Attempt_on_African_attire_with_PyTorch.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/inuwamobarak/african_atire_dictator/blob/main/Attempt_on_African_attire_with_PyTorch.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "#Introduction"
      ],
      "metadata": {
        "id": "bPAan2Ry-MKe"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "stuck accuracy approzimately 50%"
      ],
      "metadata": {
        "id": "Q3wq_JsavspN"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "\n",
        "\n",
        "---\n",
        "\n"
      ],
      "metadata": {
        "id": "WI0yQ-MvxYih"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "NnlJKhiTAlyl"
      },
      "outputs": [],
      "source": [
        "#magic function for matplotlib graphs. Graphs will be included in notebook next to the code. \n",
        "%matplotlib inline"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "RM5MaNTCAlyq"
      },
      "outputs": [],
      "source": [
        "#importing dependencies\n",
        "from __future__ import print_function, division\n",
        "\n",
        "from datetime import datetime\n",
        "from PIL import Image\n",
        "import torch\n",
        "import torch.nn as nn\n",
        "import torch.optim as optim\n",
        "from torch.optim import lr_scheduler\n",
        "import numpy as np\n",
        "import torchvision\n",
        "from torchvision import datasets, models, transforms\n",
        "from torchvision.datasets import ImageFolder\n",
        "import torchvision.transforms as T\n",
        "from torchvision.transforms import ToTensor\n",
        "import matplotlib.pyplot as plt\n",
        "import time\n",
        "import os\n",
        "import copy\n",
        "\n",
        "plt.ion()   # interactive mode"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "AWLWc3of5KC2",
        "outputId": "4afb06fd-54a9-4c6e-9b94-78de9e487b94"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Mounted at /content/drive\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "1fm8SuALAlyr"
      },
      "source": [
        "Load Data\n",
        "---------\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "n0QngBN4Ev_a"
      },
      "outputs": [],
      "source": [
        "#Dataset from drive \n",
        "DATA_FILE = \"/content/drive/MyDrive/MLProjects/dataset/african_atires\"\n",
        "train_dataset = \"/content/drive/MyDrive/Datasets/dataset/african_atires/Train/\"\n",
        "val_dataset = \"/content/drive/MyDrive/Datasets/dataset/african_atires/Validation/\""
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Data Cleaning"
      ],
      "metadata": {
        "id": "2TYT9GhdgjY3"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from pathlib import Path\n",
        "import imghdr\n",
        "\n",
        "data_dir = DATA_FILE # \"/content/drive/MyDrive/MLProjects/dataset/african_atire/\"\n",
        "image_extensions = [\".png\", \".svg\", \".gif\", \".jpg\"]  # add there all your images file extensions\n",
        "\n",
        "img_type_accepted_by_tf = [\"bmp\", \"gif\", \"jpeg\", \"png\"]\n",
        "for filepath in Path(data_dir).rglob(\"*\"):\n",
        "    if filepath.suffix.lower() in image_extensions:\n",
        "        img_type = imghdr.what(filepath)\n",
        "        if img_type is None:\n",
        "            print(f\"{filepath} is not an image\")\n",
        "        elif img_type not in img_type_accepted_by_tf:\n",
        "            print(f\"{filepath} is a {img_type}, not accepted images\")"
      ],
      "metadata": {
        "id": "RlPbLMT9gg_V"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "#Data transformation"
      ],
      "metadata": {
        "id": "iUpCsyP0AAbu"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "U-tNYVO4Alys"
      },
      "outputs": [],
      "source": [
        "#changing the format and structure of the data.\n",
        "data_transforms = {\n",
        "    'Train': transforms.Compose([\n",
        "        transforms.Resize((224, 224)),\n",
        "        transforms.RandomHorizontalFlip(),\n",
        "        transforms.ToTensor(),\n",
        "        transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])\n",
        "    ]),\n",
        "    'Validation': transforms.Compose([\n",
        "        transforms.Resize((224, 224)),\n",
        "        transforms.ToTensor(),\n",
        "        transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])\n",
        "    ]),\n",
        "}\n",
        "\n",
        "data_dir = DATA_FILE\n",
        "data_types = ['Train', 'Validation']\n",
        "image_datasets = {x: datasets.ImageFolder(os.path.join(data_dir, x),\n",
        "                                          data_transforms[x])\n",
        "                  for x in data_types}\n",
        "dataloaders = {x: torch.utils.data.DataLoader(image_datasets[x], batch_size=4,\n",
        "                                             shuffle=True, num_workers=4)\n",
        "              for x in data_types}\n",
        "dataset_sizes = {x: len(image_datasets[x]) for x in data_types}\n",
        "class_names = image_datasets['Train'].classes\n",
        "\n",
        "#checking for available processor \n",
        "device = torch.device(\"cuda:0\" if torch.cuda.is_available() else \"cpu\")"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "g8OcbPA1Alyt"
      },
      "source": [
        "# Data Visualization\n",
        "\n",
        "With the help of data visualization, we can see how the data looks and what we are dealing with from the image. \n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "def imshow(inp, title=None):\n",
        "    \"\"\"Imshow for Tensor.\"\"\"\n",
        "    inp = inp.numpy().transpose((1, 2, 0))\n",
        "    mean = np.array([0.485, 0.456, 0.406])\n",
        "    std = np.array([0.229, 0.224, 0.225])\n",
        "    inp = std * inp + mean\n",
        "    inp = np.clip(inp, 0, 1)\n",
        "    plt.imshow(inp)\n",
        "    if title is not None:\n",
        "        plt.title(title)\n",
        "    plt.pause(0.001)  # pause a bit so that plots are updated\n",
        "\n",
        "\n",
        "# Get a batch of training data\n",
        "inputs, classes = next(iter(dataloaders['Train']))\n",
        "\n",
        "# Make a grid from batch\n",
        "out = torchvision.utils.make_grid(inputs)\n",
        "\n",
        "imshow(out, title=[class_names[x] for x in classes])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 148
        },
        "id": "IChNnMCCYROh",
        "outputId": "3432eff2-c0b3-4deb-a15b-d91d2626ead7"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAACDCAYAAAB2tFtFAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOy9eZBcx33n+fllvqOquqqrbxyNgzhIkCBAHLxvSSYpkjoojeTRZcv2rGN2Ytax3ljv7Ex4dme8sZ6N3Yh17HjWMevxxIw80lCyRUuURV0UKV4iKYIABZ4AiBsE0A303V33ey8z94/3qroAghQpkSYt9Tei0FX1svLly5fvl7/f9/fNhDjnWMISlrCEJfxyQb3XDVjCEpawhCW881gy7ktYwhKW8EuIJeO+hCUsYQm/hFgy7ktYwhKW8EuIJeO+hCUsYQm/hFgy7ktYwhKW8EuIJeP+9xQi4kSkJiL/5l0+zwdE5NS7eY6fB9n1b8ze/7mI/K/vdZveDt4P/Soivy0iT77HbbhERKoiYkTkd9/LtvyyYcm4//3GNufcvwQQkYtE5Hj7gIgcF5Hb/i4bkxmLv3yLZf9IRP7ovO/WiYgVkf/v7ZzXOfdPnHP/+9v5Tdc5z+m3n1H2AyLyWPb+DzOj1P2qZZPOF3+etryV876Fsp17kF2bExHvnWzPL4rue++cO+icKwI/fm9b9cuHJeO+hPcTvgjMAp8RkfCdqPDdMmzOuf/DOVfsfgH/D7AP+Ma7cc4lLOHtYMm4/wpARLSI/ImITInIMRH5vW6PTkR+R0T2i0hFRI6KyH/7JnX9CxE5kpXdJyKffIfaKKTG/X8BYuBj5x3/ZyIyLiJjIvKPzjv2lyLyx9n7D4jIKRH55yJyBviSiKiudk+LyNdFZOCdaHdXG+4G/nvg0865WvZdhzo6v50X+P1bLvsLtHFQRL4tIgsi8iyw4bzjN4jIbhGZz/7e0HXsMRH5YxF5OotSHsjquzerb7eIXNRV/k9F5GR27DkRufmdvJYlvAU455Zefw9fgAM2vsnx48Bt2ft/QupRrgL6gYez33vZ8Y+QPugC3ArUgZ3ZsQ8Ap7rq/XVgJalj8BmgBqx4B67nZqCVte//BR7oOnYncBbYAvQAX+2+fuAvgT/uam8C/F9ACOSB3weeya4/BP4D8LV38F5cBEwDn3mze3SBdp56K2V/wXZ13+e/Ar6e9eEW4DTwZHZsgDRq+k3AAz6XfR7Mjj8GHM7GSTkbTweB27LyXwa+1HXu3wAGs2N/AJwBcm/S1seA332vn6tfpteS5/6rgX8I/Klz7pRzbhb4P7sPOue+65w74lI8DvyQ1Ni+Ds65+5xzY84565z7a+AQcM070MbfAr6fte+rwJ0iMtLV/i855152qVf8Rz+jLgv8a+dcyznXIJ3c/mV2/a3s959+JyibjD76G+DerD/elxARDXwK+FfOuZpz7mXgv3QV+QhwyDn3Fedc4pz7GnCAcyOoL2XjZB74PnDEOfewcy4B7gN2tAs65/6rc246q+tPSCfVTe/uVS6hG0vG/VcDK4GTXZ+73yMid4nIMyIyIyJzwN3A0IUqEpEvisjzIjKXld3yRmXfKkQkTxoR3AvgnPsJ8Brw+Tdo/4mfUeWkc67Z9XktcH9Xm/cDBlj2i7Q7w5+S0kh/8A7U9W5imNSLfqN+XMnr+/UEMNr1+WzX+8YFPhfbH0Tkf8qovvmsz8v8guNkCW8PS8b9VwPjpJREG6vbbzLP8xvA/w0sc871Ad8jpWjOgYisBf4j8Huk4Xof8PKFyr5NfBLoBf69iJzJuPJRUm++3f7VXeXX/Iz6zt/q9CRwl3Our+uVc86d/kUaLSK/SeoN/0PnXHyBInWg0PV5+ZtU93bK/jyYJKWr3qgfx0gnQc47/rb7KOPX/2fSiKs/Gyfz/OLjZAlvA0vG/VcDXwd+X0RGRaQP+OddxwLSkHkSSETkLuCON6inh9RwTkKaiCX13C8ISeWYv/0W2vdbwH8GtgLbs9eNwDYR2Zq1/7dFZLOIFIB//Rbq7MafA/8mm5wQkWERuecN2vyX8hbknCKyBfj3wBeccyffoNjzwOezhPadpPmMN8JbLpslN//oZ7WxG845A3wT+CMRKYjIZhYnT0gn9EtE5PMi4onIZ4DNwHfeznkylEgnkknAE5F/RTp5L+HvEEvG/VcD/5GUR38R2Ev6ICeAcc5VSFUeXydNoH0e+PaFKnHO7QP+BPgJaUi+FXjqQmVFJCBNqD3zZg0TkVHg14B/65w70/V6DvgB8FvOue8D/xZ4hDSp98hbv3QgpU6+DfxQRCpZm659g7Kr3+iazsP/SDrZfVNer3f/w6zM75Ny1nPAF4BvvUl9b6fsW23j+fg9UurkDGnC9kvtA865aeCjpPTSNKnn/VHn3NTPcZ4HSe/dQVJqp8l5VOAS3n2Ic0v/WcffR4hIk1Rd8u+cc29rdWbmnf+5c+78MPwdg4jcBPx3zrnPvVvneKeRTUgvAFe8Ac3ynkNEVgFfd87d8DML/z2AiFwM7CaNIP+pc+4v39sW/fJgybj/CiBLWH6Q1HtfRsqxP+Oc+x/e04YtYQlLeNfwrtAyInKniLwqIodF5F+8G+dYwtuCAP8bKe2yl1Qt8q/e0xYtYQlLeFfxjnvumZ72IHA7cIo05PpcxtcuYQlLWMIS/g7wbnju1wCHnXNHnXMR6aq4CyoTlrCEJSxhCe8O3g3jPsq5mfFTnLsQYglLWMISlvAu4z3bClRE/jHwjwGCwLtyw4bB7qPp/gg40v2kzoNLy/ycZ359VRes//yjr/+dZMcdkr4c1CtVfNFgLdZq6s0Y34dSoYSnFYmtkuvxiJ1gEx9xFisWi+rU130S1zlb19fu3ELdn5MYqtWIxJis1ZI1/dz2t78SFt90v1/8xXn9/IbHst85Bxe4Z+l9dN2d9/pKs3LOubTdXdfYVSA7zgXPs1jXW6Ub0/JJkuB5XvsUHWgl+IVy9qV0jrUvobMrDO6CZ+zup+56BcGd09Q3/nXnx+5C4zH9zlpDkiTgHEpplNave3a62/pGT8/5rUiv8ZyrvXDBN2l6GzaJYWEGiZJzryDbC0Xa99W59HkSR2wtiQgxjtg5LOl90FoT5kJ6enrI5XJorTvNSpKEWq1Go14njlPRUxD4FPJ5Cj09eJ6HiAIccZxQr9epVmskSVo2FwT05EJCZ3HNFnGSULOOunU4Sceo1ppisUipVMLzPJxzGGOoVCpUq1WMMXieR29viWKpiNIaax2Nep3KQpWo2UTEUcgFFMIcJk6oNJs0E4OIEAQB/f39hGGIiBDHMTMzM7RaLZRSlMtllDhOj52Zcs4NX+gWvBvG/TTnroJbxQVWuTnn/gL4C4A1q0fcV776m9lgVFgLOKHVahKEfvdvsk2GFIJ+/Zmd4Nx5o0qyCSKbEFxXGZcNlHbeoW2XFj9n5xPJDK8FpxDrgfUQ5+EQEmniVAtMyJMPv8D82Vl8a6lOe6zafBE5H4rxMP1DQn7gZbZcGbDglzh9fAjPBjS8JsaGeGbxiWmf2zqLXdxcqXPMmNdtvIRzjqkzlh8/eZLZ+SoigoiglEZEdT4D2feq82of6361y3Xekx1T7TKKtoVqG2VrLUprlM6CQknNtIigZLGdi4ZHQNI2dB8TR+e77usmO3f7mttjJn1YF+8dnfuWGiXbVb772tpl5+bnKZWKeJ6HUqpTT7k3ZPn2u9E6wGmHUoLWHoLgex4ikCQWayyJMRhrMiueTVAqHa3Kpn2hslMqpbHZ5ausybbLjIpLu9aKI01jAdZ1xkLb+XHGUpueRJkafaUcuVzIXKXK2MQ8paFRCj0D4AxW0vI4i7IOyAZ7Z7KWzoQptG9r+pQZLA6FdPctDks2yUnaenGK7JIRlV1A1+PYnD3N9Ff/DDlwOLvORaciloCaLhFhmEuaNAsecyphNo6oW6EhDh3kWDGwnKs3b+emW25icOUyCoUCURQxMzPD888/z09/+lNeO/kaYS5k6NJNbNu2jSuvuoq1a9YS+AHOWuZm5njl5Zd55cWXOPriq9RtSGFohO0b1nHtmnUwPoEcf42xqMYrzQWOtZo0lYfTIbq/zI6dO7npppsYGRmhWCwyPz/Pgw8+yFNPPYVSipUrV3LNNddw9913sXzZMJ4TXn31Ve5/8Hv8ZP+LlIs+V46s5PZt2+mdrHLs8FGemT/DwbiJEZ+dO3fyxS9+kVKpRKPRYP/+/dx3330EQcCyZcv43Oc+x8UXX8zeZx/jn/3hH7/hVhzvhnHfDVwsIutIjfpnWdwj5GciffhSA+6QdGB3DFs2KDOj/Poft53H7gd/cdC2q2k7mN1VuLb35bq8R2lPAumjagnRziDECBGiPWIHiCVIQrRxrFlR5L88/DQeZVYPjXLo+AFWDKxhYn6cQq3OzRuHUckM/XnLrFY0bR7PquzhNxe4pPN82OwilJJ0Ejzv+3P78Y0NtVIaUTo1oF3ft/vWZVGDtCe8jrfnECuIUuAMzoGxFpMkGGOx1hKEAXGSYIxBKUWSJPi+h1YO3/PwfR9jTOf+GOtQSmGtJUkStNaYxNDbWyKKYqIo6nhCyvNxzlKt1ggCH98PaDabCJmnKmCMoVQqAWBNVyd1DKsi8AO0p7HWEsdx5lnZxX7oTOhgsQgWbNbnYtDSLq+6XOIsinI2G6UW5wTb8ZYF3fHibcdwp83OooLz7pXDYq3BZZfhcDhrcVYQZxguh+xct4Ha3CxTUzOsHB3m8v71tCLh+f1HaSYRNhvPTsBZhzUOxKFUdp3pnNl5ELpHknUOJ+kOg9JVj7jUgCsRlEsnWEThlKCsQ5RD0mmt4xAYz+fVJOC1qAetMk/bWZTSaCVoEqyvmBdFPUqIPQGvSF95kK2XbGTHlTvYsmkTfbk8hXyB+UqFvXv3smvXLvbv30+lUmGgVOaqrdvZcdWVXLZ1Cz3lXqw4puYqHN63nxf3PMeRfQeYnThLzvNYvryPD67dzva+QZIjx6g89AOk1WRGOw6h2WdgrqePnpzH9Tu28YHb7mT58hX4vk+lUuHxxx/n0Ucf5fjx44RhyDXXXMOHPvQh1q9fTy6XY2Jqigd/9EN+/OPHMAsLbC0PcPOOy1kvHjNPv8yBZoNna1OcbNbIlcpcd+utfPrTn6ZQKDA+Ps7u3bt59NFHqVarrFmzhs9+9rOsW7cOYwxRK3qdvejGO27cnXOJiPwe6So1Dfxn59wrP/OHHScqfQjSQZzN/ueEtW8WDy6W7Rj4zNPDdRnAToQpXUbxXM+4+2/6QUicjxIBF2N0QiyWWBTaCkU84tlxpo4+z7IBj73Pj1EKQy5atZbJMxXGTxzhpruvoKWGmVyYYVnR4Oeh4hI8qaDRJC48j3LJHvKudnWH2u0o45zwO/O4Xm/cBWMMrVYL3/cRUVgUWussnIdqrYrWGmctWmtarYhGs0GSJBQKBZrNJp72SJIYpRRxbDJv3WFMQqPRpFgssm7DehYqFQ4fPkS53IcxhjVrVrN+zSjO2cVzArl8njhOJ4F6vY5zDq01cRLTiqLOLRWt0L5HlFgajQatOKZSr7Ni+QrqzQhByOXzNBtNGvU6rSgmiuN0Esk6S7WjFa0xxoCDnmIP42NjXLF1yzl9ufivw+IQcegswgGwzmJM6ohYY7HGYEyCJZ3gQM4zlOls6aE79Exq3G1q3LO2iQi6PeECmtT4JmKyCXfRcx8qh1y9dRUzZyY48tppCsUCLjFo6zh1+DDDpRy6PMCJU7M4C6IUohXorF3issjhdV5Ol1cvaOeQzAtfpNccokCh0KJQWlDaQ2mNdg6HRZx0IkalFIQ5YiVUPQ+tNdZZRDQmScDTxJ5DlEXlQ0aHR7h84ya2XHo5F63bQFgqUBjspZY0mJiZ5LkfPcSzTz3DxMQEzjnWrFnDRz7yEa698iqWL1uGHwTUW00OHTrEM7t28dzze5mdnCSIE5YXi9x05XYuWr4cf+oU8eEjzEzvJXYRk3nDEa/CwchSS3JIMMCmdZu465472LptEzny1Ko19u7dyw9+8AOOHDmCiLBjxw7uuOMONm3ahO/7zM/P88jDP+KRRx/h1PQYK8pFPnbd9WxOAqp7DlJNEo5IzHO1OcZsQm54mI/f8wlu/7XbcM4xPT3NN7/5TV555RWMMWzfvp1PfOITDA0N0Wq1+Na3voXEC7wZ3q3/peZ7pEvc3zKkE0JLOrCsYIzFGEsQBNkDA2/VuL/Fdp5Da1yYw21X7RCJsc6Cy9FKChw4WWXPiwfZtrHAx64pM3fqOHpaM9o7xPjyKQ4e3Me+V0/So4vc9ZGL2HnNxfzZl55h3SW93HrjIPPRcupSpC/wKUsd6coxiEjnmjs+tXRPRoJSr5+EUmN+rueuMg99fHycAwcOsHbtWuLE4CT1oq21aKVxuI73GgYBcZygPQ/lwBiH5wUUegrMz8+j/QDtC5MTUwRBQKlUJjGO4WXLCMIcJaUp9w8Q+AHNVhOlPUQJcSvBWtvht5M4IY4TfN/v8Ja5XI4oifHCNIx2IrSaTYJciIsdrTgmrz2srTE9O4sfhDQaTWjFGAfF3jIAHorYNFBtL1gEi+CsI2l79KKpN1ud6OX1OZ6MNhJNanMzg5XdF2stxpqU8sgmW4fgrM0cCsviHYMoy9C03V8RmxpdJShn0aJT+gVzTluMSzrRlLM2nYgsVGbnqVXrlPpKNJstxk+d5czJkxx6fi8LjSbLt19FYWQ9YjRexsOLcxibkkCLdt1hHYh1KOswzpFg20M/zQiJ6zhFSgTlaTytCLTC8zRBGOCHHmLb9Su0Uikd54SkFpIv+kTFJq1WCwDtaQp9PZR7S1x26RYuvuxShpctY3R4Gb1+Lm2a7zExOcmex3az+7nn2HfwVRr1On29Za6++mpuvvlmNm7cSLlcJopajE2c5YW9z7Nr1y6OHj6MNZbVhRzXXbSezatWkW9FLJw8RW3fq4TNGg1nOauEoyZmouUxI71EuZAVAyPcc8fdXLVzJ72DfTSTiD3P7eHhhx/m0KFDKKXYsGEDd9xxB9u3b8fzPCqVCvv27ePRRx/l4KsH6FHCr19+JTuHhqkc2M/piTOIVbwmPj+pzTKpPQYGVvKZ3/wc26/cQRAEnDp1invvvZcDBw6Qy+W4++67ue666+jt7eXs2bPcf//9HDx4kGt3XPrG9or3MKF6PtqcafqESMdzwVi0qNQTEMG+SR10AumOrwuQJWHSz4JNuUzVHoBkD6vKfu9QqoFLymlTqIFVKQWjfGZnPX66+wwjw5rvP7qfZ/bVadYiPnZNi8FCRDnoo+612LF1gNpoQFjYxEDPMIPhQQZ6FAts5sk9PaxYM8rfPj1LA7h41OPjN6xiwJ/A0kKpGGwuC8sTcHLevJN6T91e/ut48W5jlRWs1xskicHYlB/O9RRx1iFaESUJWmkarQilFFE7sSOCH+Y63nujFSHKwwtCrEkNW5wkzMzOYqyh0NODKEWhUKC/r5/ZuTmssfh+0LnHJkv22g73kxqHMAyz/k+9a601TqfeZhzHNFstrPOI44Senh4azWaaCIsXCMMcURyjtSafz1Or1fB8vzOJuDY5LIJS0sWrtxOSNqNI5Jw8iyhFPsijxUMwCAqbEQ4igtKklI6zOKPRzmAShxED1uFE0lRNmzNse71WpSNTUjqjw8g41/HQ23DOEbeNe1fydc+u55g4EDA4sozJiXGMSSeH00f2o6ImgvCThx/lw5/dQLHQg+97III1Jut2hfY0SglRFBEnCdZYXGJIjMHDorRCSWqkne44+wR+QC4f4mudRjRKoZS0ZzyMs1l+y3UuW3saaQnLCisYWjucRnkbN3DJpk309/XSP1Ai8QSVC5hfWOCVU0fZf/ggL+5/hbMnx2jNVwm1z+iyFdx417XsuGong6tXosKA2lyF53ftZs8zP+WnLzxPrVkh5wnXXbyRzaOrWN1sUjs9xuRTP6HiHMpYXBSzXzuOk3AaqOmQXNLD8nw/V91yHb/24Q8wNFJGK+HI4SM8+OAP2fP8XlqtFhs3buCO229n8+WbKfYUsK2Il17cx0MPP8Yr+14hiRpsWj/KbZdtpP/YDI2Hn8baFqdUlcO+40jNYV3AyIqVfOq3vsB1l2/F2oSXXn6Zv/7q1zh96hQDAwN86tc/zbpLNxFqn2P7D/FX9/8NJ8ZOsmpkhOGhbhHK6/H+Me6qTafQ8VB830Nb0NalSdAsVLRdaorzAt+0LlmkQS10lCg6+xQ7IUJhbQBWYxIwCfh+HmstfiFN2HrOA5qIckzPwoPPn+WZn8xx4nQvn71rnks3jvD4vhZPvzzN2Ml+Lss3WD56lvp0QGgvwSuG5AZX44ewSS8nnpglzg9xZt80xYKjqkucneljJq6BX+HTV5cIwxqiIsTkWcwxqE4uoE1NWWsuyLOnCa2U200NUJpsVOLRbMY4PIzz0X6Isyn3HscxYZDDOUs+VyAIAhyORr2B5wdpcrsZYY3DJEmmxBEK+Tw9hQLFYjHjby25UAg9jecFDPUPMDwwiDGG0PeJ44goUy/YjALxAw9rE0RZEtPE2gilA8TFKDE46/A9hfEVvu/RihUKodVoYhODFsXA8ACIMDU5Rb43hzOWfJgjcaC8oMODG2PwRNAoHApjEpQDTxSe9jq0TMcgAUqEYiEkl+vBU+kk0IwinMuSo5LmH6xLk5vOGmyWh5DUlcDa1NCLSsewtRaXUTkppWXSCcC6lKrJJg5rLW2/vz2e01usQAmNRp2TZ48wMz/L5OQsxVKJ2fETjJ08gRcUWDbch1YhoR8Q5kPC0Mf3vTR3oRVYUDpNEGutSBJDs9WkUW9QrzcBSz6fw/O8NKLIEspp/iTjzK0BJxjAWpdOHJYsF5NgbfpAC2n+YyFu8cV/9N+wfuMGEmfxAp9KrUrTGg6dPMWpU6c4cOAAR48eZWJigjgxeEGO5UNDXHHrdezcuZ1169dRypyIsYmz7HpuD7t3PcvY8dcIWsLakWVcs+1yNhRC/NOniXbtolKvUbMJdWuZ0R5nooh55zjt5Yi9AnEc099f5oYbbuDmm29m/br1WGeZnJriRz/6EY888ghxHLFy+Qgf+uAHuebqqymVe2lGES/ve5kfPfgQr7x6lHqlzsbBAe7ceT3DjXkajz9OZSGiRshRLPuscLphcUGeLVu28IUvfIH169dTrVZ55qe7ue/r99Gaq7Bh1Vo++/nPsXrjOhaM4dknd/Hod77HRGWSS1av4JNXXsd49B7QMj8PzlVPpEGv53koK8TGIEo6pm7Rer8erksB4TK+UjuDuATnEpqxwqhBZisBh/dNc+LYSaanZ/C81MtrtiJWrt/KVTsTVg3UEeexkAR888HneeCRszhvNS00p8Yr3LhtNXl1lpnJXu57rMkf3FNmpG8KX63mTFyDpITxoThUYd2mFTz36ixmIaDuG6KoxnDoc3Q+otkssecl4ZatedYUJwia/eASWqqFc/4F2aI2j9lWqHR/L938slKQeaRxFOOAMMzRbDbPoaTiOF6kSpKEIAhQSpHP5anX6/hdXrCxKVWmtMb3fQqFArlcgHUm9dBVyu8nScLExATFYpFCYTg1JF5qtEQ5At/HD3zi2OF5Gq1TA25Mgu8HeNqn3qwjQYCQ0nRxnMrN+vv76enp6UjgolZEPp/H0x5hGDI5OUmQL6SG1LrUA1VpjgFrU2/T81CicNZ2EvEdfjgbi/XaAg9/7z7CXIEwl8f3BOcMPYUSWim056eRjO+BdQS+h0WyqMHvDNN0ImjLE9Mx7uk8onTK1WceTdtZSaO2NIhVSuG3SUstaOWhFAwNDTM/eYwj+/cxsGwVWgxz42Mp5WSgXq1THhoklwvwfI3ne+TzOUI/JFXrZlGEpJO873vkcz7lUg/1eoR1llzOT/