examples/smile_upscaling/clip-vit-base-patch32-ablations-topk.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "import json\n",
    "import os\n",
    "from collections import defaultdict\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "from IPython.display import display\n",
    "import matplotlib\n",
    "\n",
    "sns.set_style(\"ticks\")\n",
    "\n",
    "plt.rcParams[\"font.family\"] = \"Times New Roman\"\n",
    "plt.rcParams[\"mathtext.fontset\"] = \"cm\"\n",
    "\n",
    "matplotlib.rcParams[\"pdf.fonttype\"] = 42\n",
    "matplotlib.rcParams[\"ps.fonttype\"] = 42"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_e277c_row0_col0, #T_e277c_row0_col1, #T_e277c_row0_col2, #T_e277c_row0_col3, #T_e277c_row0_col5, #T_e277c_row1_col0, #T_e277c_row1_col1, #T_e277c_row1_col2, #T_e277c_row2_col0, #T_e277c_row2_col1, #T_e277c_row2_col2, #T_e277c_row2_col9, #T_e277c_row3_col0, #T_e277c_row3_col1, #T_e277c_row3_col2, #T_e277c_row3_col4, #T_e277c_row3_col6, #T_e277c_row3_col7, #T_e277c_row3_col8, #T_e277c_row3_col10, #T_e277c_row3_col11, #T_e277c_row3_col12 {\n",
       "  background-color: #a50026;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_e277c_row0_col4, #T_e277c_row0_col7, #T_e277c_row0_col8, #T_e277c_row0_col9, #T_e277c_row0_col10, #T_e277c_row0_col11, #T_e277c_row0_col12, #T_e277c_row1_col5, #T_e277c_row1_col6, #T_e277c_row1_col7, #T_e277c_row3_col3 {\n",
       "  background-color: #006837;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_e277c_row0_col6 {\n",
       "  background-color: #108647;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_e277c_row1_col3 {\n",
       "  background-color: #e34933;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_e277c_row1_col4 {\n",
       "  background-color: #84ca66;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_e277c_row1_col8 {\n",
       "  background-color: #66bd63;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_e277c_row1_col9 {\n",
       "  background-color: #b7e075;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_e277c_row1_col10 {\n",
       "  background-color: #f1f9ac;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_e277c_row1_col11 {\n",
       "  background-color: #06733d;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_e277c_row1_col12 {\n",
       "  background-color: #87cb67;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_e277c_row2_col3 {\n",
       "  background-color: #fee999;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_e277c_row2_col4 {\n",
       "  background-color: #fdb96a;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_e277c_row2_col5 {\n",
       "  background-color: #57b65f;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_e277c_row2_col6 {\n",
       "  background-color: #cdea83;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_e277c_row2_col7 {\n",
       "  background-color: #abdb6d;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_e277c_row2_col8 {\n",
       "  background-color: #fdbb6c;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_e277c_row2_col10 {\n",
       "  background-color: #e24731;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_e277c_row2_col11 {\n",
       "  background-color: #fee695;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_e277c_row2_col12 {\n",
       "  background-color: #fdb567;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_e277c_row3_col5 {\n",
       "  background-color: #b71126;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_e277c_row3_col9 {\n",
       "  background-color: #ea5739;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_e277c\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_e277c_level0_col0\" class=\"col_heading level0 col0\" >num_params</th>\n",
       "      <th id=\"T_e277c_level0_col1\" class=\"col_heading level0 col1\" >gate_k</th>\n",
       "      <th id=\"T_e277c_level0_col2\" class=\"col_heading level0 col2\" >k</th>\n",
       "      <th id=\"T_e277c_level0_col3\" class=\"col_heading level0 col3\" >top_k</th>\n",
       "      <th id=\"T_e277c_level0_col4\" class=\"col_heading level0 col4\" >svhn</th>\n",
       "      <th id=\"T_e277c_level0_col5\" class=\"col_heading level0 col5\" >stanford_cars</th>\n",
       "      <th id=\"T_e277c_level0_col6\" class=\"col_heading level0 col6\" >resisc45</th>\n",
       "      <th id=\"T_e277c_level0_col7\" class=\"col_heading level0 col7\" >eurosat</th>\n",
       "      <th id=\"T_e277c_level0_col8\" class=\"col_heading level0 col8\" >gtsrb</th>\n",
       "      <th id=\"T_e277c_level0_col9\" class=\"col_heading level0 col9\" >mnist</th>\n",
       "      <th id=\"T_e277c_level0_col10\" class=\"col_heading level0 col10\" >dtd</th>\n",
       "      <th id=\"T_e277c_level0_col11\" class=\"col_heading level0 col11\" >sun397</th>\n",
       "      <th id=\"T_e277c_level0_col12\" class=\"col_heading level0 col12\" >average</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_e277c_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_e277c_row0_col0\" class=\"data row0 col0\" >141203712</td>\n",
       "      <td id=\"T_e277c_row0_col1\" class=\"data row0 col1\" >16</td>\n",
       "      <td id=\"T_e277c_row0_col2\" class=\"data row0 col2\" >32</td>\n",
       "      <td id=\"T_e277c_row0_col3\" class=\"data row0 col3\" >1</td>\n",
       "      <td id=\"T_e277c_row0_col4\" class=\"data row0 col4\" >0.966157</td>\n",
       "      <td id=\"T_e277c_row0_col5\" class=\"data row0 col5\" >0.748290</td>\n",
       "      <td id=\"T_e277c_row0_col6\" class=\"data row0 col6\" >0.906508</td>\n",
       "      <td id=\"T_e277c_row0_col7\" class=\"data row0 col7\" >0.983333</td>\n",
       "      <td id=\"T_e277c_row0_col8\" class=\"data row0 col8\" >0.971180</td>\n",
       "      <td id=\"T_e277c_row0_col9\" class=\"data row0 col9\" >0.995400</td>\n",
       "      <td id=\"T_e277c_row0_col10\" class=\"data row0 col10\" >0.768085</td>\n",
       "      <td id=\"T_e277c_row0_col11\" class=\"data row0 col11\" >0.715667</td>\n",
       "      <td id=\"T_e277c_row0_col12\" class=\"data row0 col12\" >88.182758</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e277c_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "      <td id=\"T_e277c_row1_col0\" class=\"data row1 col0\" >141203712</td>\n",
       "      <td id=\"T_e277c_row1_col1\" class=\"data row1 col1\" >16</td>\n",
       "      <td id=\"T_e277c_row1_col2\" class=\"data row1 col2\" >32</td>\n",
       "      <td id=\"T_e277c_row1_col3\" class=\"data row1 col3\" >2</td>\n",
       "      <td id=\"T_e277c_row1_col4\" class=\"data row1 col4\" >0.962930</td>\n",
       "      <td id=\"T_e277c_row1_col5\" class=\"data row1 col5\" >0.755130</td>\n",
       "      <td id=\"T_e277c_row1_col6\" class=\"data row1 col6\" >0.907143</td>\n",
       "      <td id=\"T_e277c_row1_col7\" class=\"data row1 col7\" >0.983333</td>\n",
       "      <td id=\"T_e277c_row1_col8\" class=\"data row1 col8\" >0.967854</td>\n",
       "      <td id=\"T_e277c_row1_col9\" class=\"data row1 col9\" >0.995200</td>\n",
       "      <td id=\"T_e277c_row1_col10\" class=\"data row1 col10\" >0.734574</td>\n",
       "      <td id=\"T_e277c_row1_col11\" class=\"data row1 col11\" >0.715315</td>\n",
       "      <td id=\"T_e277c_row1_col12\" class=\"data row1 col12\" >87.768501</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e277c_level0_row2\" class=\"row_heading level0 row2\" >2</th>\n",
       "      <td id=\"T_e277c_row2_col0\" class=\"data row2 col0\" >141203712</td>\n",
       "      <td id=\"T_e277c_row2_col1\" class=\"data row2 col1\" >16</td>\n",
       "      <td id=\"T_e277c_row2_col2\" class=\"data row2 col2\" >32</td>\n",
       "      <td id=\"T_e277c_row2_col3\" class=\"data row2 col3\" >4</td>\n",
       "      <td id=\"T_e277c_row2_col4\" class=\"data row2 col4\" >0.957322</td>\n",
       "      <td id=\"T_e277c_row2_col5\" class=\"data row2 col5\" >0.753886</td>\n",
       "      <td id=\"T_e277c_row2_col6\" class=\"data row2 col6\" >0.903333</td>\n",
       "      <td id=\"T_e277c_row2_col7\" class=\"data row2 col7\" >0.981481</td>\n",
       "      <td id=\"T_e277c_row2_col8\" class=\"data row2 col8\" >0.960095</td>\n",
       "      <td id=\"T_e277c_row2_col9\" class=\"data row2 col9\" >0.994800</td>\n",
       "      <td id=\"T_e277c_row2_col10\" class=\"data row2 col10\" >0.705851</td>\n",
       "      <td id=\"T_e277c_row2_col11\" class=\"data row2 col11\" >0.707960</td>\n",
       "      <td id=\"T_e277c_row2_col12\" class=\"data row2 col12\" >87.059108</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e277c_level0_row3\" class=\"row_heading level0 row3\" >3</th>\n",
       "      <td id=\"T_e277c_row3_col0\" class=\"data row3 col0\" >141203712</td>\n",
       "      <td id=\"T_e277c_row3_col1\" class=\"data row3 col1\" >16</td>\n",
       "      <td id=\"T_e277c_row3_col2\" class=\"data row3 col2\" >32</td>\n",
       "      <td id=\"T_e277c_row3_col3\" class=\"data row3 col3\" >8</td>\n",
       "      <td id=\"T_e277c_row3_col4\" class=\"data row3 col4\" >0.953096</td>\n",
       "      <td id=\"T_e277c_row3_col5\" class=\"data row3 col5\" >0.748539</td>\n",
       "      <td id=\"T_e277c_row3_col6\" class=\"data row3 col6\" >0.896984</td>\n",
       "      <td id=\"T_e277c_row3_col7\" class=\"data row3 col7\" >0.977407</td>\n",
       "      <td id=\"T_e277c_row3_col8\" class=\"data row3 col8\" >0.954711</td>\n",
       "      <td id=\"T_e277c_row3_col9\" class=\"data row3 col9\" >0.994900</td>\n",
       "      <td id=\"T_e277c_row3_col10\" class=\"data row3 col10\" >0.695745</td>\n",
       "      <td id=\"T_e277c_row3_col11\" class=\"data row3 col11\" >0.702368</td>\n",
       "      <td id=\"T_e277c_row3_col12\" class=\"data row3 col12\" >86.546874</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x7fc39a9edd80>"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_dir = \"/workspace/svd_subspace/outputs/ViT-B-32/ablation\"\n",
    "moe_data = defaultdict(list)\n",
    "for gate_k in [16]:\n",
    "    for k in [32]:\n",
    "        for top_k in [1, 2, 4, 8]:\n",
    "            file_name = f\"gate_k={gate_k}_k={k}_top_k={top_k}.json\"\n",
    "            file_name = os.path.join(data_dir, file_name)\n",
    "            if not os.path.exists(file_name):\n",
    "                # print(f\"File {file_name} does not exist\")\n",
    "                continue\n",
    "            with open(file_name) as f:\n",
    "                data = json.load(f)\n",
    "            moe_data[\"num_params\"].append(data[\"model_info\"][\"all_params\"])\n",
    "            moe_data[\"gate_k\"].append(gate_k)\n",
    "            moe_data[\"k\"].append(k)\n",
    "            moe_data['top_k'].append(top_k)\n",
    "            accuracies = {}\n",
    "            for key in data:\n",
    "                if key == \"model_info\":\n",
    "                    continue\n",
    "                accuracies[key] = data[key][\"accuracy\"]\n",
    "            average_accuracy = np.mean(list(accuracies.values())) * 100\n",
    "            for key in accuracies:\n",
    "                moe_data[key].append(accuracies[key])\n",
    "            moe_data[\"average\"].append(average_accuracy)\n",
    "moe_data = pd.DataFrame(moe_data)\n",
    "\n",
    "# 使用 Styler 对象并应用背景颜色渐变\n",
    "styled_df = moe_data.style.background_gradient(cmap=\"RdYlGn\")\n",
    "\n",
    "# 显示结果\n",
    "styled_df\n",
    "\n",
    "# styled_df:\n",
    "#\n",
    "#  \tnum_params\tgate_k\tk\ttop_k\tsvhn\tstanford_cars\tresisc45\teurosat\tgtsrb\tmnist\tdtd\tsun397\taverage\n",
    "# 0\t141203712\t16\t32\t1\t0.966157\t0.748290\t0.906508\t0.983333\t0.971180\t0.995400\t0.768085\t0.715667\t88.182758\n",
    "# 1\t141203712\t16\t32\t2\t0.962930\t0.755130\t0.907143\t0.983333\t0.967854\t0.995200\t0.734574\t0.715315\t87.768501\n",
