notebooks/8plot.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "!bash /home/azureuser/cloudfiles/code/blobfuse/blobfuse_raadsinformatie.sh"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import sys\n",
    "sys.path.append(\"..\")\n",
    "\n",
    "# MAKE SURE TO SET-UP PATH -> use local to run with demo data; use azure to run with complete dataset (access required)\n",
    "# Select where to run notebook: \"azure\" or \"local\"\n",
    "my_run = \"azure\"\n",
    "\n",
    "if my_run == \"azure\":\n",
    "    import config_azure as cf\n",
    "    running_demo = False\n",
    "elif my_run == \"local\":\n",
    "    import config as cf\n",
    "    running_demo = True\n",
    "\n",
    "\n",
    "import os\n",
    "if my_run == \"azure\":\n",
    "    if not os.path.exists(cf.HUGGING_CACHE):\n",
    "        os.mkdir(cf.HUGGING_CACHE)\n",
    "    os.environ[\"TRANSFORMERS_CACHE\"] = cf.HUGGING_CACHE\n",
    "\n",
    "import pandas as pd\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Notebook overview\n",
    "\n",
    "*Previous notebook: baseline*\n",
    "\n",
    "*Next notebook: none, last notebook*"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Page distribution"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "txtfiles = pd.read_pickle(f\"{cf.output_path}/txtfiles.pkl\")\n",
    "notcleaned = pd.read_pickle(f\"{cf.output_path}/txtfiles_notcleaned.pkl\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# Assuming txtfiles is your DataFrame and num_pages is the column you want to plot\n",
    "plt.hist(txtfiles['num_pages'], bins=30, edgecolor='black')  # Adjust number of bins as needed\n",
    "plt.xlabel('Number of Pages')\n",
    "plt.ylabel('Frequency')\n",
    "plt.title('Histogram of Number of Pages')\n",
    "plt.grid(True)\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# Assuming txtfiles is your DataFrame and num_pages is the column you want to plot\n",
    "plt.figure(figsize=(8, 6))  # Adjust figure size if needed\n",
    "plt.boxplot(txtfiles['num_pages'], vert=False, patch_artist=True)\n",
    "plt.xlabel('Number of Pages')\n",
    "plt.title('Boxplot of Number of Pages')\n",
    "plt.grid(True)\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count     26739.000000\n",
       "mean       2019.839298\n",
       "std        6980.052643\n",
       "min           3.000000\n",
       "25%         291.000000\n",
       "50%         516.000000\n",
       "75%        1160.000000\n",
       "max      275597.000000\n",
       "Name: token_count, dtype: float64"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "count    26737.000000\n",
       "mean         5.420541\n",
       "std         14.832611\n",
       "min          1.000000\n",
       "25%          2.000000\n",
       "50%          2.000000\n",
       "75%          4.000000\n",
       "max        502.000000\n",
       "Name: num_pages, dtype: float64"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "count     20818.000000\n",
       "mean       2120.675617\n",
       "std        7545.434859\n",
       "min          20.000000\n",
       "25%         271.250000\n",
       "50%         455.000000\n",
       "75%        1106.000000\n",
       "max      275597.000000\n",
       "Name: token_count, dtype: float64"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "count    20818.000000\n",
       "mean         5.604381\n",
       "std         16.089036\n",
       "min          1.000000\n",
       "25%          1.000000\n",
       "50%          2.000000\n",
       "75%          3.000000\n",
       "max        502.000000\n",
       "Name: num_pages, dtype: float64"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "cleaned = notcleaned.loc[notcleaned['id'].isin(txtfiles['id'])]\n",
    "\n",
    "notcleaned_removed = notcleaned.loc[~notcleaned['label'].isin(['Adviesaanvraag', 'Amendement', 'Begroting'])]\n",
    "\n",
    "display(notcleaned_removed['token_count'].describe())\n",
    "display(notcleaned_removed['num_pages'].describe())\n",
    "display(cleaned['token_count'].describe())\n",
    "display(cleaned['num_pages'].describe())\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Amount of docs exceeding limit"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "notcleaned = pd.read_pickle(f\"{cf.output_path}/txtfiles_notcleaned_llama.pkl\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count     33117.000000\n",
       "mean       5823.096416\n",
       "std       31149.520580\n",
       "min          15.000000\n",
       "25%         660.000000\n",
       "50%        1206.000000\n",
       "75%        2642.000000\n",
       "max      618067.000000\n",
       "Name: count_LlamaTokens, dtype: float64"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "notcleaned['count_LlamaTokens'].describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>count</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>min</th>\n",
       "      <th>25%</th>\n",
       "      <th>50%</th>\n",
       "      <th>75%</th>\n",
       "      <th>max</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>label</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Actualiteit</th>\n",
       "      <td>800.0</td>\n",
       "      <td>1629.756250</td>\n",
       "      <td>8662.494032</td>\n",
       "      <td>165.0</td>\n",
       "      <td>512.50</td>\n",
       "      <td>732.5</td>\n",
       "      <td>1305.00</td>\n",
       "      <td>234499.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Agenda</th>\n",
       "      <td>2544.0</td>\n",
       "      <td>1999.538522</td>\n",
       "      <td>10121.983150</td>\n",
       "      <td>74.0</td>\n",
       "      <td>516.75</td>\n",
       "      <td>1053.5</td>\n",
       "      <td>1918.00</td>\n",
       "      <td>314952.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Besluit</th>\n",
       "      <td>625.0</td>\n",
       "      <td>2182.633600</td>\n",
       "      <td>3918.344740</td>\n",
       "      <td>227.0</td>\n",
       "      <td>387.00</td>\n",
       "      <td>979.0</td>\n",
       "      <td>2619.00</td>\n",
       "      <td>56288.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Brief</th>\n",
       "      <td>1056.0</td>\n",
       "      <td>3137.704545</td>\n",
       "      <td>3082.126520</td>\n",
       "      <td>312.0</td>\n",
       "      <td>1363.00</td>\n",
       "      <td>2297.0</td>\n",
       "      <td>3841.75</td>\n",
       "      <td>56355.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Factsheet</th>\n",
       "      <td>214.0</td>\n",
       "      <td>12337.341121</td>\n",
       "      <td>26342.297934</td>\n",
       "      <td>261.0</td>\n",
       "      <td>2327.25</td>\n",
       "      <td>5927.0</td>\n",
       "      <td>11867.50</td>\n",
       "      <td>230752.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Motie</th>\n",
       "      <td>7639.0</td>\n",
       "      <td>901.439586</td>\n",
       "      <td>1680.205979</td>\n",
       "      <td>239.0</td>\n",
       "      <td>514.00</td>\n",
       "      <td>634.0</td>\n",
       "      <td>835.00</td>\n",
       "      <td>75020.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Onderzoeksrapport</th>\n",
       "      <td>1174.0</td>\n",
       "      <td>31677.848382</td>\n",
       "      <td>40610.644482</td>\n",
       "      <td>683.0</td>\n",
       "      <td>12640.00</td>\n",
       "      <td>22654.0</td>\n",
       "      <td>38087.50</td>\n",
       "      <td>618067.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Raadsadres</th>\n",
       "      <td>1621.0</td>\n",
       "      <td>2099.353486</td>\n",
       "      <td>3076.278696</td>\n",
       "      <td>108.0</td>\n",
       "      <td>737.00</td>\n",
       "      <td>1261.0</td>\n",
       "      <td>2256.00</td>\n",