vDpFSNcQZhUbVjsgnf2KTjqTtrU5WXtdlElhr4JImhEDBWmcabyjM+Nk69WuPw4UOMj40xfmYsdQTCkCAI2Lx5M5s3XcqWzZezbOUoyg9Qnkcrjnnm0HF27d7D8eefpzJ2ksG+ItdvvYSt61bQOz2P3ncAzi7QtIYFz3DAs5yxhqlWk6oTklyeSBQRkA9DbrzxRu6++25Wr16NiFCtVdm9ezcPPfQQY2NjLF++nNtvu42rt26jp1QiMoYjR07x0COPsGvPXiotw9rBHj6+YxOXWIPb+xzJ6TkwPicC2CsNTmBoKZ98oY8P3HIrn/rUpyiVSlSrVR555BG++/3vUa/VuWLrFj71qU8xsnIF9STiqe8+wvcf/iEtqfHBzRdz9+g6xh96DHvFuje1qe8j486id6oE5xQ6S975mcviujzyN6ynu8Is+aSNQbBEziO2RXbvHuOHD+3j+KE5rBGCwCOXD1m7ZhWl8hCP/XicQ/sd//R3V9JTavDcngWefHaM6mwvulQiyfmMjdfouTYkcI6FeIC/+vFx1o8u54ODq9h4kbC61MPMmTqDq4YpLO9F+/0kxz0GAMkN02x6lKnhJKDVSL34A8cV/YUSeYkRWwAd4yRGuQB4vSJm8VLPtf4C58kcFRZITEpbOGfRmaFuJ/F8P5Wcep7XUZAkmeIlDMOO3Kyt6/V9nziKcFY4c+YsWisK+RwD/QOdu7CwsECz2SSKIpYvX4ZS4PsKYyAIUgMFDq08tPaRjNdWyiOOGgga3w8AodGIKPf1UW82yeVynYSs7/soJeTzeXp6eqhWq9TrdYIgQEtKBTjnCIJUVWNNkk74Gf+sM/1xN9qTpgBxHHH69CFsxpErQGkYHhhmdm62ExmJCCZJyOVC4jhZlJ9m84RCiBOTTpK+j9aatRs2s2bDFZ3zdWSgHVWPgFKdnIFWOjPEaaXlch/hYD9OHKePHeKi5YNcuv4ijp6ZxFlhdNkgutxP4AeQ5VPSHIFDZycxJskkt3HmEKQjKPB9wKGETj4hNeIW52zqpTtHklhMYrJ6JFUWOQfYTD1lUuMOJMZQHOzjoScfobqwwMzZKQJROGMJfJ+NGzdy8cUXs379elavXs3AwADgSExChOXk+Alefnkf+17ax+mjx1Bxi5FygX9w01VsKRRxx09R/+5TqNjSiizjCPtdi+M2YepxcyEAACAASURBVMHoNGL3c1jlY5WP8nwuvfgiPvHxj3HppZeSy+Wo1Wq89NJL/OhHP+LQoUOUSiU+/OEPc9ddd1EqFomjhLHZGR565BH2PLub6bMTrOwb5JPbL2eTD82XX2RuaopIAs4GIQed4USrwZyCwC+yrjDIRz/xca6+/SY8z2N8fHxRRimKD9x6a3qucplKvcYD3/0OP338WcS0uPvKK7jZBTQefQovaSDyZvnH95FxP8duZclV7Wnq1RpBPuxogkUkk4W12fUuzXOHc180eC4baJHRNJISDz70Cl//qydo1vOAlypIIo9602Ns/BTlcj+FwS2MjcHuPSGXbfd4es8PmD/rsAsG/GWYsMTkpEfgW/K6iZYBFmq9/Nl/WmB8Z8Jvft5j1YoA5TfpG7HU9QDf/nGRx+4/wpahq3nmVIOFBUM572jZefRCjbm5hGee7qHPC+jdqvGiACFEUe3qlvP13ef3ocs4ZXXOy7nUG2+1otQYisI50zEScRyTy+UwxmCMIZ/Pd4xyW87oex4DAwOUe3tZqFRIkgQRTbOZ1lks9jAzM0Wj0aI/V2BuboHp6Wn6+/tptVrpopMwlyVwDcakEjhP+4iX/k1nb0Ernziu0mg0KZfLHDlyhDAMsbZNC6RRRK1WY2hoiMnJKfL5AiLSWYyUz+dJkjh1DnyPVqOOszb13lUqYzTGdCIfay1aL6ph2n0jGe2iM526OMDYNApKHE6StluOSQytCExsEAXiFEplMlJRKa8dpxSIszA9PcGaSxzoVDMP4GV5n5TxWTTm2lN4np+KyLN8bLhsOWNHNBcN9zFSzDE40J/WjaWvWKTZiuhbuw6UwtqEOIaGpLp8rRXO2dRYW0hskk6mOpM0GgsKtJasXw0mbmEtJNZgrMPYLGrMvrNZDktbmyZfAUgnExwkscGZhLhRZ7ivj6u3baMQ5BhduZLRlaOsWDlKEAQkSTrh1Go1jh0/xvMvvcCBI4eYmpzA1hr0ezl2rhrhyvUXsSZStA4coTY1g01azIvHKWs4bFuc0kIzzBObHEncwjqH54U4J6xctoKPfuzjXH/9TvK5EGMMe/bs4YEHHmBsbAznHDt37uQTn/gEGzduBGD/gQM88dRT7HnuWRamJxkKAj555Ra2Di/Df2Uf8ckzuEqL2SDH877HcV+YaVTRTlNQIf0Dw3z885/hqmuvQYvj6NGjfPOb32T//v3kcjnu+fjH+eAHPogOfcbHx7nva3/NwX37KfsBH73hJlbNzTO/6zlEK/ZJlHLJb4L3jXHvhKtIZ/AiglPpKytE+ox1ce7n8O9ckIs3yhElIc/tneC+v/oJLiowOlxmtjKL9nwGB4cQPKamZzFxxMKZV2kmHl/7Rotrz46ycd0WahMvc6B2lqpZRmIGiJMScVyhWJpi1J9h6yW9tGYjHtn7Kjddv40Nq8Av5DDa8uohx7/78hS5Bcst647TX9BMz2tKpWWgQWZmELOArBrm1aMDrFrfz1Bpmnxk8VwRq9KH6IIce4bzFTOLC5Myz90lRFFETzHVf4dh2ClfLBY7hj2VQLYy4y3Mzc1hjGFwcJBarYbSmkajQbVSIQzzhGGOVitNbIIwNTWNiGZubo7e3t7OuaqVKhATBD4u45nb1E0Y5giCgDiOieM4C+/h6NFjlMtlZmfnGBoa4uzZCbQXksultFIQBLRaLYIgwBhDHMcEQdAeKtSqVYaHBzNNe8qpKxHiyDA9PU1vby+tVosoiqjX65TLxXPUSmnsJxQDHyUaK3RURSZJCMMAYw0qmxQSFJ5SeIEm9Xp15pA4lPIWk7vWpTRGFDHcl6dYLuGJhyiNWEO92cK6NILpJCVVmnuxOEzi0Aqcr2kY4eSJI9SaDWbnBjCJZXJ2HmsSSr19XPShleQLhdSQmzT5HUVto5AmfI0RrDOpHDeLkLGpDt/ZVPZpjSEyMYmxGfWSLbqyFnFg3OKiO9VWBEGagU2zCCTGcNXWbVy77RpGV42yfMVKKrUqYS4kSQwLCwucPHmSQ4cOcfjwYcbGxmjMV1FJQrGYZ8eKEbZvHmVDT4nq1Fnmf7qf+QWDTjQnE8VRpTiW1JnVUC94GKXxE0s+gpYSnKcYGOzj+utv4PY7PszIyAjGxBw7dowf/OAH7NmzB+cc27Zt4/bbb2fLli34vs/s7CxPPfUU33ngO8yfnaDgCx+8ZB3XrxpFHX8Nc2A/9UbCSeM4GGpOuoR5a7CxUCLA+o5lq5Zxz2c/xdbrdpCYmP0vHuArX/kKc3NzjI6O8slPfpKrrrwKBI4eO8pXv3Ivrx04xKr+YT5347V4hw7RPHyUek7xYtTkcASX/gzz/T4z7t0evANxaTZfa0yXFPKcZT2vU5FI5+eLdWsW5hwPfv+nNJua/nIRi2VwxSjFUj9aQnyviMqNcuzYa1Rmx8iXRhibPMUzTwvlQoM1o71suPNi/vbpFi3boJUUMWiu3xnyDz62kTUDDrGGHz1d4Oj4OMYM4OfA+Q38YIRyr+I3P385W5v7+d11W4ibk0zWDBqfwOUYLJzmmisv5/kTsOvVHB+6RlFULVRSwqpWZ5GPy7TTbU62zRV3eC14neeOaGzDZjLDfLpQyE+Ta7lcniiKyOXSxNn8/Dz5fMqzW2up1Wr09vbSVy6nnLbn0dfXB0C5t49GrUVPTw/9/WUq1XmKxfR/4hMRent7Wb16FZVKhUqlChUIA0erFaWyS2NITExfv0ej0choEo9qtYZWXurpWWHV6BoazYymCYJzopdms5mqRHRqPI0xNJtNlCqQz+eoVSuMrFuf0S+p524NHU19ewJqQ0mXUivr23ozwjqwLkEh5EKfViPBZas+lc6kjCiiekpPaE+jMCgPFJrEtdLEokpVMkopxk6f4Hvf+msGhwbIhyGe5xM1G1hJ9eP5fIFcTwHPC7A21ZHjpXRPT6EHBehSkelqlShxtCbniZoRxglR0sIvQ3VuBpU/Dk5YmJlCa03UamCThCDfy8DKVSidz3T6EZFL1WkpjWcyaiUBC3HHC3eZAig15i77LlW3W7AGRZtOAkTwSZCkydzCNJ7W1I5VmKvMcnDfq9SbTSYmZzh1ZIyZuWmci+nJB5QLebZeso5Llo8y7IfkpmZQR08xPzdNq2lICDgsjmNRizFrqAU+Jt+D8yC0liiKcNaShD4rVqzg5ltu4cqrrmLl6ChxknBq/CRPPvkUTz7xBAvzc6waHeXOu+/myquvpre3TK1WZ9ezu3ngW3/LiaPHKfoBd162jYvXDqHPvkbzqafx52s0xWOfDnk+bjBeyGM8D79aI5c4XKHAtmuu5COf+CirV6zAb8S8+MIr/Kd7v8bk7CRXXrWdT3zsHtatuohqs8LhAwf5xle+wcTMNJvWruFjF1+K3fMcMj1HYhXPmSr7iRCj0G9A07bxPjHuwuIeZi4L6VID0eYn25xke6m26yrffhKly7i79pI+FFjNvhdPMH22SqU6S701TeAXGXKWvJ9j5YqVQA/V+RmEAokNaUWTlHoHqUzNIf39PPLEQfLuBBV7CTJkSHTM1ILl5u3DXLZsjlJhklwi3HLjCE8/brBuEI8ZjNMM9Cp+bWeOj9wew3OadRsdJ0738rdPNZBWi6Bo+NBNgxw9/BIvnFrLqfoAG5YPsWz5BI5Gpo5Y1FZLl/59McGcGXgyaafoDs3R3oJBUPiBTxAq8j0Bg+VB/MAniqLUq3TQqC3gKejrLbJq5XJazSa5fI4gCHHOkSQJ5VIPA329DA0NMnb6JCtWriCfz3HkcI1iMeDI0eMsW76SkcEygwNFwrymZVoEXg6HkHSSwILoVIOfGEgSQxhqisUCh48cZ/WaixARas0GzVaLkZERpqcmu7TmgjOLk32z2SQMww6lUioViaMWnqfwfMEaAIunFfm8BgTfTykLa5KUX3aptqrdl45slaYj06+7jG/OaBdZNHROHLFJF+t7DjQOSVJ+2ziXefnpGlWNkNiYqbMTTE1MIM7R01OkWa/RjFqITVd4+qFGedlCHwSyhKofBCSxxbMx2s8R2piS5FCFInP1WVQY0IgSHv3u/fT291CtVSnnfVatGiW2Mc1WxJkzFUZGV+H8gGo9SiNjgVKxmCZ6rQVJk9CCwgsDRDTOuDTa8kO09jDWoj2NSWKm584QRxFrVl9MqxmxMDfJYCFk2LNElUlePPoqhyvjNBMhqELr5BSeAJ5ioFBm58Y1XDYywgrfx6tUsfOzNF58jng+Jow0dXEc91uccQlzUY1JZamGPi7II85LZZ3ZwrswDBlZPsItt9zCzmuuZnjZCEopJs5Osnv3bp544gnOnDlDTz7PXbfdxu133MHgsmU044QXfvIsDz3+CC/tfwUdJ2xbvYrrL91MaWqMytNP0VuvE1vhFQIOGsdJiWn6Gs8a4oUGCQpd7uODH/417rzrLorFIrVajWdffJFv/M23mJma4+prd/Jbv/MFent6WZhe4JmfPsMDD3wbM2/Zecml3L3xIuSZpzALFc6g2NNscDAAG0fszOcoX2AHlm68T4w7tOmY88Tb5+q123Cv22LrTRFZxdjkBImNyIchlfkzrF7bw+98ZDPrVqxgsHcFPYURphdaPL3nBb76gzFOTczQIJW4KS9GeyGzExE2jJGmxXrC3HxMsydiYTYkX9CYIMHPl3hl7wnOjq5h8LIEk9QJcnUmq7PMmhWMDA8ips7g8jy9+ToBObbvGCKf1/z4O3torljJ3GTM87un2XZXDl/Vzu2l9oKcc5MU0PHjyaSPXfvFKOlIJ33PR+lU8RGGqWHXqi3rgzDwsSYhDPL0lorEudSoe76XcaEWpYS+vjJaK3p7e8iFPsbEOGeZmpxgYHCQKGrhnKFeW6DRbNGKWmDThTQ4i3MGz/NJkmxhGAqRVAPfbDVJbMLp8TGUCOW+Pmbm5hgeGaankKfZbKXyzTCXqjCs7Wyo1FbEiAhRK8rmvcxZUCmX7awjilokSdqGwPeoVisMD/VjjUG1NxDLkqq+TjeacpLKG7US/KA9FjMqQmsUikK+sBhBiUoNeuasaJXSE04cvla4WIFNtwHIFYr8zm//BgvzCzz59C4OHzmMacUkcZwuIsu8fWfBGUMS1bIVvMLaoRWYuSlalQatZoRXUHjFYRpegLiEuFmjEAr9g0XCvEdttkq9HrNQqTH1yssUSmUqzTrWpdFUPshl2w6kl6g9RRyZVLlDSk3FcZzKbj1NEARpcs86ojhVLU2Pv4ZCYZImZ7SQ932UiegrDTHo59F4FPwmW1ddQYkET0cUogXUdA394j6S+QYiGmuaGGlxOtBMeR5TkeKsUSTi8ANHVMhhgpAosZhWCx07eks9XHzJJdx4441cumkTfX39WIHT42M8++yzPPvss5w+fTrdLuCKK/jwnXexdv0GdBhy8tQZvvWt+3l5727ipMnqoX5u37SDFbMNaj95irloknzL51gkPCcxpy0EOsQPFVESs9Coo5RPqdTLp3/917n6hmvJ5XLMz8/z3e9+l8cffxws3HTNVXzxNz7PYO8AJ8bO8O0fPMzuJ35MRJObb9nJhyjCg0+QuIjD1vFcNMehMCLv4Kqwhx1+nqlzaYvX4f1j3F177xfVRbtIR+oGb5REvEBV5xk+8XxWrxul0HMKZQyegZuvuYSrL+2nLwRfFvAV9A2VWPahrVgV8xf33sdCZYZ8SeOsRjufOPEQV4F6i2YuYGJGUetXPPBEwuXbL8ZXimcPt3j2cMy+4xPcsilHgEZJxGtnYr7ziObDa9bDxGGOLQhDxT5yuRnWrN3Iow88Qr3hoaI8BR1xxdYRxK9gjY/DpCF9Fx8M7f5Kl8YvLm6Rjlqm27gncZqk8v3MU/d1Z20BLHq97TL1ep0kSThx4gS5XI5CoUCcJCRJQqVSoVgsMjDQT6ORJdkSSxQlhGGe8uAI01NTmZJFE8VxKm/LwnQl7XA+wdOqsyq2p9CD5ysmp8bp7+/HOcjnC4yOjpILA6qVCoVcukVDe/WyEk3ga+Ik6cg5lVIYY7AYatUFRkaG8Px8OjZwRFGEUpLKNHtKiCgajcaiSkYWnQfnHHEcIUqRy4UkcUxkXObVQqA9xKUyzSAMKfTksdYRNZu0Wk2iqNVRznha4ymXRi7OZbLMNPK86KK1bNt+Gb7WbN2+hef3vshjjzzFmYmT2T4ubcWMQjnBGsFTIfmeIh/8yMd55sGHqOlJTLOBzpfYeuPtHB87yku7n2JksAfPJUwjKPEwiaXejJicmwWlcMoDT+FlEUbo+7SSGENK0fiiMNgskWrxQw+yDe1sbPC0Io4ilChwjqgR0WxOoUXQWrEQp3r33rzH1KnTjL98HIljNvp5Rlevpzl2HJ8W+cSStCyR8pn2Fad1zBlpMuUlTOo6WnuUTUiZEvT0saBatJKYuN5Eoblo5ShXX30123fuYPny5Z3n5bVTJ3nhpZd49LFHOXPmDGEYsnHjRu666042X3YJXqGHmUbE0w8/zmPf/z7V6UkGlpX4yLYbGJ2qwXOvoqZmKCYR44FmV9RgP5ZWEDCiNCtLPRyL6zSaDZTnMTQ8whc+8zm2b9tOvrfAqVOnuP/++9m7dy9BEPCB2z7E7XfcQa5c5vjYBF/52tfZ8/KL9GuP37jmVi6qzaNf3IVNDC+omBdaTSaVodfB1TbHdikg1mJ+hov7vjHui7s5tnXAgFuUpaVl6BxbVMl0G/zzJIGZtEs7x/BgjlIPmCgm74dctmkU7VuaUQOnvWxlaA4d+Gy7eJhVAz0cn2mCqhK1DL4aSHf4S2ZwzRo2X2RqJkFd2s9f3Psq/kMJvh5gEuj1RxlvzRJH/egwDfMr9RH+w5db7B0+xG1ra+ydUNxw/XKMPcJjj+3lzNEqRin8KGLTGp9NFxs8WljjIbKYTO3eKybrOdpaZedSz1JU134eWbHEpFsJ5PM5mpFNtxeI4o70sS2FVEqliVOlqFar1Go1/IwaUzpdRXr27FnK5TKVSpVybzHjhKFc7kdlRqin2JOpUEy6cCa25IOAnkKBOM7TvsHpwh7FzPQCSikqlQrlchmt0nyA0pqk1aRUyLOQxGhPE9dqWf4hwXTJOtvKF5F0y9QVy1ZSr1dJMh7e89INqjxPs2bNWkZGloNTnevuNu5tmssBxlnEOhKbbp3QzQimu2AqrEm998QYrHEkGX0TGYt2Gi2S3h8Hts1CCiTWIomjVq0xW29R7i3SN9zPjbdeR5Avcu+9/5Xawjzt/SI97eGsYMQRas1ll13G8rVruOK2j3HgxVeYn5ll2Zo19K9ew8BoHwtjJ3GmSdyoUqtEFAoNcoUczWYLv9BDvlREbJoE9bRGlKLRauGyBUbWpusatFbkwhzNRgNj09XLuSAgNkm6b42k9FpKsSu0mEWy1WS8PEKpWCbf20cxn8OPE2ZUTFW3qLoms7k8NRfTcoaqsdSdw+ocYZxjSyNmLYqwmOeAavJaq4YxGu15LBsa5rrrruemm25icMUIVinqjTqv7j/Armef5cD+A8zMzpDP59mxYwe33norl112GaXeEguNBXa99AIPP/oEx156keU24dNbL2NbaZCFXa/iJueJlOVQAV5tNXitGTCtoKQCLpMCpVLAK7NjTEUOlQ9Zt349X/iNL7JpwyX42uPkyZN8+ctf5siRI/T393PPPfew/cod6Hye/adf42tf/RqvHTjAhr4+7rnpRvpeOsLAgRPMqRZPhy1eaDaIjbAsgZ1hiQ2x4JTiBdNE7Os3GuzG+8a4L24/AOdqXX7WDzs/Oa++RUMvcYvVy3pZuayXffos1vm0TJ0zE1VUC0JVZGLiII3IY93FmxguwuZ1q3ltcl+6cyCphtlpi44XsK0FDINMzQpaF2hKP9PVXhz9tHoCfGJUkBA1oK8vJB9CUCwzc3KAp16ZZrPKU/UV6CrifE7//9S9V5Ck13mm+Zxzfp+2KstXV3sD2zCEJwgSIEEnkAJFJ4mrGM3G3m7E3OzdXu/VzuVuxJrZjZAoxoyGEiWOCEkkRZAASHigu9EWjXblffrM351z9uLPqm6QGsZeYjKi0dXd2dnIv/78zjnf977Pu9HDZGWE18QTa9B3EINJVGiQjkXvIYrvPJHs9dPtqHfNbTbPx9UyxTUaDoYABEFIEHkwUuCoPZu/tTiOw/z8PCBYXV0liiJqtRqVSoVSqVQUupF8sFQq7e/woygesWvWEVLihT2kgIlaRJ5rpFT4vketUsZzHWQp2H8/xgLSodeNGcbDAjRmYgLfp1qpYkwhxdzcbFOvVWm22vuvOZosY2yhy4+iaP/9KKVGChlDnluCwB0tCKBEMUyGQjnD6H3tXYc74WGCguluEYVGXiqUo3AkYIthqqMcMq1J4iFxbPfdqo6QeCNI1t79LZAoVWw6pJJYtziplmt1NloZW7s7HD8wRtX3eeSBu/n5z+e42RngOpLDU5PMjtcQQuG4LpOTDeaOHcchJwgdFo6foNFNCcshvisZr47zx3/0Aq3VZUwWI0WxUC1vb7Hlx1iZkmcaYQzG5gjHJclz0JbAcwg9j05vQGaLlpDMKIaUYg9hUJzAMKMTTpqyx2V3HIERRVtTWHBct2DwiArR5HFyBWutFi9vNklTy9AUGxAtIFQwYSWnhMeE9mgol7HAx9eGbRVwWQjGZsc5efwu7r/vPo4cPUKtMYYbuGxsbvHee+/z3nvvce3aNZIkoVwu89BDD/Hcc89x1113EYbFKe7ShYv88lcv89aZs8S9AY8vzPOVg7MEizfR565iEdx0DFd0xof9IR3lUAkUJ63kaKnM0MA769tsuQG+pzl24gTf+/N/w8zsAZCCM2fP8Nc//GtW11Y5dOgQ3/72tzl69Ci+UFw8e5G/+C8/Ym11iXvHa3zvwU+Rvv0O3uIGLVxelilvm4RK7nFCW+4PfA7khrYP76g+72eaR+R/M2oZira7vaPFIESxKxpBjm4Xcrn/F24HZgB37P6LfeGeesRQKrs8/NBJ3nzlGltbCWfO3WDV7tAoVbj/ngfJsoxhr0cgNJ50Obgwi3njHP3OAKUyXCtQxBgnwZoWMo9pdgqHYk0IUrHL0fkOCT69HRecKRw3RWc5KqxSL2kcOyCThtLUJDMyxToZrvLBSZCii0w6eJ0PmVYeft4A4WJEURD2duh7p5ViPHEHovi2wn+0e7pjXqHEviKmXC7h+opme2d/SH1nQdwDjQVBsK+Y8X0fZ7Sz39sZJ0mCEIIoLLG2toEAgiAiTRLCIEQpyebWFnOz0/i+R5LHOErgu7ddrreV0MX/t841q6srzB+YHqGCs/35ghyxS+J4iOsWmnchJDovODlhGO6blfbAZEpKXNel3+8Rhh6OMwLuCj5WxPcQxI7jkKbp7dPRnmNVFBJEowuCozKCPMnQ1iCVIEWMdPsFoleMFhYpBNJkOKJoVu/do8ZqpDFYCgKnxrC8uMJ7b5xlfmEWdMzJhSnCMOCRRx5nPCxzam6KKSclGeyCdKlUxihNLyAm6uB5NOo+YeiQpRo/8MnaTdIuzIyPEy99iDUxOk/xpMexeo2aX8KRHqudDsbo4rSHwLEFi991BCZLcR1VGAqlKu4Lv/CcCGvBaBDFYFopRWZHRjkEoReQZjlWMlo0C1Vklgl2toaUPQ8pAuKwTJjGNGJDpFPKnk9DeUwrj3EtkVnKwMZseLCqBCv5gLZf5X/8d/8TU3NT2LwYhi9vbfD++bO89sqrrNxaxpGCmelpTp48zuNPPM7xU8eoV+okvZQLZ8/z8isvc/biOfJ4yJGJKT77wIMsDBPkmSu43Q7LruH9dMBFa+lqh4YJOYnHpOpSLVVYzhLe7vZYx8OXZU7cPcef//f/lsnZGeI45ddvvc3f/PBvGSYx9z1wmm9/65sszB1AAL/++S/5h7/5ewbJgCcOzfEHd51g8N7b+NtNmhbe0kPecyyOcTktAh4MHUKb01SWdxnynjWYyWn8UvR7a+onprhbsZeus9/tLOq7lCRJPNII78tiRgWe28/dF8cXg7lC3lD8EI6HzXPuuW+exz99jJ/+41t02oKnHz1NxRNMzU1z7Phxdjd2mJ6IWFnbYnaqTKnkMBxa3MBD6j5jkcd2nGLzISpL6fQiBvGQKXeZL335BH/+QoVU+/zvP2ixmw24dXmLEyfHSJwO0xXDQwd2+MJTZQ6pXfJrOVqXcI3GSAfXNnEdzdMnpnj4IISygxHBiI99p/pnpIbh9uL1MY07gJKgTLEACgHSQVtLGI4SawpyGjkWz3GQjoMFMmMIfA+MJU4SXMdBOQ5+EOyTJdM0Gy0oBYukXq1RKpdot9v0el0836McBRiTU69OEEUlcmOwWUZzt4s/G6GUwFGGLEuxFob9GGNygtAnKHkweu1U5MjR999ah7W1HYKwhtE5SmqEUKTWIpVLIS8U6DwFLFlqyHVOuVqj2W7Taveo12tIIcj7MZmXU6uP0e8nCBmQ5+wvJPunvtFP2uqRV6CQu1opiEemGJUVi6kRAuEU2m8XiRUF9AzhoK1BZ7poPY7coJHnkWMplWscOHiA6dk5ZibKHD3UwHMUqc6p+x7Pf+4xug8+jDWWweqH3Dr7KuWSTzl0STvbRFPj7GzskMcJ5cCnl3YRdoBMh5T9kO3Fq/gOaBQ60XS621gL1TDkMyfnOL/m8f6V66Ry76Qi8L0QxxkFdWQWoQS9uA+G0WzBK+5Hq5HSQVKoqKwplEg2N/SHcbHgR9G+98ARObWGZP5AwBiKVtyn29vleFVy0gR4KcWMKpNkUnJLZKzZPttZxm4O3cCnp1wiX9KJe5TjEpfPnef82XNcuHSJrdYuGMORhQU+8/TTPPzQg1RrVbzAx+iMc+cu8PJrb3Dm7Fn0sMdULeKZU8e5Jyyjr1wi2dpBOy4fWs37ac5ODK5wucvxORI6TNkc4UiuDQacaye03WJBfeDBE3zve3/CzMwMxhh+MIGc/gAAIABJREFU89prvPTSS+S55qknn+IP/vCrNCYm6HS6/N0Pf8Rbv/41ymY8du9xPj85z/A37yLTLstS8a4ectMzTBiHx1SJk77AsYYN5fLWoMcVR6IbYzz5/Kep/beyc79dqfcFfsWR3VjiYUKl4mF/57m/59XuaMsY7aCkplzXfPd7z2KMZXenRaWqUGkfm7Yx0tDeXmFj6Xoht+xbFsZD+utdhnGKsJL7Tt+NWVxkd0Oi85RcK7Z3hxxZGPL5B1ucGh+Q4vDkYyEXzw5w5ibxkiGu0gx1wrefX+C/+9I6G1ebtDYFTjaBlxsc5aEJKUcGx+uRxXXWbw2YudtBSA24/9obZL8GCXE7aGLkqBR7eFchsEgGwxjXc1GqUJMEQUS31wepUFlOrgtnp1IOeXobPQB7bs1idxrHCd4Iset5LkJalJKMj9cKnXSaEoUOvheNBrRgtUEYQ25hfWMb15VEpQCtc7I0o90eoGQRm7bdbJEby3itxuraGiBJkpRadYwk1bT7LSbHa/iOS24MJs6KXrGxJMMh9bE6jiMxRrOyuszk5CSVSpU4HtJu9xj0+gw7Par1Cp1OF6EChHCx9uNpVHdcZpyR8siMmu2OUsVOdvRnUJxEQi8Y8dyLwaIxpniukpClxX1rQCPwvIDPPvMMD3/6abzyOCaHMR+mJwOCfdIk+JFPmnv0hhm2Mo9bmabsGZQjcfOU1s3rGKeMp1w8Y/CsRRqDIy27Vz+gIg2OK8itQ4xg0B+QJDH+MKBcGnBXo0x/ZpoLGxtYCma9MaYwSnku1uZFOMto/qWkwvddkjjB2uJM6QiJEYV7WCpITbYvdDCmUD8hBNKF7dUbfHThEkFiMQqkyumWx0iiEjuZIBaWpshZyvusuTlJBKHjUR5vMH9ogemZWVzl8dpvfsX68hrL126gk5QgCHnwnnt56rHHuP/e+yjVaqRas73b4sJv3uHcb97ho6VFeiJlul7ms/c8zN1RifL5K+Rrl3CkYJWc99MOl0lJTY0TMuBhJIeUIleabaG5LhzOd5vsKg9Vcvjcs5/hG19/gXJtnDRNefnll3nppZcwxvCVL32Jr3z5y7iliGavz19+/we8+/pbjLuCL37mIWoiY+Ot1znQ8/hIe7xrY27IjLrI+YIMuCdVdDzFZddwvtfmumvQB6d46oXn2e21UK3b7vV/7fHJKe72jgbLXnQbRYqN1mDMb6cFffyxV9g+rqTce17BKnX9lInpEv/2f3iBM+9dZnXjKsnONs2VNnONeSqlOs3uOp6y6F7Cp44f4dqtd+nlCqMi3j9zlkR4KBuS5UMMLotrKaXKBE48hCTCcRRTk3X+45LPmWmBdDfYze/lvatbPHaqTqi3CE2GtGPooUIlKZ7vM1QQBAqt25hsls5WQmMhxa2ofUno777fj/96z1UpxMfVMsYKSqUSe45T13Xxg4AwMCTDmGaziRCCSqXC+soa5ahEkqRMTk4WKpksLyTWUhJGHnfffZIszxgOBvT7vQIwFkZMTk5gjSUM/X3T0F7iUhAEpJnG81yUEvh+gFKSZnOHcsUjz2FldZWNzSZhVGH51iKn778H3w/54IPzLC+vEAQRSa4Zr0UE5SpISZJbOt0evpIj7EDAcFhgCfr9PuUwhCjCmDJKOWzLbaphxNh4nThNaLUHOK5CG73fltq7nvsDfPYWz+LXcZrguLcdp1hL4DgoJI5TyAjjNC4Ku+MVp6XRt8uVikw7uL7DwSPHEE6ZwBdEJcFk1Svex2iQ20s0u92Rpj5QCN9j8tAJspWPRgYdTd7eQFQy3Po0Os9RQBR6bF67TEXlBH5IbjLSdMhgOGRje4f+oE+1UiZNKkRRykOHphlv1Dh3c5lBmo10/8W1qJZrGAzuaJdYaP4NYRjiuwVzSDqFI9eRDtpoHJVjxQja5xSfyzQvfAQPPvAE5epR7CDHCoXOE+xuk1+tLnKTDnI0wA1rFabmZ5g7dJCp+Tm0A6vNTc5uLdIedqmLEpN+EQxz+tQ9zM7McO/p+3B8h1wYbm6u8OvX3+S9t96jubGFyjNOTE3y9JG7OeZ79K5fY7C5TmY9eoHP5UGbW2lM01hKfpkncLirHFLPY7SCZeFzthVzzcb0/QC3XucPvv41vvTcZ6mXIzZ3uvz4xz/m9ddfx3VdXnzxRZ777DOM1cpcvn6L7//gP/HRhQvM+D7ffvJxptZW6S3eQGcZr9shZxLNpnKYtC7PyYijxtCLci6lKWeHlpXAp3Jqnvs+/ySL6ytsrqwwPXlnmunvPj4xxX0vvObOHfzeQFBKhda3Ax5+9yH+la9uP6Qc/UcqlOdSH/d45rl72boa8sGrb+OpiMAtY7VLfWyOOB9w70KDaGmT1bUdrqwPWGkmDOIuuarimRzEgCTLuL7S55GTAb1uj6EHSmbUyjG7A5e/+qerVL4huXhxSHMtZ6u3Q8wUUgg6qU/DKaNYQvoBB08coMw20gp0kqK0pr+d0yiX0OJ3KfbFpRKja3dbTbPnXi3WulFx14Zms8nU1BRKKWq1IszCWksQBrSbTQ4uLBCFEXGSsrO9w8TEJMZAmmTUx+o0W9uAJQwjhsMBjnLwPJeJiQa+7+2nMUEB8RKwz1Zvt9sFGsD3mJ6eQSnFcDjg6tWrTE6NMzs7hbUOg6EmSQXjjWms7jE2XkcIxcmTJxgOP2Bufp71zW1ynRNFAY7rcXNphYmJKRzAYuj3B0xPT5FlKVEYICX7LHpjNGHgEedDSuUIT3sMkyICME+TjxX22zJSiZCKNMlh5HDNRdGSiJO0cGFaCi6N41ALfMpRRFVV6TTb5FojlUOcZWAMw9F97HoB2gpWV7Z5eHqBWlQMfPc09t0kJ8kEWWIxesjqratcO3uGaj5kshJQD8eRyserhORa09vZpj47x7DXpbXZpT5eRwUKEk3e6TDst9jcbnHh1hpWGxa0RWjQOqUuJCfHxqj6h3nlgw/ppAm5cLDWpzxWIU4TsiSh0+0UrUApKEVlMqNJTSEtdhwXoVykU7h1szQrpJJGkgyHDIYDZDlgfbnJpStbpHHhesXmSB3jV2rMzU0RVcqUamVqk2PgCq5vbXB+6TJumjMVVXjo4EkOzM0yNTOPI1xOzB+CJKO92yRJM85fv8JbZ9/l/ffepbu9Qx2Hh6YPcM+JAxyWLpXrq+jFNYRJSVzLu8D1OKElLI5wOapC7ldV7tE9Equ54SsuZjnXBppdfGzgUp8c58XvfJPPPPM0vhAsLy/z1//ph3xw/jxRFPHiiy/y7LPPooTlzLkz/If/9/vsrm5wolLm648+RLSyjHdpmXomeMvTvOLF9H3JvFU8qeocSzJ2wpS33AE3E0PbK3HgkXs5+dRpzl49T2+7xWfu/RRT4ve7mD4xxf1jZdne/kkgkdIZ5YXKjz/ht1/hYxLB271oIfNRtyfCZC6WHOn1mD/eYLbxHO//7E0ufXiW48fvZavTZf6uIywsTNMbrvCNL5xkqMZ498oKP/nNe2y0QJgUbB+DZm1zm/J9E3S3LYnSRDZlyhlwasbync/fz10TF6j5fQb311DNNVI5iyl7mLGALk3cUCNFhVCFPHnfaa6/fZbWzi4TkwGdrZBSQ+FU9P7wc++HsYUZ57d/v7gOEint7eJuMpIkwfd9xsfH2bPtR1GEkoXEzXXcYriXa+JhQqdTwMEa4w12dnYRwtLpdgjDkHq9zq21ZSYmG7iewvUKbXkcFxpyYzSe6+23i+I4xhjD7PwMUkK70+Kjq9dIkoxDhyoMhym+X6CNd3d3GAxijh2bxXEKLk69XgUM1659RD9OCH1JVDpGfzBgcmoCpRQOkiQt5I5ZluL7Pp3WLtVKiaAUFv14BNYUUK9Wa5dM52idE4Y+zUF7/x66jf0t2ClpmrDnhBZSokYF3RQyEaSU+yn1aZqQZyNjltZkaUwaD1GjUygCHEehsTRbHRoz1WLgraHkO8S2WIwvX9kiTy3DJOWNf/zPnHn3dXr9hK995XlOPvEAYa0MGgrspKVqBbmFyegAiOKEoC0IrfH6Pfo3rnPr4i0+WmlRCQN8p4NnDYgIZ4REmCpV+fSpo7xy6SpKSHSa0drdJTcaYSnaekIWpmdjSJIYISU6z5HAME0KtLItkM9WG8jBdz2Ucgh8xYTj8tSDJ1nb3KIUBlgd08lalGbrlObHWV5fYTFuMlhaBmk5MHeQ5+65j7sXjjBbbVAJSnR7PbaGXW6urZIMYyqOz1uvv8H58+dZX18jy2IiJfjskaN8euEoY4OU9sVbDDodEmvoeoJrmeBWomnqHp5RnFABB1yHw6Eism2aUZlrScb5fs6ShI7McSsOdx25m2/9yXc4ce9doCSXr3zID/7