Arabic_Handwritten_Recognition.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Copy of Arabic Handwritten Recognition.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "toc_visible": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "c6TiNABwkAhe",
        "colab_type": "text"
      },
      "source": [
        "# Arabic Handwritten Recognition"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "yLgu44Fvkgdr",
        "colab_type": "text"
      },
      "source": [
        "## Overview\n",
        "The automatic recognition of text on scanned images has enabled many applications such as searching for words in large volumes of documents, automatic sorting of postal mail, and convenient editing of previously printed documents.\n",
        "\n",
        "The domain of handwriting in the Arabic script presents unique technical challenges and has been addressed more recently than other domains. Many different methods have been proposed and applied to various types of images.\n",
        "\n",
        "Here we will focus on the recognition part of handwritten Arabic letters and digits recognition that face several challenges, including the unlimited variation in human handwriting and the large public databases.\n",
        "\n",
        "In this project we will employ several deep learning models to classify a the images to an arabic letter or digit. \n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "jDxpa4JAonJF",
        "colab_type": "text"
      },
      "source": [
        "## Dataset\n",
        "The dataset for this project originates from kaggle kernels which include \n",
        "**[Arabic Digits](https://www.kaggle.com/mloey1/ahdd1)** and **[Arabic Letters](https://www.kaggle.com/mloey1/ahcd1)**.\n",
        "\n",
        "All the datasets are CSV files representing the image pixels values and their corresponding label.\n",
        "\n",
        "Here are some more details about the datasets:\n",
        "\n",
        "* **Arabic Digits Dataset represents MADBase** (modified Arabic handwritten digits database) which contains **60,000 training images, and 10,000 test images**. MADBase was **written by 700 writers**. Each writer wrote each digit (from 0 -9) ten times. To ensure including different writing styles, the database was gathered from different institutions: Colleges of Engineering and Law, School of Medicine, the Open University (whose students span a wide range of ages), a high school, and a governmental institution. MADBase is available for free and can be downloaded from [here](http://datacenter.aucegypt.edu/shazeem/).\n",
        "\n",
        "* **Arabic Letters Dataset is composed of 16,800 characters written by 60 participants**, the age range is between 19 to 40 years, and 90% of participants are right-hand. Each participant wrote each character (from ’alef’ to ’yeh’) ten times. The images were scanned at the resolution of 300 dpi. Each block is segmented automatically using Matlab 2016a to determining the coordinates for each block. **The dataset is partitioned into two sets: a training set of 13,440 characters to 480 images per class and a test set of 3,360 characters to 120 images per class**. Writers of training set and test set are exclusive. Ordering of including writers to test set are randomized to make sure that writers of test set are not from a single institution to ensure variability of the test set."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "m6_ru2oupR7h",
        "colab_type": "text"
      },
      "source": [
        "## Data Exploration\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "q2Z4_QkxppsF",
        "colab_type": "text"
      },
      "source": [
        "1. We need to mount to google drive to be able to use the uploaded dataset files.\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qNoWN1Ts6SFr",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 122
        },
        "outputId": "5a5a11e7-b58d-48ed-b04a-22105752a641"
      },
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Go to this URL in a browser: https://accounts.google.com/o/oauth2/auth?client_id=947318989803-6bn6qk8qdgf4n4g3pfee6491hc0brc4i.apps.googleusercontent.com&redirect_uri=urn%3aietf%3awg%3aoauth%3a2.0%3aoob&response_type=code&scope=email%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdocs.test%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdrive%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdrive.photos.readonly%20https%3a%2f%2fwww.googleapis.com%2fauth%2fpeopleapi.readonly\n",
            "\n",
            "Enter your authorization code:\n",
            "··········\n",
            "Mounted at /content/drive\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "jJo8eDIRp5ku",
        "colab_type": "text"
      },
      "source": [
        "2. Import libraries necessary for this project.\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "v3o6mEbK6W9v",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# Import main libraries necessary for this project\n",
        "import numpy as np\n",
        "import pandas as pd\n",
        "from IPython.display import display # Allows the use of display() for DataFrames\n",
        "\n",
        "# Import libraries needed for reading image and processing it\n",
        "import csv\n",
        "from PIL import Image\n",
        "from scipy.ndimage import rotate\n",
        "\n",
        "# Pretty display for notebooks\n",
        "%matplotlib inline"
      ],
      "execution_count": 2,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "X0ubj6mrqXku",
        "colab_type": "text"
      },
      "source": [
        "3. Load the dataset files into dataframes to be used later in the project"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "rGcWZ2BBrLrr",
        "colab_type": "text"
      },
      "source": [
        "### Loading Arabic Letters Dataset"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1MRwVjsDKpCN",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 287
        },
        "outputId": "ed64137f-e2f9-4908-b8ea-2b227519ce17"
      },
      "source": [
        "# Training letters images and labels files\n",
        "letters_training_images_file_path = \"drive/My Drive/Capstone Project/Datasets/Arabic Handwritten Characters Dataset CSV/training images.zip\"\n",
        "letters_training_labels_file_path = \"drive/My Drive/Capstone Project/Datasets/Arabic Handwritten Characters Dataset CSV/training labels.zip\"\n",
        "# Testing letters images and labels files\n",
        "letters_testing_images_file_path = \"drive/My Drive/Capstone Project/Datasets/Arabic Handwritten Characters Dataset CSV/testing images.zip\"\n",
        "letters_testing_labels_file_path = \"drive/My Drive/Capstone Project/Datasets/Arabic Handwritten Characters Dataset CSV/testing labels.zip\"\n",
        "\n",
        "# Loading dataset into dataframes\n",
        "training_letters_images = pd.read_csv(letters_training_images_file_path, compression='zip', header=None)\n",
        "training_letters_labels = pd.read_csv(letters_training_labels_file_path, compression='zip', header=None)\n",
        "testing_letters_images = pd.read_csv(letters_testing_images_file_path, compression='zip', header=None)\n",
        "testing_letters_labels = pd.read_csv(letters_testing_labels_file_path, compression='zip', header=None)\n",
        "\n",
        "# print statistics about the dataset\n",
        "print(\"There are %d training arabic letter images of 64x64 pixels.\" %training_letters_images.shape[0])\n",
        "print(\"There are %d testing arabic letter images of 64x64 pixels.\" %testing_letters_images.shape[0])\n",
        "training_letters_images.head()"
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "There are 13440 training arabic letter images of 64x64 pixels.\n",
            "There are 3360 testing arabic letter images of 64x64 pixels.\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>0</th>\n",
              "      <th>1</th>\n",
              "      <th>2</th>\n",
              "      <th>3</th>\n",
              "      <th>4</th>\n",
              "      <th>5</th>\n",
              "      <th>6</th>\n",
              "      <th>7</th>\n",
              "      <th>8</th>\n",
              "      <th>9</th>\n",
              "      <th>10</th>\n",
              "      <th>11</th>\n",
              "      <th>12</th>\n",
              "      <th>13</th>\n",
              "      <th>14</th>\n",
              "      <th>15</th>\n",
              "      <th>16</th>\n",
              "      <th>17</th>\n",
              "      <th>18</th>\n",
              "      <th>19</th>\n",
              "      <th>20</th>\n",
              "      <th>21</th>\n",
              "      <th>22</th>\n",
              "      <th>23</th>\n",
              "      <th>24</th>\n",
              "      <th>25</th>\n",
              "      <th>26</th>\n",
              "      <th>27</th>\n",
              "      <th>28</th>\n",
              "      <th>29</th>\n",
              "      <th>30</th>\n",
              "      <th>31</th>\n",
              "      <th>32</th>\n",
              "      <th>33</th>\n",
              "      <th>34</th>\n",
              "      <th>35</th>\n",
              "      <th>36</th>\n",
              "      <th>37</th>\n",
              "      <th>38</th>\n",
              "      <th>39</th>\n",
              "      <th>...</th>\n",
              "      <th>4056</th>\n",
              "      <th>4057</th>\n",
              "      <th>4058</th>\n",
              "      <th>4059</th>\n",
              "      <th>4060</th>\n",
              "      <th>4061</th>\n",
              "      <th>4062</th>\n",
              "      <th>4063</th>\n",
              "      <th>4064</th>\n",
              "      <th>4065</th>\n",
              "      <th>4066</th>\n",
              "      <th>4067</th>\n",
              "      <th>4068</th>\n",
              "      <th>4069</th>\n",
              "      <th>4070</th>\n",
              "      <th>4071</th>\n",
              "      <th>4072</th>\n",
              "      <th>4073</th>\n",
              "      <th>4074</th>\n",
              "      <th>4075</th>\n",
              "      <th>4076</th>\n",
              "      <th>4077</th>\n",
              "      <th>4078</th>\n",
              "      <th>4079</th>\n",
              "      <th>4080</th>\n",
              "      <th>4081</th>\n",
              "      <th>4082</th>\n",
              "      <th>4083</th>\n",
              "      <th>4084</th>\n",
              "      <th>4085</th>\n",
              "      <th>4086</th>\n",
              "      <th>4087</th>\n",
              "      <th>4088</th>\n",
              "      <th>4089</th>\n",
              "      <th>4090</th>\n",
              "      <th>4091</th>\n",
              "      <th>4092</th>\n",
              "      <th>4093</th>\n",
              "      <th>4094</th>\n",
              "      <th>4095</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>5 rows × 4096 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "   0     1     2     3     4     5     ...  4090  4091  4092  4093  4094  4095\n",
              "0     0     0     0     0     0     0  ...     0     0     0     0     0     0\n",
              "1     0     0     0     0     0     0  ...     0     0     0     0     0     0\n",
              "2     0     0     0     0     0     0  ...     0     0     0     0     0     0\n",
              "3     0     0     0     0     0     0  ...     0     0     0     0     0     0\n",
              "4     0     0     0     0     0     0  ...     0     0     0     0     0     0\n",
              "\n",
              "[5 rows x 4096 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 4
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "KuPNRawns2BV",
        "colab_type": "text"
      },
      "source": [
        "### Loading Arabic Digits Dataset"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Zyoid-WpLjXT",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        },
        "outputId": "504deeac-a270-4927-b4c6-34ff7d9304c9"
      },
      "source": [
        "# Training digits images and labels files\n",
        "digits_training_images_file_path = \"drive/My Drive/Capstone Project/Datasets/Arabic Handwritten Digits Dataset CSV/training images.zip\"\n",
        "digits_training_labels_file_path = \"drive/My Drive/Capstone Project/Datasets/Arabic Handwritten Digits Dataset CSV/training labels.zip\"\n",
        "# Testing digits images and labels files\n",
        "digits_testing_images_file_path = \"drive/My Drive/Capstone Project/Datasets/Arabic Handwritten Digits Dataset CSV/testing images.zip\"\n",
        "digits_testing_labels_file_path = \"drive/My Drive/Capstone Project/Datasets/Arabic Handwritten Digits Dataset CSV/testing labels.zip\"\n",
        "\n",
        "# Loading dataset into dataframes\n",
        "training_digits_images = pd.read_csv(digits_training_images_file_path, compression='zip', header=None)\n",
        "training_digits_labels = pd.read_csv(digits_training_labels_file_path, compression='zip', header=None)\n",
        "testing_digits_images = pd.read_csv(digits_testing_images_file_path, compression='zip', header=None)\n",
        "testing_digits_labels = pd.read_csv(digits_testing_labels_file_path, compression='zip', header=None)\n",
        "\n",
        "# print statistics about the dataset\n",
        "print(\"There are %d training arabic digit images of 64x64 pixels.\" %training_digits_images.shape[0])\n",
        "print(\"There are %d testing arabic digit images of 64x64 pixels.\" %testing_digits_images.shape[0])"
      ],
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "There are 60000 training arabic digit images of 64x64 pixels.\n",
            "There are 10000 testing arabic digit images of 64x64 pixels.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "LGFth_ks0VH4",
        "colab_type": "text"
      },
      "source": [
        "### Convert csv values to an image \n",
        "Writting a method to be used later if we want visualization of an image from its pixels values."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2MbTUT_mMBGH",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def convert_values_to_image(image_values, display=False):\n",
        "  image_array = np.asarray(image_values)\n",
        "  image_array = image_array.reshape(64, 64).astype('uint8')\n",
        "  # The original dataset is reflected so we will flip it then rotate for a better view only.\n",
        "  image_array = np.flip(image_array, 0)\n",
        "  image_array = rotate(image_array, -90)\n",
        "  new_image = Image.fromarray(image_array)\n",
        "  if display == True:\n",
        "    new_image.show()\n",
        "  return new_image"
      ],
      "execution_count": 7,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "MBwu1eDk0mjj",
        "colab_type": "text"
      },
      "source": [
        "### Visualizing some examples"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pX4PxwqbyQg8",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 81
        },
        "outputId": "6433de79-c1f7-4c09-afc5-e17b782b8cc4"
      },
      "source": [
        "convert_values_to_image(training_letters_images.loc[0], True)"
      ],
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<PIL.Image.Image image mode=L size=64x64 at 0x7F5B4D53E5F8>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "R0J_oTy0yeXb",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 81
        },
        "outputId": "77ab78b1-95e9-4c4e-dfcb-0a2dbbc9ab4e"
      },
      "source": [
        "convert_values_to_image(training_letters_images.loc[12], True)"
      ],
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<PIL.Image.Image image mode=L size=64x64 at 0x7F5B40CACF98>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IVwNBEIj2dcK",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 81
        },
        "outputId": "1d587abd-eee0-4c3d-c3bb-2ebf64c188b1"
      },
      "source": [
        "convert_values_to_image(training_letters_images.loc[37], True)"
      ],
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<PIL.Image.Image image mode=L size=64x64 at 0x7F5B40C760F0>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 10
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pNKxXW5-2fl9",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 81
        },
        "outputId": "d94f24dd-9059-48eb-f19e-b81cf97a5dc3"
      },
      "source": [
        "convert_values_to_image(training_letters_images.loc[3000], True)"
      ],
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<PIL.Image.Image image mode=L size=64x64 at 0x7F5B40C761D0>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 11
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FIYtePhT2iBr",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 81
        },
        "outputId": "d51afddf-7a83-4b04-c29f-a6ad3c1eb271"
      },
      "source": [
        "convert_values_to_image(training_digits_images.loc[0], True)"
      ],
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<PIL.Image.Image image mode=L size=64x64 at 0x7F5B40C76278>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 12
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zdnSlqNxYB_H",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 81
        },
        "outputId": "db1dd27a-fefe-4ec2-bf3e-ffa57066541e"
      },
      "source": [
        "convert_values_to_image(training_digits_images.loc[2], True)"
      ],
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<PIL.Image.Image image mode=L size=64x64 at 0x7F5B40C76320>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 13
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hlSt-2_2i8O1",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 81
        },
        "outputId": "464c57e0-2b7a-4c4d-f2df-aab1a0dbb5a5"
      },
      "source": [
        "convert_values_to_image(training_digits_images.loc[7], True)"
      ],
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<PIL.Image.Image image mode=L size=64x64 at 0x7F5B40C763C8>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 14
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "xsucycIri9ht",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 81
        },
        "outputId": "714e35f7-37a0-408e-c0d6-1bba94836e82"
      },
      "source": [
        "convert_values_to_image(training_digits_images.loc[9], True)"
      ],
      "execution_count": 15,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<PIL.Image.Image image mode=L size=64x64 at 0x7F5B40C76470>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 15
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ZZT4fEuLoLXw",
        "colab_type": "text"
      },
      "source": [
        "## Data Preprocessing"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "TZH8hpO_qgem",
        "colab_type": "text"
      },
      "source": [
        "### Image Normalization\n",
        "We rescale the images by dividing every pixel in the image by 255 to make them into range [0, 1]"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1HJI0WG0oNmT",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "training_digits_images_scaled = training_digits_images.values.astype('float32')/255\n",
        "training_digits_labels = training_digits_labels.values.astype('int32')\n",
        "testing_digits_images_scaled = testing_digits_images.values.astype('float32')/255\n",
        "testing_digits_labels = testing_digits_labels.values.astype('int32')\n",
        "\n",
        "training_letters_images_scaled = training_letters_images.values.astype('float32')/255\n",
        "training_letters_labels = training_letters_labels.values.astype('int32')\n",
        "testing_letters_images_scaled = testing_letters_images.values.astype('float32')/255\n",
        "testing_letters_labels = testing_letters_labels.values.astype('int32')"
      ],
      "execution_count": 16,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4GtF5Gc4sGsv",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 136
        },
        "outputId": "5f730686-de88-47dc-a308-2618ddcdd152"
      },
      "source": [
        "print(\"Training images of digits after scaling\")\n",
        "print(training_digits_images_scaled.shape)\n",
        "training_digits_images_scaled[0:5]"
      ],
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Training images of digits after scaling\n",
            "(60000, 4096)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[0., 0., 0., ..., 0., 0., 0.],\n",
              "       [0., 0., 0., ..., 0., 0., 0.],\n",
              "       [0., 0., 0., ..., 0., 0., 0.],\n",
              "       [0., 0., 0., ..., 0., 0., 0.],\n",
              "       [0., 0., 0., ..., 0., 0., 0.]], dtype=float32)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 17
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "n68sxKC0sTkq",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 136
        },
        "outputId": "15a162f4-6ba7-4b01-d8b5-81c4e2eced74"
      },
      "source": [
        "print(\"Training images of letters after scaling\")\n",
        "print(training_letters_images_scaled.shape)\n",
        "training_letters_images_scaled[0:5]"
      ],
      "execution_count": 18,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Training images of letters after scaling\n",
            "(13440, 4096)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[0., 0., 0., ..., 0., 0., 0.],\n",
              "       [0., 0., 0., ..., 0., 0., 0.],\n",
              "       [0., 0., 0., ..., 0., 0., 0.],\n",
              "       [0., 0., 0., ..., 0., 0., 0.],\n",
              "       [0., 0., 0., ..., 0., 0., 0.]], dtype=float32)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 18
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "T-X1SJjnsq8b",
        "colab_type": "text"
      },
      "source": [
        "### Encoding Categorical Labels\n",
        "From the labels csv files we can see that labels are categorical values and it is a multi-class classification problem. \n",
        "\n",
        "Our outputs are in the form of:\n",
        "* Digits from 0 to 9 have categories numbers from 0 to 9\n",
        "* Letters from ’alef’ to ’yeh’ have categories numbers from 10 to 37\n",
        "\n",
        "**Here we will encode these categories values using One Hot Encoding with keras.**\n",
        "\n",
        "One-hot encoding transforms integer to a binary matrix where the array contains only one ‘1’ and the rest elements are ‘0’.\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "uQyT17YzsU7U",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "bcc5de2e-c436-44f6-b987-f5d8d376a447"
      },
      "source": [
        "from keras.utils import to_categorical\n",
        "\n",
        "# one hot encoding\n",
        "# number of classes = 10 (digits classes) + 28 (arabic alphabet classes)\n",
        "number_of_classes = 38\n",
        "training_letters_labels_encoded = to_categorical(training_letters_labels, num_classes=number_of_classes)\n",
        "testing_letters_labels_encoded = to_categorical(testing_letters_labels, num_classes=number_of_classes)\n",
        "training_digits_labels_encoded = to_categorical(training_digits_labels, num_classes=number_of_classes)\n",
        "testing_digits_labels_encoded = to_categorical(testing_digits_labels, num_classes=number_of_classes)"
      ],
      "execution_count": 19,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Using TensorFlow backend.\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9CyoYTDl1BXD",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 136
        },
        "outputId": "417bbbb9-04d4-4a7d-f61f-78c732d241b5"
      },
      "source": [
        "print(training_digits_labels_encoded)"
      ],
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[[1. 0. 0. ... 0. 0. 0.]\n",
            " [0. 1. 0. ... 0. 0. 0.]\n",
            " [0. 0. 1. ... 0. 0. 0.]\n",
            " ...\n",
            " [0. 0. 0. ... 0. 0. 0.]\n",
            " [0. 0. 0. ... 0. 0. 0.]\n",
            " [0. 0. 0. ... 0. 0. 0.]]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "nsJzI6Gp5bbA",
        "colab_type": "text"
      },
      "source": [
        "### Reshaping Input Images to 64x64x1\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "VDldXMz76EQH",
        "colab_type": "text"
      },
      "source": [
        "When using TensorFlow as backend, Keras CNNs require a 4D array (which we'll also refer to as a 4D tensor) as input, with shape\n",
        "(nb_samples,rows,columns,channels)\n",
        "\n",
        "where nb_samples corresponds to the total number of images (or samples), and rows, columns, and channels correspond to the number of rows, columns, and channels for each image, respectively.\n",
        "\n",
        "**So we will reshape the input images to a 4D tensor with shape\n",
        "(nb_samples, 64, 64 ,1)** as we use grayscale images of 64x64 pixels."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EWTQ0nq85wWL",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        },
        "outputId": "fae3559a-8abf-4888-dddf-1bafa65b7afc"
      },
      "source": [
        "# reshape input digit images to 64x64x1\n",
        "training_digits_images_scaled = training_digits_images_scaled.reshape([-1, 64, 64, 1])\n",
        "testing_digits_images_scaled = testing_digits_images_scaled.reshape([-1, 64, 64, 1])\n",
        "\n",
        "# reshape input letter images to 64x64x1\n",
        "training_letters_images_scaled = training_letters_images_scaled.reshape([-1, 64, 64, 1])\n",
        "testing_letters_images_scaled = testing_letters_images_scaled.reshape([-1, 64, 64, 1])\n",
        "\n",
        "print(training_digits_images_scaled.shape, training_digits_labels_encoded.shape, testing_digits_images_scaled.shape, testing_digits_labels_encoded.shape)\n",
        "print(training_letters_images_scaled.shape, training_letters_labels_encoded.shape, testing_letters_images_scaled.shape, testing_letters_labels_encoded.shape)"
      ],
      "execution_count": 21,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(60000, 64, 64, 1) (60000, 38) (10000, 64, 64, 1) (10000, 38)\n",
            "(13440, 64, 64, 1) (13440, 38) (3360, 64, 64, 1) (3360, 38)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Gtsr1vVaYS9D",
        "colab_type": "text"
      },
      "source": [
        "### Merging Letters and Digits Datasets"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1bdfCcFpYpA_",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 85
        },
        "outputId": "8d6b7393-2618-42d0-bee4-98b2e02a787e"
      },
      "source": [
        "training_data_images = np.concatenate((training_digits_images_scaled, training_letters_images_scaled), axis=0) \n",
        "training_data_labels = np.concatenate((training_digits_labels_encoded, training_letters_labels_encoded), axis=0)\n",
        "print(\"Total Training images are {} images of shape\".format(training_data_images.shape[0]))\n",
        "print(training_data_images.shape, training_data_labels.shape)\n",
        "\n",
        "\n",
        "testing_data_images = np.concatenate((testing_digits_images_scaled, testing_letters_images_scaled), axis=0) \n",
        "testing_data_labels = np.concatenate((testing_digits_labels_encoded, testing_letters_labels_encoded), axis=0)\n",
        "print(\"Total Testing images are {} images of shape\".format(testing_data_images.shape[0]))\n",
        "print(testing_data_images.shape, testing_data_labels.shape)\n"
      ],
      "execution_count": 22,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Total Training images are 73440 images of shape\n",
            "(73440, 64, 64, 1) (73440, 38)\n",
            "Total Testing images are 13360 images of shape\n",
            "(13360, 64, 64, 1) (13360, 38)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "bsZR7mb89UoI",
        "colab_type": "text"
      },
      "source": [
        "## Designing Model Architecture"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "7DImv-hIKSOf",
        "colab_type": "text"
      },
      "source": [
        "Now we will make a method which creates the model architecture with the specified optimizer and activation functions."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iHOTAM-w79Ak",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from keras.models import Sequential\n",
        "from keras.layers import Conv2D, MaxPooling2D, GlobalAveragePooling2D, BatchNormalization, Dropout, Dense\n",
        "\n",
        "def create_model(optimizer='adam', kernel_initializer='he_normal', activation='relu'):\n",
        "  # create model\n",
        "  model = Sequential()\n",
        "  model.add(Conv2D(filters=16, kernel_size=3, padding='same', input_shape=(64, 64, 1), kernel_initializer=kernel_initializer, activation=activation))\n",
        "  model.add(BatchNormalization())\n",
        "  model.add(MaxPooling2D(pool_size=2))\n",
        "  model.add(Dropout(0.2))\n",
        "\n",
        "  model.add(Conv2D(filters=32, kernel_size=3, padding='same', kernel_initializer=kernel_initializer, activation=activation))\n",
        "  model.add(BatchNormalization())\n",
        "  model.add(MaxPooling2D(pool_size=2))\n",
        "  model.add(Dropout(0.2))\n",
        "\n",
        "  model.add(Conv2D(filters=64, kernel_size=3, padding='same', kernel_initializer=kernel_initializer, activation=activation))\n",
        "  model.add(BatchNormalization())\n",
        "  model.add(MaxPooling2D(pool_size=2))\n",
        "  model.add(Dropout(0.2))\n",
        "\n",
        "  model.add(Conv2D(filters=128, kernel_size=3, padding='same', kernel_initializer=kernel_initializer, activation=activation))\n",
        "  model.add(BatchNormalization())\n",
        "  model.add(MaxPooling2D(pool_size=2))\n",
        "  model.add(Dropout(0.2))\n",
        "  model.add(GlobalAveragePooling2D())\n",
        "  \n",
        "  #Fully connected final layer\n",
        "  model.add(Dense(38, activation='softmax'))\n",
        "\n",
        "  # Compile model\n",
        "  model.compile(loss='categorical_crossentropy', metrics=['accuracy'], optimizer=optimizer)\n",
        "  return model"
      ],
      "execution_count": 24,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Wuqb1snmHPqW",
        "colab_type": "text"
      },
      "source": [
        "Let’s understand above code step by step.\n",
        "\n",
        "* The first hidden layer is a convolutional layer. The layer has 16 feature maps, which with the size of 3×3 and an activation function which is relu. This is the input layer, expecting images with the structure outlined above.\n",
        "* The second layer is Batch Normalization which solves having distributions of the features vary across the training and test data, which breaks the IID assumption. We use it to help in two ways faster learning and higher overall accuracy. \n",
        "* The third layer is the MaxPooling layer. MaxPooling layer is used to down-sample the input to enable the model to make assumptions about the features so as to reduce overfitting. It also reduces the number of parameters to learn, reducing the training time.\n",
        "* The next layer is a Regularization layer using dropout. It is configured to randomly exclude 20% of neurons in the layer in order to reduce overfitting.\n",
        "* Another hidden layer with 32 feature maps with the size of 3×3 and a relu activation function to capture more features from the image.\n",
        "* Other hidden layers with 64 and 128 feature maps with the size of 3×3 and a relu activation function to capture complex patterns from the image which will decribe the digits and letters later.\n",
        "* More MaxPooling, Batch Normalization, Regularization and GlobalAveragePooling2D layers.\n",
        "* The last layer is the output layer with 10 neurons (number of output classes) and it uses softmax activation function as we have multi-classes. Each neuron will give the probability of that class.\n",
        "\n",
        "I used categorical_crossentropy as a loss function because its a multi-class classification problem. I used accuracy as metrics to improve the performance of our neural network."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_GdoRPe6LiRV",
        "colab_type": "text"
      },
      "source": [
        "## Model Summary And Visualization"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "RLAADtoKMkcB",
        "colab_type": "text"
      },
      "source": [
        "Let's see the model summary."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GzgiwjyPLmrp",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 765
        },
        "outputId": "5fa3a42f-9e06-4798-bbd2-38d5011e9547"
      },
      "source": [
        "model = create_model()\n",
        "model.summary()"
      ],
      "execution_count": 25,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential_1\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d_1 (Conv2D)            (None, 64, 64, 16)        160       \n",
            "_________________________________________________________________\n",
            "batch_normalization_1 (Batch (None, 64, 64, 16)        64        \n",
            "_________________________________________________________________\n",
            "max_pooling2d_1 (MaxPooling2 (None, 32, 32, 16)        0         \n",
            "_________________________________________________________________\n",
            "dropout_1 (Dropout)          (None, 32, 32, 16)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_2 (Conv2D)            (None, 32, 32, 32)        4640      \n",
            "_________________________________________________________________\n",
            "batch_normalization_2 (Batch (None, 32, 32, 32)        128       \n",
            "_________________________________________________________________\n",
            "max_pooling2d_2 (MaxPooling2 (None, 16, 16, 32)        0         \n",
            "_________________________________________________________________\n",
            "dropout_2 (Dropout)          (None, 16, 16, 32)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_3 (Conv2D)            (None, 16, 16, 64)        18496     \n",
            "_________________________________________________________________\n",
            "batch_normalization_3 (Batch (None, 16, 16, 64)        256       \n",
            "_________________________________________________________________\n",
            "max_pooling2d_3 (MaxPooling2 (None, 8, 8, 64)          0         \n",
            "_________________________________________________________________\n",
            "dropout_3 (Dropout)          (None, 8, 8, 64)          0         \n",
            "_________________________________________________________________\n",
            "conv2d_4 (Conv2D)            (None, 8, 8, 128)         73856     \n",
            "_________________________________________________________________\n",
            "batch_normalization_4 (Batch (None, 8, 8, 128)         512       \n",
            "_________________________________________________________________\n",
            "max_pooling2d_4 (MaxPooling2 (None, 4, 4, 128)         0         \n",
            "_________________________________________________________________\n",
            "dropout_4 (Dropout)          (None, 4, 4, 128)         0         \n",
            "_________________________________________________________________\n",
            "global_average_pooling2d_1 ( (None, 128)               0         \n",
            "_________________________________________________________________\n",
            "dense_1 (Dense)              (None, 38)                4902      \n",
            "=================================================================\n",
            "Total params: 103,014\n",
            "Trainable params: 102,534\n",
            "Non-trainable params: 480\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "F-WXi7zkLxuN",
        "colab_type": "text"
      },
      "source": [
        "Keras support plotting the model in keras.utils.vis_utils module which provides utility functions to plot a Keras model using graphviz.\n",
        "\n",
        "To be able to use that utility we should first install pydot and graphviz modules. This can be done by running the following code cell then restart the runtime environment."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "uIL81Yib39V6",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# https://pypi.python.org/pypi/pydot\n",
        "!apt-get -qq install -y graphviz && pip install -q pydot\n",
        "import pydot"
      ],
      "execution_count": 26,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "eyEMpJx3MttP",
        "colab_type": "text"
      },
      "source": [
        "Now we can plot the model and save it to a file also."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "B1zpgDtwMtEU",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "2f49025d-be39-4595-9206-a31a8740ea19"
      },
      "source": [
        "from keras.utils import plot_model\n",
        "\n",
        "plot_model(model, to_file=\"model.png\", show_shapes=True)\n",
        "from IPython.display import Image as IPythonImage\n",
        "display(IPythonImage('model.png'))"
      ],