       "      <td>35101.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Raadsnotulen</th>\n",
       "      <td>231.0</td>\n",
       "      <td>68092.839827</td>\n",
       "      <td>22624.293007</td>\n",
       "      <td>4606.0</td>\n",
       "      <td>54819.50</td>\n",
       "      <td>71582.0</td>\n",
       "      <td>84301.50</td>\n",
       "      <td>109188.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Schriftelijke Vraag</th>\n",
       "      <td>2932.0</td>\n",
       "      <td>3614.492838</td>\n",
       "      <td>12014.558830</td>\n",
       "      <td>497.0</td>\n",
       "      <td>1802.00</td>\n",
       "      <td>2474.0</td>\n",
       "      <td>3438.25</td>\n",
       "      <td>365774.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Voordracht</th>\n",
       "      <td>1982.0</td>\n",
       "      <td>1432.886983</td>\n",
       "      <td>1007.023550</td>\n",
       "      <td>376.0</td>\n",
       "      <td>918.00</td>\n",
       "      <td>1132.5</td>\n",
       "      <td>1535.25</td>\n",
       "      <td>11156.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                      count          mean           std     min       25%  \\\n",
       "label                                                                       \n",
       "Actualiteit           800.0   1629.756250   8662.494032   165.0    512.50   \n",
       "Agenda               2544.0   1999.538522  10121.983150    74.0    516.75   \n",
       "Besluit               625.0   2182.633600   3918.344740   227.0    387.00   \n",
       "Brief                1056.0   3137.704545   3082.126520   312.0   1363.00   \n",
       "Factsheet             214.0  12337.341121  26342.297934   261.0   2327.25   \n",
       "Motie                7639.0    901.439586   1680.205979   239.0    514.00   \n",
       "Onderzoeksrapport    1174.0  31677.848382  40610.644482   683.0  12640.00   \n",
       "Raadsadres           1621.0   2099.353486   3076.278696   108.0    737.00   \n",
       "Raadsnotulen          231.0  68092.839827  22624.293007  4606.0  54819.50   \n",
       "Schriftelijke Vraag  2932.0   3614.492838  12014.558830   497.0   1802.00   \n",
       "Voordracht           1982.0   1432.886983   1007.023550   376.0    918.00   \n",
       "\n",
       "                         50%       75%       max  \n",
       "label                                             \n",
       "Actualiteit            732.5   1305.00  234499.0  \n",
       "Agenda                1053.5   1918.00  314952.0  \n",
       "Besluit                979.0   2619.00   56288.0  \n",
       "Brief                 2297.0   3841.75   56355.0  \n",
       "Factsheet             5927.0  11867.50  230752.0  \n",
       "Motie                  634.0    835.00   75020.0  \n",
       "Onderzoeksrapport    22654.0  38087.50  618067.0  \n",
       "Raadsadres            1261.0   2256.00   35101.0  \n",
       "Raadsnotulen         71582.0  84301.50  109188.0  \n",
       "Schriftelijke Vraag   2474.0   3438.25  365774.0  \n",
       "Voordracht            1132.5   1535.25   11156.0  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "grouped_description = txtfiles.groupby('label')['count_LlamaTokens'].describe()\n",
    "display(grouped_description)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count     20818.000000\n",
       "mean       4340.207897\n",
       "std       15456.312555\n",
       "min          74.000000\n",
       "25%         612.000000\n",
       "50%        1031.000000\n",
       "75%        2378.000000\n",
       "max      618067.000000\n",
       "Name: count_LlamaTokens, dtype: float64"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "txtfiles['count_LlamaTokens'].describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.0027860505331924294\n"
     ]
    }
   ],
   "source": [
    "print(len(txtfiles.loc[txtfiles['count_LlamaTokens']<201])/len(txtfiles))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Fine-tuning Experiments\n",
    "- Accuracy scores for each model, and epoch\n",
    "- Runtime per doc for each model, and epoch"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "def average_runtime_per_doc(runtime):\n",
    "    return round(runtime/1100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "# incontext learing -> training on 0 epochs is the same as incontext learning\n",
    "epoch0_g = pd.read_pickle(f\"{cf.output_path}/predictionsFinal/in_context/GEITje/overview.pkl\")\n",
    "epoch0_g = epoch0_g.loc[epoch0_g['run_id']=='IC_GEITje-7B-chat-v2zeroshot_prompt_geitjeLlamaTokens200_0traintest_numEx0']\n",
    "epoch0_l = pd.read_pickle(f\"{cf.output_path}/predictionsFinal/in_context/Llama/overview.pkl\")\n",
    "epoch0_l = epoch0_l.loc[epoch0_l['run_id']=='IC_Llama-2-7b-chat-hfzeroshot_prompt_mistral_llamaLlamaTokens200_0traintest_numEx0']\n",
    "epoch0_m = pd.read_pickle(f\"{cf.output_path}/predictionsFinal/in_context/Mistral/overview.pkl\")\n",
    "epoch0_m = epoch0_m.loc[epoch0_m['run_id']=='IC_Mistral-7B-Instruct-v0.2zeroshot_prompt_mistral_llamaLlamaTokens200_0traintest_numEx0']\n",
    "epoch0 = pd.concat([epoch0_g, epoch0_l, epoch0_m])\n",
    "epoch0['base_model'] = ['GEITje', 'Llama', 'Mistral']\n",
    "epoch0['epoch'] = 0\n",
    "\n",
    "#fine-tuning\n",
    "epoch1 = pd.read_pickle(f\"{cf.output_path}/predictionsFinal/finetuning/1epochs/overview.pkl\")\n",
    "epoch1['base_model'] = ['Mistral', 'GEITje', 'Llama']\n",
    "epoch1['epoch'] = 1\n",
    "\n",
    "epoch2 = pd.read_pickle(f\"{cf.output_path}/predictionsFinal/finetuning/2epochs/overview.pkl\")\n",
    "epoch2 = epoch2.loc[epoch2['run_id']!= 'ORIGINALFT_AmsterdamDocClassificationMistral200T2Epochszeroshot_prompt_mistral_llamaLlamaTokens200_0traintest_numEx0']\n",
    "epoch2['epoch'] = 2\n",
    "epoch2['base_model'] = ['GEITje', 'Llama', 'Mistral']\n",
    "# display(epoch2)\n",
    "\n",
    "epoch3 = pd.read_pickle(f\"{cf.output_path}/predictionsFinal/finetuning/3epochs/overview.pkl\")\n",
    "epoch3 = epoch3.loc[epoch3['run_id']!='ORIGINALFT_AmsterdamDocClassificationMistral200T3Epochszeroshot_prompt_mistral_llamaLlamaTokens200_0traintest_numEx0']\n",
    "epoch3['base_model'] = ['Llama', 'GEITje', 'Mistral']\n",
    "epoch3['epoch'] = 3\n",
    "# display(epoch3)\n",
    "\n",
    "\n",
    "overview = pd.concat([epoch0, epoch1, epoch2, epoch3])\n",
    "overview['runtime_per_doc'] = overview['runtime'].apply(average_runtime_per_doc)\n",
    "# display(overview)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAbgAAAE7CAYAAACv7pTFAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjkuMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy80BEi2AAAACXBIWXMAAA9hAAAPYQGoP6dpAACH7klEQVR4nO2dd3gU1drAf7vpySZsGumF3q4QQFF6R6WIDUREAUEvAgHh2uggTS9esYBXr1dE7ycgWLAgCkiRIqJAQLqUEBJCSSGVJFvm+2PZyW6yaWQ3m909v+eZJ3POnDnzzmZn3znnvEUhSZKEQCAQCAROhtLeAggEAoFAYAuEghMIBAKBUyIUnEAgEAicEqHgBAKBQOCUCAUnEAgEAqdEKDiBQCAQOCVCwQkEAoHAKREKTiAQCAROibu9BagL9Ho9ly9fxt/fH4VCYW9xBE6IJEnk5eURGRmJUuna743ieRPYkpo8ay6h4C5fvkxMTIy9xRC4AJcuXSI6OtreYtgV8bwJ6oLqPGsuoeD8/f0BwwcSEBBgdkyj0bBlyxYGDBiAh4eHPcS7LYTcdU9lsufm5hITEyN/11wZ8bzVH5xR7po8ay6h4IzTJAEBARYfOF9fXwICAhzuCyDkrluqI7uYkhPPW33CmeWuzrPm2osFAoFAIHBahIITCAQCgVPiElOUAoFAYEv0ej0lJSX2FqMcGo0Gd3d3ioqK0Ol09hanzhEKTiAQCGpBSUkJFy5cQK/X21uUckiSRHh4OJcuXXKo9WFJkvD396e26UqFghPYnPxLKRRlZQCg1WrRJJ8n8+hh3N0NXz/voBBUMbH2FLFCHFn2+oIzf4aSJJGeno6bmxsxMTH1zgdSr9eTn5+PSqWqd7JVhCRJ5OfnU1xczLVr12rldiMUnMCm5F9KYcNdrdEVF5nVf2+y7+blzbDfT9S7HzlT2RUKCG+uIEit4OBaiStnJCSpVHYaqO0trhkrVqxg9erV/Pnnn9x///1s3LgRgGvXrjFt2jR27dpFbm4uTZo0YcGCBTzwwAPyuZcvX2b8+PHs2rWL4OBg5syZwzPPPHNbcjjy/786aLVaCgsLiYyMxNfX197ilMM4dert7e0wCg7Ay8uLoqIicnNz0el0uLm53VY/QsEJbEpRVka5H7ey6IqLKMrKqHc/cEbZ4zsquWekO6qg0ime/CyJ/Wu0JB80yO5ZzxRcZGQks2fPZtu2baSmpsr1+fn5tG/fntdff53IyEg2bdrEiBEj+P3332ndujUAjz/+OE2aNOHatWscO3aMe++9l+bNm9OzZ88ay+HI///qYFzX8vT0tLMkzofxM9VoNELBCRybo+/8C++gYPQ6Hej16PU6JOO+zrAvSXrDX+Omv1XWm+wb6