Copy_of_C2W3_Assignment.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/davosky1/CLIP/blob/main/Copy_of_C2W3_Assignment.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "f8cj-HBNoEZy"
      },
      "source": [
        "# Week 3: Transfer Learning\n",
        "\n",
        "Welcome to this assignment! This week, you are going to use a technique called `Transfer Learning` in which you utilize an already trained network to help you solve a similar problem to the one it was originally trained to solve.\n",
        "\n",
        "Let's get started!"
      ],
      "id": "f8cj-HBNoEZy"
    },
    {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {
        "id": "lbFmQdsZs5eW",
        "tags": [
          "graded"
        ]
      },
      "outputs": [],
      "source": [
        "import os\n",
        "import zipfile\n",
        "import matplotlib.pyplot as plt\n",
        "import tensorflow as tf\n",
        "from tensorflow.keras import layers\n",
        "from tensorflow.keras import Model\n",
        "from tensorflow.keras.optimizers import RMSprop\n",
        "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
        "from tensorflow.keras.utils import img_to_array, load_img"
      ],
      "id": "lbFmQdsZs5eW"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "RPvtLK1GyUWr"
      },
      "source": [
        "## Dataset\n",
        "\n",
        "For this assignment, you will use the `Horse or Human dataset`, which contains images of horses and humans. \n",
        "\n",
        "Download the `training` and `validation` sets by running the cell below:"
      ],
      "id": "RPvtLK1GyUWr"
    },
    {
      "cell_type": "code",
      "execution_count": 5,
      "metadata": {
        "id": "dIeTNcPEo79J",
        "tags": []
      },
      "outputs": [],
      "source": [
        "# Get the Horse or Human training dataset\n",
        "!wget -q -P /content/ https://storage.googleapis.com/tensorflow-1-public/course2/week3/horse-or-human.zip\n",
        "\n",
        "# Get the Horse or Human validation dataset\n",
        "!wget -q -P /content/ https://storage.googleapis.com/tensorflow-1-public/course2/week3/validation-horse-or-human.zip\n",
        "\n",
        "test_local_zip = './horse-or-human.zip'\n",
        "zip_ref = zipfile.ZipFile(test_local_zip, 'r')\n",
        "zip_ref.extractall('/tmp/training')\n",
        "\n",
        "val_local_zip = './validation-horse-or-human.zip'\n",
        "zip_ref = zipfile.ZipFile(val_local_zip, 'r')\n",
        "zip_ref.extractall('/tmp/validation')\n",
        "\n",
        "zip_ref.close()"
      ],
      "id": "dIeTNcPEo79J"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "x4OMDxYS6tmv"
      },
      "source": [
        "This dataset already has an structure that is compatible with Keras' `flow_from_directory` so you don't need to move the images into subdirectories as you did in the previous assignments. However, it is still a good idea to save the paths of the images so you can use them later on:"
      ],
      "id": "x4OMDxYS6tmv"
    },
    {
      "cell_type": "code",
      "execution_count": 4,
      "metadata": {
        "id": "lHRrmo5CpEw_",
        "lines_to_next_cell": 2,
        "tags": [
          "graded"
        ],
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "b93e4925-90a2-4dca-8a75-04695750520a"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "There are 500 images of horses for training.\n",
            "\n",
            "There are 527 images of humans for training.\n",
            "\n",
            "There are 128 images of horses for validation.\n",
            "\n",
            "There are 128 images of humans for validation.\n",
            "\n"
          ]
        }
      ],
      "source": [
        "# Define the training and validation base directories\n",
        "train_dir = '/tmp/training'\n",
        "validation_dir = '/tmp/validation'\n",
        "\n",
        "# Directory with training horse pictures\n",
        "train_horses_dir = os.path.join(train_dir, 'horses')\n",
        "# Directory with training humans pictures\n",
        "train_humans_dir = os.path.join(train_dir, 'humans')\n",
        "# Directory with validation horse pictures\n",
        "validation_horses_dir = os.path.join(validation_dir, 'horses')\n",
        "# Directory with validation human pictures\n",
        "validation_humans_dir = os.path.join(validation_dir, 'humans')\n",
        "\n",
        "# Check the number of images for each class and set\n",
        "print(f\"There are {len(os.listdir(train_horses_dir))} images of horses for training.\\n\")\n",
        "print(f\"There are {len(os.listdir(train_humans_dir))} images of humans for training.\\n\")\n",
        "print(f\"There are {len(os.listdir(validation_horses_dir))} images of horses for validation.\\n\")\n",
        "print(f\"There are {len(os.listdir(validation_humans_dir))} images of humans for validation.\\n\")"
      ],
      "id": "lHRrmo5CpEw_"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "1G5hXBB57c78"
      },
      "source": [
        "Now take a look at a sample image of each one of the classes:"
      ],
      "id": "1G5hXBB57c78"
    },
    {
      "cell_type": "code",
      "execution_count": 6,
      "metadata": {
        "id": "HgbMs7p0qSKr",
        "tags": [
          "graded"
        ],
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 573
        },
        "outputId": "8a6137fd-83cb-4c47-92d0-922877aa253f"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Sample horse image:\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAQEAAAD8CAYAAAB3lxGOAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOy92e9tW3bf9RlzztXt5tec5p5zu+rLrrgt20W5TbBlJSAlpAEChAghXiIeeMd/AeJvyAMSbyAREC9BIFmQoACJ04JctuOq6+uq299zzq/be69mNoOHOdfa+9y6t2wpKXKQz7y657ebtVc35+i+4zvGElXl5Xg5Xo4/ucP8qz6Bl+PleDn+1Y6XSuDleDn+hI+XSuDleDn+hI+XSuDleDn+hI+XSuDleDn+hI+XSuDleDn+hI8fmhIQkX9TRH5PRL4tIr/xwzrOy/FyvBz/YkN+GDwBEbHAPwf+LPAO8FvAX1PVb/1LP9jL8XK8HP9C44flCXwT+LaqvqWqE/DfAH/ph3Ssl+PleDn+BYb7Ie33deB7J+/fAX7+szaua6erVQ0oqhCjEqOCACp5IwFjBGMEEQGE2YkRmfckiIACRiT/Pv+Dat5OxGCMRcTk7cv3ImAQFr9I5uOAqh4/n/f33DvQeU/zCZx+ybIROm+iLPun7P94GXK6R5gvJd+Qssn8+nj+nGzzWQ7eck2nF/EZQz/lVT6CMJ/xc/dI9fSWn8xPnqvT7z7lzH7wyXC8b89dx3IhJ7+XT+5NTu4cz72e18xnnYssUyBl/Xz6+X7yVs7n+Wm3+LNuu4icrKdPnv2njz/Kj1/ujAjvvf07T1T14Se3+WEpgT9yiMjfAP4GwGpV8x//1V9n3bWMfuQP33vKO+9foUQAFEvTOoxVjLO0rsW5lmFKTD6iaFEOFuMMxhiqxqJJSNGgKkyjYq2h7izNqqauWkQaMA4hK4S8UIXKOqrKIWKJKaIpr+CoirMOYyw+JIwxtFXFGBKqZeUJJFXCFAComhotQqkpizZiisAYKpdfx5QQEawxGDEoCURJKS88K4KIQUiozIKl5TcWYwzW5N+rQkoRY8xzCsdag7VZCYaoJFWMmKwwVUkpAZDI15AAIwZSQlNRLKIogg+JoIoKaMwKwBhD5xyC4mNkUZgCqoIzAuaovGeFbsSACKYoZRUlqpJOFv88N2GaiN6z3p4BhhgjgVTWgF2kVOb/NTu71llMUUZZqWerYMTynOgJWGMX4bK23Ntyn0RmQ5T3Mx9pigkQjMnHSzGB5DkXEcTk70zZt60MSQUf8hyKCFYEZ/P8Rc33QJW8Jkw5flFGKSkq2WgpQLmHqJTrOzGCktfdf/Gf/soffpos/rCUwLvAmyfv3yifLUNV/ybwNwE+/8YD/Yu//os8vnyMiOX9p9e89d33Gf2Ias9+2rEfBsbR8/jBQ+6f3+d6f+Dj6xtu9xEfDbUzVLbhdh/wMVE3Doxl9IFx8ExTwodE21bcv7+lbjY06wvqboWgTNPEME6EOAsWoOBjwojQ1jVt22CsIUa42w9YFIyhzlJDSolUrKH3E8+ePMV7T7fdUDcNzhliygJR5A1rBWOyFyJicDYfO4RITNkjskaobF7AIeXFplCOJVgrRWiUEBMpZaGpyuKbfMy/L9tVkhg14n3E2iKcSQlRCTEvqhSzEugai5UiSHb2lgSxluATo0+kOPstymHwAIRUlECRLydQW8Bkc6pazl8EZwxN5aicw4jiLDRGEGOIMSEqHKbA3d2e/X6Pa9YkMxFVCEkXhVlXCWvzuYpAiiznrt4DgitKYp4nIwqqRAwqBgTq2hIiJ4rZYq1BgaoIo2piDJEs+0LSo6eqqqQsmaB5DqyF2jmcMSQEmQDMMlf5Vs2qK+9TkKKMZ2VZPC8FLZ7urIisrcAYrJSLXxSbLB7oZ40flhL4LeCrIvJFsvD/B8B/+Fkbn623fOHV1zlbn7NZbfnyF7/EL37DFUug+DhxdzhwGAY26ws22/ukas2kkMSBrYkhoTEwDBO9j+z2I4fRsxsGrq6u2R96bg4Tagznm462XbG+uEfdtMQUub7Zsdv1RciymKUi2IrgrGXV1WCEKUS6fiIWbS9GkJQXzQy0xpjQ9g95//d/m7rZsn3wmNXZBWIdImVyiyubDaQUywRSrNW8TfYEOO63KIHsCWRBikUBZUuQLYkBkiZsiYWi5OMFAV8nYkpEEQJl//NKUSDm6x+t4Ey2ZlmMKAKc/9qoSL6Q5ZxUwaRjaJCKG5Qk+3YJSyohkDVCRFFrCdYVbybff2cM6sCmgPo7vA/41SVydp+DcSTNXsscawSTBcXMnhJgl3BQv0/QYgxoivk7YxFbIcAeQa2idg4D8j5nJWIkH9dbiNmJLAqF5RjpJNzLVh7Go1tGirP7fwxBy/QUQdesbARSse5SBDrvo3hmqjlMSX6Re/MpcegPwv9/KEpAVYOI/GfA/wxY4L9S1d/+rO2trTg/e5wXWlRIAYwixmKspXIdm/WGbtMQpcG7Devz+1xuq7KHY1wXFa7vEu+894zh+g4rI51sseeRJiR8yotvAqZBYJyy65UqUmNyfF4WqM7mWiGg3GUPn6QWqg6cHt2xsvi1uNGocvHa57h78iF3H7xN8BPnr32ezcNXqZoOJ4bE0TXNr/Nxioo/ngcnMaacfpaFSxXUFBdQya7u8recV/lvtgrJcgILl+1KpK+SF6qi+Xx0Pv7RbVayO4o7xuX63LlzvDcoAbICmGP68l0of0cFEkgSJE8DBqjjhNldMd48oVeDuX+fWHdEleVM5sg3zqthxow4PZYuWMsc+mENmBPpWGLyLNVyDKhP1pgsylLNLOTLFuis0PUYj4sY/GLgZ58pb2BONHw+niwKZQ7nUlmzOdSY51QXQ2FE0ZRyOHci7TOS9TwG8v3jh4YJqOrfBv72H29b2I+KGIerKoxxoI4YHYQKbEVQxxQdY4J08Ljbp2y3a+7dW9N0pnhA+eZvWuH1Vy7oGsf19R07J4whMkZl8Amf8oIMSYkoNimpMlSuYgbptLjXKWnGBDi58WUi0slkgOS4meMNd9sLXv/Rn+KtZ+9z8/b/TX/9EWevfpHtK2+wunyAa1uss+hsrcq9mBWD6Cc+U5YFlDR/T3kNR0tyXPYnyoDjYtIioScw6GIp83ez6ObrSmSTK/NrLfejCNN81YtySmle5kepJ1vPOZY+VZ7Agv8qilVwKVJPPen2GVcfv8thGOi+/JNUTQemmN9ZMGcltKCGJ+pqObyczNVxG51vshTvb7khR+GmYBroLKglxNCCz6gct1VIxas74hKzNzKDi7MSKLhLMuWzWQXNnl5ClGV+SfM1Fyzm9L6LohmAyL/j6ImkBbv49PGvDBg8HcYY1ptzktQEaoJaklQEtahtaLsVla0gQGMFVzuqKrupimbQz2Ut7APEAIrQNi1N4/Ehgk44A7VRQkrEpIQEQRNTTIt7Oy/OqGS4qSycLHCpWO3yedLnFlYWivScAHePH6E//U3e+ge/yfTse1wdbrj54HtsH73O2cPXWF3co+422MqhIqQTj29ZwHo8uSx6gmhRUMU1lBKrswhZWhb3ceFnRZLQsmjmGXheYGcgE0AlcSI35bP8z3Hx6XLOoAsQOLvpqs8v+nxJs4qZD6RYVUz0VH4k7a453F5x9+xDDv2e9nN/Cre9h0q+dlNMtSLFD6eAcMslnSD5WgQ1g45zUJDP7Zglz2Fa4jSXYMRkoTLzNc8KN2M4s86dAbvjKO77iUKaX8x3QU4yXc/d4FnBkkO/04yOlH9n5bdkLwSMuEUJHJW5YCQDwJ81XgglEBV2scW1W0y1xpoKZyokRhJCEIckyZZE81pNSVARQhSMKq4SnFNQYRwTh37iMHimGJkKgOVjJH5C2KGgr/NaLNbNSLGkxXrmG24WAdFTwSwmQBESJ0g6Ca0sr33hq0yHPW//1m/SX39AuHlKf/Mx+ycfsbn3kM35PbqLC5rNGa5pEecQa8tCmN2+dKoLSEgBlTTHjMUNz/foVDBPrzS7o8WOLQI5r710lNvj4tZ805d05olFmS1W0qMbKpgTVzWfy2xlZw8k7zf/xii4FLDBI8OecLjD9zv6u2uG26f0u2uqN/4Uq1e/iHH1MY1mzNGaz+drivLhOGe6RACndlaW0ECeEw6FWbARRPMaY9l/vkGzIM7fyawEZpUmc1w+S+kxxs/7isu9kuV8cvgjBZxUPck2HKfvExOkReh18Tayh5fBy0TMel3kZP6+f7wQSgBA/YExeKLswdQY1xCxqDiMqzBGqJzl8aMLrBN8AD+B94l+iHivtI1FBaw1tF1DjJFYOdzZhmkKjH4kFnTduAxO7fshI+auIibFh0hKgRh8BttESGpyOmYJBw1aFlW2hkCJB0/j/Dl261zNF37i5wjDnu/+k7+LHu44fPw9hrsr+ptH9Pce0T49o1lvaLo1dbeiWW+oug5X14i1S2ptFtyks0toMoBZhDq743mt5HBFjgJBPtlUpCOVq5kFR3Sxy5jnQoJiO4slnBXl7JJbTLZYyznp0VPQ591s0YRVsNFj/ICZDujQo8Ez7O843Fwx9Hv8/gp/uKF68AbnX/lJqu35EWibhbgoqeMczB7YEThdFMQJUDkrMjkV5HKtptyonI49cc0pvz/BTnJ6dd6dWZB79Jj2k6I40Dmro6CGrMZL2lfskqa1JZUYY0ZLZqVACTtSSovSYAGYFSNzCnSeU0XEIWVVqP7/QAl4P3CYDhwmISSDSkXCoVjyaQptW9E2lst7W9rG0NSgaonJAllRxwSuElYinHUrYuqykMSUc/7AbI6CKvuDx4ecG/Y+0g8DPkx47wk+5nAh5PRZVCXGmC2v8AnhK57KAgodj2SB6vKCL339mwx3V3z8+/8PMd7iDzfcTQf81LO+eAW363BVQ7faULcrqqalbluq9Ypms6HpVlSuKp5AsedJM5Kty/ojmVQUwNGIZIt9BAnnBVngjmVhHx3O4sYXQZotTloWGSeKYFYwBS9ZlMTRYxIUlxLWjzg/wHjAH3Z4P5FCxE8Td7fPuPr4PdK4x0nCV2vWn/8x3GpbrjYLyNGSgxgFDKdozHLechSIkjvj1E0vzvLxDs2Ci5CFer5uFk/nlDgmCxBAEfojLj/fRWNyOEHhmixzpAEnFhFX7mVcgMfscR73lwU6v7bWIWW/MQVi9EUBnHgwQGRe65lbcgLNfN94IZSAoFQSaSVQ1UpSg7ERcCQMIgZrHdZM9B99h3hdU3ctddNR1W22lAhqLMYaXFNljVAfAyZVBxxBOEqslS6O56GqpLR9TpsqyhigH5QQlHHy+JBJSvt9zzB5EjCMgcEHeu/xKc5LLf9rBBHH+f3HfPFnfhm/38N7b7Hb3TAMA0/f/Tb722ds7r9GszonpUTwHtnvsdZQNS1V27A6v+T84SPa7QZX1fm65vOcBbxgFTNnYRFS1cVLyUKqOXSZMY10kh0oMeW8YGfxWvApLSEAs9U9OYbqc4tRk6J+hGEPuxu0zx7fNPb4cSDFxDSNjP2B6yfvEfZX1AbcxQPOvvJzXL7xFbBVtnbG4iqHmDLfpOV6ZE4plHNNKeYQCgoiucw6Mwg6r4G8RLJimQk5nIQJC5BaFKEWiRJzxA5kIesc1e6c21dszkI8d+9NToHGRCw3OKc3zVHJJPDRk2KkqqpixDLJzFlHTBEfhuKR2DznKWGNJZVw1lCIZ7pAnt83XgglAMqqUta1XSxpntOSMBODsaG4T31Gqg+gB8s033zJEycmhw9iG5JYFEtSQ+/hMCm+YAnWVjRNgxihaixNU9G1DcYIMWQrOZM/KiNEm5AYMFVN7aA2EYuhcRMYw1BFDv1IV0dUhCkEpskvrK+oiqjl0eMvEX/2V/mdoWffH6iryDAOXH30Xfb7HZePPsc4HKjrFls1VHVLXQ+YO8Pu+oqbjz/i/P4rXD56zPrikrbrEDOj1UcFoGnOYpBdc1mWFiDEFPN2GtGYSCm7rFEjKQaSpqPVP/lfJHs8iD2ZvyN5aeY+qIKGQOx3TDdXhP0t02FH8FMGZf1E8CPjONLvbhn6A9PdUxqjJNvQvflVvvxTP8trb76BNZbb3Z7r2z3GZcZdZjweBdhax8yghHz/l2tMIMYgxuZUWpwZhmaxqkJx3Rc33qEpHZViSgj5mvPn87F0CU3yos13JKWYgyRjStYo+4zOGOqqQUxmKvb9jpR8URAGZ6vs+lsyjmAtMUyZ+2BtxiyKsMcYF4WmqjhXQbkuTbqAnBkzesGBQRGhqevyrrhRR48PjClumikTlamVIWVgMERDUCGmnPYLyeNjYEoQYsYNRh8ZgzKV0GB2ZQWDcZamqXn44IJHr1zS1i0hRoyxnJ9XdA1UFlKbaakhgI+OmBpCSMQIPioxJFxlqBrDMEZ2O08oE9YPA7d3PcMU2H7tJzD+wD8Z9tx8/C7OVbjgGe6e8dE00Ky2NO2Wut2y2l7QtCtcVWOsYdjvCmr+MdvLB9x//DrnDx5QFc/AGAPWHr2ZE6aZKfRTKQhyipEYPTHmBRNTIsRAjCHjKXEOK4piUV08G2stztqF+qsCISZizF5Mf3fHcHPNdHNNv79jGnvCNBFCQGMi+IGhP7Df3zL2e9I0YVNErePijS/x5o99ne12zaZ1fOlzbyJi+OjJU67ubrm53XEYxkLOgvWmo3YNh/7A5EPxUgKeSEiBFHNMnGIxMPMyKx6Ew2FN9h5d5Vi3LU3rePL0hsMw4WPMSqDQkjVFRNPRbV+o3kdMKMaAoFhTCBmaaVJN13L/Yo11NdPkgQCqjH7EWkdTZXZiPyRUE9ZV2K6ldtlA9uPIOEwogrU13ltS9LiqpqpakkZASRozu1OVyjUYe6q0nx8vjBIQW5+kO05Q1xlcEQMYpgC7g3JziNwNSu/LwtXMrJr5EkcM/CTvXyxVRv3ztwiQspLo+5GPPrpi1bW0jWPVtVhzj27tcNZSNQX8aY7nfppiOl5QxirGqWKcYPDgpw3D5cDV1S1j6DjvfhGbRv7+3/mfGHygE4sw0Pc7pmHP1O1ZrXum4UC3OaPpMvXYWksInhgCw27H7uop5/cfcfnKI84vLujalqpyGOMw1uKsoa5rxAiqgjEWa0xJec2gkRI1MXlP8NmCjt6z74esvG5v6Pc7BEPTtrSrjk3XsF6t6bqWrm1p6gZVuL255u23vsPT99+lv7tjHPZM41DOORK8J0wjh8Mdh90N03BAUsRooGpqPvdjP8U3/42/xBtf+RHOtmu2mxWvv7Jl24F+dcv1IXG3i4xjIKRICDELlTXse8+T6xve/fgZT67uuNvdMYwTSGZ5xlgyEmKw1lJVFZdnW862GwRLUxsuz9Z84fWHqET+6e9+j29957vc7UeiJFQ9kopnAPmaUirK1RFTOK69Ai4mBOcszlnONh2XZ+d8+bXHbDYtHz674X1JdFboJ4cRw5uvPuRsu+bu7sA4TShK09ScbzcYDLf7A+9++DF9nxWBt4aQigEQEHEkTdRNRYyeEKacilyCjO8fL4QSAMHYfCoLvDOTJyQv3mlSphDZ9ZHrXWTXJ/opZgagHl3VWbuLsQtCO8eDRgSjxcWwp9seudkpCfv9xKGfuL458PTpHW1Xs16v2G4axIB1eRFZa0p+VsqicvkzIzgHbS00NTQepsmy6VacbyrEgPfnfOnVv8CX3rzP//jf/3d89+0/RDTXEfTjwNDfIZpo2oCfeppVz+bsgqZd48QQfAAGUor4oefu+glnF/c4v7xkvV7TdGuapmHVNdimwghMweOnhAeGoWfsB0KYmKaxgHqCnybGaeTq6hnPnj7j6uqKD9//Hv1+T12taLsV7aplvdmyPT/n3oMHvPb6azx48IAUIu+/813eefsPuLu6xvuR4EdSTKQUiTEQ/MR+d8PdzVP81EOKtM4gxvDmj/w4f+U/+k/4iZ/9ac7OGraN0BmoyeAqVnhwbuDcMWvixHERK47xzY6b/hFPbwbev7njrp+wVnj/g2v2+571esWqa3GVo+tqPv/KBWermhDhYiWsa2FlMpPxfP1FNqsNb3/vA27ubpm8R5OWkCnigxxTcBhULKpgyJhFUnBW2GxWdN2K++dbHt675Ctv3uPVi5rdsOXu8JApBD6+HUkRvvjojO2qyV5ZSARNVMbQto5+Ej5+NrBuK3b7gc2q43Z/4OpuzxQ8w5BrX2ISjDWYHD1kWUgvenZAE2kacgwqFNqqye6/mMzrBipRzjphVVt8MPSTpR8TwxTwMeZQIJbqM82eg6ocizlOPs8Ko2jsVKoIRUmSCsiUSTUpToxh4nZ34KMPM20zKw3znJKx1pUKw1wUVFUm4wmVo66qoiiEtmvZrCtcteKVy45H9/8c/e6W/+Fv/S2ur64XxTJMI9O4J6WAqzpSnEhxYntxn86cZXRYcuwZgmfyA/3ulqcfvktVN1jrqOqGtm1YbTZYa/E+pz5DCPSHA37yGWEOOQQwYuj7Pd97+w94+uRpLqKJkRBGAIzUpQoulfDDYCtLt+nomgpJAWuE1WpD07Y5bg0h38sUmcaR3e0z9jdPmYYdohFnDNa03Hv8Bn/+3//r/PKf/llWq4pKoJOsAJYsHJ/dAGOGQFcCq5XwaNXx1ccdIeZ5vP3yK8QYaZ2jKRWFRnJRk5FjFuf0OF89q3n8M6/x/o/c591nE08+vmUcPSEFlMQ4TVhjudntmEbP2dmGkJSr61uGfqKuau7du+CV+5dsN2veeNzx2rbifmupjaAbhxYRnAKEBF2V7VPOitmTTA08CVDZhvP167SN8mDt2PvIR9eBu0Pkt7/9Du9+9AQfIsY1maNRVTmc0RfcE/B+4t13vk1SaJpmQZutc9lDkAzcOGsRY7Bki941Fm2EWKrYxlCYgCESQigAlhQASUqVmM2lt9aV3GlWBooprEGzIMcpzay4kuERA1o46imDLUoG/QLCkFhotQvTUKTEhRlkzMrAUNc5h6sp8NprX+bnvvEr/MPf+nvcXD0jljLRaRwYh54YIytR/D5yFwPRj3SbC6q6xTpHMoZpPFbPAUs4YF2FczUmI0sFPMzUOucsMUT2hx39YQ8k/DTywXvvcjj0Cw6g5Gq6lG6ZxdDYjNJbY9jdWpzNIZP3PXXTcP/+I7q2JqVACpFx7BkPO6bDLZomSB5Qqqrl8pVX+fW/+Ff513/tl7i3LsBYmZnEUfA/25aV8mdgriYxQCPQuLw2Nq46+faP3l82OnDZCJd1y1cvG8KXtgsXw5Cp5wLsY66NOLMZUP5wp3zne1c0dc2bjzasGuFenX2XI9fh+VG554/9yaECjxzcuxTksqKeN1TDl88qrlUZ+0f0+z2HcWRmrQoVPjmixk/Zax4vhBJQ8g1FIGYoN6ObYcItyTbQmMMGa11mZBVChS2xrquy+gjGE8y4IPMo2KqirhpcZXBWse4ESWfOLgiGuXbbEhVCjCekm/J/eZ1KKWjUrGiSSnldQMuUacwhReJS3SfEJPjg0TIx7eqMn/vmL5GAf/QP/h56fZXR+ZgIURnHAYDt2hDHO+6eTQxDT1132IIYA1D6CmSQKOMHc1iEWFzVLAChqxw+xYzOH/aMfU8IE3Vdc35xwfX1E0LwmW4617NbW34/L5sAWERsQdUtbXtOVOX6ds/gEykcmHbXpDgiGrEaQAOkhLEVVdXy5R/7GX7253+FtqvwesxmqcklyELhWnAUkCMrIJ/FvhR3bd33L+rPEvZPs42f3LbYA2pkyTh/cmyNPPeLszP43I8/QIH2Bxz/jzpH/cT3VkpY9IkfWrLH9OqDFSm+wf6wZ7+7xk8DU4yMPocInzVeCCXgrOX+5QOSxiWuT5o9BBTqpqVy+fJFTK7vLjlaRReGXCSvIMFRVzbXCOROHjnNaEDI6TAJmcY5l8nOFrtkGjE252NTsszcupz9SvMLmJWHMUUQ7KIMfILew92g7CbBx9nTKNc3c/LLZxeXF3zzF34JEfjH//DvM334HiYEnFN8TPTjiHOWru3QFDlcD/RiWSj6YhDrqNoV3WpLu9pQuRoxpigxQ117XOWomgbFMI4jTz/+kN3dLTFM+KmnW59R1xV1XaGa8uVZh3VtYccZrGuoq5oUJ4ZhX+L+QHI5pLBVTaJm8BFr2nxPQqC2CWsSwzQhxlG3HW/+yI/xk9/402Aa3vvQUzlLSkrlhPOtoa6yrl+3wlnx1/Ncw0EzYasflOvbMX95r+Gifb6C8JPLf/4uASPF09BcWerIiqf6lN/9cYeQhf/Tjil8uvL55Mj5jXydjizkP2isEb7+esNPvvoKY0o87V/nyZMdu/5AP070/eEzf/tCKAFrLJcXF7m+PZUUocn50BDjwsYyotm6lQ47M4VTkNySLMUMQBXiiLFHauysrOd0jlDwQWOwLocIYlzR/KV7zClvW3PBzoymzyhwLFY/qSXhmKJhijAE8FHwyeSUmkipbZ+BSj2uMs1pxKp6yC//mV9je37O3/lf/xfee+e7pJRoafApME0TTVVR14L6SNKwFBGJWFLIltq7irppSCbHvnVV4aoa27S0TcfkAx9//ITbm6ccdjf4cUIEnLPs726ISXH1hhAi03iH+JGJHSlOGfgyhqpqcHWLsTXWVSQNxDgSRfFhoIqBdnVOtz5Hwj2mOGBkIoSEqmW92fLjX/8Gf/7f/ut8/is/ioql3w0MYqiqimBA1FE5cyzpbYvASs649AcYp8Tt7Q1XV0+p65amfkhVt+ynnB86byR7E0XIUUq/gqOQZRIv7Kb8XeeOqf9TnOBf5piPfVrRrSefjcBUlN1ajpjHvB2fOC8j0FhQK3Rquagtr5/X3IR77EalP/jPPJcXQgkgknsHzOymQtQR47AuU4bn2nnIKZpctDJLdtnHDPahWAvOupKmWdjkzKywmS47W5akhS8vsHD9xBxRI1WsHCsGE8IUhdsR7kaLT4akhpDMSd38kU33vA2Yqalz3t1ixOLEcnFxj5/9xi/S1C3/+9/9Tb7z7d8jDoKLhpQC3nsqZ7DGQMoKcg6lBLJFziAJIkLddKw3Z1jnOIwTH3zwAc5auqZmL4JzNdF7xmFk8jbns8VQN+sMOvoBY9zXC8AAACAASURBVBxiKqbhBj/tQBN+OmCGkoo0Ble3bM4uqZsN/eHA2O8w1lFdvoI2G6ZDRwwTMUHTrvi5X/gV/sK/89f4qZ/+Meq2QVXZHxLBZ3p234/0KeKrjDWYvSEEKcBrnjNjoaoN3WpFiAFNubBs72F/KPfEQF16Htwect3Rqs0FYj7m91GgH+HZjceK4eLc4pqjyLV8skKQ5825PPfn05f4yc8C0BcBbyi9Ayjl7WUFe47Ywx/Xe1iOUw7WIlQVbJ0wds1n/ubFUAKQPQDNRRmZuFIicJmtd6kc1NkaUyrrcjhwDA7AWYNzNrersu6k7j97Dzm+KgVBp8U2KbcSM2Yu5TTFXc9iu+CIkANWAZ8Se5+VwOkyyC0HZ+UDzy+DQrCBhYOfFm6DsF5t+Omf+XkevPKY//P/+Dv8o3/497m6ekoIQlQYfaKp5zRU1mKpkFdi8AyHHVXTsdpcsFptcK7m0O+5u73hsN/zr33zF/iJn/opnn38IVdXV3zrt/8pb731Noow9Hs0BqytWK0vGPtbht1TXL2has9Zb17Bh54UAxoDMYxEHwjTDeP+mnZ1RlWvwNb0+2s+ev8ttpt7NKv7jLvI5YOKX/kzv8pf/nf/Gl/52hdZrwRncz3duBHGCQ6HRO1qkuYQygiEkGnbqQg/ZAFOKojUNPU2k3XUsTskvM93+iqCdfl3+31mD1ZVzpvH0jpODIzjxOFwyKy7tCFuc4uxvo+cbQxtnTEBBYLClPL5tC63QqtLm7aTjPSnKoUIDBQrn7JCm5eVkAVy7uBkTz6fmc8/KDtyXF3H4zsylrD6bK7QC6IEZAbc4pEBzGyAFY2B2WRn+vVRL6aiPOYGIKiixmC1tGVKpTdAEjSZwifPJiJpVj4zzwAxBTsoqcniPZRUMEuH4px8ZQpKH4Qkspy3ppmzXT4oCmsuWll6AABzpZ0wYwXHrIJzjjfe+Dx/9s/9Wzx+9Ab/7J/9Fm+99W2mcWAuM62sIcSMbcQkBE0Ya6nbjqqAgIpwd3vD7e01h2FAFZwz3Lv3gK7tSGqo2w0PXnkT7ycO7ikhTVT1hqBCtVszDnf46QBYmu1D1tV9xuGWOI2kFECEGCaG/pp+f8twuMW4mqrbEsOIxsD2/CGby0f86q/9PH/xL/8V3njjDZwRks/ZHSxULvfis9bSNjZXiobcZzGkhJ8yOUiLe2+czXRhNPdE1MTd7UBKig953mKMxBCZwtxT0eLqCmeVyuYGomKUlEIhfll2+5F+yDn/cZy4uclp26auc8+HmCm7obQnq5xlta452zgqMTgjNLXg7PNrWWRuZpMVWOWyJ/Bp+EUgKxvIJd6zQHOy7elQcrgTNCuK9o9QRqfjxVACqsQwMbemjjFlMo7JdQFzo47TIpm0SC5kc50bPaKZeBJVmLzmbkLB4qMQYqYVJ9VCMZ6VwMIrBBJiSk1gCQ2OZaFasIki8MUSp3TEJpYqM9XC0tOl8yzo0mtg8Wggl4jO74pnA1mxrdo1X//6N3j9jc/x7X/+O3z793+X9977Lne7GwyZjBJTQo1gVWi6NZcPXsXUjvN7F4xT4vr6GcOwJ0kLOnF784xhPDCMI3d3tzx+/fN89Wv3efb0I95/9w9YbzoSNd/5zh9Q1StW2wdM/Y6mPSdpLAVeFclFatdR1R1Df4dqwE8DoPjxQIqeqlnTG4d1DQ8fPqJqVkSfuLnesd/b4opF6rqmW3elOi7PUQjH7I6EiEkRSZl3IMYgMeXmMs5Ri2OcRvw0EYaJaQpEFXxM+JktKBbTNKSYGMaR3mfWXV07mrbCucznnz3O3N6uJoVEirliL0XFT76wWJUUPZNMHA4DV08txmQmYtsW7CZ4NAaMNTRth7GSm6ciOGfoC/5gbFYKxmSq+zDmv10L0sjiDSglQ/AJyY6aPYhjqXnBGHIE/AMF/YVQAiklhv5AU9fEmPjg6TM2my2XZ2dQxMuIYGwWkKRaIv+c7zYnZZ8ziScmiEEYg6H3whiEMRRSRrEsSU9abpXo3ZQml5lqKQsjkNJ3DxSRdOLmH8cc52tRSvM85c62uZIrU5bL67lRyCL0LEVoMyiZFJwTHj96xL2Le3ztR36CDz54l7fe+ue8//53ub5+wtDvMUnpunO6zT1UlVW3ZpoSu7s9Y38L4qiqCvUH7t+/jzGGZ88+5qOPn9Kuzrj/8BXONhtqq7RtRcLyznffzi2wmzXOtVw8eIMUPf3uhqruQGwmGxWSUFW3GdXRzBAkTsRhhzeWu5SoK8O3vvV7vPb4Vb70xS/m9KMmRDTjB02Lqyo0wTRNhJitfgq5qMlag60cxlY5zAshd2ouqG8ImZMQUu4CnHkcmXRDTCieJInkB6Z+R/IDiCXUNSatic4y7g8olrptswcxDMw8idkh3+9umYYx07PFM+1vl6yLsY5uvWWo61KbEXHW4qoKV/UZlJRcZ2KtpW7qzOew5sTozUVgEH1DDPYYZgjUFbnSUorwJxgjhDCD3eCNcnUb6PvEdmtZr1/w2oGkqRRqZMHcD566TYw+J3/FzP34Z6s8RwQ5oDLGLgJIEU9rlLYCZwObWnOhUcrKYAjK4JUp5MKfUAqNQsrtsyGTkZZ0pcw5huxe5wkxnwCLjh1i8rsyFBYdPkv8rABmL6SQcpZ6fz35qc7hSA5IN5s1X/j8l3j0yiOur57wB3/w+3zrd/4Jo4+szx5Sr85R4whRwCfC2KPJszq7h59G+vFA3eQeC2PfZ9d5GvnO732LVWXQMHJ+9hBbObq6Lg0vc9XlOO5wzmGrClc1pBSzgDA3v7DU3RlhuCEZIVLR1I4QRpJ13N484d13V/ze7/4Ol9s11hhC8FR1jWtazL7H2EzI8qkIdVLEukw9RhdehLF2Rm2WqsfcUSkuZCgpTNOFRm4gxh7vJ6If0RQxrsY0LreYj8owDsQY6fc3TMPEOPQYC6v1GiWTr2IY8dNA8IIQ2d/sM/kpjRhRuu2WplsBQt2ujwrLZW8Dya3QrcscC1c5mqZDU85oGZsVdq7khBQbrMsM1JRgtOCqLPApZeH3Y8L7XDwkhcY+TpHJB3Z7IaQXXAmolphv8PTjwNXNLevNmin4vADVoKm0ZBLJ3WMWVx1E4vM0XlhCBWcovQWzFd40nJB2tJB9DGMQbkfD6HNcFVP2GHxM2as4CT+W2vO5BHVhgcnSy23pdMOJQijeCtj8UAwzC31WAmZRBicKomiFrD9SCXsSq66hqR6zWq3p+z0fPHtG3W4QW+HVYU1NjBNhuKFu1lSuZdjdYCQL1GHfM42R9WqLCtxcfczoImebjvPzM4y1NE2Ds5ZxOCDGMR3usOszSB7VEie3mbC0u/6Iul6XTjoZfc+doit2+x2WABrZHQ6888Eznj674sGDB4jNzWNSyhTZ3d1dYUmWVmnW0m02GGOZgkLI+7WpGATNNQlHWLYwOSNoSDl0Y8ohR1WhzhKngRACTbem7dYocHt1nZ/zoMI0eUSEqumQdotPibspP8fCJoOtz9isLjMz1U901RkpZsWCBoak9HcjyR9YrcbygBu4uPegKCbl7N59uq6j70dub65zdmbY46e85ptuDVhsXdGtVrSrDmsrUkoYslF0dYVxOd09DZ6+75n6AVWlbku/DVcTpsA+vOj9BBRCGBmHgfc+esKHT3dcnK+4f7airdscD2kGf2bgLY887alYgNlTKPg7c97s+MiuPDJvXLFmdtsTqxrO2yP7z0fhblA+2iX2Q2nnXNz82S87tpwuCsjkuoRcUHK6qRwVAfNpzTyEnAgSVaQoBcNR+OHYUFQLUSnFXCdf1Q5xwvm9BzzbZ7quao1GYZoOjP0dXdOgpiGGgeAH2m6L95HrqxvabsPDasWTpx9ROcfDB+esVyvWmw23tzcIQtes6Q853TdNuYGFcw131x9jXM22fcD52X1Cv8NPPXV7kZ/q4+DexYb7D17le+/8ITe3z7g4W+PjyAcfPeH9D56w2V7QdisoT3RyLjdQQUxmasb8ZCaCR2xiutsTYsK6FtetiteQSmlvUZYlo0SpO3HOYG1Og/ZDoKqg6bY0xmJMhbEVkNgddtzd3uB9JKaErSo2F2e4umMaAlEzCaqqc/2ET0KMEQMYaxCpwBa+ieQcZvQHpmlEQ07r9lPCxIQxgn9yw2oVqbsWV69IYcRWXWGIjhz6p2gyGFfRbTbUVZWBxZhyJajNYQSSA54QI30/ELwHFaq2oVltAcX7sXBcPn28EEpAVRnHA0+ePOMP3/mQ+w8fsO6apXmCnfsJWFm6xh4xgFMEnuNCeE7sNIM4BZBb+vPr/IyAIybgipWvbJ5MrwYroWAJqfQHyNz+TDsuoi45zRVLTzgjM05hFkUAc765bC/H/n/5UzkBI3U+MYwpqIXmPgpqbakbzz81LluIcYqkNJFiwhJpmgpbtUwe0jiyWp8hCPvdjocP4d7FOR9++D5jf4fEicads92eUVWOu7sdYh3NagNXHxaClhCmgdX6nOFwR4geP/YghtX6grsw4ccDxlQ063NWmy0//uM/zb37D/jH//j/Yt21iGnww44nz675UoiLp6SqBO8zfiK5v4N1FYpZOoyv1DGOE2Id1pgjUUpLGChyfFYEOfRKSUEsahyaJHsTU6Rpm9xT8tDjrODqhs35Rak7iYSQkKRImBjv7pjGgFRN7gxd12BMNgxxbuCSsQ3rqgVXsqZCjUE1pxtvb0dSOiAa0BSonKPuusKAzRTvpI4oiq0qcr8Ch9iOKQkx5GrPqp4xkNyWTTWQEhwOh8yyTVCPHtuH3FxFw9FD/pTxYiiBIsRN2/D1H/8aDx7cZ9Pl4pg5FrfGLn3m8m/mCL18tvjdue5Alu6y5QhzCm4O282xLGXOPEDpG5gy6Ng65ZWNcNkJMTqmqPQ+MgVlDFKwhJxvnh8ZllKuZ4hziLAoqvkZdmZRUHrSyWa+qAWAOlUGIjPicJIlWWBH6rYlqCUMU4EfAt1mg7OZ0ZfEY+sa2RsqZ4nBs9msiBGCDzROsAHSNGU8wHXs930JUxIpThjjcFbw0x26WrPdnPP02fv4cWDq93Tdhjgd6Iee4Eec7Rh9Yrvd8sUv/BJhPPC9d7/L41fu07RrjMA4DpyfbfNU2FLJmRSszfUJJ4rTuSo/E7BgArlsu0KskKJnLOnP2SMLIVdMZigv40oYx9JMJfgCwCljSAvBqq5b1psGFZuF2VqS6Tgc+sxXsbnjUPABUsxPvirFZ9YZQkgZGzCJVLoZietyRyU/laKs3AkJNegUMJKWepYQAtGHkiqNwMTQJzAZNBdjUJ/D08oaXNOBEUJItKalipHZDY0xkkyVn98Rw2fK3wuhBEQMq/UZ2805Vd3gSjcWTUokLVV489BiubPFnfdxXDAqwkw0OnXfj/3vDKcqgtKjP9Nvc495Le6/dYbWZcWQCrCfNKcbfYKQLD5ZBp/YD0rvFR/S8kiyU17A3LNg9gKkeDhGjg+fmD2FJS3J0UPQfHEs7bMFxLjCVBSGYUTigfOLB7Tdpix6qNViq4a2aVmtNpyd3aOyjv3uNlODfWC72dCtVlR1jWDYnj3ID21FEDUEPyHGEvxACCPW5Qx3ihOC4qqa1XpL3TTsd3eEEBk83N7e8LWvfplf/IVfRv6B8MrDV9huzzi7eEBKkbp2VHXNQsxCSgcfk+/hyb2orEO6PP+2PM5t7mPclLZj1mXhHYfA5HMBVFU5Zq9KscUxjBlvEsnCaHNhVG71TsnmCJWxrNoqY1JiqKuKqqoRZ1GUu9vbJf1YVVV2y/cHpmFAjKWqmsLi9Hk7BNs0BDFEHzFBcJVQozhKR6cUEVtTOYMPBw43T3OPAs0P1Z3EkdRijWJcjYohxEBKBuvqxTtSyJ5p4eJ+1nghlIAxhq5dFdxMCDG7YvNTdlMpwZ1VXAbg5vQdR4tBkflTL6C8yExAoTybZWEannoBM5BnBIw9EnmA3IZAYWm7nRP8QNbsPgqDN/QBbofE9SEyTIEUjy2nZl4AKT+lZg5tMrr5CWBTSuszOcUcTmDGogUOo+ejp9fc3OQ2XZuupltvqZqO+Rl2TbNBjKGqG+5dPmRzfoYYg/cTw3CgP+y42FxijBCCJ4jn7OwSRPDTgHP5uX/iaqZphw8Trl6Vfowjh36HGJut6KqjaTuur6+ZfOTDj59gq5ovf+Wr+BAIIXB+fs5me4aSOR2uqk6ah87JV4Mr7buWdAlQaVOUXw6zUow5NNKmpNUSKXmskQxsVpa6qpnGIbsEpQuQta603DLHENJYVAvvorQCT+QCqq5rUZWlpVqMgZQCtc0PjLEWulYYx8AUdzh/oN2cIc7j04RUDtomz6vNHbRjRqVJKTEV4wGan6btUyE8WVJqigEKmClgXFb6QsK6zC+MKVffkgRNIcuAEWpXg1HSNHym/L0QSgBVNMaS1hGsmSm15cLIoFgm48ztnGZknsW6LrubHee5ghCYn9k3P8BThNKRdcYJ5iaZxyKVUytcdrz8yeuytG1SxRmhq5QzFS5Xhodrw2407IfEYUwMIeUnBs8dcE9wi4xn5J4Gs2KavYG0hAXHUt75yCHBh0+u+eijD7m9eUYII5eXb9KsNkQM3gesCKtNlwGmbsXZ+WV+XLpqFvgwcbbpcA6UyNAfuLe+ZKgyyeXu7ooYY1Yi1nFIiWF/y6Y7Z706Zxp7pv4OZyts22DbFQ/OLnC24unTj3jnvfe5ubnmC5//PD/6o1/jgw8+4OLiHDGGum4zoJvvKMvjuTCFsFFChPmuL7qgZGaMEE1+uCeqqDVFOXSlz6DMzwvBuCavoaJwrTFIeUz77AVmHW+oSmn06fMVas1NQ00pzY6lkWnbrZemn5Mf8T7SNC1t7bi4d8ndzTXTeJsL4WxHVa+oKoPYEj6UBapJCVNfegNq9ty6M9SuiLl3HsPQZyJTwQ9sVSHW5JZnCXwY8H2PxtyVGFVEppwa9T+kKkIReRu4o7AhVfUbInIP+G+BLwBvA/+eql79oP0oOUc/F/bMVl5hucEzuqmqmVQRpLhzuU7AFA1Nmbg8ebP5ZvmrqaDwHPkGYrMSmX8nzI8gO3oe+YLLnpTyDLuSjThRDpnFB10lXK4dPpB5CQH2U2Kc0velHlOaw4bSH35mRMrxfEQMEgtluYzeG55d33Jz/YwQAm3b0a0vwLgc96ZIVbe4uma13nB+dsFqu1kUYQqBzbrDaIWtAlXtqJuGpumI4QpNAQ2eumkXb0RjRONITKHgBBWTHzjsr7H2Ho0PrG3NvQePGf3EoR959733eOP119hs1jx4eB+UTJ5xrmQ64vJ0nsw3yB5WceuWez7PYX7KUUKjLmFbWY+LvjYms/eMzWFlfjDH8shSKD7hfIMX1VqMixYlNLM752yOMbaEcSCxrEdrqJsGkTUxXaAxW2EjBuyGanM/d15WyaBplVH9FHLthSYPmuinieBzOrWSivPGgW2ZfMYKXB3othcY4xj7HpUAktuPJxX2veEQpfSuKD6VyRgLdvWZ8vcvwxP4NVV9cvL+N4DfVNX/UkR+o7z/z3/QDpIq/TjipxEh0dY1VVVljRsDMdoFF5gfxGBMFn4sWROiufZ/kVdZwosjeFhi7rkyEMpCK97BvOUScM9uPM8vRtXFcpyGWgtWsXwntDWsy3fzI9BSoSzPTUfGoExB8VGYkmFKFq+21DbwnIJKKaKaO/pOceL27pbdfkCxtM2K7Xa7eFYCNF1H3TVstlvuP3ycQTUnDGPuWNRWDU29QmSgW53hbI5rh6EnxAkfRqp6VYg9IyKlD+FwoG5Wue3ZFHOvwvHAwQhNvWNzfsn9B4+5uXrKt7/zNl/90hd49MpD1l3DoSDyaMxVfGEqj+HWPJlIbiknR0V86gUsDxF9bgLmmP9UUYAW4r0pgrI4dTNWo6V1+OJMFq+Ek++zC1kwnoimcOy6pKdrLBso4zI+oJjc2j6t8zmVEMaIyaQgyY8+UU2l5dsDfH+LpjH3aug9ST1N26BMxHCgkwpBWJ+11OsNh90d0Q8cDgdaEVyn9P2hENCywTGmpan+v60i/EvAr5bX/zXwv/FHKAFNET/1DOOAqFI5SyW5rYNqzo3OXJ1UwBUxQpUc0VpCyO+z9pfyMIa5CWiJq41ZsILlgRXATPrJ4wQsXFJ8lJi9fPbcmcvJXz0uyuWBkPPSOFEaZK5DilnwDxioVlRuQ7QrkuuItiVK7qk0A5JLt6WUC2Kunz3j8MHbjMPA/8vcu/zYtmVnXr8xH2utvXc8zuOe+3BeO1/Om7ZsqkpVEggJIQQCiVaJTnVpIPlfoNq0qo2QEG4gqAYNOgW0ECUkhBAyUBQlJ1WVL6ftm3kf597ziBOPvdda80VjjLkiru2bWdggnSXFiXPiROy9Y685xxzj+77xjVwSQwxcPH5GiBOtNYIPDMNICM7Q8C6ogv204/b2miDCaTkRJBKHiGdAEGrJzCedEdDQU7QzNcO4J6eFUiq76QxphVyS3qesngc5JVopjMPE+cUjnr+85oc//ilnZxPjOAJF6S4cPgRKrlsW10uArvLrKeFWHnKPmzzEbHrWsJnTUqA4ytaTZw/VevToAaDRnZkUd7HH6dJu+rASsaBkdftWflrwKTbsxJiM1gIuCGEa8FWdtO9HiwvOd72K4IMeQPvzA8v8mLSuCkyaP0ZrFTndklfH1YvXlJJUx3B2zjANDNNInhUrCENkHQeKNVrVWkhpxYevrtyH1181CDTgfxAltf+z1trvA++11j6z//8ceO9XPkhrDDEyDDtaK8Q4kVtAWtUTAz019RRXlFRAJ7I0pWIARPK9pTao669Xyy2pHRJUNFHTOyxN7LVnP0X0V+vW0vcZgiLEbGCeffdXPrt7GtAWz32M0Gaiec588frEy+VAePwb7B+fM+5Hpui2OrRUzRiMubKfVzVcWgOvXmjr3ZvbO8RFnj55zKMn7xCnPT0Q9RZqmtKXNze3PHn0SCca50yMgTZNeO8Zxx0hTuz2B3Kaubs7EcKO4JVPH1zUFl5zZ/bOmaHIyDiecTrd2f3QRptlPuKDegEUGv/0J3/C5cWO73zr1wFY5hMhBFTmq2Iv2e6ju9+YzhleAo1qYKq73/2t739L3+kgq2FJFnQ7DtTHzN/jjfduVnq4eB380e9ZE+2DeED1PgzoXwlKTXGlmjPZ9AwazJy5/yqjoetRIHi6ZYVzDvFqSDuOAxuXbTGq1Xcpz1ZqXsjm2kzJSCssp5V1ddASsiRqmsnLSqk6ByMEzxS+3pvorxoE/rXW2ici8i7wD0Xkhw//s7XWLED8uUtEfg/4PYBfe/cdxmFPrk1R0lK2G1KrpxTtj3I2UTfESAxRnW9bpWKThc0qrNfVW1puCKxDnYCDtaDKtuEejt/up/79yS6G0Os919OoL67WoFuWbyWFE7Ovcdv3pSXzxcs7fvLpLT9/sXCVJ7790bf47fff4fzCEYKGGlNHU7KWCj2OtNbHf8HpdqGtRw2EufDNb3/E5aNH7A8HnPfq7Z9Wexz1KnRAmlU5djwtUBpDCAzhTF+uH3B+olZ4+fJL9rs94zRQbQt67wkhsMx3VJJu7nxUaXIp5HRiGEYExzLfqh9eHJgOB8ZpT6qFH/z4Y1prfPMb7xGDI5iXf0OolqnY4uBh6cefPfk3Zyit05upNJ0BeuCgG9TeR9CvlHZbptazg6aofbNSqg/rqDWbZF3sANJg1Lbg8zAI2LovQnEN2UqGtrkjbViU9GzHMlfk3sLeB+tSvMc2RMDHiTju1degKTOSUyaUQjzPNClIy6RFKcq0LKT1jlITy/L1m/ivFARaa5/Y5y9E5B8A/zLwXEQ+aK19JiIfAF98zc/+PvD7AL/70XfaadUWqFwKTiox2By2FgGHdwFpyrcmU5XFGBCJ92O3rNvMyX0TRTWzD+cchUYrtsAouObuV5bdVXFuS/0xcCV4A4xMuvZwNt09a9fTTOO7awEb6PHFqyM/+vgNP/j4ml+8WpAw8L3vf5P3fu19pkkXVslsSHbHDcCcb4r1oAuktXH75g1D8NSycnFxyfnlY1wcgLaBqK0JNJ11ty4rd3e37Eykc3d3h4ueyISQoSR17HG26FvD+8rtzSuEsp180QdCHFnzifPDgTiaCKYk5anzTA5qB19xaiZq1GGtmdNS+ed/9Cmjh7/5Wx/y6GKngc2CsY7s0o1SWh9txlZ7byc+xhh85fS3zd9LCuftey2L2E7v/h4bINvresSmX+tjaoNSly9XKzF4gDvcb/xmoG73kHjoSFUN0+kH0r0cHMUaqlCkjxVT9uu+tNGRYhjQp6yVmKels/LEysvWEBNc4Q4Mh3OGMw3gtWrfxtddf+kgICIHwLXWbuzv/w7wHwH/HfDvA3/PPv+3v/rBNFWqJeOtjm/N01pWtxavwzO8eGrWGteBcehi9uFVT+QmlKZzAkEdg0/LYlmE1sohDERLEZv18W49B2LshPPUVnENatAm7x7pq9XmTmAInhA6z41F+8YyJ15eHfnxz9/wf/3xG764LSwF1pw5DBO1ZI53M7VO+O59b2YTywLLqt1hD0uCRON4M1PWGQHubm8IwZOb3M/YKxnVzA+qYsyZ4/FIa4733/vQat3GuiSoiRBA7D3RAOfY7Q68fP2KdTnZfDvFW3a7M9XClxVxkf3+nHk+asnl1dhV0sp+UpGRQwg+shsHUtHJSVc3J/73P/wpZ3HlX/9bH/Ho/LCd9LKdtJqGb9W7bVQVe/UswSg8CwLNcvjWSwn7EPH3UF/DHKj152v/NwLiEW92aSGolyViStCybbwePED3sqbmeVsboHoDHzqd6yhFsayci7VZG/hpAaMZDtMZop7+NdEMQjEx65No2Bi9glrW6xPnWm2ceQ+UZhDbBWl/3qd4u/4qmcB7wD+wVCgA/1Vr7b8Xkf8D+K9F5D8A/hT4O7/ykVrFt8wYHX7a01C1pxpYEQAAIABJREFU2BA1LZrXmTUvDHFgiAMuRhV7aFDcBkSG4PHTtIFUIajYg1V73vXmjPg4qNSyFlsQcH+KZkIIONfYphIVobZCSjrXrvsCiKDSzVDUfNNoyjU3fvbpG/7gR6/4oy8XEp4QRqRlnGSW5URrjYvLgd3u3nO+ilYh3inNSFP/OxFjEU6N45s7HJV5nnn++WfMy8yyJPqkWkQYdyM+DDqUpFZOxzsO+wM+aPlUSoaaEUEn41rdWkrl7nirXXSt8ujRY16+uiItM0HUj3B/uOB0uibnWZtefNBUtW/P1sHXhVw8eV3xPhKHiRcvnnN3d80Xxxv+m6svCWT+zX/1r7HfTfoYBpbZorjfWB1P6azMw2DQ/7+XED0rsA14P4HqnvdvZltWzcyl4mgSqBLBFRAdPioIQRq175KvUNB2sgdPG/Qg6o5U4j1xMJNUZ2xQUrobUWeiapkkqNAnrWoEWrdBsRbH7H0oVdRiLZmZbq3W92IZU22s67q1pdPMbcvGouetp+LPX3/pINBa+xnw1/+Cr78E/q3/N48lAmPU+qcv5Nqqza9L+G7AXBp+nAwrUCfi3ldN0/7znJOlaBXHgAuBaYjUqgM0wzBYCgk+johoT7uadmrUtlxTdeWl4HzDNXWUGYaR4Ad1Ri6JUoXTUVtTL88OLEvmZ5/d8Ac/ueKzWxA/chiVnpkXXcBnhwu+/1sf8e67exXpNN38JcODbFBvLven4XpK5OWG6OD1qxd88tknfPH5p7QGwQ8M48jh7DFDnLYF2XJBdjsuLx5x2E809H0ahgHnmpmAaNNMDI1lPpohS2K/P+dsmrhNM46Md81s4TVA7ncTtRSuLe2uVS3Kc15pbST6QC6J2+sbSqvc3bzheLzh7vaKN69W/pd/FPgbv/UtxlGHgtTatrr4nurt0Eyn6vrm2dbbln3pnw9Avq80GMNXHrgDfK3Pp3Q2R0I/WpeWN51+3MRT8XQfyF6etP7ozlO92s6L87QSKZu02W6m67W/4Pqpj958Pzorb3tppN6LIVqZ1mBdmzY2ga2/B2UJsCx9GKsYM9AoqTHPC/Pp60GBt0Ix6L3n7HDG3WklNz1xFPSqVBwSPDHslDoc9wr+5XWjbmopHI9Hbm9vWNeZ/W7k7OwCoRFbZQgaWLxzpFK5O97gxTHt9tqQ0tQYI8aADEG72czq3DtHWhZKTjiz8o4hMs8zd3c3VoYEYohcH1f+yU9e8I//+JpXs/baDzGAWIBKMw748MPf4IMPHjNoFyulWGnTukvx/dU3QRNI6UhwUEviZz/7CT//k59wfX1FjNo3vo4TQxi4rpUQBqZpYr/bwTQSY2BdFuIA0zAQQiWEwLoupNMR7wdDwdXxueI4v3hCDIHD2SOqeHCGMLdGjAO73YEYd6zriZJXMkIc9mrfVhrOFSiZu9sb1nVhtmYjqbpZP31xw+cvr3j6ziO9p1VdeJCvqjYfvhm9XfxBeNCXxMM3zgRecl9/9xJj4/SN8nPSB3qYhR1l2+D9d91KlAebvz4MJgit2gdiNiveAoeWWE205ED0680OmS2rESH5YJJidJhsDMSseplm71k3eQne09QwSScmC8RdMOyqZxGagcxzYN695aYiGrkG8AYIVcc0jpwPO5PKuntApenb7J3aNy3rrBtUCoezPZf+kmGcdHxZayplzIWGnm45LbTWiEPEkXHbjZKtG0mHiw545yi1MFNZqcRxwHvPPJ9I60IMnlaVrcB5fvCTL/lff/SSm6TdbjmlezqrNUpKtGKjyo+JdObVMBMDqq28caI4jhfNBEqBNBfKfMcUI1e313zy8c+4vX2DjwPiHamsLNcn8rLgw8TF5RO4uGQcJlJOOrFonHjv2Xucn+21fpdGSreATU8yrKU14XC4oJXM08dPOZ5m5lV74qPVy8Mwctifa+fk0/dpNfH65oYQR1zVIae1NZxEhlioOVPTjFRtCHNx4GbO/PRPn/O9b3/IOEZNhb3bADbXU7zusevu14sd0veQvF2tNXORNnNaemLXMwJ5+M0Poox9r+gfDz0opH2VPhYT+ejLsBLhPmW5L1Hag+ff/q1Bok/P7iWJBhNH7RiBCEk82fVCS7ENBTwDTgJNFC/QKVrcA6kPfj0NyGZD/zXXWxEEGgIuMAyOJSVqTZTUkVE9ZQvKHJS06CAP7+iSTZo2TkTndchGDGYAmhUdX9WxCCpDcArkAeua8LVTM0JJOqyT1vQ5nbDmrIyFc6R1MQGHpnH7/R4n2rb68+dv+Mc/fsGbY0ZCIC2ZlYV1XUxqqko7J8KLL7/gy89f8fjiA4Yg9K7ZlCwr8MoGgFptLzPcXd2qhZUP3N1cc7rT4aWlZL3xToum490NQ1y4bZU8z4zjRBXBB8/xyS288x7ivHZqtqI9G+jCCiGScoaW2I2Bkxd2uzN2uzPeXL9WNxvnOUx7DtNedQDBszuc86y+x+3xzrKdZptiZpVAqwVPI9JYyspsI7eXY+aPP/456/o7TIOi3dQMrnd5OkvZNYhvLm0o+LXJyzes4MGaelgy9I9uSNMsV7BNLA8CQX+Ie3OatqHzm8X9g422UYSdHrZnbFbT3T9+/z7NZnzXqljA601p22f7qKltG1uzRKclhwUf1zARmTpkVZx9lq2kqc1k6F9zvR1BoFXSckQ7xzQVctKoeaHVhVa7uYSirnoXq3WETTg/4v1Ca1mBEZtz5/2EazCMld1+p2pC50jGo5c1EWqlGaiT08qaZg0Iw7D1MiwpMYawdd3pG1445APTOPHyzcz/9sMv+NMvb1mynnR9c4agaLMTnajkhshyuuXzzz7h7GzP+x+csdt5nBeWWYOAD8oQrGtjXmGdF+a7N+wHLV2ur69YTnf4kpjvbjgtCxVh3F2wG/fQKmm5o+SV01HTyXG34/b6Nct8YhwGwuBZFjPPMJ+DdVm3zTFNkeCF3eFcG5HCHfvdQG3nLPOdTR9SCWuM+juWUkgpMQ6RtMxkAzbCTkHJXLMmWzWT1iONytX1LVfXR/aHs21Dqf5DnaFzVsoSMD8Gy9ooJiTSAFCswaxfGw8vDzYs95tr04FIf8wH9CoPsQc94u83Zr0HBjcRh2xzcDqw2VH+/jNqlvIgYDRsPfpNsdqD2lcymAd88X3XiAUA0Vbo0Crd+rZLzVPVCVjZhuKU/5/Ygf9Pr9KcAnfOOshKUgsw0dxYUAupirrbCA0fBsRpKu7iDrULV0Ao27htnKjXnXd6clb9/yEOZGcpoyh1l7M62zjnzW+u4p1SXWuuxDAQvfW5ecULlmXlRx9/yY9/8ZrTsir63iAZplByIgxaRngRSm68fvUlP/jDf8SLF8/5jW9+h+999B2ePJ20FLAYV5tShHlNOs2HQvAjy3zi6uql4hDTgXI8cTy+RKThpZJoVD+SRBiHgSGM5AqtFZbjkbTMCBfWcgo+DFoahUFLhmmnDSyjUoXTNFHSNU8fPcaPB3CRu9vrbQM6mwkpNKQmajohQ8CLnpzBBaQm47IVTKXpSHTxAx8/v+UPf/Qxl48u2Q1xMxYVzFKusYHAOv5Ms7hqfovedaTfmSbfBFqtbQrEbvtet02rJYayGboESimknM1QpmcV3eW6C7lstmW/ejCw16Pov27RLgkGtvF4tXbGyW3ydrFuyK4yFctykXv8wQANNo8M4R70bB0r6F2QVZkcgYFKppFKY33bB5Iq12+1eU66wZ0hsubXpgKgLgF1mzeAgiK9tVQpuo4YO997DwrdgGJdE2ldGAd1iFEKR7OMcdrdm5SAiV1UXrysWipMQ2AalVPPpfLZmzf84R99yevbRW90MTORku1UNVUYjSpQSmOpjdPpjtPpjvPLSxrfY11hGDQTKEXXaQhCjJ67vGrfeq3cXN/w8sVzbq5f8+rmjtO6UMRT08K4nlhaI4SJcTxomdSdlqrDPwDHS9asyQfPuDtjTZkQB6bdjjUr3jEMZ4zDjls36/cNqkk/O7s0nEb1hDEMnJ0/5umTd/n8+c9tkhNEFxiCI3pBLbASOS/M8y2pNB4fLljXwpcv3zCfZqJX1Lu2oj0T1kEKdrKi99JZ3z9SNou2ZietK6ojaa3hQ7DA34xrf4jn97Z1nSyd+zAR+z9N2zUi19p1kxhLgXaeOtFORirONYoBkeqMjOIaTvT3N4OXZgNxSyk6B9F1GvM+GyjlntoUk0C3al2Bcg8ObnSqdNCybVmJl0bzmqX0TtWvu96KIEBT2+hcCq0mghPiNKnp4rwADR8D5KLf05wBe0YntQy5UgWw6BqdQ5x2xNGKRV1hGNR9xTnBWcrmRExGfN+WqkeB1vJkXUS1aZ1VSoNWOM2Jf/bT5/ziyxtKLpRV+/OdCVfEqdR2MDCt+//1G35+fsmH3/gmwxgoVUsARCmhcYAS9HX4uhCCQ1rl6vULXnz5nKurFxxv7zillZIytVSujzNTzOx3KrASinnZTex2e84OB21vFWGaRgNidWYArXBxcUaTRl0yNUOME2N0nJ9pz7yEgWFq7HY7nJVn427Pfr/j7u6Ww+FCx7+HgKB0r3OOXJJ9zMzzLblkQtgxDAOPzwOj16ElpQx2quqizcWcdnLaOHZx2pn3sKUaK2G6dLzfvpYWay4Lxg5q9lNKZllXZSOc0x6GDWzstiaNZm26Tjo2oQ0dpVXVcxiSj9XcAptPgbf19mfLiNo0s6j9uQRqK2YEoiuw5Pvg0CXuPgTNujrAaWVRhzuboBoWCxIP3hlaZ0u+5no7goBFtsE5SvHkXLg7zYxxwPtOpWCneiObpdU4DrqorWZqtdFcM8dXpWWCqBNsPzGkZlzJCkwti4mKAi2aPFjM2WgDdrS/XZxjMO+Cioo3Xt7c8SefX7Msq5o+Lgs56810YWCKA0MMZm8FrQWC19d6dvGIb3/nI87OL1jmSo2OaRKCOWElHbTD6XjSUkgCyzpz9ea1+eN1Sa2opBq0Z8AVSlkZataso6FosvM6o9E7gus9FIH94Zx1SeymCe8dd8cjUnXxxKCYwJr1VA2jbvbxeEv00SjIPTlr9FLpqhplDdHdZ2FkTssd83xkXWd6e/CH7+749/7tv8W3f/3XOD9MlikUWg1QK2tO22nqnTnx0Cy4VRVvmRx2GAaEQMoF7xyDjWED1TY0mjo61Wan/kormQqsSQE0dS9uhgXpc+3GSUe8i2ZSiDog55LJpdvOA83+LlgJqTMxaSpKqxUzQtX7G0zi2wfbtKqGMyK9rPFbUPFmeaY27maEApvJbQcUxcqW7TXdQ5u/9HorgoD0U9MHpnFHypnTvNDEMwzBqCblj2NQRU0MOtUFdPOXYimWrjzENWsjVjUXCNKqJXV1A+188Dhxitbac3in9FyuOu48eM9uGnFm+tmBqE9fXPP89S3zaSZlnRi8rjPeB/aDnYgiFhicRXNlJz76rb/Ob//O32C/320AVi46IUkEQoCSG8fr6038tJzuePnFp6T1xGE/cntzRa0LTrRtWJHgwmk5kasyHODJpTKM11zf3NpG0iXivadVIXj9vdZUTOtuJ5tzhGHHbpfZ7c+oQPDe7OEL3kWCC4QxcIo61EOoBO8YTRqdizIQ87JwOt2R8sq8rOyGO779we/wW9/5BofDmSot08rxeIdzEJzfxsoHH9UqzHkD7RopJU7Ha7Xcco6cRoZBLcacpc8+mPV31aCh+EEzkxGtlkopumnFk0vWTeQdwTnzX0waWLwniFLGpWTFfJzOg+yHjvpSuG3defuazkvslF2AajRmr/OB1uW9XtejeJ1ULdJHsum74UWxLcEhgft+Fu4ZkJ4JaPPar95/b0UQaDaVp5ZA9/4bQtT6HtVNi1dmQFOwhmsZHTqu6X8ueiooj62bzXfSvTY7MRT0w6kEy0uhlLSxBjoVyPq4RaEjdbgVEE8tOuXGO+GYFn7x/CW3x1mFMqsyDrUkYhy0y9GidneGjTFyOL/kGx9+h+99/3cUEW/qLbAUnT4TR8cwKD6wnBZaWtQGvVXubq+5u71iiIHHl4+5evWSu5sXvD5VnJUdoEFsTStD3OHDSAiJ67sb7k63ql03+s17PZnERZaUqVXBxLQutKLf5cNgk4ZVsr2sR9b1nJJvtlp0jJEnj5/yrW9+l/l0x7yaFsMreDjnXgcn0nqi5UwcRw67gZwXTidLi6uWe1IaEqL56uvGXXNm8GFTi4r3nJ1dKCBm7EEcdoq2i24A/T+2sqRnJrVVXGnUGow90OwiWBeqExhiYBpHHWRqVmLFJhpt8oINuK8biN27HrXPoT0YLSb2b6G6exDRm0FO497/onWWxMBBJzCOwYKC4gf3HYhyDyI2aK6/OFGFpVel5b1nxp+/3oogIE7FQeuq86SjrzTXWNNCTY1ShdUpJ61uQ4Up6HFZWyFbq7BGyWAqQNNN5x4nK30eIDhKrhxPJ9Z1YYiR3W5ndZMi3qrtVsZCI2tVlNjqw3XNvHh1zWmelQ7LCiqN447d/swkoDq400nF+8AwTPzGtz7it377d7m8eKRiJNMp94UVdNAN89q4vbqj1oSESC2VN1dXHI9HaimM044QBw5eEDdzdSqsJoOW6HFxAqpuRFHA7TifQLD3UU+pGANrrqS8asYVvLIurZJyZT+NLE2Yph06JCSZ98BidtwBHweGceCDX/smL1+95NPPfqG1exwoacWXgXEYcDny2AVqEWTcc352RowR72TzhhDM2Rk2YLa0Sl5magiIBFJOgJ6wLvjNAbhPYlZtvY3k6oCZGNBowHIIo80FFGMgVIS20X/osFonCnxi/H7wCkqXWgwUvKcunXQWodgBpg1K8hUNg2wlR8epnLkrK/rfcwMDJ62r9b57slOTDaowp4T3imltjU10DKIHAL6im/iz11sRBACdIkPBidcOuJIJPpCLcsshqJa6tEahkaqQkwI7uRRy1tFRyRqHvHVQead94l4UUW5FFxUC4zgyjl1nL+S0Er32MKTN+90TnImTxNPQJpyb48rr65PqDaqyD0OIDLsDwzThw6AApBcQzXAePX6H7/7mR1xePNKpNQ7GqL3ifTx3bTDPjePdwun2muA80hrLuvDixRccb16T04J3jsP+wF0Yefei8QGFJo6E53atJKcjyHGOYRgZgqOmlbSckKp0G04976pkRBZEVI8eQmRdVnSIKSxLo6Aps/hIGC8YxgXn1cZ8t9vhvZDzyvnFY3ZXr4k+MA4KtrbsSd4RdhO1TZzyyuPzS95/9wlj9KaDVzbF91rYx00z4CwwtFYVXK2K/ueSOC2NIQ6c7c7pNXYuhWQquVobwfe6WpPmWhVnwquBSCnql+DEW1CuduL3gbHG+ADSAmHU/pMshe554C3zdPaaxTnFIEqlG6huOgFjKp1Tw5VWG871x8fEQyZlaGpf3xmwYt2IMSpAmkthXStDrIQYTNSkDEXvpNi6mL7meiuCQCmV0zzbzdIOwiVlci6U0vS08NFUVoqwlm4eiSMG5f2hkc2cwnunQgrntpvgnMfHge7p5kvv8daPcRiNYgSkbKovfKSFCDaGPJfCize33NwdSeuqnYdxYpx2xGHc0tG+kKQJ4h3vPHuPy4sLi9yyTZJpDXwEpHG8W3n16orTzRtGMsMYaa1wPN7y5uqlAWV6Gj26vODLF2e8uitMwMXkGEQ3n7fOquggBlUHvnnzii+ef86zZx+wOxz0xBddsPv9jmVd7ZTtVlhCboUwjOx2E+KsTTtnpv0ZgpjJq437CpGz83P2uz3SKjGOlHRizWpMWlvkKJesfub7z55y2E2kUvBODUBqVjYgrYlas/Z2uGCDPaLV9xlwD6YhOwNSNRgrM6HCmj4duYjvGTJOHCH6DWwG5fFP88lSaxiCY7CsU9eG37wHusxAxGtG5R54U5ryVC+hUmhbWWqHVWk6uATM4Vmz2H5w6U8WAwlBWtPsxalNnpOGk4KqCqq23ptfI80OvY2A6tnE12cB8LYEgVq5OqpbzZQgevT0HSI5V8ZxxIVgoA943wAFa7q6rEdAoffmN/Cqwe5yUR1d1zawUG9A2eydvFOUoVVwfgc4ShNyAmda9Lwq+Pfly9fcnU7UWhjiDrz6Cph2xtD6Bd8C427HNz78Ft/+zm8So2YHMaokuFQYJxXGfP7ZK/7pD37Aiy8+49sfPOPDD35NQc+0cLy74c2b12aaoqnkEAcuLy94vibepMTNqeBD4HC4JDjt5islkdKCcxG3nnjxxSdcvf4mT54+uUe7m/oilOJJpSioGEdyLRAaEcUcYvT4IGYFN5LWbN7+AyKNcRw5Pzvn8vJCh2+0ZmmuI4aRVTyHw0AMgY++9QHvvfeOofhWrzvtOixF23lbLazrTC5V8RXT1TvnCDLgCQQnWp4Y6FdK2k7kYPZdNLbUH/TexBi2+t6hMJHYxKWtk2/bQMr1VzuVc1p0l4nf3GM73dddmaEb2zi8i5QCt8fFKGvTKlj7sAiE7HCS7ssKMQbHeSvpHJ1aCL7PqLB2F9M76OvQ53TmUNP6E/yS660IAgAOtXBaswJUEgNxiExRGOOgfd/2i9uhqRVSKwrr1GptlqpcK221ykqYdhNDVF18KivBeYaoQWVedGxYCI7WhHk+EZxoj4HRTiJ1M3eoNZNy4uXrG3JujNMeH8zYsVVVe5lgSeviyDe/9V1+91/6Wzx58owQAlQ4zSZuCZXnz1/zJ3/yMX/04x/y6uVzzncj54ffUBCpZJZ55vXrL5mP15SszktOPN4Jh92Ow26gtULFM4w7pt1BFxs6ATcbDVpb4+b2mrwe8a0q9YgQUNR/P2kzUitKRYVuVBJgmkaG6Ck1MgyNadqxLIkYvc1tVG+9adwxxFEt1dJiLMyo9yllGo7dOPD+s8dMo2Z3qqZT9xuhMYyDBkun6r9SErfXt1SEGEemnZVwrdIIUAulFQMFu79gpZVESomU6wb0qay2MIq2dzvbQCOBbLhI6P6mDZCuZFX1Z2/Y6X4EYhG/awFAtjJUN2NEWlUfAAvgYhiTrpmmNDjmaCSCuZBqV6JFqk6F9rJADxubo4BJn9HfDVHGog/xpUuhv+Z6K4KAFzhMAyFEG/gJS9HmiRAExDrPnNaY4hpzOqEyYUXSWivWTGQnfy3cHI8s88yHH7xPddpvveTGEKpmGtLwHo5L5bSqIKWWxBSV0pnzjHeBcdTehVoV3QW4nVec80SnUtcQIz4qlx1DBKeU4KNHj/jOd7/Po8dPGWKgpEyubaPEPvnk5/zohz/g888+ZZ2PiMDZO0/ZTzuTGSfu7m65ev2a0+lETQsuehqB1oRp3DGOgwbPpqrJu9NCLZngFTRzNsDUh8Djx4/Z73aUWnA50ScpD8GzVk9sorUuDh8D+zNHWhMxRp3IWxrOV0oTBsB77Sx0ANPI/nBBGPbc3t6Q1tXQeA3wXeK7myYuLw7knLk73tINQ3xQEZDQAbwGrQtnYD2dmE8nQnjCbppMOKQbPqVsjIAqG0OISNA27uB18Mvt6UjJGUHIU2YIfQrRvaZfm25M6CSKH2nGqLRw7S3HrQ+rafcIf+v+BVZSWnaAfsUyC/16qxrMVQgv1hSH+Q1YZuIEbyp6zUz0sZ1oCWb5Br3FWUGEyrpYmSR+U0l+VVz11eutCAKITqPx3lM7FlB1s0tt+Fo0Mpduua2LxIeAC9HYgC7xbDgZ8cUELo8fM46jyUgjoanOoNZCCI7dqGqxZWmIAz9OTOPAmivH4wkviVZXTfWNspqXzLIkbReuWooMccCHgPhIHNTZ59GTZ3z3e9/n0aPH5tufOc0z4hyvX7/iRz/8AZ988nOOd7d2ymT2u4mnjx8xjWodPp/u+OKLz7m6esm6nGhGadamMmXnIvvdGUtCx1g1h3Ntc/oV8dSqdeOzd97j+x/9No+fPLLN4Bh3g3YgGhI/jWpLNq95s8hqQBgGQhgIQVtTRYTVFHoqqVKr+N00cb7f88Vns7I7pqnvAJrUzPl+x7Mnj/HBUXKmTyI2F34FIB1Ks4qwc5HDNLAmLRVCHAl238Wovzjoz9daSVmzQw2CIy3CmjJ3xyOl6gSrUquCzFk3TAyBaQik0k95zdSC9THkrHqB3gIMsrV6i7EGwUxE7JDGOZ2ZmJuQigFzxgwUUWBYbfBMZu09MYR7j9vO97e+iR/a6sv2fzQTymHvdS9/vXaP9kzl6663Igh0L7b+m1cauRbSrMDHNHhb0JraaNT2pKJpUrXlE52qtXyITFNgzYV1zSyp2onlGYJyuLkKJekbNQyafgrd/74RauKwU5ptWbIaZ9QEOO5MG6AbN5kHot6oYRxxPvD4yTv85ke/zdmZAmg6Qlo30PNPP+MHf/iPefXyudJW2aTNznN5ccmTR490oabE7c0tn3/+Ca9evdgm7cYYmOJETgknMI0TYzw9iPaKkzgfTCilgpunT5/y+MlThiHely210CSpoKUqk+FcQ5rOGmCAZVGabBh0Q6TsaWTdUKgIS9NiZ4+tHZfrMpPySmuiXZ1UaIWLsz37w4G9IfrVmosEedCuqyVCaVYDI+rx4A39rtks5u4NYBEhdh/Apq7VvhnvHxzusGc1UVdKyXpHLJMR9UMcvGELtZFyZoyDJdImOhOVDldD231vaUY3ZB8uK65pCYbjOC+sawaBQkPM4k1QezpvlK0zjUJtzVSrbrNc82JsjqgZioUAvX+2b4RuOOK27sRuivrLUIG3JghMuxFaI9Vqb4YmOjGqV5viM85ubiPXTq1kfG0MwTNGvZkxBk1nnc4z6N50Ws8lStHHEEFr2mCijlYoqbKsahzpLeoe9iM09exLpXJaVryDtK7k0oijkPPCYX/AOcc4jrz73vukdcE5PdVzyRxPJz759GN++M9+wOtXX2qarIIEvA/EceTy4oKz/QFaYznd8ebNFVdvXnF782abWJtyZWx2c51XM88h0tpCKsaLc586isAQI5fnF8Q4wDaerZkfuwVRAAAgAElEQVSZqqrMatdHoJSa8yo1LqWpL6PYIM7+vHEkr6um7K1sJ7r3jtM8s64n5cylKgNA1Pd8GLWuNV9+j1LEOh5cS4acMqUpil5rX/L6R7EsDoRcNNtRw862bURv4pgOTpaijz96z+DVgDYXHZCaSqKUSAjeWCYdOa+O5M76RjDBT5+BoJr/WhRXUdWhllwqMRZcU9+/tC6Ucq+DUIZLO1RrA9+c/Z6FnM06rFWijwY2GqBp8mA29qqZHfo9INmH7uhRZrqB9tC78c9fb0UQaK3x5vZIKYV5nqkI+yFycTYxTRMiwjyvrEm15D4IrYALMHhngg3tm66AlEqrWYUeYyCXpt2DNq9+GAKxwfE0czwuSkGaWCN4RZKHGJnGUW3BSt2+dlxWFSV1+5+GZRAqQwZVqL16+SXvPPM8fvyUUguvX7/kZz/7KX/6pz/l5s2VpfJ2t0TwJnh55/EThiGS1pl5nnlz9SVXV684ne7oduA68WdPsUYb5yLiB0WXtYhWV+VhIPiINE2tRXlIELGfa/Z4Rb0dDfACpcBcnwBkaLuamMiD1FjR6lLrNiwF54jjBALLclLKVRTX6Zx/SomrqzfsR5tJ6KM1WDUT+yjP7ZxY7SsGmmEKx2AjAGyTJb2v4ow6do7ByiE/KBuxFFUdHvYD0XvrVoyb5kBLC7+d+s3o5pRmHYVXVYuSS7Xf90GzGZioKDD0GYviDEDUkGEkgzEc0Kw82JaAvY86cFdNX7xt7PYAGHQdWDBMQpePPojfcA3rbDQ17S8rBeBtCQJAyoXTaeb27poYA4M7sK46OjrEkYbT6NzUNjx6rX2jl03hd3N3ZFlXxhjYT6o3L6WyrIl1TZSqTUKDafrXtHJc1PxjDIHBm2GFUY7KKXrSksmi4psYA9NO/Q6phT6qXO3IxADII4jj8jefEkLk5tpO8+tXzMc7QgyUbANRnMMPI+LUEHU3TZpKH+9Y5yOn452N9cqKh5SsJiutC1+s6cRPNFnxrtCka9CNw276c+syq9jEsiBFUUwFae2yTlShqTNjVO0Qg814FMUMhEbKyuXXVm2B60kTzXoseh0a0yqIlVq1mjFGg5JW1nlWq+1QcCFsaW2zmtw7b9kB+KCnXKW3f6vEe130pEVkyxbjNBFj0BrdJviOYzRnZLd5RkjTrsQefFRvoHMGV3OZSgZsarqvZZu0LknW3VVbd/YVIGybsot2vGhpVc0hSez5vTEPSml22bDgnd+6Bu83cL8fmvFIBw2E+9eBNVf1YqFrVJr+3Nddb0UQULq6oo0rkFLm5u6W25srdvsDu/0ZwTm8raAYZON3Y9Boua4r3rWtYy/liveZVjPrWpizttsuZi45DZExOM52O9Zsrauiffc4h5SGOy22wRxryszrHa1q96FKfZUSUttxVRI6FIh759n7PHv3PZtjX1UmHAa+8eFvAMKLLz/jeHdrXve6YS/Pz6FmPv6TP+bu9prDNHKaZ/OTF0s7FUiCZjVfQcQT/UAMIwnlnr0tapr+7GleuHqjasNaD5TW1FRblJ7tmnXdEABam3un6X2F7WSapolSK3cn67jLCraKvaZxmhjGSTeqVFxuZBk1RQ4OFwZCCIQh4pwi+MWEQs1ahtWTQcNU8A5XGmtWkVef2hNCYL8bSdGzrsmKnEbJiZSC1uvRNpixS6WWnnxpel/7llFPgWLdfNAYh8gog/ooWtu5PNj83eq+VD3Fadp3Qkk2yER7MzZk3w4YXcPBFJAFa1407YG1q2+P+dBxqIE0mnkjaFs4FgscVaoJnrr3wX0Q+SWtA29HEOjTVXZDIF5eqBosJ8a453A4t/TPVFy0rZZalpU8Z6L31nbcLPXUE905qNIQXyFbp1VtzOtKDI5xGBDR1tslFZbc6ZTMYVS6KpXMmhI1q1dAWhfujjPLsmwRPAS/1aC5VN59/0M++v7vcnZ2Tq2Fx0+ecH3zhg8+/HU++OAb3Fxf8U/+T3XgjeMB8cp1X56dcfX6BT/65z8grTPPnrzDzc01taaNAs02BKXWvBmYIDpzYYyjpp+tG690Zx115nn9+iWndeXcybbgNNNWUFZEqctWtJe+tx8jjmhBpYtPgtfUd11XnCgI1wemDnEgxJFadcPi3LbxnHOM0TOMg7W8Fqiyna46Jt5tibYTRxChlGRisU7KV/KaEK9dmeobqT+7ncTmx+cty6gWZNS5x2g7yrYuklmZedfHxykFqp2souugN6IJuKAgZHCaBdR67ymo9XrFe0HMYFTMNFfnRFZqxx1sI5fWSGtGRLOr1rTzsU8zRjorIFuA6P6Y3W8z+AfeCcaW/XJY8C0JAoiecNFrX3+uhVoSwUcFdex0yCkrb4pSQ9XoFpxaOYeoURTjbEtJGz++F9Ndl0zoxhPicF673dRs09kChzE4LVGWtKWDuMgwBfCjeRk0cJYuOzV3nHYHvvmt7/L06TvmIqw0zTDu+M3vfsQwjuS0knMmjjuGaU/O6st32O95/ukLrm+uWY+3lHXVwSvLYjfddqFhEb0ppHvd+eAZ6qAdb9U896T79lVCiEwxWsNLs2T/ftJys/LBOQ2m3ppgQjfHoG30Wa+7qZW1JJx31Jbt+cwrHzXC1Ndowh4rP3bTxG6nY9EcaiqzpJVUdOx6KeoaNIwRH0akejJl22A0NifqrpNXnMNvTk4+qqHJmpyBytV0Hiai6TJg05VUGx2HqBWco+nJbkGymVAs5Yxz2mkYrUeltUazFl+29/2+7NDgrPhGrpVlWRFp+EE1KK1hgrdiKL92tRbrcFWQV7aMrVjQ9dWGzlDN2aqZ+lWsi/Wh2vAvvt6KIOBEiKP2gq+L8t8xaB3XquCojMFtUbmUxtJRaVErpVKV9w4x0sSRU6HYGHMnQhgiNMhJTUOOc2K1kdjiPYOAC9GmGBtg4wrilBLrLaWCThgqRmtF7zRTEUeulWfP3uODD77BEM3VCNWon5+dEYfIui68fvkluVR2+3OKiUr2k+oTbu9uySmzpIXl9azW3TVZAFBGZGd9AV6655xu2GD9DSmvrGmloe3LMaozzaPLxzZWjG1cV5+x6EUQLzSxsVnidFNUwceglJVoQKjNQw2klDWEND25xYA/dVQyCbVRVL1qdS4iLmjXpQoY1KSlqWrOe8Unaq7UUkhok9gwjMSocydzSaoOFcWDclHVp27GQlpVmqtoPMYy6aTfbPReqcWANBWr+eDtMLhH1PskoFw0QKw5ba3IOakAaQjavh591IlZJoG2o1tP5FJRmy81fkl5tSAHktT8xZITYxnuqdJSeju0ajZqU2C11vvGJGNnASh1VVm1c4TmaRYIfln/wFsRBABKXrm9veHlyy9prfHo0SWtXTCNI+MQKbWxzGpBlc00MsSIGCDTEFwGcWpmqSBLtdTufly4dyO+VnLK2rnVPQuNAltTUkvspmnUEBz73ciuCbtVHYQclSF49RuQxjBEGlrXHg5n7Pb7rW70JtjBaM/T3Q2/+Phnaphasmq9pXJ5foYXYT7eaPpflQuvNVG72YXdyJQy58GTs0ekUmom+hE3DDi3KHwkJ0pelGOPA4/OLjg/O7eT3ZmhiroeOZqp0NTIw4kjVbGuNh1hpcIoyx6ky209c+3mVdoBuJgn4DhOhDDQWtYTUbTw9U42LYf3ahpaqo4zl6oqTOc9YRw2c9FqltqgTrreD8Sop2MqyYbYqkbBOcGXYq/Rs2YdlrpvFeeCBWYNSktRwNTTGC1QKn2sgrVsLcExBJWa+0AWoGgQLa2xpGLrqymO5KyHoKkXhjTdhLk11rRqt6u1eHflYKNiDiGmV1DnoGwZsMOMbIxtyLVoYBHwTVWBHt3oYsGr9LKwivXbvOVuwzknbm9ea1ocHeO0YzdOuFapJRvUJfbLacskaL93SpnVFuAYtPdfI2bmdLxlXRZ2ux273U4DyZpxThi8Y4jRalwV16RSWIq6yQjmbtNMkRYj46BDU4fguDjsVBtuQEypjWm359Hjd/BOh29uTS9VgaKSM59/9glXb67oAzFzXhko7KcdxTKhbrNda1EeuDUDtLyhzmVrUUWUIWi+Mg4joHqIIY4cT3fktDIOO87PLlS4JL12VC88Hdahx5DauisEmcFAOrOyamzFax9k0X9/h94XRcOLTjVKOh9RF3YPwlrGhDAwjhO7aUer2iuyUmm+MQ17EGFZF+ZFzUlCCDhp5LyQktm3eTUILWZDVksjlbrZuTkr77xTn4BSrFTpBJqVADln1qqS40G61bmwFmVuhiEyxmE7iUvJ2uFaMrJRgEpNSim47lDd+Xrt8KGUVWk7pwam0oqi+a77EZg/oQV6sdJMmQfDGFp/3U3NQ0uFYFITZ4FHLYvUPaOBVxHpL73eiiAASvUd9o/Y7Q5a24VJddulkmtmjIExRnMXCtYIkgAhhoiP0TjvZs0+KkrxBkRF16jSqN4AFbE3ETFjkspalHYaowpHotE36uHXTI01IqLyWKHi40hDT8knl+/w7Nn7ukmkWb933TTgyzzz2WefkkojxFHNMtaZw37i8uKcnDPzvFCzLtrukdDBuFyyuRYHNaakI9SaDbTWiMGRs9N+gvWklGKrTNPIrlObhmV0h2YRZ333ql1vovMfihl8qe69EZz221cDusXJhoWAU+A0F+Zl5fZ0NMyGTb0mNJyP1FY0CHtH84GcQUZPraMCZtaM5QwsKyVvG0WkmQNRJa3ZhDVa15e8UoZAjlpnRwMvxUHN2l2o0mIzMK26sWKM+KAiKh8co48saTV59KgMSQcbncdbC7A3XYF2rxpr0ETxBtd1BGYeag7I4pyRdfoe996LXjp0PwT1WKj0ORAbI2GYTDUWoxR1w26pbjoBMfC0OqF5r5O1f4lY4FfECBCR/1xEvhCR//vB156IyD8UkZ/Y58f2dRGR/1hEfioifygif/NfJAQ4F4jjObiBaJZYOgRSF5fKNb0pzAK5NlarpbTmjUaJKcKbqw6mCOb1DzptKCVF2ENU7rjrt7oCTERP/N006chxKQQpZtWtpiLZhCJ9A4j3G7D57J13Cd6T1sV4ZX3kXu/d3d7y5s210m3BU9aFklYO08B+d2CdZ+b5jlzS1rZacrFaXzdDLlmzI+P6+6uvJkWlmWY9BIZxAmmGrHc1nS6U+xrRoOkH1lld+upEW1n7CSqGivce+P6edZ5dUHzHOVXe9ZZcbxz+ZrhhpVBpldIcuTmqBMSPFDyGvyM+4GNUo5VlZUkZ5+/Rfn1+lcke9pNpGPQ5Quhah0owFqeWor4GvSVd2Ph+fW/TVrJMQyT6iDOWBOkORJ06VVVlb1tuQJ8oXHv3Zf/+HgBgo1G7Dd7mGNTMBNV8C0qtJDNHWVJizUnLlB4As35ecu+ULGatriKn+uD7Uq6s5mHwF13/IpnAfwH8J8Dff/C1vwv8j621vycif9f+/R8C/y7wPfv4V4D/1D7/0st5z/5wTms6+beiLzjGyBCipYPgYyQXSMvC6XTSDTAOCkDZSeiDtqcqQBtxXmWgfcZA8OqyknKiFAherDuu4o0+EtHx0cL9+PNlXclFdft9MmyICnKtKeGCZ11Xbm7ecHGp2v+UdDOPw45SCq+vrpjXlRB1iMjp7gapmUeXl4zjwDwfWZZF031dFSaKsdrY5tzPZUENNkddQLVRKZoxDROhDaRSGOLIPM+c5ju+ePGcJ++8y/sf/PoGMuqmZas7O71mmJbNZAzbdJ3adOM70ZkPIjAOkTV6PZWb/sw4Tuz2FzrgJek0KFpV0NUQfW0VDtwdZ15dvcJ52O8O2kkYJpyriBmHuKplgCrqTGfvPYNTG7mGuiHVUliXBUEYh1HFOCHgXTcc1aCaa9koSO/EsgU9D0tJeBqHcSQebD5cK2pVbptc+YVuV6YgowoCoDrDoZzgqmZdndnZTtxWTJ3Yqc+Clfj2PjcTHzUzRmkgHRwEEaMYrQRzCJMTGp7uOaAmelY+SD/m/uLrVwaB1tr/LCLf+jNf/tvAv2F//y+B/wkNAn8b+PtNj8A/EJFHIvJBa+2zX/EkRm+o8+vZftqGj6gVtKfWzHI6bp4B57u4GUMAprrTtlhqVVDFBVQFJ4YVaIReSyVX0dPewxAdMmiJsayZdV02w5LO4aopRcC5gJPC5cU5h/2eVSKpQVlnXr/6nFK+t0lQ8zJrClkLy7zy4tUrfBioKXG6uyWllbMx8uyd9xjGkZuba9Z1VjWeIdGCuih771ibgmi0Rkor07izU1o7x1JJ+DbpBvGR5isxDszLiZevnvPFF884fefIeb7El6bTlFrbUHZntGsTzH6796l3dL9nH9BlvGJagoxSZik3hmHg0aNLxmnCucKm2/C9tbVufQOlrnivjs6trnjxTNPBePdKzjNDCJSxWFmgA0urWcRpPNO0W2Jkvz9sMuTu7OwMVXetEVplRE/dZr9H96R8qMJsTTOIEAO1NIr1CfS5iH02QjHcpsOjtWrJ0ZqQKYrQm6/D/ehRdaDuxqqdQepYRc8EQDOdXipkEx1tGUdP8UUZDMTMRTtNK1gAMDjna66/LCbw3oON/Tnwnv39G8DPH3zfL+xrfy4IiMjvAb8H8I33n7HMR61dx1H59epooukN1bMuM7Ums/H2liKabBWd424TyNVaXKwPwDtCHPSG5aTTYYqmaHGKRN8FHkoP9cXRagOpKr81l5a1aBnSWuVwmDg77PnyTgEkFwKXl085HM6Iw6CS2Vbw4klr4tWrK16/eg0oE1JLQlrl8uKCR48eUWvj5vZGM4600kqy9L5Ri2itn8z+q2dMtVqg0po+p5XkTuzGvXXYJUYLAssy8/rqFTc3Nzx68pQ4ZGqNeNj0Du0B5SgIway91FgTKyM0G8F47T7dyXvHYhjGOA48unzM06cf8PrlpzixYedO27aDk+3eXJydc9hNZoGmNbDiFVEVdQGc7JR9Mb2+NBtMUotNnK73h0H3neilh6Xq3Yu/KQJK6YFomxvZTCfQdQhKCXab+J4lldI5/0yTphvNgMBer9/PK7TncdayJZZBNF2D3SJNg4lu1lLrFgi6gYhBC/o+W9t3LhqMvAGkqw1T6Q5ZPWN4aEn+dddfGRhsrTUR+WXP8XU/9/vA7wP8td/+XusW1d0UwVkkd60izTNGB6JGEmlN3F5f08RxOJzjY5/wo6Yg0VL8VtWYcsnFxkErmDbtRkRGYgxWJ0JNJlulMcZomYPdbN2LCCpqQYTDYc/Z2Z7Xpzsyjd3+wHvv/zrTbocTIeVCf/tfv37JT378I95cvaa0yul4S06J6B2PHj1mtz9TgUxWiW+rVZV46IngnSMvqzny6gLVFLQQnJCcU116a5rFuIEYBrzPhFiIMTKvC7e3V7x+/ZL33nuP3f7MZKvVjFsUI/FWCmk2I2bWqZsrG1WXS6UV1Uw4ka1OFudU0+4C5xeP+da3PmKMgVcvP2NZM+LUCDP6whDUsiu4gVq9MRFWchTNhB7OE+xzBjX91+BXi1hp91UA04tXW7hccWYCi+j7Kk1do/usyegDgQZBT/FS7v0s6qKbtYPRrQniogF63tZXLy1s7kWv/8Ws6pqOgNeyAssWVLS2Zm1nVkyHe1CvP6JlGriuOjRfxNYoWdWIWsax4RrNOkKCqWqhbTjM111/2SDwvKf5IvIB8IV9/RPg1x9834f2tV95JRsv5muDVhm9zrBzPthgkbpFdt0MmcP+jHGMiIuAzQQwo4lOAXXWgFZ0Lt4waGJWi077nZevyGudCNUpFSeYA4x3JvnsqZbjbL9nt5uo7Q4EhjgyjdqxRlPVW22FlFZ+8tMf8acf/zHOq5Q2LYvO6wuBd955xhgH1lmnIemmU2Q410wfP94tpLHZCMus/obBUmyxmrOWREozwUcV5rAQQySlxHGeef3mJcfTkbOcaHWypqSiAKA5/8g24UY58P+HuTeJtWzN8rt+X7f3Oec20bwmu2qyqgQWtstVeMDYEjOEhJggPAGBhRnYYsIIJkayLDGgmSAhGYEQA2x5hoUsITFiAkKMECBXl1lZWa/JF/Ei4sa995yz99cx+K9vn1tVLzOLspHiSK/yVbwbEfees7/1rfVf/0bvzXjUdasuzSzXDCAsrdvKQGPB/nDLp59+l+Aata58/bXhGF2HZ82F85I1am3MR8mWnT3stXWpQT1cHH1tROneZOfIiMPERgpfObOsI2qsmQJSwqZpmtiFmWmKgFdnVQvTNBFCQk7EhfOysKyriEBpYp7mDQidYhBz1YnvMDZM45APk9nRrnsnH4Fi7scKv2m2yutmiOu2Fez4vaOzgMvKsNuY4A2Y9m44Gfutg+3DcYuRUlSJ/qcf9T9rEfiHwL8J/Mf2v//Dk1//m865v48AwbufiwfYy3vLF/QeZ2KM2hrdRWmGQzcUC3a7HXOKxDSBV46f1n6YntsswpvX+giZh3Tg/f0DeVXgpogZ2AEz0QrgzAN/zF2HdFC6zbrS0EG95AWoMCzLwvn0qEPaxQ1orXL/cM/nX3zBmhu7KFtuccIDV1dXvHz+EpxnXU4cT0cdPEP4c9EHHr2gHt0gengka62mQnPQBUTW7mjLCe8nQ/nFZktJG4bFMhJqq3RztTXSLc6CWZ21noiGoIM07Kye0JdrFRYhcU7Y2I/4QIqJ/eGK588/5s2bn/Dw8EYzNR0XAufzymknC7QWtAXCjo8ufIfMRoXvjBzBHjSy5HXhdD5RmnkLelFu1yIzk+W80Mw0tXd5JAqLCByXlR36809nGcQcWlWKU1MRHIlO8zSpMxSBYhsXR9JQsO1Q70HOyNUsyLodeGsho4cWnD1rDe8aUxI7c1iR23HXGnOQtPBbAVZYqY0FFpk3Qk78xoGQpLuM4udFkvsZG8KfXwScc38PgYAfO+f+EPhbdvj/gXPurwE/Av41+/J/BPxLwO8CR+Df+tMVgMDN1Q2tFc08XiaV69rofQUfSSnhu4PWFXqZZnor1LrKFMN7vDHzOor0atZVdGApEgJBITh4fFyopSrDzgdkQgJTkBGJj4njaSFgPgG9K3Q0a1Z8/3jk7uGkEcMFallMwaiuw3vPeTnz1VdfcTqdSdNEMAKDSB+FF89ecn1zS6uF+/fvOT1YsGmtm6GlnGaGZ50VRxM11dqYolrTtVTTlzczHslM0w4fA75NhFwY5hrFOo7eqtSbXfl/g3o62mdQgWzd43vfbqNmJCWt1LAirMNZm262eYr4PtEPBw6HK1rDcgId+3k29F5o/FrkdxC9o9aVWhoxzZvqby2WFOWka3BALpnzsti+37AgW0PO82xmKPazdpjniWkSZfp0OnI8Psri2/aiS+n4aKSzNDH1xm4nv0gJsIo60zgUfWPYc9tGRVsLh3wuHT5cDD5j8PgAxezZWl2JfmgGrBhbmpEz3MUZxj8yE2qp258/ft4tdh3AmXUZ4CrKVbCzsZlOfcPrT7Md+Ks/5T/9i9/wtR34Gz/vz/wTLyexSe+BJTdTc8FujsxTwnnNZEokxlDmhguR6KSBH64yrle7qAw9tYosEVCzm7orwix4llzAe2o9s4swX+3AaWZU/pvjeDzTemPNQod9b5zPJx6PJ2EEPbCbZ66vb0wvr5b24eGBn3z1lVSCyGRj7N9D8Hzy8cfGDzjy7t0b1uWsQ18vzrIi3FyquYRBapNrWWG315xs/HZvK61cFmKakCW7iE61d05HgYS9Gf+BQfoFNfudzaebIektUmN6M+VE3UEevAQEuA0eQrZT6bxA2eub51xdP+Pu7hW9F3a7xLzfq2h23VrSD8Awx5RvhMNFAZS1OXP4qXYRTExR4TRnY1lqjebZpURLiZxXc1lWIe+t83B8x3I+i9uRItO8F6nLik0MaItggLKIadW0BYNLUbf5vzZrzfWxAyb9jdEKgI1JetdMMwHO2IkaL23oCgpG0Wdv/gKYkKuDM22DQExzg2JoGxohek1klnMRx8XYf9aC8ENhDHZ0UyL+/ZQCOWcFh/SO64XAZT/qzIxjMN6UISgwxxvCLUxAN5mzuSuYmYf3cnQt3dGybtspOKJzlIpm8SjEH4QH9FaJKRAT0BshpM0yit7Z7a+2QMxuINHbN695PB4JMZKXM7WI4VhK4fnNLZ9+6zvmx5e5u7uTY26vGytuSEljEENQVF5ZsJELp3Xh0HVjtC7nm2g8iVoz67qI6MJQBzrWvPB4fBSHoVZ87Nu6KwRjv1n3NFCAtciQJc1+Y/E5EyCVUogboOW1Suud4YAbokDCZ7fPeX/3mik6rg+zuhR7X+kCXFvvxnoTNrFmpU8HW18mb8XmyQgik9eFZTnRe7EWPTJ5xxyTLgcbKdSJYOEsMo0NvuNdw1OJ5kcxZNgpSMLeR6fTnTElA8l7YtDnPIhT+u6eJAJbd6Sz3m10E+W7dqP+tsbwh9hmesN4XEdJxgiMheFheKG4qaAYIc2KZ7cVofcX/8OLVvRPvj6MIgBaezVl0Ts6aRJVF6ORnpYTuWitEyx8c3CgXAjUXMWaqnpTa2tMQVbkrUNK8s2fkubeXB29VHbBQStqsbzCRvKauYp7dvO0gUGhBcMVJGne5mEno8ibqxtJbptirO7fv+PLn3ylA4YOqMdvBpMfffwJL19+hHdwOp14f/eWZbWI72rmGchw0tu/ty5Eu3eRnZbzsu3qHUOQ0pSE06rdhDMxJGrQOuq8WBGwFVvE6T3uugG1ivJWiOxW6mJi+hjFlByz/xOwcPyS8AnhBURP74H97sA870X48nCY45NoN7fttdWpDVCwMdKES71Ya21Ou67Te7Y5WQBmMwRd+3yFjwRnmv/emVJjv5vNnFNYjjf5eOudWs70lgnOK9zUrNSbveftyY0aN3kvXAhBl5VtN2S+N8ykRF3KnKJky7kaK7Rs5iSMt3XrIboIY7Rt7aiWYfRuhiKMFaJ31NKZ52CrdDYDlp9BE/hQikBXnFetoghZmskAACAASURBVEiuCyl6ouvsUsRHsfRK67KJil6U01w3y6beRvKqCsCatT7zPnA8LbjcOMwSgqQpmtzVGaMqAZERLJmmnRGRtEUYa5tWtRdf1sLD46N2u11MrqvrG3sgC/TC12++5v7hgXXVdkEDuw5ICpHvfOd77HYH6JWynFnOJ2pZrcK3jZmmcA49BHWIhqz1zmUxfYEUZM2EMCE4KebaSikSQnkD+XJeeHh8ZF3zBlw5QB4F0fjwzvzpnEBSzHFom4H7hhk0+lAX2UEWsKU/1H6HmYrSO8F19jthJyPjz7nOlCLeisAI9+gu0mvjeD6LLo2jV0exw1NNl1HW1VKTgo0jBe868yT3Hmefkfd2Twf9OgxegQfLj2it4ePEziLlBrBXzU5tOBPRhyjqQgfWZyTl3mbq0TGPgK4ONgSmGGhNfhXCrPQ5dANjL4JmR/cD61BXoGesm5WYCrSz0U0/m7ktt0Z1g3fxs0/fB1EEnHPs5sT5XHn39jXr6YGXH31MmueN6okPItDkFU+UuCeq1cUhXn9ttKy109V+Zp4mams8u5424KXZPnwOchQVdjjeSD0YHem+uxE/YvSXfbiBaeO/O9sZ7/c7jSSlcTqd+Ozzzzmez1pdmSDGyRyPaTfz0Ucfi1a7PLKuJ87HR1rNmzpQgKDug1KrGXu0rf0U0i9zkpCUE+i9IxfdjjF2ehORpPsRW6WDe3f3juPpZDes8h28G3toudg558elI0ejNiytHMVaVR0GE2G1y3sXQqBkR7WkZuciMU7gOrvJKxy1rnTbYKjhqxYBZ1ZeqGB65znsdjoUXRuA2nXQcNnEWiKZzfPMspx5fDzigud0XnHeadQLRqDpF0KUD0HaE6fvI02R4cNTWqWeH9WV1UzOK6fzid51mczzzDhdI0JcGRTDFOTyfhvXjFILAW2G5Iw9RrnLMDE+c4ej+YuZ6DAuxYH3Y9SQP4Lb1qbmhdydLNV7M35D2L6Xb3p9MEUAC8bczRPR7W1NVHFhJgRp/X1MlJI5nxfcPMt1xlkoZEis90fWZWW3m9jvd4QQOZ/XC3dgINwOC70Yt4Gxsgbv2lpSMbSA1lSlbTXWzK5rUEjH/rrXTGvw+vUr3t/fA+I5lOUkLNk5aivs98+4ub7RfF1FcKp13QC8YHhHzZncDQCqKgbVZsPeZayy5swhTgxDrt4ba16Vw4fEI97LZFOeAIXzcmIt8jK4aAUuxJxxqwt8Gi143zjuoxikGIVdbAIik9SqLtsNFnBOMWghRHazU2xaG5l8TbLacfBhE+qMf7x3ugQMJ/FBXdqaV6J37Hd7WneWQiQeR82V5i9jTTSDIGke5WfpSgAG9dwRQiPEJDxleVBHA2ZJXmw9qFFsWNyB2dXjbR0pqbaH7UbWV9nIYT4WKQb8pMjzkus2UnmGGvGPnpFgRXyQkbBmLMZg7FZn5CXRowXoyvn656kEP4gi0FvjfHqklszN1TXh9sY87x0hSEA07CDXHMlrsUMpkKrXTl2Oao19oBFY10LvmVLqNvuF6BUR1vsm7vExCkE25pkSaQJpTnbTmbMsirA+r5nzqrVTDAnPuu3Om83Z796947yOcElV7RjCtvZ7+fJjbq5v1FnUQi11u1mHHNXb+qh3/azeDuYAulprLDlzPp+Z573NfwFYNWIUEY0cUKvAMu91o6/rmXU5kWux1aMVOHuyuqH03jcDCvUZiZVoisDgcEWNcDVCVa/QnbgW8vdztubUz+8d3FztuL29EahYmy3aRrSX/W8bXPqGc+Jd0DR3l1popSIoD0ox9yhvxjI+cHtzJSDNmWdDLuRacG0UbKAWHIXsjjh2+BBZSyaWid00Me2ujJYy+vMuvKhkdYMO8Vns+HonSzrRrAe7Ubf2yL0YduqDjxJ9JAXobaUY/friAmaYBZeO0IPyODGDGdse9KCcBOzXhz+id+P5dR9+JzDQ5FKl4vMxEX1gnuXE22xHD44UJ1OMFdZ1peNsFQa7GfokX8KxtvFO++lWCzFOhKjZMEYLpOhsFk7ntXD/8MDVfuLgGjVnYlB2gbdVZAyOOTlurvc8e/ac0/KadV05PjxQX64seeH9wz21NmKa1N7WAkHc/8N+zy/90q/KDn05iX5rKyw3VnKmSNNtbPejCZhiqGoN+0LJmYfjkee3t4wllPcjgaeSfCRGkIpUXzHsu+7v33M6negvzZNvHH77i2u3VZPD5LsXc5FgrjrO3o/enAqoHQDtrS/GFt05uQuHwLPDzGE307uUmVh578Z/2AAy+7vGChOcWnQDY/HdRE9p4wWEEMU9MDCxdZnInHGkKRrH/sSyFHwvzHMwC/eKt1Vg7wXnD/iQcKjo9OKeZBY+UaOCdTRmY2aHNpjZpxiPwiSG119wsHlWOidT1pDM0HYF17ZD3mrfyEC5FOtM5bk5/n2oDDUOtI1dWYryGYeBTPDpp56/D6IIOLOEduPANz1cuGCQhwCcRt/IIqUBUUKiwcLCiBF6w239ZI9R7ZqlS14ZRowCXzuty84rBs+L5zdE10ihK61mMl08Du/kEOuC4/n1Fc9unvH23XtyUaJuK4X7+3uOj4+MSDDZcZsQyjl+4Re/zy/90vdtgyQEXKaVyaioek9MT2KHc7SH3QxOCtUHeq2cbU71fqyOBsZy4Ob6huPpbC48RsdtzTqvEzmvdAu10DdUaTXYg2QAYMPoqbqzHYOroIPZ7NbWH2Ejhes4PwhOY8aVPfzLF9ekFDb/g9K049G75Qzw0m05UqYG9iIDFAsLDaLayvNgSKBN8egkdOpWCEZkPfberOtCLysp7igEuUJZpwKOkk9idtJY82JWdOZUZH9vG6aivQ2MlM1SzrIaW0dbGLvARmcmcxrRtINFhnXnWLLA8RgiwQVFrjV1NMGFTWLeujZTMMZbw5Bs5EheZK1hkVZK4YNPIAI5/cwWTxWjqtf4xuuoBL2z5kwfhhrO202Un7CnKudlMQ8C3XpT8LhpZ06z0hiogFdFyxnq74Oon87rzVzOZ8p6UZJ5q/DRO24Pe57d3LDfH8j3D9S80lrm/f0Dj6dF9N21U82qzHvHp59+yq//+m/y/Plzal4Fhhl7bzdp7XkE2+1jxp0y/Gx95P3p8Jmpl2y41kyyyLAWA3THNM18/PIT7u/veHt/ZMnNeARxu/0YuPL2gEju5izQpXePd8qDjF4zb6sNtpXTcCcC38HF8b13BuvQDYAvBA77HS+e3WwmpHMKhDqo0NIAeEPD5RwlAc2as/wEcpUSsRVCk+qw9M4KqmGmpnx66FurjE15CA7vZjhk8iLUntro/YgDdvNksmIJgzQuiUGa0mSdmdaW52XZthjBOV1amKDKSXrcm4pSNsmwLRpIMbAzTYsjKejVQEp6MxOaDk4cifH81AYVSMGR+wW7GvvZ0Q2ORKUp+c15eqxZv+n1QRSB3ppstTbAyVNjMHNEPWghBChFOvtSt/28ZtSLMuvxeMbRtRmomRgUVR1cwge5EJeOWX5Ji19LUcBIa9wfH3DI0CK6jt/v5AxTKs43YkpUqhyKk+fqcMWyLJyO7zmfT9y9v+O85M0O3COtw7e//T3+0m/8ZX7hF76vD87wgZwXai5y+IWLxzxqOXtXN1CK0N7tZsRutVxYzkem6xvT5Ddx32Pg+e0tHz1/Qf/xj8nV8+7+PWmaTUizY2MK+nGkDRhtehBDsLbfvq9WismqLRQDbCcut98B7MUQTM8P1Q70NE3c3DzjtMoCrlSZxoTejcshWe4wxAwuEGxTIBako64iFOWq7IiSV6PUOkrtRh6yztBa5jHOiKiTiNOO3Zxo+SzgNGfWtXA+H6EJwEwW3xb8JMNOzDg1BB3m1uhdxXstF1cfMADZO5r5H2rDNCTAZtAaogl8bDtlBADhVYPGLYAx2MMgQNiLyNUDBFhyuWBP5ijsrSsSHqEsSvjZ9mIfRBFwpqBal7PehDlRa6Daai1GHWRZVCFudZfts6aCqNu9OaZ5J7TWbgDvPHf3R3CNq8O1Zm/zhpONtjAIAdWS/0ojLivu2twFOffGHkOW2lNwvHz2zGLVI6fTI/cPD5SyYmF5hJT49re/x2/8xl/ml375V8ilcbx7h6exno6cTmfeP9zzeHok18oI3uhGNR03t/OVVqopKI1I08U7vz8+st8fGLFmMQZuDgeur665urrm8XSiNOXj5XUxo89pAx+9WYttQRV2nrfVqT2IzkC7EDwVMd68k41b6VC6CX06m6nH2nWjxTjR3cwXb860Bof9XtLxWaBsL1myWUt3dnR6lBlJijumace+StV5Oh95XCUZ3x925AqxVvbGA6lt2GmNjXtjJBYNBqBPE2kbw87yEnBaDXpbs2qjEmk48QQscwDnmGPYbNBra0b1tg6odyZ73xx6YLUZ6sb16Nt41WqnOAMYjSGq0U+egQLFoVNYSmHEws/GiSi1SmHgLQnL4tHH6lyEu8uo8k2vD6IIeOc47BLdTCKcC+zniXAVyaXy9u6Ox9ORKcqgo/dGSklvbIfD/loyXpt9x3oPDvrQfGNKmic70iHQdHs5IxTVUknOMV3tSClYiu5lVVVqpZVmPu6Bw2HHYRd5dnPLfi/gJ6+rDlwtkjV7z7Nnt/ylX/9Nvv8rv0YIkc8++33+8A9+yPObK87373j/7i1ffPkZb9+9ZdifA5f2n+0MMnZvZRMh6UG7f3jkxe3tZvjh6FwfDtze3PDs9plZozlCnPniJ59bF6O8wDEHD0KUa91uMkt6apeHyBlxSeGjaplpT1ZhTus551Rc21i79q7vrcPnr498fXfk04+emZHmaCyCFVosZ69v4JnDk6L+oXd2U+LmcCV03FkyUM02Gkroc5h27OZZyUTek3MROQd1Cq35zZcyxkjNdvDTZF4SFzcqqUK1AdlN8rscfXgwEw9C4BKrbkiUjbC9iyDmp8BmEroZmCDbc+dsfWkjXzAPgH6Z+1vXRkgjRSR6z/V+Zin1Uqhdf7JZEqg4iupPe30QRaC1RisruymyOjbNeErSakfvqU3qr2ZOrHd3b81Mc49zCseYUrRUmGDSdkcIkzzlnbOHWMab1YwjhvbbKDIkH3BO3oJjnaOgEUyTrhXjFB03VxOn3DkcDngfeFxXnr34iJU7cs5c7ff82q/+Ob79vV9knicejyd+9KPf4we/+1tEB8e7NyznM+/vXvPu/TuGuKkOsM1hoA5sO3Tj+Q8iU0fEkNP5xGxzZnCQy4IyEWY++ehjcqmk+UCumfvHs4FrcrltTVZsrQs4HODk6Ejaplpzwvz7xfqqyW3Q2IUmxcaZIs7ozFxER3ePmR999prvf+f5VhjGLn3L3OvGf8cgEGSW4axCjPh5bKAJIdJqVM6ivXHRW4diblIheJKByiEEWhc9u7XGFCeaef6PYNRL7Jm6ytbkYOWCSE+Sdluf0di0Hs5dtP4Xca+KeujD9WfEif3RtaFDBbQP4NVbWajiVMwxEZ2n9CoGbdHfl0Yhso3EwAeq2Z4Fy9/8aa8PogjoXc8082LzQQGTzdrjGDw5q22dYmBdz7TouL55RpoPYCur2gSIhOhtrjKarVdLKADcyBvOzBztk+y944Nl0beVdQvMwPT3Nq87gYu7KfDsauL13Vk3kPMcbl/w5//id3l3/8Dp8ZGPXrzk+9//NQ77A63Dw/0DX37+GW/evCafHzk/3rOejpzPj+SybrN/rarsep5NKmKH3jEoqZfbu9bG8XRmZzjImjuPj/c8PNzx8ctP2M8zwa+ycW/f4/WbdyJQgT2klxBOurIeNicdd7lDHFI4Vtg6BOc6Pjh887gwDodus97t9xq6H2OglMoXr97TWmM3JXIpxNC2Dqg3UYI1WktApqNWjBSjz6HTzUdAnnLVK905uHkDT3trBAAnaznvVFxqGyOXGA2tiYVYaqOtq63fjKthll1+s5+rtgUZXY++l2ou18F7HUqLJxvYjT43mdX0OpKDrSj54WQ83tOhz2jbZqs32cmH4CFDc23bpF0KCzREL09RTtu1GwHtQy8CHTiXTrEcgdoNFXWQUrDAyUrv+sH2PjDPB+bdQfO8cddb02GIwZlNOPYwd1PMBROm6NGPNvcPJLs2OK+rWsTgpaprjZwzS2n4VQky+1kmHTc3e4KTt3+IkaubZ3z7019WFxMlQx2oci2Fu7u33N29Ja8ry1kBHWsW/7+3RkXbj3HriOFoayib1fvWcncDUgWeHo9nbq+vxTSslYfHO169/pKPP/4Wh90VIH7+Ry9eMM870k4BrjSz0ArCE5p5FgzbK9tsIq2C1l3d1qu5arVaDfbuI3XHa2TytsEZP38pmTVnvnj9nnd39xymWyYv0Q0m9GleNue9NZobqjlnRWVEiMm5WG9JY+jrq60ExybJ4WwjpBOg78/TSjexkZh+3TlKL7x+84ZyfiAFCDFxfXVgv59JUSa2tVZKExA4JN4hRNmHu051UiouORPXlSkGJksv0nOeL5uY0ebbGdDK0Il6bN3YxeTVm2GoDnoM+r2pD1NUGAlF47KgO6YUjYR0Caf9ptcHUQSUIqOHcNw6u2liSmmrZDdXO9LuShVzPYHNOiEIOAMuD2KX2CgGZfGFFC6+cd3Re+F0XgjOM6VAtAiq0HUzatbV2lCHcjSeUEsle4k4dMsVal5xPrHb33B7c8s8z+oqilrGUgu5FN69fcPjw73WhvagDoff8WE+RXH1Nc1mTY0s4gIMsc7Y53vWUjkej7x4dq2DWQuvXn3Bd771XaZPLLG3F+aYYH/AT1FBqraT19+nggNjzge6yFTFLN50QbXtn9bVHTjAx0DvmrW7kXe8cxRri2uptLry+t0jX7+759OPrswLT2OIbvlA7Z5cz2RLAQphFJTxzjSWXGE4ST9xhR6JUQ7DIcaaEh2wTU2ji9ieIylErw4H1tCo68LppCK9LDturq/0mdpf0rqcnYJJmXuTvmTygdWCTZe1sCyNKUb207StvLsh+cGs8pz5J3g3+C1IMt4v+o3og3IG2ujYLp+VsJdmeQXDcNWwAMOlnlKyv+n1QRQBBwQnYQU2uy/WlrUOeV05HGZurg/6octEWRe1V96b8WUWWOjHKFFsr1twRX92CEmy2NxYlrOFT2j1E53HR8/hMGuWq7pxSqlkOwCtinmYs26E3gv7ufP2vhDnTpymLfzT0eleQRDrunI6Hvnyy894eHjP8L0bfgg45cttraD3+K6VUu2DZtq3YjE+TmcPTTWw7u37B1IMXB92ODr3D3d8/sWPub6+Zbe7krORk0FlSokUpOZrtdDDyBfwW/s+MAeNCHmb1VuxdOB2ebjGCLPlPGi3afZu+u+6QR33x4Uvv77n+7/wCfMUDTwz23nTEUhyrUtBM3yn1/KE2QjeGW/fOr0KDJv5wdUfdmDj353TLV6qxX6P97UPL4ErWpzZHSwCr4iV2XpjignvlYSVhq0YDhrk2gh95BhM6gjWlbU2YreVZ4xGDvKWhaDObsianRttu+jJw3ZcQSRiFG6+hUG25q3ZGGMFOQaZsAyJtN7BMVZ98+vDKALOEeNkdlJ6sHJtuHWERSYcQTTemGjdsa7ZuO2YK64MLeVCWxgMwrUWqsmMO/p9uM6z6yt6E+jme6UX2ZnN7ElpVnyTrVymJsJSHisae0P388xHLxbePx6tVdSUJwZjJa+Zx8czOa+8evUlP/7RDzmdHhQl/QTx7WLoIO+5i+JPdAIpF8dGYLR+GrsHRVht4HntvHr7nsfTmeAl8Mn193j+7CM+/dZ3ibWTJksVsg6gFVs/2vvVrcXuQ0LcuyknhR3UUimGUAPQm9lpm3jIN9lbDQ673cNSWC4aI3Lm1ZtHcmlMCQuYZbuxnfPsZsl4Sy0S2Fj7HNIk8paBlANhb80MUYbkuAuMqyYPrjYlSCNSWNaMc+IyDBAWsHZbhCuis9jvynnJ5NKIoRB9UI5j06ZmLfKTjDEQ48RNOrA7zOx3O3EEQpCQaut03VaYhgpzqAC76UZcN7ahbaZyHolNpuZsIwYd2wBgt76lM2NgpJc13M96fRBFQHNPoLuw7ZddX2klk+bIYb/Xbr81jscjJWft8vvIJbCHoBZaPVlMtx708bDnIuVYqZU1L6qYRiIRD1xzac6VUs50J9fc6ANhDsQmrkAuXmEUFod9c5VIsVKcNz86k8PWyps3b3jz9VvSlPjyi8949eonrMuKn50i1NFBnuc93gfLGyhELyddO/EWfVY34dCFsmt0aewZ7nBalKo8Zsq390devPgBh6srbm8DrSXx4HunlUyrkdBn4wSwgWoMS7OqkIxN4DMUf9347dtBk0ioPxHu9DYOnph/y6LBAOdYcma2kWQtf8w4xBltuun72YI2WqO0Be8HDVbf82ahBlv3aAOBBFN5mHL0y/+2Tq6LqUu9rQq9ZTUa7bw3OpWROTQ2U9lXYhsRYgI8W5WmIpoT1JR27Cz5qG2Wd1ZUa8NV2/AwvCS1zdCoJ08IV1VkclEB6OgyqrXjWyX4gRPo+/a2wtXl0FhLJQUBrGOM+KbXB1EEnIMUxd5LMTJNkVYlfBlBCsE7E7o0UtKBc85teWsjaOG8NtZzllV0lG35NCdqVR7h+8cHluVMDFKgpZS4vrri9oDSjp0xA73dEhR6NkDMyB7rslDLwm6e+PTFNV89O/LVUbNpt4N5Op35nd/5x3z1k1fc3j7jD370Qx4f7ih5pYZIddL6B5/w07ANb5Rxf9bLwfNOs23nyRoNOc/q/Ru3h3kOtEazVdV5OfOPf/d3uL2+4dd+debltCfFJGrrutCmiVoL0bYCzQVbww1prFaEpY3UYVGO9WvqsLKlItXcmZIMSTaut403tUFukl8713l/f7SwVoGzKnDKfqAPcdBF2owdaLXxJphB/vpTGH4IZnJi44BHz01KSb93izITfpGCtz/7YgJildAuowmlBGVak1q04Qi9G5PSmVmJxZ3Z6lJx6vWPjG48wVuErdgmpikfcdjpe8MBRlGIVvj1Rzh6ccKa9MMynI1UCDSGbt6UxS6OfpE0f9PrgygCCnd0LOcTOXccM/v9gWm+tv1pJy9n3etOQgyBbZ1SCqU2zuvKuq4WcqGHQA6vQbfQKs3B1W7m5urAsixMKXB9uLaoqWJ24JIu4/VnY7dKs8SXVjK0svHMfXR89HzPXU10p5XSujRevfoJv/+D3+Pdu7d88YXnzdevRXf2Yu0N5ZkfqjB7L7wlzFy0cxdQUh+430At57scgBgovAVdGWg5gKXHx0f+4PM/5Nvf+i7XVzdM094cajK1ZIJhHRWPix5DNLYWdezehzGLAi9twxKVja3PSS2Zj4Haq0lZTQdQh06i4X3j1Zv3fPbFV0y/9G36iOYxMCzG8TON8Urjj/ABt20rQLN4aV2mml7gWzAKLairb61vHnsNZ76N1pK3bnO4Neq9MQfH1WGPD4nH85EYPfRIXk7kUmi147tMXbuJHUb7bctJNjswJAPvFBqD5Tniy7S/TzFtDkfBC6ylixUYgjYMSr7q0CsltD92qG2TxOVn7zZeiSg2uCbf/PogikBrjffv3nC2KLL1vKOVjA8ywxh6fxcizoDAsq6a9UPcikCtFe8du0mmIrvJM8fA6XzGNSXJpOmAaJeKDS+107C4r6a4rNIr51yMYHRJ5a21ktcTU0p05IzcWyMG2Ft+Yqud03Lid377t/jqJ1+wroVaCqfj0XAKc9atxpO3NRp24AbaH7wj2+02LKwGXtBNLdj9pNufYTsmgU/wcqFpteEMVPviyy/58idfcHN9wzwfSHOUk1ERwFlrBh8JfeDLG/zIkDjTzRjDipbrffNClALSbzP6xmP3Q6Rkh8Ru2vvjmTd39/xy+xTv5B3RnAw8vdczMeLBxmgyRQPjbDyprRjY6FldJ3j9MyXHZFmSna6W2Iv51+yAhRC20SF6jzPx8xw9+3nHbn8wq7XG6XTEeU/aX7PmM62s2zrOOdHMlX05Ng/DD8Iq1xijmii+vXVz/jH9f6+QJsm8+1PFo2c37UmhkHxmLYUGTC0+WQeOf4SteFPeeu+IBGqISjcu/2SpxP+/v1qrhOB4+eIjjqczpWYez5lO4bDfMc0zOLVYfsyb2zKxsZsjx1M3qyVTsfWGd5EQJ66vEnuzKmsIL2gNHo9HorETq6nPeqk0d4mfTtF23FVrvcOcmKcdedx+IfD8+TPuw7UOdyvcvXvDD3/4uxyP99C9eczpYWtmQjnWbQMNDiGR3WrkFN0n3gUq4oqnEOnj9sHTbR4cu3NdO7oNOm4rLs6YZI/nhR/8+Md88sm32O1vSdNOYpy8EkvBBY1BYkvbSe4CmAb/PJsseqxKFQrqTLjk6MHhvG65bcvh1XJvCVJo7DuvhS9fv+fu8ahUYCe3Z++7vU8O+fdLztvt/RihKOAoVSu2YIevNCVZLaUwR3Np7tZp2sgJjnVtlPbEX9A6jtY8S3W43Gj9xJQ8sWeiKyxZBrH7eSLsd/Qmxt5Iao4+mHGKu3wWtuLFOeP5Q2+KmotPrubWJSRrlq0gboMZpMZpC8cNoRNbZwpNKVn07ee7ZC2a/Nr+bO9hTpH8ofMEOlgufcSFiRgmvLVIOTfuH04bmOW97KbmSbn1y7IqRXd44tlN1TYwMAOS9K7rSnee3W6HcwrB0MpGgMqI18ZuYuwIeDrB6XAe9juRW0rBGVsrTTPzeeJE53xa+OyzH/Pm9VcWEjnAsWKThQ4BdK2B0HgzyCPDOQj69gD1Pto9ww68IendaLq2K1cIqvkEOmnuWxd5N/jA/f0DX716xbPblxwO1zg8OXtSXiEkJTq5sSHgUmadM+t8HUDvHYvJimXW6ky4ArU3RhJ0iJE0z+xaMRm2Zu7eNMze3Z84Hhf63nAD8zJMwRO9PofBB9nAUMxs07ZI0ZR3pdn4YWcw105t2daCRuW11jxGjwtRI419Bn4rT/OOyQAAIABJREFULmHz8W+14XvBdZGTTmfJtl9cH9jvDvRZeFU2lmdvZRtTtoxAYxs6Hy2/MG3dEFxclUstNlY1ydB7UbfAkRCSnjmTbYegqLv4RHE6nlX6WHnq49K2w2/qzG96fRBFoLVuBJHGRy9eMO12lCLnoLxmOdAYIWSKgeACvaqD8CEw7/aEMHH3/j05K4QiTYkpjtw7ORjnmsFFnF/lXRCSZsbucCEy7aYNHXfO2v8sz/0pTRwOs3bWpj5M8w7orGunNM16Dw/v+cEPfo/793cWVabWUyi+mUVyCa7Eed0ORmEd+/mGiZuaADgAt31vpo5z2hS0onBSoc2A89o+dJmzpKCQjyVnfvLqK77zre9wffOCRKCHSeuyWEjYrRknnIdaTI67MXUuYhY35v0qhyHAfB4cOKMO0/ExMe8P7K9vGclL0cnFuRSZdl7td/r9dM5L5rxkpiSga3w2dDPybN1wAh3cYU0WRoEan6etGDrGELXKNui+gwvhcLguHkkz/KlxYd8lW7c5Z16HXZsN5844L1A5hMiSM2XYrVVtGmKM261ueW6WK2kFvo/kLKtcbki5RQc+rXI4TrEoTyIkBm19MAQHs1YXjcaNZlZl3ksIF6yz/WmvD6IIjJaxFJEzUhMNMwUHKVCKY1lFVpmuJnbTXvM/zR52C6acJsBayGUlutm83gOOxNQTLsiTHZNyslVlU7C5MQPbbdyqDpKFUrQREdZ1SDqwFK9Wr1be3b3j69c/sZgv6RHGB9u6hCwqcItAIF8puVzALku5jSHgSfaANAPaRgG4AGPyN9RunoEnDB4CjeCjVpc4Si18/e4Nr9685vlH31KYZlkoJdH63kxVxMcfIwXO4brdoD6w1gWxKs35uGv8CFRk/+C0Q7dxeMS23zz/BBcmWr2jBUuNao05JW72O/2ZtZK8s5RhHbqRp9txMoBphozbTQeOYSVXauG0lCdbFI1DCgrxJlW2XX3X18SQrHBchFtjxFo7FDN2UXKQ9A7BO87LicbKbp4396lqN/RF2yEyWK3mmD1cnPr4/p11ofWPfE8gXGc3z5Qq96xcG40qJq1zeKdNRy4CNreNhNmN+Se6hrG1+mmvD6IIdHRLTcFzHom91uc0EDpr9FH94FW7fac3aF1XcNKGO+C8LOR1pSY5rXbvCUlMwMsOu4MfngNu+7BazXRTqqlSK424dVjOC851YkpMLrC2RWu6rjZW6b6e6MUa9HabFPOKK7VcFGLdUoY2pFh69MEk7N2bTsDWVV1dxEavdeAJjEhyzbb+sj4ztHh0G0KYG4/HE2/evrEOp8HI67MHB+eNxOLoTYzHQRGuRhoqa6ab3sE6e2obHnfB3JkZ8D4dz9XNCz79zi9zerjDe40FpckTzxt/oyMBWEqTZLWub4e9d9uRN1me18FrYOQU6O9OyW2ruBSVozDWeFsnA/aeDeOavrFDRwewqpLQUcaFDwkXk302WuV6L0uwEAVYe/X/m5R3+C6Wssq4xg3DWYuM84HQFR2WQjTfRIhMyEdBDMrzmllyYS2FaCQwuqE/fVxcTwa48d4zmJhtYx9+0+tPE0j63wD/MvBV7/0v2q/9R8C/A7yyL/sPe+//yP7bfwD8NeSE9O/13v+nn1sErNLJK84eRCcNeG2Fw37mepI6bF1XWre0GTP8CDGyLiun8wKI/pmsBV+KHmjnJajwISjBvXczKFFrWMtCzkW3shUckS682ZB1ehVBo3VHSiZh7YW8nsk5QVT3MG4mFyMlF0pezWtu3VZDgyqqgqQC4hAhSVRSbDNR9ACPrzMWWBi79W54iElhuxFOOmysQRl6aH+eS+Hrt+9YlxP9+rm2LjWLPem9fU3F9WKjxmqU32wr2JX1LN+9WuSP2L0oNc05vJPEN0QJwfJaKDh88Hz67e/z+vMfcj7dIT9DM/l0MvKslrgcN2lsN28BcS8C4Hyk9rBhA2qLqx2svs3BDrY1oTfyD6P9N0IS3m9jS+ydFG0r0Qpxsxh/yuaU8rLWjOujw8JWcGKt+hC3m7118F38hLyet2ezNit6fd1m9ilKiBSTqMndPu/WG1NaybYBc64bLiDOgLcuYBDfchXxyo1a8IQg9WcuAsB/C/wXwH/3x379P++9/ydPf8E59+eBfx34C8B3gf/ZOffP9oub5Te+vHO8vH1GK6t0AK4RogCzlBLem8TWZuPeK2utW0RUKVK4bbeIE9U110rO8sqPlj2QLLx04CiehsI9M75JIpziZMYauh0wwlJ3AczxNp+UX3883nM8N9Yl4mImL4rGlp6hbW3ziL5yONZ8ptZCcjJSGW5BzaylMKcf+dUFuunXB2A3tgGtVwvVcDYzq8sYfohypjEXpdXCQ53neDrx7u6e2+efkEKS7XmtlNrpa6EjMEx6e1vFrWdDsIs4BVle/ARPodHrWG0qD3DN6tgGRTfb+3t7