      "execution_count": 27,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAr8AAAgRCAYAAADnf4o+AAAABmJLR0QA/wD/AP+gvaeTAAAgAElEQVR4nOzdeVhU5f8//uewDvtiKJOIBihuuNNHUDPflpVcrqiQS28sDdFClkoFF0Sk1L7ARUK9S990laaIGPbOpS4rIktoUZEwDTFA1MQFBWGQ7f794W+mpgFk2IZxno/rmj865z7363XOGezF4T73LRFCCBARERERPfzSDLSdARERERFRV2HxS0RERER6g8UvEREREekNFr9EREREpDeMtJ0AtV1cXBxOnDih7TSIiIj0SlhYGLy8vLSdBrURn/zqsBMnTiA7O1vbaRBRByktLcX+/fu1nYbO2b9/P0pLS7WdBumJ/fv349KlS9pOg9qBT3513NixY5GWlqbtNIioA+zbtw9+fn78mdaQRCJBaGgo5s2bp+1USA9IJBJtp0DtxCe/RERERKQ3WPwSERERkd5g8UtEREREeoPFLxERERHpDRa/RERERKQ3WPwSET1kDh8+DBsbG/zvf//Tdird3rFjx7BmzRqkp6fDxcUFEokEEokEixYtUms7ZcoUWFlZwdDQEEOGDMHJkye1kLFm6urqEBsbCzc3N5iYmMDW1hZDhw5FUVFRs8fU1NRg4MCBWLt2rc7FBYDGxkbEx8fD29tbbd9nn32GLVu2oKGhoV0xSLex+CUiesgIIbSdgk7YsGEDEhMTERERAV9fX1y8eBGurq7o0aMHdu3ahUOHDqm0//LLL5GWloZp06YhPz8fo0aN0lLmrefn54ePPvoIu3fvRnV1NX777Te4urri7t27zR4TGRmJ8+fP62TcgoICPPHEEwgLC0N1dbXa/unTp0MqlWLy5Mm4fft2u2KR7uI8v0REDxkfHx/cuXNH22kAAORyOSZPnowffvhB26moeOutt7B3717k5uZCKpWq7EtMTMSiRYsQGBiI/Px82NjYaCnL9tm7dy8yMjKQm5sLDw8PAIBMJsPBgwebPeaHH37Ar7/+qpNxc3NzER0djaCgIFRVVTX7S+DKlStx8eJFTJ06FVlZWTAyYimkb/jkl4iIOs3OnTtRVlam7TRUXLhwAevWrcPGjRvVCl8A8Pb2RkhICC5fvozXXntNCxl2jHfffRejRo1SFqAPIpfL8frrryMhIUEn4w4fPhzp6elYsGABTE1NW2wbFRWF06dPtzsm6SYWv0RED5Hjx4/D2dkZEokE27dvBwAkJyfDwsIC5ubmOHjwIJ577jlYW1vDyckJe/bsUR6bmJgIqVSKnj17YtmyZZDJZJBKpfD29kZOTo6yXXBwMExMTODo6KjctmLFClhYWEAikeDGjRsAgJCQEISHh6OwsBASiQRubm4AgKNHj8La2hqbN2/uikuiJjExEUIITJ8+vdk2MTExGDBgAHbs2IFjx4612J8QAnFxcRg0aBBMTU1hZ2eHmTNn4ty5c8o2rb0HANDQ0ID169fD2dkZZmZmGDZsGFJTUzU6x9raWmRnZ2PEiBGtPiYyMhIrVqyAg4ODRrG6Q1xN2dnZYeLEiUhISOAwIT3E4peI6CEyfvx4tSEGy5cvR2hoKORyOaysrJCamorCwkK4uLhg6dKlqKurA3C/qA0ICEB1dTVWrlyJoqIinDx5EvX19Xj66adx6dIlAPeLx38uJZyUlISNGzeqbEtISMC0adPg6uoKIQQuXLgAAMqXjRobGzvlGjzIoUOH4O7uDnNz82bbmJmZ4cMPP4SBgQGWLl2KqqqqZttGRUVhzZo1iIyMRFlZGbKysnDp0iVMmDAB165dA9D6ewAAq1evxtatWxEfH4+rV69i2rRpmD9/Pn7++edWn+OVK1dQW1uLX375BZMmTVL+IjNo0CAkJSWpFXzff/89CgsLMX/+/FbH6E5x22LkyJG4fPkycnNzuzw2aReLXyIiPeLt7Q1ra2s4ODjA398fVVVVKCkpUWljZGSkfIo5ePBgJCcno7KyEikpKR2Sg4+PDyoqKrBu3boO6U8TVVVV+OOPP+Dq6vrAtl5eXggNDUVRURFWr17dZBu5XI64uDjMnj0bCxcuhI2NDTw8PPDee+/hxo0beP/999WOaeke1NTUIDk5GbNmzYKvry9sbW2xdu1aGBsba3T9FS+WOTg4YPPmzcjPz8e1a9cwc+ZMvPLKK/jkk09UziEkJATJycmt7r+7xW2L/v37AwDy8vK0Ep+0h8UvEZGeMjExAQCVp45NGTNmDMzNzVX+jK+rysrKIIRo8anv38XExMDd3R1JSUk4fvy42v78/HzcvXsXY8aMUdnu6ekJExMTleEiTfnnPTh//jyqq6sxdOhQZRszMzM4OjpqdP0VY16HDBkCb29v2Nvbw8bGBhs3boSNjY1KUR4REYGXX34ZvXv3bnX/3S1uWyi+A4qn86Q/WPwSEdEDmZqa4vr169pOo91qamoA4IEvRClIpVKkpKRAIpHgxRdfhFwuV9mvmC7L0tJS7VhbW1tUVlZqlJ9ieMXatWuVcw5LJBIUFxc3OXVXc2QyGQAox18rmJiYoG/fvigsLARwf4x4Xl4elixZolGe3S1uW5iZmQH46ztB+oPFLxERtaiurg63b9+Gk5OTtlNpN0XBo8kiB15eXggLC0NBQQE2bdqkss/W1hYAmixy23LNFC99xcfHQwih8jlx4kSr+7G0tET//v1x9uxZtX319fXK6dt27tyJr776CgYGBspCW5HD5s2bIZFINBprrK24bVFbWwvgr+8E6Q8Wv0RE1KLMzEwIITB27FjlNiMjowcOl+iOevbsCYlEovE8yJs2bcLAgQNx6tQple1Dhw6FpaWlWqGWk5OD2tpajB49WqM4ffr0gVQqxenTpzU6ril+fn44deoULl68qNxWXV2N4uJi5TRkKSkpakW24gl/ZGQkhBBqQzq6a1xNKb4DvXr16tQ41P2w+CUiIhWNjY0oLy9HfX09zpw5g5CQEDg7OyMgIEDZxs3NDbdu3UJGRgbq6upw/fp1FBcXq/Vlb2+PK1euoKioCJWVlairq8ORI0e0NtWZubk5XFxcUFpaqtFxiuEPhoaGatvDw8Nx4MAB7Nq1CxUVFcjLy0NQUBBkMhkCAwM1jrN48WLs2bMHycnJqKioQENDA0pLS3H16lUAgL+/P3r16vXA5ZXDwsLQt29fBAQEoKSkBDdv3sSqVasgl8ubfYGvJd09rqYU34HWzkdMDw8Wv0RED5Ht27fD09MTALBq1SrMmDEDycnJiI+PBwAMGzYMFy9exAcffIDw8HAAwLPPPouCggJlHzU1NfDw8ICZmRkmTJiAAQMG4JtvvlEZJ7t8+XJMmjQJzz//PNzd3bFp0ybln4+9vLyU06IFBQWhZ8+eGDx4MKZOnYpbt251yXVoiY+PD/Lz81XG73766adwc3NDYWEhPD098eqrr6odN3bsWISFhalt37BhA2JjYxEdHY1HHnkEEydORL9+/ZCZmQkLCwsA0OgeJCQkIDQ0FFu2bEGPHj0gk8kQEhKC8vJyAPf/XF9WVtbiimnA/blsv/vuOzg5OWHEiBHo3bs3fvzxRxw6dEijeXgVunvc7OxsjB8/Ho8++ihycnKQm5sLmUyGcePGISsrS639Tz/9hN69e2PYsGEa50Q6TpDOmjNnjpgzZ4620yCiDpKamiq0/c9yYGCgsLe312oOmgIgUlNTW92+oKBAGBkZiY8//rgTs+o8DQ0NYsKECWLnzp2M20Y3btwQUqlUvP322xofq+n3jbqdfXzyS0REKjR5GUwXubm5ITo6GtHR0cp5aXVFQ0MDMjIyUFlZCX9/f8Zto6ioKIwYMQLBwcEd1ifpDha/RESkd9asWYO5c+fC399f45fftCkzMxPp6ek4cuRIq+cqZlxVcXFxOH36NA4fPgxjY+MO6ZN0C4tf6haio6MxePBgWFtbw9TUFG5ubnjjjTce+FRmyZIlsLKygkQiadfb0Y2NjYiPj4e3t3eb+wCAw4cPw8bGBv/73//a1Y82ZWdnY9CgQcopiHr16oWYmBhtp6UiPT0dLi4uyimSHB0dsXDhQm2npfMiIiKQkpKCO3fu4LHHHsP+/fu1nVKn2rx5M4KDg/Hmm29qO5VWmzx5Mnbv3g1HR0fGbYODBw/i3r17yMzMhJ2dXYf0SbrHSNsJEAHA119/jVdeeQX+/v4wNjbGkSNHsHDhQuTl5eHIkSPNHrdjxw489dRTeP7559scu6CgAIsXL8b333+P4cOHt7kfAGrr1uuisWPH4rfffsOzzz6LL774AufPn1fOZdpd+Pr6wtfXF25ubrhx4wb+/PNPbaf0UIiNjUVsbKy20+hSU6ZMwZQpU7SdBnWRGTNmYMaMGdpOg7SMT36pW7C0tERgYCDs7e1hZWWFefPmYdasWTh69KjyrfHOkJubi9WrVyMoKKhNbyH/k4+PD+7cuYNp06Z1QHbtI5fL2/0ku7t4mM6FiIi0i8UvdQuff/652vyZjzzyCAA8cElPiUTS5rjDhw9Heno6FixY0OrlTnXFzp07UVZWpu00OsTDdC5ERKRdLH710Mcff4wxY8ZAKpXCwsIC/fr1Uy7ZKYRAXFwcBg0aBFNTU9jZ2WHmzJk4d+6c8vjk5GRYWFjA3NwcBw8exHPPPQdra2s4OTlhz549ynaDBg2CRCKBgYEBRo8erSxi33jjDdjY2EAqleLDDz9sNs/Lly/DzMwMjz32mHKbEALbtm2Du7s7TE1NYWNjg9dff72Dr1DbHD9+HM7OzpBIJNi+fTuA1l+rxMRESKVS9OzZE8uWLYNMJoNUKoW3tzdycnKU7YKDg2FiYqIy/m3FihWwsLCARCLBjRs3AAAhISEIDw9HYWEhJBIJ3NzcAABHjx5t8+IC3e1cNPXdd99h8ODByu+eh4cHvvjiCwD3x44rxg+7uroqV/FavHgxzM3NYWNjg88++wzA/bfP169fD2dnZ5iZmWHYsGFITU0FAGzduhXm5uawsrJCWVkZwsPD0bt3b5w/f75NORMRUSfQ8lxr1A5tmec3Pj5eABBvvvmmuHnzprh165b4z3/+IxYsWCCEEGL9+vXCxMREfPzxx+L27dvizJkzYtSoUeKRRx4Rf/75p7KfyMhIAUB89dVX4s6dO6KsrExMmDBBWFhYiNraWiGEEPX19aJfv37C2dlZ1NfXq+QRGhoq4uPjm82zqqpKWFlZieDgYJXtkZGRQiKRiP/3//6fKC8vF9XV1SIpKUkAEKdOndLoWvzT//3f/4nhw4e3q49Lly4JAOKdd95RyflB10qI+/OrWlhYiLNnz4qamhqRn58vPD09hZWVlSgpKVG2W7BggejVq5dK3G3btgkA4vr168ptvr6+wtXVVaXd559/LqysrER0dPQDz+WZZ54RAER5eXm3PBchhHB1dRU2NjYPPBchhEhLSxNRUVHi1q1b4ubNm2Ls2LGiR48eKjEMDQ3F5cuXVY6bP3+++Oyzz5T//dprrwlTU1Oxf/9+UV5eLiIiIoSBgYH46aefVK7RypUrxTvvvCNmz54tfvvtt1bl2B3m+dVF4Lyr1IX4fdN5nOdXn9TV1WHjxo2YNGkSVq9eDXt7e9jZ2eGll16Cp6cn5HI54uLiMHv2bCxcuBA2Njbw8PDAe++9hxs3buD9999X69Pb2xvW1tZwcHCAv78/qqqqUFJSAgAwNDTEypUrUVJSggMHDiiPqa6uRnp6Ol588cVmc42NjYVMJlOZZUAulyM+Ph5PPfUUwsLCYGtrCzMzM9jb23fgVeo8LV0rBSMjI+VT98GDByM5ORmVlZVISUnpkBx8fHxQUVGBdevWtauf7nAumpozZw42bNgAOzs72NvbY/r06bh58yauX78O4P5KZA0NDSr5VVRU4KeffsLUqVMB3F/5LDk5GbNmzYKvry9sbW2xdu1aGBsbq53XW2+9hVdeeQXp6ekYOHBg150oERG1iMWvHjlz5gxu376NZ555RmW7okjNz8/H3bt3MWbMGJX9np6eMDExUfmTdVNMTEwA3C+yFZYsWQIbGxskJCQot+3atQszZ86EtbV1k/0cOHAA+/btwxdffAErKyvl9gsXLqC6uhqTJ09u3Ql3Y01dq6aMGTMG5ubmKsNOuhtdPRfF/J6KBR3+9a9/YcCAAfjvf/+rnLVj79698Pf3V45HP3/+PKqrqzF06FBlP2ZmZnB0dOzQ81IMweCndR8A8PPz03oe/OjHh3QfpzrTIxUVFQDQ7LRVt2/fBnB/5oV/srW1RWVlpcYxLS0t8fLLL2Pbtm348ccf8fjjj+Pdd99tdv7QvXv3Ii4uDpmZmXj00UdV9pWWlgIAHBwcNM5Dl5mamiqfTuo6bZ7LoUOHsG3bNuTn56OiokKtWJdIJFi2bBnCwsLw1Vdf4amnnsJHH32E3bt3K9tUVVUBANauXYu1a9eqHC+TyTosV8UYYmodPz8/hISEwMvLS9upkB7w8/PTdgrUTix+9YiimFS8SPRPiqK4qSL39u3bcHJyalPc4OBgJCQkID4+HkFBQejTpw9cXV3V2r3zzjv44osv8PXXXzdZgEulUgDAvXv32pSHLqqrq2vXte9OuvpcsrKy8MsvvyA0NBQlJSWYNWsWZs+ejf/+97949NFH8c477+CNN95QOSYgIAARERHYsWMH+vTpA2tra/Tt21e5X/GLV3x8PEJCQjot93nz5nVa3w8jPz8/eHl58bpRl2Dxq/s47EGP9OvXD/b29vjyyy+b3D906FBYWlri559/Vtmek5OD2tpajB49uk1xnZycMG/ePOzfvx/r1q1TKxqEEFi1ahXy8vKQkZHRZOGryM/AwADffvttm/LQRZmZmRBCYOzYscptRkZGDxxi0B119bn88ssvsLCwAADk5eWhrq4Oy5cvh4uLC6RSaZN/vrSzs4Ofnx8yMjLw9ttvY+nSpSr7+/TpA6lU2q7VBImISLtY/OoRU1NTREREICsrC8HBwbh8+TIaGxtRWVmJs2fPQiqVIjw8HAcOHMCuXbtQUVGBvLw8BAUFQSaTITAwsM2xw8PDUV9fj/LycvzrX/9S2Xf27Fls3boVH3zwAYyNjdXGV7399tsA7j918/X1xf79+7Fz505UVFTgzJkzTb6Ip6saGxtRXl6O+vp6nDlzBiEhIXB2dkZAQICyjZubG27duoWMjAzU1dXh+vXrKC4uVuvL3t4eV65cQVFRESorK1FXV4cjR460eaqz7nYuzamrq8O1a9eQmZmpLH6dnZ0BAMeOHUNNTQ0KCgqaHcMeFBSEe/fu4fPPP1dbrEQqlWLx4sXYs2cPkpOTUVFRgYaGBpSWluLq1auaXiIiItIG7c42Qe3RlqnOhBBi+/btwsPDQ0ilUiGVSsXIkSNFUlKSEEKIxsZGsW3bNtG/f39hbGws7OzsxKxZs8T58+eVxyclJQlzc3MBQPTv318UFhaK999/X1hbWwsAom/fvuL3339Xiztp0iSxY8cOte15eXkCQLOfbdu2KdtWVlaKJUuWiB49eghLS0sxfvx4sX79egFAODk5idzcXI2uxYkTJ8S4ceOETCZTxnN0dBTe3t7i22+/1aivd955Rzg6OgoAwtzcXEyfPl2jaxUYGCiMjY1F7969hZGRkbC2thYzZ84UhYWFKnFu3rwpJk2aJKRSqXjsscfEq6++Kl5//XUBQLi5uSmnEjt58qTo27evMDMzE+PHjxd//vmnOHz4sLCyshIxMTHNnkd2drYYMmSIMDAwUF6PzZs3d6tzeffdd4Wrq2uL3xsA4sCBA8pYq1atEvb29sLW1lbMnTtXbN++XQAQrq6uKtOvCSHEyJEjxZo1a5q8Pvfu3ROrVq0Szs7OwsjISDg4OAhfX1+Rn58vtmzZIszMzAQA0adPH/Hxxx+35qujxKnO2gaceoq6EL9vOm+fRIj//7Vm0jlz584FAKSlpWk5E+oIy5YtQ1paGm7evKntVNpN18/Fx8cH27dvx98XWOkK+/btg5+fH/jPsmYkEglSU1M55pe6BL9vOi+Nwx6IuhHFtFsPA106l78Pozhz5gykUmmXF75ERNQ1WPzSQ+XcuXOtmqfR399fK/1R97Rq1SoUFBTg999/x+LFi5XLfdPD79ixY1izZg3S09Ph4uKi/JletGiRWtspU6bAysoKhoaGGDJkCE6ePKmFjDVTV1eH2NhYuLm5wcTEBLa2thg6dCiKioqaPaampgYDBw5Um85PF+IC9983iI+Ph7e3t9q+zz77DFu2bNGpX86p47H4pYfKwIEDIYR44Gfv3r1a6a85ERERSElJwZ07d/DYY481Ow+yLtDFczE3N8fAgQPx1FNPISoqCoMHD9Z2StQFNmzYgMTERERERMDX1xcXL16Eq6srevTogV27duHQoUMq7b/88kukpaVh2rRpyM/Px6hRo7SUeev5+fkp56uurq7Gb7/9BldXV9y9e7fZYyIjI3H+/HmdjFtQUIAnnngCYWFhqK6uVts/ffp0SKVSTJ48WTm3PekfFr9E3UBsbCzu3bsHIQT++OMPzJkzR9sptZkunktMTAwaGhpQUlKiNsODvpHL5U0+MdO1GA/y1ltvYe/evdi3b5/KSpIAkJiYCAMDAwQGBuLOnTtayrD99u7di4yMDKSlpeH//u//YGRkBJlMhoMHD6qsUvh3P/zwA3799VedjJubm4vVq1cjKCgII0aMaLbdypUrMXz4cEydOhX19fXtikm6icUvEREp7dy5E2VlZTofoyUXLlzAunXrsHHjRuXiOX/n7e2NkJAQXL58Ga+99poWMuwY7777LkaNGgUPD49WtZfL5Xj99ddVlqPXpbjDhw9Heno6FixYAFNT0xbbRkVF4fTp0+2OSbqJxS8RkQ4TQiAuLg6DBg2Cqakp7OzsMHPmTJw7d07ZJjg4GCYmJnB0dFRuW7FiBSwsLCCRSJSrPoaEhCA8PByFhYWQSCRwc3NDYmIipFIpevbsiWXLlkEmk0EqlcLb21tlruT2xACAo0ePdtkc1ImJiRBCYPr06c22iYmJwYABA7Bjxw4cO3asxf5acw+Sk5NhYWEBc3NzHDx4EM899xysra3h5OSEPXv2qPTX0NCA9evXw9nZGWZmZhg2bJjGS17X1tYiOzu7xSeg/xQZGYkVK1a0awl5bcXVlJ2dHSZOnIiEhATOrqKHWPwSEemwqKgorFmzBpGRkSgrK0NWVhYuXbqECRMm4Nq1awDuF3v/nJYpKSkJGzduVNmWkJCAadOmwdXVFUIIXLhwAcHBwQgICEB1dTVWrlyJoqIinDx5EvX19Xj66adx6dKldscA/podpLGxseMuTjMOHToEd3d3mJubN9vGzMwMH374IQwMDLB06VJUVVU127Y192D58uUIDQ2FXC6HlZUVUlNTUVhYCBcXFyxdulRlxpHVq1dj69atiI+Px9WrVzFt2jTMnz9fbfXNlly5cgW1tbX45ZdfMGnSJOUvLYMGDUJSUpJawff999+jsLAQ8+fPb3WM7hS3LUaOHInLly8jNze3y2OTdrH4JSLSUXK5HHFxcZg9ezYWLlwIGxsbeHh44L333sONGzc6dPVDIyMj5ZPNwYMHIzk5GZWVlUhJSemQ/n18fFBRUYF169Z1SH/Nqaqqwh9//AFXV9cHtvXy8kJoaCiKioqwevXqJtu05R54e3vD2toaDg4O8Pf3R1VVFUpKSgDcn/EgOTkZs2bNgq+vL2xtbbF27VoYGxtrdK0VL5Y5ODhg8+bNyM/Px7Vr1zBz5ky88sor+OSTT1TOISQkBMnJya3uv7vFbYv+/fsDuL/8OekXFr9ERDoqPz8fd+/exZgxY1S2e3p6wsTEpNklnDvCmDFjYG5urvKnfV1QVlYGIUSLT33/LiYmBu7u7khKSsLx48fV9rf3HpiYmAD4a67p8+fPo7q6WuXFMDMzMzg6Omp0rRVjXocMGQJvb2/Y29vDxsYGGzduhI2NjUpRHhERgZdffhm9e/dudf/dLW5bKL4DiqfzpD9Y/BIR6SjFVE2WlpZq+2xtbVFZWdmp8U1NTXH9+vVOjdHRampqAOCBL0QpSKVSpKSkQCKR4MUXX4RcLlfZ39H3QDG8Yu3atSpziRcXFzc5dVdzZDIZACjHWiuYmJigb9++KCwsBAAcP34ceXl5WLJkiUZ5dre4bWFmZgbgr+8E6Q8Wv0REOsrW1hYAmiywbt++DScnp06LXVdX1+kxOoOi4NFkkQMvLy+EhYWhoKBAbQGUjr4Hipe+4uPj1eYTP3HiRKv7sbS0RP/+/XH27Fm1ffX19bCxsQFwf+aNr776CgYGBspCW5HD5s2bIZFINBprrK24bVFbWwvgr+8E6Q8Wv0REOmro0KGwtLRUKxJycnJQW1uL0aNHK7cZGRmpvFTVXpmZmRBCYOzYsZ0WozP07NkTEolE4/l7N23ahIEDB+LUqVMq2zW5B63Rp08fSKVSnD59WqPjmuLn54dTp07h4sWLym3V1dUoLi5WTkOWkpKiVmQrnuZHRkZCCKE2pKO7xtWU4jvQq1evTo1D3Q+LXyIiHSWVShEeHo4DBw5g165dqKioQF5eHoKCgiCTyRAYGKhs6+bmhlu3biEjIwN1dXW4fv06iouL1fq0t7fHlStXUFRUhMrKSmUx29jYiPLyctTX1+PMmTMICQmBs7MzAgICOiTGkSNHumSqM3Nzc7i4uKC0tFSj4xTDHwwNDdW2t/YetDbO4sWLsWfPHiQnJ6OiogINDQ0oLS3F1atXAQD+/v7o1avXA5dXDgsLQ9++fREQEICSkhLcvHkTq1atglwub/YFvpZ097iaUnwHWjsfMT08WPwSEemwDRs2IDY2FtHR0XjkkUcwceJE9OvXD5mZmbCwsFC2W758OSZNmoTnn38e7u7u2LRpk/LPvV5eXsopy4KCgtCzZ08MHjwYU6dOxa1btwDcHxfp4eEBMzMzTJgwAQMGDMA333yjMna2vTG6io+PD/Lz81XG73766adwc3NDYWEhPD098eqrr6odN3bsWOxjmg0AACAASURBVISFhaltb809SE5ORnx8PABg2LBhuHjxIj744AOEh4cDAJ599lkUFBQAuD8dXGhoKLZs2YIePXpAJpMhJCQE5eXlAO7/ub6srAwHDx5s8Tzt7Ozw3XffwcnJCSNGjEDv3r3x448/4tChQxrNw6vQ3eNmZ2dj/PjxePTRR5GTk4Pc3FzIZDKMGzcOWVlZau1/+ukn9O7dG8OGDdM4J9JxgnTWnDlzxJw5c7SdBhF1kNTUVNEd/1kODAwU9vb22k6jWQBEampqq9sXFBQIIyMj8fHHH3diVp2noaFBTJgwQezcuZNx2+jGjRtCKpWKt99+W+NjNf2+Ubezj09+iYjogTR5Qay7c3NzQ3R0NKKjo5Xz0uqKhoYGZGRkoLKyEv7+/ozbRlFRURgxYgSCg4M7rE/SHSx+iYhI76xZswZz586Fv7+/xi+/aVNmZibS09Nx5MiRVs9VzLiq4uLicPr0aRw+fBjGxsYd0ifpFha/RETUrIiICKSkpODOnTt47LHHsH//fm2n1GE2b96M4OBgvPnmm9pOpdUmT56M3bt3w9HRkXHb4ODBg7h37x4yMzNhZ2fXIX2S7jHSdgJERNR9xcbGIjY2VttpdJopU6ZgypQp2k6DusiMGTMwY8YMbadBWsYnv0RERESkN1j8EhEREZHeYPFLRERERHqDxS8RERER6Q2+8KbjSktLsW/fPm2nQUQd4MSJEwDAn+k2UFw7IqIHkQghhLaToLaZO3fuQzXtEBERkS5ITU3FvHnztJ0GtU0ai18ioi62b98++Pn5gf/8EhF1uTSO+SUiIiIivcHil4iIiIj0BotfIiIiItIbLH6JiIiISG+w+CUiIiIivcHil4iIiIj0BotfIiIiItIbLH6JiIiISG+w+CUiIiIivcHil4iIiIj0BotfIiIiItIbLH6JiIiISG+w+CUiIiIivcHil4iIiIj0BotfIiIiItIbLH6JiIiISG+w+CUiIiIivcHil4iIiIj0BotfIiIiItIbLH6JiIiISG+w+CUiIiIivcHil4iIiIj0BotfIiIiItIbLH6JiIiISG+w+CUiIiIivcHil4iIiIj0BotfIiIiItIbLH6JiIiISG+w+CUiIiIivcHil4iIiIj0BotfIiIiItIbLH6JiIiISG+w+CUiIiIivWGk7QSIiB5mpaWl+Pe//42GhgbltvLyclhZWeHJJ59Uaevu7o7//Oc/XZwhEZF+YfFLRNSJnJycUFxcjMLCQrV93377rcp/P/HEE12VFhGR3uKwByKiTvbCCy/A2Nj4ge38/f27IBsiIv3G4peIqJMtWLAA9fX1LbYZMmQIBg8e3EUZERHpLxa/RESdzNXVFcOGDYNEImlyv7GxMf797393cVZERPqJxS8RURd44YUXYGho2OS++vp6zJ07t4szIiLSTyx+iYi6wPPPP4/Gxka17QYGBhg7diz69evX9UkREekhFr9ERF1AJpNh3LhxMDBQ/WfXwMAAL7zwgpayIiLSPyx+iYi6yKJFi9S2CSEwe/ZsLWRDRKSfWPwSEXWROXPmqIz7NTQ0xFNPPYWePXtqMSsiIv3C4peIqIvY2dnh6aefVhbAQggsXLhQy1kREekXFr9ERF1o4cKFyhffjI2NMXPmTC1nRESkX1j8EhF1oenTp8PU1BQAMG3aNFhaWmo5IyIi/cLil4ioC1lYWCif9nLIAxFR15MIIYS2kyDd0dwKVURERNowZ84cpKWlaTsN0h1pRtrOgHRPSEgIvLy8tJ0GUZvEx8cDAEJDQ7WWQ0NDA1JTUzF//nyt5aCJEydOICEhAampqdpOhUiF4ueZSBMsfkljXl5emDdvnrbTIGoTxRMibX+HZ82aBalUqtUcNJGQkKD1a0b0T3ziS23BMb9ERFqgS4UvEdHDhMUvEREREekNFr9EREREpDdY/BIRERGR3mDxS0RERER6g8UvEVEbHD58GDY2Nvjf//6n7VR00rFjx7BmzRqkp6fDxcUFEokEEokEixYtUms7ZcoUWFlZwdDQEEOGDMHJkye1kLFm6urqEBsbCzc3N5iYmMDW1hZDhw5FUVFRs8fU1NRg4MCBWLt2rc7FBYDGxkbEx8fD29tbbd9nn32GLVu2oKGhoV0xiDoCi18iojbg+kBtt2HDBiQmJiIiIgK+vr64ePEiXF1d0aNHD+zatQuHDh1Saf/ll18iLS0N06ZNQ35+PkaNGqWlzFvPz88PH330EXbv3o3q6mr89ttvcHV1xd27d5s9JjIyEufPn9fJuAUFBXjiiScQFhaG6upqtf3Tp0+HVCrF5MmTcfv27XbFImovzvNLRNQGPj4+uHPnjrbTAADI5XJMnjwZP/zwg7ZTeaC33noLe/fuRW5urtp0b4mJiVi0aBECAwORn58PGxsbLWXZPnv37kVGRgZyc3Ph4eEBAJDJZDh48GCzx/zwww/49ddfdTJubm4uoqOjERQUhKqqqmZ/MVy5ciUuXryIqVOnIisrC0ZGLEFIO/jkl4hIx+3cuRNlZWXaTuOBLly4gHXr1mHjxo1NznPs7e2NkJAQXL58Ga+99poWMuwY7777LkaNGqUsQB9ELpfj9ddfR0JCgk7GHT58ONLT07FgwQKYmpq22DYqKgqnT59ud0yi9mDxS0SkoePHj8PZ2RkSiQTbt28HACQnJ8PCwgLm5uY4ePAgnnvuOVhbW8PJyQl79uxRHpuYmAipVIqePXti2bJlkMlkkEql8Pb2Rk5OjrJdcHAwTExM4OjoqNy2YsUKWFhYQCKR4MaNGwDuLzceHh6OwsJCSCQSuLm5AQCOHj0Ka2trbN68uSsuSaskJiZCCIHp06c32yYmJgYDBgzAjh07cOzYsRb7E0IgLi4OgwYNgqmpKezs7DBz5kycO3dO2aa19wW4v+z0+vXr4ezsDDMzMwwbNkzjJZ1ra2uRnZ2NESNGtPqYyMhIrFixAg4ODhrF6g5xNWVnZ4eJEyciISGBQ4dIa1j8EhFpaPz48WpDDJYvX47Q0FDI5XJYWVkhNTUVhYWFcHFxwdKlS1FXVwfgflEbEBCA6upqrFy5EkVFRTh58iTq6+vx9NNP49KlSwDuF4r/XE44KSkJGzduVNmWkJCAadOmwdXVFUIIXLhwAQCULxY1NjZ2yjVoi0OHDsHd3R3m5ubNtjEzM8OHH34IAwMDLF26FFVVVc22jYqKwpo1axAZGYmysjJkZWXh0qVLmDBhAq5duwag9fcFAFavXo2tW7ciPj4eV69exbRp0zB//nz8/PPPrT7HK1euoLa2Fr/88gsmTZqk/OVm0KBBSEpKUiv4vv/+exQWFmL+/PmtjtGd4rbFyJEjcfnyZeTm5nZ5bCKAxS8RUYfz9vaGtbU1HBwc4O/vj6qqKpSUlKi0MTIyUj6xHDx4MJKTk1FZWYmUlJQOycHHxwcVFRVYt25dh/TXXlVVVfjjjz/g6ur6wLZeXl4IDQ1FUVERVq9e3WQbuVyOuLg4zJ49GwsXLoSNjQ08PDzw3nvv4caNG3j//ffVjmnpvtTU1CA5ORmzZs2Cr68vbG1tsXbtWhgbG2t0TxQvljk4OGDz5s3Iz8/HtWvXMHPmTLzyyiv45JNPVM4hJCQEycnJre6/u8Vti/79+wMA8vLytBKfiMUvEVEnMjExAQCVJ4xNGTNmDMzNzVX+ZP8wKSsrgxCixae+fxcTEwN3d3ckJSXh+PHjavvz8/Nx9+5djBkzRmW7p6cnTExMVIaQNOWf9+X8+fOorq7G0KFDlW3MzMzg6Oio0T1RjHkdMmQIvL29YW9vDxsbG2zcuBE2NjYqRXlERARefvll9O7du9X9d7e4baH4DiiezhN1NRa/RETdhKmpKa5fv67tNDpFTU0NADzwhSgFqVSKlJQUSCQSvPjii5DL5Sr7FdNlWVpaqh1ra2uLyspKjfJTDK9Yu3atcs5hiUSC4uLiJqfuao5MJgMA5ZhsBRMTE/Tt2xeFhYUA7o8bz8vLw5IlSzTKs7vFbQszMzMAf30niLoai18iom6grq4Ot2/fhpOTk7ZT6RSKgkeTRQ68vLwQFhaGgoICbNq0SWWfra0tADRZ5LblOipe+oqPj4cQQuVz4sSJVvdjaWmJ/v374+zZs2r76uvrldO37dy5E1999RUMDAyUhbYih82bN0MikWg01lhbcduitrYWwF/fCaKuxuKXiKgbyMzMhBACY8eOVW4zMjJ64HAJXdGzZ09IJBKN50betGkTBg4ciFOnTqlsHzp0KCwtLdUKtZycHNTW1mL06NEaxenTpw+kUilOnz6t0XFN8fPzw6lTp3Dx4kXlturqahQXFyunIUtJSVErshVP/SMjIyGEUBvS0V3jakrxHejVq1enxiFqDotfIiItaGxsRHl5Oerr63HmzBmEhITA2dkZAQEByjZubm64desWMjIyUFdXh+vXr6O4uFitL3t7e1y5cgVFRUWorKxEXV0djhw50q2mOjM3N4eLiwtKS0s1Ok4x/MHQ0FBte3h4OA4cOIBdu3ahoqICeXl5CAoKgkwmQ2BgoMZxFi9ejD179iA5ORkVFRVoaGhAaWkprl69CgDw9/dHr169Hri8clhYGPr27YuAgACUlJTg5s2bWLVqFeRyebMv8LWku8fVlOI70Nr5iIk6GotfIiINbd++HZ6engCAVatWYcaMGUhOTkZ8fDwAYNiwYbh48SI++OADhIeHAwCeffZZFBQUKPuoqamBh4cHzMzMMGHCBAwYMADffPONypjY5cuXY9KkSXj++efh7u6OTZs2Kf9U7OXlpZwWLSgoCD179sTgwYMxdepU3Lp1q0uug6Z8fHyQn5+vMn73008/hZubGwoLC+Hp6YlXX31V7bixY8ciLCxMbfuGDRsQGxuL6OhoPPLII5g4cSL69euHzMxMWFhYAIBG9yUhIQGhoaHYsmULevToAZlMhpCQEJSXlwO4/+f6srKyFldMA+7PZfvdd9/ByckJI0aMQO/evfHjjz/i0KFDGs3Dq9Dd42ZnZ2P8+PF49NFHkZOTg9zcXMhkMowbNw5ZWVlq7X/66Sf07t0bw4YN0zgnoo4gEZxlmjQgkUiQmpqqNv8oka6YO3cuACAtLU1rOSxbtgxpaWm4efOm1nLQxL59++Dn59fuRQkuXLiAQYMGISUlBQsXLuyg7LpOY2MjnnzySQQEBODFF19k3Da4efMmnJycEBMTo/wFpD26w88z6Zw0PvklItICTV78eli4ubkhOjoa0dHRynlpdUVDQwMyMjJQWVkJf39/xm2jqKgojBgxAsHBwR3WJ5GmWPySXoqOjsbgwYNhbW0NU1NTuLm54Y033njg/5CXLFkCKysrSCSSdr0Y09jYiPj4eHh7e7e5j787f/48Xn31VQwZMgRWVlYwMjKCjY0NBgwYAB8fH43eVu8srbnm6enpcHFxUZlqSiKRwMTEBD179sSTTz6Jbdu2Kf8MTbpnzZo1mDt3Lvz9/TV++U2bMjMzkZ6ejiNHjrR6rmLGVRUXF4fTp0/j8OHDMDY27pA+idpEEGkAgEhNTdV2Gu02ceJEkZSUJG7evCkqKipEamqqMDY2Fs8+++wDj92zZ48AIE6dOtWm2L///rsYN26cACCGDx/epj7+bseOHcLY2Fg88cQT4ujRo6K8vFzU1NSIwsJCsXfvXuHt7S3+85//tDtOe2lyzV1dXYWNjY0QQojGxkZRXl4uvvnmGxEQECAkEomQyWTip59+alMec+bMEXPmzGnXubTHmjVrhImJiQAg+vXrJ9LS0rSWS2ulpqaKjv7fxRdffCFWrVrVoX1S95WRkSFiY2NFfX19h/ar7Z9n0kn7jLRaeRNpiaWlJQIDA5VvkM+bNw/p6enYt28fLl26hD59+nRK3NzcXERHRyMoKAhVVVXtHkOZnZ2NwMBATJw4EV988QWMjP76kXZxcYGLiwtsbW1VXrTSlrZec4lEAltbWzz55JN48skn4ePjAz8/P/j4+OD3339Xzl+qK2JjYxEbG6vtNLRuypQpmDJlirbToC4yY8YMzJgxQ9tpEAHgsAfSU59//rna1EmPPPIIADxwNSeJRNLmuMOHD0d6ejoWLFjQ6pWuWhITE4OGhga8+eabKoXv3z3zzDN45ZVX2h2rvdpzzf9uzpw5CAgIQFlZGd57770OzZGIiB5+LH6p03388ccYM2YMpFIpLCws0K9fP+VqTUIIxMXFYdCgQTA1NYWdnR1mzpyJc+fOKY9PTk6GhYUFzM3NcfDgQTz33HOwtraGk5MT9uzZo2w3aNAgSCQSGBgYYPTo0cqC6o033oCNjQ2kUik+/PDDZvO8fPkyzMzM8Nhjjym3CSGwbds2uLu7w9TUFDY2Nnj99dc7+AqpO3r06APnaK2trcVXX32FHj164PHHH2913939mreGYi7cI0eOaHQcERERi1/qVAkJCXjhhRcwZ84cXLlyBaWlpYiIiMD58+cB3H/zd82aNYiMjERZWRmysrJw6dIlTJgwAdeuXQNwf67T0NBQyOVyWFlZITU1FYWFhXBxccHSpUuVK2D9+uuv6NevH/r06YMff/xR+ZLG1q1b8dJLL+Gtt95SWUDg76qrq/H1119j6dKlMDExUW5ft24dVq1ahcDAQFy7dg1//vlnmyaL15RiJoDGxsZm2xQXF6Ompgb9+/fXqO/ufs1bQzFn6d9XsiIiImoNFr/Uaerq6rBx40ZMmjQJq1evhr29Pezs7PDSSy/B09MTcrkccXFxmD17NhYuXAgbGxt4eHjgvffew40bN/D++++r9ent7Q1ra2s4ODjA398fVVVVKCkpAQAYGhpi5cqVKCkpwYEDB5THVFdXIz09vcV5KmNjYyGTyRATE6PcJpfLER8fj6eeegphYWGwtbWFmZkZ7O3tO/AqNc3HxwcVFRVYt25ds20qKioA3B9L21rd/Zq3lmLGjcrKSo2PJSIi/cYX3qjTnDlzBrdv38Yzzzyjsl1RMP3888+4e/eu2jrynp6eMDExQU5OTov9K54WKp5CAvenIouKikJCQoJy8vNdu3Zh5syZsLa2brKfAwcOYN++ffjyyy9hZWWl3H7hwgVUV1dj8uTJrT/pLqQoejUZL5ufn9+tr3lrKV4WbK7/ByktLcW+ffvadKw+UkyVx2tG3U1paSmcnJy0nQbpGBa/1GkUTyZtbW2b3H/79m0ATT+5tLW1bdNTPUtLS7z88svYtm0bfvzxRzz++ON49913sX///ibb7927F3FxccjMzMSjjz6qsk+x/ryDg4PGeXSFfv36QSqV4vfff2/1Md39mreW4pwHDhzYpuOzs7Ph5+fXpmP1Ga8ZdUdz5szRdgqkYzjsgTqNorC5ceNGk/sVRXFTBdft27fb/Nt8cHAwjI2NER8fj6ysLPTp0weurq5q7d555x3s2rULX3/9dZNFmFQqBQDcu3evTXl0NlNTUzzzzDO4ceMGvv/++2bb3bp1C0uWLAHQ/a95ax09ehQA8Nxzz7Xp+Dlz5kAIwU8rP6mpqQCg9Tz44eefHxa+1BYsfqnT9OvXD/b29vjyyy+b3D906FBYWlri559/Vtmek5OD2tpajB49uk1xnZycMG/ePOzfvx/r1q1DSEiIyn4hBFatWoW8vDxkZGQ0O2Z26NChMDAwwLffftumPLpCVFQUTE1NERYWBrlc3mSbX3/9VTkNWne/5q3x559/Ij4+Hk5OTi2OKSYiImoKi1/qNKampoiIiEBWVhaCg4Nx+fJlNDY2orKyEmfPnoVUKkV4eDgOHDiAXbt2oaKiAnl5eQgKCoJMJkNgYGCbY4eHh6O+vh7l5eX417/+pbLv7Nmz2Lp1Kz744AMYGxurLaX79ttvA7g/3MHX1xf79+/Hzp07UVFRgTNnzjT5UlhHO3LkyAOnOgPuz3qwe/du/Prrr5gwYQIOHz6MO3fuoK6uDn/88Qc++OADvPTSS8qlRLv7Nf87IQTu3r2LxsZGCCFw/fp1pKamYty4cTA0NERGRkabx/wSEZEeE0QaQBuWN96+fbvw8PAQUqlUSKVSMXLkSJGUlCSEuL907bZt20T//v2FsbGxsLOzE7NmzRLnz59XHp+UlCTMzc0FANG/f39RWFgo3n//fWFtbS0AiL59+4rff/9dLe6kSZPEjh071Lbn5eUJAM1+tm3bpmxbWVkplixZInr06CEsLS3F+PHjxfr16wUA4eTkJHJzczW6FidOnBDjxo0TMplMGc/R0VF4e3uLb7/9Vtnu8OHDwsrKSsTExLSq35KSEvHaa68JDw8PYWlpKQwNDYWtra0YOXKkeOmll8T333+vbNudr/lnn30mhg0bJszNzYWJiYkwMDAQAIREIhG2trbi8ccfF9HR0eLmzZutvub/xOVQNdcZyxsTdQT+PFMb7JMIIdq3virpFYlEgtTUVMybN0/bqRC1iWJGirS0NC1nojv27dsHPz8/8H8X1N3w55naII3DHoiIiIhIb7D4JWqHc+fOqY1fberj7++v7VSJiIgILH6J2mXgwIGtmo5n79692k6VSGuOHTuGNWvWID09HS4uLspfChctWqTWdsqUKbCysoKhoSGGDBmCkydPaiFjzdTV1SE2NhZubm4wMTGBra0thg4diqKiomaPqampwcCBA7F27VqdiwvcX3o9Pj4e3t7ebc7vs88+w5YtW5TLuRN1FRa/RETUaTZs2IDExERERETA19cXFy9ehKurK3r06IFdu3bh0KFDKu2//PJLpKWlYdq0acjPz8eoUaO0lHnr+fn54aOPPsLu3btRXV2N3377Da6urrh7926zx0RGRuL8+fM6GbegoABPPPEEwsLCWlxh8kH5TZ8+HVKpFJMnT1YuwEPUFbjCGxFRF5PL5Zg8eTJ++OEHnY7xIG+99Rb27t2L3Nxc5aIxComJiVi0aBECAwORn58PGxsbLWXZPnv37kVGRgZyc3Ph4eEBAJDJZDh48GCzx/zwww/49ddfdTJubm4uoqOjERQUpFxmvD35rVy5EhcvXsTUqVORlZWlnJOcqDPxyS8RURfbuXMnysrKdD5GSy5cuIB169Zh48aNaoUvAHh7eyMkJASXL1/Ga6+9poUMO8a7776LUaNGKQu8B5HL5Xj99deRkJCgk3GHDx+O9PR0LFiwAKamph2SX1RUFE6fPt3u3Ihai8UvEdEDCCEQFxeHQYMGwdTUFHZ2dpg5cybOnTunbBMcHAwTExM4Ojoqt61YsQIWFhaQSCTKZb5DQkIQHh6OwsJCSCQSuLm5ITExEVKpFD179sSyZcsgk8kglUrh7e2NnJycDokB3F8WujWLp3SExMRECCEwffr0ZtvExMRgwIAB2LFjB44dO9Zif625B8nJybCwsIC5uTkOHjyI5557DtbW1nBycsKePXtU+mtoaMD69evh7OwMMzMzDBs2TLmMc2vV1tYiOzsbI0aMaPUxkZGRWLFiBRwcHDSK1R3itpam+dnZ2WHixIlISEjgdHrUJVj8EhE9QFRUFNasWYPIyEiUlZUhKysLly5dwoQJE3Dt2jUA94u9f85/nZSUhI0bN6psS0hIwLRp0+Dq6gohBC5cuIDg4GAEBASguroaK1euRFFREU6ePIn6+no8/fTTuHTpUrtjAFC+WNTY2NhxF6cZhw4dgru7O8zNzZttY2Zmhg8//BAGBgZYunQpqqqqmm3bmnuwfPlyhIaGQi6Xw8rKCqmpqSgsLISLiwuWLl2Kuro6ZX+rV6/G1q1bER8fj6tXr2LatGmYP3++2tLfLbly5Qpqa2vxyy+/YNKkScpfWgYNGoSkpCS1Qu77779HYWEh5s+f3+oY3SluZ+UHACNHjsTly5eRm5vbJTmSfmPxS0TUArlcjri4OMyePRsLFy6EjY0NPDw88N577+HGjRsduty1kZGR8snm4MGDkZycjMrKSqSkpHRI/z4+PqioqMC6des6pL/mVFVV4Y8//oCrq+sD23p5eSE0NBRFRUVYvXp1k23acg+8vb1hbW0NBwcH+Pv7o6qqCiUlJQDuz3iQnJyMWbNmwdfXF7a2tli7di2MjY01utaKF7ccHBywefNm5Ofn49q1a5g5cyZeeeUVfPLJJyrnEBISguTk5Fb3393idkZ+Cv379wcA5OXldVmepL9Y/BIRtSA/Px93797FmDFjVLZ7enrCxMREZVhCRxszZgzMzc1V/rSvC8rKyiCEaPGp79/FxMTA3d0dSUlJOH78uNr+9t4DExMTAFA++T1//jyqq6sxdOhQZRszMzM4OjpqdK0VY16HDBkCb29v2Nvbw8bGBhs3boSNjY1KUR4REYGXX34ZvXv3bnX/3S1uZ+SnoPiuKJ7iE3UmFr9ERC1QTMFkaWmpts/W1haVlZWdGt/U1BTXr1/v1BgdraamBgBafCHq76RSKVJSUiCRSPDiiy9CLper7O/oe6AYXrF27VqVxWiKi4tbnLrrn2QyGQAox1ormJiYoG/fvigsLAQAHD9+HHl5eViyZIlGeXa3uB2d39+ZmZkB+Ou7Q9SZWPwSEbXA1tYWAJossG7fvg0nJ6dOi11XV9fpMTqDopDRZPECLy8vhIWFoaCgAJs2bVLZ19H3QPHSV3x8vNqCNCdOnGh1P5aWlujfvz/Onj2rtq++vl45fdvOnTvx1VdfwcDAQFloK3LYvHkzJBKJRmONtRW3o/P7u9raWgB/fXeIOhOLXyKiFgwdOhSWlpZqRUJOTg5qa2sxevRo5TYjIyOVl6raKzMzE0IIjB07ttNidIaePXtCIpHgzp07Gh23adMmDBw4EKdOnVLZrsk9aI0+ffpAKpXi9OnTGh3XFD8/P5w6dQoXL15UbquurkZxcbFymq+UlBS1IlvxND8yMhJCCLUhHd01bkfm93eK70qvXr06JR+iv2PxS0TUAqlUivDwcBw4cAC7du1CRUUF8vLyEBQUBJlMhsDAQGVbNzc39G55uwAAIABJREFU3Lp1CxkZGairq8P169dRXFys1qe9vT2uXLmCoqIiVFZWKovZxsZGlJeXo76+HmfOnEFISAicnZ0REBDQITGOHDnSJVOdmZubw8XFBaWlpRodpxj+YGhoqLa9tfegtXEWL16MPXv2IDk5GRUVFWhoaEBpaSmuXr0KAPD390evXr0euLxyWFgY+vbti4CAAJSUlODmzZtYtWoV5HJ5sy/wtaS7x20tTfNTfFdaO28xUXuw+CUieoANGzYgNjYW0dHReOSRRzBx4kT069cPmZmZsLCwULZbvnw5Jk2ahOeffx7u7u7YtGmT8s+4Xl5eyinLgoKC0LNnTwwePBhTp07FrVu3ANwf7+jh4QEzMzNMmDABAwYMwDfffKMydra9MbqKj48P8vPzVcbvfvrpp3Bzc0NhYSE8PT3x6quvqh03duxYhIWFqW1vzT1ITk5GfHw8AGDYsGG4ePEiPvjgA4SHhwMAnn32WRQUFAC4Px1caGgotmzZgh49ekAmkyEkJATl5eUA7v8ZvqysrMUV04D7c9R+9913cHJywogRI9C7d2/8+OOPOHTokEbz8Cp097jZ2dkYP348Hn30UeTk5CA3NxcymQzjxo1DVlZWm/P76aef0Lt3bwwbNkzj3Ik0JRGcUZo0IJFIkJqaqjbXKJGumDt3LgAgLS1Ny5moWrZsGdLS0nDz5k1tp6Jm37598PPz02gBggsXLmDQoEFISUnBwoULOzG7ztHY2Ignn3wSAQEBePHFFxm3E928eRNOTk6IiYlR/qLSWt3155m6tTQ++SUi6iY0eUGsu3Nzc0N0dDSio6OV877qioaGBmRkZKCyshL+/v6M28mioqIwYsQIBAcHd3ls0k8sfomIqFOsWbMGc+fOhb+/v8Yvv2lTZmYm0tPTceTIkVbPVcy4bRMXF4fTp0/j8OHDMDY27tLYpL9Y/BIRaVlERARSUlJw584dPPbYY9i/f7+2U+owmzdvRnBwMN58801tp9JqkydPxu7du+Ho6Mi4nejgwYO4d+8eMjMzYWdn16WxSb8ZaTsBIiJ9Fxsbi9jYWG2n0WmmTJmCKVOmaDsN6mZmzJiBGTNmaDsN0kN88ktEREREeoPFLxERERHpDRa/RERERKQ3WPwSERERkd7gC2+ksfj4eE4oTjorOzsbwF+T49ODKZae5TWj7iY7Oxtjx47VdhqkY7jCG2mE//Mjar8///wTp06d+v/Yu/e4qKr1f+Cf4TIM1wFFLkJggIooSiolqKnH8uvlaOIVbyfyVSFWiGCl4hWVMjvAQaFO5aFTmiJiWqIdj5VHzUuZioZpgOK1RFHkLgOs3x/+mBy5OAMMMzCf9+vlH+699n6e2bMGHvasvRZGjRql61SI2ryAgIB6l8QmakAai18iolbWlOWCiYioRXB5YyIiIiIyHCx+iYiIiMhgsPglIiIiIoPB4peIiIiIDAaLXyIiIiIyGCx+iYiIiMhgsPglIiIiIoPB4peIiIiIDAaLXyIiIiIyGCx+iYiIiMhgsPglIiIiIoPB4peIiIiIDAaLXyIiIiIyGCx+iYiIiMhgsPglIiIiIoPB4peIiIiIDAaLXyIiIiIyGCx+iYiIiMhgsPglIiIiIoPB4peIiIiIDAaLXyIiIiIyGCx+iYiIiMhgsPglIiIiIoPB4peIiIiIDAaLXyIiIiIyGCx+iYiIiMhgsPglIiIiIoPB4peIiIiIDAaLXyIiIiIyGCx+iYiIiMhgsPglIiIiIoPB4peIiIiIDIaJrhMgImrPFAoFSkpKVLaVlpYCAO7evauyXSKRwNbWttVyIyIyRCx+iYi06M6dO3BxcUF1dXWdfR06dFD5/7Bhw/Ddd9+1VmpERAaJwx6IiLTI0dERzz77LIyMGv9xK5FIMG3atFbKiojIcLH4JSLSslmzZj22jbGxMSZMmNAK2RARGTYWv0REWjZx4kSYmDQ8yszY2BgjR45Ex44dWzErIiLDxOKXiEjLbGxsMGrUqAYLYCEEZs6c2cpZEREZJha/REStYObMmfU+9AYAUqkUf/3rX1s5IyIiw8Til4ioFfz1r3+FhYVFne2mpqYICgqCpaWlDrIiIjI8LH6JiFqBTCbDhAkTYGpqqrJdoVBgxowZOsqKiMjwsPglImol06dPh0KhUNlmY2OD559/XkcZEREZHha/RESt5LnnnlNZ2MLU1BTTpk2DVCrVYVZERIaFxS8RUSsxMTHBtGnTlEMfFAoFpk+fruOsiIgMC4tfIqJWNG3aNOXQB0dHRwwaNEjHGRERGRYWv0RErSgwMBAuLi4AgL/97W+PXfaYiIhaVsNLDrUR27Zt03UKREQa8ff3x/Xr19GxY0f+DCOiNuWJJ55AQECArtNoFokQQug6ieaQSCS6ToGIiIjIIEyaNAlpaWm6TqM50tr8nV8ASE1NxZQpU3SdBhGR2rZv345JkybpOg2tmzx5MgC09V+WrWrbtm2YOnUq2vi9KWqHaj/PbR0HmxER6YAhFL5ERPqIxS8RERERGQwWv0RERERkMFj8EhEREZHBYPFLRERERAaDxS8RERERGQwWv0REpPf27NkDuVyOr7/+Wtep6L39+/dj0aJFSE9Ph4eHByQSCSQSCWbNmlWn7YgRI2BtbQ1jY2P07NkTJ0+e1EHGmlEoFIiNjYWXlxekUilsbW3Rq1cv5OXlNXhMRUUFvL29sWTJkjYXFwBqamoQHx+PwMDAJuf31VdfYe3ataiurm5WLu0Bi18iItJ7nPNWPcuXL0diYiIWL16MiRMn4uLFi/D09ETHjh2xadMmZGRkqLTft28f0tLSMHbsWGRlZaFv3746ylx9U6dOxWeffYbNmzejrKwMv/76Kzw9PVFSUtLgMdHR0bhw4UKbjJudnY1nn30WkZGRKCsra3J+48aNg0wmw/Dhw1FYWNisnNq6drHIBRERtW9jxozBvXv3dJ0GAKC8vBzDhw/HkSNHdJ2KinfffRdbt25FZmYmZDKZyr7ExETMmjULoaGhyMrKglwu11GWzbN161bs3LkTmZmZ8PX1BQA4Oztj165dDR5z5MgR/PLLL20ybmZmJmJiYhAWFobS0tIG/whUN7958+bh4sWLGD16NA4ePAgTE8MsA3nnl4iISAMbN25Efn6+rtNQkZOTg6VLl2LlypV1Cl8ACAwMREREBK5fv44FCxboIMOW8cEHH6Bv377KAu9xysvL8eabbyIhIaFNxu3Tpw/S09MxY8YMmJmZtUh+K1aswOnTp5udW1vG4peIiPTa4cOH4ebmBolEgg0bNgAAkpOTYWlpCQsLC+zatQujRo2CjY0NXF1dsWXLFuWxiYmJkMlkcHBwwJw5c+Ds7AyZTIbAwEAcP35c2S48PBxSqRROTk7Kba+99hosLS0hkUhw+/ZtAEBERASioqKQm5sLiUQCLy8vAMA333wDGxsbrFmzpjUuSR2JiYkQQmDcuHENtlm9ejW6deuGTz75BPv372/0fEIIxMXFoUePHjAzM4OdnR3Gjx+P8+fPK9uo+x4AQHV1NZYtWwY3NzeYm5ujd+/eSE1N1eg1VlZW4tixY/Dz81P7mOjoaLz22mvo1KmTRrH0Ia66NM3Pzs4OQ4YMQUJCgsEOJ2LxS0REem3QoEF1hhjMnTsX8+fPR3l5OaytrZGamorc3Fx4eHjglVdegUKhAPCgqA0JCUFZWRnmzZuHvLw8nDx5ElVVVXj++edx9epVAA+KxylTpqjESEpKwsqVK1W2JSQkYOzYsfD09IQQAjk5OQCgfIiopqZGK9fgcTIyMtC9e3dYWFg02Mbc3ByffvopjIyM8Morr6C0tLTBtitWrMCiRYsQHR2N/Px8HDx4EFevXsXgwYNx8+ZNAOq/BwCwcOFCvPfee4iPj8fvv/+OsWPHYvr06Thx4oTar/HGjRuorKzEzz//jGHDhin/kOnRoweSkpLqFHI//PADcnNzMX36dLVj6FNcbeUHAE899RSuX7+OzMzMVslR37D4JSKiNi0wMBA2Njbo1KkTgoODUVpaiitXrqi0MTExUd7F9PHxQXJyMoqLi5GSktIiOYwZMwZFRUVYunRpi5xPE6Wlpbh06RI8PT0f2zYgIADz589HXl4eFi5cWG+b8vJyxMXFYcKECZg5cybkcjl8fX3x4Ycf4vbt2/joo4/qHNPYe1BRUYHk5GQEBQVh4sSJsLW1xZIlS2BqaqrR9a99cKtTp05Ys2YNsrKycPPmTYwfPx6vv/46vvjiC5XXEBERgeTkZLXPr29xtZFfra5duwIAzp4922p56hMWv0RE1G5IpVIAULnrWJ/+/fvDwsJC5Wv8tio/Px9CiEbv+j5s9erV6N69O5KSknD48OE6+7OyslBSUoL+/furbPf394dUKlUZLlKfR9+DCxcuoKysDL169VK2MTc3h5OTk0bXv3bMa8+ePREYGIgOHTpALpdj5cqVkMvlKkX54sWL8eqrr8LFxUXt8+tbXG3kV6u2r9TexTc0LH6JiMggmZmZ4datW7pOo9kqKioAoNEHoh4mk8mQkpICiUSC2bNno7y8XGV/7TRYVlZWdY61tbVFcXGxRvnVDq9YsmSJcs5hiUSCy5cvNzp116OcnZ0BQDn+upZUKoW7uztyc3MBPBgjfvbsWbz88ssa5alvcVs6v4eZm5sD+LPvGBoWv0REZHAUCgUKCwvh6uqq61SarbaQ0WTxgoCAAERGRiI7OxurVq1S2WdrawsA9Ra5TblmtQ99xcfHQwih8u/o0aNqn8fKygpdu3bFuXPn6uyrqqpSTt+2ceNGfPvttzAyMlIW2rU5rFmzBhKJRKOxxrqK29L5PayyshLAn33H0LD4JSIig3PgwAEIITBgwADlNhMTk8cOl9BHDg4OkEgkGs+DvGrVKnh7e+PUqVMq23v16gUrK6s6hdrx48dRWVmJfv36aRTniSeegEwmw+nTpzU6rj5Tp07FqVOncPHiReW2srIyXL58WTnNV0pKSp0iu/YOf3R0NIQQdYZ06GvclszvYbV9xdHRUSv56DsWv0RE1O7V1NTg7t27qKqqwpkzZxAREQE3NzeEhIQo23h5eeHOnTvYuXMnFAoFbt26hcuXL9c5V4cOHXDjxg3k5eWhuLgYCoUCe/fu1dlUZxYWFvDw8MC1a9c0Oq52+IOxsXGd7VFRUdixYwc2bdqEoqIinD17FmFhYXB2dkZoaKjGcV566SVs2bIFycnJKCoqQnV1Na5du4bff/8dABAcHAxHR8fHLq8cGRkJd3d3hISE4MqVKygoKMDbb7+N8vLyBh/ga4y+x1WXpvnV9hV15y1ub1j8EhGRXtuwYQP8/f0BAG+//TZeeOEFJCcnIz4+HgDQu3dvXLx4ER9//DGioqIAACNHjkR2drbyHBUVFfD19YW5uTkGDx6Mbt264fvvv1cZJzt37lwMGzYM06ZNQ/fu3bFq1Srl18IBAQHKadHCwsLg4OAAHx8fjB49Gnfu3GmV69CYMWPGICsrS2X87pdffgkvLy/k5ubC398fb7zxRp3jBgwYgMjIyDrbly9fjtjYWMTExMDe3h5DhgxBly5dcODAAVhaWgKARu9BQkIC5s+fj7Vr16Jjx45wdnZGREQE7t69C+DB1/D5+fmNrpgGPJij9tChQ3B1dYWfnx9cXFzw448/IiMjQ6N5eGvpe9xjx45h0KBB6Ny5M44fP47MzEw4Oztj4MCBOHjwYJPz++mnn+Di4oLevXtrnHt7IBFtfIZjiUSC1NTUOvMzEhGR7k2ePBkAkJaWprMc5syZg7S0NBQUFOgsB01s27YNU6dO1WgBgpycHPTo0QMpKSmYOXOmFrPTjpqaGgwdOhQhISGYPXs242pRQUEBXF1dsXr1auUfKurSh89zC0jjnV8iImr3NHkYrC3y8vJCTEwMYmJilPO+thXV1dXYuXMniouLERwczLhatmLFCvj5+SE8PLzVY+sLgyp+/f39YWxs3KSvKJrrpZdegkwmg0QiMdipRR71/vvvKx/U+PDDD5Xb9+zZA7lcjq+//lqr8VsrjiZqamoQHx+PwMDAJp8jPT0dHh4eKlMKSSQSmJiYwN7eHs899xx27NjRglnXT90+/2i+s2bNqtNmxIgRsLa2hrGxMXr27Nli4+S0hX2bdGHRokWYPHkygoODNX74TZcOHDiA9PR07N27V+25ihm3aeLi4nD69Gns2bMHpqamrRpbnxhU8fvTTz9h2LBhOomdkpKCBQsW6CS2vlqwYEGdJUsBtNpa4/o24ic7OxvPPvssIiMjNZr78lETJ07ExYsX4enpCblcrvLUcWpqKq5fv46JEyciNTW1BbOvS90+/3C+HTt2xKZNm5CRkaHSZt++fUhLS8PYsWORlZWFvn37aivtFsG+rT8WL16MlJQU3Lt3D08++SS2b9+u65S0as2aNQgPD8c777yj61TUNnz4cGzevBlOTk6Mq0W7du3C/fv3ceDAAdjZ2bVqbH1jUMVvLYlE0uxzlJeXN+vuHDVszJgxuHfvHsaOHdti56zv/dJGnKbKzMzEwoULERYWprVvJuzs7DB8+HD84x//APBgXKEmWqPPJyYmwsjICKGhoW3qzpW6DLFv61psbCzu378PIQQuXbqESZMm6TolrRsxYgTeffddXadBeuaFF17AokWL6szuYYgMsvhtiVv9GzduRH5+fpOObYnimzTTnPerNfTp0wfp6emYMWOG2qs0NVWXLl0A/LmKk7pao88HBgYiIiIC169f5zclatL3vk1EpG8MsvjNycmBt7c3LC0tldPePLq++aFDh+Dj4wO5XA6ZTAZfX1/85z//AQBEREQgKioKubm5kEgk8PLyUh73+eefo3///pDJZLC0tESXLl1UVs8xMjJCRkYGRo0aBblcDmdnZ/zrX//S+DUkJyfD0tISFhYW2LVrF0aNGgUbGxu4urpiy5YtKm2FEIiLi0OPHj1gZmYGOzs7jB8/XmVN9ffeew8WFhawtrZGfn4+oqKi4OLigrCwMFhaWsLIyAj9+vWDo6MjTE1NYWlpib59+2Lw4MHKCcxtbW3x1ltvqX0d63P48GG4ublBIpFgw4YNAB68X4+OX63999///rdJ71d9cdS9Vppc+5b2zTffNHsu0TNnzgAAhgwZorJdX/r86tWr0a1bN3zyySfYv39/o6+Ffbv99G0iolYj2jgAIjU1Ve32w4cPFx4eHuLSpUtCoVCIX375RTzzzDNCJpOJ3377TdkuLS1NrFixQty5c0cUFBSIAQMGiI4dOyr3T5w4UXh6eqqcOz4+XgAQ77zzjigoKBB37twR//znP8WMGTOEEEJER0cLAOLbb78VhYWF4s6dO2L06NHCzMxMlJaWavzaHz7fvXv3RH5+vhg8eLCwtLQUlZWVynbLli0TUqlUfP7556KwsFCcOXNG9O3bV9jb24s//vijzvnmzZsn1q9fLyZMmCB+/fVXsXz5cgFAHD9+XJSWlorbt2+LkSNHCgAiIyND3Lp1S5SWlorw8HABQJw+fVrt65idnS0AiA8++EC57erVqwKAWL9+vbLNwoULldfo999/F3Z2diIwMFBUV1c3+f16NE5TrtXjrn1TPPPMM6JPnz717tu9e7ewtrYWMTExjz2Pp6enkMvlyv+XlZWJvXv3Cnd3dzFixAhRUlKi0l7Xfd7T01NcunRJCCHEkSNHhJGRkejSpYsyz71794oXXnhB5Rj2bf3v25MmTRKTJk3S+DhDlpqaKtrBr2dqh9rJ53lbm/90NaX4fbSwOHPmjAAgFixY0OBxsbGxAoDIz88XQtT9hVNZWSlsbW3FsGHDVI6rqqoSCQkJQog/f6mUl5cr93/22WcCgPjll1/Ufg216jtfUlKSACBycnKEEA8KHisrKxEcHKxy7I8//igAqBRR9Z1PCKEsEIqLi5Xb/v3vfwsA4uzZs3XOuXXr1gZzfvQ6qlMgPCooKEjIZDJx/vx5teOoUyA091o9eu2bqrHiVxOenp4CQJ1/vr6+4t///re4f/9+o8e3dp9/uPgVQoioqCgBQLz++utCiLrFL/t22+jb7eSXZati8Uv6qp18nreZNPfOcXvg6+sLuVyu/Dq4PrXjhBuaK/LMmTMoLCzE//3f/6lsNzY2xrx58x573pZaT14qlaqcLysrCyUlJXXWE/f394dUKsXx48ebFaeqqkq5TZ3X8rjr+Djbtm3Dl19+ibVr16J79+4tGqe51+rRa68P5HK5cmxvVVUVbt68iX379iE8PByxsbE4fPgw7O3t6z1W131+9erV2L17N5KSkjB16tQ6+9m3207fPnbsmHJyfHq82qVnec1I3xw7dgwDBgzQdRrNZpBjfutjamqq8oM9IyMDQ4cORadOnWBmZlZnvN+jioqKAAC2trZazVNTtYWPlZVVnX22trYoLi7WanxNr2NjCgoK8MYbb8Df37/OqjQtEUfX10rbTExM4OLigpdeegnvv/8+Lly4oDIdkr71eZlMhpSUFEgkEsyePVtl2VZA9+8X+zYRUdvEO794cIfnzp07cHNzAwBcuXIFQUFBmDBhAv71r3+hc+fOWL9+faO/dDp37gwAuH37dqvkrK7awqS+X26FhYVwdXXVWuymXMfGzJs3D4WFhfjuu+9UpmppqTi6vFatzdfXFwBw7tw5APrb5wMCAhAZGYn3338fq1atUn5GAfZtTei6bw8YMKCtL4faqmqXN+Y1I33TXr6N4J1fAN9//z1qamqUE+efPXsWCoUCc+fOhYeHh3KVqsZ06dIFHTp0wL59+1ojZbX16tULVlZWOHHihMr248ePo7KyEv369dNa7KZcx4ZkZGRg8+bNWLp0KXr27Knc/uabb7ZYHF1eq9b2888/A4Dy63V97vOrVq2Ct7c3Tp06pbKdfVt9htS3iYgexyCL38rKSty7dw9VVVU4efIkwsPD4e7ujpCQEABQ3l3av38/KioqkJ2dXWdMXIcOHXDjxg3k5eWhuLgYRkZGWLx4MQ4ePIjw8HBcv34dNTU1KC4uVt5d0wWZTIaoqCjs2LEDmzZtQlFREc6ePYuwsDA4OzsjNDRUa7HVuY7qKCoqwpw5c+Dn54eFCxcCACoqKnDixAmcPn26Se9XfWMXdXmt1LV3716NpzorLy9HTU0NhBC4ceMGUlJSsGTJEtjb22P+/PkA9LvP1w5/eHRidvbtB9pL3yYiajW6fuSuuaDhbA8pKSli2LBhwsHBQZiYmIiOHTuKadOmicuXL6u0e/vtt0WHDh2Era2tmDx5stiwYYMAIDw9PcWVK1fEyZMnhbu7uzA3NxeDBg1SThW0YcMG4evrK2QymZDJZOKpp54SSUlJYu3atcLc3FwAEF27dhW5ubli06ZNws7OTgAQrq6uGs34kJSUJCwsLFTO99FHHwkbGxsBQLi7uyunbqupqRHr1q0TXbt2FaampsLOzk4EBQWJCxcuKM/3cH5PPPGE+Pzzz4UQQiQkJCjjdOnSRRw6dEi8++67Qi6XCwDC0dFRbN68WWzdulU4OjoKAMLOzk5s2bLlsdcxIiJCeYylpaWYMGGCWL9+vXBychIAhIWFhRg3bpx4//336521AIAYPXp0k96vJUuW1Imj7rXS5Nqr6+jRo2LgwIHC2dlZ+dqcnJxEYGCg+N///qdst2fPHmFtbS1Wr17d4Ll27NjR4EwPZmZmomvXrmLu3LniypUrKsfpqs8/nK+9vb1ydodHvfnmm3WmOmPf1v++3U6eDm9VnO2B9FU7+TxvkwjRtheBl0gkSE1NxZQpU3SdChERPaJ2jCDHr6qvdsxvG//1TO1QO/k8pxnksAciIiIiMkwsfvXI+fPnG1zq9OF/wcHBuk6VHoPvJRHpyv79+7Fo0SKkp6fDw8ND+fNm1qxZddqOGDEC1tbWMDY2Rs+ePXHy5EkdZKwZhUKB2NhYeHl5QSqVwtbWFr169UJeXl6Dx1RUVMDb2xtLlixpc3EBoKamBvHx8QgMDGxyfl999RXWrl3b5LnI2xMWv3rE29sbQojH/tu6dauuU6XH4HtJRLqwfPlyJCYmYvHixZg4cSIuXrwIT09PdOzYEZs2bUJGRoZK+3379iEtLQ1jx45FVlaWctYjfTZ16lR89tln2Lx5M8rKyvDrr7/C09MTJSUlDR4THR2NCxcutMm42dnZePbZZxEZGYmysrIm5zdu3DjIZDIMHz5cOfe3oeI8v0RE1K6Vl5dj+PDhOHLkSJuO8Tjvvvsutm7diszMTMhkMpV9iYmJmDVrFkJDQ5GVlQW5XK6jLJtn69at2LlzJzIzM5XzlTs7O2PXrl0NHnPkyBH88ssvbTJuZmYmYmJiEBYWhtLS0gbHgaub37x583Dx4kWMHj0aBw8ehImJYZaBvPNLRETt2saNG5Gfn9/mYzQmJycHS5cuxcqVK+sUvgAQGBiIiIgIXL9+HQsWLNBBhi3jgw8+QN++fZUF3uOUl5fjzTffREJCQpuM26dPH6Snp2PGjBkwMzNrkfxWrFiB06dPNzu3tozFLxER6RUhBOLi4tCjRw+YmZnBzs4O48ePx/nz55VtwsPDIZVK4eTkpNz22muvwdLSEhKJRLnyYEREBKKiopCbmwuJRAIvLy8kJiZCJpPBwcEBc+bMgbOzM2QyGQIDA1XmUW5ODAD45ptvNJ6Xu6kSExMhhMC4ceMabLN69Wp069YNn3zyCfbv39/o+dR5D5KTk2FpaQkLCwvs2rULo0aNgo2NDVxdXbFlyxaV81VXV2PZsmVwc3ODubk5evfujdTUVI1eY2VlJY4dOwY/Pz+1j4mOjsZrr72GTp06aRRLH+KqS9P87OzsMGTIECQkJBjsjCIsfomISK+sWLECixYtQnR0NPLz83Hw4EFcvXoVgwcPxs2bNwE8KPYeneIyKSkJK1euVNmWkJCAsWPHwtPTE0II5OTkIDw8HCEhISiw0k+ZAAAgAElEQVQrK8O8efOQl5eHkydPoqqqCs8//zyuXr3a7BgAlA8W1dTUtNzFaUBGRga6d+8OCwuLBtuYm5vj008/hZGREV555RWUlpY22Fad92Du3LmYP38+ysvLYW1tjdTUVOTm5sLDwwOvvPKKyoIrCxcuxHvvvYf4+Hj8/vvvGDt2LKZPn15n1cHG3LhxA5WVlfj5558xbNgw5R8tPXr0QFJSUp1C7ocffkBubi6mT5+udgx9iqut/ADgqaeewvXr15GZmdkqOeobFr9ERKQ3ysvLERcXhwkTJmDmzJmQy+Xw9fXFhx9+iNu3b+Ojjz5qsVgmJibKO5s+Pj5ITk5GcXExUlJSWuT8Y8aMQVFREZYuXdoi52tIaWkpLl26BE9Pz8e2DQgIwPz585GXl6dcVfBRTXkPAgMDYWNjg06dOiE4OBilpaW4cuUKgAczHiQnJyMoKAgTJ06Era0tlixZAlNTU42ude2DW506dcKaNWuQlZWFmzdvYvz48Xj99dfxxRdfqLyGiIgIJCcnq31+fYurjfxqde3aFcCDpdoNEYtfIiLSG1lZWSgpKUH//v1Vtvv7+0MqlTZpGWl19e/fHxYWFipf7bcF+fn5EEI0etf3YatXr0b37t2RlJSEw4cP19nf3PdAKpUCgPLO74ULF1BWVoZevXop25ibm8PJyUmja1075rVnz54IDAxEhw4dIJfLsXLlSsjlcpWifPHixXj11Vfh4uKi9vn1La428qtV21dq7+IbGha/RESkN2qnYLKysqqzz9bWFsXFxVqNb2Zmhlu3bmk1RkurqKgAgEYfiHqYTCZDSkoKJBIJZs+ejfLycpX9Lf0e1A6vWLJkico855cvX2506q5HOTs7A4ByrHUtqVQKd3d35ObmAgAOHz6Ms2fP4uWXX9YoT32L29L5Pczc3BzAn33H0LD4JSIivWFrawsA9RZYhYWFcHV11VpshUKh9RjaUFvIaLJ4QUBAACIjI5GdnY1Vq1ap7Gvp96D2oa/4+Pg6c50fPXpU7fNYWVmha9euOHfuXJ19VVVVyunbNm7ciG+//RZGRkbKQrs2hzVr1kAikWg01lhXcVs6v4dVVlYC+LPvGBoWv0REpDd69eoFKyurOkXC8ePHUVlZiX79+im3mZiYqDxU1VwHDhyAEAIDBgzQWgxtcHBwgEQiwb179zQ6btWqVfD29sapU6dUtmvyHqjjiSeegEwmw+nTpzU6rj5Tp07FqVOncPHiReW2srIyXL58WTnNV0pKSp0iu/ZufnR0NIQQdYZ06GvclszvYbV9xdHRUSv56DsWv0REpDdkMhmioqKwY8cObNq0CUVFRTh79izCwsLg7OyM0NBQZVsvLy/cuXMHO3fuhEKhwK1bt3D58uU65+zQoQNu3LiBvLw8FBcXK4vZmpoa3L17F1VVVThz5gwiIiLg5uaGkJCQFomxd+/eVpnqzMLCAh4eHrh27ZpGx9UOfzA2Nq6zXd33QN04L730ErZs2YLk5GQUFRWhuroa165dw++//w4ACA4OhqOj42OXV46MjIS7uztCQkJw5coVFBQU4O2330Z5eXmDD/A1Rt/jqkvT/Gr7irrzFrc3LH6JiEivLF++HLGxsYiJiYG9vT2GDBmCLl264MCBA7C0tFS2mzt3LoYNG4Zp06ahe/fuWLVqlfJr3ICAAOWUZWFhYXBwcICPjw9Gjx6NO3fuAHgw3tHX1xfm5uYYPHgwunXrhu+//15l7GxzY7SWMWPGICsrS2X87pdffgkvLy/k5ubC398fb7zxRp3jBgwYgMjIyDrb1XkPkpOTER8fDwDo3bs3Ll68iI8//hhRUVEAgJEjRyI7OxvAg+ng5s+fj7Vr16Jjx45wdnZGREQE7t69C+DB1/D5+fmNrpgGPJij9tChQ3B1dYWfnx9cXFzw448/IiMjQ6N5eGvpe9xjx45h0KBB6Ny5M44fP47MzEw4Oztj4MCBOHjwYJPz++mnn+Di4oLevXtrnHt7IBFtfIZjiUSC1NTUOnMxEhGR7k2ePBkAkJaWpuNMVM2ZMwdpaWkoKCjQdSp1bNu2DVOnTtVoAYKcnBz06NEDKSkpmDlzphaz046amhoMHToUISEhmD17NuNqUUFBAVxdXbF69WrlHyrq0tfPs4bSeOeXiIgMkiYPiOk7Ly8vxMTEICYmRjnva1tRXV2NnTt3ori4GMHBwYyrZStWrICfnx/Cw8NbPba+YPFLRETUDixatAiTJ09GcHCwxg+/6dKBAweQnp6OvXv3qj1XMeM2TVxcHE6fPo09e/bA1NS0VWPrExa/RERkUBYvXoyUlBTcu3cPTz75JLZv367rlFrMmjVrEB4ejnfeeUfXqaht+PDh2Lx5M5ycnBhXi3bt2oX79+/jwIEDsLOza9XY+sZE1wkQERG1ptjYWMTGxuo6Da0ZMWIERowYoes0SM+88MILeOGFF3Sdhl7gnV8iIiIiMhgsfomIiIjIYLD4JSIiIiKDweKXiIiIiAwGi18iIiIiMhjtYoU3IiIiItK+SZMmtfkV3tr8VGepqam6ToGISCNHjx5FQkICf34RUZvzxBNP6DqFZmvzd36JiNqabdu2YerUqeCPXyKiVpfGMb9EREREZDBY/BIRERGRwWDxS0REREQGg8UvERERERkMFr9EREREZDBY/BIRERGRwWDxS0REREQGg8UvERERERkMFr9EREREZDBY/BIRERGRwWDxS0REREQGg8UvERERERkMFr9EREREZDBY/BIRERGRwWDxS0REREQGg8UvERERERkMFr9EREREZDBY/BIRERGRwWDxS0REREQGg8UvERERERkMFr9EREREZDBY/BIRERGRwWDxS0REREQGg8UvERERERkMFr9EREREZDBY/BIRERGRwWDxS0REREQGg8UvERERERkMFr9EREREZDBY/BIRERGRwWDxS0REREQGg8UvERERERkME10nQETUnt26dQtffvmlyrYTJ04AAD766COV7dbW1pg2bVqr5UZEZIgkQgih6ySIiNqr+/fvw8HBASUlJTA2NgYA1P7YlUgkynYKhQIvvvgiPv30U12kSURkKNI47IGISIvMzMwwadIkmJiYQKFQQKFQoKqqClVVVcr/KxQKAMD06dN1nC0RUfvH4peISMumT5+OysrKRtvY2triL3/5SytlRERkuFj8EhFp2bBhw9CpU6cG95uammLmzJkwMeFjGERE2sbil4hIy4yMjDBjxgyYmprWu1+hUPBBNyKiVsLil4ioFUybNk05tvdRnTt3RkBAQCtnRERkmFj8EhG1gqeffhru7u51tkulUrz44osqMz8QEZH2sPglImols2bNqjP0obKykkMeiIhaEYtfIqJWMmPGjDpDH7y8vODr66ujjIiIDA+LXyKiVuLt7Q0fHx/lEAdTU1O89NJLOs6KiMiwsPglImpFf/vb35QrvVVVVXHIAxFRK2PxS0TUiqZNm4bq6moAQN++ffHkk0/qOCMiIsPC4peIqBW5ubnhmWeeAQC8+OKLOs6GiMjwcDmhFhAXF4ejR4/qOg0iaiPu378PiUSCffv24eDBg7pOh4jaiMjISM4J3gJ457cFHD16FMeOHdN1GkTURri6usLR0REymUyrcbZv345r165pNUZ7c+zYMf48J720fft2XL16VddptAu889tCBgwYgLS0NF2nQURtRE5ODry8vLQaQyKRYP78+ZgyZYpW47QnkydPBgD+PCe9w4VwWg7v/BIR6YC2C18iIqofi18iIiIiMhgsfomIiIjIYLD4JSIiIiKDweKXiIiIiAwGi18iImrUnj17IJfL8fXXX+s6Fb23f/9+LFq0COnp6fDw8IBEIoFEIsGsWbPqtB0xYgSsra1hbGyMnj174uTJkzrIWDMKhQKxsbHw8vKCVCqFra0tevXqhby8vAaPqaiogLe3N5YsWdLm4gJATU0N4uPjERgY2OT8vvrqK6xdu1a5uiPpFotfIiJqlBBC1ym0CcuXL0diYiIWL16MiRMn4uLFi/D09ETHjh2xadMmZGRkqLTft28f0tLSMHbsWGRlZaFv3746ylx9U6dOxWeffYbNmzejrKwMv/76Kzw9PVFSUtLgMdHR0bhw4UKbjJudnY1nn30WkZGRKCsra3J+48aNg0wmw/Dhw1FYWNisnKj5OM8vERE1asyYMbh3756u0wAAlJeXY/jw4Thy5IiuU1Hx7rvvYuvWrcjMzKyzeEliYiJmzZqF0NBQZGVlQS6X6yjL5tm6dSt27tyJzMxM+Pr6AgCcnZ2xa9euBo85cuQIfvnllzYZNzMzEzExMQgLC0NpaWmDfwSqm9+8efNw8eJFjB49GgcPHoSJCUswXeGdXyIiajM2btyI/Px8XaehIicnB0uXLsXKlSvrXbUvMDAQERERuH79OhYsWKCDDFvGBx98gL59+yoLvMcpLy/Hm2++iYSEhDYZt0+fPkhPT8eMGTNgZmbWIvmtWLECp0+fbnZu1DwsfomIqEGHDx+Gm5sbJBIJNmzYAABITk6GpaUlLCwssGvXLowaNQo2NjZwdXXFli1blMcmJiZCJpPBwcEBc+bMgbOzM2QyGQIDA3H8+HFlu/DwcEilUjg5OSm3vfbaa7C0tIREIsHt27cBABEREYiKikJubi4kEolyoZBvvvkGNjY2WLNmTWtckjoSExMhhMC4ceMabLN69Wp069YNn3zyCfbv39/o+YQQiIuLQ48ePWBmZgY7OzuMHz8e58+fV7ZR9z0AgOrqaixbtgxubm4wNzdH7969kZqaqtFrrKysxLFjx+Dn56f2MdHR0XjttdfQqVMnjWLpQ1x1aZqfnZ0dhgwZgoSEBA4n0iEWv0RE1KBBgwbVGWIwd+5czJ8/H+Xl5bC2tkZqaipyc3Ph4eGBV155BQqFAsCDojYkJARlZWWYN28e8vLycPLkSVRVVeH555/H1atXATwoHh9dgjkpKQkrV65U2ZaQkICxY8fC09MTQgjk5OQAgPIhopqaGq1cg8fJyMhA9+7dYWFh0WAbc3NzfPrppzAyMsIrr7yC0tLSBtuuWLECixYtQnR0NPLz83Hw4EFcvXoVgwcPxs2bNwGo/x4AwMKFC/Hee+8hPj4ev//+O8aOHYvp06fjxIkTar/GGzduoLKyEj///DOGDRum/EOmR48eSEpKqlPI/fDDD8jNzcX06dPVjqFPcbWVHwA89dRTuH79OjIzM1slR6qLxS8RETVZYGAgbGxs0KlTJwQHB6O0tBRXrlxRaWNiYqK8i+nj44Pk5GQUFxcjJSWlRXIYM2YMioqKsHTp0hY5nyZKS0tx6dIleHp6PrZtQEAA5s+fj7y8PCxcuLDeNuXl5YiLi8OECRMwc+ZMyOVy+Pr64sMPP8Tt27fx0Ucf1TmmsfegoqICycnJCAoKwsSJE2Fra4slS5bA1NRUo+tf++BWp06dsGbNGmRlZeHmzZsYP348Xn/9dXzxxRcqryEiIgLJyclqn1/f4mojv1pdu3YFAJw9e7bV8iRVLH6JiKhFSKVSAFC561if/v37w8LCQuVr/LYqPz8fQohG7/o+bPXq1ejevTuSkpJw+PDhOvuzsrJQUlKC/v37q2z39/eHVCpVGS5Sn0ffgwsXLqCsrAy9evVStjE3N4eTk5NG1792zGvPnj0RGBiIDh06QC6XY+XKlZDL5SpF+eLFi/Hqq6/CxcVF7fPrW1xt5Fertq/U3sWn1sfil4iIWp2ZmRlu3bql6zSaraKiAgAafSDqYTKZDCkpKZBIJJg9ezbKy8tV9tdOg2VlZVXnWFtbWxQXF2uUX+3wiiVLlijnHJZIJLh8+XKjU3c9ytnZGQCU469rSaVSuLu7Izc3F8CDMeJnz57Fyy+/rFGe+ha3pfN7mLm5OYA/+w61Pha/RETUqhQKBQoLC+Hq6qrrVJqttpDRZPGCgIAAREZGIjs7G6tWrVLZZ2trCwD1FrlNuWa1D33Fx8dDCKHy7+jRo2qfx8rKCl27dsW5c+fq7KuqqlJO37Zx40Z8++23MDIyUhbatTmsWbMGEolEo7HGuorb0vk9rLKyEsCffYdaH4tfIiJqVQcOHIAQAgMGDFBuMzExeexwCX3k4OAAiUSi8TzIq1atgre3N06dOqWyvVevXrCysqpTqB0/fhyVlZXo16+fRnGeeOIJyGQynD59WqPj6jN16lScOnUKFy9eVG4rKyvD5cuXldN8paSk1Cmya+/wR0dHQwhRZ0iHvsZtyfweVttXHB0dtZIPPR6LXyIi0qqamhrcvXsXVVVVOHPmDCIiIuDm5oaQkBBlGy8vL9y5cwc7d+6EQqHArVu3cPny5Trn6tChA27cuIG8vDwUFxdDoVBg7969OpvqzMLCAh4eHrh27ZpGx9UOfzA2Nq6zPSoqCjt27MCmTZtQVFSEs2fPIiwsDM7OzggNDdU4zksvvYQtW7YgOTkZRUVFqK6uxrVr1/D7778DAIKDg+Ho6PjY5ZUjIyPh7u6OkJAQXLlyBQUFBXj77bdRXl7e4AN8jdH3uOrSNL/avqLuvMXU8lj8EhFRgzZs2AB/f38AwNtvv40XXngBycnJiI+PBwD07t0bFy9exMcff4yoqCgAwMiRI5Gdna08R0VFBXx9fWFubo7BgwejW7du+P7771XGyc6dOxfDhg3DtGnT0L17d6xatUr5tXBAQIByWrSwsDA4ODjAx8cHo0ePxp07d1rlOjRmzJgxyMrKUhm/++WXX8LLywu5ubnw9/fHG2+8Uee4AQMGIDIyss725cuXIzY2FjExMbC3t8eQIUPQpUsXHDhwAJaWlgCg0XuQkJCA+fPnY+3atejYsSOcnZ0RERGBu3fvAnjwNXx+fn6jK6YBD+aoPXToEFxdXeHn5wcXFxf8+OOPyMjI0Gge3lr6HvfYsWMYNGgQOnfujOPHjyMzMxPOzs4YOHAgDh482OT8fvrpJ7i4uKB3794a504tQyI4y3KzTZ48GQCQlpam40yIiP4kkUiQmppaZw7d1jRnzhykpaWhoKBAZzlooik/z3NyctCjRw+kpKRg5syZ2kpNa2pqajB06FCEhIRg9uzZjKtFBQUFcHV1xerVq5V/qKhLHz7P7UQa7/wSEZFWafIwWFvk5eWFmJgYxMTEKOd9bSuqq6uxc+dOFBcXIzg4mHG1bMWKFfDz80N4eHirx6Y/sfglIiJqpkWLFmHy5MkIDg7W+OE3XTpw4ADS09Oxd+9etecqZtymiYuLw+nTp7Fnzx6Ympq2amxSxeKX9NL777+vfIr6ww8/VG7fs2cP5HI5vv76a63FjomJgY+PD2xsbGBmZgYvLy+89dZbj72j8/LLL8Pa2hoSiaRZT1bX1NQgPj4egYGBTT5Heno6PDw8lFP+PG7lq7i4OEgkEhgZGcHb21tlPFtzPZqLRCKBqakpXFxcMGPGDPz6668tFutR+t6P6rs2EokEUqkUDg4OGDp0KNatW6ccm9nWLF68GCkpKbh37x6efPJJbN++XdcpadWaNWsQHh6Od955R9epqG348OHYvHkznJycGFeLdu3ahfv37+PAgQOws7Nr1dhUD0HNNmnSJDFp0iRdp9HuZGdnCwDigw8+UG7bvXu3sLGxEV999ZXW4g4ZMkQkJSWJgoICUVRUJFJTU4WpqakYOXLkY4/dsmWLACBOnTrVpNi//fabGDhwoAAg+vTp06RzPMzT01MAEE5OTqKysrLeNlVVVcLd3V0AEMOHD292zMZykcvlQgghSkpKxFdffSXc3NyElZWVOH/+vNbitoV+9PC1qampEXfv3hXff/+9CAkJERKJRDg7O4uffvpJ4xwAiNTU1Ga/FkPCn+ekr/h5bjHbeOeX2pQxY8bg3r17GDt2rNZiWFlZITQ0FB06dIC1tTWmTJmCoKAgfPPNN8onzrUhMzMTCxcuRFhYWJOeYG5Iv3798Mcff2Dnzp317k9PT2/V5UABwNLSEmPHjsU//vEPlJSUYP369a0aX5/7kUQiga2tLYYOHYqUlBRs27YNN2/eVOZMRETNw+KXDJoQAmlpaSrrr+/evbvO3Jv29vYA8NjlQCUSSZNz6dOnD9LT0zFjxgy1l0pVx9y5cwEAH3zwQb374+LiNH7quKU8/fTTAIBffvlFJ/FbSkv3o4dNmjQJISEhyM/PVxm6QURETcPiVwcSEhJgaWkJIyMj9OvXD46OjjA1NYWlpSX69u2LwYMHK1flsbW1xVtvvaVy/KFDh+Dj4wO5XA6ZTAZfX1/85z//AQB8+umnsLKygkQigZ2dHXbu3IkTJ07A3d0dxsbGmD59uka5JiYmQiaTwcHBAXPmzIGzszNkMhkCAwNx/PhxlbZCCMTFxaFHjx4wMzODnZ0dxo8fj/Pnzzep3aMOHz4MNzc3SCQSbNiwAcCDuS4tLS1hYWGBXbt2YdSoUbCxsYGrqyu2bNmicnx1dTViY2PRvXt3mJubw97eHk8++SRiY2MfO3XM9evXYW5ujieffFLldaxbtw7du3eHmZkZ5HI53nzzzcde0+b65ptvNJrQ/y9/+Qt69OiB77//HhcuXFDZ98MPP6CsrAwjRoyo91ht97WqqioAUCn2Da0fqaN2MYi9e/dqdBwREdVDh2Mu2o2mjBFbvny5ACCOHz8uSktLxe3bt8XIkSMFAJGRkSFu3bolSktLRXh4uAAgTp8+rTw2LS1NrFixQty5c0cUFBSIAQMGiI4dOyr3nzt3TlhYWIgXX3xRuW3RokXik08+adLrCw0NFZaWluLcuXOioqJCZGVlCX9/f2FtbS2uXLmibLds2TIhlUrF559/LgoLC8WZM2dE3759hb29vfjjjz80blffWM2rV68KAGL9+vXKbdHR0QKA+Pbbb8W9e/dEfn6+GDx4sLC0tFQZ57pmzRphbGwsdu3aJcrKysTPP/8sHB0dxdChQxt9/aWlpcLa2lqEh4erbI+OjhYSiUT8/e9/F3fv3hVlZWUiKSmpWWN+az3zzDMNjvndvXu3sLa2FjExMY89j6enp7h06ZL4xz/+IQCIiIgIlf1BQUEiJSVFFBcX1zvmtyX72sPjWmt9/vnnAoB48803ldsMrR81dG0eVlRUJACIJ554otEYjwLHCGqMY35JX/Hz3GK2sfhtAc0pfouLi5Xb/v3vfwsA4uzZs8ptP/74owAgtm7d2uC5YmNjBQCRn5+v3PbPf/5TABCbNm0SX3zxhYiMjNQov4eFhobW+cX8008/CQBi5cqVQgghysrKhJWVlQgODlZpV5t/baGmbjshNC9aysvLldtqi9CcnBzlNn9/f/H000+rxH311VeFkZGRuH//foOvPzo6WnTr1k0UFRUpt5WVlQkLCwvx/PPPq7Rt7gNvtRorfjVRW/wWFhYKS0tLYWdnJ8rKyoQQQuTm5gpXV1dx//79BovfRzWnrz36wNv27duFo6OjcHBwENeuXRNCGF4/qu/aNEQikQhbW9tG2zyKvyw1x+KX9BU/zy1mm4n27imTpqRSKYA/vwoGoJwLUKFQNHhcbZuHJ5J/9dVX8d///hdz5szBc8891+JTDPXv3x8WFhbKr5izsrJQUlKC/v37q7Tz9/eHVCpVDpFQt11z1V7Lh69bRUUFZDKZSrvq6mqYmprWGZtZa8eOHdi2bRv27dsHa2tr5facnByUlZVh+PDhLZKvtsnlckyfPh0ff/wxtm7dipdeegnx8fGYO3cupFIpKisr1TpPc/vavXv3IJFIYGxsDCcnJ4wePRrLly9XPnBnaP1IXaWlpRBCwMbGRuNjp06diqlTp2p8nKFrzvh9ItJvLH7boIyMDKxbtw5ZWVkoKipqsDBes2YNtm/fjvz8fK3kYWZmhlu3bgEACgsLATx4wv1Rtra2KC4u1qidNowePRrr1q3Drl27MGLECGRlZWHnzp3461//Wm/RsnXrVsTFxeHAgQPo3Lmzyr5r164BADp16qS1fFva3Llz8fHHH+PDDz9EUFAQ0tLSHjvHbkv3NblcruwD9TG0fqSu3377DQDg7e2t8bEREREICAhoUlxDFB8fDwCYP3++jjMhUsU/YlsOi9825sqVKwgKCsKECRPwr3/9C507d8b69evrPBSnUCgwb9485ZP8q1evxvLly1ssD4VCgcLCQri6ugJ4UHAAqLfoaEo7bVixYgV+/vlnhISEoKSkBM7OzpgyZUq9D46tX78e//nPf/Ddd9/VW2DV3vm7f/++1vJtaX5+fhgwYACOHTuG0NBQTJ48udHJ1nXR1wytH6nrm2++AQCMGjVK42MDAgIe+yAe/SktLQ0AeM1I77D4bTksftuYs2fPQqFQYO7cufDw8ABQ/9dzb7zxBl555RVMmDAB169fx6pVqzBixIgWuwN04MABCCEwYMAAAECvXr1gZWWFEydOqLQ7fvw4Kisr0a9fP43aaUNWVhZyc3Nx69YtmJjU3/WFEFi4cCHu3r2LnTt3NtiuV69eMDIywv/+9z+EhYVpLeeWNnfuXBw7dgzbt29HdnZ2o2110dcMrR+p448//kB8fDxcXV0xe/bsJp+HiIge4FRnbYybmxsAYP/+/aioqEB2dnad8Y1JSUlwcXHBhAkTAACxsbHw8fHBjBkzUFRU1KS4NTU1uHv3LqqqqnDmzBlERETAzc1NOQWTTCZDVFQUduzYgU2bNqGoqAhnz55FWFgYnJ2dERoaqlE7bXj99dfh5ubW6DLF586dw3vvvYePP/4YpqamdZadff/99wE8GO4wceJEbN++HRs3bkRRURHOnDmjMs+rtuzdu1ejqc4eNmXKFNjb2yMoKEhZ0DZEF33N0PrRw4QQKCkpQU1NDYQQuHXrFlJTUzFw4EAYGxtj586dTRrzS0REj9Dl43bthaZPByckJAgLCwsBQHTp0kUcOnRIvPvuu0IulwsAwtHRUWzevFls3bpVODo6CgDCzs5ObE2nurYAACAASURBVNmyRQghxNtvvy06dOggbG1txeTJk8WGDRsEAOHp6Sn8/PyERCIRHTp0EEeOHBFCCDF//nxhZGQkAAi5XC5OnDih0esLDQ0VpqamwsXFRZiYmAgbGxsxfvx4kZubq9KupqZGrFu3TnTt2lWYmpoKOzs7ERQUJC5cuKBxu7///e/K125paSkmTJgg1q9fL5ycnAQAYWFhIcaNGyeSkpKU17Jr164iNzdXfPTRR8LGxkYAEO7u7uK3334TQgjx3XffiY4dOwoAyn+mpqaiR48eIj09XQghxNmzZ1X2P/pv3bp1yhyLi4vFyy+/LDp27CisrKzEoEGDxLJlywQA4erqKjIzMzW6zkePHhUDBw4Uzs7OynhOTk4iMDBQ/O9//1O227Nnj7C2tharV69u8Fw7duxQLm1sb28vXn/9deW+t956S9k3hBBiyZIlyutqZGQkfHx8xKFDh4QQLdPXfvjhB9GtWzfla3J2dhaTJ09uMHdD6kdfffWV6N27t7CwsBBSqVR57Wpndnj66adFTEyMKCgoaLzzNAB8OlxjnO2B9BU/zy1mm0QIIbRSVRuQyZMnA/hzrFh7M2fOHKSlpaGgoEDXqTRLcnIysrOzlQ+0AEBlZSUWLlyI5ORk3L17F+bm5jrMkNqCttSPJBIJUlNTOX5VA+395zm1Xfw8t5g0jvkltTw8tVVb9McffyA8PBynT59W2S6VSuHm5gaFQgGFQqE3RQvpJ/YjIqK2j2N+DdD58+frjEGs719wcLCuU20x5ubmMDU1xcaNG3Hz5k0oFArcuHEDn3zyCZYtW4bg4OAWH09piNe5vdNFP6K2Zf/+/Vi0aBHS09Ph4eGh/JzPmjWrTtsRI0bA2toaxsbG6NmzJ06ePKmDjNW3du1aeHt7w9zcHJaWlvD29sbSpUvrjO+PiYmBj48PbGxsYGZmBi8vL7z11luNjpPXx7i1ampqEB8fj8DAwAbbKBQKxMbGwsvLC1KpFLa2tujVqxfy8vIAAF999RXWrl3b5m8ktRu6HnjRHrTnMWKLFi0SUqlUOT45LS1N1yk12cGDB8Vzzz0nbGxshLGxsZDL5SIwMFAkJSUJhUKh6/SojWhL/QgcI6ix5vw8X7ZsmRg7dqzKKn6enp7KMeK7d++uc8zevXvFCy+80OR8W9OYMWPE+++/L/Lz80VxcbHYtm2bMDU1rbPS5ZAhQ0RSUpIoKCgQRUVFIjU1VZiamoqRI0e2qbhCCPHbb7+JgQMHCgCNrroZFBQkunfvLo4dOyYUCoW4ceOGGDdunMqKrQkJCWLIkCHi7t27TcqFn+cWw+WNW0J7Ln6JqO3S9S/LsrIyERAQ0KZiNPXn+TvvvCO6deumsjy2EA+K382bNwsjIyPh4uIiCgsLVfa3peI3KCiozuubPHmyACBu3Lih3DZmzBhRVVWl0m7KlCkCgLhy5UqbiXv69GkxYcIEsWnTJuHn59dg8btlyxYhkUjEmTNnHnvO8PBwERAQ0KQ/lHX9eW5HtnHYAxERacXGjRu1tsJka8Z4nJycHCxduhQrV66ss/Q1AAQGBiIiIgLXr1/HggULdJBhy9ixY0ed11e7NPnDQwt2795dZ7VDe3t7AEBZWVmbidunTx+kp6djxowZMDMza7DdBx98gL59+8LX1/ex51yxYgVOnz6NhIQEjfOhlsPil4iIADyYazguLg49evSAmZkZ7OzsMH78eJw/f17ZJjw8HFKpFE5OTsptr732GiwtLSGRSHD79m0AD5ZVjoqKQm5uLiQSCby8vJCYmAiZTAYHBwfMmTMHzs7OkMlkCAwMVJlDujkxgAcr4jV1LuymSExMhBAC48aNa7DN6tWr0a1bN3zyySfYv39/o+dT531ITk6GpaUlLCwssGvXLowaNQo2NjZwdXXFli1bVM5XXV2NZcuWwc3NDebm5ujduzdSU1Ob96L/v+zsbNja2sLd3b3RdtevX4e5uTmefPLJNh33UZWVlTh27Bj8/PzUam9nZ4chQ4YgISEBgpNt6Y5u7zy3Dxz2QET6CBp+Tbps2TIhlUrF559/LgoLC8WZM2dE3759hb29vfjjjz+U7WbMmCEcHR1Vjl23bp0AIG7duqXcNnHiROHp6anSLjQ0VFhaWopz586JiooKkZWVJfz9/YW1tbXKV9PNibF7925hbW0tYmJi1H7ttZry89zDw0P4+PjUu8/T01NcunRJCCHEkSNHhJGRkejSpYsoKSkRQtQ/7EHd9yE6OloAEN9++624d++eyM/PF4MHDxaWlpaisrJS2W7BggXCzMxMbN++Xdy9e1csXrxYGBkZiZ9++kmj11mrsrJSXLt2Taxfv16YmZmJzz//vNH2paWlwtraWoSHhzcpnq7jCiHEM888U++wh0uXLgkAws/PTwwdOlQ4OTkJMzMz4e3tLTZs2CBqamrqHLNo0SIBQJw6dUqjHDT9PFODOOyBiIiA8vJyxMXFYcKECZg5cybkcjl8fX3x4Ycf4vbt2y26eqGJiYnyrqaPjw+Sk5NRXFyMlJSUFjn/mDFjUFRUhKVLl7bI+RpTWlqKS5cuwdPT87FtAwICMH/+fOTl5WHhwoX1tmnK+xAYGAgbGxt06tQJwcHBKC0txZUrVwDg/7F373FRlfkfwD8DzDAwXFXQEVCUURFvZLILqKFbuRpr3kDwUpE//allhNDmJTVFJLFWeHmhNuvHtuEFEMNWxVzX6PJLqDZUFtdCDO+JIjgIGAM8vz/8MdvERWZgGHA+79fLP3zOc873O2fmwJczz3ke3Lt3D8nJyZgxYwZmzZoFJycnrFmzBlKp1ODz7eHhAXd3d6xfvx5btmxBWFhYq/3j4+OhVCoRFxdnUDxTx21N47ALFxcXbNq0CYWFhbhx4wamT5+OZcuWYc+ePU32GTRoEID7S8iTabD4JSIiFBYW4u7duxgzZoxOu5+fH2QyWZOlrTvSmDFjYGtrq/O1fndRWloKIQRsbW3b1D8uLg5DhgzBzp078eWXXzbZ3t73QSaTAbg/9RYAfP/996iursbw4cO1fWxsbNCnTx+Dz/fly5dRWlqKPXv24IMPPsAjjzzS4rjrAwcOID09HZ988gns7e0NimfquK1pHAs8bNgwBAYGokePHnB0dMSGDRvg6OjY7B8rjZ+VGzduGC0vah2LXyIiQkVFBQDAzs6uyTYnJydUVlYaNb61tTVu3rxp1BjGcO/ePQBo9YGoX5LL5UhJSYFEIsGCBQtQU1Ojs72j34eqqioAwJo1a3TmF7948aJBD4EBgFQqhYuLCyZNmoR9+/ahsLAQ8fHxTfrt27cPmzdvRk5ODjw9PQ2K1RXitkapVAKAdhx6I5lMhv79+6O4uLjJPo2L4DR+dqjzsfglIiI4OTkBQLPFVUVFBdzd3Y0WW6PRGD2GsTQWMvosXhAQEIDo6GgUFRVh48aNOts6+n1wcXEBACQmJkIIofPv5MmTeh2rOSqVCpaWligsLNRp3759O1JTU3HixAn07du33XG6Stxfs7Ozw6BBg3D27Nkm2+rq6uDo6Nikvba2FgC4EqQJsfglIiIMHz4cdnZ2+Pbbb3Xa8/LyUFtbi0cffVTbZmVlpf1avSPk5ORACAF/f3+jxTAWV1dXSCQS3LlzR6/9Nm7cCG9vb+Tn5+u06/M+tIWHhwfkcnmTJbn1VVZWhrlz5zZpLyoqQn19PTw8PADcn6lixYoVKCgoQFZWVrN3sLtDXH2EhYUhPz8fFy5c0LZVV1fj4sWLzU5/1vhZ6d27d6flSLpY/BIREeRyOWJiYnDgwAGkpqZCrVajoKAAS5cuhVKpxOLFi7V9VSoVbt++jaysLGg0Gty8eRMXL15scswePXrg2rVrKCkpQWVlpbaYbWhoQHl5Oerq6nDmzBlERUWhX79+iIiI6JAY2dnZnTbVma2tLQYOHIgrV67otV/j8Idfz0urz/vQ1jjPP/889u7di+TkZKjVatTX1+PKlSu4fv06ACA8PBy9e/dudXllhUKBY8eO4cSJE1Cr1dBoNMjPz8dzzz0HhUKB6OhoAMDZs2exZcsW7Nq1C1KptMly7m+99Zb2mF05rj6io6PRv39/RERE4NKlSygrK8OKFStQU1PT7IONjZ+VtswLTMbB4peIiAAAr7/+OuLj4xEbG4tevXohKCgInp6eyMnJgUKh0PZ74YUXMHHiRMyZMwdDhgzBxo0btV/hBgQE4PLlywCApUuXwtXVFT4+Pnjqqadw+/ZtAPfHOo4YMQI2NjYYP348Bg8ejE8//VRn3Gx7Y3Sm4OBgFBYW6ozf/eijj6BSqVBcXAw/Pz+89NJLTfbz9/fXFm+/1Jb3ITk5GYmJiQCAkSNH4sKFC9i1axdiYmIAAJMnT0ZRUREAICkpCcuXL0dCQgJ69uwJpVKJqKgolJeXA7j/NXxpaSkOHjzY4muUy+UYO3YsFi5cCDc3N9jb2yM0NBSenp7Izc3VPlAn9Ji7tivHBYDc3FyMGzcOffv2RV5eHk6fPg2lUomxY8fi888/1/ZzdnbGF198AXd3d/j6+sLNzQ1ff/01Dh8+3Oz8v9988w3c3NwwcuTINudMHUsi9PnEULNCQ0MBABkZGSbOhIjoPyQSCdLS0jB79mxTp6K1ZMkSZGRkoKyszNSpNMuQn+fnz5/H0KFDkZKSgvnz5xsrNaNpaGjAhAkTEBERgQULFjCuEZWVlcHd3R1xcXHaP1Taqitez91UBu/8EhFRp9Ln4bDuQKVSITY2FrGxsTrL7XYH9fX1yMrKQmVlJcLDwxnXyNavXw9fX19ERkZ2emz6Dxa/RERE7bRq1SqEhoYiPDxc74ffTCknJweZmZnIzs5u81zFjGuYrVu34tSpUzhy5AikUmmnxiZdLH6JiKhTrF69GikpKbhz5w4GDBiA/fv3mzqlDrVp0yZERkbijTfeMHUqbfb4449j9+7d6NOnD+Ma0cGDB/Hzzz8jJycHzs7OnRqbmrIydQJERGQe4uPjm12U4GEyadIkTJo0ydRpUBczbdo0TJs2zdRp0P/jnV8iIiIiMhssfomIiIjIbLD4JSIiIiKzweKXiIiIiMwGH3jrIFeuXEF6erqp0yAi0nHy5ElTp9CtNC49y5/nRA8vrvDWAUJDQx+6KXuIiIioa+EKbx0ig8UvEVEnS09PR1hYGPjjl4io03F5YyIiIiIyHyx+iYiIiMhssPglIiIiIrPB4peIiIiIzAaLXyIiIiIyGyx+iYiIiMhssPglIiIiIrPB4peIiIiIzAaLXyIiIiIyGyx+iYiIiMhssPglIiIiIrPB4peIiIiIzAaLXyIiIiIyGyx+iYiIiMhssPglIiIiIrPB4peIiIiIzAaLXyIiIiIyGyx+iYiIiMhssPglIiIiIrPB4peIiIiIzAaLXyIiIiIyGyx+iYiIiMhssPglIiIiIrPB4peIiIiIzAaLXyIiIiIyGyx+iYiIiMhssPglIiIiIrPB4peIiIiIzAaLXyIiIiIyGyx+iYiIiMhssPglIiIiIrPB4peIiIiIzAaLXyIiIiIyG1amToCI6GF25coVPPfcc6ivr9e2lZeXw97eHhMmTNDpO2TIEPz5z3/u5AyJiMwLi18iIiNyd3fHxYsXUVxc3GTbZ599pvP/xx57rLPSIiIyWxz2QERkZM8++yykUukD+4WHh3dCNkRE5o3FLxGRkc2bNw91dXWt9hk2bBh8fHw6KSMiIvPF4peIyMi8vLwwcuRISCSSZrdLpVI899xznZwVEZF5YvFLRNQJnn32WVhaWja7ra6uDqGhoZ2cERGReWLxS0TUCebMmYOGhoYm7RYWFvD394enp2fnJ0VEZIZY/BIRdQKlUomxY8fCwkL3x66FhQWeffZZE2VFRGR+WPwSEXWSZ555pkmbEAIzZ840QTZEROaJxS8RUScJCQnRGfdraWmJJ554Aq6uribMiojIvLD4JSLqJM7OznjyySe1BbAQAvPnzzdxVkRE5oXFLxFRJ5o/f772wTepVIrp06ebOCMiIvPC4peIqBM9/fTTsLa2BgBMnToVdnZ2Js6IiMi8sPglIupECoVCe7eXQx6IiDqfRAghTJ0EdQ+hoaHYv3+/qdMgIiLSkZaWhtmzZ5s6DeoeMqxMnQF1L/7+/li+fLmp0yBql7CwMERFRSEgIMAk8evr65GWloa5c+eaJL4hEhMTAYDXP3U5YWFhpk6BuhkWv6QXd3d3/nVN3V5YWBgCAgJM+lmeMWMG5HK5yeLrKyMjAwB4/VOXw+KX9MUxv0REJtCdCl8ioocJi18iIiIiMhssfomIiIjIbLD4JSIiIiKzweKXiIiIiMwGi18iIgMdOXIEjo6O+Nvf/mbqVLq848ePY9WqVcjMzMTAgQMhkUggkUjwzDPPNOk7adIk2Nvbw9LSEsOGDcN3331ngozbLiEhAd7e3rCxsYFCoYC3tzfWrl0LtVqt0y82NhY+Pj5wcHCAtbU1VCoVXn31Vdy9e7dbxW3U0NCAxMREBAYGtthHo9EgPj4eKpUKMpkMTk5OGD58OEpKSgAAH3/8MRISElBfX9+uXIj0weKXiMhAXCOobV5//XVs27YNq1evxqxZs3DhwgV4eXmhZ8+eSE1NxeHDh3X6Hzt2DBkZGZg6dSoKCwsxevRoE2XeNl988QUWLVqES5cu4caNG9i4cSMSEhIQEhKi0+/EiRNYtmwZSkpKcOvWLcTHxyMpKQmhoaHdKi4AFBUV4bHHHkN0dDSqq6tb7BcWFoa//vWv2L17N6qrq/Hvf/8bXl5e2sL76aefhlwux+OPP46KigqD8yHSiyBqo5CQEBESEmLqNIjaDYBIS0szdRodqrq6WgQEBBjt+IZe/2+88YYYPHiwqKmp0Wn38vISu3fvFhYWFsLNzU1UVFTobM/OzhbTpk1rV86dZcaMGU1eX2hoqAAgrl27pm0LDg4WdXV1Ov1mz54tAIhLly51m7inTp0SM2fOFKmpqcLX11eMGjWq2X579+4VEolEnDlz5oHHjIyMFAEBAUKj0eidz8N4PZNRpfPOLxHRQ+D9999HaWmpqdPQcf78eaxduxYbNmxodl7jwMBAREVF4erVq3jllVdMkGHHOHDgQJPX5+bmBgA6QwsOHToES0tLnX69evUCgFbvnna1uKNGjUJmZibmzZsHa2vrFvu9/fbbGD16NEaMGPHAY65fvx6nTp1CUlKS3vkQ6YvFLxGRAb788kv069cPEokEO3bsAAAkJydDoVDA1tYWBw8exJQpU+Dg4AB3d3fs3btXu++2bdsgl8vh6uqKJUuWQKlUQi6XIzAwEHl5edp+kZGRkMlk6NOnj7btxRdfhEKhgEQiwa1btwAAUVFRiImJQXFxMSQSCVQqFQDg6NGjcHBwwKZNmzrjlDSxbds2CCHw9NNPt9gnLi4OgwcPxnvvvYfjx4+3ejwhBLZu3YqhQ4fC2toazs7OmD59Os6dO6ft09b3ALi/zPS6devQr18/2NjYYOTIkUhLS2vfi/5/RUVFcHJyQv/+/Vvtd/XqVdjY2GDAgAHdOu6v1dbWIjc3F76+vm3q7+zsjKCgICQlJXE4ERkdi18iIgOMGzcOX331lU7bCy+8gOXLl6Ompgb29vZIS0tDcXExBg4ciEWLFkGj0QC4X9RGRESguroaL7/8MkpKSvDdd9+hrq4OTz75JC5fvgzgfvH46+WEd+7ciQ0bNui0JSUlYerUqfDy8oIQAufPnwcA7UNEDQ0NRjkHD3L48GEMGTIEtra2LfaxsbHBX/7yF1hYWGDRokWoqqpqse/69euxatUqvPbaaygtLcXnn3+Oy5cvY/z48bhx4waAtr8HALBy5Ups2bIFiYmJuH79OqZOnYq5c+fi22+/Nej1ajQaXL16FTt27MDx48exfft2yGSyFvtXV1fjxIkTWLRoUav9umrc1ly7dg21tbX45z//iYkTJ2r/wBs6dCh27tzZbIH7yCOP4OrVqzh9+rRRciJqxOKXiMgIAgMD4eDgABcXF4SHh6OqqgqXLl3S6WNlZaW9i+nj44Pk5GRUVlYiJSWlQ3IIDg6GWq3G2rVrO+R4+qiqqsKPP/4ILy+vB/YNCAjA8uXLUVJSgpUrVzbbp6amBlu3bsXMmTMxf/58ODo6YsSIEXjnnXdw69YtvPvuu032ae09uHfvHpKTkzFjxgzMmjULTk5OWLNmDaRSqcHn38PDA+7u7li/fj22bNmCsLCwVvvHx8dDqVQiLi7OoHimjtuaxmEXLi4u2LRpEwoLC3Hjxg1Mnz4dy5Ytw549e5rsM2jQIABAQUGB0fIiAlj8EhEZXePdtV/edWzOmDFjYGtrq/M1fndVWloKIUSrd31/KS4uDkOGDMHOnTvx5ZdfNtleWFiIu3fvYsyYMTrtfn5+kMlkOsNFmvPr9+D7779HdXU1hg8fru1jY2ODPn36GHz+L1++jNLSUuzZswcffPABHnnkkRbHYR84cADp6en45JNPYG9vb1A8U8dtTeNY4GHDhiEwMBA9evSAo6MjNmzYAEdHx2b/WGn8rDTexScyFha/RERdiLW1NW7evGnqNNrt3r17ANDqA1G/JJfLkZKSAolEggULFqCmpkZne+M0WHZ2dk32dXJyQmVlpV75NQ6vWLNmjXbOYYlEgosXLxr0EBgASKVSuLi4YNKkSdi3bx8KCwsRHx/fpN++ffuwefNm5OTkwNPT06BYXSFua5RKJQBox6U3kslk6N+/P4qLi5vsY2NjA+A/nx0iY2HxS0TURWg0GlRUVMDd3d3UqbRbYyGjz+IFAQEBiI6ORlFRETZu3KizzcnJCQCaLXINOWcuLi4AgMTERAghdP6dPHlSr2M1R6VSwdLSEoWFhTrt27dvR2pqKk6cOIG+ffu2O05XiftrdnZ2GDRoEM6ePdtkW11dHRwdHZu019bWAvjPZ4fIWFj8EhF1ETk5ORBCwN/fX9tmZWX1wOESXZGrqyskEgnu3Lmj134bN26Et7c38vPzddqHDx8OOzu7Jg+j5eXloba2Fo8++qhecTw8PCCXy3Hq1Cm99vu1srIyzJ07t0l7UVER6uvr4eHhAeD+TBUrVqxAQUEBsrKymr2D3R3i6iMsLAz5+fm4cOGCtq26uhoXL15sdvqzxs9K7969Oy1HMk8sfomITKShoQHl5eWoq6vDmTNnEBUVhX79+iEiIkLbR6VS4fbt28jKyoJGo8HNmzdx8eLFJsfq0aMHrl27hpKSElRWVkKj0SA7O9tkU53Z2tpi4MCBuHLlil77NQ5/+PW8tHK5HDExMThw4ABSU1OhVqtRUFCApUuXQqlUYvHixXrHef7557F3714kJydDrVajvr4eV65cwfXr1wEA4eHh6N27d6vLKysUChw7dgwnTpyAWq2GRqNBfn4+nnvuOSgUCkRHRwMAzp49iy1btmDXrl2QSqU6Qy0kEgneeust7TG7clx9REdHo3///oiIiMClS5dQVlaGFStWoKamptkHGxs/K22ZF5ioPVj8EhEZYMeOHfDz8wMArFixAtOmTUNycjISExMBACNHjsSFCxewa9cuxMTEAAAmT56MoqIi7THu3buHESNGwMbGBuPHj8fgwYPx6aef6oyTfeGFFzBx4kTMmTMHQ4YMwcaNG7VfCwcEBGinRVu6dClcXV3h4+ODp556Crdv3+6U89Ca4OBgFBYW6ozf/eijj6BSqVBcXAw/Pz+89NJLTfbz9/fXFm+/9PrrryM+Ph6xsbHo1asXgoKC4OnpiZycHCgUCgDQ6z1ISkrC8uXLkZCQgJ49e0KpVCIqKgrl5eUA7n8NX1paioMHD7b4GuVyOcaOHYuFCxfCzc0N9vb2CA0NhaenJ3Jzc7UP1Okzd21XjgsAubm5GDduHPr27Yu8vDycPn0aSqUSY8eOxeeff67t5+zsjC+++ALu7u7w9fWFm5sbvv76axw+fLjZ+X+/+eYbuLm5YeTIkW3OmcgQEsHZpKmNGteBz8jIMHEmRO0jkUiQlpbWZA7dzrRkyRJkZGSgrKzMZDnow5Dr//z58xg6dChSUlIwf/58Y6VmNA0NDZgwYQIiIiKwYMECxjWisrIyuLu7Iy4uTvuHSlt1heuZupUM3vklIjIRfR4G645UKhViY2MRGxurs9xud1BfX4+srCxUVlYiPDyccY1s/fr18PX1RWRkZKfHJvPD4pc61cKFC2Fvbw+JRNLuB00eFg0NDUhMTERgYKDBx8jMzMTAgQObjOeTyWRwdXXFhAkT8Oabb2q/ziXqLKtWrUJoaCjCw8P1fvjNlHJycpCZmYns7Ow2z1XMuIbZunUrTp06hSNHjkAqlXZqbDJPLH6pU7333nvYtWuXqdPoMoqKivDYY48hOjra4LlFAWDWrFm4cOECvLy84OjoCCEEGhoaUFpaivT0dAwYMAArVqzAsGHDDF66lTrO6tWrkZKSgjt37mDAgAHYv3+/qVMyqk2bNiEyMhJvvPGGqVNps8cffxy7d+9Gnz59GNeIDh48iJ9//hk5OTlwdnbu1Nhkvlj8ErVDTU2NwXdsT58+jZUrV2Lp0qXNPvzRXhKJBE5OTpgwYQJSUlKQnp6OGzduIDg4uFvdgWtJe869qcXHx+Pnn3+GEAI//vgjQkJCTJ2S0U2aNAmbN282dRrUxUybNg2rVq1qMrsHkTGx+KVOJ5FITJ1Ch3n//fdbXEb0QUaNGoXMzEzMmzevzatgtUdISAgiIiJQWlqKd955x+jxjK09556IiMwXi18yKiEE3nzzTQwZMgTW1tZwdHTEH//4R50+W7Zsga2tLezt7VFaWoqYmBi4ubnh+++/hxACW7duxdChQ2FtbQ1nZ2dMnz4d586d0+6/bds2yOVyuLq6YsmSJVAqlZDL5QgMDEReXl6TfB50vMjISMhkMp2v/1588UUoFApIJBLtcp1RUVGIiYlBcXExJBIJVCqV5dA3RgAAIABJREFUMU4hjh492mFztTbOH5udnQ2A556IiMyQIGqjkJAQERISotc+r732mpBIJOJPf/qTKC8vF9XV1WLnzp0CgMjPz9fpB0C8/PLLYvv27WLmzJni3//+t1i3bp2QyWTiww8/FBUVFeLMmTNi9OjRolevXuKnn37S7r948WKhUCjE2bNnxb1790RhYaHw8/MT9vb24tKlS9p+bT3evHnzRO/evXVey5tvvikAiJs3b2rbZs2aJby8vPQ6J8357W9/K0aNGtXstkOHDgl7e3sRGxv7wON4eXkJR0fHFrer1WoBQHh4eGjbzPHcAxBpaWkG7WuuDLn+iToDr2fSUzrv/JLR1NTUIDExEU888QSio6Ph5OQEGxsb9OjRo8V9Nm/ejGXLliEzMxP9+/fH1q1bMXPmTMyfPx+Ojo4YMWIE3nnnHdy6dQvvvvuuzr5WVlbau4o+Pj5ITk5GZWUlUlJStPnoc7yuIjg4GGq1GmvXrm33sRpn2qisrGyyjeeeiIjMgZWpE6CH1/nz51FdXY3HH3/coP0LCwtx9+5djBkzRqfdz88PMpmsydfqvzZmzBjY2tpqv1Zv7/EeBlVVVRBCwMHBodV+5nDuT5482ekxu7PGpWfT09NNnAkRUfuw+CWjafxl6eLiYtD+FRUVAAA7O7sm25ycnJq9e/lr1tbWuHnzZocdr7v74YcfAADe3t6t9jOHc5+UlISkpKROj9vdhYWFmToFIqJ2YfFLRiOXywEAP//8s0H7Ozk5AUCzhVFFRQXc3d1b3V+j0ej0a+/xHgZHjx4FAEyZMqXVfuZw7rkcqn64vDl1VQ/TDELUOTjml4xm+PDhsLCwwGeffWbw/nZ2dk0WZcjLy0NtbS0effTRVvfPycmBEAL+/v56H8/KygoajcagvLuqn376CYmJiXB3d8eCBQta7ctzT0REDysWv2Q0Li4umDVrFvbv34/3338farUaZ86cafPDTXK5HDExMThw4ABSU1OhVqtRUFCApUuXQqlUYvHixTr9GxoaUF5ejrq6Opw5cwZRUVHo16+fdnovfY6nUqlw+/ZtZGVlQaPR4ObNm7h48WKTHHv06IFr166hpKQElZWVRinasrOz9ZrqTAiBu3fvoqGhAUII3Lx5E2lpaRg7diwsLS2RlZX1wDG/PPdERPTQMulkE9StGDLVUWVlpVi4cKHo2bOnsLOzE+PGjRPr1q0TAIS7u7s4ffq0SEhIEDY2NtopuD788EPt/g0NDeLNN98UgwYNElKpVDg7O4sZM2aI77//XifO4sWLhVQqFW5ubsLKyko4ODiI6dOni+LiYp1+bT1eWVmZmDhxopDL5WLAgAHipZdeEn/84x8FAKFSqbRTeH333Xeif//+wsbGRowbN05nyq4HOXnypBg7dqxQKpUCgAAg+vTpIwIDA8Vnn32m7XfkyBFhb28v4uLiWjzWxx9/LEaOHClsbW2FTCYTFhYWAoCQSCTCyclJ/OY3vxGxsbGirKxMZz9zPffg1Eh641Rn1FXxeiY9pUuEEMJUhTd1L115zN+SJUuQkZGBsrIyU6didrrjuZdIJBzzq6eufP2TeeP1THrK4LAHemjU19ebOgWzxXNPRETdBYtfog5y7tw5SCSSB/4LDw83dapERERmi8UvdXurV69GSkoK7ty5gwEDBmD//v0mycPb2xtCiAf+27dvn0nyM4aucu6p6zt+/DhWrVqFzMxMDBw4UPvH4DPPPNOk76RJk2Bvbw9LS0sMGzYM3333nQkybruEhAR4e3vDxsYGCoUC3t7eWLt2LdRqtU6/2NhY+Pj4wMHBAdbW1lCpVHj11Vdx9+7dbhW3UUNDAxITExEYGNhiH41Gg/j4eKhUKshkMjg5OWH48OEoKSkBAHz88cdISEjgt0fUuUw02Ji6IT7wQg8L8AEZvbXn+l+3bp2YOnWqUKvV2jYvLy/Rs2dPAUAcOnSoyT7Z2dli2rRpBufbmYKDg8Vbb70lSktLRWVlpUhPTxdSqVQ8+eSTOv2CgoLEzp07RVlZmVCr1SItLU1IpVIxefLkbhVXCCF++OEHMXbsWAFAjBo1qsV+M2bMEEOGDBG5ublCo9GIa9euiaeffloUFBRo+yQlJYmgoCBRXl5uUC68nklP6bzzS0TUyWpqalq9W9ZdYrTF5s2bsW/fPqSnp8Pe3l5n27Zt22BhYYHFixfjzp07Jsqw/WQyGV588UW4uLjAzs4OoaGhmD59Ov7+97/j+vXr2n52dnZYvHgxevToAXt7e8yePRszZszA0aNHcfny5W4T9/Tp01i5ciWWLl0KX1/fFvvt27cPWVlZyMjIwG9/+1tYWVlBqVTi4MGDGD58uLbfyy+/jFGjRuGpp55CXV2d3vkQ6YvFLxFRJ3v//fdRWlra7WM8yPnz57F27Vps2LBBu+LjLwUGBiIqKgpXr17FK6+8YoIMO8aBAweavD43NzcA0BlacOjQIVhaWur069WrFwCgurq628QdNWoUMjMzMW/ePFhbW7fY7+2338bo0aMxYsSIBx5z/fr1OHXqFJccp07B4peI6AGEENi6dSuGDh0Ka2trODs7Y/r06Th37py2T2RkJGQyGfr06aNte/HFF6FQKCCRSHDr1i0AQFRUFGJiYlBcXAyJRAKVSoVt27ZBLpfD1dUVS5YsgVKphFwuR2BgIPLy8jokBnB/eWt9Fkxpr23btkEIgaeffrrFPnFxcRg8eDDee+89HD9+vNXjteV9SE5OhkKhgK2tLQ4ePIgpU6bAwcEB7u7u2Lt3r87x6uvrsW7dOvTr1w82NjYYOXIk0tLS2vei/19RURGcnJzQv3//VvtdvXoVNjY2GDBgQLeO+2u1tbXIzc1t9c7wLzk7OyMoKAhJSUkQnIGVjM20wy6oO+GYX3pYQM8xguvWrRMymUx8+OGHoqKiQpw5c0aMHj1a9OrVS2dxjXnz5onevXvr7Pvmm28KAOLmzZvatlmzZgkvLy+dfosXLxYKhUKcPXtW3Lt3TxQWFgo/Pz9hb2+vXdijvTEOHTok7O3tRWxsbJtfeyNDrv+BAwcKHx+fZrd5eXmJH3/8UQghxFdffSUsLCyEp6enuHv3rhCi+TG/bX0fXnvtNQFA/OMf/xB37twRpaWlYvz48UKhUIja2lptv1deeUVYW1uL/fv3i/LycrF69WphYWEhvvnmG71eZ6Pa2lpx5coVsX37dmFtba2zaExzqqqqhL29vYiMjDQonqnjCiHEb3/722bH/P74448CgPD19RUTJkwQffr0EdbW1sLb21vs2LFDNDQ0NNln1apVAoDIz8/XKwd9r2cyexzzS0TUmpqaGmzduhUzZ87E/Pnz4ejoiBEjRuCdd97BrVu32rxcd1tYWVlp72r6+PggOTkZlZWVSElJ6ZDjBwcHQ61WY+3atR1yvNZUVVXhxx9/hJeX1wP7BgQEYPny5SgpKcHKlSub7WPI+xAYGAgHBwe4uLggPDwcVVVVuHTpEgDg3r17SE5OxowZMzBr1iw4OTlhzZo1kEqlBp9vDw8PuLu7Y/369diyZQvCwsJa7R8fHw+lUom4uDiD4pk6bmsah124uLhg06ZNKCwsxI0bNzB9+nQsW7YMe/bsabLPoEGDAAAFBQVGy4sI4LAHIqJWFRYW4u7duxgzZoxOu5+fH2Qymc6whI42ZswY2Nra6nyt312UlpZCCAFbW9s29Y+Li8OQIUOwc+dOfPnll022t/d9kMlkAO5PvQUA33//Paqrq3UevLKxsUGfPn0MPt+XL19GaWkp9uzZgw8++ACPPPJIi+OuDxw4gPT0dHzyySdNHgTsLnFb0zgWeNiwYQgMDESPHj3g6OiIDRs2wNHRsdk/Vho/Kzdu3DBaXkQAi18iolZVVFQAuP/E/K85OTmhsrLSqPGtra1x8+ZNo8Ywhnv37gFAqw9E/ZJcLkdKSgokEgkWLFiAmpoane0d/T5UVVUBANasWaOzCM3FixcNeggMAKRSKVxcXDBp0iTs27cPhYWFiI+Pb9Jv37592Lx5M3JycuDp6WlQrK4QtzVKpRIAtOPQG8lkMvTv3x/FxcVN9rGxsQHwn88OkbGw+CUiaoWTkxMANFtcVVRUwN3d3WixNRqN0WMYS2Mho8/iBQEBAYiOjkZRURE2btyos62j3wcXFxcAQGJiYpOFaE6ePKnXsZqjUqlgaWmJwsJCnfbt27cjNTUVJ06cQN++fdsdp6vE/TU7OzsMGjQIZ8+ebbKtrq4Ojo6OTdpra2sB/OezQ2QsLH6JiFoxfPhw2NnZ4dtvv9Vpz8vLQ21tLR599FFtm5WVlfZr9Y6Qk5MDIQT8/f2NFsNYXF1dIZFI9J6/d+PGjfD29kZ+fr5Ouz7vQ1t4eHhALpfj1KlTeu33a2VlZZg7d26T9qKiItTX18PDwwPA/ZkqVqxYgYKCAmRlZTV7B7s7xNVHWFgY8vPzceHCBW1bdXU1Ll682Oz0Z42fld69e3dajmSeWPwSEbVCLpcjJiYGBw4cQGpqKtRqNQoKCrB06VIolUosXrxY21elUuH27dvIysqCRqPBzZs3cfHixSbH7NGjB65du4aSkhJUVlZqi9mGhgaUl5ejrq4OZ86cQVRUFPr164eIiIgOiZGdnd1pU53Z2tpi4MCBuHLlil77NQ5/+PW8tPq8D22N8/zzz2Pv3r1ITk6GWq1GfX09rly5ol0gIjw8HL179251eWWFQoFjx47hxIkTUKvV0Gg0yM/Px3PPPQeFQoHo6GgAwNmzZ7Flyxbs2rULUqlUZ6iFRCLBW2+9pT1mV46rj+joaPTv3x8RERG4dOkSysrKsGLFCtTU1DT7YGPjZ6Ut8wITtQeLXyKiB3j99dcRHx+P2NhY9OrVC0FBQfD09EROTg4UCoW23wsvvICJEydizpw5GDJkCDZu3Kj9CjcgIEC7mtbSpUvh6uoKHx8fPPXUU7h9+zaA+2MdR4wYARsbG4wfPx6DBw/Gp59+qjNutr0xOlNwcDAKCwt1xu9+9NFHUKlUKC4uhp+fH1566aUm+/n7+2uLt19qy/uQnJyMxMREAMDIkSNx4cIF7Nq1CzExMQCAyZMno6ioCACQlJSE5cuXIyEhAT179oRSqURUVBTKy8sB3P8avrS0FAcPHmzxNcrlcowdOxYLFy6Em5sb7O3tERoaCk9PT+Tm5mofqBN6zF3bleMCQG5uLsaNG4e+ffsiLy8Pp0+fhlKpxNixY/H5559r+zk7O+OLL76Au7s7fH194ebmhq+//hqHDx9udv7fb775Bm5ubhg5cmSbcyYyhEToc2WQWQsNDQUAZGRkmDgTovaRSCRIS0vD7NmzTZ2K1pIlS5CRkYGysjJTp9IsQ67/8+fPY+jQoUhJScH8+fONlZrRNDQ0YMKECYiIiMCCBQsY14jKysrg7u6OuLg47R8qbdUVr2fq0jJ455eIqIvQ5+Gw7kClUiE2NhaxsbE6y+12B/X19cjKykJlZSXCw8MZ18jWr18PX19fREZGdnpsMj8sfomIyGhWrVqF0NBQhIeH6/3wmynl5OQgMzMT2dnZbZ6rmHENs3XrVpw6dQpHjhyBVCrt1Nhknlj8EhGZ2OrVq5GSkoI7d+5gwIAB2L9/v6lT6lCbNm1CZGQk3njjDVOn0maPP/44du/ejT59+jCuER08eBA///wzcnJy4Ozs3KmxyXxZmToBIiJzFx8f3+yiBA+TSZMmYdKkSaZOg7qYadOmYdq0aaZOg8wM7/wSERERkdlg8UtEREREZoPFLxERERGZDRa/RERERGQ2+MAb6SU3N1c72T1Rd5aYmMgFW/SQm5sLALz+iajbY/FLbRYQEGDqFIg6REhIiEnj//TTT8jPz8eUKVNMmoc+/P39TZ0CUbNCQkLg4eFh6jSoG+HyxkREnSw9PR1hYWHgj18iok7H5Y2JiIiIyHyw+CUiIiIis8Hil4iIiIjMBotfIiIiIjIbLH6JiIiIyGyw+CUiIiIis8Hil4iIiIjMBotfIiIiIjIbLH6JiIiIyGyw+CUiIiIis8Hil4iIiIjMBotfIiIiIjIbLH6JiIiIyGyw+CUiIiIis8Hil4iIiIjMBotfIiIiIjIbLH6JiIiIyGyw+CUiIiIis8Hil4iIiIjMBotfIiIiIjIbLH6JiIiIyGyw+CUiIiIis8Hil4iIiIjMBotfIiIiIjIbLH6JiIiIyGyw+CUiIiIis8Hil4iIiIjMBotfIiIiIjIbLH6JiIiIyGyw+CUiIiIis8Hil4iIiIjMBotfIiIiIjIbVqZOgIjoYabRaHD37l2dtqqqKgBAeXm5TrtEIoGTk1On5UZEZI5Y/BIRGdHt27fh5uaG+vr6Jtt69Oih8/+JEyfixIkTnZUaEZFZ4rAHIiIj6t27Nx577DFYWLT+41YikWDOnDmdlBURkfli8UtEZGTPPPPMA/tYWlpi5syZnZANEZF5Y/FLRGRks2bNgpVVy6PMLC0tMXnyZPTs2bMTsyIiMk8sfomIjMzBwQFTpkxpsQAWQmD+/PmdnBURkXli8UtE1Anmz5/f7ENvACCTyfCHP/yhkzMiIjJPLH6JiDrBH/7wB9ja2jZpl0qlmDFjBhQKhQmyIiIyPyx+iYg6gVwux8yZMyGVSnXaNRoN5s2bZ6KsiIjMD4tfIqJOMnfuXGg0Gp02BwcHPPnkkybKiIjI/LD4JSLqJE888YTOwhZSqRRz5syBTCYzYVZEROaFxS8RUSexsrLCnDlztEMfNBoN5s6da+KsiIjMC4tfIqJONGfOHO3Qh969e2PcuHEmzoiIyLyw+CUi6kSBgYFwc3MDADz77LMPXPaYiIg6VstLDhE1Iz093dQpEHV7fn5+uHr1Knr27MlriqidPDw8EBAQYOo0qBuRCCGEqZOg7kMikZg6BSIiIq2QkBBkZGSYOg3qPjJ455f0lpaWhtmzZ5s6DSKDhIaGAoDJf1nu378fISEhJs2hrdLT0xEWFgbeK6GupvF6JtIHB5sREZlAdyl8iYgeNix+iYiIiMhssPglIiIiIrPB4peIiIiIzAaLXyIiIiIyGyx+iYiIiMhssPglIjLAkSNH4OjoiL/97W+mTqXLO378OFatWoXMzEwMHDgQEokEEokEzzzzTJO+kyZNgr29PSwtLTFs2DB89913Jsi47RISEuDt7Q0bGxsoFAp4e3tj7dq1UKvVOv1iY2Ph4+MDBwcHWFtbQ6VS4dVXX8Xdu3e7VdxGDQ0NSExMRGBgYIt9NBoN4uPjoVKpIJPJ4OTkhOHDh6OkpAQA8PHHHyMhIQH19fXtyoVIXyx+iYgMwDlv2+b111/Htm3bsHr1asyaNQsXLlyAl5cXevbsidTUVBw+fFin/7Fjx5CRkYGpU6eisLAQo0ePNlHmbfPFF19g0aJFuHTpEm7cuIGNGzciISGhyVR2J06cwLJly1BSUoJbt24hPj4eSUlJBs9Ta6q4AFBUVITHHnsM0dHRqK6ubrFfWFgY/vrXv2L37t2orq7Gv//9b3h5eWkL76effhpyuRyPP/44KioqDM6HSG+CSA8ARFpamqnTIDJYSEiICAkJMXUaHaq6uloEBAQY7fhpaWnCkF8Xb7zxhhg8eLCoqanRaffy8hK7d+8WFhYWws3NTVRUVOhsz87OFtOmTWtXzp1lxowZTV5faGioACCuXbumbQsODhZ1dXU6/WbPni0AiEuXLnWbuKdOnRIzZ84UqampwtfXV4waNarZfnv37hUSiUScOXPmgceMjIwUAQEBQqPR6J3Pw3g9k9Gl884vEVE39/7776O0tNTUaeg4f/481q5diw0bNkAulzfZHhgYiKioKFy9ehWvvPKKCTLsGAcOHGjy+tzc3ABAZ2jBoUOHYGlpqdOvV69eANDq3dOuFnfUqFHIzMzEvHnzYG1t3WK/t99+G6NHj8aIESMeeMz169fj1KlTSEpK0jsfIkOw+CUi0tOXX36Jfv36QSKRYMeOHQCA5ORkKBQK2Nra4uDBg5gyZQocHBzg7u6OvXv3avfdtm0b5HI5XF1dsWTJEiiVSsjlcgQGBiIvL0/bLzIyEjKZDH369NG2vfjii1AoFJBIJLh16xYAICoqCjExMSguLoZEIoFKpQIAHD16FA4ODti0aVNnnJImtm3bBiEEnn766Rb7xMXFYfDgwXjvvfdw/PjxVo8nhMDWrVsxdOhQWFtbw9nZGdOnT8e5c+e0fdr6HgBAfX091q1bh379+sHGxgYjR45EWlpa+170/ysqKoKTkxP69+/far+rV6/CxsYGAwYM6NZxf622tha5ubnw9fVtU39nZ2cEBQUhKSmJw4moU7D4JSLS07hx4/DVV1/ptL3wwgtYvnw5ampqYG9vj7S0NBQXF2PgwIFYtGgRNBoNgPtFbUREBKqrq/Hyyy+jpKQE3333Herq6vDkk0/i8uXLAO4Xj7Nnz9aJsXPnTmzYsEGnLSkpCVOnToWXlxeEEDh//jwAaB8iamhoMMo5eJDDhw9jyJAhsLW1bbGPjY0N/vKXv8DCwgKLFi1CVVVVi33Xr1+PVatW4bXXXkNpaSk+//xzXL58GePHj8eNGzcAtP09AICVK1diy5YtSExMxPXr1zF16lTMnTsX3377rUGvV6PR4OrVq9ixYweOHz+O7du3QyaTtdi/uroaJ06cwKJFi1rt11XjtubatWuora3FP//5T0ycOFH7B97QoUOxc+fOZgvcRx55BFevXsXp06eNkhPRL7H4JSLqYIGBgXBwcICLiwvCw8NRVVWFS5cu6fSxsrLS3sX08fFBcnIyKisrkZKS0iE5BAcHQ61WY+3atR1yPH1UVVXhxx9/hJeX1wP7BgQEYPny5SgpKcHKlSub7VNTU4OtW7di5syZmD9/PhwdHTFixAi88847uHXrFt59990m+7T2Hty7dw/JycmYMWMGZs2aBScnJ6xZswZSqdTg8+/h4QF3d3esX78eW7ZsQVhYWKv94+PjoVQqERcXZ1A8U8dtTeOwCxcXF2zatAmFhYW4ceMGpk+fjmXLlmHPnj1N9hk0aBAAoKCgwGh5ETVi8UtEZESNd9d+edexOWPGjIGtra3O1/jdVWlpKYQQrd71/aW4uDgMGTIEO3fuxJdfftlke2FhIe7evYsxY8botPv5+UEmk+kMF2nOr9+D77//HtXV1Rg+fLi2j42NDfr06WPw+b98+TJKS0uxZ88efPDBB3jkkUdaHId94MABpKen45NPPoG9vb1B8UwdtzWNY4GHDRuGwMBA9OjRA46OjtiwYQMcHR2b/WOl8bPSeBefyJhY/BIRdRHW1ta4efOmqdNot3v37gFAqw9E/ZJcLkdKSgokEgkWLFiAmpoane2N02DZ2dk12dfJyQmVlZV65dc4vGLNmjXaOYclEgkuXrxo0ENgACCVSuHi4oJJkyZh3759KCwsRHx8fJN++/btw+bNm5GTkwNPT0+DYnWFuK1RKpUAoB2X3kgmk6F///4oLi5uso+NjQ2A/3x2iIyJxS8RUReg0WhQUVEBd3d3U6fSbo2FjD6LFwQEBCA6OhpFRUXYuHGjzjYnJycAaLbINeScubi4AAASExMhhND5d/LkSb2O1RyVSgVLS0sUFhbqtG/fvh2pqak4ceIE+vbt2+44XSXur9nZ2WHQoEE4e/Zsk211dXVwdHRs0l5bWwvgP58dImNi8UtE1AXk5ORACAF/f39tm5WV1QOHS3RFrq6ukEgkuHPnjl77bdy4Ed7e3sjPz9dpHz58OOzs7Jo8jJaXl4fa2lo8+uijesXx8PCAXC7HqVOn9Nrv18rKyjB37twm7UVFRaivr4eHhweA+zNVrFixAgUFBcjKymr2DnZ3iKuPsLAw5Ofn48KFC9q26upqXLx4sdnpzxo/K7179+60HMl8sfglIjKBhoYGlJeXo66uDmfOnEFUVBT69euHiIgIbR+VSoXbt28jKysLGo0GN2/exMWLF5scq0ePHrh27RpKSkpQWVkJjUaD7Oxsk011Zmtri4EDB+LKlSt67dc4/OHX89LK5XLExMTgwIEDSE1NhVqtRkFBAZYuXQqlUonFixfrHef555/H3r17kZycDLVajfr6ely5cgXXr18HAISHh6N3796tLq+sUChw7NgxnDhxAmq1GhqNBvn5+XjuueegUCgQHR0NADh79iy2bNmCXbt2QSqV6gy1kEgkeOutt7TH7Mpx9REdHY3+/fsjIiICly5dQllZGVasWIGamppmH2xs/Ky0ZV5govZi8UtEpKcdO3bAz88PALBixQpMmzYNycnJSExMBACMHDkSFy5cwK5duxATEwMAmDx5MoqKirTHuHfvHkaMGAEbGxuMHz8egwcPxqeffqozTvaFF17AxIkTMWfOHAwZMgQbN27Ufi0cEBCgnRZt6dKlcHV1hY+PD5566incvn27U85Da4KDg1FYWKgzfvejjz6CSqVCcXEx/Pz88NJLLzXZz9/fX1u8/dLrr7+O+Ph4xMbGolevXggKCoKnpydycnKgUCgAQK/3ICkpCcuXL0dCQgJ69uwJpVKJqKgolJeXA7j/NXxpaSkOHjzY4muUy+UYO3YsFi5cCDc3N9jb2yM0NBSenp7Izc3VPlCnz9y1XTkuAOTm5mLcuHHo27cv8vLycPr0aSiVSowdOxaff/65tp+zszO++OILuLu7w9fXF25ubvj6669x+PDhZuf//eabb+Dm5oaRI0e2OWciQ0kEZ5QmPUgkEqSlpTWZf5SouwgNDQUAZGRkmCyHJUuWICMjA2VlZSbLQR/p6ekICwvTq5g6f/48hg4dipSUFMyfP9+I2RlHQ0MDJkyYgIiICCxYsIBxjaisrAzu7u6Ii4vT/qHSVl3heqZuJ4N3fomITECfh8G6I5VKhdjYWMTGxuost9sd1NfXIysrC5WVlQgPD2dcI1u/fj18fX0RGRnZ6bHJPLH4JbOn0zmDAAAgAElEQVQUGxsLHx8fODg4wNraGiqVCq+++uoDf0kvXLgQ9vb2kEgkBj0sY2jcB/n+++/x0ksvYdiwYbC3t4eVlRUcHR0xePBgBAcHd8gT7O3VlteemZmJgQMHNhmbKJPJ4OrqigkTJuDNN9/UfjVNXduqVasQGhqK8PBwvR9+M6WcnBxkZmYiOzu7zXMVM65htm7dilOnTuHIkSOQSqWdGpvMmCDSAwCRlpZm6jTaLSgoSOzcuVOUlZUJtVot0tLShFQqFZMnT37gvnv37hUARH5+fqfGbcl7770npFKpeOyxx8TRo0dFeXm5uHfvniguLhb79u0TgYGB4s9//rPBx+8o+rx2Ly8v4ejoKIQQoqGhQZSXl4tPP/1URERECIlEIpRKpfjmm28MyiMkJESEhIS067W0x6pVq4RMJhMAhKenp8jIyDBZLm2VlpYm2vPr4pNPPhErVqzowIzoYZCVlSXi4+NFXV2dwccw9fVM3VI6i1/Sy8NS/AYHBzf5gTt79mwBQFy6dKnVfdtT/LYnbnNOnjwpLC0txe9+9zuh0Wia7XP06FGxfft2vY/d0fR57b8sfn8tIyNDWFhYCFdXV1FRUaF3Hvxlqb/2Fr9ExsLrmQyQzmEPZJYOHTrUZDqlXr16AcADV3iSSCQmiducuLg41NfX44033oCVlVWzfX7/+99j2bJl+ifbwTrqtYeEhCAiIgKlpaV45513OjRHIiJ6+LH4JaP78MMPMWbMGMjlcigUCnh6empXcBJCYOvWrRg6dCisra3h7OyM6dOn49y5c9r9k5OToVAoYGtri4MHD2LKlClwcHCAu7s79u7dq+03dOhQSCQSWFhY4NFHH9UWVK+++iocHR0hl8vxl7/8pcU8r169ChsbGwwYMEDbJoTAm2++iSFDhsDa2hqOjo744x//2KHnp7m4R48efeAcrbW1tfjHP/6Bnj174je/+U2b43X1c94WjXPhZmdn67UfERERv8civUDPYQ+JiYkCgHjjjTdEWVmZuH37tvjzn/8s5s2bJ4QQYt26dUImk4kPP/xQVFRUiDNnzojRo0eLXr16iZ9++kl7nNdee00AEP/4xz/EnTt3RGlpqRg/frxQKBSitrZWCCFEXV2d8PT0FP369Wvy9fry5ctFYmJii3lWVVUJe3t7ERkZqdP+2muvCYlEIv70pz+J8vJyUV1dLXbu3GnwsIe2xj106JCwt7cXsbGxLe77ww8/CADC399fr5hd/ZwL0fqwByGEUKvVAoDw8PDQ67ULwa9JDcFhD9RV8XomA3DML+lHn+K3trZWODk5iYkTJ+q019XViaSkJFFdXS3s7OxEeHi4zvavv/5aANAp/BoLsZqaGm1bYxF6/vx5bVtjsZ2enq5tq6qqEv369RN37txpMdfXXntNDB48WKjVam1bdXW1sLW1FU8++aRO3/aM+W1L3Lb69ttvBQDxxBNPtHmfrn7OGz2o+BVCCIlEIpycnFrt0xz+stQfi1/qqng9kwHSmx8kSNQBzpw5g4qKCvz+97/Xabe0tMTLL7+Mb7/9Fnfv3sWYMWN0tvv5+UEmkyEvL6/V48tkMgCARqPRti1cuBDr169HUlKSdvLz1NRUTJ8+HQ4ODs0e58CBA0hPT8exY8dgb2+vbT9//jyqq6vx+OOPt/1F66GluG1lZ2cHQL/xsoWFhV36nLdVVVUVhBAtHv9BcnNztbnSgzUuPctzRl1Nbm4u/P39TZ0GdTMc80tGo1arAQBOTk7Nbq+oqADwnyLul5ycnFBZWal3TDs7O/z3f/83vvrqK3z99dcAgLfffrvFydP37duHzZs3IycnB56enjrbGn/hu7i46J3Hg7QWt608PT0hl8vxww8/tHmfrn7O26rxNXt7exu0PxERmS/e+SWj6du3LwDg1q1bzW5vLIqbK7gqKirg7u5uUNzIyEgkJSUhMTERS5cuhYeHB7y8vJr02759Oz755BOcOHGi2WJQLpcDAH7++WeD8mjJg+K2lbW1NX7/+9/j4MGD+N///V+MHTu22X63b9/Gq6++ivfee6/Ln/O2Onr0KABgypQpBu3v7+/P5VD10Li8Mc8ZdTX8NoIMwTu/ZDSenp7o0aMHjh071uz24cOHw87ODt9++61Oe15eHmpra/Hoo48aFNfd3R2zZ8/G/v37sXbtWkRFRelsF0JgxYoVKCgoQFZWVotF2PDhw2FhYYHPPvvMoDx+ra1x9bF+/XpYW1sjOjoaNTU1zfb517/+pZ0Grauf87b46aefkJiYCHd3dyxYsMDg4xARkXli8UtGY21tjdWrV+Pzzz9HZGQkrl69ioaGBlRWVuLs2bOQy+WIiYnBgQMHkJqaCrVajYKCAixduhRKpRKLFy82OHZMTAzq6upQXl6O3/3udzrbzp49iy1btmDXrl2QSqVNltJ96623ANwf7jBr1izs378f77//PtRqNc6cOYN3333XoJzaGhe4P4XXg6Y6AwBfX1/s3r0b//rXvzB+/HgcOXIEd+7cgUajwY8//ohdu3bhv/7rv7TLhnb1c/5LQgjcvXsXDQ0NEELg5s2bSEtLw9ixY2FpaYmsrCyDx/wSEZH5YvFLRhUTE4MdO3YgJycHKpUKCoUCQUFByMnJAQC8/vrriI+PR2xsLHr16oWgoCB4enoiJycHCoUCwP05ZxMTEwEAI0eOxIULF7Br1y7ExMQAACZPnoyioiKduI888ggmTpyIl19+uUlOQog25/8///M/eP7557FixQq4ubnhxRdfxPjx4wEAU6dOxZkzZ9p8LH3i6mPWrFk4d+4cJk6ciJUrV8Ld3R02NjYYPXo03n77bQQFBWHOnDna/l35nP/tb3/DqFGjcP36ddy7dw+Ojo6wtLSEpaUlBg8ejK1btyIiIgKFhYUG36UmIiLzJhHG+o1MDyWJRIK0tDTMnj3b1KkQGaRxjCDHr7Zd45hf/rqgrobXMxkgg3d+iYiIiMhssPglaodz5841Gb/a3L/w8HBTp0pkMsePH8eqVauQmZmJgQMHaq+LZ555pknfSZMmwd7eHpaWlhg2bBi+++47E2TcdgkJCfD29oaNjQ0UCgW8vb2xdu1a7VSPjWJjY+Hj4wMHBwdYW1tDpVLh1Vdfxd27d8027scff4yEhATU19cblAuRwUyzuAZ1V9BzeWOiroYrQumvPSu8rVu3TkydOlVnJT8vLy/Rs2dPAUAcOnSoyT7Z2dli2rRpBufbmYKDg8Vbb70lSktLRWVlpUhPTxdSqbTJypBBQUFi586doqysTKjVapGWliakUqmYPHmyWcdNSkoSQUFBory83KB8eD2TAdJ555eIqJPV1NQgMDCw28d4kM2bN2Pfvn1IT09vspLftm3bYGFhgcWLF+POnTsmyrD9ZDIZXnzxRbi4uMDOzg6hoaGYPn06/v73v+P69evafnZ2dli8eDF69OgBe3t7zJ49G//H3p2GRXWleQD/F1BQUBSLsgooAlEkYicq3ULiqG2HcRkXFASNicsT2yUJopgYXAkQImojLUqnY9IkI4ksSiBxb0OYxImxtaMtjW0ajSsmgig7SAFnPjhULFmkiqWA+v+ehw/eOve+b917Cl5vnXtOQEAAjh49ips3b+pt3JUrV+JXv/oVpkyZgvr6eo3zIdIGi18iom724YcfoqioqNfHaMvly5exceNGvP3226oFYx7l5+eHsLAwFBYWYs2aNTrIsHNkZmY2e39OTk4AoPYV/8GDB2FoaKjWzsbGBoBmS5T3xbiRkZE4f/48EhISNM6HSBssfomInkAIgfj4eAwbNgwmJiawtrbGzJkzcenSJVWb0NBQGBsbw8HBQbXt1VdfhVwuh0QiUa10GBYWhvDwcFy5cgUSiQQeHh7YuXMnZDIZ7OzssGzZMjg6OkImk8HPzw+nT5/ulBjAw5Xx2jN/dGfYuXMnhBCYPn16q21iYmIwZMgQfPDBBzhx4kSbx2vPNUhKSoJcLoeZmRmys7MxefJkWFhYwNnZGfv27VM7XkNDAzZt2oSBAwfC1NQUI0aMQFpaWsfe9P8rKCiAlZUVBg0a1Ga7wsJCmJqaYvDgwXod19raGuPGjUNCQgJnFKHuodthF9TbgGN+qZfTZozgpk2bhLGxsdi7d68oLS0VFy5cECNHjhQ2Njbi559/VrV78cUXhb29vdq+27ZtEwBEcXGxatvs2bOFu7u7WrulS5cKuVwuLl68KGpra0V+fr7w8fERCoVC3Lhxo1NiHDx4UCgUChEVFaXR+9dmzK+bm5vw8vJq8TV3d3dx9epVIYQQ3377rTAwMBCurq6isrJSCNHymN/2XoP169cLAOLLL78UZWVloqioSIwdO1bI5XJRV1enardmzRphYmIi9u/fL+7fvy/WrVsnDAwMxJkzZzR6n03q6urErVu3RGJiojAxMRF79+5ts31VVZVQKBQiNDRUq3h9LW5ERIQAIM6dO6dRHhzzS1rgmF8iorbU1NQgPj4es2bNwvz582FpaQlvb2+89957uHv3rtYr/rXEyMhIdWfTy8sLSUlJqKioQHJycqccf+rUqSgvL8fGjRs75XitqaqqwtWrV+Hu7v7Etr6+vli1ahWuXbuGt956q8U22lwDPz8/WFhYwNbWFiEhIaiqqsKNGzcAALW1tUhKSkJAQABmz54NKysrbNiwAVKpVOtz7eLiAmdnZ0RGRmLr1q0IDg5us31sbCwcHR0RExOjVby+Fvepp54CAOTl5XUoL6L2YPFLRNSG/Px8VFZWYvTo0WrbfXx8YGxsrDYsobONHj0aZmZmal/t9wZFRUUQQsDMzKxd7WNiYjB06FDs3r0bJ0+ebPZ6R6+BsbExAECpVAIAfvjhB1RXV2P48OGqNqampnBwcND6XN+8eRNFRUX49NNP8fHHH+PZZ59tdcx1ZmYm0tPTcezYsWYPAupr3Ka+cufOnQ7lRdQeLH6JiNpQWloK4OHT64+zsrJCRUVFl8Y3MTFBcXFxl8bobLW1tQAe5t4eMpkMycnJkEgkWLx4MWpqatRe7+xrUFVVBQDYsGGD2nzc169f1+ohMACQSqWwtbWFv78/UlNTkZ+fj9jY2GbtUlNTsWXLFuTm5sLV1VWrWH0xrqmpKYBf+g5RV2LxS0TUBisrKwBoscAqLS2Fs7Nzl8VWKpVdHqMrNBUymixe4Ovri9WrV6OgoADR0dFqr3X2NbC1tQUA7NixA0IItZ9Tp05pdKyWeHh4wNDQEPn5+WrbExMTkZKSgpycHAwYMKDDcfpS3Lq6OgC/9B2irsTil4ioDcOHD4e5uTnOnj2rtv306dOoq6vDqFGjVNuMjIxUX613htzcXAghMGbMmC6L0RXs7OwgkUg0nr83Ojoanp6eOHfunNp2Ta5Be7i4uEAmk+H8+fMa7fe4kpISzJs3r9n2goICNDQ0wMXFBcDDmSrWrl2LvLw8ZGVltXgHW9/jNvUVe3v7DuVI1B4sfomI2iCTyRAeHo7MzEykpKSgvLwceXl5WL58ORwdHbF06VJVWw8PD9y7dw9ZWVlQKpUoLi7G9evXmx2zX79+uH37Nq5du4aKigpVMdvY2Ij79++jvr4eFy5cQFhYGAYOHIiFCxd2SowjR450y1RnZmZmcHNzw61btzTar2n4w+Pzw2pyDdobZ9GiRdi3bx+SkpJQXl6OhoYG3Lp1S7VQQ0hICOzt7dtcXlkul+P48ePIyclBeXk5lEolzp07hwULFkAul2P16tUAgIsXL2Lr1q3Ys2cPpFJps+XPt2/frjqmvsVt0tRXvL29W41P1FlY/BIRPcHmzZsRGxuLqKgo2NjYYNy4cXB1dUVubi7kcrmq3YoVKzBhwgTMnTsXQ4cORXR0tOprXF9fX9XKVsuXL4ednR28vLwwZcoU3Lt3D8DD8Y7e3t4wNTXF2LFjMWTIEHz11VdqY2c7GqO7TJ06Ffn5+Wrjdz/77DN4eHjgypUr8PHxweuvv95svzFjxqiKqEe15xokJSVhx44dAIARI0bgxx9/xJ49exAeHg4AmDRpEgoKCgAACQkJWLVqFeLi4tC/f384OjoiLCwM9+/fB/Dwa/iioiJkZ2e3+h5lMhmee+45vPLKK3BycoJCoUBQUBBcXV3x3XffqR6oExrMXatvcZucOXMGTk5OGDFihMb7EmlKIrTppaS3JBIJ0tLSMGfOHF2nQqSVoKAgAEBGRoaOM1G3bNkyZGRkoKSkRNepNJOeno7g4GCNiprLly9j2LBhSE5Oxvz587swu67R2NiI8ePHY+HChVi8eDHjdqGSkhI4OzsjJiZG9R+V9uqpn2fq0TJ455eIqIfQ5AGxns7DwwNRUVGIiopSW/a2N2hoaEBWVhYqKioQEhLCuF0sMjISzzzzDEJDQ7s9NuknFr9ERNQlIiIiEBQUhJCQEI0fftOl3NxcHDhwAEeOHGn3XMWMq534+HicP38ehw8fhlQq7dbYpL9Y/BIR6di6deuQnJyMsrIyDB48GPv379d1Sp3mnXfeQWhoKN59911dp9JuEydOxCeffAIHBwfG7ULZ2dl48OABcnNzYW1t3a2xSb8Z6ToBIiJ9Fxsb2+ICAX2Fv78//P39dZ0G9TAzZszAjBkzdJ0G6SHe+SUiIiIivcHil4iIiIj0BotfIiIiItIbLH6JiIiISG+w+CUiIiIivcEV3kgjEolE1ykQERGpBAYGcoU30kQGpzojjaSlpek6BaJe79SpU0hISODniagTuLi46DoF6mV455eIqJulp6cjODgY/PVLRNTtMjjml4iIiIj0BotfIiIiItIbLH6JiIiISG+w+CUiIiIivcHil4iIiIj0BotfIiIiItIbLH6JiIiISG+w+CUiIiIivcHil4iIiIj0BotfIiIiItIbLH6JiIiISG+w+CUiIiIivcHil4iIiIj0BotfIiIiItIbLH6JiIiISG+w+CUiIiIivcHil4iIiIj0BotfIiIiItIbLH6JiIiISG+w+CUiIiIivcHil4iIiIj0BotfIiIiItIbLH6JiIiISG+w+CUiIiIivcHil4iIiIj0BotfIiIiItIbLH6JiIiISG+w+CUiIiIivcHil4iIiIj0BotfIiIiItIbLH6JiIiISG+w+CUiIiIivWGk6wSIiPqy4uJifPbZZ2rbzp49CwB4//331bYrFArMnTu323IjItJHEiGE0HUSRER91YMHD2BnZ4fKykoYGhoCAJp+7UokElU7pVKJBQsW4KOPPtJFmkRE+iKDwx6IiLqQiYkJAgMDYWRkBKVSCaVSifr6etTX16v+rVQqAQDz5s3TcbZERH0fi18ioi42b9481NXVtdnGysoKv/3tb7spIyIi/cXil4ioi02YMAG2tratvi6VSjF//nwYGfExDCKirsbil4ioixkYGODFF1+EVCpt8XWlUskH3YiIugmLXyKibjB37lzV2N7HDRgwAL6+vt2cERGRfmLxS0TUDX79619j0KBBzbYbGxtjwYIFajM/EBFR12HxS0TUTV566aVmQx/q6uo45IGIqBux+CUi6iYvvvhis6EPHh4e8Pb21lFGRET6h8UvEVE38fT0hJeXl2qIg1QqxaJFi3ScFRGRfmHxS0TUjV5++WXVSm/19fUc8kBE1M1Y/BIRdaO5c+eioaEBADBy5EgMHjxYxxkREekXFr9ERN1o4MCB+M1vfgMAWLBggY6zISLSP71+OaGgoCBdp0BEpJEHDx5AIpHg+PHj+Prrr3WdDhFRu/n6+mL16tW6TqNDev2d3/379+PWrVu6ToOIqN2cnZ1hb28PmUym61S63HfffYfvvvtO12n0Krdu3cL+/ft1nQZRM9999x1OnTql6zQ6rNff+QWAVatWYc6cObpOg4io3S5fvgwPDw9dp9Hlmr6dy8jI0HEmvUd6ejqCg4N5zqjH6Svftvf6O79ERL2RPhS+REQ9EYtfIiIiItIbLH6JiIiISG+w+CUiIiIivcHil4iIiIj0BotfIiLq8Q4fPgxLS0t88cUXuk6lxztx4gQiIiJw4MABuLm5QSKRQCKR4KWXXmrW1t/fHwqFAoaGhnj66afx/fff6yDj9ouLi4OnpydMTU0hl8vh6emJjRs3ory8XK1dVFQUvLy8YGFhARMTE3h4eODNN99EZWWl3sb9/PPPERcXp1phUp+x+CUioh5PCKHrFHqFzZs3Y+fOnVi3bh1mz56NH3/8Ee7u7ujfvz9SUlJw6NAhtfbHjx9HRkYGpk2bhvz8fIwcOVJHmbfPN998gyVLluDGjRu4c+cOoqOjERcXh8DAQLV2OTk5eO2113Dt2jXcvXsXsbGxSEhI0Hqqrr4Qd/r06ZDJZJg4cSJKS0u1yqfPEL0cAJGWlqbrNIiIqAWBgYEiMDBQ12l0qurqauHr69tlx09LSxPa/Hl+9913xZAhQ0RNTY3adnd3d/HJJ58IAwMD4eTkJEpLS9VeP3LkiJgxY0aHcu4uAQEBzd5fUFCQACBu376t2jZ16lRRX1+v1m7OnDkCgLhx44Zexw0NDRW+vr5CqVRqnE8f+Tyn884vERGRBj788EMUFRXpOg01ly9fxsaNG/H222+3uHKgn58fwsLCUFhYiDVr1uggw86RmZnZ7P05OTkBgNpX/AcPHoShoaFaOxsbGwBAdXW1XseNjIzE+fPnkZCQoHE+fQWLXyIi6tFOnjyJgQMHQiKRYNeuXQCApKQkyOVymJmZITs7G5MnT4aFhQWcnZ2xb98+1b47d+6ETCaDnZ0dli1bBkdHR8hkMvj5+eH06dOqdqGhoTA2NoaDg4Nq26uvvgq5XA6JRIK7d+8CAMLCwhAeHo4rV65AIpGoFis5evQoLCws8M4773THKWlm586dEEJg+vTprbaJiYnBkCFD8MEHH+DEiRNtHk8Igfj4eAwbNgwmJiawtrbGzJkzcenSJVWb9l4DAGhoaMCmTZswcOBAmJqaYsSIEUhLS+vYm/5/BQUFsLKywqBBg9psV1hYCFNTUwwePFiv41pbW2PcuHFISEjQ2+FELH6JiKhHe/755/Htt9+qbVuxYgVWrVqFmpoaKBQKpKWl4cqVK3Bzc8OSJUugVCoBPCxqFy5ciOrqaqxcuRLXrl3D999/j/r6erzwwgu4efMmgIfF45w5c9Ri7N69G2+//bbatoSEBEybNg3u7u4QQuDy5csAoHqIqLGxsUvOwZMcOnQIQ4cOhZmZWattTE1N8dFHH8HAwABLlixBVVVVq20jIyMRERGB9evXo6ioCF9//TVu3ryJsWPH4s6dOwDafw0A4K233sLWrVuxY8cO/PTTT5g2bRrmzZuHs2fPavV+lUolCgsLsWvXLpw4cQKJiYkwNjZutX11dTVycnKwZMmSNtvpS9xnn30WhYWF+Mc//qF1Tr0Zi18iIurV/Pz8YGFhAVtbW4SEhKCqqgo3btxQa2NkZKS6i+nl5YWkpCRUVFQgOTm5U3KYOnUqysvLsXHjxk45niaqqqpw9epVuLu7P7Gtr68vVq1ahWvXruGtt95qsU1NTQ3i4+Mxa9YszJ8/H5aWlvD29sZ7772Hu3fv4v3332+2T1vXoLa2FklJSQgICMDs2bNhZWWFDRs2QCqVan3+XVxc4OzsjMjISGzduhXBwcFtto+NjYWjoyNiYmK0itfX4j711FMAgLy8vA7l1Vux+CUioj6j6S7Xo3cdWzJ69GiYmZmpfY3fWxUVFUEI0eZd30fFxMRg6NCh2L17N06ePNns9fz8fFRWVmL06NFq2318fGBsbKw2XKQlj1+DH374AdXV1Rg+fLiqjampKRwcHLQ+/zdv3kRRURE+/fRTfPzxx3j22WdbHYedmZmJ9PR0HDt2DAqFQqt4fS1uU19puouvb1j8EhGRXjIxMUFxcbGu0+iw2tpaAA/fT3vIZDIkJydDIpFg8eLFqKmpUXu9aRosc3PzZvtaWVmhoqJCo/yahlds2LBBNeewRCLB9evXtXoIDACkUilsbW3h7++P1NRU5OfnIzY2tlm71NRUbNmyBbm5uXB1ddUqVl+Ma2pqCuCXvqNvWPwSEZHeUSqVKC0thbOzs65T6bCmQkaTxQt8fX2xevVqFBQUIDo6Wu01KysrAGixyNXmnNna2gIAduzYASGE2s+pU6c0OlZLPDw8YGhoiPz8fLXtiYmJSElJQU5ODgYMGNDhOH0pbl1dHYBf+o6+YfFLRER6Jzc3F0IIjBkzRrXNyMjoicMleiI7OztIJBKUlZVptF90dDQ8PT1x7tw5te3Dhw+Hubl5s4fRTp8+jbq6OowaNUqjOC4uLpDJZDh//rxG+z2upKQE8+bNa7a9oKAADQ0NcHFxAfBwpoq1a9ciLy8PWVlZLd7B1ve4TX3F3t6+Qzn2Vix+iYioz2tsbMT9+/dRX1+PCxcuICwsDAMHDsTChQtVbTw8PHDv3j1kZWVBqVSiuLgY169fb3asfv364fbt27h27RoqKiqgVCpx5MgRnU11ZmZmBjc3N9y6dUuj/ZqGPzw+P6xMJkN4eDgyMzORkpKC8vJy5OXlYfny5XB0dMTSpUs1jrNo0SLs27cPSUlJKC8vR0NDA27duoWffvoJABASEgJ7e/s2l1eWy+U4fvw4cnJyUF5eDqVSiXPnzmHBggWQy+VYvXo1AODixYvYunUr9uzZA6lUqjbUQiKRYPv27apj6lvcJk19xdvbu9X4fRmLXyIi6tF27doFHx8fAMDatWsxY8YMJCUlYceOHQCAESNG4Mcff8SePXsQHh4OAJg0aRIKCgpUx6itrYW3tzdMTU0xduxYDBkyBF999ZXaONkVK1ZgwoQJmDt3LoYOHYro6GjV18K+vr6qadGWL18OOzs7eHl5YcqUKbh37163nIe2TJ06Ffn5+Wrjdz/77DN4eHjgypUr8PHxweuvv95svzFjxqiKqEdt3rwZsbGxiIqKgo2NDcaNGwdXV1fk5hD6SG0AACAASURBVOZCLpcDgEbXICEhAatWrUJcXBz69+8PR0dHhIWF4f79+wAefg1fVFSE7OzsVt+jTCbDc889h1deeQVOTk5QKBQICgqCq6srvvvuO9UDdZrMXatvcZucOXMGTk5OGDFihMb79gUS0ctnOJZIJEhLS2s2PyMREeleUFAQACAjI0NnOSxbtgwZGRkoKSnRWQ6aSE9PR3BwsEZFzeXLlzFs2DAkJydj/vz5XZhd12hsbMT48eOxcOFCLF68mHG7UElJCZydnRETE6P6j0p79YTPcyfI4J1fIiLq8zR5GKw38vDwQFRUFKKiotSWve0NGhoakJWVhYqKCoSEhDBuF4uMjMQzzzyD0NDQbo/dU7D4JSIi6gMiIiIQFBSEkJAQjR9+06Xc3FwcOHAAR44cafdcxYyrnfj4eJw/fx6HDx+GVCrt1tg9iV4Vvz4+PjA0NMQzzzzT7bEXLVoEmUwGiUSit/PqPW779u2qp5Tfe+891fbDhw/D0tISX3zxRZfG76447REVFQUvLy9YWFjAxMQEHh4eePPNN7W6g3PgwAG4ubk1e+jByMgINjY2+N3vfofMzMwueBfq2tvnH8/3pZdeatbG398fCoUChoaGePrpp9t8SKQnYN/uOdatW4fk5GSUlZVh8ODB2L9/v65T6lLvvPMOQkND8e677+o6lXabOHEiPvnkEzg4ODBuF8rOzsaDBw+Qm5sLa2vrbo3d0+hV8XvmzBlMmDBBJ7GTk5OxZs0ancTuqdasWYNvv/222fbuGobek4a75+Tk4LXXXsO1a9dw9+5dxMbGIiEhQTW+ShOzZ8/Gjz/+CHd3d1haWqrm0ywuLkZaWhoKCwsxe/ZspKWldcE7+UV7+/yj+fbv3x8pKSk4dOiQWpvjx48jIyMD06ZNQ35+PkaOHNlVaXcK9u2eIzY2Fg8ePIAQAlevXkVgYKCuU+py/v7+2LJli67ToB5mxowZiIiIaDa7hz7Sq+K3iUQi6fAxampq4Ofn1wnZ0OOmTp2KsrIyTJs2rdOO2dL16oo42jI3N8fSpUvRr18/KBQKzJkzBwEBATh69KjqCfOOsra2xsSJE/HHP/4RwMOHajTRHX1+586dMDAwwNKlS3vV17btpY99m4iop9HL4rczxrl8+OGHra6r/SSdUXyTZjpyvbrDwYMHm/1v3MbGBgC0Xv6zNU1LXjYtYdpe3dHn/fz8EBYWhsLCQn5T0k49vW8TEfU0eln8Xr58GZ6enpDL5ao5H0+ePKnW5ptvvoGXlxcsLS0hk8ng7e2NY8eOAQDCwsIQHh6OK1euQCKRwMPDQ7Xf3r17MXr0aMhkMsjlcri6uqotHWlgYIBDhw5h8uTJsLS0hKOjI/7yl79o/B6SkpIgl8thZmaG7OxsTJ48GRYWFnB2dsa+ffvU2gohEB8fj2HDhsHExATW1taYOXMmLl26pGqzdetWmJmZQaFQoKioCOHh4XBycsLy5cshl8thYGCAUaNGwd7eHlKpFHK5HCNHjsTYsWNVq/dYWVnhzTffbPd5bMnJkycxcOBASCQS7Nq1C8DD6/X4+NWmn7/+9a9aXa+W4rT3XGly7juisLAQpqamGDx4sGrb0aNHOzyR/oULFwAA48aNU9veU/p8TEwMhgwZgg8++AAnTpxo872wb/fOvk1EpFOilwMg0tLS2t1+4sSJws3NTVy9elUolUrxz3/+U/zmN78RMplM/Pvf/1a1y8jIEJGRkeLevXuipKREjBkzRvTv31/1+uzZs4W7u7vasXfs2CEAiHfffVeUlJSIe/fuiT//+c/ixRdfFEIIsX79egFAfPnll6K0tFTcu3dPTJkyRZiYmIiqqiqN3/ujxysrKxNFRUVi7NixQi6Xi7q6OlW7TZs2CWNjY7F3715RWloqLly4IEaOHClsbGzEzz//3Ox4K1euFImJiWLWrFniX//6l9i8ebMAIE6fPi2qqqrE3bt3xaRJkwQAcejQIVFcXCyqqqpEaGioACDOnz/f7vNYUFAgAIg//elPqm03b94UAERiYqKqzVtvvaU6Rz/99JOwtrYWfn5+oqGhQevr9Xgcbc7Vk869tqqqqoRCoRChoaFq2w8ePCgUCoWIiop64jHc3d2FpaWl6t/V1dXiyJEjYtCgQcLf319UVlaqtdd1n3d3dxdXr14VQgjx7bffCgMDA+Hq6qrK88iRI2LGjBlq+7Bv9/y+HRgYKAIDAzXeT5+lpaWJPvDnmfqgPvJ5Tu/1ny5tit9f/epXatsuXLggAIg1a9a0ul9sbKwAIIqKioQQzf/g1NXVCSsrKzFhwgS1/err60VCQoIQ4pc/KjU1NarX//u//1sAEP/85z/b/R6atHS83bt3CwDi8uXLQoiHBY+5ubkICQlR2/dvf/ubAKBWRLV0PCGEqkCoqKhQbfv4448FAJGXl9fsmKmpqa3m/Ph5bE+B8LiAgAAhk8nEpUuX2h2nPQVCR8/V4+e+I9avXy+GDBkiysvLtT6Gu7u7ANDsx9vbW3z88cfiwYMHbe7f3X3+0eJXCCHCw8MFAPHaa68JIZoXv+zbvaNv95E/lt2KxS/1VH3k85xu1NE7x32Bt7c3LC0tVV8Ht6RpnHBrE6VfuHABpaWl+M///E+17YaGhli5cuUTj6tUKjVNu0XGxsZqx8vPz0dlZSVGjx6t1s7HxwfGxsY4ffp0h+LU19ertrXnvTzpPD5Jeno6PvvsM8TFxWHo0KGdGqej5+rxc6+tzMxMpKen4/jx41AoFB06lqWlpWpsb319Pe7cuYPjx48jNDQUsbGxOHnypGps8eN03edjYmJw8OBB7N69G8HBwc1eZ9/uPX17//79fNZBCzxn1BP1hRlTWPz+P6lUqvaL/dChQ9i2bRvy8/NRXl7+xF/65eXlAAArK6suzVNTTYWPubl5s9esrKxQUVHRpfE1PY9tKSkpweuvvw4fH59mSzJ2RhxdnysASE1NRXx8PHJzczFgwIBOPbaRkRGcnJywaNEiNDQ0YMmSJXj33Xfxhz/8AUDP6/MymQzJycl4/vnnsXjxYsTFxam9ruvrxb7dfmPGjMGqVau6NEZfcurUKSQkJHT5dIREmtqxY4euU+gULH7x8A7PvXv3MHDgQADAjRs3EBAQgFmzZuEvf/kLBgwYgMTExGYPvDyqqVC5e/dut+TcXk2FSUt/3EpLS+Hs7NxlsbU5j21ZuXIlSktLkZOTozYzQmfF0eW5AoDExEQcO3YMOTk5LRYpncnb2xsAcPHiRQA9t8/7+vpi9erV2L59O6Kjo1WfUYB9WxO67tvOzs6YM2dOl8boaxISEnjOqMfJyMjQdQqdQi9ne3jcV199hcbGRtXE+Xl5eVAqlVixYgXc3NxUq1S1xdXVFf369cPx48e7I+V2Gz58OMzNzXH27Fm17adPn0ZdXR1GjRrVZbG1OY+tOXToED755BNs3LgRTz/9tGr7G2+80WlxdHWuhBBYu3Yt8vLykJWV1eWFLwD8/e9/BwDV1+s9uc9HR0fD09MT586dU9vOvt1+ujxXREQ9jV4Wv3V1dSgrK0N9fT2+//57hIaGYtCgQVi4cCEAqO4unThxArW1tSgoKGg2Jq5fv364ffs2rl27hoqKChgYGGDdunX4+uuvERoaisLCQjQ2NqKiokJ1d00XZDIZwsPDkZmZiZSUFJSXlyMvLw/Lly+Ho6Mjli5d2mWx23Me26O8vBzLli3DM888g7feegsAUFtbi7Nnz+L8+fNaXa+WvjrW1bm6ePEitm7dij179kAqlTab8mr79u2qtkeOHNF4qrOamho0NjZCCIHbt28jOTkZGzZsgI2Njeqr6J7c55uGPzw+DzL79kM9uW8TEfVIun7krqOg4WwPycnJYsKECcLOzk4YGRmJ/v37i7lz54rr16+rtVu7dq3o16+fsLKyEkFBQWLXrl0CgHB3dxc3btwQ33//vRg0aJAwNTUVzz//vGqqoF27dglvb28hk8mETCYTzz77rNi9e7eIi4sTpqamAoB46qmnxJUrV0RKSoqwtrYWAISzs7NGMz7s3r1bmJmZqR3v/fffFxYWFgKAGDRokGrqtsbGRrFt2zbx1FNPCalUKqytrUVAQID44YcfVMd7ND8XFxexd+9eIYQQCQkJqjiurq7im2++EVu2bBGWlpYCgLC3txeffPKJSE1NFfb29gKAsLa2Fvv27XvieQwLC1PtI5fLxaxZs0RiYqJwcHAQAISZmZmYPn262L59e4uzFgAQU6ZM0ep6bdiwoVmc9p4rTc59e+Tl5bX6/gCIbdu2qdoePnxYKBQKERMT0+rxMjMzW53pwcTERDz11FNixYoV4saNG2r76arPP5qvjY2NanaHx73xxhvNpjpj3+7ZfVuIPvN0eLfibA/UU/WRz3O6RIjevQi8RCJBWloax0YREfVAQUFBAPrOWMHukJ6ejuDgYPTyP8/UB/WRz3OGXg57ICIiIiL9xOK3B7l06VKrS50++hMSEqLrVOkJeC2JSFdOnDiBiIgIHDhwAG5ubqrfNy+99FKztv7+/lAoFDA0NMTTTz+N77//XgcZt19cXBw8PT1hamoKuVwOT09PbNy4UTX1YpOoqCh4eXnBwsICJiYm8PDwwJtvvonKykq9jfv5558jLi5O67nI+xQdj7voMGg45peIiLpPHxkj2K06MuZ306ZNYtq0aWqrQ7q7u4v+/fsLAOLgwYPN9mlp6fCeaurUqWL79u2iqKhIVFRUiPT0dCGVSsULL7yg1m7cuHFi9+7doqSkRJSXl4u0tDQhlUrFpEmT9DpuQkKCGDdunLh//75W+fSRz3M67/wSEVGfVlNTAz8/v14f40m2bNmC1NRUpKenN1sdcufOnTAwMMDSpUtRVlamoww7ztjYGK+++ipsbW1hbm6OoKAgzJw5E3/961/x008/qdqZm5tj6dKl6NevHxQKBebMmYOAgAAcPXoUN2/e1Nu4K1euxK9+9StMmTJFbRVLfcPil4iI+rQPP/wQRUVFvT5GWy5fvoyNGzfi7bffhkwma/a6n58fwsLCUFhYiDVr1uggw86RmZnZ7P05OTkBgNpX/AcPHmw2PWLTUu7V1dV6HTcyMhLnz59HQkKCxvn0FSx+iYioRxFCID4+HsOGDYOJiQmsra0xc+ZMXLp0SdUmNDQUxsbGcHBwUG179dVXIZfLIZFIVCsPhoWFITw8HFeuXIFEIoGHhwd27twJmUwGOzs7LFu2DI6OjpDJZPDz81ObR7kjMQDg6NGjGs/Lra2dO3dCCIHp06e32iYmJgZDhgzBBx98gBMnTrR5vPZcg6SkJMjlcpiZmSE7OxuTJ0+GhYUFnJ2dsW/fPrXjNTQ0YNOmTRg4cCBMTU0xYsSITlu+uaCgAFZWVhg0aFCb7QoLC2FqaorBgwfrdVxra2uMGzcOCQkJ+jujiG6HXXQcOOaXiKjH0maM4KZNm4SxsbHYu3evKC0tFRcuXBAjR44UNjY2qvmlhRDixRdfFPb29mr7btu2TQAQxcXFqm2zZ88W7u7uau2WLl0q5HK5uHjxoqitrRX5+fnCx8dHKBQKtTmwOxLj4MGDQqFQiKioKI3evzZjft3c3ISXl1eLr7m7u4urV68KIYT49ttvhYGBgXB1dRWVlZVCiJbH/Lb3Gqxfv14AEF9++aUoKysTRUVFYuzYsUIul4u6ujpVuzVr1ggTExOxf/9+cf/+fbFu3TphYGAgzpw5o9H7bFJXVydu3bolEhMThYmJiWr+7tZUVVUJhUIhQkNDtYrX1+JGREQIAOLcuXMa5cExv0RERJ2spqYG8fHxmDVrFubPnw9LS0t4e3vjvffew927d/H+++93WiwjIyPVnU0vLy8kJSWhoqICycnJnXL8qVOnory8HBs3buyU47WmqqoKV69ehbu7+xPb+vr6YtWqVbh27ZpqVcHHaXMN/Pz8YGFhAVtbW4SEhKCqqgo3btwA8HDVwqSkJAQEBGD27NmwsrLChg0bIJVKtT7XLi4ucHZ2RmRkJLZu3Yrg4OA228fGxsLR0RExMTFaxetrcZ966ikAD5dq10csfomIqMfIz89HZWUlRo8erbbdx8cHxsbGWi0j3V6jR4+GmZmZ2lf7vUFRURGEEDAzM2tX+5iYGAwdOhS7d+/GyZMnm73e0WtgbGwMAKqltn/44QdUV1dj+PDhqjampqZwcHDQ+lzfvHkTRUVF+PTTT/Hxxx/j2WefbXXMdWZmJtLT03Hs2LFmDwLqa9ymvnLnzp0O5dVbsfglIqIeo7S0FMDDp9cfZ2VlhYqKii6Nb2JiguLi4i6N0dlqa2sBPMy9PWQyGZKTkyGRSLB48WLU1NSovd7Z16CqqgoAsGHDBrV5zq9fv67VQ2AAIJVKYWtrC39/f6SmpiI/Px+xsbHN2qWmpmLLli3Izc2Fq6urVrH6YlxTU1MAv/QdfcPil4iIegwrKysAaLHAKi0thbOzc5fFViqVXR6jKzQVMposXuDr64vVq1ejoKAA0dHRaq919jWwtbUFAOzYsQNCCLWfU6dOaXSslnh4eMDQ0BD5+flq2xMTE5GSkoKcnBwMGDCgw3H6Uty6ujoAv/QdfcPil4iIeozhw4fD3NwcZ8+eVdt++vRp1NXVYdSoUaptRkZGqq/WO0Nubi6EEBgzZkyXxegKdnZ2kEgkGs/fGx0dDU9PT5w7d05tuybXoD1cXFwgk8lw/vx5jfZ7XElJCebNm9dse0FBARoaGuDi4gLg4UwVa9euRV5eHrKyslq8g63vcZv6ir29fYdy7K1Y/BIRUY8hk8kQHh6OzMxMpKSkoLy8HHl5eVi+fDkcHR2xdOlSVVsPDw/cu3cPWVlZUCqVKC4uxvXr15sds1+/frh9+zauXbuGiooKVTHb2NiI+/fvo76+HhcuXEBYWBgGDhyIhQsXdkqMI0eOdMtUZ2ZmZnBzc8OtW7c02q9p+MPj88Nqcg3aG2fRokXYt28fkpKSUF5ejoaGBty6dUu1UENISAjs7e3bXF5ZLpfj+PHjyMnJQXl5OZRKJc6dO4cFCxZALpdj9erVAICLFy9i69at2LNnD6RSabNl5bdv3646pr7FbdLUV7y9vVuN35ex+CUioh5l8+bNiI2NRVRUFGxsbDBu3Di4uroiNzcXcrlc1W7FihWYMGEC5s6di6FDhyI6Olr1Na6vr69qZavly5fDzs4OXl5emDJlCu7duwfg4XhHb29vmJqaYuzYsRgyZAi++uortbGzHY3RXaZOnYr8/Hy18bufffYZPDw8cOXKFfj4+OD1119vtt+YMWNURdSj2nMNkpKSsGPHDgDAiBEj8OOPP2LPnj0IDw8HAEyaNAkFBQUAgISEBKxatQpxcXHo378/HB0dERYWhvv37wN4+DV8UVERsrOzW32PMpkMzz33HF555RU4OTlBoVAgKCgIrq6u+O6771QP1AkN5q7Vt7hNzpw5AycnJ4wYMULjffsCidDmrPUgEokEaWlpmDNnjq5TISKixwQFBQEAMjIydJyJumXLliEjIwMlJSW6TqWZ9PR0BAcHa1TUXL58GcOGDUNycjLmz5/fhdl1jcbGRowfPx4LFy7E4sWLGbcLlZSUwNnZGTExMar/qLRXT/08ayiDd36JiEgvafKAWE/n4eGBqKgoREVFqS172xs0NDQgKysLFRUVCAkJYdwuFhkZiWeeeQahoaHdHrunYPFLRETUB0RERCAoKAghISEaP/ymS7m5uThw4ACOHDnS7rmKGVc78fHxOH/+PA4fPgypVNqtsXsSFr9ERKRX1q1bh+TkZJSVlWHw4MHYv3+/rlPqNO+88w5CQ0Px7rvv6jqVdps4cSI++eQTODg4MG4Xys7OxoMHD5Cbmwtra+tujd3TGOk6ASIiou4UGxvb4gIBfYW/vz/8/f11nQb1MDNmzMCMGTN0nUaPwDu/RERERKQ3WPwSERERkd5g8UtEREREeoPFLxERERHpjT7xwNupU6d0nQIREbWgaRnV9PR0HWfSezT9TeM5o57m1q1bcHZ21nUaHdYnVngjIiIioq4XGBjY61d46/V3fnt57U5Eekib5WuJiKhzcMwvEREREekNFr9EREREpDdY/BIRERGR3mDxS0RERER6g8UvEREREekNFr9EREREpDdY/BIRERGR3mDxS0RERER6g8UvEREREekNFr9EREREpDdY/BIRERGR3mDxS0RERER6g8UvEREREekNFr9EREREpDdY/BIRERGR3mDxS0RERER6g8UvEREREekNFr9EREREpDdY/BIRERGR3mDxS0RERER6g8UvEREREekNFr9EREREpDdY/BIRERGR3mDxS0RERER6g8UvEREREekNFr9EREREpDdY/BIRERGR3mDxS0RERER6g8UvEREREekNFr9EREREpDdY/BIRERGR3mDxS0RERER6g8UvEREREekNI10nQETUl926dQsLFixAQ0ODatv9+/ehUCgwfvx4tbZDhw7Fn//8527OkIhIv7D4JSLqQs7Ozrh+/TquXLnS7LX/+Z//Ufv3f/zHf3RXWkREeovDHoiIutjLL78MqVT6xHYhISHdkA0RkX5j8UtE1MVefPFF1NfXt9nm6aefhpeXVzdlRESkv1j8EhF1MXd3d4wYMQISiaTF16VSKRYsWNDNWRER6ScWv0RE3eDll1+GoaFhi6/V19cjKCiomzMiItJPLH6JiLrB3Llz0djY2Gy7gYEBxowZA1dX1+5PiohID7H4JSLqBo6OjnjuuedgYKD+a9fAwAAvv/yyjrIiItI/LH6JiLrJSy+91GybEAKzZs3SQTZERPqJxS8RUTcJDAxUG/draGiI3/3ud7Czs9NhVkRE+oXFLxFRN7G2tsYLL7ygKoCFEJg/f76OsyIi0i8sfomIutH8+fNVD75JpVLMnDlTxxkREekXFr9ERN1o+vTpMDExAQBMmzYN5ubmOs6IiEi/sPglIupGcrlcdbeXQx6IiLqfRAghdJ1EbxcUFIT9+/frOg0iIiLqw9LS0jBnzhxdp9HbZRjpOoO+YsyYMVi1apWu0yCiXqChoQFpaWmYN29el8YJDg5GWFgYfH19uzROX7Jjxw4A4O9z6nGCg4N1nUKfweK3kzg7O/N/Y0TUbgEBAZDJZF0aIzg4GL6+vvzdpIGMjAwA4DmjHofFb+fhmF8iIh3o6sKXiIhaxuKXiIiIiPQGi18iIiIi0hssfomIiIhIb7D4JSIiIiK9weKXiIjadPjwYVhaWuKLL77QdSo93okTJxAREYEDBw7Azc0NEokEEokEL730UrO2/v7+UCgUMDQ0xNNPP43vv/9eBxm3X1xcHDw9PWFqagq5XA5PT09s3LgR5eXlau2ioqLg5eUFCwsLmJiYwMPDA2+++SYqKyv1Nu7nn3+OuLg4NDQ0aJULdS4Wv0RE1CauhdQ+mzdvxs6dO7Fu3TrMnj0bP/74I9zd3dG/f3+kpKTg0KFDau2PHz+OjIwMTJs2Dfn5+Rg5cqSOMm+fb775BkuWLMGNGzdw584dREdHIy4uDoGBgWrtcnJy8Nprr+HatWu4e/cuYmNjkZCQgKCgIL2NO336dMhkMkycOBGlpaVa5UOdSFCHBQYGisDAQF2nQUSkBoBIS0vTdRqdqrq6Wvj6+nbZ8bX9ff7uu++KIUOGiJqaGrXt7u7u4pNPPhEGBgbCyclJlJaWqr1+5MgRMWPGjA7l3F0CAgKavb+goCABQNy+fVu1berUqaK+vl6t3Zw5cwQAcePGDb2OGxoaKnx9fYVSqdQ4n774edaRdN75JSKiXuPDDz9EUVGRrtNQc/nyZWzcuBFvv/12i/M3+/n5ISwsDIWFhVizZo0OMuwcmZmZzd6fk5MTAKh9xX/w4EEYGhqqtbOxsQEAVFdX63XcyMhInD9/HgkJCRrnQ52HxS8REbXq5MmTGDhwICQSCXbt2gUASEpKglwuh5mZGbKzszF58mRYWFjA2dkZ+/btU+27c+dOyGQy2NnZYdmyZXB0dIRMJoOfnx9Onz6tahcaGgpjY2M4ODiotr366quQy+WQSCS4e/cuACAsLAzh4eG4cuUKJBIJPDw8AABHjx6FhYUF3nnnne44Jc3s3LkTQghMnz691TYxMTEYMmQIPvjgA5w4caLN4wkhEB8fj2HDhsHExATW1taYOXMmLl26pGrT3msAPFxOe9OmTRg4cCBMTU0xYsQIpKWldexN/7+CggJYWVlh0KBBbbYrLCyEqakpBg8erNdxra2tMW7cOCQkJHA4kQ6x+CUiolY9//zz+Pbbb9W2rVixAqtWrUJNTQ0UCgXS0tJw5coVuLm5YcmSJVAqlQAeFrULFy5EdXU1Vq5ciWvXruH7779HfX09XnjhBdy8eRPAw+Lx8eWEd+/ejbffflttW0JCAqZNmwZ3d3cIIXD58mUAUD1E1NjY2CXn4EkOHTqEoUOHwszMrNU2pqam+Oijj2BgYIAlS5agqqqq1baRkZGIiIjA+vXrUVRUhK+//ho3b97E2LFjcefOHQDtvwYA8NZbb2Hr1q3YsWMHfvrpJ0ybNg3z5s3D2bNntXq/SqUShYWF2LVrF06cOIHExEQYGxu32r66uho5OTlYsmRJm+30Je6zzz6LwsJC/OMf/9A6J+oYFr9ERKQ1Pz8/WFhYwNbWFiEhIaiqqsKNGzfU2hgZGanuYnp5eSEpKQkVFRVITk7ulBymTp2K8vJybNy4sVOOp4mqqipcvXoV7u7uT2zr6+uLVatW4dq1a3jrrbdabFNTU4P4+HjMmjUL8+fPh6WlJby9vfHee+/h7t27eP/995vt09Y1qK2tRVJSEgICAjB79mxYWVlhw4YNkEqlWp9/FxcXODs7IzIyElu3bkVwcHCb7WNjY+Ho6IiYmBit4vW1uE899RQAIC8vr0N5kfZY/BIRUadousv16F3HlowePRpmlRtwaAAAIABJREFUZmZqX+P3VkVFRRBCtHnX91ExMTEYOnQodu/ejZMnTzZ7PT8/H5WVlRg9erTadh8fHxgbG6sNF2nJ49fghx9+QHV1NYYPH65qY2pqCgcHB63P/82bN1FUVIRPP/0UH3/8MZ599tlWx2FnZmYiPT0dx44dg0Kh0CpeX4vb1Fea7uJT92PxS0RE3c7ExATFxcW6TqPDamtrATx8P+0hk8mQnJwMiUSCxYsXo6amRu31pmmwzM3Nm+1rZWWFiooKjfJrGl6xYcMG1ZzDEokE169f1+ohMACQSqWwtbWFv78/UlNTkZ+fj9jY2GbtUlNTsWXLFuTm5sLV1VWrWH0xrqmpKYBf+g51Pxa/RETUrZRKJUpLS+Hs7KzrVDqsqZDRZPECX19frF69GgUFBYiOjlZ7zcrKCgBaLHK1OWe2trYAgB07dkAIofZz6tQpjY7VEg8PDxgaGiI/P19te2JiIlJSUpCTk4MBAwZ0OE5filtXVwfgl75D3Y/FLxERdavc3FwIITBmzBjVNiMjoycOl+iJ7OzsIJFIUFZWptF+0dHR8PT0xLlz59S2Dx8+HObm5s0eRjt9+jTq6uowatQojeK4uLhAJpPh/PnzGu33uJKSEsybN6/Z9oKCAjQ0NMDFxQXAw5kq1q5di7y8PGRlZbV4B1vf4zb1FXt7+w7lSNpj8UtERF2qsbER9+/fR319PS5cuICwsDAMHDgQCxcuVLXx8PDAvXv3kJWVBaVSieLiYly/fr3Zsfr164fbt2/j2rVrqKiogFKpxJEjR3Q21ZmZmRnc3Nxw69YtjfZrGv7w+PywMpkM4eHhyMzMREpKCsrLy5GXl4fly5fD0dERS5