/XGfQlJpyvXb3XO12s0xHdU0ndS+cfETw19J7nz80pt3X9Y1eDhhx8GICkpyUzBNW7cmBdeeEEuDxkyhBYtWrB//35at27NuXPn2LNnD+vXr8fPz4+7776bJ554glWrVt2WgnMVHGl9y1GwxmdqVwWn0WiYNm0an332GQqFgieeeILly5fLc/OmnDt3jsmTJ7N//358fX2ZOnUqL730kh2kFpRFV1xMftol8lOSybuYTF5KMvkpF8m7eIHc82fldsZpPh+1gps3Sqf5AM5/9bmdpK8YhQLuGel5a9/8YVMoFUh6iXsed6dIclzrtGvXrnHy5Enatm0LwNGjR4mIiCAsLExuk5CQwHvvvVdhH8XFxRQXF8vl3NxcwPB8a7XVewHQarVoNJrbuYU6wShbWRk1Gg2SJKHX6+utkYnxb0XyBQUF8dVXX9GrV68q++vTpw9Dhw5l6tSp1hSzHKZylx3B1eR7YlcFt2jRIvbs2cOJEycAuP/++1myZAlz5841a6fT6XjggQd48MEH+fbbbzl//jz9+/cnOjqakSNH2kN0l0Kv0VhWYCmG/cL0y1CFtVPl03x188Pg7gmevuDlp8DTFzx9FXj5Iu97+oKXb+kxVRBm8pZFoVSgCgZd0VGgRZ3cgzUpKSlhxIgRDB8+nDvvvBMwTGGq1Wqzdmq1mry8vAr7Wbp0KQsWLChXv2XLFjyuXamWLHv27MUjNb36wtuJrVu3mpXd3d0JDw8nPz/f5m4CgwcPZu/evXz99ddmyuidd95h3rx5TJgwgaVLl1o8t7L/nyRJFBYWyi8mlaHVauW1sbqgqKiIX375xexFqbCwsNrn21XBrVq1iuXLlxMREQHArFmzeOGFF8opuNOnT3P69GnmzZuHh4cHLVq0YNy4cfznP/9xKQVnK2s0vVZLweVUs5GXrMAuJlOYnoZUi7fT6kzzebcdj19kNEo3N1AqUbi5yftKpRsKNyVKNz1ubsW4uxXjpizGTVmEm1sRbsoilIqbuCluoqQQNwpRKgpRUohSKkAhFRj+YpuRlkKXaZN+bUlJSQmPPvoovr6+fPjhh3K9SqUiJyfHrG1OTk6lcf9mzJjB9OnT5bIxVuCAAQPQJJ8zMyipiG7duhLctn2N76Ou0Gg0bN26lf79+5uFjioqKuLSpUuoVCq8vb1tKoO7uzstWrRg/fr1ZkZB69ato2XLlnh6epYLjWaMvF9ZZgeFQiGHxaqODN7e3tVqWxskSSIzMxNvb2969Ohh9tnWRLnaTcFlZ2eTmppKQkKCXJeQkEBKSgo5OTk0aNBArjcOrU19IvR6PUePHrXYd2VTJpamGEz/1lfyU1PY2LktOpP7grLWaF48+OtRVNHmSk6v01GYnkb+pRQKbimu/EsXKbh00fA3LdWw3lULPPwDUMXGo4qJRRUbhyomDlVsPNqbBUQV/B2oYJpPkug21h1NhDu+oQWgyUGhyQHNDdAa93MM+/pK3pClW5sd0KJGquffH1NKSkoYNmwYJSUlfPPNN2YGEm3btuXy5ctcu3aNhg0bAoZ1vDvuuKPC/ry8vPDy8ipX7+HhgWRhucES7u7uDhEr0cPDw0xOnU6HQqFAqVTWiZXiiBEjeOedd8jLy6NBgwb89ttvANx9992yHH/88QdTp07l+PHjREZGMn36dJ5++mmUSiV6vZ558+bxn//8Bzc3N2bNmgVgJv+6detYsmQJKSkpNGvWjLfffpsuXbrIMhivY0uMv/kKhaLcZ16T74ndFFx+fj6A2XSIcd/4zzPSokUL4uPjmTt3Lq+++ipnz55l1apVFWryyqZMKjLlLTv1UN/QJJ+XlVtFa1m64mI2L5oHKNBfv4ou4zq6jGvoszKgtlEMvLxxC22IW4hhU4aE4hYSilfDAHxDvPHy0eBNHp7k4kUOXuzBU/8DQSRXPs2nUODtB965H0EdzHpo8UKDHxp8b21+FsqGv0VXs7nTezU+AZYXvCW9REE2bL+YhyZli+2FrwFarVbe9Ho9RUVFKJVKFAoFw4cPp6CggO+//76cYmrSpAldu3Zl5syZvPPOOxw7dozPPvtMdjMQ2Be1Ws19993H2rVrmTBhAqtWrWLs2LEcP34cgBs3bnDffffJU5Z79uyRDYm6d+/O6tWrWb16Nbt27SI2NpZJkyaZTV/+8MMPvPDCC3z77bckJCSwceNGhgwZwpkzZwgODrbXbd82dlNwKpUKMEx/hISEyPtAuekQDw8PvvnmG6ZNm0ZUVBTR0dGMHTuWDz74wGLflU2ZWIpubmnqob6RefQw31P1WlbhV+tuq393X1/8YmJRN45G3SiUBtFqVGF++AV54O2vwN2tAEVJJpRcR1FyFYqPQ0kmCqmKkUv5F/taISl9wKMBeDRAcm9Quu+hBnfj/q16d5NjHg3APQCUHrhj+OL7VHGt/NQU9j/3Cb3/LiHpJRTK0s9c0kuggAMbFPT994PoA9TWvdFasmjRIrOXPB8fH3r27MmCBQv45ptv8Pb2lp87gJkzZzJz5kwA1q5dy/jx4wkNDSUoKIh//vOft21B6R0UgpuXd6WuAm5e3ngHhVR4XGDO2LFjmT17NqNHj+bLL7/k2LFjvPLKKwBs2rSJ0NBQEhMTAejZsyePPvoon376Kd27d+ezzz4jMTGRli1bAvDaa6+xevVque+VK1fy4osv0qFDB8Bgjfuvf/2LH374gSeffLJub9QK2E3BBQYGEh0dTVJSEk2aNAEMUyExMTFmozcjbdq0YcuW0rfkl19+ucKHrrIpk4qUWGXH6gPu7u7VWssqa7Dh7gU+AeAb7EVg4zAaRKvxD/fDN9gLH3/w9NHi4VaIQpeNoiQN9BfMO79xa7M1YQMhoNUthaWWFZe876kG9wYo3Er/r7Y2zA5s1IROH54k79KX+GW+jZvuunxM7xFGQfAUOn34CKqY2DpbdK8u8+fPZ/78+RaPVRX+KCoqis2bN1tFDlVMLMN+PyGvHe+ZNomMw78DkPDCLOIHD3XoSCb2oG/fvowbN46FCxfSuXNnwsPD5WOpqanEx8ebtY+Li+PAgQOAwYk/Li5OPhYWFmb2W5mcnMzMmTOZN2+eXKfRaEhLS7PR3dgWuxqZjB07lsWLF9O1a1cAlixZwvjx4y22PXr0KE2aNMHDw4Pvv/+eVatW8fPPP9eluPZF0nHPSMO/q6K1rJ7PuHPHEH98g7zw8tHi7laIUmE6wrp6azOhch/caqAAz2DwCgHPUMNfr1DD5hliOHZ0OpRkVHy+TxTcsw4Ut+fMaUtUMbEQMw2kKWiv7iJp/08k3HMv7mE9CaiH8tZHVDGxsgIL79ZDVnBFmRmEtOtgT9EcEqVSyejRo1m8eDFffPGF2bHo6GiSk5PN6lJSUuRwV5GRkVy8eFE+du3aNTN7hZiYGBITE5kwYYLtbqAOsauCmzNnDpmZmbRq1QqAUaNGydMkxg/4/fffB2D9+vX8+9//pqioiHbt2rFx40bZd8cV8Cg6WuValocXhMXlA/m1vFhQqaLyNCqsisrBVSsmpTsceOJWwXT0cOt+7vhnvVRuZijckIK7k6bIo11w9/ovbz0luF2ppWTmkYN2lMSxmTZtGj179iw3izVw4ECmTJnCe++9x7PPPsvevXv54osv2LRpE2CIUrNw4UKGDh1KbGwsM2bMMDMYmTRpEs8//zx33XUXHTp04ObNm+zbt4+WLVvWKiakvbCrgvPw8GDlypWsXLmy3DGjYjOyaNEiFi1aVFei1TtqZYruoa6GojJRWEorT9VGDoVOn8HRF6HIZKrDJ8qg3CKHWvd6gnpLcLuO8n7W8T/RazQo6/HSQH0lKCiIfv36lasPDAxk8+bNPP/888yYMYPIyEjeeOMNunXrBsDTTz/NhQsX6N69u2xF+eWXX8rnDxkyhKKiIp555hnOnz+Pl5cXnTp1svgb7QiIUF0OguRWPQumwgZP4ttm2C2F1RC8gkFZD+LkRQ6FiMHlpvnESMi1UMXGofBTIRXkoysuJvvUcYLvSLC3WA7Bzp07KzxmaijSqVMn9u3bBxjM7U3Xh5VKJYsXL2bx4sVy3eTJk836GjZsGMOGDauxDPURxwkv7eK4RfahIKtiAwFJL5GfBfomL0PDvqBuBz4R9UO5GZGn+bogiWk+l0ShUOAe31guXz8spikFtkMoOAdBFdsI5Z1vY8l2UAJQKnC7821UsY3qWjSBoEa4x5UquIykQ3aURODsCAXnQEhhgynIKh8yS+ETjaLTGnzaWrZAFQjqEx6Nmsr7wtBEYEvEGpwDkbHzv8QGG95J9Do46jaOv93zoFjLEjgUplOUWcf/RFdSgpvIpyawAWIE50AoLq+X92/kNeaioq9YyxI4HMqQhniqAwFkQxOBwBYIBecgSNpiQkPPy+UdhT14aOt1Fm85YkepXIvFW46Iz9wKKBQKs8wBYh1OYCuEgnMQ8n7/D94qgwVlUQE8dawlErBgcxILNx+2r3A1wFGVxMLNh1mwOckhP/P6iKnDt1BwAlshFJyDoD/3ibz/+7EGFCpLQwXP3XTQIX5wHVVJLNx8mLmbzI0hHOUzr68EJ5SG6BKGJgJbIRScI1CSjb/PKbn4+aV25ZrU9x9cR1USluQ24gjy11eC25YqOKOhiauycPNhlJP/K75LNkAoOAdAe34tbm6G6cmcK3q+UHa12G7upoO8svEANwqLKSzRotHpq4wcXxc4gpKQJIkSrY6CYg3ZhcVcy7vJi1//VqHcRuqL/I6GKjYOL2FoIj8bEs71XWrTpg3ff1+dXO62RbgJOABXDq4g2s+wf+QPb676h1XY9vVtR3l9m3mmcw83pbx5Gv+6K/FQGvfdbh1X4OnmZtauoraeFvqU27op5X6+OXqRzw+dr0BaA3M3HeR4ejZD28ZRotWh0evR6PSUaG/91Rn+GvZ1aHQGZVRaZ6Gd3E9pW/N+TNrp9Gj1t/8iMHfTQYoKaxng2sVQKBQEJ3Tg8k5DRpCMpEOEmBieuAIVzWoAzLnf+p/Fnj17WLJkCfv370er1RIeHs69997LP/7xD+Lj41EoFPj4+JgFX+7fvz9ff/01ycnJNGrUiOzsbB5//HF2794NGLLD6/V6vL295XPy8/PlBKz2Rii4+k7BeaL9StNbbExuBS1rlgnN+KNen/n80PkqFWF9ZslPSfYWweEIaWeu4HhqnJ0lqjuqmtUA6yq57777jlGjRrFw4UJWr15Nw4YNSU9PZ926dezYsYOxY8cCsG/fPhISEirtyzRX4Pz580lKSqq3Gd/FFGV9J2WtvJt+Ss9OP9d6y7UXbsqavUTMvDfBNoI4MaHt75T3ncXQRJIkijW6Srd5mw5Wa+p73qaDVfZVrNFVuQwhSRLPP/88M2fO5Pnnn6dhw4YAREREMG3aNFm5WZP4+Hgzpbdt2zY6deqEWq2mTZs2fPvtt1a/piXECK4+I0noL3wqv4Wc/lXPqbCWlZ7y6qCOzL4vAZ1eqmB6r7LpP/Npu0qn/3R6NPrK+zmRns1f16uf6bpZaACtIwIrnAKt7lRp6ZRrJf24mU+lmrZzUypQKBSVvmWX/cyndm3CkmrfqQAgJKG8oYmjRzQp0erxnvaxVfp6dfNhXq3GmlzR8rF4eVQc7OHs2bMkJyfz2GOPWUWumnL06FGGDRvGl19+Sa9evdi3bx+DBg3iwIEDtGjRwqbXFgquPpO1H2VJKgDaEomz2a24GeNTYfNXB3WUpzXc3RS4uympuHXdUBMlYYt1h9pglKcy+Y1ym6YkEVQPVWw8XupAim9ky4YmrrYOVxdkZhpySUZGRsp1CxYsYPny5Wi1WgYOHMj69YYoScY8cUamTZvGvHnzanX9Dz74gDFjxtCnTx8AunXrxuDBg1m/fj1z5sypVd9VIaYo6zOX1si7Fw/r6TjkUab2amOxaX1UEGBQEq8O6lhpm/oqO1Quf32W2xEwGpoYEQ7ftiE42JBL8vLly3LdvHnzuHHjBi+88AIlJi4au3fv5saNG/JWW+UGkJyczPvvv49arZa3b775xkweWyFGcPUVXRFS6hdycpy/9unp8OYA7tQFlmta339oKxsJ1XfZwbL8jiC3IxCa0NGpDE083ZUULa98TWvJlqRqTT3Ovb89MwckVOualdG0aVPi4uJYv349r7zySpX9WZuYmBimTp3Ka6+9VufXFgquvnLlBxRaw7TXzRyJjLQGhCR05Ncv9ps1m3d/gkP80Dq6kphzf3t0eh2vbk5iroN85o5ASELp6NgZDE0UCkWl62EACwZ1xF2prNbUt7VkWr58OU899RS+vr6MGDGChg0bcv369Tox5//73//Offfdx7333kuPHj3QarUcOnQItVpNq1atbHptMUVZX7lUaj15dr+OyB79Ubq58euFa3L9S20DmDWgfFST+sqc+9sz7/4EFDiOYjZl1oB2fN0/1KE+8/qOqaFJ5rGjLhPRpK6nvocOHcqmTZv44YcfaN68OQEBAXTv3p2GDRuyfPlyuV2XLl1QqVTy1qlTp1pfu3379qxdu5bZs2cTGhpKVFQUc+bMobi4uNZ9V4UYwdVHiq/B1S1y8ew+PW1eGkBBsYajl7Pk+hYNHO/fN2tAO9pr0xgolESlFBUVmTnPOiumhib6khKXMjSp61mNHj160KNHjwqPV+ZuEB8fb/H4/PnzLbbX6XR4mljE9unTRzYyqUvECK4+kvoFSFoAslL1ZKZIRPXuzx8pGehuRdyIUfsR7C3ywDkL165dY+nSpXTu3BkfHx/8/Pzw8fGhc+fOLF68mKtXr9pbRJvg6oYmxpGcAseasq+M9PR0rl27RpMmTewtilBw9RIT68m/9uoJbPU3/CKj2J9cOj3ZKS7EHpIJbMDs2bNp27YtZ86cYdKkSezdu5fTp0+zd+9eJk2axLlz52jXrp3NTartRajJOpyrKTgwKDn9ivFOody2bdtG69atmTx5ss193KqD481xOTu5J+GGwcJKr5c4t19H4yf6AZitv90THwqFFy12IXAsfH19OXfuHH5+fuWOdejQgVGjRlFQUMA777xjB+lsj7MZmrgy/fr1Izs7295iyIgRXH3DxLjk8gmJwhsQ3WcAkiSx30TB3R0fagfhBLZg5syZFpWbKX5+fsyYMaOOJKpbXNXQRGB7hIKrT0g6SF0nF8/u0+Hm7U145+4kZ+ZzNe8mYPB7aR8dbC8pBTbk4MGDpKWlAYYIFFOnTuUf//iHU0dKMRqaALKhiUBgDYSCq09k/AI3DT9umiKJ5IN6Irr2wN3Hh18vlBoZtI8OxstdGJg4I2PHjuXmTcOLzIsvvsixY8c4deoUzz33nJ0lsx0KhYKQ9q69DiewDWINrj5hkjngwh96tCUQ1WcAgJmBSedGFeeDEzg2KSkpNG3aFEmS+Pbbbzl58iQ+Pj71wiLNloS060Dajm2Ac0Q0EdQPhIKrL2gL4PJGuXh2nw4wrL+BuYFJ50YN61Q0Qd3h4eFBXl4ex48fJzo6mtDQUHQ6nTyqc1ZCzCwphaGJwDoIBVdfSP8WdAUA5GdJpJ+S8IuMRt2iFTdLtCSlZspN74kXCs5ZGTJkCH369CEvL48xY8YAcOzYMaKjo+0rmI0xT51z1ClS51RG/qUUirIyKjzuHRSCKibW5nLEx8fz1ltv8eCDD9r8WvZAKLj6Qkqp79vZX3VIEkT1HYBCoeDgpQy0txy8Ixv4EhPoh1artZekAhvy/vvv88knn+Dh4cGTTz4JQFZWFnPnzrWzZLbFlSKa5F9KYcNdrdEVF1XYxs3Lm2G/n7CakuvTpw8PPvggzz//vFX6cxSEgqsP3LwM13fIxbP79ABE9+4PYOYecE98QxSKmmWbFjgOnp6ePPPMM2Z1vXv3tpM0dYfR0MR0Hc5ZFVxRVkalyg1