    "# 2\t141203712\t16\t32\t4\t0.957322\t0.753886\t0.903333\t0.981481\t0.960095\t0.994800\t0.705851\t0.707960\t87.059108\n",
    "# 3\t141203712\t16\t32\t8\t0.953096\t0.748539\t0.896984\t0.977407\t0.954711\t0.994900\t0.695745\t0.702368\t86.546874"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABI8AAAFyCAYAAAB8wfpiAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjkuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/TGe4hAAAACXBIWXMAAA9hAAAPYQGoP6dpAADsRUlEQVR4nOzdd3hT1R/H8ffN6qQUKJQ9RIZsZQ8tArJkyp4CMmQICAgKbhkFmTJkyh6yRBFU5AcUAWUKCIgMQQGhLRTobtokvz/ShKZNZ9LJ9/U8fZp777nnnpsUmn5yhmIymUwIIYQQQgghhBBCCGGHKrsbIIQQQgghhBBCCCFyLgmPhBBCCCGEEEIIIUSyJDwSQgghhBBCCCGEEMmS8EgIIYQQQgghhBBCJEvCIyGEEEIIIYQQQgiRLAmPhBBCCCGEEEIIIUSyJDwSQgghhBBCCCGEEMnSZHcDhBBCCCGEEEKInMhkMhEbG4vRaMzupgjhNCqVCq1Wi6IoaT5HwiMhhBBCCCGEECIBg8HA/fv3CQsLIzY2NrubI4TTabVa8uXLh4+PD2q1OtXyislkMmVBu4QQQgghhBBCiBzPYDBw69YtYmJiyJ8/P56enqjV6nT10hAipzKZTBgMBsLDw3n8+DEuLi6UKlUq1QBJwiMhhBBCCCGEECJeYGAgjx49onTp0ri5uWV3c4TINFFRUfz77794e3vj6+ubYlmZMFsIIYQQQgghhMDcKyMsLIz8+fNLcCTyPDc3N7y8vAgLCyO1fkUSHgkhhBBCCCGEEEBsbCyxsbF4enpmd1OEyBL58uWz/tynRMIjIYQQQgghhBACrKuqpWUCYSHyAsvPemorCkp4JIQQQgghhBBCJCCTY4unRVp/1iU8EkIIIYQQQgghhBDJkvBICCGEEEIIIYQQQiRLwiMhhBBCCCGEEEIIkSwJj4QQQgghhBBCCCFEsjTZ3QAhhBBCCCGEEELkfocOHWLVqlUEBgbi6emJoig0aNCAevXq8e233zJ37lz279/P2rVrOXHihPW8+vXr061bN9q3b5+kzmXLlrF+/XqCg4Px8PBgwIAB1KxZkxkzZnDjxg1rOZVKRfny5fn444+pU6cOAIMHD+bs2bOEhYVZyxUoUICXX36Zs2fPcuPGDUwmEwCvv/46kydPtntf33zzDYsXL+bWrVsAuLm5UaFCBbZt2+b4k5ZLKCbLMyWEEEIIIYQQQjzFoqOjuXHjBuXKlcPV1TW7m5OrrFy5kgULFjB79mxatWoFQEhICKtWreKrr76iZMmS/PzzzwCYTCbeeust6/aZM2fw8PBItu6ff/6Z0aNHs379emswBLBt2zbef/99AJYvX46fn1+Sc6OioujatSvXrl3j+eefZ926deh0OgD+/fdfxo4dy8WLFwGYOXMmnTp1SrYd8+fPZ8OGDfzwww8ULlw4Hc9OzpXWn3kZtiaEEEIIIYQQQuQABqOJX68/4Nuzd/j1+gMMxtzR1+PGjRvMnTuXvn37WoMjgIIFC/LOO+8wduxYm/KKojBq1CjrtiW8Sc61a9do06aNTXAE2Gw3aNDA7rlubm5UrlwZgNq1a1uDI4DSpUuzevVq6/YHH3zA+fPnk21HnTp1KF++fJ4JjtJDwiMhhBBCCCGEECKb/XjhLk1mHqDXit8Ys+UsvVb8RpOZB/jxwt3sblqqAgICMBgMFClSxO7xgQMHUqhQIZt9lStXpl69egBs3749xfq/++47+vfvn2S/Vqu1Plar1cmebymn0SSduSd//vzW73q9npEjRxIUFJRsPQnDp6eJhEdCCCGEEEIIIUQ2+vHCXYZvOMPdx9E2++89jmb4hjM5PkCyzIazfv167t+/n+S4Tqejfv36Sfa//vrrAOzdu9fueQDHjx/HxcWFWrVqOa/BdixYsACNRkNQUBBvvfUWer0+U6+X20h4JIQQQgghhBBCpJHJZCJSH+e0r7DoWD767iL2BqhZ9n383SXComOddk1nT33csGFDAO7cuUPnzp0JCAhIUubtt99Osq9Zs2aULFmS2NhYtmzZYrfuLVu20LNnT6e2156GDRsyZcoUAM6ePctHH32U6dfMTWS1NSGEEEIIIYQQIg1MJhNdl/7K6X8eZt01gXuh0VT/eJ/T6qxTpgDb3myIoihOqa9y5coMHjyYlStXEhQUxNChQ2nWrBkTJkygfPnyyZ6nUqno168fM2bMYMuWLQwdOtRmWNiDBw/49ddfmTp1qlPamZrevXtz7do1Nm7cyM6dO6lcubK1d9TTTsIjIYQQQgghhBAijZwTt+Q977zzDoULF2bevHlER0dz4MABDh8+TLdu3Rg7dize3t52z+vSpQsLFiwgODiYvXv32qx2tmPHDtq0aZPiSmwWvXv3TvbYv//+m+b7mDx5Mjdv3uTo0aPMmjWLihUrWntWPc0kPBJCCCGEEEIIIdJAURS2vdmQqFiD0+o8cSOEAatPplpuzcC61CtX0CnXdNOqndbrKKEBAwbQvHlz/P392b9/P3FxcWzevJl9+/Yxd+5cuyui5cuXjy5durB+/XrWrVtnDY+MRiNff/01S5YsSdO1N23aZHdCbIB3332Xb775Jk31aDQa5s+fT/fu3blx4wZjx45l+/btlCpVKk3n51USHjmJ0WgkKCgIDw+PTPlHKIQQIm8xmUxERERQpEgRVKq8OwWh/H4UQgiRHrnh96OiKLjrnPen9IsVClMsvyv3HkfbnfdIAYrmd+XFCoVRq3L+79JSpUqxePFiTp48ib+/PxcuXODBgwcMGzaMnTt32h3G1rdvXzZs2MDFixc5ffo0tWvX5pdffqFIkSJUqlQpy+/By8uLpUuX0r17dx49esSIESOSnZPpaSHhkZMEBQXh5+eX3c0QQgiRywQEBFC0aNHsbkamkd+PQgghMiKv/35MSK1S+Kh9FYZvOIMCNgGSJSr6qH2VXBEcJVS3bl22bdvG2rVrmTVrFtHR0SxdupTPP/88SdmyZcvi5+fHoUOHWLduHbVr12bLli306tUrG1r+pE0LFixg8ODBXLlyhXfffZc+ffpkW3uyW64Jjx4+fMiaNWs4dOgQ3377bbLlLl++zJw5cwgODsZoNNKuXTveeOMN1Gq13fInTpxg4cKFhIeHA9CrVy+6d++e7vZZxmAGBATg6emZ7vPzKoPBwPnz56lRo0ayr4HIHvLa5Gzy+uRcznptwsPD8fPzS9MY/txMfj8KIYRIj6fl92NirasV48u+L/DJ7kvcfRxt3V80vysfta9C62rFsrF1qduzZw8lSpSgVq1aNvtVKhUDBw7k0aNHLF26lIsXLyZbR//+/Tl06BD79+/n999/548//mDBggWZ3PKUWVZg++STT9i3b581N3ga5fjwKCwsjPXr17N582aCgoIoUaJEsmVPnTrF0KFD+fjjj+nQoQPh4eH06dOHy5cvM3fu3CTlf/jhByZPnsyXX35JgwYNCAwMpFu3bvz7779MmDAhXe20dMX39PSUN8cJGAwG3N3d8fT0lD+Acxh5bXI2eX1yLme/Nnl9KJf8fhRCCJERef33oz2tqxXjlSpFOXEjhKCwaIrkc6VeuYK5psdRQEBAkvDI4tVXX2Xp0qV4eXkle37jxo2pUKECV69eZeTIkXTq1Mlm5bXs0rt3b65fv86GDRs4duwY9erVy+4mZYucOYg0ARcXF4YMGcKqVatSLBcREcHYsWOpW7cuHTp0AMxvVCdMmMCePXvYtm2bTfm7d+/y7rvv0rlzZ+ukXb6+vowYMYIVK1Zw9OjRzLkhIYQQQgghhBDCDrVKoWH5QnSsVYKG5QvlmuAIYPPmzdy9e9fusWvXrgHQrl27FOvo168fACEhIfTs2TPVa0ZHP+mlpdfrky0XExMDQFxcXLLHYmNjkz3/vffeo3Hjxqm2Jy/L8eGRTqdDq9VSrly5FMutXr2a4OBg2rRpY7O/cePGeHh4sGDBApsflC+++ILo6Ogk5Vu2bAnAnDlznHQHQgghhBBCCCFE3vbw4UN69erFTz/9ZBPEnDx5kmnTpvHKK6+kOodRhw4d8Pb2pnHjxpQuXTrVax4/ftz6+JdffrFbJjQ0lPPnzwNw7ty5JCHRiRMnADhz5kyy17GswJZaLpGX5fhhaxZarTbF4z/88AMAVatWtdmvUqmoVq0ax48f5+jRo/j5+REXF8fPP/9st3zBggUpWbIkFy9e5Nq1azz77LNOvAshhBBCCCGEECLvef/99/Hy8uLgwYMsXboUlUrF48eP8fb25q233qJ79+6prqDn5uZGt27dkh3+ZhEQEMD06dP5559/rPvGjBlD+fLl+eSTT6hTpw4AgwcP5vfff7fOVXTy5EmaNGlCixYteP3113n77be5efMmAIMGDaJcuXLMnTuXihUrJrmmZQW2p7WjSa4Jj1Jy9+5daze44sWLJzleqlQpjh8/zpkzZ/Dz8+Ps2bOEhYXh7e2Nu7u73fK3b9/m999/TzY80uv1Nt3iLD+MBoMBg8HgjNvKEyzPhTwnOY+8NjmbvD45l7Nem7z62ib3+9ERRqOBO39eJPzRQzy9C1DiuaqoVLl3LrC8dj95ibw2QmSM/NsRr776qvVxx44dHaorLfMP+/n5pWk115UrV6Z4fM+ePWluF5hXYFu4cGG6zskr8kR4dPv2bQDc3d3tzsqfL18+AGsqaSlfpEgRu/VZylsSSHuWLVvGokWLkuw/f/683UDqaffHH39kdxNEMuS1ydnk9cm5HH1tIiMjndSSnCW5348ZdfX4MQ6sWU54yH3rPs+CPjQbMJQK9Rs57TpZJa/dT14ir40QGSP/doR4OuSJ8CgkJAQg2dDGzc0NMK/clrC8ZX9q5e0ZNmwYAwcOtG5blpSsUaOGrCaTgMFg4I8//qB69eqyYlQOI69NziavT87lrNcmry71mtzvx4y4evwY382dnmR/eMh9vps7nQ7jJueqP0zy2v3kJfLaCJEx8m9HiKdHngiPLLOjJzcvkmVogOV4esvbo9Pp7C4bqFar5Q89O+R5ybnktcnZ5PXJuRx9bfLq65rc78f0MhoNHFizPMUyP69YhKJSUOKHRiiKAvGL0igoEL/Ms2I+mGA7QTlFsZQwH1YU83HLiTbnJChnPcdc15NzFNvlpeMfm0wm9q9akuL97P/qS7yLlUClViWoQ+HJQyXVdqS1XUqC5+PJ85DwGhl7fp5UmeB6ConanbNWDkrLz9rBtcspX7e+DMMRIgH5tyPE0yVPhEdeXl5A8kvzWYYGFChQIEPlhRBCCJG17vx50WYIhD1RYaF8O3taFrUo80U+esi6d0ZmdzOyVrrCqqThYFrCKnMdCa6XKDQzxMURHZ58b3OAsAf3+XJYPzRaO8GoyZT2+01j2XTUmK7rm9LT1nQ1wfnPAWTe85D21yE9dabn8um5s/S8vumpNj2vg/2yRqMRYwpLm4P5386dPy9SqmqNdDROCJET5YnwyLJcXmhoKEajMckM7o8ePQKeTKZtKf/w4UO79SUuL4QQQoisFf7I/u/oxPL7FsUtnxeYTPF/C5n/yDGZTNa/uUyYzH8oxf+xZC6WoFyC75gS/JlkMpnPth5L8EdUOq8XFxNDTGREqvejdXFBpdE8uZ71DzzTk7/1TKb4ayTaTmu7cpJE95LDWmcjOjQ0u5sgRK6U1v/PhRA5W54Ij0qXLo2vry+BgYHcuXOHUqVK2Ry3TJDdqJF5vG2NGjXQ6XTcu3cPvV6fpHu9pXzjxo2zoPVCCCGESMzTO229f1sNG50rPtG+dfE8Wz+dnGq5zpM+yrL7MdkEXAkCHGtoZi1oE04lDqOe5FF2grbE4ZylvgTbqbUjyTaWXC/hOcm3w3bbNhw0mUzcu36Vn5envnJOiyEjKfpMBfsH0zkULzOH7qWr7vS2O30NSV/d6SqfzucvXU9Jel+btJdPf9XpqDvdz4nj7f7v6mX2fjE71fPT+v+5ECJnczg8unXrFocPH+b69evcv3+fiIgI8uXLR8mSJalatSovvvhipk8grSgK7dq1Y9WqVZw/f94mPDIYDPz1118ULlyYWrVqAeDp6UmzZs348ccfuXTpknU/wP379wkKCqJatWr4+vpmaruFEEIIYV+J56riWdAnxaFr+Qr5UOK5qlnYqozLifeTcN6jnDULUdbyKV2GX7dvTvW1qd6spczbIkQC+XwKc3jDmhz1/5oQIvOoUi9i3549e2jfvj0tW7Zk6tSpbNq0iX379nH06FF+/PFHVq1axbhx42jcuDHvvPMOf//9t0MNNRqNwJPJrBMbNGgQXl5e/PDDDzb7Dx8+TGRkJG+++abN5KQjR45Eq9UmKf/jjz8CMGrUKIfaK4QQQoiMU6nUNBswNMUyL78+NNf8MZ/X7icvkddGiIyRfztCPF3SHR7duHGDDh06MH78eG7evEnFihVp1qwZvXr1YtCgQQwfPpyBAwfSrVs3GjduTKFChfj+++9p37498+bNIy4uLkMNtYRPISEhPHjwIMlxHx8fpk6dyqFDhwgICADg7t27zJgxg1atWtG7d2+b8hUrVuSdd95h69atXLhwAYArV66wcOFCBg4cyMsvv5yhdgohhBDCOSrUb0SL5m1xNRht9rsajLRo3jbXLf+c1+4nL6lQvxEdxk3Gs6CPzf58hXxkqXEhUiD/doR4eqRr2FpAQAATJ06kWbNmTJkyheeffz5Ny/EGBwcTEBDA5s2bOXv2LF9++SXu7u5pvm6HDh34999/AfMKaS1btqR+/fosWWK75G2rVq3w9PTkiy++YMGCBYC5R1L37t2TTKIN8Prrr+Pj48OHH34ImJcX/vjjj2nTpk2a2yaEEEKIzBG6bx+6eYt42WQixMOVGK0Gl9g4CkbGoFxcRGjZZ/Fq2TK7m5lmee1+8poK9RvxzAt1uP7tLsLv/odnseKU79gJtb0V1oQQVhXqN6J83frmVTIfPcTTuwAlnqsqPY6EyGPSHB6dPXuWRYsWsXHjRp599tl0XaRw4cJ07dqVrl27snv3bt555x0WLVqU5gnpvvvuuzRfq3Hjxuma6PrVV1/l1VdfTXN5IYQQQmQ+k8FA4PQZYDKhAIUiom0LKAqB02eQr3lzFHXO/wMlr91PXhS6bx+B02dguHcPN8AA/L3oS3wnvyehnhCpUKnUuWLxAiFExqUpPIqJiWH16tWsXr3a4cmv27dvT5kyZVi3bh2vv/66Q3UJIYQQIm+KPHWauHv3ki9gMhF37x43+/VHUzDnr+QTF/IwTfdza/hwtL5FQa1CUWvM31XqZLcVtQpUahSN2vw9xW2VOZhSq83f47et+1QqsNSZlm11gmskrDMTVxLLLKH79nFnzFhIuCocEBcYaN6/YL4ESEIIIZ5qaQqPzp07x5QpU5y2alqNGjW4fz/5WfmFEEII8XSLCw5OU7noM2cyuSVZK+LwL9ndBMclDKlUCYKl5LY16vhALPG2OSizBGFJAjN1fJhlU6fqSV0227bhW8JwzYTC/cWLkwRHgHXfvY8+RtFqUbQ6m7DMfA3Nk7ZYjmni25XwuyVs06htjwmRB5gMBnPoHxyMpnBh3OvUll6UQuQxaQqP6tWr5/QLN2vWzOl1CiGEECJv0BQunKZyBQYMwKVcuUxujeNibtzg4Zo1qZbL360buuLFMBmMYDRgijOYvyfYNhkNYDCavyfcNhjAYMBkNJq/J9yOi3uy32jEZIgDS52GJ9tP6jTaPze+zhQZjebysbHYiWNyJcPDh9wePsL5FSuKbZiW6HFK+ywhm7191mMa9ZOALeExe4GXJVyzOZYoeFNrngRvlt5siYIxa0Bmr8327scSosVf12afyBUsQz4T9q7UFC0qQz6FyGPSNWF2Sm7fvs3y5cs5cuQIwcHBeHh4ULNmTQYPHkzdunWddRkhhBBCPAXc69RGU7QocYGB9nuEKAoaX19835mQKz7dNhkMhP34Y6r3U+zjj3LH/dgLqOK3TQaDOUCyBl+Jt41giHsSiNnZtg23LNeJD7cSbhuShmvJbdsEZAm29bdvEX3ufKr3rC1ZElW+fE/CtLg4c9sSBm7x+6zPRXyIlvwTaTKfE78acV4J25wmUc8tu7257AZe6kTH7ARY1t5ndgK5hAGZOplATp2oF10yPdDMPfESHEsQuqU9fEsQrOWw4aEy5FOIp4dTwqMrV67Qp08fwsLCyJ8/Pz4+PsTExHD48GGOHDnC/PnzeeWVV5xxKSGEEEI8BRS1Gt/J75n/+FAU2z9M4v9o8p38Xq4IWiAP3o9KZf4DVqvN7qY4LOL4Cf5NwzycxaZNw6N+xnrjJwnbEj227ouLsw3a4gzmYM1oNB8z2IZt5gAsYR3xx+IMiXqmGWzqtB5L2AbLMUvIZtPWOJvAzWQwWEM0kyHONjxzVrAG5nMA9HoJ1hJLNDTSpseWJkFvs4TBlUOhW9KeZyYUHm3alPyQT1kIQIg8Jc3hkcFgQJ3MP/p58+bx5ptv0q1bN7y8vKz7IyIi+Oabb5g7d66ER0IIIYRIF6+WLWHBfO5N9+d+tCcxOi9c9KH4uIZTdPK7ue7TbMv9JBne4esrwzuyUVp7ubnXqZ3ha+SlsM3ZnBasWY5JsJZzxC8EEHnqdIaDVyFEzpHm8Khfv374+/tTunTpJMcCAwPp378/2kS/ED08POjTpw+bNm1yvKVCCCGEeOoEF67FsQafEvFIb93n4a3jxcIV8UrhvJzKq2VL8jVvLhPL5iB5rVdYbiPBWvJyerAWc+0aEUePpnofaV0AQeQtR48eZefOnfzxxx+oVCoKFy6Moig0a9aM9u3b8+OPPxIbG8uAAQOyu6kijdIcHr3//vuMGzeOnj170rVrV5tjL7zwAr1796ZTp06ULFkSNzc39Ho9//33H3v37qVixYpOb7gQQggh8rbrvwfx47ILSfZHPNLz47ILtB5WjfLPF8mGljnGpKh45F2BCFVpPLxccFNU5IzZS55e0itM5EQ5PViLOH4iTeFRWhdAEHlDWFgYkyZN4siRI4wePZqPP/6YfPnyARATE8M333xDmzZtePz4Me+99142t1akR5rDoypVqrBx40ZmzJjB4cOHmTp1qnWI2rhx43jrrbf47LPPbCZvM5lMVK9enXnz5jm/5UIIIYTIs4xGE798fTXFMke2XqVczcKoVLknern+exC/fH2ViEcx1n0e3i682KNCrgzC8hLpFSZE+mTFkM+nktEA/xyD8EDw9IUyjUCVO/4fevz4Md27d+fOnTt89dVXSVZtd3FxoWfPntSoUYNevXplUytFRqVrwmwXFxc+/vhjDhw4wOuvv86kSZNo0KAB7u7urFq1ipMnT3Ls2DEePHiAl5cXdevW5aWXXsoxqwEIIYQQIne4e/WRTcBiT/jDGLZOO4FnQVe0Lmrzly7+u6vauk8Tv0/nqkbrokGjU6F10ZjL6FSo1FmzJHjyPalicnVPqrxEUatlbhYh0kiGfGaCS9/Bj5Mg9L8n+7yKQ+uZUKVD9rUrjSZPnszNmzcZOHBgkuAooSpVqjBy5MgsbJlwhgytttasWTOqV6/OlClTCAgIYPz48Wg0GurWrUvdunWd3UYhhBBCPGUiQlMOjiwe3IngwZ0Ih66l1qiehE+utoGTdV/CbXv7XBOdo1OjJOgRlVd7Ugkhnm4y5NOJLn0HW/tD4inQQ++a93dfl6MDpIMHD7J//34ABg4cmGr53r17s3fv3sxulnCiDIVHAIULF2b58uWsXbuWvn37Mn36dJ555hlntk0IIYQQTykPL5c0lavTtiz5CrkSG2OwfsUleGzvK05vQB9twGQ0v0E3xBkxxBmJjkhldaN0MvdwModJQJp6Up3+4SZFy+dH56pB56pG56qxBlrSk1sIkRM9lUM+TSaIjXRefUYD/DCRJMGR+WKAYu6R9ExT5w1h07pbe4g5w9atWwGoWLEivr6+qZb39PSkbdu2ABiNRrZv387WrVtRFIW7d+9StmxZRowYQaNGjaznREZGsnXrVlatWsXmzZu5ePEin332GcWKFWP16tV4enqya9cuNmzYQExMDDdv3kSv19O8eXOWLFnitHt9WmU4PLJ4/fXXadCgAe+++y6dOnWid+/ezmiXEEIIIZ5iRcrnJ0Jtwt0Aip3ppE2YiFBDvtqF8HLX4q7T4K5T46JRpSlkMZlMGONM5kBJbyA22hIuxRGrN5q/Rz8Jm2JjEpTRJwqkom33Wd77x+mNxOmNRIWlPZQ6sfuG3f2KAtr4QElrDZbiwyU326BJ56pB56ZG52L+rk0QROlcNai1WTNMTwjx9HiqhnyaTPBVK7h1PCsvah7K5l/KeVWWagCDfnRKgGQymThx4gQAlSpVSvN5np6eAHz++ed89dVXrFmzhoYNGxIYGEjXrl0ZPnw4+/btw9fXl6NHjzJ9+nSuXbsGwIULF/jkk08ICQkhODiY8+fPExkZyeLFi/n6668pWLAgQUFBMjzOiRwOj8D8A7JhwwY+//xzhg8fzvTp0ylQoIAzqhZCCCHEU+jUPw/52UVPx0gdJkw2AZIpPp3Z76Lny4VHbM5TKViDJPNX/GMXDe7a+H0uavtlLI+91Ljr3PFIWFarTtNwMpPJRFysMUnvp7vXH3Fsx/VUzy9QzANFAX20Obyy9JAymUAfFYc+Kg5I25C+5Kg0SjLBktocRLmo0blpbMMoV/O+hNtaV02eG2JnNJrM822FxuDh5UKxCt557h6FEM4g/y8k9PDhQ8LDwwEylANs2bIFgIYNGwLg6+tLs2bN2LJlC+fOnaNly5Y0btyY3bt388orr3D79m327t3LgQMHOHbsGKdOnaJOnTpMnDiRZ555hoIFCwJQpEgRPvnkExYvXuykO326pSs8iouLY+/evfz555+oVCoqVqzIq6++ikajQafTMWXKFH755RcGDRrE+PHjadKkSWa1WwghhBB5WFBYNFd1Rr5FT7MoLV6mJ2/UwxQTB9xiuaoz4qFTE2c0ERNnBMBogvCYOMJj4pzeJletKuXQKcF3N53aHD7pNLi7qHEppiNcZcLDmHxPqkg1DJ1SF63mSc8gSxilj7KESXHoow3m7Rjz94RBkz46LtExg/V4bIzB/BzFmYiOi3XKMD2NThXf08kSKD3p4WTtJeWmth9EJQipNLq09RjLTLISnhAiTRTF3GPHmcPW/jkGG7umXq7PdvPqa87gxGFrsbFPfp+4u7un+3w/Pz8MBoPNPksAFBn55HlWqVT4+vpy+/ZtBg0ahJubG82bN6d58+aAefjb4cOH2bJlCz169EBRFKpUqSLT6zhJmsOja9euMWzYMP777z9MCWbSX758OV999ZV1XOOLL75I1apVef/99zl06BATJ05Ep9M5v+VCCCGEyLOK5HMF4KrOyDVtDCXjVHiYFCIUE7c1RixZ0srX69KwfCHiDEaiYg1E6s1fETFxT7Zj4uL3m79H6A1E6eOsZSNtHifYjokjMtZgXUAoOtZIdKyekAzOz13BVZViT6qfXfRsnfE/8rtrcdOpcdOqcdWav9ts69S4atS46VS4uahx9VTjptPhplXjrlVTML6s5TxXrRpXrQqNohCnTxxEJXgcZbDdjk60HR9GxUYbMMSHdeaheXoiQ/UZe1LiJRyWZxssJQ6eLD2mnmzbDNVz1aDWpH9YnqyEJ4RIF0UBnYfz6ivfzLyqWuhd7M97pJiPl2/mvDmPnMjb2xuVSoXRaOThw4fpPn/+/PnWx6dPn2bnzp2cPHkSMAdCCanj59Ly8fFJUk/fvn353//+x0cffcTXX3/N0KFDadWqFePHj093m0RSaQ6PPvzwQ8qWLcvIkSMpWrQoALdv32bnzp3Mnj2bzz//3Fq2YMGCLFmyhE2bNtG3b1+mTZtGhQoVnN96IYQQQuRJ9coVpFh+V+49jsakwC2t7ZtHBSia35V65cyfTGrUKvKpVeRz1Tq1HSaTiehYY6KAKZnH8WHTk7DqyfH/HkVx9WFUqj2piNBzP8KxICY5KgWbQCnJ44Rhk5cKt0Jq3LQ66z53rZpC8cGUTlHQoaA1gtpgQmUwocSZUOKMGPVGO0FU8j2nTCZsh+U9dMKwvGR6ONn0frJMRq5Tc/jrKynWKSvhCSEylUoNrWfGr7amYBsgxf+/09o/RwZHAC4uLlSoUIG//vqL27dvZ6iOo0ePMm/ePKpUqcLQoUMpWrQoixYtSlcd9erVY9OmTXz66adcuHCBsWPH8txzz+Hv70/lypUz1C7xRJrDo0ePHrFp06Yk+zt16kS3bt3sntO7d2/q16/PlClTrOMYhRBCCCFSo1YpfNS+CsM3nEnubTQfta+COpP/mFcUxdzrR6emkAP1/Hr9Ab1W/JZqT6qpnarxbBFPomINROsNRMXGf+kNRFsfG4mOS+G4pUysObyKX1QOowki4nteZSaNSjH3kkrQA8pVp8bNVYVbPjVuOg2uWhdzaKVR4aZS4aao0KHgYgKdCWsopTZgDqZijZjiTJj0RoyxRozxk5xbgih9/Cp7ED8sLzyW6HDnrZ4X/jCGpaMOoo6f+0qlTvClUlCpVclvq+0dT7Av8ba1XOI6UjgnuTpSOUeRMEw4icwX5gRVOkD3deZV1UL/e7Lfq7g5OKrSIfvalgZt27blr7/+4uTJk0RGRqZr+NoXX3zBsmXL+PLLL3nppZccakfNmjXZsWMH//vf/5g9ezZ//vkn/fr1Y+/evRQuXNihup92aQ6PtFotBw4coFmzZtZ9cXFx7NixA29v72TPK1++POvXr3eokUIIIYR4+rSuVowv+77AJ7svcfdxtHV/0fyufNS+Cq2rFcvG1qVPWntS9apX2qmBmMlkItZgModRsQkDKIN1X5TeaA2d7AVST0IpY4qBlWV4X5zRRFhMHGGZMO9UQjq1CletCjfv+JBKo8ZTo8ZDpcJDMQdSrii4KpZQSkFjBE3C3lKxJuJC9egfpt7by2TEGlDlFYqCndDKHDApagV1ggBKUSXcVlDiwy2bMvGP1SrbbUs5JUGwlnjbXtCmTnCd9IRz2T1/1tNG5gtzoiodoPKr5jmQwgPB09c8x1EO7XGUUO/evdmwYQPBwcGsX7+eYcOGpVg+NjaWpUuX8vLLL7N48WL8/PwcDo78/f159913AWjevDkvvvgiI0aM4JdffuGHH36gf//+DtX/tEtzeDRs2DBGjBiBm5sb3t7eGAwGHj58iNFoZMWKFSmeq9U6twu5EEIIIZ4OrasV45UqRTlxI4SgsGiK5DMPVcvsHkfOll09qRRFQadR0GlU5HfLvPdjJpN50nJLmBQda0wUUD0JmZJs6xOEUykeN9droTcY0RuMhEY7FlKVilXRE5dUy33nruee2ohGpaBVFHSKgkZR0KhAa3lM/HdFQaOAGvM+tQJqRUFtAjWgURRUKKgxxX8HFQoqE6jA+l0x2T42f5lsHmM0Pyb+sfW70Twe0GR9bO91A0OcEUPm5nxZTlElDpsc76VlfzuFICs95yTpZaZK0sNNyaGhmGW+sMRzuYU/ipb5wjJKpYZyL2Z3K9LNy8uLOXPmMGTIEJYsWcLzzz9PvXr17JY1GAzMmjWL/v37c+GCeb65hJNuA+j1emtZMP+eURTFOv9y4gm2AW7cuMFvv/1GgwYNANDpdPTq1YtffvkFjcYpC80/1dL8DLZt2xYPDw9WrFjBpUuXUBSFunXrMnz4cOrWrZuZbRRCCCHEU0ytUmhY3pFBYzlDXupJlZiiKPETc6vxzsTrGONX1ku2d1SSwMqYbGAVHWfeF/g4htBII/lMSrIr4YUpJq5oDfHDC02WA+YvY5JTspbCkwQypc4J8SGUmieBlPnLHGpZt63llETlQG0yrxaotilHgsBMQUN8UMaTAE0dX06N8qSu+MfmOs3H7Ydl5n3YBGcm63NvL04xGU0YjCYMzhu5mCMkDqDs9Q5L2BvM0rMr8ba1V5fdnl5pGYZpDrwUBfat+zNJcATmnxMTJn7eeFnmC3uK1K9fn2XLljFp0iQGDRrEkCFD6Nmzp3VxLYAzZ87wzTff0L9/f0qVKgWARqPh119/Zd26ddSqVYvvv/+eo0ePAnD+/Hn0ej0NGjSgRIkS1jmVjhw5QpkyZZK04Z133mH+/PnUrl0bo9FIQEAAhQsXpnXr1lnwDORt6Yrf/Pz88PPzy6y2CCGEEELkaXmlJ1V2UamezEHlLL9ef8CHi0+kuBLeAbdYlvevTa3SBcyhhMmEwWAizmi0bscZzIFFnNGE0WbbaLNtMMXvT7AdZzRhMBifnGu01J9gO8GX9brGJ8fijCaMdreNdo8nPt/yONZotFtXTqUkCr7U1scJg6qEgZliLafYHFPs1PFkf5LQzaQ8qSNRGKex+Z4gKMP2OvZ6mVnuxx5j/OtCLEDOGUJpL3S17DeEx3H7ykNKVy6Yxa0S2aVhw4bs3buX7du38/PPP7Nt2zZ0Oh3PPPMMHh4evPTSS0yZMgVXV/OqqqVKlWLatGksWLCAhQsXUrNmTcaOHcsrr7zCm2++ydmzZ2nbti1Xr16ld+/ePHr0CIDPPvuMr7/+mu+++87m+kFBQfTu3ZvixYuTP39+KlSowNdff03BgvIz6Kg0hUe//fYbFStWdOoTnnj+JCGEEEKIp0Fe6UmVV9QrV5Dwwjq+C9bzsp2V8A66xRJeWEez53yf6pDPJnhyIDyLszk/YThmJM5gG5bZ2zaHb0ZrwGYN49IQntkLzBKWjUkmaEu6bbRORJ8pEvTGsgmaAHWCUOxJ2KUkDbFMtsFVsnUk7JGWuLdZKr3QPIwKBUzJRV1P/HEtRMKjp4ynpycDBgxgwIABaSrfqVMnOnXqlGT/6dOnbbZT6z20bNmytDZRZECawqPq1aszdepUPvvsM6eMFTx8+DBBQUEO1yOEEEIIIYQjEs5HdU0bQ4kEK+HdiV8J78ssWNkvp1OpFHRP+XOQkMmUKIiyF4gZEoRniXqbJe59llx4ljAcMxiMGEwkG54l7Z2WoGyK4ZmRmFTCNdvH5pDNJ9JEl1Bdqs9VuJJze64JIdIuTUmQh4cHr732GkOGDMHf399mzGJ6bdy4kTNnzjBnzpwM1yGEEEIIIYSzJJyP6laC+aiK5YH5qETmUBQFjVpBk/MXwco0x67eJ2Du2VTnC3u+YoFsaJ0QwtnS3I2obt269OnTh1dffZX27dvTqlUr6tSpk6aeSEFBQRw6dIjNmzdTtGhR5s2b51CjhRBCCCGEcCaZj0qI9KlfvhDzfRReCibZ+cJ+91GYKMN0hcgT0jUGrUWLFqxbt46JEyeyZcsWXF1dKVeuHM888wwFChTAw8MDgOjoaMLCwrhz5w7//PMP9+7dQ6PRMGzYMEaNGpUpNyKEEEIIIYQjZD4qIdJOrVIY1KMKc786S7MoLfnszBc2rkctCWCFyCPSPYFRlSpV2LVrF9u3b2f16tVcunSJS5cuoSi2/ymYTOa02cXFhddee40RI0ZQokQJ57RaCCGEEEIIIUS2al2tGAyCT7+7hPpBjHW+MEMhHR92qCVDPoXIQzI0+7VGo6Fnz5707NmTixcvcvLkSf766y8ePnxIbGwshQoVwtfXlzp16tCgQQNcXFyc3W4hhBBCCCGEENlMhnwK8XRweOm0qlWrUrVqVWe0RQghhBBCCCFELiNDPoXI+1TZ3QAhhBBCCCGEEEIIkXNJeCSEEEIIIYQQQgghkiXhkRBCCCGEEEIIIYRIloRHQgghhBBCCCGEECJZEh4JIYQQQgghhBBCiGQ5FB6dP3/eWe0QQgghhBBCCCGEEDmQQ+FRv379uHHjhrPaIoQQQgghhBBCCCFyGI0jJ8fExNChQwdatmxJr169qFOnjrPaJYQQQohMFh0dzW+//cb169e5f/8+ERER5MuXj5IlS1K1alVq1KiR3U0UQgghhBA5gEPhUeHChVm9ejW///47n3/+OZGRkfTs2ZOOHTvi6enprDYKIYQQwonOnj3L0qVL+fXXX9Hr9cmWK1KkCJ07d2bAgAF4e3tnXQOFEEIIIUSO4lB4tGnTJkqVKsWzzz5Lt27d+Ouvv9i6dStt27aladOm9OrVi+eee85ZbRVCCCGEA0JCQpg8eTIBAQGYTCY8PT0pV64cRYoUwc3NDRcXF6Kjo4mIiOC///7j1q1bLF26lPXr1zNp0iS6d++e3bcghBBCCCGygUPhUalSpWy2K1WqxAcffMDEiRPZvn07PXv25LnnnqNnz560bdsWnU7nUGOFEEIIkTEXL15k5MiRlC5dmmnTplGvXj1KliyZ4jlxcXFcvHiRQ4cOsXjxYk6ePIm/vz9qtTqLWi2EEEKInO7IkSMcPXqULVu2EBkZCYBGo8Hb25uoqCgKFixIlSpV6NevH3Xr1rU5d9++fZw8eZKvv/6amJgY3N3dKVCggE0ZvV5PSEgIBoOBGTNm8Nprr9kcv3r1KosXL+bixYt4enpiMBioXr06fn5+HDx4kDFjxlC0aFEAunTpwvXr14mKirKeX7hwYbp168aYMWMyVGdid+/eZfPmzfzyyy+EhoZSqFAhtFotlSpVolOnThQrVoy3336bDRs2pPi8vv/++2zbto2//vor2TIffvghX3/9tc0+d3d3Dh06RP78+VOsP70cCo/sCQsLY/369axfv56YmBjOnz+Pp6cny5cvx8/Pj4EDB1KkSBFnX1YIIYQQybhx4wbvvvsu06ZNo3Hjxmk+T6PRULNmTWrWrMnw4cNZvXo1U6ZMwd/fPxNbK4QQQojcpEmTJjRp0gR3d3cWLVpE9erV2bx5M1qtlvDwcNasWcOSJUv46aefePfddxk4cKD13JYtW9KyZUuio6PZunUrzZs3Z/bs2Umu8fDhQ0aMGJFk/8mTJxk8eDD9+/fn888/R6vVotfr+fbbb5kyZQqhoaE2odCOHTsICQmhdevWPH78mFatWjF//nxUKlWG60xoy5YtTJs2jZdeegl/f38qVapkPXb27FmmTZvG+fPnqVixYorP6b59+9i2bVuKZQIDA9m1a1eSD/W6dOni9OAIHFxtbe7cudbHDx48YPbs2bz88sssXLiQ8PBwunbtyt69e1m5ciXffvstzz33HEOHDmXz5s0ON1wIIYQQqTMYDMydO5cVK1akKzhKTKfTMWzYMF555RV27NjhxBYKIYQQwsJgNHDy3kn2/r2Xk/dOYjAasrtJaVasWDEAtFotWq0WAE9PT0aNGkX//v0BmDVrFleuXElyro+PT4p1FyhQgNGjR9vsMxqNTJo0iWeffZbx48dbr6nT6ejWrRtLly61CYUsChYsyDPPPANAgwYNbMpktE6AmTNn8tFHH9G5c2cWLVpkExwB1KpVi/Xr16e60FhgYCCzZ8/Gy8srxXKWD/UuXbpk8/X++++neF5GOdTzaPXq1RQqVIgrV66wZ88eoqOjcXd3Z8CAAQwYMABfX19rWa1WS4cOHWjWrBlt27Ylf/78tG3b1uEbEEIIIUTyfvvtN8aOHZts1+r0at68Od98841T6hJCCCHEE/v/2Y//CX8CIwOt+3zdfXm33ru0KNMiG1uWNsmFKgDdu3dn9erVGI1GfvrppyQ9b1I61+KFF17g8uXL1u2//vqLO3fuJNuLp3bt2jRt2tTuMUsoZPnuaJ0///wzX331FUWLFuW9995DURS75+t0OmbOnJlszyWTycS7777L+PHjmTlzJqGhoXbLPXr0iIMHDzJu3Di7xzODQz2PYmNj8ff3Z8eOHbi6ujJq1CgOHjzIpEmTbIKjhDw8PIiLi2PRokWOXFoIIYQQadC4cWPKly/v1Do7d+7s1PqEEEKIp93+f/Yz7tA4m+AIICgyiHGHxrH/n/3Z1DLnKFGihPXxo0ePMlSHi4sLNWvWtG4bjUYAfvnlF06dOmX3nBdffDFd18hInXq9nk8//RSAXr164ebmluI1ihcvnmxHmq+++opSpUrRqlWrFOtYt24dDx8+5N1332XXrl3WuaYyk8NzHrm5uTFy5Eh69+6d6pME5smjQkJCiI2NdfTSQgghhHCC8PBwNm7cyJEjRwgODsbDw4OaNWvSr18/ypUrl93NE0IIIXIUk8lEVFxU6gXTyGA0MOPEDEyYkl4rfp//CX/qF62PWuWcRSvcNG7J9o7JDP/++6/18bPPPpvu8z/77DM++OADm30VK1bEx8eH+/fvM3DgQN5++21ef/11mzmAevfuna7rZKTOgIAAgoKCgLSHVd26dUuy788//2T37t2pTvMTERHBhg0bePz4MXv27GHPnj3MmjWLSZMm0bFjxzRdPyMcCo80Gg3bt2+3jhdMi6JFi9K5c+cks6wLIYQQIuvdvXuXXr16ce/ePZv9Fy9eZOfOnaxcuTLVsflCCCHE08JkMtH/h/6cDT6bpdcNjAyk0ZZGTqvv+SLPs7b12iwLkBYuXAhAkSJF0h1w3Llzh59//jlJeKTVavnss89466230Ov1zJw5k2+++YaJEyemu8eRI3UeP37c+jjxPEfJ8fT0tNmOjo7mvffew9/fP9VOOVFRUUyZMoWQkBBOnjzJkSNHePDgARMnTuTWrVuMGjUqTW1IL4eGra1evTpdwRGYxzLaW15PCCGEEFlv9uzZNGvWjF27dnH69Gn+/PNPzp49y+7du+nVqxdTp07N7iYKIYQQOUpW9tjJzYxGIxcvXmTkyJH89NNPlC5dmmXLluHu7p7sOUeOHKF79+7Wr1atWtG8eXMCAwPtlm/WrBkrVqywTptz5coVBg8ezKBBg7h27VqG2p3eOi29qvLly4dGk7H+OZaMpHLlyqmW9fHxoWPHjgwcOJAlS5awd+9ea7C1cOFCjh07lqE2pMahnkd169bl5s2bHDx4kC5dulhnA79+/Tr/+9//aNmyJWXLlnVGO4UQQgiRQW+99RZTp061u2zrjRs32Llzp80+V1dXKlSowKRJk2RxCyGEECIBRVFY23qtU4etnQ48zYj/JV2GPrElzZdQ27e2U66ZmcPW/vzzT7p27cp///3HgwcPyJcvH1988QXNmjVLMkF1Yk2aNGH27Nk2+06cOMEbb7yR7DmNGjVi7969LFmyhHXr1hEbG8vRo0fp1KkTEyZMYMCAAem+h/TUGRcXB5CmaXzsOXDgAPfu3eOTTz7J0PklS5bkyy+/ZOTIkQQEBLBq1SoaNXJeLzULh8Kjq1ev0rNnTyIjI9HpdPTp0weA8uXL4+npyciRI+nSpQu9evVySmOFEEIIkX5dunRhwIABTJw4kYYNG9ocq1SpEqNGjaJr166ULFkSNzc39Ho9//33H99//7112V0hhBBCmCmKgrs2+d4z6dWoeCN83X0JigyyO++RgoKvuy+Nijdy2pxHmalw4cJs376dv//+my5duhAWFsbt27dTDY6SU69evWRXTbPw9PRk4sSJ9OrVizlz5vDDDz8QGxvLjBkzyJ8/f4YW+0hrnQUKFADg8ePH6b5GcHAw8+fPZ82aNek+NyGtVsvUqVN55ZVXOHfunEN1JcehYWvz588nJiaG6tWr06RJE5tjvr6+TJ06lU8//ZR9+/Y51EghhBBCZFzTpk1ZsWIFa9euxd/f32bRigkTJnDz5k2GDx9O+/btadGiBW3btmXw4MEcP36cjz/+OPsaLoQQQjwF1Co179Z7FzAHRQlZtifVm5QrgqOEnnnmGT777DMA5syZw2+//ZbhutK6gEepUqWYP38+q1evxtvbG4AvvvgCkylpKJdWqdVZvXp1AGJiYpIdXpecbdu28ddff9GwYUMqVapk83Xnzh0A6/bt27dTrKtIkSK89NJL1p5QzuZQePTnn3+yb98+tm7dSpkyZZIcr1y5MoUKFWLp0qWOXEYIIYQQDvLx8WHp0qWUKFGCvn37cv36dQAKFSrErl27mDFjBu3bt6dRo0a0bt2a999/nz179lCqVKlsbrl5FZqT906y9++9nLx3EoPRkN1Nckheux8hhBCOa1GmBXObzqWIexGb/b7uvsxtOpcWZVpkU8sc065dO3r06IHBYGDcuHFJFuhIq3HjxiXZFxwczPLly+2Wb9SoEfPmzQPgv//+IyQkJE3XyUidLVu2tK7Gdvjw4TRdx8LLy4ty5crZ/bLMn2TZTkvPrdKlS1O+fPl0tSGtHBq2VqBAAYoXL57scaPRiF6vz/BEVUIIIYRwrn79+tGgQQPee+89OnbsSJ8+fdBoNHTq1IlOnTpld/OS2P/PfvxP+BMY+eSTPF93X96t926ufCOd1+5HCCGE87Qo04KXS73MmaAzBEcGU9i9MC8UeSHX9ThKbMqUKZw/f54///yT0aNHs2HDBnQ6XYbqiomJ4bfffsPPzw+AgIAAhg4dardso0aN8Pb2Jjw8PMVJuhNLb53Fixene/fubN68mbVr19KpU6dUg57t27dTr149+vbtS9++fe2WadasGXfu3OHHH39Mc9uDg4MzbXEyh3oeeXh48ODBg2SPb9u2jdDQUHx8fBy5jBBCCCGcqEKFCmzYsIF///2XN998M82fxmW1/f/sZ9yhcTZBC0BQZBDjDo1j/z/7s6llGZPX7kcIIYTzqVVq6hatS9tn2lK3aN1cFRzFxMQAoNfrbfa7uLiwYMECPD09OXfuHJMmTUoyjMxoNAKkOrxs3rx51oW6AE6dOpVsb5+goCAeP35M8+bNk0xmHRVlnvA84VB+R+p85513qF69OlevXsXf3996P/YcOHCA2NhYSpcunfyNZsC9e/cIDg6mR48eTq3XwqHwqGfPngwZMoRLly7Z7A8MDMTf359PP/0URVEyNDmVI3bs2EG3bt3o2LEjTZs25cMPP0z2jfHly5cZMmQInTp1okOHDixfvhyDQbqPCyGEyNt0Oh3vvfceffv2ZdCgQenuZp3ZDEYD/if87U4catk388TMTB3yZTKZMJqMxBnj0Bv0RMdFExkbSURsBKH6UB7HPOZh9EMeRD0gODKYwIhA7kXc4074HW6F3eLf0H+5+fgmfz/6m79C/mLqb1OTvR8TJmYcn8Gj6EfEGGIwmpJ/0ykyjwwpFEKIjDGZTNY5je7fv8+tW7dsjpcpU8Y6/9HevXsZPnw4//zzj/W4pVNKcHCw3fr1ej2LFy/m4MGDPP/88zbH3nrrLVatWkVYWJh1361btxg9ejRlypThgw8+sCkfHBzMjRs3ADh9+rTdwCq9dXp4eLBy5Ur8/PzYsGEDQ4YM4dSpUzZlgoKCWLx4Mffu3XNoUbFRo0bh5+fH7t27rW2/fv06X375JXPmzLEOd3M2h2pt27Ytv//+O6+99hq+vr4UK1aMoKAgAgMDMRqNmEwmGjVqxLBhw5zV3lR99tlnHD9+nOXLl1O8eHHu3LnDkCFDrN3IChcubC176tQphg4dyscff0yHDh0IDw+nT58+XL58mblz52ZZm4UQQojs0qRJE6pUqcKHH35IQEAAkyZNynBXcmc6E3QmSQ+dhEyYuBd5j47fdsRd447RZMRgMpgDH8zvQQwmA0aT0brPaDSav5uefFmPJdpnMBnsBj2ZKSgqiBe/ftG6rVFp0Kq06NQ6dCodOrUOrUqLVq1Fp9JZjyXetpZLsG05rlVrbbbTVT7+Opm1tHN2kyGFQgiRMYsWLeLrr78mKCgIMC9d3759e8qWLcuWLVtwdXUFzPnBqVOn2LhxIwcPHuTgwYOULFmSdu3a8e233wJw/PhxWrdubTPMTK/Xc+fOHSIjIxk7dqzNtV977TVatGjB8ePHGTRoEAaDgcjISBRFoUWLFgwdOpR8+fLZlP/777+tPY++//57jh8/Trdu3RgzZkyG6rTw9vZm+fLlBAQE8N133zFp0iSioqIoW7YsXl5eVKpUia5duzo8n2Tbtm25fv06kydPZtmyZdSvX5+aNWvy0UcfoVI51D8oRYrJkWnH433//fesWrWKy5cvW5OvsmXL0qNHD/r372+dPCqz/fTTT4wePZr169dTr1496/5Lly7RuXNnXnrpJVasWAFAREQErVq1omrVqixbtsxa9pdffmHw4MFMnTqVbt26pfna4eHh1K5dm9OnT+Pp6em8m8rlDAYDZ8+epVatWln2cyDSRl6bnE1en5zLWa9NdvzeOH36NH/++ScqlYqKFStSp04dm+Nbtmxhx44dTJs2jYoVKzrlmhm9z71/72XSL5Oc0oasoFJUqFChKIr5seUrfp/BaCAiLiK7m+kUGpXGJkxKHGpZwiaNWmOzrVVrbcKw1M5Prnzi/ZbzHQm1LEMKEweGllWOcvNktULkNtn9d1V0dDQ3btygXLly1tBDiLwsrT/zTunP1K5dO9q1a0dUVBShoaF4enri4eHhjKrTZfny5Wi12iRvhqtUqULNmjU5fPgw58+fp0aNGqxevZrg4GDatGljU7Zx48Z4eHiwYMECOnfunGldvoQQQoisEhgYyJgxYzh37hxg7lquKAoNGjRg4cKF1jfnPXv2pF69ekyePJnWrVszYMCAbGtzYffCqRcCxrwwhucKPvcktEFlE94oioJaUZuP2zmmQvXkuJLoOPb3qVVqm6BIQUk1uDh57ySDfhqU6v0sbbGUGoVroDfoiTXGEmuIRW/UW7f1Bj16o55YQ6ztdvxjS3nrMUs9CbYtx2MNsXa3Y40JrmmIJc5ku+RvnDGOOGMcZM5KwBlmCbUsYZK973ZDKZWWH27+kOIQyY+OfUR4bDguahe0Kq21V1jCx9Z9ajv74h9bftaEEEKI3MapyYibm1uSiajmzp1Ljx49KFGihDMvlURkZCQXL16kSJEidrtq1a1bl3PnzvG///2PGjVq8MMPPwBQtWpVm3IqlYpq1apx/Phxjh49ap3FXQghhMitPvzwQyIiIhgxYgRFixYF4Pbt2+zdu5e5c+fy4YcfWss+88wzbNiwgfnz5zN48GD8/f2zZeGLF4q8gK+7L0GRQXb/qFdQ8HX3ZWDVgbliMtG03k+DYg1y3P0YTcakYVQaQ63E5R0JtRKHZMmFWpFxkU5/DkL1oXxw9IPUC6ZCQXkSKqm1aBSNNcxKLnCyeZyecxKUTa7OJNt2gi+NSoNKybxhECJvMBgNeW6FMCGELaeER6GhoURGRiaZUdxgMFCiRAk++ugjVq5c6YxLJSssLAyTyUR4eLjd48WLFwfME0ndvXuXa9eu2exPqFSpUhw/fpwzZ85IeCSEECLXu3nzJt9//32SZWPfeOMN+vXrl6S8RqNhwoQJHD9+nKFDh7Jz586saqqVWqXm3XrvMu7QOBQUm8DFMpRoUr1JueaPk9x8PypFhYvaBRe1S3Y3xYazQq0L9y/w0z8/pXq9it4V8Xb1Js4YZw2+LI/jjHHWayTclzjgMmEyX9+oz3E9t1KiVtTJhlmphVj2wqxMOydBOY2ikV5eWUTmCxPi6eBQeHT79m1GjRrFX3/9lWK5rBgrWqBAAbRaLREREdbxeglZgq2IiAhu374NgLu7u93hdZbJrxLO/p6YXq+3WYLQEloZDAZZrS0By3Mhz0nOI69NziavT87lrNcmK19bjUbD9evXqVy5ss3+s2fP2kxImVj9+vVZs2ZNuq6V3O/HjGhRpgVzm861+0fJpHqTct0fJXntfrKbs0Ktk/dOpik8erf+u9QtWjdddVtW60sSOBliiTWZwyt7YZTN4/heVmkpa3NO4n3pOCdx7ziDKf79bS77lZRi4JRMMJXaOan2DkvYiysDwVdu6+WV3HxhQZFBjDs0TuYLEyIPcSg8mjVrFpcvX+aZZ56hYMGCXL161WaCzdjYWO7cuZMlcybodDqaNGnCwYMH2bp1K5Mm2U6yGRhofpPm5eVFSEgIQLJvmC1D7xIuy5fYsmXLWLRoUZL958+fT/GN+NPqjz/+yO4miGTIa5OzyeuTczn62kRGOn9oTXJ69+5N586dKVu2LIUKFcJgMBAYGMjdu3eZM2dOiud6eXml61rJ/X7MqBZlWvByqZfzzHCIvHY/eUFahxS+UOSFdNetUlTWFexyE4PRkHzglCDMshtCJRd82TknufpjTbHEGdIRftkZxghYy0YRlQ3PYsaoFFXqIVYawiyHwrI0DHnUqDSoUDHjxIxk5wtTUJh5YiYvl3pZ/o8TIg9wKDw6c+YM69ats65sNmbMGKZOnWqzbN348ePp2LGjY61Mo8mTJ3Pu3DnWr1/Ps88+S6dOnTAYDOzdu9e69F/FihWJiYkBSNJ938LyaXByxwGGDRvGwIEDrdvh4eH4+flRo0YNWW0tAYPBwB9//EH16tVlxagcRl6bnE1en5zLWa+NIz1y0qtPnz4oisLKlSs5deoUAGXKlGHmzJm0bdvWqddK7vejI9Qqdbp7fORkee1+crvcPKQws6hVatQqNa7knpWmTCaTTbiUYrCVhuDLXgCWrnNSOp7gsdFkO+2H0WQkxhBDjCEmm55J5zFh4l7kPc4EnZH/84TIAxwKj3x8fKzBEUDbtm35+uuvGTx4sHVf165dWbBgAZ999pkjl0qT0qVLs2PHDhYvXszy5cvZuXMn5cuX56WXXrK+SW/atCnBwcEANt3qE7J8GlygQIFkr6XT6dDpkn6KpFar5Q89O+R5ybnktcnZ5PXJuRx9bbL6de3duze9e/fm0aNHKIpC/vz5M+U6yf1+FCInkyGFuZ+iKOZeNerkP/zNiQxGQ4rDElPr+ZVS76wMD3lMwznpERwZnEnPnhAiKzkUHqlUKkJCQihYsCAAzZs3p2PHjrz66qsUK1YMgOjoaA4fPux4S9OoePHiTJs2zWbfd999R2RkJM8//zxVq1a1zmUUGhqK0WhMsjrbo0ePrHUJIYQQudm5c+eoWLGidUi2t7e3w3WePn2a2rVrO1yPEDmJDCkU2UGtUqNGneMmpE+JyWQizhTH8bvHGb5/eKrlC7sXzoJWCSEym0MzsjVu3Jh27drx+uuvs3r1ajQajfWTzdWrV7Ns2TKmTJliHSaWHaKiopg/fz4qlYqJEycC5h5Kvr6+1jmZErNMqN2oUaMsbasQQgjhbKVKlUp1XqP0uHTpEufPn3dafULkJJYhhW2faUvdonUlOBLCDkVR0Kq0NCzWEF93X+vwziTlUCjqXjRD84UJIXIeh8KjYcOGUbJkSY4fP87u3bsBc7f4Z555hpkzZzJ//nxCQkLo3LmzUxqbETNmzODOnTuMHj2aF14w/8elKArt2rUDSPIG2GAw8Ndff1G4cGFq1aqV1c0VQgghnKpgwYLUq1ePCRMmEB0d7VBdAQEBfPHFF/Tt29dJrRNCCJFbWeYLA5IESE/rfGFC5GUODVvz9PTk66+/5tKlS5QsWRIwBzNffvklGzZs4Pbt29SsWTPLJsxOyGQyMWfOHL7++mveeOMNhg+37VI5aNAgtm3bxg8//MCrr75q3X/48GEiIyMZP368zDMihBAiT2jZsiVBQUG0aNGCAQMG0KpVK0qVKpWmcw0GA7/99hubNm0iMDCQ5cuXp7ighBBCiKeHzBcmxNPDofDo6NGjfPvttwwYMMBm4k2dTsegQYMcblxGGI1GTp48ycKFC/nrr7/4/PPP6dChQ5JyPj4+TJ06lfHjxxMQEICfnx93795lxowZtGrVit69e2dD64UQQojM0bdvX0qWLMn777/PnDlzKF26NJUqVaJcuXIUKFAADw8PwDxXYVhYGHfu3OGff/7h4sWLREdH07lzZ+bNmyeTYQshhLAh84UJ8XRwKDwaN24coaGheHt7U6VKFWe1KcOGDx/OH3/8QZEiRWjevDkLFy5MccW0Vq1a4enpyRdffMGCBQsAc4+k7t27J5lEOysYjCZO3AghKCyaIvlcqVeuIGqV/THEQgghRHo1bdqUPXv2sHLlSjZt2sQ///yDotj/PWMymZcrb9CgAWPGjOH555/PyqYKIYTIRSzzhQkh8i6HwqNnn32WS5cupdrL6OjRozRu3NiRS6XJl19+me5zGjdunCVtS82PF+7yye5L3H38ZD6KYvld+ah9FVpXK5aNLRNCCJGX5M+fn/HjxzNq1CgCAgI4ceIEV65c4eHDh8TGxlKoUCF8fX2pXbs2fn5+1mHpQgghhBDi6eVQeDR37lzGjh1LYGAgRYsWtVsmNjaWSZMmceTIEUculaf9eOEuwzecwZRo/73H0QzfcIYv+74gAZIQQgincnFxoWXLlrRs2TK7myKEEEIIIXI4h8Kj7du306BBAyZOnEjDhg3x8fGxOW4wGDh+/DgPHjxwqJF5mcFo4pPdl5IERwAmQAE+2X2JV6oUzXVD2AxGE7/9/YCT/0YR7fWABuUL57p7EEIIIYQQQgghnnYOhUcnTpzgxIkTmEwm/vnnn2TLJTefgoATN0JshqolZgLuPo6m5bwAinu7kc9VQz4XLZ6uGvNjVy35XJ48frLfXM5Vq8qW5z/JMLzjJ2UYnhBCCCGEEEIIkQs5FB69+eabnDx5kpEjR1K0aNEkS9ubTCauXr3KmjVrHLlMnhYUlnxwlND14AiuB0eku36NSnkSKMWHTl6WoClR6OQVX87TRWsTQHm6atLVY0iG4QkhhBBCCCHE0+H8+fMEBASwbds2AgMDUavVSaa1MRgMPHz4kJiYGDp37oy/v382tdZWdHQ0y5cvZ9++fSiKgkqlwsfHh5dffpkCBQpw+/Zthg0blmo9x48f54cffuDjjz9Ociw0NJS6deuiVqspXbo07u7uANy4cYPIyEhKlCiBt7c3AEFBQQQHB1OiRAkOHDjgzFt1mEPhUcOGDWnatCmjRo1KsVxOu+mcpEg+1zSVm9CyIsW93QiPiSMsOo7Q6FjCouMIj44jzPI4wbHwmDhMJogzmngUGcujyFggKsPt9NCp40MobXzApMHLTgDl4aLG/4fLeXIYnhBCCCGEEEIIWzVq1KBGjRqULVuWCRMm4OPjYzcDiI2NZebMmYSHh2dDK5OKjIzk9ddfx2AwsHTpUusiIX/88QdTp07l7NmzjBkzJk11rV27ll9//ZXx48eTL1++JMe9vb1Zv349FStWtO7r168fJ06cYPjw4XTr1g0wd8DZtWsXixYtcsIdOpdD4RHA7NmzUzy+ZcsWvv/+e0cvk2fVK1eQYvldufc42m7gogBF87syvOmz6QpbTCYTEXoDYdGxhEfHERofMlkCJtv9cYTHxMbvt4RQsYRGx6GPMwIQoTcQoTcQGBrj0P1ahuE18v8fPp4ueLho8NCp479rzN9d4reTHEu6X6tWOdQeIYQQQgghhMgpTAYDkadOExccjKZwYdzr1EZJNMInp0o8B3JiWq2Wt99+m1mzZmVRi1K2YsUKzp8/z+7du21Wl61evTpr166lZ8+eaarn1q1bHDx4EKPRyI4dOxgwYECSMv369bMJjpKjKAqdO3fmwoULab6PrOJweGTpcpWYyWTi1q1bLFmyhJYtW1KwYEFHL5UnqVUKH7WvwvANZ1DAJkCyREUfta+S7l46iqLg6WLuIUT+jLcvJs4Q37vJHCol1+PJEk5dDw7nz7thqdYbGBrjcBAFoNOo7AdMOg3uLmo87YZQT4656xKW0WTbHFFZTSYzF0IEBgbi6+ub3c0QQgghRLzQffsInD6DuHv3rPs0RYviO/k9vHLB6qgqVeof7Ht4ePDqq69mQWtSZ+kdVaRIkSTHXF1dGTlyJFevXk21nvXr1+Pq6kpkZCQbN26kf//+Ns+Fi4sLrVu3TlfbOnXqlK7yWcGh8Oi5555LU7lly5bx3nvvOXKpPK11tWJ82fcF2wmmMfc4yu4Jpl00alw81RTydElT+V+vP6DXit9SLfdx+yqU8fEgIiaOyBgD4TFxROrjCI8xxH83749I8NhSJiLGgN5g7hGljzOijzPyMDLWofu0UCkkCZgShlLuOg2eLur477ZlngRST4556NI3X1RWkMnMhRAAbdq0Yf/+/fLhjhBCCJEDhO7bx50xY8FkOx4lLjDQvH/B/FwRIKVFvXr1srsJABiN5r8p58+fb3euokaNGnHjxo0U6wgPD2fHjh3Mnj2bUaNG8e+//xIQEMDLL79sLePi4sKzzz6brrZVr149XeWzgkPhkclkwsXFhUKFCiU5ZjQaCQ8Px8vLiyNHjjhymadC62rFeKVKUU7cCCEoLJoi+VypV65gjgseUpPWYXj9GpZ16N70cUZzkKQ3EBETF/+VMGBKeCz+u/5JuYhEZSL1BgCMJgiLiSMsJi7DbUvMVauy9m5KEj5ZekTFh1OeicpYjyUIpFw0Ge+2KpOZCyEsIiMjefXVV+nWrRs9evSgRIkS2d0kIYQQIlcwmUyYojI+n2yS+gwGAqdOSxIcxV8MFAicNh2Phg2dNoRNcXPL8hEXP/30E97e3tSvX59169axePFiHj16BMD//vc/SpYsyfXr1xkyZAh37twBYMaMGbz22msA3Lt3j1WrVnHs2DH27NnD3Llz2bRpE23atOGzzz4DzBNzb9q0iR9++IGYmBgCAwOpXbs2o0aNokKFCjbtadSoEVeuXGHz5s3cu3ePTz75xKZXtoeHB0OHDk3xnnbu3EmlSpVo3rw5fn5+HDx4kA0bNtiER3mFQ+FRvnz5OHjwIJ6ennaPf/DBB7Rt25aGDRs6cpmnhlql0LB80iAuN8msYXiJ6TQqdBod3vZHTaab0WgiMtZAZIy5p5O9gMkSPFl6Rtkcs4RQlvP1BgxG891HxxqJjtVzP1zvlLZq1Yr98Clhj6eEc0jFl3HTqpj8zQWZzFwIAZgnbvT39+f333+nb9++VKxYkV69etG0adPsbpoQQgiRY5lMJv7p3Yeo33/PwouaeyBdqeu8HjtuL7xAmY0bsixACg8PZ/v27QwePBiA/v3789prr9G2bVsCAwOt5cqXL8+BAwcYN24ce/bsse5fs2YNy5YtIyQkhBIlSrBu3To2bdpEWFgY27Zt48MPP0SlUjF8+HD0ej3Lly/H09OT27dvM3jwYLp06cKSJUto0qSJtc6RI0cSEBDAjRs3OHjwIMePH2fIkCEMHDgQNze3VO/JZDKxYcMGxo4dC0CvXr04ePAgR48e5fr165QvX95Jz17O4FB4NGTIkGSDI4C33nqLVq1asXLlSmrXru3IpUQukpOH4SVHpXoyR1TSEa/pZzKZiIkzWns12QzLiw+YLPtTKxMR/zgmfvLyWIOJx1GxPI5yzlA9a5sxT2Ze6f0f0GlUaFQKGnX894SP1QpqlQqtOn6/ShW/T0GrVsV/jy+jMu9PeK5tvaqk+xLVq1ElrDPpNZ6co0paRqVCba3PfOxpmNNKZJ68NF/YokWLqFOnDn5+fowePZpDhw6xZcsWpk2bRteuXenatavdnsVCCCHEU0/eT6YoJCSE7t27W7ejo6P5+++/iY2NtYZHAJ6enpQuXdomPLIoV66czfaAAQNo0aIFzZs3Jzw8nPv37/Pbb7+xYcMG1Go1Wq2W5cuXExAQwP79+605RcmSJZk1axbdunVj/Pjx1t5PAF5eXmzatInJkydz8OBBIiMjWbBgAZs3b2b8+PGpzjt08OBBoqKieOWVVwB46aWXKFWqFLdu3WLjxo18+OGHGXn6ciyHwqPUunAVKVKEAgUKMGPGDLZv3+7IpUQuYxmG99v1YE5euELdahVz9R9Z6aUoCq5aNa5aNc760yvOYCRCb3gyLC/GkKj3U1x80GQOnyL0tkP27jyM4tbD1LvXxhlNxMUP48uLEoZNCYMqSyCVJCRT2QutEgVgiYOxhAFYMuGZvZDNcg2NWkEN/P1Aj/rOY3QaTZJgzN49aFQKqqfk31h2yGvzhdWpU8f6WKVS0axZM5o1a8a9e/fYtm0bLVu2pGnTpvTq1cumrBBCCPE0UxSFMhs3OHXYWuSpU9waOizVcqWWL8PdSb+TM3PYWsGCBdm6davNvv/++4+BAwcmbUc62lCs2JP3WyNGjECj0VhXNouNjWX16tWULFmSUqVK2ZxXo0YNatWqxdmzZ9mxYwdvvPGGTVuXLl3KTz/9xKxZs7h9+zZBQUFMmjSJb7/9li+++IJ8+fLZbc+6devo0aMHWq3Wei89evRg9uzZfPPNN4wbNy7Fzja5jcOrraXk4sWLBAYG8vDhw8y8jMih1CqFBs8UwjXUjVrPFHpqgqPMolGryO+mIr+bNkPnp3Uy84W9nqdmSW9ijUYMRhOxBst3EwajiTiD0RwwGY3EGUzxj+P3W7ctj+PLGp6UeVKXkVijCYPBRGx8eZvrGePLJL5uknqNNvVb2mMZNpiY5TgYM/Q8ZrkDv6aruErBfo+uZHqPWXto2esZlsy5yfUes9v7zBK22esZZqf3mN1wL0E7sqv32NMyX5jRaOT48eP89NNPREREsGfPHi5duoSLi4u1N5Krq2t2N1MIIYTIVoqioCSz6nhGeDRujKZoUeICA+3Pe6QoaHx98Wjc2GlzHmW14sWL07dvX4fqUMffu6enZ5L3I+fPnyckJISSJUvaPbdu3bqcPXuW06dP24RHFq1ateLll19mw4YNfPnll4SGhnLs2DFGjhzJ2rVrk7z/vHLlCqdOnWL27Nk2+7t06cIXX3xBZGQkO3bs4PXXX3fklnMUh8Kj/v37J3vs8ePHXL9+HYPBQK1atRy5jBDCCdI6mXnb6sXyRNBnMj0JkRIHYJbtuAThVJIyRiOGZAKwxEGVbbBlG7DZu26sIeGx5MoYiYyOQaXR2g3PYg32wzGjCfQGI3m481iqvcdse2glDMBsQ6wUe48lCNnUCqz85Uaemy9s48aN9OnTBwC9Xs+OHTtYtWoVd+7cwWQy0bBhQ4YOHUrDhg158OABW7dupXv37owbN07mRRJCCCGcSFGr8Z38nnlVNUWxDZDiQwvfye/l2uDIIvFQNGe6d+8eAKGhoXaPW3otRaXQY0yn0zFo0CA6duzIhAkTOHbsGMePH+e3335LMo/zunXr8PT0ZMSIEUnq8fDwQK/Xs3HjRvr3759nps1wKDw6ceJEypVrNDRu3Jjp06c7chkhhBNk1WTmOYWimIMErRpctbnvF63BYODs2bPUqlXL+ilLkjKJe4Al7LUV36MrSWiVTM8ua6CWUu8xm8As4bEnZZLrPRZnL5zLI73HLPOFnbgRkqsWPZgzZw6+vr7WVUbu378PwCuvvMKQIUNslogtVKgQw4cP57XXXqNLly58/vnnshiGEEII4UReLVvCgvkETp9BXHwQAqDx9cV38nvm47lcwsmqnc0yPCwwMBCTyZQksNHpdIC5B5TFtGnTmDJlSpK6ChUqxNKlS2nfvj3//PMPFy5csHnfExISwvfff8+2bduSrOAGcOnSJTp37sw///zD4cOH8fPzc8o9ZjeHwiONRsP8+fOpXLlykhdHq9Xi7e1tfZGEENkvN05mLpKnVimoVWpcMnUAcvax9B5LEjAleGwvALPXeyxJGaMJQxp7j10PCufXv0NSbW9QWHSqZXKSyMhI3nrrLcDcDbxTp04MGTKEZ555JtlzfH190Wg0fP755+zcuTOrmiqEEEI8FbxatiRf8+ZEnjpNXHAwmsKFca9TO9f3OLLnp59+olWrVtb5gh4/fmwz5MxgMNh8T03NmjXRarVERUXZ7SlkmUrn5Zdftu67ceMGt27dSjJHEoCLiwvNmzfnq6++In/+/DbHtm7dSsOGDe0GRwBVqlShUaNGHDt2jHXr1kl4BPDCCy/QokULZ7VFCJEFnvbJzEXukbD3GGTfm6Zfrz/g179Tny+sSL7cNxeQSqWiR48eDB06lKJFi6Za/t69e9y7dy/ZLuFCCCGEcIyiVuNRv152NyNDjEZzr3CTvXmbEjhw4AA3b94EwMfHB4Bvv/2WKlWqYDAY2LZtGz/88AMAt2/ftp5nqTcuLi5Jnd7e3nTt2pXNmzezefPmJOHRmTNnqFChgk2QYzKZmDdvHnPnzrXbzqtXr+Lu7m4TOEVGRrJ+/Xrmz5+f4j327duXY8eOceTIES5fvkzlypWTlImONn/wqNfrU6wrp1A5cvK6devs7g8JCSEyMtKRqoUQmcgymfmLpd1oIJOZC5Eiy3xhyf0rUYBi+V2pV65gVjbLYSqVijVr1vDhhx+mKTgCczfuevXqMWTIkExunRBCCCFymwcPHgAQHh5uDUYS279/PxMnTqRjx46AeXl7gLVr19K4cWMaNGjAhQsXaNeuHQArVqxgzJgxREdHc+3aNQCCg4P566+/ktQ9YcIEqlatyk8//cTatWutYdOuXbs4f/48s2bNsvZ0stizZw8jRoyw1g3mMGfhwoX89ttvTJs2jcKFCwPmcMzf35/79+/b7a2UUNmyZa2PP/rooyQfvAUHB/P3338DcOrUqVQDt5zAoZ5HRqORjz76iD179rBw4UIaN25srlSjYenSpQQHBzNlypQ8tTydEEKIp0tenS/M39+funXrpuscrVab7AdHQgghhHg6Xb58mWPHjrFhwwbA3DvnlVdesYYuYO4tFBQUxMOHD2nUqJH1g6t27dpx9epVNm/ejKIoDBgwgJEjR7Jo0SLq1q3L6NGjqVevHrNmzWLjxo2AOYfo2rUrr732Gp988on1Gp6enmzYsIFly5axbt061q1bR+HChalYsSLbtm2jRIkSNu0uUaIEq1at4vLly8yYMYOQkBCMRiNhYWFUrVqVLVu2UK1aNcDcQaZDhw4EBwcD0KZNG5o1a8acOXOSPB+ff/45mzZtsm6fPXsWPz8/xo0bR79+/ejWrRtXr161Tt69d+9eTp06RY8ePRg1apTDr0dmUUwORFxr1qzB398fMD9B7du3tzm+dOlSvvnmGzZv3kzBgrnrE9n0Cg8Pp3bt2pw+fVrCsgTSMumvyB7y2uRs8vrkPD9euJtkvrBiDswXlhN+bzx69IgTJ07QtGlT6xyF169f59y5c7Ro0QIvLy+Hr5ET7lMIIUTukd2/N6Kjo7lx4wblypVLshy8EHlRWn/mHRq2tm3bNrp168by5cut3coSeuONN7h79y7z5s1z5DJCCCFEtmtdrRhHJjVj4xt1GVs/PxvfqMuRSc1y7UTzd+/epX379owZM8Zm8uvy5ctTuXJlXn/9dQICArKxhUIIIYQQIqdwKDwC+Oyzz3jppZeSrLYG5u7tPj4+/Pzzz45eRgghhMh2eWm+sHnz5hEcHEyhQoWoUaOGzbEqVaowZcoURowYwfHjx7OphUIIIYQQIqdwKDxKrRvff//9x71794iJiXHkMkIIIYRwslOnTrF582aOHDlClSpVkhyvU6cOXl5eLFy4MBtaJ4QQQgghchKHwqMaNWqwbds2u8cePHjA22+/jclk4vnnn3fkMkIIIYRwMm9v71R/PyuKwh9//JFFLRJCCCGEEDmVQ6utjRw5ks6dO3PgwAGaNWtGsWLFCA4O5vfff2f37t1ERUWh0+kYO3ask5orhBBCCGdwcXEhMjISd3d3u8f3799PSEgIPj4+WdwyIYQQQgiR0zgUHvn4+LBmzRreeustPvjgA+u8R5YF3AoVKsTMmTOTzKUghBBCiOzVoUMHRo8ezfTp0ylSpIh1f0xMDFu3bmXu3LkoikKbNm2ysZVCCCGEECIncCg8AvOqLN9//z0HDx7kzJkzhIaG4unpSbVq1WjevLksbyiEEELkQD179uTYsWO0bNmSKlWqUKxYMYKCgrhy5QqhoaGYTCYqVarEmDFjsrupQgghhBAimzkcHgGoVCqaN29O8+bNnVGdEEIIITKZoigsWLCAFStWsGbNGs6cOWM95u7uzmuvvcbYsWPx9PTMxlYKIYQQQoicwOHw6ObNmxw8eJAuXbrg5eUFwPXr1/nf//5Hy5YtKVu2rKOXEEIIIUQmUKlUDBs2jKFDh/L333/z+PFj8uXLR7ly5dBonPL5khBCCCGEyAMcemd49epVevbsSWRkJDqdjj59+gDmoWyenp6MHDmSLl260KtXL6c0VgghhBDOpygK5cuXT7J/8eLF9OnTB29v76xvlBBCCCGEyDFUjpw8f/58YmJiqF69Ok2aNLE55uvry9SpU/n000/Zt2+fQ40UQgghRNYymUxUrlyZTz/9NLubIoQQQgghsplDPY/+/PNP9u3bR/Hixe0er1y5MoUKFWLp0qW0bNnSkUsJIYQQwokePnzI+++/z/Hjx4mMjLSulJqYLHwhhBBCCCEcCo8KFCiQbHAEYDQa0ev1XLt2zZHLCCGEEMLJZs2axf/+9z/c3Nzw9fXl4cOHFCxY0Ho8Li6OqKgoOnbsmI2tFEIIIYQQOYFD4ZGHhwcPHjygUKFCdo9v27aN0NDQFAMmIYQQQmS9o0ePMmPGDDp16oSiKIwYMYJ58+bh4uICmD8AGjt2LGPHjs3ehgohhBBCiGzn0JxHPXv2ZMiQIVy6dMlmf2BgIP7+/nz66acoikLnzp0daqQQQgghnCt//vx07twZRVEAaNGiBbt27bIeV6lUvPbaayxevDibWiiEEEIIIXIKh3oetW3blt9//53XXnsNX19fihUrRlBQEIGBgRiNRkwmE40aNWLYsGHOaq8QQgghnECn0xEdHW2d06ht27Z069aNNm3a4OXlBYCbmxs///wzkyZNys6mCiGEEEKIbOZQeAQwZcoUatasyapVqzh37px1ws2yZcvSo0cP+vfvj1qtdrihQgghhHCe6tWr07VrV2rUqEGTJk1o27YtrVu3pn///owYMQK9Xs/ChQt5/PhxdjdVCCGEEEJkM4fDI4B27drRrl07oqKiCA0NxdPTEw8PD2dULYQQQohMMGrUKHr06MHOnTs5d+4cbdu2ZciQIfz888+MGTMGAJPJJEPPhRBCCJFuV65cYevWrZw+fRqj0YhOp0NRFEqVKkWrVq2oWrUqGzduZOLEiYSGhlK3bl3UajWlS5fG3d0dgBs3bhAZGUmJEiXw9vYGICgoiODgYEqUKMGBAwes13v06BFLliwhICDAOn9j6dKl8fPzIyoqivz581sXAfnqq684ceIEBw8etJ7v7u6OTqcDoHjx4jRs2JBBgwbh4+Njc193797lu+++Y+/evVy+fNl6bsmSJdHpdERERBATE0PFihVp164drVu3RqvVZs6TnMWcEh5ZuLm54ebmlmR/37592bBhgzMvJYQQQggH+Pj48N1333H06FEqVqwImIeyrVu3jgULFnD79m1q1qzJG2+8kc0tFUIIIURuERcXx5w5c1izZg0DBgzgyy+/pGjRotbjFy5c4PPPP+e3336jTZs21v3e3t6sX7/e+p4EoF+/fpw4cYLhw4fTrVs3wPzB1q5du1i0aJG1XHBwML169aJs2bJs3rzZunrsr7/+yqeffsrff//NrFmzrOUHDRrEoEGDaNasGXfu3GHy5Mm8/vrrAFy/fp2PP/6YVatW8d1337FmzRqeffZZ67nFihVj2LBhtG3blhYtWgCwa9cuypQpYy1z8+ZNZs6cyYQJE9iyZQtLliwhf/78Tnl+s5NDE2anxalTpzhz5kxmX0YIIYQQ6XD58mVWrlxJ1apVKVu2rHW/l5cXH3zwAcuWLWPEiBHWT++EEEIIkfmMRhN3/nrIlZP3uPPXQ4xGU3Y3Kc3i4uIYOnQoX331FbNmzWLSpEk2wRFAtWrVWL16NU2bNuXBgwfW/f369bMJjpJjWZCradOm1n2ff/45d+/eZe7cudbgCKBhw4Zs3LgxSRssihUrBmDt6QRQvnx5li1bRvHixQkODua9996ze25KK8qXLVuWxYsX06BBA06dOoW/v3+q95UbZFp4FBgYyMyZMxk0aJB1HiQhhBBC5AzDhg1j6dKlrF69OrubIoQQQgjg+u9BrJt8jF3zfufnVZfYNe931k0+xvXfg7K7aWmycOFCjh49Svfu3Wnfvn2y5VQqFf7+/kRHRwPg4uJC69at03WtTp06WR8fOHAANzc3PD09k5QrWLAgAwcOTLYd9ri7u9OuXTsAzp8/z+3bt5OUSW1eZ5VKRc+ePQH4/vvviYuLS7F8buD08Oj06dO8/fbbNG/enDVr1qDX6519CSGEEEI4yNvbG41GY31jk5yrV69mUYuEEEKIp9f134P4cdkFIh7F2OyPeBTDj8su5PgA6fLlyyxfvhyVSsWIESNSLV+gQAG6d+8OmMOjhEPD0qJ69erWx0ajkbCwMFauXGm37IsvvpiuusG2Z9GjR4/SfT5gnetIr9cTFhaWoTpyEqfMeaTX6/n+++/ZsGEDf/75J2Aei+jj40PJkiU5d+6cMy4jhBBCCCeZM2cOY8aMSXVY2tChQ20mlBRCCCGediaTiTi90Wn1GY0mfvn6Soplfvn6KiUrF0SlUpxyTY1OhaI4py6ALVu2YDQaqVWrlnU4WGos8xg5qlGjRvz888/MmTOHGzdu8N577+Hl5WU9Xr58ecqXL5+uOv/9918ANBqNzfD+9Pjuu+8AKFmyJAUKFMhQHTmJQ+FRYGAgmzZtYtu2bTx8+NA6PK1y5cq88cYbtGnTBo1GYzMeUQghhBDZ78KFC/Tq1YsxY8bQpk0bChcubHM8Li6OY8eOce/evWxqoRBCCJHzmEwmdn5+hnt/P87S60Y8imHl24edVl+x8vnpPOEFpwVI+/fvB8xZQFZ79913+f3337l//z47d+7k0KFD1lVlNZr0Rx53795lx44dAPTq1cvucLiEjEbbIPHff/9l1apV/PTTT2g0GqZMmZLuNuREGQqPTp06xfr16/nf//6HwWDAZDKh0Who1qwZf//9N7t27bIpP2/ePGe0VQghhBBOsm7dOv78809MJpO113BiJpPJqZ9KCiGEEHmB/Gq0FR0dTXBwMGAeFp+cU6dO8f777xMZGWndpygKL774IlOnTs3w9UuWLMmWLVt45513+P333wkJCeHTTz9l3bp1vPvuu7z88stpqicyMpJDhw4xa9YsHj9+TLt27XjnnXdSPa9Hjx6ULFmS2NhY7ty5Q0REBABVqlTh/fffp3bt2hm+t5wkzeGRXq9n9+7dbNiwgcuXL1t7Gfn6+tKtWze6d+9OkSJF7M6d8PzzzzuvxUIIIYRw2Jtvvsno0aPp3LkzRYsWTTLxo8lk4urVq/z888/Z1EIhhBAi51EUhc4TXnDqsLX/rj7i+0WpT/XSblRNilfwdso1nTlsLTQ01Po4PDw82XJ16tRh7969fPTRR2zduhWAxYsXW5e8d0SpUqXYtGkT27dvZ/78+Tx48ICbN2/y5ptv0r59e6ZPn45Op7N77pdffsmmTZu4evUqsbGxNGnShKVLl6a5F9W2bdsoU6YMALGxsZw8eZLly5fz22+/MXfuXN577z2qVavm8D1mtzSHR3369OGPP/4AzDOHN2rUiJ49e9KsWbNUZxoXQgghRM7SsmVLGjRowIwZM1Is16xZsyxqkRBCCJE7KIqC1sV5fwOXqlIQD2+XJJNlJ+RZwIVSVZw355EzeXt7o1arMRgMqQ53V6lUtGzZ0hoeZWQy65Tq7t69O23btmXFihWsXr2amJgYdu/ejVarTfY9T9euXRkxYgRLly5l3rx5nDt3Djc3twy1QavV0qhRIxo1asTbb7/N3r176dOnD7t27aJcuXKO3F62S/Nqa1u3bmXRokXUrVvXOht6pUqVJDgSQgghcqnUgqN9+/axfv36LGqNEEII8XRSqRRe7FEhxTJNulfIkcERgE6no0aNGoB59fXUlqVP2AMotYU7MsLT05O3336b77//nipVqgCwc+dO/vnnnxTPGzZsGC+++CJhYWGMGjWKqKgoh9phGfIWHR3Nhg0bHKorJ0hzeKQoCi1atGD9+vVs3ryZsLAwOnTowNChQ/nll18ys41CCCGEyAQprYYSERHBnDlzyJ8/fxa2SAghhHg6lX++CK2HVcPD2zZM8SzgQuth1Sj/fJFsalnadOnSBYCHDx9y9OjRLL32tGnT7O4vXbo0q1atsq68dvHixRTrURSFWbNm4evry5UrV3j//fcdalfx4sUpVKgQADdv3nSorpwgQxNmV65cmenTpzNhwgQ2b97M5MmTcXd3p3fv3sTGxiYpf+vWLUqVKuVwY4UQQgjhHM2bN0/2mNFo5NGjR0RHR7N27VpGjhyZhS0TQgghnk7lny9CuZqFuXv1ERGhMXh4uVCsgneO7XGUUKdOndi8eTMXL15k9uzZNGrUCK1WmyXXPn36NOHh4XZXRStYsCANGzbkp59+StMHYgULFmTevHn079+f77//nurVqzNgwIAMtUuv11snzy5evHiG6shJ0tzzyJ6CBQsycuRIDhw4wIgRI9i9ezd//fUXc+fO5cGDB9Zyw4cPd7ihQgghhHCeO3fuJPt19+5doqKiMJlMrF27NrubKoQQQjw1VCqFEpUKULFuUUpUKpArgiMwz/XzxRdfULJkSa5cucLHH3+MwWBwqM7o6GjAHMKkJDIykqVLl9o9ZjQauX79Or6+vtSpU8fmWExMjN36a9euzdixYwGYNWsWBw4cSFKvvU4ziW3bts16D127dk21fE7nUHhkodVq6dixI9u3b2ft2rX8888/vPLKK0yZMoWlS5dy/fp1Z1xGCCGEEE7i5ubGjz/+yOXLl5N8Xbx4kU6dOvHdd99x4sSJ7G6qEEIIIXKBkiVLsmnTJl588UW2b9/OgAEDOHnypE2ZW7dusXv3bgDKli2bbF3BwcH8/fffAJw6dcq62ntyVqxYwfvvv8+dO3es+8LDw/n444/577//mDNnjs38SoGBgdac4vz58xiNtqvnDR48mKZNm2IwGBg9ejSrV6+2mQPp7t271seJ50aKjIzkq6++YsaMGSiKwoQJE6hZs2aK7c8NMjRsLSW1a9emdu3a/Pfff6xatYpFixY5+xJCCCGEcFDnzp2TfdOmVqt588036dmzJxs3bqR8+fJZ2zghhBBC5Eq+vr6sXLmSkydP8tNPP/HZZ58RHR2Nm5sbERERuLq6Ur16dZYvX57sSmvdunXj6tWr1lBm7969nDp1ih49ejBq1Kgk5atVq8b777/P2bNneffdd4mIiCA2NpbIyEjq1KnDN998Y/Oep1OnTty8edNa/9GjR3nppZfo2LGjdZJrRVGYOXMmnTt35r///sPf3585c+bQvn17KlasyK5du6z1dezYEV9fX7y9vTEYDNy/fx9vb29atWpFv379qFWrlnOe3Gzm9PDIonjx4nzwwQe0atUqw2MEhRBCCJE5PvzwwxSPly1bFnd3dz755BPWrVuXRa0SQgghRF5Qt25d6tatm6Fzt23blq7ys2fPBqBJkyZ2w6XEEgY/KfH29ubgwYN2jw0cODDN7csrMi08sqhXrx6dO3fO7MsIIYQQwonu3LlDUFAQISEh2d0UIYQQQgiRzTI9PILkl87LLFu2bOHrr79Gr9cTFxdH8eLFGTFihN3k8/Lly8yZM4fg4GCMRiPt2rXjjTfeQK1WZ2mbhRBCiKz03nvvJXvs8ePHnDp1iri4OKpUqZKFrRJCCCGEEDlRloRHWWnatGns27eP5cuXU6lSJQA2b97M66+/zqJFi2jWrJm17KlTpxg6dCgff/wxHTp0IDw8nD59+nD58mXmzp2bXbcghBBCZLpvvvkm1TLlypVj+vTpWdAaIYQQQgiRk+Wp8OjGjRusX7+eqVOnWoMjgF69erFjxw5mzZplDY8iIiIYO3YsdevWpUOHDgB4enoyYcIEBg8eTMOGDenWrVu23IcQQgiR2RRFYdKkSTz33HMoiu0ywFqtFh8fH0qVKpVNrRNCCCGEEDlJngqPLl26hMlkwtvbO8mxUqVKcejQIev26tWrCQ4Opk2bNjblGjdujIeHBwsWLKBz585oNHnqKRJCCCEAeOaZZ2RBCyGEEEIIkSaq7G6AMxUpUgSA7du3Jzl248YNm6UAf/jhBwCqVq1qU06lUlGtWjWCg4M5evRoJrZWCCGEyD7JrTQSExOTtQ0RQgghhBA5nsPdam7dusWSJUu4d+8eq1evBsw9gL777jtatWrF888/73Aj06p27do8++yzHDx4kIULF/LWW28BsH//fh4/fmydHPTu3btcu3YNgOLFiyepp1SpUhw/fpwzZ87g5+dn91p6vR69Xm/dDg8PB8BgMGAwGJx6X7mZ5bmQ5yTnkdcmZ5PXJ+dy1muT3a+tVqtl2bJlfP/990ydOpWaNWsCEBISwqpVq/D29mb48OHpXkAiud+PQgghhBAi93IoPLp58yY9evTg8ePHFC1a1Lq/SpUqlChRguHDh/PCCy8wYcIEhxuaFiqVii+++IL+/fuzaNEi7t27R4cOHfj222/Ztm0bPj4+ANy+fRsAd3d3PDw8ktSTL18+AP75559kr7Vs2TIWLVqUZP/58+dxd3d3xu3kKX/88Ud2N0EkQ16bnE1en5zL0dcmMjLSSS3JmF27djFv3jwUReHKlSvW8KhYsWK8//77TJ06lX79+rFq1Src3NzSXG9yvx+FEEIIIUTu5VB4NHfuXGJiYmjRogXXr1+3OZY/f36mTZtGmzZtKFKkCP3793eooWlVvnx5Nm3axNChQ9m+fTvfffcdixcvtgZHYP5UFUg25LG8SQ4LC0v2OsOGDWPgwIHW7fDwcPz8/KhRowaenp7OuJU8wWAw8Mcff1C9evV0f3otMpe8NjmbvD45l7Nem+zukbN27VqaNGlC8+bN6dixY5Ljb7/9NvXr12fhwoVMnDgxzfUm9/tRCCGEEELkXg6FR8ePH2f58uXUq1ePHj16JDlerlw5PD092bhxY5aFRwCPHj2ib9++GI1GZs6cyZtvvsmUKVPo06cP8GQ+B61Wa/d8y1CC5I4D6HQ6dDpdkv1qtVr+0LNDnpecS16bnE1en5zL0dcmu1/XyMhIvvnmm2SPe3h4UKRIEXbv3p2u8Ci5349CCCGEECL3cig8KlKkCPXq1QNIsswvQGhoKJGRkVk6+eaRI0dYtmwZ69atQ1EUSpQowdtvv82nn36Kh4cHnTp1wsvLC8BmToaELEMJChQokGXtFkIIIbKSZYh2ch4+fEhQUFC2h1xCCCGEECL7ObTammV1s+QsW7YMo9Fod1LqzBAYGMhbb71Fz549rWFWixYtmD9/PiqVCn9/f8LDwylXrhxgDreMRmOSeh49egTYn0xbCCGEyAueeeYZAgIC7B7T6/W89957xMXF8dxzz2Vxy4QQQgghRE7jUHjUtGlTtm3blmS/Xq9n/vz5rFq1CkVR6NSpkyOXSbPt27cTGRlJpUqVbPY3b96cnj178vDhQ86ePUvp0qXx9fUlNjaWO3fuJKnHMqF2o0aNsqTdQgghRFYbOXIkEydOZOrUqfz666/cuHGDEydOsHz5clq0aEFAQAAqlYqRI0dmd1OFEEIIIUQ2c2jYWp8+fRg9ejQHDx7k3r17zJkzh9u3b3Ps2DFCQ0MBqF+/PoMHD3ZKY1Pz+PFjACIiIpIc8/PzY9OmTYSHh6MoCu3atWPVqlWcP3+eUqVKWcsZDAb++usvChcuTK1atbKk3UIIIURWK1OmDIsWLWLMmDFs3LjR5pjJZMLFxYUPP/yQF198MZtaKIQQQgghcgqHwiOVSsXChQvZtGkTt27dYuXKlZhMJgBKlixJ9+7dGTRoEBqNQ5dJs3r16rF27VqOHDliXXLY4vbt2+h0OurUqQPAoEGD2LZtGz/88AOvvvqqtdzhw4eJjIxk/PjxMs+DEEKIPK1u3brs27ePnTt3cubMGUJDQ/H09KRatWp07NgRX1/f7G6iEEIIIYTIARxOdRRFoU+fPvTp04fo6GgeP36Mh4dHtixX36JFC1q0aMHKlSupWrUqTZs2BeDcuXMsXryYSZMm4ePjA4CPjw9Tp05l/PjxBAQE4Ofnx927d5kxYwatWrWid+/eWd5+IYQQIqt5enrSv3//LF0VVQghhBBC5C4Oh0d79+6lbdu2ALi6uuLq6pqkTGhoKCdOnKBAgQLUrl3b0UumaP78+axevZrPP/+czz77DHd3dwoWLIi/vz9+fn42ZVu1aoWnpydffPEFCxYsAMw9krp3745K5dB0UEIIIUSO9+jRI06cOEHTpk3R6XQAXL9+nXPnztGiRQvr6qRCCCGEEMlZsmQJa9eutS485eLiwoEDB6wdN1Ly33//8corrxAXFwdA0aJFURSFu3fvAubOKosWLaJFixZ2z//ss8/44YcfePDgAWBeMf3VV1/lpZdeYsaMGdy4ccNaVqVSUa5cOfz9/alRo4Z1/65du9i4cSNhYWG4u7uj0+lo0qQJ1atXZ9++fUybNg2AP//8k06dOqHVailTpgwuLi4AXLlyhdjYWMqWLYuHhwdGo5G7d+/y6NEj6tWrx/r169P5jOZMDodH69ats4ZH9vz33390794dLy8vmjRpwrfffsv48ePJnz+/o5e2S6vVMnToUIYOHZqm8o0bN6Zx48aZ0hYhhBAip7p79y7du3fn/v37fPTRR/Ts2ROA8uXLExMTw+uvv87YsWOTfPAihBBCCJHQiBEjGDp0KCNGjCAgIICYmBjWr1/P22+/neq5a9assQZH/fv3Z8qUKQAcO3aMcePG8fDhQ9555x22bt1KhQoVkpz/wQcf8N577zFo0CDi4uJYs2aN9QMxPz8/tm3bxvvvvw/AwoULk4RQn3zyCXv27GHx4sXUrVsXgHv37rFgwQIWLlxIw4YNbcqXLl2a9evXU7RoUeu+Zs2acefOHT766CProlsGg4Hly5dz7NixND2HuYHD3WsiIyOZMGECjRs35vnnn+fNN9/k5s2b1uMrVqzgwYMHzJ49m8mTJzN58mS7K7QJIYQQIuvMmzeP4OBgChUqZPPpG0CVKlWYMmUKI0aM4Pjx49nUQiGEEOLpYzQauHXxPH8eDeDWxfMYjYbsblKaaDQa/Pz8rCN4LItVpSQ0NJRt27ZZz0m42nmjRo347LPPAHPmMGLECGvPJnvXrlWrFrVr17YGRxaWOY8BmjRpYnPs119/ZdOmTYwePdoaHIG599OMGTPo0aNHkmuNGDHCJjhKjlqtZvjw4VSsWDHVsrmFQ+FReHg4d+/e5fvvv+fBgwdERUVx6NAhevbsyb179wBz1y4wf5IJ5qFtrVu3Zu/evQ42XQghhBAZderUKTZv3syRI0eoUqVKkuN16tTBy8uLhQsXZkPrhBBCiKfP1ePHWDHyDbZ+Opm9X3zO1k8ns2LkG1w9njt6r7i5udGkSRNcXFwIDQ1l69atKZbfsmUL5cqVo1ixYgDWYWAWXl5eeHl5oSgK//77L2+//TYGg/0wTavVJgmOLPstEi/kdeDAAQCKFClit84xY8bYnO/j42MTcKVFp06d0lU+J3MoPJo9ezaFChVizpw5HDp0iPPnz7Nnzx5eeOEFli1bBphTQrD9QShZsiQnTpxw5NJCCCGEcIC3tzfPP/98imUUReGPP/7IohYJIYQQT6+rx4/x3dzphIfct9kfHnKf7+ZOzzUBUqFChayBydq1a4mNjbVbTq/Xs379egYPHpxifZUrV2bUqFGAeSjbzJkzndZWy0rxy5YtIyIiIsnxQoUK2XzAVrhw4XSvRFu9enXHGpmDOBQeHT58mI0bN/Lqq69StGhRdDod5cuXZ+7cuZw7dw4wj/WzN/n07du3Hbm0EEIIIRzg4uJi/YDHnv379xMSEkK+fPmysFVCCCFEzmcymYiNjnbaV0xkBAdWL0vxmgfWLCMmMsJp17QEJ5lh0KBBqFQq7t27x+7du+2W2b17N1qtllatWqVa36hRo6zzLK9du5Zdu3Y5pZ2W+YwuXLhA165drRlGQmmZt+lp4dCE2R4eHhQqVCjJfpVKZZ30ymg02u0+FhgY6MilhRBCCOGADh06MHr0aKZPn27TXTsmJoatW7cyd+5cFEWhTZs22dhKIYQQImcxmUxs+XAi/135M0uvGx7ygEUDk87Bk1HFK1Wh5yczURTFaXValC1blhYtWrBv3z5WrVpF586dba5jMplYvXo1AwcORK1Wp6nOGTNmcOvWLf744w8++OADnnnmmSRzNqZX8+bNadu2LXv37uXvv/+mZ8+edOzYkbFjx6ZpXqOnjUM9j1xcXDhy5IjNvvDwcD7++GPKlCkDQGxsbJJPLSMjI7l/37Y7nhBCCCGyTs+ePXFzc6Nly5b07t2b8ePH069fP1566SWmT59OVFQUFStWZMyYMdndVCGEECJnyYTAJa+xDEe7du2adW4hi8OHDxMcHEyXLl3SXJ+rqyuLFy/G19cXvV7PyJEjCQoKcrids2fPZsiQIajVaoxGI9988w2tWrVi3rx5REVFOVx/XuJQz6O+ffsyePBgateuTdGiRXnw4AGXLl2icOHCuLq6smrVKu7evUuhQoUICQmhYMGCAOzcudM6gbYQQgghsp6iKCxYsIAVK1awZs0azpw5Yz3m7u7Oa6+9xtixY/H09MzGVgohhBA5i6Io9PxkJnExMU6r8/afF9jp/3Gq5V5792NKPlfNKdfUuLhkSq8ji5o1a1KnTh1OnTrFypUrad68ufXYqlWr6N27N+7u7umq09fXlyVLltCnTx+CgoJ46623WL9+vd2RTmmlVquZMGECr776KtOnT+fEiRNER0ezdOlS9uzZw+LFi6lUqVKG689LHAqPOnXqxL///suyZcuss56XKFGChQsXotPpGDx4MAaDgZ49ezJy5Ehq1aqFTqdjzZo1TJw40Sk3IIQQQoiMUalUDBs2jKFDh/L333/z+PFj8uXLR7ly5ZKsSJKdTEYTMTceYwzTo8qnw6VcfhSVfOorhBAieyiKgtbV1Wn1lan5PJ4FfZJMlp1QvkI+lKn5PCpV2oZ55QSDBw/m1KlTnDlzhtOnT1O7dm0uXLjA2bNnmTdvXobqrFatGv7+/rz99tucPXuWjz76iBkzZjjc1ueee47169ezf/9+Pv/8c27evMmtW7cYMGAAe/bssXaEeZo5/M5w9OjRdOnShXPnzuHt7U29evWsbzh3797N7du3KVeuHL169WLSpEkcPnyYqlWr0r17d4cbL4QQQgjHKYqSbI/gCRMmMHv27Cxu0RNRF+7zaPd1DI/11n3q/Dq825fHrZpPtrVLCCGEcBaVSk2zAUP5bu70ZMu8/PrQXBUcATRt2pRnn32Wa9eusWLFCmrXrs1XX31Fp06d7M6dnFZt2rTh+vXrLFy4kJ07d1K5cmWntblFixa89NJLzJs3j6+++oqQkBDWrVvH2LFjnXaN3MqhOY8sSpQoQdu2bWnUqJHNJ5VarZZy5coBUKBAAZYvX87u3bvZvHkzWq3WGZcWQgghRCa5fv06P/74Y7ZdP+rCfR5s+NMmOAIwPNbzYMOfRF2Q+ROFEELkDRXqN6LDuMl4FrT9YCRfIR86jJtMhfqNsqllGacoCoMGDQLg0KFDBAQEsG/fPus+RyRcgW3WrFmcOHEi3XUsX76c4ODgJPt1Oh2TJk2iffv2AFy8eNGxxuYRTgmPUrJ48WIePXpk3a5QoYJDYxKFEEIIkbliYmLYsGEDffr0sQ5Lz2omo4lHu6+nWObR7r8xGTNvqWEhhBAiK1Wo34ghi1fR/cPptB39Dt0/nM7gRatyZXBk0b59e4oUKYLJZGLMmDG8/PLLlC1b1il1z5gxg+rVqxMXF5eh8AggICAg2WOvvvoqAF5eXhmqO6/J1PDIZDJRuXJlPv3008y8jBBCCCGc4O7du8yePZuXXnqJadOm2Xz4k9VibjxO0uMoMcPjGGJuPM6iFgkhhBCZT6VSU6pqDZ5r7EepqjVy1VC1mJgYYmNjbfbpdDr69+8PQFRUFG+88Ybd84Ak59qrLyFXV1eWLFmCr69vsmWio6Otj+Pi4pIcX7FiBWFhYXbPvXr1KgDt2rVLtv6E10iprXmBQ+HRw4cPGTlyJHXq1KFKlSo899xzNl9VqlRh1KhRHDx40FntFUIIIYSTHT9+nLfeeotXXnmFVatW8fjxYzQaDUWLFs22NhnDUg6O0ltOCCGEEJnHZDJx6tQp/vzzzyRL3Pfs2RNPT09q165NrVq1bI5dv36dkJAQAE6fPm1z7OTJk9y4cYOHDx8me90iRYrw5Zdf4ubmZvf48ePHrY/t9U66efMmvXr14tixYxiNRuv+n3/+mSVLltCvXz9efvnlZK9/9epVa/sz2vspt3BowuxZs2bxv//9Dzc3N3x9fXn48KHNLORxcXFERUXRsWNHhxsqhBBCCOeJiYnh22+/ZcOGDdZP1kwmEz4+PvTt25cePXqQL18+/Pz8sqV9qnxpG+Ku/y8Ct2pGFE2mj8QXQgghhB3Lli1j48aNBAYGAuDn50f9+vVZuHAhAPny5aN79+7UqVPHes7Dhw/p378/N2/etIY2y5YtY8+ePXTu3JlvvvmG27dvA9CsWTMqVKjA1q1b7V6/atWq+Pv7c+3aNeu+gIAApk+fzj///GPdN3ToUJ599lmmT59OjRo1AJg/fz7h4eFs27aNmTNnolarefToEcWKFWP69OnWeZUSS9h+k8k8hH7lypX8+OOPDB06lB49emTouczJHAqPjh49yowZM+jUqROKojBixAjmzZuHi4sLAEajkbFjx8rM5EIIIUQOcefOHTZu3MiOHTsIDQ21vuGpWrUqYWFhfP/99zZzE3744YfZ0k6XcvlR59elOnQt/PBtos4G4dmkBB71i6JycXghWSGEEEKkw7Bhwxg2bFiKZSZNmmSzXaBAAXbv3p1s+VGjRqWrDa1bt7bZ9vPzS/UDsKFDh1ofd+vWLV3XS639eZFDH9Plz5+fzp07oygKYF7WbteuXU8qV6l47bXXWLx4sUONFEIIIYRjfvvtN0aOHEnLli1ZvXo1jx8/xsXFhc6dO7N161Z27NhBgQIFkixq0apVq2xpr6JS8G5fPsUy7s8XRpVPhyFUz+O9N7g74ySPf7qJQYayCSGEEEI4lUMfz+l0OqKjo3F1dQWgbdu2dOvWjTZt2lhnJHdzc+Pnn39OkjQKIYQQImv07NmTc+fOAeahaWXKlKFXr1507tyZ/PnzW8tZPgzKKdyq+VCo73M82n3dpgeSOr8L3u2fwa2aD6Y4I5G/BxF2+DZxwVGEHbxF2C+38ajti+eLJdH62J8DQQghhBBCpJ1D4VH16tXp2rUrNWrUoEmTJrRt25bWrVvTv39/RowYgV6vZ+HChTx+LCuhCCGEENll5cqVbN++nY0bN/LgwQN69+5Nly5d8PT0zO6mpcqtmg+uVQoRc+MxxjA9qnw6XMrlR1GZgy5Fo8KjblHca/sSfekBYQG30d8KI+L4PSJO3MOtmg/5XiqJrlS+bL6TBIz/b+++46oq/ziAf+4CmeLAPTINERVXzkwcmCtBTSX3Sk3LkSN3wzI1R+7UTE3cijlScaZm5uhnieYixUVOlL3ueH5/4D1y4V64wIV7gc/bFy8u5zznnO+5B+9z+J5naIG7Z4DYx4BzaaByMyAfzaZToPHaEBERGZWj5NHHH3+MgIAA7Nq1C5cuXULHjh0xdOhQHDlyBGPGjAGQ8oSza9euFgmWiIiIss7Z2RkDBw7EgAEDcOzYMQQGBmLJkiXw8/ND3759UbVqxt3DrE0ml6FIVbdMyzjUKokiNUsgOSwaMaceIPH6cyRcfoaEy89g/3pRuLSsCPs33KzbwurqXiB4EhD936tlruWA9nMBLz/rxUW8NkRERBnIUfKoZMmS2Lt3L37//Xd4eHgASOnKtmHDBixevBgPHjxAnTp1MGTIEIsES0RERNknk8ng6+sLX19fXL9+HYGBgejWrRvq1auHvn37GkxRq5eQkGBy+ltbJJPJYP96Udi/XhTqR3GIOfkA8ZeeIul2FJJuR0FV1gkuPhXgUNsdMkUeJ5Gu7gW29wcgDJdHP0xZ3nMDkxTWwmtDRESUoRwlj65fv45Dhw6he/fuKF++vLTc1dUVM2bMyHFwRERElDs8PT0xa9YsjB8/Hlu3bsWXX36J2NhYbNu2DV27dpUGzu7Vq5fBZBj5iaqME4oHVIdru8qI/S0ccRceQf0wDs+33oAi+A5c3i4Px4ZlILfL5W5JWk1KN6j945AuOQG8WrbnI+D5bUCuBGTylC+5ApDJXv0skwMyheHPcnma9WnLyFLty1i5NPtPV06W/pgG5WQmjpsqPlum06a0ODJ5bWRA8GTAsxO7sBERUaGVo+TR8OHD8eTJE8TExGD69OmWiomIiIjySPHixTFy5EgMGzYMBw4cQGBgIJYuXYo+ffqgQoUKCA0NtXaIOR6HRulWBG6dq8K1TSXE/vEQsWf+gzYyCZH7biP62D04NS0H52bloHBSmR+TEEBSNBD7JCWu2MdAzMvv0rInQOwjIO4ZjCcm0kiKBo5+bn4M+Um2kluZJbZkGSSsMku6pTpmXIRhV7V0BBAdDvz8IVD8dUChBOQqQKFKSfTJlS9fv/xZvz7168zKplv/8rWNDWJPZBLHCyMq8HKUPHJzc8Pz58/x/vvvZ1guNDQUb7zxRk4ORURERLlIqVTCz88Pfn5+uHjxIlavXo3FixdbfwY2C45DI3dUwbVNJbi0KI+4/z1GzKlwaJ8nIubYPcSeegDHN0vDpVlpKO2iU5I+qZNAMY/SJ4Y0CVk4ugxmJZAqNgXcKgJCBwjty++6lGSV0KX8gSYty+RLKitSLTeyvc7Y9ka2TXtsc85HT9omH7u8Pe+PKTeViFKkev3yZ2Ov05VNm8BKXVb56njGyhqsz0pZIzFY+3OFLIvjhREVCjlKHi1YsABjxoyBvb19huWGDRuGX3/9NSeHIiIiojxSv359rFy5Env37sWkSZOsF4ilxqERAkh4ISV+ZLGP4ax7DCfvJ0i4XwQx4R5QJ5ZB3B8PEffHAzjIf4OLMgh28rDM923vCjiXSnnSLn1P/fVy2dPrwAYzYm09DajydublbIFBUspY0iqniavMkmba9DGkKydMHDdVcuxZKHBhdebn69k55Xrq1CldEXVqQKcBtC+/p34tLUtb1shrrTrluzH6/RY0MlNJq9TJMiOtsaRlactm1IIrN8oaa2WmKpxJMY4XRlRo5Ch5dOXKFfTq1QtjxoxBhw4d4O7ubrBeo9HgzJkzePToUY6CJCIiorzn5+dnvfGOzBqHZhJQuhYQH/Gq61issa5jjwFtcrq9yAA4AnAQQJKqDmK07yFJVx8JupZISG4Je4dQuJS9AvvSashcyqRJDpVK+bJzMu98nEqmPImPfmjinGQp6ys3M/cdsj79WEjI511TdFrgxi+ZX5ueP+VuNxyd9lUiKV1ySWMiUZVZWXUW9ptZWa1hsit1Wa0mTbIs1XbG3lOhBTRaAIm5935ag5QUy6yLYnZbg2XQ4iyz1mDZLptBUozjhREVKjlKHm3YsAHXrl2DEALXrl0zWkYIYf0m70RERJQta9eutc6B754xYxya/4Cl9czfp0Mxoy2EZM6lUcS5FIo4l0ZyjCtizscg4fIzJCW8gaTbb0CV7AyXFhXgUKskZPJs3tPIFSldOLb3R/oubC/32X4O/8CyBlu5NnLFy2MUyd3j5DWd1jBBlTa5lFFizJJl07UGM9YyzFRrMiMtx4x1w5SSYgWMTG480aTTAnFPMtjw5Xhhd8/knxaVRGRSjpJHH374IUaPHo2uXbuiTJkyUCgMK1UhBEJDQ3HkyJEcBUlERESFTOxj88rJlEDRcmlaBJloJaTMuJs9ANiVBkpUAzQRCYj5LRxxfz6G+kEsnm++DmWJInBuUQFO9UtDpsrGDGJefildOIyODTKHXTusidcm9+iTYmb8/8tXdLoMWnBl1EUxKy3HMiurzWC/qVuGZaGs0aSYDtAmpXxlh7mf50Rk03KUPHrnnXfQpEkTzJ49O8NyrVu3zslhiIiIqLBxLm1euf4/A1VaWPzwyhIOKNalGlx9KyH2zH+I/eMhNBGJiPz5X0QfuQvnt8rBuXFZyB2zMEMbkJKE8OzEWYlsEa8NZYVcDsjtC2hSLINkV+pWW/f/BA6My3yf5n6eE5FNy1HyCECmiSMACAwMzOlhiIiIqDCp3MzMMYLeytUwFM52KPrOa3DxqYi4C48Q+1s4tFFJiD50FzG/PoBTozJwbl4eSrcs/AEpV7ALh63itaHCTi4H5HYA7DIvW7oWcHp+wRrLjYhMykaba0OPHz9Gv3790KFDB2nZ1atX8dlnn+HAgQMAgPLly+f0MERERFSY6MehASCNOyPJ+zGC5PYKuDQvjzKfvoliAdWhKuMIkaxF7OlwPPr2Ap5vvwH147g8iYWIyCbY2Oc0EeWuHCWPQkJC0L9/f1y4cAEJCQnSci8vL0yfPh379u3DJ598Ao2mAE4xSkRERLlLPw6Na1nD5a7lrDb9s0whh1O9Uig1pj5KDKoJuypFAZ1A/MUnePzdRTxb/w+S7kTleVxERFZhg5/TRJQ7ctRtbcmSJXB2dkbnzp3x+++/G6yzs7PDF198AR8fH7z++usYNWpUjgIlIiKiQshGx6GRyWRwqF4cDtWLI+leNGJPPkDC1QgkXn+OxOvPYVfZFS4tKqBIjeLpZmgTOoGksCjoYpIhd7GDfZWi2Z/FjYjI2mz0c5qILCtHyaO///4bmzdvhoeHBwICAtKtL126NFxdXbFr1y4mj4iIiCh7bHwcGvtKrrDv5wX103jE/haOuP89RvLdaEQEXoXS3QEuLSrAsV4pyJRyJFx5hsh9t6CNSpa2VxS1g1vnqnCoVdKKZ0FElAM2/jlNuev06dP4/fffsXXrVsTHx0vLXVxcUKRIEajVari7u+ONN95As2bN0LFjRzg5OUnlrl27hi5dukClUqFy5cqwt08ZR/DmzZtQq9V47bXX4OTkBJ1Oh4cPHyIyMhKNGjXCvHnzsGfPHhw4cADXr1+X9mdnZwc3NzckJyfD2dkZlSpVQr169dC5c2dUqVIl796YAiZHyaNKlSrBw8MDQMoTuLSePn2KmJgYJCYm5uQwRERERDZP5e6IYt3egKtvZcT+Ho7Ysw+heZqAF0GhiDpyF0WquSH+4pN022mjkhGx8RpK9K3BBBIREeU7zZs3R/PmzeHg4IDly5fD09MTa9euRYkSJQAA8fHx+O2337BixQocOHAAixYtwuzZs9GixavZUitVqoTAwECUKVNGWta6dWuEh4fj888/R7NmKQOva7VarF69GmfOnEGZMmUwfPhwdOzYEb6+vgCAdevWoUmTJpDL5RBC4MqVK9iyZQu+//57rFy5Ev369cPEiROhVOZ47rBCJ0djHpUsWRI6nc7k+gULFkAIgcqVK+fkMERERET5hsLVDkU7VEHZKY1QtEMVyF3soItONpo4Si1y320InbEZi4iIqLAQOoHEW5GI//sJEm9F5qt6oWzZlLGvHB0dpcSR/ud27dohKCgIXbt2xbNnzzBixAj89ttvUpmRI0caJI5MUSgUGDFihNSIBQDKlSsnvS5fvjzk8pQ0h0wmQ+3atfHNN99g+fLlUCqVWL9+PaZOnZrjcy2McpQ8evfdd7FixYp0y58+fYqJEydi9+7dkMlkeP/993NyGCIiIqJ8R15ECRefCig7qSGcW2Q+86w2KglJYRxsm4iosEq48gyP5p7Hsx8u4/nWG3j2w2U8mnseCVeeWTs0sygUGY9zpVQqMWvWLLz55pvQaDSYNGkSYmJiULJkSallkbm6dOli9nGBlFZM06dPBwDs2bMH+/fvz9LxKIfJIz8/P0RGRsLf3x/37t3DuHHj0LNnT7Rq1Qq//PKLVKZPnz4WCZaIiIgov5Ep5bAr52xW2aj9txF15C4SrkZAG5UEIfLPE2ciIsq+hCvPELHxmsGYeMCrrs35JYGUGYVCgUmTJgEAIiIisGXLFri7u6N06dJZ2k/t2rWzfOwePXrgjTfeAAAsXbo0y9sXdjnu6Dd9+nScOHECW7duxblz5xAVFQVHR0fUqFEDPXv2RKdOnSwRJxEREVG+JXexM6uc+r84qP+Le7Wdswp25Z2hKu/88rsLFEXtjI41SUREeUMIAaE2PXxLlvenE3ix91aGZV7svQW7am4Wm51TppJbrS7x9vZGlSpVEBYWhoMHD2LYsGF5clyZTAY/Pz8sWLAAYWFhuH79Ojw9PfPk2AWBRUaJatmyJVq2bGmJXREREREVOPZVikJR1C7dE+XU5E4quLSqAPV/cUgOj4XmSTx0sWok3niBxBsvDMrpk0n6xJLCzZ4JJSKiPCCEwNOVIUi+G52nx9VFJ+PhF39YbH92lV3h/qG31eqOOnXqICwsDDdv3kRycjLs7Mx7yJJT3t7e0usrV64weZQFOUoejRw50uiYR0RERET0ikwug1vnqojYeM1kmWJdqxnMtqZL1kL9MA7q/2KR/CAW6vBYqJ/EQRenRtLNF0i6mTqhpISqnDPsyrtIiSVFMSaUiIjINukH1NZoNIiMjESpUqXy5LglS76qZyMiIvLkmAVFjpJHx48fx2effYaPP/44zy42ERERUX7kUKskSvStgch9twxaICmK2sOt8+sGiSMAkNspYF/ZFfaVXaVlQq1F8sM4qMNjkRz+MqH0OB66OA2SQiORFBr5antHfULpVbc3RfEiTCgREeWATCaD+4feFu22lhQWhYh1/2RarsSgmrCvUtQix7RmtzUA0oxoQMpA2nkl9Tnn5XELghy9W0WKFEGpUqUwduxYlClTBn369EGDBg0sFRsRERFRgeJQqySKeJVAUlgUdDHJkLvYwb5KUbPHsJCpFLCv5Ar7SqkTSjqoH8VJyaTk/2KhfhQHXbwGSf9GIunfyFfbF1HCrrwTVOVdpG5vihJMKBERZYVMJoPMLvMZvsxV5I1imXZtVhS1R5E3illszCNr07f6sbOzg5ubW54d9/nz59JrNoDJmhwlj/r164ePP/4YH3/8Mf755x9s3LgR8+bNQ/fu3eHn55dn/RaJiIiI8guZXIYiVd0stz+VHHYVXWBX0UVaJjRpEkrhKQklkahB0q0oJN2KerV9EQXsyqUelNsZyhIOBeYPFCIiW2dO12a3zq8XqM/lK1euAADq1q1r0Aopt12+fFl6Xb9+/Tw7bkGQo+TR+PHjpdc1a9bE7NmzERkZiaCgIPTu3RtNmjRBr169UL58+RwHSkRERETmkSnlsKvgArsKaRJKj+NfJpNiUiWUtEi6HYWk26kSSvYKqcublFAqyYQSEVFuyWrX5vzsxo0buHnzJgDAz88vT4/9yy+/AAAaNGjAPEUWWbyTn5ubG959911ERkZi3bp1WLduHVq2bInly5db+lBEREREZCaZUi4lg5xQBgAgtKkTSi9bKT2Mg0jSIjksCslhqRJKdgqoyjkZjKGkdHdkQomIyEJy2rU5v1iwYAEAoFq1avD398+z4x44cAD//PMP5HI5Pvnkkzw7bkGRo+TRyZMn4ePjI/38999/Y8OGDTh8+DC0Wi2USiU6deqE/v375zhQIiIiKpx0Oh3u3r2L2NhYODs7o3LlynnaxN3SbOl8ZAo57Mo5w66cM5wapiwTWgH1k1ctlNThsVA/jINI1iL5TjSS77yanlpmJ4eqrHP6hJKiYP2hQ0SUVyzdtTkvabXaTMvMmzcPJ0+eRLFixbBo0aIMh7pJTEwEAKjV6gz3qdFoMj3uX3/9hc8++wwAMG7cODRs2DDTbchQjpJHEyZMwPr16xEaGopNmzbhypUrEEKgZMmSeP/99/H+++8bTIVHRERElBVXr15FcHAwoqNfJSxcXV3Rvn17eHl5WTGy7MkP5yNTyGBX1gl2ZZ3g9GZpACkJJc3TeMMxlP6LhUjWIfluNJLvpkooqeRQlXV6mUxygaq8M1SlmFAiIiro/vvvPwApyR6NRiPNZiaEwKVLl7B06VKcPn0aderUwfz581GpUiWT+woNDZUGtz5//rxBo5W0Hj58KL1OSEgwWPfkyRNs3boVa9asgYODA+bMmYOuXbtm+xwLsxwlj2JiYtC9e3cAKb8QNWvWRP/+/dGxY0eoVCqLBEhERESF09WrV7F9+/Z0y6Ojo7F9+3b07NnTZhIu5sjP5yNTyKAq4wRVGSegwcuEki6DhNK9GCTfi0EcXt7QK+WwkxJKKa2UVKUdIVPk3xZkRESU4o8//sDp06exceNGACmDUterVw/lypWDg4MDEhMToVKp0KBBA6xZswZvv/22yX29ePEC/fv3x507dyCEAACsWbMGwcHBGDZsGAICAqSyjx8/xt69e7F3715pmb+/P0qXLo1ixYpBp9MhISEB1atXx6RJk+Dv7w9nZ+dcehcKvhyPeSSTyfDOO++gf//+HK2ciIiILEKn0yE4ODjDMr/88gvs7OygUKRMlyyTyaQp59O+zmx9VspmZ70QAgcPHszwfIKDg+Hp6ZlvuuTJ5DKoSjtBVdoJqJ8qofQsQUomJYfHQP3fyzGU7scg+X4M4vQ7UKYkpOxSt1Aq7QiZMn+cPxERpWjatCmaNm2KiRMn5nhfxYoVw759+8wqW7p0aQwdOhRDhw7N8XEpczlKHslkMvz4449o2rRphuUmTJiA+fPn5+RQREREVIjcvXvXoGuXMfHx8dJTzoIgOjoaq1atQtGiRWFvbw87O7ssfVcqlVLCylpkchlUpRyhKuUIx3qlALxMKEUkGA7KHR4LkaSF+kEs1A9iEYdHKTtQvEooSa2UyjgxoURERGRlOUoejRgxItPE0a1btxAcHMzkEREREZktNjbWrHIuLi6wt7cHkNKFXt/EPfX3zJbldH3q7zn1+PFjPH78OFvbyuXyLCecMvqub9GVUzK5DCp3R6jcHeFY91VCSfs88WXrpFiow2OQHB4HkahJGaA7PNX11yeUyqVJKKksn1CypcHMiYiIbEmOkkejR482uS4pKQk7duzAsmXLzBp1nYiIiEjP3DEJunXrhipVquRyNObTJ5bSJpXu3LljViupFi1aoGjRokhKSkJycrJZ3/Wz0Oh0OiQmJkqz0+SUUqm0WDJKpVIZJGFkchmUJR2gLOkAxzru0nulTyi96vYWC5GQKqF04eUO5DKoSjsajKFkV9YJMlX2E175YTBzIiIia8nxmEdpPXz4EJs2bcKOHTsQHR0NIYTVm1ATERFR/lK5cmW4urpm2HXN1dUVlStXzsOoMpd6rKPUXn/9dbPOp2XLlllu6aLT6cxONJnzXf/QT6PRQKPRID4+Pmtvggl2dnbmJ51K2cG+oj3s7EpAmSSDLFINeYQaeJIM2cMkyBJ0UD+Mg/phHOL/fNlSSw6oSr0clLuCM1TlnKEq6wS5XeYJpfw8mDkREVFesFjy6Ny5c9i4cSN+/fVXaLVaCCGgUqlQokSJbDe/JiIiosJJLpejffv2Rv+g12vfvn2+6VKUm+cjl8tRpEgRFClSJCchSjQajUWTUfrWV8nJyUhOTja7S2JGZA4y2CntYCdTQikUUKllUGoVUEUoYBehgOqSEioooIIS9s4OcCjuBMeSLnAo7QrHskVRxKmIlKxSqVSZDs6e3wYzJyIisrQcJY+SkpKwZ88ebNy4EaGhoQBSmhyXLFkSffv2RUBAAFxcXODj42ORYM3RsWNH3Lp1y+T6Vq1aYeXKldLP9+/fx4IFCxAWFgadToe3334bo0ePttgNGBEREWWPl5cXevbsWWC6EuWX81EqlVAqlXB0dMzxvoQQUKvVWU46mVqn76InhECSOglJSHp1MFMNjBIB/PfyK5uio6Oxfv16uLq6Qi6XQy6XQ6FQWPR7VsuyZT/ZEo4XRlTwZSt5FB4ejk2bNiEoKEjqmgYANWvWRExMjDR1rt5nn31mmWgzceHCBdy6dQtKpRIuLi4GAz3qdDo8f/4cbdq0kZbdunULAwYMQN++fbFo0SIkJydjxIgRGDp0KH788UeDcyAiIqK85+XlBU9PzwLzR0lBO5/MyGQyqbuaueNYZSQrXfSSkpKQFJeIxKh4JMYmICkhEclJSUjWaaCWaaGGBjqZ+QOd37t3L8fxW5KlElGWTGrlZJ+Uf3G8MKLCIUvJo7NnzyIwMBAnTpyATqeDEAJFihRBhw4d0KtXL3h7eyMgICBd0qVdu3YWDdqUoKAgTJs2DQEBAdLMK3qnT5/G8OHD4evrCyDl5mPcuHEoVqwYPvzwQwApffGnTp2KTp06Yfny5fjkk0/yJG4iIiIyTS6X29Sg2DlV0M4nL1mii542OkkalDvhQTT+vXMLR/B3ptvV1FSAi3CADgI6mYBOAQi5gE6e8l3IAZ0cKetkAuLld51MvNoGupTX0EEn9N9ffukEtEIHIXTQ6nTQ6XTQ6rTQ6XRG49G9LFNQ2HpyKyvfC1OrMI4XRlR4mJ08ev/993Hp0iUAKU2FK1eujF69eqFr164oWrSoVM5aH5bJycmoX78+evbsaXR9cHAwmjRpgmLFigEA9u7di+vXr2PUqFEG5apWrYqqVati7dq1GDhwoFSeiIiIiPI/has9HFzt4VCjBFwBOP3ljjO7ryEOSYCx21gBOMEejTUekKcuoMmbeMXLfzqlDEIpg1ABQomU10qkJLEUKT9Lr/WJLQWkpJaQv0xgyQWE7FWSS0puQffqtZTQepnE0mpz9D3ta2O0Wi20Wq3UNTE/k8lk+boVWNrvpv6+0+l0HC+MqBAxO3m0Zs0a7Ny5E5s2bUJERAR69+6N9957zyJNkC3Bzs7OZOJIo9HgyJEjmDBhgrTswIEDAIBatWqlK+/t7Y1///0XBw4cQJ8+fXInYCIiIiKyOqVrETRRe+CY6jIgYJhAetmrrYnaA6U+qA27Sq4Qat3LL23Kd41O+g5p3cv1mtQ/pyqr3zbN9oZltYAOkL38J9cgJWGVmCa4tMFmSAbjGTLjRWUqecqXUvHqtUoOmTLVa5XCYBkM1r1cn2oZlLKXyS28Sm4pXia0RNYSUZZIZuV0H/rhOwyuhBDQaPIou5gHTCWZdDpdpgPgR0dH4+7du2xtSVQAmJ08cnZ2xsCBAzFgwAAcO3YMgYGBWLJkCfz8/NC3b19UrVo1N+PMkXPnziEuLg5t27YFkNJK6dy5cwCAcuXKpStfsWJFAMBff/3F5BERERFRAWZfpSiquVQAYoCzqpspLZBecoI9mqg9UM2lIuxfd4NMLgPsTI3MbXlCm4XkU9r1aZNR0noT26p1gCZVqyABiGQdRLIOedbMSiEzSFApVXKo0iSfUtYpjSauZCo5ZEXSJLzSJbP0yxSAUpbjXhP6RJKtJLNy+j2jc8xuQswSMywSkfVlecBsmUwGX19f+Pr64vr16wgMDES3bt1Qr1499O3b12hT1ISEBDg4OFgk4OzQd1lzc3MDADx79gyJiSmPbdzd3dOVd3FxAQDcvXvX5D71083q6T8U9U1uKYX+veB7Ynt4bWwbr4/tstS1KajX1lT9SGSrZHIZ3DpXRZWNyaic5I5H8kgkIAkOsEcZnRvkkMGt8+spiaO8jk0hh0whB/JoEmAhBKAR2U4+GWxntNWVNt0yaFO13NEKCK0WIjGPPh9leJVcSt06ymjCKn3yydg6hUoOpUrxMsFll7LcIVVCS2Hb4yFlJREVHh4u9ebIiK30VCGinMnWbGt6np6emDVrFsaPH4+tW7fiyy+/RGxsLLZt24auXbtKA2f36tULu3fvtkS8WabVanHkyBFMnDhRWvb8+XPptbFpaPWJrtQzBqS1atUqLFu2LN3ykJAQi0xtW9BcvnzZ2iGQCbw2to3Xx3bl9NrEx8dbKBLbYqp+JLJlDrVKokTfGojcdwvlol6Nd6koag+3zq/DoVZJK0aXd2QyGaBKaf2TV4ROGG8tZW7yKYNuhGm7AUotrPT5KoFXy/KKHFnqBphZwgoGZY0ktJTyLCU+9V3UlMrM/0wsW7Ysfvv1FGLiY02OF+bimDKjJBHlfzlKHukVL14cI0eOxLBhw3DgwAEEBgZi6dKl6NOnDypUqIDQ0FBLHCZbzp07h9jYWKnLGgAkJb1qjqxSqdJto38anHbWuNSGDx+OQYMGST/HxsbCx8cH3t7ezK6notVqcfnyZdSuXRsKRd4186bM8drYNl4f22Wpa1NQW+SYqh+JbJ1DrZIo4lUCSWFR0MUkQ+5iB/sqRa3S4qgwkcllkNkrAPu8qeuEECktnAySS1rjCSeDZJTW9LoMt9cCmlStq3SASNZCJOdh61OlLH3CKm2CKaPElbFxr1RyQC5Dk6RqKbMVZjBemMzccbaIyKZZJHkk7UyphJ+fH/z8/HDx4kWsXr0aixcvtup0lcHBwWjWrBlcXV2lZalfJycnp5vuNSEhAQCkbm7G2NnZGU0uKRQK/qFnBN8X28VrY9t4fWxXTq9NQb2upupHovxAJpehSFU3a4dBuUgmk71MpuRhd0CdALI6fpVZ3QBNbw9dqoSVRkBoNBCJpmPMrsoogTby2ibHC6ucVAxJYVH8f0VUAFg0eZRa/fr1sXLlSuzduxeTJk3KrcNkyFiXNQCoVKkSFAoFtFotXrx4gbJlyxqsf/HiBQCgfPnyeRYrEREREREVPDK5DJArUgbpziNCK9K3lkrXLTCD9UZaUyHN9tq4ZIgELaroSpkcLwwAdDHJmURLRPlBriWP9Pz8/Kw23tH58+cRExMDX19fg+X29vbw9vbGX3/9hbt376ZLHj148AAA0KxZszyLlYiIiIiIyBJkChlkCiVgn3vHSLwViWc/pIz/J4cM5XTFjJaTu7A1KlFBkOvJIwBYu3ZtXhwmneDgYLz11lsG3dT0OnfujL/++gshISFo0qSJwbqrV6/C3t4eb7/9dl6FSkRERERElG/YVykKRVE7aKNMtyxSFLWHfZWieRgVWduJEyfw448/4vHjx3B2doZMJkOTJk3QqFEj7NmzBwsXLsS0adNw/PjxdBNZ1axZExs3bjTY3+LFi7Fz5048efIEQMrkVnXq1MF///2He/fuSeWmTp2KAQMGGI1p1apV2Lx5Mx49egQgZXb1pk2bYunSpVKZHTt2YMOGDbh79y7c3d0REBCAoUOHmhyCJygoCFu3bkVycjKioqLQokULjB07FsWLFzco9/jxY7Rq1SrDWXYnT55sMF6krcq7qRTymL7LWocOHYyu7969OypUqIDg4GCD5Tdv3sTt27fRp08fFCtmPHtORERERERUmMnkMrh1rpphGbfOr3PQ+UJkzZo1GDVqFPr27YvDhw9j165d+OGHHwAAH374oTRT7axZs3Dy5EnUqlULAFCqVCkcP348XeIIAMaMGYNTp06hU6dOcHd3x5EjR/DTTz/hyJEj2LVrFypWrAgA+Pbbb/HHH38YjWv48OE4ceIE/P39UblyZZw4ccIgcbR06VJ8/vnniImJgU6nw4MHD7BgwQIsX77c6P6++uorrFu3DosXL8aePXuwadMm/Pnnn+jZsyeePn1qUHbHjh3QarWwt7dHiRIlULJkSenLxcUFANC6deusvM1WU2CTRxcuXEBMTAzatGljdL29vT3mzZuH27dvY8uWLQCAyMhIfPbZZ6hfvz7GjBmTl+ESERERERHlKw61SqJE3xpQFDXsmqYoao8SfWvAoVZJK0WWf+l0OoSFheHy5csICwuDTqezdkhmCQsLw