cu1TjOokWLsG/fPiQlJaG8vBwNDQ24desWfvrpJwBASEgI7O3t21xeWS6X4/jx48jJyUF5eTmUSiXOnTuHBQsWQC6XY/Xq1QCAixcvYuvWrdizZw+kUqnaUAuJRILt27erjqlvcZs09RVvb+9W41PXYvFLRESt2rVrF3x8fAAAa9euxYwZM5CUlIQdO3YAAEaMGIEff/wRe/bsQXh4OABg0qRJKCgoUB2jtrYW3t7eMDU1xdixYzFkyBB89dVXauNkV6xYgQkTJmDu3LkYOnQooqOjVV8L+/r6qqZFW758Oezs7ODl5YUpU6bg3r173XIe2jJ16lTk5+erjd/97LPP4OHhgStXrsDHxwevv/56s/3GjBmjKqIetXnzZsTGxiIqKgo2NjYYN24cXF1dkZubC7lcDgAaXYOEhASsWrUKcXFx6N+/PxwdHREWFob79+8DePg1fFFREbKzs1t9jzKZDM899xxeeeUVODk5QaFQICgoCK6urvjuu+9UD9RpMnetvsVtcubMGTg5OWHEiBEa70udQyI4y3KHNa3fnZGRoeNMiIh+IZFIkJaW1mwO3e60bNkyZGRkoKSkRGc5aEKb3+eXL1/GsGHDkJycjPnz53dVal2msbER48ePx8KFC7F48WLG7UIlJSVwdnZGTEyM6j8q7dUTPs99RAbv/BIRUZfS5GGw3sjDwwNRUVGIiopSW/a2N2hoaEBWVhYqKioQEhLCuF0sMjISzzzzDEJDQ7s9Nv2CxS/1SNu3b1c9SPLee++pth8+fBiWlpb44osvuix2VFQUvLy8YGFhARMTE3h4eODNN9984h+1V155BQqFAhKJRKuHS7SN25IDBw7Azc1NNe7sSZP/x8fHQyKRwMDAAJ6envj66681jtneXCQSCaRSKZycnPDiiy/iX//6V6fFelxP70ctnRuJRAJjY2PY2dlh/Pjx2LZtm+rraeq5IiIiEBQUhJCQEI0fftOl3NxcHDhwAEeOHGn3XMWMq534+HicP38ehw8fhlQq7dbY9BhBHRYYGCgCAwN1nUafU1BQIACIP/3pT6ptBw8eFBYWFuLzzz/vsrjjxo0Tu3fvFiUlJaK8vFykpaUJqVQqJk2a9MR99+3bJwCIc+fOdWvc1ri7uwsAwsHBQdTV1bXYpr6+XgwaNEgAEBMnTtQ6VntysbS0FEIIUVlZKT7//HMxcOBAYW5uLi5dutRlcXtDP3r03DQ2Nor79++Lr776SixcuFBIJBLh6Ogozpw5o3EOAERaWlqH34u2IiIihLGxsQAgXF1dRUZGhs5yaa+O/j4/duyYWLt2bSdmRH1BVlaWiI2NFfX19VofQ9ef5z4knXd+qVeZOnUqysrKMG3atC6LYW5ujqVLl6Jfv35QKBSYM2cOAgICcPToUdVDN70p7qhRo/Dzzz8jKyurxdcPHDgAJycnrY+vDblcjmnTpuGPf/wjKisrkZiY2K3xe3I/kkgksLKywvjx45GcnIz09HTcuXNHlXNvEhsbiwcPHkAIgatXryIwMFDXKXU5f39/bNmyRddpUA8zY8YMRERENJvdg3SDxS/pNSEEMjIy8P7776u2HTx4sNkvKBsbGwB44opIEolE61w6ErctK1asAAD86U9/avH1+Ph4jR+86Cy//vWvAQD//Oc/dRK/s3R2P3pUYGAgFi5ciKKiIrWhG0REpB0WvzqQkJAAuVwOAwMDjBo1Cvb29pBKpZDL5Rg5ciTGjh2rmpjcysoKb775ptr+33zzDby8vGBpaQmZTAZvb28cO3YMAPDRRx/B3NwcEokE1tbWyMrKwtmzZzFo0CAYGhq2OGl3W3bu3AmZTAY7OzssW7YMjo6OkMlk8PPza7bGvBAC8fHxGDZsGExMTGBtbY2ZM2c2Wz++ve0ed/LkSQwcOBASiQS7du0C8HC6H7lcDjMzM2RnZ2Py5MmwsLCAs7Mz9u3bp7Z/Q0MDYmNjMXToUJiamsLGxgaDBw9GbGzsE5+eLSwshKmpKQYPHqz2PrZt24ahQ4fCxMQElpaWeOONN554TjXRUtyjR49qNKfpb3/7WwwbNgxfffUVfvjhB7XX/vd//xfV1dXw9/dvcd+u7mv19fUA1JeG1bd+1B5N8+EeOXJEo/2IiKgFuhty0XdoM0Zs8+bNAoA4ffq0qKqqEnfv3hWTJk0SAMShQ4dEcXGxqKqqEqGhoQKAOH/+vGrfjIwMERkZKe7duydKSkrEmDFjRP/+/VWvX7x4UZiZmYkFCxaotkVERIgPPvhAq/e3dOlSIZfLxcWLF0Vtba3Iz88XPj4+QqFQiBs3bqjabdq0SRgbG4u9e/eK0tJSceHCBTFy5EhhY2Mjfv75Z43btTRW8+bNmwKASExMVG1bv369ACC+/PJLUVZWJoqKisTYsWOFXC5XG+f6zjvvCENDQ5GdnS2qq6vF3//+d2Fvby/Gjx/f5vuvqqoSCoVChIaGqm1fv369kEgk4g9/+IO4f/++qK6uFrt379Z6zG974x48eFAoFAoRFRX1xGO4u7uLq1evij/+8Y8CgAgLC1N7PSAgQCQnJ4uKiooWx/x2Zl97dFxrk7179woA4o033lBt07d+1Nq5eVR5ebkAIFxcXNqM8ThwjKDG+AwH9VT8PHeadBa/naAjxW9FRYVq28cffywAiLy8PNW2v/3tbwKASE1NbfVYsbGxAoAoKipSbfvzn/8sAIiUlBTx6aefitWrV2uU36OWLl3a7A/zmTNnBADx9ttvCyGEqK6uFubm5iIkJEStXVP+TYVae9sJoXnRUlNTo9rWVIRevnxZtc3Hx0f8+te/Vov7+9//XhgYGIgHDx60+v7Xr18vhgwZIsrLy1XbqqurhZmZmXjhhRfU2nbkgbf2xNVUU/FbWloq5HK5sLa2FtXV1UIIIa5cuSKcnZ3FgwcPWi1+H9eRvvb4A2/79+8X9vb2ws7OTty6dUsIoX/9qKVz0xqJRCKsrKzabPM4/rHUHItf6qn4ee406UZdd0+ZNGVsbAzgl6+CAaimQ2lrzfumNo/Opfn73/8ef/3rX7Fs2TL87ne/w/79+zs119GjR8PMzEz1FXN+fj4qKysxevRotXY+Pj4wNjZWDZFob7uOajqXj5632tpayGQytXYNDQ2QSqWtPoSQmZmJ9PR0HD9+HAqFQrX98uXLqK6uxsSJEzsl3/bG1ZalpSXmzZuHPXv2IDU1FYsWLcKOHTuwYsUKGBsbo66url3H6WhfKysrg0QigaGhIRwcHDBlyhRs3rxZ9cCdvvWj9qqqqoIQAhYWFhrve+rUKY330WdNS8+mp6frOBMi6iosfnuhQ4cOYdu2bcjPz1etOd6Sd955B/v370dRUVGX5GFiYoLi4mIAQGlpKYCHT7g/zsrKChUVFRq16wpTpkzBtm3bkJ2dDX9/f+Tn5yMrKwv/9V//1WLRkpqaivj4eOTm5mLAgAFqrzX9gbS1te30PNuK2xErVqzAnj178N577yEgIAAZGRlPnGO3s/uapaWlqg+0RN/6UXv9+9//BgB4enpqvG9CQgISEhK0iqvPgoODdZ0CEXURPvDWy9y4cQMBAQFwcHDA6dOnUVZWhri4uGbtlEolVq5cifj4eJw6dQoxMTGdmodSqURpaSmcnZ0BPCw4ALRYdGjTritERkbit7/9LRYuXAgLCwvMmjULc+bMwZ49e5q1TUxMREpKCnJyclosWJru/D148KBTc3xS3I545plnMGbMGPztb3/D0qVLERQUBGtr61bb66Kv6Vs/aq+jR48CACZPnqzxvmlpaRBC8KedP4GBgfg/9u49qqkr7R/4N0JCwh0VNIKgEgURrVqdEdSxvk5dbR3vInibUldZeCsi9K3aqlVAKtQBlgrtqnXRzljlIhQ7XlpfF2W074i9qMiL0xaxKKIVRRQELAH27w9/pMYAEiAEzPezFn+4s895Hs45wScn++y9YMECo+fBH/48+UOdh3d+e5j8/Hyo1WqsWrUKQ4YMAdD89FpvvPEGgoKCMG/ePJSWliIyMhLTp0+Hj49Pp+SRk5MDIQQmTJgAAPD29oa1tTW+//57rX5nz55FXV0dnn/+eb36GUJBQQGKiopw+/ZtmJs3f+kLIbBhwwZUVFQgKyurxX7e3t7o1asX/vWvf2HlypUdzq2tcTtq1apVyM3NxaFDh1BYWNhqX2Nca6Z2HbXFr7/+ivj4eLi4uGD58uXt3g8RET3CO789jKurKwDg5MmTePjwIQoLC3XGNyYmJsLZ2Rnz5s0D8GiieS8vLyxZsgSVlZXtitvY2IiKigrU19fj4sWLCA0Nhaurq2YKJrlcjvDwcGRmZmL//v2orKxEfn4+Vq5cCaVSieDgYL36GcKaNWvg6ura6nLBly5dQmxsLPbu3QupVKqz7OzOnTsBPBruMH/+fBw6dAj79u1DZWUlLl68qDXPqz7aGhd4NN2VPlOdPW7hwoXo27cv5s6dqyloW2KMa83UrqPHCSHw4MEDNDY2QgiB27dvIzU1FRMnToSZmRmysrLaNeaXiIieIKjD9H06OCEhQVhaWmqW/Dx9+rTYsWOHsLOzEwBEv379xGeffSZSUlJEv379BADh4OAgDh48KIQQYv369aJ3797C3t5e+Pn5iT179ggAwt3dXYwePVpIJBLRu3dv8e9//1sIIcS6detEr169BABhZ2cnvv/+e71+v+DgYCGVSoWzs7MwNzcXtra2Ys6cOaKoqEirX2Njo3j//ffF0KFDhVQqFQ4ODmLu3Lnip59+0rvf3/72N83vbmVlJebNmyd2794t+vfvLwAIS0tLMWvWLJGYmKg5lkOHDhVFRUXio48+Era2tgKAcHNzEz///LMQQojs7GzRp08fAUDzI5VKxfDhw0VGRoYQQoj8/Hyt15/8ef/99zU5VlVViddff1306dNHWFtbi0mTJoktW7YIAMLFxUXk5eW1+RjrE/fYsWPCxsZGREVFtbi/zMxMzdLGffv2FWvWrNG89tZbb2muDSGE2LRpk+a49urVS3h5eYnTp08LITrnWvvf//1fMWzYMM3volQqhZ+fX4u5m9J19MUXX4hRo0YJS0tLIZPJNMeuaWaHP/zhDyIiIkKUl5e3fgG1AHw6XG+c7YG6K76fO02aRAgOJOkoPz8/AEB6erqRMzGMFStWID09HeXl5cZOpUOSkpJQWFiI+Ph4TVtdXR02bNiApKQkVFRUQKFQGDFD6gl60nUkkUiQmpr61MU36HfP+t9z6rn4fu406RzzS23y+NRWPdGvv/6KkJAQXLhwQatdJpPB1dUVarUaarW62xQt1D3xOiIi6vk45tcE/fjjjzpjEJv7CQgIMHaqnUahUEAqlWLfvn24desW1Go1bty4gY8//hhbtmxBQEBAp4+nNMXj/KwzxnVERESdi8WvCfL09GzTtCopKSl4++23kZycjPv372Pw4MGdvlhGV7Gzs8OJEyfwf//3fxg2bBgUCgW8vLyQnJyMHTt24NNPP+30mPocZ+oZjHEdUc9y8uRJbNy4ERkZGRgyZIjmQ+6yZct0+k6fPh02NjYwMzPDiBEjcO7cOSNkrL/GxkbEx8fD19e3xT5qtRrR0dFQqVSQyWSwt7eHt7c3iouLe0zcmJgYeHp6QqFQwMrKCp6enti8ebPOw7wRERHw8vKCra0tLCwsoFKp8NZbb2k9FPvFF18gJiamx3+L+szo2jHGzyY+IEFE3RH4gIzeOvL3fMuWLWLmzJlaS1i7u7trHpA8cuSIzjbHjx8Xs2fPbne+Xe3nn38WEydOFADEc88912K/uXPnCg8PD5GbmyvUarW4ceOGmDVrlsjPz+8xcWfMmCF27twpysrKRFVVlUhLSxNSqVRnWfspU6aIxMREUV5eLiorK0VqaqqQSqXipZde0uqXkJAgpkyZIioqKvTORQi+nztRGu/8EhGRQdTW1rZ6l66nxGiLHTt2ICUlBWlpaTpLWO/atQu9evVCcHAw7t+/b6QMOy4vLw8bNmzAypUrMXr06Bb7paSkICsrC+np6fjjH/8Ic3NzKJVKHD58GN7e3j0mrkwmw+rVq+Ho6Ahra2v4+flhzpw5+J//+R/cvHlT08/a2hrBwcHo3bs3bGxssHDhQsydOxdffvklSkpKNP3Wrl2L5557Dq+88grq6+v1zoc6D4tfIiIyiH379hlsefWujPE0ly9fxubNm7Ft2zbN6o+P8/X1RWhoKEpLS/Hmm28aIcPO8dxzzyEjIwNLliyBhYVFi/0++OADjB07FiNHjuzRcTMzM3XOp7OzMwBoDWk4cuSIztLmffv2BQDU1NRotW/duhUXLlzgkuNGxuKXiIgAPFpoIy4uDsOHD4eFhQUcHBwwZ84c/Pjjj5o+ISEhkMlk6N+/v6Zt9erVsLKygkQiwZ07dwAAoaGhCA8PR1FRESQSCVQqFXbt2gW5XA4nJyesWLECSqUScrkcvr6+WguodCQG8Gg56PYuBNMeu3btghACs2bNarFPVFQUhg0bho8//hgnT55sdX9tOQ9JSUmwsrKCpaUlDh8+jJdffhm2trZwcXHBwYMHtfbX0NCALVu2wNXVFQqFAqNGjUJqamrHfukW1NXVITc3t9U7tD05bmFhIezt7eHm5tZqv9LSUigUCgwePFir3cHBAVOmTEFCQgKXLDYiFr9ERATg0V2pjRs34p133kFZWRlOnTqFkpISTJ48Gbdu3QLwqNB7cp7RxMREbNu2TastISEBM2fOhLu7O4QQuHz5MkJCQhAYGIiamhqsXbsWxcXFOHfuHOrr6/Hiiy9qviLuSAzg96kZGxsbO+/gtOLo0aPw8PCApaVli30UCgU++eQT9OrVC0FBQaiurm6xb1vOw6pVq7Bu3TrU1tbCxsYGqampKCoqwpAhQxAUFAS1Wq3Z34YNGxAbG4v4+HjcvHkTM2fOxOLFi3WWB+8MN27cQF1dHX744QdMnTpV8wFn+PDhSExMNFjBZ8i4arUapaWl2LNnD06ePIndu3dDJpO12L+mpgbZ2dkICgpqtt+YMWNQWlqKvLy8dudEHcPil4iIUFtbi7i4OMybNw9Lly6FnZ0dRo4ciQ8//BB37txp99LdzTE3N9fc1fTy8kJSUhKqqqqQnJzcKfufMWMGKisrsXnz5k7ZX2uqq6vxyy+/wN3d/al9fXx8sG7dOhQXF2PDhg3N9mnPefD19YWtrS0cHR0REBCA6upqXLt2DQDw8OFDJCUlYe7cuZg/fz7s7e2xadMmSKXSTjvej2saDuDo6Ijt27ejoKAAt27dwpw5c7BmzRocOHCg02MaOu7AgQPh4uKCrVu3IjY2Fv7+/q32j46OhlKpRFRUVLOvDx06FACQn5/f7pyoY1j8EhERCgoK8ODBA4wbN06rffz48ZDJZFrDEjrbuHHjYGlpqfW1fk9RVlYGIUSrd30fFxUVBQ8PDyQmJuKbb77Reb2j56HpTmPTnd+ffvoJNTU1Wg98KRQK9O/f3yDHu2lM7ogRI+Dr64vevXvDzs4O27Ztg52dXad+iOqquCUlJSgrK8OBAwfw6aefYsyYMS2OM8/MzERaWhq++uornQcfmzRdK0138anrsfglIiLcu3cPwKMn159kb2+Pqqoqg8a3sLDA7du3DRrDEB4+fAgArT6I9Ti5XI7k5GRIJBIsX74ctbW1Wq939nloGl6xadMmrcV1rl69qvMwVmdQKpUAoBmX3UQmk8HNzQ1FRUWdHtPQcaVSKRwdHTF9+nSkpKSgoKAA0dHROv1SUlKwY8cO5OTkYNCgQS3ur2kFyKZrh7oei18iIoK9vT0ANFtc3bt3Dy4uLgaLrVarDR7DUJoKGX0WL/Dx8UFYWBgKCwsRGRmp9VpnnwdHR0cAQHx8vM4CO2fOnNFrX21hbW2NoUOH4tKlSzqv1dfXw87OrtNjdmVclUoFMzMzFBQUaLXv3r0b+/fvR3Z2NgYMGNDqPurq6gCAy6AbEYtfIiKCt7c3rK2tdR6COnv2LOrq6vD8889r2szNzbUeqOqonJwcCCEwYcIEg8UwFCcnJ0gkEr3n742MjISnpyfOnz+v1a7PeWiLgQMHQi6X48KFC3pt1xH+/v44f/48rly5ommrqanB1atXO20aMkPHLS8vx+LFi3XaCwsL0dDQgIEDBwJ4NDPH+vXrkZ+fj6ysrGbv2D+p6Vrp16+fXjlR52HxS0REkMvlCA8PR2ZmJvbv34/Kykrk5+dj5cqVUCqVCA4O1vRVqVS4e/cusrKyoFarcfv2bVy9elVnn71798aNGzdQXFyMqqoqTTHb2NiIiooK1NfX4+LFiwgNDYWrqysCAwM7Jcbx48e7bKozS0tLDBkyBNevX9dru6bhD0/OD6vPeWhrnNdeew0HDx5EUlISKisr0dDQgOvXr2sWaggICEC/fv06bXnlsLAwuLm5ITAwENeuXUN5eTnWr1+P2tparQf9unNcKysrnDhxAtnZ2aisrIRarcb58+fx6quvwsrKCmFhYQCAS5cuITY2Fnv37oVUKtUaWiKRSLBz506dfTddK4b8IECtY/FLREQAgHfffRfR0dGIiIhA3759MWXKFAwaNAg5OTmwsrLS9Fu1ahWmTp2KRYsWwcPDA5GRkZqvcH18fDRTlq1cuRJOTk7w8vLCK6+8grt37wJ4NNZx5MiRUCgUmDx5MoYNG4avv/5aa9xsR2N0pRkzZqCgoEBr/O7nn38OlUqFoqIijB8/Hm+88YbOdhMmTNAUUY9ry3lISkpCfHw8AGDUqFG4cuUK9u7di/DwcADASy+9hMLCQgCPpoRbt24dYmJi0KdPHyiVSoSGhqKiogLAo6/hy8rKcPjw4VZ/z9zcXEyaNAkDBgzA2bNnkZeXB6VSiYkTJ+LUqVOafg4ODjh9+jRcXFwwevRoODs749tvv8XRo0e15uHtznHlcjkmTpyI119/Hc7OzrCxsYGfnx8GDRqE3NxczQOE7ZlC7bvvvoOzszNGjRql97bUOSSCsyx3mJ+fHwAgPT3dyJkQEf1OIpEgNTVVZ85cY1qxYgXS09NRXl5u7FSa1Z6/55cvX8bw4cORnJyMpUuXGio1g2lsbMQLL7yAwMBALF++nHENqLy8HC4uLoiKitJ8UGmr7vh+7qHSeeeXiIi6lD4Ph/UEKpUKERERiIiI0Fr2tidoaGhAVlYWqqqqEBAQwLgGtnXrVowePRohISFdHpt+x+KXiIiogzZu3Ag/Pz8EBATo/fCbMeXk5CAjIwPHjx9v81zFjNs+cXFxuHDhAo4dOwapVNqlsUkbi18iIuoSb7/9NpKTk3H//n0MHjwYhw4dMnZKnWr79u0ICQnBe++9Z+xU2mzatGn47LPP0L9/f8Y1oMOHD+O3335DTk4OHBwcujQ26TI3dgJERGQaoqOjm10c4Fkyffp0TJ8+3dhpUDcze/ZszJ4929hp0P/HO79EREREZDJY/BIRERGRyWDxS0REREQmg8UvEREREZkMPvDWSXJzczWToxMRdRfx8fFcgEcPubm5AMC/50TPMBa/ncDHx8fYKRBRD/Lrr7/i/PnzePnllw0aZ8GCBQbd/7NowoQJxk6BqFkLFizAwIEDjZ3GM4HLGxMRdbG0tDT4+/uDf36JiLoclzcmIiIiItPB4peIiIiITAaLXyIiIiIyGSx+iYiIiMhksPglIiIiIpPB4peIiIiITAaLXyIiIiIyGSx+iYiIiMhksPglIiIiIpPB4peIiIiITAaLXyIiIiIyGSx+iYiIiMhksPglIiIiIpPB4peIiIiITAaLXyIiIiIyGSx+iYiIiMhksPglIiIiIpPB4peIiIiITAaLXyIiIiIyGSx+iYiIiMhksPglIiIiIpPB4peIiIiITAaLXyIiIiIyGSx+iYiIiMhksPglIiIiIpPB4peIiIiITAaLXyIiIiIyGSx+iYiIiMhksPglIiIiIpPB4peIiIiITAaLXyIiIiIyGSx+iYiIiMhkmBs7ASKiZ5larcaDBw+02qqrqwEAFRUVWu0SiQT29vZdlhsRkSli8UtEZEB3796Fs7MzGhoadF7r3bu31r+nTp2K7OzsrkqNiMgkcdgDEZEB9evXD38EyssQAAAgAElEQVT605/Qq1frf24lEgkWLVrURVkREZkuFr9ERAa2bNmyp/YxMzPDvHnzuiAbIiLTxuKXiMjA5s+fD3PzlkeZmZmZ4aWXXkKfPn26MCsiItPE4peIyMBsbW3x8ssvt1gACyGwdOnSLs6KiMg0sfglIuoCS5cubfahNwCQyWT4y1/+0sUZERGZJha/RERd4C9/+QssLS112qVSKebOnQsrKysjZEVEZHpY/BIRdQG5XI558+ZBKpVqtavVaixZssRIWRERmR4Wv0REXWTx4sVQq9Vabba2tnjxxReNlBERkelh8UtE1EX+/Oc/ay1sIZVKsWjRIshkMiNmRURkWlj8EhF1EXNzcyxatEgz9EGtVmPx4sVGzoqIyLSw+CUi6kKLFi3SDH3o168fJk2aZOSMiIhMC4tfIqIu5OvrC2dnZwDAX//616cue0xERJ2r5SWHiJ5w5swZlJSUGDsNoh5v/PjxKC0tRZ8+fZCWlmbsdIh6PF9fX7i4uBg7DeohJEIIYewkqGfw8/PDoUOHjJ0GERGRltTUVCxcuNDYaVDPkM47v6SXBQsWID093dhpEHWIRCIx+n+Whw4dwoIFC4wWX19+fn4AwPc/dTsSicTYKVAPw8FmRERG0JMKXyKiZwmLXyIiIiIyGSx+iYiIiMhksPglIiIiIpPB4peIiIiITAaLXyIiIiIyGSx+iYja6dixY7Czs8M///lPY6fS7Z08eRIbN25ERkYGhgwZAolEAolEgmXLlun0nT59OmxsbGBmZoYRI0bg3LlzRshYf42NjYiPj4evr2+LfdRqNaKjo6FSqSCTyWBvbw9vb28UFxf3mLgxMTHw9PSEQqGAlZUVPD09sXnzZlRWVmr1i4iIgJeXF2xtbWFhYQGVSoW33noLDx480PT54osvEBMTg4aGBr3zIGovFr9ERO3ENYLa5t1338WuXbvw9ttvY/78+bhy5Qrc3d3Rp08f7N+/H0ePHtXqf+LECaSnp2PmzJkoKCjA2LFjjZR52xUWFuJPf/oTwsLCUFNT02I/f39//P3vf8dnn32Gmpoa/Oc//4G7u7tWQdjd454+fRpBQUG4du0abt26hcjISMTExOhM35ednY01a9aguLgYd+7cQXR0NBISEjRzRgPArFmzIJfLMW3aNNy7d0/vXIjag4tcEBG104wZM3D//n1jpwEAqK2txbRp0/Dvf//b2Klo2bFjB1JSUpCXlwe5XK712q5du7Bs2TIEBwejoKAAdnZ2RsqyY/Ly8hAREYGVK1eiurq6xQ9FKSkpyMrKQl5eHkaOHAkAUCqVOHz4cI+KK5PJsHr1as359PPzQ3p6OtLT03Hz5k0olUoAgLW1NYKDg2FmZgYAWLhwITIyMpCWloaSkhIMHDgQALB27VpcuXIFr7zyCk6dOgVzc5YmZFi880tE9AzYt28fysrKjJ2GlsuXL2Pz5s3Ytm2bTuELAL6+vggNDUVpaSnefPNNI2TYOZ577jlkZGRgyZIlsLCwaLHfBx98gLFjx2oK0J4aNzMzU+d8Ojs7A4DWneQjR45oCt8mffv2BQCdu9Rbt27FhQsXkJCQ0Ck5ErWGxS8RUTt88803cHV1hUQiwZ49ewAASUlJsLKygqWlJQ4fPoyXX34Ztra2cHFxwcGDBzXb7tq1C3K5HE5OTlixYgWUSiXkcjl8fX1x9uxZTb+QkBDIZDL0799f07Z69WpYWVlBIpHgzp07AIDQ0FCEh4ejqKgIEokEKpUKAPDll1/C1tYW27dv74pDomPXrl0QQmDWrFkt9omKisKwYcPw8ccf4+TJk63uTwiBuLg4DB8+HBYWFnBwcMCcOXPw448/avq09RwAQENDA7Zs2QJXV1coFAqMGjUKqampHfulW1BXV4fc3FyMHj3aIPs3dtzCwkLY29vDzc2t1X6lpaVQKBQYPHiwVruDgwOmTJmChIQEDicig2PxS0TUDpMmTdIZYrBq1SqsW7cOtbW1sLGxQWpqKoqKijBkyBAEBQVBrVYDeFTUBgYGoqamBmvXrkVxcTHOnTuH+vp6vPjiiygpKQHwqHhcuHChVozExERs27ZNqy0hIQEzZ86Eu7s7hBC4fPkyAGgeImpsbDTIMXiao0ePwsPDA5aWli32USgU+OSTT9CrVy8EBQWhurq6xb5bt27Fxo0b8c4776CsrAynTp1CSUkJJk+ejFu3bgFo+zkAgA0bNiA2Nhbx8fG4efMmZs6cicWLF+P777/vvIPw/924cQN1dXX44YcfMHXqVM0HnuHDhyMxMdFgBZ8h46rVapSWlmLPnj04efIkdu/eDZlM1mL/mpoaZGdnIygoqNl+Y8aMQWlpKfLy8tqdE1FbsPglIjIAX19f2NrawtHREQEBAaiursa1a9e0+pibm2vuYnp5eSEpKQlVVVVITk7ulBxmzJiByspKbN68uVP2p4/q6mr88ssvcHd3f2pfHx8frFu3DsXFxdiwYUOzfWpraxEXF4d58+Zh6dKlsLOzw8iRI/Hhhx/izp07+Oijj3S2ae0cPHz4EElJSZg7dy7mz58Pe3t7bNq0CVKptNOO/+OahgM4Ojpi+/btKCgowK1btzBnzhysWbMGBw4c6PSYho47cOBAuLi4YOvWrYiNjYW/v3+r/aOjo6FUKhEVFdXs60OHDgUA5OfntzsnorZg8UtEZGBNd7kev+vYnHHjxsHS0lLra/yeqqysDEKIVu/6Pi4qKgoeHh5ITEzEN998o/N6QUEBHjx4gHHjxmm1jx8/HjKZTGu4SHOePAc//fQTampq4O3tremjUCjQv39/gxz/pjG5I0aMgK+vL3r37g07Ozts27YNdnZ2zRbv3T1uSUkJysrKcODAAXz66acYM2ZMi+POMzMzkZaWhq+++go2NjbN9mm6Vpru4hMZCotfIqJuxMLCArdv3zZ2Gh328OFDAGj1QazHyeVyJCcnQyKRYPny5aitrdV6vWkaLGtra51t7e3tUVVVpVd+TcMrNm3apJlzWCKR4OrVq61OGdZeTTMgNI3TbiKTyeDm5oaioqJOj2nouFKpFI6Ojpg+fTpSUlJQUFCA6OhonX4pKSnYsWMHcnJyMGjQoBb3p1AoAPx+7RAZCotfIqJuQq1W4969e3BxcTF2Kh3WVMjos3iBj48PwsLCUFhYiMjISK3X7O3tAaDZIrc9x8zR0REAEB8fDyGE1s+ZM2f02ldbWFtbY+jQobh06ZLOa/X19Qab5q2r4qpUKpiZmaGgoECrfffu3di/fz+ys7MxYMCAVvdRV1cH4Pdrh8hQWPwSEXUTOTk5EEJgwoQJmjZzc/OnDpfojpycnCCRSPSeBzkyMhKenp44f/68Vru3tzesra11HkY7e/Ys6urq8Pzzz+sVZ+DAgZDL5bhw4YJe23WEv78/zp8/jytXrmjaampqcPXq1U6bhszQccvLy7F48WKd9sLCQjQ0NGjm7hVCYP369cjPz0dWVlazd+yf1HSt9OvXT6+ciPTF4peIyEgaGxtRUVGB+vp6XLx4EaGhoXB1dUVgYKCmj0qlwt27d5GVlQW1Wo3bt2/j6tWrOvvq3bs3bty4geLiYlRVVUGtVuP48eNGm+rM0tISQ4YMwfXr1/Xarmn4w5Pzw8rlcoSHhyMzMxP79+9HZWUl8vPzsXLlSiiVSgQHB+sd57XXXsPBgweRlJSEyspKNDQ04Pr167h58yYAICAgAP369eu05ZXDwsLg5uaGwMBAXLt2DeXl5Vi/fj1qa2u1HvTrznGtrKxw4sQJZGdno7KyEmq1GufPn8err74KKysrhIWFAQAuXbqE2NhY7N27F1KpVGtoiUQiwc6dO3X23XStGPKDABHA4peIqF327NmD8ePHAwDWr1+P2bNnIykpCfHx8QCAUaNG4cqVK9i7dy/Cw8MBAC+99BIKCws1+3j48CFGjhwJhUKByZMnY9iwYfj666+1xsmuWrUKU6dOxaJFi+Dh4YHIyEjN18I+Pj6aadFWrlwJJycneHl54ZVXXsHdu3e75Di0ZsaMGSgoKNAav/v5559DpVKhqKgI48ePxxtvvKGz3YQJEzRF1OPeffddREdHIyIiAn379sWUKVMwaNAg5OTkwMrKCgD0OgcJCQlYt24dYmJi0KdPHyiVSoSGhqKiogLAo6/hy8rKnroSWm5uLiZNmoQBAwbg7NmzyMvLg1KpxMSJE3Hq1ClNPwcHB5w+fRouLi4YPXo0nJ2d8e233+Lo0aNa8/B257hyuRwTJ07E66+/DmdnZ9jY2MDPzw+DBg1Cbm6u5gHC9kyh9t1338HZ2RmjRo3Se1sifUgEZ5OmNmpajz09Pd3ImRB1jEQiQWpqqs4cul1pxYoVSE9PR3l5udFy0Ed73v+XL1/G8OHDkZycjKVLlxoqNYNpbGzECy+8gMDAQCxfvpxxDai8vBwuLi6IiorSfFBpq+7wfqYeJZ13fomIjESfh8F6IpVKhYiICERERGgte9sTNDQ0ICsrC1VVVQgICGBcA9u6dStGjx6NkJCQLo9NpofFL3Wp119/HTY2NpBIJF36oEl3FBERAS8vL9ja2sLCwgIqlQpvvfVWu4qEjIwMDBkyRGdcnUwmg5OTE1544QW8//77mq9zibrKxo0b4efnh4CAAL0ffjOmnJwcZGRk4Pjx422eq5hx2ycuLg4XLlzAsWPHIJVKuzQ2mSYWv9SlPv74Y+zdu9fYaXQL2dnZWLNmDYqLi3Hnzh1ER0cjISFB8/WyPubPn48rV67A3d0ddnZ2EEKgsbERZWVlSEtLw+DBg7F+/XqMGDHCIEu3kn7efvttJCcn4/79+xg8eDAOHTpk7JQMavv27QgJCcF7771n7FTabNq0afjss8/Qv39/xjWgw4cP47fffkNOTg4cHBy6NDaZLha/RB1QW1sLX1/fdm1rbW2N4OBg9O7dGzY2Nli4cCHmzp2LL7/8UvMQU0dIJBLY29vjhRdeQHJyMtLS0nDr1i3MmDGjR92Ba0lHjr2xRUdH47fffoMQAr/88gsWLFhg7JQMbvr06dixY4ex06BuZvbs2di4caPO7B5EhsTil7qcRCIxdgqdZt++fS0u5/k0R44c0fmD37dvXwAwyApTCxYsQGBgIMrKyvDhhx92+v67WkeOPRERmS4Wv2RQQgi8//778PDwgIWFBezs7PDf//3fWn1iY2NhaWkJGxsblJWVITw8HM7Ozvjpp58ghEBcXByGDx8OCwsLODg4YM6cOfjxxx812+/atQtyuRxOTk5YsWIFlEol5HI5fH19cfbsWZ18nra/kJAQyGQyra//Vq9eDSsrK0gkEs0yoaGhoQgPD0dRUREkEglUKlWHj1dpaSkUCgUGDx6safvyyy87ba7Wpvljjx8/DoDHnoiITJAgaqMFCxaIBQsW6LXNO++8IyQSifjb3/4mKioqRE1NjUhMTBQAxPnz57X6ARBr164Vu3fvFvPmzRP/+c9/xJYtW4RMJhP/+Mc/xL1798TFixfF2LFjRd++fcWvv/6q2T44OFhYWVmJS5cuiYcPH4qCggIxfvx4YWNjI65du6bp19b9LVmyRPTr10/rd3n//fcFAHH79m1N2/z584W7u7tex6Ql1dXVwsbGRoSEhGi1HzlyRNjY2IiIiIin7sPd3V3Y2dm1+HplZaUAIAYOHKhpM8VjD0Ckpqa2a1tT1Z73P1FX4PuZ9JTGO79kMLW1tYiPj8ef//xnhIWFwd7eHgqFAr17925xmx07dmDNmjXIyMiAm5sb4uLiMG/ePCxduhR2dnYYOXIkPvzwQ9y5cwcfffSR1rbm5uaau4peXl5ISkpCVVUVkpOTNfnos7+uFh0dDaVSiaioKK32GTNmoLKyEps3b+5wjKaZNqqqqnReM+VjT0REpsPc2AnQs+vy5cuoqanBtGnT2rV9QUEBHjx4gHHjxmm1jx8/HjKZTOdr9SeNGzcOlpaWmq/VO7o/Q8rMzERaWhpOnDgBGxsbg8Wprq6GEAK2trat9jOFYx8fH88FW/SQm5sLAO2ajYSIqDth8UsG07ROu6OjY7u2v3fvHoBHsyI8yd7evtm7l0+ysLDA7du3O21/hpCSkoK4uDjk5ORgwIABBo31888/AwA8PT1b7Wcqx56IiEwPi18yGLlcDgD47bff2rW9vb09ADRbGN27dw8uLi6tbq9Wq7X6dXR/hrB792589dVXyM7ObrYw7GxffvklAODll19utZ8pHPt169ZxOVQ9cHlz6q6epRmEqGtwzC8ZjLe3N3r16oV//etf7d7e2tpaZ1GGs2fPoq6uDs8//3yr2+fk5EAIgQkTJui9P3Nzc6jV6nbl3RZCCKxfvx75+fnIysrqksL3119/RXx8PFxcXLB8+fJW+z7Lx56IiEwbi18yGEdHR8yfPx+HDh3Cvn37UFlZiYsXL7b54Sa5XI7w8HBkZmZi//79qKysRH5+PlauXAmlUong4GCt/o2NjaioqEB9fT0uXryI0NBQuLq6aqb30md/KpUKd+/eRVZWFtRqNW7fvo2rV6/q5Ni7d2/cuHEDxcXFqKqqanPRdunSJcTGxmLv3r2QSqU6yxLv3LlT0/f48eN6TXUmhMCDBw/Q2NgIIQRu376N1NRUTJw4EWZmZsjKynrqmN9n+dgTEZGJM+ZcE9SztGeqo6qqKvH666+LPn36CGtrazFp0iSxZcsWAUC4uLiIvLw8ERMTIxQKhWYKrn/84x+a7RsbG8X7778vhg4dKqRSqXBwcBBz584VP/30k1ac4OBgIZVKhbOzszA3Nxe2trZizpw5oqioSKtfW/dXXl4upk6dKuRyuRg8eLB44403xH//938LAEKlUmmm8Dp37pxwc3MTCoVCTJo0SWvKrtbk5+cLAC3+vP/++5q+x44dEzY2NiIqKqrF/X3xxRdi1KhRwtLSUshkMtGrVy8BQEgkEmFvby/+8Ic/iIiICFFeXq61nSkeeyE4NVJ7cKoz6q74fiY9pUmEEMIINTf1QN15zN+KFSuQnp6O8vJyY6dicnrisZdIJEhNTeWYXz105/c/mTa+n0lP6Rz2QM+MhoYGY6dgsnjsiYiop2DxS9RJfvzxR52xu839BAQEGDtVoi538uRJbNy4ERkZGRgyZIjm/bBs2TKdvtOnT4eNjQ3MzMwwYsQInDt3zggZ66+xsRHx8fHw9fVtsY9arUZ0dDRUKhVkMhns7e3h7e2N4uLiHhM3JiYGnp6eUCgUsLKygqenJzZv3ozKykqtfhEREfDy8oKtrS0sLCygUqnw1ltv4cGDB5o+X3zxBWJiYvgBmrqWsQdeUM/RXcf8bdy4UchkMgFADBo0SKSnpxs7JZPRU489OEZQbx15/2/ZskXMnDlTVFZWatrc3d1Fnz59BABx5MgRnW2OHz8uZs+e3e58u9rPP/8sJk6cKACI5557rsV+c+fOFR4eHiI3N1eo1Wpx48YNMWvWLJGfn99j4s6YMUPs3LlTlJWViaqqKpGWliakUql48cUXtfpNmTJFJCYmivLyclFZWSlSU1OFVCoVL730kla/hIQEMWXKFFFRUaF3LkLw/Ux6S2PxS23WXYtfIn0Z+z/Lmpoa4ePj06NitPf9/95774lhw4aJ2tparXZ3d3fx2WefiV69eglnZ2dx7949rdd7UvF74cIFMW/ePLF//34xevToFovQgwcPColEIi5evNij486dO1fnfPr5+QkA4saNG5q2GTNmiPr6eq1+CxcuFAA0D642CQkJET4+PkKtVuudj7Hfz9TjpHHYAxFRF9u3bx/Kysp6fIynuXz5MjZv3oxt27ZpFr15nK+vL0JDQ1FaWoo333zTCBl2jueeew4ZGRlYsmQJLCwsWuz3wQcfYOzYsRg5cmSPjpuZmalzPp2dnQFAa0jDkSNHYGZmptWvb9++AICamhqt9q1bt+LChQtISEjolByJWsPil4joKYQQiIuLw/Dhw2FhYQEHBwfMmTMHP/74o6ZPSEgIZDIZ+vfvr2lbvXo1rKysIJFIcOfOHQBAaGgowsPDUVRUBIlEApVKhV27dkEul8PJyQkrVqyAUqmEXC6Hr68vzp492ykxgEcr/OkzZ3RH7dq1C0IIzJo1q8U+UVFRGDZsGD7++GOcPHmy1f215TwkJSXBysoKlpaWOHz4MF5++WXY2trCxcUFBw8e1NpfQ0MDtmzZAldXVygUCowaNQqpqakd+6VbUFdXh9zcXIwePdog+zd23MLCQtjb28PNza3VfqWlpVAoFBg8eLBWu4ODA6ZMmYKEhAQITkJFBsbil4joKbZu3YqNGzfinXfeQVlZGU6dOoWSkhJMnjwZt27dAvCo0HtyqqXExERs27ZNqy0hIQEzZ86Eu7s7hBC4fPkyQkJCEBgYiJqaGqxduxbFxcU4d+4c6uvr8eKLL6KkpKTDMYDfZ+VobGzsvIPTiqNHj8LDwwOWlpYt9lEoFPjkk0/Qq1cvBAUFobq6usW+bTkPq1atwrp161BbWwsbGxukpqaiqKgIQ4YMQVBQkNZiKBs2bEBsbCzi4+Nx8+ZNzJw5E4sXL9ZZibAz3LhxA3V1dfjhhx8wdepUzQec4cOHIzEx0WAFnyHjqtVqlJaWYs+ePTh58iR2794NmUzWYv+amhpkZ2cjKCio2X5jxoxBaWkp8vLy2p0TUVuw+CUiakVtbS3i4uIwb948LF26FHZ2dhg5ciQ+/PBD3Llzp80rFraFubm55q6ml5cXkpKSUFVVheTk5E7Z/4wZM1BZWYnNmzd3yv5aU11djV9++QXu7u5P7evj44N169ahuLgYGzZsaLZPe86Dr68vbG1t4ejoiICAAFRXV+PatWsAgIcPHyIpKQlz587F/PnzYW9vj02bNkEqlXba8X5c03AAR0dHbN++HQUFBbh16xbmzJmDNWvW4MCBA50e09BxBw4cCBcXF2zduhWxsbHw9/dvtX90dDSUSiWioqKafX3o0KEAgPz8/HbnRNQWLH6JiFpRUFCABw8eYNy4cVrt48ePh0wm0xqW0NnGjRsHS0tLra/1e4qysjIIIVq96/u4qKgoeHh4IDExEd98843O6x09D013Gpvu/P7000+oqamBt7e3po9CoUD//v0NcrybxuSOGDECvr6+6N27N+zs7LBt2zbY2dl16oeoropbUlKCsrIyHDhwAJ9++inGjBnT4jjzzMxMpKWl4auvvoKNjU2zfZqulaa7+ESGwuKXiKgV9+7dAwBYW1vrvGZvb4+qqiqDxrewsMDt27cNGsMQHj58CACtPoj1OLlcjuTkZEgkEixfvhy1tbVar3f2eWgaXrFp0yatebivXr2q8zBWZ1AqlQCgGZfdRCaTwc3NDUVFRZ0e09BxpVIpHB0dMX36dKSkpKCgoADR0dE6/VJSUrBjxw7k5ORg0KBBLe5PoVAA+P3aITIUFr9ERK2wt7cHgGaLq3v37sHFxcVgsdVqtcFjGEpTIaPP4gU+Pj4ICwtDYWEhIiMjtV7r7PPg6OgIAIiPj4cQQuvnzJkzeu2rLaytrTF06FBcunRJ57X6+nrY2dl1esyujKtSqWBmZoaCggKt9t27d2P//v3Izs7GgAEDWt1HXV0dgN+vHSJDYfFLRNQKb29vWFtb6zwEdfbsWdTV1eH555/XtJmbm2s9UNVROTk5EEJgwoQJBothKE5OTpBIJLh//75e20VGRsLT0xPnz5/XatfnPLTFwIEDIZfLceHCBb226wh/f3+cP38eV65c0bTV1NTg6tWrnTYNmaHjlpeXY/HixTrthYWFaGhowMCBAwE8mplj/fr1yM/PR1ZWVrN37J/UdK3069dPr5yI9MXil4ioFXK5HOHh4cjMzMT+/ftRWVmJ/Px8rFy5EkqlEsHBwZq+KpUKd+/eRVZWFtRqNW7fvo2rV6/q7LN37964ceMGiouLUVVVpSlmGxsbUVFRgfr6ely8eBGhoaFwdXVFYGBgp8Q4fvx4l011ZmlpiSFDhuD69et6bdc0/OHJ+WH1OQ9tjfPaa6/h4MGDSEpKQmVlJRoaGnD9+nXcvHkTABAQEIB+/fp12vLKYWFhcHNzQ2BgIK5du4by8nKsX78etbW1Wg/6dee4VlZWOHHiBLKzs1FZWQm1Wo3z58/j1VdfhZWVFcLCwgAAly5dQmxsLPbu3QupVKqzzPvOnTt19t10rRjygwARwOKXiOip3n33XURHRyMiIgJ9+/bFlClTMGjQIOTk5MDKykrTb9WqVZg6dSoWLVoEDw8PREZGar7C9fHx0UxZtnLlSjg5OcHLywuvvPIK7t69C+DRWMeRI0dCoVBg8uTJGDZsGL7++mutcbMdjdGVZsyYgYKCAq3xu59//jlUKhWKioowfvx4vPHGGzrbTZgwQVNEPa4t5yEpKQnx8fEAgFGjRuHKlSvYu3cvwsPDAQAvvfQSCgsLATyaEm7dunWIiYlBnz59oFQqERoaioqKCgCPvoYvKyvD4cOHW/09c3NzMWnSJAwYMABnz55FXl4elEolJk6ciFOnTmn6OTg44PTp03BxccHo0aPh7OyMb7/9FkePHtWah7c7x5XL5Zg4cSJef/11ODs7w8bGBn5+fhg0aBByc3M1DxC2Zwq17777Ds7Ozhg1apTe2xLpQyI4mzS1kZ+fHwAgPT3dyJkQdYxEIkFqaqrOnLnGtGLFCqSnp6O8vNzYqTSrPe//y5cvY/jw4UhOTsbSpUsNlZrBNDY24oUXXkBgYCCWL1/OuAZUXl4OFxcXREVFaT6otFV3fD9Tt5bOO79ERN2EPg+H9QQqlQoRERGIiIjQWva2J2hoaEBWVhaqqqoQEBDAuAa2detWjB49GiEhIV0em0wPi18iIjKYjRs3ws/PDwEBAXo//GZMOTk5yMjIwPHjx9s8VzHjtk9cXBwuXLiAY8eOQSqVdmlsMk0sfomIjOztt99GcnIy7t+/j8GDB+PQoUPGTgbaoIEAACAASURBVKlTbd++HSEhIXjvvfeMnUqbTZs2DZ999hn69+/PuAZ0+PBh/Pbbb8jJyYGDg0OXxibTZW7sBIiITF10dHSziwM8S6ZPn47p06cbOw3qZmbPno3Zs2cbOw0yMbzzS0REREQmg8UvEREREZkMFr9EREREZDJY/BIRERGRyWDxS0REREQmg7M9kF4OHToEiURi7DSIOszf3x/+/v7GTqPH4fufiHo6Lm9MbXbmzBmUlJQYOw2iHu/MmTNISEhAamqqsVMheib4+vrCxcXF2GlQz5DO4peIqIulpaXB398f/PNLRNTl0jnml4iIiIhMBotfIiIiIjIZLH6JiIiIyGSw+CUiIiIik8Hil4iIiIhMBotfIiIiIjIZLH6JiIiIyGSw+CUiIiIik8Hil4iIiIhMBotfIiIiIjIZLH6JiIiIyGSw+CUiIiIik8Hil4iIiIhMBotfIiIiIjIZLH6JiIiIyGSw+CUiIiIik8Hil4iIiIhMBotfIiIiIjIZLH6JiIiIyGSw+CUiIiIik8Hil4iIiIhMBotfIiIiIjIZLH6JiIiIyGSw+CUiIiIik8Hil4iIiIhMBotfIiIiIjIZLH6JiIiIyGSw+CUiIiIik8Hil4iIiIhMBotfIiIiIjIZLH6JiIiIyGSw+CUiIiIik2Fu7ASIiJ5lt2/fxueff67V9v333wMAPvroI612GxsbLFq0qMtyIyIyRRIhhDB2EkREz6rffvsNTk5OePDgAczMzAAATX92JRKJpp9arcarr76KTz75xBhpEhGZinQOeyAiMiALCwssWLAA5ubmUKvVUKvVqK+vR319vebfarUaALB48WIjZ0tE9Oxj8UtEZGCLFy9GXV1dq33s7e3xX//1X12UERGR6WLxS0RkYFOnToWjo2OLr0ulUixduhTm5nwMg4jI0Fj8EhEZWK9evbBkyRJIpdJmX1er1XzQjYioi7D4JSLqAosWLdKM7X3SgAED4OPj08UZERGZJha/RERd4A9/+APc3Nx02mUyGV599VWtmR+IiMhwWPwSEXWRZcuW6Qx9qKur45AHIqIuxOKXiKiLLFmyRGfog0qlwsiRI42UERGR6WHxS0TURTw9PeHl5aUZ4iCVSvHaa68ZOSsiItPC4peIqAv99a9/1az0Vl9fzyEPRERdjMUvEVEXWrRoERoaGgAAY8eOxeDBg42cERGRaWHxS0TUhVxdXfHHP/4RAPDqq68aORsiItPD5YRIL35+fsZOgajH++233yCRSHDixAmcOnXK2OkQ9Wg+Pj4ICwszdhrUg/DOL+nl0KFDuH79urHTIGq33Nxc5ObmGjUHFxcX9OvXD3K53Kh5tNX169dx6NAhY6dBpCM3NxdnzpwxdhrUw/DOL+lt3bp1WLhwobHTIGqXpm8v0tPTjZrH5cuXoVKpjJpDW6WlpcHf39/ox4zoSfw2ktqDd36JiIygpxS+RETPGha/RERERGQyWPwSERERkclg8UtEREREJoPFLxERERGZDBa/RETtcOzYMdjZ2eGf//ynsVPp9k6ePImNGzciIyMDQ4YMgUQigUQiwbJly3T6Tp8+HTY2NjAzM8OIESNw7tw5I2Ssv8bGRsTHx8PX17fFPmq1GtHR0VCpVJDJZLC3t4e3tzeKi4t7TNyYmBh4enpCoVDAysoKnp6e2Lx5MyorK7X6RUREwMvLC7a2trCwsIBKpcJbb72FBw8eaPp88cUXiImJ0ax4SNRVWPwSEbWDEMLYKfQI7777Lnbt2oW3334b8+fPx5UrV+Du7o4+ffpg//79OHr0qFb/EydOID09HTNnzkRBQQHGjh1rpMzbrrCwEH/6058QFhaGmpqaFvv5+/vj73//Oz777DPU1NTgP//5D9zd3bUKwu4e9/Tp0wgKCsK1a9dw69YtREZGIiYmBgsWLNDql52djTVr1qC4uBh37txBdHQ0EhIStKYmmzVrFuRyOaZNm4Z79+7pnQtRe3GeXyKidpgxYwbu379v7DQAALW1tZg2bRr+/e9/GzsVLTt27EBKSgry8vJ0FvTYtWsXli1bhuDgYBQUFMDOzs5IWXZMXl4eIiIisHLlSlRXV7f4oSglJQVZWVnIy8vDyJEjAQBKpRKHDx/uUXFlMhlWr16tOZ9+fn5IT09Heno6bt68CaVSCQCwtrZGcHAwzMzMAAALFy5ERkYG0tLSUFJSgoEDBwIA1q5diytXruCVV17BqVOnYG7OsoQMj3d+iYh6uH379qGsrMzYaWi5fPkyNm/ejG3btjW7kp2vry9CQ0NRWlqKN9980wgZdo7nnnsOGRkZWLJkCSwsLFrs98EHH2Ds2LGaArSnxs3MzNQ5n87OzgCgdSf5yJEjmsK3Sd++fQFA5y711q1bceHCBSQkJHRKjkRPw+KXiEhP33zzDVxdXSGRSLBnzx4AQFJSEqysrGBpaYnDhw/j5Zdfhq2tLVxcXHDw4EHNtrt27YJcLoeTkxNWrFgBpVIJuVwOX19fnD17VtMvJCQEMpkM/fv317StXr0aVlZWkEgkuHPnDgAgNDQU4eHhKCoqgkQi0Sye8eWXX8LW1hbbt2/vikOiY9euXRBCYNasWS32iYqKwrBhw/Dxxx/j5MmTre5PCIG4uDgMHz4cFhYWcHBwwJw5c/Djjz9q+rT1HABAQ0MDtmzZAldXVygUCowaNQqpqakd+6VbUFdXh9zcXIwePdog+zd23MLCQtjb28PNza3VfqWlpVAoFBg8eLBWu4ODA6ZMmYKEhAQOJ6IuweKXiEhPkyZN0hlisGrVKqxbtw61tbWwsbFBamoqioqKMGTIEAQFBUGtVgN4VNQGBgaipqYGa9euRXFxMc6dO4f6+nq8+OKLKCkpAfCoeHxyGfHExERs27ZNqy0hIQEzZ86Eu7s7hBC4fPkyAGgeImpsbDTIMXiao0ePwsPDA5aWli32USgU+OSTT9CrVy8EBQWhurq6xb5bt27Fxo0b8c4776CsrAynTp1CSUkJJk+ejFu3bgFo+zkAgA0bNiA2Nhbx8fG4efMmZs6cicWLF+P777/vvIPw/924cQN1dXX44YcfMHXqVM0HnuHDhyMxMdFgBZ8h46rVapSWlmLPnj04efIkdu/eDZlM1mL/mpoaZGdnIygoqNl+Y8aMQWlpKfLy8tqdE1FbsfglIupkvr6+sLW1haOjIwICAlBdXY1r165p9TE3N9fcxfTy8kJSUhKqqqqQnJzcKTnMmDEDlZWV2Lx5c6fsTx/V1dX45Zdf4O7u/tS+Pj4+WLduHYqLi7Fhw4Zm+9TW1iIuLg7z5s3D0qVLYWdnh5EjR+LDDz/EnTt38NFHH+ls09o5ePjwIZKSkjB37lzMnz8f9vb22LRpE6RSaacd/8c1DQdwdHTE9u3bUVBQgFu3bmHOnDlYs2YNDhw40OkxDR134MCBcHFxwdatWxEbGwt/f/9W+0dHR0OpVCIqKqrZ14cOHQoAyM/Pb3dORG3F4peIyICa7nI9ftexOePGjYOlpaXW1/g9VVlZGYQQrd71fVxUVBQ8PDyQmJiIb775Ruf1goICPHjwAOPGjdNqHz9+PGQymdZwkeY8eQ5++ukn1NTUwNvbW9NHoVCgf//+Bjn+TWNyR4wYAV9fX/Tu3Rt2dnbYtm0b7Ozsmi3eu3vckpISlJWV4cCBA/j0008xZsyYFsedZ2ZmIi0tDV999RVsbGya7dN0rTTdxScyJBa/RETdhIWFBW7fvm3sNDrs4cOHANDqg1iPk8vlSE5OhkQiwfLly1FbW6v1etM0WNbW1jrb2tvbo6qqSq/8moZXbNq0STPnsEQiwdWrV1udMqy9mmZAaBqn3UQmk8HNzQ1FRUWdHtPQcaVSKRwdHTF9+nSkpKSgoKAA0dHROv1SUlKwY8cO5OTkYNCgQS3uT6FQAPj92iEyJBa/RETdgFqtxr179+Di4mLsVDqsqZDRZ/ECHx8fhIWFobCwEJGRkVqv2dvbA0CzRW57jpmjoyMAID4+HkIIrZ8zZ87ota+2sLa2xtChQ3Hp0iWd1+rr6w02zVtXxVWpVDAzM0NBQYFW++7du7F//35kZ2djwIABre6jrq4OwO/XDpEhsfglIuoGcnJyIITAhAkTNG3m5uZPHS7RHTk5OUEikeg9D3JkZCQ8PT1x/vx5rXZvb29YW1vrPIx29uxZ1NXV4fnnn9crzsCBAyGXy3HhwgW9tusIf39/nD9/HleuXNG01dTU4OrVq502DZmh45aXl2Px4sU67YWFhWhoaNDM3SuEwPr165Gfn4+srKxm79g/qela6devn145EbUHi18iIiNobGxERUUF6uvrcfHiRYSGhsLV1RWBgYGaPiqVCnfv3kVWVhbUajVu376Nq1ev6uyrd+/euHHjBoqLi1FVVQW1Wo3jx48bbaozS0tLDBkyBNevX9dru6bhD0/ODyuXyxEeHo7MzEzs378flZWVyM/Px8qVK6FUKhEcHKx3nNdeew0HDx5EUlISKisr0dDQgOvXr+PmzZsAgICAAPTr16/TllcOCwuDm5sbAgMDce3aNZSXl2P9+vWora3VetCvO8e1srLCiRMnkJ2djcrKSqjVapw/fx6vvvoqrKysEBYWBgC4dOkSYmNjsXfvXkilUq2hJRKJBDt37tTZd9O1YsgPAkRNWPwSEelpz549GD9+PABg/fr1mD17NpKSkhAfHw8AGDVqFK5cuYK9e/ciPDwcAPDSSy+hsLBQs4+HDx9i5MiRUCgUmDx5MoYNG4avv/5aa5zsqlWrMHXqVCxatAgeHh6IjIzUfC3s4+OjmRZt5cqVcHJygpeXF1555RXcvXu3S45Da2bMmIGCggKt8buff/45VCoVioqKMH78eLzxxhs6202YMEFTRD3u3XffRXR0NCIiItC3b19MmTIFgwYNQk5ODqysrABAr3OQkJCAdevWISYmBn369IFSqURoaCgqKioAPPoavqys7KkroeXm5mLSpEkYMGAAzp49i7y8PCiVSkycOBGnTp3S9HNwcMDp06fh4uKC0aNHw9nZGd9++y2OHj2qNQ9vd44rl8sxceJEvP7663B2doaNjQ38/PwwaNAg5Obmah4gbM8Uat999x2cnZ0xatQovbcl0pdEcEZp0oNEIkFqaqrO/KNEPYWfnx8AID093Wg5rFixAunp6SgvLzdaDvpIS0uDv7+/XkXN5cuXMXz4cCQnJ2Pp0qUGzM4wGhsb8cILLyAwMBDLly9nXAMqLy+Hi4sLoqKiNB9U2qo7vJ+px0nnnV8iIiPQ52GwnkilUiEiIgIRERFay972BA0NDcjKykJVVRUCAgIY18C2bt2K0aNHIyQkpMtjk2li8UtERAaxceNG+Pn5ISAgQO+H34wpJycHGRkZOH78eJvnKmbc9omLi8OFCxdw7NgxSKXSLo1NpovFL5mkiIgIeHl5wdbWFhYWFlCpVHjrrbf+H3t3Hhd1tf8P/DUsw7DOIIISuAEqKpjr/Qlp5pey1Iu5Bm5F3sytkMVuKpqKiFld4VLQpl+6pSmiJt3Q8lpxzVRaVERKYxH3RFEWYYzt/P7wy+Q4LDMww4Dzej4e84fncz7nvDmfD/KemfM5p9lPqF544QXY29tDIpG06EnxjRs3wtvbG9bW1rC1tYW3tzdWrVqFsrKylv4oAO4u2v/yyy9jwIABsLe3h4WFBeRyOfr06YMJEyYYZPkmXWkz5rt374aHh4fGAzJSqRQuLi547LHH8Oabb6rmZXZEK1asQHJyMkpLS9GrVy/s2rXL2CEZ1Pr16xEaGooNGzYYOxStBQQEYNu2bejatSv7NaC0tDT88ccfyMjIgKOjY5v2TSZOEOkAgEhJSTF2GK02evRokZiYKIqLi0VZWZlISUkRlpaW4qmnnmr23O3btwsA4sSJEzr3O2HCBPHWW2+JoqIiUV5eLnbu3CksLS3FE0880ZIfQwghxObNm4WlpaV49NFHxZdffilu3bol7ty5I/Lz88WOHTuEv7+/eP/991vcvr7oMuaenp5CLpcLIYSoq6sTt27dEt9++60ICQkREolEuLq6ih9//LFFcUybNk1MmzatVT+LqUlJSRH8c0HtEX+fqQV2Whg59yYyCjs7O8yfP1+1pNIzzzyD3bt3Y+fOnbh48aJqvUp9k0qlWLx4MWQyGYC7D2ukpqYiNTUVV69eVe3IpK1jx45h/vz5GD16NL766itYWPz5K+3h4QEPDw8oFAq1VQaMpaVjLpFIoFAo8Nhjj+Gxxx7DhAkTEBQUhAkTJuC3334z2AYBRET0YOK0BzJJX3zxhcZaop07dwaAZrc3lUgkLe53z549qsS3npubGwC06KGgmJgY1NbWYsOGDWqJ772efPJJvPTSS7oHq2etGfN7TZs2DSEhISgqKsJ7772n1xiJiOjBx+SXDO6TTz7BsGHDIJPJYGtri549e6q2LxVCYNOmTejXrx+srKzg6OiISZMm4cyZM6rzk5KSYGtrCxsbG6SlpWHcuHFwcHCAu7s7tm/frqrXr18/SCQSmJmZYejQoaqE6u9//zvkcjlkMhk++uijRuO8fPkyrK2t0atXL1WZEAJvvvkm+vbtCysrK8jlcrzyyit6HZ/c3FwoFAr06NFDVfbll182u0FBVVUVvv76azg5OeEvf/mL1v219zHXRv1GEPv379fpPCIiIk7iIp1Axzm/cXFxAoDYsGGDKC4uFjdv3hTvv/++mDVrlhBCiNdee01IpVLxySefiJKSEnHq1CkxZMgQ0blzZ/H777+r2omKihIAxNdffy1KS0tFUVGRGDVqlLC1tRVVVVVCCCFqampEz549Rffu3UVNTY1aHOHh4SIuLq7ROCsqKoS9vb0IDQ1VK4+KihISiUT84x//ELdu3RKVlZUiMTGxxXN+61VVVYlLly6Jt99+W1hZWYlPPvlE7fgXX3wh7O3tRXR0dKNt/PbbbwKAGDFihE59t/cxF0J9zm9DysrKBADRrVs3nX52IThHsCU455faK/4+Uwvs5P9mpBNdkt+qqiqhUCjEmDFj1MprampEfHy8qKysFHZ2diI4OFjt+A8//CAAqCV+9YmYUqlUldUnoXl5eaqy+mR7586dqrKKigrRvXt3UVpa2misUVFRok+fPqKsrExVVllZKWxsbDQeRmvNA2/1unTpIgAIJycn8c9//lOVTOrip59+EgDE448/rvU57X3M6zWX/AohhEQiEQqFosk6DeEfS90x+aX2ir/P1AI7Oe2BDObUqVMoKSnBk08+qVZubm6OJUuWICcnB7dv38awYcPUjg8fPhxSqRSZmZlNti+VSgEA1dXVqrIXXngBcrkc8fHxqrKtW7di0qRJcHBwaLCdPXv2YOfOnfjqq69gb2+vKs/Ly0NlZSUCAgK0+4F1cPHiRRQVFeHTTz/Fv/71LwwePBhFRUU6tWFnZwdAt/my7X3MtVVRUQEhRKPtN2fXrl0ay6nx1fgrKCgIAIweB1983f960JcKJMPgag9kMPVr1yoUigaPl5SUAPgzibuXQqFAeXm5zn3a2dnhxRdfxJtvvokffvgBf/nLX/Duu+82+h/kjh07sGnTJmRkZOChhx5SO3bp0iUAgLOzs85xNMfS0hLOzs4YO3YsevXqhT59+iA2NlYtgWxOz549IZPJ8Ntvv2l9Tnsfc23V/8ze3t4tOn/EiBEIDw9v0bmm6OjRo4iPj0dKSoqxQyFSExcXZ+wQqANi8ksGU5/Y3Lhxo8Hj9UlxQwlXSUkJ3N3dW9RvaGgo4uPjERcXh4ULF6Jbt27w9PTUqPf222/jq6++wjfffNNgMli/KsMff/zRoji05eXlBXNzc+Tk5Oh0npWVFZ588kmkpaXh+++/xyOPPNJgvZs3b+Lvf/87Nm/e3O7HXFtffvklAGDcuHEtOt/d3R3PPPNMi/s3RfHx8RwzandSU1ONHQJ1QJz2QAbTs2dPdOrUCQcOHGjwuI+PD+zs7PDTTz+plWdmZqKqqgpDhw5tUb/1ic2uXbuwatUqhIWFqR0XQuDVV19FdnY29u7d22gS5uPjAzMzM/z3v/9tURz3Ky4uxsyZMzXKc3NzUVtb26K1hdesWQMrKytERERAqVQ2WOf06dOqZdDa+5hr4/fff0dcXBzc3d0xd+7cFrdDRESmickvGYyVlRVWrFiBQ4cOITQ0FJcvX0ZdXR3Ky8vxyy+/QCaTITIyEnv27MHWrVtRVlaG7OxsLFy4EK6urpg/f36L+46MjERNTQ1u3bqF//mf/1E79ssvv+CNN97Ahx9+CEtLS405ZG+99RaAu9Mdpk6dil27dmHLli0oKyvDqVOn8MEHH7QoJltbWxw4cADffPMNysrKUF1djRMnTuC5556Dra0tIiIiVHX379/f7FJnADBo0CBs27YNp0+fxqhRo7Bv3z6Ulpaiuroa586dw4cffoi//e1vsLS0BIB2P+b3EkLg9u3bqKurgxAC169fR0pKCh555BGYm5tj7969LZ7zS0REJsyoz9tRh4MWbG/8zjvvCF9fXyGTyYRMJhODBw8WiYmJQoi7W9e++eabonfv3sLS0lI4OjqKyZMni7Nnz6rOT0xMFDY2NgKA6N27t8jPzxcffPCBcHBwEABEjx49xG+//abR75gxY8TmzZs1yrOzswWARl9vvvmmqm55ebl44YUXhJOTk7CzsxMjR44Ur732mgAg3N3dRVZWlk5jMXHiRNGrVy9hZ2cnrKyshKenpwgODhbZ2dlq9fbt2yfs7e1FTEyMVu1euHBBLF26VPj6+go7Ozthbm4uFAqFGDx4sPjb3/4mvv/+e1Xd9jzmn3/+uRg4cKCwsbERUqlUmJmZCQCqlR3+8pe/iOjoaFFcXKzVuDSET4frjqs9UHvF32dqgZ0SIYRoy2SbOjaJRIKUlBTO/aMOa/r06QA4V1AXO3fuRFBQEPjngtob/j5TC6Ry2gMRERERmQwmv0StcObMGa3WogwODjZ2qERGc/DgQSxfvhy7d++Gh4eH6vdizpw5GnXHjh0Le3t7mJubY8CAATh+/LgRItZdXV0d4uLi4O/v32id6upqxMbGwsvLC1KpFAqFAj4+PigsLOxQ/bakvTt37sDb2xsrV65UlX3++efYuHEjamtrWxQHUUsx+SVqBW9vbwghmn3t2LHD2KESGcXq1auRkJCAFStWYOrUqSgoKICnpyecnJywdetWpKenq9U/cOAAUlNTERgYiJycHAwZMsRIkWsvNzcXjz76KCIiIprcdCYoKAgff/wxtm3bhsrKSvz666/w9PTE7du3O1S/LWkvKioKZ8+eVSubOHEiZDIZAgICVGuQE7UFrvNLRNTGlEolAgICcOTIkQ7dR3Nef/117NixA1lZWap1s+slJCRgzpw5mD9/PnJyciCXy40UZetkZWUhOjoaCxcuVO082JAdO3Zg7969yMrKgq+vLwDA1dUVaWlpHarflrR35MgRnD59usFjS5YsQUFBAcaPH49Dhw6plmUkMiR+8ktE1Ma2bNmi83bW7bGPpuTl5WHVqlVYu3atRuILAP7+/ggLC8Ply5exdOlSI0SoHw8//DB2796NWbNmwcrKqtF67777LoYMGaJKGDtqv7q2p1Qq8corrzS5e+WaNWtw8uRJnXa4JGoNJr9ERM0QQmDTpk3o168frKys4OjoiEmTJuHMmTOqOqGhoZBKpejatauqbPHixbC1tYVEIlHtdBgWFobIyEjk5+dDIpHAy8sLCQkJkMlkcHFxwYIFC+Dq6gqZTAZ/f39kZmbqpQ/g7s542qwfrQ8JCQkQQmDixImN1omJiUGfPn2wefNmHDx4sMn2tLkGSUlJsLW1hY2NDdLS0jBu3Dg4ODjA3d0d27dvV2uvtrYWr732Grp37w5ra2sMHDjQYNs3V1VV4dixYxg0aJBB2m+rflvSXlRUFBYvXtzkNvGOjo4YPXo04uPjuaIItQkmv0REzVizZg2WL1+OqKgoFBUV4dChQ7h48SJGjRqFa9euAbib7N2/BGBiYiLWrl2rVhYfH4/AwEB4enpCCIG8vDyEhoYiJCQElZWVWLJkCQoLC3H8+HHU1NTgiSeewMWLF1vdBwDVg0V1dXX6G5xGpKeno2/fvrCxsWm0jrW1NT766COYmZlh3rx5qKioaLSuNtdg0aJFCA8Ph1KphL29PVJSUpCfnw8PDw/MmzcP1dXVqvaWLVuGN954A3Fxcbh69SoCAwMxc+ZMjd0P9eHKlSuoqqrCzz//jDFjxqje3PTr1w+JiYkGS/j03a+u7X3//ffIz89vcGfL+w0ePBiXL19GVlaWTjERtQSTXyKiJiiVSmzatAlTpkzB7NmzIZfL4evri/feew83btxo8Y5/DbGwsFB9stm/f38kJSWhvLwcycnJeml/woQJKCsrw6pVq/TSXmMqKipw7tw5eHp6NlvXz88P4eHhKCwsxLJlyxqs05Jr4O/vDwcHBzg7OyM4OBgVFRW4cOECgLsrDyQlJWHy5MmYOnUqFAoFVq5cCUtLS72N9b3qHwRzdnbG+vXrkZOTg2vXrmHSpEl46aWX8Omnn+q9T0P0q0t7SqUSYWFhSEpK0qrt3r17AwCys7N1iomoJZj8EhE1IScnB7dv38awYcPUyocPHw6pVKo2LUHfhg0bBhsbG7Wv9juCoqIiCCGa/NT3XjExMejbty8SExNx+PBhjeOtvQZSqRQAVJ/8nj17FpWVlfDx8VHVsba2RteuXQ0y1vVzcgcMGAB/f3906tQJcrkca9euhVwu1+sbKEP2q0t7K1aswIsvvgg3Nzet2q6/V+o/xScyJCa/RERNqF+Cyc7OTuOYQqFAeXm5Qfu3srLC9evXDdqHvt25cwcAmnwQ614ymQzJycmQSCSYO3culEql2nF9X4P66RUrV65UW4/7/PnzTS4Z1lKurq4AoJqTXU8qlaJHjx7Iz8/Xe5+G6FfbBjI/OAAAIABJREFU9g4fPozs7Gy88MILWrdtbW0N4M97h8iQmPwSETVBoVAAQIMJVklJCdzd3Q3Wd3V1tcH7MIT6REaXzQv8/PwQERGB3NxcrFu3Tu2Yvq9B/cNXcXFxGmtyHz16VKe2tGFnZ4fevXvjl19+0ThWU1NjsGXe9N2vtu1t2bIFX3/9NczMzFRvLOrHfP369ZBIJBpzq6uqqgD8ee8QGRKTXyKiJvj4+MDOzk7jj3VmZiaqqqowdOhQVZmFhYXaQ1WtlZGRASEERowYYbA+DMHFxQUSiQSlpaU6nbdu3Tp4e3vjxIkTauW6XANtdOvWDTKZDCdPntTpvNYICgrCiRMnUFBQoCqrrKzE+fPn9bYMWVv0q017ycnJGm8q6r+9iIqKghBCYwpL/b3SpUsXnWMi0hWTXyKiJshkMkRGRmLPnj3YunUrysrKkJ2djYULF8LV1RXz589X1fXy8sLNmzexd+9eVFdX4/r16zh//rxGm506dcKVK1dQWFiI8vJyVTJbV1eHW7duoaamBqdOnUJYWBi6d++OkJAQvfSxf//+NlnqzMbGBh4eHrh06ZJO59VPfzA3N9co1/YaaNvP888/j+3btyMpKQllZWWora3FpUuXcPXqVQBAcHAwunTporftlSMiItCjRw+EhITgwoULKC4uxquvvgqlUqn2oF9771fb9nRVf68Y8o0AUT0mv0REzVi9ejViY2MRHR2Nzp07Y/To0ejZsycyMjJga2urqrdo0SKMGTMGM2bMQN++fbFu3TrV17h+fn6qJcsWLlwIFxcX9O/fH+PHj8fNmzcB3J3v6OvrC2tra4waNQp9+vTBt99+qzZ3trV9tJUJEyYgJydHbf7uZ599Bi8vL+Tn52P48OF4+eWXNc4bMWIEIiIiNMq1uQZJSUmIi4sDAAwcOBAFBQX48MMPERkZCQB46qmnkJubC+DucnDh4eHYuHEjnJyc4OrqirCwMNy6dQvA3a/hi4qKmt0J7dixYxg5ciQeeughZGZmIisrC66urnjkkUdw6NAhVT1HR0d89913cHd3x6BBg+Dm5oYffvgB6enpauvmtvd+tW1PVz/++CPc3NwwcODAFrdBpC2J4IrSpAOJRIKUlBSNtUaJOorp06cDAFJTU40ciboFCxYgNTUVxcXFxg5Fw86dOxEUFKTTurB5eXno168fkpOTMXv2bANGZxh1dXV47LHHEBISgrlz57JfAyouLoa7uztiYmJUb1S01V5/n6ldS+Unv0RE7YQuD4i1d15eXoiOjkZ0dLRqfdiOora2Fnv37kV5eTmCg4PZr4GtWbMGgwYNQmhoaJv3TaaJyS8RERnE8uXLMX36dAQHB+v88JsxZWRkYPfu3di/f7/WaxWz35bZtGkTTp48iX379sHS0rJN+ybTxeSXiMjIVqxYgeTkZJSWlqJXr17YtWuXsUPSm/Xr1yM0NBQbNmwwdihaCwgIwLZt29C1a1f2a0BpaWn4448/kJGRAUdHxzbtm0ybhbEDICIydbGxsYiNjTV2GAYzduxYjB071thhUDvz9NNP4+mnnzZ2GGSC+MkvEREREZkMJr9EREREZDKY/BIRERGRyWDyS0REREQmgw+8kc6OHj1q7BCIWqx+G9WdO3caOZKOo/53nmNG7c2lS5fg7u5u7DCog+EOb6QTiURi7BCIiIhUpk2bxh3eSBep/OSXdML3SkSt15LtgomISD8455eIiIiITAaTXyIiIiIyGUx+iYiIiMhkMPklIiIiIpPB5JeIiIiITAaTXyIiIiIyGUx+iYiIiMhkMPklIiIiIpPB5JeIiIiITAaTXyIiIiIyGUx+iYiIiMhkMPklIiIiIpPB5JeIiIiITAaTXyIiIiIyGUx+iYiIiMhkMPklIiIiIpPB5JeIiIiITAaTXyIiIiIyGUx+iYiIiMhkMPklIiIiIpPB5JeIiIiITAaTXyIiIiIyGUx+iYiIiMhkMPklIiIiIpPB5JeIiIiITAaTXyIiIiIyGUx+iYiIiMhkMPklIiIiIpPB5JeIiIiITAaTXyIiIiIyGUx+iYiIiMhkMPklIiIiIpPB5JeIiIiITIaFsQMgInqQXbp0Cc899xxqa2tVZbdu3YK9vT0ee+wxtbp9+/bF+++/38YREhGZFia/REQG5O7ujvPnzyM/P1/j2H//+1+1fz/66KNtFRYRkcnitAciIgN79tlnYWlp2Wy94ODgNoiGiMi0MfklIjKwWbNmoaampsk6AwYMQP/+/dsoIiIi08Xkl4jIwDw9PTFw4EBIJJIGj1taWuK5555r46iIiEwTk18iojbw7LPPwtzcvMFjNTU1mD59ehtHRERkmpj8EhG1gRkzZqCurk6j3MzMDCNGjEDPnj3bPigiIhPE5JeIqA24urrikUcegZmZ+n+7ZmZmePbZZ40UFRGR6WHyS0TURubMmaNRJoTAlClTjBANEZFpYvJLRNRGpk2bpjbv19zcHI8//jhcXFyMGBURkWlh8ktE1EYcHR3xxBNPqBJgIQRmz55t5KiIiEwLk18iojY0e/Zs1YNvlpaWmDRpkpEjIiIyLUx+iYja0MSJE2FlZQUACAwMhJ2dnZEjIiIyLUx+iYjakK2trerTXk55ICJqexIhhDB2EK3R2I5JRERERKRf06ZNQ2pqqrHDaI1UC2NHoA9hYWHw8/MzdhhERFqpra1FSkoKZs6caexQDC4uLg4AEB4ebuRIOo6jR48iPj4eKSkpxg6FSE3973NH90Akv35+fnjmmWeMHQYRkdYmT54MmUxm7DAMrv4TIv4frZv4+HiOGbU7HfwTXxXO+SUiMgJTSHyJiNojJr9EREREZDKY/BIRERGRyWDyS0REREQmg8kvEREREZkMJr9ERNTu7du3D3K5HP/+97+NHUq7d/DgQSxfvhy7d++Gh4cHJBIJJBIJ5syZo1F37NixsLe3h7m5OQYMGIDjx48bIWLd1dXVIS4uDv7+/o3Wqa6uRmxsLLy8vCCVSqFQKODj44PCwsIO1W9L2rtz5w68vb2xcuVKVdnnn3+OjRs3ora2tkVxPEiY/BIRUbvXwfdjajOrV69GQkICVqxYgalTp6KgoACenp5wcnLC1q1bkZ6erlb/wIEDSE1NRWBgIHJycjBkyBAjRa693NxcPProo4iIiEBlZWWj9YKCgvDxxx9j27ZtqKysxK+//gpPT0/cvn27Q/XbkvaioqJw9uxZtbKJEydCJpMhICAAJSUlLYrlQfFArPNLREQPtgkTJqC0tNTYYQAAlEolAgICcOTIEWOHoub111/Hjh07kJWVpbGUXkJCAubMmYP58+cjJycHcrncSFG2TlZWFqKjo7Fw4UJUVFQ0+qZox44d2Lt3L7KysuDr6wsAcHV1RVpaWofqtyXtHTlyBKdPn27w2JIlS1BQUIDx48fj0KFDsLAwzTSQn/wSERHpYMuWLSgqKjJ2GGry8vKwatUqrF27tsE1pP39/REWFobLly9j6dKlRohQPx5++GHs3r0bs2bNgpWVVaP13n33XQwZMkSVMHbUfnVtT6lU4pVXXkF8fHyjddasWYOTJ082WedBx+SXiIjatcOHD6N79+6QSCR45513AABJSUmwtbWFjY0N0tLSMG7cODg4OMDd3R3bt29XnZuQkACZTAYXFxcsWLAArq6ukMlk8Pf3R2ZmpqpeaGgopFIpunbtqipbvHgxbG1tIZFIcOPGDQBAWFgYIiMjkZ+fD4lEAi8vLwDAl19+CQcHB6xfv74thkRDQkIChBCYOHFio3ViYmLQp08fbN68GQcPHmyyPSEENm3ahH79+sHKygqOjo6YNGkSzpw5o6qj7TUA7m7p/dprr6F79+6wtrbGwIEDDbZ9c1VVFY4dO4ZBgwYZpP226rcl7UVFRWHx4sVwdnZutI6joyNGjx6N+Ph4k51OxOSXiIjatZEjR2pMMVi0aBHCw8OhVCphb2+PlJQU5Ofnw8PDA/PmzUN1dTWAu0ltSEgIKisrsWTJEhQWFuL48eOoqanBE088gYsXLwK4mzzev51wYmIi1q5dq1YWHx+PwMBAeHp6QgiBvLw8AFA9RFRXV2eQMWhOeno6+vbtCxsbm0brWFtb46OPPoKZmRnmzZuHioqKRuuuWbMGy5cvR1RUFIqKinDo0CFcvHgRo0aNwrVr1wBofw0AYNmyZXjjjTcQFxeHq1evIjAwEDNnzsRPP/2kv0H4P1euXEFVVRV+/vlnjBkzRvWGp1+/fkhMTDRYwqfvfnVt7/vvv0d+fj5mzpzZbNuDBw/G5cuXkZWVpVNMDwomv0RE1KH5+/vDwcEBzs7OCA4ORkVFBS5cuKBWx8LCQvUpZv/+/ZGUlITy8nIkJyfrJYYJEyagrKwMq1at0kt7uqioqMC5c+fg6enZbF0/Pz+Eh4ejsLAQy5Yta7COUqnEpk2bMGXKFMyePRtyuRy+vr547733cOPGDXzwwQca5zR1De7cuYOkpCRMnjwZU6dOhUKhwMqVK2Fpaam38b9X/YNgzs7OWL9+PXJycnDt2jVMmjQJL730Ej799FO992mIfnVpT6lUIiwsDElJSVq13bt3bwBAdna2TjE9KJj8EhHRA0MqlQKA2qeODRk2bBhsbGzUvsbvqIqKiiCEaPJT33vFxMSgb9++SExMxOHDhzWO5+Tk4Pbt2xg2bJha+fDhwyGVStWmizTk/mtw9uxZVFZWwsfHR1XH2toaXbt2Ncj418/JHTBgAPz9/dGpUyfI5XKsXbsWcrm8weS9PfarS3srVqzAiy++CDc3N63arr9X6j/FNzVMfomIyCRZWVnh+vXrxg6j1e7cuQMATT6IdS+ZTIbk5GRIJBLMnTsXSqVS7Xj9Mlh2dnYa5yoUCpSXl+sUX/30ipUrV6rWHJZIJDh//nyTS4a1lKurKwCo5mnXk0ql6NGjB/Lz8/XepyH61ba9w4cPIzs7Gy+88ILWbVtbWwP4894xNUx+iYjI5FRXV6OkpATu7u7GDqXV6hMZXTYv8PPzQ0REBHJzc7Fu3Tq1YwqFAgAaTHJbMmb1D1/FxcVBCKH2Onr0qE5tacPOzg69e/fGL7/8onGspqbGYMu86btfbdvbsmULvv76a5iZmaneWNSP+fr16yGRSDTmVldVVQH4894xNUx+iYjI5GRkZEAIgREjRqjKLCwsmp0u0R65uLhAIpHovA7yunXr4O3tjRMnTqiV+/j4wM7OTiNhyszMRFVVFYYOHapTP926dYNMJsPJkyd1Oq81goKCcOLECRQUFKjKKisrcf78eb0tQ9YW/WrTXnJyssabivpvNKKioiCE0JjCUn+vdOnSReeYHgRMfomI6IFXV1eHW7duoaamBqdOnUJYWBi6d++OkJAQVR0vLy/cvHkTe/fuRXV1Na5fv47z589rtNWpUydcuXIFhYWFKC8vR3V1Nfbv32+0pc5sbGzg4eGBS5cu6XRe/fQHc3NzjfLIyEjs2bMHW7duRVlZGbKzs7Fw4UK4urpi/vz5Ovfz/PPPY/v27UhKSkJZWRlqa2tx6dIlXL16FQAQHByMLl266G175YiICPTo0QMhISG4cOECiouL8eqrr0KpVKo96Nfe+9W2PV3V3yuGfCPQnjH5JSKidu2dd97B8OHDAQCvvvoqnn76aSQlJSEuLg4AMHDgQBQUFODDDz9EZGQkAOCpp55Cbm6uqo07d+7A19cX1tbWGDVqFPr06YNvv/1WbZ7sokWLMGbMGMyYMQN9+/bFunXrVF8L+/n5qZZFW7hwIVxcXNC/f3+MHz8eN2/ebJNxaMqECROQk5OjNn/3s88+g5eXF/Lz8zF8+HC8/PLLGueNGDECERERGuWrV69GbGwsoqOj0blzZ4wePRo9e/ZERkYGbG1tAUCnaxAfH4/w8HBs3LgRTk5OcHV1RVhYGG7dugXg7tfwRUVFze6EduzYMYwcORIPPfQQMjMzkZWVBVdXVzzyyCM4dOiQqp6joyO+++47uLu7Y9CgQXBzc8MPP/yA9PR0tXVz23u/2ranqx9//BFubm4YOHBgi9voyCSig69wLJFIkJKSorE+IxERGd/06dMBAKmpqUaLYcGCBUhNTUVxcbHRYtDFzp07ERQUpNO6sHl5eejXrx+Sk5Mxe/ZsA0ZnGHV1dXjssccQEhKCuXPnsl8DKi4uhru7O2JiYlRvVLTVHn6f9SCVn/wSEdEDT5eHwToiLy8vREdHIzo6WrU+bEdRW1uLvXv3ory8HMHBwezXwNasWYNBgwYhNDS0zftuL0wq+R0+fDjMzc3bfMtDAHj++echk8kgkUhMdmmR+7311luqBzXee+89Vfm+ffsgl8vx73//26D9t1U/2ti4cSO8vb1hbW0NW1tbeHt7Y9WqVSgrK9O5rd27d8PDw0NtSSGJRAILCwt07twZjz/+OPbs2WOAn0Kdtvf8/fHOmTNHo87YsWNhb28Pc3NzDBgwQG/z8wyF9zYZw/LlyzF9+nQEBwfr/PCbMWVkZGD37t3Yv3+/1msVs9+W2bRpE06ePIl9+/bB0tKyTftuT0wq+f3xxx8xZswYo/SdnJyMpUuXGqXv9mrp0qUaW5YCaLO9xtvTjJ/vvvsO8+bNw4ULF3Dt2jWsW7cOGzduxLRp03Rua+rUqSgoKICnpyfkcrna078pKSm4fPkypk6dipSUFAP8JH/S9p6/N14nJyds3boV6enpanUOHDiA1NRUBAYGIicnB0OGDDFU2HrBe7v9WLFiBZKTk1FaWopevXph165dxg7JoNavX4/Q0FBs2LDB2KFoLSAgANu2bUPXrl3ZrwGlpaXhjz/+QEZGBhwdHdu07/bGpJLfehKJpNVtKJVK+Pv76yEaut+ECRNQWlqKwMBAvbXZ0PUyRD8tJZVKsXjxYjg7O8POzg7Tp0/HpEmT8J///Ef1NHRrOTo6IiAgAP/85z8B3J1XqIu2uOcTEhJgZmaG+fPnd6hPrrRlive2scXGxuKPP/6AEALnzp1r0RvKjmbs2LF4/fXXjR0GtTNPP/00li9frrG6hykyyeRXHx/1b9myBUVFRS06Vx/JN+mmNderLezZswcymUytrH6bSn3P3+vZsyeAP3dx0lZb3PP+/v4ICwvD5cuX+U2Jltr7vU1E1N6YZPKbl5cHb29v2Nraqpa9uX9/8++++w79+/eHXC6HTCaDr68vvvrqKwBAWFgYIiMjkZ+fD4lEAi8vL9V5n3zyCYYNGwaZTAZbW1v07NlTbfccMzMzpKenY9y4cZDL5XB1dcX//u//6vwzJCUlwdbWFjY2NkhLS8O4cePg4OAAd3d3bN++Xa2uEAKbNm1Cv379YGVlBUdHR0yaNEltT/U33ngDNjY2sLe3R1FRESIjI+Hm5oaFCxfC1tYWZmZmGDp0KLp06QJLS0vY2tpiyJAhGDVqlGoBc4VCgb///e9aj2NDDh8+jO7du0MikeCdd94BcPd63T9/tf71n//8p0XXq6F+tB0rXca+NXJzc6FQKNCjRw9V2ZdfftnqtURPnToFABg9erRaeXu552NiYtCnTx9s3rwZBw8ebPJn4b3dMe9tIiKjEh0cAJGSkqJ1/YCAAOHh4SHOnTsnqqurxenTp8X/+3//T8hkMvHbb7+p6qWmpoo1a9aImzdviuLiYjFixAjh5OSkOj516lTh6emp1nZcXJwAIDZs2CCKi4vFzZs3xfvvvy9mzZolhBAiKipKABBff/21KCkpETdv3hTjx48XVlZWoqKiQuef/d72SktLRVFRkRg1apSwtbUVVVVVqnqvvfaakEql4pNPPhElJSXi1KlTYsiQIaJz587i999/12hvyZIl4u233xZTpkwRv/76q1i9erUAIDIzM0VFRYW4ceOGeOqppwQAkZ6eLq5fvy4qKipEaGioACBOnjyp9Tjm5uYKAOLdd99VlV28eFEAEG+//baqzrJly1RjdPXqVeHo6Cj8/f1FbW1ti6/X/f20ZKyaG3tdVVVViUuXLom3335bWFlZiU8++UTt+BdffCHs7e1FdHR0s215enoKuVyu+ndlZaXYv3+/6NGjhxg7dqy4ffu2Wn1j3/Oenp7i3LlzQgghjhw5IszMzETPnj1Vce7fv188/fTTaufw3m7/9/a0adPEtGnTdD7PlKWkpIgH4M8zPYAekN/nnR3+t6slye/DDz+sVnbq1CkBQCxdurTR82JjYwUAUVRUJITQ/INTVVUlFAqFGDNmjNp5NTU1Ij4+Xgjx5x8VpVKpOv7xxx8LAOL06dNa/wz1GmovMTFRABB5eXlCiLsJj52dnQgODlY794cffhAA1JKohtoTQqgShPLyclXZv/71LwFAZGdna7S5Y8eORmO+fxy1SRDuN3nyZCGTycSZM2e07kebBKG1Y3X/2LdEly5dBADh5OQk/vnPf7Yqkfb09BQANF6+vr7iX//6l/jjjz+aPL+t7/l7k18hhIiMjBQAxEsvvSSE0Ex+eW93jHv7Aflj2aaY/FJ79YD8Pu+0aOUHxw8EX19fyOVy1dfBDamfJ9zYWpGnTp1CSUkJnnzySbVyc3NzLFmypNl29bWfvFQqVWsvJycHt2/f1tjXe/jw4ZBKpcjMzGxVPzU1NaoybX6W5saxOTt37sRnn32GjRs3om/fvnrtp7Vjdf/Yt8TFixdRUlKCEydOYPny5fjggw/wzTffwMXFpUXtyeVy1dzempoaXLt2DQcOHEBoaChiY2Nx+PBhdO7cucFzjX3Px8TE4IsvvkBiYiKCgoI0jvPe7jj39qVLl3R+wNKUHT16FIDuD6USGdqlS5fg7u5u7DBajcnv/7G0tFT7jz09PR1vvvkmcnJyUFZW1ux/+vXrsSoUCoPGqav6xMfOzk7jmEKhQHl5uUH713Ucm1JcXIyXX34Zw4cP19iVRh/9GHusgLv3obOzM8aOHYtevXqhT58+iI2NRXx8fKvbtrCwgJubG55//nnU1tZi3rx52LBhA/7xj38AaH/3vEwmQ3JyMkaOHIm5c+di48aNaseNfb14b2vv2LFjDb6BoaZxzKg9ehBWTDHJB97uV1NTg5s3b6J79+4AgAsXLmDy5Mno2rUrMjMzUVpaqvGH934PPfQQAODGjRsGj1cX9YlJQ3/cSkpKDPoOriXj2JQlS5agpKQEycnJaku16KsfY45VQ7y8vGBubo6cnBy9t+3r6wsA+OWXXwC033vez88PERERyM3NVXuIDuC9rQtj39vTpk1TrTfNV/Ov+jW4jR0HX3zd/3oQEl+AyS8A4Ntvv0VdXZ1q4fzs7GxUV1dj0aJF8PDwUO1S1ZSePXuiU6dOOHDgQFuErDUfHx/Y2dnhp59+UivPzMxEVVUVhg4darC+WzKOjUlPT8e2bduwatUqDBgwQFX+yiuv6K0fY41VcXExZs6cqVGem5uL2tpadOvWTe99/vzzzwCg+nq9Pd/z69atg7e3N06cOKFWzntbe8YcKyKi9sYkk9+qqiqUlpaipqYGx48fR2hoKHr06IGQkBAAUH0CfPDgQdy5cwe5ubkac+I6deqEK1euoLCwEOXl5TAzM8OKFStw6NAhhIaG4vLly6irq0N5ebnq0zVjkMlkiIyMxJ49e7B161aUlZUhOzsbCxcuhKurK+bPn2+wvrUZR22UlZVhwYIFGDRoEJYtWwYAuHPnDn766SecPHmyRderoa+OjTVWtra2OHDgAL755hvV19onTpzAc889B1tbW0RERKjq7t+/X+elzpRKJerq6iCEwJUrV5CcnIyVK1eic+fOCA8PB9C+7/n66Q/3L8zOe/uu9nxvExG1S6KDg46rPSQnJ4sxY8YIFxcXYWFhIZycnMSMGTPE+fPn1eq9+uqrolOnTkKhUIjp06eLd955RwAQnp6e4sKFC+L48eOiR48ewtraWowcOVK1VNA777wjfH19hUwmEzKZTAwePFgkJiaKjRs3CmtrawFA9O7dW+Tn54utW7cKR0dHAUC4u7vrtOJDYmKisLGxUWvvgw8+EA4ODgKA6NGjh2rptrq6OvHmm2+K3r17C0tLS+Ho6CgmT54szp49q2rv3vi6deumWmIrPj5e1U/Pnj3Fd999J15//XUhl8sFANGlSxexbds2sWPHDtVKBY6OjmL79u3NjmNYWJjqHFtbWzFlyhTx9ttvi65duwoAwsbGRkycOFG89dZbDa5aAECMHz++Rddr5cqVGv1oO1a6jL22Jk6cKHr16iXs7OyElZWV8PT0FMHBwWorDgghxL59+4S9vb2IiYlptK09e/Y0utKDlZWV6N27t1i0aJG4cOGC2nnGuufvjbdz586q1R3u98orr2gsdcZ7u/3f2w/I0+Ftiqs9UHv1gPw+75QIITr0JvASiQQpKSl45plnjB0KERHdZ/r06QCA1NRUI0fScezcuRNBQUHo4H+e6QH0gPw+p5rktAciIiIiMk1MftuRM2fONLrV6b2v4OBgY4dKzeC1JCIiap+Y/LYj3t7eWi01smPHDmOHSs3gtSQiYzl48CCWL1+O3bt3w8PDQ/Vme86cORp1x44dC3t7e5ibm2PAgAE4fvy4ESLWXV1dHeLi4uDv799onerqasTGxsLLywtSqRQKhQI+Pj4oLCzsUP22pL07d+7A29sbK1euVJV9/vnn2LhxY4s34nmQMPklIiJ6QKxevRoJCQlYsWIFpk6dioKCAnh6esLJyQlbt25Fenq6Wv0DBw4gNTUVgYGByMnJUS352Z7l5ubi0UcfRUREBCorKxutFxQUhI8//hjbtm1DZWUlfv31V3h6euL27dsdqt+WtBcVFYWzZ8+qlU2cOBEymQwBAQGqjW9MFXd4IyKiB5pSqURAQACOHDnSoftozuuvv44dO3YgKysLMplM7VhCQgLmzJmD+fPnIycnB3K53EhRtk5WVhaio6OxcOFCVFRUNPpQ4I4dO7B3715kZWWpNvVxdXVFWlpah+rh/v9AAAAgAElEQVS3Je0dOXIEp0+fbvDYkiVLUFBQgPHjx+PQoUOwsDDNNJCf/BIR0QNty5YtKCoq6vB9NCUvLw+rVq3C2rVrNRJfAPD390dYWBguX76MpUuXGiFC/Xj44Yexe/duzJo1C1ZWVo3We/fddzFkyBBVwthR+9W1PaVSiVdeeQXx8fGN1lmzZg1OnjzZZJ0HHZNfIiJqV4QQ2LRpE/r16wcrKys4Ojpi0qRJOHPmjKpOaGgopFIpunbtqipbvHgxbG1tIZFIVNtuh4WFITIyEvn5+ZBIJPDy8kJCQgJkMhlcXFywYMECuLq6QiaTwd/fX20Tkdb0AQBffvmlzpvStFRCQgKEEJg4cWKjdWJiYtCnTx9s3rwZBw8ebLI9ba5BUlISbG1tYWNjg7S0NIwbNw4ODg5wd3fH9u3b1dqrra3Fa6+9hu7du8Pa2hoDBw5UbeOsb1VVVTh27BgGDRpkkPbbqt+WtBcVFYXFixfD2dm50TqOjo4YPXo04uPjTXY5PSa/RETUrqxZswbLly9HVFQUioqKcOjQIVy8eBGjRo3CtWvXANxN9u5f3z0xMRFr165VK4uPj0dgYCA8PT0hhEBeXh5CQ0MREhKCyspKLFmyBIWFhTh+/DhqamrwxBNP4OLFi63uA4DqwaK6ujr9DU4j0tPT0bdvX9jY2DRax9raGh999BHMzMwwb948VFRUNFpXm2uwaNEihIeHQ6lUwt7eHikpKcjPz4eHhwfmzZunttvgsmXL8MYbbyAuLg5Xr15FYGAgZs6cqbHltj5cuXIFVVVV+PnnnzFmzBjVm5t+/fohMTHRYAmfvvvVtb3vv/8e+fn5mDlzZrNtDx48GJcvX0ZWVpZOMT0omPwSEVG7oVQqsWnTJkyZMgWzZ8+GXC6Hr68v3nvvPdy4cQMffPCB3vqysLBQfbLZv39/JCUloby8HMnJyXppf8KECSgrK8OqVav00l5jKioqcO7cOXh6ejZb18/PD+Hh4SgsLFRtqX2/llwDf39/ODg4wNnZGcHBwaioqMCFCxcA3F15ICkpCZMnT8bUqVOhUCiwcuVKWFpa6m2s71X/IJizszPWr1+PnJwcXLt2DZMmTcJLL72ETz/9VO99GqJfXdpTKpUICwtDUlKSVm337t0bAJCdna1TTA8KJr9ERNRu5OTk4Pbt2xg2bJha+fDhwyGVStWmJejbsGHDYGNjo/bVfkdQVFQEIUSTn/reKyYmBn379kViYiIOHz6scby110AqlQKA6pPfs2fPorKyEj4+Pqo61tbW6Nq1q0HGun5O7oABA+Dv749OnTpBLpdj7dq1kMvlen0DZch+dWlvxYoVePHFF+Hm5qZV2/X3Sv2n+KaGyS8REbUb9Usw2dnZaRxTKBQoLy83aP9WVla4fv26QfvQtzt37gBAkw9i3UsmkyE5ORkSiQRz586FUqlUO67va1A/vWLlypVqm/ycP3++ySXDWsrV1RUAVHOy60mlUvTo0QP5+fl679MQ/Wrb3uHDh5GdnY0XXnhB67atra0B/HnvmBomv0RE1G4oFAoAaDDBKikpgbu7u8H6rq6uNngfhlCfyOiyeYGfnx8iIiKQm5uLdevWqR3T9zWof/gqLi5OY6Ofo0eP6tSWNuzs7NC7d2/88ssvGsdqamoMtsybvvvVtr0tW7bg66+/hpmZmeqNRf2Yr1+/HhKJRGNudVVVFYA/7x1Tw+SXiIjaDR8fH9jZ2Wn8sc7MzERVVRWGDh2qKrOwsFB7qKq1MjIyIITAiBEjDNaHIbi4uEAikaC0tFSn89atWwdvb2+cOHFCrVyXa6CNbt26QSaT4eTJkzqd1xpBQUE4ceIECgoKVGWVlZU4f/683pYha4t+tWkvOTlZ401F/bcXUVFREEJoTGGpv1e6dOmic0wPAia/RETUbshkMkRGRmLPnj3YunUrysrKkJ2djYULF8LV1RXz589X1fXy8sLNmzexd+9eVFdX4/r16zh//rxGm506dcKVK1dQWFiI8vJyVTJbV1eHW7duoaamBqdOnUJYWBi6d++OkJAQvfSxf//+NlnqzMbGBh4eHrh06ZJO59VPfzA3N9co1/YaaNvP888/j+3btyMpKQllZWWora3FpUuXcPXqVQBAcHAwunTporftlSMiItCjRw+EhITgwoULKC4uxquvvgqlUqn2oF9771fb9nRVf68Y8o1Ae8bkl4iI2pXVq1cjNjYW0dHR6Ny5M0aPHo2ePXsiIyMDtra2qnqLFi3CmDFjMGPGDPTt2xfr1q1TfY3r5+enWrJs4cKFcHFxQf/+/TF+/HjcvHkTwN35jr6+vrC2tsaoUaPQp08ffPvtt2pzZ1vbR1uZMGECcnJy1ObvfvbZZ/Dy8kJ+fj6GDx+Ol19+WeO8ESNGICIiQqNcm2uQlJSEuLg4AMDAgQNRUFCADz/8EJGRkQCAp556Crm5uQDuLgcXHh6OjRs3wsnJCa6urggLC8OtW7cA3P0avqioqNmd0I4dO4aRI0fioYceQmZmJrKysuDq6opHHnkEhw4dUtVzdHTEd999B3d3dwwaNAhubm744YcfkJ6errZubnvvV9v2dPXjjz/Czc0NAwcObHEbHZlEdPAVjiUSCVJSUjTWYiQiIuObPn06ACA1NdXIkahbsGABUlNTUVxcbOxQNOzcuRNBQUE6rQubl5eHfv36ITk5GbNnzzZgdIZRV1eHxx57DCEhIZg7dy77NaDi4mK4u7sjJiZG9UZFW+3191lHqfzkl4iITJIuD4i1d15eXoiOjkZ0dLRqfdiOora2Fnv37kV5eTmCg4PZr4GtWbMGgwYNQmhoaJv33V4w+SUiInoALF++HNOnT0dwcLDOD78ZU0ZGBnbv3o39+/drvVYx+22ZTZs24eTJk9i3bx8sLS3btO/2hMkvERGZlBUrViA5ORmlpaXo1asXdu3aZeyQ9Gb9+vUIDQ3Fhg0bjB2K1gICArBt2zZ07dqV/RpQWloa/vjjD2RkZMDR0bFN+25vLIwdABERUVuKjY1FbGysscMwmLFjx2Ls2LHGDoPamaeffhpPP/20scNoF/jJLxERERGZDCa/RERERGQymPwSERERkclg8ktEREREJuOBeOAtLi6uoy+4TET0QDp27BiAPxfHp+bVbz3LMaP25tixYxgxYoSxw2i1Dr/DG/9zIKKO5vfff8eJEycwbtw4Y4dCRKQTPz+/BrfE7kBSO3zyS0TU0bRk+1oiItILbm9MRERERKaDyS8RERERmQwmv0RERERkMpj8EhEREZHJYPJLRERERCaDyS8RERERmQwmv0RERERkMpj8EhEREZHJYPJLRERERCaDyS8RERERmQwmv0RERERkMpj8EhEREZHJYPJLRERERCaDyS8RERERmQwmv0RERERkMpj8EhEREZHJYPJLRERERCaDyS8RERERmQwmv0RERERkMpj8EhEREZHJYPJLRERERCaDyS8RERERmQwmv0RERERkMpj8EhEREZHJYPJLRERERCaDyS8RERERmQwmv0RERERkMpj8EhEREZHJYPJLRERERCaDyS8RERERmQwmv0RERERkMpj8EhEREZHJsDB2AERED7Lq6mrcvn1brayiogIAcOvWLbVyiUQChULRZrEREZkiJr9ERAZ08+ZNuLm5oba2VuNYp06d1P49ZswYfPPNN20VGhGRSeK0ByIiA+rSpQseffRRmJk1/d+tRCLBjBkz2igqIiLTxeSXiMjA5syZ02wdc3NzTJkypQ2iISIybUx+iYgMbOrUqbCwaHyWmbm5OZ566ik4OTm1YVRERKaJyS8RkYE5ODhg3LhxjSbAQgjMnj27jaMiIjJNTH6JiNrA7NmzG3zoDQCkUin++te/tnFERESmickvEVEb+Otf/wobGxuNcktLS0yePBm2trZGiIqIyPQw+SUiagMymQxTpkyBpaWlWnl1dTVmzZplpKiIiEwPk18iojYyc+ZMVFdXq5U5ODjgiSeeMFJERESmh8kvEVEbefzxx9U2trC0tMSMGTMglUqNGBURkWlh8ktE1EYsLCwwY8YM1dSH6upqzJw508hRERGZFia/RERtaMaMGaqpD126dMHIkSONHBERkWlh8ktE1Ib8/f3h5uYGAHj22Web3faYiIj0q/Eth0hrR48excWLF40dBhF1EMOHD8fly5fh5OSEnTt3GjscIuog/P394e7ubuwwOjyJEEIYO4iObvr06di1a5exwyAiIqIHWEpKCp555hljh9HRpfKTXz2ZNm0aUlNTjR0GEXUQu3btwrRp0wzah0Qi4R9LHU2fPh0A+P85tTsSicTYITwwONmMiMgIDJ34EhFRw5j8EhEREZHJYPJLRERERCaDyS8RERERmQwmv0RERERkMpj8EhEREZHJYPJLRERN2rdvH+RyOf79738bO5R27+DBg1i+fDl2794NDw8PSCQSSCQSzJkzR6Pu2LFjYW9vD3NzcwwYMADHjx83QsS6q6urQ1xcHPz9/RutU11djdjYWHh5eUEqlUKhUMDHxweFhYUdqt+WtHfnzh14e3tj5cqVqrLPP/8cGzduRG1tbYviIP1i8ktERE3iXkjaWb16NRISErBixQpMnToVBQUF8PT0hJOTE7Zu3Yr09HS1+gcOHEBqaioCAwORk5ODIUOGGCly7eXm5uLRRx9FREQEKisrG60XFBSEjz/+GNu2bUNlZSV+/fVXeHp64vbt2x2q35a0FxUVhbNnz6qVTZw4ETKZDAEBASgpKWlRLKQ/3OSCiIiaNGHCBJSWlho7DACAUqlEQEAAjhw5YuxQ1Lz++uvYsWMHsrKyIJPJ1I4lJCRgzpw5mD9/PnJyciCXy40UZetkZWUhOjoaCxcuREVFRaNvinbs2IG9e/ciKysLvr6+AABXV1ekpaV1qH5b0t6RI0dw+vTpBo8tWbIEBQUFGD9+PA4dOgQLC6ZgxsJPfomIqMPYsmULioqKjB2Gmry8PKxatQpr167VSHwBwN/fH2FhYbh8+TKWLl1qhAj14+GHH8bu3bsxa9YsWFlZNVrv3XffxZAhQ1QJY0ftV9f2lEolXnnlFcTHxzdaZ82aNTh58mSTdcjwmPwSEVGjDh8+jO7du0MikeCdd94BACQlJcHW1hY2NjZIS0vDuHHj4ODgAHd3d2zfvl11bkJCAmQyGVxcXLBgwQK4urpCJpPB398fmZmZqnqhoaGQSqXo2rWrqmzx4sWwtbWFRCLBjRs3AABhYWGIjIxEfn4+JBIJvLy8AABffvklHBwcsH79+rYYEg0JCQkQQmDixImN1omJiUGfPn2wefNmHDx4sMn2hBDYtGkT+vXrBysrKzg6OmLSpEk4c+aMqo621wAAamtr8dprr6F79+6wtrbGwIEDkZKS0rofuhFVVVU4duwYBg0aZJD226rflrQXFRWFxYsXw9nZudE6jo6OGD16NOLj4zmdyIiY/BIRUaNGjhypMcVg0aJFCA8Ph1KphL29PVJSUpCfnw8PDw/MmzcP1dXVAO4mtSEhIaisrMSSJUtQWFiI48ePo6amBk888QQuXrwI4G7y+Mwzz6j1kZiYiLVr16qVxcfHIzAwEJ6enhBCIC8vDwBUDxHV1dUZZAyak56ejr59+8LGxqbROtbW1vjoo49gZmaGefPmoaKiotG6a9aswfLlyxEVFYWioiIcOnQIFy9exKhRo3Dt2jUA2l8DAFi2bBneeOMNxMXF4erVqwgMDMTMmTPx008/6W8Q/s+VK1dQVVWFn3/+GWPGjFG94enXrx8SExMNlvDpu19d2/v++++Rn5+PmTNnNtv24MGDcfnyZWRlZekUE+kPk18iImoxf39/ODg4wNnZGcHBwaioqMCFCxfU6lhYWKg+xezfvz+SkpJQXl6O5ORkvcQwYcIElJWVYdWqVXppTxcVFRU4d+4cPD09m63r5+eH8PBwFBYWYtmyZQ3WUSqV2LRpE6ZMmYLZs2dDLpfD19cX7733Hm7cuIEPPvhA45ymrsGdO3eQlJSEyZMnY+rUqVAoFFi5ciUsLS31Nv73qn8QzNnZGevXr0dOTg6uXbuGSZMm4aWXXsKnn36q9z4N0a8u7SmVSoSFhSEpKUmrtnv37g0AyM7O1ikm0h8mv0REpBdSqRQA1D51bMiwYcNgY2Oj9jV+R1VUVAQhRJOf+t4rJiYGffv2RWJiIg4fPqxxPCcnB7dv38awYcPUyocPHw6pVKo2XaQh91+Ds2fPorKyEj4+Pqo61tbW6Nq1q0HGv35O7oABA+Dv749OnTpBLpdj7dq1kMvlDSbv7bFfXdpbsWIFXnzxRbi5uWnVdv29Uv8pPrU9Jr9ERNTmrKyscP36dWOH0Wp37twBgCYfxLqXTCZDcnIyJBIJ5s6dC6VSqXa8fhksOzs7jXMVCgXKy8t1iq9+esXKlStVaw5LJBKcP3++ySXDWsrV1RUAVPO060mlUvTo0QP5+fl679MQ/Wrb3uHDh5GdnY0XXnhB67atra0B/HnvUNtj8ktERG2quroaJSUlcHd3N3YorVafyOiyeYGfnx8iIiKQm5uLdevWqR1TKBQA0GCS25Ixq3/4Ki4uDkIItdfRo0d1aksbdnZ26N27N3755ReNYzU1NQZb5k3f/Wrb3pYtW/D111/DzMxM9caifszXr18PiUSiMbe6qqoKwJ/3DrU9Jr9ERNSmMjIyIITAiBEjVGUWFhbNTpdoj1xcXCCRSHReB3ndunXw9vbGiRMn1Mp9fHxgZ2enkTBlZmaiqqoKQ4cO1amfbt26QSaT4eTJkzqd1xpBQUE4ceIECgoKVGWVlZU4f/683pYha4t+tWkvOTlZ401F/TcaUVFREEJoTGGpv1e6dOmic0ykH0x+iYjIoOrq6nDr1i3U1NTg1KlTCAsLQ/fu/5+9e4+Lqs7/B/46XIbLcFdABEEuKiK6fE1aofyp2bKpX++ieCnRSrwUIeQa4hWVIneB1aRSW/a7mYqAgZvatlqsWmpamoTlIl6QvIAIAgIywOf3hw+mRlAZmGGAeT0fD//ocz7nfd5zzhl6z8znfD6uCA0NVfbx8vLCnTt3kJmZCYVCgeLiYly9erVJLDs7O1y/fh1XrlxBRUUFFAoFDh48qLOpzszNzeHh4YHCwkK19msc/mBoaNikPSoqCnv37sWOHTtQXl6OnJwcLFy4EE5OTggLC1P7OHPnzsWuXbuQnJyM8vJy1NfXo7CwEDdu3AAAhISEwNHRUWPLK0dGRsLNzQ2hoaEoKChASUkJli1bhurqapUH/Tr6cVsaT12N94o2PwjQ47H4JSKiR3rvvffg7+8PAFi2bBkmTJiA5ORkJCYmAgAGDRqES5cuYdu2bYiKigIAvPDCC8jLy1PGqKmpwcCBA2FmZoZhw4ahb9+++Oqrr1TGyS5atAgjR47EjBkz0K9fP6xbt075s3BAQIByWrSFCxfCwcEBPj4+GDNmDO7cudMu5+Fxxo4di9zcXJXxu59++im8vLyQn58Pf39/vP766032Gzp0KCIjI5u0r169GnFxcYiNjUX37t0xfPhw9O7dG9nZ2ZDL5QCg1jVISkrCkiVLEB8fj27dusHJyQkREREoLS0F8OBn+KKioieuhHbixAk8++yz6NmzJ06ePIkffvgBTk5OeOaZZ3DkyBFlP1tbWxw9ehQuLi7w8/ODs7Mzvv32W+zfv19l3tyOftyWxlPXqVOn4OzsjEGDBrU6BrWNJDjLcpsFBwcDANLS0nScCRHRryRJQmpqapM5dNvTggULkJaWhpKSEp3loI7W/D2/ePEi+vfvj5SUFMyePVtbqWlNQ0MDRowYgdDQUMybN4/H1aKSkhK4uLhg/fr1yg8qLdUR3s9dRBq/+SUiIq1S52GwzsjLywuxsbGIjY1Vzg/bWdTX1yMzMxMVFRUICQnhcbVszZo18PPzQ3h4eLsfm37F4pc6pD//+c/KB0k++OADZfuBAwdgbW2Nf/7zn1o7dmxsLHx8fGBlZQUTExN4eXnhT3/60xP/p/bKK6/A0tISkiS16uGS+Ph4eHt7w8zMDHK5HN7e3li5ciXKy8vVjpWRkQEPDw/l08dPmvw/ISEBkiTBwMAA3t7eKj8lttXDuUiSBGNjYzg7O2PWrFn46aefNHash3X0+6i5cyNJEmQyGRwcHDBixAhs3LhR+fM0dVzR0dEIDg5GSEiI2g+/6VJ2djYyMjJw8ODBFs9VzOO2TkJCAs6ePYsDBw7A2Ni4XY9NDxHUZlOnThVTp07VdRpdTl5engAg3n//fWXbZ599JqysrMS+ffu0dtzhw4eLLVu2iJKSElFeXi5SU1OFsbGxeOGFF564765duwQAcebMGbWPO3bsWPHnP/9ZFBUViYqKCrFnzx5hbGws/vCHP7TmZQghhPD09BQARI8ePURtbW2zferq6oSbm5sAIEaNGtXqY7UkF2trayGEEJWVlWLfvn3C1dVVWFhYiJ9//llrx+0M99Fvz01DQ4MoLS0VX331lQgNDRWSJAknJydx6tQptXMAIFJTU9v8WlorOjpayGQyAUD07t1bpKWl6SyXlmrr3/N//etfYtmyZRrMiLqCzMxMERcXJ+rq6lodQ9fv5y5kD7/5pU5l7NixuHv3LsaNG6e1Y1hYWCAsLAx2dnawtLTEtGnTMGnSJHz++efKh260QSaTYfHixbC3t4eFhQWCg4MxceJE/Pvf/1Y+ld0aTz31FG7evInMzMxmt2dkZLR4ZSJNkcvlGDduHP7617+isrISmzdvbtfjd+T7SJIk2NjYYMSIEUhJScGePXtw69YtZc6dSVxcHO7fvw8hBC5fvoypU6fqOiWtCwoKwjvvvKPrNKiDmTBhAqKjo5vM7kG6weKX9JoQAmlpaSpLVX722WdN/kB1794dAJ64IpIkSa3OZe/evTA1NVVpayxK2zKOcNGiRQCA999/v9ntCQkJaj94oSlPP/00AODHH3/UyfE1RdP30W9NnToVoaGhKCoqUhm6QURErcPiVweSkpIgl8thYGCAp556Co6OjjA2NoZcLsfgwYMxbNgw5cTkNjY2+NOf/qSy/9GjR+Hj4wNra2uYmppi4MCB+Ne//gUA+Pvf/w4LCwtIkgRbW1tkZmbi9OnTcHNzg6GhIWbOnKlWrps2bYKpqSkcHBywYMECODk5wdTUFIGBgU3WmBdCICEhAf3794eJiQlsbW0xceLEJuvHt7Tfw44dOwZXV1dIkoT33nsPwIPpfuRyOczNzZGVlYXRo0fDysoKLi4u2LVrl8r+9fX1iIuLQ79+/WBmZobu3bvD3d0dcXFxT3x69pdffoGZmRnc3d1VXsfGjRvRr18/mJiYwNraGkuXLn3iOVVHXl4ebGxs4Obmpmz7/PPP1ZrT9LnnnkP//v3x1Vdf4cKFCyrbvv76a1RVVSEoKKjZfbV9r9XV1QFQXRpW3+6jlmicD/fgwYNq7UdERM3Q3ZCLrqM1Y8RWr14tAIiTJ0+Ke/fuidu3b4sXXnhBABD79+8XxcXF4t69eyI8PFwAEGfPnlXum5aWJtasWSPu3LkjSkpKxNChQ0W3bt2U28+fPy/Mzc3FnDlzlG3R0dFi+/btrXp9YWFhQi6Xi/Pnz4uamhqRm5sr/P39haWlpSgoKFD2W7VqlZDJZOLjjz8WZWVl4ty5c2Lw4MGie/fu4ubNm2r3a26s5rVr1wQAsXnzZmVbTEyMACAOHz4s7t69K4qKisSwYcOEXC5XGee6YcMGYWhoKLKyskRVVZX47rvvhKOjoxgxYsRjX/+9e/eEpaWlCA8PV2mPiYkRkiSJv/zlL6K0tFRUVVWJLVu2tHrMb6Pa2lpRWFgoNm/eLExMTMTHH3+ssv2zzz4TlpaWIjY29omxPD09xeXLl8Vf//pXAUBERESobJ80aZJISUkRFRUVzY751eS99ttxrY0+/vhjAUAsXbpU2aZv99Gjzs1vlZeXCwCiV69ejz3Gw8AxgmrjMxzUUfH9rDF7WPxqQFuK34qKCmXb//3f/wkAIicnR9n27bffCgBi9+7dj4wVFxcnAIiioiJl24cffigAiB07doidO3eKyMhItfL7rbCwsCb/Yz516pQAINauXSuEEKKqqkpYWFiIkJAQlX6N+TcWai3tJ4T6RUt1dbWyrbEIvXjxorLN399fPP300yrHnT9/vjAwMBD3799/5OuPiYkRffv2FeXl5cq2qqoqYW5u3uRhtLY88NbI0dFRABDdunUTf/3rXx/5oFpLNBa/ZWVlQi6XC1tbW1FVVSWEECI/P1+4uLiI+/fvP7L4fVhb7rWHH3hLT08Xjo6OwsHBQRQWFgoh9O8+au7cPIokScLGxuaxfR7G/1mqj8UvdVR8P2vMHiOtfaVMapPJZAB+/SkYgHI6lMeted/Y57dzac6fPx///ve/sWDBAjz//PNIT0/XaK5DhgyBubm58ifm3NxcVFZWNlnD3N/fHzKZTDlEoqX92qrxXP72vNXU1DQZU1tfXw9jY+NHPoSwd+9e7NmzB1988QUsLS2V7RcvXkRVVRVGjRqlkXx/69q1aygrK8OZM2cQHR2NrVu34ssvv4SDg0OrY1pbW2PmzJnYtm0bdu/ejblz5yIxMRGLFi2CTCZDbW1ti+K09V67e/cuJEmCoaEhevTogTFjxmD16tXKsc36dh+11L179yCEgJWVldr7JiYmcgEeNZw4cQLAr4tdEFHXwzG/ndD+/fsxYsQI2Nvbw8TEpMmY4EYbNmxAZWUlioqKtJKHiYkJiouLAQBlZWUAHjzh/jAbGxtUVFSo1U8bxowZg++++w5ZWVmorq7G6dOnkZmZif/93/9ttmjZvXs33nnnHWRnZ6N3794q2xrXZre3t9d4nsbGxrC3t0dQUBB2796N3NxcxMXFtTlu44NvH3zwAcrKypCWloYFCxY8dh9N32vW1tYQQqCurg6FhYX42/7IGrIAACAASURBVN/+pjKeWd/uo5b673//CwDw9vZuS+pERASA3/x2MgUFBZg0aRImT56Mv/3tb+jZsyc2b97cpChRKBR44403lE/yr1+/HqtXr9ZYHgqFAmVlZXBxcQHwoOAA0GzR0Zp+2rBmzRp89913CA0NRWVlJZycnDBt2rRmHxzbvHkz/vWvf+HLL79stsBq/Obv/v37WssXeLBylKGhIXJzc9scy8/PD0OHDsWJEycQFhaG4OBg2NraPrK/Lu41fbuPWurzzz8HAIwePVrtfZcsWcLlUNXA5eqpo2rLbEKkisVvJ5OTkwOFQoFFixbBw8MDQPNviNdffx2vvvoqJk+ejF9++QXr1q1DUFAQAgICNJJHdnY2hBAYOnQoAMDX1xcWFhY4ffq0Sr+TJ0+itrYWTz31lFr9tCE3Nxf5+fkoLi6GkVHzt74QAm+99RZKS0uRmZn5yH6+vr4wMDDAf/7zHyxcuLDNuZWUlOD111/Hzp07Vdrz8vJQX1+PXr16tfkYwINvf0+cOIH09HTk5eU9tq8u7jV9u49a4ubNm0hMTISLiwvmzZvX6jhERPQAhz10Mq6urgCAQ4cOoaamBnl5eU3GN27ZsgXOzs6YPHkygAcTzfv4+GDWrFmtWioXABoaGlBaWoq6ujqcO3cOERERcHV1VU7BZGpqiqioKOzduxc7duxAeXk5cnJysHDhQjg5OSEsLEytftrw2muvwdXV9bFz5p4/fx7vvvsutm3bBmNj4ybLzv75z38G8GC4w5QpU5Ceno6PPvoI5eXlOHfunMo8r+qQy+X44osv8OWXX6K8vBwKhQJnzpzBnDlzIJfLERkZqex78OBBtaY6+61p06ahe/fumDRpkrKgfRRd3Gv6dh/9lhAClZWVaGhogBACxcXFSE1NxTPPPANDQ0NkZma2aswvERE9RIdP23UZ6j4dnJSUJMzNzZVLfh49elS88847wtraWgAQjo6O4pNPPhG7d+9WPvlva2srdu3aJYQQYtmyZcLOzk7Y2NiI4OBg8d577wkAwtPTU/j5+QlJkoSdnZ345ptvhBBCLFmyRBgYGAgAwtraWpw+fVqt1xcWFiaMjY2Fs7OzMDIyElZWVmLixIkiPz9fpV9DQ4PYuHGj6NOnjzA2Nha2trZi0qRJ4sKFC2r3+8tf/qJ87XK5XEyePFls3rxZ9OjRQwAQ5ubmYvz48WLLli3Kc9mnTx+Rn58vtm7dKqysrAQA4ebmJv773/8KIYT48ssvRbdu3QQA5T9jY2PRv39/kZGRIYQQIicnR2X7w/82btyozLGiokK88sorolu3bsLCwkI8++yzYtWqVQKAcHFxET/88INa53n8+PHC3d1dWFhYCBMTE+Hp6SlCQkJUZv8QQogDBw4IS0tLsX79+kfG2rt3r3Jp4+7du4vXXntNue1Pf/qT8t4QQogVK1Yoz6uBgYHw8fERR48eFUJo5l77+uuvRd++fZXn0MnJSQQHBz8yd326j/bt2ycGDRokzM3NhUwmU567xpkdnn76aREbGytKSkoefeM8Bvh0uNo42wN1VHw/a8weSQghtFRX642uPkZswYIFSEtLQ0lJia5TaZPk5GTk5eUhMTFR2VZbW4u33noLycnJKC0thZmZmQ4zpM6gM91HkiQhNTWVY37V0NX/nlPnxfezxqRxzC+1yG+ntuqMbt68ifDwcJw9e1alXSaTwdXVFQqFAgqFosMULdQx8T4iIur8OOZXD/38889NxiA29y8kJETXqWqMmZkZjI2N8dFHH+HWrVtQKBS4fv06tm/fjlWrViEkJETj4yn18Tx3dbq4j6hrOnToEKKjo5GRkQEPDw/l34MXX3yxSd+goCBYWlrC0NAQAwYMwPfff6+DjNWzc+dO+Pv7w9LSEm5ubpg7dy5u3rzZYeIpFArExcXBy8sLMpkMNjY28PX1xZUrVx65T01NDby9vbFixQpl2759+xAfH9/pvyDSO7oeeNEVdOUxYtHR0UImkynHJ6elpek6pVY7cuSIeP7554WVlZUwNDQU1tbWIjAwUGzZskUoFApdp0edRGe6j8Axgmprj7/nq1atEuPGjVNZ7c/T01M5lvyzzz5rss/BgwfFhAkTtJqXpuzevVsAEPHx8aKsrEycOXNGeHh4CD8/v1a9RzQdT4gHS7v369dPnDhxQigUCnH9+nUxfvz4Js9Y/FZkZKQAIGJiYlTak5KSxPDhw0VpaWmrcmkpvp81hssba0JXLn6JqPPS9f8sq6qqREBAQKc6hrb/nr/99tuib9++KstoC/Gg+P3kk0+EgYGBcHZ2FmVlZSrbO1PxO3LkSNGzZ0/R0NCgbGt8WPbYsWM6j7dr1y4hSZI4d+5ci/f5+uuvRVBQULPFrxBChIeHi4CAAK1+ANb1+7kL2cNhD0REpBUfffSR1laYbM9jaMrFixexcuVKrF27tskS2QAQGBiIiIgI/PLLL3jzzTd1kKFmXLt2DU5OTirzgjfOVX716lWdx3v//fcxePBgDBw4sEX9q6ursXTpUiQlJT2yz5o1a3D27NnH9qGOg8UvEREBeDDXcEJCAvr37w8TExPY2tpi4sSJ+Pnnn5V9wsPDIZPJ0KNHD2Xb4sWLIZfLIUkSbt++DQCIiIhAVFQU8vPzIUkSvLy8sGnTJpiamsLBwQELFiyAk5MTTE1NERgYqDKHdFuOATxYEa+1c2Fr06ZNmyCEwPjx4x/ZZ/369ejbty+2b9+OQ4cOPTZeS65XcnIy5HI5zM3NkZWVhdGjR8PKygouLi7YtWuXSrz6+nqsWrUKrq6uMDMzw6BBg5Camqr26/Tw8GjygaRxfO6T5hfXdrza2lqcOHECfn5+Ld4nJiYGixcvfuxy9ra2thg+fDiSkpIgOIlWx6fbb567Bg57IKKOCGr+TLpq1Sohk8nExx9/LMrKysS5c+fE4MGDRffu3cXNmzeV/WbNmiUcHR1V9t24caMAIIqLi5VtU6ZMEZ6enir9wsLChFwuF+fPnxc1NTUiNzdX+Pv7C0tLS1FQUKCRY3z22WfC0tJSxMbGtvi1N9Lm33MPDw/h4+PT7DZPT09x+fJlIYQQ33zzjTAwMBC9e/cWlZWVQojmhz209HrFxMQIAOLw4cPi7t27oqioSAwbNkzI5XJRW1ur7Pfmm28KExMTkZ6eLkpLS8Xy5cuFgYGBOHXqlFqvMzs7WxgbG4tNmzaJ8vJy8eOPP4r+/fuLP/7xj2rF0Ua8y5cvCwDCz89PjBgxQvTo0UOYmJgIb29v8d5776kMrRBCiGPHjonx48cLIYQoLi5+5LAHIR48IwNAnDlzRv0X2QLqvp/pkTjsgYiIHvy0m5CQgMmTJ2P27NmwtrbGwIED8cEHH+D27dutXr2wOUZGRspvK318fJCcnIyKigqkpKRoJP7YsWNRXl6OlStXaiSeJty7dw+XL1+Gp6fnE/sGBARgyZIluHLlCt56661m+7TmegUGBsLKygr29vYICQnBvXv3UFBQAODBTAbJycmYNGkSpkyZAhsbG6xYsQLGxsZqX5fhw4dj2bJlCA8Ph5WVFXx9fVFRUYHt27erFUcb8RpXZrS3t8eGDRuQm5uLW7duYeLEiXjttddUlpivrq5GREQEkpOTWxS7T58+AB4sDU8dG4tfIiJCbm4uKisrMWTIEJV2f39/yGSyJktba9KQIUNgbm6u8nN9V1NUVAQhBMzNzVvUf/369ejXrx+2bNmCY8eONdne1uslk8kAPJjyCwAuXLiAqqoq+Pr6KvuYmZmhR48eal+XmJgYbN26FYcPH0ZlZSUuXbqEwMBABAQE4Nq1a2rF0nQ8ExMTAMCAAQMQGBgIOzs7WFtbY+3atbC2tlb50LB8+XLMnz8fzs7OLYrdeG1v3bqlVk7U/lj8EhERysrKAAAWFhZNttnY2KCiokKrxzcxMUFxcbFWj6FLNTU1AH4tvp7E1NQUKSkpkCQJ8+bNQ3V1tcp2TV+ve/fuAQBWrFihMg/51atXUVVV1eI4N27cQHx8PObPn4/nnnsOcrkc7u7u2LZtG65fv46NGzeqlZem4zk5OQGActx4I5lMBjc3N+Tn5wMAjh07hpycHLzyyistjt24uE3jtaaOi8UvERHBxsYGAJotmsrKyuDi4qK1YysUCq0fQ9caCyN1FkMICAhAZGQk8vLysG7dOpVtmr5ejQ9zJSYmQgih8u/48eMtjpOXl4f6+nr07NlTpd3Kygp2dnbIzc1VKy9Nx7OwsECfPn1w/vz5Jtvq6upgbW0N4MEsIocPH4aBgYHyg0DjOdqwYQMkScLp06dV9q+trQUArvDYCbD4JSIi+Pr6wsLCosn/0E+ePIna2lo89dRTyjYjIyPlz+WakJ2dDSEEhg4dqrVj6JqDgwMkScLdu3fV2m/dunXw9vbGmTNnVNrVuV4t0atXL5iamjZZultdjUX3jRs3VNorKipw584d5RRluooHANOnT8eZM2dw6dIlZVtVVRWuXr2qnP4sJSWlyYeAxl8mYmJiIIRoMuSk8do6OjqqnRO1Lxa/REQEU1NTREVFYe/evdixYwfKy8uRk5ODhQsXwsnJCWFhYcq+Xl5euHPnDjIzM6FQKFBcXNzsfKt2dna4fv06rly5goqKCmUx29DQgNLSUtTV1eHcuXOIiIiAq6srQkNDNXKMgwcPdripzszNzeHh4YHCwkK19msc/mBoaNikvaXXq6XHmTt3Lnbt2oXk5GSUl5ejvr4ehYWFysIzJCQEjo6Oj11e2d3dHSNHjsS2bdtw5MgRVFdX49q1a8p8Xn75ZWVfXcQDgMjISLi5uSE0NBQFBQUoKSnBsmXLUF1d/cgHDFui8dq2dP5g0h0Wv0REBABYvXo14uLiEBsbi+7du2P48OHo3bs3srOzIZfLlf0WLVqEkSNHYsaMGejXrx/WrVun/Kn3tw8hLVy4EA4ODvDx8cGYMWNw584dAA/GRA4cOBBmZmYYNmwY+vbti6+++kplPGxbj9ERjR07Frm5uSrjdz/99FN4eXkhPz8f/v7+eP3115vsN3ToUERGRjZpb8n1Sk5ORmJiIgBg0KBBuHTpErZt24aoqCgAwAsvvIC8vDwAQFJSEpYsWYL4+Hh069YNTk5OiIiIQGlpKYAHP+sXFRUhKyvrka9RkiSkpaUhJCQEL7/8MmxtbeHj44OCggJkZGRg2LBhyr66iAc8mJP36NGjcHFxgZ+fH5ydnfHtt99i//79as3/+7BTp07B2dkZgwYNanUMah+SEJyNua2Cg4MBAGlpaTrOhIjoV5IkITU1FdOmTdN1KkoLFixAWloaSkpKdJ1Ks7T59/zixYvo378/UlJSMHv2bI3H17aGhgaMGDECoaGhmDdvXpePp46SkhK4uLhg/fr1yg8WmtYR38+dVBq/+SUionalzkNfXYmXlxdiY2MRGxurnG+2s6ivr0dmZiYqKioQEhLS5eOpa82aNfDz80N4eHi7H5vUx+KXiIionURHRyM4OBghISFqP/ymS9nZ2cjIyMDBgwdbPFdxZ46njoSEBJw9exYHDhyAsbFxux6bWofFLxERtYvly5cjJSUFd+/ehbu7O9LT03Wdkk5s2LAB4eHhePvtt3WdSouNGjUKn3zyCXr06KEX8VoqKysL9+/fR3Z2Nmxtbdv12NR6RrpOgIiI9ENcXBzi4uJ0nUaHEBQUhKCgIF2nQW00YcIETJgwQddpkJr4zS8RERER6Q0Wv0RERESkN1j8EhEREZHeYPFLRERERHqDxS8RERER6Q3O9qAh6enpkCRJ12kQEamYPn06pk+frus0Oh3+PSfquri8sQYcP35cuc48EdGTHD9+HElJSUhNTdV1KkTUiQQGBsLFxUXXaXR2aSx+iYja2Z49ezB9+nTwzy8RUbtL45hfIiIiItIbLH6JiIiISG+w+CUiIiIivcHil4iIiIj0BotfIiIiItIbLH6JiIiISG+w+CUiIiIivcHil4iIiIj0BotfIiIiItIbLH6JiIiISG+w+CUiIiIivcHil4iIiIj0BotfIiIiItIbLH6JiIiISG+w+CUiIiIivcHil4iIiIj0BotfIiIiItIbLH6JiIiISG+w+CUiIiIivcHil4iIiIj0BotfIiIiItIbLH6JiIiISG+w+CUiIiIivcHil4iIiIj0BotfIiIiItIbLH6JiIiISG+w+CUiIiIivcHil4iIiIj0BotfIiIiItIbLH6JiIiISG+w+CUiIiIivcHil4iIiIj0hpGuEyAi6sqKi4vx6aefqrSdPn0aALB161aVdktLS8yYMaPdciMi0keSEELoOgkioq7q/v37cHBwQGVlJQwNDQEAjX92JUlS9lMoFJgzZw7+/ve/6yJNIiJ9kcZhD0REWmRiYoKpU6fCyMgICoUCCoUCdXV1qKurU/63QqEAAMycOVPH2RIRdX0sfomItGzmzJmora19bB8bGxs899xz7ZQREZH+YvFLRKRlI0eOhL29/SO3GxsbY/bs2TAy4mMYRETaxuKXiEjLDAwMMGvWLBgbGze7XaFQ8EE3IqJ2wuKXiKgdzJgxQzm292E9e/ZEQEBAO2dERKSfWPwSEbWDp59+Gm5ubk3aZTIZ5syZozLzAxERaQ+LXyKidvLiiy82GfpQW1vLIQ9ERO2IxS8RUTuZNWtWk6EPXl5eGDhwoI4yIiLSPyx+iYjaibe3N3x8fJRDHIyNjTF37lwdZ0VEpF9Y/BIRtaOXXnpJudJbXV0dhzwQEbUzFr9ERO1oxowZqK+vBwAMHjwY7u7uOs6IiEi/sPglImpHrq6u+P3vfw8AmDNnjo6zISLSP1xOiDQuISEBx48f13UaRB3W/fv3IUkSvvjiCxw5ckTX6RB1WJGRkZwDmzSO3/ySxh0/fhwnTpzQdRpEbZaeno7CwkKNx3VxcYGjoyNMTU01HlvXTpw4wfc/aUR6ejquXbum6zSoC+I3v6QVQ4cORVpamq7TIGoTSZKwZMkSTJs2TeOxL168CC8vL43H1bXg4GAA4Puf2owLv5C28JtfIiId6IqFLxFRZ8Dil4iIiIj0BotfIiIiItIbLH6JiIiISG+w+CUiIiIivcHil4hIyw4cOABra2v885//1HUqHd6hQ4cQHR2NjIwMeHh4QJIkSJKEF198sUnfoKAgWFpawtDQEAMGDMD333+vg4zVs3PnTvj7+8PS0hJubm6YO3cubt682WHiKRQKxMXFwcvLCzKZDDY2NvD19cWVK1ceuU9NTQ28vb2xYsUKZdu+ffsQHx+vXM2QqCNh8UtEpGVCCF2n0CmsXr0amzZtwvLlyzFlyhRcunQJnp6e6NatG3bs2IH9+/er9P/iiy+QlpaGcePGITc3F4MHD9ZR5i2TmpqKWbNmITg4GIWFhcjKysKRI0cwevRo1NXV6TweAEyfPh3/+Mc/8Mknn6Cqqgo//fQTPD09UVlZ+ch9YmJicOHCBZW28ePHw9TUFKNGjUJZWVmrciHSFha/RERaNnbsWNy9exfjxo3TdSqorq5GYGCgrtNo4p133sHu3buxZ88eWFpaqmzbtGkTDAwMEBYWhrt37+oow7b78MMP0bNnTyxduhTW1tbw8/NDZGQkzp49i5MnT+o83u7du5GZmYm0tDT8/ve/h5GREZycnJCVlQVfX99m9/nmm2/w448/NrvtjTfewO9+9zuMGTOm1cU4kTaw+CUi0iMfffQRioqKdJ2GiosXL2LlypVYu3Zts6veBQYGIiIiAr/88gvefPNNHWSoGdeuXYOTk5PK4g29evUCAFy9elXn8d5//30MHjwYAwcObFH/6upqLF26FElJSY/ss2bNGpw9e/axfYjaG4tfIiItOnbsGFxdXSFJEt577z0AQHJyMuRyOczNzZGVlYXRo0fDysoKLi4u2LVrl3LfTZs2wdTUFA4ODliwYAGcnJxgamqKwMBAlW/2wsPDIZPJ0KNHD2Xb4sWLIZfLIUkSbt++DQCIiIhAVFQU8vPzIUmScqGNzz//HFZWVtiwYUN7nJImNm3aBCEExo8f/8g+69evR9++fbF9+3YcOnTosfGEEEhISED//v1hYmICW1tbTJw4ET///LOyT0uvAQDU19dj1apVcHV1hZmZGQYNGoTU1FS1X6eHh0eTDx6N43M9PDx0Gq+2thYnTpyAn59fi/eJiYnB4sWLYW9v/8g+tra2GD58OJKSkjj8hzoMFr9ERFr07LPP4ptvvlFpW7RoEZYsWYLq6mpYWloiNTUV+fn58PDwwKuvvgqFQgHgQVEbGhqKqqoqvPHGG7hy5Qq+//571NXV4Q9/+AOuXbsG4EHx+PASzFu2bMHatWtV2pKSkjBu3Dh4enpCCIGLFy8CgPKhpIaGBq2cgyfZv38/+vXrB3Nz80f2MTMzw9///ncYGBjg1Vdfxb179x7Zd82aNYiOjkZMTAyKiopw5MgRXLt2DcOGDcOtW7cAtPwaAMBbb72Fd999F4mJibhx4wbGjRuHmTNn4vTp02q9zuXLl+PmzZvYvHkzKioqkJubi6SkJPzxj3/E0KFD1Yql6XjXr19HbW0tvvvuO4wcOVL5Qat///7YsmVLk8L166+/Rn5+PmbOnPnE2P/zP/+DX375BT/88INaORFpC4tfIiIdCgwMhJWVFezt7RESEoJ79+6hoKBApY+RkZHyW0wfHx8kJyejoqICKSkpGslh7NixKC8vx8qVKzUSTx337t3D5cuX4enp+cS+AQEBWLJkCa5cuYK33nqr2T7V1dVISEjA5MmTMXv2bFhbW2PgwIH44IMPcPv2bWzdurXJPo+7BjU1NUhOTsakSZMwZcoU2NjYYMWKFTA2Nlb7/A8fPhzLli1DeHg4rKys4Ovri4qKCmzfvl2tONqI1/hAm729PTZs2IDc3FzcunULEydOxGuvvYadO3cq+1ZXVyMiIgLJycktit2nTx8AQE5Ojtp5EWkDi18iog5CJpMBgMq3js0ZMmQIzM3NVX7G76yKiooghHjst76/tX79evTr1w9btmzBsWPHmmzPzc1FZWUlhgwZotLu7+8PmUz2xAfBHr4GFy5cQFVVlcoDX2ZmZujRo4fa5z8mJgZbt27F4cOHUVlZiUuXLiEwMBABAQHKb/F1Fc/ExAQAMGDAAAQGBsLOzg7W1tZYu3YtrK2tVT40LF++HPPnz4ezs3OLYjde28Zv3Yl0jcUvEVEnZGJiguLiYl2n0WY1NTUAfi2+nsTU1BQpKSmQJAnz5s1DdXW1yvbGabUsLCya7GtjY4OKigq18mscXrFixQrlnMOSJOHq1auoqqpqcZwbN24gPj4e8+fPx3PPPQe5XA53d3ds27YN169fx8aNG9XKS9PxnJycAEA5PryRTCaDm5sb8vPzATwYw56Tk4NXXnmlxbHNzMwA/HqtiXSNxS8RUSejUChQVlYGFxcXXafSZo2FkTqLIQQEBCAyMhJ5eXlYt26dyjYbGxsAaLbIbc05a3yYKzExEUIIlX/Hjx9vcZy8vDzU19ejZ8+eKu1WVlaws7NDbm6uWnlpOp6FhQX69OmD8+fPN9lWV1cHa2trAA9mCzl8+DAMDAyUHwQaz9GGDRsgSVKTsdC1tbUAfr3WRLrG4peIqJPJzs6GEELloSYjI6MnDpfoiBwcHCBJktrz965btw7e3t44c+aMSruvry8sLCyaFGAnT55EbW0tnnrqKbWO06tXL5iamuLs2bNq7fewxqL7xo0bKu0VFRW4c+eOcooyXcUDHixwcebMGVy6dEnZVlVVhatXryqnP0tJSWnyIaDxF4iYmBgIIZoMOWm8to6OjmrnRKQNLH6JiDq4hoYGlJaWoq6uDufOnUNERARcXV0RGhqq7OPl5YU7d+4gMzMTCoUCxcXFzc71amdnh+vXr+PKlSuoqKiAQqHAwYMHdTbVmbm5OTw8PFBYWKjWfo3DHwwNDZu0R0VFYe/evdixYwfKy8uRk5ODhQsXwsnJCWFhYWofZ+7cudi1axeSk5NRXl6O+vp6FBYWKgvPkJAQODo6PnZ5ZXd3d4wcORLbtm3DkSNHUF1djWvXrinzefnll5V9dREPACIjI+Hm5obQ0FAUFBSgpKQEy5YtQ3V19SMfMGyJxmvb0vmDibSNxS8RkRa999578Pf3BwAsW7YMEyZMQHJyMhITEwEAgwYNwqVLl7Bt2zZERUUBAF544QXk5eUpY9TU1GDgwIEwMzPDsGHD0LdvX3z11Vcq42QXLVqEkSNHYsaMGejXrx/WrVun/Jn5tw9ALVy4EA4ODvDx8cGYMWNw586ddjkPjzN27Fjk5uaqjN/99NNP4eXlhfz8fPj7++P1119vst/QoUMRGRnZpH316tWIi4tDbGwsunfvjuHDh6N3797Izs6GXC4HALWuQVJSEpYsWYL4+Hh069YNTk5OiIiIQGlpKYAHP+sXFRUhKyvrka9RkiSkpaUhJCQEL7/8MmxtbeHj44OCggJkZGRg2LBhyr66iAc8mJP36NGjcHFxgZ+fH5ydnfHtt99i//79as3/+7BTp07B2dkZgwYNanUMIk2SBGedJg0LDg4GAKSlpek4E6K2kSQJqampTebQbU8LFixAWloaSkpKdJaDOlrz/r948SL69++PlJQUzJ49W1upaU1DQwNGjBiB0NBQzJs3r8vHU0dJSQlcXFywfv165QeLluoI7z/qktL4zS8RUQenzsNgnZGXlxdiY2MRGxurnG+2s6ivr0dmZiYqKioQEhLS5eOpa82aNfDz80N4eHi7H5voUVj8EhGRzkVHRyM4OBghISFqP/ymS9nZ2cjIyMDBgwdbPFdxZ46njoSEBJw9exYHDhyAsbFxux6b6HFY/FKH9Morr8DS0hKSJLX5KevOLj4+Ht7e3jAzM4NcLoe3tzdWrlyJ8vJytWNlZGTAw8NDZb5SSZIgk8ng4OCAESNGYOPGjcqxjKRby5cvfaIYfwAAIABJREFUR0pKCu7evQt3d3ekp6frOiWt2rBhA8LDw/H222/rOpUWGzVqFD755BP06NFDL+K1VFZWFu7fv4/s7GzY2tq267GJnoTFL3VI27dvx7Zt23SdRodw9OhRvPrqqygoKMCtW7ewbt06xMfHY+rUqWrHmjJlCi5dugRPT09YW1tDCIGGhgYUFRVhz549cHd3x7JlyzBgwIAmU0VR+4uLi8P9+/chhMDly5dbdc07m6CgILzzzju6ToPaaMKECYiOjm4yGwdRR8Dil6gdVFdXIzAwsFX7ymQyLF68GPb29rCwsEBwcDAmTpyIf//7303m+GwNSZJgY2ODESNGICUlBXv27MGtW7cwduzYTvXz86O05dwTEVHXw+KXOixJknSdgsZ89NFHKCoqatW+e/fuhampqUqbs7MzAGjl4aCpU6ciNDQURUVF+OCDDzQev7215dwTEVHXw+KXOgQhBDZu3Ih+/frBxMQE1tbWWLp0qUqfd999F+bm5rC0tERRURGioqLg7OyMCxcuQAiBhIQE9O/fHyYmJrC1tcXEiRPx888/K/fftGkTTE1N4eDggAULFsDJyQmmpqYIDAzEyZMnm+TzpHjh4eGQyWQqY+kWL14MuVwOSZJw+/ZtAEBERASioqKQn58PSZLg5eXV5vOVl5cHGxsbuLm5Kds+//xzjS1U0Lh4wsGDBwHw3BMRURciiDRs6tSpYurUqWrtExMTIyRJEn/5y19EaWmpqKqqElu2bBEAxJkzZ1T6ARBvvPGG2Lx5s5g8ebL46aefxKpVq4RMJhMff/yxKCsrE+fOnRODBw8W3bt3Fzdv3lTuHxYWJuRyuTh//ryoqakRubm5wt/fX1haWoqCggJlv5bGmzVrlnB0dFR5LRs3bhQARHFxsbJtypQpwtPTU61z8rDa2lpRWFgoNm/eLExMTMTHH3+ssv2zzz4TlpaWIjY29omxPD09hbW19SO3l5eXCwCiV69eyjZ9PPcARGpqaqv21Vetef8TNYfvP9KSPfzml3SuuroaiYmJeP755xEZGQkbGxuYmZnBzs7ukfu88847eO2115CRkQE3NzckJCRg8uTJmD17NqytrTFw4EB88MEHuH37NrZu3aqyr5GRkfJbRR8fHyQnJ6OiogIpKSnKfNSJ11569eoFFxcXrFmzBu+++y6mT5+usn3s2LEoLy/HypUr23ysxpk2KioqmmzTx3NPRERdh5GuEyC6ePEiqqqqMGrUqFbtn5ubi8rKSgwZMkSl3d/fHzKZrMnP6g8bMmQIzM3NlT+rtzWetly7dg1lZWU4c+YMoqOjsXXrVnz55ZdwcHDQ+LHu3bsHIQSsrKwe208fzv306dObfNCgJ+tKY/aJqGth8Us6V1hYCACwt7dv1f5lZWUAAAsLiybbbGxsmv328mEmJiYoLi7WWDxtMDY2hr29PYKCguDu7o6+ffsiLi4OSUlJGj/Wf//7XwCAt7f3Y/vpw7mPiIhAQEBAux+3s0pMTAQALFmyRMeZUGfHD52kLSx+SecaZzK4f/9+q/a3sbEBgGYLo7KyMri4uDx2f4VCodKvrfHag5eXFwwNDZGbm6uV+J9//jkAYPTo0Y/tpw/nPiAgANOmTWv343ZWaWlpAMBzRm3G4pe0hWN+Sed8fX1hYGCA//znP63e38LCosmiDCdPnkRtbS2eeuqpx+6fnZ0NIQSGDh2qdjwjIyMoFIpW5d0SJSUlmDlzZpP2vLw81NfXo1evXho/5s2bN5GYmAgXFxfMmzfvsX278rknIqKuicUv6Zy9vT2mTJmC9PR0fPTRRygvL8e5c+da/HCTqakpoqKisHfvXuzYsQPl5eXIycnBwoUL4eTkhLCwMJX+DQ0NKC0tRV1dHc6dO4eIiAi4uroqp/dSJ56Xlxfu3LmDzMxMKBQKFBcX4+rVq01ytLOzw/Xr13HlyhVUVFS0uGiTy+X44osv8OWXX6K8vBwKhQJnzpzBnDlzIJfLERkZqex78OBBtaY6E0KgsrISDQ0NEEKguLgYqampeOaZZ2BoaIjMzMwnjvntyueeiIi6KJ1ONkFdUmumOqqoqBCvvPKK6Natm7CwsBDPPvusWLVqlQAgXFxcxA8//CDi4+OFmZmZcgqu30711dDQIDZu3Cj69OkjjI2Nha2trZg0aZK4cOGCynHCwsKEsbGxcHZ2FkZGRsLKykpMnDhR5Ofnq/RrabySkhIxcuRIYWpqKtzd3cXrr78uli5dKgAILy8v5RRe33//vXBzcxNmZmbi2WefVZmy60nGjx8v3N3dhYWFhTAxMRGenp4iJCRE5OTkqPQ7cOCAsLS0FOvXr39krH379olBgwYJc3NzIZPJhIGBgQAgJEkSNjY24umnnxaxsbGipKREZT99PffgVEtq41RnpCl8/5GW7JGEEEJ3pTd1RcHBwQB+HfvXkSxYsABpaWkoKSnRdSp6pzOee0mSkJqayvGraujI73/qXPj+Iy1J47AH0jv19fW6TkFv8dwTEZGusfglamc///wzJEl64r+QkBBdp0rU7g4dOoTo6GhkZGTAw8ND+X548cUXm/QNCgqCpaUlDA0NMWDAAHz//fc6yFg9O3fuhL+/PywtLeHm5oa5c+fi5s2bHSaeQqFAXFwcvLy8IJPJYGNjA19fX1y5cuWR+9TU1MDb2xsrVqxQtu3btw/x8fH8wEsdEotf0hvLly9HSkoK7t69C3d3d6Snp+skD29vbwghnvhv9+7dOslPGzrKuaeObfXq1di0aROWL1+OKVOm4NKlS/D09ES3bt2wY8cO7N+/X6X/F198gbS0NIwbNw65ubkYPHiwjjJvmdTUVMyaNQvBwcEoLCxEVlYWjhw5gtGjR6Ourk7n8YAH04v94x//wCeffIKqqir89NNP8PT0RGVl5SP3iYmJwYULF1Taxo8fD1NTU4waNUo5fzdRR8Hil/RGXFwc7t+/DyEELl++jKlTp+o6Jb3Bc9861dXVCAwM7PTHaIl33nkHu3fvxp49e2BpaamybdOmTTAwMEBYWBju3r2rowzb7sMPP0TPnj2xdOlSWFtbw8/PD5GRkTh79myrVi/UdLzdu3cjMzMTaWlp+P3vfw8jIyM4OTkhKysLvr6+ze7zzTff4Mcff2x22xtvvIHf/e53GDNmTKuLcSJtYPFLRNRBffTRRygqKur0x3iSixcvYuXKlVi7dq1y0ZvfCgwMREREBH755Re8+eabOshQM65duwYnJyeVpZ8b5+pubpq+9o73/vvvY/DgwRg4cGCL+ldXV2Pp0qWPXWVyzZo1OHv2rFZWoiRqLRa/REQaIoRAQkIC+vfvDxMTE9ja2mLixIn4+eeflX3Cw8Mhk8nQo0cPZdvixYshl8shSRJu374N4MGyylFRUcjPz4ckSfDy8sKmTZtgamoKBwcHLFiwAE5OTjA1NUVgYKDKN31tOQbwYIU/deaMbqtNmzZBCIHx48c/ss/69evRt29fbN++HYcOHXpsvJZch+TkZMjlcpibmyMrKwujR4+GlZUVXFxcsGvXLpV49fX1WLVqFVxdXWFmZoZBgwYhNTVV7dfp4eHR5ING4/hcDw8Pncarra3FiRMn4Ofn1+J9YmJisHjx4scuTW9ra4vhw4cjKSkJnFyKOox2n12NujzO80ldBdScZ3TVqlVCJpOJjz/+WJSVlYlz586JwYMHi+7du6vMLzxr1izh6Oiosu/GjRsFAFFcXKxsmzJlivD09FTpFxYWJuRyuTh//ryoqakRubm5wt/fX1haWirnNm7rMT777DNhaWkpYmNjW/zaG7Xm/e/h4SF8fHya3ebp6SkuX74shBDim2++EQYGBqJ3796isrJSCCHEwYMHxYQJE1T2ael1iImJEQDE4cOHxd27d0VRUZEYNmyYkMvlora2VtnvzTffFCYmJiI9PV2UlpaK5cuXCwMDA3Hq1Cm1Xmd2drYwNjYWmzZtEuXl5eLHH38U/fv3F3/84x/ViqONeJcvXxYAhJ+fnxgxYoTo0aOHMDExEd7e3uK9994TDQ0NKv2PHTsmxo8fL4QQori4WAAQMTExzcaOjo4WAMSZM2fUyknd9x9RC+3hN79ERBpQXV2NhIQETJ48GbNnz4a1tTUGDhyIDz74ALdv327xioUtYWRkpPxW08fHB8nJyaioqEBKSopG4o8dOxbl5eVYuXKlRuI9zr1793D58mV4eno+sW9AQACWLFmCK1eu4K233mq2T2uuQ2BgIKysrGBvb4+QkBDcu3cPBQUFAB7MZJCcnIxJkyZhypQpsLGxwYoVK2BsbKz2+R4+fDiWLVuG8PBwWFlZwdfXFxUVFdi+fbtacbQRr/GBNnt7e2zYsAG5ubm4desWJk6ciNdeew07d+5U9q2urkZERASSk5NbFLtPnz4AgJycHLXzItIGFr9ERBqQm5uLyspKDBkyRKXd398fMpmsVQ8gtdSQIUNgbm6u8rN+Z1FUVAQhBMzNzVvUf/369ejXrx+2bNmCY8eONdne1usgk8kAQLkM9oULF1BVVaXywJeZmRl69Oih9vmOiYnB1q1bcfjwYVRWVuLSpUsIDAxEQEAArl27plYsTcczMTEBAAwYMACBgYGws7ODtbU11q5dC2tra5UPDcuXL8f8+fPh7OzcotiN1/bWrVtq5USkLSx+iYg0oHE6JwsLiybbbGxsUFFRodXjm5iYoLi4WKvH0IaamhoAvxZfT2JqaoqUlBRIkoR58+ahurpaZbumr8O9e/cAACtWrFCZh/vq1auoqqpqcZwbN24gPj4e8+fPx3PPPQe5XA53d3ds27YN169fx8aNG9XKS9PxnJycAEA5HryRTCaDm5sb8vPzAQDHjh1DTk4OXnnllRbHNjMzA/DrtSbSNRa/REQaYGNjAwDNFldlZWVwcXHR2rEVCoXWj6EtjYWROoshBAQEIDIyEnl5eVi3bp3KNk1fh8aHuRITE5vMxX38+PEWx8nLy0N9fT169uyp0m5lZQU7Ozvk5uaqlZem41lYWKBPnz44f/58k211dXWwtrYG8GB2kMOHD8PAwED5QaDxHG3YsAGSJOH06dMq+9fW1gL49VoT6RqLXyIiDfD19YWFhUWT//GfPHkStbW1eOqpp5RtRkZGyp/VNSE7OxtCCAwdOlRrx9AWBwcHSJKk9vy969atg7e3N86cOaPSrs51aIlevXrB1NQUZ8+eVWu/hzUW3Tdu3FBpr6iowJ07d5RTlOkqHvBggYszZ87g0qVLyraqqipcvXpVOf1ZSkpKkw8Bjb84xMTEQAjRZMhJ47V1dHRUOycibWDxS0SkAaampoiKisLevXuxY8cOlJeXIycnBwsXLoSTkxPCwsKUfb28vHDnzh1kZmZCoVCguLi42XlZ7ezscP36dVy5cgUVFRXKYrahoQGlpaWoq6vDuXPnEBERAVdXV4SGhmrkGAcPHmy3qc7Mzc3h4eGBwsJCtfZrHP5gaGjYpL2l16Glx5k7dy527dqF5ORklJeXo76+HoWFhcrCMyQkBI6Ojo9dXtnd3R0jR47Etm3bcOTIEVRXV+PatWvKfF5++WVlX13EA4DIyEi4ubkhNDQUBQUFKCkpwbJly1BdXf3IBwxbovHatnT+YCJtY/FLRKQhq1evRlxcHGJjY9G9e3cMHz4cvXv3RnZ2NuRyubLfokWLMHLkSMyYMQP9+vXDunXrlD8J//ZhpYULF8LBwQE+Pj4YM2YM7ty5A+DB2MmBAwfCzMwMw4YNQ9++ffHVV1+pjJtt6zHa09ixY5Gbm6syfvfTTz+Fl5cX8vPz4e/vj9dff73JfkOHDkVkZGST9pZch+TkZCQmJgIABg0ahEuXLmHbtm2IiooCALzwwgvIy8sDACQlJWHJkiWIj49Ht27d4OTkhIiICJSWlgJ48LN+UVERsrKyHvkaJUlCWloaQkJC8PLLL8PW1hY+Pj4oKChARkYGhg0bpuyri3jAgzl5jx49ChcXF/j5+cHZ2Rnffvst9u/fr9b8vw87deoUnJ2dMWjQoFbHINIkSQjOOk2aFRwcDABIS0vTcSZEbSNJElJTUzFt2jRdp6K0YMECpKWloaSkRNepNKs17/+LFy+if//+SElJwezZs7WVmtY0NDRgxIgRCA0Nxbx587p8PHWUlJTAxcUF69evV36waKmO+P6jLiGN3/wSEXUy6jwc1hl4eXkhNjYWsbGxyvlmO4v6+npkZmaioqICISEhXT6eutasWQM/Pz+Eh4e3+7GJHoXFLxER6Vx0dDSCg4MREhKi9sNvupSdnY2MjAwcPHiwxXMVd+Z46khISMDZs2dx4MABGBsbt+uxiR6HxS8RUSexfPlypKSk4O7du3B3d0d6erquU9KoDRs2IDw8HG+//bauU2mxUaNG4ZNPPkGPHj30Il5LZWVl4f79+8jOzoatrW27HpvoSYx0nQAREbVMXFwc4uLidJ2GVgUFBSEoKEjXaVAbTZgwARMmTNB1GkTN4je/RERERKQ3WPwSERERkd5g8UtEREREeoPFLxERERHpDT7wRlpRWFiIPXv26DoNojY7fvy4rlPoVBqXsuX7n4g6Kq7wRhoXHBzc5aZgIiKi9scV3kgL0lj8EhG1sz179mD69Ongn18ionbH5Y2JiIiISH+w+CUiIiIivcHil4iIiIj0BotfIiIiItIbLH6JiIiISG+w+CUiIiIivcHil4iIiIj0BotfIiIiItIbLH6JiIiISG+w+CUiIiIivcHil4iIiIj0BotfIiIiItIbLH6JiIiISG+w+CUiIiIivcHil4iIiIj0BotfIiIiItIbLH6JiIiISG+w+CUiIiIivcHil4iIiIj0BotfIiIiItIbLH6JiIiISG+w+CUiIiIivcHil4iIiIj0BotfIiIiItIbLH6JiIiISG+w+CUiIiIivcHil4iIiIj0BotfIiIiItIbLH6JiIiISG+w+CUiIiIivcHil4iIiIj0BotfIiIiItIbLH6JiIiISG8Y6ToBIqKurLCwEHPmzEF9fb2yrbS0FJaWlhgxYoRK3379+uHDDz9s5wyJiPQLi18iIi1ycXHB1atXkZ+f32Tbf/7zH5X//n//7/+1V1pERHqLwx6IiLTspZdegrGx8RP7hYSEtEM2RET6jcUvEZGWzZo1C3V1dY/tM2DAAPj4+LRTRkRE+ovFLxGRlnl6emLQoEGQJKnZ7cbGxpgzZ047Z0VEpJ9Y/BIRtYOXXnoJhoaGzW6rq6tDcHBwO2dERKSfWPwSEbWDGTNmoKGhoUm7gYEBhg4dit69e7d/UkREeojFLxFRO3BycsIzzzwDAwPVP7sGBgZ46aWXdJQVEZH+YfFLRNROXnzxxSZtQghMnjxZB9kQEeknFr9ERO1k6tSpKuN+DQ0N8fzzz8PBwUGHWRER6RcWv0RE7cTW1hZ/+MMflAWwEAKzZ8/WcVZERPqFxS8RUTuaPXu28sE3Y2NjTJw4UccZERHpFxa/RETtaPz48TAxMQEAjBs3DhYWFjrOiIhIv7D4JSJqR3K5XPltL4c8EBG1P0kIIXSdBHVuwcHBSE9P13UaRETUxaWmpmLatGm6ToM6tzQjXWdAXcPQoUOxZMkSXadB1CrHjx9HUlISUlNT2+V49fX1SE1NxcyZM9vleNoyffp0REREICAgQNepkB6YPn26rlOgLoLFL2mEi4sLP41Tp5aUlNSu9/CkSZNgamrabsfThunTpyMgIIDvfWoXLH5JUzjml4hIBzp74UtE1Fmx+CUiIiIivcHil4iIiIj0BotfIiIiItIbLH6JiIiISG+w+CUi0pADBw7A2toa//znP3WdSqd06NAhREdHIyMjAx4eHpAkCZIk4cUXX2zSNygoCJaWljA0NMSAAQPw/fff6yBj9ezcuRP+/v6wtLSEm5sb5s6di5s3b3aYeAqFAnFxcfDy8oJMJoONjQ18fX1x5cqVR+5TU1MDb29vrFixQtm2b98+xMfHo76+vtW5EGkTi18iIg35/+zda1hTV9o38H+AkEBIOIgCgpzxVK2MytMay1DqjNY6HhAPtLZzaQ9Sa4sIUhTUWkSqxQt4sNBWdJgZaRVUijMq1lcd6jg6to5SEUeLeMIjIAIBgpzu94MPqRHQBAJBcv+uiw9de+217uy9ijc7a6/FewZ13ieffILk5GRERUUhMDAQly9fhoeHB/r164eMjAzs27dPrf7Bgwexc+dOTJ06FYWFhRg9erSeItdMZmYm5s2bh9mzZ+PGjRvYs2cPjh49ismTJ6OpqUnv7QEPlxL761//im+++QZ1dXX473//Cw8PD9TU1HR4TnR0NC5evKhWNm3aNIjFYkyYMAGVlZWdioWx7sTJL2OM6ciUKVNQVVWFqVOn6jsUKJVKyOVyfYehkfXr12PHjh3IysqCVCpVO5acnAwjIyMEBwejqqpKTxF23ddff42BAwciIiIClpaW8Pb2RlhYGPLz83Hy5Em9t7djxw7k5ORg586deOGFF2BiYgIHBwfs2bMHI0aMaPec48eP49y5c+0eW7JkCUaNGoXXXnut08k4Y92Fk1/GGOuDtm7ditLSUn2H8VSXLl3CqlWr8Omnn7a79rFcLkdoaChu3ryJZcuW6SFC3SgpKYGDgwMEAoGqbNCgQQCAa9eu6b29L7/8EqNHj8bIkSM1qq9UKhEREYGkpKQO66xZswb5+flPrMOYPnDyyxhjOnDs2DE4OztDIBDgiy++AACkpqZCIpHA3Nwce/bsweTJkyGTyeDk5ITt27erzk1OToZYLMaAAQPw/vvvw8HBAWKxGHK5XO0pXkhICExNTWFvb68qW7x4MSQSCQQCAcrLywEAoaGhCA8PR3FxMQQCATw9PQEABw4cgEwmw7p163rikmgkOTkZRIRp06Z1WCc2NhaDBw/Gli1bcOjQoSe2R0RISEjAsGHDIBKJYG1tjRkzZuDChQuqOpreF+DhVtSrV6+Gs7MzzMzM8Pzzz3dqG2x3d/c2f4y0zs91d3fXa3sNDQ3497//DW9vb43PiY6OxuLFi9G/f/8O61hbW8PPzw9JSUk8JYj1Kpz8MsaYDrz00ks4fvy4WtkHH3yApUuXQqlUQiqVIjMzE8XFxXB3d8d7772HxsZGAA+T2vnz56Ourg5LlizB1atXcfr0aTQ1NeH3v/89SkpKADxMFB/fSjglJQWffvqpWllSUhKmTp0KDw8PEBEuXboEAKoXkFpaWrrlGnTGvn37MGTIEJibm3dYx8zMDH/+859hZGSE9957D7W1tR3WXbNmDVasWIHo6GiUlpbi6NGjKCkpga+vL+7evQtA8/sCAMuXL8fnn3+OxMRE3L59G1OnTsUbb7yBU6dOafU5o6KicOfOHWzatAkKhQKFhYVISkrCpEmT8OKLL2rVlq7bu3XrFhoaGvCf//wH/v7+qj++hg0bhpSUlDaJ67/+9S8UFxfjjTfeeGrbv/nNb3Dz5k38/PPPWsXEWHfi5JcxxnqAXC6HTCZD//79ERQUhNraWly/fl2tjomJieqJ5fDhw5GamgqFQoH09HSdxDBlyhRUV1dj1apVOmmvq2pra3HlyhV4eHg8te64ceOwdOlSXL16FcuXL2+3jlKpREJCAmbOnIk333wTlpaWGDlyJL766iuUl5dj8+bNbc550n2pr69HamoqAgICEBgYCCsrK6xcuRJCoVDre+Ln54fIyEiEhIRAJpNhxIgRUCgU2LJli1btdEd7rS+09e/fH+vWrUNhYSHu3r2LGTNm4MMPP8S3336rqqtUKhEaGorU1FSN2vby8gIAFBQUaB0XY92Fk1/GGOthpqamAKD2hLE9Y8eOhbm5udpX9n1JaWkpiOiJT30fFRsbiyFDhiAlJQXHjh1rc7ywsBA1NTUYO3asWrmPjw9MTU2f+iLY4/fl4sWLqKurU3vhy8zMDPb29lrfk+joaGzevBmHDx9GTU0NLl++DLlcjnHjxqme7OurPZFIBAB47rnnIJfLYWNjA0tLS3z66aewtLRU+6MhKioKCxcuhKOjo0Ztt97b1qfujPUGnPwyxlgvJhKJUFZWpu8wukV9fT2AX5OvpxGLxUhPT4dAIMDbb78NpVKpdrx1WS0LC4s251pZWUGhUGgVX+v0ipUrV6rWHBYIBLh27Rrq6uo0buf27dvYsGEDFi5ciFdeeQUSiQRubm5IS0vDrVu3EB8fr1Vcum7PwcEBAFRzxluZmprCxcUFxcXFAB7Oay8oKMC7776rcdtmZmYAfr3XjPUGnPwyxlgv1djYiMrKSjg5Oek7lG7RmhhpsxnCuHHjEBYWhqKiIqxdu1btmJWVFQC0m+R25jq2vsyVmJgIIlL7OXHihMbtFBUVobm5GQMHDlQrl8lksLGxQWFhoVZx6bo9CwsLeHl54fz5822ONTU1wdLSEsDDFUQOHz4MIyMj1R8Crddo3bp1EAgEbeZCNzQ0APj1XjPWG3DyyxhjvVReXh6ISO0FJhMTk6dOl3hWDBgwAAKBQOv1e9euXYuhQ4fizJkzauUjRoyAhYVFmwTs5MmTaGhowJgxY7TqZ9CgQRCLxcjPz9fqvMe1Jt23b99WK1coFKioqFAtUaav9oCHG1ycOXMGly9fVpXV1dXh2rVrquXP0tPT2/wR0PqtRHR0NIiozZST1ntrZ2endUyMdRdOfhljrJdoaWnB/fv30dTUhLNnzyI0NBTOzs6YP3++qo6npycqKiqQk5ODxsZGlJWVtbuuq42NDW7duoWrV69CoVCgsbERubm5vWqpM3Nzc7i7u+PGjRtandc6/cHY2LhNeXh4OLKzs5GRkYHq6moUFBRg0aJFcHBwQHBwsNb9LFiwANu3b0dqaiqqq6vR3NyMGzduqBLPoKAg2NnZPXF7ZTc3N/j7+yMtLQ1Hjx6FUqlESUmJKp533nlHVVcf7QFAWFgYXFxcMH/+fFy/fh337t1DZGQso7h0AAAgAElEQVQklEplhy8YaqL13mq6fjBjPYGTX8YY04EvvvgCPj4+AIDIyEhMnz4dqampSExMBAA8//zzuHz5MtLS0hAeHg4AePXVV1FUVKRqo76+HiNHjoSZmRl8fX0xePBg/OMf/1CbE/vBBx/A398fr7/+OoYMGYK1a9eqvlJ+9GWnRYsWYcCAARg+fDhee+01VFRU9Mh10NaUKVNQWFioNn/3u+++g6enJ4qLi+Hj44OPPvqozXkvvvgiwsLC2pR/8skniIuLQ0xMDGxtbeHn5wdXV1fk5eVBIpEAgFb3JSkpCUuXLsWGDRvQr18/ODg4IDQ0FPfv3wfw8Gv90tJS7Nmzp8PPKBAIsHPnTgQFBeGdd96BtbU1hg8fjuvXr2P37t3w9fVV1dVHe8DDNXn/+c9/wsnJCd7e3nB0dMSPP/6Iffv2abX+7+N++uknODo64vnnn+90G4zpHDHWRbNmzaJZs2bpOwzGOi0zM5P0/eswODiYbGxs9BqDtgBQZmZml9ooKioiExMT2rZtm46i6lnNzc3k6+tLW7duNYj2tFFeXk5isZg2btyok/Z0Md4YI6IsfvLLGGO9hDYvfvUVnp6eiImJQUxMjGq92WdFc3MzcnJyoFAoEBQU1Ofb09aaNWvg7e2NkJCQHu+bsSfh5JexR8TExGD48OGQyWQQiUTw9PTExx9//NR/lN99911IpVIIBIIuvxwDPPz6e+jQoVi5cmWX2rl48SI++ugjPPfcc5BKpTAxMYGlpSUGDx6MKVOmaPXGenfR5Jrv3r0b7u7uastNCQQCmJqaYsCAAXj55ZcRHx+v+iqaPVtWrFiB2bNnIygoSOuX3/QpLy8Pu3fvRm5ursZrFT/L7WkjISEB+fn52L9/P4RCYY/2zdhT6fvZM3v29aVpD35+fpSSkkL37t2j6upqyszMJKFQSK+++upTz92+fTsBoDNnznQ5jrCwMAJA0dHRnW5jy5YtJBQK6be//S0dOHCA7t+/T/X19VRcXEw7duwguVxOX3/9dZdj7SptrrmHhwdZWloSEVFLSwvdv3+f/vGPf9D8+fNJIBCQg4MD/fTTT1rHoO9pDytWrCBTU1MCQK6urrRz5069xaIN6Phr6O+//54iIyN11h7Tj5ycHIqLi6Ompiadtqvr8cYMVpaJPhNvxnobCwsLBAcHq94inzNnDnbv3o2srCyUlJR0agkhbR0/fhznzp3rUhv//ve/ERwcDD8/P3z//fcwMfn1f3V3d3e4u7vDyspK7WUrfensNRcIBLCyssLLL7+Ml19+GVOmTMHcuXMxZcoU/PLLL6q1SZ8FcXFxiIuL03cYejdx4kRMnDhR32GwLpo+fTqmT5+u7zAY6xBPe2DsEXv37m2zfJKtrS0APHVHJ4FA0OX+lUolIiIikJSU1KV2YmNj0dzcjM8++0wt8X3UpEmT8OGHH3apH13oyjV/1KxZszB//nyUlpbiq6++0mmMjDHG+g5OfpnebNu2DWPHjoVYLIZEIoGrq6tqxyYiQkJCAoYNGwaRSARra2vMmDEDFy5cUJ2fmpoKiUQCc3Nz7NmzB5MnT4ZMJoOTkxO2b9+uqjds2DAIBAIYGRlhzJgxqoTq448/hqWlJcRiMf785z93GOfNmzdhZmYGNzc3VRkRIT4+HkOGDIFIJIKlpSUiIiK6fE2io6OxePFi1a5Jjztw4MBT12ltaGjA4cOH0a9fP/zP//yPxn339muuidb1cHNzc7U6jzHGmOHg5JfpRVJSEv74xz9i1qxZuHXrFm7cuIGoqChcvHgRwMO3hFesWIHo6GiUlpbi6NGjKCkpga+vL+7evQvg4XqnS5cuhVKphFQqRWZmJoqLi+Hu7o733ntPtQvWuXPn4OrqikGDBuHHH39Uvfjx+eef45133sH69evVNhF4VF1dHY4cOYL33nsPpqamqvJVq1YhMjISwcHBuHv3Lu7cudOlheAB4F//+heKi4vxxhtvdFindTWAlpaWDutcu3YN9fX18PLy0qr/3n7NNdG6Humju1QxxhhjavQ86Zj1Adq+8NbQ0EBWVlbk7++vVt7U1ERJSUlUV1dHFhYWFBQUpHb8xx9/JAAUExOjKouOjiYApFQqVWUpKSkEgC5duqQqS0xMJACUlZWlKqutrSVnZ2eqqqrqMNbo6GgaPHgwVVdXq8rq6urI3Nycfv/736vV7coLb3V1dTR27Fi6ceMGERGVlZV1+oW3U6dOEQD63e9+p1X/vfmat3r0hbeOCAQCsrKyemKdx+n7hbdnFfgFJNaDeLwxHeEX3ljPO3v2LCorKzFp0iS1cmNjYyxZsgSnTp1CTU1Nmz3ifXx8YGpqipMnTz6x/danha1PIYGHS5GtWbMGSUlJmD17NgAgIyMDM2bMgEwma7ed7OxsZGVl4eDBg5BKparyS5cuoa6uDhMmTND8Qz9FVFQUFi5cCEdHxy63ZWFhAUC7+bKFhYW9+pprqra2FkTUYftPk5WV1anzDFlvWC6PMca0wckv63HV1dUAACsrq3aPV1ZWAvg1iXuUlZUVFAqF1n1aWFhg4cKFiI+Px48//oj/+Z//wZdffoldu3a1W3/Hjh1ISEhAXl4eBg4cqHasda/6jublauvYsWMoKChAQkKCTtpzdXWFWCzGL7/8ovE5vf2aa6r1Mw8dOrRT58+dO7dT5xmypKSkLr+gyRhjPYnn/LIe15rYlJeXt3u8NSluL+GqrKyEk5NTp/oNCQmBUChEYmIijh49ikGDBsHDw6NNvU2bNiEjIwNHjhxpNwkTi8UAgAcPHnQqjsdt3boVhw8fhpGRkWrzhtbEet26dRAIBDh16pTG7YlEIkyaNAnl5eX417/+1WG9iooKvPvuuwB6/zXX1IEDBwAAkydP7tT5RMQ/WvwAQGZmpt7j4B/D+GFMVzj5ZT3O1dUVNjY2OHjwYLvHR4wYAQsLizYJ38mTJ9HQ0IAxY8Z0ql8nJyfMmTMHu3btwqpVqxAaGqp2nIgQGRmJgoIC5OTktPsUtDU+IyMj/PDDD52K43Hp6eltfsmXlZUBeLj6AxG1mY7wNGvWrIFIJEJYWBiUSmW7dc6dO6daBq23X3NN3LlzB4mJiXBycsLbb7/d6XYYY4z1bZz8sh4nEokQFRWFo0ePIiQkBDdv3kRLSwsUCgXOnz8PsViM8PBwZGdnIyMjA9XV1SgoKMCiRYvg4OCA4ODgTvcdHh6OpqYm3L9/H6+88orasfPnz+Pzzz9HWloahEJhm610N27cCODhdIfAwEDs2rULW7duRXV1Nc6ePYvNmzd36bpoIjc396lLnQEPVz345ptvcO7cOfj6+mL//v2oqqpCY2Mjrly5grS0NLzzzjuqbUd7+zV/FBGhpqYGLS0tqj8UMjMzMX78eBgbGyMnJ6fTc34ZY4z1fZz8Mr0IDw/HF198gby8PHh6ekIikcDPzw95eXkAgE8++QRxcXGIiYmBra0t/Pz84Orqiry8PEgkEgAP15xNTEwEADz//PO4fPky0tLSEB4eDgB49dVX2+xg9pvf/Ab+/v5YsmRJm5i0+VrtT3/6ExYsWIDIyEg4Ojpi8eLF8PX1BQBMnToVZ8+e1fqa6FpgYCAuXLgAf39/LF++HE5OTjAzM8Po0aPx5Zdfws/PD6+//rqqfm++5n//+98xatQo3L59G/X19bC0tISxsTGMjY0xePBgJCQkYP78+SgsLOz0U2rGGGOGQUA8kYZ1Ueub/Dt37tRzJIx1TlZWFubOncvzCrUkEAiQmZmJOXPm6DsUZgB4vDEd2clPfhljjDHGmMHg5JexbnDhwoU281fb+wkKCtJ3qIwxxphB4eSXsW4wdOhQjZbu2bFjh75DZUwvDh06hBUrVmD37t1wd3dX/UH41ltvtak7ceJESKVSGBsb47nnnsPp06f1ELF2vv32W/j4+EAqlcLFxQULFizAnTt3ek17wMNt0hMTEyGXy9s9HhMTg+HDh0Mmk0EkEsHT0xMff/wxampqtI7vb3/7GzZs2KDaop0xfeLklzHGWI/65JNPkJycjKioKAQGBuLy5cvw8PBAv379kJGRgX379qnVP3jwIHbu3ImpU6eisLAQo0eP1lPkmsnMzMS8efMwe/Zs3LhxA3v27MHRo0cxefJkNDU16b09ACgqKsJvf/tbhIWFdbgb5JEjR/Dhhx/i6tWrKC8vR1xcnNqOjdrEN23aNIjFYkyYMEG1qQ5j+sLJL2OM9QJKpbLDJ3DPUh9Ps379euzYsQNZWVlttrBOTk6GkZERgoODUVVVpacIu+7rr7/GwIEDERERAUtLS3h7eyMsLAz5+flP3Sq8J9r7+eefsXz5cixatAje3t4d1rOwsEBwcDBsbGwglUoxZ84cBAQE4MCBAygpKdE6viVLlmDUqFF47bXXOp20M6YLnPwyxlgvsHXrVpSWlj7zfTzJpUuXsGrVKnz66aeqnRIfJZfLERoaips3b2LZsmV6iFA3SkpK4ODgAIFAoCobNGgQAODatWt6b2/UqFHYvXs35s2bB5FI1GG9vXv3wtjYWK3M1tYWANSeFmsT35o1a5Cfn89bYjO94uSXMcY6gYiQkJCAYcOGQSQSwdraGjNmzMCFCxdUdUJCQmBqagp7e3tV2eLFiyGRSCAQCFRbfIeGhiI8PBzFxcUQCATw9PREcnIyxGIxBgwYgPfffx8ODg4Qi8WQy+VqT9O60gfwcEtoTTZO0YXk5GQQEaZNm9ZhndjYWAwePBhbtmzBoUOHntieJvcgNTUVEokE5ubm2LNnDyZPngyZTAYnJyds375drb3m5masXr0azs7OMDMzw/PPP4/MzEytP6e7u3ubPzJa57+6u7vrvb2uuHnzJszMzODm5tap+KytreHn54ekpCReWpDpDzHWRbNmzaJZs2bpOwzGOi0zM5O0/XW4evVqMjU1pW3btlFlZSWdPXuWRo8eTba2tnTnzh1VvXnz5pGdnZ3aufHx8QSAysrKVGWBgYHk4eGhVi84OJgkEgmdP3+e6uvrqbCwkHx8fEgqldL169d10sfevXtJKpVSTEyMVp+fiAgAZWZmalzf3d2dhg8f3u4xDw8PunLlChERHT9+nIyMjMjV1ZVqamqIiCg3N5emT5+udo6m9yA6OpoA0OHDh6mqqopKS0vJ19eXJBIJNTQ0qOotW7aMRCIR7dq1i+7fv09RUVFkZGREP/30k8afkYgoLy+PhEIhJScnU3V1NZ07d46GDRtGkyZN0qqd7mrvUS+88AKNGjVKo7q1tbUklUopJCSkS/GtWLGCANCZM2e0ilXb8cZYB7L4yS9jjGlJqVQiISEBM2fOxJtvvglLS0uMHDkSX331FcrLy3W61bWJiYnqyebw4cORmpoKhUKB9PR0nbQ/ZcoUVFdXY9WqVTppryO1tbW4cuUKPDw8nlp33LhxWLp0Ka5evYrly5e3W6cz90Aul0Mmk6F///4ICgpCbW0trl+/DgCor69HamoqAgICEBgYCCsrK6xcuRJCoVDra+3n54fIyEiEhIRAJpNhxIgRUCgU2LJli1btdFd7nRUXFwcHBwfExsZ2KT4vLy8AQEFBQbfHzFh7OPlljDEtFRYWoqamBmPHjlUr9/HxgampaadeQtLU2LFjYW5urvbV/rOgtLQURARzc3ON6sfGxmLIkCFISUnBsWPH2hzv6j0wNTUFADQ2NgIALl68iLq6OowYMUJVx8zMDPb29lpf6+joaGzevBmHDx9GTU0NLl++DLlcjnHjxqm9KKav9jojOzsbWVlZ+P7779u8qKhtfK1j4O7duz0SO2OP4+SXMca01LpUk4WFRZtjVlZWUCgU3dq/SCRCWVlZt/aha/X19QDwxBesHiUWi5Geng6BQIC3334bSqVS7biu70FtbS0AYOXKlWob0Vy7dq3DpcDac/v2bWzYsAELFy7EK6+8AolEAjc3N6SlpeHWrVuIj4/XKi5dt9cZO3bswPr165GXlwdXV9cux2dmZgbg1zHBWE/j5JcxxrRkZWUFAO0mWJWVlXBycuq2vhsbG7u9j+7QmvBos8nBuHHjEBYWhqKiIqxdu1btmK7vQf/+/QEAiYmJbTajOXHihMbtFBUVobm5GQMHDlQrl8lksLGxQWFhoVZx6bo9bW3atAkZGRk4cuRImxg6G19DQwOAX8cEYz2Nk1/GGNPSiBEjYGFhgVOnTqmVnzx5Eg0NDRgzZoyqzMTERPXVui7k5eWBiPDiiy92Wx/dYcCAARAIBFqv37t27VoMHToUZ86cUSvX5h5oYtCgQRCLxcjPz9fqvMe1Jt23b99WK1coFKioqFAtAaav9jRFRIiMjERBQQFycnLafcLe2fhax4CdnZ2Oo2ZMM5z8MsaYlsRiMcLDw5GdnY2MjAxUV1ejoKAAixYtgoODA4KDg1V1PT09UVFRgZycHDQ2NqKsrKzdtVltbGxw69YtXL16FQqFQpXMtrS04P79+2hqasLZs2cRGhoKZ2dnzJ8/Xyd95Obm9shSZ+bm5nB3d8eNGze0Oq91+sPj681qcw807WfBggXYvn07UlNTUV1djebmZty4cUOV2AUFBcHOzu6J2yu7ubnB398faWlpOHr0KJRKJUpKSlTxvPPOO6q6+mhPU+fPn8fnn3+OtLQ0CIVCtakgAoEAGzdu1Dq+Vq1jYOTIkV2Ok7FO0eNSE6yP4KXO2LOuM0udtbS0UHx8PHl5eZFQKCRra2sKCAigixcvqtW7d+8e+fv7k1gsJjc3N/roo48oIiKCAJCnp6dqybLTp0+Ti4sLmZmZ0UsvvUR37tyh4OBgEgqF5OjoSCYmJiSTyWjGjBlUXFyssz72799PUqmUYmNjtb5u0HLpqZCQEBIKhVRXV6cqy87OJg8PDwJAtra29OGHH7Z7bkRERJulzjS5BykpKWRubk4AyMvLi4qLi2nz5s0kk8kIALm4uNAvv/xCREQPHjygyMhIcnZ2JhMTE+rfvz8FBgZSYWEhEREFBAQQAFq9evUTP2d5eTmFhoaSp6cniUQisrCwoPHjx9N3332nVk9f7Z04cYLGjx9PDg4OBIAAkL29Pcnlcvrhhx+IiKigoEB1rL2f+Ph4reNrNWXKFHJ0dKSWlpYnxvk4bccbYx3I4uSXdRknv+xZ15nktycEBweTjY2NvsPokLbJSFFREZmYmNC2bdu6Maru09zcTL6+vrR161aDaK87lJeXk1gspo0bN2p9Lie/TEd4nV/GGOvNtHlBrLfz9PRETEwMYmJiUFNTo+9wtNLc3IycnBwoFAoEBQX1+fa6y5o1a+Dt7Y2QkBB9h8IMGCe/jDHGesyKFSswe/ZsBAUFaf3ymz7l5eVh9+7dyM3N1Xit4me5ve6QkJCA/Px87N+/H0KhUN/hMAPGyS9jjPVCUVFRSE9PR1VVFdzc3LBr1y59h6Qz69atQ0hICD777DN9h6KxCRMm4JtvvoG9vb1BtKdre/bswYMHD5CXlwdra2t9h8MMnIm+A2CMMdZWXFwc4uLi9B1Gt5k4cSImTpyo7zBYD5k+fTqmT5+u7zAYA8BPfhljjDHGmAHh5JcxxhhjjBkMTn4ZY4wxxpjB4OSXMcYYY4wZDH7hjenEv//9b8yePVvfYTDWKa3brfIY1l5iYiJ27typ7zAYY0xjnPyyLhs3bpy+Q2CsS5ycnDBr1qwe6+/OnTs4c+YMJk+e3GN9doeevGaMzZo1C4MGDdJ3GKwPEBAR6TsIxhgzJFlZWZg7dy741y9jjPW4nTznlzHGGGOMGQxOfhljjDHGmMHg5JcxxhhjjBkMTn4ZY4wxxpjB4OSXMcYYY4wZDE5+GWOMMcaYweDklzHGGGOMGQxOfhljjDHGmMHg5JcxxhhjjBkMTn4ZY4wxxpjB4OSXMcYYY4wZDE5+GWOMMcaYweDklzHGGGOMGQxOfhljjDHGmMHg5JcxxhhjjBkMTn4ZY4wxxpjB4OSXMcYYY4wZDE5+GWOMMcaYweDklzHGGGOMGQxOfhljjDHGmMHg5JcxxhhjjBkMTn4ZY4wxxpjB4OSXMcYYY4wZDE5+GWOMMcaYweDklzHGGGOMGQxOfhljjDHGmMHg5JcxxhhjjBkMTn4ZY4wxxpjB4OSXMcYYY4wZDE5+GWOMMcaYweDklzHGGGOMGQxOfhljjDHGmMEw0XcAjDHWlzU2NqKmpkatrLa2FgBw//59tXKBQAArK6sei40xxgwRJ7+MMdaNKioq4OjoiObm5jbHbGxs1P7b398fR44c6anQGGPMIPG0B8YY60Z2dnb47W9/CyOjJ/+6FQgEeP3113soKsYYM