AV1xEUVZGnYzinBmh4OoDqZ8DhhHa9Qt6blyWUCiVRPbqC4j1N2fn6aefrla7VatW2VgS++IMhiaSJKGvwo+vun5+upISdNUISKz09LTKS29KSgrjxo0jKSkJrVZLly5dWLlyJfHx8QCMGTMGNzc3cnJy+PHHH4mLi+Pzzz9n7969LFy4kOLiYhYsWMDEiRMBOHz4MImJiZw4cQI3Nzf69evHihUr5Px0dYFwE7A3kgQpn8lF4+gtpMOdeAcGIUmSeQQToeCcjgYNGsibUqlkzZo1XLlyBS8vL65evcratWvNsixXxYoVK7jzzjvx8vIqt7aSm5vLyJEjCQgIICwsjIULF9bouC1xBkMTfUkJH4f7Vbp9N6Bbtfr6bkC3Kvv6ONyvSoVabdn1eqZPn86lS5e4ePEivr6+5WYTNmzYwLRp07hx4wZ33XUXQ4cO5dy5c5w/f55169Yxbdo0OW6qUqnktdde4+rVqxw7doy0tLQ6z0cnRnD2JucI5J0EQK9XcO43c+vJS9kFpOcWAuCuVNAxRsSgdDZM05U8+uijfPHFFwwePFiu27RpU41Gb5GRkcyePZtt27aRmppqdiwxMZGsrCxSUlK4du0a/fr1Iy4ujqeeeqpax22Jqxma1Dfi4+Pl0Zq3tzezZs3innvuQa/Xo1QaxkKDBg2ia9euAAwfPpz//e9/LFiwAE9PT/r27UuDBg34888/CQsLo1270owhYWFhTJ8+nRdffLFO70mM4OyNSWDltBMKivIM+6XuAaUO3gnRwfh4incSZ2bLli0MHDjQrO6+++5j69at1e7j4Ycf5sEHHyQkxPxlqLCwkHXr1rFo0SLUajXNmzcnMTGRjz76qFrHbY2IaGJfrl+/zsiRI4mJiSEgIIAePXpQXFxMXl6e3CYsrNQH19fXF39/f3x8fMzq8vPzATh79ixDhw4lMjKSgIAARo0aRUZGxZajtkD8WtoTvRZS18vF07sM8+2eAQ0I7WhINLg/+bp8XKy/OT/x8fGsWrWK8ePHy3WrV68mLi6u1n2fPn2akpISEhIS5LqEhASWLFlSreMVUVxcbJa80hhWTKPRoNFozNoay2XrjQS160D6rp8BuHrwdxq0+lv1bs7GVCS3RqMxrLvp9ej1enB3Z/TlPEtdyGQeTeL7+7pXec3BP+4muG1C1cK5uxuubQFjDjdJkmQ5y2KU/ZVXXqGgoIA//viD0NBQkpKS6NixIzqdDr1eX66Psn/L9jdhwgSaNWvGsWPHUKvVbNy4kaeffrpCWSuSW6PRmE3RV/TdsYRQcPbk2jYoNigwrc6LS0mGH4nIXn1Ruhv+NaYjOBHBxPlZsWIFQ4cO5a233iIuLo6LFy+SlpbGN998U+u+8/Pz8fPzw9299LFXq9XyG3pVxyti6dKlLFiwoFz9li1b8PX1tXhORSPSfFWAvH/4u42cDapf3/mycru7uxMeHk5+fj4l1VwLu6mpfruC6mS9rkYbnU5Hfn4+166VrucrFAr0ej03b94kNzeXrKwsPDw8UCqVJCcny64peXl5KJVKNBoNJSUl8gtMYWEhkiSZxUk17S87O1tOenrixAlef/31cu2roqioiF9++cXMLaqwsLDa5wsFZ09MMgeknlKh0xr+8Ub3gGKNjsPCwdul6NGjBxcuXOC7774jPT2dyMhIBg0aRGBgYK37VqlUFBYWotVqZSWWk5ODv79/tY5XxIwZM5g+fbpczs3NJSYmhgEDBhAQEGDWVqPRsHXrVvr374+Hh0e5vpJ1Reza9DUAftkZ5aZr7UVFchcVFXHp0iVUKlW1M7ArY+Nw8/Ku0g8uODYOVZnPr6ZIkkReXh5ubm7MnTvXzJ8yLi4OpVKJj48PAQEBLFq0iLFjx9KoUSOio6N5/vnn2bRpE/7+/gQEBODh4YGnp6f8P/X19UWhUJj9j037W758Oc899xwfffQRzZs354knnmDhwoXlvhMVyZ2ZmYm3tzc9evQw+2xroiCFgrMXJTcg/Tu5+Oe3pSM1Y/zJQ5cyKNEahvNh/j7EB6vqVESBfVCr1bKTtzVp0aIFHh4eHDlyhI4dDRaLSUlJ3HHHHdU6XhFeXl54eXmVq/fw8LCoxCo7Fn5rah4g+8QxlJJUrwxNysqt0+lQKBQolUrZEKMqAmLjGfb7iTqJZGKcDtyxY0eV8rVp04YDBw6Y1U2YMEHe/+STT8yO9enThxs3bpjVJScny/s9evTg+HHzddR//OMfNZJboVCU+8wr+k5ZQig4e3F5I+gNUwsaQrl6xmDt1qBZC/xjDestpgGW72kkHLxdgby8PJYvX87BgwfLTQ1u3769Wn1otVp50+v1FBUVoVQq8fX15bHHHmPOnDmsXbuWa9eu8e6778quAFUdrwtUsfF4BQZRnJ3l1BFNVDGxwom7DhAKzl6YJja9EA4YFJzRehKEg7cr8uSTT3Lp0iUeeeSRKpOgVsSiRYvM1sR8fHzo2bMnO3fuZMWKFfz9738nOjoaHx8fJk+ebOYCUNVxW6NQKAhJ6OASEU0EtkcoOHtQkAyZe+Ti0a/T5P3oPv3l/V/LhOgSOD87d+4kJSWlWusUFTF//nzmz59v8VhAQABr1661eKw6x+sCs4gmhw86ZEQTQf1A+MHZg0ulWbu1Pu24+mc6YAi5E961JwCp2QWk3jBkF3BTKrgzVjh4uwIxMTE1MoN2RswimhwRyU8Ft4/dFJxGo2Hy5MkEBgYSFBREYmJihRHy09LSePDBBwkODiYkJIThw4dz/fp1i23rPZJk5tx9Nb2RvB/euRset6alTNff2kUF4edV/YVVgeMyefJkHnvsMbZu3crRo0fNNlfBUkQTgeB2sJuCW7RoEXv27OHEiRMcP36c3bt3V+hQOmnSJAAuXrzIhQsXKCoqYsqUKXUprvXI/h0Kzhn2lZ6c3FKqqE3X38pmEBC4Bs899xzbt2/n3nvvJSEhQd7at3eddSijoQmIiCaC2mE3Bbdq1Spmz55NREQEERERzJo1q8KQQOfPn2f48OGoVCr8/f157LHH+PPPP+tYYithMnrTh97HpZ2/yuWo3ibrb8mmDt5CwbkKxigQZTedTmdv0eoMo6GJkYwkMU0puD3sYmSSnZ1NampquZBAKSkp5OTk0KBBA7P206dPZ8OGDQwaNAhJkli7di1DhgypsH9rhg6yKrpi3FM3YDT2z7jRGt3NDQD4NAzHv0UrQ7QArY6DKaU+Mh2jg+wrtxVxVLmhctmtfT9XrlwhNTWVmJgYs/h/roIwNBFYA7soOGMwTrVaLdcZ9/Py8sopuK5du/Lhhx/K0Rw6d+7MjBkzKuzfmqGDrEmEdIBO3ACgGH+2/a80JYi+WUs2b94MwJkcDcW3HLwDPBScPrCbMxX4wNWF3LbAUeUGy7LXJHxQZWRkZPDEE0+wdetWvLy8KCkpoX///vzvf/8jNDTUKtdwBIShSd2TkpJC69atSUtLK/cbbAtWr17NW2+9RVJSks2uYRcFp1IZInLk5OTIEc9zcnIAyoUF0uv19O/fn+HDh8s/LPPnz2fAgAHs37/fYv/WDB1kTdz++D+4NfPoHj8Sr+TtGH8WO40aTeNbYYnO7ToB/A5At2ZRDBrUt1xfdSm3NXFUuaFy2WsSPqgyEhMTCQoKIi0tjYiICK5cucL06dOZPHkyn3/+uVWu4Qi4TOocSQcZe6H4CniFQ0hXUFQ/91916dOnD7t27WLr1q3069dPrl+2bBkvvfQSU6dO5a233pIHH5WRnJxMo0aNyM7ONhuk1EfsouACAwOJjo4mKSmJJk2aAIaQQDExMeXeHLKysrh48SJTpkyRR1+JiYksW7aMjIyMcilBwLqhg6xGcQZcK33zL/G7l+yT7xgKCgWx/e6Tr//7pdL4k12bhFcql83lthGOKjdYlt1a97J9+3YuXLggf9fDw8P5z3/+Q+PGja3Sv6PgEhFNLn8DR1+EolI/WLyjoO0yiBxq9cu1aNGCjz/+2EzBffzxx7Rs2dLq19JoNPXi+babkcnYsWNZvHgxV65c4cqVKyxZssQsRYiRkJAQmjZtysqVKykqKqKoqIiVK1cSHR1tUbnVW9K+BOnWOo1/Sy79ni4fCmnXAZ+Q0ukn4eDtunh7e5OdnW1Wd+PGDYsvbM6M0xuaXP4GDjxhrtwAii4b6i/XPntEWUaMGMHmzZvl2bLffvsNgLvvvhswjMwUCoUcX3Lr1q20bdsWf39/wsLCeO655wDo1MkQLzQ6OhqVSsVnn33Gzp07UavV/Pvf/yY2NpYuXboAMGrUKDkfXMeOHdmxY4fV76sy7Kbg5syZQ+fOnWnVqhWtWrWia9euzJw5EzAE+DQN8vnNN99w6NAhoqKiiIiI4MCBA3z77bf2Ev32MLGeJOZxUn8uHc1FmUQvSc8p5GKWYZpAqVDQKd511l0EMHLkSAYOHMgXX3zB77//zoYNGxgyZAhPPPGEvUWrc0LamSi4wwcraVmPkCTQFVe+aQvh6AuAZKkDw5+jLxraVdWXrthwzWqgVqu577775Eg1q1atYuzYsRW2Hz16NC+++CJ5eXmcP39eDgBuDMicmppKfn6+/N3My8vjyJEjnDp1il27dgHQt29fTp48SWZmJiNGjODRRx+tMv2SNbFbqC4PDw9WrlzJypUryx17//33zcqtW7fmp59+qivRrE/eGcj+41ZBgT5yGGk7l8mHTcNzmTp43xEZiEo4eLsUCxcuxMfHh5kzZ5Kamkp0dDSjRo2q1KjKWXFIQxN9CXwXVMtOJMPI7vtqvtwOyQK36o3wx44dy+zZsxk9ejRffvklx44d45VXXrHY1sPDg7Nnz3L9+nVCQ0PlUVlF6PV6XnvtNTNDPlMF+uKLL7JkyRKOHj1K165dqyVvbRGhuuoCk8DKhPQg88w1irMM62weKhUN7+osHxYO3q6Nu7s7c+fO5cyZMxQWFnLmzBnmzp1bL9Yz6pqQ9qUKTkQ0sQ59+/YlPT2dhQsX0rlzZ8LDwyts+/XXX3Ps2DFatGhB+/btWb9+faV9+/v7mxmd6PV6Zs2aRbNmzQgICECtVpOTk0NGRsVpgqyNUHC2RtKbK7jYkaRu3yIXI7r3NrMOExkEXJstW7Zw+vRps7pTp045tFvF7aKKiRMRTayMUqlk9OjRvPbaa5VOTwJ06NCBL7/8koyMDObMmcPIkSO5evVqhXnlytavWbOGNWvWsGnTJnJycrhx4wYNGjRAquaUqjUQ2QRsTcYeuHnJsO/mCxFDSdte6qRuGp5Lo9PzR0pp6K57hIJzORITE/n555/N6vz8/EhMTOTUqVN2kso+lEudc/hg/bekVHoapgwrQ9LBtnZQlI7ldTgFeEdCv6TquQwoa+Y+MW3aNHr27EnPnj0rbFNSUsLnn3/O4MGDCQwMlEdm7u7uBAQEoFQqOXfunJwY1xK5ubl4enoSEhJCSUkJr7/+ep2uv4EYwdkeU+OSiAcouSlx9fdS/73ovqUK7mhaFjc1hpBMgb5eNG9oe2dLQf0iPT2d6Ohos7qYmBjS0tIqOMO5MTM0cQRLSoXCsB5W2ebuC23fMJ5QtgPDn7bLDO2q6svNy3DNGhAUFES/fv2qnPZes2YNTZs2xd/fn8TERNasWUNwcDA+Pj7MmzeP+++/H7VazZo1ayyeP3r0aNq0aUNcXByNGzfGx8en3Hfb1ogRnC3RFsLlr0vLsY9zefcOpFtZEwIaNSGgURP58K8XSuNP3hMfKjJ4uyBRUVEkJSWZhbE7cuQIkZGR9hPKjjikoUl1iBwKnT4r7wfnEwV3/NPqfnDbt2+vcGpx9erV8r7p9KExspIl5s6dy9y5c83qjO4FRvz8/Pjqq6/M6l588UV5f8yYMYwZM6YKyWuHUHC25Mr3oL0VGcA7HEJ7k/pzaRaEKJPpSTC3oOzcyPXiDwoM2QQeffRRFixYQNOmTTl79iwLFiyQfZBcDUuGJk4T0SRyKEQMrpNIJq6KUHC2JMVk6B49AgklaSYGJtFlFJy5g7fwf3NFEhMTAVi8eDEXL14kPj6eSZMmOW56qFpiNDRx2ogmCjcI7WFvKZwWoeBsRVE6XDMxFogdSe75s+RdvACAwt2diG6li7zX8m5yPsOwAKtQwN3CRcAlUSgUTJkyxWUVWlkc0tBEUG8QRia24tJ6wJARgAZtIaANqT+Xjt7COnXG0yTws6n/W5vwQAJ8nGQaRlBj/vrrLxYvXszkyZMBOH36tEtl9C6LwxmaCOoNQsHZClPft5iRAGb+b2WnJ03X34R7gOvy3Xffcdddd3Hy5Ek+/fRTwJA/8YUXXrCzZPYjpP2d8r5TGZoIbI6YorQFOUch91bGcYUbRA9DV1JC+p6dchNT9wAQDt4CAzNnzuSHH36gS5cucv7D9u3b2zRnVn3HEVLn1KXzsqtgjc9UKDhbYDp6a9gPvMO5umcn2oICALxDQgk2WUfQ6vQcuGji4C3W31yW1NRUOeaf0U3Ew8MDnU5nT7HsSn02NPHw8EChUMjxGuuba49er6ekpISioqIK3QTqG5IkUVxczPXr13F3d8ezFi8zQsFZG70WLpkkprw1PZlmsv4W1asfCpMv27H0bApLDL5xDXw8aRmmrhNRBfWP5s2bs2vXLrMoE7/88gutWrWyo1T2pT4bmri5uREdHU1qairJycn2FqcckiRx8+ZNfHx86p3yrQxJksjKyuLOO++slWIWCs7aXN8Bxca03QEQMQiA1O2lsQTLT0+WOnjfHReKUuk4X0SBdVm8eDEPPfQQY8eOpbi4mJkzZ/Lxxx/LKU5clZB2Jgou6RCMtrNAJqhUKpo1a4ZGo7G3KOXQaDT88ssv9OjRw6ECdkuSxF9//VVrmYWCszamobmiHgI3HwqvXSXz6OHS6l79zE7Zf6F0elKsv7k2/fr1Y+fOnfznP/+hd+/eZGdns3nzZrPIJq5IfTc0cXNzw82t/jlou7m5odVq8fb2digFZ62XBaHgrIkmF9K/Ky0bpydvvXkCBLVpi294hNlppiM4EcFE0LZtW1asWGFvMeoVjmBoIqh/OMaqo6NweSPobhr2feMg2GAsYBq9xDR7N0BmfhF/Xc+VyyKDt2uzatUqjh07BsDRo0dp164dHTt25M8//7SzZPalXOqck8fsLJHAERAKzpqYWk9GjwCFEkmvJ3VHxetvpv5vrcLVBPpWLzOvwDlZtGgRYWGGUfxLL73Evffey9ChQ10+sonR0MSIcPgWVAcxRWktClMg45fScuzjgGE6pei6QYm5+/oSfk83s9PMHLyFe4DLk5GRQWhoKEVFRfz666988803uLu789Zbb9lbNLtTnw1NBPUToeCshalrQGAnUDUDMAvPFdG1J25e5iM04eAtMCUwMJC//vqLY8eO0bFjR7y8vCgqKkKv19tbNLtT3w1NBPUPoeCsgSSZW0/GjpR3TcNzRZWZntTp9fyWLBy8BaVMnTpVtpj85JNPANizZ49L+8EZEYYmgpoiFJw1uHEQ8s8Y9hUeEPUwAJr8fK7u3ys3Kxt/8kT6DfKLDeaw/t4etI5Q14m4gvrL9OnTGTJkCG5ubjRu3BiA2NhYPvzwQztLZn/KRTQ5ecwsELNAUBZhZGINUkyMS8LvB89gANL37ER/y59DFR1Lg6bNzU4znZ7sFBeKm4OE0hHYlmbNmsnKDQzRTf72t7/ZUaL6gTA0EdQU8YtaW/QlkLahtBxjOj1Zaj0Z1XdAuVA55hm8xfSkQFAVIQmlGb6FghNUhVBwteXqFijJNOx7BEH4vfKhytLjgHDwFghqipmCE4YmgioQCq62pJgYl0QPA6Vh0Tvv4gVyz/0FgMLNjciefcxOyy4s5tTVHLl8d5xw8BbUHWlpaTz44IMEBwcTEhLC8OHDuX7dYPCk0WiYPHkygYGBBAUFkZiYiFartbPEBiwZmggEFVFjBdemTRvefPNNrl27VnVjZ6ckC678UFqOeVzeNZ2eDO3YCa8GarNTfzOZnmzesAHBKm+biSkQlGXSpEkAXLx4kQsXLlBUVCQ7ky9atIg9e/Zw4sQJjh8/zu7du1myZIk9xZUREU0ENaHGCm7OnDls2bKF+Ph4HnzwQb755hvXzVWV9hVIt4KCqppBYKmfjtn0ZF9L05PCwVtQnuPHj9OnTx8aNGggB/BVKpVWD+R7/vx5hg8fjkqlwt/fn8cee0wOB7Zq1Spmz55NREQEERERzJo1i48++siq179dhKGJoCbU2E1gxIgRjBgxgrS0ND799FNeeeUV/v73v/Pkk08yduxYWrdubQs56yemvm8xI+GWEYleo+Hyru3yIUvrb/uFg7fAAqNHj6Zjx44sWrQIPz8/m11n+vTpbNiwgUGDBiFJEmvXrmXIkCFkZ2eTmppqlr0gISGBlJQUcnJyaNCgQbm+iouLKS4ulsu5uYbYqhqNplxUeGO5NtHig9q2lyOaXDv0B01G2j6kiTXktgfOKHdN7uW2/eCioqKYMWMGXbp04fnnn+fNN9/k/fffp0OHDrz99tvOn94j/yxk/VZajhkh71774zc0eYaH3EsdaBaBAUCvl/jNNIO3UHCCW5w5c4YDBw7YPPty165d+fDDDwkMDASgc+fOzJgxg5wcw7qwWq2W2xr38/LyLCq4pUuXsmDBgnL1W7ZswdfX1+L1t27darG+OhRrJXn//O4dZP/wQyWtrUtt5LYnziR3YWFhtc+/LQV38eJFPvnkEz799FO0Wi1jxoxh48aNNGzYkHfeeYdHHnmEc+fO3U7XjoNpYOXg7uAbKxdNpycje/VFWWZ66dTVG+TcNCyO+3m687eIQNvKKnAYunTpwokTJ2zq96bX6+nfvz/Dhw+Xf0Dmz5/PgAED2Lx5MwA5OTmEhITI+wD+/v4W+5sxYwbTp0+Xy7m5ucTExDBgwAACAgLM2mo0GrZu3Ur//v1vOz9Z/t/a8OXKNwz3knaJe/v1s3lEE2vIbQ+cUW7jDEF1qLGC6927NwcOHOCBBx7gvffeo3///mb+XS+//DKLFy+uabeOhaQ3V3Cxj5sdTqvSPaB0evKuuFDc3YQxq8BAhw4dGDhwIKNGjSI8PNzsmLUyCmRlZXHx4kWmTJkij7ASExNZtmwZOp2O6OhokpKSaNKkCQBJSUnExMRYHL0BeHl54eVVPguGh4dHhT+qlR2rCnXjJmYRTfLPnq6ziCa1kdueOJPcNbmPGiu4hx56iC+//JKgoKAK29y4caOm3ToWmb9C4UXDvtIbIh+SDxVlZnD98EG5bHH9TTh4Cyrg119/pUmTJvz6669m9QqFwmoKLiQkhKZNm7Jy5UrmzZsHwMqVK4mOjiYkJISxY8eyePFiunbtCsCSJUsYP368Va5tDYyGJqaZBUTILoElaqzgHnnkkXJ12dnZ3Lx5k8jISACbrx/YHVPjksgh4FE6DZO2a7sh+DKgbtkav6jocqeLDAKCitixY0edXOebb75h2rRpREVFodfrad++Pd9++y1gsJTOzMyUAzyPGjWKmTNn1olc1SUkoaNInSOokhoruIcffpj//Oc/ZiO4ixcv8txzz5V763RKdDcN7gFGYiqbnjTP3g2Qc7OEE1ey5fLdwkVAYAdat27NTz/9ZPGYh4cHK1euZOXKlXUsVfUREU0E1aHGCu7UqVO0a9fOrK5du3acPHnSakLVa9I3gfbWIqdXQwjtKx+SJKnK8FwHkq8ZB3g0CQmgob+PTcUV1H9iY2NJSUkBDPngysYsNZKVlVWXYtVrROocQXWosYILDAzk6tWrhIWVxk68evUqKpXKqoLVW0yNS6IfA2XpR5h98jiF6ZcBcPPyIrxLj3Kn/2qawbuRCM8lgDVrSqe8N27caD9BHAiROkdQHWqs4AYPHszYsWP58MMPiYqKIi0tjQkTJvDAAw/YQr76RdFVuGbil2GS2BTM3QPCu3TH3af86MzMwTteBFgWQLdu3eT9nj172lESx0GhUBDSviNpt0LiCUMTgSVqbA2ydOlSvL29iYmJwdfXl9jYWDw8PHj99ddtIV/9InUDSLfCkgW0gQZtzQ6brr9FWZie1Osl9icLB2+BOe+//36VwYx1Oh0ffPBBHUnkGJgqNBGyS2CJGo/g/P39+eqrr7h69SopKSnExcXRsKGL/FCXDc1lgrawkCv7dstlS+tvf13PIbvQENLIx8ONtlEVu1oIXIcjR47QpEkTHnvsMfr06UPLli0JCAggNzeXU6dOsWPHDtatW8fAgQPtLWq9wjw33MFKWgpcldsO1RUWFma2Duf05B6HnCO3CkrD+psJ6ft+QXcrHp9vZBSBrdqU66Ksg7eHcPAWAP/+9785e/YsH3zwAVOnTuWvv/6SDU2aNWvG4MGD2bZtG82aNbOzpPULM0OTE38KQxNBOWqs4IqLi3nzzTfZuXMnGRkZSFJpXLhDh5x4msDUuKRhH/CJMDuctqN0bS66d3+LlnD7RQYBQQU0bdqUZcuWsWzZMm7evEl2djaBgYH4WFjHFRgQhiaCqqjxEGL69On83//9HwMHDuTUqVOMHj2awsJChg4dagv56geSDi6tKy2XmZ4E8/xvltbfwNyCUjh4CyrCx8eHyMhIodyqwGhoYkSswwnKUmMFt3HjRjZt2sTUqVPx8PBg6tSpfP311+zcubNG/dQka7BKpTLbPDw8aNu2rcW2NuH6TihKN+y7qyBiiNnh/NRL3Dh1wlBQKIjq1Zey5BWVcOxyqYO3MDARCGqPMDQRVEaNFVxBQQHx8fEAeHt7U1RURKtWrTh4sGaLvDXJGpyfn2+2tWrVihEjRlhsaxNSTENzPQju5ilATKcnQzvciXdQcLkufr+Ygf7WdG58sIrwAMtpRAQCQfURhiaCyqixgmvWrBlHjhiMLe644w6WL1/Ov//9bzm1RnW53azBBw4c4MSJE4wZM6amot8e2nxI/7a0bHF60sQ9oHf58FwAv164Ku+L9TeBwDpYMjQRCIzU2MhkyZIl5OfnAwafuMcff5y8vLwa+ejcTtZgIx999BH333+/HNjZEtbMMKxI/Qp3nSHBnuQdjbbBPWDSVq/TkbbzZ7kc3qOPxb72nS9VcJ1iQ2qdYdcZM/XWd6yVZbg6SJLElStXiIiIqLqxC6OKicMrKJjirExhaCIoR40UnE6no6SkhB49DCGo7rzzTv76668aX9SoIGuSNRgM06Pr1q3j008/rbR/a2YY7iKtwBhQ66+iDpzc/KPZcc3ZM5TcMKytKXx8+eNaJooyGYYlSWLPX5lyWZt2hh9+uFDpPVQXZ8rU6yjUNstwZeTl5TFp0iQ+//xz3N3dKSgoYOPGjRw6dIhXX33VKtdwJuTUOSKiicACNVJwbm5ujBgxgry8vFpd1Bi3siZZgwE2bNiAr68vgwYNqrR/q2UYvpmG+/YTcrFRj1k08m9h1iRp2WFu3NqP6d2P3hZClp29nkvutq8B8PZwY+KwwXi6u5VrVxOcMVNvfcdaWYYrIzExEZ1Ox7Fjx+jUqRMA99xzDy+//LJQcBUQ0s5cwYnUOQIjNZ6i7NKlCwcPHqRjx45VN66AwMDAGmcNBvjvf//L6NGjcXevXGyrZRi+8CVwy89P3RGPoL+VOy99V+n0ZEy/+yz2/0dqaRT4jjEh+Pl4Vyp/TXCmTL2OQm2zDFfGjz/+yIULF/Dx8ZF9KcPDw7l69WoVZ7ouwtBEUBE1VnAJCQkMGjSIkSNHEhsba5bctCYZh2uaNfj06dPs27ePjz/+uKYi3x6SZO7cHVveuKQ45wbXDx6Qy9F9K/B/Ew7egmri5eVVzl0mMzPTLP+iwBwR0URQETVWcAcOHKBVq1YcPnyYw4cPy/UKhaJGCq6yrMETJkwADEFojXz00Ud079697sIV5SRB3q0cdwp3iHq0XJPLu7Yj6QzBlxs0bY5/bLzFrvaLDN6CajJkyBAmTpzIihUrAIPB1EsvvcRDDz1kZ8nqL8LQRFARNVZwO3bssMqFK8sabKrYjPzzn/+0ynWrTcpnpfth94JXeTeI6rgHFBRrOHq5dIpSOHgLKuP1119n7NixBAcHo9frUalUPPjgg7zzzjv2Fq3eIgxNBBUhov1aQq8xpMYxYsH3TZIk0n42zd5tWcH9kZKBTm9Yx4sJ9CNK7WddWQVOhZ+fH+vXryc9PZ0DBw6QmprKhg0b8PMT35vKMItocliswwkM1HgEp1QqLQYSBoMbgVNwbRuUZBj2PdQQfn+5Jjl/nSY/NQUApYcHEd16WexKrL8JbofQ0FBCQ0XG9+piZmhyRITsEhiosYIzXXcDuHz5MsuWLavb0Fm2xjQ0V9Qj4FbeItN0ejLsnq543HJ9KMt+EWBZUAPOnDnDSy+9xKFDh2R/USNZWVkVnCUQhiYCS9RYwbVr165cuWPHjvTr149nn33WaoLZjZJsuLKptGxhehIg1Wx60rL1pCRJIkSXoEY8/vjjtGvXjg8++KDCoASC8ghDE4ElbjvhqSne3t5cvHjRGl3Zn8tfg/5WmC+/xhB0d7kmuuJi0vfukssVuQdcyMzjWl4RAJ7uSjrE1Cxep8D1OHv2LL///ruZ+42gaoShicASNVZwZa25CgsL+eqrr+jVq5e1ZLIvpr5vMSPBwnrjlf170N28CYBPwzCC2lhO3WPqHtAhOgQvj9pFLxE4P/fffz+//fYbnTt3trcoDodZRJPDB2F0xX61Ategxgru66+/NiurVCruv/9+pk2bZjWh7EbBecjcV1qOsbyuaDo9GdW7P4oK3rbNDEzE+pugGrz//vv06tWLNm3aEBYWZnbszTfftJNUjoEwNBGUxW5+cPUS06zdwV3Ar5HFZmmm2bsr8H8DYWAiqDlTp07lypUrtGjRguzs7KpPEMiYKjhhaCKA21BwW7ZsIS4ujhYtSoMOnzp1ikuXLtG/f8U/9vWesqG5KjAuKbySTtbxo3K5Iv+3myVaklJLMwgIAxNBdfjqq6/466+/CA8Pt7coDocqJlYYmgjMqPFKdmJiYjmnUz8/PxITE60mlD1QZB8wTFECKL0g0nJopFST7N3BbdvjE2pZcR28lIH2loN3ZANfYgKFo66gamJjY/Hx8bG3GA6J0dDESEaSmKZ0dWqs4NLT04mOjjari4mJIS0tzWpC2QNFmsn0ZMRg8FRbbFed6CVQ3sG7Iud4gcCUiRMnMnz4cLZt28bRo0fNNkHViIgmAlNqPEUZFRVFUlKSWTbuI0eOVJphu76jlEpQXv6qtCLmcYvtJL2etJ3b5HJUBe4BIAIsC24P40xI2aSqCoXCeSIF2RBhaCIwpcYK7rnnnuPRRx9lwYIFNG3alLNnz7JgwQKee+45W8hXJ4RxGIXWkHAVzxBo2M9iu8yjhynKNITw8lCpCOvUxWI7g4O3UHCCmqPX6+0tgkMjDE0EptRYwRnfMBcvXszFixeJj49n0qRJNUqVU2+QdCgyd9OMb0rrooeD0nLySlP3gIhuvSp8cFKy80nPLQTAXakQDt4CQR0hDE0EptRYwRnzvjmkQjPl8jdw9EXci9IINK33iarwFNP4kxWF5wLz6cn2MSH4eFolYIzASXniiSf47DNDeqaHHnqowvXar776ymK9oJRyEU0OHxQKzoVxTTeBy9/AgScAqfyx47MN/m+RQ82qS3JzuXrgV7kcVW0DExERXlA5f/vb3+R907Vtwe1hFtFEWFK6NLc1Rfnzzz+b1RndBE6dOmU1wWyGpIOjL2JRuRn58yWDJaWiNLRW+p6dSFotAP5xjQho3LTC080dvMMqbCcQAMyYMYNBgwaxadMm5s2bZ29xHJ6Q9nfK+8LQxLVxPTeBjL1QVJmsEtxMNbQzwSw8V58BFU4jFWm0HLokHLwFNWP37t32FsH23FrzjpL2ocjcbXjZtAGmU5JGQxOBa1JjBWd0EzDFodwEiq/cVjuz9bdK3AMOX8pEozNYwoX5+xAfbDlPnEBgb7799lsSEhLw8/MjMjKS999/H4Dc3FxGjhxJQEAAYWFhLFy4sPYXu/wN/NQK9/1DuJMVuO8fAj+1MtRbGaOhCSAbmghcE6u4CcyfP5+JEyfaQj7r41XNEEgm7XIvnCMv2RDlROHuTmT33hWeVjbAsnDwFlSHkpIS3n33XSSp4qlzaxp2/fjjj0ycOJH/+7//o3v37uTm5nL1qiF3YWJiIllZWaSkpHDt2jX69etHXFwcTz311O1drKI176LLhvpOn5Vb864NwtBEYMQqbgKTJ08ulwi13hLSFbyjDA+XxXU4hcGSMqSrXGM6PRl21z14BgRU2P2vIsCy4DbQarWVWkkarZetxZw5c5g7d66c5iowMJDAwEAKCwtZt24de/fuRa1Wo1arSUxM5KOPPro9BVfpmrcEKCyuedcWYWgigFq6CVy+fJnVq1fz7rvvkp6eTkFBgS1ktC4KN2i77NYbpQLzB+/WaOuOf5o9bKbTk1GVuAeAiGAiuD18fX3rLFNHQUEBBw8eZODAgTRv3pzc3Fy6d+/OO++8w5UrVygpKTGz5kxISGDJkiUV9ldcXExxcbFczs3NBUCj0aC9ugv3aqx5a6/uQgruXttbkwm8I0Hev550EI1GU6v+jOfXtp+6xhnlrsm91FjB6XQ6vvvuOz766CN++ukndDods2fPdiy/uMihhmmRoy+aG5z4RBmUm8l0ia6khPTdpT88lcWfTM0uIPWGQcm7KRV0FA7egnpIdnY2kiSxceNGtm7dSnBwMBMmTGDUqFHMnz8fPz8/3N1LfxrUajV5eXkV9rd06VIWLFhQrn7Lli0080niTgvnlCVp/0+kKSq+Rk3RZd2Q9zOPHWXTN9+g8LAcwKEmlA2h5ig4k9yFhYXVPr/aCu706dN89NFHfPrppxQUFPDQQw/xww8/MGrUKCZNmkRwcHC1L1oviBwKEYPRXt1F0v6fSLjnXtzDepabJrn2+69o8vMB8AoKJriSuXxT94B2UUH4edX+gRK4BpWtvVkblcpg+DRlyhTi4uIAWLBgAc2aNePVV1+lsLAQrVYrK7mcnBz8/f0r7G/GjBlMnz5dLufm5hITE8OAAQNooAmD/SuqlCnhnntpZ8URnCRJfL5kNsVZmaDTck98TKXPblVoNBq2bt1K//798bCCoqwrnFFu4wxBdai2gmvVqhXBwcG8/vrrPPbYY3LKHIc2olC4IQV3J02RZ3i4LKwBmLkH9OqH0q3idYJfL1yV94V7gKAmVDZCsjZqtZrY2FiLx+644w48PDw4cuQIHTsa4jomJSVxxx13VNifl5cXXl5e5eo9PDxwD+pZrTVvSy+XtcXU0OTGsaOE33l3rfv08PBwKEVhxJnkrsl9VNtNYMyYMRQVFTFr1izmz5/PsWOuYXqbapK9uzL3ABAZvAWOw7PPPsu7775LWloaN2/e5NVXX6Vv374EBATw2GOPMWfOHHJycvjrr7949913GT9+/O1dyLjmbShYaCCVW/O2Fmapc4ShiUtSbQW3atUq0tPTWbhwIXv37qVdu3Z06NCBvLw8cnJybCmj3bh5/RqZJpEQonpXvP5WotVxMMXEwVsoOEE95pVXXqFv3760a9eOmJgYCgsL+d///gfAihUraNCgAdHR0XTt2pVx48bdvosAlK55e1vwlVU1h4gHbr/vShARTQQ1cvRWqVSMHz+effv28eeff9K7d2/8/Pxo167d7b/h1WPSdpaGJAtsfQd+ERU7syelZlKsNURmCFF50ySkYlcCgcDeuLm58a9//YuMjAwyMjLYsGED4eEG38+AgADWrl1LXl4e165dY+7cubW/YORQuPck2nu+4ziPl05W5p+Bqz