the optimizer using those training samples.  The progress is printed every 100 iterations."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 147,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Counter for total number of iterations performed so far.\n",
    "total_iterations = 0\n",
    "\n",
    "def optimize(num_iterations):\n",
    "    # Ensure we update the global variable rather than a local copy.\n",
    "    global total_iterations\n",
    "\n",
    "    # Start-time used for printing time-usage below.\n",
    "    start_time = time.time()\n",
    "    \n",
    "    best_val_loss = float(\"inf\")\n",
    "    patience = 0\n",
    "\n",
    "    for i in range(total_iterations,\n",
    "                   total_iterations + num_iterations):\n",
    "\n",
    "        # Get a batch of training examples.\n",
    "        # x_batch now holds a batch of images and\n",
    "        # y_true_batch are the true labels for those images.\n",
    "        x_batch, y_true_batch, _, cls_batch = data.train.next_batch(train_batch_size)\n",
    "        x_valid_batch, y_valid_batch, _, valid_cls_batch = data.valid.next_batch(train_batch_size)\n",
    "\n",
    "        # Convert shape from [num examples, rows, columns, depth]\n",
    "        # to [num examples, flattened image shape]\n",
    "\n",
    "        x_batch = x_batch.reshape(train_batch_size, img_size_flat)\n",
    "        x_valid_batch = x_valid_batch.reshape(train_batch_size, img_size_flat)\n",
    "\n",
    "        # Put the batch into a dict with the proper names\n",
    "        # for placeholder variables in the TensorFlow graph.\n",
    "        feed_dict_train = {x: x_batch,\n",
    "                           y_true: y_true_batch}\n",
    "        \n",
    "        feed_dict_validate = {x: x_valid_batch,\n",
    "                              y_true: y_valid_batch}\n",
    "\n",
    "        # Run the optimizer using this batch of training data.\n",
    "        # TensorFlow assigns the variables in feed_dict_train\n",
    "        # to the placeholder variables and then runs the optimizer.\n",
    "        session.run(optimizer, feed_dict=feed_dict_train)\n",
    "        \n",
    "\n",
    "        # Print status at end of each epoch (defined as full pass through training dataset).\n",
    "        if i % int(data.train.num_examples/batch_size) == 0: \n",
    "            val_loss = session.run(cost, feed_dict=feed_dict_validate)\n",
    "            epoch = int(i / int(data.train.num_examples/batch_size))\n",
    "            \n",
    "            print_progress(epoch, feed_dict_train, feed_dict_validate, val_loss)\n",
    "            \n",
    "            if early_stopping:    \n",
    "                if val_loss < best_val_loss:\n",
    "                    best_val_loss = val_loss\n",
    "                    patience = 0\n",
    "                else:\n",
    "                    patience += 1\n",
    "\n",
    "                if patience == early_stopping:\n",
    "                    break\n",
    "\n",
    "    # Update the total number of iterations performed.\n",
    "    total_iterations += num_iterations\n",
    "\n",
    "    # Ending time.\n",
    "    end_time = time.time()\n",
    "\n",
    "    # Difference between start and end-times.\n",
    "    time_dif = end_time - start_time\n",
    "\n",
    "    # Print the time-usage.\n",
    "    print(\"Time elapsed: \" + str(timedelta(seconds=int(round(time_dif)))))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Helper-function to plot example errors"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Function for plotting examples of images from the test-set that have been mis-classified."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 148,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_example_errors(cls_pred, correct):\n",
    "    # cls_pred is an array of the predicted class-number for\n",
    "    # all images in the test-set.\n",
    "\n",
    "    # correct is a boolean array whether the predicted class\n",
    "    # is equal to the true class for each image in the test-set.\n",
    "\n",
    "    # Negate the boolean array.\n",
    "    incorrect = (correct == False)\n",
    "    \n",
    "    # Get the images from the test-set that have been\n",
    "    # incorrectly classified.\n",
    "    images = data.valid.images[incorrect]\n",
    "    \n",
    "    # Get the predicted classes for those images.\n",
    "    cls_pred = cls_pred[incorrect]\n",
    "\n",
    "    # Get the true classes for those images.\n",
    "    cls_true = data.valid.cls[incorrect]\n",
    "    \n",
    "    # Plot the first 9 images.\n",
    "    plot_images(images=images[0:9],\n",
    "                cls_true=cls_true[0:9],\n",
    "                cls_pred=cls_pred[0:9])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Helper-function to plot confusion matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 149,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_confusion_matrix(cls_pred):\n",
    "    # cls_pred is an array of the predicted class-number for\n",
    "    # all images in the test-set.\n",
    "\n",
    "    # Get the true classifications for the test-set.\n",
    "    cls_true = data.valid.cls\n",
    "    \n",
    "    # Get the confusion matrix using sklearn.\n",
    "    cm = confusion_matrix(y_true=cls_true,\n",
    "                          y_pred=cls_pred)\n",
    "\n",
    "    # Print the confusion matrix as text.\n",
    "    print(cm)\n",
    "\n",
    "    # Plot the confusion matrix as an image.\n",
    "    plt.matshow(cm)\n",
    "\n",
    "    # Make various adjustments to the plot.\n",
    "    plt.colorbar()\n",
    "    tick_marks = np.arange(num_classes)\n",
    "    plt.xticks(tick_marks, range(num_classes))\n",
    "    plt.yticks(tick_marks, range(num_classes))\n",
    "    plt.xlabel('Predicted')\n",
    "    plt.ylabel('True')\n",
    "\n",
    "    # Ensure the plot is shown correctly with multiple plots\n",
    "    # in a single Notebook cell.\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Helper-function for showing the performance"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Function for printing the classification accuracy on the test-set.\n",
