Stock Market Analysis.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4601fe61",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Predicting Profitable Stocks using LSTM and Financial Data by Aryan Chopra"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "4c04a1b5",
   "metadata": {
    "collapsed": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Requirement already satisfied: pandas in ./anaconda3/lib/python3.10/site-packages (1.5.3)\n",
      "Requirement already satisfied: numpy in ./anaconda3/lib/python3.10/site-packages (1.23.5)\n",
      "Requirement already satisfied: matplotlib in ./anaconda3/lib/python3.10/site-packages (3.7.0)\n",
      "Requirement already satisfied: scikit-learn in ./anaconda3/lib/python3.10/site-packages (1.2.1)\n",
      "Requirement already satisfied: keras in ./anaconda3/lib/python3.10/site-packages (2.13.1rc0)\n",
      "Requirement already satisfied: python-dateutil>=2.8.1 in ./anaconda3/lib/python3.10/site-packages (from pandas) (2.8.2)\n",
      "Requirement already satisfied: pytz>=2020.1 in ./anaconda3/lib/python3.10/site-packages (from pandas) (2022.7)\n",
      "Requirement already satisfied: fonttools>=4.22.0 in ./anaconda3/lib/python3.10/site-packages (from matplotlib) (4.25.0)\n",
      "Requirement already satisfied: cycler>=0.10 in ./anaconda3/lib/python3.10/site-packages (from matplotlib) (0.11.0)\n",
      "Requirement already satisfied: pillow>=6.2.0 in ./anaconda3/lib/python3.10/site-packages (from matplotlib) (9.4.0)\n",
      "Requirement already satisfied: pyparsing>=2.3.1 in ./anaconda3/lib/python3.10/site-packages (from matplotlib) (3.0.9)\n",
      "Requirement already satisfied: contourpy>=1.0.1 in ./anaconda3/lib/python3.10/site-packages (from matplotlib) (1.0.5)\n",
      "Requirement already satisfied: kiwisolver>=1.0.1 in ./anaconda3/lib/python3.10/site-packages (from matplotlib) (1.4.4)\n",
      "Requirement already satisfied: packaging>=20.0 in ./anaconda3/lib/python3.10/site-packages (from matplotlib) (22.0)\n",
      "Requirement already satisfied: scipy>=1.3.2 in ./anaconda3/lib/python3.10/site-packages (from scikit-learn) (1.10.0)\n",
      "Requirement already satisfied: threadpoolctl>=2.0.0 in ./anaconda3/lib/python3.10/site-packages (from scikit-learn) (2.2.0)\n",
      "Requirement already satisfied: joblib>=1.1.1 in ./anaconda3/lib/python3.10/site-packages (from scikit-learn) (1.1.1)\n",
      "Requirement already satisfied: six>=1.5 in ./anaconda3/lib/python3.10/site-packages (from python-dateutil>=2.8.1->pandas) (1.16.0)\n"
     ]
    }
   ],
   "source": [
    "!pip install pandas numpy matplotlib scikit-learn keras"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "2c4718f3",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "from keras.layers import LSTM, Dense, Dropout\n",
    "from sklearn.model_selection import TimeSeriesSplit\n",
    "from sklearn.metrics import mean_squared_error, r2_score\n",
    "import matplotlib.dates as mandates\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "from sklearn import linear_model\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense\n",
    "import keras.backend as K\n",
    "from keras.callbacks import EarlyStopping\n",
    "from keras.optimizers import Adam\n",
    "from keras.models import load_model\n",
    "from keras.layers import LSTM\n",
    "from tensorflow.keras.utils import plot_model\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "230d75e5",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "('aau.us.csv', <http.client.HTTPMessage at 0x1664ac3d0>)"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import urllib.request\n",
    "\n",
    "url = 'https://storage.googleapis.com/kagglesdsdata/datasets/4538/7213/Stocks/aau.us.txt?X-Goog-Algorithm=GOOG4-RSA-SHA256&X-Goog-Credential=gcp-kaggle-com%40kaggle-161607.iam.gserviceaccount.com%2F20230717%2Fauto%2Fstorage%2Fgoog4_request&X-Goog-Date=20230717T141848Z&X-Goog-Expires=259200&X-Goog-SignedHeaders=host&X-Goog-Signature=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'\n",
    "local_filename = 'aau.us.csv'\n",
    "\n",
    "urllib.request.urlretrieve(url, local_filename)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "1384692f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>Volume</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2005-02-25</th>\n",
       "      <td>1.58</td>\n",
       "      <td>1.58</td>\n",
       "      <td>1.53</td>\n",
       "      <td>1.58</td>\n",
       "      <td>8570</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2005-02-28</th>\n",
       "      <td>1.64</td>\n",
       "      <td>1.64</td>\n",
       "      <td>1.54</td>\n",
       "      <td>1.64</td>\n",
       "      <td>11000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2005-03-01</th>\n",
       "      <td>1.53</td>\n",
       "      <td>1.61</td>\n",
       "      <td>1.53</td>\n",
       "      <td>1.53</td>\n",
       "      <td>5450</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2005-03-02</th>\n",
       "      <td>1.53</td>\n",
       "      <td>1.53</td>\n",
       "      <td>1.53</td>\n",
       "      <td>1.53</td>\n",
       "      <td>3333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2005-03-03</th>\n",
       "      <td>1.54</td>\n",
       "      <td>1.54</td>\n",
       "      <td>1.49</td>\n",
       "      <td>1.54</td>\n",
       "      <td>3200</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            Open  High   Low  Close  Volume\n",
       "Date                                       \n",
       "2005-02-25  1.58  1.58  1.53   1.58    8570\n",
       "2005-02-28  1.64  1.64  1.54   1.64   11000\n",
       "2005-03-01  1.53  1.61  1.53   1.53    5450\n",
       "2005-03-02  1.53  1.53  1.53   1.53    3333\n",
       "2005-03-03  1.54  1.54  1.49   1.54    3200"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df=pd.read_csv(\"aau.us.csv\",na_values=['null'],index_col='Date',parse_dates=True,infer_datetime_format=True)\n",
    "df = df.drop(df.columns[-1], axis=1)\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "0806eeb3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Open      3177\n",
       "High      3177\n",
       "Low       3177\n",
       "Close     3177\n",
       "Volume    3177\n",
       "dtype: int64"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.count()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "ca0434ee",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "DatetimeIndex: 3177 entries, 2005-02-25 to 2017-11-10\n",
      "Data columns (total 5 columns):\n",
      " #   Column  Non-Null Count  Dtype  \n",
      "---  ------  --------------  -----  \n",
      " 0   Open    3177 non-null   float64\n",
      " 1   High    3177 non-null   float64\n",
      " 2   Low     3177 non-null   float64\n",
      " 3   Close   3177 non-null   float64\n",
      " 4   Volume  3177 non-null   int64  \n",
      "dtypes: float64(4), int64(1)\n",
      "memory usage: 148.9 KB\n"
     ]
    }
   ],
   "source": [
    "df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "5ebceffc",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>Volume</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>3177.000000</td>\n",
       "      <td>3177.000000</td>\n",
       "      <td>3177.000000</td>\n",
       "      <td>3177.000000</td>\n",
       "      <td>3.177000e+03</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>1.774635</td>\n",
       "      <td>1.825294</td>\n",
       "      <td>1.718543</td>\n",
       "      <td>1.769668</td>\n",
       "      <td>1.869007e+05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.916737</td>\n",
       "      <td>0.942470</td>\n",
       "      <td>0.884982</td>\n",
       "      <td>0.913004</td>\n",
       "      <td>2.813646e+05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.330000</td>\n",
       "      <td>0.400000</td>\n",
       "      <td>0.330000</td>\n",
       "      <td>0.390000</td>\n",
       "      <td>1.950000e+02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>1.060000</td>\n",
       "      <td>1.100000</td>\n",
       "      <td>1.030000</td>\n",
       "      <td>1.060000</td>\n",
       "      <td>5.740000e+04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>1.500000</td>\n",
       "      <td>1.540000</td>\n",
       "      <td>1.460000</td>\n",
       "      <td>1.500000</td>\n",
       "      <td>1.099890e+05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>2.470000</td>\n",
       "      <td>2.530000</td>\n",
       "      <td>2.400000</td>\n",
       "      <td>2.460000</td>\n",
       "      <td>2.159960e+05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>5.130000</td>\n",
       "      <td>5.350000</td>\n",
       "      <td>4.770000</td>\n",
       "      <td>5.120000</td>\n",
       "      <td>5.136239e+06</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              Open         High          Low        Close        Volume\n",
       "count  3177.000000  3177.000000  3177.000000  3177.000000  3.177000e+03\n",
       "mean      1.774635     1.825294     1.718543     1.769668  1.869007e+05\n",
       "std       0.916737     0.942470     0.884982     0.913004  2.813646e+05\n",
       "min       0.330000     0.400000     0.330000     0.390000  1.950000e+02\n",
       "25%       1.060000     1.100000     1.030000     1.060000  5.740000e+04\n",
       "50%       1.500000     1.540000     1.460000     1.500000  1.099890e+05\n",
       "75%       2.470000     2.530000     2.400000     2.460000  2.159960e+05\n",
       "max       5.130000     5.350000     4.770000     5.120000  5.136239e+06"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "e9206140",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(3177, 5)\n"
     ]
    }
   ],
   "source": [
    "print(df.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "a14cdffc",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "False"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.isnull().values.any()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "f3f1fe25",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: xlabel='Date'>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df['Close'].plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "715a2b00",
   "metadata": {},
   "outputs": [],
   "source": [
    "output_var = pd.DataFrame(df['Close'])\n",
    "features = ['Open', 'High', 'Low', 'Volume']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "1f6f24c4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Volume</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2005-02-25</th>\n",
       "      <td>0.260417</td>\n",
       "      <td>0.238384</td>\n",
       "      <td>0.270270</td>\n",
       "      <td>0.001631</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2005-02-28</th>\n",
       "      <td>0.272917</td>\n",
       "      <td>0.250505</td>\n",
       "      <td>0.272523</td>\n",
       "      <td>0.002104</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2005-03-01</th>\n",
       "      <td>0.250000</td>\n",
       "      <td>0.244444</td>\n",
       "      <td>0.270270</td>\n",
       "      <td>0.001023</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2005-03-02</th>\n",
       "      <td>0.250000</td>\n",
       "      <td>0.228283</td>\n",
       "      <td>0.270270</td>\n",
       "      <td>0.000611</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2005-03-03</th>\n",
       "      <td>0.252083</td>\n",
       "      <td>0.230303</td>\n",
       "      <td>0.261261</td>\n",
       "      <td>0.000585</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                Open      High       Low    Volume\n",
       "Date                                              \n",
       "2005-02-25  0.260417  0.238384  0.270270  0.001631\n",
       "2005-02-28  0.272917  0.250505  0.272523  0.002104\n",
       "2005-03-01  0.250000  0.244444  0.270270  0.001023\n",
       "2005-03-02  0.250000  0.228283  0.270270  0.000611\n",
       "2005-03-03  0.252083  0.230303  0.261261  0.000585"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "scaler = MinMaxScaler()\n",
    "feature_transform = scaler.fit_transform(df[features])\n",
    "feature_transform= pd.DataFrame(columns=features, data=feature_transform, index=df.index)\n",
    "feature_transform.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "3f28a484",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>Volume</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Open</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.997757</td>\n",
       "      <td>0.997408</td>\n",
       "      <td>0.996172</td>\n",
       "      <td>0.199877</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>High</th>\n",
       "      <td>0.997757</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.996833</td>\n",
       "      <td>0.998370</td>\n",
       "      <td>0.225004</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Low</th>\n",
       "      <td>0.997408</td>\n",
       "      <td>0.996833</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.997695</td>\n",
       "      <td>0.182336</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Close</th>\n",
       "      <td>0.996172</td>\n",
       "      <td>0.998370</td>\n",
       "      <td>0.997695</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.209032</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Volume</th>\n",
       "      <td>0.199877</td>\n",
       "      <td>0.225004</td>\n",
       "      <td>0.182336</td>\n",
       "      <td>0.209032</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            Open      High       Low     Close    Volume\n",
       "Open    1.000000  0.997757  0.997408  0.996172  0.199877\n",
       "High    0.997757  1.000000  0.996833  0.998370  0.225004\n",
       "Low     0.997408  0.996833  1.000000  0.997695  0.182336\n",
       "Close   0.996172  0.998370  0.997695  1.000000  0.209032\n",
       "Volume  0.199877  0.225004  0.182336  0.209032  1.000000"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.corr()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "e4d629be",
   "metadata": {},
   "outputs": [],
   "source": [
    "timesplit= TimeSeriesSplit(n_splits=10)\n",
    "for train_index, test_index in timesplit.split(feature_transform):\n",
    "        X_train, X_test = feature_transform[:len(train_index)], feature_transform[len(train_index): (len(train_index)+len(test_index))]\n",
    "        y_train, y_test = output_var[:len(train_index)].values.ravel(), output_var[len(train_index): (len(train_index)+len(test_index))].values.ravel()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "2711128d",
   "metadata": {},
   "outputs": [],
   "source": [
    "trainX =np.array(X_train)\n",
    "testX =np.array(X_test)\n",
    "X_train = trainX.reshape(X_train.shape[0], 1, X_train.shape[1])\n",
    "X_test = testX.reshape(X_test.shape[0], 1, X_test.shape[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "c2189e0d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2889"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(X_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "2b6d64b8",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "You must install pydot (`pip install pydot`) and install graphviz (see instructions at https://graphviz.gitlab.io/download/) for plot_model to work.\n"
     ]
    }
   ],
   "source": [
    "lstm = Sequential()\n",
    "lstm.add(LSTM(32, input_shape=(1, trainX.shape[1]), activation='relu', return_sequences=False))\n",
    "lstm.add(Dense(1))\n",
    "lstm.compile(loss='mean_squared_error', optimizer='adam')\n",
    "plot_model(lstm, show_shapes=True, show_layer_names=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "9d58f9be",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Requirement already satisfied: pydot in ./anaconda3/lib/python3.10/site-packages (1.4.2)\n",
      "Requirement already satisfied: pyparsing>=2.1.4 in ./anaconda3/lib/python3.10/site-packages (from pydot) (3.0.9)\n",
      "Note: you may need to restart the kernel to use updated packages.\n"
     ]
    }
   ],
   "source": [
    "pip install pydot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "5db9b489",
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "os.environ[\"PATH\"] += os.pathsep + '/Users/aryanchopra/anaconda3/share/graphviz/bin'\n",
    "import pydotplus\n",
    "pydotplus.graphviz.graph_from_dot_data = pydotplus.graph_from_dot_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "41fa7e2c",
   "metadata": {
    "collapsed": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "lstm = Sequential()\n",
    "lstm.add(LSTM(32, input_shape=(1, trainX.shape[1]), activation='relu', return_sequences=False))\n",
    "lstm.add(Dense(1))\n",
    "lstm.compile(loss='mean_squared_error', optimizer='adam')\n",
    "\n",
    "plot_model(lstm, show_shapes=True, show_layer_names=True, to_file='model_plot.png')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "3409776c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[31mERROR: Could not find a version that satisfies the requirement python-graphviz (from versions: none)\u001b[0m\u001b[31m\n",
      "\u001b[0m\u001b[31mERROR: No matching distribution found for python-graphviz\u001b[0m\u001b[31m\n",
      "\u001b[0mNote: you may need to restart the kernel to use updated packages.\n"
     ]
    }
   ],
   "source": [
    "pip install python-graphviz"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "480b2aae",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Requirement already satisfied: pydotplus in ./anaconda3/lib/python3.10/site-packages (2.0.2)\n",
      "Requirement already satisfied: pyparsing>=2.0.1 in ./anaconda3/lib/python3.10/site-packages (from pydotplus) (3.0.9)\n",
      "Note: you may need to restart the kernel to use updated packages.\n"
     ]
    }
   ],
   "source": [
    "pip install pydotplus\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "2526c07a",
   "metadata": {
    "collapsed": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      "362/362 [==============================] - 0s 625us/step - loss: 0.0080\n",
      "Epoch 2/100\n",
      "362/362 [==============================] - 0s 596us/step - loss: 0.0073\n",
      "Epoch 3/100\n",
      "362/362 [==============================] - 0s 598us/step - loss: 0.0067\n",
      "Epoch 4/100\n",
      "362/362 [==============================] - 0s 600us/step - loss: 0.0062\n",
      "Epoch 5/100\n",
      "362/362 [==============================] - 0s 594us/step - loss: 0.0058\n",
      "Epoch 6/100\n",
      "362/362 [==============================] - 0s 633us/step - loss: 0.0054\n",
      "Epoch 7/100\n",
      "362/362 [==============================] - 0s 670us/step - loss: 0.0051\n",
      "Epoch 8/100\n",
      "362/362 [==============================] - 0s 663us/step - loss: 0.0048\n",
      "Epoch 9/100\n",
      "362/362 [==============================] - 0s 640us/step - loss: 0.0046\n",
      "Epoch 10/100\n",
      "362/362 [==============================] - 0s 610us/step - loss: 0.0044\n",
      "Epoch 11/100\n",
      "362/362 [==============================] - 0s 590us/step - loss: 0.0042\n",
      "Epoch 12/100\n",
      "362/362 [==============================] - 0s 601us/step - loss: 0.0041\n",
      "Epoch 13/100\n",
      "362/362 [==============================] - 0s 599us/step - loss: 0.0040\n",
      "Epoch 14/100\n",
      "362/362 [==============================] - 0s 585us/step - loss: 0.0038\n",
      "Epoch 15/100\n",
      "362/362 [==============================] - 0s 592us/step - loss: 0.0037\n",
      "Epoch 16/100\n",
      "362/362 [==============================] - 0s 599us/step - loss: 0.0036\n",
      "Epoch 17/100\n",
      "362/362 [==============================] - 0s 603us/step - loss: 0.0035\n",
      "Epoch 18/100\n",
      "362/362 [==============================] - 0s 595us/step - loss: 0.0034\n",
      "Epoch 19/100\n",
      "362/362 [==============================] - 0s 598us/step - loss: 0.0034\n",
      "Epoch 20/100\n",
