7.4. Tabnet Gpu Oversample.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "RgYw4D9wZXVw",
    "outputId": "fdbac52c-bbf0-4fcc-9426-3ce8cd4225bb"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using PyTorch version: 1.12.1+cu113, Device: cuda\n"
     ]
    }
   ],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "from IPython.display import display\n",
    "from tqdm import tqdm\n",
    "import os\n",
    "import random\n",
    "import pickle\n",
    "import warnings\n",
    "import gc\n",
    "\n",
    "from sklearn.preprocessing import MinMaxScaler, StandardScaler, LabelEncoder\n",
    "from sklearn.model_selection import train_test_split, cross_val_score, StratifiedKFold\n",
    "from sklearn.metrics import (precision_recall_curve, roc_curve, make_scorer, log_loss,\n",
    "                             f1_score, confusion_matrix, auc, roc_auc_score, accuracy_score)\n",
    "\n",
    "import torch\n",
    "from torch import nn\n",
    "from torch.utils.data import DataLoader, Dataset\n",
    "import torch.optim as optim\n",
    "import torch.nn.functional as F\n",
    "from pytorch_tabnet.tab_model import TabNetClassifier\n",
    "from pytorch_tabnet.metrics import Metric\n",
    "from pytorch_tabnet.pretraining import TabNetPretrainer\n",
    "from torch.optim.lr_scheduler import ReduceLROnPlateau\n",
    "\n",
    "import optuna\n",
    "from optuna.samplers import TPESampler\n",
    "\n",
    "DEVICE = torch.device('cuda') if torch.cuda.is_available() else torch.device('cpu')\n",
    "RANDOM_STATE = 4158\n",
    "CHUNKSIZE = 50000\n",
    "\n",
    "with open(\"../Data/dtypes.pkl\", 'rb') as f:\n",
    "    dtypes = pickle.load(f)\n",
    "\n",
    "pd.set_option('display.max_columns', None)\n",
    "warnings.filterwarnings('ignore')\n",
    "print(f\"Using PyTorch version: {torch.__version__}, Device: {DEVICE}\")\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "19h9pJhCZXVx"
   },
   "outputs": [],
   "source": [
    "def seed_everything(seed_value):\n",
    "    np.random.seed(seed_value)\n",
    "    torch.manual_seed(seed_value)\n",
    "    os.environ['PYTHONHASHSEED'] = str(seed_value)\n",
    "    \n",
    "    if torch.cuda.is_available(): \n",
    "        torch.cuda.manual_seed(seed_value)\n",
    "        torch.cuda.manual_seed_all(seed_value)\n",
    "        torch.backends.cudnn.deterministic = True\n",
    "        torch.backends.cudnn.benchmark = False\n",
    "        \n",
    "seed_everything(RANDOM_STATE)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "M8MpMFkkyI1M"
   },
   "outputs": [],
   "source": [
    "def viewinfo(df:pd.DataFrame) -> None:\n",
    "    print(df.shape)\n",
    "    print()\n",
    "    df.info()\n",
    "    print()\n",
    "    display(df.head())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "WH4BHNyMZXVy",
    "outputId": "49ee530a-3614-4738-e527-e98709c8a5c3"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'loan_limit': 'uint64', 'loan_rate': 'float16', 'credit_score': 'uint16', 'yearly_income': 'uint64', 'income_type': 'uint8', 'employment_type': 'uint8', 'houseown_type': 'uint8', 'desired_amount': 'uint64', 'purpose': 'uint8', 'personal_rehabilitation_yn': 'bool', 'personal_rehabilitation_complete_yn': 'bool', 'existing_loan_cnt': 'uint16', 'existing_loan_amt': 'uint64', 'enter_birth': 'uint8', 'gender': 'bool', 'k_lending_rate': 'float16', 'us_lending_rate': 'float16', 'btc_price': 'uint32', 'btc_variance': 'float16', 'eth_price': 'uint32', 'eth_variance': 'float16', 'lunc_price': 'float32', 'lunc_variance': 'float16', 'Close': 'float16', 'Open': 'float16', 'High': 'float16', 'Low': 'float16', 'Volume': 'uint64', 'Change': 'float16', '실업자': 'uint16', '실업률': 'float16', '현재가계부채CSI': 'uint8', '가계부채전망CSI': 'uint8', '소비심리지수': 'float16', 'loanapply_insert_time_rank': 'uint8', 'loan_limit_avg_sub': 'float64', 'loan_rate_avg_sub': 'float16', 'bank_diversity': 'uint8', 'product_diversity': 'uint8', 'missing_value_cnt': 'uint8', 'working_year': 'uint8', 'working_month': 'uint16', 'age': 'uint8', 'holiday': 'bool', 'weekday': 'uint8', 'hour': 'uint8', 'weekday_sin': 'float16', 'weekday_cos': 'float16', 'hour_sin': 'float16', 'hour_cos': 'float16', 'foreign': 'bool', 'product_cnt': 'uint16', 'event': 'uint16', 'loan_cnt': 'uint16', 'duration': 'uint32', 'visit': 'uint16', 'segment': 'uint8', 'loan_limit_mean_x': 'float64', 'loan_rate_mean_x': 'float16', 'loan_limit_min_x': 'float64', 'loan_rate_min_x': 'float16', 'loan_limit_max_x': 'float64', 'loan_rate_max_x': 'float16', 'bank_label': 'uint8', 'loan_limit_min_y': 'float64', 'loan_rate_min_y': 'float16', 'loan_limit_max_y': 'float64', 'loan_rate_max_y': 'float16', 'loan_limit_mean_y': 'float64', 'loan_rate_mean_y': 'float16', 'product_label': 'uint8', 'is_applied': 'bool'}\n"
     ]
    }
   ],
   "source": [
    "print(dtypes)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "UKxLGTuPZXVz",
    "outputId": "036fceb8-40ca-482d-990b-c51a35566dd9"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "96"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gc.collect()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "SlvZxiz2rJS2"
   },
   "source": [
    "# Data Load"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "JK6pqugaS_L3"
   },
   "outputs": [],
   "source": [
    "def loaddata(fname:str, chunksize:int, dtype:dict=None, columns:list=None):  \n",
    "    df = pd.DataFrame()\n",
    "    for chunk in tqdm(pd.read_csv(fname, engine='python', low_memory=True, chunksize=chunksize, dtype=dtype)):\n",
    "        df = pd.concat([df, chunk], axis=0)\n",
    "        del chunk\n",
    "        gc.collect()\n",
    "\n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "d3oxUtH2ZXV0",
    "outputId": "1921015a-85d2-4221-cff4-99ec4a935a52"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "109it [04:45,  2.62s/it]\n",
      "19it [00:47,  2.50s/it]\n"
     ]
    }
   ],
   "source": [
    "train = loaddata(\"../Data/oversample_train.csv\", chunksize=CHUNKSIZE, dtype=dtypes)\n",
    "valid = loaddata(\"../Data/master_public_data.csv\", chunksize=CHUNKSIZE, dtype=dtypes)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "3jbqcaaZ-bdD",
    "outputId": "f88802b3-cb96-4fef-905b-99249635677e"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(5437418, 72)\n",
      "\n",
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 5437418 entries, 0 to 5437417\n",
      "Data columns (total 72 columns):\n",
      " #   Column                               Dtype  \n",
      "---  ------                               -----  \n",
      " 0   loan_limit                           uint64 \n",
      " 1   loan_rate                            float16\n",
      " 2   credit_score                         uint16 \n",
      " 3   yearly_income                        uint64 \n",
      " 4   income_type                          uint8  \n",
      " 5   employment_type                      uint8  \n",
      " 6   houseown_type                        uint8  \n",
      " 7   desired_amount                       uint64 \n",
      " 8   purpose                              uint8  \n",
      " 9   personal_rehabilitation_yn           bool   \n",
      " 10  personal_rehabilitation_complete_yn  bool   \n",
      " 11  existing_loan_cnt                    uint16 \n",
      " 12  existing_loan_amt                    uint64 \n",
      " 13  enter_birth                          uint8  \n",
      " 14  gender                               bool   \n",
      " 15  k_lending_rate                       float16\n",
      " 16  us_lending_rate                      float16\n",
      " 17  btc_price                            uint32 \n",
      " 18  btc_variance                         float16\n",
      " 19  eth_price                            uint32 \n",
      " 20  eth_variance                         float16\n",
      " 21  lunc_price                           float32\n",
      " 22  lunc_variance                        float16\n",
      " 23  Close                                float16\n",
      " 24  Open                                 float16\n",
      " 25  High                                 float16\n",
      " 26  Low                                  float16\n",
      " 27  Volume                               uint64 \n",
      " 28  Change                               float16\n",
      " 29  실업자                                  uint16 \n",
      " 30  실업률                                  float16\n",
      " 31  현재가계부채CSI                            uint8  \n",
      " 32  가계부채전망CSI                            uint8  \n",
      " 33  소비심리지수                               float16\n",
      " 34  loanapply_insert_time_rank           uint8  \n",
      " 35  loan_limit_avg_sub                   float64\n",
      " 36  loan_rate_avg_sub                    float16\n",
      " 37  bank_diversity                       uint8  \n",
      " 38  product_diversity                    uint8  \n",
      " 39  missing_value_cnt                    uint8  \n",
      " 40  working_year                         uint8  \n",
      " 41  working_month                        uint16 \n",
      " 42  age                                  uint8  \n",
      " 43  holiday                              bool   \n",
      " 44  weekday                              uint8  \n",
      " 45  hour                                 uint8  \n",
      " 46  weekday_sin                          float16\n",
      " 47  weekday_cos                          float16\n",
      " 48  hour_sin                             float16\n",
      " 49  hour_cos                             float16\n",
      " 50  foreign                              bool   \n",
      " 51  event                                uint16 \n",
      " 52  loan_cnt                             uint16 \n",
      " 53  duration                             uint32 \n",
      " 54  visit                                uint16 \n",
      " 55  segment                              uint8  \n",
      " 56  product_cnt                          uint16 \n",
      " 57  loan_limit_mean_x                    float64\n",
      " 58  loan_rate_mean_x                     float16\n",
      " 59  loan_limit_min_x                     float64\n",
      " 60  loan_rate_min_x                      float16\n",
      " 61  loan_limit_max_x                     float64\n",
      " 62  loan_rate_max_x                      float16\n",
      " 63  bank_label                           uint8  \n",
      " 64  loan_limit_min_y                     float64\n",
      " 65  loan_rate_min_y                      float16\n",
      " 66  loan_limit_max_y                     float64\n",
      " 67  loan_rate_max_y                      float16\n",
      " 68  loan_limit_mean_y                    float64\n",
      " 69  loan_rate_mean_y                     float16\n",
      " 70  product_label                        uint8  \n",
      " 71  is_applied                           bool   \n",
      "dtypes: bool(6), float16(24), float32(1), float64(7), uint16(8), uint32(3), uint64(5), uint8(18)\n",
      "memory usage: 1.0 GB\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "\n",
       "  <div id=\"df-5c04fbd5-5456-4969-8a4e-6801e2f4ad57\">\n",
       "    <div class=\"colab-df-container\">\n",
       "      <div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>loan_limit</th>\n",
       "      <th>loan_rate</th>\n",
       "      <th>credit_score</th>\n",
       "      <th>yearly_income</th>\n",
       "      <th>income_type</th>\n",
       "      <th>employment_type</th>\n",
       "      <th>houseown_type</th>\n",
       "      <th>desired_amount</th>\n",
       "      <th>purpose</th>\n",
       "      <th>personal_rehabilitation_yn</th>\n",
       "      <th>personal_rehabilitation_complete_yn</th>\n",
       "      <th>existing_loan_cnt</th>\n",
       "      <th>existing_loan_amt</th>\n",
       "      <th>enter_birth</th>\n",
       "      <th>gender</th>\n",
       "      <th>k_lending_rate</th>\n",
       "      <th>us_lending_rate</th>\n",
       "      <th>btc_price</th>\n",
       "      <th>btc_variance</th>\n",
       "      <th>eth_price</th>\n",
       "      <th>eth_variance</th>\n",
       "      <th>lunc_price</th>\n",
       "      <th>lunc_variance</th>\n",
       "      <th>Close</th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Volume</th>\n",
       "      <th>Change</th>\n",
       "      <th>실업자</th>\n",
       "      <th>실업률</th>\n",
       "      <th>현재가계부채CSI</th>\n",
       "      <th>가계부채전망CSI</th>\n",
       "      <th>소비심리지수</th>\n",
       "      <th>loanapply_insert_time_rank</th>\n",
       "      <th>loan_limit_avg_sub</th>\n",
       "      <th>loan_rate_avg_sub</th>\n",
       "      <th>bank_diversity</th>\n",
       "      <th>product_diversity</th>\n",
       "      <th>missing_value_cnt</th>\n",
       "      <th>working_year</th>\n",
       "      <th>working_month</th>\n",
       "      <th>age</th>\n",
       "      <th>holiday</th>\n",
       "      <th>weekday</th>\n",
       "      <th>hour</th>\n",
       "      <th>weekday_sin</th>\n",
       "      <th>weekday_cos</th>\n",
       "      <th>hour_sin</th>\n",
       "      <th>hour_cos</th>\n",
       "      <th>foreign</th>\n",
       "      <th>event</th>\n",
       "      <th>loan_cnt</th>\n",
       "      <th>duration</th>\n",
       "      <th>visit</th>\n",
       "      <th>segment</th>\n",
       "      <th>product_cnt</th>\n",
       "      <th>loan_limit_mean_x</th>\n",
       "      <th>loan_rate_mean_x</th>\n",
       "      <th>loan_limit_min_x</th>\n",
       "      <th>loan_rate_min_x</th>\n",
       "      <th>loan_limit_max_x</th>\n",
       "      <th>loan_rate_max_x</th>\n",
       "      <th>bank_label</th>\n",
       "      <th>loan_limit_min_y</th>\n",
       "      <th>loan_rate_min_y</th>\n",
       "      <th>loan_limit_max_y</th>\n",
       "      <th>loan_rate_max_y</th>\n",
       "      <th>loan_limit_mean_y</th>\n",
       "      <th>loan_rate_mean_y</th>\n",
       "      <th>product_label</th>\n",
       "      <th>is_applied</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>14000000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>710</td>\n",
       "      <td>28000000</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>40000000</td>\n",
       "      <td>7</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>4</td>\n",
       "      <td>67000000</td>\n",
       "      <td>29</td>\n",
       "      <td>True</td>\n",
       "      <td>1.50</td>\n",
       "      <td>1.00</td>\n",
       "      <td>39120756</td>\n",
       "      <td>-0.026001</td>\n",
       "      <td>2640599</td>\n",
       "      <td>-0.080688</td>\n",
       "      <td>0.000322</td>\n",
       "      <td>-0.999512</td>\n",
       "      <td>2550.0</td>\n",
       "      <td>2570.0</td>\n",
       "      <td>2592.0</td>\n",
       "      <td>2546.0</td>\n",
       "      <td>900330000</td>\n",
       "      <td>-0.016296</td>\n",
       "      <td>76</td>\n",
       "      <td>2.300781</td>\n",
       "      <td>102</td>\n",
       "      <td>101</td>\n",
       "      <td>94.0000</td>\n",
       "      <td>10</td>\n",
       "      <td>-4.161290e+06</td>\n",
       "      <td>-1.806641</td>\n",
       "      <td>22</td>\n",
       "      <td>31</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>9</td>\n",
       "      <td>31</td>\n",
       "      <td>False</td>\n",
       "      <td>4</td>\n",
       "      <td>11</td>\n",
       "      <td>-0.433838</td>\n",
       "      <td>-0.900879</td>\n",
       "      <td>0.258789</td>\n",
       "      <td>-0.965820</td>\n",
       "      <td>False</td>\n",
       "      <td>10</td>\n",
       "      <td>5</td>\n",
       "      <td>141</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>1.668727e+07</td>\n",
       "      <td>10.851562</td>\n",
       "      <td>1000000.0</td>\n",
       "      <td>5.898438</td>\n",
       "      <td>100000000.0</td>\n",
       "      <td>16.296875</td>\n",
       "      <td>2</td>\n",
       "      <td>6000000.0</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>20000000.0</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>1.355632e+07</td>\n",
       "      <td>9.054688</td>\n",
       "      <td>1</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>9000000</td>\n",
       "      <td>6.601562</td>\n",
       "      <td>580</td>\n",
       "      <td>36000000</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>3000000</td>\n",
       "      <td>7</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>5</td>\n",
       "      <td>78000000</td>\n",
       "      <td>49</td>\n",
       "      <td>False</td>\n",
       "      <td>1.50</td>\n",
       "      <td>0.50</td>\n",
       "      <td>50162048</td>\n",
       "      <td>-0.027496</td>\n",
       "      <td>3790690</td>\n",
       "      <td>-0.029800</td>\n",
       "      <td>81.518517</td>\n",
       "      <td>-0.071899</td>\n",
       "      <td>2716.0</td>\n",
       "      <td>2720.0</td>\n",
       "      <td>2720.0</td>\n",
       "      <td>2702.0</td>\n",
       "      <td>665350000</td>\n",
       "      <td>0.000100</td>\n",
       "      <td>47</td>\n",
       "      <td>1.599609</td>\n",
       "      <td>102</td>\n",
       "      <td>101</td>\n",
       "      <td>91.8125</td>\n",
       "      <td>1</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>-0.657227</td>\n",
       "      <td>7</td>\n",
       "      <td>7</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>13</td>\n",
       "      <td>51</td>\n",
       "      <td>False</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>-0.433838</td>\n",
       "      <td>-0.900879</td>\n",
       "      <td>0.866211</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>True</td>\n",
       "      <td>6</td>\n",
       "      <td>2</td>\n",
       "      <td>354</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>1.668727e+07</td>\n",
       "      <td>10.851562</td>\n",
       "      <td>1000000.0</td>\n",
       "      <td>5.898438</td>\n",
       "      <td>100000000.0</td>\n",
       "      <td>16.296875</td>\n",
       "      <td>2</td>\n",
       "      <td>2000000.0</td>\n",
       "      <td>6.500000</td>\n",
       "      <td>20000000.0</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>1.212646e+07</td>\n",
       "      <td>6.613281</td>\n",
       "      <td>1</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1000000</td>\n",
       "      <td>12.500000</td>\n",
       "      <td>780</td>\n",
       "      <td>25000000</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>40000000</td>\n",
       "      <td>7</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>3</td>\n",
       "      <td>237000000</td>\n",
       "      <td>33</td>\n",
       "      <td>False</td>\n",
       "      <td>1.25</td>\n",
       "      <td>0.25</td>\n",
       "      <td>48605860</td>\n",
       "      <td>0.003300</td>\n",
       "      <td>3289621</td>\n",
       "      <td>0.013397</td>\n",
       "      <td>87.549965</td>\n",
       "      <td>0.050507</td>\n",
       "      <td>2714.0</td>\n",
       "      <td>2736.0</td>\n",
       "      <td>2736.0</td>\n",
       "      <td>2702.0</td>\n",
       "      <td>765680000</td>\n",
       "      <td>-0.012199</td>\n",
       "      <td>65</td>\n",
       "      <td>3.099609</td>\n",
       "      <td>103</td>\n",
       "      <td>102</td>\n",
       "      <td>93.8125</td>\n",
       "      <td>1</td>\n",
       "      <td>-8.391304e+06</td>\n",
       "      <td>1.521484</td>\n",
       "      <td>20</td>\n",
       "      <td>23</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>16</td>\n",
       "      <td>35</td>\n",
       "      <td>True</td>\n",
       "      <td>6</td>\n",
       "      <td>22</td>\n",
       "      <td>-0.781738</td>\n",
       "      <td>0.623535</td>\n",
       "      <td>-0.500000</td>\n",
       "      <td>0.866211</td>\n",
       "      <td>False</td>\n",
       "      <td>9</td>\n",
       "      <td>3</td>\n",
       "      <td>301</td>\n",
       "      <td>1</td>\n",
       "      <td>6</td>\n",
       "      <td>4</td>\n",
       "      <td>1.668727e+07</td>\n",
       "      <td>10.851562</td>\n",
       "      <td>1000000.0</td>\n",
       "      <td>5.898438</td>\n",
       "      <td>100000000.0</td>\n",
       "      <td>16.296875</td>\n",
       "      <td>2</td>\n",
       "      <td>1000000.0</td>\n",
       "      <td>5.898438</td>\n",
       "      <td>100000000.0</td>\n",
       "      <td>16.296875</td>\n",
       "      <td>2.030726e+07</td>\n",
       "      <td>13.828125</td>\n",
       "      <td>0</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>40000000</td>\n",
       "      <td>7.898438</td>\n",
       "      <td>990</td>\n",
       "      <td>40000000</td>\n",
       "      <td>5</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>270000000</td>\n",
       "      <td>3</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>1</td>\n",