/Wvn8LmEU0OX4UnYmlp8A1qHEkEyOtW7fmX//6F2lpafzvf/8jPT3dmnLVC9K2Vy97N5g7eHcWGbwFtUCr1bJ3714+//xzwGDW7xAuOFVxa837rGIIUtSI0voT80Gyfh48s4gmGo2IaOKC3LaCM+Lu7s4jjzzCpk2brCFPvUGSJFJ3VH/9rWwEE4Hgdjh16hStWrXiscceY9y4cQD8/PPPTjdDomv+CihuGQvkHoPUDVa/hjGiiRGxDud61FrBOStZx45y86ohHqWbjw9h93SrtL0wMBFYg4kTJzJ16lRSU1Nla7FevXqxZ88eO0tmZXzjIf7p0vKpRaC3vjOyaYZvoeBcD6HgKsA0eklElx64e3tX2DY9p5CLWQZfOaVCwV1xIgec4PZISkqS47oap7kDAgLq1I2gzmjxErj5GvYLzsPFT61+CTMFJwxNXA6h4CogrZrZA8B89HZHZCAq4eAtuE3CwsJITk42qztz5ky5FFVOgXc4NDEJ0n56KehuWvUSwtDEtREKzgKaggKu7C/NB1dV/Enh4C2wFs899xyPPPII3333HTqdji1btvDkk08yefJke4tmG5o+Dx5qw35ROpx/36rdC0MT10YoOAuk79mJ/laSRL+oGNTNW1bafv+F6/K+WH8T1IYpU6bwzDPPMGPGDHQ6Hc8//zxPPvkkEyZMsLdotsEzEJo9X1o+8yZorOdXKwxNXBuh4CyQtmObvB/dt+Ls3QAanZ4/UkoVnLCgFNSWiRMncuzYMfLz8zlx4oTzjt6MNJ4IXreeG00W/PW2VbsXhiaui1BwFjDL3l3F9OSR1ExuagwO3kG+XjRv2MCmsgmcmy1btnD69GmzulOnTjlsNPhq4e4HLV4pLZ9bAcXXKm5fQ4ShiesiFFwZ8lIukvOX4QdGoVQS2bNPpe1NDUxEBm9BbUlMTJQDmRvx8/OTEw07LfFjwdeQ2QBdAZxeVnn7GiAMTVyXGueDc3ZMrSdDO96Flzqw0vZmDt71yMBEkiS0Wi06nc7eopih0Whwd3enqKio3slWH0hPTy9nMRkTE0NaWmVJQ50ApSe0nAWHnjWUk/8LTRPB13LWg5pgNDQpzsqUDU1MR3UC50UouDLUJHs31E8H75KSEtLT02uUGLCukCSJ8PBwLl265JCjXU9PT5v2HxUVRVJSkllG7SNHjhAZWXEcVKchZgT8tRzyToK+BE4tgQ61t6o0GpoYU+dkJB0SCs5FEArOBL1Wy+Vd2+VyVetv1/Jucj7D4ICrUECneuDgrdfruXDhAm5ubkRGRuLp6VmvFIleryc/Px+VSoVS6Tgz5JIkcf36dYKCgtDpdDbLrfXcc8/x6KOPsmDBApo2bcrZs2dZsGABzz33nE2uV69QuEGreXDgVpzKlM+g2TTwb1HrrkMSOpopOIFrIBScCdf++I2SXIOJsmcDNaEd7qq0/X6T6ck24YEE+Nj27b46lJSUoNfriYmJwdfX197ilEOv11NSUoK3t7dDKTiA4OBgrl+/jlartdk1jGttixcv5uLFi8TFxTFp0iSmTJlis2vWKyIGQ+CdkP0HoIeTC6HT/9W6W2Fo4poIBWdCmklw5ahefVG6V/7x1OcAy46mPBwB40hYkixlprbeNaZMmeI6Cq0sCgW0ng97BxvKl7+GG4dB3b5W3VoyNHGzkIVc4FzY9VdQo9EwefJkAgMDCQoKIjExsdK342+//ZaEhAT8/PyIjIzk/fetG/WgputvphFM6sv6m8DxKSgo4PDhw/zyyy9mm8sQ2htCe5WWT8yrdZcioolrYlcFt2jRIvbs2cOJEyc4fvw4u3fvZsmSJRbb/vjjj0ycOJG33nqL3Nxcjh8/Tq9evawmS3F2FhmH/pDL0ZVk7wbQ6vT8npIhl+uTBaWjo1ar2blzZ7Xa9urVi7feesum8tQl69evJyIigs6dOzN06FB5e/DBB+0tWt3SekHp/rWf4XrtFLxCoSC0vXD4djXsquBWrVrF7NmziYiIICIiglmzZvHRRx9ZbDtnzhzmzp1Lr169cHNzIzAwkJYtKw+hVRPSf9mOpDckXWzQvCWqmMrNk/+8nEVhiWG0qfbxpGWY2mqyOAq9evVCoVCwbds2s/ply5ahUCh4/vnn7SOYA/PSSy+xcuVKioqKyM7OlresrCx7i1a3BN4JEQ+Ulk/Oh1pODQe3EyG7XA27rcFlZ2eTmppqZg6dkJBASkoKOTk5NGhQGhGkoKCAgwcPMnDgQJo3b05ubi7du3fnnXfeISIiolzfxcXFFJs4c+bm5gKGKVGNxjznlLF86efS6cnIXn3LtSvL3nOlGcw7xYWg02mpS7cuo3yW7keSJPR6PXq99bMkl6VFixasWrWKPn1KHeI//vhjWrZsKcthinH9ytIxU2oif1V9WQuj7FqttsLvUW3Jyclh1KhRVunL4Wk1F9K/AyTI+g2ubIaIgbfdnTA0cT3spuDy8w3509RqtVxn3M/LyzNTcNnZ2UiSxMaNG9m6dSvBwcFMmDCBUaNG8fPPP5fre+nSpSxYsKBc/ZYtWyxaFkqSRPJPpRnJ01Rqfvjhh0rl//JYrrwfWHKjyva2omwIJ3d3d8LDw8nPz6fkVsBoW6HVannwwQf54IMPuHTpEg0aNOCPP/5Ar9fToUMHSkpKyM3N5fDhw7zyyiucOnWKiIgIXnjhBR599FHAoMiWLl3KJ598gpubG9OnT0eSJAoLC+UXky+//JI333yT1NRUmjRpwtKlS7n77rtlGYqKiuS2tsT4ee7bt6/cWrG1fA5HjRrFV199xSOPPGKV/hyagFYQ8zhcWmMon5wP4feB4vYmnoShiethNwWnUqkAwxtrSEiIvA/g7+9vse2UKVOIizOE81mwYAHNmjWjoKCgXGijGTNmMH36dLmcm5tLTEwMAwYMICAgwKytRqNh8+qP0GcbpoCUXl4MmTId9ypM7F9I+howjBKf7N+FAS2jqn3v1kCj0bB161b69+9v5pNVVFTEpUuXUKlUeFeSpNUauLu7ExYWxn333cemTZuYMGEC69ev5+mnn+b48eN4enqi1+sZNmwYc+fO5e9//zt79+7lgQceoHnz5nTr1o1Vq1axbt06du7cSWxsLJMnTyY/Px9fX18CAgL44YcfmDdvHhs3biQhIYGNGzcycuRITp06RXBwMO7u7nh7e5f7v9qCmzcNucq6dOkifyeNWEvBXrx4kSeeeIJ//etfhIeHmx376quvrHINh6LlbEjdAJIGco9D6nqDQ/htICKauB52W4MLDAwkOjqapKQkuS4pKYmYmBiz0RsYRnaxsZbXxCyZbHt5eREQEGC2AXh4eFjcSv4slSG8czd8GjSosK2Hhwc5xTrOXi/9QevSJKLS9rbaKronhUKBUqm0+QaGxfunn36aTz75hOLiYr766iueeuopFAoFCoWCzZs3ExoaypQpU/Dy8qJXr148+uij/O9//0OpVLJ27VoSExNp3bo1KpWK119/Hb1eL1/j3//+Ny+++CJ33nkn7u7uPProo7Rs2ZIff/zRTIa6uF+jm4C7u3uF/4/a0rFjR2bMmMG9995Lu3btzDaXxC8O4seVlk8uMkQ5uQ2EoYnrYVc/uLFjx7J48WK6du0KwJIlSxg/frzFts8++yzvvvsu9913H0FBQbz66qv07du33Jv07VBy7Ii8X1X0EoDfTMJztQpXE+jr2tMcffv2Zdy4cSxcuJDOnTubjTxSU1OJj483ax8XF8eBAwcAuHz5sjwqB0NGay+TaaPk5GRmzpzJvHmlpuIajcZpYzOa3qfgFi1ehJRPQVcIhRfg4qfQyPLvRFUEt+tA6q31dqHgnB+7Krg5c+aQmZlJq1atAMP6w8yZMwHkBI9GX7dXXnmFrKws+U22d+/e/O9//6u1DNqbN9GcPiGXq6Pg6muAZXuhVCoZPXo0ixcv5osvvjA7Fh0dTXJyslldSkqKHFA4MjKSixcvyseuXbtmZiAUExNDYmKi8yb8tMCOHTtYs2YNV65c4bvvvuOPP/4gLy+P3r1721s0++AdDk0mwpk3DOXTSyFmJLjXPFKP6ZTk9aSD1pJQUE+xq5uAh4cHK1eulE2h3333XdxvRQ95//33zRy53dzc+Ne//kVGRgYZGRls2LCh3BpFdcm/lELGkUNkHDnEX2s+AY1hysM7OASdpoT8SymVnm+q4ISDt4Fp06axZcsWhgwZYlY/cOBArl27xnvvvYdWq2X37t188cUXPPnkkwA8/vjjrFy5ktOnT3Pz5k1mzJhhFoVl0qRJLFu2jIMHD8rGJ9u2bSM1NbVO76+u+O9//8uTTz5JWFiY7Nzt4eHB3Llz7SyZnWn6PHioDftFV+DC7QV5MDU0yT7xp0id4+S4XKiu/EspbLirNbrionLHijIz+Kb33bh5eTPs9xMWfeF0ej0HLpZm8BYKzkBQUBD9+vUrVx8YGMjmzZt5/vnnmTFjBpGRkbzxxht069YNgKeffpoLFy7QvXt33NzcmDVrFl9++aV8/pAhQygqKuKZZ57h/PnzeHl50alTJ1auXFln91aX/POf/2TLli20bt1avsfWrVtz8uRJO0tmZzwDDYGXjVFNzrwJcU+Dp7pG3ahiYvEODqEoM0MYmrgALqfgirIyLCo3U3TFRRRlZVhUcMfTs8kvNvg8+Xt70CpcbQsxHYLKoo2sXr1a3u/UqRP79u0DDG4BphaHSqWSxYsXs3jxYrlu8uTJZn0NGzaMYcOG1VgGRyQzM5PWrVsDpbEvjQY7Lk/jiXDuPSi+CppsOPsOtK7ZyNaYOsd0HU4oOOdFROStIfsvlI7e7o4LxU0ENRZYkXbt2pmNYMEQg7VDhw4VnOFCuPtCi1dKy+dWQNHVittXgIho4jq43AiutpgGWK5vGQQEjs8bb7zBgAEDWLNmDYWFhYwcOZLt27fz008/2Vu0+kH8GDj7NhQmg64AziyDtm/UqAthaOI6iOFHDTHP4B1mR0kEzkiHDh04fvw4nTt3Zvz48bRr145Dhw65rh9cWZSe0HJWaTn5Iyis3CisLMLQxHUQI7gakFVQxKmrOXL57nqQwVvgPGi1WmJiYrhw4QIvvPCCvcWpv8Q8Bn+9CXknDU7fpxZDhw+qfbowNHEdxAiuBvyWXLr+1rxhA4JVtg2FJXAt3N3dUalUNo8h6vAo3AxJUY2krIG8U9U//ZahiRGxDue8CAVXA0ynJ4WDt8AWzJo1i7Fjx/Lnn39y48YNcnNz5U1gQvggCLzrVkEPJxfW6HRhaOIauNwUpXdQCG5e3pW6Crh5eeMdFFKuXjh4C2zN008/DcDXX38tuwZIkoRCoUBXl/mY6jsKhWEUt3eQoXx5I2QfgsDqWZsKQxPXwOUUnComlmG/n6Aoy5CNW6vVsmfPXrp16ypHUfEOCinnA6fXS2YxKIWCE9iCCxcu2FsExyG0F4T2hus7DOUT86Drd9U61ZKhiUid43y45BSlKiaWkHYdCGnXgeC27fGIb0xw2/ZynSUH75NXb5BbZHDw9vN0p01EYF2LXacs3HwY5eT/snDz4Tq/dnx8PBs3bqzz69YH4uLiiIuLIzY2Fk9PT7lsGpBaYEJrk7yP17fD9V3VOs1oaALIhiYC58MlFdztsN9kerJTfCjubs770S3cfJi5mw4iAXM3HbSZkuvVqxdvvfWWTfp2VHJzc3nqqafw9vamadOmAGzcuNFmsShv3rxJ06ZNzRIP5+bmMnLkSAICAggLC2Phwpqtb9UpgR0hYmhp+cR8sJBCqyzC0MQ1cN5faStj5uDtxAYmRuVmii2VnMCcKVOmoNPpOHbsGJ6engDcc889fP755za53ty5c8uNDhMTE8nKyiIlJYXdu3fz4Ycf8umnn9rk+lah1Rzkn7LsA3Dlh2qdZmpoItbhnBOh4KqJKxiYWFJuRuyl5FJSUujfvz+hoaEEBgYyaNAgs/Q7Y8aMYdy4cTz66KOoVCratGnDsWPH+OCDD4iOjiY0NJT33ntPbn/48GG6detGUFAQoaGhPP7442RmZtb5fVXEjz/+yH//+1+aNWsmG5mEh4dz9WrNQ1JVxcGDB/nxxx95+eWX5brCwkLWrVvHokWLUKvVNG/enMTERD766COrX99qBLSCmMdLyyfng1S1QY6poYkYwTknLmdkcjvcKCzmxJUbcvluBxrBSZJEiVZfZbslW5J4tQoFNnfTQbR6PTMHJFTaztNdabXgwHq9nunTp9O7d29KSkoYN24czzzzDFu3bpXbbNiwgc2bN7Nu3TrGjx/P0KFDeeSRRzh//jy7d+9m4MCBPPLII4SFhaFUKnnttde4++67ycrKYtiwYbzyyit8+OGHVpG3tnh5eaHVas3qMjMzCQoKsup1tFotzzzzDCtXrkSvL/1+nD59mpKSEhISEuS6hIQElixZUmFfxcXFZjn8jC4NGo0GjUZj1tZYLltfa5q+hHvqehSSBnJPoE1eixT9WKWnqP/WVt7PPvEnRfn5FRqa2ExuG+OMctfkXoSCqwam6XGahATQ0N/HjtLUjBKtHu9pH1utv1c3H65SERYtH4uXh5tVrhcfHy9nBPf29mbWrFncc8896PV6OW/coEGD5Kzww4cP53//+x8LFizA09OTvn370qBBA/7880/CwsLMQl6FhYUxffp0XnzxRavIag2GDBnCxIkTWbFiBWBQHi+99BIPPfSQVa+zbNky2rdvT48ePcwyMuTn5+Pn5ydbFAOo1Wry8vIq7Gvp0qUsWLCgXP2WLVvw9bWclNT0BcVa3CH1oTGGmJ3FR+bw8xEfJEXFP3GSJKHwD0DKy0Wv0fD9R//BI75Jpdewhdx1gTPJXVhYWO3zhYKrBmYO3o1EeK665Pr160ydOpXdu3eTk2MIk1ZcXExeXh4NGjQADIrKiK+vL/7+/vj4+JjV5efnA3D27Fn+8Y9/8Pvvv5Ofn49er8fDw6MO76hyXn/9dcaOHUtwcDB6vR6VSsWDDz7IO++8Y7VrnD17lvfff5/Dh8u/qKhUKgoLC9FqtbKSy8nJwd/fv8L+ZsyYwfTp0+Vybm4uMTExDBgwgICAALO2Go2GrVu30r9/f+t/7sV3Iu1IQKErxI9rDPrbdfRx4yo9Zeund3N5h+FHtJXKl+YDB1psZ1O5bYgzyl2ToAdCwVUDs/W3eBFguS6ZMWMGhYWFHDp0iNDQUJKSkmjfvj1SNSzlLDFhwgSaN2/OJ598glqtZuPGjYwZM8a6QteQI0eOyCNLPz8/1q9fT0ZGBsnJycTExJgpcGuwZ88erl69SvPmzQHDj0leXh4hISF8+eWXeHh4cOTIETp2NKxRJSUlcccdd1TYn5eXF14WpvY8PDwq/FGt7Nht4xEFTSYZMgwAbmeX4Rb/lCHNTgWEtu8oK7isP5OqlMkmctcBziR3Te5DKLgqMDh4m2Twbuw4629gWA8rWj62ynbVWYMDmHt/+2qtwVUXrVZLUVFpVJmya3e5ubn4+vqiVqvJzMy0OBVWE3Jzc/H39ycgIIBLly6xbNmyWvVnDbp37y6/lTZr1oy//vqLkJAQQkLKR9OxBsOHDzfLvv7rr78yfvx4kpKSaNiwIY899hhz5sxh7dq1XLt2jXfffbd+uwqY0vR5uPBfQ0LUoitw/t/Q/B8VNheGJs6NsKKsgjPXcsguNCyg+3i4cUekdRf7bY1CocDLw63KbcGgjrw6qPKI6q8O6siCQR2r7KsmBiYvvvgiPj4+8taiRQuz4wsWLODs2bMEBgbStWtX7r///tv6HIy8+eabfP/99wQEBMjGKPZGpVKxb98+cnNzuXLlCnl5eWYxKK0di9LX15fo6Gh5Cw0NRaFQEB0djaenJytWrKBBgwZER0fTtWtXxo0bx1NPPWW169sUTzU0m1Za/ms5lNyosHmoiYITqXOcDzGCqwLT9be74kLxcGIH7zn3twew6Crw6qCO8nFrYWrcUBGtWrXiwIEDZnXPPvusvL969WqzY7169eLGjRtmdaZuBd26deP48eNmx03Xj+zBggUL6Nevn2yJaOp0DbaPRVn2MwsICGDt2rU2uVad0Pg5OLcSiq8aRnJn34bW8yw29YuOEalznBjn/bW2Eqbrb87s4G1kzv3ty43kbKHcBKU888wz5ObmkpycjI+PD+fPnzfbLly4wPnz5+0tpuPg7gstXiktn1sJRZb9CEVEE+dGKLgqMI1g4qwO3mUxKjkFQrnVFe7u7sTExLBnzx6z+JMiFuVtEj8GfOMN+7oC2fDEEiKiifMipigrIa+ohGPp2XL5HhdRcGBQckKx1T0JCQn88ssvHDx4sJzvma3iUTolSk9oNRsOjjeUL/wXmiSCX/kXBWFo4rwIBVcJBy5el+O2xgerCA+o2NxYILAGL7/8MitXrqRLly74+fnJ9daKDONSRA+Hv96E3BMgaeDUYuj4n3LNLBmaiNQ5zoFQcJWwX/i/CeqYVatW8ccff9CyZUt7i2ITFm85wqtbrzPX/QjzB91p24sp3KDVfPhtuKF8aa3BwjKglVkzYWjivIg1uEowMzAREUwEdYCPjw+NGze2txg2YeHmwyzYnIQELNicVDfBu8MHQmCnWwU9nCzvz1fW0OT6YbEO5ywIBVcBkiSZuQh0biRGcALbM2/ePGbOnGkWvNgZsFsaJoUCWs8vLad/A9nlFZipoUnGEbEO5ywIBVcBZ6/nkllg+JHx9nCjXZRjOXgLHJMXXniBt956Cz8/P4KCgsw2R8XuaZhCe0Jon9LyifI+caHtS6dLhaGJ8yDW4CrAdHqyY0wInu7WiY4vqBkpKSm0bt2atLQ0ObiyLVm9ejVvvfUWSUlJNr+WJTZu3GiX69qKypSbEeNxm1rttl4Au7Yb9q/vgOs7IbSXfDjEZAQnDE2cB6HgKsAsg4ALOHgD5F9KoSgro8Lj3kEhqGJirXa9Xr16sWvXLrZu3WoWG3HZsmW89NJLTJ06lbfeekvOBFAZycnJNGrUiOzs7HKRQByJnj172lsEq/H61iMs2XG6Wm1truQCO0DEUMMUJcCJ+dBjh2EKE2Fo4qwIBVcBrpDB25T8SylsuKs1uuKiCtu4eXkz7PcTVlVyLVq04OOPPzZTcB9//LFNrAg1Gk29jahenXQ4U6ZMqQNJrMeSn5LAs/q5E+dtOmjjUdxcSP8O0EP273BlE0QMBkoNTVJ/3gIYDE2EgnN8xBqcBQqKNRxNy5LLruDgXZSVUalyA9AVF1U6wrsdRowYwebNm+Vcb7/99hsAd999N2AYmSkUCjlW4tatW2nbti3+/v6EhYXx3HPPAdCpk8FSLjo6GpVKxWeffcbOnTtRq9X8+9//JjY2li5dugAwatQoIiMjCQgIoGPHjuzYscOq93Q7fP3115Vujjh1OfPehBq1/0ffilPyWAX/lhA7srR8Yj5IpfE9haGJ8yFGcBb4/eJ19Lc8vGMC/YhS+1VxRv1FkiT0JSVVttNVo42xXVUR15WentV2TFar1dx3332sXbuWCRMmsGrVKsaOHVsuILKR0aNH8/rrr/Pkk09SUFDAkSNHADhw4ACNGjUiNTVVnqLcuXMneXl5HDlyhFOnTsl99O3bl5UrV+Lr68tbb73Fo48+SnJycqVJPW1NfVCy1ubl/u3w9lVVuQZn5N1dJyjW6ni5fzvbPXMtZ0HqetCXQN5JuPS5rPSEoYnzIRScBfab5n9z8NGbvqSEj8Ot92Px3YBuVbYZe6WgRgv0Y8eOZfbs2YwePZovv/ySY8eO8corr1hs6+HhwdmzZ7l+/TqhoaHyqKwi9Ho9r732Gr6+pVFoxo4tzY/34osvsmTJEo4ePUrXrl2rLbOgelSWoaIsxVod7+46wX/2nubZri1so+h8YyF+nCFPHBiim0Q/CkpPYWjihIgpSguYBlh2FQMTe9K3b1/S09NZuHAhnTt3Jjw8vMK2X3/9NceOHaNFixa0b9+e9evXV9q3v7+/mdGJXq9n1qxZNGvWjICAANRqNTk5OWRkWHfqVVCKpQwVRuYP7MDaMb1pFa6W64yKrsn89SSu30fajQLrCtT8RXC7pTgLkyF5NVBqaAKg12jIOvGnda8rqHOEgiuDJEllDEyEg7etUSqVjB49mtdee81sdGWJDh068OWXX5KRkcGcOXMYOXIkV69eRam0/FUuW79mzRrWrFnDpk2byMnJ4caNGzRo0ADJGHRUYBMqSsM0b2AHRtzZhD9nPmxR0a345QSN539uXUXnHQZNJpWWT78G2gKROscJEVOUZbiQmcf1fIOxhae7kvbRwXaWqHYoPT0Ze6XqH4aMo4erNf04ZMseQtpWbumm9PSstnxGpk2bRs+ePSs1ky8pKeHzzz9n8ODBBAYGyiMzd3d3AgICUCqVnDt3jo4dK7Z+y83NxdPTk5CQEEpKSnj99dfLRe0X2IY597dHp9fx6uYk5t6fYGYx6aZUMuLOJgzr0IgvDl9gwebDnLxyA4ASrZ4Vv5zgP/tO8WyXlrwywApTl02nwoUPDQlRi68apiybv0BIQkfZklIYmjg+YgRXBtPRW4foELw8HNvBW6FQ4OblVfVWTaXk5ulZZV+3E/k+KCiIfv36VWnGv2bNGpo2bYq/vz+JiYmsWbOG4OBgfHx8mDdvHvfffz9qtZo1a9ZYPH/06NG0adOGuLg4GjdujI+PD9HR0TWWV3B7zBrQjq/7hzJrQDuLx92USh7raBjRrRtrPqIzKjqrjOg81dDcJJP7X8uhJFukznEyxAiuDPvNAiy7zvqbd1AIbl7eVfrBeQeFWO2aO3furPDY6tWr5X3T6cPNmzdXeM7cuXPL5UwzuhcY8fPz46uvvjKre/HFF+X9MWPGMGbMmIqFFtQJRkX3aHsbjugaTbiV7fsKaG7A2bcJafe0fFgYmjg+QsGVwdUcvI2oYmIZ9vuJOo1kIhBUhU0VnbsvtHgFjjxvKJ9biV+/CWYRTbJO/GnmPiBwLISCM6GwRMuRtEy57EoKDgxKTigwQX3EZooubjT89TYUXgBdIYozy8wimmQkHRIKzoERa3AmHEzJQKs3TIdFNvAl2oEdvAUCZ8Tqa3RKT2g1u7Sc/BGRdzWVi8LQxLERCs4E8/xvDW/LWEIgENiesoqudW0UXfRwCGhj2Jc0NG55Rj4kDE0cG7spOI1Gw+TJkwkMDCQoKIjExES0Wq3FtmPGjMHT0xOVSiVvv/76q9VlEg7eAoFjYVR0R2uj6BRKaFWaI85P2oM60vByazQ0ETgmdlNwixYtYs+ePZw4cYLjx4+ze/dulixZUmH7iRMnkp+fL2+dO3e2qjzlHbyFghMIHIVaK7rwgRBkCPCtQKLTCENoNxHRxLGxm4JbtWoVs2fPJiIigoiICGbNmsVHH31kL3FIyS7gSu5NANyVCjrEWM8cXiAQ1A3VVXST1+8jNdtE0SkU0Gq+XIy9Q0toI8MoTkxT1j2Ltxzhoa3XWbzlSK36sYsVZXZ2NqmpqSQkJMh1CQkJpKSkkJOTYzFz86effsqnn35KREQETz/9NNOmTaswPFNxcTHFJtMKubm5gGFaVKPRmLU1lveeSy+VJToId4VUrm19wiibpfuRJAm9Xo9er7eHaJVi9GkzyuhIGGXXarUVfo8E9YOyVpevbj7MCROry5W/nODDfad4pktLXunfjuhAPwjtAQ37wrWfAbjzEXc2v6Eh48ghmo6qPIScwHos3HyYBZuTAFiwOQk3pdtt5wm0i4IzZmg2DYJr3M/Lyyun4KZMmcKyZcsICgri999/Z/jw4SiVSqZNm2ax/6VLl7JgwYJy9Vu2bDGLKm/KF3sOy/thUiE//PBDTW7JbmzdutWs7O7uTnh4OPn5+ZRUMwWOPbBHeKzOnTszb9487rvvvts63/h57tu3r9x6cWFhYa3lE1ifGiu61gtkBRfVRklkKwUZSYdYvOUIr269zlz3I8wfJNwGbMXCzYfLZZ6oTbZ3hWSHKLPZ2dkEBQVx9uxZmjRpAsDZs2dp1qyZHPy2Mt577z0+/fRT9u/fb/G4pRFcTEwMGRkZBAQEmLXVaDRs3bqVRaf0/HHJ4AP36ZPdGdGhcW1u0eYY5e7fv79ZeKuioiIuXbpEfHw83t7et9e5pIPMfYYID97hENwFFNYJWSZJEnl5efj7+6NQKNizZw9Lly7lt99+Q6vVEh4ezoABA5g+fTrx8fG4ubnh4+NjNlrv168fX331FcnJyTRp0oTMzExGjhzJnj17AIMi0uv1ZvdvHMXXhps3b3Lq1CkaN26MSqUyO5abm0tISAg5OTnlvmOuRm5uLg0aNLD4WWg0Gn744QcGDhxol+zqer3EF0kXWPDDIVnRGfF0V/JMl5b8M/JtfDO+B+DaOT3fLJWYOGw5WjeDvK8O6mjbzONWxN6fd02wpNxMMX7ulX2/ymKXEVxgYCDR0dEkJSXJCi4pKYmYmJgqlRuUjxBfFi8vL7wshNfx8PCw+E8u0UkcuZwtl7s3jaz3XwYjZe9Jp9OhUChQKpVVfk4WufwNHH0RitJK67yjoO0yiBxaa3mN05IKhYJNmzYxcuRIFi5cyCeffELDhg1JT09n3bp17Nq1i8aNDS8Z+/btM5vONmK8P6VSyY8//ijXz58/n6SkJKtnwTa6jbi7u5f7fjjK98XVUSoVDO/QmEcTGpVTdMYR3e6Adhy+6weUCj0NmyiJb6sh6sZlLgbHAbUbUQjKU6zRMfO733lz+7FK2xk/96ldm1S7b7tFMhk7diyLFy+Wk0wuWbKE8ePHW2y7fv167rvvPvz9/Tl48CCvvfYakyZNstj2djifp0WjM/zwhvn7EBekquIMJ+XyN3DgCaDMoL7osqG+02dWUXJgGMlNmTKFmTNn8vzzz8v1ERERFU4915b4+HjeeustHnzwQQC2bdvGzJkzOXPmDFFRUSxdupQHHnjAJtcW1C8qU3RHcxuyOv0uno78DYA7H3Gj0XfJsoIDx1Fy9pxaLdbouJJXSHrOTS7nFMh/L+cUkp57qy73Jhn5Fce/LcvcTQcpKsyvdnu7Kbg5c+aQmZlJq1atABg1ahQzZ84EYMKECQC8//77AKxYsYJnn30WrVZLVFQUEydO5B//+IfVZDl1o9RAwOkcvCUJ9NVYi5N0cPQFyik3w0FAYRjZNexf9XSl0tNglVYJZ86cITk5mccee6xq2WzA0aNHGTZsGF9++SW9evVi3759DBo0iAMHDtCiRQu7yCSoeypSdAvO38eTDQ/g4S4RGKVk2h37CA9QkV4SwO7sJuhR1nslt+iHP9i5fwOPheWyc/9fLFLA7IG1V3K2UFymKNHTPfAcEZ65Zp+3kSU/JVW7L7spOA8PD1auXMnKlSvLHTMqNiO//PKLTWU5k1Oq4Jwug4C+BL4LskJHkmHa8vvQqpsOyQK3yiOwGzNoR0ZGynULFixg+fLlaLVaBg4cKGfr7t69O25upUp12rRpzJs3j9rwwQcfMGbMGPr06QNAt27dGDx4MOvXr2fOnDm16lvgeJgqusdXb2f9IdhwpiUjW58E4IGuV3iATwG4VKRm6umH+fp6O+ZuOsjSLUk09PdB5eWBv5cH/t6lf8vWqcodd8ffy1Ous1Z6rg3fv8Xo/GXM7nhDrruU+39s+P5Fhg1+3uI5ZRXX5ZxC0nMKzRTX5ZxCMgts5/j+UOgR3m7xFTHeJnKbfN4AM+9NYMmH1etPBFsGTueUWsQJB++6ISTE4Gd4+fJlea1t3rx5zJs3T15DM7J7926La3C1ITk5me3bt/Pxxx/LdVqt1uUNRFwdpVLBhkMXANhX1IzHpRPlZnSivG7wRdtVPHr0ab6+3o6bGh0Xs6o/bVYZHm5KC0rR/da+Z4V1BkVpOOfg/lWMVs6HMu+YUV43eEQzi8Uf51AYPLBOFJevpzuRDXyJbOBLRMCtv7fKpnVv7TzGud3v8nHb9eX6MH7eY38fTtPuU5jatQkVhwQxx+UV3Mvf/E5msWH9zU2poKNw8K4TmjdvTlxcHOvXr+eVV16p8+vHxMQwdepUXnvttTq/tqB+s2BQR+Zv+p2XW/8ClJ9qVyoMM/9ftF1Foc7DYhuborm1WdSpEn9z06Cg/CqBUe4Z6n9SWLIcfBTgA4TdhgwKUKBAoQClQnHreoayAgVKRalRVrlP5+at7Upp1Ry9Hu4yuNqUfaFQKkDSS6xs+jm0mUtNvGddWsEt3HyY5TtPyOWGKm/8vJzMGk7paZgyrApJB9vaQVE6ltfhFOAdCf2SqrcGVwUKhYK3336bUaNG4evry4gRI2jYsCHXr1/n+PHjVctbS/7+979z3333ce+999KjRw+0Wi2HDh1CrVbL68IC12TO/e2J0x8iRnOjwjaGH3JQuTuWg7/N5ZbK/K0mCqh03V6hVKAKgpzL21Goh1W7X5fNJmDJ5yI99yYLNx+u4AwHRaEwrIdVtbn7Qts3jCeV7cTwp+0yQ7uq+qqmkc7QoUPZtGkTP/zwA82bNycgIIDu3bvTsGFDli9fLrfr0qWLWaDtTp061fpjad++PWvXrmX27NmEhoYSFRXFnDlzzPwnnZXi4mKeeeYZGjVqhL+/Py1btmTVqlXy8dzcXEaOHElAQABhYWEsXLjQjtLah+HR9TdIgiuj0GVW3cgElxzBVeZQWN8to2xK5FCDK0BZPzifKLjjn1ZzETClR48e9OjRo8LjlcUhiI+Pt3h8/vz5FtvrdDo8PUtHl3369JGNTFwJrVZLREQE27Zto3Hjxvz222/cf//9REdHM2DAABITE8nKyiIlJYVr167Rr18/4uLieOqpp+wtep1RfNOT6oRJyA19lYAOj9tcnurw1o5jnNr5Ee93+b7KthP2DaZlr3E83/tvFo/rNCVoCgrQFhSgKchHW3jrb0EB2sLCW/v5hjaFBWgKTdoW5KMpLDQ7V19FVKWGTRT0m1z1zI/kFlyjCWGXU3BVecuDUHJEDIaMvVB8BbzCIaSr1SKZ2Iv09HSuXbsmBxZwZfz8/Hj11Vfl8j333EPv3r3Zs2cP3bp1Y926dezduxe1Wo1arSYxMZGPPvrIpRScxrst+VkSfmrD9FhZJL1EQTacv1CC/9U9dS+gBYYAUm5oteT+28ErtLz6XzZ/lIcmLw9Nfj6a/Dw0BYa/VSkka3PxkFQtuTXxbalaDZbiUgru9a1HWLLjdLXaurSSU7gZAs86Cdu2bWPYsGFMnjxZ+LhZoKioiAMHDjBy5EhOnz5NSUlJuUDolaWyup3g5vU9OLVWJ7F/jZa+k9yR9JLZj66kl0AB+9dqST5Yv1xK/IH9Z5VVyu19aC/JdZwkQaFU4u6nwkOlwkPlj7tKhYefYV9fUsL+NduqlPtvCRKKGnx3XErBLfkpCTx9qt1+3qaDrqngnIx+/fqRnZ1ddUMXRJIkxo8fT7NmzXj44YfZu3cvfn5+uLuX/jSo1epKg2PfTnDzskHC6xua5PPcOKjn55Va7hnpjsrElbQg26jc6mc2jGRrya1QoPD2QeHtXeavT2nZyxuFj0/5+lttlSb1eHqWs5DU3do0yee5MX9LlXLf2LMXzZlz1f4sXErBzbw3odojODCYCwsEzookSUycOJHTp0+zbds2lEolKpWKwsJCtFqtrORycnLw9/evsJ8ZM2Ywffp0uWwMbj5gwIAKg5uXDRJe38g8epjvMSiLi4dKCG+uwEet4OYNiStnJIxLvyF33o2nf/3xnSzJyyXjj9+qlLvVhCmom7c0jKb8VLir/G/tG/66+6lw9/Wts6hO1f28u3Xrike8A8SitAcv92+Ht6+qyjU4cKyI4QJBTZEkiUmTJvHbb7/x888