      "362/362 [==============================] - 0s 592us/step - loss: 0.0033\n",
      "Epoch 21/100\n",
      "362/362 [==============================] - 0s 591us/step - loss: 0.0032\n",
      "Epoch 22/100\n",
      "362/362 [==============================] - 0s 602us/step - loss: 0.0032\n",
      "Epoch 23/100\n",
      "362/362 [==============================] - 0s 598us/step - loss: 0.0031\n",
      "Epoch 24/100\n",
      "362/362 [==============================] - 0s 601us/step - loss: 0.0031\n",
      "Epoch 25/100\n",
      "362/362 [==============================] - 0s 593us/step - loss: 0.0031\n",
      "Epoch 26/100\n",
      "362/362 [==============================] - 0s 592us/step - loss: 0.0030\n",
      "Epoch 27/100\n",
      "362/362 [==============================] - 0s 601us/step - loss: 0.0030\n",
      "Epoch 28/100\n",
      "362/362 [==============================] - 0s 598us/step - loss: 0.0030\n",
      "Epoch 29/100\n",
      "362/362 [==============================] - 0s 615us/step - loss: 0.0030\n",
      "Epoch 30/100\n",
      "362/362 [==============================] - 0s 631us/step - loss: 0.0029\n",
      "Epoch 31/100\n",
      "362/362 [==============================] - 0s 630us/step - loss: 0.0029\n",
      "Epoch 32/100\n",
      "362/362 [==============================] - 0s 612us/step - loss: 0.0029\n",
      "Epoch 33/100\n",
      "362/362 [==============================] - 0s 595us/step - loss: 0.0028\n",
      "Epoch 34/100\n",
      "362/362 [==============================] - 0s 601us/step - loss: 0.0028\n",
      "Epoch 35/100\n",
      "362/362 [==============================] - 0s 606us/step - loss: 0.0028\n",
      "Epoch 36/100\n",
      "362/362 [==============================] - 0s 599us/step - loss: 0.0028\n",
      "Epoch 37/100\n",
      "362/362 [==============================] - 0s 598us/step - loss: 0.0027\n",
      "Epoch 38/100\n",
      "362/362 [==============================] - 0s 603us/step - loss: 0.0027\n",
      "Epoch 39/100\n",
      "362/362 [==============================] - 0s 596us/step - loss: 0.0027\n",
      "Epoch 40/100\n",
      "362/362 [==============================] - 0s 599us/step - loss: 0.0027\n",
      "Epoch 41/100\n",
      "362/362 [==============================] - 0s 595us/step - loss: 0.0027\n",
      "Epoch 42/100\n",
      "362/362 [==============================] - 0s 596us/step - loss: 0.0026\n",
      "Epoch 43/100\n",
      "362/362 [==============================] - 0s 601us/step - loss: 0.0026\n",
      "Epoch 44/100\n",
      "362/362 [==============================] - 0s 596us/step - loss: 0.0026\n",
      "Epoch 45/100\n",
      "362/362 [==============================] - 0s 596us/step - loss: 0.0026\n",
      "Epoch 46/100\n",
      "362/362 [==============================] - 0s 600us/step - loss: 0.0026\n",
      "Epoch 47/100\n",
      "362/362 [==============================] - 0s 601us/step - loss: 0.0026\n",
      "Epoch 48/100\n",
      "362/362 [==============================] - 0s 598us/step - loss: 0.0026\n",
      "Epoch 49/100\n",
      "362/362 [==============================] - 0s 599us/step - loss: 0.0025\n",
      "Epoch 50/100\n",
      "362/362 [==============================] - 0s 603us/step - loss: 0.0025\n",
      "Epoch 51/100\n",
      "362/362 [==============================] - 0s 605us/step - loss: 0.0025\n",
      "Epoch 52/100\n",
      "362/362 [==============================] - 0s 593us/step - loss: 0.0025\n",
      "Epoch 53/100\n",
      "362/362 [==============================] - 0s 614us/step - loss: 0.0025\n",
      "Epoch 54/100\n",
      "362/362 [==============================] - 0s 605us/step - loss: 0.0025\n",
      "Epoch 55/100\n",
      "362/362 [==============================] - 0s 652us/step - loss: 0.0025\n",
      "Epoch 56/100\n",
      "362/362 [==============================] - 0s 679us/step - loss: 0.0025\n",
      "Epoch 57/100\n",
      "362/362 [==============================] - 0s 642us/step - loss: 0.0025\n",
      "Epoch 58/100\n",
      "362/362 [==============================] - 0s 624us/step - loss: 0.0024\n",
      "Epoch 59/100\n",
      "362/362 [==============================] - 0s 635us/step - loss: 0.0024\n",
      "Epoch 60/100\n",
      "362/362 [==============================] - 0s 618us/step - loss: 0.0024\n",
      "Epoch 61/100\n",
      "362/362 [==============================] - 0s 609us/step - loss: 0.0024\n",
      "Epoch 62/100\n",
      "362/362 [==============================] - 0s 619us/step - loss: 0.0024\n",
      "Epoch 63/100\n",
      "362/362 [==============================] - 0s 615us/step - loss: 0.0024\n",
      "Epoch 64/100\n",
      "362/362 [==============================] - 0s 597us/step - loss: 0.0024\n",
      "Epoch 65/100\n",
      "362/362 [==============================] - 0s 602us/step - loss: 0.0024\n",
      "Epoch 66/100\n",
      "362/362 [==============================] - 0s 592us/step - loss: 0.0024\n",
      "Epoch 67/100\n",
      "362/362 [==============================] - 0s 591us/step - loss: 0.0024\n",
      "Epoch 68/100\n",
      "362/362 [==============================] - 0s 609us/step - loss: 0.0024\n",
      "Epoch 69/100\n",
      "362/362 [==============================] - 0s 601us/step - loss: 0.0023\n",
      "Epoch 70/100\n",
      "362/362 [==============================] - 0s 599us/step - loss: 0.0023\n",
      "Epoch 71/100\n",
      "362/362 [==============================] - 0s 589us/step - loss: 0.0023\n",
      "Epoch 72/100\n",
      "362/362 [==============================] - 0s 583us/step - loss: 0.0023\n",
      "Epoch 73/100\n",
      "362/362 [==============================] - 0s 636us/step - loss: 0.0023\n",
      "Epoch 74/100\n",
      "362/362 [==============================] - 0s 651us/step - loss: 0.0023\n",
      "Epoch 75/100\n",
      "362/362 [==============================] - 0s 629us/step - loss: 0.0023\n",
      "Epoch 76/100\n",
      "362/362 [==============================] - 0s 632us/step - loss: 0.0023\n",
      "Epoch 77/100\n",
      "362/362 [==============================] - 0s 616us/step - loss: 0.0023\n",
      "Epoch 78/100\n",
      "362/362 [==============================] - 0s 663us/step - loss: 0.0023\n",
      "Epoch 79/100\n",
      "362/362 [==============================] - 0s 614us/step - loss: 0.0023\n",
      "Epoch 80/100\n",
      "362/362 [==============================] - 0s 604us/step - loss: 0.0023\n",
      "Epoch 81/100\n",
      "362/362 [==============================] - 0s 609us/step - loss: 0.0023\n",
      "Epoch 82/100\n",
      "362/362 [==============================] - 0s 598us/step - loss: 0.0023\n",
      "Epoch 83/100\n",
      "362/362 [==============================] - 0s 610us/step - loss: 0.0022\n",
      "Epoch 84/100\n",
      "362/362 [==============================] - 0s 650us/step - loss: 0.0022\n",
      "Epoch 85/100\n",
      "362/362 [==============================] - 0s 615us/step - loss: 0.0022\n",
      "Epoch 86/100\n",
      "362/362 [==============================] - 0s 601us/step - loss: 0.0022\n",
      "Epoch 87/100\n",
      "362/362 [==============================] - 0s 611us/step - loss: 0.0022\n",
      "Epoch 88/100\n",
      "362/362 [==============================] - 0s 612us/step - loss: 0.0022\n",
      "Epoch 89/100\n",
      "362/362 [==============================] - 0s 646us/step - loss: 0.0022\n",
      "Epoch 90/100\n",
      "362/362 [==============================] - 0s 679us/step - loss: 0.0022\n",
      "Epoch 91/100\n",
      "362/362 [==============================] - 0s 679us/step - loss: 0.0022\n",
      "Epoch 92/100\n",
      "362/362 [==============================] - 0s 642us/step - loss: 0.0022\n",
      "Epoch 93/100\n",
      "362/362 [==============================] - 0s 604us/step - loss: 0.0022\n",
      "Epoch 94/100\n",
      "362/362 [==============================] - 0s 600us/step - loss: 0.0022\n",
      "Epoch 95/100\n",
      "362/362 [==============================] - 0s 594us/step - loss: 0.0022\n",
      "Epoch 96/100\n",
      "362/362 [==============================] - 0s 591us/step - loss: 0.0022\n",
      "Epoch 97/100\n",
      "362/362 [==============================] - 0s 609us/step - loss: 0.0022\n",
      "Epoch 98/100\n",
      "362/362 [==============================] - 0s 598us/step - loss: 0.0022\n",
      "Epoch 99/100\n",
      "362/362 [==============================] - 0s 590us/step - loss: 0.0022\n",
      "Epoch 100/100\n",
      "362/362 [==============================] - 0s 597us/step - loss: 0.0022\n"
     ]
    }
   ],
   "source": [
    "history=lstm.fit(X_train, y_train, epochs=100, batch_size=8, verbose=1, shuffle=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "b2b79e45",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "9/9 [==============================] - 0s 485us/step\n"
     ]
    }
   ],
   "source": [
    "#Prediction\n",
    "y_pred= lstm.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "048eb057",
   "metadata": {
    "collapsed": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAkIAAAHFCAYAAAAe+pb9AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy88F64QAAAACXBIWXMAAA9hAAAPYQGoP6dpAADA8UlEQVR4nOydd3wb9f3/n3fakm3J246dYRKyQwZhQ0nYK4yU1UAhIaWldNACLaNAKbusQgv8yhcKYRUIZe8NAcIKISEhgUwnTuK9bW3d/f74nCTLK3YsDyWf5+Ohh3WfWx/J0ul176nouq4jkUgkEolEsgeiDvYEJBKJRCKRSAYLKYQkEolEIpHssUghJJFIJBKJZI9FCiGJRCKRSCR7LFIISSQSiUQi2WORQkgikUgkEskeixRCEolEIpFI9likEJJIJBKJRLLHIoWQRCKRSCSSPRYphCSSIcqiRYtQFCX2MJvNFBcXs2DBArZv3z4gcxg1ahTz58+PLX/00UcoisJHH33Uq+MsXbqU66+/noaGhg7rZs2axaxZs/o0z10h+lr+97//9et5ov/HZcuWdbtdWVkZF198MWPHjsXhcJCVlcWUKVO48MILKSsro7S0NOHz0N2jtLQ09voURWHRokWdnvOII45AURRGjRqV/BcukaQI5sGegEQi6Z5HH32U8ePH4/P5WLJkCbfeeisff/wxq1atwuVyDehcZsyYweeff87EiRN7td/SpUv529/+xvz58/F4PAnrHnjggSTOMDXZtm0bM2bMwOPxcNlllzFu3DgaGxtZs2YNixcvZtOmTRx44IF8/vnnCftdfPHFNDY28tRTTyWMFxYWUlpaCkB6ejr/+c9/EgQtwObNm/noo4/IyMjoz5cmkQx5pBCSSIY4kydPZubMmQDMnj2bSCTCjTfeyEsvvcQ555zT6T5erxen05n0uWRkZHDggQcm9Zi9FVW7Iw899BA1NTV89dVXlJSUxMZPPfVUrr76ajRNQ1XVDu99RkYGwWCw2//JWWedxcMPP8z69evZe++9Y+OPPPIIRUVFTJkyhTVr1iT/RUkkKYJ0jUkkKUb0R2/Lli0AzJ8/n7S0NFatWsUxxxxDeno6Rx55JADBYJCbbrqJ8ePHY7PZyM3NZcGCBVRXVyccMxQK8ec//5mCggKcTieHHnooX331VYdzd+Ua+/LLL5kzZw7Z2dnY7XZGjx7NH/7wBwCuv/56/vSnPwFQUlISc9dEj9GZa6yuro6LL76YoqIirFYre+21F3/5y18IBAIJ2ymKwm9/+1ueeOIJJkyYgNPpZOrUqbz22ms9fj/9fj+XXnopBQUFOBwODj/8cL799tvY+ieeeAJFUTpYYwBuuOEGLBYLO3bs6PH5OqO2thZVVcnLy+t0varu+qX66KOPZvjw4TzyyCOxMU3TeOyxxzj//PP7dGyJZHdAfgMkkhRjw4YNAOTm5sbGgsEgJ598MkcccQQvv/wyf/vb39A0jVNOOYXbbruNefPm8frrr3Pbbbfx7rvvMmvWLHw+X2z/Cy+8kDvvvJPzzjuPl19+mZ/+9KfMnTuX+vr6nc7n7bff5rDDDmPr1q3cfffdvPnmm1xzzTVUVlYC8Itf/ILf/e53ALzwwgt8/vnnfP7558yYMaPT4/n9fmbPns3jjz/OpZdeyuuvv865557L7bffzty5czts//rrr3Pfffdxww038Pzzz5OVlcVpp53Gpk2bevR+Xn311WzatImHH36Yhx9+mB07djBr1qzY/meddRYFBQXcf//9CfuFw2EefPBBTjvtNIYNG9ajc3XFQQcdhKZpzJ07l7fffpumpqY+Ha8tqqoyf/58Hn/8cSKRCADvvPMO27ZtY8GCBUk7j0SSsugSiWRI8uijj+qA/sUXX+ihUEhvbm7WX3vtNT03N1dPT0/XKyoqdF3X9fPPP18H9EceeSRh/6effloH9Oeffz5h/Ouvv9YB/YEHHtB1XdfXrl2rA/of//jHhO2eeuopHdDPP//82NiHH36oA/qHH34YGxs9erQ+evRo3efzdfla7rjjDh3QN2/e3GHd4Ycfrh9++OGx5X//+986oC9evDhhu7///e86oL/zzjuxMUDPz8/Xm5qaYmMVFRW6qqr6rbfe2uV82r6WGTNm6JqmxcZLS0t1i8Wi/+IXv4iN/fWvf9WtVqteWVkZG3v22Wd1QP/444+7PU/0//j11193uY2mafqvfvUrXVVVHdAVRdEnTJig//GPf+z0PYty+OGH65MmTer29T333HP6pk2bdEVR9Ndee03XdV0/44wz9FmzZum6rusnnniiPnLkyG5fg0SyOyMtQhLJEOfAAw/EYrGQnp7OSSedREFBAW+++Sb5+fkJ2/30pz9NWH7ttdfweDzMmTOHcDgce0ybNo2CgoKYa+rDDz8E6BBvdOaZZ2I2dx9GuG7dOjZu3MjChQux2+19fKWCDz74AJfLxemnn54wHg32ff/99xPGZ8+eTXp6emw5Pz+fvLy8mOtwZ8ybNw9FUWLLI0eO5OCDD469LwC//vWvARHLE+W+++5jypQp/OQnP+nZC+sGRVH497//zaZNm3jggQdYsGABoVCIf/zjH0yaNImPP/64T8cvKSlh1qxZPPLII9TW1vLyyy9zwQUX9HneEsnugAyWlkiGOI8//jgTJkzAbDaTn59PYWFhh22cTmeH7J/KykoaGhqwWq2dHrempgYQ8SkABQUFCevNZjPZ2dndzi0aa1RcXNyzF9MDamtrKSgoSBAnAHl5eZjN5th8o3Q2R5vNluD66472rzs6tnLlythyfn4+Z511Fg8++CBXXnkl33//PZ988gkPPvhgj87RU0aOHBkTXQCLFy/mZz/7GX/60586jdnqDQsXLmTBggXcfffdOByODkJTItlTkUJIIhniTJgwIZY11hXtRQNATk4O2dnZvPXWW53uE7WiRIVERUUFRUVFsfXhcLiD6GhPNE5p27Zt3W7XG7Kzs/nyyy/RdT3hdVVVVREOh8nJyUnauUC87s7G2gusSy65hCeeeIKXX36Zt956C4/H02XWXrI488wzufXWW1m9enWfjzV37lx+85vfcNttt3HhhRficDiSMEOJJPWRrjGJZDflpJNOora2lkgkwsyZMzs8xo0bBxDL2Gpfi2bx4sWEw+FuzzF27FhGjx7NI4880iGjqy02mw2gR1aaI488kpaWFl566aWE8ccffzy2Ppk8/fTT6LoeW96yZQtLly7tkMm27777cvDBB/P3v/+dp556ivnz5yetjlN5eXmn4y0tLZSVlfU5GBvA4XBw3XXXMWfOnASrk0SypyMtQhLJbsrZZ5/NU089xQknnMAll1zC/vvvj8ViYdu2bXz44YeccsopnHbaaUyYMIFzzz2Xe+65B4vFwlFHHcXq1au58847e1Rs7/7772fOnDkceOCB/PGPf2TEiBFs3bqVt99+OyaupkyZAsC9997L+eefj8ViYdy4cQmxPVHOO+887r//fs4//3xKS0uZMmUKn376KbfccgsnnHACRx11VFLfp6qqKk477TQuvPBCGhsb+etf/4rdbueqq67qsO0ll1zCWWedhaIoXHzxxb06zwcffBArctiWE044gZtvvpnPPvuMs846i2nTpuFwONi8eTP33XcftbW13HHHHbv68hK49NJLufTSS5NyLIlkd0EKIYlkN8VkMvHKK69w77338sQTT3DrrbfG2nQcfvjhMXEC8J///If8/HwWLVrEP//5T6ZNm8bzzz/P2WefvdPzHHvssSxZsoQbbriB3//+9/j9foqLizn55JNj28yaNYurrrqKxx57jIceeghN0/jwww87ba1ht9v58MMP+ctf/sIdd9xBdXU1RUVFXH755fz1r39NynvTlltuuYWvv/6aBQsW0NTUxP77788zzzzD6NGjO2x76qmnYrPZmD17dkJxwp5wxRVXdDq+efNmfv7znwPwzDPPcMcdd9DY2EhWVhb77rsvb7zxBscff3zvX5hEIukRit7WJiyRSCSSLnn11Vc5+eSTef311znhhBMGezoSiSQJSCEkkUgkO2HNmjVs2bKFSy65BJfLxfLlyzsNUJdIJKmHDJaWSCSSnXDxxRdz8sknk5mZydNPPy1FkESyGyEtQhKJRCKRSPZYpEVIIpFIJBLJHosUQhKJRCKRSPZYpBCSSCQSiUSyx7LH1RHSNI0dO3aQnp4uAx4lEolEIkkRdF2nubmZYcOGoarJs+PscUJox44dDB8+fLCnIZFIJBKJZBcoKytLaqPnPU4IRUv6l5WV9ah9gEQikUgkksGnqamJ4cOHd9qapy/scUIo6g7LyMiQQkgikUgkkhQj2WEtMlhaIpFIJBLJHosUQhKJRCKRSPZYpBCSSCQSiUSyx7LHxQhJJBKJZOgTiUQIhUKDPQ3JAGO1WpOaGt8TpBCSSCQSyZBB13UqKipoaGgY7KlIBgFVVSkpKcFqtQ7YOaUQkkgkEsmQISqC8vLycDqdsvDtHkS04HF5eTkjRowYsP+9FEISiUQiGRJEIpGYCMrOzh7s6UgGgdzcXHbs2EE4HMZisQzIOWWwtEQikUiGBNGYIKfTOcgzkQwWUZdYJBIZsHNKISSRSCSSIYV0h+25DMb/XgohiUQikUgkeyxSCEkkEolEIgFg1qxZ/OEPfxjsaQwoUghJJBKJRLKLKIrS7WP+/PkDMo85c+Zw1FFHdbru888/R1EUli9fPiBzSTVk1phEIklJ/KEIFpOKSZXxJJLBo7y8PPb82Wef5brrruPHH3+MjTkcjoTtQ6FQv2RDLVy4kLlz57JlyxZGjhyZsO6RRx5h2rRpzJgxI+nn3R2QFiGJRJJyeINhfnL7h5z78JeDPRXJHk5BQUHs4Xa7URQltuz3+/F4PCxevJhZs2Zht9t58sknuf7665k2bVrCce655x5GjRqVMPboo48yYcIE7HY748eP54EHHuhyHieddBJ5eXksWrQoYdzr9fLss8+ycOFCamtr+dnPfkZxcTFOp5MpU6bw9NNPd/v6FEXhpZdeShjzeDwJ59m+fTtnnXUWmZmZZGdnc8opp1BaWtrtcYcSUghJJJKUY0utl6rmAMu31g/2VCT9iK7reIPhQXnoup6013HFFVfw+9//nrVr13Lsscf2aJ+HHnqIv/zlL9x8882sXbuWW265hWuvvZbHHnus0+3NZjPnnXceixYtSpj7c889RzAY5JxzzsHv97Pvvvvy2muvsXr1an75y1/y85//nC+/3PUbCq/Xy+zZs0lLS2PJkiV8+umnpKWlcdxxxxEMBnf5uAOJdI1JJJKUo9kfBiAQ1tB1XaZb76b4QhEmXvf2oJx7zQ3H4rQm5yfyD3/4A3Pnzu3VPjfeeCN33XVXbL+SkhLWrFnDgw8+yPnnn9/pPhdccAF33HEHH330EbNnzwaEW2zu3LlkZmaSmZnJ5ZdfHtv+d7/7HW+99RbPPfccBxxwwC69tmeeeQZVVXn44Ydj38NHH30Uj8fDRx99xDHHHLNLxx1IpBCSSCQpR5Mv3owzENawW0yDOBuJpHtmzpzZq+2rq6spKytj4cKFXHjhhbHxcDiM2+3ucr/x48dz8MEH88gjjzB79mw2btzIJ598wjvvvAOIIoW33XYbzz77LNu3bycQCBAIBHC5XLv2woBvvvmGDRs2kJ6enjDu9/vZuHHjLh93IJFCSCKRpBxNfimE9gQcFhNrbuiZK6k/zp0s2gsNVVU7uN6iVbVB9NwC4R5rb6kxmbqf18KFC/ntb3/L/fffz6OPPsrIkSM58sgjAbjrrrv4xz/+wT333MOUKVNwuVz84Q9/6NaFpSjKTue677778tRTT3XYNzc3t9u5DhWkEJJIJClHgkUoFAHHwPQkkgwsiqIkzT01lMjNzaWioiLBrbtixYrY+vz8fIqKiti0aRPnnHNOr4595plncskll/Df//6Xxx57jAsvvDB2jk8++YRTTjmFc889FxAiZv369UyYMKHbubbNjFu/fj1erze2PGPGDJ599lny8vLIyMjo1VyHCjJYWiKRpBzRGCEQFiGJJJWYNWsW1dXV3H777WzcuJH777+fN998M2Gb66+/nltvvZV7772XdevWsWrVKh599FHuvvvubo+dlpbGWWedxdVXX82OHTsS6hiNGTOGd999l6VLl7J27Vp+9atfUVFR0e3xjjjiCO677z6WL1/OsmXLuOiiixLS/8855xxycnI45ZRT+OSTT9i8eTMff/wxl1xyCdu2bev9mzMISCEkkUhSjrauMX9o4JozSiTJYMKECTzwwAPcf//9TJ06la+++iohiBngF7/4BQ8//DCLFi1iypQpHH744SxatIiSkpKdHn/hwoXU19dz1FFHMWLEiNj4tddey4wZMzj22GOZNWsWBQUFnHrqqd0e66677mL48OH85Cc/Yd68eVx++eUJTXGdTidLlixhxIgRzJ07lwkTJnDBBRfg8/lSxkKk6MnMEUwBmpqacLvdNDY2psw/SSKRJHLF/77j2WVlALz2u0OZXNR1AKkkdfD7/WzevJmSkhLsdvtgT0cyCHT3Geiv329pEZJIJCmHtAhJJJJkIYWQRCJJOWSMkEQiSRZSCEkkkpSjyR/ChQ8TEWkRkkgkfWL3y0uUSCS7PdbWCj6y/ZEtegFV4VcGezoSiSSFkRYhiUSScvzC/wi5ShMz1XUEAr7Bno5EIklhpBCSSCQphd64nSO0eJNIU0vlIM5GIpGkOlIISSSSlCK89H6sSjxY2tRS3s3WEolE0j1SCEkkkpRCq1ybsGz1dl8ZVyKRSLpDCiGJRJJSRMKhhGWbV1qEJBLJriOFkEQiSSnCYeEW26bnAOD0yxghiWQg+Oijj1AUhYaGhsGeSlKRQkgikaQUWkRYhLZLISQZIsyfP7/bnl3ffvstJ510Enl5edjtdkaNGsVZZ51FTU0N119/PYqidPsoLS2NbXfcccd1OP7tt9+OoijMmjWr0/N/8803KIrCp59+2un6Y489lpNPPnlXXvpugRRCEokkpYhERAHFbXouAGnBqsGcjkTSLVVVVRx11FHk5OTw9ttvs3btWh555BEKCwvxer1cfvnllJeXxx7FxcXccMMNCWPDhw8HoLCwkA8//LBDV/dHH300oblqe/bdd1+mTp3Ko48+2mFdWVkZ7733HgsXLkzuC08hpBCSSCQphW5YhKKusYxg9WBORyLplqVLl9LU1MTDDz/M9OnTKSkp4YgjjuCee+5hxIgRpKWlUVBQEHuYTCbS09M7jAHk5eVxzDHH8NhjjyUcv6amhhNPPLHbeSxcuJDFixfT2tqaML5o0SJyc3M58cQTefLJJ5k5c2bs/PPmzaOqqusbjeuvv55p06YljN1zzz2MGjUqYezRRx9lwoQJ2O12xo8fzwMPPNCDd27gkEJIIpGkFJphEYq6xjLCtRAJd7eLJFXRdQi2Ds5D15PyEgoKCgiHw7z44ovoSTjmBRdcwKJFi2LLjzzyCOeccw5Wq7Xb/c455xxCoRDPPfdcbEzXdRYtWsT555+P2WwmGAxy4403snLlSl566SU2b97M/Pnz+zTfhx56iL/85S/cfPPNrF27lltuuYVrr702QcwNNrLFhkQiSSl0TQihRlM2Id2ERYlASyW4iwZ5ZpKkE/LCLcMG59xX7wCrq8+HOfDAA7n66quZN28eF110Efvvvz9HHHEE5513Hvn5+b0+3kknncRFF13EkiVL2HfffVm8eDGffvopjzzySLf7ZWVlceqpp/Loo4/GxM1HH33Epk2buOCCCwBifwH22msv/vnPf7L//vvT0tJCWlpar+cKcOONN3LXXXcxd+5cAEpKSlizZg0PPvgg559//i4dM9lIi5BEIkkpdE1Yf9KcTirJFINNOwZxRhJJ99x8881UVFTw73//m4kTJ/Lvf/+b8ePHs2rVql4fy2KxcO655/Loo4/y3HPPMXbsWPbZZ58e7btw4UKWLFnChg0bAGFNOuSQQxg3bhwggrpPOeUURo4cSXp6eiz4euvWrb2eJ0B1dTVlZWUsXLiQtLS02OOmm25i48aNu3TM/kBahCQSSUoRtQilOe1UeLMoVmqgWQqh3RKLU1hmBuvcSSQ7O5szzjiDM844g1tvvZXp06dz55137pKL6IILLuCAAw5g9erVCVacnXHUUUcxcuRIFi1axJ///GdeeOEF7rvvPgBaW1s55phjOOaYY3jyySfJzc1l69atHHvssQSDwU6Pp6pqB3dfKBSv86VpGiDcYwcccEDCdtG4p6GAFEISiSSlUAyLULrDRoWeJQalRWj3RFGS4p4aalitVkaPHt0hcLmnTJo0iUmTJvHdd98xb968Hu+nKAoLFizg4Ycfpri4GFVVOfPMMwH44YcfqKmp4bbbbotlqS1btqzb4+Xm5lJRUYGu6yiKAsCKFSti6/Pz8ykqKmLTpk2cc845vXyVA4cUQhKJJKVQdHGXabdaqNSla0wyNGhsbEwQASDicr777jueeeYZzj77bMaOHYuu67z66qu88cYbnaaz95QPPviAUCiEx+Pp1X4LFizghhtu4Oqrr+bss8/G5RJCc8SIEVitVv71r39x0UUXsXr1am688cZujzVr1iyqq6u5/fbbOf3003nrrbd48803ycjIiG1z/fXX8/vf/56MjAyOP/54AoEAy5Yto76+nksvvbTXr7s/kDFCEokkpVARQshqtVKrGxdcb+0gzkgiEYHH06dPT3hcd911TJw4EafTyWWXXca0adM48MADWbx4MQ8//DA///nPd/l8Lper1yIIhOA56qijqK+vT3Cr5ebmsmjRIp577jkmTpzIbbfdxp133tntsSZMmMADDzzA/fffz9SpU/nqq6+4/PLLE7b5xS9+wcMPP8yiRYuYMmUKhx9+OIsWLaKkpKTXc+8vFD0Z+XwpRFNTE263m8bGxgTVKpFIUoPaG0aTrdXw2JTHWLP8U/5ueQj2PhbOWTzYU5P0Eb/fz+bNmykpKcFutw/2dCSDQHefgf76/ZYWIYlEklKoiGBpm9VGnZ4uBr01gzgjiUSSykghJJFIUgpVF0LIbjXHXWOtUghJJJJdQwohiUSSUkRjhOw2K3UIi5AuY4QkEskuIoWQRCJJKWIWIZuNWt0NgBJsgZBvMKclkUhSFCmEJBJJShG1CDlsVppxENSNwmzSPbbbsIfl8EjaMBj/eymEJBJJSmEyhJDNakVRFOqIptBLIZTqWCwWALxe7yDPRDJYRKtYD2TlaVlQUSKRpBTRrDGT2YzNrFKnZ1Cg1EOrjBNKdUwmEx6Ph6qqKgCcTmesYrFk90fTNKqrq3E6nZjNAydPpBCSSCSpg65jNixCJtWEzWyiNixT6HcnCgoKAGJiSLJnoaoqI0aMGFABLIWQRCJJHYz2GgCq2YLdolIXjqbQVw/SpCTJRFEUCgsLycvLS2jgKdkzsFqtqOrARu1IISSRSFIHo/M8gMlkFhYhWUtot8RkMg2pDuWS3ZdBDZZesmQJc+bMYdiwYSiKwksvvbTTfQKBAH/5y18YOXIkNpuN0aNH88gjj/T/ZCUSyeCjx4VQ1CIU7zcmhZBEIuk9g2oRam1tZerUqSxYsICf/vSnPdrnzDPPpLKykv/85z+MGTOGqqoqwuFwP89UIpEMCbT4d91sEjFC0aKKMlhaIpHsCoMqhI4//niOP/74Hm//1ltv8fHHH7Np0yaysrIAGDVqVD/NTiKRDDnauMZUkxEjJC1CEomkD6RUHaFXXnmFmTNncvvtt1NUVMTYsWO5/PLL8flkRVmJZI+gjRAyW6IxQsIiFGySWUYSiaT3pFSw9KZNm/j000+x2+28+OKL1NTUcPHFF1NXV9dlnFAgECAQCMSWm5qaBmq6Eokk2bSJETKpJuwWlR1GQcVAYxXhYBinNaUuaxKJZJBJKYuQpmkoisJTTz3F/vvvzwknnMDdd9/NokWLurQK3Xrrrbjd7thj+PDhAzxriUSSLLSwSKcO6SZMqpKQNZau+KhvahnM6UkkkhQkpYRQYWEhRUVFuN3u2NiECRPQdZ1t27Z1us9VV11FY2Nj7FFWVjZQ05VIJEkmYgRLa6iYVRWzSaEJJ2FdXMpCzdI9JpFIekdKCaFDDjmEHTt20NISv+tbt24dqqpSXFzc6T42m42MjIyEh0QiSU20iHCNhVExmRSWldajo1JPmhhvkZljEomkdwyqEGppaWHFihWsWLECgM2bN7NixQq2bt0KCGvOeeedF9t+3rx5ZGdns2DBAtasWcOSJUv405/+xAUXXIDD4RiMlyCRSAaQsOEai6BiVhXOP3gkAA1GwLQmhZBEIuklgyqEli1bxvTp05k+fToAl156KdOnT+e6664DoLy8PCaKANLS0nj33XdpaGhg5syZnHPOOcyZM4d//vOfgzJ/iUQysGhhYRHSUDGpCucfPIrnf30wfotwl2veusGcnkQiSUEGNb1i1qxZ6Lre5fpFixZ1GBs/fjzvvvtuP85KIpEMVSLhIABhTJgUBYtJZd+RmXxhyoAI0IkQCoQj2MyyVYNEIumclIoRkkgkezaaFrcIqWq8O3WrSViEFF+iELrnvXXsc/07rN7eOHCTlEgkKYUUQhKJJGWIRETWWKTdpctnFkJIbSeEvthUSyCsSSEkkUi6RAohiUSSMmjhaPp8oqsrYMQImQL1CeMtAbF9IKwNwOwkEkkqIoWQRCJJGaIWIa3dpStgzQTAHGhgQ1ULb39fAUCLX2zvD0WQSCSSzpBCSCKRpAxa1DWmJF66QlYPALZgPZcuXsGvnviGHyuaaQkIASQtQhKJpCtkUx6JRJIyaJHOXWNhmwcAa6iRCr8fgB2NPloCou6QtAhJJJKukEJIIpGkDFrEKKjYziIUsQvXmD3USGtIiKX61iD+kLAESYuQRCLpCukak0gkKUM0fV5vZxGKOLIBcESaCYREraHyRn9svbQISSSSrpBCSCKRpAzRXmNaO4uQYvfEnrt10Ytwe4MvNiYtQhKJpCukEJJIJCmDbrjG2scI2WxWGnUnAB5FCKEdbYSQtAhJJJKukEJIIpGkDF1ZhOwWE/VG49VMmgEob4i7xqIWofrW4EBMUyKRpBBSCEkkkpShq6wxu0WlHkMIRS1CjYkWoae+3ML0G9/l5RXbB2i2EokkFZBCSCKRpAy6JoSQrrQTQmYT9XoaAJmKsAg1G8UUAaPNRhMA325tGICZSiSSVEEKIYlEkjJEXWN6e9eY1UQ9hhCipcN+gVAEX1AIo5qWQD/PUiKRpBJSCEkkkpQhahHSOrEINehx15iNILeYH+IY9WsA/CENb1CIqNoWGSckkUjiSCEkkUhSBj1aR6i9ELKo1LUJlj5R/YJ55g+5zvIEAIFwBJ+ROVbbKi1CEokkjqwsLZFIUgY90oVFyGJihy6KKo5Ty9BQAChWasinDn/ITqvRiV5ahCQSSVukRUgikaQMUdcY7YSQw2JiqTYJgH2UTcwyrYytm6GuJxCOYPHX8mvTKwS9DUQ0fcDmLJFIhjZSCEkkkpQhahHqECxtMVFJFj9qxaiKTqFSF1s3Q12PP6RxuvdZrrA8w29NL1In6wlJJBIDKYQkEknK0F2MEMAn2pQO+0QtQqMipQAcrH4v44QkEkkMKYQkEknK0LUQEsufaPvExpZEhCiarGzGrIcYqYlCipOULTTU1QzEdCUSSQoghZBEIkkdYgUVE/M8bGZxKftSG09AtwDwQuQwavQMbEqYg9Q15Cn1AKiKDluWDuCkJRLJUEYKIYlEkjJELUKo7brPKwp2i4ofGw9GTmSZNpb3tRl8o40F4GzTBwnbu8q/HJD5SiSSoY8UQhKJJHXowiIEcffY3eEzOT14Pc04+VKbAMDR6jcAhHVxycut/WogZiuRSFIAKYSSRbAVVj8PK58Z7JlIJLsvUYuQ0vHSZTfH44Zy0mwAfGEIIbMius9/oE0HIK91Hfib+nOmEokkRZBCKFn4G+F/F8BLF4Mua5RIJP1B3DVm6rAumjkGUOi2A7BWH0Gj7oqNL9UmsU3PQUWDHd/272QlEklKIIVQsrA4xV89AhFZo0Qi6Rd0wzWmdu0aAygwhBCKykrTpNj4Rn0YK7W9xMKO5f03T4lEkjJIIZQsrPG7TkLewZuHRLI7owkXV6eusTZCKGoRSrOaWW2J1xbaqA3jO220WNguhZBEIpFCKHmYLKCKtF2CUghJJP2C1p1FKH45i1qE0uxm1tinAdCkO2mx5fGdHrUISdeYRCKRTVeTi9UpYoWkRUgi6R90ESOkdBoj1NEi5LKZKbeP4bLqi6gkk6L8NFaVl6DpCmpjGbRUQ1ruwMxdIpEMSaRFKJlE44SCrYM7D4lkdyWWNdaJEDKyxlQF9h2Rhc2sMmOEB7tF5XntJ3yqTSEv3YZmTWeTXih2knFCEskejxRCySQqhKRFSCLpF5RY1lhHY7bDKoSQy2ZmRLaT5dcezd9/ug+2Nmn1LpuJkdkuVkbdYzJOSCLZ45FCKJlYoxYhKYQkkn5B33n6vMsqRJLLZo5VnI7isJgpyXHyvVYiBqq+79/5SiSSIY+MEUomFiNzLCRdYxJJf6AYwdKdCaGo5cdpM3U6DuC0msjPsLFRzxEDTeX9M1GJRJIySItQMpEWIYmkf9FF+rzSTR2hqEUoPh6/zDmtJkbluKjQM8VAsxRCEsmejhRCySQWIyQtQhJJf6AYBRU7zxozXGPdWIQcVhMlOS4q9Cwx0FwRD8AeivjqB3sGEslujxRCySRaVFFahCSS/sGwCHUaLN2FRcjWziI0MttJDW4iuiJijlqr+2++feHr/8DfR8Gq/w32TCSS3RophJJJzCLkG9x5SCS7KdGssc4sQvkZonZQrL2GQaJFyExumg271Uo1HjHYtKN/JttXohltpZ8M7jwkkt0cGSydTKwyWFoi6U+UaNaYqaMQOmFKIXaLygEl2QnjCTFCFhOKoog4oZpMCpT6oRsnFGgUf+tLB3UaEsnujrQIJROLDJaWSPqTqBBSO3GNWc0qx00uJNNlTRhvnzUGMCrHRWU0TmioWoT8UghJJAOBFELJxCoLKkok/YnSTYuNrkioI2QIoZJsF5VDPXPM3yT+NpRBJDS4c5FIdmOkEEomssWGRNKvxIVQz736iRYhsV9CCv1QrSUUMISQHoHGbYM7F4lkN0YKoWQiW2xIJP2KGhVCncQIdUX7OkIAo7KdcddY81B1jTXFn0v3mETSb0ghlExkQUWJpF9Ruimo2BXt6wiBYRFCWIS0oW4RAqjfPHjzkEh2c6QQSiayxYZE0q+oRC1CPRdCnVmEsl1Wmi25AOhDUQiF/BAJxpelRUgi6TekEEom0iIkkfQraixrrOeusYQYIYsQUIqiYM8aDoAp2DT04vqiGWNRpBCSSPoNKYSSScwiJIWQRNIfxFxju2gRirrGAPLzcmnRjeKLQ80q1NYtBlAnXWMSSX8hhVAyScX0+e3L4dETYNuywZ6JRLJT1G7qCHVF1CJkVhWs5vglb1S2kx26UXyxYQtfl9Zx1Qvf0RoIJ2/CXfD+8w/yzlN3d72Bv50Qqi8FXe/XOUkkeyqysnQyScWCiqufhy2fwarnoHjmYM9GIukWld5bhPLSbVhMCkUeR8L4qGwXm/RhjGU7kep1nPGyH4Cx+eksOKQkeZNuR6vPz6Hf/QULYWqq5pGTV9Bxo2hV6ay9oG6TsBD5G8CR2W/zkkj2VKRFKEmsq2zmrEe/EwthH2ja4E6op/gbxN9A86BOQyLpCVGLkKkXQijTZeXFiw/hiYUHJIyPynGxUS8EYPv6lbHxUKR/v7tbK6qwKSFURSfUXNX5RlGLUFpB/AbL19Cv85JI9lSkEEoShW4739e0MamninssKoCCLYM7D4mkB8SzxnoeLA0wucjN8CxnwlhJjouN2jAAGsq+j433sw5iR2VN7Hm4pa7zjYxgad2WDnZ3wphEIkkuUggliXS7hckj8+MDqSKEoneeASmEJEMf1QiW7o1FqCsynRYqrCMAKAhuZZ7pfd63Xoa5qazPx+6Oypo2Qshb3/lGRrD0Kz+20qIYSRhSCEkk/YIUQklk1vh8vLpNLAy1dNyukBYhSQoRtQipSRBCiqIwZaqIi8tTGrjMvJjRajnFNUv6fOzuqKmtjT2PtHYhhIwblAbNwZZW47VKISSR9AtSCCWR2ePy8CKEUMCXIsIiIC1CktQhGiytmixJOd5Vpx2Ani7ihLIVcVNg93cRt5MkGhri7jCtq7gf43vZjJPKoJHiL4WQRNIvSCGURMbmpxFUxEVrdekQ7V/UnphFSAZLS4Y+pphFqHcxQt2h5OydsOwMVCft2J3R1NgQX/A3dLqNboieZt1JE0ZskxRCEkm/IIVQElEUBdUm/Pk/bq0c5Nn0EBkjJEkh4jFCybEIAZAzNmExPdh/QqjZH0JrUyNI6UIIBVqEy6wZJ026yxjrIrBaIpH0CSmEkkxUCLW2NO1kyyGAFon3RZMxQpIUIO4aS55FqL0QygjVdLFh39lS68Wl+GPLahdWnmBLAwBWVyYBcxoAvuYu4okkEkmfkEIo2Rg1P4L+FBAWbcv4R4IQDna9rUQyBDAbQshkTmIt2OEHAApeVzEA7kht99v3gc01raQRF0KmYOdCKOIT4y53Fpo1A4BQV4HVEomkT0ghlGSiFqGgr5ussUj/l/DvEe2LKEqrkGSIE7MIJVMIDZsGF33KqiOfACBdb4GQr9tdPvyhil8+vozalkCvTlVa04pLiR/bHOzCcmzcpHgys9GNOkJdBlZLJJI+IYVQkjHbhRk70pVF6LN/wq3FUPbVAM6qC6QQkqQQmqbHgqXNyYwRAiiYjJo5Ml7+orn7JqyPLi3lnTWVvL+2dxlmVc0BXG0sQtZQ50LIEhLfzezsHFSHRwzKYGmJpF+QQijJWB1CCBFsRdM6aZK4+nnRgmPL0oGdWGe0b+woA6YlQ5iwpmOKWoR60XS1pzisZip1j1horuh2W6/RmLXe2zt3ciAcIY24Rcga7iRbU9dxaOK7mJ+bh9kl+oup7TvSSySSpCCFUJKxOtMBsBGgyR9KWOf3edErjVL+viGQASItQpIUIhLRMCtGjJAl+ULIaTVRhdHUtKn78hf+sLBM1XtD3W7XnmBYSwiWtoc7ihst0BoTfEWFhdjSxJwsXViPJBJJ3xhUIbRkyRLmzJnDsGHDUBSFl156qcf7fvbZZ5jNZqZNm9Zv89sVzA7hzy9Ryqlpid8t7mjwcd6tj6JoxoXTOxSEUHuLkKwlJBm6hNvE1iWjxUZ7nFYzFXqWWNiJRcgfEkKl0dc7i1AwoiVYhOyat0PMYE2tyFoL6ypFuVk4M8ScbGF5oyKR9AeDKoRaW1uZOnUq9913X6/2a2xs5LzzzuPII4/sp5n1gfEnAXC8+hUt29fGhj9eV83ekQ3x7XxDIAOkvRCSFiHJEEaLRGLPTeYkxwgBDquJSl1YXyJN3ccI+UNiLg29tgjpCRYhcbDE2J8dFeLcrYoLi9lEmjtHzE/vKJokEknfGVQhdPzxx3PTTTcxd+7cXu33q1/9innz5nHQQQf108z6wLBpLLMdgEnRyVn+z9jwt1vrmaJsim83FCxCMkZIkkKE25R3MKlJrCNk4LSaYjFCkcaduMZCUddY7y1CbYOlxcEaEhbrKkXT12ZLNgAZnqz4ShknJJEknZSLEXr00UfZuHEjf/3rXwd7Kl3yTu4CAIaVvQatwsy9oqyBfdTNsW3Crf1Xq6THyBghSQrR1iKk9oNFyGJSqVWE6NB3FiNkuMZ6bxGKkKa0S81vlxbfUrNdbGsXlqBsd1o8m60XmWOrtzfy8ortvZqfRLInknxHez+yfv16rrzySj755BPMPawjEggECATitT6amvr/jqo1ezI7yrIYptRBwxaaTW62VtUx1loW2ybUXDP4b76MEZKkEOFwG9GhJN8iBNBgFuJDadlZjNCuusbiFqGAbsamhMGf6CYPNAjXmO7KAyDbZaMJJ04ChFobsGQlHpMf34L0AlEPqQ2XLV7Jj5XNjC/IYFxBeq/mKZHsSaSMRSgSiTBv3jz+9re/MXbs2J3vYHDrrbfidrtjj+HDh/fjLAXZaTaqo2m4LVWs2tbIOLZiVjTCipA/1lAD6J2k1w8k0iIkSSG0tvEx/eAaA6iziE70lqZtMWtue8IRjbCmAzqH+9+DTR/3+PjCNSYsQuW6cH21twjRImoTWT1iLm6HhWZdVKxvbmxnSW7YCk+fBc+c0+FcOxp8CX8lEknnpIwQam5uZtmyZfz2t7/FbDZjNpu54YYbWLlyJWazmQ8++KDT/a666ioaGxtjj7Kysk63SyY5aVaqdZE9Rksl35Y1MFwRjRyrHGMAMOmRwbfARGOEnMYFOdhNNWyJZJAJt3GNofTPpavFls93WgmKHobVL3S6jT8s3GLHq1/xd/UBePzkHh9fDwWwKuJ17IgKoTYxQpqmY/WLa0Va9jAAVFXBaxL1yZob2gmhmvXib9N20TvQIBTRaDZqHTX0MrNNItnTGHTvTE/JyMhg1apVCWMPPPAAH3zwAf/73/8oKSnpdD+bzYbNZhuIKcbIdrWxCDVXsqKsgVGKuIA1uUaS1boRuxIStYTsGQM6tyi6rrO+bAdjATKGgbeW1uYG5t3/GQ3eIEUeBw+dNxOXLWU+IpLdHM0Ilg6jYlaUfjmHw2rihchhIp5v5dNwwC87bCPcYjq/M78UHww0g23n7idzOH6zUU5Hi1BFk58sXSxn5BTFD29KhzD4mtslWTRuM57oIn7IKfxmjb64y6637juJZE9jUC1CLS0trFixghUrVgCwefNmVqxYwdatWwFhzTnvvPMAUFWVyZMnJzzy8vKw2+1MnjwZl8s1WC+jA9lpVqqJW4Q2VLUwzBBCAecw6jEumIOYObat3kegRcQmRNKECb68qpqVZQ1sqfWydGMtX5cOgcw2icQgYliEIvSPWwxE5tirkYPQFBPsWA7VP3bYxheMcIDyAxPVLfHBndQdihIVQhGTnTrduA60sQiV1rSSq4iAaFNGQWw8bBXbBlsSv5N6YxsLd5vrSVvx09uijxLJnsagCqFly5Yxffp0pk+fDsCll17K9OnTue666wAoLy+PiaJUIifNSpVRj0RvqaS80RcTQuH0Ihp0ow3HIFaXrvcGY4XdfPZ8APyticHTTX5Zs0QydIgYwdJaP162HFYztbipzD1UDKx9tcM2gXCEc83vJg7upDdZFEvEEEIWF426uHnT21iENte2kqsYy2n5sfGIVdxYhVrj2z7w0QZeWdKmZ2Gb2mRtCz029jLFXyLZ0xhUITRr1ix0Xe/wWLRoEQCLFi3io48+6nL/66+/PmZNGkrkpNliMULhxgr8IY1CQwiRMYz6qBDyDl5RxXpviHQjjTeaKaP5mxmjbGN4unA7NPnknaRk6BBNn+9PIeS0CGtTVfp4APTGjunn/pBGvtLuu9tDi5A14hXHtaTRgLgO6G0sOWXVjWQrRuxgGyGkdNKB/oO1VeRFquMHb3NjVd/axjUmv8cSSbekTLB0KpFht1CnCItQuElcIItVIYRMmSOoZ/AtQg3eIOmGRahaEbEK+7CO92x/5n/Kn5molHbolSaRDCbRrLH+do0BvF0qAqK3b9/SYRt/KBL/7uhGjF8PLUJRIYQtjRrjZklriYuZuiohvDTFDI7M2Ljq9ABg9sUz2eq8QYYpbTLb2rrGfNI1JpH0FCmE+gFVVVDSRQ0Qs68aG0GyEG4na+YIGvTBjxFqbm3FpogL5A4tM2FdfrCM563XQ0P/Z9hJJD0lKoT61zUmhNBmn0hXbytSovhD8RT4TbrI7GInLTmi2DQhhBRbOg1KNI6wKrY+Ytw4Be3ZoMZfZzhvHwBGt3wTa7PR0OKPW5ohwTXW4JWuMYmkp0gh1E9kGKmvFi3AGMUwr1ucON3Z1EWDpQfRIuRtaog93xJKzFzzm9JxKEHya74Y4FlJJF0TiVqE+qmYIsQtQlFrjSPYsQK8LxSvDr1RE4kGPbEIhSMaTqOYomJLo8UsbkBUb9yqY/IKURRx5iXsa9rrUOr0NDK0Rti6lIimY/HXxFLxxcQ6D5aWrjGJpHukEOonCnNzaNYdAExTN4pBdzHpDkssWFobRItQ0LjTbdIdbPNaY+Mhk5ONw0Tj2NyWHwZlbhJJZ+gDYhES5SJqjKzPtFDHOD5/KBKrDr0xahFqroAd38KWz7s8dtvO84otnRajRLQaaoGQGLf5hShS0/MT9s33pPFuZKZYWPMKjb4Qw2hX8LGNRahtD7T61iDrKpu58bU1XP/K93z4YxUSiSSOFEL9xKgcVyxgep9os9WMItLtlliwdKSl88q1A4FunLtOz2BDm/ZF4aL9acmZBkCBb/0gzEwi6ZxwRFg29H4qpghxi1CtEfvj0L0xkRIl6PeJ1hi0EUL1m2HRHFFcsX0z4+h+bdprqPZ0NEs6Ad2o09VShTcYxh0RYsbsLkjYt8Bt5y1tPwC0ta9S1+KPZaLG6CJGyOav4ZXFj1K+9GmeXrqOi574hvpW6S6TSKLIann9REmOk2o87EUF+8QsQkVYzSotJnGR1Qex8apimOPrSOebihAYNSftex1M2DoJgOHBTaBpCbEKEslgEQgY6fNK/122okKoCWe8F1hrNXhGxLaJtOnRF4sRaqmMH6RpR6eFUoNhjTTFEEK2dBxWMzVeN0XUQmsNtXpuLHXenJEohNJsZlZaptGsO0hvqSCwbQVFRrX6kG7CokQSXGONhmtsvLKVV61/wVIbASuU6flcFvwlz3w9ll/PGr2L75JEsnvR61+4xsZG/ve//3HnnXdy11138cILLwxII9NUY2S2K1ZdeoJqBB27RZ+zoNUITvYNXvq8yS8umrV6BhE9/jFQJp2KKW8sft2CQ/eJO12JZAjgDxpWjH60CDks0fgjJeYeo13AtO4X6e0+bFTo7TugIoRTJwTC8RghrC7sVlMsFonWKmpaArFiikp6QYf9s93pfK+PAiBcsTZmEVqvF4sNOnGNzTEtxaJEqNAzqdI9DFcquc/6L55YuplwROvyfZBI9iR6dUV58sknGTlyJGeeeSZ//vOf+dOf/sTpp5/OyJEjefbZZ/trjinJ8ExnvN9YlAxRMj9i8wBg8g+eELIExLnrDBfAcYHbeG7KQ5A7jnSnnR90ozltxaquDiGRDCh+wyKE2n8WobYtZaLusfbCRvOJGz+f4iSIhQbatdZo7TwGJxjRcClxIeSwqAnnqG0JtimmmNdh/wK3nU1GcLapfgNFRur899pIACoqyllRJvaPBksfoa4A4LbQzzhFuQfdZCVPaUBtLuPdNZVIJJJeCKHly5ezYMECTj31VL799lt8Ph9er5dly5YxZ84cfv7zn7Ny5cr+nGtKYTWrBB25iYN5EwHQ7R4ATOFWiAxORocjaAgh4yL+gz6CvCmzAVEHaY1xcaXiu0GZn0TSHn8wIJ70U+d5iKfPQzxzTG9JFAx6UFiEQmZRGbq8XfmJ9hakKMGwhgPjNVhdOCxtLEItVdS2BsjFCNhLy++wf0GGnU26EEL2xk2MV4SleZk+DgCX1szTX4pK/A3eIIXUMkHdiqYrfKztQ15OLkquKBQ5SSnl43Wdz1Mi2dPosRD617/+xamnnsqiRYuYOnUqNpsNu93OjBkzePzxxzn55JO59957+3OuKUcoY1TsuX70jVC8LwBmZxtL0SB0oNc0nbRIAwBhe3ZsfFqxB4AMh4U1hgk+Ui6FkGRoEAiKmwalH4WQ09JRCIWaEi08SqAFAM2ahsNiolJvJ4S6cI2FIhrOtkLIaqKWuEWopjlAXjcWoUJ3XAjlNK5muCrO85kmYvrSFR+rt1YS9PtoDUaYbVoBwLf6GOrJYFSOCwpFPaJJainfbm3o/s2QSPYQeiyEPvvsM371q191uf6iiy7i008/Tcqkdhdqi4/ij8Ffc67r3yiH/D427nTYadWN6GR/Yxd79x/N/jCZCAHm8IgL7l45LtxOCwDpNjNrdGERUnasAF0f8DlKJNXNAbY3xDO2wgHxXDfZ+u2c9jZCKCpSQk2JFiElKIRQ2JzGyGxnJ0KoC9dYWMMZdY1ZnNjbWoRaq2lqqsepGELJ1VEI5bcRQp6gKLxYoRawXc9B00VbnMcbF6Dcvx+ZNHGU+g0AH0REL8eR2S4omArAJGUL66qaaQnIfoISSY+F0I4dOxg7dmyX68eOHcv27R378uzJjMjz8KJ2GHpmScJ4us1CM6Jy7WAIoQZfkGxFxDlEhdC04Z7YelVV2GoZQ1hXUb3V0CT/r5KBRdd1bv7nfdz+jzvxBUXRQD1qibG4+u28ihJ/3mIWgdCR5kRhYwoZQsjioiTHxX8jR7DDM5Nn1ROM7bt2jTnbucaq27jGwo1CcAVNTrClddi/0G1nm55LuE2Lkdr0ceioNCLek2ylGUtzGY9Z/84RhkXoHU3UHyrJccYsQvuYtjBMr2bdmhU9eVskkt2aHgshr9eL3W7vcr3NZsPv9ydlUrsLJ+5TyJHj8/jFoXsljKfbzTTrhhAKDHzGXYM3RJbR2PGgyeOYNS6XBYckijWbw8UPupEyvP2bgZ6iZA+nxR/kntAN3KvcSc3WtQDohiVGt3YUCcliSpGbM2cWc9Xx4/HbjIywdq6uqBDSLGmMzHaxUh/D72038r5fxOoEGztvwBqIdIwRausa01ui7TVyOt0/P8NOGDPbiMcPWYunce6BI1Ccidlr+6gi23PZsHNiWWWjsl2QPxlQyKWO92x/YtJrJw+Ke14iGUr0Kv3i7bffxu12d7quoaEhGfPZrchLt/Of+ft1GE+3W2gaRItQfWuA8Ubvs71GjmTRtJEdtkm3m1nZMprJaqkQQhNPGeBZSvZkGhviuVjalqUwZjJqqFUMWPvTIqRw++nCfXTNsmwIIKyibTAb89AsacLKAizbUs8Mo3eY0kWMUDDcJmvM4sRhTXSNmRD7tW+vEaXQLSrVr48UMsq0wxjch5sOnQIP5cF2ESi9Xitib3U7G8xjWDX+D7BJFEYtyXGBzQrZo6F2Aw4lCJEgVKyGkQf15m2SSHYreiWEzj///G7XK23typIuSbAIdVGFtj9pbW6IVcbF1fndZ4bdwkp9L87hfdi+fABnJ5FAa1O8OKDJyFxUQ6JhqdqJ26g/CDtzoBksvsQK8JaIsAhhSxdWFoNo3SGzv/NCqe1dY3aLQm1UCHnrcKhGLFIngdIAmU4LVrPKJj1eY8haPE08aWPVOTt4DYepqwgVHcERLnGdcTsseJxGK53CqVC7Iba9XrEKRQohyR5Mj11jmqbt9BGJRHZ+IIkQQoi7u8FwjfkaRMxDULF1eXed4TCzUjMqz+5YAZr830oGjpY2QshVJVyzprAQQib7wAghzSh/YQ02xDq+A1iMeejWdGFlMYhad8wRHxjxTG0JhiJx15jFSYbdTB3pRu80neJQqVjl7lhMEcSNZnGmI1bNulrPwJ1r1Puq+TG2XS1uXtIOJTO3gOFZQghNLGxT6frAi4mMPpr3tBkAeMtkZqhkz0b2ThgE0u0WmnTjAjoIrrGwUeek1ezpcpsMu4X1ejEhkwOCzVAj+45JBg5fc0PsubvpRwi0YI4Il5Spk/YV/YHqyiagm1HQob40Nm7TxDwUexq56TZcRu0hL3a80WzQ0k/hh9cTjhcOBUQrDACri4nDMtBQqTecgBMUcQ6rp7DLOZ06rYhPIlOo19N4RTuUDIdh5TnhTgC2HHIr15w4gRtOmcSlR49jv1GZ/L9zZnD76fvED1I8E9PP/8eyjKMBCG6XQkiyZ9NjIbRhwwa++SYxaPb9999n9uzZ7L///txyyy1Jn9zuSobd3CZrbOAtQprRcDVgzexymwyHBQ2VCqcIAGXHtwMxNYkEgEBLvOq6qkfQty/DGhHp81Znele7JZV0p43lmpEpu/nj2LgtEhVCGSiKItLSDWKVop8+C56Zl1CZXQu2xg9udTFpmBuLSWGTJoKfJylbgI59xtrys/1HsIMcpgf+jxtD56KqRjjCzIVw2Y+MPPpifnHYXpx30CiyXFYUReH4KYUxy1Bb7MUiFsrV+KO0+Er2aHoshP70pz/x0ksvxZY3b97MnDlzsFqtHHTQQdx6663cc889/TDF3Y8Mh4UmffCCpWkVQiho66RPkkG6XYSP1ZqNi3KLLMcvGThCrQ2Jy6VfxPp0WR0DJITsFj7VJouFTR/Fxu2aEGQmu5jHKCNgOtNpifcni1Idd1lhCKEwZjBZsFtMTCjM4BtDbKmKUa+rk6rSUXLTbYzM7ihqUFXopD9ZdxSPmYxPt2LV/FAnewpK9lx6LISWLVvGCSecEFt+6qmnGDt2LG+//Tb33nsv99xzD4sWLeqPOe52pNvN8ayxwMALIc0QQjg7D5QG4RoDYvVJ8Df086wkkjjhdpZSrXwVrqgQGiCLUIbdzGdRIbR5Scxq4tRFjJDZIUTPcZMLsZgUfnX4aGr0dm67hi2xp1GLUNDkiI1NG+6JCaEYXQRLR/nP+TPxOC384tCSbrfbGdNGZPOjkVoflhXkJXswPRZCNTU1FBcXx5Y//PBD5syZE1ueNWsWpaWlSZ3c7orLZqZZFxdDfRBcY9G6KHZ31xfcDIewCNVHY5nadLaWSPob3SduEMK6uETp3rpY6rliGyAh5LCwSi+hVXGJG4Fy0UvRoQuLkNkhRM/JU4ex/uYTOHu/4fF0+CgNW2NPlYAQQiF1Z0Koa4sQwJi8dL655miuOWnirrysGHvluNigjgKg6ft3ZAV5yR5Lj4VQVlYW5eXlgMggW7ZsGQcccEBsfTAYRJdfpB5hM6s0GZaW6AV/oPAGw9hDDQCkZ3d9wY1ahGojhuXK19DPM5NI4ihGNuU23Whc7G+Iucb6s45QWzLsFiKYWGUxAo03fQiAC2ERsjgTrT9uh4VCU0PiQdoIIYxss5AaL0w7bbiHOjLYqLUJkHa1a9bcCSa176VKVFWhKmt/ALJ+eBqeOkPGCkn2SHoshA4//HBuvPFGysrKuOeee9A0jdmzZ8fWr1mzhlGjRvXHHHc7bGZT3CI0wK6x0hovhYg6J47Moi63y3AIIVQeNC7a0jUmGUCi/by26cJ9q/gb4324BkoIGd+Br5giBjYvIRSOtHHReRK2VxSFb2wHJh6koSy+PmgIIXM8xqckx4XHaYlbhZzZYLIk8VV0j3/cadwcmkdIscKGd6HsywE7t0QyVOixELr55ptZu3Yto0aN4oorruD222/H5YpfkJ544gmOOOKIfpnk7obFpNCiiIuhMsCusdLaVsaoRu+w3HFdbjd5mBuTqrCx2bgoS9eYZAAxh0SBwKhFyBRojAkQ+rHFRlui7uHPIsb3pOxrfK2NsRR4q6uji+6b7JP4TfD3fDjreTHQWBZzOUUrY0faxAgpisKNp0zGPvpgMbATt1iymTYyk4ciJ7FKHS8G2gg3iWRPoceVpUtKSli7di1r1qwhNzeXYcOGJaz/29/+lhBDJOkaRVHwm8TFXAk0iQvlAFXl3l5RwQlKg1jI2bvL7dxOC/uPyqJxczRGqKHf5yaRRLGFhUVou2ERsoRbSIs2Gx0gi1C64R5e4S+EdA/4G9A2fxKfo7NjPaN8Txovbj6QqXoxsxUVwn5oqYL0fFTDNRYxJ2Z9zZk6DMb9Gp7/Biad1n8vqBOmFnsA2BT0MMMENG0b0PNLJEOBXhVUtFgsTJ06tYMIApg6dSrZ2dlJm9juTtAk7iYVLSQulv3A019t5e531yXEbgXKRQPLFmsu2DvvGxfl6In5MmtMMuDouh4rWhgVQkCbYoQDZBEySkgEIhAZLlpQmDe8C0CrbkdRTR32yc8QruQdzRH0dBH3883KFQCoYRFk3V4IAeK7eM5zMG1eUl/DzshOszEiy8kO3Sil0bh9QM+filQ2+bnif9+xevsglD6R9As9tghdcMEFnY673W7GjRvHueeeS1rawFygdgciZieRkIJJ0UUtIYtj5zv1kqteEMXcZo3LZcYIUTxRNSpE+9yj2dl/6+iJ+TzwmthK9zehaBHo5OIvkSST5kCYNCMgudnkplW34YrGB8GAWYRcVjNOqwlvMEJZxnRG8Sa2zYYQUhx0NouiTPE93lzTSq2lgBy28+7nX7PvocfGWoRo5uR/1/vC9BEeKlYZN7FNUgjtjBe/3c6zy8oIaRp3nzltsKcjSQI9tgjV19d3+lixYgXXXXcd48aNY9OmTf05190Kq9VMS7TfWD/ECUW0uBVoY1W871Fay0YAlG7ig6IMz3JSkC9iFhT0wSn+KNnjaPSGSEdYTyxOd9wqCYQV64AFE6uqwqnTRULB0xXC7W9prQDAq3Ri1QH2KRJW1pXbGthuxDc5vEJcmIzK2FpnFqFBZNpwj7QI9YLqZiHK61qDgzwTSbLosUXoxRdf7HKdz+fjvPPO48orr2Tx4sVJmdjujs1sohknbrzo/gY0TU9KSmwUbzDeJLLK+OK2BMIMC20FE7iKJvXoOGMKs2htMO7I/Q3g7LoatUSSDOq9QYYpQjSodjeNPhfDFNGENWRy9PyilQTmHzyK/365lUc2pXNFmisW8Fyt5jKqk+0nFGZgNas0eEN8TRpTgZxQJf5QBEtUCA2QRaunTBvu4RldWIT0pm0MTLRi6lJvCKB6b2iQZyJJFklpuupwOLjiiiv44osvknG4PQKbWY01Xr37tWUceddH+EPJq+HhC8aPtaNBXIA3VbcwRtkBgGPYhB4dJ81ujt+Ry4BpyQDQ4A2RbrjGzE53rOYWQHiArSlj89M5dEwOId3E+4W/pD7vQG4Pncmdrss63d5qVpk8TARR/xgQNw0jlEoqGv1YIuI1YRlaFqGJwzKoUYX1SvHVg5HmL+mcOq8QQo1eaRHaXUha9/msrCwaGhqSdbjdHrvFRLPhGtuyvZzSWi/b6pN3AfK2EUKlteIu9pVvNjFcqRKDOTt3jYHInGnUjWgimUIvGQCaWlqwKcKiaXF5aNTjQihiHnhryvyDRwFw2daD+GvmbTwQOZW8wuFdbj9tuIjHW6OJ/Q5Rv6d545dYjB5lWIaWRchmNpGbk0uLbtQMk3FC3ZLR+CMfWv/IjNZPB3sqkiSRNCG0dOlSRo8enazD7fYIi5C4M3TpQqg0+cPd7dIrokJoL2UHd2z/Of7PH2L58mWYFJ2QJWOn/YyipNnMMnNMMqB4m+piz51p7gQhpA2CiJg9Po8RWU6a/GFeWSksqucdNLLL7aeN8ADwvT6KFyKHoio6Iz67AkdExOop1qFlEQIo8DgoN9xjNMoU+u7Yr+VDStRKztZeS4jFlKQuPXa3f/dd5035Ghsb+frrr7nlllu46aabkjax3R3RZkNcEKNugOakCiFxrFNMSxmmV1H76YPsFT4CLGAunNTjukUZdnP8h0i6xiQDgK+lAQC/6sTttCUESw9UxlhbTKrC+QeP4sbX1gAwuSiDmSMzu9x++nBP7PlNoXOZpa4gq2kd+xljim3oZdcWZNgp17PYm+3SIrQTckLloMBUZQNNzc1kujvWk5KkFj0WQtOmTUNRlE77ieXm5nLFFVdw0UUXJXVyuzOizYYhhIzA0GZ/8oLvohahqYrIEvO0buAg1cgAK57Z4+Ok2c00SNeYZAAJGkIoaE4jJ91GlR4XDuogiYgzZhZz1zs/4g1GmH9wCUo3NxLFmQ5y0mzUtAQYXjycNyoO4Fzz+7H1yhALlgYocNvZEbMISSHUFYFwhEK9EhSwKWHqSr+GqUcO9rQkfaTHQmjz5s2djrvdbjweT7Lms8dgs6jU6uJOotDIiGlJumtMZx9VCCETOiepRh+hon17fJx0m4VN0jUmGUB0o5xEyJzOnH2G8cnqYjB+m91uz6DMKcNu4d6zp7NqeyOnTutYULYtiqJw5xn7sHp7I+l2C+teT6y4r9qGoBDKsMddY7K6dJc0eEPxOEtA37JUCqHdgB4LoZEju/aJS3qPzayyThcXyL0VceFJpmvMFwpTrFSTpcRrCNkUw+JU1HOLULp0jUkGGqPzfMTiIttl5eQDJsELYpVi69jfa6A4emI+R0/sWS+wWePymDUuj3e+r+AtPVEImYaia8xt5xukRWhnNNTXMk5pji3bd8hM6d2BpAVLS3qHzWxinS4yT8Yq21DQdsk1FvI2Unr7Iaz579UJ495ghKlKJwUuXXng7nlPuMT0eekak/Q/atAQQlYj9sLhia8cgm6l7ihw21mnpYBFyG2nWjda7rRWD+5khjC+KnFN1XThGk2vXg6R5N3ASgYHKYQGCZtZZYueTxALDiXIcKWa5kDvv1Brv3iTUd7V7PXj/1HfEK/87AtGmGq4xdZpRfEdivbtVYPXDLslbhGSlaUlA4ApJKyYus0QQnZPfOUA9RlLFgVuO7W4qdHjAbVm+9B7DYUZDrxG+rwm6wh1SahWhIis1kfRpDtFbajqHwZ5VpK+IoXQIGGzqGiolKrCKjROKdsl11hL1RYA7EqITz94NTbubSOE0g75Jbpi/KuLex4fBCJ9vsHoSqZLi5BkALCEhOtBj7rBUtgilOOyYVYVNujxmxGLY+gJoQyHOVasUgu0DvJshiCttfDtk5hrRa/GLXo+26MxVa1V3ewoSQWkEBokbGbRvPRHw2w+Vtm2S66xcH1Z7HnL928RjmgAeAMhJimlAAybdjRKNEB61GG9On5amxgh3SuF0EAQimgEwsmrMp5qWMPih1i1d2IRGoLxNd2hqgr5GXa2avG6XUPRIqQoCq50wzUWbOl+4z2Rj2+Dl3/D9PX/BGCrnke9bgh1b103O0pSASmEBgmbWbz134fFneI4ddcsQqbmHbHn+4aW88EP4u7E1rKdNMVPRDFDzlj46X/g3OdhxIG9Or7FpOI3G194mTU2IJz+/5Zy+O0fiVpQ3jrY8e1gT2lAMUeEEDI5OosRGnoiYmcUuO1sMxqwAliHoEUIIM0QQmrYC52USdmjWf18wmK5WkC9YSmXQij16VHWWGZmZrd1M9pSVyc/FD0hKoR+0OKZYy27ECPk9FfEno9Vt/P6D2s4ZlIB7pYNANQ7R5FjskDmSPHYBTSbG8KIhpORMJgGsu3lnkUwrLFym4jF+uz7Uo5+4zAIeeG3yyBn70Ge3cBgi3hBAXPUImSyiLYUodaUc42BKLC4pSxuEbJahub3x+3OhEpQ9QiEA2CxD/aUhg5FM2H927HFcMZI6htE6AHe2kGalCRZ9Ogbec8998Se19bWctNNN3Hsscdy0EEHAfD555/z9ttvc+211/bLJHdHbBbhGluniRih0coOvD5/r44RDGtkRmpAgbBqx6z5cW5+FziCLK/IbmhKG0NOH+eq2DIgqtECTbIDfT/S1j1a9MX1QgQB1KzvKIR0vVeB76mAruvYNS+YwOxskyrv8KSsEDr3wJEc++n+rNLe4Ed9BD9Vh+b/LCvTE18Itkoh1JZgYtyUNbeEuoaoa0wKoVSnR0Lo/PPPjz3/6U9/yg033MBvf/vb2Njvf/977rvvPt577z3++Mc/Jn+WuyFRi9AOsqnT08hSWjjU9wFwVI+PUVbXSjHCAtc65ee4Vz7E7OZXCIRuJM8nhFCre0yf5+p02PG22HAqAZE5JoVQvyHcozoLTW8wsTIe/N7WLfnemkpKqj9g9NIrYO6DMO74AZ9nf+EPabgUcUNgc7ZpXTDpNPjxDSjYZ5BmtuuMynFRUpDNnIpbADh9iIrXfLcLv27BroSE6IzWFZJAMF476AttAp6CEurXSSG0u9DrGKG3336b4447rsP4sccey3vvvZeUSe0JRIOlQeHf4TkA/CbyJBhVdXvC9u1l2JQQGgoZx1yJFxtjlW2ULX+bgmApAL7MnnWZ7440u5lmHGIh0PP5SXpPkz/EFeZnuNbyVOIKI2OvotHPhU8sY8OHj0GgEX54bRBm2X+0BMK4EC1nLI42QujYm+H33ybGC6UQvz2i7zck/U1+ho1WDCtQsI+ZY9U/wpM/hc1L+j6xoYDxfvzaciNnB6+lJC8jHiztk+EgqU6vhVB2djYvvvhih/GXXnqJ7Gx5B9FTohYhgEcjx7NRKyRXaSSy5K4eH6N2h+H+MmWhuHL4PO0YAKxfP0hRSGSThbL6LoTSbZZYX7TeCDVJ72lu9XGh6XUAbg7No3XqArHCCMhcV9mMrsNoTZRNoLaTopkpjDcYxkkAANU+eFWkk82JUwq5/af78MTC/Qd7Kl2SZrPEagn1SQhpEbh/f9jwHrx99c63TwUCIpOu3CecKCU5adQjPp96q7QIpTq9jtr729/+xsKFC/noo49iMUJffPEFb731Fg8//HDSJ7i7YrPEhVAIM7eHz+ZB6z9Qvn0SjrxGBIjuBG+1+DH02fPxANvG/hyWv8qImo8B8OsWlMxRfZ5rut1ME4YQkhahfiXQUI5Z0QjpJh6OnMAZgaWMhZhFqLS2FRtBSpRysUPdxkGba3/QGoiQbTQhTsUMsa5QFIUz9xs+2NPoFodVxYtNLPQlhf6bRfHnu4tIMN6P+ogNRYG9cl3UGRYhzVuLqbt9JUOeXluE5s+fz9KlS/F4PLzwwgs8//zzuN1uPvvsM+bPn98PU9w9ibvGBO9r06nRM1B9NQS/WsT2m/fh+0W/6/YYkQbRoyySLppAjhw3nafDs2PrN+hFOOzWPs81zW5uYxGS1aX7k3CjKIdQhQcdldJW4/9nCKHNNa3srWzDpBjpzS2VEGju7FApiTcYxoWRNJBiNYNSHbvFhDfmGtvF6tKREN63/9ZmOZCwOhzRuGzxSh74aMMuznIQ0LSYhcyr25k23EOG3ULQ5gFA6Sp9/uuH4Yt/D9AkJX1hl/I4DzjgAJ566qmdbyjpkrauMYAwZl6NHMQC89tY376cIiC39L8QvK3LTBlrq7AKmDwiBX9ykZvfh3/GUaZvyFWaWKcXM8na91TddLuFZrpwjS1/HNa9DXMfAquzz+fa01GaRTmEKj0TgFJvohDaUutlgro1cae6TVA4dcDm2J+0+IOkGcHSWHcf11gq4LCYqNejFqFdc41Vb1xObrjNzZK3VogqqxMiIZZv2MHzy7ehKnDO/iNxO3du+R50Ql5A3Hi0YGf2OFEKQXdkgQ/UiD/+GqP4GuD1y8V+034GdveAT1vSc3apoOLGjRu55pprmDdvHlVVooDfW2+9xffff5/Uye3O2C0djakvRQ5JWLYSRtv4cZfHcIcqAbBkCZN7tsuK35zBn0IXsVEr5PnIYTitfTfaptvMNOldBEu/8jsRsLv6f30+jwTUFiGEGi2i6MGOoPG+R11jNa2MV8oSd6rdfdxjAW8b61YKpsqnMg5rW4vQrrnGqtZ8CsCSyBTCZsOiV7seXv4t3DGamc9MZ4qyCU2HTzakSHNX473QdAUftpgQsjkyCOjGjWb7zLG6TUTFE601AzRRya7SayH08ccfM2XKFL788kuef/55WlrEh+S7777jr3/9a9InuLvS3iIEsFIfTUv6XgDs0EWKemDtm2JlSxUtb92I3iRcJ8GwRrYmvnyOnBGAiEMoyLDzkTaNI4N38Zk2BUcyhJDdTBMdG69WV25rs9XQTAlONSxeIW7DznwAyrzGHbqvnnBEY2udl/GKsAj5EdYif+X6gZ9oPxH0CqGtoYLFMciz2bNwWEyxrLFIYNeEkFa2DIBv9b1psYvPMF89BN8+Af5GVD3CHNPnjFXKKHz/99BQ1s3RhgiGdawVOzlpNiYNE9mMHpc1FjDduRAykJWnhzy9FkJXXnklN910E++++y5Wazz+ZPbs2Xz++edJndzuTNtgaYjWxVP4bOa/uDfrGq4OLQTAtPE90HV+fP4m0r64k8g900HXqWsNMkIR1jhHTknsOAXuxCJoybAIdRYj9MWmWv5475PxjXStz+eRgM0v/qdqRiEAW3xRIdTA9gYfYU1jgiqC5Jfo0wB4+5NP+WFTqYhlSHHCPmER8quO3a5Y5FDHbjHhM1xjId+uxZ1lN6wE4FttDNWqsJyE14osyCaTcPceqq7mGvOT7Nv4Lvond/d12v2PEYPXip3Dx+ahGgUxPU5r1yn0dZvjz2WdoSFPr4XQqlWrOO200zqM5+bmUlsr/+E9pX2wdLZLiMpycxHP+fblc20SPt2KtXUHVK3FXvkNAGbND98tpq6ulgJFuEvU3HiNkoKMRCFkNyfDImTpUEdo9fbGWFNXMb77BOwOJq6gcBfYs4pQFKjTDEtcoJHSqkZyaSRLaSGiK7wVmgHAKSxh7OPTaHnjmsGadtKI/gAHVRlvNtDYzGrMNRbx9z5GKNxSy7DIdgBWaKNj3dnNfiES/h0Q5T0mqls4RF0NQOjHd4Z+XzPDNdaq25kx0hMb9jgsscyxDlafBIuQ/F0c6vRaCHk8HsrLyzuMf/vttxQVFSVlUnsC7V1jueniAlTbGmRHg48AVr7QJoiVpZ8QicS7kevvXktgh7iQNChucGTG1hW2sQg5LKbY3UtfSLOZO9QRavSFmGhYJgAphJJERkjEE5jdw8h2WWkkHidTXlnJCEW4zirVPH7UR8TWqeio3z4uekSlMJrx+QqZZXzQQKMoCkGTuOHRduH7vP17ER+0SSuggXQ2BDIT1n8c2YfvNdHvMJr1aG3ZJtrHDGXauMZKsuOfy0ynpU3j1W5cY7Lg4pCn10Jo3rx5XHHFFVRUVKAoCpqm8dlnn3H55Zdz3nnn9cccd0vaC6G8dGGSXrOjCc24QfpRN+qO1G7EYwRGAygtlWSsfwGASktibZK2rrFkuMUAMuzmeNaYYRGq9waZqEghlGwyI+KCanYPIzfdTgQTYYu42FZXVTBcERajBmshm/WChH2dkWbY8P7ATjjJaH5x9x2WQmhQCJnE91wL9N4i1Lj+CwDWW8cDsNYbz5QK6ibW68V8qk2On0s3rk8b3t3V6Q4IEZ+45rXqDkblxD+X7rausW5jhJJrEXrn+wreXNXRGCHZdXothG6++WZGjBhBUVERLS0tTJw4kZ/85CccfPDBXHNN6pvmBwqzScXUxlqTawih77bHg5G36EawYfUPZGniruJrbSwARdvfAqDeEbcKQKJrLBmB0iBcY02GRUg3YoRaW5rZS9kR30gWWuw7IR/puhAC1qyi2GciYBY/KI11VRQbQiiUMRwvdpZkn4lvzIk8FT4SgMh3iwdh4knE6OkUMUvX2GAQFUL6LgRL67XCsmMpnIzVpLI1Eu9JuE4fThALn2pTxLao/Cdi9MhbP7SFUEODCEHwKY6E66vHYaGOTlxjgWZorYovJ1EINXiDXPzUcn739Le0BMI730HSI3othCwWC0899RTr1q1j8eLFPPnkk/zwww888cQTmEyyvmZvaGsVilqEqpuFa8NiUiiN3vFv+xoQlaI/jEwDwB5qAKA1bVTCMfvDIuRxxusI6YbrIqNxXbyoH0iLUDIwagj5dCuujOzYZ6LVJC62wZbamEVo7NhJXD9nIjN++f+wn/MUr5iOBkD58Y3UboNiuCE0iyymOBhEDNeYvgt1hOw+8eNvySxmRLaT7XpObN332iiuOXECjr1nEZj8M7ZOv5znIoeLlVs+g5Cv75PvJ+obDJFjS0sINch0WTq3CLUNlIakZo2tKGsgrOmENZ3yhqH7nqUau1xtb/To0YwePTqZc9njsJlVvEER+zM2P7F43LThHraWiqwLUdALtus5rNQT3/OgpyRhua0QciShmCKIbJKAKZo+L35ks7ztatdIIdRnIk3lmIBKPZMMhyUmhJqVdPIArbU2ZhFy5O3F/Knx/304fx+2VuQygmrYvgxGHzEIr6DvqCFhidBlDaFBQbM4wQ9KqPdCKN2Ib7N4hjEq28mHVZloqKho/KCUcO0hJfzisL2Ag1DrvGz8/APq9HSyIs1QtRaKZiT51SSH5qYGAMz2RHHudlip142xmvUQCYPJnOgWg6QLoSgVTX72bve7Idk1evRLeemll/b4gHffnQLpkEMEkTkWAuCEKYW8vqqcd9eIWKAD98rm/tJsgroZqyJMoNv1HFZpeyUcQ8neO2E5N82GqoCmg7OToo27iupwQ8ioohoO4gyIi16tnk620ixdY0kgULcdJ1BJJsPslphrrEEXosASbGS4yTC5exJdomMLMthYPkwIocZtpCqmqCXCJi/wg0HEIj5rSqiXLTZ0PRbfZs8qZpTXQQQTO9QCirUdVGZMTrCmFLjtqIrCWm0Eh5i+h8rvh6wQ8jaLcACbK7E6dKbTwjf6WAK6BVvlKnjtD3Dyvyhdv5pRQMhVgKW1oseusbXlTVy2eCWXHTOWIyfkd7pNWyFU3ujfhVcj6YweCaFvv/22RwdTZN2PXhGtJWQ1qVjNKveePY1zH/6S77Y1cuI+hfzrgw1s1fMYY8Ti1JnzsNmz2BgoZLRajqYr2PISLURmk0peup2KJn/SXGMAVqcbouFLgSZcoVpQoFQvMISQtAj1lUC9EEK1ZGI1i/8jQG1E/DhlK00UKsbdZebIhH3H5aexI+qKSIUidV1gigghpEghNCjohhAy9dYiFGjCYfSIy8gbwWQ9CMDFvl+zt7KdYMm0hM0tJpWCDDtrW0dwCN9D5eo+z72/CHrFhc+Z7kkY9zitbNPz+F3otzxouxfl2ydg9Gzqf/yEUcCP5vFMpudC6IMfqlhT3sQL327vVAjpus7KmBDSqYwKIV2H1c9D8UxIQpPtPZEeCaEPP/ywv+exRxKNEbIbgshpNfPcRQfT7A/hcVrxOC1sCeUzBiGEmm0FTBvmYeX60YymnO16DpnujA7HzXcLIZSsYGmAdKed5gYH6YqPUGs9Hq0eTLBZL2Rf1kshlASiDVcbzaL+StQiVBkW8VnjlDIsSgRMVkhLzBgbV5DBJ1EhlMIWIUtY/ACb7DJGaFCICqFw7yxCwYbtWIFG3Um2x8NxuSqZr1r4zjua7/TRXJjTMfi9ONPJDy2GZbNy6LZnCvuEuzY9I9EilGEXP5/vaPvhO+APOL+4G/2Dm9nHJ1xjT+tHcTMfifY4mgZq9yG5rUbwc21L5yUwttR6afAGeMF6PTZCPNPwqFix4T14fiGMOgzmv7arL3OPZpd6jUmSQ7SoYlvBYlIVPE5RXLEgwx7PHAN8zkKmDfewXBPusHV6cawQY1sKjcyGZFqEMp3WWFHFlsY6cpUGADZpogKyFEJ9R28R7sYWixBC0Rih8oD4f05RjSBM9/AOF9Wx+WmxAnaR+nZNWVMIa0T8AJvs0iI0GCg2IYTMkd4JoeYqYYWs0jPJcJixW0zMOyDuvm2bdh6lKNPBWqOuEJWrh2RhxVBEAyNuzePJSlhnNqkxMVQxYQGYHSh1GzGh80lkMi/UGK9fj0CgkZ0RjRetbQl2uv7bsnqKlBpmqBuYpG6huPw9saJ8hfi7bRlokU73lXTPLgmhr7/+mj//+c+cffbZzJ07N+HRG5YsWcKcOXMYNmwYiqLw0ksvdbv9Cy+8wNFHH01ubi4ZGRkcdNBBvP3227vyEoYEUYuQo4tYnkJ3ohAKuoqYPtzD4sgs7gqdzq3hn5Gd1lEIRQOmXbbkBEuDkTlmpNB7m+vIMfxkm/Q2Qmg3aPEwmCheEQgdsIpCdFGLUFVEvO/RSuLt44MAstNstDiGARBOYSFk00QmjMXR0dIp6X8Uq7DEmSP+Xn2ffbXCCllnyomFSJx74MhYiZBR2R2FUHGmgw36MCKYhNWkaUeHbQab7fU+XLr4TGa4Mzusj9601ulposu8waLIsfg0M5FoPaweBEz7okKotXMhtGpbE6OVeP2ggxpeEU9qNoi/Yd/QL045ROm1EHrmmWc45JBDWLNmDS+++CKhUIg1a9bwwQcf4Ha7d36ANrS2tjJ16lTuu+++Hm2/ZMkSjj76aN544w2++eYbZs+ezZw5c3ocwzTUiMYIddaJHkQmWVshhLuYKcVuQoqFf0XmssMyEmcnmWFnzCzmsL1zOG168ip9u50WmowUel9zPbmKEEKxFH902IVME0kck0/EEoTt4s7TZTPjsppo0Nu5idrFB8X29xQDYGkpT0lRqmk6duNHx+KUQmgwUG3is6agix/WHhKsF601mizxlPlCt4NrTpzAadOL2L8kq8M+RR4HAayUR4vCDkH32ObaVpyKiMWJvjdt8TgtADR4Q3DQbwmqdn7QhvOhNh0Ar1EDrCdCyBsSQqjeGyQc6fj9rWsNJNRu2ye8GqrXQW0b8VO+smcvTJJAr00Gt9xyC//4xz/4zW9+Q3p6Ovfeey8lJSX86le/orCwsFfHOv744zn++ON7vP0999zTYS4vv/wyr776KtOnT+/VuYcCnbnG2jJtuIe3DSGk6QrmrGLS7RbG5qXzY2Vzp9YggEnD3Dyx8ICkzjXTaY1ZhMIN20kzLg7b9BzCmDATEan1Msh1l7EExMVSc2THxvIy7HxTszf1pJOJ4X7sxCIEYMssJlyjYiYELZWQ0bvv42DjC0VwIX58bS4phAYDs92Bpiuoii5qOvWwjIHWJCwVXltewviCQ0o62xwQMUIA6/QRFFMq3GNjj9m1ifcHIR81ZesYbgSBY+1MCIlrcIMvBNmj+ZXnIZbtCDC2wM0PFc3U6Wmi5GIPAqZ9QREjpOtQ7w3FLMJRWoMRZiqJVrPwN49hjlqEQAihqWf1/DVKgF2wCG3cuJETTzwRAJvNRmtrK4qi8Mc//pH/+7//S/oEu0PTNJqbm8nK6ni3ESUQCNDU1JTwGCrszDU2bYSHUr2AR8LHcVf4DLIyhMiYNtwDQLbL1ul+/YHHES+qaKoXwYA+bLTgoDXWkFXGCfUFe1C4vnRn/K56xohM6sngJP9NfK2NRUeFEQd3un+u20UFxnchBQOmW4NhXMaPjlVahAYFh9WCF+O6Eux5dWlTqxBCIWfnad+dUZQprhvLg8KSSemnwpL53WKo3djNngPEW1fy00/nMEY1xEdnFiFH1CIUJBCO8FmluE4uOGQUAOWhqGts50IoGiMEUNvaMWDaGwyzl+Eai1qcWPW/xPijiu92eh5JR3othLKysmhuFj94RUVFrF5tNP9saMDr7WXtiT5y11130drayplnntnlNrfeeitutzv2GD58eJfbDjQ7jxFykJ9h54bwedwfOZUcwwIUNTMPzxq4NgQiRkhcuByNQgg1qpmAEhuXQqgPBFuxaEIEmNLiQmj2+FwAtpPLGcG/svrcFTDyoE4PUei2x6v5NqZenFBrIBKzNCqd3H1L+h+7xYQvJoR67uq2ekV9K61dNmN3DPPYsZgUXgvtKwY2fgDvXw8vXAgvXtTj4/QbZV+j0sZF1alFKO4a+2RdDcGIRrbLyon7DMOkKpSHjGt0DxqvJgihTgKmWwMR9lKFEHrbIYwR5lZRjT7Wt618ZUq6xQebXguhww47jHffFb1hzjzzTC655BIuvPBCfvazn3HkkUcmfYJd8fTTT3P99dfz7LPPkpeX1+V2V111FY2NjbFHWdnQqbESjQ2yd5PdFbX+QNwCdMq0Ydx5xlT+csKEfp1fWzxOK7WIu/Ss5rUA+G3ChRPtQyaLKvaBVpExFtDNZGXFXWOHjcklXodOIS+v6zvuAredHUbmWCpahJp8oZhrrLO7b0n/47CaaNWN6vTBnt/YOgNCCJncw3q8j81s4vjJhZTqhWxwTgd0+OxesXL7N70SYv1CS0XicjeusXpvkEVLSwE4fd9i0mxmZozwxOP7euQaiwuhmk5S6PVAUyxhotIzje+1eKzgF9oEAljENbihdKfnkiTSayF03333cfbZZwNCZFx++eVUVlYyd+5c/vOf/yR9gp3x7LPPsnDhQhYvXsxRRx3V7bY2m42MjIyEx1BhZxYhgOkj4pkK0Zggs0nl9H2LE9pp9Dcep4XVmvD3O8PCFBty5GK3qLG0emkR6gNeo1I3GeS7HbFht9PCviPFZ8CkKuSkde0OLWhrEUrBoorNviAuxfgBsMpYs8HAYTHhJSqEeuga0zQywuKH3prZuwSN+YYL6YGmQxJX6BHYvrxXx0oq4WBH8dKNa2xZaT2fbqhBVUS2HMCscXnURXuRGTc6hDuvEQTgDcWbqHZmEcr2i5ubkD0Hd2YOn2mTY+vW68X8oBnejnLpHustu+QaGzZMqH5VVfnzn//MK6+8wt13301mZsf0wmTz9NNPM3/+fP773//GYpVSFZshgLoTQp1ZhAaDTKeVb7XEdh4RZy7ZLhst0jXWd1rFRbdOz6CwncCdNU5YPPPTbbF05M4oyLDHqkvrjYMshMKdpwB3R2tLm1gHaREaFBwWE62Ga+wfr37Jn57rQRZSazUmNCK6Qlp27wL0pw/3MLXYzevhmfiMDKsGxci0KvuyV8dKBg89vZgrrr2CWdc/23GlpWPgeKZLCKEfK8W17+iJ+bGQhVnjcmPfR62uVKS23zYS3ryy03P7dhIjVBAS7u5w5mgK3A6WapNi6zbphWzQDRFaK1Poe0uvhdAbb7zRae2ed955hzfffLNXx2ppaWHFihWsWLECgM2bN7NixQq2bhX/8Kuuuorzzjsvtv3TTz/Neeedx1133cWBBx5IRUUFFRUVNDbuvFjVUGT6cA9mVWHGSE+X20wb7qHI42DqcE9SK0X3FrfDQg1utmq5sTE9LZ/sNCstMYuQdI3tKpEWUUOoTk+nICNRCJ08dRhuh4VZ47t2AQPkZ8RdY5H6QRRCjdvgjtHw2h97tZvPEEIRVDAPnLVTEsduNbHFKIlxbsO/WbV8aZu2Dl3QLOJWanCTld67ZrmKorDgkBICWPmt/meeL/oz/wyeLFaWfdXb6fcJXdc59odr+Lvp3xysd1KSpZPK0FOLPbHOABaTwkWHx1seTSzMoN4hrEOhqnUiBirsgx/f6PT8O4sRGhYRFiEte28mFKbzlTY+Fhu0UR/Gxmhx27ZZZJIe0WshdOWVVxKJdKxeqWkaV17ZudLtimXLljF9+vRY6vull17K9OnTue666wAoLy+PiSKABx98kHA4zG9+8xsKCwtjj0suuaS3L2NIcPyUQlb/7VhOm17c5TZ2i4mP/jSL5y/qPEB2oLBbTNgtKsv1uFVITc/H7bBIi1ASaK0XzXbrySC7nftreJaT5dcezS2nTen2GFazSsBhiKWWyn6ZZ4/YsUKI4o29a82jG3FNrSYPyL6Fg4LDYuK20M9Yo40kV2nkv9abeOPDj7rdRzcKIVboWV2W9OiOE6YUkptu4/3WEi7bOI1l2lhx3G1fDWjgb1V9IyMU8b05wfkDADW6EUqRN6nTffbKTeOba47m86uOYPm1RyeEMiiKwvCx+wBg81XGhV3DVggl1mjSdR1fqG2MUKIQCkU0CnVxs6RmjeLYSQU4XBn8K3wa70Zm8I02lk26EZ8lLUK9ptdCaP369UycOLHD+Pjx49mwoXdKdNasWei63uGxaNEiABYtWsRHH30U2/6jjz7qdvtUpKtiim2xmFTMpsHvhtLePWZ1FxitN6LB0lII7Sq+BhGY6bdmder+6s4l1hY1Q9zNm3y1EAnvZOt+Ivo5iMZF7ITvtjXQ4A2SUbsKgArXuP6amWQnOCwmanBzdvAaVmijyVJaWLDxj9Rs7zqd3W8UU6zSM7uNYesKq1nl3APigb9r9JH4dCuKrx5qB866UV76Y+z5TFUIoWXaOB6c9r9ue3i5bGYK3Q7S7ZYO66aMGRUXU+vfMUb1Dq/LH9ISOoy0d415gxE8iJgtS3oedouJcw4YwT8jc7kwdDme9DQ2G5a8SPV6qhp9olq3pEf0+tfV7XazadOmDuMbNmzA5eqdWVSSWrgdllifMwBH5rCEtHrpGtt1Qk0i6yZaVXpXcbrzCOuqqAzcWp2MqfWe6Ocg2Nzhzrc9a3Y0cfJ9n/GHZ1eQ1SRKcdRkdH73Lel/ou73JlzMD/6Z9VoRBUodzS9d3uU+fqO9Rq2a1aMbu86Yd8AIrMbNXhgz3+l7iRXRPloDQOOOuCXFHhAxe5W6h8zi8eDcte/luPz0Nm2I2lwfq39M2M4bTLxpae8a8wbDZCniBsOcJtzf5xw4ErNxg3Tq9CJK9QI0FEzBJj66/zfw91Gw5pVdmveeRq+F0Mknn8wf/vAHNm6M3yFs2LCByy67jJNPPjmpk5MMLTxOC2v1ETTrDjQU8opH43FY2sQISYvQrqJHrSeu3O433An5Hic1GMGm7dN/B4q2F/ydiLGyepGivb6yhcIWUZahKWuffpuapHvaJm40kM4l2h+J6Aol1R/A1i863SfYIFxj7atK94bcdBs3nTqZnx84klnjcinTjWMNYBmIYHVH61Nu4UhOnLLrFdr3zk9js97J/jXrEhbbxgdBxw70rYEInmhleUOU5WfYufm0ySw4ZBQ/238EAaxs10Rw9mmBl8S2yx/b5bnvSfRaCN1xxx24XC7Gjx9PSUkJJSUlTJgwgezsbO68887+mKNkiOBxWAlj5oLgn/hgyh2o7kI8TquMEUoC0T5jpvS+CaGCDDvVelQIVSWufP5CeOzk/u9Q7e+5EPIbcRH+1nrygiIe0JcrhdBg0d6ikz58Ms9GZomFd6/rtEN8NEaoN1WlO+PM/YZz46mTGZefToVuxNoMYCNWtWFLh7ETDp7ep+bVdospFjCdQDshFI0PsholVVqDkYQsMm8wTKZilDNwxuuMnbXfCP46ZxIjspxYTErM+mRRjH03L5HX5R7Q6/+w2+1m6dKlvPvuu6xcuRKHw8E+++zDT37yk/6Yn2QIEU0R/Vofz8MniJ5AHqe0CCUDq9Few+HZ9btqgAK3gyo9EyiFZmER2lzTylMfruCa7xeLjepLIXt0V4foO20/By09E0JjIxtRTTrb9Bxsnp5XJ5Ykl/aZqdNHZHLP5tM5w/wJlrIv+cczr3PM4YcxaVi8wba51QjMT1Jvu1E5Lr7XDVdUc3n3GycRV2snmZbpff8sRrLGQNQ4a3eDv1E0SwXWVTbz2NJSDts7h2KlmtusT1KtOqjTXOiL/wsHXgBjjsTrD5KBUeDS0dFNZ1IVRma72FRXyOG0qSMUCbLkzWf4yakX9vl17M7sktRVFIVjjjmGY44ZQg3yJP3OvP1HcPMbazl6osgWA6P1RlQI+VOzjMFQIC0shFBaVt8uvPkZNrbpHrFgZI498fkWVi3/gmjnBFqr+1kI9cYiJLKC9lFE3OFKbS/cnQSdSgYGuznRSTBtuId/k8kKZTz76aupXf0e/6aIf/0s3uTa4ReWR4und8UUu2JUtov3okKoaXtSjrkzNE0nO1QO7XMSetEypCuchePjQmj8SbDiKREsrUV4+JNNLF62jc01rZxreo9Dta+Fn0YFNgAtW2HMkYRaa0UjXABH5/X6Tp02jOrPRoBhDKoiizzqqP3mRRqPmY/bKb9XXdFj19iXX37ZoU7Q448/TklJCXl5efzyl78kEOi6aqYk9fn5QSN5dP5+PHDOjNiYx2mlPlo9tQdl5CWdEPJh10WPrcycnrco6IxMp5XqWIyQEEKVzX7Gqm1iLdq7zJJNW4vQToRQ1CUwVRUxh99po8lw7LorQtI3zCY1FrQMMHW4+CwtCY4H4CD1e3Y0tAmADwdwRcQNkCu76zIgvWFUjpMKox6WPkCuscomL8MR3xc9vY1lK73vVq68kePivcD2PlrUyIoEoGELLbXb+bXpFSqra5iqiO/ACtch/Dd8hNi+Zh1oEcLN4traqrjA1Pn347dH7M2fzzkptnx98FwAjlC/pbyh5w1090R6LISuv/56vvsubnJbtWoVCxcu5KijjuLKK6/k1Vdf5dZbb+2XSUqGBnaLidnj87C0uVB6HJZ4eqi3Nul1PzStY0zC7kY0UDqkm8jN6VuMkMdpMVxjoBuusdqWAHsrbYRQq9Egs7/e217ECIk4CJ39jHTl5dreZEiL0KASLRCY5bJSkGHHalZjVYwPVNdS2dCmB5nhuvLrFjKz++bWjZKfbqfObLSKaanapSrlvWVbWSl2JUQEFaXECPNQLbucLdaWsYVZfK2Nw6db8RXuD9lG5m3lGk6vfoArLM9wpvcZJqubAVg/6fdcE76AoGKFsB/qS9GMyvOtpp20iCreD71wGv/TZ/OWtj9B3YRb8VJXkXpNmAeSHguhFStWJDRVfeaZZzjggAN46KGHuPTSS/nnP//J4sWL+2WSkqFLptNKPYZFSNeSWruiujnA/re8z/WvfJ+0Yw5FGpvEHXUr9oQ+Y7uCx2ml2nCNac3iDre2JcjYtkKopZrrXl7Ngbe+T11rP/zI9MIi5A9HKFEqyFWaCOgWVuqjSbdLi9BgEo0TKsiwoygKBRl2vtNH06rbyFaacbesZ0eDj4NvfZ+n3xdFAiv1TAo9zqScX1UV3FkFBHSzKAMxANmPDdtF6nytOQ9yREFH0guSUthzVI6LC7WrOTBwHxNuX8HXRgkSfc3L7B9aBsDZpvdJV3wEFDsjx81AQ2VztGVG9Q9oXtG9PtqGpEusLpRffczzRVeioVJuWNa8VaV9fh27Mz0WQvX19eTnx7MCPv74Y4477rjY8n777TekOrtLBoYMh4UwZhp0o4aUt2dF9HrCN1vqqWkJ8P4Pg1gleQBoaRZCyK/YY1kju4rLaqJO9QBtLEKtQfZOcI1V8t6aSqqaA6wt74faT72JEQpG2N+wBn2rjyGIpdPCdJKBw2kVQjTa1LnAbSeEma814R47gO957et1/LH1HrJWPwpABVlJbQI9PNtF5QBmjmm1Ikat0V4MuUZBT/fwpBzbYlI5Zp/hNCL65927QxQkVlYtJk0Rbka3IqxsZfaxTBmejUlVWBMx3OTVP6D6hBDyWzw9OucZM4tJt5uFsAOCddIi1B09vurm5+ezebMw3QWDQZYvX85BB8XbPjQ3N2OxyAvYnoZJVciwm6mNuseSWMSvsknEzTS2BqBp4LJHBpqAt1X8VfreVFdRFPxGPRe1tYpwOALeGnKVRHFS7w0BHeuXJIUE11j3wtgXirC/KuoHfaWNw2Ex9VkMSvpGNIU+JoSM3nefGe6xo9RvcK1+kjPMSziWpQBUk0m2q/ftNbqiIMNOOUaa+AAETNuaxU18q7MYxh4Hs6+BY29K2vH/cdY01t5wHPuOzGRpZDytls5dbtudE3BYTYwvSGe9ZliEqn5A9QtLe6CHQmjujGJWXX8sjuwRYmAA6zGlIj2+4hx33HFceeWVfPLJJ1x11VU4nU4OO+yw2PrvvvuO0aP7MRNFMmTJdFmpJSqExA+fLxjh3TWVCbUwekt5oxBC10YegLvHw6aP+zzXoUjIL1xJQSU5d9QRp4gzUiMBGupr2ZvEHxKtuTIWpNy+om2f0fUEi1CgsaLbWCR/SOMAwyL0lTZBBkoPARxGjFChIYAKDUH0RuQAAA42reHwhhcT9mm25KL2sA1MTyhw26mIZY714CZI0+CV38GXD+7S+dK8wmISSB8BJgsc/ico2neXjtUVDquJBYeMQkPlzfDM2PhnkXgl9cp08XzacA/rdSP4vPoHzAEhhMI2T6/OqbvFMaytA5N9l6r0WAjddNNNmEwmDj/8cB566CEeeughrNb4HcAjjzwi0+n3UDwOC3XtLEKPfLaZCx9fxiOfbd7l41Y2+TlUXcUZ5iViYOP7fZ3qkCToExahkJocIeRyuWjURbxGU01ZzC1WrwvTvN4ma6wvQrVTgi1AXPiYfbW8833XP2QO7w6KlRrCuspybW/pFhsCRP8HwzwiXi1qGdpObuxHu1hJtPxa7Mltr1SQYac8JoR64BqrWgPLH4f3b+i06OPOcAfEOSLuUb3etzccO6mAQred5wP7AdCi27kpfG5sfa1bNFaePiKTdVEhVLMOa0C4xsK2zlPnu8KcJSxCaf5BqjKfIvT49is3N5dPPvmExsZG0tLSMJkSC28999xzpKWlJX2CkqGPx2mlrl0K/ertIu5lW723q912SnVDIzeYH40PVK7Z5WMNZcIBkdoaMvUtUDqK22GlSs/ErXhprdnBJKUUgKXaRE40fYXSxn2ZdNeYESitoaKiYVJ01pdu5bgpnZcFyPOJtgbr9OF4sZMhA6UHnd8dMYaR2U6OnSxq6ERdYwAvaIdxiEkkL7wb2ZeZ6o9kKi1UeqYldQ6FbjurYzFCPbBmGJmQBFtEwkYvs71yg0IIKVmjerVfb7GYVI6ckMeTX0zkr5EL2BTJY60+gnvCc1EAJUPEJU0b7qFMz8OvW7CH/RR5hdVU76KGUFc4c0VV68xwm5IZzRWiOrVJ3nRE2aWmq+1FEEBWVlaChUiy5+BxWqgh0SJUWisEUJNv110vk2vfZS+1gmC0BkfV7imEIn7xXkVMybEIZTotsTYbofoyDlFFM9M3IgcCoIa9OBFux6iLLGkY8UGNujNmgWqs6fqHzBoSgjlagiHDIS/Og83MUVnccMpk0myJQdMApflH0aKL5ecjhzE7cBcLg5dRX3BYp8faVQrc9ljGU49qCRlZVQA09DIwONCCR28AwJrb/+Ed04ZnAgqPhY7iE20fQOGe8On8I3w6TiNjb68cFy67lY26uIEoDopgbr2TqtLd4SkUzWsL9GrhBt+2DO6eAM/NT9bL2S2QUYmSPpPoGqtB13W21Ap3T5M/1OPjPPzJJm5+fQ26rqPrOscFRAHPhyMnig2atic1PX+ooAXFexVJkkXI47SwWi8BoHjDfxmhVhPSTXyoTYudI0cRAiTpMUKGRahZd9BkZK9567s2y9tCQjg1IVwrsobQ0KOtEJq+VxF/DF3MPeG5/OA+DJxZvK/tS4EnOZ/dtueMxgjpPQn0bRuU39jL7OX6UvFHTyPDk939tklg2nBPl+uipQtUVWHacA/faGMT1qu9tHS5coRrzK14WblhGyte+3+izMkPrxFa9wF/e/V7Fi+T2d5SCEn6jMdpbZM1VkNVc8Bwueg0+Xv2Q6vrOn9/6wc++fRjIneMJfTfc5imbCCom3gkfDytDqPCa9Xa/nkRg4gWFBYh3ZIsIWTl1YjI6MxrWgXACn00XuwEbOJCn0NUCCXXIqQbbVZacGLziHIbYxuXondRaNMeFsIpWn5B1hAaeuSm2chyWUmzmTl4TDbvajO5J3w6I3MzOGSMKHw4riA9qed0Ws14beLYSmv1zuN+2pbtaOjdD7tWJ+IYt+p5A9KGYq8cV8LnvK072Nmm19v04R4+1KYl7Kum5/TqXIo9gyYjbf+6J94mr/yD2Lqm165m0Web+OvL3xOOJLcQbqohhZCkz3iclnjWmLeGzTWtHKV+w0rbhUxp/rRHx2gOhAlFdH6irsTsrcK6/nUA3tb2owY3tc4xYsPK3bC4omER0s3JKUjncVpYpZdQZY7H5XymTQbAaxV3lLmGRSjZwdJlFaLmU6vixDP9FADmK68ReuJ0qNvUYXtHRAihxqhFSLrGhhxmk8pLFx/Cy789hL1y4nGgJTkubp07hed/fTCHjundD3RPsGQIIa1ooZ1bgtu29+mlRShQJVpbbNXz8Dj6P7wjau2JMrkoXiTRYYmLomkjPHyuTSRI/DthSeu9xareIsppHKMuY5hShw8bzbqD7Ka1nKR+gS8UYX3Znp1VJoWQpM9kJliEqtlS28pPTUtwK15mBL7q0TEaWoULbZgSv6CFdZVF4WMBKLcJX/duGScUMno3WZMkhBxWQOET209iY58bQqjJJIIt466x5AqhDVtFPIfV5cZ+2O+5w/wrAroZ6+b34f4DYOMHCdu7NCGEpGtsaDMi28no3LQEN9nIbCcZdgv7jsxESUIF5vZkezJi2Y8d6pO9/Bu4d1q8Pk5b11gvYoRCEY2QUUyxXC0YsBpW07sQQm0tQtOGZ+LHxkqtJDZmTe99C54mmwh6n29+C4Dy3EP4v7AIN7jcvJhbzA8xYdFk+OH1Xh97d0EKIUmfcTvbxAh56yitbmam+iMAmZGeNWJt8IlWD0WGEFo59hKODf6db3RR5XWLRWQ/UPopLHs0sWhfiqOGhWtMsSRHCGUa5v03OBQNhUbdSa1HpOU2qoYQ6ifX2DbDIpThyQZFYUXBXE4M3kKdZzJEgrDu7YTtnZrImNPtHkC6xoY6dosJj/H5GpWT3JT59hRm2KnRExsIA+BvhBX/hfrN8M41YqxtsHQPLUJPfF7KuGvepG7bOgBqLUXJmHaPmDbCA4i+bnu1eR/bCqEsl5WR2U5WaGPi6129z8xuMKpLR4uqFuz3U55S51CtuxmpVjHP/KHYcN1bvT727oIUQpI+43FYqIv2G0NH2b4s9qXLpR5/DzKTopWOhynizm6TqYSNelGs1c86RoknNevgtT/AJ3cl8RUMLkpYWIQUW3KEUDTO4btgIZfZruPc4NWMyhcCqMboTJ+rNFBILX8tuwA+TF6z5LBXCKy0DHG+UdkuNujFfOsRlr1ok04QcWFpuhBCk8eMYniWg4NH93+wqqRv/HRGMePy05k5snep3L0l322PfV5pU/uKLUtFwC/A9y+KQqu9jBEKhCPc+/56NB1MDVsAaHIMnBA6cK9sJhZmMHdGMTlp8Yry0fYmUQ4encPjkaPRdIUvtfEd1veEvSbuF1/I3hvnPnO4+Jh9eNp5TuKGA9DKZKgib78kfSbTaSWCiQY9DY/Swl41H8bW5St1NPlDsbL9XdHgFRahqGtsc0hcZEuyXWyqaeWHcBHMXAhbPxfusc27T5VpsyGEzLbk3GFnOkWcQ4M3yFvqRHx6hINz03hvbVUsJXlfdT2/ML/B8PAW+PYJmH1VUs5tiwhhozrED1iJcbe7KZDOkZBQJTgU0fEgtj9i2lhOPvuIpMxB0r9ce9LEATlPodseayCc4BrbbBRYNdtFd/ZP7kyMEfLVibg7a9ffp9dWluNoLeMW8ysM08pBAa+rOPkvogucVjNvXCJKDny7tb7NeOJ1cs7UQp7+Kp/9Aw/gxcYyW/fX0c4oOuKXUJAHWXvBsBmgqvziMDccfCPet028/flyTjN9RqR2I70/+u6BtAhJ+kzUVB6tBbOff2lsXa7SRFOLb6fHaPSFcOInUxE/jKtahAl4fKGwNDX4Q3DS3fjPegYAvXxlYpfzFMYUETV9TEkSQtH/Ryiix+oEjc4T7+eH6kH4sDNB3cr5JsNN1bQ9FrDdV+yacPOZDCE0Klu8ph9ajNfWXEEgHOGF5dvYWufFrYjz2nYhCFSye1PgtsfqYSVYhAwhtGn0eWK5YlXMNaYjTMjfff89SzfWQGsNTU+dT+N9s6BGFO/UNY3gezfxnvVPzDN/iEnR+SwyiUj6wFmE2pJoEUqUIgeUZKMqwpLrxY7dvAtSxWyFfc6E4pmgtvnJN5lxnnATj7kWAKA2bIVIz8ud7E5IISTpM+l2C4pCLHNspJLYLd5Xv3OTa31riELDGtSkO/lsm/hCTijIiK0HuGFJE2VaLoquQVnPArGHOhZNCCGzPTmV2R0WE1aTmrBcbNR52ea38YIqWuGYlTYps51kdPWWiKbj1IWwMTvE/60kVwig5fVGkG1zOa98u51LF6/kpte+JwNj+7T+dbNIUo9hbkcsRijcbFxTWqqhUhQIPe87o0eXrx50Ifi3KCJT8l8vfMD1j7yMdv+BZKx/CXfNt4QeOhq2fUPle//kZ76nsSlhPo1M4ozAdZwTuhq3s+9Nj3eFnDQbJlXBpCqktYuRM6kKR4zPiy0ns59blOLhJfh0K4oe6X0xyt0EKYQkfcakKrgdFrZqeQnjzYr4EQw17FwINfiCMbfYdj2bYFgjy2Xl6EkihbbRF6K+NcgLy7fxlT5e7LRlaVeHSymshhCyOJJjEVIUJWYVAjh6Yj6eqLvMF+LB0HEE9HZe8doNfT6vLxQhDWH9s7qEsNkrx8XY/DS2hY07ey1EdZVwj22trMOmiDpTSi9bB0h2f8bkpRFyCEthXYWRHWa4xNdqw9mm5+KL1hdD3EBtCBuBwVoVP1PeRvVWs14rYrU2CkugDh45ltzPRVf5xVm/4lfKtXytjweUmEt5oHFYTdx5xj7cNndKpzFAFxwissailb6TzbBMF6W6uM4m44YoFZFCSJIUPA4LN4XPZcmI3/BxZB+etP2McrPomxNp2HmNigZvKCaEonEsP9t/OPnpwpLQEgjz5Bdb8Ic0vtR2MyGkCyFkcySvKJ3WpgDd/ENGxYRRdXOArSEP14UX8HzkMN7WRUfxZAghbzBMuiJcYxansAgpisL8g0sIYabeCHwNNwphHGwWAa5hVLDKPoWSREyqwrjRImPKW18uOsx/di8AH2jTAai1j4xtX6ens1kXwmi6sp7DVFFM9M7wGZwVvJY3I/uBFsKkh3gvMp3gfhczMjv+ufMMQDHFrjhtejFnzBze6bqDx+Tw318cwIsXH9wv5850WtmiixR7KYQkkj7gcVppJI2nrT/l/NCVvJ23gGaLUfOiJd5iwReMcOHjy/jvl4km2AZvMJYxtkPPxqQqnHvgSDIclljm2MOfigqwX0WF0PZvIOTv3xc2ANj1AAC2JFmEAGpagrHn04d7OtztPhuZzWWhX/NdxPghqd3Y53P6ghHSDYuQYo/XRjltehFuh4VyzSMGjOwUl5Ex1kwa9EMdGknqc8AU8V23BWr4vwfvhorvCKhOHg6fAMB2Uzyup4503ovsC8CJpi8Zo+4goit8rk2iFQe/Dv2BJwqv5snIUfwp9Ctmjc9jVE48U9M9hIt5Hjwmh73zk1u9O0qWy0JpVAgl4TqQikghJEkK0bupjdXixy3bZcVnE0LI1EYIfbqhhnfXVHLfB+vFQEMZtFTR4AvFagjt0HP46YwiCt0OTKoSK7LX6BNxQqV6AX5sEAlAc2qnfIYjGnaEELIbVpRksOCQUQD846ypKIqC3aKS7YqLIbtFfPU3a9ELYDIsQhHSFSMw3ha/aDusJk6ZNoxKIwPI07KR28z/x6HGHXuL2j8XeEnqU1wshHoOjRxZ/jAAD2snUW/EI27Q49XT6/R0vtbHUa94cCriO7VSH00TLqYUuQGFazdP5prQBeTkDaM40xkL5gcGzTU22GQ6rdI1NtgTkOweeIy7qdIa4RrJTrPhdwp/vcUbz/gorRHBsTsa/fhrt8IDB8F/jqaxNcAwhEXo/OMP5ebTpsSP3cZkLQIHFZox7uQCLf32mgaC1mAER1QIuZJnEbry+PF8/KdZnDZdpAQrisKZ+8VN74VuETwddSUk407QG4zHCGFPFHWjsl2xJppntDzF2eaPuMIsMgBbVekWk3SBS9xMWZQIo9VyWnQ7D/iPja1e7Y/HJdbpGWiovBaO180R3d1h/sGj+OVP9oqNzzYCkNsWhRxM19hgkuWyxi1CddIiJJHsMtFg3KDRvC87zUrYKb5cDn88i2xzbTxNu3XpfyDYDPWluLzbYjFC+cPHYGmT9RQ9tllV+O0RImagWTeykIKpLYS8Ph8WRWS8JDNGyGY2MTI7UVj9/MB4PEV1cwCbWY3fCfrqEqvz7gJ+vz92J44tUQgVuu1UIQKiHQh3ptV43T5pEZJ0hcUObdysS7R9aCXenPjrlniPszrDSvR65IDY2CcR0VpmXEE6Vx43ntOmF2Ezq5w2XbjUShKE0B5qEXJZKdWM60D9lj0yhV4KIUlSaH83leOyoaWJL5craFR9jYSorBLWITNhXN8/Fdt+7+CaWIwQ7sSgwai16YQphYwz/OQtUSGU4rWEfK1t5p+kFhtdMczjwGEUttR1HafVhA87tSZx17129bd9On6wtSG+YEsUN/luO5V655lhPrMUQpJucMWtPtvzZwMww2hRsSGQgV8R14KIXXy+vtLG86U2nm+1MazQx6AoIgNNVRX+cdY0Vv/tWCYUGqU+suPfuT3WIuS0UkEWTbpDlCGo/nGwpzTgSCEkSQrt/evZaVbUDOG/zwgbAuf5X/Cv8nmMVco4Wv0Guz9eLXae6X2sSgQ9LR/ciRVez95vONNHePjDUXvjtJqwmBRadOOuMMWFkN8rLFphVDD1/x3pq787lGnDPdx15rRYqu6PIfFDU1f2fZ+OHfaJtioBxQamxB+VQrc9FiPUHr85ebFRkt2QcCD29PATz2FKkZuLZ40xYt4U1keE5blkhLB4aqicFbyO3zhuJ4yZUdmuhMr2ba3NuWk25s4o4sR9ChNi6PYk3A4LqqKwVjcsxhXfDe6EBgEphCRJof3dVHaaDbNbCCGX1gKtteg/vI4TPxeY3uQX5jfEhob1Z6YqGh8qIw/pkEF0/JRCXrz4EPbKTTNq5Fjj5vEUd435vYZ4wDYgmVNj8tJ46TeHcNzkAhxGFdsNmvg/ORv7Fh8Q9ok+Yz61Y6xTbpqNajq3CAUtUghJuqExnmE6tmQkr/7uUI6amE9xprgG/Cd8AsuZQPH+cxJ2O3ycsHSO6ybbSlEU7j5zGvfPm4Gyh2YuqqqoofS9NkoMlEshJJHsEu1TT7NdVpzpmTTqhun5qwdRNOF7PsP0Mfuq6wliheNuSzzQqEN2ei6Pw0ILUddYaguhoE/ETAVU+4CfO1rOf70u4iUyWvqWMaJFLUKdCCGzSSXkKogt/6DF3Z8hKYQk3XGMKIDI8bcnDBdnimvLS9qhPDnh3+QPi8fAmVSFS48ex1kzh3PJUXsP2FRTlUxXGyEkLUISya7RmWssw2nlfW2GGPj0ntg6kyKK/f3PdCyMORJdadM/Z9RhPTpX624SIxT0CSEXUgZeCEXjhTYYQiiztW9CSPcLi1DQ3HkWmN2dT4PuIqSb+E/k+Nh42OrudHuJBIADfg2/Ww4H/CphuCgzHjQ9a3xerFUFCAtkbrqNv5++TyweSNI1WU4r3+ujxELFKlG8cg9CCiFJUmjrGnNaTTitZjLsFl6LHCgGI8LPv0wbC4BPt3J36wn4dCvNaaKEfKPqgZyxOz2X22mhJZo+n+KusXBAWIRCg2gR2qAJIeQJ7OhbgUpDlIbMnZcByHM7mR+8gp+HruL9yIzYeMTm2fVzSnZ/TGbIHt1hOOoaUxX4yd45mFSFvHTRL6zAPfDfp1Qm02Vhgz6MiGqFQBM0lA72lAYUKYQkScHjiFuEstPE8wyHmU+1KTTp8cyMy0IX8W3BmVyj/IYa3Gypa6UqTVSP3ejcp0dxMh6HZWhkjdVvgaq1fTpE2G8ES5sGQQgZvYuqcdOgu1DR+lRYUQkI11jI0rlFqNDtYIU+hi+0iXgtHrYbrVTCdtl5XtJ7JhcJS+IhY3Jiqe9RAVQohVCvyHJZCWOmxinKk+xpcUJSCEmSQrrdTLQxck6auCtLs5kJYuEdbSYA5aYitugFbN7/r2zIOQoQBRaX5c7le20kXxX8rEfnynRZaWWQhZCmwSPHwv/NAsMltCtEAqIAZcTcv6nzneGMZdIoMfcY1T/s8vFUwzqnWToPTm17lz4mL42rQ7/g3vBp1GdN3eVzSvZcZozI5PlfH8y9Z0+PjRVkiM9YfoYUQr0hGtqwzRYVQisHcTYDjxRCkqSgGh3oAbJdQgiZTSoep4VF4WOI2Nws5mhAVHMdnSesBmvLm/kiuBcnBm8lULBf5wdvh9thoWWws8aad0BzOYT90LjzprJdoRuuMW0wLELWeGzWei0qhHa9hogpJERpxNqFEGrz47R3Xjofa1P5R/gMHJ103JZIesK+IzPJ+v/t3Xl8VOWhN/DfmZkzezLZNxKSsCTsuygogliDKIqX9l7stRVwqbQqVdD2Wm8LrX0L7W29ttpq3xblWq1wrcuLBSu4sIlYAomsIkogARICIetMZn/eP87MJEMWJiGZJfP7fj75mDnnzOSZx8Pkl2dtN+190mBlZuJE3zpDFBp/HZ5QFygHLnwRucJEAIMQ9Rn/XxVp5rYPpvxUEw6JIdh48278t7UEgLLdwoS8JABAWVUDyqsaAADj80IbNJts1LZbRyhCQaj9B4Wtrtcv43UpLULefl5MsTPtA0igRehC74OQxuX7f6ELrUXIr/0aL0RX4r4Zhfj0Rzdi/oRBl7+YAvyf3W17D8bXVhsMQtRnLL4B06ntglChb+XWHV8oiyom6jVINsqBIFR68iJO1ilhwH/scpKMclvXmDP8XWNCCFw4eajtQOsVbE3hVN57f68q3Zn2LUJtXWOhBaHaJjua7MFL8Wvd/iDU+Syd4BahtiBk0DIIUd+QJIndYr3gbxE65lLWXkJ9RVzNHGMQoj7j/6vC3zUGILDf1fYvlFWkC9NMkCQJI7ISodWoYHMq+00NSTOFvNePMlg6citLv/zJKWz6aEfbgStoEYLLt0mpbOj+un7QPgh95d989WIFIES3z2u2uzDr19vw9T/sDjqu8yhBSNJ33rLXvkVoeLtF7tgiRBRZyb4g9EVrEqDSKF3+zWcjW6gwYhCiPnPX1YMxfWgqSkZnBo75NzW8aHUCaNvtWatRYeygtl+YobYGAcrmiIExQhHoGvu8pglDpXYfElcShJzKGCGNPvw7sLdviakTvlYcjyNQpq5UN9phc3pwvLYFja1trUI6j/I8taHzFiG9rMYDM4fgXyfnoiDVCLNv1hqDEFFkpfj+CD3f6gGSfAtTXryydcViCYMQ9ZkbR2bir/dfE1jxFWgLPn7td0RvH34m9GBwY5JRDiyoKCIwWLre6sJQVXXbgSvYtd3r6xrTGjpfe6c/tW8RaoUODvjWgrpMsLM63AAAM2xo3vnHQHAyeH1ByNj1WK8n5o7Ef/3reEhS25ovCXoOliaKpBTfcAa7ywtXkrKuWzyNE2IQon5VmBr8C74wrS0ktQ9CE/M634eqM8ntWoQktx3wuC7zjL5ltzYgW2oXfq4gCEluJQgZjOFf/dYgtw8gEuqFr7vqMmOe/N2Zj2s2IHf3k8DO3yivJ5T3IhtCG/T+43mj8OANQzEhN6lH5SaivmXWaVCUqbRKV0n+bnIGIaI+YTHKSG636nRBu2A0OT8ZGpWERL0GI7K73hjxUnpZBZe63eDiMI8TMrecCj7Qy64xp9sL2aOs5Gw0h79rLCNRaZEZku7rvvQHocu8nxaHG4DA19T7lQMVOwEAZl8Q0pqTQvr5N4zIwONzRkClis/NLomiyaziDABAuTVFOXCxIoKlCS8GIep37bvDCtt1leUkGfDyPVPxyn1XQ1aHfitKkgSNrIVd+AJWb7rHTn4MNNf0/Hn7X8bXrRsAAF74foH3MghdbHFgjEr5sNGnDenVa1yJiXlJeGbhBDz3TWW7i3rhC2OXaeGyOd0okk5jkOR732fL4HXYYIIy8Ftr4t5hRLFmVrEyY+zDWt8fRHHUNcbOeep3hWkmlFc1wGKQO8wMmz4srVevqZfVaHEaoIer5wOmzx0G1t2ifL+yQZm5pdEBqssM2j13BNj4MG7wPfxS5KJIqup1EGo+fQjDpQtwQIau8LpevcaVkCQJd0wcBCEEVBJQD3+LUPdByOrw4AZVedsBrwvOil3QS0qXmcEcejcnEUWHKfkpMOs0ONCaCujQNoVeNfDbSwb+O6SI83eHXTpw+kroZXWvd6C3VbZbPv7I/wN+NxFYe1O3z9mwtxLPr/0jAOCcSMJn3iF43jXP94K9GyOk+up9AMABzThAG/51hPwkSYJRqwm5a8zqcGOWSqlDF5TwKL7YAgDwCgkGE3f7Joo1Wo0K1w5LxRmRBo+kjqsp9AxC1O+uHZYKtUrCzOG9a/3pjF5Wtdtmo2dB6IRvAUcAwOuLgJYa4Mw+wN4EnNkPnNjW4Tnr91ZhbOteAMDz7tsx3/lzfOid2Pbz3c4ev4eEqg8BAJ+bp/b4uX3NoFWjHv6use6DkMvWiCkqZeHFV903AgA0vlDXAgNUak6HJ4pFUwtT4YEatXKucuDckbaTl1lfLJYxCFG/m1KQggMrS7C8pLjPXlMvq3u9lpDD2sUmqQ2ngFcWAH9ZADSfU163oRJCCFTVnMdVvl/+273KJqFNMEJIvn9CPV1d2t6E1DplsPGplGt79tx+YNSqQ541ltDwOWTJgzMiFW96ZgAA5EZlrJNVilzLFhFdmUFJymfqF5JvzKJ/89WGKuB3E4C3lkamYP2MQYjCwqTr2+FoV9I15mzt4vrKPUBrPSA8wLlDwBv3As+MheOl23GXZyN0khtnkIEKoezHI6CCW5ukPLeLVhR3Uy2qN62Bt+WS8yc+glq4cdKbCZEytEfl7w8GWR1y15il5UsAwDFvHo6IfLg0bTP+bAxCRDEr27f6e7nbt6hidbnSEvTyfKD+JPDZaxErW39iEKKYFNQi1MNZY+5W5foT3ix4x3wDGDRFOdG+S+zcIeArpetKX7kDj8pvAAC2uccCaJvu7eguCHm9OP/incjeuxrH1v8w+NznmwEA73snBe3NFilGrRoNCG3WWIpVmU3yhciDGxp8kvFvgXNOifs8EcUqfxDa0+rrGqs+AJS/GrymkMcdgZL1LwYhikl6jQotgRahngUhYVdahLZ6J6Nl3gtA3tXKCd96OACAIxsBjxPQJuBQzr+iViTBLVR4w9cV5NcqJynfdBYeyv6C7IZ9AIBBZ7coHyB1XwFuB3D8PaUMnilIa7c3W6QED5buPghl2pWl96XMkQCAJ2tmBs6N9H7RPwUkon6XZtZBo5JwyFugHGisBN7/afBF9oZwF6vfMQhRTNLLalh7OVjavy2HVRjQYncDSYOVE452Y4fOlCr/zR6HPyc+iKsdz2G040XsF0VBr9Wi8q2Zc2mL0LnDwJYfBx4mehsh3rwfeHYS8PupQGs9mqQElIqiqGgRMrQfI2Sr63pgpBDIcZ4EAAwfMxWDkgyosmnwduJdAIDNpn8JQ2mJqD+oVBIyE/VohhH2hALloLUWMKYCat8fbK31EStff2EQopikl1WwondjhCSXsieWFTpllWR/EOpM1jgcO9cCARUc6BhYmqWOrShfHimD9U+3Ao5G7BdF+F+30mIiHX5TuaD+JABgpzQFHqiRao6GFiE1LvrXEepu49WWWiR4m+AREqS0Inx7mjKW4JHaW3CH42fYmHpPmEpMRP0hJ0n5XL2YOCJwbE/yPIgEZWwkgxBRlDDIajSL3s0aU7uVX/I26C8bhDyZY/BVrfL6mnZbQWg1yj+dtkUI21qEWjc+BpO7HjXGIix2PIZ3vNPaXjB9JKBSVsTe5FJWdE41Rb5FyKhVoxU6uCVfWbqaOVarTKc9JTJhMJlx51V5vs1bJZSLYUhJ5mKKRLEsy6J8rp7WK63fHqiw/KvJaAhxwdVYxJWlKSbpZTUaezlYWuPb6DTQNZaRF3yBNiHQ3XZGPxxOzwUYtWoMTTfj4Bml+yw/xYjjtS2oE5esvWO7iJF2ZVr8Q66H0QQzPvGOQqU3HRkaK87e+EfkqusgTu/Hu+8pywlERdeYrAEgwaaxINF1Xnk/nQXE858DUAZKD9JqkGTU4n8fmIb9lfXQqlWYMzorvAUnoj7lHzC9TzcVV6k0+JtrBs4iDY0wIxlgixBRtND1cvq8xyug8yp7Ylmhg9XhBvQWQJ8EABBqLYR/uwuVjD1NyiKQI7MTAx8QQNv+aac9/g0KlVkV7iN/hwZeHPHmo7Q5FQAgqWXc5vw/mGX/DWavO43b/q7B7uy7IaCCUauGURv5v0eUVh3AqulizBOgjBuq3AMAOCZyYdQpzxkzyIK7pxXgzqmDkRwFrVtE1Hv+z7mDrhzsu/MzPOG6FwBQ5/EtjcEgRBQdglaW7kEQamx1wQRlx3crDGh2+KaC+lo/qlVZeO2Ub4uIjJH44Ljyj35mUTqy2gWhwjTlQ+Ggp0A5cO4w4HHDcfAtAMBmT9tq0bOKM9AIM2q8Ssg4dq4ZS9Ypq1RHQ2sQoAyWBoAWle+92y75sBMCeH8VcORtAMAe7yiYoiDAEVHf8gehsw127K9xwOuLCTUuBiGiqKLXqNEiet41dtHqhFHyBSGhV7rGgEAQOmxPw18bx8Kr0sI9+uv4+EulZWRWcXAQ8u+bdsSeCmjNyr48p/fCUKVMwX/X2xaEpg9NxdTCFORY9HjhW5NwzZCUwLkbijN6+M77h79FqEnyBaHj7wVafwAo33/8DADgKddd+MQ7GiYdt9IgGmiyfWOEahrtKK9qCByvsvtnjXGMEFFUUKbP93wdoYtWJ4p8LUI2f9cYAKQOAwB8KQbhkBiCF2/4FKNyEtHi+BRpZi3G5FjwZW3bzyn0dY3V2z1AwTigcjewfQ1UXheOeXPxlRgUuLYgzYQN37kGXgGoVRJuHpONJrsLEoAEvXwFtdB3/EGowR+EDr4OHH4LeHgfkFwAnNwFAHAU3461n93qew4/PogGGn+LUG2zHaUn21p/al1GQAZbhIiihUHbu66xiy2OQNdYizCgxeHGgdMNODV8Ec5N+QHWum8BAJRXNWDbsfMAgOuL0qFSSZ22CNmcHrgzxygHfStT/8M7FWnturwKU02QJAnqdrPOEvVy1IQgADD4Qs1W3U1AwQwgIRvwuoFy35L6p5WuvPNJykazelkV9H6IaGBI9S2q6BVAbbMDKglINspoEMpnHoNQH9uxYwduu+025OTkQJIkvP3225d9zvbt2zF58mTo9XoMGTIEL7zwQv8XlKJOUNeYywp4vSE9r7HFClnyAFCmz5+ub8U3nv8E//bXEzgy/H5cgDKOp6yyAVsO1wBQxvgAbRsS6jQqZCXqA0GgJXl00M/Y7JmK28crLUJatQqDkg1X8E7DwygrLUJHMARY/Heg5OfKifK/Al4PXFVKEFq5X3kv5j7eO46IooNaJWFEdtv+gcVZiSjOSmi3VAi7xvqU1WrF+PHjsWTJEnz961+/7PUVFRW45ZZbcP/99+OVV17Bxx9/jO9973tIT08P6fk0cAR1jQHKOCF94mWf19Lc9teMDTocPtsIp8eLc00OVJxvW0TwTIMys8ys02D2CCUI5aeasOzG4ci26KFSSRicYkTFBStOaIZjku95lVIOjok8PDUmC5mJOqQn6CCro7/h1T8DzOZUQiJG3AroLMoS+2V/gWy/CIfQYGdztnI9u8WIBqyX77kamw9Wo7yqAQsmDsI7B87iWMXAbRGK6KfZ3LlzMXfu3JCvf+GFFzB48GA888wzAICRI0eitLQUv/71rxmE4oxOVsEBGW6ooYFH6R4LIQjZmpsAAK3QwgM1Ki/aAucOnWnscP03JucGtX4sv6lti42iTDMqLljxmT0DkzR6wG3HO66rAEgoSDNiamFKh9eLVv5g0+oPQrIBGPt1oPRFYOtKAMBhUQAnZN/1HChNNFClmLT41jX5+NY1ysrxB8804lP/psytDZErWD+J/j9V2/nkk09QUlISdGzOnDkoLS2Fy+Xq9DkOhwNNTU1BXxT79LKymrGth4sqtlqVsOOQlOd5222pdehsxyDk/yDoTFGm0lR87HwrMOxr8GoMeNNzHUxaNdKjYNuMnvAHG5uz3c7S1zwIaAyBTRbLvMMDp0zsGiOKGwVpJtT7F491NA64HehjKgjV1NQgMzMz6FhmZibcbjcuXLjQ6XNWr14Ni8US+MrLy+v0OootBt+YlrYB06EFIadNCcIujbHDOf+ssOEZyj/4GcPTMMz3fWf8QeiLc804NO2/cbP0Ar4SgzB6kAWSFFsDiZONyuDuxlYXGmxO5WDaMOCW/wpcU+4dGvieQYgofuQlG9EEU9uBS3egb6kF1s4B9v8lrOXqKzEVhAB0+AUjfLtkd/WL54knnkBjY2Pgq6qqqt/LSP1P7wtCgdWlQ9yB3tmqXOeVTR3O+VuH7p8xBL+9cwJ+e+fEbl+rOMsfhFrw4P8exhfNMoamm7BmwdiQyhJN0hN0GJ5hhlcAO4+3+6Ni4rdwfPi9KPUWYZt3QuCwiV1jRHEjQa+BB2o0iS4WVTz6DlC1B/j0j+EvXB+IqT/rsrKyUFNTE3SstrYWGo0GqampnT5Hp9NBp4utbgq6PL2sZPhmoQckhD6F3tdyJGm7bulJT9SFtNBhQaoJGpWEFocbLQ43zDoN3vzetbAYomdafE/cMCIDx2tbsO3YedwyNhvVja1QSRJeMi7BX503QqdRAW5ldh4HSxPFD3/Xeb0wI1GydQxCvs2YUV+hrEIfYy3iMfVpNm3aNLzzzjtBx7Zs2YIpU6ZAlmPzlw/1jl6j/MPs6Q70apcyM0zSdR2E0kyhBWetRoUh6SZ8cU752XPHZMVsCAKAWUXp+L87TmDbsVrMeWZH0AKSAPC1UZnYdKAaALiqNFEc8XeFN8CMfNR2nEJ/7rDyX2cLYL0AmNPDXMIrE9GusZaWFpSXl6O8vByAMj2+vLwclZWVAJRurbvvvjtw/dKlS3Hq1CksX74cR48exYsvvoi1a9fisccei0TxKYL0gTFC/q6x0IKQxqNMi9fouwlCCaHv/zU8s229jX+ZNKibK6PflIIUmLRq1Fmd+LK2BWqVBFktQSUBeSkGLJpWELiWY4SI4odOoyyg2tjZoopCAOeOtD2+eCK8hesDEf00Ky0txQ033BB4vHz5cgDAokWLsG7dOlRXVwdCEQAUFhZi8+bNePTRR/H73/8eOTk5+N3vfsep83FIp1EyvDXQIhTabEDZYwPUgGzoeqp9Sg92UC/OTMAmVCPHosc1hZ13z8YKrUaF64an4b3D55Co1+CtB6/F0PS2wGh3eSBJyucexwgRxQ9JkmDUqtHg8U+hb9ci1HRGmUnmV18BDL46vAW8QhENQrNmzQoMdu7MunXrOhybOXMm9u/f34+loligUknQaVQ92m/M7fFCL5R1gzSGBOg0KjjcXmg1Kjh9Y18S9RroNKH/kp8/IQfvHqrB0plDoBoAW058d9Yw2JweLLtxeFAIApRWuByLAWcaWjlGiCjOmLQa1Nl8f0C21LadaN8aBAAXK8JXqD7CTzOKWXpZjWZX6OsItbo8MMEBQAlCZp0GDrcTeckG1DY70Gx3I62H6//kp5rw7vdn9Ljs0WpCXhL+cm/Xf80NyzDjTEMrkoyxOxaKiHrOqFOjxupbJLa5uu3EuUPBF9YzCBGFjV5WweoMfbB0q8sDo2/DVbU+AWa9BnVWJ7Isyr5hzfaWHgehePP4nGKMzE5EyeisSBeFiMLIrNOgRiQrD5rOtp3wzxjLGgfUHOAYIaJwCtpvLIQxQq1OD8ySMlha0poDW2dkJuqhVqnwxbkWpJpDHx8Uj8YMsmDMIEuki0FEYWbUqnEOnbUI+YLQyNt8QSj2WoRibkFFIj+9Rt02fT7ErjGjr2sMWlNg5lNWoh45FiVQMQgREXVk0l7SIiQE4LQC5z9Xjo28Xfmv7QJgj62trBiEKGbptWpYe7DFhs3pgUlSusagMwfW/Mm26DF7RAYsBhmzii6/kCIRUbwx6jSoEb4WIZcNsDcCZ/YDwgNvQg6e/kwFl953PsbGCbFrjGKWPmjW2OVXlrY7PUiA0jUGrRkPXD8EyUYZ88blINmkxU2jMmNujzAionAw69SwQwe7JhF6d5PSPXZ6LwDgc80I/O6D47jFnIERuAhcOA5kj49wiUPHFiGKWXpZjZYedI3ZnB6kSr4mW0MyphSk4FffGI9k37pBDEFERJ3zL5nRJPtWjW46C+fJTwEAb5/PAQD80zFYOVf1z7CX70owCFHM0suqHu0+77C3Igu+FVGT8vuxZEREA4t/EdVGTRoA4Mixz9H05W4AQKlnGABgr6dIubhqT/gLeAUYhChmKS1C7Xaf72ZxTgCQGqugkgTskh4wpYWhhEREA4PRN7nkolpZQd9z8mOkSU1wCA0qdUUoyjSj1FusXFxzKOT9H6MBgxDFLIPcbrC08CoD+LohNyvbtdTJ2TG3OzIRUST5Z9lekJQgVFT3IQCgLmEEdj15M742MhPVSEWDnAEID3CmNGJl7SkGIYpZelkNG3QQ8IWay/wFomuuAgDUa3P6u2hERAOKv2vsHJQp9DqvMvGkIeNq6GU1xuUq64t9Jo1QnlD5afgL2UsMQhSzdLIKgASH2qgcuMyAaYPtNACgSc8gRETUE/7B0oEp9AC8QsKF4m8CAMbmJgEAttmGKCerGISI+p3etzmqQ+ULQpdZXdrsC0Itxtx+LRcR0UBj0imft6fdyYFjH3gnwpQ1FACQY9EjxaTFP/0Dpk/vBbyesJezNxiEKGbpZeUfZqvkD0LdryVksZ9RrjcxCBER9YR/jFClJylw7GVPCVJNyv6MkiRhzCALPhd5OFR4D3DH88rYzRjAIEQxyyArt69NZVYO2Bu7vT7JoWwUaE8Y3K/lIiIaaEy+rrFqpwnuqd/Fy+6bsNM7FinttiUamZ0AD9RYb7kHGDkPUMuRKm6PMAhRzPIvhHjR62sRaq3v+uLWehi9yhgiTwJbhIiIesLoGyxtdbpxbvpK/MS9BFq1Ggm6tg0qRmUnAgCOVl9+pf9owiBEMWu8b3DeabtvLaHWhq4vrj8FADgvLNAYEvq3YEREA4zZF3jsLi9qm5Q9G1NM2qAV+Uf6gtDn1U3wertf1y2aMAhRzMpPNSLZKKM+lBah+pMAgCqRHvjLhoiIQmPUtX1uVtUrU+dT23WLAcCQNBO0ahWsTg+q6rtf1y2aMAhRzJIkCePzktAgfGOEugtC1eUAgK+8OTDIDEJERD2hVaugUSmtP1UXlZCTatYFXaNRqzA8U/k8Plrd/SzeaMIgRDFtYl4yGuAfLN3Q9YUndwEA9nhHwcAWISKiHpEkKdCaHghCJm2H60bG4DghBiGKaRMGJ6FRmJQHXbUIOZqBM/sBAHu8I9kiRETUC/5xQpUhBSG2CBGFxYTcJDRACUJu68VOr9m8+S1AeHAaGTiD9MAKqUREFDrjJUEoxdxZEFImoxytYRAiCguLUYY+QdlJ3m3t2CIkhMDp/VsAALvdIwGALUJERL3g32/stG+wdJpJ1+GaUdmJyEzUYVi6GW5PbCyoyD+NKfYZkgAHoHY0dDjlcHsxVToCAPjEO0q5nGOEiIh67NLW9JROusaSjFp8+qOvhatIfYItQhTzXFqlT1p2NQMed9C5FrsLo6STAIBSoeyBwyBERNRzJl1wELp0+nysYhCimOfWJrU9uGSbDVtLA7SSsvFfrVA2C2TXGBFRz5l0wZ+daeaOXWOxiEGIYp4sa9EsDMqDS6bQ25suAABahRYOaKHVqKBWSSAiop4ZkZUY+F6rUQ2YIMQxQhTzDFo1GmFCAlo7TKF3NiszyfxrDXFVaSKi3lk6cwimFqZg/6l6DMswD5hhBgxCFPP0GhUahBm50oUOQcjdUgcAaPCtNSSr2QhKRNQbkiRhcn4yJucnR7oofYq/FSjm6WV1u0UVG4LO+dcWavS1CF20OsNZNCIiinIMQhTz9LIqsKjipS1CwqY89u9H5omhHZGJiKj/MQhRzFNahLrYeNXuD0KmMJeKiIhiAYMQxTy9rAyWBtBh1pjK9ziwMSsREVE7DEIU83S+wdIAOrQIaXyrTftbjMblWsJZNCIiinKcNUYxTy+r21p8LglCslNZYHFScQGemzAR1wxJDXfxiIgoijEIUczrbtaY1qUEIZUxBfPG5YS5ZEREFO3YNUYxTy+r2sYItV4MOmdwNwEAJOPAWveCiIj6BoMQxTy9Ro2LIkF5YAsOQkaPEoRUppRwF4uIiGIAgxDFPL3cLgi1XgS8XuV7IWDyNgMAZAYhIiLqBIMQxTxlQUVfEBLetin0rlZo4QIAyAkcJE1ERB0xCFHM08tquKBBC4zKAX/3mG8GmUuoYTRx2jwREXXEIEQxTy8rt3G9v1XIpmy06g9CDTDBpJcjUTQiIopyDEIU83QaNQDgon9RRV8QEr4ZZI3CDJNOHZGyERFRdGMQopinl5WQU+cNbhFyNCv/bYAZZh2XzCIioo4YhCjmGbRKELq0a8zpC0KNwgSDzBYhIiLqiEGIYp5eo9zGbWsJ1QF710J7aD0AoEWVAEmSIlU8IiKKYuwvoJinUaugUUmo9wehs2XA7t9B7ztvVXPGGBERdY4tQjQg6GU1Lvq7xqr+CQBw6ZLxrucqbNHfHMGSERFRNGMQogFBL6vaWoQ8DgBAXcY1+K7rUdQbCyJXMCIiimoMQjQg6NrvN+bT4AtARi17gImIqHMMQjQg6GVVW9eYT52+AABg4tR5IiLqAoMQDQh6Wd3WNeZzTjsYAGDmYopERNQFBiEaEPSyGo0wQaBtmvwZ9SAAbBEiIqKuMQjRgKCXVfBADZecqBxIzMV5h9ISlGjgPmNERNQ5BiEaEPS+/cbs2mTlQNpwfHGuGQAwNN0cqWIREVGUYxCiAcG/31irRlk8UaQNx7EaJQgVZyZ0+TwiIopvDEI0IOhk5VZu0WYo/7UUod7mgiQBwzLYIkRERJ3jKFIaEPwtQrvyvoOhY67GodQ5AA6iINUU2JSViIjoUmwRogHBP0aoWh4MzHwcRy54AABFmWwNIiKirjEI0YCg93WN2V1KAPrCPz4oKzFiZSIioujHIEQDgr9rzOFWgtCxcxwoTUREl8cgRAOCwReE7C4vvF6B4/4glMWuMSIi6lrEg9Af/vAHFBYWQq/XY/Lkydi5c2e317/66qsYP348jEYjsrOzsWTJEtTV1YWptBSt/F1jDTYnvr+hHFanB0atGgWppgiXjIiIollEg9CGDRvwyCOP4Mknn0RZWRlmzJiBuXPnorKystPrd+3ahbvvvhv33nsvDh8+jNdffx179+7FfffdF+aSU7TR+VqEPjp2Hu98dhYalYSf3zEGGnXEsz4REUWxiP6WePrpp3Hvvffivvvuw8iRI/HMM88gLy8Pzz//fKfX79mzBwUFBVi2bBkKCwtx3XXX4YEHHkBpaWmYS07Rxj9GyO+lJVdhwaTcCJWGiIhiRcSCkNPpxL59+1BSUhJ0vKSkBLt37+70OdOnT8fp06exefNmCCFw7tw5/O1vf8Ott97a5c9xOBxoamoK+qKBR69pu5XnjsnCjOHpESwNERHFiogFoQsXLsDj8SAzMzPoeGZmJmpqajp9zvTp0/Hqq69i4cKF0Gq1yMrKQlJSEp599tkuf87q1athsVgCX3l5eX36Pig6qFVtu85//2vDI1gSIiKKJREfQCFJUtBjIUSHY35HjhzBsmXL8JOf/AT79u3DP/7xD1RUVGDp0qVdvv4TTzyBxsbGwFdVVVWflp+iw9TCFIzOScT3Zg3FCK4dREREIYrYFhtpaWlQq9UdWn9qa2s7tBL5rV69Gtdeey0ef/xxAMC4ceNgMpkwY8YM/PznP0d2dnaH5+h0Ouh0ur5/AxRVEvQyNi2bEeliEBFRjIlYi5BWq8XkyZOxdevWoONbt27F9OnTO32OzWaDShVcZLVaGSQrhOifghIREdGAFdGuseXLl+PPf/4zXnzxRRw9ehSPPvooKisrA11dTzzxBO6+++7A9bfddhvefPNNPP/88zhx4gQ+/vhjLFu2DFOnTkVOTk6k3gYRERHFqIjuPr9w4ULU1dXhZz/7GaqrqzFmzBhs3rwZ+fn5AIDq6uqgNYUWL16M5uZmPPfcc1ixYgWSkpIwe/Zs/PKXv4zUWyAiIqIYJok461NqamqCxWJBY2MjEhM5qJaIiCgW9Nfv74jPGiMiIiKKFAYhIiIiilsMQkRERBS3GISIiIgobjEIERERUdxiECIiIqK4xSBEREREcYtBiIiIiOIWgxARERHFLQYhIiIiilsR3WssEvw7ijQ1NUW4JERERBQq/+/tvt4ZLO6CUHNzMwAgLy8vwiUhIiKinmpubobFYumz14u7TVe9Xi/Onj2LhIQESJLUp6/d1NSEvLw8VFVVcUPXHmLd9Q7rrfdYd73Deus91l3v+OutsrISkiQhJycHKlXfjeyJuxYhlUqF3Nzcfv0ZiYmJvMl7iXXXO6y33mPd9Q7rrfdYd71jsVj6pd44WJqIiIjiFoMQERERxS0GoT6k0+mwcuVK6HS6SBcl5rDueof11nusu95hvfUe6653+rve4m6wNBEREZEfW4SIiIgobjEIERERUdxiECIiIqK4xSBEREREcYtBqI/84Q9/QGFhIfR6PSZPnoydO3dGukhRZ9WqVZAkKegrKysrcF4IgVWrViEnJwcGgwGzZs3C4cOHI1jiyNixYwduu+025OTkQJIkvP3220HnQ6knh8OBhx9+GGlpaTCZTLj99ttx+vTpML6LyLhc3S1evLjDPXjNNdcEXROPdbd69WpcddVVSEhIQEZGBu644w4cO3Ys6Bredx2FUm+85zr3/PPPY9y4cYHFJadNm4Z33303cD6c9xuDUB/YsGEDHnnkETz55JMoKyvDjBkzMHfuXFRWVka6aFFn9OjRqK6uDnwdPHgwcO5Xv/oVnn76aTz33HPYu3cvsrKycNNNNwX2h4sXVqsV48ePx3PPPdfp+VDq6ZFHHsFbb72F9evXY9euXWhpacG8efPg8XjC9TYi4nJ1BwA333xz0D24efPmoPPxWHfbt2/Hgw8+iD179mDr1q1wu90oKSmB1WoNXMP7rqNQ6g3gPdeZ3NxcrFmzBqWlpSgtLcXs2bMxf/78QNgJ6/0m6IpNnTpVLF26NOjYiBEjxH/8x39EqETRaeXKlWL8+PGdnvN6vSIrK0usWbMmcMxutwuLxSJeeOGFMJUw+gAQb731VuBxKPXU0NAgZFkW69evD1xz5swZoVKpxD/+8Y+wlT3SLq07IYRYtGiRmD9/fpfPYd0pamtrBQCxfft2IQTvu1BdWm9C8J7rieTkZPHnP/857PcbW4SukNPpxL59+1BSUhJ0vKSkBLt3745QqaLX8ePHkZOTg8LCQtx55504ceIEAKCiogI1NTVB9ajT6TBz5kzWYzuh1NO+ffvgcrmCrsnJycGYMWNYlwC2bduGjIwMFBUV4f7770dtbW3gHOtO0djYCABISUkBwPsuVJfWmx/vue55PB6sX78eVqsV06ZNC/v9xiB0hS5cuACPx4PMzMyg45mZmaipqYlQqaLT1VdfjZdffhnvvfce/vSnP6GmpgbTp09HXV1doK5Yj90LpZ5qamqg1WqRnJzc5TXxau7cuXj11Vfx4Ycf4je/+Q327t2L2bNnw+FwAGDdAcrYjOXLl+O6667DmDFjAPC+C0Vn9QbwnuvOwYMHYTabodPpsHTpUrz11lsYNWpU2O+3uNt9vr9IkhT0WAjR4Vi8mzt3buD7sWPHYtq0aRg6dCj+53/+JzB4kPUYmt7UE+sSWLhwYeD7MWPGYMqUKcjPz8emTZuwYMGCLp8XT3X30EMP4cCBA9i1a1eHc7zvutZVvfGe61pxcTHKy8vR0NCAN954A4sWLcL27dsD58N1v7FF6AqlpaVBrVZ3SKC1tbUd0iwFM5lMGDt2LI4fPx6YPcZ67F4o9ZSVlQWn04n6+vouryFFdnY28vPzcfz4cQCsu4cffhgbN27ERx99hNzc3MBx3nfd66reOsN7ro1Wq8WwYcMwZcoUrF69GuPHj8dvf/vbsN9vDEJXSKvVYvLkydi6dWvQ8a1bt2L69OkRKlVscDgcOHr0KLKzs1FYWIisrKygenQ6ndi+fTvrsZ1Q6mny5MmQZTnomurqahw6dIh1eYm6ujpUVVUhOzsbQPzWnRACDz30EN588018+OGHKCwsDDrP+65zl6u3zvCe65oQAg6HI/z3Wy8Hd1M769evF7Isi7Vr14ojR46IRx55RJhMJnHy5MlIFy2qrFixQmzbtk2cOHFC7NmzR8ybN08kJCQE6mnNmjXCYrGIN998Uxw8eFB885vfFNnZ2aKpqSnCJQ+v5uZmUVZWJsrKygQA8fTTT4uysjJx6tQpIURo9bR06VKRm5sr3n//fbF//34xe/ZsMX78eOF2uyP1tsKiu7prbm4WK1asELt37xYVFRXio48+EtOmTRODBg2K+7r77ne/KywWi9i2bZuorq4OfNlstsA1vO86uly98Z7r2hNPPCF27NghKioqxIEDB8SPfvQjoVKpxJYtW4QQ4b3fGIT6yO9//3uRn58vtFqtmDRpUtD0SVIsXLhQZGdnC1mWRU5OjliwYIE4fPhw4LzX6xUrV64UWVlZQqfTieuvv14cPHgwgiWOjI8++kgA6PC1aNEiIURo9dTa2ioeeughkZKSIgwGg5g3b56orKyMwLsJr+7qzmaziZKSEpGeni5kWRaDBw8WixYt6lAv8Vh3ndUZAPHSSy8FruF919Hl6o33XNfuueeewO/M9PR0ceONNwZCkBDhvd8kIYToWRsSERER0cDAMUJEREQUtxiEiIiIKG4xCBEREVHcYhAiIiKiuMUgRERERHGLQYiIiIjiFoMQERERxS0GISLqN6tWrcKECRMiXYx+sW3bNkiShIaGhkgXhYiuAIMQEfWKJEndfi1evBiPPfYYPvjgg4iU74033sDVV18Ni8WChIQEjB49GitWrIhIWYgoemkiXQAiik3V1dWB7zds2ICf/OQnOHbsWOCYwWCA2WyG2WwOe9nef/993HnnnfjFL36B22+/HZIk4ciRIxELZUQUvdgiRES9kpWVFfiyWCyQJKnDsUu7xhYvXow77rgDv/jFL5CZmYmkpCT89Kc/hdvtxuOPP46UlBTk5ubixRdfDPpZZ86cwcKFC5GcnIzU1FTMnz8fJ0+e7LJsf//733Hdddfh8ccfR3FxMYqKinDHHXfg2WefDbpu48aNmDJlCvR6PdLS0rBgwYLAuVdeeQVTpkxBQkICsrKy8O///u+ora3ttk52796N66+/HgaDAXl5eVi2bBmsVmvolUpEYccgRERh9eGHH+Ls2bPYsWMHnn76aaxatQrz5s1DcnIyPv30UyxduhRLly5FVVUVAMBms+GGG26A2WzGjh07sGvXLpjNZtx8881wOp2d/oysrCwcPnwYhw4d6rIcmzZtwoIFC3DrrbeirKwMH3zwAaZMmRI473Q68dRTT+Gzzz7D22+/jYqKCixevLjL1zt48CDmzJmDBQsW4MCBA9iwYQN27dqFhx56qHcVRUThcYUbyBIRiZdeeklYLJYOx1euXCnGjx8feLxo0SKRn58vPB5P4FhxcbGYMWNG4LHb7RYmk0m89tprQggh1q5dK4qLi4XX6w1c43A4hMFgEO+9916n5WlpaRG33HKLACDy8/PFwoULxdq1a4Xdbg9cM23aNHHXXXeF/B7/+c9/CgCiublZCNG20319fb0QQohvf/vb4jvf+U7Qc3bu3ClUKpVobW0N+ecQUXixRYiIwmr06NFQqdo+ejIzMzF27NjAY7VajdTU1EA31L59+/Dll18iISEhMOYoJSUFdrsdX331Vac/w2QyYdOmTfjyyy/xn//5nzCbzVixYgWmTp0Km80GACgvL8eNN97YZTnLysowf/585OfnIyEhAbNmzQIAVFZWdnr9vn37sG7dukAZzWYz5syZA6/Xi4qKih7VERGFDwdLE1FYybIc9FiSpE6Peb1eAIDX68XkyZPx6quvdnit9PT0bn/W0KFDMXToUNx333148sknUVRUhA0bNmDJkiUwGAxdPs9qtaKkpAQlJSV45ZVXkJ6ejsrKSsyZM6fL7jiv14sHHngAy5Yt63Bu8ODB3ZaTiCKHQYiIotqkSZOwYcMGZGRkIDExsdevU1BQAKPRGBi8PG7cOHzwwQdYsmRJh2s///xzXLhwAWvWrEFeXh4AoLS09LLlPHz4MIYNG9brMhJR+LFrjIii2l133YW0tDTMnz8fO3fuREVFBbZv347vf//7OH36dKfPWbVqFX7wgx9g27ZtqKioQFlZGe655x64XC7cdNNNAICVK1fitddew8qVK3H06FEcPHgQv/rVrwAoLTharRbPPvssTpw4gY0bN+Kpp57qtpw//OEP8cknn+DBBx9EeXk5jh8/jo0bN+Lhhx/u2wohoj7FIEREUc1oNGLHjh0YPHgwFixYgJEjR+Kee+5Ba2trly1EM2fOxIkTJ3D33XdjxIgRmDt3LmpqarBlyxYUFxcDAGbNmoXXX38dGzduxIQJEzB79mx8+umnAJQut3Xr1uH111/HqFGjsGbNGvz617/utpzjxo3D9u3bcfz4ccyYMQMTJ07Ej3/8Y2RnZ/dthRBRn5KEECLShSAiIiKKBLYIERERUdxiECIiIqK4xSBEREREcYtBiIiIiOIWgxARERHFLQYhIiIiilsMQkRERBS3GISIiIgobjEIERERUdxiECIiIqK4xSBEREREcYtBiIiIiOLW/wcZfAHIzwS1hAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Predicted vs True \n",
    "plt.plot(y_test, label='True Value')\n",
    "plt.plot(y_pred, label='LSTM Value')\n",
    "plt.title(\"Prediction by LSTM\")\n",
    "plt.xlabel('Time Scale')\n",
    "plt.ylabel('Scaled USD')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "631c31f5",
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import pandas as pd\n",
    "\n",
    "def convert_files():\n",
    "    sPath = \"/Users/aryanchopra/Downloads/archive-3/Data/Stocks/\"\n",
    "    for sName in os.listdir(sPath):\n",
    "        if sName.lower().endswith('.txt'):\n",
    "            file_path = os.path.join(sPath, sName)\n",
    "            if os.path.getsize(file_path) > 0:  # Check if file is not empty\n",
    "                df = pd.read_csv(file_path, delimiter='\\t')\n",
    "                if not df.empty:  # Check if the DataFrame has columns\n",
    "                    new_file_path = os.path.join(sPath, sName[:-4] + '.csv')\n",
    "                    df.to_csv(new_file_path, index=False)\n",
    "                    os.remove(file_path)  # Remove the original .txt file\n",
    "\n",
    "convert_files()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "83384ffd",
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import pandas as pd\n",
    "\n",
    "# Define the folder path containing the CSV files\n",
    "folder_path = '/Users/aryanchopra/Downloads/archive-3/Data/Stocks/'\n",
    "\n",
    "# Define the new header\n",
    "new_header = ['Date','Open', 'High', 'Low', 'Close', 'Volume']\n",
    "\n",
    "# Iterate through the CSV files\n",
    "csv_files = [file for file in os.listdir(folder_path) if file.endswith('.csv')]\n",
    "for csv_file in csv_files:\n",
    "    file_path = os.path.join(folder_path, csv_file)\n",
    "    \n",
    "    # Read the CSV file\n",
    "    df = pd.read_csv(file_path)\n",
    "    \n",
    "    # Drop the rightmost column\n",
    "    df = df.iloc[:, :-1]\n",
    "    \n",
    "    # Rename the header\n",
    "    df.columns = new_header\n",
    "    \n",
    "    # Save the updated DataFrame to the same CSV file\n",
    "    df.to_csv(file_path, index=False)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "id": "ea11f006",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/10\n",
      "52/52 [==============================] - 0s 1ms/step - loss: 607887.6875\n",
      "Epoch 2/10\n",
      "52/52 [==============================] - 0s 1ms/step - loss: 9865.2393\n",
      "Epoch 3/10\n",
      "52/52 [==============================] - 0s 982us/step - loss: 77.8265\n",
      "Epoch 4/10\n",
      "52/52 [==============================] - 0s 973us/step - loss: 2.5225\n",
      "Epoch 5/10\n",
      "52/52 [==============================] - 0s 1ms/step - loss: 0.5147\n",
      "Epoch 6/10\n",
      "52/52 [==============================] - 0s 1ms/step - loss: 0.5035\n",
      "Epoch 7/10\n",
      "52/52 [==============================] - 0s 1ms/step - loss: 0.4912\n",
      "Epoch 8/10\n",
      "52/52 [==============================] - 0s 1ms/step - loss: 0.4997\n",
      "Epoch 9/10\n",
      "52/52 [==============================] - 0s 1ms/step - loss: 0.5523\n",
      "Epoch 10/10\n",
      "52/52 [==============================] - 0s 1ms/step - loss: 0.5582\n",
      "52/52 [==============================] - 0s 596us/step\n",
      "Profitable stocks: []\n"
     ]
    }
   ],
   "source": [
    "import os\n",
    "import pandas as pd\n",
    "from sklearn.model_selection import train_test_split\n",
    "from tensorflow.keras.models import Sequential\n",
    "from tensorflow.keras.layers import LSTM, Dense\n",
    "import numpy as np\n",
    "\n",
    "growth_threshold = 1.5 \n",
    "csv_files = os.listdir('/Users/aryanchopra/Downloads/archive-3/Data/Stocks/')\n",
    "profitable_stocks = []\n",
    "\n",
    "csv_file = csv_files[0]\n",
    "df = pd.read_csv(os.path.join('/Users/aryanchopra/Downloads/archive-3/Data/Stocks/', csv_file))\n",
    "\n",
    "df['Close_pct_change'] = df['Close'].pct_change()\n",
    "\n",
    "trainY = df['Close_pct_change'].shift(-1)  \n",
    "trainY = trainY.apply(lambda x: 1 if x > 0 else 0) \n",
    "trainX = df[['Open', 'High', 'Low', 'Close', 'Volume']]  \n",
    "trainX = np.expand_dims(trainX, axis=2)\n",
    "\n",
    "lstm = Sequential()\n",
    "lstm.add(LSTM(32, input_shape=(trainX.shape[1], trainX.shape[2]), activation='relu', return_sequences=False))\n",
    "lstm.add(Dense(1))\n",
    "lstm.compile(loss='mean_squared_error', optimizer='adam')\n",
    "lstm.fit(trainX, trainY, epochs=10, batch_size=32)\n",
    "\n",
    "future_data = df[['Open', 'High', 'Low', 'Close', 'Volume']]  \n",
    "future_data = np.expand_dims(future_data, axis=2)\n",
    "future_predictions = lstm.predict(future_data)\n",
    "\n",
    "current_price = df['Close'].iloc[-1]\n",
    "future_price = future_predictions[-1]\n",
    "growth_rate = future_price / current_price\n",
    "\n",
    "if growth_rate >= growth_threshold:\n",
    "    profitable_stocks.append(csv_file)\n",
    "\n",
    "print(\"Profitable stocks:\", profitable_stocks)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "id": "dc80fac4",
   "metadata": {
    "collapsed": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/10\n",
      "52/52 [==============================] - 0s 1ms/step - loss: 73643296.0000\n",
      "Epoch 2/10\n",
      "52/52 [==============================] - 0s 1ms/step - loss: 6323165.5000\n",