+5Lz8U5dUZfIxTuIztmOHMBtxXLD6yxizDOEQbYu7ePvHcDXKIEYXrA9nQwNhutIoYiahtFN0U2D36thDB1oT9SlSHAWPLuwo+MpvdOWZftZdvO0HUDfG71pfAtR3oh6HvrmGDyQvdqNINaRtNr4B3RHw9soGShN9HQpL8WCHZbmIUSiU+Ea9u31ZzAFx+tPk0r8vzjnvv9z/yS9/hXg7/feF+CHzrnfBf4F4H/9Wb/Je880TZzyymlZaOczU0om1HCyC1fJA6T5r61vKjjnmnnNsd3ywheM5lnNCuu8sriVWlccTr/fydFlMMtCjHIWktsjJRfWVSEe1dhboTuWLGfhXk5c7W6IRYaSb96+5t3dO1azqGpd/ghi/6lwLOuZMKp9N/FNM2FLq4RQ1R6arVkgao9vM2OthcqFRamup2/OvcEHFQFLsGmliPFnCPp5WXn7/j3frZVQFWNWKxwfH6n1SIiB6MM2I9dWWc8LeTmTc1HXdTwCorfGEOWE4zwpyfOvtxHlnqnLSj4/EssDn754wd2bLzgeFeMmgVHbRrKx5VjXlTy89J0INQAtL0C2wuyM0jti6S8HRy0/bE7UtnURrdlqiq3TrN1QwasC72orpswzxqSF1ZzO6jDmlPQ92OfQQf5/rRhbVSEmDivIIRngWOnUbZPVzYil907F0ftZ6tco5WfpneOiUWBsVpwz923bIgWz22vFsKoqrgvwU8NMnr7+STCBv+mc+zeA/wP493vvb4HvAf/bk6/5Q/u1P/Fyzv114K8DfO/bn8pwsyk+PHrHFBUSUZr29KL/zgYGAb2RV8tyc45cKx5vnPmx51aLnWK0FlAPhHzYnTkEFw5X1+z3exyO03lhWTMphu2WXS1y7IJH6Fm7Pszs0wH8zGfHzv3xni+++Exy4VYhiE2nD9vApvVErdWKQBeN2NZDS17orehQTUKYex+acu3EPY6C2nDnR/ClAXAGbvoQ8K5Rq3UTbZhXXOLU7h/kfpSmGdcbj48PvPn6axJwuH7GfLiFINtqUFufz3LYrUum1wxO66yAtBQxBrpPG7svqWaTXce1lbvXn3F9SLx8fstyfkuthfvHew57+9ksQbk2S1geIxlKG9Kvl61ge4t8dzgZnRiz0eE3otCTpk+Xrq3LhxnpMDkF4/XrD6S2zroW1jqMRDChmoDho1sYxi77OREDtF5Yzo3dvONgxqTeuAg+RHpPFOugLsKmJuAyykVLVmMwT9pY1NrIeeW8ZqqBfq5L1zJP0yUc1rHN/iqu9rW0DTP5aa8/axH4L4G/jd7fvw38p8C//f/lD+i9/13g7wL883/hz/UUPKuh6C44rVOEbZpBpCyhainmKdc2brmMExrTTsSKIXihY6wvVccwRx0+WynNweODyEjBCz0/Pj5QW+X25hqH1F9TEDrtC5Qu3CB6ZH4aAmupnI8n3j1mvn71JWteJNqQnk5BKVmztR52PWwyFFlU8W3HLJQ7E7In+Nm6H5tH3QW46l0ferSOojZ1QyN403Z8aogdFoqq3ytJtNSAtTTWfOb155/z6vMf8ezqwPd/5Z8jOk8LiWI8h7yuckguFdcrwfXt+6UWXFCBqrWoCKEOwlNwPVNr5u7d1/iauNkfeBUCuzmx3+0IQezNdTmLhNM7U0yWMyCvwhEhN01x+7W8lm1c0HahbXO/37oCthWns3HIeXE/uq38nG0chsnnUETG5GTlZSKNgDNLcoPhWle3mc+k4ERp7pV1WSm7RaYkuytGMEgMnv00qxhUjavndZWJjhOo2F0QvtLFNWhc0oxLqeQ+pNCV1ApTiAYWX4BPuITnOj+6IsdPe/2ZikDv/Sfj351z/xXwP9r/+xnwi0++9Bfs137myznd/KdeyU1+cquh4lpxXObe2kQCykUMrtaRkMV1nFtF5gmR2QemSRU/54JJ0qhde9Q0SxPuEGuvlsqyrJKEenkeEmZCSqQ4mQahkVLkdDrjemGeErV1zscjp/s73r5N3N19rQ7DB/AF79I2t3pbT2peU9Cl99LTxzThQ2RZHum1y24sBK097dbo1tpFs66SGEfzn+TYiYsI5gJ6hqSgktpNGOUcU1J7XVvj/uE9P/rBb/Hbv/1/8tHLT/nok1/k9tmCpxNdYMu3A/UkragIoHMWzb5K2gFk72agXG4FWqGsC6XB6ZSZDnvmaWaedvgwg0tMk2Oa9/Z9dhXWbNuVpkM4x0D0OvCdYBF0th1xY703nqmycRrEXzACk2u4psPlDYfQ+CMQeJi4hNZIyDbOB7eNEcF7iG3zASilULMwmmomr60VHo8FeTdoFPBo/RpnFecaIz5nBjlsBKiEHtXONxGDQPP/fkr4nYr74INELwOXYtsGFey+jQ2lWEq1V8rUT3v9mYqAc+47vfcv7P/9V4H/y/79HwL/vXPuP0PA4D8D/O8//0+UC9DV1RUHVLnXXC3J1m37VG9fJ751Y62jxVFlPq0ZkMf9btY2oeRMTAlRAwRyCZkNgNFqQ+J0OvH+4cR5zexiJ3qlwdRSmOeZoYxPKemwNa0TSy0iFq33fPbjt9y9+5pSFrWqzUMSMj1NM6lrzmutsq4LIBsr7wKtFdb1TO+6qZrN4r7WzR9ugIV6oBHXfxiTGM25tSFbrZtpydCZxzBrFImBedqpfSyZt2/f8tnnP+bN3R23z75lgGhmjolBxnFdoS2yvjRJsK0S8WNDgd3MYwQS1lHXE/n0aCpHK8I2bp3XzLSotY0x4GOgGdf9eF7sgVc3t4bKnC6bg2at7/YUGUBWikBT58YIJBv2KSa0Uqxbdp8aRgWMYK683mGYiEBl3HB00oovWpCo8JxMiUqxbmaEMt6DZckcw6N9/jJ6Cd5TauWcV5acDYgUe3REtQn+kt2d68YoddpkRO+YpvTkItCYuq4r3ZKIvak7QwhyWW7FSEvf/PrTrAj/HvBXgI+dc38I/C3grzjnflNvIb8P/Lv2Ifzfzrl/APw/yCXpb/y8zcB4TSmJB1BV/edJs+DxdOTx4d7AIa1rShHDUOYRbUOM03hTSqGaGANEs8yr1H9TVIVtpeLniRATy3nleFrowO3NLXU9avaMsGbFoTsDrhSEKUJR8I4QE1dXV9weIq+//AOOxwfT+XdcSEYtxkg8tv5jcN4jMUgc1ey2izFeCEC2Rjrs96QEwUM1M9RtFfkENBNRJWySV7zZeDcz8hykmKqgFrWNleP9nZyW0ox3bL4GtIoPjuD65kpUe1V+4ugLWsN17eY91p1gXgStQlcXsJ7Pm/PTIEXt5kCjkmvGF7eZnyqAJphatOJjAhcpZSGfMm5zzx0P9lipPuHPDyyg67kJtt9fcibXylr0fYLYoaVWuouMPL8UAptvUleB2ejHdikBBJeIHop31KL/Pk07ma+UldIay3oWONknpmnWKJMzy7psoGVrYvzVIKJW8M0MTTq+jwLr5DNhSsYRoDNkx+qMCj14epRj9ig66oS/+fWn2Q781W/45f/6Z3z93wH+zs/7c5++Wus8nlY7tH2jXZZSeP/+Hb5rTXc6N3J34BMhTOzMi37yEod4b9FOvRF6xQW/yVhpmRC0Foqu282bCYb6T9ERgSlAmSZaLTaGxG0159DtJu57pIVZyK4L7HcT63qkGSKMIcGuNQM4K95cY1Ka8c4Eu30QOTzeAkSX5axMOq8uYjjKhjgRMSTYlGk6oHYhuyi02QlB3uZpECehVSXxOM/5fGJdV2IU2y2micP+Gu+4rAJbZYqmCRC2jesZx4XMVZG1m2i68uoLTj9Pr8Wwh4VaMhCI045SjsTQuTrMyBFJm5xc6+bF55x4Ha3rcDjblPSmYla75fEN5mTvijfzjjlFUzGauMmA2dN51a4dpStNk3IoHBqPvA+sJZtTkD5/7/W1pCS/gDB0/+p+qm2sphQpWdT2ed7hWqfWxLAuW7LCV9V9aqPiXSabzFcgtHE9wFbGegbgogEQx6BRbKOi78PZgZfHhMdUlA6CE6jennRMf/z1wTAGq7VR9IY3l9/eYb87qFIXtTnJy0GVXkn2QAcHOyMC0So0CXFkgtFNty33ltYdOK0WS6lUm217Bx8nOp4QZ5vVdReU1tVWO6nkomVPtFIMjW3SMKSLcUR3juCi+dl6hkHIIKvo5Yhpxtk4sCwnhndANWvrjqPklRZktx62vHmtsjwiHWnnbVZVflhKPd2ZjxtDOHmthZxXhu/+9f6GvHaOpzOP9/fCJNbFHnpvKHPRLRnDRrIJ5k3YWsH3jiNa59GpZSWvC72qNceLtXhaTsyz5/pqt0mGa3N2O3dG0rDDE+NkqzAxArufjB3ojOFZTAna9SyEIOwgDLqxiGjZeCPOeabkmW0dHHwixaQuE2g9cc6Z83LWCrJaylUMJPMFHISk4AQGytxUHpbDTGSOiXmapRAshbWsLOtCaZXJxHGH/Z5cLFvRe4IL+BE6Yh2jN2NZZRM85TN0K+6YCclFmt67+APBC+Bkk3B/8+uDKAIb7bFddOi96UP34UAtK70LaIneMadJTKhWWddMsSXR1V6Bph1Zdo9qnUsh58apyCrKGSPNdTHbxFgzkMj4Bq03hT7g5Ftn82ewal1tnm69sxroNk1BM7TtfwlJ2QkW1Kn0o2a3opSOoUemaUfOZ7xXWMZwmam213dM23w/otoHzTf4i95/2Gy13jetQatVSDkwcum865zPJx3Q3rm+vuX6cMPxrIf/1ZvX/PK6ME07M2w1wRIWsAFspijO2m+dU0mKjR/RaqGbcYls4T25Vh5PD/zit2ZePr8iRTNL7Rq3Ugzsppl5SmBbgdZHaVRxl/kLZIvmKrVwvZvZTZOpAZ2ZyIqEExmGHVGZha6zmxNTmpU5GCTiWnNhrZVungweyaOrjTrJglQFRFpX4L1xUUbeAuRS8VTSLhJDMn2CJ5eVktXKTykxz/IQWFdP9DLBET9En5eA74FhlI263Z1Wx947M0t1tiK8+E5Uc3vy+Cf70W9+fRBFAMAP2aefBPwhNdZmqIHIOK0u+K6cNt1GgTRJ3Xc2ZlqgM00TKSXmORGC9szndSVXDKXXr8/eE6fBI7islCJuUxh2oOYsK7Mus9MQ4zYbp+C52k08vzY1XBcvXLr6QqkLkUlrNNN+pxDE3y8rrhdpFoznPRRguRZaH0QcHYRaq1aBabJCoNEm+GAfflXBbDCSbcYTYN7B1O44nU48PD5we/sx+/01t7fPeTivNOf5+u4d7999zWG3Z12D2VaZ/VjJ9rk4uykvopjuOz56epPys5UztSz6GZ0jhok1n3Cu8clHL3nx/JZ5nkwMJjxkN00cdnumKdGreB21Xww6W+v0FPXerpYUVPV7D7vdE36ALglHJ9n/X0JlrQI45yCC2ogUq2aXdlpWpT75P4pRjFQs57opvOUT6Zwjn08s61n2dQ8P0AqfvHhGbzIk3c070rTj8fheseh0cq2k7khpplZhM7IM0WfpekdR4xbhXitLMcmxG8+Qlz2dFQJJqR1xSrSmgiu+RuRnVYEPowgYsaehfXRf1E4O/nRwnZubPfs68/79W7786hWexrPbW2La4yh4BKikGAioDXp4f888yVBkv99JkHNeGdYbrRXOi263lCIB2yPbDTTkv90YJMGD96rqHiAmcs54V7k+zLy4OYi81Br4pg4G8cljmCxkxGBF55nnHbU2lvW0tbgxBjvo0VR6Zi5qnopb1FWIpOQNU/AEZ+6zdd1SluhajdGHUlPvqXMy37i/f0/+eOGwv+Hm+ob9/T3de0pZ+eLVV7x88VJWXN7mTjquVwGxTsKqgmZzFWqPMxy41kzOC2U5s65najddSK9EV/nep895fnOtg9gaQ35cauXxdKbY5y5Blq2IvXwZIW76/RojxXVybZyWhXmelSpscuHWLpRt79SJyUYds+yS5VluxspDeot5nghexWa8l03sK+NhmIrUdc6rCh2t0Gvhej9ze31rlHeZwUTn2E2JWovAyWw6AGcs2K4OodRK9PWibKzySlCXIpKR1oOBGgf/INrqt25pRCF4Wwu6J3/PN78+iCKgXXC/gHi9MYIYMCpmXXUTRtfkq+8cV1fXxLhjmrR/TykRLcLs7uFIrp22FErtlkXniLvBaNOuOJcmY5LWmaKnuXGA9H+2pB/fcV0EFL/dCqpfxb7nw2Emesi5EJvDW7S1sIhqPgSIBGO7/bqurHnVvj1Euo+2b3aGCXT9d6AGIea+eYId8LH/naJ9lE7IfwxSWDpDtsFWjJue3nE6PnI6PnK1v+Xq6sDN4Rp8IueVt3cPvHnzmm99miS8cpL/emuzcRqVomfj43vXTQrd6XU1hWJlydrm7KbE6Zy52ie+8/EzUnACfY3go/VeprRC7gFHxbeF6L1EYcbXGLfeWBe7JpflXDLzuqjVnmbz7vN2ozpT9MmOvtRqRqr6qBU6K6n3nBL7eWJK1ik4Ty3FVoADaVDXlkuld8cUE92MOw6zHJ1PqxiwjovhSTJ36+MCDw8PnM+rvqI2mcBUUdh3k2TkI7jmvGaGy7Nznuy7zFtzYUqJKUam4CEYKGpfNycL1v2nTRb6p/1yTuy7tVSCh+iTrW1U+c7LyrKc+H+pe5df2dIsP+i3vtfeEXHOfWTefFRmVVd2UdUPF27aRrIHniAxAgaeWUwsjJA8gQESAyz+Ao+QPEJqiQGWkAAJJBgwQUgMGMDAFraRLNPIcjfdVLurMvM+zonY+3usxeC3vn3SUNndohrpElIqM8+590SciL3Xt9Zv/R6X04LzuvAkSpnkEmQSgLphWINYxVBgWRYs6wkAP/juiC8jmaIjqZNXLpR7jrn4ctCHOlefI+XYy5pOUwcnY8SMdngS0mJsTcUvMKFwxCIkRHRX7Jnn37UpyXWnGEZM0/+PYSzAlAa3ViGWIbmAkzjHpJwyTyY3QVEDTBKFNpNQo24FDkN29H3bNzw+PuDZPXXxHzx/hlR2CrL2DV+/fofnz56j5OJg44DowLCBiabOOHOFQAftwejYQ5XhvlV0FcScIaKo7YZXH604X1Zs+0YCjlvNw0k/Aj1Ud6INDR2tMWQ0pnyAlUPJjY8BUF+XAnDtAU9uSsT7MRoNpQaA2oThoZ7B9/LjGAG2bcO+bcQGUoJ4LiNlk7xmaFTC1eFQgUjCuvCQ+IOffY29D5QS3f6LgG5JAael4D4yVfjh8Yrrvh+Essn6q334+pj3RE5Pm4xZCKevQx8cI8TIZowR7jbN7RIZnO/5dgCAi3aiS2px0FtFDHd3F/cYoElHWU6+DopUvo0OhIghiYCeDp5ekxYaATHFEEFrbK1IOGLL35QhjrOVNfCmX5eCpZwgh7edMZkIVJXtdUcMpILyA45HsQBwrHtUB4LSzFNNucHw+T8GAVJC7ZVWYBIPujHCNB8NrkJTzPZjmCHYNF+lwOV6u6J3CoKmocSRJwDujF2PiAjB1hr2nYGoJResJcJwQu2CmwJvHyvevnmD07qiLGeXNNNRWJ3HHxyR50HHkURHRa07W9/OMJAcBLXdcNve4Xz6AIaAvTai7oW07fGNcQnG1lxC9Nn3yVbcIBTlQJGjrystMaXHaBZq0UNZOhmHDPvk6HeYc+pUZCokJef5c1Nz2xg0M2POghuEMFrtSadA9H/43+P4Mvx1A0y7Gr5RAuhKdKsNJfneKEaudR39p01G9BBcAo1DyRFZS8apkGswzWmHW6GZb2TmgcGK6IxJd1j6tsd7UgQEtenBhpuo87qu0DFQMkE2dWeYiIRcSACqe/U8wXz42yO4gSdBBeQUyJbTAaAiOJ2yO7tOnWFFGiZcugmEyLQbgD9rtIbenf5aG7PyhKvMJEDKGWvJePe4cUQQ2qFjTIvop05DwLgoBIPJIGtwDEgwPBlECGaOwZH3N8FLwFdGkYaXvqacgton81E3oQxP+nqbTkRquN5uaPuGAJKClqgQjZBlRW0dX75+xOn8BhdVIvY+SKn7H8CUxhYQjF79Brqh7hWtDtTODidax5t3P8Veb1iWhJBovLLkjCUn5Bj8ZhHU0RGGKyAlYmBKjoNvTQSjkaIbBMwQyBG3PXA+rx2LGVL0wFbHUmrlrj5FsgdzAoYy0DYlMgsPPog9ORiZb1tkEpAQUHs7VrgDlAGXlJiQpcoDIQQ3TomIge9caw21D7x7eASDVbmtyhEHUW7e2NEBbIN4QAsdtp+FhNNC74EYZsiOv07fZHGL40YxeHKk+nmP96II8LR0A80QKMwYXP+ZKVI+I8SMoUBt1PhPU5BSEnJZuE4anfv+mCAxUNppU1SB+V6xxTa3qlJFyQmX0xklUlZsLpwZSm2Cu/k7+SQDQXGOEa3TB+5h2wHteNwqo9Ej6coxCVpTQMgfpyqM8V4UjhK86mOmCw2okOACB9lEgBkwmmLiSeSz5/x6jJmx3y4kUedF6KAzDedIgSkJJZyT2WG8uz5g228oZYEpkNDp5SAJMUWUElFrR8qbdwHj6MqOUIvoBqlDAeX1POVHAAAgAElEQVTGoNWK2lgoSlCgX/Hm7ZcAmP1Q3EDDQkSjPRNXY6J0ivbbbQzDbkCDIQWP5w4T2wB990IkU9LEI7uoukuxo+SFRh+Ah9QQHyq5PBFtDtB0blP47N31/STucMNAazl2ZNFvcndVw7oknEvhCFsr+t6IaZlgDOdvBJeF+1HQhyFKRIoUEXU1GtGMAYvDWakJOdAhe28Dbx4fsbfqBbS4xwG70JAiBOkoCBMc/iNYw+9HEVAd2DwgdMpjTSmhBBTXWyAV0yIsBEhkTBcz5BljRSTdW1IVxElQ8UquU1ChtMQy3z9DuJudsWAkgAUGewxeOK2RzJEiK/oaEgYirhujtnfz5F73GUgxYPcLewpiNDfklFEb21woFV6td4aRAIcCL6V8UFSDRScFJcBdcs3nxmE8IWduYZCImBNaqxhj7u/Vde/iICvn2uSFYKs7tu0B95d7f78MCRtCWLB4CIuOwdiryW6cHIlBIc0Yc71qPjPzZoF2lEC69t4f0PqGlCPO6wpF8M1MQ4rimgCObXD2W4THhteBZhPEI0ibAskwAYBrFiFQJCFQqU6dtjgcCPaiYoLagRg6uyYfGflwfwKRg2mpNp5abmPRSIldp6rCRmeoqgTcNupNeueq8bQWGNyQBASEk3M5YiTwrT6KJi9syQwWg5u5KGptKEJGYyzOiBVuPdrgKzaN0Egh1Bxp6G6V/Xmf8JKf93gvikAfiq/eXXFeMtaS4KlySLkgzjbNEXam3wb0TvBmJvOa0YiDck6mDU27LFI16dseM7cDW93ciptzWWtEfhuAMTai+croMtWBnBIupxNnOQF65QqM0WmFz+F77IlliG8XoNNl1lt79VNaKeeF6dNNOgyjy9Hm9VYxXWLCJLD46mvSgofvg2c2IFNyBiallsQpYfETkp7ExUeqwG27odcr137O9+es7aesKjAyeuNJO9yX0Css9tqOE7UPxV65djOl4MiMUVvDFJec8fLFHU7rCdeNKbsSEznxMCwxARbc9DRgMQGw47o3P5kFIgMagtOKDfuo3JungqVkB/kaKcTuM6imiHLG3klzVu9mevcOQKafgMDArsrGoEMUeJpDBCW7h+JguvNpXZA9ILX2gdo7Hm+d74EpUnLTkcn/D0LUzj+v4UKmoS6DDr7Xl3B4P0IVAYpTSQcfJESXS4MzSjd1frgeB4Pa7F4XSPj2W/29KAICHC4sMDrlMk9PEJC4QmkMAOmteULwQG8bZkhICoJ8SQfASEceItd0ryHvXAQOwC2YuXTDq/EYng032rGj775BiL7jUUSMXrFVGoCUGHE+Ldhrw3nNpOU6jXf0xgupA6k39LazxTeepK01zm1u9gFEtL5jDCL+rY9DbjwjxeBFJCIg5eSUY2BYJ4UXU0tvALjyAnjxPeEffH9nOMjjbcN2e8T95TlSWtD7DLYcmHLZdwYyCP21IwgRc6MYS/w52Ap34hQwtsII2DtP07tzwbP7C+7OJ+Sc0Bo7gT5YDNWcP29slYc9YSIw1/SncHQ/8PldhI7QKVHefbs9YowG80KfJCJHw7koVHlDkpdA3IYliHTc3qdhCNV7poLpFG0IjsY/Kf3aXrn6S5E3nGSUTA+JWpX24wFgb8J/DjK3s1UPIljg/x+uVzOcFAR2l1yO9CTY9A8cmLoDgLqKYA4CuzDpvR8HYgh4ebfitJ6wV4WaOMnCcNs2Z0xNQQUwc9khwly8oRgBwOMDBIbzenH6sJNYwuTuy8G+6i4TDq7Ciimj5Jk5yAKx1wq06kSecHjG18ZiBFB49OyyYMvkrBumqaXvdAPn3NYqYkh0RA6RuXxTL+58CHV7MO6xKU3l95mlxxmcwhbuuw3RUXAFU4xotkKPAqhzGGCQSHrr5GD4hg9mwHW74d3tHe7WMzTkQ46sLmJx5j7goqcxFBFUPwY4IKUKDLitWcCuzTkVAbUz+yGFiMu6kNsPQyoJmvj3m3CGnclF8NGjdXYUp8LwjhSEVvQH+g0gRqSYkWM6UPnssV6c7uKBnAfzMBcFqk32nRwblBCmHyW7gQwWT0ba800L7lcSHLgdw12Zd9q9mRpOa8Fazhz3BgtiikBKlMmbZzWmKIgSQY/VqRzldVNywGmhspM8mAUxcT087cXJ1Xa/RB96+AO4ujRwzHzvLcdjDHh2ORE8S8D1Vg8qJddqnKOZQeesC1MEM+RpJ20DrQ9crxuCJBTwYhKbZp+dvoOD2YBDqd5al4UAy7rydPDWurmtFm8Il+T6U6cUsZY7lIXATC4FS1e8+vAlzuuKh0fqAARAdKMJ89VlDOuxCg2B1mViQHUnYjWF6JNFNOm2CSYG1YZh4TCvEBUY3IBVIkYj52Dy93ij82SB8HRP8UmDHiIxhT4Ur98+4IPzHUpy6rP4SawDkIRFEoBwUID5GUyhF881hsgQnE0pMo1IDaYkvIQYcH9ZsRT6CLJl7c41oA1cjL6OMxqrFjWsFNYDyvFqdm4EOaPP8IbH2+MhJya3nxqR3gcFUSF41BlxgBnfNbvBOoG04HHog7LiFAIsiRODAHxDx98HO0uIOAmMuMPbR5KcliXDeiNxzYReEyIe5OrXmBN7ZnZAbw0xM/8hOocmpeyd7QDc1UHVr08AMEMM6qOHARiHh4AGO9bWP+/xXhQBtvQRMWW0vdMxeFC7HkUA+IgwxB1sm89FXIkxlonGi8OA2gbauMFMmT9nQOuG2pk8PLX0swUP0pFzpyMODKOre8U3jhUu1GEh2ZFixGk9YV0X947jRff8/oLLiTqGfduREkM09zpcwEJzCiriApawABAMJ/gQ6ac9Wnc9QAxPYI84yUnVDvdk59oBINvuYJWF6BJohYA+9cc4Aeb6wQATcg7ebTvePj7iw2dU1XUPA2W34cVxNJ6CED8RKXOeGwODZxQKDgFOVcUYNJBNUfDhyzus63r8/NapGaH6j9RontzJ992KIM7fCONY54qoR5/Rd2IoAVy1qUQkmKYh4XrbCNgmCotiiIetuc53z1dz/qZ4ziDNUhtI0FnzbLkFtXc08/wDF54tZQEQ0FpDM8W273jYdhY5Z5iKEkhMOaHNNeMkSTnBZ4bKjqF4+/iI8+jsCMwALEh58dFvgqAG82KqTmBSD82JQtxnyLfberwXRWCu6tre6CYjntPu3JThZqMpJeRTwbbvCIEoOqOo/CYY1KX3XhGFSq0AcY04V4u1dSAI1rJirx2vH7lOutUNp3UhiDSMzD4biBJc3eW8eANC5E683jokJEgkCntZM57fnZ1pSDefyYCj3XRDbQzKzDECpmghwBwEQ8xkwpkrxrw49t6REpAlMclGHFByMGm0BokRcNYdfVw8RelgOlI4RdNMPdyIbAKLBnz5+ID700L9udHwMueIGAyCCigO1VwMBWUpqNroKjwaugeRjj4gMToOsePNw1sM7bi7O+P73/0YL57dQ/zzCpnmsbRKJ+HGtP0z1wbEtQpIADqRGVfbmbkjc0w4nVai5q2idpqHjAHsleKvkhNKimjoB1quBnY8PiamaWFmcB8HwVoSGag5k+4Nbmb6UNTesc0OAeKMz+jMV39umd3GgA1AanfXqgo3tX5ihgZBSK4YNW6PHh1AVVWcfDMUU0HOBCT32tAUaJ2U6eLUYRYJQuPvvZ8AAHz1+mv87Ot3XBVGVtb7uxPWnJFSwWlZnANOgEqCbw2mnt/BpCVnrMt6OArPiz5KQAtMaD2dTljXFT97/Yjt7SO+frzi1hIurSKH5DZPgesc4YmRckIOCaZTjuwGpyq0etaBJQEv71eUFLE5NZepMwGtEVFPMUE8GKX3yhYvMnKqteB2ZeSiz3TfeULN03aGfR5MNnFGu6oDnwbr9SBeqdGoNGCSlZzVKHBAlIj/V+8MawC+84GSKgy4rNaQQkawhhQKefI6kKSj6Y69XlkYR0PKEUBH70AbFa8f3uHd9QEGwcu7gs8+/gClUDmYSIokUUoHw2CN/APx35eno/P3dX6PrffwzicYI+pTyqBtnGBrHdtePU+RnJIYAtbVOSXGlXFv01HY0AdQezv4BzEE3C0Zd6cFOWYvPI4xKX/GyQx3amj6ZPddcoYio+SBy4meCWrsRLfaqJfQzuKGJ3r6+IZJSq0NpRSUkn2cCz5+DKTRD11EcIt3BIKJObO4T3Un8UP70/cY/NN+qBm+fvMW18dHxAAs7gdoOoCwODMuIOXF51Tqt6GdNO7Ay5sgoPvlc6kGMYaJrKvvftWjn3tDxMCSDTCu2q5bQ5DqgozigZJw40uFhpleK8CgYsu8PTQY7i5nfP7Jc/z93w7AAwHOnAOWUqBaHcWFr+Zsft5uw+WMP9fpiwjEAz1M7PCiA4BmhmS0EsvJiUJ9OuE8+QtwbHf/ushNSPQ9fvBfQ+HxYgBqA37y9VusOeLD+wvJSQ6GlZSQgsHQob2j9R290oqs1RtCTBi9UXfg2vzXjw/48uEGBQ1TP/voDs8uBB6n94HZgIIbgMmpIPGoHah2wYKQeKLFlJ0YNMk9ngfhjsyqA3vrDrSxkymJhWFdCi7rCgNp1t3HgsU4Rt62jsrlO05LxmXJOC8JS06YjMhhipQSRDiipkB3q1V4KOwOIBKroFCMRCM4Nbt5PoGi7ZuveuVwiuI2I5E23DvO5xNOJ2pgRm883d31KUlCCsmpbE5scIwBEIg9nf5zc/DzHu9FERAAzy9nvHrxzNeFA02BIRF9AE3Zzt2ljJIzo8gc4VdECAJUO3rdcAUgQsPG82nFmtNhkQUTmA3U9kTLPC2F6HQfPlNNYgidZkUC4iQDdYJc3KO7CUTi3n3iBL/02Ud4frfip1/Nk1ncMmtKQxtUI4IQRb/VDa1Vxzl4sZp1+hgYQVMn+5K2KpSRUjhSnZFGz/qU6UMf/m9CJ4GhNbbYIQRPqVUIpk8BUeU2FG0M/O6Xr7G1hld3K0oAtGTANqTICK3WBm77jhDZidU+tyYVdQzcWseb64aHSvvwFBKeXwr+zA8+xnkRWL8hljNCjhjDDumz2ThAxBQYPx9Dcl9C5zoYCTy9U47LNW+lH6A4t2IMxBBxudyhpIDio0kSOElK0F3zcFrLgXN0A1qjvmD1KDu+Ll9DxojoeoMgTMSm/ThxmAA6XPE4skM4dIwWZhhDXDMSEc4X7HXnurm1A3ua4G2QgPO64O7yDIBg266eQOS8E82AdLewC2S5OqA+U5QS+cr49hLwnhSBGCNeffgKEMG+7bjuO4UjIYLkByrihipKCUinE7obLAz1+YywvXcKOBJyhwFNwZleHDBTWlrX1kiscfqtGQAxRKUSULC4IIP2WZOTnYPA8ITSsx0jnfezT1/h+995id/7yU9x2ytMA9YcUVNAbd1ZkEb6scnR8s6QiOA59nMWjt55lFCOteAkC03X3Zm2G6FQi4iSeGMogbJJMJprToWnIlnzU+MptgowXJviJ2+veNgb7jJwWcjtL4lZdwBn3aF87uve0HTgtjfsvhKjeQulzjEGfO/TO3z2yUtmMNSNpB8X48zlVhCBxIAUC81lYgEwmXt2+ALM/MHpvz+p4RKmucgJIS0+VzthCbSR31tD7R17J5kqxfXYvS8SoIX4Dccrc/GNb1lSgYLGKha8yLr9Fx9PeQffZLDqoUN4eu0wQ3DnampZOlo3DDFf4z6NfLCBnBdoWSG9uuKVTNPgATdPndEkkk22ox0A6rc93osiADM8bs3ZWYCkBVnCUdmJpoQjVaUs1JUX4xtAD7kEs3Lk35XESk4lnxwW03ut2Orm4Zo7fQaHa/19AaykdrHq3ritoLebOmXTkELkhxQTEOj+o2p4fn+Pf+HX/jn8g//td5kao4qS2TLulShua+b4QMGSE/Zu0MEuIMbE0w10+k2JO21+zfFg49gTIxF0NUXK/G8SRLrvtNkTH7vjeaL5uqj7OlSPUcM9HMAk5a+vFa+1U9sOmnwMY7ho7wTSamvYXOMxRxAd3jIjIAnw4fMFv/Gr38OHL+8PdaegAWoUCE0TFEmHH6NIAgLB0zkaWG8IkdbgKTEinPhR8tQlLkYlRL5/Isdnx+i2jq7dPz8AiDTtHOaxXcP1GQQADWQ6koEXEcUgDq4ealDgUPQd5B+w8AUPvvG3xUdHdh1bq3jcmFnZWwMtz4hrBCiutxtutwf0Ez+bVakBSDEdOE9tFcExJHWQl7bmUyw2CUfhST/zcx7vRRFQM7x7fITEiCVzbaaDs+J0hoGI6wkazoM57TkGhCSQYZDeUbuTYUwxlz8COsDsvR95cpSVAqdlxfJsId1WqTIbY/jPoQ8ByTDqbXRCbTdYUAgChgoQaHaqg+mzkIDvf+87+JUvvoN3jze0xovxtETcakCrHUqYncCd8ENNfkPjoBbzhh9RMVNsp6y1tuqMN9JaS6EYZvQKCQndqJE3A+cH//wPFF7kKADBBRaTkz6Gk1gK28gYi++kFerP2TyWjUWgu5rSjo0On1gRBXh2XvGbv/4Ffu2Hv8QkKXmK3DKZuZJOkBHyPxg20oDKdvyyJtxdLlAf2WplZN0QmnRMg03ab+MYG6A7TU8n+JZXSBhIsXmXJB5hp1D/nDTak/5eHUb109kG39+cEjtHgHO3cb33JFSjYpSpxl4cwFm+N7oZb23gtlWOoc5JWUrCutItuSwn9FDRdODt9Yp3tyu3Sj4SR3G/xSmFhhzYQJDZkczu9/8HRQAATqfVKbqT809wiCGgbAU569KGGiEjJMHM+DM3mAgCNON6je0zYEKX2m6DFFwoShacTwtO6wID0Lti8bCTecLNtQwdbTPWteD+bkUSzl7NeeWj8cbgzptrux998Tl+9/f/ED99/Q4GYF0y1q1iNA+9dK73dOpJxWOmHODLKXohasfNHCeaBzLBiGI/Md6iqww5Jw9MP0M4WSl62OcEWsXAOdfJVuogm7VKJqUz8aYOv/aKp9GBnQQ3MyTdDjXnM9Au/HIu+PUffoa/8Od/Hff3dzCTJxed4MlJcONWH4fM9/wsSFTb9W6o++ZmM9NG3Ln27q40T7ypYZB5MoqvSrUjgCEkZgWwzNnazJOZ6DlIzgH8pOfmYgbfMOWIZJ0UA+vrrBVwEw9Mnwq3p/NiPN/b7iQkip0UKgrFwOiKx65o3dCW5AEjhfbtwg72cdCW/+5EgDNHt2S37kDvE52aG6EJeiueRFL/z8d7UQTE0dToAZOUeWYHoVx1582wuBtwrZy3g98YKWcElWNuvm03XP3Uu1zu3A+fxUSd3PLukdUY0SOiIqPNJRmAGxmGxmzAbR8QGUjnMyxlGHw7YARpbvuGx9sVfTDK+vPPPsGPf+X7+Dv/62/junWoAvenFb13+uy3hqoUFbHtp2FkyhmTPAJ8w7QC4je3u/26qCVOirCvyUIIMCdY2JhW0yQHzZnfZwBMV1vz5+mHMEiQcsKS6KeoSmv1vTX/01y9zsCUodR98OJjAVhywucf3+Mv/MYv49kd7eBKWf1CfTLzmFRXM2AmDPM1KlWbkXZrX379NcYAWYUpYkaRezk6bkB2O07ACdGltYxOozGqu/gEckzIUqSPRIrZNQ7d/4z3kyqAFzozF+h8A3CbjlM4CoCHtBhgoOhp+DW35ITzaUFrDbcAlN7RkmBYQBtw74tw5A+GKbUWhVlD7x0PVyfRLTywtt4ABwdj9OgXoR16Cumgxn/b470oAqqG2745Gs45Zsl0C46hO2JvTv6YzjCCvaqLNvhr9NbQu1tsR57WMEPbrgSdZssmgiFu2TUoKU0huTzUzR275xe6Kq+1hnePD3jz7oFbh7JiLYvP3JzLB4DuVtAvnl/wmz/+Ab76+g3+0e/8AVqngcTdqZDb0Gn2WduGMMgySw4ukuI7/L3hXN/VZ2if8WOg9JnW0kCOK9WKMMSUICERvJpe9d7SzyQdUzczCd+YaW3iCQx+lcI6cbiJwY6sApttcJirKM7TKZLbnxLwyauX+KXvfoKXz+4QY+ZIB4/6nhHg5voAYzvLck+3oBwjWm+41R1/+NMv8Tu//zN8/ulH+OzTV4cbcXd9f/ZMBvpRTJediKRO2LLhbMQA+kV2P8E71C3J4eSzINNpmuMPPHwVPnMHMWDgWMEZ1Nvt6T/ALZP5tT0NQ4eTuDIKemcXogqCummBRbeigxygZwnckCkSTNSLCsNspmJ02+tBoyZ5jkUrinMHIjch3/Z4L4qAmWHbd0AEWfmSxjBIaD5fzdkMh/tLCoK8kFNt0+VGAzTF40bqfsGyb4x+cwT0RtcfikV4+oh3CeZIffZY8ElEmm01/QB2lFRgNriuVKYUn3JCDp4LHwK+8/GH+HM//iF++vVb/PTrR9TODL51YTINnMlGMhBbQ3QXsPhce0SpiQARiIguEU1PcuPO4sdQFIpYcopACL4yopjnm4lw5h0Bo8cd4IIc4BIA7rg7Qb8+hnvmsyugWxFHN/W2OyXBklgERBJiWlDKGSEyeksHba56b7jtGwFBCc75oL//FDiVnOAfI2IwLGtCSoaUaToqElCV40gUO5SjfZCIU1Jkr+7tO9l/TyPTPLXhIpzeOxQBIRaIBHRtriPx7mnO1/ACoTOklVqNaet1cHJ8V682vS7saNPHvqP1jttOQNoMdHEOgbb00UNGzWCo6M0LuUSokTfS1ZOTU0TYBddt58o60z5/6MymaMiJ3dm3Pd6LIiCu4psUUFUKPuAoq4AnVYIhJRJocubclPNCJluMMBHcbjvePlyxO0gnLqWFRKjy1AoOJFG6Kq6w4kmRQkRIGYaA2349QLglJxg4p8dAMdO2bdj9+zlGxEKmGESQI/GFH/3wC/zO//lTfP33fhsPt52niMCLjCchuZXWcKdeAXyF5849IoeOwByzYKs3c+YMXTuAgJTc2huKFDJscNthSmGSeJAna4th5vHRr4C/m8hT8CuXBU83C3kHPmYYuL5KdOpJKQES0Y0Ku4NUg8ibxUjH7m78KSGiZO7eaxvAAEwSyTAQSEhYU8B5KXh2d4fPPv7o4FJMeTXjzxlB1npDCoZ1PTmDdBZYXjNqXG2KdBerRf/95Pi5ekxLyi6LFw8QDEmCC5SIWbADZK3hqAB07xh403I0YF4GBxdzlmttJLql9FTw4WxETFAvuLOV0L7O3ASn1g7tN2zb7piWS6879RFBInIiS7X1gdoMt+0X0A6IyPcA/G0An/jH/ltm9rdE5AMA/zmAL8BQ0r9iZl8L+6K/BeBfBXAF8NfM7O/+MU+CnDk3xhAxtLth5nxTuOuldZYeK5cJhpjS2jrlhNP5TNnr4w02bywHiXTKag8yjae7dO7Lc8oIJUG9xZptaghkm8HkUB/OC6aUhfiv0S0mRGbwMVnY8OL+Dr/0+cf4e//od/DV2xt6t+PCSolHndQnuan46QF1TNlcChp5yx6mnt5ySjD/OekQzcDMBTLkohtoW0X1HCdvdRR56tLFV400oUi08RL4SedbGqOqLrjiLQiLtCrfi+ZmGWaKpVA8FZP4nBrB0ZZrLmTu0nPgBcvknQiJmSlMYiwiwU9gMZwCU5JhcJMZZQJxIH8hBZ/xQ/LPgXF1KgKoj3+HBFsRejiw+xAijVsDD5RSCkZ3p2p1hyYzBMQjJdr8tPcf4nALC2rtJGepkWTWj65uYEaVxSDIOSKEDIEdgi2Cejh8MgCjolFY/IdbqbWh2N5e6bVZd8RAsZdkemxmNx4hd2Zed/8vigCYLvzvmdnfFZF7AH9HRP47AH8NwH9vZn9TRP4GgL8B4N8H8K+AkeQ/AvAXAfxH/u8/ogYw3dfU0JSBpKaK7HHRMQLRXVR00MWG8UsMu1QTbEMxHm84n88IIaKkhAEg6gwPIdI+Bokgw7XyAW4KWS5IqSBHQW87feszX1MIQMlcA7bBSs5YqMZINGEab5CIJAINIMrtheG0erahRAxvuaHmYJTAkvq8zFmdwiPOdjFFRA3H9wnyhadrT4JnN/I94+8EQB29pp0BwThPTKLvIMegyXYjWMcCN0G2vVH22npD79UZktFNNUjLHuqS49YRIzyTMSGliLvLirVkJkFLxPDvwU9p7Y1FxMyp3pH2aTkceY+tP8390TuuIEyeTgEoOaEpN0HdCMrmFBGG+vs0uRR8juZ4T4jKFXMg+GajA60iSEAu1EfkspCs0+ohwDFVdF9HQp7o6pQ1w5WRA9MKfHoikJsxOzC6EgURJHeDiikhAT7ysqoocIxflCkDsMqItRj99q3owlEHqthvzH7MeSoygRID9BeREpvZTwD8xP/7nYj8QwCfA/jLAP4l/2P/CYD/ASwCfxnA3zaWvf9JRF6IyHf853zbs8C0Heh0d8+1GCnmCfC2zug8XFJCKlTu9VrRFMfqbHevwqAD0xJKfNQISLDIlm6Mjm2j7v+8Fr8LdqBTmJSzOLkGaJ1OOQwwiRAb0K6IsvBUdD57EHcVDlR3iQ73nKMf3NwZdw3QQVpM8tk6Yjijj9sAyBMOklImp3625YKjVTdhRyNg+zm0I6bkfHWi72KKYMNlp04f9qMrCOOv4RsawJxmTQXb6JWdWDDaeXmKrkQWNkr954XJQk1gN+HunBHRAW2QCG4yZBJoBvhfA3A3J6hi3w2tG9aVneEYimn5RWtWzycI7OjWFJAU2LcNre1ojQlT8JXo7Jz8KnNuxQAG4+tLGljXC9QCHq4bT1QRrOuC83rG4vFmw23j1JmKAiLxU7w1TT0mFyWnBLN+2IapKkFWAWDBcRNSwLv7CS6+VSmOCd1qw0NvUBMH+TJQBoJwhOQBuQBgEdJeIUby0ZRnh8CU6Jj+lLYDIvIFgD8H4H8G8Mk3buw/AMcFgAXi//jGX/s9/9q3FoFZSftgospQRcgM54Spt0QKaCPZpnegtoPIIxLILvN3mQgt+QUxkOllAGKiuGi4BfjkzcdAy3JtO9okiIDt21C2doKBU1gpT5YMKQv35c5DGF1yUdwAACAASURBVGbY2o7t9ggzPfa4EYLbreK6dwwNjggL2qAWYAQ5/AIgQIzmnYUTTvziggQSoGQqAN1bca7qHJyj3PbJ/Qb2jbyMoNDDGOMpyccZPlh9xSjBOwQ8ufMGOJHIZ3GGWiRiDJYdm+A4klLBaV3w/LJCtaNVQ0wGCWypp5efKj9fOicZBkjyInnLXZZcOSchI6ZCqy8BnJfnAp7IXIOd14+3SDB/zXNNCt8iBBjaqO4tQMdqVZp3hMBMy3ePG6+fyx3zLJNAO9DHNGeJhwiL4xlI3AHHJYlA0kBti2suRHCEpSQ3PxGf/dug1mBJCSUT/LW9OjELxEdKwWVhcXm3dXC6VYg0xJS5pDUao5DSTIHbMIX1PwVnIRG5A/BfAvh3zeztNxlIZmZCt4w/8UNE/jqAvw4An33yCmaCUk4wkJa5FubN8eKiI0xKC4jQV5hVroSMVE6uhSLUxC9Y3/WWjMwei+2RMlwiBKr7zAbUAk82AMEBG4EezK7eKoha8QZeFwfBVIFRsW0PB8GkNqK0po1OvX3gZ18/Yqv6jQI0DhVhN0McvBCpRfexR3mS1z4wTPwI8SLgsLkp3YrgJJSUsqvRmAgcvBgSP3FKdgzgcG4AFKpsySFyJA3BOO5EGQjRoKOigRsJtqHxWBGmRGl08SIUQ0SAYskBgo6vX3+FUhaklBjJHWcwiqB2evOFQ6br6dDegdHyraDW3XECBriYeMqwRCpJR0U0xZoj1pKPbuPYvAiviXm9mCWYLe5DQHBSQAByIOC27Xi4XtF7x611YNtRUkT2gjPXrFQIck2X3OTDZzZMuXPw9wlwTb/AFZnBx1IjMBl8ODMFLB5WdClGvwaUBxRz5XxbJFANyGWFjgoLwOiTBv4Ub09o6RfsBEQkgwXgPzWz/8q//E9nmy8i3wHwh/713wfwvW/89e/61/6Zh5n9FoDfAoA/+6u/bFEYGVXO1HunyBl13240rEBALCuZZGpIOWNZTh54wSosGEjgCgWgR/3dZUXOhTtpt49asqDrpOl6NHavnNFygQwyuAx0ohnigI+xjW0uNkqleKfCoqGqh92ZGdBHx+N2w5dvHtAtcndvDD4RN+pUrpzRhyGXFdmLQfQYttjU51i3C3dEOgb7hvx07oiNxqYQUmrz08w6LbhLWUiZdb/EmcQUINBREUCBFUzd/JUGJpOELd0AobGlAJAgWGLxtabTgNsVr56vuJwKWhsohXyFrfLvpcNu2zCcuzE5+zFlb9sV4t4OgomfRMZ79wZY9JWjUZvvG4CYIkpOkJhcFzJ/V2IEEhJEshuwery3R46LRJhEiHbYWHAF0NTQbhVLpi9ijoHuwseMzcNFMTc8iubekEOBvXVfUzuY7bbgCK4tGONwB+5D8XDbYUatSQgBd+ezm4YM7PuGBkOJEzdQT7pOgAo0CXoX9No8O1HRfev1C9GGHe3/jwH8QzP7D7/xrf8GwL8B4G/6v//rb3z93xGR/wwEBN/80XiAnyAO