8KFC9G3b1+0a9dOWl68eHFMnDgRY8eONShvZ2eHN998EwBQp06dDBtuyGQyNGzYEA0aNIC7u7u0vGbNmliyZAkAQKPRYOzYsbh//77JfTRo0AC1atUymB392rVr+Ouvv3Ds2DGcOHEC58+fh7+/P4CUZJhGozHYz6FDh7Bx40Z89tlnKFeuHICUsZLnz5+P+/fvY+rUqVJZrVaLEydOYM2aNbh06RLOnDmD33//Xfrq3bs3vLy8ULlyZZPnbkvypNuaNei7rOmzecbUr18fGzduxPz587Ft2zYAQMeOHTFgwADY2+diB2EiIiIiIqICwKFWSRTxKoGksCjoYpIhd7GDfZWibHGUDVevXkVwcDCio6OlZa6urmjfvj28vLysGFnmTp48Ca1Wi1KlShldP2jQIPz6668Gy/RJnLSznxtjb28PBweHdMtdXV3h4OAAjUaDyMhIjBw5Elu3boWTk1O6siqVKt2MsHfu3MGKFSukGBwdHfH111/j5MmTiIyMRExMjEFia/Xq1VCpVFLiS8/Lywt16tTBqVOnEBISAm9vb1y9ehXTp09H3bp1jZ5TcHAw3nvvvUzP3VYU2OTRF198YVa5WrVqYf369bkaCxERERERUUElk8tQpKqbtcPI165evYrt27enWx4dHY3t27ejZ8+eNp1AEkIAAAIDA9G5c2eULGnY6szOzg6NGzfOlWMXL14co0aNwuTJk3Hz5k1MnjwZS5YsMTleUWrGhrmxs7NDpUqVUKpUKYPEUXx8PP755x+UKlUKcnn6TlwNGzbEpUuXcOzYMXh7e6N27domj3vt2jXcvXvX5DA7tqjAdlsjIiIiIiIisjQhBJKTky32lZiYiIMHD2Z4zODgYCQmJlrsmPpkj6U0bdoUABAeHo6uXbvi5MmT6cp88sknFj1mal27dsUHH3wAADh8+LDJ8YrMkZycjPv372PatGkGy2NiYiCEQGxsrNHt9N3Ybt26lekxgoODUbNmTVSqVCnbcea1AtvyiIiIiIiIiMiShBBYu3atybF1ckt0dDTmzJljsf1VrFgRgwcPNqt1jjk8PT3xwQcfYM2aNXjy5AmGDRuG1q1bY8KECahaNeOB1S1l/PjxuHXrFn799VcsW7YMnp6e8PX1zfJ+lixZgnHjxqWblb1YsWJQqVSIi4tDWFgYqlSpYrBePz5VXFxcpsfIb13WALY8IiIiIiIiIqIcmjhxIqZMmSKNH3T8+HH4+fnhiy++QGRkZK4fXy6XY8GCBfDw8IAQAhMnTkRoaKhZ26rVavz5558YPnw4fvjhB3z//fc4duyYQRk7Ozs0b94cAIx2MXz8+DGAlHGYMnL9+nXcuXMnX3VZA9jyyGL0zf5MNWErrLRaLeLj4xEbGwuFQmHtcCgVXhvbxutjuyx1bfT1haWbjdsa1o9ERJQVtl4/ymQyDB48GGq12mL7vHv3LjZt2pRpuT59+lhsZi6VSmWxVkepDRw4EG3atMGcOXNw9OhRaDQabNmyBYcPH8bChQvTteaxNCcnJ6xcuRLdu3fH8+fPMXLkSOzYsQNubm4ZbvfixQuEh4ejTJkycHR0xH///YdRo0ZhzZo1aNasmVRu6tSpuHTpEgIDA1GtWjV06dIFWq0WBw4cwJ49ewAAHh4eGR7r4MGDqFWrFipWrJjj881LTB5ZiL5pmo+Pj5UjISKi/CQuLi7DmUHzO9aPRESUHbZcP8pksnSzduVE1apV4erqajDLWlqurq6oWrWq0YGabU3FihWxfPlyXLhwAXPmzMGVK1cQERGB4cOHY9euXVI3tqwkr4QQZpcvX748li1bhgEDBuDevXv45JNPsGbNmgy3KVWqFPz9/eHv74+PP/4YI0eOREhICJYuXWqQPKpUqRKCgoKwfPlyrF69WjqfFi1aSInPli1bZnisQ4cOoXv37madiy1h8shCSpUqhZMnT8LJySlXMrj5VWxsLHx8fHDy5ElpKkayDbw2to3Xx3ZZ6toIIRAXF2dyStuCgvWjcfw/brt4bWwbr4/tYv2YfXK5HO3btzfaFUqvffv2+SJxlFrDhg2xY8cO/PTTT/j222+RmJiIlStXYt68eQAAe3t7ACkDVGcmMTFR6g5njgYNGuCrr77C5MmTcebMGcydOxeenp5mbevu7o7Zs2ejU6dOuH79err15cqVw6xZswyW7d27F/Hx8ahXrx5q1qxpct83btxAWFgY2rdvb/a52AomjyxELpejTJky1g7DZjk7O7OCt1G8NraN18d2WeLa2OoTVUti/Zgx/h+3Xbw2to3Xx3axfsweLy8v9OzZE8HBwQYtkFxdXdG+fXt4eXlZMbrM7d+/H+XLl0fdunUNlsvlcgwaNAiRkZFYuXIl/vnnH2ldsWLFAABPnz7NdP8RERFZTih27doV//77L9asWYOffvoJzZo1Q+nSpc3atlq1anj99dfx/PnzTMsmJCRg0aJFkMvl+PTTTzMsGxwcjNq1a6NChQpmxWFLmDwiIiIiIiIisjIvLy94enri7t27iI2NhbOzMypXrpxvWhydPHkyXfJIr1OnTli5cqXBYNL6lkA3b96ETqfL8DwvXbqEQYMGZTmm8ePH4/bt2zh+/DjOnDmDrl27mr2tnZ0dGjVqlGm52bNnIzw8HGPHjkX9+vUzLBscHJwvu6wBnG2NiIiIiIiIyCbI5XJUqVIFtWvXRpUqVfJN4ggAtmzZgocPHxpd9++//wIA3n33XWlZrVq10LRpU8TGxuLgwYMm93vy5Em8ePHCYOwhvaSkpAwHL5fL5Zg/fz6qV69u7mkAAB48eIB79+5hxIgRJssIITB//nxs27YNQ4YMybAskJIku337dr7ssgYweUS5zM7ODh9//LFFB5Qjy+C1sW28PraL14Ysgb9HtovXxrbx+tguXht68eIFevXqhUOHDhkkdC5cuIBZs2ahbdu26NWrl8E233zzDV577TVMnz4du3fvhkajkdY9e/YMy5Ytw9y5c/Hdd98ZTaRduHABz549w71790zG5eTkhO+//x4lSpRIt27q1Kno168fDh06BJ1OBwAIDw/Hp59+irlz5xrtLqjT6XDu3Dn069cP27Ztw7x58zLtrgaktDqqU6cOypcvn2lZWyQTtjoHIhEREREREVEeSkxMRFhYGKpUqZKlAZoLu/379+P58+dwdXXFH3/8gRs3bkAulyMqKgpubm7o3r07evbsaTQBFBUVhQ0bNuDEiROIiIiAm5sbVCoVFAoF2rVrh+7du6cbB+v06dP46quvcPfuXQghYG9vj9deew0//fSTNJZSWhcvXsTu3bsxc+ZMadmqVauwefNmaUylmjVrokKFCujXrx/KlSuXbh8jRozA5cuXUapUKbRp0wa9e/c2eby0OnbsiB49emSr+11uMvd3nskjIiIiIiIiIjB5RIWPub/z7LZGREREREREREQmMXlEREREREREREQmMXlEREREREREREQmKa0dABU8sbGxWLx4MQ4fPoyIiAiUKVMGnTp1wogRI9hv2AYJITBkyBDcuXMHx48ft3Y4lMrdu3exceNG3L9/H5UqVYKPjw/eeusta4dVaG3duhXbtm1DcnIyNBoNypUrh5EjR6Jhw4bWDo3yCdaP+QvrR9vF+tH2sI4kKviYPCKLio+PR58+fXD9+nWUKlUKcrkc9+/fx8qVK3HmzBls2LABDg4O1g6TUlm/fj1+//33fDtlZEGk1Wrx3XffISgoCNOmTcO0adOsHVKhN2vWLBw+fBirV69G9erVAQBbtmzBgAEDsGzZMrRu3drKEZKtY/2Y/7B+tD2sH20T60iiwoHd1sii1q1bhxIlSuDo0aP47bffcPHiRUyYMAEymQwhISH48ccfrR0ipXL9+nXs2LHD2mFQKklJSRg2bBh2796NLVu24N1337V2SIVeWFgYAgMDMWrUKOmmGAB69eoFLy8vfPvtt1aMjvIL1o/5C+tH28P60TaxjiQqPJg8Iou6evUqvv/+e1SsWBEAoFQqMXToUPTs2RMA2OzbhiQlJWHSpEmYPXu2tUOhl3Q6HSZMmICzZ8/i+++/x2uvvWbtkAgpn2tCCLi5uaVbV7FiRTx+/Djvg6J8h/Vj/sH60fawfrRdrCOJCg8mj8hihBAYNmwY7O3t063r1KkTAECtVud1WGTC3Llz0bFjR9SpU8faodBLmzdvxuHDh9G/f3/Url3b2uHQS6VKlQIA7Ny5M926sLAwvP3223kdEuUzrB/zF9aPtof1o+1iHUlUeDB5RBYjk8lM3miVLFkSAAyas5L1nDhxArdu3cLQoUOtHQq9FBUVhaVLl0KpVGLw4MHWDodSadCgAapVq4Zff/0VS5culZYfPXoUUVFRmDJlihWjo/yA9WP+wfrR9rB+tG0FuY4UQlg7BKI8Ye7vOgfMpjxx+/ZtAEDXrl2tHAk9e/YM3377LdatWwe5nPljWxEcHIzIyEg0bNgQf/zxB06dOoVbt25Bp9Ohbdu2GDp0qNFWC5T75HI5lixZgv79+2PZsmV49OgR/Pz8sGfPHuzYsUP6458oO1g/2g7Wj7aJ9aNtK4h1pP7/v1artXIkRHlD/7ueWd3HmpHyxMGDB9G4cWNOo2oDpk6dinHjxqF06dLWDoVSOXz4MAAgOTkZZcuWxbx587Bp0ybUqVMHS5cuxbBhw6DRaKwcZeFVtWpVbN68Ga+99hp27tyJDz74AD169MiXN8VkW1g/2g7Wj7aJ9aPtK2h1pEqlgkqlQmxsrLVDIcoTMTEx0u99Rpg8olz377//4vTp0/jqq6+sHUqhFxgYiPLly8PX19faoVAaYWFhAIBRo0ahYcOGkMlkcHR0xOeff44qVarg7Nmz2Lp1q5WjLNwiIyPRt29fTJ06FVqtFh9++CE2bdpk7bAoH2P9aDtYP9ou1o/5Q0GqI2UyGVxcXBAVFYWEhARrh0OUqxISEhAdHQ0XFxfIZLIMy7LbGuUqjUaDzz77DF9//TUqV65s7XAKtdDQUOzevRubN2+2dihkxNOnTwEA7u7uBssVCgW6d++OefPm4dChQ+jbt681wiv0Tp8+jVWrVmHDhg2QyWQoX748PvnkE8ycORNOTk7o0qWLtUOkfIb1o+1g/WjbWD/avoJYR5YsWRIJCQm4d+8eXF1d4eLiAoVCkekf10T5gRACWq0WMTExiI6Ohr29vVktBZk8olw1f/58vPPOO3jnnXesHUqht379ely5cgXe3t5G14eHh0sDtt64cSMvQyMADg4OSE5OhlKZ/mO5Vq1aAIAnT57kdVgE4PHjxxg1ahS+/vpr6abR19cXixYtwscff4w5c+bA19cXzs7OVo6U8hPWj7aD9aNtY/1o2wpqHalQKFCxYkU8e/YMMTExiIyMtHZIRBanUqng5uaGkiVLQqFQZFqeySPKNZs2bYK9vT0GDhxo7VAIQPHixVGlShWj68LCwqBUKlGxYsU8jor0qlSpgr///huPHj1CtWrVDNbpn7YWLVrUGqEVejt37kR8fHy62bDatGmD999/H5s3b8bff/+N5s2bWylCym9YP9oW1o+2jfWjbSvIdaRCoUDp0qVRqlQpqNVq6HQ6a4dEZDFyuRwqlSpLremYPKJcsW/fPoSFhWH69Onp1iUkJMDBwcEKURVu48ePx/jx442uq169OkqXLo3g4OA8jor02rRpg7///hvnzp1Ld4P1/PlzAECTJk2sEVqhFxUVBQCIi4tLt87HxwebN2/moJpkNtaPtof1o21j/WjbCkMdKZPJYGdnZ+0wiKyOA2aTxf3yyy+4cOECpk2blm7d+fPnsWjRorwPisjG9e7dGyVKlMCuXbuQmJhosO706dNwdHREnz59rBRd4daoUSMAKdchrQcPHsDOzg5vvvlmXodF+RDrR6KsY/1o21hHEhUebHlEFrVjxw58/vnnqFixIjp06CAtF0IgJiYGERERWL9+vfUCJLJRzs7OWLJkCQYNGoRJkybh22+/hb29PU6fPo3AwEDMmTOH00dbia+vL3x9fbFmzRrUrFkTLVu2BABcunQJy5cvx6RJk/LtdMSUd1g/EmUP60fbxjqSqPCQCSGEtYOggmH//v0YN25chmXc3d1x6tQpyOVs9GZLqlevjvLly+P48ePWDqXQCwkJwYIFC3D79m2UKFEC7u7u+Oijj1C3bl1rh1aoqdVqrFu3Dnv27EFiYiIcHR1RvHhxDB48GD4+PtYOj2wc68f8i/Wj7WD9aLtYRxIVDkweERERERERERGRSXy8RUREREREREREJjF5REREREREREREJjF5REREREREREREJjF5REREREREREREJjF5REREREREREREJjF5REREREREREREJjF5REREREREREREJjF5REREREREREREJjF5REREREREREREJjF5REREREREREREJjF5REREREREREREJimtHQAR5Z1r166hS5cuUKlUqFy5Muzt7QEAN2/ehFqtxmuvvQYnJyfodDo8fPgQkZGRaNSoEQIDA/M81pMnT2L27NkICwuTlsnlclStWhXffPMNvL29cf78eUyePBnh4eFSGRcXF8yYMQP+/v55HjMREeVPrB+JiIgyJhNCCGsHQUR549q1axg9ejQCAwNRpkwZaXnr1q0RHh6OdevWoVmzZgAArVaL1atX48yZM1a5OdbbuXMnpk2bBgD4/vvv0bp1a4P1Op0OjRo1glarxcSJE9GzZ08olcyLExGR+Vg/EhERZYw1CFEhM3LkSIMbY1MUCgVGjBiBZ8+e5UFUpjVo0EB6rb9xT2379u0oUqQI1qxZA09Pz7wMjYiIChDWj0RERKYxeURUiJQsWdLoDWZGunTpkjvBmEmlUkmv0z4xDQ4OxsqVK7Fhwwa8/vrreR0aEREVEKwfiYiIMsbkEVEh4u7unuVtateunQuR5NyePXuwcOFCrFu3jjfGRESUI6wfiYiIMsbZ1ogoU4mJiVi+fDnef/99dO3aFS1atMDUqVPx33//GZRLSkrCpk2b0KpVKzx48AA3b95Ev379UKdOHbRv3x579uyxSDzr16/Hd999h/Xr1/PGmIiIrIb1IxERFRZseUREGYqNjUX//v1RtWpVbNiwAXZ2drh+/ToGDx6MY8eOYf369ahRowbOnTuHWbNm4caNGwCA8PBwjBkzBkWKFIFarUZYWBg+/fRTaDQavPfee9mKRQiBuXPn4tChQwgMDETFihUteapERERmY/1IRESFCVseEVGGvv32W4SFheHzzz+HnZ0dAMDT0xMzZsxAZGQkxo4dC7VajcaNG2P37t2oXLkygJSnn+vXr8eJEydw6tQp1K9fHwAwd+5cxMfHZyuW8ePHY+3atahfvz7Kly9vmRMkIiLKBtaPRERUmDB5REQmPX/+HLt27ULt2rXh7OxssK5du3YoXbo07ty5g2PHjgEA5HI5SpUqBQAYPXq0NLtLyZIlsWjRIqhUKkRFReHs2bPZiqdq1aoAgH379mHKlCnQ6XTZPTUiIqJsY/1IRESFDZNHRGTS6dOnoVarUbx48XTr5HK5NE3w//73P2m5TCYDALi4uBiUL126NJo2bQoACAsLy1Y8H330ESZPngwA2L17NyZOnAitVputfREREWUX60ciIipsOOYREZn06NEjAEB0dLTR9WXKlAEAJCQkmLW/atWq4dSpU9IN7aeffoqQkJB05b799lt4e3sb3cegQYPg6OiIL774Ar/88gvUajUWLFhgMGUxERFRbmL9SEREhQ2TR0RkkpOTEwDg4cOHRtfrx3goV66cWfvTP3XV31Q/fPjQ6FPWzG62AwIC4ODggClTpuDQoUPQaDRYtGiRFA8REVFuYv1IRESFDZNHRGRSo0aNAKQ0o7937x4qVapksP7FixcAgFatWqXbVgiRbtmDBw8gl8ul/QYGBmY7Nj8/Pzg6OmLs2LE4duwYPv74YyxduhT29vbZ3icREZE5WD8SEVFhwzGPiAiJiYkAALVabbD8jTfeQIsWLSCEwJYtW9Jt99dff6F58+aoUaNGunWPHz82+Dk6Ohpnz55Fx44dpSerWYkNADQajcE6X19frFixAiqVCidPnsTAgQMRGRlp9r6JiIgywvqRiIgoBZNHRIVcaGgonj9/DgA4f/58uvVff/01ypcvjw0bNuDgwYPS8hUrViAhIQFfffWV0f0uXLhQ2m9SUhK++OILuLm5Yfr06VmK79y5c9LrM2fOpFvfokUL1KpVCwBw8eJF+Pv74/DhwxwolIiIcoT1IxER0SsyYaztLBEVeC9evED//v1x584dJCcnS8srVKiAYcOGISAgQFr2/PlzLFu2DEePHoWjoyOKFi2K+vXrY+jQoelmmunXrx/Onz+PGTNmYPfu3VCr1UhISECzZs0wevRoozPTGHPy5El88803uHv3rtTEXyaToWrVqpg9e7Y0YGjnzp1x8+bNdNu7urrik08+Qe/evbP83hARUeHF+pGIiCg9Jo+IyKL0N8fHjh1DhQoVrB0OERGRTWD9SERE+Rm7rRERERERERERkUlMHhGRRenHUmCjRiIioldYPxIRUX7G5BERWUxSUhLCw8MBADdu3LByNERERLaB9SMREeV3HPOIiCxi9erV+P777xEfHw8AUKlUqFWrFrZu3WrlyIiIiKyH9SMRERUETB4REREREREREZFJ7LZGREREREREREQmMXlEREREREREREQmMXlEREREREREREQmMXlEREREREREREQmMXlEREREREREREQmMXlEREREREREREQmMXlEREREREREREQmMXlEREREREREREQmMXlEREREREREREQmMXlEBZpOp8vyNkKIXIiEiIjItmSnjgRYTxIRERVGTB5RgbV3716sXbs2y9stWLAAZ8+ezYWIiIiIbEPqOvLp06dYtmwZ3n77bZw7dy7TbU+dOoXVq1cjOTk5t8OkfCAyMhLz5s1Djx49ULduXXTr1g0HDx60dlgmaTQaHDx4EH379sWUKVOsHU6WhIaG4uuvv0ajRo3w4MEDa4dDRIUMk0dU4Gg0GsyYMQPR0dH44IMPsrz92LFjERQUhNWrV1sspqFDh+Lu3bsW2x8REVF2pK0jjx8/jqlTp2Lp0qV48uSJWfvw8fFB1apVMWTIELx48cIice3fvx+LFy+2yL4o7zx58gQ9e/ZEq1atsGPHDnz33Xe4fv06xo4di6NHj1o7vHSeP3+OmTNn4osvvsCFCxfyVSu6AwcOYPHixQgMDERUVJS1wyGiQojJIypwZsyYAWdnZ/Tt2zdb2yuVSnzzzTc4ceIEVq5cmeN4bt26hd9++w2bNm3K8b6IiIhyIm0d2bp1a/zwww947bXXsrSfNm3aoGPHjhg4cCASExNzHNdPP/2E7du3szVTPjN79mzIZDK8+eabAIBWrVph2rRpcHNzQ5EiRawcXXrFixfHzJkzMXjwYIvuNzY2FteuXbPoPtPq2LEjli5dCkdHx1w9DhGRKUweUYGyadMm/PbbbxgzZkyO9qNSqTB79mysWLECf/zxR472tWHDBgghsGvXLsTHx+doX0RERNmVUR1ZvHjxLO+vV69eKFq0KL788sscxXXp0iVcunQJz549s+nuTmQoLi4OR44cgZubm8HyPn364Ny5c2jevLl1AjNDdn7fM7J169ZcTx4BgEwmg4uLS64fh4jIGCaPqMB4+PAh5s+fj379+lnkaVflypXRvn17TJo0KdtPQqOiovDLL7/AwcEBMTEx2L17d47jIiIiyqrM6kilUpmt/Y4YMQK7du3Cr7/+mu3YfvrpJzg5OQEANm7cmO39UN66f/8+1Go1VCqVtUPJMoVCYbF9/fvvvxZpqW6u7P5fJSLKKSaPqMBYtWoV4uPj4evra7F9durUCY8fP8aOHTuytf327dvRvHlzdOnSBQDYdY2IiKwiN+pIAGjcuDHc3d2xfPnybG3/+PFjnDx5Et999x0AICQkBCEhIZYMkXJJdHQ0gJTWMIXV/fv3MWzYMMTExFg7FCKiXMfkERUIMTExCAoKQpkyZVC1alWjZZKSkrBmzRp06dIF77zzDho3box+/frh1KlTJvfboEEDyOVyrFmzJsuDKmq1WmzevBmDBg1C//79IZPJ8O+//xrtBterVy9Ur15d+qpZsyZ27dolrd+3bx/q1auH6tWro1atWvj777+ldS9evMCcOXPg7++PRo0aoVWrVli4cCGSkpIMjnHmzBkMGTIEU6ZMQXJyMqZMmYL69evj66+/lspcvXoVI0eOhJ+fHxo3bozWrVtj1qxZiI2NNXqO586dw6BBg+Dj44OGDRuif//++N///mfyPTE31oxkNcYrV67g448/Rrt27fDWW2/hvffewy+//GK07J07d/Dpp5+iXbt2aN68OTp37oxNmzZBq9UCAEaPHg0vLy/pOulnOklISEDTpk2l5f369TPYb26895md1+jRow1+p2rVqoVVq1ZJ658/f45GjRpJ65ctW2bGu09E+ZE5dWRqDx48wMiRI1GvXj34+PhgyZIlJlvgyuVy1K9fH5cvX85WN+9NmzahU6dO8PHxQb169QAAgYGBZm379OlTfPnll2jXrh18fHzQvn17rFixwugYTDExMViwYAE6deqE1q1bw9fXF7Nnz5YGHl67di3q1q0rfSamroOHDh2KGjVqSOtSu3XrFqZPn4727dsDSOmq3rhxY/Tp00fqqq6Ps3PnzvDx8UGzZs0wZswYhIWFZeu8Nm3aZPD5Xr16dQwfPtxgH7179zZZJ5ny8OFDzJw5E35+fnj77bfRpk0bfPPNN4iMjDQot2fPHrRt2xbjx48HkNLtsG3btmjbti1mzZpl1rH+/fdfjBs3Du+++y7q1auHzp07Y+fOnenKZfXeLSkpCatXr4afnx/atGmDVq1aYerUqXj8+HGG8Rw/fhz+/v6oW7cuhgwZgqdPn2Z6Dv/88w/GjBkjvT/z589H27Ztpd8FIOX3buHChfD390fr1q3RpEkTDBs2zGiC9Pnz55g0aRI6dOiAhg0bStdv/fr1mcYyYsQIg9+HZs2amXUORERZIogKgJ9//ll4eHiIDz74wGSZwYMHiwYNGojbt28LIYS4fv26ePPNN4WXl5cIDQ01uV2bNm2Eh4eHuHLlSpZiOnDggAgICJB+/uCDD4SHh4cYMWJEurJarVZ89913wsPDQ3h4eIi//vorXZmQkBBRr149ce/ePWlZeHi48PX1Fdu2bRM6nU7Ex8eLTz/9VHh4eIgBAwYIjUYjHj16JAYPHiyqV68uPDw8xKRJk8T06dNF/fr1hYeHh6hevbpITk4W58+fFzVr1hRffvml0Gq1QqPRiBkzZggPDw8xbty4dPHs3r1beHp6iq+++kpotVoRFxcnBg8eLDw9PUWzZs1Eu3btRNeuXbMUa2ayGuPPP/8sGjVqJE6cOCGEECI+Pl5069ZNeHh4iKVLlxqU/f3338Wbb74pfv75Z6HT6URycrIYMWKE8PDwEJMnT5bKhYaGCi8vL+Hh4SHu379vsI/ly5cLDw8P0bdvXyGEyLX33pzz0mg0YtasWdLvlL5sajExMaJ69erihx9+yPS9J6L8y5w6sm/fvsLDw0Ps3LlTNGvWTLRo0ULUrVtX+gwZOXKkyW2XLVsmPDw8xFdffZWluBISEkSTJk3ErVu3hBBC7N+/X3h4eIiaNWuKp0+fZrjttWvXRNOmTcWqVauEWq0WOp1OfPnll8LDw0P0799f6HQ6qWx4eLh45513xMyZM0VSUpIQQojVq1cLDw8P0alTJxEfHy+ESPlMbN++vfDw8BBBQUEGx7tw4YL0XuhNnz5deHt7Cw8PD9GqVSuxd+9e0aBBA6ncmTNnREREhHj77bdFjx49RExMjBDi1fVo1aqVFE9Wz+vgwYNS3TJ37lyj71G3bt3EiBEjhFqtzvRaXL58