    "\n",
    "It takes a while to compute the classification for all the images in the test-set, that's why the results are re-used by calling the above functions directly from this function, so the classifications don't have to be recalculated by each function.\n",
    "\n",
    "Note that this function can use a lot of computer memory, which is why the test-set is split into smaller batches. If you have little RAM in your computer and it crashes, then you can try and lower the batch-size."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 150,
   "metadata": {},
   "outputs": [],
   "source": [
    "def print_validation_accuracy(show_example_errors=False,\n",
    "                        show_confusion_matrix=False):\n",
    "\n",
    "    # Number of images in the test-set.\n",
    "    num_test = len(data.valid.images)\n",
    "   \n",
    "    \n",
    "\n",
    "    # Allocate an array for the predicted classes which\n",
    "    # will be calculated in batches and filled into this array.\n",
    "    cls_pred = np.zeros(shape=num_test, dtype=np.int)\n",
    "    print(cls_pred)\n",
    "\n",
    "    # Now calculate the predicted classes for the batches.\n",
    "    # We will just iterate through all the batches.\n",
    "    # There might be a more clever and Pythonic way of doing this.\n",
    "\n",
    "    # The starting index for the next batch is denoted i.\n",
    "    i = 0\n",
    "\n",
    "    while i < num_test:\n",
    "        # The ending index for the next batch is denoted j.\n",
    "        j = min(i + batch_size, num_test)\n",
    "        \n",
    "        print(j)\n",
    "        print(i)\n",
    "        # Get the images from the test-set between index i and j.\n",
    "        images = data.valid.images[i:j, :].reshape(train_batch_size,49152 )\n",
    "        \n",
    "\n",
    "        # Get the associated labels.\n",
    "        labels = data.valid.labels[i:j, :]\n",
    "        print(labels)\n",
    "\n",
    "        # Create a feed-dict with these images and labels.\n",
    "        feed_dict = {x: images,\n",
    "                     y_true: labels}\n",
    "\n",
    "        # Calculate the predicted class using TensorFlow.\n",
    "        cls_pred[i:j] = session.run(y_pred_cls, feed_dict=feed_dict)\n",
    "        print(cls_pred[i:j])\n",
    "\n",
    "        # Set the start-index for the next batch to the\n",
    "        # end-index of the current batch.\n",
    "        i = j\n",
    "        \n",
    "        \n",
    "    \n",
    "    cls_true = np.array(data.valid.cls)\n",
    "    print(cls_true)\n",
    "    cls_pred = np.array([classes[x] for x in cls_pred]) \n",
    "    print(cls_pred)\n",
    "\n",
    "    # Create a boolean array whether each image is correctly classified.\n",
    "    correct = (cls_true == cls_pred)\n",
    "\n",
    "    # Calculate the number of correctly classified images.\n",
    "    # When summing a boolean array, False means 0 and True means 1.\n",
    "    correct_sum = correct.sum()\n",
    "\n",
    "    # Classification accuracy is the number of correctly classified\n",
    "    # images divided by the total number of images in the test-set.\n",
    "    acc = float(correct_sum) / num_test\n",
    "\n",
    "    # Print the accuracy.\n",
    "    msg = \"Accuracy on Test-Set: {0:.1%} ({1} / {2})\"\n",
    "    print(msg.format(acc, correct_sum, num_test))\n",
    "\n",
    "    # Plot some examples of mis-classifications, if desired.\n",
    "    if show_example_errors:\n",
    "        print(\"Example errors:\")\n",
    "        plot_example_errors(cls_pred=cls_pred, correct=correct)\n",
    "\n",
    "    # Plot the confusion matrix, if desired.\n",
    "    if show_confusion_matrix:\n",
    "        print(\"Confusion Matrix:\")\n",
    "        plot_confusion_matrix(cls_pred=cls_pred)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Performance after 1 optimization iteration"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 151,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1 --- Training Accuracy:  56.2%, Validation Accuracy:  59.4%, Validation Loss: 0.675\n",
      "Time elapsed: 0:00:01\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 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 0 0]\n",
      "32\n",
      "0\n",
      "[[0. 1.]\n",
      " [0. 1.]\n",
      " [1. 0.]\n",
      " [0. 1.]\n",
      " [0. 1.]\n",
      " [1. 0.]\n",
      " [1. 0.]\n",
      " [1. 0.]\n",
      " [1. 0.]\n",
      " [0. 1.]\n",
      " [1. 0.]\n",
      " [1. 0.]\n",
      " [1. 0.]\n",
      " [1. 0.]\n",
      " [0. 1.]\n",
      " [1. 0.]\n",
      " [0. 1.]\n",
      " [1. 0.]\n",
      " [1. 0.]\n",
      " [0. 1.]\n",
      " [1. 0.]\n",
      " [1. 0.]\n",
      " [1. 0.]\n",
      " [0. 1.]\n",
      " [0. 1.]\n",
      " [1. 0.]\n",
      " [1. 0.]\n",
      " [1. 0.]\n",
      " [1. 0.]\n",
      " [0. 1.]\n",
      " [0. 1.]\n",
      " [0. 1.]]\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 0 0 0 0 0 0 0]\n",
      "64\n",
      "32\n",
      "[[1. 0.]\n",
      " [0. 1.]\n",
      " [1. 0.]\n",
      " [0. 1.]\n",
      " [1. 0.]\n",
      " [0. 1.]\n",
      " [1. 0.]\n",
      " [0. 1.]\n",
      " [0. 1.]\n",
      " [1. 0.]\n",
      " [1. 0.]\n",
      " [0. 1.]\n",
      " [1. 0.]\n",
      " [1. 0.]\n",
      " [0. 1.]\n",
      " [0. 1.]\n",
      " [1. 0.]\n",
      " [1. 0.]\n",
      " [0. 1.]\n",
      " [0. 1.]\n",
      " [1. 0.]\n",
      " [1. 0.]\n",
      " [1. 0.]\n",
      " [0. 1.]\n",
      " [0. 1.]\n",
      " [0. 1.]\n",
      " [0. 1.]\n",
      " [0. 1.]\n",
      " [0. 1.]\n",
      " [1. 0.]\n",
      " [1. 0.]\n",
      " [0. 1.]]\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 0 0 0 0 0 0 0]\n",
      "['002.Laysan_Albatross' '002.Laysan_Albatross'\n",
      " '001.Black_footed_Albatross' '002.Laysan_Albatross'\n",
      " '002.Laysan_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '002.Laysan_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '002.Laysan_Albatross' '001.Black_footed_Albatross'\n",
