RNN,LSTM,GRU.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "e3fikzDRtSEc"
      },
      "source": [
        "One protein sequence can have many functions and can thus be classified into any number of terms. Each term is uniquely identified by a GO Term ID. Thus our model has to predict all the GO Term IDs for a protein sequence. This means that the task at hand is a multi-label classification problem."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ha42g4Gl-FIz"
      },
      "source": [
        "# **Data Preparation for Modeling**"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "#### Mount Google Drive to access datasets"
      ],
      "metadata": {
        "id": "rvQkUu5kEt28"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "IWXCqJRe-Jdw",
        "outputId": "3de74128-fa6b-477c-c771-b577ca80b288"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Mounted at /content/drive\n"
          ]
        }
      ],
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "####  Import libraries\n"
      ],
      "metadata": {
        "id": "55o7cK1mEnbV"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#Environnement Setup\n",
        "\n",
        "import numpy as np\n",
        "import pandas as pd\n",
        "import tensorflow as tf\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "from IPython.display import Image\n",
        "from tensorflow.keras.models import load_model\n",
        "from tensorflow.keras.utils import plot_model\n",
        "from tensorflow.keras import layers, Sequential\n",
        "from tensorflow.keras.callbacks import ModelCheckpoint\n",
        "\n",
        "from sklearn.metrics import precision_recall_curve, average_precision_score\n",
        "from sklearn.metrics import f1_score, precision_recall_curve, auc\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.metrics import confusion_matrix, ConfusionMatrixDisplay\n"
      ],
      "metadata": {
        "id": "DQ4Vn58ABtUv"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "#### converting embeddings numpy array(train_embeds) into pandas dataframe"
      ],
      "metadata": {
        "id": "df61QfWvFEEA"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "FC4SBi9MGEBI"
      },
      "outputs": [],
      "source": [
        "# Paths to the files\n",
        "path_train_embeds = '/content/drive/My Drive/DATASET/EMB/train_embeds.npy'\n",
        "# loading data\n",
        "train_embeds = np.load(path_train_embeds)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 291
        },
        "id": "spD8DrX-CDiu",
        "outputId": "e3e96b49-fe09-4447-ec31-26e97669fe2e"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Loaded data shape: (142246, 1024)\n",
            "(142246, 1024)\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "   Column_1  Column_2  Column_3  Column_4  Column_5  Column_6  Column_7  \\\n",
              "0  0.049488 -0.032935  0.032473 -0.033723 -0.059505  0.075936 -0.045860   \n",
              "1 -0.044616  0.064925 -0.080263 -0.075338 -0.004731  0.025408 -0.024685   \n",
              "2 -0.020128 -0.049779  0.007894 -0.000829 -0.047737  0.086453 -0.038107   \n",
              "3 -0.007515  0.060628  0.004045  0.027056 -0.021542  0.010380 -0.025064   \n",
              "4  0.013468  0.041516  0.018435 -0.035595  0.008770  0.018699 -0.015452   \n",
              "\n",
              "   Column_8  Column_9  Column_10  ...  Column_1015  Column_1016  Column_1017  \\\n",
              "0 -0.104476 -0.072112   0.038806  ...    -0.046516    -0.028773    -0.021878   \n",
              "1 -0.016568 -0.038980  -0.033870  ...    -0.040173     0.003925     0.006130   \n",
              "2 -0.036379  0.029611   0.045029  ...     0.029388     0.008456     0.000697   \n",
              "3 -0.055834  0.068238   0.027764  ...     0.020792     0.023307     0.009009   \n",
              "4 -0.038092 -0.038326  -0.012299  ...    -0.044742    -0.025432    -0.060886   \n",
              "\n",
              "   Column_1018  Column_1019  Column_1020  Column_1021  Column_1022  \\\n",
              "0    -0.097883     0.056475     0.080950    -0.020938    -0.043532   \n",
              "1     0.007362    -0.066848     0.106882    -0.030134     0.026724   \n",
              "2     0.013502    -0.005968    -0.011571     0.005704    -0.036103   \n",
              "3     0.018211     0.020820    -0.031738     0.013279    -0.018357   \n",
              "4    -0.026915     0.026342     0.017237     0.014731    -0.016861   \n",
              "\n",
              "   Column_1023  Column_1024  \n",
              "0     0.096463     0.073070  \n",
              "1     0.027879    -0.048430  \n",
              "2     0.007693     0.106234  \n",
              "3     0.008087     0.010917  \n",
              "4    -0.016272     0.037054  \n",
              "\n",
              "[5 rows x 1024 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-d906588a-8ac9-4409-ac7e-68d7a293b032\" class=\"colab-df-container\">\n",
              "    <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>Column_1</th>\n",
              "      <th>Column_2</th>\n",
              "      <th>Column_3</th>\n",
              "      <th>Column_4</th>\n",
              "      <th>Column_5</th>\n",
              "      <th>Column_6</th>\n",
              "      <th>Column_7</th>\n",
              "      <th>Column_8</th>\n",
              "      <th>Column_9</th>\n",
              "      <th>Column_10</th>\n",
              "      <th>...</th>\n",
              "      <th>Column_1015</th>\n",
              "      <th>Column_1016</th>\n",
              "      <th>Column_1017</th>\n",
              "      <th>Column_1018</th>\n",
              "      <th>Column_1019</th>\n",
              "      <th>Column_1020</th>\n",
              "      <th>Column_1021</th>\n",
              "      <th>Column_1022</th>\n",
              "      <th>Column_1023</th>\n",
              "      <th>Column_1024</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0.049488</td>\n",
              "      <td>-0.032935</td>\n",
              "      <td>0.032473</td>\n",
              "      <td>-0.033723</td>\n",
              "      <td>-0.059505</td>\n",
              "      <td>0.075936</td>\n",
              "      <td>-0.045860</td>\n",
              "      <td>-0.104476</td>\n",
              "      <td>-0.072112</td>\n",
              "      <td>0.038806</td>\n",
              "      <td>...</td>\n",
              "      <td>-0.046516</td>\n",
              "      <td>-0.028773</td>\n",
              "      <td>-0.021878</td>\n",
              "      <td>-0.097883</td>\n",
              "      <td>0.056475</td>\n",
              "      <td>0.080950</td>\n",
              "      <td>-0.020938</td>\n",
              "      <td>-0.043532</td>\n",
              "      <td>0.096463</td>\n",
              "      <td>0.073070</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>-0.044616</td>\n",
              "      <td>0.064925</td>\n",
              "      <td>-0.080263</td>\n",
              "      <td>-0.075338</td>\n",
              "      <td>-0.004731</td>\n",
              "      <td>0.025408</td>\n",
              "      <td>-0.024685</td>\n",
              "      <td>-0.016568</td>\n",
              "      <td>-0.038980</td>\n",
              "      <td>-0.033870</td>\n",
              "      <td>...</td>\n",
              "      <td>-0.040173</td>\n",
              "      <td>0.003925</td>\n",
              "      <td>0.006130</td>\n",
              "      <td>0.007362</td>\n",
              "      <td>-0.066848</td>\n",
              "      <td>0.106882</td>\n",
              "      <td>-0.030134</td>\n",
              "      <td>0.026724</td>\n",
              "      <td>0.027879</td>\n",
              "      <td>-0.048430</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>-0.020128</td>\n",
              "      <td>-0.049779</td>\n",
              "      <td>0.007894</td>\n",
              "      <td>-0.000829</td>\n",
              "      <td>-0.047737</td>\n",
              "      <td>0.086453</td>\n",
              "      <td>-0.038107</td>\n",
              "      <td>-0.036379</td>\n",
              "      <td>0.029611</td>\n",
              "      <td>0.045029</td>\n",
              "      <td>...</td>\n",
              "      <td>0.029388</td>\n",
              "      <td>0.008456</td>\n",
              "      <td>0.000697</td>\n",
              "      <td>0.013502</td>\n",
              "      <td>-0.005968</td>\n",
              "      <td>-0.011571</td>\n",
              "      <td>0.005704</td>\n",
              "      <td>-0.036103</td>\n",
              "      <td>0.007693</td>\n",
              "      <td>0.106234</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>-0.007515</td>\n",
              "      <td>0.060628</td>\n",
              "      <td>0.004045</td>\n",
              "      <td>0.027056</td>\n",
              "      <td>-0.021542</td>\n",
              "      <td>0.010380</td>\n",
              "      <td>-0.025064</td>\n",
              "      <td>-0.055834</td>\n",
              "      <td>0.068238</td>\n",
              "      <td>0.027764</td>\n",
              "      <td>...</td>\n",
              "      <td>0.020792</td>\n",
              "      <td>0.023307</td>\n",
              "      <td>0.009009</td>\n",
              "      <td>0.018211</td>\n",
              "      <td>0.020820</td>\n",
              "      <td>-0.031738</td>\n",
              "      <td>0.013279</td>\n",
              "      <td>-0.018357</td>\n",
              "      <td>0.008087</td>\n",
              "      <td>0.010917</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>0.013468</td>\n",
              "      <td>0.041516</td>\n",
              "      <td>0.018435</td>\n",
              "      <td>-0.035595</td>\n",
              "      <td>0.008770</td>\n",
              "      <td>0.018699</td>\n",
              "      <td>-0.015452</td>\n",
              "      <td>-0.038092</td>\n",
              "      <td>-0.038326</td>\n",
              "      <td>-0.012299</td>\n",
              "      <td>...</td>\n",
              "      <td>-0.044742</td>\n",
              "      <td>-0.025432</td>\n",
              "      <td>-0.060886</td>\n",
              "      <td>-0.026915</td>\n",
              "      <td>0.026342</td>\n",
              "      <td>0.017237</td>\n",
              "      <td>0.014731</td>\n",
              "      <td>-0.016861</td>\n",
              "      <td>-0.016272</td>\n",
              "      <td>0.037054</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>5 rows × 1024 columns</p>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-d906588a-8ac9-4409-ac7e-68d7a293b032')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-d906588a-8ac9-4409-ac7e-68d7a293b032 button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-d906588a-8ac9-4409-ac7e-68d7a293b032');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-ff13bdf1-2b6a-4950-9377-8d1a5375da55\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-ff13bdf1-2b6a-4950-9377-8d1a5375da55')\"\n",
              "            title=\"Suggest charts\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-ff13bdf1-2b6a-4950-9377-8d1a5375da55 button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "\n",
              "    </div>\n",
              "  </div>\n"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "dataframe",
              "variable_name": "train_df"
            }
          },
          "metadata": {},
          "execution_count": 5
        }
      ],
      "source": [
        "# Print the shape of the loaded data to confirm\n",
        "print(\"Loaded data shape:\", train_embeds.shape)\n",
        "\n",
        "# converting embeddings numpy array(train_embeds) into pandas dataframe.\n",
        "column_num = train_embeds.shape[1]\n",
        "train_df = pd.DataFrame(train_embeds, columns = [\"Column_\" + str(i) for i in range(1, column_num+1)])\n",
        "print(train_df.shape)\n",
        "\n",
        "train_df.head()"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Loaded Label dataframe and split the Dataset"
      ],
      "metadata": {
        "id": "XiqXarxpFZYi"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Define the path to the pickle file\n",
        "pickle_path = \"/content/drive/My Drive/DATASET/labels_df.pkl\"\n",
        "\n",
        "# Directly load labels_df from the pickle file\n",
        "labels_df = pd.read_pickle(pickle_path)\n",
        "print(\"Loaded labels_df from pickle.\")\n",
        "\n",
        "# Proceed with splitting the datasets\n",
        "# First, split the data into a training set and a temporary testing set\n",
        "X_train, X_temp, y_train, y_temp = train_test_split(\n",
        "    train_df, labels_df, test_size=0.2, random_state=42\n",
        ")\n",
        "\n",
        "# Next, split the temporary set into validation and test sets\n",
        "X_val, X_test, y_val, y_test = train_test_split(\n",
        "    X_temp, y_temp, test_size=0.5, random_state=42\n",
        ")\n",
        "\n",
        "# Display the dimensions of the sets to verify proper splitting\n",
        "print(f\"Training set dimensions: {X_train.shape}, {y_train.shape}\")\n",
        "print(f\"Validation set dimensions: {X_val.shape}, {y_val.shape}\")\n",
        "print(f\"Test set dimensions: {X_test.shape}, {y_test.shape}\")\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "stsunqsRWD8X",
        "outputId": "b4f28494-61db-4cea-f6ba-e331e3e8d161"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Loaded labels_df from pickle.\n",
            "Training set dimensions: (113796, 1024), (113796, 1500)\n",
            "Validation set dimensions: (14225, 1024), (14225, 1500)\n",
            "Test set dimensions: (14225, 1024), (14225, 1500)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# **Modeling**"
      ],
      "metadata": {
        "id": "ucWGqDKN1MBD"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "#--------------------> 1. RNN, LSTM, GRU"
      ],
      "metadata": {
        "id": "ciGNI8681lT8"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Data configuration\n",
        "INPUT_SHAPE = [X_train.shape[1]]\n",
        "BATCH_SIZE = 512\n",
        "num_of_labels = y_train.shape[1]\n"
      ],
      "metadata": {
        "id": "Iec3TwSQCYYF"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "#### Model Definition"
      ],
      "metadata": {
        "id": "BUwsnYidFxpu"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Model Definition\n",
        "\n",
        "# Define the Simple RNN model with Dropout to prevent overfitting\n",
        "model_rnn_simple = Sequential([\n",
        "    layers.Reshape((1, X_train.shape[1]), input_shape=(X_train.shape[1],)),  # Reshape input for RNN compatibility\n",
        "    layers.SimpleRNN(64, return_sequences=True),  # First RNN layer with return sequences\n",
        "    layers.Dropout(0.2),  # Dropout for regularization\n",
        "    layers.SimpleRNN(64),  # Second RNN layer\n",
        "    layers.Dropout(0.2),  # Additional Dropout for regularization\n",
        "    layers.Dense(256, activation='relu'),  # Dense layer for feature interpretation\n",
        "    layers.Dropout(0.3),  # Dropout after dense layer\n",
        "    layers.Dense(num_of_labels, activation='sigmoid')  # Output layer\n",
        "])\n",
        "model_rnn_simple.compile(optimizer='adam',\n",
        "                         loss='binary_crossentropy',\n",
        "                         metrics=['binary_accuracy', tf.keras.metrics.AUC()])\n",
        "\n",
        "# Define the LSTM model with Dropout to improve generalization\n",
        "model_lstm = Sequential([\n",
        "    layers.Reshape((1, X_train.shape[1]), input_shape=(X_train.shape[1],)),  # Reshape input for LSTM compatibility\n",
        "    layers.LSTM(64, return_sequences=True),  # First LSTM layer with return sequences\n",
        "    layers.Dropout(0.2),  # Dropout for regularization\n",
        "    layers.LSTM(64),  # Second LSTM layer\n",
        "    layers.Dropout(0.2),  # Additional Dropout for regularization\n",
        "    layers.Dense(256, activation='relu'),  # Dense layer for feature interpretation\n",
        "    layers.Dropout(0.3),  # Dropout after dense layer\n",
        "    layers.Dense(num_of_labels, activation='sigmoid')  # Output layer\n",
        "])\n",
        "model_lstm.compile(optimizer='adam',\n",
        "                   loss='binary_crossentropy',\n",
        "                   metrics=['binary_accuracy', tf.keras.metrics.AUC()])\n",
        "\n",
        "# Define the GRU model with Dropout to enhance model robustness\n",
        "model_gru = Sequential([\n",
        "    layers.Reshape((1, X_train.shape[1]), input_shape=(X_train.shape[1],)),  # Reshape input for GRU compatibility\n",
        "    layers.GRU(64, return_sequences=True),  # First GRU layer with return sequences\n",
        "    layers.Dropout(0.2),  # Dropout for regularization\n",
        "    layers.GRU(64),  # Second GRU layer\n",
        "    layers.Dropout(0.2),  # Additional Dropout for regularization\n",
        "    layers.Dense(256, activation='relu'),  # Dense layer for feature interpretation\n",
        "    layers.Dropout(0.3),  # Dropout after dense layer\n",
        "    layers.Dense(num_of_labels, activation='sigmoid')  # Output layer\n",
        "])\n",
        "model_gru.compile(optimizer='adam',\n",
        "                  loss='binary_crossentropy',\n",
        "                  metrics=['binary_accuracy', tf.keras.metrics.AUC()])\n"
      ],
      "metadata": {
        "id": "3pDuTvOXl99_"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "#### Model Training"
      ],
      "metadata": {
        "id": "hb-Is39SF46K"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Setup ModelCheckpoint callbacks\n",
        "checkpoint_rnn = ModelCheckpoint('/content/drive/My Drive/DATASET/model_rnn_simple_best.h5', monitor='val_loss', save_best_only=True, verbose=1)\n",
        "checkpoint_lstm = ModelCheckpoint('/content/drive/My Drive/DATASET/model_lstm_best.h5', monitor='val_loss', save_best_only=True, verbose=1)\n",
        "checkpoint_gru = ModelCheckpoint('/content/drive/My Drive/DATASET/model_gru_best.h5', monitor='val_loss', save_best_only=True, verbose=1)\n",
        "# Train the Simple RNN model with checkpointing\n",
        "history_rnn_simple = model_rnn_simple.fit(\n",
        "    X_train, y_train,\n",
        "    validation_data=(X_val, y_val),\n",
        "    batch_size=BATCH_SIZE,\n",
        "    epochs=10,\n",
        "    callbacks=[checkpoint_rnn]\n",
        ")\n",
        "\n",
        "# Train the LSTM model\n",
        "history_lstm = model_lstm.fit(\n",
        "    X_train, y_train,\n",
        "    validation_data=(X_val, y_val),\n",
        "    batch_size=BATCH_SIZE,\n",
        "    epochs=10,\n",
        "    callbacks=[checkpoint_lstm]\n",
        ")\n",
        "\n",
        "# Train the GRU model\n",
        "history_gru = model_gru.fit(\n",
        "    X_train, y_train,\n",
        "    validation_data=(X_val, y_val),\n",
        "    batch_size=BATCH_SIZE,\n",
        "    epochs=10,\n",
        "    callbacks=[checkpoint_gru]\n",
        ")\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "SZaI5nK9nfPH",
        "outputId": "ffc9b919-16bd-4f6b-b5e6-77cbbc2a6eaa"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/10\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.4923 - binary_accuracy: 0.8899 - auc: 0.5504\n",
            "Epoch 1: val_loss improved from inf to 0.09155, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best.h5\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3103: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\r23/23 [==============================] - 10s 198ms/step - loss: 0.4884 - binary_accuracy: 0.8908 - auc: 0.5507 - val_loss: 0.0915 - val_binary_accuracy: 0.9800 - val_auc: 0.6477\n",
            "Epoch 2/10\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0978 - binary_accuracy: 0.9795 - auc: 0.7287\n",
            "Epoch 2: val_loss improved from 0.09155 to 0.08759, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best.h5\n",
            "23/23 [==============================] - 2s 77ms/step - loss: 0.0977 - binary_accuracy: 0.9795 - auc: 0.7293 - val_loss: 0.0876 - val_binary_accuracy: 0.9800 - val_auc: 0.7924\n",
            "Epoch 3/10\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0832 - binary_accuracy: 0.9798 - auc: 0.8084\n",
            "Epoch 3: val_loss improved from 0.08759 to 0.07871, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best.h5\n",
            "23/23 [==============================] - 2s 68ms/step - loss: 0.0832 - binary_accuracy: 0.9798 - auc: 0.8084 - val_loss: 0.0787 - val_binary_accuracy: 0.9800 - val_auc: 0.8286\n",
            "Epoch 4/10\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0804 - binary_accuracy: 0.9798 - auc: 0.8142\n",
            "Epoch 4: val_loss improved from 0.07871 to 0.07820, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best.h5\n",
            "23/23 [==============================] - 1s 65ms/step - loss: 0.0804 - binary_accuracy: 0.9798 - auc: 0.8142 - val_loss: 0.0782 - val_binary_accuracy: 0.9800 - val_auc: 0.8311\n",
            "Epoch 5/10\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0799 - binary_accuracy: 0.9799 - auc: 0.8170\n",
            "Epoch 5: val_loss improved from 0.07820 to 0.07787, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best.h5\n",
            "23/23 [==============================] - 1s 64ms/step - loss: 0.0799 - binary_accuracy: 0.9799 - auc: 0.8170 - val_loss: 0.0779 - val_binary_accuracy: 0.9799 - val_auc: 0.8337\n",
            "Epoch 6/10\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0796 - binary_accuracy: 0.9799 - auc: 0.8204\n",
            "Epoch 6: val_loss improved from 0.07787 to 0.07766, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best.h5\n",
            "23/23 [==============================] - 1s 65ms/step - loss: 0.0796 - binary_accuracy: 0.9799 - auc: 0.8204 - val_loss: 0.0777 - val_binary_accuracy: 0.9799 - val_auc: 0.8357\n",
            "Epoch 7/10\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0793 - binary_accuracy: 0.9799 - auc: 0.8231\n",
            "Epoch 7: val_loss improved from 0.07766 to 0.07745, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best.h5\n",
            "23/23 [==============================] - 2s 68ms/step - loss: 0.0793 - binary_accuracy: 0.9799 - auc: 0.8231 - val_loss: 0.0774 - val_binary_accuracy: 0.9799 - val_auc: 0.8374\n",
            "Epoch 8/10\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0791 - binary_accuracy: 0.9799 - auc: 0.8247\n",
            "Epoch 8: val_loss improved from 0.07745 to 0.07728, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best.h5\n",
            "23/23 [==============================] - 2s 71ms/step - loss: 0.0791 - binary_accuracy: 0.9799 - auc: 0.8247 - val_loss: 0.0773 - val_binary_accuracy: 0.9799 - val_auc: 0.8386\n",
            "Epoch 9/10\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0788 - binary_accuracy: 0.9799 - auc: 0.8268\n",
            "Epoch 9: val_loss improved from 0.07728 to 0.07707, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best.h5\n",
            "23/23 [==============================] - 2s 68ms/step - loss: 0.0788 - binary_accuracy: 0.9799 - auc: 0.8268 - val_loss: 0.0771 - val_binary_accuracy: 0.9799 - val_auc: 0.8400\n",
            "Epoch 10/10\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0785 - binary_accuracy: 0.9799 - auc: 0.8288\n",
            "Epoch 10: val_loss improved from 0.07707 to 0.07679, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best.h5\n",
            "23/23 [==============================] - 1s 65ms/step - loss: 0.0786 - binary_accuracy: 0.9799 - auc: 0.8288 - val_loss: 0.0768 - val_binary_accuracy: 0.9800 - val_auc: 0.8419\n",
            "Epoch 1/10\n",
            "23/23 [==============================] - ETA: 0s - loss: 0.6787 - binary_accuracy: 0.9498 - auc_1: 0.5477\n",
            "Epoch 1: val_loss improved from inf to 0.62521, saving model to /content/drive/My Drive/DATASET/model_lstm_best.h5\n",
            "23/23 [==============================] - 6s 130ms/step - loss: 0.6787 - binary_accuracy: 0.9498 - auc_1: 0.5477 - val_loss: 0.6252 - val_binary_accuracy: 0.9800 - val_auc_1: 0.6082\n",
            "Epoch 2/10\n",
            " 1/23 [>.............................] - ETA: 1s - loss: 0.6252 - binary_accuracy: 0.9801 - auc_1: 0.5999"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3103: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "22/23 [===========================>..] - ETA: 0s - loss: 0.3510 - binary_accuracy: 0.9799 - auc_1: 0.5683\n",
            "Epoch 2: val_loss improved from 0.62521 to 0.09242, saving model to /content/drive/My Drive/DATASET/model_lstm_best.h5\n",
            "23/23 [==============================] - 1s 57ms/step - loss: 0.3483 - binary_accuracy: 0.9799 - auc_1: 0.5689 - val_loss: 0.0924 - val_binary_accuracy: 0.9799 - val_auc_1: 0.6811\n",
            "Epoch 3/10\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0953 - binary_accuracy: 0.9797 - auc_1: 0.7391\n",
            "Epoch 3: val_loss improved from 0.09242 to 0.08272, saving model to /content/drive/My Drive/DATASET/model_lstm_best.h5\n",
            "23/23 [==============================] - 2s 77ms/step - loss: 0.0952 - binary_accuracy: 0.9797 - auc_1: 0.7397 - val_loss: 0.0827 - val_binary_accuracy: 0.9800 - val_auc_1: 0.8078\n",
            "Epoch 4/10\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0811 - binary_accuracy: 0.9799 - auc_1: 0.8128\n",
            "Epoch 4: val_loss improved from 0.08272 to 0.07874, saving model to /content/drive/My Drive/DATASET/model_lstm_best.h5\n",
            "23/23 [==============================] - 2s 70ms/step - loss: 0.0812 - binary_accuracy: 0.9799 - auc_1: 0.8128 - val_loss: 0.0787 - val_binary_accuracy: 0.9799 - val_auc_1: 0.8267\n",
            "Epoch 5/10\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0796 - binary_accuracy: 0.9799 - auc_1: 0.8188\n",
            "Epoch 5: val_loss improved from 0.07874 to 0.07813, saving model to /content/drive/My Drive/DATASET/model_lstm_best.h5\n",
            "23/23 [==============================] - 2s 69ms/step - loss: 0.0796 - binary_accuracy: 0.9799 - auc_1: 0.8188 - val_loss: 0.0781 - val_binary_accuracy: 0.9799 - val_auc_1: 0.8307\n",
            "Epoch 6/10\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0791 - binary_accuracy: 0.9799 - auc_1: 0.8230\n",
            "Epoch 6: val_loss improved from 0.07813 to 0.07781, saving model to /content/drive/My Drive/DATASET/model_lstm_best.h5\n",
            "23/23 [==============================] - 2s 70ms/step - loss: 0.0792 - binary_accuracy: 0.9799 - auc_1: 0.8230 - val_loss: 0.0778 - val_binary_accuracy: 0.9800 - val_auc_1: 0.8339\n",
            "Epoch 7/10\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0788 - binary_accuracy: 0.9799 - auc_1: 0.8264\n",
            "Epoch 7: val_loss improved from 0.07781 to 0.07760, saving model to /content/drive/My Drive/DATASET/model_lstm_best.h5\n",
            "23/23 [==============================] - 2s 70ms/step - loss: 0.0788 - binary_accuracy: 0.9799 - auc_1: 0.8264 - val_loss: 0.0776 - val_binary_accuracy: 0.9799 - val_auc_1: 0.8359\n",
            "Epoch 8/10\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0786 - binary_accuracy: 0.9799 - auc_1: 0.8282\n",
            "Epoch 8: val_loss improved from 0.07760 to 0.07744, saving model to /content/drive/My Drive/DATASET/model_lstm_best.h5\n",
            "23/23 [==============================] - 2s 71ms/step - loss: 0.0786 - binary_accuracy: 0.9799 - auc_1: 0.8282 - val_loss: 0.0774 - val_binary_accuracy: 0.9800 - val_auc_1: 0.8374\n",
            "Epoch 9/10\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0785 - binary_accuracy: 0.9799 - auc_1: 0.8297\n",
            "Epoch 9: val_loss improved from 0.07744 to 0.07731, saving model to /content/drive/My Drive/DATASET/model_lstm_best.h5\n",
            "23/23 [==============================] - 2s 72ms/step - loss: 0.0785 - binary_accuracy: 0.9799 - auc_1: 0.8297 - val_loss: 0.0773 - val_binary_accuracy: 0.9799 - val_auc_1: 0.8386\n",
            "Epoch 10/10\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0783 - binary_accuracy: 0.9799 - auc_1: 0.8309\n",
            "Epoch 10: val_loss improved from 0.07731 to 0.07721, saving model to /content/drive/My Drive/DATASET/model_lstm_best.h5\n",
            "23/23 [==============================] - 2s 70ms/step - loss: 0.0783 - binary_accuracy: 0.9799 - auc_1: 0.8309 - val_loss: 0.0772 - val_binary_accuracy: 0.9800 - val_auc_1: 0.8397\n",
            "Epoch 1/10\n",
            "23/23 [==============================] - ETA: 0s - loss: 0.6334 - binary_accuracy: 0.9067 - auc_2: 0.5475\n",
            "Epoch 1: val_loss improved from inf to 0.35879, saving model to /content/drive/My Drive/DATASET/model_gru_best.h5\n",
            "23/23 [==============================] - 6s 118ms/step - loss: 0.6334 - binary_accuracy: 0.9067 - auc_2: 0.5475 - val_loss: 0.3588 - val_binary_accuracy: 0.9799 - val_auc_2: 0.5834\n",
            "Epoch 2/10\n",
            " 3/23 [==>...........................] - ETA: 0s - loss: 0.3110 - binary_accuracy: 0.9803 - auc_2: 0.5796"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3103: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "22/23 [===========================>..] - ETA: 0s - loss: 0.1408 - binary_accuracy: 0.9797 - auc_2: 0.6409\n",
            "Epoch 2: val_loss improved from 0.35879 to 0.09859, saving model to /content/drive/My Drive/DATASET/model_gru_best.h5\n",
            "23/23 [==============================] - 1s 65ms/step - loss: 0.1404 - binary_accuracy: 0.9797 - auc_2: 0.6420 - val_loss: 0.0986 - val_binary_accuracy: 0.9799 - val_auc_2: 0.7485\n",
            "Epoch 3/10\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0874 - binary_accuracy: 0.9797 - auc_2: 0.7898\n",
            "Epoch 3: val_loss improved from 0.09859 to 0.07935, saving model to /content/drive/My Drive/DATASET/model_gru_best.h5\n",
            "23/23 [==============================] - 2s 70ms/step - loss: 0.0873 - binary_accuracy: 0.9798 - auc_2: 0.7900 - val_loss: 0.0793 - val_binary_accuracy: 0.9799 - val_auc_2: 0.8253\n",
            "Epoch 4/10\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0804 - binary_accuracy: 0.9799 - auc_2: 0.8167\n",
            "Epoch 4: val_loss improved from 0.07935 to 0.07833, saving model to /content/drive/My Drive/DATASET/model_gru_best.h5\n",
            "23/23 [==============================] - 1s 66ms/step - loss: 0.0804 - binary_accuracy: 0.9799 - auc_2: 0.8167 - val_loss: 0.0783 - val_binary_accuracy: 0.9799 - val_auc_2: 0.8299\n",
            "Epoch 5/10\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0797 - binary_accuracy: 0.9799 - auc_2: 0.8188\n",
            "Epoch 5: val_loss improved from 0.07833 to 0.07791, saving model to /content/drive/My Drive/DATASET/model_gru_best.h5\n",
            "23/23 [==============================] - 2s 66ms/step - loss: 0.0796 - binary_accuracy: 0.9799 - auc_2: 0.8189 - val_loss: 0.0779 - val_binary_accuracy: 0.9799 - val_auc_2: 0.8330\n",
            "Epoch 6/10\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0793 - binary_accuracy: 0.9799 - auc_2: 0.8225\n",
            "Epoch 6: val_loss improved from 0.07791 to 0.07768, saving model to /content/drive/My Drive/DATASET/model_gru_best.h5\n",
            "23/23 [==============================] - 1s 65ms/step - loss: 0.0793 - binary_accuracy: 0.9799 - auc_2: 0.8225 - val_loss: 0.0777 - val_binary_accuracy: 0.9799 - val_auc_2: 0.8352\n",
            "Epoch 7/10\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0790 - binary_accuracy: 0.9799 - auc_2: 0.8256\n",
            "Epoch 7: val_loss improved from 0.07768 to 0.07751, saving model to /content/drive/My Drive/DATASET/model_gru_best.h5\n",
            "23/23 [==============================] - 2s 66ms/step - loss: 0.0790 - binary_accuracy: 0.9799 - auc_2: 0.8257 - val_loss: 0.0775 - val_binary_accuracy: 0.9800 - val_auc_2: 0.8367\n",
            "Epoch 8/10\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0788 - binary_accuracy: 0.9799 - auc_2: 0.8268\n",
            "Epoch 8: val_loss improved from 0.07751 to 0.07737, saving model to /content/drive/My Drive/DATASET/model_gru_best.h5\n",
            "23/23 [==============================] - 2s 67ms/step - loss: 0.0788 - binary_accuracy: 0.9799 - auc_2: 0.8268 - val_loss: 0.0774 - val_binary_accuracy: 0.9800 - val_auc_2: 0.8380\n",
            "Epoch 9/10\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0786 - binary_accuracy: 0.9799 - auc_2: 0.8285\n",
            "Epoch 9: val_loss improved from 0.07737 to 0.07726, saving model to /content/drive/My Drive/DATASET/model_gru_best.h5\n",
            "23/23 [==============================] - 2s 68ms/step - loss: 0.0786 - binary_accuracy: 0.9799 - auc_2: 0.8285 - val_loss: 0.0773 - val_binary_accuracy: 0.9800 - val_auc_2: 0.8390\n",
            "Epoch 10/10\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0785 - binary_accuracy: 0.9799 - auc_2: 0.8297\n",
            "Epoch 10: val_loss improved from 0.07726 to 0.07719, saving model to /content/drive/My Drive/DATASET/model_gru_best.h5\n",
            "23/23 [==============================] - 2s 67ms/step - loss: 0.0785 - binary_accuracy: 0.9799 - auc_2: 0.8297 - val_loss: 0.0772 - val_binary_accuracy: 0.9799 - val_auc_2: 0.8397\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Paths of models\n",
        "path_rnn = '/content/drive/My Drive/DATASET/model_rnn_simple_best.h5'\n",
        "path_lstm = '/content/drive/My Drive/DATASET/model_lstm_best.h5'\n",
        "path_gru = '/content/drive/My Drive/DATASET/model_gru_best.h5'\n",
        "\n",
        "# load models\n",
        "model_rnn_simple = load_model(path_rnn)\n",
        "model_lstm = load_model(path_lstm)\n",
        "model_gru = load_model(path_gru)\n"
      ],
      "metadata": {
        "id": "tS9mpcputya4"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "#### Models Summary"
      ],
      "metadata": {
        "id": "pQ_10FymF_Ze"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Display the summary of the RNN model\n",
        "print(\"RNN Model Summary:\")\n",
        "model_rnn_simple.summary()\n",
        "\n",
        "# Display the summary of the LSTM model\n",
        "print(\"\\nLSTM Model Summary:\")\n",
        "model_lstm.summary()\n",
        "\n",
        "# Display the summary of the GRU model\n",
        "print(\"\\nGRU Model Summary:\")\n",
        "model_gru.summary()\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "mK60Zr7FNPQm",
        "outputId": "94aa5378-9cb0-49d8-9196-3988d0267c41"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "RNN Model Summary:\n",
            "Model: \"sequential\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " reshape (Reshape)           (None, 1, 1024)           0         \n",
            "                                                                 \n",
            " simple_rnn (SimpleRNN)      (None, 1, 64)             69696     \n",
            "                                                                 \n",
            " dropout (Dropout)           (None, 1, 64)             0         \n",
            "                                                                 \n",
            " simple_rnn_1 (SimpleRNN)    (None, 64)                8256      \n",
            "                                                                 \n",
            " dropout_1 (Dropout)         (None, 64)                0         \n",
            "                                                                 \n",
            " dense (Dense)               (None, 256)               16640     \n",
            "                                                                 \n",
            " dropout_2 (Dropout)         (None, 256)               0         \n",
            "                                                                 \n",
            " dense_1 (Dense)             (None, 1500)              385500    \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 480092 (1.83 MB)\n",
            "Trainable params: 480092 (1.83 MB)\n",
            "Non-trainable params: 0 (0.00 Byte)\n",
            "_________________________________________________________________\n",
            "\n",
            "LSTM Model Summary:\n",
            "Model: \"sequential_1\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " reshape_1 (Reshape)         (None, 1, 1024)           0         \n",
            "                                                                 \n",
            " lstm (LSTM)                 (None, 1, 64)             278784    \n",
            "                                                                 \n",
            " dropout_3 (Dropout)         (None, 1, 64)             0         \n",
            "                                                                 \n",
            " lstm_1 (LSTM)               (None, 64)                33024     \n",
            "                                                                 \n",
            " dropout_4 (Dropout)         (None, 64)                0         \n",
            "                                                                 \n",
            " dense_2 (Dense)             (None, 256)               16640     \n",
            "                                                                 \n",
            " dropout_5 (Dropout)         (None, 256)               0         \n",
            "                                                                 \n",
            " dense_3 (Dense)             (None, 1500)              385500    \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 713948 (2.72 MB)\n",
            "Trainable params: 713948 (2.72 MB)\n",
            "Non-trainable params: 0 (0.00 Byte)\n",
            "_________________________________________________________________\n",
            "\n",
            "GRU Model Summary:\n",
            "Model: \"sequential_2\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " reshape_2 (Reshape)         (None, 1, 1024)           0         \n",
            "                                                                 \n",
            " gru (GRU)                   (None, 1, 64)             209280    \n",
            "                                                                 \n",
            " dropout_6 (Dropout)         (None, 1, 64)             0         \n",
            "                                                                 \n",
            " gru_1 (GRU)                 (None, 64)                24960     \n",
            "                                                                 \n",
            " dropout_7 (Dropout)         (None, 64)                0         \n",
            "                                                                 \n",
            " dense_4 (Dense)             (None, 256)               16640     \n",
            "                                                                 \n",
            " dropout_8 (Dropout)         (None, 256)               0         \n",
            "                                                                 \n",
            " dense_5 (Dense)             (None, 1500)              385500    \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 636380 (2.43 MB)\n",
            "Trainable params: 636380 (2.43 MB)\n",
            "Non-trainable params: 0 (0.00 Byte)\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "#### Models Predicton § Evaluation"
      ],
      "metadata": {
        "id": "6kw9_wK0GH6e"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "def evaluate_model(model, X, y):\n",
        "    \"\"\"Evaluate the model using Fmax and AUPR metrics.\"\"\"\n",
        "    # Get model predictions as probabilities\n",
        "    y_pred_probs = model.predict(X)\n",
        "\n",
        "    # Ensure y and y_pred_probs are numpy arrays\n",
        "    if isinstance(y, pd.DataFrame):\n",
        "        y = y.values\n",
        "    if isinstance(y_pred_probs, pd.DataFrame):\n",
        "        y_pred_probs = y_pred_probs.values\n",
        "\n",
        "    fmax_scores = []\n",
        "    aupr_scores = []  # Using average_precision_score for simplicity\n",
        "\n",
        "    # Calculate Fmax and AUPR for each label\n",
        "    for i in range(y.shape[1]):\n",
        "        precision, recall, thresholds = precision_recall_curve(y[:, i], y_pred_probs[:, i])\n",
        "        if len(precision) > 1 and len(recall) > 1:\n",
        "            f1_scores = 2 * (precision * recall) / (np.maximum(precision + recall, 1e-5))\n",
        "            fmax = np.max(f1_scores)\n",
        "            fmax_scores.append(fmax)\n",
        "            aupr_score = average_precision_score(y[:, i], y_pred_probs[:, i])  # Direct calculation of AUPR\n",
        "            aupr_scores.append(aupr_score)\n",
        "        else:\n",
        "            fmax_scores.append(0)\n",
        "            aupr_scores.append(0)\n",
        "\n",
        "    average_fmax = np.mean(fmax_scores)\n",
        "    average_aupr = np.mean(aupr_scores)  # AUPR directly calculated\n",
        "\n",
        "    return average_fmax, average_aupr\n",
        "\n",
        "# Evaluate each model on the test set\n",
        "fmax_rnn, aupr_rnn = evaluate_model(model_rnn_simple, X_test, y_test)\n",
        "fmax_lstm, aupr_lstm = evaluate_model(model_lstm, X_test, y_test)\n",
        "fmax_gru, aupr_gru = evaluate_model(model_gru, X_test, y_test)\n",
        "\n",
        "# Print the evaluation results\n",
        "print(\"Evaluation Results:\")\n",
        "print(f\"Simple RNN - Fmax: {fmax_rnn:.4f}, AUPR: {aupr_rnn:.4f}\")\n",
        "print(f\"LSTM - Fmax: {fmax_lstm:.4f}, AUPR: {aupr_lstm:.4f}\")\n",
        "print(f\"GRU - Fmax: {fmax_gru:.4f}, AUPR: {aupr_gru:.4f}\")\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "xfxdH1i0oL8i",
        "outputId": "a69b1c3d-00cf-41fb-d529-d197284b4474"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "445/445 [==============================] - 1s 2ms/step\n",
            "445/445 [==============================] - 2s 2ms/step\n",
            "445/445 [==============================] - 1s 2ms/step\n",
            "Evaluation Results:\n",
            "Simple RNN - Fmax: 0.0641, AUPR: 0.0333\n",
            "LSTM - Fmax: 0.0612, AUPR: 0.0309\n",
            "GRU - Fmax: 0.0620, AUPR: 0.0316\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Comparative Analysis\n",
        "print(f\"Model Performance Comparison:\")\n",
        "print(f\"{'Model':<20}{'Fmax':<10}{'AUPR':<10}\")\n",
        "print(f\"{'-'*40}\")\n",
        "print(f\"{'Simple RNN':<20}{fmax_rnn:<10.4f}{aupr_rnn:<10.4f}\")\n",
        "print(f\"{'LSTM':<20}{fmax_lstm:<10.4f}{aupr_lstm:<10.4f}\")\n",
        "print(f\"{'GRU':<20}{fmax_gru:<10.4f}{aupr_gru:<10.4f}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "1_BdBocnoqiJ",
        "outputId": "be2388bf-3fb7-40a6-9e37-d86d59cba713"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model Performance Comparison:\n",
            "Model               Fmax      AUPR      \n",
            "----------------------------------------\n",
            "Simple RNN          0.0641    0.0333    \n",
            "LSTM                0.0612    0.0309    \n",
            "GRU                 0.0620    0.0316    \n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "####Visualise the results\n"
      ],
      "metadata": {
        "id": "-MHJ6F32GWL-"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Data preparation\n",
        "models = ['Simple RNN', 'LSTM', 'GRU']\n",
        "fmax_scores = [fmax_rnn, fmax_lstm, fmax_gru]\n",
        "aupr_scores = [aupr_rnn, aupr_lstm, aupr_gru]\n",
        "\n",
        "# Plotting\n",
        "fig, ax = plt.subplots(1, 2, figsize=(14, 5))\n",
        "\n",
        "# Fmax Scores\n",
        "ax[0].bar(models, fmax_scores, color='blue')\n",
        "ax[0].set_title('Fmax Scores Comparison')\n",
        "ax[0].set_ylabel('Fmax Score')\n",
        "ax[0].set_ylim([0, 0.1])\n",
        "\n",
        "# AUC Scores\n",
        "ax[1].bar(models, aupr_scores, color='green')\n",
        "ax[1].set_title('AUPR Scores Comparison')\n",
        "ax[1].set_ylabel('AUPR Score')\n",
        "ax[1].set_ylim([0, 0.1])\n",
        "\n",
        "plt.show()\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 468
        },
        "id": "joYFMpL3ox4-",
        "outputId": "33cb799f-e50c-4326-f49c-f44e931da2db"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1400x500 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "#------------------> 2. RNN, LSTM, GRU OPTIMIZATION"
      ],
      "metadata": {
        "id": "yTSL3bu-13Zm"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "#### Models definition § training"
      ],
      "metadata": {
        "id": "4rb7-XkKGuQV"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Define the model architectures\n",
        "def build_rnn_model(input_shape, num_of_labels):\n",
        "    model = Sequential([\n",
        "        layers.Reshape((1, input_shape), input_shape=(input_shape,)),\n",
        "        layers.SimpleRNN(128, return_sequences=True),\n",
        "        layers.Dropout(0.3),\n",
        "        layers.SimpleRNN(128),\n",
        "        layers.Dropout(0.3),\n",
        "        layers.Dense(512, activation='relu'),\n",
        "        layers.Dropout(0.4),\n",
        "        layers.Dense(num_of_labels, activation='sigmoid')\n",
        "    ])\n",
        "    model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['binary_accuracy', tf.keras.metrics.AUC()])\n",
        "    return model\n",
        "\n",
        "def build_lstm_model(input_shape, num_of_labels):\n",
        "    model = Sequential([\n",
        "        layers.Reshape((1, input_shape), input_shape=(input_shape,)),\n",
        "        layers.LSTM(128, return_sequences=True),\n",
        "        layers.Dropout(0.3),\n",
        "        layers.LSTM(128),\n",
        "        layers.Dropout(0.3),\n",
        "        layers.Dense(512, activation='relu'),\n",
        "        layers.Dropout(0.4),\n",