/y0HOW7RogYeHB0eOHKFjR8MLXlJSEnfccUeFfdU0uHlVx+oDpiNYSYL00xKWbN+7vfEuIe061KFklZNx5BAbexmsjCuTu8WIUfVKblXDcNy8vNEVF1Uot5uXN6qG4ehr8L1xKQUHpWtq1UnL4Mg4WjJRR8BosekMsUonT57M3r172b59O4GBgXK9r68vjz32GHPmzGHt2rVcu3aNd9991yVdBQR1hyomlmG/n6AoyxDCT6vVsmfPXrp16yq/bHgHhaCKia1R6iuXU3BQuZJzdOXm6emJUqnk8uXLhIaG4mlh3tue6PV6SkpKKCoqur10PnZCkiQyMzPNpu4clYsXL/Lee+/h5eVFXFxphPxRo0bx/vvvs2LFCv7+978THR2Nj48PkydPdikLSjD8mBpHFBXh5uWNd1D9inzkqHKDQcmpYmIBw1S2R2o6wW3b12qk79hPai2wpOQcXbmBITdao0aNSE9P5/Lly/YWpxySJHHz5k18fHzqleKtLllZWWaBnx2RuLi4Sv0LAwICWLt2bR1KVP+oyYiiPuGoctsKl1VwYFByOr2OVzcnMff+BIdXbkY8PT2JjY1Fq9Wi0+nsLY4ZGo2GX375hR49etTrNZiKOHPmjL1FENQRthhR1AWOKrctcGkFBzBrQDvaa9MYOKCdvUWxKgqFol4u5Lu5uaHVavH29q53slVFfffdEggE5jjOIohAIBAIBDVAKDiBQCAQOCVCwQkEAoHAKXGJNTijxZgl/wmNRkNhYSG5ubkOtSYk5K57KpPd+N2qzDrRVRDPW/3BGeWuybPmEgrOGEcvJibGzpIInJ28vDw5KoirIp43QV1QnWdNIbnAK6der+fy5cv4+/uX870yxs27dOlSubh59Rkhd91TmeySJJGXl0dkZKRDObDbAvG81R+cUe6aPGsuMYJTKpVER0dX2iYgIMChvgBGhNx1T0Wyu/rIzYh43uofziZ3dZ81137VFAgEAoHTIhScQCAQCJwSl1dwXl5ezJs3z2K6j/qMkLvucWTZ6wuO+hkKuesWa8ntEkYmAoFAIHA9XH4EJxAIBALnRCg4gUAgEDglQsEJBAKBwClxaQWn0WiYPHkygYGBBAUFkZiYiFartbdYVbJixQruvPNOvLy8ePDBB+0tTrUoLi7mmWeeoVGjRvj7+9OyZUtWrVplb7GqTWJiIjExMQQEBBAVFcXzzz9PSUmJvcVyKBzxeXPEZw0c+3mz5rPm0gpu0aJF7NmzhxMnTnD8+HF2797NkiVL7C1WlURGRjJ79myeeeYZe4tSbbRaLREREWzbto3c3FxWr17NP/7xD7Zs2WJv0arFxIkTOXXqFLm5uRw5coQjR47wz3/+095iORSO+Lw54rMGjv28WfVZk1yY6OhoacOGDXJ5/fr1UmxsrB0lqhnz5s2Thg4dam8xbpuHHnpImjNnjr3FqDHXrl2T+vTpIz311FP2FsWhcOTnzdGfNUlyzOetts+ay47gsrOzSU1NJSEhQa5LSEggJSWFnJwc+wnmIhQVFXHgwAHatm1rb1GqzWuvvYZKpaJhw4YcOXKExMREe4vkMIjnzb442vNmrWfNZRVcfn4+AGq1Wq4z7hujoQtsgyRJjB8/nmbNmvHwww/bW5xq88orr5Cfn8+JEyeYMGEC4eHh9hbJYRDPm/1wxOfNWs+ayyo4lUoFYPb2aNz39/e3i0yugCRJTJw4kdOnT7Nx40aHjLzfqlUr2rVrx5gxY+wtisMgnjf74OjPW22fNce6WysSGBhIdHQ0SUlJcl1SUhIxMTEiKryNkCSJSZMm8dtvv7FlyxaH/pw1Gg1//fWXvcVwGMTzVvc4y/NWm2fNZRUcwNixY1m8eDFXrlzhypUrLFmyhPHjx9tbrCrRarUUFRWh1WrR6/UUFRU5hMn65MmT2bt3L1u3biUwMNDe4lSb/Px8Pv74Y27cuIEkSfz5558sWrSIe++9196iORSO+Lw56rMGjvm8Wf1Zs5Kxi0NSUlIiTZw4UVKr1ZJarZYmT54saTQae4tVJfPmzZMAs61nz572FqtSkpOTJUDy8vKS/Pz85O3vf/+7vUWrkvz8fKlfv35SUFCQ5OfnJzVq1Eh64YUXpIKCAnuL5lA44vPmiM+aJDnu82btZ00EWxYIBAKBU+LSU5QCgUAgcF6EghMIBAKBUyIUnEAgEAicEqHgBAKBQOCUCAUnEAgEAqdEKDiBQCAQOCVCwQkEAoHAKREKTiAQCAROiVBwtWT+/Pl2z/S7f/9+Wrdujb+/P++8845Nr9WmTRu+//77arVdsmQJjz/+uE3lsQe9evXirbfesrcYLol43ipGPG/lcSoF16tXL9zc3Dh69Khcd+PGDRQKBcnJyfYTzMbMnj2bxx9/nLy8PKZMmWJ27P7770elUqFSqfD09MTd3V0uGyO814Tjx48zePDgarWdOXMma9eurfE1qotCocDX19fsfh566CGbXU9gjnjexPNW3583p1JwYIhaPmPGDHuLcVtIkoROp6vxeRcuXOCOO+6weGzz5s3k5+eTn5/PzJkzGTx4sFw25ugyotVqb0tue7Jv3z6z+/n666/tLZJLIZ43c8TzVr9wOgU3ceJE9u7dyy+//GLx+JgxY3j++eflctk3zjFjxjBu3DgeffRRVCoVbdq04dixY3zwwQdER0cTGhrKe++9Z9anVqtl3LhxBAQE0KxZM7N/uiRJvPPOO7Rs2RK1Wk2vXr04efKkfDw+Pp6lS5dyzz334Ovry4kTJ8rJnJeXx7PPPktERAQRERFMmDCBgoICAMLDw7lw4QKPP/44KpWKM2fOVPuz6tWrFy+99BIDBgzAz8+PzZs3s2XLFu68804aNGhAREQEEydO5ObNm2bybty4EYDVq1eTkJDAwoULadiwIWFhYWZTCWWnkxQKBe+//z5/+9vfCAgI4IEHHjDLD/bLL79wxx134O/vz8MPP8y4ceNuOw/Uzp07UavVvPvuu0RERBAeHs68efMwDb36f//3f7Rq1Qq1Wk23bt04dOiQfKykpIS5c+fSpEkT/P39ueOOO8yOX716lXvvvRd/f386dOjAn3/+KR978803iY2Nxd/fn/j4eP773//e1j04AuJ5E88b1OPnzUpBoOsFPXv2lJYvXy4tWbJE6ty5syRJkpSdnS0B0oULFyRJkqTRo0dLU6dOlc+xdNzf31/as2ePpNFopNGjR0uNGzeWXnzxRam4uFjatm2b5OnpKV25ckWSJEO0cTc3N+n999+XNBqN9O2330peXl7S2bNnJUmSpJUrV0pt27aVzpw5I2k0Guntt9+WmjRpIhUXF0uSJElxcXFS8+bNpVOnTklarVauN2Xs2LFS7969pYyMDOn69etSz549pWeeeUY+HhcXJ3399ddVfj7z5s2Thg4davZ5hYaGSr/99puk1+ulwsJC6ZdffpEOHTokabVa6dy5c1LLli2lRYsWWbzWxx9/LLm7u0tvvPGGVFJSIu3YsUNyd3eX773s9QCpd+/e0tWrV6Xs7Gypffv20rx58yRJkqSsrCypQYMG0kcffSRpNBpp06ZNkpeXlzR69OgK7weQDh8+bPHYjh07JKVSKY0ZM0YqKCiQTp48KUVHR0urV6+WJEmSdu3aJalUKmnXrl1SSUmJtHz5cik0NFS6ceOGJEmSNG3aNKljx47SmTNnJL1eL506dUpKTk6WP7eoqCgpKSlJ0mg00jPPPCNHmD99+rTk4+MjnTx5UpIkSbpy5Yp05MiRKv83joh43ipHPG/2f96cUsEVFhZKkZGR0tdff31bD9yIESPk45s2bZKUSqVUWFgo14WGhkpbt26VJMnwpWrVqpWZHPfdd5+0cOFCSZIkqXXr1tLGjRvNjkdGRkq//PKLJEmGL/Dy5csrvCedTid5enpK+/fvl+v27t0reXl5STqdTu7jdh8408/CEsuXL5f69esnl8s+cOHh4WbtmzZtKn3xxRcWrwdImzdvlsuLFi2SBg8eLEmSJH366adSmzZtzPoaOHBglQ+cSqWSGjRoIG/z58+XJMnwwAHS1atX5favvfaa1LdvX0mSJGn8+PHShAkTzPpr3ry59Nlnn0l6vV7y9fWVdu3aZfG6PXv2lF5++WW5vGfPHkmlUkmSJElnz56VvL29pS+++MLsO+OMiOetcsTzZv/nzemmKAF8fHyYN28eM2fOvK059rCwMHnf19cXf39/fHx8zOpM59Pj4uLMzo+LiyMtLQ2A5ORkRo0ahVqtlrfs7GxSU1Pl9rGxsRXKcv36dUpKSoiPj5frGjduTHFxMRkZGTW+t7KUvfbvv/9Ov379CAsLIyAggJkzZ1Z6HdPPCsDPz4+8vLwK24eHh1tse/nyZWJiYiqVzRK7d+/mxo0b8jZv3jz5mLe3Nw0bNpTLpv+X1NRUs88UoFGjRqSmpnL9+nUKCwtp1qxZte/D+H1o0qQJn3zyCStWrCAsLIwBAwaYZbF2RsTzVn3E81ZKXTxvTqngAMaNG4der+eTTz4xq1epVBQWFsrl9PT0Wl/r4sWLZuWUlBSioqIAiImJYcOGDWZfisLCQjNzXqWy4n9DaGgonp6eZlZpycnJeHl5ERISUmvZy1778ccfp3fv3pw/f57c3FyWLFliNo9uKyIjI7l06ZJZXUpKSq36LCoq4tq1a2b9Gf8v0dHR5Sz9kpOT5XUfX19fzp49e1vXHT58ODt27ODq1au0a9eOJ5988rbvwVEQz1v1EM9bKXXxvDmtgnNzc2Px4sUsWbLErL5Dhw789NNPpKenk5eXx4IFC2p9rTNnzvDhhx+i1WrZtGkT27dv57HHHgNg0qRJzJ07l9OnTwOQm5vLN998U+lblylKpZKRI0cya9YssrKyyMzMZObMmTz55JOVPqi3S25uLmq1Gj8/P06ePMm///1vq1/DEoMGDeLSpUusXr0arVbLjz/+yPbt22vVp1KpZMaMGdy8eZPTp0+zcuVKnnjiCQBGjRrFZ599xt69e9Fqtbz77rtkZmYycOBAFAoFzzzzDP/4xz84e/YskiRx+vTpcj+sljh9+jRbt27l5s2beHp6olKpcHd3r9V9OALiebs9xPNm2+fNaRUcwCOPPELTpk3N6kaNGkXPnj1p2bIlCQkJDBo0qNbXue+++9i/fz9BQUFMnTqV//u//5OH25MnT2bMmDE8/PDDBAQE0KpVK9asWVOj/t9++23i4+Np3bo1bdq0oWnTprz55pu1ltsSH3zwAW+88QYqlYoJEyYwYsQIm1ynLEFBQWzcuJE33ngDtVrNf/7zH4YNG4aXl1el53Xp0sXML6dTp07yMX9/fxISEmjcuDE9evTgqaeeYvTo0QD07NmTd999l3HjxhEcHMy6devYvHkzarUagNdff52+ffvSr18/AgICGDZsGFlZWVXeR0lJCXPmzCEsLIzg4GC2b9/O6tWrb/tzcSTE81ZzxPOmBmz3vCmkuhgPCwS3wb333kuPHj2YNWtWjc/duXMnDz74IDdu3LC+YAKBE+KMz5tTj+AEjsWWLVvIyMhAq9Wybt06tm/fzsMPP2xvsQQCp8QVnjfnXxwQOAwHDx7kiSeeoLCwkEaNGrF27VpatWplb7EEAqfEFZ43MUUpEAgEAqdETFEKBAKBwCkRCk4gEAgETolQcAKBQCBwSoSCEwgEAoFTIhScQCAQCJwSoeAEAoFA4JQIBScQCAQCp0QoOIFAIBA4JULBCQQCgcAp+X89u61CH9B2HwAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 450x325 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# First plot\n",
    "plt.figure(figsize=(4.5, 3.25))  # Adjust figsize as needed\n",
    "colors = ['#006B9B', '#9B1003',  '#FFAF0A']\n",
    "\n",
    "markers = ['D', 's', 'o']  # Square, Star, Diamond\n",
    "marker_size = 14\n",
    "line_width = 2.2\n",
    "\n",
    "\n",
    "plt.subplot(1, 2, 1)  # Subplot with 1 row, 2 columns, position 1\n",
    "df = overview.copy()\n",
    "# for idx, model in enumerate(df['base_model'].unique()):\n",
    "for idx, model in enumerate(['Llama', 'Mistral', 'GEITje']):\n",
    "\n",
    "    model_df = df[df['base_model'] == model]\n",
    "    plt.plot(model_df['epoch'], model_df['accuracy'], marker=markers[idx], label=model, color=colors[idx % len(colors)], linewidth=line_width)\n",
    "\n",
    "fs = 9\n",
    "plt.xticks(fontsize=fs)\n",
    "plt.yticks(fontsize=fs)\n",
    "plt.xticks([0,1, 2, 3],fontsize=fs)\n",
    "plt.xlabel('Number of Training Epochs', fontsize=fs)\n",
    "plt.ylabel('Accuracy', fontsize=fs)\n",
    "plt.legend(title='Model', fontsize=fs, title_fontsize=fs)\n",
    "plt.grid(True)\n",
    "\n",
    "# Second plot\n",
    "plt.subplot(1, 2, 2)  # Subplot with 1 row, 2 columns, position 2\n",
    "df = overview.copy()\n",
    "for idx, model in enumerate(df['base_model'].unique()):\n",
    "    model_df = df[df['base_model'] == model]\n",
    "    plt.plot(model_df['epoch'], model_df['runtime_per_doc'], marker=markers[idx], label=model, color=colors[idx % len(colors)], linewidth=line_width)\n",
    "\n",
    "plt.xticks(fontsize=fs)\n",
    "plt.yticks(fontsize=fs)\n",
    "plt.xticks([0,1, 2, 3],fontsize=fs)\n",
    "plt.xlabel('Number of Training Epochs', fontsize=fs)\n",
    "plt.ylabel('Inference Time (in seconds)', fontsize=fs)\n",
    "plt.legend(title='Model', fontsize=fs, title_fontsize=fs)\n",
    "plt.grid(True)\n",
    "\n",
    "plt.tight_layout()  # Ensures plots are nicely spaced\n",
    "plt.show()\n"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.19"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}