y+yxe/4i5uTm+9a1v8eCDD+J5Hm+++To/+MH36TQ73D03y4v334O+8C75ToumCricas5nCmtdTkvB/VKx4DrsOPA+KefijGEQcfqzj1M7Ns0r535NliV88dPP8sLjz/Puyz//vRX1E1Tc1UjWt5c/WQywlCzmg9Zo/muIS2tvy9buoILsjVkRIweksUUakFQ5lhTlJgxzh+sb29Qrilfef5N7T5/m0IkZbt68xNi8w6wf4kRl7n/8QU4/1eD7f/UWFy7mxLmHIqHTGSLSgF7LxY0jlIByBMdnYj51T5tpR2MXz/D5w/dxYXtA3I9p72wQjTXopxrHaDBjLC33mXn+JNsHfHY7Owx2fZQ3xfZOQiMs3I8Ig7aFkWQPQVC8f3tHC2GPSS7uKO7Fc13X3Uf57qF7C8qmYXlpiUqlghAOw2HC1tYWjuNQLlW5efMmx48fp7nbo++lhEGV3d0uYRghhdgnR/q+V3zgR/+u1prhcEgcx1SrVaSE7e0t4mHKqVN30Wp1uHF9ifFGjZmZGmmScvDgXGHzt4UbtrD9G44cOczObpNKfQxXWiqVEtpoXDcuIFbG3NamWxgbq6PQeI4ahT8blIBqpUQzz4iTIdoYarUx2u2YNEvvQCR/vA3mqEJJUrzXopi7ytnvq2d5kdUaBcH+aSrXGkdIgiDizpBso0egMWu4de0yDz/yANJxMCji1NLtZgwHMdvLa1x8/y38yQP0U0HuBmQeHDj9MJXZWdA5jlCYvEDigsF1CjaPtYAROLY4TXnBOHP1Bidjxdsf3mBsdort5jqm2WPCZmR5sSiXkoRGZYwnTh7m4tIa1nFx/IBhkqBlITXFQL/fx5DjhgEWGPT7H29nGYOVRR5DpVzBVQ5JlmLQ7G4vc/bDG7R6PYzOMXlMQkpeDrjr6SfwUs3JygQn732MQ0cOc2D+ACUnIslSOkmfC2sfce6jK1xfXmJ9aYOaCLHtmPVbNwlcwXR5jPtPnebUeJ1yc5vk3fcY9nqUCFmxcF4n3BKWvhEo6zDlaU5Jj3ttSElbOplmxZe8n8dspBkdY8mVpFTyefrpJ/jaV/6IyZkptNa89dZb/MX3/4p+u8O9J07x9X/zJxw9chQ/hTdee52/+Osf0Gxt8/DJozw3NUf7jTcJBwNa2vJu1uWiL5GZ4lN5xOlIMG4sbQyv6h6XTUYaRDz56afwDk7w3pXzBEHEH37xBR49/iC+CbC/e6D/2OOTU9xtUbwK0WyRm1Tslgq5mdb5HUoa2N/F/5ZF9fZTbn84pZVYsycHSJGkWG3I+n0unGnS6mVUxxQHTxzgmec+RbN7i9q4ISz5WKERsoOnBjz7lM+hic/w7/7nn9PZ0cg8YagFrbQDcgOdVjBejC0prJB0Yp/58ZDD90wzMAmHnIPs3lhjpjTD0QNVzn20S9l1EEnIIK/wL6+c5at/+Bhn/+E1lm4tcagyBW2feqOG8nOQOUoWJD9LjtWjQj/i8BljsSNlx15bpoA3yX2yw3A4LOSQnltosZXiyOHDOI4z4qRbjBGsr6/vFzqlFJ7rI6XC80Ly3JKlFpNDPLL4Z1nG3sKSpnFBIEzT/bi8IAzptLtcv36DU6fuwvc8PNelVq9y5EiRRJNmMXMHFjAWHFk4HwsssIOjHCYaU9Qb4+h0gBQC11HkWQLCRQmFpSjEclTg8zxD6xzH8QGLdBVCRDhugLE5wyTG8yJc1ynkgHcU9UIKWSycwzQpcA5SoI3FZoZKVKJULiOkYGdnh0TnZN1e4cQNAtJco4VFKkUSD7BGF5RIaxGZxNgh586+y/j0LJ//8ldx0Qit2VjrsbWxRXt9g6vvvErj1Gma/Q5hrUo01mDuwAy1UkQ2TBkOemAseZZQqY0hneI+6LZauMoZXSMX6RRDyZMPPsDJ+x8gzQcYJWjubJJsd4j7CdUwoFatMGY0E+VxDk7U+WizzTDrFicUCpiZNQVDxxpLGic4rnsbj8wojWnvEyglaZ6O3NEgA8lOv0tr0KU37O/7LKz1cGLB8w8+zmP3PoBvBcp1ijDtLOGDW+dYXF7k5sotVjdW6cUDHN9lZn6Kchqws77Ls8eO88D8FGNKkq3v0H3tEokpTtNL0rKRDllLNS0lUEJwxMKpqMRBEREBmc25rgwX8owlJLeyAbk1+IHP3Nw0L7zwAp965BHK5TKdfo9f/epX/OQnP6Hf7/PoI4/wnT/+ExoH58i7A3760s/523/4Mb18wBPHjvKFmTn0m2cIujmL0uWD4ZBVpzAVPm4jHk0drLJcczM+GnS55hqyWoUnP/8ZKCveuXaB8dIY33z2y9y9cAxrFe/cuMDa7ubvLamfnOL+rzxkoWnEcQquyZ399P+aycn81pOKouYWu0EMUuYom6KHfbKO5vKZ88xNVDk0U2OiPkVr5zJOJcZzClmcI32EdVDWxQ6HHJ2r87kn57nxX9YQ2UF0FtGJDUpBP09wnB4yH2OQRPzT+5bakz7G93inH3Lu1Ys8/dQJJho16ptlrG4RlkUxVIqmefuDs5x+ImB84S52r7+PXLvCwUAwLOWEs4pMdrHSwRFRkdHpalzhIKTE2gJq5XrqDktTIQ3dQwREpRKlckgY+DiuxAiKwWg8KEI5hj7Gqn0m+t6Mowg/VrieRKki1Nn3XVzPQ0lvxK0pYhGlLPJQtTYoVxVSOeVgNOxud1iYO0ye5AxsFylzjhyeAxHjeFCvV7l69Trd7oBTxxeYnZ4svgfKpdMesHhrjepYmbnpBmJBEPkB5TBCC8UwTnE8F4ndDxJJ8wwrBTlFEcmModXtcO3aMtMzk2RZQrXi4LkeVt8e/u4X+b1Cb2/fTHu8GQfotdujk6FFKQFCkmYpwyQGIDMWJLgjsqmUzmjxsAgriNMhv/rZTxmrznJg5gBRyWV5aYPdjU2GvTYb7V16Ny8g3QCBZbzWoBZ6NJfWUI5D69oV/EpELBzSXDJMc1xXYjs7hEqRpwlD6aOkYnrhAGOliLGJBq2mRsg6tlQmbW9za32VsNmn0U/QQjGmFVOlEjuVgFZ2m2sfBGExPE9jsjQhSzIslnq9zmBQLLjaFFgIJLijAXWlUsNojSMSBjqlXKngjpK+yuUyY2NjVColStZCxeHcyk02d3bY3thkafUWG70tyC21UoWDs3NIC63dHdY/+ojLNzY5RMRzd9+DOfcBw85aMfSNBUsq5KLN+UiClR5jXs4hG3PCkRx0Q0pGY/OQFZVzxna5JhIGnkevl5KrAlp4991386d/+qccO3aMXr/P4uIiL730Em+//TYAzz77LN/47ncIxxvYzQ5/9x9/wMtvvYYwmq/edx+fcgTuyx/gDC0XPcmraY8dV1Mm5TEV8LRxqCiH97ThlbTDuqNxG2M8/6UvsGv6vLl8gQNTh/mz57/FydokOk35p3Ov8rOr73Mf5d9bPz8xxX3vFHy7u2Bvt1tGR/zbwKzfVo7sf3X7N23RljFYhFBF/J7MkeQ4aYqXataX1zg4X2es5jM7OY6LxKEz+vcDRB6Spx79Ts7u9jb1SgPXgfuPHCZyF0nyHjb3Ge64lHXEsK0Y8xJcPaBUGuf/+TuD0/NY3FnjRz8PeWSuhFP10P51hLdAr9shKhlMK4dglsHODj/+h6t894XH2G122GhdJ1q8zlQkGTt2mNS1DLOMQX/IME/ITOEwzLLC8m4MdFsBUgZIe9ulKmUBEguCEKUc0uw2dsBiWVldZeHAAcIoRGtBksSMjY+h85xKpczxEyfwfJdDhxawBpIk4+TJ41QqZXw/2IeGFeYiva/JF0KOLPqW4TBmdnYWYwyXr1xkdnaaAwfmC9ZNnnHj+nV8r4TOd2g0pslzkMIhNxnaFOz3AwsLtHtN+oMB2miCKKJUKdPtJ0VsHgWBMYpKJEmC6/qAJM8MRhcDXs8LCIKA8bExpBKsrW4Ww9ARemAPrby3sCmlqFeqxY50NLgWI8eqVKowP0lV/FoIcqtHNn47ojf6OEJhdQ5CIV1Fv99BjHYhJd+jXq3S7/bpdxMsOdVawLVba0wdu4th2kOPzFMH52ZJuz2u/PoNKvOzeElMu7NLHJYIhI90nSIWcnODfjIgqNRZ2ljj1tUPeeRznyaYmscLXUqmgml1SHKLjuq404Lm8g3UYMhsqul1uzAc4guPNEuwxqKUQ1QqiolyFPGg+D5rY0iTBJNrCgWoQchC8KZzjUDQ7/cAQeRZsizjO9/5DuVymTiOaTQalEol+sMusZvyozf+hV9efI9hmlCyipmoxNHpGUIV0Nloc/PNS/R2u6S9BCtcJmoLLMzMcXV1mwkr2bVllpIhayalJQakSMLcYY6UU5HHIRlQtYJM+CwJh3edHovZgL6nSEVE2o1xc4UKXR594nG+973vMTExQZZlXLxwgb//8Y9ZWlqiVqvxxS9+kc997nPUKjU2Ftf56//r/+atS+/hlV3+/LFnOLTVpXf2fWQsuOIIfpl32RLQ0Ion/THuzhxSJ+EVN+aDbkbX8ykdbPCprz3L0u4mq7du8Oipu/jWMy8yWWqwkXT4xTuv8s6lMziBIgr+FYn0HY9PTHH/eMEuvr4zeuz/30vsySGL3iuj3qMu1McYkSJNRtbpk63tgO5z+FiFyfEKE/UZOtsJSvbJjcLmHrvrKVfO30TmionxWYyaR3mGuu8QeT65oxHktFuLeDJla3uGel0w0BlJug35AXqxh0Jj85yea9mQAUJk9BPo7BrGyi75sIXjT5L6sywvr/HOhx3izjjHp3y21t4jmW2werNGFglkIHFGDkJHhIhAUrSoRrCSLEXIFLmv25akSUFQ9P1g1IsvAselcrAUZMQwKhOEEa1WG9dzuO++e1hbWyMqhYyNj9Fu77CwMEe326fZbHHo8ALDQZ8kKZKOhNAjz5gdoQFEsejYIootz2Nc16HTbjM/N8/8gdn97+/2dhPXCWk2u0jhglWsrGxw+MhBoiig3eqwubXFzMwBhCywxcMkYXN3m1pjkt3OGrku+uuVSoU0TWi1mvieu+/gTZJCZ+44LsNhzGAwwA9uD33v1PTvewYAIQXlehXX8Wi3WsXAGEjyDGEhT9N9eaCSDpaC8miFxlEe4+NTRegGOQJJr9splCQUdvveoM+V828zMzaN47o0pqa5dOMyJc9l6Ef04l4RKagE7W6H7Y1tlpcW6Vw8y31PfQYXn167j6wXbaxeliLdEJ0M2dzdpd9u095cptPs0rFbDHoDeu0OnVabLDMoaZmv1hi/725qoYc0hs1mk+YwpqdCTFAtNP1IOu1Wwa1XEodR62SUUev5Ho6UaJWz01mnXC0TOlWyTAMFyC03lsAprvmpU6dot9sjoqQgcBS9zi7Tqc8Tk8eLFluc09nYZOXSR+xstEj6Bsf4jNdnOXj3CWrjs+wO+yyv3WK1s8qsk7LZz2laj9yXBKQcAU5Kl4MipeIYYuFwLTN8OIxZ0pZmkIPromRInmZkUUgeSJ564CH+7M/+jEajwXA45NVXX+VHP/oRnW6XhYUFvvOd73D69GmEEFy9eIm//D/+A7eWrjMzU+Vrjz3G1LVl0iuLaCl428t5Q/fpe5L5GB4QAXflIdYa3nAGvCJjdCXknoNHufeLn+aNtcs0d7d49vSjvPDks5RkwHp7k7898ypnb37IdLXC/RMHyFa2fm85/AQV999ttAjB/m7Kdd2PKUI+9jf3ZYAFf9rY29Fp1hjQOUbmGBsjdYYT57RXmgQzhokxH2VTjLA4Xg1tdzEmJ+kPWFtpsnjjFkfnD6OTId3+KkpE1KtVJhuztAcO2lG0soyKu8Bf/d2Qr3z1FMstl1+9dovvfnWKFz+vUGnIfYc8/vM/GC4sO8jJU7z+epfdtTJHjzqY4Rr+WIO8PI5tWd45v8H8/FF2+hs8e+puxqoRc7P3sDZIkV6O53fR1mCMGNnwVQF40gaBQYjsYzFxwzjB90N0bnHDInM0y3NczyPJUnrDmFvLy4yPjaNzTZ5nOI5ifn4O13Vot1rU6mV6vS5B6NFQYywt3STwAxyliKIQ5Uhsko3kkBbPC3DdYjhrrWBleYV+vUKtVsXzPVqtNnE8pNPt0OsPMFqOzESGOImRyuPmrRUOLMyysr5OVCkRZyl+FFKpVdFGk6QpZ86exQoHz/GJBzk6y/C9ov2Ra8NgkIwi+orTnzFQLpdYXFqiWimjlF8MCkcPZ5TYtHdK1LlmZ3MboUY4BF2AyoTRCFkEhAspR+oYO8oQLWZGEkWntTsa1AIUc59SuYTOR7Ok3PD2Gz/DGw4Zr9SYnZ4BLYmQ1EzMUFpyPwAFveGAq0urjB08RP+DDc6//iuqk3OISo1gNsOXxYB1c7eL3m0hMGwtfsT0gTl22n2EcQlcjywIcMYVnjbMlCQ1naFTwyAestVuc6vdhbCMlQ6+4xCVIqR0Ciia0WhrQRXB4r7nI2RR4JNhTGuwTS/dZbDVZX6ihs4NxmrSJEFZRaVeY2dnh2azycrKCsYYtre3aTV32dzeIM0LGNzSjZuk3SIaUXk+5coER+85ysT0YZJMsL6xy8Ur10EbQkcxOTdPL98k6WxxIFfMasFhv8KEI/CEJgNuWsOMOT14AAAgAElEQVT5OOWaKRQzfingnvmTDDLL1ZUNbBhQCx2effpJvv7C1wnDkF6vx09+8hNeeuklQPDggw/xx3/8XQ4ePEie57z++uv88Ic/ZGtng8PzDf784ccIzn9ItrZM7BjODzWviZiu43IiFnzaC5kTkr6jeC9PuRxD7nrMP3qKg59+nNcuniXJ+vzhE5/liw89gzEelzuL/PztX3JtaZHDjSlmwjFe/skveODI4d9bUT8xxd1YC0IWvI0CSQSiGBBiIUcT5+k+JnWvD2qsHn0YKXbpUoGjcJG4KIxysKFFS43Exxn4xQewFBJM5ES+gbbL6soiRpUYm/aIk4xyZYx772uwfHORbm+LMCih9IA88/E9y9zUOB9dlVgp6A18SmGVf3xth5+eXyHJp4mqEzx8cMDRiT6JWOeFiRO8c87nf/k/t5jzPS5d2eLUwUNMTEnceB0/HBBMRvRzj/5wlfUmtFcT7p0dZ8ykdN9/k1LlMKIUoMctTiCwwsVIDyt8hHRxpEC5mwjiOwaDhVY9CCJ2dnYIowDPczHWoNyULMuo1upYYGNrC991mZ6eAgzBaGc7NlbD88EPxkjiItZufLxOrzsgTbMRl0URlYJRzukIH5AmGF0M3+699z58r3CO5jpDSgjDkEqlBKNA9MEwx/X9EZ5X4DiCdrtLVCpz/EQdrQXKc3DQJFlGEIaMNcbpD1Ia9QZrK8vU6zUElqXFRebm5tDa0O32RqcLiTGWiYkJhGhgbI7rhKRJ0ZPXWu9neO5dOymLAHUrBK7nEngejlLkWUamc1xVaPyNsDhSoXUGRuAGflH4hSBPUnSe47guwzRBKRcXi+u7WMD3xwk9j62lm/QuneXQ9BzlMKKaJkS1kB3tkDsOnufTTTKazQ7+wlGGrTY7zQ6+UaAlVhctsF5/yO5Om93NVWpRQFid5NrNJQ7dVaXV2sVXMF3xmXEUZtii0+vQ7DXZaffZ1RKqU1RqY3QHQ9Isw8ZFYEzhWs738l8AsEmMcAucdJ4Z2t02pXoZrML1DcJqjJZ4noOjJE888xmUG/Hv//3/ym5zF2yButDWEFuN6yhCPyQKKzQWZpmamqMxOUe/K9ndzbhypcUwGRKUHA4dmmO87JF011i6dY52a4lnKgGnUkM5tngmpY/PR3nOYm5ZzIbsln28uWkevP8klcjjxpnrrKxt4Xg+JxYO8aff+AanDh8m8QW7rRY//OHf8Oabb+A4Dl/4wvN87Q+/Qa0c0d7Z4eVfvMyP//GfGeQZx48c5MW7j5G/9Tpqe8COkbzf13woJMPIcrfQfM4pM21c2qWQd+IB5+IBWTniwcdPU3/4CC9fep8KId/7wos8evg4uYk5t32dv3/nX2jvtjg6Nk/QE7z2T/9Mp7eDPXLw99bUT0xx19YU/XVVSNf2ghGkLAYv0il2745yihvsjti9PbSs1cXrJFlOloLKIMeSu4bcDqgEZeKWZaY+Q1bfIaOFyCzxToorfPyGiwoEXl4BXFxfML8wxsatDp5XQqoyjqpiHUs58pBaY9EM+jmNckoiDN3BGJ6YYGoiZX62jKe3QUwgVcIXvljlpf9tnGs3hpjEMJbk1BqThDg4juIzXzzJL//xHP2dcbp9TTKs8OGWZHreYHaWuHV5nbnDJ5EdTXWmwczBQwS1BsqvoGSAgyKOzyLE1h2SUMEwjgnDiGarXRyF/aK4O56H4zr4fgjW4jgGiaFer1Eul0ZhFgopBIaYIBjp4W2R75kkKdUR2nevCJoRhrSIvfPIMo1yXDzfA5sX+GCnwN9KJYogC2kR0iHLTVEcpcRzI7Q2ZHmKHA0ljTEIIzDCUKsVzsnMglI+WZZz5MgxrM7Y3FhjYmKC3d0mCwsH96Fpxc7djPJjDQgXKdxRylGRBCWVAp0ji5sQAOlISpUqrucXOnelyAYxSasJviKLi750qV4hjvukwxR/xJQPw5BSGLG5UexKzYjDIoBEF2A01/dQno9TG6M77HN9dYmZsTHmJmcIPI+GE9KSkmaWE5ar3Lz4AdVqRLOfE7iSajkqzFK2yCcrOYquzWlMz+J5IcuLi9Qm5rBGUydj1veIbELS7LK+vc1Op8fOcIAOaohGAyUK/n8QRehcgy3yT4PQZxAn5GSU/BqYAoLWjTuAZjDIyRPDsJdTGvfIvSFGCPRAIGzRThsfq1MdmyJJC969MQXRUylJIC3CwKnj9zM/dzerKz02N2NuLi9jtaAalTk4N061OkGaNVlfv8LZqyu0Wzv4nmJm+ghz5TJye5sN02M7zlhKEjaEpB05BPOz3PPAXdTnp1jfWufylSsMtjoEjsuzTz3FV7/0ZaZmptnY2eLqxav84he/4MqVK3iex5e//GW+8KUvElRK7Cyu89c/+ht++u6vCRA8e/wuPjc7h/7Nezg7bW65it+kCdcR+FgeEB5PuB5TOWw4krd6Xa5jMPUyD37hKbyZGmfOnmU8rPPdP/gjHjx4hGGvxQdLl/jpmV+z1drl+Oxx7HbGG798h2yYcerkw4yPTf7emvrJKO5CMN4Yp9tpF9pmoymisC1aQ6YLw0TuFAOEbBQ5ttdT3tMnGyzCSkLhIDwBvsQTAowhtwU4Ke5nrC4tk223qZYDXMcjtTnGZviRC6IYNjpSYLOMer1EZ7MobMjbAReu62FMYWTqdxNCN0Rma5gsQpUymruGn715lWNzGZWshAxiorILWqCth8ZltxMj3YhGo8xi07C8vswTzy3wzy+tEScWQ8o7H7TpNLs8ff8UF668TFguU7FV4kqANjU6XU3e7WBtH4xkt90ZXdI99IAlTmIq1RpZP0N5HkK5+E6hZHBVsTvXo9xL5RTFVusMJS2uEriOA/gIit6rEYUkME1TfNfDYslyTRBWKPwEMBgMR2RESa5TpCpMV8bqgjMyUo7keYrnFoxyz/HwnOLkZnRS3BsWPC8oNO2OxJVFohEWTJYxW6+y0x6ysdvnxvULlEOPe+86SafdojExjus7+KE3kmha0Ja9GEJjCtSFECCUixZusblQIyMTaiT5i4njDNcrWja+56KMxOqcfr83Cvw2dDstrChSpga9LtZCHA/wHKcAmgmJVJLQ9zF6b7db0EhVZgg8j3xsnDyLWe506HQ6zE2N06hPMVWt4eNQ8z1O3Huaa+/8gnoQIGyEzgyO54xakgbrFGAv292l1elhERx64jni7RXmXI3tNlls7bKxvctunpEqD9mYJM4g6bQJHUXZD0iSpGhTYUjzlE6safc20SKlMT6JyR3yPCHJuwR+hPIlIoVBkqKbGfGgTxRF+CpAIhlmGZ7n02g0mJ+f31dllctlGhPjlCtlxurThMEU77+/yMZaSr02xtG5SRpTDta2WFk5y5WPVuh2d3A9Rb02zaFjT1IqT7Kz0+ONtUVKqc/2sMNASTq+g2yMc/RTx5k7PMfaxibnLl1CoTh26BQdu87zT3+Opx9/Et/3aQ97/PI3r/GLn/+MdqtFpVLhm9/8Jo8//jiO73Hh0jle+osfceHWTZyyyx+dvo9HE0v22m+o9CU3TJl/0V2uuDklBx6wLo9SYnwguOUIfpX1uCYtjP9/1L1ZjKTZeab3nOVfIyL3zKqsytq6qqu6eqlmFbvJblJq0qRIjsgZUSIsCYZhyIah8YU8gDGA4Lnz7QC+MmDA0MAYyLJnZAljbUOKi0RRJLub3c1eqrp6rX3Nyso9M5Z/OZsvTmSyNaMhBflG/oFEZUYGIjOjIr5zzve97/N2Of/Fz2C6mvfee5fjswv8k89/mZMnHmGrGfDWrff4/is/xOB45pGPcf+Dm7z/47dJ2pzz5z/P1MxxRqPrP7Ws/oMo7sF7VtceMuz36ZQ5WgnEXhi0iJK+RKho3BGCQqVoOZY3eoEg9hqRAd0YiiAZaRjpgHCeLEikTHBac3v1Aae7BxhYaNcdJvfUgwrRzRnVuyRpQMgCoT2tHVKUEql81GwnLvakGQcRqMisMS4lTTKmupLthxafjmjsHD+6lfCFZo5H1G1cO82797r0a4HOJrCuYns4YlTD9EzK9bWED684iqcVRx6d5ub1ljwdsrM74O6G5OKVNc6eO8v2cI00TwmDPo2Ni2HrLF6ktBaaan3MmBkvfiFmdzofwyT2htVpmgFRDbKXYGSMIYwj67I0IU2S6GLdVzLFPrZxnrqq6XS7MfRjvPPa08TrceFUSkUZ67gPr3W0S9d1vY8a1kruz1K01kxOTjIY1kj1E/yu9/EUFxk3kk7Z3YeYdcqCyUmFk4rJ3knmZ6bIUo11FXnZQ0r1N/AMe7MbEBjTonW6H0ZirUVKtT/nET4WfmsanJe0LaRpgjEjOnmJTgTOOFKdITNwvkEKhc7T2CKxHudarLBkObix0MtaQ57nCAmmAZ0pGhtdtTLJIMnJpxfY3dlgc32TuX6fQzMz9PIJwsZD5o6cwgw/RrP5EJl1IolSaRwhJlu1DSma1gW603PMHz9DXVc02xusrqyxsrPFtnIw0cUpSIsUmSna3SGqdQgSGlez2w4i8lcB2iJSQTfNUaFAJRZVBBIPeehSZiWZTunVBc5CYxpaY7AOpPAoFdBFjOTr9Xr89m//Nh9++CHWWg4cOECaFlQjh/WGUVXRm+5y584qu1u79Hfu8corHzCqtlCyYHryCGfOP8/01Azb/QG3lx9w5eZVvBAspCl13mFV9ikP9njyiccop6dY217h4uW3KFXBJx4/xzNPfZwiKdjd2uSJ02fIlObG9Rv80Z/+CT+++BbSOB555BG+9rWvcfLkSfI85/333+Nf/+7/zv2Hmzwyvch/df7jTN+5TnP3GjvB8p7QvIbllhActimfSFJOYymt53oI/LCp+EAKkoU5Xvjal7g32OLqlSucPrjEVz/zRR5bOsry6j1ev32Vv37rxyRJypnDx7j6ymVuv3sN5RKe+fjPk+VLXLx4l6cf/+l19WcWdyHEvwb+MbAaQnhyfNsM8AfAceAW8GshhC0RG73/C/BlYAT81yGEN/8uBR4JD1dXOHF0CTEu5CLsjVnjcTPGjYG2AdG0NLWhbR3VsEXKBCU84eEGWWeC/PgMzVQS2RXeY0JL0BqLpW4rJFCv1rS9jNFghFGWwsZIOiE1TagwdoRUBjeW4wXnsM5iiMO8EAJoiQ0lIkmZmNTwMOCpaQm89L7ln//PW/zzXznED3+0zJ+8vUMlZtBihpB0aN2A/tCR5x6kw9g53n1nyOzSFFlvl45Iefz8Sfor9ymnE966+iInDhXMFYeYO5jTtveBFusDKpuirj3BRwLfXnGHKEn7aHRelASm+xzzPaUIxFORlJIiL+IiO4aQWWsJYpxUROyXLy8/IJmZBdh3wcaBJPtD8P3HbQ0iVWPJZHTKCiHIsgzTNgRia82YuBB8FOK1V5Cdi3+HcR7nA1lRgoga/MUyizZ4IRgO+zSmJqdL0zT7f5+1dr89sw8Ec9H0FeP0TDTV+J8otKRWdOe7BO/xQqCkIniJ0PHFmKHJx+2oqrI46cnSnETkSC8ZtFuMmj6mMRjrSZKUpjYUoWR6ZhKPoRZt1H2rgJUNprVYoaGQGN1hWyu2t1aZtBs82p0hlFPYILBJj2ZQcfzcaRpjUalCKs3E7DSjnQ3KNCUtS25de5/Hnv4U1+7c4Nr2Q8JUQSM8eaHJtELh8NKSTyTIVuKGjsY6isksDoITgUxywKKliqexto2STyFQQiFCoB71o2PWBooyo+yWEMQYTyAQISptsixjaWmJ7bH6KLKIkniiokWlLbW9z4uv/hHVoCERKZNTszxy8gVmZ05AmGBjdcCliytU7Q4T0ylnHzuKTj3bD5ZZ29jm+a98BlcEbt+7xd0HN0lVxtOnzvL8hWdZnDtI07asrq/zo4uvc3/9AccWFvk3v/t73Ll9h16vx3MvfJIv/MIXmJ+PXotXXnmFP/i3v89oc5PTRw/yG+efYf7HV7BrK6ymgjel50222M0yjljFZ3zKGScwZcYlZXjdVNzPNJPHj/LJL7zA7Y173F9+wONHT/BrX/5lDnanub/9kB9dfo1X33uXotvj1OIJ3vyrl1m9+YBU9nj22c9iTc7ldz8kzUrKIuWnXX+XnfvvAv8r8Hsfue1fAN8NIfxLIcS/GH/9PwK/CDw6/vgk8L+N//3p13hwZZ1BhEBwMR0m+kai6iOE6PZzrWX4cAe3UVNXLU3tMSaQZSWj9XX08hp94+g89yji3BLdTgnC4oWldi3ZRMb6wzXKdoQfOpbXV1C5wjpwvsUHRSIFPliUCgQc3kf8rbcWayzGW6qqwhhB0AorEgwCnQWChKA8+Ix6u+RSo/n+2xX3h6dY3tgZx9XlqKyLcAM2tob0JjPM7R1EWCMPS2yul3QmWuptzeHFgnR+i+m85Etf+i+Znqgpkkm8VkAFvkURJYZazzAYuHGPXOwXZTluJe3tpPdCpPckf3qcFhRlaRnz8/N4ZzHGk6XxJSKlxPmwz1LZ47t77/Zj+rz3DIdDirJDXUeXqrWWpmnodDrocfDD3uBSjU8SSZLgfMA0kcyYpDHQO01TQojtHwClY19eqkBjHVoKjI1I2+DbuGsNHqkVZa+HkBJw+4W6KAqUUjEUfDy7cUiSMbnRWssebdN5jxaxaA/bbbSO6IFgA1JLMpWT5RlKBZwd4PDozFE3FUMTWTmBgE8DQlh0AppYSDtTKY2tGLmGspPg9ty2fkgxp0iNRCUC0ziKsYnIdkq2dmrWd1foDY+gfWB3c4VidpHdrV26c4skSlJkCdeX72NGFbtrDxmsP2Dh1FmatmLD7JIvTqEzTWYNBFAu4JsW6y1BRZZ/UzfxNFBpEqXJkgTfgncOi6d1hizJsCbOWrwAfOTVp3kCxkVZZPBYD64xYAK20CRaUxQFdV1TVRWbm5usrKwwGlUsL6+wsrrM2sYqq+tbHJg7xeITh8iSLs512NxyvPfhJo1dpltqjj06Tbe7wO72MnduX2Z19R7OCBoluHz9OqEUeGN4+vRTfPLCczxy7DjCO/qjXS5df4fX3rnErVv3ef3tt5gQKTsPVjm2tMSv/eqv8djjT6KUYmdnh7/8y7/k29/+Ns2o5vzRU3z57Cl45WX6O33uJ5KLVcOtkcF2cx51gs+JggXVsJ0GboaU75sR23nOyUdP88hnL/D2w+uMHq7x3Jlz/NIvfIVub5LtuuI7b/2Iy1cvMj9zgIMTB/nRn/0F23fXmZ6b59nnv8j9uy0Pltfo9Uoee/wgwd77qWX1Zxb3EMIPhBDH/4Obvwp8dvz5/wH8NbG4fxX4vRDPva8IIaaEEIshhAc/6+dYLOVEB5GnUYvtAeKATwSBcpA0gXpjwJ2LH7DxYIvFA4cJBhQpQQaqe6vo/gCXZexcucvWndtMzc3Qm5vDpBUyVfRv19x47zaHCsdk1hKEo20tXTWLMR6dFwRj0T6hk+aUmULJaBLSXmKCxzqNqXYRziLcLMIL6saRSoEkxTiNVhVO9JCkXLu7y+ljOb0Ag6DwDBACvMrY3BYcPVCQtduIzhCRWBCeTpmz25/mB9+9wm/8ygQTnS1m5xcQqcbaGukThNR44Qg64BqDbWFYuT3EzLilEXsBe47TvaIah4oRR5Dn+X6P3ow57tFHFnezjHfrgZ+4Nq21dDoder3ePq9m78MYQ6fT2Vee7BXNKNvkb6Q+eWdJtEJpiQ9xobDWRsftWJL40dBrYz2I+PjOewbDEZMTJWkaQ0esCzTtkIBme3uHLM32zVx7u/89cFoIkIylj3v5q1L+xPwVSWEBmXvyIkN5iVQJhpYQDFJLkHHgKL1DpgKVZFhvqcw2qcoRaMoiOoLb1pIIjQ2WYV1FeaGBYVMjqSNSuSiiw7c1JEJhTIWXAnJFOtth0AxZvXeDhcVDzB85xshYirzAGEt3/DwcOXmSi7fvMH9ggaleh4nFIwy31um3Q0SeQWOROmAaQ5n3EEoychX1ToPTkqzIInxMaVzrCa0nmNgSlUqhrN9PxiIIsjRD6QQBOFq0dJjGkKQKhcQoDQK8j22vNMu4eu0av/M7v8Pu9g5t3YwDzi1CeXwQ+NDhzPFH2V6TPHy4jfNblJ2UxUOTTM9OgRhx7867vH3pFqaJKVPOWawXeKnor23zS//5V3ny9BkWJ2ZJfM5oWHNl+UN+/OEbfHDnKo01HF04SuYlt668z4lTx/mt/+6f0S0nsBruXb7OH/7hH/H67fcpleeFkyf5zNETtK9dhNEuV/PAm3Wfe8ahVYcz0vJzqeLYoOZhWfCm9FwdtWwlKac+fo4nnjnHG8sfsjHs89lnPsPXPv1ZppOMld11/vCH3+Lqg6scmj/IvFzk5T9+he21HRYPL3D+E1/i2tVtNtYGTE93Ofv4AV599es88cjBn1pT/7499wMfKdgrwIHx54eBux+5373xbT+7uJuWNE1jgnjZibCnsNdTF0gfqIcj7t28jU5LZo5OceqJsyzfv4NWkKcJabLAytWGKgQ6IWeiSdi5NuTOOw9Z7d9hpxqxsgz1DhxZ7PG1XzyMkiPyTkl3vofIC4yJBcg7Q1JExkld1ySioZtGfa8PCdZUJNJibI2kwLlAqVOkSGiDJogK8hLvPCsPWl74mKagYSckSFfhQ2SI31/u88IzHWT9EFc7RmKLI8dz6tGA2ZkuD+4P+Tf/55v8s996HhIDsoNWUcPsSfEywVKDVtjG4myEpO3tVvei56y1+71uMd7RA/s71r0WzZ4kEO+jkgWxf2oSIcoag4yxfW3b0jTNfvEuioKJiQmWH6ygdUKn09mnRMaTgt0v1HstmeA1eZZQNy3GxCzW4MN4aL4X0RYdkj5EFJwbzwW6nQ7OtjR1S6pzmqplULVImVG3htXVdY4sHR7jomO/v2nbSG30nhCiXV6rn/TlpVT7TB5C/PndXkmiBM54PFFLT/Ds1lWM7pOCVGua1pCkBXmaEfAUuiBLM6rREOEdWgS0hERmiHwSISX94YjaeKSpEUoiQgz3SHTCqKqjptyBtTUWga8qDnU9D6+8y8yBQ3Q7E6zd+JDt9WVGR45xcG6Gu9c+ZOnUadbu3qQzOc/W+hqmrpCJIksTetPTtHZI6ASKNEe4QOIyesbjVXR1752WtE7RUlHv1DTWolIoJzK01OgkwduA0vHUHULc3WeppswLGtNQNRVF0YnJXomOsb8hkOc5TV2zu72DNQatM2RICK4GEdAqcOfWNRYPP8rCYo/ZmQ5Cjtjduc3FN2+ws7U1ztL9ibN4sjfNkWPHOf34WQ4eWuTcx84xNTvF8soKd9be4/K1d2Jv3gcOTR1iOp/h4fU7vP7WRZq65uDsAnmnwIaWr//xN3jluy+yurnN7EzBPzpzhqcNuB+/TNIE7iB4bVixIiRaSp7MM55UOZO24mohuNgMuRkkg07K08+dZ+H0cV56702ss/z657/CC+c+iRSeD1Zv8/Uf/QXvr97kxOJJpk3Jy9/6JtVmxZGlR3nqwgu8884dtja3OXpknqXDE7z04p+zs72GPHn4p9bU/88D1RBCEP8pXONPuYQQ/xT4pwDTU11cY9FBsrO+RXkg5jTKEBUL4KNJpNQsPXWSplGoZAJZaBYPnEQkHoRn8fRBlj52jLXVXa5+eJfhsEXqjE4xRacaUashE1OK3eEO9/qeYSo5ujSLziROeLJcQR3ffEoOSTKDDS21gU6hKcuSvCwZjCyBQFkqqqZGiJwQHHkeZZsIsK2LO2sf2NpuyHJNkowIlRqbjzQhOHZ3d8mKaYJ1+JHB+QG2vw7tKkeXjrBZGG5ch9//gyv85v/wSSZ7AW093jlMkDjhaEOLtylK+FgEhR+3JGTc6ZISQopzkhBisLSxDjFuP/jgUY2iNS1TUxM0bUy7z7SmNSCkwrbRIBZkgvWB/qBG6owkLwltw71798jzgunpaT68GimSU1mO9/HN3