        "        layers.Dense(num_of_labels, activation='sigmoid')\n",
        "    ])\n",
        "    model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['binary_accuracy', tf.keras.metrics.AUC()])\n",
        "    return model\n",
        "\n",
        "def build_gru_model(input_shape, num_of_labels):\n",
        "    model = Sequential([\n",
        "        layers.Reshape((1, input_shape), input_shape=(input_shape,)),\n",
        "        layers.GRU(128, return_sequences=True),\n",
        "        layers.Dropout(0.3),\n",
        "        layers.GRU(128),\n",
        "        layers.Dropout(0.3),\n",
        "        layers.Dense(512, activation='relu'),\n",
        "        layers.Dropout(0.4),\n",
        "        layers.Dense(num_of_labels, activation='sigmoid')\n",
        "    ])\n",
        "    model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['binary_accuracy', tf.keras.metrics.AUC()])\n",
        "    return model\n",
        "\n",
        "# Initialize models\n",
        "model_rnn_simple = build_rnn_model(X_train.shape[1], num_of_labels)\n",
        "model_lstm = build_lstm_model(X_train.shape[1], num_of_labels)\n",
        "model_gru = build_gru_model(X_train.shape[1], num_of_labels)\n",
        "\n",
        "# Setup ModelCheckpoint callbacks\n",
        "checkpoint_rnn = ModelCheckpoint('/content/drive/My Drive/DATASET/model_rnn_simple_best2.h5', monitor='val_loss', save_best_only=True, verbose=1)\n",
        "checkpoint_lstm = ModelCheckpoint('/content/drive/My Drive/DATASET/model_lstm_best2.h5', monitor='val_loss', save_best_only=True, verbose=1)\n",
        "checkpoint_gru = ModelCheckpoint('/content/drive/My Drive/DATASET/model_gru_best2.h5', monitor='val_loss', save_best_only=True, verbose=1)\n",
        "\n",
        "# Model training with checkpointing\n",
        "history_rnn_simple = model_rnn_simple.fit(\n",
        "    X_train, y_train,\n",
        "    validation_data=(X_val, y_val),\n",
        "    batch_size=BATCH_SIZE,\n",
        "    epochs=50,  # Adjusted number of epochs\n",
        "    callbacks=[checkpoint_rnn]\n",
        ")\n",
        "\n",
        "history_lstm = model_lstm.fit(\n",
        "    X_train, y_train,\n",
        "    validation_data=(X_val, y_val),\n",
        "    batch_size=BATCH_SIZE,\n",
        "    epochs=50,  # Adjusted number of epochs\n",
        "    callbacks=[checkpoint_lstm]\n",
        ")\n",
        "\n",
        "history_gru = model_gru.fit(\n",
        "    X_train, y_train,\n",
        "    validation_data=(X_val, y_val),\n",
        "    batch_size=BATCH_SIZE,\n",
        "    epochs=50,  # Adjusted number of epochs\n",
        "    callbacks=[checkpoint_gru]\n",
        ")\n",
        "\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "-poJCPJV2UuT",
        "outputId": "7c51ade2-7313-4c36-e43f-1e14ed059276"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.3483 - binary_accuracy: 0.9136 - auc_3: 0.5770\n",
            "Epoch 1: val_loss improved from inf to 0.12769, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3103: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\r23/23 [==============================] - 6s 138ms/step - loss: 0.3462 - binary_accuracy: 0.9142 - auc_3: 0.5777 - val_loss: 0.1277 - val_binary_accuracy: 0.9794 - val_auc_3: 0.7041\n",
            "Epoch 2/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.1004 - binary_accuracy: 0.9789 - auc_3: 0.7738\n",
            "Epoch 2: val_loss improved from 0.12769 to 0.08026, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 2s 75ms/step - loss: 0.1002 - binary_accuracy: 0.9789 - auc_3: 0.7740 - val_loss: 0.0803 - val_binary_accuracy: 0.9799 - val_auc_3: 0.8230\n",
            "Epoch 3/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0819 - binary_accuracy: 0.9797 - auc_3: 0.8048\n",
            "Epoch 3: val_loss improved from 0.08026 to 0.07826, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 2s 66ms/step - loss: 0.0819 - binary_accuracy: 0.9797 - auc_3: 0.8049 - val_loss: 0.0783 - val_binary_accuracy: 0.9800 - val_auc_3: 0.8284\n",
            "Epoch 4/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0802 - binary_accuracy: 0.9798 - auc_3: 0.8160\n",
            "Epoch 4: val_loss improved from 0.07826 to 0.07754, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 1s 66ms/step - loss: 0.0802 - binary_accuracy: 0.9798 - auc_3: 0.8160 - val_loss: 0.0775 - val_binary_accuracy: 0.9799 - val_auc_3: 0.8360\n",
            "Epoch 5/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0796 - binary_accuracy: 0.9798 - auc_3: 0.8226\n",
            "Epoch 5: val_loss improved from 0.07754 to 0.07721, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 2s 67ms/step - loss: 0.0796 - binary_accuracy: 0.9798 - auc_3: 0.8225 - val_loss: 0.0772 - val_binary_accuracy: 0.9800 - val_auc_3: 0.8391\n",
            "Epoch 6/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0792 - binary_accuracy: 0.9798 - auc_3: 0.8254\n",
            "Epoch 6: val_loss improved from 0.07721 to 0.07689, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 1s 66ms/step - loss: 0.0792 - binary_accuracy: 0.9798 - auc_3: 0.8255 - val_loss: 0.0769 - val_binary_accuracy: 0.9800 - val_auc_3: 0.8416\n",
            "Epoch 7/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0789 - binary_accuracy: 0.9799 - auc_3: 0.8280\n",
            "Epoch 7: val_loss improved from 0.07689 to 0.07653, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 2s 66ms/step - loss: 0.0789 - binary_accuracy: 0.9799 - auc_3: 0.8280 - val_loss: 0.0765 - val_binary_accuracy: 0.9800 - val_auc_3: 0.8440\n",
            "Epoch 8/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0784 - binary_accuracy: 0.9799 - auc_3: 0.8315\n",
            "Epoch 8: val_loss improved from 0.07653 to 0.07602, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 2s 67ms/step - loss: 0.0783 - binary_accuracy: 0.9799 - auc_3: 0.8316 - val_loss: 0.0760 - val_binary_accuracy: 0.9800 - val_auc_3: 0.8470\n",
            "Epoch 9/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0777 - binary_accuracy: 0.9799 - auc_3: 0.8357\n",
            "Epoch 9: val_loss improved from 0.07602 to 0.07520, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 2s 67ms/step - loss: 0.0777 - binary_accuracy: 0.9799 - auc_3: 0.8357 - val_loss: 0.0752 - val_binary_accuracy: 0.9800 - val_auc_3: 0.8527\n",
            "Epoch 10/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0770 - binary_accuracy: 0.9799 - auc_3: 0.8412\n",
            "Epoch 10: val_loss improved from 0.07520 to 0.07445, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 2s 68ms/step - loss: 0.0769 - binary_accuracy: 0.9799 - auc_3: 0.8413 - val_loss: 0.0744 - val_binary_accuracy: 0.9800 - val_auc_3: 0.8594\n",
            "Epoch 11/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0762 - binary_accuracy: 0.9799 - auc_3: 0.8470\n",
            "Epoch 11: val_loss improved from 0.07445 to 0.07385, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 2s 66ms/step - loss: 0.0762 - binary_accuracy: 0.9799 - auc_3: 0.8470 - val_loss: 0.0739 - val_binary_accuracy: 0.9800 - val_auc_3: 0.8644\n",
            "Epoch 12/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0757 - binary_accuracy: 0.9800 - auc_3: 0.8514\n",
            "Epoch 12: val_loss improved from 0.07385 to 0.07345, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 2s 68ms/step - loss: 0.0757 - binary_accuracy: 0.9800 - auc_3: 0.8514 - val_loss: 0.0734 - val_binary_accuracy: 0.9800 - val_auc_3: 0.8669\n",
            "Epoch 13/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0752 - binary_accuracy: 0.9800 - auc_3: 0.8549\n",
            "Epoch 13: val_loss improved from 0.07345 to 0.07296, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 1s 66ms/step - loss: 0.0751 - binary_accuracy: 0.9800 - auc_3: 0.8549 - val_loss: 0.0730 - val_binary_accuracy: 0.9801 - val_auc_3: 0.8700\n",
            "Epoch 14/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0747 - binary_accuracy: 0.9800 - auc_3: 0.8577\n",
            "Epoch 14: val_loss improved from 0.07296 to 0.07247, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 1s 66ms/step - loss: 0.0747 - binary_accuracy: 0.9800 - auc_3: 0.8578 - val_loss: 0.0725 - val_binary_accuracy: 0.9802 - val_auc_3: 0.8731\n",
            "Epoch 15/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0741 - binary_accuracy: 0.9801 - auc_3: 0.8612\n",
            "Epoch 15: val_loss improved from 0.07247 to 0.07184, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 1s 66ms/step - loss: 0.0741 - binary_accuracy: 0.9801 - auc_3: 0.8613 - val_loss: 0.0718 - val_binary_accuracy: 0.9802 - val_auc_3: 0.8770\n",
            "Epoch 16/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0735 - binary_accuracy: 0.9801 - auc_3: 0.8652\n",
            "Epoch 16: val_loss improved from 0.07184 to 0.07107, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 2s 68ms/step - loss: 0.0735 - binary_accuracy: 0.9801 - auc_3: 0.8653 - val_loss: 0.0711 - val_binary_accuracy: 0.9803 - val_auc_3: 0.8816\n",
            "Epoch 17/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0728 - binary_accuracy: 0.9802 - auc_3: 0.8697\n",
            "Epoch 17: val_loss improved from 0.07107 to 0.07043, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 1s 66ms/step - loss: 0.0728 - binary_accuracy: 0.9802 - auc_3: 0.8696 - val_loss: 0.0704 - val_binary_accuracy: 0.9803 - val_auc_3: 0.8856\n",
            "Epoch 18/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0722 - binary_accuracy: 0.9802 - auc_3: 0.8734\n",
            "Epoch 18: val_loss improved from 0.07043 to 0.06994, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 2s 66ms/step - loss: 0.0722 - binary_accuracy: 0.9802 - auc_3: 0.8734 - val_loss: 0.0699 - val_binary_accuracy: 0.9803 - val_auc_3: 0.8885\n",
            "Epoch 19/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0717 - binary_accuracy: 0.9802 - auc_3: 0.8769\n",
            "Epoch 19: val_loss improved from 0.06994 to 0.06952, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 1s 65ms/step - loss: 0.0717 - binary_accuracy: 0.9802 - auc_3: 0.8769 - val_loss: 0.0695 - val_binary_accuracy: 0.9804 - val_auc_3: 0.8911\n",
            "Epoch 20/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0713 - binary_accuracy: 0.9802 - auc_3: 0.8792\n",
            "Epoch 20: val_loss improved from 0.06952 to 0.06917, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 2s 66ms/step - loss: 0.0713 - binary_accuracy: 0.9802 - auc_3: 0.8792 - val_loss: 0.0692 - val_binary_accuracy: 0.9804 - val_auc_3: 0.8929\n",
            "Epoch 21/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0709 - binary_accuracy: 0.9803 - auc_3: 0.8816\n",
            "Epoch 21: val_loss improved from 0.06917 to 0.06880, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 2s 67ms/step - loss: 0.0709 - binary_accuracy: 0.9803 - auc_3: 0.8816 - val_loss: 0.0688 - val_binary_accuracy: 0.9804 - val_auc_3: 0.8952\n",
            "Epoch 22/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0705 - binary_accuracy: 0.9803 - auc_3: 0.8841\n",
            "Epoch 22: val_loss improved from 0.06880 to 0.06830, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 1s 65ms/step - loss: 0.0705 - binary_accuracy: 0.9803 - auc_3: 0.8840 - val_loss: 0.0683 - val_binary_accuracy: 0.9804 - val_auc_3: 0.8976\n",
            "Epoch 23/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0700 - binary_accuracy: 0.9803 - auc_3: 0.8864\n",
            "Epoch 23: val_loss improved from 0.06830 to 0.06779, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 1s 65ms/step - loss: 0.0701 - binary_accuracy: 0.9803 - auc_3: 0.8864 - val_loss: 0.0678 - val_binary_accuracy: 0.9805 - val_auc_3: 0.9003\n",
            "Epoch 24/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0696 - binary_accuracy: 0.9804 - auc_3: 0.8891\n",
            "Epoch 24: val_loss improved from 0.06779 to 0.06729, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 2s 67ms/step - loss: 0.0696 - binary_accuracy: 0.9804 - auc_3: 0.8891 - val_loss: 0.0673 - val_binary_accuracy: 0.9805 - val_auc_3: 0.9029\n",
            "Epoch 25/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0692 - binary_accuracy: 0.9804 - auc_3: 0.8912\n",
            "Epoch 25: val_loss improved from 0.06729 to 0.06689, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 2s 69ms/step - loss: 0.0692 - binary_accuracy: 0.9804 - auc_3: 0.8912 - val_loss: 0.0669 - val_binary_accuracy: 0.9806 - val_auc_3: 0.9049\n",
            "Epoch 26/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0689 - binary_accuracy: 0.9804 - auc_3: 0.8930\n",
            "Epoch 26: val_loss improved from 0.06689 to 0.06659, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 2s 67ms/step - loss: 0.0689 - binary_accuracy: 0.9804 - auc_3: 0.8931 - val_loss: 0.0666 - val_binary_accuracy: 0.9806 - val_auc_3: 0.9061\n",
            "Epoch 27/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0685 - binary_accuracy: 0.9805 - auc_3: 0.8946\n",
            "Epoch 27: val_loss improved from 0.06659 to 0.06635, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 1s 66ms/step - loss: 0.0685 - binary_accuracy: 0.9805 - auc_3: 0.8946 - val_loss: 0.0664 - val_binary_accuracy: 0.9806 - val_auc_3: 0.9080\n",
            "Epoch 28/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0682 - binary_accuracy: 0.9805 - auc_3: 0.8964\n",
            "Epoch 28: val_loss improved from 0.06635 to 0.06611, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 2s 67ms/step - loss: 0.0682 - binary_accuracy: 0.9805 - auc_3: 0.8963 - val_loss: 0.0661 - val_binary_accuracy: 0.9807 - val_auc_3: 0.9085\n",
            "Epoch 29/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0681 - binary_accuracy: 0.9805 - auc_3: 0.8974\n",
            "Epoch 29: val_loss improved from 0.06611 to 0.06590, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 2s 67ms/step - loss: 0.0680 - binary_accuracy: 0.9805 - auc_3: 0.8973 - val_loss: 0.0659 - val_binary_accuracy: 0.9807 - val_auc_3: 0.9091\n",
            "Epoch 30/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0678 - binary_accuracy: 0.9805 - auc_3: 0.8982\n",
            "Epoch 30: val_loss improved from 0.06590 to 0.06573, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 2s 69ms/step - loss: 0.0678 - binary_accuracy: 0.9805 - auc_3: 0.8983 - val_loss: 0.0657 - val_binary_accuracy: 0.9807 - val_auc_3: 0.9103\n",
            "Epoch 31/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0676 - binary_accuracy: 0.9806 - auc_3: 0.8994\n",
            "Epoch 31: val_loss improved from 0.06573 to 0.06554, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 2s 66ms/step - loss: 0.0676 - binary_accuracy: 0.9806 - auc_3: 0.8994 - val_loss: 0.0655 - val_binary_accuracy: 0.9807 - val_auc_3: 0.9114\n",
            "Epoch 32/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0674 - binary_accuracy: 0.9806 - auc_3: 0.9003\n",
            "Epoch 32: val_loss improved from 0.06554 to 0.06535, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 2s 69ms/step - loss: 0.0674 - binary_accuracy: 0.9806 - auc_3: 0.9004 - val_loss: 0.0654 - val_binary_accuracy: 0.9807 - val_auc_3: 0.9122\n",
            "Epoch 33/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0672 - binary_accuracy: 0.9806 - auc_3: 0.9011\n",
            "Epoch 33: val_loss improved from 0.06535 to 0.06516, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 2s 67ms/step - loss: 0.0673 - binary_accuracy: 0.9806 - auc_3: 0.9011 - val_loss: 0.0652 - val_binary_accuracy: 0.9807 - val_auc_3: 0.9128\n",
            "Epoch 34/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0670 - binary_accuracy: 0.9806 - auc_3: 0.9023\n",
            "Epoch 34: val_loss improved from 0.06516 to 0.06501, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 2s 68ms/step - loss: 0.0670 - binary_accuracy: 0.9806 - auc_3: 0.9023 - val_loss: 0.0650 - val_binary_accuracy: 0.9808 - val_auc_3: 0.9134\n",
            "Epoch 35/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0669 - binary_accuracy: 0.9806 - auc_3: 0.9029\n",
            "Epoch 35: val_loss improved from 0.06501 to 0.06484, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 2s 67ms/step - loss: 0.0669 - binary_accuracy: 0.9806 - auc_3: 0.9029 - val_loss: 0.0648 - val_binary_accuracy: 0.9808 - val_auc_3: 0.9137\n",
            "Epoch 36/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0667 - binary_accuracy: 0.9806 - auc_3: 0.9037\n",
            "Epoch 36: val_loss improved from 0.06484 to 0.06470, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 2s 68ms/step - loss: 0.0667 - binary_accuracy: 0.9806 - auc_3: 0.9038 - val_loss: 0.0647 - val_binary_accuracy: 0.9808 - val_auc_3: 0.9144\n",
            "Epoch 37/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0665 - binary_accuracy: 0.9806 - auc_3: 0.9045\n",
            "Epoch 37: val_loss improved from 0.06470 to 0.06455, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 2s 66ms/step - loss: 0.0666 - binary_accuracy: 0.9806 - auc_3: 0.9045 - val_loss: 0.0646 - val_binary_accuracy: 0.9808 - val_auc_3: 0.9153\n",
            "Epoch 38/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0664 - binary_accuracy: 0.9806 - auc_3: 0.9054\n",
            "Epoch 38: val_loss improved from 0.06455 to 0.06440, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 2s 67ms/step - loss: 0.0664 - binary_accuracy: 0.9806 - auc_3: 0.9054 - val_loss: 0.0644 - val_binary_accuracy: 0.9808 - val_auc_3: 0.9154\n",
            "Epoch 39/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0662 - binary_accuracy: 0.9806 - auc_3: 0.9061\n",
            "Epoch 39: val_loss improved from 0.06440 to 0.06427, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 2s 68ms/step - loss: 0.0662 - binary_accuracy: 0.9807 - auc_3: 0.9062 - val_loss: 0.0643 - val_binary_accuracy: 0.9808 - val_auc_3: 0.9160\n",
            "Epoch 40/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0661 - binary_accuracy: 0.9807 - auc_3: 0.9064\n",
            "Epoch 40: val_loss improved from 0.06427 to 0.06418, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 2s 67ms/step - loss: 0.0661 - binary_accuracy: 0.9807 - auc_3: 0.9064 - val_loss: 0.0642 - val_binary_accuracy: 0.9808 - val_auc_3: 0.9170\n",
            "Epoch 41/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0660 - binary_accuracy: 0.9807 - auc_3: 0.9069\n",
            "Epoch 41: val_loss improved from 0.06418 to 0.06404, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 2s 68ms/step - loss: 0.0660 - binary_accuracy: 0.9807 - auc_3: 0.9069 - val_loss: 0.0640 - val_binary_accuracy: 0.9808 - val_auc_3: 0.9175\n",
            "Epoch 42/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0659 - binary_accuracy: 0.9807 - auc_3: 0.9078\n",
            "Epoch 42: val_loss improved from 0.06404 to 0.06391, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 2s 70ms/step - loss: 0.0659 - binary_accuracy: 0.9807 - auc_3: 0.9078 - val_loss: 0.0639 - val_binary_accuracy: 0.9808 - val_auc_3: 0.9176\n",
            "Epoch 43/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0657 - binary_accuracy: 0.9807 - auc_3: 0.9083\n",
            "Epoch 43: val_loss improved from 0.06391 to 0.06383, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 2s 67ms/step - loss: 0.0657 - binary_accuracy: 0.9807 - auc_3: 0.9083 - val_loss: 0.0638 - val_binary_accuracy: 0.9808 - val_auc_3: 0.9181\n",
            "Epoch 44/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0656 - binary_accuracy: 0.9807 - auc_3: 0.9090\n",
            "Epoch 44: val_loss improved from 0.06383 to 0.06370, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 2s 67ms/step - loss: 0.0656 - binary_accuracy: 0.9807 - auc_3: 0.9090 - val_loss: 0.0637 - val_binary_accuracy: 0.9808 - val_auc_3: 0.9187\n",
            "Epoch 45/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0655 - binary_accuracy: 0.9807 - auc_3: 0.9095\n",
            "Epoch 45: val_loss improved from 0.06370 to 0.06359, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 2s 67ms/step - loss: 0.0655 - binary_accuracy: 0.9807 - auc_3: 0.9095 - val_loss: 0.0636 - val_binary_accuracy: 0.9808 - val_auc_3: 0.9192\n",
            "Epoch 46/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0653 - binary_accuracy: 0.9807 - auc_3: 0.9099\n",
            "Epoch 46: val_loss improved from 0.06359 to 0.06348, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 2s 68ms/step - loss: 0.0654 - binary_accuracy: 0.9807 - auc_3: 0.9099 - val_loss: 0.0635 - val_binary_accuracy: 0.9809 - val_auc_3: 0.9198\n",
            "Epoch 47/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0653 - binary_accuracy: 0.9807 - auc_3: 0.9105\n",
            "Epoch 47: val_loss improved from 0.06348 to 0.06336, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 2s 67ms/step - loss: 0.0653 - binary_accuracy: 0.9807 - auc_3: 0.9104 - val_loss: 0.0634 - val_binary_accuracy: 0.9809 - val_auc_3: 0.9197\n",
            "Epoch 48/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0651 - binary_accuracy: 0.9807 - auc_3: 0.9110\n",
            "Epoch 48: val_loss improved from 0.06336 to 0.06332, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 2s 69ms/step - loss: 0.0651 - binary_accuracy: 0.9807 - auc_3: 0.9110 - val_loss: 0.0633 - val_binary_accuracy: 0.9809 - val_auc_3: 0.9201\n",
            "Epoch 49/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0651 - binary_accuracy: 0.9807 - auc_3: 0.9113\n",
            "Epoch 49: val_loss improved from 0.06332 to 0.06322, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 2s 69ms/step - loss: 0.0650 - binary_accuracy: 0.9807 - auc_3: 0.9113 - val_loss: 0.0632 - val_binary_accuracy: 0.9809 - val_auc_3: 0.9209\n",
            "Epoch 50/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0649 - binary_accuracy: 0.9807 - auc_3: 0.9120\n",
            "Epoch 50: val_loss improved from 0.06322 to 0.06310, saving model to /content/drive/My Drive/DATASET/model_rnn_simple_best2.h5\n",
            "23/23 [==============================] - 2s 69ms/step - loss: 0.0649 - binary_accuracy: 0.9807 - auc_3: 0.9120 - val_loss: 0.0631 - val_binary_accuracy: 0.9809 - val_auc_3: 0.9214\n",
            "Epoch 1/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.6573 - binary_accuracy: 0.9379 - auc_4: 0.5560\n",
            "Epoch 1: val_loss improved from inf to 0.44357, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3103: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\r23/23 [==============================] - 7s 169ms/step - loss: 0.6556 - binary_accuracy: 0.9384 - auc_4: 0.5561 - val_loss: 0.4436 - val_binary_accuracy: 0.9799 - val_auc_4: 0.6152\n",
            "Epoch 2/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.1646 - binary_accuracy: 0.9796 - auc_4: 0.6316\n",
            "Epoch 2: val_loss improved from 0.44357 to 0.10104, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 84ms/step - loss: 0.1640 - binary_accuracy: 0.9796 - auc_4: 0.6328 - val_loss: 0.1010 - val_binary_accuracy: 0.9799 - val_auc_4: 0.7529\n",
            "Epoch 3/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0865 - binary_accuracy: 0.9797 - auc_4: 0.7962\n",
            "Epoch 3: val_loss improved from 0.10104 to 0.07917, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 79ms/step - loss: 0.0864 - binary_accuracy: 0.9797 - auc_4: 0.7963 - val_loss: 0.0792 - val_binary_accuracy: 0.9799 - val_auc_4: 0.8296\n",
            "Epoch 4/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0798 - binary_accuracy: 0.9799 - auc_4: 0.8197\n",
            "Epoch 4: val_loss improved from 0.07917 to 0.07794, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 77ms/step - loss: 0.0798 - binary_accuracy: 0.9799 - auc_4: 0.8197 - val_loss: 0.0779 - val_binary_accuracy: 0.9800 - val_auc_4: 0.8330\n",
            "Epoch 5/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0789 - binary_accuracy: 0.9799 - auc_4: 0.8256\n",
            "Epoch 5: val_loss improved from 0.07794 to 0.07757, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 79ms/step - loss: 0.0789 - binary_accuracy: 0.9799 - auc_4: 0.8256 - val_loss: 0.0776 - val_binary_accuracy: 0.9800 - val_auc_4: 0.8360\n",
            "Epoch 6/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0785 - binary_accuracy: 0.9799 - auc_4: 0.8292\n",
            "Epoch 6: val_loss improved from 0.07757 to 0.07734, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 80ms/step - loss: 0.0786 - binary_accuracy: 0.9799 - auc_4: 0.8291 - val_loss: 0.0773 - val_binary_accuracy: 0.9799 - val_auc_4: 0.8383\n",
            "Epoch 7/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0784 - binary_accuracy: 0.9799 - auc_4: 0.8307\n",
            "Epoch 7: val_loss improved from 0.07734 to 0.07719, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 81ms/step - loss: 0.0784 - binary_accuracy: 0.9799 - auc_4: 0.8307 - val_loss: 0.0772 - val_binary_accuracy: 0.9799 - val_auc_4: 0.8396\n",
            "Epoch 8/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0782 - binary_accuracy: 0.9799 - auc_4: 0.8324\n",
            "Epoch 8: val_loss improved from 0.07719 to 0.07706, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 81ms/step - loss: 0.0782 - binary_accuracy: 0.9799 - auc_4: 0.8324 - val_loss: 0.0771 - val_binary_accuracy: 0.9800 - val_auc_4: 0.8409\n",
            "Epoch 9/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0781 - binary_accuracy: 0.9799 - auc_4: 0.8333\n",
            "Epoch 9: val_loss improved from 0.07706 to 0.07697, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 81ms/step - loss: 0.0781 - binary_accuracy: 0.9799 - auc_4: 0.8333 - val_loss: 0.0770 - val_binary_accuracy: 0.9799 - val_auc_4: 0.8418\n",
            "Epoch 10/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0780 - binary_accuracy: 0.9799 - auc_4: 0.8342\n",
            "Epoch 10: val_loss improved from 0.07697 to 0.07690, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 79ms/step - loss: 0.0780 - binary_accuracy: 0.9799 - auc_4: 0.8342 - val_loss: 0.0769 - val_binary_accuracy: 0.9800 - val_auc_4: 0.8425\n",
            "Epoch 11/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0779 - binary_accuracy: 0.9799 - auc_4: 0.8353\n",
            "Epoch 11: val_loss improved from 0.07690 to 0.07683, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 81ms/step - loss: 0.0779 - binary_accuracy: 0.9799 - auc_4: 0.8353 - val_loss: 0.0768 - val_binary_accuracy: 0.9800 - val_auc_4: 0.8432\n",
            "Epoch 12/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0778 - binary_accuracy: 0.9799 - auc_4: 0.8359\n",
            "Epoch 12: val_loss improved from 0.07683 to 0.07677, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 79ms/step - loss: 0.0778 - binary_accuracy: 0.9799 - auc_4: 0.8359 - val_loss: 0.0768 - val_binary_accuracy: 0.9800 - val_auc_4: 0.8438\n",
            "Epoch 13/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0778 - binary_accuracy: 0.9799 - auc_4: 0.8363\n",
            "Epoch 13: val_loss improved from 0.07677 to 0.07673, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 83ms/step - loss: 0.0778 - binary_accuracy: 0.9799 - auc_4: 0.8363 - val_loss: 0.0767 - val_binary_accuracy: 0.9800 - val_auc_4: 0.8441\n",
            "Epoch 14/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0777 - binary_accuracy: 0.9799 - auc_4: 0.8369\n",
            "Epoch 14: val_loss improved from 0.07673 to 0.07670, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 82ms/step - loss: 0.0777 - binary_accuracy: 0.9799 - auc_4: 0.8370 - val_loss: 0.0767 - val_binary_accuracy: 0.9799 - val_auc_4: 0.8443\n",
            "Epoch 15/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0776 - binary_accuracy: 0.9799 - auc_4: 0.8375\n",
            "Epoch 15: val_loss improved from 0.07670 to 0.07666, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 81ms/step - loss: 0.0776 - binary_accuracy: 0.9799 - auc_4: 0.8375 - val_loss: 0.0767 - val_binary_accuracy: 0.9799 - val_auc_4: 0.8448\n",
            "Epoch 16/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0776 - binary_accuracy: 0.9799 - auc_4: 0.8379\n",
            "Epoch 16: val_loss improved from 0.07666 to 0.07663, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 80ms/step - loss: 0.0776 - binary_accuracy: 0.9799 - auc_4: 0.8379 - val_loss: 0.0766 - val_binary_accuracy: 0.9800 - val_auc_4: 0.8451\n",
            "Epoch 17/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0775 - binary_accuracy: 0.9799 - auc_4: 0.8383\n",
            "Epoch 17: val_loss improved from 0.07663 to 0.07660, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 82ms/step - loss: 0.0775 - binary_accuracy: 0.9799 - auc_4: 0.8383 - val_loss: 0.0766 - val_binary_accuracy: 0.9800 - val_auc_4: 0.8455\n",
            "Epoch 18/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0775 - binary_accuracy: 0.9799 - auc_4: 0.8386\n",
            "Epoch 18: val_loss improved from 0.07660 to 0.07658, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 83ms/step - loss: 0.0775 - binary_accuracy: 0.9799 - auc_4: 0.8387 - val_loss: 0.0766 - val_binary_accuracy: 0.9799 - val_auc_4: 0.8455\n",
            "Epoch 19/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0774 - binary_accuracy: 0.9799 - auc_4: 0.8389\n",
            "Epoch 19: val_loss improved from 0.07658 to 0.07655, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 81ms/step - loss: 0.0775 - binary_accuracy: 0.9799 - auc_4: 0.8389 - val_loss: 0.0765 - val_binary_accuracy: 0.9799 - val_auc_4: 0.8460\n",
            "Epoch 20/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0774 - binary_accuracy: 0.9799 - auc_4: 0.8396\n",
            "Epoch 20: val_loss improved from 0.07655 to 0.07652, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 80ms/step - loss: 0.0774 - binary_accuracy: 0.9799 - auc_4: 0.8396 - val_loss: 0.0765 - val_binary_accuracy: 0.9799 - val_auc_4: 0.8460\n",
            "Epoch 21/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0774 - binary_accuracy: 0.9799 - auc_4: 0.8398\n",
            "Epoch 21: val_loss improved from 0.07652 to 0.07649, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 83ms/step - loss: 0.0774 - binary_accuracy: 0.9799 - auc_4: 0.8398 - val_loss: 0.0765 - val_binary_accuracy: 0.9799 - val_auc_4: 0.8462\n",
            "Epoch 22/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0774 - binary_accuracy: 0.9799 - auc_4: 0.8398\n",
            "Epoch 22: val_loss improved from 0.07649 to 0.07647, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 81ms/step - loss: 0.0774 - binary_accuracy: 0.9799 - auc_4: 0.8398 - val_loss: 0.0765 - val_binary_accuracy: 0.9800 - val_auc_4: 0.8463\n",
            "Epoch 23/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0773 - binary_accuracy: 0.9799 - auc_4: 0.8403\n",
            "Epoch 23: val_loss improved from 0.07647 to 0.07642, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 82ms/step - loss: 0.0773 - binary_accuracy: 0.9799 - auc_4: 0.8403 - val_loss: 0.0764 - val_binary_accuracy: 0.9800 - val_auc_4: 0.8468\n",
            "Epoch 24/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0772 - binary_accuracy: 0.9799 - auc_4: 0.8408\n",
            "Epoch 24: val_loss improved from 0.07642 to 0.07637, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 81ms/step - loss: 0.0772 - binary_accuracy: 0.9799 - auc_4: 0.8409 - val_loss: 0.0764 - val_binary_accuracy: 0.9800 - val_auc_4: 0.8471\n",
            "Epoch 25/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0772 - binary_accuracy: 0.9799 - auc_4: 0.8411\n",
            "Epoch 25: val_loss improved from 0.07637 to 0.07631, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 82ms/step - loss: 0.0772 - binary_accuracy: 0.9799 - auc_4: 0.8410 - val_loss: 0.0763 - val_binary_accuracy: 0.9799 - val_auc_4: 0.8474\n",
            "Epoch 26/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0771 - binary_accuracy: 0.9799 - auc_4: 0.8416\n",
            "Epoch 26: val_loss improved from 0.07631 to 0.07623, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 84ms/step - loss: 0.0771 - binary_accuracy: 0.9799 - auc_4: 0.8416 - val_loss: 0.0762 - val_binary_accuracy: 0.9800 - val_auc_4: 0.8476\n",
            "Epoch 27/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0770 - binary_accuracy: 0.9799 - auc_4: 0.8421\n",
            "Epoch 27: val_loss improved from 0.07623 to 0.07610, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 81ms/step - loss: 0.0770 - binary_accuracy: 0.9799 - auc_4: 0.8422 - val_loss: 0.0761 - val_binary_accuracy: 0.9800 - val_auc_4: 0.8484\n",
            "Epoch 28/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0769 - binary_accuracy: 0.9799 - auc_4: 0.8429\n",
            "Epoch 28: val_loss improved from 0.07610 to 0.07596, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 84ms/step - loss: 0.0769 - binary_accuracy: 0.9800 - auc_4: 0.8429 - val_loss: 0.0760 - val_binary_accuracy: 0.9800 - val_auc_4: 0.8489\n",
            "Epoch 29/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0767 - binary_accuracy: 0.9800 - auc_4: 0.8438\n",
            "Epoch 29: val_loss improved from 0.07596 to 0.07570, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 81ms/step - loss: 0.0767 - binary_accuracy: 0.9800 - auc_4: 0.8437 - val_loss: 0.0757 - val_binary_accuracy: 0.9800 - val_auc_4: 0.8502\n",
            "Epoch 30/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0765 - binary_accuracy: 0.9800 - auc_4: 0.8449\n",
            "Epoch 30: val_loss improved from 0.07570 to 0.07533, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 80ms/step - loss: 0.0765 - binary_accuracy: 0.9800 - auc_4: 0.8449 - val_loss: 0.0753 - val_binary_accuracy: 0.9800 - val_auc_4: 0.8525\n",
            "Epoch 31/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0761 - binary_accuracy: 0.9800 - auc_4: 0.8475\n",
            "Epoch 31: val_loss improved from 0.07533 to 0.07483, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 82ms/step - loss: 0.0761 - binary_accuracy: 0.9800 - auc_4: 0.8476 - val_loss: 0.0748 - val_binary_accuracy: 0.9800 - val_auc_4: 0.8562\n",
            "Epoch 32/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0757 - binary_accuracy: 0.9800 - auc_4: 0.8503\n",
            "Epoch 32: val_loss improved from 0.07483 to 0.07436, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 80ms/step - loss: 0.0757 - binary_accuracy: 0.9800 - auc_4: 0.8503 - val_loss: 0.0744 - val_binary_accuracy: 0.9800 - val_auc_4: 0.8606\n",
            "Epoch 33/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0754 - binary_accuracy: 0.9799 - auc_4: 0.8536\n",
            "Epoch 33: val_loss improved from 0.07436 to 0.07408, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 79ms/step - loss: 0.0754 - binary_accuracy: 0.9799 - auc_4: 0.8535 - val_loss: 0.0741 - val_binary_accuracy: 0.9800 - val_auc_4: 0.8637\n",
            "Epoch 34/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0751 - binary_accuracy: 0.9799 - auc_4: 0.8562\n",
            "Epoch 34: val_loss improved from 0.07408 to 0.07384, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 80ms/step - loss: 0.0751 - binary_accuracy: 0.9799 - auc_4: 0.8563 - val_loss: 0.0738 - val_binary_accuracy: 0.9800 - val_auc_4: 0.8659\n",
            "Epoch 35/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0748 - binary_accuracy: 0.9799 - auc_4: 0.8581\n",
            "Epoch 35: val_loss improved from 0.07384 to 0.07366, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 78ms/step - loss: 0.0748 - binary_accuracy: 0.9799 - auc_4: 0.8581 - val_loss: 0.0737 - val_binary_accuracy: 0.9800 - val_auc_4: 0.8672\n",
            "Epoch 36/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0747 - binary_accuracy: 0.9799 - auc_4: 0.8595\n",
            "Epoch 36: val_loss improved from 0.07366 to 0.07345, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 80ms/step - loss: 0.0746 - binary_accuracy: 0.9799 - auc_4: 0.8596 - val_loss: 0.0735 - val_binary_accuracy: 0.9800 - val_auc_4: 0.8688\n",
            "Epoch 37/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0745 - binary_accuracy: 0.9799 - auc_4: 0.8606\n",
            "Epoch 37: val_loss improved from 0.07345 to 0.07334, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 78ms/step - loss: 0.0745 - binary_accuracy: 0.9800 - auc_4: 0.8606 - val_loss: 0.0733 - val_binary_accuracy: 0.9800 - val_auc_4: 0.8692\n",
            "Epoch 38/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0743 - binary_accuracy: 0.9800 - auc_4: 0.8620\n",
            "Epoch 38: val_loss improved from 0.07334 to 0.07319, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 77ms/step - loss: 0.0743 - binary_accuracy: 0.9800 - auc_4: 0.8620 - val_loss: 0.0732 - val_binary_accuracy: 0.9800 - val_auc_4: 0.8702\n",
            "Epoch 39/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0742 - binary_accuracy: 0.9800 - auc_4: 0.8628\n",
            "Epoch 39: val_loss improved from 0.07319 to 0.07307, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 80ms/step - loss: 0.0742 - binary_accuracy: 0.9800 - auc_4: 0.8628 - val_loss: 0.0731 - val_binary_accuracy: 0.9800 - val_auc_4: 0.8709\n",
            "Epoch 40/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0740 - binary_accuracy: 0.9800 - auc_4: 0.8639\n",
            "Epoch 40: val_loss improved from 0.07307 to 0.07292, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 76ms/step - loss: 0.0740 - binary_accuracy: 0.9800 - auc_4: 0.8639 - val_loss: 0.0729 - val_binary_accuracy: 0.9800 - val_auc_4: 0.8717\n",
            "Epoch 41/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0739 - binary_accuracy: 0.9800 - auc_4: 0.8648\n",
            "Epoch 41: val_loss improved from 0.07292 to 0.07278, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 76ms/step - loss: 0.0739 - binary_accuracy: 0.9800 - auc_4: 0.8647 - val_loss: 0.0728 - val_binary_accuracy: 0.9800 - val_auc_4: 0.8725\n",
            "Epoch 42/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0737 - binary_accuracy: 0.9800 - auc_4: 0.8659\n",
            "Epoch 42: val_loss improved from 0.07278 to 0.07266, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 81ms/step - loss: 0.0737 - binary_accuracy: 0.9800 - auc_4: 0.8659 - val_loss: 0.0727 - val_binary_accuracy: 0.9800 - val_auc_4: 0.8731\n",
            "Epoch 43/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0736 - binary_accuracy: 0.9800 - auc_4: 0.8665\n",
            "Epoch 43: val_loss improved from 0.07266 to 0.07254, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 77ms/step - loss: 0.0736 - binary_accuracy: 0.9800 - auc_4: 0.8665 - val_loss: 0.0725 - val_binary_accuracy: 0.9800 - val_auc_4: 0.8742\n",
            "Epoch 44/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0735 - binary_accuracy: 0.9800 - auc_4: 0.8675\n",
            "Epoch 44: val_loss improved from 0.07254 to 0.07245, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 79ms/step - loss: 0.0735 - binary_accuracy: 0.9800 - auc_4: 0.8674 - val_loss: 0.0725 - val_binary_accuracy: 0.9800 - val_auc_4: 0.8749\n",
            "Epoch 45/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0733 - binary_accuracy: 0.9800 - auc_4: 0.8685\n",
            "Epoch 45: val_loss improved from 0.07245 to 0.07236, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 78ms/step - loss: 0.0733 - binary_accuracy: 0.9800 - auc_4: 0.8685 - val_loss: 0.0724 - val_binary_accuracy: 0.9800 - val_auc_4: 0.8754\n",
            "Epoch 46/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0733 - binary_accuracy: 0.9800 - auc_4: 0.8690\n",
            "Epoch 46: val_loss improved from 0.07236 to 0.07227, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 77ms/step - loss: 0.0733 - binary_accuracy: 0.9800 - auc_4: 0.8690 - val_loss: 0.0723 - val_binary_accuracy: 0.9800 - val_auc_4: 0.8757\n",
            "Epoch 47/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0732 - binary_accuracy: 0.9800 - auc_4: 0.8695\n",
            "Epoch 47: val_loss improved from 0.07227 to 0.07221, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 77ms/step - loss: 0.0732 - binary_accuracy: 0.9800 - auc_4: 0.8695 - val_loss: 0.0722 - val_binary_accuracy: 0.9800 - val_auc_4: 0.8763\n",
            "Epoch 48/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0730 - binary_accuracy: 0.9800 - auc_4: 0.8702\n",
            "Epoch 48: val_loss improved from 0.07221 to 0.07214, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 78ms/step - loss: 0.0731 - binary_accuracy: 0.9800 - auc_4: 0.8701 - val_loss: 0.0721 - val_binary_accuracy: 0.9800 - val_auc_4: 0.8766\n",
            "Epoch 49/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0730 - binary_accuracy: 0.9800 - auc_4: 0.8707\n",
            "Epoch 49: val_loss improved from 0.07214 to 0.07208, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 76ms/step - loss: 0.0730 - binary_accuracy: 0.9800 - auc_4: 0.8707 - val_loss: 0.0721 - val_binary_accuracy: 0.9800 - val_auc_4: 0.8770\n",
            "Epoch 50/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0729 - binary_accuracy: 0.9800 - auc_4: 0.8711\n",
            "Epoch 50: val_loss improved from 0.07208 to 0.07202, saving model to /content/drive/My Drive/DATASET/model_lstm_best2.h5\n",
            "23/23 [==============================] - 2s 77ms/step - loss: 0.0729 - binary_accuracy: 0.9800 - auc_4: 0.8711 - val_loss: 0.0720 - val_binary_accuracy: 0.9801 - val_auc_4: 0.8772\n",
            "Epoch 1/50\n",
            "23/23 [==============================] - ETA: 0s - loss: 0.5063 - binary_accuracy: 0.9069 - auc_5: 0.5539\n",
            "Epoch 1: val_loss improved from inf to 0.09619, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3103: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\r23/23 [==============================] - 7s 159ms/step - loss: 0.5063 - binary_accuracy: 0.9069 - auc_5: 0.5539 - val_loss: 0.0962 - val_binary_accuracy: 0.9799 - val_auc_5: 0.6672\n",
            "Epoch 2/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.1086 - binary_accuracy: 0.9791 - auc_5: 0.7308\n",
            "Epoch 2: val_loss improved from 0.09619 to 0.08302, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 80ms/step - loss: 0.1083 - binary_accuracy: 0.9792 - auc_5: 0.7314 - val_loss: 0.0830 - val_binary_accuracy: 0.9800 - val_auc_5: 0.8182\n",
            "Epoch 3/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0820 - binary_accuracy: 0.9798 - auc_5: 0.8146\n",
            "Epoch 3: val_loss improved from 0.08302 to 0.07864, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 71ms/step - loss: 0.0819 - binary_accuracy: 0.9798 - auc_5: 0.8146 - val_loss: 0.0786 - val_binary_accuracy: 0.9799 - val_auc_5: 0.8294\n",
            "Epoch 4/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0797 - binary_accuracy: 0.9798 - auc_5: 0.8179\n",
            "Epoch 4: val_loss improved from 0.07864 to 0.07760, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 70ms/step - loss: 0.0797 - binary_accuracy: 0.9798 - auc_5: 0.8180 - val_loss: 0.0776 - val_binary_accuracy: 0.9799 - val_auc_5: 0.8354\n",
            "Epoch 5/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0790 - binary_accuracy: 0.9799 - auc_5: 0.8255\n",
            "Epoch 5: val_loss improved from 0.07760 to 0.07734, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 70ms/step - loss: 0.0791 - binary_accuracy: 0.9798 - auc_5: 0.8255 - val_loss: 0.0773 - val_binary_accuracy: 0.9800 - val_auc_5: 0.8382\n",
            "Epoch 6/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0788 - binary_accuracy: 0.9799 - auc_5: 0.8283\n",
            "Epoch 6: val_loss improved from 0.07734 to 0.07717, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 73ms/step - loss: 0.0788 - binary_accuracy: 0.9799 - auc_5: 0.8284 - val_loss: 0.0772 - val_binary_accuracy: 0.9800 - val_auc_5: 0.8398\n",
            "Epoch 7/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0786 - binary_accuracy: 0.9799 - auc_5: 0.8301\n",
            "Epoch 7: val_loss improved from 0.07717 to 0.07704, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 71ms/step - loss: 0.0786 - binary_accuracy: 0.9799 - auc_5: 0.8300 - val_loss: 0.0770 - val_binary_accuracy: 0.9800 - val_auc_5: 0.8411\n",
            "Epoch 8/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0784 - binary_accuracy: 0.9799 - auc_5: 0.8316\n",
            "Epoch 8: val_loss improved from 0.07704 to 0.07692, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 73ms/step - loss: 0.0784 - binary_accuracy: 0.9799 - auc_5: 0.8316 - val_loss: 0.0769 - val_binary_accuracy: 0.9800 - val_auc_5: 0.8423\n",
            "Epoch 9/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0782 - binary_accuracy: 0.9799 - auc_5: 0.8328\n",
            "Epoch 9: val_loss improved from 0.07692 to 0.07684, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 72ms/step - loss: 0.0782 - binary_accuracy: 0.9799 - auc_5: 0.8328 - val_loss: 0.0768 - val_binary_accuracy: 0.9800 - val_auc_5: 0.8430\n",
            "Epoch 10/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0782 - binary_accuracy: 0.9799 - auc_5: 0.8338\n",
            "Epoch 10: val_loss improved from 0.07684 to 0.07678, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 72ms/step - loss: 0.0781 - binary_accuracy: 0.9799 - auc_5: 0.8338 - val_loss: 0.0768 - val_binary_accuracy: 0.9799 - val_auc_5: 0.8436\n",
            "Epoch 11/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0781 - binary_accuracy: 0.9799 - auc_5: 0.8345\n",
            "Epoch 11: val_loss improved from 0.07678 to 0.07672, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 73ms/step - loss: 0.0781 - binary_accuracy: 0.9799 - auc_5: 0.8345 - val_loss: 0.0767 - val_binary_accuracy: 0.9800 - val_auc_5: 0.8440\n",
            "Epoch 12/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0779 - binary_accuracy: 0.9799 - auc_5: 0.8349\n",
            "Epoch 12: val_loss improved from 0.07672 to 0.07667, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 71ms/step - loss: 0.0780 - binary_accuracy: 0.9799 - auc_5: 0.8350 - val_loss: 0.0767 - val_binary_accuracy: 0.9799 - val_auc_5: 0.8446\n",
            "Epoch 13/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0779 - binary_accuracy: 0.9799 - auc_5: 0.8359\n",
            "Epoch 13: val_loss improved from 0.07667 to 0.07663, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 72ms/step - loss: 0.0779 - binary_accuracy: 0.9799 - auc_5: 0.8360 - val_loss: 0.0766 - val_binary_accuracy: 0.9800 - val_auc_5: 0.8449\n",