       "      <td>55000000</td>\n",
       "      <td>35</td>\n",
       "      <td>True</td>\n",
       "      <td>1.50</td>\n",
       "      <td>1.00</td>\n",
       "      <td>37052276</td>\n",
       "      <td>-0.055389</td>\n",
       "      <td>2468705</td>\n",
       "      <td>-0.083191</td>\n",
       "      <td>0.000061</td>\n",
       "      <td>-0.195312</td>\n",
       "      <td>2626.0</td>\n",
       "      <td>2634.0</td>\n",
       "      <td>2644.0</td>\n",
       "      <td>2618.0</td>\n",
       "      <td>870270000</td>\n",
       "      <td>0.002100</td>\n",
       "      <td>65</td>\n",
       "      <td>1.700195</td>\n",
       "      <td>103</td>\n",
       "      <td>100</td>\n",
       "      <td>92.5000</td>\n",
       "      <td>7</td>\n",
       "      <td>1.133333e+07</td>\n",
       "      <td>-2.546875</td>\n",
       "      <td>18</td>\n",
       "      <td>21</td>\n",
       "      <td>2</td>\n",
       "      <td>15</td>\n",
       "      <td>165</td>\n",
       "      <td>50</td>\n",
       "      <td>False</td>\n",
       "      <td>3</td>\n",
       "      <td>14</td>\n",
       "      <td>0.433838</td>\n",
       "      <td>-0.900879</td>\n",
       "      <td>-0.500000</td>\n",
       "      <td>-0.866211</td>\n",
       "      <td>False</td>\n",
       "      <td>10</td>\n",
       "      <td>5</td>\n",
       "      <td>503</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>7</td>\n",
       "      <td>2.464846e+07</td>\n",
       "      <td>12.179688</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4.800781</td>\n",
       "      <td>150000000.0</td>\n",
       "      <td>19.906250</td>\n",
       "      <td>2</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4.800781</td>\n",
       "      <td>150000000.0</td>\n",
       "      <td>19.906250</td>\n",
       "      <td>2.418366e+07</td>\n",
       "      <td>12.664062</td>\n",
       "      <td>0</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>6000000</td>\n",
       "      <td>15.796875</td>\n",
       "      <td>910</td>\n",
       "      <td>24000000</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>10000000</td>\n",
       "      <td>7</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>12</td>\n",
       "      <td>111000000</td>\n",
       "      <td>40</td>\n",
       "      <td>False</td>\n",
       "      <td>1.75</td>\n",
       "      <td>1.00</td>\n",
       "      <td>39541500</td>\n",
       "      <td>0.057404</td>\n",
       "      <td>2493979</td>\n",
       "      <td>0.082214</td>\n",
       "      <td>0.000127</td>\n",
       "      <td>0.234253</td>\n",
       "      <td>2670.0</td>\n",
       "      <td>2662.0</td>\n",
       "      <td>2672.0</td>\n",
       "      <td>2654.0</td>\n",
       "      <td>469800000</td>\n",
       "      <td>0.012001</td>\n",
       "      <td>71</td>\n",
       "      <td>2.699219</td>\n",
       "      <td>102</td>\n",
       "      <td>101</td>\n",
       "      <td>93.3125</td>\n",
       "      <td>14</td>\n",
       "      <td>-7.500000e+06</td>\n",
       "      <td>4.148438</td>\n",
       "      <td>26</td>\n",
       "      <td>36</td>\n",
       "      <td>1</td>\n",
       "      <td>9</td>\n",
       "      <td>100</td>\n",
       "      <td>49</td>\n",
       "      <td>False</td>\n",
       "      <td>1</td>\n",
       "      <td>18</td>\n",
       "      <td>0.781738</td>\n",
       "      <td>0.623535</td>\n",
       "      <td>-1.000000</td>\n",
       "      <td>-0.000000</td>\n",
       "      <td>False</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>900</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>1.861316e+07</td>\n",
       "      <td>13.226562</td>\n",
       "      <td>0.0</td>\n",
       "      <td>7.199219</td>\n",
       "      <td>50000000.0</td>\n",
       "      <td>19.906250</td>\n",
       "      <td>2</td>\n",
       "      <td>0.0</td>\n",
       "      <td>13.398438</td>\n",
       "      <td>50000000.0</td>\n",
       "      <td>19.906250</td>\n",
       "      <td>2.121015e+07</td>\n",
       "      <td>15.804688</td>\n",
       "      <td>0</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>\n",
       "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-5c04fbd5-5456-4969-8a4e-6801e2f4ad57')\"\n",
       "              title=\"Convert this dataframe to an interactive table.\"\n",
       "              style=\"display:none;\">\n",
       "        \n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "       width=\"24px\">\n",
       "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
       "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
       "  </svg>\n",
       "      </button>\n",
       "      \n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      flex-wrap:wrap;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "      <script>\n",
       "        const buttonEl =\n",
       "          document.querySelector('#df-5c04fbd5-5456-4969-8a4e-6801e2f4ad57 button.colab-df-convert');\n",
       "        buttonEl.style.display =\n",
       "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "        async function convertToInteractive(key) {\n",
       "          const element = document.querySelector('#df-5c04fbd5-5456-4969-8a4e-6801e2f4ad57');\n",
       "          const dataTable =\n",
       "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                     [key], {});\n",
       "          if (!dataTable) return;\n",
       "\n",
       "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "            + ' to learn more about interactive tables.';\n",
       "          element.innerHTML = '';\n",
       "          dataTable['output_type'] = 'display_data';\n",
       "          await google.colab.output.renderOutput(dataTable, element);\n",
       "          const docLink = document.createElement('div');\n",
       "          docLink.innerHTML = docLinkHtml;\n",
       "          element.appendChild(docLink);\n",
       "        }\n",
       "      </script>\n",
       "    </div>\n",
       "  </div>\n",
       "  "
      ],
      "text/plain": [
       "   loan_limit  loan_rate  credit_score  yearly_income  income_type  \\\n",
       "0    14000000   9.000000           710       28000000            0   \n",
       "1     9000000   6.601562           580       36000000            1   \n",
       "2     1000000  12.500000           780       25000000            0   \n",
       "3    40000000   7.898438           990       40000000            5   \n",
       "4     6000000  15.796875           910       24000000            0   \n",
       "\n",
       "   employment_type  houseown_type  desired_amount  purpose  \\\n",
       "0                3              3        40000000        7   \n",
       "1                3              3         3000000        7   \n",
       "2                3              3        40000000        7   \n",
       "3                3              2       270000000        3   \n",
       "4                3              2        10000000        7   \n",
       "\n",
       "   personal_rehabilitation_yn  personal_rehabilitation_complete_yn  \\\n",
       "0                       False                                False   \n",
       "1                       False                                False   \n",
       "2                       False                                False   \n",
       "3                       False                                False   \n",
       "4                       False                                False   \n",
       "\n",
       "   existing_loan_cnt  existing_loan_amt  enter_birth  gender  k_lending_rate  \\\n",
       "0                  4           67000000           29    True            1.50   \n",
       "1                  5           78000000           49   False            1.50   \n",
       "2                  3          237000000           33   False            1.25   \n",
       "3                  1           55000000           35    True            1.50   \n",
       "4                 12          111000000           40   False            1.75   \n",
       "\n",
       "   us_lending_rate  btc_price  btc_variance  eth_price  eth_variance  \\\n",
       "0             1.00   39120756     -0.026001    2640599     -0.080688   \n",
       "1             0.50   50162048     -0.027496    3790690     -0.029800   \n",
       "2             0.25   48605860      0.003300    3289621      0.013397   \n",
       "3             1.00   37052276     -0.055389    2468705     -0.083191   \n",
       "4             1.00   39541500      0.057404    2493979      0.082214   \n",
       "\n",
       "   lunc_price  lunc_variance   Close    Open    High     Low     Volume  \\\n",
       "0    0.000322      -0.999512  2550.0  2570.0  2592.0  2546.0  900330000   \n",
       "1   81.518517      -0.071899  2716.0  2720.0  2720.0  2702.0  665350000   \n",
       "2   87.549965       0.050507  2714.0  2736.0  2736.0  2702.0  765680000   \n",
       "3    0.000061      -0.195312  2626.0  2634.0  2644.0  2618.0  870270000   \n",
       "4    0.000127       0.234253  2670.0  2662.0  2672.0  2654.0  469800000   \n",
       "\n",
       "     Change  실업자       실업률  현재가계부채CSI  가계부채전망CSI   소비심리지수  \\\n",
       "0 -0.016296   76  2.300781        102        101  94.0000   \n",
       "1  0.000100   47  1.599609        102        101  91.8125   \n",
       "2 -0.012199   65  3.099609        103        102  93.8125   \n",
       "3  0.002100   65  1.700195        103        100  92.5000   \n",
       "4  0.012001   71  2.699219        102        101  93.3125   \n",
       "\n",
       "   loanapply_insert_time_rank  loan_limit_avg_sub  loan_rate_avg_sub  \\\n",
       "0                          10       -4.161290e+06          -1.806641   \n",
       "1                           1        0.000000e+00          -0.657227   \n",
       "2                           1       -8.391304e+06           1.521484   \n",
       "3                           7        1.133333e+07          -2.546875   \n",
       "4                          14       -7.500000e+06           4.148438   \n",
       "\n",
       "   bank_diversity  product_diversity  missing_value_cnt  working_year  \\\n",
       "0              22                 31                  0             2   \n",
       "1               7                  7                  2             2   \n",
       "2              20                 23                  2             2   \n",
       "3              18                 21                  2            15   \n",
       "4              26                 36                  1             9   \n",
       "\n",
       "   working_month  age  holiday  weekday  hour  weekday_sin  weekday_cos  \\\n",
       "0              9   31    False        4    11    -0.433838    -0.900879   \n",
       "1             13   51    False        4     4    -0.433838    -0.900879   \n",
       "2             16   35     True        6    22    -0.781738     0.623535   \n",
       "3            165   50    False        3    14     0.433838    -0.900879   \n",
       "4            100   49    False        1    18     0.781738     0.623535   \n",
       "\n",
       "   hour_sin  hour_cos  foreign  event  loan_cnt  duration  visit  segment  \\\n",
       "0  0.258789 -0.965820    False     10         5       141      1        1   \n",
       "1  0.866211  0.500000     True      6         2       354      1        4   \n",
       "2 -0.500000  0.866211    False      9         3       301      1        6   \n",
       "3 -0.500000 -0.866211    False     10         5       503      1        1   \n",
       "4 -1.000000 -0.000000    False      2         0       900      1        3   \n",
       "\n",
       "   product_cnt  loan_limit_mean_x  loan_rate_mean_x  loan_limit_min_x  \\\n",
       "0            4       1.668727e+07         10.851562         1000000.0   \n",
       "1            4       1.668727e+07         10.851562         1000000.0   \n",
       "2            4       1.668727e+07         10.851562         1000000.0   \n",
       "3            7       2.464846e+07         12.179688               0.0   \n",
       "4            2       1.861316e+07         13.226562               0.0   \n",
       "\n",
       "   loan_rate_min_x  loan_limit_max_x  loan_rate_max_x  bank_label  \\\n",
       "0         5.898438       100000000.0        16.296875           2   \n",
       "1         5.898438       100000000.0        16.296875           2   \n",
       "2         5.898438       100000000.0        16.296875           2   \n",
       "3         4.800781       150000000.0        19.906250           2   \n",
       "4         7.199219        50000000.0        19.906250           2   \n",
       "\n",
       "   loan_limit_min_y  loan_rate_min_y  loan_limit_max_y  loan_rate_max_y  \\\n",
       "0         6000000.0         9.000000        20000000.0        10.000000   \n",
       "1         2000000.0         6.500000        20000000.0        10.000000   \n",
       "2         1000000.0         5.898438       100000000.0        16.296875   \n",
       "3               0.0         4.800781       150000000.0        19.906250   \n",
       "4               0.0        13.398438        50000000.0        19.906250   \n",
       "\n",
       "   loan_limit_mean_y  loan_rate_mean_y  product_label  is_applied  \n",
       "0       1.355632e+07          9.054688              1       False  \n",
       "1       1.212646e+07          6.613281              1       False  \n",
       "2       2.030726e+07         13.828125              0       False  \n",
       "3       2.418366e+07         12.664062              0       False  \n",
       "4       2.121015e+07         15.804688              0       False  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "viewinfo(train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "JEUBjYFfe5ji",
    "outputId": "da04e204-769e-48ee-8eb1-64694870f245"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(924301, 72)\n",
      "\n",
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 924301 entries, 0 to 924300\n",
      "Data columns (total 72 columns):\n",
      " #   Column                               Non-Null Count   Dtype  \n",
      "---  ------                               --------------   -----  \n",
      " 0   loan_limit                           924301 non-null  uint64 \n",
      " 1   loan_rate                            924301 non-null  float16\n",
      " 2   credit_score                         924301 non-null  uint16 \n",
      " 3   yearly_income                        924301 non-null  uint64 \n",
      " 4   income_type                          924301 non-null  uint8  \n",
      " 5   employment_type                      924301 non-null  uint8  \n",
      " 6   houseown_type                        924301 non-null  uint8  \n",
      " 7   desired_amount                       924301 non-null  uint64 \n",
      " 8   purpose                              924301 non-null  uint8  \n",
      " 9   personal_rehabilitation_yn           924301 non-null  bool   \n",
      " 10  personal_rehabilitation_complete_yn  924301 non-null  bool   \n",
      " 11  existing_loan_cnt                    924301 non-null  uint16 \n",
      " 12  existing_loan_amt                    924301 non-null  uint64 \n",
      " 13  enter_birth                          924301 non-null  uint8  \n",
      " 14  gender                               924301 non-null  bool   \n",
      " 15  k_lending_rate                       924301 non-null  float16\n",
      " 16  us_lending_rate                      924301 non-null  float16\n",
      " 17  btc_price                            924301 non-null  uint32 \n",
      " 18  btc_variance                         924301 non-null  float16\n",
      " 19  eth_price                            924301 non-null  uint32 \n",
      " 20  eth_variance                         924301 non-null  float16\n",
      " 21  lunc_price                           924301 non-null  float32\n",
      " 22  lunc_variance                        924301 non-null  float16\n",
      " 23  Close                                924301 non-null  float16\n",
      " 24  Open                                 924301 non-null  float16\n",
      " 25  High                                 924301 non-null  float16\n",
      " 26  Low                                  924301 non-null  float16\n",
      " 27  Volume                               924301 non-null  uint64 \n",
      " 28  Change                               924301 non-null  float16\n",
      " 29  실업자                                  924301 non-null  uint16 \n",
      " 30  실업률                                  924301 non-null  float16\n",
      " 31  현재가계부채CSI                            924301 non-null  uint8  \n",
      " 32  가계부채전망CSI                            924301 non-null  uint8  \n",
      " 33  소비심리지수                               924301 non-null  float16\n",
      " 34  loanapply_insert_time_rank           924301 non-null  uint8  \n",
      " 35  loan_limit_avg_sub                   924301 non-null  float64\n",
      " 36  loan_rate_avg_sub                    924301 non-null  float16\n",
      " 37  bank_diversity                       924301 non-null  uint8  \n",
      " 38  product_diversity                    924301 non-null  uint8  \n",
      " 39  missing_value_cnt                    924301 non-null  uint8  \n",
      " 40  working_year                         924301 non-null  uint8  \n",
      " 41  working_month                        924301 non-null  uint16 \n",
      " 42  age                                  924301 non-null  uint8  \n",
      " 43  holiday                              924301 non-null  bool   \n",
      " 44  weekday                              924301 non-null  uint8  \n",
      " 45  hour                                 924301 non-null  uint8  \n",
      " 46  weekday_sin                          924301 non-null  float16\n",
      " 47  weekday_cos                          924301 non-null  float16\n",
      " 48  hour_sin                             924301 non-null  float16\n",
      " 49  hour_cos                             924301 non-null  float16\n",
      " 50  foreign                              924301 non-null  bool   \n",
      " 51  event                                924301 non-null  uint16 \n",
      " 52  loan_cnt                             924301 non-null  uint16 \n",
      " 53  duration                             924301 non-null  uint32 \n",
      " 54  visit                                924301 non-null  uint16 \n",
      " 55  segment                              924301 non-null  uint8  \n",
      " 56  product_cnt                          924301 non-null  uint16 \n",
      " 57  loan_limit_mean_x                    924301 non-null  float64\n",
      " 58  loan_rate_mean_x                     924301 non-null  float16\n",
      " 59  loan_limit_min_x                     924301 non-null  float64\n",
      " 60  loan_rate_min_x                      924301 non-null  float16\n",
      " 61  loan_limit_max_x                     924301 non-null  float64\n",
      " 62  loan_rate_max_x                      924301 non-null  float16\n",
      " 63  bank_label                           924301 non-null  uint8  \n",
      " 64  loan_limit_min_y                     924301 non-null  float64\n",
      " 65  loan_rate_min_y                      924301 non-null  float16\n",
      " 66  loan_limit_max_y                     924301 non-null  float64\n",
      " 67  loan_rate_max_y                      924301 non-null  float16\n",
      " 68  loan_limit_mean_y                    924301 non-null  float64\n",
      " 69  loan_rate_mean_y                     924301 non-null  float16\n",
      " 70  product_label                        924301 non-null  uint8  \n",
      " 71  is_applied                           924301 non-null  bool   \n",
      "dtypes: bool(6), float16(24), float32(1), float64(7), uint16(8), uint32(3), uint64(5), uint8(18)\n",
      "memory usage: 176.3 MB\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "\n",
       "  <div id=\"df-f10d442e-218f-49df-ad00-6fb51439c960\">\n",
       "    <div class=\"colab-df-container\">\n",
       "      <div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>loan_limit</th>\n",
       "      <th>loan_rate</th>\n",
       "      <th>credit_score</th>\n",
       "      <th>yearly_income</th>\n",
       "      <th>income_type</th>\n",
       "      <th>employment_type</th>\n",
       "      <th>houseown_type</th>\n",
       "      <th>desired_amount</th>\n",
       "      <th>purpose</th>\n",
       "      <th>personal_rehabilitation_yn</th>\n",
       "      <th>personal_rehabilitation_complete_yn</th>\n",
       "      <th>existing_loan_cnt</th>\n",
       "      <th>existing_loan_amt</th>\n",
       "      <th>enter_birth</th>\n",
       "      <th>gender</th>\n",
       "      <th>k_lending_rate</th>\n",
       "      <th>us_lending_rate</th>\n",