      " '002.Laysan_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '002.Laysan_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '002.Laysan_Albatross'\n",
      " '002.Laysan_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '002.Laysan_Albatross'\n",
      " '002.Laysan_Albatross' '002.Laysan_Albatross'\n",
      " '001.Black_footed_Albatross' '002.Laysan_Albatross'\n",
      " '001.Black_footed_Albatross' '002.Laysan_Albatross'\n",
      " '001.Black_footed_Albatross' '002.Laysan_Albatross'\n",
      " '001.Black_footed_Albatross' '002.Laysan_Albatross'\n",
      " '002.Laysan_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '002.Laysan_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '002.Laysan_Albatross' '002.Laysan_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '002.Laysan_Albatross' '002.Laysan_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '002.Laysan_Albatross'\n",
      " '002.Laysan_Albatross' '002.Laysan_Albatross' '002.Laysan_Albatross'\n",
      " '002.Laysan_Albatross' '002.Laysan_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '002.Laysan_Albatross']\n",
      "['001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross']\n",
      "Accuracy on Test-Set: 53.1% (34 / 64)\n"
     ]
    }
   ],
   "source": [
    "optimize(num_iterations=1)\n",
    "print_validation_accuracy()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Performance after 100 optimization iterations\n",
    "\n",
    "After 100 optimization iterations, the model should have significantly improved its classification accuracy."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 154,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 22 --- Training Accuracy:  65.6%, Validation Accuracy:  59.4%, Validation Loss: 0.671\n",
      "Epoch 23 --- Training Accuracy:  65.6%, Validation Accuracy:  78.1%, Validation Loss: 0.625\n",
      "Epoch 24 --- Training Accuracy:  62.5%, Validation Accuracy:  56.2%, Validation Loss: 0.676\n",
      "Epoch 25 --- Training Accuracy:  62.5%, Validation Accuracy:  78.1%, Validation Loss: 0.615\n",
      "Epoch 26 --- Training Accuracy:  65.6%, Validation Accuracy:  59.4%, Validation Loss: 0.671\n",
      "Epoch 27 --- Training Accuracy:  65.6%, Validation Accuracy:  78.1%, Validation Loss: 0.609\n",
      "Epoch 28 --- Training Accuracy:  71.9%, Validation Accuracy:  68.8%, Validation Loss: 0.655\n",
      "Epoch 29 --- Training Accuracy:  78.1%, Validation Accuracy:  71.9%, Validation Loss: 0.606\n",
      "Epoch 30 --- Training Accuracy:  81.2%, Validation Accuracy:  68.8%, Validation Loss: 0.643\n",
      "Epoch 31 --- Training Accuracy:  81.2%, Validation Accuracy:  71.9%, Validation Loss: 0.597\n",
      "Epoch 32 --- Training Accuracy:  75.0%, Validation Accuracy:  68.8%, Validation Loss: 0.643\n",
      "Epoch 33 --- Training Accuracy:  71.9%, Validation Accuracy:  71.9%, Validation Loss: 0.584\n",
      "Epoch 34 --- Training Accuracy:  75.0%, Validation Accuracy:  68.8%, Validation Loss: 0.641\n",
      "Epoch 35 --- Training Accuracy:  81.2%, Validation Accuracy:  71.9%, Validation Loss: 0.578\n",
      "Epoch 36 --- Training Accuracy:  81.2%, Validation Accuracy:  71.9%, Validation Loss: 0.626\n",
      "Epoch 37 --- Training Accuracy:  84.4%, Validation Accuracy:  68.8%, Validation Loss: 0.577\n",
      "Epoch 38 --- Training Accuracy:  81.2%, Validation Accuracy:  71.9%, Validation Loss: 0.619\n",
      "Epoch 39 --- Training Accuracy:  81.2%, Validation Accuracy:  71.9%, Validation Loss: 0.567\n",
      "Epoch 40 --- Training Accuracy:  81.2%, Validation Accuracy:  71.9%, Validation Loss: 0.621\n",
      "Epoch 41 --- Training Accuracy:  81.2%, Validation Accuracy:  71.9%, Validation Loss: 0.560\n",
      "Epoch 42 --- Training Accuracy:  84.4%, Validation Accuracy:  71.9%, Validation Loss: 0.613\n",
      "Epoch 43 --- Training Accuracy:  81.2%, Validation Accuracy:  68.8%, Validation Loss: 0.560\n",
      "Epoch 44 --- Training Accuracy:  81.2%, Validation Accuracy:  68.8%, Validation Loss: 0.608\n",
      "Epoch 45 --- Training Accuracy:  81.2%, Validation Accuracy:  68.8%, Validation Loss: 0.555\n",
      "Epoch 46 --- Training Accuracy:  81.2%, Validation Accuracy:  71.9%, Validation Loss: 0.612\n",
      "Epoch 47 --- Training Accuracy:  81.2%, Validation Accuracy:  68.8%, Validation Loss: 0.553\n",
      "Epoch 48 --- Training Accuracy:  81.2%, Validation Accuracy:  68.8%, Validation Loss: 0.604\n",
      "Epoch 49 --- Training Accuracy:  84.4%, Validation Accuracy:  68.8%, Validation Loss: 0.553\n",
      "Epoch 50 --- Training Accuracy:  81.2%, Validation Accuracy:  68.8%, Validation Loss: 0.605\n",
      "Epoch 51 --- Training Accuracy:  81.2%, Validation Accuracy:  68.8%, Validation Loss: 0.548\n",
      "Epoch 52 --- Training Accuracy:  84.4%, Validation Accuracy:  68.8%, Validation Loss: 0.600\n",
      "Epoch 53 --- Training Accuracy:  84.4%, Validation Accuracy:  71.9%, Validation Loss: 0.548\n",
      "Epoch 54 --- Training Accuracy:  87.5%, Validation Accuracy:  68.8%, Validation Loss: 0.593\n",
      "Epoch 55 --- Training Accuracy:  87.5%, Validation Accuracy:  68.8%, Validation Loss: 0.542\n",
      "Epoch 56 --- Training Accuracy:  90.6%, Validation Accuracy:  68.8%, Validation Loss: 0.590\n",
      "Epoch 57 --- Training Accuracy:  90.6%, Validation Accuracy:  71.9%, Validation Loss: 0.539\n",
      "Epoch 58 --- Training Accuracy:  90.6%, Validation Accuracy:  68.8%, Validation Loss: 0.580\n",
      "Epoch 59 --- Training Accuracy:  90.6%, Validation Accuracy:  71.9%, Validation Loss: 0.533\n",
      "Epoch 60 --- Training Accuracy:  90.6%, Validation Accuracy:  68.8%, Validation Loss: 0.575\n",
      "Epoch 61 --- Training Accuracy:  90.6%, Validation Accuracy:  71.9%, Validation Loss: 0.526\n",
      "Epoch 62 --- Training Accuracy:  90.6%, Validation Accuracy:  68.8%, Validation Loss: 0.563\n",
      "Epoch 63 --- Training Accuracy:  90.6%, Validation Accuracy:  71.9%, Validation Loss: 0.519\n",
      "Epoch 64 --- Training Accuracy:  90.6%, Validation Accuracy:  68.8%, Validation Loss: 0.557\n",
      "Epoch 65 --- Training Accuracy:  90.6%, Validation Accuracy:  71.9%, Validation Loss: 0.511\n",