1yc/DLGWDd76623nlrH2NgYM2fO7IFoGGPMsHHyyxhj3SwwMBAmJh3PMjM2Nsarr76Kfv369WBUjDFmmDj5ZYyxbiaTyTB58uQOE2AiwptvvtnDUTHGmGHi5JcxxnrAm2++2e5LbwBgamqKP/zhDz0cEWOMGSZOfhljrAf84Q9/gLm5eZtyoVCIgIAASCQSPUTFGGOGh5NfxhjrAWKxGDNnzoRQKFQrb2xsxLx58/QUFWOMGR5OfhljrIe88cYbaGxsVCuTyWT4/e9/r6eIGGPM8HDyyxhjPeR3v/ud2sYWQqEQr7/+OkxNTfUYFWOMGRZOfhljrIeYmJjg9ddfV019aGxsxBtvvKHnqBhjzLBw8ssYYz3o9ddfV019sLOzw0svvaTniBhjzLBw8ssYYz1ILpfD0dERAPDHP/7xqdseM8YY062Otxx6hp04cQIlJSX6DoMxxtrl4+ODmzdvol+/fsjKytJ3OIwx1i65XA4nJyd9h6FzAiIifQeha7Nnz8auXbv0HQZjjDHG2DMrMzMTc+bM0XcYurazTz75BYBZs2Zh586d+g6DMcbatWvXLsyaNUvfYehFVlYW5s6diz747KVbCQSCvpqMsF5IIBDoO4Ruw5PNGGNMDww18WWMMX3j5JcxxhhjjBkMTn4ZY4wxxpjB4OSXMcYYY4wZDE5+GWOMMcaYweDklzHGGGOMGQxOfhljjD2T9u/fD0tLS/z973/Xdyi93qFDh7BixQrs3r0b7u7uEAgEEAgEeOutt9rUnThxIqRSKYyNjfHcc8/h9OnTeohYO99++y18fHwglUrh4uKCBQsW4M6dO72mPQBoaWlBYmIi5HJ5u8djYmIwfPhwyGQyiEQieHp64uOPP0ZNTY3W8f3tb3/Dhg0b0Nzc3KWY+ypOfhljjD2TeJ1gzXzyySdITk5GVFQUAgMDcfnyZXh4eKBfv37IyMjAvn371OofPHgQO3fuxNSpU1FYWIjRo0frKXLNZGZmYt68eZg9ezZu3LiBPXv24OjRo5g8eTKampr03h4AFBUV4be//S3CwsJQV1fXbp0jR47gww8/xNWrV1FeXo64uDgkJSVh9uzZWsc3bdo0iMViTJgwAZWVlZ2KuS/j5JcxxtgzacqUKaiqqsLUqVP1HQqUSmWHT/T0af369dixYweysrIglUrVjiUnJ8PIyAjBwcGoqqrSU4Rd9/XXX2PgwIGIiIiApaUlvL29ERYWhvz8fJw8eVLv7f38889Yvnw5Fi1aBG9v7w7rWVhYIDg4GDY2NpBKpZgzZw4CAgJw4MABlJSUaB3fkiVLMGrUKLz22mudTtr7Kk5+GWOMsS7aunUrSktL9R2GmkuXLmHVqlX49NNPIRaL2xyXy+UIDQ3FzZs3sWzZMj1EqBslJSVwcHBQ25Fs0KBBAIBr167pvb1Ro0Zh9+7dmDdvHkQiUYf19u7dC2NjY7UyW1tbAFB7WqxNfGvWrEF+fj6SkpK0jrsv4+SXMcbYM+fYsWNwdnaGQCDAF198AQBITU2FRCKBubk59uzZg8mTJ0Mmk8HJyQnbt29XnZucnAyxWIwBAwbg/fffh4ODA8RiMeRyudqTs5CQEJiamsLe3l5VtnjxYkgkEggEApSXlwMAQkNDER4ejuLiYggEAnh6egIADhw4AJlMhnXr1vXEJWkjOTkZRIRp06Z1WCc2NhaDBw/Gli1bcOjQoSe2R0RISEjAsGHDIBKJYG1tjRkzZuDChQuqOpreAwBobm7G6tWr4ezsDDMzMzz//PPIzMzU+nO6u7u3+cOjdf6ru7u73tvrips3b8LMzAxubm6dis/a2hp+fn5ISkriaUKPoj5o1qxZNGvWLH2HwRhjrB2ZmZmki39+SkpKCABt2rRJVRYdHU0A6PDhw1RVVUWlpaXk6+tLEomEGhoaVPWCg4NJIpHQ+fPnqb6+ngoLC8nHx4ekUildv35dVW/evHlkZ2en1m98fDwBoLKyMlVZYGAgeXh4qNXbu3cvSaVSiomJ6fJnJSICQJmZmRrXd3d3p+HDh7d7zMPDg65cuUJERMePHycjIyNydXWlmpoaIiLKzc2l6dOnq52zevVqMjU1pW3btlFlZSWdPXuWRo8eTba2tnTnzh1VPU3vwbJly0gkEtGuXbvo/v37FBUVRUZGRvTTTz9p/BmJiPLy8kgoFFJycjJVV1fTuXPnaNiwYTRp0iSt2umu9h71wgsv0KhRozSqW1tbS1KplEJCQroU34oVKwgAnTlzRqtYtR1vz5AsfvLLGGOsz5HL5ZDJZOjfvz+CgoJQW1uL69evq9UxMTFRPcUcPnw4UlNToVAokJ6erpMYpkyZgurqaqxatUon7WmjtrYWV65cgYeHx1Prjhs3DkuXLsXVq1exfPnydusolUokJCRg5syZePPNN2FpaYmRI0fiq6++Qnl5OTZv3tzmnCfdg/r6eqSmpiIgIACBgYGwsrLCypUrIRQKtb7+fn5+iIyMREhICGQyGUaMGAGFQoEtW7Zo1U53tddZcXFxcHBwQGxsbJfi8/LyAgAUFBR0e8zPCk5+GWOM9WmmpqYAgMbGxifWGzt2LMzNzdW+xn9WlZaWgohgbm6uUf3Y2FgMGTIEKSkpOHbsWJvjhYWFqFeelDwAACAASURBVKmpwdixY9XKfXx8YGpq+tQXwR6/BxcvXkRdXR1GjBihqmNmZgZ7e3utr390dDQ2b96Mw4cPo6amBpcvX4ZcLse4cePUXhTTV3udkZ2djaysLHz//fdtXlTUNr7WMXD37t0eif1ZwMkvY4wx9n9EIhHKysr0HUaX1dfXA8ATX7B6lFgsRnp6OgQCAd5++20olUq1463LZVlYWLQ518rKCgqFQqv4amtrAQArV65UrTksEAhw7dq1DpcCa8/t27exYcMGLFy4EK+88gokEgnc3NyQlpaGW7duIT4+Xqu4dN1eZ+zYsQPr169HXl4eXF1duxyfmZkZgF/HBOPklzHGGAPw8KlkZWUlnJyc9B1Kl7UmPNpscjBu3DiEhYWhqKgIa9euVTtmZWUFAO0muZ25Zv379wcAJCYmgojUfk6cOKFxO0VFRWhubsbAgQPVymUyGWxsbFBYWKhVXLpuT1ubNm1CRkYGjhw50iaGzsbX0NAA4NcxwTj5ZYwxxgAAeXl5ICK8+OKLqjITE5OnTpfojQYMGACBQKD1+r1r167F0KFDcebMGbXyESNGwMLCAqdOnVIrP3nyJBoaGjBmzBit+hk0aBDEYjHy8/O1Ou9xrUn37du31coVCgUqKipUS4Dpqz1NEREiIyNRUFCAnJycdp+wdza+1jFgZ2en46ifXZz8MsYYM0gtLS24f/8+mpqacPbsWYSGhsLZ2Rnz589X1fH09ERFRQVycnLQ2NiIsrKydtd6tbGxwa1bt3D16lUoFAo0NjYiNzdXb0udmZubw93dHTdu3NDqvNbpD4+vNysWixEeHo7s7GxkZGSguroaBQUFWLRoERwcHBAcHKx1PwsWLMD27duRmpqK6upqNDc348aNG6rELigoCHZ2dk/cXtnNzQ3+/v5IS0vD0aNHoVQqUVJSoornnXfeUdXVR3uaOn/+PD7//HOkpaVBKBSqTQURCATYuHGj1vG1ah0DI0eO7HKcfYYel5roNrzUGWOM9V66WOps06ZNZG9vTwDI3Nycpk2bRikpKWRubk4AyMvLi4qLi2nz5s0kk8kIALm4uNAvv/xCRA+XOhMKheTo6EgmJiYkk8loxowZVFxcrNbPvXv3yN/fn8RiMbm5udFHH31EERERBIA8PT1Vy6KdPn2aXFxcyMzMjF566SW6c+cO7d+/n6RSKcXGxnbps7aClktPhYSEkFAopLq6OlVZdnY2eXh4EACytbWlDz/8sN1zIyIi2ix11tLSQvHx8eTl5UVCoZCsra0pICCALl68qKqjzT148OABRUZGkrOzM5mYmFD//v0pMDCQCgsLiYgoICCAANDq1auf+DnLy8spNDSUPD09SSQSkYWFBY0fP56+++47tXr6au/EiRM0fvx4cnBwIAAEgOzt7Ukul9MPP/xAREQFBQWqY+39xMfHax1fqylTppCjoyO1tLQ8Mc7HaTveniFZnPwyxhjrUbpa57crgoODycbGRq8xaEvbZKSoqIhMTExo27Zt3RhV92lubiZfX1/aunWrQbTXHcrLy0ksFtPGjRu1PrcvJ7887YExxphB0uZlsGeRp6cnYmJiEBMTg5qaGn2Ho5Xm5mbk5ORAoVAgKCioz7fXXdasWQNvb2+EhIToO5RehZNfPFyn0NjYGN7e3j3e94IFCyAWiyEQCHgZkv+zceNG1csaX331lap8//79sLS0xN///vdu7b+n+umM+vp6DB06FCtXrtT63N27d8Pd3b3NXDITExPY2trid7/7HbKzs7shanWajvnH433rrbfa1Jk4cSKkUimMjY3x3HPP6WTuXXfisc162ooVKzB79mwEBQVp/fKbPuXl5WH37t3Izc3VeK3iZ7m97pCQkID8/Hzs378fQqFQ3+H0Kpz8Avjpp5/g7++vl77T09OxbNkyvfTdWy1btgzHjx9vU049tC95T/XTGdHR0bh48WKnzg0MDMTly5fh4eEBS0tL1bJCZWVlyMzMxM2bNxEYGIjMzEwdR61O0zH/aLz9+vVDRkYG9u3bp1bn4MGD2LlzJ6ZOnYrCwkKMHj26u8LWCR7bvUNUVBTS09NRVVUFNzc37Nq1S98hdat169YhJCQEn332mb5D0diECRPwzTffwN7e3iDa07U9e/bgwYMHyMvLg7W1tb7D6XU4+X2EQCDochtKpRJyuVwH0bDHTZkyBVVVVZg6darO2mzvfnVHP7pw/PhxnDt3TuftWltbY8KECfjf//1fAEBWVpZW5/fEmE9OToaRkRGCg4OfqadXmjL0sd3T4uLi8ODBAxARrly5glmzZuk7pG43ceJErF+/Xt9hsB4yffp0rFixos2qHewhTn4foYuvBbZu3YrS0tJOnauL5Jtppyv3qycplUpEREQgKSmp2/po3UmodScnTfXEmJfL5QgNDcXNmzf5mxINPStjmzHGehonv4+4dOkShg4dColEAjMzM/j6+rbZ4/yf//wnhg8fDktLS4jFYowcORLff/89ACA0NBTh4eEoLi6GQCCAp6en6rxt27Zh7NixEIvFkEgkcHV1VdtBx8jICPv27cPkyZNhaWkJBwcH/OlPf9L6M6SmpkIikcDc3Bx79uzB5MmTIZPJ4OTkhO3bt6vVJSIkJCRg2LBhEIlEsLa2xowZM9T2Vf/8889hbm4OqVSK0tJShIeHw9HREYsWLYJEIoGRkRHGjBkDOzs7CIVCSCQSjB49Gr6+vqpFzK2srPDxxx9rfB3bc+zYMTg7O0MgEOCLL74A8PB+PT5/tfXn//2//9ep+9VeP5peK22uvbaio6OxePFi1a5Ijztw4ECX1xM9e/YsAMDPz0+tvLeM+djYWAwePBhbtmzBoUOHnvhZeGw/O2ObMcZ6nJ6WmehWnVnqbMKECeTu7k5XrlyhxsZGOnfuHL3wwgskFotVaxISEe3cuZPWrFlDFRUVdO/ePXrxxRepX79+quOBgYHk4eGh1nZiYiIBoM8++4zu3btHFRUV9PXXX9O8efOIiCg6OpoA0OHDh6myspIqKirotddeI5FIRLW1tVp//kfbq6qqotLSUvL19SWJREINDQ2qeqtXryZTU1Patm0bVVZW0tmzZ2n06NFka2tLd+7cadPekiVLaNOmTTRz5kz673//S5988gkBoJMnT1JtbS2Vl5fTq6++SgBo3759VFZWRrW1tRQSEkIAKD8/X+PrWFRURADoyy+/VJWVlJQQANq0aZOqzvLly1XX6Pbt22RtbU1yuZyam5s7fb8e76cz1+pp114bx44do2nTphERUVlZGQGg6OhotTp79+4lqVRKMTExT23Pw8ODLC0tVf9dV1dHubm55OLiQhMnTqSamhq1+voe8x4eHnTlyhUiIjp+/DgZGRmRq6urKs7c3Nw265Hy2O7dY7s3LHX2LELfXXqK9UJ9eLzxOr+tJkyYQKNGjVIrO3v2LAGgZcuWdXheXFwcAaDS0lIiavsPTkNDA1lZWZG/v7/aeU1NTZSUlEREv/6jolQqVcf/+te/EgA6d+6cVp+jo/ZSUlIIAF26dImIHiY8FhYWFBQUpHbujz/+SADUkqj22iMiVYKgUChUZX/5y18IABUUFLRpc8eOHR3G/Ph11CRBeFxAQACJxWK6cOGCxv1okiB09Vo9fu21UVdXR2PHjqUbN24QUcfJrzZaF7h//GfkyJH0l7/8hR48ePDE83t6zD+a/BIRhYeHEwDV4vyPJ788tnv/2Obkt3P6cDLCeqE+PN6yTLry1LivGzlyJCwtLVVfB7endZ5wR+tFnj17FpWVlZg0aZJaubGxMZYsWfLUdnW1p7ypqalae4WFhaipqcHYsWPV6vn4+MDU1BQnT57sUj9NTU2qMk0+y9Ou49NkZWXhu+++w4YNGzBkyBCd9tPVa/X4tddGVFQUFi5cCEdHR63PfRJLS0vV3N6mpibcvXsXBw8eREhICOLi4nDs2DHY2tq2e66+x3xsbCz27t2LlJQUzJ07t81xHtvPxtgGgNmzZ3fqPEOWmJiInTt36jsMxp5pPOf3KYRCodov9n379uHll19G//79IRKJ2sz3e1x1dTUAwMrKqlvj1FZr4mNhYdHmmJWVFRQKRbf2r+11fJJ79+7ho48+go+PD8LDw3Xej76u1bFjx1BQUIB33323W9pvZWJiAkdHRyxYsAAbN27ExYsX1ZZE6m1jXiwWIz09HQKBAG+//TaUSqXacR7bmtP3tWKMMX3gJ79P0NTUhIqKCjg7OwMArl+/joCAAMycORN/+tOfMHDgQGzatOmJ/+gMHDgQAFBeXt4jMWuqNTFp7x+3yspKODk5dVvfnbmOT7JkyRJUVlbiyJEjasu66KoffV2rrVu34vDhwzAyavs36rp167Bu3Tr89NNPbZ7adcXIkSMBAOfPnwfQe8f8uHHjEBYWho0bN2Lt2rWq/0cBHtva0Oe1AsBPMLUkEAiwdOlSzJkzR9+hMAPQl1eg4ie/T/CPf/wDLS0tqoXzCwoK0NjYiA8++ADu7u6qXaqexNXVFTY2Njh48GBPhKyxESNGwMLCAqdOnVIrP3nyJBoaGjBmzJhu67sz17Ej+/btwzfffINVq1bhueeeU5VHRETorB99Xav09HTVRhT0yIYUwMPVH4hIp4kvAPznP/8BANXX6715zK9duxZDhw7FmTNn1Mp5bGtOn9eKMcb0hZPfRzQ0NKCqqgpNTU04ffo0QkJC4OLigvnz5wOA6unSoUOHUF9fj6KiojZz4mxsbHDr1i1cvXoVCoUCRkZGiIqKwtGjRxESEoKbN2+ipaUFCoVC9XRNH8RiMcLDw5GdnY2MjAxUV1ejoKAAixYtgoODA4KDg7utb02uoyaqq6vx/vvvw9vbG8uXLwfwcPvfU6dOIT8/v1P3q725i/q8VprKzc3VeqkzpVKJlpYWEBFu3bqF9PR0rFy5Era2tli6dCmA3j3mW6c/PL6IO4/th/rK2GaMMZ3T5+t23aUzqz2kp6eTv78/DRgwgExMTKhfv370+uuv07Vr19TqRUZGko2NDVlZWdHs2bPpiy++IADk4eFB169fp9OnT5OLiwuZmZnRSy+9pFoq6IsvvqCRI0eSWCwmsVhMv/nNbyglJYU2bNhAZmZmBIC8vLyouLiYMjIyyNramgCQk5OTVis+pKSkkLm5uVp7mzdvJplMRgDIxcVFtXRbS0sLxcfHk5eXFwmFQrK2tqaAgAC6ePGiqr1H4xs0aBBt27aNiIiSkpJU/bi6utI///lPWr9+PVlaWhIAsrOzo2+++YZ27NhBdnZ2BICsra1p+/btT72OoaGhqnMkEgnNnDmTNm3aRPb29gSAzM3Nadq0abRx48Z2Vy0AQK+99lqn7tfKlSvb9KPptdLm2ndWR6s97N+/n6RSKcXGxnZ4bnZ2docrPYhEIvLy8qIPPviArl+/rnaevsb8o/Ha2tqqVnd4XERERJulznhs9+6xzas9dA767tv3rBfqw+MtS0DU9zZ7b32DmOeTMcZY75OVlYW5c+eiD/7z060EAgEyMzN5zi/rEX14vO3kaQ+MMcYYY8xgcPL7DLhw4UKHW50++hMUFKTvUNlT8L1kjOnDoUOHsGLFCuzevRvu7u6q3zVvvfVWm7oTJ06EVCqFsbExnnvuOZw+fVoPEWvn22+/hY+PD6RSKVxcXLBgwQLcuXOn17QHAC0tLUhMTIRcLm/3eExMDIYPHw6ZTAaRSARPT098/PHHqKmp0Tq+v/3tb9iwYUOn1xfv8/Q876JbdGbOL2OMsZ7Bc347B52cg7l69WqaOnUqVVdXq8o8PDyoX79+BID27t3b5pz2tg3vrXbs2EEAaMOGDVRZWUlnzpwhd3d38vb2psbGRr23R0T0yy+/0Pjx4wlAm91kW/n5+VFKSgrdu3ePqqurKTMzk4RCIb366qudii8pKYn8/Pzo/v37nYq5s+PtGZDFT34ZY4wZHKVS2eETuGepj6dZv349duzYgaysLEilUrVjycnJMDIyQnBwMKqqqvQUYdd9/fXXGDhwICIiImBpaQlvb2+EhYUhPz+/U6ut6Lq9n3/+GcuXL8eiRYvg7e3dYT0LCwsEBwfDxsYGUqkUc+bMQUBAAA4cOICSkhKt41uyZAlGjRqF1157TW1nSsbTHhhjjBmgrVu3orS09Jnv40kuXbqEVatW4dNPP4VYLG5zXC6XIzQ0FDdv3sSyZcv0EKFulJSUwMHBQW2t60GDBgEArl27pvf2Ro0ahd27d2PevHkQiUQd1tu7d2+bpRtbt5mvq6vrVHxr1qxBfn4+kpKStI67L+PklzHGWK9HREhISMCwYcMgEolgbW2NGTNm4MKFC6o6ISEhMDU1hb29vaps8eLFkEgkEAgEql0HQ0NDER4ejuLiYggEAnh6eiI5ORlisRgDBgzA+++/DwcHB4jFYsjlcrWnaV3pAwAOHDig9ZrcnZWcnAwiwrRp0zqsExsbi8GDB2PLli04dOjQE9vT5B6kpqZCIpHA3Nwce/bsweTJkyGTyeDk5ITt27ertdfc3IzVq1fD2dkZZmZmeP7555GZman153R3d2/zR0br/Fd3d3e9t9cVN2/ehJmZGdzc3DoVn7W1Nfz8/JCUlMSrqzxKv9MuugfP+WWMsd6rM3N+V69eTaamprRt2zaqrKyks2fP0ujRo8nW1la1tjQR0bx588jOzk7t3Pj4eAJAZWVlqrLAwEDy8PBQqxccHEwSiYTOnz9P9fX1VFhYSD4+PiSVStXWv+5KH3v37iWpVEoxMTFafX4i7edguru70/Dhw9s95uHhQVeuXCEiouPHj5ORkRG5urpSTU0NEbU/51fTexAdHU0A6PDhw1RVVUWlpaXk6+tLEomEGhoaVPWWLVtGIpGIdu3aRffv36eoqCgyMjKin376SePPSESUl5dHQqGQkpOTqbq6ms6dO0fDhg2jSZMmadVOd7X3qBdeeKHDOb+Pq62tJalUSiEhIV2Kb8WKFQSAzpw5o1Ws2o63ZwjP+WWMMda7KZVKJCQkYObMmXjzzTdhaWmJkSNH4quvvkJ5eTk2b96ss75MTExUTzaHDx+O1NRUKBQKpKen66T9KVOmoLq6GqtWrdJJex2pra3FlStX4OHh8dS648aNw9KlS3H16lXVjoKP68w9kMvlkMlk6N+/P4KCglBbW4vr168DeLhjYWpqKgICAhAYGAgrKyusXLkSQqFQ62vt5+eHyMhIhISEQCaTYcSIEVAoFNiyZYtW7XRXe50VFxcHBwcHxMbGdik+Ly8vAA+3X2cPcfLLGGOsVyssLERNTQ3Gjh2rVu7j4wNTU9NOvYSkqbFjx8Lc3Fztq/1nQWlpKYgI5ubmGtWPjY3FkCFDkJKSgmPHjrU53tV7YGpqCgCqbbYvXryIuro6jBgxQlXHzMwM9vb2Wl/r6OhobN68GYcPH0ZNTQ0uX74MuVyOcePGqb0opq/2OiM7OxtZWVn4/vvv27yoqG18rWPg7t27PRL7s4CTX8YYY71aZWUlgIdvwz/OysoKCoWiW/sXiUQoKyvr1j50rb6+HgCe+ILVo8RiMdLT0yEQCPD2229DqVSqHdf1PaitrQUArFy5Um2N82vXrqm93PU0t2/fxoYNG7Bw4UK88sorkEgkcHNzQ1paGm7duoX4+Hit4tJ1e52xY8cOrF+/Hnl5eXB1de1yfGZmZgB+HROMk1/GGGO9nJWVFQC0m2BVVlbCycmp2/pubGzs9j66Q2vCo80mB+PGjUNYWBiKioqwdu1atWO6vgf9+/cHACQmJoKI1H5OnDihcTtFRUVobm7GwIED1cplMhlsbGxQWFioVVy6bk9bmzZtQkZGBo4cOdImhs7G19DQAODXMcE4+WWMMdbLjRgxAhYWFjh16pRa+cmTJ9HQ0IAxY8aoykxMTFRfretCXl4eiAgvvvhit/XRHQYMGACBQKD1+r1r167F0KFDcebMGbVybe6BJgYNGgSxWIz8/Hytzntca9J9+/ZttXKFQoGKigrVEmD6ak9TRITIyEgUFBQgJyen3SfsnY2vdQzY2dnpOOpnFye/jDHGejWxWIzw8HBkZ2cjIyMD1dXVKCgowKJFi+Dg4IDg4GBVXU9PT1RUVCAnJweNjY0oKytrd21WGxsb3Lp1C1evXoVCoVAlsy0tLbh//z6amppw9uxZhIaGwtnZGfPnz9dJH7m5uT2y1Jm5uTnc3d1x48YNrc5rnf7w+Hqz2twDTftZsGABtm/fjtTUVFRXV6O5uRk3btxQJXZBQUGws7N74vbKbm5u8Pf3R1paGo4ePQqlUomSkhJVPO+8846qrj7a09T58+fx+eefIy0tDUKhsM2W9xs3btQ6vlatY2DkyJFdjrPP0ONSE92GlzpjjLHeqzNLnbW0tFB8fDx5eXmRUCgka2trCggIoIsXL6rVu3fvHvn7+5NYLCY3Nzf66KOPKCIiggCQp6enasmy06dPk4uLC5mZmdFLL71Ed+7coeDgYBIKheTo6EgmJiYkk8loxowZVFxcrLM+9u/fT1KplGJjY7W+btBy6amQkBASCoVUV1enKsvOziYPDw8CQLa2tvThhx+2e25ERESbpc40uQcpKSlkbm5OAMjLy4uKi4tp8+bNJJPJCAC5uLjQL7/8QkREDx48oMjISHJ2diYTExPq378/BQYGUmFhIRERBQQEEABavXr1Ez9neXk5hYaGkqenJ4lEIrKwsKDx48fTd999p1ZPX+2dOHGCxo8fTw4ODgSAAJC9vT3J5XL64YcfiIiooKBAday9n/j4eK3jazVlyhRydHSklpaWJ8b5OG3H2zMki5NfxhhjPaozyW9PCA4OJhsbG32H0SFtk5GioiIyMTGhbdu2dWNU3ae5uZl8fX1p69atBtFedygvLyexWEwbN27U+ty+nPzytAfGGGPs/2jzglhv5+npiZiYGMTExKCmpkbf4WilubkZOTk5UCgUCAoK6vPtdZc1a9bA29sbISEh+g6lV+HklzHGGOujVqxYgdmzZyMoKEjrl9/0KS8vD7t370Zubq7GaxU/y+11h4SEBOTn52P//v0QCoX6DqdX4eSXMcaYwYuKikJ6ejqqqqrg5uaGXbt26TsknVm3bh1CQkLw2Wef6TsUjU2YMAHffPMN7O3tDaI9XduzZw8ePHiAvLw8WFtb6zucXsdE3wEwxhhj+hYXF4e4uDh9h9FtJk6ciIkTJ+o7DNZDpk+fjunTp+s7jF6Ln/wyxhhjjDGDwckvY4wxxhgzGJz8MsYYY4wxg8HJL2OMMcYYMxic/DLGGGOMMYPRZ1d72LVrFwQCgb7DYIwx1gH+Ha29uXPnYu7cufoO4/+zd+dxUVf7/8BfAwwMywyLiqCIsriR2zXpKmZmlqV+MXGL1EorxS1EzFxwReSqeIGLwTWRL30fWgqIgdetrhpxTfNX11TCqyFuuAUoyLApy/n94YO5joAyMAs6r+fjMX90Puecz3s+c5zefOZ8ziF6pkmEEMLQQWjbiRMnkJeXZ+gwiIgadOLECURHRyMpKcnQoRARNcrHxwcuLi6GDkPbUp7L5JeIqDVLTk7GO++8A379EhHpXQrn/BIRERGR0WDyS0RERERGg8kvERERERkNJr9EREREZDSY/BIRERGR0WDyS0RERERGg8kvERERERkNJr9EREREZDSY/BIRERGR0WDyS0RERERGg8kvERERERkNJr9EREREZDSY/BIRERGR0WDyS0RERERGg8kvERERERkNJr9EREREZDSY/BIRERGR0WDyS0RERERGg8kvERERERkNJr9EREREZDSY/BIRERGR0WDyS0RERERGg8kvERERERkNJr9EREREZDSY/BIRERGR0WDyS0RERERGg8kvERERERkNJr9EREREZDSY/BIRERGR0WDyS0RERERGg8kvERERERkNJr9EREREZDTMDB0AEdHzrKCgAN98841a2S+//AIA2Lp1q1q5XC7Hu+++q7fYiIiMkUQIIQwdBBHR8+r+/ftwdHREaWkpTE1NAQB1X7sSiURVr6qqCh988AG+/PJLQ4RJRGQsUjjtgYhIhywsLDBhwgSYmZmhqqoKVVVVqK6uRnV1teq/q6qqAACTJ082cLRERM8/Jr9ERDo2efJkPHjw4Il17Ozs8Nprr+kpIiIi48Xkl4hIx4YNG4Z27do1elwqlWLq1KkwM+NjGEREusbkl4hIx0xMTDBlyhRIpdIGj1dVVfFBNyIiPWHyS0SkB++++65qbu/jOnTogEGDBuk5IiIi48Tkl4hID1566SV07ty5Xrm5uTk++OADtZUfiIhId5j8EhHpyXvvvVdv6sODBw845YGISI+Y/BIR6cmUKVPqTX3w9PRE7969DRQREZHxYfJLRKQnPXr0gJeXl2qKg1QqxfTp0w0cFRGRcWHyS0SkR++//75qp7fq6mpOeSAi0jMmv0REevTuu++ipqYGANC/f3+4ubkZOCIiIuPC5JeISI9cXV3x5z//GQDwwQcfGDgaIiLjw+2EdOTEiROIjIw0dBhE1Ardv38fEokE3333HTIzMw0dDhG1QikpKYYO4bnFO786kpeXh927dxs6DCJqhVxcXNC+fXvIZDK9n/unn37CTz/9pPfzPsuuX7/O73PSG4433eOdXx3jX25E1JCLFy/C09NT7+edOHEiAH43aSI5ORnvvPMOrxnpRd14I93hnV8iIgMwROJLRERMfomIiIjIiDD5JSIiIiKjweSXiIiIiIwGk18iIiIiMhpMfomISGMHDhyAra0t/vGPfxg6lFbv8OHDWLp0KVJTU+Hu7g6JRAKJRIL33nuvXt0RI0ZALpfD1NQUL7zwAk6dOmWAiDXz9ddfw9vbG3K5HJ07d8b06dNx+/btVtMfANTW1iIqKgo+Pj4NHg8NDYWXlxcUCgUsLCzg6emJzz77DKWlpRrHt3fvXmzYsEG1kyO1Pkx+dZmvegAAIABJREFUiYhIY0IIQ4fwTFi1ahViYmKwbNkyjB8/HpcuXYKHhwfatGmDHTt2YP/+/Wr1v/vuO6SkpMDX1xfZ2dno37+/gSJvmqSkJEyZMgUTJ07E9evXkZ6ejszMTIwcORLV1dUG7w8AcnJy8MorryA4OBjl5eUN1jl69CjmzZuHK1euoLCwEOHh4YiOjlYtDahJfGPGjIFMJsPw4cNRXFzcrJhJt5j8EhGRxkaPHo179+7B19fX0KGgoqKi0Tt6hrR+/Xrs2rULycnJkMvlasdiYmJgYmKCgIAA3Lt3z0ARttwXX3yBDh06YNGiRbC1tUW/fv0QHByM06dP4+TJkwbv78yZM1iyZAlmz56Nfv36NVrPxsYGAQEBcHBwgFwux6RJk+Dn54dDhw4hLy9P4/jmz5+Pvn37YtSoUc1O2kl3mPwSEdEzLSEhAfn5+YYOQ83FixexYsUKrFmzpsGd/Hx8fBAUFIQbN27g008/NUCE2pGXlwdnZ2dIJBJVWadOnQAAV69eNXh/ffv2RWpqKqZMmQILC4tG6+3btw+mpqZqZW3btgUAtbvFmsS3evVqnD59GtHR0RrHTbrF5JeIiDRy7NgxuLq6QiKR4PPPPwcAxMXFwdraGlZWVkhPT8fIkSOhUCjg4uKCnTt3qtrGxMRAJpPB0dERs2bNgrOzM2QyGXx8fNTunAUGBsLc3BxOTk6qsrlz58La2hoSiQSFhYUAgKCgICxcuBC5ubmQSCSqzUMOHToEhUKBdevW6eOS1BMTEwMhBMaMGdNonbCwMHTr1g3btm3D4cOHn9ifEAKRkZHo2bMnLCwsYG9vj7Fjx+L8+fOqOk39DACgpqYGK1euhKurKywtLdGnTx8kJSVp/D7d3d3r/eFRN//V3d3d4P21xI0bN2BpaQk3N7dmxWdvb4+hQ4ciOjqa04RaG0E6kZSUJHh5iai1mTBhgpgwYUKL+8nLyxMAxObNm1VlISEhAoA4cuSIuHfvnsjPzxdDhgwR1tbW4sGDB6p6AQEBwtraWpw7d05UVlaK7Oxs4e3tLeRyubh27Zqq3pQpU0T79u3VzhsRESEAiIKCAlXZ+PHjhYeHh1q9ffv2CblcLkJDQ1v8Xpvzfe7u7i68vLwaPObh4SEuX74shBDi+PHjwsTERHTp0kWUlpYKIYQ4ePCgePvtt9XarFy5Upibm4vt27eL4uJicfbsWdG/f3/Rtm1bcfv2bVW9pn4Gn376qbCwsBC7d+8WRUVFYtmyZcLExET8/PPPGr3PjIwMIZVKRUxMjCgpKRG//fab6Nmzp3jzzTc16kdX/T3qz3/+s+jbt2+T6paVlQm5XC4CAwNbFN/SpUsFAPHrr782OU7mDzqXzDu/RESkVT4+PlAoFGjXrh38/f1RVlaGa9euqdUxMzNT3cX08vJCXFwclEolEhMTtRLD6NGjUVJSghUrVmilP02UlZXh8uXL8PDweGrdQYMGYcGCBbhy5QqWLFnSYJ2KigpERkZi3LhxmDp1KmxtbdG7d29s2bIFhYWF2Lp1a702T/oMKisrERcXBz8/P4wfPx52dnZYvnw5pFKpxtd/6NChWLx4MQIDA6FQKNCrVy8olUps27ZNo3501V9zhYeHw9nZGWFhYS2Kr2vXrgCArKwsncdMTcfkl4iIdMbc3BwAUFVV9cR6AwYMgJWVldrP+M+q/Px8CCFgZWXVpPphYWHo3r07YmNjcezYsXrHs7OzUVpaigEDBqiVe3t7w9zc/KkPgj3+GVy4cAHl5eXo1auXqo6lpSWcnJw0vv4hISHYunUrjhw5gtLSUly6dAk+Pj4YNGiQ2oNihuqvOfbs2YPk5GR8++239R5U1DS+ujHwxx9/6CV2ahomv0RE1CpYWFigoKDA0GG0WGVlJQA88QGrR8lkMiQmJkIikeDDDz9ERUWF2vG65bJsbGzqtbWzs4NSqdQovrKyMgDA8uXLVWsOSyQSXL16tdGlwBpy69YtbNiwATNnzsRrr70Ga2truLm5IT4+Hjdv3kRERIRGcWm7v+bYtWsX1q9fj4yMDHTp0qXF8VlaWgL475ig1oHJLxERGVxVVRWKi4vh4uJi6FBarC7h0WSTg0GDBiE4OBg5OTlYu3at2jE7OzsAaDDJbc41a9euHQAgKioKQgi114kTJ5rcT05ODmpqatChQwe1coVCAQcHB2RnZ2sUl7b709TmzZuxY8cOHD16tF4MzY3vwYMHAP47Jqh1YPJLREQGl5GRASEEBg4cqCozMzN76nSJ1sjR0RESiUTj9XvXrl2LHj164Ndff1Ur79WrF2xsbPDLL7+olZ88eRIPHjzAiy++qNF5OnXqBJlMhtOnT2vU7nF1SfetW7fUypVKJe7evataAsxQ/TWVEAKLFy9GVlYW0tLSGrzD3tz46sZA+/bttRw1tQSTXyIi0rva2loUFRWhuroaZ8+eRVBQEFxdXTFt2jRVHU9PT9y9exdpaWmoqqpCQUFBg2u9Ojg44ObNm7hy5QqUSiWqqqpw8OBBgy11ZmVlBXd3d1y/fl2jdnXTHx5fb1Ymk2HhwoXYs2cPduzYgZKSEmRlZWH27NlwdnZGQECAxueZPn06du7cibi4OJSUlKCmpgbXr19XJXb+/v5o3779E7dXdnNzw7BhwxAfH4/MzExUVFQgLy9PFc9HH32kqmuI/prq3Llz2LhxI+Lj4yGVStWmgkgkEmzatEnj+OrUjYHevXu3OE7SIgMuNfFc41IlRNQaaWOps82bNwsnJycBQFhZWYkxY8aI2NhYYWVlJQCIrl27itzcXLF161ahUCgEANG5c2fx+++/CyEeLnUmlUpFx44dhZmZmVAoFGLs2LEiNzdX7Tx37twRw4YNEzKZTLi5uYlPPvlELFq0SAAQnp6eqmXRTp06JTp37iwsLS3Fyy+/LG7fvi0OHDgg5HK5CAsLa9F7FaJ53+eBgYFCKpWK8vJyVdmePXuEh4eHACDatm0r5s2b12DbRYsW1VvqrLa2VkRERIiuXbsKqVQq7O3thZ+fn7hw4YKqjiafwf3798XixYuFq6urMDMzE+3atRPjx48X2dnZQggh/Pz8BACxcuXKJ77PwsJCERQUJDw9PYWFhYWwsbERgwcPFt98841aPUP1d+LECTF48GDh7OwsAAgAwsnJSfj4+IgffvhBCCFEVlaW6lhDr4iICI3jqzN69GjRsWNHUVtb+8Q4H8X8QeeSeXV1hIOXiFojba3z2xIBAQHCwcHBoDFoojnf5zk5OcLMzExs375dR1HpVk1NjRgyZIhISEgwiv50obCwUMhkMrFp0yaN2jF/0Dmu80tERPqnycNgzyJPT0+EhoYiNDQUpaWlhg5HIzU1NUhLS4NSqYS/v/9z35+urF69Gv369UNgYKChQ6HHMPklIiLSgaVLl2LixInw9/fX+OE3Q8rIyEBqaioOHjzY5LWKn+X+dCEyMhKnT5/GgQMHIJVKDR0OPYbJLz0zNm3apHqKesuWLaryAwcOwNbWFv/4xz90du7Q0FB4eXlBoVDAwsICnp6e+Oyzz556R+fjjz+GXC6HRCJp8ZPVwMO1Inv06IHly5dr3DY1NRXu7u6qhzietvNVZGQkJBIJTExM0KNHD2RmZjY37KfGIpFIIJVK0bFjR0yZMgX/+c9/tHaux7X2cdTQtZFIJDA3N4ejoyNeffVVREREoKioSGdx6tKyZcuQmJiIe/fuwc3NDbt37zZ0SDq1bt06BAYG4i9/+YuhQ2my4cOH46uvvoKTk5NR9Kdt6enpuH//PjIyMmBvb2/ocKghhp548bzinB3dyMnJEQDE3//+d1XZvn37hEKhEHv37tXZeYcOHSpiY2PFnTt3RElJiUhKShJSqVS89dZbT227c+dOjfd2b0xwcLAAIEJCQprdR90DN05OTuLBgwcN1qmurhadO3cWAMTw4cObfa6mxGJrayuEEKK0tFTs3btXuLq6ChsbG3H+/HmdnfdZGEePXpva2lpRVFQkvv/+ezFt2jQhkUiEs7Oz+PnnnzWOoTXM+X3W8Puc9InjTec455eefaNHj8a9e/fg6+urs3PY2NggICAADg4OkMvlmDRpEvz8/HDo0CG9bbl5/Phx/Pbbb1rp68UXX8Tt27eRlpbW4PHU1FR07NhRK+dqKmtra/j6+uJvf/sbSktLsXnzZr2evzWPI4lEAjs7O7z66qtITExEcnIy/vjjD1XMRETUdEx+iR4jhEBKSgq2bt2qKtu3b1+9tTfbtm0LAE/dDlQikbQ4poqKCixatAjR0dEt7gsA5syZAwD4+9//3uDxyMhILFy4UCvn0tRLL70EAFpL9A1F2+PoURMmTMC0adOQn5+vNnWDiIiejslvKxEdHQ1ra2uYmJjgxRdfRPv27SGVSmFtbY3+/ftjyJAhql157Ozs8Nlnn6m1/9e//gUvLy/Y2tpCJpOhd+/e+PbbbwEAX375JWxsbCCRSGBvb4+0tDT88ssv6Ny5M0xNTTF58mSNYo2JiYFMJoOjoyNmzZoFZ2dnyGQy+Pj44OTJk2p1hRCIjIxEz549YWFhAXt7e4wdOxbnz59vVr3HHTt2DK6urpBIJPj8888BAHFxcbC2toaVlRXS09MxcuRIKBQKuLi4YOfOnWrta2pqEB4eju7du8PS0hJt27aFm5sbwsPDMWnSpCee+8aNG7C0tISbm5va+4iIiED37t1hYWEBW1tbLFq06KnX9GlCQkIwd+5c1bakjzt06JBGC/q/9tpr6NmzJ77//ntcuHBB7diPP/6I8vJyjBgxosG2uh5r1dXVAAALCwtVmbGNo6ao2wzi4MGDGrUjIjJ6Bpxz8VxrzpydVatWCQDi5MmToqysTBQWFoq33npLABD79+8XBQUFoqysTAQGBgoA4vTp06q2KSkpYvXq1eLu3bvizp07YuDAgaJNmzaq4+fOnRNWVlbigw8+UJUtXbpUbNu2rVnvLyAgQFhbW4tz586JyspKkZ2dLby9vYVcLlctPC+EECtXrhTm5uZi+/btori4WJw9e1b0799ftG3bVty+fVvjeg3N1czLyxMAxObNm1VlISEhAoA4cuSIuHfvnsjPzxdDhgwR1tbWavNc161bJ0xNTUV6erooLy8X//73v0X79u3Fq6+++sT3X1ZWJuRyuQgMDFQrDwkJERKJRPz1r38VRUVFory8XMTGxrZozu+xY8fEmDFjhBBCFBQUNDjnd9++fUIul4vQ0NCn9ufh4SEuX74s/va3vwkAIigoSO24n5+fSExMFEqlssE5v9oca4/Oa62zfft2AUAsWrRIVWZs46ixa/OokpISAUB06tTpied4HOf8ao5zMEmfON50jptc6EpLkl+lUqkq+7//+z8BQGRlZanK/t//+38CgNi1a1ejfYWHhwsAIj8/X1X2xRdfCABix44d4uuvvxbBwcEaxfeogICAev9j/vnnnwUAsWbNGiGEEOXl5cLGxkb4+/ur1auLvy5Ra2o9ITRPWioqKlRldUnoxYsXVWXe3t7ipZdeUjvvzJkzhYmJibh//36j7z8kJER069ZNlJSUqMrKy8uFlZWVeOONN9TqtuSBt/LycjFgwABx/fp1IUTjya8m6pLf4uJiYW1tLezt7VW7UOXm5goXFxdx//79RpPfx7VkrD3+wNvu3btF+/bthaOjo+o9G9s4aujaNEYikQg7O7sn1nkck1/NMRkhfeJ407lkM93cTyZtMTc3B/Dfn4IBqNYMrKqqarRdXZ1HF5KfOXMm/vnPf2LWrFl4/fXXtb7E0IABA2BlZaX6iTk7OxulpaUYMGCAWj1vb2+Ym5urpkg0tV5L1V3LR69bZWUlZDKZWr2amhpIpdJ6czPr7NmzB8nJyfjuu+8gl8tV5RcvXkR5eTmGDx+ulXiBh8tCzZw5UycPn9na2mLy5MmIj4/Hrl27MH36dERFRWHOnDkwNzfHgwcPmtRPS8favXv3IJFIYGpqCicnJ4waNQqrVq1SvWdjG0dNVVZWBiEEFAqFxm13796tlbnoxobXjOj5wOT3ObF//35EREQgOzsbJSUljSbG69atw+7du5Gfn6+TOCwsLFBQUAAAKC4uBvDwCffH2dnZQalUalRPF0aNGoWIiAikp6djxIgRyM7ORlpaGv7nf/6nwaRl165diIyMREZGBjp06KB27Pr16wDQ6LxcTR07dgxZWVmIjIzUSn8NmTNnDuLj47Flyxb4+fkhJSXlqWvsanus2draqsZAQ4xtHDXV77//DgDo0aOHxm0HDhyIBQsWNOu8xujEiROIjo5GUlKSoUMhI1A33kh3mPw+B65duwY/Pz+MGzcO//u//4sOHTpg8+bN9R6Kq6qqwvz581VP8oeFhWHVqlVai6OqqgrFxcVwcXEB8DDhANBg0tGcerqwevVq/Pvf/8a0adNQWloKZ2dnTJo0qcEHxzZv3oxvv/0WR48ebTDBqrvzd//+fa3ElpCQgCNHjsDEpP5zqevWrcO6devw888/17vTqYl+/fph4MCB+OmnnxAQEICJEyc+cVF2Q4w1YxtHTXXo0CEAwMiRIzVu6+Li8tQH8UhddHQ0rxnpDZNf3WLy+xzIyspCVVUV5syZA3d3dwAN/zz3ySefYMaMGRg3bhxu3LiBtWvXYsSIERg0aJBW4sjIyIAQAgMHDgQA9OrVCzY2Nvjll1/U6p08eRIPHjzAiy++qFE9XcjOzkZubi4KCgpgZtbwPwchBJYsWYKioiKkpaU1Wq9Xr14wMTHBDz/8gNmzZ7c4tsTERCQmJqqVFRYWol27dggJCUFYWFiLzwE8vPv7008/Yffu3cjJyXliXUOMNWMbR01x+/ZtREVFwcXFBR9++GGz+yEiMkZc6uw54OrqCgA4fPgwKisrkZOTU29+Y2xsLDp27Ihx48YBAMLDw+Hl5YUpU6agpKSkWeetra1FUVERqqurcfbsWQQFBcHV1VW1BJNMJsPChQuxZ88e7NixAyUlJcjKysLs2bPh7OyMgIAAjerpwrx58+Dq6vrEbYrPnTuHjRs3Ij4+HlKptN62s5s2bQLwcLrD+PHjsXv3biQkJKCkpARnz55VW+dVVw4ePKjRUmePmjRpEtq2bQs/Pz9VQtsYQ4w1YxtHjxJCoLS0FLW1tRBCoKCgAElJSRg8eDBMTU2RlpbWrDm/RERGzZCP2z3PNH1aMzo6WlhZWQkAokuXLuJf//qXWL9+vbC1tRUARPv27cVXX30ldu3aJdq3by8ACHt7e7Fz504hhBCLFy8WDg4Ows7OTkycOFF8/vnnAoDw8PAQ/fr1ExKJRDg4OIjjx48LIYRYsGCBMDExEQCEra2t+OWXXzR6fwEBAUIqlYqOHTsKMzMzoVAoxNixY0Vubq5avdraWhERESG6du0qpFKpsLe3F35+fuLChQsa1/vrX/+qeu/W1tZi3LhxYvPmzcLJyUkAEFZWVmLMmDEiNjZWdS27du0qcnNzxdatW4VCoRAAROfOncXvv/8uhBDi6NGjok2bNgKA6iWVSkXPnj1FamqqEEKIrKwsteOPvyIiIlQxKpVK8fHHH4s2bdoIGxsb8fLLL4uVK1cKAMLFxUWcOXNGo+v8uMZWezhw4ICQy+UiLCys0bZ79uxRbW3ctm1bMW/ePNWxzz77TDU2hBBi+fLlqutqYmIivLy8xL/+9S8hhHbG2o8//ii6deumuobOzs5i4sSJjcZuTONo7969ok+fPsLKykqYm5urrl3dyg4vvfSSCA0NFXfu3Gl8oDwBV3vQHJ++J33ieNO5ZIkQQugkqzZyycnJeOedd/C8Xt5Zs2YhJSUFd+7cMXQoLRIXF4ecnBxERUWpyh48eIAlS5YgLi4ORUVFsLS0NGCE9Cx4lsbRxIkTAQApKSkGjuTZ8bx/n1PrwvGmcymc80vN9ujSVs+i27dvIzAwEKdPn1YrNzc3h6urK6qqqlBVVdVqkhZqnTiOiIieLZzzSwCA8+fP15uD2NDL39/f0KFqjaWlJaRSKRISEvDHH3+gqqoKN2/exLZt27By5Ur4+/trfT6lMV7n550hxhE9Ww4fPoylS5ciNTUV7u7uqn/n7733Xr26I0aMgFwuh6mpKV544QWcOnXKABE3X2VlJXr06IHly5e3qv5qa2sRFRUFHx+fBo+HhobCy8sLCoUCFhYW8PT0xGeffdbgPP6vv/4a3t7ekMvl6Ny5M6ZPn47bt2+rju/duxcbNmx45m8QPdcMOuviOfY8z9lZunSpMDc3V81PTklJMXRIzZaZmSlef/11oVAohKmpqbC1tRU+Pj4iNjZWVFVVGTo8ekY8S+OIc34115Lv85UrVwpfX1+1Xfw8PDxUc8T37dtXr83BgwfF22+/3ex4DSk4OLjFu1Bqu7/ff/9dDB48WAAQffv2bbDO0KFDRWxsrLhz544oKSkRSUlJQiqVirfeekut3q5duwQAsWHDBlFcXCx+/fVX4e7uLvr166f2bz06OloMHTpUFBUVaRzv85w/tBLc4Y00Fx4ejvDwcEOHoRVDhgzBP//5T0OHQc84jiPNVFRUYPjw4Th+/PgzfY6nWb9+PXbt2oUzZ87U2wEwJiYG7733HgICApCdnQ1bW1sDRak9x48fx2+//daq+jtz5gxCQ0Mxe/Zs1a6IDbGxsUFAQIBqU5pJkyYhNTUVycnJyMvLQ6dOnQAAX3zxBTp06IBFixZBIpGgX79+CA4Oxrx583Dy5EkMHjwYADB//nxcunQJo0aNQmZmZouWNiTt47QHIiLSq4SEBJ3tMqnPczzJxYsXsWLFCqxZs6Ze4gsAPj4+CAoKwo0bN/Dpp58aIELtqqiowKJFi7S2OYO2+uvbty9SU1MxZcoUWFhYNFpv37599XZjbNu2LQCgvLxcVZaXlwdnZ2e19c3rEuOrV6+qtV+9ejVOnz7NDStaISa/RET0REIIREZGomfPnrCwsIC9vT3Gjh2L8+fPq+oEBgbC3NwcTk5OqrK5c+fC2toaEokEhYWFAICgoCAsXLgQubm5kEgk8PT0RExMDGQyGRwdHTFr1iw4OztDJpPBx8dHbR3plpwDeLgrXnPXw9ZUTEwMhBAYM2ZMo3XCwsLQrVs3bNu2DYcPH35if035DOLi4mBtbQ0rKyukp6dj5MiRUCgUcHFxwc6dO9X6q6mpwcqVK+Hq6gpLS0v06dOnRds3h4SEYO7cuVrb3l3b/TXHjRs3YGlpCTc3N1WZu7t7vT+q6ub7Pr5Our29PYYOHYro6Giu3NDKMPklIqInWr16NZYuXYqQkBDk5+cjMzMTeXl5GDJkCP744w8AD5O9x7f/jY2NxZo1a9TKoqOj4evrCw8PDwghcPHiRQQGBmLatGkoLy/H/PnzceXKFZw6dQrV1dV44403kJeX1+JzAP9doaa2tlZ7F6cR+/fvR/fu3WFlZdVoHUtLS3z55ZcwMTHBjBkzUFZW1mjdpnwGc+bMwYIFC1BRUQG5XI6kpCTk5ubC3d0dM2bMQFVVlaq/JUuWYOPGjYiKisKtW7fg6+uLyZMn19shsSl+/PFH5ObmYvLkyRq31Ud/zVFeXo6jR49ixowZMDc3V5UvW7YMt2/fxubNm6FUKpGdnY3o6Gi8+eabqt1NH/WnP/0JN27cwJkzZ/QZPj0Fk18iImpURUUFIiMjMW7cOEydOhW2trbo3bs3tmzZgsLCQq3uYGhmZqa6s+nl5YW4uDgolcp623w31+jRo1FSUoIVK1Zopb/GlJWV4fLly/Dw8Hhq3UGDBmHBggW4cuUKlixZ0mCd5nwGPj4+UCgUaNeuHfz9/VFWVoZr164BeLiCQlxcHPz8/DB+/HjY2dlh+fLlkEqlGl/riooKBAUFIS4uTqN2+uqvucLDw+Hs7FxvG/mhQ4di8eLFCAwMhEKhQK9evaBUKrFt27YG++natSuAh1vDU+vB5JeIiBqVnZ2N0tJSDBgwQK3c29sb5ubm9ba31qYBAwbAyspK7af9Z0F+fj6EEE+86/uosLAwdO/eHbGxsTh27Fi94y39DOruXNbd+b1w4QLKy8vRq1cvVR1LS0s4OTlpfK2XLVuGmTNnomPHjhq101d/zbFnzx4kJyfj22+/hVwuVzsWEhKCrVu34siRIygtLcWlS5fg4+ODQYMGqX6heFTdGKi7O0+tA5NfIiJqVHFxMYCHT8M/zs7ODkqlUqfnt7CwQEFBgU7PoW2VlZUA8MQHrB4lk8mQmJgIiUSCDz/8EBUVFWrHtf0Z1E2vWL58udr64levXlV7uOtpjh07hqysLHz88ccanV9f/TXHrl27sH79emRkZKBLly5qx27duoUNGzZg5syZeO2112BtbQ03NzfEx8fj5s2biIiIqNdf3eY2dWOCWgcmv0RE1Cg7OzsAaDDBKi4uhouLi87OXVVVpfNz6EJdwqPJJgeDBg1CcHAwcnJysHbtWrVj2v4M6h4ii4qKghBC7XXixIkm95OQkIAjR47AxMRElUDX9b1u3TpIJBKN5hBruz9Nbd68GTt27MDRo0fRoUOHesdzcnJQU1NT75hCoYCDgwOys7PrtXnw4AEAcIfHVobJLxERNapXr16wsbGpl3ScPHkSDx48wIsvvqgqMzMzU3uoqqUyMjIghFB7kEjb59AFR0dHSCQS3Lt3T6N2a9euRY8ePfDrr7+qlWvyGTRFp06dIJPJ6m3JranExMR6yXPdXfqQkBAIIepN1dBnf00lhMDixYuRlZWFtLS0Bu+wA1D9kXHr1i21cqVSibt376qWPHtU3Rho3769lqOmlmDyS0REjZLJZFi4cCH27NmDHTt2oKSkBFlZWZg9ezacnZ0REBCgquvp6Ym7d+8iLS0NVVVVKCgoqLf2KQA4ODjg5s2buHLlCpRKpSqZra2tRVFREaqrq3H27FkEBQXB1dUV06ZN08o5Dh48qJelzqysrODu7o7r169r1K5mGuZ3AAAgAElEQVRu+sPj681q8hk09TzTp0/Hzp07ERcXh5KSEtTU1OD69euqxM7f3x/t27fX2vbKrbm/c+fOYePGjYiPj4dUKq233fymTZsAAG5ubhg2bBji4+ORmZmJiooK5OXlqa7/Rx99VK/vujHQu3fvFsdJ2sPkl4iInmjVqlUIDw9HaGgo2rZti6FDh6JLly7IyMiAtbW1qt6cOXMwbNgwvPvuu+jevTvWrl2r+rn30QeCZs+eDUdHR3h5eWHUqFG4e/cugIfzInv37g1LS0sMGTIE3bp1w/fff682d7al59CX0aNHIzs7W23+7jfffANPT0/k5ubC29sbn3zySb12AwcORHBwcL3ypnwGcXFxiIqKAgD06dMHly5dQnx8PBYuXAgAeOutt5CTkwPg4XJwCxYswIYNG9CmTRs4OzsjKCgIRUVFAB7+XJ+fn4/09HStXA9D9ffTTz/h5ZdfRocOHXDy5EmcOXMGzs7OGDx4MDIzMwGgyWvwSiQSpKSkwN/fHx999BHs7e3h5eWFa9euITU1FUOGDKnX5ueff0bHjh3Rp08fzd8k6Y7edlI2Mtybm4haowkTJogJEyYYOox6AgIChIODg6HDaFBzvs9zcnKEmZmZ2L59u46i0q2amhoxZMgQkZCQYBT96UJhYaGQyWRi06ZNGrVj/qBzybzzS0RErYImD4i1dp6enggNDUVoaChKS0sNHY5GampqkJaWBqVSCX9//+e+P11ZvXo1+vXrh8DAQEOHQo9h8ktERKQDS5cuxcSJE+Hv76/xw2+GlJGRgdTUVBw8eLDJaxU/y/3pQmRkJE6fPo0DBw5AKpUaOhx6DJNfIiIyqGXLliExMRH37t2Dm5sbdu/ebeiQtGbdunUIDAzEX/7yF0OH0mTDhw/HV199BScnJ6PoT9vS09Nx//59ZGRkwN7e3tDhUAPMDB0AEREZt/DwcISHhxs6DJ0ZMWIERowYYegwSE/efvttvP3224YOg56Ad36JiIiIyGgw+SUiIiIio8Hkl4iIiIiMBpNfIiIiIjIafOBNx5KTkw0dAhGRSt12q/xuaroTJ04A4DUj/agbb6Q7EiGauK8faSQ5ORnvvPOOocMgIiKiZxDTM51JYfJLRKRndX8c8+uXiEjvUjjnl4iIiIiMBpNfIiIiIjIaTH6JiIiIyGgw+SUiIiIio8Hkl4iIiIiMBpNfIiIiIjIaTH6JiIiIyGgw+SUiIiIio8Hkl4iIiIiMBpNfIiIiIjIaTH6JiIiIyGgw+SUiIiIio8Hkl4iIiIiMBpNfIiIiIjIaTH6JiIiIyGgw+SUiIiIio8Hkl4iIiIiMBpNfIiIiIjIaTH6JiIiIyGgw+SUiIiIio8Hkl4iIiIiMBpNfIiIiIjIaTH6JiIiIyGgw+SUiIiIio8Hkl4iIiIiMBpNfIiIiIjIaTH6JiIiIyGgw+SUiIiIio8Hkl4iIiIiMBpNfIiIiIjIaTH6JiIiIyGgw+SUiIiIio8Hkl4iIiIiMhpmhAyAiep5dv34dH3zwAWpqalRlRUVFkMvlePXVV9Xqdu/eHV988YWeIyQiMi5MfomIdMjFxQVXr15Fbm5uvWM//PCD2n+/8sor+gqLiMhocdoDEZGOvf/++5BKpU+t5+/vr4doiIiMG5NfIiIdmzJlCqqrq59Y54UXXoCXl5eeIiIiMl5MfomIdMzDwwN9+vSBRCJp8LhUKsUHH3yg56iIiIwTk18iIj14//33YWpq2uCx6upqTJw4Uc8REREZJya/RER68O6776K2trZeuYmJCQYOHIguXbroPygiIiPE5JeISA+cnZ0xePBgmJiof+2amJjg/fffN1BURETGh8kvEZGevPfee/XKhBAYN26cAaIhIjJOTH6JiPRkwoQJavN+TU1N8frrr8PR0dGAURERGRcmv0REemJvb4833nhDlQALITB16lQDR0VEZFyY/BIR6dHUqVNVD75JpVKMHTvWwBERERkXJr9ERHo0ZswYWFhYAAB8fX1hY2Nj4IiIiIwLk18iIj2ytrZW3e3llAciIv2TCCGEoYOgZ1dycjLeeecdQ4dBRERGgmkLtVCKmaEjoOdDUlKSoUMgapGoqCgAwIIFC3R+rpqaGiQlJWHy5Mk6P5cunThxAtHR0fz3T3pRN96IWorJL2nFpEmTDB0CUYukpKQA0N9Y9vPzg0wm08u5dCk6Opr//klvmPySNnDOLxGRATwPiS8R0bOIyS8RERERGQ0mv0RERERkNJj8EhEREZHRYPJLREREREaDyS8RkRYdOHAAtra2+Mc//mHoUFq9w4cPY+nSpUhNTYW7uzskEgkkEgnee++9enVHjBgBuVwOU1NTvPDCCzh16pQBIm6+yspK9OjRA8uXL29V/dXW1iIqKgo+Pj4NHg8NDYWXlxcUCgUsLCzg6emJzz77DKWlpfXqfv311/D29oZcLkfnzp0xffp03L59W3V879692LBhA2pqaloUM1FLMfklItIiLsDfNKtWrUJMTAyWLVuG8ePH49KlS/Dw8ECbNm2wY8cO7N+/X63+d999h5SUFPj6+iI7Oxv9+/c3UOTNExISggsXLrSq/nJycvDKK68gODgY5eXlDdY5evQo5s2bhytXrqCwsBDh4eGIjo7GxIkT1eolJSVhypQpmDhxIq5fv4709HRkZmZi5MiRqK6uBvBwa2+ZTIbhw4ejuLi4RbETtQSTXyIiLRo9ejTu3bsHX19fQ4eCioqKRu/oGdL69euxa9cuJCcnQy6Xqx2LiYmBiYkJAgICcO/ePQNFqF3Hjx/Hb7/91qr6O3PmDJYsWYLZs2ejX79+jdazsbFBQEAAHBwcIJfLMWnSJPj5+eHQoUPIy8tT1fviiy/QoUMHLFq0CLa2tujXrx+Cg4Nx+vRpnDx5UlVv/vz56Nu3L0aNGqVKion0jckvEdFzKiEhAfn5+YYOQ83FixexYsUKrFmzpsG1jn18fBAUFIQbN27g008/NUCE2lVRUYFFixZpbXMGbfXXt29fpKamYsqUKbCwsGi03r59+2BqaqpW1rZtWwBQu1ucl5cHZ2dnSCQSVVmnTp0AAFevXlVrv3r1apw+fZobVpDBMPklItKSY8eOwdXVFRKJBJ9//jkAIC4uDtbW1rCyskJ6ejpGjhwJhUIBFxcX7Ny5U9U2JiYGMpkMjo6OmDVrFpydnSGTyeDj46N25ywwMBDm5uZwcnJSlc2dOxfW1taQSCQoLCwEAAQFBWHhwoXIzc2FRCKBp6cnAODQoUNQKBRYt26dPi5JPTExMRBCYMyYMY3WCQsLQ7du3bBt2zYcPnz4if0JIRAZGYmePXvCwsIC9vb2GDt2LM6fP6+q09TPAHi49fTKlSvh6uoKS0tL9OnTp0XbN4eEhGDu3Llo165ds/vQZX/NcePGDVhaWsLNzU1V5u7uXu8Prbr5vu7u7mrl9vb2GDp0KKKjozlNiAyCyS8RkZa8/PLLOH78uFrZnDlzsGDBAlRUVEAulyMpKQm5ublwd3fHjBkzUFVVBeBhUjtt2jSUl5dj/vz5uHLlCk6dOoXq6mq88cYbqp+YY2Ji6m0nHBsbizVr1qiVRUdHw9fXFx4eHhBC4OLFiwCgetiotrZWJ9fgafbv34/u3bvDysqq0TqWlpb48ssvYWJighkzZqCsrKzRuqtXr8bSpUsREhKC/Px8ZGZmIi8vD0OGDMEff/wBoOmfAQAsWbIEGzduRFRUFG7dugVfX19MnjwZv/zyi8bv9ccff0Rubi4mT56scVt99Ncc5eXlOHr0KGbMmAFzc3NV+bJly3D79m1s3rwZSqUS2dnZiI6OxptvvomBAwfW6+dPf/oTbty4gTNnzugzfCIATH6JiPTGx8cHCoUC7dq1g7+/P8rKynDt2jW1OmZmZqq7mF5eXoiLi4NSqURiYqJWYhg9ejRKSkqwYsUKrfSnibKyMly+fBkeHh5PrTto0CAsWLAAV65cwZIlSxqsU1FRgcjISIwbNw5Tp06Fra0tevfujS1btqCwsBBbt26t1+ZJn0FlZSXi4uLg5+eH8ePHw87ODsuXL4dUKtX4+ldUVCAoKAhxcXEatdNXf80VHh4OZ2dnhIWFqZUPHToUixcvRmBgIBQKBXr16gWlUolt27Y12E/Xrl0BAFlZWTqPmehxTH6JiAyg7q7Zo3cdGzJgwABYWVmp/Yz/rMrPz4cQ4ol3fR8VFhaG7t27IzY2FseOHat3PDs7G6WlpRgwYIBaube3N8zNzdWmizTk8c/gwoULKC8vR69evVR1LC0t4eTkpPH1X7ZsGWbOnImOHTtq1E5f/TXHnj17kJycjG+//bbeg4ohISHYunUrjhw5gtLSUly6dAk+Pj4YNGiQ2oNxderGQN3deSJ9YvJLRNTKWVhYoKCgwNBhtFhlZSUAPPEBq0fJZDIkJiZCIpHgww8/REVFhdrxuuWybGxs6rW1s7ODUqnUKL666RXLly9XrTkskUhw9erVRpcCa8ixY8eQlZWFjz/+WKPz66u/5ti1axfWr1+PjIwMdOnSRe3YrVu3sGHDBsycOROvvfYarK2t4ebmhvj4eNy8eRMRERH1+rO0tATw3zFBpE9MfomIWrGqqioUFxfDxcXF0KG0WF3Co8kmB4MGDUJwcDBycnKwdu1atWN2dnYA0GCS25xrVvcQWVRUFIQQaq8TJ040uZ+EhAQcOXIEJiYmqgS6ru9169ZBIpFoNIdY2/1pavPmzdixYweOHj2KDh061Duek5ODmpqaescUCgUcHByQnZ1dr82DBw8A/HdMEOkTk18iolYsIyMDQgi1h4bMzMyeOl2iNXJ0dIREItF4/d61a9eiR48e+PXXX9XKe/XqBRsbm3qJ38mTJ/HgwQO8+OKLGp2nU6dOkMlkOH36tEbtHpeYmFgvea67cx8SEgIhRL2pGvrsr6mEEFi8eDGysrKQlpbW4B12AKo/Mm7duqVWrlQqcffuXdWSZ4+qGwPt27fXctRET8fkl4ioFamtrUVRURGqq6tx9uxZBAUFwdXVFdOmTVPV8fT0xN27d5GWloaqqioUFBTUW0sVABwcHHDz5k1cuXIFSqUSVVVVOHjwoMGWOrOysoK7uzuuX7+uUbu66Q+Przcrk8mwcOFC7NmzBzt27EBJSQmysrIwe/ZsODs7IyAgQOPzTJ8+HTt37kRcXBxKSkpQU1OD69evqxI7f39/tG/fXmvbK7fm/s6dO4eNGzciPj4eUqlUbSqIRCLBpk2bAABubm4YNmwY4uPjkZmZiYqKCuTl5amu/0cffVSv77ox0Lt37xbHSaQpJr9ERFry+eefw9vbGwCwePFivP3224iLi0NUVBQAoE+fPrh06RLi4+OxcOFCAMBbb72FnJwcVR+VlZXo3bs3LC0tMWTIEHTr1g3ff/+92jzZOXPmYNiwYXj33XfRvXt3rF27VvXz8aMPGM2ePRuOjo7w8vLCqFGjcPfuXb1chycZPXo0srOz1ebvfvPNN/D09ERubi68vb3xySef1Gs3cOBABAcH1ytftWoVwsPDERoairZt22Lo0KHo0qULMjIyYG1tDQAafQbR0dFYsGABNmzYgDZt2sDZ2RlBQUEoKioC8PDn+vz8fKSnp2vlehiqv59++gkvv/wyOnTogJMnT+LMmTNwdnbG4MGDkZmZCaDpW3VLJBKkpKTA398fH330Eezt7eHl5YVr164hNTUVQ4YMqdfm559/RseOHdGnTx/N3yRRSwmiFkhKShIcRvQ8mDBhgpgwYYJBYwgICBAODg4GjUETzfn3n5OTI8zMzMT27dt1FJVu1dTUiCFDhoiEhASj6E8XCgsLhUwmE5s2bdKoHf9/Q1qSzDu/REStiCYPgz2LPD09ERoaitDQUJSWlho6HI3U1NQgLS0NSqUS/v7+z31/urJ69Wr069cPgYGBhg6FjBSTXzK4jz/+GHK5HBKJpMUPmjxvKisr0aNHDyxfvlzjtqmpqXB3d683T8/c3ByOjo549dVXERERofo5l0hfli5diokTJ8Lf31/jh98MKSMjA6mpqTh48GCT1yp+lvvThcjISJw+fRoHDhyAVCo1dDhkpJj8ksFt27YN8fHxhg6jVQoJCcGFCxea1Xb8+PG4dOkSPDw8YGtrCyEEamtrkZ+fj+TkZLi5uWHx4sV44YUXdLpMEjXNsmXLkJiYiHv37sHNzQ27d+82dEg6tW7dOgQGBuIvf/mLoUNpsuHDh+Orr76Ck5OTUfSnbenp6bh//z4yMjJgb29v6HDIiDH5JdKyiooK+Pj4tLif48eP47ffftNCRP8lkUhgZ2eHV199FYmJiUhOTsYff/yB0aNHP1N34BqjrWtvCOHh4bh//z6EELh8+TImTJhg6JB0bsSIEVi/fr2hwyA9efvtt7F06dJ6q3YQ6RuTX2oVJBKJoUPQmoSEBOTn57eoj4qKCixatAjR0dFaiqphEyZMwLRp05Cfn48tW7bo9Fz6oI1rT0REzzcmv6R3QghERESge/fusLCwgK2tLRYtWqRWZ+PGjbCysoJcLkd+fj4WLlyIjh074sKFCxBCIDIyEj179oSFhQXs7e0xduxYnD9/XtU+JiYGMpkMjo6OmDVrFpydnSGTyeDj44OTJ0/Wi+dp/QUGBsLc3Fzt58S5c+fC2toaEokEhYWFAICgoCAsXLgQubm5kEgk8PT0bNY1CgkJwdy5c1W7OD3u0KFDWlurtW792IMHDwLgtSciouecAZeaoOdAc5aeCQkJERKJRPz1r38VRUVFory8XMTGxgoA4tdff1WrB0DMnz9fbN68WYwbN0785z//EStXrhTm5uZi+/btori4WJw9e1b0799ftG3bVty+fVvVPiAgQFhbW4tz586JyspKkZ2dLby9vYVcLhfXrl1T1Wtqf1OmTBHt27dXey8RERECgCgoKFCVjR8/Xnh4eGh0TR517NgxMWbMGCGEEAUFBQKACAkJUauzb98+IZfLRWho6FP78/DwELa2to0eLykpEQBEp06dVGXGeO1bw1JnzxouPUX6xPFGWsKlzki/KioqEBUVhddffx3BwcGws7ODpaUlHBwcGm2zfv16zJs3D6mpqejcuTMiIyMxbtw4TJ06Fba2tujduze2bNmCwsJCbN26Va2tmZmZ6q6il5cX4uLioFQqkZiYqIpHk/50raKiAkFBQYiLi3tivdGjR6OkpAQrVqxo8TnrVtpQKpX1jhnTtSciIuNgZugAyLhcvHgR5eXlGD58eLPaZ2dno7S0tN4+9t7e3jA3N6/3s/rjBgwYACsrK9XP6i3tT9uWLVuGmTNnomPHjno7Z1lZGYQQUCgUT6z3vF974OGWq8nJyXo/77PqxIkTAMBrRnpRN96IWorJL+lV3X7ujc1lfZri4mIAgI2NTb1jdnZ2Dd69fJyFhQUKCgq01p+2HDt2DFlZWYiMjNTbOQHg999/BwD06NHjifWe52tf56effsI777yj9/M+63jNiOhZwmkPpFcymQwAcP/+/Wa1t7OzA4AGE6Pi4mK4uLg8sX1VVZVavZb2p00JCQk4cuQITExMVBtS1P2RsG7dOkgkEp2sx3vo0CEAwMiRI59Y73m+9nUmTJgAIQRfTXwlJSUBgMHj4Ms4XnXjjailmPySXvXq1QsmJib44Ycfmt3exsamXhJ48uRJPHjwAC+++OIT22dkZEAIgYEDB2rcn5mZGaqqqpoVd1MkJibW+7Kvu0saEhICIUS9KQItdfv2bURFRcHFxQUffvjhE+s+z9eeiIiMB5Nf0qt27dph/Pjx2L17NxISElBSUoKzZ882+eEmmUyGhQsXYs+ePdixYwdKSkqQlZWF2bNnw9nZGQEBAWr1a2trUVRUhOrqapw9exZBQUFwdXVVLe+lSX+enp64e/cu0tLSUFVVhYKCAly9erVejA4ODrh58yauXLkCpVKpk6Tt4MGDGi11JoRAaWkpamtrVUl1UlISBg8eDFNTU6SlpT11zi+vPRERPRcEUQs0Z+kZpVIpPv74Y9GmTRthY2MjXn75ZbFy5UoBQLi4uIgzZ86IDRs2CEtLS9USXNu3b1e1r62tFREREaJr165CKpUKe3t74efnJy5cuKB2noCAACGVSkXHjh2FmZmZUCgUYuzYsSI3N1etXlP7u3Pnjhg2bJiQyWTCzc1NfPLJJ2LRokUCgPD09FQt4XXq1CnRuXNnYWlpKV5++WW1Jbs01dhSZwcOHBByuVyEhYU12nbv3r2iT58+wsrKSpibmwsTExMBQEgkEmFnZydeeuklERoaKu7cuaPWzlivPZc60xyXniJ94ngjLUmWCCGEoRJvevYlJyfjnXfeQWscRrNmzUJKSgru3Llj6FCMzrN47SdOnAgASElJMXAkz47W/O+fnj8cb6QlKZz2QM+1mpoaQ4dgtHjtiYioNWLyS6RD58+fV63c8KSXv7+/oUMlIiIyCkx+6bm0bNkyJCYm4t69e3Bzc8Pu3bsNEkePHj2atITPrl27DBKfLrSWa0+t3+HDh7F06VKkpqbC3d1d9cfge++9V6/uiBEjIJfLYWpqihdeeAGnTp0yQMTNV1lZiR49emD58uWtqr/a2lpERUXBx8enweOhoaHw8vKCQqGAhYUFPD098dlnn6G0tLRe3a+//hre3t6Qy+Xo3Lkzpk+fjtu3b6uO7927Fxs2bOCvQmRwTH7puRQeHo779+9DCIHLly9jwoQJhg7JaPDaU1OsWrUKMTExWLZsGcaPH49Lly7Bw8MDbdq0wY4dO7B//361+t999x1SUlLg6+uL7Oxs9O/f30CRN09ISAguXLjQqvrLycnBK6+8guDgYJSXlzdY5+jRo5g3bx6uXLmCwsJChIeHIzo6WjVHvk5SUhKmTJmCiRMn4vr160hPT0dmZiZGjhyJ6upqAMCYMWMgk8kwfPhw1SY3RIbA5JeIqJWoqKho9A7cs3SOp1m/fj127dqF5ORkyOVytWMxMTEwMTFBQEAA7t27Z6AItev48eP47bffWlV/Z86cwZIlSzB79mz069ev0Xo2NjYICAiAg4MD5HI5Jk2aBD8/Pxw6dAh5eXmqel988QU6dOiARYsWwdbWFv369UNwcDBOnz6ttlX5/Pnz0bdvX4waNUqVFBPpG5NfIqJWIiEhAfn5+c/8OZ7k4sWLWLFiBdasWaPa8fFRPj4+CAoKwo0bN/Dpp58aIELtqqiowKJFixAdHd2q+uvbty9SU1MxZcoUWFhYNFpv3759MDU1VStr27YtAKjdLc7Ly4OzszMkEomqrFOnTgBQb03u1atX4/Tp01q7JkSaYvJLRNRMQghERkaiZ8+esLCwgL29PcaOHYvz58+r6gQGBsLc3BxOTk6qsrlz58La2hoSiQSFhYUAgKCgICxcuBC5ubmQSCTw9PRETEwMZDIZHB0dMWvWLDg7O0Mmk8HHx0ftblpLzgE83OJak01TWiImJgZCCIwZM6bROmFhYejWrRu2bduGw4cPP7G/pnwGcXFxsLa2hpWVFdLT0zFy5EgoFAq4uLhg586dav3V1NRg5cqVcHV1haWlJfr06dOibXVDQkIwd+5c1VblLaXt/prjxo0bsLS0hJubm6rM3d293h9VdfN93d3d1crt7e0xdOhQREdHc9kyMggmv0REzbR69WosXboUISEhyM/PR2ZmJvLy8jBkyBD88ccfAB4me5MmTVJrFxsbizVr1qiVRUdHw9fXFx4eHhBC4OLFiwgMDMS0adNQXl6O+fPn48qVKzh16hSqq6vxxhtvqH52bsk5gP8uS1dbW6u9i9OI/fv3o3v37rCysmq0jqWlJb788kuYmJhgxowZKCsra7RuUz6DOXPmYMGCBaioqIBcLkdSUhJyc3Ph7u6OGTNmqO0EuGTJEmzcuBFRUVG4desWfH19MXny5HrbcDfFjz/+iNzcXEyePFnjtvrorznKy8tx9OhRzJgxA+bm5qryZcuW4fbt29i8eTOUSiWys7MRHR2NN998U7Wl+aP+9Kc/4caNGzhz5ow+wycCwOSXiKhZKioqEBkZiXHjxmHq1KmwtbVF7969sWXLFhQWFjZ5y+6mMDMzU93Z9PLyQlxcHJRKJRITE7XS/+jRo1FSUoIVK1Zopb/GlJWV4fLly/Dw8Hhq3UGDBmHBggW4cuUKlixZ0mCd5nwGPj4+UCgUaNeuHfz9/VFWVoZr164BeLiCQlxcHPz8/DB+/HjY2dlh+fLlkEqlGl/riooKBAUFIS4uTqN2+uqvucLDw+Hs7IywsDC18qFDh2Lx4sUIDAyEQqFAr169oFQqsW3btgb76dq1KwAgKytL5zETPY7JLxFRM2RnZ6O0tBQDBgxQK/f29oa5ubnatARtGzBgAKysrNR+2n8W5OfnQwjxxLu+jwoLC0P37t0RGxuLY8eO1Tve0s+g7s5l3Z3fCxcuoLy8HL3+P3t3Hld1lf8P/HVZL9tlUVkEUZZcENSpKEUdNctSUnPBcGnGLQmbCLVCxVxQyW2EHybTiA49RlMBMShT66FGjJM6WS6EkyKuuAQKsiPb+f3hlzteL+C9cBfgvp6Ph394Pudzzvuez+cDbz73fM7H11dex8LCAs7OzmqP9dKlSzFv3jy4urqqtZ+u2muJ/fv3Izk5Gd9++63Sg4qRkZHYtm0bjh49irKyMly5cgUBAQEYNGiQwoNxDRrOgYa780S6xOSXiKgFGpZqsra2VtpmZ2eH0tJSrfZvbm6OgoICrfahaVVVVQDQ7ANWj5NKpUhMTIREIsHs2bNRWVmpsF3Tx6BhesWyZcsUXkJz/fr1JpcCa8zx48eRlZWFuXPnqtW/rtprib1792LdunXIyMhAjx49FLbduXMH69evx7x58/DSSy/BysoKHh4eSEhIwO3bt7Fx40al9iwsLAD875wg0iUmv0RELWBnZwcAjSZYDx48gJubm9b6rqmp0Xof2u0r+DQAACAASURBVNCQ8KjzkoNBgwZh4cKFyMnJwerVqxW2afoYNDxEFhMTo/QimhMnTqjczo4dO3D06FEYGRnJE+iGtteuXQuJRKLWHGJNt6euLVu2YNeuXTh27Bi6du2qtD0nJwd1dXVK22QyGRwcHJCdna20T3V1NYD/nRNEusTkl4ioBXx9fWFtba2UdJw6dQrV1dV47rnn5GUmJiYKD1W1VkZGBoQQCg8SaboPbXB0dIREIlF7/d7Vq1ejd+/eOHPmjEK5OsdAFd26dYNUKsXZs2fV2u9JiYmJSslzw136yMhICCGUpmrosj1VCSEQERGBrKwspKWlNXqHHYD8j4w7d+4olJeWlqKwsFC+5NnjGs4BJycnDUdN9HRMfomIWkAqlWLRokXYv38/du3ahZKSEmRlZSE0NBQuLi4ICQmR1/X29kZhYSHS0tJQU1ODgoICpbVPAcDBwQG3b9/GtWvXUFpaKk9m6+vrUVRUhNraWpw/fx7h4eFwd3fHzJkzNdLHoUOHdLLUmaWlJTw9PZGXl6fWfg3TH55cb1adY6BqP7NmzcKePXsQHx+PkpIS1NXVIS8vT57YBQcHw8nJSWOvV27L7V24cAEbNmxAQkICTE1NFaaCSCQSbNq0CQDg4eGBESNGICEhAZmZmaisrMTNmzfl4z9nzhylthvOAT8/v1bHSaQuJr9ERC20YsUKREdHIyoqCp07d8awYcPQo0cPZGRkwMrKSl5v/vz5GDFiBKZOnYpevXph9erV8q97H38gKDQ0FI6OjvDx8cGYMWNQWFgI4NG8SD8/P1hYWGDo0KHo2bMnvv/+e4W5s63tQ1cCAwORnZ2tMH/3yy+/hLe3N3Jzc+Hv74/33ntPab+BAwdi4cKFSuWqHIP4+HjExMQAAPr164crV64gISEBixYtAgC89tpryMnJAfBoObgFCxZg/fr16NSpE1xcXBAeHo6ioiIAj76uz8/PR3p6ukbGQ1/tnTx5EkOGDEHXrl1x6tQpnDt3Di4uLhg8eDAyMzMBQOU1eCUSCVJSUhAcHIw5c+bA3t4ePj4+uHHjBlJTUzF06FClfX766Se4urqiX79+6n9IotYSRK2QlJQkeBpRRzB58mQxefJkfYehJCQkRDg4OOg7jEa15PrPyckRJiYmYufOnVqKSrvq6urE0KFDxY4dOwyiPW24d++ekEqlYtOmTWrtx983pCHJvPNLRNTGqfOAWFvn7e2NqKgoREVFoaysTN/hqKWurg5paWkoLS1FcHBwh29PW1auXIkBAwYgLCxM36GQgWLyS0REOrVkyRIEBQUhODhY7Yff9CkjIwOpqak4dOiQymsVt+f2tGHz5s04e/YsDh48CFNTU32HQwaKyS8RURu1dOlSJCYmori4GB4eHti3b5++Q9KYtWvXIiwsDJ988om+Q1HZyJEj8cUXX8DZ2dkg2tO09PR0PHz4EBkZGbC3t9d3OGTATPQdABERNS46OhrR0dH6DkNrRo0ahVGjRuk7DNKR8ePHY/z48foOg4h3fomIiIjIcDD5JSIiIiKDweSXiIiIiAwGk18iIiIiMhh84I00IigoSN8hELXKyZMnAfBcVkfDK2o5ZqQL6r4Wm6gpEiFUfH8hUSNOnDiBzZs36zsMonbl7t27OHPmDEaPHq3vUIjanZSUFH2HQO1bCpNfIiIdS05Oxptvvgn++CUi0rkUzvklIiIiIoPB5JeIiIiIDAaTXyIiIiIyGEx+iYiIiMhgMPklIiIiIoPB5JeIiIiIDAaTXyIiIiIyGEx+iYiIiMhgMPklIiIiIoPB5JeIiIiIDAaTXyIiIiIyGEx+iYiIiMhgMPklIiIiIoPB5JeIiIiIDAaTXyIiIiIyGEx+iYiIiMhgMPklIiIiIoPB5JeIiIiIDAaTXyIiIiIyGEx+iYiIiMhgMPklIiIiIoPB5JeIiIiIDAaTXyIiIiIyGEx+iYiIiMhgMPklIiIiIoPB5JeIiIiIDAaTXyIiIiIyGEx+iYiIiMhgMPklIiIiIoPB5JeIiIiIDAaTXyIiIiIyGEx+iYiIiMhgmOg7ACKijqympgZlZWUKZeXl5QCAoqIihXKJRAI7OzudxUZEZIiY/BIRaVFhYSFcXV1RV1entM3BwUHh/yNGjMCxY8d0FRoRkUHitAciIi1ycnLCH//4RxgZNf/jViKRYOrUqTqKiojIcDH5JSLSsrfeeuupdYyNjTFx4kQdRENEZNiY/BIRadmkSZNgYtL0LDNjY2O89tpr6NSpkw6jIiIyTEx+iYi0TCaTYfTo0U0mwEIIzJgxQ8dREREZJia/REQ6MGPGjEYfegMAMzMzvP766zqOiIjIMDH5JSLSgddffx2WlpZK5aamppgwYQKsrKz0EBURkeFh8ktEpANSqRQTJ06EqampQnlNTQ2mT5+up6iIiAwPk18iIh2ZNm0aampqFMpkMhleeeUVPUVERGR4mPwSEenIyy+/rPBiC1NTU0ydOhVmZmZ6jIqIyLAw+SUi0hETExNMnTpVPvWhpqYG06ZN03NURESGhckvEZEOTZ06VT71wcnJCUOGDNFzREREhoXJLxGRDgUEBMDV1RUA8Kc//emprz0mIiLNUlpxPS8vDz/++KM+YiEiMgj+/v64desWOnXqhOTkZH2HQ0TUYU2ZMkWpTCKEEI8XJCcn480339RZUERERERE2vBEmgsAKU2+bL6RykREpCH79u3D5MmT9R0GtVBQUBAAICUlRc+RtB8NN9eYX5AuNHczl5PNiIj0gIkvEZF+MPklIiIiIoPB5JeIiIiIDAaTXyIiIiIyGEx+iYiIiMhgMPklIiIiIoPB5JeIiEhPDh48CFtbW3z99df6DqXNO3LkCJYsWYLU1FR4enpCIpFAIpHgrbfeUqo7atQo2NjYwNjYGH379sUvv/yih4hbrqqqCr1798ayZcvaVHv19fWIiYlBQEBAo9ujoqLg4+MDmUwGc3NzeHt746OPPkJZWZlS3d27d8Pf3x82Njbo3r07Zs2ahbt378q3f/XVV1i/fj3q6upaFXNjmPwSERHpCde8Vc2KFSsQFxeHpUuXYtKkSbhy5Qq8vLzQqVMn7Nq1C998841C/e+++w4pKSkYO3YssrOz8eyzz+op8paJjIzExYsX21R7OTk5+OMf/4iFCxeioqKi0TrHjh3DX/7yF1y7dg337t1DdHQ0YmNj5etiN0hKSsL06dMRFBSEvLw8pKenIzMzE6NHj0ZtbS0AYNy4cZBKpRg5ciQePHjQqtifxOSXiIhITwIDA1FcXIyxY8fqOxRUVlY2eUdPn9atW4e9e/ciOTkZNjY2Ctvi4uJgZGSEkJAQFBcX6ylCzfrxxx/x66+/tqn2zp07h8WLFyM0NBQDBgxosp61tTVCQkLg4OAAGxsbTJkyBRMmTMDhw4dx8+ZNeb2///3v6Nq1Kz788EPY2tpiwIABWLhwIc6ePYtTp07J673//vvo378/xowZI0+KNYHJLxEREWHHjh3Iz8/XdxgKLl++jI8//hirVq2CVCpV2h4QEIDw8HDcunULH3zwgR4i1KzKykp8+OGHiI2NbVPt9e/fH6mpqZg+fTrMzc2brHfgwAEYGxsrlHXu3BkAFO4W37x5Ey4uLpBIJPKybt26AQCuX7+usP/KlStx9uxZjY0JwOSXiIhIL44fPw53d3dIJBJ8+umnAID4+HhYWVnB0tIS6enpGD16NGQyGdzc3LBnzx75vnFxcZBKpXB0dMQ777wDFxcXSKVSBAQEKNw5CwsLg5mZGZydneVl7777LqysrCCRSHDv3j0AQHh4OBYtWoTc3FxIJBJ4e3sDAA4fPgyZTIa1a9fqYkiUxMXFQQiBcePGNVlnzZo16NmzJ7Zv344jR440254QAps3b0afPn1gbm4Oe3t7vPHGG/jtt9/kdVQ9BgBQV1eH5cuXw93dHRYWFujXrx+SkpJa/HkjIyPx7rvvokuXLi1uQ5vttcStW7dgYWEBDw8PeZmnp6fSH1oN8309PT0Vyu3t7TFs2DDExsZqbJoQk18iIiI9GDJkCH788UeFsvnz52PBggWorKyEjY0NkpKSkJubC09PT7z99tuoqakB8CipnTlzJioqKvD+++/j2rVr+OWXX1BbW4tXXnlF/hVzXFwcpkyZotDH1q1bsWrVKoWy2NhYjB07Fl5eXhBC4PLlywAgf9iovr5eK2PwNN988w169eoFS0vLJutYWFjg888/h5GREd5++22Ul5c3WXflypVYsmQJIiMjkZ+fj8zMTNy8eRNDhw7F77//DkD1YwAAixcvxoYNGxATE4M7d+5g7NixmDZtGk6fPq32Z/33v/+N3NxcTJs2Te19ddFeS1RUVODYsWN4++23YWZmJi9funQp7t69iy1btqC0tBTZ2dmIjY3Fq6++ioEDByq184c//AG3bt3CuXPnNBIXk18iIqI2KCAgADKZDF26dEFwcDDKy8tx48YNhTomJibyu5g+Pj6Ij49HaWkpEhMTNRJDYGAgSkpK8PHHH2ukPXWUl5fj6tWr8PLyemrdQYMGYcGCBbh27RoWL17caJ3Kykps3rwZEydOxIwZM2Braws/Pz989tlnuHfvHrZt26a0T3PHoKqqCvHx8ZgwYQImTZoEOzs7LFu2DKampmqPf2VlJcLDwxEfH6/Wfrpqr6Wio6Ph4uKCNWvWKJQPGzYMERERCAsLg0wmg6+vL0pLS7F9+/ZG23nmmWcAAFlZWRqJi8kvERFRG9dw1+zxu46Nef7552FpaanwNX57lZ+fDyFEs3d9H7dmzRr06tULW7duxfHjx5W2Z2dno6ysDM8//7xCub+/P8zMzBSmizTmyWNw8eJFVFRUwNfXV17HwsICzs7Oao//0qVLMW/ePLi6uqq1n67aa4n9+/cjOTkZ3377rdKDipGRkdi2bRuOHj2KsrIyXLlyBQEBARg0aJDCg3ENGs6BhrvzrcXkl4iIqAMxNzdHQUGBvsNotaqqKgBo9gGrx0mlUiQmJkIikWD27NmorKxU2N6wXJa1tbXSvnZ2digtLVUrvobpFcuWLZOvOSyRSHD9+vUmlwJrzPHjx5GVlYW5c+eq1b+u2muJvXv3Yt26dcjIyECPHj0Utt25cwfr16/HvHnz8NJLL8HKygoeHh5ISEjA7du3sXHjRqX2LCwsAPzvnGgtJr9EREQdRE1NDR48eAA3Nzd9h9JqDQmPOi85GDRoEBYuXIicnBysXr1aYZudnR0ANJrktmTMGh4ii4mJgRBC4d+JEydUbmfHjh04evQojIyM5Al0Q9tr166FRCJRaw6xpttT15YtW7Br1y4cO3YMXbt2Vdqek5ODuro6pW0ymQwODg7Izs5W2qe6uhrA/86J1mLyS0RE1EFkZGRACKHw0JCJiclTp0u0RY6OjpBIJGqv37t69Wr07t0bZ86cUSj39fWFtbW1UuJ36tQpVFdX47nnnlOrn27dukEqleLs2bNq7fekxMREpeS54c59ZGQkhBBKUzV02Z6qhBCIiIhAVlYW0tLSGr3DDkD+R8adO3cUyktLS1FYWChf8uxxDeeAk5OTRmJl8ktERNRO1dfXo6ioCLW1tTh//jzCw8Ph7u6OmTNnyut4e3ujsLAQaWlpqKmpQUFBgdJaqgDg4OCA27dv49q1aygtLUVNTQ0OHTqkt6XOLC0t4enpiby8PLX2a5j+8OR6s1KpFIsWLcL+/fuxa9culJSUICsrC6GhoXBxcUFISIja/cyaNQt79uxBfHw8SkpKUFdXh7y8PHliFxwcDCcnJ429Xrktt3fhwgVs2LABCQkJMDU1VZgKIpFIsGnTJgCAh4cHRowYgYSEBGRmZqKyshI3b96Uj/+cOXOU2m44B/z8/FodJ8Dkl4iISC8+/fRT+Pv7AwAiIiIwfvx4xMfHIyYmBgDQr18/XLlyBQkJCVi0aBEA4LXXXkNOTo68jaqqKvj5+cHCwgJDhw5Fz5498f333yvMk50/fz5GjBiBqVOnolevXli9erX86+PHHzAKDQ2Fo6MjfHx8MGbMGBQWFupkHJoTGBiI7Oxshfm7X375Jby9vZGbmwt/f3+89957SvsNHDgQCxcuVCpfsWIFoqOjERUVhc6dO2PYsGHo0aMHMjIyYGVlBQBqHYPY2FgsWLAA69evR6dOneDi4oLw8HAUFRUBePR1fX5+PtLT0zUyHvpq7+TJkxgyZAi6du2KU6dO4dy5c3BxccHgwYORmZkJQPVXdUskEqSkpCA4OBhz5syBvb09fHx8cOPGDaSmpmLo0KFK+/z0009wdXVFv3791P+QjRFPSEpKEo0UExER0f+ZPHmymDx5sl5jCAkJEQ4ODnqNQR0tyS9ycnKEiYmJ2Llzp5ai0q66ujoxdOhQsWPHDoNoTxvu3bsnpFKp2LRpk1r7NXO+JfPOLxERUTulzsNg7ZG3tzeioqIQFRWFsrIyfYejlrq6OqSlpaG0tBTBwcEdvj1tWblyJQYMGICwsDCNtamz5HfTpk3yyeufffaZWvv6+/vD2NgYAwYM0FsM1DY0dQwPHjwIW1tbfP3