       "      <th>btc_price</th>\n",
       "      <th>btc_variance</th>\n",
       "      <th>eth_price</th>\n",
       "      <th>eth_variance</th>\n",
       "      <th>lunc_price</th>\n",
       "      <th>lunc_variance</th>\n",
       "      <th>Close</th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Volume</th>\n",
       "      <th>Change</th>\n",
       "      <th>실업자</th>\n",
       "      <th>실업률</th>\n",
       "      <th>현재가계부채CSI</th>\n",
       "      <th>가계부채전망CSI</th>\n",
       "      <th>소비심리지수</th>\n",
       "      <th>loanapply_insert_time_rank</th>\n",
       "      <th>loan_limit_avg_sub</th>\n",
       "      <th>loan_rate_avg_sub</th>\n",
       "      <th>bank_diversity</th>\n",
       "      <th>product_diversity</th>\n",
       "      <th>missing_value_cnt</th>\n",
       "      <th>working_year</th>\n",
       "      <th>working_month</th>\n",
       "      <th>age</th>\n",
       "      <th>holiday</th>\n",
       "      <th>weekday</th>\n",
       "      <th>hour</th>\n",
       "      <th>weekday_sin</th>\n",
       "      <th>weekday_cos</th>\n",
       "      <th>hour_sin</th>\n",
       "      <th>hour_cos</th>\n",
       "      <th>foreign</th>\n",
       "      <th>event</th>\n",
       "      <th>loan_cnt</th>\n",
       "      <th>duration</th>\n",
       "      <th>visit</th>\n",
       "      <th>segment</th>\n",
       "      <th>product_cnt</th>\n",
       "      <th>loan_limit_mean_x</th>\n",
       "      <th>loan_rate_mean_x</th>\n",
       "      <th>loan_limit_min_x</th>\n",
       "      <th>loan_rate_min_x</th>\n",
       "      <th>loan_limit_max_x</th>\n",
       "      <th>loan_rate_max_x</th>\n",
       "      <th>bank_label</th>\n",
       "      <th>loan_limit_min_y</th>\n",
       "      <th>loan_rate_min_y</th>\n",
       "      <th>loan_limit_max_y</th>\n",
       "      <th>loan_rate_max_y</th>\n",
       "      <th>loan_limit_mean_y</th>\n",
       "      <th>loan_rate_mean_y</th>\n",
       "      <th>product_label</th>\n",
       "      <th>is_applied</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>3000000</td>\n",
       "      <td>9.898438</td>\n",
       "      <td>680</td>\n",
       "      <td>30000000</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>23000000</td>\n",
       "      <td>6</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>8</td>\n",
       "      <td>251000000</td>\n",
       "      <td>43</td>\n",
       "      <td>True</td>\n",
       "      <td>1.25</td>\n",
       "      <td>0.5</td>\n",
       "      <td>52932200</td>\n",
       "      <td>-0.015701</td>\n",
       "      <td>4000174</td>\n",
       "      <td>0.001300</td>\n",
       "      <td>94.531815</td>\n",
       "      <td>-0.087280</td>\n",
       "      <td>2700.0</td>\n",
       "      <td>2706.0</td>\n",
       "      <td>2712.0</td>\n",
       "      <td>2686.0</td>\n",
       "      <td>948360000</td>\n",
       "      <td>0.001700</td>\n",
       "      <td>62</td>\n",
       "      <td>1.599609</td>\n",
       "      <td>103</td>\n",
       "      <td>101</td>\n",
       "      <td>91.8125</td>\n",
       "      <td>5</td>\n",
       "      <td>-5.200000e+06</td>\n",
       "      <td>-4.500000</td>\n",
       "      <td>10</td>\n",
       "      <td>10</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>23</td>\n",
       "      <td>46</td>\n",
       "      <td>False</td>\n",
       "      <td>5</td>\n",
       "      <td>18</td>\n",
       "      <td>-0.975098</td>\n",
       "      <td>-0.222534</td>\n",
       "      <td>-1.000000</td>\n",
       "      <td>-0.000000</td>\n",
       "      <td>False</td>\n",
       "      <td>6</td>\n",
       "      <td>3</td>\n",
       "      <td>186</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>7</td>\n",
       "      <td>2.464846e+07</td>\n",
       "      <td>12.179688</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4.800781</td>\n",
       "      <td>150000000.0</td>\n",
       "      <td>19.90625</td>\n",
       "      <td>2</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4.800781</td>\n",
       "      <td>150000000.0</td>\n",
       "      <td>19.906250</td>\n",
       "      <td>2.418366e+07</td>\n",
       "      <td>12.664062</td>\n",
       "      <td>0</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>41000000</td>\n",
       "      <td>10.703125</td>\n",
       "      <td>777</td>\n",
       "      <td>40000000</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>140000000</td>\n",
       "      <td>4</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>2</td>\n",
       "      <td>62199999</td>\n",
       "      <td>38</td>\n",
       "      <td>False</td>\n",
       "      <td>1.25</td>\n",
       "      <td>0.5</td>\n",
       "      <td>57062564</td>\n",
       "      <td>0.008102</td>\n",
       "      <td>4311752</td>\n",
       "      <td>0.002399</td>\n",
       "      <td>116.432358</td>\n",
       "      <td>0.031204</td>\n",
       "      <td>2758.0</td>\n",
       "      <td>2730.0</td>\n",
       "      <td>2758.0</td>\n",
       "      <td>2722.0</td>\n",
       "      <td>1020000000</td>\n",
       "      <td>0.006599</td>\n",
       "      <td>47</td>\n",
       "      <td>1.599609</td>\n",
       "      <td>102</td>\n",
       "      <td>101</td>\n",
       "      <td>91.8125</td>\n",
       "      <td>16</td>\n",
       "      <td>8.256410e+06</td>\n",
       "      <td>-1.407227</td>\n",
       "      <td>29</td>\n",
       "      <td>39</td>\n",
       "      <td>5</td>\n",
       "      <td>13</td>\n",
       "      <td>142</td>\n",
       "      <td>51</td>\n",
       "      <td>False</td>\n",
       "      <td>1</td>\n",
       "      <td>22</td>\n",
       "      <td>0.781738</td>\n",
       "      <td>0.623535</td>\n",
       "      <td>-0.500000</td>\n",
       "      <td>0.866211</td>\n",
       "      <td>True</td>\n",
       "      <td>9</td>\n",
       "      <td>3</td>\n",
       "      <td>300</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>3.244987e+07</td>\n",
       "      <td>12.773438</td>\n",
       "      <td>3000000.0</td>\n",
       "      <td>7.199219</td>\n",
       "      <td>150000000.0</td>\n",
       "      <td>17.50000</td>\n",
       "      <td>2</td>\n",
       "      <td>3000000.0</td>\n",
       "      <td>7.199219</td>\n",
       "      <td>150000000.0</td>\n",
       "      <td>17.500000</td>\n",
       "      <td>3.819088e+07</td>\n",
       "      <td>11.851562</td>\n",
       "      <td>0</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2000000</td>\n",
       "      <td>19.406250</td>\n",
       "      <td>880</td>\n",
       "      <td>3000000</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>50000000</td>\n",
       "      <td>4</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>1</td>\n",
       "      <td>39400000</td>\n",
       "      <td>26</td>\n",
       "      <td>False</td>\n",
       "      <td>1.25</td>\n",
       "      <td>0.5</td>\n",
       "      <td>57062564</td>\n",
       "      <td>0.008102</td>\n",
       "      <td>4311752</td>\n",
       "      <td>0.002399</td>\n",
       "      <td>116.432358</td>\n",
       "      <td>0.031204</td>\n",
       "      <td>2758.0</td>\n",
       "      <td>2730.0</td>\n",
       "      <td>2758.0</td>\n",
       "      <td>2722.0</td>\n",
       "      <td>1020000000</td>\n",
       "      <td>0.006599</td>\n",
       "      <td>137</td>\n",
       "      <td>6.398438</td>\n",
       "      <td>105</td>\n",
       "      <td>103</td>\n",
       "      <td>91.1875</td>\n",
       "      <td>15</td>\n",
       "      <td>-1.233333e+07</td>\n",
       "      <td>6.640625</td>\n",
       "      <td>30</td>\n",
       "      <td>39</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>13</td>\n",
       "      <td>28</td>\n",
       "      <td>False</td>\n",
       "      <td>1</td>\n",
       "      <td>8</td>\n",
       "      <td>0.781738</td>\n",
       "      <td>0.623535</td>\n",
       "      <td>0.866211</td>\n",
       "      <td>-0.500000</td>\n",
       "      <td>False</td>\n",
       "      <td>8</td>\n",
       "      <td>4</td>\n",
       "      <td>284</td>\n",
       "      <td>1</td>\n",
       "      <td>6</td>\n",
       "      <td>1</td>\n",
       "      <td>4.320983e+06</td>\n",
       "      <td>16.437500</td>\n",
       "      <td>2000000.0</td>\n",
       "      <td>4.500000</td>\n",
       "      <td>15000000.0</td>\n",
       "      <td>19.90625</td>\n",
       "      <td>2</td>\n",
       "      <td>2000000.0</td>\n",
       "      <td>4.500000</td>\n",
       "      <td>15000000.0</td>\n",
       "      <td>19.906250</td>\n",
       "      <td>4.320983e+06</td>\n",
       "      <td>16.437500</td>\n",
       "      <td>0</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3000000</td>\n",
       "      <td>14.796875</td>\n",
       "      <td>850</td>\n",
       "      <td>20000000</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>30000000</td>\n",
       "      <td>7</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>1</td>\n",
       "      <td>13000000</td>\n",
       "      <td>42</td>\n",
       "      <td>True</td>\n",
       "      <td>1.50</td>\n",
       "      <td>1.0</td>\n",
       "      <td>37914376</td>\n",
       "      <td>0.002899</td>\n",
       "      <td>2492835</td>\n",
       "      <td>-0.011597</td>\n",
       "      <td>0.000069</td>\n",
       "      <td>0.095581</td>\n",
       "      <td>2618.0</td>\n",
       "      <td>2622.0</td>\n",
       "      <td>2630.0</td>\n",
       "      <td>2602.0</td>\n",
       "      <td>712610000</td>\n",
       "      <td>0.004398</td>\n",
       "      <td>65</td>\n",
       "      <td>1.700195</td>\n",
       "      <td>103</td>\n",
       "      <td>100</td>\n",
       "      <td>90.6875</td>\n",
       "      <td>1</td>\n",
       "      <td>-1.333333e+06</td>\n",
       "      <td>0.933105</td>\n",
       "      <td>9</td>\n",
       "      <td>9</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>14</td>\n",
       "      <td>44</td>\n",
       "      <td>False</td>\n",
       "      <td>3</td>\n",
       "      <td>14</td>\n",
       "      <td>0.433838</td>\n",
       "      <td>-0.900879</td>\n",
       "      <td>-0.500000</td>\n",
       "      <td>-0.866211</td>\n",
       "      <td>False</td>\n",
       "      <td>12</td>\n",
       "      <td>3</td>\n",
       "      <td>376</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "      <td>14</td>\n",
       "      <td>1.766154e+07</td>\n",
       "      <td>14.460938</td>\n",
       "      <td>1000000.0</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>200000000.0</td>\n",
       "      <td>19.90625</td>\n",
       "      <td>2</td>\n",
       "      <td>1000000.0</td>\n",
       "      <td>10.601562</td>\n",
       "      <td>65000000.0</td>\n",
       "      <td>19.906250</td>\n",
       "      <td>4.057580e+06</td>\n",
       "      <td>14.476562</td>\n",
       "      <td>0</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>11000000</td>\n",
       "      <td>14.898438</td>\n",
       "      <td>652</td>\n",
       "      <td>42000000</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>30000000</td>\n",
       "      <td>6</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>4</td>\n",
       "      <td>76200000</td>\n",
       "      <td>39</td>\n",
       "      <td>True</td>\n",
       "      <td>1.50</td>\n",
       "      <td>0.5</td>\n",
       "      <td>49672932</td>\n",
       "      <td>-0.001200</td>\n",
       "      <td>3676750</td>\n",
       "      <td>-0.006599</td>\n",
       "      <td>90.530510</td>\n",
       "      <td>0.013901</td>\n",
       "      <td>2704.0</td>\n",
       "      <td>2704.0</td>\n",
       "      <td>2708.0</td>\n",
       "      <td>2690.0</td>\n",
       "      <td>1240000000</td>\n",
       "      <td>-0.008598</td>\n",
       "      <td>62</td>\n",
       "      <td>1.599609</td>\n",
       "      <td>101</td>\n",
       "      <td>100</td>\n",
       "      <td>93.0000</td>\n",
       "      <td>6</td>\n",
       "      <td>-1.291837e+07</td>\n",
       "      <td>1.983398</td>\n",
       "      <td>34</td>\n",
       "      <td>49</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>29</td>\n",
       "      <td>42</td>\n",
       "      <td>False</td>\n",
       "      <td>0</td>\n",
       "      <td>11</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.258789</td>\n",
       "      <td>-0.965820</td>\n",
       "      <td>False</td>\n",
       "      <td>8</td>\n",
       "      <td>5</td>\n",
       "      <td>214</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>1.327673e+07</td>\n",
       "      <td>14.156250</td>\n",
       "      <td>0.0</td>\n",
       "      <td>8.000000</td>\n",
       "      <td>30000000.0</td>\n",
       "      <td>19.50000</td>\n",
       "      <td>2</td>\n",
       "      <td>0.0</td>\n",
       "      <td>14.000000</td>\n",
       "      <td>30000000.0</td>\n",
       "      <td>19.296875</td>\n",
       "      <td>1.174694e+07</td>\n",
       "      <td>16.109375</td>\n",
       "      <td>0</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>\n",
       "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-f10d442e-218f-49df-ad00-6fb51439c960')\"\n",
       "              title=\"Convert this dataframe to an interactive table.\"\n",
       "              style=\"display:none;\">\n",
       "        \n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "       width=\"24px\">\n",
       "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
       "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
       "  </svg>\n",
       "      </button>\n",
       "      \n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      flex-wrap:wrap;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "      <script>\n",
       "        const buttonEl =\n",
       "          document.querySelector('#df-f10d442e-218f-49df-ad00-6fb51439c960 button.colab-df-convert');\n",
       "        buttonEl.style.display =\n",
       "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "        async function convertToInteractive(key) {\n",
       "          const element = document.querySelector('#df-f10d442e-218f-49df-ad00-6fb51439c960');\n",
       "          const dataTable =\n",
       "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                     [key], {});\n",
       "          if (!dataTable) return;\n",
       "\n",
       "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "            + ' to learn more about interactive tables.';\n",
       "          element.innerHTML = '';\n",
       "          dataTable['output_type'] = 'display_data';\n",
       "          await google.colab.output.renderOutput(dataTable, element);\n",
       "          const docLink = document.createElement('div');\n",
       "          docLink.innerHTML = docLinkHtml;\n",
       "          element.appendChild(docLink);\n",
       "        }\n",
       "      </script>\n",
       "    </div>\n",
       "  </div>\n",
       "  "
      ],
      "text/plain": [
       "   loan_limit  loan_rate  credit_score  yearly_income  income_type  \\\n",
       "0     3000000   9.898438           680       30000000            2   \n",
       "1    41000000  10.703125           777       40000000            0   \n",
       "2     2000000  19.406250           880        3000000            0   \n",
       "3     3000000  14.796875           850       20000000            3   \n",
       "4    11000000  14.898438           652       42000000            0   \n",
       "\n",
       "   employment_type  houseown_type  desired_amount  purpose  \\\n",
       "0                1              2        23000000        6   \n",
       "1                3              2       140000000        4   \n",
       "2                3              0        50000000        4   \n",
       "3                1              3        30000000        7   \n",
       "4                3              3        30000000        6   \n",
       "\n",
       "   personal_rehabilitation_yn  personal_rehabilitation_complete_yn  \\\n",
       "0                       False                                False   \n",
       "1                       False                                False   \n",
       "2                       False                                False   \n",
       "3                       False                                False   \n",
       "4                       False                                False   \n",
       "\n",
       "   existing_loan_cnt  existing_loan_amt  enter_birth  gender  k_lending_rate  \\\n",
       "0                  8          251000000           43    True            1.25   \n",
       "1                  2           62199999           38   False            1.25   \n",
       "2                  1           39400000           26   False            1.25   \n",
       "3                  1           13000000           42    True            1.50   \n",
       "4                  4           76200000           39    True            1.50   \n",
       "\n",
       "   us_lending_rate  btc_price  btc_variance  eth_price  eth_variance  \\\n",
       "0              0.5   52932200     -0.015701    4000174      0.001300   \n",
       "1              0.5   57062564      0.008102    4311752      0.002399   \n",
       "2              0.5   57062564      0.008102    4311752      0.002399   \n",
       "3              1.0   37914376      0.002899    2492835     -0.011597   \n",
       "4              0.5   49672932     -0.001200    3676750     -0.006599   \n",
       "\n",
       "   lunc_price  lunc_variance   Close    Open    High     Low      Volume  \\\n",
       "0   94.531815      -0.087280  2700.0  2706.0  2712.0  2686.0   948360000   \n",
       "1  116.432358       0.031204  2758.0  2730.0  2758.0  2722.0  1020000000   \n",
       "2  116.432358       0.031204  2758.0  2730.0  2758.0  2722.0  1020000000   \n",
       "3    0.000069       0.095581  2618.0  2622.0  2630.0  2602.0   712610000   \n",
       "4   90.530510       0.013901  2704.0  2704.0  2708.0  2690.0  1240000000   \n",
       "\n",
       "     Change  실업자       실업률  현재가계부채CSI  가계부채전망CSI   소비심리지수  \\\n",
       "0  0.001700   62  1.599609        103        101  91.8125   \n",
       "1  0.006599   47  1.599609        102        101  91.8125   \n",
       "2  0.006599  137  6.398438        105        103  91.1875   \n",
       "3  0.004398   65  1.700195        103        100  90.6875   \n",
       "4 -0.008598   62  1.599609        101        100  93.0000   \n",
       "\n",
       "   loanapply_insert_time_rank  loan_limit_avg_sub  loan_rate_avg_sub  \\\n",
       "0                           5       -5.200000e+06          -4.500000   \n",
       "1                          16        8.256410e+06          -1.407227   \n",
       "2                          15       -1.233333e+07           6.640625   \n",
       "3                           1       -1.333333e+06           0.933105   \n",
       "4                           6       -1.291837e+07           1.983398   \n",
       "\n",
       "   bank_diversity  product_diversity  missing_value_cnt  working_year  \\\n",
       "0              10                 10                  2             3   \n",
       "1              29                 39                  5            13   \n",
       "2              30                 39                  3             2   \n",
       "3               9                  9                  3             2   \n",
       "4              34                 49                  4             3   \n",
       "\n",
       "   working_month  age  holiday  weekday  hour  weekday_sin  weekday_cos  \\\n",
       "0             23   46    False        5    18    -0.975098    -0.222534   \n",
       "1            142   51    False        1    22     0.781738     0.623535   \n",
       "2             13   28    False        1     8     0.781738     0.623535   \n",
       "3             14   44    False        3    14     0.433838    -0.900879   \n",
       "4             29   42    False        0    11     0.000000     1.000000   \n",
       "\n",
       "   hour_sin  hour_cos  foreign  event  loan_cnt  duration  visit  segment  \\\n",
       "0 -1.000000 -0.000000    False      6         3       186      1        4   \n",
       "1 -0.500000  0.866211     True      9         3       300      1        1   \n",
       "2  0.866211 -0.500000    False      8         4       284      1        6   \n",
       "3 -0.500000 -0.866211    False     12         3       376      2        4   \n",
       "4  0.258789 -0.965820    False      8         5       214      1        0   \n",
       "\n",
       "   product_cnt  loan_limit_mean_x  loan_rate_mean_x  loan_limit_min_x  \\\n",
       "0            7       2.464846e+07         12.179688               0.0   \n",
       "1            2       3.244987e+07         12.773438         3000000.0   \n",