      "Epoch 3/10\n",
      "52/52 [==============================] - 0s 1ms/step - loss: 13007.0420\n",
      "Epoch 4/10\n",
      "52/52 [==============================] - 0s 1ms/step - loss: 83.6697\n",
      "Epoch 5/10\n",
      "52/52 [==============================] - 0s 1ms/step - loss: 12.0181\n",
      "Epoch 6/10\n",
      "52/52 [==============================] - 0s 1ms/step - loss: 56.6736\n",
      "Epoch 7/10\n",
      "52/52 [==============================] - 0s 1ms/step - loss: 11.7869\n",
      "Epoch 8/10\n",
      "52/52 [==============================] - 0s 1ms/step - loss: 7.1866\n",
      "Epoch 9/10\n",
      "52/52 [==============================] - 0s 1ms/step - loss: 938.0260\n",
      "Epoch 10/10\n",
      "52/52 [==============================] - 0s 1ms/step - loss: 2.6037\n",
      "52/52 [==============================] - 0s 549us/step\n",
      "Processed file 1 of 10\n",
      "Epoch 1/10\n",
      "29/29 [==============================] - 0s 1ms/step - loss: 97096072.0000\n",
      "Epoch 2/10\n",
      "29/29 [==============================] - 0s 1ms/step - loss: 14120783.0000\n",
      "Epoch 3/10\n",
      "29/29 [==============================] - 0s 1ms/step - loss: 229484.9219\n",
      "Epoch 4/10\n",
      "29/29 [==============================] - 0s 1ms/step - loss: 4884.4941\n",
      "Epoch 5/10\n",
      "29/29 [==============================] - 0s 1ms/step - loss: 160.5340\n",
      "Epoch 6/10\n",
      "29/29 [==============================] - 0s 1ms/step - loss: 15.2710\n",
      "Epoch 7/10\n",
      "29/29 [==============================] - 0s 1ms/step - loss: 1.5310\n",
      "Epoch 8/10\n",
      "29/29 [==============================] - 0s 1ms/step - loss: 0.5238\n",
      "Epoch 9/10\n",
      "29/29 [==============================] - 0s 1ms/step - loss: 0.4166\n",
      "Epoch 10/10\n",
      "29/29 [==============================] - 0s 1ms/step - loss: 0.4091\n",
      "29/29 [==============================] - 0s 654us/step\n",
      "Processed file 2 of 10\n",
      "Epoch 1/10\n",
      "4/4 [==============================] - 0s 2ms/step - loss: 1217.3096\n",
      "Epoch 2/10\n",
      "4/4 [==============================] - 0s 2ms/step - loss: 1038.5619\n",
      "Epoch 3/10\n",
      "4/4 [==============================] - 0s 2ms/step - loss: 837.4832\n",
      "Epoch 4/10\n",
      "4/4 [==============================] - 0s 2ms/step - loss: 624.7638\n",
      "Epoch 5/10\n",
      "4/4 [==============================] - 0s 2ms/step - loss: 556.3495\n",
      "Epoch 6/10\n",
      "4/4 [==============================] - 0s 1ms/step - loss: 496.4242\n",
      "Epoch 7/10\n",
      "4/4 [==============================] - 0s 2ms/step - loss: 441.3772\n",
      "Epoch 8/10\n",
      "4/4 [==============================] - 0s 1ms/step - loss: 413.1730\n",
      "Epoch 9/10\n",
      "4/4 [==============================] - 0s 1ms/step - loss: 365.5919\n",
      "Epoch 10/10\n",
      "4/4 [==============================] - 0s 2ms/step - loss: 25.1090\n",
      "4/4 [==============================] - 0s 985us/step\n",
      "Processed file 3 of 10\n",
      "Epoch 1/10\n",
      "100/100 [==============================] - 0s 1ms/step - loss: 18035.3359\n",
      "Epoch 2/10\n",
      "100/100 [==============================] - 0s 1ms/step - loss: 6.5445\n",
      "Epoch 3/10\n",
      "100/100 [==============================] - 0s 1ms/step - loss: 5.9420\n",
      "Epoch 4/10\n",
      "100/100 [==============================] - 0s 1ms/step - loss: 0.7724\n",
      "Epoch 5/10\n",
      "100/100 [==============================] - 0s 1ms/step - loss: 0.3293\n",
      "Epoch 6/10\n",
      "100/100 [==============================] - 0s 1ms/step - loss: 0.3127\n",
      "Epoch 7/10\n",
      "100/100 [==============================] - 0s 1ms/step - loss: 0.4555\n",
      "Epoch 8/10\n",
      "100/100 [==============================] - 0s 1ms/step - loss: 0.3097\n",
      "Epoch 9/10\n",
      "100/100 [==============================] - 0s 1ms/step - loss: 0.2763\n",
      "Epoch 10/10\n",
      "100/100 [==============================] - 0s 1ms/step - loss: 0.3286\n",
      "100/100 [==============================] - 0s 498us/step\n",
      "Processed file 4 of 10\n",
      "Epoch 1/10\n",
      "1/1 [==============================] - 0s 316ms/step - loss: 234968.2969\n",
      "Epoch 2/10\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 34906.4883\n",
      "Epoch 3/10\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 6268.5391\n",
      "Epoch 4/10\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 64421.1055\n",
      "Epoch 5/10\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 94298.6172\n",
      "Epoch 6/10\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 73545.3438\n",
      "Epoch 7/10\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 33618.6797\n",
      "Epoch 8/10\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 5104.1597\n",
      "Epoch 9/10\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 1684.5323\n",
      "Epoch 10/10\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 17595.5762\n",
      "1/1 [==============================] - 0s 50ms/step\n",
      "Processed file 5 of 10\n",
      "Epoch 1/10\n",
      "16/16 [==============================] - 0s 2ms/step - loss: 77150328.0000\n",
      "Epoch 2/10\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 64325608.0000\n",
      "Epoch 3/10\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 52969268.0000\n",
      "Epoch 4/10\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 45674004.0000\n",
      "Epoch 5/10\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 15710.7080\n",
      "Epoch 6/10\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 5661.9175\n",
      "Epoch 7/10\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 9277.4229\n",
      "Epoch 8/10\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 109034.5859\n",
      "Epoch 9/10\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 31109.0059\n",
      "Epoch 10/10\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 3669.3850\n",
      "16/16 [==============================] - 0s 728us/step\n",
      "Processed file 6 of 10\n",
      "Epoch 1/10\n",
      "41/41 [==============================] - 0s 1ms/step - loss: 4526428672.0000\n",
      "Epoch 2/10\n",
      "41/41 [==============================] - 0s 1ms/step - loss: 2969712384.0000\n",
      "Epoch 3/10\n",
      "41/41 [==============================] - 0s 986us/step - loss: 1992837888.0000\n",
      "Epoch 4/10\n",
      "41/41 [==============================] - 0s 970us/step - loss: 2418744.7500\n",
      "Epoch 5/10\n",
      "41/41 [==============================] - 0s 977us/step - loss: 6980822.0000\n",
      "Epoch 6/10\n",
      "41/41 [==============================] - 0s 965us/step - loss: 2048947.6250\n",
      "Epoch 7/10\n",
      "41/41 [==============================] - 0s 990us/step - loss: 1315278.8750\n",
      "Epoch 8/10\n",
      "41/41 [==============================] - 0s 939us/step - loss: 455096.0312\n",
      "Epoch 9/10\n",
      "41/41 [==============================] - 0s 966us/step - loss: 378965.3125\n",
      "Epoch 10/10\n",
      "41/41 [==============================] - 0s 954us/step - loss: 355770.4062\n",
      "41/41 [==============================] - 0s 550us/step\n",
      "Processed file 7 of 10\n",
      "Epoch 1/10\n",
      "100/100 [==============================] - 0s 873us/step - loss: 3730235136.0000\n",
      "Epoch 2/10\n",
      "100/100 [==============================] - 0s 758us/step - loss: 16511110.0000\n",
      "Epoch 3/10\n",
      "100/100 [==============================] - 0s 753us/step - loss: 133801.9062\n",
      "Epoch 4/10\n",
      "100/100 [==============================] - 0s 759us/step - loss: 122013.5000\n",
      "Epoch 5/10\n",
      "100/100 [==============================] - 0s 734us/step - loss: 110715.3125\n",
      "Epoch 6/10\n",
      "100/100 [==============================] - 0s 749us/step - loss: 99812.1797\n",
      "Epoch 7/10\n",
      "100/100 [==============================] - 0s 756us/step - loss: 91422.9062\n",
      "Epoch 8/10\n",
      "100/100 [==============================] - 0s 773us/step - loss: 78497.9766\n",
      "Epoch 9/10\n",
      "100/100 [==============================] - 0s 745us/step - loss: 66471.2969\n",
      "Epoch 10/10\n",
      "100/100 [==============================] - 0s 749us/step - loss: 59402.6641\n",
      "101/101 [==============================] - 0s 453us/step\n",
      "Processed file 8 of 10\n",
      "Epoch 1/10\n",
      "100/100 [==============================] - 0s 880us/step - loss: 3086712832.0000\n",
      "Epoch 2/10\n",
      "100/100 [==============================] - 0s 756us/step - loss: 43160052.0000\n",
      "Epoch 3/10\n",
      "100/100 [==============================] - 0s 780us/step - loss: 253971.1875\n",
      "Epoch 4/10\n",
      "100/100 [==============================] - 0s 760us/step - loss: 475743.9062\n",
      "Epoch 5/10\n",
      "100/100 [==============================] - 0s 765us/step - loss: 137373.2812\n",
      "Epoch 6/10\n",
      "100/100 [==============================] - 0s 780us/step - loss: 4971635.0000\n",
      "Epoch 7/10\n",
      "100/100 [==============================] - 0s 761us/step - loss: 58298.7695\n",
      "Epoch 8/10\n",
      "100/100 [==============================] - 0s 770us/step - loss: 35688.7148\n",
      "Epoch 9/10\n",
      "100/100 [==============================] - 0s 758us/step - loss: 13981.1738\n",
      "Epoch 10/10\n",
      "100/100 [==============================] - 0s 747us/step - loss: 15890.8789\n",
      "101/101 [==============================] - 0s 462us/step\n",
      "Processed file 9 of 10\n",
      "Epoch 1/10\n",
      "48/48 [==============================] - 0s 1ms/step - loss: 30750074880.0000\n",
      "Epoch 2/10\n",
      "48/48 [==============================] - 0s 1ms/step - loss: 5075818496.0000\n",
      "Epoch 3/10\n",
      "48/48 [==============================] - 0s 980us/step - loss: 3613992.5000\n",
      "Epoch 4/10\n",
      "48/48 [==============================] - 0s 990us/step - loss: 27385.7422\n",
      "Epoch 5/10\n",
      "48/48 [==============================] - 0s 984us/step - loss: 193.9369\n",
      "Epoch 6/10\n",
      "48/48 [==============================] - 0s 965us/step - loss: 2.4084\n",
      "Epoch 7/10\n",
      "48/48 [==============================] - 0s 960us/step - loss: 1.6520\n",
      "Epoch 8/10\n",
      "48/48 [==============================] - 0s 979us/step - loss: 1.6511\n",
      "Epoch 9/10\n",
      "48/48 [==============================] - 0s 1ms/step - loss: 1.6502\n",
      "Epoch 10/10\n",
      "48/48 [==============================] - 0s 985us/step - loss: 1.6494\n",
      "48/48 [==============================] - 0s 519us/step\n",
      "Processed file 10 of 10\n",
      "Epoch 1/10\n",
      "47/47 [==============================] - 0s 1ms/step - loss: 6775983.0000\n",
      "Epoch 2/10\n",
      "47/47 [==============================] - 0s 1ms/step - loss: 3693186.5000\n",
      "Epoch 3/10\n",
      "47/47 [==============================] - 0s 1ms/step - loss: 1298787.6250\n",
      "Epoch 4/10\n",
      "47/47 [==============================] - 0s 1ms/step - loss: 127.9224\n",
      "Epoch 5/10\n",
      "47/47 [==============================] - 0s 1ms/step - loss: 52.0254\n",
      "Epoch 6/10\n",
      "47/47 [==============================] - 0s 970us/step - loss: 381.5074\n",
      "Epoch 7/10\n",
      "47/47 [==============================] - 0s 961us/step - loss: 4262.6421\n",
      "Epoch 8/10\n",
      "47/47 [==============================] - 0s 993us/step - loss: 506.5240\n",
      "Epoch 9/10\n",
      "47/47 [==============================] - 0s 936us/step - loss: 17.9102\n",
      "Epoch 10/10\n",
      "47/47 [==============================] - 0s 924us/step - loss: 6.7270\n",
      "47/47 [==============================] - 0s 515us/step\n",
      "Processed file 11 of 10\n",
      "Epoch 1/10\n",
      "28/28 [==============================] - 0s 1ms/step - loss: 2139987.0000\n",
      "Epoch 2/10\n",
      "28/28 [==============================] - 0s 1ms/step - loss: 264525.4062\n",
      "Epoch 3/10\n",
      "28/28 [==============================] - 0s 1ms/step - loss: 124004.1953\n",
      "Epoch 4/10\n",
      "28/28 [==============================] - 0s 1ms/step - loss: 33076.9297\n",
      "Epoch 5/10\n",
      "28/28 [==============================] - 0s 980us/step - loss: 8235.5518\n",
      "Epoch 6/10\n",
      "28/28 [==============================] - 0s 1ms/step - loss: 2496.5261\n",
      "Epoch 7/10\n",
      "28/28 [==============================] - 0s 989us/step - loss: 502.4800\n",
      "Epoch 8/10\n",
      "28/28 [==============================] - 0s 988us/step - loss: 157.0715\n",
      "Epoch 9/10\n",
      "28/28 [==============================] - 0s 975us/step - loss: 130.5587\n",
      "Epoch 10/10\n",
      "28/28 [==============================] - 0s 986us/step - loss: 472.9890\n",
      "28/28 [==============================] - 0s 632us/step\n",
      "Processed file 12 of 10\n",
      "Epoch 1/10\n",
      "20/20 [==============================] - 0s 2ms/step - loss: 1903755136.0000\n",
      "Epoch 2/10\n",
      "20/20 [==============================] - 0s 1ms/step - loss: 39388720.0000\n",
      "Epoch 3/10\n",
      "20/20 [==============================] - 0s 1ms/step - loss: 955130.8125\n",
      "Epoch 4/10\n",
      "20/20 [==============================] - 0s 1ms/step - loss: 113246.8125\n",
      "Epoch 5/10\n",
      "20/20 [==============================] - 0s 1ms/step - loss: 53732.1211\n",
      "Epoch 6/10\n",
      "20/20 [==============================] - 0s 1ms/step - loss: 6936.9951\n",
      "Epoch 7/10\n",
      "20/20 [==============================] - 0s 1ms/step - loss: 380.5829\n",
      "Epoch 8/10\n",
      "20/20 [==============================] - 0s 1ms/step - loss: 159.8281\n",
      "Epoch 9/10\n",
      "20/20 [==============================] - 0s 1ms/step - loss: 87.4440\n",
      "Epoch 10/10\n",
      "20/20 [==============================] - 0s 1ms/step - loss: 19.8799\n",
      "20/20 [==============================] - 0s 656us/step\n",
      "Processed file 13 of 10\n",
      "Epoch 1/10\n",
      "68/68 [==============================] - 0s 1ms/step - loss: 140618.0625\n",
      "Epoch 2/10\n",
      "68/68 [==============================] - 0s 961us/step - loss: 42841.2695\n",
      "Epoch 3/10\n",
      "68/68 [==============================] - 0s 956us/step - loss: 12229.7412\n",
      "Epoch 4/10\n",
      "68/68 [==============================] - 0s 921us/step - loss: 1835.2483\n",
      "Epoch 5/10\n",
      "68/68 [==============================] - 0s 951us/step - loss: 191.5625\n",
      "Epoch 6/10\n",
      "68/68 [==============================] - 0s 952us/step - loss: 66.7221\n",
      "Epoch 7/10\n",
      "68/68 [==============================] - 0s 984us/step - loss: 42.3390\n",
      "Epoch 8/10\n",
      "68/68 [==============================] - 0s 977us/step - loss: 25.9578\n",
      "Epoch 9/10\n",
      "68/68 [==============================] - 0s 960us/step - loss: 18.9498\n",
      "Epoch 10/10\n",
      "68/68 [==============================] - 0s 1ms/step - loss: 26.6405\n",
      "68/68 [==============================] - 0s 497us/step\n",
      "Processed file 14 of 10\n",
      "Epoch 1/10\n",
      "97/97 [==============================] - 0s 1ms/step - loss: 17129.3496\n",
      "Epoch 2/10\n",
      "97/97 [==============================] - 0s 985us/step - loss: 387.3496\n",
      "Epoch 3/10\n",
      "97/97 [==============================] - 0s 954us/step - loss: 0.2927\n",
      "Epoch 4/10\n",
      "97/97 [==============================] - 0s 956us/step - loss: 0.3049\n",
      "Epoch 5/10\n",
      "97/97 [==============================] - 0s 945us/step - loss: 0.3055\n",
      "Epoch 6/10\n",
      "97/97 [==============================] - 0s 928us/step - loss: 0.3805\n",
      "Epoch 7/10\n",
      "97/97 [==============================] - 0s 954us/step - loss: 1.7325\n",
      "Epoch 8/10\n",
      "97/97 [==============================] - 0s 960us/step - loss: 8.9619\n",
      "Epoch 9/10\n",
      "97/97 [==============================] - 0s 938us/step - loss: 3.4783\n",
      "Epoch 10/10\n",
      "97/97 [==============================] - 0s 943us/step - loss: 6.5313\n",
      "97/97 [==============================] - 0s 474us/step\n",
      "Processed file 15 of 10\n",
      "Epoch 1/10\n",
      "2/2 [==============================] - 0s 3ms/step - loss: 57701636.0000\n",
      "Epoch 2/10\n",
      "2/2 [==============================] - 0s 2ms/step - loss: 55051240.0000\n",
      "Epoch 3/10\n",
      "2/2 [==============================] - 0s 2ms/step - loss: 53069656.0000\n",
      "Epoch 4/10\n",
      "2/2 [==============================] - 0s 3ms/step - loss: 49910576.0000\n",
      "Epoch 5/10\n",
      "2/2 [==============================] - 0s 2ms/step - loss: 48230556.0000\n",
      "Epoch 6/10\n",
      "2/2 [==============================] - 0s 3ms/step - loss: 45034772.0000\n",
      "Epoch 7/10\n",
      "2/2 [==============================] - 0s 3ms/step - loss: 43557264.0000\n",
      "Epoch 8/10\n",
      "2/2 [==============================] - 0s 3ms/step - loss: 41415960.0000\n",
      "Epoch 9/10\n",
      "2/2 [==============================] - 0s 3ms/step - loss: 39342104.0000\n",
      "Epoch 10/10\n",
      "2/2 [==============================] - 0s 2ms/step - loss: 37452840.0000\n",
      "2/2 [==============================] - 0s 1ms/step\n",
      "Processed file 16 of 10\n",
      "Epoch 1/10\n",
      "94/94 [==============================] - 0s 1ms/step - loss: 395856256.0000\n",
      "Epoch 2/10\n",
      "94/94 [==============================] - 0s 976us/step - loss: 13496660.0000\n",
      "Epoch 3/10\n",
      "94/94 [==============================] - 0s 939us/step - loss: 14419.7188\n",
      "Epoch 4/10\n",
      "94/94 [==============================] - 0s 976us/step - loss: 599.0616\n",
      "Epoch 5/10\n",
      "94/94 [==============================] - 0s 940us/step - loss: 4.8272\n",
      "Epoch 6/10\n",
      "94/94 [==============================] - 0s 971us/step - loss: 2.0173\n",
      "Epoch 7/10\n",
      "94/94 [==============================] - 0s 959us/step - loss: 1.2086\n",
      "Epoch 8/10\n",
      "94/94 [==============================] - 0s 974us/step - loss: 0.8904\n",
      "Epoch 9/10\n",
      "94/94 [==============================] - 0s 957us/step - loss: 0.7016\n",
      "Epoch 10/10\n",
      "94/94 [==============================] - 0s 964us/step - loss: 0.6812\n",
      "94/94 [==============================] - 0s 472us/step\n",
      "Processed file 17 of 10\n",
      "Epoch 1/10\n",
      "90/90 [==============================] - 1s 1ms/step - loss: 628991.8750\n",
      "Epoch 2/10\n",
      "90/90 [==============================] - 0s 964us/step - loss: 5567.8608\n",
      "Epoch 3/10\n",
      "90/90 [==============================] - 0s 971us/step - loss: 2607.5967\n",
      "Epoch 4/10\n",
      "90/90 [==============================] - 0s 942us/step - loss: 2803.9998\n",
      "Epoch 5/10\n",
      "90/90 [==============================] - 0s 953us/step - loss: 28.8703\n",
      "Epoch 6/10\n",
      "90/90 [==============================] - 0s 961us/step - loss: 12.9189\n",
      "Epoch 7/10\n",
      "90/90 [==============================] - 0s 955us/step - loss: 2.5053\n",
      "Epoch 8/10\n",
      "90/90 [==============================] - 0s 978us/step - loss: 41.8110\n",
      "Epoch 9/10\n",
      "90/90 [==============================] - 0s 959us/step - loss: 2059865.3750\n",
      "Epoch 10/10\n",
      "90/90 [==============================] - 0s 937us/step - loss: 292952.4375\n",