tT9yFgRIFAYaxFExISxfgw9CwgciQYhE4bDKn7fGLIsI88yGlszGIzY3d5m6fAiQnmcIRYJD2mW01pPfzCkaRwPVjZxIeDHA1ljTNyBChn11dbS7XaZnJkZt41AiPHcQMQ5QH97gBICryytMWCh1CU4wWa9Q13XpIlECsXk1GSU5gIGS5Iq8iROU+uqZmCGTE5OgRA0jYmpTElOp8ixwURZqAIXLEozBpYF0qCwlaCuWkJQ9KZm2H24zPSixG/eZeXGNt//8z/jyIF5epniwJnz5GWXweYqs8ePs7WyzoGjC5A4RtUOxjYEoLUjEq3x0jMx3SFgaG1L0pFIr6mHI7QuSDoQnKBpW4Y7fYSA3kSXRKdUdU2aa5zxQATGbfUbjA/kpSRJI9s9zfU4hwAOHDjA/Pw8G6vrFGXBRHeaiYkZFg5OcvDwApPT83TKWbZ2+3z/xb/iztsrNM2A4Fqk8EA8hZZlyeLiIc6de4qTJ08yM38AJQShNVQbOzzob/LvX/kedzav4EzFwal5jvSW6N/o887Ft1gZ3USojJPHP0aqFvi3/9e/QzDk9e//iCYYnj5zlF889Rjltbu4W3fRtuKO0LxSVdyXgZ6UPJ6kfCIYek3BLTp8Xw64qgyqW3Lh889RHJjltQ8vM5X2+JXPfolnT57BtiPeWrvJN17/Pmtba5w+8gjZTsoPvvldfL3GsaMneeSRZ3j91V1q4zl95hBZMeQHL/8Jw37F2dPnmZk98LfW1r3r71vcH+61W4QQi8Aenuw+8NF4kKXxbf/RFUL4V8C/AjhyZCHIICnSgv7WLkrEF0HcOY7RrzhqalRHUU5mpEJinAfRwQtN3XpGyiDmu8juAY7MHOfOtWV65TTTMzOcKxKaapuN1bu8+tIrDPs1DzfXWDp1GKdbpIgRdEWakY3TcSYmYWPVcPzEFJ3iKJ2iC0rR+obaVJR5wdaoTxAdnLfkRYoQPvYfbSAYBxLaVtGYikSPCCYBHwuv1JrgAzoRKK0wxmH8iMFWw+GDigDhQCUAACAASURBVHb3DpMT06yGjEvvjvj333zAL331CDrI+F8nEwQKJSwuhFjcpcQSlTHeRzKfThRSQZppEFF77b2l0y2RUmKModvtjtsVDVvbOwgC9XBIXQ3ROiYtTUxMIqRiOD5GQ4xti9mqnqYZUFUts3PzVE3Dgwcr9Do5ZVmQpnosqYQ4MJXUdYVUGil/Erx99OgRrLVsbO4g8Dxy4gRvvfUWkxNLLC4epDYNKytrKJUwMdEhLySJEHSzEqkS6tbQGkvrHKurG9y8cZsky6maGh8EQia0TYPWAp1kIBTDqmF6doapqSkiNFqyhzBSSjHbnWLUrzDSEoxFiQS0ZDQc0N8euyizlFFdIYUgzzP0OHGqdRXBxhbHaFhhvUUmCT4IVCIplMIb8KHFhgrfCLyPcsw0ScE7dgd9Op0uRdnjwPwpjhw5zRMvfIqth+s8uHYFWw+RWZfzU3OUaYJMEqRSHH/qAr3ZSSbnF7n4V9+n6Exx7eYbaBWxEHvDa5GMe/zeYk0dh9thLC1WgVG1g5ARftbrdJnoFggRWTLO1SAMLZ4s0yiR4C1k1pNLRVKEuDB3FTQeIeQ+gfQ3f/M3+eC990mUZmH+IN3uFEEYhHY4Ak1jeXjxDisPbtHf6UcuvEjI84zFQ4d48qmnePrpp5mfn48B6UTZtLCe4aDi2s2bvPnwJjuJ4UTnOPNpwfbDVS6+9CLbK300OUeOneXRRy+wer/i+od3eefSh5RlzUSW8svPPs85J3A/eJO2P2BTBa6GlA/qih2lWRKBjwXPo6ZGZZL3heLNkWclycgPTXPuc89gksCbH7zDwuwBfv1Lv8y5gydpmwEv3niL71x+nY1qwPkTZ7FXl3ntr76Dt5Inz77A0uFzvHHpLsZ2eOaZM2xtvcsrr/yQIAxPnfsEB+fPUlU/vSHy9y3ufwb8BvAvx//+6Udu/++FEP83cZC683fptxMiQS9JU1AamSRjpUwA1JgELFC6pHEB7xTD2tK20NQOYzR5MUkxdQRddOlNZhw9WvDU0xLXWvp+jf7uBoPbK+wObzC30CAwdMucZlRR9jRCBOragMvxbYU3Na5KWF0uKct5srKDM462qhm1LaOmYm5mjvXdDVrTMhoNKMuCECoCDgik0mGkw5k0DvFoEVaDb4HY17TOgfQkqaa2ghAcTV2RJz0GGw9JRInMM2or+NGrIy48X/LIE4tIWqwXbG1tUw22o+LDNmitaMaclxigUZNlggMLkxw8MIlOJHmWYp0dDzTjtQfLGlUNiU7QUrIwO4VWkrapMcZQlh1a6xBSsrm9w2hUo1VGkqSkSbkf3KBU7JNX1YhuGUFdWkuKPN9vs8TfK0PrBOuiOmZzc3PcwpHMzvZYXJxlcmqKj51/HCGgrnbxQtHWFdNTJYmSJFqRJ4osk1GW6j3SO1xbMzk5QafbpWraMcQs7Lep/kOURWTPxBfjXvA4gLOe3c1dWt+QZtDpZjTDhuCHdCYTkD0IkuA8edKJWbF1QEiHFAbnLcHUdHvTFFlB1VZ4awnBUXuPDgpJgnFxWN3t5CRpfO6CaxFa0i0msc7iE8vS8RMcefQsqtullxVk8wdoL3ySajBCI9FC4ZREqCIGtiSCkKQcv3Ce9eou71/98dgzoPHGkuYdkNCYlkSnOBelxQJBmiUEpTFa4Y1FSIGWCrTHNYLWNsgQ5y9JTiSnElCFZqLsUg8afPC0rSU4Sy8t91HKSimeOncOGaCuanrdiXjSDZrWNNT1iMFwQJEVdPIe0idM9aY4c/o0z5x/htNnT9GZKNBa45yL7UFjuHdvmetXrnL50kVuLd+hSgNf+tVfYef9B7z11qvs7mwiE83R42c4fvpxhsOU9y7fpx7G18uJEzPU7QanDkxwZBgIb7+PGIxYTjQXTcOHUkEiOUbgEzrjOAEbDBeF5xWzy0a3x8LBQ5z/wie4unuP5dsPOXvsEb76xa9wcHaB7bbPy1ff5JsXf0gwgWePPc7d1z/g3uvvI7zk48++wFT3ad546z5OJTzz/AIP7l/knbffRKcF589/BqVmuHjxBh9/PPupZfXvIoX8feLwdE4IcQ/4n4hF/Q+FEP8tcBv4tfHd/5wog7xGlEL+Nz/r8cc/hLzbQ0nJzPwBRJJFxUfQeJ+jnUaSoJOUVCYgEnQnReBQMlBkEoKjGlasLF/h3sYK1XAdGRqgRRQJo36g6Q9JypSHleDBsOT57jyJt/iqxQhHlk6TJF2kMzx8sE3Tz9D+NL3ePIZdXGMYtZYHq6tUdc3cRE63VKxvxRi7paVTwDAWCGcQNDhrqOto8pESFAotHQ6J24N94Qky4EU8rVjTUhQlq8OKPO2SJV1a0bC+Y/jWn69xZCbFjK5QjQI7W30Gw1UOLR6gVxYUecqob/ZJmcPBAC0hS2Lxy3ON1pClSUSzEqWG1lrSNInKCA9KCiZ7XRItCb5D8J66iXMRYz1t22IaC2O1iZRybHpJSdK4WCqlonplHwMce6R7M4A9/k1eZBGjqyUTE12sc4Rgx1ybdWame9R1TVHkCJnSzM2SpTl5luJMhc5jIXPBR85JsBFPLFM6ZYnSCYNRha2b+FyPa/feQmOMwZt2f8gaF70xITJ4GmWYnOuR5p6qP0JkCUWZUhvD5HSBSBTBOXwb/+ZElRjb0B9sEyRMll10R5GiSYtuDPC2liA8pgJvJM47Ot2CTlnQmECW5igJdVsRvAIpcF5w5NFHmT52jKYx0W9gPMZCUDkOFVtQXsT3hhIkQdDfacgn55krHM4GWm8okoI0T8kyRWNaNJBqjSxzIODt2JcQIjytcfFJ88GhFKRa4URClmUkaY7UARU0QULVtjjfUOYdRm1DkmcomZGg93l+QgjyLMN5z+7uLlVV4UNgc3OX1bUNrt+6yYMHy2zvbnH60VN8/MIFDs4vcGTpMKlOyTsFxsc229bWFjdu3ODSpUu8e/k91jc3sSLOR0Lf8cq/+zr9tgIfWDpxnNOnLtC6HleurDHsb6MVnHl8kTSrefedN9jeWGdwJeH0/BKZgstFy+u1YbNVZEnFmULySZVxqNWsJiWvu5oP2wF9mXHgiROc/PR5Lt67zmBrh088+Sy/+sIXmM1LduoR33rzh3z/w7cQwPNHn+DaD9/mw7ev4JKMn3vhyyRhmtcvX8UjuXDhOLfvvsb7b79BkU7x3Ce/xHCkuHl9lYnJg5Tl3obkb7/+LmqZ/+I/8a3P/y33DcBv/azH/I9/CLh6mjTrMlEoMmbIVImQoDKJJiCxGFNRtyNstUO1vcPucJfBcAdjKoJv9znj0zMps/MFadIlSRQu67K+NiQJmrnZHun0PNPzh8jrW5jte6QuIIvAyNSYjQqMpx7MkauDpMkczbBiUG2w1a9589YGf/XWayyv7rJ4Yod87MLc7DuenGhQwoFskLZAGgE1NCHuJGkrnNW4UBNEggotI6dQukMu4rLgtCBVO4jRFDuNpzcfsKFEJiX4AVeu3uVPv7PJbHaLxI0IRuJczVRX0snnSZMCwXD/qd3a2hoHmgS0jMMf7xxa6dh+8J6mauKQMwiyJInmITyJluAdgkj9CyFQtwYpAoI4cFWJAhGj69JMjdkxIQ4VVUQXqI/gc/fs5ns7rrzI9ge8nW5JVY/i/VwEgBV5jlaRhFlXI5I0LoBFnhGCheAJY926FHEQuWtH+ABplqGTBOX8fhTecDSi2+mSZRmj0QjvPZ1OB9MwHv76vwEP00rQ6Qic74NNKJKExkdInZAh9uUDCB2/trWP+ZxNTZ6mpEnKVK+LFxadgpIJNrg4Y7CWNM0our1o4JCOwWB33DIxkeOTpMgA5UQXrXIOLizQ6WgYuQgfEz+JBQwxXSSSFRDYNjpi80IyM6FALJDmXXY21kFremWBNw4lBanKkEGSZx2adkiWJzjrCCYQU+oFo6YiUR6MQZPgBLTeoNPYRk11DCmRHqQMWAup7mBaR2N28eiIjNAOhaTabvn2N77D3Xu36dd9dqpdhv0aZzy2HUtR8Rw5coxf+OKX2dnZQWAwvsEMW5aXV7ly5QpvvPEGt+/cpq7qeEIToIQkS0rmZ2Z59JFHqEtNZ2KBekdw436fjbX7ZFnGEycP0e0Fbty5yO27V9nZHTI7c4ilIyd5d7fm9dpwywh2nWMukTxTJFyQKfNBcSuFl0zFB62l7mYcf+YJpp48zqW7H5AMKr7285/n55//HEUQbO5u8o1Xv8dLt99lstfh7OQib33nhzy89pBOMc35Fz7D7ijn5gf3mSwUH3v8KFevvsrV6xeZKEo+9ekvsLIGN26vsHCgw5nTBbZd/6ll9R+EQ1UISS9bAhxCeeqqTxO28K4lmIqm6WPaAU29g3MVE52SRKfkmaDTLRCqjIM6HTG2Efu79wFtkLR+wOVLl3ju4xeYLLvMTypau8YwbdhuU+RIQyMINsG5lG4+QT30bGxfoz94wOb2XaqQ8N2Xb3Jrsw8uox716ZQlwnuGA8fMZEaewUBGfrgQOiYOZYYk65HnEpkoglURXiY8JkQJX6pAAdY7NBWuaRg2Fp3mBNlg/S7KlTTDCT68sc3jh1tyvwVGQzAMR4pZ7NhaLxAyukRHo4putxcxAxK8F2idjpUxKc5bpFDUdY3WGUkiWV9f48jSYZQUkYUf/NggBMFFOZsElIquzHHAPT5EGmWWR1bNnuRRjjM692SIe0aqPWlibA4FBoM+c3NzMaFHpKQqxcvIYrGtRScpzhqG/T7dTo+yzMauTwlhzOwPcWAbQqBtxsoYFwmhUgiKceAEENU6IYZ7aBlPI9ZGaabf07wDxtWoMUlT+WQ8p7AgIbiAaWvSPA7KJQlKOLqdBGcCE50J8jSh8nFY2rQjGm8ITqKsYjCqKBYn4mKZCJKQYsctkKmJCZACLTWmNWgpmOpN0MsVWVqQZS16CMMKbOMjtz7E95MEslwwM5XSKRU6CUhdcPjwCap6HZVE57AMkZsf21YS03q8VTTG0jQ1iUgRMs5DsrxAInBYvIxgPhEUpoE8y9CqZNDskucFOMdo2KCzBOEUMkiyJEcAioCSEuc971+9wq071xFAJqJyRyHRQaJVQkgCdx7cIik09VrD5sMV1pfvcvmdd7n8wYfs7OzsK7+klCAFM5MznDp1iqeffprjx48jpWR5u+Fb336RjYcVMkiOHzvEgYUO926+y1uXrtDYmonpac4++Ql80Fy5+4B2Y4cyeCZkxpPecr7scDIIkIpLwfBKtcP1AGJ6iqf/s2foLC1w6doHZAi++rlf5HMXngcCF1du8O03X+L27Vs8MneAA8kUL3/jh6zev8f8zCKffP5L3HvouHV3hTz1PP3kcd59+0Vu3bpGZ3KBT3/6c9x/sMvt+1scP3GUxQMpr7z451x4/P8H4LAQAg8e3gRqqtE6ZenpdCSKgHaSLJV0coGcLGhqyHuTtCFFaxmt7TLyTrwzxBjkmFgfAgQPmdBMJDlhZNi6v0lR5oiRZJIcMz3HbruEut0wWQsou4ycZG1lne3tZXb7t0iyXZ44N8/k3GH+4uIt7mxmBFJM2zIzNYFrRvQ3HInMKToeGqKjVXqETjCqZX3Lc+zRQ3yw7XG1xtsIvGrbiiAtZVfByBGsZ2Z6ht3dPqYyzEwKkrAbN2RNQ42hP5xi/mDGwd4EqShQEnpzFqTfD8qIxMxoec/yFOcMqUpiu8sF0iQjBEVrDQCJzhBjBjwikJcFVWsJzlOWJSLUpGnC9mAXrRKcVzHQJHjwAqGiFt4aQ5Z60jTBGgfBU2QR5WvMT1oxEHu1Pvi4u1ZxaLu/c3bgxqHoIQicA51IlJZMTHYZDndBlCgVKLPe2N0YryzNaIcjWjekdRWjumFnZ5csLTHWYZQlLxxVPUIrzWhg0cpH9YpzZIkmiHjaETIiFQgeawJtbdA6x1poXUuZ5ZRpDjjCmKSJ8yA9KEGnLKnrAU7GIA4RPK42oFLKMicvOtR2hFYp3jiEEsgQYWLeWfCSkW3IkpQ00+hx8HiqQJV5TIYKFisFtQpRyqtAB8fEVE5RRqNZ2wZSrTlx9Aw3b7xFpnK8cWyubdCbmUTIuAC2xiIljIYVCIELJmazpil5onDBE6TGG0k1qMlFRpJo6qqBoPdlidYHpEj28Q91NaKjUnQQlFbihGBydppjjx7nzoPr4AVN0ExPdZmZ6LJ0YIHDCwvMTZeoxPDmX3yTV968zDvvvM9ga5vGGZz+iRGvKAqOHDnChY9f4MmnnmJhYWHfu9A0Db2RZq4zxdJjh/DOsLJygx+8/yGV3aIspzl/9ucoOzPcvXebh2v3qH1g7sAC08KSrd7j+dkZHhtaTBC8pBwvtjWbKiVfmOSZX/w5BtLw9uWLTKcdfuUrX+XsmbMMneHq3Sv8Pxdf4tZwm8ceOc7MmuGlP/4mm0PL/PQJPvWpL/DB1TXWtwVL83McO5zw8stfZ3NrlfnFk5y/8Hlu3tlidX3EI6cOMT8reel7X8c3Q0T4yXvpb7v+QRR38BTdh2gdZVOTE106e72/EAguFg4RBNZ76iBJ8gQRDInwOCkIKkHYgGbMKGAcByMFqZV0gubUwWPkNqVwJdWqpRoepTn4GKE8Rj6xRTC3WN68z+bDbXa375PmmywuWU49OsPSkR7WDXj0ZIcP7iuqocE2jm4HpHRsbQ4Y9S150SDqlERJvGswXuJpeO+DdZ4+fwheX0VIHXfjVlHXUe8+MZkSHhhQkqnJSUarm5G/LgYcPdTl3vKAJvRp6mU2V3r0d+DQtEaikTIQsHgfOTIQe6NCaFpjSVONkAFEDFOOZEaNswGpJHzEpbr8YJm5+QVu3brLG29cZG5ujrOPnWV2KiMvUzrdDqPaMaoMQqYkQpCnCutqRAgkKiVNNEoCSiCExzuLEDEcGthvI+wxcIKP/oFBfxBPEaHFWsfMzMw+72VUNyAVvvWsb24wMzMdj986i87ZaGfGecfO7g5F2UGEQNEtqRtLrzeBFCnKuHEweAy3KMsC24JpK4wxmNYiy2Ic1xjhdQmK4FUMJ29djI8LjiItyVOFCgKDJ9fZWLgLWV4iAgwG2zSmxkkTFTI2MF1OMCRQuYoiSRCJxFMjgkALTZZkEBTbg11AohW44Mg6OVoqbGvGJz9FrhQ20dGgRGxlISydskCnkUcffHx9VJsNme6R6S6maUh1weKhozS2oWprrDI0TT0Gu6XkWUY9qhi1hjxL8SbQjGqSTBFQqCSGZ9t6hGka2qaN7ouJBG+ij8Abi1YJ3aJAivHC5SVeC4SGz/78p5lPJcG0TM0fYLE3STaqkZs7DFcesH1xmfvbG6wIxZXhiDVjcDLm2SZSMjMTd+nPP/88Z86coex0kCoGrgwGA7a3t3nw4AE//vGbmNbTDAPvvneZph3Qmyx54olPsnjwMZbvDrlxfRkfPAcXjjM1lXLvxjUu3/yACTfkaDJPt9fh8qjiFdPyMCk4evAwH/vsea4N73NveZnH5pf49S/+ExYPHqJ2hlc/vMgPLr1If1Dx3MmzDG9v8K1v/QBGFceOn+XJJ77C5UvrbA/g6CM9Ds8rXv7+N9gZbnP05Gkee+xTvPfOHepW8/gTj6PVQ37413+MbyoUeTxt/ZTrH0ZxF5DlFUpqtEgZ9odkqoxRXSpuv+N7zWNCHFbqFmSoEcGAyqIschzpJoQnhPGwMgg2HqyxeuMhZeiSGk1mcgKB3brk/rtbPHZ+iYc777J27xJDM0Tblt7kDoePwqnTU/QmIMuG9KTk3LkeL70TuF/tYtuY7C4yQ209G2s7TE4Aaw4tPdi4QwvBsXrPkH3CIHGYYEC2BNshhC6DUUO3mxN8BSrBGUsik2jgshnPnD/J40+s872/fpNmWHKgt8bxY2fo9VoSCpQMiGSHEBKyLPImlFI0TbT/K633Wwx7vBfnLMY6nDdj446iaWs2N7cRQmOM58KFZxkORtS148HyJlMzPSanJxjVfZQOYz5MStPWONfEdoDwtKZFI3GuZSItcMHirQURX257v0vEHUgQcXBXFB3KMka5hRD2j9vW2kgXzDJcCMzMxGg2AGMsaZLgx61hiBF4PkCedOhkPXZFRZ6mEKL9X8io7dcKxDgPIPLeIx3zo2qZPWWHlimWgAgxmEIh6HW6SGFilqgDKVQEsqlxqpPSDAY7JImiql1sHVlPpgShbfDe4p3Ae0HbVKRk6CRDO0maQJLkSOKJBeUouxO0jWHQ34nzpbyDczGQRhMIY52+FIAx+OCphxWjUcXWxjoq76FkRvAR1pdmKc57RlVsR6EdMlHRoxGgweCFABKCUFRNQ0KCN4EgXWS9j2ceSo3nOkJimxZvo0w2zVOEsCRC4IMiCCJVVMTssM9+/Fm2X36V+t4a4YOb3N8dIIc1WeMpgkCLEE9BhSaoSAwtOiUnjh/n4+cv8NRTT40zWHtRBSWi63p1dZW3336b1157jdu3b1FVO5GCmpbknUkee+Ic07MHWN1wXHrrJm3TsrS4wIGD0zxYvsnbr3+ftj9AK9iRmu9ub7J6uOA92TJMJKefPM3T5z/Gezc/YLW/wXOnz/OPXvgsh+fn2DVDXn7nLf7yzZdwEj6xdJb7L17hnTfex5DysWef49D8cS69e5vhUHHuyZMUxQo/+N4fMxpYHj/3KY6ceJJLF9/HNpZz559gNLrH6y9/B1Nvk2YFWd4h/0jIzN92/cMo7gFMlRG0QuDZ2V5jZqIXDRxWQLDjdosjtUMknnrUJ/ghvU7UF+tiCmNjnzsWGeLOPUC9M2CwsU13sotyEjuwlN2SXq9P7+Fl6suvIu0dOpMNeT1BkTuWjs5w6KhiZlaRJRrfOrAjjh/t0psYIlZcJN6h8TrQOuIwZjJDtBYhWkRhUSJGsNUDia36YBtcMKAsghTnPNtb1b4rlBDNLjQGESRKOt55+20+9/lP8Otfy3jv7bt8+Zee5fDRnLLwpKKkGvYZeUtda3Si95/TCAXzf6MVYowZW/k91pqxecjTtrGd0+10UTKhNzPDndt3mZ6eYXt7l4Nz0wyHDcZustPfjSEqCGTc+CMSRZqkWBtQSbSbS52Q5XmM2hunRO1hBGLARzwtIKJLtqpakqQes3Di7/VRXICxFqUSpNTjvNZA00SFS920BCFwQeA89AcVEkc9arGto20MzoXxY3jyXJNnGq0EwUewmfcxiF0rhWn3pJKBeLixTJQdWtXinUWFLCpYnKdtDN5L2rah71t6nRSVKzwNQguCUAQvKDsZwYIRoKQkSQryNKdxFcKCkwbpJUnepdApSsU4vzTTBOnodafZ3a7YureM3V5jcnaWxgWEEtTOUznP9vYIFRxZgFRLvHEYHxi1LdOHuqhCobXEOEddD6MzlIAgJlU1bQNWkCkRAzlCJKGOmnj6EjJBAt5GkJ7zFlPX6AS81AilccbStIY0S/Eu4JoaUWQ45XECrAInQAXQVcPgg5u0N27HBSpNGUjJai4ZOsvIG9Z8y6ax5IcP89XnPsWjR45y5PBhetPT+5sVKSVVVXHt+nVe+/FrXL58ma2trY/A8qLaaH76ED/3c/+Yq9fWePfyDk3bcOjwJEePddncuMmbb3yPnZ1dWjEiEQFlE2YOLHHq7BmmJibofXCFjz9+nOzRab5190eIOvCFJ57nK899jqxMebC+/v9S917Pdp3nmefvSyvsdCLOQU5EIkAwR9mULGlst1OXW3ZP25Q1bU+Xp+Zu/ghfzsX8Ca5217Rkyy57Su1ujSTTFEWRYo4ACIIgwgFwcthxpS/MxbcJya4p3Uwoed0cHOSzzt7ver/3fZ7fw48+epWffPIO2hgeWT7Htdc/5Ppb1zB0+OLTX8S0D/HGW7fxasQjTx6D+jNeffFlbN3w1LPPM7twkQ/evYPwKb/0zFlurb3HRx++AU2DkYqZ2QX+9e//Aara/bll9RejuAvASIIArVSEQAmLcCAtYPcwchJNTU2BYA6pc5pxiag9svaYkGBMjkeCzwk+wTbxxSrTWWR7k+3xGnPJfmbSkqK+QjATOrMrVGGP3qwm66SokJOYhv0HuywsaHRSgQxgAt4JDvVaLLYtl2VKqWKRliEQQpfdepPF9j4yC4WZkPgMIWtSoxhahRUdeuYOZdNG+BzpCpzdw9eLjPoKi8bYmBxfNA1tDzZJubWX85d/e5V//8fn+F9+7QLtNGBtEQtitUlZThBS0kkbJs3kPkelruv7rJTBYBCNHiHQarXuyxON0VOJ4hT9a+KY6PatG4zGRUxt8p6NjXW6My0aG+ffBIW1DluP8KFm39Is7XabzY1dVlc3yPKcqpqgpGKYRmfqaDhEScni4uL9RauQNTdu3CaEwMLCIkVRAkSevBL3cQhaRwRwFSzOedZWN5iZ6TE7O8fm5g7W1WR5m7X1TRaXlrHWMdjts7mxQZ5HjPRkPGFmvkvAceLkEYyROBcAw95un8Fgl2Zqu7+PCRYCncSHofclygist1grkSrBW0mQGhcCSZ6TKokyEqE8eZJSW0fw0Om0SIyk0RKpBaoSNLYh+BpfVLjSoloGGQJVU9LOM7IEhBeEYFEyQLB4gNpSb29SGUnIuzTDAUElhNrS3x2ivaPrG8gUKutSjidYpfFegJdxF2TA2moaPl7icVjrmPRL8qxNWY+QMr6O4j7LI0SgEYokzaNO3tXUZU09qklmOgQP4+GYVp4RpCcERVVYMp1hK0c1Pd0FNBKJVdA6chi1sMDmp7fYFY5xM2YgPQUeG6AJLtYFJzm7sJ9//zv/ho3+HjozaKNobM1wOOTmzRu88/Y7fHjpI8bj8fT1XAPRbb20tJ8L5x/m6NGz3L2zx97GGjN5h+MX9zMptnj3rbfY2V4jhPj9SlROez7n4vmHefixx0nznKaqOHrqGFt5wz98/BbOBn73ua/yG48/j/GBa9u3+D/f/iHX7n5Gt93l7PIpPvhvb3Pr9ie0Ox2e+8KvUhUzPWm6hQAAIABJREFUfHBpBZ0Ennj0JOt33uPjSx8gQsZzX/5dvFrk3XdvoILiuWfOc+3Tf+Ta9Q+oy5osyZibW+CFF/6ICw8/zAdv/vDnltVfjOKOIJAihMQ7T6q7BJdAKHBsk5gKX5d4b8AmBJWhsxmkj4lCUEbokB8BFUUxoqolVeMJEixDzjxymM8ufcbt1cv8ytnTqOweSla0e6PpaAI2N8YEN8GkOY0NTApDS2lSnUR2dw25gl4au71JUaIlSBw+KIaTMQ8ePAV2E6sVyjbIEJeKNVBZRSuR6InCNgJXj/BhlyxZZjTYIZqqBVm7zWC9pLGTeHbFkbYCUg8py4CfeJyvpjGEHm08IEjTwGC3vH9XrbUgBEVRsLuzQ6vVwrnIyjly5DDGRCVMNB856qbk2PHDDIcTQmjozXS5dOkypx44zdGDBzGpZjga0ZtZ5Ac/+DFSapRqyFqaiw+fJcsy6spy+eotZuYWGA/7ZGmL/cuLuOCYm51jeXnpPp1xbW2Npm6YnZlne2f7p+OWMGWjTx9Mny/NvA/UTWAwGLBvX/x7+v0BSgkW9y2xsbVD3mqjteHWnTucOHSMwd4OSSI5fuwo3ls+/fQ6WiXM9DpU1QR8wNqGxcX5aPqRn3sPfnrF8ZWjGo4gxASqROSkWRdnEoSSaOeoG4tWAesFWpl4uqChrmqyrE1wFRriA8RLlJRIB74WtJMZ6rokbecIAY2tkAqcL6lrT2IMVVUQhKCqKnw5odzZoHV0lqaqMC2Nl3D06GGapmZ84xplLel05hiPx9DqYoPAKBhXE8pmiKwF3hPHSgKawuJ2S6oOyBQUgqIsQE6loVKQJxkTN4nxgs5jQkqW5iQkEVvcaNpJmxoLSpLmHSbDMUoryqrEC2iUIGuisqXMUgbHF/jJexMmQYEEGQKpg45QLEjFggShEpL1e4TJLqIFW8NdBjd3eeedd7h69SpbW1v3X++fJ251u11OnDjBuXPnOH/+IVp5j6YOzPQ2mO21uXVrhffefYntnfVpQxR3DXOzPS5cOM+zzz7D/Pz8fTPgYHePenfA1asfc6DV47d+9as8euY8XgSurN7ge6+9yO2tu+yfW+RQZ5m3f/BDNlZWmOnt5wvP/gYbG4G7dzfodjMuXJjn2pU3uX71A+bmlvilL/4rVrcrbtz4hE6rw6MXT/H++z/k9u3LOO9JTItDh4/wjW98g9OnT5NlWeQf/ZzrF6S4g0ktmQIdGuYfmKOpBpGa53ukecqkGCB8Rpp1wUAx6aOdIniPdSUyLdFOkKk2rTylykwMPeh02emP0TKw7+A+Vm59wkQ4Du67gyt2WEpCHCGInPmewagOSdYjy+MbM3gX5XTWIn2DsyVaBWQI2MJSjSqELxHU7O0JZufbpMk9KhShqRDe4RqL9IGirElNhkLRuBpfTRC6QDCmKrcRtAlBk+UZKjMII3DlHvOtPr/162c5edQi3QAlWqACLjiYSiqjssNjbRyvCCGnenUYj0ccOnSQXq9HmqbcuHGDdrtFXRV451DTBVWSJGysrbK4uA+tF9jY3OaRRy9GRnox4syRBxDCT5epklNnzlGXeywt9WhnGVIpDh44wNGjQ7Rp4WpHtzXDZ9ducP7cKU4/cAyTJAxHI370yqu02m0efPACCwuapaXFOGYpq9it+8jq8d7fDxSJ5iNPbR0Oze76Dv29XbrdeKIYjAYs7ltiOChYX91gsDlgPB5y+PB+jJGkaYeZXgfnQElot1NGownY6dKXME0Bm4KmA3jnGW710YnB4lHCI0WKC5KqaSI1UokoU08UtqlRQqO1j/eXgEkMHo93gaqqCRLM5/jlIOkmLVRQTCYOISzCB6o6BpMEwFWe8WTCelincZHdUlRD/NaE7MBJ0rzDaGcLObePUDXkWcpeWeGFjgErAfAOHxo+u3mFnc3NOFZrJFmaERoBStDJelgVdwreRbyC1AqjYpBL7SqMiolPUUuvcI1jJu/EUPJE0U4TAqClpihKsgzaM23KuqaddgmAlwIVAOcRWvDA4xdJX/wBc3ueFopZJZnXihkEXeFRvqLvYWPQ541XX+FeKnnlxz9hc2UtQv1+JjdACMGBAwd45plneOihh5ifn4+/HjTWekKoSVLPndUrvP3em/ebiBAC7XabJ554gmeffYYDB/bH8JTpuGcwGHDl4yu889pPWB/t8cIffZ0vnXuEPV/x5o0rvPjmS2xvrXF0cT9LySyv/92LDHa2mNs/x5eff4GPPxyxvTNicbHD+QtzvPX633P3zj1mFw7x1V/9PT7+dJ3VtQ0W5lqcOrXIK699m821u9N0NMPZs2d54YUXOHbsGHmeT3MKPD/v+gUp7gI7nsGGilaa0dQFPrQorMfojO3RGOUCxpQESqSrSNyQxCt6rYzGD2jEEBVa1JMKKxRp3mNSjfGpxwg/VbLMg/ZUbhXvt0l1gVAKLTOchSTTJIkCU09layoGEXhwTcQdDMclQSjyzODrJtqy/RjnMna2E4IsSNMSVSeo4DECSudx1lM3DqWTmMOpNUEJWu2cPE9pGocMGUoYJsN6uhYs6OUlv/PrD/LUE/Mk6W0Suohp1oQIKgZziGgeiomCnuCjZnk0mqC0YGF+jtmZHnmeU9eRey6B3CRY6bA+IFBUpeXoocMYrbm+soZJ8/hQE4GqCQyHIw4s7+P9Dy7R2II791aw9ZhTJw6RRN0Lwgiaasxn128SrKN17jQnjxzgoXMn8CGwt7PLux9+hE5anDx9gdp6XBiitaJpLFppQoiF9nM9fFnG04iQAm080mh+8MPXKSY1Fy+c5fKb7+Fqw/nzp7hy+Q0S00IbyQeXr3P4yBLaZFz66BKLC4txbJEIQqjx1pImkXYppcRoEbtjH5n4UVYqSNMeaIlKHbZukEqTZgkoh7cxHrCpm7hTgOmcuiaIhtpFXLTA0DQVQhiyNI943ODxHqoqKmVUSgypQTCejPGVR+sUbxXFZEI93mBSjLDeMRkVSGEZ3vmMxQuPYNfuEsZDRDpDPfY0LhAaT2U9JDlKp1y+/GOur3yI7Vs0GmVS0JHX4rylLGsWlxdACsajAm0Ce8WAqvBIF/DO06gK0zVY68iSnMY7+mUf0xh0I1Amwv5UqkF6vI8hL0IpfIingcwLai2oRCARki8fe4i77UPIvTUSJBqQwSKCY6wUqzpjr3FsTTxXvv1fGWjF3niAkj/d4bTbbU6cOMFjjz3G+QsXWFiIITKfs5Mm45rbKyt88MF7vP/+u2xurZNkGu9iJvD58+f5yle+wgMPPEC3253iSErKsuTjjz/m5Zdf5tKly5RVRVDw3e/8PRfOnWVVFXznnZfpD7Y4f+wkYWvMq9//EaO7I44ev8CFZ36F995eZ7Qn2X+ow/HTgh//+Fusr9xl/8EzPPf8b/PRJ3e4t7bHkQNLHD/S4aUf/i39/l1CkCTJPE88+Sj/9t/+G44cOXK/0Xn77bdYW/3/hi3z/+4VPAl92rpksHaLpp6wsP8gVnrSNMOFEdoXmMQxqfoYNRNBWbqgcQN8cGidoVSOz6IxJoSSTjbDaHcQ5/hVtHy32paZniANEi0yhEoILkGSIVQWXZu4mByDAhT4iLotbIvSS9ozs5i8YlLXSCVp5QJbWAZ9gw1jkqREVV1CU2GriqAtwcPe3oDe7Bx+a4CUCcq3UJnAiWw6lkqQwKS/xmIvcO7oAZ5/5gALvQ7KlEAOtg2mQgaFCJrYuQeEcNMAigKIRb4oJqRpSpqmbG1t3dcDR5WKmOJqJThPUUV3ZrvVZn19jaqqUEkal6UiLjjLsiRdmkdrw+zcLHuDPkcOLdDupBgjqF2NkIaZ2TY6SVmcm8dT8+CZE0jl2dsdcm9tnYsXL9JYz717d1hfX+W5p59gOBySZXmUTAYXs2oFU5wwmCTB2RggQpC42lEWMdbOOSjqBi8l99bXmZ9bZn6hAwJa7Rb79y9zYHmexCTs27eESaaZu15OF8Kf5/X6+wOZ++wZASpPyfOUsh4gpEQoQW0LJrUlM11oBJ6AqyoIktrV0wwA8F4hSJFOUIdAlimsi6c5lMQ3Ueqpg8Q5i/ZQ1ZZq4glFQOYK5z3BSZrJhHfffIk51aUcDzESuHOde85HgudoQNKZZ1Q6xk1DNzcMdrbZunuL22u3qE3FoBhRFZ7EBHJhsZMoQy2bCoPEijgjX5iZZzDcI1iHCwLTNrQ7OYiGwhfsjftYW6O0odPpoaSJDZGGVEgKNyFL2zS2ZlI16CQjSVtRfYSgFuCit40DvXn2z81xc/ceExUoQ2DPVez5hoFTjJscG6D2EEqLyQ2pMgRpmZ+f55FHHuGJJ57g0KFDtNrt+8ykz8Njrl27xhtvvM1HH11iOOxH3X9TkyQZDz10nkcffZQLFy4wOztLq9UihGio++ijS7z66o95//33GY1GcV9gDE56Gi159Z03GLYCksD5Q6cY3lrj6msfEsaacw9/geOnHuXtt25TFSWnzxxhcV/gxX/8a0a7Wxw9cpZHnvwK71++wc7ukLNnH2CuLfnef/02tR0ggiLP2zz11NP8u//+9zlwcAmlFOPxmJ/85Cd8+9vf5uK5fwEmJiFgtjei44fo1g7JrCDJtzEhI6BJTY6oM3zhMK6L0Iu0FjJcs0ph+yjRY1K0SI0jbzX4pmKvP0YKyd5Oxdq92zQu5fDxoxw5Eui2SpSNZoyqCngbrfZpYohncUlAAToqbqjxPjCYGLYHNXPLB8Bs4yUIA61MM5g0lGWb0ajP8vIMW31HQJEag8kMoU4YDocs7juO/WwPKRNoMtBQOINFoExDKi3Hjxi++NQJjixnpK3dyLVHAF2QPYLeQqERPgcRjUsKj1SCquoDGSEExpMJaZqR5xkER5bFn5+ZiVhaS+wStdYkU9BYCIGqLKeda4jSPteQJFFJ4n0gzzOefPIpdvtjZrqSLE/wvoHg0MawtLxAt7dAf2eXUX+HNIuhE1evXWNpaX8MmJaOyWCXhx48S6fT5eOPP+bYsaMkSY7Sauoy/hxopfGuQUkJWiNxsUgIiURMZ60aFzyNs1RNTRBRw1+VRaRPYhkOBtN76VlYnAHiIvlz5o34HOT2z6SQwjmq8STm6PoABppQU7sGlECrlKqeILxCWMGknMQHblAkyTRu0Id4ahPQ2ChTDI3D1g4lFXiDVVFFZF3AVoFUJzgC6ECqUlLT4vKHrxLGjuMo8tlZUmuo735KNY0o3P34YzYrz9bKZ+QtxWA0YlBMEIuzVFZST6r4fSJBI+M+wTlyY6IjGdCpxuiEXjJPp9fFCUeQAa0itM2VFje22GpC3m6RziuapsQ5G4UMzlK7krr0dLpdpKhRwmN0TG2qFKQVaAu18nBohs1Ti7xzo6YpHI2TVDqh0K2pE1ZicVjhsK4EE3jw3GnOnD7L+QfPxz1OqxVn0FIQvGd7e5sPP/yQN996i9srKxSTAmddjJEUitS0OXvqIf7n/+lPyVoZzkei6mg0YmVlhZdeepG33nor7iuYRj36wPLCfs4+9hCnz55grpOzOt5icd8C7778BrcufUrWpJy7+DRzR87x6kefYIuaCw8t0+1u8+I/vkhdwJlTv8yD5x7l3Q9vMKod5y+cItUF//iDv6epJiA0Rvf49V/7DX7zt3+DbjfKfoui4NVXX+Wb3/wmo9GIEP6FFPf5GQhjQXv5AE4rhoMS5ROq2pEYT5orQi1JmAc0IkzwYYzzoKUh1aBlhS9il6najt3tPcrBHgy2EEFQb+fsm08xcspadwqjU4JUuGn2o5YG4yVOSLyMx/JQNdixZdjMMHu4y8W5GX6jqrmzssHMgZzWKqR7Da6E119d4/lf/jI3Vz5kz88Q9iakc1BIyZ07E86ekqS6xtAlGAPs4f0CStaYesDyAc9v/+ZjzLcrcj3EyZomAMFASJHaIuUcwQlQCVprwFI3Y8p6ROMqIAcC49GY7kyPJElJExN13NOIOx88ksD0K7y/aLXOorTChypK3US0i/vpn3PWMhyO2B0PGYxKJiPFvtk2QcVTgHeB7a0ttrcHBOfZN98jSQ3XP/0UIWPRLiZjbF3x4JkHmJmdoyrryLXPErq9Ft43SEyULzYNWZqSZxkheKyz7A1KvK8JvsY2BVp6SluihCdPNQKLUWC0YDQesLp6h6oYoZWmri1pmrK8fz7KRIUkBDvdT0TJpfcBKX66yA0+joKsl1TDuNRURuKDRGYJu9t9ZBBoE/k8oZTUVUyzSmc1aSujaEZTOV6UFgYlom5eCVxwlD6eFrQwZFlCXY0weUKQ8VQhTcRDu9JRuSEDn1A0LdI6AvXU1HpPucv6jVtMRhNCk1I3FY0I5KliuLWLF9BOEtpJlEA6BToxBF/jgo9h6AqCdFhZIXKJD5YkyXFNoJGQJC1O7D/JXjkkeEtTNlh8/H8oTeVrApK69BSqQEmBDIHRYBAdtc7iIscOJSTWpCyce4jtH75GIyJKWkhNr9tl//4DjIdDVu6sUFc1dV0hheRP//g/0E4yApDlGQjBZDJhZeUmlz/6kPff/5B765uMbUMjBPNeRTzGXMqD587xxAMPM+MzWls7zJ4/xvWNdUZru7z3k7d5+fXX2Rps/5MYym63y4XzF3ju2V+i1evQNBXBexZKz3tvvsu9Dz9DhTYPP/vfkbWXee+Dz/Cu5sEHF7DFKq+8/gbWKi4+/CX2H3yYn7x7laIsee7Zc1TFGq+//A/YpgQpMUbxta/9Lr/5W79FkhqM1uzs7PDiiy/yne985z4D6eevU39BinsIAYcnS1M6eY9hWeO1wg0rElHg6oqgW4wmDXVVsrTvGO3uPkylmUxuUpTX6eazSNmCuofWHYalx9mKbjuwcEKT5ODCNr0FSyIU2HbsrkhROiEoQ10GJtUYJQxai8hOcZKyluyMNaGzRK8zT2t5jgNHjxO8w9mGRx9d5/bNMf09Sdrps29fieY2YXwTrTLmFjKGmzVbGy3e0wPU1jZpqnBlyWC4w3f+98vs3V2nqPZhZ9N4pA+BsgQrM7yIVL0865KYHCkTjI543aKIc8HRpB/Z5hoqwn0NeDf02NraYjIeYm2UFHY6bY4eOYIghnsMxxM8gp2dnfusFSlEtJBLCT7EQA+lmEwmDAYD3v/wOnlnjtFAcebEMVqJRkhNWTnu3l1lMCxJlObYoV/CWZifX2R+X4crV66At5w/e4a5mR5BwOrODvv27YuLLSw+WFzjsE1EAXjb4K2KxmMqhGiYn2vRbiVIUXPk8D4kLTqp5OnHHmIyrpHacvHiBe6t3iIEz7kHz6KE5NatO9R15M1EeWUM4NAqJRAoiyJCf38meq+oK4zzNJVjNC4xps1M2iVVAaykGji0MrTnW1hvabe7DOoB3U6bRBucq/E4lNRIoWl8jTKaYjgh0/FUpeL8C5kpimqMSHwc59EQrMW7BF9bpIsLzbEMjKsyFmmVxJMWMcMXFWhsQ2UVtfXIToKjwRiBaWcoL6cpSjVpZmJ+KjUqEyRZdMRKGe6P+2wdk5LSVsQR+zpC44KJUL+ysVjlkVOqqXcu+j9k9B3YqkJWltl2L+Y0+GlTERlfyCB5/PGn+P7JF6fhGwfuq1wWFhb5q7/8NteufxpNyB6Kccm1a9f5wjNPs7G5yc7eLvfW1njjjTd47+NLjKrYpbe0JgfmkMztP8zjFy/w8JF9mP4Oq++9ycrtVbY/OMWX/8c/4vXXX+EnP36DO3fWKHUExnnvmZub58knn+ALX/gC+/fvx7lwP9VsZWWFl1/6EVcuX6HbmueXf+V32RsbPvroBt7VXDx/jO3Na1y78gGJznjq6S/SmjnGW+98gBCBZ5+5yMbaVT547xVEiFhqk6S88MILfOlLX6LbbVNVFZubm/zd3/0dL7300lRYEOJpO/1/iPz9/+MSQiCMpBgPKKsBOm0xWN9ksdVCJRUmy9gbFhBiyEHWm6E9d5LcHkMNatbXbuBDh6KK1L1QS3a2FcJ1mO3ukakhJq3o726T4vDjOfrbhrkDApkoCAZJQp5rsgxGTcBWQ1wxoiFlq5ljnB2gTuapGoMWgtR4jBQINE8+dpwnHw94n+PlgHv3hoRqnbBXk8zN86++/Gv8zX99n93Bfj77eIJqPmX2wC6b/RH1Xsb17QZvWggxg5Q5Wdqh0zZkKgGtCCKuKz/PFy3rglKUhGmmqPUVrXZCliu6nYzhXpxbV1WFFILJeMzhw4dRUnH33r0oOfWxI5yU48hlCYIsy9nc3GJmpkeztoPQhvFoTKIFShrm5+bZWL8XY8oOHaQ3s8TWxj0GgwndPEMqiXPQbs9y+OgSu1vbeCT9/ggtDf3hEAE8dOEhZjotgrfs7vZZXV1naWkeCPjpm6cpXeSkh5+eKjqdFjoBgeXk8cPMzi2wvbnG8r4jHFg+yurqCvuPH6KuYXXtFgePnkPpANRx7IJCIBmPx5RVhS9cDI9QcUE4Ho9pmhpnY4hF8FGNYGuPSj0qSBbnF1AiiV16IhmPK5aXlijKiqydUduCxBjSfB6dGBLTIgQBFdRNyWC8G3G/Y7CVQGUKgqX2NVJpWtpg65rGVSgTOUn1qGLSDNnXWyLNBZ1eD19ZBoOS2aaDMhYJWBsVOUqqaRauxQuJ6hlqUdOaTXBAXdbsjrZJdEruc8phgckkSkYDYFHVJDrOrGUQJNpM6ZcVOkkRStCfDNHdjFaaIssCEww6jZJBoxTWeZQSsdALyaRf0FL59A3/U7Gpn97jY8eO8Wd/9mfMzMzc3wk1TYNznvPnz9Pr9ajrml6vh1KKy598zPmHL/DKm6/z4QcfsHL7NraxqEbSkikTVSEyxelTR/nSo49xcO4w1c1V+t/9MfW9FaowoQol1z9peOd//d+4ur1O4QLNlFPVbmdcvHCBL33pSxw7diwiO5RmNBpRFAWvv/46//AP/8Du7i7OenIjWL27x/pONC4+8tgR7t29xLXLV0lMziOPfwEvenzw4U1MInn44n4+vfoGN29cRgSPtTAz0+UPX3iB5557jlardR9n/K1vfYvXXnsN7/395fFXv/pVDu3r/Ny6+gtR3AF8CNSugFBTuor+YBddVRw6NofzEmcly/sO0h8Hdvur1FahwxzFRKPFUariAHVQIB1lOcSFhG5vHzKsIaVFKYd0Bm1buHKG1VsjZvYlaK0oioBtGrrdBKRHJwlKtmhCYDRJGIslXOcBrOyipEILEZEIIhYgjcHLvch4EYKl+UUuXjjDSzufMvGa7tw8D118mNfe62DrwEPnc6R6i1s3t6idJogMgiJJQeu4PIzMFENwFS6MiRxyF5e9OsQOScRs0xAiiVEpR6eTA1G6WU3GtNI0IpM9+OCZjEuyrIvzinExjrZ7JFVVcGB5mes3Vjh+4jhSCAa7OzFjVSlOH9vH7c9uMioqXNDIEBChoWkcn9y4yey+BRJt2NjZZmdviMo6FI3l7sYmvXaKciXS1zz16EMRR2sdw+GEz26skKQdNjZ2kcpQjG3U70/drJ93z0IKdlc3cDaqOnIjmWlldA4dQSooyj4zsz22dnaROmF2337u3l1ld2fAvn0zXL9+l6a0lEWFI9Dvl9ORi2AyGTIc7BJ8TZYY8A4tBXXdxJm7j0v2SVkwPzeLVAItofCTKWgsZXY+p/ElysBoEkOfkzyn3ZYYmaBEtOY722BLiwsR1mbrCkJMbxIKismYxjagPHVZYUQaIXM6wSQSYTxlsFSuImkso8kEZQyKSFnEezKTUvvYUTuT0DSeYC2Nc3gi0KsKgTxPkULGQGjRkMiEZhxoQoGcNUCUe/pQIbWgqSw2NPhgIY289GFZIRF4YNAfxvxTmcZMVQK2rgERv+6m+b9973++6D948GCsBd7fB8gRHOfPn+dP/uRPWF5eZm5uLrKJjOK1N37Cf/n7v6cuS5SQMWIkBObaOU+feZBfevwiB4xm49JlVr/3In6zT2oTGqG4pXM+bmrWfZ/mXp9Wb4bQeBKhuXjyNF/66i9z6vSpfxL03u/vceXKVb73ve9x+/ZtiqKYig08ldtjUq2xb2kfh48c5aOPfsCdlU9J0wM88fSX6A8c99bu0e3Oc/HiET5477vcvHkNgcNIw/7Dx3nhD/+ARx97mCSJCJE7d+7wF3/xF3z44Yd47zHG0O12+drXvsYXv/hF3vuXYGJy3nPn9oiFrA0yYWt3hFMLjBpN6eYYjws8KWXlaecZ4/Eeg9FlyonGh4p+v8bbLRpvOXp8gaUDGWnVYq59hmL9FsL3qOsxe5vQSTqURUUVGpTuEbzg3somETSmEKrGJIY6KPpNlz5z0D5J7WfxQkS8bXAgiMAyNNpJghTgUrzXCAUnTx/mpTeuMvQ11z67zCPnH+L1S7eomOHaPc+R/cep6s9wQoBoIYLG1iVKzaC1IAiHl4IgKiCySgLuviXe+8guqStPkmQRoeoq0jQDGqxtorFLCEyiuLd6l53dPtYHujMzjMZj8Pb+aWA8HmOMYf/+/dy8eYutrS16vR7GGIbDIe+9/z4nH3ggUhKlpGkqhv1tjHD0tzdYvXODxcVF7q3coNvOGPZ3yPM23jmufXoN7SuOnzjM1tYmTWPZ2d6LXJi8jfXxlDEZjwnTh9hoVBCQ9x2yxkRujpSaXi9nZeUOVz++QgiCffsWaOWGIATGGBprWd9Yp7GSdiel281JtGQ0GAOO7sxsJBWWMaAjz9ocO3YM2xSs3ruLc5HHE1lFHmcDyhKdqlQ0Ltrwy2JMVXpaWUrja5ybZtj6QJpG3X/dVCQdg1KgSFgMszRZTTUOGJ1GXHCSYPKMwpUE6zFovBBUgxh0kaVttGoQMpBmOYOdUdwPpJJBUURImVRxOWqjcsYjGJWWdruH8gkESzEckbVzlEmQBtI0p5W1YpCKimMYhGBQjtBSRNlfcAglaWyDbSyutlNSZo6fBuQILWnqino0ifjkTNNKcia2wmQtBjumDmKMAAAgAElEQVQD2u3uz6ypI3mUwNQ8xH1+/ucfPzevSSlZWlria1/72v08gLquscExnoz4/uwcOxsbYD0Hl5d54qlHuHjyBO29IYO3LnHr408w4wltbekLuOpKbjSeVQl9BELW9NBkdcOBY0d49Kln+OKTz9FemInL2Wlq2M2bN/n+97/PpUtX7ktz9TSn4MSJ4zz19GPsP3wCLwzf/+53ubdyl0TO8NQv/yo7/YaV1R2WFma5+OBhXvnhf2Ft/RqRTh44evQoL7zwx5w5cyrSVK3lxo0b/Pmf/zmfffbZ/a+71WrxjW98g8cee4xWq8U/N9v98+sXorgTJHduCDZ8n/asIZ89TEgESiVsDioEBu8sZX9Enpf4csze+p1oBsnmUAqybkGiZ+jvrLPv0AxGapwUtOaPUPevUgxLtrcLlpcUg2KTxcMZHsveTsXqvV3OnDkZpU7eEyZjStFh185QZ0ep/QyBBCHraVAFBDG1naMR3uJcQ1MFiqKhP9rmytVLKNvBMs8n767y/IUv0uYqu7VkJHPu3ckQvodiEgt2MOAtiQGYdkehgWAhyKlxRCGCAARS2chDcYIkMXHpJwx+GjMUGTIi5pdqT2IUVV1HzknwVFXBeDik3W6Tpindbpe9vT2KsiLPc44dO3Y/mLrb6ZBqxbgoSNOMbrdNv284sH+Z44ePsLl9jyQRFINtDi7OsLD/EFeu3UDqhMXFOVLhMKJBKUmrlQIRAtY0ju2dAc5ZOt02c/NzJEm0jW9v32J2dpa6rrl9+y7GJJw7d4Z2O8Kx2u02RjvG4wmCQLfTJgBBKBrnaY0zJmVFK085dHABowV3fc3+/cdAxbzaVqvFcDhmMinwvoUxhr29Peq6IZkyerRWdHsZVV3Q6WQYI5FBxdeBi2lCOlHxIeRslINqidSGPM2obc2k3CUxREdvAkhPLjNCLUnn50FKGuGoPZHSKUEqg84y8iwl1wbra1xhmYiCTjtFGY1XBXvDCa0yx0xTk5zzMYUrUQyrkm4roZ23cXVDvtBCpoqiqrEu0DSekRkTgkBJTSIEmTZoI/BVSbAVKjUIrRiPRriiwFlAahIlSZI0ulsbga8tmW6RJQnt9iyViPz5dquL9knU8YufUSHFd9HP/PBnPmc6qp0W+4jKiN+PMM0JqG3N6ZOnePKRR1n57AYXz57j3MlTpDdusPPX32W0sYVyFqcta0nFeilZsY4VFShzSdd6TjrBomzTlSlpb4Hf+dM/heUFpIUwFSTcunWLN998k7fffpuiLKObI8Qd1MLCAl947gs88sgjEXJIQ1HXdDtd9s2c4cJDT3N9bYudvYKlAwc5c2yOH7/0t2xvrCBDVHo9eP4cX3/hf+CBB86hlGA8HnDz5k3++tvf5tNPP0VPC/vC4gJff+HrPP7443Q6HfzUhfvzrl+I4i6lYv7A46TJHTK1h6hq5uZS6sk2iQCHIO8kSCnQomQ02KRlHOMGer0u1o/RKOY7kqEo2NmyyCSneyRBqicYb7/HeLCDzCRl8ORdQ7eXUAxr9rZ3WF5KyXNLsJ7GWqxs0xfzDMIype3FYAo/wlpFIxS28TRNzWBUsLvbZzSc0LiKdr5J3mqjVOTMe6GRWNbuwnZ5BW32UG4GaQyDyYSYR5MiAwRn6PQUX3n+AbKmQId41HUhyhd+Oief6tOdwFUOYxK0AGujnM82cRFYVTVKKTqdDvOzLUKwtLtdpDYgJEJIXNMwHA6ZmZkB4gMB4hurrmu01tEUpRSN9/FYX0b1w/kHT3Pn9gqDfovjRw8hZMxt9V5w/cZnLC3MsLmzF8MElaCVt2l3IsY5eI/WcuomXObexjZChpgfW1k++eQqZSnodqPjttVqU5YVd+7c5eTJA1H9k2ZoFRfHghgiwrTb+1za2mnnhNAgRDQSHVhepK4dlYuRh4E4DinLAjt16lrr4iK6/fl82CNMhStKUpGh0CihkZkkb2zsqhNNE0oG/THBRMplu90hVQl2OtqpmgYhHSoIggNCIGsZ8k6O8w7hBS2VUowmMFVztXoprTzF15ZcdrHWMmEadVhXBC0Za0G/KDAmIYjApKkjjz9NGdiaacIKrbkOeStDZZK9wQiFZlyNInQvSQjNGCksnVaLunaU1pOlGkLA1g2hdlGpU3hMbjCpwfsYDemsi9F3aSuGeRhNbSsqW6EqhUwin19LQ+zaY8cZfqbACxEQLtyPQASi18ToKX6ZqTwXhJIoclqZ4T/8wTdY+egdPvnxy9x46zXy7W1S66mFZl0JPnENNwnUXkVRQqg55DzHVMIBLUlDRV+UrE02+ezaxzx+9FfZqwr2hnu88sorvPrqq+zu7k1ZQ9Bg6c33ePTRx3j++ecjlto59oZDmrJg5fZt1tZ2OHvhl1jZrNnaqDm+PMuZQ11+9NL/wVZ/jWCi8fDRxx7jha+/wOHDRxAyUEwmXPnoEn/1V3/J3Tt30SJ6UJYOHuAPv/F1HnvkUXKTIH3g+rVP2d3Z+bl19RekuEsuXDhHVdVI50kaTz2Y4IxAKtjaHWKSDnlqGO5uUxU1IuQoaZibPYiUEybjMWm6zPpWyaQuOXa8i7H3GO6tYetNFhbnyAxo7cjyXpS8NRM6nR7ORY1wWZYEIbCmx1jux3KcSdlFeIv2E0pKiqKi3+8zHo0JBLK8xeLyIr1eD611tJRLydPPXODd996haUompWZrd5u5Oc2drZqinBDKUcQSywzpM4I0tLqaUw8eo/YNZWHxhAi2DEQGjIhdO1Ug+DRaqp2lLC0xlxKa+8W9RCk1lT82hOBIUkNtI+sGov66aRrG4zF5HhOKZAhMJpN/kpQUwtRCHwIiOEb9Ae3ccPbMKbZ39n4GhQt17VlcmI3qm611Duxfom5q0rkun+eXBqYjADytdkaSGopijPeOqiqYneuxuTFkc3PjfnfSaqVoLdnc3CTPI+r08yN8XVV4l01NXSLG/8kYSlKWFa6JoSBZklM0MbQ5IEBEE5P3jqau0DLuYKx10yIT72VRTjCJxoiEumpIs8g0SnOD8466rBmOhgz6I9rzcyRosiTF1g3WOhCxC8y0IVU5MqQ4lVBLRWI0RTmJOAFhCWpa3TyITFBSRTMUljRLyYNBI8kTRYGlaXvWVncRAYzWFGVFkJ7ZNMMSmFucpXYWm1qaLNBQUYUCJSTeWDKdIrykGBTILGr0tUnRMkEoKOtmGrkFxuQxa9dkeA/jSUGQcRGbyQQXoAwNVTmg9iXCBsajIShNUA5vsukkQfwUq3w/PWtavLlf7qe/KxBEuN/Jf/4btHaEpGBxPuGD99+nefcSohgyEW1uB8EnfsSadDQOUidZEBXLynNS58y7QHAVlahYNYL1ACvliNvf+28sPHSR99+/xMsvvczq6ur9nY+1lna7zbNPPsazzzzDkaNH4z6qmZJV+yPeef01fvTqq4zGgvXVCUsHTnHm9BGWZhK+94O/pZjs4HAIoXj22af5/d/7PQ4ePEiSGMajEa+//hO+/a2/ZHt7OzZq1nL69Gn+3R+9wKkzZ8izjHJScPnDj/jWf/4mD5xY+rl19ReiuAO0dM322l2M2CRJclo5NMozKSvylibPDMJ7MmWYOEmnO0sv7zEaDzGJwVSBjZ0d6iA4ffwksy2BGF8lqa5yYBlcmZEYRVNPYpGrAkkWjSujSUl/OCGZLjIHQtJePoJQp8nDLEY35EkR5595Ou0yLQFHYxucEz8FFgWPd44nq4v8x//4FlVd03jP9lrNweWcjz4e4PARY6wyhBcgcoSGB84dZvbAPCIZonyY0iantu/GRtb49HMlDTMz8zEndCpR3NjYZDSpIASqqrrP3PDeYYymGI3xQFTISpRSkYE9/XcinMvf57l8Pm9MkgQXPFpF7VpTV0zGI/I8j0f86ThByGhSKYoJS0tLLC7M4WyNVgKlZFwIh9idRTqgp64r6rqMH6uKXq9Lp9tiNKoZjyuMifTIsio5dOgExgT29gaEIMjzLBZvV8V7bx0IRd3E5dfBg8fY29mlKhoSIynLGms9lqilBjmd52ta7RbeRn5NVVWE4KfZABKtFFhBWdQ4AaGuoPEYrREe+rsj6romz9ux25wC3fp7e4zLgsykFEU8WdACk+dIC7nJ2B30sc2YrbUNvIBUJZjUoNOMpJUwqWuM06RaQeNpfI2WCTU1tW9oUkcZatjdi1p3F9BJhPi2UoWZMahgojhgNET4Bls3NNIhtUQlCaIUUDlUlpHkClREbtgAQsfTjPMerxWmY0gSjbKSlkiZ2Ia6bhDtFiZLSISnLAuaSU0n7aBMRByPqor0/n3/p5f4vIGXnxf6aWc/LeSCn45pxBQLIfBUusG3FOfOPsj6j9/ljiq4Wzm20eykBoRgv/WcERmLSaBNIPceS+BuqFltKm6EhF0ChdIkwzF/8Z/+Ezc+vUFVTR/MRAPTuXPn+MpXvsLZc6cRUpJlGbaxTEZjLn90iVd/8I98eus6ZbAErRmO1zgk51nszfPSD79DYydY4cmSjOeff57f+te/zcLiAoTA3u4ub7z+Ot/8z99kNBjGXYNSPHTxIi98/QWOHjtGkqWMR2PeeP11/uZv/oadrW1OPbD8c2vqL0ZxDwFl/y/m3uzXsrQ88/x9wxr3dKaYh5wjB8gkExISnIAZPDDZCFy4bcpWudVWtVrdF5a6VV0q1R9QVy3VjUuy3C2V25iiXUDjoWzJUGCgjJmSIU0mEBmRMUecIc45e1rTN/XFt87JyAQSl7ouWFJoh/Zee+191l7r/d73eZ/3eRasDhTNssFbh5AKaysQinKQ402HCh4VLGkWKCYaErh56wrjyYjZ9Dq1Nxw9tUYiG1zlUX5JwR5WQjHZwJiKJEso8igk1HQNo9GQwcqEuu7i0EwAJTra9ianHngMPVhDakDUQB8IsYA7nA51PmYY9Jmu95677g6sHymZLuaxwXe94cwjOUXSUNnYT2iDBhQyyRhOEt7xrjdy30P3kJdxgjA4Bz7BuSgC1rZtnMbszaaKokBrTZ7nCCG4zzm+983vcevqc7RdF83BnaFpNc4HOhtIEhWz7OAIQJrpXhypoessUmiyVOFdzEiVUFjjGGcpSkQhrY21o5w6dZL9/X2kt6g+YxdK46zgwpVtFkayvn4EfMT+nW3wOo0ywTZKJjgnqBYtTRM/lxDQ0PPONadObUS+vYmLiRIWqVLSdMStWze57/4VLrx4iTwpaayKDA0RqFvLhcublNkKpu6olh02SzC2IWrF5NHQhUCqctrK8IMfvMDKZIyx0NQWvAAfKZFFNkCkiiwb4kP0VDWd6Y2cDYN8Be+nDMoyDr4pwe50xmJRoUN0BbOtQI0KsrwgSIHUAecs9F6tWhY0y45p2zEcSgbrkmpakQ4GDIqSolQgFcI2dKaOC29d0xlLtrbO5uVNNIJMalbyEa97+mme/e7fMe/mOCfBp3SLBhEcSSrRec54OCZNC3bbKUePbtB2Fc2ywWWSQZr21FHQSpKNC2rTELxg2cwok5LOOtq6pSiHiFThhI3aPMbjbcCJqCmfDkpUnuKbhpfhLnduQsR+UZ+cx1I0Pi/CK3eVILIIf6aW0RvexPc/81c8uzNnJjvGznCfl9wjc06nkjUhEZ1iV8L3teG8r7npAj4taYWgGA554twDPPzgg5RJitureOHGdZRSHDt2jLe97W08+eST5FlGnmdYY1jM5ly7do3P/s3f8NxzzzNfVDgVF44ykTzx0IO87tx9fPurX0Q1u3RCIRLJL/zCu/nA+9/PaDJBScXu9g6f+9x/5rOf/WzUEyKg04TXPvYYv/Vbv8WJEydQQlAvljzzzDN8/OMfZ29/H6UVP+FMHm4/G8GdgDQtwrRoHKnUtNWcxbxCZasY1zLf30O7hiPDjDyT7O1P6cKU4aiga/aZjDV2e49RSBgohxIOb2OGm6Q5IQiqes7qak6SB/ysJRsoBis5pvNkZUEIEik1x4TkyvZV5ltf5vh9T+HVOsZrhDSH3Od45fVj8rJ3pOmHSEIIjCcDTp9e44UXbxOU4tb1itc/vc4oFzRzhSRBkICUOBzjyYBHzp3BmJpqvybYBOk8CIXUOWmSko9GJEmEfnQiiailQMiY0SRAkmcx++wz7tu3d7h27RonT55k48h63E/pvkmVYjrL+RfOc2TjGErpl8yrD6hoBLTSSKJi40F5+q1vfYeTJ44zHg2RUjKfLbh+c5O6aVhfW+PCD3/Ia1/zSG95FtibLZBa45xnuj8nBEFZDjAm6mdrLakWSwZ5SvCOarlE6RytJKZXrwzeIYJBI+jqJW01w3cNL16+wc72NaQClMRawe3tbTh3D2trA6bTPS5ducLK6hrHjh5l0EsOW2twrqPI0p6bHb9LtaxfijABtFBRqsJG5ykpM5RWVIslRTYkSXKCChRlRmMakIJqViFsXBiLNGFjfZ12WYHyTEYD6kUHOnKifZBMNlYockfTi8ZlRUnnTFScTCVVX2954UEEqtmSarrAisBrH/k5LoXz3Lr8IoPJGk/80nt58uffw8IsubD7Q0zVMR6WpPmArq0Ivc9qvayYz2uEjwqVaT6gXbR4a3Ayxdsu8tW1QieaNE1ZTGtwsHAVSknwjv3be9y+vcvK+gTjOpSXOO9ougZnHYOBxAbDynjUoyp3NFIPIsBBMA93QDY9Jn/g3HX4nIjGILlPwHf4u05y7OffyLOfvM7dSnI6TbjXZqx2HoNlU1umSnGxqbkYHDtlRpsWjPMhD509wlve9EZOHjvBsBxGFUaRsPz8Z7nv/vt58sknOXnyJMPhkBACpmnY3d3li1/6Il/9+69ye/c2AYHKUgSSc2dO8fOPPspg2XLhc59D726zlmbcTjTvev/7+cB73sPKcAxCsru7y6f+4yf54he/GDF9KcjyjKeeeoqPfvSjrK6uopRivjflG1//On/8sY8xW8wJUtA528tT/+TtZyO4B3BNi206irxAhMgsWF3dwDBCKrBNRTubUdU1t3ctTh0nHw6BlJXxkIvf/wduXVhw5bs3eM8HT1MOPHUjUGrAsJwwn9cY26HTDBdals2c8eoYmQSE8yAlqYrWdknnEYstrJLsXM/YOPs2tFzBMQPh+85+LyoWIgc60ItN4QlCkmSKM2ePodIXkE6yt9chRUeWRO0a0EipkQKsN+R5ILiKeinJRjlZOqBMUmQa8IooWCUCUliE9ASvCOGlUvXwVIoIsVgbKWt5njMej2nb9kcgGClfWhicc1RV3fcMDurkuK/SMsrA9gHfOcfrX/969vf36JqK4AMXLlxgsrpGXhSoYHnrm97AdDbHC4VIEpaNgxAz5wh7EOmCIkI2VbVkMio5MPj23nPt6lVOnDjOrevXydKEjfUJibTYxlFoydG1FRYnjyEs3HPPqejiJ0GqnIceOMdwKCiLMbe2djl16iSzeR1VFk2HQkblQaFwIiB6uGg4HL7UWO6ZEdYFvLMYGTNTrTRSJ4Q0IR2l+K5jMMg4MIbwDibDVRZmDrnAS0GZ5awkOa10zBZTcJK2aSA4ms4ilabIS06uHKXuWoIMdPUS42pmaYUQkkE5QHpPcAHTWEQHKlN0IvDR/+lfc+3KZUajFdZPbDBvWpZtw7JqUFKDlngbjTe6tkMrz9LVGGcp0pLV4iTeGIKNJi7eBOp5jROecqWks4bZ/pJgYTQe470lTTISleMHktoYRADXtmTpAKGTOByXajpnUERU6pX3PfR0yIOQ/gpMXvwYGCdAXExNIDMWCnjPO59i+OVvUFzbwhuLF3AjUVyXnmuuZSs1hEyBVZzUAx565FEefeJxjt27htKSPC3I8yFeKB49ssGDr38dhMD169fJ8xxjDPt7e3zjq3/PV77yFba2tvsZh5QQBMPRCm9/y1M8fPw4N7/2DW58+7uM6gZVFCxqzwfe/37e9SsfZDIuUC6ws73Lpz/1Kb78pS8Rel0btOKd7343v/qrv8rq2lqvorrL33/xy/zpJz6B9dF7ASl48NwDh3MBP2n7mQjuAU+z3MW7FRo3oXUNWZmAl+zv7DGfbTFIJ7gwYHdZ0WlFUpymto56uks9X9DVCilTmsZw5fyMM2fW+dLnr/PD7+/wi+97HavHOorSkqUW15ZU04SV9QwPUbfDSoSU8aIqHOMyQZiK5Y3nqcsT6LWon0LwCB9tu3yQuCDQQuH6zEMEAT7qsdx1+hSJaLBizLTt2N2xHF2rubKnMMqShwrvxgRtKYYVx8+OOHZsDaE9QjoQ7SE+HnHG2DAMHgQtSgiE0BAUIfQiZ33mE8JL+ijeB4bDYe9qdGCtFwN8mqYMh0O6riNJkl4FMqC1PMTH27ahMd2h8Nj169cpioL1tVXm0z22trZiVrdYcO36DR576AFwlmtXLnP05CnKNMM7S6IVbdMCsbR2vRzuYaUuBOPJhK6tqKqajSNHMcZw+swZmqoiS1OGg4xmueTuu86QZylnz5wmT4cUhSLIKFxmnYxBh0X8e4jqksPBiOAju0McCK7JWJ0slktWi5y8x/EP1ssQAjYIBJI01bRVhVTRozXJcuq2oqtbAlEdsO0sw3KVIBQyTZEqaqnkStFKy6JeEpwlOIX1HUkiaa0jlYpMglUeioTFch/TdCipqedTkiShTDOs9+gAg8EIoRKMsjRuRjYacezM/ZiuYe/2bbrZLhKL1B60oGnj9w6JIpMFTYjuS23XEKxk1+9SFgPy8YiWJuK+QtO2FctlhW0sicjJBz30JGBZVWCjZR9KEoKmKAuciVOpAYsXikRlSB0ro0gJuDM7j1u0gpR9PIgqrAfRgVfsSwhoB0EImlxCornvxFmqJ97AD3a/wGY95RodNwRM0TSiwCeB4aDkkQcf4umfe5p77703Ij+JPNSPybKoURMAnR/BG0tXV2xfu861a1f5/Bc+z5Wbl2iNwIUERcracMRrH3mQJx5/jPbii/zwTz+NvD0jd5p9kXK5qdi2kuf/4QU++NEhno4bO7f4Dx/7JM9861u01kRjkcGYX3zPL/O+X/kAk9EYnGe+t8/f/NVf85/+8i9jUzd4pJI89sQT/MZHf5Mr3//2q8bVn4ngDoK9bo2qFRQakB2NCQzTASmSXFlsDe0yTteVOUgxw8hAXsAgz8GssHVji8FowKWLc7AFN69NuX6140t/+x3e9s4V7rpvDeFyrPEs5jPSZAWBRUqBlyEOVkhwQlEOB3QLg2z32bv6LMfL4zA6iQ8CHzzWRE9XIRRdZ2malqpaMp/Pmc/ndK3D+QatPG3rMBhu7yy56/Q6375aYa04HNgheAbDPHpseo84TGkOLuqDasHFxYjDBOela//wLaGnJL7EgS3L4tBs+mDz3ked66qiqWvSNIqNJanG2pYsyw79VZ3zlMMBUkRt7OFwSJ7nNG2LlJIsyzhz5gzT+YK8KEnzEiETBoMhiVIIa0ilwDoXVRd7Ma5Y6cQbO89zbt28yemTx9AqQgDee+pqydrKCazp0ElKWpRcu3mRU6dO0HSWRdWwN52TDzZASlrrCShubG2zdfMy999/H3VlWVnZ4Nq1W4BgePo4ljic05lAazxrG0e4uXmT9cla9Ja9A/8VOkESZWSzrEQqiTEdQiRU9TIadGhFs6wYDsYQwLqaLJeUWU7b1ohE0NQ1bdfQ1QYpep2WIDCdQSWB2nfYxSIqXRqPEhmDrEQrh0olwjusb0l0NIKWSqGCZ7bc4ub15+iWhmo6xZmW5dZNqq1NJBBk/Fucd1gRGI4KtEvYX+xgmihT7GSHTROCcCRKRVX1XNFZQ2g9uHgcb+Nv6KzDd4JhnoNqcdKjZIRX6qpBCpisTJCZxnuHTCTS+582d8NBrh7Ej2buBxWq9x6PwClQQaJ9gGLAmbe9mb/8uy/xYuPYxrFIE5xUFGnJ44+9hve99z2cOnXq0OKRvupVSh0yyw41hQR4rdnZ3+Uv/uzPuXLpMm3bIlxAaU02SHn8oXO8+w2Po2ZTrv75Z2gvX2etc1RFynO24oJrWIRAiWDnuW/x1U98gjf9wjv4sz/9f/na178eEywpmYzHfPjDH+Z1r3+CIoum6PvTKZ/61Kf48he/GMXmvEdqzeuffAO//Tv/jLX1dW5dfO5Vz+TPRHAPCJwume9co15OEaGiKBJknuPsDlk2xdGRrUKSNLiuYjlXlMNVykHCYrnHeGPMibuOsbe5hV16blyrOX36DIvlLVZXLKZW/PDZhjS33HPfOmU+IBEl0ul4I3sXJ9IIeCnJhyNstU2hLM3yMvMrXyO5750Yn9H1xhWz/dvs3t5me2eX5XLBcDjiyJEjbGxsMBwM0Wqd/+uPPk1dC7x33Lo544nXnSGR2zRWYG0giA4IDAeDKJ0bIvs9RhYVmRd95A69RG8IcViHIGJzCdlPrhJ7Av0F6noZ2CRJ2Nvf4+ixI3dk5LH5C6B1pEQmSRpV92SEF4RwROsx3XPoLePxmPF4zPb2NvP5jHP334fzjps3b/EPzz3P23/+nexs77CsWoajFRIpUcGiAjQmTsQefL+ouqdomw4lA5cuX2ZjfQVnYxWxWCxYX11lc3OTploicFy9Kbh8Y5PaOParKZcv3WC6V0Oq0InAekfXBS5eukY3rwnuGuVwgEoGCCGYzudkewN0b8yRJAU7W1NMV3NkYwPhBcuui56hvZ5PVS8i1CY03iogQlpt07BcRm9fH+JN7wGpJZ1pESK6EolcM60XNKZBepBC95O3Fusd9DRV2zhEa1grBohREbXmUWxkOa007M53MbZFFwpC6I09YLm8zd9+4WOMwwQz71ju7+Bn+wwKxXhtxJ5vqUOEk2wwLDvBSGVoJxmInLIcIzMNaUwcWucIBoTWWB+VKofZgMYYfM/N1SisCCzqJSQOLwODwRBBQA81UgvqqqJQ8boWEvIshx7q+mnbS0H9Jcz9zscgo8m2dpLEB4xKGTx0D+LRh9n8ZksjPeUm128AACAASURBVOPVCXedPs07n34bjz/2KKPR6PC6P5iA1UmClPKQHXanYNzWzjZ//bnP8oMXL0AvopeLgpMnj/Kut76JByZDrv6XL7H3/R8yaEDi2Uw8L3RzXsAxSzUj7zktAmek49nPfIZvP/MdvnXrBsBh/+pDH/oQTz31VM/Ysmze3uVjH/sY3/j612P1DaAkP/fWp/nIRz7C0Y0NdJLgejbPT9p+JoK7IJCzSeGvslrU4OYMB2OMKyhX16jrjs3dPTRw5OiEZtkyKHJUkdHJjqzULOdLNk6uI+hop4K9zRlH11d48OExk8mArRv7XLq0Q9d6th89ybmHT7JzLTCd75KVKUeOr5IVupdwd+hMIRNJQofwU7rbP2CrTZjZIbUfItOSIpecOnOchx5+OI6Qp8khThsC+CA5eXqD7VszlArcvD7lrW+RTMqOxTKPvj2iRQrN+vpar8InDoOelLEjHnoDa5DRZUlIQpCEIDDWHpbOs9mcze1NIGrOpFmBEyLa++UDbtyMZtHORT30JEki/o2gKLPe8DpOddquBWIAW19fRwBNtaStNetrqwxyjVsf0zY1N29tMpmscO7ee2iXM04fX0dKSLVAi0CRabSSjOht43orQESEOoy1mKDZvn2U85e3aOoG0Bw9ssGgyGhqybUbO1zbnOG8ZWVljVndcvGbl3HWc2TjGNYFvvWd7/DQww9z69Y2s/19RHBMlzOyMmd7Z/vwxt7a3MI6g3OO4XAUNVkEVHVNmWY4b/DBHbSrEV5QdRVeZaRqGGHEpgWhUDrBmZpERUcj7wNdt6TtWpIyw4mOpm0QIaCIswJpGqu/1nQkUpAXKdYZpI6OWkZ6skyihcQLTeU6GlNhbYuSirbtGGclWZKwNBVBJxgz5+aVizS7FalKKVSKt5AZj9fQ+g7nW7yUOKch1YiQk6cpWmlUquh8R5CW0DmaqqUcj5hMVmLFVjUMByNaYyNpwPdyAc4jpaJIFEWR0bmW0WCE0pKhHSLkgRWmxSdxXuMAEnxlv+gwHtwxnRq38COvyxCVJY2SJDZEZcy1Vd754Q9zfdbw8CMP8a53vJ2TGxuUSYJT9J2xuMm+AXDQlDz4Hoe89s5w5OhRXv/kG7h48SJmXrExmvC2J9/Ckw/cy/Z3v8Xzf/kdRtMFqwYu5Rk/oOKiczRSUdqMc43idAJHhSMPns2u4/mLF6hSHUkUfcb+jne8I/obW8eVK1f45Kc+ybPPPksQsXGaZinvfte7ed/73sfGxgYSwfT2HnVVvWpc/ZkI7hDIdMP6GApR41VH2+7jdHRmX3abrBxNWB1NqOotjG8ZlEcoV44gzRLRaaQXOOM4dc8atjI8cE6z2KvIF1HhcTASnL1ryHze0tQ1mzf2uHLlGsu6Ji0THn/TQ9z34F1472JTT2pUnqF0i6sN2s8p20ucuuuNpEceoQoZXhi0NGAP+Fs9Y8Z5EILBSPDoYw/wnW98gxA8+7sWbyvWVxQ3b2tQGi8bpJRMRiOE8H0Qj9i5teA9KJX1evMea6PhsjX2kLmTplF35eTJEyx35vhwlbptKIoBnbF0NhpCLKuWznjS/oZuGosQmjRTIDwrqxP2dlqOHT12iLW3dcPKZIW2rejyqKvurWNlUtC2gqp1TEZDXvvIQ1y+chmlFGurA4bDgmo5Z5DnlEWGFALn6aWMG7Is62UXHCLJ2KsCt6dzfOgHhVTHE48do8wFw0HGjVu7LFsYZgnz+ZLhqOSuu+7DOsPKeIIQkjTJSVTKeDjk+NGjWNewXC6QSaxOnA19pibI0oTlcom3FoLHWUNZFuhExwDk/WFMCc4jQ4SklDDY1lC3LVqlUU9GpJgWkKCSQOdiFeSMw6lAqmPmqlSGTQqMbSJE4uMgmZCBetmghUQqyQJJ11qkh0E+pkgT2sbRmhatIsPK9vzvIi0QqaZaLChVSlAmlvFY2s6RtTllkSJ9zHatsyA9VbdAjzKkBOkFxncRkuw6MpGQlGWEcoja87rIUBqyRFHXsaILaIbDCUIEVN47aCkFWtK6aDpvmw5jOgKCTEVBrB8N7D8+wMMBc/JHM/3Eg5HQptExLbGKRGieeN1j3Pevz7K2shLNR/r3WzyqX1iA3pxFvOzIB4mVFIKgHDpL+OVf+mW+981vMQ4J73zDU4SdHZ7/9J/jr9+gUIIdqbgiDM91NfMA3iccd3CvTDmGIPOeRqdcsR0X6oodLfAq5/jJk3zwgx+MGXtZopTi2pWr/MdPfIJvf/e7IAWOQFrk/JOPfISff/vbGJax+tzf3eMLn/88uzvXXjWq/tTgLoQ4A/wRcKw/y38QQvi3Qog14BPA3cAl4NdDCHsi/ir/FngfUAG/E0J45tU/Q5KUa+xuCm5s71CUGZRDJuvHSHVKujYi1xmJzmnajHIlIMixIcEGxaVLOxxZOcHq6hhJA53AVTOKImG0zNnbmdJ1LWkJa0c0Td2yrHYpC8Xq2ir3P3w366dWQTZI/dLkRFLkUUu+NcjQUdpbNDvnyVZOo7IjBOlx3qCEPJzECD4QRJzW1GngdY+/hj/Nv0tXdzRGcP78edYnIyTghUbIgJCO8XgYreUQOBfFqoyxWAemc5jOApI0ycjygkGZkCY60tEEHNwAUiustdRVzWAwiq4/3qNkLDuNMYclqJaqb9TGwR2AvCjZ25+xurrCsqpZXdtgsaxwtqEcDLi9fZsj62tUlSfgaduWk6dO0HYN4/GQuqpp2xZjWqrFjPGZ05GT3ba9lAJ0XXfHoyMvCwiKPEuZrBylWtYMc0WeQfA1eZZyz90neP78VURQmK5lMbOsroxZGa+ys3Ob2XSGc44bN26wu7vLuXPnkFKwXC5RqmU4GNM0UTJ2Op+Rpil5njOdTllbW0PJOAhWrq3CAWQESCFJlMJYjxMeZxvaLmpqOyz4aGJijMO4QKoLymyIsxbhPFmaQaJp6wZnAlXbIDBoLdGJRCcZy/l+v8hJ2mAQ1pCpHJRCFzkWT90FZrOOYSFJCoXvpSGUt5ycHMFnK5j5DWQdmC6XIGJma9uW0EqKQU6zrMlQ1HVNPhmC9jhavBBIIfAIMp2SJSnGBTyOMksx1tEZF2V4nSHPC3SZgI+9CKEEJLEqUokmOB8VJNs2VgVSI7U+nLB+tay9jzl3vHbAkXz5pkMcxrMabJBkDjIHMg0UJ1YhCIxzOCRe9MTjoA6D+2F18IrjHgqWKYnxnvXVVf7V//Yv2H3uAl/9879i/tw/MFk2pB6u+Y6vhSVXUtAOznYZ9xrNSQmpqAnasS8yvt92nLeeKstIteeBe87yzl9+H+9+97sPJ1EvXbrEJz7+J3zv2X84bKUVRcF73/9+nn76aUbjMfjA5tYmX/r8F/irv/hLnnrTa18trP6jMncL/K8hhGeEECPgm0KIvwF+B/hcCOHfCCH+JfAvgf8deC/wQP/vKeDf9Y8/cQsEuqTDpDVOKbZuO/zcIgtLqi3V3owyLzh+ckBRrDJZX6Nb7tM2t9DeM1k5QZvcQz7cRTS3MbVDMgQFsvSoYzlNLanrFtM58lLStkuKoeb4qXXWjmZkWegZFJB6SaBDpwLrErI8J7SGrGmZTS8x23qW4enXI+0AfIaXcfzfe0HfZsUTEDLjdY/dzfq6YesaeJVzc0vxyOOK8tsdMzUiQSDlDqoU7O+2eDxIDwqUluhUUgwTJmmJVkkcLBEymjPjIgovDlqs8XMPz2voF5s+WB1cSAdTqEHEzL9IMpqmYTwekyYJdVVz/eZm1NMOAmMdeRppgvPpgrIcMJ7kmK5FqBStJZubtxiNRmR5RpalpIlC9pXMgR57ELJXTYy61M45slSTSB1dn0zHlcsXSZRm/cxxUhX6datjMi6YT3fo2sCgLAk+4sNd1zAaDvj+89/n0UcfZW9vj6KIujBaacbjMVJqvAsURXEorSCEYDabMRqNMMbEaiJNMcaQZtGpSvXQQL1smC8aymGOSALWdhgX9V+0D8hM4oJFZQnWGZbLGUJArqMGS9W1+BAwbQc+xAG9zqFVymx3TioEqkhxChKfMtBDvPVkRYqQgsViiQ2BldGEskxJZULVLPqKwpPJkgceuJfdZMAPuh9Q+4A3hkRqBAHXdtRasZpPepcpRaoke4spy3aJTATjwShOlIYIXzgMra1JXIY1kX7ZtQ7jPOkgo20MeBPdl0TS+wx76JvmwQeUiFWSsQ6F6KUVXh1vvxMe6Z/hR4N7wMQRE0Y9T8Bk/eRziJClEKCV4kDNQRAJE688krzj8w4bqkSZjTQIhFYM1se8cP5Z3A+/Sy6iaffzTcWLiWFfeVY7w11K8toi4yQWGyQ3dMklBS+2HUsBLpdsHF3n0dc/zmsfeYRja0dxLprnXLt2LSpAnn+hjyOeohzwwQ9/mDe88UnWxqtgPVe3rvKpT3+Sr/+XryDNfwMnphDCTeBm//+5EOJ54BTwQeAd/W7/HvgCMbh/EPijEH+dvxdCrAghTvTH+QkfIihdzkgPGGyssSNnZKsKxy4vXJ2SejC3bnJ7usX68Q1YWLrFDqlKcCbl5LHT5KO7Ed0WzhkUAqVjCSiVQ2kdh4DyAmsj48Zaj9Ypk9UJRZkhZL+Kh1jKhV7+VCmF0AneetCBlDnLW9+jGEzQqw/TiBwp6tjg6svAONUHwjlOnhhz7twJbly5AAKuXN3h3ENrTEpFvbuDthYtJpTpGlkmycsyMiNUz1lXvb5GiBLA4o5M5pBPEHooh5ckU5O+UeTv4I1LKUnTNDrGW4fQAiHAWUvbdty6dYtzD5xjNl8ymy8YTya4APNlxaAYs1wumU6nZIkmhAFCeJZ1Q54fZW1tBSEEbdtQ1xVC5JFV4RyJit/FWE/X1YeLTOTXxyoneM+Z06fojGcyHqG9JVUKb+PvMEgz3vzkG9m5PSXLM7a3t8izBGsNW9s7lGXJcDhkMpmwu7tL27acvesETdvQdYZAvOHTNI0sma5jMBigtY4GEzrhQE75QDBNJ3HoarFsSHUSXa9awbJegFSMBnkP4UiUBBkcMkhUiI06gPligdJJLxYmqJYLhuMhQSa0TUvqFPlA07kOiYg2ezojSSUuGNpqhhaBUZ6TlCO0lixnc2zTRNqtlCTpgCQfUWwc59i9ljXg5vmLmMWURAvKoKg7h5UBnWesTDKcr8majNossbXFYjGqwwlJXTW0tkZoEConUQlaOpbWoBJFEI5BWeBaT1pEOqjvotyxsR3BRqtGlSUorUkygXdgzY9RMXyJc/rykPDK13lFiBd3PPSTrLHJLA/fol72hpc++9D8nJew9zuDOwTwOrKDBIg85dG3vpmr3/w2P3zxRZ4PczaHHmUcJ6zgXLbGWS/JuoDJEq4Fz3Om5YbzmEQik5QH77mLd/3SL3L81EnKckCaZMxmM0II/OEf/iEXL15EhshoKsuSj/z6r/PUz72FcjDA1BW3rlzlM5/5DN965hmk0LieXvFq238V5i6EuBt4AvgqcOyOgH2LCNtADPxX73jbtf65nxjchRA4MySIdWSec+SUpsGwNA0nTh5lsbfLZEWRpBmLVrLzwpxJNqGrGpp2n86c58TpARuTDmUdivhbex8zRykFaaajv2PrkCIhTSO3VakQhyvES8HyzpxBKoVQMjIHAiTtgkHXMb30NcbZKmJ4N9IdSO0KZIgmy9Z0dIs5e9OGu8+eRCeXcF6zqFqcnXLXacOH33eSxfYWF67scc89ipX1PIpaCUcQccAIGcXDBDIGe6I7Tzj8hvSaLX0G32OMSkU8tzH2cPrv4AI2xhxKiUoZm3uTyYSyLLi9u8d0Oou87iTrbb1iZz/VSb9fyerKCnWzRKje/q7XYmnamAFXVc18us/q5KXMOI6Qvnxz1mK6DhGitVmS5kgZ8J1FOKJWSpIgQsCZaF5d5BnBWepqiRCBRGsGg8Hh8NHW1hZVVbE/3UFrhTGWPB/gbOw3WGN6SeS4yJRlibWO2XTKaBClfw98VEMIBCNI0xzvJfPFgrTMMG2L1qCTlNFKhnearjWkStI1LZUzmMYyGA0ZjjO8cKQ64fSp09R1jZOSRCq0kljZ4ruOLEnJ8wxHS9tGnaIyLRAhapJIndC2DmsUUiTM9hYIBecvXQKrKTPF2Ydfx/Vbtxifsuycf45UygizFCkhk4hEMaumIGt0qkn1AOUsWEHVdgil8Aa6FhrTEQYdFAmB2EdS/SxEnkjSpKDTgiAzgov6+Ma1BGdAJBjToWyKVAlKJwxGo8P7/afGmleJFT9uuxNu+Wn7H3ijwkvB/ZXHARDBR6vABDYefojhax7m0vnzbKfR+/URUfJaOWTiAqlUbCvHD9pFnILVCUZLThxZ4Vd/5QOcOXOWlcmE4XBImuXREavrWCwWh9eis47BaMR/95u/wZvf/GaUVtSLBXu3t/nYx/+YF753ASlSDI5kVDBZXXnVc/iPDu5CiCHwSeD3Qgizl01FhhCEeKUCxE893j8H/jnA2uqYqUtZyHWK5Dj4ApxDsUto9lA+o5pVtLZltHqctmu5sb9LU9VoseS+ewqOrSzBLMB5lAhxTEIEkCE2VbxAKkGSqF6QSJCkAoQjBIvgYIgiZgAH2YAUAi/igJPQCmksqq0Q/iaLm9+nODFg3mTUdcvu/pT5sjosSzMNRTnizU+9iT/75DNMpwbroF5YTq1bnn6i5vSRVRb+BEfOBYToYtYhZHQh6nFQ4WNXXwRJzySEPpBb42jajtl0zs7OLjfOXz+cXNVKQdtGfN/7w4BOCNgQIATSLAUEOzs73HXXXWgZpYCVUuze3ukHnyysDhmUA65dvs7RjXXm8yVKczixqlWsk5MkIUkS8ixDCZBCobUizws8ssf2oyQBROhEK4Vdtpz//g/Zmy/Jiox7Tp7m3rvPgFS4EDDG8eyz32M6nzFZmZBojdax7D+yMeT8+Re5ffs2D9x/P2ura9x9991IBTdv3ohNaR2xX5UktHVFlmVxcCtNaduW3b19VldX2JvuM0x1bA7KgJKS8WiASDTN7SWmhcFK0WPaBtMZqqUjyRO8dszbBk1C3Ri0VHgCe3t75PmA1fEKnYkGHNpLhqMJorVM2wVOWgZpQtXUdFKDEHStRwZFRPcDSRAYF6JmvIK8LNCZ5vZ0j7S5wCOveYQLV7bZ3t7k1uZN8iLDBsdqlnLmgcfYlo4XXngO0y4JokNpjS40IpGoQtE6gyKlWxqypKSYjBFeMl+0JJkCJTHGgfcsXUAWsKgbUqXiCL6CZlGDIEo72MCg5/13xlAk2ati7XfEhh957s6ge+DQdUj//QmZ+CuOymFf6o6A/kpmTugzIY9CA5qARWLyAW/+4If4z9/8FtnlazyoSk54Regci1zxgmh5sV6yUyQsVMrq6gZvedMbeeLxRzhz6hRZklJmOXmWRZMd78nznOFwyG//9m/z+7//+/gQ+LVf+zAPP/wwWRah0p3Nbf7g3/8BFy9dxitFEjzH19f4p7/+Iarmv4HkrxAiIQb2j4UQPtU/vXkAtwghTgBb/fPXgTN3vP10/9zLthDCHwB/AHD2zMlgnKdzjuV8n1AbCu1pqm2qxS4iNBjbkA9WoJsTwhKyijyxlBrWxxmF77C9yL4ChBIEJD7QT40KhIvpeQgCgsS6DlRASN+b9YoYNHGHdaGUfWA/gDiSjBA0ifO02y9ybXPKpj2O0ClrG0c4e+YEWV6QZCkoj5SO4ajl5ImUanodGQr2dha89cmznChb1nVHoRPK1B/COgGB930hGQSJSgg2Ugi71jCfL7i9vcvO1g7z+TL6ZqoEnSSELqLu1kYDEWdapEhQUiGIi57Wiq7rYjNXKLwPHD9+LAoxdS2nThyh6zq6LnqPJsmAalGTyISizJkv5gzLAu0kdVdFX8u+Yds2LaYzdE3LbDpldbKCMS6aUqPi+H9fMVhr6ZwlSTwCiTOBohghs4S9qmHWdWjp8cay6BzojLWjR1hWSxpruLG1zZkzZ1gsK1ZXVzh+/DiL5YK8lzUIAdquX3gAHxxNY0mUom0biqLEORunV0dDOmOQwfcSDg5n43DSYDLBGIfyGulaQuvRQqGDRmnNbDqnsCkqic3s5aJB5RkEQSpz6naOsgbvWppmiemTDe8UMlFM8jEDo5nvz+iWHSRRZRcf+fKIQG1aMu1JZILQIfL5hUOKlBOnH+TtT7yZ0WTEpS99lZlxbC/2GCWSHMVqgGZzl7d/+Ne4dv06bb1EqgwpE5RuscFEPrtU4AWJTsiKHKcUzjYkOo68K60xbUsQEoun7QXeTGfpOoNKE5Je4sJbi1aCtooSCGlRUMjsJf76T83e73z9R/PGV2vG/tdsL1ebjFz2WKVoVBCoEAhS4aTkxD3n+M3f+Agv/LtPMNqds13AjVxyfTHnQmKoyxSfl7znl36JN772NZw+doQ0zymLaE5zwKcXQpL3g4YhBB555BF+93d/F+sd995/P8OyZLlccuPqNf74j/5vrl95ESElnQicWF/lX/zT34QXL/BMtfOqf9s/hi0jgP8TeD6E8H/c8dKfAf8M+Df942fueP5/EUL8B2IjdfqqeDsRCNHWkYUG4ePQUc4e6ZpmushxpmO26MhHsD/f5sTxB7i5neHcTTaOKorBPn55HUVCEHk//CF6UabYZEEcoNKC0Cv1KS0idatnmxzqpd/xgJBxH+VQXhJ0EZs1nUK6BVJIjj/2evLRGhaF8wIhHF40dMoRfM3qEcXv/d4H+NrfPsNy1jJcNbzhsTEbw5uIbkqWFBRJShOXJYIXGBto2o5qXjG9vc/29g6L+SI67UjFICsZDsasnzhCkmSonpVw3VzhRf/yFd0HhxIS3bsL+eBQOl7Mps9eu84SQiBPNCuTCV0XRcJicE9ZVEsa27Kolxw7fpTxypimqRhn48NM6GA4pCxLBFFfPBp/tHTGkOXxAnfOHTY1hUxwQmG8Q6UZq2vrdN6xv7fLomoYlQrnYbaYY5zl6InjmFv2sLSu6wZrDZOVIRtHV2HbUZQlUkfxtGNH1tjd3cXbljwvWNQVQUmUiIbs3keoRkgVjUuGA7LkwGIvjo+1LlDNO4blgPFwhB5oFvPbJDKj7Qym8chgGYwVwUpciBldmuf9hK8jEJjXC9quBpmSaoEJNbnMSKVCKcm+8wwHY+Zdcygb4fumuAuezrWkeRKvAwEiJL0h+YR3fuBXMHiS9bv467/4NNdvvUCrDCHJaKxDLOY0S0sQBdZFYE/6QPCCtrXITpJlBcJ7ijLHCqLypbUMR3FeouoWmK7BGEGuUpRXFMMhrWsRSdRIigVvtJ2UXrBc1DS1JZEFrulejpXz8gz75dDKPx4I+GnMm/4uiPdWeOW6Ig7g+jseZVTg9B7pZTxPPZHhzU+9kfwL3+Z73/02X1H7XPYumpokKaePn+KJN76RX/zFdzIZJmgpyLNR74msI+RFgAAyvCSv7ZzjiSeeiJz2Ime2t8/mzZt8/GN/wqVLl0hMIEkVZ8+c5X/8Jx9h+e1nmP/t31K+9fFXPS//mMz9aeC3gWeFEAdiBv+KGNT/HyHE/wBcBn69f+0/EWmQLxCpkP/9T/uAANigmM5aMDWjXGDkgM4NUIOctBWUa8dZilWOHZ2gs4S71yWimnIk30W1S7AFyMgBDv1PGUKIahUHPUgREDL0v6Dv5z1jtk/wEZ456IjG3RFIvEoIqUAEhySghCcNHhM8ynbYzW9C/ga69CROp2SuQQeL8QlpCIjU8+QTEx6blKynnk4VlGlF21bUwUCtaKYJW9OKrc3b7GzvMd1fUtUdOgjGg5LV1QnrxzbI8oQsTdAqQYoIhcQyNcoEJP0E6oHJbp7lcaFTUfmv67o+CAukDLE34QyDwZC9vV1GR47Q1hWXr1xmbXUN6ywnjp+gbipcmrK6scZsMcPYGq1UtFIbDOjaCCl1bXvIb5AyqgNmaQYhRGu4XoAMItXLOkdnPfPFkqau2blwAeMNbVdz5eoG9911huADFy9eZLHY5+KFBQCrKyts3rrJjevXaduK1ZUh43HB7ds3OZmeZL/aZ+vmJolSHDl6lLKMTJljayPqpmGxWHBkfZ2m173XOqPQUeUw10lUohAHJX5HlinytCQAWZGxnO3TNh1SKkbFSrTHcyIOD+kOXSik1CwXC+q2OaTdyf4a9AgWpsGrjkbGgJLqjGrZoLLYWIv+rRohJWmS4hpHU1e01iFVgqla8nzIYFzghacTAidzLl+9RGdqxqMJG2tnWe2WBKHYu71HVVX4oNBSMV8uKVINTuBwGNGRpRmd73Be46ylyDKci+5VwVmEjbo6rTWERU2q4wK2XNYoHW0glYwkBOUlqXTk5QCQJDLpK+cfxcdfGeRfPsT0kzLyO+mSLz/Oj2LvBzTIl1Mr76RcHkCFBFDBYgQEqXopCkdIAun6CbIP/TJ/d/WHbO9XeOFoJxlvferNvPe97+Xo0WOkaYLWur+uIoHh4J/v4dADgoNz7nDmI9UKYw2bN2/wqT/+E269eBWLwmjPa8+e4Xff/yt033iGy1/5AiJMMeL/54RqCOHLLzsbL9/e/WP2D8D//NOO+/JNEHSOSMfIZINZO0O6ltFkHeNGCGdZ7i8QY0ORztDeslpa8BblWryxgO6D3UFPsf+RXyYh+sqP/XF82/7HP0jghYiwjI/WdEIIpAC0JHiBC5757euowRH00SFBruCDJgiJ8AFvUhKZUparWOWwdhchMrxzCAmt0eBSvvAXn2XTTRiNVlhb3eD0iWNkaU6iNMOyjKJPwcahEykQMgXR62F4DyiUBpWmKKV48MEHD70ndb+Pcz7i5yIuaj5Exoroz8V4UCKjMA1nT53uOfTRuX5lNMKYjrLI403eN6u9F1SL5SHu3lQ1y8UiMpSUYm9vDxmikw0yNrWNMVgb4RDrLAhJIOXhB+6lswYvHDZYMhVY7u/T1C0bk2HEvkPofV1HtKeOkiYJSgmUx8572wAAEKVJREFUDP9fe+caa1lS1fHfqtqPc859d0/P0D0Pe3qmAQeVgUED0QiDIyAxYhCNiEEjwWhQiY8YJ3wyfjIxQUyM0Wj84Ac1gFEyRkGBGJSRZlAyzPQM87a7h25m+nUf555z9q6q5Yeqfe45997uuUOgb/ft/U9Oztm16+xde9WqVbVXrQd5nnPj4hxlWeKc446bD9IpO2RZRu1qrIkTXNDooNTEFcnzPE7iyWwU1ZSdLsbWybNAXs7jR1EAZEXO0tISg9E64pVRVTOzb4b+YJ3SZqhGdZYRoT8cYooMr57gXRTiHYMSI206X+OoUQ9+PU4meWkJgxoflEKEqq7IjFDXHieBudl5hqMKn8PaYMTRO1/FfLegCJAbIc9j5EZrOrzi9qPcttCjckKx/wYy6ylsSTUY4D2MRAk+qh2tyfEKwfkY298AJicQqFxNXXvEKbnYWDfAxdV+zLLk4q6VzaP+3dgCHwILS/twlVKJYst8Sj++ITOa4bhNlFOmRPFUfTaVb13tSzPMt5kcZItmqHlzNyaFIYiBeZC0GBEr2G6HO1/3Wo7eczen/usLHDywj7e/88e47833kieB3sSqMSbu7zWCHYjjC8ar9mYSy7IMV1eM1gcc++9jPH3yJC4Do3Dr4SP84nt+mgsPHuPslx/COVhmFp0OF7UFV4WHqoiQzy3Sc4cQX1ANXoAwQotZZBStGc6eOcdg/QxLc5a8WCOMRsx3K4Ib4HxApUj23lvNrZqXvPG3bHzvVEcXmc4i4hABYxRjlcx7itEq/dNPsNCZp5zPcczhvEXdkPU16K+tcv7MN8lX18lywboeEoSsu8DQryNZzmu+7y5e1Tk8zp6kGsPrGhv142g0Y2yyxnsxiFiMteSdnMzG30XvHBBN/qy1WGMQTTG5MzueoCINQpqsTHKggkxi7srUiHG5iMGKRPr6RIOk15bk+KGq7N+3D9dYKRFVbkbiJm7AYm02rZYh5lN1dRQIQQOIw4uPbxwhbiQ7FLUGdT7tjcT2GWOjEEqDhfFkldQDGgW0MZ3x781sH7TGYKO9vYIfO6fGV+dq5LClI6jDFjmOit7iDKwpRV5ibzDkZUaxliNeGGmgyArqah0fHJlmSf8/SvGDHLUTClNijWEUUo7V2rCwOMdyvRr3RMSyljYobVbgNVp+aQjMz83iyoBS8Oojr8QPK7QoqOohVVij6HTIyw533/s2jtxxJ84HTp6KaeOMQFnmyEhZPn+R2V6BLQuKomS9v85wWAExmflwNIiTuHhMlpFpRqgq1Ag2E/KUlnB2ZpbgazyeajQi82U0N+6WrLoBCzfsRySuXDevyCV12HbqlZ2Nzp2pcL6VjdzpcmUUPPl8j3f93M+gsx3ue8fbOHz77VC58WKhiTIZ+W/DGxY2Nn8nJ7ImlHYhgeBL7n3zvTx8/DgnvnGCQ4tL/Oq7f4rh1x7j/LGHsG7E82XB/1xc5TVstT6bxFUh3BXFKVQ+w8osWc9gGVHRx3IaK4a5TkmPnAO5Yb6zTq83gKqPr4aIyUCy6Hyw+eJjHZ8kYZ5mfm3SvaUIhdEkZUsHT7krmww1TchcH3XvmaEYOsLgHGsnH6WeGfDEKc/JMyu8ePoE3zx1lpGzLM4qP/zdGYv7SzI7QCmp3I185nMv8MLK87z6niO88nsLZnqzKDVKCiss0cEFBClL8iyPbxK2i82KuImqjaop6rCbdjcCNyMmAvA+mvdJIkRjNKDEqH8Q0+RNBmhqNsCsAsFHCxiTwv8paLJcsBN69yLPCWh0AvIOa2xS0QjG2hgMSRsPUEFdTSY5vh4mtZjDGgUVrGaxDMUHh6hiRNLACektQjCYaCAaYqQ9QrQxadS36t2YHaKA3zB4jZNQVMvFN/O4gos/o9v+oO5jiwzJoAZWVlbpFSUmg9zmIAGnjm6nR8eVuOBx3lGUJV4d3rm0qRYiD9YKuVIFR+0c6mGm26PyFcNqSJmXDFYHqMQ4QUYNM2VJqD04ZdQf0ZudQSXjmWefpezM4vMezzz5BBKGqARuue0OTi8PWFgfcNPSHDNzRfR6ZkBZFHSy2ZhrNtSM6uhEVo8CtQt0uwXOVRiiaa8YKLMOoYpmk7U6Ot0ZjLFkwTBaW4/b9cbHTdci8kh/fQ0VZbW/DFn5suTCy1l4TQrNl3PNS6l+mrGz5ZrWIFa47c4j/MqvfyhGPVUg+dXAhjCP6tINob65HZOfWBdyYzh04CDve/8v8Kl//iQ//5a3wCOPc/4/vkS3gucyw4PrFzhT5By9fK6Oq0O4C+B9zG8ptsBmgFi8BpyZwVWOoZmLnoFhROVXyXUAviYGd3qJp3yJ+T/K+ihAVKOQ3w5xlWijflWabKRR9ZH7EdXaGR5/9CRfeHiZcv4w++bmuO31N9NZWKTXyalPPcqLZy+wdJNBaovIQY49/DgPfvU5qn96knt+4Cj3f+S3mVvoIKaGLL7KZnmBtTmGjCagWFDBIeMNt6b9ISRBH01txiENNCRbmYlHC5qMPzVOfiYJc6WxHohet5JedUTi/RrBLGOnqkTHZkUSAlmaRYXItMGnXvAh3rPZvFYIGicksSb1QU5yniWkTTAQjMq4a2KZGXdtSKs/hRj2OP037aNN84DEiT22NxZ5BG1WV+kGTXatoevTMT2GwxFFWeJqR5nnrPZXqYfLZBLjqZgshn/IssD6eh/RGNNGiGEAjFEq5yhsjhvWuJHHdHJQoZeVFHnJml/FakaohNx2qdwQW2R08oKgOf1qQNkpKTpdMhVGYcS/fOZTfPWR43jg+WceI9QVtQaW+8scf+RrfP/dd9E1sH9+gdnFJVYvxnj5GTWV1gzrEb4OaD1ATfSDiPl7azQTau/oFAWj2lFqxsL8LOsjjxeDzXJGVR/NosmwryuMyUCIYYMN2Nwg6vHD4fYDcIt+fLMA3v5vk3+d/s+lxvvm8sb88fITyaQ1jVWllrgf0ykKshBVJ7WdniSigDdjPov30anzDWzjcyImhmkwcOTIHfzaBz/Iyn9+kacf/AL5qOIFzfniyjrnujkHaqV7OcJwlQh3BdZWl0GjmqCuk7ON7+HcbVT9Cwxdn7KAi/5FlmaV2seoWjbmeImDccK1p1nMqiZNRhrwOnFusgE62dHjsgYTK9lkb27ispWggskyjK/Rao07Dt3E0ivuZs3fiDU5Ygf4TKgGnmPHnmLluXN0Z+fp7s94/vwZjj9zBlfUVHXNk0+f4OOffICffPc7OXTLfmwR0DxDbUwxpJIhZKixWIkr0BA0Dp7EnNZudOlYv7nJOmBDdzlBE7bRWDZCUZuEGhO1NjQj42tNd+rEsRKdsqbqpetu4Ybplc7WGlsH4bhpCWHiHrqpwnhSmejvieZMXDPpRr1nNByQlTFD0Vq/j9gO9bDGlwEpcuphTQhCaYThqE/ZySlCgdYpsbQPSPAMBp75xTkG/ZrgY5wWrSu6lEgurOuIteEQKpAso64rFheXcMYxO9tl6KAONubaVGFxZo7gAssXv8G5889jTWCx08VqzOX5zdMnOXzrPZQ2p8gsM70OC/v2s3L+FEZsNK8bVdik6jl/YYWiW+JTIvOyyFGnBK9UAXxdIwZqVyN5zurKGj0Lg/4Q2zFIUeAcGCuUWc5oGCNd4h15XtDtzkTKbl4tj/ua6fLpHt4Wjdpj43grJ2+1kGFT/U19rxuhfxvVZDOMTFpKqsT9vIx4bW82q5rMxqbwxDhqELOdmSkVjTHRMk8slKYgdzllp8dzReAFM+L4es2ZskspjjcUOQfyy78JyU5iK3+ncfPBG/V3fuuXoElekVbjo6rGuRG5NeSlIbOQy4iMFURjxMTJuM+XxKYOHI/1ZgYdD/gJ3djE38d8oTFY1ljYKajGDvRKzCqEIZgutRZJ6xyvUI0qLpw9G+3VvcGrMqiifXGjs46ekYH9B/Zz6NaDyXNWxgJp41mFcSIPnW7j+mqfC+eWaZwxICaVn95AapY7Or7mxpVlepHbqGgmJahMfU2T+hLlmy67RSBPXXDbk5e643aYkNibSprS7bml6fuNs7WvWB2tYJN3rXM+DlwENVHnj4++Bdaa8eY0IgTXxByKCVh80Oiw5T1GYhJyEbDEDTjX7BuopHAYMaF5ICQnseYtKNYxzYqyeXvREJ3QvE87T4alfQd5xY37ya3BB+X/Tp6g318eq9Wcd9jkHVnXcbNdVTGZic+sKf65xFAR1piYZAQheI/JLOo8xsZn9iGqvDKbnAXTxGGswYrllptuYXZuIfGDXLZLt0uxd4ku276Td4ItjKAbb26XuNT4L3qJGpecSDbdUJs3yCZxTXoLDzEtjwRPVjsunjnNytoafRUqgY4YljCw0OMPPvZXX1HVN2x3v6tCuIvIKvD13W7HNYAbgMt7LrSAlk47RUunneFqptN3qeqB7U5cFWoZ4OuXmn1abEBEHmrp9NJo6bQztHTaGa5VOr3UTmSLFi1atLgG0Qr3Fi1atNiDuFqE+1/sdgOuEbR02hlaOu0MLZ12hmuSTlfFhmqLFi1atPj24mpZubdo0aJFi28jdl24i8g7ROTrIvKUxFys1yVE5FYR+byIHBeRR0Xkw6l8n4j8m4g8mb6XUrmIyJ8kuj0sIq/f3Se4shARKyL/KyIPpOPbReRLiR5/LyJFKi/T8VPp/OHdbPeVhMQUl58QkcdF5DEReVPLT1shIr+ZxtwjIvK3ItLZC/y0q8JdRCzwp8Sk2ncB7xWRu3azTbuIJhH5XcAbgQ8lWvweMRH5UeCz6RimE5H/MjER+fWEDwOPTRz/IfBRVb0TuAB8IJV/ALiQyj+a6l0v+Bjwr6r6auC1RHq1/DQBEbkZ+A3gDar6PYAFfpa9wE+TrrZX+gO8Cfj0xPH9wP272aar5UNMfvKjROeug6nsINEnAODPgfdO1B/X2+sfYnavzwJvBR4g+gSeBbJ0fsxXwKeBN6XfWaonu/0MV4BGC8Czm5+15actdGpyPu9L/PEA8Pa9wE+7rZa5VDLt6xrpVe9bTUR+PeCPgd9lI77zfuCi6jjC9SQtxnRK55dT/b2O24EXgb9O6qu/FJEZWn6agqo+D/wRcAI4TeSPr7AH+Gm3hXuLTZBNicgnz2lcLlzX5k0i8uPAC6r6ld1uy1WODHg98Geq+jqgz4YKBmj5CSDtObyLOBkeAmaAd+xqo75N2G3hvqNk2tcL5DKJyNP5l52IfA/iB4GfEJHngL8jqmY+BiyKSBNOY5IWYzql8wvAuSvZ4F3CKeCUqn4pHX+CKOxbfprGfcCzqvqiqtbAPxB57Jrnp90W7l8Gjqad6YK4kfGpXW7TrkBi+MXLJSKHrYnI35+sHN7IDhKR7wWo6v2qeouqHibyy+dU9X3A54H3pGqb6dTQ7z2p/p5frarqGeCkiLwqFf0IcJyWnzbjBPBGEemlMdjQ6drnp91W+hOTaT8BPA18ZLfbs4t0+CHiK/LDwFfT551Efd5ngSeBfwf2pfpCtDR6Gvgacbd/15/jCtPsLcAD6fcR4BgxMfvHgTKVd9LxU+n8kd1u9xWkz93AQ4mn/hFYavlpWzr9PvA48AjwN0C5F/ip9VBt0aJFiz2I3VbLtGjRokWL7wBa4d6iRYsWexCtcG/RokWLPYhWuLdo0aLFHkQr3Fu0aNFiD6IV7i1atGixB9EK9xYtWrTYg2iFe4sWLVrsQfw/5DoY0wlVNzYAAAAASUVORK5CYII=\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print(\"List of Directories:\", os.listdir(data_dir))\n",