+tVmGL0hirqxB5CE4FEMgtPp5PMWZ9QufLNLmC613JunGHE+n1BShKlrAHwDro5qs92XA+AyTPRboBLZwgqOtV6M+cjMGzogXlRCoOMN+QLinSfNR949XPGHX7FTGMbKz501Tz5RntQxca/OEC6u/LhW4mw3o7PFV3RmgAhR5BADV1yOoouLZ8Q3G+wAIoJO5DlQEafUpZNI4rp3J2TJ4UvgOMYYyDGiK7sSCpuoi4+JXdVQPQDE73/+AT569RzrsvoNMfjPaNgd37htHaUknC9nAMCSFyzG35ebG87X8z3iDZeA6C7SgzZswTENSEQw2otn8WHKKbV9cAQYHipbEp2peAIY4XTt0LFDtGIpEYrCaDc1jJEhNhCD0dDUC0KKtLmDmouRhuMCLAj7Tup79uIETBszUpI5jU0dCcNMSEdvEBC8lKUg9oabMjiXilGCq1GcqBQTNDgFPQT0FrBtFU86hl9sHPhLAP4qgH8gIv+Lf+0/8Jv/vxCRfwvA7wD4K/69/xZcD/7v4Irw3/zjnkDAbDnA22gRVDXs7YYlFdd9n/D8xXOkvGJ0urKkRCDLjJ05lO3fu4dH1H1D2FhVWQk9/ccIlMGRZDMhWu7tvjg5ScBswVQWmHKmM4l8LsPTutJ93YYKQlywxgI1ipXGUHz9bsc//fI1Hh8fHWBsfgInlLz6Qsnc3ZYXSPdW3VTpSmwcdZgrQGFswKDpSUxOqIkODs6UI6fkytQGNDD1yDzhCLAYjpmZxKlA++9pyhFcB58SMujsnGM41q45RdzGwF47+qjUtNeKTz844cc/+j6eP3sONeD169e4Xm8HPyNGuEy8491DxXI6wZTFJ6YEk8RAGVUWukCfPN0r04QMh3Hm7Kowd/dGVmOru285DDPvgUAv0fquitCeot1ilGPOp+mo4rrt5JvAPB9iEn52jD4QMHBeC86nFSkW3zgRsAxGGnQuBVEHDyMz7HVgbxT85LRhzdnBS3ZlKXgHPNQVjODqNDAtWTA/p+hEpI4QCARrCOAiKaM3div7Xv16+/b770+yHfgfAXxbL/Ev/5w/bwD+7T/u537zIbM1UtqAhbQgYsEYmfM7DKUUv+AFubjbkLp/vxp6o0srk3mA675hDCbxhJQBPIV/xEDOwPCkYL5wPU42to7iLrPUMAwnu0waaAoOPHkmPS/CSC73XrFXKsR+9voBj9uO1isJIuAJYBJJiJFAqqinIwWBn8zGdtfYLcD6ET9GhJxzNcSr/6E9n3JiQnvBcKgvBWPKBXwnTQLQ5bQiRrcQD8J9dx8Es8AxN6QE1EoS17Hz5rpNRKB9c46B4Aff+x6e3V1QJ8grhrtzRskLRTs6XKfBk/hUIsR3BeTt+5rYg2MAnpCP2w5BdSKTx7SDvHuAc/KYmFB/WnfGEFEKQ2eCDigqtsprY46IKUcsOfvzAfug3b15kxADyMB0AdAw0G3KOkwqTos4zZ1SaHalQO60j08x0iVZeVMaBFvj9ZMi34/kdMOuVAeqKqBKpqTMYh+4pUoJuRS0MaCdmgLEDFji758Hci64piu2vR5r8p/3eC8Yg+oIMCSCZAoSg0o5IUlHiQE5ZRzGCKYe4NHR9oqH2469VgwzrMv6jVksIGRBGLT2br73zvHp4vBN3CHOAdgHiBuBPj4+Pmn5HeNh+0XBhhhXhzFynbWP6mhuREpkhHH1NwUlNMlofUCtHco5cyVbSgmjuaegt5QMC42e9dfh2T8wBITgqU0yvfTcK0A5PqkxuCNHdjmkWzxJpVMMWN3T/7p51HgICEJ14PGO2DQlobpPdEA1uFYD1NaL4fOPX+DHP/o+LufVPfEMd3f3SMFVnXXHsiwuTLIj5XeSWWx0xChYMxF79U1HiQE9x0Npetu501+X5LwHRfPNSfQ9u4HP0c3/Lzobz+jhtzV9Knx9YKv6RCIzEm/uTucnNWbggaCditHbjSaf+rijD0PJTDEuOeKyEHjMydCHszhFsJZyeEUOl2XPM3YapBgMpoKUGGM2NzNmipRWX08mJG/hqsjhqSCzAA1S1DnNcSX+bY/3ogiIz2hmBE5IDvEKHDNdYpeVpxGAtm3QTi+3pobaFO9uFQ8P7/DsXBDSQt53jLS+6op932DWAATs3j4v2Y7KrY6up0yvvNE79sEPG25uQgIPP8xuA5ILUsw0qDBFF57cBsFpWZBixN355PvaAOuVIRUAUlo5g4MGIbQzo1hkes0R5WeRiDEgR99sOPVXtSNnIYFJyFBMjmHY6KRJB8OaA9aSsaRCVx3XLMSSkLOr6MKUCM+ZPx3dgoGR7nDGG5sFqhSnM1AfhqVE/OB7n+Kzzz528w+Ks1pXaORNKXFBTKt3GDRDDc4KfDLhIAyZEnkIB1sRihzYZhPP6Rz3As022xiO1FM0w729ovWGmfAbfXTIiSNGPq/ow3h9YDhpiTflWgruLyfMwSPmEyxkHgqjY1sf8PrNG7TemEkASrsHDCUXRO8ek+v5Q4iQlJzzolAVdDHnQphvClhsy3ripkYbBNR4qARkNzwZvWJI8jUpMx5IFlOCjk7MiikjxURtzrc83psikHw1FpGQdfDmKqt7sAssJNwer1BV1G2D2MD5smJJgr4mbHXmwTeccsHd3T1UB2q9HVhAStGppgMdDa1tmHudFCOW9cyL2gx779i2G1Q7YvAbJWWUkskikImgU+w0OQMhZuQQkWKgYemyIAXx3bsjAAb43gaHxz46mWVOPjmoukq7NLIJOaiTlOTsyckHAC/wyynj1YsLluwya+u4Xm/elnMMSCf/S8JIrRjIXAuBfHPeiH4zyAx9GQfNOGSuIgU8wefK7HJe8P3PP0YIAbc2+HuDGwSqBwNUEm7NoH2HmeK8JuSQHcBinZkyYf5cFic1B7y0I4eAngJzBoYCKSJKxDT0jNFFZBBo66BHAfkiUQRlOeH+fPa4b87oa2Exut12hMTV3GkpWPKCHNlJKBRqXBdaTIhyRpCZkjXzDshh6A4EqypOObrPJMG8EHhYdMNhOT5ViiZerKKTOwJg6lF7ElwS7RteMyB4fkZgJkd3nUIMPDRyyri73Pt6/ec/3osiMJR6cR10er2sC5bnL3B3OVFCXBtGa+it47bdsN0esC4F61hQ1sys+xDw7LLA1HA6XzDUcL3dKFjpjfl5vk4yV4h14w02Jaa1NWyuVWc7yNVVkOgrmeFItKA2gnwxJsRUIOApLSEiS3btAE+GUugI05LjCka2GEEmhdjw0z8cq7uDrDTmjtmJO66j4A3prX+9olXgow+e41//1/4Svv/ZJ9j2G75+/Rq//Y//Cf7J7++47VTcqQN6Qw2X0+oedD5rBmB3/4IYonMTpqKN/ocpshU9Ep4xzT2AX/3l7+JHP/guQhCOZ7NFNRduBaAEdmfDAIkFAwndhBe460Do2MQbhoQoxp1JykjGlj7FgC02jyJn4nNOmcSrA1cBoB1LFizLeoCAOS/UJsQGk4KLBcSc0FvHY76SoReY3MzilMjYdGMbMXJQcoxId89wd1qhHhunqsS0irsat53hoTM+fbAoqT4lacfI151sFtTIFSrMtzP8GjEYJznljElBIDGuH+NsjHRu2vcd205uyHufQNR7x1dv3uK0kERx3dRltxmn9QSkeJxKr9++IRi1FLTekTVjyRn5Ob3k5ilb9x02MupgCkuDoHeafGJSat1thWowQ9eB0Rg4kmNAzryw6FOnmAGlqp2x4E56mRReptY4GQiEmEb3yi3eKlqGOOWTjDFgyqVN3UvPDCTkjEMKPW8IiQIIwy/FaOc1lNuNz17d4c/+6hd49fI5fvbVl9C+4+NXH+DNwxX25gGt0VKblFNg38mNqOprL3mSxPbhwSspOWBKhmZ0yjXTmLjqyjnj1ctn+Iu/+WfwycevcARj6iCxR8RxHWc6girAPgb2rpCYIREoQVBidMFLByTBQoaO5vNwQo7FXY8BSMTeO2qt6KNRRgtiIdnEZdsnV0aKt+t0Xtr7zjVfWcixgGHJGaZnbg4kQkD0vU7XqNHdhp7XRojFR5kChIDogDX9MBnAsqb1KNiqimlTMVGdp3GI16ljvQcIzVme15kO4j4xTMK1QNSzI7wjC6BXQpCIHjwpqXeuFr/l8d4UgXUpuDtfcFry8UG1WrHmhBgLZ14AKRcqDAPtobatYlmAvBTAU2lghph4Au+1cU2yFtRKa+9HuaFVeg9OBlYIM757ICTy9VOcXQGOE1yVQqQQ3OxSArRXPxlpTMn5jh+Saj/IMdFbbvr8T6caFzg5+49Ms3Dsg0mh9dTZIMec3rq69XjHFAR/8uFzBBFcvT2NIeLli2f4/NOPAAjePdwgwghtahR4UXPkIOFm2pXXyuy+k5+gKTrJBQTV9kap97qsePHsBf75X/kCv/bD7+NyOqM2pjqL5wTO9lYCnYFkydgdqB0WIZJJsNFO4QuZExB3Xg7mllyOAYSUcb6/IOcd724Ve6WDcR9c+aVEt+bWGU0+I9tEptUaWZYKQ6u7k4wY9pmTE6/9OU1mduGA2kCtvMFSzIBQwGbgOJXLk0Oy9E7HqMCQWkhAkIEuAktMOpZoCCYeozety4n3dJ0HwISSQaNTecqMsNHhQml2Cgfoy46PxjSkOA97z4HBJSd89Pwed5cL1ssFIhGtMW4KgazyWhu264YcBMuzZ4eUFRBsW8Vto67/dDl560VX3JIjUshImUhr6wOX6xXvHh5wvd1Qe3OlIEk6OdFrbl3PyMuJ/P7e0Q4N/nSR4WihQZwT0B0HKPQjMGrXz6cFz+5WmlMaELXDAglR8J2yuEsQMHnn7iKjiiGMxprbiegMuzEaWm1YckCOQEkZn3/6CiEGmmOkjLIW5C1jWVZ89OEHEHmNGCMerjdIZsvOvIDhpCJX9/nY0GqH2c621QNDJjhKWmvB6XTCx69e4F/8jV/Bhx98wBs+ddhokBhYsAEWZ2He5GldkWtAzZmbHKcFmwwWTmP7nMQVAjkjlEh6dGuYZqK04Iq4v7vD9RaxbbeDHBODK/nqDWYsRMvCuVh7BRV9w3/fgZhWpFSwloIYs/slCpIk6KiQnKGlwDrp6rPrEKPr0xhAHxE5FnZUPs703knCEM8ocDJbCAlrOeNI0jJ1rkTnetA4Kqbgo6O5pRkcfxrECyZZP7jXgRoPCnZcEeacFeA97wRSSnj5/DmVY6czEAJS6xh1hxnYivVGeSnise+v1aPFXU9+u15xt290Y43ic5c6FbU7om4oKeNyuUeQSP96NW/32QKHGDxRh8pFlpqJWtMmbL4GU7ocpVhoYyXBKbxMrbm7nPHRy2dI6TXi4DqRKi8n3MTE9lbGUbzMqcRPTjU8IZhBD595Bd2moSVwPi149fK576gdHBMGeNxfThjDsJZH3LZ6WLhlVwvO+VkVqI2g0nTEuek8nQwxPAFvKSbknHFaFvzoi0/xy9/7hKs/FyEFyXzdJWNapEEEIuT257LCUGF1Rw4KC1xvcqxqCIGfXRMGeCLg2BhICLhd3/l6MuLudMJpOeG2PdB70oVCMEPrbPXP5xXrckYzwcObn6Htj+4bQGv0tj2gQWCnM3K5YGY1du9QgghyCNAUHDx2+W8MyCgYvSFgwEYFPK5OhIWEvpdcBYaZJeEWaim6wGioE4TMR8yASWnPvj2ZWwtutb0d8AJhRmekyeGgwzKJYs1m+O633H//393af/JHCAGpsOWX3mFiqLcbHt685akHinZOpzOkDWzbDXXfj+gt8URjhOj8ANJ3g8ihxR6tY9t31F6hbhXOgJNJKMYR9qCjM9HXBsaoaK37jObuwGE24I7oulkHZ0m3M0sZvXfklPDZRy9wWX6C7jLh6dEvgdr1EDKsP9ldcWRkwZnOtZgrNOu+dhLAPekEES/uL3j+7B6mHbUHtN4B4wlfyoKcKy6n9WiZm8eotT7cWclgPjsfcylAA8v4xMxM0+DTaKn94tkZf/7HP8CzuxU6NgBUV4aUUZyL0Uf3G5tt+b53qHH7MpSOwynI4UwU3VWp9k67eGRag4khpYKhTHIeOpBKPlJ7YrhgLATpxNe1i7pdWWFYKXQgR4FFIKWApazcipQKQUKImeu2uvO6mPkOOpBygGCBdh4+0Z2Yg6+cSRobvtrlOKUWYUqHn8kIDKk4vZtO1jCmbbVBUDTKdE6Obnnn7425D4a645AXArpv+YEAyuvrYDhMV2JfEt5zezHWeEFrHe22o7fK1N/9hmGKy+WebXYISDlQ6SeKvGScL3fMC/RwStHuO1jeOLS49pjvnKG1YW8NHR11u8G0Ys0F67JiWVbSgTs59czcq9j8gjOd1ZUX0GxzE7w1nAGhh+23Yl0yfumzD/Hh8wsed0UI7ZDr0v6c/AjxaDGArd1MCuIHy9Wgtp2dSHDZreAItPj0ow/w4YcvEENA3Tf0uvuowYt4WQrOlxPfEwW21tFa960IwaMcyW9/vHFOle6GosDTetBNQAwKCYYf//AzfPD87CdzgcijG2CcEVMEEm294GEtIsDehhuVNOx1Q5CApWQXJRnKcoJBkZtvdoxlmhFu9GEMMSJmRoEfGQBgV0ZJOjkhEkjiYqZEB6wh5YIYGe0Fp9xKXOj6NLjInQlPgb34xPG4u5eA4eY2QUAyko+I9Lqk5h9QDKMTdl5WdpyjOiDMvIkxBmq9oikFZ7Np91p8FOVpaqJGw93aXbMiEYKOc4kOmKqrF+mSzWvKHYu/5fF+FAEz3B7eYWvc2TLqi0yu7bpRS306e5tjyIUtYckJ6+kE1YGvro94fHxAAHnap2UhnqCCbe/YqyLmDMSE7eEd9roBMORISyuV6BJlI9DnwRq1MeJ67qzr4E68DAJ9ORWIO93O5J8Aw1DGTsdU8MVnH+OTlxf83pePrPAhIgghHeYM8G1Qfx5Yo0TYaGppOrkAvgqSeCjPhvspfPrhCzy7XBADkf3sxKqTZKjLkOFGo60rcNvplNQ7Zgo03XV3TDdhCqn8337KDJ1GqhEfPLvgu9/5EKcl4Xy6YFlW3qimCOAWoruunci4QgKZoSkmpNMZS8kHOg4IxBqGC5v0Gyw3NUfHwfCWmM8oIXGVCPIOpn+kOUuyaTu8I5IbbExKL4laAmvdb7Z4MPlMB/btSoxppet1ScVj59gB8HPy3Er4qewd3jBFygs7hZDQFQgICIExcfxduYbeR8dtr4e/5OjD2azBTVECgvnKFASmJ9agAkiMaLWiNuBuXRBSQcoJMVIYthqJXA/X27fefu9FEVAzvLtSrVW7IgbqyAMGTgs/kG2/5wcM2nxZ7+gSsF1vGNrR9o1Ot+UOMUSspxNNQmHIhXPmvl1hkvDixUvOaq2i1up72UguuA5EEKza9npYW63Lisvdc1y3zUMkO0riB8uVGmB9oORArzoxDCHd99UHL/GjLz7B3//HPz3oyph+/kNRyvS/q9w2qHPyvTqMoYjRabXeDpLGxjn6tBR88tFLnNYVfVSkFGl3lQLyUOR0YorPumBJmRdXulKKDOB629gST2GNc9hzSh4xRicjHRyJhikup4wvvvsxvvjux3jx/AVOpxNR+U6PwOYOytxyROx1oI4BcWQ9iKsnp4GqE46GEbiz0dF681VsoqpTEkpyJqBW7FtFSYVJxS59HhowFNgaO77ByvYNJ96A7hiLgdZo9EHgSBQlQKC4uxR/fTxsghupmLkFcpjaDyL28M0OAjkPIUQoAvoAtkpwdM2Jug8Q95mF1YTXUUkZFdQylExZdQhPgbTmHVGIATHz5m/tSr7GYIrVMnlgIsgx8xoJSo/Mb3m8F0UghIDT+Z6VsNFmqmS21jkCAH3wH96+427ewTOe9M7KWk84nz9AiAXdE3RQK/3mPVjkfDqhLAtiTGit4najJn2rO6qLj4Iw2JJhGQVrSMgp4u7+HpfLPb56/QavXzdYIJ05pwLTgRh8Tem73pwSpFM2XJaML77zIS5LxNsrW9Qg8LgtT5j1VpLehe594EQms4FWO6YTkvpqUVy2fFoLXtzfs62VQPmu6x3omusgqiTSf0NAyOym3sUHiqkcFwghYVKWY4hYSznUklODD6cIf/7RCwQB3r57QG8Ny8Kb5bZvGAqspNoh54LLGhEbxVWbm7SEQPOM3hNidEFWb4B2xMDsCYnEeQJIzhE/baMo+QOtIYR7qIOqoxNd32tHG3xXxWjBLf1JO2EOVAK8ecQ1ByVFpOABtmFKcAP2rlBrSNG5+RhuSsLfOThfZK6pu3bUnXTiicNQivwUERYEWHKmilnp1JTSgjEGkluXpUTlYG0Vj7cbzAliOmhdl2LAuhQsLoCK0W3jHVOgvdlwcdTPf7wXRQAQnE7/F3NvEmvZmqZnPX+71t77nIgbEbfJuplVWZ1dVS4kykLyABgyR8xgAAMQ9gCEkBjhEZLlGRgxQjJiggRCSDBAyBKCIaITBtPYZWxTLqsyq7K/ESfO3qv5Owbvt3Zk4bqJm0nsHGTmvXHinLP3Wv/6vrc901slXS40s4KWogAJF9RvLxCr4YYp5apGqBSVwoLzvH//pJxB44Zvi9aJT1690gdsyTaH0WOcJ+od/NOpHCxXLoTE5eGRPGXzNjge5hPldKH3QowZ7yPe0nkGugCGcxY1pcOh152f+/QFX76+8OPnYmuHHJDOqsEOo8+h1T8UhRjfe7QXHyUqw6YAnOOzV498+vqlgWOGEHetG22o5ALfmBo89oaLEbUX6f3ovbOuG0/XnX2XMlMy3yN0xOFM9Tb6IOA5nyZ++JO3/JW/9jf5/PVLPnv9kmkSE+FDJKbMddVa9epF4uF04nKeeb5Frstmun8bqYcjuEm5km5AV7lKiBG8Zy32u3TDfNzx9NbhHpM+m+AyNXT6tooOLIWtWMhoqeqFDNKRxOjJ06QIuEPXb2CyD4qBC0PZhsdqWJuxUXbNYgDeQRnGNOG7kPrWGnsRDjG6Ck97U5JRShnvAikFUjxESQojjd7TYtAkOBqt7pbGJIzhAKlzTrw6PUjQ5rBcSLtemoRKB9VZq7QNX/f6KA6BgaSzMXjWZeH5+YnRm3b++UTIE90l9h1GqXg/xL3WwpQTp9NEa53bbWHbFh0ezfHw8ILpdKHuC9tyZW36OyUh1X/6IbrxyhcwCw2n6URIyZJzs+iiItVYilGFtKMz6gZedVWMjusG1g3zPLSCY/D6xYk/9gtv+OvffcdRAgrShXcvlLhVKwkB4GgM6lYeguK0wMRLeu9yjvzxb3+DNy8fRJklXQzdQlZ7d+y1UGu/TyiXsxDoeZqYcyJ6z9t3T3zvqyu3rdw17zgbva0MVWtPZ8oS//zgJ2/Z94XldmVZFj7/9DXeB87nQDDL7XJbyeEJ1yshmax7SjiX7pVtcpHWu54j+JN2+VaNanPgPGVfoRepOaeZo7ZujEYKiZgzy1boeWarnXJbKftuSL6UIm6NhQAAIABJREFUit3yE6JFlwXnCIC3EppSK9u2su07zsGUZ6acid5Rdvk7PljOZfY5xvYYEzlPNk0paUgCVtGCUg0K9wlRJp8QlZ7UWrv/rtHyI9teKa3Qmg6BKU+EcJSXTuQ8Sw2JvC9iLpvJrI+ylXAHdb/u9XEcAr2zLFfonffXZ2ptnCZFVsWo8fR229iKxXgjlDV4mOYHcHJRuZBYasePwSmfxNOfT5QUKGVj3Xeer1dev/I8Pj5ye1p5er4q2tky/0qpBBw5K9GmWsloa41SB7VjOXqDVu2JPga+e+OFNTKvpfJ8fcbVlZcPJ8Dx5acvuGTPu+PiCarlvrMRFnwBxtMfe6YT8DQMDFT5hVaFh9PMt7/8nFILb5+eLAMPE/h0jtZfgGH1X6DDo9ZKzxMvX1wsil3W2MMX4AImJebuo/dBh2gIgVIbz7eV73zvJ7y/rqx75bbud7oy58h5znjnpBDcdCBO08SUZ0ABKK03+hBANxEZQ6Nuq9KJpDjjYiKMCoeQKM0K1rADpDetOTEEMpHTSYeo5MK6IUQHK93n0EgcQN4AOSq9Z10L19tC7535NCSq6o193+40nFR66qE4aF/nPSll02EoO9FZ+E0Ilt/YO6fzhRyTqVQxq7Ke+j4m0XmjsbVAadXchf2DoMx0BKUWGtKaNLs2vBcboc/b0pn616cKwUdyCNTW+N73fx8P5JSYpyzZ7+lMnuQknLNcUnLXW8gGolm2bcWNIbOP6hjvIQ6K1JJOfsqJh4cL+XTG54lpnpjmk4l2RE9dl4U+BrdlJ6V+57GP1pzR2h0z8MGbweenT1qtAu+vz6y397w8Twyvm/0bX7zmyzcXfvT2WR9caIxR7EaohjQbfzxMW46h5oeqbAAITdf473i+LvyN3/mOZLmWPlPNMDXlCW8tTgf1dtS2FcMi9tLY9sJtUxCK8gx1cDycZoVcWGRWiFGlrU5GndIGb59uXG8r19vGj7564uXjmW989imfPF44n2cuF7icgiUaCQFXWl6gjWbIOKy7AaXJRFEWgR68ItNiSsqICBY7nyR5jl2lrJqkGtFnpujx5xMpeHrX4ROd9TyOfn8/iyUHpQiTh9PpDO5oXAYXM+u+qd9CbjO8G5zmTM7Z9vXCthfa8OzrrmsxJWKQupEpW0CMnvTTdCKnbEIfCc7GGHYASArej9JUL7C62WRYmzQH6z2gxsDLqMM5e3dPcNJ02dhLo329YPDjOATuQp3ROc/5Tt/MlxfWNViJURfhsq74oUip4D/o+McYROf47M2nLNf3xOhJ+cS+V8q+s6wrwTs+mR9x3rHcFlptnCcZVFqrls7aaG3IHJJPdBvhatOToJXdKDU5tcYY+KERrQ+ZgEprRApfvHnJ5fxI95Gyb3zxReLXf/Vb/K3v/IhlUyhmG41qhSgHAnzYiY83pw8FggR/kMcaGFofLFvlr//Od/jkxVvtjeYtCF7BKA+Xk3npMymfRE32xpx1EdbaWJaNdVeq88FMHGPrFpvZrxX8GWKmdsdauxpv/MFYeJ6vK6N33r59x/d/+BWfvnrJ609e8OrlA5+++YSHy0VPNOfJkzNmRboIvNeNYlHmwXmc0w1A0Hjt7Uk6uiLEfPDElKG0eymNwE2xNSkE/DwzRlbKcFeS0mHI2c3Nc7zbrQNegPDrV69oHfZa2F1jo0EcuJBIwSmIxTlC1Mp6XRbWqlLc0ZRY3Rw6qIjsZac3AZ7r8kyw3+fuaejNYH15Ptrw7FUMypx16LbeWHclVN2WVYIv6+bANeJSeHFKXOYJUJDOVgrrvltM2h/9+igOgeA9r168lDXXa2/LhpoOa83xdOYILqEwhqZ22LLtpDyR88R0mgXs+WFyYEVbXU4T8/SZRrp1AVb2IpNGM9FKa0WouYOHUyZPJ9KkHPrWOm7OLOvC9flKniayVYiXslM2az2m09ow485ruSDPF1xItLpzOnd+5dvf4MVl4u3zQooKqIhxUCRaVUQ1cBeJjmH9CiaH+UPyUTEFT7eVYsWhbhxJvY6cE2+fnnXxO0/OiRwDU86keLj5NNaX5jifHhjjhxafDrTG+9vVFGeBKScGolJnc68pw8KJ/tuVb9B75937hffvr7x9es+bVy94/3zj5eOZKWceHh9o3VmXZL4DmsMPs9VGo/e0DqUh/UCwSrRg0ee9DapRarXsdBfAB0Zpd81FPZSFXrgCYIeBcJGQJ5TNKOCxbQuOLqHTqCQPPif9t5NGIKWsaHRLctIT2ZNKo03JnH/6lJSPKHBXRa6R1irbtuhB4vx9zbtnMLoPeMgY7j71pWjVcmNQUrByXH3dujduyy7rslMt+7LuPC2b3suPXSzkvTjco/3XR7XV3N6/v8siGYNaJeVMKenPx0QnkNKEc46yF/I8kfJMpNNqY4RBq6uBZUfgZJd0dHTr5PN4y9x3iNLK2Uw9Q0/O3ivBDT7/9A3ny4WYInRY943b+2fOJ9WUH9uXUP9BKZVIE1rtG28+ecG3vviEH7xdIBzE4PE+HPt8MVHIsX+Cyi1NlNKHXXjqHljXDbqQfuxr6INaV6Oz9DrkuLLkKnjjKAJ5ePyCzz/7Jt/53g+5rT9mDNgt38E7RxjCBdZtZRqZKc5q0fmgcaPWxnUd6iD0agJ+9/7KthXevbvy+Hjm9ctHPm+Dh8dOrxl3akxTJpDwSQDbVgvPt2fzN0gWnIIOaN2/Hgy1L/vKqEWuOQL4LFFYr+y7boBamyZHW9kOFV70MnwFb9NeLwbJOGrRA8IbdRhTko4/Ze39Mclm3QoMmHLnoasfUEo94Vetd5ZtF63tPcMCR3zZJZGOJg/GgYuiN7crDlGyMWZ6q+yjoCLemRchKlmoHyG7jXNubDWyl8L7ZcE7x3UpfPW0stVm9O4f/fooDoGBUUJTtmARndraZazPrVTWfeHxPHO5PKhC2jt78muPanuh7c4MHBIj++AYQ7qAYoCXdNi6vXyQ6ENBLgf44mBURlNl+uE0y3kiOE8ru2EDgQBcTjMpp3v82b4IXZZUtcHIpHiibBsP58yv/Pwbfvt3f0QbjZ6yPfEgmBDIoSfcsCxAU0BjxOCd3+7j2Am5j7RHj2GMB+gmUwpON1fwjtumC2rOmT6aRCZz483rL/js0y94+/7prps4RC29F+sBGKQGo8vyq0ouzHXnqbviy6Yc2WqDBc5TY9kKT7eVZSvUDp+3xuOLR8ApryF4SgM/GrXulu7bCHSql936oD9brwQ7JI/AGB8sTr5vOrCDtUgHT3CaFpyTLVvMQ6Naqo/PWfQr5tu4n+S6iWXxlgYDp7DR2AbbLo+KHmCREL0sxG5QirwhMfIhq6LJAl72TTXDZm1OedKBUVfKthh17Ondkp27HI+tOx7Omug8ELq1VgfHKSfOo/O8eJ5vN5zXQy8EiOZI/LrXx3EIDBhxVsJOLXikptsspEMcbTPkNWnkjuY69+CmiCeKVw5KZfEdaIFaKk9P71mXZ7yNkmPk+03jfeAchcbu8nLQe2W9bYaI61A6tAW9FtZyeNCTcfIC4cp10WhdNppFi4sBaFyvN5Zl4ZQdv/kr3+Iv/9Xv8re//57ag/kiji573eoyMFmdVGv3n/c4DECCkBSVPSBNfjUBidKApPyTrbV1rQ3BJoNgSUq1VaaUOZ1XTvPEt7/5LX74o+/z469+LFXjHQvX904hcDnNzNMEfSd6z7oXjjKO3rWKVWuKXrbClDOX00T0np+8e89Xb59YvvkNPtsaD5czWx2cTjN5mnFI51/NJ7G7wwQzpD8IagWOVjfnRsAHkKf2oFA13c3zidQqrTlLijIx07qy9QVnN/QRsRasMaiZg7K3ZpFlR+OQZREOx+AmENM8Jx/yFjqtbELuTeLb79oPGb5UG+eppRJSJteiuPba7hrxfuQZFDU0T3limhIpSaLehg7M4ayh0YsivJy8mac6c+5yYA55Rb7u9VEcAvLHjzvAV3uxPWqQYuY0n0WzRNFa6yZ9+0GzhVoUY2U1U+AEzjRVd6cUqSWxLDd6L6RSmabMlETTpBQpBVovH5RzQ3z/cBpr51NmPp9opXG73XDOcTo/qHSjKrp6WTfWfVMzsYFzOWe2Uln3Hcfg4Xzml3/xW/zGL/8dvvOjZ2r3tC6gynUh1ilF20kx/3zlELgdYSTOzCu9CxlYtxXQBFB6Z12Vx3g5nThNmeEG5ykZ26G1q7ZmASMbP/rqK778ZuGb3/gGt9uv87/91f+dd++fOPwzMSbO5zOfPL7gF778gnL7MXsVyHescrXqyDh0+s7J4v10vfK9H31FToE5R96+lwT8/bLx6uULbsvKq1cvuFx2HWBV1W/uOAj3ovXnMER5JR6lI/glyKvfUBJPihHXFXdet8WCXjxhmgkxcpkj8/SSMD2w395p6uvyacCHw7f1ZpSq3mtJu3W41dopbZDyzjwt98AYHSbSZmxbk7jHeebzWUW2XUUoroo5CF1yZ4/arw4xmEpsO7jAOXumeWaeTioldaZmbA5XoRse5H24G+0O5sO744FVvvb++ygOAWm2q0AtE1k4H5gCzKcTMU8qHEH9A1PK8tMX6cv3ihVhKgTT2YVUtp3gA48PL5jyTEwT1+XGut4YdOu3c9ScpefvnXZcfC4QswQp275RS6WWRogZHxLDVIIhRm7LQu9S/U1JtGXv+lnPlzNx3zlNCpuoTe69X/ryNZ88TCxfPSNEO5gO3eHaEYelp1MN1TzpA9WIdhuLtd/vVbXd59PE+TRzXVZGahYmYYeakxQ2R/Hk2UJW3NVLP3F75nZ9x4vPf45f/va3efv0lr/5u7/DbRWA+cWbz3n96jWfffoZp9j5yfYjrr0TnbuDVc4pRuyQQrXROGUBvK0Nphw5zUK6b2uh9bds22bsTGFdLrx8KX3BQeG1rpG8NYGbKUP30Mdu8t8EiBdvVs5ae4XoLXm5KYvAW1uQZSL6EO8dCGIsjgJT6Q5Axpt1r7KQ+5/S+w/HVgd7haVs8le0RquVx4eZ8/kMzhHTgbtgWRMOhseTJDIbEvUcXv/RKy4meh/s286yV1yIPJweyWmS76ENMLt6CMqsCNbFKXGXJzqpUEvduK37Pfb+614fxSFwdLo7oPZgSGbndJotDtvhp4lSK8vtqrHn8ZHpfCZUOcUwiMqp01yuuKBGHMdg9CNjP0NT1qAsmZVTa5zmk75XsyCHPJmiSz312tUq0zTxyatX1LLrKZRPtHFlW1dylNpruE5MUaKYSYqz1iq1FMpVjrFPP33JZ5+c+MFX7yU9jcIGaJXgrGaMQZgmOyBkX3XjcKRpRO0dYRVe5pvr7caUMzme7yNtdFCBbv2GvfU72PXy4Uy4CRm/Pb+jvX7Dw3nit/7Er/P6k0/4vT/4fV48PPKrv/TLPDy8ZNsWfv/3ftvWiMC6V4bVhsWgLsIYlZYT48AjTj/nzClnU7DBbVl4/9x49/6ZZV14f73y2ZvXjAGfvHqhclTUFVBq57mqgisOyy4c6k7MZhAKhsdIBQgguhbniD4afiP2ZKD499vynm272VNcmoTalMN47O9tQM6JaPkTsjRXfHUWKa7qes/g4eHCN774VC3VrdPrzW5yPblDnAz4rbRqrlNFqKJ49HJXUV7XjWWr5DxwXqWvxwF10IIhBbwvhskcWZSeVja22lj2wrIV2Bvz9JF7B3ofPK/yf+/7RmtVrbZevHXKjel0URZerQKG1sjDi0Sez+Jkq54MvWxQHDSFR8Q52V69U7cb27aZ1VasQprOnGfl6JV9oZv++hBdpGlifnhF223kc4OQM7TJHHKOaUosy1C9dFTKcK0dtzdi2knWZVBrJeWJT/LEL/1C4Nd/8ff4f777Exrg00Spu6UPW/S4BZniFUapxlnlIg4TEF1OmZeXCecD755v1Fa4zBMxZm7LSkwJHwMZSa6PQJPSO65rQsgx0Dts2419W0nxgVcvX3I6nfjWz31BCBPTrKfb73/3d7hd35oHI7JshXfPN3JKPMaZGNxdNTmM/tqbmqUwoLM2yVxr61yXjXUrbKVTmv5dbY1PX3/CNCXrC6zkqjUgT4nrdbnz67Ws1HREws33Qlhl/wnzOay5OH/fz/da2cvOulWjaYU1tD5Y98K27jjn+OTlAy8fH5TIHBIN7e7rVFi31SaVyMNp4htvXhNT4u27J3rb7+Y3kFoxBQWsMjytJzrCbmpTrNiyFakTUXU9XvhW2TcLd4lU64nszqGIzKDrwZSHvQ5pA9aVvSh/sXbHXj5yihAkZ133lduy0nvj5cOJHBtlqGlotI6Pkel00gg5hpqDQ6PVRtlWRUH3jgtH2Wcl5ySw0HhzvGevjZwyl9Msb0Gt3G5XS2ZVTRkgLwDOJgn540eH2GR0Cfb3zXkixMS2rcQ06QNdbjxfb9Syc3k4qxEmeE7zRb/HlPlHfu0X+O//j7/N+naX6MVhWXuHYhAb9xx0T7MMuSMpWe61QXSDPho5OhpSSi7XnW0vPJwnojX0xiPA1ExDtZjxaYhTX7crz8tCyjOTixb2MRNCBhdYl2fevf0++76ZQMVbhZnG1NK1lwanz9PQC0IIEtGYOzRZsUnOmdI6t7XA2/f3xNxatea9fLxwPp+IKZu11mrYbGcW7jIIYbCXokgxqzY7Is5itI5LM2Kpdi7SkQTa54m67RYmL1yhi6ohJ8eLy4mHOVnxaKY76fynuHOZM7V3WdadFeqOztP7J+bkmONMTIkxFEQaLYxmDE9OJ7oLtKqQ1Ntyo7Qbe6nElMjRM3nHbMlJ+74RE5SyM4C6FdZtYYqaOEMMuODpHJhFxTmvVWD8NJX7d7/+XlqJfx74j4Av0Kr3F8cY/55z7t8C/mXgh/ZH/+wY4y/Z1/ybwL+EovL+tTHGf/2zv4duOOkEuD9Jwl3fbe663vFpgujZ9p2cK32v7OuNdVlkAY0BHwwJ7Z1aNsVKh0h+mJl7Yd93IJBzFlMTFDE+m1FE5RfKLmi9si/vcSbyaG1Q6kqvO/PpRAg7fTTO8yyKqSkPkVGhd8q+sq/GRRHp+4YrjdEGn716wZdvHvn+01fyKJiTzA2Ptx1PmYWJ5qNGz70zLNhEWIJBhb3jB/eG37FLJHNblDw8m8T1qE1XTr3KTZwJW2oXGCVJjiRaIWSFruB59/SOut8MsFOOn3Oe0/TI0/XGXto9UBWnCO2tVrq18xyJztih6qxebPTBXhvvnq4mY97ZS+G6rLx++YKHx0dSjtQuDb0POvxakbS3dlFg27YJKI5YcIgOT3kxGq5bpVuOnIMn54m5GWVnaHzrjTIN9joTbOJqfeiQ9tFE640cZQMutZHN9jucBESPlxNzjkw5Gw5wxNfZARUiw+uQVcxZxjn1HqaUmLMSnqIlPaUYjW4ebNtCaypmjV7v4/X6nuADaZrlhTCF6xjd0rvDBwXqP8ghgNbJf2OM8b865x6Bv+yc+2/s3/27Y4x/+w/f0O5PAP8s8JvAl8B/65z742OMr51HBCih0E0/01uTu81Ueb12fJDDrNSGq8pvO862TqAMz3J9Jgfk+rJAylYU8eSdwkhC9ExpNnlsQRVfKn+YcraK7WEJwgW3iZ6hN/zohBTY1p29rOBUjtq7yjHmlNj2VZZT54lZpZs5Z412A0Zt1FG5rQuexhevzszfecdShgmjkv1O0vUD5nVXeYd3geakFDtKU0KIDD+4xGjZATvLJlS41MYRKuq9at69+EcLCuky/myFx3Ax89CwxiKPC8p4rLWx3N6hoFVPaZ05RHJOFnoSGXXcY6y883djjz8mGC+6MpmvY3Sl6e5OaHyphdttFbA4nEW06TN+8XhmmpJivG2VcKhavtZOivrzvXei0Zqjd4bTxNBxpDjZGuMUL54cM1CnTNsVaVe6KuDvLEXrvHte8MtOtGBRj7wmpW7CF0YjGXjtvOfl4wuS4VKtiSoeLtCdUo/8CDpw6n63kTsGU4rUAac5GXNlvYoHx2+msrZupKSVb5gJ6ui72PeVVjdhG7vlMkQVsv4DHwJjjD8A/sD+93vn3G8D3/wZX/JPA//pGGMD/rZz7m8Bfwr4H77uCxwwRSuojA5SYtjIN8+ZfD7dp4TgPSPNdJ+lWDMZ2WgT78tC6QXfGoTMqHrKhaNVeFfnuw/hrrGf5pkYEvMslLl3qfyGUTe9Ntq2yY7q5NtODyemHCUQCskkqhGYSWumjffUshNilMUYBUUKzCpQ1ajEqHzx5swcYS2KrzZmX4pDs5Zqtz5G2eM90z6vjDrxzDF41nVj2wspZjqDvelCbq2Tpogf2jflLtTYPsA86wG8sgCi1wVbWme4SB+FdXm+/0zn+WQHabMgFN3Qow/2Xsk2uqcYOSzWzZ74SumxfAIg+mbHnVBxPTm9tTdNnCzPYZoi0etgbX1XmUm1tbD7e7qud4grtymp3Jt+ZvCTpZ9HUcjdmrBTIvhBbIEOpDi4bSq6PXIeQlD3QwpR+oYi8VWMx2HYxP0z6Hgs1xSHtRmjz9D1Tm27rQGLfC4h8MnLl+ZSFeh4Z1nsoReCVqlt09fqwBXOEUMULTuO8NJu02InBAW0fN3r7wsTcM79IvAngf8J+CeAf9U59y8A/wuaFr5CB8T/+FNf9h1+9qGBc/BwPkvEsa8Cblrj4gM+XBhDp3MvG3PWk9UHnZB1b9B2PI3L6USrMiDVez9gJ0VPK4V9vemCdWptzSnzgL8XdQSj92LwJgPO9A77ttwvguEceT4xzd6y5eUNV0Q1pGnChcRXP/ohzrjb0jplXRitQR9Mp5l5mskp8/nrl7w8Rd5etw9Boz5Ynt7BeYBZF7m3Ixt3PaXAtmkqucyz9mLntb70qoTj3tks7Tb4wLoXllLNpKNxt3dYlhu1VHDB4qtF17Yhp+F1WTUxJT1rj5KTZk3G8svr563N2ffzNKtda62zNjUVn+eJ9zflPCrYZNyr22vvxJQotbLvheteKW+v1AGX04TzAsoU4GINPh7txfa0VbGqMBxvunznHViR6gh6P513JB+BQC3Qg4DT82XmsQ/Vym83UY9DD6brVnhe35GCfp5eK7vZuFvrlN4IfseZ29OZL8ExCDEzRmVZb/dmoxgmcj4zzRfzEmiNXNeFVndNX5gnBMeRiN1NOtz6YES1O+s1DKA9WCVNWV/3+ns+BJxzD8B/DvzrY4wn59y/D/w5dFj9OeDfAf7Fv4+/708DfxrgW9/4jDTNZqBwilEqO9u2MN4Ncp44ny9479m3jV47KekNXpdVfG8TQDhMGeYY9xacmKb7vre3rlDRtRCKZLVzEj6QW77n/2H0z/GE3XbzdZeKi9p9h2UBOC+9AFE/v0w6SVl5dzGIRt6jW9F5x3x+5Hw58/pl5u/88EnRX6Yj7+NDypAfspH29kH6KT+Mv+vecwomEw04j7oAm0CivVZyispFDCaFtZvN6bMw8LOYBkEqOfHzWquXdaOPxpQntrKzrtu94gpgOH9XNY7BPdN/DDEBHf2zboqcZVMWw170szlEteE8oVk9eB+U1tn2hpsCy3o8eU2d5yKgw+aoazNbgEBAYHSFj6Yk09EhLPJj2BMenK1gIWb86GBxcdHCQmpSbNlWGy1BSrIO133jelsJwWEuYz3Fg9eKyYEFgXh9TzA78BFqK4l3Ik8CQOkCfUfM+FB0DwzVwPe+cC+IGQJRjyBVleDsHDXq6Ui8Rr2Zyz9sK7FzLqED4D8eY/wX+qDH93/q3/8HwH9l//e7wM//1Jd/y/7ZH3qNMf4i8BcBfus3fnW4EEg+MDXdqNdt57qs1Kayj1ZMs99V5VRDICRp31vrtudpzPd2JQQvU1Bfb9R1NReXgi7dJL9+qYWUErUPxr7rgw+R3irLsoIbvHh4YD6fTC1WuL5/x2meiGmCIVy5d6BaQEitpBgo3YvZGHCaEj5lCMEAvspWNkZvvPnkRA6dd2W1+Opg6bK6eNRA3AzD0JgXDMz03jNF4SMqF1WKzWXOLNuOPzm2KjFJHXB03CkzUMBrrerxm1Ikp6hk31IUtjJMxblv+FHYto1SKtd1tact9+zBvbZ7j4FzMK6QU+R8mjnUjArA1Mhe2+Gy0w3UBooY80f3orewl8acDGjzUaq72vDRk3IipUlrlkGcyjpQlyO+KadPDjJAxTKOTpqynvBF7VXBVKcgSrc3ic0IycpWhmlaEnvaKVNkL7q51B5kSsPWSZOCZwawrgpynadkhzbA4Egcis7dzWBuSCbemvoz9moCqDBIB6BrMnHhr8fBa1mV9nnpZ+qsRXkCW/mH8A44fZf/EPjtMcZf+Kl//nOGFwD8M8D/Zf/7vwT+E+fcX0DA4B8D/uef9T16bzy9e0eKFhKKcuP7kLKt7jvXXQWS0zyD9yx7IWLR1UGgiB9e0wDWquudBZFKgPN4muQNGNBDotUP0dZbV1lHCInhHOu2cdsqU07kPDPPJ41bdujcnq84vxr/LNS61Hq/yeBDHmDvldIDl2m+B0ekkxDfh4cL3/z8Uz55/D4/eloUf+VBXXz2gY4PH7gqrpRYG0LgcpbnYt/FUlRUY30EhQavSrfVKLRweCsMpVdlVSBGb844ueXwAU/XrwZs28JtuelpE7wmoUNTbyO5F/bHXixFt+s9ud4WBlI3zjkzulaKQ/Kr7j0BlA9ztlt52I5dWJeFW9TO68OZKQfmOTHNMy9fvuJ0eWBfr3hrkDp8HqM7kybIO9Dv/QSD1oraiQ/qzoJF75Vdxpz4mEghmZlL71eMmRgDz7dOXQbRwykp3eeYTBSgWtlrZSsyGgHW6iwloYJNdbh1w24cmClK2MhpntWh2RRN3wcM+3MhBqY86XPwDo+/dzMcSktVrzfFzP+DHgJo9//ngf/TOfdX7J/9WeCfc879lt4xfhf4MwBjjL/qnPvPgL+GmIV/5WcxA/Y1PL9/p+BO5/HB8dmrlzifYAz29aqw0aTcAHxgKatcvlrz8C5NwYckAAAgAElEQVQSYwCnMRPnNI6bAy/FKGcVQ0raCCM4SlVwBENqrK08W8y2Yz6daKXy/nolBnXZKdxzMFqxSCs9VWrbiUSWZdGB5ALdOTMSiZNufdy7DUcvvLycaZ99xvNt55uffoe/87231O6VJw9wH7D1kgXWHgFDLUQpBNWlMe47cGntroMXMq+WnJwSl9NMqe1e07XXRnMKE83TmZgsajt6apMkeFmrCjKqfOl+2A3amrLtnC6wEIRTpGgpQm4QvXwEpcqIdcifcVYWGgPNpoRsuQIp6hBr1gRc9p19XygJenaEfGLOmcfzxJwilJ2EWas5mqn7/R2sFpBQqxmbykpAfYkxZYZ3eH9ShqOD1nboAx8UdhqiHhh13+2Ad+ZSFd3nXCenJKDYRbo9EPbiLNMC8wNUalWMWXKOlBLTdCKlQ0nYrAK9GluQOZ+SkoZvz5ZDOCz34sAatBLoYHb3B9ABDKbgqEHXy9e9/l7Ygf8O/tC1eLz+0s/4mj8P/Pn/v7/7w8tRh9DUKQkVvkwzzcAqNdoKBd2N/y61SbI6OsEHTvNJhY0u0nshGlDljovhaJph4KI9ZcPAx4bbFkrZ7AL2TPnCUds0GEwx3JNszuczMU/gYZSi3oNSzfRiHfWtUcZuoZZDDba9s+8bzlJ+Kp3eC/M888Vnb/i5z16R4ndZW7mzA3KTDXtye7oT5z3M8qqxWWi92pNFVXa4G38OEU0MnuDgelvIKRIYxBS1ag1Hx3O6fMY0P9L70HrUG8Mnat3Z9u0ef1btYhzAVusdlR9VnQX+WGWGE0gZguTuRksGp4OwOmeON8gh0t0BRzjDTiqchwJns8Rdl7NSiy+nE9M0cXl8Ad5z++qH9v5Gu1qHoesas30ItFrYy4YblRgGrS44KhWl8JzPL3RI0ZnyGW9twa0pbam2alLsTmu7QOqYGEOHYTIdgADZpuCR7olBVmU5TFVt13uXxNyoameszBjqq4AjYMXrMApaUa7LjWVdxXaVQVuUNDRG53zKpKDaejxShHpHzvEuq/6jXh+HYtAdo+7RjafkXe2Ykm8GonajOmiWIz9sjzoSftQ1Nxg0ehdSLJWK5JoSHA16VyiFsIOKCxXfOzl4ziFYFp+SZ8UUnOh4u9kr+Cheeb8JMCo7P3l6otVCjt6kv/ZktPqzGDyl7GzbzvnyyHx+oMbG2FZiUGx4zhG3qRrM5mwh34YYY7QaFj/dxxA9aHXVx2FfzAIbjgjrMWimmFNOg0beNmQqUuR25sWLN6rW9jIkNacQkoHoPdPqAHpbex/EZA6+3u0G92a2GcIonFfSMY6Yojkfj8IWaeeD9+zmFDxeA9GrY8BR2DJZ2Ms0TZzmSTmCfWM0R/IDn4MUff1wV+o9ED7XmaJ5VJxKcCWi0XheBzybLyWmZO3IiOMrx76NsQweH5XbqLXH4uU6DER3xhiZxyD/VHlt71J/dDSVlFLkOo2NlANzmnQveH9ncsYYjNZJSFTUWmfbdW/0prAb0b2dddkgR7zPtv7ogPWOe13dH/X6KA4BhyMFxT7HmEjRc1s33r6/EnyXWIdq4RbVElcmnN3YdDm4jhps74N+6W4UUQfnB304MQYGrA037OL2+HQmuE4OjmmaTasv9mDfNrW+BJ30oxdKWdmWBVDmQA6O3TrmY1BxyhGW6UA4QBDtte0bR6BETOKYHy8Tp6y026YrwRSDtiF3BYF4AxY95e6ac6YbP2i44HTztN6lssP0FfZ+e6cnvSqwO4PI60+/5PHFG5yT8CQ6f4/MHi5S/UwdScJac/d5C7g8YsP1PexiG6LflnUjRmE34ehYMLpM68tgK1W5/SZ2gp9SNo7BulXePd84zxPns8Rkx9fW9YYzsRdYMZMbtCF1J1GR6oOG9wnvB3QVuSh3cdbEUytl3+hEOwBuKiU5JMjhqBa34BEXbAevmsC8t+tTN7z3npwSrXqLCWuQkwGiHzID1m0lj4nhCmnqHJVso0vkBkchrVaLEMV0LLuo2eg8U3a05tTabcDiwFOHu4OQHzKv/u7XR3EIDAYhRqaUmM9KYs15I4bAut7UTtuGAVfekNgmYCVlGJnalETDvS1YAN3h/NL3kYX0QMU/JMZYP5yFaNTecU3fQ2m0nexm8pQ/yF4Zlgw77mUiR15eshPb24UW0sTl5WsCOrG3deV2fU8pO9N8Yp7PfPr6Fa9ePPD9twtgeXMcwOD4QOUdFJfz5CTl3PBOB6E10mhy+nCjrnthZ5jXXaGec4rU4bhtOzG/4PHVN6m9cw7WW2CKweETPg4uD29YX3zK7f3vWxqxBFi9SW7dLAXHm3zWBeEZ3oDDFI+SVq0ZOIloGDo8jpCTw/DTDQG/ripWPZ0mhgvsBZa9EzcZY+bhCK3AUDTYYLDthauZcQ7jTYyBOSe8Scs11isFaS870cslUkqhbSr8SFEHVzwoYDRNHRX3xxM+eE9Ik4WF6GYbBkKO0eh0o0o9OZ3BrjHGwY401vUZP9q9caiNccc3Wu926FlTldNDzXVnxbgCJ33Q79CGDgM3nIHKw67+P/r1URwCzkbUHKTNrrUpwaUVC5XDwkStVdcipQ7RjN0hBhD2eyqP9mr7V5Z4E7x2NobFatsM3fpghKg0l20zv7230c4y5XIGHGVfWZaFVgrbtir70DlO8wOX85mcJ66LcglO88w0n4k5waiEoQrrcZoZDIlGqmUoSB10/7yGqRqPj/AYE4ddEEsZTFGR3MOUZthF0vowYQnMOd2RYhB3Xltn78IrPnvzBY8PjxbR5YkeXAh0Ao3AdAq8ehO4Pv2Q29MfyLtu420plsjjj5veWRKQft7WOt6r23GY8s47jendCbWdsmTR3Tt80H7ugHVzKovJMhBJiiy/wFYshNMHou8Ee9IdY7wLWdNA2Yi2G7euvoGgBFtqH2zFaGh7dvRucWC79BVKG2qGq1gYacfWg37n5B18mJyGcgalDRkckJoOAeVhtibcYC2Nbb/pY29qtQoh3PUAzQ73bu7Bo2otBxjWIEVvVDOVaXU82CRJV3QtfeTrgHbgiAsTe2k8395bRJMSYYbtx3ebavDsTifeXirOzDrOCTntHQjOkmAtT65uYgGwHbtXA6rkfVdnm3jhMCXm05kUE6WopxCgrAvzacZPiegfuS0Lte7EeCblmcvDC+bT+e4d770ZLTcYbWU05cl3A/PGGFKFjWA3MCYr5QPXe3+XjObAmToPrvvOJXPHLkY3e6ldCA5VaQmoO04YvZe1W79gzMQ0M5wqwT0aM0NwluKTSc7zMj3w7uVnvP9RZvTC1qrJUgUUBKddWTJu/fzJjCuHNr6Ynv00RVI00Q/iyKtJew9qa993cgq6YRG9V1tltMi+b6LLmqa0aLhG9Hpvjh7BcE7UbNHjXqGqrXdyzPgQDDvp9KGm5vs02HRgtiLBkR/2Pbx+r9YFjkrWnHE0+thMgRgs719YljdVoj46TZpGpuhzQlbmYeBjShZpb8DwMFGQmJlh74f9ru4oRPXsu36mGFVC6hwyWtkk9jMiBj+SQ8A5vvjsDSkk1m3jdJrZnKO0GwxIIZGCySKdUTTGB3frenOMe7HmcEK9dTV2S5vRB+ucykR6qzpYjt0TjEf27A1uy8ZpNq4fPUkYXZ4A47YZH2qqp2lW8+sY7Gsx3MFqo9ebBUaqU68jnvhgG3I+8XzdPsRnWY7AkW93pPZgYh/vPTmfCSNT+o05eFtb9D1jiuAGpSjUk4Hx1KL2gunM67Cqc1uflOuvJp/h7InqVJbh2iBP2RJ24521CM7fQSfnPbggqmoMcgo0D6UYQJd1k4MUdrVpxtmq8ehOcfL4wDxl6+k7shcrpayUCG6x0tYkX8IB3rqQwOjYcIScWBeA9wJFK4MSI8lp3cIOKG8HSHDao1vXYnk0Q22WGYmHuovKO1tuogxNArG1ylmM+J2/V2/Csu2w7vfP0HtvU5VuYMaQ9yB4Yk6GJ5nS0MJ0D3Dy0KwwjhGyyzaeJkopnE4TfcC+y3rcfqrm/f/7+igOgUPNtbed0UVp3cZRoqE8OUk4bS8bzT44oEND4+3oHR+TgXLCGQZRVB0WXx2U8hNCvNNww3h9RgY8e3O8fV64rTs5QnBwOqmpSD52kyu3Rp6y3Ilmf5bTq9IseagPqGWjFKXcKANRk0dphZzzPfCiDadeuxEkPGmN4QwYQmUcztaOn/+FX8WNzvb2u4zbd4ih3SlQfmqFaLXfDyQJZ6zGbRwU4zAdgyOFIWPN0L7v7BAI0bOVlbY9AU2uSJsuc4z39qbNrjPvFe0lLMPda7F99xaoWbmtm6LMTccQvL/bXr2BXzGG+yS37zvbtpKDMxqy3qXF8zRxPsn23LuSpg6/w7o3yr6Z9iExTRlYpTg9evqCRn2HQmuCG8wp0jwMPK0Wyq7k616ULeiNijtlxc21dRPrMiAm3bTe6ff1OJZSeF6MNgabtLTe1do4avA4ItWGYsw6Ul2mIGxiGF3svKOXdgcup5x48XghxsRtUSCGZOMCh9f1o88YlGDn6f2VVjdev3rF4/mkU97cd95UeYds1nsp8bC9XsWQBx3TaNX2IAe2KRKcbujDaaELrNPaZqOgkO6G4zTP+BhY9oU5CgeQmlC1VsMQ197Uae8FPDAYWmX2DXImhiSAMIS72aM7Rxse7yMpnWhO5RG1O13Iw8nk4j6MqACOwHy68Md//U/yG7/5j0EffP/3/gbf+xsLdfkB1Rx+rVnE9nFQ2jQRopVW2lgsRN9SmXoDp/w9K2eUj63t1M2zrc9c3//YfkfLfnC6+I6yUm/BrsEQ7lKbRn53rCOD0bA6Mj2d9l1syumcyCbfPtise76gmYUU1zZxOZ81jQEdbwzIANcZfVdjlY9MSW1JtQ7KvrGVja3ICBY9vHyYuJwuonGPduayMUVJy9uI9NFYN1spTYvvvJB4j1YJsRidlCx9ejTcUFqRQyDrvlc9lQ/JqtHiMTo7sLXKOhd1IFRRwLUOtr0Ro2fOOnDHKPRu7dSW1hRj5LYWBlVpQnUXIJuiQknK8rX33kdxCIwBtcK+Fx4fX3A6nWn7xhKCRBTeM1qjt1071/B4n4CG843g411pJUDKqDm7mA66xhkt54ZujmqahGhAVRsWyumTFaJ6aIW9rNRaeLg8MgZsu1SB5UhvCZ6pVx1CLhBLoBTd5DEGcjozxmDfdSH5fCLlzPr8JHrLeSG6YSZljDMXWHiHAvROkWLi1atPuZwf2NaVaToR0oXtKnUezpOjVITNEGphp+KNe+9K9rGuwdYb6+0dZV/w7kKMnlL0PgwDaTsL+/LM9f1XgLoEj/LO4wZw3nM+5buQCEPDHZiiTevMUbF2mibU2NSYsjFDOVPtoGytU4rENWMMtf74QIxKGMopMpxjq/EeKjJ6JaWsOPQxiCPiQ8IzWILA58PIddsFEp+mi6rNnTPqOdk146x5WrFiOSNZcRPZFpAO4un5au8UnM5njfR1u8uXj56DOy2NgL0jglwM7rh/TzE/HyY0VbHJlbmXblOnpiHnFC9eSpVALDgulwcGhd4K86zp+Xmt95LSP+r1URwCzsue+zolHs5nPI19SGbpOgqybNqz/ZD01zOkCkwTHWj7et+1Ys4EF81ZJpZAh4bafYaxBQyFdoaYRc14U5s5jfZyaMnxlVOiD1jXhdrKndceQ39XbcbZRyzAQXkAy20hpsD58kCc/F18E1LA95ltE0X1fNtxYQa2O8LMMdQbXXTXexw5hJaNcNwgOQUp/ew9+rDjdnPVaVffdlFgxwV4u71nWRbcIY32O5QGbVgXI3z19CwwlXH3MOhn0eganG7cMSQ17kbR6vdwZnr5EDUWo8CvfXdS1IXAslVwfBDJjKH3Z9spZWfdC5cqz77iyxN5ms2zICXiEcYBEGxddkx6GqaAR7jMsgp4fnp+z6lmtS3nGefUJAxQTAOQYrI6emNZeldqVa3clhvQOc2zymDHISHig0gIYQMxHq7CoAllwBTkXehNJ84B2MbgLcfB4u6caMPeBr5J9+G8qvgOQmn0xlTlgNy2SkqB03xmrdKjfN3r4zgEnON8Pin/b1nYloV132kDJq/T7igg1YVttc7GJ9d9M/08xDQRvAoXgjsYNwMO72OCxrTgOtGsxsHGs6N5FixUpBXtYyFIv76rLrsOKeRsGqPshbdv390vwJSyLrzeWZ4lDro8vLhTPGXdCDgezhNvnwrrViQprUpCPrQBP0UQ3jnE4TxxnsmnE70VfvydF/ivAkc+33BwW3bTDABdCTMK2OjU0iyGzYxGdeP5+Zl16+QpkKYTLlZKc4Tu2IrCSEdb7hzDGEoRCl4qT+8dU9I6hbkbe6v3TIHe9f7LTTdZQvQgT0qAAk+wYigX9DXLXjh7Lzpz3ymlyFjWJ9unPecs9+GRWO1ZjZoUJlBaJcbAlDzeB5oFzTycErV1tm3RkzsEStZ/55wZtbIX9V/46YhRV+Rdsomm9o73mhJjDNAahESI6R60wlCXpHMWfYfA1NK0ysSopCJRrGqkrk6/08DjYxDYadfV8b73MRR1ZmIgxSNUvnp6FtXpHE/PC+vWqMN9/DVkoODOVnduN2mjx4DaB6E3ppDwNn52a57pXQr6w9CRghDxFJN8AeaVP/hsJ5TLgKBoLrOmyKvxQanVTWTjTKcfQsKhfAJ6JUdvgSUDXNQe2hpbKbgqNZz3nhhWyY+dAlH2fWfaFnzMDB9VilF2G80767qwLldToB15AoqrJh55f5E3n/0cn376BafLI8FrL3z52bd494P/m319S6mNrVSjsYAxyFG3bh9YWi33TIHWOy460vQgjcSQstK5IHR+RPbWKO2DXbiPwbKoSTnndJ8MjthrSZUhOIlXYgjUoci13eLGRIPp73TOhF3OhC8hyuNlIR19aOS93Ram6Mk58MiDzDshqCexqbG3Nf1dT++vbNtOiOb3iDKjHc5Oj9J/dz/u3gD6YJon9h2T9FbhUbbKVFtr5hTJIRFzMDm4GqhqqzhTaB6UoFYxc/3ZGO+HY0oOvH5PaV6klXA4aB37cQQgOtHYYlX0+RyJ06WJASMpncg7R9mrQE9TURrJ9LWvj+IQGL2zrRvvrzduy2YX4lBvW9WBkFMSvTT6/RBwwXhmHC5K6eaNu279UBA6Uw4KJ9CFIFFRt5SZNnYYySYBmZN8iCqNQGO2p+OcRuopJBNzKAW21CvUaqOf9Pq9wboX/R1BOXx130k+EqeAI7ED67ZzvS2smwpM5IW3qcRBTjOPL14rqqwPvvzmL/HNb36b03wGOnXfePjkS3w8M/qP5cfPwZxsql1zJlHdazVE3iS2XiPn6y9+jcvLb1Bb4Om5WNRaF1DqAmN4YjwxXKb1m56AznTz42jTHbaOySiTc1Jyrx3Gx58Pzt8bihQnF+0zFagp3a9u5FoGLqpk87CWl9bY9sJtsVDRu5TaPA7DU7u1Qd2uqoB7vjGfTjjnOZ2yxnJ7snrncVE3cmmFvg4rbtXPES0ItbSuItEOMQwSesLHkCSH7s3wvrsIQBOn4VlTFkY0sCCRoAlocNTBb9SyMxDONE2RYxYMxgAodxGqMERG76RY9X1tlY33HIpBsvQt781t+zWvj+IQ6L1zW1e2XVHT3eROznvWvVKvC+dZ4Y45eBttBoxj5HLGUQ9qK/LPbzfGaJzmM6fTBQ9MU6YPWK5PNnZjCjaPGzqNR5eScDNFVw6RnC7gDh++J08TPkZKqdQmX8G27rTa+Pzzb+BwvH3/FWN0LucHYjqRoyefJkKKgCHHPgrdeHpHq/ICeGuVcW7HO8+Ll5/yj/+T/xSv33zKbVm4XF7w8PhS2vJaaKXikW6gxcTojXVXnZp6B4dlzv9UfqE7yjsH+fIFLz/7RU7nB1wI1NpZ1mGRYY3hig6TfGF6/JJaroyyM5D6LkZ9TemVOcsurcm43fUNh8mr2yEhKm7cAcvWlVxcW5ebsPc79eWcY9sKS9w5bYV9ntlLY992Vqshm9zAu8aRWXi9bSzrRkqRfdu4rYqxn08nbjcp8PI0kUwBCjo8nImxOjogTymRU1QSUMMYAWzVKKQxTOac7q5Tx7izLwoRlc4gx4kYEr1XYxcU0XboLVKMii1HGoXJNCeMca84E0YArcs9yujEywXnA3sbrNtO9IrTUwCNaQoGlP71o8BHcQgMNGq1IWPLUSOt5ilRQO9vK6ccSOeLntJ9/zDye6XP9FpZ942jniunE48PD8ynR+q24LyjmYvsSNNNFiElBPZwZCmbrx90Xm/aw4OoyrLvsKs6q++b0mfNQ1+2RRe1hUc8nmdCUvZ/SAlSkLFpNynoaOQUyfGIhR4WQto5TTO/8PO/yK/8yq/x+tUb2xPlyX9+vrKvMjGVXam3tUsMpJ3R34G8Q8Ai7792aYbUia8+/ZLHxzfENOGojO6llPOBbvHqe+3k0wMvv/h16vqOuP4AkCTa+8CyLqybburd0HxMkOSdxZgZzRls943xGFVFN7Ym1qIaQObdvfyBMRxrUblpuG2kmAnPiyU9bVweTgqBTYkO/O7vfY+nt1/xi9/6nFevX9D7UCpRiDwvG09Pz8ynzsuX6c5mdHuQpOiIDpJXavWcM60VoN5DQ2ptrF0r15Sd+idsR9fUNahtk+S3N0arMpcN7fwezzBruzQgmj5CiAyn/ADHIKVAdMdDzzEI+Jho3ZlrthJiIvhICOC62ITgHZtDYHXTwRs/ekxgWBjGviGQI96biB5OJ07nR56uz+zrQmmV0I7z20Ik2gehzJQC0zyT8kyvg946y/Ud3cb1NqQCkxkpm/4AQDuuc53kEbiIx9uFOXohnyfpCAbSeQddXBHUTBsjraiZJuVHYpwZvUpGyqBeN4WSukBrwyinjVpWPdXrTikqZ00x8a1v/RK/9Y/+Kd68fq0Unq4dcBi/P3pjW496tJm3rQOeeZrVOrOrYGSMQXaJmCJbKYxabSQ/8fjJl5qQUKNuNPqwmxKtd7X2kibOr77JevsN1h/cCGMlepOxTpnRK8HL0bZvSj8epv5zTtXc3n2IEXf9kItjgNiBKfDB/mzf23vHVlQq43y44w+tzYB6BKv5Dfrg/23vXGIsW6+7/lvfYz9O1anqvg9fjDGxHRmhTCBWhDyIMgTiiWGWERkgMSESDBgYZZIpkWCAhJBARAoIkQkgMkHiISRGBAJyHIfIiYFE4Jj78E139TlnP74Xg7X26ca6fSFXvl0l3bOkUlefqu766jt7f3s9/g+WZeHN1/f0XSDESDXa7ulw4vfeecKaM288cuyvr5VcFEUtx6pwmnWMeLsfVYEKiLFnHyJryhznVfs/TtcQQjy7HetEAyUCeU9rkZYqBSUmtVroO/1+RfEV62MFg68HvGkHbJmrtwNCmaQoytUUhXJazrDn4MSan51Nw9QnY02L4URefvs9iEPAOeHmelD9PmuQzfOkHdOho6RZTSpC4Hg6Iq0HnqupbniaVrVk2BhhjUrJSljRBhaAZ1nUvqqWbOIcejGqpHRUmGpTQM+SK2k5cXt9xdV+zzQnptPCJmAZfFA2ovUtfBe4ur6mOQ9E5uMdaZ2pZqCqECWhZAXmzMvMO+99j7tnByWVKM6Hvtvz2T/6x/j0H/4swWmKrD2SQlpX0rKe/6w5U5uSU1RLAXUlCtrQzDmTSkaSO0uPIULsBpwLrLnR2/SqgI79StEDsELsB9bSCKFnvHmL0/duqPNBSVeo0EYpVcszQUVWG9ZMVTSlGL4ghGAYCP2aEnt0Vyj1rIvQ7P31PqjcVm08O0wGfNHfvZSdPVkbfR/pu0hp8MajK67GCAhPnx5RbINiAx7dXDMMHa1B8EI/9IohMVzGtKzkpKjDJa3aOjA2ofIPjLbtBIdCwUW0abvMJ1KatffhAzHqdZowezkrN5TMo6Arqj6oWm1UUfv1PgaTMNN1+fNkADCehQqZBEorSFEErY8ecVEPothRt+nQBkt+STyIQ0DEMVxf80bsqG+/wzwdAQVJHKeJvuics+8iy6zW5N5krXRevSnTZnqnN5kvKtXsnVdV2eoZrnZmOhF5+v5EKZpGFxeMwFOfz7VxpNJMhNPTdypCsdvvmNIT5lMiSaV5UUZfydAyfdcTe2fjt0rXBYIfSWmhlMJpPvHk7qhMsM6TskJi1edQoBWbfMDhcMfdkyeqIydC2gwrs0qbKaiksrn2KjtOVBXXxlPbIyCngnfWfLMnfGlCKip31dARW8BBMyJSWvFRYbFrLTgqod/T7V7n6ZPfpYvaeC11u8Abpeihnoty9v0Z9qyEFgVWPn/fWtWphWoR6MW6rip7Ls7RdaZVIMrdP00zrRa64EhrsNpX7edDF2mrmsYua+bZM/VJ6INBsr3jegzc7HeI1+up78QMVLU8GaIn5ciSilqHeYWti6ihaRcUnqzpv/u/dBrULk6BS1BVv6AfgQ0zUBDxNtOv5wNEO/46ieg7XtCBtH6AlXGx37EsM/P8RLMnpxMFH/RgKkUURYvS41vD5Po8zj90PYGmqiwhRl5/7RYh8+x0oDRPCAM+BnZDz9B3LH1kmo7GmdZTdMmL3ji1QcB04RREs6zZ1IAL67wQO6HrO25e/xSn45F5ns/AFgz0s6RCxnFzvefmeqcgGKfOu7F3vPb669z5wPHuXdRqWtPZZVnIKdHnjPcLy7owTTO1VaZ5JpdErVpaDJ3qGnRB3WbefLzn8c0NRIVQ39zsdfyUFT0IRnEt2TrRClDQZps8dxt26kWwZjXwmCbtgTT0ZgdR04rWeHY8ktaJZU2qtksjVYczAZeUGy5UhdE6OBgwpoU94iKp6EjOGaOtipw5GX3nzlZY2n/RwycVgzbXBk6sg93O1Fphq18VG56zyYx7zrJkNfpztqKQYqPxNmEYemIYmVfFym/6E6oCrH6M/TDgXCTGjmhP7CamvJwmhhAQl00qrrHScAzDYyYAABb9SURBVE4h4V2MKneHrlvVrRrNQexGlfrOiwnAyZnUlLPulfICitK/7dqHcpYOb6gKdK6VGnuGQUlKiMOFDlmTgeAqQTZiUKOaV0VOqx0aCmdnk1N/6NOBUgpP33uXvtdf+q0336B/EribVkQ8uyFysx/pvMPVlZwjNVdKaQYzNZsuWfTkoxJCz5qVMyCWQtWqXWJxO/r9DS4O1KdPWaZnLOvKs8OBw+FIqY233nqD25traHC4e0IqK+uimgLduONRu6KuB3JSCqnbSDTrYhc/HI4HdahtEKNaQTnB0r1A8Lr9pVZCd8UX//iPUv01qTY+97kv8MZbn+Fmf4v3UWvqoiNSERXKWOticOpKdI1k6ab3jmrS4JvM9zYbyFUbRwaR0Ebiqg5LWRydN2kucXhDuC1WxjgDUImPiI+UtFovRs6Nv0YzQcxq4zBD7YnW+tWgW5VqzVdrZK0KkroeOpsOWcptGVIMKoeuhYIKtlZzViq23070UFSZ7krfd/pvfKTmhXWZGcZRFYXiaHgO3R9vMFx8b+m8rimXEzEk7SH5QGf1e2tbQ9tGww0lXLmoMudn+rbTpqvzkGZqUldttoyrbFB3d8ZDLKlQ24STE9e7K66vdgpgqnfQKsMwkBblAqRSaGXVKYUIiO5JCM40MVRSvrYH3hhsrXI63RH8LbWq7PXtzZ6uV3XXq+sdsYusp5mcTWYMpcKmsum7Feq60ruOq6trrm4fs8wTwQvT6cTxeMC3QoweXAfDY/LxbVrRpto8L8zzzDBG9lfXvPHoEbvdFXiPd5WcMrHvlOTRCmHs2d/ecnj6BMwopBSFt276+6VkhqGnNp0e+GClSTM5cO/196mVx699ii//0J9guHkT5yPj1TUhduTcOJ1mUlnYCFEuBFpa9SZqTZ8yadJ1iFN5au/xtdJ16ohULAvpnWNZE+lcPhRqWahZ9QSqMTlLVak3fULrTVCqOuOU2liL4M7TCi17sKdiLus5pT2/x8jZF1GsjBEr49ZUSSlxezWc694Xb3Ztmmn2YNwvHdElvbBraXTdwKPHj6mt8d77d7z//jOGALvda4TdLb//3jtEH+li95xDQqO0SlpUoEadoRUzUEo+IyNBdQKcAZxatTFgU8uvbKCoJptMvFOCjxNlvqNy7tbpYE0ztSQVNVl1EhWCP8uSp1xsTOnJh8Zhmhn6jptxRz8MKkA6OPM80J6BN9KT2pEvSK3UtOCc4mfkoTcGQRljy5pwISNxxIXIaGOq1hwlV1JSHEErBSeRWpMKj9RGWk/0rnF1dc24uyJQiNfXNCD0I/14xfzsKbUYzPjJ91iPT4nRM4wjzZ7Wwatm/Kb86lphuLqhrCoRWdesPHHvDcqpOga1ZGveZVJJxrDriFEbcxt8NHbbRajOMKu5+fTdI/rrK7rrPXEY8VHtwE/VuATGJtq60DlrDV+yiqronFzIVXsKIThy3tiHdvM4RQE653EpgRcOxwP97qjOPM7RUa1ub9Ac2fQQa9nYkbNOGCSQ6gsz7GbahEH9DrPx11VCXPUXUql6CBswSCwzKaUwGn++2pgzGgdgG7u11liS4uLdorX6rqkfxPV+z243KiZk6Hl8c8v7j/ekZTL58kUnOUFNbmsttKJaC86wykpULHinZjetNmJw1OpVlNQpfmNz/VUnpU3pJ1FyVlMb5xXtibBpQ4vRXktVTcetpCu5UpuQmyeviSGaKrTfMArR0KOaYb6fE8MyMfTKU2ji8C5Qi/40vc7EiEs66hZJ+p6I/4C7TuNBHAJiNNGcE4ejerSJAKYmsyn05ALTvLKkhdwWYwHqSd11PV105Oa4u7ujjx3D/pbYdfjgGK/3SMmcDk9Yn75HLe/gQ2S3v1Fz0dcfUestyzSxLrORQ2a94asZUgY9lEJRz/i2zlC1IdjIZwBH51WTToUlKtFEJ1TPUKHIpVV9kpj5iZSJ9XSHiztNnWP3fI7cVOHC+YCrlVSyMe303+a0Ii3bk1JTbRp0XUfOCWo2UxOtpXdDT99FcvO0snA6nRiGXuG7Rf0XaZmae03dWyIvi1FUEzXrjDobYaiLJuOGcD0OpJw5nKYzSjAXrY/jpnUlz4VQ15TP5h9rUrScOHdGGjZx5qGgI6BWFfkYk45us00cYnQK/25N38/XHpHyDTnNhlMQLQ1NQp1SDGKufRLIZwQioHdGyTTrv0hRUFaxLKaxkcGMaNa00Gn5eRe/GdEreGsGWuawTVMUd6BCMs0EY7to2hliDlOhZ9NNXFLmOB05TScV5I3ReiOVtWQcxWzoVCk7540pWhVW/5J4EIeAdonNwadUhaBWWJejSo7TCDYHXCwN7vvAm48f4eL4XPMtz5SkRo6CIz29Y9wNDOOOWtVuzHcjy7Mn+BjZ3T7G4chrJo49LnY43+HEhChTxkklREfL+iT3QXULtsPCGburZJXbil1Hyco8PLvvigOn82xVgVWsQIxKa85dIdTCND9l6a7AB8Q4BrVavYnKeUvQEWbsvPYg5hMlLzqiM/jq0PeINQCXpZBatVpWwTvBayc+tExb3+N0GFjWnuura1o9aGc6RJZV1Y7UvTgz58Y8HVgPv0dNB6JXnEYxfnz0TpWUgb7rjWq9sbgazrvnaknoE1Vvkuez8ayIKJUpo22AED0Ma8KFrYZXWfZnxxNvv/MeLSfkzYKXR8RwpboHVWW+nDiueiUdheGa6XTgeHim1NzTTF4Xui5Qxyti6DZEMVp7aclWyDSjo1dr7lVTEc7mtOxcf5Z+22zZxb5Pm7jVSked2+eiUvmlNrrOM44j+/EK7wNpnU0UV1Gg0VdKzqy1seREnRfzPFTgvPfavA3o/ioMWkuWUDIlPHAqMcZ+y6XQpOKS4bWrgnscDfFCa6opt7/e8ejx61zfvA7eU5aJdc3kxbPMkKxWamUlpR7xjcPvv2vEGQVe98NIP460rJRcXEAk4jvHKM1uPoGyKg2zAQg1zSynE9N0opiWnBpBQh8H5jUzGa3ZG1EGh1lHO8SrVDo1o008swYjU9anlPkW34/UKeC6gYYSo7wLRozQp47JVUDN5OVADI0WOybLYlpTO6vohdL0xtGbX8dTnfeUVlnzHdPT36W/epOjiWR45xgGrdObyXDNy0ReDkx3b5NPb+NtPOtNHVgM2baxBcFm/KpHrmk51syzp3yt9bnUeNvq73aW+hbBUHYawSTeHN70AwxQlBPT6cTTu6Dd9BCprZDSwvF0xPvA1ajj4bwuSKt4HzkcDuScGAfVjjge76wsDAzjFeIiIp6cF2pV81Un6oSsoCdlQtZSiLFDvDuXa2IgqNoaLStFupaikx7jcqhvQFUmatXyyPmo6M6o7ISNz6L3tmkfFn1v1HVIUa+1NJZSmZYNa2FNc+tgBnngjUERlWNa1pXSil40UZs4azPjBRdY0kJwwhAjwWm67aSpkm9teBkVFbdM5NwYxx0hRvJ64u5wQCgMw3AGcSyHA9FszVptkFbyPJOTzqL7cYfrOlpOpHXFeUWfLfOJaTqeASR97CwN87RVb27nFTOv0GCHMz376BvSCpyd2RQU43whrM9YlgPrcoXDE/Dq0Ik+GZWYpOVPKcUo1AnKEVqi6yKpZBM0VahuyuncZW9N3Xx0ZGgNr5ZgfZclH5kO11TXE2LH1A1AI0hlPt0xnZ7i28y6TgSnKsCKTVAAzVXfcb3rVTZctAmYSj1LwuWcCc3syOzG8UY5VtCRNksV0t3A1ItEhOrUW8Ibv6DUxtBrbV/NVSilQE4r0zzR94oF0YOwIiRC0Hl/SwtTWgkO3ny81xvOQEDPTgeWeWaaVAcwdAOylTS5sK4zgkO8AF4nHXagee9x53GtTmxq24a7+nQ+g4Wska1NPeVxFKNLz+tM9NE0IDJO8tnbQM1INcNY1qRgrF7VjDaBkdWQsVj2WIHmlDX5sngQhwDoPPd2vzdXXmczYuXBt6rNlBAi3ikaqzRYpxN963Bdh/faJY4x4Bhog2PY30LLcFAASQhBL5684ltROm/JjFd7KMrdrDWznI6WRjb6UXXivXO0kijrrPPm0NHFzmq8alqC2bj1KutUW8VtBCUDw5SyaLPLdVTU3PJ2uGbcQ3h65LtP78jzHu86QqyUlFSQw9h1rVVqTupmNE/azS8T0YtxzCG3hhdYbSqhnXh9OqSsrLLg1BSUVhm7wJomfJnJqeJLR12c8RkazBPBtAF2fbRGoHa0uy6Si17ga9J9LjnTR2HoFXV5miaagyWtmuY3wKmLtBhvw5s+ADZiDEGzL31Jqb6ba9CyLByOahk39IGh86zjQMpKolnmRacK4hjHkV0fGfpRJzF5JTrtlwRnCV7DqNeNNZtepDRKmkwUNpimYGXzO1xsRN1qIcRgh0W1EkEPBzgr2emDIDiDQmdUzLRBE2LQG7iUzDIdqTFaQ7zQmiOX1YBUzRyGszY1EevL6MHrg2YNJWsGopoL3h6WDxwxiKgIpHMej4pmlJKpadVOsahIaOw7nes6hVSG4JHgaDmd3Wud1xtQWoU8Ic7T9R2Pbm9Z1kWloFslxBFx3nT59CKurVAMjy3OsS6Lps79aH+f1JMPUVimKKpQoclYdzxSVe5ByUkNq4Wx5pn2+XJWr7ouRkLs8LXRu0IsT5imPX68odm6nKHSNiy74sgNLJMXaj7RB8+SNNU34V76qI7EpYFHewGpVlKqSITWtKGn2YEy20L01kRzhkoTak3m16A4doWhqspvHxzRDC+XNbGsiuALdoA0sad32QAxzUQ+Ojavgs0vYTPvDCaqWQ1k5EWdbUspiBGh1C1a4b/rsrIsi4qRTitLv7AbBsah176FvccxBGZRw9FlTkjf2/Wk5cz12DNGNR3p+o7WKnfHhWR24F3UIkzn+JsT0Xa4KpI12FP8nAMYp0U/1ybdshZS1jG3c4pGFbRxnNbVSgR9IG1+GeIcqRpbMOnFlBuUpaj3hDcAdTOGaGu4pmhCTHjkZfEwDoHGGTtdqqH71omSV3rv1DknRvb7Pd4FlpTNrSVsbhGmTgNp0Rt9XRLDqmAjoRH7gWWZmE5H+n6g4c8XmfPqZSBhoDf117SuIJG0Fmo9sayKAXBGYIGtQVQpOLMqE4ZBiUSw0WYtszGBUi8Ocap91/UDyzIznY4gqrhzs5uYp4PKn5dMMxBIKibrhUppaVbTcZJC9FiXuljtWbVjz3NZr5yrAWr0902l4Gxur6CaeqZKhxDPRKvj6aRqvXYQ51JZk9Kwo9ebJ7iNsVZMFquZzLcjmjimd57JQCxbmp+zrtcbq9F71T9wBjDY8BS5FGPL6f/TxUjf9WcpMb1eZpblRBp6lpw5rSs+Rrp+JK0z66rCH7UWy1wqss5UHxAXzatR5dcVtKOSckPnYW0UFOSlMN8MaUHQQ6ukxNxM3j4ognHTHZS6NUHViXpOmeOcqEUbzkOvLBfvnBraWqZRyjZaBpw2kOe1qSBtMREcGqXC2grBBFU1npcYaUkMfXw+mfmAeBCHQANTSFGo6WmaKenEftR0dVobY+hpTfDDyG5UjHyjUk7qWqzjvh0tK54dtzItC61OOrKLAUT5+zFGQhfPF7ce0QUoOO/ZjSOLE/pxZJ5mDsdnFAmE8UYltOajNt9KRlpRfHxQ7fv9zR5wHI8naklGk9UneteNmiM06IcRrOkWYq/KQstMWg601CHNnp55peAoJr8Voyf2wZyahVJVozCvxzMCruTMmq3edIohr2RK1tm9sybeJg+Wcyb6QIjhbKyRq2rWp1zMw0Ep1ZtAaUPHd/OatOln/IFN3NT7cPZd0As24L1xHIIwLzpF2PwHnQh97PV/bjx3S2ITijXfCQO/5Fp4djgS/BX0g1Gfq+E19Dm85JXptKhEXD8goBTckolBWXlrXTktR5IpEiFC54Xd2JNFYcvamDOuo/PgVR/B2dpK1h5AzsX6O+6M66iGg8h5c9IuZhDTyGvimAprl+mCBxp9ZyzJks1tWKxv0JgXLQU2C71NfAbTh8gmrS/Gwmqt4cWfFZ5eFg/iEBDR1GadJmpT7LwT0bQrr4zjDUM/6EVcK9JH8mnm/XffYU0zn/rUH6Lb7ZAWaG1GfGB/dcNwOLAuJ32immho6DpCvyOlcp5ApVmzjporoR8JXWS4vqHkxOl0IlXh0ac/yzCO1FqYv/ddvFso2VsZo6d5Nwx0fY+KVXbkNCmFOeqsOwRPrY6WFzWZ0DxaSwmE1ITTPJPWZ4S80HJSWUyvrsaauirHItrpLlSmZSZqOqVZU1D15SUrtqEJpJI0nRaVrI7emVyWt/fAIaXqXB4V2VytJ1OKsHkuIMqzRzjPoXPW2td7tR3LRZGJ18OOWgqnebFywxuIxfQCRJmYrWqZsOaklt7izjDebOq6avCicuxCs8lAYU3ZyEHPZc6cGCQ3q1BprUlNXYKnjz3BBeZl1pq+VU7TzNh3uGgHo1fQjk6sqk099ABRxp47KynXpuNUcWJ8Lc28NvPcWqGYJfmSTXnYyg+VVcPKAyUndfE5YKpUowCLPfFXk75zNjGxLEpk0xo0xWI0M3z+oGjIh7AIZWN43WeIyLvAEXjvvtfyQrzBZT0fFg9tPfDw1vTQ1vNDrbU3v//FB3EIAIjIr7bWfuy+17HFZT0fHg9tPfDw1vTQ1vOyePnw8BKXuMQnIi6HwCUu8QmPh3QI/L37XsD3xWU9Hx4PbT3w8Nb00NbzgfFgegKXuMQl7iceUiZwiUtc4h7i3g8BEfmzIvItEfm2iHztntbwOyLy6yLydRH5VXvtNRH51yLy2/bn4495Db8gIu+IyDdfeO0D1yAaf9v27Bsi8qVXtJ6fE5Hv2D59XUS+8sLX/rqt51si8mc+hvV8VkT+nYj8VxH5DRH5K/b6vezRh6zn3vboI8fm/nofHyik/b8BXwA64NeAH7mHdfwO8Mb3vfbzwNfs868Bf+NjXsNPAF8Cvvn/WgPwFeBfolinLwO/8orW83PAX/uA7/0Re+964PP2nvof8Ho+DXzJPt8Dv2U/91726EPWc2979FE/7jsT+FPAt1tr/721tgK/BHz1nte0xVeBX7TPfxH4cx/nD2ut/Xvg/f/PNXwV+IdN4z8Aj0Tk069gPS+LrwK/1FpbWmv/A/g2+t7+INfz3dbaf7HPnwG/CXyGe9qjD1nPy+Jj36OPGvd9CHwG+J8v/P1/8eEb+XFFA/6ViPxnEflL9tpbrbXv2uf/G3jrHtb1sjXc5779jKXXv/BCifRK1yMinwN+FPgVHsAefd964AHs0R8k7vsQeCjx4621LwE/CfxlEfmJF7/YNJ+71zHKQ1gD8HeBHwb+JPBd4G++6gWIyDXwT4G/2lq7e/Fr97FHH7Cee9+jP2jc9yHwHeCzL/z9j9hrrzRaa9+xP98B/jmapr29pY/25zuvel0fsoZ72bfW2tuttdJUV+vv8zydfSXrEZGI3nD/uLX2z+zle9ujD1rPfe/RR4n7PgT+E/BFEfm8iHTATwG//CoXICJXIrLfPgf+NPBNW8dP27f9NPAvXuW6LF62hl8G/oJ1wL8MPH0hJf7Y4vtq6j+P7tO2np8SkV5EPg98EfiPP+CfLcA/AH6ztfa3XvjSvezRy9Zzn3v0keO+O5NoF/e30G7pz97Dz/8C2rX9NeA3tjUArwP/Fvht4N8Ar33M6/gnaPqY0HrxL75sDWjH++/Ynv068GOvaD3/yH7eN9CL+tMvfP/P2nq+Bfzkx7CeH0dT/W8AX7ePr9zXHn3Ieu5tjz7qxwUxeIlLfMLjvsuBS1ziEvccl0PgEpf4hMflELjEJT7hcTkELnGJT3hcDoFLXOITHpdD4BKX+ITH5RC4xCU+4XE5BC5xiU94/B+JkcEhOdJv7gAAAABJRU5ErkJggg==\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "Sample human image:\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ],
      "source": [
        "print(\"Sample horse image:\")\n",
        "plt.imshow(load_img(f\"{os.path.join(train_horses_dir, os.listdir(train_horses_dir)[0])}\"))\n",
        "plt.show()\n",
        "\n",
        "print(\"\\nSample human image:\")\n",
        "plt.imshow(load_img(f\"{os.path.join(train_humans_dir, os.listdir(train_humans_dir)[0])}\"))\n",
        "plt.show()"
      ],
      "id": "HgbMs7p0qSKr"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "LBnbnY0c8Zd0"
      },
      "source": [
        "`matplotlib` makes it easy to see that these images have a resolution of 300x300 and are colored, but you can double check this by using the code below:"
      ],
      "id": "LBnbnY0c8Zd0"
    },
    {
      "cell_type": "code",
      "execution_count": 7,
      "metadata": {
        "id": "4lIGjHC5pxua",
        "tags": [
          "graded"
        ],
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "047ef114-651f-4c13-9e0b-2f5edec58ae6"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Each image has shape: (300, 300, 3)\n"
          ]
        }
      ],
      "source": [
        "# Load the first example of a horse\n",
        "sample_image  = load_img(f\"{os.path.join(train_horses_dir, os.listdir(train_horses_dir)[0])}\")\n",
        "\n",
        "# Convert the image into its numpy array representation\n",
        "sample_array = img_to_array(sample_image)\n",
        "\n",
        "print(f\"Each image has shape: {sample_array.shape}\")"
      ],
      "id": "4lIGjHC5pxua"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "4fYwAYyd8zEm"
      },
      "source": [
        "As expected, the sample image has a resolution of 300x300 and the last dimension is used for each one of the RGB channels to represent color."
      ],
      "id": "4fYwAYyd8zEm"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "6HcE1TSqNRY2"
      },
      "source": [
        "## Training and Validation Generators\n",
        "\n",
        "Now that you know the images you are dealing with, it is time for you to code the generators that will fed these images to your Network. For this, complete the `train_val_generators` function below:\n",
        "\n",
        "**Important Note:** The images have a resolution of 300x300 but the `flow_from_directory` method you will use allows you to set a target resolution. In this case, **set a `target_size` of (150, 150)**. This will heavily lower the number of trainable parameters in your final network, yielding much quicker training times without compromising the accuracy!"