      "90/90 [==============================] - 0s 482us/step\n",
      "Processed file 18 of 10\n",
      "Epoch 1/10\n",
      "2/2 [==============================] - 0s 3ms/step - loss: 173.0031\n",
      "Epoch 2/10\n",
      "2/2 [==============================] - 0s 2ms/step - loss: 961.1654\n",
      "Epoch 3/10\n",
      "2/2 [==============================] - 0s 3ms/step - loss: 987.0402\n",
      "Epoch 4/10\n",
      "2/2 [==============================] - 0s 2ms/step - loss: 945.8072\n",
      "Epoch 5/10\n",
      "2/2 [==============================] - 0s 3ms/step - loss: 890.2830\n",
      "Epoch 6/10\n",
      "2/2 [==============================] - 0s 3ms/step - loss: 829.4645\n",
      "Epoch 7/10\n",
      "2/2 [==============================] - 0s 2ms/step - loss: 775.1694\n",
      "Epoch 8/10\n",
      "2/2 [==============================] - 0s 2ms/step - loss: 724.2557\n",
      "Epoch 9/10\n",
      "2/2 [==============================] - 0s 2ms/step - loss: 664.6038\n",
      "Epoch 10/10\n",
      "2/2 [==============================] - 0s 2ms/step - loss: 618.1197\n",
      "2/2 [==============================] - 0s 1ms/step\n",
      "Processed file 19 of 10\n",
      "Epoch 1/10\n",
      "100/100 [==============================] - 0s 872us/step - loss: 20588942.0000\n",
      "Epoch 2/10\n",
      "100/100 [==============================] - 0s 760us/step - loss: 5395.1982\n",
      "Epoch 3/10\n",
      "100/100 [==============================] - 0s 758us/step - loss: 1.2467\n",
      "Epoch 4/10\n",
      "100/100 [==============================] - 0s 749us/step - loss: 1.1593\n",
      "Epoch 5/10\n",
      "100/100 [==============================] - 0s 756us/step - loss: 1.2459\n",
      "Epoch 6/10\n",
      "100/100 [==============================] - 0s 735us/step - loss: 1.1624\n",
      "Epoch 7/10\n",
      "100/100 [==============================] - 0s 734us/step - loss: 1.1808\n",
      "Epoch 8/10\n",
      "100/100 [==============================] - 0s 740us/step - loss: 1.2346\n",
      "Epoch 9/10\n",
      "100/100 [==============================] - 0s 748us/step - loss: 1.1456\n",
      "Epoch 10/10\n",
      "100/100 [==============================] - 0s 752us/step - loss: 1.1328\n",
      "101/101 [==============================] - 0s 466us/step\n",
      "Processed file 20 of 10\n",
      "Epoch 1/10\n",
      "86/86 [==============================] - 0s 1ms/step - loss: 3098806.5000\n",
      "Epoch 2/10\n",
      "86/86 [==============================] - 0s 1000us/step - loss: 2194.9412\n",
      "Epoch 3/10\n",
      "86/86 [==============================] - 0s 969us/step - loss: 12.7150\n",
      "Epoch 4/10\n",
      "86/86 [==============================] - 0s 962us/step - loss: 5.1182\n",
      "Epoch 5/10\n",
      "86/86 [==============================] - 0s 947us/step - loss: 4.7834\n",
      "Epoch 6/10\n",
      "86/86 [==============================] - 0s 975us/step - loss: 8.6328\n",
      "Epoch 7/10\n",
      "86/86 [==============================] - 0s 979us/step - loss: 5.1184\n",
      "Epoch 8/10\n",
      "86/86 [==============================] - 0s 969us/step - loss: 9.6543\n",
      "Epoch 9/10\n",
      "86/86 [==============================] - 0s 972us/step - loss: 6.4445\n",
      "Epoch 10/10\n",
      "86/86 [==============================] - 0s 968us/step - loss: 15.4650\n",
      "86/86 [==============================] - 0s 468us/step\n",
      "Processed file 21 of 10\n",
      "Epoch 1/10\n",
      "5/5 [==============================] - 0s 2ms/step - loss: 215618.4844\n",
      "Epoch 2/10\n",
      "5/5 [==============================] - 0s 2ms/step - loss: 180997.5781\n",
      "Epoch 3/10\n",
      "5/5 [==============================] - 0s 2ms/step - loss: 142135.6094\n",
      "Epoch 4/10\n",
      "5/5 [==============================] - 0s 2ms/step - loss: 111319.3359\n",
      "Epoch 5/10\n",
      "5/5 [==============================] - 0s 1ms/step - loss: 94282.8203\n",
      "Epoch 6/10\n",
      "5/5 [==============================] - 0s 1ms/step - loss: 69185.4062\n",
      "Epoch 7/10\n",
      "5/5 [==============================] - 0s 1ms/step - loss: 55021.3242\n",
      "Epoch 8/10\n",
      "5/5 [==============================] - 0s 2ms/step - loss: 40189.3438\n",
      "Epoch 9/10\n",
      "5/5 [==============================] - 0s 1ms/step - loss: 30248.2363\n",
      "Epoch 10/10\n",
      "5/5 [==============================] - 0s 1ms/step - loss: 25053.8105\n",
      "5/5 [==============================] - 0s 809us/step\n",
      "Processed file 22 of 10\n",
      "Epoch 1/10\n",
      "100/100 [==============================] - 0s 882us/step - loss: 1781274.5000\n",
      "Epoch 2/10\n",
      "100/100 [==============================] - 0s 760us/step - loss: 87740.1797\n",
      "Epoch 3/10\n",
      "100/100 [==============================] - 0s 745us/step - loss: 303.4869\n",
      "Epoch 4/10\n",
      "100/100 [==============================] - 0s 745us/step - loss: 2.8472\n",
      "Epoch 5/10\n",
      "100/100 [==============================] - 0s 762us/step - loss: 2.8781\n",
      "Epoch 6/10\n",
      "100/100 [==============================] - 0s 762us/step - loss: 2.7968\n",
      "Epoch 7/10\n",
      "100/100 [==============================] - 0s 751us/step - loss: 2.6532\n",
      "Epoch 8/10\n",
      "100/100 [==============================] - 0s 774us/step - loss: 2.6298\n",
      "Epoch 9/10\n",
      "100/100 [==============================] - 0s 761us/step - loss: 2.5369\n",
      "Epoch 10/10\n",
      "100/100 [==============================] - 0s 750us/step - loss: 2.4622\n",
      "101/101 [==============================] - 0s 467us/step\n",
      "Processed file 23 of 10\n",
      "Epoch 1/10\n",
      "100/100 [==============================] - 0s 1ms/step - loss: 10010621.0000\n",
      "Epoch 2/10\n",
      "100/100 [==============================] - 0s 971us/step - loss: 4487530.0000\n",
      "Epoch 3/10\n",
      "100/100 [==============================] - 0s 947us/step - loss: 1553195.5000\n",
      "Epoch 4/10\n",
      "100/100 [==============================] - 0s 977us/step - loss: 329616.9375\n",
      "Epoch 5/10\n",
      "100/100 [==============================] - 0s 971us/step - loss: 48960.8516\n",
      "Epoch 6/10\n",
      "100/100 [==============================] - 0s 966us/step - loss: 4812.1914\n",
      "Epoch 7/10\n",
      "100/100 [==============================] - 0s 983us/step - loss: 1260.2531\n",
      "Epoch 8/10\n",
      "100/100 [==============================] - 0s 937us/step - loss: 272.6274\n",
      "Epoch 9/10\n",
      "100/100 [==============================] - 0s 955us/step - loss: 2278.6501\n",
      "Epoch 10/10\n",
      "100/100 [==============================] - 0s 963us/step - loss: 508.2315\n",
      "100/100 [==============================] - 0s 413us/step\n",
      "Processed file 24 of 10\n",
      "Epoch 1/10\n",
      "43/43 [==============================] - 0s 1ms/step - loss: 1112212.6250\n",
      "Epoch 2/10\n",
      "43/43 [==============================] - 0s 1ms/step - loss: 3850.7446\n",
      "Epoch 3/10\n",
      "43/43 [==============================] - 0s 965us/step - loss: 36.5218\n",
      "Epoch 4/10\n",
      "43/43 [==============================] - 0s 1ms/step - loss: 0.4564\n",
      "Epoch 5/10\n",
      "43/43 [==============================] - 0s 1ms/step - loss: 0.2637\n",
      "Epoch 6/10\n",
      "43/43 [==============================] - 0s 1ms/step - loss: 0.2609\n",
      "Epoch 7/10\n",
      "43/43 [==============================] - 0s 1ms/step - loss: 0.2620\n",
      "Epoch 8/10\n",
      "43/43 [==============================] - 0s 1ms/step - loss: 0.2657\n",
      "Epoch 9/10\n",
      "43/43 [==============================] - 0s 975us/step - loss: 0.2615\n",
      "Epoch 10/10\n",
      "43/43 [==============================] - 0s 1ms/step - loss: 0.2606\n",
      "43/43 [==============================] - 0s 550us/step\n",
      "Processed file 25 of 10\n",
      "Epoch 1/10\n",
      "35/35 [==============================] - 0s 1ms/step - loss: 2374629.7500\n",
      "Epoch 2/10\n",
      "35/35 [==============================] - 0s 1ms/step - loss: 426864.0000\n",
      "Epoch 3/10\n",
      "35/35 [==============================] - 0s 1ms/step - loss: 9225.8652\n",
      "Epoch 4/10\n",
      "35/35 [==============================] - 0s 991us/step - loss: 947.6151\n",
      "Epoch 5/10\n",
      "35/35 [==============================] - 0s 1ms/step - loss: 188.0406\n",
      "Epoch 6/10\n",
      "35/35 [==============================] - 0s 1ms/step - loss: 78.0272\n",
      "Epoch 7/10\n",
      "35/35 [==============================] - 0s 1ms/step - loss: 42.0367\n",
      "Epoch 8/10\n",
      "35/35 [==============================] - 0s 1ms/step - loss: 15.8148\n",
      "Epoch 9/10\n",
      "35/35 [==============================] - 0s 1ms/step - loss: 7.4688\n",
      "Epoch 10/10\n",
      "35/35 [==============================] - 0s 1ms/step - loss: 4.4527\n",
      "35/35 [==============================] - 0s 607us/step\n",
      "Processed file 26 of 10\n",
      "Epoch 1/10\n",
      "100/100 [==============================] - 0s 878us/step - loss: 1131812.5000\n",
      "Epoch 2/10\n",
      "100/100 [==============================] - 0s 750us/step - loss: 37.0939\n",
      "Epoch 3/10\n",
      "100/100 [==============================] - 0s 767us/step - loss: 47.6459\n",
      "Epoch 4/10\n",
      "100/100 [==============================] - 0s 763us/step - loss: 29.8251\n",
      "Epoch 5/10\n",
      "100/100 [==============================] - 0s 757us/step - loss: 67.5537\n",
      "Epoch 6/10\n",
      "100/100 [==============================] - 0s 749us/step - loss: 96.7536\n",
      "Epoch 7/10\n",
      "100/100 [==============================] - 0s 755us/step - loss: 39.4014\n",
      "Epoch 8/10\n",
      "100/100 [==============================] - 0s 753us/step - loss: 81.3627\n",
      "Epoch 9/10\n",
      "100/100 [==============================] - 0s 745us/step - loss: 1726.1173\n",
      "Epoch 10/10\n",
      "100/100 [==============================] - 0s 755us/step - loss: 218.7713\n",
      "101/101 [==============================] - 0s 468us/step\n",
      "Processed file 27 of 10\n",
      "Epoch 1/10\n",
      "36/36 [==============================] - 0s 1ms/step - loss: 2884.5903\n",
      "Epoch 2/10\n",
      "36/36 [==============================] - 0s 1ms/step - loss: 9.8749\n",
      "Epoch 3/10\n",
      "36/36 [==============================] - 0s 1ms/step - loss: 0.5631\n",
      "Epoch 4/10\n",
      "36/36 [==============================] - 0s 998us/step - loss: 0.3213\n",
      "Epoch 5/10\n",
      "36/36 [==============================] - 0s 1ms/step - loss: 0.2940\n",
      "Epoch 6/10\n",
      "36/36 [==============================] - 0s 1ms/step - loss: 0.8814\n",
      "Epoch 7/10\n",
      "36/36 [==============================] - 0s 996us/step - loss: 0.5267\n",
      "Epoch 8/10\n",
      "36/36 [==============================] - 0s 1ms/step - loss: 0.4277\n",
      "Epoch 9/10\n",
      "36/36 [==============================] - 0s 990us/step - loss: 0.3200\n",
      "Epoch 10/10\n",
      "36/36 [==============================] - 0s 1ms/step - loss: 0.8841\n",
      "36/36 [==============================] - 0s 580us/step\n",
      "Processed file 28 of 10\n",
      "Epoch 1/10\n",
      "20/20 [==============================] - 0s 1ms/step - loss: 47107.1211\n",
      "Epoch 2/10\n",
      "20/20 [==============================] - 0s 873us/step - loss: 22626.7832\n",
      "Epoch 3/10\n",
      "20/20 [==============================] - 0s 917us/step - loss: 9912.2979\n",
      "Epoch 4/10\n",
      "20/20 [==============================] - 0s 915us/step - loss: 3991.4915\n",
      "Epoch 5/10\n",
      "20/20 [==============================] - 0s 916us/step - loss: 1556.7537\n",
      "Epoch 6/10\n",
      "20/20 [==============================] - 0s 846us/step - loss: 342.0754\n",
      "Epoch 7/10\n",
      "20/20 [==============================] - 0s 816us/step - loss: 46.4471\n",
      "Epoch 8/10\n",
      "20/20 [==============================] - 0s 836us/step - loss: 11.0318\n",
      "Epoch 9/10\n",
      "20/20 [==============================] - 0s 781us/step - loss: 2.6665\n",
      "Epoch 10/10\n",
      "20/20 [==============================] - 0s 830us/step - loss: 1.0758\n",
      "21/21 [==============================] - 0s 701us/step\n",
      "Processed file 29 of 10\n",
      "Epoch 1/10\n",
      "100/100 [==============================] - 0s 874us/step - loss: 95603504.0000\n",
      "Epoch 2/10\n",
      "100/100 [==============================] - 0s 760us/step - loss: 1448.0663\n",
      "Epoch 3/10\n",
      "100/100 [==============================] - 0s 750us/step - loss: 66.6816\n",
      "Epoch 4/10\n",
      "100/100 [==============================] - 0s 749us/step - loss: 65.5327\n",
      "Epoch 5/10\n",
      "100/100 [==============================] - 0s 754us/step - loss: 64.5833\n",
      "Epoch 6/10\n",
      "100/100 [==============================] - 0s 758us/step - loss: 63.8988\n",
      "Epoch 7/10\n",
      "100/100 [==============================] - 0s 760us/step - loss: 63.1746\n",
      "Epoch 8/10\n",
      "100/100 [==============================] - 0s 743us/step - loss: 61.9993\n",
      "Epoch 9/10\n",
      "100/100 [==============================] - 0s 739us/step - loss: 62.0700\n",
      "Epoch 10/10\n",
      "100/100 [==============================] - 0s 747us/step - loss: 60.7185\n",
      "101/101 [==============================] - 0s 469us/step\n",
      "Processed file 30 of 10\n",
      "Epoch 1/10\n",
      "39/39 [==============================] - 0s 1ms/step - loss: 11527213.0000\n",
      "Epoch 2/10\n",
      "39/39 [==============================] - 0s 1ms/step - loss: 1661896.7500\n",
      "Epoch 3/10\n",
      "39/39 [==============================] - 0s 1ms/step - loss: 9517.1826\n",
      "Epoch 4/10\n",
      "39/39 [==============================] - 0s 1ms/step - loss: 2010.0433\n",
      "Epoch 5/10\n",
      "39/39 [==============================] - 0s 993us/step - loss: 2239.1562\n",
      "Epoch 6/10\n",
      "39/39 [==============================] - 0s 1ms/step - loss: 1515.1685\n",
      "Epoch 7/10\n",
      "39/39 [==============================] - 0s 964us/step - loss: 1199.7738\n",
      "Epoch 8/10\n",
      "39/39 [==============================] - 0s 1ms/step - loss: 1192.7051\n",
      "Epoch 9/10\n",
      "39/39 [==============================] - 0s 981us/step - loss: 1008.0997\n",
      "Epoch 10/10\n",
      "39/39 [==============================] - 0s 991us/step - loss: 539.0266\n",
      "39/39 [==============================] - 0s 552us/step\n",
      "Processed file 31 of 10\n",
      "Epoch 1/10\n",
      "1/1 [==============================] - 0s 299ms/step - loss: 643396672.0000\n",
      "Epoch 2/10\n",
      "1/1 [==============================] - 0s 2ms/step - loss: 598613184.0000\n",
      "Epoch 3/10\n",
      "1/1 [==============================] - 0s 2ms/step - loss: 555491712.0000\n",
      "Epoch 4/10\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 514059648.0000\n",
      "Epoch 5/10\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 474339456.0000\n",
      "Epoch 6/10\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 436351520.0000\n",
      "Epoch 7/10\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 400110752.0000\n",
      "Epoch 8/10\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 365629152.0000\n",
      "Epoch 9/10\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 332912768.0000\n",
      "Epoch 10/10\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 301963104.0000\n",
      "1/1 [==============================] - 0s 49ms/step\n",
      "Processed file 32 of 10\n",
      "Epoch 1/10\n",
      "79/79 [==============================] - 0s 1ms/step - loss: 48144.7539\n",
      "Epoch 2/10\n",
      "79/79 [==============================] - 0s 971us/step - loss: 5.4265\n",
      "Epoch 3/10\n",
      "79/79 [==============================] - 0s 950us/step - loss: 2.2420\n",
      "Epoch 4/10\n",
      "79/79 [==============================] - 0s 988us/step - loss: 7.1887\n",
      "Epoch 5/10\n",
      "79/79 [==============================] - 0s 984us/step - loss: 3.6321\n",
      "Epoch 6/10\n",
      "79/79 [==============================] - 0s 973us/step - loss: 9.2445\n",
      "Epoch 7/10\n",
      "79/79 [==============================] - 0s 987us/step - loss: 4.2682\n",
      "Epoch 8/10\n",
      "79/79 [==============================] - 0s 977us/step - loss: 13.7217\n",
      "Epoch 9/10\n",
      "79/79 [==============================] - 0s 973us/step - loss: 5.1956\n",
      "Epoch 10/10\n",
      "79/79 [==============================] - 0s 971us/step - loss: 6.9742\n",
      "79/79 [==============================] - 0s 481us/step\n",
      "Processed file 33 of 10\n",
      "Epoch 1/10\n",
      "44/44 [==============================] - 0s 1ms/step - loss: 4954208768.0000\n",
      "Epoch 2/10\n",
      "44/44 [==============================] - 0s 1ms/step - loss: 2395092224.0000\n",
      "Epoch 3/10\n",
      "44/44 [==============================] - 0s 985us/step - loss: 168583.3438\n",
      "Epoch 4/10\n",
      "44/44 [==============================] - 0s 988us/step - loss: 1516.1451\n",
      "Epoch 5/10\n",
      "44/44 [==============================] - 0s 996us/step - loss: 79.2284\n",
      "Epoch 6/10\n",
      "44/44 [==============================] - 0s 951us/step - loss: 83.7341\n",
      "Epoch 7/10\n",
      "44/44 [==============================] - 0s 980us/step - loss: 11.5545\n",
      "Epoch 8/10\n",
      "44/44 [==============================] - 0s 994us/step - loss: 83.1661\n",
      "Epoch 9/10\n",
      "44/44 [==============================] - 0s 963us/step - loss: 150.8401\n",
      "Epoch 10/10\n",
      "44/44 [==============================] - 0s 984us/step - loss: 25.8274\n",
      "44/44 [==============================] - 0s 540us/step\n",
      "Processed file 34 of 10\n",
      "Epoch 1/10\n",
      "22/22 [==============================] - 0s 2ms/step - loss: 44332.2617\n",
      "Epoch 2/10\n",
      "22/22 [==============================] - 0s 1ms/step - loss: 7127.7310\n",
      "Epoch 3/10\n",
      "22/22 [==============================] - 0s 1ms/step - loss: 2058.4053\n",
      "Epoch 4/10\n",
      "22/22 [==============================] - 0s 1ms/step - loss: 682.9529\n",
      "Epoch 5/10\n",
      "22/22 [==============================] - 0s 1ms/step - loss: 13.4635\n",
      "Epoch 6/10\n",
      "22/22 [==============================] - 0s 1ms/step - loss: 5.1694\n",
      "Epoch 7/10\n",
      "22/22 [==============================] - 0s 1ms/step - loss: 0.7968\n",
      "Epoch 8/10\n",
      "22/22 [==============================] - 0s 1ms/step - loss: 1.1136\n",
      "Epoch 9/10\n",
      "22/22 [==============================] - 0s 1ms/step - loss: 0.9630\n",
      "Epoch 10/10\n",
      "22/22 [==============================] - 0s 1ms/step - loss: 0.6843\n",
      "22/22 [==============================] - 0s 682us/step\n",
      "Processed file 35 of 10\n",
      "Epoch 1/10\n",
      "85/85 [==============================] - 0s 1ms/step - loss: 48635460.0000\n",
      "Epoch 2/10\n",
      "85/85 [==============================] - 0s 974us/step - loss: 20198.9219\n",
      "Epoch 3/10\n",
      "85/85 [==============================] - 0s 987us/step - loss: 4.0056\n",
      "Epoch 4/10\n",
      "85/85 [==============================] - 0s 961us/step - loss: 0.4932\n",
      "Epoch 5/10\n",
      "85/85 [==============================] - 0s 963us/step - loss: 0.4887\n",
      "Epoch 6/10\n",
      "85/85 [==============================] - 0s 943us/step - loss: 0.4852\n",