            "Epoch 14/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0778 - binary_accuracy: 0.9799 - auc_5: 0.8367\n",
            "Epoch 14: val_loss improved from 0.07663 to 0.07658, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 72ms/step - loss: 0.0778 - binary_accuracy: 0.9799 - auc_5: 0.8366 - val_loss: 0.0766 - val_binary_accuracy: 0.9800 - val_auc_5: 0.8451\n",
            "Epoch 15/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0776 - binary_accuracy: 0.9800 - auc_5: 0.8373\n",
            "Epoch 15: val_loss improved from 0.07658 to 0.07650, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 74ms/step - loss: 0.0777 - binary_accuracy: 0.9799 - auc_5: 0.8373 - val_loss: 0.0765 - val_binary_accuracy: 0.9799 - val_auc_5: 0.8456\n",
            "Epoch 16/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0776 - binary_accuracy: 0.9799 - auc_5: 0.8378\n",
            "Epoch 16: val_loss improved from 0.07650 to 0.07642, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 75ms/step - loss: 0.0776 - binary_accuracy: 0.9799 - auc_5: 0.8378 - val_loss: 0.0764 - val_binary_accuracy: 0.9800 - val_auc_5: 0.8460\n",
            "Epoch 17/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0775 - binary_accuracy: 0.9799 - auc_5: 0.8384\n",
            "Epoch 17: val_loss improved from 0.07642 to 0.07626, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 73ms/step - loss: 0.0775 - binary_accuracy: 0.9799 - auc_5: 0.8383 - val_loss: 0.0763 - val_binary_accuracy: 0.9800 - val_auc_5: 0.8469\n",
            "Epoch 18/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0772 - binary_accuracy: 0.9799 - auc_5: 0.8401\n",
            "Epoch 18: val_loss improved from 0.07626 to 0.07588, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 72ms/step - loss: 0.0772 - binary_accuracy: 0.9799 - auc_5: 0.8401 - val_loss: 0.0759 - val_binary_accuracy: 0.9800 - val_auc_5: 0.8490\n",
            "Epoch 19/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0768 - binary_accuracy: 0.9799 - auc_5: 0.8422\n",
            "Epoch 19: val_loss improved from 0.07588 to 0.07521, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 73ms/step - loss: 0.0768 - binary_accuracy: 0.9799 - auc_5: 0.8423 - val_loss: 0.0752 - val_binary_accuracy: 0.9800 - val_auc_5: 0.8534\n",
            "Epoch 20/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0761 - binary_accuracy: 0.9799 - auc_5: 0.8474\n",
            "Epoch 20: val_loss improved from 0.07521 to 0.07446, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 71ms/step - loss: 0.0761 - binary_accuracy: 0.9799 - auc_5: 0.8474 - val_loss: 0.0745 - val_binary_accuracy: 0.9800 - val_auc_5: 0.8601\n",
            "Epoch 21/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0755 - binary_accuracy: 0.9799 - auc_5: 0.8527\n",
            "Epoch 21: val_loss improved from 0.07446 to 0.07398, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 69ms/step - loss: 0.0755 - binary_accuracy: 0.9799 - auc_5: 0.8527 - val_loss: 0.0740 - val_binary_accuracy: 0.9800 - val_auc_5: 0.8646\n",
            "Epoch 22/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0751 - binary_accuracy: 0.9799 - auc_5: 0.8558\n",
            "Epoch 22: val_loss improved from 0.07398 to 0.07366, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 70ms/step - loss: 0.0751 - binary_accuracy: 0.9799 - auc_5: 0.8558 - val_loss: 0.0737 - val_binary_accuracy: 0.9800 - val_auc_5: 0.8668\n",
            "Epoch 23/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0748 - binary_accuracy: 0.9800 - auc_5: 0.8583\n",
            "Epoch 23: val_loss improved from 0.07366 to 0.07341, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 70ms/step - loss: 0.0748 - binary_accuracy: 0.9800 - auc_5: 0.8583 - val_loss: 0.0734 - val_binary_accuracy: 0.9800 - val_auc_5: 0.8684\n",
            "Epoch 24/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0745 - binary_accuracy: 0.9800 - auc_5: 0.8603\n",
            "Epoch 24: val_loss improved from 0.07341 to 0.07317, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 71ms/step - loss: 0.0745 - binary_accuracy: 0.9800 - auc_5: 0.8603 - val_loss: 0.0732 - val_binary_accuracy: 0.9800 - val_auc_5: 0.8699\n",
            "Epoch 25/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0743 - binary_accuracy: 0.9800 - auc_5: 0.8616\n",
            "Epoch 25: val_loss improved from 0.07317 to 0.07296, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 70ms/step - loss: 0.0743 - binary_accuracy: 0.9800 - auc_5: 0.8616 - val_loss: 0.0730 - val_binary_accuracy: 0.9800 - val_auc_5: 0.8715\n",
            "Epoch 26/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0741 - binary_accuracy: 0.9800 - auc_5: 0.8634\n",
            "Epoch 26: val_loss improved from 0.07296 to 0.07281, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 69ms/step - loss: 0.0741 - binary_accuracy: 0.9800 - auc_5: 0.8634 - val_loss: 0.0728 - val_binary_accuracy: 0.9800 - val_auc_5: 0.8722\n",
            "Epoch 27/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0740 - binary_accuracy: 0.9800 - auc_5: 0.8641\n",
            "Epoch 27: val_loss improved from 0.07281 to 0.07267, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 69ms/step - loss: 0.0740 - binary_accuracy: 0.9800 - auc_5: 0.8642 - val_loss: 0.0727 - val_binary_accuracy: 0.9800 - val_auc_5: 0.8733\n",
            "Epoch 28/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0738 - binary_accuracy: 0.9800 - auc_5: 0.8653\n",
            "Epoch 28: val_loss improved from 0.07267 to 0.07256, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 69ms/step - loss: 0.0738 - binary_accuracy: 0.9800 - auc_5: 0.8653 - val_loss: 0.0726 - val_binary_accuracy: 0.9800 - val_auc_5: 0.8741\n",
            "Epoch 29/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0737 - binary_accuracy: 0.9800 - auc_5: 0.8662\n",
            "Epoch 29: val_loss improved from 0.07256 to 0.07251, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 69ms/step - loss: 0.0737 - binary_accuracy: 0.9800 - auc_5: 0.8661 - val_loss: 0.0725 - val_binary_accuracy: 0.9800 - val_auc_5: 0.8740\n",
            "Epoch 30/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0736 - binary_accuracy: 0.9800 - auc_5: 0.8668\n",
            "Epoch 30: val_loss improved from 0.07251 to 0.07237, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 68ms/step - loss: 0.0736 - binary_accuracy: 0.9800 - auc_5: 0.8669 - val_loss: 0.0724 - val_binary_accuracy: 0.9801 - val_auc_5: 0.8751\n",
            "Epoch 31/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0734 - binary_accuracy: 0.9800 - auc_5: 0.8680\n",
            "Epoch 31: val_loss improved from 0.07237 to 0.07220, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 72ms/step - loss: 0.0734 - binary_accuracy: 0.9800 - auc_5: 0.8679 - val_loss: 0.0722 - val_binary_accuracy: 0.9801 - val_auc_5: 0.8763\n",
            "Epoch 32/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0732 - binary_accuracy: 0.9800 - auc_5: 0.8689\n",
            "Epoch 32: val_loss improved from 0.07220 to 0.07194, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 68ms/step - loss: 0.0732 - binary_accuracy: 0.9800 - auc_5: 0.8689 - val_loss: 0.0719 - val_binary_accuracy: 0.9801 - val_auc_5: 0.8776\n",
            "Epoch 33/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0729 - binary_accuracy: 0.9800 - auc_5: 0.8705\n",
            "Epoch 33: val_loss improved from 0.07194 to 0.07145, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 70ms/step - loss: 0.0729 - binary_accuracy: 0.9800 - auc_5: 0.8705 - val_loss: 0.0715 - val_binary_accuracy: 0.9802 - val_auc_5: 0.8801\n",
            "Epoch 34/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0724 - binary_accuracy: 0.9801 - auc_5: 0.8735\n",
            "Epoch 34: val_loss improved from 0.07145 to 0.07078, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 68ms/step - loss: 0.0724 - binary_accuracy: 0.9801 - auc_5: 0.8735 - val_loss: 0.0708 - val_binary_accuracy: 0.9802 - val_auc_5: 0.8842\n",
            "Epoch 35/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0719 - binary_accuracy: 0.9801 - auc_5: 0.8766\n",
            "Epoch 35: val_loss improved from 0.07078 to 0.07028, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 69ms/step - loss: 0.0719 - binary_accuracy: 0.9801 - auc_5: 0.8766 - val_loss: 0.0703 - val_binary_accuracy: 0.9803 - val_auc_5: 0.8872\n",
            "Epoch 36/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0715 - binary_accuracy: 0.9802 - auc_5: 0.8788\n",
            "Epoch 36: val_loss improved from 0.07028 to 0.06990, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 68ms/step - loss: 0.0715 - binary_accuracy: 0.9802 - auc_5: 0.8788 - val_loss: 0.0699 - val_binary_accuracy: 0.9803 - val_auc_5: 0.8896\n",
            "Epoch 37/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0712 - binary_accuracy: 0.9802 - auc_5: 0.8806\n",
            "Epoch 37: val_loss improved from 0.06990 to 0.06964, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 68ms/step - loss: 0.0712 - binary_accuracy: 0.9802 - auc_5: 0.8806 - val_loss: 0.0696 - val_binary_accuracy: 0.9803 - val_auc_5: 0.8912\n",
            "Epoch 38/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0710 - binary_accuracy: 0.9802 - auc_5: 0.8821\n",
            "Epoch 38: val_loss improved from 0.06964 to 0.06942, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 70ms/step - loss: 0.0710 - binary_accuracy: 0.9802 - auc_5: 0.8820 - val_loss: 0.0694 - val_binary_accuracy: 0.9803 - val_auc_5: 0.8923\n",
            "Epoch 39/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0708 - binary_accuracy: 0.9802 - auc_5: 0.8834\n",
            "Epoch 39: val_loss improved from 0.06942 to 0.06927, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 70ms/step - loss: 0.0708 - binary_accuracy: 0.9802 - auc_5: 0.8834 - val_loss: 0.0693 - val_binary_accuracy: 0.9803 - val_auc_5: 0.8931\n",
            "Epoch 40/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0705 - binary_accuracy: 0.9802 - auc_5: 0.8847\n",
            "Epoch 40: val_loss improved from 0.06927 to 0.06910, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 68ms/step - loss: 0.0705 - binary_accuracy: 0.9802 - auc_5: 0.8847 - val_loss: 0.0691 - val_binary_accuracy: 0.9803 - val_auc_5: 0.8942\n",
            "Epoch 41/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0704 - binary_accuracy: 0.9802 - auc_5: 0.8856\n",
            "Epoch 41: val_loss improved from 0.06910 to 0.06897, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 67ms/step - loss: 0.0704 - binary_accuracy: 0.9802 - auc_5: 0.8856 - val_loss: 0.0690 - val_binary_accuracy: 0.9803 - val_auc_5: 0.8948\n",
            "Epoch 42/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0702 - binary_accuracy: 0.9802 - auc_5: 0.8867\n",
            "Epoch 42: val_loss improved from 0.06897 to 0.06885, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 69ms/step - loss: 0.0702 - binary_accuracy: 0.9802 - auc_5: 0.8868 - val_loss: 0.0689 - val_binary_accuracy: 0.9803 - val_auc_5: 0.8952\n",
            "Epoch 43/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0701 - binary_accuracy: 0.9803 - auc_5: 0.8872\n",
            "Epoch 43: val_loss improved from 0.06885 to 0.06874, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 69ms/step - loss: 0.0701 - binary_accuracy: 0.9802 - auc_5: 0.8872 - val_loss: 0.0687 - val_binary_accuracy: 0.9803 - val_auc_5: 0.8959\n",
            "Epoch 44/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0700 - binary_accuracy: 0.9803 - auc_5: 0.8882\n",
            "Epoch 44: val_loss improved from 0.06874 to 0.06865, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 70ms/step - loss: 0.0700 - binary_accuracy: 0.9803 - auc_5: 0.8882 - val_loss: 0.0686 - val_binary_accuracy: 0.9803 - val_auc_5: 0.8960\n",
            "Epoch 45/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0699 - binary_accuracy: 0.9803 - auc_5: 0.8887\n",
            "Epoch 45: val_loss improved from 0.06865 to 0.06854, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 68ms/step - loss: 0.0699 - binary_accuracy: 0.9803 - auc_5: 0.8887 - val_loss: 0.0685 - val_binary_accuracy: 0.9803 - val_auc_5: 0.8965\n",
            "Epoch 46/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0697 - binary_accuracy: 0.9803 - auc_5: 0.8894\n",
            "Epoch 46: val_loss improved from 0.06854 to 0.06840, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 68ms/step - loss: 0.0697 - binary_accuracy: 0.9803 - auc_5: 0.8893 - val_loss: 0.0684 - val_binary_accuracy: 0.9804 - val_auc_5: 0.8972\n",
            "Epoch 47/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0696 - binary_accuracy: 0.9803 - auc_5: 0.8901\n",
            "Epoch 47: val_loss improved from 0.06840 to 0.06823, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 70ms/step - loss: 0.0696 - binary_accuracy: 0.9803 - auc_5: 0.8900 - val_loss: 0.0682 - val_binary_accuracy: 0.9804 - val_auc_5: 0.8981\n",
            "Epoch 48/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0694 - binary_accuracy: 0.9803 - auc_5: 0.8908\n",
            "Epoch 48: val_loss improved from 0.06823 to 0.06807, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 68ms/step - loss: 0.0694 - binary_accuracy: 0.9803 - auc_5: 0.8909 - val_loss: 0.0681 - val_binary_accuracy: 0.9804 - val_auc_5: 0.8986\n",
            "Epoch 49/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0693 - binary_accuracy: 0.9803 - auc_5: 0.8919\n",
            "Epoch 49: val_loss improved from 0.06807 to 0.06787, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 68ms/step - loss: 0.0693 - binary_accuracy: 0.9803 - auc_5: 0.8919 - val_loss: 0.0679 - val_binary_accuracy: 0.9804 - val_auc_5: 0.8995\n",
            "Epoch 50/50\n",
            "22/23 [===========================>..] - ETA: 0s - loss: 0.0690 - binary_accuracy: 0.9804 - auc_5: 0.8926\n",
            "Epoch 50: val_loss improved from 0.06787 to 0.06764, saving model to /content/drive/My Drive/DATASET/model_gru_best2.h5\n",
            "23/23 [==============================] - 2s 67ms/step - loss: 0.0691 - binary_accuracy: 0.9804 - auc_5: 0.8925 - val_loss: 0.0676 - val_binary_accuracy: 0.9805 - val_auc_5: 0.9010\n",
            "445/445 [==============================] - 1s 2ms/step\n",
            "445/445 [==============================] - 2s 2ms/step\n",
            "445/445 [==============================] - 2s 2ms/step\n",
            "Evaluation Results:\n",
            "Simple RNN - Fmax: 0.2055, AUC: 0.1271\n",
            "LSTM - Fmax: 0.0948, AUC: 0.0504\n",
            "GRU - Fmax: 0.1377, AUC: 0.0777\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Paths of models\n",
        "path_rnn = '/content/drive/My Drive/DATASET/model_rnn_simple_best2.h5'\n",
        "path_lstm = '/content/drive/My Drive/DATASET/model_lstm_best2.h5'\n",
        "path_gru = '/content/drive/My Drive/DATASET/model_gru_best2.h5'\n",
        "\n",
        "# load models\n",
        "model_rnn_simple = load_model(path_rnn)\n",
        "model_lstm = load_model(path_lstm)\n",
        "model_gru = load_model(path_gru)"
      ],
      "metadata": {
        "id": "Yr6bk-OLIIIn"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Display the summary of the RNN model\n",
        "print(\"RNN 2 Model Summary:\")\n",
        "model_rnn_simple.summary()\n",
        "\n",
        "# Display the summary of the LSTM model\n",
        "print(\"\\nLSTM 2 Model Summary:\")\n",
        "model_lstm.summary()\n",
        "\n",
        "# Display the summary of the GRU model\n",
        "print(\"\\nGRU 2 Model Summary:\")\n",
        "model_gru.summary()\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "N4olnxanOwJQ",
        "outputId": "42627499-f3f5-4ee1-dc80-871b2c8b7394"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "RNN 2 Model Summary:\n",
            "Model: \"sequential_3\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " reshape_3 (Reshape)         (None, 1, 1024)           0         \n",
            "                                                                 \n",
            " simple_rnn_2 (SimpleRNN)    (None, 1, 128)            147584    \n",
            "                                                                 \n",
            " dropout_9 (Dropout)         (None, 1, 128)            0         \n",
            "                                                                 \n",
            " simple_rnn_3 (SimpleRNN)    (None, 128)               32896     \n",
            "                                                                 \n",
            " dropout_10 (Dropout)        (None, 128)               0         \n",
            "                                                                 \n",
            " dense_6 (Dense)             (None, 512)               66048     \n",
            "                                                                 \n",
            " dropout_11 (Dropout)        (None, 512)               0         \n",
            "                                                                 \n",
            " dense_7 (Dense)             (None, 1500)              769500    \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 1016028 (3.88 MB)\n",
            "Trainable params: 1016028 (3.88 MB)\n",
            "Non-trainable params: 0 (0.00 Byte)\n",
            "_________________________________________________________________\n",
            "\n",
            "LSTM 2 Model Summary:\n",
            "Model: \"sequential_4\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " reshape_4 (Reshape)         (None, 1, 1024)           0         \n",
            "                                                                 \n",
            " lstm_2 (LSTM)               (None, 1, 128)            590336    \n",
            "                                                                 \n",
            " dropout_12 (Dropout)        (None, 1, 128)            0         \n",
            "                                                                 \n",
            " lstm_3 (LSTM)               (None, 128)               131584    \n",
            "                                                                 \n",
            " dropout_13 (Dropout)        (None, 128)               0         \n",
            "                                                                 \n",
            " dense_8 (Dense)             (None, 512)               66048     \n",
            "                                                                 \n",
            " dropout_14 (Dropout)        (None, 512)               0         \n",
            "                                                                 \n",
            " dense_9 (Dense)             (None, 1500)              769500    \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 1557468 (5.94 MB)\n",
            "Trainable params: 1557468 (5.94 MB)\n",
            "Non-trainable params: 0 (0.00 Byte)\n",
            "_________________________________________________________________\n",
            "\n",
            "GRU 2 Model Summary:\n",
            "Model: \"sequential_5\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " reshape_5 (Reshape)         (None, 1, 1024)           0         \n",
            "                                                                 \n",
            " gru_2 (GRU)                 (None, 1, 128)            443136    \n",
            "                                                                 \n",
            " dropout_15 (Dropout)        (None, 1, 128)            0         \n",
            "                                                                 \n",
            " gru_3 (GRU)                 (None, 128)               99072     \n",
            "                                                                 \n",
            " dropout_16 (Dropout)        (None, 128)               0         \n",
            "                                                                 \n",
            " dense_10 (Dense)            (None, 512)               66048     \n",
            "                                                                 \n",
            " dropout_17 (Dropout)        (None, 512)               0         \n",
            "                                                                 \n",
            " dense_11 (Dense)            (None, 1500)              769500    \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 1377756 (5.26 MB)\n",
            "Trainable params: 1377756 (5.26 MB)\n",
            "Non-trainable params: 0 (0.00 Byte)\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "#### Models Prediction § Evaluation"
      ],
      "metadata": {
        "id": "w0UgD9crG4h6"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "def evaluate_model(model, X, y):\n",
        "    \"\"\"Evaluate the model using Fmax and AUPR metrics.\"\"\"\n",
        "    # Get model predictions as probabilities\n",
        "    y_pred_probs = model.predict(X)\n",
        "\n",
        "    # Ensure y and y_pred_probs are numpy arrays\n",
        "    if isinstance(y, pd.DataFrame):\n",
        "        y = y.values\n",
        "    if isinstance(y_pred_probs, pd.DataFrame):\n",
        "        y_pred_probs = y_pred_probs.values\n",
        "\n",
        "    fmax_scores = []\n",
        "    aupr_scores = []  # Using average_precision_score for simplicity\n",
        "\n",
        "    # Calculate Fmax and AUPR for each label\n",
        "    for i in range(y.shape[1]):\n",
        "        precision, recall, thresholds = precision_recall_curve(y[:, i], y_pred_probs[:, i])\n",
        "        if len(precision) > 1 and len(recall) > 1:\n",
        "            f1_scores = 2 * (precision * recall) / (np.maximum(precision + recall, 1e-5))\n",
        "            fmax = np.max(f1_scores)\n",
        "            fmax_scores.append(fmax)\n",
        "            aupr_score = average_precision_score(y[:, i], y_pred_probs[:, i])  # Direct calculation of AUPR\n",
        "            aupr_scores.append(aupr_score)\n",
        "        else:\n",
        "            fmax_scores.append(0)\n",
        "            aupr_scores.append(0)\n",
        "\n",
        "    average_fmax = np.mean(fmax_scores)\n",
        "    average_aupr = np.mean(aupr_scores)  # AUPR directly calculated\n",
        "\n",
        "    return average_fmax, average_aupr\n",
        "\n",
        "# Evaluate each model on the test set\n",
        "fmax_rnn, aupr_rnn = evaluate_model(model_rnn_simple, X_test, y_test)\n",
        "fmax_lstm, aupr_lstm = evaluate_model(model_lstm, X_test, y_test)\n",
        "fmax_gru, aupr_gru = evaluate_model(model_gru, X_test, y_test)\n",
        "\n",
        "# Print the evaluation results\n",
        "print(\"Evaluation Results:\")\n",
        "print(f\"Simple RNN - Fmax: {fmax_rnn:.4f}, AUPR: {aupr_rnn:.4f}\")\n",
        "print(f\"LSTM - Fmax: {fmax_lstm:.4f}, AUPR: {aupr_lstm:.4f}\")\n",
        "print(f\"GRU - Fmax: {fmax_gru:.4f}, AUPR: {aupr_gru:.4f}\")\n"
      ],
      "metadata": {
        "id": "-Rg8ip3NH57t"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Comparative Analysis\n",
        "print(f\"Model Performance Comparison:\")\n",
        "print(f\"{'Model':<20}{'Fmax':<10}{'AUPR':<10}\")\n",
        "print(f\"{'-'*40}\")\n",
        "print(f\"{'Simple RNN 2':<20}{fmax_rnn:<10.4f}{aupr_rnn:<10.4f}\")\n",
        "print(f\"{'LSTM 2':<20}{fmax_lstm:<10.4f}{aupr_lstm:<10.4f}\")\n",
        "print(f\"{'GRU 2':<20}{fmax_gru:<10.4f}{aupr_gru:<10.4f}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "gjPuFbNLrHE2",
        "outputId": "43a8b29f-bd72-4c29-c3d0-276c03ce10e4"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model Performance Comparison:\n",
            "Model               Fmax      AUPR      \n",
            "----------------------------------------\n",
            "Simple RNN 2        0.2055    0.1301    \n",
            "LSTM 2              0.0948    0.0515    \n",
            "GRU 2               0.1377    0.0796    \n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# models paths\n",
        "path_rnn = '/content/drive/My Drive/DATASET/model_rnn_simple_best2.h5'\n",
        "path_lstm = '/content/drive/My Drive/DATASET/model_lstm_best2.h5'\n",
        "path_gru = '/content/drive/My Drive/DATASET/model_gru_best2.h5'\n",
        "\n",
        "# model loading\n",
        "model_rnn = load_model(path_rnn)\n",
        "model_lstm = load_model(path_lstm)\n",
        "model_gru = load_model(path_gru)\n",
        "\n",
        "# if y_test is a  numpy\n",
        "if isinstance(y_test, pd.DataFrame):\n",
        "    y_test = y_test.values\n",
        "\n",
        "# the function to  calculate  confusion matrix\n",
        "def display_confusion_matrix(model, model_name):\n",
        "    # do the predicitons\n",
        "    predictions = model.predict(X_test)\n",
        "\n",
        "\n",
        "    # Threshold for converting probabilities into binary labels\n",
        "    binary_predictions = (predictions > 0.5).astype(int)\n",
        "\n",
        "   # Initialize a global confusion matrix\n",
        "    global_confusion_matrix = np.zeros((2, 2), dtype=int)\n",
        "\n",
        "   # Calculate confusion matrices for each label and sum them\n",
        "    NUM_LABELS = 1500\n",
        "    for i in range(NUM_LABELS):\n",
        "        cm = confusion_matrix(y_test[:, i], binary_predictions[:, i])\n",
        "        global_confusion_matrix += cm\n",
        "\n",
        "   # Display the global confusion matrix\n",
        "    disp = ConfusionMatrixDisplay(confusion_matrix=global_confusion_matrix, display_labels=['Negative', 'Positive'])\n",
        "    plt.figure(figsize=(8, 6))\n",
        "    disp.plot(cmap=plt.cm.Blues)\n",
        "    plt.title(f'Global Confusion Matrix for All Labels ({model_name})')\n",
        "    plt.show()\n",
        "\n",
        "   # Display the global confusion matrix as a table\n",
        "    print(f\"Global Confusion Matrix for {model_name}:\")\n",
        "    print(\"                          Predicted\")\n",
        "    print(\"               ------------------------------\")\n",
        "    print(f\"               Negative     | Positive\")\n",
        "    print(f\"Actual Negative    {global_confusion_matrix[0, 0]}        | {global_confusion_matrix[0, 1]}\")\n",
        "    print(f\"Actual Positive    {global_confusion_matrix[1, 0]}        | {global_confusion_matrix[1, 1]}\")\n",
        "\n",
        "# Calculate and display confusion matrices for each model\n",
        "display_confusion_matrix(model_rnn, \"RNN\")\n",
        "display_confusion_matrix(model_lstm, \"LSTM\")\n",
        "display_confusion_matrix(model_gru, \"GRU\")\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "Z6192FJQR7uA",
        "outputId": "1f8d1507-5380-4e2c-b15c-b2a622c0bbc5"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "445/445 [==============================] - 1s 2ms/step\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 800x600 with 0 Axes>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 2 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjUAAAHHCAYAAABHp6kXAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAABuOElEQVR4nO3deVxN+f8H8NettC+itJBClEhZRmNnRGVn7KGyzNjGmnXGvjTGIHuGIZmsY1/GFtl3wtgjQovQoprWe35/+HW/jlt0daN7vZ7zOI9xP+dzPudzbrd69/58PudIBEEQQERERKTiNL50B4iIiIiUgUENERERqQUGNURERKQWGNQQERGRWmBQQ0RERGqBQQ0RERGpBQY1REREpBYY1BAREZFaYFBDREREaoFBjZqaPn06JBLJJx1rZ2eHdu3aKa0vjx8/hkQiQXBwsNLaLC7x8fHo2rUrypYtC4lEgsDAQKWfQyKRYPr06UpvV1X5+vrCzs5Oae3l5ORg/PjxsLGxgYaGBjp16qS0tosqODgYEokEjx8/lpU1b94czZs3/6z98PX1haGhoVLbVPZ1XLx4Edra2njy5InS2lS227dvQ0tLC//++++X7gr9PwY1KiQqKgrDhw9HtWrVoK+vD319fTg5OWHYsGG4cePGl+6e0sTHx8Pf3x+Ojo7Q19eHgYEB6tati9mzZyMpKalYzz169GgcOnQIkyZNwoYNG+Dp6Vms5/uc8gJdDQ0NPH36VG5/SkoK9PT0IJFIMHz4cIXbT09Px/Tp0xEeHq6E3n66tWvXYv78+ejatSvWr1+P0aNHf7Zz169fHxKJBCtXrlR628r+Y6Ok+/nnn9GrVy/Y2trKypo3bw6JRCLb9PT0UKtWLQQGBkIqlYqOz/tjSiKRYPv27XLt530/vHz5Ulbm6+sLiUSCWrVqIb8nCL3/veHk5IS2bdti6tSpyrhkUgKtL90BKpx9+/ahR48e0NLSgre3N1xcXKChoYG7d+9ix44dWLlyJaKiokQ/AFTRpUuX0KZNG6SmpqJPnz6oW7cuAODy5cv49ddfcfLkSRw+fLjYzn/s2DF07NgR/v7+xXaO//77D1paX+5bT0dHB5s2bcL48eNF5Tt27ChSu+np6ZgxYwYAKPQX++rVq+V+IRXFsWPHUL58eSxatEhpbRbGgwcPcOnSJdjZ2SE0NBRDhgz5rOdXJxERETh69CjOnj0rt69ChQoICAgAALx8+RIbN27E6NGjkZCQgDlz5uTb3syZM9GlS5dCZ69v3ryJHTt24Pvvv/9o3cGDB6NNmzZ4+PAhqlSpUqj2qfgwqFEBDx8+RM+ePWFra4uwsDBYWVmJ9s+bNw8rVqyAhoZqJ96SkpLQuXNnaGpq4tq1a3B0dBTtnzNnDlavXl2sfXjx4gVKly5drOfQ1dUt1vY/pk2bNvkGNRs3bkTbtm3z/au2OKSlpcHAwAClSpVSarvK/hpKpVJkZWV99Ov2119/oVy5cliwYAG6du2Kx48fK3VY7Wuybt06VKxYEd9++63cPhMTE/Tp00f2evDgwXB0dMTSpUsxc+ZMaGpqiuq7uroiIiICO3fuRJcuXT56bj09PdjY2BQ6EHJ3d4epqSnWr1+PmTNnFvIKqbio9m/Br8Rvv/2GtLQ0rFu3Ti6gAQAtLS2MGDECNjY2H2wnJycHs2bNQpUqVaCjowM7OztMnjwZmZmZ+dY/fPgwXF1doaurCycnJ7m/5F+/fg1/f384OzvD0NAQxsbG8PLywvXr1z/pOletWoXnz59j4cKFcgENAFhYWOCXX34Rla1YsQI1atSAjo4OrK2tMWzYMLkhqubNm6NmzZq4ffs2WrRoAX19fZQvXx6//fabrE7eXAdBELB8+XJZ2hooeH5SfvMjLl++DA8PD5iZmUFPTw+VKlVC//79RcflN6fm2rVr8PLygrGxMQwNDdGyZUucP38+3/OdOXMGY8aMgbm5OQwMDNC5c2ckJCQU+L6+r3fv3oiIiMDdu3dlZXFxcTh27Bh69+4tVz8rKwtTp05F3bp1YWJiAgMDAzRp0gTHjx+X1Xn8+DHMzc0BADNmzJC9f3nXmTeH4+HDh2jTpg2MjIzg7e0t2/fuL/9p06ZBQ0MDYWFhon788MMP0NbWLvDzlTfccPz4cdy6dUvWh7zhsLS0NIwdOxY2NjbQ0dGBg4MDfv/9d7lhhrwhhtDQUNln6+DBgx99Xzdu3IiuXbuiXbt2MDExwcaNGz96jLKdOnUK3bp1Q8WKFaGjowMbGxuMHj0a//33X771Hz16BA8PDxgYGMDa2hozZ86Uez+kUikCAwNRo0YN6OrqwsLCAj/++CMSExM/2p+lS5eiRo0a0NfXh6mpKerVq1eo92XXrl347rvvCpVZ0dXVxTfffIM3b97gxYsXcvt79uyJatWq5Xtt+dHQ0MAvv/yCGzduYOfOnR+tX6pUKTRv3hy7d+/+aF0qfgxqVMC+fftgb28PNze3IrUzcOBATJ06FXXq1MGiRYvQrFkzBAQEoGfPnnJ1Hzx4gB49esDLywsBAQHQ0tJCt27dcOTIEVmdR48eYdeuXWjXrh0WLlyIcePG4ebNm2jWrBliYmIU7t+ePXugp6eHrl27Fqr+9OnTMWzYMFhbW2PBggX4/vvvsWrVKrRu3RrZ2dmiuomJifD09ISLiwsWLFgAR0dHTJgwAf/88w8AoGnTptiwYQMAoFWrVtiwYYPsdWG9ePECrVu3xuPHjzFx4kQsXboU3t7ecsHJ+27duoUmTZrg+vXrGD9+PKZMmYKoqCg0b94cFy5ckKv/008/4fr165g2bRqGDBmCvXv3KjQHpmnTpqhQoYLol8uWLVtgaGiItm3bytVPSUnBmjVr0Lx5c8ybNw/Tp09HQkICPDw8EBERAQAwNzeXzSPp3Lmz7P179y/jnJwceHh4oFy5cvj9998LTO3/8ssvcHV1xYABA/DmzRsAwKFDh7B69WpMnToVLi4u+R5nbm6ODRs2wNHRERUqVJD1oXr16hAEAR06dMCiRYvg6emJhQsXwsHBAePGjcOYMWPk2jp27BhGjx6NHj16YPHixR/NuFy4cAGRkZHo1asXtLW10aVLF4SGhn7wmOKwbds2pKenY8iQIVi6dCk8PDywdOlS9OvXT65ubm4uPD09YWFhgd9++w1169bFtGnTMG3aNFG9H3/8EePGjUOjRo2wePFi+Pn5ITQ0FB4eHnLfZ+9avXo1RowYAScnJwQGBmLGjBlwdXXN9zP9rufPnyM6Ohp16tQp9HXnBbT5Zeg0NTXxyy+/4Pr164UKUoC3gX/VqlULHQjVrVsX//77L1JSUgrdZyomApVoycnJAgChU6dOcvsSExOFhIQE2Zaeni7bN23aNOHdL29ERIQAQBg4cKCoDX9/fwGAcOzYMVmZra2tAEDYvn27qB9WVlZC7dq1ZWUZGRlCbm6uqL2oqChBR0dHmDlzpqgMgLBu3boPXqupqang4uLywTp5Xrx4IWhrawutW7cW9WHZsmUCAGHt2rWysmbNmgkAhJCQEFlZZmamYGlpKXz//feidgEIw4YNE5W9/17mWbdunQBAiIqKEgRBEHbu3CkAEC5duvTBvgMQpk2bJnvdqVMnQVtbW3j48KGsLCYmRjAyMhKaNm0qdz53d3dBKpXKykePHi1oamoKSUlJHzxv3nUkJCQI/v7+gr29vWzfN998I/j5+eX7HuTk5AiZmZmithITEwULCwuhf//+srKEhAS5a8vj4+MjABAmTpyY7z5bW1tR2c2bNwVtbW1h4MCBQmJiolC+fHmhXr16QnZ29gevURDefr1r1KghKtu1a5cAQJg9e7aovGvXroJEIhEiIyNlZQAEDQ0N4datWx89V57hw4cLNjY2sq/L4cOHBQDCtWvXRPXe/8zk9bdZs2YfPYetra3Qtm3bD9Z592dAnoCAAEEikQhPnjyRleV9PX766SdZmVQqFdq2bStoa2sLCQkJgiAIwqlTpwQAQmhoqKjNgwcPypW/fx0dO3aU+zoUxtGjRwUAwt69e+X2NWvWTHB0dJT9zLt7964wbtw4AYDce5P3c2f+/PlCTk6OULVqVcHFxUX2NXr3++Hd98XAwEAQBEFYv369AEDYsWOHbH9+Px8EQRA2btwoABAuXLig8PWScjFTU8LlRf75Lb9s3rw5zM3NZdvy5csLbOfAgQMAIPdX6dixYwEA+/fvF5VbW1ujc+fOstfGxsbo168frl27hri4OABvJ5zmzePJzc3Fq1evYGhoCAcHB1y9elXRS0VKSgqMjIwKVffo0aPIysrCqFGjRHOJBg0aBGNjY7nrMTQ0FI3Da2tro379+nj06JHC/SxI3l+J+/bt++BfsO/Kzc3F4cOH0alTJ1SuXFlWbmVlhd69e+P06dNyf/398MMPorR8kyZNkJubq9DS1969eyMyMhKXLl2S/T+/oSfg7V+62traAN4ORbx+/Ro5OTmoV6+ewl/nwk6erVmzJmbMmIE1a9bAw8MDL1++xPr16z95gvWBAwegqamJESNGiMrHjh0LQRBkGbs8zZo1g5OTU6HazsnJwZYtW9CjRw/Z1+W7775DuXLlPnu2Rk9PT/bvtLQ0vHz5Eg0bNoQgCLh27Zpc/XczfHnDbllZWTh69CiAt5kfExMTtGrVCi9fvpRtdevWhaGhoWgI8n2lS5fGs2fPcOnSJYWu4dWrVwAAU1PTfPffvXtX9jPP0dER8+fPR4cOHT54y4h3szW7du0qVD+8vb0Lna3J6+u7K6noy2BQU8Ll/ZJPTU2V27dq1SocOXIEf/3110fbefLkCTQ0NGBvby8qt7S0ROnSpeV+Idrb28uNZ1erVg0AZHNIpFIpFi1ahKpVq0JHRwdmZmYwNzfHjRs3kJycXOhrzGNsbCwbbijM9QCAg4ODqFxbWxuVK1eWu54KFSrIXY+pqWmh5gUUVrNmzfD9999jxowZMDMzQ8eOHbFu3boC5ywBQEJCAtLT0+WuAwCqV68OqVQqt/y6YsWKotd5P1AVuZbatWvD0dERGzduRGhoKCwtLfHdd98VWH/9+vWoVasWdHV1UbZsWZibm2P//v0KfZ21tLRQoUKFQtcfN24cXFxccPHiRUybNq3QQUZ+njx5Amtra7mguXr16rL976pUqVKh2z58+DASEhJQv359REZGIjIyElFRUWjRogU2bdqk1JVdHxMdHQ1fX1+UKVMGhoaGMDc3R7NmzQBA7muloaEhCqQB+e/xBw8eIDk5GeXKlRP9AWVubo7U1NR857DkmTBhAgwNDVG/fn1UrVoVw4YNw5kzZwp9LQUFEnZ2djhy5AgOHTqEFStWoHz58khISPjoRG5vb2/Y29sXekgpLxCKiIj4aCCU196n3hvsY06ePIn27dvD2toaEomk0IFZnrx5ge9vBgYGxdLfL4mrn0o4ExMTWFlZ5Xtzp7w5Nu9OVP0YZX7TzZ07F1OmTEH//v0xa9YslClTBhoaGhg1atQn/SB3dHREREQEsrKyZJkBZXl/RUSewvxwK+g9y83Nlav3999/4/z589i7dy8OHTqE/v37Y8GCBTh//rzSbnZWlGt5V+/evbFy5UoYGRmhR48eBa6e++uvv+Dr64tOnTph3LhxKFeuHDQ1NREQEICHDx8W+nzvZvYK49GjR3jw4AGAt0tsP6d3Mx4fk5eN6d69e777T5w4gRYtWiilXx+Sm5uLVq1a4fXr15gwYQIcHR1hYGCA58+fw9fX95O+J6VS6QczTnmTw/NTvXp13Lt3D/v27cPBgwexfft2rFixAlOnTpUt/c9P2bJlARQcpBsYGMDd3V32ulGjRqhTpw4mT56MJUuWFNhuXpDi6+tb6Em93t7emDVrFmbOnPnBmzjm9dXMzKxQ7SoqLS0NLi4u6N+/f6FWcL3P398fgwcPFpW1bNkS33zzjbK6WGIwU6MC2rZti8jISFy8ePGT27C1tYVUKpX9ksgTHx+PpKQkufvbREZGyv2SvH//PgDIJk3+/fffaNGiBf7880/07NkTrVu3hru7+yffIK99+/b477//CrWkOK+/9+7dE5VnZWUp/X49eZmQ96+roOGeb7/9FnPmzMHly5cRGhqKW7duYfPmzfnWNTc3h76+vtx1AG/T7BoaGh9d1fapevfujdjYWNy/f7/AoSfg7de5cuXK2LFjB/r27QsPDw+4u7sjIyNDVE+ZAbNUKoWvry+MjY0xefJkbNq0qUj30bG1tUVMTIxcJjBvBdinfl7S0tKwe/du9OjRA9u2bZPbrKysPtsQ1M2bN3H//n0sWLAAEyZMQMeOHeHu7g5ra+t860ulUrnh1/e/x6tUqYJXr16hUaNGcHd3l9sKmrSdx8DAAD169MC6desQHR2Ntm3bYs6cOXKfnXflrXyMiooq1HXXqlULffr0wapVqxAdHf3Bun369IG9vT1mzJihcLbmQ4FQVFQUNDQ0ZJkuZfPy8sLs2bNFUwLelZmZCX9/f5QvXx4GBgZwc3MT3QTT0NAQlpaWsi0+Ph63b9/GgAEDiqW/XxKDGhUwfvx46Ovro3///oiPj5fbX5hvzjZt2gCA3G3/Fy5cCAByq15iYmJEKwVSUlIQEhICV1dXWFpaAnj7Df/+ubdt24bnz59//KLyMXjwYFhZWWHs2LGyH67vevHiBWbPng3g7b0htLW1sWTJElEf/vzzTyQnJ+e7iudT5d1Q6+TJk7KytLQ0rF+/XlQvMTFR7v1wdXUFgAKHoDQ1NdG6dWvs3r1blHGLj4/Hxo0b0bhxYxgbGyvhKuRVqVIFgYGBCAgIQP369Qusl5cZevfaLly4gHPnzonq6evrA5AP/j7FwoULcfbsWfzxxx+YNWsWGjZsiCFDhnzynIU2bdogNzcXy5YtE5UvWrQIEokEXl5en9Tuzp07kZaWhmHDhqFr165yW7t27bB9+/YPDkEqS35fJ0EQsHjx4gKPeff9EAQBy5YtQ6lSpdCyZUsAb7NPubm5mDVrltyxOTk5H/xa582NyaOtrQ0nJycIgvDBOWfly5eHjY0NLl++XGCd940fPx7Z2dmyn2cFeTdI2bNnT6HafjcQKsiVK1dQo0YNmJiYFLrPyjR8+HCcO3cOmzdvxo0bN9CtWzd4enrK/RGbZ82aNahWrRqaNGnymXta/Dj8pAKqVq2KjRs3olevXnBwcJDdUVgQBERFRWHjxo3Q0ND44HwFFxcX+Pj44I8//kBSUhKaNWuGixcvYv369ejUqZNcerxatWoYMGAALl26BAsLC6xduxbx8fFYt26drE67du0wc+ZM+Pn5oWHDhrh58yZCQ0PlxukLy9TUFDt37kSbNm3g6uoquqPw1atXsWnTJjRo0ADA2wzHpEmTMGPGDHh6eqJDhw64d+8eVqxYgW+++UY0KbioWrdujYoVK2LAgAEYN24cNDU1sXbtWpibm4v+Mly/fj1WrFiBzp07o0qVKnjz5g1Wr14NY2NjWVCZn9mzZ+PIkSNo3Lgxhg4dCi0tLaxatQqZmZmie+kUh5EjR360Trt27bBjxw507twZbdu2RVRUFIKCguDk5CSa66WnpwcnJyds2bIF1apVQ5kyZVCzZk3UrFlToT7duXMHU6ZMga+vL9q3bw/g7T16XF1dMXToUGzdulWxi8TbLGCLFi3w888/4/Hjx3BxccHhw4exe/dujBo16pPvBBsaGoqyZcuiYcOG+e7v0KEDVq9ejf3793/SsMH7IiMjZYH9u2rXro3WrVujSpUq8Pf3x/Pnz2FsbIzt27cXOIyjq6uLgwcPwsfHB25ubvjnn3+wf/9+TJ48WTas1KxZM/z4448ICAhAREQEWrdujVKlSuHBgwfYtm0bFi9eXOAtGFq3bg1LS0s0atQIFhYWuHPnDpYtW4a2bdt+dEFAx44dsXPnTgiCUKgMoJOTE9q0aYM1a9ZgypQpsiGs/OQNKeXdjuBjNDU18fPPP8PPzy/f/dnZ2Thx4gSGDh1aqPaULTo6WpYJy8vK+fv74+DBg1i3bh3mzp0rqp+RkYHQ0FBMnDjxS3S3+H3WtVZUJJGRkcKQIUMEe3t7QVdXV9DT0xMcHR2FwYMHCxEREaK6+S1Dzs7OFmbMmCFUqlRJKFWqlGBjYyNMmjRJyMjIENXLWzp66NAhoVatWoKOjo7g6OgobNu2TVQvIyNDGDt2rGBlZSXo6ekJjRo1Es6dOye3tLOwS7rzxMTECKNHjxaqVasm6OrqCvr6+kLdunWFOXPmCMnJyaK6y5YtExwdHYVSpUoJFhYWwpAhQ4TExERRnfyW+ApC/kuJUcCSzStXrghubm6Ctra2ULFiRWHhwoVyy3OvXr0q9OrVS6hYsaKgo6MjlCtXTmjXrp1w+fJluXO8v+z56tWrgoeHh2BoaCjo6+sLLVq0EM6ePSuqk3e+95eMHz9+XAAgHD9+XK7f78pvCWt+3n8PpFKpMHfuXMHW1lbQ0dERateuLezbty/f9+/s2bNC3bp1BW1tbdF1vrtU9n3vtpOTkyN88803QoUKFeSWqC9evFgAIGzZsuWD/S/o6/3mzRth9OjRgrW1tVCqVCmhatWqwvz580XL4/O7/oLEx8cLWlpaQt++fQusk56eLujr6wudO3cWBKHoS7oB5LsNGDBAEARBuH37tuDu7i4YGhoKZmZmwqBBg4Tr16/Lff/lfT0ePnwotG7dWtDX1xcsLCyEadOmyd2mQRAE4Y8//hDq1q0r6OnpCUZGRoKzs7Mwfvx4ISYmpsDrWLVqldC0aVOhbNmygo6OjlClShVh3Lhxct/D+bl69aoAQDh16pSovKCvrSAIQnh4uOgz9+6S7vflfR3e/34o6HOanZ0tVKlSJd/Pxj///CMAEB48ePDR61IGAMLOnTtlr/ft2ycAEAwMDESblpaW0L17d7njN27cKGhpaQlxcXGfpb+fm0QQFJxdSEREVMxatmwJa2trhW+C+bl16tQJEomk0Df2K6q8c+VNXN6yZQu8vb1x69YtuUUEeXNp3tWyZUsYGxt/tv5+bhx+IiKiEmfu3Llo0qQJZs+eXWIf1Hvnzh3s27ev0ENZxaF27drIzc3FixcvPjpHJioqCsePHy/0fCJVxKCGiIhKHDc3N2RlZX3pbnxQ9erVkZOTU+znSU1NRWRkpOx1VFQUIiIiUKZMGVSrVg3e3t7o168fFixYgNq1ayMhIQFhYWGoVauWaNHE2rVrYWVl9ckT41UBh5+IiIhKsPDw8HzvdeTj44Pg4GBkZ2dj9uzZCAkJwfPnz2FmZoZvv/0WM2bMgLOzM4C3S/htbW3Rr18/zJkz53NfwmfDoIaIiIjUAu9TQ0RERGqBQQ0RERGpBU4ULoGkUiliYmJgZGRUbA9IIyKi4iEIAt68eQNra2uFnnemqIyMDKVNptbW1v7oQ0FVAYOaEigmJqbYnvdDRESfx9OnTxV6Mr0iMjIyoGdUFshJV0p7lpaWiIqKUvnAhkFNCZR3C3FtJx9INJX7tGqikiI6/Pcv3QWiYvEmJQX2lWw++jiIosjKygJy0qHj5AMU9fdEbhbibq9HVlYWgxpSvrwhJ4mmNoMaUlvF9aBOopLis0wf0NIt8u8JQaI+02sZ1BAREakqCYCiBk9qNHWTQQ0REZGqkmi83YrahppQnyshIiKirxozNURERKpKIlHC8JP6jD8xqCEiIlJVHH4SUZ8rISIioq8aMzVERESqisNPIgxqiIiIVJYShp/UaNBGfa6EiIiIilVAQAC++eYbGBkZoVy5cujUqRPu3bv30eO2bdsGR0dH6OrqwtnZGQcOHBDtFwQBU6dOhZWVFfT09ODu7o4HDx4o3D8GNURERKoqb/ipqFshnThxAsOGDcP58+dx5MgRZGdno3Xr1khLSyvwmLNnz6JXr14YMGAArl27hk6dOqFTp074999/ZXV+++03LFmyBEFBQbhw4QIMDAzg4eGBjIwMxd4OQRAEhY6gYpeSkgITExPoOA/iYxJIbSVeWvalu0BULFJSUmBR1gTJycnF9jgQ2e+Jb8ZAoqVTpLaEnExkXlr4Sf1NSEhAuXLlcOLECTRt2jTfOj169EBaWhr27dsnK/v222/h6uqKoKAgCIIAa2trjB07Fv7+/gCA5ORkWFhYIDg4GD179ix0f5ipISIiIqSkpIi2zMzMjx6TnJwMAChTpkyBdc6dOwd3d3dRmYeHB86dOwcAiIqKQlxcnKiOiYkJ3NzcZHUKi0ENERGRqlLi8JONjQ1MTExkW0BAwAdPLZVKMWrUKDRq1Ag1a9YssF5cXBwsLCxEZRYWFoiLi5PtzysrqE5hcfUTERGRqlLizfeePn0qGn7S0fnwsNawYcPw77//4vTp00U7vxIxU0NERKSqlJipMTY2Fm0fCmqGDx+Offv24fjx46hQocIHu2hpaYn4+HhRWXx8PCwtLWX788oKqlNYDGqIiIioUARBwPDhw7Fz504cO3YMlSpV+ugxDRo0QFhYmKjsyJEjaNCgAQCgUqVKsLS0FNVJSUnBhQsXZHUKi8NPREREquozP/tp2LBh2LhxI3bv3g0jIyPZnBcTExPo6ekBAPr164fy5cvL5uSMHDkSzZo1w4IFC9C2bVts3rwZly9fxh9//PH29BIJRo0ahdmzZ6Nq1aqoVKkSpkyZAmtra3Tq1EmhS2FQQ0REpKokEiUENYW/T83KlSsBAM2bNxeVr1u3Dr6+vgCA6OhoaGj8r08NGzbExo0b8csvv2Dy5MmoWrUqdu3aJZpcPH78eKSlpeGHH35AUlISGjdujIMHD0JXV1exS+F9akoe3qeGvga8Tw2pq896n5qGkyDRUuwX//uEnAxkng0o1v5+LszUEBERqSoNydutqG2oCQY1REREquozz6kp6dTnSoiIiOirxkwNERGRqlLwgZQFtqEmGNQQERGpKg4/iajPlRAREdFXjZkaIiIiVcXhJxEGNURERKqKw08iDGqIiIhUFTM1IuoTnhEREdFXjZkaIiIiVcXhJxEGNURERKqKw08i6hOeERER0VeNmRoiIiKVpYThJzXKbzCoISIiUlUcfhJRn/CMiIiIvmrM1BAREakqiUQJq5/UJ1PDoIaIiEhVcUm3iPpcCREREX3VmKkhIiJSVZwoLMKghoiISFVx+EmEQQ0REZGqYqZGRH3CMyIiIvqqMVNDRESkqjj8JMKghoiISFVx+ElEfcIzIiIi+qoxU0NERKSiJBIJJMzUyDCoISIiUlEMasQ4/ERERERqgZkaIiIiVSX5/62obagJBjVEREQqisNPYhx+IiIiIrXATA0REZGKYqZGjJkaIiIiFZUX1BR1U8TJkyfRvn17WFtbQyKRYNeuXR+s7+vrm+85a9SoIaszffp0uf2Ojo4Kvx8MaoiIiFTUlwhq0tLS4OLiguXLlxeq/uLFixEbGyvbnj59ijJlyqBbt26iejVq1BDVO336tEL9Ajj8RERERArw8vKCl5dXoeubmJjAxMRE9nrXrl1ITEyEn5+fqJ6WlhYsLS2L1DdmaoiIiFSVREnbZ/Tnn3/C3d0dtra2ovIHDx7A2toalStXhre3N6KjoxVum5kaIiIiFaXMicIpKSmiYh0dHejo6BSt7ffExMTgn3/+wcaNG0Xlbm5uCA4OhoODA2JjYzFjxgw0adIE//77L4yMjArdPjM1REREBBsbG9lQkYmJCQICApR+jvXr16N06dLo1KmTqNzLywvdunVDrVq14OHhgQMHDiApKQlbt25VqH1maoiIiFSURAIlZGre/u/p06cwNjaWFSs7SyMIAtauXYu+fftCW1v7g3VLly6NatWqITIyUqFzMFNDRESkoiRQwuqn/49qjI2NRZuyg5oTJ04gMjISAwYM+Gjd1NRUPHz4EFZWVgqdg0ENERERFVpqaioiIiIQEREBAIiKikJERIRsYu+kSZPQr18/ueP+/PNPuLm5oWbNmnL7/P39ceLECTx+/Bhnz55F586doampiV69einUNw4/ERERqagvcUfhy5cvo0WLFrLXY8aMAQD4+PggODgYsbGxciuXkpOTsX37dixevDjfNp89e4ZevXrh1atXMDc3R+PGjXH+/HmYm5sr1DcGNURERKrqCzylu3nz5hAEocD9wcHBcmUmJiZIT08v8JjNmzcr1okCcPiJiIiI1AIzNURERKpKCcNPgho90JJBDRERkYpSxpyaIs/JKUEY1BAREakoBjVinFNDREREaoGZGiIiIlX1BVY/lWQMaoiIiFQUh5/EOPxEREREaoGZGiIiIhXFTI0YgxoiIiIVxaBGjMNPREREpBaYqSEiIlJRzNSIMaghIiJSVVzSLcLhJyIiIlILzNQQERGpKA4/iTGoISIiUlEMasQY1BAREakoBjVinFNDREREaoGZGiIiIlXF1U8iDGqIiIhUFIefxDj8RERERGqBmZqPsLOzw6hRozBq1Kgv3ZWvxmjf1mjXwgVVbS2QkZmNizceYfqy3Yh88qLAYxwrW2LSj+3g6miDitZlMWnh3wjaFF7sfe3YsjYmD26LilZl8ehpAqYv3YUjZ2/L9ideWpbvcVMX78TSv8KKvX+kHhYFH8bM5XswuGdzBIztKtonCAK6jVyJsHO38df8QWjb3AUAsHHveQyb+Ve+7d0/FADzMkYAgNVbT2DNtpOIjn2NChamGNvfAz3busnqtvsxEGeuRsq10apRDWwNHKKsS6RPxEyN2BcNanx9fbF+/XoEBARg4sSJsvJdu3ahc+fOEAThs/UlODgYo0aNQlJSkqj80qVLMDAw+Gz9IKBhHXus2XYS124/gZamJqYMbY8dS4fj2+6zkZ6Rle8xerraePL8JXYfvYY5Y7oopR+N6lTFiml94NJxWr7769eqhDWzfTFz+R4cOv0vunrWw1+//4DmfefhzsNYAICD5yTRMe4Na2DpL72x53iEUvpI6u/qrScI3nkGNaqWz3f/yk3Hkd/vpM6t6qBlAydR2bAZG5CRlS0LaP78+xRmrdiLwMm9UMfJFlduP8aoOZtgYqQPr6bOAIANvw1CVnaurI3XyWlo4h2ATi1rK+kKqSgkUEJQo0aTar748JOuri7mzZuHxMTEL92VfJmbm0NfX/9Ld+Or0m3ECmzadwF3H8Xh3wfPMXTGX7CxKgPX6jYFHnPtdjSmLtmFHUeuICsrJ986EokEo31bI2LXdMScWohToRPR4TvXT+7njz2bI+zcHSz9Kwz3H8djbtB+XL/7FIO6NZPVefHqjWhr09QZp648wJPnrz75vPT1SE3PxA9Tg7F4ci+UNtKT23/z3jMsDz2GZVP6yO3T09WGhZmxbNPUlODk5fvo07GhrM6WAxfh07kRurSuC7sKZvi+dT34dGqExSFHZHVMTQxE7YRfuAt9XW10dGdQQyXPFw9q3N3dYWlpiYCAgALrnD59Gk2aNIGenh5sbGwwYsQIpKWlyfbHxsaibdu20NPTQ6VKlbBx40bY2dkhMDBQVmfhwoVwdnaGgYEBbGxsMHToUKSmpgIAwsPD4efnh+TkZFkqb/r06QAgaqd3797o0aOHqG/Z2dkwMzNDSEgIAEAqlSIgIACVKlWCnp4eXFxc8Pfffyvhnfp6GRvqAgASU9KL1M4Y39bo0aY+xvy6BQ16zsGKTcexaqYPGtax/6T26jtXQvilu6KyY+fv4Btnu3zrm5cxQuvGNfHX7nOfdD76+oz7bQtaN6qJ5m6OcvvSM7IwaEow5o/vDgsz44+2tXn/RejpaqPjO4F8VnYOdLVLierp6pbC1VtPkJ2Ti/xs2HMWXVrVgYGejmIXQ8Ui73dWUTd18cWDGk1NTcydOxdLly7Fs2fP5PY/fPgQnp6e+P7773Hjxg1s2bIFp0+fxvDhw2V1+vXrh5iYGISHh2P79u34448/8OKFeP6FhoYGlixZglu3bmH9+vU4duwYxo8fDwBo2LAhAgMDYWxsjNjYWMTGxsLf31+uL97e3ti7d68sGAKAQ4cOIT09HZ07dwYABAQEICQkBEFBQbh16xZGjx6NPn364MSJE0p5v742EokEAWO64nzEQ9mQzqfQLqWF0X6t8dOsUBw7fwdPnr/Cpn0XsPWfS/Dr3PiT2ixX1hgJr96IyhJev0G5svn/gunV1g2paRnYy6EnKoTthy/j+t2nmDqsQ777Jy/cjvq1KqFNs1qFau+vPefQ1aMe9HS1ZWXffVsdG3afRcSdaAiCgGu3n2DDrrPIzsnFq6RUuTau3HqMOw9j0bdTQ7l99IVIlLSpiRIxUbhz585wdXXFtGnT8Oeff4r2BQQEwNvbWzZRt2rVqliyZAmaNWuGlStX4vHjxzh69CguXbqEevXqAQDWrFmDqlWritp5d6KvnZ0dZs+ejcGDB2PFihXQ1taGiYkJJBIJLC0tC+ynh4cHDAwMsHPnTvTt2xcAsHHjRnTo0AFGRkbIzMzE3LlzcfToUTRo0AAAULlyZZw+fRqrVq1Cs2bN8m03MzMTmZmZstcpKSmFe+O+Ar+P747qVazgNWhRkdqpbGMGAz0d7Fg2XFSuXUoTN+79L5h+emKB7N+aGhLoaGuJyrb9cwljft38SX3w7vAtth28jMwChseI8jyLS8SkBduxY9lw6OqUktt/4MQNnLp8Hyf+mpjP0fIu3niEe1FxCJrRT1Q+boAnXrxKQSu/3yEAKFfGCD3buWFJyFFo5PPX+4bd5+Bkb426New+5bKIil2JCGoAYN68efjuu+/kMiTXr1/HjRs3EBoaKisTBAFSqRRRUVG4f/8+tLS0UKdOHdl+e3t7mJqaito5evQoAgICcPfuXaSkpCAnJwcZGRlIT08v9JwZLS0tdO/eHaGhoejbty/S0tKwe/dubN789pdcZGQk0tPT0apVK9FxWVlZqF274PHngIAAzJgxo1B9+Jr8Nq4bPJrURJsfAhHzIqlIbeWlynuMXonY99rKyv5fkNHU+3/DoPVq2mHa8I5oP3ixrOxNWobs3y9epcC8rJGoLfMyRnjxSj4obeBaBdXsLDFg8roiXQd9Ha7fjUbC6zdo3neerCw3V4qz1x5i9baT6P99Y0Q9ewm778aJjus3YQ0auFbBvlWjROUbdp+Dc7UKcK1eUVSup6uNZVP7YNHkXnjxKgWWZiYI3nkGRga6MDM1FNVN+y8TOw5fweQf2yr3YqlIuPpJrMQENU2bNoWHhwcmTZoEX19fWXlqaip+/PFHjBgxQu6YihUr4v79+x9t+/Hjx2jXrh2GDBmCOXPmoEyZMjh9+jQGDBiArKwshSYCe3t7o1mzZnjx4gWOHDkCPT09eHp6yvoKAPv370f58uKVCjo6BY8/T5o0CWPGjJG9TklJgY1NwZNivwa/jeuGts1d0H7wYkTHFH1S7b2oOGRkZsPGwhRn81memifq2UvZv63LmSI3Vyoqe9fFm1Fo9o2DaOl4CzdHXLr5WK5un44NcO12NP598PyTr4G+Hk2/ccCZTZNFZcNn/oWqdhYY2a8VypY2hO97w6aNes3F3NHfw7NJTVF5anomdh29iikFDGMBQCktTZS3ePuH4I7DV9C6cQ1oaIhnJ+w+eg1Z2Tno7vVNUS6NlIxBjViJCWoA4Ndff4WrqyscHBxkZXXq1MHt27dhb5//ZE4HBwfk5OTg2rVrqFu3LoC3GZN3V1NduXIFUqkUCxYskH2jbt26VdSOtrY2cnPznxj3roYNG8LGxgZbtmzBP//8g27duqFUqbfpYScnJ+jo6CA6OrrAoab86OjofDDo+dr8PqE7unrUQ2//P5CanoFy/58NSUnNQEZmNgBg5fS+iE1IxszlewC8/aHsUPnt0GGpUlqwNi+NmtXKIy09E1HPXiI1PRPL/grDnDHfQ6KhgfMRD2FsqAs3lyp4k5aBzfsvKNzPVZvDsW/VKAzz/g6HT99Cl9Z14Vq9IkbN3SSqZ2Sgi44ta2NK4M6ivC30FTEy0IWTvbWoTF9PG2VMDGTl+U0OrmBpCtvyZqKynUeuICdXih75BCORT+Jx5dYT1Ktph6Q36Vgeegx3HsVg5fS+cnU37DmHNs1qoUxpQ7l99OVIJMh3Sb+ibaiLEhXUODs7w9vbG0uWLJGVTZgwAd9++y2GDx+OgQMHwsDAALdv38aRI0ewbNkyODo6wt3dHT/88ANWrlyJUqVKYezYsdDT05NFn/b29sjOzsbSpUvRvn17nDlzBkFBQaJz29nZITU1FWFhYXBxcYG+vn6BGZzevXsjKCgI9+/fx/Hjx2XlRkZG8Pf3x+jRoyGVStG4cWMkJyfjzJkzMDY2ho+PTzG8a+pnQNemAID976XQh87YgE373gYfFSzLQPrOfYwszU1wKvR/94T5qa87furrjtNXHsiGj+YE7cPLpFSM9m0Fu/K9kPzmP1y/9xSL1h36pH5evBGFQb8E4+ch7TBlaHs8epqAPv5/yE1o7tK6LiQSCbYfuvxJ5yEqig27z6FdcxeYGMn/PMuVClgeegyRT+KhpaWJJvWq4dCasahoXVZU78HjeJyPeIgdy4Z9rm4TfRKJ8DnvcPceX19fJCUlYdeuXbKyx48fw8HBAVlZWbKb7126dAk///wzzp07B0EQUKVKFfTo0QOTJ79Nz8bGxmLAgAE4duyYbHn4qFGjMHPmTPz4448AgEWLFmH+/PlISkpC06ZN4e3tjX79+iExMRGlS5cGAAwZMgTbtm3Dq1evMG3aNEyfPj3fOwrfuXMHTk5OsLW1RVRUlCh1JwgClixZgpUrV+LRo0coXbo06tSpg8mTJ6Np06aFel9SUlJgYmICHedBkGhqf/wAIhVU0N2WiVRdSkoKLMqaIDk5GcbGH19u/6nnMDExQeWf/oaGTtFuECvNTMOjpV2Ltb+fyxcNaorLs2fPYGNjg6NHj6Jly5ZfujsKY1BDXwMGNaSuPmtQM+JvaBYxqMnNTMOjJeoR1JSo4adPdezYMaSmpsLZ2RmxsbEYP3487OzsCp0ZISIiItX3xW++pwzZ2dmYPHkyatSogc6dO8Pc3Bzh4eGyCbxERETq6EvcUfjkyZNo3749rK2tIZFIRFNI8hMeHp7vOePi4kT1li9fDjs7O+jq6sLNzQ0XL15U9O1Qj0yNh4cHPDw8vnQ3iIiIPqsvsfopLS0NLi4u6N+/P7p0KfwDhO/duyca3ipXrpzs31u2bMGYMWMQFBQENzc3BAYGwsPDA/fu3RPV+xi1CGqIiIjo8/Dy8oKXl5fCx5UrV062MOd9CxcuxKBBg+Dn5wcACAoKwv79+7F27VpMnFi4O2cDajL8RERE9DXS0JAoZQPeTj5+d3v38T3K4OrqCisrK7Rq1QpnzpyRlWdlZeHKlStwd3d/57o04O7ujnPnFHsAMIMaIiIiFZU3/FTUDQBsbGxgYmIi2wICAj588kKysrJCUFAQtm/fju3bt8PGxgbNmzfH1atXAQAvX75Ebm4uLCwsRMdZWFjIzbv5GA4/EREREZ4+fSqa86KsO907ODiInhTQsGFDPHz4EIsWLcKGDRuUco48DGqIiIhUlDKf/WRsbPzZ7lNTv359nD59GgBgZmYGTU1NxMfHi+rEx8fD0tJSoXY5/ERERKSilDn89DlFRETAysoKwNtnL9atWxdhYWGy/VKpFGFhYWjQoIFC7TJTQ0REpKK+xFO6U1NTERkZKXsdFRWFiIgIlClTBhUrVsSkSZPw/PlzhISEAAACAwNRqVIl1KhRAxkZGVizZg2OHTuGw4cPy9oYM2YMfHx8UK9ePdSvXx+BgYFIS0uTrYYqLAY1REREVGiXL19GixYtZK/HjBkDAPDx8UFwcDBiY2MRHR0t25+VlYWxY8fi+fPn0NfXR61atXD06FFRGz169EBCQgKmTp2KuLg4uLq64uDBg3KThz9GLZ/9pOr47Cf6GvDZT6SuPuezn2pM2K2UZz/dmteRz34iIiKiL+dL3FG4JONEYSIiIlILzNQQERGpKAmUMFEY6pOqYVBDRESkojj8JMbhJyIiIlILzNQQERGpqC9xn5qSjEENERGRiuLwkxiHn4iIiEgtMFNDRESkojj8JMaghoiISEVx+EmMQQ0REZGKYqZGjHNqiIiISC0wU0NERKSqlDD8pEY3FGZQQ0REpKo4/CTG4SciIiJSC8zUEBERqSiufhJjUENERKSiOPwkxuEnIiIiUgvM1BAREakoDj+JMaghIiJSURx+EuPwExEREakFZmqIiIhUFDM1YgxqiIiIVBTn1IgxqCEiIlJRzNSIcU4NERERqQVmaoiIiFQUh5/EGNQQERGpKA4/iXH4iYiIiNQCMzVEREQqSgIlDD8ppSclA4MaIiIiFaUhkUCjiFFNUY8vSTj8RERERGqBmRoiIiIVxdVPYgxqiIiIVBRXP4lx+ImIiEhFaUiUsyni5MmTaN++PaytrSGRSLBr164P1t+xYwdatWoFc3NzGBsbo0GDBjh06JCozvTp02UBWt7m6Oio4LvBoIaIiIgUkJaWBhcXFyxfvrxQ9U+ePIlWrVrhwIEDuHLlClq0aIH27dvj2rVrono1atRAbGysbDt9+rTCfePwExERkaqSKGH4SMHDvby84OXlVej6gYGBotdz587F7t27sXfvXtSuXVtWrqWlBUtLS8U68x5maoiIiFRU3kThom4AkJKSItoyMzOLpc9SqRRv3rxBmTJlROUPHjyAtbU1KleuDG9vb0RHRyvcNoMaIiIigo2NDUxMTGRbQEBAsZzn999/R2pqKrp37y4rc3NzQ3BwMA4ePIiVK1ciKioKTZo0wZs3bxRqm8NPREREKkry//8VtQ0AePr0KYyNjWXlOjo6RWo3Pxs3bsSMGTOwe/dulCtXTlb+7nBWrVq14ObmBltbW2zduhUDBgwodPsMaoiIiFTUp6xeyq8NADA2NhYFNcq2efNmDBw4ENu2bYO7u/sH65YuXRrVqlVDZGSkQufg8BMREREVq02bNsHPzw+bNm1C27ZtP1o/NTUVDx8+hJWVlULnYaaGiIhIRX2Jm++lpqaKMihRUVGIiIhAmTJlULFiRUyaNAnPnz9HSEgIgLdDTj4+Pli8eDHc3NwQFxcHANDT04OJiQkAwN/fH+3bt4etrS1iYmIwbdo0aGpqolevXgr1rVBBzZ49ewrdYIcOHRTqABEREX2aL/GYhMuXL6NFixay12PGjAEA+Pj4IDg4GLGxsaKVS3/88QdycnIwbNgwDBs2TFaeVx8Anj17hl69euHVq1cwNzdH48aNcf78eZibmyvUt0IFNZ06dSpUYxKJBLm5uQp1gIiIiFRH8+bNIQhCgfvzApU84eHhH21z8+bNRezVW4UKaqRSqVJORkRERMqjIZFAo4ipmqIeX5IUaU5NRkYGdHV1ldUXIiIiUgCf0i2m8Oqn3NxczJo1C+XLl4ehoSEePXoEAJgyZQr+/PNPpXeQiIiI8vf+QyA/dVMXCgc1