        "classes = os.listdir(data_dir + \"/Train\")\n",
        "print(\"List of classes:\", classes)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "jMQI7lQ1yw62",
        "outputId": "230a7371-7327-4fcc-bd2d-2df696312c2d"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "List of Directories: ['Train', 'Validation']\n",
            "List of classes: ['Idoma', 'Zulu', 'Tswana-Shweshwe', 'Tiv', 'Igala', 'Idgo', 'Adire', 'Xhosa-South Africa']\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import matplotlib.pyplot as plt\n",
        "\n",
        "def show_example(img, label):\n",
        "    print('Label: ', train_dataset.classes[label], \"(\"+str(label)+\")\")\n",
        "    plt.imshow(img.permute(1, 2, 0))"
      ],
      "metadata": {
        "id": "BypEq5nkO6OO"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "import random\n",
        "random_value = random.randint(1, 800)\n",
        "show_example(*train_dataset[2])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 200
        },
        "id": "DyFv_60tO-CX",
        "outputId": "e2c5f376-487c-474c-eb35-a7b4f6ca74fa"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "error",
          "ename": "TypeError",
          "evalue": "ignored",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mTypeError\u001b[0m                                 Traceback (most recent call last)",
            "\u001b[0;32m<ipython-input-21-867afbcb7210>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mrandom\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      2\u001b[0m \u001b[0mrandom_value\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mrandom\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrandint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m800\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 3\u001b[0;31m \u001b[0mshow_example\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0mtrain_dataset\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m2\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
            "\u001b[0;31mTypeError\u001b[0m: show_example() missing 1 required positional argument: 'label'"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "random_value = random.randint(1, 50)\n",
        "show_example(*train_dataset[random_value])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 183
        },
        "id": "fwd3dJuISN13",
        "outputId": "f500e768-0681-42c1-eb49-4c4f2c407e26"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "error",
          "ename": "TypeError",
          "evalue": "ignored",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mTypeError\u001b[0m                                 Traceback (most recent call last)",
            "\u001b[0;32m<ipython-input-22-bd8497ab1c6e>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[0mrandom_value\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mrandom\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrandint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m50\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 2\u001b[0;31m \u001b[0mshow_example\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0mtrain_dataset\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mrandom_value\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
            "\u001b[0;31mTypeError\u001b[0m: show_example() missing 1 required positional argument: 'label'"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "random_value = random.randint(1, 2000)\n",
        "show_example(*train_dataset[random_value])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 183
        },
        "id": "fwHgfCqISfy5",
        "outputId": "96610cc7-3cb8-4185-dbb0-2de5690b55a6"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "error",
          "ename": "IndexError",
          "evalue": "ignored",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mIndexError\u001b[0m                                Traceback (most recent call last)",
            "\u001b[0;32m<ipython-input-23-7a18f308e063>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[0mrandom_value\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mrandom\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrandint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m2000\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 2\u001b[0;31m \u001b[0mshow_example\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0mtrain_dataset\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mrandom_value\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
            "\u001b[0;31mIndexError\u001b[0m: string index out of range"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "random_value = random.randint(1, 2000)\n",
        "show_example(*train_dataset[random_value])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 183
        },
        "id": "wmib1-prUnIQ",
        "outputId": "edfbd01f-6ee3-4a3c-cd17-87b9c9dd9b97"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "error",
          "ename": "IndexError",
          "evalue": "ignored",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mIndexError\u001b[0m                                Traceback (most recent call last)",
            "\u001b[0;32m<ipython-input-24-7a18f308e063>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[0mrandom_value\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mrandom\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrandint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m2000\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 2\u001b[0;31m \u001b[0mshow_example\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0mtrain_dataset\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mrandom_value\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
            "\u001b[0;31mIndexError\u001b[0m: string index out of range"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "\n",
        "#Tensorboarding"
      ],
      "metadata": {
        "id": "2WA8xOFPL3xq"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "WZVwalrCNB27"
      },
      "outputs": [],
      "source": [
        "#tensorboard logging\n",
        "#to track various metrics such as accuracy and log loss on training or validation set\n",
        "from torch.utils.tensorboard import SummaryWriter\n",
        "\n",
        "TB_DIR = f'runs/exp_{datetime.now().strftime(\"%Y%m%d-%H%M%S\")}'  \n",
        "tb_train_writer = SummaryWriter(f'{TB_DIR}/Train')\n",
        "tb_val_writer = SummaryWriter(f'{TB_DIR}/Validation')\n",
        "\n",
        "%load_ext tensorboard"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "5zAbS8g2Alyu"
      },
      "source": [
        "#Model Building\n"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Helper function for training the model"
      ],
      "metadata": {
        "id": "6qVvJi-xVP9x"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "3O3ArQKtAlyu"
      },
      "outputs": [],
      "source": [
        "def train_model(model, criterion, optimizer, scheduler, num_epochs=4):\n",
        "    since = time.time()\n",
        "\n",
        "    best_model_wts = copy.deepcopy(model.state_dict())\n",
        "    best_acc = 0.0\n",
        "\n",
        "    for epoch in range(num_epochs):\n",
        "        print('Epoch {}/{}'.format(epoch, num_epochs - 1))\n",
        "        print('-' * 10)\n",
        "\n",
        "        # Each epoch has a training and validation phase\n",
        "        for phase in ['Train', 'Validation']:\n",
        "            if phase == 'Train':\n",
        "                model.train()  # Set model to training mode\n",
        "            else:\n",
        "                model.eval()   # Set model to evaluate mode\n",
        "\n",
        "            running_loss = 0.0\n",
        "            running_corrects = 0\n",
        "\n",
        "            # Iterate over data.\n",
        "            for inputs, labels in dataloaders[phase]:\n",
        "                inputs = inputs.to(device)\n",
        "                labels = labels.to(device)\n",
        "\n",
        "                # zero the parameter gradients\n",
        "                optimizer.zero_grad()\n",
        "\n",
        "                # forward\n",
        "                # track history if only in train\n",
        "                with torch.set_grad_enabled(phase == 'Train'):\n",
        "                    outputs = model(inputs)\n",
        "                    _, preds = torch.max(outputs, 1)\n",
        "                    loss = criterion(outputs, labels)\n",
        "\n",
        "                    # backward + optimize only if in training phase\n",
        "                    if phase == 'Train':\n",
        "                        loss.backward()\n",
        "                        optimizer.step()\n",
        "\n",
        "                # statistics\n",
        "                running_loss += loss.item() * inputs.size(0)\n",
        "                running_corrects += torch.sum(preds == labels.data)\n",
        "            if phase == 'Train':\n",
        "                scheduler.step()\n",
        "\n",
        "            epoch_loss = running_loss / dataset_sizes[phase]\n",
        "            epoch_acc = running_corrects.double() / dataset_sizes[phase]\n",
        "\n",
        "            if phase == 'Train':\n",
        "              tb_writer = tb_train_writer\n",
        "            else:\n",
        "              tb_writer = tb_val_writer\n",
        "\n",
        "            tb_writer.add_scalar(f'Loss', epoch_loss, epoch)\n",
        "            tb_writer.add_scalar(f'Accuracy', epoch_acc, epoch)\n",
        "\n",
        "            print('{} Loss: {:.4f} Acc: {:.4f}'.format(\n",
        "                phase, epoch_loss, epoch_acc))\n",
        "\n",
        "            # deep copy the model\n",
        "            if phase == 'Validation' and epoch_acc > best_acc:\n",
        "                best_acc = epoch_acc\n",
        "                best_model_wts = copy.deepcopy(model.state_dict())\n",
        "\n",
        "        print()\n",
        "\n",
        "    time_elapsed = time.time() - since\n",
        "    print('Training complete in {:.0f}m {:.0f}s'.format(\n",
        "        time_elapsed // 60, time_elapsed % 60))\n",
        "    print('Best val Acc: {:4f}'.format(best_acc))\n",
        "\n",
        "    tb_train_writer.close()\n",
        "    tb_val_writer.close()\n",
        "\n",
        "    # load best model weights\n",
        "    model.load_state_dict(best_model_wts)\n",
        "    return model"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "5LWqU4mmAlyv"
      },
      "source": [
        "## Visualizing Predictions\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "QYZ5GSZNAlyv"
      },
      "outputs": [],
      "source": [
        "def visualize_model(model, dataset, num_images=6):\n",
        "    was_training = model.training\n",
        "    model.eval()\n",
        "    images_so_far = 0\n",
        "    fig = plt.figure()\n",
        "\n",
        "    with torch.no_grad():\n",
        "        for i, (inputs, labels) in enumerate(dataloaders[dataset]):\n",
        "            inputs = inputs.to(device)\n",
        "            labels = labels.to(device)\n",
        "\n",
        "            outputs = model(inputs)\n",
        "            _, preds = torch.max(outputs, 1)\n",
        "\n",
        "            for j in range(inputs.size()[0]):\n",
        "                images_so_far += 1\n",
        "                ax = plt.subplot(num_images//2, 2, images_so_far)\n",
        "                ax.axis('off')\n",
        "                ax.set_title('predicted: {}'.format(class_names[preds[j]]))\n",
        "                imshow(inputs.cpu().data[j])\n",
        "\n",
        "                if images_so_far == num_images:\n",
        "                    model.train(mode=was_training)\n",
        "                    return\n",
        "        model.train(mode=was_training)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "4817UOzEAlyw"
      },
      "source": [
        "##Finetuning the Convnet\n",
        "----------------------\n",
        "Load a pretrained model and reset final fully connected layer.\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 156,
          "referenced_widgets": [
            "4ba027926cca41b9b24d857c4c8f5913",
            "f1598173524b4ccd9ee37a7f078a197f",
            "3414708488c5445c98bde8e4c337ee53",
            "d8b0bcbdf2974e3eb201bcfd91e43909",
            "5c51a2141ad94814a899475d24442f08",
            "cc654541b1a84d598827a9396826febc",
            "1196abc6814c4a9a9f1e4c1cac7c40b5",
            "6ea82114e667485ca8be17b862bf6675",
            "ff2198461ab940759745a6196f7424a5",
            "10ef03ee47d34107aad37a81d499a5e2",
            "4a616f72c18a45818cdfee30715babec"
          ]
        },
        "id": "pbs81lcCAlyw",
        "outputId": "b3710cb9-eb62-4e3e-cbe6-5bf4925eab56"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.8/dist-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead.\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.8/dist-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=ResNet18_Weights.IMAGENET1K_V1`. You can also use `weights=ResNet18_Weights.DEFAULT` to get the most up-to-date weights.\n",
            "  warnings.warn(msg)\n",
            "Downloading: \"https://download.pytorch.org/models/resnet18-f37072fd.pth\" to /root/.cache/torch/hub/checkpoints/resnet18-f37072fd.pth\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "  0%|          | 0.00/44.7M [00:00<?, ?B/s]"
            ],
            "application/vnd.jupyter.widget-view+json": {
              "version_major": 2,
              "version_minor": 0,
              "model_id": "4ba027926cca41b9b24d857c4c8f5913"
            }
          },
          "metadata": {}
        }
      ],
      "source": [
        "model_ft = models.resnet18(pretrained=True)\n",
        "num_ftrs = model_ft.fc.in_features\n",
        "# Here the size of each output sample is set to number of classes.\n",
        "# Alternatively, it can be generalized to nn.Linear(num_ftrs, len(class_names)).\n",
        "model_ft.fc = nn.Linear(num_ftrs, len(class_names))\n",
        "\n",
        "model_ft = model_ft.to(device)\n",
        "\n",
        "criterion = nn.CrossEntropyLoss()\n",
        "\n",
        "# Observe that all parameters are being optimized\n",
        "optimizer_ft = optim.SGD(model_ft.parameters(), lr=0.001, momentum=0.9)\n",
        "\n",
        "# Decay LR by a factor of 0.1 every 7 epochs\n",
        "exp_lr_scheduler = lr_scheduler.StepLR(optimizer_ft, step_size=7, gamma=0.1)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "qiu4u1RxAlyx"
      },
      "source": [
        "## Training and evaluation\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "wOcM17ZIAlyx",
        "outputId": "8cadd83d-0d4c-43e3-df31-9ffa1b089ad3"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 0/23\n",
            "----------\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.8/dist-packages/torch/utils/data/dataloader.py:554: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.\n",
            "  warnings.warn(_create_warning_msg(\n",