      "Epoch 66 --- Training Accuracy:  90.6%, Validation Accuracy:  68.8%, Validation Loss: 0.545\n",
      "Epoch 67 --- Training Accuracy:  90.6%, Validation Accuracy:  71.9%, Validation Loss: 0.504\n",
      "Epoch 68 --- Training Accuracy:  90.6%, Validation Accuracy:  68.8%, Validation Loss: 0.538\n",
      "Epoch 69 --- Training Accuracy:  90.6%, Validation Accuracy:  75.0%, Validation Loss: 0.496\n",
      "Epoch 70 --- Training Accuracy:  90.6%, Validation Accuracy:  71.9%, Validation Loss: 0.525\n",
      "Epoch 71 --- Training Accuracy:  93.8%, Validation Accuracy:  75.0%, Validation Loss: 0.489\n",
      "Epoch 72 --- Training Accuracy:  93.8%, Validation Accuracy:  78.1%, Validation Loss: 0.516\n",
      "Epoch 73 --- Training Accuracy:  96.9%, Validation Accuracy:  71.9%, Validation Loss: 0.483\n",
      "Epoch 74 --- Training Accuracy:  96.9%, Validation Accuracy:  78.1%, Validation Loss: 0.503\n",
      "Epoch 75 --- Training Accuracy:  96.9%, Validation Accuracy:  71.9%, Validation Loss: 0.477\n",
      "Epoch 76 --- Training Accuracy:  96.9%, Validation Accuracy:  78.1%, Validation Loss: 0.494\n",
      "Epoch 77 --- Training Accuracy: 100.0%, Validation Accuracy:  68.8%, Validation Loss: 0.475\n",
      "Epoch 78 --- Training Accuracy: 100.0%, Validation Accuracy:  78.1%, Validation Loss: 0.487\n",
      "Epoch 79 --- Training Accuracy: 100.0%, Validation Accuracy:  65.6%, Validation Loss: 0.472\n",
      "Epoch 80 --- Training Accuracy: 100.0%, Validation Accuracy:  78.1%, Validation Loss: 0.481\n",
      "Epoch 81 --- Training Accuracy: 100.0%, Validation Accuracy:  65.6%, Validation Loss: 0.470\n",
      "Epoch 82 --- Training Accuracy: 100.0%, Validation Accuracy:  78.1%, Validation Loss: 0.477\n",
      "Epoch 83 --- Training Accuracy: 100.0%, Validation Accuracy:  65.6%, Validation Loss: 0.470\n",
      "Epoch 84 --- Training Accuracy: 100.0%, Validation Accuracy:  81.2%, Validation Loss: 0.474\n",
      "Epoch 85 --- Training Accuracy: 100.0%, Validation Accuracy:  65.6%, Validation Loss: 0.471\n",
      "Epoch 86 --- Training Accuracy: 100.0%, Validation Accuracy:  81.2%, Validation Loss: 0.468\n",
      "Epoch 87 --- Training Accuracy: 100.0%, Validation Accuracy:  68.8%, Validation Loss: 0.474\n",
      "Epoch 88 --- Training Accuracy: 100.0%, Validation Accuracy:  81.2%, Validation Loss: 0.467\n",
      "Epoch 89 --- Training Accuracy: 100.0%, Validation Accuracy:  68.8%, Validation Loss: 0.481\n",
      "Epoch 90 --- Training Accuracy: 100.0%, Validation Accuracy:  81.2%, Validation Loss: 0.467\n",
      "Epoch 91 --- Training Accuracy: 100.0%, Validation Accuracy:  71.9%, Validation Loss: 0.488\n",
      "Epoch 92 --- Training Accuracy: 100.0%, Validation Accuracy:  81.2%, Validation Loss: 0.464\n",
      "Epoch 93 --- Training Accuracy: 100.0%, Validation Accuracy:  71.9%, Validation Loss: 0.493\n",
      "Epoch 94 --- Training Accuracy: 100.0%, Validation Accuracy:  84.4%, Validation Loss: 0.463\n",
      "Epoch 95 --- Training Accuracy: 100.0%, Validation Accuracy:  71.9%, Validation Loss: 0.498\n",
      "Epoch 96 --- Training Accuracy: 100.0%, Validation Accuracy:  84.4%, Validation Loss: 0.465\n",
      "Epoch 97 --- Training Accuracy: 100.0%, Validation Accuracy:  71.9%, Validation Loss: 0.502\n",
      "Epoch 98 --- Training Accuracy: 100.0%, Validation Accuracy:  81.2%, Validation Loss: 0.465\n",
      "Epoch 99 --- Training Accuracy: 100.0%, Validation Accuracy:  71.9%, Validation Loss: 0.509\n",
      "Epoch 100 --- Training Accuracy: 100.0%, Validation Accuracy:  81.2%, Validation Loss: 0.469\n",
      "Epoch 101 --- Training Accuracy: 100.0%, Validation Accuracy:  71.9%, Validation Loss: 0.523\n",
      "Epoch 102 --- Training Accuracy: 100.0%, Validation Accuracy:  81.2%, Validation Loss: 0.479\n",
      "Epoch 103 --- Training Accuracy: 100.0%, Validation Accuracy:  71.9%, Validation Loss: 0.538\n",
      "Epoch 104 --- Training Accuracy: 100.0%, Validation Accuracy:  81.2%, Validation Loss: 0.486\n",
      "Epoch 105 --- Training Accuracy: 100.0%, Validation Accuracy:  71.9%, Validation Loss: 0.548\n",
      "Epoch 106 --- Training Accuracy: 100.0%, Validation Accuracy:  81.2%, Validation Loss: 0.490\n",
      "Epoch 107 --- Training Accuracy: 100.0%, Validation Accuracy:  71.9%, Validation Loss: 0.557\n",
      "Epoch 108 --- Training Accuracy: 100.0%, Validation Accuracy:  81.2%, Validation Loss: 0.497\n",
      "Epoch 109 --- Training Accuracy: 100.0%, Validation Accuracy:  71.9%, Validation Loss: 0.572\n",
      "Epoch 110 --- Training Accuracy: 100.0%, Validation Accuracy:  81.2%, Validation Loss: 0.513\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 111 --- Training Accuracy: 100.0%, Validation Accuracy:  71.9%, Validation Loss: 0.582\n",
      "Epoch 112 --- Training Accuracy: 100.0%, Validation Accuracy:  81.2%, Validation Loss: 0.516\n",
      "Epoch 113 --- Training Accuracy: 100.0%, Validation Accuracy:  71.9%, Validation Loss: 0.589\n",
      "Epoch 114 --- Training Accuracy: 100.0%, Validation Accuracy:  81.2%, Validation Loss: 0.526\n",
      "Epoch 115 --- Training Accuracy: 100.0%, Validation Accuracy:  71.9%, Validation Loss: 0.603\n",
      "Epoch 116 --- Training Accuracy: 100.0%, Validation Accuracy:  81.2%, Validation Loss: 0.537\n",
      "Epoch 117 --- Training Accuracy: 100.0%, Validation Accuracy:  71.9%, Validation Loss: 0.611\n",
      "Epoch 118 --- Training Accuracy: 100.0%, Validation Accuracy:  81.2%, Validation Loss: 0.545\n",
      "Epoch 119 --- Training Accuracy: 100.0%, Validation Accuracy:  71.9%, Validation Loss: 0.626\n",
      "Epoch 120 --- Training Accuracy: 100.0%, Validation Accuracy:  78.1%, Validation Loss: 0.563\n",
      "Epoch 121 --- Training Accuracy: 100.0%, Validation Accuracy:  71.9%, Validation Loss: 0.639\n",
      "Time elapsed: 0:02:04\n"
     ]
    }
   ],
   "source": [
    "optimize(num_iterations=100)  # We already performed 1 iteration above."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 156,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[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 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 0 0]\n",