111rrOyoqCj4+PpDJZDA3N4e3tzc++uijp/5Anjt3LmxsbCCRSFr1YER9fT1iYmIQEBDQ4jYed+nSJbz33nvo27cvZDIZzMzM0KVLF/Tu3RsTJ07El19+Ka/bnq7f3bt3QyKRaGyc2qrU1FR4enoqzKkzNTWFq6srpk+fjv/+979a67utX4eNjY1EIoGZmRkcHR0xfPhwbNy4Uf7VNLVtS5YsQVBQEIKDg9V++E2fMjIykJqaikOHDqm8VnF7bk8bNm/ejLNnz+LgwYMwNTXVXMNq3CZutZycHAFA/O1vf1N735EjR4r+/fvrNQZqGxo7hgcOHBAymUx89dVXWut32LBhYuvWreL+/fuipKREJCUlCVNTU/Haa689dd89e/YIAOLMmTMt6vvSpUti8ODBAoBGroPExERhZmYmhgwZIg4fPiyKiopEVVWVyM3NFV9//bUIDAwUISEhCvu0l+s3MDBQeHl5CQAiJyen1X22dV5eXsLW1lYIIURZWZn46quvhLu7u7C2tha//fab1vptD9fh42NTX18vioqKxPfffy9mzpwpJBKJcHFxET/99FOL4tD3tIclS5YIMzMzAUD06NFDpKSk6C0WVbU2v/j2229FRESEBiOitiwtLU1ER0eL2traFu3fYaY9SCQSfYdAbVRgYCCKi4sxduxYrfVhbW2NkJAQODg4wMbGBlOmTMGECRNw+PDhRt9Ioynnzp3D4sWLERoa2uq7p8CjBxfmzp2LgIAAfP/993j11VdhZ2cHc3NzeHp64vXXX0dcXJwGIleki+v3/v37uHDhAlatWgUA+Oc//6n1PtsSKysrjB07Fv/v//0/lJWVYcuWLTrtvy1fhxKJBHZ2dhg+fDgSExORnJyM33//XR5zexMdHY2HDx9CCIGrV69i8uTJ+g5J60aNGoV169bpOwzSkfHjx2PJkiVKq3ZogtaSXyEEUlJSGn1Xdktp9Ja3gdDGcTAEjY3bgQMHlC7Czp07A8BT3+bTmsSvf//+SE1NxfTp01V+01Fz1q5di7q6OnzyyScwMTFptI6np6fGpwbp4vpNTk5GYGAgxo0bB6lUip07d6r8BLK+afJafeGFFwAAv/76a6vb0idNX4ePmzx5MmbOnIn8/HxOgyMyMBpJfuvq6hAdHY1evXrBwsICnTt3hoeHB6KjozFlypRm9xVCYPPmzejTpw/Mzc1hb2+PN954o9H3Yl++fBm9e/eGlZWVfFmXJ9/f/a9//Qs+Pj6wtbWFVCqFn58fvv32W018zGbb7tOnDyQSCYyMjPDcc8/Jfwh/9NFH8vqff/45gEfjtXz5cri7u8PCwgL9+vVDUlISAGDDhg2wtLSEjY0N8vPzsWjRIri6uuLixYtP/WyqHofm+ldVXFwcpFIpHB0d8c4778DFxQVSqRQBAQFK70dX9Rircy487vjx43B3d4dEIsGnn34K4NFSNVZWVrC0tER6ejpGjx4NmUwGNzc37NmzR2H/1py/t27dgoWFBTw8PBQ+x8aNG9GrVy+Ym5vD1tYWH3744VPHtLUOHz781PU4q6urceTIETg4OCgsgt9Sbe363b17NyZOnAgbGxuMGjUK165dw7/+9S+FOoZwrdbW1gJQfC2soV2HqmhYC/fQoUNq7UdE7ZwacySatHbtWmFsbCzS09NFRUWF+Pnnn4WTk5MYPny4Qr3G5ogtX75cmJmZiZ07d4oHDx6I8+fPi2effVZ07txZ3L17V15v5MiRwtPTU1y9elXU1NSIX3/9Vbz44otCKpWKS5cuyeulpKSIlStXisLCQnH//n0xcOBA0alTp2ZjUFVzbdfW1ooePXoId3d3pfkpCxYsEDExMfL/f/DBB8Lc3Fzs27dPFBUViaVLlwojIyP53LPIyEgBQLz//vtiy5YtYuLEieK///3vUz+bqsfhaf2rKiQkRFhZWYkLFy6IqqoqkZ2dLfz9/YWNjY24ceOGvJ6qx1jVeo0dw5s3bwoAYsuWLfKyhnE8evSoKC4uFvn5+WLo0KHCyspKVFdXqz1uTyovLxc2NjYiLCxMoTwyMlJIJBLx17/+VRQVFYmKigqxdevWVs35bfDiiy82OXf2wIEDwsbGRkRFRTW5/6VLlwQAMXDgQLX7buvX7/Xr10WXLl3k19/OnTsFADFnzhyFeh3tWn18XmuDhs/+4Ycfqn2sOsp12NTYPK6kpEQAEN26dWu2j8boe85ve8SlVEmXmpvzq5Hk19/fX7zwwgsKZfPmzRNGRkbi4cOH8rInf1hWVFQIa2trERwcrLDvf/7zHwFA4Zd4Yw/MnD9/XgAQH3zwQZOxRUdHCwAiPz+/0Rha48m2Y2JiBACRnJwsr1NeXi7c3d1FcXGxEEKIyspKYWlpqfCZKyoqhLm5uZg/f74Q4n+/LCorK9XqX5XjoEr/qgoJCVH6xfLTTz8JAGLVqlXytlU5xuqcC+r+0n18HBuS0MuXL8vLVD1/nxQZGSl69uwpSkpK5GUVFRXC0tJSvPLKKwp1W/vAW4Pmkl9VnD59WgAQL7/8str7tvXr95NPPhGzZs2S/7+4uFiYm5sLmUwmKioqFOp2pGv1yQfe9u3bJ5ycnISjo6PIy8uT72dI12FjY9MUiUQi7Ozsmq3TGCa/6mPyS7rUXPLb+IQ/NVVVVUEqlSqU1dXVwdTUtNmJytnZ2SgrK1N6x7S/vz/MzMyUvj5/kp+fH2xtbXH+/Pkm6zTMM9TGWohPtj137lysXLkSsbGxCAoKAgDs2rULb7zxBmQyGQDg4sWLqKiogK+vr7wdCwsLODs7P/Vrxaf1r8px0GT/jXn++edhaWkpb0vVY9zac0FVZmZmAKDwnvuWnL/79+9HcnIyvvvuO9jY2MjLL1++jIqKCowcOVIj8Wpaw7vWy8vLG92enJyMiIgIXLt2DQDQu3dv/PDDD3B0dFSq29au3927dyM6Olr+f5lMhlGjRuHrr79Genq6wjqWHe1aLS4uhkQigbGxMZydnTFmzBisWLECrq6uAAzvOlRVeXk5hBDyY66ukydPys8ferqGV9RyzEgXmnsttkbm/I4ZMwY///wz0tPTUVlZidOnTyMtLQ2vv/56s8nvgwcPAPzvF/Lj7OzsUFpa+tS+TU1NFX6AfvPNNxg+fDi6dOkCc3NzfPTRRy34RI17WtvW1taYN28efvzxR/znP/8BAPztb39TWJi5IelYtmyZwvqT169ff+rDGk/rX5Xj0Jr+VWVubo6CggIAqh9jTZwLLaXu+bt3716sW7cOGRkZ6NGjh8K2houtS5cuWou3Nbp37w5zc3Ncvny50e1TpkzB1atX0b17dzg5OeG///1vo4kv0Lau319//RVZWVkYO3aswnndsN7sk6s+dLRr1dbWFkII1NbWIi8vD//4xz/QvXt3+XZDuw5VdenSJQCP/sgjIsOhkTu/K1euxM8//4yZM2eirKwMLi4umDJlSrMP3gCPfpgCaPQH6oMHD+Dm5tbs/rW1tSgsLIS7uzsA4MaNG5gwYQImTpyIf/zjH+jatSu2bNmikQRY1bbDwsIQGxuLmJgYhIaGolu3bvDy8pJvb0iKYmJiEB4ertH+VTkOLe1fVTU1NQrHTtVj3NpzoTXUOX+3bNmCb7/9FseOHWs0QWi4c/Xw4UOtxdsaUqkUL7/8Mr755hucPHmyVQ+9taXr94svvsDUqVOxe/duhfKioiK4urriu+++w927d+Hs7CzfZkjXqqFdh6o6fPgwAGD06NEt2n/gwIFISUlpcf+GJjk5GW+++SbHjHSi4XxrjEaS3+zsbOTm5qKgoKDJpZMa4+vrC2tra5w+fVqh/NSpU6iursZzzz3X7P7ff/896uvr8eyzzwIAsrKyUFNTg/nz58PT0xOA5tYWVbVtNzc3TJkyBUlJSbh9+zZWrFihsL1bt26QSqVqv+lLlf5VOQ4t7V9VGRkZEELIkypVj3Frz4XWUGXchBBYvHgxioqKkJaW1mQ9X19fGBkZ4YcffkBoaKjWYm6NVatW4bvvvsOHH36IY8eOtXgJsrZy/QohsHfvXuzatUtpm729PYKCgvDPf/4Tu3fvxsKFC+XbDOlaNbTrUBV3795FTEwM3NzcMHv27Ba3Q0Ttj0amPfzlL3+Bu7u72u/dlkqlWLRoEfbv349du3ahpKQEWVlZCA0NhYuLC0JCQhTqV1dXo7i4GLW1tfjll18QFhaG7t27y5erabiDdOTIEVRVVSEnJ0djc9TUaXvRokWora1FUVERXnrpJaXPPGvWLOzZswfx8fEoKSlBXV0d8vLycOfOnVb1r8pxaGn/Tamvr0dRURFqa2tx/vx5hIeHw93dXX5MVD3G6p4LmqTKuF24cAEbNmxAQkICTE1NlV6bumnTJgCP7tZNZlfNegAAIABJREFUmjQJ+/btw44dO1BSUoLz58/rZJ3lQ4cOPXWpMwB47rnnsHPnTvz8888YPnw4Dh8+jDt37qC2thbXr1/Hzp07UVhY+NT+2sr1++OPP0Imk2Hw4MGNbm/4I6SxF14YyrVqaNfh44QQKCsrQ319PYQQKCgoQFJSEgYPHgxjY2OkpaW1eM4vEbVTajwd16Rjx46JTp06CQDyf6ampqJPnz4iNTVVCCHEX//6V+Hk5CQACCsrKzFx4kQhxKNXTm7cuFE888wzwtTUVNjb24sJEyaIixcvKvSRmJgoRowYIRwdHYWJiYno1KmTmDp1qrh+/bpCvYiICOHg4CDs7OxEUFCQ+PTTTwUA4eXlJcLDwxuNQVXNtf340l5CCDFixAixffv2Rtt5+PChiIiIEO7u7sLExER06dJFTJo0SWRnZ4v169cLCwsL+fI7O3fuVLl/VY7D0/pXR0hIiDA1NRWurq7CxMREyGQy8cYbb4jc3FyFeqoeY1XqNXYebdmyRTg7OwsAwtLSUowbN05s3bpVWFpaCgDimWeeEbm5uWLbtm1CJpMJAKJ79+7yJbZUGbesrCyF7U/+27hxozzG0tJSMXfuXNGpUydhbW0thgwZIpYvXy4ACDc3N3Hu3Dm1xvnEiRNi8ODBwsXFRd6fs7OzCAgIED/88IO83sGDB4WNjY1Ys2aNSu1evXpVhIeHi759+worKyshlUqFh4eHGDp0qFi8eLHIzMxsdtzVObbaun7nzJkjrKyshImJiejfv7/45ZdfFNpbvXq1wri5urqKrVu3KtRpr9fqv//9b9GzZ0/5/i4uLiIoKKjJ421I1+FXX30l+vXrJywtLYWZmZkwMjISAOQrO7zwwgsiKipK3L9/v8nxehqu9qA+rvZAuqT1pc62bt0qwsPDFcoePnwoFixYIMzNzZWWGSLt0PVxCAkJEQ4ODhptUx94/pKu8ZxT1t7GhMmv+pj8ki5pdamzu3fvIiwsTGlempmZGdzd3VFTU4OamhpYWFi0titqhr6OgzaWkNMlnr+kazznlHFMiEiXWj3n18LCAqamptixYwd+//131NTU4Pbt29i+fTuWL1+O4ODgNjuf6rffflOaM9bYv8fXB22rNHEcOtJ4qEof568hjjP9T3v+maktHBPShyNHjmDJkiVITU2Fp6en/GfvW2+9pVR31KhRsLGxgbGxMfr27YtffvlFDxGrr76+HjExMQgICGh0e1RUFHx8fCCTyWBubg5vb2989NFHjc693717N/z9/WFjY4Pu3btj1qxZuHv3rnz7V199hfXr17ePm2Jq3CZuUmZmpnj55ZeFTCYTxsbGwtbWVgQEBIitW7eKmpqaFt6wJnXp8jgsWbJEmJmZCQCiR48eIiUlRaPt6xLPX9I1nnPK2tuYcNqD+trStIfly5eLsWPHKrwZ0MvLSz7v/MCBA0r7HDp0SIwfP16XYbbKpUuXxODBgwWAJt8MOmzYMLF161Zx//59UVJSIpKSkoSpqal47bXXFOrt3btXABDr168XDx48EGfOnBGenp5iwIABCtdnbGysGDZsmCgqKtLqZ1OF1uf8EhERGZK2kPxWVFSIQYMGtZs+2kp+8cknn4iePXsqvZbcy8tLfPHFF8LIyEi4urqKBw8eKGxvT8nv2bNnxcSJE8WuXbvEgAEDmkx+AwMDRW1trULZlClTBACFB/lHjBghunbtKurr6+VlDQ/yHj9+XGH/sLAwMWjQIL3/0dpc8quRpc6IiIhIt3bs2IH8/Px234cuXb58GR9//DFWrVql9DptAAgICEB4eDhu3bqFDz74QA8Rakb//v2RmpqK6dOnw9zcvMl6Bw4cUHqDYufOnQFA4U2SN2/ehIuLi8Ka5d26dQMAXL9+XWH/lStX4uzZs4iNjW3159AWJr9EREQ6IITA5s2b0adPH5ibm8Pe3h5vvPEGfvvtN3mdsLAwmJmZKbyN8N1334WVlRUkEgnu3bsHAAgPD8eiRYuQm5sLiUQCb29vxMXFQSqVwtHREe+88w5cXFwglUoREBCgsNZ0a/oAHr0ZT5U1xduiuLg4CCEwbty4JuusWbMGPXv2xPbt23HkyJFm21PlmMbHx8PKygqWlpZIT0/H6NGjIZPJ4Obmhj179ii0V1dXh+XLl8Pd3R0WFhbo168fkpKSWveh1XTr1i1YWFjAw8NDXubp6an0R1DDfN+Gl/k0sLe3x7BhwxAbGwshhPYDbgk1bhMTERGRaNm0h+XLlwszMzOxc+dO8eDBA3H+/Hnx7LPPis6dO4u7d+/K602fPl04OTkp7Ltx40YBQBQUFMjLJk2aJLy8vBTqhYSECCsrK3HhwgVRVVUlsrOzhb+/v7CxsVH4Grs1fRw4cEDY2NiIqKgotT5/W8gvPD09hY+PT6PbvLy8xNWrV4UQQvz444/CyMhI9OjRQ5SVlQkhGp/2oOoxjYyMFADE0aNHRXFxscjPzxdDhw4VVlZWorq6Wl7vgw8+EObm5mLfvn2iqKhILF26VBgZGYmffvqpxZ/5xRdfbHLaw5PKy8uFjY2NCAsLUyjPyMgQpqamIi4uTpSUlIhff/1V9OnTR7z66quNtrNkyRIBQJw5c6bFcbcWpz0QERHpUWVlJTZv3oyJEydixowZsLW1hZ+fHz777DPcu3dPo2+BNDExkd+J9PHxQXx8PEpLS5GYmKiR9gMDA1FSUoKPP/5YI+3pSnl5Oa5evQovL6+n1h00aBAWLFiAa9euYfHixY3WackxDQgIgEwmQ5cuXRAcHIzy8nLcuHEDAFBVVYX4+HhMmDABkyZNgp2dHZYtWwZTU1ONHbuniY6OhouLC9asWaNQPmzYMERERCAsLAwymQy+vr4oLS3F9u3bG23nmWeeAfDode9tEZNfIiIiLcvOzkZZWRmef/55hXJ/f3+YmZmp9Crvlnr++edhaWmp8FW8IcrPz4cQApaWlirVX7NmDXr16oWtW7fi+PHjSttbe0zNzMwAADU1NQCAixcvoqKiAr6+vvI6FhYWcHZ21smx279/P5KTk/Htt9/CxsZGYVtkZCS2bduGo0ePoqysDFeuXEFAQAAGDRqEmzdvKrXVMMa///671uNuCSa/REREWvbgwQMAgLW1tdI2Ozs7lJaWarV/c3NzFBQUaLWPtq6qqgoAmn0A7HFSqRSJiYmQSCSYPXs2KisrFbZr+piWl5cDAJYtW6aw5vv169cVHj7Thr1792LdunXIyMhAjx49FLbduXMH69evx7x58/DSSy/BysoKHh4eSEhIwO3bt7Fx40al9hpeSNMw5m0Nk18iIiIts7OzA4BGE6IHDx7Azc1Na33X1NRovY/2oCEhU+clDIMGDcLChQuRk5OD1atXK2zT9DHt0qULACAmJgZCCIV/J06cUKstdWzZsgW7du3CsWPH0LVrV6XtOTk5qKurU9omk8ng4OCA7OxspX2qq6sBoM2+lZHJLxERkZb5+vrC2toap0+fVig/deoUqqur8dxzz8nLTExM5F+Fa0JGRgaEEBg4cKDW+mgPHB0dIZFIUFxcrNZ+q1evRu/evXHmzBmFcnWOqSq6desGqVSq9JpvbRFCICIiAllZWUhLS2v0DjYAeRJ/584dhfLS0lIUFhbKlzx7XMMYOzk5aThqzWDyS0REpGVSqRSLFi3C/v37sWvXLpSUlCArKwuhoaFwcXFBSEiIvK63tzcKCwuRlpaGmpoaFBQUKK2lCgAODg64ffs2rl27htLSUnkyW19fj6KiItTW1uL8+fMIDw+Hu7s7Zs6cqZE+Dh061C6XOrO0tISnpyfy8vLU2q9h+sOT6+Gqc0xV7WfWrFnYs2cP4uPjUVJSgrq6OuTl5ckTz+DgYDg5OWnk9coXLlzAhg0bkJCQAFNTU4WpFhKJBJs2bQIAeHh4YMSIEUhISEBmZiYqKytx8+ZN+eebM2eOUtsNY+zn59fqOLWByS8REZEOrFixAtHR0YiKikLnzp0xbNgw9OjRAxkZGbCyspLXmz9/PkaMGIGpU6eiV69eWL16tfzr48cfMAoNDYWjoyN8fHwwZswYFBYWAng0z9LPzw8WFhYYOnQoevbsie+//15hrmtr+2ivAgMDkZ2drTB/98svv4S3tzdyc3Ph7++P9957T2m/gQMHYuHChUrlqhzT+Ph4xMTEAAD69euHK1euICEhAYsWLQIAvPbaa8jJyQEAxMbGYsGCBVi/fj06deoEFxcXhIeHo6ioCMCj6QT5+flIT09v9nOePHkSQ4YMQdeuXXHq1CmcO3cOLi4uGDx4MDIzMwFA5TV4JRIJUlJSEBwcjDlz5sDe3h4+Pj64ceMGUlNTMXToUKV9fvrpJ7i6uqJfv34q9aFrEvHEp09OTsabb77ZdhcmJiIi0rOgoCAAQEpKip4jUfTOO+8gJSUF9+/f13coStpCfnH58mX06dMHiYmJmDFjht7iaKn6+noMHz4cM2fOxOzZs/UdTqPu378PNzc3rFmzRp7g60Mz51sK7/wSERF1IOo80GVovL29ERUVhaioKJSVlek7HLXU1dUhLS0NpaWlCA4O1nc4TVq5ciUGDBiAsLAwfYfSJCa/REREZDCWLFmCoKAgBAcHq/3wmz5lZGQgNTUVhw4dUnmtYl3bvHkzzp49i4MHD8LU1FTf4TSJyS8REVEHsHTpUiQmJqK4uBgeHh7Yt2+fvkNqs9auXYuwsDB88skn+g5FZSNHjsQXX3wBZ2dnfYfSqPT0dDx8+BAZGRmwt7fXdzjNMtF3AERERNR60dHRiI6O1ncY7caoUaMwatQofYfRYYwfPx7jx4/Xdxgq4Z1fIiIiIjIYTH6JiIiIyGAw+SUiIiIig8Hkl4iIiIgMBpNfIiIiIjIYTa72IJFIdBkHERFRu8PflerjmJG+KSW/AQEBSEpK0kcsREQG4cSJE4iNjeXPWiIiPZAIfb5km4jIADXzznkiItKuFM75JSIiIiKDweSXiIiIiAwGk18iIiIiMhhMfomIiIjIYDD5JSIiIiKDweSXiIiIiAwGk18iIiIiMhhMfomIiIjIYDD5JSIiIiKDweSXiIiIiAwGk18iIiIiMhhMfomIiIjIYDD5JSIiIiKDweSXiIiIiAwGk18iIiIiMhhMfomIiIjIYDD5JSIiIiKDweSXiIiIiAwGk18iIiIiMhhMfomIiIjIYDD5JSIiIiKDweSXiIiIiAwGk18iIiIiMhhMfomIiIjIYDD5JSIiIiKDweSXiIiIiAwGk18iIiIiMhhMfomIiIjIYDD5JSIiIiKDweSXiIiIiAwGk18iIiIiMhhMfomIiIjIYJjoOwAioo6soKAAX375pULZ6dOnAQDbtm1TKLexscHUqVN1FhsRkSGSCCGEvoMgIuqoHj58CEdHR5SVlcHY2BgA0PBjVyKRyOvV1NTgz3/+Mz7//HN9hElEZChSOO2BiEiLzM3NMXnyZJiYmKCmpgY1NTWora1FbW2t/P81NTUAgGnTpuk5WiKijo/JLxGRlk2bNg3V1dXN1rGzs8NLL72ko4iIiAwXk18iIi0bMWIEunTp0uR2U1NTzJgxAyYmfAyDiEjbmPwSEWmZkZERpk+fDlNT00a319TU8EE3IiIdYfJLRKQDU6dOlc/tfVLXrl0xaNAgHUdERGSYmPwSEenACy+8gO7duyuVm5mZ4c9//rPCyg9ERKQ9TH6JiHTkrbfeUpr6UF1dzSkPREQ6xOSXiEhHpk+frjT1wdvbG35+fnqKiIjI8DD5JSLSkd69e8PHx0c+xcHU1BSzZs3Sc1RERIaFyS8RkQ796U9/kr/prba2llMeiIh0jMkvEZEOTZ06FXV1dQCAZ599Fh4eHnqOiIjIsDD5JSLSIXd3d7z44osAgD//+c96joaIyPDwdULUIZ04cQKbN2/WdxhEjXr48CEkEgm+++47ZGZm6jscokalpKToOwQireCdX+qQbt68iX379uk7DGqHTp48iZMnT2q1Dzc3Nzg5OUEqlWq1H13Jy8vj9daB8HhSR8c7v9Sh8c4FqSsoKAiA9s+dy5cvw9vbW6t96EpycjLefPNNXm8dRMPxJOqoeOeXiEgPOkriS0TU3jD5JSIiIiKDweSXiIiIiAwGk18iIiIiMhhMfomIiIjIYDD5JSLSgoMHD8LW1hZff/21vkNp844cOYIlS5YgNTUVnp6ekEgkkEgkeOutt5Tqjho1CjY2NjA2Nkbfvn3xyy+/6CFi9dXX1yMmJgYBAQGNbo+KioKPjw9kMhnMzc3h7e2Njz76CGVlZUp1d+/eDX9/f9jY2KB79+6YNWsW7t69K9/+1VdfYf369fI3CRKRIia/RERaIITQdwjtwooVKxAXF4elS5di0qRJuHLlCry8vNCpUyfs2rUL33zzjUL97777DikpKRg7diyys7Px7LPP6ily1eXk5OCPf/wjFi5ciIqKikbrHDt2DH/5y19w7do13Lt3D9HR0YiNjZUvvdcgKSkJ06dPR1BQEPLy8pCeno7MzEyMHj0atbW1AIBx48ZBKpVi5MiRePDggdY/H1F7w+SXiEgLAgMDUVxcjLFjx+o7FFRWVjZ5x1Gf1q1bh7179yI5ORk2NjYK2+Li4mBkZISQkBAUFxfrKcLWO3fuHBYvXozQ0FAMGDCgyXrW1tYICQmBg4MDbGxsMGXKFEyYMAGHDx/GzZs35fX+/ve/o2vXrvjwww9ha2uLAQMGYOHChTh79ixOnTolr/f++++jf//+GDNmjDwpJqJHmPwSEXVwO3bsQH5+vr7DUHD58mV8/PHHWLVqVaNvugsICEB4eDhu3bqFDz74QA8Rakb//v2RmpqK6dOnw9zcvMl6Bw4cgLGxsUJZ586dAUDhbvHNmzfh4uICiUQiL+vWrRsA4Pr16wr7r1y5EmfPnkVsbGyrPwdRR8Lkl4hIw44fPw53d3dIJBJ8+umnAID4+HhYWVnB0tIS6enpGD16NGQyGdzc3LBnzx75vnFxcZBKpXB0dMQ777wDFxcXSKVSBAQEKNzZCwsLg5mZGZydneVl7777LqysrCCRSHDv3j0AQHh4OBYtWoTc3FxIJBL5yzUOHz4MmUyGtWvX6mJIlMTFxUEIgXHjxjVZZ82aNejZsye2b9+OI0eONNueEAKbN29Gnz59YG5uDnt7e7zxxhv47bff5HVUPQYAUFdXh+XLl8Pd3R0WFhbo168fkpKSWveh1XTr1i1YWFjAw8NDXubp6an0h0zDfF9PT0+Fcnt7ewwbNgyxsbGchkP0GCa/REQaNmTIEPz4448KZfPnz8eCBQtQWVkJGxsbJCUlITc3F56ennj77bdRU1MD4FFSO3PmTFRUVOD999/HtWvX8Msvv6C2thavvPKK/CvwuLg4TJkyRaGPrVu3YtWqVQplsbGxGDt2LLy8vCCEwOXLlwFA/jBUfX29Vsbgab755hv06tULlpaWTdaxsLDA559/DiMjI7z99tsoLy9vsu7KlSuxZMkSREZGIj8/H5mZmbh58yaGDh2K33//HYDqxwAAFi9ejA0bNiAmJgZ37tzB2LFjMW3aNJw+fVpzg9CMiooKHDt2DG+//TbMzMzk5UuXLsXdu3exZcsWlJaWIjs7G7GxsXj11VcxcOBApXb+8Ic/4NatWzh37pxO4iZqD5j8EhHpWEBAAGQyGbp06YLg4GCUl5fjxo0bCnVMTEzkdzF9fHwQHx+P0tJSJCYmaiSGwMBAlJSU4OOPP9ZIe+ooLy/H1atX4eXl9dS6gwYNwoIFC3Dt2jUsXry40TqVlZXYvHkzJk6ciBkzZsDW1hZ+fn747LPPcO/ePWzbtk1pn+aOQVVVFeLj4zFhwgRMmjQJdnZ2WLZsGUxNTTU2/k8THR0NFxcXrFmzRqF82LBhiIiIQFhYGGQyGXx9fVFaWort27c32s4zzzwDAMjKytJ6zETtBZNfIiI9arir9/hdx8Y8//zzsLS0VPgav73Kz8+HEKLZu76PW7NmDXr16oWtW7fi+PHjStuzs7NRVlaG559/XqHc398fZmZmCtNFGvPkMbh48SIqKirg6+srr2NhYQFnZ2edjP/+/fuRnJyMb7/9VulBwMjISGzbtg1Hjx5FWVkZrly5goCAAAwaNEjhwbgGDWPccPebiJj8EhG1G+bm5igoKNB3GK1WVVUFAM0+APY4qVSKxMRESCQSzJ49G5WVlQrbG5bzsra2VtrXzs4OpaWlasXXML1i2bJl8jWHJRIJrl+/3uRSZZqyd+9erFu3DhkZGejRo4fCtjt37mD9+vWYN28eXnrpJVhZWcHDwwMJCQm4ffs2Nm7cqNSehYUFgP+NOREx+SUiahdqamrw4MEDuLm56TuUVmtIyNR5CcOgQYOwcOFC5OTkYPXq1Qrb7OzsAKDRJLclY9alSxcAQExMDIQQCv9OnDihVlvq2LJlC3bt2oVjx46ha9euSttzcnJQV1entE0mk8HBwQHZ2dlK+1RXVwP435gTEZNfIqJ2ISMjA0IIhYeaTExMnjpdoi1ydHSERCJRe/3e1atXo3fv3jhz5oxCua+vL6ytrZUeRjt16hSqq6vx3HPPqdVPt27dIJVKcfbsWbX2aykhBCIiIpCVlYW0tLRG72ADkCfxd+7cUSgvLS1FYWGhfMmzxzWMsZOTk4ajJmq/mPwSEbVB9fX1KCoqQm1tLc6fP4/w8HC4u7tj5syZ8jre3t4oLCxEWloaampqUFBQoLTWKwA4ODjg9u3buHbtGkpLS1FTU4NDhw7pbakzS0tLeHp6Ii8vT639GqY/PLkerlQqxaJFi7B//37s2rULJSUlyMrKQmhoKFxcXBASEqJ2P7NmzcKePXsQHx+PkpIS1NXVIS8vT554BgcHw8nJSSOvV75w4QI2bNiAhIQEmJqaKky1kEgk2LRpEwDAw8MDI0aMQEJCAjIzM1FZWYmbN2/KP9+cOXOU2m4YYz8/v1bHSdRRMPklItKwTz/9FP7+/gCAiIgIjB8/HvHx8YiJiQEA9OvXD1euXEFCQgIWLVoEAHjttdeQk5Mjb6Oqqgp+fn6wsLDA0KFD0bNnT3z//fcK82Tnz5+PESNGYOrUqejVqxdWr14t/3r78QegQkND4ejoCB8fH4wZMwaFhYU6GYfmBAYGIjs7W2H+7pdffglvb2/k5ubC398f7733ntJ+AwcOxMKFC5XKV6xYgejoaERFRaFz584YNmwYevTogYyMDFhZWQGAWscgNjYWCxYswPr169GpUye4uLggPDwcRUVFAB5NJ8jPz0d6enqzn/PkyZMYMmQIunbtilOnTuHcuXNwcXHB4MGDkZmZCUD1V2FLJBKkpKQgODgYc+bMgb29PXx8fHDjxg2kpqZi6NChSvv89NNPcHV1Rb9+/VTqg8gQSARXvqYOKDk5GW+++SYXdie1BQUFAQBSUlL0FsM777yDlJQU3L9/X28xqKMl19vly5fRp08fJCYmYsaMGVqMTjvq6+sxfPhwzJw5E7Nnz9Z3OI26f/8+3NzcsGbNGnmCrwr+/KQOLoV3fomI2iB1HgZrj7y9vREVFYWoqCiUlZXpOxy11NXVIS0tDaWlpQgODtZ3OE1auXIlBgwYgLCwMH2HQtSmMPklIiK9WLJkCYKCghAcHKz2w2/6lJGRgdTUVBw6dEjltYp1bfPmzTh79iwOHjwIU1NTfYdD1KYw+SVqwty5c2FjYwOJRKKzp761pb6+HjExMQgICGhxG6mpqfD09FR6GMfMzAyOjo4YPnw4Nm7cKJ8TSS2zdOlSJCYmori4GB4eHti3b5++Q9KqtWvXIiwsDJ988om+Q1HZyJEj8cUXX8DZ2VnfoTQqPT0dDx8+REZGBuzt7fUdDlGbw+SXqAnbt29HQkKCvsNotZycHPzxj3/EwoULW7VA/6RJk3DlyhV4eXnB1tYWQgjU19cjPz8fycnJ8PDwQEREBPr27au05BSpLjo6Gg8fPoQQAlevXsXkyZP1HZLWjRo1CuvWrdN3GB3G+PHjsWTJEqVVMYjoESa/RB3YuXPnsHjxYoSGhmLAgAEab18ikcDOzg7Dhw9HYmIikpOT8fvvvyMwMLBdfY1NRESGg8kvUTMkEom+Q2iV/v37IzU1FdOnT1f5VbKtMXnyZMycORP5+fn47LPPtN4fERGRupj8Ev0fIQQ2btyIXr16wdzcHLa2tvjwww+V6tXV1WH58uVwd3eHhYUF+vXrh6SkJACP1hG1srKCpaUl0tPTMXr0aMhkMri5uWHPnj0K7fzwww944YUXYGlpCZlMBj8/P5SUlDy1D204fPiwxl540PAShkOHDsnLOuKYERFR+8Tkl+j/fPzxx4iIiEBISAh+//133L17F4sXL1aqt3jxYmzYsAExMTG4c+cOxo4di2nTpuH06dOYP38+FixYgMrKStjY2CApKQm5ubnw9PTE22+/LX8VbXl5OcaNG4fJkyejsLAQOTk56NmzJ6qrq5/ahzY0LKtVX1/f6rYapldcuXJFXtYRx4yIiNopQdQBJSUlCXVO74qKCmFpaSleeeUVhfI9e/YIAOLMmTNCCCEqKyuFpaWlCA4OVtjX3NxczJ8/XwghRGRkpAAgKisr5XW2bt0qAIjLly8LIYT49ddfBQBx4MABpVhU6aMlXnzxRdG/f/8W79/Ay8tL2NraNltHIpEIOzs7IUT7G7PJkyeLyZMnq7WPoVP3eqO2jceTOrhkE30l3URtyeXLl1FRUYGRI0c2W+/ixYuoqKiAr6+vvMzCwgLOzs747bffmtzPzMwMAOR3MT09PeHo6IgZM2bg/fffx8yZM9GjR49W9dFWlJeXQwgBmUwGoH2O2b59+9r9fG994JgRUXvA5JcIQF5eHgCgS5cuzdYrLy8HACxbtgzLli1T2Obi4qJyfxYWFjh27BgWL16MtWvXIioqClOmTEFiYqLG+tCXS5cuAQB69+4NoH2O2cAmom3YAAAN3klEQVSBA7FgwQK19zNUJ06cQGxsLOdYdxANx5Ooo2LySwRAKpUCAB4+fNhsvYbkOCYmBuHh4a3qs2/fvvj6669RUFCAzZs3Y926dejbt6/8dama6EMfDh8+DAAYPXo0gPY5Zm5ubpgyZUqr2zEksbGxHLMOhMkvdWR84I0IgK+vL4yMjPDDDz80W69bt26QSqWtfuPb7du3ceHCBQCPksNPPvkEzz77LC5cuKCxPvTh7t27iImJgZubG2bPng2AY0ZERG0Lk18iPEqmJk2ahH379mHHjh0oKSnB+fPnsW3bNoV6UqkUs2bNwp49exAfH4+SkhLU1dUhLy8Pd+7cUbm/27dv45133sFvv/2G6upqnDlzBtevX8fAgQM11oc6Dh06pNZSZ0IIlJWVob6+HkIIFBQUICkpCYMHD4axsTHS0tLkc3476pgREVE7pecn7oi0oiVPK5eWloq5c+eKTp06CWtrazFkyBCxfPlyAUC4ubmJc+fOCSGEePjwoYiIiBDu7u7CxMREdOnSRUyaNElkZ2eLrVu3CktLSwFAPPPMMyI3N1ds27ZNyGQyAUB0795dXLp0SVy7dk0EBAQIe3t7YWxsLLp27SoiIyNFbW3tU/tQx4kTJ8TgwYOFi4uLACAACGdnZxEQECB++OEHeb2DBw8KGxsbseb/t3f3MVWWbxzAv7ycF84BBBMBeQvOocDQzKQF1czcaObiJWKwtKW2wmYRgcxAIQKyiI0aTVZqY01TAiVxJZtpY9ZyriZMOk1TfAEy5EV5FQTh+v3ROHl+kHIAOQLfz8Y/93M/9309z30Ouzg857pzcv5zrIMHD8rChQtFo9GIUqkUa2trAWCs7PDYY49JVlaWtLa2Djt3Kt0zVnswH6sDTC9cT5rmSqxERCyUdxPdNSUlJYiNjQVf3mSumJgYAEBpaamFI5k6+H6bXrieNM2V8rEHIiIiIpoxmPwSTSGnT5+GlZXVHX+Gqh8QTVVHjhxBamoq9u/fDz8/P+Nr++WXXx7WNywsDA4ODrCxscFDDz2EkydPWiDi0cvNzUVAQADs7Oyg1WoREBCA9PR041bdt9qzZw+Cg4Ph4OAAHx8frF27Fo2NjcbjBw8eRG5urnGXRiK6Mya/RFNIQEAAROSOP8XFxZYOlWjM3nvvPRQUFCAtLQ3R0dE4f/48dDod7rvvPuzevRvff/+9Sf/Dhw+jtLQUzz//PAwGAxYvXmyhyEfnp59+wmuvvYa6ujpcuXIF2dnZyM3NxYsvvmjS75tvvsGqVasQExODhoYGlJeX49ixY1ixYgVu3rwJAAgPD4darcby5cvR1tZmicshmnKY/BIR3WN6enoQGho65ecYi48++gjFxcUoKSmBg4ODybGCggJYW1sjPj4e7e3tFopw/JRKJTZs2AAXFxfY29sjJiYGkZGR+OGHH0yqk3zxxReYN28eUlJSMGvWLCxatAhJSUmorq7GiRMnjP3efvttPPzww3juueeMSTER/Tcmv0RE95gvv/wSTU1NU34Oc507dw7p6el4//33jRvP3Co0NBSJiYn466+/sHHjRgtEODHKysqGXZ+HhwcAoKury9hWX18Pd3d3k22jvby8AACXLl0yOT8zMxPV1dXcnIJoFJj8EhGNk4ggPz8fgYGBUKlUcHZ2RmRkJE6fPm3sk5CQAKVSCTc3N2Pbhg0boNVqYWVlhZaWFgBAYmIikpOTUVtbCysrK+j1ehQUFECtVmPu3LlYv3493N3doVarERoaavIJ4HjmAP7Znc+ces8TraCgACKC8PDw/+yTk5ODBx54ADt37sSRI0duO95o1qWwsBBarRYajQbl5eVYsWIFHB0d4enpib1795qMNzAwgIyMDHh7e8POzg4LFy6csC2dz549CycnJ/j4+Bjb/Pz8hv2BMvS8r5+fn0m7s7Mzli5dik8//ZRVGojuZLKLqxFNBtappLEaS53fjIwMUSqVsmvXLmlra5NTp07J4sWLZc6cOdLY2Gjst2rVKnF1dTU5Ny8vTwBIc3OzsS06Olp0Op1Jv/j4eNFqtfLHH39Ib2+vGAwGCQ4OFgcHB6mrq5uQOb777jtxcHCQrKwss65/ot5vfn5+Mn/+/BGP6XQ6uXDhgoiI/PLLL2JtbS3333+/dHV1iYhIRUWFREREmJwz2nXZvHmzAJCjR49Ke3u7NDU1yVNPPSVarVb6+vqM/TZu3CgqlUr27dsn165dk7S0NLG2tpZff/11TNfb19cnDQ0N8tlnn4lKpZJdu3aZHK+srBSFQiEFBQXS0dEhv//+uwQGBsqzzz474nipqakCQKqqqsYUzxD+/qRproSf/BIRjUNPTw/y8/PxwgsvYPXq1Zg1axYWLFiAzz//HC0tLcN2CRwPW1tb46eY8+fPR2FhITo7O1FUVDQh469cuRIdHR1IT0+fkPHM0d3djQsXLkCn092xb0hICN555x1cvHgR77777oh9xrIuoaGhcHR0hIuLC+Li4tDd3Y26ujoAQG9vLwoLCxEVFYXo6Gg4OTlhy5YtUCgUY77/Xl5e8PT0RGZmJj7++GPExsaaHF+6dCk2bdqEhIQEODo6IigoCJ2dndi5c+eI4/n7+wMAampqxhQP0UzB5JeIaBwMBgO6urqwZMkSk/bg4GAolUqTxxIm2pIlS6DRaEz+jT9VNTU1QUSg0WhG1T8nJwcPPvggtm3bhp9//nnY8fGui1KpBAD09/cDAM6cOYPr168jKCjI2MfOzg5ubm5jvv/19fVoamrCnj178NVXX+GRRx4xecxh8+bN2L59O44ePYquri6cP38eoaGhCAkJQX19/bDxhu7dlStXxhQP0UzB5JeIaByGykvZ29sPO+bk5ITOzs67Or9KpUJzc/NdnWMy9Pb2AvjnekZDrVajqKgIVlZWWLduHXp6ekyOT/S6dHd3AwC2bNliUlP70qVLuH79ulljDVEoFHBxcUFYWBiKi4thMBiwdetWAMDff/+N3NxcvP7663jmmWeg1Wrh6+uLHTt24PLly8jLyxs2np2dHYB/7yURjYzJLxHRODg5OQHAiMlUW1sbPD0979rc/f39d32OyTKUuJmzWUNISAiSkpJw9uxZZGdnmxyb6HVxcXEBAHzyySfD6mofP37crLFGotfrYWNjA4PBAOCfL8ANDAxg3rx5Jv0cHR0xe/ZsY79b9fX1Afj3XhLRyJj8EhGNQ1BQEOzt7fHbb7+ZtJ84cQJ9fX149NFHjW22trbGf6NPhMrKSogIHn/88bs2x2SZO3curKyszK7fm52djYCAAFRVVZm0m7Muo+Hl5QW1Wo3q6mqzzvt/ra2teOmll4a1DyW7Q6XMhpLzW+v+Av8k81evXjX2u9XQvXN1dR1XjETTHZNfIqJxUKvVSE5ORllZGXbv3o2Ojg7U1NTgjTfegLu7O+Lj44199Xo9rl69igMHDqC/vx/Nzc3D6rUCwOzZs3H58mVcvHgRnZ2dxmR2cHAQ165dw82bN3Hq1CkkJibC29sba9asmZA5KioqLFbqTKPRwM/PDw0NDWadN/T4g42NzbD20a7LaOdZu3Yt9u7di8LCQnR0dGBgYAANDQ3GBDUuLg6urq633V5Zq9Xi8OHD+PHHH9HR0YH+/n5UVVXhlVdegVarRVJSEgDA19cXy5Ytw44dO3Ds2DH09PSgvr7eGPerr746bOyhe7dgwQKzro1oxrFkrQmiu4WlemisxlLqbHBwUPLy8sTf318UCoU4OztLVFSUnDlzxqRfa2urLFu2TNRqtfj6+spbb70lKSkpAkD0er2xZNnJkyfFx8dH7Ozs5Mknn5TGxkaJj48XhUIhHh4eYmtrK46OjhIZGSm1tbUTNsehQ4fEwcFBcnJyzLr+iXq/JSQkiEKhkOvXrxvbysrKRKfTCQCZM2eOvPnmmyOem5KSMqzU2WjWZdu2baLRaASA+Pv7S21trWzfvl0cHR0FgPj4+Miff/4pIiI3btyQTZs2ibe3t9ja2oqLi4tER0eLwWAQEZGoqCgBIBkZGbe9zvDwcPH19RV7e3tRqVSi0+kkLi5OampqTPq1tLRIYmKi6PV6UalUYm9vL0888YR8++23I467cuVK8fDwkMHBwdvOfyf8/UnTXImVCKth0/RTUlKC2NhYFnsns8XExAAASktLLRyJqfXr16O0tBStra2WDmWYiXq/nTt3DoGBgSgqKsLq1asnKLrJMzg4iKeffhpr1qzBunXrJnXu1tZWeHp6IicnB8nJyeMai78/aZor5WMPRERThDlfBpuK9Ho9srKykJWVZbLN71QwMDCAAwcOoLOzE3FxcZM+f2ZmJhYtWoSEhIRJn5toqmHyS0RE94zU1FTExMQgLi7O7C+/WVJlZSX279+PioqKUdcqnij5+fmorq7GoUOHoFAoJnVuoqmIyS8R0T0uLS0NRUVFaG9vh6+vL/bt22fpkO6qDz74AAkJCfjwww8tHcqoLV++HF9//TXc3Nwmdd7y8nLcuHEDlZWVcHZ2ntS5iaYqW0sHQEREt7d161bj5gczRVhYGMLCwiwdxj0vIiICERERlg6DaErhJ79ERERENGMw+SUiIiKiGYPJLxERERHNGEx+iYiIiGjG4BfeaForKSmxdAg0xQxtEcvXzugdP34cAO/ZdDG0nkTTFXd4o2lpaIciIiIaG6YHNE2VMvklIiIiopmC2xsTERER0czB5JeIiIiIZgwmv0REREQ0YzD5JSIiIqIZ43+2xaGkLx4oqAAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<IPython.core.display.Image object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "XKburGrWjUug",
        "colab_type": "text"
      },
      "source": [
        "## Parameters Tuning"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Oz6ZOYT5jde-",
        "colab_type": "text"
      },
      "source": [
        "We will tune the parameters optimizer, kernel_initializer and activation."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iqd8pUpo1vC3",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "e8b9bd6f-c30f-4325-e368-877abff2e31f"
      },
      "source": [
        "# fix random seed for reproducibility\n",
        "seed = 7\n",
        "np.random.seed(seed)\n",
        "\n",
        "# define the grid search parameters\n",
        "optimizer = ['RMSprop', 'Adam', 'Adagrad', 'Nadam']\n",
        "kernel_initializer = ['normal', 'uniform']\n",
        "activation = ['relu', 'linear', 'tanh']\n",
        "\n",
        "param_grid = dict(optimizer=optimizer, kernel_initializer=kernel_initializer, activation=activation)\n",
        "\n",
        "# count number of different parameters values combinations\n",
        "parameters_number = 1\n",
        "for x in param_grid:\n",
        "  parameters_number = parameters_number * len(param_grid[x]) \n",
        "print(\"Number of different parameter combinations = {}\".format(parameters_number))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Number of different parameter combinations = 24\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "rn5e1UWniPmD",
        "colab_type": "text"
      },
      "source": [
        "We will try different models with different parameters to find the best parameter values."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "kbtGq865gbEW",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 5341
        },
        "outputId": "628f25ba-bd1f-4101-f64b-00f07395a0e4"
      },
      "source": [
        "epochs = 5\n",
        "batch_size = 20 # 20 divides the training data samples\n",
        "\n",
        "#creating the models with different hyperparameters\n",
        "for a,b,c in [(x,y,z) for x in optimizer for z in activation for y in kernel_initializer]:\n",
        "    params = {'optimizer' : a , 'kernel_initializer' : b , 'activation' : c}\n",
        "    print(params)\n",
        "    curr_model = create_model(a, b, c)\n",
        "    curr_model.fit(training_data_images, training_data_labels, \n",
        "                    validation_data=(testing_data_images, testing_data_labels),\n",
        "                    epochs=epochs, batch_size=batch_size, verbose=1)\n",
        "    print(\"=============================================================================\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "{'optimizer': 'RMSprop', 'kernel_initializer': 'normal', 'activation': 'relu'}\n",
            "Train on 73440 samples, validate on 13360 samples\n",
            "Epoch 1/5\n",
            "73440/73440 [==============================] - 84s 1ms/step - loss: 0.3349 - acc: 0.9089 - val_loss: 2.4386 - val_acc: 0.5037\n",
            "Epoch 2/5\n",
            "73440/73440 [==============================] - 82s 1ms/step - loss: 0.1106 - acc: 0.9670 - val_loss: 0.1605 - val_acc: 0.9507\n",
            "Epoch 3/5\n",
            "73440/73440 [==============================] - 81s 1ms/step - loss: 0.0865 - acc: 0.9738 - val_loss: 0.1138 - val_acc: 0.9665\n",
            "Epoch 4/5\n",
            "73440/73440 [==============================] - 81s 1ms/step - loss: 0.0788 - acc: 0.9772 - val_loss: 2.5557 - val_acc: 0.5233\n",
            "Epoch 5/5\n",
            "73440/73440 [==============================] - 81s 1ms/step - loss: 0.0710 - acc: 0.9788 - val_loss: 0.1519 - val_acc: 0.9559\n",
            "=============================================================================\n",
            "{'optimizer': 'RMSprop', 'kernel_initializer': 'uniform', 'activation': 'relu'}\n",
            "Train on 73440 samples, validate on 13360 samples\n",
            "Epoch 1/5\n",
            "73440/73440 [==============================] - 83s 1ms/step - loss: 0.3012 - acc: 0.9181 - val_loss: 0.3428 - val_acc: 0.8807\n",
            "Epoch 2/5\n",
            "73440/73440 [==============================] - 82s 1ms/step - loss: 0.1075 - acc: 0.9672 - val_loss: 0.1243 - val_acc: 0.9611\n",
            "Epoch 3/5\n",
            "73440/73440 [==============================] - 82s 1ms/step - loss: 0.0875 - acc: 0.9734 - val_loss: 0.0698 - val_acc: 0.9793\n",
            "Epoch 4/5\n",
            "73440/73440 [==============================] - 81s 1ms/step - loss: 0.0793 - acc: 0.9761 - val_loss: 7.7491 - val_acc: 0.3073\n",
            "Epoch 5/5\n",
            "73440/73440 [==============================] - 81s 1ms/step - loss: 0.0727 - acc: 0.9780 - val_loss: 3.6094 - val_acc: 0.5719\n",
            "=============================================================================\n",
            "{'optimizer': 'RMSprop', 'kernel_initializer': 'normal', 'activation': 'linear'}\n",
            "Train on 73440 samples, validate on 13360 samples\n",
            "Epoch 1/5\n",
            "73440/73440 [==============================] - 81s 1ms/step - loss: 0.6358 - acc: 0.8246 - val_loss: 0.5407 - val_acc: 0.8210\n",
            "Epoch 2/5\n",
            "73440/73440 [==============================] - 80s 1ms/step - loss: 0.2895 - acc: 0.9135 - val_loss: 0.3693 - val_acc: 0.8883\n",
            "Epoch 3/5\n",
            "73440/73440 [==============================] - 80s 1ms/step - loss: 0.2099 - acc: 0.9368 - val_loss: 0.1959 - val_acc: 0.9409\n",
            "Epoch 4/5\n",
            "73440/73440 [==============================] - 80s 1ms/step - loss: 0.1700 - acc: 0.9478 - val_loss: 2.2486 - val_acc: 0.7231\n",
            "Epoch 5/5\n",
            "73440/73440 [==============================] - 80s 1ms/step - loss: 0.1494 - acc: 0.9535 - val_loss: 1.5786 - val_acc: 0.6550\n",
            "=============================================================================\n",
            "{'optimizer': 'RMSprop', 'kernel_initializer': 'uniform', 'activation': 'linear'}\n",
            "Train on 73440 samples, validate on 13360 samples\n",
            "Epoch 1/5\n",
            "73440/73440 [==============================] - 82s 1ms/step - loss: 0.5886 - acc: 0.8377 - val_loss: 0.4105 - val_acc: 0.8868\n",
            "Epoch 2/5\n",
            "73440/73440 [==============================] - 80s 1ms/step - loss: 0.2662 - acc: 0.9196 - val_loss: 0.3536 - val_acc: 0.8877\n",
            "Epoch 3/5\n",
            "73440/73440 [==============================] - 80s 1ms/step - loss: 0.1943 - acc: 0.9400 - val_loss: 10.3813 - val_acc: 0.1775\n",
            "Epoch 4/5\n",
            "73440/73440 [==============================] - 80s 1ms/step - loss: 0.1643 - acc: 0.9488 - val_loss: 1.0918 - val_acc: 0.6720\n",
            "Epoch 5/5\n",
            "73440/73440 [==============================] - 80s 1ms/step - loss: 0.1426 - acc: 0.9559 - val_loss: 0.3167 - val_acc: 0.8961\n",
            "=============================================================================\n",
            "{'optimizer': 'RMSprop', 'kernel_initializer': 'normal', 'activation': 'tanh'}\n",
            "Train on 73440 samples, validate on 13360 samples\n",
            "Epoch 1/5\n",
            "73440/73440 [==============================] - 83s 1ms/step - loss: 0.4914 - acc: 0.8685 - val_loss: 2.6090 - val_acc: 0.5469\n",
            "Epoch 2/5\n",
            "73440/73440 [==============================] - 82s 1ms/step - loss: 0.1908 - acc: 0.9409 - val_loss: 1.9058 - val_acc: 0.6582\n",
            "Epoch 3/5\n",
            "73440/73440 [==============================] - 82s 1ms/step - loss: 0.1584 - acc: 0.9510 - val_loss: 0.5578 - val_acc: 0.8564\n",
            "Epoch 4/5\n",
            "73440/73440 [==============================] - 82s 1ms/step - loss: 0.1438 - acc: 0.9553 - val_loss: 0.3582 - val_acc: 0.8997\n",
            "Epoch 5/5\n",
            "73440/73440 [==============================] - 82s 1ms/step - loss: 0.1337 - acc: 0.9591 - val_loss: 0.7828 - val_acc: 0.8272\n",
            "=============================================================================\n",
            "{'optimizer': 'RMSprop', 'kernel_initializer': 'uniform', 'activation': 'tanh'}\n",
            "Train on 73440 samples, validate on 13360 samples\n",
            "Epoch 1/5\n",
            "73440/73440 [==============================] - 84s 1ms/step - loss: 0.4230 - acc: 0.8841 - val_loss: 0.8043 - val_acc: 0.7874\n",
            "Epoch 2/5\n",
            "73440/73440 [==============================] - 83s 1ms/step - loss: 0.1744 - acc: 0.9474 - val_loss: 1.0021 - val_acc: 0.7788\n",
            "Epoch 3/5\n",
            "73440/73440 [==============================] - 82s 1ms/step - loss: 0.1511 - acc: 0.9531 - val_loss: 0.4386 - val_acc: 0.8782\n",
            "Epoch 4/5\n",
            "73440/73440 [==============================] - 82s 1ms/step - loss: 0.1372 - acc: 0.9571 - val_loss: 3.2566 - val_acc: 0.3754\n",
            "Epoch 5/5\n",
            "73440/73440 [==============================] - 82s 1ms/step - loss: 0.1308 - acc: 0.9600 - val_loss: 0.4176 - val_acc: 0.8881\n",
            "=============================================================================\n",
            "{'optimizer': 'Adam', 'kernel_initializer': 'normal', 'activation': 'relu'}\n",
            "Train on 73440 samples, validate on 13360 samples\n",
            "Epoch 1/5\n",
            "73440/73440 [==============================] - 90s 1ms/step - loss: 0.3496 - acc: 0.9075 - val_loss: 6.2671 - val_acc: 0.2249\n",
            "Epoch 2/5\n",
            "73440/73440 [==============================] - 89s 1ms/step - loss: 0.1066 - acc: 0.9679 - val_loss: 0.2868 - val_acc: 0.9138\n",
            "Epoch 3/5\n",
            "73440/73440 [==============================] - 89s 1ms/step - loss: 0.0800 - acc: 0.9755 - val_loss: 0.1065 - val_acc: 0.9688\n",
            "Epoch 4/5\n",
            "73440/73440 [==============================] - 89s 1ms/step - loss: 0.0692 - acc: 0.9782 - val_loss: 0.0776 - val_acc: 0.9775\n",
            "Epoch 5/5\n",
            "73440/73440 [==============================] - 89s 1ms/step - loss: 0.0581 - acc: 0.9820 - val_loss: 0.2886 - val_acc: 0.9206\n",
            "=============================================================================\n",
            "{'optimizer': 'Adam', 'kernel_initializer': 'uniform', 'activation': 'relu'}\n",
            "Train on 73440 samples, validate on 13360 samples\n",
            "Epoch 1/5\n",
            "73440/73440 [==============================] - 92s 1ms/step - loss: 0.3284 - acc: 0.9116 - val_loss: 1.2440 - val_acc: 0.7041\n",
            "Epoch 2/5\n",
            "73440/73440 [==============================] - 90s 1ms/step - loss: 0.1081 - acc: 0.9666 - val_loss: 1.4073 - val_acc: 0.7209\n",
            "Epoch 3/5\n",
            "73440/73440 [==============================] - 90s 1ms/step - loss: 0.0826 - acc: 0.9748 - val_loss: 0.5529 - val_acc: 0.8179\n",
            "Epoch 4/5\n",
            "73440/73440 [==============================] - 90s 1ms/step - loss: 0.0669 - acc: 0.9794 - val_loss: 0.0744 - val_acc: 0.9786\n",
            "Epoch 5/5\n",
            "73440/73440 [==============================] - 90s 1ms/step - loss: 0.0618 - acc: 0.9812 - val_loss: 0.0576 - val_acc: 0.9831\n",
            "=============================================================================\n",
            "{'optimizer': 'Adam', 'kernel_initializer': 'normal', 'activation': 'linear'}\n",
            "Train on 73440 samples, validate on 13360 samples\n",
            "Epoch 1/5\n",
            "73440/73440 [==============================] - 91s 1ms/step - loss: 0.6370 - acc: 0.8235 - val_loss: 2.8734 - val_acc: 0.6292\n",
            "Epoch 2/5\n",
            "73440/73440 [==============================] - 88s 1ms/step - loss: 0.2778 - acc: 0.9138 - val_loss: 0.3445 - val_acc: 0.8853\n",
            "Epoch 3/5\n",
            "73440/73440 [==============================] - 88s 1ms/step - loss: 0.2002 - acc: 0.9372 - val_loss: 0.2580 - val_acc: 0.9189\n",
            "Epoch 4/5\n",
            "73440/73440 [==============================] - 89s 1ms/step - loss: 0.1637 - acc: 0.9491 - val_loss: 2.0659 - val_acc: 0.4595\n",
            "Epoch 5/5\n",
            "73440/73440 [==============================] - 88s 1ms/step - loss: 0.1404 - acc: 0.9550 - val_loss: 0.1368 - val_acc: 0.9581\n",
            "=============================================================================\n",
            "{'optimizer': 'Adam', 'kernel_initializer': 'uniform', 'activation': 'linear'}\n",
            "Train on 73440 samples, validate on 13360 samples\n",
            "Epoch 1/5\n",
            "73440/73440 [==============================] - 92s 1ms/step - loss: 0.5804 - acc: 0.8396 - val_loss: 0.9485 - val_acc: 0.7478\n",
            "Epoch 2/5\n",
            "73440/73440 [==============================] - 90s 1ms/step - loss: 0.2529 - acc: 0.9237 - val_loss: 8.6808 - val_acc: 0.1462\n",
            "Epoch 3/5\n",
            "73440/73440 [==============================] - 88s 1ms/step - loss: 0.1825 - acc: 0.9431 - val_loss: 0.2915 - val_acc: 0.9082\n",
            "Epoch 4/5\n",
            "73440/73440 [==============================] - 87s 1ms/step - loss: 0.1507 - acc: 0.9539 - val_loss: 0.2655 - val_acc: 0.9148\n",
            "Epoch 5/5\n",
            "73440/73440 [==============================] - 87s 1ms/step - loss: 0.1319 - acc: 0.9585 - val_loss: 0.5280 - val_acc: 0.8577\n",
            "=============================================================================\n",
            "{'optimizer': 'Adam', 'kernel_initializer': 'normal', 'activation': 'tanh'}\n",
            "Train on 73440 samples, validate on 13360 samples\n",
            "Epoch 1/5\n",
            "73440/73440 [==============================] - 92s 1ms/step - loss: 0.5181 - acc: 0.8634 - val_loss: 2.0401 - val_acc: 0.3782\n",
            "Epoch 2/5\n",
            "73440/73440 [==============================] - 89s 1ms/step - loss: 0.1895 - acc: 0.9414 - val_loss: 6.7160 - val_acc: 0.3058\n",
            "Epoch 3/5\n",
            "73440/73440 [==============================] - 87s 1ms/step - loss: 0.1503 - acc: 0.9522 - val_loss: 2.1043 - val_acc: 0.6128\n",
            "Epoch 4/5\n",
            "73440/73440 [==============================] - 89s 1ms/step - loss: 0.1358 - acc: 0.9579 - val_loss: 3.8762 - val_acc: 0.6211\n",
            "Epoch 5/5\n",
            "73440/73440 [==============================] - 89s 1ms/step - loss: 0.1225 - acc: 0.9610 - val_loss: 2.0894 - val_acc: 0.5716\n",
            "=============================================================================\n",
            "{'optimizer': 'Adam', 'kernel_initializer': 'uniform', 'activation': 'tanh'}\n",
            "Train on 73440 samples, validate on 13360 samples\n",
            "Epoch 1/5\n",
            "73440/73440 [==============================] - 93s 1ms/step - loss: 0.4551 - acc: 0.8800 - val_loss: 0.4224 - val_acc: 0.8805\n",
            "Epoch 2/5\n",
            "73440/73440 [==============================] - 90s 1ms/step - loss: 0.1675 - acc: 0.9494 - val_loss: 7.1558 - val_acc: 0.2093\n",
            "Epoch 3/5\n",
            "73440/73440 [==============================] - 89s 1ms/step - loss: 0.1372 - acc: 0.9566 - val_loss: 0.2988 - val_acc: 0.9068\n",
            "Epoch 4/5\n",
            "73440/73440 [==============================] - 90s 1ms/step - loss: 0.1215 - acc: 0.9623 - val_loss: 0.4133 - val_acc: 0.8790\n",
            "Epoch 5/5\n",
            "73440/73440 [==============================] - 89s 1ms/step - loss: 0.1142 - acc: 0.9644 - val_loss: 1.3391 - val_acc: 0.6110\n",
            "=============================================================================\n",
            "{'optimizer': 'Adagrad', 'kernel_initializer': 'normal', 'activation': 'relu'}\n",
            "Train on 73440 samples, validate on 13360 samples\n",
            "Epoch 1/5\n",
            "73440/73440 [==============================] - 87s 1ms/step - loss: 0.4422 - acc: 0.8875 - val_loss: 1.2394 - val_acc: 0.6644\n",
            "Epoch 2/5\n",
            "73440/73440 [==============================] - 83s 1ms/step - loss: 0.1884 - acc: 0.9509 - val_loss: 0.2345 - val_acc: 0.9350\n",
            "Epoch 3/5\n",
            "73440/73440 [==============================] - 83s 1ms/step - loss: 0.1396 - acc: 0.9618 - val_loss: 0.1238 - val_acc: 0.9647\n",
            "Epoch 4/5\n",
            "73440/73440 [==============================] - 83s 1ms/step - loss: 0.1167 - acc: 0.9677 - val_loss: 0.2001 - val_acc: 0.9427\n",
            "Epoch 5/5\n",
            "73440/73440 [==============================] - 83s 1ms/step - loss: 0.1036 - acc: 0.9707 - val_loss: 0.1052 - val_acc: 0.9710\n",
            "=============================================================================\n",
            "{'optimizer': 'Adagrad', 'kernel_initializer': 'uniform', 'activation': 'relu'}\n",
            "Train on 73440 samples, validate on 13360 samples\n",
            "Epoch 1/5\n",
            "73440/73440 [==============================] - 87s 1ms/step - loss: 0.3526 - acc: 0.9115 - val_loss: 0.1827 - val_acc: 0.9504\n",
            "Epoch 2/5\n",
            "73440/73440 [==============================] - 83s 1ms/step - loss: 0.1402 - acc: 0.9631 - val_loss: 0.1405 - val_acc: 0.9637\n",
            "Epoch 3/5\n",
            "73440/73440 [==============================] - 84s 1ms/step - loss: 0.1052 - acc: 0.9715 - val_loss: 0.1036 - val_acc: 0.9696\n",
            "Epoch 4/5\n",
            "73440/73440 [==============================] - 83s 1ms/step - loss: 0.0890 - acc: 0.9755 - val_loss: 0.0966 - val_acc: 0.9773\n",
            "Epoch 5/5\n",
            "73440/73440 [==============================] - 84s 1ms/step - loss: 0.0780 - acc: 0.9790 - val_loss: 0.0855 - val_acc: 0.9796\n",
            "=============================================================================\n",
            "{'optimizer': 'Adagrad', 'kernel_initializer': 'normal', 'activation': 'linear'}\n",
            "Train on 73440 samples, validate on 13360 samples\n",
            "Epoch 1/5\n",
            "73440/73440 [==============================] - 86s 1ms/step - loss: 0.7344 - acc: 0.8080 - val_loss: 0.6743 - val_acc: 0.8025\n",
            "Epoch 2/5\n",
            "73440/73440 [==============================] - 82s 1ms/step - loss: 0.4684 - acc: 0.8702 - val_loss: 0.5411 - val_acc: 0.8420\n",
            "Epoch 3/5\n",
            "73440/73440 [==============================] - 84s 1ms/step - loss: 0.3942 - acc: 0.8899 - val_loss: 0.4227 - val_acc: 0.8801\n",
            "Epoch 4/5\n",
            "73440/73440 [==============================] - 88s 1ms/step - loss: 0.3465 - acc: 0.9031 - val_loss: 0.3989 - val_acc: 0.8791\n",
            "Epoch 5/5\n",
            "73440/73440 [==============================] - 85s 1ms/step - loss: 0.3127 - acc: 0.9124 - val_loss: 0.4164 - val_acc: 0.8647\n",
            "=============================================================================\n",
            "{'optimizer': 'Adagrad', 'kernel_initializer': 'uniform', 'activation': 'linear'}\n",
            "Train on 73440 samples, validate on 13360 samples\n",
            "Epoch 1/5\n",
            "73440/73440 [==============================] - 88s 1ms/step - loss: 0.6657 - acc: 0.8253 - val_loss: 0.6107 - val_acc: 0.8263\n",
            "Epoch 2/5\n",
            "73440/73440 [==============================] - 84s 1ms/step - loss: 0.4114 - acc: 0.8864 - val_loss: 0.5026 - val_acc: 0.8509\n",
            "Epoch 3/5\n",
            "73440/73440 [==============================] - 83s 1ms/step - loss: 0.3377 - acc: 0.9060 - val_loss: 0.3457 - val_acc: 0.9032\n",
            "Epoch 4/5\n",
            "73440/73440 [==============================] - 84s 1ms/step - loss: 0.2942 - acc: 0.9185 - val_loss: 0.3007 - val_acc: 0.9116\n",
            "Epoch 5/5\n",
            "73440/73440 [==============================] - 83s 1ms/step - loss: 0.2638 - acc: 0.9275 - val_loss: 0.3070 - val_acc: 0.9059\n",
            "=============================================================================\n",
            "{'optimizer': 'Adagrad', 'kernel_initializer': 'normal', 'activation': 'tanh'}\n",
            "Train on 73440 samples, validate on 13360 samples\n",
            "Epoch 1/5\n",
            "73440/73440 [==============================] - 88s 1ms/step - loss: 0.6322 - acc: 0.8420 - val_loss: 0.7225 - val_acc: 0.7845\n",
            "Epoch 2/5\n",
            "73440/73440 [==============================] - 86s 1ms/step - loss: 0.3128 - acc: 0.9185 - val_loss: 0.6741 - val_acc: 0.8092\n",
            "Epoch 3/5\n",
            "73440/73440 [==============================] - 85s 1ms/step - loss: 0.2363 - acc: 0.9372 - val_loss: 0.5194 - val_acc: 0.8432\n",
            "Epoch 4/5\n",
            "73440/73440 [==============================] - 86s 1ms/step - loss: 0.1975 - acc: 0.9462 - val_loss: 0.4444 - val_acc: 0.8707\n",
            "Epoch 5/5\n",
            "73440/73440 [==============================] - 86s 1ms/step - loss: 0.1727 - acc: 0.9527 - val_loss: 0.3372 - val_acc: 0.8906\n",
            "=============================================================================\n",
            "{'optimizer': 'Adagrad', 'kernel_initializer': 'uniform', 'activation': 'tanh'}\n",
            "Train on 73440 samples, validate on 13360 samples\n",
            "Epoch 1/5\n",
            "73440/73440 [==============================] - 90s 1ms/step - loss: 0.5182 - acc: 0.8685 - val_loss: 0.7889 - val_acc: 0.7621\n",
