notebooks/vietnamese-newspaper-category-classification.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Setup Working directory"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "\n",
    "if os.path.basename(os.getcwd()) == \"notebooks\":\n",
    "    os.chdir(\"../\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "execution": {
     "iopub.execute_input": "2025-01-15T04:58:33.678481Z",
     "iopub.status.busy": "2025-01-15T04:58:33.678008Z",
     "iopub.status.idle": "2025-01-15T04:58:41.466368Z",
     "shell.execute_reply": "2025-01-15T04:58:41.465292Z",
     "shell.execute_reply.started": "2025-01-15T04:58:33.678451Z"
    },
    "id": "gxGgev7JY2SS",
    "outputId": "9678930f-c08d-4ca1-c6fc-c35ad9a07c18",
    "trusted": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m20.9/20.9 MB\u001b[0m \u001b[31m60.6 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m:00:01\u001b[0m00:01\u001b[0m\n",
      "\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m657.8/657.8 kB\u001b[0m \u001b[31m24.7 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
      "\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m1.2/1.2 MB\u001b[0m \u001b[31m40.6 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
      "\u001b[?25h"
     ]
    }
   ],
   "source": [
    "!pip install -q underthesea\n",
    "# !pip install -U sentence-transformers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "execution": {
     "iopub.execute_input": "2025-01-15T00:30:12.051983Z",
     "iopub.status.busy": "2025-01-15T00:30:12.051766Z",
     "iopub.status.idle": "2025-01-15T00:30:13.378762Z",
     "shell.execute_reply": "2025-01-15T00:30:13.377690Z",
     "shell.execute_reply.started": "2025-01-15T00:30:12.051965Z"
    },
    "id": "_Nr5BsSODkr3",
    "outputId": "d76656a7-08a7-471b-c5c3-4d2bbfbb447a",
    "trusted": true
   },
   "outputs": [],
   "source": [
    "# !cp -r ../input/d/avaholmes/vnnews-classification/ABC/* ."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-01-15T05:00:58.145752Z",
     "iopub.status.busy": "2025-01-15T05:00:58.144596Z",
     "iopub.status.idle": "2025-01-15T05:01:04.754188Z",
     "shell.execute_reply": "2025-01-15T05:01:04.752426Z",
     "shell.execute_reply.started": "2025-01-15T05:00:58.145690Z"
    },
    "trusted": true
   },
   "outputs": [],
   "source": [
    "# !cp -r ../input/vietnamese-newspaper-category-classification-1/* ."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-01-15T05:10:40.474022Z",
     "iopub.status.busy": "2025-01-15T05:10:40.473570Z",
     "iopub.status.idle": "2025-01-15T05:10:40.484821Z",
     "shell.execute_reply": "2025-01-15T05:10:40.483376Z",
     "shell.execute_reply.started": "2025-01-15T05:10:40.473988Z"
    },
    "id": "FS-BSrmy4qQx",
    "trusted": true
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import time\n",
    "import os\n",
    "import numpy as np\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.preprocessing import OneHotEncoder\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "from underthesea import word_tokenize\n",
    "\n",
    "import pickle\n",
    "import json\n",
    "import tensorflow as tf\n",
    "from tensorflow.keras.preprocessing.text import Tokenizer\n",
    "from tensorflow.keras.preprocessing.sequence import pad_sequences\n",
    "\n",
    "import tensorflow as tf\n",
    "from tensorflow.keras.models import Model\n",
    "from tensorflow.keras.layers import Input, Bidirectional, LSTM, Dense, concatenate, Dropout, BatchNormalization, GRU, RNN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 423
    },
    "execution": {
     "iopub.execute_input": "2025-01-15T05:07:06.328988Z",
     "iopub.status.busy": "2025-01-15T05:07:06.328465Z",
     "iopub.status.idle": "2025-01-15T05:07:09.255551Z",
     "shell.execute_reply": "2025-01-15T05:07:09.254486Z",
     "shell.execute_reply.started": "2025-01-15T05:07:06.328950Z"
    },
    "id": "TjYWqSqC0alG",
    "outputId": "dd3817a7-050e-4146-db16-b69c2dd27622",
    "trusted": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>content</th>\n",
       "      <th>label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>&lt;sos&gt; Quán cuối tuần : Nhà_hàng Edo Nhà_hàng n...</td>\n",
       "      <td>Doi song</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>&lt;sos&gt; Kết_quả nghiên_cứu của một nhóm nghiên_c...</td>\n",
       "      <td>Doi song</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>&lt;sos&gt; Khăn choàng Bạn được tặng nhiều khăn_quà...</td>\n",
       "      <td>Doi song</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>&lt;sos&gt; Chuyện của anh Kha rất giống gia_đình tô...</td>\n",
       "      <td>Doi song</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>&lt;sos&gt; Cá hấp_Món cá_hấp có vị béo_ngọt của thị...</td>\n",
       "      <td>Doi song</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26054</th>\n",
       "      <td>&lt;sos&gt; Đạn súng_hơi gây thương_tích cho một em ...</td>\n",
       "      <td>Suc khoe</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26055</th>\n",
       "      <td>&lt;sos&gt; Không_khí sạch giúp xua_đuổi virus gây c...</td>\n",
       "      <td>Suc khoe</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26056</th>\n",
       "      <td>&lt;sos&gt; Mất tiếng do ho nhiều Một_số người bị ho...</td>\n",
       "      <td>Suc khoe</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26057</th>\n",
       "      <td>&lt;sos&gt; Bệnh_nhân Hong_Kong không chết vì SARS h...</td>\n",
       "      <td>Suc khoe</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26058</th>\n",
       "      <td>&lt;sos&gt; Uống nước như_thế_nào khi tập_luyện ? Kh...</td>\n",
       "      <td>Suc khoe</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>26059 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                 content     label\n",
       "0      <sos> Quán cuối tuần : Nhà_hàng Edo Nhà_hàng n...  Doi song\n",
       "1      <sos> Kết_quả nghiên_cứu của một nhóm nghiên_c...  Doi song\n",
       "2      <sos> Khăn choàng Bạn được tặng nhiều khăn_quà...  Doi song\n",
       "3      <sos> Chuyện của anh Kha rất giống gia_đình tô...  Doi song\n",
       "4      <sos> Cá hấp_Món cá_hấp có vị béo_ngọt của thị...  Doi song\n",
       "...                                                  ...       ...\n",
       "26054  <sos> Đạn súng_hơi gây thương_tích cho một em ...  Suc khoe\n",
       "26055  <sos> Không_khí sạch giúp xua_đuổi virus gây c...  Suc khoe\n",
       "26056  <sos> Mất tiếng do ho nhiều Một_số người bị ho...  Suc khoe\n",
       "26057  <sos> Bệnh_nhân Hong_Kong không chết vì SARS h...  Suc khoe\n",
       "26058  <sos> Uống nước như_thế_nào khi tập_luyện ? Kh...  Suc khoe\n",
       "\n",
       "[26059 rows x 2 columns]"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Read Convert CSV\n",
    "os.makedirs(\"dataset\", exist_ok=True)\n",
    "converted_df = pd.read_csv(\"dataset/converted_train_dataset.csv\")\n",
    "converted_df['content'] = \"<sos> \" + converted_df['content'] + \" <eos>\"\n",
    "converted_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-01-15T05:07:09.257438Z",
     "iopub.status.busy": "2025-01-15T05:07:09.257028Z",
     "iopub.status.idle": "2025-01-15T05:07:09.531306Z",
     "shell.execute_reply": "2025-01-15T05:07:09.530223Z",
     "shell.execute_reply.started": "2025-01-15T05:07:09.257400Z"
    },
    "trusted": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>content</th>\n",
       "      <th>label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>&lt;sos&gt; Những phát_hiện thú_vị về đàn_ông Đàn_ôn...</td>\n",
       "      <td>Doi song</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>&lt;sos&gt; Kỹ_thuật trồng sung Mỹ trong chậu_Anh Ng...</td>\n",
       "      <td>Doi song</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>&lt;sos&gt; Khổ vì bạn đời quá sạch Lau_chùi ba lượt...</td>\n",
       "      <td>Doi song</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>&lt;sos&gt; Chả lá lốt - món dân dã_vạn người mê Khô...</td>\n",
       "      <td>Doi song</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>&lt;sos&gt; Mực nhồi thịt sốt cà_chua Công_thức và c...</td>\n",
       "      <td>Doi song</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2395</th>\n",
       "      <td>&lt;sos&gt; Xu_hướng tình_dục ở trẻ ' đáng báo_động ...</td>\n",
       "      <td>Suc khoe</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2396</th>\n",
       "      <td>&lt;sos&gt; Bệnh gan bí_ẩn tàn_phá sức khỏe bệnh_nhi...</td>\n",
       "      <td>Suc khoe</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2397</th>\n",
       "      <td>&lt;sos&gt; Điều cần tránh khi uống trà xanh Theo bá...</td>\n",
       "      <td>Suc khoe</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2398</th>\n",
       "      <td>&lt;sos&gt; Sốt_xuất_huyết bùng_phát ở phía Nam Hệ_t...</td>\n",
       "      <td>Suc khoe</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2399</th>\n",
       "      <td>&lt;sos&gt; Tổn_thương đa_tạng do đồng_nhiễm xoắn_kh...</td>\n",
       "      <td>Suc khoe</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2400 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                content     label\n",
       "0     <sos> Những phát_hiện thú_vị về đàn_ông Đàn_ôn...  Doi song\n",
       "1     <sos> Kỹ_thuật trồng sung Mỹ trong chậu_Anh Ng...  Doi song\n",
       "2     <sos> Khổ vì bạn đời quá sạch Lau_chùi ba lượt...  Doi song\n",
       "3     <sos> Chả lá lốt - món dân dã_vạn người mê Khô...  Doi song\n",
       "4     <sos> Mực nhồi thịt sốt cà_chua Công_thức và c...  Doi song\n",
       "...                                                 ...       ...\n",
       "2395  <sos> Xu_hướng tình_dục ở trẻ ' đáng báo_động ...  Suc khoe\n",
       "2396  <sos> Bệnh gan bí_ẩn tàn_phá sức khỏe bệnh_nhi...  Suc khoe\n",
       "2397  <sos> Điều cần tránh khi uống trà xanh Theo bá...  Suc khoe\n",
       "2398  <sos> Sốt_xuất_huyết bùng_phát ở phía Nam Hệ_t...  Suc khoe\n",
       "2399  <sos> Tổn_thương đa_tạng do đồng_nhiễm xoắn_kh...  Suc khoe\n",
       "\n",
       "[2400 rows x 2 columns]"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "converted_test_df = pd.read_csv(\"dataset/converted_test_dataset.csv\")\n",
    "converted_test_df['content'] = \"<sos> \" + converted_test_df['content'] + \" <eos>\"\n",
    "converted_test_df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "jZwlKqA51fXX"
   },
   "source": [
    "# Tokenizer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-01-12T14:06:22.418578Z",
     "iopub.status.busy": "2025-01-12T14:06:22.418212Z",
     "iopub.status.idle": "2025-01-12T14:06:22.422976Z",
     "shell.execute_reply": "2025-01-12T14:06:22.421794Z",
     "shell.execute_reply.started": "2025-01-12T14:06:22.418550Z"
    },
    "id": "qFQpaYlOrh7R",
    "trusted": true
   },
   "outputs": [],
   "source": [
    "# tokenizer = Tokenizer()\n",
    "tokenizer = Tokenizer(filters='!\"#$%&()*+,-./:;=?@[\\\\]^`{|}~\\t\\n')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "z-ba_CD5uk-g",
    "trusted": true
   },
   "outputs": [],
   "source": [
    "tokenizer.fit_on_texts( list(converted_df['content']))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "zB1LAwR6wDl5",
    "trusted": true
   },
   "outputs": [],
   "source": [
    "# Save Tokenize to Json files\n",
    "os.makedirs('tokenizers', exist_ok=True)\n",
    "\n",
    "tokenizer_json = tokenizer.to_json()\n",
    "\n",
    "\n",
    "with open('tokenizers/tokenizer.pkl', 'wb') as f:\n",
    "    pickle.dump(tokenizer_json, f)\n",
    "\n",
    "# with open('tokenizers/s_tokenizer.pkl', 'wb') as f:\n",
    "#     pickle.dump(s_tokenizer_json, f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-01-15T03:27:48.641289Z",
     "iopub.status.busy": "2025-01-15T03:27:48.640978Z",
     "iopub.status.idle": "2025-01-15T03:27:49.448633Z",
     "shell.execute_reply": "2025-01-15T03:27:49.447913Z",
     "shell.execute_reply.started": "2025-01-15T03:27:48.641259Z"
    },
    "id": "hemz7rOrwGCH",
    "trusted": true
   },
   "outputs": [],
   "source": [
    "# Load Exists Tokenizers\n",
    "\n",
    "tokenizer_file = open('tokenizers/tokenizer.pkl', 'rb')\n",
    "tokenizer = pickle.load(tokenizer_file)\n",
    "\n",
    "tokenizer = tf.keras.preprocessing.text.tokenizer_from_json(tokenizer)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "execution": {
     "iopub.execute_input": "2025-01-15T00:45:21.272044Z",
     "iopub.status.busy": "2025-01-15T00:45:21.271747Z",
     "iopub.status.idle": "2025-01-15T00:45:21.277131Z",
     "shell.execute_reply": "2025-01-15T00:45:21.276389Z",
     "shell.execute_reply.started": "2025-01-15T00:45:21.272020Z"
    },
    "id": "1b8ZREHXvxcn",
    "outputId": "cdeb66b4-d2fc-4a1f-87c2-fb030798a6d7",
    "trusted": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "188876"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(tokenizer.word_index) # 188876"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-01-15T03:27:49.588234Z",
     "iopub.status.busy": "2025-01-15T03:27:49.587952Z",
     "iopub.status.idle": "2025-01-15T03:27:51.290022Z",
     "shell.execute_reply": "2025-01-15T03:27:51.289117Z",
     "shell.execute_reply.started": "2025-01-15T03:27:49.588212Z"
    },
    "trusted": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean number of words in the content: 434.77\n",
      "1-sigma interval: (92.33625778489187, 777.2043999533175)\n",
      "2-sigma interval: (-250.097813299321, 1119.6384710375305)\n",
      "3-sigma interval: (-592.5318843835337, 1462.0725421217433)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.10/dist-packages/seaborn/_oldcore.py:1119: FutureWarning: use_inf_as_na option is deprecated and will be removed in a future version. Convert inf values to NaN before operating instead.\n",
      "  with pd.option_context('mode.use_inf_as_na', True):\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA/wAAALVCAYAAAB9deqxAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAAEAAElEQVR4nOzdd1gUxxsH8O/d0cvRkSJNRMTeoiLYDdhbbNjAGhONLTYSe2LvsRfAHiMq9i427KIkVlQECyAWpLfjbn5/8LuV5e4A9fAMeT/Pc4/e7ty7787eArM7OyNgjDEQQgghhBBCCCGkXBFqOgFCCCGEEEIIIYSoHzX4CSGEEEIIIYSQcoga/IQQQgghhBBCSDlEDX5CCCGEEEIIIaQcogY/IYQQQgghhBBSDlGDnxBCCCGEEEIIKYeowU8IIYQQQgghhJRD1OAnhBBCCCGEEELKIWrwE0IIIYQQQggh5RA1+AkhhKg0c+ZMCASCL7KtFi1aoEWLFtz7c+fOQSAQYM+ePV9k+wEBAXB2dv4i2/pUGRkZGDp0KGxsbCAQCDB27FhNp/TRih5noojqiBBCiLpQg58QQv4jNm/eDIFAwL309PRgZ2cHX19f/PHHH0hPT1fLdhISEjBz5kxERUWpJZ46fc25lcbcuXOxefNm/PDDD9i2bRsGDBigtFy1atVQu3ZtheVhYWEQCARo3ry5wrrg4GAIBAKcPHlS7Xl/KqlUipCQELRo0QLm5ubQ1dWFs7MzBg0ahJs3b2o6PQDA/fv3MXPmTMTFxWk6lRLFxMTg+++/R6VKlaCnpwexWAwvLy+sWLEC2dnZZbbdL1lHWVlZmDlzJs6dO1fm2yKEkH8DLU0nQAgh5MuaPXs2XFxcIJFI8OrVK5w7dw5jx47F0qVLcfDgQdSqVYsrO3XqVEyZMuWj4ickJGDWrFlwdnZGnTp1Sv25L9HQLC63jRs3QiaTlXkOnyM8PByNGzfGjBkzii3n7e2NoKAgpKamwsTEhFt+6dIlaGlp4caNG5BIJNDW1uatE4lE8PT0LLP8P0Z2dja6d++O48ePo1mzZvjll19gbm6OuLg47N69G1u2bMHz589RsWJFjeZ5//59zJo1Cy1atFBbD5GyOBeOHDmCnj17QldXFwMHDkSNGjWQl5eHiIgITJw4Effu3cOGDRvUvl2gbOpIlaysLMyaNQsAqJcEIYSAGvyEEPKf065dOzRo0IB7HxgYiPDwcHTs2BGdO3fGgwcPoK+vDwDQ0tKCllbZ/qrIysqCgYEBdHR0ynQ7JSnc+P1avX79GtWqVSuxnLe3NzZu3IjLly+jXbt23PJLly6hV69e2LlzJyIjI9G4cWNuXUREBGrVqgVjY+PPyjEzMxOGhoafFQMAJk6ciOPHj2PZsmUKjy7MmDEDy5Yt++xtfK3UfS7ExsaiT58+cHJyQnh4OGxtbbl1I0eOxJMnT3DkyBG1bpMQQsjXgbr0E0IIQatWrTBt2jQ8e/YM27dv55Yre4b/1KlT8Pb2hqmpKYyMjODu7o5ffvkFQMFz99988w0AYNCgQdzjA5s3bwZQcMetRo0aiIyMRLNmzWBgYMB9VtVzy1KpFL/88gtsbGxgaGiIzp0748WLF7wyzs7OCAgIUPhs4Zgl5absGf7MzEz8/PPPcHBwgK6uLtzd3bF48WIwxnjlBAIBRo0ahf3796NGjRrQ1dVF9erVcfz4ceUVXsTr168xZMgQVKhQAXp6eqhduza2bNnCrZePZxAbG4sjR45wuavqIu3t7Q2goIEvl5OTg1u3bqF79+6oVKkSb92bN2/w6NEj7nMAcPv2bbRr1w5isRhGRkZo3bo1rl69ytuO/DGR8+fP48cff4S1tTXvjvuGDRvg6uoKfX19NGzYEBcvXixVfbx8+RLr16/Ht99+q3ScApFIhAkTJvC2VZp8VY1JId+PwvXp7OyMjh07IiIiAg0bNoSenh4qVaqErVu38j7Xs2dPAEDLli254yLvTn7z5k34+vrC0tIS+vr6cHFxweDBg0vcf1XjWezevRtz5sxBxYoVoaenh9atW+PJkyclxlu4cCEyMjIQFBTEa+zLVa5cGWPGjOHe5+fn47fffoOrqyv3GMUvv/yC3Nxc3ufUUUcAcOzYMTRt2hSGhoYwNjZGhw4dcO/ePd62AgICYGRkhPj4eHTt2hVGRkawsrLChAkTIJVKAQBxcXGwsrICAMyaNYvb1syZM0usI0IIKa+owU8IIQQAuOfBi+tOfO/ePXTs2BG5ubmYPXs2lixZgs6dO3ONRw8PD8yePRsAMHz4cGzbtg3btm1Ds2bNuBjv3r1Du3btUKdOHSxfvhwtW7YsNq85c+bgyJEjmDx5MkaPHo1Tp06hTZs2H/3McWlyK4wxhs6dO2PZsmVo27Ytli5dCnd3d0ycOBHjx49XKB8REYEff/wRffr0wcKFC5GTk4PvvvsO7969Kzav7OxstGjRAtu2bUO/fv2waNEimJiYICAgACtWrOBy37ZtGywtLVGnTh0ud3njpqhKlSrBzs4OERER3LIbN24gLy8PTZo0QZMmTXgN/suXLwP4cKHg3r17aNq0Kf7++29MmjQJ06ZNQ2xsLFq0aIFr164pbO/HH3/E/fv3MX36dO4RkKCgIHz//fewsbHBwoUL4eXlpfRijTLHjh1Dfn6+yjEKivrYfEvryZMn6NGjB7799lssWbIEZmZmCAgI4BqjzZo1w+jRowEAv/zyC3dcPDw88Pr1a/j4+CAuLg5TpkzBypUr0a9fP4WLEB9j/vz5CAsLw4QJExAYGIirV6+iX79+JX7u0KFDqFSpEpo0aVKq7QwdOhTTp09HvXr1sGzZMjRv3hzz5s1Dnz59FMp+Th0BwLZt29ChQwcYGRlhwYIFmDZtGu7fvw9vb2+FC1pSqRS+vr6wsLDA4sWL0bx5cyxZsoR7FMHKygpr164FAHTr1o3bVvfu3Uu134QQUi4xQggh/wkhISEMALtx44bKMiYmJqxu3brc+xkzZrDCvyqWLVvGALA3b96ojHHjxg0GgIWEhCisa968OQPA1q1bp3Rd8+bNufdnz55lAJi9vT1LS0vjlu/evZsBYCtWrOCWOTk5MX9//xJjFpebv78/c3Jy4t7v37+fAWC///47r1yPHj2YQCBgT5484ZYBYDo6Orxlf//9NwPAVq5cqbCtwpYvX84AsO3bt3PL8vLymKenJzMyMuLtu5OTE+vQoUOx8eR69uzJ9PX1WV5eHmOMsXnz5jEXFxfGGGNr1qxh1tbWXNkJEyYwACw+Pp4xxljXrl2Zjo4Oi4mJ4cokJCQwY2Nj1qxZM26Z/Dvl7e3N8vPzeflbW1uzOnXqsNzcXG75hg0bGADeMVFm3LhxDAC7fft2qfa1tPkW/T4X3Y/Y2FhumZOTEwPALly4wC17/fo109XVZT///DO3LDQ0lAFgZ8+e5cUMCwsr8XxTRdW54OHhwavPFStWMADszp07KmOlpqYyAKxLly6l2nZUVBQDwIYOHcpbLv+OhIeHc8s+t47S09OZqakpGzZsGG/5q1evmImJCW+5v78/A8Bmz57NK1u3bl1Wv3597v2bN28YADZjxoxS7S8hhJR3dIefEEIIx8jIqNjR+k1NTQEABw4c+OQB7nR1dTFo0KBSlx84cCDvufIePXrA1tYWR48e/aTtl9bRo0chEom4u5NyP//8MxhjOHbsGG95mzZt4Orqyr2vVasWxGIxnj59WuJ2bGxs4Ofnxy3T1tbG6NGjkZGRgfPnz39S/t7e3sjOzkZkZCSAgu798ju8Xl5eeP36NR4/fsytc3FxgZ2dHaRSKU6ePImuXbuiUqVKXDxbW1v07dsXERERSEtL421r2LBhEIlE3PubN2/i9evXGDFiBO959ICAAN4ggqrI45dmPIFPybe0qlWrhqZNm3Lvrays4O7uXuIxBT6cK4cPH4ZEIvmk7Rc1aNAgXn3Kcysun4+pSwDceVW0F8vPP/8MAArP+n9OHZ06dQopKSnw8/PD27dvuZdIJEKjRo1w9uxZhc+MGDGC975p06al2hYhhPxXUYOfEEIIJyMjo9iGQe/eveHl5YWhQ4eiQoUK6NOnD3bv3v1RjX97e/uPGpTMzc2N914gEKBy5cplPsXXs2fPYGdnp1Af8q7Iz5494y13dHRUiGFmZob379+XuB03NzcIhfxfyaq2U1qFn+NnjOHy5cvw8vICANSoUQNisRiXLl1CTk4OIiMjufJv3rxBVlYW3N3dFWJ6eHhAJpMpdMt3cXFR2CdA8dhpa2vzGuWqiMViACjVVJGfkm9pfeoxBYDmzZvju+++w6xZs2BpaYkuXbogJCRE4Tn4z8nHzMwMAIrN52PqEig4dkKhEJUrV+Ytt7Gxgampqdq+9wC4C06tWrWClZUV73Xy5Em8fv2aV15PT0/hMZbSbosQQv6raJR+QgghAAoGSktNTVX4Q78wfX19XLhwAWfPnsWRI0dw/Phx/PXXX2jVqhVOnjzJu8tbXAx1UzYQG1Bw97c0OamDqu2wIgP8fSm1a9eGsbExIiIi0L59eyQnJ3N3+IVCIRo1aoSIiAi4uroiLy+PN2Dfx1L3Ma1atSoA4M6dOx81tWNJivueKPM5x1QgEGDPnj24evUqDh06hBMnTmDw4MFYsmQJrl69CiMjo9In/hn5iMVi2NnZ4e7dux+1LVV1pY6c5OQXCrdt2wYbGxuF9UVnCPlS5zIhhJQndIefEEIIgII/ugHA19e32HJCoRCtW7fG0qVLcf/+fcyZMwfh4eFc99vSNhRKS34XUI4xhidPnvBG1DczM0NKSorCZ4vejfyY3JycnJCQkKBwZ/Thw4fcenVwcnLC48ePFXpJfO52RCIRGjdujEuXLiEiIgJisRg1a9bk1ssH7pMP3idv8FtZWcHAwADR0dEKMR8+fAihUAgHB4cS9wlQPHYSiQSxsbEl5t6uXTuIRCLejBGqfEy+8jviRb8rn9qLAij5O9W4cWPMmTMHN2/exI4dO3Dv3j3s2rXrk7f3KTp27IiYmBhcuXKlxLJOTk6QyWQKxy4pKQkpKSmf9H1UVUfyR2Csra3Rpk0bhZeyWTs+dVuEEPJfRQ1+QgghCA8Px2+//QYXF5diR/1OTk5WWCa/Ayvvqiyfg11ZA/xTbN26ldfo3rNnDxITE3nzy7u6uuLq1avIy8vjlh0+fFihK/fH5Na+fXtIpVKsWrWKt3zZsmUQCAS87X+O9u3b49WrV/jrr7+4Zfn5+Vi5ciWMjIzQvHnzT47t7e2NN2/eICQkBI0aNeI9NtCkSRNER0fjwIEDsLCw4B4hEIlE8PHxwYEDB3iPTSQlJWHnzp3w9vbmuomr0qBBA1hZWWHdunW8Y7J58+ZS1b2DgwOGDRuGkydPYuXKlQrrZTIZlixZgpcvX35UvvIG5oULF7hymZmZvCkQP5aq79T79+8V7nIXPVe+lEmTJsHQ0BBDhw5FUlKSwvqYmBhuRoj27dsDAJYvX84rs3TpUgBAhw4dPnr7qurI19cXYrEYc+fOVTrOwZs3bz56WwYGBkq3RQgh/1XUpZ8QQv5jjh07hocPHyI/Px9JSUkIDw/HqVOn4OTkhIMHD0JPT0/lZ2fPno0LFy6gQ4cOcHJywuvXr7FmzRpUrFiRu0Ps6uoKU1NTrFu3DsbGxjA0NESjRo0UnvMuLXNzc3h7e2PQoEFISkrC8uXLUblyZQwbNowrM3ToUOzZswdt27ZFr169EBMTg+3bt/MG0fvY3Dp16oSWLVvi119/RVxcHGrXro2TJ0/iwIEDGDt2rELsTzV8+HCsX78eAQEBiIyMhLOzM/bs2YNLly5h+fLlpR5sTRn5Mbly5YrCXOSNGzeGQCDA1atX0alTJ96d0d9//x2nTp2Ct7c3fvzxR2hpaWH9+vXIzc3FwoULS9yutrY2fv/9d3z//fdo1aoVevfujdjYWISEhJTqGX4AWLJkCWJiYjB69Gjs27cPHTt2hJmZGZ4/f47Q0FA8fPiQmyautPn6+PjA0dERQ4YMwcSJEyESiRAcHAwrKys8f/68VHkVVadOHYhEIixYsACpqanQ1dVFq1atsHPnTqxZswbdunWDq6sr0tPTsXHjRojFYq5R/aW4urpi586d6N27Nzw8PDBw4EDUqFEDeXl5uHz5MkJDQxEQEACg4FEQf39/bNiwASkpKWjevDmuX7+OLVu2oGvXriVOo6mMqjqytrbG2rVrMWDAANSrVw99+vThjsWRI0fg5eWlcMGtJPr6+qhWrRr++usvVKlSBebm5qhRowZq1Kjx0XkTQki5oLkJAgghhHxJ8qnH5C8dHR1mY2PDvv32W7ZixQre9G9yRacxO3PmDOvSpQuzs7NjOjo6zM7Ojvn5+bFHjx7xPnfgwAFWrVo1pqWlxZsGr3nz5qx69epK81M1Fdmff/7JAgMDmbW1NdPX12cdOnRgz549U/j8kiVLmL29PdPV1WVeXl7s5s2bCjGLy63otHyMFUwbNm7cOGZnZ8e0tbWZm5sbW7RoEZPJZLxyANjIkSMVclI1XWBRSUlJbNCgQczS0pLp6OiwmjVrKp068GOm5WOMsczMTG4/T548qbC+Vq1aDABbsGCBwrpbt24xX19fZmRkxAwMDFjLli3Z5cuXeWVKmupxzZo1zMXFhenq6rIGDRqwCxcuKD0mquTn57NNmzaxpk2bMhMTE6atrc2cnJzYoEGDFKbsK02+jDEWGRnJGjVqxHR0dJijoyNbunSpymn5lNW1svw3btzIKlWqxEQiETf93K1bt5ifnx9zdHRkurq6zNramnXs2JHdvHmzxP1WdS6EhobyysXGxqqcZlKZR48esWHDhjFnZ2emo6PDjI2NmZeXF1u5ciXLycnhykkkEjZr1izm4uLCtLW1mYODAwsMDOSVYezz66jw/vn6+jITExOmp6fHXF1dWUBAAK+u/P39maGhocK2lE21ePnyZVa/fn2mo6NDU/QRQv7zBIxpaDQhQgghhBBCCCGElBl6hp8QQgghhBBCCCmHqMFPCCGEEEIIIYSUQ9TgJ4QQQgghhBBCyiFq8BNCCCGEEEIIIeUQNfgJIYQQQgghhJByiBr8hBBCCCGEEEJIOaSl6QT+DWQyGRISEmBsbAyBQKDpdAghhBBCCCGElHOMMaSnp8POzg5C4afdq6cGfykkJCTAwcFB02kQQgghhBBCCPmPefHiBSpWrPhJn6UGfykYGxsDAGJjY2Fubq6WmBKJBCdPnoSPjw+0tbUpHsX7V8cri5gUj+JpMl5ZxKR4FE+T8coiJsWjeJqOSfEonibjlUXMovHS0tLg4ODAtUc/BTX4S0Hejd/Y2BhisVhluSsvriBXmgtdkS48HTyLjSmRSGBgYACxWKy2LwcXLzQUyMoCDAyAvn0/P95n5CevExETld3+libemyuALBcQ6gJWisemTI9HoXgvrryANFcKka4IDp6l7zXypfL7mmJSPIqnyXhlEZPiUTxNxiuLmBSP4mk6JsWjeJqMVxYxVcX7nMfKqcGvRj1DeyI+PR72xvZ4Of6l5hKZNAmIjwfs7T+5wa8uhetktetqzSUS0RPIjgf07YFumjs2oT1DkR6fDmN7Y4x/OV5jeRBCCCGEEELKPxqlnxBCCCGEEEIIKYfoDr8ajfxmJNJy0yDWVd3t/79GXieG2oZAigYTqTISkKQB2po9Nt+M/Aa5abnQFetqNA9CCCGEEEJI+UcNfjUKbBqo6RS+OvI6kUgkOHr0qOYSqf51HJumgU01nQIhhJQbMpkMeXl5n/RZiUQCLS0t5OTkQCqVfnYuX3u8sohJ8SiepmNSPIqnyXjqiqmjo/PJU+6VBjX4CSGEEPKvk5eXh9jYWMhksk/6PGMMNjY2ePHixWcNhvRviVcWMSkexdN0TIpH8TQZT10xhUIhXFxcoKOjo5aciqIGPyGEEEL+VRhjSExMhEgkgoODwyfdGZHJZMjIyICRkZFa7qx87fHKIibFo3iajknxKJ4m46kjpkwmQ0JCAhITE+Ho6KiWnIqiBj8hhBBC/lXy8/ORlZUFOzs7GBgYfFIM+eMAenp6avtD8muOVxYxKR7F03RMikfxNBlPXTGtrKyQkJCA/Px8teRUFDX41ajBhgZ4lfEKNkY2uDn8pqbT+SrI66SCYQVMt5muuUSONwCyXwH6NkBbzR2bDQ02IONVBoxsjDD85nCN5UEIIf9m8ucky6r7IyGEEPKlyH+XSaVSiEQitcenBr8avcp4hfj0eE2n8VX5auok+xWQrfk8Ml5lID0+XdNpEEJIuaCuZzAJIYQQTSnr32XU4FcjGyMb3r+aS8SG/68GyeuigmEFzSaib8P/V0OMbIx4/xJCCCGEEEJIWaEGvxp9Nd34b34leeBDnWh8Wj4NduMvjLrxE0IIIYQQQr6UspvwjxBCCCGEcAICAiAQCDBixAiFdSNHjoRAIEBAQMCXT+wjjRgxAgKBACtWrOAt79y5MxwdHaGnpwdbW1sMGDAACQkJSmM8efIExsbGMDU1LXZbmzdvhkAgUPp6/fo1gA/1WvRVs2ZNtewvIYT8m1GDnxBCCCHkC3FwcMCuXbuQnZ3NLcvJycHOnTvLbEomdQoLC8PVq1dhZ2ensK5ly5bYvXs3oqOjsXfvXsTExKBHjx4K5SQSCfz8/NC0adMSt9e7d28kJibyXr6+vmjevDmsra0BACtWrOCtf/HiBczNzZVumxBC/muowU8IIYQQ8oXUq1cPDg4O2LdvH7ds3759cHR0RN26dXllZTIZ5s+fDxcXF+jr66N27drYs2cPt14qlWLIkCHcend3d4W77gEBAejatSsWL14Me3t7VKpUCaNGjYJEIvno3OPj4/HTTz9hx44d0NbWVlg/btw4NG7cGE5OTmjSpAmmTJmCq1evKmxr6tSpqFq1Knr16lXiNvX19WFjY8O9RCIRwsPDMWTIEK6MiYkJr8zNmzfx/v37f0VvCUIIKWv0DL8azbs4D2m5aRDrihHYNFBziXz/PZCcDJibA+vXay4PfKgTQ21D1IQGu9bdmwdI0gBtMVBdc8fm4ryLyE3Lha5YF00DS76zQQghpPwZPHgwQkJC0K9fPwBAcHAwBg0ahHPnzvHKLV26FHv37sW6devg5uaGCxcuoH///rCyskLz5s0hk8lQsWJFhIaGwsLCApcvX8bw4cNha2vLa0yfPXsWtra2OHPmDP755x8MGTIEdevWxbBhwwAAM2fOxObNmxEXF6cyZ5lMhgEDBmDixImoXr16ifuYnJyMHTt2oEmTJryLA+Hh4QgNDUVUVBTvokdpbd26FQYGBsXevQ8KCkKbNm3g5OSEtLS0j94GIYSUJ9TgV6PVN1YjPj0e9sb2mm3wHzkCxMcD9vaay+H/CtfJatfVmkvk0eqCafn07TXa4L+x+gbS49NhbG9MDX5CCFG3pUsLXiWpVw/Yv5+/rHNn4Natkj87fnzB6zP0798fgYGBePbsGQDg0qVL2LVrF6/Bn5ubi2XLluHkyZPw8vICAFSqVAkRERFYv349mjdvDm1tbcyaNYv7jIuLC65cuYLdu3fzGvxmZmZYtWoVBAIB7Ozs0L59e5w5c4Zr8FtaWsLV1bXYnBcsWAAtLS2MHj262HKTJ0/GqlWrkJWVhcaNG+Pw4cPcunfv3iEgIADbt2+HWCwuXWUVERQUhL59+0JfX1/p+oSEBBw7dgw7d+78pPiEEFLeUIOfEEIIIeVDWlrBBe+SODgoLnvzpnSfVcMdYysrK3To0AGbN28GYwwdOnSApaUlr8yTJ0+QlZUFX19f3vK8vDxe1//Vq1cjODgYz58/R3Z2NvLy8lCnTh3eZ6pXrw6RSASZTAYAsLW1xd27d7n1o0aNwqhRo1TmGxkZiRUrVuDWrVslzhc9ceJEDBkyBM+ePcOsWbMwcOBAHD58GAKBAMOGDUPfvn3RrFmzYmOocuXKFTx48ADbtm1TWWbLli0wNTVF165dP2kbhBBS3lCDX41Ce4YiV5oLXZGuplP5asjrRMRESP4nWXOJeIcCslxAqNlj0zO0J6S5Uoh0RRrNgxBCyiWxuHS926yslC8rzWc/8c50UYMHD+Ya2atXK/aAy8jIAAAcOnQIDkUuUOjqFvwu27VrFyZMmIAlS5bA09MTxsbGWLRoEa5du8YrX/R5e4FAwDX+S+PixYt4/fo1b1BBqVSKCRMmYNmyZbxHASwtLWFpaYkqVarAw8MDDg4OuHr1Kjw9PREeHo6DBw9i8eLFAADGGGQyGbS0tLBhw4YSn7nftGkT6tSpg/r16ytdzxhDcHAwBgwYAB0dnY/aR0IIKa+owa9Gng6emk7hqyOvE4lEgqP/HNVcIlZfx7Fx8FRyV4kQQoh6fEx3+6KNwYMH1Z9PMdq2bYu8vDwIBAKFu/gAUK1aNejq6uL58+do2bKl0hiXLl1CkyZN8OOPP3LLYmJi1J7rgAED0KZNG94yX19f9O/fv9hn6eUN7tzcXAAFd+ilUim3/sCBA1iwYAEuX74M+xIutmRkZGD37t2YN2+eyjLnz5/HkydPeAP6EULIfx01+AkhhBBCvjCRSIQHDx5w/y/K2NgYo0aNws8//wwA8Pb2RmpqKi5dugSxWAx/f3+4ublh69atOHHiBFxcXLBt2zbcuHEDLi4uH5XLqlWrEBYWhjNnzihdb2FhAQsLC94ybW1t2NjYwM3NDQBw7do13LhxA97e3jAzM0NMTAymTZsGV1dXeHoWXHT38PDgxbh58yaEQiFq1KgBoOACweHDh/H777/j4cOHvLJ//fUX8vPz0b9/f5X7ERQUhEaNGnHxCCGE0LR8hBBCCCEaIRaLix287tdff8XUqVMxb948eHh4oG3btjhy5AjXoP/+++/RvXt39O7dG40aNcK7d+94d/tL6+3bt5/dM8DAwAD79u1D69at4e7ujiFDhqBWrVo4f/489whCaaSlpSE6OlpheVBQELp37w5TU1Oln0tNTcXevXvp7j4hhBRBd/jVKPptNPJl+dASasHd0l3T6XwV5HXCZEyziaRFA7J8QKgFiDV3bN5Gv4UsXwahlhCW7pYlf4AQQki5sXnz5mLX7y8yc4BAIMDo0aMxduxYpeV1dXUREhKCkJAQ3vLC3d6VbXPZsmUQCj/c85k5cyZmzpxZbG5FxcXFQSaTcdPe1axZE+Hh4R8VIyAgQOG5/b59+2LEiBEKZS9fvlxsLBMTE2RlZX3U9gkh5L+AGvxq1Hpra24KupfjX2o6na9C4TrR6LR8Z1p/mJavm+aOzdbWW7lp+ca//LxpnQghhBBCCCGkONSlnxBCCCGEEEIIKYfoDr8a+dXww/uc9zDTM9NwIn7A+/eAmYbzwIc6MdExAXI1mIizH5D3HtDRbJ3U8KuBnPc50DPT02gehBBCCCGEkPKPGvxqtMhnkaZTKLDoK8kDH+pEIpHg6FENTstX9+uoE59FPppOgRBCCCGEEPIfQQ3+MtbEuymSXr9WWK6rq4t5v/+G2vXqw9TEBJcjLmogO0IIIYQQQggh5RU1+MtY0uvX2Hz+lsJyJs3H+8hzWHfsIga1aqiBzAghhBBCCCGElGc0aB8hhBBCCCGEEFIO0R1+NWq7vS1eZ76GtaE1jvc/rrlEqlYFEhIAOzvg4UPN5YEPdWJlYIUfxT9qLpGzbYGc14CeNdBSc8dme9vtyHydCUNrQ/Q/3l9jeRBCCCGEEELKP2rwq9Hd13e5Oec1KiMDSE8v+FfDeHUi1mAiKXeB7HhAX7PH5vXd10iPT4exvbFG8yCEEEIIIYSUf9SlX410RDrcixT4aupEqPPhpUEiHRH3IoQQQj6Xs7Mzli9fruk0PsqZM2fg4eEBqVSq6VQIIaTUpkyZgp9++knTaXw0avCr0dMxT5E7NRdPxzzVdCpfDXmdRP8YrdlEujwF+uQW/KtBY56OwdTcqRjzdIxG8yCEEPLlXbhwAZ06dYKdnR0EAgH279//2TFv3LiB4cOHf35yX9CkSZMwdepUiEQfLn6vXr0a1atXh62tLTw8PLB161beZzZu3IimTZvCzMwMZmZmaNOmDa5fv17sdiIiIuDr6wsrKyvo6+ujatWqWLZsGa/M2rVrUatWLYjFYojFYnh6euLYsWMKsa5cuYI2bdrA3t4epqamaNasGbKzswEAcXFxGDJkCFxcXKCvrw9XV1fMmDEDeXl53Ofj4uIgEAh4L5FIhBs3bnBlWrRooVBGIBCgQ4cOXBll6wUCARYVmpL50aNH6NKlCywtLSEWi+Ht7Y2zZ88qraN3796hYsWKEAgESElJ4a3bsWMHvL29YWRkBFtbWwwePBjv3r3j1kskEsyePRuurq7Q09ND7dq1cfw4/7HJ0nznMzIyMGrUKFSsWBH6+vqoVq0a1q1bV2zdyV+hoaFK90uZgIAApTGqV6/OlalUqRLMzMwgEol4ZUaOHPnJuUgkEkyZMgU1a9aEoaEh7OzsMHDgQCQkJPDKJScno1+/fhCLxTA1NcWQIUOQoaK37pMnT2BsbAxTU1Pe8n379qFBgwYwNTWFoaEh6tSpg23btvHKJCUlISAgAHZ2djAyMkKPHj3w+PFjXpmYmBh069YNVlZWEIvF6NWrF5KSknhlbt26hW+//RampqawsLDA8OHDFfJVVk+7du3i1kdERMDLywsWFhYqz9GgoCDUqVOn2HN0w4YNaNGiBcRisdLv8rlz57hzrujxlZ+Dqo7t1atXuTgTJkzAli1b8PTpv6utRw1+QgghhJAvIDMzE7Vr18bq1avVFtPKygoGBgZqi1fWIiIiEBMTg++++45btnbtWgQGBmL69Om4cuUKZsyYgZEjR+LQoUNcmXPnzsHPzw9nz57FlStX4ODgAB8fH8THx6vclqGhIYYNG4Zz587hwYMHmDp1KqZOnYoNGzZwZSpWrIj58+cjMjISN2/eRKtWrdClSxfcu3ePK3PlyhW0bdsW3377LU6fPo1r165h1KhREAoL/ox++PAhZDIZ1q9fj3v37mHZsmVYt24dfvnlF4WcTp8+jcTERCQmJiI+Ph516tTh1u3bt49bl5iYiLt370IkEqFnz55cmcLrExMTERwcDIFAwKvPzp07Iz8/H+Hh4YiMjETt2rXRsWNHvHr1SiGfIUOGoFatWgrLL126hICAAAwYMAB37txBaGgorl+/jmHDhnFlpk6divXr12PlypW4f/8+RowYgW7duuH27dtcmdJ858ePH4/jx49j+/btePDgAcaOHYtRo0bh4MGDAAAHBweF/Z41axaMjIzQrl07lXGLWrFiBS/GixcvYG5uzqvfa9eu4eHDh4iPj0diYiJOnToFAFyZT8klKysLt27dwrRp03Dr1i3s27cP0dHR6Ny5M69cv379cO/ePZw6dQqHDx/GhQsXlF7Mk0gk6NevH5o2baqwztzcHL/++iuuXLmCf/75B4MGDcKgQYNw4sQJAABjDF27dsXTp09x4MABREZGomLFivDx8UFmZiaAgmPm4+MDgUCA8PBwXLp0CXl5eejUqRNkMhkAICEhAW3atEHlypVx7do1HD9+HPfu3cOgQYMUcgoJCeHVV9euXbl1hoaGGDVqFC5cuKDyHLWzs8PcuXOLPUezsrLQtm1bpeccADRp0oQ75+THd+jQoXBxcUGDBg14ZQufo4mJiahfvz63ztLSEr6+vli7dq3S7Xy1GClRamoqA8Devn370Z+t5FaFXUjIUHidf5HC9u/fz86/SGGV3Kp8do55eXls//79LC8vjzF7e8aAgn/VEU8NKF75jlcWMSkexdNkvLKISfHUFy87O5vdv3+fZWdnf3I8qVTK3r9/z6RSqVry+9h4AFhYWFiJ8fLz89mMGTOYg4MD09HRYba2tuynn37iyjk5ObFly5Zx7x88eMC8vLyYrq4u8/DwYKdOneK2JZVK2d9//80AsL/++ot5e3szPT091qBBAxYdHc2uX7/O6tevzwwNDVnbtm3Z69evubjXr19nbdq0YRYWFkwsFrNmzZqxGzdufHQdjhw5kvXo0YO3zNPTk02YMIFXh+PHj2deXl4q4+Tn5zNjY2O2ZcuWEuuwcH7dunVj/fv3LzZHMzMztmnTJu59o0aN2NSpUz/qGC9cuJC5uLhw72NjYxkAdvv27WLzK2zZsmXM2NiYZWRkqNxOly5dWKtWrbh4T548YQDYhQsXuDJpaWkMADt16hTvs2vWrGHNmzdnZ86cYQDY+/fvuXWLFi1ilSpV4uX3xx9/MPtCf1va2tqyVatW8WJ2796d9evXT2muANjevXsV9rl69eps9uzZvLL16tVjv/76q8r9rlOnDhs8ePBnncdhYWFMIBCwuLg4blnReGPGjGGurq5MJpOVmIsyqvK7fv06A8CePXvGGGPs/v37DAC7ceMGV+bYsWNMIBCw+Ph4XrzRo0ezfv36sZCQEGZiYlLiftatW5dNnTqVMcZYdHQ0A8Du3r3LxXv37h2zsrJiGzduZIwxduLECSYUCllqaioXIyUlhQkEAu47tH79emZtbc3br3/++YcBYJGRkdzykn7OKVP4HFVVf0XPUbmzZ88qfJcLk8fLyclhVlZWvO+dsnNUmS1btrCKFSsqxPyc3yWFf6cV/d0pb4cWPh4fi+7wE0IIIYR8pfbu3Ytly5Zh/fr1ePz4Mfbv34+aNWsqLSuVStG1a1cYGBjg2rVr2LBhA3799VelZWfMmIGpU6fi1q1b0NLSQt++fTFp0iSsWLECFy9exJMnTzB9+nSufHp6Ovz9/REREYGrV6/Czc0NHTt2RHp6OlcmICAALVq0KHZ/Ll68qHBHLTc3F3p6erxl+vr6uH79OiQSidI4WVlZkEgkMDc3L3Z7hd2+fRuXL19G8+bNla6XSqXYtWsXMjMz4enpCQB4/fo1rl27Bmtra3h7e6NKlSpo2bIlIiIiit1Wamqq0tw6d+7MxZLfwVYlKCgIffr0gaGhodL1SUlJOHLkCIYMGcItMzc3h7u7O7Zu3YrMzEzk5+dj/fr1sLa25t2pvH//PmbPno2tW7dyPRUK8/T0xIsXL3Dy5EkwxpCUlIQ9e/agffv2XBlVx62kuimqSZMmOHjwIOLj48EYw9mzZ/Ho0SP4+PgoLR8ZGYmoqCjefn+KoKAgtGnTBk5OTkrX5+XlYfv27Rg8eDAEAoFac0lNTYVAIOC65F+5cgWmpqa8c6NNmzYQCoW4du0atyw8PBwHDhzAqlWrStwGYwxnzpxBdHQ0mjVrBqDgmAHgHTehUAhdXV3uuOXm5kIgEEBXV5cro6enB6FQyCujo6PD++7o6+sDAK8LPACMHDkSlpaWaNiwIYKDg8EYU5nzp5yjn+LgwYN49+6d0h4JJZ2jDRs2xMuXLxEXF/fJ2//SaJR+NdoQuQEZeRkw0jHC8Pr/rufpyoq8TvRF+rCDneYSebIBkGQA2kZAZc0dm8gNkcjLyIOOkQ7qD69f8gcIIYSU2tIrS7H0ylIAwPbu29HCuQW3LvZ9LJqGFHSB7Va1G1a0XcH7bOc/O+NW4i0AwMvxL3nrNkdtxtTwqQCAP9r9ge4e3ctqFxS8ePECNjY2aNOmDbS1teHo6IiGDRsqLXvq1CnExMTg3LlzsLGxAQDMmTMH3377rULZCRMmwNfXFwAwZswY+Pn54cyZM/Dy8gJQ0NV78+bNXPlWrVrxPr9hwwaYmpri0qVL6NWrFwDA1taW6/KryrNnz2Bnx/97wNfXF5s2bULnzp1RuXJl3Lx5E5s2bYJEIsHbt29ha2urEGfy5Mmws7NDmzZtit0eADg6OuLNmzfIz8/HzJkzMXToUN76O3fuwNPTEzk5OTAyMkJYWBiqVasGANyzujNnzsTChQvh5uaGffv2oXXr1rh79y7c3NwUtvfkyROsXLkSixcv5pYZGRlhyZIl8PLyglAoxN69e9G9e3ds374dffr0UYhx/fp13L17F0FBQSr3a8uWLTA2Nkb37h++jwKBACdPnkT37t1hbGwMoVAIa2trHD9+HGZmZgAKGmt+fn5YtGgRHB0dlT6P7OXlhW3btmHIkCHo168f8vPz0alTJ17XfF9fXyxduhTNmjWDq6srzpw5g3379n30YIwrV67E8OHDUbFiRWhpaUEoFGLjxo1cI7WooKAgeHh4oEmTJiV+31RJSEjAsWPHsHPnTpVl9u/fj5SUFAQEBKgsUziX0srJycHkyZPh5+cHsbhgCqtXr17B2tqaV05LSwvm5ubcoxjv3r3D4MGDsXbtWu5zyqSmpsLe3h65ubkQiURYs2YN9zOgatWqcHR0RGBgINavXw99fX0sX74cL1++RGJiIgCgcePGMDQ0xOTJkzF37lwwxjBlyhRIpVKuTKtWrTB+/HgsWrQIY8aMQWZmJqZMmcLti9zs2bPRqlUrGBgY4OTJk/jxxx+RkZGB0aNH83KuWLFiieeol5eX0nP0UwQHB8PX1xcVK1bklik7R7t27Yr9+/fzHr+Q//x69uwZnJ2dPzmHL4ka/Go0+/xsbgo6avAXKFwnq13V98ziR7sz+8O0fBps8J+ffZ6blo8a/IQQol5puWmITy94pjs3P5e3Tsqk3Lr3Oe8VPvsm6w23vqjMvExuXZYkS50p88ydOxdz587l3l+5cgU9evTAihUrUKlSJbRt2xbt27dHp06doKWl+CdcdHQ0HBwcuMY+AJUXBwo/t12hQgUA4PUcqFChAl6/fs29T0pKwtSpU3Hu3Dm8fv0aUqkUWVlZePnyw8WRefPmlbiP2dnZCneFp02bhlevXqFJkyZgjKFChQrw9/fHwoULld59nj9/Pnbt2oVz584pxFLm/PnzyMrKwtWrVzFlyhRUrlwZfn5+3Hp3d3dERUUhNTUVe/bsgb+/P86fP49q1apxDcrvv/8egwYNQlpaGry9vREeHo7g4GCFfY6Pj0fbtm3Rs2dP3vPulpaWGD9+PPf+m2++QXx8PFauXKm0wR8UFISaNWuqPH5AQaOlX79+vDpgjGHUqFGwtrbGxYsXoa+vj02bNqFTp064ceMGbG1tERgYCA8PD/Tv319l7Pv372PcuHGYOHEiOnfujKSkJEycOBEjRozgLkKsWLECw4YNQ9WqVSEQCODq6opBgwYhODhYZVxlVq5ciatXr+LgwYNwcnLChQsXMHLkSKUXdLKzs7Fz505Mmzbto7ZR1JYtW2Bqasp7nryooKAgtGvXTuEC1efkIpFI0KtXLzDGPvo58GHDhsHPz4+7KKeKsbExoqKikJGRgTNnzmD8+PGoVKkSWrRoAW1tbezbtw9DhgyBubk5RCIRWrRogbZt23Kft7KyQmhoKH744Qf88ccfEAqF8PPzQ7169bjzsXr16tiyZQvGjx+PwMBAiEQijB49GhUqVOCds4Xrpm7dusjMzMSiRYsUGvwXL15ERkbGJ52jHys+Ph4nT57E7t27ecuVnaMJCQlYtGgRr8Ev78mQlVV2vwvUjRr8hBBCCCkXxLpi2BvbAwB0tXR560QCEbfOTM9M4bNWBlbc+qIMdQy5dQbaZTdA3ogRI7i75TKZDObm5jA3N0d0dDROnz6NU6dO4ccff8SiRYtw/vx5aGtrf/K2Cn9W3l256LLCd0/9/f3x7t07rFixAk5OTtDV1YWnp6fKLveqWFpa4v17/gUXfX19BAcHY+3atXjy5AmqVKmCTZs2wdjYGFZWVryyixcvxvz583H69Gmlg80p4+LiAqFQiJo1ayIpKQkzZ87kNSZ0dHRQuXJlAED9+vVx48YNrFixAuvXr+d6FxRtWHh4eOD58+e8ZQkJCWjZsiWaNGnCG3RMlUaNGnGDwhWWmZmJXbt2Yfbs2So/e/HiRURHR+Ovv/7iLb9w4QKOHDmC9+/fc3eB16xZg1OnTmHLli2YMmUKwsPDcefOHezZswcAuC7WlpaW+PXXXzFr1izMmzcPTZo0wejRoyEWiyEUCmFoaIimTZvi999/h62tLaysrLB//37k5OTg3bt3sLOzw5QpU1CpUqUS910uOzsbv/zyC8LCwrjZCGrVqoWoqCgsXrxYocG/Z88eZGVlYeDAgaXeRlGMMQQHB2PAgAHQ0VE+XfOzZ89w+vRp7Nu3T2Wcj81F3th/9uwZwsPDeXfpbWxseBfYACA/Px/JycncBbzw8HAcPHgQS5Ys4fZDJpNBS0sLGzZswODBgwEUdNGXf5/r1KmDBw8eYN68edzjNvXr1+cazzk5OdDV1YWvry/vcQIfHx/ExMTg7du30NLSgqmpKWxsbHjHtm/fvujbty+SkpJgaGgIgUCApUuXFnvXu1GjRvjtt9+Qm5vLe2TAxcUFAD7pHP1YO3fuhIWFhcKgiaryLXqOJicnA4DCz6avGTX41Whdx3XIlmRDX1tfw4msA7KzAX0N54EPdaIt0AYel1y+zDRcB0izAZFm66Tjuo6QZEugrf/pf6QRQghRbrzneIz3HK90nYuZC6+rftGuwAf9VD9PHVAnAAF1AtSSY3HkDXygIL+0tDQABQ3iTp06oVOnThg5ciSqVq2KO3fuoF69erzPu7u748WLF0hKSuLu2hee9u1zXLp0CWvWrOGe4X7x4gXevn370XHq1q2L+/fvK12nra0Ne3t7iEQi7Nq1Cx07duTdLVy4cCHmzJmDEydOKIwDUFoymYx7jrk0ZZydnWFnZ4foaP70wo8ePeKNyh4fH4+WLVuifv36CAkJUdozoaioqCjuOBUWGhqK3NzcYu/ABwUFoX79+qhduzZvufyuY9HtC4VC7ju/d+9ebkpBoOA7MnjwYFy8eBGurq5cnMLTJgLg3hd9BltPTw/29vaQSCTYu3cvd9GqNCQSCSQSiUK+IpFIaXf9oKAgdO7c+bMaW+fPn8eTJ0+Kfe5+8+bNsLa25k2J+Dm5SCQS9OnTB48fP8bZs2dhYWHBW+/p6YmUlBRERkZyYy2Eh4dDJpOhUaNGAAp6/EgkEmRkZMDIyAiHDh3CggULcPnyZdjbK79YCaj+zpuYmMDY2Bi3b9/GzZs38dtvvymUsbS05HJ5/fq10kay/DscHBwMPT09tGzZUmUuUVFRMDMz4zX2S5vvx5ZRhjGGHTt2YMCAAaW6YBoVFaXwSNHdu3ehra3Nm87xa0cNfjXqWKWjplMo0PEryQMf6kQikeDo46OaS8T+66iTKh2raDoFQgghGpKRkYEnT55w72NjYxEVFQVzc3M4Ojoq/czmzZvBGEOjRo1gYGCA7du3Q19fX+lAY99++y1cXV257vDp6emYOrVg7AFVg46VlpubG7Zt24YGDRogLS0NEydO5Lq2ygUGBiI+Ph5bt25VGcfX1xdbtmzhLXv06BGuX7+Ob775Bi9fvsSGDRtw9+5dXrkFCxZg+vTp2LlzJ5ydnbnnhI2MjGBkZKR0+2vWrIGFhQXXFfnChQtYvHgxrztxYGAg2rVrB0dHR6Snp2Pnzp04d+4cN42ZQCDAxIkTMWPGDNSsWROVK1fGvn378PDhQ+4OeXx8PFq0aAEnJycsXrwYb9684eLL785u2bIFOjo6qFu3LoCCKfhCQkLwxx9/KNRRUFAQunbtqtAolEtLS0NoaCh3p7ewhg0bwszMDP7+/pg+fTr09fWxceNGxMbGco1XeaNeTn7hxsPDgxtErlOnThg2bBgaNWqELl26ICkpCWPHjkXDhg25Lu7Xrl3jphaMj4/HzJkzIZPJMGnSJC520e98XFwc7ty5AwcHBzg7O0MsFqN58+bc98nJyQnnz5/H1q1bsXTpUl6eT548wYULF3D06Of9PRkUFIRGjRqhRo0aStfLZDJs3rwZ/v7+Sh+dKU0uVatWxbx589ClSxdIJBIMHDgQt2/fxuHDhyGVSrnvr7m5OXR0dODh4YG2bdti2LBhWLduHSQSCUaNGoU+ffpw9e3h4cFdCBSLxbh16xaEQiFvP+bNm4cGDRrA1dUVubm5OHr0KLZt28Z7fCA0NBRWVlZwdHTE33//jTFjxqBLly68QRJDQkLg4eEBKysrXLlyBWPGjMG4cePg7u7OlVm1ahWaNGkCIyMjnDp1ChMnTsS8efNgYmICADh06BCSkpLQuHFj6Onp4dSpU5g7dy4mTJjAxVi9ejUcHR1RtWpVAFB6js6aNQtdunSBs7Oz0nMUKBg34NWrV9x37c6dOzA2NoajoyNv8Mzw8HA8e/ZM6cUeZedocHAwNm3axCt38eJFNG3aVOHn39eMGvyEEEIIIV/AzZs3eXe/5M+L+vv78wbIK8zU1BQLFy7E+PHjIZVKUbNmTRw6dEhpY1AkEmH//v0YOnQovvnmG1SqVAmLFi1Cp06dSvWse3GCgoIwfPhw1KtXDw4ODgp/uAMFc8QX7eZeVL9+/TBp0iRER0dzjQepVIolS5YgOjoaWlpaaNmyJS5fvszrGrx27Vrk5eWhR48evHgzZszAzJkzlW5fJpNh9uzZeP78ObS0tODq6ooFCxbg+++/58q8fv0aAwcORGJiIkxMTFCrVi2cOHGCN9Dh2LFjkZOTg59//hnJycmoXbs2Tp06xTWcT506hSdPnuDJkye8QcAA/t3w3377Dc+ePYOWlhaqVq2KP//8U2Ek+ujoaERERODkyZMq63DXrl1gjPG6PMtZWFjg6NGjmDZtGlq1agWJRILq1avjwIEDCr0BihMQEIC0tDSsXbsW06ZNg6mpKVq1aoUFCxZwZXJycjB16lQ8ffoURkZGaN++PbZt28ZdNAAUv/M///wzAGDgwIHcBZ1du3YhMDAQ/fr1Q3JyMpycnDBnzhyMGDGCl1NwcDA3Z7wyzs7OCAgI4L4PyqSmpmLv3r1YsWKFyjLnzp3D8+fPuS7yypSUS3R0NFJTUwEUfC8PHToEoKCLfWFnz57lutrv2LEDo0aNQuvWrSEUCvHdd98pvSBUnMzMTPz44494+fIl9PX1UbVqVWzfvh29e/fmyiQmJmL8+PFISkqCra0tevfurXB3Pzo6GoGBgUhOToazszN+/fVXjBs3jlfm+vXrmDFjBjIyMlC1alWsX78e/fr143omaWtrY/Xq1Rg3bhwYY6hcuTKWLl3KG9tCJpMhMDAQsbGxKs/Rt2/fIiAgoNhzdN26dZg1axb3Xj7gY0hICG/QxeDgYDRs2JC7wFBU0XP0r7/+UviZs2vXrmK/Y1+lT57Q7z9EPv/h27dvP/qzldyqsAsJGQqv8y9S2P79+9n5FymskluVz87x3zR/MsUrf/HKIibFo3iajFcWMSme+uIVnrP4U6lj7uR/Q7yIiAgGgD158uSryXHChAls+PDhaoun7vwoXtnEK4uY8njp6elMT0+PnT179qvMj+KVj3hHjx5lHh4eTCKRqDXHwr/Tiv7ulLdDU1NTPzk+3eFXo3dZ7yBjMggFQlgYKO+G9UVERgJ5eYCODlBfsyPBy+tEmv9xU7SoXe47gMkAgRDQ1dyxyXqXBSZjEAgFMLAou4GfCCGE/DeFhYXByMgIbm5uePLkCcaMGQMvLy+4urp+8hRm6vbrr79izZo1kMlkpXrWnZCSnD17Fq1ateLulhNSFjIzMxESEqLyUY+v1b8r269c7XW1uSnois7h+0V16QLExwP29sBLDeYBfp1odFq+o7U/TMvXTXN1sq72Om5avvEvlQ8sRQghhHyq9PR0TJ48Gc+fP4elpSXatGmj9FlvTTI1NcUvv/yi6TRIOdKhQwd06tRJ02mQcq5o9/5/C2rwE0IIIYSUEwMHDvysKcsIIYSUL9TgV6NvXb/Fu6x3mu3O/5WR14myOY+/KNtvC7r1a7A7PwC4fuuKrHdZ1J2fEEIIIYQQUuaowa9GIV1CNJ3CV0deJxKJ5LOnUfksjb+OY9MlpIumUyCEEEIIIYT8R9BIKYQQQgghhBBCSDlEDX5CCCGEEEIIIaQcogY/IYQQQgghhBBSDtEz/GrUb18/vM16C0sDS+zovkPT6XwV5HVirmeOPtp9NJfIpX5A7ltA1xLw0tyx2ddvH7LeZsHA0gDdd3TXWB6EEEIIIYSQ8o/u8KvR+bjzOBlzEufjzms6la+GvE4uPr+o2URenwdenSz4V4Pizsch5mQM4s7HaTQPQggh5YNAIMD+/fs1ncZHCQoKgo+Pj6bTIISQzxYYGIhJkyZpOo1iUYOfEEIIIeQLmDdvHr755hsYGxvD2toaXbt2RXR09GfFTExMRLt27dSUYdnLycnBtGnTMGPGjGLLlaauWrRoAYFAwHuNGDGCV+bFixfo2LEjDAwMYG1tjYkTJyI/P7/YbScnJ6Nfv34Qi8UwNTXFkCFDkJGRwSuze/du1KlTBwYGBnBycsKiRYt46wMCAhRyEwgEqF69OldGKpVizpw5cHV1hb6+PlxdXfHbb7+BMcaV2bdvH3x8fGBhYQGBQICoqCiFfL///nu4urrC0NAQlStXRteuXfHw4UNu/d9//w0/Pz84ODhAX18fHh4eWLFihUKc1atXw8PDA/r6+nB3d8fWrVt56yUSCWbPng1XV1fo6emhdu3aOH78OK/M2rVrUatWLYjFYojFYnh6euLYsWO8up00aRK3HUdHR4wePRqpqam8OKNHj0b9+vWhq6uLOnXqKORaGvfu3cN3330HZ2dnCAQCLF++XKHMhQsX0KlTJ9jZ2am8eFaaYxATE4Nu3brBysoKYrEYvXr1QlJSUok5xsfHo3///rCwsIC+vj5q1qyJmzdvcutnzpyJqlWrwtDQEGZmZvDx8eGtB4BHjx6hS5cusLS0hFgshre3N86ePat0e+/evUPFihUhEAiQkpLCLR80aFCJ39fS1BVjDHPnzoW9vT309fXRpk0bPH78mFemc+fOcHR0hJ6eHmxtbTFgwAAkJCTw9lm+fZFIBDMzM4hEIhgaGnJlNm/erJCrnp4ebztF665Nmza4du0ar0xJ53pcXJzSerl69SpX5ueff8auXbvw9OlTpXX+NaAGvxo9GPkAqVNS8WDkAw0n8gBITS34V8PkdfLP8H80m0jHB0DP1IJ/NWjkg5GYkjoFIx+M1GgehBBCvrzz589j5MiRuHr1Kk6dOgWJRAIfHx9kZmZ+ckwbGxvo6uqqMcuytWfPHojFYnh5eRVbrrR1NWzYMCQmJnKvhQsXcuukUil69+6NvLw8XL58GVu2bMHmzZsxffr0Yrfdr18/3Lt3D6dOncLhw4dx4cIFDB8+nFt/7Ngx9OvXDyNGjMDdu3exZs0aLFu2DKtWreLKrFixgpfXixcvYG5ujp49e3JlFi5ciODgYPzxxx948OABFixYgIULF2LlypVcmczMTHh7e2PBggUq861fvz5CQkJw79497N27F4wx+Pj4QCqVAgAiIyNhbW2N7du34969e/j1118RGBjIy3ft2rUIDAzEzJkzce/ePcyaNQs//fQTr7E+depUrF+/HitXrsT9+/cxYsQIdOvWDbdv3+bKVKxYEfPnz0dkZCRu3ryJVq1aoUuXLrh37x4AICEhAa9evcLChQtx9+5dbN68GcePH8eQIUMU9mvw4MHo3bt3sceqOFlZWahUqRLmz58PGxsbpWUyMzNRu3ZtrF69WmWcko5BZmYmfHx8IBAIEB4ejkuXLiEvLw+dOnWCTCZTGff9+/fw8vKCtrY2jh07hvv372PJkiUwMzPjylSpUgWrVq3CnTt3EBERAScnJ3Tv3h1v3rzhynTs2BH5+fkIDw9HZGQkateujY4dO+LVq1cK2xwyZAhq1aqlsHz58uUlfl9LU1eLFi3C+vXrsWbNGly7dg2Ghobw9fVFTk4OV6Zly5bYvXs3oqOjsXfvXsTExKBHjx7c+gkTJnB5xMfH4+HDh6hWrRovFwAQi8W8nJ89e8ZbX7TunJ2d0bZtW7x9+5YrU9K5Lnf69GneturXr8+ts7S0RKtWrbBu3TqV9aJxjJQoNTWVAWBv37796M9WcqvCLiRkKLzOv0hh+/fvZ+dfpLBKblU+O8e8vDy2f/9+lpeX99mxKB7F+xpiUjyKp8l4ZRGT4qkvXnZ2Nrt//z7Lzs7+5HhSqZS9f/+eSaVSteT3KfFev37NALDz58+rjJednc1GjhzJbGxsmK6uLnN0dGRz587lygFgYWFh3PtLly6x2rVrM11dXVa/fn0WFhbGALDbt28zqVTKDh06xACw48ePszp16jA9PT3WsmVLlpSUxI4ePcqqVq3KjI2NmZ+fH8vMzOTiHjt2jHl5eTETExNmbm7OOnTowB49evTR+9yhQwc2YcIEpeuKq0NlddW8eXM2ZswYlds6fPgwEwqFLCEhgVu2du1aJhaLWW5urtLP3L9/nwFgN27c4JYdO3aMCQQC9uLFC/b+/XvWp08f1qNHD97n/vjjD1axYkUmk8mUxg0LC2MCgYDFxcVxy9q3b8/69+/P29/u3buzfv36KXw+NjaWO46qyOvv9u3bDAB78uSJyrI//vgja9myJffe09NT4biMGzeONWrUiMvP1taWrVq1ildGVb6FmZmZsU2bNvFyLLzPu3fvZjo6OkwikSh8dsaMGax27doqY5f2vHNycmLLli0rtgwAtnfvXpXxVB2DEydOMKFQyFJTU7llKSkpTCAQsBMnTqiMN3nyZObt7V1sTkW9f/+eAWAnT55kjDH25s0bBoBduHCBK5OWlsYAsFOnTvE+u2bNGta8eXN25swZBoDLS1l+yr6vhRX9ucMYYzKZjNnY2LDZs2dz8VJSUpiuri77888/Ve7TgQMHmEAgUPq7QiqVsgsXLijsY0hICDMxMVEZUxl5e27//v1MKpUWe67Hx8czxkp/3q1du5ZVrFjxo/IprPDvtKK/O+V5F/5+fSy6w08IIYQQogHybszm5uYqy6xcuRIHDx7k7ojt2LEDzs7OSsumpaWhU6dOqFmzJm7duoXffvsNkydPVlp25syZWLVqFS5fvowXL16gV69eWL58OXbu3IkjR47g5MmTCneax48fj5s3b+LMmTMQCoX47rvveHcwW7RogYCAgGL3OSIiAg0aNCi2jDKq6mrHjh2wtLREjRo1EBgYiKysLG7d1atXUa1aNVSoUIFb5uvri7S0NO6Oc1FXrlyBqakpL8c2bdpAKBRy3YFzc3MVug/r6+vj5cuXCncZ5YKCgtCmTRs4OTlxy5o0aYLz58/j0aNHAAq63kdERHzWIxqZmZnYvHkzXFxc4ODgoLJcamoqry5V7dOtW7cgkUiKLRMREaF0G1KpFLt27UJmZiY8PT2LzUUsFkNL6985lnhubi4EAgGvp42enh6EQiEuXbqk8nMHDx5EgwYN0LNnT1hbW6Nu3brYuHGjyvJ5eXnYuHEjxGIxateuDQCwsLDgHr/IzMxEfn4+1q9fD2tra95d6Pv372P27NnYunUrhMKSm3/Kvq8liY2NxatXr9CiRQtumYmJCRo1aoQrV64o/UxycjJ27NiBJk2aQFtbW2mZbdu2oUqVKmjatClveUZGBpycnODg4MDrRaJMXl4eNmzYABMTE9SoUQNA6c51uc6dO8Pa2hre3t44ePCgQvx69erh5cuXiIuLU5mDJlGDnxBCCCHlw4OlQFjFglfSOf66jNgP627+pPjZ850/rC/q6eYP617sU0uqMpkMY8eOhZeXF/cHqDLPnz+Hm5sbvL294eTkBG9vb/j5+Sktu3PnTggEAmzcuBHVqlVDu3btMHHiRKVlf//9d3h5eaFu3boYMmQIzp8/j7Vr16Ju3bpo2rQpevTowXsO+LvvvkP37t1RuXJl1KlTB8HBwbhz5w7vWXFHR0fY2tqq3JeUlBSkpqbCzs6upOrhUVVXffv2xfbt23H27FkEBgZi27Zt6N+/P7f+1atXsLa25sWSN/6VdXdW9RktLS2Ym5tzn/Hx8cG+fftw5swZyGQyPHr0CEuWLAFQMKZCUQkJCTh27BiGDh3KWz558mR0794d1apVg7a2NurWrYuxY8eiX79+pa0azpo1ayAWi1GxYkUcP34cp06dgo6OjtKyly9fxl9//cXruuzr64tNmzYhMjISjDHcvHkTQUFBkEgkXBdoX19fLF26FI8fP4ZMJsOpU6ewb98+hX2+c+cOjIyMoKurixEjRiAsLAzVqlVTmsvbt2/x22+/Ke1G/W/RuHFjGBoaYvLkycjKykJmZiYmTJgAqVSq9Psg9/TpU6xduxZubm44ceIEfvjhB4wePRpbtmzhlTt8+DCMjIygp6eH5cuXIywsDJaWlgAKBu08ffo0bt++DWNjY+jp6WHp0qU4fvw492hAbm4u/Pz8sGjRIjg6Opa4P6q+ryWRnx9WVla85RUqVFA43yZPngxDQ0NYWFjg+fPnOHDggNKYOTk5CA0NxeDBg3nL3d3dERwcjAMHDmD79u2QyWRo0qQJXr58yStXuO6WLVuGEydOwMLCgsu3pHPdyMgIS5YsQWhoKI4cOQJvb2907dpVodEvf2RE1QU/TaMGvxrte7AP2//Zjn0P1PPHwCdbuhSYObPgXw2T10nYwzDNJvJiHxC7XW1/qH2qB/se4J/t/+DBPs2Pr0AIIeWOJA3Iji94SXP565j0w7q894qfzXnzYX1R+Zkf1uVnKa7/BCNHjsTdu3exa9cubtmIESNgZGQEIyMjiMViAIC/vz+ioqLg7u6O0aNH4+TJkypjRkdHo1atWry7sA0bNlRatvBzvBUqVICBgQEqVarEW/b69Wvu/ePHj+Hn54dKlSpBLBZzvQwK/4G9detWzJs3T2V+2dnZAMDL7+LFi7x93r17t8LnlNUVAAwfPhy+vr6oWbMm+vXrh61btyIsLAwxMTEqc1CHYcOGYdSoUejYsSN0dHTQuHFj9OlTMPWwsrunW7ZsgampKbp27cpbvnv3boSGhmL79u24desWtmzZgsWLFys0+EqjX79+iIyMxOHDh+Hm5oZevXrxnpuWu3v3Lrp06YIZM2bwZkqYNm0a2rVrh8aNG0NbWxtdunTBwIEDefu0YsUKuLm5oWrVqtDR0cGoUaMwaNAghX12d3dHVFQUrl27hh9++AH+/v64f/++Qi5paWno0KEDqlWrhpkzZ370Pn8trKysEBoaikOHDsHIyAgmJiZISUlBvXr1ir2bLpPJUK9ePcydOxd169bF8OHDMWzYMIVnwVu2bImoqChcvnwZvr6+GDRoEHduMsYwcuRIWFtb4+LFi7h+/Tq6du2KTp06cRcbAgMD4eHhwbsYVhxV31d1mjhxIm7fvo2TJ09CJBJh4MCBvMEq5cLCwpCRkcF9F+U8PT0xcOBA1KlTB82bN8e+fftgZWWF9evX88oVrru2bduiT58+vPEPSmJpaYnx48ejUaNG+OabbzB//nz0799fYZBOfX19AOD1MPqaUINfjUYfG40BYQMw+thozSaydCkwa9ZX0eCX18n4U+M1m8jN0cCVAQX/atCx0ccQNiAMx0YfK7kwIYSQj6MtBvTtC16iIgPZCUQf1umYKX5Wz+rD+qK0DD+s0zL47DRHjRqFw4cP4+zZs6hY8UOPgtmzZyMqKgpRUVG4desWgIKuorGxsfjtt9+QnZ2NXr168Qa4+lSFu88KBAKF7rQCgYDXXb9Tp05ITk7Gxo0bce3aNa7Lq7y7d2nIRzl///7DBZcGDRrw9rlod3ZVdaVMo0aNAABPnjwBUHDXrfBFCwDcyOmqBnFT9pn8/HwkJydznxEIBFiwYAEyMjLw7NkzvHr1iruwUviiCVDQIAsODsaAAQMU7rhPnjwZY8eORZ8+fVCzZk0MGDAA48aNK/aiiSomJiZwc3ODl5cXQkND8fDhQ4SF8W+23L9/H61bt8bw4cMxdepU3jp9fX0EBwcjKysLcXFxeP78OZycnGBsbMzdsbWyssL+/fuRmZmJZ8+e4eHDhzAyMlLYZx0dHVSuXBn169fHvHnzULt2bYVZAdLT09G2bVsYGxsjLCxMZXfufwsfHx/ExMTg9evXePv2LbZt24b4+HiFuinM1tZWoeeDh4cHnj9/zlsmn32hcePG2LRpE7S0tBAcHAwACA8Px+HDh7Fr1y54eXmhXr16WLNmDfT19bkLR+Hh4QgNDYWWlha0tLTQunVrAAWN2aIXWor7vpZEfn4UbVAnJSUpnG+WlpaoUqUKvv32W+zatQtHjx7ljXwvFxwcDF9fX95jOcrIe8jIz325wnUXFBQELS0tbNu2jcu3pHNdmUaNGilsR/4zrWjvhq/Fv/NhGUIIIYSQojzGF7yUMXIBuhXq7ll09Ozmis9lcioFFLw+E2MMP/30E8LCwnDu3Dm4uLjw1ltbW3NdTGUyGdLS0gAUjEbdu3dv9O7dGz169EDbtm2RnJys8Dy7u7s7tm/fjtzcXO554hs3bnx23u/evUN0dDQ2btzIPUer6rnt4ujo6KBatWq4f/8+d3dZX18flStXBsDf55LqShn5dGnyxwoaN26MuXPn4vXr19wf8KdOnYJYLFbZxdzT0xMpKSmIjIzknoEODw+HTCbjLijIiUQi2NsXXCD6888/4enpqfAH//nz5/HkyROlo9BnZWUp3AEWiUTFjuxeGowxMMaQm/uhl8u9e/fQqlUr+Pv7Y86cOSo/q62tzV1Y2b17N3x8fBRy1NPTg729PSQSCfbu3YtevXoVm49MJuPlkpaWht69e0NXVxcHDx5UGBfg30ze1T48PByvX79Gp06dVJb18vJSmGry0aNHJT43X7g+5XeUix4joVDIfY/27t3L9a4BCn4mDB48GBcvXlQ4r4r7vpbExcUFNjY2OH/+PDcLR1paGtfTo7j9AcD7jgAFYwKcPXsWO3fuLHHbUqkUd+7cQfv27YstJ5PJkJeXB+DjzvXCoqKiFB5devDgAbS1tXnTGH5NqMGvRr+3+h2ZeZkw1DEsufB/hLxO9ER6wMuSy5eZ2r8XdMnU0uyxafV7K+Rl5kHH8OOumhJCCPn3GzlyJHbu3IkDBw7A2NiYe07UxMSE6xJa1LJly2BnZ4e6detCKBQiNDQUNjY2MDU1VSjbt29f/Prrrxg+fDimTJmC58+fY/HixQAK7kp/KjMzM1hYWGDDhg2wtbXF8+fPMWXKFIVyAwcOhL29fbF3qH19fREREYGxY8cWu82S6iomJgY7d+5E+/btYWFhgX/++Qfjxo1Ds2bNuMcVfHx84O7ujoEDB2LRokV49eoVpk6dipEjR3IXRK5fv46BAwfizJkzsLe3h4eHB9q2bct1rZZIJBg1ahT69OkDOzs7pKWl4e3bt9i3bx9atGiBnJwchISEIDQ0FOfPn1fYj6CgIDRq1EjpOA0dO3bE0qVLUaVKFdSsWRO3b9/G0qVLec8rJycn4/nz59w85fIGoo2NDWxsbPD06VP89ddf3Dzx0dHRWLVqFfT19bnGz927d9GqVSv4+vpi/PjxXF2KRCLuAsWjR49w/fp1NGrUCO/fv8fSpUtx9+5d3jgO165dQ3x8POrUqYP4+HjMnDkTMpkMkyZN4soEBgaiXbt2cHR0RHp6Onbu3Ilz587hxIkTAAoagN999x1yc3Oxfft2pKWlcRd5rKysIBKJABT00sjIyMCrV6+QnZ3NXcypVq1aqe885+XlcY8S5OXlIT4+HlFRUTAyMuIuMmVkZPDu1sbFxeHOnTtwcHDgHlsp6RgAQEhICDw8PGBlZYUrV65gzJgxGDduHNzd3bn9a926Nbp164ZRo0YBAMaNG4cmTZpg7ty56NWrF65fv44NGzZgw4YNAAoGYJwzZw46d+4MW1tbvH37FqtWrUJiYiLXy8fT0xNmZmbw9/fH9OnToa+vj40bNyI2NhYdOnQAALi6uvLqRT4mg4eHB8RiMZcfUPz3tWhdxcbGIioqCubm5nB0dIRAIMCYMWMwf/581KhRA66urpg2bRrs7Oy4xwOuXbuGGzduwNvbG2ZmZoiJicG0adPg6uqqMLBjcHAwbG1t8e233yrkMnv2bDRu3BiVK1dGSkoKFi1ahGfPnnHjDiiru9WrVyM+Ph5dunTh9r+4cx0oeLxBR0cHdevWBQDs27cPwcHB2LRpEy+fK1euoGnTpip/jmvcJ4/v/x/yr5uWz96eMaDgX3XEUwOKV77jlUVMikfxNBmvLGJSPPXF+7dOywdA6SskJERlvHXr1rE6deowQ0NDJhaLWevWrdmtW7d4MYtOy1erVi2mo6PD6tevz3bu3MkAsIcPH/Km5Xv//j33GWVTXBWdDu3UqVPMw8OD6erqslq1arFz584xAGz79u3cPjdv3pz5+/sXW0/37t1j+vr6LCUlpdg6LKmunj9/zpo1a8bMzc2Zrq4uq1y5Mps4cSJv6iqpVMr+/vtv1rZtW6avr88sLS3Zzz//zJv+7ezZswwAi42N5Za9e/eO+fn5MSMjIyYWi9mgQYNYeno6l19SUhJr3LgxMzQ0ZAYGBqx169bs6tWrCvuTkpLC9PX12YYNG5TWRUpKChsxYgRzdHRkenp6rFKlSuzXX3/lTRkYEhKitB5mzJjBGGMsPj6etWvXjllbWzNtbW1mZ2fH/Pz82MOHD7kYM2bMUBrDycmJK3P//n1Wp04dpq+vz8RiMevSpQu7f/8+7zt97tw57jtgYWHBBgwYwE1fJjd48GDm5OTEdHR0mJWVFWvdujU3hRxjjJsSTtmr8DFo3rx5iWUAsKCgoBKn0Sv6at68ucLxL/oaOHBgqY8BYwVT7FWoUIFpa2szNzc3tmTJEiaTyXjfaScnJ95nGGPs0KFDrEaNGkxXV5dVrVqV913Jzs5m3bp1Y3Z2dkxHR4fZ2tqyTp06sTNnzvD298aNG8zHx4eZm5szY2Nj1rhxY3b06FGF+ii6z0Wn5Svp+6qqrgqf8/n5+WzixImsQoUKTFdXl7Vu3ZpFR0dz6//55x/WsmVL7rx1dnZmI0aMYC9fvuRtSyqVsooVK7LAwEClx3fs2LHM0dGR6ejosAoVKrD27dvzfi4qq7vOnTuzq1ev8uKpOtflNm/ezDw8PJiBgQETi8WsYcOGLDQ0VCFXNzc3tmPHDpV1XpKynpaPGvylQA3+z0fxyne8sohJ8SieJuOVRUyKp754/9YGvybibd++nWlra7OsrKyvJscePXqwuXPnqi2euvOjeGUTT50xnz59yrS0tNjDhw+/6n2meOU7HmOMHT58mLm7u/Mu1H2ssm7w06B9hBBCCCHlxNatWxEREYHY2Fjs378fkydPRq9evb6qrqaLFi2CkZGRptMg/2JHjx7F8OHD4ebmpulUyH9cZmYmVq1aBS2tr/dJ+a83M0IIIYQQ8lFevXqF6dOn49WrV7C1tUXPnj2LHaRNE5ydnfHTTz9pOg3yLzZy5EgA+OwBDgn5XD169OCNg/A1oga/GlVcWhHx6fGwN7bHy/GaHKHu61G4Tla7rtZcImEVC+ZP1rfnj9L8hS2tuBTp8ekwtjfG+JcanqqQEEJIuTNp0iTeIGqEEEL+26hL/xeUK5UhLl2CXCnTdCqEEEIIIYQQQso5usOvRvVs68HBxAFWBlZK1ydmSRGfmY+sfBk8xKIyTKQe4OAAWCnP40uS14mFvoVmEzGvB+Q4AHqarRPberYwcTCBgZWBRvMghBBCCCGElH/U4Fejg34Hi10vkRXc2X+fK0Oa5NPnwy05keLz+JLkdSKRSHD06FHNJdL866gTv4N+mk6BEEIIIYQQ8h9BXfq/oMI9+V9k0iAjhBBCCCGEEELKDjX4vyCZ7EOLP1XCkKlrosFsCCGEEEIIIYSUZ9Tg/4Lkd/h1hQXd+d+YOIIxGsCPEEIIIYQQQoj6UYNfjX46+hP67+uPn44qn1tW+v/GvYORFoQAsvXESJGUQYO/c2fA07PgXw2T18nYE2M1m8jNn4DL/Qv+1aCjPx3Fvv77cPQnDY5nQAghpNwQCATYv3+/ptP4KEFBQfDx8dF0GoQQUuaOHz+OOnXqQCbT3OPc1OBXo7CHYdhxZwfCHoYpXS+/w6+vJYCNfkHVPy+LZ/lv3QKuXi34V8PkdXLg0QHNJvIiDIjbUfCvBj0Me4g7O+7gYdhDjeZBCCHky1u7di1q1aoFsVgMsVgMT09PHDt27LNiJiYmol27dmrKsOzl5ORg2rRpmDFjRrHlZs6cCYFAwHtVrVqVVyYmJgbdunWDlZUVxGIxevXqhaSkJG59XFwcfvrpJ7i6ukJfXx+urq6YMWMG8vLyit12ixYtFLY9YsQIbv27d+/Qtm1b2NnZQVdXFw4ODhg1ahTS0tK4MomJiejbty+qVKkCoVCIsWPHKt3W2rVr4eHhAX19fTg4OGDcuHHIycnh1kulUkybNg0uLi7cPvz2228qe4iOGzcOIpEIy5cv5y3v3LkzHB0doaenB1tbWwwYMAAJCQm8MidOnEDjxo1hbGwMKysr9OjRA8+fP+fWR0REwMvLCxYWFtDX10fVqlWxbNkyXozSHLexY8fCzc0N+vr6sLKyQpcuXfDw4Ye/i0pTvwEBARAIBBCJRDAzM4NIJIJAIED16tWV1osqGzZsQIsWLSAWiyEQCJCSkqKybG5uLurUqQOBQICoqCjeOsYYFi9ejCpVqkBXVxf29vaYM2cOt37fvn349ttvue+qp6cnTpw4UWJ+//zzD5o2bQo9PT04ODhg0aJFvPXKvqsCgQAdOnQAUDBo9uTJk1GzZk0YGhrCzs4OAwcOVDj2lSpVUogxf/58pTk9efIExsbGMDU15S3fuHEjmjdvDmdnZ1hYWKBNmza4fv06r8y+ffvg4+MDCwsLpfWYnJyMn376Ce7u7tDX14ezszMmT56M1NRUXrnRo0ejfv360NXVRZ06dRRyVPY9FAgEMDY2VrpPu3btgkAgQNeuXRXiVK1aFYaGhjAzM0ObNm1w7do1XpkuXbqgRo0aMDAwUHputW3bFtra2tixY4fSbX8J1OD/guR3+EUCAewNhBDIpMjMZ3Bq0FTDmRFCCCGkrFWsWBHz589HZGQkbt68iVatWqFLly64d+/eJ8e0sbGBrq6uGrMsW3v27IFYLIaXl1eJZatXr47ExETuFRERwa3LzMyEj48PBAIBwsPDcenSJeTl5aFTp07cnbSHDx9CJpNh7dq1uHfvHpYtW4Z169bhl19+KXHbw4YN42174cKF3DqhUIguXbrg4MGDePToETZv3ozTp0/zLgrk5ubCysoKU6dORe3atZVuY+fOnZg1axamTZuGBw8eICgoCH/99RcvvwULFmDt2rVYtWoVHjx4gAULFmDhwoVYuXKlQrywsDDcvHkTdnZ2CutatmyJ3bt3Izo6Gnv37kVMTAx69OjBrY+NjUWXLl3QqlUrREVF4cSJE3j37h0GDBjAlTE0NMSoUaNw4cIFPHjwAFOnTsXUqVOxYcMG3raKO24AUKdOHQQFBeHBgwc4ceIEGGPw8fGBVCotdf2uWLECiYmJiI+Px8OHD/Hs2TOYm5ujZ8+eSutalaysLLRt27ZU34lJkyYprVsAGDNmDDZt2oTFixfj4cOHOHjwIBo2bMitv3jxIr799lscPXoUkZGRaNmyJTp16oTbt2+r3F5aWhp8fHzg5OSEyMhILFq0CLNmzcLmzZu5Mvv27ePV9d27dyESibh6yMrKwq1btzBt2jTcunUL+/btQ3R0NDor6QU8e/ZsXqyfflLsFSuRSODn54emTRXbLufOnUOfPn1w6NAhXLp0CQ4ODvDx8UF8fDxXJjMzE97e3liwYIHSfU5ISEBCQgIWL16Mu3fvIjg4GGfOnMHQoUMVyg4ePBi9e/dWGmfChAm8fUlMTES1atV433m5uLg4TJgwQek+ValSBatWrcKdO3cQEREBZ2dn+Pj44M2bN1yZli1bIiQkBA8ePFB6bgEFF6j++OMPpbl+EYyUKDU1lQFgb9++Lbbc0+Sn7PG7x+xp8lNuWSW3KuxCQga7kJDBFtx+w+bdesNOPk9j51+ksA1nb7F5t96wCWFXmEwm+6wc8/Ly2P79+1leXh5j9vaMAQX/qiPeZ5DXSfTraLXE++T80p8ylva44F91xPvE/JKfJrN3j9+x5KfJaomn7vy+ppgUj+JpMl5ZxKR46ouXnZ3N7t+/z7Kzsz85nlQqZe/fv2dSqVQt+X1qPDMzM7Zp0yaV8bKzs9nIkSOZjY0N09XVZY6Ojmzu3LlcOQAsLCyMe3/p0iVWu3Ztpqury+rXr8/CwsIYAHb79m0mlUrZoUOHGAB2/PhxVqdOHaanp8datmzJkpKS2NGjR1nVqlWZsbEx8/PzY5mZmVzcY8eOMS8vL2ZiYsLMzc1Zhw4d2KNHjz56nzt06MAmTJigdF3hOpwxYwarXbu2yjgnTpxgQqGQpaamcstSUlKYQCBgp06dUognt3DhQubi4lJsjs2bN2djxowpNr+iVqxYwSpWrPhR8X788UfWrFkzXrzx48czLy8v7n2HDh3Y4MGDeZ/r3r0769evH2/Zy5cvmb29Pbt8+TJzcnJiy5YtK2YPGTtw4AATCATc+RkaGsq0tLR4uezfv58JBAKWk5OjMk63bt1Y//79ufclHTdldfj3338zAOzJkycqP6eqfuXx9u7dywQCAYuLi1MZozhnz55lANi7d++UHmP5uXHv3j3ufJK7f/8+09LSYg8fPlSZn7LvTLVq1disWbNU5rRmzRpmZmbGcnNzuWWTJk1ibm5uKs+5ZcuWMWNjY5aRkaEy7vXr1xkA9uzZMy6/0nxn5Nvv378/CwkJYSYmJgrrC+9vfn4+MzY2Zlu2bFEoFxsbq1CPykilUhYSEsJ0dHSYRCJRWF/S900uKiqKAWDnzp3jHY/8/HzWpEkTtmnTJubv78+6dOlSbBx5m/D06dNK95kxxXOLMcaePXtW7He88O+0or875dss/LPuY2n0Dn9J3X5ycnIwcuRIWFhYwMjICN999x2vqxYAPH/+HB06dICBgQGsra0xceJE5Ofn88qcO3cO9erVg66uLipXrsy7MqZOLmYuqGxeGS5mLgrrGGOQD9Iv+v+gfRZp8RACsHCqjDc50jLJSdO4OjFVrJMvysgFMK5c8K8GmbmYwbyyOcxczDSaByGEEM2SSqXYtWsXMjMz4enpqbLcypUrcfDgQe7u7I4dO+Ds7Ky0bFpaGjp16oSaNWvi1q1b+O233zB58mSlZWfOnIlVq1bh8uXLePHiBXr16oXly5dj586dOHLkCE6ePMm7i5yZmYnx48fj5s2bOHPmDIRCIb777jvec6ktWrRAQEBAsfsdERGBBg0aFFtG7vHjx7Czs0OlSpXQr18/Xvfy3NxcCAQCXu8GPT09CIVChTvKhaWmpsLc3LzEbe/YsQOWlpaoUaMGAgMDkZWVpbJsQkIC9u3bh+bNm5dqv+Q8PT0RFRXFdXt++vQpjh49ivbt23NlmjRpgjNnzuDRo0cAgL///hsRERG8xzhkMhkGDBiACRMmwMPDo8TtJicnY8eOHWjSpAm0tbUBAPXr14dQKERISAikUilSU1Oxfft2tGjRgitT1O3bt3H58mWF/S7uuBWVmZmJkJAQuLi4wMHBQWmZ0tRvcHAw2rRpAycnp5J2/6MlJSVh2LBh2LZtGwwMDBTWHzp0CJUqVcLhw4fh4uICZ2dnDB06FMnJySpjymQypKenF/tdvHLlCpo1awYdHR1umY+PDx4/foz3798r/UxQUBD69OkDQ0NDlXFTU1MhEAgUuuTPnz8fFhYWqFu3LhYtWqTQngoPD0doaChWr16tMnZhWVlZkEgkpTrfipOWlgaxWAwtLa1PjrFp0yZUqVJF4S7+7NmzYW1tjSFDhpQYIy8vDxs2bICJiYnKXjvKzi0AcHR0RIUKFXDx4sVP3ofP8ek1pybVq1fH6dOnufeFD+a4ceNw5MgRhIaGwsTEBKNGjUL37t1x6dIlAAW/LDt06AAbGxtcvnwZiYmJGDhwILS1tTF37lwABV2UOnTogBEjRmDHjh1ctxBbW1v4+vp+sf2UFnrUSigAwAARk0JPBGRJgQyJDNb6XywdQgghpNy5svQKriy9AgDovr07nFs4c+vex75HSNMQAEDVblXRdkVb3mf/7PwnEm8lAgDGvxzPWxe1OQrhU8MBAO3+aAeP7iU3qlS5c+cOPD09kZOTAyMjI4SFhaFatWoqyz9//hxubm7w9vaGQCAotkGzc+dOCAQCbNy4EXp6eqhWrRri4+MxbNgwhbK///47161+yJAhCAwMRExMDCpVqgQA6NGjB86ePctdMPjuu+94nw8ODoaVlRUePnyIxo0bAyj4o9bW1lZlfikpKUhNTVXZLbqwRo0aYfPmzXB3d0diYiJmzZqFpk2b4u7duzA2Nkbjxo1haGiIyZMnY+7cuWCMYcqUKZBKpUhMTFQa88mTJ1i5ciUWL15c7Lb79u0LJycn2NnZ4Z9//sHkyZMRHR2NPXv28Mr5+fnhwIEDyM7ORqdOnbBp06YS96voduLj49GsWTMwxpCfn48RI0bwupdPmTIFaWlpqFq1KkQiEaRSKebMmYN+/fpxZRYsWAAtLS389NNPSE9PV7m9yZMnY9WqVcjKykLjxo1x+PBhbp2LiwtOnjyJXr164fvvv4dUKoWnpyf+/PNPhTgVK1bEmzdvkJ+fj5kzZ/K6W5d03OTWrl2LyZMnIzMzE+7u7jh16hSvYfsx9ZuYmIjjx49j586dxdT2p2GMISAgACNGjECDBg0QFxenUObp06d49uwZQkNDsXXrVkilUowbNw49evTgtXEKW7x4MTIyMtCrVy+V23716hVcXPg3qipUqMCts7Cw4K27fv067t69i6CgIJUxc3JyMHnyZPj5+UEsFnMX7H766SfUr18f5ubmuHz5MgIDA5GYmIilS5cCKBhXISAgANu3b4dYLFYZv7DJkyfDzs4Obdq0KVV5Zd6+fYtFixYp/RlWWjk5OdixYwemTJnCWx4REYGgoCCFcQSKOnz4MPr06YOsrCzY2tri1KlTsLS05JWZMWMGNm3apPTckrOzs8OzZ88+eT8+h8af4dfS0oKNjQ33kldgamoqgoKCsHTpUrRq1Qr169dHSEgILl++jKtXrwIATp48ifv372P79u2oU6cO2rVrh99++w2rV6/mBmRZt24dXFxcsGTJEnh4eGDUqFHo0aOHwiAjZU1WuMFfaLn2/+/2Z0o0N3IjIYQQUh7kpuUiPT4d6fHpyM/l351iUsaty3mfo/DZrDdZ3Pqi8jLzuHWSLMln5eju7o6oqChcu3YNP/zwA/z9/XH//n0AwIgRI2BkZAQjIyPuj2p/f39ERUXB3d0do0ePxsmTJ1XGjo6ORq1ataCnp8ctK/wccWG1atXi/l+hQgUYGBhwjX35stevX3PvHz9+DD8/P1SqVAlisZjrZfDy5UuuzNatWzFv3jyV+WVnZwMAL7+LFy/y9nn37t0AgHbt2qFnz56oVasWfH19cfToUaSkpHDrraysEBoaikOHDsHIyAgmJiZISUlBvXr1IBQq/nkbHx+Ptm3bomfPniU2HoYPHw5fX1/UrFkT/fr1w9atWxEWFoaYmBheuWXLluHWrVs4cOAAYmJiMH78eBURlTt37hyWLl2KVatWcc9XHzlyBL/99htXZvfu3dixYwd27tyJW7duYcuWLVi8eDG2bNkCAIiMjMSKFSuwefNmCASCYrc3ceJE3L59GydPnoRIJMLAgQO5wf9evXqFYcOGwd/fHzdu3MD58+eho6MDf39/hQECL168iJs3b2LdunVYvnw576JAScdNrm/fvrh9+zbOnz+PKlWqoFevXrzBCj+mfv/880+YmpoqDLimDitXrkR6ejoCAwNVlpHJZMjNzcXWrVvRtGlTtGjRAkFBQTh79iyio6MVysvHbti9ezesra3VlmtQUBBq1qyp8pyXSCTo1asXGGNYu3Ytb924cePQokUL1KpVCyNGjMCSJUuwcuVK5ObmAigY06Jv375o1qxZqXJZsGABdu3ahbCwMN75/jHkPZbc3d1LHOSzOGFhYUhPT4e/vz+3LD09HQMGDMDGjRsVGu9FtWzZElFRUbh8+TLatm2LXr168X42AgWDCEZGRio9t+T09fWL7SlUljR+h1/e7UdPTw+enp6YN28eHB0dERkZCYlEwrsqVLVqVTg6OuLKlSto3Lgxrly5gpo1a3JXuwDA19cXP/zwA+7du4e6deviypUrCleWfH19VY6W+jnOxZ1Dbn4udLV00cK5BW/dhwH7CqbQkX8FtP//Oykrv3w2+OV1IoJIs4kknQOkuYBIF6jQQmNpxJ2LQ35uPrR0tXh3ngghhHw+XbEujO0L7iJq6fL/xBGIBNw6PTPFP0ANrAy49UXpGOpw67QNlHdvLi0dHR1UrlwZQEE36hs3bmDFihVYv349Zs+ejQkTJgAAd+etXr16iI2NxbFjx3D69Gn06tULbdq0Ubjb/LEKdzcVCAQK3bYFAgGvu36nTp3g5OSEjRs3ws7ODjKZDDVq1IBEUvoLIPKRuQt3R27QoAF3h00mk0FfX3l3R1NTU1SpUgVPnjzhlvn4+CAmJgZv376FlpYWTE1NYWNjw7twARR0CW/VqhWaNGmiMMBcaTRq1AhAQQ8BKysrbrn8ZlXVqlVhbm6Opk2bYtq0acX2cihsxowZ6NWrF4YOHQqhUIiaNWsiMzMTw4cPx6+//gqhUIiJEydiypQp6NOnDwCgZs2aePbsGebNmwd/f39cvHgRr1+/hqOjIxdXKpXi559/xvLly3l3pC0tLWFpaYkqVarAw8MDDg4OuHr1Kjw9PbF69WqYmJjwBifcunUrnJyccO3aNTRp0oRbLr/rXLNmTSQlJWHmzJnw8/NTuo/KjhsAmJiYwMzMDG5ubmjcuDHMzMwQFhbGi1Oa+mWMYceOHejfv79CDwF1CA8Px5UrVxQGxmzQoAH69euHLVu2wNbWFlpaWqhSpQq3Xv5oxfPnz3n57tq1C0OHDkVoaGiJd75tbGwUHmWWv7exseEtz8zMxK5duzB79mylseSN/WfPniE8PLzEu/SNGjVCfn4+4uLi4O7ujvDwcBw8eJDrHcMYg0wmg5aWFjZs2IDBgwdzn125ciWWLFmC06dP8y4sfoz09HS0bdsWxsbG2L59u8rHSkpj06ZN6NixIypUqMD9TIuJiUFcXBw6derElZOv09LSQnR0NFxdXQEUDFZZuXJlVK5cGY0bN4abmxuCgoJ4F4EsLCwgFotRtWpVhXNLLjk5mffz40vSaIO/uG4/r169go6OjsLzJRUqVMCrV68AFFyNLNzYl6+XryuuTFpaGrKzs5X+YsnNzeWuaAHgpgGRSCTF/mLrv68/4tPjYW9sj9ifYgEAurq6YNJ85Od/aPAzaT6YtODOg/b/m/7pefkf9UuzKPlnJRIJtAD8/6kB5H9izMLxPoe8TuyM7LCm8prPjif3sflpXe4PQXY8mL498jvGfna8T81vX/99SI9Ph7G9MX6KVRz99GPjqTu/rykmxaN4moxXFjEpnvriSaVS7g/Owo3SRmMbodHYRtz7wutMnEww9vlY7r387os8Tu/9vZV+DgBqDayFWgNrqVyvLF5pyWQy5OTkQCaTcQ0yeZz09HQwxmBkZISePXuiZ8+e6N69O9q3b4+3b99yz8bK68HNzQ3bt29HdnY210CRTyElk8l4d5wK113Rfwvvj0wmw7t37xAdHY3169dzz8AWfk6+tPuspaWFatWq4d69e1xjR1dXl2ugF97novEyMjIQExOD/v37K6yT18Pp06fx+vVrdOzYkdvfhIQEdO3aFfXq1eO6On/sfNi3/j/NsbyRpSw/+fPO2dnZKr8fRZdnZWVBKBTy1snv0stHrJffESz8WaFQyB2/fv36oVWrVtw2MjMz0bNnT/Tv3x8BAQEq97VovpmZmVzcwtuR56IqjlQqRW5ursr1RY+bsvNEfj6rqjtl+cqdO3cOT58+xaBBgz5rnnP5Z4vmt3z5cl4jOiEhAe3atcOff/6JRo0aQSaTwdPTE/n5+Xj8+DHXSJRPMyi/ECO/MDF06FDs3LkT7dq1KzHfRo0aYdq0acjNzeUavKdOnYKbmxtMTU15n//rr7+Qm5uLvn37KsSVSCTo3bs3njx5gjNnzsDMzEzl/srdunULQqEQlpaWkMlkuHTpEvedBICDBw9i4cKFiIiIgL29PffZhQsXYtGiRTh27Bjq1aunch8L/8wpWiYtLQ3t2rWDrq4uwsLCeD/viyr8c0qZ2NhYnD17Fvv37+d9/9zd3fH333/zyk6bNg0ZGRlYtmwZb5+U5S7/ma2sDpV9V3NychATE4PatWsrjSvPTSKRcOvV+TtUow3+wgOO1KpVC40aNYKTkxN2796t8grvlzBv3jzMmjVLYfnZs2eVDtYhJ++KlJOTg6NHjxbE+v03vI88h0xdMVChJpCbhfeRH6bgyH8VB5g543HcS2TeevzZuZ86dQo+OTnQ/38eJ/+fx+fE+xzyOpFfQPnceEWVNl5p66Ss81P2HfmceJ9L3fHKIibFo3iajFcWMSne58eTPw6YkZFR4pzqJSnumWd1x5s1axbatGkDBwcHpKenY8+ePTh37hz27t3Lm2O8sPnz56NChQqoVasWhEIh/vzzT1SoUAFCoZD7THZ2NtLS0tCxY0dMnToVgwcPxtixY/Hy5UvujlxmZiYvt/T0dK5Bl5OTA8YYL4fc3FxIpVKkpaVBJBLB3Nwca9asgbGxMV6+fMn7O0ked8SIEbC1tS22+22LFi1w7tw5DBo0qNg6nDZtGtq2bQsHBwckJiZi/vz5EAqF6NChA5fnjh07UKVKFVhaWuL69esIDAzEjz/+CFtbW6SlpSEhIQGdOnWCg4MDpk+fjqdPn3LbkN8Mkl8QWLt2LerXr4/Y2Fjs2bMH3377LczNzXH37l38+uuvaNKkCXdne+/evXjz5g3q1q0LIyMjPHjwADNmzECjRo1gbm7O5Xfnzh0ABY+pJiQk4NKlS9DW1uYGqP7222+xZs0a1KpVCw0aNMDTp0+5/c7MzARQ0Ct17ty5sLS0hIeHB/755x8sXboU/fr1Q1paGrS1tXl39wFAJBLB1NSUq4ebN2/i1q1b8PT0hImJCeLi4jBnzhy4uLigevXqSEtLQ4sWLbB8+XJMnToV3333HTIyMvDbb7/BwcEBlStXRlpaGjZu3IiKFStyd7IvX76MJUuWYPjw4dw+l3Tc4uLisG/fPrRq1QoWFhZISEjA8uXLoaenh6ZNmyItLQ0nT54sVf0CwIYNG9CgQQM4OTmpPIeKk5SUhNevX+Pu3bsACi6QGRsbo2LFijAzM4OpqanCzUeg4OKPWCxGWloaGjZsiNq1ayMgIADz5s2DTCbDxIkT0bJlS+7ufnBwMH788UfMmzcP1apVw+PHBX/z6+npwcTEhNuXI0eO4MCBAwCAjh07Yvbs2fD398eYMWPw4MEDrFy5EnPmzFH4ObNx40a0b98e2travHqQSCTw9/fH33//jV27diElJQUpKSkAADMzM+jo6OD69euIjIyEt7c3jI2Ncf36dfz666/o1asXRCIR0tLSYG9vz9ueubk5BAIB991LS0vD8uXLMW/ePK6bvHwfDQ0NYWRkBAB4//49Xr58yY2zERUVhczMTFhbW3M3ZL/77jtkZWVh+/bt3Hz2SUlJsLS0hEhU0GP46dOnyMzMxPPnz5GZmcmN7+bu7s7r6bFu3TrY2NjAy8tLoV6KnjeGhobIz8+Ho6MjcnJy8O7dOyxZsgTt2rVDhQoVkJycjE2bNnGPB5X23AIKLpDq6urylhWWl5eH7OxsXLhwgbtgIP/dqZbHAD55fP8y0qBBAzZlyhR25swZBoC9f/+et97R0ZEtXbqUMcbYtGnTFKZiePr0KQPAbt26xRhjrGnTpgpToQQHBzOxWKwyh5ycHJaamsq9Xrx4wQCwxMRElpeXp/K18OJCNvX0VLbw4kJumUeNmuz8ixQW9iSZzbv1hv3xz1t2/kUKOxf3lu3fv5/tji6Yqu/PR++LjV3SKzMzk+3fv59lZmay/IULWf7UqSx/4UK1xPucvOR1Mu/cPLXE+9T88u8uZPm3p7L8u8rrRF37W1K8iwsvstNTT7OLCy9q5HiUVbx/Q44Uj+JpOibFU1+8tLQ0du/ePZaZmcmkUuknvfLz89n79+9Zfn7+J8f42HiDBg1iTk5OTEdHh1lZWbHWrVuz48ePFxtv3bp1rE6dOszQ0JCJxWLWunVrdvPmTa4cALZ3717u/cWLF1mtWrWYjo4Oq1+/Ptu+fTsDwO7fv8/y8/O5afnevXvHfSYoKIiZmJjwtj99+nRWu3Zt7v2JEyeYh4cH09XVZbVq1WLh4eEMANu+fTu3z82bN2cDBw4stp7u3LnD9PX1WXJycrF12KtXL2Zra8t0dHSYvb0969WrF3v06BGv/KRJk1iFChWYtrY2c3NzY4sXL+bVf1BQEENBh0eFl7xMTEwMA8DOnDnDpFIpi4uLY82aNWPm5uZMV1eXVa5cmU2YMIHL6/379+zUqVPM09OTmZiYMD09Pebm5sYmTZrEq1P5sSn6cnJy4tbn5OSwKVOmMFdXV6anp8ccHBzYDz/8wIuTkpLCRo8ezRwdHZmenh6rVKkS++WXX1h2drbK+nNycmJLly7llkdFRbGWLVty++Ts7My+//579vz5c97nd+zYwerWrcsMDQ2ZlZUV69SpE7t27RpXpytWrGDVq1dnBgYGTCwWs7p167JVq1YxiUTCxSjpuD1//py1adOGWVtbM21tbVaxYkXm5+fH7t+/z5U5ffp0qeo3OTmZ6evrs+XLlys97+THv7jv4/Tp05Uep6CgIKXl5d+XyMhI3vIXL16wbt26MSMjI1ahQgXm7+/P3rx5wx2TZs2aKd1O4fNl+vTpvO+HVCplt2/fZt7e3kxXV5fZ29uzuXPnKvycuX//PjfVpqp8lb3OnDnD8vPz2blz51jDhg25+vbw8GBz5sxhWVlZKutN2c8MJycnpduZPn16ieekvIy8/afsFRMTw8Vp3rx5iWUkEgmrWLEiCwwMLNXP6YEDB7LOnTtz7zMzM1nXrl2ZnZ0d09HRYba2tqxTp07s6tWrvHOrRYsWzMzMrNhza9iwYWz48OEq6zMzM5Pdu3ePpaWlKfzufPv27WdPy/dVNfjT09OZmZkZW7FiBUtJSWHa2tpsz5493PqHDx8yAOzKlSuMsYI5MYVCIUtKSuLKrF+/nonFYm7O0EmTJrEaNWrwtuPn58d8fX1LnZd8/sO3b99+9D5VcqvCLiRksL0xKWzerTdszd137EJCBjv/IoXt37+f7f//hYCgBx83L3tReXn/nvmTKV75i1cWMSkexdNkvLKISfHUF6/wnMWfSipVPT92eYq3fft2pq2tzf3x/jXk2KNHDzZ37ly1xVN3fhSvbOKVRczi4k2fPp01b978q82P4v3745Um5ps3b5i5uTl7+vSpyhiFf6cV/d0pb4d+ToNfo6P0T5gwAefPn0dcXBwuX76Mbt26QSQSwc/PDyYmJhgyZAjGjx+Ps2fPIjIyEoMGDYKnpyc3/YuPjw+qVauGAQMG4O+//8aJEycwdepUjBw5knt2bcSIEXj69CkmTZqEhw8fYs2aNdi9ezfGjRv3RfdVPi2fsMgAqvJR+rPyGQghhBBCPsfWrVsRERGB2NhY7N+/H5MnT0avXr00+qhkUYsWLeK6+BJSVo4dO8YbhJAQTYiLi8OaNWsUplj8kjT6DP/Lly/h5+eHd+/ewcrKCt7e3rh69So3guGyZcsgFArx3XffITc3F76+vlizZg33eZFIhMOHD+OHH36Ap6cnDA0N4e/vzxtcw8XFBUeOHMG4ceOwYsUKVKxYEZs2bYKvr+8X3Vf5tHyiIlOmcKP0SwoGayhpShVCCCGEEFVevXqF6dOn49WrV7C1tUXPnj0xZ84cTafF4+zsjJ9+Kv3AtYR8iuvXr2s6BULQoEEDNGjQQKM5aLTBv2vXrmLX6+npYfXq1Vi9erXKMk5OTiUOftaiRQvcvn272DJlrfC0fIXJG/wyADlSBn0tNTT409MBxgCBADBWPsUQIYQQQsqfSZMmYdKkSZpOgxBCyFdCow3+8qbW2lpIzEiErZEt/vnhH946qYo7/EKBALkZadA1EiMzXwZ9LTU8ZeHhAcTHA/b2wMuXnx/vM8jrxMbQBr/b/665RI7WArITAX1boP0/JZcvI2trrUVGYgaMbI3wwz8/aCwPQgghhBBCSPlHDX41Ss5Oxtust9AV6Sqsk8qU3+EHgOy0lIIGv0QGS72yzvLLKq5OvqjcZCD3LSDUbB7ZydnIepsFka5Io3kQQkh5wBiNf0MIIeTfrax/l1GDX42cTZ2hp6UHGyMbhXXcHX4lN/CzU5NhaudYLgfuk9dJBcMKmk3EyBkQ6QH6isfmSzJ1NoWWnhaMbGiwIkII+VTyuZjz8vK+qsHoCCGEkI+Vl5cH4MPvNnWjBr8aRQyOULlO/gy/UMmgfNmp7wEAmfmysklMg+R1IpFIShxroUx9q/rYfEmDIwZrOgVCCPnX09LSgoGBAd68eQNtbW0IhR//OJxMJkNeXh5ycnI+6fP/tnhlEZPiUTxNx6R4FE+T8dQRUyaT4c2bNzAwMICWlhby8/PVkldh1OD/Qj48w6+4Ljs1GUDBSP2EEEIIKZ5AIICtrS1iY2Px7NmzT4rBGEN2djb09fXVMkPO1x6vLGJSPIqn6ZgUj+JpMp66YgqFQjg6OpbZbG3U4P9CZNwo/f+tO/yEEEJIWdDR0YGbmxvXFfJjSSQSXLhwAc2aNYO2tvZn5/O1xyuLmBSP4mk6JsWjeJqMp66YOjo6autxoAw1+L+Q4u7w53AN/vL3DD8hhBBSVoRCIfT0Pm20W5FIhPz8fOjp6anlD7+vPV5ZxKR4FE/TMSkexdNkvLKKqW7U4FejWedmITU3FSa6JpjRYgZvnfT/N++V3+Evv1365XVipG2E+qivuUTuzALyUgEdE6DmjJLLl5Fzs84hNzUXuia6aDGjhcbyIIQQQgghhJR/1OBXo423NiI+PR72xvaKDX55l35lo/Snld8u/YXrpL6rBhv8TzYC2fGAvr1GG/y3Nt5Cenw6jO2NqcFPCCGEEEIIKVNl97AA4Sl20L6U/9/hL4cNfkIIIYQQQgghmkF3+NXoQJ8DyJPmQUekw1suYwzyp/OLG7RPIgPypAw6yq4KfFQiB4C8PEBHp+SyZUxeJ0ImxKvbrzSXSPMDgDQPEGm2Tvoc6ANpnhQinbKZZ5MQQgghhBBC5KjBr0b17ZR3WZcWGotPqKQtL8nJgpYAyGcFd/l1RJ/ZGKyvwa7zRcjrRCKR4Ojto5pLxPzrqBO7+naaToEQQgghhBDyH0Fd+r8A+ZR8AgBCFfMrGmgXHIry+Bw/IYQQQgghhJAvjxr8X8CHEfpVlzHU+n+DvxyO1E8IIYQQQggh5MujLv1qdCfpDiQyCbSF2qhZoSa3nBuwT1l//v8z0CpYl5XPVJYptcOHgexsQF8f6Njx8+N9BnmdQNPXMVLuADIJINQGTGuWXL6MJN1Jgkwig1BbiAo1K2gsD0IIIYQQQkj5Rw1+NWq3ox03Bd3L8S+55dyUfMXd4Vdnl/4RI4D4eMDeHnj5suTyZahwnax2Xa25RM62+zAtXzfN1cmOdju4afnGvxyvsTwIIYQQQggh5R916f8CPkzJp7rFL+/ST1PzEUIIIYQQQghRB7rDr0b+tf2RkpMCUz1T3vLS3OE3KKfP8MvrxFjHGMjSYCKV/IG8FEDHVINJALX9ayMnJQd6pnoazYMQQgghhBBS/lGDX43mtJ6jdPmHO/yqP8sN2lfO7vDL60QikeDoUQ1Oy1db+bH50lrPaa3pFAghhBBCCCH/EdSl/wuQ3+FXNSUfABhoq3HQPkIIIYQQQggh/3nU4P8CuGn5iqltmpaPEEIIIYQQQog6UYP/C5Bxz/CXPGhfjpRxPQIIIYQQQgghhJBPRc/wq1GrLa2QlJmECoYVEO4fzi0vzTP8+loCCAAwFIzUb6wtKtNcvxR5nVgbWGOs2VjNJXKmFZCTBOhVAFqHl1y+jGxptQWZSZkwrGAI/3B/jeVBCCGEEEIIKf+owa9Gj949Qnx6PFJzUnnLSzMtn0AggIGWAJn5DJkSBmPtssz0y+HqxDgVMNNgImmPgOx4IC+15LJl6N2jd0iPT0dOao5G8yCEEEIIIYSUf9SlX42MdIxgrGMMIx0j3vLSTMsHfJiaL+tzR+o3MgKMjQv+1TB5nRhqG2o2EW0jQMu44F8N0jHSgY6xDnSMdDSaByGEEEIIIaT8ozv8avRw1EOly7lB+0po8BtqC/EmR/r5A/c9VJ6HJsjrROPT8nX8Oupk1MNRmk6BEEIIIYQQ8h9Bd/i/AG5aPmHxLX5Ddd3hJ4QQQgghhBDyn0cN/i+gNIP2AYCBVkGBzHwapZ8QQgghhBBCyOehBv8XUJpp+YCCLv0APr9LPyGEEEIIIYSQ/zx6hl+NVl1fhfTcdBjrGmNUww/Papf+Dr+auvRPnAi8fw+YmQGLFn1erM8krxMDLQM4w1lziUSvAvLTCwbuc9fcc/TXV11HbnoudI110XBUQ43lQQghhBBCCCn/qMGvRvMj5iM+PR72xvYqGvyle4Y/83Mb/H/+CcTHA/b2Gm/wF66T1a6rNZfI/fkF0/Lp22u0wR8xPwLp8ekwtjemBj8hhBBCCCGkTFGX/jKmpavH/b80o/QDQBY9w08IIYQQQggh5DPRHX41Cu4SjJz8HOhpfWjka+vpc/8vYZB+btC+LIkMjDEISugR8G8grxMtaEHyUKK5RBoHA9IcQKRXctky1CW4C/Jz8qGlR6ceIYQQQgghpGxRq0ONfFx9FJZp6RkAKGjsl9SAlz/DLwOQI2XQ1/r3N/jldSKRSHD04VHNJWKreGw0wdXHVdMpEEIIIYQQQv4jqEt/GdP+f4O/pO78AKAlFEBXJJ+aj0bqJ4QQQgghhBDy6ajBX8bkXfpLGrBPjhu4j6bmI4QQQgghhBDyGahLvxolpidCyqQQCUSwNbYFAGjrl/4OP1DwHH9ybvkZuE9eJzKphi9gZCcCTAoIRIC+rcbSSE9MB5MyCEQCGNsaaywPQgghhBBCSPlHDX41+mbjN9wUdC/HvwTwCXf4tdU0Nd9XonCdaHRavuPffJiWr9tLjaWx8ZuN3LR841+O11gehBBCCCGEkPKPuvSXMe4Z/lLWtLxLfxZ16SeEEEIIIYQQ8hnoDr8adXDrgOScZJjrmXPLSnOHPynpFVyruAMA6n43CA16DUXQjl0YtHEhV6aCtTUuR1wsZSIdgORkwNy85LJlTF4npjqmmk3EvgOQmwzoarZO3Dq4ISc5B3rmmp0ekBBCCCGEEFL+UYNfjdZ3Wq+wTLvQtHyqyGQMm8/fAgC8yspHTJoEXh26YWj/PlyZgOb1PiIRxTw0RV4nEokER49qcFq+hl9HnXRa30nTKRBCCCGEEEL+I6hLfxn72EH7tP9/ZUAiKx+D9hFCCCGEEEII0Qxq8Jexjx207/9j9oEe4SeEEEIIIYQQ8jmowV/GuEH7SnmHX+v/d/jz6Q4/IYQQQgghhJDPQM/wq1HP0J54k/kGVoZWCO0ZCgDQkt/hL+4h/kK0/t8TIJ8BjDEIStkzgKdBA+DVK+B/7N15eFTl3f/xz5k12yQQlgABBVEQFRdQKxYqiIKKloKPrYp1rVYftGpabW19rFp93KrWBUWtij6Ktvxqq1VcKAhqxaqI+4I7koSwJ5NltjPn98dkhrDD5EzuLO/XdXFNMjnzPZ+ciZfXd+773HefPtLbb+/6612UviY98nvo9LzTzQV59SQpuloK9pLGzDEWY85Jc9SwukGFvQp10pyTjOUAAAAA0PnR8Lto8XeLM3vOpwV28R5+X4s5F7Yj+bLo97VypVRZmcUL3dfympw+2GDDv2ax1FQp5Zfv+Ngc+m7xdwpXhhUqDxnNAQAAAKDzY0p/jvl28R5+j2VlVvRn4T4AAAAAQLYY4XfRl7/4covnNi7at/N1fJalmOMo0Qn6/fQ1icfjmv/SfHNBfrjle2PCL778hekIAAAAALoIGn4XBX3BLZ7buGjfznf8fo8US3aOhfvS18TjGJ5M4t3yvTHBF+Q/OQAAAABtgyn9OZZu+HdyzT5JrNQPAAAAAGg9Gv4c8+/ion3SxoX6EskcBAIAAAAAdAnML3bR7A9mqzHeqAJ/gU4dfqqSjiNfIDWVfGe35ZNajPA7HX+EP31NAp6ASlRiLsg3s6VEo+QrkAaeaizGB7M/ULwxLn+BX8NPHW4sBwAAAIDOj4bfRZfPuzyzBd2pw09VzN7YsO/SCH9zw98ZVulveU1mDJ5hLsjSyzduy2ew4Z93+bzMtnw0/AAAAAByiSn9ORRrbtgtpbbb21lM6QcAAAAAtBYj/C66+eibM1P6pY0j9Lsyui+5MKX/5pulxkapoCC717sofU0CnoD0rcEgB928cUq/QUfffHRmSj8AAAAA5BINv4tOHb7pVPH0lP5d2ZJParlKf7ZBzE1Z31z6msTjcc39dq65IAan8bfENH4AAAAAbYUp/TmUntLv2cWrvHFKf8e/hx8AAAAAYAYNfw7FspzS7+9Eq/QDAAAAAMxgSr+Loolo5uugL9iKKf2px0RSchxH1i6+Xp99JiUSks8nDR26a691WfqaxBNxozlkb3xv5A0ai5GIJjJf+4L85wcAAAAgd+g4XDT4zsGZLehWVKzIeoTf19zgO5KSzq6/XuPHS5WVUnm5tGLFLr7YXS2vidFt+Z4ZvHFbvinmrsmdg+/MbMtXsaLCWA4AAAAAnR9T+nMo2xF+j5Xayk9iWj8AAAAAIDuM8Lto1IBRWt2wWr0Ke0mS4s2r7O/qCL1lWfJ5Uq+PJ6Wg1+WgbSh9TXrk9zAbpOcoKbpaCvYyGmPAqAFqWN2gwl6FRnMAAAAA6Pxo+F0056Q5m3yfmdLv2dU5+amt+eJJp8Ov1J++JvF4XHPnGtyWb8ycHR/TBk6ac5LpCAAAAAC6CKb051B6Sn8W/b78VnqlfjcTAQAAAAC6Chr+HMp20T6p5Ur9dPwAAAAAgF1Hw59DGxv+LKb0p0f4afgBAAAAAFngHn4X/fyfP9e6yDqV5pXqvhPua7FK/67Xyozwd/B+P31NugW66Xgdby7Imz+XouukYKl06H3GYvzz5/9UZF1EeaV5OuG+E4zlAAAAAND50fC76LnPn8vsOS9J8daM8Hs6xwh/y2ty/GCDDX/lc1JTpZRfbi6DpM+f+1zhyrBC5SGjOQAAAAB0fkzpz6HW3cPfORp+AAAAAIAZjPC76K1z35Lt2PJaXkkbV+nPalu+5pfEs+n333pLsm3J683ixe5KX5OkndTSV5aaC3LMW5JjS5bZa3LuW+fKsR1Z2XwKBAAAAAC7gIbfRX1DfTf5Pj3Cn822fK0a4e/bd8fHtJH0NYnH41oqgw1/fvu4JqG+TOUHAAAA0DaY0p8jjuO0WLSvNffwuxoLAAAAANBF0PDniO1I6bH5bGZv+5tfk3C4hx8AAAAAsOuY0u+il758SZFERHm+PI3e7ajM861ZtC/pSMldbfrvv1+qr5eKiqTzztv1k7sofU18pv/Uql+S7IjkzZP6TjAW48uXvlQikpAvz6fBEwYbywEAAACg86Phd9HZT5+d2YLuw+nfSpLikSZZVv4u12r5IcEuT+u/9lqpslIqLzfe8Le8JjMGzzAX5I2zN27LN2WFsRhPn/10Zlu+ihUVxnIAAAAA6PyY0p8j6QX74pGmrF5vWVZmpX625gMAAAAA7CpG+F30m9G/UTgaVigYUry5SU9Es2v4pdS0/oTtdOj7+NPXpMBXIK0yGGSf30iJsOQzu0r+6N+MVjQcVTAUNJoDAAAAQOdHw++iCw+9MPP18nBckmQn4lnX83kk2VK8A6/Un74m8Xhcc+fONRdk6IU7PqYNHHrhoaYjAAAAAOgimNKfI3bzqHyyNQ2/ld6ar+OO8AMAAAAAzKDhzxG7uUe3460Z4W9u+DvwlH4AAAAAgBk0/Dnixgi/v/nd2eVV+gEAAAAAXR738Lto77v3VlW4Sv1C/fTUtPckSclEIut6nWFKf/qa9C3qq5t3u9lckGf3lhqrpIJ+0vGfGotx9953K1wVVqhfSBd+2j7WFQAAAADQOTHC76L6WL3CsbDqY/Wym5t0Ox7Lut7GKf2uxDMifU0a4g1mg8TrU6v0x+uNxojVxxQLxxSrz/7vAgAAAAB2BiP8LhrSY4hK8kpUVli28R7+1ozwZ6b072LHP2SIVFIilZVlfW63pK9J74LeZoMUD5ECJVKe2WvSY0gP5ZXkqbCs0GgOAAAAAJ0fDb+LFpyxIPP1ktVNkqSk3fpV+uO72vAvWLDjY9pI+poY35ZvfPu4JmcsOMN0BAAAAABdBFP6c8SdVfpTjx15Sj8AAAAAwAwa/hxJ38PfmlX6M/fwd+BF+wAAAAAAZtDw50h6hN+NVfptR7Is3ioAAAAAwM7jHn4X/W7+77QhskHd8rppwrDfSmrtKv0bvw4UFu38C6dNk9askXr2lB5/POvzuyF9TUKBkEZplLkg7/1Oim2QAt2kA643FmP+7+YrsiGivG55Gn/9eGM5AAAAAHR+NPwueuS9R1QZrlR5qFzj925u+Fsxwu+xLHmt1Ah/sKh451+4aJFUWSmVl2d9bre0vCajBhts+L96RGqqlPLLjTb87z3ynsKVYYXKQzT8AAAAAHKKeeI5Yjutv4df2jitf5cafgAAAABAl8cIv4uen/a84sm4/B6/KqOp5+zWNvweKZrsuA1/+pooKX339nfmgox7XkrGJY/fXAZJ056fpmQ8KY+fz9oAAAAA5BYNv4uGlw3PfL3827AkF0b4PZYkp8M2/OlrEo/H9Z0MNvzdhu/4mDZQNrzMdAQAAAAAXQTDjDmS3knPjrvR8Et5HbThBwAAAACYQcOfIwnX7uFPPQaLQq2NBAAAAADoQpjS76IlVUsUs2MKeAOynb0ktf4efn/zCH+wsGOO8Keviccx/NnSuiWSHZO8Aal0pLEYVUuqZMdseQNe9RvZz1gOAAAAAJ1fuxnhv/HGG2VZli655JLMc5FIRNOnT1ePHj1UVFSkE088UTU1NZu8bvny5Zo0aZIKCgrUu3dvXXbZZUpsthXewoULNWLECAWDQe25556aNWtWTn6HyU9O1uEPHa7JT05WMunSCH/zO9RR7+FPX5MT/9+JZoMsmizNOzz1aNCTk5/UQ4c/pCcnP2k0BwAAAIDOr100/G+99Zbuu+8+7b///ps8f+mll+qf//yn5syZo0WLFqmqqkpTp07N/Ny2bU2aNEmxWEyvv/66HnnkEc2aNUtXXXVV5pivv/5akyZN0rhx4/Tuu+/qkksu0c9+9jO9+OKLOf2dEul7+N3alq+QKf0AAAAAgJ1nfEp/fX29pk2bpgceeEDXXXdd5vna2lo9+OCDmj17to488khJ0sMPP6xhw4bpjTfe0GGHHaaXXnpJH3/8sf71r3+prKxMBx54oP7whz/o17/+ta6++moFAgHNnDlTgwYN0q233ipJGjZsmF577TXdfvvtmjhxoqu/y7kjzlVttFYlwRIlM/fwJ3bwqu1LL9q3SyP8554r1dZKJSWtOrcb0tekyF8khQ0G2fNcKVYrBcxekxHnjlC0NqpgSdBoDgAAAACdn/GGf/r06Zo0aZKOOuqoTRr+JUuWKB6P66ijjso8t/fee2u33XbT4sWLddhhh2nx4sUaPny4yso2bnU2ceJEXXDBBfroo4900EEHafHixZvUSB/T8taBzUWjUUWj0cz3dXV1klJby8W3s+r+b7//28zXj31ZL0nyyJFjb9n0p59z7ITy8/O2eowkeZ2kJCkvVLLdc6d/Fo/Hpd/+tuUPtvma7dmkXiukr0k8Hte8efNaXS9tl/Ptvf1r4tbvu6N63//t97c4pjX1suV2vVzUpB71TNbLRU3qUc9kvVzUpB71TNekHvVM1stFzc3ruVHXcpzmoWgDnnzySV1//fV66623lJeXp7Fjx+rAAw/Un/70J82ePVtnnXXWJo23JB166KEaN26cbrrpJp133nn69ttvN5me39jYqMLCQs2dO1fHHnushgwZorPOOktXXHFF5pi5c+dq0qRJamxsVH5+/ha5rr76al1zzTVbPD979mwVFBTs1O/2VZ8DFQ0UasCqD1UUqd3ZS7KFiL9AX/c9SF47piGVb2VdBwAAAADQcTQ2NurUU09VbW2tiouzW9PN2Aj/d999p4svvljz5s1TXl6eqRhbdcUVV6iioiLzfV1dnQYMGKBx48apR48eO1Xj4c/DikaT+vMDf9Y1f5qxxc8dO6EN776mbgeO1tRDhunv73y+1TpR29HX6xKKy6tjjz1WVvM9/ZtLj6AfffTR8vv9O5Vxe6hHPdM1qUc9k/VyUZN61DNZLxc1qUc90zWpRz2T9XJRc/N66ZnmrWGs4V+yZIlWrVqlESNGZJ6zbVuvvPKK7r77br344ouKxWLasGGDunXrljmmpqZGffr0kST16dNHb7755iZ106v4tzxm85X9a2pqVFxcvNXRfUkKBoMKBre8x9rv9+/0G2k3z5uINDXK8m77Mlten5qaIts8xu9xJCXk8XrleH0KeLe/zuKuZNwZ1KOe6ZrUo57JermoST3qmayXi5rUo57pmtSjnsl6uaiZrudGTWOr9I8fP14ffPCB3n333cy/gw8+WNOmTct87ff7NX/+/MxrPvvsMy1fvlyjRo2SJI0aNUoffPCBVq1alTlm3rx5Ki4u1j777JM5pmWN9DHpGm4a/dBo7Xnnnhr90GglXVql32tZmTepKbGTd1/07y9ZVurRsPQ1GfvoWLNB5o2Wntkz9WjQQ6Mf0p173qmHRj9kNAcAAACAzs/YCH8oFNJ+++23yXOFhYXq0aNH5vlzzjlHFRUVKi0tVXFxsS666CKNGjVKhx12mCRpwoQJ2mefffTTn/5UN998s1auXKkrr7xS06dPz4zQn3/++br77rt1+eWX6+yzz9aCBQv017/+Vc8995zrv9M3G75RZbhSkUREdmaV/tYvtODzSLGkFLGNLbeQtZbXRD0NBqn/RmqqlOyIwRDShm82KFwZViLSut0bAAAAAGBHjK/Svz233367PB6PTjzxREWjUU2cOFH33HNP5uder1fPPvusLrjgAo0aNUqFhYU644wzdO2112aOGTRokJ577jldeumluuOOO9S/f3/9+c9/dn1LPkkqzS9V1I6qNL9UdmpxfdkurKzo81iKJR1FEslW12pr6WvSPa+72SDBUikZTT0alF+aLztqK79067eTAAAAAIBb2lXDv3Dhwk2+z8vL04wZMzRjxpaL3qXtvvvumjt37nbrjh07VkuXLnUj4na9f8H7ma//+O4aSS6N8FuWJEdNHXCEP31N4vH4Dt+nnDru/R0f0wYueP8C0xEAAAAAdBHG7uHv7NK9uTsj/KnHiN3xRvgBAAAAAGbQ8OdA0nGUHot35x7+1FZ8kZ1dtA8AAAAA0OXR8OdAy5n3dqL1i7OlpvSrQ07pBwAAAACY0a7u4e/oblt8m+qidcr3h6SCMyRJdiLW6rqZKf0dcNG+9DUp9BVqiIaYC/LJbVK8TvIXS8MqjMVYfNtiReuiChYHNarC/a0hAQAAACCNht9Fty2+TZXhSvULleuicamG37HtVtdNT+nviCP86WtSHirXjMHbXnwx5z69LbUtX3658YY/XBlWqDxEww8AAAAgp5jSn0Ney506/uY6LNoHAAAAANhZjPC76LGpjymaiCqW9OmjBslrudPx7/KifY89JkWjUjDoyvlbI31NvPKq4aMGc0EOf0yyo5LX7DWZ+thUJaIJ+YL8pwcAAAAgt+g6XDR24FhJ0pqmhD76dINrI/y7PKV/7Fh3TuyC9DWJx+Oa+9Fcc0HKxpo7dwsDxw40HQEAAABAF8GU/hxI9+WujfCnp/R3wEX7AAAAAABm0PDngO2kOn6vS1c3PcKfcKR4suMt3AcAAAAAaHtM6XfR1+u/lu3YqmlKSip1bYTfa0lJOyGP16dIIil/wLv9FyxcuPEefsPT+9PXJGl6wcH6ryXHliyvVDTIWIz1X6+XYzuyvJa6D+puLAcAAACAzo+G30VjHh6jynCl+hSV6+Ij33XtHn7LshRtCCu/uLsitqPQjl5w2mlSZaVUXi6tWOFOiCylr4nxbfnmjdm4Ld8Uc9fk4TEPZ7blq1hhbntAAAAAAJ0fU/pzID3p3q0RfkmK1ocl7cLCfQAAAACALo0RfhdN2XuK1kfWy/IUS3LvHn5JitbXSep4C/elr0lJoESyDQYZMEWKrZcCZqfR7z1lb0XWR5TXPc9oDgAAAACdHw2/i+467i5J0ifro3r6m7DLI/yphr+jjfCnr0k8HtfcuQa35Tv4LnPnbuG4u44zHQEAAABAF8GU/hzIrNLvXr/fYUf4AQAAAABm0PDnQHoQ3pOLhr+DjfADAAAAAMyg4c8BO5lqyn0udvzRhuYp/QkafgAAAADAjnEPv4t++MQPtbpxtfzeUh13wCM5WaU/Yno/+12UviY98nvo3MJzzQVZ9EMpslrK6yUd8YyxGE/88Ak1rm5UQa8CnfLMKcZyAAAAAOj8aPhd9E71O6oMV6pXYT8dJ3en9EfqO+YIf/qalIfKpcEGg6x7R2qqlPLLDYaQqt+pVrgyrFB5yGgOAAAAAJ0fU/pzyJeDVfo72gg/AAAAAMAMRvhdtKJihSTplaoGvV7TlJNF+3ZqW74VK9w7cSulr4nxbfmmtI9rUrGiwnQEAAAAAF0EI/w5kO7J3dyWL9aQuoc/2sGm9AMAAAAAzKDhzwHbcX+V/vQ9/NGkk6kPAAAAAMC20PDnQHqE380p/bGG+szXjPIDAAAAAHaEe/hdNOvdWWqINeiLsFdlpf/l6rZ8TtJW0GspajtqspMq8G/ns5prrpFqa6WSEun3v3ctQzbS1yTPm6ee6mkuyFezpESD5CuU9jjTWIx3Z72rWENMgcKADjzzQGM5AAAAAHR+NPwuunLBlaoMV6pHQT/96qj/cvUefknKa274IztauO+BB6TKSqm83HjDn74m5aFyzRg8w1yQ967cuC2fwYZ/wZULMtvy0fADAAAAyCWm9OeQ1805/ZLyvam3q4kp/QAAAACAHWCE30V3HnunGuONem9d6nufi1P6JSnPl6oXsZOu1s2l9DUJWAHpK4NBDr5TSjRKvgKDIaRj7zxW8ca4/AV+ozkAAAAAdH40/C6aOmyqJCnwZa2+qou7umifJOU33yPQkUb409ckHo9r7ldzzQUZMNXcuVsYNnWY6QgAAAAAugim9OdAegDe9Xv4fam3qyON8AMAAAAAzKDhzwHbSY3Au38Pf8cb4QcAAAAAmMGUfheFo2E5clQfC0vKd32EP+hN38PfcRr+9DVJxBNmg8TDkhxJluQPGYsRDUczMYKhoLEcAAAAADo/Gn4XDZsxTJXhSnXP76vLj35fXpcX7cvvgFP609fE+LZ8zw7buC3flBXGYswYNiOzLV/FigpjOQAAAAB0fkzpz4H0+Lvr9/CnR/iZ0g8AAAAA2AFG+F10xMAjtKZxjdbHiiUpZyP8TTsa4T/iCGnNGqlnT1fPn430NSnNKzUbpPcRUnSNFDR7TQYeMVCNaxpV0NPs9oAAAAAAOj8afhc9PvVxSdI9H65TXTwpr8vzJ3Z6hP/xx909cSukr0k8HtfcuQa35ft++7gmUx9vH9sDAgAAAOj8mNKfA4n0Kv05u4ffkeMwrR8AAAAAsG00/DmQXkQ/V/fwO5KiHWilfgAAAABA26Phz4Fkjkb4fR5L/uZ3rCNtzQcAAAAAaHvcw++is54+S2sb16qqqVD/ddBdrt/DL0l5Xo/iyaSa7KS6ybv1g448UqqpkcrKpAUL3A+xC9LXpHted031GLx//Y2zpOhaKdhDOuxhYzGePutpNa5tVEGPAk1+eLKxHAAAAAA6Pxp+F837cp4qw5Uqzusryf0Rfik1rT8c38HCfcuWSZWVUm2t6+ffVelrUh4q19TBBhv+6nlSU6WUX24ug6Qv532pcGVYofKQ0RwAAAAAOj+m9OeQ2/fwSy235mNKPwAAAABg2xjhd9F757+nxkRC93+SGlnP1Qi/JEUSSddr58J757+npJOUnbD1n4X/MRfkuPckJylZZj/jOv+98+UkHVmeHHwaBAAAAAAt0PC7qEdBD+XHkyoMpO6tz8UIf56vueHvICP8PQp6SJLi8bjZIMEeZs/frKBHgekIAAAAALoIpvS7zG5eod9jSVYORvjzm1cCbOogI/wAAAAAADNo+F2WHnjPxei+1GJKfwcZ4QcAAAAAmMGUfhc9u+xZrWqs1ycrkzqofGJOztHRFu17dtmzaoo3yW/5ZcngfeuVz0p2k+TNl8qPNxZj2bPLFG+Ky5/v15DjhxjLAQAAAKDzo+F30fnPnp/Zlu/g/rlp+Dvaon3pa1IeKteMwTPMBXnz/I3b8k1ZYSzGs+c/m9mWr2JFhbEcAAAAADo/pvTnSC5W6Jc63qJ9AAAAAAAzGOF30VVHXKUV4Vp9uN4rb44+StmpRfuuukqqr5eKinITYhdcdcRVqo/VK9+bL1UbDDL8KileL/nNXpMjrjpCsfqYAkUBozkAAAAAdH40/C46b+R5+iYc05Nf1LXJCL/jOFvfCeC883Jy7mycNzKVJR6Pa271XHNB9mwf12TkeSNNRwAAAADQRTCl32XJHK/Snx7htx0p3jFu4wcAAAAAGEDD7zLbSXX8uRrh93s2vmkRm44fAAAAALB1TOl3WboHz9U9/JZlKc9nqTHhKGI7Kt7aQdXVkm1LXq/Ut29uggAAAAAA2jUafhftccceWlFXqaK8vrpnyns5O0++16PGhL3thfsOOUSqrJTKy6UV5ragk1LXJL0t3+0DbzcX5Ok9Nm7LN/krYzHu2OOOzLZ8F391sbEcAAAAADo/pvS7KGbHFE/GZCdjObuHX+pYW/PF7Fjmn1HJ2MZ/BtkxO/MPAAAAAHKJEX4X7dd7P4WCPWVbpTm7h1+S8po/TYgk2n/Dv1/v/dS7sLd6FfQyG6TbflJe79Q/g3rv11uFvQtV2LvQaA4AAAAAnR8Nv4teOO0Fvb2qSf+qbMjpCH++LzUxo6kDLNr3wmkvSGrelm+uwW35xr1g7twtnPbCaaYjAAAAAOgimNLvsswq/R5G+AEAAAAA5tDwuyx9Wz0j/AAAAAAAk2j4XZYZ4W+Le/g7wKJ9AAAAAAAzuIffRZe9dJk+XrtaYbtIh4y9MWfnyfc2j/B3gCn9l710mdZH1qskUKIf6Afmgiy9TIqtlwLdpYNuMRbjpcteUmR9RHnd8zThlgnGcgAAAADo/Gj4XfTEh0+oMlyp4ry+8lo35ew8G7fla/9T+tPXpDxUrh8MNtjwf/OE1FQp5Zcbbfg/fOJDhSvDCpWHaPgBAAAA5BRT+nPEk8Mry6J9AAAAAIAdYYTfRfNPn6+XK2v1eV1Svhzew59etG+b9/DPny8lEpLP/Ns7//T5SiQTcpKOvvzPl+aCjJ8vJROSx+w1OX3+6UomkvL4+KwNAAAAQG6Z7wg7kaE9h+rzhrA2JKM5XaU/PcIfSzqyk86WWwAOHZq7k++ioT1TWeLxuL6UwYa/uH1ck55De5qOAAAAAKCLYJjRZXYy96v0B1t8msBK/QAAAACAraHhd1m6//bm8Mp6LCszyt+YaP8L9wEAAAAA2h5T+l20+LvFer96rWoiXnl2G5fTcxX6PYrYthoSSfXa/IezZ0uNjVJBgXTqqTnNsSOLv1usqB2V1/EazaHVi6VkVPIEpV6jjMX4bvF3sqO2vEGvBowaYCwHAAAAgM6Pht9FJ805KbMt34UHfZXTcxX4LK2V1Li1lfovv1yqrJTKy403/OlrUh4q14zBM8wFee2kjdvyTVlhLMack+ZktuWrWFFhLAcAAACAzo8p/TmSy3v4JamweZX3hjhT+gEAAAAAW2KE30XTD5muf1evUkKF2nzh/NaqqVmpwUM2rjR/+FmXat9j/kt/vOsevf2X+xUMBnXDdX/Q0ROP0UJ3T90q0w+ZrrponQr9hdIGg0GGTJfidZK/2GAI6ZDphyhaF1WwOGg0BwAAAIDOj4bfRVeMuUIPf7peNU22fC53/Mmko1mL3sl8/119XMvrE5o47Wxd+N/ny7ETWr9koVatXu3qeVvrijFXSEptyzd37lxzQfa9wty5WxhzxRjTEQAAAAB0EUzpd1l6lX63R/g3528+QTzJtnwAAAAAgC3R8LvMdlINeK7v4afhBwAAAABsDw2/y+zmNfR8OR/hTz2yZh8AAAAAYGuyuof/q6++0h577OF2lg7v4PsP1hfrq1QU7K2zhr6d03MF0iP8tiPHab+j/Afff7BW1q9UWWGZrupzlbkgLxwsNa2U8vtIx+T2vdme+w++X/Ur61XUp0jnvX2esRwAAAAAOr+sRvj33HNPjRs3To899pgikYjbmTqslfUrVRupVji6St4cz51IT+lPauO6Ae3RyvqVqgxXqqahxmyQppVSU2Xq0aD6lfUKV4ZVv7LeaA4AAAAAnV9Wbek777yj/fffXxUVFerTp49+/vOf680333Q7W4fTp6iPSvL6KhTsLV+O7+H3eqzMwoBb3Mffp49UXp56NKxPUR+Vh8pVVlhmNkh+Hym/PPVoUFGfIoXKQyrqU2Q0BwAAAIDOL6sp/QceeKDuuOMO3XrrrXrmmWc0a9YsjR49WkOGDNHZZ5+tn/70p+rVq5fbWdu9t897Wze/u0ZJJ/er9EupUf6o7SiedJTnbRnE3JT1zb19XiqL8W35DE7jb4lp/AAAAADaSqsmnvt8Pk2dOlVz5szRTTfdpC+++EK/+tWvNGDAAJ1++umqrq52K2eH4DiO0oPtuR7hl1i4DwAAAACwba1q+N9++23993//t/r27avbbrtNv/rVr/Tll19q3rx5qqqq0uTJk93K2SG0vJfe0wb7H7A1HwAAAABgW7Ka0n/bbbfp4Ycf1meffabjjjtOjz76qI477jh5mrvcQYMGadasWRo4cKCbWds9u8Vq+d42GOEPbNLwt8E9BAAAAACADiOrhv/ee+/V2WefrTPPPFN9+/bd6jG9e/fWgw8+2KpwHc1Nr92oBStqlOcP6fID/5Dz86VH+GObN/w//7m0bp1UWirdd1/Oc2zPDa/eoLponQr9hRqu4eaCfHSDFK+T/MXSvlcYi/HqDa8qWhdVsDioMVeMMZYDAAAAQOeXVcP/+eef7/CYQCCgM844I5vyHdZ9S+5RVbhSxXl95bGuy/n5Nk7p3+wHzz0nVVamVuo3bMZbM1QZrlR5qFwzBs8wF2TZjNS2fPnlRhv+t2a8pXBlWKHyEA0/AAAAgJzK6k7zhx9+WHPmzNni+Tlz5uiRRx5pdaiOrq0m12cW7bO5hx8AAAAAsKmsRvhvuOEG3beVqeK9e/fWeeed1+VG9tMemvyk/vbVWuX7gm1yvo6waN+ck+Yoakfldbxa9/46c0FGz5GSUcnTNu/Ntpw05yTZUVveoHfHBwMAAABAK2TV8C9fvlyDBg3a4vndd99dy5cvb3Wojuqgfodpad0GFfjaZow/4G3/Df+oAaMkSfF4XHPfn2suSK9R5s7dwoBRA0xHAAAAANBFZDWlv3fv3nr//fe3eP69995Tjx49Wh2qo0r33W2xQr+0cYQ/4UhJp/02/QAAAACAtpdVw3/KKafoF7/4hV5++WXZti3btrVgwQJdfPHFOvnkk93O2GEkmjt+bxvdxN9yIsEWC/cBAAAAALq0rKb0/+EPf9A333yj8ePHy+dLlUgmkzr99NP1v//7v64G7EiWrf1UNXW1iucFJB2a8/NZliW/J9Xst9eG/7M1nymRTMgxfdtB3WdSMiF5fFLxUGMx1ny2RslEUh6fRz2H9jSWAwAAAEDnl1XDHwgE9Je//EV/+MMf9N577yk/P1/Dhw/X7rvv7na+DmXa/ztWK+sr1T2/r/7ne1Vtcs6Ax1I86SjeTqf0j390fPvYlm/++I3b8k1ZYSzGo+MfzWzLV7GiwlgOAAAAAJ1fVg1/2pAhQzRkyBC3snR4G1vuttqYL30fv6N4UvK32VkBAAAAAO1dVg2/bduaNWuW5s+fr1WrVimZ3HQ++YIFC1wJ19FMGvJjvb9mlXoXdG+zc7bcmi/T8J9yirR+vdS97XJsyyn7naL1kfUqCZRIUYNBBp4ixdZLAbPXZL9T9lNkfUR53fOM5gAAAADQ+WXV8F988cWaNWuWJk2apP32209WlqvS33vvvbr33nv1zTffSJL23XdfXXXVVTr22GMlSZFIRL/85S/15JNPKhqNauLEibrnnntUVlaWqbF8+XJdcMEFevnll1VUVKQzzjhDN9xwQ2ZtAUlauHChKioq9NFHH2nAgAG68sordeaZZ2aVeXt++f0b9My3Ye1e1HZj7Rsb/hZP3nJLm51/R26ZkMoSj8c1d67BbfkOah/XZMItE0xHAAAAANBFZNXwP/nkk/rrX/+q4447rlUn79+/v2688UbttddechxHjzzyiCZPnqylS5dq33331aWXXqrnnntOc+bMUUlJiS688EJNnTpV//73vyWlZhpMmjRJffr00euvv67q6mqdfvrp8vv9mcUDv/76a02aNEnnn3++Hn/8cc2fP18/+9nP1LdvX02cOLFV+TeXaL6P3pvV3gfZ8TefK256UTwAAAAAQLuS9aJ9e+65Z6tPfsIJJ2zy/fXXX697771Xb7zxhvr3768HH3xQs2fP1pFHHilJevjhhzVs2DC98cYbOuyww/TSSy/p448/1r/+9S+VlZXpwAMP1B/+8Af9+te/1tVXX61AIKCZM2dq0KBBuvXWWyVJw4YN02uvvabbb7/d9YY/3XN7s5zxkI2tjvADAAAAALq8rBr+X/7yl7rjjjt09913Zz2df3O2bWvOnDlqaGjQqFGjtGTJEsXjcR111FGZY/bee2/ttttuWrx4sQ477DAtXrxYw4cP32SK/8SJE3XBBRfoo48+0kEHHaTFixdvUiN9zCWXXLLNLNFoVNHoxhvO6+rqJKWmpcfj8W2+LpZISJIsx8kcFwwG5diJLY5NP+fYCeXn5231mJa2dYxfqU4/1vxpQzAY3G7GnZWu4UYt6nX+ermoST3qmayXi5rUo57JermoST3qma5JPeqZrJeLmpvXc6Ou5Ti7vp/blClT9PLLL6u0tFT77ruv/P5N71l/6qmndrrWBx98oFGjRikSiaioqEizZ8/Wcccdp9mzZ+uss87apPGWpEMPPVTjxo3TTTfdpPPOO0/ffvutXnzxxczPGxsbVVhYqLlz5+rYY4/VkCFDdNZZZ+mKK67IHDN37lxNmjRJjY2Nys/P3yLT1VdfrWuuuWaL52fPnq2CgoJt/i5Xfnuj1iUbVOIt1g27XbbT16A1mvyF+qbvgfIlYtqr6i1J0pHTpytv3TpFSku1YIbBrfAkXfPlNapN1KrEV6LfD/69sRyHRa5R0KlV1CrRG3nmcnx5zZdK1CbkK/Fp8O8HG8sBAAAAoH1rbGzUqaeeqtraWhUXF2dVI6sR/m7dumnKlClZnXBzQ4cO1bvvvqva2lr9v//3/3TGGWdo0aJFrtTO1hVXXKGKio17pNfV1WnAgAEaN26cevTosc3XnXvHf2t1Q5WiBf0y6xscMGKkZj7/6hbHOnZCG959Td0OHK2phwzT39/5fLuZpozYa6vHFNiOvlmXkO0LyJF0zR+u0w8lWU1N8klZr7MQj8c1b948HX300Vt8oLMrpt81XZVNlepX1E+SWl0v23y+Z6fLaqqUk1++1Wvi1u+7o3p3Tb9LTZVNCpWHdum9aat87akm9ahnsl4ualKPeibr5aIm9ahnuib1qGeyXi5qbl4vPdO8NbJq+B9++OFWnzit5XoAI0eO1FtvvaU77rhDP/nJTxSLxbRhwwZ169Ytc3xNTY369OkjSerTp4/efPPNTerV1NRkfpZ+TD/X8pji4uKtju5LqWnxwWBwi+f9fv9230ifJyCvJyC/J5A5LhqNyvJu+zJbXp+amiLbPUbSNo8JeBxJCTmSbI8vdb507ebMrbGj33lHAt6AAt6Agr6gK/U2t9P1PAHJE5DV4r1pVb0s83kD3sy/bM5j7PoZrEk96pmsl4ua1KOeyXq5qEk96pmuST3qmayXi5rpem7UzKrhl6REIqGFCxfqyy+/1KmnnqpQKKSqqioVFxerqKgo60DJZFLRaFQjR46U3+/X/PnzdeKJJ0qSPvvsMy1fvlyjRo2SJI0aNUrXX3+9Vq1apd69e0uS5s2bp+LiYu2zzz6ZYzbfDm7evHmZGm564uQPtbimSQf3ars91j2WJa8l2Y5ke9puO8Cd9dXFX0lKfVpldFu+yV+ZO3cLF391sekIAAAAALqIrBr+b7/9Vsccc4yWL1+uaDSqo48+WqFQSDfddJOi0ahmzpy5U3WuuOIKHXvssdptt90UDoc1e/ZsLVy4UC+++KJKSkp0zjnnqKKiQqWlpSouLtZFF12kUaNG6bDDDpMkTZgwQfvss49++tOf6uabb9bKlSt15ZVXavr06ZkR+vPPP1933323Lr/8cp199tlasGCB/vrXv+q5557L5lffLtvAKv1SaqV+23aU8La/hh8AAAAAYEZWDf/FF1+sgw8+WO+9994m97RPmTJF55577k7XWbVqlU4//XRVV1erpKRE+++/v1588UUdffTRkqTbb79dHo9HJ554oqLRqCZOnKh77rkn83qv16tnn31WF1xwgUaNGqXCwkKdccYZuvbaazPHDBo0SM8995wuvfRS3XHHHerfv7/+/Oc/u74lnyQlmlfK97Ztv6+Ax1LEdmTT8AMAAAAAmmXV8L/66qt6/fXXFQgENnl+4MCBqqys3Ok6Dz744HZ/npeXpxkzZmjGdlaa33333Xc4VXzs2LFaunTpTufKVjI9wu9p6xH+1GPCE9j+gQAAAACALiOrhj+ZTMq27S2eX7FihUKhUKtDdVTPffqQvqqr1YYN3fX9Phe22Xn9zVMK2uOU/vuX3K/6WL3yvfnqp37mgnxxvxSvl/xF0p7nGYux5P4litXHFCgKaOR5I43lAAAAAND5ZdXwT5gwQX/60590//33S5Isy1J9fb1+//vfZ70NXGfwxHs3aW1jld4u7Kcbx7Vhw988o6A9Tum/dtG1qgxXqjxUrhmDtz1TI+c+uFZqqpTyy402/IuuXaRwZVih8hANPwAAAICcyqrhv/XWWzVx4kTts88+ikQiOvXUU/X555+rZ8+eeuKJJ9zO2GE4hs6bbvgT7XCVfgAAAACAGVk1/P3799d7772nJ598Uu+//77q6+t1zjnnaNq0advc274rOO/Q27W8PqzDyrq36XkDm4/wz5wpNTVJ7eC9mHn8TDXFm+S3/NLnBoMcOlOymySv2Wty/MzjFW+Ky5/PhzMAAAAAciurhl+SfD6fTjvtNDezdHgHlk9USTiucbsXtel5Ny7a19xEHn98m55/e44fksoSj8c19/PtL66YU+Xt45oMOX6I6QgAAAAAuoisGv5HH310uz8//fTTswrT0dnpVfqttl6lP71oH6v0AwAAAABSsmr4L7744k2+j8fjamxsVCAQUEFBQRdu+FMdv7dt+/1Mw+94vPIF8tr25AAAAACAdimrhn/9+vVbPPf555/rggsu0GWXXdbqUB1VbdM61UdjqovEJfVts/N6LckjKSkpr7ibtGSJFItJgYA00uxK8Gsb1yrpJGUnttzGsU1F10pOUrI8UrCHsRiNaxvlJB1ZHksFPQqM5QAAAADQ+WV9D//m9tprL91444067bTT9Omnn7pVtkO56sXRWt9UrQeLylX9yxVtdl7LsuTzSLFkc8M/ebJUWSmVl0sr2i7H1hww84D2sS3f3AM2bss3xdw1mXnAzMy2fBUrKozlAAAAAND5edws5vP5VFVV5WbJDsXUtnzSxpX680pKDaYAAAAAALQXWY3wP/PMM5t87ziOqqurdffdd+v73/++K8E6or17j1VtZJ2G9yxr83OnV+rPK+7W5ufenqMHH621jWvVPa9ttyrcQt+jU9P6DU7nl6TBRw9W49pGpvMDAAAAyLmsGv4f/ehHm3xvWZZ69eqlI488UrfeeqsbuTqkaQffrfp4UmcO7dbm5043/PnFhhvrzTw8+WFJzdvyzTW4Ld9hD5s7dwuTH55sOgIAAACALiKrhj+ZTLqdo1Owk6lJ/b42XqVfkvyWJclRXkn7avgBAAAAAGa4eg9/V2c338Tv9bR9x99ep/QDAAAAAMzIaoS/omLnVxe/7bbbsjlFh2Q7qY7fa2KEv/lDhvY2pR8AAAAAYEZWDf/SpUu1dOlSxeNxDR06VJK0bNkyeb1ejRgxInOcZRnofA1xHEePv32+GmNrtfTLvvrLf81u0/MH0vfwd+/ZpufdkWlPTdOaxjUqzSvVyf6TzQX59zQpukYK9pS+/7ixGE9Ne0qNaxpV0LNAUx+faiwHAAAAgM4vq4b/hBNOUCgU0iOPPKLu3VMjyuvXr9dZZ52lMWPG6Je//KWrITuCpKSv176uuki1GiLlbX7+QPO0gsLSXnIktZePWhZ9s0iV4UqVh8p18mCDDf+qRVJTpZTf9u9NS98s+kbhyrBC5SGjOQAAAAB0flndw3/rrbfqhhtuyDT7ktS9e3ddd911XXaVfrvFOoYmmu1g8zvpC+YZODsAAAAAoL3JaoS/rq5Oq1ev3uL51atXKxwOtzpUR2Q7jiqOfF2O4+jS/dt+r3ePZclrx2R7A1q95AP1zvdK7eCWik+mfyJHjhLxhF6d/6q5IMd/IrWDuQ/TP5neHmIAAAAA6AKyavinTJmis846S7feeqsOPfRQSdJ//vMfXXbZZZo6tWvel2w7UtBXJEkqDpqZru1vbvhr8wrUuzhoJMPmQs3XIu6Jmw3ibx9T6IOh9vG+AAAAAOj8smr4Z86cqV/96lc69dRTFY+nGjmfz6dzzjlHt9xyi6sBO4r0Cv0+y9xihb5EVAoUKRxL7vhgAAAAAECnllXDX1BQoHvuuUe33HKLvvzyS0nS4MGDVVhY6Gq4jiR9D7/X4DR6vx2VJNXFafgBAAAAoKvLquFPq66uVnV1tX7wgx8oPz9fjuN0qa34WrIdRx9WPStLTXoqUKapw9r+1gZ/IiZJ6nXPnZIvKhUXSxUVbZ6jpac+eUqN8UYFrICCMjid/bunpESj5CuQBpi77eSTpz5RvDEuf4Ffw6YOM5YDAAAAQOeXVcO/du1a/fjHP9bLL78sy7L0+eefa4899tA555yj7t27d8mV+m1H+ueHv1VdpFovf1ZupOH3NY/wD/rz3VJNlVRebrzh/8Xzv8hsyzdj8AxzQd7+xcZt+Qw2/M//4vnMtnw0/AAAAAByKatt+S699FL5/X4tX75cBQUFmed/8pOf6IUXXnAtXEeSvoffJH8i1fAnZT4LAAAAAMCsrEb4X3rpJb344ovq37//Js/vtdde+vbbb10J1tHYjjRh7yvksyKauFsvIxnS9/C3g88eMq478jo1xBqU582TVhgMcsB1UqJB8pldZ+LI645UrCGmQGHAaA4AAAAAnV9WDX9DQ8MmI/tp69atUzDYNbcds5OORu52inrleXXmsO5GMvjsmJJJu12N75954JmSpHg8rrkr5poLsseZ5s7dwoFnHmg6AgAAAIAuIqsp/WPGjNGjjz6a+d6yLCWTSd18880aN26ca+E6Eru5y/Z6zC1aaElq2rDO2PkBAAAAAO1HViP8N998s8aPH6+3335bsVhMl19+uT766COtW7dO//73v93O2CGk7+H3Gd6koHHdKrMBAAAAAADtQlYj/Pvtt5+WLVum0aNHa/LkyWpoaNDUqVO1dOlSDR482O2MHUJ6hN9jeFvChrWrjZ4fAAAAANA+7PIIfzwe1zHHHKOZM2fqd7/7XS4ydUh20tENL+2vuki1Lg+Va0WFmRXqGtrZCH//2/q3j235/t5/47Z8U8ytHnhb/9sy2/JVrDC7ZSIAAACAzm2XR/j9fr/ef//9XGTp0JLtZKW8xnWM8AMAAAAAsryH/7TTTtODDz6oG2+80e08HVbCcVResr/6FPXX0NI+xnI0rF2lmr33V7Rvf/XazVyOtBF9R2hAyQD1yO9hNkjpCCkyQMozs2ViWt8RfVUyoEQFvbbc5QIAAAAA3JRVw59IJPTQQw/pX//6l0aOHKnCwk33Nr/ttttcCdeR2I50+vce0z7dg/rhwJCxHA3rVutvf3pMIb9H0/crNZYj7ZlTnpHUvC3fXIPb8h3xjLlzt3DKM6eYjgAAAACgi9ilhv+rr77SwIED9eGHH2rEiBGSpGXLlm1yjGV40TpTks2r9HsN//oNa2skSfXxpJKOY3wRQQAAAACAGbvU8O+1116qrq7Wyy+/LEn6yU9+ojvvvFNlZWU5CdeRJJKpR6/hBjtSt0EeK7WmQDieVEnAazQPAAAAAMCMXVq0z3E2XZnu+eefV0NDg6uBOqrMCH9WGx26yHEU8qdChGNJw2EAAAAAAKZkdQ9/2uYfAHRltiM988Fv9C9PvZ4p7aW7jrvLWJYTfjFN1uo1Ki4vk55/1lgOSbpo7kVaH1mvkkCJJmiCuSBvXyTF1kuB7tLB5t6buRfNVWR9RHnd83TcXccZywEAAACg89ulht+yrC3u0e+q9+xvLuE4+qh6ruoi1fp0VbnRhr/Hx+8rf2WVomv6GcuQ9vdP/67KcKXKQ+WaMNhgw//d36WmSim/3GjD/+nfP1W4MqxQeYiGHwAAAEBO7VLD7ziOzjzzTAWDQUlSJBLR+eefv8Uq/U899ZR7CTuIZDua7JC+q4AJGAAAAADQde1Sw3/GGWds8v1pp53mapiOLJF09PPv/1OH9Aro4N5FRrOkJ120hzv4Xz3rVdmOraSd1Cevf2IuyNGvSo4tWWYXMTzr1bPk2I4s09s5AAAAAOj0dqnhf/jhh3OVo8NLOlJp4e7avXuhBnXPN5olfZtFe1hjYVD3QZKkeDyuT2Sw4S8aZO7cLXQf1N10BAAAAABdhOk15TsNO71KfzsYuGVKPwAAAACAht8liebm2tsOFjH0tJjSn2hPiwsAAAAAANpMq7blw0bJpKOv1vxbbzgebagPaezAsaYjSZLC8aS6B83dt77wm4WKJqLyyuy986pZKNlRyRuUysYai/HNwm+UiCbkC/o0cOxAYzkAAAAAdH40/C6xHekv71ygByLVKg+Va0XFCmNZWs4xqIvZRhv+0546LbMt34zBM4zl0OunbdyWb4q59+ap057KbMtXsaLCWA4AAAAAnR9T+l1it9Mb5uti7WGtfgAAAABAW2OE3yW2I40efIEGF8U0sMTwSuwVFVq2fLW+8xYoP2624a8YVaG6aJ0KfYXSWoNB9q6Q4nWSv9hgCGlUxShF66IKFgeN5gAAAADQ+dHwu8R2HI0ZfIF+MrhYg4oDZsNUVKimukFvrWzSgYZH+CtGpaatx+NxzZ0711yQYe1j+vyoilGmIwAAAADoIpjS7xK7ua9uD6v0S1KxP3XffjhuG04CAAAAADCBht8l6Xv4ve3kioYCqSDcww8AAAAAXVM7aU87Prt5zb52McIfDqukqV6BhnrVGb6HHwAAAABgBg2/SxJJR3e8fIT2u7uv9r93f7Nhhg1Tjz499LMTD1fUdhS1zTX9+9+7v3rd0ksjHhhhLIMkae7+0t96pR4Nunf/e3VLr1t07/73Gs0BAAAAoPNj0T6X2I7UGF+vusga5fnaxwrs6bkG4VhSwXwzn+2sa1qnNY1rFPQavibRdVJ0jeQxm6NpXZMa1zTKG/QazQEAAACg86Phd0nCcdS9YIB65BWoX6iP6TiSJE9zx18XT6pnvpkMA7sNVJ4vT2WFZWYCpBUNlLx5Ur7Z96bbwG7y5flU1KfIaA4AAAAAnR8Nvwscx1HSkc4f/Zx+sV+pCvzt404Jq3mMP2xw4b7Xzn5NUjvYlu/o18ydu4WzXzvbdAQAAAAAXUT76Ew7uISz8ev2skq/tHGEf0OMrfkAAAAAoKtpR+1px2UnN3b8vvawSn+zTMMfpeEHAAAAgK6Ght8FdosRfo/hfn/1qhpV19RIkmo3bJAkvbLkfQ0eMjTz7/DRYwwmBAAAAAC0Be7hd0HCSXX8Cz67RatWxVUSLNHvx/7eSJak46i0V2+pukqhkm6SpD577q2HFy6R1Tz74Mwj2m6LvGsWXqPaaK2K/EUaqZFtdt4tfHCNFKuVAiXScDPvjSQtvGahorVRBUuCGvv7scZyAAAAAOj8aPhdkN7m/s1v/0/zPqtWeajcWMPfUnqyge2k1hnwG5h98MA7D6gyXKnyULlGDjbY8H/xgNRUKeWXG23433ngHYUrwwqVh2j4AQAAAOQUU/pb4fDRYzR4yFBNmHS8JCmZTHX+K2tqMtPna2pWmQtoSYHmdzjScmVBAAAAAECnxwh/K9SsWqVZi95ROJ7U+2ujOufwxzWkxJLfE9DQSw6SJE0c0rfNc3388F9kxWJyAgHleT2KJZOK2EmFDHy+8/TJTytmx+RxPFq5dGWbnz/jiKclOyZ5A+YySDr56ZNlx2x5A16jOQAAAAB0fjT8LnCa7+HfvfuB2q9XnuE0Uv3+B2W+zquNqS4uRWwzI/wj+6Wm8cfjcc1dOtdIBklSqcHbCVroN7Kf6QgAAAAAugim9LsgvSuf6RX6tybPmwrVxJR+AAAAAOhSaPhd0Lxmn9phv59p+E2N8AMAAAAAzGBKvwuaZ/Sruu4T5cuSz+PT4NL9jOUpnfe8PJEmJfPyVT92oiQpkt5KoI19UPOB4sn4xk9FTNnwgZSMSx6/1G24sRg1H9QoGU/K4/eobHiZsRwAAAAAOj8afhckmzv+e//9Y21oqlavwn7627RlxvLs+ZuLFayuUrRvP6166zNJUjwp2UlH3ja+7+DYx4/NbMs3Y/CMNj33Jl4+duO2fFNWGIvx+LGPZ7blq1hRYSwHAAAAgM6PKf0uaM+T5X0eS77mHp9p/QAAAADQdTDC74L0on2jdj9ZQU+9QoESs4E2k+ezVB93FLEdFfrb9txnHHCGNkQ2KBQISY1te+5N7HGGFNsgBboZDCEdcMYBimyIKK+b+d0cAAAAAHRuNPwuSDf8U/f/Hw3tZnaf963J83pUH7fVlEhKatv9368ff72k5m355hrclu+A682du4Xx1483HQEAAABAF8GUfhckmyf1t8dV+iVW6gcAAACAroiG3wXpVfrbeD28nZbvo+EHAAAAgK6Ght8FyXbe8GdG+BM0/AAAAADQVXAPvwvSDf9NCyarKb5a3fN7647jDd6vvpk8b+pznWjSyWwh2FaOfORI1TTUqHdBb13S/ZI2Pfcm5h8pRWqkvDJp/AJjMR458hE11DSosKxQZyw4w1gOAAAAAJ0fDb8LnOZ7+FeGv9T6pio1xOoMJ9qU35OafZB02n5a/7K1y1QZrlRtqFbq3qan3lTdMqmpUorVGgwhrV22VuHKsCK1EaM5AAAAAHR+TOl3QXqEP89fqAJ/SPm+QqN57IJCJYpCsgtSOSzLMjatvyhQpFAgpEK/2Wsif5HkC6UeDQoUBRQIBRQoan+7OQAAAADoXBjhd0F6lvytx7+p/kVtvNH9Vix5dekWz+V7LTUmnDYf4f/0wk8ltYNt+Y7/1Ny5W7jw0wtNRwAAAADQRTDC74Jk86PHaqer9knK86Xe6oid3MGRAAAAAIDOgIbfBemF8NrrKv1Si5X62ZoPAAAAALoEGn4XpKf0t+N+P9PwN7E1HwAAAAB0CdzD74L0JPkXPrtfXjUq31+kE/c731ieQdf+Tr7aDUqUdNPXV10vScrzpRr+qO3Istruc56737xb4WhYBb4CDdTANjvvFj67W0qEUwv3DTV3H/2bd7+paDiqYCioQy881FgOAAAAAJ0fDb8L0qv0/+OjP2ltY5V6FfYz2vD3enqOgtVVivbtl2n4gx5LliRHUmHP3m2W5cbXblRluFLloXLNGDyjzc67hY9vTG3Ll19utOF/7cbXFK4MK1QeouEHAAAAkFNM6XdB+h7+9syyLAWbp/UXl5UbTgMAAAAAyDVG+F2Qbvd/cfjd8nviCvjyjObZlnyfpYjttGnD/9DkhxRJROSTT/FP42123i0c9pBkRySv2fdm8kOTlYgk5MvjPz0AAAAAuUXX4YL0lP6R/cerOOA1G2Y70gv3Fffp32bnnDB4giQpHo9r7qdz2+y8W+g7wdy5Wxg8YbDpCAAAAAC6CKb0uyCzSr/Vntfpl/K8qbebKf0AAAAA0PnR8Lsg2Typv71fzI0j/DT8AAAAANDZMaXfBekp/eubVqoh6sjj8apnQR+zobYivTVfcVm5HMdpkxkJ1eFq2Y6tpJ3c8cG51FQtObZkeaX8vsZihKvDcmxHltdSqG/IWA4AAAAAnR8NvwvSU/p/8cxYrWnelu9v05aZDbUV6RF+f16B6uNJhdpgvYFDHjikfWzL98IhG7flm7LCWIwHDnkgsy1fxYoKYzkAAAAAdH7tfRZ6u+c4jgyPXe80j2Upv3mUv6bJNpwGAAAAAJBLjPC3ktPi68N2m6j66HqFgt2N5ZGkdeMnyrdhvRLdtsxR5POoKWGrpimhPUsCOc8yaa9JWhdZp26Bbjk/13aVT5Ki66RgqdEYe03aS5F1EeWVts+tGwEAAAB0HjT8rZRs0fH/csyd8raDlfq/uPmubf6s0G9pdURa2Zhokyz3nXCfpOZt+eYa3Jbv0PvMnbuFE+47wXQEAAAAAF0EU/pbqeUIf0e4mEW+VMqaprZp+AEAAAAAZnSEHrVdS4/wW1KbrHrfWoX+1FteF0uqKdFRVh8AAAAAAOwqow3/DTfcoEMOOUShUEi9e/fWj370I3322WebHBOJRDR9+nT16NFDRUVFOvHEE1VTU7PJMcuXL9ekSZNUUFCg3r1767LLLlMisekI9sKFCzVixAgFg0HtueeemjVrliu/g9O8RH8H6PUlST6PpdqVqVXqa9poWj8AAAAAoO0ZbfgXLVqk6dOn64033tC8efMUj8c1YcIENTQ0ZI659NJL9c9//lNz5szRokWLVFVVpalTp2Z+btu2Jk2apFgsptdff12PPPKIZs2apauuuipzzNdff61JkyZp3Lhxevfdd3XJJZfoZz/7mV588cVW/w7pMXKPpKvmnaZf/PMYXTXvtFbXbY0DjxmjQ0cO0YHHjNnqz9d+nfpQZWUbTOs/ac5JGjtrrE5+6uScn2u7Xj1J+tfY1KNBc06ao1ljZ2nOSXOM5gAAAADQ+RldtO+FF17Y5PtZs2apd+/eWrJkiX7wgx+otrZWDz74oGbPnq0jjzxSkvTwww9r2LBheuONN3TYYYfppZde0scff6x//etfKisr04EHHqg//OEP+vWvf62rr75agUBAM2fO1KBBg3TrrbdKkoYNG6bXXntNt99+uyZOnNiq3yE9pd9jSR+telOrG6rUq7Bfq2q2VmB1jYLVVdv8+Zqvl2mPUePbZIR/8XeLVRmuVHmoXKcPPj3n59umNYulpkopv9xcBknfLf5O4cqwQuUhozkAAAAAdH7tapX+2tpaSVJpaWrrtCVLligej+uoo47KHLP33ntrt9120+LFi3XYYYdp8eLFGj58uMrKyjLHTJw4URdccIE++ugjHXTQQVq8ePEmNdLHXHLJJVvNEY1GFY1GM9/X1dVJSq00H4/HM88Hg0El7VTTbEkbV/BzJKf5+fz8vMzXLaWfc+zENo9paUfHpH+Wl5cnxe0tcrRUu+JrSamV+lv+Pi2ln9/Wz3dV+tYHt+rtaj6fUu+RIymxlde4/fvuTL1dOZeJfKZrUo96Juvloib1qGeyXi5qUo96pmtSj3om6+Wi5ub13KhrOelOzLBkMqkf/vCH2rBhg1577TVJ0uzZs3XWWWdt0nxL0qGHHqpx48bppptu0nnnnadvv/12k+n5jY2NKiws1Ny5c3XsscdqyJAhOuuss3TFFVdkjpk7d64mTZqkxsZG5efnb1L/6quv1jXXXLNFxtmzZ6ugoGCT5xqCxVpeNlyBeKN2q3wz87zf48/+YrTShHPOUf7atWrq0UMvPfjgFj9PeHz6vP/3JElDvntDXsfOWZZ4cuMfqclr4nE25kha5nIk4xsXSvT4WTMTAAAAwNY1Njbq1FNPVW1trYqLi7Oq0W5G+KdPn64PP/ww0+ybdMUVV6iioiLzfV1dnQYMGKBx48apR48emecPGDFSNz69SKq15csrUO9Djt6i1pQRe+nv73y+xfOOndCGd19TtwNHa+ohw7Z6zM7U2bzef190kb7yByVJHn9Q3UeO3eLY848do+lPvKxw3NFBPxiv/oVb/hnE43HNmzdPRx99tPz+1jfI1Ovc9XJRk3rUM1kvFzWpRz2T9XJRk3rUM12TetQzWS8XNTevl55p3hrtouG/8MIL9eyzz+qVV15R//79M8/36dNHsVhMGzZsULdu3TLP19TUqE+fPplj3nzzzU3qpVfxb3nM5iv719TUqLi4eIvRfSk1VT8YDG7xvN/v3+SNjEajSlpeSbY8HkuWd8vL2dQU2erzaZbXt8NjdqZOWiQSkXzNGS1t9TXRaFRlBX6Fa2NaE5MGddv2H+fmv3NrUa9z18tFTepRz2S9XNSkHvVM1stFTepRz3RN6lHPZL1c1EzXc6Om0TnFjuPowgsv1N///nctWLBAgwYN2uTnI0eOlN/v1/z58zPPffbZZ1q+fLlGjRolSRo1apQ++OADrVq1KnPMvHnzVFxcrH322SdzTMsa6WPSNVr1OzQ/etRB9uVr1ic/9UFATRus1A8AAAAAaHtGR/inT5+u2bNn6+mnn1YoFNLKlSslSSUlJcrPz1dJSYnOOeccVVRUqLS0VMXFxbrooos0atQoHXbYYZKkCRMmaJ999tFPf/pT3XzzzVq5cqWuvPJKTZ8+PTNKf/755+vuu+/W5ZdfrrPPPlsLFizQX//6Vz333HOt/h2SzUsgeCxp3hd/VTTRqKCvQEfv+eNW186lsgKvJOV8pf7ZH8xWY7xRAU9AJSrJ6bm265vZUqJR8hVIA081FuOD2R8o3hiXv8Cv4acON5YDAAAAQOdntOG/9957JUljx47d5PmHH35YZ555piTp9ttvl8fj0YknnqhoNKqJEyfqnnvuyRzr9Xr17LPP6oILLtCoUaNUWFioM844Q9dee23mmEGDBum5557TpZdeqjvuuEP9+/fXn//851ZvySdJ6SUPLUua+Z8rM9vytfeGPz3CvyZiK5505PfkZobC5fMuz2zLN2PwjJycY6csvXzjtnwGG/55l8/LbMtHww8AAAAgl4w2/DuzQUBeXp5mzJihGTO23Szuvvvumjt37nbrjB07VkuXLt3ljDuSbP4VOtp660V+jwp8lhoTjlY3JdSv0NzK9QAAAAAA97WLRfs6sqTSU/otnf+96zJT+k36+nfXydPUqGT+tnNYlqU++T59FY6rJocN/81H35yZ0q9vc3KKnXPQzRun9Bt09M1HZ6b0AwAAAEAu0fC3Ussp/e1lGv/qqTuXo6wg1fCvzOF9/KcOT02fj8fjmvvt9mdh5JTBafwtMY0fAAAAQFvpaDPR252OOqVfSjX8klTTaBtOAgAAAABwW0fsU9uVZPOj1bF25ZO0ceG+1ZGE7OSO11MAAAAAAHQcTOlvpY3b8lmK2dHM8wFv0FQk5X+xTJadkOP1qWnPIds8riTgUdBrKWo7WhOxMyP+boomUtcknoi7XnuXtHhvZPC9SUQ33j7hC/KfHwAAAIDcoeNoJafFlP5Tnhye2Zbvb9OWGcs0/CfHK1hdpWjffnpzybZzWJalsnyfltenFu7LRcM/+M7B7WNbvmcGb9yWb8oKYzHuHHxnZlu+ihUVxnIAAAAA6PyY0t9K6Sn9OdrGPuf6NDf5uVy4DwAAAADQ9hjhb6WWq/Tv2/tQbYisUbe8nmZD7YKyfK8kqTpHDf+oAaO0umG1euT3yEn9ndZzlBRdLQV7GY0xYNQANaxuUGGvQqM5AAAAAHR+NPytlLmHX5auPfoxw2l2Tk3NSg0eMlSSVNSzTKfMeEqVdRENHX6AEtGIJGlA//669OJftPpcc06aI6l5W765BrflGzPH3LlbOGnOSaYjAAAAAOgiaPhbqSNO6U8mHc1a9I4kyXEcvb06qpjPpz/+81V1C6ZG/H8+YZTJiAAAAACAVuIe/lZqOaW/I7IsS8WB1J9BXTy5g6MBAAAAAB0FDX8rpbev93TUjl/a2PDHaPgBAAAAoLNgSn8rJZXq+C1Jt7xykcLR9QoFu+uyH9xlNtguKPanGv5wPKmk47j64cXP//lzrYusU7dANx2v412ru8ve/LkUXScFS6VD7zMW458//6ci6yLKK83TCfedYCwHAAAAgM6Phr+VnMwIv/TGdy9qdUOVehX2MxtqFxX4LPksKeFIDXFHoYB7Df9znz+nynClykPlOn6wwYa/8jmpqVLKLzeXQdLnz32ucGVYofKQ0RwAAAAAOj+m9LdSskXD31FZlqVQ5j5+23AaAAAAAIAbGOFvpcwq/bJ035RXlEza8ni8RjMtnfuKLNuW4935HMUBj9ZHk6qLJVXu4hbxb537lmzHVtJOaukrS90rvKuOeUtybMky+96c+9a5cmxHlrcDf0IEAAAAoEOg4W8lp3lOv2VJPQv6GE6TEi/b9Rzp+/jrYsnM7+SGvqG+qUzxuJbKYMOf39fcuVsI9WUqPwAAAIC2wZT+VuoMU/olqcjvkUep+/ibbPcafgAAAACAGTT8rbRxSn/H5rEsFbE9HwAAAAB0Gkzpb6X07HfLsvTmivmKJSIK+PJ0aP/xxjL1eewheRsaZBcWauVpZ+/064r9HtXFkq42/C99+ZIiiYh8pv/Uql+S7IjkzZP6TjAW48uXvlQikpAvz6fBEwYbywEAAACg86Phbw3LUnryu8eSblp0QWZbvr9NW2Ys1m6336hgdZWiffvtUsNfEvBoRYO7I/xnP312Zlu+GYNnuFZ3l71x9sZt+aasMBbj6bOfzmzLV7GiwlgOAAAAAJ1fR5+JbpTX58983cFv4ZckhZoX7osmHRWU9jKcBgAAAADQGozwt4I3EMx87bGkUw+oUFO8Xvn+IoOpsuf1WCryWapPOCobOtyVmr8Z/RuFo2EV+AqkVa6UzM4+v5ESYclndpX80b8ZrWg4qmAouOODAQAAAKAVaPhbYfMR/hP3O99cGJcUBzyqT9jqPWR/V+pdeOiFklLb8s2dO9eVmlkZeqG5c7dw6IWHmo4AAAAAoItgSn8reP2pht+j1KJ9nUFxwCtJ6u3SCD8AAAAAwAwa/lbw+gOSpE7S60vaeB9/9/6DZHuYAAIAAAAAHRUNfyukG35PJ2r4A15L+d7UL9QYNHu/OwAAAAAgewzhtoInM6U/1SCf9peDtKZxpXoW9NFjP1lqMlqrhAIeNTXZagoUt7rW3nfvrapwlfoW9dXNu93sQrosPbu31FglFfSTjv/UWIy7975b4aqwQv1CuvDT9rGuAAAAAIDOiRH+VvD6Np3S35RoUGM8rKZEg8FUrVfcPK2/yYUR/vpYvcKxsBrihq9JvD61Sn+83miMWH1MsXBMsfqY0RwAAAAAOj9G+Fshs2hfc8Pfv2RPFQaK1T2/t8FUUtMeeyoRKla8V3Y5QoHmhj9QpKTjtCrLkB5DVJJXot4FZq+JiodIgRIpr8xojB5DeiivJE+FZYVGcwAAAADo/Gj4WyEzwt/8/R3HG9x2roUP5rQuR77XUqyhXoHCIq2OJNU/kH2tBWcskNQOtuUbv8DcuVs4Y8EZpiMAAAAA6CKY0t8K3kB60b5OtGqfUlsMrvnqE0lSVWPCcBoAAAAAQDZo+FvB69t0Sn9nsvqLjyVJ1Y224SQAAAAAgGzQ8LeCJ70tn+EcubD6i48kSdVNNPwAAAAA0BFxD38rpBftS8/of+DNqxWO1SoUKNG5h15tLNfQ6WfLv26t4qU99NmMh7KqsfrL1JT+DbGkGuJJFfqz+1jjd/N/pw2RDQoFQhqlUVnVcMV7v5NiG6RAN+mA643FmP+7+YpsiCivW57GXz/eWA4AAAAAnR8Nfyt4/Zvew//C57O1uqFKvQr7GW34S954TcHqKkX79su6RryxQYF4o2L+AlU1xrVXSTCrOo+894gqw5UqD5Vr1GCDDf9Xj0hNlVJ+udGG/71H3lO4MqxQeYiGHwAAAEBOdcbZ6G0msy2f4Ry5kh8NS5IqG1i4DwAAAAA6Gkb4WyGzLV/zlP6bj31KiWRCPk/nuKz50bBqi8pU1YqG//lpzyuejEtJ6bu3v3Mx3S4a97yUjEsev7kMkqY9P03JeFKeLG+RAAAAAICd1Tk6U0MyI/zNDf/g0v0MpnFfQaxOklTdGFfScbLafnB42XBJUjwe13cy2PB3G27u3C2UDS8zHQEAAABAF8EwYyt40iP86oT78kkKxJsU8EjxpLSK1foBAAAAoEOh4W8FXyC9aJ/hIDliSeqbn5oEUtUQNxsGAAAAALBLmNLfCh7fplP6P1u9VPFkTH5PQEN7HWQwmXv6Fnj1bUNClQ0Jjei1669fUrVEMTsmj2P4s6V1SyQ7JnkDUulIYzGqllTJjtnyBrzqNzL7XRQAAAAAYEdo+FshvS1f+tb23770k8y2fH+btsxgMvf0K/BKkiqzHOGf/OTkzLZ8MwbPcDParlk0eeO2fFNWGIvx5OQnM9vyVayoMJYDAAAAQOfHlP5W2LgtXyed0y+pb0HqM6ENsaQa40nDaQAAAAAAO4sR/lZIj/Cnp/Qfv/eZaojVqTBQbDCVtPLUM+UN18kOtT5HntdSjzyv1kZsVTbGtVdJcJdef+6Ic1UbrVWRv0gKtzpO9vY8V4rVSoESgyGkEeeOULQ2quAuXkcAAAAA2FU0/K3g3ewe/rNG/tZgmo2W/9LdHP0KfFobsVXVkNjlhv/3Y38vKbUt39y5c13NtUuG/97cuVsY+/uxpiMAAAAA6CKY0t8KmXv4DefItfLC1AcblQ0Jw0kAAAAAADuLhr8VPOl7+K3O3fKXF6YmglQ3xpV0HMNpAAAAAAA7g4a/Fby+TVfp70xWr6qRJB0wYqQOO3A/RRvCiielUcf+SIOHDNXgIUN1+OgxhlMCAAAAALaFe/hbwRtoXrSv+fvpTx+tdU01Ks0v04zJ84zlOnTkEAWrqxTt209vLslue8D0SP7M51+V5fXp4/VRrY8mdendj2Sm+J95xIgd1hn90GitrF+pssIyXd7z8qyyuGLeaKlppZTfRzr6NWMxHhr9kOpX1quoT5HOfu1sYzkAAAAAdH40/K2w+aJ9K+u/1eqGKsXsiMFUuVHs92h9NKm6WFLlhTv/um82fKPKcKUiiYjUM3f5dqj+G6mpUjL83mz4ZoPClWElIqyHAAAAACC3aPhbwdt8D7/VPKc/FOyumB1VKNjdZKycKAl4JSVUF0vKcZzM77wjpfmlitpRdc8zfE2CpVIymno0KL80X3bUVn5pvtEcAAAAADo/Gv5W8Pg2ndI/67/+Yy5MjhX6LXksKeFITQlHBf6da/jfv+B9Se1gW77j3jd37hYueP8C0xEAAAAAdBEs2tcKvvQ9/J1w0b7NeSxLIX/qz6U2njScBgAAAACwIzT8reDZ7B7+zq44kPpzqYvR8AMAAABAe0fDnyXHceQLBCVJlrpGx1/sTzf8tpzmVfwBAAAAAO0T9/BnKdmi302P8P/l/bvUEKtTYaBYP9n/IjPBcigU8MiSFEtKUXvnGv7bFt+mumidCn2FGqIhuQ24PZ/cJsXrJH+xNKzCWIzFty1WtC6qYHFQoypGGcsBAAAAoPOj4c9SosUId3rB+r9+cJdWN1SpV2G/Ttnwey1LRX6PwvGkandyWv9ti29TZbhS5aFyzRg8I8cJt+PT21Lb8uWXG2/4w5VhhcpDNPwAAAAAcoop/VmyW/S7XekiZu7jZ+E+AAAAAGjXGOHPUnqE35Iye9JfOe7Pitsx+b0Bg8mkz+76s6xYTE7A/RwlAY8qG3Z+4b7Hpj6maCIqr7xq+KjB9Tw77fDHJDsqeYPmMkia+thUJaIJ+YL8pwcAAAAgt+g6spS+hd1qsV7fQf1+YCbMZmoPz12O9NZ8EdtRQfeeOzx+7MCxkqR4PK65H83NWa4dKhtr7twtDBw70HQEAAAAAF1EV5qN7iq7edW+rnYBfR5Lhb7Upxx99j7AcBoAAAAAwLZ0tX7VNYnmEX5P19iRbxPp+/j7DKPhBwAAAID2iin9WbKb7+H3tJjTX1X3jZKOLY/lVb/igYaSSSWvv5K5hz8X0/tLAl5VN9rqO+zAHR779fqvZTu2krbhRf7qv5YcW7K8UtEgYzHWf71eju3I8lrqPqi7sRwAAAAAOj8a/iwlkhsX7Uu76J8TMtvy/W3aMjPBJA296GcKVlcp2ref3lzifo70CH/pboPVlEgq37ftiSJjHh7TPrblmzdm47Z8U1YYi/HwmIcz2/JVrDC3PSAAAACAzo8p/Vmyu/CUfr/HUr439Yt/Vx83nAYAAAAAsDWM8GcpM8LfouEfM/AEhaMbFAp2MxOqDRUHPGpqsrWiIaEh3ba91d2UvadofWS9SgIlkt2GATc3YIoUWy8FzE6j33vK3oqsjyive57RHAAAAAA6Pxr+LGVG+FtM6r/k+7caStP2SgIe1TTZ+iYck1S4zePuOu4uSc3b8s01uC3fwXeZO3cLx911nOkIAAAAALoIpvRnaeOifYaDGNIt6JWTTGpVk61w3OTQPQAAAABga2j4s5RedN7qog2/32Np9VefSpK+quM+fgAAAABob2j4s5RIj/AbzmHSd+++IUn6ui5mOAkAAAAAYHPcw5+l9KJ9nhZD/L954ceqjaxRSV5P3XjMX01FazMr3n1DI//rbH0djivpOJtci7QfPvFDrW5crR75PXRu4bkGUjZb9EMpslrK6yUd8YyxGE/88Ak1rm5UQa8CnfLMKcZyAAAAAOj8aPizlF60r2WP+/nad7W6oUq9CvuZCdXGVn/xifK9lppsR5UNCQ0o8m9xzDvV76gyXKnyULk02EDItHXvSE2VUn65wRBS9TvVCleGFSoPGc0BAAAAoPPryjPSW6WrL9onSY6T1KDigCTpK6b1AwAAAEC7wgh/ltKL9rX8xORv05YZybK5N5e0XY49iv36eH1UX9XFdES/LbfnW1GxQlI72JZvygpz526hYkWF6QgAAAAAughG+LOUXrTP6qrL9DcbFEqN8Nc02aqPJw2nAQAAAACk0fBnKX0Pf1ee0i9JhX6P+uSnJoqwWj8AAAAAtB80/FnKrNJvOEd7sEdxarE+7uMHAAAAgPaDe/iztHGV/o1D/M9/9piaEg3K9xXq2KGnGUom7Xbr/8obrpMdKtbyX/425+fbozig12uatro936x3Z6kh1qA8b556qmfOs2zTV7OkRIPkK5T2ONNYjHdnvatYQ0yBwoAOPPNAYzkAAAAAdH40/Fna2ir9f3772sy2fCYb/j6zZylYXaVo335t0vD3K/Qpz2spYjuqakiof4vt+a5ccGVmW74Zg2fkPMs2vXflxm35DDb8C65ckNmWj4YfAAAAQC4xIz1LTOmXampWavCQodpr6N766NV/SZIqbrxDg4cMzfxbWVNjOCUAAAAAdE2M8Gdp45T+jc/94vBbFE00KejLNxOqjSWTjmYtekeSVNOY0Bd1cY35yVm6aPoFmWMWff207rz6l7rt1tukr0wllXTwnVKiUfIVGAwhHXvnsYo3xuUv8O/4YAAAAABoBRr+LGVG+Ft0/EcMmmwqjnHdg15JcTUkHMVsRwFv6rocMWiyHv7mGk3Ze4rmfjXXXMABU82du4VhU4eZjgAAAACgi+jKM9JbJbMtn9kY7UbAa6nQl2ryN8Rsw2kAAAAAAPSrWUqP8Lec0t/VpUb5pXURGn4AAAAAMI0p/VnKjPC3aPgbY2E5cmTJUkEgZCaYQaVBr1Y0JLQhlsxsz9cYCyvpsxWOhs2Gi4clOZIsyW/uvYmGo5kYwVDQWA4AAAAAnR8Nf5Yy2/JpY8f/0zkjM9vy/W3aMlPRjCnyW/J7pHhSqosl1S3oTV2Tn1Rp//v3N7st37PDNm7LN2WFsRgzhs3IbMtXsaLCWA4AAAAAnR9T+rNkJ1OPTOnfyLKsjdP6o0zrBwAAAACTGOHPUiI9wt+i4T+g72jVRtaqJK+HoVQptYeNln/dWsVL2z5HadCrVU221kWTGuQ4OqDvaP37lWc15ugxbZ5lE72PkKJrpGBPozEGHjFQjWsaVdDT7PaAAAAAADo/Gv4sbZzSv9FVRz5kJsxmPpthLke3QOomh6jtqCnh6KojH9KZ14zQo3c/qrlzDW7L9/3HzZ27hamPt4/tAQEAAAB0fkzpz1IiM6WfOf0teT2WugVSf1ZM6wcAAAAAc2j4s5B0HDUv0r/JlH6kdM9L38efNJwEAAAAALouGv4spLfkk7iAW1MaTF2VcDypeNLZwdEAAAAAgFzgHv4sJFo0sS1n9N+w8HzVRtapJK9UV4ydaSBZyvCTjpN/9SrFe/XWB3Pa/r75oNejQp+lhoSj617+uVYeUamfPfszTfUYvH/9jbOk6Fop2EM67GFjMZ4+62k1rm1UQY8CTX54srEcAAAAADo/Gv4spEf4k0lbLWf0v125QKsbqtSrsJ+RXGn5X32hYHWVouE6Yxm6B71qSCS0tOplNfZv0Pyv52vqYIMNf/U8qalSyi83l0HSl/O+VLgyrFB5yGgOAAAAAJ0fM9KzkB7hT8bjLNq3DaXN9/E7zOgHAAAAACOMNvyvvPKKTjjhBPXr10+WZekf//jHJj93HEdXXXWV+vbtq/z8fB111FH6/PPPNzlm3bp1mjZtmoqLi9WtWzedc845qq+v3+SY999/X2PGjFFeXp4GDBigm2++uVW501vy2fHYJs8/dOJiPfPTr/XQiYtbVb8zKPJZ8nuki8Yu1Pc+OlFvn/O22UDHvSdNXZV6NOj8987Xr1b9Sue/d77RHAAAAAA6P6MNf0NDgw444ADNmDFjqz+/+eabdeedd2rmzJn6z3/+o8LCQk2cOFGRSCRzzLRp0/TRRx9p3rx5evbZZ/XKK6/ovPPOy/y8rq5OEyZM0O67764lS5bolltu0dVXX637778/69zpKf12Ir7J8yV5PdQtv5dK8npkXbuzsCxL3YNeFQZKteeB49WjwPA1CfaQ8nqlHg0q6FGgwl6FKuhRYDQHAAAAgM7P6D38xx57rI499tit/sxxHP3pT3/SlVdeqcmTU4ubPfrooyorK9M//vEPnXzyyfrkk0/0wgsv6K233tLBBx8sSbrrrrt03HHH6Y9//KP69eunxx9/XLFYTA899JACgYD23Xdfvfvuu7rttts2+WBgV6Sn9Nvx+A6O7NpKg16tarK1+8jvy2FuPwAAAAC0qXa7aN/XX3+tlStX6qijjso8V1JSou9973tavHixTj75ZC1evFjdunXLNPuSdNRRR8nj8eg///mPpkyZosWLF+sHP/iBAoFA5piJEyfqpptu0vr169W9e/ctzh2NRhWNRjPf19WlFr+Lx+OKx+OKxhOSpGQiLsdObPf3yM/P2+ox6eccO7HNY3amzub18vLypLjd/KS2+pqdOV9eXt4mdbPJVOJz5JEU6t1PVfWp2x/iLn1Ikq5DvfZRLxc1qUc9k/VyUZN61DNZLxc1qUc90zWpRz2T9XJRc/N6btS1nHYy9GpZlv7+97/rRz/6kSTp9ddf1/e//31VVVWpb9++meN+/OMfy7Is/eUvf9H//u//6pFHHtFnn322Sa3evXvrmmuu0QUXXKAJEyZo0KBBuu+++zI///jjj7Xvvvvq448/1rBhw7bIcvXVV+uaa67Z4vnZs2eroKBADcESLS/bT8FYg/ZY+W7m52/VvqWYE1PACuiQkkNaeUWyN+Gcc5S/dq2aevTQSw8+aCyHJD3vVCnsD6g00qCjfLsby1GWeEtexWQroBqfufem9q1aOTFHVsBSySElxnIAAAAAaN8aGxt16qmnqra2VsXFxVnVaLcj/CZdccUVqqioyHxfV1enAQMGaNy4cerRo4e+Cse1/NtGVa74TiO/NzZz3P1P/LdWN1apV0E/TTjyMknSlBF76e/vfL75KeTYCW149zV1O3C0ph4ybKvHtLStOpvX+++LLtJX/qAkyeMPqvvIsVscu6NaknTK4cN1z113qduBo2V5t/5nsjN15swepnVN1eqW31fjh96rCUcfLb/fv93X7Ix4PK558+bp6J2s53t2uqymSjn55Uoc9/tW18s2313T78psy3fK709pdT2387WnmtSjnsl6uahJPeqZrJeLmtSjnuma1KOeyXq5qLl5vfRM89Zotw1/nz59JEk1NTWbjPDX1NTowAMPzByzatWqTV6XSCS0bt26zOv79OmjmpqaTY5Jf58+ZnPBYFDBYHCL5/1+v/x+vxxPMnWuWHTTZtja+Jh+vqkpss2GWUodt6NjdqZOWiQSkXz+LXLsaq30woiW17fNY3emjqd528KkI0X8hZlr6JZdrWc1v8atejuyvXrZnMf09TNRk3rUM1kvFzWpRz2T9XJRk3rUM12TetQzWS8XNdP13KjZbhv+QYMGqU+fPpo/f36mwa+rq9N//vMfXXDBBZKkUaNGacOGDVqyZIlGjhwpSVqwYIGSyaS+973vZY753e9+p3g8nrlg8+bN09ChQ7d6//7OsJNb35bvjBG/UVO8Qfn+wqzqumX5pb+Rt6FBdqHZHJJ05sjf6NV33lK/wQcqXNDTXJDhV0nxeslfZC6DpCOuOkKx+pgCRYEdHwwAAAAArWC04a+vr9cXX3yR+f7rr7/Wu+++q9LSUu2222665JJLdN1112mvvfbSoEGD9D//8z/q169f5j7/YcOG6ZhjjtG5556rmTNnKh6P68ILL9TJJ5+sfv36SZJOPfVUXXPNNTrnnHP061//Wh9++KHuuOMO3X777VnnzmzLt9kiCj8cdnbWNd208rT2kUNKXZOlsz7SoeNPV10iYm61/j2z25HBbSPPG2k6AgAAAIAuwmjD//bbb2vcuHGZ79P3zZ9xxhmaNWuWLr/8cjU0NOi8887Thg0bNHr0aL3wwguZFeQl6fHHH9eFF16o8ePHy+Px6MQTT9Sdd96Z+XlJSYleeuklTZ8+XSNHjlTPnj111VVXZb0lnyTZzU2rnWBbvp3x3dLF8llS3JenVZGk+jO4DQAAAAA5Z7ThHzt27HZHfC3L0rXXXqtrr712m8eUlpZq9uzZ2z3P/vvvr1dffTXrnJtLpG7h32JKP7YuEY1oj5Bfy+ri+qw2pv7FeTt+EQAAAACgVTymA3RE6RH+pIt7OLrJX7NSgapK+WtWmo6SMbQktX7CZ7Vxc9P6AQAAAKALabeL9rVn2xrh/8kT+2lNQ5V6FvbTX0750ECylIOO+4GC1VWK9u2nN5csM5ZDSl2T6pO/1fGP7avzx7yhurhXK5sS6lvg7sqYO/T0HlJTpZRfLk3+qm3P3cIde9yR2Zbv4q8uNpYDAAAAQOfHCH8WMvfwb9bwJ5IxxZMxJZJM9U9LJGOS11E8GVdR03pJ0qfrDVyfZGzjP4PsmJ35BwAAAAC5RMOfhT2KAxrbr0DL33l9k+cHdd9He/U4QIO672MoWfszqPs+CqwLat9e+6q4cY0k6ZMN0baf1t9tP6n7QalHg3rv11t9Duqj3vv1NpoDAAAAQOfHlP4sDCjya0CRX5UfvLXJ83887h9mArVjfzzuHzrzphF69o/P6tnnX5DfI9XFkqpuTKhfYRtO6x/3QtudaztOe+E00xEAAAAAdBGM8CPnampW6oARI+Vxklr2+gJJ0q9uvVeDhwzN/Dt89BjDKQEAAACgc6HhR84lk45mPp/aFvGYCUdLkg6ePE0PL1yiWYve0axF76hm1SqTEQEAAACg06HhR5vqHrDktaRY0lE4njQdBwAAAAA6Le7hd9E9b/xO4egGhYLd9N+HXW86Trtwzxu/U2xCRPe++T862TteHstSadCr1RFbqyO2igPetgmy9DIptl4KdJcOuqVtzrkVL132kiLrI8rrnqcJt0wwlgMAAABA58cIv4vmfzlHz332iOZ/Ocd0lHZj/pdzZO8f1/wv/5Z5rld+qslfG7HbbrX+b56Qvnww9WjQh098qKUPLtWHT3xoNAcAAACAzo+GH22uJOCRz5LiSak2xrR+AAAAAMgFpvS76PZJz8pOJuT1mL2sH/zlWVl2Qo7X/Nt7+6Rndc6xo3Xbs09LX1ZKkjyWpR55XtU0pab1dwu2wbT+8fOlZEIy/N6cPv90JRNJeXx81gYAAAAgt8x3hJ3Ibt2GmI4gSWras33kkFLXxLPWq9267aX1qsw83ys/1fCvjdgaXNwG0/qLh+b+HDuh59CepiMAAAAA6CIYZoQRxX6PAh7JdqT1Uab1AwAAAIDbaPhhhGVZ6pmXmsq/JmIbTgMAAAAAnQ9T+l30Yc1/FLej8nuD2q/se8Zy9Hrqr/I0NSqZX6DVU39sLIeUuib2gIQ+rHlT5Zv9rGeeT1WNttZFbfmCebkNsnqxlIxKnqDUa1Ruz7Ud3y3+TnbUljfo1YBRA4zlAAAAAND50fC76Pf/+qlWN1SpV2E//W3aMmM5Bl1/pYLVVYr27We84f/9v36q2E+adPWCM/XAkHs2+VmR31Ke11LEdrTbiO/nNshrJ0lNlVJ+uTRlRW7PtR1zTpqjcGVYofKQKlZUGMsBAAAAoPNjSj+MaTmtf/Dh4w2nAQAAAIDOhRF+F03Z5zw1xMMq9IdMR2k3puxznh66+3/1o+k/kxJb/rxnnlcrGhIacNAoRRJJ5eVqu7oh06V4neQvzk39nXTI9EMUrYsqWBw0mgMAAABA50fD76LTDvqV6QjtzmkH/Ur/9+qtOu3+Cq1fsnCLnxf6PSrwWWpUQJ/VxnRAjxzdy7/vFbmpu4vGXDHGdAQAAAAAXQRT+mFcr+Zp/R+sjRhOAgAAAACdBw0/jOuV75OTTGpFQ0IbomzRBwAAAABuoOGHcUGvpcoP3pYkfbCOUX4AAAAAcAMNv4vOfWqMTnx8iM59ivu00859aoyafl6v854eu93jlr3yvCTpw3VROY7jfpAXDpb+3j/1aND9B9+v2/rfpvsPvt9oDgAAAACdH4v2uWhdU41WN1SZjtGurGuqkUKO1jWu2u5x37y5SAGPpdpYUt81JLRbkd/dIE0rpaZKd2tmoX5lvcKVYdMxAAAAAHQBNPwuKs0v2+TRlFivsk0eTSrNL9PqldUq7dl7u8fZsaj27hbQ++ui+nBtxP2GP7/Ppo+GFPUp2uQRAAAAAHKFht9FD0x91XQESdK7L7SPHFLqmkwc0lf3f7Jwq9vytbRfjzy9vy6qTzfEdPQAR36P5V6QY952r1YrnPf2eaYjAAAAAOgiuIcf7caAQp9KAh7Fko6WbYiajgMAAAAAHRoNP9oNy7K0X2lQkvTBOhp+AAAAAGgNGn60K8NL8yRJ34TjqovZhtMAAAAAQMfFPfwuemzpH9UQD6vQH9JpB/3KWI49L79Ivg3rlejWXV/cfJexHFLqmsTHRPXYe7dpkkbs8PhuQa/6F/q0oiGhj9ZFNapPgTtBPrpBitdJ/mJp3yvcqZmFV294VdG6qILFQY25gu0bAQAAAOQODb+L/v7x/VrdUKVehf2MNvyl819UsLpK0b79jGVI+/vH9yvxvZj+8fGfNWnIPTv1muE98rSioV4frIvqsLJ8WZYLi/ctm5Hali+/3GjD/9aMtxSuDCtUHqLhBwAAAJBTNPxoF2pqVmrwkKGSJH9+gU6992mtU4HG/vgMrXjvP5Kkst699fpr7WcHAgAAAABoz2j4XXTNUf+nuB2V3xs0HaXduOao/9Mlpx6vq/9vlrSicZvHJZOOZi16J/P913UxVTXaOvnqP2UW8jvziB3fErBNo+dIyajkMfvenDTnJNlRW96g12gOAAAAAJ0fDb+L9iv7nukI7c5+Zd+T9zuf9is7VOtXLNzp1/Ut9Kmq0VZtLKn6eFJF/lauL9lrVOte75IBowaYjgAAAACgi2CVfrRLeV6PeualRsGrGhKG0wAAAABAx0PDj3arX2FqAsqaiK2onTScBgAAAAA6Fqb0u2j5hmWykwl5PT7t1m2I6TjtwvINy5TsYWv5hs8V2sXXhvweFfs9qosnVdVgty5I3WdSMiF5fFLx0NbVaoU1n61RMpGUx+dRz6E9jeUAAAAA0PnR8Lvo0ueOz2zL97dpy0zHaRcufe54Rc9qVMXzk/XATm7L11J5oU91G2KqaUrIn1+QfZD54zduyzdlRfZ1WunR8Y9mtuWrWFFhLAcAAACAzo8p/WjXugc9yvdash1p7yNPMB0HAAAAADoMRvhdNH7wSQpHNygU7GY0x+rJJ8lXu0GJErM5pNQ1mfPYPRo/7cSsXm9ZlvoV+vRlXVz7Hfdj2Y4jr2XteqGBp0ix9VKge1Y53LLfKfspsj6ivO55RnMAAAAA6Pxo+F3034ddbzqCJOnrq9pHDil1TZ4+/SFdcMcftH7Jwqxq9M73anl9XEU9++jT9VHtW5pFs3zQLVmd220TbplgOgIAAACALoIp/Wj3PJalvgWpz6ZeqW5UPOkYTgQAAAAA7R8NPzqEfgU+1a+pUW0sqcUrG03HAQAAAIB2j4YfHYLXY2nxI3dKkv6zqknroq3cpg8AAAAAOjnu4XfRr+b+SOubVqt7fi/98bh/GMsxcsxBCtSsVKysj5a8utRYDil1TSI/bdBlL5yo3/a6qFW1vnlzofYI+fVVOK75VRHt0iZ9Lx8jRVZJeb2lcS+0KkdrPHbMY2pY1aDC3oU67YXTjOUAAAAA0PnR8Lvo6/Ufa3VDlXoV9jOaw9vYIF99WHYoZDSHlLomTllSX6//ROrV+npHDyjSnz9Zr+UNCZUX9Nj5F274UGqqlPLLWx+iFVZ9uErhyrBC5ebfGwAAAACdG1P6XeTzBOT3BOTzBExHaTd8noCUkHxevyv1uge9OqwsX5JU022QYvZOLuDnCWz8Z5A34M38AwAAAIBcYoTfRX855UPTEdqdv5zyoSYO6asnP3kv6235NndYWYE+WhfVBgX1+qqIjt5tJ5r4yV+5cu7Wuviri01HAAAAANBFMMKPDsfvsXRk3zxJ0jtrY6pujBtOBAAAAADtDw0/OqRBIb9CDWvkSHr667AidtJ0JAAAAABoV2j40WHU1KzU4CFDNXjIUB0wYqT6rvtC9atXakMsqYtmztHgIUN1+OgxpmMCAAAAQLvAPfwueuaTh9QUb1C+v1A/HHa26TjtwjOfPKT4yJie+XSWxmhgq2olk45mLXpHkuTYCa1fslCj9irXhxtsDT58vI4+5hhdceyhW3/xF/dL8XrJXyTteV6rcrTGkvuXKFYfU6AooJHnjTSWAwAAAEDnR8PvokfeuTGzLR8Nf8oj79yoxLioHl16s8YMucf1+iG/R7uHLH0TTujrurhKd99r6wd+cO3GbfkMNvyLrl2U2ZaPhh8AAABALjGlHx1evwKfKUSLaAAAXA5JREFUugc9ciSNv+RaRbmfHwAAAAAY4XfTL0ffoajdpKA332iOL268Q55Ik5J5ZnNIqWty1fSfquKu26TVuTmHZVnaqySgd9dE1a3fbnrxuwadsHuRLMvaeNChMyW7STL83hw/83jFm+Ly5/uN5gAAAADQ+dHwu+jw3Y81HUGStO7o9pFDSl0T7zK/Dt/tGK1fvTBn5/F7LA3t5td7qxv18XqpR55X3+9TsPGA8uNzdu5dMeT4IaYjAAAAAOgimNKPTqM44NW/H7xVkvRqdaM+WhcxnAgAAAAAzKHhR6fy6fxn9L3eqWn7c5fX67v6uOFEAAAAAGAGU/pdVBtZK8dJyrI8KsnrYSxH0ftLZcVicgIB1e9/kLEcUvM1yU+qNrKuzc45tl+BNsRsfbYhpr99VafTh3RTqbVBcpKS5ZGC5t6bxrWNcpKOLI+lgh4FO34BAAAAAGSJht9FZ/9tVGZbvr9NW2Ysxz5n/UTB6ipF+/bTm0vM5ZBS1yQyvUHn/H20HsjBtnxbY1mWjt89pLpYraobE5rzVa3O/ewAedLb8k1Z0SY5tmbmATMz2/JVrKgwlgMAAABA58eUfnQqNTUrNXjIUO2999665bRjFV5VrfXRpMKR1NT+6poaDR4yVIePHmM4KQAAAADkFiP8Ljq4/EjVRtapJK/UdJR24+DyI/Xi3Cc08tixbXK+ZNLRrEXvZL5vTCT18bqYvi0+UgX2egV69dSsRffpzCNGtEmezQ0+erAa1zYynR8AAABAztHwu+iKsTNNR2h3rhg7UwvPe1pX3HCP1i9Z2ObnL/B5tH+PoF71zlA4npQkDWwwt5Df5IcnGzs3AAAAgK6FKf3o9AJeS/uVBtQ73ytJ+iac0BEX/E624xhOBgAAAAC5Q8OPLsFjWdqz2K9BIb8kacjY4zT323o5NP0AAAAAOikafnQZlmWpX6FPe3cLKGkn9NH6qP5V2UDTDwAAAKBTouF30bULztYv507WtQvONh2l3bh2wdmKntioPyw812iOoR+frf3enayhH5+tHnleLbzneknSktUR/XtlU5vleGraU3ps4mN6atpTbXZOAAAAAF0Ti/a56L3q17S6oUq9CvuZjtJuvFf9mpKDbL1X/W8pNM1YjpINrykYrVI0mHpvvnztJd1w082at6JBr61slN/Ka5Mc3yz6RuHKsELloTY5HwAAAICuixF+dFkje+VrdJ/U9ngvV0e0obCX4UQAAAAA4B5G+F30fyctkSNHliyjOZYsWiI5jmSZzSGlrsmPRuylR9/6j6IfLDGWY8mhSyQ5UvN7U1OzUoOHDJUkHXbGxRp+3I9VXbqX/qviKn34Ymq6fVnv3nr9tVddzTH9k+ktYwAAAABAztDwu6gg0D6madtF7SOHlLomVsxSQSCkqMEctm/Ta5JMOpq16B1JkuM4+rI2pppIUqPO/qVOvvjX6l/o15lHjHA9RzAUdL0mAAAAAGwNU/rR5VmWpT2KPOpR+50k6dtwQt+G44ZTAQAAAEDr0PADSjX9vWuXa7fC1H8SKxoS+v45v2TLPgAAAAAdFlP6XbTo66cVTTQp6MvXEYMmG8tRft9d8obrZIeKVfnzi4zlkFLXJDEsrle+eUbDVWwsR4/VT8tjNynpzdfaXtt+b/oXeOXzevVVXVz7TJiqBz/doFFl+RrWPSiPC2sifPLUJ4o3xuUv8GvY1GGtrgcAAAAA20LD76I7X78ssy2f0Yb//rsUrK5StG8/4w3/na9fpvikiO5c/Bs9MOQeYzkGf35ZZlu+7TX8ktS3wCefJX1YvV5rVKR/fluv11Y26rCyAu3XPSivJ/vG//lfPJ/Zlo+GHwAAAEAuMaUf2Ipe+T7d8ePReuvJ+xSp26D10aSeX16v38//WCdefr32Pfh7GjxkqA4fPcZ0VAAAAADYKkb4XfSzg69SU6JB+b5C01HajZ8dfJX++D8X65xrfieFzeX4ZtBV8toNsr07/9401dXpkksukZ10tLIpocqGhIp6lumQU36uQ0/5uXrmeXV/xZm7lOPI645UrCGmQGFgF38DAAAAANg1NPwuOnboaaYjtDvHDj1Nf3r3Mh07ZJrWL1loLMeqvtm/N16PpfJCv/oW+LQmYqu6IaH6hKPVEVtT/vdBPfrZBh1alq8hJYEd3ud/4JkHZp0DAAAAAHYFDT+wkzyWpd75PvXO9ykcS6q6MaGV4SZVNUr/+DqsbgGPDu2dr+E98kxHBQAAAADu4QeyEQp4NKRbQE9Mn6rDy/KV57W0IZbUSysadM+H6/TvmogSHr/pmAAAAAC6MEb4gVb4ZtknOmvsCPmCeRpyxHEaPulkqU+53lgdlVV+sH5+52P67t8vaN5TfzEdFQAAAEAXQ8PvohMfH5LZlu9v05aZjtMunPj4EDX9Kqz/emIfo9vyHfr6kMy2fG8e7t57k0w6mrXoncz3juNobTSpyvq46hMeDTnyBA058gTN+bJW+5bm6YXhMzLb8lWsqHAtBwAAAABsjoYfcJFlWeqZ51WpL6mqj5aqbsD+Wh9z9GVdXF/WxRWLJyVJCUeyk468nu0v8gcAAAAA2aLhd9FePQ5U78L+KsnraTRH/X4HKtqvv+KlZnNIqWuy5vOV2uuw/Y3mqC86UNFgf8X9bXNNLMtSQbRO5SU+nXbMERo15aca/P2jlL93L6msSNHu+bru9W+0fMm/9fV/FmrFe/9Rz+7d9Pprr7ZJPgAAAACdHw2/i2485q+mI0iSPn6kfeSQUtdk4i/66ob/edLotnwf72/umqz59itd+N/ny3Ec1T+6h1ZHEloTsRVPSnv94Bjt9YNj5LGkL16fr2/DMe1W5Je1g+39AAAAAGBHaPiBNmJZlkIBS6FAQINCjsLxpNZEbK2LJBVNOtpj1Hg98UWdyvK9OrR3vvbuHpSXxh8AAABAltiWDzDAsiwVB7zaozigkb2C2r9HUB+/9JR8llTTZOuf39Zr5kfr9eaqJsVsx3RcAAAAAB0QDT9gmGVZCvk9eup/L9ND5xynt/9yvxo3rFU4ntSCygZd98rnOuW3NyppeXT0xGNMxwUAAADQQTCl30V/+vcvFY5uUCjYTZd8/1ZjOfY548fyr1ujeGlP4/fz/+nfv1TsuCb9afFlOiMwyViOwct+KV9igxK+bvpyiLn3Ztnv5iteG5G/JE9Drh+/yc+SSUf3Pfty6mvH0aomW5UNCalbqQ4+9QJ9YcfUc+RYRe2kgl4+qwMAAACwfXQNLnr1m39q3hd/0avf/NNojqIP31XxkjdV9OG7RnNIqWti75PQa988ZzRHjzX/VO+av6jHGrPvzeoXPlfNU59o9Qufb/c4j2WpT4FPB/UMas9iv4IeyfYGdPCpF+iuD9bpqa/q9PG6qKJ2so2SAwAAAOhoGOEH2jGPZamswKeeAUfffvaJPo0GVdJvNy2rjWlZbUw+SxpUHNCQkoAGFwdU4OczPAAAAAApNPwuuuuEl5R0bHksr+ko7cZdJ7ykM476nu56aa607BtjOd476CVZji3H8Hsz4u8ny7EdWd5dW33fY1nq1rBKfz7vPA3c/2Dtcdg47XHYkSrpt5s+r43p89qYnGRSNcs+0LpP39VtV1ysHkEv2/sBAAAAXRgNv4v6FQ80HaHd6Vc8UJ4NHvUNDdR6fWMsRzR/oLFzt5S/W7dWvT7pOLrt0dS6DI7jqDHhaG3E1rqorYaER332PkB99j5Af/5kg7oHPdqzOKA9SwIaUOSXh+YfAAAA6FJo+IEOyrIsFfotFfo92i3kV9ROal00qUWLXtHAEaO0PprUW6sjemt1REGvpbJ8n0qDXnUPehTySRF/oWpjSRUoKb/Xks8SMwIAAACATqRLNfwzZszQLbfcopUrV+qAAw7QXXfdpUMPPdR0LMAVQa9HfQs8euTS09R/9z1Uvv8h2n3kaA0YMUoq7q7l9XEtr49vfEHfA/XnZeHMt5akAp+lkoBXJQGPipsfSwJelQRTj34PHwgAAAAAHUWXafj/8pe/qKKiQjNnztT3vvc9/elPf9LEiRP12WefqXfv3q6cY2nVK4rbMfm9AR3U7weu1Ozolla9IntgQkurX9VAgzlK1r8iKxmT4wmotru592b968uVjNnyBLzqfvhuOTlHMunogRdfy3zvOI7qE46aEkk1JRxFbEcfvv+eygcPUSRhyx/MTx0nqSHhqCGRUFXj1msX+CwV+T3yWZa8HqUeLUs+j2Q5jqpK99S/qpoU8EaV7/Oo0OdRvi81C6HA51GBz1LAYzGTAAAAAGgDXabhv+2223TuuefqrLPOkiTNnDlTzz33nB566CH95je/ceUc1738M61uqFKvwn7627RlrtTs6K57+WeK/VeTrl/4cz0w5B5jOYZ+8jMFo1WKBvvpzcPNvTcfXzRX0ep6BfsW6ftLzm+Tc1qWpZDfUqjFCv6/vujHeuiBB9R95FjJ45XtSLYjxZOOonbqQ4GondTCl15QzwEDFerVV4GCIjUmHDUm7G2frKhM762LbTeP15IKfakPAPxeySNLHkvyNN9S4FHqazmOqkv30osrGuXzept/ruafp15jqfm55telv7aaa6SOsWRZkmPbqi3opU83xOT3JWU1v952pETSafHo7NyFdZJaX1SmD9fHFPDZ8jZ/COK1LHmtjY+e5g9EvM2ZpdSHK5uUcqREIqmYN6jaWFK+5MZrvLVjN/l+iyNS4glbUV++appseWKp39NOOtraRo4eS1t8iJP5uvkx6aTOlHQcJbdyjXb0Ec7OfsjjbFZ7i9+/xZMtf5ZIOkpaHsWTjpR0dnidWtZJf5l+Lv11LJ5U3BtQXSwpb9LO/M2k//5a/s2lv7Z24XdF7mz+d7Q1O/qb3tyO3lWnuZ7jOJnz87eA1tqZv+WWx27+N7jTr93Kc7v638jOnMN2HDmyZDuOPDtTcweH2MnmeklHnqSzMy/ZrkTSUVKWEklHVnL7lXbmPPHN/t+UjZaXqeX/6xy79e9J3E7Vi9mOkp7Wb/Uctx3ZllfRbdTb1cSJ5noR25FtuZdva/WyvZrxRFIJj09NiaQSO5FxR+dJ12tMJOW3kvr/7Z13WBTX+7efBWmCIoLSm4AgTZQWjVJsIGJJNDFqVGKJvcf2jVHsJUZjjb0QiZrExBZrEIiVJlgBFQsYBJQifYHdz/sH786PBTSzsMiK574ur8udHe49s/M8s3POnFJcUf/j/iAq/GVlZRQXF0cLFy7ktikpKVGvXr3o+vXrNfYXCoUkFAq5169fvyYiopycHKn9mjVrRoV5/7dNXCImKiUSK4m57epqqlL7SIBIRMXFxaSSl/vGfaryX/tIfKqqqpRfLiI1IhKKxLX+DZ/PU1VV5conUK59Zns+Hsl3IlISvdUn63dQ9fur6nuTJ79ITGplRMKKyu+k+n61+epzXt50fktFpSSkUoKoWY2/e9vnVT2/9Y0VorefX7X//4+I6MTSaRR6OZ6IiCrExVQmpsofBar8ARIT0e51y2nMvEUkFoFKMlJJTd+MxAIBVYCoXFz5L/3fNGreSpfrTVD01tJVRYMyn7/9WGRC3ZAeJ2XJz6eqT0+T5ejTtqV7N9Pl52tpQ/dvvZCfr7UzJUT9K1df/I3n8vPpONKtaDmWj4hI247uyvOcyPuYma/ePnnHtELnSEM4mU/hfAod00RErZ3kXEYnSpDntb+1k3x/S+T929QQvhh5+hzo9ofkIyJqZU93YuV4r1DFJyyqHH4rayNeVQSoz1+/J6Snp5OxsTFdu3aNunTpwm2fN28eRUZGUlRUlNT+wcHBtHTp0nddTAaDwWAwGAwGg8FgMKRIS0sjExOTOv3tB/GEX1YWLlxIs2fP5l7n5eWRubk5paamkra2tlw+Iz8/n0xNTSktLY1atmzJfMz3Xvsawsl8zNeYvoZwMh/zNaavIZzMx3yN7WQ+5mtMX0M4q/sAUEFBARkZGdXZ+UFU+PX09EhZWZkyMzOltmdmZpKBgUGN/dXU1EhNTa3Gdm1tbbkFh4SWLVvK1cl8zNeYvoZwMh/zNaavIZzMx3yN6WsIJ/MxX2M7mY/5GtPXEM6qvvo+cFb6713ef1RVVcnV1ZXCwsK4bWKxmMLCwqS6+DMYDAaDwWAwGAwGg9FU+CCe8BMRzZ49m0aPHk1ubm7k4eFBP/74IxUVFXGz9jMYDAaDwWAwGAwGg9GU+GAq/EOHDqWXL1/S4sWLKSMjg1xcXOjcuXOkr6//n3+rpqZGS5YsqbWbf12Rt5P5mK8xfQ3hZD7ma0xfQziZj/ka09cQTuZjvsZ2Mh/zNaavIZwNUcYPYpZ+BoPBYDAYDAaDwWAwPjQ+iDH8DAaDwWAwGAwGg8FgfGiwCj+DwWAwGAwGg8FgMBhNEFbhZzAYDAaDwWAwGAwGownCKvwMBoPBYDAYDAaDwWA0QViFvxGQ9zyJH5qPwWAwGAwGg8FgMBj/DZulvxbEYjFFRUXRw4cPiYioY8eO1L59e9LQ0GjkkjEYio9YLCYiIiUl+bQnKqIPAAkEghr/VxQfQ7FRxJiWt4/lCENWFD1mGtInL1huNG0UPaYbuozyoCF9inpOiFiFvwYFBQW0bNky2rx5M7Vp04ZsbGwoPz+fhEIhde7cmaZOnUoeHh519qekpFBkZCRpaWmRoaEhGRkZkbGxMamrqyuELzo6mi5evEgaGhrUpk0bMjExIRsbGzIxMSEi2YNP3j4Jip5gH+LNqVAoJGVlZWrWrJnUdrFYXKeKg6L7srOzqUWLFqSqqqqQvqqIRCIiqqzAySMWFdH3PuSwosc0yxHFiml5+96HHFH0mJG379mzZ6Sjo0MikYiUlZVJU1OTlJWVSSwWk0AgqNN32hBOIsVsWJS3j+VI/X0N4ZR3TMvb9z6cEwJDip07d8LKygp///03ACAxMREXLlzAhg0b0K9fP5ibm2PPnj11cm/btg3W1tawtrZG69atIRAIoKenh8GDB+PSpUvcfmKxuFF8y5cvh6WlJRwcHGBsbIxWrVrB1NQUvr6++PnnnxvdV51///0XFRUVNbaLRKIm5xMKhXjw4AFycnKQmZmJ/Px87ntrbF9hYSFCQkIwcuRIODs7w8XFBVOnTkV4eLjM5XoffC9fvsSmTZvQo0cP6OrqwszMDJ999hlCQkIgFAob3VeVvLy8WrfXNaYV3femnKtt27v0KXpMsxypiaLE9IeSI4oeM/L2paamIjg4GPb29lBWVoaenh68vLywfPlyPHr0SGZfQzmBN8dgXe/dFN3HcqTu11VFzxN5+96HcyKBVfirERgYiDlz5tT6XmlpKWbNmgV7e3ukpaXJ5M3MzETbtm2xZs0aJCcnQygUIj09HQcOHEDv3r2hrq6OWbNmNZovIyMDOjo62LNnD3JzcwEABQUFOHnyJIYPHw4VFRV88sknKC4ubhRfVRITEzF16lS4urpCU1MTVlZWGDVqFE6ePCmz633wRUVFYfjw4TAxMYFAIEDbtm3Rp08fbNq0CTk5OY3umzdvHuzs7NCnTx+sXLkS06dPh5ubGzQ0NGBra4ujR482Kd/kyZPh5OSEsWPH4pdffsG6devQr18/tGzZEtra2li7dq1MF2Z5+4DKG4zFixejf//+sLKygqurK+bPn4/Y2FiZPO+LT9FzWNFjmuWI4sX0h5Yjih4z8vaNHj0arq6uCA4OxvXr13H48GGMHj0aRkZGUFZWxpQpU2T+PZa3U9FjmuWIYsV0QzjlHdPy9r0P50QCq/BXY+7cuejTpw+ysrJqfT8nJwd2dnY4dOiQTN6DBw/CwcGBe129Ze/gwYPQ1dXFX3/91Si+0NBQKV9ZWZnU+xERETA2NkZoaGij+KrSvXt3eHh4YNWqVTh27BhWrlwJX19fqKmpwdzcHL///nuT8rm4uKBPnz745ZdfEBcXh5CQEAwZMgStW7eGpqYmNmzYIFNLtrx9mpqaOHv2LPdaKBQiJycHV69excSJE2FqaooLFy40GV/z5s1x5cqVGtvT0tKwevVqGBsb47fffms0HwAEBATA1tYWX3/9NX744QdMmjQJjo6O0NDQQLdu3RAVFdWkfIqew4oe0yxHFC+mP7QcUfSYkbdPQ0PjjRXTw4cPw8TEBLt27eLtawinosc0y5FKFCWmG8Ip75iWt+99OCcSWIW/GgkJCTA3N8eECRNw8+bNGi0pmZmZ0NLSwr1792TyhoSEwN7eHomJidw2kUgk5Q8KCsK4ceMaxXf8+HFYW1vj4sWLUr7i4mKuMWHSpEkYOnRoo/gkXLt2Dbq6usjPz6/x3v379zFp0iRYWVnh7t27TcJ3/fp16Orq1mgwASq74a5cuRImJia4ceNGo/ju3r2Ltm3b4sGDB7W+X1hYiE8//RSjRo3i1Sqp6L6kpCQYGhoiJiaG2yYSibgGEqFQiIkTJ6Jfv34oKip65z6gMgZbt26Nf//9l/OVlZUhKysLZ86cQf/+/eHk5IQnT540GZ8i57CixzTLEcWM6Q8pRxQ9ZuTtS0lJgZmZGdfIJhaLUVZWxuWXSCTCd999h65du9b6Hb8L5/sQ0yxHFCemG8Ip75iWt+99OCdVYRX+Wvjzzz9hZWUFDQ0N+Pr6Ijg4GIcOHcL//vc/uLq6onfv3jI7S0tL4erqCk9PT5w/f77GySorK0OPHj2waNGiRvEBlcMZOnTogCNHjqCgoEDqvYKCAvTo0QPffvtto/kAYNeuXXBycuKGCYjFYqmESE9PR69evTB//vwm4Tt06BDs7Oy4H62KigoIhULuApWbm4tPP/2Ud8OOvH3Z2dno2bMnRo4c+cbxaL/++ivMzc2bhK+kpASffPIJ+vXr98bhKKdOnYKxsXGj+ABg8+bNcHNzq7VRBwAePnwIDw8PrF+/vkn4FD2HFT2mWY7UpLFj+kPLEUWPGXn7xGIxxowZg65duyIzM7PWfc6fPw99fX1evoZwKnpMsxypSWNfVxU9T+Ttex/OSVXkMzVmE2PQoEH06NEj+vXXX8nMzIwOHjxIM2bMoGvXrlGPHj1o165dMjvV1NTol19+oTZt2tCMGTOof//+NGnSJNqwYQOdPHmSRo4cSc+ePaMRI0Y0io+IaMeOHdSlSxeaOXMmubi4kL+/P82bN4+2b99O/fr1o5cvX9Lo0aMbzUdE5O/vT+Xl5bR27VqqqKgggUAgNWuxoaEhWVtb07Nnz5qEr1+/fqSpqUkrVqygwsJCUlZWJlVVVW7mzlatWpGRkRHl5eU1iq9169Y0ceJEunr1Kg0ePJj27dtH0dHRlJqaSkREGRkZFBoaSp6enk3Cp66uTtOmTaPk5GTq1q0bfffdd/Tnn3/S3bt3iYgoKSmJNm/eTN7e3o3iIyLy9fWl7Oxs2rRpU63vW1tbk4ODAyUkJDQJn6LnsKLHNMuRmjR2TH9oOaLoMSNvn0AgoGnTplFxcTG5uLjQiBEjaNu2bXTt2jUqLy+niIgIWr58OfXt25eXryGcih7TLEcUK6YbwinvmJa37304J1LUqZngA0QoFCIjI6PenpcvXyI0NBQTJkyAv78/XF1d0bp1a3h7eyM6OrrRfUBlt+81a9ZgxIgR8PDwgIWFBUaMGIHk5GSF8O3cuRNGRkbo1asXVq5cidOnT+P+/fsAgBMnTsDMzAxHjhxpMr5jx47B1NQUtra2GDduHHbv3o3o6GgIhUJs374dZmZmMo3pkbcPqOwVExAQAHNzczg7O2PAgAHo3r07NDQ0EBgYKPMkOorui4uLw9SpU+Hh4YFOnTqhe/fuMDMzQ7NmzTB8+HDe3fQawicSibgVMsaMGYPQ0FDExcUhNTUVQGU+WlhY4JdffmkSPkDxcxhQ/JhmOaI4Mf2h5ogix0xD+LKysrBhwwYEBgbC3d0dDg4O0NDQgLq6OiZNmvTGYTbvwqnoMc1yRDFjuiGc8s4Tefveh3MCAAIAqFtTwYcJZFxHs/r+ubm5pKmpSaqqqpSTk0O5ublkaGhIzZs35+WXt68qr169ooyMDDIwMCA9PT0Si8VUUVEhtQ5kY/qqcubMGTpy5Ajdu3ePBAIBNWvWjB4+fEhlZWU0e/Zs+t///kdqampNxpeYmEiHDx+mqKgoysnJoby8PHry5Anp6urSggULaMqUKTXW63yXvqreS5cu0b1790hbW5ucnJyoZ8+epK+vL7PrffA9f/6cYmJiKCUlhVq0aEHt2rWjLl26kJaWVqP6KioqaOfOnXTkyBF69uwZtW3blgwMDOjZs2f0+PFjGjt2LK1bt47U1dWbhI9I8XNYgqLHNMsRxYjpDzlHFDVmGsqXn59PSUlJ9Pz5c9LU1CQ9PT3q2LEjNWvWrE4+eTkVPaZZjihuTDeEU955Im+fop8TVuF/B9y5c4cWLlxIt2/fJlNTUzI1NSU7Ozvy8fGh7t27k7KycqP6rly5QtOnT6fU1FSysrIiTU1NMjU1JR8fH+rTpw8ZGxs3qu9tPH78mOLi4ig9PZ2MjY3JwsKC3NzcmqwvJyeHkpOTKTs7m3R1dUlPT49sbGwazZeSkkLh4eHcOW7fvj21bduWiOrWmKPovujoaLpw4QKpq6uTvr4+2drakqOjI9fAJivy9tVGfHw8RURE0JMnT8jS0pIcHBzI19eXVFRUmqRP0XJY0WOa5Yjix3RTzxFFjxl5+06fPk3nzp0jVVVV0tHRIRsbG3JzcyNra2tuH1lzryGcVVH0mGY5IhsNcV1V9DyRt+99OCdVYRX+BiYhIYGGDRtG5ubmNHjwYHr8+DElJSXR48ePqaCggHr06EGrVq3ibrDetS82NpY+++wz8vX1pU8//ZSePHlCT548oQcPHlBaWhpZWFjQxo0bqV27do3iq87Ro0cpPj6eWrVqRaampuTg4EBOTk4yN3K8D77y8nLasmUL3b59m7S1talNmzbk4OBA7u7uZGJiwu3H9wIlb9/27dtp48aNRERc75JWrVqRl5cXTZs2jXr27NmkfCtWrKB9+/ZR8+bNKS8vjwoLC0lTU5OsrKxo3LhxNGrUqEb1Sfjnn38oLi6O2rZtS5aWlmRvb0+tWrWqk+t98ClyDit6TLMcUcyY/pByRNFjRt6+b775hv744w8yNDSk8vJyEgqFJBKJqGXLltSvXz+aM2cOqaury3SeG8Kp6DHNckRxYrohnPKOaXn73odzUoM6DQRg8GbKlCkYOHBgjRkXs7Oz8fPPP8Pa2hqBgYGN5ps6dSoGDhxYY3bmvLw8nD17Fp07d4a7uztKS0sbxSchPz8fI0eOhKGhIdzc3GBiYgIdHR2YmJjA399fpjWi3wdfXl4eBgwYAHNzc/j7+6NLly7o2LEj7O3t4ePjg927dzeqLzMzE23btsWaNWuQnJwMoVCI9PR0HDhwAL1794a6ujpmzZrVZHwZGRnQ0dHBnj17uFl3CwoKcPLkSQwfPhwqKir45JNP3jizakP7gMrl02bMmAF9fX3Y29tDR0cHAoEA+vr6GD16NG7dugUA3CzB77tP0XNY0WOa5YjixfSHliOKHjMN4WvRogWOHTvGbSsqKkJERATmzJkDHR0ddO7cGVlZWbx8DeFU9JhmOaJYMd1QZZRnTDeET9HPSW2wCn8D88knn2D27NkAKicZKSkpgUgk4t4PCwuDjY0NTp8+3Si+UaNGYcKECdzr8vJyqQtlYmIi7O3teU8sIm+fhD179qBDhw5SF8uXL1/iyJEj8PPzg6qqqkxL/Cm6b/fu3ejQoQNu3rzJbcvNzcW5c+fw1VdfQVVVFSNGjGg038GDB+Hg4MC9rt7Ac/DgQejq6uKvv/5qEr7Q0FApX/WlgCIiImBsbIzQ0NBG8QHA/v370aFDB4SGhiI7OxtisRjPnj3D1q1b4e7ujlatWmHXrl1NxqfoOazoMc1yRPFi+kPLEUWPGXn7jh49KuUrLy+Xev/p06ewtbXFtm3bePkawqnoMc1yRLFiuiGc8o5pefveh3NSG6zC38AcOHAABgYGCA8Pl9peNeAsLS1x4sSJRvGdOHEC2tra2L9/v1TDQdVKuqWlJf74449G8Un44osvMHXqVACVDR3VE3bjxo2wtrZGXFxck/CNHj2aazgRiUQ1bsaPHz8Oa2trRERENIovJCQE9vb2SExM5LaJRCIIhULudVBQEMaNG9ckfJLv5+LFi1K+4uJi7rucNGkShg4d2ig+ABg8eDCmTZvGva56jsvLyzFz5ky0b9+e9+oYiu5T9BxW9JhmOaJ4Mf2h5Yiix4y8fWFhYbCwsMD+/fu5bRUVFSgpKeF8s2bNQv/+/Xn5GsKp6DHNckSxYrohnPKOaXn73odzUhtKdRsIwODLZ599Rh9//DH17duXvvzySzp79iwJhUJq1qwZvXjxgrZu3UqvX7/mxkr+F4MHD5arr3///jR+/HiaPn06+fv70759++jJkyeUl5dHt27domXLllFxcTH16dOnUXwSbG1t6dq1a5Senk5KSkrUrFkzEolEVFJSQkREkydPJm1tbbp582aT8Lm4uNDFixcpNjaWlJSUSFlZmSoqKqi0tJSIiAYMGEA6Ojrc+pzv2vf555+ThoYGBQUF0YULF6i4uJiUlJS4mf3Ly8spNTWVDAwMmoRv4MCBZGdnR9OnT6ejR49SYWEhKSkpkYaGBikrK1NhYSElJydLTf7yLn1ERMbGxpSYmEgikYiIiJSVlUkkElFZWRk1a9aMgoODSUtLi2JiYpqET9FzWNFjmuWI4sX0h5Yjih4z8vb16NGD+vbtS0uXLqVNmzZRbm4uKSsrk7q6OikrK1NWVhbFx8eTs7MzL19DOBU9plmOKFZMN4RT3jEtb9/7cE5qpV7NBYy3UvWpdkhICLy8vKCtrQ1VVVWYmJjAwcEBNjY22LBhg8zugwcPolu3btDS0oKqqioMDAzq5QsLC8OQIUOgq6sLgUAAPT092NraokOHDjhw4AAvR9WW1gsXLmDIkCFo06ZNnX1VSU9PR/v27eHr64uzZ8/WmAMgJSUF2traUl3W32dffn4+fHx80LFjR4SEhEiN3RGJRLh9+zZ0dHQazQcAycnJCAwMhJ2dHXr06IGJEyfihx9+wIkTJzB06FBYWVlJPT3k4+vfvz86dOigkL7nz59j3LhxMDIygpWVFfz8/DB37lxs27YNXl5ecHJykmn91rS0NIwZMwZGRkawtrauty8xMRE6Ojr4/PPPceXKlRrdwv7991+ZYlDRfYqewwDLEZYjLEf+i+fPn2PMmDEwMDCQS8xUjUF5xIy8fa9evcKiRYtgaGgIFRUVtG/fHqNGjcLChQtha2uLjz76CI8fP+btAyq7jC9cuBAGBgb1dip6TLMcUbwcaQinvPNEnjkiOV5FPyfVYbP0v0NevHhBqamplJaWRhkZGVRaWkoBAQFkb2/P6+/j4uJIT0+PzM3NiYioqKiIUlNTKSkpifLz8ykjI4MCAwPJwcGBl6+0tFTqiQ5R5UzNT548oYcPH1J5eTn17t2b9xMeIqLi4mJuCQmRSERZWVmUnp7OrTvau3dvMjQ05O0j+r9ZKePj4+m7776juLg40tHRoXbt2pGTkxOpqanRb7/9RmZmZnT27FnevoSEBPruu+/o5s2bpK2tXWefWCwmJSUlSkhIoMWLF1NcXFy9fJLyPX78mFauXEnHjx+n4uJibgkRZWVlOnfuHHXr1o0OHz7Mu3wpKSm0fPlyOnXqVL18VXn16hVduHCB/vnnH3r27Bm9fPmSnjx5Qk5OTvT999+Tu7u7TL7U1FT6+++/KTo6mtLS0ujly5eUkpJSZ196ejqdP3+eoqKiKDU1lTIyMujZs2d19lVUVNA///xD0dHRdPfuXXrw4AG9fPmSPv74Y1q8eDG1b99eJh8R0aVLlyg6Opru379PSUlJlJWVRd26dZPJJ4mZsLAwWrJkCb1+/ZpMTEzIysqKbGxsSEtLi/bv308A6OrVq7x9ly5doiVLllB+fj4ZGRnV2SeJQXn75JVzEp/kGhMfH18vX1VYjrAcYTny39y4cYMiIyPpzp079PDhQ8rKyqpzzJSVldHly5cpJiaGEhMT6cGDB5SZmUldu3ats+/KlStcDCYmJlJ2dnadfUREjx49otjYWLpx4wbduXOH8vLyyN/fnyZPnlznpYtTUlIoJiaGoqOj6datWzI7WY6wHKlPTL8PeVLfHKmOop+TqrAKfwNSWFhIt2/fJiUlJXJzc6NmzZrV2ZWRkUGdO3cmX19fMjIyIkdHR3JycqJOnTqRsbExXb9+nWsI4ENOTg5t2LCBhg8fzjU4vHr1ikpLS6WWaOOLSCSiy5cvU0hICAGg6dOnU6dOnUgkEtV5WZPaqKiooIiICLp69SolJibSo0ePqLS0lIYNG0YTJkwgPT09mXy5ubl0+fJlSkhIoOTkZHr06BEVFxfTF198UWfflStXKD4+nkvYsrKyOvuIiO7cuUM3b96k6OhoSkxMJCUlJfryyy9pyJAhpKWlVWdfbGws3bt3TyYfqi0JkpubS5qamqSqqsot62VoaMg1+lTf/798VUlLS6OioiIyMzPj7SOqjJHquQaAsrOzqbS0lFq3bi2T78GDB2RqakoaGhpS28ViMQkEAqm/5+MjInr+/Dm9ePGCXFxcuHWCqzaWyeqrytOnT+nChQt048YNSk9Pp/T0dMrKyqLBgwfT/PnzyczMTCbfvXv3KCwsjBISEigzM5P+/fdfysjIqLMvKSmJLl68SPHx8ZSenk7//vsvvXz5ss6+vLw8unz5Mt28eVMu14TCwkIKDw+XymGhUMjbx3Lk/2A5wnLkv3j16hVlZGSQgYEB6enpkVgspoqKCqkHEXU5xxJyc3NJIBBwS7bx9RUVFZGmpmaN7WVlZaSioiJzTMfExJCBgQHp6+tLHVtt8D1eAHT79m1KSkqiLl26vDUuZP0OWY6wHOHjU/Q8aagcUeRz8kbq1T+A8UYSExMRFBQEgUAAgUCAzp074/r161L7lJWV1ZjQ401ERUVBRUUFXbt2RdeuXeHo6AgvLy8EBARAIBBg27ZtSEhIQFFRES/f9u3b4eDggJycHJSWlmLHjh34+OOPYW1tDVtbWwQHByMvL4/38Z46dQouLi5wd3eHl5cX+vfvj/j4eEyePBkODg6YPHkyHj16xNtXHbFYjAcPHuDWrVvIyckBAKlJAeuLSCTivgtZyc/Pr7Gt+qR4shAREYHMzEy5HV9BQQHCwsIQGhqKwsLCevtu376Nfv36wdTUFF27dsXQoUOxZMkShIeH1+m4Hz9+jHHjxuGjjz7C0qVL8eLFC6n3JUNj+C6zs3r1apw8eRLJycnIzc3lhi/s2LFDys3XZ2dnh71796KiogJlZWWIiIjATz/9hK1btyI6Olrm87R9+3a0aNECbdq0wbBhw1BcXIycnBz89ddfiIyM5MrLt3xVKS4uxsuXL7nrysuXL5GWliazpzYqKirw9OlTPHnyROa/rd7tEqicYCkvLw/p6eky++7evVursy7fGQCUlpYiKSkJ165d47ZVn8BOFliOsByRlQ8tRy5fvoxOnTpBV1cXHh4e8PX1xahRo7Bv3z48f/5cZt/t27fRs2dPtG/fHpMnT67RPVfW4540aRI2bdqECxcu4MGDB9x9R3BwMFJSUmT2Wlpa4tChQxCLxSgoKMDRo0cxd+5czJkzBydPnkRJSYnM5Vy6dCmaN28OW1tb9O3bF3l5eUhJScG2bdtw9OhRbokvWY+d5QjLEb4oep7IO0feh3PyJliFv4EYNWoU/Pz8cPXqVWRmZiIwMBABAQEoKiribvjOnz/Pe5mFkpISLF++HFOmTEFeXh4iIyOxceNG2Nvbo0WLFnB3d4eHhwd++OEHXj4/Pz/Mnz8fALB+/Xp4enoiKCgIhw4dwqJFi+Dg4MDbJfFJlgvMz8+Hv78/3N3d4evrixUrVsDJyQnTp08HIFsA5+fnY8+ePWjTpg309fXRpUsX9OjRA19//TV+//13Lln5UlJSgmPHjsHOzg7t27fH0qVLkZ2dLbWPLOW7f/8+goKCsGvXLoSHhyMlJYVrNJg5c6bM62Zeu3YNVlZWiI2NBQA8evQI27Ztw7hx47Bw4UJERkbKdPOcmpqKESNGoFmzZrC1tcXMmTNRUlKCa9euYeXKlThz5gx3AeVDfHw87Ozs4Ofnh127dmHBggUYNGgQnJ2dYWlpibFjxyIzM5O3LzExEd27d4eLiwumT58OBwcHbjUByZKOIpEIr1+/5uWLiYmBQCCAuro6WrVqBT8/P3z//fc4f/481NTUcOPGDd5lA4DIyEgYGhqipKQEBQUFWLBgAQQCAezt7dG1a1d4eXkhLCyMt+/69euwsLDAjh07cPbsWbi6umL37t3Q1dWFi4sLunbtiu3bt8tUxuLiYvz5559wdHSEsbExunfvjuHDh2PNmjW8ZwKuSllZGcLDw9GnTx94enrip59+qtGoJUuOPHr0CMHBwbhw4QJSUlKQm5vL/f369et5N3pKuHHjBmxtbXH58mUAlWMm//zzT6xatQo7d+7E/fv3ZfK9ePEC06dPh0AggIGBAZYvX46ysjI8fvwYoaGhiImJkamMLEdYjgAsR95GTEwMLCws8NVXX+HUqVPYvHkzZs2ahX79+sHZ2RkDBgyQqiz8F3fu3IGrqyt69eqFtWvXwsnJCYGBgQD+78FARUUFbt++zeu8JCQkQCAQwMbGBhYWFvDx8cHs2bOxc+dOCAQCnDlzptaK4puIjIyEvr4+KioqkJubiwkTJqBZs2bw9/eHn58fPDw88PPPP/P2AcDVq1dhYWGBY8eO4erVq/D29sby5cuhra2N7t27o3Pnztw9Hh9YjrAckSVHAMXPE3nnyPtwTt4Gq/A3ELq6urh69Sr3+tGjR7CyspK6UXF1dcWiRYt4O0UiEXr27IlJkyZx2zw9PTFr1iz89ddfmDBhAu/l+Hr27MlV6B0cHKTWLS0vL8eSJUvQrVs3PH36lJfP0NCQu3ACgJmZGZYuXcol+5YtW+Ds7Iz4+HhePgkrVqyAra0tVq1ahXPnzmHz5s2YNm0a+vTpgw4dOuDrr7/mfaMLVD45cnR0xPjx47Fy5UpYWlrif//7HwBwrbAZGRm8l3RZu3YtBAIBbG1tYWlpid69e2Pp0qWYMWMGWrRogUePHsl0kZ82bRoGDhwIoLLy37NnT7Rs2RKfffYZevbsCW9vb1y5coW3b/78+ejZsydu3LiB48ePo2fPnpg6dSp0dHTg6+sLV1dX7N27l7dvypQpGDhwYI2GjOzsbPz888+wtrbmLlh8mDVrFj755BPuqeLJkydhaGiILVu2cPucOnUK9vb2vHz379+Hv78/1q1bhwsXLmDQoEFo3bo1BAIBlJWVMXToUGzbtg23bt3i5fv222/h5+cHAPjpp5/g4uKCsLAwlJeXIyoqCkOHDoWxsTHvCtz06dPx+eefc68XLlwIfX197Nq1CwkJCZg6dSq0tbVlusH64YcfYGNjg8mTJ+PgwYOYP38+AgMD0aFDB7Rr1w6LFy+GUCjk/YOxf/9+ODo6IjAwEJMnT4aNjQ22bt0KAFxjVn5+Pu/eRIsXL4ZAIICGhgb09PQwYMAA7Nu3D6tXr4axsTFevXrF+1iByu9QkiPx8fEYMGAABAIBPDw84Onpif79+/POX6DyHHft2hXnz5/Hnj170KtXLyxduhRaWlpwdXXlJl/iC8sRliMAy5G3MXXqVAwcOLBGb5e8vDycPXsWnTt3hru7O+8ed7NmzcKQIUO4XolxcXFo164d99sOVOaIjY0NL19aWhoGDRqEqVOn4urVq5g2bRrc3d2hq6sLNTU1BAYGYsGCBTh9+jQv39KlS9G7d28AlfdCbm5uiI+Ph0gkQnJyMmbNmgU9PT2ZJuWaMWMGPvvsM+71ypUroa+vj99//x3Pnj3DmjVr0Lp1a6mlvt4GyxGWI7LkCKD4eSLvHHkfzsnbYBX+BuDBgwdo165dja4ZISEhaNu2LTIyMiAWi6GlpYU7d+7wckousqWlpfD398e8efMAAGpqaoiOjpa5jDt37kT37t2Rl5eHXr164fjx41Lv5+TkwNDQkFdrVXZ2Nry8vLBy5UqUlZUhLS0NAoFA6tiys7NhaGiIhw8fylROGxsb7Nixo8b2rKwsHD58GG3atMGUKVN4+xwdHbF9+3auEn7kyBHo6enhr7/+4vYZM2YM77Wno6Ki8PHHH2Pfvn24ePEivvrqK3Tu3BnNmzeHjo4OvvjiCyxYsIB3Q4e3tzdWrlwJAOjXrx/Gjx+PrKwsCIVC3L17F7169UL37t1rHUZQG3Z2dlKrInh6esLf3x/Xrl3D06dPMXLkSDg6OvJulfzkk0+4nhwikQglJSVSPQ7CwsJgY2PD+wLfsWNH7Ny5U2rbpk2bYGNjwzU2TZgwQaYKUlRUFJydnaW61Q0aNAidOnVCjx49IBAIeMfMzp07MWjQIACVvXYWLFgg9X5hYSF69OiB3bt38/L1798fwcHB3OuAgACpspSWlsLPz0+m3jXW1tb46aefatyIpaenY+PGjdDR0cGyZcsA8Hui4uTkhB9//BEFBQUoKyvD+vXrYWBggKioKG6fMWPGYOLEibzKd/r0afTq1QunT5/Gr7/+il69eqFly5YQCATQ0tLC119/jZ07d/LuMurp6Ynvv/8eQOV5/eKLL/D06VMUFRUhMjISHTt2RP/+/Xl3nbezs8PBgwe5166urvDx8cG5c+cQHx+PHj16wN3dHVlZWbx8LEdYjgAsR97GqFGjuF4qwP/1VJGQmJgIe3t7HDlyhJfP1dW1Rq+PX3/9FWZmZkhISAAAfP311zKtb//8+XN4enpKleHTTz+Fp6cnRo8eDT09PSxcuJCX68iRI+jTpw+Aykb9OXPmSL0vFosRGBiIjRs38i7f559/zt0HAoC/vz8mT54stc+gQYOwYsUKXj6WIyxHZM0RQLHzRN458r6ckzfBKvwNQHR0NLp06cI9ba96sxcQEICgoCDExcVBQ0NDJq/kwhMTE4PPPvsMs2fPhq6uLnJyciASiWTq6v3w4UO4urpi2LBhGDFiBPr16yc1/iQkJAS6urq8fT/++CPatWsHf39/dOvWDc7Ozti3bx/3/sWLF6Gnp8fbB1R2CevcubOUp/rT8iNHjqBDhw5ISkr6T19RURF0dXVrdNWaM2cO3NzcuJZmyRgivly6dAkODg64d+8et01fXx+jRo3C0KFD0bJlS1y6dImX67vvvuMaG1xdXXHq1Cmp91NTU+Hk5MSrkUcoFMLZ2Rl//vknt61FixY4efIk9/rFixdwcHDg3WvgwIEDMDAwQHh4uNT2qufF0tKSV08ToVAIT09PqWEtkhjv1q0bpk2bBgCwsLDgfQGV5MCZM2fg5eWFyMhIAICRkZFUmfjOT/Hvv//CwMAAGzduxKZNmzBx4sQa8yC0b98ex44d4+XbsGEDHB0d8ffff2PXrl1o3bo199RD8h3K4hMKhbC1teX2LysrQ2lpqdSP0Pr169GxY0deYyaLioqgra1dowV92LBh8PPz47yWlpYyda3bu3cvOnbsiIKCAm5b27Zt8emnn6JLly4QCAQ4d+7cf3rEYjGmT5/O/Yjb2triwoULUvskJCTA2dmZ1xPq0tJStG/fXqqFX1tbWypHHjx4AHt7e6kb1bfBcoTlCMBy5G2cOHEC2tra2L9/v9R9S9VzYmlpiT/++OM/XWVlZfDy8qrRSxGobHwbNmwYgMocOXz4MK/yScoUHR0NX19f7lgNDQ2lylT1XL2NvLw8mJubY/r06Vi3bh1GjBghNRxRLBbD2toav/32Gy8fAOzbtw9mZmbYvXs3goODYWhoiHXr1kntY2Njw8vJcoTlCCBbjgCKnyfyzBHg/Tgnb4NV+BsAkUiExMRE7mm2WCzmbtIiIyO5bh99+/at82eEhISgefPm+Oqrr+o8tuPhw4cICAiAkpISBAIBnJycEBQUBDc3Nzg5OWH16tW8XcXFxTh06BAmTpyIsLAwbN26FV26dMHp06exatUqeHp6YurUqTKVTywWY+nSpTAyMnpjhfTp06dQV1fn1S0sPj4enTt35lyShC0oKICDgwM2bNiAV69eQSAQ8J7cTuLYv38/+vTpg5SUFLx48QIqKipSN8x8z9G1a9ego6ODjRs3Sv0gSXjw4AE0NDR4lU8oFGLmzJno1KkT9uzZg4kTJ6JVq1bcDT5Q2YDA1wdU/pAPHjwY6urqGDFiBM6cOcN1X0pPT8eWLVvQunVrXr78/HwsWrSI675U9Tu6cuUKjI2NcebMGSgpKfH+wZB4xGIxFi1ahKCgIDx+/BgqKipISkqq04Rpp06dgrm5OWxsbCAQCLB48WJERUXh2rVr+Pbbb2FiYsK7W2JaWhrc3d3RvHlzeHl5Ye7cufD29uaO/ejRo9DX1+ftKysrw6xZs+Do6FjjyYbkWJ89ewZNTU1ezpiYGLi4uHDdpSXxnZmZCTMzM/z888/IycmRKUckrFmzBoGBgcjIyOBypKqDb46cPXsWenp6+PPPPzF79mysWrVK6n1ZjreoqAhjxoxBnz59EB4ejoULF6J58+aIiYnhyvT8+XPePqDyeiKvHHn9+jX+97//cU9I5JEjkvyQZ46cPHlSbjny9OlTqRyZN2+ewuVIp06dFD5H2rRpo7A5IhaL8c0336BFixbo3bs39u7di8ePHyMnJwcJCQlYunQp9PX1eX1/RUVFWLVqVa1zBN25cwempqYIDQ2t0+8IUNmbpn///rh9+zZUVVWRnJxcpxyJi4uDm5sbLC0tIRAI8OWXX+L333/HH3/8gYkTJ8La2pr39wcAubm5GDp0KNq1a4eAgACsX78eTk5OePbsGXJycrBr1y4YGxvzcgqFQsyYMUNuORIdHQ1nZ2e55YhknhJAfjny119/ye13pLCwEEFBQQqbI8XFxVi9erXcc6SqSxHzJC8vT245AlTG8dy5c6GpqQlfX1/s2rWr3tetGTNmAKj8nZJ8X/U9J2+CVfjfIZKTOXPmTAgEAt4T9r2JlJQUmSaIeBOvX7/G+fPnMX78eHh7e2PSpEk4efJkvWaaFwqFWLhwITQ0NGBmZob58+fj2bNnMnsyMzMxYMAAtGzZEkOGDMGRI0eQmpqKrKws/PXXXxg5ciQ8PT15udLT0zF58mSEhIQAkG6I2bt3Lzp37ozVq1fD1NRUpjKKRCKUl5djypQpWLBgAXbu3In27dujpKSkTt/hoUOH0L59e2hra0MgEGDu3Lk4fvw4tm3bhgEDBnBjnPhw69YteHl5wcrKCnPmzOHGF5eUlOD+/fuYMWMGunTpwstV9UciJCQEXl5e0NbWhqqqKkxMTODg4AAbGxts2LCBd/kKCgrw8uXLGn6gcvZXgUAAFxcX3r6qlJSUICgoCKqqqnB1da3TCgySm4yrV69i/PjxcHBw4LoQWllZwcnJCUePHpXZGx8fj0ePHiE/Px8fffQRjI2N4eXlBUtLS5kmkQEqu5G5uLhAW1sbkydPlprYMTY2FhMnTkSnTp14uVJSUjB27Fj8/vvvACqPXxLD69atQ9euXbFlyxaYmJjwLp+kcvn69WsMHz4cP/74Iw4cOAAbGxsUFxfXKUdWrVqFtm3bcmMFf/rpJ9y8eRPHjh3D8OHD0a1bN96uv//+GyYmJtDU1MT48eMxbNgwbp6UrKwsLFy4EG5ubjKX8eDBg+jWrRu0tLSgqqoKAwODOuWISCTixqdWVFRweSISieqdI0VFRRg5ciRUVVXRsWPHOuWIhCtXrmDs2LGws7ODkpIS1NXV65UjcXFx3KzPH330EfT19dG9e/c65cj9+/fRsWNH6OjoYMqUKYiIiKhzjjx8+BBjx47lngqJxWLuyUxdcqRqA0x+fj6GDx+OjRs31jlHJPGxcuVKGBgYQE9PD6qqqnLJES0tLbnmSFhYGIYMGQJdXV0IBALo6enB1tYWHTp0kBqKxgfJOai+YsV3333HrZRUV+bMmcM56pMj9+7d474rgUAATU1N2NrawsPDg3cvwOo8ePAAL168QEVFBT799FO0aNECTk5OsLS05NVVWfI9paSkSOVIfX5HXrx4gcmTJ0vlSH1/RyQeSY7I43ekao7U93fk8uXLcsuRqsdy4cIFDBkyBG3atKl3jki8kvNanxyprWHlm2++qXOeVG+EWLhwIdzd3eucJ1W/w8TERLx48QJisbhOOVK9fLGxsZg0aRLXKKGjo4P27dvX6ZxUn9RQ8jmLFi2q93WrOqzC3wgUFhbiwIED3PIVioQ8l4CQ+GSZVK8qVRP2yJEj8PPzQ6tWraCkpAQDAwNYWFigb9++UuNQ+VB96INYLEZZWRkGDhwIgUCA7777rk7lffXqFQIDAyEQCPDNN9/UySH5/q9du4aZM2dyFcw2bdrAxsYGo0aN4j1hSdXjlLS2Pn36FPb29tDT04O5uTk8PDykup3JQnp6Om7cuIHffvsNW7Zswffffy81rEFWqp+X+Ph46OjoSE1gIivZ2dmYOXMm9u/fX2eHhPLycjx79gw3b97EH3/8gaNHj/Iej/c20tPTsWLFCowaNQoHDx6s042LUCjEhg0b4OrqCk1NTaipqcHQ0BBmZmbw9vbG+fPnebvy8/O5lRuqnpNXr17Bx8cHAoGgzuckMTERvXr1gkAgwMyZM+vkkHDmzBmMGDECFhYWEAgEaNGiBUxMTDBo0CBulQtZSEhIQGFhIaKjo2FoaAgrKyvY2trCxcWFd3f52NhYqYlOCwsLcf/+ffzxxx84cOAA1qxZg7t37/IuU0xMzFuX+omNjUWrVq14n4/Y2NgaT/DKy8sxa9YsrpFHFmJjY5Gamiq1LTc3FwkJCbh48SJ++eUXmVYkqP79ScjKysLmzZsxYsSIeuXI+vXr0alTJ7Ro0QJqamowMjKCubm5zDny+vVrLkckZRGLxcjOzoavr2+9cuT+/fvo06cPBAIBZs2aVSeHhFOnTmH48OGwsrLicsTMzAyffPJJnXIkLi4O+fn5iI2NhaGhIaytrWFnZydTjpSUlNRYpiw7OxuxsbE4fPgwQkJCaiw7+TaKioq4ykVt9y+SOZX4NhKVlJRwjQdV42zz5s28j7G6r7aZyl+9eoW///4bERERMq/mU1xcXGuFqri4GHv37sWcOXNkmiAuIiICQGXlY+3atXBzc4OWlhbU1NRgamoKCwsLmXLk6tWr+Pfff2tdQi03N5ebJ4Rvjvzzzz9Sx5uUlMT9jtTlXisyMpIr2/Hjx/Hll19ylbdWrVrB3Nxcphyp2msyOjoaRUVF3O+IpCIoS44AkHrqXFFRgfT0dC5HDh48KPMyhG976ixrjkioHrdisZjLk7oumVqd7OxsXLx4sU55UtsxS3Jk9uzZMuWIWCxGeHg4tm7dimXLlmH//v24cOECEhIScO7cOYSGhsp0TioqKnDp0iVs27YNCxYsQEhIiNRw45SUlDqdk7fBKvyM94bS0lKkpaVxFS5ZbtD4cO7cOejp6UmtrsAXScUoOzsbK1euxO3bt+VSJrFYjFevXiEiIoLXPAVvo2oZd+/ejWXLlvFehQGofBp/9epVXL9+XeYlcP7LV/0GsOpTB76z777JV15eLtPSg7X56vNUpzZfXdZcro0nT55gz549+Pnnn1FSUoKMjAzcuHEDJ06cwNGjR/Hzzz/zHosNVK73fvDgQYSEhNS6dvDhw4ehqqqK69ev8y7f/v37ERISwh1zUlISJk6ciEePHvEuV1Xfvn37uHGf5eXleP78Oe7evYvIyEj8888/MlUGHz9+jN27d9eYsyM5OZkbViPplvlfvHjxAoaGhhg+fDi++eYbHDhwAHFxcRCLxTA0NJQp12rz7du3j4ttQ0NDbmzp06dPeeVIbb4rV65AJBJBT0+P99jSN/n27t2LqKgolJeXw9jYWOY1iav79uzZw+WKsbExEhMTZfIBlde6PXv2SD39LSgoQHh4OE6cOIGQkBDs3buXd468evUKe/bsqbG9aoUuJCQEKioqvHJEUr7q14GbN29iwoQJMjegSnzVG08fPHiAO3fu4MKFC/j7779550j1768qycnJWLNmDSZNmsQ7R7Kzs/Htt99KHVd91nivzZeVlVWjN2FJSQmvbra1+V68eFGjUas+5Xvx4oXM14L/cmZmZtZpXXvJ31pYWODHH3/ktqWlpeHKlSs4duwYduzYgX379vF+cJOZmQlzc3Nu3o3a+Pnnn3n/jkjKV3VVEgC4e/cuxo4dyzv2qvs2bdrEbSsqKsLTp08RHx+P48ePIywsjPecWJLjrTqRnORv7927h8WLF8uUIxUVFQgPD8dXX32FoKAg3Lx5k9teF97kq94IVVpaynsI0pucdb0vrO6TfFe13YPUxSdpuKnrfdzr168xf/58aGlpoXPnzjAzM4NAIICFhQUGDhzIe0Wbqr558+bV8Jmbm2PAgAHcMrZCoVDmYWFvg1X4GQrJ69ev8erVq7dedMvLy3nPVi/xvemiKbnh4nuT+rbylZWVQSQSoaCggHey8jleyTrX8vIB/CdTSUxMRFBQEAQCAdfNqPqPdVlZGe8LfnWfq6trjQpHeXk57x9dPj7J0Iu6+qquTy7pxivP4xUKhTL9YF6/fh3+/v5QU1ODiooKxo8fj7y8PBw5cgTTpk3jPQt8VZ+fnx/nGzNmDDIzM3Ho0CFMnTqVq2TzfUJdvXxjxozBixcvcPToUYwbNw4HDhyo1/F+9dVXePXqFUJDQzFjxgycOXNG5uOt7svMzMTBgwcxY8YMmbuhR0VFQUVFBV27dkXXrl3h6OgILy8vBAQEQCAQYNu2bUhISOA9XrCq7+OPP4aTkxN8fHzQr18/CAQC/PTTT0hISODdmFXd5+zsDB8fH/Tt2xfKysrYuXMn7ty5U6/yeXt7c8e7detW3Lp1S27Hu2PHDsTHx8s0znnjxo3o2LEj9/rRo0f4/vvvMW/ePKxdu1bmYWYbN26Es7Mz9zo5ORkrV67E7NmzsWbNGu7Gj+/qO9XLl5ycjGXLlmH+/PkIDg6Wef3z6r6kpCSsXr0ac+bMwffffy9zxbq6LzExEStWrMDMmTPx/fffy7xW+fbt2+Hg4ICcnByUlpZix44d+Pjjj2FtbQ1bW1sEBwfL1EBZm69bt25cz4Ply5dLTfpVV5+VlRU6dOiAFStW8L7n+K/ySY5X1h6QfJyy9B7dvn0714uwd+/eMp/T//K9KRf45kh1X9WHKaWlpTJNVv1fvrpQ3Sdr5a86p06dgouLC9zd3eHl5YX+/fsjPj4ekydPhoODAyZPnixTY3l134ABA5CQkIApU6bA0dER06ZNk3n1rDc5J0+eDCcnJ0yfPr3GKmWy+uLj4zFlyhQ4ODhgypQpcjtmiU+WJTB37twJR0dHrmHjzJkz8PDwwNSpU+Hj4wMLCwuZhuTU5vP09MS0adPg4+MDa2trrtIvT1iFn6GQzJs3DwEBAdi4cSP++ecfpKamorCwUOrivmXLFt7duar70tLSUFxcLPVkZdOmTfXyFRUVSTUobN26td4+RTneUaNGwc/PD1evXkVmZiYCAwMREBAgdcznz5/nPS8FH5+km1R9fZJKpbzL15g+oHIm1xEjRiAvLw/37t2Dn58f+vbtCyMjI/Ts2ROWlpa1Po3k6/P390e/fv1gbGyMnj17wsrKCnv37q23z8jICL1794a1tbXUChx18QUEBMDY2Bg9evSAhYWFXMpX9Xj5LiUHVDbILV++HFOmTEFeXh4iIyOxceNG2Nvbo0WLFnB3d4eHhwfv5eTepU9LS0uhy1cXH1C5zOncuXMBVC7r5eHhgbZt28LDwwM2Njaws7OTqUdXdZ+npycMDAzg6ekJGxsb2Nvb8+798iafvr4+3N3dYWdnBwcHB6mGxvocr6QSLMsQOD7Hy3eFFwDw8/PjuqiuX78enp6eCAoKwqFDh7Bo0SI4ODjIdH4/NF9DOL28vBAcHIyoqCjY29vDysqK6zEGVDZEy1KprupzcHCAtbU1QkNDuaer1Zcyq4vv0KFDXPduWZ98Vz9eiU9yvNVXJ6hL+UJDQzlfeXm5TGX08/PjlnXNz8+Hv78/3N3d4evrixUrVnAVaoBfD0E+PslqL3yP+23O5cuXN0gZG9MXGBiIJUuWSG0bP348Fi1aBAAYPHgw/P395erz8/Pj7eMLq/AzFJJWrVqhffv20NHRgUAggKWlJUaOHIk9e/YgLi4O6enpcHR0lFqvub4+BwcHufrkXb7G9Onq6krdGD969AhWVlZSa4i6urpyFyzma1ifxHn9+nXuB8HU1BTjx49HdnY2xGIxhg4disDAQN5PpPj4AgICeD+RaozyNaYPqOxF0rNnT26yJqByvedZs2bhr7/+woQJE3gtx8d88vE1b96c6/rs5eWFsWPHIjs7G0BlV9yOHTti/PjxcvVJllWtry8jI6NByteYvp49e3KVUQcHhxpLUi1ZsgTdunXj3eX9Q/M1hFNDQ4PrbZaRkYExY8agXbt2WLZsGe8yMZ/8fIaGhrh8+TL32szMDEuXLuW64G/ZsgXOzs6Ij49vFN/7UEZ5+7744gvMmjVLahhE+/btsXnzZgCVk486OjrybjyWt48vrMLPUDiePXsGZ2dn7klJUlIS1q5di65du0JNTQ2qqqpwdHSEQCDgNa6T+ernk0zoUr2LVkhICNq2bYuMjAyIxWJoaWnx6qbHfPXzAZXn1NTUlOuSXFRUBIFAIDWuU/IEg88wFearnw/4v5b40tJS+Pv7c0tqqqmpITo6mpeD+eTnu3v3LgQCAS5evIjHjx/DxMQEDx8+lHpicujQIXTu3JnX5JvMVz8fUNmVtXv37sjLy0OvXr1w/PhxqfdzcnJgaGjIe/WhD80nb2d8fDxatGgh9dQ9JycHwcHB0NDQgKurq9Ta8szXsL7s7Gx4eXlh5cqVKCsrQ1paGgQCgdR9QXZ2NgwNDXl1w5e3730oY0Mc8/Hjx2FqaorffvsNd+7cwbJly6ChocE1fr5+/Rp6enpITk5uFB9fWIWfoXA8efIE33777Rtnj79+/Tr69esHfX195nsHvujoaHTp0oV7sla1e19AQACCgoIQFxcHDQ0N5nsHPqByFYeJEydy49BSU1OxcuVKKXdERASMjIyY7x34JEi6bsbExOCzzz7D7Nmzoauri5ycHIhEIpnHmzJf3X0XL16Eg4MDOnXqBH19/VrXGA8PD4eZmRnzvQMfULmkoaurK4YNG4YRI0agX79+Ug1qISEh0NXVZb535Fy8eDG8vb0B1Ow6fO/ePfTp0wcdOnTgvYIH89XPBwA//vgj2rVrB39/f3Tr1g3Ozs5SQ98uXrwIPT29RvO9D2WUty83Nxfjx4+HQCCAuro6XFxcuEkkKyoq8Pvvv0NHR6fRfHxhFX6GQlJ1wruKiooaE6RNmjQJXl5ezPcOfCKRCImJiVxraNU1dSMjI9G5c2e4u7ujb9++zPcOfBJndnb2WyddnDNnDgICApjvHfhqIyQkBM2bN8dXX30ll3F4zCcbIpEImZmZuHLlClasWIHly5fXmG19zpw5MuUx89XdJ+Hhw4cICAiAkpISBAIBnJycEBQUBDc3Nzg5OWH16tXM946cZWVlNVb4EIvFXMPas2fPMHDgQIwePZr53oEPqFw27tChQ5g4cSLCwsKwdetWdOnSBadPn8aqVavg6emJqVOnNprvfShjQxwzUDnPzI0bN6SevKempmLBggXcXCeN6fsvWIWf8d4hFAoxbtw4btZw5mscn6TSOnPmTAgEApkmnGM++fuA/3vK8Ntvv8HV1RW//fYb8zWiLyUlRabuuszXcL7qq34cOXIE9vb2dVpbnfnq73v9+jXOnz+P8ePHw9vbG5MmTcLJkyfrvPzYh+ZrKGd1KioquK7GzPfufUKhEAsXLoSGhgbMzMwwf/58mVcXaUjf+1DGhjhmCQUFBbh37x5evnypkL6qCACAGIz3jKKiIlJXVydlZWXma2RfUVER/f777zRgwADS0dFhvkb2ERHl5+fT/fv3ycXFhdTV1ZmvkX0MxeP169d0+/ZtcnNzIw0NDeZrRB8AEggE9S7Th+p7X5zMVz9XQUEBtWzZUiF9DeFUdN/7BqvwMxgMBoPBYDAYDAaD0QRRauwCMBgMBoPBYDAYDAaDwZA/rMLPYDAYDAaDwWAwGAxGE4RV+BkMBoPBYDAYDAaDwWiCsAo/g8FgMBgMBoPBYDAYTRBW4WcwGAwG4wPk6dOnJBAIKCEhobGLwpGUlEQfffQRqaurk4uLS2MXp1Z8fHxo5syZjV0MBoPBYDB4wSr8DAaDwWA0AkFBQSQQCGjNmjVS248fPy73JbbeF5YsWUKampqUnJxMYWFhNd7fsWMHtWjRgioqKrhthYWFpKKiQj4+PlL7RkREkEAgoJSUlIYuNoPBYDAYCgur8DMYDAaD0Uioq6vT2rVrKTc3t7GLIjfKysrq/LcpKSnUrVs3Mjc3J11d3Rrv+/r6UmFhIcXGxnLbLl++TAYGBhQVFUWlpaXc9vDwcDIzMyMrKyuZywFAqlGBwWAwGIz3FVbhZzAYDAajkejVqxcZGBjQ6tWr37hPcHBwje7tP/74I1lYWHCvg4KCaNCgQbRq1SrS19enVq1a0bJly6iiooLmzp1LrVu3JhMTE9q/f38Nf1JSEnXt2pXU1dXJ0dGRIiMjpd6/e/cu9e3bl7S0tEhfX59GjhxJr1694t738fGhqVOn0syZM0lPT4/8/PxqPQ6xWEzLli0jExMTUlNTIxcXFzp37hz3vkAgoLi4OFq2bBkJBAIKDg6u4bC1tSVDQ0OKiIjgtkVERNDAgQPJ0tKSbty4IbXd19eXiIiEQiFNnz6d2rZtS+rq6tStWzeKiYmR2lcgENDZs2fJ1dWV1NTU6MqVK1RUVESjRo0iLS0tMjQ0pB9++KFGmbZv3042Njakrq5O+vr6NGTIkFqPn8FgMBiMxoBV+BkMBoPBaCSUlZVp1apVtGXLFnr+/Hm9XJcuXaL09HT6559/aMOGDbRkyRIKDAwkHR0dioqKookTJ9KECRNqfM7cuXNpzpw5FB8fT126dKH+/ftTdnY2ERHl5eVRjx49qFOnThQbG0vnzp2jzMxM+vzzz6UcBw8eJFVVVbp69Srt2LGj1vJt2rSJfvjhB1q/fj3dvn2b/Pz8aMCAAfTw4UMiInrx4gU5ODjQnDlz6MWLF/TNN9/U6vH19aXw8HDudXh4OPn4+JC3tze3vaSkhKKiorgK/7x58+jYsWN08OBBunnzJllbW5Ofnx/l5ORIuRcsWEBr1qyhxMREcnZ2prlz51JkZCSdOHGCLly4QBEREXTz5k1u/9jYWJo+fTotW7aMkpOT6dy5c+Tl5fWf54rBYDAYjHcGGAwGg8FgvHNGjx6NgQMHAgA++ugjjBkzBgDw559/ourP85IlS9CxY0epv924cSPMzc2lXObm5hCJRNw2W1tbdO/enXtdUVEBTU1NHD58GADw5MkTEBHWrFnD7VNeXg4TExOsXbsWALB8+XL06dNH6rPT0tJAREhOTgYAeHt7o1OnTv95vEZGRli5cqXUNnd3d0yePJl73bFjRyxZsuStnt27d0NTUxPl5eXIz89Hs2bNkJWVhV9++QVeXl4AgLCwMBARnj17hsLCQqioqCA0NJRzlJWVwcjICOvWrQMAhIeHg4hw/Phxbp+CggKoqqri119/5bZlZ2dDQ0MDM2bMAAAcO3YMLVu2RH5+/n8eP4PBYDAYjQF7ws9gMBgMRiOzdu1aOnjwICUmJtbZ4eDgQEpK//ezrq+vT05OTtxrZWVl0tXVpaysLKm/69KlC/f/Zs2akZubG1eOW7duUXh4OGlpaXH/7OzsiIikJsNzdXV9a9ny8/MpPT2dPv74Y6ntH3/8sczH7OPjQ0VFRRQTE0OXL1+m9u3bU5s2bcjb25sbxx8REUHt2rUjMzMzSklJofLycqnPVlFRIQ8Pjxqf7ebmxv0/JSWFysrKyNPTk9vWunVrsrW15V737t2bzM3NqV27djRy5EgKDQ2l4uJimY6HwWAwGIyGhFX4GQwGg8FoZLy8vMjPz48WLlxY4z0lJSUCILWtvLy8xn4qKipSrwUCQa3bxGIx73IVFhZS//79KSEhQerfw4cPpbqua2pq8nbWF2trazIxMaHw8HAKDw8nb29vIiIyMjIiU1NTunbtGoWHh1OPHj1kdst6HC1atKCbN2/S4cOHydDQkBYvXkwdO3akvLw8mT+bwWAwGIyGgFX4GQwGg8FQANasWUOnTp2i69evS21v06YNZWRkSFX6ExIS5Pa5VSe6q6iooLi4OOrQoQMREXXu3Jnu3btHFhYWZG1tLfVPlspxy5YtycjIiK5evSq1/erVq2Rvby9zmX19fSkiIoIiIiKkluPz8vKis2fPUnR0NDd+38rKiptfQEJ5eTnFxMS89bOtrKxIRUWFoqKiuG25ubn04MEDqf2aNWtGvXr1onXr1tHt27fp6dOndOnSJZmPicFgMBiMhqBZYxeAwWAwGAwGkZOTE40YMYI2b94std3Hx4devnxJ69atoyFDhtC5c+fo7Nmz1LJlS7l87rZt28jGxoY6dOhAGzdupNzcXBozZgwREU2ZMoV2795Nw4YNo3nz5lHr1q3p0aNHdOTIEdqzZw8pKyvz/py5c+fSkiVLyMrKilxcXGj//v2UkJBAoaGhMpfZ19eXpkyZQuXl5dwTfiIib29vmjp1KpWVlXEVfk1NTZo0aRK3WoGZmRmtW7eOiouLaezYsW/8DC0tLRo7dizNnTuXdHV1qW3btvTtt99KDZs4ffo0PX78mLy8vEhHR4fOnDlDYrFYqts/g8FgMBiNCavwMxgMBoOhICxbtoyOHj0qta1Dhw60fft2WrVqFS1fvpwGDx5M33zzDe3atUsun7lmzRpas2YNJSQkkLW1NZ08eZL09PSIiLin8vPnz6c+ffqQUCgkc3Nz8vf3l6r48mH69On0+vVrmjNnDmVlZZG9vT2dPHmSbGxsZC6zr68vlZSUkJ2dHenr63Pbvb29qaCggFu+r+oxisViGjlyJBUUFJCbmxudP3+edHR03vo533//PTesoUWLFjRnzhx6/fo1936rVq3ojz/+oODgYCotLSUbGxs6fPgwOTg4yHxMDAaDwWA0BAJUHxjIYDAYDAaDwWAwGAwG472HjeFnMBgMBoPBYDAYDAajCcIq/AwGg8FgMBgMBoPBYDRBWIWfwWAwGAwGg8FgMBiMJgir8DMYDAaDwWAwGAwGg9EEYRV+BoPBYDAYDAaDwWAwmiCsws9gMBgMBoPBYDAYDEYThFX4GQwGg8FgMBgMBoPBaIKwCj+DwWAwGAwGg8FgMBhNEFbhZzAYDAaDwWAwGAwGownCKvwMBoPBYDAYDAaDwWA0QViFn8FgMBgMBoPBYDAYjCYIq/AzGAwGg8FgMBgMBoPRBPl/mJoeAPiDoqIAAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 1200x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# Calculate word count\n",
    "converted_df['word_count'] = converted_df['content'].apply(lambda x: len(x.split()))\n",
    "\n",
    "# Calculate mean and standard deviation\n",
    "mean_word_count = converted_df['word_count'].mean()\n",
    "std_word_count = converted_df['word_count'].std()\n",
    "\n",
    "# Calculate 1-sigma, 2-sigma, and 3-sigma intervals\n",
    "one_sigma_interval = (mean_word_count - std_word_count, mean_word_count + std_word_count)\n",
    "two_sigma_interval = (mean_word_count - 2 * std_word_count, mean_word_count + 2 * std_word_count)\n",
    "three_sigma_interval = (mean_word_count - 3 * std_word_count, mean_word_count + 3 * std_word_count)\n",
    "\n",
    "# Display the intervals\n",
    "print(f\"Mean number of words in the content: {mean_word_count:.2f}\")\n",
    "print(f\"1-sigma interval: {one_sigma_interval}\")\n",
    "print(f\"2-sigma interval: {two_sigma_interval}\")\n",
    "print(f\"3-sigma interval: {three_sigma_interval}\")\n",
    "\n",
    "# Plot the distribution of word counts\n",
    "plt.figure(figsize=(12, 8))\n",
    "sns.histplot(converted_df['word_count'], bins=100, kde=True, color='skyblue', edgecolor='black')\n",
    "plt.axvline(mean_word_count, color='red', linestyle='dashed', linewidth=2, label=f'Mean: {mean_word_count:.2f}')\n",
    "\n",
    "# Highlight 1-sigma, 2-sigma, and 3-sigma intervals\n",
    "plt.axvline(one_sigma_interval[0], color='green', linestyle='dotted', linewidth=2, label=f'1-sigma: {one_sigma_interval}')\n",
    "plt.axvline(one_sigma_interval[1], color='green', linestyle='dotted', linewidth=2)\n",
    "\n",
    "plt.axvline(two_sigma_interval[0], color='orange', linestyle='dotted', linewidth=2, label=f'2-sigma: {two_sigma_interval}')\n",
    "plt.axvline(two_sigma_interval[1], color='orange', linestyle='dotted', linewidth=2)\n",
    "\n",
    "plt.axvline(three_sigma_interval[0], color='purple', linestyle='dotted', linewidth=2, label=f'3-sigma: {three_sigma_interval}')\n",
    "plt.axvline(three_sigma_interval[1], color='purple', linestyle='dotted', linewidth=2)\n",
    "\n",
    "# Customize X-axis labels to appear every 20 units\n",
    "start = 0\n",
    "end = converted_df['word_count'].max()\n",
    "increment = 200\n",
    "xticks_locations = range(start, int(end) + 1, increment)\n",
    "plt.xticks(xticks_locations, rotation=75)\n",
    "plt.xlim(start, end)  # Ensure the plot shows the entire data\n",
    "\n",
    "plt.title('Distribution of Word Counts in Content')\n",
    "plt.xlabel('Number of Words')\n",
    "plt.ylabel('Frequency')\n",
    "plt.legend()\n",
    "plt.grid(True)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-01-15T05:10:23.090336Z",
     "iopub.status.busy": "2025-01-15T05:10:23.089851Z",
     "iopub.status.idle": "2025-01-15T05:10:33.494223Z",
     "shell.execute_reply": "2025-01-15T05:10:33.493025Z",
     "shell.execute_reply.started": "2025-01-15T05:10:23.090298Z"
    },
    "id": "S4U0etUEwW_w",
    "trusted": true
   },
   "outputs": [],
   "source": [
    "encode_contents = tokenizer.texts_to_sequences(converted_df['content'])\n",
    "contents_max_length = max([len(el) for el in encode_contents])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-01-15T05:10:33.495852Z",
     "iopub.status.busy": "2025-01-15T05:10:33.495549Z",
     "iopub.status.idle": "2025-01-15T05:10:33.500267Z",
     "shell.execute_reply": "2025-01-15T05:10:33.499202Z",
     "shell.execute_reply.started": "2025-01-15T05:10:33.495826Z"
    },
    "trusted": true
   },
   "outputs": [],
   "source": [
    "VOCAB_SIZE = len(tokenizer.word_index)\n",
    "MAX_LEN = contents_max_length"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-01-15T05:10:33.502486Z",
     "iopub.status.busy": "2025-01-15T05:10:33.502086Z",
     "iopub.status.idle": "2025-01-15T05:10:34.347609Z",
     "shell.execute_reply": "2025-01-15T05:10:34.346476Z",
     "shell.execute_reply.started": "2025-01-15T05:10:33.502442Z"
    },
    "trusted": true
   },
   "outputs": [],
   "source": [
    "encode_contents = pad_sequences(encode_contents, maxlen=MAX_LEN, padding='post')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-01-15T05:10:34.350039Z",
     "iopub.status.busy": "2025-01-15T05:10:34.349622Z",
     "iopub.status.idle": "2025-01-15T05:10:35.781666Z",
     "shell.execute_reply": "2025-01-15T05:10:35.780680Z",
     "shell.execute_reply.started": "2025-01-15T05:10:34.349997Z"
    },
    "trusted": true
   },
   "outputs": [],
   "source": [
    "X_test_contents = tokenizer.texts_to_sequences(converted_test_df['content'])\n",
    "contents_test_max_length = max([len(el) for el in X_test_contents])\n",
    "X_test_contents = pad_sequences(X_test_contents, maxlen=MAX_LEN, padding='post')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-01-15T05:09:57.806880Z",
     "iopub.status.busy": "2025-01-15T05:09:57.806310Z",
     "iopub.status.idle": "2025-01-15T05:09:57.824524Z",
     "shell.execute_reply": "2025-01-15T05:09:57.823277Z",
     "shell.execute_reply.started": "2025-01-15T05:09:57.806842Z"
    },
    "trusted": true
   },
   "outputs": [],
   "source": [
    "# One-hot y\n",
    "y = converted_df['label']\n",
    "y_test = converted_test_df['label']\n",
    "# Create the one-hot encoder\n",
    "enc = OneHotEncoder(handle_unknown='ignore', sparse_output=False)\n",
    "\n",
    "# Fit the encoder on the category names\n",
    "enc.fit(np.array(y).reshape(-1, 1))\n",
    "y = enc.transform(np.array(y).reshape(-1, 1))\n",
    "y_test = enc.transform(np.array(y_test).reshape(-1, 1))\n",
    "\n",
    "# Save to Json and Json file\n",
    "category_to_one_hot_dict = dict(zip(list(range(enc.categories_[0].shape[0])), enc.categories_[0]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-01-15T05:24:15.281520Z",
     "iopub.status.busy": "2025-01-15T05:24:15.281024Z",
     "iopub.status.idle": "2025-01-15T05:24:15.288742Z",
     "shell.execute_reply": "2025-01-15T05:24:15.287351Z",
     "shell.execute_reply.started": "2025-01-15T05:24:15.281481Z"
    },
    "trusted": true
   },
   "outputs": [],
   "source": [
    "with open(\"saved_models/label_dict.json\", \"w\") as file:\n",
    "    json.dump(category_to_one_hot_dict, file)\n",
    "file.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-01-15T05:10:35.783374Z",
     "iopub.status.busy": "2025-01-15T05:10:35.782993Z",
     "iopub.status.idle": "2025-01-15T05:10:36.368443Z",
     "shell.execute_reply": "2025-01-15T05:10:36.367388Z",
     "shell.execute_reply.started": "2025-01-15T05:10:35.783343Z"
    },
    "id": "tGH30L2311km",
    "trusted": true
   },
   "outputs": [],
   "source": [
    "\n",
    "X_train_contents, X_val_contents, y_train, y_val = train_test_split(\n",
    "    encode_contents,  # Contents feature\n",
    "    y, stratify=y,\n",
    "    test_size=0.2,\n",
    "    random_state=12345\n",
    ")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-01-15T05:10:36.369755Z",
     "iopub.status.busy": "2025-01-15T05:10:36.369438Z",
     "iopub.status.idle": "2025-01-15T05:10:36.374548Z",
     "shell.execute_reply": "2025-01-15T05:10:36.373230Z",
     "shell.execute_reply.started": "2025-01-15T05:10:36.369725Z"
    },
    "id": "NFbjvKvnS7oL",
    "trusted": true
   },
   "outputs": [],
   "source": [
    "def convert_to_datast_tensor(X_contents, yy):\n",
    "\n",
    "  return tf.convert_to_tensor(X_contents, dtype=tf.int32),\\\n",
    "          tf.convert_to_tensor(yy, dtype=tf.float64)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-01-15T05:10:36.375814Z",
     "iopub.status.busy": "2025-01-15T05:10:36.375525Z",
     "iopub.status.idle": "2025-01-15T05:10:38.688615Z",
     "shell.execute_reply": "2025-01-15T05:10:38.687289Z",
     "shell.execute_reply.started": "2025-01-15T05:10:36.375789Z"
    },
    "id": "t-6TmnERBM8S",
    "trusted": true
   },
   "outputs": [],
   "source": [
    "# prompt: create tensorfrom slice ((X_train_title, X_train_summary,X_train_contents), y\n",
    "BATCH_SIZE = 92\n",
    "\n",
    "train_dataset = tf.data.Dataset.from_tensor_slices((X_train_contents, y_train)).\\\n",
    "                      map(convert_to_datast_tensor, num_parallel_calls=tf.data.AUTOTUNE).\\\n",
    "                      batch(BATCH_SIZE)\n",
    "\n",
    "val_dataset = tf.data.Dataset.from_tensor_slices((X_val_contents, y_val)).\\\n",
    "                      map(convert_to_datast_tensor, num_parallel_calls=tf.data.AUTOTUNE).\\\n",
    "                      batch(BATCH_SIZE)\n",
    "\n",
    "test_dataset = tf.data.Dataset.from_tensor_slices((X_test_contents, y_test)).\\\n",
    "                      map(convert_to_datast_tensor, num_parallel_calls=tf.data.AUTOTUNE).\\\n",
    "                      batch(BATCH_SIZE)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-01-15T03:28:25.243086Z",
     "iopub.status.busy": "2025-01-15T03:28:25.242757Z",
     "iopub.status.idle": "2025-01-15T03:28:25.249114Z",
     "shell.execute_reply": "2025-01-15T03:28:25.248294Z",
     "shell.execute_reply.started": "2025-01-15T03:28:25.243057Z"
    },
    "trusted": true
   },
   "outputs": [],
   "source": [
    "embedding_dim = 192\n",
    "# VOCAB_SIZE = \n",
    "def create_model():\n",
    "    # Define input layers\n",
    "    content_input = Input(shape=(MAX_LEN,), name='content_input')\n",
    "    \n",
    "    # Define embedding layers (you might need to adjust embedding dimension)    \n",
    "    content_embedding = tf.keras.layers.Embedding(VOCAB_SIZE,                                                  \n",
    "                                                  embedding_dim,\n",
    "                                                  # weights = [embedding_matrix],\n",
    "                                                  trainable=True)(content_input)\n",
    "    \n",
    "    content_lstm = Bidirectional(GRU(64, return_sequences=True))(content_embedding)\n",
    "    # content_lstm = LSTM(64, return_sequences=True)(content_embedding)\n",
    "    content_lstm = Dropout(0.2)(content_lstm)\n",
    "    content_lstm = BatchNormalization()(content_lstm)\n",
    "    \n",
    "    content_lstm = Bidirectional(GRU(64, return_sequences=False))(content_lstm)\n",
    "    # content_lstm = LSTM(220, return_sequences=False)(content_lstm)\n",
    "    content_lstm = Dropout(0.2)(content_lstm)\n",
    "    content_lstm = BatchNormalization()(content_lstm)\n",
    "    \n",
    "    # Define dense layers for classification\n",
    "    dense1 = Dense(192, activation='relu')(content_lstm)\n",
    "    dropout = Dropout(0.2)(dense1)\n",
    "    output = Dense(y_train.shape[1], activation='softmax')(dense1)  # Output layer with softmax\n",
    "    \n",
    "    # Create the model\n",
    "    model = Model(inputs=content_input, outputs=output)\n",
    "\n",
    "    return model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 321
    },
    "execution": {
     "iopub.execute_input": "2025-01-15T03:29:57.016966Z",
     "iopub.status.busy": "2025-01-15T03:29:57.016615Z",
     "iopub.status.idle": "2025-01-15T03:29:57.167010Z",
     "shell.execute_reply": "2025-01-15T03:29:57.166244Z",
     "shell.execute_reply.started": "2025-01-15T03:29:57.016935Z"
    },
    "id": "RJ4wGlpm8Co1",
    "outputId": "20142340-33ab-47c1-a2b0-47a6c90c92b0",
    "trusted": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\">Model: \"functional_1\"</span>\n",
       "</pre>\n"
      ],
      "text/plain": [
       "\u001b[1mModel: \"functional_1\"\u001b[0m\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\">┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━┓\n",
       "┃<span style=\"font-weight: bold\"> Layer (type)                         </span>┃<span style=\"font-weight: bold\"> Output Shape                </span>┃<span style=\"font-weight: bold\">         Param # </span>┃\n",
       "┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━┩\n",
       "│ content_input (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">InputLayer</span>)           │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">8054</span>)                │               <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
       "├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n",
       "│ embedding_1 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Embedding</span>)              │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">8054</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">192</span>)           │      <span style=\"color: #00af00; text-decoration-color: #00af00\">36,264,192</span> │\n",
       "├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n",
       "│ bidirectional_2 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Bidirectional</span>)      │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">8054</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">128</span>)           │          <span style=\"color: #00af00; text-decoration-color: #00af00\">99,072</span> │\n",
       "├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n",
       "│ dropout_3 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dropout</span>)                  │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">8054</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">128</span>)           │               <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
       "├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n",
       "│ batch_normalization_2                │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">8054</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">128</span>)           │             <span style=\"color: #00af00; text-decoration-color: #00af00\">512</span> │\n",
       "│ (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">BatchNormalization</span>)                 │                             │                 │\n",
       "├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n",
       "│ bidirectional_3 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Bidirectional</span>)      │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">128</span>)                 │          <span style=\"color: #00af00; text-decoration-color: #00af00\">74,496</span> │\n",
       "├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n",
       "│ dropout_4 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dropout</span>)                  │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">128</span>)                 │               <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