      ],
      "id": "6HcE1TSqNRY2"
    },
    {
      "cell_type": "code",
      "execution_count": 8,
      "metadata": {
        "cellView": "code",
        "id": "AX5Q3NL_FXMT",
        "tags": [
          "graded"
        ]
      },
      "outputs": [],
      "source": [
        "# GRADED FUNCTION: train_val_generators\n",
        "def train_val_generators(TRAINING_DIR, VALIDATION_DIR):\n",
        "  \"\"\"\n",
        "  Creates the training and validation data generators\n",
        "  \n",
        "  Args:\n",
        "    TRAINING_DIR (string): directory path containing the training images\n",
        "    VALIDATION_DIR (string): directory path containing the testing/validation images\n",
        "    \n",
        "  Returns:\n",
        "    train_generator, validation_generator: tuple containing the generators\n",
        "  \"\"\"\n",
        "  ### START CODE HERE\n",
        "\n",
        "  # Instantiate the ImageDataGenerator class \n",
        "  # Don't forget to normalize pixel values and set arguments to augment the images \n",
        "  train_datagen = ImageDataGenerator(rescale=1.0/255.0)\n",
        "\n",
        "  # Pass in the appropriate arguments to the flow_from_directory method\n",
        "  train_generator = train_datagen.flow_from_directory(directory=TRAINING_DIR,\n",
        "                                                      batch_size=32, \n",
        "                                                      class_mode='binary',\n",
        "                                                      target_size=(150, 150))\n",
        "\n",
        "  # Instantiate the ImageDataGenerator class (don't forget to set the rescale argument)\n",
        "  # Remember that validation data should not be augmented\n",
        "  validation_datagen = ImageDataGenerator(rescale=1.0/255.0)\n",
        "\n",
        "  # Pass in the appropriate arguments to the flow_from_directory method\n",
        "  validation_generator = validation_datagen.flow_from_directory(directory=VALIDATION_DIR,\n",
        "                                                                batch_size=32, \n",
        "                                                                class_mode='binary',\n",
        "                                                                target_size=(150, 150))\n",
        "  ### END CODE HERE\n",
        "  return train_generator, validation_generator\n"
      ],
      "id": "AX5Q3NL_FXMT"
    },
    {
      "cell_type": "code",
      "execution_count": 9,
      "metadata": {
        "id": "8FLUUqMKFwVR",
        "tags": [
          "graded"
        ],
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "960f995c-5d35-450f-b494-9ff5db1622e1"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Found 1027 images belonging to 2 classes.\n",
            "Found 256 images belonging to 2 classes.\n"
          ]
        }
      ],
      "source": [
        "# Test your generators\n",
        "train_generator, validation_generator = train_val_generators(train_dir, validation_dir)"
      ],
      "id": "8FLUUqMKFwVR"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "TszKWhunQaj4"
      },
      "source": [
        "**Expected Output:**\n",
        "```\n",
        "Found 1027 images belonging to 2 classes.\n",
        "Found 256 images belonging to 2 classes.\n",
        "```"
      ],
      "id": "TszKWhunQaj4"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Izx51Ju1rXwd"
      },
      "source": [
        "## Transfer learning - Create the pre-trained model\n",
        "\n",
        "Download the `inception V3` weights into the `/tmp/` directory:"
      ],
      "id": "Izx51Ju1rXwd"
    },
    {
      "cell_type": "code",
      "execution_count": 10,
      "metadata": {
        "id": "-lEzPAqxrPcU",
        "tags": [],
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "cc2deeb7-254b-4dae-9387-d7f2177efff8"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "--2022-11-26 21:51:39--  https://storage.googleapis.com/mledu-datasets/inception_v3_weights_tf_dim_ordering_tf_kernels_notop.h5\n",
            "Resolving storage.googleapis.com (storage.googleapis.com)... 142.251.16.128, 172.217.164.144, 172.253.62.128, ...\n",
            "Connecting to storage.googleapis.com (storage.googleapis.com)|142.251.16.128|:443... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 87910968 (84M) [application/x-hdf]\n",
            "Saving to: ‘/tmp/inception_v3_weights_tf_dim_ordering_tf_kernels_notop.h5’\n",
            "\n",
            "/tmp/inception_v3_w 100%[===================>]  83.84M   278MB/s    in 0.3s    \n",
            "\n",
            "2022-11-26 21:51:39 (278 MB/s) - ‘/tmp/inception_v3_weights_tf_dim_ordering_tf_kernels_notop.h5’ saved [87910968/87910968]\n",
            "\n"
          ]
        }
      ],
      "source": [
        "# Download the inception v3 weights\n",
        "!wget --no-check-certificate \\\n",
        "    https://storage.googleapis.com/mledu-datasets/inception_v3_weights_tf_dim_ordering_tf_kernels_notop.h5 \\\n",
        "    -O /tmp/inception_v3_weights_tf_dim_ordering_tf_kernels_notop.h5"
      ],
      "id": "-lEzPAqxrPcU"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_zlXNulm9USZ"
      },
      "source": [
        "Now load the `InceptionV3` model and save the path to the weights you just downloaded:"
      ],
      "id": "_zlXNulm9USZ"
    },
    {
      "cell_type": "code",
      "execution_count": 11,
      "metadata": {
        "id": "zfmRpsMf7E3-",
        "tags": [
          "graded"
        ]
      },
      "outputs": [],
      "source": [
        "# Import the inception model  \n",
        "from tensorflow.keras.applications.inception_v3 import InceptionV3\n",
        "\n",
        "# Create an instance of the inception model from the local pre-trained weights\n",
        "local_weights_file = '/tmp/inception_v3_weights_tf_dim_ordering_tf_kernels_notop.h5'"
      ],
      "id": "zfmRpsMf7E3-"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ZPQb0PkT9_3w"
      },
      "source": [
        "Complete the `create_pre_trained_model` function below. You should specify the correct `input_shape` for the model (remember that you set a new resolution for the images instead of the native 300x300) and make all of the layers non-trainable:"
      ],
      "id": "ZPQb0PkT9_3w"
    },
    {
      "cell_type": "code",
      "execution_count": 12,
      "metadata": {
        "cellView": "code",
        "id": "x2JnQ6m8r5oe",
        "tags": [
          "graded"
        ]
      },
      "outputs": [],
      "source": [
        "# GRADED FUNCTION: create_pre_trained_model\n",
        "def create_pre_trained_model(local_weights_file):\n",
        "  \"\"\"\n",
        "  Initializes an InceptionV3 model.\n",
        "  \n",
        "  Args:\n",
        "    local_weights_file (string): path pointing to a pretrained weights H5 file\n",
        "    \n",
        "  Returns:\n",
        "    pre_trained_model: the initialized InceptionV3 model\n",
        "  \"\"\"\n",
        "  ### START CODE HERE\n",
        "  pre_trained_model = InceptionV3(input_shape = (150, 150, 3),\n",
        "                                  include_top = False, \n",
        "                                  weights = None) \n",
        "\n",
        "  pre_trained_model.load_weights(local_weights_file)\n",
        "\n",
        "  # Make all the layers in the pre-trained model non-trainable\n",
        "  for layer in pre_trained_model.layers:\n",
        "    layer.trainable = False\n",
        "\n",
        "  ### END CODE HERE\n",
        "\n",
        "  return pre_trained_model\n",
        "  "
      ],
      "id": "x2JnQ6m8r5oe"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "phE00SCr-RCT"
      },
      "source": [
        "Check that everything went well by comparing the last few rows of the model summary to the expected output:"
      ],
      "id": "phE00SCr-RCT"
    },
    {
      "cell_type": "code",
      "execution_count": 13,
      "metadata": {
        "id": "ve7eh9iztT4q",
        "tags": [
          "graded"
        ],
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "95449d25-9223-4a22-9281-628a1eb0dde9"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"inception_v3\"\n",
            "__________________________________________________________________________________________________\n",
            " Layer (type)                   Output Shape         Param #     Connected to                     \n",
            "==================================================================================================\n",
            " input_1 (InputLayer)           [(None, 150, 150, 3  0           []                               \n",
            "                                )]                                                                \n",
            "                                                                                                  \n",
            " conv2d (Conv2D)                (None, 74, 74, 32)   864         ['input_1[0][0]']                \n",
            "                                                                                                  \n",
            " batch_normalization (BatchNorm  (None, 74, 74, 32)  96          ['conv2d[0][0]']                 \n",
            " alization)                                                                                       \n",
            "                                                                                                  \n",
            " activation (Activation)        (None, 74, 74, 32)   0           ['batch_normalization[0][0]']    \n",
            "                                                                                                  \n",
            " conv2d_1 (Conv2D)              (None, 72, 72, 32)   9216        ['activation[0][0]']             \n",
            "                                                                                                  \n",
            " batch_normalization_1 (BatchNo  (None, 72, 72, 32)  96          ['conv2d_1[0][0]']               \n",
            " rmalization)                                                                                     \n",
            "                                                                                                  \n",
            " activation_1 (Activation)      (None, 72, 72, 32)   0           ['batch_normalization_1[0][0]']  \n",
            "                                                                                                  \n",
            " conv2d_2 (Conv2D)              (None, 72, 72, 64)   18432       ['activation_1[0][0]']           \n",
            "                                                                                                  \n",
            " batch_normalization_2 (BatchNo  (None, 72, 72, 64)  192         ['conv2d_2[0][0]']               \n",
            " rmalization)                                                                                     \n",
            "                                                                                                  \n",
            " activation_2 (Activation)      (None, 72, 72, 64)   0           ['batch_normalization_2[0][0]']  \n",
            "                                                                                                  \n",
            " max_pooling2d (MaxPooling2D)   (None, 35, 35, 64)   0           ['activation_2[0][0]']           \n",
            "                                                                                                  \n",
            " conv2d_3 (Conv2D)              (None, 35, 35, 80)   5120        ['max_pooling2d[0][0]']          \n",
            "                                                                                                  \n",
            " batch_normalization_3 (BatchNo  (None, 35, 35, 80)  240         ['conv2d_3[0][0]']               \n",
            " rmalization)                                                                                     \n",
            "                                                                                                  \n",
            " activation_3 (Activation)      (None, 35, 35, 80)   0           ['batch_normalization_3[0][0]']  \n",
            "                                                                                                  \n",
            " conv2d_4 (Conv2D)              (None, 33, 33, 192)  138240      ['activation_3[0][0]']           \n",
            "                                                                                                  \n",
            " batch_normalization_4 (BatchNo  (None, 33, 33, 192)  576        ['conv2d_4[0][0]']               \n",
            " rmalization)                                                                                     \n",
            "                                                                                                  \n",
            " activation_4 (Activation)      (None, 33, 33, 192)  0           ['batch_normalization_4[0][0]']  \n",
            "                                                                                                  \n",
            " max_pooling2d_1 (MaxPooling2D)  (None, 16, 16, 192)  0          ['activation_4[0][0]']           \n",
            "                                                                                                  \n",
            " conv2d_8 (Conv2D)              (None, 16, 16, 64)   12288       ['max_pooling2d_1[0][0]']        \n",
            "                                                                                                  \n",
            " batch_normalization_8 (BatchNo  (None, 16, 16, 64)  192         ['conv2d_8[0][0]']               \n",
            " rmalization)                                                                                     \n",
            "                                                                                                  \n",
            " activation_8 (Activation)      (None, 16, 16, 64)   0           ['batch_normalization_8[0][0]']  \n",
            "                                                                                                  \n",
            " conv2d_6 (Conv2D)              (None, 16, 16, 48)   9216        ['max_pooling2d_1[0][0]']        \n",
            "                                                                                                  \n",
            " conv2d_9 (Conv2D)              (None, 16, 16, 96)   55296       ['activation_8[0][0]']           \n",
            "                                                                                                  \n",
            " batch_normalization_6 (BatchNo  (None, 16, 16, 48)  144         ['conv2d_6[0][0]']               \n",
            " rmalization)                                                                                     \n",
            "                                                                                                  \n",
            " batch_normalization_9 (BatchNo  (None, 16, 16, 96)  288         ['conv2d_9[0][0]']               \n",
            " rmalization)                                                                                     \n",
            "                                                                                                  \n",
            " activation_6 (Activation)      (None, 16, 16, 48)   0           ['batch_normalization_6[0][0]']  \n",
            "                                                                                                  \n",
            " activation_9 (Activation)      (None, 16, 16, 96)   0           ['batch_normalization_9[0][0]']  \n",
            "                                                                                                  \n",
            " average_pooling2d (AveragePool  (None, 16, 16, 192)  0          ['max_pooling2d_1[0][0]']        \n",
            " ing2D)                                                                                           \n",
            "                                                                                                  \n",
            " conv2d_5 (Conv2D)              (None, 16, 16, 64)   12288       ['max_pooling2d_1[0][0]']        \n",
            "                                                                                                  \n",
            " conv2d_7 (Conv2D)              (None, 16, 16, 64)   76800       ['activation_6[0][0]']           \n",
            "                                                                                                  \n",
            " conv2d_10 (Conv2D)             (None, 16, 16, 96)   82944       ['activation_9[0][0]']           \n",
            "                                                                                                  \n",
            " conv2d_11 (Conv2D)             (None, 16, 16, 32)   6144        ['average_pooling2d[0][0]']      \n",
            "                                                                                                  \n",
            " batch_normalization_5 (BatchNo  (None, 16, 16, 64)  192         ['conv2d_5[0][0]']               \n",
            " rmalization)                                                                                     \n",
            "                                                                                                  \n",
            " batch_normalization_7 (BatchNo  (None, 16, 16, 64)  192         ['conv2d_7[0][0]']               \n",
            " rmalization)                                                                                     \n",
            "                                                                                                  \n",
            " batch_normalization_10 (BatchN  (None, 16, 16, 96)  288         ['conv2d_10[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " batch_normalization_11 (BatchN  (None, 16, 16, 32)  96          ['conv2d_11[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " activation_5 (Activation)      (None, 16, 16, 64)   0           ['batch_normalization_5[0][0]']  \n",
            "                                                                                                  \n",
            " activation_7 (Activation)      (None, 16, 16, 64)   0           ['batch_normalization_7[0][0]']  \n",
            "                                                                                                  \n",
            " activation_10 (Activation)     (None, 16, 16, 96)   0           ['batch_normalization_10[0][0]'] \n",
            "                                                                                                  \n",
            " activation_11 (Activation)     (None, 16, 16, 32)   0           ['batch_normalization_11[0][0]'] \n",
            "                                                                                                  \n",
            " mixed0 (Concatenate)           (None, 16, 16, 256)  0           ['activation_5[0][0]',           \n",
            "                                                                  'activation_7[0][0]',           \n",
            "                                                                  'activation_10[0][0]',          \n",
            "                                                                  'activation_11[0][0]']          \n",
            "                                                                                                  \n",
            " conv2d_15 (Conv2D)             (None, 16, 16, 64)   16384       ['mixed0[0][0]']                 \n",
            "                                                                                                  \n",
            " batch_normalization_15 (BatchN  (None, 16, 16, 64)  192         ['conv2d_15[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " activation_15 (Activation)     (None, 16, 16, 64)   0           ['batch_normalization_15[0][0]'] \n",
            "                                                                                                  \n",
            " conv2d_13 (Conv2D)             (None, 16, 16, 48)   12288       ['mixed0[0][0]']                 \n",
            "                                                                                                  \n",
            " conv2d_16 (Conv2D)             (None, 16, 16, 96)   55296       ['activation_15[0][0]']          \n",
            "                                                                                                  \n",
            " batch_normalization_13 (BatchN  (None, 16, 16, 48)  144         ['conv2d_13[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " batch_normalization_16 (BatchN  (None, 16, 16, 96)  288         ['conv2d_16[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " activation_13 (Activation)     (None, 16, 16, 48)   0           ['batch_normalization_13[0][0]'] \n",
            "                                                                                                  \n",
            " activation_16 (Activation)     (None, 16, 16, 96)   0           ['batch_normalization_16[0][0]'] \n",
            "                                                                                                  \n",
            " average_pooling2d_1 (AveragePo  (None, 16, 16, 256)  0          ['mixed0[0][0]']                 \n",
            " oling2D)                                                                                         \n",
            "                                                                                                  \n",
            " conv2d_12 (Conv2D)             (None, 16, 16, 64)   16384       ['mixed0[0][0]']                 \n",
            "                                                                                                  \n",
            " conv2d_14 (Conv2D)             (None, 16, 16, 64)   76800       ['activation_13[0][0]']          \n",
            "                                                                                                  \n",
            " conv2d_17 (Conv2D)             (None, 16, 16, 96)   82944       ['activation_16[0][0]']          \n",
            "                                                                                                  \n",
            " conv2d_18 (Conv2D)             (None, 16, 16, 64)   16384       ['average_pooling2d_1[0][0]']    \n",
            "                                                                                                  \n",
            " batch_normalization_12 (BatchN  (None, 16, 16, 64)  192         ['conv2d_12[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " batch_normalization_14 (BatchN  (None, 16, 16, 64)  192         ['conv2d_14[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " batch_normalization_17 (BatchN  (None, 16, 16, 96)  288         ['conv2d_17[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " batch_normalization_18 (BatchN  (None, 16, 16, 64)  192         ['conv2d_18[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " activation_12 (Activation)     (None, 16, 16, 64)   0           ['batch_normalization_12[0][0]'] \n",
            "                                                                                                  \n",
            " activation_14 (Activation)     (None, 16, 16, 64)   0           ['batch_normalization_14[0][0]'] \n",
            "                                                                                                  \n",
            " activation_17 (Activation)     (None, 16, 16, 96)   0           ['batch_normalization_17[0][0]'] \n",
            "                                                                                                  \n",
            " activation_18 (Activation)     (None, 16, 16, 64)   0           ['batch_normalization_18[0][0]'] \n",
            "                                                                                                  \n",
            " mixed1 (Concatenate)           (None, 16, 16, 288)  0           ['activation_12[0][0]',          \n",
            "                                                                  'activation_14[0][0]',          \n",
            "                                                                  'activation_17[0][0]',          \n",
            "                                                                  'activation_18[0][0]']          \n",
            "                                                                                                  \n",
            " conv2d_22 (Conv2D)             (None, 16, 16, 64)   18432       ['mixed1[0][0]']                 \n",
            "                                                                                                  \n",
            " batch_normalization_22 (BatchN  (None, 16, 16, 64)  192         ['conv2d_22[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " activation_22 (Activation)     (None, 16, 16, 64)   0           ['batch_normalization_22[0][0]'] \n",
            "                                                                                                  \n",
            " conv2d_20 (Conv2D)             (None, 16, 16, 48)   13824       ['mixed1[0][0]']                 \n",
            "                                                                                                  \n",
            " conv2d_23 (Conv2D)             (None, 16, 16, 96)   55296       ['activation_22[0][0]']          \n",
            "                                                                                                  \n",
            " batch_normalization_20 (BatchN  (None, 16, 16, 48)  144         ['conv2d_20[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " batch_normalization_23 (BatchN  (None, 16, 16, 96)  288         ['conv2d_23[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " activation_20 (Activation)     (None, 16, 16, 48)   0           ['batch_normalization_20[0][0]'] \n",
            "                                                                                                  \n",
            " activation_23 (Activation)     (None, 16, 16, 96)   0           ['batch_normalization_23[0][0]'] \n",
            "                                                                                                  \n",
            " average_pooling2d_2 (AveragePo  (None, 16, 16, 288)  0          ['mixed1[0][0]']                 \n",
            " oling2D)                                                                                         \n",
            "                                                                                                  \n",
            " conv2d_19 (Conv2D)             (None, 16, 16, 64)   18432       ['mixed1[0][0]']                 \n",
            "                                                                                                  \n",
            " conv2d_21 (Conv2D)             (None, 16, 16, 64)   76800       ['activation_20[0][0]']          \n",
            "                                                                                                  \n",
            " conv2d_24 (Conv2D)             (None, 16, 16, 96)   82944       ['activation_23[0][0]']          \n",
            "                                                                                                  \n",
            " conv2d_25 (Conv2D)             (None, 16, 16, 64)   18432       ['average_pooling2d_2[0][0]']    \n",
            "                                                                                                  \n",
            " batch_normalization_19 (BatchN  (None, 16, 16, 64)  192         ['conv2d_19[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " batch_normalization_21 (BatchN  (None, 16, 16, 64)  192         ['conv2d_21[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " batch_normalization_24 (BatchN  (None, 16, 16, 96)  288         ['conv2d_24[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " batch_normalization_25 (BatchN  (None, 16, 16, 64)  192         ['conv2d_25[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " activation_19 (Activation)     (None, 16, 16, 64)   0           ['batch_normalization_19[0][0]'] \n",
            "                                                                                                  \n",
            " activation_21 (Activation)     (None, 16, 16, 64)   0           ['batch_normalization_21[0][0]'] \n",
            "                                                                                                  \n",
            " activation_24 (Activation)     (None, 16, 16, 96)   0           ['batch_normalization_24[0][0]'] \n",
            "                                                                                                  \n",
            " activation_25 (Activation)     (None, 16, 16, 64)   0           ['batch_normalization_25[0][0]'] \n",
            "                                                                                                  \n",
            " mixed2 (Concatenate)           (None, 16, 16, 288)  0           ['activation_19[0][0]',          \n",
            "                                                                  'activation_21[0][0]',          \n",
            "                                                                  'activation_24[0][0]',          \n",
            "                                                                  'activation_25[0][0]']          \n",
            "                                                                                                  \n",
            " conv2d_27 (Conv2D)             (None, 16, 16, 64)   18432       ['mixed2[0][0]']                 \n",
            "                                                                                                  \n",
            " batch_normalization_27 (BatchN  (None, 16, 16, 64)  192         ['conv2d_27[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " activation_27 (Activation)     (None, 16, 16, 64)   0           ['batch_normalization_27[0][0]'] \n",
            "                                                                                                  \n",
            " conv2d_28 (Conv2D)             (None, 16, 16, 96)   55296       ['activation_27[0][0]']          \n",
            "                                                                                                  \n",
            " batch_normalization_28 (BatchN  (None, 16, 16, 96)  288         ['conv2d_28[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " activation_28 (Activation)     (None, 16, 16, 96)   0           ['batch_normalization_28[0][0]'] \n",
            "                                                                                                  \n",
            " conv2d_26 (Conv2D)             (None, 7, 7, 384)    995328      ['mixed2[0][0]']                 \n",
            "                                                                                                  \n",
            " conv2d_29 (Conv2D)             (None, 7, 7, 96)     82944       ['activation_28[0][0]']          \n",
            "                                                                                                  \n",
            " batch_normalization_26 (BatchN  (None, 7, 7, 384)   1152        ['conv2d_26[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " batch_normalization_29 (BatchN  (None, 7, 7, 96)    288         ['conv2d_29[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " activation_26 (Activation)     (None, 7, 7, 384)    0           ['batch_normalization_26[0][0]'] \n",
            "                                                                                                  \n",
            " activation_29 (Activation)     (None, 7, 7, 96)     0           ['batch_normalization_29[0][0]'] \n",
            "                                                                                                  \n",
            " max_pooling2d_2 (MaxPooling2D)  (None, 7, 7, 288)   0           ['mixed2[0][0]']                 \n",
            "                                                                                                  \n",
            " mixed3 (Concatenate)           (None, 7, 7, 768)    0           ['activation_26[0][0]',          \n",
            "                                                                  'activation_29[0][0]',          \n",
            "                                                                  'max_pooling2d_2[0][0]']        \n",
            "                                                                                                  \n",
            " conv2d_34 (Conv2D)             (None, 7, 7, 128)    98304       ['mixed3[0][0]']                 \n",
            "                                                                                                  \n",
            " batch_normalization_34 (BatchN  (None, 7, 7, 128)   384         ['conv2d_34[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " activation_34 (Activation)     (None, 7, 7, 128)    0           ['batch_normalization_34[0][0]'] \n",
            "                                                                                                  \n",
            " conv2d_35 (Conv2D)             (None, 7, 7, 128)    114688      ['activation_34[0][0]']          \n",
            "                                                                                                  \n",
            " batch_normalization_35 (BatchN  (None, 7, 7, 128)   384         ['conv2d_35[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " activation_35 (Activation)     (None, 7, 7, 128)    0           ['batch_normalization_35[0][0]'] \n",
            "                                                                                                  \n",
            " conv2d_31 (Conv2D)             (None, 7, 7, 128)    98304       ['mixed3[0][0]']                 \n",
            "                                                                                                  \n",
            " conv2d_36 (Conv2D)             (None, 7, 7, 128)    114688      ['activation_35[0][0]']          \n",
            "                                                                                                  \n",
            " batch_normalization_31 (BatchN  (None, 7, 7, 128)   384         ['conv2d_31[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " batch_normalization_36 (BatchN  (None, 7, 7, 128)   384         ['conv2d_36[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " activation_31 (Activation)     (None, 7, 7, 128)    0           ['batch_normalization_31[0][0]'] \n",
            "                                                                                                  \n",
            " activation_36 (Activation)     (None, 7, 7, 128)    0           ['batch_normalization_36[0][0]'] \n",
            "                                                                                                  \n",
            " conv2d_32 (Conv2D)             (None, 7, 7, 128)    114688      ['activation_31[0][0]']          \n",
            "                                                                                                  \n",
            " conv2d_37 (Conv2D)             (None, 7, 7, 128)    114688      ['activation_36[0][0]']          \n",
            "                                                                                                  \n",
            " batch_normalization_32 (BatchN  (None, 7, 7, 128)   384         ['conv2d_32[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " batch_normalization_37 (BatchN  (None, 7, 7, 128)   384         ['conv2d_37[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " activation_32 (Activation)     (None, 7, 7, 128)    0           ['batch_normalization_32[0][0]'] \n",
            "                                                                                                  \n",
            " activation_37 (Activation)     (None, 7, 7, 128)    0           ['batch_normalization_37[0][0]'] \n",
            "                                                                                                  \n",
            " average_pooling2d_3 (AveragePo  (None, 7, 7, 768)   0           ['mixed3[0][0]']                 \n",
            " oling2D)                                                                                         \n",
            "                                                                                                  \n",
            " conv2d_30 (Conv2D)             (None, 7, 7, 192)    147456      ['mixed3[0][0]']                 \n",
            "                                                                                                  \n",
            " conv2d_33 (Conv2D)             (None, 7, 7, 192)    172032      ['activation_32[0][0]']          \n",
            "                                                                                                  \n",
            " conv2d_38 (Conv2D)             (None, 7, 7, 192)    172032      ['activation_37[0][0]']          \n",
            "                                                                                                  \n",
            " conv2d_39 (Conv2D)             (None, 7, 7, 192)    147456      ['average_pooling2d_3[0][0]']    \n",
            "                                                                                                  \n",
            " batch_normalization_30 (BatchN  (None, 7, 7, 192)   576         ['conv2d_30[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " batch_normalization_33 (BatchN  (None, 7, 7, 192)   576         ['conv2d_33[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " batch_normalization_38 (BatchN  (None, 7, 7, 192)   576         ['conv2d_38[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " batch_normalization_39 (BatchN  (None, 7, 7, 192)   576         ['conv2d_39[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " activation_30 (Activation)     (None, 7, 7, 192)    0           ['batch_normalization_30[0][0]'] \n",
            "                                                                                                  \n",
            " activation_33 (Activation)     (None, 7, 7, 192)    0           ['batch_normalization_33[0][0]'] \n",
            "                                                                                                  \n",
            " activation_38 (Activation)     (None, 7, 7, 192)    0           ['batch_normalization_38[0][0]'] \n",
            "                                                                                                  \n",
            " activation_39 (Activation)     (None, 7, 7, 192)    0           ['batch_normalization_39[0][0]'] \n",
            "                                                                                                  \n",
            " mixed4 (Concatenate)           (None, 7, 7, 768)    0           ['activation_30[0][0]',          \n",
            "                                                                  'activation_33[0][0]',          \n",
            "                                                                  'activation_38[0][0]',          \n",
            "                                                                  'activation_39[0][0]']          \n",
            "                                                                                                  \n",
            " conv2d_44 (Conv2D)             (None, 7, 7, 160)    122880      ['mixed4[0][0]']                 \n",
            "                                                                                                  \n",
            " batch_normalization_44 (BatchN  (None, 7, 7, 160)   480         ['conv2d_44[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " activation_44 (Activation)     (None, 7, 7, 160)    0           ['batch_normalization_44[0][0]'] \n",
            "                                                                                                  \n",
            " conv2d_45 (Conv2D)             (None, 7, 7, 160)    179200      ['activation_44[0][0]']          \n",
            "                                                                                                  \n",
            " batch_normalization_45 (BatchN  (None, 7, 7, 160)   480         ['conv2d_45[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " activation_45 (Activation)     (None, 7, 7, 160)    0           ['batch_normalization_45[0][0]'] \n",
            "                                                                                                  \n",
            " conv2d_41 (Conv2D)             (None, 7, 7, 160)    122880      ['mixed4[0][0]']                 \n",
            "                                                                                                  \n",
            " conv2d_46 (Conv2D)             (None, 7, 7, 160)    179200      ['activation_45[0][0]']          \n",
            "                                                                                                  \n",
            " batch_normalization_41 (BatchN  (None, 7, 7, 160)   480         ['conv2d_41[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " batch_normalization_46 (BatchN  (None, 7, 7, 160)   480         ['conv2d_46[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " activation_41 (Activation)     (None, 7, 7, 160)    0           ['batch_normalization_41[0][0]'] \n",
            "                                                                                                  \n",
            " activation_46 (Activation)     (None, 7, 7, 160)    0           ['batch_normalization_46[0][0]'] \n",
            "                                                                                                  \n",
            " conv2d_42 (Conv2D)             (None, 7, 7, 160)    179200      ['activation_41[0][0]']          \n",
            "                                                                                                  \n",
            " conv2d_47 (Conv2D)             (None, 7, 7, 160)    179200      ['activation_46[0][0]']          \n",
            "                                                                                                  \n",
            " batch_normalization_42 (BatchN  (None, 7, 7, 160)   480         ['conv2d_42[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " batch_normalization_47 (BatchN  (None, 7, 7, 160)   480         ['conv2d_47[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " activation_42 (Activation)     (None, 7, 7, 160)    0           ['batch_normalization_42[0][0]'] \n",
            "                                                                                                  \n",
            " activation_47 (Activation)     (None, 7, 7, 160)    0           ['batch_normalization_47[0][0]'] \n",
            "                                                                                                  \n",
            " average_pooling2d_4 (AveragePo  (None, 7, 7, 768)   0           ['mixed4[0][0]']                 \n",
            " oling2D)                                                                                         \n",
            "                                                                                                  \n",
            " conv2d_40 (Conv2D)             (None, 7, 7, 192)    147456      ['mixed4[0][0]']                 \n",
            "                                                                                                  \n",
            " conv2d_43 (Conv2D)             (None, 7, 7, 192)    215040      ['activation_42[0][0]']          \n",
            "                                                                                                  \n",
            " conv2d_48 (Conv2D)             (None, 7, 7, 192)    215040      ['activation_47[0][0]']          \n",
            "                                                                                                  \n",
            " conv2d_49 (Conv2D)             (None, 7, 7, 192)    147456      ['average_pooling2d_4[0][0]']    \n",
            "                                                                                                  \n",
            " batch_normalization_40 (BatchN  (None, 7, 7, 192)   576         ['conv2d_40[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " batch_normalization_43 (BatchN  (None, 7, 7, 192)   576         ['conv2d_43[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " batch_normalization_48 (BatchN  (None, 7, 7, 192)   576         ['conv2d_48[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " batch_normalization_49 (BatchN  (None, 7, 7, 192)   576         ['conv2d_49[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " activation_40 (Activation)     (None, 7, 7, 192)    0           ['batch_normalization_40[0][0]'] \n",
            "                                                                                                  \n",
            " activation_43 (Activation)     (None, 7, 7, 192)    0           ['batch_normalization_43[0][0]'] \n",
            "                                                                                                  \n",
            " activation_48 (Activation)     (None, 7, 7, 192)    0           ['batch_normalization_48[0][0]'] \n",
            "                                                                                                  \n",
            " activation_49 (Activation)     (None, 7, 7, 192)    0           ['batch_normalization_49[0][0]'] \n",
            "                                                                                                  \n",
            " mixed5 (Concatenate)           (None, 7, 7, 768)    0           ['activation_40[0][0]',          \n",
            "                                                                  'activation_43[0][0]',          \n",
            "                                                                  'activation_48[0][0]',          \n",
            "                                                                  'activation_49[0][0]']          \n",
            "                                                                                                  \n",
            " conv2d_54 (Conv2D)             (None, 7, 7, 160)    122880      ['mixed5[0][0]']                 \n",
            "                                                                                                  \n",
            " batch_normalization_54 (BatchN  (None, 7, 7, 160)   480         ['conv2d_54[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " activation_54 (Activation)     (None, 7, 7, 160)    0           ['batch_normalization_54[0][0]'] \n",
            "                                                                                                  \n",
            " conv2d_55 (Conv2D)             (None, 7, 7, 160)    179200      ['activation_54[0][0]']          \n",
            "                                                                                                  \n",
            " batch_normalization_55 (BatchN  (None, 7, 7, 160)   480         ['conv2d_55[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " activation_55 (Activation)     (None, 7, 7, 160)    0           ['batch_normalization_55[0][0]'] \n",
            "                                                                                                  \n",
            " conv2d_51 (Conv2D)             (None, 7, 7, 160)    122880      ['mixed5[0][0]']                 \n",
            "                                                                                                  \n",
            " conv2d_56 (Conv2D)             (None, 7, 7, 160)    179200      ['activation_55[0][0]']          \n",
            "                                                                                                  \n",
            " batch_normalization_51 (BatchN  (None, 7, 7, 160)   480         ['conv2d_51[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " batch_normalization_56 (BatchN  (None, 7, 7, 160)   480         ['conv2d_56[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " activation_51 (Activation)     (None, 7, 7, 160)    0           ['batch_normalization_51[0][0]'] \n",
            "                                                                                                  \n",
            " activation_56 (Activation)     (None, 7, 7, 160)    0           ['batch_normalization_56[0][0]'] \n",
            "                                                                                                  \n",
            " conv2d_52 (Conv2D)             (None, 7, 7, 160)    179200      ['activation_51[0][0]']          \n",
            "                                                                                                  \n",
            " conv2d_57 (Conv2D)             (None, 7, 7, 160)    179200      ['activation_56[0][0]']          \n",
            "                                                                                                  \n",
            " batch_normalization_52 (BatchN  (None, 7, 7, 160)   480         ['conv2d_52[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " batch_normalization_57 (BatchN  (None, 7, 7, 160)   480         ['conv2d_57[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " activation_52 (Activation)     (None, 7, 7, 160)    0           ['batch_normalization_52[0][0]'] \n",
            "                                                                                                  \n",
            " activation_57 (Activation)     (None, 7, 7, 160)    0           ['batch_normalization_57[0][0]'] \n",
            "                                                                                                  \n",
            " average_pooling2d_5 (AveragePo  (None, 7, 7, 768)   0           ['mixed5[0][0]']                 \n",
            " oling2D)                                                                                         \n",
            "                                                                                                  \n",
            " conv2d_50 (Conv2D)             (None, 7, 7, 192)    147456      ['mixed5[0][0]']                 \n",
            "                                                                                                  \n",
            " conv2d_53 (Conv2D)             (None, 7, 7, 192)    215040      ['activation_52[0][0]']          \n",
            "                                                                                                  \n",
            " conv2d_58 (Conv2D)             (None, 7, 7, 192)    215040      ['activation_57[0][0]']          \n",
            "                                                                                                  \n",
            " conv2d_59 (Conv2D)             (None, 7, 7, 192)    147456      ['average_pooling2d_5[0][0]']    \n",
            "                                                                                                  \n",
            " batch_normalization_50 (BatchN  (None, 7, 7, 192)   576         ['conv2d_50[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " batch_normalization_53 (BatchN  (None, 7, 7, 192)   576         ['conv2d_53[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " batch_normalization_58 (BatchN  (None, 7, 7, 192)   576         ['conv2d_58[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " batch_normalization_59 (BatchN  (None, 7, 7, 192)   576         ['conv2d_59[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " activation_50 (Activation)     (None, 7, 7, 192)    0           ['batch_normalization_50[0][0]'] \n",
            "                                                                                                  \n",
            " activation_53 (Activation)     (None, 7, 7, 192)    0           ['batch_normalization_53[0][0]'] \n",
            "                                                                                                  \n",
            " activation_58 (Activation)     (None, 7, 7, 192)    0           ['batch_normalization_58[0][0]'] \n",
            "                                                                                                  \n",
            " activation_59 (Activation)     (None, 7, 7, 192)    0           ['batch_normalization_59[0][0]'] \n",
            "                                                                                                  \n",
            " mixed6 (Concatenate)           (None, 7, 7, 768)    0           ['activation_50[0][0]',          \n",
            "                                                                  'activation_53[0][0]',          \n",
            "                                                                  'activation_58[0][0]',          \n",
            "                                                                  'activation_59[0][0]']          \n",
            "                                                                                                  \n",
            " conv2d_64 (Conv2D)             (None, 7, 7, 192)    147456      ['mixed6[0][0]']                 \n",
            "                                                                                                  \n",
            " batch_normalization_64 (BatchN  (None, 7, 7, 192)   576         ['conv2d_64[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " activation_64 (Activation)     (None, 7, 7, 192)    0           ['batch_normalization_64[0][0]'] \n",
            "                                                                                                  \n",
            " conv2d_65 (Conv2D)             (None, 7, 7, 192)    258048      ['activation_64[0][0]']          \n",
            "                                                                                                  \n",
            " batch_normalization_65 (BatchN  (None, 7, 7, 192)   576         ['conv2d_65[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " activation_65 (Activation)     (None, 7, 7, 192)    0           ['batch_normalization_65[0][0]'] \n",
            "                                                                                                  \n",
            " conv2d_61 (Conv2D)             (None, 7, 7, 192)    147456      ['mixed6[0][0]']                 \n",
            "                                                                                                  \n",
            " conv2d_66 (Conv2D)             (None, 7, 7, 192)    258048      ['activation_65[0][0]']          \n",
            "                                                                                                  \n",
            " batch_normalization_61 (BatchN  (None, 7, 7, 192)   576         ['conv2d_61[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " batch_normalization_66 (BatchN  (None, 7, 7, 192)   576         ['conv2d_66[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " activation_61 (Activation)     (None, 7, 7, 192)    0           ['batch_normalization_61[0][0]'] \n",
            "                                                                                                  \n",
            " activation_66 (Activation)     (None, 7, 7, 192)    0           ['batch_normalization_66[0][0]'] \n",
            "                                                                                                  \n",
            " conv2d_62 (Conv2D)             (None, 7, 7, 192)    258048      ['activation_61[0][0]']          \n",
            "                                                                                                  \n",
            " conv2d_67 (Conv2D)             (None, 7, 7, 192)    258048      ['activation_66[0][0]']          \n",
            "                                                                                                  \n",
            " batch_normalization_62 (BatchN  (None, 7, 7, 192)   576         ['conv2d_62[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " batch_normalization_67 (BatchN  (None, 7, 7, 192)   576         ['conv2d_67[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " activation_62 (Activation)     (None, 7, 7, 192)    0           ['batch_normalization_62[0][0]'] \n",
            "                                                                                                  \n",
            " activation_67 (Activation)     (None, 7, 7, 192)    0           ['batch_normalization_67[0][0]'] \n",
            "                                                                                                  \n",
            " average_pooling2d_6 (AveragePo  (None, 7, 7, 768)   0           ['mixed6[0][0]']                 \n",
            " oling2D)                                                                                         \n",
            "                                                                                                  \n",
            " conv2d_60 (Conv2D)             (None, 7, 7, 192)    147456      ['mixed6[0][0]']                 \n",
            "                                                                                                  \n",
            " conv2d_63 (Conv2D)             (None, 7, 7, 192)    258048      ['activation_62[0][0]']          \n",
            "                                                                                                  \n",
            " conv2d_68 (Conv2D)             (None, 7, 7, 192)    258048      ['activation_67[0][0]']          \n",
            "                                                                                                  \n",
            " conv2d_69 (Conv2D)             (None, 7, 7, 192)    147456      ['average_pooling2d_6[0][0]']    \n",
            "                                                                                                  \n",
            " batch_normalization_60 (BatchN  (None, 7, 7, 192)   576         ['conv2d_60[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " batch_normalization_63 (BatchN  (None, 7, 7, 192)   576         ['conv2d_63[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " batch_normalization_68 (BatchN  (None, 7, 7, 192)   576         ['conv2d_68[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " batch_normalization_69 (BatchN  (None, 7, 7, 192)   576         ['conv2d_69[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " activation_60 (Activation)     (None, 7, 7, 192)    0           ['batch_normalization_60[0][0]'] \n",
            "                                                                                                  \n",
            " activation_63 (Activation)     (None, 7, 7, 192)    0           ['batch_normalization_63[0][0]'] \n",
            "                                                                                                  \n",
            " activation_68 (Activation)     (None, 7, 7, 192)    0           ['batch_normalization_68[0][0]'] \n",
            "                                                                                                  \n",
            " activation_69 (Activation)     (None, 7, 7, 192)    0           ['batch_normalization_69[0][0]'] \n",
            "                                                                                                  \n",
            " mixed7 (Concatenate)           (None, 7, 7, 768)    0           ['activation_60[0][0]',          \n",
            "                                                                  'activation_63[0][0]',          \n",
            "                                                                  'activation_68[0][0]',          \n",
            "                                                                  'activation_69[0][0]']          \n",
            "                                                                                                  \n",
            " conv2d_72 (Conv2D)             (None, 7, 7, 192)    147456      ['mixed7[0][0]']                 \n",
            "                                                                                                  \n",
            " batch_normalization_72 (BatchN  (None, 7, 7, 192)   576         ['conv2d_72[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " activation_72 (Activation)     (None, 7, 7, 192)    0           ['batch_normalization_72[0][0]'] \n",
            "                                                                                                  \n",
            " conv2d_73 (Conv2D)             (None, 7, 7, 192)    258048      ['activation_72[0][0]']          \n",
            "                                                                                                  \n",
            " batch_normalization_73 (BatchN  (None, 7, 7, 192)   576         ['conv2d_73[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " activation_73 (Activation)     (None, 7, 7, 192)    0           ['batch_normalization_73[0][0]'] \n",
            "                                                                                                  \n",
            " conv2d_70 (Conv2D)             (None, 7, 7, 192)    147456      ['mixed7[0][0]']                 \n",
            "                                                                                                  \n",
            " conv2d_74 (Conv2D)             (None, 7, 7, 192)    258048      ['activation_73[0][0]']          \n",
            "                                                                                                  \n",
            " batch_normalization_70 (BatchN  (None, 7, 7, 192)   576         ['conv2d_70[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " batch_normalization_74 (BatchN  (None, 7, 7, 192)   576         ['conv2d_74[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " activation_70 (Activation)     (None, 7, 7, 192)    0           ['batch_normalization_70[0][0]'] \n",
            "                                                                                                  \n",
            " activation_74 (Activation)     (None, 7, 7, 192)    0           ['batch_normalization_74[0][0]'] \n",
            "                                                                                                  \n",
            " conv2d_71 (Conv2D)             (None, 3, 3, 320)    552960      ['activation_70[0][0]']          \n",
            "                                                                                                  \n",
            " conv2d_75 (Conv2D)             (None, 3, 3, 192)    331776      ['activation_74[0][0]']          \n",
            "                                                                                                  \n",
            " batch_normalization_71 (BatchN  (None, 3, 3, 320)   960         ['conv2d_71[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " batch_normalization_75 (BatchN  (None, 3, 3, 192)   576         ['conv2d_75[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " activation_71 (Activation)     (None, 3, 3, 320)    0           ['batch_normalization_71[0][0]'] \n",
            "                                                                                                  \n",
            " activation_75 (Activation)     (None, 3, 3, 192)    0           ['batch_normalization_75[0][0]'] \n",
            "                                                                                                  \n",
            " max_pooling2d_3 (MaxPooling2D)  (None, 3, 3, 768)   0           ['mixed7[0][0]']                 \n",
            "                                                                                                  \n",
            " mixed8 (Concatenate)           (None, 3, 3, 1280)   0           ['activation_71[0][0]',          \n",
            "                                                                  'activation_75[0][0]',          \n",
            "                                                                  'max_pooling2d_3[0][0]']        \n",
            "                                                                                                  \n",
            " conv2d_80 (Conv2D)             (None, 3, 3, 448)    573440      ['mixed8[0][0]']                 \n",
            "                                                                                                  \n",
            " batch_normalization_80 (BatchN  (None, 3, 3, 448)   1344        ['conv2d_80[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " activation_80 (Activation)     (None, 3, 3, 448)    0           ['batch_normalization_80[0][0]'] \n",
            "                                                                                                  \n",
            " conv2d_77 (Conv2D)             (None, 3, 3, 384)    491520      ['mixed8[0][0]']                 \n",
            "                                                                                                  \n",
            " conv2d_81 (Conv2D)             (None, 3, 3, 384)    1548288     ['activation_80[0][0]']          \n",
            "                                                                                                  \n",
            " batch_normalization_77 (BatchN  (None, 3, 3, 384)   1152        ['conv2d_77[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " batch_normalization_81 (BatchN  (None, 3, 3, 384)   1152        ['conv2d_81[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " activation_77 (Activation)     (None, 3, 3, 384)    0           ['batch_normalization_77[0][0]'] \n",
            "                                                                                                  \n",
            " activation_81 (Activation)     (None, 3, 3, 384)    0           ['batch_normalization_81[0][0]'] \n",
            "                                                                                                  \n",
            " conv2d_78 (Conv2D)             (None, 3, 3, 384)    442368      ['activation_77[0][0]']          \n",
            "                                                                                                  \n",
            " conv2d_79 (Conv2D)             (None, 3, 3, 384)    442368      ['activation_77[0][0]']          \n",
            "                                                                                                  \n",
            " conv2d_82 (Conv2D)             (None, 3, 3, 384)    442368      ['activation_81[0][0]']          \n",
            "                                                                                                  \n",
            " conv2d_83 (Conv2D)             (None, 3, 3, 384)    442368      ['activation_81[0][0]']          \n",
            "                                                                                                  \n",
            " average_pooling2d_7 (AveragePo  (None, 3, 3, 1280)  0           ['mixed8[0][0]']                 \n",
            " oling2D)                                                                                         \n",
            "                                                                                                  \n",
            " conv2d_76 (Conv2D)             (None, 3, 3, 320)    409600      ['mixed8[0][0]']                 \n",
            "                                                                                                  \n",
            " batch_normalization_78 (BatchN  (None, 3, 3, 384)   1152        ['conv2d_78[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " batch_normalization_79 (BatchN  (None, 3, 3, 384)   1152        ['conv2d_79[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " batch_normalization_82 (BatchN  (None, 3, 3, 384)   1152        ['conv2d_82[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " batch_normalization_83 (BatchN  (None, 3, 3, 384)   1152        ['conv2d_83[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " conv2d_84 (Conv2D)             (None, 3, 3, 192)    245760      ['average_pooling2d_7[0][0]']    \n",
            "                                                                                                  \n",
            " batch_normalization_76 (BatchN  (None, 3, 3, 320)   960         ['conv2d_76[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " activation_78 (Activation)     (None, 3, 3, 384)    0           ['batch_normalization_78[0][0]'] \n",
            "                                                                                                  \n",
            " activation_79 (Activation)     (None, 3, 3, 384)    0           ['batch_normalization_79[0][0]'] \n",
            "                                                                                                  \n",
            " activation_82 (Activation)     (None, 3, 3, 384)    0           ['batch_normalization_82[0][0]'] \n",
            "                                                                                                  \n",
            " activation_83 (Activation)     (None, 3, 3, 384)    0           ['batch_normalization_83[0][0]'] \n",
            "                                                                                                  \n",
            " batch_normalization_84 (BatchN  (None, 3, 3, 192)   576         ['conv2d_84[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " activation_76 (Activation)     (None, 3, 3, 320)    0           ['batch_normalization_76[0][0]'] \n",
            "                                                                                                  \n",
            " mixed9_0 (Concatenate)         (None, 3, 3, 768)    0           ['activation_78[0][0]',          \n",
            "                                                                  'activation_79[0][0]']          \n",
            "                                                                                                  \n",
            " concatenate (Concatenate)      (None, 3, 3, 768)    0           ['activation_82[0][0]',          \n",
            "                                                                  'activation_83[0][0]']          \n",
            "                                                                                                  \n",
            " activation_84 (Activation)     (None, 3, 3, 192)    0           ['batch_normalization_84[0][0]'] \n",
            "                                                                                                  \n",
            " mixed9 (Concatenate)           (None, 3, 3, 2048)   0           ['activation_76[0][0]',          \n",
            "                                                                  'mixed9_0[0][0]',               \n",
            "                                                                  'concatenate[0][0]',            \n",
            "                                                                  'activation_84[0][0]']          \n",
            "                                                                                                  \n",
            " conv2d_89 (Conv2D)             (None, 3, 3, 448)    917504      ['mixed9[0][0]']                 \n",
            "                                                                                                  \n",
            " batch_normalization_89 (BatchN  (None, 3, 3, 448)   1344        ['conv2d_89[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " activation_89 (Activation)     (None, 3, 3, 448)    0           ['batch_normalization_89[0][0]'] \n",
            "                                                                                                  \n",
            " conv2d_86 (Conv2D)             (None, 3, 3, 384)    786432      ['mixed9[0][0]']                 \n",
            "                                                                                                  \n",
            " conv2d_90 (Conv2D)             (None, 3, 3, 384)    1548288     ['activation_89[0][0]']          \n",
            "                                                                                                  \n",
            " batch_normalization_86 (BatchN  (None, 3, 3, 384)   1152        ['conv2d_86[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " batch_normalization_90 (BatchN  (None, 3, 3, 384)   1152        ['conv2d_90[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " activation_86 (Activation)     (None, 3, 3, 384)    0           ['batch_normalization_86[0][0]'] \n",
            "                                                                                                  \n",
            " activation_90 (Activation)     (None, 3, 3, 384)    0           ['batch_normalization_90[0][0]'] \n",
            "                                                                                                  \n",
            " conv2d_87 (Conv2D)             (None, 3, 3, 384)    442368      ['activation_86[0][0]']          \n",
            "                                                                                                  \n",
            " conv2d_88 (Conv2D)             (None, 3, 3, 384)    442368      ['activation_86[0][0]']          \n",
            "                                                                                                  \n",
            " conv2d_91 (Conv2D)             (None, 3, 3, 384)    442368      ['activation_90[0][0]']          \n",
            "                                                                                                  \n",
            " conv2d_92 (Conv2D)             (None, 3, 3, 384)    442368      ['activation_90[0][0]']          \n",
            "                                                                                                  \n",
            " average_pooling2d_8 (AveragePo  (None, 3, 3, 2048)  0           ['mixed9[0][0]']                 \n",
            " oling2D)                                                                                         \n",
            "                                                                                                  \n",
            " conv2d_85 (Conv2D)             (None, 3, 3, 320)    655360      ['mixed9[0][0]']                 \n",
            "                                                                                                  \n",
            " batch_normalization_87 (BatchN  (None, 3, 3, 384)   1152        ['conv2d_87[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " batch_normalization_88 (BatchN  (None, 3, 3, 384)   1152        ['conv2d_88[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " batch_normalization_91 (BatchN  (None, 3, 3, 384)   1152        ['conv2d_91[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " batch_normalization_92 (BatchN  (None, 3, 3, 384)   1152        ['conv2d_92[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " conv2d_93 (Conv2D)             (None, 3, 3, 192)    393216      ['average_pooling2d_8[0][0]']    \n",
            "                                                                                                  \n",
            " batch_normalization_85 (BatchN  (None, 3, 3, 320)   960         ['conv2d_85[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " activation_87 (Activation)     (None, 3, 3, 384)    0           ['batch_normalization_87[0][0]'] \n",
            "                                                                                                  \n",
            " activation_88 (Activation)     (None, 3, 3, 384)    0           ['batch_normalization_88[0][0]'] \n",
            "                                                                                                  \n",
            " activation_91 (Activation)     (None, 3, 3, 384)    0           ['batch_normalization_91[0][0]'] \n",
            "                                                                                                  \n",
            " activation_92 (Activation)     (None, 3, 3, 384)    0           ['batch_normalization_92[0][0]'] \n",
            "                                                                                                  \n",
            " batch_normalization_93 (BatchN  (None, 3, 3, 192)   576         ['conv2d_93[0][0]']              \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " activation_85 (Activation)     (None, 3, 3, 320)    0           ['batch_normalization_85[0][0]'] \n",
            "                                                                                                  \n",
            " mixed9_1 (Concatenate)         (None, 3, 3, 768)    0           ['activation_87[0][0]',          \n",
            "                                                                  'activation_88[0][0]']          \n",
            "                                                                                                  \n",
            " concatenate_1 (Concatenate)    (None, 3, 3, 768)    0           ['activation_91[0][0]',          \n",
            "                                                                  'activation_92[0][0]']          \n",
            "                                                                                                  \n",
            " activation_93 (Activation)     (None, 3, 3, 192)    0           ['batch_normalization_93[0][0]'] \n",
            "                                                                                                  \n",
            " mixed10 (Concatenate)          (None, 3, 3, 2048)   0           ['activation_85[0][0]',          \n",
            "                                                                  'mixed9_1[0][0]',               \n",
            "                                                                  'concatenate_1[0][0]',          \n",
            "                                                                  'activation_93[0][0]']          \n",
            "                                                                                                  \n",
            "==================================================================================================\n",
            "Total params: 21,802,784\n",
            "Trainable params: 0\n",
            "Non-trainable params: 21,802,784\n",
            "__________________________________________________________________________________________________\n"
          ]
        }
      ],
      "source": [
        "pre_trained_model = create_pre_trained_model(local_weights_file)\n",
        "\n",
        "# Print the model summary\n",
        "pre_trained_model.summary()"
      ],
      "id": "ve7eh9iztT4q"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "4cAY2gQytr0-"
      },
      "source": [
        "**Expected Output:**\n",
        "```\n",
        "batch_normalization_v1_281 (Bat (None, 3, 3, 192)    576         conv2d_281[0][0]                 \n",
        "__________________________________________________________________________________________________\n",
        "activation_273 (Activation)     (None, 3, 3, 320)    0           batch_normalization_v1_273[0][0] \n",
        "__________________________________________________________________________________________________\n",
        "mixed9_1 (Concatenate)          (None, 3, 3, 768)    0           activation_275[0][0]             \n",
        "                                                                activation_276[0][0]             \n",
        "__________________________________________________________________________________________________\n",
        "concatenate_5 (Concatenate)     (None, 3, 3, 768)    0           activation_279[0][0]             \n",
        "                                                                activation_280[0][0]             \n",
        "__________________________________________________________________________________________________\n",
        "activation_281 (Activation)     (None, 3, 3, 192)    0           batch_normalization_v1_281[0][0] \n",
        "__________________________________________________________________________________________________\n",
        "mixed10 (Concatenate)           (None, 3, 3, 2048)   0           activation_273[0][0]             \n",
        "                                                                mixed9_1[0][0]                   \n",
        "                                                                concatenate_5[0][0]              \n",
        "                                                                activation_281[0][0]             \n",
        "==================================================================================================\n",
        "Total params: 21,802,784\n",
        "Trainable params: 0\n",
        "Non-trainable params: 21,802,784\n",
        "\n",
        "\n",
        "```"
      ],
      "id": "4cAY2gQytr0-"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "MRHkV9jo-hkh"
      },
      "source": [
        "To check that all the layers in the model were set to be non-trainable, you can also run the cell below:"
      ],
      "id": "MRHkV9jo-hkh"
    },
    {
      "cell_type": "code",
      "execution_count": 14,
      "metadata": {
        "id": "VASOaB8xDbhU",
        "tags": [
          "graded"
        ],
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "e03d5499-5bee-4f80-8d1e-317e64187b7e"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "There are 21,802,784 total parameters in this model.\n",
            "There are 0 trainable parameters in this model.\n"
          ]
        }
      ],
      "source": [
        "total_params = pre_trained_model.count_params()\n",
        "num_trainable_params = sum([w.shape.num_elements() for w in pre_trained_model.trainable_weights])\n",
        "\n",
        "print(f\"There are {total_params:,} total parameters in this model.\")\n",
        "print(f\"There are {num_trainable_params:,} trainable parameters in this model.\")"
      ],
      "id": "VASOaB8xDbhU"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "mRioO7FH5a8I"
      },
      "source": [
        "**Expected Output:**\n",
        "```\n",
        "There are 21,802,784 total parameters in this model.\n",
        "There are 0 trainable parameters in this model.\n",
        "```"
      ],
      "id": "mRioO7FH5a8I"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "dFtwDyKj-4GR"
      },
      "source": [
        "## Creating callbacks for later\n",
        "\n",
        "You have already worked with callbacks in the first course of this specialization so the callback to stop training once an accuracy of 99.9% is reached, is provided for you:"
      ],
      "id": "dFtwDyKj-4GR"
    },
    {
      "cell_type": "code",
      "execution_count": 15,
      "metadata": {
        "id": "SeVjZD2o7gWS",
        "tags": [
          "graded"
        ]
      },
      "outputs": [],
      "source": [
        "# Define a Callback class that stops training once accuracy reaches 99.9%\n",
        "class myCallback(tf.keras.callbacks.Callback):\n",
        "  def on_epoch_end(self, epoch, logs={}):\n",
        "    if(logs.get('accuracy')>0.999):\n",
        "      print(\"\\nReached 99.9% accuracy so cancelling training!\")\n",
        "      self.model.stop_training = True"
      ],
      "id": "SeVjZD2o7gWS"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "lHZnFl-5_p3a"
      },
      "source": [
        "## Pipelining the pre-trained model with your own\n",
        "\n",
        "Now that the pre-trained model is ready, you need to \"glue\" it to your own model to solve the task at hand.\n",
        "\n",
        "For this you will need the last output of the pre-trained model, since this will be the input for your own. Complete the `output_of_last_layer` function below.\n",
        "\n",
        "**Note:** For grading purposes use the `mixed7` layer as the last layer of the pre-trained model. However, after submitting feel free to come back here and play around with this."