            "/usr/local/lib/python3.8/dist-packages/PIL/Image.py:959: UserWarning: Palette images with Transparency expressed in bytes should be converted to RGBA images\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.9094 Acc: 0.7050\n",
            "Validation Loss: 5.0017 Acc: 0.4550\n",
            "\n",
            "Epoch 1/23\n",
            "----------\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.8/dist-packages/PIL/Image.py:959: UserWarning: Palette images with Transparency expressed in bytes should be converted to RGBA images\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.6199 Acc: 0.8048\n",
            "Validation Loss: 5.4219 Acc: 0.4984\n",
            "\n",
            "Epoch 2/23\n",
            "----------\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.8/dist-packages/PIL/Image.py:959: UserWarning: Palette images with Transparency expressed in bytes should be converted to RGBA images\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.4822 Acc: 0.8493\n",
            "Validation Loss: 6.4376 Acc: 0.4760\n",
            "\n",
            "Epoch 3/23\n",
            "----------\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.8/dist-packages/PIL/Image.py:959: UserWarning: Palette images with Transparency expressed in bytes should be converted to RGBA images\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.4078 Acc: 0.8759\n",
            "Validation Loss: 5.8308 Acc: 0.4748\n",
            "\n",
            "Epoch 4/23\n",
            "----------\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.8/dist-packages/PIL/Image.py:959: UserWarning: Palette images with Transparency expressed in bytes should be converted to RGBA images\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.3153 Acc: 0.9018\n",
            "Validation Loss: 6.1894 Acc: 0.5085\n",
            "\n",
            "Epoch 5/23\n",
            "----------\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.8/dist-packages/PIL/Image.py:959: UserWarning: Palette images with Transparency expressed in bytes should be converted to RGBA images\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.2721 Acc: 0.9133\n",
            "Validation Loss: 5.3384 Acc: 0.5039\n",
            "\n",
            "Epoch 6/23\n",
            "----------\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.8/dist-packages/PIL/Image.py:959: UserWarning: Palette images with Transparency expressed in bytes should be converted to RGBA images\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.2435 Acc: 0.9304\n",
            "Validation Loss: 5.5488 Acc: 0.4957\n",
            "\n",
            "Epoch 7/23\n",
            "----------\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.8/dist-packages/PIL/Image.py:959: UserWarning: Palette images with Transparency expressed in bytes should be converted to RGBA images\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.1427 Acc: 0.9610\n",
            "Validation Loss: 5.5125 Acc: 0.5050\n",
            "\n",
            "Epoch 8/23\n",
            "----------\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.8/dist-packages/PIL/Image.py:959: UserWarning: Palette images with Transparency expressed in bytes should be converted to RGBA images\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.1040 Acc: 0.9715\n",
            "Validation Loss: 5.9554 Acc: 0.5093\n",
            "\n",
            "Epoch 9/23\n",
            "----------\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.8/dist-packages/PIL/Image.py:959: UserWarning: Palette images with Transparency expressed in bytes should be converted to RGBA images\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.0931 Acc: 0.9745\n",
            "Validation Loss: 5.9414 Acc: 0.5062\n",
            "\n",
            "Epoch 10/23\n",
            "----------\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.8/dist-packages/PIL/Image.py:959: UserWarning: Palette images with Transparency expressed in bytes should be converted to RGBA images\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.0998 Acc: 0.9719\n",
            "Validation Loss: 5.4425 Acc: 0.5093\n",
            "\n",
            "Epoch 11/23\n",
            "----------\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.8/dist-packages/PIL/Image.py:959: UserWarning: Palette images with Transparency expressed in bytes should be converted to RGBA images\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.0802 Acc: 0.9797\n",
            "Validation Loss: 5.4063 Acc: 0.5101\n",
            "\n",
            "Epoch 12/23\n",
            "----------\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.8/dist-packages/PIL/Image.py:959: UserWarning: Palette images with Transparency expressed in bytes should be converted to RGBA images\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.0673 Acc: 0.9822\n",
            "Validation Loss: 5.4766 Acc: 0.5105\n",
            "\n",
            "Epoch 13/23\n",
            "----------\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.8/dist-packages/PIL/Image.py:959: UserWarning: Palette images with Transparency expressed in bytes should be converted to RGBA images\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.0636 Acc: 0.9821\n",
            "Validation Loss: 5.4198 Acc: 0.5136\n",
            "\n",
            "Epoch 14/23\n",
            "----------\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.8/dist-packages/PIL/Image.py:959: UserWarning: Palette images with Transparency expressed in bytes should be converted to RGBA images\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.0661 Acc: 0.9819\n",
            "Validation Loss: 5.7225 Acc: 0.5140\n",
            "\n",
            "Epoch 15/23\n",
            "----------\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.8/dist-packages/PIL/Image.py:959: UserWarning: Palette images with Transparency expressed in bytes should be converted to RGBA images\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.0640 Acc: 0.9831\n",
            "Validation Loss: 5.5556 Acc: 0.5159\n",
            "\n",
            "Epoch 16/23\n",
            "----------\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.8/dist-packages/PIL/Image.py:959: UserWarning: Palette images with Transparency expressed in bytes should be converted to RGBA images\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.0666 Acc: 0.9831\n",
            "Validation Loss: 5.5762 Acc: 0.5147\n",
            "\n",
            "Epoch 17/23\n",
            "----------\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.8/dist-packages/PIL/Image.py:959: UserWarning: Palette images with Transparency expressed in bytes should be converted to RGBA images\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.0596 Acc: 0.9837\n",
            "Validation Loss: 5.4863 Acc: 0.5155\n",
            "\n",
            "Epoch 18/23\n",
            "----------\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.8/dist-packages/PIL/Image.py:959: UserWarning: Palette images with Transparency expressed in bytes should be converted to RGBA images\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.0653 Acc: 0.9829\n",
            "Validation Loss: 5.4856 Acc: 0.5163\n",
            "\n",
            "Epoch 19/23\n",
            "----------\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.8/dist-packages/PIL/Image.py:959: UserWarning: Palette images with Transparency expressed in bytes should be converted to RGBA images\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.0624 Acc: 0.9827\n",
            "Validation Loss: 5.4184 Acc: 0.5147\n",
            "\n",
            "Epoch 20/23\n",
            "----------\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.8/dist-packages/PIL/Image.py:959: UserWarning: Palette images with Transparency expressed in bytes should be converted to RGBA images\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.0563 Acc: 0.9842\n",
            "Validation Loss: 5.6183 Acc: 0.5143\n",
            "\n",
            "Epoch 21/23\n",
            "----------\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.8/dist-packages/PIL/Image.py:959: UserWarning: Palette images with Transparency expressed in bytes should be converted to RGBA images\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.0656 Acc: 0.9831\n",
            "Validation Loss: 5.7027 Acc: 0.5155\n",
            "\n",
            "Epoch 22/23\n",
            "----------\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.8/dist-packages/PIL/Image.py:959: UserWarning: Palette images with Transparency expressed in bytes should be converted to RGBA images\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.0683 Acc: 0.9822\n",
            "Validation Loss: 5.5706 Acc: 0.5182\n",
            "\n",
            "Epoch 23/23\n",
            "----------\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.8/dist-packages/PIL/Image.py:959: UserWarning: Palette images with Transparency expressed in bytes should be converted to RGBA images\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.0621 Acc: 0.9830\n",
            "Validation Loss: 5.8770 Acc: 0.5093\n",
            "\n",
            "Training complete in 55m 53s\n",
            "Best val Acc: 0.518217\n"
          ]
        }
      ],
      "source": [
        "model_ft = train_model(model_ft, criterion, optimizer_ft, exp_lr_scheduler,\n",
        "                       num_epochs=24)"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "#Showing TensorBoard for metrices learnt"
      ],
      "metadata": {
        "id": "9_AIOH4AF9Kr"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "ybRAPqSeOdci"
      },
      "outputs": [],
      "source": [
        "%tensorboard --logdir='./runs'"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ZaNzMBpRsEdZ"
      },
      "source": [
        "# Model Testing"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "_VvKDMlnV394"
      },
      "outputs": [],
      "source": [
        "def check_accuracy(test_loader, model, device):\n",
        "    num_correct = 0\n",
        "    total = 0\n",
        "    model.eval()\n",
        "\n",
        "    with torch.no_grad():\n",
        "        for data, labels in test_loader:\n",
        "            data = data.to(device=device)\n",
        "            labels = labels.to(device=device)\n",
        "\n",
        "            outputs = model(data)\n",
        "            _, predictions = torch.max(outputs, 1)\n",
        "            num_correct += (predictions == labels).sum()\n",
        "            total += labels.size(0)\n",
        "        print(f\"Test Accuracy of the model: {float(num_correct)/float(total):.2f}\")\n",
        "\n",
        "def visualize_test_errors(test_loader, model, device):\n",
        "    model.eval()\n",
        "    incorrect_examples=[]\n",
        "\n",
        "    with torch.no_grad():\n",
        "        for data, labels in test_loader:\n",
        "            data = data.to(device=device)\n",
        "            labels = labels.to(device=device)\n",
        "\n",
        "            outputs = model(data)\n",
        "            _, predictions = torch.max(outputs, 1)\n",
        "            idxs_mask = (predictions != labels).view(-1)\n",
        "            num_true_idxs = idxs_mask.sum() \n",
        "            if num_true_idxs > 0:\n",
        "                found_falses = data[idxs_mask].cpu().numpy()\n",
        "                false_predictions = predictions[idxs_mask].cpu().numpy()\n",
        "                false_labels = labels[idxs_mask].cpu().numpy()\n",
        "                for j in range(found_falses.shape[0]):\n",
        "                    incorrect_examples.append((found_falses[j,:,:,:], \n",
        "                                               false_predictions[j],\n",
        "                                               false_labels[j]))\n",
        "    return incorrect_examples\n",
        "\n",
        "def show_image(inp, title=None):\n",
        "    inp = inp.transpose((1, 2, 0))\n",
        "    mean = np.array([0.485, 0.456, 0.406])\n",
        "    std = np.array([0.229, 0.224, 0.225])\n",
        "    inp = std * inp + mean\n",
        "    inp = np.clip(inp, 0, 1)\n",
        "    plt.imshow(inp)\n",
        "    if title is not None:\n",
        "        plt.title(title)\n",
        "    plt.pause(0.001) \n",
        "\n",
        "def visualize_error_images(error_examples):\n",
        "    for img, prediction, label in error_examples:\n",
        "        print(f'gt label: {class_names[label]}')\n",
        "        print(f'prediction: {class_names[prediction]}')\n",
        "        show_image(img.squeeze())\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "YbxL85UxsNct"
      },
      "outputs": [],
      "source": [
        "error_images = check_accuracy(dataloaders['Validation'], model_ft, device)\n",
        "\n",
        "error_examples = visualize_test_errors(dataloaders['Validation'], model_ft, device)\n",
        "#shows some prediction errors. Might show nothing if prediction is 100%\n",
        "visualize_error_images(error_examples)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "UUwPBc9P2Dah"
      },
      "source": [
        "# Model Saving"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "s1nNfRN_HgJj"
      },
      "outputs": [],
      "source": [
        "torch.save(model_ft, 'modelv_1_0.pt')"
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "cRRo9UTNTzf-"
      },
      "execution_count": null,
      "outputs": []
    }
  ],
  "metadata": {
    "accelerator": "GPU",
    "colab": {
      "provenance": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "display_name": "Python 3",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.6.13"
    },
    "widgets": {
      "application/vnd.jupyter.widget-state+json": {
        "4ba027926cca41b9b24d857c4c8f5913": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HBoxModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HBoxModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HBoxView",
            "box_style": "",
            "children": [
              "IPY_MODEL_f1598173524b4ccd9ee37a7f078a197f",
              "IPY_MODEL_3414708488c5445c98bde8e4c337ee53",
              "IPY_MODEL_d8b0bcbdf2974e3eb201bcfd91e43909"
            ],
            "layout": "IPY_MODEL_5c51a2141ad94814a899475d24442f08"
          }
        },
        "f1598173524b4ccd9ee37a7f078a197f": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_cc654541b1a84d598827a9396826febc",
            "placeholder": "​",
            "style": "IPY_MODEL_1196abc6814c4a9a9f1e4c1cac7c40b5",
            "value": "100%"
          }
        },
        "3414708488c5445c98bde8e4c337ee53": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "FloatProgressModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "FloatProgressModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "ProgressView",
            "bar_style": "success",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_6ea82114e667485ca8be17b862bf6675",
            "max": 46830571,
            "min": 0,
            "orientation": "horizontal",
            "style": "IPY_MODEL_ff2198461ab940759745a6196f7424a5",
            "value": 46830571
          }
        },
        "d8b0bcbdf2974e3eb201bcfd91e43909": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_10ef03ee47d34107aad37a81d499a5e2",
            "placeholder": "​",
            "style": "IPY_MODEL_4a616f72c18a45818cdfee30715babec",
            "value": " 44.7M/44.7M [00:00&lt;00:00, 221MB/s]"
          }
        },
        "5c51a2141ad94814a899475d24442f08": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "cc654541b1a84d598827a9396826febc": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "1196abc6814c4a9a9f1e4c1cac7c40b5": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "6ea82114e667485ca8be17b862bf6675": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "ff2198461ab940759745a6196f7424a5": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "ProgressStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "bar_color": null,
            "description_width": ""
          }
        },
        "10ef03ee47d34107aad37a81d499a5e2": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "4a616f72c18a45818cdfee30715babec": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        }
      }
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}