       "├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n",
       "│ batch_normalization_3                │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">128</span>)                 │             <span style=\"color: #00af00; text-decoration-color: #00af00\">512</span> │\n",
       "│ (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">BatchNormalization</span>)                 │                             │                 │\n",
       "├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n",
       "│ dense_2 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dense</span>)                      │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">192</span>)                 │          <span style=\"color: #00af00; text-decoration-color: #00af00\">24,768</span> │\n",
       "├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n",
       "│ dense_3 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dense</span>)                      │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">8</span>)                   │           <span style=\"color: #00af00; text-decoration-color: #00af00\">1,544</span> │\n",
       "└──────────────────────────────────────┴─────────────────────────────┴─────────────────┘\n",
       "</pre>\n"
      ],
      "text/plain": [
       "┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━┓\n",
       "┃\u001b[1m \u001b[0m\u001b[1mLayer (type)                        \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1mOutput Shape               \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1m        Param #\u001b[0m\u001b[1m \u001b[0m┃\n",
       "┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━┩\n",
       "│ content_input (\u001b[38;5;33mInputLayer\u001b[0m)           │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m8054\u001b[0m)                │               \u001b[38;5;34m0\u001b[0m │\n",
       "├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n",
       "│ embedding_1 (\u001b[38;5;33mEmbedding\u001b[0m)              │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m8054\u001b[0m, \u001b[38;5;34m192\u001b[0m)           │      \u001b[38;5;34m36,264,192\u001b[0m │\n",
       "├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n",
       "│ bidirectional_2 (\u001b[38;5;33mBidirectional\u001b[0m)      │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m8054\u001b[0m, \u001b[38;5;34m128\u001b[0m)           │          \u001b[38;5;34m99,072\u001b[0m │\n",
       "├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n",
       "│ dropout_3 (\u001b[38;5;33mDropout\u001b[0m)                  │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m8054\u001b[0m, \u001b[38;5;34m128\u001b[0m)           │               \u001b[38;5;34m0\u001b[0m │\n",
       "├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n",
       "│ batch_normalization_2                │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m8054\u001b[0m, \u001b[38;5;34m128\u001b[0m)           │             \u001b[38;5;34m512\u001b[0m │\n",
       "│ (\u001b[38;5;33mBatchNormalization\u001b[0m)                 │                             │                 │\n",
       "├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n",
       "│ bidirectional_3 (\u001b[38;5;33mBidirectional\u001b[0m)      │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m128\u001b[0m)                 │          \u001b[38;5;34m74,496\u001b[0m │\n",
       "├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n",
       "│ dropout_4 (\u001b[38;5;33mDropout\u001b[0m)                  │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m128\u001b[0m)                 │               \u001b[38;5;34m0\u001b[0m │\n",
       "├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n",
       "│ batch_normalization_3                │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m128\u001b[0m)                 │             \u001b[38;5;34m512\u001b[0m │\n",
       "│ (\u001b[38;5;33mBatchNormalization\u001b[0m)                 │                             │                 │\n",
       "├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n",
       "│ dense_2 (\u001b[38;5;33mDense\u001b[0m)                      │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m192\u001b[0m)                 │          \u001b[38;5;34m24,768\u001b[0m │\n",
       "├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n",
       "│ dense_3 (\u001b[38;5;33mDense\u001b[0m)                      │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m8\u001b[0m)                   │           \u001b[38;5;34m1,544\u001b[0m │\n",
       "└──────────────────────────────────────┴─────────────────────────────┴─────────────────┘\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Total params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">36,465,096</span> (139.10 MB)\n",
       "</pre>\n"
      ],
      "text/plain": [
       "\u001b[1m Total params: \u001b[0m\u001b[38;5;34m36,465,096\u001b[0m (139.10 MB)\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Trainable params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">36,464,584</span> (139.10 MB)\n",
       "</pre>\n"
      ],
      "text/plain": [
       "\u001b[1m Trainable params: \u001b[0m\u001b[38;5;34m36,464,584\u001b[0m (139.10 MB)\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Non-trainable params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">512</span> (2.00 KB)\n",
       "</pre>\n"
      ],
      "text/plain": [
       "\u001b[1m Non-trainable params: \u001b[0m\u001b[38;5;34m512\u001b[0m (2.00 KB)\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "model = create_model()\n",
    "# model = create_model_rnn()\n",
    "# Compile the model\n",
    "model.compile(optimizer=tf.keras.optimizers.Adam(learning_rate=5e-4),\n",
    "              loss='categorical_crossentropy', metrics=['accuracy'])\n",
    "\n",
    "# Print model summary\n",
    "model.summary()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-01-15T03:29:45.125025Z",
     "iopub.status.busy": "2025-01-15T03:29:45.124693Z",
     "iopub.status.idle": "2025-01-15T03:29:45.401603Z",
     "shell.execute_reply": "2025-01-15T03:29:45.400446Z",
     "shell.execute_reply.started": "2025-01-15T03:29:45.125002Z"
    },
    "trusted": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/lib/python3.10/pty.py:89: RuntimeWarning: os.fork() was called. os.fork() is incompatible with multithreaded code, and JAX is multithreaded, so this will likely lead to a deadlock.\n",
      "  pid, fd = os.forkpty()\n"
     ]
    }
   ],
   "source": [
    "# !rm -rf saved_models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-01-15T03:30:00.286145Z",
     "iopub.status.busy": "2025-01-15T03:30:00.285853Z",
     "iopub.status.idle": "2025-01-15T03:30:00.291025Z",
     "shell.execute_reply": "2025-01-15T03:30:00.290190Z",
     "shell.execute_reply.started": "2025-01-15T03:30:00.286123Z"
    },
    "trusted": true
   },
   "outputs": [],
   "source": [
    "# Callback\n",
    "os.makedirs(\"saved_models/weights\", exist_ok=True)\n",
    "my_callbacks = [\n",
    "    tf.keras.callbacks.EarlyStopping(monitor=\"val_loss\", patience=5,\n",
    "    mode=\"min\", restore_best_weights=True),\n",
    "    tf.keras.callbacks.BackupAndRestore(\n",
    "    backup_dir='saved_models/weights', delete_checkpoint=False\n",
    "    )\n",
    "]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "trusted": true
   },
   "outputs": [],
   "source": [
    "# del model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "execution": {
     "iopub.execute_input": "2025-01-15T03:30:02.706290Z",
     "iopub.status.busy": "2025-01-15T03:30:02.705997Z",
     "iopub.status.idle": "2025-01-15T04:28:46.949423Z",
     "shell.execute_reply": "2025-01-15T04:28:46.948650Z",
     "shell.execute_reply.started": "2025-01-15T03:30:02.706267Z"
    },
    "id": "9g-5dTT5WkEl",
    "outputId": "394a5cde-9d57-4cde-d511-e52df863b17d",
    "trusted": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/15\n",
      "\u001b[1m227/227\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m399s\u001b[0m 2s/step - accuracy: 0.4486 - loss: 1.5151 - val_accuracy: 0.6278 - val_loss: 1.5007\n",
      "Epoch 2/15\n",
      "\u001b[1m227/227\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m391s\u001b[0m 2s/step - accuracy: 0.9490 - loss: 0.1674 - val_accuracy: 0.8421 - val_loss: 0.4998\n",
      "Epoch 3/15\n",
      "\u001b[1m227/227\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m390s\u001b[0m 2s/step - accuracy: 0.9944 - loss: 0.0269 - val_accuracy: 0.9290 - val_loss: 0.2553\n",
      "Epoch 4/15\n",
      "\u001b[1m227/227\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m391s\u001b[0m 2s/step - accuracy: 0.9983 - loss: 0.0104 - val_accuracy: 0.9382 - val_loss: 0.2394\n",
      "Epoch 5/15\n",
      "\u001b[1m227/227\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m391s\u001b[0m 2s/step - accuracy: 0.9989 - loss: 0.0067 - val_accuracy: 0.9390 - val_loss: 0.2410\n",
      "Epoch 6/15\n",
      "\u001b[1m227/227\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m391s\u001b[0m 2s/step - accuracy: 0.9990 - loss: 0.0049 - val_accuracy: 0.9392 - val_loss: 0.2485\n",
      "Epoch 7/15\n",
      "\u001b[1m227/227\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m391s\u001b[0m 2s/step - accuracy: 0.9992 - loss: 0.0031 - val_accuracy: 0.9367 - val_loss: 0.2728\n",
      "Epoch 8/15\n",
      "\u001b[1m227/227\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m390s\u001b[0m 2s/step - accuracy: 0.9997 - loss: 0.0021 - val_accuracy: 0.9405 - val_loss: 0.2701\n",
      "Epoch 9/15\n",
      "\u001b[1m227/227\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m390s\u001b[0m 2s/step - accuracy: 0.9997 - loss: 0.0020 - val_accuracy: 0.9355 - val_loss: 0.2952\n"
     ]
    }
   ],
   "source": [
    "history = model.fit(train_dataset,\n",
    "          epochs=15, callbacks=my_callbacks,\n",
    "          validation_data=val_dataset)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Save Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-01-15T04:32:49.596104Z",
     "iopub.status.busy": "2025-01-15T04:32:49.595820Z",
     "iopub.status.idle": "2025-01-15T04:32:50.682697Z",
     "shell.execute_reply": "2025-01-15T04:32:50.681998Z",
     "shell.execute_reply.started": "2025-01-15T04:32:49.596083Z"
    },
    "trusted": true
   },
   "outputs": [],
   "source": [
    "model.save('saved_models/bidirectional-GRU.h5')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Evaluation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-01-15T04:29:32.054635Z",
     "iopub.status.busy": "2025-01-15T04:29:32.054306Z",
     "iopub.status.idle": "2025-01-15T04:29:32.499802Z",
     "shell.execute_reply": "2025-01-15T04:29:32.498830Z",
     "shell.execute_reply.started": "2025-01-15T04:29:32.054612Z"
    },
    "trusted": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x400 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# Plot training & validation loss values\n",
    "plt.figure(figsize=(12, 4))\n",
    "\n",
    "plt.subplot(1, 2, 1)\n",
    "plt.plot(history.history['loss'], label='Train Loss')\n",
    "plt.plot(history.history['val_loss'], label='Validation Loss')\n",
    "plt.title('Model Loss')\n",
    "plt.ylabel('Loss')\n",
    "plt.xlabel('Epoch')\n",
    "plt.legend()\n",
    "\n",
    "# Plot training & validation accuracy values\n",
    "plt.subplot(1, 2, 2)\n",
    "plt.plot(history.history['accuracy'], label='Train Accuracy')\n",
    "plt.plot(history.history['val_accuracy'], label='Validation Accuracy')\n",
    "plt.title('Model Accuracy')\n",
    "plt.ylabel('Accuracy')\n",
    "plt.xlabel('Epoch')\n",
    "plt.legend()\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-01-15T04:29:36.927662Z",
     "iopub.status.busy": "2025-01-15T04:29:36.927318Z",
     "iopub.status.idle": "2025-01-15T04:30:12.850916Z",
     "shell.execute_reply": "2025-01-15T04:30:12.850130Z",
     "shell.execute_reply.started": "2025-01-15T04:29:36.927636Z"
    },
    "trusted": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m57/57\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m36s\u001b[0m 629ms/step - accuracy: 0.9372 - loss: 0.2545\n",
      "Loss of Validation set: 0.23940232396125793, Accuracy of Validation set: 0.9382194876670837\n"
     ]
    }
   ],
   "source": [
    "val_loss, val_acc = model.evaluate(val_dataset)\n",
    "print(f\"Loss of Validation set: {val_loss}, Accuracy of Validation set: {val_acc}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-01-15T04:30:12.852077Z",
     "iopub.status.busy": "2025-01-15T04:30:12.851841Z",
     "iopub.status.idle": "2025-01-15T04:30:29.636278Z",
     "shell.execute_reply": "2025-01-15T04:30:29.635332Z",
     "shell.execute_reply.started": "2025-01-15T04:30:12.852056Z"
    },
    "trusted": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m27/27\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m17s\u001b[0m 618ms/step - accuracy: 0.6614 - loss: 1.7429\n",
      "Loss of Test set: 1.171366810798645, Accuracy of Test set: 0.7616666555404663\n"
     ]
    }
   ],
   "source": [
    "test_loss, test_acc = model.evaluate(test_dataset)\n",
    "print(f\"Loss of Test set: {test_loss}, Accuracy of Test set: {test_acc}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Inference"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-01-15T05:01:39.975526Z",
     "iopub.status.busy": "2025-01-15T05:01:39.974769Z",
     "iopub.status.idle": "2025-01-15T05:01:39.981019Z",
     "shell.execute_reply": "2025-01-15T05:01:39.979260Z",
     "shell.execute_reply.started": "2025-01-15T05:01:39.975488Z"
    },
    "trusted": true
   },
   "outputs": [],
   "source": [
    "import re"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-01-15T05:01:53.961316Z",
     "iopub.status.busy": "2025-01-15T05:01:53.960798Z",
     "iopub.status.idle": "2025-01-15T05:01:55.186100Z",
     "shell.execute_reply": "2025-01-15T05:01:55.185064Z",
     "shell.execute_reply.started": "2025-01-15T05:01:53.961276Z"
    },
    "trusted": true
   },
   "outputs": [],
   "source": [
    "# Load Model\n",
    "model = tf.keras.models.load_model('saved_models/bidirectional-GRU.h5')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load Tokenizer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-01-15T05:02:39.481891Z",
     "iopub.status.busy": "2025-01-15T05:02:39.481277Z",
     "iopub.status.idle": "2025-01-15T05:02:40.625049Z",
     "shell.execute_reply": "2025-01-15T05:02:40.624062Z",
     "shell.execute_reply.started": "2025-01-15T05:02:39.481849Z"
    },
    "trusted": true
   },
   "outputs": [],
   "source": [
    "# Load Exists Tokenizers\n",
    "\n",
    "tokenizer_file = open('tokenizers/tokenizer.pkl', 'rb')\n",
    "tokenizer = pickle.load(tokenizer_file)\n",
    "\n",
    "tokenizer = tf.keras.preprocessing.text.tokenizer_from_json(tokenizer)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load Label Dict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "trusted": true
   },
   "outputs": [],
   "source": [
    "# Load the JSON file back into a dictionary\n",
    "with open(\"saved_models/label_dict.json\", 'r', encoding='utf-8') as file:\n",
    "    label_dict = json.load(file)\n",
    "file.close()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Real News Inference"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "trusted": true
   },
   "outputs": [],
   "source": [
    "MAX_LEN = 8054\n",
    "characters_to_replace = r'!\"#$%&()*+,-./:;=?@\\[\\\\\\]^`{|}~\\t\\n'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2025-01-15T05:32:56.644824Z",
     "iopub.status.busy": "2025-01-15T05:32:56.644404Z",
     "iopub.status.idle": "2025-01-15T05:32:58.525188Z",
     "shell.execute_reply": "2025-01-15T05:32:58.524033Z",
     "shell.execute_reply.started": "2025-01-15T05:32:56.644792Z"
    },
    "trusted": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Type of this news is: The thao\n"
     ]
    }
   ],
   "source": [
    "# Ground Truth: The thao\n",
    "content = \"\"\"\n",
    "Liverpool và Chelsea cùng mất điểm.\n",
    "Toàn bộ các đại gia ra sân ở vòng 21 Premier League sáng 15-1 đều bị cầm hòa. Trong đó, Liverpool và Chelsea phải rất vất vả mới lấy lại được 1 điểm.\n",
    "Liverpool và Chelsea cùng mất điểm.\n",
    "Liverpool (phải) vất vả giành lại 1 điểm từ tay Nottingham Forest\n",
    "\n",
    "Liverpool phải làm khách trước đội bóng duy nhất đánh bại họ ở Premier League mùa này là Nottingham Forest. Được xem là \"ngựa ô\" của mùa giải năm nay, Nottingham Forest chỉ kém ngôi đầu bảng của Liverpool có 6 điểm. Và họ hoàn toàn có thể tham gia cuộc đua vô địch.\n",
    "\n",
    "Nỗi sợ với CĐV \"lữ đoàn đỏ\" tái hiện rất sớm. Ngay phút thứ 8, Nottingham Forest ghi bàn mở tỉ số với pha lên bóng đầu tiên của họ.\n",
    "\n",
    "Liverpool và Chelsea cùng mất điểm.\n",
    "Wood ghi bàn mở tỉ số từ sớm.\n",
    "\n",
    "Từ một pha tranh chấp giữa sân, Elanga có bóng rồi chọc khe cho Chris Wood thoát xuống dứt điểm ở góc khá hẹp. Tiền đạo người New Zealand vẫn tung được cú dứt điểm hiểm hóc, hạ gục thủ thành Alisson.\n",
    "\n",
    "Trước và sau bàn thua này, Liverpool hoàn toàn làm chủ thế trận. Dù vậy, các chân sút của họ bất lực trong việc tạo ra cơ hội trước hàng thủ chơi quá chắc chắn của Nottingham Forest.\n",
    "\n",
    "Liverpool và Chelsea cùng mất điểm.\n",
    "Liverpool đỉnh bảng, Man United vẫn hạng 14 sau nửa mùa Giải Ngoại hạng Anh.\n",
    "ĐỌC NGAY\n",
    "Độ nguy hiểm của Nottingham Forest cũng khiến các cầu thủ phòng ngự Liverpool không dám dâng cao. Vì vậy, Liverpool đành chấp nhận tỉ số 0-1 sau 45 phút đầu tiên.\n",
    "\n",
    "Liverpool tăng tốc ở hiệp 2 như dự đoán, và bước ngoặt đến từ pha thay người ở phút 66 của HLV Arne Slot. Hai cầu thủ được tung vào sân là Tsimikas và Jota lập công ngay trong cú chạm bóng đầu tiên của họ.\n",
    "\n",
    "Liverpool và Chelsea cùng mất điểm\n",
    "Jota ghi bàn gỡ hòa\n",
    "\n",
    "Trên chấm phạt góc, Tsimikas tạt bóng chuẩn xác để Jota ập vào đánh đầu cận thành tung lưới đối thủ, ghi bàn gỡ hòa.\n",
    "\n",
    "Sau bàn thắng này, trận đấu cũng cởi mở hơn khi Nottingham Forest bắt đầu dâng cao đội hình. Nhờ vậy, Liverpool tạo ra hàng loạt cơ hội ăn bàn.\n",
    "\n",
    "Khá nhàn hạ trong hiệp 1, thủ thành Matz Sels trở thành tâm điểm của hiệp 2 với một chuỗi những pha cứu thua khó tin. Chỉ trong khoảng 20 phút cuối trận, thủ thành người Bỉ liên tiếp từ chối 4 cơ hội ghi bàn mười mươi của Jota, Salah và Gakpo.\n",
    "\n",
    "Nottingham Forest còn gặp may khi cú sút cực căng của Salah sau tình huống phạt góc cuối trận bị hậu vệ Ola Aina phá ngay trước vạch vôi.\n",
    "\n",
    "Không thể ghi thêm bàn thắng, Liverpool chấp nhận chia điểm với Nottingham Forest. Liverpool hiện có 47 điểm nhưng mới đá 20 trận, nhiều hơn 6 điểm so với vị trí nhì bảng của chính Nottingham Forest (đá 21 trận).\n",
    "\n",
    "Liverpool không phải quá lo lắng bởi cả hai đại gia cùng ra sân sáng 15-1 là Man City và Chelsea đều bị cầm hòa. Riêng Chelsea phải rất vất vả mới có thể giành lại 1 điểm dù được đá trên sân nhà trước Bournemouth.\n",
    "\n",
    "Chelsea mở đầu trận đấu với kịch bản vô cùng quen thuộc. Cole Palmer tỏa sáng ghi bàn mở tỉ số, để rồi hàng thủ kém cỏi khiến họ phải trả giá.\n",
    "\n",
    "Liverpool và Chelsea cùng mất điểm\n",
    "Tình huống ghi bàn gỡ hòa của Chelsea\n",
    "Từ phút 50 đến 68, Bournemouth ghi liền 2 bàn để dẫn lại 2-1. Chelsea chơi tấn công dồn dập trong những phút cuối trận và cũng có được bàn gỡ hòa nhờ pha sút phạt đẳng cấp của Reece James.\n",
    "\n",
    "Với trận hòa này, Chelsea mới có 37 điểm sau 21 trận, kém Liverpool 10 điểm và chỉ hơn đội hạng 5 là Newcastle vỏn vẹn 2 điểm. Chelsea có thể văng khỏi top 4 nếu Newcastle thắng vòng đấu này.\n",
    "\n",
    "\"\"\"\n",
    "\n",
    "content = re.sub(f\"[{re.escape(characters_to_replace)}]\", \"\", content.replace(\"\\n\", \" \"))\n",
    "content = \"<sos> \"+ word_tokenize(content.replace(\"\\n\", \"\"), format=\"text\") + \" <eos>\"\n",
    "content = tokenizer.texts_to_sequences([content])\n",
    "content = pad_sequences(content, maxlen=MAX_LEN, padding='post')\n",
    "content_predict = model.predict(content, verbose=0)\n",
    "result = np.argmax(content_predict, axis=1)\n",
    "print(f\"Type of this news is: {label_dict[str(result[0])]}\")"
   ]
  }
 ],
 "metadata": {
  "colab": {
   "provenance": []
  },
  "kaggle": {
   "accelerator": "none",
   "dataSources": [
    {
     "datasetId": 6476682,
     "sourceId": 10461457,
     "sourceType": "datasetVersion"
    },
    {
     "sourceId": 217722818,
     "sourceType": "kernelVersion"
    }
   ],
   "dockerImageVersionId": 30823,
   "isGpuEnabled": false,
   "isInternetEnabled": true,
   "language": "python",
   "sourceType": "notebook"
  },
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}