WTRp0kSsXLlSOu7u3buFl5eXaNGihcG9h97Zs2cN6jxznTp1SrRq1UpcuHBBCCHEw4cPhb+/v9G6OSv3blFRUaJ79+5i1KhR0nut/5166623xJMnT6SyQUFBUr28a9cuUbt2bdGqVSvpPR06dKjZ5zNp0iSjvzOJiYmic+fOolWrVtLv85kzZ4SXl5do0KCBiIiIkMqq1WrRrVs3MWfOHOl+6PDhw8Lb21usW7fOYL+tWrVKdw+SkJAgvL29xZAhQ6T3iojI0tjyiAoE/dPOKlWqGF1//fp1nD59Gp6enlKZ6tWrw9fXFxqNBr///rvJfVeoUAEAcPr06SzF9NNPP2HQoEHSzwMHDgQA/Prrr1KLFT25XI7Ro0ejcuXKAJDuKR+Q0tLk/fffR8WKFaVln376KZo1a4aePXtCJpPBwcEBM2fOhJubG/744w/s3r0bpUuXxo8//ohhw4YBSOkSUL16dZw6dQpDhgzBuHHjoFKpsH79eqjVarzzzjuQy+VQKBTSk8y0T/hiYmIwc+ZMODo6YsKECZDL5XB0dMRnn30GIGWspw0bNhg8uTUn1sxkJcZr165h+vTpGDJkCHx8fAAADg4OaNeuHYCUp8x6jx8/xpgxY9ChQwd06dIFMpkMKpUKnTp1AgDs2rVLmo66WrVqcHd3Nxqf/qm5Xm689+ael0KhwIQJE1CpUiUAxn+n7t69i6JFi6J///7G33AiKhAyqyNT27x5M1atWoWTJ0/ijz/+QM+ePQEAR48eNdlqU19P/vbbb1mKa+/evahTpw5ef/11AEC7du1Qrlw5qNVqbNu2zeR2CQkJGDlyJGrUqIFhw4ZBqVRCJpPBz88PAHD27FlcvXoVACCEwJgxY6BUKjFt2jTY2dkBgNSdPDQ0VIrb2dkZtWvXNnrMtJ/vAPDVV19J3fXi4+Pxv//9D+fOncPnn3+O7t27o379+ti3bx8eP36M5s2bw9nZWTp2+fLlER4ejtu3b2frvNq3b4/OnTsDgNFp2zUaDe7fv4+PPvoo0zFyEhMTMW7cOFStWhXDhw+Xjuvv74/hw4fj0aNHGD16tNQKNydevHiBCRMmYMKECdIMbWXKlMHMmTMBAMuXL8fdu3cBZP3ebcaMGXjw4AG++eYb6b1u164dHBwc8PTpU6O/v5cvX8bFixdx6tQpHD9+HIGBgZDJZDh16hSeP3+eo3P97bffcOPGDTRs2BAlS5YEADRt2hT169dHTEwMLl68KJX9+++/ceXKFXTt2lUaj6lt27YYOnRopsdJSkrC2LFj0aNHD6xevdqs/+dERNnB5BEVCPoZLvSVc1rFihWDs7MzatSoYbC8dOnSAJBhX/Vy5coBSGmebK7Lly/j2bNnaNu2rbTsrbfewhtvvAGdTofNmzen20Yul6N3794AUsZKSmvHjh3o1auX9HNISAguXLggJQ307O3tpSmXDx06JC3XJ52cnJzQt29fODk54dNPP5USG5UqVYKjo6PBdM1lypQBkP79uXDhAmJjY1GpUiWDgVcrV66MGjVqQK1WG/wRkdVYTclKjN9//z3UajW6d+9usNzPzw+tWrUyeC/Xr1+P6Oho9OjRw6Csj48POnXqhG7duqFo0aLScrnc+EenqeWWfO+zcl52dnbSMbZu3Zourr1798LPz0/6Y4qICqbM6sjUJkyYgFq1agEAihQpgpkzZ8LDwwMADB4IpFa+fHkAKV1/4+LizI4rMDDQ4CGLQqFAnz59AKR8Zmk0GqPbBQUFITw8PN1ndu3atREQEIC2bdtKD2NOnDiBkJAQdOvWzeAz2t3dHR9++CFatGgBb29vabmpz3FTAyzrP9+TkpIwevRoKBQK9O7dG7NmzYK9vT0qVqwIuVye7v7D2Gd8Vs4LAEaOHAm5XI4DBw5I4w/pnTx5EmXLlkXNmjWNxp1aUFAQ7t69a3DPojdw4EDY2dnh6tWrOHr0aKb7MudYcXFxaN26tcFyT09PAIBOp8ORI0cAZO3e7ebNmwgODkaHDh2kxBGQct3Gjx+PZs2aoWnTpuniKV++PL766itpxriGDRuiSpUqEELg4cOHOTrXsmXLQqVSmbz2qbulR0REAEg/NmaPHj0yHFMqJiYGQ4YMQa1atTB9+nSTv79ERJbA4fqpQND3ZdfP1pJW6dKlcfbsWWlGkMjISOzevVuaElhkMJ6Ro6MjAODZs2dmx/PTTz9hwIAB6Srx/v37Y8aMGQgKCsLo0aPTzXjTrVs3LFq0CCdOnEB4eLh0Qx4SEoISJUoYtDo6e/YsAOCzzz5L91QxNjYWbm5uBjeT+htfU+NdTJo0SWoJAwB//vkngoKCAKR/f9RqtclzL1euHP755x+D9yursZpiboxarRa//fYbXF1d003HW6ZMmXSzopw8eRIADJI3QMpT6IULF2YaV2Ys9d5n9bwAwN/fH4sXL8bFixfx119/SU/Pk5OTsWfPHrPGUiCi/C2zOjK1tPWWTCZD//79MX36dKnVS1r6ehJI+SPYnOOcOXMGKpUKjRs3Nljes2dPLF++HE+ePMHhw4fRsWPHdNua+sxWKBRSC5a0ZVMnXfQ++eSTTOPMjP7zvVixYkanf2/dujVCQkKkz/e7d+9i+/bt+PfffwGkJEvSxmrOeQEpLcnatGmDI0eOYMuWLQbjHu3cuRPvvfeeWeegb5Gjv+dIzdXVFfXq1cO5c+dw4sSJdA+Bsurs2bMQQkgtv1LTJ3D0iZSs3LtldJ379etnctwnY9dMH0dCQoJ5J2VCzZo18ddff0nxP378GEFBQdI9UeprX7duXdjb22Pr1q24ffs2xo8fj7p166J06dIYMGCA0f0/e/YMH330ETp37owPPvggR7ESEZmDySMqEPRPb+zt7U2WUalUePToEVatWoWIiAh069YNHTp0wPfff5/hvh0cHAC8upnJzJMnT3D06FGEhISke4Kk0+mgVCoRGRmJffv2pXu66Orqis6dO2P79u3YunWrNBjlli1bpFZJevonYqtWrTJrAFRzKJVKHDhwAEFBQahXrx7GjBlj9Emzt7c3FAoF/vvvv3Tr9Ddzqbt2WTJWc2KMjIxEfHy8WX/AAEB4eDgAWKRJfnblxnkBKa2P+vXrh4ULF2LNmjVSF4vjx4+jfPny0tNeIiq4zKkjM6LvVmaqlW7qByERERFSd9mM/PTTT3jy5InB4MJ6+ocMgYGBRpNHWfnM1pdN/Yd6XlOpVAgJCcGaNWvg5uaGgIAAhISE4Pz58wblslMXDRkyBEeOHJFacdnZ2eHp06c4e/YsZs+ebdY+bt26BcD0rGlVq1bFuXPnctwSB0i5HyhatCiCg4PNKm/uvZslr7P+988S+1KpVLh165Y0YUWPHj1w7949/PzzzwblSpcujYULF2Ly5Mk4f/48AgIC0Lx5c0yePBlvvPGG0X3PmTMH169fN5owIyLKDWzbSAWC/qlnRrN27du3Dz179kTr1q2xZMkStGzZ0mQz9NT0T2HN7dqzefNmBAQE4PDhwwgODjb4Onz4sJQEMjWbjL7J/o4dO5CcnIzo6GhcunRJGt9GT39zaWq2lqx6/vw5hgwZgsOHD2Px4sX4+OOPpabVaZUtWxYDBw5EZGQkzpw5Iy0XQiA0NBQODg4G8VoqVnNj1N/wxcXFmTVmgb78vXv3chRfduXWeen16tULjo6OOH78uDS+RlBQkNlPpYkofzOnjsxIiRIlAAAuLi5G16duraRvZZGRO3fu4J9//sGvv/6arp4MDg7Ghg0bAAAXL1402topK5/Z1v581+l0mDt3LqZNm4bx48dj5syZJruSZSfWevXqoX79+nj69Kk0duDu3bvh4+MjtaDJjL6FjakZyVxdXQHAoDtYdmk0Gjx//tys1saA+fduuXGdM2qVbq61a9di6NCh6NevH7799ls0bNjQZFlfX18EBwejd+/eUKlUOH36NLp27YrDhw8bLT9p0iS4u7vj0KFDRlseExFZGpNHVCDob2yNTc8LAL///js+/fRTDB06FG+//XaW9q0fv0F/85SRpKQkbN++3WDcmbTef/99AMCNGzfSPXUEUvr9N2jQAC9evMD+/fuxe/du+Pv7p+tKoG/Zo2++bUxWpk0eNWoUbt68iW+//dasG8SJEyciICAAs2bNwv3796HRaPD999/j3r17mD59unRNLBmruTEWK1ZMesJu6pghISG4efMmgFfjD5gq+/DhwwwHVc+p3DovPVdXV/To0QM6nQ5r167Fo0eP8Oeff0qDrRJRwZZZHZkZ/R/6acdu0dNPSQ/AYHw4UwIDA9G9e3eTiaYaNWpI4xAZe9CS2Wd2XFwcDhw4YFZZnU5n0Mozo/FlsmPJkiVYu3Yt5s+fn2kLkaycV2pDhgwBkJKo0Ol0CAoKSjcuXkb0LcVMPeDRPwDSj32VE6VKlYIQwuQ5CiFw7tw5AFm7d9O/d0ePHjXZtX7nzp05iDzrduzYgblz52L69OkmB2JPq2TJkvj8889x8OBBvPXWW1Cr1Zg2bZrRscTc3d2xaNEiqFQqLF68ON0EG0RElsbkERUI+n76plpjbNu2DTqdThr8Oq2Mmibrm/ub091q79698PT0TDdeQWpVq1ZFgwYNAGTe+mjjxo34+eefjd4E6mcp2b9/v9FBLG/fvo1jx45lGjOQMtDkn3/+CXd3d5MtrNK+R2fPnsXBgwdRpUoVDBkyBB06dMDff/+NdevWpYvXErFmJUalUon69esDANatW2fwhw2QMmbTsmXLpBv5Ro0aAUi50Xv69Gm6/c6fP9/g+uu7aKSd4UbfpSOjMaHy8rxSGzhwIJRKJfbs2YPVq1ejTZs2JlsREFHBklkdmZkrV64AAN59912j6/WffQ4ODkbHzUlbds+ePdKDFFP0s7zt378/Xdz6z+yjR4+mS5YDwIoVK1C2bFmDsleuXJHGxUlt69atBp+9+gS9qc93IGXMOHNt2bIFgPHxhADDujUr55VamzZtUKVKFYSFheHbb79FUlISmjVrZnaMrVq1ApCStDLWZU7fXa1Dhw4Gy7PTMkd/P/Ddd9/h/v376db//PPPePLkCYCs3bvp37snT55gx44d6cqePHnS7OEHsspUwlF/7c2JPzg42GBW34oVK+KHH35A7dq1ER0dLXUtTOvNN9/EpEmToNPpMGHCBGmmOiKi3MDkERUI+hk0Hjx4YHS9voLeuHEjEhMTIYTA0aNHsWfPHgApU8dGRERIT7tS09/cZPbUSKPRYM2aNdL07hl55513AKTcIBq7IXjnnXfg7u6OK1euoEqVKgatePSaNGmCGjVqQAiBsWPHYuHChQgNDUV4eDh++eUXDB06FN26dUv3HhgbAFK/7vr169IsaU+ePMEXX3whlXnx4oX0xDM+Ph5jx45FmzZtMHfuXBw4cADBwcFYvXq10dlMshqrMVmNUT+Dz/379zFs2DBcv34darUa165dw/Dhw1GjRg3pj4R+/fpBqVQiJiYGgwcPxsWLF6FWqxEWFobx48dDJpMZdCPT3whu27YNWq0WOp0O+/btk8Y0CA8Ph0ajkW6sLfneZ+W8UitXrhzat2+P5ORkbNq0KV2CT6vVYsqUKRg4cKDJrgtElD9lVkemljYhoNVqsXXrVtSpU8foIMfAq3rSy8sr09meNmzYAE9PT5PdovX09WRSUlK6gf179OgBZ2dnaLVaDBs2DKdOnUJycjLCw8Mxa9YsXLx4UZocoH379tJn9sSJE/HLL78gMTERT58+xYoVK7BhwwbpWMCrz/d9+/ZJLa4uXbqEKVOmSGXu3bsnvU/676YGV9av/+GHHwCktP5at26dNIPrixcvEBISggcPHmTpvFKTyWQYPHgwgJQHC127ds3SrFsDBw6Eq6srnj59mm4souTkZJw7dw7vvvtuurF39A8wsjLDXkBAABwdHfHixQsEBARgy5YtuH//PsLCwvDDDz9g1apV0kxsWbl3e/PNN6XWarNnz8amTZsQGxuLyMhIbN68GZ9//rnBOJP6mfwyGl/K1Gx/aenr3LTdQvXXft26ddBqtVCr1QgKCpJaCL148QJ3796VfhfWrVtn8P9PoVBIyTb9DHPAq+SlPvZ+/frBz88PUVFRGDFiRLqxyc6ePYsePXpI3UGJiLKLySMqEPTNme/cuWN0ffPmzQGkzO7y1ltvoUmTJti8ebN0I7xz50707dvX6KCEd+7cgUKhQJs2bUweX6vVYsGCBbhz5w5evHiRabz6m0ydTocpU6akm8lNpVJJT13TDpStJ5PJsGDBApQoUQJqtRqrVq3Cu+++i9atW2P8+PHo1asXvLy8AKTcwPz9998AgAsXLqRLDrz++usoV64cdDodhg4dipYtW6J9+/Zo1qyZ1I2qY8eO0h8IL168QHR0NH7++WfUr18fNWvWhJeXF6pXrw5vb2907NjR4GY/K7GaktUYfXx8pJvpCxcuwN/fH7Vq1UKXLl2QnJyMkSNHSvv29PTE5MmTAaS0BOrVqxdq1aqF9u3b49q1a5g2bZpBLPrZZrZt2yZN/7tr1y6MHj0aQEryx9/fH8ePH7f4e5+V80pLPxtLxYoV081wdPXqVezatQt//PEH9u/fn+G1IKL8JbM6EgAqVKgAAJg1axYuX74MIKWumjp1KhQKBVasWGFynED9fo1N9Z7axYsXsXr1aiQmJqZrOZlWbGysdLx169ZJyXUgpbvOnDlzoFQq8fDhQwwdOhS1a9dG69atcfDgQcydO1cqa29vj4ULF8LR0RFRUVEYP3486tSpg+bNm+OHH37A3LlzDVoe+fr6QqFQ4J9//sHbb7+NFi1aYMyYMVIdAaT8sa5/WHDx4kUAKa26Uo8BqKe//1i5cqV0/xERESG1lJkxYwZWrFiB8uXLZ+m80urSpQtKliwJmUyW6cOYtEqWLImFCxfCzs4OX375pdSNPD4+HlOmTEGZMmUwffp0g210Oh2OHDkCAAgNDTXZMiatMmXKYNasWVAqlYiIiMAXX3wBX19ftG/fHkuWLMGsWbOkSSGyeu82b948uLu7Izk5GTNnzkSDBg3QuHFjfPPNN5g+fbo0s1rqrnHXr183SPzFxcVJrXf01zYz+uPry2/btg3Pnj2T4t+9ezeaNWuGJk2a4Ny5c9L/k2XLluHTTz9FtWrVAACnT5/G5MmTpZZ2jx49wtGjRxEQECAlj27duiXdM166dEmKYdq0aZDJZLh16xY++OADg3uNdevWISQkBN99951Z50NEZJIgKiC6d+8uatSoIeLi4tKt02g0Ys6cOaJx48aiSZMmYuHChUKtVot79+6J5s2bi27duol///033Xb37t0THh4e4sMPPzR53ISEBPHmm28KDw8P6atp06bizp07Rss3a9bMoKyHh4fw8vISK1euNCj36NEj4e/vn+l5P3z4UEyePFk0bdpU1KxZU/j7+4s9e/YY7KdRo0bpjrdo0SKD/Vy+fFl069ZNeHt7i4CAAHHlyhUhhBBz5swRDRo0EEuXLjUov3PnTtGsWTPRqlUrUa9ePVGjRo1057V+/fosxZqZrMYohBBBQUGic+fOombNmqJly5Zi0aJFIikpyej+jx8/Lnr27Clq164t3nrrLfHVV1+J6Ohoo2UXLVokGjVqJBo1aiRmzZolEhMTxdmzZ0XLli1FYGCgSExMzLX3PqvnlVrPnj3F8uXL0y1PSEgQPXv2FM2bNxc3b97MdD9ElL9kVEfqHTlyRAwfPlw0atRItGrVSvTo0UOsXr1aJCYmZrjv3r17i1q1aonHjx+bLDNjxgyDz0Jvb2+xatUqo2Vnz54tatasma5O6dixo0G5//3vf2LAgAGiTp06olGjRmLSpEni0aNHRvcZGhoqRowYIerXry/q1asnPvroI6P1vhBCBAcHizZt2oi6deuKESNGiP/++08IIUTdunXFnDlzxJMnT4QQKe9p2hj79etnsK+IiAgxcuRIUbduXdG2bVuxb98+IYQQhw4dEvXr1xejR48WsbGx2T6v1ObPny8GDBiQaTlTbty4IUaNGiUaN24sWrduLfz9/cWqVatEfHy8Qblff/1VNG7c2OC8a9SoIVq3bi1Onjxp1rEuXrwoBg8eLOrVqyfq1q0rBg0aJC5dumRQJjv3bg8fPhQTJ04UjRo1Et7e3mLgwIHi77//ltZHRkamq5fr1q0rtm3bJlauXCnq1q1rsM7Pzy/Tc0lISBCjRo0S3t7eYvjw4dLx4uLixOTJk0WDBg1EixYtxLp166Rzb9SokRgwYID0u3Tw4EHpmNWrVxctW7YUfn5+YvPmzUKr1QohhBg3bpzBvZanp6d49913hRBCjBw50iDumjVriunTpwshhNizZ4+oV6+e+PLLL826NkREpsiEsMBUAkQ24OTJkxg2bBiWL18OX19fi+xz06ZNmDlzJnbs2CE1h6aUllYfffQRpk6dmm5KZv1MKjt37sSxY8cQFBRkpSgprYSEBPj4+GDv3r2ZdhkhooIlN+pIIKWFUOPGjREQEIDPPvvMYvulrBs0aBC6devGyRCIiChXsNsaFRg+Pj5o3rw5tm/fbrF97tq1C++99x4TR2msXr0a7u7u6RJHQMqgzqVKlcLAgQPNHi+A8kZwcDC8vb2ZOCIqhHKjjgSAPXv2oGjRohg1apRF90tZc//+fVy9etVg/CYiIiJLYvKICpR58+YhNDTU6LgDWXX8+HHExsYaDJJJQGRkJFauXGly7Au9X375BS1atMijqCgzWq0W69atQ0BAgLVDISIrsWQdCaQM/vzjjz9i1qxZKFasmEX2Sdmzbt06dO7c2eiECURERJagtHYARJZUvHhxrFixAlOnTsWyZcsynTLYlPDwcCxZsgRr1qzhdOZpPHv2DImJidi2bRuKFi2KgIAAg2loo6OjsWPHDuzYscPiT7jJfEIITJ06FXfu3ME777wjTbVtye4qRJS/WKqO1Js6dSqGDh0qTfVOeScwMBA///wz3nrrLdjb22PXrl04dOiQtcMiIqICjGMeUYF0//59LFy4ECNHjjQ6g1pGQkJCEBgYiIkTJ6JUqVK5FGH+NmvWLIMpX93c3FC0aFEkJibiyZMnqFatGlatWpXjP0wo+54/fy5Nzw0ATk5O2LRpE2rUqGHFqIjIFuSkjgRSxjmaN28eWrRokeFMpJR73n33XYSGhko/f/311wZT0RMREVkak0dUYOl0Opw/fx5NmjTJ0nanTp3C22+/DZlMlkuRFQxnzpzBli1b8Ndff+HFixdwdnbGG2+8gY4dO6JHjx5QqVTWDrHQ+/LLL7F7927Url0bU6dOhaenp7VDIiIbkd06Ekh5yFKxYkV2VbOi3bt3Y9asWShVqhRGjx6Ndu3aWTskIiIq4Jg8IiIiIiIiIiIikzhgNhERERERERERmcTkERERERERERERmcTkERERERERERERmcTkERERERERERERmcTkERERERERERERmcTkERERERERERERmcTkERERERERERERmcTkERERERERERERmfR/KKytHwYejpQAAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 1200x350 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "matplotlib.rcParams[\"font.size\"] = 17\n",
    "\n",
    "\n",
    "# Extract the relevant data\n",
    "top_k_values = moe_data[\"top_k\"].unique()\n",
    "average_accuracies = moe_data.groupby(\"top_k\")[\"average\"].mean()\n",
    "tasks = [\n",
    "    \"svhn\",\n",
    "    \"stanford_cars\",\n",
    "    \"resisc45\",\n",
    "    \"eurosat\",\n",
    "    \"gtsrb\",\n",
    "    \"mnist\",\n",
    "    \"dtd\",\n",
    "    \"sun397\",\n",
    "]\n",
    "\n",
    "# Create subplots\n",
    "fig, axes = plt.subplots(\n",
    "    1,\n",
    "    2,\n",
    "    figsize=(12, 3.5),\n",
    "    sharey=True,\n",
    ")\n",
    "\n",
    "# Plot the average accuracy\n",
    "axes[0].plot(top_k_values, average_accuracies, marker=\"o\", linestyle=\"-\")\n",
    "# axes[0].set_title(\"Average Accuracy across different top_k values\")\n",
    "axes[0].set_xlabel(\"Top-K\")\n",
    "axes[0].set_ylabel(\"Average Accuracy (%)\")\n",
    "axes[0].grid(True)\n",
    "\n",
    "# Plot the accuracy for each task\n",
    "for task in tasks:\n",
    "    axes[1].plot(\n",
    "        moe_data[\"top_k\"],\n",
    "        moe_data[task] * 100,\n",
    "        marker=\"o\",\n",
    "        linestyle=\"-\",\n",
    "        label=task.upper()\n",
    "        .replace(\"STANFORD_CARS\", \"Cars\")\n",
    "        .replace(\"EUROSAT\", \"EuroSAT\"),\n",
    "    )\n",
    "\n",
    "# axes[1].set_title(\"Accuracy of Each Task across Different top_k Values\")\n",
    "axes[1].set_xlabel(\"Top-K\")\n",
    "axes[1].set_ylabel(\"Accuracy (%)\")\n",
    "axes[1].legend(\n",
    "    # place at the right of the plot\n",
    "    loc=\"center left\",\n",
    "    bbox_to_anchor=(1.05, 0.5),\n",
    ")\n",
    "# axes[1].grid(True)\n",
    "# Add text below each subfigure\n",
    "fig.text(0.25, -0.01, \"(a) Average accuracy.\", ha=\"center\", fontsize=18)\n",
    "fig.text(0.65, -0.01, \"(b) Accuracy of each task.\", ha=\"center\", fontsize=18)\n",
    "\n",
    "# Show the plot\n",
    "plt.tight_layout()\n",
    "plt.savefig(\"clip-vit-b-32_ablations-topk.pdf\", bbox_inches=\"tight\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "base",
   "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": "3.10.14"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}