       "2            1       4.320983e+06         16.437500         2000000.0   \n",
       "3           14       1.766154e+07         14.460938         1000000.0   \n",
       "4            4       1.327673e+07         14.156250               0.0   \n",
       "\n",
       "   loan_rate_min_x  loan_limit_max_x  loan_rate_max_x  bank_label  \\\n",
       "0         4.800781       150000000.0         19.90625           2   \n",
       "1         7.199219       150000000.0         17.50000           2   \n",
       "2         4.500000        15000000.0         19.90625           2   \n",
       "3         5.000000       200000000.0         19.90625           2   \n",
       "4         8.000000        30000000.0         19.50000           2   \n",
       "\n",
       "   loan_limit_min_y  loan_rate_min_y  loan_limit_max_y  loan_rate_max_y  \\\n",
       "0               0.0         4.800781       150000000.0        19.906250   \n",
       "1         3000000.0         7.199219       150000000.0        17.500000   \n",
       "2         2000000.0         4.500000        15000000.0        19.906250   \n",
       "3         1000000.0        10.601562        65000000.0        19.906250   \n",
       "4               0.0        14.000000        30000000.0        19.296875   \n",
       "\n",
       "   loan_limit_mean_y  loan_rate_mean_y  product_label  is_applied  \n",
       "0       2.418366e+07         12.664062              0       False  \n",
       "1       3.819088e+07         11.851562              0       False  \n",
       "2       4.320983e+06         16.437500              0       False  \n",
       "3       4.057580e+06         14.476562              0       False  \n",
       "4       1.174694e+07         16.109375              0       False  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "viewinfo(valid)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "kckBQt3NrNEo"
   },
   "source": [
    "# Preprocessing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "R6VPH5ysLlUT"
   },
   "outputs": [],
   "source": [
    "# train.drop('holiday', axis=1, inplace=True)\n",
    "# valid.drop('holiday', axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "km_R0G1dRxRF"
   },
   "outputs": [],
   "source": [
    "cat_feats = ['income_type', 'employment_type', 'houseown_type', 'purpose', 'personal_rehabilitation_yn', 'personal_rehabilitation_complete_yn', \n",
    "             'gender', 'bank_diversity', 'product_diversity', 'holiday', 'weekday', 'foreign', 'segment', 'bank_label', 'product_label']\n",
    "cat_idxs = [i for i, col in enumerate(train.columns.values) if col in cat_feats]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "6Gq1qTakVZbM"
   },
   "outputs": [],
   "source": [
    "for col in train.columns.values:\n",
    "    if train[col].dtype == 'bool':\n",
    "        train[col] = train[col].astype('uint8')\n",
    "    if valid[col].dtype == 'bool':\n",
    "        valid[col] = valid[col].astype('uint8')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "UhEVMfzUYtsy"
   },
   "outputs": [],
   "source": [
    "for col in cat_feats:\n",
    "    if train[col].min() != 0:\n",
    "        train[col] -= train[col].min()\n",
    "    if valid[col].min() != 0:\n",
    "        valid[col] -= valid[col].min()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "nYt2kRNjQ6Zy"
   },
   "outputs": [],
   "source": [
    "cat_dims = [train[col].nunique() for col in cat_feats]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "ayyoqz_zTCqT",
    "outputId": "48734729-0320-4bf3-dbd2-9743f789279c"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[4, 5, 6, 8, 9, 10, 14, 37, 38, 43, 44, 50, 55, 63, 70]\n",
      "[6, 4, 4, 8, 2, 2, 2, 56, 79, 2, 7, 2, 7, 3, 2]\n"
     ]
    }
   ],
   "source": [
    "print(cat_idxs)\n",
    "print(cat_dims)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Wy6d2Oz1ZXV1"
   },
   "source": [
    "# Evaluation Metric"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "_wfCc03dZXV2"
   },
   "outputs": [],
   "source": [
    "class F1Score(Metric):\n",
    "    def __init__(self):\n",
    "        self._name = \"f1\"\n",
    "        self._maximize = True\n",
    "    \n",
    "    def __call__(self, y_true, y_score):\n",
    "        y_score = (y_score[:, 1] >= 0.5).astype(np.uint8)\n",
    "        return f1_score(y_true, y_score)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "B6-JsG1EZXV2"
   },
   "source": [
    "# Pretraining"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Q-nCsQk18-x3"
   },
   "source": [
    "## Split X, y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "lqjqzXbGrYku"
   },
   "outputs": [],
   "source": [
    "X_train = train.drop('is_applied', axis=1)\n",
    "y_train = train['is_applied']\n",
    "\n",
    "X_valid = valid.drop('is_applied', axis=1)\n",
    "y_valid = valid['is_applied']\n",
    "\n",
    "X_tr, X_val, y_tr, y_val = train_test_split(X_train, y_train, test_size=0.3, shuffle=True, random_state=RANDOM_STATE, stratify=y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "n-NlGbOcPACy",
    "outputId": "40dde9bf-56f8-4507-f85c-f25efacd07e4"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Unique categories\n",
      "\ttrain income_type -> [0 1 2 3 4 5]\tvalid income_type -> [0 1 2 3 4 5]\n",
      "\ttrain employment_type -> [0 1 2 3]\tvalid employment_type -> [0 1 2 3]\n",
      "\ttrain houseown_type -> [0 1 2 3]\tvalid houseown_type -> [0 1 2 3]\n",
      "\ttrain purpose -> [0 1 2 3 4 5 6 7]\tvalid purpose -> [0 1 2 3 4 5 6 7]\n",
      "\ttrain personal_rehabilitation_yn -> [0 1]\tvalid personal_rehabilitation_yn -> [0 1]\n",
      "\ttrain personal_rehabilitation_complete_yn -> [0 1]\tvalid personal_rehabilitation_complete_yn -> [0 1]\n",
      "\ttrain gender -> [0 1]\tvalid gender -> [0 1]\n",
      "\ttrain bank_diversity -> [ 0  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20 21 22 23\n",
      " 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47\n",
      " 48 49 50 51 52 53 54 55]\tvalid bank_diversity -> [ 0  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20 21 22 23\n",
      " 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47\n",
      " 48 49 50 51 52 53 54 55]\n",
      "\ttrain product_diversity -> [ 0  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20 21 22 23\n",
      " 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47\n",
      " 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71\n",
      " 72 73 74 75 76 77 78]\tvalid product_diversity -> [ 0  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20 21 22 23\n",
      " 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47\n",
      " 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71\n",
      " 72 73 74 75 76 77 78]\n",
      "\ttrain holiday -> [0 1]\tvalid holiday -> [0 1]\n",
      "\ttrain weekday -> [0 1 2 3 4 5 6]\tvalid weekday -> [0 1 2 3 4 5 6]\n",
      "\ttrain foreign -> [0 1]\tvalid foreign -> [0 1]\n",
      "\ttrain segment -> [0 1 2 3 4 5 6]\tvalid segment -> [0 1 2 3 4 5 6]\n",
      "\ttrain bank_label -> [0 1 2]\tvalid bank_label -> [0 1 2]\n",
      "\ttrain product_label -> [0 1]\tvalid product_label -> [0 1]\n"
     ]
    }
   ],
   "source": [
    "print(f\"Unique categories\")\n",
    "for col in cat_feats:\n",
    "    print(f\"\\ttrain {col} -> {np.sort(X_tr[col].unique())}\", end='')\n",
    "    print(f\"\\tvalid {col} -> {np.sort(X_val[col].unique())}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "3wEl3jP-G0jw"
   },
   "source": [
    "## Hyper-parameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "uhmNxF5lZXV3"
   },
   "outputs": [],
   "source": [
    "## constant values for hyper-parameter\n",
    "BATCH_SIZE = 2**11\n",
    "GHOST_BATCH = 2**7\n",
    "# N_D = N_A = 12\n",
    "MAX_EPOCH = 100\n",
    "# MASK_TYPE = \"entmax\"\n",
    "N_STEPS = 5\n",
    "LR = 0.01\n",
    "PATIENCE = 8\n",
    "# PRETRAINING_RATIO = 0.7\n",
    "NROWS = 1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "dAW9T-oK9B0L"
   },
   "source": [
    "## Pretrain model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "vcdDfZ0-ZXV3",
    "outputId": "219f7af5-6c79-4114-f61a-979087d6e4cd"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch 0  | loss: 109.95198| val_0_unsup_loss_numpy: 75.85262298583984|  0:03:53s\n",
      "epoch 1  | loss: 7.72789 | val_0_unsup_loss_numpy: 7.831409931182861|  0:07:44s\n",
      "epoch 2  | loss: 5.05535 | val_0_unsup_loss_numpy: 6.36890983581543|  0:11:34s\n",
      "epoch 3  | loss: 4.18821 | val_0_unsup_loss_numpy: 4.250140190124512|  0:15:25s\n",
      "epoch 4  | loss: 3.65581 | val_0_unsup_loss_numpy: 3.586590051651001|  0:19:20s\n",
      "epoch 5  | loss: 3.67806 | val_0_unsup_loss_numpy: 3.0711100101470947|  0:23:13s\n",
      "epoch 6  | loss: 3.93225 | val_0_unsup_loss_numpy: 3.0656700134277344|  0:27:05s\n",
      "epoch 7  | loss: 3.58761 | val_0_unsup_loss_numpy: 9.622770309448242|  0:30:56s\n",
      "epoch 8  | loss: 3.76628 | val_0_unsup_loss_numpy: 3.355340003967285|  0:34:47s\n",
      "epoch 9  | loss: 3.29962 | val_0_unsup_loss_numpy: 3.016160011291504|  0:38:33s\n",
      "epoch 10 | loss: 3.32598 | val_0_unsup_loss_numpy: 3.1162500381469727|  0:42:20s\n",
      "epoch 11 | loss: 3.19814 | val_0_unsup_loss_numpy: 3.0869998931884766|  0:46:12s\n",
      "epoch 12 | loss: 3.25716 | val_0_unsup_loss_numpy: 2.7756900787353516|  0:50:02s\n",
      "epoch 13 | loss: 3.10488 | val_0_unsup_loss_numpy: 2.9546899795532227|  0:53:52s\n",
      "epoch 14 | loss: 3.13644 | val_0_unsup_loss_numpy: 2.9227800369262695|  0:57:43s\n",
      "epoch 15 | loss: 3.00394 | val_0_unsup_loss_numpy: 2.9556500911712646|  1:01:33s\n",
      "epoch 16 | loss: 2.9682  | val_0_unsup_loss_numpy: 2.852560043334961|  1:05:22s\n",
      "epoch 17 | loss: 2.93932 | val_0_unsup_loss_numpy: 2.8449299335479736|  1:09:12s\n",
      "\n",
      "Early stopping occurred at epoch 17 with best_epoch = 12 and best_val_0_unsup_loss_numpy = 2.7756900787353516\n"
     ]
    }
   ],
   "source": [
    "tabnet_params = dict(\n",
    "    cat_idxs=cat_idxs,\n",
    "    cat_dims=cat_dims,\n",
    "    seed=RANDOM_STATE,\n",
    "    optimizer_params=dict(lr=LR), \n",
    "    # scheduler_params=dict(mode=\"min\", patience=5, min_lr=1e-5, factor=0.9),\n",
    "    # scheduler_fn=torch.optim.lr_scheduler.ReduceLROnPlateau,\n",
    "    device_name=str(DEVICE)\n",
    ")\n",
    "\n",
    "pretrainer = TabNetPretrainer(**tabnet_params)\n",
    "\n",
    "pretrainer.fit(\n",
    "    X_train=X_train.values, eval_set=[X_valid.values], \n",
    "    max_epochs=MAX_EPOCH, \n",
    "    patience=PATIENCE, \n",
    "    batch_size=BATCH_SIZE, \n",
    "    virtual_batch_size=GHOST_BATCH, \n",
    "    num_workers=0, \n",
    "    drop_last=False, \n",
    "    # pretraining_ratio=PRETRAINING_RATIO\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "cmEweYbt9FA-"
   },
   "source": [
    "## Plot results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "gtAwFnZvtQVw"
   },
   "outputs": [],
   "source": [
    "pretrain_explain_matrix, pretrain_masks = pretrainer.explain(X_valid.values)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 399
    },
    "id": "qLeF6kOcISGm",
    "outputId": "0b72259f-7feb-4137-e7c3-b0ae4b131f90"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 2160x720 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(30, 10))\n",
    "for key in pretrain_masks:\n",
    "    plt.subplot(int(f\"{NROWS}{len(pretrain_masks)//NROWS}{key+1}\"))\n",
    "    plt.imshow(pretrain_masks[key][:50], cmap='gray')\n",
    "    # plt.axis('off')\n",
    "    plt.title(f\"Pretrained mask {key}\")\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 438
    },
    "id": "qH8UwT0atlmK",
    "outputId": "9004d6e0-4248-4966-add3-ed64fecb8e1a"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x864 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "mask_agg = np.zeros(shape=pretrain_masks[0].shape)\n",
    "for key in pretrain_masks:\n",
    "    mask_agg += pretrain_masks[key]\n",
    "mask_agg /= len(pretrain_masks)\n",
    "\n",
    "plt.figure(figsize=(8, 12))\n",
    "plt.imshow(mask_agg[:50], cmap='gray')\n",
    "plt.title(f\"Aggregated Mask\")\n",
    "# plt.axis('off')\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "oz_NIwHJvvSf",
    "outputId": "9814c33c-8242-4bb7-86dd-0f1252c3057f"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "9313"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "del pretrain_explain_matrix, pretrain_masks, mask_agg\n",
    "gc.collect()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "wgcIIE3qrP3O"
   },
   "source": [
    "## Save Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 54
    },
    "id": "_Yp5EN-RzVyY",
    "outputId": "b5f0cf1a-dc27-4344-d511-9da663fcf848"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Successfully saved model at /content/drive/MyDrive/projects/빅콘테스트/models/pretrainer_smote_default.zip\n"
     ]
    },
    {
     "data": {
      "application/vnd.google.colaboratory.intrinsic+json": {
       "type": "string"
      },
      "text/plain": [
       "'/content/drive/MyDrive/projects/빅콘테스트/models/pretrainer_smote_default.zip'"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "with open(\"../Model/pretrainer_smote_default.pkl\", 'wb') as f:\n",
    "    pickle.dump(pretrainer, f, protocol=pickle.HIGHEST_PROTOCOL)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "9aLt0EUDrWpm"
   },
   "source": [
    "# Supervised Training"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "swWNW_XEMW98"
   },
   "source": [
    "## With Pretraining"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "pA-IqPYhrU8P"
   },
   "source": [
    "### Load Pretrained Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "uR02JD2h0qP2"
   },
   "outputs": [],
   "source": [
    "with open(\"../Model/pretrainer_smote_default.pkl\", 'rb') as f:\n",
    "    pretrainer = pickle.load(f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "zEFYTkRh9JGd",
    "outputId": "418340b8-4344-4303-b5eb-33bd1a278e81"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "48"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gc.collect()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Ou6tIiHrsu9G"
   },
   "source": [
    "### Model Implementation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "QW2peg0zs553"
   },
   "outputs": [],
   "source": [
    "clf = TabNetClassifier(\n",
    "    cat_idxs=cat_idxs,\n",
    "    cat_dims=cat_dims,\n",
    "    seed=RANDOM_STATE,\n",
    "    optimizer_params=dict(lr=LR), \n",
    "    # scheduler_params=dict(mode=\"min\", patience=5, min_lr=1e-5, factor=0.9),\n",
    "    # scheduler_fn=torch.optim.lr_scheduler.ReduceLROnPlateau,\n",
    "    device_name=str(DEVICE)\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "wK8MktGw8BLd",
    "outputId": "1aa456ce-1068-4256-9987-afa02dd80fa3"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch 0  | loss: 0.2201  | train_logloss: 0.12724 | train_f1: 0.94852 | valid_logloss: 0.17169 | valid_f1: 0.26205 |  0:04:36s\n",
      "epoch 1  | loss: 0.12664 | train_logloss: 0.11042 | train_f1: 0.95436 | valid_logloss: 0.16028 | valid_f1: 0.27237 |  0:09:09s\n",
      "epoch 2  | loss: 0.11757 | train_logloss: 0.10969 | train_f1: 0.95514 | valid_logloss: 0.16872 | valid_f1: 0.26783 |  0:13:38s\n",
      "epoch 3  | loss: 0.1138  | train_logloss: 0.11996 | train_f1: 0.94898 | valid_logloss: 0.15345 | valid_f1: 0.18282 |  0:18:16s\n",
      "epoch 4  | loss: 0.1108  | train_logloss: 0.12375 | train_f1: 0.94948 | valid_logloss: 0.20403 | valid_f1: 0.30996 |  0:22:52s\n",
      "epoch 5  | loss: 0.10805 | train_logloss: 0.10003 | train_f1: 0.95916 | valid_logloss: 0.1509  | valid_f1: 0.26406 |  0:27:24s\n",
      "epoch 6  | loss: 0.10528 | train_logloss: 0.10049 | train_f1: 0.95935 | valid_logloss: 0.15557 | valid_f1: 0.29124 |  0:31:54s\n",
      "epoch 7  | loss: 0.10382 | train_logloss: 0.10331 | train_f1: 0.95722 | valid_logloss: 0.15162 | valid_f1: 0.28732 |  0:36:21s\n",
      "epoch 8  | loss: 0.10225 | train_logloss: 0.09859 | train_f1: 0.95935 | valid_logloss: 0.14903 | valid_f1: 0.16799 |  0:40:44s\n",
      "epoch 9  | loss: 0.10124 | train_logloss: 0.09735 | train_f1: 0.96037 | valid_logloss: 0.15279 | valid_f1: 0.29081 |  0:45:06s\n",
      "epoch 10 | loss: 0.10089 | train_logloss: 0.11297 | train_f1: 0.95421 | valid_logloss: 0.19483 | valid_f1: 0.33266 |  0:49:27s\n",
      "epoch 11 | loss: 0.10003 | train_logloss: 0.10039 | train_f1: 0.95942 | valid_logloss: 0.16028 | valid_f1: 0.34388 |  0:53:53s\n",
      "epoch 12 | loss: 0.09985 | train_logloss: 0.09939 | train_f1: 0.95892 | valid_logloss: 0.15359 | valid_f1: 0.33117 |  0:58:21s\n",
      "epoch 13 | loss: 0.09887 | train_logloss: 0.09899 | train_f1: 0.95987 | valid_logloss: 0.16323 | valid_f1: 0.30857 |  1:02:49s\n",
      "epoch 14 | loss: 0.09944 | train_logloss: 0.09586 | train_f1: 0.96072 | valid_logloss: 0.15018 | valid_f1: 0.27194 |  1:07:21s\n",
      "epoch 15 | loss: 0.09818 | train_logloss: 0.11083 | train_f1: 0.95451 | valid_logloss: 0.1818  | valid_f1: 0.32817 |  1:11:55s\n",
      "epoch 16 | loss: 0.09827 | train_logloss: 0.11367 | train_f1: 0.95212 | valid_logloss: 0.18469 | valid_f1: 0.33472 |  1:16:28s\n",
      "epoch 17 | loss: 0.09724 | train_logloss: 0.09824 | train_f1: 0.95959 | valid_logloss: 0.15753 | valid_f1: 0.26868 |  1:20:55s\n",
      "epoch 18 | loss: 0.09728 | train_logloss: 0.1068  | train_f1: 0.95531 | valid_logloss: 0.17379 | valid_f1: 0.30834 |  1:25:22s\n",
      "epoch 19 | loss: 0.09812 | train_logloss: 0.11968 | train_f1: 0.95416 | valid_logloss: 0.17811 | valid_f1: 0.31235 |  1:29:49s\n",
      "\n",
      "Early stopping occurred at epoch 19 with best_epoch = 11 and best_valid_f1 = 0.34388\n"
     ]
    }
   ],
   "source": [
    "%time\n",
    "clf.fit(\n",
    "    X_train=X_train.values, \n",
    "    y_train=y_train.values,\n",
    "    eval_set=[(X_train.values, y_train.values), (X_valid.values, y_valid.values)],\n",
    "    eval_name=['train', 'valid'],\n",
    "    eval_metric=['logloss', 'f1'],\n",
    "    max_epochs=MAX_EPOCH, \n",
    "    batch_size=BATCH_SIZE, \n",
    "    virtual_batch_size=GHOST_BATCH,\n",
    "    patience=8,\n",
    "    num_workers=0, \n",
    "    drop_last=False,\n",
    "    from_unsupervised=pretrainer\n",
    ") "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "M-LEg7mhCmL5",
    "outputId": "f4a26049-43db-48af-f8b5-e86b3877cdc1"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "22"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gc.collect()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "lIOr9gbPtCOf"
   },
   "source": [
    "### Plot Evaluation Metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 441
    },
    "id": "pzE9FcPItEdS",
    "outputId": "1d0f9843-a57c-47e5-d20b-6d9756e32593"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(18, 6))\n",
    "\n",
    "plt.subplot(121)\n",
    "plt.plot(clf.history['train_logloss'], label=\"Train\")\n",
    "plt.plot(clf.history['valid_logloss'], label=\"Valid\")\n",
    "plt.legend()\n",
    "plt.grid(True)\n",
    "\n",
    "plt.subplot(122)\n",
    "plt.plot(clf.history['train_f1'], label=\"Train\")\n",
    "plt.plot(clf.history['valid_f1'], label=\"Valid\")\n",
    "plt.legend()\n",
    "plt.grid(True)\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "GjGGOhEgtFNN"
   },
   "source": [
    "### Plot Feature Importances"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 1000
    },
    "id": "cNJkVaz_tI-l",
    "outputId": "e80c8a1e-ba62-4c96-f753-42ebb4742730"