      "32\n",
      "0\n",
      "[[0. 1.]\n",
      " [0. 1.]\n",
      " [1. 0.]\n",
      " [0. 1.]\n",
      " [0. 1.]\n",
      " [1. 0.]\n",
      " [1. 0.]\n",
      " [1. 0.]\n",
      " [1. 0.]\n",
      " [0. 1.]\n",
      " [1. 0.]\n",
      " [1. 0.]\n",
      " [1. 0.]\n",
      " [1. 0.]\n",
      " [0. 1.]\n",
      " [1. 0.]\n",
      " [0. 1.]\n",
      " [1. 0.]\n",
      " [1. 0.]\n",
      " [0. 1.]\n",
      " [1. 0.]\n",
      " [1. 0.]\n",
      " [1. 0.]\n",
      " [0. 1.]\n",
      " [0. 1.]\n",
      " [1. 0.]\n",
      " [1. 0.]\n",
      " [1. 0.]\n",
      " [1. 0.]\n",
      " [0. 1.]\n",
      " [0. 1.]\n",
      " [0. 1.]]\n",
      "[1 0 1 1 1 1 0 0 1 1 0 0 1 0 1 0 0 0 1 0 0 0 0 1 1 0 0 1 0 1 1 1]\n",
      "64\n",
      "32\n",
      "[[1. 0.]\n",
      " [0. 1.]\n",
      " [1. 0.]\n",
      " [0. 1.]\n",
      " [1. 0.]\n",
      " [0. 1.]\n",
      " [1. 0.]\n",
      " [0. 1.]\n",
      " [0. 1.]\n",
      " [1. 0.]\n",
      " [1. 0.]\n",
      " [0. 1.]\n",
      " [1. 0.]\n",
      " [1. 0.]\n",
      " [0. 1.]\n",
      " [0. 1.]\n",
      " [1. 0.]\n",
      " [1. 0.]\n",
      " [0. 1.]\n",
      " [0. 1.]\n",
      " [1. 0.]\n",
      " [1. 0.]\n",
      " [1. 0.]\n",
      " [0. 1.]\n",
      " [0. 1.]\n",
      " [0. 1.]\n",
      " [0. 1.]\n",
      " [0. 1.]\n",
      " [0. 1.]\n",
      " [1. 0.]\n",
      " [1. 0.]\n",
      " [0. 1.]]\n",
      "[0 1 0 1 0 1 0 1 1 1 0 0 0 0 1 0 0 0 0 0 0 0 0 1 0 0 1 1 1 0 0 1]\n",
      "['002.Laysan_Albatross' '002.Laysan_Albatross'\n",
      " '001.Black_footed_Albatross' '002.Laysan_Albatross'\n",
      " '002.Laysan_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '002.Laysan_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '002.Laysan_Albatross' '001.Black_footed_Albatross'\n",
      " '002.Laysan_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '002.Laysan_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '002.Laysan_Albatross'\n",
      " '002.Laysan_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '002.Laysan_Albatross'\n",
      " '002.Laysan_Albatross' '002.Laysan_Albatross'\n",
      " '001.Black_footed_Albatross' '002.Laysan_Albatross'\n",
      " '001.Black_footed_Albatross' '002.Laysan_Albatross'\n",
      " '001.Black_footed_Albatross' '002.Laysan_Albatross'\n",
      " '001.Black_footed_Albatross' '002.Laysan_Albatross'\n",
      " '002.Laysan_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '002.Laysan_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '002.Laysan_Albatross' '002.Laysan_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '002.Laysan_Albatross' '002.Laysan_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '002.Laysan_Albatross'\n",
      " '002.Laysan_Albatross' '002.Laysan_Albatross' '002.Laysan_Albatross'\n",
      " '002.Laysan_Albatross' '002.Laysan_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '002.Laysan_Albatross']\n",
      "['002.Laysan_Albatross' '001.Black_footed_Albatross'\n",
      " '002.Laysan_Albatross' '002.Laysan_Albatross' '002.Laysan_Albatross'\n",
      " '002.Laysan_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '002.Laysan_Albatross'\n",
      " '002.Laysan_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '002.Laysan_Albatross'\n",
      " '001.Black_footed_Albatross' '002.Laysan_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '002.Laysan_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '002.Laysan_Albatross' '002.Laysan_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '002.Laysan_Albatross' '001.Black_footed_Albatross'\n",
      " '002.Laysan_Albatross' '002.Laysan_Albatross' '002.Laysan_Albatross'\n",
      " '001.Black_footed_Albatross' '002.Laysan_Albatross'\n",
      " '001.Black_footed_Albatross' '002.Laysan_Albatross'\n",
      " '001.Black_footed_Albatross' '002.Laysan_Albatross'\n",
      " '001.Black_footed_Albatross' '002.Laysan_Albatross'\n",
      " '002.Laysan_Albatross' '002.Laysan_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '002.Laysan_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '002.Laysan_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '002.Laysan_Albatross' '002.Laysan_Albatross' '002.Laysan_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '002.Laysan_Albatross']\n",
      "Accuracy on Test-Set: 75.0% (48 / 64)\n",
      "Confusion Matrix:\n",
      "[[27  7]\n",
      " [ 9 21]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1c34c09deb8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "print_validation_accuracy(show_confusion_matrix=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Performance after 1000 optimization iterations"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 10 --- Training Accuracy:  31.2%, Validation Accuracy:  25.0%, Validation Loss: 2.172\n",
      "Epoch 11 --- Training Accuracy:  37.5%, Validation Accuracy:  25.0%, Validation Loss: 2.159\n",
      "Epoch 12 --- Training Accuracy:  40.6%, Validation Accuracy:  28.1%, Validation Loss: 2.150\n",
      "Epoch 13 --- Training Accuracy:  37.5%, Validation Accuracy:  28.1%, Validation Loss: 2.140\n",
      "Epoch 14 --- Training Accuracy:  37.5%, Validation Accuracy:  28.1%, Validation Loss: 2.145\n",
      "Epoch 15 --- Training Accuracy:  43.8%, Validation Accuracy:  25.0%, Validation Loss: 2.150\n",
      "Epoch 16 --- Training Accuracy:  43.8%, Validation Accuracy:  25.0%, Validation Loss: 2.154\n",
      "Epoch 17 --- Training Accuracy:  43.8%, Validation Accuracy:  25.0%, Validation Loss: 2.158\n",
      "Epoch 18 --- Training Accuracy:  53.1%, Validation Accuracy:  25.0%, Validation Loss: 2.166\n",
      "Epoch 19 --- Training Accuracy:  53.1%, Validation Accuracy:  25.0%, Validation Loss: 2.175\n",
      "Epoch 20 --- Training Accuracy:  53.1%, Validation Accuracy:  28.1%, Validation Loss: 2.188\n",
      "Epoch 21 --- Training Accuracy:  53.1%, Validation Accuracy:  28.1%, Validation Loss: 2.200\n",
      "Epoch 22 --- Training Accuracy:  53.1%, Validation Accuracy:  28.1%, Validation Loss: 2.213\n",
      "Epoch 23 --- Training Accuracy:  53.1%, Validation Accuracy:  28.1%, Validation Loss: 2.221\n",