      "Epoch 7/10\n",
      "85/85 [==============================] - 0s 947us/step - loss: 0.4037\n",
      "Epoch 8/10\n",
      "85/85 [==============================] - 0s 967us/step - loss: 0.2677\n",
      "Epoch 9/10\n",
      "85/85 [==============================] - 0s 949us/step - loss: 0.2619\n",
      "Epoch 10/10\n",
      "85/85 [==============================] - 0s 978us/step - loss: 0.2573\n",
      "85/85 [==============================] - 0s 425us/step\n",
      "Processed file 36 of 10\n",
      "Epoch 1/10\n",
      "98/98 [==============================] - 0s 1ms/step - loss: 5208390.5000\n",
      "Epoch 2/10\n",
      "98/98 [==============================] - 0s 966us/step - loss: 339.0597\n",
      "Epoch 3/10\n",
      "98/98 [==============================] - 0s 954us/step - loss: 1.8543\n",
      "Epoch 4/10\n",
      "98/98 [==============================] - 0s 947us/step - loss: 0.3917\n",
      "Epoch 5/10\n",
      "98/98 [==============================] - 0s 950us/step - loss: 0.3365\n",
      "Epoch 6/10\n",
      "98/98 [==============================] - 0s 927us/step - loss: 0.2752\n",
      "Epoch 7/10\n",
      "98/98 [==============================] - 0s 948us/step - loss: 0.2619\n",
      "Epoch 8/10\n",
      "98/98 [==============================] - 0s 938us/step - loss: 0.2603\n",
      "Epoch 9/10\n",
      "98/98 [==============================] - 0s 938us/step - loss: 0.2569\n",
      "Epoch 10/10\n",
      "98/98 [==============================] - 0s 955us/step - loss: 0.2717\n",
      "98/98 [==============================] - 0s 442us/step\n",
      "Processed file 37 of 10\n",
      "Epoch 1/10\n",
      "22/22 [==============================] - 0s 1ms/step - loss: 120626688.0000\n",
      "Epoch 2/10\n",
      "22/22 [==============================] - 0s 1ms/step - loss: 12217321.0000\n",
      "Epoch 3/10\n",
      "22/22 [==============================] - 0s 1ms/step - loss: 160822592.0000\n",
      "Epoch 4/10\n",
      "22/22 [==============================] - 0s 1ms/step - loss: 553149120.0000\n",
      "Epoch 5/10\n",
      "22/22 [==============================] - 0s 1ms/step - loss: 1836452.0000\n",
      "Epoch 6/10\n",
      "22/22 [==============================] - 0s 1ms/step - loss: 32223.7051\n",
      "Epoch 7/10\n",
      "22/22 [==============================] - 0s 1ms/step - loss: 3060.0156\n",
      "Epoch 8/10\n",
      "22/22 [==============================] - 0s 1ms/step - loss: 2034.7961\n",
      "Epoch 9/10\n",
      "22/22 [==============================] - 0s 1ms/step - loss: 346.6594\n",
      "Epoch 10/10\n",
      "22/22 [==============================] - 0s 1ms/step - loss: 43.9717\n",
      "22/22 [==============================] - 0s 686us/step\n",
      "Processed file 38 of 10\n",
      "Epoch 1/10\n",
      "27/27 [==============================] - 0s 1ms/step - loss: 7875415.0000\n",
      "Epoch 2/10\n",
      "27/27 [==============================] - 0s 1ms/step - loss: 748289.4375\n",
      "Epoch 3/10\n",
      "27/27 [==============================] - 0s 1ms/step - loss: 3358.0151\n",
      "Epoch 4/10\n",
      "27/27 [==============================] - 0s 1ms/step - loss: 1335.7150\n",
      "Epoch 5/10\n",
      "27/27 [==============================] - 0s 1ms/step - loss: 455.7574\n",
      "Epoch 6/10\n",
      "27/27 [==============================] - 0s 1ms/step - loss: 406.1674\n",
      "Epoch 7/10\n",
      "27/27 [==============================] - 0s 1ms/step - loss: 416.7289\n",
      "Epoch 8/10\n",
      "27/27 [==============================] - 0s 1ms/step - loss: 404.1670\n",
      "Epoch 9/10\n",
      "27/27 [==============================] - 0s 1ms/step - loss: 393.7885\n",
      "Epoch 10/10\n",
      "27/27 [==============================] - 0s 1ms/step - loss: 400.8731\n",
      "27/27 [==============================] - 0s 639us/step\n",
      "Processed file 39 of 10\n",
      "Epoch 1/10\n",
      "49/49 [==============================] - 0s 1ms/step - loss: 1140888960.0000\n",
      "Epoch 2/10\n",
      "49/49 [==============================] - 0s 1ms/step - loss: 147013248.0000\n",
      "Epoch 3/10\n",
      "49/49 [==============================] - 0s 965us/step - loss: 599839.8750\n",
      "Epoch 4/10\n",
      "49/49 [==============================] - 0s 971us/step - loss: 11000.0176\n",
      "Epoch 5/10\n",
      "49/49 [==============================] - 0s 959us/step - loss: 228.3947\n",
      "Epoch 6/10\n",
      "49/49 [==============================] - 0s 974us/step - loss: 169.8550\n",
      "Epoch 7/10\n",
      "49/49 [==============================] - 0s 965us/step - loss: 168.9243\n",
      "Epoch 8/10\n",
      "49/49 [==============================] - 0s 987us/step - loss: 174.4375\n",
      "Epoch 9/10\n",
      "49/49 [==============================] - 0s 972us/step - loss: 159.3935\n",
      "Epoch 10/10\n",
      "49/49 [==============================] - 0s 998us/step - loss: 156.5145\n",
      "49/49 [==============================] - 0s 544us/step\n",
      "Processed file 40 of 10\n",
      "Epoch 1/10\n",
      "98/98 [==============================] - 0s 1ms/step - loss: 2254251.7500\n",
      "Epoch 2/10\n",
      "98/98 [==============================] - 0s 978us/step - loss: 12252.2236\n",
      "Epoch 3/10\n",
      "98/98 [==============================] - 0s 965us/step - loss: 60620.0703\n",
      "Epoch 4/10\n",
      "98/98 [==============================] - 0s 971us/step - loss: 698.6609\n",
      "Epoch 5/10\n",
      "98/98 [==============================] - 0s 955us/step - loss: 41.2290\n",
      "Epoch 6/10\n",
      "98/98 [==============================] - 0s 958us/step - loss: 0.8120\n",
      "Epoch 7/10\n",
      "98/98 [==============================] - 0s 945us/step - loss: 0.7958\n",
      "Epoch 8/10\n",
      "98/98 [==============================] - 0s 968us/step - loss: 1.1837\n",
      "Epoch 9/10\n",
      "98/98 [==============================] - 0s 948us/step - loss: 0.7348\n",
      "Epoch 10/10\n",
      "98/98 [==============================] - 0s 948us/step - loss: 138.7099\n",
      "98/98 [==============================] - 0s 475us/step\n",
      "Processed file 41 of 10\n",
      "Epoch 1/10\n",
      "72/72 [==============================] - 0s 1ms/step - loss: 646513664.0000\n",
      "Epoch 2/10\n",
      "72/72 [==============================] - 0s 971us/step - loss: 715317.9375\n",
      "Epoch 3/10\n",
      "72/72 [==============================] - 0s 1ms/step - loss: 1672.5354\n",
      "Epoch 4/10\n",
      "72/72 [==============================] - 0s 1ms/step - loss: 24.6207\n",
      "Epoch 5/10\n",
      "72/72 [==============================] - 0s 1ms/step - loss: 20.4863\n",
      "Epoch 6/10\n",
      "72/72 [==============================] - 0s 1ms/step - loss: 19.7342\n",
      "Epoch 7/10\n",
      "72/72 [==============================] - 0s 1ms/step - loss: 19.2847\n",
      "Epoch 8/10\n",
      "72/72 [==============================] - 0s 1ms/step - loss: 18.3035\n",
      "Epoch 9/10\n",
      "72/72 [==============================] - 0s 1ms/step - loss: 17.5325\n",
      "Epoch 10/10\n",
      "72/72 [==============================] - 0s 1ms/step - loss: 16.7642\n",
      "72/72 [==============================] - 0s 514us/step\n",
      "Processed file 42 of 10\n",
      "Epoch 1/10\n",
      "5/5 [==============================] - 0s 2ms/step - loss: 173595.0625\n",
      "Epoch 2/10\n",
      "5/5 [==============================] - 0s 2ms/step - loss: 145211.0938\n",
      "Epoch 3/10\n",
      "5/5 [==============================] - 0s 2ms/step - loss: 118032.8047\n",
      "Epoch 4/10\n",
      "5/5 [==============================] - 0s 2ms/step - loss: 95616.6172\n",
      "Epoch 5/10\n",
      "5/5 [==============================] - 0s 1ms/step - loss: 75185.3516\n",
      "Epoch 6/10\n",
      "5/5 [==============================] - 0s 1ms/step - loss: 61329.1953\n",
      "Epoch 7/10\n",
      "5/5 [==============================] - 0s 1ms/step - loss: 52184.4766\n",
      "Epoch 8/10\n",
      "5/5 [==============================] - 0s 1ms/step - loss: 44187.0664\n",
      "Epoch 9/10\n",
      "5/5 [==============================] - 0s 1ms/step - loss: 37262.4492\n",
      "Epoch 10/10\n",
      "5/5 [==============================] - 0s 1ms/step - loss: 30812.1113\n",
      "5/5 [==============================] - 0s 888us/step\n",
      "Processed file 43 of 10\n",
      "Epoch 1/10\n",
      "100/100 [==============================] - 0s 1ms/step - loss: 4021073.0000\n",
      "Epoch 2/10\n",
      "100/100 [==============================] - 0s 961us/step - loss: 2322.9773\n",
      "Epoch 3/10\n",
      "100/100 [==============================] - 0s 968us/step - loss: 10.9417\n",
      "Epoch 4/10\n",
      "100/100 [==============================] - 0s 943us/step - loss: 9.9974\n",
      "Epoch 5/10\n",
      "100/100 [==============================] - 0s 958us/step - loss: 8.8128\n",
      "Epoch 6/10\n",
      "100/100 [==============================] - 0s 970us/step - loss: 7.8878\n",
      "Epoch 7/10\n",
      "100/100 [==============================] - 0s 961us/step - loss: 7.2265\n",
      "Epoch 8/10\n",
      "100/100 [==============================] - 0s 965us/step - loss: 7.1688\n",
      "Epoch 9/10\n",
      "100/100 [==============================] - 0s 961us/step - loss: 5.3628\n",
      "Epoch 10/10\n",
      "100/100 [==============================] - 0s 960us/step - loss: 5.1973\n",
      "100/100 [==============================] - 0s 453us/step\n",
      "Processed file 44 of 10\n",
      "Epoch 1/10\n",
      "100/100 [==============================] - 0s 881us/step - loss: 553680256.0000\n",
      "Epoch 2/10\n",
      "100/100 [==============================] - 0s 749us/step - loss: 2887.8691\n",
      "Epoch 3/10\n",
      "100/100 [==============================] - 0s 759us/step - loss: 734.9278\n",
      "Epoch 4/10\n",
      "100/100 [==============================] - 0s 767us/step - loss: 684.3064\n",
      "Epoch 5/10\n",
      "100/100 [==============================] - 0s 754us/step - loss: 560.8406\n",
      "Epoch 6/10\n",
      "100/100 [==============================] - 0s 752us/step - loss: 1222.0428\n",
      "Epoch 7/10\n",
      "100/100 [==============================] - 0s 760us/step - loss: 753.7509\n",
      "Epoch 8/10\n",
      "100/100 [==============================] - 0s 748us/step - loss: 863.1708\n",
      "Epoch 9/10\n",
      "100/100 [==============================] - 0s 766us/step - loss: 896.9843\n",
      "Epoch 10/10\n",
      "100/100 [==============================] - 0s 757us/step - loss: 3072.9287\n",
      "101/101 [==============================] - 0s 467us/step\n",
      "Processed file 45 of 10\n",
      "Epoch 1/10\n",
      "88/88 [==============================] - 0s 1ms/step - loss: 60426592.0000\n",
      "Epoch 2/10\n",
      "88/88 [==============================] - 0s 969us/step - loss: 3404110.2500\n",
      "Epoch 3/10\n",
      "88/88 [==============================] - 0s 950us/step - loss: 15686.9512\n",
      "Epoch 4/10\n",
      "88/88 [==============================] - 0s 957us/step - loss: 7019.5903\n",
      "Epoch 5/10\n",
      "88/88 [==============================] - 0s 966us/step - loss: 3380.9822\n",
      "Epoch 6/10\n",
      "88/88 [==============================] - 0s 957us/step - loss: 1964.5948\n",
      "Epoch 7/10\n",
      "88/88 [==============================] - 0s 955us/step - loss: 1152.2657\n",
      "Epoch 8/10\n",
      "88/88 [==============================] - 0s 982us/step - loss: 632.9636\n",
      "Epoch 9/10\n",
      "88/88 [==============================] - 0s 939us/step - loss: 321.7886\n",
      "Epoch 10/10\n",
      "88/88 [==============================] - 0s 947us/step - loss: 163.0205\n",
      "88/88 [==============================] - 0s 485us/step\n",
      "Processed file 46 of 10\n",
      "Epoch 1/10\n",
      "20/20 [==============================] - 0s 2ms/step - loss: 1872445312.0000\n",
      "Epoch 2/10\n",
      "20/20 [==============================] - 0s 1ms/step - loss: 1574154240.0000\n",
      "Epoch 3/10\n",
      "20/20 [==============================] - 0s 1ms/step - loss: 1336928768.0000\n",
      "Epoch 4/10\n",
      "20/20 [==============================] - 0s 1ms/step - loss: 425514752.0000\n",
      "Epoch 5/10\n",
      "20/20 [==============================] - 0s 1ms/step - loss: 335926912.0000\n",
      "Epoch 6/10\n",
      "20/20 [==============================] - 0s 1ms/step - loss: 173504688.0000\n",
      "Epoch 7/10\n",
      "20/20 [==============================] - 0s 1ms/step - loss: 37416404.0000\n",
      "Epoch 8/10\n",
      "20/20 [==============================] - 0s 1ms/step - loss: 9665375.0000\n",
      "Epoch 9/10\n",
      "20/20 [==============================] - 0s 1ms/step - loss: 1923579.2500\n",
      "Epoch 10/10\n",
      "20/20 [==============================] - 0s 1ms/step - loss: 299978.4688\n",
      "20/20 [==============================] - 0s 646us/step\n",
      "Processed file 47 of 10\n",
      "Epoch 1/10\n",
      "100/100 [==============================] - 0s 874us/step - loss: 47584036.0000\n",
      "Epoch 2/10\n",
      "100/100 [==============================] - 0s 754us/step - loss: 722.8549\n",
      "Epoch 3/10\n",
      "100/100 [==============================] - 0s 749us/step - loss: 0.4973\n",
      "Epoch 4/10\n",
      "100/100 [==============================] - 0s 749us/step - loss: 0.4814\n",
      "Epoch 5/10\n",
      "100/100 [==============================] - 0s 748us/step - loss: 0.4911\n",
      "Epoch 6/10\n",
      "100/100 [==============================] - 0s 740us/step - loss: 0.5060\n",
      "Epoch 7/10\n",
      "100/100 [==============================] - 0s 748us/step - loss: 0.4871\n",
      "Epoch 8/10\n",
      "100/100 [==============================] - 0s 743us/step - loss: 0.4414\n",
      "Epoch 9/10\n",
      "100/100 [==============================] - 0s 759us/step - loss: 0.5213\n",
      "Epoch 10/10\n",
      "100/100 [==============================] - 0s 760us/step - loss: 0.5389\n",
      "101/101 [==============================] - 0s 469us/step\n",
      "Processed file 48 of 10\n",
      "Epoch 1/10\n",
      "100/100 [==============================] - 0s 877us/step - loss: 51341208584192.0000\n",
      "Epoch 2/10\n",
      "100/100 [==============================] - 0s 748us/step - loss: 1402136559616.0000\n",
      "Epoch 3/10\n",
      "100/100 [==============================] - 0s 748us/step - loss: 3173472.7500\n",
      "Epoch 4/10\n",
      "100/100 [==============================] - 0s 735us/step - loss: 255.4233\n",
      "Epoch 5/10\n",
      "100/100 [==============================] - 0s 755us/step - loss: 132.9034\n",
      "Epoch 6/10\n",
      "100/100 [==============================] - 0s 755us/step - loss: 132.7367\n",
      "Epoch 7/10\n",
      "100/100 [==============================] - 0s 736us/step - loss: 132.8133\n",
      "Epoch 8/10\n",
      "100/100 [==============================] - 0s 755us/step - loss: 132.5698\n",
      "Epoch 9/10\n",
      "100/100 [==============================] - 0s 738us/step - loss: 132.9014\n",
      "Epoch 10/10\n",
      "100/100 [==============================] - 0s 742us/step - loss: 133.2772\n",
      "101/101 [==============================] - 0s 466us/step\n",
      "Processed file 49 of 10\n",
      "Epoch 1/10\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 3827030.0000\n",
      "Epoch 2/10\n",
      "11/11 [==============================] - 0s 2ms/step - loss: 1610574.0000\n",
      "Epoch 3/10\n",
      "11/11 [==============================] - 0s 1ms/step - loss: 364529.5625\n",
      "Epoch 4/10\n",
      "11/11 [==============================] - 0s 1ms/step - loss: 252229.3750\n",
      "Epoch 5/10\n",
      "11/11 [==============================] - 0s 1ms/step - loss: 2725.3525\n",
      "Epoch 6/10\n",
      "11/11 [==============================] - 0s 1ms/step - loss: 8517.2568\n",
      "Epoch 7/10\n",
      "11/11 [==============================] - 0s 1ms/step - loss: 3047.7451\n",
      "Epoch 8/10\n",
      "11/11 [==============================] - 0s 1ms/step - loss: 329.6898\n",
      "Epoch 9/10\n",
      "11/11 [==============================] - 0s 1ms/step - loss: 256.0119\n",
      "Epoch 10/10\n",
      "11/11 [==============================] - 0s 1ms/step - loss: 172.8857\n",
      "11/11 [==============================] - 0s 799us/step\n",
      "Processed file 50 of 10\n",
      "Profitable stocks: ['split_ueps.us.csv', 'split_siri.us.csv', 'split_pbmd.us.csv', 'split_mmlp.us.csv', 'split_ltrx.us.csv', 'split_jack.us.csv', 'split_erm.us.csv', 'split_dtea.us.csv', 'split_atacu.us.csv']\n"
     ]
    }
   ],
   "source": [
    "import os\n",
    "import pandas as pd\n",
    "from sklearn.model_selection import train_test_split\n",
    "from tensorflow.keras.models import Sequential\n",
    "from tensorflow.keras.layers import LSTM, Dense\n",
    "import numpy as np\n",
    "\n",
    "# Define the desired growth threshold\n",
    "growth_threshold = 1.3  # 30% growth\n",
    "\n",
    "# Iterate through the CSV files\n",
    "csv_files = os.listdir('/Users/aryanchopra/Downloads/archive-3/Data/Stocks/')\n",
    "profitable_stocks = []\n",
    "\n",
    "# Process the first 50 CSV files\n",
    "for i, csv_file in enumerate(csv_files[:50]):\n",
    "    df = pd.read_csv(os.path.join('/Users/aryanchopra/Downloads/archive-3/Data/Stocks/', csv_file))\n",
    "    # Calculate the percentage change in Close prices\n",
    "    df['Close_pct_change'] = df['Close'].pct_change()\n",
    "    # Split the data into features (trainX) and target (trainY)\n",
    "   \n",
    "    trainY = df['Close_pct_change'].shift(-1)  # Shift the percentage change by one step to predict future increase\n",
    "    trainY = trainY.apply(lambda x: 1 if x > 0 else 0) \n",
    "    trainY = trainY[:-1]  # Remove the last row since there is no corresponding future value\n",
    "\n",
    "    trainX = df[['Open', 'High', 'Low', 'Close', 'Volume']]  \n",
    "    trainX = trainX[:-1]\n",
    "    # Reshape the data to have a third dimension of size 1\n",
    "    trainX = np.expand_dims(trainX, axis=2)\n",
    "    \n",
    "    # Train and evaluate the LSTM model\n",
    "    lstm = Sequential()\n",
    "    lstm.add(LSTM(32, input_shape=(trainX.shape[1], trainX.shape[2]), activation='relu', return_sequences=False))\n",
    "    lstm.add(Dense(1))\n",
    "    lstm.compile(loss='mean_squared_error', optimizer='adam')\n",
    "    lstm.fit(trainX, trainY, epochs=10, batch_size=32)\n",
    "\n",
    "    # Make predictions for the next year\n",
    "    future_data = df[['Open', 'High', 'Low', 'Close', 'Volume']]  # Use the same data for debugging\n",
    "    future_data = np.expand_dims(future_data, axis=2)\n",
    "    future_predictions = lstm.predict(future_data)\n",
    "\n",
    "    # Calculate the projected growth rate\n",
    "    current_price = df['Close'].iloc[-1]\n",
    "    future_price = future_predictions[-1]\n",
    "    growth_rate = future_price / current_price\n",
    "\n",
    "    # Check if the projected growth meets the threshold\n",
    "    if growth_rate >= growth_threshold:\n",
    "        profitable_stocks.append(csv_file)\n",
    "\n",
    "    print(f\"Processed file {i+1} of 50\")\n",
    "profitable_stocks = sorted(profitable_stocks, reverse=True)\n",
    "print(\"Profitable stocks:\", profitable_stocks)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "319ec4a8",
   "metadata": {},
   "outputs": [],
   "source": [
    "from the first 50 files profitable stocks from most listed first is shown above (can be done for all 7000 just training time will go up) the training data is taken from https://www.kaggle.com/datasets/borismarjanovic/price-volume-data-for-all-us-stocks-etfs?select=Stocks"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}