c+bMQXBwMH777Tdoa2vLymvWrIk1a9YotXNEREREhaVwUBMSEoI//vgD3t7e0NTUlJW7uLjg7t27Su0cERERFUyZz35SBwrPqXn+/Dns7e3lyqVSKbKzs5XSKSIiIvo4ThQWUzhT4+TkhFOnTsmV//3336JHiBMRERF9TgpnaqZOnQofHx88f/4cUqkUO3bswL179xASEoJ9+/YVRx+JiIgoH5L/34rahrpQOFPTsWNH7N27F0ePHoWBgQGmTp2KO3fuYO/evWjVqlVx9JGIiIjywdVPYp90n5omTZrgyJEjyu4LERER0Sf75JvvXb58GXfu3AHwdp5N3bp1ldYpIiIi+jgNydutqG2oC4WDmryHTp05cwalS5cGACQlJaFhw4bYvHkzKlSooOw+EhERUT6+xFO6SzKF59QMHDgQ2dnZuHPnDl6/fo3Xr1/jzp07kEqlGDhwYHH0kYiIiOijFM7UnDhxAmfPnoWDg4OszMHBAUuXLkWTJk2U2jkiIiL6MDVKtBSZwkGNjY1NvjfZy83NhbW1tVI6RURERB/H4ScxhYef5s+fj59++gmXL1+WlV2+fBkjR47E77//rtTOERERUcHyJgoXdVMXhcrUmJqaiiK5tLQ0uLm5QUvr7eE5OTnQ0tJC//790alTp2LpKBEREdGHFCqoCQwMLOZuEBERkaI4/CRWqKDGx8enuPtBRERECuJjEsQ++eZ7AJCRkYGsrCxRmbGxcZE6RERERPQpFA5q0tLSMGHCBGzduhWvXr2S25+bm6uUjhEREdGHaUgk0Cji8FFRjy9JFF79NH78eBw7dgwrV66Ejo4O1qxZgxkzZsDa2hohISHF0UciIiLKh0SinE1dKJyp2bt3L0JCQtC8eXP4+fmhSZMmsLe3h62tLUJDQ+Ht7V0c/SQiIiL6IIUzNa9fv0blypUBvJ0/8/r1awBA48aNcfLkSeX2joiIiAqUt/qpqJu6UDioqVy5MqKiogAAjo6O2Lp1K4C3GZy8B1wSERFR8ePwk5jCQY2fnx+uX78OAJg4cSKWL18OXV1djB49GuPGjVN6B4mIiIgKQ+E5NaNHj5b9293dHXfv3sWVK1dgb2+PWrVqKbVzREREVDCufhIr0n1qAMDW1ha2trbK6AsREREpQBnDR2oU0xQuqFmyZEmhGxwxYsQnd4aIiIgKj49JECtUULNo0aJCNSaRSBjUEBER0RdRqKAmb7UTfV5Rx+bzsRNERFQgDXzCip982lAXRZ5TQ0RERF8Gh5/E1ClAIyIioq8YMzVEREQqSiIBNLj6SYaZGiIiIhWlIVHOpoiTJ0+iffv2sLa2hkQiwa5duz56THh4OOrUqQMdHR3Y29sjODhYrs7y5cthZ2cHXV1duLm54eLFi4p1DAxqiIiISAFpaWlwcXHB8uXLC1U/KioKbdu2RYsWLRAREYFRo0Zh4MCBOHTokKzOli1bMGbMGEybNg1Xr16Fi4sLPDw88OLFC4X69klBzalTp9CnTx80aNAAz58/BwBs2LABp0+f/pTmiIiI6BN8iQdaenl5Yfbs2ejcuXOh6gcFBaFSpUpYsGABqlevjuHDh6Nr166i28UsXLgQgwYNgp+fH5ycnBAUFAR9fX2sXbtWob4pHNRs374dHh4e0NPTw7Vr15CZmQkASE5Oxty5cxVtjoiIiD6RMoefUlJSRFve7/eiOnfuHNzd3UVlHh4eOHfuHAAgKysLV65cEdXR0NCAu7u7rE5hKRzUzJ49G0FBQVi9ejVKlSolK2/UqBGuXr2qaHNERERUAtjY2MDExES2BQQEKKXduLg4WFhYiMosLCyQkpKC//77Dy9fvkRubm6+deLi4hQ6l8Krn+7du4emTZvKlZuYmCApKUnR5oiIiOgTKfPZT0+fPhXd8FVHR6doDX8BCgc1lpaWiIyMhJ2dnaj89OnTqFy5srL6RURERB+hzKd0GxsbF8td7C0tLREfHy8qi4+Ph7GxMfT09KCpqQlNTc1861haWip0LoWHnwYNGoSRI0fiwoULkEgkiImJQWhoKPz9/TFkyBBFmyMiIqJPpKGkrTg1aNAAYWFhorIjR46gQYMGAABtbW3UrVtXVEcqlSIsLExWp7AUztRMnDgRUqkULVu2RHp6Opo2bQodHR34+/vjp59+UrQ5IiIiUiGpqamIjIyUvY6KikJERATKlCmDihUrYtKkSXj+/DlCQkIAAIMHD8ayZcswfvx49O/fH8eOHcPWrVuxf/9+WRtjxoyBj48P6tWrh/r16yMwMBBpaWnw8/NTqG8KBzUSiQQ///wzxo0bh8jISKSmpsLJyQmGhoaKNkVERERFoMw5NYV1+fJltGjRQvZ6zJgxAAAfHx8EBwcjNjYW0dHRsv2VKlXC/v37MXr0aCxevBgVKlTAmjVr4OHhIavTo0cPJCQkYOrUqYiLi4OrqysOHjwoN3n4o9ciCIKg2OVQcUtJSYGJiQliEpL4lG5SW5pFvbc7UQmVkpICi7ImSE5OLraf4Xm/J8b9fRU6BkVLKmSmpWJ+1zrF2t/PReFMTYsWLT54o55jx44VqUNEREREn0LhoMbV1VX0Ojs7GxEREfj333/h4+OjrH4RERHRR3yJ4aeSTOGg5t3bGr9r+vTpSE1NLXKHiIiIqHA+5YGU+bWhLpS2kqtPnz4KP6OBiIiISFkUztQU5Ny5c9DV1VVWc0RERPQREgmKfPO9r3r4qUuXLqLXgiAgNjYWly9fxpQpU5TWMSIiIvowzqkRUzioMTExEb3W0NCAg4MDZs6cidatWyutY0RERESKUCioyc3NhZ+fH5ydnWFqalpcfSIiIqJC4ERhMYUmCmtqaqJ169Z8GjcREVEJIFHSf+pC4dVPNWvWxKNHj4qjL0RERKSAvExNUTd1oXBQM3v2bPj7+2Pfvn2IjY1FSkqKaCMiIiL6Ego9p2bmzJkYO3Ys2rRpAwDo0KGD6HEJgiBAIpEgNzdX+b0kIiIiOZxTI1booGbGjBkYPHgwjh8/Xpz9ISIiokKSSCQffB5jYdtQF4UOavIe5t2sWbNi6wwRERHRp1JoSbc6RXNERESqjsNPYgoFNdWqVftoYPP69esidYiIiIgKh3cUFlMoqJkxY4bcHYWJiIiISgKFgpqePXuiXLlyxdUXIiIiUoCGRFLkB1oW9fiSpNBBDefTEBERlSycUyNW6Jvv5a1+IiIiIiqJCp2pkUqlxdkPIiIiUpQSJgqr0aOfFJtTQ0RERCWHBiTQKGJUUtTjSxIGNURERCqKS7rFFH6gJREREVFJxEwNERGRiuLqJzEGNURERCqK96kR4/ATERERqQVmaoiIiFQUJwqLMaghIiJSURpQwvCTGi3p5vATERERqQVmaoiIiFQUh5/EGNQQERGpKA0UfchFnYZs1OlaiIiI6CvGoIaIiEhFSSQSpWyKWr58Oezs7KCrqws3NzdcvHixwLrNmzfP95xt27aV1fH19ZXb7+npqXC/OPxERESkoiQo+kO2FT1+y5YtGDNmDIKCguDm5obAwEB4eHjg3r17KFeunFz9HTt2ICsrS/b61atXcHFxQbdu3UT1PD09sW7dOtlrHR0dBXvGTA0REZHKyrujcFE3RSxcuBCDBg2Cn58fnJycEBQUBH19faxduzbf+mXKlIGlpaVsO3LkCPT19eWCGh0dHVE9U1NTxd8PhY8gIiIitZOSkiLaMjMz5epkZWXhypUrcHd3l5VpaGjA3d0d586dK9R5/vzzT/Ts2RMGBgai8vDwcJQrVw4ODg4YMmQIXr16pfA1MKghIiJSYZIibnlsbGxgYmIi2wICAuTO9fLlS+Tm5sLCwkJUbmFhgbi4uI/29eLFi/j3338xcOBAUbmnpydCQkIQFhaGefPm4cSJE/Dy8kJubm7h3oT/xzk1REREKkqZ96l5+vQpjI2NZeWfMqflY/788084Ozujfv36ovKePXvK/u3s7IxatWqhSpUqCA8PR8uWLQvdPjM1REREBGNjY9GWX1BjZmYGTU1NxMfHi8rj4+NhaWn5wfbT0tKwefNmDBgw4KN9qVy5MszMzBAZGanQNTCoISIiUlGfe0m3trY26tati7CwMFmZVCpFWFgYGjRo8MFjt23bhszMTPTp0+ej53n27BlevXoFKyurQvcNYFBDRESksjSUtClizJgxWL16NdavX487d+5gyJAhSEtLg5+fHwCgX79+mDRpktxxf/75Jzp16oSyZcuKylNTUzFu3DicP38ejx8/RlhYGDp27Ah7e3t4eHgo1DfOqSEiIqJC69GjBxISEjB16lTExcXB1dUVBw8elE0ejo6OhoaGOFS6d+8eTp8+jcOHD8u1p6mpiRs3bmD9+vVISkqCtbU1WrdujVmzZik8r0ciCILw6ZdGxSElJQUmJiaISUgSTdoiUieaGmr0FD2id6SkpMCirAmSk5OL7Wd43u+JdafuQt/QqEhtpae+gV8Tx2Lt7+fCTA0REZGK+hJ3FC7JOKeGiIiI1AIzNURERCrqUx9I+X4b6oJBDRERkYr6lNVL+bWhLhjUEBERqShmasTUKUAjIiKirxgzNURERCqKq5/EGNQQERGpKGU+0FIdcPiJiIiI1AIzNURERCpKAxJoFHEAqajHlyQMaoiIiFQUh5/EOPxEREREaoGZGiIiIhUl+f//itqGumBQQ0REpKI4/CTG4SciIiJSC8zUEBERqSiJElY/cfiJiIiIvjgOP4kxqCEiIlJRDGrEOKeGiIiI1AIzNURERCqKS7rFGNQQERGpKA3J262obagLDj8RERGRWmCmhoiISEVx+EmMQQ0REZGK4uonMQ4/ERERkVpgpoaIiEhFSVD04SM1StQwqCEiIlJVXP0kxuEnIiIiUgtfbaYmPDwcLVq0QGJiIkqXLl1gPTs7O4waNQqjRo36bH0jsbXbTyF4x2lEx7wGADhWtoT/AE+4N6whq3PpZhTmrNyLq7eeQENDAzWrlce2xUOhp6uN6JhXWLD2IE5dvo8Xr9/A0swEXT3rYYyfB7RLvf0WePAkHv6/bsH9qDikpP0HSzMTfO9RD+MGeqGUlqbsPEGbjmPdjtN4Hp+IMiYGaP+dK6YM7QBdnVIAgNxcKX5bfQDbDl6SnatnWzeM7e8BiTrNxqNiVavDVDyNfS1XPqBrE/w+oQfiX6Zg6pKdCL9wF6npmbC3LYex/T3Q4bvaAIDomFeY/+dBnLx8Hy9epcDSzATdvb7B2P7/+8xnZGZjTMBmRNyNxv3H8fBoXBOhv/8gOl/cy2T8ErgDEXei8ejpS/zYoxkCxnYt/jeACo2rn8RKfFDj6+uL9evXAwBKlSqFihUrol+/fpg8eTK0tD69+w0bNkRsbCxMTEwAAMHBwRg1ahSSkpJE9S5dugQDA4NPPg8VnXW50pgytAMq25hDALBl/wX0HbcaxzdMgGNlK1y6GYXuI1dglE8r/OrfDZqaGrj14Dk0/j+n+uBJPKRSAQsm9kQlG3PceRiLMXM3If2/LMwc2RkAUEpLEz3a1EcthwowMdLHvw+eY8zcTZBKpfhlaAcAwN+HLmPWij1Y/Is36jtXwsPoFxg+6y9IJBLMHtUFALBkwxGs23Eay6b2gWNlK0TcicZPs0NhbKiLH3o0/xJvH6mgY+vHITdXkL2+8zAGnYcvQyf3t0HLkOkhSH7zHzYu/BFlTQzx96HL8Ju0FsdDxqOWgw3uP46HVCrFokk9UbmCOW4/jMGouZuQ/l8mZv3/ZzVXKoWubin82KM59h6LyLcfWVk5MCttBP/+nlix8XixXzcpjqufxEp8UAMAnp6eWLduHTIzM3HgwAEMGzYMpUqVwqRJkz65TW1tbVhaWn60nrm5+Sefg5TDs4mz6PXPQ9pj3Y7TuPzvYzhWtsIvi3bgh+7NMNKntaxOVVsL2b9bNnBCywZOstd25c0Q+SQewTtOy4Iau/JmsCtvJqtjY1UGZ64+wPmIh7KySzceoX6tyujqUQ8AUNG6LLq0rour/z6R1bl4IwpeTZ3RunFNWZ0dh6/g6u3/1SH6GDNTI9HrwPWHUamCGRrVqQoAuHjjEX6f2BN1a9gBAPwHeGLFpmOIuPMUtRxs4N7QCe4N3/nMVzBDZPQLrP37lCyoMdDTwcKJPQEAF64/QnLqf3L9qGhdFr/6v83M/LXnnNKvk4pOgqJP9FWjmEY15tTo6OjA0tIStra2GDJkCNzd3bFnzx4kJiaiX79+MDU1hb6+Pry8vPDgwQPZcU+ePEH79u1hamoKAwMD1KhRAwcOHADwdvhJIpEgKSkJ4eHh8PPzQ3JyMiQSCSQSCaZPnw7g7fBTYGAgAKB3797o0aOHqG/Z2dkwMzNDSEgIAEAqlSIgIACVKlWCnp4eXFxc8Pfffxf/m/SVyM2VYsfhK0j/Lwvf1LRDwus3uHLrMczKGMFr4EJU95yM9oMXi4KR/LxJy0BpY/0C9z96moBj5+6g4f//EgGAb2pVxvW7T3H11mMAwOPnL3H07G24N/rfL4/6tSrh5OX7iIx+AQD49/4zXLj+SBRUESkiKzsHW/+5BO8ODWRDmPVrVcbOI1eQmJwGqVSK7YcvIzMzB43rVi2wnZTU/2BqUvBnnkgdqESm5n16enp49eoVfH198eDBA+zZswfGxsaYMGEC2rRpg9u3b6NUqVIYNmwYsrKycPLkSRgYGOD27dswNDSUa69hw4YIDAzE1KlTce/ePQDIt563tze6deuG1NRU2f5Dhw4hPT0dnTu//Ys/ICAAf/31F4KCglC1alWcPHkSffr0gbm5OZo1a5bv9WRmZiIzM1P2OiUlpcjvkbq5HRkDr4ELkJGVAwM9HayfNxAOla1w+WYUAOC31QcwY0Rn1KxWHlsOXESX4ctwauMkVKlYTq6tR08TsHrrCcwY0Ulun9fAhbhx7ykys3LQr1MjTPyhjWxfV496eJ2UirY/BEIQBOTkSuHbpTFG+3rI6ozs1wpv0jLQoPtsaGpIkCsV8PPgdujm+Y3y3xT6KuwPv4Hk1P/Qu52brGxdQH/0n7wWld0nQEtTA3q62tgwfxAq2+SfWX70NAF/bDmBWf+fmST1oQEJNIo4fqShRrkalcjU5BEEAUePHsWhQ4dQsWJF7NmzB2vWrEGTJk3g4uKC0NBQPH/+HLt27QIAREdHo1GjRnB2dkblypXRrl07NG3aVK5dbW1tmJiYQCKRwNLSEpaWlvkGNR4eHjAwMMDOnTtlZRs3bkSHDh1gZGSEzMxMzJ07F2vXroWHhwcqV64MX19f9OnTB6tWrSrwugICAmBiYiLbbGxsiv5mqRl723I4vmEiDv05Fn5dGmP4zL9w71EspMLbeQc+nRuhd/tvUcvBBnNGfw9723LYuPe8XDuxL5LQY9QKdGhZG/06NZLbv2aOH46tH49VM31w5MwtLAs9Jtt3+soDBAYfxm/ju+NYyASsnzcQR87cwu9/HpTV2XX0Gv4+eBmrZvrgWMgELJ/aB8tDw7B5/4VieFfoa/DXnrNwb+AEK/PSsrI5QfuQ/OY/7Fr+E46FjMcw7+/gN2ktbkU+lzs+5kUSuo5Yjk7uteHTWf4zT6pNoqRNUcuXL4ednR10dXXh5uaGixcvFlg3ODhYNgqSt+nq6orqCIKAqVOnwsrKCnp6enB3dxeNvBSWSgQ1+/btg6GhIXR1deHl5YUePXrA19cXWlpacHP7318vZcuWhYODA+7cuQMAGDFiBGbPno1GjRph2rRpuHHjRpH6oaWlhe7duyM0NBQAkJaWht27d8Pb2xsAEBkZifT0dLRq1QqGhoayLSQkBA8fFjwcMmnSJCQnJ8u2p0+fFqmf6ki7lBYq25jDtXpFTBnWATWqWmPVlhOwMDMGAFSrZCWqX9XOAs/iE0VlsQnJ6Dh0Cb5xroRFk3rme57yFqZwqGyF7z3qYcqwDpi/+gByc6UAgF9X7UM3r/ro27EhnOyt0ba5C34e0h6L1x+GVPq2zvSluzCyXyt0aV0XTvbW6N6mPgb3aoHA9YeV/ZbQVyA69jXCL95Dv04NZWVRzxKweutJLJ3SB83qO8C5WgVMGNQGtatXxJptJ0XHxyYkocOQxahfqzICJ/f63N0nNbVlyxaMGTMG06ZNw9WrV+Hi4gIPDw+8ePGiwGOMjY0RGxsr2548Ec8z/O2337BkyRIEBQXhwoULMDAwgIeHBzIyMhTqm0oENS1atEBERAQePHiA//77D+vXry/U8tiBAwfi0aNH6Nu3L27evIl69eph6dKlReqLt7c3wsLC8OLFC+zatQt6enrw9PQEAKSmpgIA9u/fj4iICNl2+/btD86r0dHRgbGxsWijD5NKBWRlZ6OiVVlYmpvg4ZN40f5H0QmwsTSVvY59kYSOQxbDxdEGS6f0gYbGxz/6giBFdk6uLBuUnpEtW1GVR/P/X/9/FfyXkQWJXB0NSKUCiBS1ce85mJsaoXWj/92+ID0jCwDkP4uaEgjvfM5iXiSh/eDFcHGsiOVTC/eZJxX0BVI1CxcuxKBBg+Dn5wcnJycEBQVBX18fa9euLbib74yEWFpawsLif4s5BEFAYGAgfvnlF3Ts2BG1atVCSEgIYmJiZCMvhaUSc2oMDAxgb28vKqtevTpycnJw4cIFNGz49q+YV69e4d69e3By+t+kTBsbGwwePBiDBw/GpEmTsHr1avz0009y59DW1kZubu5H+9KwYUPY2Nhgy5Yt+Oeff9CtWzeUKvX2HiVOTk7Q0dFBdHR0gfNnSHGzlu9By4ZOqGBhitT0TGw/dBlnrkZi2+KhkEgkGO7dEvNWH0CNquVRs1oFbNl/AQ+exGNtQH8AbwOaDkOWwMbKFDNHdMbLpFRZ2xZl3waQ2w5eQiktTThVsYa2thYi7kRj1oq96NSqjuw+NR5NamLlxuNwrlYBdWvaIurpS/z6x360blITmpoasjqL1h1GBQtTOFa2ws37z7By03H0bv/tZ37XSNVJpVKE7j2Pnm3doPXOvZKq2Vmiso05RgdswqyRnVHGxAD7w2/g+IV72LxoMID/BTQ2lmUwa2RnvEx85zNv9r8/mu4+ikV2di4SU9KQmp6Jm/eeAQCcHSrI6uSVpf2XiZeJqbh57xlKldKEY2VxdpS+jM99n5qsrCxcuXJFtPpYQ0MD7u7uOHeu4BVyqampsLW1hVQqRZ06dTB37lzUqPE2WI+KikJcXBzc3d1l9U1MTODm5oZz586hZ8/8M+v5UYmgJj9Vq1ZFx44dMWjQIKxatQpGRkaYOHEiypcvj44dOwIARo0aBS8vL1SrVg2JiYk4fvw4qlevnm97dnZ2SE1NRVhYGFxcXKCvrw99/fxXCvTu3RtBQUG4f/8+jh//370bjIyM4O/vj9GjR0MqlaJx48ZITk7GmTNnYGxsDB8fH+W/EV+Bl4lvMGzGBsS/TIGxoS6c7K2xbfFQNHdzBAAM7tUCmVnZ+CVwB5JS0lGjann8vWQYKlV4O2ky/OJdRD1LQNSzBDi3nyJu+8LbzJ2WpgaWhBzFw6cvAEFABcsyGNi1KQb3aiGrO9bPAxIJELBqH2ITklG2tCE8GtfEz0PayeoEjO2GX1ftx/j5W/EyMRWWZibw6dwI/gM8i/ttIjUTfvEensUlok8HcUBcSksTWwOHYMay3eg1ZhXS0jNRycYcK6b3lWV0wi/cxaOnCXj0NAE12v4iOj7x0jLZv7uPWim6yV/TPr/K1ckrA4CIO0/x96HLsLEqgxt7ZirvYqlEeH+Rio6ODnR0dERlL1++RG5urijTAgAWFha4e/duvu06ODhg7dq1qFWrFpKTk/H777+jYcOGuHXrFipUqIC4uDhZG++3mbevsCSCIJTovLivry+SkpLyTUElJiZi5MiR2LNnD7KystC0aVMsXboUVau+Xdb4008/4Z9//sGzZ89gbGwMT09PLFq0CGXLls33jsJDhgzBtm3b8OrVK0ybNg3Tp0/P947Cd+7cgZOTE2xtbREVFSUaChMEAUuWLMHKlSvx6NEjlC5dGnXq1MHkyZPznaScn5SUFJiYmCAmIYlDUaS2NNXpgTNE70hJSYFFWRMkJycX28/wvN8TYRHRMDQq2jlS36SgpWtFufK834PviomJQfny5XH27Fk0aNBAVj5+/HicOHECFy58fFFEdnY2qlevjl69emHWrFk4e/YsGjVqhJiYGFhZ/S8D2L17d0gkEmzZsqXQ11LiMzXBwcEF7jM1NZXdHyY/H5o/07x5c7wfz61cuRIrV64UlT1+/Fju2OrVq8sdm0cikWDkyJEYOXJkgecmIiJSBmXefO/p06eiIOz9LA0AmJmZQVNTE/Hx4nmM8fHxhbqhLfD26QC1a9dGZGQkAMiOi4+PFwU18fHxcHV1VeBKVGSiMBERERWv9xes5BfUaGtro27duggLC5OVSaVShIWFiTI3H5Kbm4ubN2/KAphKlSrB0tJS1GZKSgouXLhQ6DbzlPhMDRERERXgCzwnYcyYMfDx8UG9evVQv359BAYGIi0tDX5+fgCAfv36oXz58ggICAAAzJw5E99++y3s7e2RlJSE+fPn48mTJxg4cODb00skGDVqFGbPno2qVauiUqVKmDJlCqytrdGpUyeF+saghoiISEV9iad09+jRAwkJCZg6dSri4uLg6uqKgwcPyib6RkdHi24hkJiYiEGDBiEuLg6mpqaoW7cuzp49K1qpPH78eKSlpeGHH35AUlISGjdujIMHD8rdpO+j11LSJwp/jThRmL4GnChM6upzThQOv/FUKROFm9eyKdb+fi6cU0NERERqgcNPREREKuoLTKkp0RjUEBERqSpGNSIcfiIiIiK1wEwNERGRivoSq59KMgY1REREKkoiebsVtQ11weEnIiIiUgvM1BAREakozhMWY1BDRESkqhjViHD4iYiIiNQCMzVEREQqiqufxBjUEBERqSiufhJjUENERKSiOKVGjHNqiIiISC0wU0NERKSqmKoRYVBDRESkojhRWIzDT0RERKQWmKkhIiJSUVz9JMaghoiISEVxSo0Yh5+IiIhILTBTQ0REpKqYqhFhUENERKSiuPpJjMNPREREpBaYqSEiIlJRXP0kxqCGiIhIRXFKjRiDGiIiIlXFqEaEc2qIiIhILTBTQ0REpKK4+kmMQQ0REZGqUsJEYTWKaTj8REREROqBmRoiIiIVxXnCYgxqiIiIVBWjGhEOPxEREZFaYKaGiIhIRXH1kxgzNURERCoq7zEJRd0UtXz5ctjZ2UFXVxdubm64ePFigXVXr16NJk2awNTUFKampnB3d5er7+vrC4lEIto8PT0V7heDGiIiIiq0LVu2YMyYMZg2bRquXr0KFxcXeHh44MWLF/nWDw8PR69evXD8+HGcO3cONjY2aN26NZ4/fy6q5+npidjYWNm2adMmhfvGoIaIiEhFSZS0KWLhwoUYNGgQ/Pz84OTkhKCgIOjr62Pt2rX51g8NDcXQoUPh6uoKR0dHrFmzBlKpFGFhYaJ6Ojo6sLS0lG2mpqYK9oxBDRERkepSYlSTkpIi2jIzM+VOl5WVhStXrsDd3V1WpqGhAXd3d5w7d65QXU5PT0d2djbKlCkjKg8PD0e5cuXg4OCAIUOG4NWrV4V+G2R9UfgIIiIiKhEkSvoPAGxsbGBiYiLbAgIC5M738uVL5ObmwsLCQlRuYWGBuLi4QvV5woQJsLa2FgVGnp6eCAkJQVhYGObNm4cTJ07Ay8sLubm5Cr0fXP1EREREePr0KYyNjWWvdXR0lH6OX3/9FZs3b0Z4eDh0dXVl5T179pT929nZGbVq1UKVKlUQHh6Oli1bFrp9ZmqIiIhUlARKWP30/20ZGxuLtvyCGjMzM2hqaiI+Pl5UHh8fD0tLyw/29ffff8evv/6Kw4cPo1atWh+sW7lyZZiZmSEyMlKRt4NBDRERkar63BOFtbW1UbduXdEk37xJvw0aNCjwuN9++w2zZs3CwYMHUa9evY+e59mzZ3j16hWsrKwU6B2DGiIiIlLAmDFjsHr1aqxfvx537tzBkCFDkJaWBj8/PwBAv379MGnSJFn9efPmYcqUKVi7di3s7OwQFxeHuLg4pKamAgBSU1Mxbtw4nD9/Ho8fP0ZYWBg6duwIe3t7eHh4KNQ3zqkhIiJSUZ9687z321BEjx49kJCQgKlTpyIuLg6urq44ePCgbPJwdHQ0NDT+lzNZuXIlsrKy0LVrV1E706ZNw/Tp06GpqYkbN25g/fr1SEpKgrW1NVq3bo1Zs2YpPK9HIgiCoNjlUHFLSUmBiYkJYhKSRJO2iNSJpob63Jqd6F0pKSmwKGuC5OTkYvsZnvd74vbjBBgV8RxvUlLgZGderP39XDj8RERERGqBw09EREQq6ksMP5VkDGqIiIhU1Kc85iC/NtQFh5+IiIhILTBTQ0REpKI4/CTGoIaIiEhFvfvspqK0oS4Y1BAREakqTqoR4ZwaIiIiUgvM1BAREakoJmrEGNQQERGpKE4UFuPwExEREakFZmqIiIhUFFc/iTGoISIiUlWcVCPC4SciIiJSC8zUEBERqSgmasQY1BAREakorn4S4/ATERERqQVmaoiIiFRW0Vc/qdMAFIMaIiIiFcXhJzEOPxEREZFaYFBDREREaoHDT0RERCqKw09iDGqIiIhUFB+TIMbhJyIiIlILzNQQERGpKA4/iTGoISIiUlF8TIIYh5+IiIhILTBTQ0REpKqYqhFhUENERKSiuPpJjMNPREREpBaYqSEiIlJRXP0kxqCGiIhIRXFKjRiDGiIiIlXFqEaEc2qIiIhILTBTQ0REpKK4+kmMQQ0REZGK4kRhMQY1JZAgCACAN29SvnBPiIqPpoYa/SQleseblLc/u/N+lhenlJSi/55QRhslBYOaEujNmzcAAIfKFb9wT4iI6FO9efMGJiYmxdK2trY2LC0tUbWSjVLas7S0hLa2tlLa+pIkwucIJUkhUqkUMTExMDIygkSd8oIlVEpKCmxsbPD06VMYGxt/6e4QKR0/45+XIAh48+YNrK2toaFRfOtxMjIykJWVpZS2tLW1oaurq5S2viRmakogDQ0NVKhQ4Ut346tjbGzMH/ik1vgZ/3yKK0PzLl1dXbUIRJSJS7qJiIhILTCoISIiIrXAoIa+ejo6Opg2bRp0dHS+dFeIigU/4/S14ERhIiIiUgvM1BAREZFaYFBDREREaoFBDREREakFBjVECrKzs0NgYOCX7gbRR4WHh0MikSApKemD9fiZJnXBoIZKFF9fX0gkEvz666+i8l27dn32uysHBwejdOnScuWXLl3CDz/88Fn7Quot73MvkUigra0Ne3t7zJw5Ezk5OUVqt2HDhoiNjZXdCI6faVJ3DGqoxNHV1cW8efOQmJj4pbuSL3Nzc+jr63/pbpCa8fT0RGxsLB48eICxY8di+vTpmD9/fpHazHs+0Mf+IOBnmtQFgxoqcdzd3WFpaYmAgIAC65w+fRpNmjSBnp4ebGxsMGLECKSlpcn2x8bGom3bttDT00OlSpWwceNGuRT7woUL4ezsDAMDA9jY2GDo0KFITU0F8DZt7+fnh+TkZNlf0NOnTwcgTtX37t0bPXr0EPUtOzsbZmZmCAkJAfD2WV4BAQGoVKkS9PT04OLigr///lsJ7xSpEx0dHVhaWsLW1hZDhgyBu7s79uzZg8TERPTr1w+mpqbQ19eHl5cXHjx4IDvuyZMnaN++PUxNTWFgYIAaNWrgwIEDAMTDT/xM09eAQQ2VOJqampg7dy6WLl2KZ8+eye1/+PAhPD098f333+PGjRvYsmULTp8+jeHDh8vq9OvXDzExMQgPD8f27dvxxx9/4MWLF6J2NDQ0sGTJEty6dQvr16/HsWPHMH78eABv0/aBgYEwNjZGbGwsYmNj4e/vL9cXb29v7N27VxYMAcChQ4eQnp6Ozp07AwACAgIQEhKCoKAg3Lp1C6NHj0afPn1w4sQJpbxfpJ709PSQlZUFX19fXL58GXv27MG5c+cgCALatGmD7OxsAMCwYcOQmZmJkydP4ubNm5g3bx4MDQ3l2uNnmr4KAlEJ4uPjI3Ts2FEQBEH49ttvhf79+wuCIAg7d+4U8j6uAwYMEH744QfRcadOnRI0NDSE//77T7hz544AQLh06ZJs/4MHDwQAwqJFiwo897Zt24SyZcvKXq9bt04wMTGRq2draytrJzs7WzAzMxNCQkJk+3v16iX06NFDEARByMjIEPT19YWzZ8+K2hgwYIDQq1evD78Z9NV493MvlUqFI0eOCDo6OkKnTp0EAMKZM2dkdV++fCno6ekJW7duFQRBEJydnYXp06fn2+7x48cFAEJiYqIgCPxMk/rjU7qpxJo3bx6+++47ub8mr1+/jhs3biA0NFRWJggCpFIpoqKicP/+fWhpaaFOnTqy/fb29jA1NRW1c/ToUQQEBODu3btISUlBTk4OMjIykJ6eXuj5BVpaWujevTtCQ0PRt29fpKWlYffu3di8eTMAIDIyEunp6WjVqpXouKysLNSuXVuh94PU2759+2BoaIjs7GxIpVL07t0bXbp0wb59++Dm5iarV7ZsWTg4OODOnTsAgBEjRmDIkCE4fPgw3N3d8f3336NWrVqf3A9+pkmVMaihEqtp06bw8PDApEmT4OvrKytPTU3Fjz/+iBEjRsgdU7FiRdy/f/+jbT9+/Bjt2rXDkCFDMGfOHJQpUwanT5/GgAEDkJWVpdCkSW9vbzRr1gwvXrzAkSNHoKenB09PT1lfAWD//v0oX7686Dg+h4fe1aJFC6xcuRLa2tqwtraGlpYW9uzZ89HjBg4cCA8PD+zfvx+HDx9GQEAAFixYgJ9++umT+8LPNKkqBjVUov36669wdXWFg4ODrKxOnTq4ffs27O3t8z3GwcEBOTk5uHbtGurWrQvg7V+X766munLlCqRSKRYsWAANjbdTy7Zu3SpqR1tbG7m5uR/tY8OGDWFjY4MtW7bgn3/+Qbdu3VCqVCkAgJOTE3R0dBAdHY1mzZopdvH0VTEwMJD7TFevXh05OTm4cOECGjZsCAB49eoV7t27BycnJ1k9GxsbDB48GIMHD8akSZOwevXqfIMafqZJ3TGooRLN2dkZ3t7eWLJkiaxswoQJ+PbbbzF8+HAMHDgQBgYGuH37No4cOYJly5bB0dER7u7u+OGHH7By5UqUKlUKY8eOhZ6enmxpq729PbKzs7F06VK0b98eZ86cQVBQkOjcdnZ2SE1NRVhYGFxcXKCvr19gBqd3794ICgrC/fv3cfz4cVm5kZER/P39MXr0aEilUjRu3BjJyck4c+YMjI2N4ePjUwzvGqmLqlWromPHjhg0aBBWrVoFIyMjTJw4EeXLl0fHjh0BAKNGjYKXlxeqVauGxMREHD9+HNWrV8+3PX6mSe196Uk9RO96d8JknqioKEFbW1t49+N68eJFoVWrVoKhoaFgYGAg1KpVS5gzZ45sf0xMjODl5SXo6OgItra2wsaNG4Vy5coJQUFBsjoLFy4UrKysBD09PcHDw0MICQkRTaoUBEEYPHiwULZsWQGAMG3aNEEQxJMq89y+fVsAINja2gpSqVS0TyqVCoGBgYKDg4NQqlQpwdzcXPDw8BBOnDhRtDeL1EZ+n/s8r1+/Fvr27SuYmJjIPqv379+X7R8+fLhQpUoVQUdHRzA3Nxf69u0rvHz5UhAE+YnCgsDPNKk3iSAIwheMqYg+i2fPnsHGxgZHjx5Fy5Ytv3R3iIioGDCoIbV07NgxpKamwtnZGbGxsRg/fjyeP3+O+/fvy+YGEBGReuGcGlJL2dnZmDx5Mh49egQjIyM0bNgQoaGhDGiIiNQYMzVERESkFviYBCIiIlILDGqIiIhILTCoISIiIrXAoIaIiIjUAoMaIsqXr68vOnXqJHvdvHlzjBo16rP3Izw8HBKJBElJSQXWkUgk2LVrV6HbnD59OlxdXYvUr8ePH0MikSAiIqJI7RCR8jCoIVIhvr6+kEgkkEgk0NbWhr29PWbOnImcnJxiP/eOHTswa9asQtUtTCBCRKRsvE8NkYrx9PTEunXrkJmZiQMHDmDYsGEoVaoUJk2aJFc3KysL2traSjlvmTJllNIOEVFxYaaGSMXo6OjA0tIStra2GDJkCNzd3bFnzx4A/xsymjNnDqytrWVPN3/69Cm6d++O0qVLo0yZMujYsSMeP34sazM3NxdjxoxB6dKlUbZsWYwfPx7v38Lq/eGnzMxMTJgwATY2NtDR0YG9vT3+/PNPPH78GC1atAAAmJqaQiKRwNfXFwAglUoREBCASpUqQU9PDy4uLvj7779F5zlw4ACqVasGPT09tGjRQtTPwpowYQKqVasGfX19VK5cGVOmTEF2drZcvVWrVsHGxgb6+vro3r07kpOTRfvXrFmD6tWrQ1dXF46OjlixYoXCfSGiz4dBDZGK09PTQ1ZWlux1WFgY7t27hyNHjmDfvn3Izs6Gh4cHjIyMcOrUKZw5cwaGhobw9PSUHbdgwQIEBwdj7dq1OH36NF6/fo2dO3d+8Lz9+vXDpk2bsGTJEty5cwerVq2CoaEhbGxssH37dgDAvXv3EBsbi8WLFwMAAgICEBISgqCgINy6dQujR49Gnz59cOLECQBvg68uXbqgffv2iIiIwMCBAzFx4kSF3xMjIyMEBwfj9u3bWLx4MVavXo1FixaJ6kRGRmLr1q3Yu3cvDh48iGvXrmHo0KGy/aGhoZg6dSrmzJmDO3fuYO7cuZgyZQrWr1+vcH+I6DP5gg/TJCIFvfs0Z6lUKhw5ckTQ0dER/P39ZfstLCyEzMxM2TEbNmwQHBwcRE9azszMFPT09IRDhw4JgiAIVlZWwm+//Sbbn52dLVSoUEH05OhmzZoJI0eOFARBEO7duycAEI4cOZJvP/N7OnRGRoagr68vnD17VlR3wIABQq9evQRBEIRJkyYJTk5Oov0TJkyQa+t9AISdO3cWuH/+/PlC3bp1Za+nTZsmaGpqCs+ePZOV/fPPP4KGhoYQGxsrCIIgVKlSRdi4caOonVmzZgkNGjQQBOHt0+MBCNeuXSvwvET0eXFODZGK2bdvHwwNDZGdnQ2pVIrevXtj+vTpsv3Ozs6ieTTXr19HZGQkjIyMRO1kZGTg4cOHSE5ORmxsLNzc3GT7tLS0UK9ePbkhqDwRERHQ1NREs2bNCt3vyMhIpKeno1WrVqLyrKws1K5dGwBw584dUT8AoEGDBoU+R54tW7ZgyZIlePjwIVJTU5GTkwNjY2NRnYoVK6J8+fKi80ilUty7dw9GRkZ4+PAhBgwYgEGDBsnq5OTkwMTEROH+ENHnwaCGSMW0aNECK1euhLa2NqytraGlJf42NjAwEL1OTU1F3bp1ERoaKteWubn5J/VBT09P4WNSU1MBAPv37xcFE8DbeULKcu7cOXh7e2PGjBnw8PCAiYkJNm/ejAULFijc19WrV8sFWZqamkrrKxEpF4MaIhVjYGAAe3v7QtevU6cOtmzZgnLlysllK/JYWVnhwoULaNq0KYC3GYkrV66gTp06+dZ3dnaGVCrFiRMn4O7uLrc/L1OUm5srK3NycoKOjg6io6MLzPBUr15dNuk5z/nz5z9+ke84e/YsbG1t8fPPP8vKnjx5IlcvOjoaMTExsLa2lp1HQ0MDDg4OsLCwgLW1NR49egRvb2+Fzk9EXw4nChOpOW9vb5iZmaFjx444deoUoqKiEB4ejhEjRuDZs2cAgJEjR+LXX3/Frl27cPfuXQwdOvSD95ixs7ODj48P+vfvj127dsna3Lp1KwDA1tYWEokE+/btQ0JCAlJTU2FkZAR/f3+MHj0a69evx8OHD3H16lUsXbpUNvl28ODBePDgAcaNG4d79+5h48aNCA4OVuh6q1atiujoaGzevBkPHz7EkiVL8p30rKurCx8fH1y/fh2nTp3CiBEj0L17d1haWgIAZsyYgYCAACxZsgT379/HzZs3sW7dOixcuFCh/hDR58OghkjN6evr4+TJk6hYsSK6dOmC6tWrY8CAAcjIyJBlbsaOHYu+ffvCx8cHDRo0gJGRETp37vzBdleuXImuXbti6NChcHR0xKBBg5CWlgYAKF++PGbMmIGJEyfCwsICw4cPBwDMmjULU6ZMQUBAAKpXrw5PT0/s378flSpVAvB2nsv27duxa9cuuLi4ICgoCHPnzlXoejt06IDRo0dj+PDhcHV1xdmzZzFlyhS5evb29ujSpQvatGmD1q1bo1atWqIl2wMHDsSaNWuwbt06ODs7o1mzZggODpb1lYhKHolQ0ExAIiIiIhXCTA0RERGpBQY1REREpBYY1BAREZFaYFBDREREaoFBDREREakFBjVERESkFhjUEBERkVpgUENERERqgUENERERqQUGNURERKQWGNQQERGRWmBQQ0RERGrh/wAxL9JIsJySmAAAAABJRU5ErkJggg==\n"
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Global Confusion Matrix for RNN:\n",
            "                          Predicted\n",
            "               ------------------------------\n",
            "               Negative     | Positive\n",
            "Actual Negative    20852104        | 44797\n",
            "Actual Positive    362388        | 78211\n",
            "445/445 [==============================] - 2s 2ms/step\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 800x600 with 0 Axes>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 2 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjUAAAHHCAYAAABHp6kXAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAABs7klEQVR4nO3dd1gUV9sG8HsBWTqiUkRpioJYwEo0KhpRwBJL7KhgjaixYs2rYizEJHaNGBuSYDf2WFHsGnuMHUWxADaKQKg73x9+TJwAysKi7Hr/cs0V98yZM2eWBR6ec86MTBAEAURERERqTutjd4CIiIhIFRjUEBERkUZgUENEREQagUENERERaQQGNURERKQRGNQQERGRRmBQQ0RERBqBQQ0RERFpBAY1REREpBEY1GiIoKAgyGSyIh1rb2+P9u3bq6wvDx48gEwmQ2hoqMraLCnx8fHo2rUrypcvD5lMhoULF6r8HDKZDEFBQSpvV135+/vD3t5eZe1lZ2djwoQJsLGxgZaWFjp16qSytosrNDQUMpkMDx48EMtatGiBFi1afNB++Pv7w8jISKVtqvo6/vzzT+jq6uLhw4cqa/NDCwkJga2tLTIyMj52Vz5ZDGpKsejoaIwYMQLVq1eHgYEBDAwM4OLiguHDh+Ovv/762N1Tmfj4eAQGBsLZ2RkGBgYwNDRE/fr1MWvWLCQmJpbouceMGYMDBw5g8uTJ+PXXX+Ht7V2i5/uQcgNdLS0tPHr0KM/+5ORk6OvrQyaTYcSIEUq3n5aWhqCgIERGRqqgt0W3Zs0a/Pjjj+jatSvWrVuHMWPGfLBzN2rUCDKZDMuXL1d526r+Y6O0+/bbb9GrVy/Y2dmJZS1atECtWrXee+zJkyfh4+ODSpUqQU9PD7a2tujQoQPWr18P4E1QJ5PJ3rv5+/uL55XJZKhWrVq+5zt06JB4zNatW8Vyf39/ZGZmYsWKFcV4J6g4dD52Byh/e/bsQY8ePaCjowNfX1+4urpCS0sLt27dwu+//47ly5cjOjpa8gNAHZ0/fx5t27ZFSkoK+vTpg/r16wMALly4gO+//x7Hjx/HwYMHS+z8R44cQceOHREYGFhi5/jnn3+go/PxvtXkcjk2bNiACRMmSMp///33YrWblpaGGTNmAIBSf7GvXLkSCoWiWOd+25EjR1CpUiUsWLBAZW0Wxt27d3H+/HnY29sjPDwcAQEBH/T8muTKlSs4fPgwTp8+rfSxW7ZsQY8ePeDm5oZRo0bBzMwM0dHROH78OFauXInevXvj66+/hqenp3hMdHQ0pk2bhiFDhqBZs2ZiedWqVcV/6+npISoqCn/++ScaNWokOWd4eDj09PSQnp4uKdfT04Ofnx/mz5+Pb775psjZcyo6BjWl0L1799CzZ0/Y2dkhIiICFStWlOyfO3cufv75Z2hpqXeiLTExEZ07d4a2tjYuX74MZ2dnyf7Zs2dj5cqVJdqHZ8+eoWzZsiV6Dj09vRJt/33atm2bb1Czfv16tGvXDtu2bfsg/UhNTYWhoSHKlCmj0nZV/TVUKBTIzMx879ftt99+g4WFBebNm4euXbviwYMHKh1W+5SsXbsWtra2+Oyzz5Q+NigoCC4uLjh79ix0dXUl+549ewYAaNy4MRo3biyWX7hwAdOmTUPjxo3Rp0+ffNutWrUqsrOzsWHDBklQk56eju3btxf4vdO9e3f88MMPOHr0KL744gulr4eKR71/K2qoH374AampqVi7dm2egAYAdHR0MHLkSNjY2LyznezsbMycORNVq1aFXC6Hvb09pkyZUuB478GDB+Hm5gY9PT24uLjk+Uv+1atXCAwMRO3atWFkZAQTExP4+Pjg6tWrRbrOFStW4MmTJ5g/f36egAYALC0t8b///U9S9vPPP6NmzZqQy+WwtrbG8OHD8wxR5aasb9y4gZYtW8LAwACVKlXCDz/8INbJnesgCAKWLVsmppKBgucn5Tc/4sKFC/Dy8kKFChWgr68PBwcHDBgwQHJcfnNqLl++DB8fH5iYmMDIyAitWrXC2bNn8z3fqVOnMHbsWJibm8PQ0BCdO3fG8+fPC3xf/6t37964cuUKbt26JZbFxcXhyJEj6N27d576mZmZmDZtGurXrw9TU1MYGhqiWbNmOHr0qFjnwYMHMDc3BwDMmDFDfP9yrzN3Dse9e/fQtm1bGBsbw9fXV9z39i//6dOnQ0tLCxEREZJ+DBkyBLq6ugV+vnLnbh09ehTXr18X+5A7HJaamopx48bBxsYGcrkcTk5O+OmnnyAIgqSd3OG38PBw8bO1f//+976v69evR9euXdG+fXuYmpqKQx0f0okTJ9CtWzfY2tpCLpfDxsYGY8aMwT///JNv/fv378PLywuGhoawtrbGd999l+f9UCgUWLhwIWrWrAk9PT1YWlri66+/RkJCwnv7s2TJEtSsWRMGBgYwMzNDgwYNCvW+7NixA1988UWRMhv37t1Dw4YN8wQ0AGBhYaF0e2/r1asXNm3aJMks7t69G2lpaejevXu+x9SvXx/lypXDzp07i3VuKhoGNaXQnj174OjoCHd392K1M2jQIEybNg316tXDggUL4OHhgeDgYPTs2TNP3bt376JHjx7w8fFBcHAwdHR00K1bNxw6dEisc//+fezYsQPt27fH/PnzMX78eFy7dg0eHh54+vSp0v3btWsX9PX10bVr10LVDwoKwvDhw2FtbY158+bhq6++wooVK9CmTRtkZWVJ6iYkJMDb2xuurq6YN28enJ2dMXHiROzbtw8A0Lx5c/z6668AgNatW+PXX38VXxfWs2fP0KZNGzx48ACTJk3CkiVL4Ovrmyc4+a/r16+jWbNmuHr1KiZMmICpU6ciOjoaLVq0wLlz5/LU/+abb3D16lVMnz4dAQEB2L17t1JzYJo3b47KlStLfrls2rQJRkZGaNeuXZ76ycnJWLVqFVq0aIG5c+ciKCgIz58/h5eXF65cuQIAMDc3F+eRdO7cWXz/unTpIraTnZ0NLy8vWFhY4KeffsJXX32Vb//+97//wc3NDQMHDsTr168BAAcOHMDKlSsxbdo0uLq65nucubk5fv31Vzg7O6Ny5cpiH2rUqAFBEPDll19iwYIF8Pb2xvz58+Hk5ITx48dj7Nixedo6cuQIxowZgx49emDRokXvzbicO3cOUVFR6NWrF3R1ddGlSxeEh4e/85iSsGXLFqSlpSEgIABLliyBl5cXlixZgn79+uWpm5OTA29vb1haWuKHH35A/fr1MX36dEyfPl1S7+uvv8b48ePx+eefY9GiRejfvz/Cw8Ph5eWV5/vsbStXrsTIkSPh4uKChQsXYsaMGXBzc8v3M/22J0+eICYmBvXq1SvSe5Cb0X78+HGRjn+X3r17IzY2VjJvbP369WjVqtU7A6Z69erh1KlTKu8PFYJApUpSUpIAQOjUqVOefQkJCcLz58/FLS0tTdw3ffp04e0v55UrVwQAwqBBgyRtBAYGCgCEI0eOiGV2dnYCAGHbtm2SflSsWFGoW7euWJaeni7k5ORI2ouOjhbkcrnw3XffScoACGvXrn3ntZqZmQmurq7vrJPr2bNngq6urtCmTRtJH5YuXSoAENasWSOWeXh4CACEsLAwsSwjI0OwsrISvvrqK0m7AIThw4dLyv77XuZau3atAECIjo4WBEEQtm/fLgAQzp8//86+AxCmT58uvu7UqZOgq6sr3Lt3Tyx7+vSpYGxsLDRv3jzP+Tw9PQWFQiGWjxkzRtDW1hYSExPfed7c63j+/LkQGBgoODo6ivsaNmwo9O/fP9/3IDs7W8jIyJC0lZCQIFhaWgoDBgwQy54/f57n2nL5+fkJAIRJkyblu8/Ozk5Sdu3aNUFXV1cYNGiQkJCQIFSqVElo0KCBkJWV9c5rFIQ3X++aNWtKynbs2CEAEGbNmiUp79q1qyCTyYSoqCixDICgpaUlXL9+/b3nyjVixAjBxsZG/LocPHhQACBcvnxZUu+/n5nc/np4eLz3HHZ2dkK7du3eWeftnwG5goODBZlMJjx8+FAsy/16fPPNN2KZQqEQ2rVrJ+jq6grPnz8XBEEQTpw4IQAQwsPDJW3u378/T/l/r6Njx455vg6FcfjwYQGAsHv37jz78vva/tfq1asFAIKurq7QsmVLYerUqcKJEyfy/Kx62/nz59/5M+rt8zZo0EAYOHCgIAhvvg90dXWFdevWCUePHhUACFu2bMlz/JAhQwR9ff139ptKBjM1pUxycjIA5Lv8skWLFjA3Nxe3ZcuWFdjOH3/8AQB5/iodN24cAGDv3r2Scmtra3Tu3Fl8bWJign79+uHy5cuIi4sD8GbCae48npycHLx8+RJGRkZwcnLCpUuXlL1UJCcnw9jYuFB1Dx8+jMzMTIwePVoyl2jw4MEwMTHJcz1GRkaSsXJdXV00atQI9+/fV7qfBcmdx7Fnz553/gX7tpycHBw8eBCdOnVClSpVxPKKFSuid+/eOHnypPgZyDVkyBBJWr5Zs2bIyclRaulr7969ERUVhfPnz4v/z2/oCQC0tbXFVL5CocCrV6+QnZ2NBg0aKP11Luzk2Vq1amHGjBlYtWoVvLy88OLFC6xbt67IE6z/+OMPaGtrY+TIkZLycePGQRAEMWOXy8PDAy4uLoVqOzs7G5s2bUKPHj3Er8sXX3wBCwuLD56t0dfXF/+dmpqKFy9eoEmTJhAEAZcvX85T/+0MX+6wW2ZmJg4fPgzgTebH1NQUrVu3xosXL8Stfv36MDIykgxB/lfZsmXx+PFjnD9/XqlrePnyJQDAzMxMqeNyDRgwAPv370eLFi1w8uRJzJw5E82aNUO1atWKNPH4v3r37o3ff/8dmZmZ2Lp1K7S1tSU/K/NjZmaGf/75B2lpacU+PymHQU0pk/tLPiUlJc++FStW4NChQ/jtt9/e287Dhw+hpaUFR0dHSbmVlRXKli2b5xeio6NjnvHs6tWrA4A4h0ShUGDBggWoVq0a5HI5KlSoAHNzc/z1119ISkoq9DXmMjExEYcbCnM9AODk5CQp19XVRZUqVfJcT+XKlfNcj5mZWaHmBRSWh4cHvvrqK8yYMQMVKlRAx44dsXbt2nfeo+L58+dIS0vLcx0AUKNGDSgUijzLr21tbSWvc3/4K3MtdevWhbOzM9avX4/w8HBYWVm9cxLjunXrUKdOHejp6aF8+fIwNzfH3r17lfo66+jooHLlyoWuP378eLi6uuLPP//E9OnTCx1k5Ofhw4ewtrbOEzTXqFFD3P82BweHQrd98OBBPH/+HI0aNUJUVBSioqIQHR2Nli1bYsOGDSpd2fU+MTEx8Pf3R7ly5WBkZARzc3N4eHgAQJ6vlZaWliSQBvJ+j9+9exdJSUmwsLCQ/AFlbm6OlJQUceJtfiZOnAgjIyM0atQI1apVw/Dhw5UaghH+M7dHGV5eXjhw4AASExNx/PhxDB8+HA8fPkT79u3f2efC6NmzJ5KSkrBv3z6Eh4ejffv27/1jLPdaVLX66fjx4+jQoQOsra0hk8mwY8cOpY7PnSf4383Q0FAl/StNuPqplDE1NUXFihXx999/59mXO8fm7Ymq76PKJYVz5szB1KlTMWDAAMycORPlypWDlpYWRo8eXaQf5M7Ozrhy5QoyMzPzneRXHNra2vmWF+YHZ0HvWU5OTp56W7duxdmzZ7F7924cOHAAAwYMwLx583D27FmV3eysONfytt69e2P58uUwNjZGjx49Clw999tvv8Hf3x+dOnXC+PHjYWFhAW1tbQQHB+PevXuFPt/bmb3CuH//Pu7evQsAuHbtWqGPU4W3Mx7vk5uNKWii6LFjx9CyZUuV9OtdcnJy0Lp1a7x69QoTJ06Es7MzDA0N8eTJE/j7+xfpe1KhULwz45Q7OTw/NWrUwO3bt7Fnzx7s378f27Ztw88//4xp06aJS//zU758eQDKBekFMTAwQLNmzdCsWTNUqFABM2bMwL59++Dn51fkNitWrIgWLVpg3rx5OHXqVKFWCyYkJMDAwECpz9W7pKamwtXVFQMGDJDMWyuswMBADB06VFLWqlUrNGzYUCX9K00Y1JRC7dq1w6pVq/K9P0Jh2dnZQaFQ4O7du+Jfp8CbG90lJibmub9NVFQUBEGQ/EK/c+cOAIiTJrdu3YqWLVti9erVkmMTExNRoUIFpfvYoUMHnDlzBtu2bUOvXr3eez0AcPv2bclfm5mZmYiOjpbcg6K4cjMhiYmJkqXCBQ33fPbZZ/jss88we/ZsrF+/Hr6+vti4cSMGDRqUp665uTkMDAxw+/btPPtu3boFLS2t965qK6revXtj2rRpiI2Nfeek6K1bt6JKlSr4/fffJZ+H/04oVWXArFAo4O/vDxMTE4wePRpz5sxB165di/QDHHjzeTl8+DBev34t+as6dwVYUe/vlJqaip07d6JHjx75TnAfOXIkwsPDP0hQc+3aNdy5cwfr1q2TTAx+e3L/2xQKBe7fvy9mZ4C83+NVq1bF4cOH8fnnnxfpF7KhoSF69OiBHj16IDMzE126dMHs2bMxefLkApfI5658jI6OVvp879KgQQMAQGxsbLHb6t27NwYNGoSyZcuibdu2760fHR0t+blbXD4+PvDx8Slwf0ZGBr799lts2LABiYmJqFWrFubOnSveP8rIyEjyR9bVq1dx48YNhISEqKyPpQWHn0qhCRMmwMDAAAMGDEB8fHye/YX5Cz33G++/t/2fP38+AORZ9fL06VNs375dfJ2cnIywsDC4ubnBysoKwJuMwX/PvWXLFjx58uT9F5WPoUOHomLFihg3bpz4w/Vtz549w6xZswAAnp6e0NXVxeLFiyV9WL16NZKSkvJdxVNUuTfgOn78uFiWmpqKdevWSeolJCTkeT/c3NwAoMAhKG1tbbRp0wY7d+6UZNzi4+Oxfv16NG3aFCYmJiq4iryqVq2KhQsXIjg4+J3Bcm5m6O1rO3fuHM6cOSOpZ2BgAAAquevz/Pnzcfr0afzyyy+YOXMmmjRpgoCAALx48aJI7bVt2xY5OTlYunSppHzBggWQyWTv/AXxLtu3b0dqaiqGDx+Orl275tnat2+Pbdu2fZDb5Of3dRIEAYsWLSrwmLffD0EQsHTpUpQpUwatWrUC8Cb7lJOTg5kzZ+Y5Njs7+51f69y5Mbl0dXXh4uICQRDeOeesUqVKsLGxwYULFwqs8y7/vRVArtx5hfkN9Sqra9eumD59On7++edCZZUvXbqEJk2aFPu8hTVixAicOXMGGzduxF9//YVu3brB29tbzHz+16pVq1C9enXJjQc1BTM1pVC1atWwfv169OrVC05OTuIdhQVBQHR0NNavXw8tLa13zldwdXWFn58ffvnlFyQmJsLDwwN//vkn1q1bh06dOuX5S7J69eoYOHAgzp8/D0tLS6xZswbx8fFYu3atWKd9+/b47rvv0L9/fzRp0gTXrl1DeHh4nnH6wjIzM8P27dvRtm1buLm5Se4ofOnSJWzYsEG8YZa5uTkmT56MGTNmwNvbG19++SVu376Nn3/+GQ0bNizwBlpF0aZNG9ja2mLgwIEYP348tLW1sWbNGpibmyMmJkast27dOvz888/o3LkzqlatitevX2PlypUwMTF5519zs2bNwqFDh9C0aVMMGzYMOjo6WLFiBTIyMiT30ikJo0aNem+d9u3b4/fff0fnzp3Rrl07REdHIyQkBC4uLpK5Xvr6+nBxccGmTZtQvXp1lCtXDrVq1SrUbe3fdvPmTUydOhX+/v7o0KEDgDf36HFzc8OwYcOwefNm5S4Sb7KALVu2xLfffosHDx7A1dUVBw8exM6dOzF69GjJnWOVER4ejvLlyxf4C+vLL7/EypUrsXfv3iJnmd4WFRUlBvZvq1u3Ltq0aYOqVasiMDAQT548gYmJCbZt21bgMI6enh72798PPz8/uLu7Y9++fdi7dy+mTJkiDit5eHjg66+/RnBwMK5cuYI2bdqgTJkyuHv3LrZs2YJFixYVeAuGNm3awMrKCp9//jksLS1x8+ZNLF26FO3atXvvHJSOHTti+/btebLFwJt5aPm9Bw4ODvD19UXHjh3h4OCADh06oGrVqkhNTcXhw4exe/duNGzYUPxMFYepqWmhn9928eJFvHr1Ch07diz2eQsjJiYGa9euRUxMDKytrQG8GW7av38/1q5dizlz5kjqp6enIzw8HJMmTfog/fvgPvRyKyq8qKgoISAgQHB0dBT09PQEfX19wdnZWRg6dKhw5coVSd38liFnZWUJM2bMEBwcHIQyZcoINjY2wuTJk4X09HRJvdylowcOHBDq1KkjyOVywdnZOc9SxfT0dGHcuHFCxYoVBX19feHzzz8Xzpw5k2dpZ2GXdOd6+vSpMGbMGKF69eqCnp6eYGBgINSvX1+YPXu2kJSUJKm7dOlSwdnZWShTpoxgaWkpBAQECAkJCZI6BS0DzW8pMfJZ0i0IgnDx4kXB3d1d0NXVFWxtbYX58+fnWZ576dIloVevXoKtra0gl8sFCwsLoX379sKFCxfynOO/y54vXbokeHl5CUZGRoKBgYHQsmVL4fTp05I6uef775Lx3KWkR48ezdPvt729pPtd/vseKBQKYc6cOYKdnZ0gl8uFunXrCnv27Mn3/Tt9+rRQv359QVdXV3Kdfn5+gqGhYb7ne7ud7OxsoWHDhkLlypXzLFFftGiRAEDYtGnTO/tf0Nf79evXwpgxYwRra2uhTJkyQrVq1YQff/xRsjw+v+svSHx8vKCjoyP07du3wDppaWmCgYGB0LlzZ0EQir+kG0C+W+4S4xs3bgienp6CkZGRUKFCBWHw4MHC1atX83z/5X497t27J7Rp00YwMDAQLC0thenTp+e79PmXX34R6tevL+jr6wvGxsZC7dq1hQkTJghPnz4t8DpWrFghNG/eXChfvrwgl8uFqlWrCuPHj8/zPZyfS5cuCQCEEydOSMpzb8+Q39aqVStBEARhw4YNQs+ePYWqVasK+vr6gp6enuDi4iJ8++23QnJycr7nU2ZJd0EKWtI9ceJEwdbWNs/nTFUACNu3bxdf79mzRwAgGBoaSjYdHR2he/fueY5fv369oKOjI8TFxZVI/z42mSAUY8o5ERGRCrRq1QrW1tZK3wSzNMnIyIC9vT0mTZpUqKxoUchkMmzfvl18Gv2mTZvg6+uL69ev51lUYGRkJE4fyNWqVSuYmJhIphtoEg4/ERHRRzdnzhw0a9YMs2bNUtsH9a5duxZlypTJs9KoJNWtWxc5OTl49uzZe+fIREdH4+jRo9i1a9cH6t2Hx0wNERFRKZaSkoKoqCgAb4KY+fPno2XLlihXrhxsbW3Rp08fnDp1CvPmzUPdunXx/PlzREREoE6dOpJFFFOnTsWaNWsQExNT4K0i1B2DGiIiolIsMjIy39sE+Pn5ITQ0FFlZWZg1axbCwsLw5MkTVKhQAZ999hlmzJiB2rVrA3izpN/Ozg79+vXD7NmzP/QlfDAMaoiIiEgj8D41REREpBEY1BAREZFG4OqnUkihUODp06cwNjZW6a3oiYio5AmCgNevX8Pa2lqp558pKz09HZmZmSppS1dXt8BHWagTBjWl0NOnT0vs+T9ERPRhPHr0SKkn1SsjPT0d+sblgew0lbRnZWWF6OhotQ9sGNSUQrm3FNd18YNMW7VPryYqLWIif/rYXSAqEa+Tk+HoYPPex0MUR2ZmJpCdBrmLH1Dc3xM5mYi7sQ6ZmZkMakj1coecZNq6DGpIY5XUgzuJSosPMn1AR6/YvycEmeZMr2VQQ0REpK5kAIobPGnQ1E0GNUREROpKpvVmK24bGkJzroSIiIg+aczUEBERqSuZTAXDT5oz/sSghoiISF1x+ElCc66EiIiIPmnM1BAREakrDj9JMKghIiJSWyoYftKgQRvNuRIiIiIqUcHBwWjYsCGMjY1hYWGBTp064fbt2+89bsuWLXB2doaenh5q166NP/74Q7JfEARMmzYNFStWhL6+Pjw9PXH37l2l+8eghoiISF3lDj8VdyukY8eOYfjw4Th79iwOHTqErKwstGnTBqmpqQUec/r0afTq1QsDBw7E5cuX0alTJ3Tq1Al///23WOeHH37A4sWLERISgnPnzsHQ0BBeXl5IT09X7u0QBEFQ6ggqccnJyTA1NYW89mA+JoE0VsL5pR+7C0QlIjk5GZblTZGUlFRijwMRf080HAuZjrxYbQnZGcg4P79I/X3+/DksLCxw7NgxNG/ePN86PXr0QGpqKvbs2SOWffbZZ3Bzc0NISAgEQYC1tTXGjRuHwMBAAEBSUhIsLS0RGhqKnj17Fro/zNQQERERkpOTJVtGRsZ7j0lKSgIAlCtXrsA6Z86cgaenp6TMy8sLZ86cAQBER0cjLi5OUsfU1BTu7u5incJiUENERKSuVDj8ZGNjA1NTU3ELDg5+56kVCgVGjx6Nzz//HLVq1SqwXlxcHCwtLSVllpaWiIuLE/fnlhVUp7C4+omIiEhdqfDme48ePZIMP8nl7x7WGj58OP7++2+cPHmyeOdXIWZqiIiI1JUKMzUmJiaS7V1BzYgRI7Bnzx4cPXoUlStXfmcXraysEB8fLymLj4+HlZWVuD+3rKA6hcWghoiIiApFEASMGDEC27dvx5EjR+Dg4PDeYxo3boyIiAhJ2aFDh9C4cWMAgIODA6ysrCR1kpOTce7cObFOYXH4iYiISF194Gc/DR8+HOvXr8fOnTthbGwsznkxNTWFvr4+AKBfv36oVKmSOCdn1KhR8PDwwLx589CuXTts3LgRFy5cwC+//PLm9DIZRo8ejVmzZqFatWpwcHDA1KlTYW1tjU6dOil1KQxqiIiI1JVMpoKgpvD3qVm+fDkAoEWLFpLytWvXwt/fHwAQExMDLa1/+9SkSROsX78e//vf/zBlyhRUq1YNO3bskEwunjBhAlJTUzFkyBAkJiaiadOm2L9/P/T09JS7FN6npvThfWroU8D71JCm+qD3qWkyGTId5X7x/5eQnY6M08El2t8PhZkaIiIidaUle7MVtw0NwaCGiIhIXX3gOTWlneZcCREREX3SmKkhIiJSV0o+kLLANjQEgxoiIiJ1xeEnCc25EiIiIvqkMVNDRESkrjj8JMGghoiISF1x+EmCQQ0REZG6YqZGQnPCMyIiIvqkMVNDRESkrjj8JMGghoiISF1x+ElCc8IzIiIi+qQxU0NERKS2VDD8pEH5DQY1RERE6orDTxKaE54RERHRJ42ZGiIiInUlk6lg9ZPmZGoY1BAREakrLumW0JwrISIiok8aMzVERETqihOFJRjUEBERqSsOP0kwqCEiIlJXzNRIaE54RkRERJ80ZmqIiIjUFYefJBjUEBERqSsOP0loTnhGREREnzRmaoiIiNSUTCaDjJkaEYMaIiIiNcWgRorDT0RERKQRmKkhIiJSV7L/34rbhoZgUENERKSmOPwkxeEnIiIi0gjM1BAREakpZmqkGNQQERGpKQY1Uhx+IiIiUlO5QU1xN2UcP34cHTp0gLW1NWQyGXbs2PHO+v7+/vmes2bNmmKdoKCgPPudnZ2Vfj8Y1BAREVGhpaamwtXVFcuWLStU/UWLFiE2NlbcHj16hHLlyqFbt26SejVr1pTUO3nypNJ94/ATERGRuvoIS7p9fHzg4+NT6PqmpqYwNTUVX+/YsQMJCQno37+/pJ6Ojg6srKyU68x/MFNDRESkplQ5/JScnCzZMjIySqTPq1evhqenJ+zs7CTld+/ehbW1NapUqQJfX1/ExMQo3TaDGiIiIoKNjY2YVTE1NUVwcLDKz/H06VPs27cPgwYNkpS7u7sjNDQU+/fvx/LlyxEdHY1mzZrh9evXSrXP4SciIiI1JZNBBauf3vzv0aNHMDExEYvlcnnx2s3HunXrULZsWXTq1ElS/vZwVp06deDu7g47Ozts3rwZAwcOLHT7DGqIiIjUlAwqWNL9/1GNiYmJJKhRNUEQsGbNGvTt2xe6urrvrFu2bFlUr14dUVFRSp2Dw09ERERU4o4dO4aoqKhCZV5SUlJw7949VKxYUalzMFNDRESkpj7GzfdSUlIkGZTo6GhcuXIF5cqVg62tLSZPnownT54gLCxMctzq1avh7u6OWrVq5WkzMDAQHTp0gJ2dHZ4+fYrp06dDW1sbvXr1UqpvDGqIiIjU1UdY0n3hwgW0bNlSfD127FgAgJ+fH0JDQxEbG5tn5VJSUhK2bduGRYsW5dvm48eP0atXL7x8+RLm5uZo2rQpzp49C3Nzc6X6xqCGiIiICq1FixYQBKHA/aGhoXnKTE1NkZaWVuAxGzduVEXXGNQQERGpLRUMPwka9OwnBjVERERqShVzaoq/eqr0YFBDRESkphjUSHFJNxEREWkEZmqIiIjU1UdY/VSaMaghIiJSUxx+kuLwExEREWkEZmqIiIjUFDM1UgxqiIiI1BSDGikOPxEREZFGYKaGiIhITTFTI8WghoiISF1xSbcEh5+IiIhIIzBTQ0REpKY4/CTFoIaIiEhNMaiRYlBDRESkphjUSHFODREREWkEZmqIiIjUFVc/STCoISIiUlMcfpLi8BMRERFpBGZq3sPe3h6jR4/G6NGjP3ZXPhlj/NugfUtXVLOzRHpGFv786z6Clu5E1MNnBR7jXMUKk79uDzdnG9hal8fk+VsRsiGyxPvasVVdTBnaDrYVy+P+o+cIWrIDh07fEPcnnF+a73HTFm3Hkt8iSrx/pH5Wbz2BNdtO4FHsKwBvPtvjB/qg9ec1S+R8giAgeMVehO04jaSUf+BepwrmTeqBqrYWeepmZGbB0/8n/H33CY7/Ngm1nSqXSJ+o8JipkfqomRp/f3/IZDJ8//33kvIdO3Z88Dc5NDQUZcuWzVN+/vx5DBky5IP25VPXpJ4jVm05jjYDfkKXEUtRRkcbvy8ZAQM93QKP0dfTxcMnLzBj6S7EvUhSST8+r1cNV3fOKHB/ozoOWDXLH7/tPAOPPt9j77Gr+O2nIahRtaJYx8l7smQb/t1vUCgU2HX0ikr6SJrH2qIspo/oiKNhE3Bk3Xg0a1AdvoG/4Oa92CK19/0vezEs6NcC9y8KO4wVm45h/uSeOLQ2EAb6uvjqm2VIz8jKU3f64p2wMjctUj+oZMggEwObIm8aNKnmow8/6enpYe7cuUhISPjYXcmXubk5DAwMPnY3PindRv6MDXvO4db9OPx99wmGzfgNNhXLwa2GTYHHXL4Rg2mLd+D3QxeRmZmdbx2ZTIYx/m1wZUcQnp6YjxPhk/DlF25F7ufXPVsg4sxNLPktAncexGNOyF5cvfUIg7t5iHWevXwt2do2r40TF+/i4ZOXRT4vaTaf5rXR5vOaqGprAUc7S0wd9iUMDeS48Hc0ACDpdRpGzgqHY+tJsG0RiC8DFuPancdFOpcgCAjZcBSBA7zQ1qMOalWrhOUz+iHuRRL2HrsqqXvo1HUcPXcTM0d1LvY1EpWUjx7UeHp6wsrKCsHBwQXWOXnyJJo1awZ9fX3Y2Nhg5MiRSE1NFffHxsaiXbt20NfXh4ODA9avXw97e3ssXLhQrDN//nzUrl0bhoaGsLGxwbBhw5CSkgIAiIyMRP/+/ZGUlCRGrkFBQQAgaad3797o0aOHpG9ZWVmoUKECwsLCAAAKhQLBwcFwcHCAvr4+XF1dsXXrVhW8U58uEyM9AEBCclqx2hnr3wY92jbC2O83oXHP2fh5w1Gs+M4PTeo5Fqm9RrUdEHn+lqTsyNmbaFjbPt/65uWM0aZpLfy280yRzkefnpwcBbYdvIC0fzLRsLYDAMB/0mo8f/UaWxYF4GjYBLg6VUanYUuQkJT6ntbyevjkJeJfJqNFI2exzNRIH/Vr2uP8Xw/EsmcvkzF6zgaEzOj3zowpfXjFztKoYPiqNPnoQY22tjbmzJmDJUuW4PHjvH9t3Lt3D97e3vjqq6/w119/YdOmTTh58iRGjBgh1unXrx+ePn2KyMhIbNu2Db/88guePZPOv9DS0sLixYtx/fp1rFu3DkeOHMGECRMAAE2aNMHChQthYmKC2NhYxMbGIjAwME9ffH19sXv3bjEYAoADBw4gLS0NnTu/+eslODgYYWFhCAkJwfXr1zFmzBj06dMHx44dU8n79amRyWQIHtsVZ6/cK3L6HQB0y+hgTP82+GZmOI6cvYmHT15iw55z2LzvPPp3blqkNi3Km+D5y9eSsuevXsOivEm+9Xu1c0dKajp2c+iJ3uN61BNUbj4Wlp+PxtjgTfj1x8FwrlIRZ67cw8XrDxH6/UDUdbFDVVsLzBzdBabG+tgZcVnp88S/TAYAmJc3lpRblDfGs//fJwgChs34Df27NEVdF7viXxyplkxFm4YoFROFO3fuDDc3N0yfPh2rV6+W7AsODoavr684UbdatWpYvHgxPDw8sHz5cjx48ACHDx/G+fPn0aBBAwDAqlWrUK1aNUk7b0/0tbe3x6xZszB06FD8/PPP0NXVhampKWQyGaysrArsp5eXFwwNDbF9+3b07dsXALB+/Xp8+eWXMDY2RkZGBubMmYPDhw+jcePGAIAqVarg5MmTWLFiBTw8PPJtNyMjAxkZGeLr5OTkwr1xn4CfJnRHjaoV4TN4QbHaqWJTAYb6cvy+dISkXLeMNv66/W8w/ejYPPHf2loyyHV1JGVb9p3H2O83FqkPvl9+hi37LyCjgOExolzV7CxxPHwyklP+wc6IyxgW9Cv2rBiFv+88Ruo/GajqOVFS/5+MLEQ/eQEAOH05Ct1H/Szuy8zKgSAI2HXk36Bn/uRe6O7TsFB9+WXTMaSkpWOMfxsVXBlRySoVQQ0AzJ07F1988UWeDMnVq1fx119/ITw8XCwTBAEKhQLR0dG4c+cOdHR0UK9ePXG/o6MjzMzMJO0cPnwYwcHBuHXrFpKTk5GdnY309HSkpaUVes6Mjo4OunfvjvDwcPTt2xepqanYuXMnNm5880suKioKaWlpaN26teS4zMxM1K1bt8B2g4ODMWNGwRNSP1U/jO8Gr2a10HbIQjx9llistgz15QCAHmOWI/Y/bWVm/RtkNPf9dxi0QS17TB/RER2GLhLLXqemi/9+9jI5z1+45uX+/Qv3bY3dqqK6vRUGTllbrOugT4NuGR1UsTEHALjVsMXlGzEI2RgJ+0oVYFXBFLtDRuU5xtRYHwBQt4YtjodPFstXbIpE7LNEBH3TSSwzL/fmc2v5/1nF5y9fw6rCvxOAn718jdrV36xsOn7hDs5fi4bl56Ml52vp9wO6eTfA8qB+xb9gKjKufpIqNUFN8+bN4eXlhcmTJ8Pf318sT0lJwddff42RI0fmOcbW1hZ37tx5b9sPHjxA+/btERAQgNmzZ6NcuXI4efIkBg4ciMzMTKUmAvv6+sLDwwPPnj3DoUOHoK+vD29vb7GvALB3715UqlRJcpxcLi+wzcmTJ2Ps2LHi6+TkZNjYFDwp9lPww/huaNfCFR2GLkLM0+JPqr0dHYf0jCzYWJrh9KWoAutFP34h/tvawgw5OQpJ2dv+vBYNj4ZOkqXjLd2dcf7agzx1+3RsjMs3YvD33SdFvgb6dCkEAZmZ2XB1roz4l8nQ0daCrXX5fOvq6+mKAREAmJkY4HVKuqQsl12l8rAsb4Jj52+Ly7OTU/7BxesPMKDrm2HZ7wO74tuh7cVj4l4k4atvlmHNnP6oX9NehVdJRcGgRqrUBDUA8P3338PNzQ1OTk5iWb169XDjxg04OuY/mdPJyQnZ2dm4fPky6tevD+BNxuTt1VQXL16EQqHAvHnzoKX1ZhrR5s2bJe3o6uoiJyfnvX1s0qQJbGxssGnTJuzbtw/dunVDmTJlAAAuLi6Qy+WIiYkpcKgpP3K5/J1Bz6fmp4nd0dWrAXoH/oKUtHRY/H82JDklXVxmujyoL2KfJ+G7ZbsAAGV0tOFU5c3QYZkyOrA2L4ta1SshNS0D0Y9fICUtA0t/i8DssV9BpqWFs1fuwcRID+6uVfE6NR0b955Tup8rNkZiz4rRGO77BQ6evI4uberDrYYtRs/ZIKlnbKiHjq3qYurC7cV5W+gTMWPpTng2qQkbKzO8TkvH1v0XcPLiXWxbMgwtGjmjYW0H+Ab+ghkjO8HR1gKxz5Nw8NTfaN/CVek5LzKZDEN7tcRPa/ajio057CqVx5yQvbCqYIp2Hq4AABurcpJjjAze/KxyqGSOSpZmedqkD0sme7MVtw1NUaqCmtq1a8PX1xeLFy8WyyZOnIjPPvsMI0aMwKBBg2BoaIgbN27g0KFDWLp0KZydneHp6YkhQ4Zg+fLlKFOmDMaNGwd9fX0x+nR0dERWVhaWLFmCDh064NSpUwgJCZGc297eHikpKYiIiICrqysMDAwKzOD07t0bISEhuHPnDo4ePSqWGxsbIzAwEGPGjIFCoUDTpk2RlJSEU6dOwcTEBH5+fiXwrmmegV2bAwD2rhgtKR8241ds2PMm+KhsVQ4KQRD3WZmb4sRbKfdv+nrim76eOHnxrjh8NDtkD14kpmCMf2vYV+qFpNf/4OrtR1iw9kCR+vnnX9EY/L9QfBvQHlOHdcD9R8/RJ5/7iXRpUx8ymQzbDlwo0nno0/IiIQUBQWGIf5EMEyM91HSshG1LhqGlew0AwOaFAZi1fDdGfPcbXiSkwKK8CZrUdYR5ufwnqL/PqH6eSPsnA2PmbEBSyj/4zLUqti4eBj15GVVeFtEHIROEt34zfGD+/v5ITEzEjh07xLIHDx7AyckJmZmZyO3a+fPn8e233+LMmTMQBAFVq1ZFjx49MGXKFABvlnQPHDgQR44cEZeHjx49Gt999x2+/vprAMCCBQvw448/IjExEc2bN4evry/69euHhIQE8aZ7AQEB2LJlC16+fInp06cjKCgo3zsK37x5Ey4uLrCzs0N0dLQkdScIAhYvXozly5fj/v37KFu2LOrVq4cpU6agefPmhXpfkpOTYWpqCnntwZBpc/kkaaaC7rZMpO6Sk5NhWd4USUlJMDEpWrBZmHOYmpqiyjdboSU3LFZbioxU3F/StUT7+6F81KCmpDx+/Bg2NjY4fPgwWrVq9bG7ozQGNfQpYFBDmuqDBjUjt0K7mEFNTkYq7i/WjKCmVA0/FdWRI0eQkpKC2rVrIzY2FhMmTIC9vX2hMyNERESk/j76zfdUISsrC1OmTEHNmjXRuXNnmJubIzIyUpzAS0REpIk+xh2Fjx8/jg4dOsDa2hoymUwyhSQ/kZGR+Z4zLi5OUm/ZsmWwt7eHnp4e3N3d8eeffyr7dmhGpsbLywteXl4fuxtEREQf1MdY/ZSamgpXV1cMGDAAXbp0KfRxt2/flgxvWVj8+yT4TZs2YezYsQgJCYG7uzsWLlwILy8v3L59W1LvfTQiqCEiIqIPw8fHBz4+PkofZ2FhIS7M+a/58+dj8ODB6N+/PwAgJCQEe/fuxZo1azBp0qRCn0Mjhp+IiIg+RVpaMpVswJvJx29vbz++RxXc3NxQsWJFtG7dGqdOnRLLMzMzcfHiRXh6er51XVrw9PTEmTPKPQCYQQ0REZGayh1+Ku4GADY2NjA1NRW34ODgd5+8kCpWrIiQkBBs27YN27Ztg42NDVq0aIFLly4BAF68eIGcnBxYWlpKjrO0tMwz7+Z9OPxEREREePTokWTOi6rudO/k5CR5UkCTJk1w7949LFiwAL/++qtKzpGLQQ0REZGaUuWzn0xMTD7YfWoaNWqEkydPAgAqVKgAbW1txMfHS+rEx8fDyspKqXY5/ERERKSmVDn89CFduXIFFStWBPDm2Yv169dHRESEuF+hUCAiIgKNGzdWql1maoiIiNTUx3hKd0pKCqKiosTX0dHRuHLlCsqVKwdbW1tMnjwZT548QVhYGABg4cKFcHBwQM2aNZGeno5Vq1bhyJEjOHjwoNjG2LFj4efnhwYNGqBRo0ZYuHAhUlNTxdVQhcWghoiIiArtwoULaNmypfh67NixAAA/Pz+EhoYiNjYWMTEx4v7MzEyMGzcOT548gYGBAerUqYPDhw9L2ujRoweeP3+OadOmIS4uDm5ubti/f3+eycPvo5HPflJ3fPYTfQr47CfSVB/y2U81J+5UybOfrs/tyGc/ERER0cfzMe4oXJpxojARERFpBGZqiIiI1JQMKpgoDM1J1TCoISIiUlMcfpLi8BMRERFpBGZqiIiI1NTHuE9NacaghoiISE1x+EmKw09ERESkEZipISIiUlMcfpJiUENERKSmOPwkxaCGiIhITTFTI8U5NURERKQRmKkhIiJSVyoYftKgGwozqCEiIlJXHH6S4vATERERaQRmaoiIiNQUVz9JMaghIiJSUxx+kuLwExEREWkEZmqIiIjUFIefpBjUEBERqSkOP0lx+ImIiIg0AjM1REREaoqZGikGNURERGqKc2qkGNQQERGpKWZqpDinhoiIiDQCMzVERERqisNPUgxqiIiI1BSHn6Q4/EREREQagZkaIiIiNSWDCoafVNKT0oFBDRERkZrSksmgVcyoprjHlyYcfiIiIiKNwEwNERGRmuLqJykGNURERGqKq5+kOPxERESkprRkqtmUcfz4cXTo0AHW1taQyWTYsWPHO+v//vvvaN26NczNzWFiYoLGjRvjwIEDkjpBQUFigJa7OTs7K/luMKghIiIiJaSmpsLV1RXLli0rVP3jx4+jdevW+OOPP3Dx4kW0bNkSHTp0wOXLlyX1atasidjYWHE7efKk0n3j8BMREZG6kqlg+EjJw318fODj41Po+gsXLpS8njNnDnbu3Indu3ejbt26YrmOjg6srKyU68x/MFNDRESkpnInChd3+5AUCgVev36NcuXKScrv3r0La2trVKlSBb6+voiJiVG6bWZqiIiICMnJyZLXcrkccrlc5ef56aefkJKSgu7du4tl7u7uCA0NhZOTE2JjYzFjxgw0a9YMf//9N4yNjQvdNjM1REREakqmov8AwMbGBqampuIWHBys8v6uX78eM2bMwObNm2FhYSGW+/j4oFu3bqhTpw68vLzwxx9/IDExEZs3b1aqfWZqiIiI1FRRVi/l1wYAPHr0CCYmJmK5qrM0GzduxKBBg7BlyxZ4enq+s27ZsmVRvXp1REVFKXUOZmqIiIgIJiYmkk2VQc2GDRvQv39/bNiwAe3atXtv/ZSUFNy7dw8VK1ZU6jzM1BAREampj3HzvZSUFEkGJTo6GleuXEG5cuVga2uLyZMn48mTJwgLCwPwZsjJz88PixYtgru7O+Li4gAA+vr6MDU1BQAEBgaiQ4cOsLOzw9OnTzF9+nRoa2ujV69eSvWtUEHNrl27Ct3gl19+qVQHiIiIqGg+xmMSLly4gJYtW4qvx44dCwDw8/NDaGgoYmNjJSuXfvnlF2RnZ2P48OEYPny4WJ5bHwAeP36MXr164eXLlzA3N0fTpk1x9uxZmJubK9W3QgU1nTp1KlRjMpkMOTk5SnWAiIiI1EeLFi0gCEKB+3MDlVyRkZHvbXPjxo3F7NUbhQpqFAqFSk5GREREqqMlk0GrmKma4h5fmhRrTk16ejr09PRU1RciIiJSAp/SLaX06qecnBzMnDkTlSpVgpGREe7fvw8AmDp1KlavXq3yDhIREVH+/vsQyKJumkLpoGb27NkIDQ3FDz/8AF1dXbG8Vq1aWLVqlUo7R0RERFRYSgc1YWFh+OWXX+Dr6wttbW2x3NXVFbdu3VJp54iIiKhg6vjsp5Kk9JyaJ0+ewNHRMU+5QqFAVlaWSjpFRERE78eJwlJKZ2pcXFxw4sSJPOVbt26VPEKciIiI6ENSOlMzbdo0+Pn54cmTJ1AoFPj9999x+/ZthIWFYc+ePSXRRyIiIsqH7P+34rahKZTO1HTs2BG7d+/G4cOHYWhoiGnTpuHmzZvYvXs3WrduXRJ9JCIionxw9ZNUke5T06xZMxw6dEjVfSEiIiIqsiLffO/ChQu4efMmgDfzbOrXr6+yThEREdH7acnebMVtQ1MoHdTkPnTq1KlTKFu2LAAgMTERTZo0wcaNG1G5cmVV95GIiIjy8TGe0l2aKT2nZtCgQcjKysLNmzfx6tUrvHr1Cjdv3oRCocCgQYNKoo9ERERE76V0pubYsWM4ffo0nJycxDInJycsWbIEzZo1U2nniIiI6N00KNFSbEoHNTY2NvneZC8nJwfW1tYq6RQRERG9H4efpJQefvrxxx/xzTff4MKFC2LZhQsXMGrUKPz0008q7RwREREVLHeicHE3TVGoTI2ZmZkkkktNTYW7uzt0dN4cnp2dDR0dHQwYMACdOnUqkY4SERERvUuhgpqFCxeWcDeIiIhIWRx+kipUUOPn51fS/SAiIiIl8TEJUkW++R4ApKenIzMzU1JmYmJSrA4RERERFYXSQU1qaiomTpyIzZs34+XLl3n25+TkqKRjRERE9G5aMhm0ijl8VNzjSxOlVz9NmDABR44cwfLlyyGXy7Fq1SrMmDED1tbWCAsLK4k+EhERUT5kMtVsmkLpTM3u3bsRFhaGFi1aoH///mjWrBkcHR1hZ2eH8PBw+Pr6lkQ/iYiIiN5J6UzNq1evUKVKFQBv5s+8evUKANC0aVMcP35ctb0jIiKiAuWufirupimUDmqqVKmC6OhoAICzszM2b94M4E0GJ/cBl0RERFTyOPwkpXRQ079/f1y9ehUAMGnSJCxbtgx6enoYM2YMxo8fr/IOEhERERWG0nNqxowZI/7b09MTt27dwsWLF+Ho6Ig6deqotHNERERUMK5+kirWfWoAwM7ODnZ2dqroCxERESlBFcNHGhTTFC6oWbx4caEbHDlyZJE7Q0RERIXHxyRIFSqoWbBgQaEak8lkDGqIiIjooyhUUJO72ok+rPsRP/CxE0REVCAtFGHFTz5taIpiz6khIiKij4PDT1KaFKARERHRJ4yZGiIiIjUlkwFaXP0kYqaGiIhITWnJVLMp4/jx4+jQoQOsra0hk8mwY8eO9x4TGRmJevXqQS6Xw9HREaGhoXnqLFu2DPb29tDT04O7uzv+/PNP5ToGBjVERESkhNTUVLi6umLZsmWFqh8dHY127dqhZcuWuHLlCkaPHo1BgwbhwIEDYp1NmzZh7NixmD59Oi5dugRXV1d4eXnh2bNnSvWtSEHNiRMn0KdPHzRu3BhPnjwBAPz66684efJkUZojIiKiIvgYD7T08fHBrFmz0Llz50LVDwkJgYODA+bNm4caNWpgxIgR6Nq1q+R2MfPnz8fgwYPRv39/uLi4ICQkBAYGBlizZo1SfVM6qNm2bRu8vLygr6+Py5cvIyMjAwCQlJSEOXPmKNscERERFZEqh5+Sk5MlW+7v9+I6c+YMPD09JWVeXl44c+YMACAzMxMXL16U1NHS0oKnp6dYp7CUDmpmzZqFkJAQrFy5EmXKlBHLP//8c1y6dEnZ5oiIiKgUsLGxgampqbgFBwerpN24uDhYWlpKyiwtLZGcnIx//vkHL168QE5OTr514uLilDqX0qufbt++jebNm+cpNzU1RWJiorLNERERURGp8tlPjx49ktzwVS6XF6/hj0DpoMbKygpRUVGwt7eXlJ88eRJVqlRRVb+IiIjoPVT5lG4TE5MSuYu9lZUV4uPjJWXx8fEwMTGBvr4+tLW1oa2tnW8dKysrpc6l9PDT4MGDMWrUKJw7dw4ymQxPnz5FeHg4AgMDERAQoGxzREREVERaKtpKUuPGjRERESEpO3ToEBo3bgwA0NXVRf369SV1FAoFIiIixDqFpXSmZtKkSVAoFGjVqhXS0tLQvHlzyOVyBAYG4ptvvlG2OSIiIlIjKSkpiIqKEl9HR0fjypUrKFeuHGxtbTF58mQ8efIEYWFhAIChQ4di6dKlmDBhAgYMGIAjR45g8+bN2Lt3r9jG2LFj4efnhwYNGqBRo0ZYuHAhUlNT0b9/f6X6pnRQI5PJ8O2332L8+PGIiopCSkoKXFxcYGRkpGxTREREVAyqnFNTWBcuXEDLli3F12PHjgUA+Pn5ITQ0FLGxsYiJiRH3Ozg4YO/evRgzZgwWLVqEypUrY9WqVfDy8hLr9OjRA8+fP8e0adMQFxcHNzc37N+/P8/k4fdeiyAIgnKXQyUtOTkZpqamePIsgU/pJo2lo817f5JmSk5OhmV5UyQlJZXYz/Dc3xPjt16C3LB4SYWM1BT82LVeifb3Q1E6U9OyZct33qjnyJEjxeoQERERUVEoHdS4ublJXmdlZeHKlSv4+++/4efnp6p+ERER0Xt8jOGn0kzpoObt2xq/LSgoCCkpKcXuEBERERVOUR5ImV8bmkJlg9p9+vRR+hkNRERERKqidKamIGfOnIGenp6qmiMiIqL3kMlQ7JvvfdLDT126dJG8FgQBsbGxuHDhAqZOnaqyjhEREdG7cU6NlNJBjampqeS1lpYWnJyc8N1336FNmzYq6xgRERGRMpQKanJyctC/f3/Url0bZmZmJdUnIiIiKgROFJZSaqKwtrY22rRpw6dxExERlQIyFf2nKZRe/VSrVi3cv3+/JPpCRERESsjN1BR30xRKBzWzZs1CYGAg9uzZg9jYWCQnJ0s2IiIioo+h0HNqvvvuO4wbNw5t27YFAHz55ZeSxyUIggCZTIacnBzV95KIiIjy4JwaqUIHNTNmzMDQoUNx9OjRkuwPERERFZJMJnvn8xgL24amKHRQk/swbw8PjxLrDBEREVFRKbWkW5OiOSIiInXH4ScppYKa6tWrvzewefXqVbE6RERERIXDOwpLKRXUzJgxI88dhYmIiIhKA6WCmp49e8LCwqKk+kJERERK0JLJiv1Ay+IeX5oUOqjhfBoiIqLShXNqpAp9873c1U9EREREpVGhMzUKhaIk+0FERETKUsFEYQ169JNyc2qIiIio9NCCDFrFjEqKe3xpwqCGiIhITXFJt5TSD7QkIiIiKo2YqSEiIlJTXP0kxaCGiIhITfE+NVIcfiIiIiKNwEwNERGRmuJEYSkGNURERGpKCyoYftKgJd0cfiIiIiKNwEwNERGRmuLwkxSDGiIiIjWlheIPuWjSkI0mXQsRERF9whjUEBERqSmZTKaSTVnLli2Dvb099PT04O7ujj///LPAui1atMj3nO3atRPr+Pv759nv7e2tdL84/ERERKSmZCj+Q7aVPX7Tpk0YO3YsQkJC4O7ujoULF8LLywu3b9+GhYVFnvq///47MjMzxdcvX76Eq6srunXrJqnn7e2NtWvXiq/lcrmSPWOmhoiISG3l3lG4uJsy5s+fj8GDB6N///5wcXFBSEgIDAwMsGbNmnzrlytXDlZWVuJ26NAhGBgY5Alq5HK5pJ6ZmZny74fSRxAREZHGSU5OlmwZGRl56mRmZuLixYvw9PQUy7S0tODp6YkzZ84U6jyrV69Gz549YWhoKCmPjIyEhYUFnJycEBAQgJcvXyp9DQxqiIiI1JismFsuGxsbmJqailtwcHCec7148QI5OTmwtLSUlFtaWiIuLu69ff3zzz/x999/Y9CgQZJyb29vhIWFISIiAnPnzsWxY8fg4+ODnJycwr0J/49zaoiIiNSUKu9T8+jRI5iYmIjlRZnT8j6rV69G7dq10ahRI0l5z549xX/Xrl0bderUQdWqVREZGYlWrVoVun1maoiIiAgmJiaSLb+gpkKFCtDW1kZ8fLykPD4+HlZWVu9sPzU1FRs3bsTAgQPf25cqVaqgQoUKiIqKUuoaGNQQERGpqQ+9pFtXVxf169dHRESEWKZQKBAREYHGjRu/89gtW7YgIyMDffr0ee95Hj9+jJcvX6JixYqF7hvAoIaIiEhtaaloU8bYsWOxcuVKrFu3Djdv3kRAQABSU1PRv39/AEC/fv0wefLkPMetXr0anTp1Qvny5SXlKSkpGD9+PM6ePYsHDx4gIiICHTt2hKOjI7y8vJTqG+fUEBERUaH16NEDz58/x7Rp0xAXFwc3Nzfs379fnDwcExMDLS1pqHT79m2cPHkSBw8ezNOetrY2/vrrL6xbtw6JiYmwtrZGmzZtMHPmTKXn9cgEQRCKfmlUEpKTk2FqaoonzxIkk7aINImONhPFpJmSk5NhWd4USUlJJfYzPPf3xNoTt2BgZFysttJSXqN/M+cS7e+HwkwNERGRmvoYdxQuzfinEhEREWkEZmqIiIjUVFEfSPnfNjQFgxoiIiI1VZTVS/m1oSkY1BAREakpZmqkNClAIyIiok8YMzVERERqiqufpBjUEBERqSlVPtBSE3D4iYiIiDQCMzVERERqSgsyaBVzAKm4x5cmDGqIiIjUFIefpDj8RERERBqBmRoiIiI1Jfv//4rbhqZgUENERKSmOPwkxeEnIiIi0gjM1BAREakpmQpWP3H4iYiIiD46Dj9JMaghIiJSUwxqpDinhoiIiDQCMzVERERqiku6pRjUEBERqSkt2ZutuG1oCg4/ERERkUZgpoaIiEhNcfhJikENERGRmuLqJykOPxEREZFGYKaGiIhITclQ/OEjDUrUMKghIiJSV1z9JMXhJyIiItIIn2xQExkZCZlMhsTExHfWs7e3x8KFCz9In0h5i8IOwfyzkfh2wbYSO4cgCPj+l72o2e5/sPEYh69GLMW9mGeSOvU6BcH8s5GSbVHYoRLrE31aFoQehFnDEZg8b6tY1v7rhTBrOEKyjQneIDnuv/vNGo7AtoMXJHU27zuPpr2DYd10DJy9p2DEd7/hVWKKuP/mvVj0m7ASdb6cBrOGI7B8/dGSvVhSikxF/2mKUj/85O/vj3Xr1gEAypQpA1tbW/Tr1w9TpkyBjk7Ru9+kSRPExsbC1NQUABAaGorRo0fnCXLOnz8PQ0PDIp+HSs7lGw8Rtv0UajpaF6udH1b+gZjYV1g6rU+++5f8ehgrNx/H0mm+sK1YHt//shc9Ri/HyQ1ToCcvI9abNKQt+nRsIr42MpAXq19EAHDp+kOEbj+FmtUq5dnn16kJJn/dXnytr1cmT51l0/qgVWMX8bWpsb7477NX7yEgKAxzxnwF72a18PR5EsYGb8So2Rvw64+DAQD/pGfCrlIFdPSsi2/n/67KSyMV4OonKbXI1Hh7eyM2NhZ3797FuHHjEBQUhB9//LFYberq6sLKygqy93w1zc3NYWBgUKxzkeqlpGVg6PQwzJ/cC6bG0q9P0us0jJ69Hs7ek+HwxXh0Hr4Ef999UqTzCIKAFZuOYWz/NvBpXgc1q1XCsul9EfciCfuO/yWpa2ggh2V5E3Ez1GdQQ8WTkpaBIdNCsWhKL5R9KxjJpa+nC8sKJuJmYpS3jqmxvqTO24H4+b+iYVuxPL7u2QJ2lSqgsVtV9O/yOS7deCjWqVfTDjNHdcZXbRpAV7fU/x38yZGpaNMUahHUyOVyWFlZwc7ODgEBAfD09MSuXbuQkJCAfv36wczMDAYGBvDx8cHdu3fF4x4+fIgOHTrAzMwMhoaGqFmzJv744w8A0uGnyMhI9O/fH0lJSZDJZJDJZAgKCgIgHX7q3bs3evToIelbVlYWKlSogLCwMACAQqFAcHAwHBwcoK+vD1dXV2zduhWkWhN/2oLWn9eERyOnPPsGTlmLFwkp2LggAIdDx6OOkw2+GrEUCUmpSp/n4dOXePYyGc0b/nseEyN91Ktph/PXHkjqLg47jOptJqFlv7lY+lsEsrNzlD4f0dvG/7AJbT6vhRbuzvnu37L/Aqp6TkTjHrMxY+lOpKVn5tPGZlT1nIhWfj/it11nIAiCuK9hHQc8iU/AwVPXIQgCnr1Mxs6IK2jdxCVPO0TqQC3Dbn19fbx8+RL+/v64e/cudu3aBRMTE0ycOBFt27bFjRs3UKZMGQwfPhyZmZk4fvw4DA0NcePGDRgZGeVpr0mTJli4cCGmTZuG27dvA0C+9Xx9fdGtWzekpKSI+w8cOIC0tDR07twZABAcHIzffvsNISEhqFatGo4fP44+ffrA3NwcHh4e+V5PRkYGMjIyxNfJycnFfo802fZDF3Ht9iMcXBOYZ9/ZK/dw6cZD3Nw3G3LdN3+RzhjZCX8c/wu7j15Bv06fK3WuZy/ffC3MyxlLys3LGYv7AGBw9+ao42SDsiYGOH8tGrOW70b8iyTMHN1F2csjAgBsO3gBV289wpF1E/Ld39WrAWwqloOVuSmu332KGUt3IurhM3HYCACmfN0OzRpWh4GeLo6cvYXAuZuQmpaBr3u2AAB85loVv8z0w8Apa5CekYXsHAW8m9XCjxN75HtOKn20IINWMcePtDQoV6NWQY0gCIiIiMCBAwfg4+ODHTt24NSpU2jS5M08hvDwcNjY2GDHjh3o1q0bYmJi8NVXX6F27doAgCpVquTbrq6uLkxNTSGTyWBlZVXg+b28vGBoaIjt27ejb9++AID169fjyy+/hLGxMTIyMjBnzhwcPnwYjRs3Fs958uRJrFixosCgJjg4GDNmzCjy+/IpeRKfgG/n/44ti4dJ0ui5rkc9Qeo/GajuNVlSnp6RhQePXwAAzly5h55jlov7srJyIAgCdh+9IpbNm9gDXb0bFrpfAb2/EP9ds1ollCmjjcDvN+F/wzqIwRVRYT2OS8Dkedvw+9IR+X7OAcC/S1Px3zUdK8Gqggk6DluC6MfP4VDZHAAwfpCPWKeOkw3S/snA4l8Pi0HNrfuxmDxvK8YP8sEXn9VA/IskTFu8A2ODN2LJVN+Su0BSGVUMHxXl+GXLluHHH39EXFwcXF1dsWTJEjRq1CjfuqGhoejfv7+kTC6XIz09XXwtCAKmT5+OlStXIjExEZ9//jmWL1+OatWqKdUvtQhq9uzZAyMjI2RlZUGhUKB3797o0qUL9uzZA3d3d7Fe+fLl4eTkhJs3bwIARo4ciYCAABw8eBCenp746quvUKdOnSL3Q0dHB927d0d4eDj69u2L1NRU7Ny5Exs3bgQAREVFIS0tDa1bt5Ycl5mZibp16xbY7uTJkzF27FjxdXJyMmxsbIrcT0129dYjPE94jVb+/86pyslR4MyVe1i99QSmfN0OluVNsOPnkXmOzZ0g6eZsg6NhE8XylZuPIfZ5EqYN/1Iss/j/zIxFeRMAwPNXr2FVwVTc//zVa9SqVrnAftavaY/sHAUexb6Co51lEa+WPlVXb8Xg+avXaNF3rliWk6PA6cv3sHLLccSfWghtbensgfq17AEA9x/9G9T8V/1a9vhx9X5kZGZBrlsGC0IPwt21Kkb29QQA1KpWCQb6crQdvADfBrSXfOaJcm3atAljx45FSEgI3N3dsXDhQnh5eeH27duwsLDI9xgTExNxJARAnvmsP/zwAxYvXox169bBwcEBU6dOhZeXF27cuAE9Pb1C900tgpqWLVti+fLl0NXVhbW1NXR0dLBr1673Hjdo0CB4eXlh7969OHjwIIKDgzFv3jx88803Re6Lr68vPDw88OzZMxw6dAj6+vrw9vYGAKSkvFkGuXfvXlSqJF2pIJcXPGlULpe/cz/9q3mD6jgePklSNnLWelSzs8A3fT0R/yIZz169ho62Fmyty+fbhr6eLqrY/PtD38zEAK9T0yVlueysy8OivAlOnL+D2tXfBDGvU//BpesP0f+tv5T/6+87j6GlJUMFM+MC6xAVpHlDJ5zaMEVSNuK731DN3hKj+rXOE9AAwLU7jwEAlu8IRK7deYyyJgZi9vCf9EzoaGtL6mj//53Y3p57Q6XYR0jVzJ8/H4MHDxazLyEhIdi7dy/WrFmDSZMm5XvMu0ZCBEHAwoUL8b///Q8dO3YEAISFhcHS0hI7duxAz549C903tQhqDA0N4ejoKCmrUaMGsrOzce7cOXH46eXLl7h9+zZcXP6d5GZjY4OhQ4di6NChmDx5MlauXJlvUKOrq4ucnPdP7GzSpAlsbGywadMm7Nu3D926dUOZMm9+QLi4uEAulyMmJqbAoSYqHiNDPdSoKl3CbaCnCzNTQ9Soag3nKhXRoJY9+k1chekjvkRVGwvEvUjCoVM30K5FHbjVsFXqfDKZDF/38MD80AOoYmMOW+s3S7qtKpjCp/mbrN/5a9G4eP0BmtavDiMDOS5ci8bURdvR1bshyppw5Rwpz9hQDy7/uVWBgb4uypkawsXRGtGPn2Pr/gto/XlNlDM1xN93n+DbBb+jSV1H1Pr/pd/7jl/D81ev0aCWPfTkZXD03C0sWHsQI/q0Etv0blYbo2avx+qtJ9DqsxqIe5mEKfO2oX5NO1Q0LwsAyMzKxu37cQCArKxsPH2eiGu3H8PQQJ7vHwL0YX3op3RnZmbi4sWLmDz53yF+LS0teHp64syZMwUel5KSAjs7OygUCtSrVw9z5sxBzZo1AQDR0dGIi4uDp6enWN/U1BTu7u44c+aM5gU1+alWrRo6duyIwYMHY8WKFTA2NsakSZNQqVIlMdIbPXo0fHx8UL16dSQkJODo0aOoUaNGvu3Z29sjJSUFERERcHV1hYGBQYFLuXv37o2QkBDcuXMHR4/+eyMqY2NjBAYGYsyYMVAoFGjatCmSkpJw6tQpmJiYwM/PT/VvBEnIZDJsnD8Us0P2YOSs9XiZkAKL8iZo7FY1z2TfwvqmryfS0jMx9vuNSE75B+51qmDTwgBxroNuGR3sOHQJP67aj8ysbNhWLIeve7ZAQK+Wqrw0IlEZHR1E/nkbyzceRdo/mahkaYYOX7ghcIDXW3W0sWrLcXy7YBsEQYBDZXPMGtMFfp3+vZdS7w6fISUtHas2H8PUhb/D1FgfzRo4IeibjmKduOdJaN7ne/H10t8isPS3CHxezxF7Voz+INdLH8Z/F6nkN4rw4sUL5OTkwNJSOqxuaWmJW7du5duuk5MT1qxZgzp16iApKQk//fQTmjRpguvXr6Ny5cqIi4sT2/hvm7n7CksmlPIco7+/PxITE7Fjx448+xISEjBq1Cjs2rULmZmZaN68OZYsWSJOLPrmm2+wb98+PH78GCYmJvD29saCBQtQvnx5REZGomXLlkhISEDZsmUBAAEBAdiyZQtevnyJ6dOnIygoCPb29hg9ejRGjx4tnvfmzZtwcXGBnZ0doqOjJWODgiBg8eLFWL58Oe7fv4+yZcuiXr16mDJlCpo3b16oa05OToapqSmePEuAiYlJkd87otJMJ58hFCJNkJycDMvypkhKSiqxn+G5vycirsTAyLh450h5nYxWbnmz2Lm/B9/29OlTVKpUCadPnxYXxADAhAkTcOzYMZw7d+6958vKykKNGjXQq1cvzJw5E6dPn8bnn3+Op0+fomLFimK97t27QyaTYdOmTYW+llKfqQkNDS1wn5mZmXh/mPwsWbKkwH0tWrTIM2a8fPlyLF++XFL24MGDPMfWqFGjwPFmmUyGUaNGYdSoUQWem4iISBVUOaXm0aNHkiAsv7meFSpUgLa2NuLj4yXl8fHx71w9/LYyZcqgbt26iIqKAgDxuPj4eElQEx8fDzc3NyWuRE1uvkdEREQly8TERLLlF9To6uqifv36iIiIEMsUCgUiIiIkmZt3ycnJwbVr18QAxsHBAVZWVpI2k5OTce7cuUK3mavUZ2qIiIioAB9h9dPYsWPh5+eHBg0aoFGjRli4cCFSU1PF1VD9+vVDpUqVEBwcDAD47rvv8Nlnn8HR0RGJiYn48ccf8fDhQwwaNOjN6WUyjB49GrNmzUK1atXEJd3W1tbo1KmTUn1jUENERKSmPvTqJwDo0aMHnj9/jmnTpiEuLg5ubm7Yv3+/ONE3JiYGWlr/DgQlJCRg8ODBiIuLg5mZGerXr4/Tp09LVipPmDABqampGDJkCBITE9G0aVPs379fqXvUAGowUfhTxInC9CngRGHSVB9yonDkX49UMlG4RR2bEu3vh8KfKkRERKQROPxERESkpj7Ws59KKwY1RERE6opRjQSHn4iIiEgjMFNDRESkpj7G6qfSjEENERGRmpLJ3mzFbUNTcPiJiIiINAIzNURERGqK84SlGNQQERGpK0Y1Ehx+IiIiIo3ATA0REZGa4uonKQY1REREaoqrn6QY1BAREakpTqmR4pwaIiIi0gjM1BAREakrpmokGNQQERGpKU4UluLwExEREWkEZmqIiIjUFFc/STGoISIiUlOcUiPF4SciIiLSCMzUEBERqSumaiQY1BAREakprn6S4vATERERaQRmaoiIiNQUVz9JMaghIiJSU5xSI8WghoiISF0xqpHgnBoiIiLSCMzUEBERqSmufpJiUENERKSuVDBRWINiGg4/ERERkWZgpoaIiEhNcZ6wFIMaIiIidcWoRoLDT0RERKQRmKkhIiJSU1z9JMVMDRERkZrKfUxCcTdlLVu2DPb29tDT04O7uzv+/PPPAuuuXLkSzZo1g5mZGczMzODp6Zmnvr+/P2QymWTz9vZWul8MaoiIiKjQNm3ahLFjx2L69Om4dOkSXF1d4eXlhWfPnuVbPzIyEr169cLRo0dx5swZ2NjYoE2bNnjy5Imknre3N2JjY8Vtw4YNSveNQQ0REZGakqloU8b8+fMxePBg9O/fHy4uLggJCYGBgQHWrFmTb/3w8HAMGzYMbm5ucHZ2xqpVq6BQKBARESGpJ5fLYWVlJW5mZmZK9oxBDRERkfpSYVSTnJws2TIyMvKcLjMzExcvXoSnp6dYpqWlBU9PT5w5c6ZQXU5LS0NWVhbKlSsnKY+MjISFhQWcnJwQEBCAly9fFvptEPui9BFERERUKshU9B8A2NjYwNTUVNyCg4PznO/FixfIycmBpaWlpNzS0hJxcXGF6vPEiRNhbW0tCYy8vb0RFhaGiIgIzJ07F8eOHYOPjw9ycnKUej+4+omIiIjw6NEjmJiYiK/lcrnKz/H9999j48aNiIyMhJ6enljes2dP8d+1a9dGnTp1ULVqVURGRqJVq1aFbp+ZGiIiIjUlgwpWP/1/WyYmJpItv6CmQoUK0NbWRnx8vKQ8Pj4eVlZW7+zrTz/9hO+//x4HDx5EnTp13lm3SpUqqFChAqKiopR5OxjUEBERqasPPVFYV1cX9evXl0zyzZ3027hx4wKP++GHHzBz5kzs378fDRo0eO95Hj9+jJcvX6JixYpK9I5BDRERESlh7NixWLlyJdatW4ebN28iICAAqamp6N+/PwCgX79+mDx5slh/7ty5mDp1KtasWQN7e3vExcUhLi4OKSkpAICUlBSMHz8eZ8+exYMHDxAREYGOHTvC0dERXl5eSvWNc2qIiIjUVFFvnvffNpTRo0cPPH/+HNOmTUNcXBzc3Nywf/9+cfJwTEwMtLT+zZksX74cmZmZ6Nq1q6Sd6dOnIygoCNra2vjrr7+wbt06JCYmwtraGm3atMHMmTOVntcjEwRBUO5yqKQlJyfD1NQUT54lSCZtEWkSHW0mikkzJScnw7K8KZKSkkrsZ3ju74kbD57DuJjneJ2cDBd78xLt74fCnypERESkETj8REREpKY+xvBTacaghoiISE0V5TEH+bWhKTj8RERERBqBmRoiIiI1xeEnKQY1REREaurtZzcVpw1NwaCGiIhIXXFSjQTn1BAREZFGYKaGiIhITTFRI8WghoiISE1xorAUh5+IiIhIIzBTQ0REpKa4+kmKQQ0REZG64qQaCQ4/ERERkUZgpoaIiEhNMVEjxaCGiIhITXH1kxSHn4iIiEgjMFNDRESktoq/+kmTBqAY1BAREakpDj9JcfiJiIiINAKDGiIiItIIHH4iIiJSUxx+kmJQQ0REpKb4mAQpDj8RERGRRmCmhoiISE1x+EmKQQ0REZGa4mMSpDj8RERERBqBmRoiIiJ1xVSNBIMaIiIiNcXVT1IcfiIiIiKNwEwNERGRmuLqJykGNURERGqKU2qkGNQQERGpK0Y1EpxTQ0RERBqBmRoiIiI1xdVPUgxqiIiI1BQnCksxqCmFBEEAALx+nfyRe0JUcnS0OfpNmul18puf3bk/y0tScnLxf0+ooo3SgkFNKfT69WsAgHNVu4/cEyIiKqrXr1/D1NS0RNrW1dWFlZUVqjnYqKQ9Kysr6OrqqqStj0kmfIhQkpSiUCjw9OlTGBsbQ6ZJecFSKjk5GTY2Nnj06BFMTEw+dneIVI6f8Q9LEAS8fv0a1tbW0NIquYxkeno6MjMzVdKWrq4u9PT0VNLWx8RMTSmkpaWFypUrf+xufHJMTEz4A580Gj/jH05JZWjepqenpxGBiCpxUJuIiIg0AoMaIiIi0ggMauiTJ5fLMX36dMjl8o/dFaISwc84fSo4UZiIiIg0AjM1REREpBEY1BAREZFGYFBDREREGoFBDZGS7O3tsXDhwo/dDaL3ioyMhEwmQ2Ji4jvr8TNNmoJBDZUq/v7+kMlk+P777yXlO3bs+OB3Vw4NDUXZsmXzlJ8/fx5Dhgz5oH0hzZb7uZfJZNDV1YWjoyO+++47ZGdnF6vdJk2aIDY2VrwRHD/TpOkY1FCpo6enh7lz5yIhIeFjdyVf5ubmMDAw+NjdIA3j7e2N2NhY3L17F+PGjUNQUBB+/PHHYrWZ+3yg9/1BwM80aQoGNVTqeHp6wsrKCsHBwQXWOXnyJJo1awZ9fX3Y2Nhg5MiRSE1NFffHxsaiXbt20NfXh4ODA9avX58nxT5//nzUrl0bhoaGsLGxwbBhw5CSkgLgTdq+f//+SEpKEv+CDgoKAiBN1ffu3Rs9evSQ9C0rKwsVKlRAWFgYgDfP8goODoaDgwP09fXh6uqKrVu3quCdIk0il8thZWUFOzs7BAQEwNPTE7t27UJCQgL69esHMzMzGBgYwMfHB3fv3hWPe/jwITp06AAzMzMYGhqiZs2a+OOPPwBIh5/4maZPAYMaKnW0tbUxZ84cLFmyBI8fP86z/969e/D29sZXX32Fv/76C5s2bcLJkycxYsQIsU6/fv3w9OlTREZGYtu2bfjll1/w7NkzSTtaWlpYvHgxrl+/jnXr1uHIkSOYMGECgDdp+4ULF8LExASxsbGIjY1FYGBgnr74+vpi9+7dYjAEAAcOHEBaWho6d+4MAAgODkZYWBhCQkJw/fp1jBkzBn369MGxY8dU8n6RZtLX10dmZib8/f1x4cIF7Nq1C2fOnIEgCGjbti2ysrIAAMOHD0dGRgaOHz+Oa9euYe7cuTAyMsrTHj/T9EkQiEoRPz8/oWPHjoIgCMJnn30mDBgwQBAEQdi+fbuQ+3EdOHCgMGTIEMlxJ06cELS0tIR//vlHuHnzpgBAOH/+vLj/7t27AgBhwYIFBZ57y5YtQvny5cXXa9euFUxNTfPUs7OzE9vJysoSKlSoIISFhYn7e/XqJfTo0UMQBEFIT08XDAwMhNOnT0vaGDhwoNCrV693vxn0yXj7c69QKIRDhw4Jcrlc6NSpkwBAOHXqlFj3xYsXgr6+vrB582ZBEAShdu3aQlBQUL7tHj16VAAgJCQkCILAzzRpPj6lm0qtuXPn4osvvsjz1+TVq1fx119/ITw8XCwTBAEKhQLR0dG4c+cOdHR0UK9ePXG/o6MjzMzMJO0cPnwYwcHBuHXrFpKTk5GdnY309HSkpaUVen6Bjo4OunfvjvDwcPTt2xepqanYuXMnNm7cCACIiopCWloaWrduLTkuMzMTdevWVer9IM22Z88eGBkZISsrCwqFAr1790aXLl2wZ88euLu7i/XKly8PJycn3Lx5EwAwcuRIBAQE4ODBg/D09MRXX32FOnXqFLkf/EyTOmNQQ6VW8+bN4eXlhcmTJ8Pf318sT0lJwddff42RI0fmOcbW1hZ37tx5b9sPHjxA+/btERAQgNmzZ6NcuXI4efIkBg4ciMzMTKUmTfr6+sLDwwPPnj3DoUOHoK+vD29vb7GvALB3715UqlRJchyfw0Nva9myJZYvXw5dXV1YW1tDR0cHu3bteu9xgwYNgpeXF/bu3YuDBw8iODgY8+bNwzfffFPkvvAzTeqKQQ2Vat9//z3c3Nzg5OQkltWrVw83btyAo6Njvsc4OTkhOzsbly9fRv369QG8+evy7dVUFy9ehEKhwLx586Cl9WZq2ebNmyXt6OrqIicn5719bNKkCWxsbLBp0ybs27cP3bp1Q5kyZQAALi4ukMvliImJgYeHh3IXT58UQ0PDPJ/pGjVqIDs7G+fOnUOTJk0AAC9fvsTt27fh4uIi1rOxscHQoUMxdOhQTJ48GStXrsw3qOFnmjQdgxoq1WrXrg1fX18sXrxYLJs4cSI+++wzjBgxAoMGDYKhoSFu3LiBQ4cOYenSpXB2doanpyeGDBmC5cuXo0yZMhg3bhz09fXFpa2Ojo7IysrCkiVL0KFDB5w6dQohISGSc9vb2yMlJQURERFwdXWFgYFBgRmc3r17IyQkBHfu3MHRo0fFcmNjYwQGBmLMmDFQKBRo2rQpkpKScOrUKZiYmMDPz68E3jXSFNWqVUPHjh0xePBgrFixAsbGxpg0aRIqVaqEjh07AgBGjx4NHx8fVK9eHQkJCTh69Chq1KiRb3v8TJPG+9iTeoje9vaEyVzR0dGCrq6u8PbH9c8//xRat24tGBkZCYaGhkKdOnWE2bNni/ufPn0q+Pj4CHK5XLCzsxPWr18vWFhYCCEhIWKd+fPnCxUrVhT09fUFLy8vISwsTDKpUhAEYejQoUL58uUFAML06dMFQZBOqsx148YNAYBgZ2cnKBQKyT6FQiEsXLhQcHJyEsqUKSOYm5sLXl5ewrFjx4r3ZpHGyO9zn+vVq1dC3759BVNTU/GzeufOHXH/iBEjhKpVqwpyuVwwNzcX+vbtK7x48UIQhLwThQWBn2nSbDJBEISPGFMRfRCPHz+GjY0NDh8+jFatWn3s7hARUQlgUEMa6ciRI0hJSUHt2rURGxuLCRMm4MmTJ7hz5444N4CIiDQL59SQRsrKysKUKVNw//59GBsbo0mTJggPD2dAQ0SkwZipISIiIo3AxyQQERGRRmBQQ0RERBqBQQ0RERFpBAY1REREpBEY1BBRvvz9/dGpUyfxdYsWLTB69OgP3o/IyEjIZDIkJiYWWEcmk2HHjh2FbjMoKAhubm7F6teDBw8gk8lw5cqVYrVDRKrDoIZIjfj7+0Mmk0Emk0FXVxeOjo747rvvkJ2dXeLn/v333zFz5sxC1S1MIEJEpGq8Tw2RmvH29sbatWuRkZGBP/74A8OHD0eZMmUwefLkPHUzMzOhq6urkvOWK1dOJe0QEZUUZmqI1IxcLoeVlRXs7OwQEBAAT09P7Nq1C8C/Q0azZ8+GtbW1+HTzR48eoXv37ihbtizKlSuHjh074sGDB2KbOTk5GDt2LMqWLYvy5ctjwoQJ+O8trP47/JSRkYGJEyfCxsYGcrkcjo6OWL16NR48eICWLVsCAMzMzCCTyeDv7w8AUCgUCA4OhoODA/T19eHq6oqtW7dKzvPHH3+gevXq0NfXR8uWLSX9LKyJEyeievXqMDAwQJUqVTB16lRkZWXlqbdixQrY2NjAwMAA3bt3R1JSkmT/qlWrUKNGDejp6cHZ2Rk///yz0n0hog+HQQ2RmtPX10dmZqb4OiIiArdv38ahQ4ewZ88eZGVlwcvLC8bGxjhx4gROnToFIyMjeHt7i8fNmzcPoaGhWLNmDU6ePIlXr15h+/bt7zxvv379sGHDBixevBg3b97EihUrYGRkBBsbG2zbtg0AcPv2bcTGxmLRokUAgODgYISFhSEkJATXr1/HmDFj0KdPHxw7dgzAm+CrS5cu6NChA65cuYJBgwZh0qRJSr8nxsbGCA0NxY0bN7Bo0SKsXLkSCxYskNSJiorC5s2bsXv3buzfvx+XL1/GsGHDxP3h4eGYNm0aZs+ejZs3b2LOnDmYOnUq1q1bp3R/iOgD+YgP0yQiJb39NGeFQiEcOnRIkMvlQmBgoLjf0tJSyMjIEI/59ddfBScnJ8mTljMyMgR9fX3hwIEDgiAIQsWKFYUffvhB3J+VlSVUrlxZ8uRoDw8PYdSoUYIgCMLt27cFAMKhQ4fy7Wd+T4dOT08XDAwMhNOnT0vqDhw4UOjVq5cgCIIwefJkwcXFRbJ/4sSJedr6LwDC9u3bC9z/448/CvXr1xdfT58+XdDW1hYeP34slu3bt0/Q0tISYmNjBUEQhKpVqwrr16+XtDNz5kyhcePGgiC8eXo8AOHy5csFnpeIPizOqSFSM3v27IGRkRGysrKgUCjQu3dvBAUFiftr164tmUdz9epVREVFwdjYWNJOeno67t27h6SkJMTGxsLd3V3cp6OjgwYNGuQZgsp15coVaGtrw8PDo9D9joqKQlpaGlq3bi0pz8zMRN26dQEAN2/elPQDABo3blzoc+TatGkTFi9ejHv37iElJQXZ2dkwMTGR1LG1tUWlSpUk51EoFLh9+zaMjY1x7949DBw4EIMHDxbrZGdnw9TUVOn+ENGHwaCGSM20bNkSy5cvh66uLqytraGjI/02NjQ0lLxOSUlB/fr1ER4enqctc3PzIvVBX19f6WNSUlIAAHv37pUEE8CbeUKqcubMGfj6+mLGjBnw8vKCqakpNm7ciHnz5ind15UrV+YJsrS1tVXWVyJSLQY1RGrG0NAQjo6Oha5fr149bNq0CRYWFnmyFbkqVqyIc+fOoXnz5gDeZCQuXryIevXq5Vu/du3aUCgUOHbsGDw9PfPsz80U5eTkiGUuLi6Qy+WIiYkpMMNTo0YNcdJzrrNnz77/It9y+vRp2NnZ4dtvvxXLHj58mKdeTEwMnj59Cmtra/E8WlpacHJygqWlJaytrXH//n34+voqdX4i+ng4UZhIw/n6+qJChQro2LEjTpw4gejoaERGRmLkyJF4/PgxAGDUqFH4/vvvsWPHDty6dQvDhg175z1m7O3t4efnhwEDBmDHjh1im5s3bwYA2NnZQSaTYc+ePXj+/DlSUlJgbGyMwMBAjBkzBuvWrcO9e/dw6dIlLFmyRJx8O3ToUNy9exfjx4/H7du3sX79eoSGhip1vdWqVUNMTAw2btyIe/fuYfHixflOetbT04Ofnx+uXr2KEydOYOTIkejevTusrKwAADNmzEBwcDAWL16MO3fu4Nq1a1i7di3mz5+vVH+I6MNhUEOk4QwMDHD8+HHY2tqiS5cuqFGjBgYOHIj09HQxczNu3Dj07dsXfn5+aNy4MYyNjdG5c+d3trt8+XJ07doVw4YNg7OzMwYPHozU1FQAQKVKlTBjxgxMmjQJlpaWGDFiBABg5syZmDp1KoKDg1GjRg14e3tj7969cHBwAPBmnsu2bduwY8cOuLq6IiQkBHPmzFHqer/88kuMGTMGI0aMgJubG06fPo2pU6fmqefo6IguXbqgbdu2aNOmDerUqSNZsj1o0CCsWrUKa9euRe3ateHh4YHQ0FCxr0RU+siEgmYCEhEREakRZmqIiIhIIzCoISIiIo3AoIaIiIg0AoMaIiIi0ggMaoiIiEgjMKghIiIijcCghoiIiDQCgxoiIiLSCAxqiIiISCMwqCEiIiKNwKCGiIiINAKDGiIiItII/wcQYcYX1J7PegAAAABJRU5ErkJggg==\n"
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Global Confusion Matrix for LSTM:\n",
            "                          Predicted\n",
            "               ------------------------------\n",
            "               Negative     | Positive\n",
            "Actual Negative    20866654        | 30247\n",
            "Actual Positive    395018        | 45581\n",
            "445/445 [==============================] - 1s 2ms/step\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 800x600 with 0 Axes>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Global Confusion Matrix for GRU:\n",
            "                          Predicted\n",
            "               ------------------------------\n",
            "               Negative     | Positive\n",
            "Actual Negative    20861387        | 35514\n",
            "Actual Positive    381353        | 59246\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Visualization"
      ],
      "metadata": {
        "id": "sHVOcX7qIBH3"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Data preparation\n",
        "models = ['Simple RNN', 'LSTM', 'GRU']\n",
        "fmax_scores = [fmax_rnn, fmax_lstm, fmax_gru]\n",
        "aupr_scores = [aupr_rnn, aupr_lstm, aupr_gru]\n",
        "\n",
        "# Plotting\n",
        "fig, ax = plt.subplots(1, 2, figsize=(14, 5))\n",
        "\n",
        "# Fmax Scores\n",
        "ax[0].bar(models, fmax_scores, color='blue')\n",
        "ax[0].set_title('Fmax Scores Comparison')\n",
        "ax[0].set_ylabel('Fmax Score')\n",
        "ax[0].set_ylim([0, 0.3])\n",
        "\n",
        "# AUC Scores\n",
        "ax[1].bar(models, aupr_scores, color='green')\n",
        "ax[1].set_title('AUPR Scores Comparison')\n",
        "ax[1].set_ylabel('AUPR Score')\n",
        "ax[1].set_ylim([0, 0.3])\n",
        "\n",
        "plt.show()"
      ],
      "metadata": {
        "id": "4sgrvspHsyKz"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Define the data for model comparison\n",
        "models = ['Simple RNN', 'LSTM', 'GRU', 'Simple RNN 2', 'LSTM 2', 'GRU 2']\n",
        "fmax_scores = [0.0641, 0.0612, 0.0620, 0.2055, 0.0948, 0.1377]\n",
        "aupr_scores = [0.0333, 0.0309, 0.0316, 0.1301, 0.0515, 0.0796]\n",
        "\n",
        "# Prepare positions on the x-axis for each model's bar set\n",
        "x = np.arange(len(models))  # the label locations\n",
        "\n",
        "# Create the bar plots\n",
        "fig, ax = plt.subplots(figsize=(12, 8))\n",
        "rects1 = ax.bar(x - 0.2, fmax_scores, 0.4, label='Fmax', color='skyblue')\n",
        "rects2 = ax.bar(x + 0.2, aupr_scores, 0.4, label='AUPR', color='orange')\n",
        "\n",
        "# Add some text for labels, title and custom x-axis tick labels\n",
        "ax.set_ylabel('Scores')\n",
        "ax.set_title('Model Performance Comparison Before and After Optimization')\n",
        "ax.set_xticks(x)\n",
        "ax.set_xticklabels(models)\n",
        "ax.legend()\n",
        "\n",
        "# Function to add labels to the bars\n",
        "def autolabel(rects):\n",
        "    \"\"\"Attach a text label above each bar in *rects*, displaying its height.\"\"\"\n",
        "    for rect in rects:\n",
        "        height = rect.get_height()\n",
        "        ax.annotate('{}'.format(round(height, 4)),\n",
        "                    xy=(rect.get_x() + rect.get_width() / 2, height),\n",
        "                    xytext=(0, 3),  # 3 points vertical offset\n",
        "                    textcoords=\"offset points\",\n",
        "                    ha='center', va='bottom')\n",
        "\n",
        "autolabel(rects1)\n",
        "autolabel(rects2)\n",
        "\n",
        "# Show the plot\n",
        "fig.tight_layout()\n",
        "\n",
        "plt.show()\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 807
        },
        "id": "H_ZeJJzlxmrb",
        "outputId": "81c78f49-d0b2-4d44-a14a-edd634893fbf"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1200x800 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAABKUAAAMWCAYAAAAgRDUeAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAACicElEQVR4nOzdeVxU9f7H8fcAAiqCCwjuuOVSKoZKmIYmiWY3KTKXSkUyK5eUsrRrotnvQm5hV3Mpt0rTbLFS0xTDbkkuKFmaXjXNUgEpBQUFhfP7owdzGwFFxDOIr+fjMY+c7/nO93zOcGaMt+f7PRbDMAwBAAAAAAAAJnKwdwEAAAAAAAC49RBKAQAAAAAAwHSEUgAAAAAAADAdoRQAAAAAAABMRygFAAAAAAAA0xFKAQAAAAAAwHSEUgAAAAAAADAdoRQAAAAAAABMRygFAAAAAAAA0xFKAQCui8Vi0aRJk675dUePHpXFYtGSJUtKvabr8d5776l58+aqUKGCqlatau9ycJMrq+f5jXbw4EF1795dHh4eslgsWr16tb1LMlWXLl3UpUuXUh0zJSVFjzzyiGrUqCGLxaLY2NhSHb+sGzx4sHx9fUt1zBvxcyrL+wWAsohQCgDKgSVLlshischisejbb78tsN0wDNWrV08Wi0UPPPCAHSosufj4eOuxWSwWVahQQY0aNdLAgQP1yy+/lOq+9u/fr8GDB6tx48Z6++23tWDBglId/1aVlJSkxx9/XPXq1ZOLi4uqV6+u4OBgLV68WLm5ufYu75bm6+tr8/lydXVV06ZNNXbsWP35558lHnfQoEH68ccf9X//939677331K5du1KsuvzJzc1V7dq1ZbFY9OWXXxbaZ8yYMdqwYYPGjx+v9957Tz169NC6detK9I8CpWHNmjXq0aOHatSoIVdXV91222164YUX9Mcff5R4zBMnTmjSpElKSkoqvULtZN++fZo0aZKOHj1q71IAoExzsncBAIDS4+rqquXLl6tTp0427Vu2bNHvv/8uFxcXO1V2/UaNGqX27dvr4sWL2rVrlxYsWKC1a9fqxx9/VO3atUtlH/Hx8crLy9OsWbPUpEmTUhnzVvfOO+/o6aeflre3t5544gk1bdpUZ8+eVVxcnCIiInTy5Em9/PLL9i7zhmnQoIHOnz+vChUq2LuUIvn5+en555+XJF24cEGJiYmKjY3Vli1btH379mse7/z580pISNA///lPjRgxorTLLZc2b96skydPytfXV8uWLVPPnj0L7dO7d2+98MIL1rbZs2drzpw5pgdTL7zwgmbMmKE2bdropZdeUvXq1bVr1y7Nnj1bK1asUFxcnJo1a3bN4544cUKTJ0+Wr6+v/Pz8bLa9/fbbysvLK6Uj+MtXX31VquP93b59+zR58mR16dKlwBVeN3K/AHCzIZQCgHLk/vvv16pVq/Tmm2/Kyel/X/HLly+Xv7+/0tLS7Fjd9encubMeeeQRSVJ4eLhuu+02jRo1SkuXLtX48eOva+zMzExVrlxZqampklSq0/aysrJUqVKlUhvvZvL999/r6aefVmBgoNatW6cqVapYt40ePVo7d+7UTz/9ZMcKb5xLly4pLy9Pzs7OcnV1tXc5V1SnTh09/vjj1udPPvmk3NzcNH36dB08eFBNmza9pvFOnTolqXQ/R/mf0fLq/fff15133qlBgwbp5ZdfLvR4U1NTTZlSbBiGLly4oIoVKxa6/YMPPtCMGTPUt29fLVu2TI6OjtZtgwcPVteuXdWnTx/t2rXL5u+h63Ujgl1nZ+dSH7Ms7xcAyiKm7wFAOdK/f3/98ccf2rhxo7UtJydHH330kQYMGFDoazIzM/X8889bp1Y1a9ZM06dPl2EYNv2ys7M1ZswYeXl5qUqVKnrwwQf1+++/Fzrm8ePHNWTIEHl7e8vFxUW33367Fi1aVHoHKunee++VJB05csTa9uWXX6pz586qXLmyqlSpol69emnv3r02rxs8eLDc3Nx0+PBh3X///apSpYoee+wx+fr6KioqSpLk5eVVYK2st956S7fffrtcXFxUu3ZtDR8+XGfOnLEZu0uXLrrjjjuUmJioe+65R5UqVdLLL79sXVdo+vTpmjNnjho1aqRKlSqpe/fu+u2332QYhqZMmaK6deuqYsWK6t27d4GpU5999pl69eql2rVry8XFRY0bN9aUKVMKTH/Lr2Hfvn3q2rWrKlWqpDp16mjq1KkF3sMLFy5o0qRJuu222+Tq6qpatWrp4Ycf1uHDh6198vLyFBsbq9tvv12urq7y9vbWsGHDdPr06av+jCZPniyLxaJly5bZBFL52rVrp8GDB1ufF/dctFgsGjFihFatWqWWLVuqYsWKCgwM1I8//ihJmj9/vpo0aSJXV1d16dKlwPSZv/+cOnbsqIoVK6phw4aaN2+eTb+cnBxNnDhR/v7+8vDwUOXKldW5c2d9/fXXNv3+/vONjY1V48aN5eLion379hW6plRycrLCw8NVt25dubi4qFatWurdu3eBOq/lnCvOz/ta+Pj4SFKBUGH//v165JFHVL16dbm6uqpdu3b6/PPPrdsnTZqkBg0aSJLGjh0ri8Vic5XI7t271bNnT7m7u8vNzU3dunXT999/b7OP/OnIW7Zs0bPPPquaNWuqbt261u3F+ZwX5s8//9QLL7ygVq1ayc3NTe7u7urZs6d++OEHm375U4Y//PBD/d///Z/q1q0rV1dXdevWTYcOHSow7oIFC9S4cWNVrFhRHTp00H/+85+r1vJ358+f16effqp+/frp0Ucf1fnz5/XZZ58VeD8Mw9CcOXOsUy0HDx6sOXPmSJLNFMx8xf3s+vr66oEHHtCGDRvUrl07VaxYUfPnzy+y3smTJ6tatWpasGCBTSAlSR06dNBLL72kH3/8UR999JG1vTifufj4eLVv317SX//wkH88+Z+dy9eUKo3v1cvXdrp8OuvfH/Hx8ZKkX3/9Vc8++6yaNWumihUrqkaNGurTp4/N53fJkiXq06ePJKlr164FxihsTanU1FRFRETI29tbrq6uatOmjZYuXWrT5+/HnH/eubi4qH379tqxY0eRPzMAKMu4UgoAyhFfX18FBgbqgw8+sE7/+PLLL5Wenq5+/frpzTfftOlvGIYefPBBff3114qIiJCfn582bNigsWPH6vjx43rjjTesfZ988km9//77GjBggDp27KjNmzerV69eBWpISUnRXXfdZQ0OvLy89OWXXyoiIkIZGRkaPXp0qRxrfnBSo0YNSX8tUD5o0CCFhITo9ddfV1ZWlubOnatOnTpp9+7dNr/MXLp0SSEhIerUqZOmT5+uSpUqafDgwXr33Xf16aefau7cuXJzc1Pr1q0l/fWL9uTJkxUcHKxnnnlGBw4c0Ny5c7Vjxw599913Nv+C/8cff6hnz57q16+fHn/8cXl7e1u3LVu2TDk5ORo5cqT+/PNPTZ06VY8++qjuvfdexcfH66WXXtKhQ4f073//Wy+88IJNkLdkyRK5ubkpMjJSbm5u2rx5syZOnKiMjAxNmzbN5r05ffq0evTooYcffliPPvqoPvroI7300ktq1aqV9bzIzc3VAw88oLi4OPXr10/PPfeczp49q40bN+qnn35S48aNJUnDhg3TkiVLFB4erlGjRunIkSOaPXu2du/eXeDY/y4rK0txcXG65557VL9+/av+PK/lXJSk//znP/r88881fPhwSVJ0dLQeeOABvfjii3rrrbf07LPP6vTp05o6daqGDBmizZs3F3iP7r//fj366KPq37+/PvzwQz3zzDNydnbWkCFDJEkZGRl655131L9/fw0dOlRnz57VwoULFRISou3btxeYXrR48WJduHBBTz31lHXtrMKmG4WFhWnv3r0aOXKkfH19lZqaqo0bN+rYsWPW8/Razrni/Lyv5OLFi9arKC9cuKDdu3dr5syZuueee9SwYUNrv7179+ruu+9WnTp1NG7cOFWuXFkffvihQkND9fHHH+uhhx7Sww8/rKpVq2rMmDHq37+/7r//frm5uVlf37lzZ7m7u+vFF19UhQoVNH/+fHXp0kVbtmxRQECATV3PPvusvLy8NHHiRGVmZkq6ts/55X755RetXr1affr0UcOGDZWSkqL58+crKChI+/btKzANOCYmRg4ODnrhhReUnp6uqVOn6rHHHtO2bdusfRYuXKhhw4apY8eOGj16tH755Rc9+OCDql69uurVq3fV916SPv/8c507d079+vWTj4+PunTpomXLlln/IeGee+7Re++9pyeeeEL33XefBg4cKElq3LixTpw4oY0bN+q9994rMO61fHYPHDig/v37a9iwYRo6dGiRU+8OHjyoAwcOaPDgwXJ3dy+0z8CBAxUVFaU1a9aoX79+1varfeZatGihV199VRMnTtRTTz2lzp07S5I6dux4xffver5XLxcbG6tz587ZtL3xxhtKSkqy/l2zY8cObd26Vf369VPdunV19OhRzZ07V126dNG+fftUqVIl3XPPPRo1apTefPNNvfzyy2rRooUkWf97ufPnz6tLly46dOiQRowYoYYNG2rVqlUaPHiwzpw5o+eee86m//Lly3X27FkNGzZMFotFU6dO1cMPP6xffvmlTE8VBoBCGQCAm97ixYsNScaOHTuM2bNnG1WqVDGysrIMwzCMPn36GF27djUMwzAaNGhg9OrVy/q61atXG5KM1157zWa8Rx55xLBYLMahQ4cMwzCMpKQkQ5Lx7LPP2vQbMGCAIcmIioqytkVERBi1atUy0tLSbPr269fP8PDwsNZ15MgRQ5KxePHiKx7b119/bUgyFi1aZJw6dco4ceKEsXbtWsPX19ewWCzGjh07jLNnzxpVq1Y1hg4davPa5ORkw8PDw6Z90KBBhiRj3LhxBfYVFRVlSDJOnTplbUtNTTWcnZ2N7t27G7m5udb22bNnW+vKFxQUZEgy5s2bZzNu/rF6eXkZZ86csbaPHz/ekGS0adPGuHjxorW9f//+hrOzs3HhwgVrW/779nfDhg0zKlWqZNMvv4Z3333X2padnW34+PgYYWFh1rZFixYZkoyZM2cWGDcvL88wDMP4z3/+Y0gyli1bZrN9/fr1hbb/3Q8//GBIMp577rki+/xdcc9FwzAMSYaLi4tx5MgRa9v8+fMNSYaPj4+RkZFhbc9/j//eN/89mjFjhrUtOzvb8PPzM2rWrGnk5OQYhmEYly5dMrKzs23qOX36tOHt7W0MGTLE2pb/83V3dzdSU1Nt+l9+np8+fdqQZEybNq3I96Ik59zVft5FadCggSGpwOPuu+8u8Bnu1q2b0apVK5vzLS8vz+jYsaPRtGnTAsd8+TGGhoYazs7OxuHDh61tJ06cMKpUqWLcc8891rb877NOnToZly5dsrZfy+e8MBcuXLB5P/NrdXFxMV599VVrW/53TosWLWx+/rNmzTIkGT/++KNhGIaRk5Nj1KxZ0/Dz87Ppt2DBAkOSERQUdMV68j3wwAPG3XffbfN6JyenAueSJGP48OE2bcOHDzcK+9/5a/ns5p8D69evv2qt+Z/TN95444r93N3djTvvvNP6vLifuR07dhT598KgQYOMBg0aWJ+XxvdqUFDQFX9OH374oSHJ5vwo7Ls4ISGhwOdw1apVhiTj66+/LtD/8v3GxsYakoz333/f2paTk2MEBgYabm5u1u+0/GOuUaOG8eeff1r7fvbZZ4Yk44svvijyWACgrGL6HgCUM/nTP9asWaOzZ89qzZo1RU7dW7dunRwdHTVq1Cib9ueff16GYVjvArVu3TpJKtDv8queDMPQxx9/rH/84x8yDENpaWnWR0hIiNLT07Vr164SHdeQIUPk5eWl2rVrq1evXsrMzNTSpUvVrl07bdy4UWfOnFH//v1t9uno6KiAgIAC060k6ZlnninWfjdt2qScnByNHj1aDg7/+2tz6NChcnd319q1a236u7i4KDw8vNCx+vTpIw8PD+vz/CtDHn/8cZtpUgEBAcrJydHx48etbX9f3+Xs2bNKS0tT586dlZWVpf3799vsx83NzWaNIGdnZ3Xo0MHmboUff/yxPD09NXLkyAJ15k8BWrVqlTw8PHTffffZvK/+/v5yc3Mr9H3Nl5GRIUmFTtsrTHHPxXzdunWzuSom/70MCwuz2Wd+++V3anRyctKwYcOsz52dnTVs2DClpqYqMTFRkuTo6Ghd+yUvL09//vmnLl26pHbt2hV6HoeFhcnLy+uKx1mxYkU5OzsrPj6+yCmQ13rOFefnfSUBAQHauHGjNm7cqDVr1uj//u//tHfvXj344IM6f/68pL+mvm3evFmPPvqo9fxLS0vTH3/8oZCQEB08eNDmfL1cbm6uvvrqK4WGhqpRo0bW9lq1amnAgAH69ttvrefM34/379PDSvI5/zsXFxfr+5mbm6s//vhDbm5uatasWaE/z/DwcJu1f/Kv3Ml/X3fu3KnU1FQ9/fTTNv0GDx5s8zm/kj/++EMbNmxQ//79rW1hYWHW6YMlda2f3YYNGyokJOSq4549e1bS1T/XVapUKfDzLM5nriSu53v1Svbt26chQ4aod+/emjBhgrX979/FFy9e1B9//KEmTZqoatWqJf77bd26dfLx8bE5DypUqKBRo0bp3Llz2rJli03/vn37qlq1atbnl5+bAHAzYfoeAJQzXl5eCg4O1vLly5WVlaXc3FzrAuGX+/XXX1W7du0Cv2DkTzH49ddfrf91cHCwTunKd/kUj1OnTunMmTNasGCBFixYUOg+8xcTv1YTJ05U586d5ejoKE9PT7Vo0cL6C8fBgwcl/W+dqctdPs3EycnJZo2aK8l/Dy4/VmdnZzVq1Mi6PV+dOnWKXMT28mls+b9IXT7NJ7/976HF3r17NWHCBG3evLnAL3vp6ek2z+vWrWuztowkVatWTXv27LE+P3z4sJo1a3bFhYgPHjyo9PR01axZs9DtV/pZ5r/n+b/EXk1xz8V81/NeSlLt2rULLCR92223Sfpr3Za77rpLkrR06VLNmDFD+/fv18WLF619/z6t7Uptl3NxcdHrr7+u559/Xt7e3rrrrrv0wAMPaODAgdZ1nK71nCvOz/tKPD09FRwcbH3eq1cvNWvWTI888ojeeecdjRw5UocOHZJhGHrllVf0yiuvFDpOamqq6tSpU+i2U6dOKSsrq9BpYS1atFBeXp5+++033X777db2y9/Pa/2cXy7/zppvvfWWjhw5YrMeW/7UrL+7/BzLDwHyz6X8n8PlC8FXqFDBJni7kpUrV+rixYtq27atzXpVAQEBWrZsmXV66rW61s9ucc5d6X9h1NU+12fPni2w7+J+5q7V9X4XFCYjI0MPP/yw6tSpo3fffdfm83X+/HlFR0dr8eLFOn78uM2ad5d/FxfXr7/+qqZNm9qE0FLxv/8uPzcB4GZCKAUA5dCAAQM0dOhQJScnq2fPnqbcsUmSdf2cxx9/XIMGDSq0T/46TdeqVatWNr84F7bf9957z/qL/d9dHrz8/YqJ0lbUHaskFVgU+Grt+b/snDlzRkFBQXJ3d9err76qxo0by9XVVbt27dJLL71UYN2iq41XXHl5eapZs6aWLVtW6PYrXRXUpEkTOTk5WRcfL20lfS+vxfvvv6/BgwcrNDRUY8eOVc2aNeXo6Kjo6GibxeDzXeln/3ejR4/WP/7xD61evVobNmzQK6+8oujoaG3evFlt27a95jpL85jzdevWTZL0zTffaOTIkdZz7IUXXijyipomTZqUeH+Fufz9vNbP+eX+9a9/6ZVXXtGQIUM0ZcoUVa9eXQ4ODho9enSha3/diPf1cvmfrbvvvrvQ7b/88kuxA66/u9bPbnHP3fyQ5EqB56+//qqMjAy1bNmymNVenxvxXTB48GCdOHFC27dvLxB2jhw5UosXL9bo0aMVGBgoDw8PWSwW9evXr9Dz6EYw49wEALMQSgFAOfTQQw9p2LBh+v7777Vy5coi+zVo0ECbNm3S2bNnba5QyZ8Oln8XrQYNGigvL896dU2+AwcO2IyXf2e+3NzcIgOkGyH/Cq6aNWuW+n7z34MDBw7Y/HKYk5OjI0eOmHKc8fHx+uOPP/TJJ5/onnvusbb//c6D16px48batm2bLl68WOTCuI0bN9amTZt09913F/uX1nyVKlXSvffeq82bN+u333676qLPxT0XS8uJEyeUmZlpc+XGf//7X0myTgv86KOP1KhRI33yySc2V0rk36XxejRu3FjPP/+8nn/+eR08eFB+fn6aMWOG3n///TJxzl26dEmSrIs+59dRoUKFEu3fy8tLlSpVKvCdIf31M3ZwcLjqOXK9n/OPPvpIXbt21cKFC23az5w5I09Pz2seL//ndPDgQZurty5evKgjR46oTZs2V3z9kSNHtHXrVo0YMUJBQUE22/Ly8vTEE09o+fLlNlPHLnf5FXL5ruezeyW33XabbrvtNq1evVqzZs0qdBrfu+++K0l64IEHbNqL85kr6njMFBMTo9WrV+uTTz5R8+bNC2z/6KOPNGjQIM2YMcPaduHChQJ3xryWY2nQoIH27NmjvLw8m38wuVHffwBQlrCmFACUQ25ubpo7d64mTZqkf/zjH0X2u//++5Wbm6vZs2fbtL/xxhuyWCzWO3fl//fyu/fFxsbaPHd0dFRYWJg+/vhj/fTTTwX2d+rUqZIczlWFhITI3d1d//rXv2ymWJXGfoODg+Xs7Kw333zT5l+hFy5cqPT09ELvQFja8v9V/O/7z8nJ0VtvvVXiMcPCwpSWllbgZ//3/Tz66KPKzc3VlClTCvS5dOlSgV/CLhcVFSXDMPTEE08UuKOVJCUmJlpveV7cc7G0XLp0yea29zk5OZo/f768vLzk7+8vqfD3fdu2bUpISCjxfrOysnThwgWbtsaNG6tKlSrKzs6WVDbOuS+++EKSrMFKzZo11aVLF82fP18nT54s0P9qnzFHR0d1795dn332mY4ePWptT0lJ0fLly9WpU6erTr+73s+5o6NjgStJVq1aVew1hi7Xrl07eXl5ad68ecrJybG2L1my5KqfDel/V0m9+OKLeuSRR2wejz76qIKCgoq80ilffsBz+f6u97N7JRMnTtTp06f19NNP20yBlP76TL/++uu64447FBYWVmC/V/vMFXU8Ztm0aZMmTJigf/7znwoNDS20T2Hn0b///e8C78W1HMv999+v5ORkm39EunTpkv7973/Lzc2tQGgJAOUJV0oBQDlV1PS5v/vHP/6hrl276p///KeOHj2qNm3a6KuvvtJnn32m0aNHW69M8PPzU//+/fXWW28pPT1dHTt2VFxcnM0aKPliYmL09ddfKyAgQEOHDlXLli31559/ateuXdq0aZP+/PPPUj9Wd3d3zZ07V0888YTuvPNO9evXT15eXjp27JjWrl2ru+++u9DwpTi8vLw0fvx4TZ48WT169NCDDz6oAwcO6K233lL79u1tFpi+UTp27Khq1app0KBBGjVqlCwWi957773rmqoxcOBAvfvuu4qMjNT27dvVuXNnZWZmatOmTXr22WfVu3dvBQUFadiwYYqOjlZSUpK6d++uChUq6ODBg1q1apVmzZpV5Hpl+XXPmTNHzz77rJo3b64nnnhCTZs21dmzZxUfH6/PP/9cr732mqTin4ulpXbt2nr99dd19OhR3XbbbVq5cqWSkpK0YMEC65VjDzzwgD755BM99NBD6tWrl44cOaJ58+apZcuWhYZsxfHf//5X3bp106OPPqqWLVvKyclJn376qVJSUtSvXz9J5p9zx48f1/vvvy/pr6Dghx9+0Pz58wsshD9nzhx16tRJrVq10tChQ9WoUSOlpKQoISFBv//+u3744Ycr7ue1117Txo0b1alTJz377LNycnLS/PnzlZ2dralTp161zuv9nD/wwAN69dVXFR4ero4dO+rHH3/UsmXLSjQ9TvrrqrHXXntNw4YN07333qu+ffvqyJEjWrx4cbHGXLZsmfz8/Iq8QuzBBx/UyJEjtWvXLt15552F9skPc0aNGqWQkBA5OjqqX79+1/3ZvZLHHntMO3bs0KxZs7Rv3z499thjqlatmnbt2qVFixapRo0a+uijjwpcgVmcz1zjxo1VtWpVzZs3T1WqVFHlypUVEBBQ7DWvrlf//v3l5eWlpk2bWj8T+e677z55e3vrgQce0HvvvScPDw+1bNlSCQkJ2rRpU4F1yfz8/OTo6KjXX39d6enpcnFx0b333lvoOl9PPfWU5s+fr8GDBysxMVG+vr766KOP9N133yk2NrbYN4wAgJuSmbf6AwDcGPm3UN+xY8cV+zVo0MDo1auXTdvZs2eNMWPGGLVr1zYqVKhgNG3a1Jg2bZqRl5dn0+/8+fPGqFGjjBo1ahiVK1c2/vGPfxi//fabIcmIioqy6ZuSkmIMHz7cqFevnlGhQgXDx8fH6Natm7FgwQJrn/xbWxd26++/y789+6pVq676Pnz99ddGSEiI4eHhYbi6uhqNGzc2Bg8ebOzcudPaZ9CgQUblypULfX1UVJQhyTh16lSBbbNnzzaaN29uVKhQwfD29jaeeeYZ4/Tp0zZ9goKCjNtvv73Aa/OPddq0acU6tsJ+nt99951x1113GRUrVjRq165tvPjii8aGDRsK3HK8qBouv526Yfx1a/N//vOfRsOGDa0/p0ceecQ4fPiwTb8FCxYY/v7+RsWKFY0qVaoYrVq1Ml588UXjxIkTBfZTmMTERGPAgAHWc6xatWpGt27djKVLlxq5ubnWfsU9FyUZw4cPt2m7lvc4/z3auXOnERgYaLi6uhoNGjQwZs+ebfPavLw841//+pfRoEEDw8XFxWjbtq2xZs2aIm9Nf/m+/74t/zxPS0szhg8fbjRv3tyoXLmy4eHhYQQEBBgffvhhgddezzlX2M+7MA0aNDAkWR8ODg5GzZo1jf79+xuHDh0q0P/w4cPGwIEDDR8fH6NChQpGnTp1jAceeMD46KOPivV+7Nq1ywgJCTHc3NyMSpUqGV27djW2bt1q0+dq32fF+ZwX5sKFC8bzzz9v1KpVy6hYsaJx9913GwkJCUZQUJARFBRkM35hn8uivrPeeusto2HDhoaLi4vRrl0745tvvikw5uUSExMNScYrr7xSZJ+jR48akowxY8YYhlH4eX/p0iVj5MiRhpeXl2GxWIzL/9e+OJ/dwv5eKI7Vq1cb9913n1GtWjXDxcXFaNKkifH8888X+v1Z3M+cYRjGZ599ZrRs2dJwcnKyeb+L+7m7lu/Vy39Of/8sXP7I/549ffq0ER4ebnh6ehpubm5GSEiIsX//fqNBgwbGoEGDbPb59ttvG40aNTIcHR1txijs/EhJSbGO6+zsbLRq1arAuXalz1ZhfxcDwM3AYhisiAcAAG4dXbp0UVpaWqFTTAGUPj5zAICisKYUAAAAAAAATEcoBQAAAAAAANMRSgEAAAAAAMB0rCkFAAAAAAAA03GlFAAAAAAAAExHKAUAAAAAAADTOdm7gJtVXl6eTpw4oSpVqshisdi7HAAAAAAAgDLBMAydPXtWtWvXloND0ddDEUqV0IkTJ1SvXj17lwEAAAAAAFAm/fbbb6pbt26R2wmlSqhKlSqS/nqD3d3d7VwNAAAAAABA2ZCRkaF69epZs5OiEEqVUP6UPXd3d0IpAAAAAACAy1xtuSMWOgcAAAAAAIDpCKUAAAAAAABgOkIpAAAAAAAAmI41pQAAAAAAQLmVl5ennJwce5dRrlSoUEGOjo7XPQ6hFAAAAAAAKJdycnJ05MgR5eXl2buUcqdq1ary8fG56mLmV0IoBQAAAAAAyh3DMHTy5Ek5OjqqXr16cnBgBaPSYBiGsrKylJqaKkmqVatWiccilAIAAAAAAOXOpUuXlJWVpdq1a6tSpUr2LqdcqVixoiQpNTVVNWvWLPFUPmJCAAAAAABQ7uTm5kqSnJ2d7VxJ+ZQf9F28eLHEYxBKAQAAAACAcut61jxC0UrjfSWUAgAAAAAAgOkIpQAAAAAAAGA6FjoHAAAAAAC3jJjdaabub1xbz2t+zeDBg7V06dIC7QcPHlSTJk1Ko6wygVAKAAAAAACgjOnRo4cWL15s0+bl5WWnam4Mpu8BAAAAAACUMS4uLvLx8bF5ODo6qkuXLho5cqRGjx6tatWqydvbW2+//bYyMzMVHh6uKlWqqEmTJvryyy+tY+Xm5ioiIkINGzZUxYoV1axZM82aNcu6/cKFC7r99tv11FNPWdsOHz6sKlWqaNGiRTfsGAmlAAAAAAAAbiJLly6Vp6entm/frpEjR+qZZ55Rnz591LFjR+3atUvdu3fXE088oaysLElSXl6e6tatq1WrVmnfvn2aOHGiXn75ZX344YeSJFdXVy1btkxLly7VZ599ptzcXD3++OO67777NGTIkBt2HIRSAAAAuC5z5syRr6+vXF1dFRAQoO3btxfZ9+2331bnzp1VrVo1VatWTcHBwQX6G4ahiRMnqlatWqpYsaKCg4N18OBBmz6+vr6yWCw2j5iYGOv2o0ePFthusVj0/fffl+7BAwBwg6xZs0Zubm7WR58+fazb2rRpowkTJqhp06YaP368XF1d5enpqaFDh6pp06aaOHGi/vjjD+3Zs0eSVKFCBU2ePFnt2rVTw4YN9dhjjyk8PNwaSkmSn5+fXnvtNT355JMaPXq0fv31V7399ts39BgJpQAAAFBiK1euVGRkpKKiorRr1y61adNGISEhSk1NLbR/fHy8+vfvr6+//loJCQmqV6+eunfvruPHj1v7TJ06VW+++abmzZunbdu2qXLlygoJCdGFCxdsxnr11Vd18uRJ62PkyJEF9rdp0yabPv7+/qX7BgAAcIN07dpVSUlJ1sebb75p3da6dWvrnx0dHVWjRg21atXK2ubt7S1JNn8fz5kzR/7+/vLy8pKbm5sWLFigY8eO2ezz+eef12233abZs2dr0aJFqlGjxo06PEmEUgAAALgOM2fO1NChQxUeHq6WLVtq3rx5qlSpUpHrTyxbtkzPPvus/Pz81Lx5c73zzjvKy8tTXFycpL+ukoqNjdWECRPUu3dvtW7dWu+++65OnDih1atX24xVpUoVm3U2KleuXGB/NWrUsOlToUKFUn8PAAC4ESpXrqwmTZpYH7Vq1bJuu/zvM4vFYtNmsVgk/TVtT5JWrFihF154QREREfrqq6+UlJSk8PBw5eTk2IyTmpqq//73v3J0dCxwlfKNQCgFAACAEsnJyVFiYqKCg4OtbQ4ODgoODlZCQkKxxsjKytLFixdVvXp1SdKRI0eUnJxsM6aHh4cCAgIKjBkTE6MaNWqobdu2mjZtmi5dulRg/AcffFA1a9ZUp06d9Pnnn5fkMAEAuOl999136tixo5599lm1bdtWTZo00eHDhwv0GzJkiFq1aqWlS5fqpZde0s8//3xD63K6oaMDAACg3EpLS1Nubq51ikA+b29v7d+/v1hjvPTSS6pdu7Y1hEpOTraOcfmY+dskadSoUbrzzjtVvXp1bd26VePHj9fJkyc1c+ZMSZKbm5tmzJihu+++Ww4ODvr4448VGhqq1atX68EHHyzxMQMAcDNq2rSp3n33XW3YsEENGzbUe++9px07dqhhw4bWPnPmzFFCQoL27NmjevXqae3atXrsscf0/fffy9nZ+YbURSgFAAAAu4iJidGKFSsUHx8vV1fXa3ptZGSk9c+tW7eWs7Ozhg0bpujoaLm4uMjT09OmT/v27XXixAlNmzaNUAoAcMsZNmyYdu/erb59+8pisah///569tln9eWXX0qS9u/fr7Fjx2rhwoWqV6+eJOmtt95S69at9corr+j111+/IXURSgEAAKBEPD095ejoqJSUFJv2lJQU+fj4XPG106dPV0xMjDZt2mSzWGv+61JSUmzWzkhJSZGfn1+R4wUEBOjSpUs6evSomjVrVmSfjRs3Xu2wAADl3Li2nvYu4aqWLFlS5Lb4+PgCbUePHi3QZhiG9c8uLi5avHixFi9ebNMnOjpaktS8eXNlZWXZbKtatWqBhdBLG2tKAQAAoEScnZ3l7+9vXaRcknXR8sDAwCJfN3XqVE2ZMkXr169Xu3btbLY1bNhQPj4+NmNmZGRo27ZtVxwzKSlJDg4Oqlmz5hX7/D3oAgAA9sWVUgAAACixyMhIDRo0SO3atVOHDh0UGxurzMxMhYeHS5IGDhyoOnXqWP8l9vXXX9fEiRO1fPly+fr6WteJcnNzk5ubmywWi0aPHq3XXntNTZs2VcOGDfXKK6+odu3aCg0NlSQlJCRo27Zt6tq1q6pUqaKEhASNGTNGjz/+uKpVqyZJWrp0qZydndW2bVtJ0ieffKJFixbpnXfeMfkdAgAARSGUAgAAQIn17dtXp06d0sSJE5WcnCw/Pz+tX7/eulD5sWPH5ODwv4vz586dq5ycHD3yyCM240RFRWnSpEmSpBdffFGZmZl66qmndObMGXXq1Enr16+3rjvl4uKiFStWaNKkScrOzlbDhg01ZswYmzWkJGnKlCn69ddf5eTkpObNm2vlypUF9gsAAOzHYvx9kiGKLSMjQx4eHkpPT5e7u7u9ywEAAAAAAH9z4cIFHTlyRA0bNrzmG2rg6q70/hY3M2FNKQAAAAAAAJiOUAoAAAAAAACmI5QCAAAAAACA6QilAAAAAAAAYDpCKQAAAAAAAJiOUAoAAAAAAACmc7J3AQAAACg7Ynan2buEMmdcW097lwAAKE3LLebub4BRopclJCSoU6dO6tGjh9auXWttj4+PV9euXXX69GlVrVrV5jW+vr4aPXq0Ro8eLUmyWP53rO7u7rrjjjs0ZcoU3XvvvZKkwYMHa+nSpZIkJycn1a1bV3369NGrr74qV1fXEtV9LbhSCgAAAAAAoIxZuHChRo4cqW+++UYnTpwo8TiLFy/WyZMn9d1338nT01MPPPCAfvnlF+v2Hj166OTJk/rll1/0xhtvaP78+YqKiiqNQ7gqQikAAAAAAIAy5Ny5c1q5cqWeeeYZ9erVS0uWLCnxWFWrVpWPj4/uuOMOzZ07V+fPn9fGjRut211cXOTj46N69eopNDRUwcHBNttvJEIpAAAAAACAMuTDDz9U8+bN1axZMz3++ONatGiRDKNk0wD/rmLFipKknJycQrf/9NNP2rp1q5ydna97X8XBmlIAAAAAAABlyMKFC/X4449L+mt6XXp6urZs2aIuXbqUeMysrCxNmDBBjo6OCgoKsravWbNGbm5uunTpkrKzs+Xg4KDZs2df7yEUC6EUAAAAAABAGXHgwAFt375dn376qaS/FiDv27evFi5cWKJQqn///nJ0dNT58+fl5eWlhQsXqnXr1tbtXbt21dy5c5WZmak33nhDTk5OCgsLK63DuSJCKQAAAAAAgDJi4cKFunTpkmrXrm1tMwxDLi4umj17ttzd3SVJ6enpBe6+d+bMGXl4eNi0vfHGGwoODpaHh4e8vLwK7K9y5cpq0qSJJGnRokVq06aNFi5cqIiIiFI+soJYUwoAAAAAAKAMuHTpkt59913NmDFDSUlJ1scPP/yg2rVr64MPPlDTpk3l4OCgxMREm9f+8ssvSk9P12233WbT7uPjoyZNmhQaSF3OwcFBL7/8siZMmKDz58+X6rEVur8bvgcAAAAAAABc1Zo1a3T69GlFRETojjvusHmEhYVp4cKFqlKlip588kk9//zz+vzzz3XkyBF98803euyxx3TXXXepY8eO11VDnz595OjoqDlz5pTSURWNUAoAAAAAAKAMWLhwoXWq3eXCwsK0c+dO7dmzR7NmzdKgQYP00ksv6fbbb9fgwYPVunVrffHFF7JYLNdVg5OTk0aMGKGpU6cqMzPzusa6GotRGvcUvAVlZGTIw8ND6enp1vmcAAAAN7uY3Wn2LqHMGdfW094lAABK4MKFCzpy5IgaNmwoV1dXe5dT7lzp/S1uZsKVUgAAAAAAADAdoRQAAAAAAABMRygFAAAAAAAA05WJUGrOnDny9fWVq6urAgICtH379iL7vv322+rcubOqVaumatWqKTg4uEB/wzA0ceJE1apVSxUrVlRwcLAOHjxo0+fPP//UY489Jnd3d1WtWlURERE6d+7cDTk+AAAAAAAA2LJ7KLVy5UpFRkYqKipKu3btUps2bRQSEqLU1NRC+8fHx6t///76+uuvlZCQoHr16ql79+46fvy4tc/UqVP15ptvat68edq2bZsqV66skJAQXbhwwdrnscce0969e7Vx40atWbNG33zzjZ566qkbfrwAAAAAAAAoA3ffCwgIUPv27TV79mxJUl5enurVq6eRI0dq3LhxV319bm6uqlWrptmzZ2vgwIEyDEO1a9fW888/rxdeeEGSlJ6eLm9vby1ZskT9+vXTzz//rJYtW2rHjh1q166dJGn9+vW6//779fvvv6t27dpX3S933wMAAOURd98riLvvAcDNKf/ucL6+vqpYsaK9yyl3srKy9Ouvv17X3fecbnSRV5KTk6PExESNHz/e2ubg4KDg4GAlJCQUa4ysrCxdvHhR1atXlyQdOXJEycnJCg4Otvbx8PBQQECAEhIS1K9fPyUkJKhq1arWQEqSgoOD5eDgoG3btumhhx4qsJ/s7GxlZ2dbn2dkZFzz8QIAAAAAAHNUqFBBFotFp06dkpeXlywWi71LKhcMw1BOTo5OnTolBwcHOTs7l3gsu4ZSaWlpys3Nlbe3t027t7e39u/fX6wxXnrpJdWuXdsaQiUnJ1vHuHzM/G3JycmqWbOmzXYnJydVr17d2udy0dHRmjx5crFqAgAAAAAA9uXo6Ki6devq999/19GjR+1dTrlTqVIl1a9fXw4OJV8Zyq6h1PWKiYnRihUrFB8fX+BSsdI2fvx4RUZGWp9nZGSoXr16N3SfAAAAAACg5Nzc3NS0aVNdvHjR3qWUK46OjnJycrruq8/sGkp5enrK0dFRKSkpNu0pKSny8fG54munT5+umJgYbdq0Sa1bt7a2578uJSVFtWrVshnTz8/P2ufyhdQvXbqkP//8s8j9uri4yMXFpdjHBgAAAAAA7M/R0VGOjo72LgOFsOvd95ydneXv76+4uDhrW15enuLi4hQYGFjk66ZOnaopU6Zo/fr1NutCSVLDhg3l4+NjM2ZGRoa2bdtmHTMwMFBnzpxRYmKitc/mzZuVl5engICA0jo8AAAAAAAAFMHu0/ciIyM1aNAgtWvXTh06dFBsbKwyMzMVHh4uSRo4cKDq1Kmj6OhoSdLrr7+uiRMnavny5fL19bWuAeXm5iY3NzdZLBaNHj1ar732mpo2baqGDRvqlVdeUe3atRUaGipJatGihXr06KGhQ4dq3rx5unjxokaMGKF+/foV6857AAAAAAAAuD52D6X69u2rU6dOaeLEiUpOTpafn5/Wr19vXaj82LFjNotmzZ07Vzk5OXrkkUdsxomKitKkSZMkSS+++KIyMzP11FNP6cyZM+rUqZPWr19vs+7UsmXLNGLECHXr1k0ODg4KCwvTm2++eeMPGAAAAAAAALIYhmHYu4ibUUZGhjw8PJSeni53d3d7lwMAAFAqYnan2buEMmdcW097lwAAwE2luJmJXdeUAgAAAAAAwK2JUAoAAAAAAACmI5QCAAAAAACA6QilAAAAAAAAYDpCKQAAAAAAAJiOUAoAAAAAAACmI5QCAAAAAACA6QilAAAAAAAAYDpCKQAAAAAAAJiOUAoAAAAAAACmI5QCAAAAAACA6QilAAAAAAAAYDpCKQAAAAAAAJiOUAoAAAAAAACmI5QCAAAAAACA6QilAAAAAAAAYDpCKQAAAAAAAJiOUAoAAAAAAACmI5QCAAAAAACA6QilAAAAAAAAYDpCKQAAAAAAAJiOUAoAAAAAAACmI5QCAAAAAACA6QilAAAAAAAAYDpCKQAAAAAAAJiOUAoAAAAAAACmI5QCAAAAAACA6QilAAAAAAAAYDpCKQAAAAAAAJiOUAoAAAAAAACmI5QCAAAAAACA6QilAAAAAAAAYDpCKQAAAAAAAJiOUAoAAAAAAACmI5QCAAAAAACA6QilAAAAAAAAYDpCKQAAAAAAAJiOUAoAAAAAAACmI5QCAAAAAACA6QilAAAAAAAAYDpCKQAAAAAAAJiOUAoAAAAAAACmI5QCAAAAAACA6QilAAAAAAAAYDpCKQAAAAAAAJiOUAoAAAAAAACmI5QCAAAAAACA6QilAAAAAAAAYDpCKQAAAAAAAJiOUAoAAAAAAACmI5QCAAAAAACA6QilAAAAAAAAYDpCKQAAAAAAAJiOUAoAAAAAAACmI5QCAAAAAACA6QilAAAAAAAAYDpCKQAAAAAAAJiOUAoAAAAAAACmI5QCAAAAAACA6QilAAAAAAAAYDpCKQAAAAAAAJiOUAoAAAAAAACmI5QCAAAAAACA6QilAAAAAAAAYDpCKQAAAAAAAJiOUAoAAAAAAACmI5QCAAAAAACA6QilAAAAAAAAYDpCKQAAAAAAAJiOUAoAAAAAAACmI5QCAAAAAACA6QilAAAAAAAAYDpCKQAAAAAAAJjO7qHUnDlz5OvrK1dXVwUEBGj79u1F9t27d6/CwsLk6+sri8Wi2NjYAn3yt13+GD58uLVPly5dCmx/+umnb8ThAQAAAAAAoBB2DaVWrlypyMhIRUVFadeuXWrTpo1CQkKUmppaaP+srCw1atRIMTEx8vHxKbTPjh07dPLkSetj48aNkqQ+ffrY9Bs6dKhNv6lTp5buwQEAAAAAAKBIdg2lZs6cqaFDhyo8PFwtW7bUvHnzVKlSJS1atKjQ/u3bt9e0adPUr18/ubi4FNrHy8tLPj4+1seaNWvUuHFjBQUF2fSrVKmSTT93d/dSPz4AAAAAAAAUzm6hVE5OjhITExUcHPy/YhwcFBwcrISEhFLbx/vvv68hQ4bIYrHYbFu2bJk8PT11xx13aPz48crKyrriWNnZ2crIyLB5AAAAAAAAoGSc7LXjtLQ05ebmytvb26bd29tb+/fvL5V9rF69WmfOnNHgwYNt2gcMGKAGDRqodu3a2rNnj1566SUdOHBAn3zySZFjRUdHa/LkyaVSFwAAAAAAwK3ObqGUGRYuXKiePXuqdu3aNu1PPfWU9c+tWrVSrVq11K1bNx0+fFiNGzcudKzx48crMjLS+jwjI0P16tW7MYUDAAAAAACUc3YLpTw9PeXo6KiUlBSb9pSUlCIXMb8Wv/76qzZt2nTFq5/yBQQESJIOHTpUZCjl4uJS5DpWAAAAAAAAuDZ2W1PK2dlZ/v7+iouLs7bl5eUpLi5OgYGB1z3+4sWLVbNmTfXq1euqfZOSkiRJtWrVuu79AgAAAAAA4OrsOn0vMjJSgwYNUrt27dShQwfFxsYqMzNT4eHhkqSBAweqTp06io6OlvTXwuX79u2z/vn48eNKSkqSm5ubmjRpYh03Ly9Pixcv1qBBg+TkZHuIhw8f1vLly3X//ferRo0a2rNnj8aMGaN77rlHrVu3NunIAQAAAAAAbm12DaX69u2rU6dOaeLEiUpOTpafn5/Wr19vXfz82LFjcnD438VcJ06cUNu2ba3Pp0+frunTpysoKEjx8fHW9k2bNunYsWMaMmRIgX06Oztr06ZN1gCsXr16CgsL04QJE27cgQIAAAAAAMCGxTAMw95F3IwyMjLk4eGh9PR0ubu727scAACAUhGzO83eJZQ549p62rsEAABuKsXNTOy2phQAAAAAAABuXYRSAAAAAAAAMB2hFAAAAAAAAExHKAUAAAAAAADTEUoBAAAAAADAdIRSAAAAAAAAMB2hFAAAAAAAAExHKAUAAAAAAADTEUoBAAAAAADAdIRSAAAAAAAAMB2hFAAAAAAAAExHKAUAAAAAAADTEUoBAAAAAADAdIRSAAAAAAAAMB2hFAAAAAAAAExHKAUAAAAAAADTEUoBAAAAAADAdIRSAAAAAAAAMB2hFAAAAAAAAExHKAUAAAAAAADTEUoBAAAAAADAdIRSAAAAAAAAMB2hFAAAAAAAAExHKAUAAAAAAADTEUoBAAAAAADAdIRSAAAAAAAAMB2hFAAAAAAAAExHKAUAAAAAAADTEUoBAAAAAADAdIRSAAAAAAAAMB2hFAAAAAAAAExHKAUAAAAAAADTEUoBAAAAAADAdIRSAAAAAAAAMB2hFAAAAAAAAExHKAUAAAAAAADTEUoBAAAAAADAdIRSAAAAAAAAMB2hFAAAAAAAAExHKAUAAAAAAADTEUoBAAAAAADAdIRSAAAAAAAAMB2hFAAAAAAAAExHKAUAAAAAAADTEUoBAAAAAADAdIRSAAAAAAAAMB2hFAAAAAAAAExHKAUAAAAAAADTEUoBAAAAAADAdIRSAAAAAAAAMB2hFAAAAAAAAExHKAUAAAAAAADTEUoBAAAAAADAdIRSAAAAAAAAMB2hFAAAAAAAAExHKAUAAAAAAADTEUoBAAAAAADAdIRSAAAAAAAAMB2hFAAAAAAAAExHKAUAAAAAAADTEUoBAAAAAADAdIRSAAAAAAAAMB2hFAAAAAAAAExHKAUAAAAAAADTEUoBAAAAAADAdIRSAAAAAAAAMB2hFAAAAAAAAExHKAUAAAAAAADTEUoBAAAAAADAdIRSAAAAAAAAMB2hFAAAAAAAAExHKAUAAAAAAADTEUoBAAAAAADAdHYPpebMmSNfX1+5uroqICBA27dvL7Lv3r17FRYWJl9fX1ksFsXGxhboM2nSJFksFptH8+bNbfpcuHBBw4cPV40aNeTm5qawsDClpKSU9qEBAAAAAACgCHYNpVauXKnIyEhFRUVp165datOmjUJCQpSamlpo/6ysLDVq1EgxMTHy8fEpctzbb79dJ0+etD6+/fZbm+1jxozRF198oVWrVmnLli06ceKEHn744VI9NgAAAAAAABTNrqHUzJkzNXToUIWHh6tly5aaN2+eKlWqpEWLFhXav3379po2bZr69esnFxeXIsd1cnKSj4+P9eHp6Wndlp6eroULF2rmzJm699575e/vr8WLF2vr1q36/vvvS/0YAQAAAAAAUJDdQqmcnBwlJiYqODj4f8U4OCg4OFgJCQnXNfbBgwdVu3ZtNWrUSI899piOHTtm3ZaYmKiLFy/a7Ld58+aqX7/+FfebnZ2tjIwMmwcAAAAAAABKxm6hVFpamnJzc+Xt7W3T7u3treTk5BKPGxAQoCVLlmj9+vWaO3eujhw5os6dO+vs2bOSpOTkZDk7O6tq1arXtN/o6Gh5eHhYH/Xq1StxjQAAAAAAALc6uy90Xtp69uypPn36qHXr1goJCdG6det05swZffjhh9c17vjx45Wenm59/Pbbb6VUMQAAAAAAwK3HyV479vT0lKOjY4G73qWkpFxxEfNrVbVqVd122206dOiQJMnHx0c5OTk6c+aMzdVSV9uvi4vLFdexAgAAAAAAQPHZ7UopZ2dn+fv7Ky4uztqWl5enuLg4BQYGltp+zp07p8OHD6tWrVqSJH9/f1WoUMFmvwcOHNCxY8dKdb8AAAAAAAAomt2ulJKkyMhIDRo0SO3atVOHDh0UGxurzMxMhYeHS5IGDhyoOnXqKDo6WtJfi6Pv27fP+ufjx48rKSlJbm5uatKkiSTphRde0D/+8Q81aNBAJ06cUFRUlBwdHdW/f39JkoeHhyIiIhQZGanq1avL3d1dI0eOVGBgoO666y47vAsAAAAAAAC3HruGUn379tWpU6c0ceJEJScny8/PT+vXr7cufn7s2DE5OPzvYq4TJ06obdu21ufTp0/X9OnTFRQUpPj4eEnS77//rv79++uPP/6Ql5eXOnXqpO+//15eXl7W173xxhtycHBQWFiYsrOzFRISorfeesucgwYAAAAAAIAshmEY9i7iZpSRkSEPDw+lp6fL3d3d3uUAAACUipjdafYuocwZ19bT3iUAAHBTKW5mUu7uvgcAAAAAAICyj1AKAAAAAAAApiOUAgAAAAAAgOkIpQAAAAAAAGA6QikAAAAAAACYjlAKAAAAAAAApiOUAgAAAAAAgOkIpQAAAAAAAGA6QikAAAAAAACYjlAKAAAAAAAApiOUAgAAAAAAgOkIpQAAAAAAAGA6QikAAAAAAACYjlAKAAAAAAAApiOUAgAAAAAAgOkIpQAAAAAAAGA6QikAAAAAAACYjlAKAAAAAIAyas6cOfL19ZWrq6sCAgK0ffv2Ivvu3btXYWFh8vX1lcViUWxsbIE+c+fOVevWreXu7i53d3cFBgbqyy+/tG4/evSoLBZLoY9Vq1ZJkpYsWVJkn9TU1FJ/D1B+EUoBAAAAAFAGrVy5UpGRkYqKitKuXbvUpk0bhYSEFBn8ZGVlqVGjRoqJiZGPj0+hferWrauYmBglJiZq586duvfee9W7d2/t3btXklSvXj2dPHnS5jF58mS5ubmpZ8+ekqS+ffsW6BMSEqKgoCDVrFnzxrwZKJcshmEY9i7iZpSRkSEPDw+lp6fL3d3d3uUAAACUipjdafYuocwZ19bT3iUAuEUFBASoffv2mj17tiQpLy9P9erV08iRIzVu3LgrvtbX11ejR4/W6NGjr7qf6tWra9q0aYqIiCh0e9u2bXXnnXdq4cKFhW4/deqU6tSpo4ULF+qJJ5646v5Q/hU3M+FKKQAAAAAAypicnBwlJiYqODjY2ubg4KDg4GAlJCSUyj5yc3O1YsUKZWZmKjAwsNA+iYmJSkpKKjKwkqR3331XlSpV0iOPPFIqdeHW4WTvAgAAAAAAgK20tDTl5ubK29vbpt3b21v79++/rrF//PFHBQYG6sKFC3Jzc9Onn36qli1bFtp34cKFatGihTp27FjkeAsXLtSAAQNUsWLF66oLtx6ulAIAAAAA4BbSrFkzJSUladu2bXrmmWc0aNAg7du3r0C/8+fPa/ny5Ve8SiohIUE///zzFfsAReFKKQAAAAAAyhhPT085OjoqJSXFpj0lJaXIRcyLy9nZWU2aNJEk+fv7a8eOHZo1a5bmz59v0++jjz5SVlaWBg4cWORY77zzjvz8/OTv739dNeHWxJVSAAAAAACUMc7OzvL391dcXJy1LS8vT3FxcUWu/1RSeXl5ys7OLtC+cOFCPfjgg/Ly8ir0defOndOHH37IVVIoMa6UAgAAAACgDIqMjNSgQYPUrl07dejQQbGxscrMzFR4eLgkaeDAgapTp46io6Ml/bU4ev40vJycHB0/flxJSUlyc3OzXhk1fvx49ezZU/Xr19fZs2e1fPlyxcfHa8OGDTb7PnTokL755hutW7euyPpWrlypS5cu6fHHH78Rh49bAKEUAAAAAABlUN++fXXq1ClNnDhRycnJ8vPz0/r1662Lnx87dkwODv+bAHXixAm1bdvW+nz69OmaPn26goKCFB8fL0lKTU3VwIEDdfLkSXl4eKh169basGGD7rvvPpt9L1q0SHXr1lX37t2LrG/hwoV6+OGHVbVq1dI7aNxSLIZhGPYu4maUkZEhDw8Ppaeny93d3d7lAAAAlIqY3Wn2LqHMGdfW094lAABwUyluZsKaUgAAAAAAADAdoRQAAAAAAABMRygFAAAAAAAA0xFKAQAAAAAAwHSEUgAAAAAAADAdoRQAAAAAAABM52TvAgAAAAAAuBXE7E6zdwllzri2nvYuAXbElVIAAAAAAAAwHaEUAAAAAAAATEcoBQAAAAAAANMRSgEAAAAAAMB0hFIAAAAAAAAwHaEUAAAAAAAATEcoBQAAAAAAANMRSgEAAAAAAMB0hFIAAABAKZszZ458fX3l6uqqgIAAbd++vci+e/fuVVhYmHx9fWWxWBQbG1ugz9y5c9W6dWu5u7vL3d1dgYGB+vLLL236XLhwQcOHD1eNGjXk5uamsLAwpaSk2PQZNWqU/P395eLiIj8/v9I4VAAASoxQCgAAAChFK1euVGRkpKKiorRr1y61adNGISEhSk1NLbR/VlaWGjVqpJiYGPn4+BTap27duoqJiVFiYqJ27type++9V71799bevXutfcaMGaMvvvhCq1at0pYtW3TixAk9/PDDBcYaMmSI+vbtWzoHCwDAdbAYhmHYu4ibUUZGhjw8PJSeni53d3d7lwMAAFAqYnan2buEMmdcW89r6h8QEKD27dtr9uzZkqS8vDzVq1dPI0eO1Lhx4674Wl9fX40ePVqjR4++6n6qV6+uadOmKSIiQunp6fLy8tLy5cv1yCOPSJL279+vFi1aKCEhQXfddZfNaydNmqTVq1crKSnpmo4NwPXhO7aga/2Oxc2huJkJV0oBAAAApSQnJ0eJiYkKDg62tjk4OCg4OFgJCQmlso/c3FytWLFCmZmZCgwMlCQlJibq4sWLNvtt3ry56tevX2r7BQCgtDnZuwAAAACgvEhLS1Nubq68vb1t2r29vbV///7rGvvHH39UYGCgLly4IDc3N3366adq2bKlJCk5OVnOzs6qWrVqgf0mJydf134BALhRuFIKAAAAuAk0a9ZMSUlJ2rZtm5555hkNGjRI+/bts3dZAACUGKEUAAAAUEo8PT3l6OhY4K53KSkpRS5iXlzOzs5q0qSJ/P39FR0drTZt2mjWrFmSJB8fH+Xk5OjMmTOlvl8AAG4UQikAAACglDg7O8vf319xcXHWtry8PMXFxVnXfyoteXl5ys7OliT5+/urQoUKNvs9cOCAjh07Vur7BQCgtLCmFAAAAFCKIiMjNWjQILVr104dOnRQbGysMjMzFR4eLkkaOHCg6tSpo+joaEl/LY6ePw0vJydHx48fV1JSktzc3NSkSRNJ0vjx49WzZ0/Vr19fZ8+e1fLlyxUfH68NGzZIkjw8PBQREaHIyEhVr15d7u7uGjlypAIDA23uvHfo0CGdO3dOycnJOn/+vPXuey1btpSzs7NZbxEAAJIIpQAAAIBS1bdvX506dUoTJ05UcnKy/Pz8tH79euvi58eOHZODw/8mLJw4cUJt27a1Pp8+fbqmT5+uoKAgxcfHS5JSU1M1cOBAnTx5Uh4eHmrdurU2bNig++67z/q6N954Qw4ODgoLC1N2drZCQkL01ltv2dT25JNPasuWLdbn+fs9cuSIfH19S/utAADgiiyGYRj2LuJmlJGRIQ8PD6Wnp8vd3d3e5QAAAJSKmN1p9i6hzBnX1tPeJQAoJ/iOLYjv2PKpuJkJa0oBAAAAAADAdIRSAAAAAAAAMB2hFAAAAAAAAExHKAUAAAAAAADTEUoBAAAAAADAdIRSAAAAAAAAMJ2TvQsAAAAAyrTlFntXUPYMMOxdAQCgHOBKKQAAAAAAAJiOUAoAAAAAAACmI5QCAAAAAACA6QilAAAAAAAAYDpCKQAAAAAAAJiOUAoAAAAAAACmI5QCAAAAAACA6QilAAAAAAAAYDpCKQAAAAAAAJiOUAoAAAAAAACmI5QCAAAAAACA6eweSs2ZM0e+vr5ydXVVQECAtm/fXmTfvXv3KiwsTL6+vrJYLIqNjS3QJzo6Wu3bt1eVKlVUs2ZNhYaG6sCBAzZ9unTpIovFYvN4+umnS/vQAAAAAAAAUAS7hlIrV65UZGSkoqKitGvXLrVp00YhISFKTU0ttH9WVpYaNWqkmJgY+fj4FNpny5YtGj58uL7//ntt3LhRFy9eVPfu3ZWZmWnTb+jQoTp58qT1MXXq1FI/PgAAAAAAABTOyZ47nzlzpoYOHarw8HBJ0rx587R27VotWrRI48aNK9C/ffv2at++vSQVul2S1q9fb/N8yZIlqlmzphITE3XPPfdY2ytVqlRksAUAAAAAAIAby25XSuXk5CgxMVHBwcH/K8bBQcHBwUpISCi1/aSnp0uSqlevbtO+bNkyeXp66o477tD48eOVlZVVavsEAAAAAADAldntSqm0tDTl5ubK29vbpt3b21v79+8vlX3k5eVp9OjRuvvuu3XHHXdY2wcMGKAGDRqodu3a2rNnj1566SUdOHBAn3zySZFjZWdnKzs72/o8IyOjVGoEAAAAAAC4Fdl1+t6NNnz4cP3000/69ttvbdqfeuop659btWqlWrVqqVu3bjp8+LAaN25c6FjR0dGaPHnyDa0XAAAAAADgVmG36Xuenp5ydHRUSkqKTXtKSkqprPU0YsQIrVmzRl9//bXq1q17xb4BAQGSpEOHDhXZZ/z48UpPT7c+fvvtt+uuEQAAAAAA4FZlt1DK2dlZ/v7+iouLs7bl5eUpLi5OgYGBJR7XMAyNGDFCn376qTZv3qyGDRte9TVJSUmSpFq1ahXZx8XFRe7u7jYPAAAAAAAAlIxdp+9FRkZq0KBBateunTp06KDY2FhlZmZa78Y3cOBA1alTR9HR0ZL+Whx937591j8fP35cSUlJcnNzU5MmTST9NWVv+fLl+uyzz1SlShUlJydLkjw8PFSxYkUdPnxYy5cv1/33368aNWpoz549GjNmjO655x61bt3aDu8CAAAAAADArceuoVTfvn116tQpTZw4UcnJyfLz89P69euti58fO3ZMDg7/u5jrxIkTatu2rfX59OnTNX36dAUFBSk+Pl6SNHfuXElSly5dbPa1ePFiDR48WM7Oztq0aZM1AKtXr57CwsI0YcKEG3uwAAAAAAAAsLIYhmHYu4ibUUZGhjw8PJSens5UPgAAUG7E7E6zdwllzrifvexdQtkzgF8hgJLgO7agcW097V0CboDiZiZ2W1MKAAAAAAAAty5CKQAAAAAAAJiOUAoAAAAAAACmI5QCAAAAAACA6QilAAAAAAAAYDpCKQAAAAAAAJiOUAoAAAAAAACmI5QCAAAAAACA6QilAAAAAAAAYDpCKQAAAAAAAJiOUAoAAAAAAACmI5QCAAAAAACA6QilAAAAAAAAYDpCKQAAAAAAAJiOUAoAAAAAAACmI5QCAAAAAACA6QilAAAAAAAAYDpCKQAAAAAAAJiOUAoAAAAAAACmI5QCAAAAAACA6QilAAAAAAAAYDpCKQAAAAAAAJiOUAoAAAAAAACmI5QCAAAAAACA6QilAAAAAAAAYDpCKQAAAAAAAJiOUAoAAAAAAACmI5QCAAAAAACA6QilAAAAAAAAYDpCKQAAAAAAAJiuVEKpjIwMrV69Wj///HNpDAcAAAAAAIByrkSh1KOPPqrZs2dLks6fP6927drp0UcfVevWrfXxxx+XaoEAAAAAAAAof0oUSn3zzTfq3LmzJOnTTz+VYRg6c+aM3nzzTb322mulWiAAAAAAAADKnxKFUunp6apevbokaf369QoLC1OlSpXUq1cvHTx4sFQLBAAAAAAAQPlTolCqXr16SkhIUGZmptavX6/u3btLkk6fPi1XV9dSLRAAAAAAAADlj1NJXjR69Gg99thjcnNzU/369dWlSxdJf03ra9WqVWnWBwAAAAAAgHKoRKHUs88+qw4dOui3337TfffdJweHvy64atSoEWtKAQAAAAAA4KpKFEpJUrt27dS6dWsdOXJEjRs3lpOTk3r16lWatQEAAAAAAKCcKtGaUllZWYqIiFClSpV0++2369ixY5KkkSNHKiYmplQLBAAAAAAAQPlTolBq/Pjx+uGHHxQfH2+zsHlwcLBWrlxZasUBAAAAAACgfCrR9L3Vq1dr5cqVuuuuu2SxWKztt99+uw4fPlxqxQEAAAAAAKB8KtGVUqdOnVLNmjULtGdmZtqEVAAAAAAAAEBhShRKtWvXTmvXrrU+zw+i3nnnHQUGBpZOZQAAAAAAACi3SjR971//+pd69uypffv26dKlS5o1a5b27dunrVu3asuWLaVdIwAAAAAAAMqZEl0p1alTJ/3www+6dOmSWrVqpa+++ko1a9ZUQkKC/P39S7tGAAAAAAAAlDPXfKXUxYsXNWzYML3yyit6++23b0RNAAAAAAAAKOeu+UqpChUq6OOPP74RtQAAAAAAAOAWUaLpe6GhoVq9enUplwIAAAAAAIBbRYkWOm/atKleffVVfffdd/L391flypVtto8aNapUigMAAAAAAED5VKJQauHChapataoSExOVmJhos81isRBKAQAAAAAA4IpKFEodOXKktOsAAAAAAADALaREa0r9nWEYMgyjNGoBAAAAAADALaLEodS7776rVq1aqWLFiqpYsaJat26t9957rzRrAwAAAAAAQDlVoul7M2fO1CuvvKIRI0bo7rvvliR9++23evrpp5WWlqYxY8aUapEAAAAAAAAoX0oUSv373//W3LlzNXDgQGvbgw8+qNtvv12TJk0ilAIAAAAAAMAVlWj63smTJ9WxY8cC7R07dtTJkyevuygAAAAAAACUbyUKpZo0aaIPP/ywQPvKlSvVtGnT6y4KAAAAAAAA5VuJpu9NnjxZffv21TfffGNdU+q7775TXFxcoWEVAAAAAAAA8HclulIqLCxM27Ztk6enp1avXq3Vq1fL09NT27dv10MPPVTaNQIAAAAAAKCcKdGVUpLk7++v999/vzRrAQAAAAAAwC2iRFdKrVu3Ths2bCjQvmHDBn355ZfXXRQAAAAAAADKtxKFUuPGjVNubm6BdsMwNG7cuOsuCgAAAAAAAOVbiUKpgwcPqmXLlgXamzdvrkOHDl13UQAAAAAAACjfShRKeXh46JdffinQfujQIVWuXPm6iwIAAAAAAED5VqJQqnfv3ho9erQOHz5sbTt06JCef/55Pfjgg6VWHAAAAAAAAMqnEoVSU6dOVeXKldW8eXM1bNhQDRs2VPPmzVWjRg1Nnz69tGsEAAAAAABAOeNUkhd5eHho69at2rhxo3744QdVrFhRbdq0UefOnUu7PgAAAAAAAJRD13SlVEJCgtasWSNJslgs6t69u2rWrKnp06crLCxMTz31lLKzs29IoQAAAAAAACg/rimUevXVV7V3717r8x9//FFDhw7Vfffdp3HjxumLL75QdHR0qRcJAAAAAACA8uWaQqmkpCR169bN+nzFihXq0KGD3n77bUVGRurNN9/Uhx9+WOpFAgAAAAAAoHy5plDq9OnT8vb2tj7fsmWLevbsaX3evn17/fbbb6VXHQAAAAAAAMqlawqlvL29deTIEUlSTk6Odu3apbvuusu6/ezZs6pQoULpVggAAAAAAIBy55pCqfvvv1/jxo3Tf/7zH40fP16VKlWyuePenj171Lhx41IvEgAAAAAAAOXLNYVSU6ZMkZOTk4KCgvT222/r7bfflrOzs3X7okWL1L1792sqYM6cOfL19ZWrq6sCAgK0ffv2Ivvu3btXYWFh8vX1lcViUWxsbInGvHDhgoYPH64aNWrIzc1NYWFhSklJuaa6AQAAAAAAUHLXFEp5enrqm2++0enTp3X69Gk99NBDNttXrVqlqKioYo+3cuVKRUZGKioqSrt27VKbNm0UEhKi1NTUQvtnZWWpUaNGiomJkY+PT4nHHDNmjL744gutWrVKW7Zs0YkTJ/Twww8Xu24AAAAAAABcn2sKpfJ5eHjI0dGxQHv16tVtrpy6mpkzZ2ro0KEKDw9Xy5YtNW/ePFWqVEmLFi0qtH/79u01bdo09evXTy4uLiUaMz09XQsXLtTMmTN17733yt/fX4sXL9bWrVv1/fffF7t2AAAAAAAAlFyJQqnSkJOTo8TERAUHB/+vGAcHBQcHKyEh4YaNmZiYqIsXL9r0ad68uerXr1/i/QIAAAAAAODaONlrx2lpacrNzZW3t7dNu7e3t/bv33/DxkxOTpazs7OqVq1aoE9ycnKRY2dnZys7O9v6PCMjo0Q1AgAAAAAAwI5XSt1soqOj5eHhYX3Uq1fP3iUBAAAAuEbXcqMl6a91c5s3by5XV1e1atVK69ats9mekpKiwYMHq3bt2qpUqZJ69OihgwcPFjqWYRjq2bOnLBaLVq9ebbNtx44d6tatm6pWrapq1aopJCREP/zww3UdKwCUdXYLpTw9PeXo6FjgrncpKSlFLmJeGmP6+PgoJydHZ86cuab9jh8/Xunp6dbHb7/9VqIaAQAAANjHtd5oaevWrerfv78iIiK0e/duhYaGKjQ0VD/99JOkv0Km0NBQ/fLLL/rss8+0e/duNWjQQMHBwcrMzCwwXmxsrCwWS4H2c+fOqUePHqpfv762bdumb7/9VlWqVFFISIguXrxYum8CAJQhdgulnJ2d5e/vr7i4OGtbXl6e4uLiFBgYeMPG9Pf3V4UKFWz6HDhwQMeOHbvifl1cXOTu7m7zAAAAAHDzuNYbLc2aNUs9evTQ2LFj1aJFC02ZMkV33nmnZs+eLUk6ePCgvv/+e82dO1ft27dXs2bNNHfuXJ0/f14ffPCBzVhJSUmaMWNGofvav3+//vzzT7366qtq1qyZbr/9dkVFRSklJUW//vpr6b8RAFBG2HX6XmRkpN5++20tXbpUP//8s5555hllZmYqPDxckjRw4ECNHz/e2j8nJ0dJSUlKSkpSTk6Ojh8/rqSkJB06dKjYY3p4eCgiIkKRkZH6+uuvlZiYqPDwcAUGBuquu+4y9w0AAAAAYIqS3GgpISHBpr8khYSEWPvnrznr6upqM6aLi4u+/fZba1tWVpYGDBigOXPmFDo7o1mzZqpRo4YWLlyonJwcnT9/XgsXLlSLFi3k6+tb4mMGgLLObgudS1Lfvn116tQpTZw4UcnJyfLz89P69eutC5UfO3ZMDg7/y81OnDihtm3bWp9Pnz5d06dPV1BQkOLj44s1piS98cYbcnBwUFhYmLKzsxUSEqK33nrLnIMGAAAAYLqS3GgpOTm50P75N0jKv4v3+PHjNX/+fFWuXFlvvPGGfv/9d508edL6mjFjxqhjx47q3bt3ofupUqWK4uPjFRoaqilTpkiSmjZtqg0bNsjJya6/sgHADWX3b7gRI0ZoxIgRhW7LD5ry+fr6yjCM6xpT+utfMubMmaM5c+ZcU60AAAAAkK9ChQr65JNPFBERoerVq8vR0VHBwcHq2bOn9feWzz//XJs3b9bu3buLHOf8+fOKiIjQ3XffrQ8++EC5ubmaPn26evXqpR07dqhixYpmHRIAmMruoRQAAAAA3GgludGSj4/PVfv7+/srKSlJ6enpysnJkZeXlwICAtSuXTtJ0ubNm3X48GFVrVrVZpywsDB17txZ8fHxWr58uY4ePaqEhATrTJHly5erWrVq+uyzz9SvX7/rPXwAKJPsuqYUAAAAAJihJDdaCgwMtOkvSRs3biy0v4eHh7y8vHTw4EHt3LnTOlVv3Lhx2rNnj3Vt3KSkJEl/LSmyePFiSX+tOeXg4GBzZ77853l5edd13ABQlnGlFAAAAIBbQmRkpAYNGqR27dqpQ4cOio2NLXCjpTp16ig6OlqS9NxzzykoKEgzZsxQr169tGLFCu3cuVMLFiywjrlq1Sp5eXmpfv36+vHHH/Xcc88pNDRU3bt3l/TX1VaFXYlVv359NWzYUJJ03333aezYsRo+fLhGjhypvLw8xcTEyMnJSV27dr3RbwsA2A2hFAAAAIBbwrXeaKljx45avny5JkyYoJdffllNmzbV6tWrdccdd1j7nDx5UpGRkUpJSVGtWrU0cOBAvfLKK9dUV/PmzfXFF19o8uTJCgwMlIODg9q2bav169erVq1apXPwAFAGWYzirByOAjIyMuTh4aH09HS5u7vbuxwAAIBSEbM7zd4llDnjfvaydwllzwB+hQBKgu/Ygsa19bR3CbgBipuZsKYUAAAAAAAATEcoBQAAAAAAANMRSgEAAAAAAMB0hFIAAAAAAAAwHaEUAAAAAAAATEcoBQAAAAAAANM52bsAAAAAACiJmN1p9i6hzBnX1tPeJQBAsXGlFAAAAAAAAExHKAUAAAAAAADTEUoBAAAAAADAdIRSAAAAAAAAMB2hFAAAAAAAAExHKAUAAAAAAADTEUoBAAAAAADAdIRSAAAAAAAAMB2hFAAAAAAAAExHKAUAAAAAAADTEUoBAAAAAADAdIRSAAAAAAAAMB2hFAAAAAAAAExHKAUAAAAAAADTEUoBAAAAAADAdIRSAAAAAAAAMB2hFAAAAAAAAExHKAUAAAAAAADTEUoBAAAAAADAdIRSAAAAAAAAMB2hFAAAAAAAAExHKAUAAAAAAADTEUoBAAAAAADAdIRSAAAAAAAAMB2hFAAAAAAAAExHKAUAAAAAAADTEUoBAAAAAADAdIRSAAAAAAAAMB2hFAAAAAAAAExHKAUAAAAAAADTEUoBAAAAAADAdIRSAAAAAAAAMB2hFAAAAAAAAExHKAUAAAAAAG4ac+bMka+vr1xdXRUQEKDt27dfsf+qVavUvHlzubq6qlWrVlq3bp3NdovFUuhj2rRp1j67du3Sfffdp6pVq6pGjRp66qmndO7cuQL7WrJkiVq3bi1XV1fVrFlTw4cPL52DLqcIpQAAAAAAwE1h5cqVioyMVFRUlHbt2qU2bdooJCREqamphfbfunWr+vfvr4iICO3evVuhoaEKDQ3VTz/9ZO1z8uRJm8eiRYtksVgUFhYmSTpx4oSCg4PVpEkTbdu2TevXr9fevXs1ePBgm33NnDlT//znPzVu3Djt3btXmzZtUkhIyA17L8oDi2EYhr2LuBllZGTIw8ND6enpcnd3t3c5AAAApSJmd5q9Syhzxv3sZe8Syp4BZeNXCM7Xgsa19bR3CbgCztmCrvWcDQgIUPv27TV79mxJUl5enurVq6eRI0dq3LhxBfr37dtXmZmZWrNmjbXtrrvukp+fn+bNm1foPkJDQ3X27FnFxcVJkhYsWKBXXnlFJ0+elIPDX9f2/Pjjj2rdurUOHjyoJk2a6PTp06pTp46++OILdevW7ZqOqTwqbmbClVIAAAAAAKDMy8nJUWJiooKDg61tDg4OCg4OVkJCQqGvSUhIsOkvSSEhIUX2T0lJ0dq1axUREWFty87OlrOzszWQkqSKFStKkr799ltJ0saNG5WXl6fjx4+rRYsWqlu3rh599FH99ttvJTvYWwShFAAAAAAAKPPS0tKUm5srb29vm3Zvb28lJycX+prk5ORr6r906VJVqVJFDz/8sLXt3nvvVXJysqZNm6acnBydPn3aelXWyZMnJUm//PKL8vLy9K9//UuxsbH66KOP9Oeff+q+++5TTk5OiY+5vCOUAgAAAAAAkLRo0SI99thjcnV1tbbdfvvtWrp0qWbMmKFKlSrJx8dHDRs2lLe3t/Xqqby8PF28eFFvvvmmQkJCdNddd+mDDz7QwYMH9fXXX9vrcMo8QikAAAAAAFDmeXp6ytHRUSkpKTbtKSkp8vHxKfQ1Pj4+xe7/n//8RwcOHNCTTz5ZYNuAAQOUnJys48eP648//tCkSZN06tQpNWrUSJJUq1YtSVLLli2tr/Hy8pKnp6eOHTt2bQd6CyGUAgAAAAAAZZ6zs7P8/f2