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAvoAAAV+CAYAAAAHmZqVAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOzdebxf073/8dc7CYIQQ45cQ0kbxEyJuTjI7Wwqmqs6REsa3HJ7De3vVokWVdqa2wptQpGGKEJrKr4aiYhEZiLtlbiKRmJIHSXN8Pn9sdepna/vmZLjfM93n/fz8TiPs/faa6/12XudxyOfvb5rf6OIwMzMzMzMiqVbtQMwMzMzM7P250TfzMzMzKyAnOibmZmZmRWQE30zMzMzswJyom9mZmZmVkBO9M3MzMzMCsiJvpmZdWqS/kfSjdWOoxYpM1LSm5ImVzue1pBUknRyteMwKwIn+mZmBSZpgaR3JTXkfrZohzYHtVeMLYmISyKiUyR+koZLuqXacbTBJ4B/B7aKiH06smNJ9+f+5pZJ+mdu/5er2WY/SSHpD2Xlt0ga3so2OvTv16yaelQ7ADMz+9AdERF/rHYQjST1iIjl1Y6jrSTV4r+Z2wALIuKdSgc/zLGIiM/k+hkF/DUizmun5veVdEBETGyn9swKyTP6ZmZdkKTekn4l6VVJL0u6SFL3dKy/pEclvS5psaRbJW2Ujv0G2Bq4N83MniupXtJfy9r/16xpmgUfm2Zd/w4Maa7/CrH+axY9N6N7kqSX0pKUYZL2ljRT0luSrs2dO0TSBEnXSloiaa6kw3PHt5A0TtIbkv4i6ZSyfvNxDwP+Bxicrn1GqneSpOckvS3pBUnfzLVRL+mvks6S9Fq63pNyx9eV9FNJL6b4npC0bjq2n6SJ6ZpmSKovu64XUp/zJZ1Y4b59A7gR2D/Fe2Eunu9I+hswUtI6kq6U9Er6uVLSOmXxn5uL/2hJn5U0L923/2nub61CXBtLuk/SojR+90naqqxaf0mTJf1d0j2SNik7fhlwcTN9fF7S9HTvJkraLZV/4O+3LbGb1Ron+mZmXdMoYDmwLfBx4JNA4/IYAT8CtgB2BD4CDAeIiK8A/0f2KUGviLislf0dBYwFNgJubaH/1tgX2A4YDFwJfA8YBOwMfFHSIWV1/xfoA1wA/C6XOP4W+Gu61uOASyQd1kTcvwIuAcaka9891XkN+DywIXAScIWkPXNt/BvQG9gS+AZwnaSN07GfAHsBBwCbAOcCKyVtCfweuCiVnw3cKalO0vrA1cBnImKDdO708hsUEb8iezh5MsV7QS6eTchm+4eme7cfsAewO7APkJ95/zegZ4r/fOAG4Msp7oOA70v6aHn/zegGjEz9bw28C1xbVuerwNeBzcn+Tq4uO/5zYHtVWIIj6ePAr4FvApsC1wPjJK2zBn+/ZjXJib6ZWfHdnWY235J0t6S+wGeB/4qIdyLiNeAK4D8AIuIvEfFwRCyNiEXAz4BDmm6+VZ6MiLsjYiVZQtxk/630w4h4LyIeAt4BRkfEaxHxMjCe7OGh0WvAlRGxLCLGAM8Dn5P0EeBA4DuprelkM+BfrRR3RLxbKZCI+H1E/G9kHgceIkuAGy0DfpD6/wPQAAyQ1I0smT0zIl6OiBURMTEilpIl0n+IiD+kvh8GpqT7BrAS2EXSuhHxakTMacO9WwlckMb3XeDEFN9rabwvBL5SFv/FEbGM7MGoD3BVRLyd+n2W7AGhVSLi9Yi4MyL+ERFvk83Ml/99/SYiZqclR98ne3jLf+LzbjrvogpdDAWuj4in0j29CVhK9jBj1qU40TczK76jI2Kj9HM02UzqWsCrjQ8AZLOemwFI6ivpt2lJzd+BW8iSuzXxUm672f5baWFu+90K+71y+y9HROT2XySbwd8CeCMlm/ljWzYRd0WSPiNpUlrG8hZZMp6/X6+XrYP/R4qvD9lM+f9WaHYb4PjcA9pbZC/Wbp6S38Fks/WvSvq9pB1aijNnUUS8l9vfguy6GzXen3z8K9J248NOc/e7WZLWk3R9Wq70d+BPwEZliXz+vr9I9vdS/jd4I9BX0hFl5dsAZ5Xdu4+UXZNZl+BE38ys63mJbIazT+4BYMOI2DkdvwQIYNeI2JBsdlm582PV5ngHWK9xJyVsdWV18ue01H9721JSPv6tgVfSzyaSNig79nITcX9gP61lv5NsCU7fiNgI+AOr3q+mLAbeA/pXOPYS2az2Rrmf9SPiUoCIeDAi/p1sactcsuU0rVV+Ta+QJceNGu/Ph+UsYACwb/r7OjiV5+/ZR8riWUZ2v/4lIv5J9unDD8vOfYnsE4j8vVsvIkY3ntp+l2LWuTnRNzPrYiLiVbLlJT+VtKGkbspewG1cPrEB2fKSJWmt+DllTSwEPpbbnwf0lPQ5SWuRre9eZw36b2+bAWdIWkvS8WTvHfwhIl4CJgI/ktQzvbD5DbJPMJqyEOiXlt0ArE12rYuA5ZI+Q/a+QYvSMqZfAz9T9lJwd0n7p4eHW4AjJH0qlfdML8ZulT5xOSqt1V9KNlYr23hP8kYD56X1/33I1uF/mF8hugHZpwBvpXclLqhQ58uSdpK0HvADYGzuU4W835B9KvLpXNkNwDBJ+yqzfvrbbHygK//7NSssJ/pmZl3TV8mS1GeBN8leON08HbsQ2BNYQvZC6O/Kzv0RWWL4lqSzI2IJcBrZUoqXyWb4/0rzmuu/vT1F9uLuYrJ13cdFxOvp2AlAP7IZ7LvI1q4391Wkd6Tfr0t6Ji37OQO4new6vgSMa0NsZwOzgKeBN4AfA93SQ8hRZN/ys4hslvocsn+3uwH/nWJ+g2x9+6lt6LPcRWTr/2emWJ6h8tr39nIlsC7ZeEwCHqhQ5zdkL2z/jSyRP6NSQyn5P5/s5eLGsinAKWQv+L4J/AUYkjttlb/fNbsUs85Nqy5bNDMzKw5JQ4CTI+IT1Y7FzKyjeUbfzMzMzKyAnOibmZmZmRWQl+6YmZmZmRWQZ/TNzMzMzArIib6ZmZmZWQH1qHYAZp3NRhttFNtuu221w7B28s4777D++utXOwxrRx7TYvF4FovHs+NNnTp1cUSU/yeFgBN9sw/o27cvU6ZMqXYY1k5KpRL19fXVDsPakce0WDyexeLx7HiSXmzymF/GNVvV9v23jQlnD692GNZOZvRZj90X/6PaYVg78pgWi8ezWDye76s79csd0o+kqRExsNIxr9E3MzMzMysgJ/pmZmZmZgXkRN/MzMzMrICc6FvNkrSFpLEt1JmYfveT9KWOiczMzMys+pzoW82KiFci4rgW6hyQNvsBTvTNzMysy3CibzVB0qWSTs/tD5d0tqTZaX9nSZMlTZc0U9J2qbwhnXIpcFA6/u2OvwIzMzOzjuXv0bdaMQa4Ergu7X8R+CYwJO0PA66KiFslrQ10Lzv/u8DZEfH5So1LGgoMBairq2NGn/XaN3qrmnd7dPN4FozHtFg8nsXi8Xxfj1Kp2iE40bfaEBHTJG0maQugDngTeClX5Unge5K2An4XEX9uY/sjgBGQfY++vwO4OPydzsXjMS0Wj2exeDzfV3f8F6odgpfuWE25AzgOGEw2w/8vEXEbcCTwLvAHSYd1fHhmZmZmnYdn9K2WjAFuAPoAhwDrNB6Q9DHghYi4WtLWwG7Ao7lz3wY26MBYzczMzKrKM/pWMyJiDlmy/nJEvFp2+IvAbEnTgV2Am8uOzwRWSJrhl3HNzMysK/CMvtWUiNg1t72ALKknIi4l+2ad8vq90u9lgJfzmJmZWZfhGX0zMzMzswJyom9mZmZmVkBeumNWRj26U3fql6sdhrWTHqVSp/iKM2s/HtNi8XgWi8ezc/GMvpmZmZlZATnRNzMzMzMrICf6ZmZmZmYF5DX6ZmVi+XIW/fJX1Q7D2snyTXt7PAvGY1ostTyedcO+Ue0QzJrlGX0zMzMzswJyom9mZmZmVkBO9M3MzMzMCsiJvpmZmZlZATnR70QkNVQ7hjxJ9ZIOqHYca0rScElnVzsOMzMzs47kRL+Lk9TcNy/VAzWf6JuZmZl1RU70OyFlLpc0W9IsSYNTeS9Jj0h6JpUflcr7SXpO0g2S5kh6SNK6zbRfknSlpCnAmZKOkPSUpGmS/iipr6R+wDDg25KmSzpIUp2kOyU9nX4ObKaPfSQ9mdqcKGlAKp8kaeeyWAamth9O8d8o6UVJfZpoe31Jv5c0I92jxvuzoPGc1GYpd9ruKZ4/SzqlVQNhZmZmVsP8Pfqd0xeAPYDdgT7A05L+BCwCjomIv6eEdpKkcemc7YATIuIUSbcDxwK3NNPH2hExEEDSxsB+ERGSTgbOjYizJP0SaIiIn6R6twFXRMQTkrYGHgR2bKL9ucBBEbFc0iDgkhTTGOCLwAWSNgc2j4gpkq4FHo2IH0n6NNDclxN/GnglIj6X4urdTN1GuwH7AesD0yT9PiJeaTwoaSgwFKCuro4Zm7amSasF7/bo7vEsGI9psdTyePYolaodQqfT0NBAyfel03Ci3zl9AhgdESuAhZIeB/YG7gcukXQwsBLYEuibzpkfEdPT9lSgXwt9jMltbwWMSYn32sD8Js4ZBOwkqXF/Q0m9IqLSuwW9gZskbQcEsFYqvx14CLiALOEfm7vmYwAi4gFJbzYT+yzgp5J+DNwXEeObqdvonoh4F3hX0mPAPsDdjQcjYgQwAmD7/v1j99eXtKJJqwUzNu2Nx7NYPKbFUsvjWXf8cdUOodMplUrU19dXOwxLvHSntpwI1AF7RcQewEKgZzq2NFdvBS0/xL2T274GuDYidgW+mWuzXDeymf890s+WTST5AD8EHouIXYAjGtuMiJeB1yXtBgxm1QeOVomIecCeZAn/RZLOT4eW8/7fdPk1RAv7ZmZmZoXiRL9zGg8MltRdUh1wMDCZbJb8tYhYJulQYJt26q838HLa/lqu/G1gg9z+Q8C3Gnck7dHKNoeUHRsDnAv0joiZqWwC2Qw/kj4JbNxUw5K2AP4REbcAl5Ml/QALgL3S9rFlpx0lqaekTcleMn66mdjNzMzMap4T/c7pLmAmMAN4lGzN/N+AW4GBkmYBXyVbB98ehgN3SJoKLM6V3wsc0/gyLnBG6n+mpGfJXtZtymXAjyRN44OfLowF/oNsGU+jC4FPSpoNHA/8jexBo5JdgcmSppMtAboo18ZV6SXjFWXnzAQeAyYBP8yvzzczMzMrIq/R70Qiolf6HcA56Sd/fDGwfxOn75Kr95MW+qkv278HuKdCvXlkL7HmDW6u7dy5TwLb54rOyx1byAf/9pYAn0ov7+4P7B0RS6kgIh4kexG4vHx8WZ+N5cNbE7OZmZlZkTjRt85ia+B2Sd2AfwL+CkwzMzOzNeBEv8AkXQeUf9f9VRExsh37OAk4s6x4QkSc3pZ2IuLPwMfL2t4UeKRC9cMj4vU2BdoG6tGDumHNfbun1ZIepZK/GaNgPKbF4vE0+/A40S+wtibbq9nHSKDdHhzK2n6d7P8TMDMzM7M28su4ZmZmZmYF5ETfzMzMzKyAvHTHrEwsX86iX15X7TCsnSzfdLNCjGfdsA99JZ6ZmRWMZ/TNzMzMzArIib6ZmZmZWQE50TczMzMzKyAn+mZmZmZmBeRE31abpCGSrl2Dc7dooU5J0sAW6iyQ1KeN/a5WzGZmZma1xIm+fYCk7h3QzRCg2UTfzMzMzFafE/0uRlI/SXMl3SrpOUljJa2XZsZ/LOkZ4HhJJ0iaJWm2pB/nzj9J0jxJk4EDc+WjJB2X22/IbX8ntTVD0qWp3kDgVknTJa3birh/IWmKpDmSLiw7fG5qf7KkbVP9Okl3Sno6/RxYoVkzMzOzwvL36HdNA4BvRMQESb8GTkvlr0fEnmlJzSRgL+BN4CFJRwNPARem8iXAY8C05jqS9BngKGDfiPiHpE0i4g1J/wmcHRFTWhnz99J53YFHJO0WETPTsSURsaukrwJXAp8HrgKuiIgnJG0NPAjs2EycQ4GhAHV1dczYdLNWhmWd3bs9ehRiPHuUStUOodNoaGig5PtRGB7PYvF4di5O9LumlyJiQtq+BTgjbY9Jv/cGShGxCEDSrcDB6Vi+fAywfQt9DQJGRsQ/ACLijdWM+YspGe8BbA7sBDQm+qNzv6/I9buTpMbzN5TUq6nGI2IEMAJg+/79Y/fXX1vNMK2zmbHpZhRhPOuOP77aIXQapVKJ+vr6aodh7cTjWSwez87FiX7XFE3sv7MGbS4nLQWT1A1Yew3aWoWkjwJnA3tHxJuSRgE9c1WiwnY3YL+IeK+srfYKy8zMzKxT8xr9rmlrSfun7S8BT5QdnwwcIqlPWipzAvA42dKdQyRtKmktID/FuIBsSQ/AkcBaafth4CRJ6wFI2iSVvw1s0Mp4NyR7CFkiqS/wmbLjg3O/n0zbDwHfaqwgaY9W9mVmZmZWCE70u6bngdMlPQdsDPwifzAiXgW+S7YGfwYwNSLuSeXDyZLpCcBzudNuIHsImAHsT/p0ICIeAMYBUyRNJ5uZBxgF/LI1L+NGxAyydwHmArelvvM2ljQTOBP4dio7AxgoaaakZ4FhLd0UMzMzsyLx0p2uaXlEfLmsrF9+JyJG8/7a93z5SGBkhfKFwH65ou/kjl0KXFpW/07gzuaCjIj63PaQJuo0xv2dsvLFvD/Tny8fRfaQYWZmZlZontE3MzMzMysgz+h3MRGxANil2nHkSboL+GhZ8Xci4sGqxNOjB3XDTq9G1/Yh6FEq+RtrzMysS3Kib1UXEcdUOwYzMzOzovHSHTMzMzOzAnKib2ZmZmZWQF66Y1Ymli9n4S8ur3YY/9L31HOqHYKZmZnVIM/om5mZmZkVkBN9MzMzM7MCcqJvZmZmZlZATvTNzMzMzArIiX4NkdRQ7RjyJNVLOqDacZiZmZnZBznRt2ZJau6bmeoBJ/pmZmZmnZAT/RqkzOWSZkuaJWlwKu8l6RFJz6Tyo1J5P0nPSbpB0hxJD0lat5n2S5KulDQFOFPSEZKekjRN0h8l9ZXUDxgGfFvSdEkHSaqTdKekp9PPgc30MVzSTZLGS3pR0hckXZbifkDSWqne+amt2ZJGpGvvkcrqU50fSbq4iX56S3pe0oC0P1rSKatz383MzMxqiSKi2jFYK0lqiIheko4lS7I/DfQBngb2BRYB60XE3yX1ASYB2wHbAH8BBkbEdEm3A+Mi4pYm+ikBz0bEaWl/Y+CtiAhJJwM7RsRZkoYDDRHxk1TvNuDnEfGEpK2BByNixyb6GA4MAg4FdgKeBI6NiPsl3QXcFBF3S9okIt5I5/wGuD0i7pW0MzAW+BZwObBvRPyzib7+HfgBcBUwJCI+XaHOUGAoQF1dn71uu+7aSk1VRY+6vtUOoaY1NDTQq1evaodh7chjWiwez2LxeHa8Qw89dGpEDKx0zP9hVm36BDA6IlYACyU9DuwN3A9cIulgYCWwJdCYJc6PiOlpeyrQr4U+xuS2twLGSNocWBuY38Q5g4CdJDXubyipV0Q09W7B/RGxTNIsoDvwQCqflYvvUEnnAusBmwBzgHsjYk5K/O8D9m8qyQeIiIclHQ9cB+zeRJ0RwAiA7fv3j10X/19TzXW4vscPrnYINa1UKlFfX1/tMKwdeUyLxeNZLB7PzsWJfrGcCNQBe6UEegHQMx1bmqu3Amhy6U7yTm77GuBnETEuLZcZ3sQ53YD9IuK9Vsa7FCAiVkpaFu9/vLQS6CGpJ/Bzsk8iXkqfAvTMnb8r8BawWXOdSOoG7Aj8A9gY+Gsr4zMzMzOrWV6jX5vGA4MldZdUBxwMTAZ6A6+lJP9QsiU77aE38HLa/lqu/G1gg9z+Q2RLaQCQtMca9tuY1C+W1As4Ltf2F8hm+A8GrpG0UTPtfBt4DvgSMLJx/b+ZmZlZkTnRr013ATOBGcCjwLkR8TfgVmBgWgrzVWBuO/U3HLhD0lRgca78XuCYxpdxgTNS/zMlPUv2HsFqi4i3gBuA2cCDZO8ikN4/uBQ4OSLmAdeSrb//gPQS7snAWRExHvgTcN6axGVmZmZWC7x0p4ZERK/0O4Bz0k/++GJg/yZO3yVX7yct9FNftn8PcE+FevOA3cqKW7WgPCKGl+33qnQsIs6jcmK+fa7O1c308zzZsp3G/f9uTXxmZmZmtc4z+mZmZmZmBeQZ/S5M0nVA+XfdXxURI9uxj5OAM8uKJ0TE6e3VR66vu4CPlhV/JyIebFM7PXrQ99RzWq5oZmZm1ok50e/CPoxku0IfI4F2e3Booa9jOqIfMzMzs1rgpTtmZmZmZgXkRN/MzMzMrIC8dMesTCxfxqs/P7+qMWx+2g+q2r+ZmZnVPs/om5mZmZkVkBN9MzMzM7MCcqJvZmZmZlZATvTNzMzMzArIib6ZmZmZWQH5W3es5kgaDuwHLE9FPYBJTZRRqTwihndErGZmZmbV4kTfatV/RMRbAJI2Av6ribKm6pqZmZkVmhN9M0DSUGAoQF1dH2bX7VDVeJ4vlaraf5E0NDRQ8v0sFI9psXg8i8Xj2bk40TcDImIEMAJg+/4fi10Wza1qPJsf/6Wq9l8kpVKJ+vr6aodh7chjWiwez2LxeHYufhnXzMzMzKyAnOibmZmZmRWQE30zMzMzswJyom9mZmZmVkB+Gddq0WvAzZJWpv1uwANNlNFMuZmZmVlhOdG3mhMRPwd+XuFQpbLmys3MzMwKy4m+WRn1WIvNT/tBtcMwMzMzWyNeo29mZmZmVkBO9M3MzMzMCsiJvlmZWP7PaodgZmZmtsac6JuZmZmZFZATfTMzMzOzAnKib2ZmZmZWQE70zczMzMwKyIl+jZHU0EH9jJJ0XNq+UdJObTh3oKSr03a9pAPaKaY/SNqoPdoyMzMzKzr/h1nWoog4uY31pwBT0m490ABMbIc4PrumbZiZmZl1FZ7Rr1HKXC5ptqRZkgan8l6SHpH0TCo/KpX3k/ScpBskzZH0kKR1W9lXSdLAtN2Q+p0j6Y+S9knHX5B0ZKpTL+k+Sf2AYcC3JU2XdFAT7Y+S9AtJk1I79ZJ+neIdlau3QFKftlyLpB6SnpZUn/Z/JOni1t1lMzMzs9rlGf3a9QVgD2B3oA/wtKQ/AYuAYyLi75L6AJMkjUvnbAecEBGnSLodOBa4pY39rg88GhHnSLoLuAj4d2An4CagsS8iYoGkXwINEfGTFtrdGNgfODK1cSBwcrquPSJieln9Vl1LRCyXNAQYK+lbwKeBfcvrSRoKDAWoq+tDqVRqIVyrFQ0NDR7PgvGYFovHs1g8np2LE/3a9QlgdESsABZKehzYG7gfuETSwcBKYEugbzpnfi5hngr0W41+/wk8kLZnAUsjYpmkWavZXqN7IyJSOwsjYhaApDmp3fJEv9XXEhFzJP0GuA/YPyI+8D9iRcQIYATA9v0/GvX19WtwKdaZlEolPJ7F4jEtFo9nsXg8Oxcn+sVzIlAH7JUS8AVAz3Rsaa7eCqBVS3fKLIuISNsrG9uMiJWS1uTvqTG2lawa50oq/5229Vp2Bd4CNlvdAM3MzMxqidfo167xwGBJ3SXVAQcDk4HewGspyT8U2KaaQQJvAxtUMwBJXwA2IbtH1/ibe8zMzKwrcKJfu+4CZgIzgEeBcyPib8CtwMC0BOarwNzqhQjAvcAxzb2M+2FK7ylcCpwcEfOAa4GrOjoOMzMzs47mpTs1JiJ6pd8BnJN+8scXk73UWskuuXrNvhwbEUNy2/Xl/aft4U3EVgJKaXsesFsb+lpQFmf+WL+0ubi115Lux/a5/aubi8XMzMysKDyjb2ZmZmZWQJ7R7+IkXUf2VZZ5V0XEyA+hr+8Bx5cV3xER7fK99h15LWZmZmadnRP9Li4iTu/Avi4GPrT/rKq9rkU91m6PZszMzMyqykt3zMzMzMwKyIm+mZmZmVkBOdE3MzMzMysgJ/pmZmZmZgXkRN/MzMzMrICc6JuZmZmZFZATfTMzMzOzAnKib2ZmZmZWQE70OxFJJUkD26GdhjbWHy7p7ArlW0gam7brJd2Xto+U9N20fbSknVrRxyr1JP1A0qC2xGlmZmZmredEv5Ukdar/RVhS9w+7j4h4JSKOq1A+LiIuTbtHAy0m+uX1IuL8iPhj+0RqZmZmZuU6VfL6YZPUD3gAmArsCcwBvgrsCPwM6AUsBoZExKuSSsB04BPAaEn/B1wArACWRMTBknoCvwAGAsuB/46IxyQNAY4E1gP6A3dFxLkpjl8AewPrAmMj4oJWxt8AXA8MAk5P13MGsDbwFHBaRKxIdS8GPg+8CxwVEQslHQGcl+q/DpwYEQtT87tLehLoA1wWETek9u+LiF3K4hiSrve2dI2HSDoPOBY4DBia+vgL8BVgjwr1vp/aHivpcOAnZH+PTwOnRsRSSQuAm4AjgLWA4yNiboX70g14HjggIhal/XnA/sDlwN9TvP8GnBsRYyu0MTTFTV1dHaVSqclxsNrS0NDg8SwYj2mxeDyLxePZuXSpRD8ZAHwjIiZI+jVwOnAMWTK8SNJg4GLg66n+2hExEEDSLOBTEfGypI3S8dOBiIhdJe0APCRp+3RsD+DjwFLgeUnXRMRLwPci4o00K/+IpN0iYmYrYl8feCoizpK0I/Ad4MCIWCbp58CJwM2p3qSI+J6ky4BTgIuAJ4D9IiIknQycC5yV2t4N2C+dO03S71sKJiImShpHStjTPXorIm5I2xele31NhXqk3z2BUcDhETFP0s3AqcCVqZvFEbGnpNOAs4GTK8SxUtIt6fqvJHsQmpHGE2Bzsoe1HYBxwAcS/YgYAYwAGDBgQNTX17d0+VYjSqUSHs9i8ZgWi8ezWDyenUtXXLrzUkRMSNu3AJ8CdgEeljSdbMZ7q1z9MbntCcAoSacAjUtnPpHaIc02vwg0JvqPRMSSiHgPeBbYJpV/UdIzwDRgZ1q39AWyTxLuTNuHA3sBT6e4Dwc+lo79E7gvbU8F+qXtrYAH0wPLOanvRvdExLsRsRh4DNinlTGV20XS+NTHiWV9VDIAmB8R89L+TcDBueO/q3Adlfya7NMZyB7SRuaO3R0RKyPiWaBvy5dgZmZmVvu64ox+lO2/DcyJiP2bqP/Ov06MGCZpX+BzwFRJe7XQ19Lc9gqgh6SPks1M7x0Rb4wwXdMAACAASURBVEoaBfRsZezvNS7NAQTcFBH/r0K9ZRHReJ0reH+crwF+FhHjJNUDw3PnlN+X8v3WGgUcHREz0hKf+tVsp1HjPcxfxwdExEuSFko6jOwh5cQKbUB238zMzMwKryvO6G8tqTGp/xIwCahrLJO0lqSKs9CS+kfEUxFxPrAI+AgwnpRUpiU7W5OtF2/KhmQPD0sk9QU+s5rX8QhwnKTNUt+bSNqmhXN6Ay+n7a+VHTtKUk9Jm5Il50+3Mo63gQ1y+xsAr0pai1WT7fJ6jZ4H+knaNu1/BXi8lX2Xu5Hs05U7cg9EZmZmZl1SV0z0nyd7kfU5YGOyWe7jgB9LmkH28u0BTZx7uaRZkmYDE4EZwM+BbmmpyhiyF3mXNnE+ETGDbMnOXLKXWSc0Vbc5aRnKeWTvBMwEHiZbi96c4cAdkqaSvXScN5Nsyc4k4IcR8UorQ/ktcI6kaZL6k71k+xTZdc1tpl7jdbwHnJTimgWsBH7Zyr7LjSN7oXpkSxXNzMzMik7vr/Aovqa+RcaKIf0fBFdExEFr0s6AAQPi+eeb+1DGaolfDCsej2mxeDyLxePZ8SRNbfzimHJdcY2+FVD6D7xOZdXlQmZmZmZdVpdK9CNiAdk37HRqkp4C1ikr/kpEzKpGPJ2JpJOAM8uKJ0TE6cClFU4xMzMz65K6VKJfKyJi32rH0FlFxEi8Bt/MzMysRV3xZVwzMzMzs8Jzom9mZmZmVkBO9M3MzMzMCsiJvpmZmZlZATnRNzMzMzMrICf6ZmZmZmYF5ETfOoykfpJmt6H+KEnHpe0bJe1Uoc4QSde2Z5xmZmZmReDv0beaEBEnVzsGMzMzs1riGX3raN0l3SBpjqSHJK0raQ9JkyTNlHSXpI3LT5JUkjQwbZ8kaZ6kycCBuTpHSHpK0jRJf5TUV1I3SX+WVJfqdJP0l8Z9MzMzs6LyjL51tO2AEyLiFEm3A8cC5wLfiojHJf0AuAD4r0onS9ocuBDYC1gCPAZMS4efAPaLiJB0MnBuRJwl6RbgROBKYBAwIyIWlbU7FBgKUFdXR6lUas9rtipqaGjweBaMx7RYPJ7F4vHsXJzoW0ebHxHT0/ZUoD+wUUQ8nspuAu5o5vx9gVJjoi5pDLB9OrYVMCY9DKwNzE/lvwbuIUv0vw6MLG80IkYAIwAGDBgQ9fX1q3Vx1vmUSiU8nsXiMS0Wj2exeDw7Fy/dsY62NLe9AtioHdu+Brg2InYFvgn0BIiIl4CFkg4D9gHub8c+zczMzDolJ/pWbUuANyUdlPa/AjzeTP2ngEMkbSppLeD43LHewMtp+2tl590I3ALcEREr1jxsMzMzs87Nib51Bl8DLpc0E9gD+EFTFSPiVWA48CQwAXgud3g4cIekqcDislPHAb2osGzHzMzMrIi8Rt86TEQsAHbJ7f8kd3i/CvWH5Lbrc9sjqbzO/h6ytfiV7E72Eu7cNoZtZmZmVpOc6FvhSfoucCrZN++YmZmZdQleumOFFxGXRsQ2EfFEtWMxMzMz6yhO9M3MzMzMCsiJvpmZmZlZATnRNzMzMzMrICf6ZmZmZmYF5ETfzMzMzKyAnOibmZmZmRWQE30zMzMzswJyom9mZmZmVkBO9M3MzMzMCsiJfhchqaHaMbSGpBsl7VTtOMzMzMxqXY9qB2DWSFL3iDi52nGYmZmZFYEiotoxWAeQ1BARvSTVA2dHxOdT+bXAlIgYJWkBcBNwBLAWcHxEzJXUC7gGGAgEcGFE3Fmhj2FA/4g4J+0PAQZGxH9Kuhv4CNATuCoiRjTGBVwPDAJOBy5K8U2R9Atgb2BdYGxEXJDOaVOckj4JXAisA/wvcFJErPIJh6ShwFCAurq6vW6//fbVvdXWyTQ0NNCrV69qh2HtyGNaLB7PYvF4drxDDz10akQMrHTMM/pWbnFE7CnpNOBs4GTg+8CSiNgVQNLGTZx7J/AkcE7aHwxcnLa/HhFvSFoXeFrSnRHxOrA+8FREnJXazrf3vXROd+ARSbtFxMy2xCmpD3AeMCgi3pH0HeC/gR/kO0oPHiMABgwYEPX19W26adZ5lUolPJ7F4jEtFo9nsXg8Oxev0bdyv0u/pwL90vYg4LrGChHxZqUTI2IR8IKk/SRtCuwATEiHz5A0A5hENrO/XSpfQfaAUMkXJT0DTAN2BvJr91sb537pvAmSpgNfA7Zpoj8zMzOzwvCMfteznFUf8HqWHV+afq9g9f4+fgt8EZgL3BURkZYLDQL2j4h/SCrl+n0vIlaUNyLpo2Qz9XtHxJuSRpXF2to4BTwcESesxrWYmZmZ1SzP6Hc9LwI7SVpH0kbA4a0452Gy9fNAs0t3AO4CjgJOIEv6AXoDb6YkfweyWfaWbAi8AyyR1Bf4zGrGOQk4UNK2qWx9Sdu3oi0zMzOzmuZEv4uJiJeA24HZ6fe0Vpx2EbCxpNlp+c2hzbT/JvAcsE1ETE7FDwA9JD0HXEqWfLcU54wU21zgNt5fAtSmONNyoiHAaEkzyd4h2KEVbZmZmZnVNC/d6SIioldu+1zg3Ap1+uW2pwD1abuBbG17a/v6fNn+UpqYkc/Hlfbrc9tDmjinTXFGxKNk395jZmZm1mV4Rt/MzMzMrIA8o2+rRdJTZN9Ln/eViJhVjXjMzMzMbFVO9G21RMS+1Y7BzMzMzJrmpTtmZmZmZgXkRN/MzMzMrICc6JuZmZmZFZATfTMzMzOzAnKib2ZmZmZWQE70zczMzMwKyIm+mZmZmVkBOdG31SJpgaQ+FconViMeMzMzM1uVE31rM0ndmzoWEQd0ZCyrQ5L/ozgzMzMrPCf6XYykcySdkbavkPRo2j5M0q2STpA0S9JsST/Ondcg6aeSZgD758rXlXS/pFMa66Xf9ZJKksZKmpvaVjr22VQ2VdLVku5rItZukv4sqS63/xdJdennTklPp58DU519JD0paZqkiZIGpPIhksal632k/e+smZmZWefimc2uZzxwFnA1MBBYR9JawEHAPODHwF7Am8BDko6OiLuB9YGnIuIsgJSz9wJ+C9wcETdX6OvjwM7AK8AE4EBJU4DrgYMjYr6k0U0FGhErJd0CnAhcCQwCZkTEIkm3AVdExBOStgYeBHYE5gIHRcRySYOAS4BjU5N7ArtFxBvlfUkaCgwFqKuro1QqtXQfrUY0NDR4PAvGY1osHs9i8Xh2Lk70u56pwF6SNgSWAs+QJfwHAfcCpYhYBCDpVuBg4G5gBXBnWVv3AJdFxK1N9DU5Iv6a2poO9AMagBciYn6qM5qUYDfh16mfK4GvAyNT+SBgp/TAAbChpF5Ab+AmSdsBAayVa+vhSkk+QESMAEYADBgwIOrr65sJyWpJqVTC41ksHtNi8XgWi8ezc/HSnS4mIpYB84EhwESyGf5DgW2BBc2c+l5ErCgrmwB8Wrlsu8zS3PYKVuPBMiJeAhZKOgzYB7g/HeoG7BcRe6SfLSOiAfgh8FhE7AIcAfTMNfdOW/s3MzMzq1VO9Lum8cDZwJ/S9jBgGjAZOERSn/TC7QnA4820cz7ZEp/r2tD388DHJPVL+4Nbcc6NwC3AHbmHjYeAbzVWkLRH2uwNvJy2h7QhLjMzM7NCcaLfNY0HNgeejIiFwHvA+Ih4Ffgu8BgwA5gaEfe00NaZwLqSLmtNxxHxLnAa8ICkqcDbwJIWThtH9j7AyFzZGcBASTMlPUv2sAJwGfAjSdPw0jQzMzPrwpwIdUER8Qi5tesRsX1uezTZuvnyc3qV7ffL7Z5UXi8iSkApV/6fufqPRcQOacnPdcCUFkLenewl3Lm59hZT4dOAiHgS2D5XdF4qHwWMaqEfMzMzs8LwjL5Vwynp5dw5ZEttrm+qoqTvkr0E/P86KDYzMzOzQvCMvnW4iLgCuCJfJukksmVAeRMi4nTg0o6KzczMzKwonOhbpxARI1l1Db6ZmZmZrQEv3TEzMzMzKyAn+mZmZmZmBeRE38zMzMysgJzom5mZmZkVkBN9MzMzM7MCcqJvZmZmZlZATvTNzMzMzArIib6tQtICSX0qlE+sRjxtJale0gG5/VGSjqtmTGZmZmbV4ETf/kVS96aORcQBTR3rZOqBWonVzMzM7EPjRL8gJJ0j6Yy0fYWkR9P2YZJulXSCpFmSZkv6ce68Bkk/lTQD2D9Xvq6k+yWd0lgv/a6XVJI0VtLc1LbSsc+msqmSrpZ0XzPxDpd0k6Txkl6U9AVJl6UYH5C0Vqp3uKRpqfzXktZJ5QskXSjpmXRsB0n9gGHAtyVNl3RQ6u5gSRMlveDZfTMzM+sqelQ7AGs344GzgKuBgcA6KVk+CJgH/BjYC3gTeEjS0RFxN7A+8FREnAWQcvZewG+BmyPi5gp9fRzYGXgFmAAcKGkKcD1wcETMlzS6FTH3Bw4FdgKeBI6NiHMl3QV8TtIDwCjg8IiYJ+lm4FTgynT+4ojYU9JpwNkRcbKkXwINEfGTdD3fADYHPgHsAIwDxpYHImkoMBSgrq6OUqnUivCtFjQ0NHg8C8ZjWiwez2LxeHYuTvSLYyqwl6QNgaXAM2QJ/0HAvUApIhYBSLoVOBi4G1gB3FnW1j3AZRFxaxN9TY6Iv6a2pgP9gAbghYiYn+qMJiXOzbg/IpZJmgV0Bx5I5bNSmwOA+RExL5XfBJzO+4n+73LX/oVm+rk7IlYCz0rqW6lCRIwARgAMGDAg6uvrWwjdakWpVMLjWSwe02LxeBaLx7Nz8dKdgoiIZcB8YAgwkWyG/1BgW2BBM6e+FxErysomAJ9uXJJTwdLc9gpW/4FxKUBKwpdFRKTyla1sszGOlmLIx9vUNZmZmZkVihP9YhkPnA38KW0PA6YBk4FDJPVJL9yeADzeTDvnky3xua4NfT8PfCytkwcY3KbIm26zn6Rt0/5XaD5ugLeBDdqhbzMzM7Oa5kS/WMaTrUd/MiIWAu8B4yPiVeC7wGPADGBqRNzTQltnAutKuqw1HUfEu8BpwAOSppIl3EtW7zL+1eZ7wEnAHWl5z0rgly2cdi9wTNnLuGZmZmZdjtfoF0hEPAKsldvfPrc9mmzdfPk5vcr2++V2TyqvFxEloJQr/89c/cciYoe05Oc6YEozsQ5vKo78sXRNH69wfr/c9hSyr9UkreffLVd1fFP9mJmZmRWZZ/StPZ2SXs6dA/Qm+xYeMzMzM6sCz+hbu4mIK4Ar8mWSTiJbBpQ3ISJO77DAzMzMzLogJ/r2oYqIkcDIasdhZmZm1tV46Y6ZmZmZWQE50TczMzMzKyAn+mZmZmZmBeRE38zMzMysgJzom5mZmZkVkBN9MzMzM7MCcqJvZmZmZlZATvTNzMzMzArIib7VHEl3S5oqaY6koansG5LmSZos6QZJ16byOkl3Sno6/RxY3ejNzMzMOob/Z1yrRV+PiDckrQs8Len3wPeBPYG3gUeBGanuVcAVEfGEpK2BB4EdqxG0mZmZWUdyom+16AxJx6TtjwBfAR6PiDcAJN0BbJ+ODwJ2ktR47oaSekVEQ77B9MnAUIC6ujpKpdKHewXWYRoaGjyeBeMxLRaPZ7F4PDsXJ/pWUyTVkyXv+0fEPySVgLk0PUvfDdgvIt5rrt2IGAGMABgwYEDU19e3V8hWZaVSCY9nsXhMi8XjWSwez87Fa/St1vQG3kxJ/g7AfsD6wCGSNpbUAzg2V/8h4FuNO5L26NBozczMzKrEib7VmgeAHpKeAy4FJgEvA5cAk4EJwAJgSap/BjBQ0kxJzwLDOjxiMzMzsyrw0h2rKRGxFPhMebmkKRExIs3o3wXcneovBgZ3bJRmZmZm1ecZfSuK4ZKmA7OB+aRE38zMzKyr8oy+FUJEnF3tGMzMzMw6E8/om5mZmZkVkBN9MzMzM7MCcqJvZmZmZlZATvTNzMzMzArIib6ZmZmZWQE50TczMzMzKyAn+mZmZmZmBeRE38zMzMysgJzom5mZmZkVkBP9GiTpSEnfXY3zJn4Y8bSh/yGSrq1i//0kfala/ZuZmZl1JCf6NSgixkXEpatx3gEfRjw1pB/gRN/MzMy6BCf6nUyadZ4raZSkeZJulTRI0gRJf5a0T35mXNLxkmZLmiHpT6lsZ0mTJU2XNFPSdqm8If2ul1SSNDb1daskpWOfTWVTJV0t6b4m4uwmaYGkjXJlf5bUV9IRkp6SNE3SHyX1rXD+KEnH5fYbctvnSHo6xX5hC/frq6neDEm/ybV9taSJkl7I9XMpcFC6L99uzXiYmZmZ1aoe1Q7AKtoWOB74OvA02Sz0J4Ajgf8B7s7VPR/4VES8nEu6hwFXRcStktYGulfo4+PAzsArwATgQElTgOuBgyNivqTRTQUYESsl3QMcA4yUtC/wYkQslPQEsF9EhKSTgXOBs1pz4ZI+CWwH7AMIGCfp4Ij4U4W6OwPnAQdExGJJm+QOb052z3YAxgFjge8CZ0fE5yu0NRQYClBXV0epVGpNuFYDGhoaPJ4F4zEtFo9nsXg8Oxcn+p3T/IiYBSBpDvBISppnkS0/yZsAjJJ0O/C7VPYk8D1JWwG/i4g/V+hjckT8NfUxPbXbALwQEfNTndGk5LcJY8geNEYC/5H2AbYCxkjaHFgbmF/59Io+mX6mpf1eZIn/BxJ94DDgjohYDBARb+SO3R0RK4FnK32iUC4iRgAjAAYMGBD19fVtCNk6s1KphMezWDymxeLxLBaPZ+fipTud09Lc9src/krKHs4iYhjZrPZHgKmSNo2I28hm/98F/iDpsBb6WFHebis9CWwrqQ44mvcfNK4Bro2IXYFvAj0rnLuc9PcnqRvZAwFks/g/iog90s+2EfGr1Ygtf31ajfPNzMzMapoT/RonqX9EPBUR5wOLgI9I+hjZzPzVwD3Abq1s7nngY5L6pf3BzVWOiADuAn4GPBcRr6dDvYGX0/bXmjh9AbBX2j4SWCttPwh8XVKvdH1bStqsiTYeBY6XtGmqu0kT9Rq9DWzQQh0zMzOzQnCiX/sulzRL0mxgIjAD+CIwOy3J2QW4uTUNRcS7wGnAA5KmkiXGS1o4bQzwZd5ftgMwHLgjtbG4ifNuAA6RNAPYH3gnxfAQcBvwZFqqNJYmkvOImANcDDye2vlZC7HOBFakF3f9Mq6ZmZkVmrJJWbOMpF4R0ZC+hec64M8RcUW14+pIAwYMiOeff77aYVg78XrR4vGYFovHs1g8nh1P0tSIGFjpmGf0rdwp6ZOAOWRLcK6vcjxmZmZmthr8rTu2ijR7v8oMvqSTgDPLqk6IiNM7Iqa0Bv+RCocOz70XYGZmZmY5TvStRRExkuwrNKvV/+vAHtXq38zMzKwWeemOmZmZmVkBOdE3MzMzMysgJ/pmZmZmZgXkRN/MzMzMrICc6JuZmZmZFZATfTMzMzOzAnKib2ZmZmZWQE70rcuQdLSknaodh5mZmVlHcKJvXcnRgBN9MzMz6xKc6FunIunLkiZLmi7pekmnS7o8d3yIpGubqNs9lTdIuljSDEmTJPWVdABwJHB5qt+/OldoZmZm1jEUEdWOwQwASTsClwFfiIhlkn4OPAV8PyK2TXXuBy4GXq9Qd1JE/H/27j1Kr7K++//7Iwc5DCLKPBT10bQig4hKTZCT4sQf9dBakQIPpXgIHlJbD/WE+LRUUxe0oNQTHmi0ECuUElQkagVRvDFEjpEcQA66JM9SiwgegAGNGL6/P+6dejudmYRkMnPPzvu1Vlaufe1rX9d339f88b33fHfyb0kKeGlVfTHJ+4B7q+qUJIuAL1XVZ8dYez4wH2BwcHD24sWLt/wNa0qMjIwwMDAw3WFoErmn7eJ+tov7OfXmzp27vKrmjHVu26kORprA/wfMBq5LArAj8BPg+0kOAr4L7AMsA94wzliAXwNfatrLgT/a0MJVtRBYCDA0NFTDw8OTckOafp1OB/ezXdzTdnE/28X97C8m+uonAT5dVf/3dzqTVwP/B7gFuKiqKt3s/n+MbTxYv/1V1Tr8OZckSVsha/TVT74OHJ3kfwEkeUySJwEXAUcAxwH/sYGxE7kP2GWLRC5JktRnTPTVN6rqO8DJwFeTrAIuA/asqp8DNwNPqqprJxq7gSX+AzgxyQ2+jCtJktrOkgb1laq6ALhgjP6XPIyxAz3tzwKfbdrL8J/XlCRJWwmf6EuSJEktZKIvSZIktZCJviRJktRCJvqSJElSC5noS5IkSS1koi9JkiS1kIm+JEmS1EIm+pIkSVILmehLkiRJLWSiL0mSJLWQib76UpJOkjkbGDMvyUenKiZJkqSZxERfkiRJaiETfU2KJCcmeXPT/mCSy5v285Ocl+QFSa5K8u0kFyYZaM7PTnJFkuVJLk2y56h5H5FkUZJTmuMTktyW5Frg0J5xf5rkmiQ3JPlakj2aa7+bZLBnru+tP5YkSWqzVNV0x6AWSHIQ8PaqOibJUuCRdBPxvwV+BfwJ8OKquj/JSc35fwKuAI6oqruSHAu8sKpenaQDvAv4G+DGqjq1+RJwDTAbuAf4BnBDVb0xyW7AL6qqkrwWeGpVvT3Je4B7qupDSV4A/GVVHTVG/POB+QCDg4OzFy9evKU+Kk2xkZERBgYGpjsMTSL3tF3cz3ZxP6fe3Llzl1fVmOXO2051MGqt5cDsJI8C1gLfBuYAzwWWAPsCy5IAbA9cBQwB+wGXNf3bAHf0zPkvwOKqOrU5PhDoVNVdAEkuAPZuzj0BuKD5MrA9cHvTfzZwMfAh4NXAOWMFX1ULgYUAQ0NDNTw8vIkfg/pNp9PB/WwX97Rd3M92cT/7i4m+JkVVPZjkdmAe8C1gFTAX2Itu0n1ZVR3Xe02SpwM3VdXB40z7LWBukn+uql9tIIQzgQ9U1ZIkw8CCJq4fJLkzyfOBZwPHb8r9SZIkzTTW6GsyLQXeAXyzab8euAG4Gjg0yV4ASXZOsjdwKzCY5OCmf7skT+uZ71+B/wQWJ9mWbtnO85I8Nsl2wDE9Y3cFftS0XzUqrk8B5wIXVtW6SbtbSZKkPmair8m0FNgTuKqq7qRbm7+0KbWZB5yfZBXdsp19qurXwNHA6UlWAiuAQ3onrKoP0P2y8BngTrpP6q8ClgE39wxdAFyYZDlw96i4lgADjFO2I0mS1EaW7mjSVNXXge16jvfuaV8OHDDGNSuAw8boH+5pv6fn1DmMkbBX1cV0a/HH8kxgZVXdssGbkCRJagkTfbVakncBf4W1+ZIkaStj6Y5arapOq6onVdWV0x2LJEnSVDLRlyRJklrIRF+SJElqIRN9SZIkqYVM9CVJkqQWMtGXJEmSWshEX5IkSWohE31JkiSphUz0JUmSpBYy0VdfSzIryY3THYckSdJMY6KvrU6Sbac7BkmSpC3NRF8zwTZJPpnkpiRfTbJjkv2TXJ1kVZKLkuwGkKSTZE7T3j3JmqY9L8mSJJcDX5++W5EkSZoaPtnUTPAU4Liqel2SxcBRwDuBN1XVFUneC7wHeMsG5nkW8Iyq+tnoE0nmA/MBBgcH6XQ6kxm/ptHIyIj72TLuabu4n+3ifvYXE33NBLdX1YqmvRx4MvDoqrqi6fs0cOFGzHPZWEk+QFUtBBYCDA0N1fDw8OZFrL7R6XRwP9vFPW0X97Nd3M/+YumOZoK1Pe11wKMnGPsbfvtzvcOoc/dPZlCSJEn9zERfM9E9wM+TPLc5fgWw/un+GmB20z56iuOSJEnqG5buaKZ6FXBWkp2A7wMnNP1nAIubmvsvT1dwkiRJ081EX32tqtYA+/Ucn9Fz+qAxxt8CPKOn6+SmfxGwaEvEKEmS1I8s3ZEkSZJayERfkiRJaiETfUmSJKmFTPQlSZKkFjLRlyRJklrIRF+SJElqIRN9SZIkqYVM9CVJkqQWMtGXJEmSWshEX5IkSWohE31JkiSphUz0NWmSdJLM2cCYeUk+uoXjmJPkI1tyDUmSpH637XQHIE22qroeuH6645AkSZpOPtHfiiU5Mcmbm/YHk1zetJ+f5LwkL0hyVZJvJ7kwyUBzfnaSK5IsT3Jpkj1HzfuIJIuSnNIcn5DktiTXAof2jPvTJNckuSHJ15Ls0Vz73SSDPXN9b/3xGPdwTJIbk6xM8s2mbzjJl5r2giRnN79t+P76+5UkSWo7n+hv3ZYCbwc+AswBHplkO+C5wCrgZODwqro/yUnA25L8E3AmcERV3ZXkWOBU4NXNnNsC5wE3VtWpzZeAfwBmA/cA3wBuaMZeCRxUVZXktcA7q+rtSc4Fjgc+BBwOrKyqu8a5h3cDL6yqHyV59Dhj9gHmArsAtyb5RFU92DsgyXxgPsDg4CCdTmfDn55mhJGREfezZdzTdnE/28X97C8m+lu35cDsJI8C1gLfppvwPxdYAuwLLEsCsD1wFTAE7Adc1vRvA9zRM+e/AIur6tTm+ECgsz5RT3IBsHdz7gnABc2Xge2B25v+s4GL6Sb6rwbOmeAelgGLkiwGPj/OmC9X1VpgbZKfAHsAP+wdUFULgYUAQ0NDNTw8PMGSmkk6nQ7uZ7u4p+3ifraL+9lfTPS3YlX1YJLbgXnAt+g+xZ8L7EU36b6sqo7rvSbJ04Gbqurgcab9FjA3yT9X1a82EMKZwAeqakmSYWBBE9cPktyZ5PnAs+k+3R/vHl6f5EDgT4DlSWaPMWxtT3sd/txLkqStgDX6Wgq8A/hm03493dKaq4FDk+wFkGTnJHsDtwKDSQ5u+rdL8rSe+f4V+E9gcZJtgWuA5yV5bFMWdEzP2F2BHzXtV42K61PAucCFVbVuvOCTPLmqrqmqdwN3Af/7YX8CkiRJLWSir6XAnsBVVXUn8CtgaVNqMw84P8kqumU7+1TVr4GjgdOTrARWAIf0TlhVH6D7ZeEzwJ10n9RfRbfM5uaeoQuAC5MsB+4eFdcSYICJy3YA3p9kdZIb6f42YeVG37kkSVKLWcKwlauqrwPb9Rzv3dO+HDhgjGtWAIeN0T/c035Pz6lzGCNhr6qL6dbij+WZdF/CvWUD8f/ZGN2d5g9VWhN8UQAAIABJREFUtWDU+P0mmk+SJKktTPTVd5K8C/grJqjNlyRJ0sRM9NV3quo04LTeviR/x+/W90O3fv9UJEmS9D+Y6GtGaBJ6k3pJkqSN5Mu4kiRJUguZ6EuSJEktZKIvSZIktZCJviRJktRCJvqSJElSC5noS5IkSS1koi9JkiS1kIm+NluSTpI5GxgzL8lHpyomSZKkrZ2JviRJktRCJvpboSQnJnlz0/5gksub9vOTnJfkBUmuSvLtJBcmGWjOz05yRZLlSS5NsueoeR+RZFGSU5rjE5LcluRa4NCecX+a5JokNyT5WpI9mmu/m2SwZ67vrT8e4x72SHJRkpXNn0Oa/rclubH585amb+ckX27G3Zjk2En/UCVJkvpMqmq6Y9AUS3IQ8PaqOibJUuCRdBPxvwV+BfwJ8OKquj/JSc35fwKuAI6oqruaZPmFVfXqJB3gXcDfADdW1anNl4BrgNnAPcA3gBuq6o1JdgN+UVWV5LXAU6vq7UneA9xTVR9K8gLgL