      ],
      "id": "lHZnFl-5_p3a"
    },
    {
      "cell_type": "code",
      "execution_count": 18,
      "metadata": {
        "id": "CFsUlwdfs_wg",
        "tags": [
          "graded"
        ]
      },
      "outputs": [],
      "source": [
        "# GRADED FUNCTION: output_of_last_layer\n",
        "def output_of_last_layer(pre_trained_model):\n",
        "  \"\"\"\n",
        "  Gets the last layer output of a model\n",
        "  \n",
        "  Args:\n",
        "    pre_trained_model (tf.keras Model): model to get the last layer output from\n",
        "    \n",
        "  Returns:\n",
        "    last_output: output of the model's last layer \n",
        "  \"\"\"\n",
        "  ### START CODE HERE\n",
        "  last_desired_layer = pre_trained_model.get_layer('mixed7')\n",
        "  print('last layer output shape: ', last_desired_layer.output_shape)\n",
        "  last_output = last_desired_layer.output\n",
        "  print('last layer output: ', last_output)\n",
        "  ### END CODE HERE\n",
        "\n",
        "  return last_output\n"
      ],
      "id": "CFsUlwdfs_wg"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "13AEzKG2A6_J"
      },
      "source": [
        "Check that everything works as expected:"
      ],
      "id": "13AEzKG2A6_J"
    },
    {
      "cell_type": "code",
      "execution_count": 19,
      "metadata": {
        "id": "zOJPUtMN6PHo",
        "tags": [
          "graded"
        ],
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "dac066bc-0bde-45c4-abc6-5f2f64c4cc78"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "last layer output shape:  (None, 7, 7, 768)\n",
            "last layer output:  KerasTensor(type_spec=TensorSpec(shape=(None, 7, 7, 768), dtype=tf.float32, name=None), name='mixed7/concat:0', description=\"created by layer 'mixed7'\")\n"
          ]
        }
      ],
      "source": [
        "last_output = output_of_last_layer(pre_trained_model)"
      ],
      "id": "zOJPUtMN6PHo"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "XqIWKZ_h7CuY"
      },
      "source": [
        "**Expected Output (if `mixed7` layer was used):**\n",
        "```\n",
        "last layer output shape:  (None, 7, 7, 768)\n",
        "last layer output:  KerasTensor(type_spec=TensorSpec(shape=(None, 7, 7, 768), dtype=tf.float32, name=None), name='mixed7/concat:0', description=\"created by layer 'mixed7'\")\n",
        "```"
      ],
      "id": "XqIWKZ_h7CuY"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "0Rp-J6JuwJTq"
      },
      "source": [
        "Now you will create the final model by adding some additional layers on top of the pre-trained model.\n",
        "\n",
        "Complete the `create_final_model` function below. You will need to use Tensorflow's [Functional API](https://www.tensorflow.org/guide/keras/functional) for this since the pretrained model has been created using it. \n",
        "\n",
        "Let's double check this first:"
      ],
      "id": "0Rp-J6JuwJTq"
    },
    {
      "cell_type": "code",
      "execution_count": 20,
      "metadata": {
        "id": "cKQknB4j7K9y",
        "tags": [
          "graded"
        ],
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "585c9aca-1b37-4ead-d473-9f1032eda891"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "The pretrained model has type: <class 'keras.engine.functional.Functional'>\n"
          ]
        }
      ],
      "source": [
        "# Print the type of the pre-trained model\n",
        "print(f\"The pretrained model has type: {type(pre_trained_model)}\")"
      ],
      "id": "cKQknB4j7K9y"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Kt7AU7jP7LW9"
      },
      "source": [
        "To create the final model, you will use Keras' Model class by defining the appropriate inputs and outputs as described in the first way to instantiate a Model in the [docs](https://www.tensorflow.org/api_docs/python/tf/keras/Model).\n",
        "\n",
        "Note that you can get the input from any existing model by using its `input` attribute and by using the Funcional API you can use the last layer directly as output when creating the final model."
      ],
      "id": "Kt7AU7jP7LW9"
    },
    {
      "cell_type": "code",
      "execution_count": 21,
      "metadata": {
        "cellView": "code",
        "id": "BMXb913pbvFg",
        "tags": [
          "graded"
        ]
      },
      "outputs": [],
      "source": [
        "# GRADED FUNCTION: create_final_model\n",
        "def create_final_model(pre_trained_model, last_output):\n",
        "  \"\"\"\n",
        "  Appends a custom model to a pre-trained model\n",
        "  \n",
        "  Args:\n",
        "    pre_trained_model (tf.keras Model): model that will accept the train/test inputs\n",
        "    last_output (tensor): last layer output of the pre-trained model\n",
        "    \n",
        "  Returns:\n",
        "    model: the combined model\n",
        "  \"\"\"\n",
        "  # Flatten the output layer to 1 dimension\n",
        "  x = layers.Flatten()(last_output)\n",
        "\n",
        "  ### START CODE HERE\n",
        "\n",
        "  # Add a fully connected layer with 1024 hidden units and ReLU activation\n",
        "  x = layers.Dense(1024, activation='relu')(x)\n",
        "  # Add a dropout rate of 0.2\n",
        "  x = layers.Dropout(0.2)(x)  \n",
        "  # Add a final sigmoid layer for classification\n",
        "  x = layers.Dense  (1, activation='sigmoid')(x)        \n",
        "\n",
        "  # Create the complete model by using the Model class\n",
        "  model = Model(inputs=pre_trained_model.input, outputs=x)\n",
        "\n",
        "  # Compile the model\n",
        "  model.compile(optimizer = RMSprop(learning_rate=0.0001), \n",
        "                loss = 'binary_crossentropy',\n",
        "                metrics = ['accuracy'])\n",
        "\n",
        "  ### END CODE HERE\n",
        "  \n",
        "  return model\n"
      ],
      "id": "BMXb913pbvFg"
    },
    {
      "cell_type": "code",
      "execution_count": 22,
      "metadata": {
        "id": "cL6ga5Z1783H",
        "tags": [
          "graded"
        ],
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "7944784d-b6d8-4d9b-d987-c5f78d784281"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "There are 47,512,481 total parameters in this model.\n",
            "There are 38,537,217 trainable parameters in this model.\n"
          ]
        }
      ],
      "source": [
        "# Save your model in a variable\n",
        "model = create_final_model(pre_trained_model, last_output)\n",
        "\n",
        "# Inspect parameters\n",
        "total_params = model.count_params()\n",
        "num_trainable_params = sum([w.shape.num_elements() for w in model.trainable_weights])\n",
        "\n",
        "print(f\"There are {total_params:,} total parameters in this model.\")\n",
        "print(f\"There are {num_trainable_params:,} trainable parameters in this model.\")"
      ],
      "id": "cL6ga5Z1783H"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "J4d3zlcQDrvm"
      },
      "source": [
        "**Expected Output:**\n",
        "```\n",
        "There are 47,512,481 total parameters in this model.\n",
        "There are 38,537,217 trainable parameters in this model.\n",
        "```"
      ],
      "id": "J4d3zlcQDrvm"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_eqwHj5xEBZ7"
      },
      "source": [
        "Wow, that is a lot of parameters!\n",
        "\n",
        "After submitting your assignment later, try re-running this notebook but use the original resolution of 300x300, you will be surprised to see how many more parameters are for that case.\n",
        "\n",
        "Now train the model:"
      ],
      "id": "_eqwHj5xEBZ7"
    },
    {
      "cell_type": "code",
      "execution_count": 23,
      "metadata": {
        "id": "Blhq2MAUeyGA",
        "tags": [],
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "246ad35c-b9d6-44f3-f5fd-b8fd72dcb266"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/100\n",
            "33/33 - 17s - loss: 0.0268 - accuracy: 0.9864 - val_loss: 0.0304 - val_accuracy: 0.9844 - 17s/epoch - 530ms/step\n",
            "Epoch 2/100\n",
            "\n",
            "Reached 99.9% accuracy so cancelling training!\n",
            "33/33 - 7s - loss: 5.7326e-05 - accuracy: 1.0000 - val_loss: 0.0912 - val_accuracy: 0.9766 - 7s/epoch - 201ms/step\n"
          ]
        }
      ],
      "source": [
        "# Run this and see how many epochs it should take before the callback\n",
        "# fires, and stops training at 99.9% accuracy\n",
        "# (It should take a few epochs)\n",
        "callbacks = myCallback()\n",
        "history = model.fit(train_generator,\n",
        "                    validation_data = validation_generator,\n",
        "                    epochs = 100,\n",
        "                    verbose = 2,\n",
        "                    callbacks=callbacks)"
      ],
      "id": "Blhq2MAUeyGA"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Y94djl4t0sK5"
      },
      "source": [
        "The training should have stopped after less than 10 epochs and it should have reached an accuracy over 99,9% (firing the callback). This happened so quickly because of the pre-trained model you used, which already contained information to classify humans from horses. Really cool!\n",
        "\n",
        "Now take a quick look at the training and validation accuracies for each epoch of training:"
      ],
      "id": "Y94djl4t0sK5"
    },
    {
      "cell_type": "code",
      "execution_count": 24,
      "metadata": {
        "id": "C2Fp6Se9rKuL",
        "tags": [],
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 298
        },
        "outputId": "f9cb15d7-9f6e-4fcb-868d-91dc9592404b"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYAAAAEICAYAAABWJCMKAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3deXxU1f3/8deHsIkgaFBZAgQruxASAihWRdFKxa8ICoIrsXWt+tPWWm2tWq31a6XfKt9WLVYTwQUrrUpF1Lp9sVWWsAoIChglbCL7DoHz++PchDshCQNZJjPzfj4e82DmnnvvnDsZ7mfOOfeejznnEBGR5FMn1hUQEZHYUAAQEUlSCgAiIklKAUBEJEkpAIiIJCkFABGRJKUAICXMbIqZXVPV68aSmRWY2bnVsF9nZicHz582s19Hs+4RvM8VZvbukdZTpCKm+wDim5ltC71sBOwG9gWvb3DOvVjztao9zKwA+LFz7r0q3q8DOjjnllbVumaWDnwF1HPOFVVFPUUqUjfWFZDKcc41Ln5e0cnOzOrqpCK1hb6PtYO6gBKUmfU3s0Iz+4WZrQFyzexYM3vTzNaZ2cbgeVpom4/M7MfB81Fm9m8zGx2s+5WZ/fAI121vZlPNbKuZvWdmfzazF8qpdzR1fMjM/hPs710zax4qv8rMvjaz9Wb2qwo+n75mtsbMUkLLhpjZ/OB5HzP71Mw2mdlqM/uTmdUvZ195Zvbb0OufB9usMrNrS607yMzmmNkWM1thZg+EiqcG/24ys21mdlrxZxvavp+ZzTSzzcG//aL9bA7zcz7OzHKDY9hoZq+Hygab2dzgGJaZ2cBgeUR3m5k9UPx3NrP0oCvsR2b2DfBBsPzV4O+wOfiOdAttf5SZ/SH4e24OvmNHmdlkM7u11PHMN7MhZR2rlE8BILG1AI4D2gHX4//eucHrtsBO4E8VbN8XWAI0B34PPGtmdgTrvgTMAFKBB4CrKnjPaOp4OZADnADUB+4EMLOuwFPB/lsF75dGGZxz04HtwDml9vtS8HwfcEdwPKcBA4CbK6g3QR0GBvU5D+gAlB5/2A5cDTQDBgE3mdnFQdmZwb/NnHONnXOfltr3ccBkYExwbP8DTDaz1FLHcNBnU4ZDfc7j8V2K3YJ9/TGoQx9gHPDz4BjOBArK+zzKcBbQBTg/eD0F/zmdAMwGwl2Wo4FeQD/89/guYD/wPHBl8UpmlgG0xn82cjicc3okyAP/H/Hc4Hl/YA/QsIL1ewIbQ68/wnchAYwClobKGgEOaHE46+JPLkVAo1D5C8ALUR5TWXW8N/T6ZuDt4Pl9wIRQ2dHBZ3BuOfv+LfBc8LwJ/uTcrpx1bwdeC712wMnB8zzgt8Hz54D/Dq3XMbxuGft9HPhj8Dw9WLduqHwU8O/g+VXAjFLbfwqMOtRnczifM9ASf6I9toz1/lJc34q+f8HrB4r/zqFjO6mCOjQL1mmKD1A7gYwy1msIbMSPq4APFE/W9P+3RHioBZDY1jnndhW/MLNGZvaXoEm9Bd/l0CzcDVLKmuInzrkdwdPGh7luK2BDaBnAivIqHGUd14Se7wjVqVV438657cD68t4L/2t/qJk1AIYCs51zXwf16Bh0i6wJ6vE7fGvgUCLqAHxd6vj6mtmHQdfLZuDGKPdbvO+vSy37Gv/rt1h5n02EQ3zObfB/s41lbNoGWBZlfctS8tmYWYqZ/XfQjbSFAy2J5sGjYVnvFXynXwGuNLM6wEh8i0UOkwJAYit9idfPgE5AX+fcMRzociivW6cqrAaOM7NGoWVtKli/MnVcHd538J6p5a3snFuEP4H+kMjuH/BdSYvxvzKPAX55JHXAt4DCXgImAW2cc02Bp0P7PdQleavwXTZhbYGVUdSrtIo+5xX4v1mzMrZbAXyvnH1ux7f+irUoY53wMV4ODMZ3kzXFtxKK6/AdsKuC93oeuALfNbfDleouk+goACSXJvhm9aagP/n+6n7D4Bd1PvCAmdU3s9OA/6qmOk4ELjSz7wcDtg9y6O/4S8D/w58AXy1Vjy3ANjPrDNwUZR3+Bowys65BACpd/yb4X9e7gv70y0Nl6/BdLyeVs++3gI5mdrmZ1TWzy4CuwJtR1q10Pcr8nJ1zq/F9808Gg8X1zKw4QDwL5JjZADOrY2atg88HYC4wIlg/G7g0ijrsxrfSGuFbWcV12I/vTvsfM2sVtBZOC1prBCf8/cAf0K//I6YAkFweB47C/7qaBrxdQ+97BX4gdT2+3/0V/H/8shxxHZ1zC4Gf4E/qq/H9xIWH2Oxl/MDkB86570LL78SfnLcCzwR1jqYOU4Jj+ABYGvwbdjPwoJltxY9Z/C207Q7gYeA/5q8+OrXUvtcDF+J/va/HD4peWKre0TrU53wVsBffCvoWPwaCc24GfpD5j8Bm4P840Cr5Nf4X+0bgN0S2qMoyDt8CWwksCuoRdifwGTAT2AA8SuQ5axzQHT+mJEdAN4JJjTOzV4DFzrlqb4FI4jKzq4HrnXPfj3Vd4pVaAFLtzKy3mX0v6DIYiO/3ff1Q24mUJ+heuxkYG+u6xDMFAKkJLfCXKG7DX8N+k3NuTkxrJHHLzM7Hj5es5dDdTFIBdQGJiCQptQBERJJUXE0G17x5c5eenh7raoiIxJVZs2Z955w7vvTyuAoA6enp5Ofnx7oaIiJxxcxK30EOqAtIRCRpKQCIiCQpBQARkSQVV2MAZdm7dy+FhYXs2rXr0CtLUmjYsCFpaWnUq1cv1lURqdXiPgAUFhbSpEkT0tPTKT9XiSQL5xzr16+nsLCQ9u3bx7o6IrVaVF1AZvacmX1rZgvKKTczG2NmS4PUbFmhsmvM7MvgcU1oeS8z+yzYZkwFmaYqtGvXLlJTU3XyFwDMjNTUVLUIRaIQ7RhAHjCwgvIf4tO6dcCnHnwKSlLY3Y9PF9gHuN/Mjg22eQq4LrRdRfuvkE7+Eqbvg0h0ogoAzrmp+OlYyzMYGOe8afjMQi3xeT//5Zwrzi70L2BgUHaMc26a83NRjAMuLnfvIiLJyDmYNg3uuAP27q3y3VfVVUCtiUyDVxgsq2h5YRnLD2Jm15tZvpnlr1u3roqqW3XWr19Pz5496dmzJy1atKB169Ylr/fs2VPhtvn5+dx2222HfI9+/fpVVXVFJB6sWgWPPgpdu8Jpp8Ff/gKffVblb1PrB4Gdc2MJpnzNzs6udTPXpaamMnfuXAAeeOABGjduzJ133llSXlRURN26ZX/M2dnZZGdnH/I9Pvnkk6qpbA3at28fKSnlpRoWkYPs3g2TJkFeHrz9NuzfD6efDn/9KwwbBsccU+VvWVUtgJVE5kFNC5ZVtDytjOUJYdSoUdx444307duXu+66ixkzZnDaaaeRmZlJv379WLJkCQAfffQRF154IeCDx7XXXkv//v056aSTGDNmTMn+GjduXLJ+//79ufTSS+ncuTNXXHEFxbO5vvXWW3Tu3JlevXpx2223lew3rKCggDPOOIOsrCyysrIiAsujjz5K9+7dycjI4O677wZg6dKlnHvuuWRkZJCVlcWyZcsi6gxwyy23kJeXB/ipOn7xi1+QlZXFq6++yjPPPEPv3r3JyMjgkksuYccOnxd+7dq1DBkyhIyMDDIyMvjkk0+47777ePzxx0v2+6tf/Yonnnii0n8LkVrNOZg9G269FVq1guHDYd48uPtu+OIL+Pe/4Uc/qpaTP1RdC2AScIuZTcAP+G52zq02s3eA34UGfn8A3OOc22BmW4KUd9OBq4H/rXQtbr8dgl/jVaZnTwidmKJVWFjIJ598QkpKClu2bOHjjz+mbt26vPfee/zyl7/k73//+0HbLF68mA8//JCtW7fSqVMnbrrppoOuZZ8zZw4LFy6kVatWnH766fznP/8hOzubG264galTp9K+fXtGjhxZZp1OOOEE/vWvf9GwYUO+/PJLRo4cSX5+PlOmTOGNN95g+vTpNGrUiA0b/HDPFVdcwd13382QIUPYtWsX+/fvZ8WKFWXuu1hqaiqzZ88GfPfYddddB8C9997Ls88+y6233sptt93GWWedxWuvvca+ffvYtm0brVq1YujQodx+++3s37+fCRMmMGPGjMP+3EXiwrffwosvQm6u79pp0ACGDIGcHBgwAGqo9RxVADCzl4H+QHMzK8Rf2VMPwDn3ND5Z9QX4HKg78DlDCU70D+FzegI86JwrHky+GX910VH4BNRTKn84tcewYcNKukA2b97MNddcw5dffomZsbecwZxBgwbRoEEDGjRowAknnMDatWtJS0uLWKdPnz4ly3r27ElBQQGNGzfmpJNOKrnufeTIkYwde3CipL1793LLLbcwd+5cUlJS+OKLLwB47733yMnJoVGjRgAcd9xxbN26lZUrVzJkyBDA31wVjcsuu6zk+YIFC7j33nvZtGkT27Zt4/zzzwfggw8+YNy4cQCkpKTQtGlTmjZtSmpqKnPmzGHt2rVkZmaSmpoa1XuKxIW9e+Gtt/xJf/JkKCqCPn3gySdhxAg49thD76OKRRUAnHNl/6Q8UO7wybjLKnsOeK6M5fnAKdG8f9SO4Jd6dTn66KNLnv/617/m7LPP5rXXXqOgoID+/fuXuU2DBg1KnqekpFBUVHRE65Tnj3/8IyeeeCLz5s1j//79UZ/Uw+rWrcv+/ftLXpe+3j583KNGjeL1118nIyODvLw8Pvroowr3/eMf/5i8vDzWrFnDtddee9h1E6mVPvvMn/RfeAHWrYMTT/S9FaNGQbduMa2a5gKqAZs3b6Z1a3+RU3F/eVXq1KkTy5cvp6CgAIBXXnml3Hq0bNmSOnXqMH78ePbt2wfAeeedR25ubkkf/YYNG2jSpAlpaWm8/rpP3bt792527NhBu3btWLRoEbt372bTpk28//775dZr69attGzZkr179/Liiy+WLB8wYABPPfUU4AeLN2/eDMCQIUN4++23mTlzZklrQSQubdgAf/4zZGdDjx7wpz/BGWfAP/8JhYXw2GMxP/mDAkCNuOuuu7jnnnvIzMw8rF/s0TrqqKN48sknGThwIL169aJJkyY0bdr0oPVuvvlmnn/+eTIyMli8eHHJr/WBAwdy0UUXkZ2dTc+ePRk9ejQA48ePZ8yYMfTo0YN+/fqxZs0a2rRpw/DhwznllFMYPnw4mZmZ5dbroYceom/fvpx++ul07ty5ZPkTTzzBhx9+SPfu3enVqxeLFi0CoH79+px99tkMHz5cVxBJ/CkqgilT/EBuy5Zwyy2wbx888YS/rPPvf4cLL4RyrgqMhbjKCZydne1KJ4T5/PPP6dKlS4xqVHts27aNxo0b45zjJz/5CR06dOCOO+6IdbUOy/79+0uuIOrQoUOl9qXvhdSYJUt8F8/48f5En5oKV17pu3h69ox17QAws1nOuYOuOVcLIEE888wz9OzZk27durF582ZuuOGGWFfpsCxatIiTTz6ZAQMGVPrkL1LtNm+GsWOhXz/o3BlGj4Zevfyv/FWr/HhkLTn5V0QtAElI+l5Ildu/Hz74wN+o9Y9/wM6d/k7dnBz/i79Fi1jXsFzltQBqT2eUiEhttGwZPP+8f3zzDTRr5rt3cnL8IG8cTz6oACAiUtq2bTBxou/bnzrVn+R/8AP4/e9h8GA4gkuoayMFABER8NMyfPyxP+m/+ips3w4dOsDvfgdXXQWlbspMBAoAIpLcvvnGd+/k5cHy5dC4sb8zNyfHD/LGcRfPoegqoEo6++yzeeeddyKWPf7449x0003lbtO/f3+KB7MvuOACNm3adNA6DzzwQMn1+OV5/fXXS66hB7jvvvt47733Dqf6Islp504/F89550F6Otx3H7RrB+PGwZo1fgbO009P6JM/KABU2siRI5kwYULEsgkTJpQ7IVtpb731Fs2aNTui9y4dAB588EHOPffcI9pXrBTfjSxS7YqTq9xwg79i58orYelSuP9++Oorf4XPVVdBaDqTRKcAUEmXXnopkydPLkn+UlBQwKpVqzjjjDO46aabyM7Oplu3btx///1lbp+ens53330HwMMPP0zHjh35/ve/XzJlNFDmtMqffPIJkyZN4uc//zk9e/Zk2bJljBo1iokTJwLw/vvvk5mZSffu3bn22mvZvXt3yfvdf//9ZGVl0b17dxYvXnxQnTRttCSU0slVXngBLr7Yn/CXLfMBID091rWMiYQaA4jFbNDHHXccffr0YcqUKQwePJgJEyYwfPhwzIyHH36Y4447jn379jFgwADmz59Pjx49ytzPrFmzmDBhAnPnzqWoqIisrCx69eoFwNChQ8ucVvmiiy7iwgsv5NJLL43Y165duxg1ahTvv/8+HTt25Oqrr+app57i9ttvB6B58+bMnj2bJ598ktGjR/PXv/41YntNGy1xrzi5Sm4uvPOOv4b/+9+v1uQq8UgtgCoQ7gYKd//87W9/Iysri8zMTBYuXBjRXVPaxx9/zJAhQ2jUqBHHHHMMF110UUnZggULOOOMM+jevTsvvvgiCxcurLA+S5YsoX379nTs2BGAa665hqlTp5aUDx06FIBevXqVTCAXtnfvXq677jq6d+/OsGHDSuod7bTRxeUVKT1tdFnH98EHH5SMpRRPG52enl4ybfS7776raaPlAOdg1qzI5CqffXYgucrHH1drcpV4lFAtgFjNBj148GDuuOMOZs+ezY4dO+jVqxdfffUVo0ePZubMmRx77LGMGjXqoKmTo3W40yofSvGU0uVNJ61poyWu1JLkKvFILYAq0LhxY84++2yuvfbakl//W7Zs4eijj6Zp06asXbuWKVMqzndz5pln8vrrr7Nz5062bt3KP//5z5Ky8qZVbtKkCVu3bj1oX506daKgoIClS5cCflbPs846K+rj0bTRUuvt3QtvvOH78lu3hp/+FI46Cp56yl/F8/LL/sYtnfwrpABQRUaOHMm8efNKAkBGRgaZmZl07tyZyy+/nNNPP73C7bOysrjsssvIyMjghz/8Ib179y4pK29a5REjRvDYY4+RmZnJsmXLSpY3bNiQ3Nxchg0bRvfu3alTpw433nhj1MeiaaOl1vrsM3+yb93an/ynT4c77oAFC/zzG2/0UzVIVDQZnMSdaKaN1vcigWzYAC+95G/UmjUL6tWDiy7y8/EMHFir5tevrTQdtCQETRudJIqKfP7c4uQqt94amVxl4sRal1wlHunTk7jStWtXli9fHutqSHVZvNj/0h83DlavhubN4aabalVylUSSEAHAOYcl+C3bEr146tYUfHKVV17xV/FMm+YHbi+4wF/FM2gQ1K8f6xomrLgPAA0bNmT9+vWkpqYqCAjOOdavX39El65KDSpOrpKb65Or7Nrl79R97LFan1wlkcR9AEhLS6OwsJB169bFuipSSzRs2JC0BJy6NyEsW+a7eJ5/Hlas8Ffs5OQkRHKVeBT3AaBevXq0b98+1tUQkfKUl1zlsccSKrlKPIr7ACAitVASJleJRwoAIlJ1SidXadIERo70V/EkeHKVeKQAICKVs2MHvPaaP+m//77/9X/OOfDAAzB0aFLNrx9vFABE5PAVJ1fJy4MJE2DLFj+n/v33wzXXJO38+vFGAUBEordqFYwf70/8ixdDo0Zw6aX+Kp4zz4Q6mlwgnigAiEjFykuu8uyzPrlKkyaxrqEcIQUAETmYczB7tj/pv/QSbNzor9y55x7fxaN5mBKCAoCIHPDttz5nbl6ekqskAQUAkWS3dy9MnuxP+pMn+5k4+/TxyVVGjND8+gksqhEbMxtoZkvMbKmZ3V1GeTsze9/M5pvZR2aWFip71MwWBI/LQsvzzOwrM5sbPDTVn0hNCidXGTLkQHKVhQuVXCVJHLIFYGYpwJ+B84BCYKaZTXLOhTOcjwbGOeeeN7NzgEeAq8xsEJAF9AQaAB+Z2RTn3JZgu5875yZW4fGISEWKk6vk5vo+/uLkKjk5cP75ml8/yUTTAugDLHXOLXfO7QEmAINLrdMV+CB4/mGovCsw1TlX5JzbDswHBla+2iISteLkKsOGHUiu4lxkcpVBg3TyT0LRBIDWwIrQ68JgWdg8YGjwfAjQxMxSg+UDzayRmTUHzgbahLZ7OOg2+qOZNSjrzc3sejPLN7N8zfgpchgWL4a774a2bf0J/qOPfHKVuXP9r//bbvMJVyRpVVXIvxP4k5mNAqYCK4F9zrl3zaw38AmwDvgU2Bdscw+wBqgPjAV+ATxYesfOubFBOdnZ2cr0IVIRJVeRwxBNAFhJ5K/2tGBZCefcKoIWgJk1Bi5xzm0Kyh4GHg7KXgK+CJavDjbfbWa5+CAiIoerrOQq3brB6NE+ucqJJ8a6hlJLRRMAZgIdzKw9/sQ/Arg8vELQvbPBObcf/8v+uWB5CtDMObfezHoAPYB3g7KWzrnV5tN4XQwsqKJjEkkOZSVXufZaP/OmkqtIFA4ZAJxzRWZ2C/AOkAI855xbaGYPAvnOuUlAf+ARM3P4LqCfBJvXAz4OUjVuAa50zhUFZS+a2fGAAXOBG6vusEQS1LZtfn79vDyfXKVOHSVXkSNm8ZRAOzs72+Xn58e6GiI1yzl/ss/Li0yukpMDV1/tr+MXqYCZzXLOZZderuu+RGqr8pKr5OTAaaepi0cqTQFApDYpTq6Sm+sHdouTq/zmN/5uXSVXkSqkACASa8XJVXJz/SWcW7ZA+/Y+o9bVVyu5ilQbBQCRWFFyFYkxBQCRmqTkKlKLKACIVLeKkquMGgUnnxzrGkqSUgAQqS7FyVVyc2HBAp9cZehQ38VzzjlKriIxpwAgUpXKSq7St6+Sq0itpAAgUhXmz/cn/RdegHXroEULn1xl1Cjo2jXWtRMpkwKAyJFSchWJc/qGihyOoiJ4911/0p80CfbsgcxMGDPG36Wr+fUljigAiERj8WJ/0h8/Hlav9if6m2/2XTwZGbGuncgRUQAQKc/mzTBhgu/bL06uMmiQP+kruYokAAUAkTAlV5EkogAgAuUnV8nJgV69NPOmJCQFAElexclVcnPh448PJFcZPdpfzaPkKpLgFAAkuZSVXKVjR/jd75RcRZKOAoAkh6+/hnHjlFxFJEQBQBJXRclVhg710y+LJDEFAEksFSVXueYaaNcu1jUUqTUUACQxrFp1oItnyRL/637YMH/NvpKriJRJAUDi165dfjqGvLwDyVXOOAPuukvJVUSioAAg8cU5mDXLn/SVXEWkUhQAJD6UTq7SsCEMGaLkKiKVoAAgtVdxcpXcXHjrrQPJVZ5+Gi67TMlVRCpJAUBqn/nz/Un/xRcPJFf56U/9VTxKriJSZRQApHZYvx5eflnJVURqkP5XSexUlFzl8sshNTXWNRRJaAoAUvOUXEWkVlAAkJpRXnKVnBy44AIlVxGJAQUAqT779vk5ePLylFxFpBZSAJCqt3SpT6yi5CoitZoCgFSNrVth4kQlVxGJI1HNkGVmA81siZktNbO7yyhvZ2bvm9l8M/vIzNJCZY+a2YLgcVloeXszmx7s8xUzUydwvNm/H/7v//zgbcuW/lf+2rXwyCPwzTcwZQoMH66Tv0gtdcgAYGYpwJ+BHwJdgZFmVvpunNHAOOdcD+BB4JFg20FAFtAT6AvcaWbHBNs8CvzROXcysBH4UeUPR2rE11/Dgw9Chw7Qv7/v37/8cvjPf/wVPnffrcxaInEgmhZAH2Cpc265c24PMAEYXGqdrsAHwfMPQ+VdganOuSLn3HZgPjDQzAw4B5gYrPc8cPGRH4ZUux07/Fw8557r59e//37/7/jxsGYNjB0L/fqpf18kjkQTAFoDK0KvC4NlYfOAocHzIUATM0sNlg80s0Zm1hw4G2gDpAKbnHNFFexTYs05+PRTuP5638Vz1VU+neIDD0BBAbz3nr+aR5m1ROJSVQ0C3wn8ycxGAVOBlcA+59y7ZtYb+ARYB3wK7DucHZvZ9cD1AG3btq2i6kqFVq70v+xLJ1fJyfHz7Su5ikhCiCYArMT/ai+WFiwr4ZxbRdACMLPGwCXOuU1B2cPAw0HZS8AXwHqgmZnVDVoBB+0ztO+xwFiA7OxsF/WRyeEpTq6Sm+unZyhOrvKLX8Cllyq5ikgCiiYAzAQ6mFl7/El6BHB5eIWge2eDc24/cA/wXLA8BWjmnFtvZj2AHsC7zjlnZh8Cl+LHFK4B3qiiY5JolZVcpU0b+OUv/cybSq4iktAOGQCcc0VmdgvwDpACPOecW2hmDwL5zrlJQH/gETNz+C6gnwSb1wM+9mO+bAGuDPX7/wKYYGa/BeYAz1bdYUmF1q71Uy0ruYpIUjPn4qdXJTs72+Xn58e6GvFpzx6fVKV0cpWcHCVXEUlwZjbLOZddernuBE505SVXGTUKunSJde1EJIYUABLR+vW+Tz8v70BylcGD/a/9H/xAyVVEBFAASBxFRfDOO/6kX5xcJStLyVVEpFwKAPHu88/9Sb90cpWcHOjRI9a1E5FaTAEgHm3aBK+84vv2p09XchUROSIKAPGiOLlKbi689pq/ceuUU+APf4ArrlByFRE5bAoAtd3Spb6LZ9w4n1zl2GPhRz/yV/EouYqIVIICQG2k5CoiUgMUAGqL/fth6lT/a3/iRNi+HTp29MlVrrpK8+uLSJVTAIi1r7/2uXPz8uCrr/yka5df7gd0Tz1VXTwiUm0UAGJhxw6fRSsvzw/sgp+D56GH/Jw8ml9fRGqAAkBNKU6ukpvrL+HcuhVOOgl+8xu4+mpo1y7WNRSRJKMAUN2UXEVEaikFgOqg5CoiEgcUAKpKcXKV3Fx4+WUlVxGRWk8BoLLWroUXXvBdPMXJVYYO9TdqKbmKiNRiCgBHYs8emDzZn/QnT/bTNJx6Kjz9tJKriEjcUAA4HPPm+ZP+Cy/Ad9/55Co/+5mSq4hIXFIAOJTi5Cq5uTBnjp9p86KLlFxFROKezl5lKU6ukpvrr+bZu9cnV/nf/4WRI5VcRUQSggJA2Oef+5P++PGwZg0cfzzccovv4lFyFRFJMAoAmzbBhAm+b3/6dN+lM2iQP+kruYqIJLDkDABKriIikmQBoLzkKjk5vo9fM2+KSBJJjgDw4ov+Gv1//9vPvXP++f7X/n/9l5KriEjSSo4A8OabsG6dkquIiIQkRwAYOxYaN1YXj6SnuVIAABDgSURBVIhISHIEAM2+KSJyEE1GLyKSpBQARESSlAKAiEiSUgAQEUlSCgAiIklKAUBEJElFFQDMbKCZLTGzpWZ2dxnl7czsfTObb2YfmVlaqOz3ZrbQzD43szFm/mL8YL0lZjY3eJxQdYclIiKHcsgAYGYpwJ+BHwJdgZFm1rXUaqOBcc65HsCDwCPBtv2A04EewClAb+Cs0HZXOOd6Bo9vK3swIiISvWhaAH2Apc655c65PcAEYHCpdboCHwTPPwyVO6AhUB9oANQD1la20iIiUnnRBIDWwIrQ68JgWdg8YGjwfAjQxMxSnXOf4gPC6uDxjnPu89B2uUH3z6+Lu4ZKM7PrzSzfzPLXrVsXRXVFRCQaVTUIfCdwlpnNwXfxrAT2mdnJQBcgDR80zjGzM4JtrnDOdQfOCB5XlbVj59xY51y2cy77+OOPr6LqiohINAFgJdAm9DotWFbCObfKOTfUOZcJ/CpYtgnfGpjmnNvmnNsGTAFOC8pXBv9uBV7CdzWJiEgNiSYAzAQ6mFl7M6sPjAAmhVcws+ZmVryve4Dnguff4FsGdc2sHr518HnwunmwbT3gQmBB5Q9HRESidcjZQJ1zRWZ2C/AOkAI855xbaGYPAvnOuUlAf+ARM3PAVOAnweYTgXOAz/ADwm875/5pZkcD7wQn/xTgPeCZqj20A+69F5Yvh1NPhb59oWdPaNCgut5NRCQ+mHMu1nWIWnZ2tsvPzz/s7X72M5/3fdUq/7p+fcjMPBAQTj0V0tOVLkBEEpOZzXLOZR+0PBkCQLHCQpg+HaZN8//m58POnb7s+OMPBIO+faF3b2jatIoqLiISQ+UFgORICBNIS/OPSy7xr/fuhQULDgSE6dN99kjwrYEuXSKDQrduUDepPjERSWRJ1QKIxsaNMHPmgaAwbRps2ODLjj4asrMPBIS+faFVq2qtjohIpakL6Ag5B8uWRXYdzZ3rWw8AbdpEthKysqBRoxqtoohIhRQAqtCuXTBnTmRQKCjwZXXrQo8ekUGhQweoo3lXRSRGFACq2dq1kQFh5kzYutWXHXss9OlzICj06QOpqbGtr4gkDwWAGrZvH3z+eWRQWLDAdymBbxWEWwk9evjLU0VEqpoCQC2wdau/9DQcFNas8WUNGkCvXpFBoW1b3ZsgIpWnAFALOQfffHPgEtRp02D2bD/GANCixYGrjU491V+B1KRJbOssIvFH9wHUQmbQrp1/DB/ul+3ZA/PnR7YS3njDl9Wp4+9FCAeFLl0gJSV2xyAi8UstgDiwfj3MmHEgKMyY4e9XAN8i6N07suvoxBNjW18RqV3UBZRAnIMvv4y8WW3+fCgq8uXp6ZGthMxMaNgwplUWkRhSAEhwO3b48YPweMKKII9bvXp+BtRwUPje9zTALJIsFACS0KpVBwJC8b0J27f7stTUyIDQpw80axbb+opI9VAAEIqKYNGiyK6jzz8/cG9Cp06RU2R3767J70QSgQKAlGnzZn9vQjgorFvny446yl96Gm4ppKXFtr4icvh0GaiUqWlTGDDAP8C3BgoKIqfIHjPGX54KfvbT8Gyo2dl+llQRiT8KABLBDNq394+RI/2y3bth3rzIoPCPf/iylBQ45ZTIrqNOnTT5nUg8UBeQHJF16/z9COGgsGWLL2va1N+bEG4pHH98bOsrksw0BiDVav9+WLIk8g7m+fP9coCTTooMCD17+vmPRKT6KQBIjdu+HWbNihxgXrXKl9Wv729QC3cdpafr3gSR6qAAILVCYWFkKyE/H3bu9GXHHx85pUXv3r47SUQqR1cBSa2QluYfl1ziX+/d6/MkhIPCm2/6MjM/2V04KHTrpnsTRKqKWgBS62zc6O9aDgeF9et92dFH+0tPw+MJrVrFtr4itZ26gCRuOQfLlkUGhLlzfesBoE2byFZCVhY0ahTbOovUJgoAklB27YI5cyKDQkGBL6tb16fYDAeFDh10b4IkLwUASXhr10bOhjpzpk/DCXDssX7Cu+KA0KePnxBPJBkoAEjS2bfPT3YXDgoLFx64N6FDh8hWQo8e/vJUkUSjACCCbxHk50cGhTVrfFnDhn78IBwU2rbVvQkS/xQARMrgnE+cE75ZbfZsP8YA0KJF5Gyo2dk+DadIPNF9ACJlMPO/8tu2heHD/bK9e/00FuGg8MYbvqxOHX8vQriV0KWLnxRPJN6oBSAShfXr/eR34QxrGzf6siZN/F3L4aBw4omxra9ImLqARKqQc/Dll5Gzoc6b57OugZ/XKNx1lJnpxxhEYqFSAcDMBgJPACnAX51z/12qvB3wHHA8sAG40jlXGJT9HhgE1AH+Bfw/55wzs15AHnAU8Fbx8orqoQAgtdnOnX78INx1tGKFL6tXz8+AGg4K3/ueBpilZhxxADCzFOAL4DygEJgJjHTOLQqt8yrwpnPueTM7B8hxzl1lZv2Ax4Azg1X/DdzjnPvIzGYAtwHT8QFgjHNuSkV1UQCQeLN6deTNajNn+llSwd+HEA4IffpAs2axra8kpsoMAvcBljrnlgc7mgAMBhaF1ukK/DR4/iHwevDcAQ2B+oAB9YC1ZtYSOMY5Ny3Y5zjgYqDCACASb1q2hIsv9g/wXUSLFkV2HU2Z4ruUwGdTC0+R3b27Jr+T6hPNV6s1sCL0uhDoW2qdecBQfDfREKCJmaU65z41sw+B1fgA8Cfn3Odmlh3sJ7zP1mW9uZldD1wP0LZt2yiqK1J7FU9T0aMHXH+9X7Zli28ZFAeFKVPg+ed92VFH+UtPwy2FtLTY1V8SS1X9trgT+JOZjQKmAiuBfWZ2MtAFKP7K/svMzgB2Rrtj59xYYCz4LqAqqq9IrXHMMTBggH+Abw0UFER2HY0ZA3v2+PJWrSJnQ83O9rOkihyuaALASqBN6HVasKyEc24VvgWAmTUGLnHObTKz64BpzrltQdkU4DRgPAeCQpn7FElWZtC+vX+MGOGX7d7trzIKB4V//MOXpaTAKadEdh116qTJ7+TQohkErosfBB6AP0nPBC53zi0MrdMc2OCc229mDwP7nHP3mdllwHXAQHwX0NvA4865f5YxCPy/zrm3KqqLBoFFDli3zt+bUBwQZsyAzZt9WdOmflC5uJXQt6/PuCbJ6YgHgZ1zRWZ2C/AO/jLQ55xzC83sQSDfOTcJ6A88YmYO3wX0k2DzicA5wGf4AeG3nXP/DMpu5sBloFPQALDIYTn+eBg0yD/AT3K3ZEnkPEePPOInxQM46aTIVkJGBjRoELv6S+zpRjCRBLZ9O8yaFdl1tDLobK1f39+gFg4K6em6NyER6U5gEQGgsDCylZCf729iA9+qCE9p0bu3706S+KbJ4EQE8JeRpqXBJZf410VF8NlnkUHhzTd9mZmf7C581VG3bro3IVGoBSAiB9m0KXLyu2nT/IR44C85zc6ODAqtWsW2vlIxdQGJyBFzDpYvj5znaO5cP3U2QJs2kV1HWVnQqFFs6ywHKACISJXatcsHgXBQKCjwZcV3PIeDQocOujchVhQARKTarV0bmTNhxgyfhhPg2GP9vQnFAaFPHz8hnlQ/BQARqXH79sHixZGT3y1Y4O9ZAN8qCLcSevTwl6dK1VIAEJFaYds2f+lpuOtozRpf1rChHz8IB4W2bXVvQmUpAIhIreScT5wTvllt1iw/xgDQokXkbKjZ2T4Np0RP9wGISK1k5n/lt20Lw4b5ZXv3wvz5kV1Hb7zhy+rU8fcihFsJXbr4SfHk8KgFICJxYcOGyMnvpk+HjRt9WZMm/q7lcFA48cTY1rc2UReQiCQU5+DLLyO7jubN83c2g5/XKNx1lJnpxxiSkQKAiCS8nTth9uzIoPDNN76sXj3o2TOylfC97yXHALMCgIgkpdWrI6e0mDnTz5IK/j6EcCuhTx9o1iy29a0OCgAiIvh7ExYujGwlLFrku5QAOneODArdu8f/5HcKACIi5diyxbcMwkHh22992VFH+UtPw0EhLa3i/dU2ugxURKQcxxwDAwb4B/jWwNdfR96sNmYM7Nnjy1u1ikyk06uXnyU13igAiIiUYuavIkpPhxEj/LLdu/1VRuHxhH/8w5elpMApp0QGhU6dav/kd+oCEhE5QuvWHcibMG2af755sy9r2tQPKhd3HfXt6zOuxYLGAEREqtn+/fDFF5FdR5995geeAU46KbKVkJEBDRpUf70UAEREYmD7dj+3UbjraOVKX1a/vr9BLRwU0tOr/t4EBQARkVqisDAyb8LMmf4mNvDdROGb1Xr39t1JlaGrgEREaom0NP+45BL/uqjI50kIdx29+aYvM/OT3U2c6P+tSgoAIiIxVreun6aiZ0+48Ua/bNMm3zIoDgqtW1fD+1b9LkVEpLKaNYPzzvOP6lLLr1IVEZHqogAgIpKkFABERJKUAoCISJJSABARSVIKACIiSUoBQEQkSSkAiIgkqbiaC8jM1gFfH+HmzYHvqrA68UDHnBx0zImvssfbzjl30GTUcRUAKsPM8suaDCmR6ZiTg4458VXX8aoLSEQkSSkAiIgkqWQKAGNjXYEY0DEnBx1z4quW402aMQAREYmUTC0AEREJUQAQEUlSCRcAzGygmS0xs6VmdncZ5Q3M7JWgfLqZpdd8LatWFMf8UzNbZGbzzex9M2sXi3pWpUMdc2i9S8zMmVlcXzIYzfGa2fDg77zQzF6q6TpWtSi+123N7EMzmxN8ty+IRT2rkpk9Z2bfmtmCcsrNzMYEn8l8M8uq1Bs65xLmAaQAy4CTgPrAPKBrqXVuBp4Ono8AXol1vWvgmM8GGgXPb0qGYw7WawJMBaYB2bGudzX/jTsAc4Bjg9cnxLreNXDMY4GbguddgYJY17sKjvtMIAtYUE75BcAUwIBTgemVeb9EawH0AZY655Y75/YAE4DBpdYZDDwfPJ8IDDAzq8E6VrVDHrNz7kPn3I7g5TQgrYbrWNWi+TsDPAQ8CuyqycpVg2iO9zrgz865jQDOuW9ruI5VLZpjdsAxwfOmwKoarF+1cM5NBTZUsMpgYJzzpgHNzKzlkb5fogWA1sCK0OvCYFmZ6zjnioDNQGqN1K56RHPMYT/C/4KIZ4c85qBp3MY5N7kmK1ZNovkbdwQ6mtl/zGyamQ2ssdpVj2iO+QHgSjMrBN4Cbq2ZqsXU4f5/r5CSwicRM7sSyAbOinVdqpOZ1QH+BxgV46rUpLr4bqD++BbeVDPr7pzbFNNaVa+RQJ5z7g9mdhow3sxOcc7tj3XF4kWitQBWAm1Cr9OCZWWuY2Z18U3H9TVSu+oRzTFjZucCvwIucs7trqG6VZdDHXMT4BTgIzMrwPeVTorjgeBo/saFwCTn3F7n3FfAF/iAEK+iOeYfAX8DcM59CjTET5qWyKL6/x6tRAsAM4EOZtbezOrjB3knlVpnEnBN8PxS4AMXjK7EqUMes5llAn/Bn/zjvW8YDnHMzrnNzrnmzrl051w6ftzjIudcfmyqW2nRfK9fx//6x8ya47uEltdkJatYNMf8DTAAwMy64APAuhqtZc2bBFwdXA10KrDZObf6SHeWUF1AzrkiM7sFeAd/FcFzzrmFZvYgkO+cmwQ8i28qLsUPtoyIXY0rL8pjfgxoDLwajHd/45y7KGaVrqQojzlhRHm87wA/MLNFwD7g5865uG3ZRnnMPwOeMbM78APCo+L8xxxm9jI+kDcPxjbuB+oBOOeexo91XAAsBXYAOZV6vzj/vERE5AglWheQiIhESQFARCRJKQCIiCQpBQARkSSlACAikqQUAEREkpQCgIhIkvr/rnse2KTPLHYAAAAASUVORK5CYII=\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 0 Axes>"
            ]
          },
          "metadata": {}
        }
      ],
      "source": [
        "# Plot the training and validation accuracies for each epoch\n",
        "\n",
        "acc = history.history['accuracy']\n",
        "val_acc = history.history['val_accuracy']\n",
        "loss = history.history['loss']\n",
        "val_loss = history.history['val_loss']\n",
        "\n",
        "epochs = range(len(acc))\n",
        "\n",
        "plt.plot(epochs, acc, 'r', label='Training accuracy')\n",
        "plt.plot(epochs, val_acc, 'b', label='Validation accuracy')\n",
        "plt.title('Training and validation accuracy')\n",
        "plt.legend(loc=0)\n",
        "plt.figure()\n",
        "\n",
        "plt.show()"
      ],
      "id": "C2Fp6Se9rKuL"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "g-4-4i9U1a0s"
      },
      "source": [
        "You will need to submit this notebook for grading. To download it, click on the `File` tab in the upper left corner of the screen then click on `Download` -> `Download .ipynb`. You can name it anything you want as long as it is a valid `.ipynb` (jupyter notebook) file."
      ],
      "id": "g-4-4i9U1a0s"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "7w54-pbB1W9r"
      },
      "source": [
        "**Congratulations on finishing this week's assignment!**\n",
        "\n",
        "You have successfully implemented a convolutional neural network that leverages a pre-trained network to help you solve the problem of classifying humans from horses.\n",
        "\n",
        "**Keep it up!**"
      ],
      "id": "7w54-pbB1W9r"
    }
  ],
  "metadata": {
    "accelerator": "GPU",
    "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": "3.7.4"
    },
    "colab": {
      "provenance": [],
      "include_colab_link": true
    }
  },
  "nbformat": 4,
  "nbformat_minor": 5
}