tC5BLf12hFBcXp8DAwEJfExgYaNNf+mv9p8L6L1y4UP7+/mrTpk2RNXh7e8vNzU0rV66Uq6ur7rvvPknS3XffLUk6cOCAte+ff/6ptLQ0NWjQoPgHeYtxsncBAAAAAAAAxREZGalBgwapXbt26tChg2JjY5WZmanw8HBJ0sCBA1WnTh1FR0dLkp577jkFBQVpxowZ6tWrl1asWKGdO3dqwYIFNuNmZGRo1apVmjFjRqH7nT17tjp27Cg3Nzdt3LhRY8eOVUxMjKpWrSpJuu2229S7d28999xzWrBggdzd3TV+/Hg1b95cXbt2vXFvyE2OUAoAAAAAANwU+vbtq1OnTmnixIlKTk6Wn5+f1q9fb13M/NixYzZ3yevYsaOWL1+uCRMm6OWXX1bTpk21evVq3XHHHTbjrlixQoZhqH///oXud/v27YqKitK5c+fUvHlzzZ8/X0888YRNn3fffVdjxoxRr1695ODgoKCgIK1fv14VKlQo5Xeh/LAYhmHYu4ibUUZGhjw8PJSeni53d3d7lwMAAFAqYnan2buEMmfcz172LqHsGVA2foXgfC1oXFtPe5eAK+CcLYhztnwqbmbCmlIAAAAAAAAwHaEUAAAAAAAATEcoBQAAAAAAANMRSgEAAAAAAMB0hFIAAAAAAAAwHaEUAAAAAAAATOdk7wIAAAAAAMAtarnF3hWUPQMMe1dgGq6UAgAAAAAAgOkIpQAAAAAAAGA6QikAAAAAAACYjlAKAAAAAAAApiOUAgAAAAAAgOkIpQAAAAAAAGA6QikAAAAAAACYjlAKAAAAAAAApiOUAgAAAAAAgOkIpQAAAAAAAGC6MhFKzZkzR76+vnJ1dVVAQIC2b99+xf6rVq1S8+bN5erqqlatWmndunU22y0WS6GPadOmWfv4+voW2B4TE3NDjg8AAAAAAAC27B5KrVy5UpGRkYqKitKuXbvUpk0bhYSEKDU1tdD+W7duVf/+/RUREaHdu3crNDRUoaGh+umnn6x9Tp48afNYtGiRLBaLwsLCbMZ69dVXbfqNHDnyhh4rAAAAAAAA/mL3UGrmzJkaOnSowsPD1bJlS82bN0+VKlXSokWLCu0/a9Ys9ejRQ2PHjlWLFi00ZcoU3XnnnZo9e7a1j4+Pj83js88+U9euXdWoUSObsapUqWLTr3Llyjf0WAEAAAAAAPAXu4ZSOTk5SkxMVHBwsLXNwcFBwcHBSkhIKPQ1CQkJNv0lKSQkpMj+KSkpWrt2rSIiIgpsi4mJUY0aNdS2bVtNmzZNly5duo6jAQAAAAAAQHE52XPnaWlpys3Nlbe3t027t7e39u/fX+hrkpOTC+2fnJxcaP+lS5eqSpUqevjhh23aR40apTvvvFPVq1fX1q1bNX78eJ08eVIzZ84sdJzs7GxlZ2dbn2dkZFz1+AAAAAAAAFA4u4ZSZli0aJEee+wxubq62rRHRkZa/9y6dWs5Oztr2LBhio6OlouLS4FxoqOjNXny5BteLwAAAAAAwK3ArtP3PD095ejoqJSUFJv2lJQU+fj4FPoaHx+fYvf/z3/+owMHDujJJ5+8ai0BAQG6dOmSjh49Wuj28ePHKz093fr47bffrjomAAAAAAAACmfXUMrZ2Vn+/v6Ki4uztuXl5SkuLk6BgYGFviYwMNCmvyRt3Lix0P4LFy6Uv7+/2rRpc9VakpKS5ODgoJo1axa63cXFRe7u7jYPAAAAAAAAlIzdp+9FRkZq0KBBateunTp06KDY2FhlZmYqPDxckjRw4EDVqVNH0dHRkqTnnntOQUFBmjFjhnr16qUVK1Zo586dWrBggc24GRkZWrVqlWbMmFFgnwkJCdq2bZu6du2qKlWqKCEhQWPGjNHjjz+uatWq3fiDBgAAAAAAuMXZPZTq27evTp06pYkTJyo5OVl+fn5av369dTHzY8eOycHhfxd0dezYUcuXL9eECRP08ssvq2nTplq9erXuuOMOm3FXrFghwzDUv3//Avt0cXHRihUrNGnSJGVnZ6thw4YaM2aMzTpTAAAAAAAAuHEshmEY9i7iZpSRkSEPDw+lp6czlQ8AAJQbMbvT7F1CmTPuZy97l1D2DCgbv0JwvhY0rq2nvUvAFXDOFsR3bCHKyHfs9ShuZmLXNaUAAAAAAABwayKUAgAAAAAAgOkIpQAAAAAAAGA6QikAAAAAAACYjlAKAAAAAAAApiOUAgAAAAAAgOkIpQAAAAAAAGA6QikAAAAAAACYjlAKAAAAAAAApiOUAgAAAAAAgOkIpQAAAAAAAGA6QikAAAAAAACYjlAKAAAAAAAApiOUQonMmTNHvr6+cnV1VUBAgLZv337F/qtWrVLz5s3l6uqqVq1aad26dQX6/Pzzz3rwwQfl4eGhypUrq3379jp27FiBfoZhqGfPnrJYLFq9erXNtlGjRsnf318uLi7y8/O7nkMEAAAAAAA3EKEUrtnKlSsVGRmpqKgo7dq1S23atFFISIhSU1ML7b9161b1799fERER2r17t0JDQxUaGqqffvrJ2ufw4cPq1KmTmjdvrvj4eO3Zs0evvPKKXF1dC4wXGxsri8VSZH1DhgxR3759r/9AAQAAAADADeNk7wJw85k5c6aGDh2q8PBwSdK8efO0du1aLVq0SOPGjSvQf9asWerRo4fGjh0rSZoyZYo2btyo2bNna968eZKkf/7zn7r//vs1depU6+saN25cYKykpCTNmDFDO3fuVK1atQpsf/PNNyVJp06d0p49e67/YAEAAAAAwA3BlVK4Jjk5OUpMTFRwcLC1zcHBQcHBwUpISCj0NQkJCTb9JSkkJMTaPy8vT2vXrtVtt92mkJAQ1axZUwEBAQWm5mVlZWnAgAGaM2eOfHx8SvfAAAAAAACAqQilcE3S0tKUm5srb29vm3Zvb28lJycX+prk5OQr9k9NTdW5c+cUExOjHj166KuvvtJDDz2khx9+WFu2bLG+ZsyYMerYsaN69+5dykcFAAAAAADMxvQ92F1eXp4kqXfv3hozZowkyc/PT1u3btW8efMUFBSkzz//XJs3b9bu3bvtWSoAAAAAACglXCmFa+Lp6SlHR0elpKTYtKekpBQ5pc7Hx+eK/T09PeXk5KSWLVva9GnRooX17nubN2/W4cOHVbVqVTk5OcnJ6a88NSwsTF26dCmNQwMAAAAAACYilMI1cXZ2lr+/v+Li4qxteXl5iouLU2BgYKGvCQwMtOkvSRs3brT2d3Z2Vvv27XXgwAGbPv/973/VoEEDSdK4ceO0Z88eJSUlWR+S9MYbb2jx4sWldXgAUGbNmTNHvr6+cnV1VUBAgLZv337F/qtWrVLz5s3l6uqqVq1aad26dQX6/Pzzz3rwwQfl4eGhypUrq3379tZ/DPjzzz81cuRINWvWTBUrVlT9+vU1atQopaen35DjAwAAwK2H6Xu4ZpGRkRo0aJDatWunDh06KDY2VpmZmda78Q0cOFB16tRRdHS0JOm5555TUFCQZsyYoV69emnFihXauXOnFixYYB1z7Nix6tu3r+655x517dpV69ev1xdffKH4+HhJf11tVdiVWPXr11fDhg2tzw8dOqRz584pOTlZ58+ft4ZXLVu2lLOz8w16RwDgxlq5cqUiIyM1b948BQQEKDY2ViEhITpw4IBq1qxZoP/WrVvVv39/RUdH64EHHtDy5csVGhqqXbt26Y477pAkHT58WJ06dVJERIQmT54sd3d37d27V66urpKkEydO6MSJE5o+fbpatmypX3/9VU8//bROnDihjz76yNTjBwAAQPlkMQzDsHcRN6OMjAx5eHgoPT1d7u7u9i7HdLNnz9a0adOUnJwsPz8/vfnmmwoICJAkdenSRb6+vlqyZIm1/6pVqzRhwgQdPXpUTZs21dSpU3X//ffbjLlo0SJFR0fr999/V7NmzTR58uQrLmpusVj06aefKjQ01NrWpUsXm8XR8x05ckS+vr7XdcwAYC8BAQFq3769Zs+eLemvK1Tr1aunkSNHaty4cQX69+3bV5mZmVqzZo217a677pKfn5/mzZsnSerXr58qVKig9957r9h1rFq1So8//rgyMzOt06hR/sTsTrN3CWXOuJ+97F1C2TOgbPwKwfla0Li2nvYuAVfAOVsQ37GFKCPfsdejuJkJ0/dQIiNGjNCvv/6q7Oxsbdu2zRpISVJ8fLxNICVJffr00YEDB5Sdna2ffvqpQCAlSUOGDNHBgwetVzhd7S57hmHYBFL5+zYMo8CDQApmT32SpAULFqhLly5yd3eXxWLRmTNnbF5/9OhRRUREqGHDhqpYsaIaN26sqKgo5eTklMoxo3zIyclRYmKigoODrW0ODg4KDg5WQkJCoa9JSEiw6S9JISEh1v55eXlau3atbrvtNoWEhKhmzZoKCAjQ6tWrr1hL/v9UEEgBAACgNBBKASj38qc+RUVFadeuXWrTpo1CQkKUmppaaP/8qU8RERHavXu3QkNDFRoaqp9++snaJ3/qU/PmzRUfH689e/bolVdesU59kqSsrCz16NFDL7/8cqH72b9/v/Ly8jR//nzt3btXb7zxhubNm1dkf9ya0tLSlJubK29vb5t2b29vJScnF/qa5OTkK/ZPTU3VuXPnFBMTox49euirr77SQw89pIcffrjQq03z65gyZYqeeuqpUjgqAAAAgDWlANwCZs6cqaFDh1rXPZs3b57Wrl2rRYsWFTr1adasWerRo4fGjh0rSZoyZYo2btyo2bNnW6c+/fOf/9T999+vqVOnWl/XuHFjm3FGjx4tSda10S7Xo0cP9ejRw/q8UaNGOnDggObOnavp06eX+HiBq8nLy5Mk9e7dW2PGjJEk+fn5aevWrZo3b56CgoJs+mdkZKhXr15q2bKlJk2aZHa5AAAAKKe4UgpAuVaWpj4VR3p6uqpXr37d46D88PT0lKOjo1JSUmzaU1JSCr0BhPTXzSGu1N/T01NOTk5q2bKlTZ8WLVrYTEGVpLNnz6pHjx6qUqWKPv30U1WoUOF6DwkAAACQRCgFoJwrK1OfiuPQoUP697//rWHDhpV4DJQ/zs7O8vf3V1xcnLUtLy9PcXFxCgwMLPQ1gYGBNv0laePGjdb+zs7Oat++vQ4cOGDT57///a8aNGhgfZ6RkaHu3bvL2dlZn3/+uc30VAAAAOB6MX0PAK7RtU59Ko7jx4+rR48e6tOnj4YOHVqq9eLmFxkZqUGDBqldu3bq0KGDYmNjlZmZaZ2SOnDgQNWpU0fR0dGSpOeee05BQUGaMWOGevXqpRUrVmjnzp1asGCBdcyxY8eqb9++uueee9S1a1etX79eX3zxhXW6aX4glZWVpffff18ZGRnKyMiQJHl5ecnR0dHcNwEAAADlDqEUgHLN7KlP33777TXXeOLECXXt2lUdO3a0CQ2AfH379tWpU6c0ceJEJScny8/PT+vXr7de0Xfs2DE5OPzv4ueOHTtq+fLlmjBhgl5++WU1bdpUq1ev1h133GHt89BDD2nevHmKjo7WqFGj1KxZM3388cfq1KmTJGnXrl3atm2bJKlJkyY29Rw5coS7mgIAAOC6EUpBMbvT7F1CmTOurae9S0Ap+fvUp9DQUEn/m/o0YsSIQl+TP/Upf6FyqWRTn4rj+PHj6tq1q/z9/bV48WKbYAH4uxEjRhR5zha2mH6fPn3Up0+fK445ZMgQDRkypNBtXbp0kWEY11wnAAAAUFyEUgDKPXtMfZL+WpsqOTlZhw4dkiT9+OOPqlKliurXr6/q1avr+PHj6tKlixo0aKDp06fr1KlT1tcWdRUXAAAAAJQXhFIAyj17TH2SpHnz5mny5MnW5/fcc48kafHixRo8eLA2btyoQ4cO6dChQ6pbt65NzVyhAgAAAKC8sxj85lMiGRkZ8vDwUHp6utzd3e1dznVh+l5BTN8DANyq+P+Cgsb97GXvEsqeAWXjVwjO14L4/9iyjXO2IL5jC1FGvmOvR3EzExYvAQAAAAAAgOkIpQAAAAAAAGA6QikAAAAAAACYjoXOAQC4gVg7oiDWOwEAAIDElVIAAAAAAACwA0IpAAAAAAAAmI7pewBuKkyFKoipUAAAAABuRlwpBQAAAAAAANMRSgEAAAAAAMB0hFIAAAAAAAAwHaEUAAAAAAAATEcoBQAAAAAAANMRSgEAAAAAAMB0hFIAAAAAAAAwHaEUAAAAAAAATEcoBQAAAAAAANMRSgEAAAAAAMB0hFIAAAAAAAAwHaEUAAAAAAAATEcoBQAAAAAAANMRSgEAAAAAAMB0hFIAAAAAcIubM2eOfH195erqqoCAAG3fvv2K/VetWqXmzZvL1dVVrVq10rp162y2Dx48WBaLxebRo0cPmz7/93//p44dO6pSpUqqWrVqofu5fAyLxaIVK1Zc17ECKDsIpQAAAADgFrZy5UpFRkYqKipKu3btUps2bRQSEqLU1NRC+2/dulX9+/dXRESEdu/erdDQUIWGhuqnn36y6dejRw+dPHnS+vjggw9stufk5KhPnz565plnrljf4sWLbcYJDQ29ruMFUHYQSgEAAADALWzmzJkaOnSowsPD1bJlS82bN0+VKlXSokWLCu0/a9Ys9ejRQ2PHjlWLFi00ZcoU3XnnnZo9e7ZNPxcXF/n4+Fgf1apVs9k+efJkjRkzRq1atbpifVWrVrUZx9XV9foOGECZQSgFAAAAALeonJwcJSYm6v/bu/e4qqr8/+NvQOGIhoQkqFHe7yCoiVjzU5OExkaZMcJLgxGDZVEYaQaDoDmmXbym5cNKpUcaZpnfhnFIRKkmURMk0686ZSWZHrylBn1FhPP7o4enThwRFPcBeT0fj/MYWfuz116LWZ1l7/bZJzQ01Nrm7Oys0NBQ5eXl2T0nLy/Ppl6SwsLCqtTn5uaqdevW6tatmyZNmqRTp05d1Rgff/xxeXt7a8CAAVqxYoUsFstV9QOg/mni6AEAAAAAABzj5MmTqqiokI+Pj027j4+PDhw4YPccs9lst95sNlt/Dg8P11/+8hd16NBBhw4dUnJysu69917l5eXJxcWlxuN77rnndPfdd8vd3V2bNm3SY489ppKSEj355JO1mCWA+opQCgAAAABQp8aMGWP9s7+/vwICAtSpUyfl5uZq2LBhNe5n+vTp1j8HBQWptLRUL730EqEUcIPg43sAAAAA0Eh5e3vLxcVFxcXFNu3FxcXy9fW1e46vr2+t6iWpY8eO8vb21tdff31N4w0ODtaRI0dUVlZ2Tf0AqB8IpQAAAACgkXJ1dVW/fv2Uk5NjbausrFROTo5CQkLsnhMSEmJTL0nZ2dmXrZekI0eO6NSpU2rTps01jbewsFA333yz3NzcrqkfAPUDH98DAAAAgEYsMTFREyZMUP/+/TVgwAAtXLhQpaWliomJkSRFR0erXbt2mjNnjiQpISFBgwcP1rx58zRixAhlZGRo165dWr58uSSppKREM2fO1OjRo+Xr66tDhw7pmWeeUefOnRUWFma9blFRkU6fPq2ioiJVVFSosLBQktS5c2e1aNFC//znP1VcXKyBAwfKZDIpOztbzz//vKZMmWLsLwjAdUMoBQAAAACNWFRUlE6cOKHU1FSZzWYFBgYqKyvL+jDzoqIiOTv/+iGbQYMGac2aNUpJSVFycrK6dOmiDRs2qHfv3pIkFxcX7dmzR+np6Tpz5ozatm2r4cOHa9asWTZ3OKWmpio9Pd36c1BQkCRp69atGjJkiJo2baqlS5fqqaeeksViUefOnTV//nzFxcUZ8WsBYABCKQAAAABo5OLj4xUfH2/3WG5ubpW2yMhIRUZG2q1v1qyZPvrooytec9WqVVq1atVlj4eHhys8PPyK/QBouHimFAAAAAAAAAxHKAUAAAAAAADDEUoBAAAAAADAcIRSAAAAAAAAMByhFAAAAAAAAAxHKAUAAAAAAADDNXH0AAAAAAAAdWSNk6NHUP+Mszh6BAAuo17cKbV06VK1b99eJpNJwcHB2rlzZ7X169atU/fu3WUymeTv76+NGzfaHH/ooYfk5ORk8woPD7epOX36tMaPHy8PDw95enoqNjZWJSUldT43AAAAAAAAVOXwUGrt2rVKTExUWlqaCgoK1KdPH4WFhen48eN267dt26axY8cqNjZWu3fvVkREhCIiIrR3716buvDwcB07dsz6euedd2yOjx8/Xvv27VN2drYyMzP1ySefaOLEiddtngAAAAAAAPiVw0Op+fPnKy4uTjExMerZs6eWLVsmd3d3rVixwm79okWLFB4erqlTp6pHjx6aNWuW+vbtqyVLltjUubm5ydfX1/q6+eabrcf279+vrKwsvfHGGwoODtZdd92lV155RRkZGTp69Oh1nS8AAAAAAAAcHEpduHBB+fn5Cg0NtbY5OzsrNDRUeXl5ds/Jy8uzqZeksLCwKvW5ublq3bq1unXrpkmTJunUqVM2fXh6eqp///7WttDQUDk7O2vHjh11MTUAAAAAAABUw6EPOj958qQqKirk4+Nj0+7j46MDBw7YPcdsNtutN5vN1p/Dw8P1l7/8RR06dNChQ4eUnJyse++9V3l5eXJxcZHZbFbr1q1t+mjSpIm8vLxs+vmtsrIylZWVWX8+d+5creYKAAAAAACAX92Q3743ZswY65/9/f0VEBCgTp06KTc3V8OGDbuqPufMmaOZM2fW1RABAAAAAAAaNYd+fM/b21suLi4qLi62aS8uLpavr6/dc3x9fWtVL0kdO3aUt7e3vv76a2sfv3+Q+sWLF3X69OnL9pOUlKSzZ89aX99///0V5wcAAAAAAAD7HBpKubq6ql+/fsrJybG2VVZWKicnRyEhIXbPCQkJsamXpOzs7MvWS9KRI0d06tQptWnTxtrHmTNnlJ+fb63ZsmWLKisrFRwcbLcPNzc3eXh42LwAAAAAAABwdRz+7XuJiYl6/fXXlZ6erv3792vSpEkqLS1VTEyMJCk6OlpJSUnW+oSEBGVlZWnevHk6cOCAZsyYoV27dik+Pl6SVFJSoqlTp2r79u367rvvlJOTo1GjRqlz584KCwuTJPXo0UPh4eGKi4vTzp079dlnnyk+Pl5jxoxR27Ztjf8lAAAAAAAANDIOf6ZUVFSUTpw4odTUVJnNZgUGBiorK8v6MPOioiI5O/+anQ0aNEhr1qxRSkqKkpOT1aVLF23YsEG9e/eWJLm4uGjPnj1KT0/XmTNn1LZtWw0fPlyzZs2Sm5ubtZ/Vq1crPj5ew4YNk7Ozs0aPHq3FixcbO3kAAAAAAIBGyuGhlCTFx8db73T6vdzc3CptkZGRioyMtFvfrFkzffTRR1e8ppeXl9asWVOrcQIAAAAAAKBuOPzjewAAAAAAAGh8CKUAAAAAAABgOEIpAAAAAAAAGI5QCgAAAAAAAIYjlAIAAAAAAIDhCKUAAAAAAABgOEIpAAAAAAAAGI5QCgAAAAAAAIYjlAIAAAAAAIDhCKUAAAAAAABgOEIpAAAAAAAAGI5QCgAAAAAAAIYjlAIAAAAAAIDhCKUAAAAAAABgOEIpAAAAAAAAGI5QCgAAAAAAAIYjlAIAAAAAAIDhCKUAAAAAAABgOEIpAAAAAAAAGI5QCgAAAAAAAIYjlAIAAAAAAIDhCKUAAAAAAABgOEIpAAAAAAAAGI5QCgAAAAAAAIYjlAIAAAAAAIDhCKUAAAAAAABgOEIpAAAAAAAAGI5QCgAAAAAAAIYjlAIAAAAAAIDhCKUAAAAAAABgOEIpAAAAAAAAGI5QCgAAAAAAAIYjlALqyNKlS9W+fXuZTCYFBwdr586d1davW7dO3bt3l8lkkr+/vzZu3GhzfMaMGerevbuaN2+um2++WaGhodqxY4dNzciRI3XbbbfJZDKpTZs2+utf/6qjR49ajx88eFBDhw6Vj4+PTCaTOnbsqJSUFJWXl9fdxAEAAAAAuAqEUkAdWLt2rRITE5WWlqaCggL16dNHYWFhOn78uN36bdu2aezYsYqNjdXu3bsVERGhiIgI7d2711rTtWtXLVmyRF9++aX+85//qH379ho+fLhOnDhhrRk6dKjeffddHTx4UO+//74OHTqk+++/33q8adOmio6O1qZNm3Tw4EEtXLhQr7/+utLS0q7fLwMAAAAAgBpwslgsFkcPoiE6d+6cWrZsqbNnz8rDw8PRw7kmc3efdPQQ6p1ng7xrVR8cHKw77rhDS5YskSRVVlbKz89PTzzxhJ599tkq9VFRUSotLVVmZqa1beDAgQoMDNSyZcvsXuPSmtu8ebOGDRtmt+bDDz9URESEysrK1LRpU7s1iYmJ+vzzz/Xpp5/Wao71Beu1qtquVxiLNVsVa7Z+Y81W9ez+Wxw9hPpnXP34VwjWa1WsVzvqyXqVWLP2sGbtqEdr9mrVNDPhTingGl24cEH5+fkKDQ21tjk7Oys0NFR5eXl2z8nLy7Opl6SwsLDL1l+4cEHLly9Xy5Yt1adPH7s1p0+f1urVqzVo0KDLBlJff/21srKyNHjw4JpMDQAAAACA64ZQCrhGJ0+eVEVFhXx8fGzafXx8ZDab7Z5jNptrVJ+ZmakWLVrIZDJpwYIFys7Olre37R0G06ZNU/PmzdWqVSsVFRXpf/7nf6pcb9CgQTKZTOrSpYv+8Ic/6LnnnruaqQIAAAAAUGcIpYB6bOjQoSosLNS2bdsUHh6uBx54oMpzqqZOnardu3dr06ZNcnFxUXR0tH7/qdy1a9eqoKBAa9as0b/+9S+9/PLLRk4DAAAAAIAqCKWAa+Tt7S0XFxcVFxfbtBcXF8vX19fuOb6+vjWqb968uTp37qyBAwfqzTffVJMmTfTmm29WuX7Xrl11zz33KCMjQxs3btT27dttavz8/NSzZ0+NHTtWc+fO1YwZM1RRUXG1UwYAwzniG05nz56tQYMGyd3dXZ6enpe91qpVqxQQECCTyaTWrVvr8ccfv+p5AgAANCaEUsA1cnV1Vb9+/ZSTk2Ntq6ysVE5OjkJCQuyeExISYlMvSdnZ2Zet/22/ZWVl1R6XdMWa8vJyay0A1HeO+obTCxcuKDIyUpMmTbrs2ObPn6+///3vevbZZ7Vv3z5t3rxZYWFhdTd5AACAGxihFFAHEhMT9frrrys9PV379+/XpEmTVFpaqpiYGElSdHS0kpKSrPUJCQnKysrSvHnzdODAAc2YMUO7du1SfHy8JKm0tFTJycnavn27Dh8+rPz8fD388MP64YcfFBkZKUnasWOHlixZosLCQh0+fFhbtmzR2LFj1alTJ2u4tXr1ar377rvav3+/vvnmG7377rtKSkpSVFTUZR+GjsbDEXeenD59WuPHj5eHh4c8PT0VGxurkpISm5p3331XgYGBcnd31+23366XXnqpbiaMBmv+/PmKi4tTTEyMevbsqWXLlsnd3V0rVqywW79o0SKFh4dr6tSp6tGjh2bNmqW+fftavyFVksaNG6fQ0FB17NhRvXr10vz583Xu3Dnt2bPHWjNz5kw99dRT8vf3t3udH3/8USkpKXrrrbc0btw4derUSQEBARo5cmTd/gIAAABuUIRSQB2IiorSyy+/rNTUVAUGBqqwsFBZWVnWh5kXFRXp2LFj1vpBgwZpzZo1Wr58ufr06aP33ntPGzZsUO/evSVJLi4uOnDggEaPHq2uXbvqT3/6k06dOqVPP/1UvXr1kiS5u7tr/fr1GjZsmLp166bY2FgFBATo448/lpubmySpSZMmeuGFFzRgwAAFBARo5syZio+P1xtvvGHwbwj1jaPuPBk/frz27dun7OxsZWZm6pNPPtHEiROtx//9739r/PjxevTRR7V37169+uqrWrBggU2YgMalvnzDqT3Z2dmqrKzUDz/8oB49eujWW2/VAw88oO+//77GfQAAADRmTpbfPxEZNXLu3Dm1bNlSZ8+elYeHh6OHc03m7j7p6CHUO88GeV+5CA7Beq3qatZrcHCw7rjjDmvYU1lZKT8/Pz3xxBN69tlnq9RHRUWptLRUmZmZ1raBAwcqMDBQy5Yts3uNS++Tmzdv1rBhw7R//3717NlTn3/+ufr37y9JysrK0h//+EcdOXJEbdu21bhx41ReXq5169ZZ+3nllVf04osvqqioSE5OTrWeq6OxZquqzZo9evSo2rVrp23bttl8xPmZZ57Rxx9/XOVuPOmXj1Wnp6dr7Nix1rZXX31VM2fOtHmeX2ZmpsaMGaOff/5Zbdq00YYNG3THHXdU6W/VqlWaPHmyzpw5Y9M+d+5cpaamqmPHjlq0aJFatmyplJQUHTlyRHv27JGrq2uN51mfsGarenb/LY4eQv0zrn78KwTrtSrWqx31ZL1KrFl7WLN21KM1e7VqmplwpxQANDKOuvMkLy9Pnp6e1kBKkkJDQ+Xs7GwNFsrKymQymWz6atasmY4cOaLDhw/XfrJANWryDafVufSMvsWLFyssLEwDBw7UO++8o6+++kpbt269jiMHAAC4MRBKAUAjc/LkSVVUVFg/XnqJj4+PzGaz3XPMZnON6jMzM9WiRQuZTCYtWLBA2dnZ8vb2tvbRunVrm/omTZrIy8vL2k9YWJjWr1+vnJwcVVZW6r///a/mzZsnSTYfgUXj4ehvOK1OmzZtJEk9e/a0tt1yyy3y9vZWUVFRjfsBAABorAilAAB15lrvPImLi1N8fLzuu+8+ubq6auDAgRozZoykX+7mQuNTn77h9PfuvPNOSdLBgwetbadPn9bJkyd1++2317gfAACAxoq/4QNAI+OoO098fX2rBFQXL17U6dOnrf04OTnphRdeUElJiQ4fPiyz2awBAwZIkjp27Hj1k0aD5ohvOJV++ZKKwsJCFRUVqaKiQoWFhSosLLR+Y2TXrl01atQoJSQkaNu2bdq7d68mTJig7t27a+jQoQb+hgAAABomQikAaGQcdedJSEiIzpw5o/z8fOvxLVu2qLKyUsHBwTbnubi4qF27dnJ1ddU777yjkJAQ3XILD8FsrBzxDaeSlJqaqqCgIKWlpamkpERBQUEKCgrSrl27rDVvvfWWgoODNWLECA0ePFhNmzZVVlaWmjZtatBvBwAAoOFq4ugBAACMl5iYqAkTJqh///4aMGCAFi5cWOXOk3bt2mnOnDmSfrnzZPDgwZo3b55GjBihjIwM7dq1S8uXL5f0y50ns2fP1siRI9WmTRudPHlSS5cutbnzpEePHgoPD1dcXJyWLVum8vJyxcfHa8yYMWrbtq2kX5539d5772nIkCE6f/68Vq5cqXXr1unjjz92wG8J9Ul8fLz1Tqffy83NrdIWGRlpc9fTb5lMJq1fv/6K11y1apVWrVpVbY2Hh4fefPPNWj2LCgAAAL8glAKARigqKkonTpxQamqqzGazAgMDq9x58ttnOF268yQlJUXJycnq0qWL3TtP0tPTdfLkSbVq1Up33HFHlTtPVq9erfj4eA0bNkzOzs4aPXq0Fi9ebDO29PR0TZkyRRaLRSEhIcrNzbV+hA8AAADAjYNQCrBnjZOjR1D/jLM4egSoY46488TLy0tr1qy57HFvb2/l5eVdsR8AAAAADR/PlAIAAAAAAIDhCKUAAAAAAABgOEIpAAAAAAAAGI5QCgAAAAAAAIbjQecAAMBYfJlEVXyZBAAAaIS4UwoAAAAAAACGI5QCAAAAAACA4fj4HgA0dHwUyj4+DgUAAADUa9wpBQAAAAAAAMMRSgEAAAAAAMBwhFIAAAAAAAAwHKEUAAAAAAAADEcoBQAAAAAAAMMRSgEAAAAAAMBwhFIAAAAAAAAwHKEUAAAAAAAADEcoBQAAAAAAAMMRSgEAAAAAAMBwhFIAAAAAAAAwHKEUAAAAAAAADEcoBQAAAAAAAMMRSgEAAAAAAMBw9SKUWrp0qdq3by+TyaTg4GDt3Lmz2vp169ape/fuMplM8vf318aNG63HysvLNW3aNPn7+6t58+Zq27atoqOjdfToUZs+2rdvLycnJ5vX3Llzr8v8AAAAAAAAYMvhodTatWuVmJiotLQ0FRQUqE+fPgoLC9Px48ft1m/btk1jx45VbGysdu/erYiICEVERGjv3r2SpJ9//lkFBQWaPn26CgoKtH79eh08eFAjR46s0tdzzz2nY8eOWV9PPPHEdZ0rAAAAAAAAfuHwUGr+/PmKi4tTTEyMevbsqWXLlsnd3V0rVqywW79o0SKFh4dr6tSp6tGjh2bNmqW+fftqyZIlkqSWLVsqOztbDzzwgLp166aBAwdqyZIlys/PV1FRkU1fN910k3x9fa2v5s2bX/f5AgAAAAAAwMGh1IULF5Sfn6/Q0FBrm7Ozs0JDQ5WXl2f3nLy8PJt6SQoLC7tsvSSdPXtWTk5O8vT0tGmfO3euWrVqpaCgIL300ku6ePHiZfsoKyvTuXPnbF4AAAAAAAC4Ok0cefGTJ0+qoqJCPj4+Nu0+Pj46cOCA3XPMZrPderPZbLf+/PnzmjZtmsaOHSsPDw9r+5NPPqm+ffvKy8tL27ZtU1JSko4dO6b58+fb7WfOnDmaOXNmbaYHAAAAAACAy3BoKHW9lZeX64EHHpDFYtFrr71mcywxMdH654CAALm6uuqRRx7RnDlz5ObmVqWvpKQkm3POnTsnPz+/6zd4AAAAAACAG5hDQylvb2+5uLiouLjYpr24uFi+vr52z/H19a1R/aVA6vDhw9qyZYvNXVL2BAcH6+LFi/ruu+/UrVu3Ksfd3NzshlUAAAAAAACoPYc+U8rV1VX9+vVTTk6Ota2yslI5OTkKCQmxe05ISIhNvSRlZ2fb1F8KpL766itt3rxZrVq1uuJYCgsL5ezsrNatW1/lbAAAAAAAAFBTDv/4XmJioiZMmKD+/ftrwIABWrhwoUpLSxUTEyNJio6OVrt27TRnzhxJUkJCggYPHqx58+ZpxIgRysjI0K5du7R8+XJJvwRS999/vwoKCpSZmamKigrr86a8vLzk6uqqvLw87dixQ0OHDtVNN92kvLw8PfXUU3rwwQd18803O+YXAQAAAAAA0Ig4PJSKiorSiRMnlJqaKrPZrMDAQGVlZVkfZl5UVCRn519v6Bo0aJDWrFmjlJQUJScnq0uXLtqwYYN69+4tSfrhhx/04YcfSpICAwNtrrV161YNGTJEbm5uysjI0IwZM1RWVqYOHTroqaeesnlmFAAAAAAAAK4fh4dSkhQfH6/4+Hi7x3Jzc6u0RUZGKjIy0m59+/btZbFYqr1e3759tX379lqPEwAAAAAAAHXDoc+UAgAAAAAAQONEKAUAAAAAAADDEUoBAAAAAADAcIRSAAAAAAAAMByhFAAAAAAAAAxHKAUAAAAAAADDEUoBAAAAAADAcIRSAAAAAAAAMByhFAAAAAAAAAxHKAUAAAAAAADDEUoBAAAAAADAcIRSAAAAAAAAMByhFAAAAAAAAAxHKAUAAAAAAADDEUoBAAAAAADAcIRSAAAAAAAAMByhFAAAAAAAAAxHKAUAAAAAAADDEUoBAAAAAADAcIRSAAAAAAAAMByhFAAAAAAAAAxHKAUAAAAAAADDEUoBAAAAAADAcIRSAAAAAAAAMByhFAAAAAAAAAxHKAUAAAAAAADDEUoBAAAAAADAcIRSAAAAAAAAMByhFAAAAAAAAAxHKAUAAAAAAADDEUoBAAAAAADAcIRSAAAAAAAAMByhFAAAAAAAAAxHKAUAAAAAAADDEUoBAAAAAADAcIRSAAAAAAAAMByhFAAAAAAAAAxHKAUAAAAAAADDEUoBAAAAAADAcIRSAAAAAAAAMByhFAAAAAAAAAxHKAUAAAAAAADDEUoBAAAAAADAcIRSAAAAAAAAMByhFAAAAAAAAAxHKAUAAAAAAADDEUoBAAAAAADAcIRSAAAAAAAAMByhFAAAAAAAAAxHKAUAAAAAAADDEUoBAAAAAADAcIRSAAAAAAAAMByhFAAAAAAAAAxHKAUAAAAAAADDEUoBAAAAAADAcIRSAAAAAAAAMByhFAAAAAAAAAxHKAUAAAAAAADDEUoBAAAAAADAcIRSAAAAAAAAMByhFAAAAAAAAAxHKAUAAAAAAADDEUoBAAAAAADAcIRSAAAAAAAAMByhFAAAAAAAAAxHKAUAAAAAAADDEUoBAAAAAADAcIRSAAAAAAAAMByhFAAAAAAAAAxHKAUAAAAAAADDEUoBAAAAAADAcIRSAAAAAAAAMByhFAAAAAAAAAxXL0KppUuXqn379jKZTAoODtbOnTurrV+3bp26d+8uk8kkf39/bdy40ea4xWJRamqq2rRpo2bNmik0NFRfffWVTc3p06c1fvx4eXh4yNPTU7GxsSopKanzuQEAAAAAAKAqh4dSa9euVWJiotLS0lRQUKA+ffooLCxMx48ft1u/bds2jR07VrGxsdq9e7ciIiIUERGhvXv3WmtefPFFLV68WMuWLdOOHTvUvHlzhYWF6fz589aa8ePHa9++fcrOzlZmZqY++eQTTZw48brPFwAAAAAAAPUglJo/f77i4uIUExOjnj17atmyZXJ3d9eKFSvs1i9atEjh4eGaOnWqevTooVmzZqlv375asmSJpF/uklq4cKFSUlI0atQoBQQE6K233tLRo0e1YcMGSdL+/fuVlZWlN954Q8HBwbrrrrv0yiuvKCMjQ0ePHjVq6gAAAAAAAI1WE0de/MKFC8rPz1dSUpK1zdnZWaGhocrLy7N7Tl5enhITE23awsLCrIHTt99+K7PZrNDQUOvxli1bKjg4WHl5eRozZozy8vLk6emp/v37W2tCQ0Pl7OysHTt26M9//nOV65aVlamsrMz689mzZyVJ586dq/3E65nzJT85egj1zrmfHT2CeqierHXWa1Ws18tgzdZbrFk76sl6lViz9rBm7agna5b1WhXr1Y56sl4l1qw9rFk76tGavVqXshKLxVJtnUNDqZMnT6qiokI+Pj427T4+Pjpw4IDdc8xms916s9lsPX6prbqa1q1b2xxv0qSJvLy8rDW/N2fOHM2cObNKu5+f3+Wmhwas6v/TUFxLR48Al8F6vQzWbL3FmrWD9VqvsWbtYM3WW6xXO1iv9Rpr1o4baM3+9NNPatny8vNxaCjVkCQlJdncoVVZWanTp0+rVatWcnJycuDIUNfOnTsnPz8/ff/99/Lw8HD0cIBqsV7R0LBm0dCwZtGQsF7R0LBmb1wWi0U//fST2rZtW22dQ0Mpb29vubi4qLi42Ka9uLhYvr6+ds/x9fWttv7S/xYXF6tNmzY2NYGBgdaa3z9I/eLFizp9+vRlr+vm5iY3NzebNk9Pz+oniAbNw8ODN0Y0GKxXNDSsWTQ0rFk0JKxXNDSs2RtTdXdIXeLQB527urqqX79+ysnJsbZVVlYqJydHISEhds8JCQmxqZek7Oxsa32HDh3k6+trU3Pu3Dnt2LHDWhMSEqIzZ84oPz/fWrNlyxZVVlYqODi4zuYHAAAAAAAA+xz+8b3ExERNmDBB/fv314ABA7Rw4UKVlpYqJiZGkhQdHa127dppzpw5kqSEhAQNHjxY8+bN04gRI5SRkaFdu3Zp+fLlkiQnJydNnjxZ//jHP9SlSxd16NBB06dPV9u2bRURESFJ6tGjh8LDwxUXF6dly5apvLxc8fHxGjNmzBVvLQMAAAAAAMC1c3goFRUVpRMnTig1NVVms1mBgYHKysqyPqi8qKhIzs6/3tA1aNAgrVmzRikpKUpOTlaXLl20YcMG9e7d21rzzDPPqLS0VBMnTtSZM2d01113KSsrSyaTyVqzevVqxcfHa9iwYXJ2dtbo0aO1ePFi4yaOesvNzU1paWlVPq4J1EesVzQ0rFk0NKxZNCSsVzQ0rFk4Wa70/XwAAAAAAABAHXPoM6UAAAAAAADQOBFKAQAAAAAAwHCEUgAAAAAAADAcoRTqLScnJ23YsOG6X2fIkCGaPHnydb8OAACoHns/AACNC6EUHOLEiROaNGmSbrvtNrm5ucnX11dhYWH67LPPrDXHjh3Tvffe68BR1lz79u3l5OQkJycnubu7y9/fX2+88YZNTW5urpycnNSrVy9VVFTYHPP09NSqVauq9Ld9+3abusmTJ2vIkCHXaxpwgIceekgRERF2j33xxRcaOXKkWrduLZPJpPbt2ysqKkrHjx/XjBkzrGvucq9L/Ts5OenRRx+t0v/jjz8uJycnPfTQQ9dxhmgMzGazEhIS1LlzZ5lMJvn4+OjOO+/Ua6+9pp9//llSzd4nV61aJU9PT7vXMCqswPXD3n999v4vvvhCY8eOlZ+fn5o1a6YePXpo0aJF1zw/XD/V7f1S/dj/c3NzNWrUKLVp00bNmzdXYGCgVq9efa1Txw2iJvu+dH33/vLyck2bNk3+/v5q3ry52rZtq+joaB09erSupgmDEErBIUaPHq3du3crPT1d//3vf/Xhhx9qyJAhOnXqlLXG19e3QX016HPPPadjx45p7969evDBBxUXF6d///vfVeq++eYbvfXWW1fsz2Qyadq0addjqGgATpw4oWHDhsnLy0sfffSR9u/fr5UrV6pt27YqLS3VlClTdOzYMevr1ltvta7BS69L/Pz8lJGRof/7v/+ztp0/f15r1qzRbbfd5ojp4QbyzTffKCgoSJs2bdLzzz+v3bt3Ky8vT88884wyMzO1efNma21N3ydxY2Lvvz57f35+vlq3bq23335b+/bt09///nclJSVpyZIlteoH9UN92f+3bdumgIAAvf/++9qzZ49iYmIUHR2tzMzM6zZ3NAy12fel67f3//zzzyooKND06dNVUFCg9evX6+DBgxo5cuQ19w2DWQCD/fjjjxZJltzc3GrrJFk++OADi8VisXz77bcWSZa1a9da7rrrLovJZLL079/fcvDgQcvOnTst/fr1szRv3twSHh5uOX78uLWPCRMmWEaNGmWZMWOGxdvb23LTTTdZHnnkEUtZWZm1ZvDgwZaEhATrz+fPn7c8/fTTlrZt21rc3d0tAwYMsGzdurXasd5+++2WBQsW2LR5eXlZnnrqKevPW7dutUiyTJ061eLn52c5f/689VjLli0tK1eutOnvySeftLi6ulr+9a9/WdsTEhIsgwcPrnYsaFgurdHf++CDDyxNmjSxlJeX16gfe2vwt/337t3b8vbbb1vbV69ebQkICLCMGjXKMmHChKscPWCxhIWFWW699VZLSUmJ3eOVlZUWi6Vm75MrV660tGzZ0m4/v90T0PCw9xu79z/22GOWoUOH1uocGOdye7/FUr/3/z/+8Y+WmJiYWp2DG09N932Lxfi9f+fOnRZJlsOHD9f4HDged0rBcC1atFCLFi20YcMGlZWV1erctLQ0paSkqKCgQE2aNNG4ceP0zDPPaNGiRfr000/19ddfKzU11eacnJwc7d+/X7m5uXrnnXe0fv16zZw587LXiI+PV15enjIyMrRnzx5FRkYqPDxcX331VY3GWFlZqffff18//vijXF1dqxyfPHmyLl68qFdeeaXafjp06KBHH31USUlJqqysrNG1cePw9fXVxYsX9cEHH8hisVxzfw8//LBWrlxp/XnFihWKiYm55n7RuJ06dUqbNm3S448/rubNm9utufRRkt+60vskbjzs/cbu/WfPnpWXl9dVnw/Hqc/7P+sKV7vvS8bs/WfPnpWTk9NlPw6I+olQCoZr0qSJVq1apfT0dHl6eurOO+9UcnKy9uzZc8Vzp0yZorCwMPXo0UMJCQnKz8/X9OnTdeeddyooKEixsbHaunWrzTmurq5asWKFevXqpREjRui5557T4sWL7f5lr6ioSCtXrtS6dev0hz/8QZ06ddKUKVN011132Wzo9kybNk0tWrSQm5ub7r//ft18883629/+VqXO3d1daWlpmjNnjs6ePVttnykpKfr222/5DH8jNHDgQCUnJ2vcuHHy9vbWvffeq5deeknFxcVX1d+DDz6o//znPzp8+LAOHz6szz77TA8++GAdjxqNzddffy2LxaJu3brZtHt7e1tDiN9+FKmm75O48bD3G7f3b9u2TWvXrtXEiROv6nw4Vn3d/9999119/vnn/AetRq62+75k3N5//vx5TZs2TWPHjpWHh0ed94/rh1AKDjF69GgdPXpUH374ocLDw5Wbm6u+ffvaPPDTnoCAAOuffXx8JEn+/v42bcePH7c5p0+fPnJ3d7f+HBISopKSEn3//fdV+v/yyy9VUVGhrl27Wt9YW7RooY8//liHDh2qdmxTp05VYWGhtmzZouDgYC1YsECdO3e2WxsbG6tWrVrphRdeqLbPW265RVOmTFFqaqouXLhQbS1uPLNnz5bZbNayZcvUq1cvLVu2TN27d9eXX35Z675uueUWjRgxQqtWrdLKlSs1YsQIeXt7X4dRA9LOnTtVWFioXr162dwVU5v3Sdx42Puv/96/d+9ejRo1SmlpaRo+fHitzkX9Ud/2/61btyomJkavv/66evXqVesx4MZ3uX1fMmbvLy8v1wMPPCCLxaLXXnutTvvG9dfE0QNA42UymXTPPffonnvu0fTp0/W3v/1NaWlp1X4TSNOmTa1/vnRr6O/bruV295KSErm4uCg/P18uLi42x1q0aFHtud7e3urcubM6d+6sdevWyd/fX/3791fPnj2r1DZp0kSzZ8/WQw89pPj4+Gr7TUxM1KuvvqpXX3219hNCg9eqVStFRkYqMjJSzz//vIKCgvTyyy8rPT291n09/PDD1vW2dOnSuh4qGqHOnTvLyclJBw8etGnv2LGjJKlZs2Y27Vd6n/Tw8FBpaakqKyvl7Pzrfzc7c+aMJKlly5bXcTYwAnv/9dv7//d//1fDhg3TxIkTlZKSUuPzUD/Vl/3/448/1p/+9CctWLBA0dHRtb42biy13fel67/3XwqkDh8+rC1btnCXVAPEnVKoN3r27KnS0tI67/eLL76w+daR7du3q0WLFvLz86tSGxQUpIqKCh0/ftz65nnp5evrW+Nr+vn5KSoqSklJSZetiYyMVK9evap9xoX0y1+Ip0+frtmzZ+unn36q8Rhw43F1dVWnTp2u+p+T8PBwXbhwQeXl5QoLC6vj0aExatWqle655x4tWbKk1uvS3vtkt27ddPHiRRUWFtrUFhQUSJK6du16zWNG/cLeb19t9/59+/Zp6NChmjBhgmbPnl3jMaNhcNT+n5ubqxEjRuiFF17g46CQdG37vlT3e/+lQOqrr77S5s2b1apVq1qPCY7HnVIw3KlTpxQZGamHH35YAQEBuummm7Rr1y69+OKLGjVqVJ1f78KFC4qNjVVKSoq+++47paWlKT4+3iaJv6Rr164aP368oqOjNW/ePAUFBenEiRPKyclRQECARowYUePrJiQkqHfv3tq1a5f69+9vt2bu3Lk1+svBxIkTtWDBAq1Zs0bBwcE1HgMahrNnz1bZiL/88kt99NFHGjNmjLp27SqLxaJ//vOf2rhx4xWfcXI5Li4u2r9/v/XPQF149dVXdeedd6p///6aMWOGAgIC5OzsrM8//1wHDhxQv379Lnvu798ne/XqpeHDh+vhhx/WvHnz1LFjRx08eFCTJ09WVFSU2rVrZ+DMUJfY+39V13v/3r17dffddyssLEyJiYkym82Sfnmfv+WWW2o8dhjL3t7fqlUrffHFF8rIyHD4/r9161bdd999SkhI0OjRo63rytXVlYedN3LXsu9Ldbf3l5eX6/7771dBQYEyMzNVUVFhXadeXl58kUoDQigFw7Vo0cL6eeJDhw6pvLxcfn5+iouLU3Jycp1fb9iwYerSpYv+3//7fyorK9PYsWM1Y8aMy9avXLlS//jHP/T000/rhx9+kLe3twYOHKj77ruvVtft2bOnhg8frtTUVG3cuNFuzd133627775bmzZtqravpk2batasWRo3blytxoCGITc3V0FBQTZtQ4cOVefOnfX000/r+++/l5ubm7p06aI33nhDf/3rX6/6WtzSjLrWqVMn7d69W88//7ySkpJ05MgRubm5qWfPnpoyZYoee+yxy55r731y7dq1SktL0yOPPKKjR4/q1ltv1Z///GdNnz7dqCnhOmDv/1Vd7/3vvfeeTpw4obfffltvv/22tf3222/Xd999V6vxwzj29v7Y2FglJyfL3d3d4ft/enq6fv75Z82ZM0dz5syxtg8ePFi5ublXPQ40fNey70t1t/f/8MMP+vDDDyVJgYGBNse2bt2qIUOGXNM8YRwnS1181yhQTz300EM6c+aMNmzY4OihAAAAA7D3AwDQcPBMKQAAAAAAABiOUAoAAAAAAACG4+N7AAAAAAAAMBx3SgEAAAAAAMBwhFIAAAAAAAAwHKEUAAAAAAAADEcoBQAAAAAAAMMRSgEAAAAAAMBwhFIAAAAAAAAwHKEUAAAAAAAADEcoBQAAAAAAAMMRSgEAAAAAAMBw/x9ChJjHTHh4EQAAAABJRU5ErkJggg==\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Chemins des modèles\n",
        "model_paths = {\n",
        "    'RNN_v1': '/content/drive/My Drive/DATASET/model_rnn_simple_best.h5',\n",
        "    'RNN_v2': '/content/drive/My Drive/DATASET/model_rnn_simple_best2.h5',\n",
        "    'LSTM_v1': '/content/drive/My Drive/DATASET/model_lstm_best.h5',\n",
        "    'LSTM_v2': '/content/drive/My Drive/DATASET/model_lstm_best2.h5',\n",
        "    'GRU_v1': '/content/drive/My Drive/DATASET/model_gru_best.h5',\n",
        "    'GRU_v2': '/content/drive/My Drive/DATASET/model_gru_best2.h5'\n",
        "}\n",
        "\n",
        "# Fonction pour charger, visualiser et afficher l'architecture de chaque modèle\n",
        "def visualize_model_architecture(model_path, model_name):\n",
        "    # Charger le modèle\n",
        "    model = load_model(model_path)\n",
        "\n",
        "    # Nom du fichier image\n",
        "    image_file = f'{model_name}_structure.png'\n",
        "\n",
        "    # Visualiser l'architecture\n",
        "    plot_model(model, to_file=image_file, show_shapes=True, show_layer_names=True)\n",
        "\n",
        "    # Afficher l'image\n",
        "    display(Image(filename=image_file))\n",
        "\n",
        "# Visualiser chaque modèle\n",
        "for model_name, path in model_paths.items():\n",
        "    print(f\"Visualizing {model_name}...\")\n",
        "    visualize_model_architecture(path, model_name)\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "z8M1zEWqWYx8",
        "outputId": "11d9ccba-d6cf-4e4c-f94f-ea3d111cbd48"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Visualizing RNN_v1...\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<IPython.core.display.Image object>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Visualizing RNN_v2...\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAWQAAAO/CAYAAAD4fbzlAAAABmJLR0QA/wD/AP+gvaeTAAAgAElEQVR4nOzdeVQUZ74//ncBTS/YzSKrKAqNSnAdo0YwDsn1jhP1SESjMlETzc0NGg0SFQkiiogL4lGv28xJomaiGQWFgwYk5qoHDKNxkhsdEQPigqhEARXZZfv8/vDb/bNtlmZpujCf1zn9B09VPfWp6qc+VD/1VJVARATGGGMmZ2bqABhjjD3DCZkxxkSCEzJjjIkEJ2TGGBMJixcLzp8/j61bt5oiFsYY+91YunQpfHx8dMr0zpDv3LmDo0ePdllQ7OVw9OhR3L1719RhiNqPP/6IH3/80dRhMBE4evQo7ty5o1eud4asceTIEaMGxF4ugiDg008/xcyZM00dimjNmDEDAB9b7Nnx0hTuQ2aMMZHghMwYYyLBCZkxxkSCEzJjjIkEJ2TGGBMJUSbk+fPnQyaTQRAE1NTUmDqcVsXGxsLLywtyuRxWVlbw8vJCZGQkysrK2lTPiRMnYG1tjW+//dZIkYrb7337m7JgwQIIgqD9zJkzR2+eU6dOITw8HImJifDw8NDOO3fuXL15J0yYAKVSCXNzcwwaNAi//PJLV2xGuzQ2NmLbtm3w9fVtcnpmZibGjh0LhUIBFxcXhIWF4enTp9rp0dHR8Pb2hkqlglQqhaenJ1asWIGKiopm11lTUwMvLy+sWrUKAHD8+HHExsaioaFBZ77k5GSd78Xe3r4TtlikCXn//v1Yvny5qcMw2A8//ID//u//RkFBAR48eIB169YhNjYW77zzTpvq+b0/eO/3vv3NsbOzQ1paGnJzc7F3716daWvWrMGOHTuwcuVKTJ8+HTdv3oRarUbPnj1x8OBBpKam6sz//fff48iRI5gyZQqys7MxYsSIrtwUg+Xl5eGPf/wjli5diqqqKr3p2dnZmDBhAsaPH4/i4mIkJSVh3759WLhwoXaeM2fOYPHixcjPz0dJSQk2bNiA7du3a4cfNiUiIgK5ubnav/39/SGTyTB+/HiUlpZqy99++23cvXsXZ8+exaRJkzppqwHQC+Lj46mJ4i4XERFBAKi6utrUobQqICBAL84ZM2YQACosLDRRVG1XVVVFPj4+7VoWAMXHx3dyRF2rI9tviHfeeYfeeeedNi0TFBRErq6uTU7buHEjDRgwQK/tqdVq+uabb8jMzIxcXV2ptLRUZ3paWhq9/fbbbQu+C126dImmTZtGBw8epOHDh9OwYcP05pk1axa5u7tTY2OjtiwuLo4EQaBff/2ViIgmT55M9fX1OsvNnDmTAFBBQYFenf/85z9pwoQJBIAiIiJ0pgUHB5OPjw/V1dXpLbdkyRLq2bNnm7axueNFlGfIz2tuALWYJCUlQSaT6ZS5uroCQIs/j8Rm7969KCoqMnUYJtOdtv/69euIjIzE2rVr9doeAPj6+iIkJAT37t3rVr82AWDYsGFITEzE7NmzIZVK9abX19cjNTUVfn5+Ovlh4sSJICIcO3YMAJCSkgJzc3OdZTVdCy+edVdXVyM0NBTbt29vMqaoqChcunSp2emdpcMJefPmzVAoFFAqlSgqKsKyZcvg6uqK3NxcNDQ0YPXq1XBzc4NcLsfQoUMRHx+vXTYjIwOjR4+GQqGASqXCkCFDdPpdzczMkJqaiokTJ8La2houLi7Yt2+fdvoPP/wAb29vWFtbQyaTYciQITh58iQAYMeOHZDJZHB0dMSCBQvg4uICmUwGX19fXLhwQVtHazG2V15eHmxsbNC3b1+D5s/MzISbmxsEQcCuXbsAAHv27IGVlRUUCgWOHTuGiRMnQqVSoXfv3jh06JDB2xkcHAxLS0s4Oztr17do0SJYWVlBEASUlJQgJCQEy5Ytw40bNyAIAjw9PTu8D9pCrNv/3XffQaVSYf369V26P1qzY8cOEBH8/f2bnScmJgYDBgzAl19+iVOnTjU7HxFh69ateOWVVyCVSmFra4upU6ciJycHgGHfA2C8Y+lFN2/eREVFBdzc3HTK1Wo1AODy5cvNLnvv3j3I5XK4u7vrlEdERGDRokVwcHBocjlbW1v4+flh+/btxu1ae/GUuT1dFpruhSVLltDOnTtp2rRp9Ouvv9Ly5ctJKpXS0aNH6fHjx7Ry5UoyMzOjn376iSoqKkilUlFsbCxVV1fT/fv3adq0aVRcXKxT5+nTp6m0tJQePXpEkyZNIqlUSpWVlUREdOTIEYqKiqJHjx7Rw4cPacyYMTo/HYKCgsjKyoquXr1KNTU1lJ2dTaNGjSKlUqn9ydJSjG1VW1tLd+/epZ07d5JUKqUDBw60afk7d+4QANq5c6fevj19+jQ9efKEioqKaNy4cWRlZUW1tbUGb+fs2bPJyclJZ31xcXEEQLvPp0+fTmq1us3bTdQ5XRZi3P6UlBRSKpUUHR3doW0j6twuCw8PD/L29m5yGbVaTbdu3SIionPnzpGZmRn169ePKioqiEi/y2L16tVkaWlJBw4coNLSUrp8+TKNGDGC7O3t6f79+0Rk2PfQmceSxmuvvabXZZGRkUEAKC4uTm9+uVxO48ePb7KuyspKUiqVFBwcrFOemZlJ/v7+RERUXFzcZJcFEVF4eDgBoIsXL+qUi7bLYtOmTVi8eDESExPRr18/7NmzBwEBAZg+fTpsbGywatUqSCQS7N+/H/n5+SgrK8OgQYMgk8ng5OSExMREvauVvr6+sLa2hq2tLQIDA/H06VPcunULAPDOO+9gzZo1sLW1hZ2dHfz9/fHw4UMUFxdrl7ewsND+5/f29saePXtQXl6O/fv3o6ampsUY26pPnz7o3bs3oqKisHnzZsyaNatjO/SF/aBSqeDg4IDAwEBUVlaioKDAoO18GZhq+ydPnoyysjJERkZ2dBM6TWVlJW7duqU9I2yJj48PPv30U+Tn5+Ozzz7Tm15dXY2tW7di2rRpmDNnDqytrTFkyBD87W9/Q0lJCT7//HOd+Zv7Hjr7WGqJZiTFi90RACCRSFBdXd3kchs2bICLiwtiYmK0ZdXV1QgJCcGePXtaXW///v0BAFlZWe0J2yBG60POzc1FVVUVBg8erC2Ty+VwdnZGTk4OPDw84OjoiDlz5iAqKgr5+fmt1imRSAAAdXV1LU5/cYjK80aOHAmFQoGcnJxWY2yrO3fuoKioCP/4xz/w97//HX/4wx+M0idpaWkJoPn9AOhu58vm9779RUVFICIoFAqD5o+JicHAgQOxe/duZGZm6kzLzs5GRUUFRo4cqVM+atQoWFpa6nTvvej576Gzj6WWaPrM6+vr9abV1tZCLpfrlSclJSEhIQEnT56EUqnUlq9cuRIfffSR9ppPSzT7+8GDB+0NvVVGS8iVlZUAgFWrVumM17t9+zaqqqogl8tx5swZvP7661i/fj08PDwQGBjY7H+3pqSmpuKNN96Ag4MDpFIpVqxYYdByUqkUxcXFrcbYVhKJBA4ODpgwYQIOHz6M7OxsbNiwoc31dBbNdv5evazbrxmb39QFr6bIZDLs378fgiDggw8+0DnGNEO5evToobecjY0NysvLDVpHZx9LLdFcC3hxnH9VVRVqamrg4uKiU3748GFs2rQJ6enp6Nevn7Y8MzMTWVlZ+PDDDw1arybRG/PeCKMlZE3n+LZt20BEOp/z588DAAYNGoRvv/0WhYWFCAsLQ3x8PLZs2WJQ/QUFBQgICICzszMuXLiAJ0+eIDY2ttXl6urqUFpait69exsUY3t5enrC3Nwc2dnZHaqnvZ7fzt+jl3n7NYmhpV+CL/Lx8cHSpUuRl5eHdevWacttbGwAoMnE25b9Z8xj6UXu7u5QKpW4ffu2Tvn169cBAEOHDtWW7dy5EwcPHsSZM2fQq1cvnfn37t2L06dPw8zMTPsPRLMd69evhyAI+Pnnn7Xz19bWAkCTZ+CdxWgJuU+fPpDJZLh06VKT0wsLC3H16lUAz77MjRs3YsSIEdqy1mRlZaGurg4ff/wxPDw8tHf2tSY9PR1EhDFjxrQaoyEePnyId999V688Ly8PDQ0N6NOnT7vr7ojntxN41sfa0k/8l83LvP2Ojo4QBAFPnjxp03Lr1q2Dl5cXLl68qC0bPHgwevTooZN4AODChQuora3Fq6++alDdnXEsGcrCwgKTJk3C2bNn0djYqC1PS0uDIAjw9/cHESEsLAxZWVlITk5u8hfA/v379f55aH5RRUREgIh0unI0+9vJyclo22a0hCyTyTB//nwcOnQIe/bsQVlZGRoaGnD37l389ttvKCwsxIIFC5CTk4Pa2lpcvHgRt2/f1h5ArdEMeTl16hRqamqQl5fXZH9XY2MjHj9+jPr6ely+fBkhISFwc3PDvHnzWo3REFZWVvj+++9x5swZlJWVoa6uDhcvXsT7778PKysrLF261PCd1gEtbSfw7Iz90aNHSE5ORl1dHYqLi/XOMOzs7FBYWIj8/HyUl5d3qwRmrO1PS0sT3bA3hUIBDw+PNr+hRdN18fzFMJlMhmXLliEpKQkHDx5EWVkZsrKysHDhQri4uCAoKMjguls7lgIDA+Hk5NQpt2tHRkbiwYMHWLNmDSorK3H+/HnExcVh3rx5GDhwIK5evYrNmzfjiy++gEQi0elGEQTB4F/iz9Ps7yFDhnQ4/ma9OOyircPeYmNjSS6XEwDq06ePzlCvp0+fUlhYGLm5uZGFhQU5ODjQ9OnTKTs7m/Lz88nX15dsbW3J3NycevXqRREREVRfX69TZ//+/enGjRt08OBBsrW1JQDUu3dvunLlCoWFhZGdnR3Z2NjQjBkzaNeuXQSA1Go1FRQUUFBQEEkkEnJ1dSULCwtSqVQ0depUunHjhkExGsrf35/c3d2pR48eJJVKSa1WU2BgIGVlZRlcx86dO8nZ2ZkAkEKhIH9/f9q9ezcpFAqd/fD555+TSqUiANS3b1+6du2aQdv58OFDevPNN0kmk5G7uzt98sknFBoaSgDI09OTCgoK6JdffqG+ffuSXC6n119/XTvkyRDo4LA3sW7/iRMnSKlUUkxMTLu3TaMzh70FBweTRCKhqqoqbVlSUhKp1WoCQPb29rR48eIm6wwNDdUZ9tbY2EhxcXHUv39/kkgkZGtrSwEBAZSbm0tEZPD30NqxFBAQQABo9erVLW7z+fPnaezYseTi4kIACAA5OzuTr68vZWRkaOfLyMig0aNHk1QqJRcXFwoNDaWamhoiIsrKytIu29SnqSFzRC0Pe5s8eTK5urrq3B1I1LnD3kR763RnCAoKIjs7O1OHYXRi2M6OJuSOEMP2G6IzE3JeXh5ZWFi0eay7KTU0NNC4ceNo7969pg6lzUpKSkgmk9GWLVv0pol2HLIYteXCR3f2e9nO5rzM219dXY2TJ08iLy9Pe2HJ09MT0dHRiI6O7ha35zc0NCA5ORnl5eUIDAw0dThtFhUVheHDhyM4OBjAs7sbCwsLkZmZqb2Y2Ble+oTcXjk5OXr9Tk19WmtcnVUP+/169OgR3nrrLQwYMAAffPCBtjw8PBwzZsxAYGBgmy/wdbX09HQkJiYiLS3N4PHTYrF161ZcunQJJ06c0N7rcOzYMbi6umLcuHF6T9TrkBdPmV+WLovw8HCytLQkANSvXz86cuSIqUMyCrFsJ0zUZSGW7TdEe7osDHHy5EkKCwvr9HoZUXJyMm3YsEHvqXEd1dzxIvy/iVoJCQmYNWsWP5uWtYkgCIiPj8fMmTNNHYpoaZ7De+TIERNHwkytueOFuywYY0wkOCEzxphIcEJmjDGR4ITMGGMiwQmZMcZEwqK5Cd3hXXZMXGbNmtWpD+V/WfGxxZrTbEI2xruw2Mtr1qxZCAkJgY+Pj6lDEa1t27YBAD799FMTR8JMrbkTl2YTMo8nZW0xa9Ys+Pj4cLtpgWb8Me8j1lxC5j5kxhgTCU7IjDEmEpyQGWNMJDghM8aYSHBCZowxkRBVQv7xxx/xyiuvaN8C6+TkhJiYGFOHhcTERHh4eGifXezs7Iw5c+aYOiz2O7BgwQKd52Y31e5OnTqF8PBwvXY6d+5cvXknTJgApVIJc3NzDBo0qFPeb2csjY2N2LZtG3x9fZucnpmZibFjx0KhUMDFxQVhYWF4+vSpdnp0dDS8vb2hUqkglUrh6emJFStWtPhA/5qaGnh5eWHVqlUAgOPHjyM2NlbvBQjJyck634u9vX0nbDHE+TzkP//5zwSAHj9+bNI4XqRWq8na2trUYYgSTPgKp+6iva9wsrOzo7S0NMrNzdW+M05j9erVNGXKFCorK9OWqdVq6tmzJwGglJQUvTrT0tJ03qknRteuXaOxY8cSABo2bJje9CtXrpBcLqfIyEiqqKigc+fOkb29Pc2fP187j5+fH+3evZsePnxIZWVlFB8fTxKJhN56661m17t06VK9d+pt376d/Pz8dPJRY2Mj3b17l86ePUuTJk3iVzgZU3V1dbP/lZm4GPO7Eks7kMvl2jeGSKVSbfmmTZtw+PBhJCQkQKlU6iyzY8cOmJmZISgoSPRvE3nRv//9b3z22WdYuHAhhg8f3uQ869atg7OzM9auXQsrKyv4+PggLCwMX331FXJycgAAPXr0QFBQEOzs7KBUKjFz5kwEBATgu+++w507d/TqPHfuHK5cuaJXvmTJEgwbNgyTJk1CfX09gGd3W2reGNK/f/9O23ZOyE3Yu3cvioqKTB0GM4Axvysxt4Pr168jMjISa9euhUwm05vu6+uLkJAQ3Lt3D8uXLzdBhO03bNgwJCYmYvbs2Tr/gDTq6+uRmpoKPz8/ndvQJ06cCCLCsWPHAAApKSkwNzfXWVbTtVBVVaVTXl1djdDQUGzfvr3JmKKionDp0qVmp3eWbpGQ9+zZAysrKygUChw7dgwTJ06ESqVC7969cejQIQDPzghkMhkcHR2xYMECuLi4QCaTwdfXFxcuXAAABAcHw9LSEs7Oztq6Fy1aBCsrKwiCgJKSEoSEhGDZsmW4ceMGBEGAp6dnm+P94Ycf4O3tDWtra8hkMgwZMgQnT54EAHz44Yfafie1Wo2LFy8CAObPnw+FQgFra2scP34cDQ0NWL16Ndzc3CCXyzF06FDt7eybN2+GQqGAUqlEUVERli1bBldXV+Tm5nZoP3clIsLWrVvxyiuvQCqVwtbWFlOnTtWe3bT3uzJ2O/juu++gUqmwfv36Ltxb+nbs2AEigr+/f7PzxMTEYMCAAfjyyy9x6tSpZudr7bsw5PgD0GKb7Uw3b95ERUUF3NzcdMrVajUA4PLly80ue+/ePcjlcri7u+uUR0REYNGiRXBwcGhyOVtbW/j5+WH79u3GfZvSi30YYu1DjoiIIAB0+vRpevLkCRUVFdG4cePIysqKamtriehZf5uVlRVdvXqVampqKDs7m0aNGkVKpZIKCgqIiGj27Nnk5OSks764uDgCQMXFxURENH36dFKr1XpxGdqHfOTIEYqKiqJHjx7Rw4cPacyYMTp9TNOnTydzc3O6d++eznLvvvsuHT9+nIiIli9fTlKplI4ePUqPHz+mlStXkpmZGf300086+2PJkiW0c+dOmjZtGv3666+txmYsaGMf8urVq8nS0pIOHDhApaWldPnyZRoxYgTZ29vT/fv3iaj935Ux20FKSgoplUqKjo42eFs12tuH7Orqqlfu4eFB3t7eTS6jVqvp1q1bRER07tw5MjMzo379+lFFRQUR6fchG/JdGHL8tdZm2+O1117T60POyMggABQXF6c3v1wup/HjxzdZV2VlJSmVSgoODtYpz8zMJH9/fyIiKi4u1utD1ggPDycAdPHiRZ3yJUuW/H77kH19faFSqeDg4IDAwEBUVlaioKBAO93CwkL7n97b2xt79uxBeXk59u/f32UxvvPOO1izZg1sbW1hZ2cHf39/PHz4EMXFxQCAhQsXoqGhQSemsrIy/PTTT5g0aRJqamqwZ88eBAQEYPr06bCxscGqVasgkUj0tmPTpk1YvHgxEhMT4eXl1WXb2BHV1dXYunUrpk2bhjlz5sDa2hpDhgzB3/72N5SUlODzzz/v8DqM1Q4mT56MsrIyREZGdjjG9qqsrMStW7e0Z4Qt8fHxwaeffor8/Hx89tlnetPb+l00d/y1pc12lGYkxYvdEQAgkUhQXV3d5HIbNmyAi4uLzsit6upqhISEYM+ePa2uV9NXnJWV1Z6wDdLtEvLzLC0tAQB1dXXNzjNy5EgoFArtzy9T0Lw6XDN05j/+4z8wYMAA7Nu3T/vz5/DhwwgMDIS5uTlyc3NRVVWFwYMHa+uQy+VwdnY26XZ0luzsbFRUVGDkyJE65aNGjYKlpaW2a6EziaEddJaioiIQERQKhUHzx8TEYODAgdi9ezcyMzN1pnXku3j++OvKNqvpM9dcYHtebW0t5HK5XnlSUhISEhJw8uRJnQugK1euxEcffQRXV9dW16vZ3w8ePGhv6K3q1gnZUFKpVHt22hVSU1PxxhtvwMHBAVKpFCtWrNCZLggCFixYgJs3b+L06dMAgK+//hr/9V//BeDZGRAArFq1Smes4+3bt/UuRnRHpaWlAJ5dBX+RjY0NysvLjbLerm4HxlJTUwMATV7waopMJsP+/fshCAI++OADnTPIzvouurLNavr+y8rKdMqrqqpQU1MDFxcXnfLDhw9j06ZNSE9PR79+/bTlmZmZyMrKwocffmjQejWJXrP/jeGlT8h1dXUoLS1F7969jbqes2fPYtu2bSgoKEBAQACcnZ1x4cIFPHnyBLGxsXrzz5s3DzKZDF9++SVyc3OhUqnQt29fANBeWNi2bRuISOdz/vx5o25HV7CxsQGAJg92Y31XXdUOuoImMbx4s0JLfHx8sHTpUuTl5WHdunXa8s76Lrqyzbq7u0OpVOL27ds65devXwcADB06VFu2c+dOHDx4EGfOnEGvXr105t+7dy9Onz6tvRFNEATtdqxfvx6CIODnn3/Wzl9bWwsATZ6Bd5aXPiGnp6eDiDBmzBgAz/oWW+riaK//+7//g5WVFbKyslBXV4ePP/4YHh4ekMlkTb4hwtbWFrNmzUJycjK2bNmC//7v/9ZO69OnD2QyGS5dutTpcYrB4MGD0aNHD53GDgAXLlxAbW0tXn31VQCd+111VTvoCo6OjhAEoc3ji9etWwcvLy/tyB7A8O+iNV3ZZi0sLDBp0iScPXsWjY2N2vK0tDQIggB/f38QEcLCwpCVlYXk5OQmfwHs379f75+H5hdUREQEiEinK0ezv52cnIy2bS9dQm5sbMTjx49RX1+Py5cvIyQkBG5ubpg3bx4AwNPTE48ePUJycjLq6upQXFys95/Wzs4OhYWFyM/PR3l5eYsHbl1dHR48eID09HRYWVlph+KcOnUKNTU1yMvLa7YfbuHChXj69ClSUlIwZcoUbblMJsP8+fNx6NAh7NmzB2VlZWhoaMDdu3fx22+/dXAPmZ5MJsOyZcuQlJSEgwcPoqysDFlZWVi4cCFcXFwQFBQEoGPflbHaQVpamsmHvSkUCnh4eODu3bttWk7TdfH8xTBDvwtD6m6tzQYGBsLJyalTbteOjIzEgwcPsGbNGlRWVuL8+fOIi4vDvHnzMHDgQFy9ehWbN2/GF198AYlEotONIggCtmzZ0uZ1avb3kCFDOhx/s14cdmHKYW8//vgjDRo0iMzMzAgAOTs70/r162n37t2kUCgIAPXv359u3LhBn3/+OalUKgJAffv2pWvXrlFQUBBJJBJydXUlCwsLUqlUNHXqVLpx44Z2HQ8fPqQ333yTZDIZubu70yeffEKhoaEEgDw9PamgoIB++eUX6tu3L8nlcnr99dfpr3/9K6nVagLQ4icpKYmIiMLCwsjOzo5sbGxoxowZtGvXLgJAarVaO+xK4w9/+AOFh4fr7YunT59SWFgYubm5kYWFBTk4OND06dMpOzubYmNjSS6XEwDq06cPHThwwLhfjAHQxmFvjY2NFBcXR/379yeJREK2trYUEBBAubm52nna813dv3/faO3g/v37dOLECVIqlRQTE9PmfdSZw96Cg4NJIpFQVVWVtiwpKUnbTu3t7Wnx4sVN1hkaGqoz7K2178LQ46+lNktEFBAQQABo9erVLW7z+fPnaezYseTi4qI9tpydncnX15cyMjK082VkZNDo0aNJKpWSi4sLhYaGam8tz8rKavFYbWrIHFHLw94mT55Mrq6u1NjYqFPemcPeRJWQO0pz3393MmnSJLp586apw+iwtiZkYxJrO+jMhJyXl0cWFhai+GdsqIaGBho3bhzt3bvX1KG0WUlJCclkMtqyZYvetN/1OOTWtOVChyk83/1x+fJlyGQyvbuGWMeJvR20RXV1NU6ePIm8vDzthSVPT09ER0cjOjq6xaeXiUVDQwOSk5NRXl6OwMBAU4fTZlFRURg+fDiCg4MBPLu7sbCwEJmZmdqLiZ3hpUvIYhcWFoa8vDxcu3YN8+fP17nizVhTHj16pH240AcffKAtDw8Px4wZMxAYGCj6Bwilp6cjMTERaWlpBo+fFoutW7fi0qVLOHHihPaegmPHjmkfLpSamtp5K3vxlLm7dlmEh4eTpaUlAaB+/frRkSNHTB1SkyIiIsjMzIz69OmjvU36ZQCRdFmIuR20p8vCECdPnqSwsLBOr5cRJScn04YNG6i+vr5T623ueBH+30SthIQEzJo1y7gP0GAvHUEQEB8fz6+4b8GMGTMAAEeOHDFxJMzUmjteuMuCMcZEghMyY4yJBCdkxhgTCU7IjDEmEhbNTUhISOjKONhL4GV48JExaW695WOLNevFYReaYW/84Q9/+MMf430MGvbGWHejGTrEZ56su+M+ZMYYEwlOyIwxJhKckBljTCQ4ITPGmEhwQmaMMZHghMwYYyLBCZkxxkSCEzJjjIkEJ2TGGBMJTsiMMSYSnJAZY0wkOCEzxphIcEJmjDGR4ITMGGMiwQmZMcZEghMyY4yJBCdkxhgTCU7IjDEmEpyQGWNMJDghM8aYSHBCZowxkeCEzBhjIsEJmV1ijf4AACAASURBVDHGRIITMmOMiQQnZMYYEwlOyIwxJhKckBljTCQ4ITPGmEhwQmaMMZHghMwYYyLBCZkxxkSCEzJjjIkEJ2TGGBMJC1MHwFhbnD17FufPn9cpy8nJAQDExsbqlPv4+OCPf/xjl8XGWEcJRESmDoIxQ50+fRr/+Z//CYlEAjOzpn/gNTY2oq6uDqdOncL48eO7OELG2o8TMutWGhsb4ezsjOLi4hbns7e3x/3792Fubt5FkTHWcdyHzLoVMzMzzJ49G5aWls3OY2lpiTlz5nAyZt0OJ2TW7fzlL39BbW1ts9Nra2vxl7/8pQsjYqxzcJcF65b69euH27dvNzmtT58+uH37NgRB6OKoGOsYPkNm3dLcuXMhkUj0yiUSCebNm8fJmHVLfIbMuqWcnBy88sorTU67cuUKBg0a1MURMdZxfIbMuiUvLy8MGjRI70zY29ubkzHrtjghs27rvffe0xlJIZFI8P7775swIsY6hrssWLd1584d9O3bF5omLAgCbt68iX79+pk2MMbaic+QWbfVp08fvPbaazAzM4OZmRlee+01TsasW+OEzLq1uXPnQhAEmJmZYe7cuaYOh7EO4S4L1q2VlJTA2dkZAFBYWAhHR0cTR8RY+xk9IfN4UMbYy8LY569d8vjNkJAQ+Pj4dMWq2P9z/vx5bN++HfHx8aYOxejOnj0LQRAwbty4Ni87a9Ysbp+sVZrjydi65Aw5Pj4eM2fONOZq2AsSEhIwa9Yso/9HF4Py8nIAgFKpbPOy3D6ZIbrqeOIH1LNurz2JmDEx4lEWjDEmEpyQGWNMJDghM8aYSHBCZowxkXjpE/L8+fMhk8kgCAJqampMHU6rYmNj4eXlBblcDisrK3h5eSEyMhJlZWUmiefEiROwtrbGt99+a5L1vwxOnTqF8PBwJCYmwsPDA4IgQBCEJu8snDBhApRKJczNzTFo0CD88ssvJojYMI2Njdi2bRt8fX2bnJ6ZmYmxY8dCoVDAxcUFYWFhePr0qXZ6dHQ0vL29oVKpIJVK4enpiRUrVqCioqLZddbU1MDLywurVq0CABw/fhyxsbFoaGjo3I0zFTIyABQfH2/s1bQoIiKCAFB1dbVJ4zDE5MmTacuWLVRUVETl5eWUkJBAEomE/vSnP7Wpnvj4eOqMrzclJYVUKhUdP368w3WJkbHb5+rVq2nKlClUVlamLVOr1dSzZ08CQCkpKXrLpKWl0dtvv220mDrDtWvXaOzYsQSAhg0bpjf9ypUrJJfLKTIykioqKujcuXNkb29P8+fP187j5+dHu3fvpocPH1JZWRnFx8eTRCKht956q9n1Ll26lABQRESEtmz79u3k5+dHjx8/7tyNfE5nHU+teenPkLsbS0tLLFq0CA4ODujRowdmzJiBqVOn4n//93/x22+/dXk8kydPxpMnTzBlypQuXzcAVFdXN3sGJnabNm3C4cOHkZCQoDc0b8eOHTAzM0NQUBCePHliogjb59///jc+++wzLFy4EMOHD29ynnXr1sHZ2Rlr166FlZUVfHx8EBYWhq+++go5OTkAgB49eiAoKAh2dnZQKpWYOXMmAgIC8N133+HOnTt6dZ47dw5XrlzRK1+yZAmGDRuGSZMmob6+vnM3tov9rhJyd7iNOykpCTKZTKfM1dUVAFr8Kfey2rt3L4qKikwdRptdv34dkZGRWLt2rd73CQC+vr4ICQnBvXv3sHz5chNE2H7Dhg1DYmIiZs+eDalUqje9vr4eqamp8PPz0znmJk6cCCLCsWPHAAApKSl6bwa3t7cHAFRVVemUV1dXIzQ0tNm75aKionDp0qUuuZvOmESVkDdv3gyFQgGlUomioiIsW7YMrq6uyM3NRUNDA1avXg03NzfI5XIMHTpU57bgjIwMjB49GgqFAiqVCkOGDNHpdzUzM0NqaiomTpwIa2truLi4YN++fdrpP/zwA7y9vWFtbQ2ZTIYhQ4bg5MmTAJ6dzchkMjg6OmLBggVwcXGBTCaDr68vLly4oK2jtRjbKy8vDzY2Nujbt2+H62qLzMxMuLm5QRAE7Nq1CwCwZ88eWFlZQaFQ4NixY5g4cSJUKhV69+6NQ4cOATBsfwUHB8PS0lL7YCAAWLRoEaysrCAIAkpKShASEoJly5bhxo0bEAQBnp6eAIDvvvsOKpUK69ev79L90RY7duwAEcHf37/ZeWJiYjBgwAB8+eWXOHXqVLPzERG2bt2KV155BVKpFLa2tpg6dar2TNOQ7wQwXvt80c2bN1FRUQE3NzedcrVaDQC4fPlys8veu3cPcrkc7u7uOuURERHaX45NsbW1hZ+fH7Zv39697041dp8I2thHp+nvXbJkCe3cuZOmTZtGv/76Ky1fvpykUikdPXqUHj9+TCtXriQzMzP66aefqKKiglQqFcXGxlJ1dTXdv3+fpk2bRsXFxTp1nj59mkpLS+nRo0c0adIkkkqlVFlZSURER44coaioKHr06BE9fPiQxowZQz179tTGFRQURFZWVnT16lWqqamh7OxsGjVqFCmVSiooKCAiajHGtqqtraW7d+/Szp07SSqV0oEDB9q0fGf1ed25c4cA0M6dO7Vlz+/PJ0+eUFFREY0bN46srKyotraWiAzbX7NnzyYnJyed9cXFxREA7Xc3ffp0UqvVOvOkpKSQUqmk6OjoDm9fW9unoTw8PMjb27vJaWq1mm7dukVEROfOnSMzMzPq168fVVRUEJF+H/Lq1avJ0tKSDhw4QKWlpXT58mUaMWIE2dvb0/3794nIsO+kM9unxmuvvabXh5yRkUEAKC4uTm9+uVxO48ePb7KuyspKUiqVFBwcrFOemZlJ/v7+RERUXFys14esER4eTgDo4sWL7d2cZnVVH7JoE/LzF+Cqq6tJoVBQYGCgtqyqqoqkUil9/PHHdOXKlWYvkDRX59dff00A6MqVK00us2HDBgJARUVFRPQswVhbW+vM89NPPxEAWrt2basxtpWTkxMBoJ49e9L//M//aA8qQ3VFQn5+f+7evZsA0PXr14mo9f1F1P6E3JmMkZArKipIEASaMmVKk9OfT8hERMuWLSMAtHjxYiLSTchVVVXUo0cPnXZFRPSvf/2LAGj/KbX2nXR2+9RoKiF///33BIC2bt2qN79KpSJfX98m64qIiKABAwboXACtqqqikSNH0t27d4mo5YS8b98+AkBff/11u7enOXxR7zm5ubmoqqrC4MGDtWVyuRzOzs7IycmBh4cHHB0dMWfOHERFRSE/P7/VOjWvkK+rq2txekvDaUaOHAmFQoGcnJxWY2yrO3fuoKioCP/4xz/w97//HX/4wx9E3ZdqaWkJoPn9Cejur5dZUVERiAgKhcKg+WNiYjBw4EDs3r0bmZmZOtOys7NRUVGBkSNH6pSPGjUKlpaWOl1mL3r+O+ns9tkSTZ95UxfYamtrIZfL9cqTkpKQkJCAkydP6lwAXblyJT766CPtdZSWaPb3gwcP2hu6yXWLhFxZWQkAWLVqlXYMpyAIuH37NqqqqiCXy3HmzBm8/vrrWL9+PTw8PBAYGIjq6mqD15Gamoo33ngDDg4OkEqlWLFihUHLSaVSFBcXtxpjW0kkEjg4OGDChAk4fPgwsrOzsWHDhjbXIzaa/fUy04x3b+qCV1NkMhn2798PQRDwwQcf6LTb0tJSAM9GJLzIxsZG+6S71nR2+2yJ5rrAi2Pnq6qqUFNTAxcXF53yw4cPY9OmTUhPT9d5BVdmZiaysrLw4YcfGrReTaLvDvcbNKdbJGRNR/62bdtAz7pZtJ/z588DAAYNGoRvv/0WhYWFCAsLQ3x8PLZs2WJQ/QUFBQgICICzszMuXLiAJ0+eIDY2ttXl6urqUFpait69exsUY3t5enrC3Nwc2dnZHarH1J7fXy8zTWJoy80KPj4+WLp0KfLy8rBu3TptuY2NDQA0mXjbsi+N2T5f5O7uDqVSidu3b+uUX79+HQAwdOhQbdnOnTtx8OBBnDlzBr169dKZf+/evTh9+jTMzMy0/0A027F+/XoIgoCff/5ZO39tbS0ANHkG3l10i4Tcp08fyGQyXLp0qcnphYWFuHr1KoBnDW/jxo0YMWKEtqw1WVlZqKurw8cffwwPDw/tnX2tSU9PBxFhzJgxrcZoiIcPH+Ldd9/VK8/Ly0NDQwP69OnT7rrF4Pn9BQAWFhYtdnF0V46OjhAEoc3ji9etWwcvLy9cvHhRWzZ48GD06NFDJ/EAwIULF1BbW4tXX33VoLo7o30aysLCApMmTcLZs2fR2NioLU9LS4MgCPD39wcRISwsDFlZWUhOTm7yF8D+/fv1/nlofl1FRESAiHS6cjT728nJychbaDzdIiHLZDLMnz8fhw4dwp49e1BWVoaGhgbcvXsXv/32GwoLC7FgwQLk5OSgtrYWFy9exO3bt7UHfms0w3NOnTqFmpoa5OXlNdk319jYiMePH6O+vh6XL19GSEgI3NzcMG/evFZjNISVlRW+//57nDlzBmVlZairq8PFixfx/vvvw8rKCkuXLjV8p4lAS/sLeHbm/+jRIyQnJ6Ourg7FxcV6Z1V2dnYoLCxEfn4+ysvLUVdXh7S0NFEPe1MoFPDw8MDdu3fbtJym6+L5sbkymQzLli1DUlISDh48iLKyMmRlZWHhwoVwcXFBUFCQwXW31j4DAwPh5OTUKbdrR0ZG4sGDB1izZg0qKytx/vx5xMXFYd68eRg4cCCuXr2KzZs344svvoBEItHpRhEEweBft8/T7O8hQ4Z0OH6TMfZVQ7ThKnZsbCzJ5XICQH369NEZ6vX06VMKCwsjNzc3srCwIAcHB5o+fTplZ2dTfn4++fr6kq2tLZmbm1OvXr0oIiKC6uvrders378/3bhxgw4ePEi2trYEgHr37k1XrlyhsLAwsrOzIxsbG5oxYwbt2rWLAJBaraaCggIKCgoiiURCrq6uZGFhQSqViqZOnUo3btwwKEZD+fv7k7u7O/Xo0YOkUimp1WoKDAykrKwsw3c6dc5V4Z07d5KzszMBIIVCQf7+/rR7925SKBQ6+/Pzzz8nlUpFAKhv37507do1g/bXw4cP6c033ySZTEbu7u70ySefUGhoKAEgT09PKigooF9++YX69u1LcrmcXn/9dbp//z6dOHGClEolxcTEdGj7iIw37C04OJgkEglVVVVpy5KSkkitVhMAsre3146qeFFoaKjOsLfGxkaKi4uj/v37k0QiIVtbWwoICKDc3FwiIoO/k9baZ0BAAAGg1atXt7ht58+fp7Fjx5KLiwsBIADk7OxMvr6+lJGRoZ0vIyODRo8eTVKplFxcXCg0NJRqamqIiCgrK0u7bFOfpobMEbU8ymLy5Mnk6upKjY2NLcbfHr/bYW9iFRQURHZ2dqYOw2Bd1YCa0132l7HaZ15eHllYWLR5/LgpNTQ00Lhx42jv3r2mDqXNSkpKSCaT0ZYtW4xSPw97E6GX5olSXeT3vL88PT0RHR2N6OjobnHLe0NDA5KTk1FeXo7AwEBTh9NmUVFRGD58OIKDg00dSodwQu4COTk5en1kTX2644HAmhceHo4ZM2YgMDBQ9A8QSk9PR2JiItLS0gwePy0WW7duxaVLl3DixAnt/QPdFSdkA6xcuRL79+/HkydP4O7ujqNHj7ZpeS8vL72rxU19Dh8+bKQt6Fod3V8vk/Xr1yM4OBgbN240dSgtGj9+PL755hudZ4t0B8eOHcPTp0+Rnp4OW1tbU4fTYQKRcZ/Ewa9ZN42uem15d8ftkxmiq44nPkNmjDGR4ITMGGMiwQmZMcZEghMyY4yJhEVXrKSzH17CWqfZ5wkJCSaORPy4fbLWdFUb6ZJRFowx9jJ4KUZZxMfHGzQOlz+d99G8K83UcYj9w+2TP4Z8jPHuwaZwHzJjjIkEJ2TGGBMJTsiMMSYSnJAZY0wkOCEzxphIcEJmjDGR6BYJOTExER4eHnrPD5bJZHB3d8cHH3yAW7dudcq65s+fr33JaXd+nTgzvlOnTiE8PFyvfc6dO1dv3gkTJkCpVMLc3ByDBg3qlPfWGUtjYyO2bdsGX19fo9eVmZmJsWPHQqFQwMXFBWFhYXj69Kl2enR0NLy9vaFSqSCVSuHp6YkVK1a0+ND/mpoaeHl5YdWqVQCA48ePIzY2tnu8MIGMDJ34ihy1Wk3W1tZE9Ox1Mw8ePKCvv/6aFAoFOTo6UklJSaesJyIiggBQdXV1p9RnCqZ+hVN30d72uXr1apoyZQqVlZVpy9RqNfXs2ZMAUEpKit4yaWlpOu/KE6Nr167R2LFjCQANGzbMqHVduXKF5HI5RUZGUkVFBZ07d47s7e1p/vz52nn8/Pxo9+7d9PDhQyorK6P4+HiSSCT01ltvNbvepUuX6r13b/v27eTn50ePHz9u17bwK5xaYWZmBkdHR8ydOxeLFy9GUVERTp06ZeqwGIDq6upOObvq6roNtWnTJhw+fBgJCQlQKpU603bs2AEzMzMEBQWJ/i0hL/r3v/+Nzz77DAsXLsTw4cONXte6devg7OyMtWvXwsrKCj4+PggLC8NXX32FnJwcAECPHj0QFBQEOzs7KJVKzJw5EwEBAfjuu+9w584dvTrPnTuHK1eu6JUvWbIEw4YNw6RJk1BfX9+hbTOmbpuQn+fp6QkAuH//fqfWy7d9t8/evXtRVFTU7eo2xPXr1xEZGYm1a9dCJpPpTff19UVISAju3buH5cuXmyDC9hs2bBgSExMxe/ZsSKVSo9ZVX1+P1NRU+Pn56RxnEydOBBHh2LFjAICUlBSYm5vrLGtvbw8AqKqq0imvrq5GaGgotm/f3mRMUVFRuHTpUrPTxeClSMh5eXkAnjUCjYaGBqxevRpubm6Qy+UYOnSozu2PGRkZGD16NBQKBVQqFYYMGYKysjLtdDMzM6SmpmLixImwtraGi4sL9u3bp53+ww8/wNvbG9bW1pDJZBgyZAhOnjwJ4NlZkkwmg6OjIxYsWAAXFxfIZDL4+vriwoULBsfY1YgIW7duxSuvvAKpVApbW1tMnTpVe7YSHBwMS0tLndf8LFq0CFZWVhAEASUlJQgJCcGyZctw48YNCIIAT09Pg/ZHe+sGgO+++w4qlQrr1683+j7asWMHiAj+/v7NzhMTE4MBAwbgyy+/bPFXW2v7e8+ePbCysoJCocCxY8cwceJEqFQq9O7dG4cOHdLWI7Z2ZIibN2+ioqICbm5uOuVqtRoAcPny5WaXvXfvHuRyOdzd3XXKIyIisGjRIjg4ODS5nK2tLfz8/LB9+3YQifRNOsbuE4GR+pCJiB4/fkxfffUVKRQKmjx5ss68y5cvJ6lUSkePHqXHjx/TypUryczMjH766SeqqKgglUpFsbGxVF1dTffv36dp06ZRcXExEf3/fcinT5+m0tJSevToEU2aNImkUilVVlYSEdGRI0coKiqKHj16RA8fPqQxY8ZQz549tesPCgoiKysrunr1KtXU1FB2djaNGjWKlEolFRQUtBpjR7Wnz2v16tVkaWlJBw4coNLSUrp8+TKNGDGC7O3t6f79+0RENHv2bHJyctJZLi4ujgBo99/06dNJrVbrzGPI/mhv3SkpKaRUKik6OrpN20vU9vbp4eFB3t7eTU5Tq9V069YtIiI6d+4cmZmZUb9+/aiiooKI9PuQDdnfz7fFJ0+eUFFREY0bN46srKyotraWiIzTjl577bUO9yG3VFdGRgYBoLi4OL355XI5jR8/vsm6KisrSalUUnBwsE55ZmYm+fv7ExFRcXGxXh+yRnh4OAGgixcvtmkbuqoPudslZAA6H0EQKCYmRts4iYiqq6tJoVBQYGCgtqyqqoqkUil9/PHHdOXKlWYvvBA1fVHv66+/JgB05cqVJpfZsGEDAaCioiIiepaAnv/nQUT0008/EQBau3ZtqzF2VFsbUFVVFfXo0UMnHiKif/3rXwRAm+w6kpBb2h8dqbsj2tI+KyoqSBAEmjJlSpPTn0/IRETLli0jALR48WIi0k3Ihu7vptri7t27CQBdv37daO3I2An5+++/JwC0detWvflVKhX5+vo2WVdERAQNGDBA52JqVVUVjRw5ku7evUtELSfkffv2EQD6+uuv27QNfFGvGdbW1tonMIWGhoKIYG1trfP679zcXFRVVWHw4MHaMrlcDmdnZ+Tk5MDDwwOOjo6YM2cOoqKikJ+f3+p6NfXX1dW1OL2loTUjR46EQqFATk5OqzF2tezsbFRUVGDkyJE65aNGjYKlpaVOV0tneX5/dAdFRUUgIigUCoPmj4mJwcCBA7F7925kZmbqTOvI/ra0tATwrC2KrR0ZStP/3tQFttraWsjlcr3ypKQkJCQk4OTJkzoXU1euXImPPvoIrq6ura5X8909ePCgvaEbVbdLyM+LjIyEs7MzVq5cqXPFtbKyEgCwatUqnXHLt2/fRlVVFeRyOc6cOYPXX38d69evh4eHBwIDA1FdXW3wulNTU/HGG2/AwcEBUqkUK1asMGg5qVSK4uLiVmPsaqWlpQCeXdV+kY2NDcrLy42yXs3+6A4049INveAlk8mwf/9+CIKADz74QKd9ddb+Fls7MpTmWsHz122AZxfqampq4OLiolN++PBhbNq0Cenp6ejXr5+2PDMzE1lZWfjwww8NWq8m0Yv1HoNunZCVSiU2bdqE8vJyfPzxx9pyTaf+tm3b9J5rqnny/6BBg/Dtt9+isLAQYWFhiI+Px5YtWwxab0FBAQICAuDs7IwLFy7gyZMniI2NbXW5uro6lJaWonfv3gbF2JVsbGwAoMlEoIm5sz2/P7oDzcHclhsMfHx8sHTpUuTl5WHdunXa8s7a32JrR4Zyd3eHUqnE7du3dcqvX78OABg6dKi2bOfOnTh48CDOnDmDXr166cy/d+9enD59GmZmZtp/Rpp9sn79egiCgJ9//lk7f21tLQA0eQYuBt06IQPAe++9h9deew0pKSna1xX16dMHMpkMly5danKZwsJCXL16FcCzBr1x40aMGDFCW9aarKws1NXV4eOPP4aHh4f2zr7WpKeng4gwZsyYVmPsaoMHD0aPHj10Gi8AXLhwAbW1tXj11VcBABYWFs1227TV8/ujs+s2BkdHRwiC0ObxxevWrYOXlxcuXryoLTN0f7dGbO3IUBYWFpg0aRLOnj2LxsZGbXlaWhoEQYC/vz+ICGFhYcjKykJycnKTvyb279+v949I84srIiICRKTTLaT57pycnIy8he3T7ROyIAjYsWMHBEFAcHAwHj9+DJlMhvnz5+PQoUPYs2cPysrK0NDQgLt37+K3335DYWEhFixYgJycHNTW1uLixYu4ffu2NjG0RjNU59SpU6ipqUFeXl6TfX6NjY14/Pgx6uvrcfnyZYSEhMDNzQ3z5s1rNcauJpPJsGzZMiQlJeHgwYMoKytDVlYWFi5cCBcXFwQFBQF4Nub70aNHSE5ORl1dHYqLi/XOcuzs7FBYWIj8/HyUl5drk2xL+6MjdaelpXXJsDeFQgEPDw/cvXu3Tctpui6eH09r6P42pO7W2lFgYCCcnJw65XbtzqwrMjISDx48wJo1a1BZWYnz588jLi4O8+bNw8CBA3H16lVs3rwZX3zxBSQSid6jEwz9Rfs8zXc3ZMiQDsdvFMa+aohOGGXxz3/+kwYMGKAdWdGrVy9asGCBzjzz5s0jAGRjY0MbN26kp0+fUlhYGLm5uZGFhQU5ODjQ9OnTKTs7m/Lz88nX15dsbW3J3NycevXqRREREVRfX0+xsbEkl8sJAPXv359u3LhBBw8eJFtbWwJAvXv3pitXrlBYWBjZ2dmRjY0NzZgxg3bt2kUASK1WU0FBAQUFBZFEIiFXV1eysLAglUpFU6dOpRs3bmhjbinGjmrPVeHGxkaKi4uj/v37k0QiIVtbWwoICKDc3FztPA8fPqQ333yTZDIZubu70yeffEKhoaEEgDw9PamgoIB++eUX6tu3L8nlcnr99dfp/v37Bu2P9tZ94sQJUiqVFBMT0+b91Nb2GRwcTBKJhKqqqrRlSUlJ2hFA9vb22lEVLwoNDdUZ9tba/t69ezcpFAqdtvj555+TSqUiANS3b1+6du1aq+0oICCAANDq1atb3Lbz58/T2LFjycXFRXusOTs7k6+vL2VkZHR6XUTPhr+NHj2apFIpubi4UGhoKNXU1BARUVZWlt6oquc/TQ2ZI2p5lMXkyZPJ1dWVGhsbW4z/RTzsrZsLCgoiOzs7k61fbM+yMPX+aE5b22deXh5ZWFjQgQMHjBhV52poaKBx48bR3r17RVVXVyspKSGZTEZbtmxp87I87O0l0C2eLtWFXob94enpiejoaERHR7f4xDGxaGhoQHJyMsrLyxEYGCiaukwhKioKw4cPR3BwsKlDaRYnZMbaKDw8HDNmzEBgYKDoHyCUnp6OxMREpKWlGTx+uivq6mpbt27FpUuXcOLECZ17FsSGE7IRrFy5Evv378eTJ0/g7u6Oo0ePmjokk3oZ98f69esRHByMjRs3mjqUFo0fPx7ffPONzjNCxFBXVzp27BiePn2K9PR02NramjqcFglExn3KhiAIiI+Px8yZM425GvaChIQEzJo1S7wPUREJbp/MEF11PPEZMmOMiQQnZMYYEwlOyIwxJhKckBljTCS65KLemDFjus0DZF4Wd+/exY8//oh33nnH1KGI2tGjR7l9slZpjidjX9QzekKeMWOGMatnDFlZWQBE/HwC9tI4cuSIUes3ekJmzNg0Q9Y0T/tjrLviPmTGGBMJTsiMMSYSnJAZY0wkOCEzxphIcEJmjDGR4ITMGGMiwQmZMcZEghMyY4yJBCdkxhgTCU7IjDEmEpyQGWNMJDghM8aYSHBCZowxkeCEzBhjIsEJmTHGRIITMmOMiQQnZMYYEwlOyIwxJhKckBljTCQ4ITPGmEhwQmaMMZHghMwYYyLBCZkxxkSCEzJjjIkEJ2TGGBMJyMJdIgAAIABJREFUTsiMMSYSnJAZY0wkOCEzxphIcEJmjDGR4ITMGGMiwQmZMcZEghMyY4yJBCdkxhgTCYGIyNRBMGaor7/+Glu3bkVDQ4O2rKSkBABgb2+vLTM3N8fSpUvx3nvvdXmMjLUXJ2TWrVy7dg0DBw40aN7c3FwMGDDAyBEx1nm4y4J1KwMGDMCwYcMgCEKz8wiCgGHDhnEyZt0OJ2TW7bz33nswNzdvdrqFhQXef//9LoyIsc7BXRas2yksLESfPn3Q2NjY5HRBEHDnzh24urp2cWSMdQyfIbNup1evXvD19YWZmX7zNTMzw9ixYzkZs26JEzLrlubOndtkuSAIPLKCdVvcZcG6pcePH8PJyQl1dXU65RYWFrh//z569uxposgYaz8+Q2bdkq2tLf70pz/pXNwzNzfHn//8Z07GrNvihMy6rTlz5uhc2CMizJkzx4QRMdYx3GXBuq2qqir07NkTNTU1AACZTIaSkhJYWVmZODLG2ofPkFm3pVAoEBAQAIlEAolEgoCAAE7GrFvjhMy6tXfffRd1dXWoq6vDu+++a+pwGOsQC0NnTEhIMGYcjLVLQ0MDFAoFiAhlZWXcTpkozZw506D5DO5DbunZAYwxxppn6KW6NnVZxMfHg4j48xJ/4uPjtQ2ou3zS09ORkZHRpevk44E/hnw0x5OhDO6yYEysxo0bZ+oQGOsUnJBZt9fUMy0Y6464JTPGmEhwQmaMMZHghMwYYyLBCZkxxkTCJAn5xIkTsLa2xrffftsl69uyZQscHR0hCAL+9re/dck6u1J0dDS8vb2hUqkglUrh6emJFStWoKKiwmQxdfV33J2cOnUK4eHhSExMhIeHBwRBgCAITT7jecKECVAqlTA3N8egQYPwyy+/mCBiwzQ2NmLbtm3w9fU1el2ZmZkYO3YsFAoFXFxcEBYWhqdPn2qnt+eYqKmpgZeXF1atWgUAOH78OGJjY3XecG5sJknIRF37PKPly5fj3LlzXbrOrnTmzBksXrwY+fn5KCkpwYYNG7B9+3bMmDHDZDF19XfcXaxZswY7duzAypUrMX36dNy8eRNqtRo9e/bEwYMHkZqaqjP/999/jyNHjmDKlCnIzs7GiBEjTBR5y/Ly8vDHP/4RS5cuRVVVlVHrys7OxoQJEzB+/HgUFxcjKSkJ+/btw8KFC7XztOeYiIiIQG5urvZvf39/yGQyjB8/HqWlpR3aJoORgQBQfHy8obOLTl5eHgGgv/71r6YOpdNNnjyZ6uvrdcpmzpxJAKigoKBNdcXHx1MbmoVoVVVVkY+Pj9Hqb8/xsHHjRhowYABVV1frlKvVavrmm2/IzMyMXF1dqbS0VGd6Wloavf322x2O2VguXbpE06ZNo4MHD9Lw4cNp2LBhRq1r1qxZ5O7uTo2NjdqyuLg4EgSBfv31VyJq+zHxz3/+kyZMmEAAKCIiQmdacHAw+fj4UF1dXZu3p63HE/chvwRSUlL03sJsb28PAB0+W+mu9u7di6KiIlOHoXX9+nVERkZi7dq1kMlketN9fX0REhKCe/fuYfny5SaIsP2GDRuGxMREzJ49G1Kp1Kh11dfXIzU1FX5+fjqPc5g4cSKICMeOHQPQtmOiuroaoaGh2L59e5MxRUVF4dKlS81O70xGS8gZGRkYPXo0FAoFVCoVhgwZgrKyMmRmZsLNzQ2CIGDXrl0AgO3bt8PKygpmZmZ49dVX4eT0/7F352FRXOn+wL+NNL3ZzSIIiIAsbrj+jMaAcRjj3ESHG0dAI0aTUWcmmIlpcSVuxCAuiFGuRMbrcrn3qhGMEEyMJvOog44T45M8gZFgVKKioCKgQLMKwvv7w9s9dti66aa72ryf5+nn4TlV59Spc6peqk+fqnKHWCyGQqHAmDFjMHHiRHh7e0MqlcLJyQkrV67UbWfnzp2QSqXo27cvFi5cCE9PT0ilUoSEhODixYud1rGlpQVxcXHw8fGBTCbDyJEjjbrVcevWrZDL5VAqlSgrK8OyZcvg5eWFxYsXQ6FQQC6X49ixY5g6dSpUKhX69++Pw4cPAwBSU1O7XMcUd+7cgUwmg5+fn8llGau9PjZkfw3pS7VaDQcHB3h4eOi2984770ChUEAkEqGiogIxMTFYtmwZrl+/DpFIhMDAQADAl19+CZVKhY0bN1q4RZ7sGxFh2rRpHa6TkJCAQYMGYd++fTh16lSH6xERtm/fjqFDh0IikcDZ2RnTp0/HlStXABh+bJl6/FvDjRs3UFtbCx8fH730gIAAAMClS5c6zNvRObFmzRq88847cHNzazefs7MzQkNDkZyc3PNDcYZeSsOIr2i1tbWkUqkoMTGRGhoaqLS0lCIiIqi8vJyIiIqLiwkApaSk6PK8//77BIAuXrxIdXV1VFFRQVOmTCEA9MUXX1B5eTnV1dWRWq0mAJSXl6fLGx0dTQqFgi5fvkyNjY1UUFBA48aNI6VSqft60t6QxfLly0kikdDRo0epsrKSVq9eTXZ2dvTtt98a2iy0Zs0aAkCLFy+mlJQUioiIoB9//FGXfvr0aaqurqaysjKaOHEiKRQKampq0svb2TrdUVdXR0qlktRqtdF5zTVk0V4fG7K/hvTlnDlzyN3dXW97SUlJBEB3jEVGRlJAQIDeOsePHyelUknx8fEm758x5wMRkb+/PwUFBbW7LCAggG7evElERF9//TXZ2dnRgAEDqLa2lojaDlnExcWRg4MDHThwgKqqqujSpUs0ZswYcnV1pdLSUiIyrK3Ncfz/3Pjx400asuiqrLNnzxIASkpKarO+TCajyZMnt1tWR+fE+fPnadq0aUREVF5e3u6QBRHRqlWrCADl5uYatQ+CGLIoKiqCRqPBsGHDIJVK4e7ujszMTN1Xhs4EBQVBLpejT58+mD17NgDAx8cHrq6ukMvlulf0aK8GtOzt7XVXDEFBQUhNTUVNTQ3S0tLa3U5jYyNSU1MRHh6OyMhIODk5Ye3atRCLxR3m6cyWLVuwaNEiZGZmYsiQIbr0kJAQqFQquLm5ISoqCnV1dbh9+7ZeXkPWMcamTZvg6emJhISEbpfRk7raX2P70lBhYWHQaDRYt26dqbtglLq6Oty8eVN3FdeZ4OBgLFmyBEVFRXjvvffaLG9oaMD27dsRERGBuXPnwtHRESNGjMDu3btRUVGBPXv26K3fUVub+/i3FO1Mip8PRwCAWCxGQ0NDu/naOycaGhoQExOD1NTULrc7cOBAAEB+fn53qm2wHgnI/v7+6Nu3L+bOnYv169ejqKioW+U4ODgAeDJupCUWiwGgzduGf27s2LGQy+VtArfW1atXUV9fj+HDh+vSZDIZPDw8OsxjKu3+dFZ3Q9bpTFZWFo4cOYKvvvoKSqWyW2VYkiH721VfCl1ZWRmICHK53KD1ExISMHjwYOzatQvnz5/XW1ZQUIDa2lqMHTtWL33cuHFwcHDodJju6ba2xvFvDtrx96djglZTUxNkMlmb9I7OidWrV+Ott96Cl5dXl9vV9t39+/e7W3WD9EhAlslkOHPmDF588UVs3LgR/v7+iIqK6vC/V0+RSCQoLy9vd1ldXR0AYO3atbp5oCKRCLdu3bLZH8LS09OxZcsW5OTkYMCAAdaujll11pdCp33nn6E/eEmlUqSlpUEkEmHBggV65412+lXv3r3b5HNyckJNTY1B27DV41/724FGo9FLr6+vR2NjIzw9PfXSOzonzp8/j/z8fPzxj380aLvaQK/ty57SYz/qDRs2DJ9//jnu3r2L2NhYZGRkYNu2bT21uTaam5tRVVWF/v37t7tcO4C/Y8eONs8wvXDhgsXqaS4pKSk4ePAgzpw5g379+lm7OmbVVV8KnfZkNuYGg+DgYCxduhSFhYXYsGGDLt3JyQkA2g28xrSRrR7/fn5+UCqVuHXrll76Tz/9BAAYOXKkLq2zc2L//v04ffo07OzsdP+MtG2yceNGiEQifPfdd7r1m5qaAKDdK3Bz6pGAfPfuXVy+fBnAk47fvHkzxowZo0uzhJycHBARXnjhhXaXa2dt5OXlWaxOPYGIEBsbi/z8fGRnZ7d75WTrft6X9vb23R7SsQbtXaLV1dVG5duwYQOGDBmC3NxcXdrw4cPRu3dvvWABABcvXkRTUxOee+45g8q21ePf3t4ev/3tb3Hu3Dm0trbq0k+ePAmRSIRp06YZdE6kpaW1+Uek/Qa2Zs0aEJHesJC279zd3Xt0/3osIC9cuBBXrlxBU1MTcnNzcevWrQ6Dozm0traisrISjx8/xqVLlxATEwMfHx/Mmzev3fWlUinmz5+Pw4cPIzU1FRqNBi0tLSgpKcG9e/d6rJ7mdvnyZWzduhV79+6FWCzW+/opEoks+q3EXLrqy8DAQDx8+BDZ2dlobm5GeXl5mysmFxcX3L17F0VFRaipqUFzczNOnjxplWlvcrkc/v7+KCkpMSqfduji6R+wpFIpli1bhqysLBw8eBAajQb5+fl4++234enpiejoaIPL7ur4j4qKgru7u1lu1zZnWevWrcP9+/fx/vvvo66uDhcuXEBSUhLmzZuHwYMH98g5oe27ESNGmFz/Thk6HQNGTPMpKiqikJAQcnZ2pl69elG/fv1ozZo19PjxY0pJSSEPDw8CQHK5nKZNm0bJyckkl8sJAA0YMID+/ve/05YtW8jR0ZEAkLu7Ox06dIjS09PJ3d2dAJCzszMdPnyYiJ5MlRKLxeTl5UX29vakUqlo+vTpdP36dSIi+vDDD3X5FAoFRUREEBHRo0ePKDY2lnx8fMje3p7c3NwoMjKSCgoKDNrPxMREkslkBIC8vb3pwIEDRES0a9cu3f4MHDiQrl+/Tnv27CGVSkUAyNfXl2JiYrpc59q1a13WIT8/nwB0+GlvelBnzDHtrb0+NqRNrl271mVfEhE9ePCAJk2aRFKplPz8/Ojdd9+lFStWEAAKDAyk27dv0/fff0++vr4kk8noxRdfpNLSUjpx4gQplUpKSEgwaf+IjJ/2plarSSwWU319vS4tKyuLAgICCAC5urrSokWL2s27YsUKvWlvra2tlJSURAMHDiSxWEzOzs4UHh5OV69eJSLDjr9r1651efyHh4cTAIqLi+t03y5cuEATJkwgT09P3XHn4eFBISEhdPbsWbOXRfRk+tvzzz9PEomEPD09acWKFdTY2EhE3T8nOpv2FhYWRl5eXnp3BxrC2PPpmbh1Ojo6mlxcXKxdjWeCtW+dtpW+NPZ8KCwsJHt7e90/bVvQ0tJCEydOpP379wuqLEurqKggqVRK27ZtMzqvIOYhW4Mln8jEetaz2JeBgYGIj49HfHy8VZ/CZ6iWlhZkZ2ejpqYGUVFRginLGtavX4/Ro0dDrVb3+LaemYBsTleuXGkz7tTepycPLiHUgZnXqlWrMHPmTERFRRn9A5+l5eTkIDMzEydPnjR4/rQlyrK07du3Iy8vDydOnNDdA9GjDL2UhkCHLFatWkUODg668edPPvnE2lWyadYcsrClvjTlfPjqq68oNjbWzDVi5padnU2bNm1q89Q4Yxh7PomIDHtahkgkQkZGBl577bWe++/ArO7IkSOYNWsWP8+4C3w+MEMYez7xkAVjjAkEB2TGGBMIDsiMMSYQHJAZY0wgjPpR74UXXrDZB7www5SUlOCbb77BjBkzrF0VQTt69CifD6xL2vOJf9RjjDEbw9PemB6e9mYYPh+YIXjaG2OM2SgOyIwxJhAckBljTCA4IDPGmEBwQGaMMYGwSkD+5z//iaioKPj5+UEikcDV1RWjRo1CQkKCbp0TJ07A0dERn3/+uUXqtG3bNt27z3bv3m1QnszMTPj7+7d5JKZUKoWfnx8WLFiAmzdvdrj+G2+80abMl19+GUqlEr169cKwYcOwZs0ao/OY4zU5zLJOnTqFVatWdesYEXJ/t7a2YseOHQgJCenRsuLj4xEUFASVSgWJRILAwECsXLmyzbOnP/74Y4wbNw5KpRK+vr6YP38+SktLAQCfffYZEhMTrfs8bkMfCwczPX7z0qVLJJfLafHixXTz5k1qaGigq1ev0sqVK2ny5Mm69Y4fP04qlYo+++wzk7dpqMLCQgJAf/nLX4zKFxAQQI6OjkT05M0I9+/fp//93/8luVxOffv2pYqKijbr9+nThwDQ8ePH25R38uRJvVf2dDdPd1j7jSG2wlznAxFRXFwcvfrqq6TRaHRplurvnnTt2jWaMGECAaBRo0b1aFmhoaG0a9cuevDgAWk0GsrIyCCxWExTpkzRrZOenk4AKDExkaqqqig3N5f8/f1p9OjR1NzcTEREycnJFBoaSpWVlSbVV0vwr3B68803qV+/fm3SHz16RP/+7/9ucvmmMEdAftrKlSsJAKWnp7dZ/9ChQ2RnZ0deXl5UVVWlt7yjgGxsnu6wZkCur6+n4OBgmyjbXOfD5s2badCgQdTQ0KCXbqn+7il5eXkUERFBBw8epNGjR5sUkA0pKywsrM1zi1977TUCQLdv3yYiokmTJlG/fv303ov30UcfEQA6f/68Lk2tVlNwcLAuSJtC8K9wevDgAaqrq/Hw4UO9dAcHB4sNT1hKYGAgAOi+Ej0tJCQEMTExuHPnDpYvX25Qed3JY0v279+PsrIymyu7u3766SesW7cOH3zwAaRSaZvlttzfo0aNQmZmJubMmQOJRNLjZR0/flzv7dwA4OrqCgCor68HABQXF8PT0xMikUi3jre3NwDovbV8/fr1yMvLQ3Jyskn17g6LB+Rx48ahrq4OL730Ev7xj3+0u8758+fh4+MDkUiEjz76CACQnJwMhUIBOzs7PPfcc3B3d4dYLIZCocCYMWMwceJEeHt7QyqVwsnJCStXrtSVt3PnTkilUvTt2xcLFy6Ep6cnpFIpQkJCcPHixU7r29LSgri4OPj4+EAmk2HkyJHIyMgwaF8LCwsBPDmg2pOQkIBBgwZh3759OHXqlEFldidPTyMibN++HUOHDoVEIoGzszOmT5+OK1euAADUajUcHBzg4eGhy/POO+9AoVBAJBKhoqICMTExWLZsGa5fvw6RSITAwECD+q27ZQPAl19+CZVKhY0bN1qwtf5l586dICJMmzatw3UM7e+u+iA1NRUKhQJyuRzHjh3D1KlToVKp0L9/fxw+fFhXjinHu9DcuXMHMpkMfn5+AAB/f/82/5S1F0v+/v66NGdnZ4SGhiI5Odnyd6waeikNM31Fq6+vp7Fjx+peyR0UFESJiYn04MEDvfWKi4sJAKWkpOjS3n//fQJAFy9epLq6OqqoqKApU6YQAPriiy+ovLyc6urqSK1WEwDKy8vT5Y2OjiaFQkGXL1+mxsZGKigooHHjxpFSqdR9pWlvyGL58uUkkUjo6NGjVFlZSatXryY7Ozv69ttvdev8fMiisrKS/vu//5vkcjmFhYW1aYOAgAC6efMmERF9/fXXZGdnRwMGDKDa2loi6njIwtg83dGdIYu4uDhycHCgAwcOUFVVFV26dInGjBlDrq6uVFpaSkREc+bMIXd3d718SUlJBIDKy8uJiCgyMpICAgL01jGk37pb9vHjx0mpVFJ8fLxR+0tknvPB39+fgoKC2l1mbH8b0gdr1qwhAHT69Gmqrq6msrIymjhxIikUCmpqaiIiw453Y40fP97kMWRjy6qrqyOlUklqtVqXlpOTQ2KxmHbu3EkajYZ++OEHGjp0KL3yyitt8q9atYoAUG5urkn1FfyQhUwmw9dff43/+I//wJAhQ3D58mXExsZi6NChOHv2rEFlBAUFQS6Xo0+fPpg9ezYAwMfHB66urpDL5Zg7dy4A6K4OtOzt7XVXEEFBQUhNTUVNTQ3S0tLa3U5jYyNSU1MRHh6OyMhIODk5Ye3atRCLxW3yVFdX634Zd3Z2xvz587F69Wp8+umnne5LcHAwlixZgqKiIrz33nsG7X938vSUhoYGbN++HREREZg7dy4cHR0xYsQI7N69GxUVFdizZ4/J2zC23wwVFhYGjUaDdevWmVxHY9XV1eHmzZsICAjoct2u+tvYPggJCYFKpYKbmxuioqJQV1eH27dvG3W8C92mTZvg6empN3MrNDQUsbGxUKvVUKlUGD58OGpqarBv3742+QcOHAgAyM/Pt1idAStNexOLxVCr1fjxxx/xzTffYPr06SgrK8PMmTNRWVlpVFkODg4AgMePH+uVDwDNzc2d5h07dizkcnmbwK119epV1NfXY/jw4bo0mUwGDw+PNnkcHR1BT34kxYoVK0BEcHR0NOhNtQkJCRg8eDB27dqF8+fPd7l+d/P0hIKCAtTW1mLs2LF66ePGjYODg0OXQ0Ld0VW/2YKysjIQkcFvYe6sv03pA+3509zcbNTxLmRZWVk4cuQIvvrqKyiVSl36mjVrsGfPHpw+fRq1tbW4ceMGQkJCEBwcjOLiYr0ytP1y//59i9bd6jeGjB8/Hp9++inefvttlJeX429/+5tFty+RSFBeXt7usrq6OgDA2rVr9eYZ37p1S/dDQXvWrVsHDw8PrF69uk1Ht0cqlSItLQ0ikQgLFixAQ0NDj+TpCVVVVQCA3r17t1nm5OSEmpqaHtluZ/1mCxobGwHA4B+8Outvc/VBd493IUlPT8eWLVuQk5ODAQMG6NLv3buHxMREvPXWW3jppZegUCjg5+eHvXv34u7du0hKStIrRyaTAfhXP1mKxQNyZGSk3tWslnYCvCU7vrm5GVVVVR0+ZNzNzQ0AsGPHDt3Vr/Zz4cKFDstVKpXYsmULampq8Oc//9mgugQHB2Pp0qUoLCzEhg0beiyPuTk5OQFAuyd9Z21riq76zRZoT3hjbkLoqL/N1QfdPd6FIiUlBQcPHsSZM2fQr18/vWWFhYVoaWlpk65SqeDi4oKCggK99KamJgD/6idLsXhAfvToES5fvtwm/erVqwCAkSNHWqwuOTk5ICK88MIL7S7XztrIy8szuuw333wT48ePx/Hjx3HkyBGD8mzYsAFDhgxBbm6uwdvpTh5zGj58OHr37o3vvvtOL/3ixYtoamrCc889B+DJOHBXQ0iG+nm/mbNsS9HeFVpdXW1Uvvb629A+6Iopx7s1ERFiY2ORn5+P7Ozsdr8paP8p3bt3Ty+9pqYGDx8+1E1/09L2i7u7ew/Vun1WGbIIDw/HkSNHUFVVherqahw7dgzvvfcefve73/VoQG5tbUVlZSUeP36MS5cuISYmBj4+Ppg3b16760ulUsyfPx+HDx9GamoqNBoNWlpaUFJS0qZjf04kEmHnzp0QiURQq9UGjY1rv5b+fD6lufOYk1QqxbJly5CVlYWDBw9Co9EgPz8fb7/9Njw9PREdHQ3gyZzshw8fIjs7G83NzSgvL9eb+wkALi4uuHv3LoqKilBTU6MLsl31W3fLPnnypNWmvcnlcvj7+6OkpMSofO31t6F9YEjZXR3vUVFRcHd3N8vt2uYq6/Lly9i6dSv27t0LsVjc5lEG27Ztg5+fHyZNmoS9e/fi3LlzaGhoQHFxsa5t/vCHP+iVqe2XESNGmFQ3oxk6HQNmmvb217/+lWbNmkUBAQEkkUjIwcGBBg8eTOvXr6fGxkYiIkpJSSEPDw8CQHK5nKZNm0bJyckkl8sJAA0YMID+/ve/05YtW8jR0ZEAkLu7Ox06dIjS09PJ3d2dAJCzszMdPnyYiJ5MnxKLxeTl5UX29vakUqlo+vTpdP36dSIi+vDDD3X5FAoFRUREENGTOwhjY2PJx8eH7O3tyc3NjSIjI6mgoID+8Y9/0KBBg3RT+Pr160cLFy7U29958+YRAHJycqLnn3+eAgICCAC5urrSokWL2m2jFStW6KY0ZWVlGZ3HFN2Z9tba2kpJSUk0cOBAEovF5OzsTOHh4XT16lXdOg8ePKBJkyaRVColPz8/evfdd2nFihUEgAIDA+n27dv0/fffk6+vL8lkMnrxxReptLS0y34zpewTJ06QUqmkhIQEo9vJHOeDWq0msVhM9fX1urTu9ndXfbBr1y7d+TNw4EC6fv067dmzh1QqFQEgX19funbtWqfHOxFReHg4AaC4uLhO9+3ChQs0YcIE8vT01J0fHh4eFBISQmfPnjVrWfn5+br09j5JSUlERFRRUUExMTEUGBhIEomEevfuTRMmTKBPP/20zTbDwsLIy8tL766+7hD8rdPWEh0dTS4uLtauhuAJ7VkWQu03c5wPhYWFZG9vTwcOHDBTrXpeS0sLTZw4kfbv3y+ossypoqKCpFIpbdu2zeSyBD8P2Zqs+hQn1m3Par8FBgYiPj4e8fHxbZ5KJkQtLS3Izs5GTU0NoqKiBFOWua1fvx6jR4+GWq22+LZ/UQGZMaFZtWoVZs6ciaioKKN/4LO0nJwcZGZm4uTJkwbPn7ZEWea0fft25OXl4cSJEwbdQ2Buv4iAvHr1aqSlpaG6uhp+fn44evSotavEDPBL6beNGzdCrVZj8+bN1q5KpyZPnoxDhw7pPTdECGWZy7Fjx/Do0SPk5OTA2dnZKnUQERn29Ax+7fkvg7GvLf+l4vOBGcLY8+kXcYXMGGO2gAMyY4wJBAdkxhgTCA7IjDEmEByQGWNMIIyaZcEYY8x4hs6ysDe0QFt9rxZ79u3YsQMAsGTJEivXhDHTGHyFzJhQaecCG/qYU8aEiseQGWNMIDggM8aYQHBAZowxgeCAzBhjAsEBmTHGBIIDMmOMCQQHZMYYEwgOyIwxJhAckBljTCA4IDPGmEBwQGaMMYHggMwYYwLBAZkxxgSCAzJjjAkEB2TGGBMIDsiMMSYQHJAZY0wgOCAzxphAcEBmjDGB4IDMGGMCwQGZMcYEggMyY4wJBAdkxhgTCA7IjDEmEByQGWNMIDggM8aYQHBAZowxgeCAzBhjAsEBmTHGBIIDMmOMCQQHZMYYEwgOyIwxJhD21q4AY8aoqKiARqPRS6urqwMA3LhxQy9dpVLB1dXVYnVjzFQiIiJrV4IxQ6Wd9lAGAAAgAElEQVSlpWHBggUGrftf//VfmD9/fg/XiDHz4YDMbEp1dTXc3NzQ3Nzc6XpisRjl5eVwdHS0UM0YMx2PITOb4ujoiN/+9rewt+94tM3e3h5hYWEcjJnN4YDMbM7cuXPR0tLS4fLW1lbMnTvXgjVizDx4yILZnMbGRri6uup+zPs5uVyOiooKyGQyC9eMMdPwFTKzOVKpFBERERCLxW2WicVizJgxg4Mxs0kckJlNev3119v9Ya+5uRmvv/66FWrEmOl4yILZpMePH8Pd3R0PHz7US3dyckJ5eXmnP/oxJlR8hcxskr29PWbPnq03bCEWizF37lwOxsxmcUBmNmv27Nl6wxbNzc2YPXu2FWvEmGl4yILZLCKCt7c37ty5AwDw9PTEnTt3IBKJrFwzxrqHr5CZzRKJRHjjjTfg4OAABwcH/P73v+dgzGwaXyEzm3bp0iWMGjVK9/eIESOsXCPGus/iv37MnDnT0ptkz7jevXsDAOLj461cE/as+eSTTyy6PYsPWRw9ehQlJSWW3iz7P89i+/v6+mLAgAFmK++bb77BN998Y7bymO0pKSnB0aNHLb5diw9ZiEQiZGRk4LXXXrPkZtn/eRbbX/scZH9/f7OUp/0WZ+mrIyYcR44cwaxZs2DpEV2esMlsnrkCMWPWxrMsGGNMIDggM8aYQHBAZowxgeCAzBhjAmFzAfmPf/wjlEolRCIR8vLyrF0di2hubkZcXBz8/f3h4OAALy8vLF++HA0NDVapz4kTJ+Do6IjPP//cKtsXulOnTmHVqlXIzMyEv78/RCKR7q7Cn3v55ZehVCrRq1cvDBs2DN9//70VamyY1tZW7NixAyEhIT1aVnx8PIKCgqBSqSCRSBAYGIiVK1eitrZWb72PP/4Y48aNg1KphK+vL+bPn4/S0lIAwGeffYbExMRO3ywjSGRhACgjI8OkMg4fPkwAKDc310y1ErY///nPJJVK6fDhw6TRaOhvf/sbqVQqev31140uyxztf/z4cVKpVPTZZ5+ZVI5QzZgxg2bMmNGtvHFxcfTqq6+SRqPRpQUEBFCfPn0IAB0/frxNnpMnT9Lvfve7btfXEq5du0YTJkwgADRq1KgeLSs0NJR27dpFDx48II1GQxkZGSQWi2nKlCm6ddLT0wkAJSYmUlVVFeXm5pK/vz+NHj2ampubiYgoOTmZQkNDqbKy0ug6ZmRkkBXCI3FAtqD6+noKDg42Ks/169fJzs6O3nrrLb30tWvXEgC6fPmyUeWZo/2trTvtaIzuBuTNmzfToEGDqKGhQS89ICCADh06RHZ2duTl5UVVVVV6y4UekPPy8igiIoIOHjxIo0ePNikgG1JWWFgYPX78WC/ttddeIwB0+/ZtIiKaNGkS9evXj1pbW3XrfPTRRwSAzp8/r0tTq9UUHBysC9KGslZAtrkhCwA2+wCZ/fv3o6yszKg83377LVpbWzF+/Hi99ClTpgAAvvrqK7PVz1Z0px172k8//YR169bhgw8+gFQqbbM8JCQEMTExuHPnDpYvX26FGnbfqFGjkJmZiTlz5kAikfR4WcePH0evXr300lxdXQEA9fX1AIDi4mJ4enrqxQJvb28AwK1bt3Rp69evR15eHpKTk02qt6UIPiATEZKSkjB48GBIJBI4OjpixYoVuuVbt26FXC6HUqlEWVkZli1bBi8vL1y9ehVEhO3bt2Po0KGQSCRwdnbG9OnTceXKFQDAzp07IZVK0bdvXyxcuBCenp6QSqUICQnBxYsX9erQWTlqtRoODg7w8PDQ5XnnnXegUCggEolQUVGBmJgYLFu2DNevX4dIJEJgYKBB+29n96SLfv6OuIEDBwIAfvzxx260avedP38ePj4+EIlE+OijjwAAqampUCgUkMvlOHbsGKZOnQqVSoX+/fvj8OHDAAxra1Pa8csvv4RKpcLGjRst2h5aO3fuBBFh2rRpHa6TkJCAQYMGYd++fTh16lSH63V1vBnS3gDQ0tKCuLg4+Pj4QCaTYeTIkcjIyDDfTlvQnTt3IJPJ4OfnB+DJzUA//6esHT9++kYhZ2dnhIaGIjk52eJ33XWLpS/JYeRX5jVr1pBIJKIPP/yQKisrqb6+nnbt2qU3ZLFmzRoCQIsXL6aUlBSKiIigH3/8keLi4sjBwYEOHDhAVVVVdOnSJRozZgy5urpSaWkpERFFR0eTQqGgy5cvU2NjIxUUFNC4ceNIqVTqvh4ZUs6cOXPI3d1dr+5JSUkEgMrLy4mIKDIykgICAoxqr0uXLhEAWrdunV7648ePCQCFh4cbVZ6x7d+e4uJiAkApKSm6NG0fnD59mqqrq6msrIwmTpxICoWCmpqaiMiwtu5uOx4/fpyUSiXFx8ebtG9E3Ruy8Pf3p6CgoHaXBQQE0M2bN4mI6OuvvyY7OzsaMGAA1dbWElHbIQtDjjdD2nv58uUkkUjo6NGjVFlZSatXryY7Ozv69ttvjW0SnfHjx5s8hmxsWXV1daRUKkmtVuvScnJySCwW086dO0mj0dAPP/xAQ4cOpVdeeaVN/lWrVhk9xMlDFu1oaGjAjh078Jvf/AZLly6Fk5MTZDIZXFxc2l1/y5YtWLRoETIzM+Hr64vt27cjIiICc+fOhaOjI0aMGIHdu3ejoqICe/bs0eWzt7fXXY0EBQUhNTUVNTU1SEtLQ0NDg8Hl9IQRI0ZgypQp2LVrF86cOYPGxkaUlpYiKysLIpGo3Rd9WlNISAhUKhXc3NwQFRWFuro63L59W7e8s7Y2RVhYGDQaDdatW2fqLhitrq4ON2/eREBAQJfrBgcHY8mSJSgqKsJ7773XZrmxx1tH7d3Y2IjU1FSEh4cjMjISTk5OWLt2LcRiscltbWmbNm2Cp6cnEhISdGmhoaGIjY2FWq2GSqXC8OHDUVNTg3379rXJr/02mZ+fb7E6d5egA/JPP/2E+vp6TJ482ei8BQUFqK2txdixY/XSx40bBwcHB70hiZ8bO3Ys5HI5rly5YlI55pKeno6ZM2fizTffhIuLCyZMmIBPP/0URIQ+ffr0+Pa7y8HBAQA6/afxdFvbqrKyMhAR5HK5QesnJCRg8ODB2LVrF86fP6+3zJTj7en2vnr1Kurr6zF8+HDdcplMBg8PD5tq66ysLBw5cgRfffUVlEqlLn3NmjXYs2cPTp8+jdraWty4cQMhISEIDg5GcXGxXhnafrl//75F694dgg7I2sdEurm5GZ23qqoKwL+elfs0Jycn1NTUdJpfIpGgvLzc5HLMwdHREbt370ZJSQnq6+tx/fp1fPjhhwCAfv369fj2e5q2rW1VY2MjABj8g5dUKkVaWhpEIhEWLFigN5/cXMdbXV0dAGDt2rW6edAikQi3bt3S/TAmdOnp6diyZQtycnL0Hq967949JCYm4q233sJLL70EhUIBPz8/7N27F3fv3kVSUpJeOdrfX7T9JGSCDsjaX6sfPXpkdF4nJycAaPcArqqqQv/+/TvM29zcrFvHlHJ60rfffgsAmDRpklW2by5Pt7Wt0p7wxtyEEBwcjKVLl6KwsBAbNmzQpZvreNNexOzYsQP0ZHqr7nPhwgWD62ktKSkpOHjwIM6cOdPmoqOwsBAtLS1t0lUqFVxcXFBQUKCX3tTUBKDtD+NCJOiAPHz4cNjZ2eHs2bPdytu7d2989913eukXL15EU1MTnnvuuQ7z5uTkgIjwwgsvGFyOvb29Rcdz9+7dCz8/P4SGhlpsmz3h6bYGLN+O5tC3b1+IRCJUV1cblW/Dhg0YMmQIcnNzdWmmHLdP8/b2hlQqtbm7WYkIsbGxyM/PR3Z2drvfFLT/lO7du6eXXlNTg4cPH+qmv2lp+8Xd3b2Ham0+gg7Ibm5umDFjBo4ePYr9+/dDo9Hg0qVLBv2QJpVKsWzZMmRlZeHgwYPQaDTIz8/H22+/DU9PT0RHR+vWbW1tRWVlJR4/foxLly4hJiYGPj4+mDdvnsHlBAYG4uHDh8jOzkZzczPKy8v15kMCgIuLC+7evYuioiLU1NQYHHief/553Lp1C48fP0ZRURGWL1+OU6dOYf/+/bpxQ1vRWVsD3W/HkydPWm3am1wuh7+/v9FvYtEOXTw959aY47arsufPn4/Dhw8jNTUVGo0GLS0tKCkp0QWyqKgouLu7m+V2bXOVdfnyZWzduhV79+6FWCzWG24RiUTYtm0b/Pz8MGnSJOzduxfnzp1DQ0MDiouLdW3zhz/8Qa9Mbb/YxPsWLT2tA0ZOu6qpqaE//elP1KdPH+rduze9+OKLFBcXRwCof//+NGfOHJLJZASAvL296cCBA7q8ra2tlJSURAMHDiSxWEzOzs4UHh5OV69e1a0THR1NYrGYvLy8yN7enlQqFU2fPp2uX79uVDkPHjygSZMmkVQqJT8/P3r33XdpxYoVBIACAwPp9u3b9P3335Ovry/JZDJ68cUXdVOYuvJv//Zv5OTkRPb29uTs7ExhYWHdnrpkbPv/XEpKCnl4eBAAksvlNG3aNNq1axfJ5XICQAMHDqTr16/Tnj17SKVSEQDy9fWla9euGdTW3W3HEydOkFKppISEhG7vm1Z3pr2p1WoSi8VUX1+vS8vKyqKAgAACQK6urrRo0aJ2865YsUJv2ltXx5uh7f3o0SOKjY0lHx8fsre3Jzc3N4qMjKSCggIiIgoPDycAFBcX1+m+XbhwgSZMmECenp4EgACQh4cHhYSE0NmzZ81aVn5+vi69vU9SUhIREVVUVFBMTAwFBgaSRCKh3r1704QJE+jTTz9ts82wsDDy8vLSu6uvK3zrtJVER0eTi4uLtathMdZsf1tp6+4E5MLCQrK3t9e7IBC6lpYWmjhxIu3fv19QZZlTRUUFSaVS2rZtm1H5eB6yFdncE6Fs2LPa1oGBgYiPj0d8fHybp5IJUUtLC7Kzs1FTU4OoqCjBlGVu69evx+jRo6FWq61dFYNwQLaSK1eutBkfa+8jtAOcdWzVqlWYOXMmoqKijP6Bz9JycnKQmZmJkydPGjx/2hJlmdP27duRl5eHEydOQCwWW7s6BvlFB+TVq1cjLS0N1dXV8PPzs+hrv4cMGdJmOlJ7n/T0dIvVqSdZs60taePGjVCr1di8ebO1q9KpyZMn49ChQ3rPDRFCWeZy7NgxPHr0CDk5OXB2drZ2dQwmIrLsEzeexdfQ2xJu/67NnDkTAPDJJ59YuSbMWo4cOYJZs2ZZ/IFEv+grZMYYExIOyIwxJhAckBljTCA4IDPGmEBwQGaMMYGwyiwLxhizBZaeZWFv0a39n5iYGAQHB1tj0794s2bN4vbvwo4dOwAAS5YssXJNmLVcuHDBKi9GtUpADg4O5nmwVjJr1ixu/y5o5x9zG/2yWSMg8xgyY4wJBAdkxhgTCA7IjDEmEByQGWNMIDggM8aYQNhsQM7MzIS/v3+b5wc7ODigb9+++PWvf42kpCRUVlZau6rsF+bUqVNYtWpVm2P0jTfeaLPuyy+/DKVSiV69emHYsGFmeb9dT2ltbcWOHTsQEhLSo2XFx8cjKCgIKpUKEokEgYGBWLlyZZsH/3/88ccYN24clEolfH19MX/+fJSWlgIAPvvsMyQmJtreCxEs/YoSmPkVQgEBAeTo6EhET95FVllZSX/7299o3rx5JBKJyNPTs9vvn3sWmbv9n0XdeYWTVlxcHL366quk0Wh0aQEBAdSnTx8CQMePH2+T5+TJk3rv1BOia9eu0YQJEwgAjRo1qkfLCg0NpV27dtGDBw9Io9FQRkYGicVimjJlim6d9PR0AkCJiYlUVVVFubm55O/vT6NHj6bm5mYiIkpOTqbQ0FCqrKw0uo78CiczEIlEcHJywq9//WukpaXhyJEjuH//PsLCwgT/Boefa2hoMMuViFD05P4Ipa22bNmC9PR0HDlyBEqlUm/Zzp07YWdnh+joaJs7Fv/5z3/ivffew9tvv43Ro0f3eFm9e/dGdHQ0XFxcoFQq8dprryE8PBxffvkliouLAQD/+Z//iX79+mHFihVwdHTE6NGjsXTpUuTl5eHixYsAgMWLF2PUqFH47W9/i8ePH5tUb0t5pgLyz82YMQPz5s1DWVkZdu/ebe3qGGX//v0oKyuzdjXMpif3Rwht9dNPP2HdunX44IMPIJVK2ywPCQlBTEwM7ty5g+XLl1uhht03atQoZGZmYs6cOZBIJD1e1vHjx9GrVy+9NFdXVwBAfX09AKC4uBienp56j2Lw9vYGANy6dUuXtn79euTl5VnlJo/ueKYDMgDMmzcPAHDy5Els3boVcrkcSqUSZWVlWLZsGby8vHD16lUQEbZv346hQ4dCIpHA2dkZ06dPx5UrVwA8ucKRSqXo27cvFi5cCE9PT0ilUoSEhOj+IwPoshy1Wg0HBwe919288847UCgUEIlEqKioQExMDJYtW4br169DJBIhMDDQcg32Mz21P4a0pylt9eWXX0KlUmHjxo0WaaedO3eCiDBt2rQO10lISMCgQYOwb98+nDp1qsP1umrz1NRUKBQKyOVyHDt2DFOnToVKpUL//v1x+PBhXTktLS2Ii4uDj48PZDIZRo4ciYyMDPPttAXduXMHMpkMfn5+AAB/f/82/4S148f+/v66NGdnZ4SGhiI5Odniz6XoFkuPkaAHx5Dbo9FoCAB5e3sTEdGaNWsIAC1evJhSUlIoIiKCfvzxR4qLiyMHBwc6cOAAVVVV0aVLl2jMmDHk6upKpaWlRPTkNfYKhYIuX75MjY2NVFBQQOPGjSOlUkm3b98mIjKonDlz5pC7u7tePZOSkggAlZeXExFRZGQkBQQEmK2dtIxt/57cH0Pas7tlHz9+nJRKJcXHxxu8r1rdGUP29/enoKCgdpcFBATQzZs3iYjo66+/Jjs7OxowYADV1tYSUdsxZEPaXHscnz59mqqrq6msrIwmTpxICoWCmpqaiIho+fLlJJFI6OjRo1RZWUmrV68mOzs7k35TGT9+vMljyMaWVVdXR0qlktRqtS4tJyeHxGIx7dy5kzQaDf3www80dOhQeuWVV9rkX7VqFQGg3Nxcg+vGY8g9RKlUQiQSoaamRi99y5YtWLRoETIzM+Hr64vt27cjIiICc+fOhaOjI0aMGIHdu3ejoqICe/bs0eWzt7fXXbkEBQUhNTUVNTU1SEtLQ0NDg8Hl2AJL7E9n7WmKsLAwaDQarFu3zuQ6dqWurg43b95EQEBAl+sGBwdjyZIlKCoqwnvvvddmubFtHhISApVKBTc3N0RFRaGurg63b99GY2MjUlNTER4ejsjISDg5OWHt2rUQi8Umt62lbdq0CZ6enkhISNClhYaGIjY2Fmq1GiqVCsOHD0dNTQ327dvXJv/AgQMBAPn5+Rarc3c98wG5rq4ORASVStXhOgUFBaitrcXYsWP10seNGwcHBwe9IYmfGzt2LORyOa5cuWJSOUJkjf15uj1tRVlZGYgIcrncoPUTEhIwePBg7Nq1C+fPn9dbZkqbOzg4AACam5tx9epV1NfXY/jw4brlMpkMHh4eNtW2WVlZOHLkCL766iu9H0rXrFmDPXv24PTp06itrcWNGzcQEhKC4OBg3Q9/Wtp+uX//vkXr3h3PfEC+du0aAGDIkCEdrlNVVQXgya+7P+fk5NTm6vrnJBIJysvLTS5HaKy1P9r2tBWNjY0AYPAPXlKpFGlpaRCJRFiwYAEaGhp0y8zV5nV1dQCAtWvX6s3Tv3Xrlu6HMaFLT0/Hli1bkJOTgwEDBujS7927h8TERLz11lt46aWXoFAo4Ofnh7179+Lu3btISkrSK0cmkwH4Vz8J2TMfkL/88ksAwNSpUztcx8nJCQDaPdirqqrQv3//DvM2Nzfr1jGlHCGyxv483Z62QnvCG3MTQnBwMJYuXYrCwkJs2LBBl26uNndzcwPw5NnORKT3uXDhgsH1tJaUlBQcPHgQZ86cQb9+/fSWFRYWoqWlpU26SqWCi4sLCgoK9NKbmpoA/KufhOyZDsilpaXYsWMH+vfvjwULFnS43vDhw9G7d2989913eukXL15EU1MTnnvuuQ7z5uTkgIjwwgsvGFyOvb09mpubTdgzy7DG/jzdnuYuu6f07dsXIpHI6PnFGzZswJAhQ5Cbm6tLM+VYfJq3tzekUiny8vKMqpO1ERFiY2ORn5+P7Ozsdr8paP8p3bt3Ty+9pqYGDx8+1E1/09L2i7u7ew/V2nyeiYBMRKitrUVrayuICOXl5cjIyMCECRPQq1cvZGdndzqGLJVKsWzZMmRlZeHgwYPQaDTIz8/H22+/DU9PT0RHR+vWbW1tRWVlJR4/foxLly4hJiYGPj4+mDdvnsHlBAYG4uHDh8jOzkZzczPKy8v15k4CgIuLC+7evYuioiLU1NRYJShZYn86a09Tyj558qTFpr3J5XL4+/ujpKTEqHzaoYun59wacyx2Vfb8+fNx+PBhpKamQqPRoKWlBSUlJbpAFhUVBXd3d7Pcrm2usi5fvoytW7di7969EIvFbR6NsG3bNvj5+WHSpEnYu3cvzp07h4aGBhQXF+va5g9/+INemdp+GTFihEl1swhLT+uAmaa9ffbZZzRy5EiSy+Xk4OBAdnZ2BIBEIhE5OTnR888/T/Hx8fTgwQNdnsTERJLJZLppcAcOHNAta21tpaSkJBo4cCCJxWJydnam8PBwunr1qm6d6OhoEovF5OXlRfb29qRSqWj69Ol0/fp1o8p58OABTZo0iaRSKfn5+dG7775LK1asIAAUGBhIt2/fpu+//558fX1JJpPRiy++qJvuZCpj278n98eQ9uxu2SdOnCClUkkJCQlGt1F3pr2p1WoSi8VUX1+vS8vKyqKAgAACQK6urrRo0aJ2865YsUJv2ltXbb5r1y6Sy+UEgAYOHEjXr1+nPXv2kEqlIgDk6+tL165do0ePHlFsbCz5+PiQvb09ubm5UWRkJBUUFBARUXh4OAGguLi4TvftwoULNGHCBPL09CQABIA8PDwoJCSEzp49a9ay8vPzdentfZKSkoiIqKKigmJiYigwMJAkEgn17t2bJkyYQJ9++mmbbYaFhZGXlxe1trZ2WrenWWvam80GZGuIjo4mFxcXa1fDJEJqf6G2Z3cCcmFhIdnb2+v9kxe6lpYWmjhxIu3fv19QZZlTRUUFSaVS2rZtm1H5eB6yjbC5p0cJ3LPSnoGBgYiPj0d8fHybp5IJUUtLC7Kzs1FTU4OoqCjBlGVu69evx+jRo6FWq61dFYNwQGbMTFatWoWZM2ciKipK8A8QysnJQWZmJk6ePGnw/GlLlGVO27dvR15eHk6cOAGxWGzt6hiEA7KBVq9ejbS0NFRXV8PPzw9Hjx61dpVs2rPanhs3boRarcbmzZutXZVOTZ48GYcOHdJ7TogQyjKXY8eO4dGjR8jJyYGzs7O1q2MwEZFln7ghEomQkZHBr1i3Em7/rs2cORMA8Mknn1i5Jsxajhw5glmzZln8gUR8hcwYYwLBAZkxxgSCAzJjjAkEB2TGGBMIe2ts1BYebvIs4/bvnPZW2yNHjli5JsxarHWOWGWWBWOM2QJLz7Kw+BWypXeQPfu0U/j4ipbZOh5DZowxgeCAzBhjAsEBmTHGBIIDMmOMCQQHZMYYEwgOyIwxJhAckBljTCA4IDPGmEBwQGaMMYHggMwYYwLBAZkxxgSCAzJjjAkEB2TGGBMIDsiMMSYQHJAZY0wgOCAzxphAcEBmjDGB4IDMGGMCwQGZMcYEggMyY4wJBAdkxhgTCA7IjDEmEByQGWNMIDggM8aYQHBAZowxgeCAzBhjAsEBmTHGBIIDMmOMCQQHZMYYEwgOyIwxJhAckBljTCA4IDPGmEBwQGaMMYGwt3YFGDPGuXPncOHCBb20K1euAAASExP10oODg/GrX/3KYnVjzFQiIiJrV4IxQ50+fRq/+c1vIBaLYWfX/he81tZWNDc349SpU5g8ebKFa8hY93FAZjaltbUVHh4eKC8v73Q9V1dXlJaWolevXhaqGWOm4zFkZlPs7OwwZ84cODg4dLiOg4MD5s6dy8GY2RwOyMzmzJ49G01NTR0ub2pqwuzZsy1YI8bMg4csmE0aMGAAbt261e4yb29v3Lp1CyKRyMK1Ysw0fIXMbNIbb7wBsVjcJl0sFmPevHkcjJlN4itkZpOuXLmCoUOHtrvshx9+wLBhwyxcI8ZMx1fIzCYNGTIEw4YNa3MlHBQUxMGY2SwOyMxmvfnmm3ozKcRiMX7/+99bsUaMmYaHLJjNKi4uhq+vL7SHsEgkwo0bNzBgwADrVoyxbuIrZGazvL29MX78eNjZ2cHOzg7jx4/nYMxsGgdkZtPeeOMNiEQi2NnZ4Y033rB2dRgzCQ9ZMJtWUVEBDw8PAMDdu3fRt29fK9eIMRNQN2VkZBAA/vCHP/zhz1OfjIyM7oZVMvnxmxkZGaYWwQRux44dAIAlS5ZYuSbtO3fuHEQiESZOnGi1Oly4cAHJycl8PvzCzZo1y6T8Jgfk1157zdQimMB98sknAITb11OnTgUAKJVKq9YjOTlZsG3ELMPqAZkxa7N2IGbMXHiWBWOMCQQHZMYYEwgOyIwxJhAckBljTCAEE5BPnDgBR0dHfP755xbZ3rZt29C3b1+IRCLs3r3bItu0pMTERAwZMgQymQwKhQJDhgzBunXroNForFIfS/evrTl16hRWrVqFzMxM+Pv7QyQSQSQStXv34csvvwylUolevXph2LBh+P77761QY8O0trZix44dCAkJ6dGy4uPjERQUBJVKBYlEgsDAQKxcuRK1tbV663388ccYN24clEolfH19MX/+fJSWlgIAPvvsMyQmJqKlpcXkunaXYAKypW8YXL58Ob7++muLbtOS/v73v+NPf/oTbt++jfv372PDhg1ITEzEjBkzrFIfviG0Y++//z527tyJ1atXIzIyEjdu3EBAQAD69OmDgwcP4osvvtBb/69//Ss++eQTvPrqq+KM2psAACAASURBVCgoKMCYMWOsVPPOFRYW4le/+hWWLl2K+vr6Hi3rzJkzWLRoEYqKilBRUYFNmzYhOTkZM2fO1K2TkZGBOXPmYObMmSgpKcGxY8dw7tw5TJ06FY8fP8a0adMglUoxefJkVFVVmVTfbjP1Tj1bVlhYSADoL3/5i7WrYnbh4eHU0NCglzZz5kwCQHfv3jWqrBkzZtCMGTPMWT2Lq6+vp+Dg4B4rv7vnw+bNm2nQoEFt+iogIIAOHTpEdnZ25OXlRVVVVXrLT548Sb/73e9MqnNPysvLo4iICDp48CCNHj2aRo0a1aNlhYWF0ePHj/XSXnvtNQJAt2/fJiKiSZMmUb9+/ai1tVW3zkcffUQA6Pz587o0tVpNwcHB1NzcbHRdYeKdeoK5QmbmlZWVBalUqpfm5eUFAG2+xv0S7N+/H2VlZdauhp6ffvoJ69atwwcffNCmrwAgJCQEMTExuHPnDpYvX26FGnbfqFGjkJmZiTlz5kAikfR4WcePH2/zlnFXV1cA0F1RFxcXw9PTU++lBt7e3gCg937G9evXIy8vD8nJySbVuzssGpDPnj2L559/HnK5HCqVCiNGjIBGo8H58+fh4+MDkUiEjz76CMCTu54UCgXs7Ozw3HPPwd3dHWKxGAqFAmPGjMHEiRPh7e0NqVQKJycnrFy5UrednTt3QiqVom/fvli4cCE8PT0hlUoREhKCixcvdlrHlpYWxMXFwcfHBzKZDCNHjjTqdtitW7dCLpdDqVSirKwMy5Ytg5eXFxYvXgyFQgG5XI5jx45h6tSpUKlU6N+/Pw4fPgwASE1N7XIdUxQWFsLJyQm+vr4ml2WM9vrXkH01pB/VajUcHBx0DxgCgHfeeQcKhQIikQgVFRWIiYnBsmXLcP36dYhEIgQGBgIAvvzyS6hUKmzcuNGi7aG1c+dOEBGmTZvW4ToJCQkYNGgQ9u3bh1OnTnW4HhFh+/btGDp0KCQSCZydnTF9+nRcuXIFgOHHlqnHv5DcuXMHMpkMfn5+AAB/f/82/5S148f+/v66NGdnZ4SGhiI5OdnyQ23dvbQ29itabW0tqVQqSkxMpIaGBiotLaWIiAgqLy8nIqLi4mICQCkpKbo877//PgGgixcvUl1dHVVUVNCUKVMIAH3xxRdUXl5OdXV1pFarCQDl5eXp8kZHR5NCoaDLly9TY2MjFRQU0Lhx40ipVOq+wrQ3ZLF8+XKSSCR09OhRqqyspNWrV5OdnR19++23Bu/rmjVrCAAtXryYUlJSKCIign788Udd+unTp6m6uprKyspo4sSJpFAoqKmpSS9vZ+sYo6mpiUpKSiglJYUkEgkdOHDA6DLMMWTRXv8asq+G9OOcOXPI3d1db3tJSUkEQHd8RUZGUkBAgN46x48fJ6VSSfHx8SbtG1H3hiz8/f0pKCio3WUBAQF08+ZNIiL6+uuvyc7OjgYMGEC1tbVE1HbIIi4ujhwcHOjAgQNUVVVFly5dojFjxpCrqyuVlpYSkWHtbY7j/+fGjx9v0pBFd8qqq6sjpVJJarVal5aTk0NisZh27txJGo2GfvjhBxo6dCi98sorbfKvWrWKAFBubq5R9YOtDFkUFRVBo9Fg2LBhkEqlcHd3R2Zmpu5rRWeCgoIgl8vRp08fzJ49GwDg4+MDV1dXyOVyzJ07FwB0VwNa9vb2uiuGoKAgpKamoqamBmlpae1up7GxEampqQgPD0dkZCScnJywdu1aiMXiDvN0ZsuWLVi0aBEyMzMxZMgQXXpISAhUKhXc3NwQFRWFuro63L59Wy+vIesYwtvbG/3798f69euxdetWk++17wld7aux/WiosLAwaDQarFu3ztRdMFpdXR1u3ryJgICALtcNDg7GkiVLUFRUhPfee6/N8oaGBmzfvh0RERGYO3cuHB0dMWLECOzevRsVFRXYs2eP3vodtbe5j39r2rRpEzw9PZGQkKBLCw0NRWxsLNRqNVQqFYYPH46amhrs27evTf6BAwcCAPLz8y1WZ8CCQxb+/v7o27cv5s6di/Xr16OoqKhb5Tg4OAAAHj9+rEvTvg6+ubm507xjx46FXC5vE7i1rl69ivr6egwfPlyXJpPJ4OHh0WEeU2n3p7O6G7JOR4qLi1FWVoaPP/4Y//M//4P/9//+n+DGUp9myL521Y+2oKysDEQEuVxu0PoJCQkYPHgwdu3ahfPnz+stKygoQG1tLcaOHauXPm7cODg4OHQ6TPd0e1vj+O8JWVlZOHLkCL766iu955ysWbMGe/bswenTp1FbW4sbN24gJCQEwcHBKC4u1itD2y/379+3aN0tFpBlMhnOnDmDF198ERs3boS/vz+ioqLQ0NBgqSoAACQSCcrLy9tdVldXBwBYu3atbh6oSCTCrVu3TJ62Yy1isRhubm54+eWXkZ6ejoKCAmzatMna1TJZZ/1oCxobGwHA4B+8pFIp0tLSIBKJsGDBAr3zRjtFq3fv3m3yOTk5oaamxqBtPAvHf3p6OrZs2YKcnBy913ndu3cPiYmJeOutt/DSSy9BoVDAz88Pe/fuxd27d5GUlKRXjkwmA/CvfrIUi/6oN2zYMHz++ee4e/cuYmNjkZGRgW3btlls+83NzaiqqkL//v3bXe7m5gbgyfN/iUjvc+HCBYvVs6cEBgaiV69eKCgosHZVTNJVP9oC7QlvzE0IwcHBWLp0KQoLC7FhwwZdupOTEwC0G3iNaSdbP/5TUlJw8OBBnDlzBv369dNbVlhYiJaWljbpKpUKLi4ubc6JpqYmAP/qJ0uxWEC+e/cuLl++DOBJx2/evBljxozRpVlCTk4OiAgvvPBCu8u1szby8vIsVqee8ODBA7z++utt0rUHpXaqj636eT/a29t3azjHmrR3iVZXVxuVb8OGDRgyZAhyc3N1acOHD0fv3r3x3Xff6a178eJFNDU14bnnnjOobFs9/okIsbGxyM/PR3Z2drvfFLT/lO7du6eXXlNTg4cPH7Y5J7T94u7u3kO1bp9FA/LChQtx5coVNDU1ITc3F7du3eowOJpDa2srKisr8fjxY1y6dAkxMTHw8fHBvHnz2l1fKpVi/vz5OHz4MFJTU6HRaNDS0oKSkpI2HSlkCoUCf/3rX3HmzBloNBo0NzcjNzcXv//976FQKLB06VJrV9EoXfVjYGAgHj58iOzsbDQ3N6O8vFxvXikAuLi44O7duygqKkJNTQ2am5tx8uRJq017k8vl8Pf3R0lJiVH5tEMXT8+5lUqlWLZsGbKysnDw4EFoNBrk5+fj7bffhqenJ6Kjow0uu6vjPyoqCu7u7ma5XdtcZV2+fBlbt27F3r17IRaL9YZbRCIRtm3bBj8/P0yaNAl79+7FuXP/n707j4riSvsH/m2k6QW6WQQBUZDFfYFjNAmNDjG+k8RwYkRcSDQzmsyEODEdohLEhRhEI8EgL0Ym4zK+J2oUFIOGiJlXM5hxNL7JiQSCURE3XBBQoFlle35/+KPHthF7AbraPJ9zOMdzq+r2U7duP1bfulX1HRobG1FaWqptmzfeeEOnzo7jMnr0aLNiM5qp0zOMneZz+fJlUqlU5OzsTH369KH+/fvT8uXLqbW1lTZu3EgeHh4EgORyOU2dOpVSU1NJLpcTABo0aBD961//onXr1pGjoyMBIHd3d9q1axft2bOH3N3dCQA5OzvT7t27iejedCmxWExeXl5ka2tLSqWSpk2bRiUlJURE9Mknn2i3s7e3p+nTpxMR0d27dyk2Npa8vb3J1taW3NzcKCIigoqKigzaz6SkJJLJZASABg4cqJ1mtmnTJu3+DB48mEpKSmjz5s2kVCoJAPn4+FB0dPQj1zl//rxBcUydOpV8fX3JwcGBJBIJ+fv7U2RkJBUWFhp8zDqYO+2ts+NrSHucP3/+kceRiOj27ds0adIkkkql5OvrS++88w7FxMQQAAoICKCrV6/STz/9RD4+PiSTyWjChAlUVlZGhw4dIoVCQYmJiSbvWwdTpr2p1WoSi8XU0NCgLdu/fz/5+/sTAHJ1daWFCxd2um1MTIzOtLf29nZKTk6mwYMHk1gsJmdnZwoPD6dz584RkWH97/z584/s/+Hh4QSA4uPju9y3kydPUkhICHl6emrfNefh4UEqlYqOHTvWrXUVFhZ2+Y675ORkIiKqrKyk6OhoCggIIIlEQg4ODhQSEkJffvml3meGhYWRl5eXzl19hoCZ094e21uno6KiyMXFxdJhPBYseeu0tRxHU74PxcXFZGtra9LccEtpa2ujiRMn0rZt2wRVV3eqrKwkqVRK69evN3pbcxPyY33rtCWf2sS6z+N6HAMCApCQkICEhASruJ29ra0N2dnZqK2tRWRkpGDq6m6rVq1CUFAQ1Gp1r3/2Y52Qu9PZs2f1xqY6++vJziWEGFj3iouLw8yZMxEZGWn0Bb7elpeXh6ysLOTm5ho8f7o36upOKSkpyM/Px6FDh7T3N/QqU0+thTxkERcXR3Z2dtrx571791o6JKtmqSELazqO5n4fvvnmG4qNje3GiJixsrOzae3atXpPjTMGzByyEP3/SoyWmZmJ2bNn83NufwM6nim7d+9eC0ciXPx9YAAgEomQkZGBWbNmmbQ9D1kwxphAcEJmjDGB4ITMGGMCwQmZMcYEwtbcCjIzM7sjDiZgHbeR8rF+uI6H73AbMXOYnZCF+MBz1jP4WD8atxEzh9kJmaf5PP542tuj8bQ3BkDnBaqm4DFkxhgTCE7IjDEmEJyQGWNMIDghM8aYQHBCZowxgeCEzBhjAiGIhPzzzz8jMjISvr6+kEgkcHV1RWBgIBITE7XrHDp0CI6Ojvjqq696Jab169drX0T52WefGbRNVlYW/Pz89J5PLJVK4evri9dffx2XLl166PqvvfaaXp3PPfccFAoF+vTpg5EjR2L58uVGb9Md7z9jlnfkyBHExcWZ1G+E3Afa29uxYcMGqFQqvWUJCQkYMWIElEolJBIJAgIC8P777+s90P+LL77A+PHjoVAo4OPjg/nz56OsrAwAcPDgQSQlJVnHiw5MfW5ndz0PuaCggORyOb377rt06dIlamxspHPnztH7779PkydP1q6Xk5NDSqWSDh48aPZnGqq4uJgA0F//+lejtvP39ydHR0ciuveamlu3btHnn39Ocrmc+vXrR5WVlXrr9+3blwBQTk6OXn25ubk6708zdRtTWfIVTtaip58PHh8fTy+99BJpNBptWW/2gZ5y/vx5CgkJIQAUGBiotzw0NJQ2bdpEt2/fJo1GQxkZGSQWi+mFF17QrrNnzx4CQElJSVRdXU2nT58mPz8/CgoKopaWFiIiSk1NpdDQUKqqqurR/YG1v8Jp/fr1cHJyQmpqKgYNGgSpVIohQ4Zg9erVkMlk2vXCwsJQU1ODl156yYLRGs/Gxgb9+vXDa6+9hoULF6K8vBxHjhzRWy8tLQ02NjaIiooy+M0RpmxjbRobGzs9cxJ63d1p3bp12LNnDzIzM6FQKHSWWXMf+Pnnn7F06VIsWLAAQUFBna7j4OCAqKgouLi4QKFQYNasWQgPD8fhw4dRWloKAPjb3/6G/v37IyYmBo6OjggKCsKiRYuQn5+PU6dOAQDeffddBAYG4sUXX0Rra2uv7aOxLJ6Qb9++jZqaGty5c0en3M7OrteGJ3pLQEAAAGh/St1PpVIhOjoa169fx5IlSwyqz5RtrM22bdtQXl5udXV3lwsXLmDlypX48MMPIZVK9ZZbcx8IDAxEVlYW5syZA4lE0uk6OTk56NOnj06Zq6srAKChoQEAUFpaCk9PT5275AYOHAgAuHLlirZs1apVyM/PR2pqarfuR3eyeEIeP3486uvr8eyzz+Lf//53p+scP34c3t7eEIlE+PTTTwEAqampsLe3h42NDZ544gm4u7tDLBbD3t4eY8eOxcSJEzFw4EBIpVI4OTnh/fff19aXlpYGqVSKfv364a233oKnpyekUilUKpX2f9SHaWtrQ3x8PLy9vSGTyTBmzBhkZGQYtK/FxcUA7nXEziQmJmLIkCHYunVrp2fR3bVNbyAipKSkYPjw4ZBIJHB2dsa0adNw9uxZAIBarYadnR08PDy027z99tuwt7eHSCRCZWUloqOjsXjxYpSUlEAkEiEgIMCgY2dq3QBw+PBhKJVKrFmzphdb6+HS0tJARJg6depD1zG0DzzqmKSnp8Pe3h5yuRwHDhzAlClToFQqMWDAAOzevVtbjznfge5w/fp1yGQy+Pr6AgD8/Pz0/mPtOOnx8/PTljk7OyM0NBSpqanCvcXd1LGO7hoza2hooHHjxhEAAkAjRoygpKQkun37ts56paWlBIA2btyoLfvggw8IAJ06dYrq6+upsrKSXnjhBQJAX3/9NVVUVFB9fT2p1WoCQPn5+dpto6KiyN7ens6cOUNNTU1UVFRE48ePJ4VCQVevXiWizseQlyxZQhKJhPbt20dVVVW0bNkysrGxoR9++EG7zv1jyEREVVVV9D//8z8kl8spLCxMrw38/f3p0qVLRER04sQJsrGxoUGDBlFdXR0RPXwM2dhtTGXKGHJ8fDzZ2dnRjh07qLq6mgoKCmjs2LHk6upKZWVlREQ0Z84ccnd319kuOTmZAFBFRQUREUVERJC/v7/OOoYcO1PrzsnJIYVCQQkJCUbtb0+NIfv5+dGIESM6XWZsHzDkmCxfvpwA0NGjR6mmpobKy8tp4sSJZG9vT83NzURk2HfAWE899VSnY8gPqq+vJ4VCQWq1WluWl5dHYrGY0tLSSKPR0C+//ELDhw+n559/Xm/7uLg4AkCnT582OdauwNrHkGUyGU6cOIH//u//xrBhw3DmzBnExsZi+PDhOHbsmEF1jBgxAnK5HH379sUrr7wCAPD29oarqyvkcjnmzp0LANozgQ62trbas4URI0YgPT0dtbW12L59e6ef09TUhPT0dISHhyMiIgJOTk5YsWIFxGKx3jY1NTXaq+DOzs6YP38+li1bhi+//LLLfQkODsZ7772Hy5cvY+nSpQbtvynb9KTGxkakpKRg+vTpmDt3LhwdHTF69Gh89tlnqKysxObNm83+DGOPnaHCwsKg0WiwcuVKs2M0V319PS5dugR/f/9HrvuoPmDsMVGpVFAqlXBzc0NkZCTq6+tx9epVo74DPWHt2rXw9PTUmYEVGhqK2NhYqNVqKJVKjBo1CrW1tdi6dave9oMHDwYAFBYW9nisprB4QgYAsVgMtVqNX3/9Fd9//z2mTZuG8vJyzJw5E1VVVUbVZWdnBwA6A/cdr/NuaWnpcttx48ZBLpfrJe4O586dQ0NDA0aNGqUtk8lk8PDw0NvG0dERRAQiQkxMDIgIjo6OBr1aPDExEUOHDsWmTZtw/PjxR65v6jY9paioCHV1dRg3bpxO+fjx42FnZ/fIYSFTPOrYWaPy8nIQEeRyuUHrd9UHzDkmHd+plpYWo74D3W3//v3IzMzEN998o3Nxc/ny5di8eTOOHj2Kuro6XLx4ESqVCsHBwdoLfx062vLWrVs9GqupBJGQ7/fUU0/hyy+/xIIFC1BRUYF//vOfvfr5EokEFRUVnS6rr68HAKxYsUJnnvGVK1e0Fxg6s3LlSnh4eGDZsmV6HaQzUqkU27dvh0gkwuuvv47GxsYe2aanVFdXA7h3hfxBTk5OqK2t7ZHP7erYWaOmpiYAeOgFrwd11Qe665iY+h0w1549e7Bu3Trk5eVh0KBB2vKbN28iKSkJb775Jp599lnY29vD19cXW7ZswY0bN5CcnKxTT8fMrY62FRqLJ+SIiIhOp6F0THbvyYP8oJaWFlRXV2PAgAGdLndzcwMAbNiwQXv22/HX8caIzigUCqxbtw61tbX4y1/+YlAswcHBWLRoEYqLi7F69eoe26YnODk5AUCnX/Ku2tccjzp21qgjeRhzQ8PD+kB3HRNTvwPm2LhxI3bu3Ilvv/0W/fv311lWXFyMtrY2vXKlUgkXFxcUFRXplDc3NwOAzpRaIbF4Qr579y7OnDmjV37u3DkAwJgxY3otlry8PBARnn766U6Xd8zayM/PN7ruP/zhD3jqqaeQk5Nj8Gt+Vq9ejWHDhuH06dMGf44p23S3UaNGwcHBAT/++KNO+alTp9Dc3IwnnngCwL1x4EcNIxnqwWPXnXVbSsedosbOL+6sDxh6TB7FnO+AsYgIsbGxKCwsRHZ2dqdn9x3/kdy8eVOnvLa2Fnfu3NFOf+vQ0Zbu7u49FLV5LJ6QASA8PByZmZmorq5GTU0NDhw4gKVLl+Lll1/u0YTc3t6OqqoqtLa2oqCgANHR0fD29sa8efM6XV8qlWL+/PnYvXs30tPTodFo0NbWhmvXrul1iAeJRCKkpaVBJBJBrVYbNDbe8RP0wXmY3b1Nd5NKpVi8eDH279+PnTt3QqPRoLCwEAsWLICnpyeioqIA3JuXfefOHWRnZ6OlpQUVFRU680YBwMXFBTdu3MDly5dRW1urTbKPOnam1p2bmyuYaW9yuRx+fn7adxoaqrM+YOgxMaTuR30HIiMj4e7ubvbt2mfOnMHHH3+MLVu2QCwW6z2SYP369fD19cWkSZOwZcsWfPfdd2hsbERpaal2f9544w2dOjvacvTo0WbF1mNMnZ7RXdN8/vGPf9Ds2bPJ39+fJBIJ2dnZ0dChQ2nVqlXU1NREREQbN24kDw8PAkByuZymTp1KqampJJfLCQANGjSI/vWvf9G6devI0dGRAJC7uzvt2rWL9uzZQ+7u7gSAnJ2daffu3UR0b+qUWCwmLy8vsrW1JaVSSdOmTaOSkhIiIvrkk0+029nb29P06dOJiOju3bsUGxtL3t7eZGtrS25ubhQREUFFRUX073//m4YMGaKdwte/f3966623dPZ33rx5BICcnJzoySefJH9/fwJArq6utHDhwk7bKCYmRjt9af/+/UZvYy5Tpr21t7dTcnIyDR48mMRiMTk7O1N4eDidO3dOu87t27dp0qRJJJVKydfXl9555x2KiYkhABQQEEBXr16ln376iXx8fEgmk9GECROorKzskcfOnLoPHTpECoWCEhMTjdrfnpr2plarSSwWU0NDg7bM1D7wqGOyadMm7Xdq8ODBVFJSQps3byalUkkAyMfHh86fP9/ld4CIKDw8nABQfHx8l/t28uRJCgkJIU9PT+13xsPDg1QqFR07dowKCwu15Z39JScnExFRZWUlRUdHU0BAAEkkEnJwcKCQkBD68ssv9T4zLCyMvLy8qL293bgDYSCYOe3N4gnZUqKiosjFxcXSYVgFoT3LQojHrqe+D8XFxWRra0s7duzo9rp7SltbG02cOJG2bdtm6VB0VFZWklQqpfXr1/fYZ5ibkAUxZGEpVvH0J9ap38qxCwgIQEJCAhISEvSecCZEbW1tyM7ORm1tLSIjIy0djo5Vq1YhKCgIarXa0qE81G86ITNmDeLi4jBz5kxERkYK/gFCeXl5yMrKQm5ursHzp3tDSkoK8vPzcejQIYPuBbCU32RCXrZsGbZv346amhr4+vpi3759lg6JGei3euzWrFkDtVqNjz76yNKhdGny5MnYtWuXznNELO3AgQO4e/cu8vLy4OzsbOlwuiT6/+MeRsvMzMTs2bOF+5AO1m1mzpwJANi7d6+FIxEu/j4w4N5sqoyMDMyaNcuk7X+TZ8iMMSZEnJAZY0wgOCEzxphAcEJmjDGBsDW3go4LPuzx9f333wPgY92VjltyuY2YOUyeZXHy5EmkpKR0dzyMGa3jYeOCfT4B+01ZtGgRgoODTdrW5ITMmFB0TDEy9Cl6jAkVjyEzxphAcEJmjDGB4ITMGGMCwQmZMcYEghMyY4wJBCdkxhgTCE7IjDEmEJyQGWNMIDghM8aYQHBCZowxgeCEzBhjAsEJmTHGBIITMmOMCQQnZMYYEwhOyIwxJhCckBljTCA4ITPGmEBwQmaMMYHghMwYYwLBCZkxxgSCEzJjjAkEJ2TGGBMITsiMMSYQnJAZY0wgOCEzxphAcEJmjDGB4ITMGGMCwQmZMcYEghMyY4wJBCdkxhgTCE7IjDEmEJyQGWNMIDghM8aYQIiIiCwdBGOG+vzzz5GSkoK2tjZtWWVlJQDA1dVVW9anTx8sWrQIf/jDH3o9RsZMxQmZWZXz589j6NChBq177tw5DBkypIcjYqz78JAFsypDhgxBYGAgRCLRQ9cRiUQIDAzkZMysDidkZnX+8Ic/oE+fPg9dbmtriz/+8Y+9GBFj3YOHLJjVuXHjBgYOHIj29vZOl4tEIpSWlsLLy6uXI2PMPHyGzKxO//79oVKpYGOj331tbGwQEhLCyZhZJU7IzCq99tprnZaLRCKeWcGsFg9ZMKtUVVUFd3d3tLS06JTb2tqirKwMffv2tVBkjJmOz5CZVXJ2dsbvf/97nYt7ffr0wfPPP8/JmFktTsjMas2dO1fnwh4RYe7cuRaMiDHz8JAFs1oNDQ3o27cvmpqaAABSqRSVlZWwt7e3cGSMmYbPkJnVksvlCA8Ph1gshlgsRnh4OCdjZtU4ITOr9uqrr6KlpQUtLS149dVXLR0OY2axtXQA165dw4kTJywdBrNSbW1tkMvlICJoNBpkZmZaOiRmpVQqFQYMGGDRGCw+hpyZmYnZs2dbMgTGGENGRgZmzZpl0Rgsfobcga8tCptIJBJEh+3MsWPHIBKJ8Lvf/c6iccycORMAsHfvXovGwYzX1cOqepNgEjJjppo4caKlQ2CsW3BCZlavs2daMGaNuCczxphAcEJmjDGB4ITMGGMCwQmZMcYE4rFIyH/605+gUCggEomQn59v6XB6VXt7OzZs2ACVStXp8uPHjyMkJARyuRyenp6IjY3F3bt3eznKew4dOgRHR0d89dVXFvl8oTty5Aji4uKQlZUFPz8/iEQiiESiTp/9/Nxzz0GhUKBPnz4YOXIkfvrpJwtEbJiu+mhCQgJGjBgBpVIJiUSCjP4vpwAAIABJREFUgIAAvP/++6irq9NZ74svvsD48eOhUCjg4+OD+fPno6ysDABw8OBBJCUl6byJ3GqRhWVkZFB3hLF7924CQKdPn+6GqKzD+fPnKSQkhABQYGCg3vJffvmFZDIZrVy5kurq6ujEiRPk6upK8+fPN/qzAFBGRoZZ8ebk5JBSqaSDBw+aVY9QzZgxg2bMmGHStvHx8fTSSy+RRqPRlvn7+1Pfvn0JAOXk5Ohtk5ubSy+//LLJ8faGR/XR0NBQ2rRpE92+fZs0Gg1lZGSQWCymF154QbvOnj17CAAlJSVRdXU1nT59mvz8/CgoKIhaWlqIiCg1NZVCQ0OpqqrKpDi7o393B07IFtbQ0EDBwcFGb5efn0/Tp0+nnTt3UlBQUKedffbs2eTr60vt7e3asuTkZBKJRPTrr78a9XlC6bDmMLWtDWVqQv7oo49oyJAh1NjYqFPu7+9Pu3btIhsbG/Ly8qLq6mqd5UJPyIb00bCwMGptbdUpmzVrFgGgq1evEhHRpEmTqH///jr9+NNPPyUAdPz4cW2ZWq2m4OBgbZI2hlD692MxZAEI504bY23btg3l5eVGbxcYGIisrCzMmTMHEolEb3lrayu+/vprhIaG6rTNlClTQEQ4cOCAWXFbI1PbuidduHABK1euxIcffgipVKq3XKVSITo6GtevX8eSJUssEKHpHtVHASAnJ0fvDeKurq4A7j1eFQBKS0vh6emp048HDhwIALhy5Yq2bNWqVcjPz0dqamq37kdvssqETERITk7G0KFDIZFI4OjoiJiYGO3yjz/+GHK5HAqFAuXl5Vi8eDG8vLxw7tw5EBFSUlIwfPhwSCQSODs7Y9q0aTh79iwAIC0tDVKpFP369cNbb70FT09PSKVSqFQqnDp1SieGrupRq9Wws7ODh4eHdpu3334b9vb2EIlEqKysRHR0NBYvXoySkhKIRCIEBAR0WxtdvHgRdXV18Pb21in39/cHABQUFHTbZxni+PHj8Pb2hkgkwqeffgoASE9Ph729PeRyOQ4cOIApU6ZAqVRiwIAB2L17NwDDjoc5bX348GEolUqsWbOmV9ujQ1paGogIU6dOfeg6iYmJGDJkCLZu3YojR448dL1H9UlD2hu498Cm+Ph4eHt7QyaTYcyYMcjIyOi+nX6E69evQyaTwdfXFwDg5+en9x9px/ixn5+ftszZ2RmhoaFITU213kcxWPDsnIhMG7JYvnw5iUQi+uSTT6iqqooaGhpo06ZNOkMWy5cvJwD07rvv0saNG2n69On066+/Unx8PNnZ2dGOHTuourqaCgoKaOzYseTq6kplZWVERBQVFUX29vZ05swZampqoqKiIho/fjwpFArtzyhD6pkzZw65u7vrxJ6cnEwAqKKigoiIIiIiyN/f36w2fOqpp/R+Dh47dowAUHJyst76MpmMJk+ebNRnoBt+0pWWlhIA2rhxo7as4zgdPXqUampqqLy8nCZOnEj29vbU3NxMRIYdD1PbOicnhxQKBSUkJJi1b0SmDVn4+fnRiBEjOl3m7+9Ply5dIiKiEydOkI2NDQ0aNIjq6uqISH/IwpA+aUh7L1myhCQSCe3bt4+qqqpo2bJlZGNjQz/88IOxTaLVWR/tTH19PSkUClKr1dqyvLw8EovFlJaWRhqNhn755RcaPnw4Pf/883rbx8XFmTR02R39uztY3RlyY2MjNmzYgP/6r//CokWL4OTkBJlMBhcXl07XX7duHRYuXIisrCz4+PggJSUF06dPx9y5c+Ho6IjRo0fjs88+Q2VlJTZv3qzdztbWVnumMWLECKSnp6O2thbbt29HY2OjwfVYSsdMigd/DgKAWCxGY2Njb4fUJZVKBaVSCTc3N0RGRqK+vh5Xr17VLu/qeJgjLCwMGo0GK1euNHcXjFZfX49Lly5pf7V0JTg4GO+99x4uX76MpUuX6i03tk8+rL2bmpqQnp6O8PBwREREwMnJCStWrIBYLDa7rQ2xdu1aeHp6IjExUVsWGhqK2NhYqNVqKJVKjBo1CrW1tdi6dave9oMHDwYAFBYW9nisPcHqEvKFCxfQ0NCAyZMnG71tUVER6urqMG7cOJ3y8ePHw87OTmdI4kHjxo2DXC7H2bNnzaqnt3SMR7a2tuota25uhkwm6+2QDGZnZwcAem+Uvt/9x8NalZeXg4ggl8sNWj8xMRFDhw7Fpk2bcPz4cZ1l5vTJ+9v73LlzaGhowKhRo7TLZTIZPDw8eryt9+/fj8zMTHzzzTdQKBTa8uXLl2Pz5s04evQo6urqcPHiRahUKgQHB6O0tFSnjo62vHXrVo/G2lOsLiFfu3YNAODm5mb0ttXV1QAABwcHvWVOTk6ora3tcnuJRIKKigqz6+kNHeOpGo1Gp7yhoQFNTU3w9PS0RFjdquN4WKuOdwE+7ILXg6RSKbZv3w6RSITXX39d51dOd/XJ+vp6AMCKFSu086BFIhGuXLmivcjWE/bs2YN169YhLy8PgwYN0pbfvHkTSUlJePPNN/Hss8/C3t4evr6+2LJlC27cuIHk5GSdejpONDra1tpYXULuOPMz5eYGJycnAOi0c1ZXV3f5toCWlhbtOubU01t8fX2hUCh0rkID935hAMCYMWMsEVa3uf94WKuO5GHMDQ3BwcFYtGgRiouLsXr1am15d/XJjhOdDRs2gO5Ni9X+nTx50uA4jbFx40bs3LkT3377Lfr376+zrLi4GG1tbXrlSqUSLi4uKCoq0ilvbm4GAEH/AuyK1SXkUaNGwcbGBseOHTNpWwcHB/z444865adOnUJzczOeeOKJh26bl5cHIsLTTz9tcD22trZd/uzuSba2tnjxxRfx3Xffob29XVuem5sLkUjU5VV9a3D/8QAs29am6tevH0QiEWpqaozabvXq1Rg2bBhOnz6tLTOnb99v4MCBkEqlvXLHKxEhNjYWhYWFyM7O7vTsvuM/kps3b+qU19bW4s6dO9rpbx062tLd3b2Hou5ZVpeQ3dzcMGPGDOzbtw/btm2DRqNBQUGBQRfSpFIpFi9ejP3792Pnzp3QaDQoLCzEggUL4OnpiaioKO267e3tqKqqQmtrKwoKChAdHQ1vb2/MmzfP4HoCAgJw584dZGdno6WlBRUVFXpnrC4uLrhx4wYuX76M2trabk0qK1euxK1bt/DBBx+gvr4eJ0+eRHJyMubNm4ehQ4d22+f0hq6OB2B6W+fm5lps2ptcLoefn592GM5QHUMX91+wNaZvP6ru+fPnY/fu3UhPT4dGo0FbWxuuXbumTYqRkZFwd3c3+3btM2fO4OOPP8aWLVsgFot1hkhEIhHWr18PX19fTJo0CVu2bMF3332HxsZGlJaWavfnjTfe0Kmzoy1Hjx5tVmwWY7kJHveYMu2ttraW/vznP1Pfvn3JwcGBJkyYQPHx8QSABgwYQHPmzCGZTEYAaODAgbRjxw7ttu3t7ZScnEyDBw8msVhMzs7OFB4eTufOndOuExUVRWKxmLy8vMjW1paUSiVNmzaNSkpKjKrn9u3bNGnSJJJKpeTr60vvvPMOxcTEEAAKCAigq1ev0k8//UQ+Pj4kk8lowoQJ2ulJj3Ly5EkKCQkhT09PAkAAyMPDg1QqFR07dky73rFjx+jJJ58kiURCnp6eFBMTQ01NTUa1N5H504I2btxIHh4eBIDkcjlNnTqVNm3aRHK5nADQ4MGDqaSkhDZv3kxKpZIAkI+PD50/f96g42FqWx86dIgUCgUlJiaavG8dTJn2plarSSwWU0NDg7Zs//795O/vTwDI1dWVFi5c2Om2MTExOtPeHtUnDW3vu3fvUmxsLHl7e5OtrS25ublRREQEFRUVERFReHg4AaD4+Pgu9+1RfbSwsFBb3tlfx5TNyspKio6OpoCAAJJIJOTg4EAhISH05Zdf6n1mWFgYeXl56dzVZwhz+3d3scqE3NOioqLIxcXF0mEIiiU7rLUcD1MScnFxMdna2uqcNAhdW1sbTZw4kbZt22bpUHRUVlaSVCql9evXG72tUBKy1Q1Z9JbH4slRj5HH9XgEBAQgISEBCQkJek84E6K2tjZkZ2ejtrYWkZGRlg5Hx6pVqxAUFAS1Wm3pUEzGCVlAzp49qzeO1tmf0L4IzDxxcXGYOXMmIiMjjb7A19vy8vKQlZWF3Nxcg+dP94aUlBTk5+fj0KFDEIvFlg7HZJyQH7Bs2TJs374dNTU18PX1xb59+3rts4cNG6Y31aizvz179vRaTJZmyePRm9asWQO1Wo2PPvrI0qF0afLkydi1a5fOc0Ms7cCBA7h79y7y8vLg7Oxs6XDMIiKy7FM4MjMzMXv2bOt9GMhvhEgkQkZGBmbNmmXpUARr5syZAIC9e/daOBJmLKH0bz5DZowxgeCEzBhjAsEJmTHGBIITMmOMCYStpQPo0HFBhAnXhg0b+IJVF77//nsA3JeZ6fgMmTHGBEIwZ8h85iVsIpEI7733nsWnBQkZT3uzXkJ5STKfITPGmEBwQmaMMYHghMwYYwLBCZkxxgSCEzJjjAnEY52Qs7Ky4Ofnp/f4Sjs7O/Tr1w/PPPMMkpOTUVVVZelQGcORI0cQFxen129fe+01vXWfe+45KBQK9OnTByNHjjT7dUo9qb29HRs2bIBKpdJblpCQgBEjRkCpVEIikSAgIADvv/++3rOhv/jiC4wfPx4KhQI+Pj6YP38+ysrKAAAHDx5EUlLS4/HMbMs8F/8/euONIf7+/uTo6EhE915zU1VVRf/85z9p3rx5JBKJyNPTk3744YcejcHaQSBvVBAyU94Y0iE+Pp5eeukl0mg02jJ/f3/q27cvAaCcnBy9bXJzc3Ve4SRE58+fp5CQEAJAgYGBestDQ0Np06ZNdPv2bdJoNJSRkUFisZheeOEF7Tp79uwhAJSUlETV1dV0+vRp8vPzo6CgIGppaSEiotTUVAoNDaWqqiqT4hRK/36sz5A7IxKJ4OTkhGeeeQbbt29HZmYmbt26hbCwMME/HPxBjY2NnZ51PE56ch+F0n7r1q3Dnj17kJmZCYVCobMsLS0NNjY2iIqKsrr++fPPP2Pp0qVYsGABgoKCOl3HwcEBUVFRcHFxgUKhwKxZsxAeHo7Dhw+jtLQUAPC3v/0N/fv3R0xMDBwdHREUFIRFixYhPz8fp06dAgC8++67CAwMxIsvvojW1tZe28fu9ptLyA+aMWMG5s2bh/Lycnz22WeWDsco27ZtQ3l5uaXD6FE9uY9CaL8LFy5g5cqV+PDDDyGVSvWWq1QqREdH4/r161iyZIkFIjRdYGAgsrKyMGfOHEgkkk7XycnJ0Xl7NgC4uroCABoaGgAApaWl8PT01Ll5Y+DAgQCg82bxVatWIT8/H6mpqd26H73pN5+QAWhfJZ+bm4uPP/4YcrkcCoUC5eXlWLx4Mby8vHDu3DkQEVJSUjB8+HBIJBI4Oztj2rRpOHv2LIB7ZzNSqRT9+vXDW2+9BU9PT0ilUqhUKu3/5AAeWY9arYadnZ3OWxnefvtt2NvbQyQSobKyEtHR0Vi8eDFKSkogEokQEBDQew1mgJ7aR0Pa2Jz2O3z4MJRKJdasWdMr7ZSWlgYiwtSpUx+6TmJiIoYMGYKtW7fiyJEjD13vUW2enp4Oe3t7yOVyHDhwAFOmTIFSqcSAAQOwe/dubT1tbW2Ij4+Ht7c3ZDIZxowZg4yMjO7b6Ue4fv06ZDIZfH19AQB+fn56/3F2jB/7+flpy5ydnREaGorU1FTrfeGFJcdLiHp/DLkzGo2GANDAgQOJiGj58uUEgN59913auHEjTZ8+nX799VeKj48nOzs72rFjB1VXV1NBQQGNHTuWXF1dqaysjIjuvSHZ3t6ezpw5Q01NTVRUVETjx48nhUJBV69eJSIyqJ45c+aQu7u7TpzJyckEgCoqKoiIKCIigvz9/bu9vToDI8fYenIfDWljU+vOyckhhUJBCQkJBu9rB1PGkP38/GjEiBGdLvP396dLly4REdGJEyfIxsaGBg0aRHV1dUSkP4ZsSJt39O2jR49STU0NlZeX08SJE8ne3p6am5uJiGjJkiUkkUho3759VFVVRcuWLSMbGxuzrrM89dRTnY4hP6i+vp4UCgWp1WptWV5eHonFYkpLSyONRkO//PILDR8+nJ5//nm97ePi4ggAnT592qj4jO3fPYXPkAEoFAqIRCLU1tbqlK9btw4LFy5EVlYWfHx8kJKSgunTp2Pu3LlwdHTE6NGj8dlnn6GyshKbN2/Wbmdra6s9SxkxYgTS09NRW1uL7du3o7Gx0eB6rFVv7GNXbWyOsLAwaDQarFy50uwYH6W+vh6XLl2Cv7//I9cNDg7Ge++9h8uXL2Pp0qV6y41tc5VKBaVSCTc3N0RGRqK+vh5Xr15FU1MT0tPTER4ejoiICDg5OWHFihUQi8Vmt60h1q5dC09PTyQmJmrLQkNDERsbC7VaDaVSiVGjRqG2thZbt27V237w4MEAgMLCwh6PtSdwQsa9LwYRQalUPnSdoqIi1NXVYdy4cTrl48ePh52dnc6QxIPGjRsHuVyOs2fPmlWPtbDEPt7fxtaivLwcRGTw25sTExMxdOhQbNq0CcePH9dZZk6b29nZAQBaWlpw7tw5NDQ0YNSoUdrlMpkMHh4ePd62+/fvR2ZmJr755hudi5vLly/H5s2bcfToUdTV1eHixYtQqVQIDg7WXvjr0NGWt27d6tFYewonZADnz58HcO+tzw9TXV0N4N5V4Qc5OTnpnV0/SCKRoKKiwux6rIGl9rGjja1FU1MTADz0gteDpFIptm/fDpFIhNdffx2NjY3aZd3V5vX19QCAFStW6Mzdv3LlivYiW0/Ys2cP1q1bh7y8PAwaNEhbfvPmTSQlJeHNN9/Es88+C3t7e/j6+mLLli24ceMGkpOTdeqRyWQA/tO21oYTMu5dyAGAKVOmPHQdJycnAOi0Y1dXV2PAgAEP3balpUW7jjn1WAtL7OP9bWwtOpKHMTc0BAcHY9GiRSguLsbq1au15d3V5m5ubgDuvYyAiHT+Tp48aXCcxti4cSN27tyJb7/9Fv3799dZVlxcjLa2Nr1ypVIJFxcXFBUV6ZQ3NzcD+E/bWpvffEIuKyvDhg0bMGDAALz++usPXW/UqFFwcHDAjz/+qFN+6tQpNDc344knnnjotnl5eSAiPP300wbXY2tri5aWFjP2zHIssY/3t3F3191T+vXrB5FIZPT84tWrV2PYsGE4ffq0tsyc/nm/gQMHQiqVIj8/36iYTEFEiI2NRWFhIbKzszs9u+/4j+TmzZs65bW1tbhz5452+luHjrZ0d3fvoah71m8mIRMR6urq0N7eDiJCRUUFMjIyEBISgj59+iA7O7vLMWSpVIrFixdj//792LlzJzQaDQoLC7FgwQJ4enoiKipKu257ezuqqqrQ2tqKgoICREdHw9vbG/PmzTO4noCAANy5cwfZ2dloaWlBRUWFzpxLAHBxccGNGzdw+fJl1NbWCiYB9cY+dtXG5tSdm5vba9Pe5HI5/Pz8cO3aNaO26xi6uH/+rjH981F1z58/H7t370Z6ejo0Gg3a2tpw7do1bVKMjIyEu7u72bdrnzlzBh9//DG2bNkCsVis94iD9evXw9fXF5MmTcKWLVvw3XffobGxEaWlpdr9eeONN3Tq7GjL0aNHmxWbxVhododWT057O3jwII0ZM4bkcjnZ2dmRjY0NASCRSEROTk705JNPUkJCAt2+fVu7TVJSEslkMu00uB07dmiXtbe3U3JyMg0ePJjEYjE5OztTeHg4nTt3TrtOVFQUicVi8vLyIltbW1IqlTRt2jQqKSkxqp7bt2/TpEmTSCqVkq+vL73zzjsUExNDACggIICuXr1KP/30E/n4+JBMJqMJEyZopzb1BBg5Lagn99GQNja17kOHDpFCoaDExESj28iUaW9qtZrEYjE1NDRoy/bv30/+/v4EgFxdXWnhwoWdbhsTE6Mz7e1Rbb5p0yaSy+UEgAYPHkwlJSW0efNmUiqVBIB8fHzo/PnzdPfuXYqNjSVvb2+ytbUlNzc3ioiIoKKiIiIiCg8PJwAUHx/f5b6dPHmSQkJCyNPTkwAQAPLw8CCVSkXHjh2jwsJCbXlnf8nJyUREVFlZSdHR0RQQEEASiYQcHBwoJCSEvvzyS73PDAsLIy8vL2pvbzfqOBjbv3vKY52QLSEqKopcXFwsHUa3E0qHJRJuG5uSkIuLi8nW1lbnP36ha2tro4kTJ9K2bdssHYqOyspKkkqltH79eqO3FUr//s0MWfSmx+KpUwL3uLRxQEAAEhISkJCQoPeEMyFqa2tDdnY2amtrERkZaelwdKxatQpBQUFQq9WWDsVknJAZs7C4uDjMnDkTkZGRgn+AUF5eHrKyspCbm2vw/OnekJKSgvz8fBw6dAhisdjS4ZiME3I3WrZsGbZv346amhr4+vpi3759lg7psfO4tvGaNWugVqvx0UcfWTqULk2ePBm7du3SeU6IpR04cAB3795FXl4enJ2dLR2OWUREln0KR2ZmJmbPnm29DwP5jRCJRMjIyMCsWbMsHYpgzZw5EwCwd+9eC0fCjCWU/s1nyIwxJhCckBljTCA4ITPGmEBwQmaMMYHghMwYYwJha+kAOtz/viwmTLNnz8bs2bMtHYbgcV9mprL4tLdr167hxIkTlgyBWbkNGzYAAN577z0LR8KsmUqlsvjjWy2ekBkzV8fc0czMTAtHwph5eAyZMcYEghMyY4wJBCdkxhgTCE7IjDEmEJyQGWNMIDghM8aYQHBCZowxgeCEzBhjAsEJmTHGBIITMmOMCQQnZMYYEwhOyIwxJhCckBljTCA4ITPGmEBwQmaMMYHghMwYYwLBCZkxxgSCEzJjjAkEJ2TGGBMITsiMMSYQnJAZY0wgOCEzxphAcEJmjDGB4ITMGGMCwQmZMcYEghMyY4wJBCdkxhgTCE7IjDEmEJyQGWNMIDghM8aYQHBCZowxgeCEzBhjAmFr6QAYM0ZlZSU0Go1OWX19PQDg4sWLOuVKpRKurq69Fhtj5hIREVk6CMYMtX37drz++usGrfv3v/8d8+fP7+GIGOs+nJCZVampqYGbmxtaWlq6XE8sFqOiogKOjo69FBlj5uMxZGZVHB0d8eKLL8LW9uGjbba2tggLC+NkzKwOJ2RmdebOnYu2traHLm9vb8fcuXN7MSLGugcPWTCr09TUBFdXV+3FvAfJ5XJUVlZCJpP1cmSMmYfPkJnVkUqlmD59OsRisd4ysViMGTNmcDJmVokTMrNKr776aqcX9lpaWvDqq69aICLGzMdDFswqtba2wt3dHXfu3NEpd3JyQkVFRZcX/RgTKj5DZlbJ1tYWr7zyis6whVgsxty5czkZM6vFCZlZrVdeeUVn2KKlpQWvvPKKBSNizDw8ZMGsFhFh4MCBuH79OgDA09MT169fh0gksnBkjJmGz5CZ1RKJRHjttddgZ2cHOzs7/PGPf+RkzKwanyEzq1ZQUIDAwEDtv0ePHm3hiBgzndVd/Th58iRSUlIsHQYTEAcHBwBAQkKChSNhQrJo0SIEBwdbOgyjWN2QRWlpKfbt22fpMKzevn37cO3aNUuH0S18fHwwaNCgbq/3+++/x/fff9/t9bKet2/fPpSWllo6DKNZ3Rlyh71791o6BKsmEonw3nvvYdasWZYOxWwdz0H28/Pr1npnzpwJgPuaNbLWawlWm5AZ69DdiZgxS7G6IQvGGHtccUJmjDGB4ITMGGMCwQmZMcYE4jeZkP/0pz9BoVBAJBIhPz/f0uEYpaWlBWvXrkVAQADs7Ozg5OSEUaNG4fLly70ey6FDh+Do6Iivvvqq1z/bGhw5cgRxcXHIysqCn58fRCKR9u7CBz333HNQKBTo06cPRo4ciZ9++skCERumvb0dGzZsgEql0luWkJCAESNGQKlUQiKRICAgAO+//z7q6up01vviiy8wfvx4KBQK+Pj4YP78+SgrKwMAHDx4EElJSV2+FeZx9ZtMyFu3bsWWLVssHYZJZs+ejc8//xy7du1CQ0MDfv31V/j7++t1+N7AN3k+3AcffIC0tDQsW7YMERERuHjxIvz9/dG3b1/s3LkTX3/9tc76//jHP7B371689NJLKCoqwtixYy0UedeKi4vxu9/9DosWLUJDQ4Pe8m+//RYLFy7E5cuXUVlZibVr1yI1NVU7hRAAMjIyMGfOHMycORPXrl3DgQMH8N1332HKlClobW3F1KlTIZVKMXnyZFRXV/fm7lkeWZmMjAzqjrB3795NAOj06dPdEFXv2L17N4lEIiooKDC7LgCUkZHRDVFZTkNDAwUHB/dY/TNmzKAZM2YYvd1HH31EQ4YMocbGRp1yf39/2rVrF9nY2JCXlxdVV1frLM/NzaWXX37ZrJh7Un5+Pk2fPp127txJQUFBFBgYqLdOWFgYtba26pTNmjWLANDVq1eJiGjSpEnUv39/am9v167z6aefEgA6fvy4tkytVlNwcDC1tLQYHau19u/f5BkyYJ0Tx//6179i7Nix/LyG/2/btm0oLy+3dBg6Lly4gJUrV+LDDz+EVCrVW65SqRAdHY3r169jyZIlFojQdIGBgcjKysKcOXMgkUg6XScnJwd9+vTRKXN1dQUA7Rl1aWkpPD09db6DAwcOBABcuXJFW7Zq1Srk5+cjNTW1W/dDyH4TCZmIkJycjKFDh0IikcDR0RExMTE667S1tSE+Ph7e3t6QyWQYM2YMMjIyAADp6emwt7eHXC7HgQMHMGXKFCiVSgwYMAC7d+/W1nHs2DE8+eSTkMvlUCqVGD16NDQazSPrN0RzczO+//57BAUFdUOLmO/48ePw9vaGSCTCp59+CsCwdkpLS4NUKkW/fv3w1ltvwdPTE1KpFCqVCqdOnQIAqNVq2NnZwcPDQ/t5b7/9Nuzt7SESiVBZWYno6GgsXrwYJSUlEIlECAgIAABwG7YOAAAgAElEQVQcPnwYSqUSa9as6eUWgXb/iAhTp0596DqJiYkYMmQItm7diiNHjjx0PSJCSkoKhg8fDolEAmdnZ0ybNg1nz54FYHi/NLfvmev69euQyWTw9fUFcO9Gngf/I+0YP77/Jh9nZ2eEhoYiNTX1tzM8ZtkTdOOZMmSxfPlyEolE9Mknn1BVVRU1NDTQpk2bdIYslixZQhKJhPbt20dVVVW0bNkysrGxoR9++EFbBwA6evQo1dTUUHl5OU2cOJHs7e2pubmZ6urqSKlUUlJSEjU2NlJZWRlNnz6dKioqDKr/US5dukQAKCgoiJ555hny8PAgiURCw4YNo08//VTn558h0A0/6UpLSwkAbdy4UVv2qHYiIoqKiiJ7e3s6c+YMNTU1UVFREY0fP54UCoX2Z+2cOXPI3d1d5/OSk5MJgLZNIyIiyN/fX2ednJwcUigUlJCQYNa+EZk2ZOHn50cjRozodJm/vz9dunSJiIhOnDhBNjY2NGjQIKqrqyMi/SGL+Ph4srOzox07dlB1dTUVFBTQ2LFjydXVlcrKyojIsPY2t+915qmnnup0yOJB9fX1pFAoSK1Wa8vy8vJILBZTWloaaTQa+uWXX2j48OH0/PPP620fFxdn0tBid/RvS3jsE3JDQwPJ5XL6/e9/r1N+/xhyY2MjyeVyioyM1NlOIpHQX/7yFyL6T8e/f1ywI6lfuHCBfvnlFwJAOTk5ejEYUv+jFBYWEgD6/e9/T//+97/p9u3bVF1dTUuXLiUAtHPnToPbhKjnE/LD2onoXkJ2dHTUqeuHH34gAPThhx8SkekJuTsZm5Dr6upIJBLRSy+91Ony+xMyEdHixYsJAC1cuJCIdBNyQ0MDOTg46PQZIqL/+7//IwDa/3Ae1d7d0fc6Y2hCXr58OQ0ZMoQ0Go1O+YoVKwiA9m/AgAFUWlqqt/3f//53AkCff/65UfFZa0J+7IcsLly4gIaGBkyePPmh65w7dw4NDQ0YNWqUtkwmk8HDw0P787AzdnZ2AO5NRfPz80O/fv0wd+5crFq1Smcamqn1369jzG7kyJFQqVRwcXGBo6MjPvzwQzg6OmLz5s0G1WMJ97fTw4wbNw5yudzg9hCi8vJyEBHkcrlB6ycmJmLo0KHYtGkTjh8/rrOsqKgIdXV1GDdunE75+PHjYWdnpx3e6cz97d0dfc9U+/fvR2ZmJr755hsoFApt+fLly7F582YcPXoUdXV1uHjxIlQqFYKDg/We0NbRlrdu3erRWIXisU/IHY+YdHNze+g69fX1AIAVK1Zo54qKRCJcuXKl06k9nZHJZPj2228xYcIErFmzBn5+foiMjERjY2O31O/p6QkAqKys1Cm3s7ODj48PSkpKDKpHyCQSCSoqKiwdhsmampoA4KEXvB4klUqxfft2iEQivP7662hsbNQu65ju1fGs5/s5OTmhtrbWoM/ojr5nij179mDdunXIy8vTeTTqzZs3kZSUhDfffBPPPvss7O3t4evriy1btuDGjRtITk7WqUcmkwH4T9s+7h77hNxxpfvu3bsPXacjWW/YsAF0bxhH+3fy5EmDP2vkyJH46quvcOPGDcTGxiIjIwPr16/vlvodHBwwePBgnDlzRm9Za2srHB0dDY5TiFpaWlBdXY0BAwZYOhSTdSQPY25oCA4OxqJFi1BcXIzVq1dry52cnACg08RrTDt1V982xsaNG7Fz5058++236N+/v86y4uJitLW16ZUrlUq4uLigqKhIp7y5uRnAf9r2cffYJ+RRo0bBxsYGx44de+g6AwcOhFQqNeuuvRs3bmiTpZubGz766COMHTsWZ86c6Zb6gXs3hZw+fVr7/F/g3lSiK1euWP1UuLy8PBARnn76aQCAra1tl0McQtSvXz+IRCLU1NQYtd3q1asxbNgwnD59Wls2atQoODg44Mcff9RZ99SpU2hubsYTTzxhUN3d1fcMQUSIjY1FYWEhsrOzOz277/iP5ObNmzrltbW1uHPnjnb6W4eOtnR3d++hqIXlsU/Ibm5umDFjBvbt24dt27ZBo9GgoKBAZ8xVKpVi/vz52L17N9LT06HRaNDW1oZr167pdZyHuXHjBt566y2cPXsWzc3NOH36NK5cuYKnn366W+oH7r2SxsfHB/PmzcPVq1dx+/ZtxMbGorGxEUuXLjW6bSypvb0dVVVVaG1tRUFBAaKjo+Ht7Y158+YBAAICAnDnzh1kZ2ejpaUFFRUVOnNUAcDFxQU3btzA5cuXUVtbi5aWFuTm5lps2ptcLoefn5/Rb2LpGLq4f/6uVCrF4sWLsX//fuzcuRMajQaFhYVYsGABPD09ERUVZXDdj+p7kZGRcHd3N/t27TNnzuDjjz/Gli1bIBaLdYZIRCIR1q9fD19fX0yaNAlbtmzBd999h8bGRpSWlmr354033tCps6Mtrf2Ew2AWuJBoFlOmvdXW1tKf//xn6tu3Lzk4ONCECRMoPj5ee3X3559/prt371JsbCx5e3uTra0tubm5UUREBBUVFdGmTZtILpcTABo8eDCVlJTQ5s2bSalUEgDy8fGh//3f/yWVSkXOzs7Up08f6t+/Py1fvlx711JX9RujtLSUXnnlFXJ2diaJREJPPvkk5ebmGlUHkflXoTdu3EgeHh4EgORyOU2dOtWgdjp//jxFRUWRWCwmLy8vsrW1JaVSSdOmTaOSkhJt/bdv36ZJkyaRVColX19feueddygmJoYAUEBAAF29epV++ukn8vHxIZlMRhMmTKCysjI6dOgQKRQKSkxMNHnfOpgy7U2tVpNYLKaGhgZt2f79+8nf358AkKurq3ZWxYNiYmJ0pr21t7dTcnIyDR48mMRiMTk7O1N4eDidO3eOiMjg9n5U3wsPDycAFB8f3+W+nTx5kkJCQsjT01M7O8LDw4NUKhUdO3ZMOxPoYX/JyclERFRZWUnR0dEUEBBAEomEHBwcKCQkhL788ku9zwwLCyMvLy+LTOu0hN9EQmb6LNlho6KiyMXFxSKfbQxTEnJxcTHZ2trSjh07eiiq7tfW1kYTJ06kbdu2WToUHZWVlSSVSmn9+vVGb2utCfmxH7JgwvS4PskrICAACQkJSEhIsMgDn4zV1taG7Oxs1NbWIjIy0tLh6Fi1ahWCgoKgVqstHUqv4YRsYWfPntUba+vsT2hfFvZwcXFxmDlzJiIjI42+wNfb8vLykJWVhdzcXIPnT/eGlJQU5Ofn49ChQxCLxZYOp9dwQrawYcOG6U1H6uxvz549lg61Wyxbtgzbt29HTU0NfH19sW/fPkuH1CPWrFkDtVqNjz76yNKhdGny5MnYtWuXznNDLO3AgQO4e/cu8vLy4OzsbOlwehW/dZr1qrVr12Lt2rWWDqNXPPfcc3juuecsHYbVefnll/Hyyy9bOgyL4DNkxhgTCE7IjDEmEJyQGWNMIDghM8aYQHBCZowxgbDaWRbW+E48oZk9ezZmz55t6TAEj/sa6y1Wm5B7851gj6PZs2cjOjoawcHBlg5FsDZs2AAAeO+99ywcCTOWtZ5oWG1CnjVrlqVDsGqzZ89GcHAwt2MX9u7dC4D7mjWy1oTMY8iMMSYQnJAZY0wgOCEzxphAcEJmjDGB4ITMGGMC8ZtOyFlZWfDz89N79rCdnR369euHZ555BsnJyaiqqrJ0qMwKHTlyBHFxcXr