            "Epoch 2/5\n",
            "73440/73440 [==============================] - 86s 1ms/step - loss: 0.2390 - acc: 0.9377 - val_loss: 0.5195 - val_acc: 0.8433\n",
            "Epoch 3/5\n",
            "73440/73440 [==============================] - 85s 1ms/step - loss: 0.1814 - acc: 0.9515 - val_loss: 0.2427 - val_acc: 0.9240\n",
            "Epoch 4/5\n",
            "73440/73440 [==============================] - 85s 1ms/step - loss: 0.1508 - acc: 0.9596 - val_loss: 0.2552 - val_acc: 0.9205\n",
            "Epoch 5/5\n",
            "73440/73440 [==============================] - 85s 1ms/step - loss: 0.1354 - acc: 0.9627 - val_loss: 0.1798 - val_acc: 0.9451\n",
            "=============================================================================\n",
            "{'optimizer': 'Nadam', 'kernel_initializer': 'normal', 'activation': 'relu'}\n",
            "Train on 73440 samples, validate on 13360 samples\n",
            "Epoch 1/5\n",
            "73440/73440 [==============================] - 102s 1ms/step - loss: 0.2761 - acc: 0.9219 - val_loss: 0.1491 - val_acc: 0.9564\n",
            "Epoch 2/5\n",
            "73440/73440 [==============================] - 99s 1ms/step - loss: 0.1010 - acc: 0.9686 - val_loss: 0.2383 - val_acc: 0.9087\n",
            "Epoch 3/5\n",
            "73440/73440 [==============================] - 98s 1ms/step - loss: 0.0823 - acc: 0.9745 - val_loss: 0.1861 - val_acc: 0.9542\n",
            "Epoch 4/5\n",
            "73440/73440 [==============================] - 98s 1ms/step - loss: 0.0711 - acc: 0.9775 - val_loss: 0.2912 - val_acc: 0.9115\n",
            "Epoch 5/5\n",
            "73440/73440 [==============================] - 98s 1ms/step - loss: 0.0644 - acc: 0.9803 - val_loss: 1.5884 - val_acc: 0.7040\n",
            "=============================================================================\n",
            "{'optimizer': 'Nadam', 'kernel_initializer': 'uniform', 'activation': 'relu'}\n",
            "Train on 73440 samples, validate on 13360 samples\n",
            "Epoch 1/5\n",
            "73440/73440 [==============================] - 103s 1ms/step - loss: 0.2668 - acc: 0.9253 - val_loss: 1.5238 - val_acc: 0.7397\n",
            "Epoch 2/5\n",
            "73440/73440 [==============================] - 98s 1ms/step - loss: 0.1046 - acc: 0.9674 - val_loss: 0.1860 - val_acc: 0.9345\n",
            "Epoch 3/5\n",
            "73440/73440 [==============================] - 96s 1ms/step - loss: 0.0821 - acc: 0.9743 - val_loss: 1.4487 - val_acc: 0.6641\n",
            "Epoch 4/5\n",
            "73440/73440 [==============================] - 96s 1ms/step - loss: 0.0736 - acc: 0.9772 - val_loss: 0.4209 - val_acc: 0.8993\n",
            "Epoch 5/5\n",
            "73440/73440 [==============================] - 96s 1ms/step - loss: 0.0671 - acc: 0.9787 - val_loss: 0.1993 - val_acc: 0.9399\n",
            "=============================================================================\n",
            "{'optimizer': 'Nadam', 'kernel_initializer': 'normal', 'activation': 'linear'}\n",
            "Train on 73440 samples, validate on 13360 samples\n",
            "Epoch 1/5\n",
            "73440/73440 [==============================] - 100s 1ms/step - loss: 0.5131 - acc: 0.8516 - val_loss: 6.4008 - val_acc: 0.1222\n",
            "Epoch 2/5\n",
            "73440/73440 [==============================] - 97s 1ms/step - loss: 0.2208 - acc: 0.9295 - val_loss: 0.8846 - val_acc: 0.7870\n",
            "Epoch 3/5\n",
            "73440/73440 [==============================] - 105s 1ms/step - loss: 0.1642 - acc: 0.9481 - val_loss: 0.2441 - val_acc: 0.9202\n",
            "Epoch 4/5\n",
            "73440/73440 [==============================] - 101s 1ms/step - loss: 0.1412 - acc: 0.9546 - val_loss: 5.7308 - val_acc: 0.2032\n",
            "Epoch 5/5\n",
            "73440/73440 [==============================] - 96s 1ms/step - loss: 0.1227 - acc: 0.9603 - val_loss: 3.0600 - val_acc: 0.7353\n",
            "=============================================================================\n",
            "{'optimizer': 'Nadam', 'kernel_initializer': 'uniform', 'activation': 'linear'}\n",
            "Train on 73440 samples, validate on 13360 samples\n",
            "Epoch 1/5\n",
            "73440/73440 [==============================] - 102s 1ms/step - loss: 0.4945 - acc: 0.8586 - val_loss: 7.1451 - val_acc: 0.3430\n",
            "Epoch 2/5\n",
            "73440/73440 [==============================] - 97s 1ms/step - loss: 0.2141 - acc: 0.9326 - val_loss: 0.9167 - val_acc: 0.7728\n",
            "Epoch 3/5\n",
            "73440/73440 [==============================] - 97s 1ms/step - loss: 0.1628 - acc: 0.9480 - val_loss: 3.7391 - val_acc: 0.7296\n",
            "Epoch 4/5\n",
            "73440/73440 [==============================] - 96s 1ms/step - loss: 0.1380 - acc: 0.9554 - val_loss: 0.1289 - val_acc: 0.9596\n",
            "Epoch 5/5\n",
            "73440/73440 [==============================] - 97s 1ms/step - loss: 0.1233 - acc: 0.9607 - val_loss: 1.4198 - val_acc: 0.6144\n",
            "=============================================================================\n",
            "{'optimizer': 'Nadam', 'kernel_initializer': 'normal', 'activation': 'tanh'}\n",
            "Train on 73440 samples, validate on 13360 samples\n",
            "Epoch 1/5\n",
            "73440/73440 [==============================] - 104s 1ms/step - loss: 0.4371 - acc: 0.8779 - val_loss: 2.2470 - val_acc: 0.5817\n",
            "Epoch 2/5\n",
            "73440/73440 [==============================] - 99s 1ms/step - loss: 0.2140 - acc: 0.9309 - val_loss: 9.5446 - val_acc: 0.0624\n",
            "Epoch 3/5\n",
            "73440/73440 [==============================] - 98s 1ms/step - loss: 0.1971 - acc: 0.9371 - val_loss: 1.6289 - val_acc: 0.5686\n",
            "Epoch 4/5\n",
            "73440/73440 [==============================] - 99s 1ms/step - loss: 0.1771 - acc: 0.9431 - val_loss: 0.4576 - val_acc: 0.8562\n",
            "Epoch 5/5\n",
            "73440/73440 [==============================] - 98s 1ms/step - loss: 0.1667 - acc: 0.9460 - val_loss: 0.1962 - val_acc: 0.9391\n",
            "=============================================================================\n",
            "{'optimizer': 'Nadam', 'kernel_initializer': 'uniform', 'activation': 'tanh'}\n",
            "Train on 73440 samples, validate on 13360 samples\n",
            "Epoch 1/5\n",
            "73440/73440 [==============================] - 105s 1ms/step - loss: 0.3933 - acc: 0.8903 - val_loss: 6.5998 - val_acc: 0.1526\n",
            "Epoch 2/5\n",
            "73440/73440 [==============================] - 98s 1ms/step - loss: 0.2030 - acc: 0.9367 - val_loss: 5.0520 - val_acc: 0.4349\n",
            "Epoch 3/5\n",
            "73440/73440 [==============================] - 99s 1ms/step - loss: 0.1776 - acc: 0.9434 - val_loss: 0.7476 - val_acc: 0.7984\n",
            "Epoch 4/5\n",
            "73440/73440 [==============================] - 99s 1ms/step - loss: 0.1678 - acc: 0.9457 - val_loss: 0.8716 - val_acc: 0.7469\n",
            "Epoch 5/5\n",
            "73440/73440 [==============================] - 98s 1ms/step - loss: 0.1530 - acc: 0.9510 - val_loss: 2.0540 - val_acc: 0.5690\n",
            "=============================================================================\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "TgVoF_U9ifx0",
        "colab_type": "text"
      },
      "source": [
        "**From the above results we can see that best parameters are:**\n",
        "* **Optimizer: Adam**\n",
        "* **Kernel_initializer: uniform**\n",
        "* **Activation: relu** "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Nt9JJvWzjKpf",
        "colab_type": "text"
      },
      "source": [
        "Let's create the model with the best parameters obtained."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TQ61tTgthNX3",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model = create_model(optimizer='Adam', kernel_initializer='uniform', activation='relu')"
      ],
      "execution_count": 28,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "SAyJLJOsJdKY",
        "colab_type": "text"
      },
      "source": [
        "## Training the Model\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "9REBW24cmdKI",
        "colab_type": "text"
      },
      "source": [
        "### Fitting the Model"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "i8GiTDtRkZTp",
        "colab_type": "text"
      },
      "source": [
        "Train the model using batch_size=20 to reduce used memory and make the training more quick.\n",
        "We will train the model first on 10 epochs to see the accuracy that we will obtain."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fOtKJTQ51qC5",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 734
        },
        "outputId": "7cad5b4b-6c0d-467a-d3e8-b39cf0fb08a9"
      },
      "source": [
        "from keras.callbacks import ModelCheckpoint  \n",
        "\n",
        "# using checkpoints to save model weights to be used later instead of training again on the same epochs.\n",
        "checkpointer = ModelCheckpoint(filepath='weights.hdf5', verbose=1, save_best_only=True)\n",
        "history = model.fit(training_data_images, training_data_labels, \n",
        "                    validation_data=(testing_data_images, testing_data_labels),\n",
        "                    epochs=10, batch_size=20, verbose=1, callbacks=[checkpointer])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Train on 73440 samples, validate on 13360 samples\n",
            "Epoch 1/10\n",
            "73440/73440 [==============================] - 94s 1ms/step - loss: 0.3189 - acc: 0.9153 - val_loss: 10.5339 - val_acc: 0.1437\n",
            "\n",
            "Epoch 00001: val_loss improved from inf to 10.53390, saving model to weights.hdf5\n",
            "Epoch 2/10\n",
            "73440/73440 [==============================] - 89s 1ms/step - loss: 0.1049 - acc: 0.9681 - val_loss: 0.3283 - val_acc: 0.9012\n",
            "\n",
            "Epoch 00002: val_loss improved from 10.53390 to 0.32826, saving model to weights.hdf5\n",
            "Epoch 3/10\n",
            "73440/73440 [==============================] - 90s 1ms/step - loss: 0.0797 - acc: 0.9757 - val_loss: 0.6523 - val_acc: 0.7985\n",
            "\n",
            "Epoch 00003: val_loss did not improve from 0.32826\n",
            "Epoch 4/10\n",
            "73440/73440 [==============================] - 89s 1ms/step - loss: 0.0702 - acc: 0.9782 - val_loss: 0.2071 - val_acc: 0.9394\n",
            "\n",
            "Epoch 00004: val_loss improved from 0.32826 to 0.20706, saving model to weights.hdf5\n",
            "Epoch 5/10\n",
            "73440/73440 [==============================] - 89s 1ms/step - loss: 0.0592 - acc: 0.9811 - val_loss: 0.0752 - val_acc: 0.9769\n",
            "\n",
            "Epoch 00005: val_loss improved from 0.20706 to 0.07522, saving model to weights.hdf5\n",
            "Epoch 6/10\n",
            "73440/73440 [==============================] - 89s 1ms/step - loss: 0.0545 - acc: 0.9833 - val_loss: 0.9768 - val_acc: 0.6818\n",
            "\n",
            "Epoch 00006: val_loss did not improve from 0.07522\n",
            "Epoch 7/10\n",
            "73440/73440 [==============================] - 88s 1ms/step - loss: 0.0497 - acc: 0.9845 - val_loss: 0.1118 - val_acc: 0.9671\n",
            "\n",
            "Epoch 00007: val_loss did not improve from 0.07522\n",
            "Epoch 8/10\n",
            "73440/73440 [==============================] - 89s 1ms/step - loss: 0.0473 - acc: 0.9854 - val_loss: 0.0576 - val_acc: 0.9829\n",
            "\n",
            "Epoch 00008: val_loss improved from 0.07522 to 0.05756, saving model to weights.hdf5\n",
            "Epoch 9/10\n",
            "73440/73440 [==============================] - 89s 1ms/step - loss: 0.0453 - acc: 0.9863 - val_loss: 0.0800 - val_acc: 0.9767\n",
            "\n",
            "Epoch 00009: val_loss did not improve from 0.05756\n",
            "Epoch 10/10\n",
            "73440/73440 [==============================] - 89s 1ms/step - loss: 0.0432 - acc: 0.9862 - val_loss: 0.2793 - val_acc: 0.9097\n",
            "\n",
            "Epoch 00010: val_loss did not improve from 0.05756\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "etE1oULFCnSn",
        "colab_type": "text"
      },
      "source": [
        "### Plotting Loss and Accuracy Curves with Epochs"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZWyxaeKxCn9O",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "\n",
        "def plot_loss_accuracy(history):\n",
        "  # Loss Curves\n",
        "  plt.figure(figsize=[8,6])\n",
        "  plt.plot(history.history['loss'],'r',linewidth=3.0)\n",
        "  plt.plot(history.history['val_loss'],'b',linewidth=3.0)\n",
        "  plt.legend(['Training loss', 'Validation Loss'],fontsize=18)\n",
        "  plt.xlabel('Epochs ',fontsize=16)\n",
        "  plt.ylabel('Loss',fontsize=16)\n",
        "  plt.title('Loss Curves',fontsize=16)\n",
        "\n",
        "  # Accuracy Curves\n",
        "  plt.figure(figsize=[8,6])\n",
        "  plt.plot(history.history['acc'],'r',linewidth=3.0)\n",
        "  plt.plot(history.history['val_acc'],'b',linewidth=3.0)\n",
        "  plt.legend(['Training Accuracy', 'Validation Accuracy'],fontsize=18)\n",
        "  plt.xlabel('Epochs ',fontsize=16)\n",
        "  plt.ylabel('Accuracy',fontsize=16)\n",
        "  plt.title('Accuracy Curves',fontsize=16) "
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LE9WKyG7GNSs",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 805
        },
        "outputId": "d610c15c-18ac-428c-9193-40a87bccce15"
      },
      "source": [
        "plot_loss_accuracy(history)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfIAAAGKCAYAAAD34MN8AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n5AAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<matplotlib.figure.Figure at 0x7fa74977aac8>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfQAAAGKCAYAAAD6/rM7AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<matplotlib.figure.Figure at 0x7fa748f28630>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "p8QOKiQpmWDC",
        "colab_type": "text"
      },
      "source": [
        "### Load the Model with the Best Validation Loss\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "OUunssvzmQUb",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model.load_weights('weights.hdf5')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Aijsry3Xmmv-",
        "colab_type": "text"
      },
      "source": [
        "## Test the Model\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8h7E8G7CsCFB",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 68
        },
        "outputId": "ce7d4c98-e32c-4d49-8c4d-4a7bef7f3c30"
      },
      "source": [
        "# Final evaluation of the model\n",
        "metrics = model.evaluate(testing_data_images, testing_data_labels, verbose=1)\n",
        "print(\"Test Accuracy: {}\".format(metrics[1]))\n",
        "print(\"Test Loss: {}\".format(metrics[0]))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "13360/13360 [==============================] - 5s 343us/step\n",
            "Test Accuracy: 0.9828592814371258\n",
            "Test Loss: 0.05756271482567684\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Xz04uhKPnFft",
        "colab_type": "text"
      },
      "source": [
        "**We get test accuracy of 98.286% after training on 10 epochs only.**\n",
        "\n",
        "**What about increasing the epochs we train on ?**"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Rit6CswqnrEC",
        "colab_type": "text"
      },
      "source": [
        "## Training More on the best model"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BZAP7VW1npA-",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 2202
        },
        "outputId": "0fc06279-2faa-4d2b-f358-edda4c304338"
      },
      "source": [
        "epochs = 20\n",
        "batch_size = 20\n",
        "\n",
        "checkpointer = ModelCheckpoint(filepath='weights.hdf5', verbose=1, save_best_only=True)\n",
        "\n",
        "history = model.fit(training_data_images, training_data_labels, \n",
        "                    validation_data=(testing_data_images, testing_data_labels),\n",
        "                    epochs=epochs, batch_size=batch_size, verbose=1, callbacks=[checkpointer])\n",
        "          \n",
        "model.load_weights('weights.hdf5')\n",
        "plot_loss_accuracy(history)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Train on 73440 samples, validate on 13360 samples\n",
            "Epoch 1/20\n",
            "73440/73440 [==============================] - 89s 1ms/step - loss: 0.0447 - acc: 0.9859 - val_loss: 0.4244 - val_acc: 0.8557\n",
            "\n",
            "Epoch 00001: val_loss improved from inf to 0.42443, saving model to weights.hdf5\n",
            "Epoch 2/20\n",
            "73440/73440 [==============================] - 89s 1ms/step - loss: 0.0414 - acc: 0.9863 - val_loss: 0.0576 - val_acc: 0.9830\n",
            "\n",
            "Epoch 00002: val_loss improved from 0.42443 to 0.05760, saving model to weights.hdf5\n",
            "Epoch 3/20\n",
            "73440/73440 [==============================] - 89s 1ms/step - loss: 0.0396 - acc: 0.9872 - val_loss: 0.0546 - val_acc: 0.9844\n",
            "\n",
            "Epoch 00003: val_loss improved from 0.05760 to 0.05457, saving model to weights.hdf5\n",
            "Epoch 4/20\n",
            "73440/73440 [==============================] - 89s 1ms/step - loss: 0.0400 - acc: 0.9875 - val_loss: 0.4204 - val_acc: 0.8711\n",
            "\n",
            "Epoch 00004: val_loss did not improve from 0.05457\n",
            "Epoch 5/20\n",
            "73440/73440 [==============================] - 89s 1ms/step - loss: 0.0367 - acc: 0.9881 - val_loss: 2.2243 - val_acc: 0.5124\n",
            "\n",
            "Epoch 00005: val_loss did not improve from 0.05457\n",
            "Epoch 6/20\n",
            "73440/73440 [==============================] - 89s 1ms/step - loss: 0.0366 - acc: 0.9881 - val_loss: 0.0553 - val_acc: 0.9842\n",
            "\n",
            "Epoch 00006: val_loss did not improve from 0.05457\n",
            "Epoch 7/20\n",
            "73440/73440 [==============================] - 89s 1ms/step - loss: 0.0349 - acc: 0.9889 - val_loss: 0.0449 - val_acc: 0.9877\n",
            "\n",
            "Epoch 00007: val_loss improved from 0.05457 to 0.04494, saving model to weights.hdf5\n",
            "Epoch 8/20\n",
            "73440/73440 [==============================] - 89s 1ms/step - loss: 0.0320 - acc: 0.9895 - val_loss: 0.0504 - val_acc: 0.9858\n",
            "\n",
            "Epoch 00008: val_loss did not improve from 0.04494\n",
            "Epoch 9/20\n",
            "73440/73440 [==============================] - 89s 1ms/step - loss: 0.0326 - acc: 0.9897 - val_loss: 0.1189 - val_acc: 0.9697\n",
            "\n",
            "Epoch 00009: val_loss did not improve from 0.04494\n",
            "Epoch 10/20\n",
            "73440/73440 [==============================] - 89s 1ms/step - loss: 0.0300 - acc: 0.9904 - val_loss: 0.0488 - val_acc: 0.9871\n",
            "\n",
            "Epoch 00010: val_loss did not improve from 0.04494\n",
            "Epoch 11/20\n",
            "73440/73440 [==============================] - 89s 1ms/step - loss: 0.0309 - acc: 0.9899 - val_loss: 0.0511 - val_acc: 0.9853\n",
            "\n",
            "Epoch 00011: val_loss did not improve from 0.04494\n",
            "Epoch 12/20\n",
            "73440/73440 [==============================] - 89s 1ms/step - loss: 0.0294 - acc: 0.9907 - val_loss: 0.0419 - val_acc: 0.9886\n",
            "\n",
            "Epoch 00012: val_loss improved from 0.04494 to 0.04191, saving model to weights.hdf5\n",
            "Epoch 13/20\n",
            "73440/73440 [==============================] - 89s 1ms/step - loss: 0.0288 - acc: 0.9906 - val_loss: 2.7113 - val_acc: 0.7689\n",
            "\n",
            "Epoch 00013: val_loss did not improve from 0.04191\n",
            "Epoch 14/20\n",
            "73440/73440 [==============================] - 89s 1ms/step - loss: 0.0277 - acc: 0.9911 - val_loss: 0.0617 - val_acc: 0.9846\n",
            "\n",
            "Epoch 00014: val_loss did not improve from 0.04191\n",
            "Epoch 15/20\n",
            "73440/73440 [==============================] - 89s 1ms/step - loss: 0.0258 - acc: 0.9918 - val_loss: 0.0939 - val_acc: 0.9754\n",
            "\n",
            "Epoch 00015: val_loss did not improve from 0.04191\n",
            "Epoch 16/20\n",
            "73440/73440 [==============================] - 89s 1ms/step - loss: 0.0265 - acc: 0.9912 - val_loss: 0.2457 - val_acc: 0.9448\n",
            "\n",
            "Epoch 00016: val_loss did not improve from 0.04191\n",
            "Epoch 17/20\n",
            "73440/73440 [==============================] - 89s 1ms/step - loss: 0.0251 - acc: 0.9918 - val_loss: 0.0505 - val_acc: 0.9866\n",
            "\n",
            "Epoch 00017: val_loss did not improve from 0.04191\n",
            "Epoch 18/20\n",
            "73440/73440 [==============================] - 89s 1ms/step - loss: 0.0271 - acc: 0.9916 - val_loss: 0.0441 - val_acc: 0.9883\n",
            "\n",
            "Epoch 00018: val_loss did not improve from 0.04191\n",
            "Epoch 19/20\n",
            "73440/73440 [==============================] - 89s 1ms/step - loss: 0.0251 - acc: 0.9917 - val_loss: 0.5342 - val_acc: 0.8984\n",
            "\n",
            "Epoch 00019: val_loss did not improve from 0.04191\n",
            "Epoch 20/20\n",
            "73440/73440 [==============================] - 89s 1ms/step - loss: 0.0239 - acc: 0.9925 - val_loss: 0.0499 - val_acc: 0.9877\n",
            "\n",
            "Epoch 00020: val_loss did not improve from 0.04191\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfQAAAGKCAYAAAD6/rM7AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<matplotlib.figure.Figure at 0x7fa74b8b1ac8>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfQAAAGKCAYAAAD6/rM7AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<matplotlib.figure.Figure at 0x7fa748e4bcc0>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "c8GWs-8bNsxR",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 805
        },
        "outputId": "acef3123-a33c-49bc-f759-5ce10a9d50a7"
      },
      "source": [
        "plot_loss_accuracy(history)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfQAAAGKCAYAAAD6/rM7AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<matplotlib.figure.Figure at 0x7fa748ea72e8>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfQAAAGKCAYAAAD6/rM7AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<matplotlib.figure.Figure at 0x7fa748d77320>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "RC6yRcCHoMa7",
        "colab_type": "text"
      },
      "source": [
        "## Testing the Model again"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Bm5y2MgXoP3F",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 68
        },
        "outputId": "4c262179-d450-4c3e-e621-a86814f8caa7"
      },
      "source": [
        "# Final evaluation of the model\n",
        "metrics = model.evaluate(testing_data_images, testing_data_labels, verbose=1)\n",
        "print(\"Test Accuracy: {}\".format(metrics[1]))\n",
        "print(\"Test Loss: {}\".format(metrics[0]))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "13360/13360 [==============================] - 5s 346us/step\n",
            "Test Accuracy: 0.988622754491018\n",
            "Test Loss: 0.04191313635113142\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3cVhqFs8oYwz",
        "colab_type": "text"
      },
      "source": [
        "After training the model on more epochs we gained a better model which can classify complex patterns . So when we tested it on our test dataset we had better results than before.\n",
        "\n",
        "**Test accuracy is improved from 98.286% to 98.862% As we train the model on 20 more epochs.**\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "RUes2bqFqycc",
        "colab_type": "text"
      },
      "source": [
        "## Saving the Final Model"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "dA6vbYZKp0bB",
        "colab_type": "text"
      },
      "source": [
        "Let's save the model on json format to be used later instead of creating the model again from scratch."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZtanngD4oPgv",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from keras.models import model_from_yaml"
      ],
      "execution_count": 23,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EVsTrDhop_XA",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model_yaml = model.to_yaml()\n",
        "with open(\"model.yaml\", \"w\") as yaml_file:\n",
        "    yaml_file.write(model_yaml)"
      ],
      "execution_count": 29,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ozgwPPrArhoY",
        "colab_type": "text"
      },
      "source": [
        "Save the model weights to file."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jAVASfDAo_lN",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "efc9bb9b-444a-4191-a15f-9fed644fef7f"
      },
      "source": [
        "# serialize weights to HDF5\n",
        "model.save_weights(\"model.h5\")\n",
        "print(\"Saved model to disk\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Saved model to disk\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "xv1bAB-ur15m",
        "colab_type": "text"
      },
      "source": [
        "if we want to load the model with the last obtained weights at anytime, we will run the following code cell."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "xD9ogMjqsCvJ",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "de36a321-39bd-41fc-df6a-895f579646a9"
      },
      "source": [
        "# load YAML and create model\n",
        "yaml_file = open('model.yaml', 'r')\n",
        "loaded_model_yaml = yaml_file.read()\n",
        "yaml_file.close()\n",
        "loaded_model = model_from_yaml(loaded_model_yaml)\n",
        "# load weights into new model\n",
        "loaded_model.load_weights(\"model.h5\")\n",
        "print(\"Loaded model from disk\")\n",
        " \n",
        "# compile the loaded model\n",
        "loaded_model.compile(loss='binary_crossentropy', optimizer='Adam', metrics=['accuracy'])"
      ],
      "execution_count": 30,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Loaded model from disk\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "bvOiQNT-u6DQ",
        "colab_type": "text"
      },
      "source": [
        "## Benchmark Model"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "VljY9PAbwWfo",
        "colab_type": "text"
      },
      "source": [
        "We will use a very simple (vanilla) CNN model as benchmark and Train/test it using the same data that you have used for our model solution.\n",
        "Then Compare the results between the vanilla model and our complex model."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "r3uAvHHzwW8D",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 275
        },
        "outputId": "e70b620d-73ef-4b69-dfb1-9719dc8ff6a0"
      },
      "source": [
        "baseline_model = Sequential()\n",
        "baseline_model.add(Conv2D(filters=16, kernel_size=3, padding='same', input_shape=(64, 64, 1), activation='relu')) # Input layer\n",
        "baseline_model.add(GlobalAveragePooling2D())\n",
        "baseline_model.add(Dense(38, activation = 'softmax')) # Output layer => output dimension = 38 as it is multi-class\n",
        "\n",
        "# Compile the baseline model\n",
        "baseline_model.compile(loss='categorical_crossentropy', metrics=['accuracy'], optimizer='Adam')\n",
        "\n",
        "# Fit the baseline model with training dataset\n",
        "epochs = 5\n",
        "batch_size = 20\n",
        "\n",
        "baseline_model.fit(training_data_images, training_data_labels, \n",
        "                  validation_data=(testing_data_images, testing_data_labels),\n",
        "                  epochs=epochs, batch_size=batch_size, verbose=1)\n",
        "\n",
        "# Test the baseline model\n",
        "baseline_metrics = baseline_model.evaluate(testing_data_images, testing_data_labels, verbose=1)\n",
        "print(\"Baseline Model Test Accuracy: {}\".format(baseline_metrics[1]))\n",
        "print(\"Baseline Model Test Loss: {}\".format(baseline_metrics[0]))\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Train on 73440 samples, validate on 13360 samples\n",
            "Epoch 1/5\n",
            "73440/73440 [==============================] - 45s 619us/step - loss: 2.7163 - acc: 0.1814 - val_loss: 2.7079 - val_acc: 0.2183\n",
            "Epoch 2/5\n",
            "73440/73440 [==============================] - 40s 550us/step - loss: 2.4709 - acc: 0.2514 - val_loss: 2.6120 - val_acc: 0.2372\n",
            "Epoch 3/5\n",
            "73440/73440 [==============================] - 40s 549us/step - loss: 2.3945 - acc: 0.2754 - val_loss: 2.5453 - val_acc: 0.2608\n",
            "Epoch 4/5\n",
            "73440/73440 [==============================] - 41s 552us/step - loss: 2.3237 - acc: 0.3069 - val_loss: 2.4661 - val_acc: 0.2959\n",
            "Epoch 5/5\n",
            "73440/73440 [==============================] - 40s 550us/step - loss: 2.2357 - acc: 0.3444 - val_loss: 2.3761 - val_acc: 0.3237\n",
            "13360/13360 [==============================] - 2s 175us/step\n",
            "Baseline Model Test Accuracy: 0.32372754491017963\n",
            "Baseline Model Test Loss: 2.376121672327647\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "x2vaf6pNSiBj",
        "colab_type": "text"
      },
      "source": [
        "**We get test accuracy of 32.37% from the baseline Model (vanilla). **"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "VKGKogPCMhZM",
        "colab_type": "text"
      },
      "source": [
        "## Predict Image Classes"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_TEpbiLtEtaW",
        "colab_type": "text"
      },
      "source": [
        "Making a method which takes a model, data and its true labels (optional for using in testing). Then it gives the predicted classes of the given data using the given model."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dXQEPxpGBJao",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def get_predicted_classes(model, data, labels=None):\n",
        "  image_predictions = model.predict(data)\n",
        "  predicted_classes = np.argmax(image_predictions, axis=1)\n",
        "  true_classes = np.argmax(labels, axis=1)\n",
        "  return predicted_classes, true_classes"
      ],
      "execution_count": 31,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "mA_tBIt3A3xr",
        "colab_type": "text"
      },
      "source": [
        "## Comparing Evaluation Metrics between Benchmark Model and Final Model"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "LSPG55usTdyZ",
        "colab_type": "text"
      },
      "source": [
        "Making a method which will print all metrics (precision, recall, f1-score and support) with each class in the dataset."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rS1nOxCQSImh",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from sklearn.metrics import classification_report\n",
        "\n",
        "def get_classification_report(y_true, y_pred):\n",
        "  print(classification_report(y_true, y_pred))"
      ],
      "execution_count": 32,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "FdzM8mCUT0m1",
        "colab_type": "text"
      },
      "source": [
        "### Evaluating Final Model using the specified metrics"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HgR-5MsXS0O3",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "y_pred, y_true = get_predicted_classes(model, testing_data_images, testing_data_labels)\n",
        "get_classification_report(y_true, y_pred)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "MlYyVo8uUJqf",
        "colab_type": "text"
      },
      "source": [
        "### Evaluating Baseline (Benchmark) Model using the specified metrics"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "AbG3WPovTFQD",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 802
        },
        "outputId": "ed6edb73-41ae-4766-ece2-3143af66f389"
      },
      "source": [
        "y_pred, y_true = get_predicted_classes(baseline_model, testing_data_images, testing_data_labels)\n",
        "get_classification_report(y_true, y_pred)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "             precision    recall  f1-score   support\n",
            "\n",
            "          0       0.95      0.84      0.89      1000\n",
            "          1       0.25      0.90      0.39      1000\n",
            "          2       0.29      0.23      0.26      1000\n",
            "          3       0.24      0.28      0.25      1000\n",
            "          4       0.31      0.29      0.30      1000\n",
            "          5       0.52      0.73      0.61      1000\n",
            "          6       0.18      0.37      0.24      1000\n",
            "          7       0.28      0.64      0.39      1000\n",
            "          8       0.00      0.00      0.00      1000\n",
            "          9       0.16      0.03      0.04      1000\n",
            "         10       0.00      0.00      0.00       120\n",
            "         11       0.00      0.00      0.00       120\n",
            "         12       0.00      0.00      0.00       120\n",
            "         13       0.00      0.00      0.00       120\n",
            "         14       0.00      0.00      0.00       120\n",
            "         15       0.00      0.00      0.00       120\n",
            "         16       0.00      0.00      0.00       120\n",
            "         17       0.48      0.18      0.27       120\n",
            "         18       0.00      0.00      0.00       120\n",
            "         19       0.00      0.00      0.00       120\n",
            "         20       0.00      0.00      0.00       120\n",
            "         21       0.00      0.00      0.00       120\n",
            "         22       0.00      0.00      0.00       120\n",
            "         23       0.00      0.00      0.00       120\n",
            "         24       0.00      0.00      0.00       120\n",
            "         25       0.00      0.00      0.00       120\n",
            "         26       0.00      0.00      0.00       120\n",
            "         27       0.00      0.00      0.00       120\n",
            "         28       0.00      0.00      0.00       120\n",
            "         29       0.00      0.00      0.00       120\n",
            "         30       0.00      0.00      0.00       120\n",
            "         31       0.00      0.00      0.00       120\n",
            "         32       0.00      0.00      0.00       120\n",
            "         33       0.00      0.00      0.00       120\n",
            "         34       0.00      0.00      0.00       120\n",
            "         35       0.00      0.00      0.00       120\n",
            "         36       0.00      0.00      0.00       120\n",
            "         37       0.00      0.00      0.00       120\n",
            "\n",
            "avg / total       0.24      0.32      0.26     13360\n",
            "\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/sklearn/metrics/classification.py:1135: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples.\n",
            "  'precision', 'predicted', average, warn_for)\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "X_Kf3t3fpvwD",
        "colab_type": "text"
      },
      "source": [
        "## A Demo to see how the model works on examples of the testing data"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "gfa_fZNbsNSf",
        "colab_type": "text"
      },
      "source": [
        "Let's recall:\n",
        "* Digits from 0 to 9 were encoded to categorical labels from 0 to 9\n",
        "* Letters from ’alef’='أ' to ’yeh’='ى' were encoded to categorical labels from 10 to 37\n",
        "\n",
        "**Recall that we have used One Hot Encoding which transforms integer to a binary matrix where the array contains only one ‘1’ and the rest elements are ‘0’.**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "28J4tcx60-lO",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def get_non_zero_index_from_one_hot_encoding(one_hot_encoding):\n",
        "  non_zero_index = np.where(one_hot_encoding == 1)[0]\n",
        "  assert(len(non_zero_index) == 1)\n",
        "  non_zero_index = non_zero_index[0]\n",
        "  return non_zero_index"
      ],
      "execution_count": 163,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "xuAddVtcsCK4",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def convert_categorical_label_to_real_label(categorical_label):\n",
        "  real_labels = []\n",
        "  real_labels.extend([x for x in range(10)])\n",
        "  real_labels.extend(['أ', 'ب', 'ت', 'ث', 'ج', 'ح', 'خ', 'د', 'ذ', 'ر', 'ز', 'س', 'ش', 'ص', 'ض', 'ط', 'ظ', 'ع', 'غ', 'ف', 'ق', 'ك', 'ل', 'م', 'ن', 'ه', 'و', 'ى'])\n",
        "  return real_labels[categorical_label]"
      ],
      "execution_count": 164,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vyyZNtjn16Be",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def show_example(sample_index):\n",
        "  y_pred, y_true = get_predicted_classes(model, testing_data_images[[sample_index]], testing_data_labels[[sample_index]])\n",
        "  non_zero_index = get_non_zero_index_from_one_hot_encoding(testing_data_labels[sample_index])\n",
        "  y_true = y_true[0]\n",
        "  y_pred = y_pred[0]\n",
        "  assert y_true == non_zero_index\n",
        "  true_label = convert_categorical_label_to_real_label(y_true)\n",
        "  predicted_label = convert_categorical_label_to_real_label(y_pred)\n",
        "  print(\"The following image has the written character '{}' but the model predicted it as '{}'\".format(true_label, predicted_label))\n",
        "  return true_label == predicted_label"
      ],
      "execution_count": 165,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bBMKkXB6ruso",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 98
        },
        "outputId": "5309c76d-c266-4d35-fbb3-d0276acec7d0"
      },
      "source": [
        "sample_index = 5\n",
        "show_example(sample_index)\n",
        "convert_values_to_image(testing_digits_images.loc[sample_index], True)"
      ],
      "execution_count": 166,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "The following image has the written character '5' but the model predicted it as '5'\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<PIL.Image.Image image mode=L size=64x64 at 0x7F5A9EC47278>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 166
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dQIeiAXi4TPh",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 98
        },
        "outputId": "a6005a3f-1e64-47ab-b5b8-101aefb9bc70"
      },
      "source": [
        "sample_index = 10002\n",
        "show_example(sample_index)\n",
        "convert_values_to_image(testing_letters_images.loc[sample_index - testing_digits_images.shape[0]], True)"
      ],
      "execution_count": 167,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "The following image has the written character 'ب' but the model predicted it as 'ب'\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<PIL.Image.Image image mode=L size=64x64 at 0x7F5B4154CD30>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 167
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "o5KU-tgO5bIt",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 98
        },
        "outputId": "e8fd2100-61bf-44e2-ef27-636d1cc0e102"
      },
      "source": [
        "sample_index = 11414\n",
        "show_example(sample_index)\n",
        "convert_values_to_image(testing_letters_images.loc[sample_index - testing_digits_images.shape[0]], True)"
      ],
      "execution_count": 168,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "The following image has the written character 'د' but the model predicted it as 'د'\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAEAAAABACAAAAACPAi4CAAABMElEQVR4nO2TTU/CMBjH2+6l3QCxsIAm4zCJ4sWL3vj8mnjyYDQawfgalBGCLNsYG+38AuuaLXrr7/o8/T3/J20BUCgUCsWfAaUdGkGMZVxU1qUCfICTeLWtLYDNYSddPAgjyAQQOeMRmvysgcAgXwHApktuAn9TT5Dz5TVwvTG6EgiQbH4ePj9+8eHpvlZ8YfIVkpl1eeaOKN6yWgkACxdPn1aPNrTCslwAQDL7Bi2HmoXF8hWgjpDZ6LmdPMl4Xl0ADZvg7vHJRZ+/vc+LH6NYoBG73e7aOvUGDpvezaOqCcihd340sDIG4tuP+xc/ryqAOm7t2UYar/3p62S5KT5f8p0J7XvUSFdBEMVxlDKBQJxgF/kMsyQI40zYU5oA6pqJcp6xnWC2QqFQKBT/wC/3OG0e9vIt+QAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<PIL.Image.Image image mode=L size=64x64 at 0x7F5B4154CDA0>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 168
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "FMrOKy46Ua1d",
        "colab_type": "text"
      },
      "source": [
        "## Conclusion"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "T1UPKIl9Uew7",
        "colab_type": "text"
      },
      "source": [
        "**We built a CNN model which can classify the arabic handwritten images into digits and letters. We tested the model on more than 13000 image with all possible classes and got very high accuracy of 98.86%.**"
      ]
    }
  ]
}