6vqqHHu4QLgqmbsNsAAsBewCDgISLP+y4E/AF5UVa9rrt21qu4ZNd98YD7A4ODg7MWLF2/WZ6z+MTIywsDAwHSHoUnknraL+9ku7ufUmzt37vKqGrOEetupDkZ9YTkwO8mjgLXAt4E5wHOBJcC+wLIkANsDVwFDwH7AZU3/NsAdPXP+C7C4qk5tjg8EOlV1F/x3Yr53c+4JwAXNl4Htgdub/rOBi4EPAa8GzpngHp4PvBKgqtYB9yR5DnBRVd3frPn55p4uAf45yenAl6pq6ejJqmohsBBgaGiohoeHJ1haM0mn08H9bBf3tF3cz3ZxP/uLpTtboap6kG5yPQ/4FrAUmEv3ifjtwGVVtX/zZ9+qeg3dJ+Q39fQ/vape0DPtt4C5SXbYiBDOBD5aVU8H/hLYoYnrB8CdSZ4PPBv4yiTd723As4DVwClJ3j0Z80qSJPUzE/2t11LgHcA3m/brgRuAq4FDk+wF/13fvjdwKzCY5OCmf7skT+uZ71+B/wQWJ9mWbtnM85I8Nsl2wDE9Y3cFftS0XzUqrk8B5wIXNk/qx/N14K+aWLZJsmtzHy9LslOSnYEjgaVJHgc8UFXnAu+nm/RLkiS1mon+1mspsCfdOvc76dbmL21KbeYB5ydZRbdsZ5+q+jVwNHB6kpXACuCQ3gmr6gN0vyx8BrgTWNBcvwy4uWfoAuDCJMuBu0fFtYRuvf1EZTvQfR9gbpLVdEuR9q2qb9Ot0b+W7heNT1XVDcDTgWuTrADeA5yygbklSZJmPGv0t1JV9XVgu57jvXvalwMHjHHNCuCwMfqHe9rv6Tl1DmMk7FV1Md1a/LE8E1hZVbdsIP47gSPG6P8A8IFRfZcCl040nyRJUtuY6KtvJHkX3XKc46c7FkmSpJnORF99o6pOA07r7Uvyd/xufT906/dPRZIkSeMy0VdfaxJ6k3pJkqSHyZdxJUmSpBYy0ZckSZJayERfkiRJaiETfUmSJKmFTPQlSZKkFjLRlyRJklrIRF+SJElqIRN9Tboks5LcOM0x/GeSR09nDJIkSdPJ/zBLM0KSbavqNxs7vqr+eEvGI0mS1O98oq8tZZskn0xyU5KvJtkxyf5Jrk6yKslFSXYDSNJJMqdp755kTdOel2RJksuBr4+1SJI9k3wzyYokNyZ5btO/pplrVpKbR8cyNR+BJEnS9PGJvraUpwDHVdXrkiwGjgLeCbypqq5I8l7gPcBbNjDPs4BnVNXPxjn/F8ClVXVqkm2AnTYylnN7BySZD8wHGBwcpNPpbNRNqv+NjIy4ny3jnraL+9ku7md/MdHXlnJ7Va1o2suBJwOPrqormr5PAxduxDyXTZDkA1wHnJ1kO+ALPWtOFMus0QOqaiGwEGBoaKiGh4c3IjTNBJ1OB/ezXdzTdnE/28X97C+W7mhLWdvTXgdM9GLsb/jtz+IOo87dP9EiVfVN4DDgR8CiJK/ciFj8gitJklrPRF9T5R7g5+tr6IFXAOuf7q8BZjftox/OpEmeBNxZVZ8EPkW31EeSJGmr55NNTaVXAWcl2Qn4PnBC038GsLipk//yw5xzGDgxyYPACDDWE31JkqStjom+Jl1VrQH26zk+o+f0QWOMvwV4Rk/XyU3/ImDRBtb6NN16/9H9s5rm3RPEIkmS1FqW7kiSJEkt5BN9zQhJng58ZlT32qo6cDrikSRJ6ncm+poRqmo1sP90xyFJkjRTWLojSZIktZCJviRJktRCJvqSJElSC5noS5IkSS1koi9JkiS1kIm+JEmS1EIm+pIkSVILmehr0iSZleTG6Y5DkiRJJvrqc0n8T90kSZI2gYm+Jts2ST6Z5KYkX02yY5L9k1ydZFWSi5LsBpCkk2RO0949yZqmPS/JkiSXA18fb6EkJyVZnWRlktOavvHWenOS7zT9/7GlPwRJkqTplqqa7hjUEklmAd8D5lTViiSLgSXAO4E3VdUVSd4LPKqq3pKkA7yjqq5PsjtwfVXNSjIPOAV4RlX9bJy1Xgz8PXB4VT2Q5DFV9bMkq8ZZ67+A36+qtUkeXVW/GDXffGA+wODg4OzFixdP9sejaTIyMsLAwMB0h6FJ5J62i/vZLu7n1Js7d+7yqpoz1jnLIjTZbq+qFU17OfBk4NFVdUXT92ngwo2Y57LxkvzG4cA5VfUAQJPk7zrBWquA85J8AfjC6MmqaiGwEGBoaKiGh4c3IkTNBJ1OB/ezXdzTdnE/28X97C+W7miyre1prwMePcHY3/Dbn8EdRp27fzKDAv4E+BjwLOA6a/8lSVLbmehrS7sH+HmS5zbHrwDWP3FfA8xu2kc/zHkvA05IshNAU7oz5lpJHgH876r6BnASsCvg7xUlSVKr+VRTU+FVwFlNUv594ISm/wxgcVMf/+WHM2FVXZJkf+D6JL8G/hP423HW2gY4tyntCfCR0TX6kiRJbWOir0lTVWuA/XqOz+g5fdAY428BntHTdXLTvwhYtBHrnQacNqpvxVhrAc/Z0HySJEltYumOJEmS1EI+0VdfS/J04DOjutdW1YHTEY8kSdJMYaKvvlZVq4H9pzsOSZKkmcbSHUmSJKmFTPQlSZKkFjLRlyRJklrIRF+SJElqIRN9SZIkqYVM9CVJkqQWMtGXJEmSWshEX1MqyZuT3JzkvM2c56VJ3jVZcUmSJLWN/2GWptpfA4dX1Q83NDDJtlX1m7HOVdUSYMlkBydJktQWPtHXlElyFvAHwFeSvD3JF5KsSnJ1kmc0YxYk+UySZcBnkgwm+VyS65o/hzbj5iX5aNN+cjPH6iSnJBlp+oeTdJJ8NsktSc5Lkmm6fUmSpCnlE31Nmap6fZIXAXOB9wA3VNXLkjwf+Ddg/2bovsBzquqXSf4d+GBVXZnkicClwFNHTf1h4MNVdX6S148694fA04D/ApYBhwJXjo4tyXxgPsDg4CCdTmfzb1h9YWRkxP1sGfe0XdzPdnE/+4uJvqbLc4CjAKrq8iSPTfKo5tySqvpl0z4c2LfnQfyjkgyMmutg4GVN+9+BM3rOXbu+TCjJCmAWYyT6VbUQWAgwNDRUw8PDm35n6iudTgf3s13c03ZxP9vF/ewvJvrqR/f3tB8BHFRVv+od8DAqcNb2tNfhz7wkSdpKWKOv6bIUOB66tfTA3VV17xjjvgq8af1Bkv3HGHM1zW8HgD+f3DAlSZJmJhN9TZcFwOwkq4DTgFeNM+7NwJzmpd3vAKNr8AHeArytmWsv4J4tEK8kSdKMYhmDplRVzeo5fNkY5xeMOr4bOHaMcYuARc3hj+iW91SSPweGmjEdoNNzzRs3I3RJkqQZxURfbTAb+GjzT2f+Anj1NMcjSZI07Uz0NeNV1VLgmdMdhyRJUj+xRl+SJElqIRN9SZIkqYVM9CVJkqQWMtGXJEmSWshEX5IkSWohE31JkiSphUz0JUmSpBYy0ZckSZJayERfkiRJaiET/T6UZGS6Y+iVZDjJIdMdhyRJkjaeib4ASLLtBKeHARN9SZKkGSRVNd0xaJQkI1U1kCTA+4AXAwWcUlUXJBkALgZ2A7YDTq6qi5PMAr4CXEk3Mf8RcERV/XKcdTrACuA5wPnAbcDJwPbAT4HjgR2Bq4F1wF3Am4BbgLOAJzZTvaWqlo2zxgLg94E/aMa/FTiouacfAX9aVQ8mmQ18ABgA7gbmVdUdSV4HzG9i+h7wiqp6IMki4F5gDvB7wDur6rPjxPBvwOer6gvN8XnA4qq6uGfM/GYdBgcHZy9evHisqTQDjYyMMDAwMN1haBK5p+3ifraL+zn15s6du7yq5ox1zkS/D/Uk+kcBrwdeBOwOXAccSDfh3qmq7k2yO91E/CnAk+gmw3OqakWSxcCSqjp3nHU6wHeq6q+b492AX1RVJXkt8NSqenuTrI9U1RnNuH8HPl5VVyZ5InBpVT11nDUWAIcDc4F9gauAo6rqK0kuAj4NfBm4gu6XkruSHAu8sKpeneSxVfXTZq5TgDur6swm0d8ZOBbYp7nPvcaJ4XnAW6vqZUl2pfvl5ilV9Zuxxg8NDdWtt9461inNQJ1Oh+Hh4ekOQ5PIPW0X97Nd3M+pl2TcRH+icg1Nv+cA51fVOuDOJFcAB9B9av+PSQ4DHgIeD+zRXHN7Va1o2suBWRtY44Ke9hOAC5LsSfcJ+u3jXHM4sG/3Fw4APCrJQFWN927BV5qn9quBbYBLmv7VTXxDwH7AZc2c2wB3NGP2axL8R9N92n9pz7xfqKqHgO8k2YNxVNUVST6eZBA4CvjceEm+JElSW5joz0zHA4PA7CaBXgPs0Jxb2zNuHd3Sm4nc39M+E/hAVS1JMgwsGOeaRwAHVdWvNjLetQBV9VCSB+u3v0Z6iO7PYICbqurgMa5dBLysqlYmmUf3fYHfmbcRJvZvwMuBPwdO2Mi4JUmSZixfxu1vS4Fjk2zTPI0+DLgW2BX4SZPkz6VbsjMZdqVbNw/wqp7++4Bdeo6/SrdWH4Ak+2/murcCg0kObubbLsnTmnO7AHck2Y7uF5xNtQh4C0BVfWcz5pEkSZoRTPT720XAKmAlcDndF05/DJwHzGlKYV5J9+XYybAAuDDJcrovxK73ReDIJCuSPBd4c7P+qiTfofsewSarql8DRwOnJ1lJt4Z+/b/y8/fANcAyNuM+q+pO4GbgnM2JVZIkaaawdKcPVdVA83cBJzZ/es/fDYxV5gLdWvf1487YwDrDo44vpvuv+YwedxvwjFHdx040d8+1C0YdD4x1rnmv4LAxrv8E8Ikx+ueNN+9YkuxE94Xl8zcmbkmSpJnOJ/pqvSSH032af2ZV3TPd8UiSJE0Fn+hvBZJ8DDh0VPeHq2rSyliSnAD8zajuZVX1hslaYyNieDrwmVHda6vqQCbvPQZJkqQZwUR/KzAVyXbzpWFa69+rajWwuS8GS5IktYKlO5IkSVILmehLkiRJLWSiL0mSJLWQib4kSZLUQib6kiRJUguZ6EuSJEktZKIvSZIktZCJviZdkpHpjmEsSd6SZKfpjkOSJGkqmOhra/IWwERfkiRtFVJV0x2DWibJSFUNJAnwPuDFQAGnVNUFSQaAi4HdgO2Ak6vq4iSzgK8AVwKHAD8CjqiqX46zzl7AWcAgsA44BvjfwALgbmA/YDnwcuBNwBnArcDdVTV31FzzgfkAg4ODsxcvXjwpn4Wm38jICAMDA9MdhiaRe9ou7me7uJ9Tb+7cucuras5Y50z0Nel6Ev2jgNcDLwJ2B64DDgTuAnaqqnuT7A5cDTwFeBLwPWBOVa1IshhYUlXnjrPONcBpVXVRkh3o/obq2XS/RDwN+C9gGXBiVV2ZZE0z990TxT80NFS33nrrZn4K6hedTofh4eHpDkOTyD1tF/ezXdzPqZdk3ETf0h1tSc8Bzq+qdVV1J3AFcAAQ4B+TrAK+Bjwe2KO55vaqWtG0lwOzxpo4yS7A46vqIoCq+lVVPdCcvraqflhVDwErxptDkiSpzbad7gC0VTqebrnN7Kp6sHnSvkNzbm3PuHXAjpsw/+g5/DmXJElbHZ/oa0taChybZJskg8BhwLXArsBPmiR/Lt2SnYelqu4DfpjkZQBJHrkR/6LOfcAuD3ctSZKkmchEX1vSRcAqYCVwOfDOqvoxcB4wJ8lq4JXALZs4/yuANzclQN8Cfm8D4xcClyT5xiauJ0mSNGNY0qBJV1UDzd8FnNj86T1/N3DwOJfv1zPujA2s813g+aO6vw90esa8sad9JnDmBm9AkiSpBXyiL0mSJLWQT/TV95J8DDh0VPeHq+qc6YhHkiRpJjDRV9+rqjdMdwySJEkzjaU7kiRJUguZ6EuSJEktZKIvSZIktZCJviRJktRCJvqSJElSC5noS5IkSS1koi9JkiS1kIl+H0kyMkXrLEpydNP+VJJ9H8a1c5J8pGkPJzlkS8UpSZKkTed/mLWVq6rXPszx1wPXN4fDwAjwrUkOS5IkSZvJJ/p9KF3vT3JjktVJjm36B5J8Pcm3m/4jmv5ZSW5O8skkNyX5apIdN3KtTpI5TXukWfemJF9L8uzm/PeTvLQZM5zkS0lmAa8H3ppkRZLnjjP/oiSfSHJ1M89wkrObeBf1jHtBkquae7swyUDT/+4k1zWfxcIk6Yn79CTXJrltvPWbsd9Msn/P8ZVJnrkxn48kSdJM5RP9/vRnwP7AM4HdgeuSfBO4Cziyqu5NsjtwdZIlzTVPAY6rqtclWQwcBZz7MNfdGbi8qk5MchFwCvBHwL7Ap4H1a1FVa5KcBYxU1RkbmHc34GDgpc0chwKvbe5rf+CHwMnA4VV1f5KTgLcB7wU+WlXvBUjyGeAlwBebebetqmcn+WPgPcDh46z/r8A84C1J9gZ2qKqVvQOSzAfmAwwODtLpdDZwS5opRkZG3M+WcU/bxf1sF/ezv5jo96fnAOdX1TrgziRXAAcAXwH+MclhwEPA44E9mmtur6oVTXs5MGsT1v01cEnTXg2sraoHk6zexPnW+2JVVTPPnVW1GiDJTc28T6D7ZWJZ88B+e+Cq5tq5Sd4J7AQ8BriJ3yb6n2/+3tD9Xgj8fZITgVcDi0YPqKqFwEKAoaGhGh4e3oTbVD/qdDq4n+3inraL+9ku7md/MdGfWY4HBoHZTQK+BtihObe2Z9w6YKNKd0Z5sKqqaT+0fs6qeijJ5vysrI/tIX43zofo/gyuAy6rquN6L0qyA/BxYE5V/SDJAn57v73zrmOCn+WqeiDJZcARwP8BZm/6rUiSJM0M1uj3p6XAsUm2STIIHAZcC+wK/KRJ8ucCT5rOIIH7gF0mYZ6rgUOT7AWQZOf1JTbN+bubmv2jN2ONTwEfAa6rqp9vVrSSJEkzgIl+f7oIWAWsBC4H3llVPwbOA+Y0JTCvBG6ZvhCBbgnNkRO9jLsxquouujX05ydZRbdsZ5+q+gXwSeBG4FLgus1YYzlwL3DOps4hSZI0k1i600eqaqD5u4ATmz+95++m+1LrWPbrGTfhy7FVNa+nPTx6/aa9YJzYOkCnad8GPONhrLVmVJy95y6n+x7C6OtPpvui7uj+3rjvZgPvECR5HN0vtl+daJwkSVJb+ERfrZfklcA1wN9V1UPTHY8kSdJU8Il+iyX5GN1/yrLXh6tq0stXkvwdcMyo7gur6tTJXmuCGF4InD6q+/aqOhL4t6mKQ5IkqR+Y6LdYVb1hCtc6FZiypH6cGC6lW8svSZK01bN0R5IkSWohE31JkiSphUz0JUmSpBYy0ZckSZJayERfkiRJaiETfUmSJKmFTPQlSZKkFjLR38okmZfko5tx7eMexvjhJF9q2i9N8q5NWXcj13pcks827f2T/PGWWkuSJGkmMNFviSTbTMEy84CNTvR7VdWSqjptcwNIMuZ/8lZV/1VVRzeH+wMm+pIkaatmoj8DJJmV5JYk5yW5Oclnk+yUZE2S05N8GzgmyXFJVie5McnpPdefkOS2JNcCh/b0L0pydM/xSE/7pGaulUlOa8bNAc5LsiLJjuPE+qIm1m8Df9bTPy/JR5PsmuT/JXlE079zkh8k2S7Jk5NckmR5kqVJ9umJ86wk1wDvS/K8JoYVSW5IskvzGd2YZHvgvcCxzfljk3w3yWAz1yOSfG/9sSRJUluN+XRUfWkIeE1VLUtyNvDXTf9Pq+pZTUnN1cBs4OfAV5O8DLgG+Iem/x7gG8ANEy2U5MXAEcCBVfVAksdU1c+SvBF4R1VdP851OwCfBJ4PfA+4YPSYqronyQrgeU0sLwEuraoHkywEXl9V301yIPDxZi6AJwCHVNW6JF8E3tB8FgPAr3rm/3WSdwNzquqNTVz7AMcDHwIOB1ZW1V2jYp8PzAcYHByk0+lM9BFpBhkZGXE/W8Y9bRf3s13cz/5ioj9z/KCqljXtc4E3N+31yfQBQGd9ApvkPOCw5lxv/wXA3htY63DgnKp6AKCqfraRMe4D3F5V323WOpcmeR7lAuBYuon+nwMfbxL2Q4ALk6wf98ieay6sqnVNexnwgeYeP19VP+y5ZixnAxfTTfRfDZwzekBVLQQWAgwNDdXw8PAGb1YzQ6fTwf1sF/e0XdzPdnE/+4ulOzNHjXN8/2bM+Ruan4GmlGb7zZjr4VgCvCjJY+j+puHyJo5fVNX+PX+e2nPNf99nU+v/WmBHYNn6Ep/xVNUPgDuTPB94NvCVyb0dSZKk/mOiP3M8McnBTfsvgCtHnb8WeF6S3ZsXc48DrqBbuvO8JI9Nsh1wTM81a+gm2gAvBbZr2pcBJyTZCaBJyAHuA3aZIMZbgFlJntwcHzfWoKoaAa4DPgx8qarWVdW9wO1JjmnWTJJnjnV9kidX1eqqOr2ZZ3SiP1acn6L7m5De3wxIkiS1lon+zHEr8IYkNwO7AZ/oPVlVdwDvolsOsxJYXlUXN/0LgKvolrzc3HPZJ+l+CVgJHEzz1LyqLqH71P36pp7+Hc34RcBZ472MW1W/oluq8+XmZdyfTHA/FwAv53fr+I8HXtPEcxPd9wTG8pbmxdtVwIP8zyf03wD2Xf8ybtO3BBhgjLIdSZKkNrJGf+b4TVW9fFTfrN6DqjofOH/0hVV1DmPXpd8JHNTTdVLPudOA00aN/xzwuYmCbL4k/I9SmqpaRPeLwvrjzwIZNeZ24EVjXDtv1PGbxlh6DbBfc/5ndN9Z6PVMui/h3jJR/JIkSW1hoq/Wa/6jrr+i+xsDSZKkrYKJ/gxQVWtonlb3iyQXAb8/qvukqrp0OuKZyFi/nZAkSWo7E31tkqo6crpjkCRJ0vh8GVeSJElqIRN9SZIkqYVM9CVJkqQWMtGXJEmSWshEX5IkSWohE31JkiSphUz0tVVIMivJX0x3HJIkSVPFRF9bi1mAib4kSdpqmOhr0iTZOcmXk6xMcmOSY5PMTnJFkuVJLk2yZzP2gCSrkqxI8v4kNzb985J8IcllSdYkeWOStyW5IcnVSR7TjHtykkuaeZcm2afpX5TkI0m+leT7SY5uwjsNeG6z3lun4/ORJEmaSib6mkwvAv6rqp5ZVfsBlwBnAkdX1WzgbODUZuw5wF9W1f7AulHz7Af8GXBAM/6BqvpD4Crglc2YhcCbmnnfAXy85/o9gecAL6Gb4AO8C1haVftX1Qcn64YlSZL61bbTHYBaZTXwz0lOB74E/Jxu0n5ZEoBtgDuSPBrYpaquaq77d7pJ+XrfqKr7gPuS3AN8sWf+ZyQZAA4BLmzmBXhkz/VfqKqHgO8k2WNjAk8yH5gPMDg4SKfT2fi7Vl8bGRlxP1vGPW0X97Nd3M/+YqKvSVNVtyV5FvDHwCnA5cBNVXVw77gm0Z/I2p72Qz3HD9H9mX0E8IvmtwEbuj7jjBkd+0K6vyVgaGiohoeHN+YyzQCdTgf3s13c03ZxP9vF/ewvlu5o0iR5HN0ym3OB9wMHAoNJDm7Ob5fkaVX1C7pP6w9sLv3zh7NOVd0L3J7kmGbeJHnmBi67D9jl4awjSZI0k5noazI9Hbg2yQrgPcC7gaOB05OsBFbQLbkBeA3wyWbszsA9D3Ot44HXNPPeBByxgfGrgHXNi8K+jCtJklrP0h1Nmqq6FLh0jFOHjdF3U1U9AyDJu4DrmzkWAYt65pzV0/7vc1V1O92Xf0fHMG/U8UDz94PA8zf2XiRJkmY6E31Nlz9J8n/p/gz+P2De9IYjSZLULib6mhZVdQFwwXTHIUmS1FbW6EuSJEktZKIvSZIktZCJviRJktRCJvqSJElSC5noS5IkSS1koi9JkiS1kIm+JEmS1EIm+pIkSVILmehLkiRJLWSiry0mybwkH92Max83yfH87WTOJ0mS1M9M9PWwJdlmCpaZB0xqog+Y6EuSpK2Gib5+R5JZSW5Jcl6Sm5N8NslOSdYkOT3Jt4FjkhyXZHWSG5Oc3nP9CUluS3ItcGhP/6IkR/ccj/S0T2rmWpnktGbcHOC8JCuS7DhOrAck+VZz3bVJdml+E/D5JJck+W6S9zVjTwN2bOY7b9I/OEmSpD6z7XQHoL40BLymqpYlORv466b/p1X1rKak5mpgNvBz4KtJXgZcA/xD038P8A3ghokWSvJi4AjgwKp6IMljqupnSd4IvKOqrh/nuu2BC4Bjq+q6JI8Cftmc3h/4Q2AtcGuSM6vqXUneWFX7jzPffGA+wODgIJ1OZ4MfkmaGkZER97Nl3NN2cT/bxf3sLyb6GssPqmpZ0z4XeHPTvqD5+wCgU1V3ATRPyA9rzvX2XwDsvYG1DgfOqaoHAKrqZxsZ4xBwR1Vd11x3b7MmwNer6p7m+DvAk4AfTDRZVS0EFgIMDQ3V8PDwRoahftfpdHA/28U9bRf3s13cz/5i6Y7GUuMc378Zc/6G5uctySOA7Tdjrg1Z29Neh19oJUnSVshEX2N5YpKDm/ZfAFeOOn8t8Lwkuzcv5h4HXEG3dOd5SR6bZDvgmJ5r1tAt6QF4KbBd074MOCHJTgBJHtP03wfsMkGMtwJ7JjniuRNfAAAgAElEQVSguW6XJBtK6B9s4pIkSWo9E32N5VbgDUluBnYDPtF7sqruAN5FtwZ/JbC8qi5u+hcAVwHLgJt7Lvsk3S8BK4GDaX47UFWXAEuA65OsAN7RjF8EnDXey7hV9WvgWODMZs7LgB02cF8LgVW+jCtJkrYGljRoLL+pqpeP6pvVe1BV5wPnj76wqs4Bzhmj/07goJ6uk3rOnQacNmr854DPTRRkU59/0KjuRc2f9WNe0tM+qXddSZKkNvOJviRJktRCPtHX76iqNcB+0x1HryQXAb8/qvukqrp0OuKRJEmaCUz01feq6sjpjkGSJGmmsXRHkiRJaiETfUmSJKmFTPQlSZKkFjLRlyRJklrIRF+SJElqIRN9SZIkqYVM9CVJkqQWMtHvI0lGpmidRUmObtqfSrLvw7h2TpKPNO3hJIdsqTglSZK06fwPs7ZyVfXahzn+euD65nAYGAG+NclhSZIkaTP5RL8Ppev9SW5MsjrJsU3/QJKvJ/l2039E0z8ryc1JPpnkpiRfTbLjRq7VSTKnaY80696U5GtJnt2c/36SlzZjhpN8Kcks4PXAW5OsSPLcceZflOQTSa5u5hlOcnYT76KecS9IclVzbxcmGWj6353kuuazWJgkPXGfnuTaJLeNt34z9q1Jzm7aT2/m2mljPh9JkqSZyif6/enPgP2BZwK7A9cl+SZwF3BkVd2bZHfg6iRLmmueAhxXVa9Lshg4Cjj3Ya67M3B5VZ2Y5CLgFOCPgH2BTwPr16Kq1iQ5CxipqjM2MO9uwMHAS5s5DgVe29zX/sAPgZOBw6vq/iQnAW8D3gt8tKreC5DkM8BLgC82825bVc9O8sfAe4DDx1n/w0AnyZHA3wF/WVUP9A5IMh+YDzA4OEin09nALWmmGBkZcT9bxj1tF/ezXdzP/mKi35+eA5xfVeuAO5NcARwAfAX4xySHAQ8Bjwf2aK65vapWNO3lwKxNWPfXwCVNezWwtqoeTLJ6E+db74tVVc08d1bVaoAkNzXzPoHul4llzQP77YGrmmvnJnknsBPwGOAmfpvof775e8L7raqHkswDVgH/UlXLxhizEFgIMDQ0VMPDw5t4q+o3nU4H97Nd3NN2cT/bxf3sLyb6M8vxwCAwu0nA1wA7NOfW9oxbB2xU6c4oD1ZVNe2H1s/ZJMqb87OyPraH+N04H6L7M7gOuKyqjuu9KMkOwMeBOVX1gyQL+O399s67jg3/LD+F7vsEj9uUG5AkSZpprNHvT0uBY5Nsk2QQOAy4FtgV+EmT5M8FnjSdQQL3AbtMwjxXA4cm2Qsgyc5J9ua3Sf3dTc3+0ZsyeZJdgY/Q/Rwfu/5fHJIkSWozE/3+dBHdMpOVwOXAO6vqx8B5wJymBOaVwC3TFyLQLaE5cqKXcTdGVd0FzAPOT7KKbtnOPlX1C+CTwI3ApcB1m7jEB4GPVdVtwGuA05L8r02NV5IkaSbIbys1JEG3Rv/WW2+d7jA0SawXbR/3tF3cz3ZxP6dekuVVNWescz7RlyRJklrIl3FbLMnH6P5Tlr0+XFXnbIG1/g44ZlT3hVV16mSvNUEMLwROH9V9e1UdOVUxSJIk9QsT/RarqjdM4VqnAlOW1I8Tw6V0a/klSZK2epbuSJIkSS1koi9JkiS1kIm+JEmS1EIm+pIkSVILmehLkiRJLWSiL0mSJLWQib4kSZLUQib6kiRJUguZ6PehJCPTHUOvJMNJDpnuOCRJkrTxTPQFQJKJ/pfkYcBEX5IkaQZJVU13DBolyUhVDSQJ8D7gxUABp1TVBUkGgIuB3YDtgJOr6uIks4CvAFfSTcx/BBxRVb8cZ50OsAJ4DnA+cBtwMrA98FPgeGBH4GpgHXAX8CbgFuAs4InNVG+pqmXjrLEA+H3gD5rxbwUOau7pR8CfVtWDSWYDHwAGgLuBeVV1R5LXAfObmL4HvKKqHkiyCLgXmAP8HvDOqvrsODEcCbwROLwZewVwWFX9uGfM/GYdBgcHZy9evHisqTQDjYyMMDAwMN1haBK5p+3ifraL+zn15s6du7yq5ox1zkS/D/Uk+kcBrwdeBOwOXAccSDfh3qmq7k2yO91E/CnAk+gmw3OqakWSxcCSqjp3nHU6wHeq6q+b492AX1RVJXkt8NSqenuTrI9U1RnNuH8HPl5VVyZ5InBpVT11nDUW0E2w5wL7AlcBR1XVV5JcBHwa+DLd5PuIqrorybHAC6vq1UkeW1U/beY6Bbizqs5sEv2dgWOBfZr73GuCz/Tc5nN6EXBeVZ0/3tihoaG69dZbxzutGabT6TA8PDzdYWgSuaft4n62i/s59ZKMm+hPVK6h6fcc4PyqWgfcmeQK4AC6T+3/MclhwEPA44E9mmtur6oVTXs5MGsDa1zQ034CcEGSPek+Qb99nGsOB/bt/sIBgEclGaiq8d4t+Erz1H41sA1wSdO/uolvCNgPuKyZcxvgjmbMfk2C/2i6T/sv7Zn3C1X1EPCdJHswsTcBNwJXT5TkS5IktYWJ/sx0PDAIzG4S6DXADs25tT3j1tEtvZnI/T3tM4EPVNWSJMPAgnGueQRwUFX9aiPjXQtQVQ8lebB++2ukh+j+DAa4qaoOHuPaRcDLqmplknl03xf4nXkbYWJPaNbbI8kjmi8IkiRJreXLuP1tKXBskm2SDAKHAdcCuwI/aZL8uXRLdibDrnTr5gFe1dN/H7BLz/FX6T4hByDJ/pu57q3AYJKDm/m2S/K05twuwB1JtqP7Bedha140Phs4DrgZeNtmxitJktT3TPT720XAKmAlcDndF05/DJwHzGlKYV5J9+XYybAAuDDJcrovxK73ReDIJCuSPBd4c7P+qiTfofsewSarql8DRwOnJ1lJ9wXh9f/Kz98D1wDL2PT7/FtgaVVdSTfJf22SMd8pkCRJagtLd/pQVQ00fxdwYvOn9/zdwFhlLtCtdV8/7owNrDM86vhiuv+az+hxtwHPGNV97ERz91y7YNTxwFjnmvcKDhvj+k8Anxijf954844x9r097fvovrwrSZLUaj7RlyRJklrIJ/pbgSQfAw4d1f3hqjpnEtc4AfibUd3LquoNk7XGRsTwdOAzo7rXVtWBUxWDJElSvzDR3wpMRbLdfGmYtC8OmxjDamBzXwyWJElqBUt3JEmSpBYy0ZckSZJayERfkiRJaiETfUmSJKmFTPQlSZKkFjLRlyRJklrIRF+SJElqIRP9GSrJyHTH0CvJcJJDtvAan0qy75ZcQ5IkqS38D7O00ZJsW1W/Gef0MDACfGtLrV9Vr91Sc0uSJLVNqmq6Y9AmSDJS9f+zd+/hclZ1mve/NwflsCFAZzdD0wOxeWFDpCFKQEBIV2x0hFYRgY4amw4KkRHloIDzNkyb7gEbhbeVRg4TuEhQkCFBkcAMJ4ECDCIhkAPhaJs4HlokIIctEjC53z9q7bao3sdkJ3vXk/tzXXXVetazDr+qtf/41VPrqe0OSQK+ChwOGDjX9vWSOoCbgO2BzYFzbN8kaRxwK/AD4GDgF8CRtn/Xxzx1YBFwCHAd8DRwDvAW4HlgKrAl8CCwGngO+BzwJHA5sEsZ6jTb8/uYYwbwNuDPSvvTgQPLa/oF8EHbb5RYzrD9cPlG4yLgA8Dvymt4to/xbwK+Y/ubkj4NTLI9taXNdGA6QGdn535z5szpbahoQ93d3XR0dIx0GDGMsqbVkvWslqznhjd58uSFtif2di5X9NvfR4AJwL7AWGCBpPtoJNxH2X5Z0ljgQUnzSp/dgY/ZPlHSHOBo4Jp+5nhLzx+QpO2BA21b0gnAWba/IOlyoNv2haXdt4Gv2f6BpF2A24G9+pljN2AyMB74IXC07bMk3Qj8FfC9lvZbAw/aPlvSV4ETgXP7GHs6MF/ScuALND5EvIntmcBMgK6uLtdqtX5CjXZSr9fJelZL1rRasp7VkvUcXZLot79DgOtsrwaelXQvsD+Nq/ZfljQJWAPsDOxY+iy3vaiUFwLjBpjj+qbynwLXS9qJxlX95X30OQwY3/jCAYBtJXXY7uveglvLVfulwKbAbaV+aR/xvQ7c0vQa3ttX8LaflfT3wD00Pvy80FfbiIiIiKpIol9dU4FOYL+SQK8AtijnVjW1W01j601/fttUvhj4Z9vzJNWAGX302YTGlf/XBhnvKgDbayS94T/sKVtD73+nzW1W99Gm2Z/T2Gr0J4OMJyIiIqKt5Vd32t/9wBRJm0rqBCYBDwFjgF+XJH8ysOswzTeGxr55gL9tqn8F2Kbp+A4ae/UBkDRhmOYfMkkH0Njv/w7gDElvG6lYIiIiIjaUJPrt70ZgCbAYuJvGnvlfAdcCE8tWmONo3Bw7HGYAcyUtBFY21d8MHCVpkaRDgVPK/EskPQ6cNEzzD4mktwJXAJ+0/Usae/SvUtOeooiIiIgqytadNmW7ozwbOLM8ms+vBA7qo/veTe0uHGCeWsvxTTR+zae13dPAPi3VU/obu6nvjJbjjt7ONcfS0uYG4IY+xl5F40blnuN5wLze2kZERERUSa7oR0RERERUUK7oBwCSLgHe3VJ9ke1ZwzjH8cCpLdXzbZ88TOOfDRzbUj3X9nnDMX5EREREO0miHwAMV7I9wByzgGH74NDL+OcBSeojIiIiyNadiIiIiIhKSqIfEREREVFBSfQjIiIiIiooiX5ERERERAUl0Y+IiIiIqKAk+hERERERFZREPyIiIiKigpLotwFJ3RtontmSjinlKyWNH0LfiZL+pZRrkg5eX3FGRERExMDyD7OiV7ZPGGL7h4GHy2EN6AYeGOawIiIiImKQckW/jajhAkmPSVoqaUqp75B0l6RHSv2RpX6cpCckXSFpmaQ7JG05yLnqkiaWcneZd5mk70s6oJz/iaQPlTY1SbdIGgecBJwuaZGkQ/sYf7akyyQ9WMapSbqqxDu7qd1lkh4uc/9DqRsj6SlJXeX4Okkn9jHPrpKekTRW0iaS7pf0vkG94RERERFtLFf028tHgAnAvsBYYIGk+4DngKNsvyxpLPCgpHmlz+7Ax2yfKGkOcDRwzRDn3Rq42/aZkm4EzgXeC4wHrgZ65sL2CkmXA922Lxxg3O2Bg4APlTHeDZxQXtcE24uAs22/IGlT4C5J+9heIumzwGxJFwHb276itwls/1TSV4DLgIeAx23f0dpO0nRgOkBnZyf1en2Qb02Mdt3d3VnPismaVkvWs1qynqNLEv32cghwne3VwLOS7gX2B24FvixpErAG2BnYsfRZXhJmgIXAuLWY93XgtlJeCqyy/YakpWs5Xo+bbbuM86ztpQCSlpVxFwF/XZLwzYCdaHy4WGL7TknHApfQ+ODTJ9tXlrYn0fig1FubmcBMgK6uLtdqtXV4WTGa1Ot1sp7VkjWtlqxntWQ9R5ck+tUwFegE9isJ+Apgi3JuVVO71cCgtu60eMO2S3lNz5i210hal7+hntjW8OY41wCbSXobcAawv+3flC09WwBI2gTYC3iVxjcDP+9rEklbAX9aDjuAV9Yh5oiIiIi2kD367eV+YIqkTSV1ApNobEcZA/y6JPmTgV1HMkgaifQ2wzDOtsBvgZck7Qgc3nTudOAJ4OPALEmb9zPOV4Brgb8Het3iExEREVE1SfTby43AEmAxcDdwlu1f0UhiJ5YtMMcBT45ciADcDBzV3824g2F7MfAojdfzbWA+QLkJ9wTgC7bvB+4DzultDEl/QWN701dsXwu8Lun4tY0pIiIiol1k604bsN1Rng2cWR7N51fSuKm1N3s3tev35ljb05rKtdb5S3lGH7HVgXopPw3sM4S5VrTEOa23cou9mtp8vp957gUObDr+SH9xRURERFRFruhHRERERFRQruhvhCRdQuOnLJtdZHvWepjrbODYluq5ts9bD3P9CHhrS/Xf9PyaT0RERMTGJIn+Rsj2yRtwrvOAYU/q+5jrXRtinoiIiIh2kK07EREREREVlEQ/IiIiIqKCkuhHRERERFRQEv2IiIiIiApKoh8RERERUUFJ9CMiIiIiKiiJfkREREREBSXRj0GRNE7SY8MwzjRJ3xhk2xmSzhigzWxJxwxh/mF5HRERERGjXRL9iIiIiIgKSqIfQ7GZpGslPSHpBklbSfp7SQskPSZppiQBSKpL+oqkhyQ9LenQ1sEk/ZWkH0oaO9DEkk4s8yyW9B1JWzWdPkzSw2WeD5T2m0q6oPRZIunTw/YuRERERLSBzUY6gGgrXcCnbM+XdBXwGeAbtv8RQNK3gA8AN5f2m9k+QNIRwJeAw3oGknQU8HngCNu/GcTc37V9Rel7LvAp4OJybhxwALAbcI+k/wc4DnjJ9v6S3grMl3QH4N4GlzQdmA7Q2dlJvV4fzPsRbaC7uzvrWTFZ02rJelZL1nN0SaIfQ/Ez2/NL+RrgFGC5pLOArYAdgGX8IdH/bnleSCMZ7/EeYCLwPtsvD3LuvUuCvx3QAdzedG6O7TXAM5J+AuwJvA/Yp2n//hhgd+Dp3ga3PROYCdDV1eVarTbIsGK0q9frZD2rJWtaLVnPasl6ji5J9GMoWq+GG7gUmGj7Z5JmAFs0nV9Vnlfz5r+1fwX+DNgDeHiQc88GPmx7saRpQG2AuAR8znbzBwIkjRvkfBERERFtLXv0Yyh2kXRQKX8c+EEpr5TUAQz2129+ChwNfFPS2wfZZxvg3yRtDkxtOXespE0k7UbjA8RTNK74/9fSHkl7SNp6kHNFREREtL1c0Y+heAo4uezPfxy4DNgeeAz4FbBgsAPZflLSVGCupA/a/tcBuvx34EfAc+V5m6Zz/xd4CNgWOMn2a5KupLFd6JFyg/BzwIcHG19EREREu0uiH4NiewWNve+tzimP1va1pvJKyh5927NpbMPB9qPA+H7mnNFUvozGB4vWNtP66LsG+LvyaPYSsHdfc0ZERERURbbuRERERERUUK7ox4iTdDZwbEv1XNvnjUQ8EREREVWQRD9GXEnok9RHREREDKNs3YmIiIiIqKAk+hERERERFZREPyIiIiKigpLoR0RERERUUBL9iIiIiIgKSqIfEREREVFBSfQjIiIiIiooiX6bkdS9geaZLemYUr5S0vgh9J0o6V9KuSbp4GGK6f9I2m44xoqIiIiouvzDrBiQ7ROG2P5h4OFyWAO6gQeGIY4j1nWMiIiIiI1Frui3KTVcIOkxSUslTSn1HZLukvRIqT+y1I+T9ISkKyQtk3SHpC0HOVdd0sRS7i7zLpP0fUkHlPM/kfSh0qYm6RZJ44CTgNMlLZJ0aB/jz5Z0maQHyzg1SVeVeGc3tVshaexQXouk3SQ90nS8e/NxRERERFXlin77+ggwAdgXGAsskHQf8BxwlO2XJY0FHpQ0r/TZHfiY7RMlzQGOBq4Z4rxbA3fbPlPSjcC5wHuB8cDVQM9c2F4h6XKg2/aFA4y7PXAQ8KEyxruBE8rrmmB7UUv7Qb0W2/8q6aWmMY4HZrW2kzQdmA7Q2dlJvV4f6H2INtHd3Z31rJisabVkPasl6zm6JNFvX4cA19leDTwr6V5gf+BW4MuSJgFrgJ2BHUuf5U0J80Jg3FrM+zpwWykvBVbZfkPS0rUcr8fNtl3Gedb2UgBJy8q4rYn+UF7LlcDxkj4PTAEOaG1geyYwE6Crq8u1Wm3tX0mMKvV6naxntWRNqyXrWS1Zz9ElW3eqZyrQCexnewLwLLBFObeqqd1q1u6D3hu2Xcpresa0vWYtx+vRE9sa3hxnX+MO5bV8Bzgc+ACw0Pbz6xBnRERERFtIot++7gemSNpUUicwCXgIGAP8ulxlnwzsOpJBAq8A24xkALZfA24HLqOXbTsRERERVZREv33dCCwBFgN3A2fZ/hVwLTCxbIE5Dnhy5EIE4GbgqP5uxt1ArqXx7cAdIxhDRERExAaTPfptxnZHeTZwZnk0n19J46bW3uzd1K7fm2NtT2sq11rnL+UZfcRWB+ql/DSwzxDmWtESZ/O5caW4ciivpTgEmFXuaYiIiIiovCT6UXnl14F2A94z0rFEREREbChJ9Ddyki6h8VOWzS6yPex72SWdDRzbUj3X9nnDNH5fr+Wo4Rg/IiIiop0k0d/I2T55A851HjAsSX0f42+w1xIREREx2uVm3IiIiIiICkqiHxERERFRQUn0IyIiIiIqKIl+REREREQFJdGPiIiIiKigJPoRERERERWURD8iIiIiooKS6LcpSd0jHUMzSTVJB6/nOa6UNH59zhERERFRFfmHWTFokjaz/fs+TteAbuCB9TW/7RPW19gRERERVZMr+m1ODRdIekzSUklTSn2HpLskPVLqjyz14yQ9IekKScsk3SFpy37Gr0v6uqSHgVMlfVDSjyQ9Kun7knaUNA44CThd0iJJh0rqlPQdSQvK4939zDFD0tWS7pf0U0kfkfTVEvdtkjZvimViKXdLOk/SYkkPStqxj7G3kbS8aYxtm48jIiIiqipX9NvfR4AJwL7AWGCBpPuA54CjbL8saSzwoKR5pc/uwMdsnyhpDnA0cE0/c7zFdk+CvT1woG1LOgE4y/YXJF0OdNu+sLT7NvA12z+QtAtwO7BXP3PsBkwGxgM/BI62fZakG4G/Ar7X0n5r4EHbZ0v6KnAicG7roLZfkVRvGuOjwHdtv9HcTtJ0YDpAZ2cn9Xq9n1CjnXR3d2c9KyZrWi1Zz2rJeo4uSfTb3yHAdbZXA89KuhfYH7gV+LKkScAaYGeg56r3ctuLSnkhMG6AOa5vKv8pcL2knYC3AMv76HMYMF5Sz/G2kjps93Vvwa2235C0FNgUuK3UL+0jvteBW5pew3v7if9K4Cwaif7xND4UvIntmcBMgK6uLtdqtX6Gi3ZSr9fJelZL1rRasp7VkvUcXZLoV9dUoBPYryTQK4AtyrlVTe1WA31u3Sl+21S+GPhn2/Mk1YAZffTZhMaV/9cGGe8qANtrJL1h26V+Db3/nTa3Wd1HG8qY88uWpRqwqe3HBhlTRERERNvKHv32dz8wRdKmkjqBScBDwBjg1yXJnwzsOkzzjQF+Ucp/21T/CrBN0/EdwOd6DiRNGKb519Y3gW8Ds0Y4joiIiIgNIol++7sRWAIsBu6msWf+V8C1wMSyFeY44Mlhmm8GMFfSQmBlU/3NwFE9N+MCp5T5l0h6nMbNuiPpWmB74LoRjiMiIiJig8jWnTZlu6M8GzizPJrPrwQO6qP73k3tLhxgnlrL8U3ATb20exrYp6V6Sn9jN/Wd0XLc0du55lha2twA3DDANIcAN9h+cTAxRURERLS7JPpReZIuBg4HjhjpWCIiIiI2lCT6AYCkS4DW37q/yPaw7WmXdDxwakv1fNsnD9P4ZwPHtlTPtf253tpHREREVFkS/QBguJLtAeaYxXq8Gdb2ecB562v8iIiIiHaSm3EjIiIiIiooiX5ERERERAUl0Y+IiIiIqKAk+hERERERFZREPyIiIiKigpLoR0RERERUUBL9iIiIiIgKSqLfBiR1b6B5Zks6ppSvlDR+CH0nSvqXUq5JOnh9xRkRERERA8s/zIpe2T5hiO0fBh4uhzWgG3hgmMOKiIiIiEHKFf02ooYLJD0maamkKaW+Q9Jdkh4p9UeW+nGSnpB0haRlku6QtOUg56pLmljK3WXeZZK+L+mAcv4nkj5U2tQk3SJpHHAScLqkRZIO7WP82ZIuk/RgGacm6aoS7+ymdpdJerjM/Q+lboykpyR1lePrJJ3YxzyflPT1puMTJX1tMO9BRERERDvLFf328hFgArAvMBZYIOk+4DngKNsvSxoLPChpXumzO/Ax2ydKmgMcDVwzxHm3Bu62faakG4FzgfcC44GrgZ65sL1C0uVAt+0LBxh3e+Ag4ENljHcDJ5TXNcH2IuBs2y9I2hS4S9I+tpdI+iwwW9JFwPa2r+hjjjnA2ZLOtP0GcDzw6dZGkqYD0wE6Ozup1+uDeV+iDXR3d2c9KyZrWi1Zz2rJeo4uSfTbyyHAdbZXA89KuhfYH7gV+LKkScAaYGdgx9JneUmYARYC49Zi3teB20p5KbDK9huSlq7leD1utu0yzrO2lwJIWlbGXQT8dUnCNwN2ovHhYontOyUdC1xC44NPr2x3S7ob+ICkJ4DNe+ZpaTcTmAnQ1dXlWq22Di8rRpN6vU7Ws1qyptWS9ayWrOfokkS/GqYCncB+JQFfAWxRzq1qarcaGNTWnRZv2HYpr+kZ0/YaSevyN9QT2xreHOcaYDNJbwPOAPa3/ZuypWcLAEmbAHsBr9L4ZuDn/cxzJfB3wJPArHWINyIiIqJtZI9+e7kfmCJpU0mdwCTgIWAM8OuS5E8Gdh3JIIFXgG2GYZxtgd8CL0naETi86dzpwBPAx4FZkjbvaxDbPwL+c2l73TDEFRERETHqJdFvLzcCS4DFwN3AWbZ/BVwLTCxbYI6jceV6JN0MHNXfzbiDYXsx8CiN1/NtYD5AuQn3BOALtu8H7gPOGWC4OcB8279Z23giIiIi2km27rQB2x3l2cCZ5dF8fiWNm1p7s3dTu35vjrU9ralca52/lGf0EVsdqJfy08A+Q5hrRUuc03ort9irqc3n+5urOATIr+1ERETERiNX9KPSJG0n6Wngd7bvGul4IiIiIjaUXNHfCEm6hMZPWTa7yPaw36gq6Wzg2JbqubbPWw9z/Qh4a0v139jeY7jnioiIiBjtkuhvhGyfvAHnOg8Y9qS+j7netSHmiYiIiGgH2boTEREREVFBSfQjIiIiIiooiX5ERERERAUl0Y+IiIiIqKAk+hERERERFZREPyIiIiKigpLoR0RERERUUBL9iIiIiIgKSqLfRiR1j3QMzSTVJB080nFERERExH+URD/6Jam//55cA5LoR0RERIxCsj3SMcQgSeq23SFJwFeBwwED59q+XlIHcBOwPbA5cI7tmySNA24FfkAjMf8FcKTt3/UxTx1YBBwCXAc8DZwDvAV4HpgKbAk8CKwGngM+BzwJXA7sUpWIztMAACAASURBVIY6zfb8PuaYAbwN+LPS/nTgwPKafgF80PYbkv4e+GCZ7wHg08CmwA+BM23XJf0TsMb22b3M8x7gFNsfLsfvBT5j+6iWdtOB6QCdnZ37zZkzp7ewow11d3fT0dEx0mHEMMqaVkvWs1qynhve5MmTF9qe2Nu5JPptpCnRPxo4CXg/MBZYALyLRsK9le2XJY2lkYjvDuwK/BiYaHuRpDnAPNvX9DFPHXjc9mfK8fbAi7Yt6QRgL9tfKMl6t+0LS7tvA5fa/oGkXYDbbe/VxxwzgMOAycB4Gon70bZvlXQjcLXt70nawfYLpc+3gDm2b5b0duAGGh8wLgDeZfv1XuYR8ARwqO3nSozX2b65r/e5q6vLTz31VF+no83U63VqtdpIhxHDKGtaLVnPasl6bniS+kz0+9uWEaPXITSS1dXAs5LuBfancdX+y5ImAWuAnYEdS5/ltheV8kJg3ABzXN9U/lPgekk70biqv7yPPocB4xu5NQDbSuqw3de9BbeWq/ZLaVylv63UL22Kb7Kks4CtgB2AZcDNtpeVxP8W4KDeknyA8uHkW8AnJM0CDgKO6++FR0RERFRBEv1qmQp0AvuVBHoFsEU5t6qp3WoaW2H689um8sXAP9ueJ6kGzOijzybAgbZfG2S8qwBsr5H0hv/w9dIaYDNJWwCX0vgm4mflW4Atmvr/OfAi8McDzDMLuBl4DZhr+/eDjC8iIiKibeVm3PZ0PzBF0qaSOoFJwEPAGODXJcmfTGPLznAYQ2PfPMDfNtW/AmzTdHwHja00AEiasI7z9iT1K8v9B8c0jf0RGlf4JwEXS9qur0Fs/xL4JY37DGatY0wRERERbSGJfnu6EVgCLAbuBs6y/SvgWmBi2QpzHI2bY4fDDGCupIXAyqb6m4GjJC2SdChwSpl/iaTHadxHsNZsvwhcATwG3E7jXgTK/QfnAyfYfhr4BnDRAMNdC/zM9hPrElNEREREu8jWnTZiu6M8GzizPJrPr6SxB703eze1u3CAeWotxzfR+DWf1nZPA/u0VE/pb+ymvjNajjt6O2f7HBpX4lvt0dTmXwYx5SE0PjREREREbBSS6EfllW8ifgt8YaRjiYiIiNhQkuhvxCRdAry7pfoi28O2j13S8cCpLdXzbZ88XHM0zXUjjd/mb/ZF2/sN91wRERERo10S/Y3Y+ki2e5ljFhvoBtjWf4IVERERsTHLzbgRERERERWURD8iIiIiooKS6EdEREREVFAS/YiIiIiICkqiHxERERFRQUn0IyIiIiIqKIl+rHeSZkg6YxjG2U7SZ5qO/0TSDes6bkREREQVJdGPUUVSf//bYTvg3xN927+0fcz6jyoiIiKi/STRj/VC0tmSnpb0A6Cr1NUlTSzlsZJWlPI0SfMk3Q3cJalD0l2SHpG0VNKRZdjzgd0kLZJ0gaRxkh4rY2whaVZp/6ikyU1jf1fSbZKekfTVDfxWRERERIyI/GfcGHaS9gM+Ckyg8Tf2CLBwgG7vBPax/UK5qn+U7ZcljQUelDQP+G/A3rYnlHnGNfU/GbDtP5e0J3CHpD3KuQnAO4BVwFOSLrb9s5aYpwPTATo7O6nX62v34mPU6e7uznpWTNa0WrKe1ZL1HF2S6Mf6cChwo+1XAUqSPpA7bb9QygK+LGkSsAbYGdhxgP6HABcD2H5S0k+BnkT/LtsvlVgeB3YF3pTo254JzATo6upyrVYbRMjRDur1OlnPasmaVkvWs1qynqNLEv3YkH7PH7aLbdFy7rdN5alAJ7Cf7TfKFp/W9kOxqqm8mvzdR0RExEYge/RjfbgP+LCkLSVtA3yw1K8A9ivl/m6iHQP8uiT5k2lcgQd4Bdimjz730/iAQNmyswvw1Fq/goiIiIg2l0Q/hp3tR4DrgcXArcCCcupC4L9KehQY288Q1wITJS0FjgOeLOM+D8yX9JikC1r6XApsUvpcD0yzvYqIiIiIjVS2MMR6Yfs84LxeTu3TVD6ntJ0NzG7quxI4qI9xP95StXepfw04vpf2rWN/YODoIyIiItpfruhHRERERFRQEv2IiIiIiApKoh8RERERUUFJ9CMiIiIiKiiJfkREREREBSXRj4iIiIiooCT6EREREREVlEQ/IiIiIqKCkuhHRERERFRQEv2IiIiIiApKoh8RERERUUFJ9GOtSereQPPMlnRMKV8pafwQ+k6U9C+lXJN08PqKMyIiImI02WykA4gYCtsnDLH9w8DD5bAGdAMPDHNYEREREaNOrujHOlPDBZIek7RU0pRS3yHpLkmPlPojS/04SU9IukLSMkl3SNpykHPVJU0s5e4y7zJJ35d0QDn/E0kfKm1qkm6RNA44CThd0iJJh66P9yIiIiJitMgV/RgOHwEmAPsCY4EFku4DngOOsv2ypLHAg5LmlT67Ax+zfaKkOcDRwDVDnHdr4G7bZ0q6ETgXeC8wHrga6JkL2yskXQ50276wdSBJ04HpAJ2dndTr9SGGEqNVd3d31rNisqbVkvWslqzn6JJEP4bDIcB1tlcDz0q6F9gfuBX4sqRJwBpgZ2DH0me57UWlvBAYtxbzvg7cVspLgVW235C0dKjj2Z4JzATo6upyrVZbi3BiNKrX62Q9qyVrWi1Zz2rJeo4uSfRjfZoKdAL7lQR8BbBFObeqqd1qYFBbd1q8YdulvKZnTNtrJOVvOyIiIjZq2aMfw+F+YIqkTSV1ApOAh4AxwK9Lkj8Z2HUkgwReAbYZ4RgiIiIiNogk+jEcbgSWAIuBu4GzbP8KuBaYWLbSHAc8OXIhAnAzcFRuxo2IiIiNQbY3xFqz3VGeDZxZHs3nVwIH9dF976Z2/+Hm2JZxpjWVa63zl/KMPmKrA/VSfhrYp7+5IiIiIqoiV/QjIiIiIiooV/Rj1JB0CfDuluqLbM8aiXgiIiIi2lkS/Rg1bJ880jFEREREVEW27kREREREVFAS/YiIiIiICkqiHxERERFRQUn0IyIiIiIqKIl+REREREQFJdGPiIiIiKigJPoRERERERWURL/CJI2T9NgwjDNN0jfWsm93ef4TSTesaywDzPV/JG1XHp9Zn3NFREREjHZJ9GODsP1L28es6ziS+vwnb7aPsP0isB2QRD8iIiI2akn0q28zSddKekLSDZK2kvT3khZIekzSTEkCkFSX9BVJD0l6WtKhrYNJ+itJP5Q0trfJJL2tnF8q6dym+n//dkHSg5Le3nSuLmmipK0lXVXmf1TSkeX8NEnzJN0N3CVpJ0n3SVpUXsOhpd2KEtf5wG7l/AWSvinpw03zXdszdkRERERV9Xl1NCqjC/iU7fmSrqJxpfsbtv8RQNK3gA8AN5f2m9k+QNIRwJeAw3oGknQU8HngCNu/6WO+i4DLbH9T0sl9tLke+GvgS5J2Anay/bCkLwN32/6kpO2AhyR9v/R5J7CP7RckfQG43fZ5kjYFtmoZ/78Be9ueUOL+C+B04HuSxgAHA3/b3EHSdGA6QGdnJ/V6vY/Qo910d3dnPSsma1otWc9qyXqOLkn0q+9ntueX8jXAKcBySWfRSJB3AJbxh0T/u+V5ITCuaZz3ABOB99l+uZ/53g0cXcrfAr7SS5s5wB00Pkj8NdCzd/99wIcknVGOtwB2KeU7bb9QyguAqyRtDnzP9qJ+4sH2vZIuldRZYvuO7d+3tJkJzATo6upyrVbrb8hoI/V6naxntWRNqyXrWS1Zz9ElW3eqz70cXwocY/vPgStoJNQ9VpXn1bz5g+C/AtsAe6zFnG8+af8CeF7SPsAUGlf4AQQcbXtCeexi+4ly7rdN/e8DJgG/AGZLOm4QMX0T+ARwPHDVINpHREREtLUk+tW3i6SDSvnjwA9KeaWkDmCwN8j+lMbV8G8276/vxXzgo6U8tZ921wNnAWNsLyl1twOfa7pn4B29dZS0K/Cs7SuAK2ls62n2Co0PJc1mA6cB2H68n7giIiIiKiGJfvU9BZws6Qlge+AyGlfxH6ORWC8Y7EC2n6SRvM+VtFsfzU4t8y0Fdu5nuBtofCCY01T3P4DNgSWSlpXj3tSAxZIepfGNwEUtcT4PzC836l5Q6p4FngBm9RNTRERERGVkj36F2V4B7NnLqXPKo7V9ram8krJH3/ZsGlfEsf0oML6fOZcDBzVVndMUy95N7Z6l5e/P9u+AT/cy5r/PX46vBq7upd24pvLHm89J2grYHbiur9gjIiIiqiRX9KPyJB1G42r+xbZfGul4IiIiIjaEXNGPtSLpbODYluq5ts8biXj6Y/v7wK4jHUdERETEhpREP9ZKSehHXVIfEREREQ3ZuhMRERERUUFJ9CMiIiIiKiiJfkREREREBSXRj4iIiIiooCT6EREREREVlEQ/IiIiIqKCkuhHRERERFRQEv0YFpLGSXqsl/rTJG21nud+YH2OHxEREdGOkujH+nYasF4SfUmbAdg+eH2MHxEREdHOkujHcNpM0rWSnpB0g6RTgD8B7pF0D4Ck90t6RNJiSXf1NZCkGZK+JemHkp6RdGKpr0m6X9I84PFS193U74uSlpbxzy91u0m6TdLC0nfP9fgeRERERIwKm410AFEpXcCnbM+XdBXwFuCXwGTbKyV1AlcAk2wvl7TDAOPtAxwIbA08Kul/l/p3AnvbXt7cWNLhwJHAu2y/2jT+TOAk289IehdwKfCelr7TgekAnZ2d1Ov1tXn9MQp1d3dnPSsma1otWc9qyXqOLkn0Yzj9zPb8Ur4GOKXl/IHAfT0Juu0XBhjvJtu/A35XvhE4AHgReKg1yS8OA2bZfrVnfEkdwMHAXEk97d7a2tH2TBofCOjq6nKtVhsgtGgX9XqdrGe1ZE2rJetZLVnP0SWJfgwnD3A8XOP9dghjbAK8aHvCOsYSERER0VayRz+G0y6SDirljwM/AF4Btil1DwKTJL0NYBBbd46UtIWkPwJqwIIB2t8JHN/zKz+SdrD9MrBc0rGlTpL2HeLrioiIiGg7SfRjOD0FnCzpCWB74DIa22Fuk3SP7edo7IP/rqTFwPUDjLcEuIfGB4T/YfuX/TW2fRswD3hY0iLgjHJqKvCpMucyGvv4IyIiIiotW3diWNheAfT2azYXl0dPu1uBWwc57BLbx7XMUwfqLXUdTeXzgfNbzi8H3j/IOSMiIiIqIVf0IyIiIiIqKFf0Y0RJOh44taV6vu2TRyKeiIiIiKpIoh8jyvYsYNZIxxERERFRNdm6ExERERFRQUn0IyIiIiIqKIl+REREREQFJdGPiIiIiKigJPoRERERERWURD8iIiIiooKS6EdEREREVFAS/VgvJHWX5z+RdEMpT5B0xMhGFhEREbFxSKIfgyZpyP9gzfYvbR9TDicAGzTRX5uYIyIiIqogiX68iaTjJC2RtFjStyTNlnS5pB8BX5W0m6TbJC2UdL+kPUu/t0n6oaSlks5tGm+cpMckvQX4R2CKpEWSpvQx/1+U84skPSppm1L/xTL2Yknnl7oJkh4s8d4oaftSX5f0dUkPA6dK2k/SvSXm2yXttH7fxYiIiIiRJ9sjHUOMEpLeDtwIHGx7paQdgH8GxgJH2l4t6S7gJNvPSHoX8E+23yNpHnCD7W9KOhn4iu0OSeOAW2zvLWkaMNH2Z/uJ4WbgfNvzJXUArwHvBf47cJjtVyXtYPsFSUuAz9m+V9I/AtvaPk1SHXjc9mckbQ7cW+J/rnzA+C+2P9ky73RgOkBnZ+d+c+bMGZb3NEZed3c3HR0dIx1GDKOsabVkPasl67nhTZ48eaHtib2dy7aGaPYeYK7tlQAlmabUrS6J98HA3FIP8Nby/G7g6FL+FvCVtYxhPvDPkq4Fvmv755IOA2bZfrUprjHAdrbvLf2uBuY2jXN9ee4C9gbuLDFvCvxb66S2ZwIzAbq6ulyr1dYy/Bht6vU6Wc9qyZpWS9azWrKeo0sS/RiM35bnTYAXbU/oo906fz1k+3xJ/5vGXv75kv7LWg7VE7OAZbYPWtfYIiIiItpJ9uhHs7uBYyX9EUDZuvPvbL8MLJd0bDkvSfuW0/OBj5by1D7GfwXYpr8AJO1me6ntrwALgD2BO4HjJW3VE5ftl4DfSDq0dP0bGlt0Wj0FdEo6qPTdvGxRioiIiKi0JPrx72wvA84D7pW0mMb+/FZTgU+V88uAI0v9qcDJkpYCO/cxxT3A+P5uxgVOKzfvLgHeAG61fRswD3hY0iLgjNL2b4ELStsJNG72bX1NrwPHAF8pMS+isf0oIiIiotKydSfexPbVNPa793V+OfD+Puqbt8ecU+pX0Ngjj+0XgP0HmP9zfdSfD5zfUrcIOLCXtrVe2k3qb96IiIiIqskV/YiIiIiICsoV/RgRko6nsd2n2XzbJ49EPBERERFVk0Q/RoTtWcCskY4jIiIioqqydSciIiIiooKS6EdEREREVFAS/YiIiIiICkqiHxERERFRQUn0IyIiIiIqKIl+REREREQFJdGPiIiIiKigJPoxJJLqkiauRb8H1kc8EREREdG7JPoxaJI2Xdu+tg8ezlgiIiIion9J9CtK0j9KOq3p+DxJp0o6U9ICSUsk/UPT+e9JWihpmaTpTfXdkv4/SYuBg5rqPynp603HJ0r6Wj/xdJfnWvlW4AZJT0q6VpLKuf0lPSBpsaSHJG0jaQtJsyQtlfSopMml7bQS852SVkj6rKTPlzYPStqhtNtN0m3ltd0vac9heYMjIiIiRjnZHukYYj2QNA74ru13StoEeAb4O+AvgU8DAuYBX7V9n6QdbL8gaUtgAfAXtp+XZGCK7Tll3DpwBvAksBjY0/YbZWvOp20v7SOebtsdkmrATcDbgV8C84EzgYfKmFNsL5C0LfAqcCrwdtufLEn6HcAewEeBc4B3AFsAPwa+aPvy8oHjp7a/Luku4CTbz0h6F/BPtt/TS3zTgekAnZ2d+82ZM2ct3vUYjbq7u+no6BjpMGIYZU2rJetZLVnPDW/y5MkLbfe6rXqzDR1MbBi2V0h6XtI7gB2BR4H9gfeVMkAHsDtwH3CKpKNK/X8u9c8Dq4Hv9DJ+t6S7gQ9IegLYvK8kvxcP2f45gKRFwDjgJeDfbC8o479czh8CXFzqnpT0UxqJPsA9tl8BXpH0EnBzqV8K7COpAzgYmFu+NAB4ax/v10xgJkBXV5drtdogX0qMdvV6naxntWRNqyXrWS1Zz9EliX61XQlMA/4TcBWNq/n/ZPt/NjcqV9kPAw6y/Wq5ar9FOf2a7dX9jP93NK7EzxpCXKuayqtZ+7/D5nHWNB2vKWNuArxoe8Jajh8RERHRtrJHv9puBN5P40r+7eXxyXKlG0k7S/pjYAzwm5Lk7wkcOJjBbf+IxtX/jwPXrWOsTwE7Sdq/xLaNpM2A+4GppW4PYJfSdjDxvQwsl3Rs6S9J+65jnBERERFtIVf0K8z265LuoXFVezVwh6S9gB+WrSzdwCeA24CTyhacp4AHhzDNHGCC7d8MQ6xTgIvLfQK/o/Etw6XAZZKWAr8Hptle1bQVZyBTS/9zgM2B/0Xj3oKIiIiISkuiX2HlJtwDgWN76mxfBFzUS/PDexvDdkfLca2lySFAn7+20zqO7TpQb6r/bFN5Ab1/m3B8L+PNBmY3HY/r7Zzt5TS+1YiIiIjYqGTrTkVJGk/jl2jusv3Mehh/O0lPA7+zfddwjx8RERER6yZX9CvK9uPAn63H8V/kD79+A4CkPwJ6S/r/0vbz6yuWiIiIiPiPkujHsCnJfH7hJiIiImIUyNadiIiIiIgKSqIfEREREVFBSfQjIiIiIiooiX5ERERERAUl0Y+IiIiIqKAk+hERERERFZREPyIiIiKigpLoR0RERERUUBL9jZykB0Y6hoFI+rCk8SMdR0REREQ7SaK/kbN98EjHMAgfBpLoR0RERAyBbI90DDGCJHXb7pBUA2YAK4G9gYXAJ2xb0v7ARcDWwCrgL4E3gMuAicDvgc/bvkfSNBqJ+dbA7sCFwFuAvyl9j7D9gqTdgEuATuBV4ETbT/YS38HALcBL5XE0MNf2O8v53YHrbb9T0gpgDnA48Dvg47Z/LKkTuBzYpQx7mu35LfNMB6YDdHZ27jdnzpy1e0Nj1Onu7qajo2Okw4hhlDWtlqxntWQ9N7zJkycvtD2xt3ObbehgYlR7B/B24JfAfODdkh4Crgem2F4gaVsaSfSpgG3/uaQ9gTsk7VHG2buMtQXwY+CLtt8h6WvAccDXgZnASbafkfQu4FLgPa0B2X5A0jzgFts3AEh6SdIE24uA44FZTV1eKjH1zPMBGh9Svmb7B5J2AW4H9mqZZ2aJia6uLtdqtbV+E2N0qdfrZD2rJWtaLVnPasl6ji5J9KPZQ7Z/DiBpETCOxlX0f7O9AMD2y+X8IcDFpe5JST8FehL9e2y/Arwi6SXg5lK/FNhHUgdwMDBXUs/cbx1CnFcCx0v6PDAFOKDp3HVNz18r5cOA8U1zbSupw3b3EOaMiIiIaCtJ9KPZqqbyatb+76N5nDVNx2vKmJsAL9qesJbjfwf4EnA3sND2803n3Et5E+BA26+t5XwRERERbSc348ZAngJ2Kvv0kbSNpM2A+4GppW4PGvvfnxrMgOVbgeWSji39JWnffrq8AmzT1P81GttvLuPN23agcYW/5/mHpXwH8LmeBpLW9gNGRERERNtIoh/9sv06jaT5YkmLgTtp7L2/FNhE0lIae/in2V7V90j/wVTgU2XMZcCR/bT9X8CZkh4tN/ECXEvjG4I7WtpuL2kJjXsITi91pwATJS2R9Dhw0hDijIiIiGhL2bqzkbPdUZ7rQL2p/rNN5QXAgb10P76X8WYDs5uOx/V2zvZy4P2DjHE+//HnNQ8BZtle3VJ/ge0vtvRfyR+u9EdERERsFJLoR9uRdCOwG738Sk9ERERENCTRj1FD0tnAsS3Vc22f11xh+6je+jd/exARERGxsUuiH6NGSejPG7BhRERERAwoN+NGRERERFRQEv2IiIiIiApKoh8RERERUUFJ9CMiIiIiKiiJfkREREREBSXRj4iIiIiooCT6EREREREVlER/IyKpe6RjaCZpnKSPD9BmgqQjNlRMEREREVWRRD9G0jig30QfmAAk0Y+IiIgYItke6RiiF5I+AZwCvAX4EfAZ4CXgMhqJ778Bfwd8FdgFOM32PEnTgKOAMcDOwDW2/6GM2W27Q5JKv8MBA+favl7SN4Hv2v5eaX8tMAfYHvgwsDWwO3BhietvgFXAEbZfkLQbcAnQCbwKnGj7SUmzgZeBicB/As6yfYOkB4G9gOXA1ba/1vIevAX4MbAl8Avgn4BzgYNtPydpE+Bp4CDgAuC1Mse2wOdt3yJpU+B8oAa8FbjE9v/s5f2eDkwH6Ozs3G/OnDmDWqcY/bq7u+no6BjpMGIYZU2rJetZLVnPDW/y5MkLbU/s9aTtPEbZg0byezOweTm+FDiORlJ+eKm7EbgD2BzYF1hU6qfR+BDwRzQS5MeAieVcd3k+GrgT2BTYEfi/wE7AXwDfK23G0EjANytj/hjYhkYS/xJwUmn3NRofMgDuAnYv5XcBd5fybGAujW+QxgM/LvU14JYB3otpwDeajr/UNN/7gO80zXFbmWN34OfAFjSS93NKm7cCDwNv62/OPfbYw1Ed99xzz0iHEMMsa1otWc9qyXpueMDD7iOn2azX7D9G2l8C+wELGhff2RL4NfA6jWQWYCmwyvYbkpbS2AbT407bzwNI+i5wCI0Et8chwHW2VwPPSroX2N+NbwQuldRJ48PAd2z/vsRwj+1XgFckvUTjg0hPHPtI6gAOBuaW9tBIrHt8z/Ya4HFJO67De3MVcBPwdeCTwKymc3PKHM9I+gmwJ40PA/tIOqa0GUPjg8DydYghIiIiYtRLoj86icZWlv/3TZXSGeWTG8AaGttmsL1GUvNatu7HGsr+rG8CnwA+ChzfVL+qqbym6XgNjb+jTYAXbU/oY9zm/uqjzYBs/0zSs5LeAxwATG0+3dq8zPU527ev7ZwRERER7Sg3445OdwHHSPpjAEk7SNp1CP3fW/psSWNv/fyW8/cDUyRtWq7eTwIeKudmA6cB2H58sBPafhlYLunYErMk7TtAt1dobAcaapsrgWuAueVbiR7HStqk3CvwZ/z/7N15mB5Vnf7/980mS5BFWgZHISNCAwYNEtnBjjAuIwoIigqMASQyMKKjuIIaEFTE8atEGY3zg6Cgk7AjDjs8LGELgewQmCE4juiwR5olYrh/f9RpfXjoNXTS3ZX7dV19ddWps3yqTv/xqXpOPQ2LgKuAf5K0Zolra0nr9eukIiIiIkawJPrDUEmwTwSuljSXaj39ZgPo4k7gQmAu1fKbu1qOX1yOzQGup3o59g9l7P8D7uWlS2L66xDgSElzgAXAfn3UnwsskzRH0r/0UOcGYDtJsyUdXMouA0Z1E+P/UJ37FVTvEDxPdVOwELhb0nzgJ+STrIiIiFgFJOEZpmxPA6a1FI9qOj6ppX7zK+7/a3v/bvocVX4b+Hz5eQlJ61KtYf9lU7upVE/6u/ZHd3fM9mLgPd2MO6GHOF4A3tlav6XuE8DbW4rfCsyxfV9L+bW2j25p/yLVtxN9pbdxIiIiIuomT/TjLyTtQ/U0f7LtJUMdT3ckfYnq04ov91U3IiIiYlWWJ/o10/r0fYBtrwUG8i7AoJH0buC0luLFtg9oLrD9barvxaelfMKKiy4iIiJi5EmiH8NC+VacfDNORERExCDJ0p2IiIiIiBpKoh8RERERUUNJ9CMiIiIiaiiJfkREREREDSXRj4iIiIiooST6ERERERE1lEQ/IiIiIqKGkujXhKTRkuYPdRwrgqQOSbsNdRwRERERI0kS/RgJOoAk+hEREREDkES/XlaX9FNJCyRdLWkdSWMl3S5prqSLJW0EIKkhaVzZ3kTSQ2X7zZLulDS7tNmqlB/aVP4TSauX8o9KmidpvqTTStmHJH2vbH9a0oNlFbBhtwAAIABJREFU+42SZpTthySdJOnu0n6b7k5I0mjgaOBfyth7Slosac1y/NVd++WcflDqzZe0U6mznqSzSvz3SNpvhVz9iIiIiGFkjaEOIAbVVsBHbR8laTpwIPAF4FO2b5R0MvB14DO99HE08APb50lai+rmYVvgYGB32y9IOhM4RNK1wGnAjsCTwNWS9gduLuMC7Ak8Lulvy/ZNTWM9Zvttko4Bjgc+0RqM7Yck/RjotP1dqG5SgPcBlwAfAS4qcQGsa3uspL2As4AxwAnA9baPkLQhcKeka20/0zWOpInARIC2tjYajUbvVzpGjM7OzsxnzWRO6yXzWS+Zz+EliX69LLY9u2zPArYENrR9Yyk7Bzi/jz5uA06Q9HqqBPoBSXtTJfMzSzK9DvAI8HagYftRAEnnAXvZvkTSKEnrA28AfgHsRZXoX9Q0Vtf2LOCDAzjPf6e6kbgEOBw4qunYLwFs31Se9m8IvAv4gKTjS521gc2Be7sa2Z4CTAFob293R0fHAMKJ4azRaJD5rJfMab1kPusl8zm8JNGvl6VN28uADXup+2f+unRr7a5C27+QdAfVE/P/lPRJQMA5tr/c3EEfS2BupUrCF1E94T8C2BX4XDfxLmMAf4u2Z5SXjzuA1W03v4Ts1uol/gNtL+rvGBEREREjXdbo19sS4ElJe5b9w4Cup/sPUT2lBzioq4GkNwIP2j4DuBR4C3AdcJCk15Y6G0vaArgTeEdZ47868NGm/m+mWo5zE3APMB5YanvJcpzH08D6LWU/o/qk4OyW8oNLjHsAS8p4VwGfUvk4QtIOyxFDRERExIiSRL/+Pg6cLmkuMBY4uZR/F/gnSfcAmzTV/zAwX9JsqvXtP7O9EDiRag3+XOAaYDPbvwe+BNwAzAFm2b609HMz1bKdm2wvA34L3LKc5/Ar4ICul3FL2XnARpSlOk2eL+f0Y+DIUvYNYE1grqQFZT8iIiKi1rJ0pyZsP0SVmHftf7fp8C7d1L+P6ml9lxNL+beBb3dTfxowrZvyX/LyZBvb/021ZKZr/10tx0c3bd9F9RWa3bJ9f0usAHsAF9h+qqX8XNsvednY9nPAJ3vqPyIiIqKOkujHiCNpMvBe4B+GOpaIiIiI4SqJfgwbkg4HPt1SPMP2sc0Ftj/VXXvbHSsotIiIiIgRJ4l+DBu2z+blL9dGRERExHLIy7gRERERETWURD8iIiIiooaS6EdERERE1FAS/YiIiIiIGkqiHxERERFRQ0n0IyIiIiJqKIl+REREREQNJdGvEUmTJB0/wDbjJJ3xCsd9SNImr6SPlUHS/pK2G+o4IiIiIlaGJPqrONt32T6utVxSHf+Z2v5AEv2IiIhYJSTRH+EknSDpfkm3AO2lbEtJV0qaJelmSduU8g9Jmi9pjqSbSlmHpMvL9iRJP5c0A/i5pDZJF0qaWX52L/VeI+lqSQsk/TugPmK8pMSyQNLEpvJOSaeX8msl7SSpIelBSR8oddaWdLakeZLukTS+lE+Q9MOmvi6X1NHU76nlPG+XtKmk3YAPAKdLmi1py8GZgYiIiIjhqY5PbVcZknYEPgKMpZrLu4FZwBTgaNsPSNoZOBN4J/A14N22fydpwx663Q7Yw/Zzkn4B/D/bt0jaHLgK2Bb4OnCL7ZMlvQ84so9Qj7D9hKR1gJmSLrT9OLAecL3tz0u6GDgF+PsSwznAZcCxgG1vX25Yrpa0dR/jrQfcbvsESd8BjrJ9iqTLgMttX9DNtZwITARoa2uj0Wj0MUSMFJ2dnZnPmsmc1kvms14yn8NLEv2RbU/gYtvPApREdm1gN+B86S8P2l9Vfs8ApkqaDlzUQ5+X2X6ubO8DbNfUz6sljQL2Aj4IYPvXkp7sI87jJB1Qtt8AbAU8DvwJuLKUzwOW2n5B0jxgdCnfA5hcxrpP0m+AvhL9PwGXl+1ZVDcPvbI9heoGifb2dnd0dPTVJEaIRqNB5rNeMqf1kvmsl8zn8JJEv35WA56yPbb1gO2jyxP+9wGzyicCrZ5p6WsX2883V2hK/PtUltPsA+xq+1lJDaqbEYAXbLtsvwgsLXG+2I93BP7MS5eerd203dzvMvJ3HhEREaugrNEf2W4C9pe0jqT1gfcDzwKLJX0IQJW3lu0tbd9h+2vAo1RP13tzNfCprh1JXTcPNwEfK2XvBTbqpY8NgCdLkr8NsMsAz/Fm4JAy1tbA5sAi4CFgrKTVJL0B2KkffT0NrD/A8SMiIiJGpCT6I5jtu4FpwBzgCmBmOXQIcKSkOcACYL9Sfnp5qXU+cGtp15vjgHGS5kpaCBxdyk8C9pK0gGoJz//00seVwBqS7gW+Ddw+kHOker9gtbKcZxowwfZSqmVIi4GFwBlU7yf05T+Az5eXevMybkRERNRaljSMcLZPBU7t5tB7uqn7wW7qNcoPtie11H8MOLibfh4H3tXP+JYC7+3h2Kim7daxR5XfzwOHd9PWlCf9ffR7AXBB2Z5Bvl4zIiIiVhF5oh8RERERUUN5oh+DQtJrgOu6ObR3+QQgIiIiIlaiJPoxKEoy/7Jv+omIiIiIoZGlOxERERERNZREPyIiIiKihpLoR0RERETUUBL9iIiIiIgaSqIfEREREVFDSfQjIiIiImooiX5ERERERA0l0Y9hSVL+x0NERETEK5BEP1YYSaMl3SfpPEn3SrpA0rqSHpK0SakzTlKjbE+S9HNJM4CfS5og6VJJDUkPSPp6U9+flTS//HymlK0n6deS5pTyg0v5jpJulDRL0lWSNlv5VyMiIiJi5cpT01jR2oEjbc+QdBZwTB/1twP2sP2cpAnATsAY4FlgpqRfAwYOB3YGBNwh6UbgjcDDtt8HIGkDSWsCk4H9bD9akv9TgSMG+0QjIiIihpMk+rGi/db2jLJ9LnBcH/Uvs/1c0/41th8HkHQRsAdVon+x7WeayvcErgT+VdJpwOW2b5Y0hupG4RpJAKsDv28dVNJEYCJAW1sbjUZjec41hqHOzs7MZ81kTusl81kvmc/hJYl+rGjuZv/P/HXZ2Notx5/pR/vuB7Lvl/Q24B+AUyRdB1wMLLC9a69B2lOAKQDt7e3u6OjorXqMII1Gg8xnvWRO6yXzWS+Zz+Ela/RjRdtcUleS/THgFuAhYMdSdmAf7f9e0saS1gH2B2YANwP7l/X+6wEHADdLeh3wrO1zgdOBtwGLgLauGCStKenNg3d6EREREcNTnujHirYIOLasz18I/BtwJ/D/SfoG0Oij/Z3AhcDrgXNt3wUgaWo5BvDvtu+R9G7gdEkvAi8A/2T7T5IOAs6QtAHV3/z3gQWDeI4RERERw04S/VjR/mz70Jaym4GtWyvantRN+/+1vX83db8HfK+l7Crgqm7qzgb2GkDMERERESNelu5ERERERNRQnujHCmP7IapvvFne9lOBqYMUTkRERMQqJU/0IyIiIiJqKIl+REREREQNJdGPiIiIiKihJPoRERERETWURD8iIiIiooaS6EdERERE1FAS/YiIiIiIGkqiHxERERFRQ0n0IyIiIiJqaMQn+pIaksYNQj+dA6w/SdLx3ZS/TtIFZbtD0uVl+wOSvlS295e0XT/GeEk9SSdL2mcgcQ4nfV1jSRtKOmZlxRMRERFRZys00Ze0xorsf6Akrb6ix7D9sO2Duim/zPa3y+7+QJ+Jfms921+zfe3gRDosbQgk0Y+IiIgYBH0m+pJGS7pP0nmS7pV0gaR1Je0o6UZJsyRdJWmzUr8h6fuS7gI+LelDkuZLmiPpplJnbUlnS5on6R5J40v5BEkXSbpS0gOSvtMUx79JukvSAkkn9fcEJXVK+ldJc4BdJR0q6U5JsyX9pDn5l3RqifN2SZuWsvdLuqPEeW1XefFWSbeVWI9qul7zu4ljgqQfStoN+ABweolhS0lHSZpZxr6wXN/u6k2VdFDpb+8S0zxJZ0l6VSl/SNJJku4ux7bp5dqMapqHuZIOLOUfLWXzJZ3Wci1PL3NwraSdynw/KOkDTed5aSl/QNLXexj78+Wc5zbN57eBLcv5nt5Lve76O1nSZ1rm8tOqPlVplL/brr9j9dRPRERERF3094l7O3Ck7RmSzgKOBQ4A9rP9qKSDgVOBI0r9tWyPA5A0D3i37d9J2rAcPxaw7e1LInq1pK3LsbHADsBSYJGkybZ/C5xg+4mSmF8n6S225/Yj9vWAO2x/TtK2wBeB3W2/IOlM4BDgZ6Xe7bZPKDcYRwGnALcAu9i2pE8AXwA+V/p+C7BLaXuPpF/3FYztWyVdBlxuu2uJz1O2f1q2TynXenI39Si/1wamAnvbvl/Sz4B/Ar5fhnnM9ttULYM5HvhED+F8FVhie/vS70aSXgecBuwIPEk1N/vbvqSc5/W2Py/p4nJ9/p7qU4dzgMtKvzsBY4BngZmSfm37rq5BJb0L2KrUE3CZpL2ALwFjbI/trZ7tm7o5l7OAi4DvS1oN+Ehptz3V39ObgYeBGcDuVPP6F5ImAhMB2traaDQaPVyyGGk6OzsznzWTOa2XzGe9ZD6Hl/4m+r+1PaNsnwt8hSqRu6Ykn6sDv2+qP61pewYwVdJ0qkQMYA9gMoDt+yT9BuhK9K+zvQRA0kJgC+C3wIdLMrYGsBlVctmfRH8ZcGHZ3psqgZ1Z4l4HeKQc+xNwedmeRZXAArwemKbqE4u1gMVNfV9q+zngOUk3UCWWs/sRU6sxJcHfEBgFXNVH/XZgse37y/45VDdPXYl+13WeBXywl372oUqIAbD9ZEm4G7YfBZB0HrAXcAnVNbqyVJ8HLC03TPOA0U39XmP78dL+Iqr5vqvp+LvKzz1lfxRVQv8/LfH1VO9lib7thyQ9LmkHYFPgHtuPl3m+0/b/lnhml1hvaWk/BZgC0N7e7o6OjpdfrRiRGo0Gmc96yZzWS+azXjKfw0t/E3237D8NLLC9aw/1n/lLQ/toSTsD7wNmSdqxj7GWNm0vA9aQ9HdUT6bfXpLRqcDa/Yz9edvLyraAc2x/uZt6L9juOs9l/PXaTAa+Z/sySR3ApKY2rdeldb+/pgL7254jaQLQsZz9dOm6hs3nMRiar9GLXePYflEvfR+jr+si4Fu2f/KSQml0f+r14t+BCcDfUD3h7/Kyv6l+9hcRERExYvX3ZdzNJXUl9R8DbgfausokrSnpzd01lLSl7Ttsfw14FHgDcDPVkhnKkp3NgUW9jP9qqpuHJarWyL+3n3G3ug44SNJry9gbS9qijzYbAL8r2x9vObafqvcNXkOVnM/sZxxPA+s37a8P/F7SmpTr0kO9LouA0ZLeVPYPA27s59jNrqH6JAColu4AdwLvkLRJWSb10eXo++/LtV2H6oXiGS3HrwKOkDSqjPu3ZU5az7enej25GHgP8Hb6/lQkIiIiotb6m+gvAo6VdC+wEdVT7oOA01S95Dob2K2Htqd3vdgJ3ArMAc4EVitLPqYBE2wv7aE9tudQLd+4D/gFL08c+8X2QuBEqnXnc6kS3c36aDYJOF/SLOCxlmNzgRuobny+YfvhfobyH8DnVb1MuyXVWvk7qM7rvl7qdZ3H88DhJa55VE/Xf9zPsZudAmyk8rI0MN7276nWyt9ANVezbF86wH7vpFouNRe4sHl9fon/aqp5vK3EfwGwflnuM6PEc3pP9Xoa1PafStzTmz7FiYiIiFgl6a8rMXqoUC2nuNz2mJURUIxsZenRONv/PARjrwbcDXzI9gPL2097e7sXLertA6YYSbJetH4yp/WS+ayXzOfKJ2lW15fgtBrx/zArAkDVPxb7L6qXuZc7yY+IiIioiz5fSrT9ENU37Axrku4AXtVSfJjteUMRz3Ai6XDg0y3FM2wf2139V8L2VKqXi1eI8j7Edd0c2tv2G1fUuBEREREjTW2+fcT2zkMdw3Bl+2zg7KGOYzCUdfxjhzqOiIiIiOEuS3ciIiIiImooiX5ERERERA0l0Y+IiIiIqKEk+hERERERNZREPyIiIiKihpLoR0RERETUUBL9iIiIiIgaSqJfI5JeJ+mCXo5vKOmY/tZfzhgmSTp+MPscDJLGSvqHoY4jIiIiYmVJol8jth+2fVAvVTYEjhlA/ToZCyTRj4iIiFVGEv0RQNKhku6UNFvSTyTtLGmupLUlrSdpgaQxkkZLml/avLmpzVxJWwHfBrYsZae31J8g6SJJV0p6QNJ3msY/UtL9pb+fSvphP+MeK+n2Mv7FkjYq5UdJmilpjqQLJa1byqdKOkPSrZIelNTrTYikL0qaV/r5dilrSDqtxHq/pD0lrQWcDBxczv3ggc9CRERExMiyxlAHEL2TtC1wMLC77RcknQm0A5cBpwDrAOfani9pdFPTo4Ef2D6vJLqrA18CxtgeW/purg/VU+8dgKXAIkmTgWXAV4G3AU8D1wNz+hn+z4BP2b5R0snA14HPABfZ/mmJ4RTgSGByabMZsAewTTnHbpcWSXovsB+ws+1nJW3cdHgN2zuVpTpft72PpK8B42z/cw/9TQQmArS1tdFoNPp5ijHcdXZ2Zj5rJnNaL5nPesl8Di9J9Ie/vYEdgZmSoErsH6F6Qj0TeB44rpt2twEnSHo9VWL9QGnfm+tsLwGQtBDYAtgEuNH2E6X8fGDrvjqStAGwoe0bS9E5wPlle0xJ8DcERgFXNTW9xPaLwEJJm/YyxD7A2bafBeiKr7io/J4FjO4r1tJ+CjAFoL293R0dHf1pFiNAo9Eg81kvmdN6yXzWS+ZzeEmiP/wJOMf2l19SKG1GlSSvCawNPNN83PYvJN0BvA/4T0mfBB7sY6ylTdvLWHF/H1OB/W3PkTQB6Oghhj7vTHrQ1ceKPIeIiIiIYS1r9Ie/64CDJL0WQNLGkrYAfkK1pOY84LTWRpLeCDxo+wzgUuAtVEtv1h/g+DOBd0jaSNIawIH9aVQ+GXhS0p6l6DCg6+n++sDvJa0JHDLAeLpcAxzetL5/4z7qL8+5R0RERIxYedo5zNleKOlE4GpJqwEvUCXuL5Sn9qsDt0p6Jy99Yv9h4DBJLwB/AL5p+wlJM8oLuFcAP+rH+L+T9E3gTuAJ4D5gST/D/zjw45KMPwgcXsq/CtwBPFp+DzgBt32lpLHAXZL+BPwn8JVemtwAfEnSbOBbtqcNdMyIiIiIkUS2hzqGGOYkjbLdWZ7oXwycZfvioY5rRWlvb/eiRYuGOowYJFkvWj+Z03rJfNZL5nPlkzTL9rjujmXpTvTHpPIkfD6wGLhkiOOJiIiIiD5k6U70yfbL/tOtpBOAD7UUn2/71MEcW9L2wM9bipfa3nkwx4mIiIiomyT6sVxKQj+oSX0P48yj+n7/iIiIiBiALN2JiIiIiKihJPoRERERETWURD8iIiIiooaS6EdERERE1FAS/YiIiIiIGkqiHxERERFRQ0n0IyIiIiJqKIl+REREREQNJdEfwSS9TtIFvRzfUNIx/a2/nDFMkvSy/5w73EjqkLTbUMcRERERsbIk0R/BbD9s+6BeqmwIHDOA+nXWASTRj4iIiFWGbA91DNFC0qHAccBawB3AWcBPgZ2A1YE7gYOBTuBy22MkvRk4u7RZDTgQ+AawH7AIuAb4UVP9CcAHgHWBLYGLbX+hjH8k8EXgKWAOsNT2P/cQ6ySg0/Z3JY0Fflz6/G/gCNtPSjoKmFhi+y/gMNvPSpoK/BEYB/wN8AXb3X7iIGkUcCmwEbAmcKLtSyWNBq4EbqdK5GeW63AS8FrgEOCRcnwZ8CjwKds3t/Q/scRIW1vbjtOnT+8ujBiBOjs7GTVq1FCHEYMoc1ovmc96yXyufOPHj59le1x3x5LoDzOStgW+A3zQ9guSzqRKUrcG1gbWAf7X9rdKktuVuE8Gbrd9nqS1qG4INu06Xvpurj8B+BqwA7CU6mZgD6pk+FbgbcDTwPXAnH4m+nOpkugbJZ0MvNr2ZyS9xvbjpf4pwP/ZnlwS/fWoblq2AS6z/aYexlkDWNf2HyVtUq7JVsAWVDcPOwALqBL9OcCRVDcyh9vevznOvuagvb3dixYt6qtajBCNRoOOjo6hDiMGUea0XjKf9ZL5XPkk9Zjor7Gyg4k+7Q3sCMyUBFVi/whwMlUS+zzV0/5WtwEnSHo9cJHtB0r73lxnewmApIVUSfMmwI22nyjl51PdZPRK0gbAhrZvLEXnAOeX7TElwd8QGAVc1dT0EtsvAgslbdrbEMA3Je0FvAj8LdWNDMBi2/NKHAvKeVnSPGB0X7FHRERE1FES/eFHwDm2v/ySQmkzqiR5Taon+880H7f9C0l3AO8D/lPSJ4EH+xhradP2Mlbc38NUYH/bc8onCR09xNDbnckhQBuwY/mk4yGq69Dax4tN+y+Sv/GIiIhYReVl3OHnOuAgSa8FkLSxpC2AnwBfBc4DTmttJOmNwIO2z6Bay/4WqqU36w9w/JnAOyRtVJbLHNifRuWTgScl7VmKDgO6nu6vD/xe0ppUCfvy2AB4pCT546k+fRiI5bkWERERESNWnnYOM7YXSjoRuFrSasALVIn7C+Wp/erArZLeyUuf2H8YOEzSC8AfgG/afkLSDEnzgSuoXsbta/zfSfom1Qu/TwD3AUv6Gf7HgR9LWrfEdngp/yrVS8WPlt/Lk3CfB/yqLMe5q8Q1EL8CLpC0H928jBsRERFRN3kZN15G0ijbneWJ/sXAWbYvHuq4Vpa8jFsveTGsfjKn9ZL5rJfM58rX28u4WboT3ZkkaTYwH1gMXDLE8URERETEAGXpTryM7Zf9p1tJJwAfaik+3/apgzm2pO2Bn7cUL7W982COExEREVF3SfSjX0pCP6hJfQ/jzAPGruhxIiIiIuouS3ciIiIiImooiX5ERERERA0l0Y+IiIiIqKEk+hERERERNZREPyIiIiKihpLoR0RERETUUBL9iIiIiIgaSqIfg0rS/pK2G6S+Jknq7p93vU7SBT20GS3pY037EyT9cDDiiYiIiBhJkujHYNsfGFCiL2lA/7jN9sO2D+qhn9HAx17WKCIiImIVI9tDHUMMc5IOBY4D1gLuAI4BlgA/APYFngP2A7YELi/HlgAHli5+BLQBzwJH2b5P0lTgeWAHYIbtz3Yz7qTS55uATYDv2P6ppNHA5bbHSJoAfBAYBawOvArYFlgMnAM8CXwAWLf0dbHtL3Qz1kRgIkBbW9uO06dPX55LFcNQZ2cno0aNGuowYhBlTusl81kvmc+Vb/z48bNsj+vu2ICepMaqR9K2wMHA7rZfkHQmcAiwHnC77RMkfYcqgT9F0mVUSfgFpf11wNG2H5C0M3Am8M7S/euB3Wwv6yWEtwC7lPHukfTrbuq8DXiL7SckdQDH2963jD8BGEt1Q7EUWCRpsu3fNndgewowBaC9vd0dHR0DuEoxnDUaDTKf9ZI5rZfMZ71kPoeXJPrRl72BHYGZkgDWAR4B/kT19B5gFvD3rQ0ljQJ2A84vbaF64t7l/D6SfIBLbT8HPCfpBmAnYHZLnWtsP9FLH9fZXlJiWghsAfy2l/oRERERI14S/eiLgHNsf/klhdLx/uu6r2V0/7e0GvCU7bE99P1MP8ZvXVvW3VqzvvpZ2rTdU6wRERERtZKXcaMv1wEHSXotgKSNJW3RS/2ngfUBbP8RWCzpQ6WtJL11gOPvJ2ltSa8BOoCZfdT/y/gRERERq7Ik+tEr2wuBE4GrJc0FrgE266XJfwCfl3SPpC2p1vMfKWkOsIDqpd2BmAvcANwOfMP2w/2ov0zSHEn/MsCxIiIiImojSxiiT7anAdNaikc1Hb8AuKBsz+DlX6/5nm76nNCPcSf1UP4QMKZsTwWmNh17gb++7Nul+fi+fY0bERERUQd5oh8RERERUUN5oh9DTtLhwKdbimfYPnYo4omIiIiogyT6MeRsnw2cPdRxRERERNRJlu5ERERERNRQEv2IiIiIiBpKoh8RERERUUNJ9CMiIiIiaiiJfkREREREDSXRj4iIiIiooST6MeJJmirpoKGOIyIiImI4SaIfqxxJ+f8RERERUXtJ9GOlkvRVSYsk3SLpl5KOl7SlpCslzZJ0s6RtSt2pks6QdKukB7ue2qvyw9LPtcBrm/rfUdKNpa+rJG1WyhuSvi/pLl7+X3gjIiIiaidPNmOlkfR24EDgrcCawN3ALGAKcLTtByTtDJwJvLM02wzYA9gGuAy4ADgAaAe2AzYFFgJnSVoTmAzsZ/tRSQcDpwJHlL7Wsj1uhZ9oRERExDCQRD9Wpt2BS20/Dzwv6VfA2sBuwPmSuuq9qqnNJbZfBBZK2rSU7QX80vYy4GFJ15fydmAMcE3pa3Xg9019TespMEkTgYkAbW1tNBqN5T7JGF46OzsznzWTOa2XzGe9ZD6HlyT6MdRWA56yPbaH40ubttVDnebjC2zv2sPxZ3pqaHsK1ScLtLe3u6Ojo4+hYqRoNBpkPuslc1ovmc96yXwOL1mjHyvTDOD9ktaWNArYF3gWWCzpQ/CX9fdv7aOfm4CDJa1e1uCPL+WLgDZJu5a+1pT05hVyJhERERHDXBL9WGlsz6RaZz8XuAKYBywBDgGOlDQHWADs10dXFwMPUK3N/xlwW+n/T8BBwGmlr9lUy4IiIiIiVjlZuhMr23dtT5K0LtWT+Vm2FwPvaa1oe0LL/qjy28A/d9e57dlUa/hbyzteceQRERERI0gS/VjZpkjajuol3HNs3z3UAUVERETUURL9WKlsf2yoY4iIiIhYFWSNfkREREREDSXRj4iIiIiooST6ERERERE1lEQ/IiIiIqKGkuhHRERERNRQEv2IiIiIiBpKoh8RERERUUNJ9CMiIiIiaiiJfkREREREDSXRrylJoyXN70e9DkmXD9KYEyT9sGwfLekfB6Pf5YzjdUMxdkRERMRwscZQBxD1ZPvHK7J/SavbXtbD4QnAfODhFRlDRERExHCWJ/qrAElvlHSPpLf3UW89SWdJurPU36+UT5B0kaQrJT0g6TtNbQ6XdL+kO4HORn/CAAAgAElEQVTdm8onSTq+bDcknVb6vV/SnqV8XUnTJS2UdLGkOySN6yW+Tkn/KmkOsKukr0maKWm+pCmqHASMA86TNFvSOpJ2lHSjpFmSrpK02Su5nhEREREjQZ7o15ykduA/gAm25/RR/QTgettHSNoQuFPSteXYWGAHYCmwSNJk4M/AScCOwBLgBuCeHvpew/ZOkv4B+DqwD3AM8KTt7SSNAWb3Ed96wB22P1fObaHtk8v2z4F9bV8g6Z+B423fJWlNYDKwn+1HJR0MnAoc0XKdJgITAdra2mg0Gn2EEiNFZ2dn5rNmMqf1kvmsl8zn8JJEv97agEuBD9pe2I/67wI+0PUkHlgb2LxsX2d7CVQJNrAFsAnQsP1oKZ8GbN1D3xeV37OA0WV7D+AHALbnS5rbR3zLgAub9sdL+gKwLrAxsAD4VUubdmAMcI0kgNWB37d2bHsKMAWgvb3dHR0dfYQSI0Wj0SDzWS+Z03rJfNZL5nN4SaJfb0uA/6FKqPuT6As40PailxRKO1M9ye+yjIH/7XS1X562XZ7vWpcvaW3gTGCc7d9KmkR1Y9JKwALbuy7nmBEREREjUtbo19ufgAOAf5T0sX7Uvwr4lMqjb0k79FH/DuAdkl5Tlsh8aIDxzQA+XMbaDth+AG27kvrHJI0CDmo69jSwftleBLRJ2rWMs6akNw8wzoiIiIgRJ0/0a872M5L2pVq60mn7sl6qfwP4PjBX0mrAYmDfXvr+fXmSfhvwFH2vsW91JnBOWQp0H9XSmyX9aWj7KUk/pfp2nT8AM5sOTwV+LOk5YFeqm4AzJG1A9Tf//TJWRERERG0l0a8p2w9RrU3H9lNAt9+4Y7sBNMr2c8Anu6kzlSp57trft2n7bODsbtpMatruaNp+jL+u0X8eONT285K2BK4FftPLOY1q2T8ROLGbehfy0rX8s4G9euo3IiIioo6S6MdQWhe4oSz7EXCM7T8NcUwRERERtZBEfxUh6d3AaS3Fi20fMBTxANh+muo7719C0h3Aq1qKD7M9b6UEFhEREVEDSfRXEbavonrZdtizvfNQxxAREREx0uVbdyIiIiIiaiiJfkREREREDSXRj4iIiIiooST6ERERERE1lEQ/IiIiIqKGkuhHRERERNRQEv2IiIiIiBpKor+KkzRB0g8Hqa9Jko4v2ydL2mcw+l2OOD4jad2hGDsiIiJiuEiiHyuE7a/ZvnZF9K1Kb3+7nwGS6EdERMQqLYn+CCdptKT5TfvHlyfrx0laKGmupP/oZ19tki6UNLP87F7KJ0k6S1JD0oOSjmtqc4Kk+yXdArQ3lU+VdFDZfkjSSZLuljRP0jZN410jaYGkf5f0G0mb9HKeiyT9DJgPvEHSv0m6q7Q/qdQ7DngdcIOkG0rZuyTdVsY/X9KoAV3kiIiIiBFojaEOIFaYLwF/Z3uppA372eYHwP+zfYukzYGrgG3LsW2A8cD6wCJJ/wa8BfgIMJbqb+luYFYPfT9m+22SjgGOBz4BfB243va3JL0HOLKP+LYCPm77dqhuMmw/IWl14DpJb7F9hqTPAuNtP1ZuHE4E9rH9jKQvAp8FTm7uWNJEYCJAW1sbjUajH5crRoLOzs7MZ81kTusl81kvmc/hJYl+fc0FzpN0CXBJP9vsA2wnqWv/1U1Pv39teymwVNIjwKbAnsDFtp8FkHRZL31fVH7PAj5YtvcADgCwfaWkJ/uI7zddSX7x4ZKgrwFsBmxHdd7NdinlM8p5rQXc1tqx7SnAFID29nZ3dHT0EUqMFI1Gg8xnvWRO6yXzWS+Zz+Elif7I92deugRr7fL7fcBewPuBEyRtb/vPffS1GrCL7eebC0uCvLSpaBkD/9vpar88bbs80xTT31F9MvB2209Kmspfz72ZgGtsf3Q5x4yIiIgYkbJGf+T7P+C1kl4j6VXAvlTz+gbbNwBfBDYA+rMu/WrgU107ksb2Uf8mYH9J60han+qmYiBmAB8uY70L2GgAbV9NlfgvkbQp8N6mY09TLTECuB3YXdKbyjjrSdp6gHFGREREjDh5oj/C2X5B0snAncDvgPuA1YFzJW1A9UT7DNtP9aO744AfSZpL9bdxE3B0L2PfLWkaMAd4BJg5wPBPAn4p6TCq5TR/oErS+2R7jqR7qM73t1Q3DV2mAFdKetj2eEkTyjivKsdPBO4fYKwRERERI0oS/RqwfQZwxnK2nQpMLduPAQd3U2dSy/6Ypu1TgVO7aTOhaXt00/ZdQEfZXQK82/afJe1KtQyneYlQc38PAWNayib0UHcyMLlp/3rg7d3VjYiIiKirJPoxlDYHppfvxP8TcNQQxxMRERFRG0n0VxGSDgc+3VI8w/axQxEPgO0HgB2ayyS9Brium+p72358pQQWERERUQNJ9FcRts8Gzh7qOPpSkvm+XgKOiIiIiD7kW3ciIiIiImooiX5ERERERA0l0Y+IiIiIqKEk+hERERERNZREPyIiIiKihpLoR0RERETUUBL9iIiIiIgaSqK/ipM0WtL8bso/I2ndIYhnnKQzVva4EREREXWTRD968hlgpSb6ktawfZft41bmuBERERF1lEQ/ANaQdJ6keyVdIOk44HXADZJuAJD0Hkl3S5oj6bruOpG0mqSHJG3YVPaApE0lvV/SHZLukXStpE3L8UmSfi5pBvBzSR2SLi/HdpJ0W2lzq6T2Uj5B0kWSriz9f6dpvJfFKWk9SWdJurP0td8Kuo4RERERw8YaQx1ADAvtwJG2Z0g6C1gLeBgYb/sxSW3AT4G9bC+WtHF3ndh+UdKlwAHA2ZJ2Bn5j+/8k3QLsYtuSPgF8AfhcabodsIft5yR1NHV5H7Cn7T9L2gf4JnBgOTYW2AFYCiySNBl4voc4TwCut31EuQm5U9K1tp/pGkjSRGAiQFtbG41GY7kuZAw/nZ2dmc+ayZzWS+azXjKfw0sS/QD4re0ZZftcoHXpzC7ATbYXA9h+ope+pgFfA84GPlL2AV4PTJO0GdWNxOKmNpfZfq6bvjYAzpG0FWBgzaZj19leAiBpIbAFsFEPcb4L+ICk48v+2sDmwL1dndmeAkwBaG9vd0dHRy+nGCNJo9Eg81kvmdN6yXzWS+ZzeMnSnYAqie5tfyBuA95UPgXYH7iolE8Gfmh7e+CTVMl2l2fo3jeAG2yPAd7f0mZp0/Yyer9pFXCg7bHlZ3Pb9/ZSPyIiImLES6IfAJtL2rVsfwy4BXgaWL+U3Q7sJenvAHpaugNg28DFwPeAe20/Xg5tAPyubH+8n3E1t5nQj/o9xXkV8ClJKuU79HP8iIiIiBEriX4ALAKOlXQv1fKXf6NaxnKlpBtsP0q1fv0iSXP463KcnkwDDm2pNwk4X9Is4LF+xvUd4FuS7qEfy8x6ifMbVMt+5kpaUPYjIiIiai1r9Fdxth8Ctunm0OTy01XvCuCKfvZ5F9VymeayS4FLu6k7qWW/ATTK9m3A1k2HTyzlU4GpTW327S3Osv7/k/2JPSIiIqIu8kQ/IiIiIqKG8kQ/loukw4FPtxTPsH3sUMQTERERES+VRD+Wi+2zqb5CMyIiIiKGoSzdiYiIiIiooST6ERERERE1lEQ/IiIiIqKGkuhHRERERNRQEv2IiIiIiBpKoh8RERERUUNJ9CMiIiIiaiiJfrxikjqHOoZmkjok7TbUcUREREQMpST6MSJJ6u2fvXUASfQjIiJilZZEPwaNKqdLmi9pnqSDS/koSddJuruU71fKR0u6V9JPJS2QdLWkdXrpvyHp+5LuAj4t6f2S7pB0j6RrJW0qaTRwNPAvkmZL2lNSm6QLJc0sP7uvhMsRERERMaRke6hjiBFOUqftUZIOpEqy3wNsAswEdgYeBda1/UdJmwC3A1sBWwD/BYyzPVvSdOAy2+f2ME4DWGj7mLK/EfCUbUv6BLCt7c9JmgR02v5uqfcL4Ezbt0jaHLjK9rYtfU8EJgK0tbXtOH369MG7QDGkOjs7GTVq1FCHEYMoc1ovmc96yXyufOPHj59le1x3x3pb/hAxUHsAv7S9DPg/STcCbweuAL4paS/gReBvgU1Lm8W2Z5ftWcDoPsaY1rT9emCapM2AtYDFPbTZB9hOUtf+qyWNsv2XdwtsTwGmALS3t7ujo6OPMGKkaDQaZD7rJXNaL5nPesl8Di9J9GNlOARoA3a0/YKkh4C1y7GlTfWWAT0u3SmeadqeDHzP9mWSOoBJPbRZDdjF9vMDjDsiIiJixMoa/RhMNwMHS1pdUhuwF3AnsAHwSEnyx1Mt2RkMGwC/K9sfbyp/Gli/af9q4FNdO5LGDtL4EREREcNWEv0YTBcDc4E5wPXAF2z/ATgPGCdpHvCPwH2DNN4k4HxJs4DHmsp/BRzQ9TIucFwZf66khVTvEURERETUWpbuxCtme1T5beDz5af5+GPArj00H9NU77t9jNPRsn8pcGk39e4H3tJSfHBvfUdERETUTZ7oR0RERETUUJ7ox7Aj6UdA63fd/8D22UMRT0RERMRIlEQ/hh3bxw51DBEREREjXZbuRERERETUUBL9iIiIiIgaSqIfEREREVFDSfQjIiIiImooiX5ERERERA0l0Y+IiIiIqKEk+hERERERNZREPyIiIiKihpLoxwoj6StN26MlzV8BY4yTdMZg9xsREREx0iXRjxXpK31XWX6S1rB9l+3jVuQ4ERERESORbA91DFEDkg4FjgPWAu4A/gh8FpgHLABOAK4AbgF2A34H7Gf7uR76awBzgHcAawBH2L5T0iRgS+CNwP8APwGOt72vpFHAZGAcYOAk2xdKehdwEvAq4L+Bw213tow3EZgI0NbWtuP06dMH4arEcNDZ2cmoUaOGOowYRJnTesl81kvmc+UbP378LNvjuju2xsoOJupH0rbAwcDutl+QdCZVgv+c7bGlzmhgK+Cjto+SNB04EDi3l67XtT1W0l7AWcCYUr4dsIft5yR1NNX/KrDE9vZlzI0kbQKcCOxj+xlJX6S6ATm5eSDbU4ApAO3t7e7oaO42RrJGo0Hms14yp/WS+ayXzOfwkkQ/BsPewI7ATEkA6wCPdFNvse3ZZXsWMLqPfn8JYPsmSa+WtGEpv6yHTwL2AT7StWP7SUn7Ut0YzCixrQXc1p+TioiIiBjJkujHYBBwju0vv6RQOr6l3tKm7WVUNwS9aV1X1rX/zABju8b2RwfQJiIiImLEy8u4MRiuAw6S9FoASRtL2gJ4QdKar6Dfg0t/e1AtyVnSR/1rgGO7diRtBNwO7C7pTaVsPUlbv4KYIiIiIkaEJPrxitleSLUO/mpJc6kS7s2o1rzPlXTecnb9vKR7gB8DR/aj/inARpLmS5oDjLf9KDAB+GWJ7TZgm+WMJyIiImLEyNKdGBS2pwHTWopvB77YtD+mqf53+9HtubY/0zLOpJb9BtAo253Ax7uJ7Xrg7f0YLyIiIqI28kQ/IiIiIqKG8kQ/hpSkHwG7txT/wHbHEIQTERERURtJ9GNI2T6271oRERERMVBZuhMRERERUUNJ9CMiIiIiaiiJfkREREREDSXRj4iIiIiooST6ERERERE1lEQ/IiIiIqKGkuhHRERERNRQEv14CUlfadoeLWn+Shz7aEn/uLLGi4iIiKizJPrR6it9Vxl8ktaw/WPbPxuK8SMiIiLqRraHOoYYIpIOBY4D1gLuAP4IfBaYBywATgCuAG4BdgN+B+xn+7lu+toG+Jntncr+aOBXtreX9DXg/cA6wK3AJ21bUgOYDewB/BJYH+i0/V1JRwETS2z/BRxm+1lJU0uc44C/Ab5g+4Iy5heBQ4EXgStsf0nSlsCPgDbgWeAo2/d1E//EMh5tbW07Tp8+fXkuaQxDnZ2djBo1aqjDiEGUOa2XzGe9ZD5XvvHjx8+yPa67Y2us7GBieJC0LXAwsLvtFySdSZXgP2d7bKkzGtgK+KjtoyRNBw4Ezm3tz/Z9ktaS9He2F5e+p5XDP7R9cunz58C+wK/KsbW6/jglTWrq8iLbPy3lpwBHApPLsc2obg62AS4DLpD0XmA/YOdyQ7BxqTsFONr2A5J2Bs4E3tlN/FNKXdrb293R0dGPqxgjQaPRIPNZL5nTesl81kvmc3hJor/q2hvYEZgpCaqn7Y90U2+x7dllexYwupc+p1Ml+N8uvw8u5eOl/7+9uw+3rC7rP/7+DKMgDEHEwScMCvUY8hSMiCH+DoVWago5JobpmDlZmpGpZWGNpqlZKT4gjhVgPg0iCGqCCm5FE5gZhXlAkVBMEsEn0KMyAnP//thrZHM8Z2afYc7e+6x5v67La9b+ru9a617nHvWz1/nuPXkpsCuwF93fFmwO+iunnqRxUBPw9wQWARf17PtgVW0Crk5y32bsOOCMqvoRQFV9N8kiur+JeH9zjwA7b6F+SZKk1jDo77gCnFVVL7vbYPLiKfM29mzfSfcNwUxW0g3V5wLVPEXfhe5T9MVV9fXmqf0uPcf8cIZznQkcX1VXJVkKTMxQU5jZAuCWzb+hkCRJ2pH4Ydwd18XAkiT7ACTZK8l+wO1J7rUtJ6yq6+i+GXg5dz2p3xzqv908YV/S5+l2B25sajmpj/kfB56dZFfo3k9VfR/4apKnNmNJcmif15ckSZrXDPo7qKq6GjgF+FiStXSD8v3prlNfm+Td23jqlXQ/EHt2c51bgHcA6+kuv1nV53leTvcDwp8FfubDs1NV1YV01+uvTnIlsPk3EycBz0lyFd0lQ0/u+04kSZLmMZfu7MCqaiU/u0b+MuCvel4f1DP/n/s45z8D/zxl7BS6byqmzp2Y8np5z/bbgLdNc8zSKa8X9Wy/lu7nA3r3fxX4ra3VLUmS1DY+0ZckSZJayCf6mrUkbwWOnjJ8alWdMYx6JEmS9LMM+pq1qnr+sGuQJEnSlrl0R5IkSWohg74kSZLUQgZ9SZIkqYUM+pIkSVILGfQlSZKkFjLoS5IkSS1k0JckSZJayKCvWUkyOewaeiV5XpJnDrsOSZKkUeM/mKV5K8nCqjp92HVIkiSNIp/oa5skmUjy4Z7Xb0mytNm+Pskrknw+ybokD2vGFyU5oxlbm+QpWzj/ZJI3JNmQ5OIkY814J8kbk6wG/jzJ8iQvbvY9OMknklzVXPuAZvwlSVY113zF3P1UJEmSRodP9DVXvl1Vhyf5U+DFwB8BLwduraqDAZL8/BaO3w1YXVV/keTvgL8HXtDsu3dVLW7OsbznmHcDr62q85LsAixI8jjgIcCRQIALkjymqj7de7Eky4BlAGNjY3Q6nXtw6xolk5OT9rNl7Gm72M92sZ+jxaCvuXJu8+ca4Heb7eOAEzdPqKrvbeH4TcDKZvtdPeejZ/ynkuwOPLCqzmvOfVsz/jjgccAXmqmL6Ab/uwX9qloBrAAYHx+viYmJLd6c5o9Op4P9bBd72i72s13s52gx6Gtb3cHdl37tMmX/xubPO9k+f8+qZ/uHszguwGuq6u3boQZJkqR5wzX62lZfAw5MsnOSPYHf6OOYjwPP3/xiK0t3FgBLmu3fBz6zpRNX1Q+AG5Ic35x75yS7AhcBf5hkUTP+wCT79FGrJEnSvGbQ1zapqq8DZwPrmz+/sOUjAHgV8PNJ1ie5Cjh2C3N/CByZZD3w68Ar+zj/HwAvTLIW+G/gflX1MeA9wOeSrAPOAXbv41ySJEnzmkt3NCtVtahn+6XAS6eZs3/P9mpgotmeBJ41i2u9aJqxiSmvl/dsX0v3TcHUY04FTu33upIkSW3gE31JkiSphXyir6FKcjmw85ThP+j9zYEkSZJmz6CvoaqqRw67BkmSpDZy6Y4kSZLUQgZ9SZIkqYUM+pIkSVILGfQlSZKkFjLoS5IkSS1k0JckSZJayKCvkZPkfknel+S6JGuS/FeShyZZP+zaJEmS5gu/R18jJUmA84CzqurEZuxQ4L5DLUySJGme8Ym+Rs2xwO1Vdfrmgaq6Cvj65tdJdklyRpJ1Sb6Q5Nhm/OFJrkhyZZK1SR7SjD+jZ/ztSXYa9E1JkiQNmkFfo+YgYM1W5jwfqKo6GHg6cFaSXYDnAadW1WHAYuCGJL8CPA04uhm/EzhpzqqXJEkaES7d0Xz0aODNAFX1pSRfAx4KfA742yT7AudW1bVJfgM4AljVXRXEfYCbp54wyTJgGcDY2BidTmcQ96EBmJyctJ8tY0/bxX62i/0cLQZ9jZoNwJJtObCq3pPkcuAJwH8l+WMgdNf7v2wrx64AVgCMj4/XxMTEtpSgEdTpdLCf7WJP28V+tov9HC0u3dGouQTYuXnCDkCSQ4AH9cy5lGb5TZKHAr8IXJPkl4GvVNWbgPOBQ4CLgSVJ9mnm75Vkv4HciSRJ0hAZ9DVSqqqAE4Djmq/X3AC8Bvhmz7TTgAVJ1gErgaVVtRH4PWB9kivprvV/Z1VdDZwCfCzJWuDjwP0Hd0eSJEnD4dIdjZyq+gbd0D7VQc3+24BnT3Pca4HXTjO+ku4bAkmSpB2GT/QlSZKkFjLoS5IkSS1k0JckSZJayKAvSZIktZBBX5IkSWohg74kSZLUQgZ9SZIkqYUM+pIkSVILGfQlSZKkFjLoS5IkSS1k0NfISbJvkvOTXJvkuiSnJrn3sOuSJEmaTwz6GilJApwLfLCqHgI8FFgEvHqohUmSJM0zC4ddgDTFrwO3VdUZAFV1Z5K/AL6a5KvAbwJ7AA8E3lVVrwBI8gzghcC9gcuBP22OnQROBZ4I/Bh4clXdNOibkiRJGjSf6GvUPBxY0ztQVd8H/pfuG9MjgacAhwBPTbI4ya8ATwOOrqrDgDuBk5rDdwMuq6pDgU8Dzx3IXUiSJA2ZT/Q133y8qr4DkORc4NHAHcARwKruyh/uA9zczP8J8OFmew3w2OlOmmQZsAxgbGyMTqczR+Vr0CYnJ+1ny9jTdrGf7WI/R4tBX6PmamBJ70CSnwN+kW6grynzCwhwVlW9bJrz3V5Vm4+5kxn+zlfVCmAFwPj4eE1MTGxr/RoxnU4H+9ku9rRd7Ge72M/R4tIdjZqLgV2TPBMgyU7AvwBnAj8CHptkryT3AY4HPtscsyTJPs0xeyXZbxjFS5IkjQqDvkZK8/T9BLrr768FvgzcBvxNM+UK4APAWuADVbW6qq4GTgE+lmQt8HHg/gMvXpIkaYS4dEcjp6q+DvzO1PFm/f0NVXX8NMesBFZOM76oZ/sc4JztWqwkSdKI8om+JEmS1EI+0de8UVVn0l2rL0mSpK3wib4kSZLUQgZ9SZIkqYUM+pIkSVILGfQlSZKkFjLoS5IkSS1k0JckSZJayKAvSZIktZBBX5IkSWohg75GWpLJKa+XJnlLs/28JM/cyvE/nS9JkrQj8V/G1bxVVacPuwZJkqRR5RN9zVtJlid5cbP9iCRrk1yZ5PVJ1vdMfUCSC5Ncm+SfhlSuJEnSQPlEX6PuPkmu7Hm9F3DBNPPOAJ5bVZ9L8top+w4DfhXYCFyT5M1V9fW5KVeSJGk0GPQ16n5cVYdtfpFkKbC4d0KSPYHdq+pzzdB7gCf2TLm4qm5t5l4N7Ad8fco5lgHLAMbGxuh0Otv3LjQ0k5OT9rNl7Gm72M92sZ+jxaCvHcHGnu07mebvfVWtAFYAjI+P18TExGAq05zrdDrYz3axp+1iP9vFfo4W1+hr3quqW4AfJHlkM3TiMOuRJEkaBQZ9tcVzgHc06/l3A24dcj2SJElD5dIdjbSqWjTl9ZnAmc328p5dG6rqEIAkfw2snjq/ed27dl+SJKm1DPpqiyckeRndv9NfA5YOtxxJkqThMuirFapqJbBy2HVIkiSNCtfoS5IkSS1k0JckSZJayKAvSZIktZBBX5IkSWohg74kSZLUQgZ9SZIkqYUM+pIkSVILGfQlSZKkFjLoS5IkSS1k0Ne8kmRy2DVIkiTNBwZ9SZIkqYUM+pr3khyW5LIka5Ocl+Tnk+yTZE2z/9AkleQXm9fXJdl1uFVLkiTNrVTVsGuQ+pZksqoWTRlbC/xZVX0qySuBn6uqk5NsAB4FPBN4FvBG4DPA+6rqUVPOsQxYBjA2NnbE2WefPYC70SBMTk6yaNGirU/UvGFP28V+tov9HLxjjz12TVUtnm7fwkEXI21PSfYA9qyqTzVDZwHvb7b/GzgaeAzwj8BvAQEunXqeqloBrAAYHx+viYmJuS1cA9PpdLCf7WJP28V+tov9HC0u3VGbfRo4BtgPOB84FHg00wR9SZKktjHoa16rqluB7yU5phn6A2Dz0/1LgWcA11bVJuC7wOPpLt+RJElqNZfuaL7ZNckNPa//le76+9ObD9h+BXg2QFVdnyR0n+xDN+DvW1XfG2TBkiRJw2DQ17xSVTP9FuqoGeY/qGf7H+mu1ZckSWo9l+5IkiRJLWTQlyRJklrIoC9JkiS1kEFfkiRJaiGDviRJktRCBn1JkiSphQz6kiRJUgsZ9CVJkqQWMuhLkiRJLWTQlyRJklrIoK85l+STSX5zytjJSd42w/zrk+w9mOokSZLayaCvQXgvcOKUsRObcUmSJM0Bg74G4RzgCUnuDZBkf+ABwAOTrEuyPsnrph6UZP8k63tevzjJ8ma7k+QNSVYn+WKSRyQ5N8m1SV7Vc8wzklyR5Mokb0+y09zeqiRJ0mhYOOwC1H5V9d0kVwC/DZxP92n+J4DXAUcA3wM+luT4qvrgLE79k6panOTPm/MeAXwXuC7JG4B9gKcBR1fV7UlOA04C3jn1REmWAcsAxsbG6HQ623azGjmTk5P2s2XsabvYz3axn6PFoK9B2bx8Z3PQPw/oVNW3AJK8G3gMMJugf0Hz5zpgQ1Xd2JzrK8CDgEfTDf+rkgDcB7h5uhNV1QpgBcD4+HhNTEzMogyNsk6ng/1sF3vaLvazXeznaDHoa1DOB96Q5HBgV+BK4ICtHHMHd19etsuU/RubPzf1bG9+vRAIcFZVvWxbi5YkSZqvXKOvgaiqSeCTwH/Qfbp/BfD/kuzdrJt/OvCpKTKH0EkAABceSURBVIfdBOyT5BeS7Aw8cZaXvRhYkmQfgCR7JdnvntyHJEnSfOETfQ3Se+ku2Tmxqm5M8td0w3+Aj1TV+b2Tm3X1r6T7puD/gC/N5mJVdXWSU+iu/18A3A48H/jaPb8VSZKk0WbQ18A0H7RNz+v3Ms1XbFbV/j3bbwLeNM2ciZ7tDtCZYd9KYOU9q1ySJGn+cemOJEmS1EIGfUmSJKmFDPqSJElSCxn0JUmSpBYy6EuSJEktZNCXJEmSWsigL0mSJLWQQV+SJElqIYO+JEmS1EIGfUmSJKmFFg67AO0YktwPeCPwCOAW4Cbgg8CTquqJw6xNkiSpjXyirzmXJMB5QKeqDqiqI4CXAfcdbmWSJEntZdDXIBwL3F5Vp28eqKqrgEuBRUnOSfKlJO9u3hSQ5O+SrEqyPsmKnvFOktcluSLJl5Mc04zvmuTsJFcnOS/J5UkWN/sel+RzST6f5P1JFg38JyBJkjRgLt3RIBwErJlh368CDwe+AXwWOBr4DPCWqnolQJL/BJ4IfKg5ZmFVHZnk8cDfA8cBfwp8r6oOTHIQcGVz7N7AKcBxVfXDJH8FvAh4ZW8RSZYBywDGxsbodDrb4741AiYnJ+1ny9jTdrGf7WI/R4tBX8N2RVXdAJDkSmB/ukH/2CQvBXYF9gI2cFfQP7f5c00zH+DRwKkAVbU+ydpm/CjgQOCzzS8F7g18bmoRVbUCWAEwPj5eExMT2+v+NGSdTgf72S72tF3sZ7vYz9Fi0NcgbACWzLBvY8/2ncDCJLsApwGLq+rrSZYDu0xzzJ1s/e9wgI9X1dNnXbUkSdI85hp9DcIlwM7N8hgAkhwCHDPD/M2h/tvNevqZ3iT0+izwe825DwQObsYvA45O8uBm325JHjr7W5AkSZpfDPqac1VVwAnAcUmuS7IBeA3wzRnm3wK8A1gPXASs6uMypwFjSa4GXkX3twi3VtW3gKXAe5vlPJ8DHnbP7kiSJGn0uXRHA1FV36B54j7FO3rmvKBn+xS6H6Kdep6Jnu1vc9ca/duAZ1TVbUkOAD4BfK2Zdwnd7++XJEnaYRj01Ra7Ap9Mci+66/L/tKp+MuSaJEmShsagr1aoqh8Ai4ddhyRJ0qhwjb4kSZLUQgZ9SZIkqYUM+pIkSVILGfQlSZKkFjLoS5IkSS1k0JckSZJayKAvSZIktZBBX5IkSWohg74kSZLUQgZ9SZIkqYUWDrsAabaSLAeOAu5ohhYCl80wxnTjVbV8ELVKkiQNi0Ff89WJVXULQJI9gZNnGJtpriRJUqsZ9CUgyTJgGcDY2BidTme4BWm7mZyctJ8tY0/bxX62i/0cLQZ9CaiqFcAKgPHx8ZqYmBhuQdpuOp0O9rNd7Gm72M92sZ+jxQ/jSpIkSS1k0JckSZJayKAvSZIktZBBX5IkSWohP4yr+ehm4J1JNjWvFwAXzjDGFsYlSZJay6CveaeqTgNOm2bXdGNbGpckSWotl+5IkiRJLWTQlyRJklrIoC9JkiS1kEFfkiRJaiGDviRJktRCBn1JkiSphQz6kiRJUgsZ9CVJkqQWMuhr3kpyvyTvS3JdkjVJ/ivJQ5MsSPKmJOuTrEuyKskvNcdcn2TvYdcuSZI01/yXcdW3JMuBo4A7mqGFwGUzjDHH468AzgPOqqoTm/oOBe4LHAE8ADikqjYl2Rf44TbetiRJ0rxk0NdsnVhVtwAk2RM4eYaxmeZur/Fjgdur6vTNhVXVVc2cFwE3VtWmZvyG7fsjkCRJGn0Gfc1XBwFrZth3NvCZJMcAFwPvqqovbOlkSZYBywDGxsbodDrbsVQN0+TkpP1sGXvaLvazXeznaDHoq3Wq6oYk48CvN/+5OMlTq+riLRyzAlgBMD4+XhMTEwOpVXOv0+lgP9vFnraL/WwX+zlaDPqarzYAS2baWVUbgY8CH01yE3A83af7kiRJOwS/dUfz1SXAzs2SGwCSHJLkmCSHJ3lAM7YAOAT42pDqlCRJGgqDvualqirgBOC45us1NwCvAb4J7AN8KMl6YC3db+x5y9CKlSRJGgKX7mjeqqpvAL83za5rgQtnOGb/uaxJkiRpVBj0NRs3A+9Msql5vYBuoJ5ujAGMS5IkaQYGffWtqk4DTptm13RjgxiXJEnSDFyjL0mSJLWQQV+SJElqIYO+JEmS1EIGfUmSJKmFDPqSJElSCxn0JUmSpBYy6EuSJEktZNCXJEmSWsigr3kryf2SvC/JdUnWJPmvJA9NsiDJm5KsT7Iuyaokv9Qcc32SvYdduyRJ0lzzX8ZV35IsB44C7miGFgKXzTDGHI+/AjgPOKuqTmzqOxS4L3AE8ADgkKralGRf4IfbeNuSJEnzkkFfs3ViVd0CkGRP4OQZxmaau73GjwVur6rTNxdWVVc1c14E3FhVm5rxG7bvj0CSJGn0GfQ1Xx0ErJlh39nAZ5IcA1wMvKuqvrClkyVZBiwDGBsbo9PpbMdSNUyTk5P2s2XsabvYz3axn6PFoK/WqaobkowDv9785+IkT62qi7dwzApgBcD4+HhNTEwMpFbNvU6ng/1sF3vaLvazXeznaDHoa77aACyZaWdVbQQ+Cnw0yU3A8XSf7kuSJO0Q/NYdzVeXADs3S24ASHJIkmOSHJ7kAc3YAuAQ4GtDqlOSJGkoDPqal6qqgBOA45qv19wAvAb4JrAP8KEk64G1dL+x5y1DK1aSJGkIXLqjeauqvgH83jS7rgUunOGY/eeyJkmSpFFh0Nds3Ay8M8mm5vUCuoF6ujEGMC5JkqQZGPTVt6o6DThtml3TjQ1iXJIkSTNwjb4kSZLUQgZ9SZIkqYUM+pIkSVILGfQlSZKkFjLoS5IkSS1k0JckSZJayKAvSZIktZBBX5IkSWohg74kSZLUQv7LuBqYJMuBo4A7mqGFwGUzjDGX41W1/J7ejyRJ0igz6GvQTqyqWwCS7AmcPMPYTHO357gkSVJrGfQlIMkyYBnA2NgYnU5nuAVpu5mcnLSfLWNP28V+tov9HC0GfQmoqhXACoDx8fGamJgYbkHabjqdDvazXexpu9jPdrGfo8UP40qSJEktZNCXJEmSWsigL0mSJLWQQV+SJElqIT+Mq0G6GXhnkk3N6wXAhTOMMYBxSZKk1jLoa2Cq6jTgtGl2TTc2iHFJkqTWcumOJEmS1EIGfUmSJKmFDPqSJElSCxn0JUmSpBYy6EuSJEktZNCXJEmSWsigL0mSJLWQQV+SJElqodYG/SSTw65hS7alviSLk7xpLurpucbxSQ7cypylSR7Q8/rftnbMMCU5M8mSYdchSZI0SK0N+m1UVaur6oVzdf4kC4Hjga2F9qXAT4N+Vf1RVV09V3Vt1tQnSZKkPrQ+6Kfr9UnWJ1mX5GnN+KIkFyf5fDP+5GZ8/yRfTPKOJBuSfCzJfZp9z02yKslVST6QZNdm/MwkpydZneTLSZ7YjC9Ncn6STpJrk/z9NPW9M8nxPa/fvbmWaeZOJPlws708yX805/5Kkhc247sl+UhT4/qe+z0iyaeSrElyUZL7N+OdJG9Mshr4K+BJwOuTXJnkgGlqWAIsBt7dzLlPc47Fzf7J5ue9IcknkhzZU+OTmjk7NXNWJVmb5I+30L+JJJcmuQC4uhn7YHMfG5Is65k7meTVzb1fluS+05zvH5p+7TTTNSVJktpgR3hC+rvAYcChwN7AqiSfBr4FnFBV30+yN3BZEyYBHgI8vaqem+Rs4CnAu4Bzq+odAEleBTwHeHNzzP7AkcABwCeTPLgZPxI4CPhRc+2PVNXqnvr+HfgL4INJ9gB+DXhWn/f2MOBYYHfgmiRvA34L+EZVPaGpc48k92rqfHJVfasJ/68G/rA5z72ranNQfwjw4ao6Z7oLVtU5SV4AvHjzfSTpnbIbcElVvSTJecCrgMfS/S3BWcAFzc/t1qp6RJKdgc8m+VhVfXWG+zwcOKhn/x9W1XebN2Crknygqr7TXPuyqvrbJP8EPLe5Pk2dr29+Vs+uquq9QPOGYRnA2NgYnU5nhlI030xOTtrPlrGn7WI/28V+jpYdIeg/GnhvVd0J3JTkU8AjgI8C/5jkMcAm4IHA5ifAX62qK5vtNXRDPMBBTcDfE1gEXNRznbOrahNwbZKv0A3hAB9vQihJzm3q+WnQr6pPJTktyRjdNxQfqKo7+ry3j1TVRmBjkpub+tcB/5LkdXQD+6VJDqL7ZuPjTSjfCbix5zwr+7xeP34CXNhsrwM2VtXtSdZx18/xccAhuWvd/B5031zNFPSvmPIm4IVJTmi2H9Qc+53m2h9uxtfQfYOx2cuBy6tqGdOoqhXACoDx8fGamJjYym1qvuh0OtjPdrGn7WI/28V+jpYdIejP5CRgDDiiCaLXA7s0+zb2zLsTuE+zfSZwfFVdlWQpMNEz725PiHtezzTe653AM4ATgWf3fQc/W+fCqvpyksOBxwOvSnIxcB6woaoeNcN5fjiLa27N7T1PyzdtrrGqNuWuNfYB/qyqLpruBFuqL8kEcBzwqKr6UZIOd/Wt99p3cve/36uAI5LsVVXfneU9SZIkzTutX6MPXAo8rVkXPgY8BriC7lPkm5uQfyywXx/n2h24sVkKc9KUfU9NsqBZ1/7LwDXN+GOT7NUsMzke+Ow05z0TOBngnn6oNd1vw/lRVb0LeD3dZS/XAGNJHtXMuVeSh89wih/Qvc8t6WfOllwE/EnzcyTJQ5Ps1uexewDfa0L+w4Cj+jzuQuC1wEeS3JPaJUmS5oUd4Yn+ecCjgKvoPk1/aVV9M8m7gQ81S0pWA1/q41wvBy6nu77/cu4edv+X7huInwOeV1W3NctkrgA+AOwLvGvK+nwAquqmJF8EPrhtt3g3B9P9MO0m4HbgT6rqJ80ymTc1nwNYCLwR2DDN8e8D3tF8uHdJVV03zZwzgdOT/Jjuz3a2/o3uMp7Pp/tD+hbdN0H9uBB4XvPzuga4rN+LVtX7m5B/QZLHV9WPZ1e2JEnS/JEpn0nUNkhyJtN8gLVZ3rO4ql6wleN3pbue/fCqunWu6lR/xsfH65prrtn6RM0LrhdtH3vaLvazXezn4CVZs/lLVabaEZbujLQkxwFfBN5syJckSdL2siMs3ZlzVbV0hvEz6S5z2dKxn2DK5wOS/CbwuilTv1pVJzBASd4KHD1l+NSqOmMOrnUw8J9ThjdW1SO397UkSZJ2BAb9EdR8G02/30gzl3U8f4DXWkf33zuQJEnSduDSHUmSJKmFDPqSJElSCxn0JUmSpBYy6EuSJEktZNCXJEmSWsigL0mSJLWQQV+SJElqIYO+JEmS1EIG/RGQZHJA1zkzyZJm+9+SHDiLYxcneVOzPZHk1+aqzu2t974lSZJ2FP7LuDuoqvqjWc5fDaxuXk4Ak8B/b+eyJEmStJ2kqoZdww4vyWRVLUoS4J+A3wYKeFVVrUyyCDgf+HngXsApVXV+kv2BjwKfAX4N+D/gyVX14xmucybw4ao6J0kHeHFVrW5+o/A24PHAjcDfNHX8InByVV2QZAJ4MfAC4DLgTuBbwJ9V1aXTXOt3gFOAewPfAU5q5n8FOKyqbmnmXQs8GlgEvBvYrbnXk6tq0Qz3cX9gJfBzdN+s/klVXbr559jMWQI8saqWNvd9G7C4OeZFVfXhKedcBiwDGBsbO+Lss8+e7tKahyYnJ1m0aNq/Spqn7Gm72M92sZ+Dd+yxx66pqsXT7fOJ/mj5XeAw4FBgb2BVkk/TDcgnVNX3k+wNXJbkguaYhwBPr6rnJjkbeArwrlledzfgkqp6SZLzgFcBjwUOBM4CNl+Lqro+yenAZFX98xbO+RngqKqqJH8EvLSq/jLJ+cAJwBlJHgl8rapuSvLvwKlV9d4kz9tKvb8PXFRVr06yE7BrH/e4P3AkcADwySQPrqrbeu5rBbACYHx8vCYmJvo4peaDTqeD/WwXe9ou9rNd7OdocY3+aHk08N6qurOqbgI+BTwCCPCPSdYCnwAeCNy3OearVXVls72GbqCdrZ8AFzbb64BPVdXtzfa2nA9gX+CiJOuAlwAPb8ZXAk9rtk9sXgM8Cnh/s/2erZx7FfDsJMuBg6vqB33Uc3ZVbaqqa+n+VuFhfd2FJEnSPGXQnx9OAsaAI6rqMOAmYJdm38aeeXeybb+lub3uWsO1afM5q2rTNp4P4M3AW6rqYOCPuavezwEPTjIGHA+cO9sTV9WngcfQXap0ZpJnbt7VM22XqYdt5bUkSVKrGPRHy6XA05Ls1AThxwBXAHsAN1fV7UmOBfYbZpHAD4DdtzJnD7pBHOBZmwebNxTnAf8KfLGqvtPsuozusiPoPumfUZL9gJuq6h3AvwGHN7tuSvIrSRbQXR7U66lJFiQ5APhl4Jqt1C9JkjSvGfRHy3nAWuAq4BK669q/SfdDqoubZTDPBL40vBIB+BBwQpIrkxwzw5zlwPuTrAG+PWXfSuAZ3LVsB+Bk4EXN8qQHA7du4foTwFVJvkB3GdCpzfhfAx+m+21AN0455n/pvmn6KPC83vX5kiRJbeSHcUfA5m+KaZ52v6T5T+/+b9Ndwz6dg3rmbenDsVTV0p7tianXb7aXz1BbB+g0218GDtnKtc6n++050+1bTfdzB73+j7s+vHsiML6Fc59F90PCU8fPAc6ZZnzplmqVJElqI4O+RsURwFuarxi9BfjDIdcjSZI0rxn0WyjJW4GjpwyfWlVnzMG1/hZ46pTh91fVq2dznua7+A+dcu6Dgf+cMnVjVT1y1oVKkiTtYAz6LVRVzx/gtV4NzCrUz+Lc6+j+uwKSJEmaJf9lXGmKJD/Ab+Vpk7352Q+Ea36zp+1iP9vFfg7eflU1Nt0On+hLP+uamf4pac0/SVbbz3axp+1iP9vFfo4Wv15TkiRJaiGDviRJktRCBn3pZ60YdgHaruxn+9jTdrGf7WI/R4gfxpUkSZJayCf6kiRJUgsZ9LXDSvJbSa5J8j9J/nqa/TsnWdnsvzzJ/oOvUv3qo5+PSfL5JHckWTKMGtW/Pvr5oiRXJ1mb5OIk+w2jTvWnj34+L8m6JFcm+UySA4dRp/qztX72zHtKkkrit/AMiUFfO6QkOwFvBX4bOBB4+jT/x/Ic4HtV9WDgDcDrBlul+tVnP/8XWAq8Z7DVabb67OcXgMVVdQhwDvBPg61S/eqzn++pqoOr6jC6vfzXAZepPvXZT5LsDvw5cPlgK1Qvg752VEcC/1NVX6mqnwDvA548Zc6TgbOa7XOA30iSAdao/m21n1V1fVWtBTYNo0DNSj/9/GRV/ah5eRmw74BrVP/66ef3e17uBvgBwtHVz/9/AvwD3Qdktw2yON2dQV87qgcCX+95fUMzNu2cqroDuBX4hYFUp9nqp5+aP2bbz+cAH53TinRP9NXPJM9Pch3dJ/ovHFBtmr2t9jPJ4cCDquojgyxMP8ugL0mat5I8A1gMvH7Yteieqaq3VtUBwF8Bpwy7Hm2bJAvoLr36y2HXIoO+dlz/Bzyo5/W+zdi0c5IsBPYAvjOQ6jRb/fRT80df/UxyHPC3wJOqauOAatPszfa/n+8Djp/TinRPbK2fuwMHAZ0k1wNHARf4gdzhMOhrR7UKeEiSX0pyb+BE4IIpcy4AntVsLwEuKf/hiVHVTz81f2y1n0l+FXg73ZB/8xBqVP/66edDel4+Abh2gPVpdrbYz6q6tar2rqr9q2p/up+heVJVrR5OuTs2g752SM2a+xcAFwFfBM6uqg1JXpnkSc20fwd+Icn/AC8CZvwKMQ1XP/1M8ogkNwBPBd6eZMPwKtaW9Pnfz9cDi4D3N1/J6Bu7EdVnP1+QZEOSK+n+7+2zZjidhqzPfmpE+C/jSpIkSS3kE31JkiSphQz6kiRJUgsZ9CVJkqQWMuhLkiRJLWTQlyRJklrIoC9JkiS1kEFfkiRJaiGDviRJktRC/x8afQ7DFVIMqgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 720x1800 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "imp = pd.Series(data=clf.feature_importances_, index=X_train.columns.values).sort_values(ascending=False)\n",
    "\n",
    "plt.figure(figsize=(10, 25))\n",
    "sns.barplot(x=imp, \n",
    "            y=imp.index)\n",
    "plt.title(\"Feature importances from TabNet\")\n",
    "plt.grid(True)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "BFiv6b8g1k2k"
   },
   "source": [
    "### Save Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "DV0caiex1mRG"
   },
   "outputs": [],
   "source": [
    "try:\n",
    "    clf.save_model(\"../Model/clf_smote_pretrain_1\")\n",
    "    print()\n",
    "except:\n",
    "    with open(\"../Model/clf_smote_pretrain_1.pkl\", \"wb\") as f:\n",
    "        pickle.dump(clf, f, protocol=pickle.HIGHEST_PROTOCOL)"
   ]
  }
 ],
 "metadata": {
  "accelerator": "GPU",
  "colab": {
   "collapsed_sections": [
    "2AlF6scurEP6",
    "SlvZxiz2rJS2",
    "Q-nCsQk18-x3"
   ],
   "machine_shape": "hm",
   "provenance": []
  },
  "gpuClass": "standard",
  "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": "3.9.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}