97LXXXtNb97nnnoNCoUCfPn0wcuRIs99v1xMSExM7fVb3qFGj9NZtb2/Hhg0boFKp9JYlJCRgxIgRUCqVkEgkCAgIwPvvv699oP/BgweRlJT02L7EoCu/6YQcERGBixcvwt/fH46OjiAitLe3o7y8HJmZmfD19UVsbCxGjhyp9/ZfxrrywQcfIC0tDcuWLdPpZ3379sXOnTvx9ddf66z/j3/8A3v37sVLL72EoqIijB071kKRm6+4uBi/+93vsGjRIjQ0NOgt//bbb7Fw4UJcvnwZlZWVWLt2LVJTUzFz5kwAwNSpUyGVSjF58mRUV1f3dvgW9ZtOyJ0RiURwcnLCM888g+3btyMzMxO3bt1CWFiY4N/+YA0aGxs7PWsSet3GWLduHfbs2YPMzEwoFAqdZWlpabCxsUFUVJRV9qcdO3bovTzhl19+0S7/+eefsXTpUixYsABBQUGd1uHg4ICoqCi4uLhAoVBg1qxZCA8Px+HDh1FaWgoAePfddxEYGIgXX3wRra2tvbJvQsAJ+RFmzJiBefPmoby8HJ999pmlw7F627ZtQ3l5udXVbagLFy5g5cqV+PDDDyGVSvWWq1QqREdH4/r161iyZIkFIuxZgYGByMrKwpw5cyCRSDpdJycnB3369NEpc3V1BQCdM+pVq1YhPz8fqampPRewwHBCNsC8efMAALm5uQDuvRgyPj4e3t7ekMlkGDNmjPZW7vT0dNjb20Mul+PAgQOYMmUKlEolBgwYgN27d2vrPHbsGJ588knI5XIolUqMHj0aGo3mkfVbChEhJSUFw4cPh0QigbOzM6ZNm4azZ88CANRqNezs7HReBfT222/D3t4eIpEIlZWViI6OxuLFi1FSUgKRSISAgACkpaVBKpWiX79+eOutt+Dp6QmpVAqVSoVTp06ZVTcAHD58GEqlEmvWrOmVdkpLSwMRYerUqQ9dJzExEUOGDMHWrVtx5MiRh673qDY3tK8JsT896Pr165DJZPD19dWWOTs7IzQ0FKmpqSAiC0bXi3r7NdfmysjIoO4O29/fnxwdHR+6XKPREAAaOHAgEREtWbKEJBIJ7du3j6qqqmjZsmVkY2NDP/zwAxERLV++nADQ0aNHqaamhsrLy2nixIlkb29Pzc3NVFdXR0qlkpKSkqixsZHKyspo+vTpVFFRYVD93QFGviY9Pj6e7OzsaMeOHVRdXU0FBQU0duxYcnV1pbKyMiIimjNnDrm7u+tsl5ycTAC0+xYREUH+/v4660RFRZG9vT2dOXOGmpqaqKioiMaPH0+K/9fevQdFcWb9A/8OzAwzAzMCcpWLwOAlKt6iUVBj8rpryliKiBd2NRtjaaEbQ4xoCF6IAbwFFyxdeFMmLlWreQUUC40Rs2t8ccuSV91SIuJ6jYBKEDAgV2Fkzu+P/Jg4AWQGBrpHz6eKf7qfPn36mYdD0/1Mt1pNpaWlPYp9/PhxUqvVFB8fb/Kxtpk3bx7NmzfPrG0CAgJo2LBhHa7TarV09+5dIiI6d+4c2djYkJ+fH9XX1xMRUW5uLoWGhhram9LnXY01IsuNp4SEBPL29iZHR0eSyWTk5+dHoaGhdOHChQ7bT5gwgUaNGtVl3IaGBlKr1RQVFdVuXWxsLAGgy5cvm5WrueNbLLggU9cFmYhIIpGQo6MjNTU1kUqlooiICMO6xsZGsrOzoz//+c9E9OsvSVNTk6FNamoqAaDbt2/T1atXCQAdP3683X5MiW8J5gzYxsZGcnBwMMqJiOjChQsEwFDselKQf9v/Fy9eJAD02Wef9Sh2T5hbkOvr60kikdCsWbM6XP9sQSYiio6OJgC0atUqIjIuyKb2eVdjzZLjqbS0lC5dukR1dXXU3NxM+fn5NGbMGFIqlXT16tV27U0tyBs2bKDBgwdTbW1tu3V/+9vfCAD9/e9/NytXay3IfMnCBA0NDSAiaDQa3LhxA42NjUZTfZRKJTw8PAz/SnZELpcDAHQ6HQICAuDm5obFixdj8+bNKC4uNrTrbvzeVFRUhPr6eowbN85o+fjx4yGXyw2XFixp3LhxUKlUgh1zd1RUVICIoFKpTGqfmJiIIUOGIDU1FWfPnjVa15M+f3asWXI8+fj4YMyYMXBwcIBcLsfEiRORnp6OpqYmpKammhWrzZEjR5CVlYXvvvuu3Q1QAIa+fPjwYbfiWxsuyCa4efMmAGDo0KFoaGgAAGzcuNFoLmZJSUmHU3w6olQqcfr0aUyePBlbtmxBQEAAIiIi0NTUZJH4ltY29cjBwaHdOkdHR9TV1fXKfu3s7FBZWdkrsXvDkydPAKDTm1m/pVAokJ6eDolEgqVLl6KpqcmwzlJ93tvjKSgoCLa2tobfEXNkZGRg+/btyMvLg5+fX4dtlEolgF/79kXHBdkEJ0+eBADMmDEDrq6uAH55Vi79ZvpPfn6+yTGHDx+Ob775BmVlZYiJiUFmZiZ27txpsfiW5OjoCAAdFoGamhp4e3tbfJ86na7XYveWtuJhzhcagoODsWbNGty6dQsJCQmG5Zbq894eT3q9Hnq93uQ/Qm327NmDAwcO4PTp0xgwYECn7VpaWgD82rcvOi7IXSgvL0dKSgq8vb2xdOlS+Pj4QKFQoKCgoNsxy8rKcO3aNQC//MJs27YNY8eOxbVr1ywS39JGjBgBBweHdl+OOX/+PFpaWvDqq68CAKRSKXQ6nUX2mZeXByLCxIkTLR67t7i5uUEikZg9vzghIQFDhw7F5cuXDctM7fOuWHI8vfXWW+2WXbx4EURk8osOiAgxMTEoLCxETk5Oh/8BPKutL93d3c1P2ApxQf7/iAj19fXQ6/UgIlRWViIzMxOTJk2Cra0tcnJyoNFooFAo8N577+HgwYNIS0tDbW0tWltbcf/+ffz0008m7ausrAwrVqzA9evX0dLSgsuXL6OkpAQTJ060SHxLUygUiI6OxpEjR3DgwAHU1taisLAQK1euhKenJyIjIwEAgYGB+Pnnn5GTkwOdTofKykqUlJQYxXJ2dkZZWRmKi4tRV1dnKLJ6vR7V1dV4+vQprly5gtWrV8PX19cw5bC7sXNzc/ts2ptKpUJAQADu379v1nZtly6enZtrap+bErur8RQREQF3d/cuv6794MEDZGRkoKamBjqdDvn5+Vi2bBl8fX2xcuVKk/K5du0aPv/8c3z55ZeQyWTtvoa9c+dOo/ZtfRkUFGRSfKvXxzcRe8ySsyyOHTtGI0eOJJVKRXK5nGxsbAiAYUbFa6+9RvHx8fTo0SOj7ZqbmykmJoZ8fX1JKpWSq6srhYeHU1FREaWmppJKpSIANGjQILpz5w7t3buXNBoNAaCBAwfSP//5TwoJCSEnJyeytbWlAQMG0IYNG+jp06ddxrcUmHkXWq/XU1JSEg0aNIhkMhk5OTlRWFgY3bhxw9Dm0aNH9Oabb5JCoSB/f3/64IMPaN26dQSAAgMDDXfpBw4cSEqlkiZPnkzl5eUUGRlJMpmMvLy8SCqVkkajoTlz5tCdO3d6HPvEiROkVqspMTHR7D7qzrS3qKgokslk1NjYaFh25MgR0mq1BIBcXFwMsyp+a926dUbT3rrqc1PG2s2bN7scT2FhYQSA4uLinnts0dHRpNVqyd7enqRSKXl7e9Py5cuprKzM0CY/P58mTZpEnp6eBIAAkIeHB4WEhNCZM2eosLDQsLyjn6SkJKN9zpw5k7y8vEiv15v1OZg7vsXipS7ILzMxDdjIyEhydnYWOo12ulOQb926RVKplPbv399LWVlea2srTZkyhfbt2yd0KkaqqqpIoVDQzp07zd5WTOPbHHzJgonCi/Jkr8DAQMTHxyM+Pt7w9DIxa21tRU5ODurq6hARESF0OkY2b96M0aNHIyoqSuhU+gwXZMYsLDY2FvPnz0dERIToHyCUl5eH7Oxs5Obmmjx/ui8kJyejoKAAJ06cgEwmEzqdPsMFmQlq/fr1SE9Px+PHj+Hv74/Dhw8LnZJFbNmyBVFRUdi2bZvQqTzXtGnT8PXXXxs9J0RoR48eRXNzM/Ly8uDk5CR0On1KKnQC7OW2detWbN26Veg0esX06dMxffp0odOwOqGhoQgNDRU6DUHwGTJjjIkEF2TGGBMJLsiMMSYSXJAZY0wkrPamXlZWltApWD2hHlZkLdq+tstjjfUVCZF1vRslKysLCxcuFDoNxpjIZWZmYsGCBUKnYRarK8iM/VbbLx2fyTJrx9eQGWNMJLggM8aYSHBBZowxkeCCzBhjIsEFmTHGRIILMmOMiQQXZMYYEwkuyIwxJhJckBljTCS4IDPGmEhwQWaMMZHggswYYyLBBZkxxkSCCzJjjIkEF2TGGBMJLsiMMSYSXJAZY0wkuCAzxphIcEFmjDGR4ILMGGMiwQWZMcZEggsyY4yJBBdkxhgTCS7IjDEmElyQGWNMJLggM8aYSHBBZowxkeCCzBhjIsEFmTHGRIILMmOMiQQXZMYYEwkuyIwxJhJckBljTCSkQifAmDn+9a9/IT8/32jZ9evXAQA7duwwWh4cHIzXX3+9z3JjrKckRERCJ8GYqb7//nv87ne/g0wmg41Nx//g6fV66HQ6nDp1CtOmTevjDBnrPi7IzKro9Xp4eHigsrLyue1cXFxQXl4OW1vbPsqMsZ7ja8jMqtjY2GDRokWQy+WdtpHL5Vi8eDEXY2Z1uCAzq/OHP/wBLS0tna5vaWnBH/7whz7MiDHL4EsWzCr5+fmhpKSkw3U+Pj4oKSmBRCLp46wY6xk+Q2ZW6Z133oFMJmu3XCaTYcmSJVyMmVXiM2Rmla5fv45XXnmlw3VXr17F8OHD+zgjxnqOz5CZVRo6dCiGDx/e7kx42LBhXIyZ1eKCzKzWn/70J6OZFDKZDO+++66AGTHWM3zJglmte/fuYeDAgWgbwhKJBD/++CP8/PyETYyxbuIzZGa1fHx8MGHCBNjY2MDGxgYTJkzgYsysGhdkZtXeeecdSCQS2NjY4J133hE6HcZ6hC9ZMKtWVVUFDw8PAEBZWRnc3NwEzoix7hO8IGdlZWHhwoVCpsAYY8jMzMSCBQsEzUE0j9/MzMwUOgX2HAsXLsTq1asRHBwsdCrt/Otf/4JEIsGUKVMEzSMlJQUA8NFHHwmaBzOfWE4KRVOQhf7LxJ5v4cKFCA4OFuXnNGPGDACAWq0WNI9Dhw4B4LFsjbggM2YhQhdixiyFZ1kwxphIcEFmjDGR4ILMGGMiwQWZMcZE4oUoyMuWLYNarYZEIkFBQYHQ6fQpvV6PlJQUhISE9KhNXzhx4gT69euHb775RtA8xOrUqVOIjY1FdnY2AgICIJFIIJFIOvwG4vTp06FWq2Fra4vhw4fj0qVLAmT8fImJiYZjePZnxIgR7do+b4zGx8dj2LBh0Gg0sLOzQ2BgID7++GPU19cDAI4dO4YdO3agtbW114+pt70QBfmrr77Cl19+KXQafe7WrVt4/fXXsWbNGjQ2Nna7TV/hL4V27tNPP8Xu3buxfv16hIeH48cff4RWq0X//v1x4MABfPvtt0bt//GPf+DQoUOYNWsWioqKMHbsWIEy77muxujp06exatUqFBcXo6qqClu3bsWuXbswf/58AMDs2bOhUCgwbdo01NTU9HX6FvVCFGRr1tTU1K0z1x9++AGffPIJVq5cidGjR3e7TV+aOXMmHj9+jFmzZgmy/+72dW/bvn07MjIykJWV1W4K3+7du2FjY4PIyEg8fvxYoAy7b//+/SAio5+rV68a1psyRh0cHBAZGQlnZ2eo1WosWLAAYWFhOHnyJO7duwcA+PDDDzFq1Ci8/fbbePr0aZ8cW294YQqytb6yZ9++faioqDB7u1GjRiE7OxuLFi2CnZ1dt9u8TLrb173p9u3b2LRpEz777DMoFIp260NCQrB69Wo8ePAAa9euFSDD3mXKGD1+/Hi7N4i7uLgAgNEZ9ebNm1FQUIBdu3b1XsK9zCoLMhEhKSkJQ4YMgZ2dHfr164d169YZ1n/++edQqVRQq9WoqKhAdHQ0vLy8cOPGDRARkpOT8corr8DOzg5OTk6YM2cOrl+/DuCXMxKFQgE3NzesWLECnp6eUCgUCAkJwfnz541yeF6cqKgoyOVyw4NvAOD999+Hvb09JBIJqqqqsHr1akRHR+POnTuQSCQIDAzsox7se2fPnoWvry8kEgn++te/AgDS0tJgb28PlUqFo0ePYsaMGdBoNPD29sbBgwcBmPZ59KSvT548CY1Ggy1btvRxj8BwfESE2bNnd9omMTERgwcPxldffYVTp0512q6rMWlKfwNAa2sr4uLi4OvrC6VSiZEjR4ru0QYPHjyAUqmEv7+/YZmTkxOmTp2KXbt2We/lMRJYZmYmmZvGhg0bSCKR0F/+8heqrq6mxsZGSk1NJQB0+fJlQxsA9OGHH9KePXto7ty59J///Ifi4uJILpfT/v37qaamhq5cuUJjx44lFxcXKi8vJyKiyMhIsre3p2vXrtGTJ0+oqKiIxo8fT2q1mkpLS4mITIqzaNEicnd3N8o9KSmJAFBlZSUREYWHh5NWq+1RH06YMIFGjRrV4zbPA4AyMzO7vT0R0b179wgA7dmzx7Cs7XP6/vvv6fHjx1RRUUFTpkwhe3t7amlpISLTPo/u9vXx48dJrVZTfHx8j46NiGjevHk0b948s7YJCAigYcOGdbhOq9XS3bt3iYjo3LlzZGNjQ35+flRfX09ERLm5uRQaGmpob8qYNKW/165dS3Z2dnT48GGqrq6m9evXk42NDV28eNGsY0tISCBvb29ydHQkmUxGfn5+FBoaShcuXOiwvaljtKGhgdRqNUVFRbVbFxsba1QHTGWJ8W0JVneG3NTUhJSUFPzud7/DmjVr4OjoCKVSCWdn5w7bb9++HatWrUJ2djYGDhyI5ORkzJ07F4sXL0a/fv0QFBSEL774AlVVVdi7d69hO6lUajjTGDZsGNLS0lBXV4f09HQ0NTWZHIeZJiQkBBqNBq6uroiIiEBDQwNKS0sN65/3efTEzJkzUVtbi02bNvX0EMzW0NCAu3fvQqvVdtk2ODgYH330EYqLi/HJJ5+0W2/umOysv588eYK0tDSEhYUhPDwcjo6O2LhxI2Qymdl9/e677+LYsWO4d+8e6uvrcfDgQZSWlmLq1KkoKioyK9aztm7dCk9PTyQmJrZbN2jQIABAYWFht+MLyeoK8u3bt9HY2Ihp06aZvW1RURHq6+sxbtw4o+Xjx4+HXC43uiTxW+PGjYNKpcL169d7FId1TS6XAwB0Ol2nbZ79PJSJ0KIAACAASURBVKxVRUUFiAgqlcqk9omJiRgyZAhSU1Nx9uxZo3U9GZPP9veNGzfQ2NhoNDVNqVTCw8PD7L728fHBmDFj4ODgALlcjokTJxpOaFJTU82K1ebIkSPIysrCd9991+EzTNr68uHDh92KLzSrK8j3798HALi6upq9bduUGAcHh3brHB0dUVdX99zt7ezsUFlZ2eM4zDLaPg9r9eTJEwAw+YarQqFAeno6JBIJli5diqamJsM6S43JhoYGAMDGjRuN5g6XlJRYZNpkUFAQbG1tcfPmTbO3zcjIwPbt25GXl9fpq7qUSiWAX/vW2lhdQW67E93c3Gz2to6OjgDQ4eCsqamBt7d3p9vqdDpDm57EYZbx7OdhrdqKhzlfaAgODsaaNWtw69YtJCQkGJZbaky2neikpKS0m66Wn59vcp6d0ev10Ov1Zs/62bNnDw4cOIDTp09jwIABnbZraWkB8GvfWhurK8gjRoyAjY0Nzpw5061tHRwc8O9//9to+fnz59HS0oJXX321023z8vJARJg4caLJcaRS6XP/7Wbd9+znAVhnX7u5uUEikZg9vzghIQFDhw7F5cuXDct6Mraf5ePjA4VCYZFvvL711lvtll28eBFEZPKLDogIMTExKCwsRE5OTof/ATyrrS/d3d3NT1gErK4gu7q6Yt68eTh8+DD27duH2tpaXLlyxaQbaQqFAtHR0Thy5AgOHDiA2tpaFBYWYuXKlfD09ERkZKShrV6vR3V1NZ4+fYorV65g9erV8PX1xZIlS0yOExgYiJ9//hk5OTnQ6XSorKxESUmJUU7Ozs4oKytDcXEx6urqrK6o9JXnfR5A9/s6NzdXsGlvKpUKAQEBhstwpmq7dPHs3FxzxnZXsd977z0cPHgQaWlpqK2tRWtrK+7fv4+ffvoJABAREQF3d/cuv6794MEDZGRkoKamBjqdDvn5+Vi2bBl8fX2xcuVKk/K5du0aPv/8c3z55ZeQyWTtvoa9c+dOo/ZtfRkUFGRSfNERbH7H/9edaW91dXW0fPly6t+/Pzk4ONDkyZMpLi6OAJC3tzctWrSIlEolASAfHx/av3+/YVu9Xk9JSUk0aNAgkslk5OTkRGFhYXTjxg1Dm8jISJLJZOTl5UVSqZQ0Gg3NmTOH7ty5Y1acR48e0ZtvvkkKhYL8/f3pgw8+oHXr1hEACgwMpNLSUrp06RINHDiQlEolTZ482TA9qSv5+fk0adIk8vT0JAAEgDw8PCgkJITOnDljchtToYfTgvbs2UMeHh4EgFQqFc2ePZtSU1NJpVIRABo0aBDduXOH9u7dSxqNhgDQwIED6ebNmyZ9Ht3t6xMnTpBarabExMRuH1ub7kx7i4qKIplMRo2NjYZlR44cIa1WSwDIxcWFVq1a1eG269atM5r21tWYNLW/m5ubKSYmhnx9fUkqlZKrqyuFh4dTUVERERGFhYURAIqLi3vusUVHR5NWqyV7e3uSSqXk7e1Ny5cvp7KyMkObrsZoYWGhYXlHP0lJSUb7nDlzJnl5eZFerzfrc+jp+LYUqyzIvS0yMpKcnZ2FTkNUhByw1vJ5dKcg37p1i6RSqdFJg9i1trbSlClTaN++fUKnYqSqqooUCgXt3LnT7G3FUpCt7pJFX3kRnhz1InlRP4/AwEDEx8cjPj7e8PQyMWttbUVOTg7q6uoQEREhdDpGNm/ejNGjRyMqKkroVLqNC7KIXL9+vcPHFf72R2y/CKxnYmNjMX/+fERERIj+AUJ5eXnIzs5Gbm6uyfOn+0JycjIKCgpw4sQJyGQyodPpNi7Iv7F+/Xqkp6fj8ePH8Pf3x+HDh/ts30OHDm031aijn4yMjD7LSWhCfh59acuWLYiKisK2bduETuW5pk2bhq+//trouSFCO3r0KJqbm5GXlwcnJyeh0+kRCZGwT+HIysrCwoULrfdhIC8JiUSCzMxMfsX9c7Q9n/fQoUMCZ8LMJZbxzWfIjDEmElyQGWNMJLggM8aYSHBBZowxkZAKnUCbrKwsoVNgXbDEw2VeZG1f2+WxzLpLNLMsGGNMSGKYZSGaM2Se9iZuYpkWJGY87c16ieUlyXwNmTHGRIILMmOMiQQXZMYYEwkuyIwxJhJckBljTCS4IDPGmEi80AU5OzsbAQEB7Z4nLJfL4ebmhjfeeANJSUmorq4WOlXGcOrUKcTGxrYbt++88067ttOnT4darYatrS2GDx/e5fvthJCYmNjh87xHjBjRrq1er0dKSgpCQkLarYuPj8ewYcOg0WhgZ2eHwMBAfPzxx4YH+h87dgw7dux4IV5i8EIX5PDwcPz444/QarXo168fiAh6vR4VFRXIysqCv78/YmJiMHz48HZv62WsL3366afYvXs31q9fbzRu+/fvjwMHDuDbb781av+Pf/wDhw4dwqxZs1BUVISxY8cKlHnP3bp1C6+//jrWrFmDxsbGdutPnz6NVatWobi4GFVVVdi6dSt27dplmPc9e/ZsKBQKTJs2DTU1NX2dvkW90AW5IxKJBI6OjnjjjTeQnp6OrKwsPHz4EDNnzhT92xp+q6mpqcMzihdJbx6jWPpv+/btyMjIQFZWFtRqtdG63bt3w8bGBpGRkVY3PgFg//797V6wcPXqVcP6H374AZ988glWrlyJ0aNHdxjDwcEBkZGRcHZ2hlqtxoIFCxAWFoaTJ0/i3r17AIAPP/wQo0aNwttvv42nT5/2ybH1hpeuIP/WvHnzsGTJElRUVOCLL74QOh2z7Nu3DxUVFUKn0at68xjF0H+3b9/Gpk2b8Nlnn0GhULRbHxISgtWrV+PBgwdYu3atABn2rlGjRiE7OxuLFi2CnZ1dh22OHz8OW1tbo2UuLi4AYHRGvXnzZhQUFGDXrl29l3Ave+kLMgAsWbIEAJCbm4vPP/8cKpUKarUaFRUViI6OhpeXF27cuAEiQnJyMl555RXY2dnByckJc+bMwfXr1wH8cjajUCjg5uaGFStWwNPTEwqFAiEhITh//rxhf13FiYqKglwuN3pNzvvvvw97e3tIJBJUVVVh9erViI6Oxp07dyCRSBAYGNh3HWaC3jpGU/q4J/138uRJaDQabNmypU/6affu3SAizJ49u9M2iYmJGDx4ML766iucOnWq03Zd9XlaWhrs7e2hUqlw9OhRzJgxAxqNBt7e3jh48KAhTmtrK+Li4uDr6wulUomRI0ciMzPTcgdtAQ8ePIBSqYS/v79hmZOTE6ZOnYpdu3ZZ76MY+ur11p3JzMyk3k5Dq9VSv379Ol1fW1tLAMjHx4eIiDZs2EAA6MMPP6Q9e/bQ3Llz6T//+Q/FxcWRXC6n/fv3U01NDV25coXGjh1LLi4uVF5eTkS/vLLe3t6erl27Rk+ePKGioiIaP348qdVqKi0tJSIyKc6iRYvI3d3dKM+kpCQCQJWVlUREFB4eTlqt1uL91RGY+Zr03jxGU/q4u7GPHz9OarWa4uPjTT7WNvPmzaN58+aZtU1AQAANGzasw3VarZbu3r1LRETnzp0jGxsb8vPzo/r6eiIiys3NpdDQUEN7U/q8bWx///339PjxY6qoqKApU6aQvb09tbS0EBHR2rVryc7Ojg4fPkzV1dW0fv16srGxoYsXL5p1bAkJCeTt7U2Ojo4kk8nIz8+PQkND6cKFCx22nzBhAo0aNarLuA0NDaRWqykqKqrdutjYWAJAly9fNitXc8d3b+EzZABqtRoSiQR1dXVGy7dv345Vq1YhOzsbAwcORHJyMubOnYvFixejX79+CAoKwhdffIGqqirs3bvXsJ1UKjWcpQwbNgxpaWmoq6tDeno6mpqaTI5jrfriGJ/Xxz0xc+ZM1NbWYtOmTT3OsSsNDQ24e/cutFptl22Dg4Px0Ucfobi4GJ988km79eb2eUhICDQaDVxdXREREYGGhgaUlpbiyZMnSEtLQ1hYGMLDw+Ho6IiNGzdCJpOZ3bfvvvsujh07hnv37qG+vh4HDx5EaWkppk6diqKiIrNiPWvr1q3w9PREYmJiu3WDBg0CABQWFnY7vpC4IOOXXwwigkaj6bRNUVER6uvrMW7cOKPl48ePh1wuN7ok8Vvjxo2DSqXC9evXexTHWghxjM/2sbWoqKgAEUGlUpnUPjExEUOGDEFqairOnj1rtK4nfS6XywEAOp0ON27cQGNjo9HUNKVSCQ8PD7P71sfHB2PGjIGDgwPkcjkmTpxoOClJTU01K1abI0eOICsrC9999127G6AADH358OHDbsUXGhdkADdv3gQADB06tNM2bdNpHBwc2q1zdHRsd3b9W3Z2dqisrOxxHGsg1DG29bG1ePLkCQB0ejPrtxQKBdLT0yGRSLB06VI0NTUZ1lmqzxsaGgAAGzduNJo7XFJS0uGUNHMFBQXB1tbW8DtnjoyMDGzfvh15eXnw8/PrsI1SqQTwa99aGy7I+OVGDgDMmDGj0zaOjo4A0OHArqmpgbe3d6fb6nQ6Q5uexLEWQhzjs31sLdqKhzlfaAgODsaaNWtw69YtJCQkGJZbqs9dXV0BACkpKe2mq1nijTF6vR56vd7kP0Jt9uzZgwMHDuD06dMYMGBAp+1aWloA/Nq31ualL8jl5eVISUmBt7c3li5d2mm7ESNGwMHBod0XSM6fP4+Wlha8+uqrnW6bl5cHIsLEiRNNjiOVSqHT6XpwZMIR4hif7WNLx+4tbm5ukEgkZs8vTkhIwNChQ3H58mXDsp6Mz2f5+PhAoVCgoKDArJw68tZbb7VbdvHiRRARgoODTYpBRIiJiUFhYSFycnI6/A/gWW196e7ubn7CIvDSFGQiQn19PfR6PYgIlZWVyMzMxKRJk2Bra4ucnJznXkNWKBSIjo7GkSNHcODAAdTW1qKwsBArV66Ep6cnIiMjDW31ej2qq6vx9OlTXLlyBatXr4avry+WLFlicpzAwED8/PPPyMnJgU6nQ2VlJUpKSoxycnZ2RllZGYqLi1FXVyeaAtQXx/i8Pu5J7Nzc3D6b9qZSqRAQEGB4F5+p2i5dPDs315zx2VXs9957DwcPHkRaWhpqa2vR2tqK+/fv46effgIAREREwN3dvcuvaz948AAZGRmoqamBTqdDfn4+li1bBl9fX6xcudKkfK5du4bPP/8cX375JWQyWbuvYe/cudOofVtfBgUFmRRfdISZ3PGr3pz2duzYMRo5ciSpVCqSy+VkY2NDAEgikZCjoyO99tprFB8fT48ePTJss2PHDlIqlYZpcPv37zes0+v1lJSURIMGDSKZTEZOTk4UFhZGN27cMLSJjIwkmUxGXl5eJJVKSaPR0Jw5c+jOnTtmxXn06BG9+eabpFAoyN/fnz744ANat24dAaDAwEAqLS2lS5cu0cCBA0mpVNLkyZMNU5t6A8ycFtSbx2hKH3c39okTJ0itVlNiYqLZfdSdaW9RUVEkk8mosbHRsOzIkSOk1WoJALm4uNCqVas63HbdunVG09666vPU1FRSqVQEgAYNGkR37tyhvXv3kkajIQA0cOBAunnzJjU3N1NMTAz5+vqSVColV1dXCg8Pp6KiIiIiCgsLIwAUFxf33GOLjo4mrVZL9vb2JJVKydvbm5YvX05lZWWGNvn5+TRp0iTy9PQkAASAPDw8KCQkhM6cOUOFhYWG5R39JCUlGe1z5syZ5OXlRXq93qzPwdzx3Vte6IIshMjISHJ2dhY6DYsTy4AlEm8fd6cg37p1i6RSqdEffrFrbW2lKVOm0L59+4ROxUhVVRUpFArauXOn2duKZXy/NJcs+tKL8NQpsXtR+jgwMBDx8fGIj483PL1MzFpbW5GTk4O6ujpEREQInY6RzZs3Y/To0YiKihI6lW7jgsyYwGJjYzF//nxERESI/gFCeXl5yM7ORm5ursnzp/tCcnIyCgoKcOLECchkMqHT6TYuyBa0fv16pKen4/Hjx/D398fhw4eFTumF86L28ZYtWxAVFYVt27YJncpzTZs2DV9//bXRc0KEdvToUTQ3NyMvLw9OTk5Cp9MjEiJhn8KRlZWFhQsXWu/DQF4SEokEmZmZWLBggdCpiFbb83kPHTokcCbMXGIZ33yGzBhjIsEFmTHGRIILMmOMiQQXZMYYEwmp0Am0abshwsQrJSWFb1g9x//93/8B4LHMuk/wWRb5+flITk4WMgVm5doeRm61zy9gorBmzRqTH3rUWwQvyIz1VNtUpaysLIEzYaxn+BoyY4yJBBdkxhgTCS7IjDEmElyQGWNMJLggM8aYSHBBZowxkeCCzBhjIsEFmTHGRIILMmOMiQQXZMYYEwkuyIwxJhJckBljTCS4IDPGmEhwQWaMMZHggswYYyLBBZkxxkSCCzJjjIkEF2TGGBMJLsiMMSYSXJAZY0wkuCAzxphIcEFmjDGR4ILMGGMiwQWZMcZEggsyY4yJBBdkxhgTCS7IjDEmElyQGWNMJLggM8aYSHBBZowxkeCCzBhjIsEFmTHGRIILMmOMiYSEiEjoJBgz1d///nckJyejtbXVsKyqqgoA4OLiYlhma2uLNWvW4E9/+lOf58hYd3FBZlbl5s2bGDJkiEltb9y4gcGDB/dyRoxZDl+yYFZl8ODBGDVqFCQSSadtJBIJRo0axcWYWR0uyMzq/OlPf4KtrW2n66VSKd59990+zIgxy+BLFszqlJWVwcfHB3q9vsP1EokE9+7dg5eXVx9nxljP8BkyszoDBgxASEgIbGzaD18bGxtMmjSJizGzSlyQmVV65513OlwukUh4ZgWzWnzJglml6upquLu7Q6fTGS2XSqUoLy9H//79BcqMse7jM2RmlZycnPD73//e6Oaera0t3nrrLS7GzGpxQWZWa/HixUY39ogIixcvFjAjxnqGL1kwq9XY2Ij+/fvjyZMnAACFQoGqqirY29sLnBlj3cNnyMxqqVQqhIWFQSaTQSaTISwsjIsxs2pckJlV++Mf/widTgedToc//vGPQqfDWI9IhU6gK/fv38e5c+eEToOJVGtrK1QqFYgItbW1yMrKEjolJlIhISHw9vYWOo3nEv015KysLCxcuFDoNBhjVi4zMxMLFiwQOo3nEv0ZchuR/90QPYlEYhUDsjvOnDkDiUSC119/vUdx5s+fDwA4dOiQJdJiIvK8h1GJidUUZMY6M2XKFKFTYMwiuCAzq9fRMy0Ys0Y8khljTCS4IDPGmEhwQWaMMZHggswYYyLxUhTkZcuWQa1WQyKRoKCgQOh0TPbGG29AIpF0+OPg4NDn+Zw4cQL9+vXDN9980+f7tkanTp1CbGwssrOzERAQYPjsOnqW8/Tp06FWq2Fra4vhw4fj0qVLAmT8fImJiR2OxREjRrRrq9frkZKSgpCQkHbr4uPjMWzYMGg0GtjZ2SEwMBAff/wx6uvrAQDHjh3Djh07jN4s/rJ4KQryV199hS+//FLoNCxq8uTJfb5Pngtuuk8//RS7d+/G+vXrER4ejh9//BFarRb9+/fHgQMH8O233xq1/8c//oFDhw5h1qxZKCoqwtixYwXKvOdu3bqF119/HWvWrEFjY2O79adPn8aqVatQXFyMqqoqbN26Fbt27TLMA589ezYUCgWmTZuGmpqavk5fUC9FQbZWCoUCtbW1ICKjn8jISHz88cd9ns/MmTPx+PFjzJo1q8/3DQBNTU0dnnGJzfbt25GRkYGsrCyo1Wqjdbt374aNjQ0iIyPx+PFjgTLsvv3797cbj1evXjWs/+GHH/DJJ59g5cqVGD16dIcxHBwcEBkZCWdnZ6jVaixYsABhYWE4efIk7t27BwD48MMPMWrUKLz99tt4+vRpnxybGLw0BdlavqnzrJMnT7b7hb537x6uXr2K//qv/xIoK+Hs27cPFRUVQqfxXLdv38amTZvw2WefQaFQtFsfEhKC1atX48GDB1i7dq0AGfauUaNGITs7G4sWLYKdnV2HbY4fP97ureEuLi4AYHRGvXnzZhQUFGDXrl29l7DIvJAFmYiQlJSEIUOGwM7ODv369cO6deuM2rS2tiIuLg6+vr5QKpUYOXIkMjMzAQBpaWmwt7eHSqXC0aNHMWPGDGg0Gnh7e+PgwYOGGGfOnMFrr70GlUoFjUaDoKAg1NbWdhm/J7Zv344PP/ywx3HMdfbsWfj6+kIikeCvf/0rANP6affu3VAoFHBzc8OKFSvg6ekJhUKBkJAQnD9/HgAQFRUFuVwODw8Pw/7ef/992NvbQyKRoKqqCqtXr0Z0dDTu3LkDiUSCwMBAAL/80dJoNNiyZUsf90jHdu/eDSLC7NmzO22TmJiIwYMH46uvvsKpU6c6bUdESE5OxiuvvAI7Ozs4OTlhzpw5uH79OgDTx2lvjUVLevDgAZRKJfz9/Q3LnJycMHXqVOzatevluVxGIpeZmUnmprlhwwaSSCT0l7/8haqrq6mxsZFSU1MJAF2+fJmIiNauXUt2dnZ0+PBhqq6upvXr15ONjQ1dvHjREAMAff/99/T48WOqqKigKVOmkL29PbW0tFB9fT1pNBrasWMHNTU1UXl5Oc2dO5cqKytNit8d9+/fp2HDhlFra6vZ2wKgzMzMbu+biOjevXsEgPbs2WNY1lU/ERFFRkaSvb09Xbt2jZ48eUJFRUU0fvx4UqvVVFpaSkREixYtInd3d6P9JSUlEQBDn4aHh5NWqzVqc/z4cVKr1RQfH9+jYyMimjdvHs2bN69HMQICAmjYsGEdrtNqtXT37l0iIjp37hzZ2NiQn58f1dfXExFRbm4uhYaGGtrHxcWRXC6n/fv3U01NDV25coXGjh1LLi4uVF5eTkSm9b+lxmJCQgJ5e3uTo6MjyWQy8vPzo9DQULpw4UKH7SdMmECjRo3qMm5DQwOp1WqKiopqty42Ntbo97a7LDH++8ILV5AbGxtJpVLR73//e6PlBw8eNHywTU1NpFKpKCIiwmg7Ozs7+vOf/0xEvw70pqYmQ5u2on779m26evUqAaDjx4+3y8GU+N2xatUq+u///u9ubdvbBbmzfiL6pSD369fPKNbFixcJAH322WdE1P2CbEk9Lcj19fUkkUho1qxZHa5/tiATEUVHRxMAWrVqFREZF+TGxkZycHAwGkNERBcuXCAAhj9AXfW/JcdiaWkpXbp0ierq6qi5uZny8/NpzJgxpFQq6erVq+3am1qQN2zYQIMHD6ba2tp26/72t78RAPr73/9uVq6/ZS0F+YW7ZHH79m00NjZi2rRpnba5ceMGGhsbjabrKJVKeHh4GP4d7IhcLgcA6HQ6BAQEwM3NDYsXL8bmzZtRXFzc4/jPU1ZWhmPHjmHJkiXd2r4vPdtPnRk3bhxUKlW3+0OMKioqQERQqVQmtU9MTMSQIUOQmpqKs2fPGq0rKipCfX09xo0bZ7R8/PjxkMvlhss9HXm2/y05Fn18fDBmzBg4ODhALpdj4sSJSE9PR1NTE1JTU82K1ebIkSPIysrCd9991+5+CQBDXz58+LBb8a3NC1eQ79+/DwBwdXXttE1DQwMAYOPGjUbzKUtKSjqcptMRpVKJ06dPY/LkydiyZQsCAgIQERGBpqYmi8T/rR07dmD58uUd3iiyVnZ2dqisrBQ6DYtpe7dfZzezfkuhUCA9PR0SiQRLly5FU1OTYV3bdK+O5ps7Ojqirq7OpH30xlh8VlBQEGxtbXHz5k2zt83IyMD27duRl5cHPz+/DtsolUoAv/bti+6FK8htBau5ubnTNm3FOiUlpd0Unvz8fJP3NXz4cHzzzTcoKytDTEwMMjMzsXPnTovFb1NeXo7/+Z//wZ///GeztxUrnU6Hmpoa0b/BwRxtxcOcLzQEBwdjzZo1uHXrFhISEgzLHR0dAaDDwmtOv1l6LP6WXq+HXq83+Y9Qmz179uDAgQM4ffo0BgwY0Gm7lpYWAL/27YvuhSvII0aMgI2NDc6cOdNpGx8fHygUih59a6+srAzXrl0D8Mug37ZtG8aOHYtr165ZJP6zduzYgcWLF8PZ2dki8cQgLy8PRISJEycCAKRS6XMvcVgDNzc3SCQSs+cXJyQkYOjQobh8+bJh2YgRI+Dg4IB///vfRm3Pnz+PlpYWvPrqqybFtuRYfOutt9otu3jxIogIwcHBJsUgIsTExKCwsBA5OTldfuO0rS/d3d3NT9gKvXAF2dXVFfPmzcPhw4exb98+1NbW4sqVK9i7d6+hjUKhwHvvvYeDBw8iLS0NtbW1aG1txf379/HTTz+ZtJ+ysjKsWLEC169fR0tLCy5fvoySkhJMnDjRIvHbPHz4EH/729/w0UcfmbWd2Oj1elRXV+Pp06e4cuUKVq9eDV9fX8M18cDAQPz888/IycmBTqdDZWUlSkpKjGI4OzujrKwMxcXFqKurg06nQ25urmimvalUKgQEBBgum5mq7dLFs3NzFQoFoqOjceTIERw4cAC1tbUoLCzEypUr4enpicjISJNjdzUWIyIi4O7u3uXXtR88eICMjAzU1NRAp9MhPz8fy5Ytg6+vL1auXGlSPteuXcPnn3+OL7/8EjKZrN3XsHfu3GnUvq0vg4KCTIpv9fr+PqJ5ujPtra6ujpYvX079+/cnBwcHmjx5MsXFxREA8vb2ph9++IGam5spJiaGfH19SSqVkqurK4WHh1NRURGlpqaSSqUiADRo0CC6c+cO7d27lzQaDQGggQMH0j//+U8KCQkhJycnsrW1pQEDBtCGDRvo6dOnRETPjW+ONWvW0OLFi83apiPo4V3mPXv2kIeHBwEglUpFs2fPNqmfbt68SZGRkSSTycjLy4ukUilpNBqaM2cO3blzxxD/0aNH9Oabb5JCoSB/f3/64IMPaN26dQSAAgMDDXf4Bw4cSEqlkiZPnkzl5eV04sQJUqvVlJiY2OM+ssS0t6ioKJLJZNTY2GhYduTIEdJqtQSAXFxcDLMqfmvdunVG0970ej0lJSXRoEGDSCaTkZOTE4WFhdGNGzeIiEzu/67GYlhYGAGguLi45x5bN5UXgAAADoFJREFUdHQ0abVasre3J6lUSt7e3rR8+XIqKysztMnPz6dJkyaRp6cnASAA5OHhQSEhIXTmzBkqLCw0LO/oJykpyWifM2fOJC8vL9Lr9eZ9EL/R0/HfV17IgszaE3JARkZGkrOzsyD7NoclCvKtW7dIKpXS/v37LZRV72ttbaUpU6bQvn37hE7FSFVVFSkUCtq5c2ePY1lLQX7hLlkwcXpZntwVGBiI+Ph4xMfHG55eJmatra3IyclBXV0dIiIihE7HyObNmzF69GhERUUJnUqf4YLcx65fv97pIzWf/RHbLwczXWxsLObPn4+IiAjRP0AoLy8P2dnZyM3NNXn+dF9ITk5GQUEBTpw4AZlMJnQ6fYYLch8bOnRou+lHHf1kZGQInapFrF+/Hunp6Xj8+DH8/f1x+PBhoVPqE1u2bEFUVBS2bdsmdCrPNW3aNHz99ddGzxER2tGjR9Hc3Iy8vDw4OTkJnU6f4rdOs161detWbN26Veg0BDF9+nRMnz5d6DSsTmhoKEJDQ4VOQxB8hswYYyLBBZkxxkSCCzJjjIkEF2TGGBMJq7mp1/YCRNZ9KSkpOHTokNBpiNb//d//AeCxxoTDZ8iMMSYSVnOGzGd2PSORSPDRRx9hwYIFQqciWm1nxjzWXjzW8pJjPkNmjDGR4ILMGGMiwQWZMcZEggsyY4yJBBdkxhgTiZeqIGdnZyMgIKDdoy7lcjnc3NzwxhtvICkpCdXV1UKnyqzQqVOnEBsb226cvfPOO+3aTp8+HWq1Gra2thg+fHiXr08Skl6vR0pKCkJCQtqtS0xM7PDxsSNGjDBqd/bsWUyaNAkqlQqenp6IiYlp9yLi57U5duwYduzY8eI/V1uwR+ObqDfeGKLVaqlfv35E9Mtrcqqrq+l///d/acmSJSSRSMjT05MuXrxo0X0KDVbyxgQh9eSNIXFxcTRr1iyqra01LNNqtdS/f38CQMePH2+3TW5urtErm8To5s2bNGnSJAJAo0aNarc+ISGhw1cxDR8+3NDm6tWrpFQqadOmTVRfX0/nzp0jFxcXeu+998xqs2vXLpo6dSpVV1ebfRzWMv5f+oL8W4cOHSIbGxtyc3Ojmpoai+5XSEINyMbGRgoODraK2N0tyNu2baPBgwdTU1OT0XKtVktff/012djYkJeXV7vxJPaCXFBQQHPnzqUDBw7Q6NGjOy3IXb2uauHCheTv72/0XrykpCSSSCT0n//8x+Q2RL+8szA4OJh0Op1Zx2ItBfmlumRhinnz5mHJkiWoqKjAF198IXQ6Vm/fvn2oqKiwutimun37NjZt2oTPPvsMCoWi3fqQkBCsXr0aDx48wNq1awXIsPtGjRqF7OxsLFq0CHZ2dt2K8fTpU3z77beYOnWq0ZczZsyYASLC0aNHTWrTZvPmzSgoKMCuXbu6f2AixgW5A22vps/NzQXwy3vH4uLi4OvrC6VSiZEjRyIzMxMAkJaWBnt7e6hUKhw9ehQzZsyARqOBt7c3Dh48aIh55swZvPbaa1CpVNBoNAgKCkJtbW2X8YVCREhOTsYrr7wCOzs7ODk5Yc6cObh+/ToAICoqCnK53OhNE++//z7s7e0hkUhQVVWF1atXIzo6Gnfu3IFEIkFgYCB2794NhUIBNzc3rFixAp6enlAoFAgJCcH58+d7FBsATp48CY1Ggy1btvRJP+3evRtEhNmzZ3faJjExEYMHD8ZXX32FU6dOddquqz43dayJaTz9+OOPqK+vh6+vr9FyrVYLALhy5YpJbdo4OTlh6tSp2LVrF4iol7MXgJCn56bo60sWRES1tbUEgHx8fIiIaO3atWRnZ0eHDx+m6upqWr9+PdnY2BiuM2/YsIEA0Pfff0+PHz+miooKmjJlCtnb21NLSwvV19eTRqOhHTt2UFNTE5WXl9PcuXOpsrLSpPiWADP/ZYuLiyO5XE779++nmpoaunLlCo0dO5ZcXFyovLyciIgWLVpE7u7uRtslJSURAMOxhYeHk1arNWoTGRlJ9vb2dO3aNXry5AkVFRXR+PHjSa1WU2lpaY9iHz9+nNRqNcXHx5t8rG26c8kiICCAhg0b1uE6rVZLd+/eJSKic+fOkY2NDfn5+VF9fT0Rtb9kYUqfdzXWiHpnPE2YMKHTSxbe3t7k6OhIMpmM/Pz8KDQ0lC5cuEBERGfOnCEAlJSU1G5bpVJJ06ZNM6nNs2JjYwkAXb582eT8zR3/QuEz5A6o1WpIJBLU1dXhyZMnSEtLQ1hYGMLDw+Ho6IiNGzdCJpMhPT3daLuQkBBoNBq4uroiIiICDQ0NKC0tRXFxMWprazF8+HAoFAq4u7sjOzsbLi4uZsXvK01NTUhOTsbcuXOxePFi9OvXD0FBQfjiiy9QVVWFvXv39ngfUqnUcCY4bNgwpKWloa6ursfHPHPmTNTW1mLTpk09zrErDQ0NuHv3ruFM7nmCg4Px0Ucfobi4GJ988km79eb2eWdjra/H07vvvotjx47h3r17qK+vx8GDB1FaWoqpU6eiqKjIMEvC1ta23bYymQxNTU0mtXnWoEGDAACFhYWWPhzBcUHuQENDA4gIGo0GN27cQGNjo9E0HqVSCQ8PD8O/kh2Ry+UAAJ1Oh4CAALi5uWHx4sXYvHkziouLDe26G783FRUVob6+HuPGjTNaPn78eMjlcsOlBUsaN24cVCqVYMfcHRUVFSAik9/WnJiYiCFDhiA1NRVnz541WteTPn92rPX1ePLx8cGYMWPg4OAAuVyOiRMnIj09HU1NTUhNTTVcV3/69Gm7bVtaWqBUKk1q86y2/n748KGlD0dwXJA7cPPmTQC/vCG6oaEBALBx40ajeZYlJSVobGw0KZ5SqcTp06cxefJkbNmyBQEBAYiIiEBTU5NF4ltaTU0NAMDBwaHdOkdHR9TV1fXKfu3s7FBZWdkrsXvDkydPAMDkG14KhQLp6emQSCRYunSp0ZmfpfpcDOMpKCgItra2uHnzpuE+QNv9kjaNjY148uQJPD09TWrzrLYC3db/LxIuyB04efIkgF/u8rq6ugL45eHu9Ms0QcNPfn6+yTGHDx+Ob775BmVlZYiJiUFmZiZ27txpsfiW5OjoCAAdFoGamhp4e3tbfJ86na7XYveWtsJgzpcVgoODsWbNGty6dQsJCQmG5ZbqczGMJ71eD71eDzs7O/j7+0OtVqOkpMSoze3btwEAI0eONKnNs1paWgCg3Znzi4AL8m+Ul5cjJSUF3t7eWLp0KXx8fKBQKFBQUNDtmGVlZbh27RqAX35htm3bhrFjx+LatWsWiW9pI0aMgIODA/79738bLT9//jxaWlrw6quvAvjlOrBOp7PIPvPy8kBEmDhxosVj9xY3NzdIJBI8fvzYrO0SEhIwdOhQXL582bDM1D7vSl+Pp7feeqvdsosXL4KIEBwcDKlUirfffhv/+te/oNfrDW1yc3MhkUgwe/Zsk9o8q62/3d3de+mohPPSFmQiQn19PfR6PYgIlZWVyMzMxKRJk2Bra4ucnBxoNBooFAq89957OHjwINLS0lBbW4vW1lbcv38fP/30k0n7Kisrw4oVK3D9+nW0tLTg8uXLKCkpwcSJEy0S39IUCgWio6Nx5MgRHDhwALW1tSgsLMTKlSvh6emJyMhIAEBgYCB+/vln5OTkQKfTobKyst1ZjrOzM8rKylBcXIy6ujpDkdXr9aiursbTp09x5coVrF69Gr6+voYph92NnZub22fT3lQqFQICAnD//n2ztmu7dPHsTSxT+9yU2F2Np4iICLi7u1vk69oPHjxARkYGampqoNPpkJ+fj2XLlsHX1xcrV64EAGzatAkPHz7Ep59+ioaGBuTn5yMpKQlLlizBkCFDTG7Tpq2/g4KCepy/6PT1tA5zWXLa27Fjx2jkyJGkUqlILpeTjY0NASCJREKOjo702muvUXx8PD169Mhou+bmZoqJiSFfX1+SSqXk6upK4eHhVFRURKmpqaRSqQgADRo0iO7cuUN79+4ljUZDAGjgwIH0z3/+k0JCQsjJyYlsbW1pwIABtGHDBnr69GmX8S0FZk770ev1lJSURIMGDSKZTEZOTk4UFhZGN27cMLR59OgRvfnmm6RQKMjf358++OADWrduHQGgwMBAKi0tpUuXLtHAgQNJqVTS5MmTqby8nCIjI0kmk5GXlxdJpVLSaDQ0Z84cunPnTo9jnzhxgtRqNSUmJprdR92Z9hYVFUUymYwaGxsNy44cOUJarZYAkIuLC61atarDbdetW2c07a2rPjdlrN28ebPL8RQWFkYAKC4u7rnHlp+fT5MmTSJPT0/DV6I9PDwoJCSEzpw5Q0RE0dHRpNVqyd7enqRSKXl7e9Py5cuprKzMKNaZM2fotddeIzs7O/L09KR169bRkydPzG5DRDRz5kzy8vIy+lZfV8wd/0J5qQryy0xMAzIyMpKcnZ2FTqOd7hTkW7dukVQq7fLrw2LS2tpKU6ZMoX379gmditmqqqpIoVDQzp07zdpOTOP/eV7aSxZMWC/KU7sCAwMRHx+P+Ph41NfXC51Ol1pbW5GTk4O6ujpEREQInY7ZNm/ejNGjRyMqKkroVHoFF2TGeig2Nhbz589HRESE2Tf4+lpeXh6ys7ORm5tr8vxpsUhOTkZBQQFOnDgBmUwmdDq9ggsy61Pr169Heno6Hj9+DH9/fxw+fFjolCxiy5YtiIqKwrZt24RO5bmmTZuGr7/+2ug5Idbg6NGjaG5uRl5eHpycnIROp9dIhU6AvVy2bt2KrVu3Cp1Gr5g+fTqmT58udBovpNDQUISGhgqdRq/jM2TGGBMJLsiMMSYSXJAZY0wkuCAzxphIcEFmjDGRsJpZFs++a4t1z8KFC7Fw4UKh0xA9HmtMKBIicb+Y6v79+zh37pzQaTDGrFxISIjoH+8q+oLMGGMvC76GzBhjIsEFmTHGRIILMmOMiYQUwCGhk2CMMQb8P45Jg98g2E05AAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<IPython.core.display.Image object>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Visualizing LSTM_v1...\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<IPython.core.display.Image object>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Visualizing LSTM_v2...\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<IPython.core.display.Image object>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Visualizing GRU_v1...\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<IPython.core.display.Image object>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Visualizing GRU_v2...\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<IPython.core.display.Image object>"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "#Train Models With Updated Label Dataframe"
      ],
      "metadata": {
        "id": "7eaE_e6IMFzC"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "\n",
        "# Define the path to the pickle file\n",
        "pickle_path = \"/content/drive/My Drive/DATASET/updated00_labels_df.pkl\"\n",
        "\n",
        "# load labels_df from the pickle file\n",
        "labels_df = pd.read_pickle(pickle_path)\n",
        "print(\"Loaded labels_df from pickle.\")\n",
        "\n",
        "\n",
        "# Ensure train_embeds has the shape (num_samples, sequence_length, embedding_dim)\n",
        "# Adding a new axis to have shape (142246, 1024, 1)\n",
        "train_embeds = np.expand_dims(train_embeds, axis=-1)\n",
        "\n",
        "# Proceed with splitting the datasets\n",
        "# First, split the data into a training set and a temporary testing set\n",
        "X_train, X_temp, y_train, y_temp = train_test_split(\n",
        "    train_df, labels_df, test_size=0.2, random_state=42\n",
        ")\n",
        "\n",
        "# Next, split the temporary set into validation and test sets\n",
        "X_val, X_test, y_val, y_test = train_test_split(\n",
        "    X_temp, y_temp, test_size=0.5, random_state=42\n",
        ")\n",
        "\n",
        "# Display the dimensions of the sets to verify proper splitting\n",
        "print(f\"Training set dimensions: {X_train.shape}, {y_train.shape}\")\n",
        "print(f\"Validation set dimensions: {X_val.shape}, {y_val.shape}\")\n",
        "print(f\"Test set dimensions: {X_test.shape}, {y_test.shape}\")\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "pQmssSXJKjJn",
        "outputId": "bc0b98ce-2090-48bb-ab2f-1fb1b87df58c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Loaded labels_df from pickle.\n",
            "Training set dimensions: (113796, 1024), (113796, 1500)\n",
            "Validation set dimensions: (14225, 1024), (14225, 1500)\n",
            "Test set dimensions: (14225, 1024), (14225, 1500)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        " #Data configuration\n",
        "INPUT_SHAPE = [X_train.shape[1]]\n",
        "BATCH_SIZE = 512\n",
        "num_of_labels = y_train.shape[1]\n",
        "\n",
        "# Define the model architectures\n",
        "def build_rnn_model(input_shape, num_of_labels):\n",
        "    model = Sequential([\n",
        "        layers.Reshape((1, input_shape), input_shape=(input_shape,)),\n",
        "        layers.SimpleRNN(128, return_sequences=True),\n",
        "        layers.Dropout(0.3),\n",
        "        layers.SimpleRNN(128),\n",
        "        layers.Dropout(0.3),\n",
        "        layers.Dense(512, activation='relu'),\n",
        "        layers.Dropout(0.4),\n",
        "        layers.Dense(num_of_labels, activation='sigmoid')\n",
        "    ])\n",
        "    model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['binary_accuracy', tf.keras.metrics.AUC()])\n",
        "    return model\n",
        "\n",
        "def build_lstm_model(input_shape, num_of_labels):\n",
        "    model = Sequential([\n",
        "        layers.Reshape((1, input_shape), input_shape=(input_shape,)),\n",
        "        layers.LSTM(128, return_sequences=True),\n",
        "        layers.Dropout(0.3),\n",
        "        layers.LSTM(128),\n",
        "        layers.Dropout(0.3),\n",
        "        layers.Dense(512, activation='relu'),\n",
        "        layers.Dropout(0.4),\n",
        "        layers.Dense(num_of_labels, activation='sigmoid')\n",
        "    ])\n",
        "    model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['binary_accuracy', tf.keras.metrics.AUC()])\n",
        "    return model\n",
        "\n",
        "def build_gru_model(input_shape, num_of_labels):\n",
        "    model = Sequential([\n",
        "        layers.Reshape((1, input_shape), input_shape=(input_shape,)),\n",
        "        layers.GRU(128, return_sequences=True),\n",
        "        layers.Dropout(0.3),\n",
        "        layers.GRU(128),\n",
        "        layers.Dropout(0.3),\n",
        "        layers.Dense(512, activation='relu'),\n",
        "        layers.Dropout(0.4),\n",
        "        layers.Dense(num_of_labels, activation='sigmoid')\n",
        "    ])\n",
        "    model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['binary_accuracy', tf.keras.metrics.AUC()])\n",
        "    return model\n",
        "\n",
        "# Initialize models\n",
        "model_rnn_simple = build_rnn_model(X_train.shape[1], num_of_labels)\n",
        "model_lstm = build_lstm_model(X_train.shape[1], num_of_labels)\n",
        "model_gru = build_gru_model(X_train.shape[1], num_of_labels)\n",
        "\n",
        "# Setup ModelCheckpoint callbacks\n",
        "checkpoint_rnn = ModelCheckpoint('/content/drive/My Drive/DATASET/model_rnn_simple_best0.h5', monitor='val_loss', save_best_only=True, verbose=1)\n",
        "checkpoint_lstm = ModelCheckpoint('/content/drive/My Drive/DATASET/model_lstm_best0.h5', monitor='val_loss', save_best_only=True, verbose=1)\n",
        "checkpoint_gru = ModelCheckpoint('/content/drive/My Drive/DATASET/model_gru_best0.h5', monitor='val_loss', save_best_only=True, verbose=1)\n",
        "\n",
        "\n",
        "history_gru = model_gru.fit(\n",
        "    X_train, y_train,\n",
        "    validation_data=(X_val, y_val),\n",
        "    batch_size=BATCH_SIZE,\n",
        "    epochs=50,\n",
        "    callbacks=[checkpoint_gru]\n",
        ")\n",
        "\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "nUAFBlzNK2b1",
        "outputId": "a6778246-4cf6-4e1d-a657-ae3eb76daaec"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/50\n",
            "23/23 [==============================] - ETA: 0s - loss: 0.6187 - binary_accuracy: 0.7421 - auc_5: 0.7365\n",
            "Epoch 1: val_loss improved from inf to 0.47864, saving model to /content/drive/My Drive/DATASET/model_gru_best0.h5\n",
            "23/23 [==============================] - 48s 2s/step - loss: 0.6187 - binary_accuracy: 0.7421 - auc_5: 0.7365 - val_loss: 0.4786 - val_binary_accuracy: 0.7908 - val_auc_5: 0.8209\n",
            "Epoch 2/50\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3103: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "23/23 [==============================] - ETA: 0s - loss: 0.4526 - binary_accuracy: 0.7851 - auc_5: 0.8236\n",
            "Epoch 2: val_loss improved from 0.47864 to 0.44035, saving model to /content/drive/My Drive/DATASET/model_gru_best0.h5\n",
            "23/23 [==============================] - 44s 2s/step - loss: 0.4526 - binary_accuracy: 0.7851 - auc_5: 0.8236 - val_loss: 0.4403 - val_binary_accuracy: 0.7908 - val_auc_5: 0.8320\n",
            "Epoch 3/50\n",
            "23/23 [==============================] - ETA: 0s - loss: 0.4413 - binary_accuracy: 0.7893 - auc_5: 0.8312\n",
            "Epoch 3: val_loss improved from 0.44035 to 0.43711, saving model to /content/drive/My Drive/DATASET/model_gru_best0.h5\n",
            "23/23 [==============================] - 47s 2s/step - loss: 0.4413 - binary_accuracy: 0.7893 - auc_5: 0.8312 - val_loss: 0.4371 - val_binary_accuracy: 0.7908 - val_auc_5: 0.8325\n",
            "Epoch 4/50\n",
            "23/23 [==============================] - ETA: 0s - loss: 0.4394 - binary_accuracy: 0.7897 - auc_5: 0.8325\n",
            "Epoch 4: val_loss improved from 0.43711 to 0.43629, saving model to /content/drive/My Drive/DATASET/model_gru_best0.h5\n",
            "23/23 [==============================] - 47s 2s/step - loss: 0.4394 - binary_accuracy: 0.7897 - auc_5: 0.8325 - val_loss: 0.4363 - val_binary_accuracy: 0.7908 - val_auc_5: 0.8330\n",
            "Epoch 5/50\n",
            "23/23 [==============================] - ETA: 0s - loss: 0.4384 - binary_accuracy: 0.7899 - auc_5: 0.8334\n",
            "Epoch 5: val_loss improved from 0.43629 to 0.43555, saving model to /content/drive/My Drive/DATASET/model_gru_best0.h5\n",
            "23/23 [==============================] - 46s 2s/step - loss: 0.4384 - binary_accuracy: 0.7899 - auc_5: 0.8334 - val_loss: 0.4355 - val_binary_accuracy: 0.7908 - val_auc_5: 0.8333\n",
            "Epoch 6/50\n",
            "23/23 [==============================] - ETA: 0s - loss: 0.4369 - binary_accuracy: 0.7902 - auc_5: 0.8348\n",
            "Epoch 6: val_loss improved from 0.43555 to 0.43330, saving model to /content/drive/My Drive/DATASET/model_gru_best0.h5\n",
            "23/23 [==============================] - 45s 2s/step - loss: 0.4369 - binary_accuracy: 0.7902 - auc_5: 0.8348 - val_loss: 0.4333 - val_binary_accuracy: 0.7908 - val_auc_5: 0.8358\n",
            "Epoch 7/50\n",
            "23/23 [==============================] - ETA: 0s - loss: 0.4307 - binary_accuracy: 0.7911 - auc_5: 0.8422\n",
            "Epoch 7: val_loss improved from 0.43330 to 0.42137, saving model to /content/drive/My Drive/DATASET/model_gru_best0.h5\n",
            "23/23 [==============================] - 45s 2s/step - loss: 0.4307 - binary_accuracy: 0.7911 - auc_5: 0.8422 - val_loss: 0.4214 - val_binary_accuracy: 0.7953 - val_auc_5: 0.8508\n",
            "Epoch 8/50\n",
            "23/23 [==============================] - ETA: 0s - loss: 0.4217 - binary_accuracy: 0.7940 - auc_5: 0.8508\n",
            "Epoch 8: val_loss improved from 0.42137 to 0.41771, saving model to /content/drive/My Drive/DATASET/model_gru_best0.h5\n",
            "23/23 [==============================] - 47s 2s/step - loss: 0.4217 - binary_accuracy: 0.7940 - auc_5: 0.8508 - val_loss: 0.4177 - val_binary_accuracy: 0.7985 - val_auc_5: 0.8539\n",
            "Epoch 9/50\n",
            "23/23 [==============================] - ETA: 0s - loss: 0.4183 - binary_accuracy: 0.7963 - auc_5: 0.8536\n",
            "Epoch 9: val_loss improved from 0.41771 to 0.41593, saving model to /content/drive/My Drive/DATASET/model_gru_best0.h5\n",
            "23/23 [==============================] - 45s 2s/step - loss: 0.4183 - binary_accuracy: 0.7963 - auc_5: 0.8536 - val_loss: 0.4159 - val_binary_accuracy: 0.7994 - val_auc_5: 0.8552\n",
            "Epoch 10/50\n",
            "23/23 [==============================] - ETA: 0s - loss: 0.4164 - binary_accuracy: 0.7983 - auc_5: 0.8551\n",
            "Epoch 10: val_loss improved from 0.41593 to 0.41467, saving model to /content/drive/My Drive/DATASET/model_gru_best0.h5\n",
            "23/23 [==============================] - 48s 2s/step - loss: 0.4164 - binary_accuracy: 0.7983 - auc_5: 0.8551 - val_loss: 0.4147 - val_binary_accuracy: 0.8000 - val_auc_5: 0.8563\n",
            "Epoch 11/50\n",
            "23/23 [==============================] - ETA: 0s - loss: 0.4146 - binary_accuracy: 0.8001 - auc_5: 0.8566\n",
            "Epoch 11: val_loss improved from 0.41467 to 0.41297, saving model to /content/drive/My Drive/DATASET/model_gru_best0.h5\n",
            "23/23 [==============================] - 45s 2s/step - loss: 0.4146 - binary_accuracy: 0.8001 - auc_5: 0.8566 - val_loss: 0.4130 - val_binary_accuracy: 0.8008 - val_auc_5: 0.8578\n",
            "Epoch 12/50\n",
            "23/23 [==============================] - ETA: 0s - loss: 0.4121 - binary_accuracy: 0.8015 - auc_5: 0.8588\n",
            "Epoch 12: val_loss improved from 0.41297 to 0.40885, saving model to /content/drive/My Drive/DATASET/model_gru_best0.h5\n",
            "23/23 [==============================] - 45s 2s/step - loss: 0.4121 - binary_accuracy: 0.8015 - auc_5: 0.8588 - val_loss: 0.4088 - val_binary_accuracy: 0.8015 - val_auc_5: 0.8612\n",
            "Epoch 13/50\n",
            "23/23 [==============================] - ETA: 0s - loss: 0.4083 - binary_accuracy: 0.8024 - auc_5: 0.8619\n",
            "Epoch 13: val_loss improved from 0.40885 to 0.40547, saving model to /content/drive/My Drive/DATASET/model_gru_best0.h5\n",
            "23/23 [==============================] - 48s 2s/step - loss: 0.4083 - binary_accuracy: 0.8024 - auc_5: 0.8619 - val_loss: 0.4055 - val_binary_accuracy: 0.8029 - val_auc_5: 0.8639\n",
            "Epoch 14/50\n",
            "23/23 [==============================] - ETA: 0s - loss: 0.4053 - binary_accuracy: 0.8041 - auc_5: 0.8643\n",
            "Epoch 14: val_loss improved from 0.40547 to 0.40246, saving model to /content/drive/My Drive/DATASET/model_gru_best0.h5\n",
            "23/23 [==============================] - 45s 2s/step - loss: 0.4053 - binary_accuracy: 0.8041 - auc_5: 0.8643 - val_loss: 0.4025 - val_binary_accuracy: 0.8050 - val_auc_5: 0.8663\n",
            "Epoch 15/50\n",
            "23/23 [==============================] - ETA: 0s - loss: 0.4014 - binary_accuracy: 0.8066 - auc_5: 0.8675\n",
            "Epoch 15: val_loss improved from 0.40246 to 0.39830, saving model to /content/drive/My Drive/DATASET/model_gru_best0.h5\n",
            "23/23 [==============================] - 45s 2s/step - loss: 0.4014 - binary_accuracy: 0.8066 - auc_5: 0.8675 - val_loss: 0.3983 - val_binary_accuracy: 0.8080 - val_auc_5: 0.8695\n",
            "Epoch 16/50\n",
            "23/23 [==============================] - ETA: 0s - loss: 0.3986 - binary_accuracy: 0.8084 - auc_5: 0.8697\n",
            "Epoch 16: val_loss improved from 0.39830 to 0.39624, saving model to /content/drive/My Drive/DATASET/model_gru_best0.h5\n",
            "23/23 [==============================] - 47s 2s/step - loss: 0.3986 - binary_accuracy: 0.8084 - auc_5: 0.8697 - val_loss: 0.3962 - val_binary_accuracy: 0.8092 - val_auc_5: 0.8711\n",
            "Epoch 17/50\n",
            "23/23 [==============================] - ETA: 0s - loss: 0.3967 - binary_accuracy: 0.8097 - auc_5: 0.8712\n",
            "Epoch 17: val_loss improved from 0.39624 to 0.39519, saving model to /content/drive/My Drive/DATASET/model_gru_best0.h5\n",
            "23/23 [==============================] - 45s 2s/step - loss: 0.3967 - binary_accuracy: 0.8097 - auc_5: 0.8712 - val_loss: 0.3952 - val_binary_accuracy: 0.8097 - val_auc_5: 0.8721\n",
            "Epoch 18/50\n",
            "23/23 [==============================] - ETA: 0s - loss: 0.3952 - binary_accuracy: 0.8106 - auc_5: 0.8723\n",
            "Epoch 18: val_loss improved from 0.39519 to 0.39361, saving model to /content/drive/My Drive/DATASET/model_gru_best0.h5\n",
            "23/23 [==============================] - 48s 2s/step - loss: 0.3952 - binary_accuracy: 0.8106 - auc_5: 0.8723 - val_loss: 0.3936 - val_binary_accuracy: 0.8110 - val_auc_5: 0.8731\n",
            "Epoch 19/50\n",
            "23/23 [==============================] - ETA: 0s - loss: 0.3936 - binary_accuracy: 0.8115 - auc_5: 0.8734\n",
            "Epoch 19: val_loss improved from 0.39361 to 0.39196, saving model to /content/drive/My Drive/DATASET/model_gru_best0.h5\n",
            "23/23 [==============================] - 47s 2s/step - loss: 0.3936 - binary_accuracy: 0.8115 - auc_5: 0.8734 - val_loss: 0.3920 - val_binary_accuracy: 0.8117 - val_auc_5: 0.8744\n",
            "Epoch 20/50\n",
            "23/23 [==============================] - ETA: 0s - loss: 0.3919 - binary_accuracy: 0.8126 - auc_5: 0.8747\n",
            "Epoch 20: val_loss improved from 0.39196 to 0.38972, saving model to /content/drive/My Drive/DATASET/model_gru_best0.h5\n",
            "23/23 [==============================] - 47s 2s/step - loss: 0.3919 - binary_accuracy: 0.8126 - auc_5: 0.8747 - val_loss: 0.3897 - val_binary_accuracy: 0.8127 - val_auc_5: 0.8760\n",
            "Epoch 21/50\n",
            "23/23 [==============================] - ETA: 0s - loss: 0.3899 - binary_accuracy: 0.8134 - auc_5: 0.8762\n",
            "Epoch 21: val_loss improved from 0.38972 to 0.38845, saving model to /content/drive/My Drive/DATASET/model_gru_best0.h5\n",
            "23/23 [==============================] - 52s 2s/step - loss: 0.3899 - binary_accuracy: 0.8134 - auc_5: 0.8762 - val_loss: 0.3884 - val_binary_accuracy: 0.8132 - val_auc_5: 0.8770\n",
            "Epoch 22/50\n",
            "23/23 [==============================] - ETA: 0s - loss: 0.3886 - binary_accuracy: 0.8140 - auc_5: 0.8771\n",
            "Epoch 22: val_loss improved from 0.38845 to 0.38686, saving model to /content/drive/My Drive/DATASET/model_gru_best0.h5\n"
          ]
        },
        {
          "output_type": "error",
          "ename": "OSError",
          "evalue": "[Errno 5] Unable to create file (unable to open file: name = '/content/drive/My Drive/DATASET/model_gru_best0.h5', errno = 5, error message = 'Input/output error', flags = 13, o_flags = 242)",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mOSError\u001b[0m                                   Traceback (most recent call last)",
            "\u001b[0;32m<ipython-input-8-5c69272cbcf1>\u001b[0m in \u001b[0;36m<cell line: 76>\u001b[0;34m()\u001b[0m\n\u001b[1;32m     74\u001b[0m \u001b[0;31m#)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     75\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 76\u001b[0;31m history_gru = model_gru.fit(\n\u001b[0m\u001b[1;32m     77\u001b[0m    \u001b[0mX_train\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my_train\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     78\u001b[0m    \u001b[0mvalidation_data\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mX_val\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my_val\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.10/dist-packages/keras/src/utils/traceback_utils.py\u001b[0m in \u001b[0;36merror_handler\u001b[0;34m(*args, **kwargs)\u001b[0m\n\u001b[1;32m     68\u001b[0m             \u001b[0;31m# To get the full stack trace, call:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     69\u001b[0m             \u001b[0;31m# `tf.debugging.disable_traceback_filtering()`\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 70\u001b[0;31m             \u001b[0;32mraise\u001b[0m \u001b[0me\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mwith_traceback\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfiltered_tb\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32mfrom\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     71\u001b[0m         \u001b[0;32mfinally\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     72\u001b[0m             \u001b[0;32mdel\u001b[0m \u001b[0mfiltered_tb\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.10/dist-packages/h5py/_hl/files.py\u001b[0m in \u001b[0;36m__init__\u001b[0;34m(self, name, mode, driver, libver, userblock_size, swmr, rdcc_nslots, rdcc_nbytes, rdcc_w0, track_order, fs_strategy, fs_persist, fs_threshold, fs_page_size, page_buf_size, min_meta_keep, min_raw_keep, locking, alignment_threshold, alignment_interval, meta_block_size, **kwds)\u001b[0m\n\u001b[1;32m    565\u001b[0m                                  \u001b[0mfs_persist\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mfs_persist\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mfs_threshold\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mfs_threshold\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    566\u001b[0m                                  fs_page_size=fs_page_size)\n\u001b[0;32m--> 567\u001b[0;31m                 \u001b[0mfid\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mmake_fid\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mname\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmode\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0muserblock_size\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mfapl\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mfcpl\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mswmr\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mswmr\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    568\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    569\u001b[0m             \u001b[0;32mif\u001b[0m \u001b[0misinstance\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mlibver\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtuple\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.10/dist-packages/h5py/_hl/files.py\u001b[0m in \u001b[0;36mmake_fid\u001b[0;34m(name, mode, userblock_size, fapl, fcpl, swmr)\u001b[0m\n\u001b[1;32m    235\u001b[0m         \u001b[0mfid\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mh5f\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcreate\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mname\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mh5f\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mACC_EXCL\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mfapl\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mfapl\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mfcpl\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mfcpl\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    236\u001b[0m     \u001b[0;32melif\u001b[0m \u001b[0mmode\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0;34m'w'\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 237\u001b[0;31m         \u001b[0mfid\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mh5f\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcreate\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mname\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mh5f\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mACC_TRUNC\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mfapl\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mfapl\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mfcpl\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mfcpl\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    238\u001b[0m     \u001b[0;32melif\u001b[0m \u001b[0mmode\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0;34m'a'\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    239\u001b[0m         \u001b[0;31m# Open in append mode (read/write).\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32mh5py/_objects.pyx\u001b[0m in \u001b[0;36mh5py._objects.with_phil.wrapper\u001b[0;34m()\u001b[0m\n",
            "\u001b[0;32mh5py/_objects.pyx\u001b[0m in \u001b[0;36mh5py._objects.with_phil.wrapper\u001b[0;34m()\u001b[0m\n",
            "\u001b[0;32mh5py/h5f.pyx\u001b[0m in \u001b[0;36mh5py.h5f.create\u001b[0;34m()\u001b[0m\n",
            "\u001b[0;31mOSError\u001b[0m: [Errno 5] Unable to create file (unable to open file: name = '/content/drive/My Drive/DATASET/model_gru_best0.h5', errno = 5, error message = 'Input/output error', flags = 13, o_flags = 242)"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "ga0CE10vLknb"
      },
      "execution_count": null,
      "outputs": []
    }
  ],
  "metadata": {
    "colab": {
      "provenance": [],
      "gpuType": "T4",
      "machine_shape": "hm"
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU"
  },
  "nbformat": 4,
  "nbformat_minor": 0
}