      "Epoch 24 --- Training Accuracy:  53.1%, Validation Accuracy:  28.1%, Validation Loss: 2.236\n",
      "Epoch 25 --- Training Accuracy:  53.1%, Validation Accuracy:  28.1%, Validation Loss: 2.244\n",
      "Epoch 26 --- Training Accuracy:  56.2%, Validation Accuracy:  28.1%, Validation Loss: 2.270\n",
      "Epoch 27 --- Training Accuracy:  62.5%, Validation Accuracy:  28.1%, Validation Loss: 2.289\n",
      "Epoch 28 --- Training Accuracy:  62.5%, Validation Accuracy:  28.1%, Validation Loss: 2.298\n",
      "Epoch 29 --- Training Accuracy:  65.6%, Validation Accuracy:  28.1%, Validation Loss: 2.318\n",
      "Epoch 30 --- Training Accuracy:  71.9%, Validation Accuracy:  28.1%, Validation Loss: 2.333\n",
      "Epoch 31 --- Training Accuracy:  71.9%, Validation Accuracy:  28.1%, Validation Loss: 2.353\n",
      "Epoch 32 --- Training Accuracy:  71.9%, Validation Accuracy:  28.1%, Validation Loss: 2.377\n",
      "Epoch 33 --- Training Accuracy:  75.0%, Validation Accuracy:  31.2%, Validation Loss: 2.391\n",
      "Epoch 34 --- Training Accuracy:  75.0%, Validation Accuracy:  31.2%, Validation Loss: 2.414\n",
      "Epoch 35 --- Training Accuracy:  75.0%, Validation Accuracy:  34.4%, Validation Loss: 2.431\n",
      "Epoch 36 --- Training Accuracy:  75.0%, Validation Accuracy:  34.4%, Validation Loss: 2.446\n",
      "Epoch 37 --- Training Accuracy:  75.0%, Validation Accuracy:  34.4%, Validation Loss: 2.467\n",
      "Epoch 38 --- Training Accuracy:  75.0%, Validation Accuracy:  34.4%, Validation Loss: 2.489\n",
      "Epoch 39 --- Training Accuracy:  75.0%, Validation Accuracy:  34.4%, Validation Loss: 2.508\n",
      "Epoch 40 --- Training Accuracy:  78.1%, Validation Accuracy:  34.4%, Validation Loss: 2.512\n",
      "Epoch 41 --- Training Accuracy:  78.1%, Validation Accuracy:  34.4%, Validation Loss: 2.536\n",
      "Epoch 42 --- Training Accuracy:  75.0%, Validation Accuracy:  34.4%, Validation Loss: 2.579\n",
      "Epoch 43 --- Training Accuracy:  75.0%, Validation Accuracy:  34.4%, Validation Loss: 2.586\n",
      "Epoch 44 --- Training Accuracy:  78.1%, Validation Accuracy:  34.4%, Validation Loss: 2.628\n",
      "Epoch 45 --- Training Accuracy:  78.1%, Validation Accuracy:  34.4%, Validation Loss: 2.622\n",
      "Time elapsed: 0:07:48\n"
     ]
    }
   ],
   "source": [
    "optimize(num_iterations=500)  # We performed 100 iterations above."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[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 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 0 0]\n",
      "32\n",
      "0\n",
      "[[0. 0. 0. 0. 1. 0. 0. 0. 0. 0.]\n",
      " [0. 0. 0. 0. 1. 0. 0. 0. 0. 0.]\n",
      " [0. 0. 1. 0. 0. 0. 0. 0. 0. 0.]\n",
      " [0. 0. 0. 0. 0. 1. 0. 0. 0. 0.]\n",
      " [0. 0. 0. 0. 0. 0. 0. 0. 0. 1.]\n",
      " [0. 0. 0. 1. 0. 0. 0. 0. 0. 0.]\n",
      " [0. 0. 0. 0. 0. 1. 0. 0. 0. 0.]\n",
      " [0. 0. 0. 0. 0. 0. 0. 1. 0. 0.]\n",
      " [0. 0. 0. 0. 0. 0. 0. 1. 0. 0.]\n",
      " [1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      " [0. 0. 0. 0. 0. 0. 0. 0. 0. 1.]\n",
      " [0. 0. 0. 0. 0. 0. 1. 0. 0. 0.]\n",
      " [0. 0. 0. 0. 0. 0. 0. 0. 1. 0.]\n",
      " [0. 0. 0. 0. 1. 0. 0. 0. 0. 0.]\n",
      " [0. 0. 0. 0. 0. 1. 0. 0. 0. 0.]\n",
      " [1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      " [0. 0. 0. 0. 0. 0. 0. 0. 1. 0.]\n",
      " [0. 0. 0. 0. 0. 0. 0. 0. 0. 1.]\n",
      " [0. 0. 0. 0. 0. 0. 0. 0. 1. 0.]\n",
      " [0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      " [0. 0. 0. 0. 0. 0. 1. 0. 0. 0.]\n",
      " [0. 0. 0. 0. 0. 0. 1. 0. 0. 0.]\n",
      " [0. 0. 0. 0. 0. 1. 0. 0. 0. 0.]\n",
      " [0. 0. 0. 0. 0. 0. 1. 0. 0. 0.]\n",
      " [0. 0. 0. 0. 0. 0. 0. 0. 1. 0.]\n",
      " [0. 0. 0. 0. 0. 0. 1. 0. 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",
      " [0. 0. 0. 0. 0. 1. 0. 0. 0. 0.]\n",
      " [0. 0. 0. 0. 0. 1. 0. 0. 0. 0.]\n",
      " [0. 0. 0. 0. 0. 0. 0. 0. 1. 0.]\n",
      " [1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]]\n",
      "[0 1 0 0 3 8 6 6 0 0 0 6 9 3 9 6 0 3 2 0 6 6 2 8 3 6 0 9 1 1 6 0]\n",
      "64\n",
      "32\n",
      "[[0. 0. 0. 0. 0. 0. 1. 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. 0. 0. 0. 0. 0. 0. 1. 0.]\n",
      " [0. 0. 0. 0. 0. 0. 0. 1. 0. 0.]\n",
      " [0. 0. 0. 0. 0. 0. 0. 1. 0. 0.]\n",
      " [0. 0. 0. 0. 0. 0. 0. 0. 0. 1.]\n",
      " [0. 0. 0. 0. 0. 1. 0. 0. 0. 0.]\n",
      " [0. 0. 0. 0. 0. 0. 0. 1. 0. 0.]\n",
      " [0. 0. 0. 0. 0. 0. 0. 0. 0. 1.]\n",
      " [0. 0. 0. 0. 0. 0. 0. 0. 1. 0.]\n",
      " [0. 0. 0. 0. 1. 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. 0. 1.]\n",
      " [0. 0. 0. 0. 0. 1. 0. 0. 0. 0.]\n",
      " [0. 0. 0. 0. 1. 0. 0. 0. 0. 0.]\n",
      " [0. 0. 0. 1. 0. 0. 0. 0. 0. 0.]\n",
      " [0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      " [0. 0. 0. 0. 0. 0. 0. 1. 0. 0.]\n",
      " [0. 0. 1. 0. 0. 0. 0. 0. 0. 0.]\n",
      " [0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      " [0. 0. 0. 0. 0. 0. 0. 0. 1. 0.]\n",
      " [0. 0. 0. 0. 0. 0. 0. 0. 1. 0.]\n",
      " [0. 0. 0. 0. 0. 0. 0. 0. 0. 1.]\n",
      " [0. 0. 0. 0. 0. 0. 1. 0. 0. 0.]\n",
      " [0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      " [0. 0. 0. 1. 0. 0. 0. 0. 0. 0.]\n",
      " [0. 0. 0. 1. 0. 0. 0. 0. 0. 0.]\n",
      " [0. 0. 0. 0. 0. 0. 0. 0. 0. 1.]\n",
      " [0. 0. 0. 0. 1. 0. 0. 0. 0. 0.]\n",
      " [0. 0. 0. 1. 0. 0. 0. 0. 0. 0.]\n",
      " [0. 0. 0. 0. 0. 0. 0. 0. 1. 0.]]\n",
      "[8 0 0 0 2 2 2 1 0 6 3 6 2 9 0 0 0 2 0 0 0 2 0 9 2 6 9 3 9 0 3 3]\n",
      "['005.Crested_Auklet' '005.Crested_Auklet' '003.Sooty_Albatross'\n",
      " '006.Least_Auklet' '010.Red_winged_Blackbird' '004.Groove_billed_Ani'\n",
      " '006.Least_Auklet' '008.Rhinoceros_Auklet' '008.Rhinoceros_Auklet'\n",
      " '001.Black_footed_Albatross' '010.Red_winged_Blackbird'\n",
      " '007.Parakeet_Auklet' '009.Brewer_Blackbird' '005.Crested_Auklet'\n",
      " '006.Least_Auklet' '001.Black_footed_Albatross' '009.Brewer_Blackbird'\n",
      " '010.Red_winged_Blackbird' '009.Brewer_Blackbird' '002.Laysan_Albatross'\n",
      " '007.Parakeet_Auklet' '007.Parakeet_Auklet' '006.Least_Auklet'\n",
      " '007.Parakeet_Auklet' '009.Brewer_Blackbird' '007.Parakeet_Auklet'\n",
      " '009.Brewer_Blackbird' '004.Groove_billed_Ani' '006.Least_Auklet'\n",
      " '006.Least_Auklet' '009.Brewer_Blackbird' '001.Black_footed_Albatross'\n",
      " '007.Parakeet_Auklet' '002.Laysan_Albatross' '002.Laysan_Albatross'\n",
      " '009.Brewer_Blackbird' '008.Rhinoceros_Auklet' '008.Rhinoceros_Auklet'\n",
      " '010.Red_winged_Blackbird' '006.Least_Auklet' '008.Rhinoceros_Auklet'\n",
      " '010.Red_winged_Blackbird' '009.Brewer_Blackbird' '005.Crested_Auklet'\n",
      " '008.Rhinoceros_Auklet' '010.Red_winged_Blackbird' '006.Least_Auklet'\n",
      " '005.Crested_Auklet' '004.Groove_billed_Ani' '002.Laysan_Albatross'\n",
      " '008.Rhinoceros_Auklet' '003.Sooty_Albatross' '002.Laysan_Albatross'\n",
      " '009.Brewer_Blackbird' '009.Brewer_Blackbird' '010.Red_winged_Blackbird'\n",
      " '007.Parakeet_Auklet' '002.Laysan_Albatross' '004.Groove_billed_Ani'\n",
      " '004.Groove_billed_Ani' '010.Red_winged_Blackbird' '005.Crested_Auklet'\n",
      " '004.Groove_billed_Ani' '009.Brewer_Blackbird']\n",
      "['001.Black_footed_Albatross' '002.Laysan_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '004.Groove_billed_Ani' '009.Brewer_Blackbird' '007.Parakeet_Auklet'\n",
      " '007.Parakeet_Auklet' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '007.Parakeet_Auklet' '010.Red_winged_Blackbird' '004.Groove_billed_Ani'\n",
      " '010.Red_winged_Blackbird' '007.Parakeet_Auklet'\n",
      " '001.Black_footed_Albatross' '004.Groove_billed_Ani'\n",
      " '003.Sooty_Albatross' '001.Black_footed_Albatross' '007.Parakeet_Auklet'\n",
      " '007.Parakeet_Auklet' '003.Sooty_Albatross' '009.Brewer_Blackbird'\n",
      " '004.Groove_billed_Ani' '007.Parakeet_Auklet'\n",
      " '001.Black_footed_Albatross' '010.Red_winged_Blackbird'\n",
      " '002.Laysan_Albatross' '002.Laysan_Albatross' '007.Parakeet_Auklet'\n",
      " '001.Black_footed_Albatross' '009.Brewer_Blackbird'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '003.Sooty_Albatross' '003.Sooty_Albatross'\n",
      " '003.Sooty_Albatross' '002.Laysan_Albatross' '001.Black_footed_Albatross'\n",
      " '007.Parakeet_Auklet' '004.Groove_billed_Ani' '007.Parakeet_Auklet'\n",
      " '003.Sooty_Albatross' '010.Red_winged_Blackbird'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '003.Sooty_Albatross'\n",
      " '001.Black_footed_Albatross' '001.Black_footed_Albatross'\n",
      " '001.Black_footed_Albatross' '003.Sooty_Albatross'\n",
      " '001.Black_footed_Albatross' '010.Red_winged_Blackbird'\n",
      " '003.Sooty_Albatross' '007.Parakeet_Auklet' '010.Red_winged_Blackbird'\n",
      " '004.Groove_billed_Ani' '010.Red_winged_Blackbird'\n",
      " '001.Black_footed_Albatross' '004.Groove_billed_Ani'\n",
      " '004.Groove_billed_Ani']\n",
      "Accuracy on Test-Set: 17.2% (11 / 64)\n",
      "Confusion Matrix:\n",
      "[[2 0 0 0 0 0 1 0 0 0]\n",
      " [4 0 1 0 0 0 1 0 0 0]\n",
      " [2 0 0 0 0 0 0 0 0 0]\n",
      " [1 0 0 2 0 0 0 0 1 2]\n",
      " [3 1 0 1 0 0 1 0 0 0]\n",
      " [2 3 1 0 0 0 1 0 0 1]\n",
      " [0 0 1 0 0 0 4 0 2 0]\n",
      " [3 0 3 0 0 0 1 0 0 0]\n",
      " [4 0 2 3 0 0 1 0 0 1]\n",
      " [1 0 1 2 0 0 1 0 0 3]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1c35741cac8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "print_validation_accuracy(show_confusion_matrix=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Test on Sample Image"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x25c01562ba8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.axis('off')\n",
    "\n",
    "test_cat = cv2.imread('abc.jpg')\n",
    "test_cat = cv2.resize(test_cat, (img_size, img_size), cv2.INTER_LINEAR) / 255\n",
    "\n",
    "preview_cat = plt.imshow(test_cat.reshape(img_size, img_size, num_channels))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x25c015c5f98>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "test_dog = cv2.imread('def.jpg')\n",
    "test_dog = cv2.resize(test_dog, (img_size, img_size), cv2.INTER_LINEAR) / 255\n",
    "\n",
    "preview_dog = plt.imshow(test_dog.reshape(img_size, img_size, num_channels))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "metadata": {},
   "outputs": [
    {
     "ename": "ValueError",
     "evalue": "Cannot feed value of shape (1, 2) for Tensor 'y_true_1:0', which has shape '(?, 17)'",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mValueError\u001b[0m                                Traceback (most recent call last)",
      "\u001b[1;32m<ipython-input-120-3a50bbdc648a>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m()\u001b[0m\n\u001b[0;32m      9\u001b[0m     \u001b[1;32mreturn\u001b[0m \u001b[0mclasses\u001b[0m\u001b[1;33m[\u001b[0m\u001b[0mtest_pred\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;36m0\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     10\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 11\u001b[1;33m \u001b[0mprint\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;34m\"Predicted class for test_cat: {}\"\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mformat\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0msample_prediction\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mtest_cat\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     12\u001b[0m \u001b[0mprint\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;34m\"Predicted class for test_dog: {}\"\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mformat\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0msample_prediction\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mtest_dog\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m<ipython-input-120-3a50bbdc648a>\u001b[0m in \u001b[0;36msample_prediction\u001b[1;34m(test_im)\u001b[0m\n\u001b[0;32m      6\u001b[0m     }\n\u001b[0;32m      7\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m----> 8\u001b[1;33m     \u001b[0mtest_pred\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0msession\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mrun\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0my_pred_cls\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mfeed_dict\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mfeed_dict_test\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m      9\u001b[0m     \u001b[1;32mreturn\u001b[0m \u001b[0mclasses\u001b[0m\u001b[1;33m[\u001b[0m\u001b[0mtest_pred\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;36m0\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     10\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\lib\\site-packages\\tensorflow\\python\\client\\session.py\u001b[0m in \u001b[0;36mrun\u001b[1;34m(self, fetches, feed_dict, options, run_metadata)\u001b[0m\n\u001b[0;32m    903\u001b[0m     \u001b[1;32mtry\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    904\u001b[0m       result = self._run(None, fetches, feed_dict, options_ptr,\n\u001b[1;32m--> 905\u001b[1;33m                          run_metadata_ptr)\n\u001b[0m\u001b[0;32m    906\u001b[0m       \u001b[1;32mif\u001b[0m \u001b[0mrun_metadata\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    907\u001b[0m         \u001b[0mproto_data\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mtf_session\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mTF_GetBuffer\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mrun_metadata_ptr\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\lib\\site-packages\\tensorflow\\python\\client\\session.py\u001b[0m in \u001b[0;36m_run\u001b[1;34m(self, handle, fetches, feed_dict, options, run_metadata)\u001b[0m\n\u001b[0;32m   1114\u001b[0m                              \u001b[1;34m'which has shape %r'\u001b[0m \u001b[1;33m%\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1115\u001b[0m                              (np_val.shape, subfeed_t.name,\n\u001b[1;32m-> 1116\u001b[1;33m                               str(subfeed_t.get_shape())))\n\u001b[0m\u001b[0;32m   1117\u001b[0m           \u001b[1;32mif\u001b[0m \u001b[1;32mnot\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mgraph\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mis_feedable\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0msubfeed_t\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1118\u001b[0m             \u001b[1;32mraise\u001b[0m \u001b[0mValueError\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;34m'Tensor %s may not be fed.'\u001b[0m \u001b[1;33m%\u001b[0m \u001b[0msubfeed_t\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;31mValueError\u001b[0m: Cannot feed value of shape (1, 2) for Tensor 'y_true_1:0', which has shape '(?, 17)'"
     ]
    }
   ],
   "source": [
    "def sample_prediction(test_im):\n",
    "    \n",
    "    feed_dict_test = {\n",
    "        x: test_im.reshape(1, img_size_flat),\n",
    "        y_true: np.array([[1, 0]])\n",
    "    }\n",
    "\n",
    "    test_pred = session.run(y_pred_cls, feed_dict=feed_dict_test)\n",
    "    return classes[test_pred[0]]\n",
    "\n",
    "print(\"Predicted class for test_cat: {}\".format(sample_prediction(test_cat)))\n",
    "print(\"Predicted class for test_dog: {}\".format(sample_prediction(test_dog)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Helper-function for plotting the output of a convolutional layer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Input Images"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Helper-function for plotting an image."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Plot an image from the test-set which will be used as an example below."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Plot another example image from the test-set."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Convolution Layer 1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now plot the filter-weights for the first convolutional layer.\n",
    "\n",
    "Note that positive weights are red and negative weights are blue."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Applying each of these convolutional filters to the first input image gives the following output images, which are then used as input to the second convolutional layer. Note that these images are down-sampled to about half the resolution of the original input image."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The following images are the results of applying the convolutional filters to the second image."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Convolution Layer 2"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now plot the filter-weights for the second convolutional layer.\n",
    "\n",
    "There are 16 output channels from the first conv-layer, which means there are 16 input channels to the second conv-layer. The second conv-layer has a set of filter-weights for each of its input channels. We start by plotting the filter-weigths for the first channel.\n",
    "\n",
    "Note again that positive weights are red and negative weights are blue."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There are 16 input channels to the second convolutional layer, so we can make another 15 plots of filter-weights like this. We just make one more with the filter-weights for the second channel. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "It can be difficult to understand and keep track of how these filters are applied because of the high dimensionality.\n",
    "\n",
    "Applying these convolutional filters to the images that were ouput from the first conv-layer gives the following images.\n",
    "\n",
    "Note that these are down-sampled yet again to half the resolution of the images from the first conv-layer."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "And these are the results of applying the filter-weights to the second image."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Write Test Predictions to CSV"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Close TensorFlow Session"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We are now done using TensorFlow, so we close the session to release its resources."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "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": "3.6.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}