neural_network.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import tensorflow as tf\n",
    "from tensorflow.keras.layers import Input, LSTM, BatchNormalization, Dense, Dropout\n",
    "from tensorflow.keras.callbacks import EarlyStopping\n",
    "from sklearn.model_selection import train_test_split\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "pd.set_option('display.max_columns', None)\n",
    "pd.set_option('display.max_rows', None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv(\"modified_QUANTT_dataset.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [],
   "source": [
    "# df = df[df[\"Average Value\"] < 20]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = df[['Stock price','Strike price','T','Risk-Free Rate','Implied Volatility']]\n",
    "targets = df['Average Value']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test  = train_test_split(data, targets, test_size=0.2, shuffle=True)\n",
    "X_test, X_val, y_test, y_val = train_test_split(X_test, y_test, test_size=0.5, shuffle=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(20616, 5) (2577, 5) (2577, 5)\n"
     ]
    }
   ],
   "source": [
    "print(X_train.shape, X_test.shape, X_val.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m9s\u001b[0m 7ms/step - MAE: 20.3253 - loss: 1540.8948 - val_MAE: 17.6022 - val_loss: 1318.1340\n",
      "Epoch 2/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 5ms/step - MAE: 18.8454 - loss: 1433.2595 - val_MAE: 16.4618 - val_loss: 1221.6069\n",
      "Epoch 3/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - MAE: 18.0393 - loss: 1380.4138 - val_MAE: 15.2355 - val_loss: 1067.4231\n",
      "Epoch 4/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - MAE: 16.4120 - loss: 1146.9521 - val_MAE: 14.2212 - val_loss: 873.0052\n",
      "Epoch 5/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - MAE: 15.3233 - loss: 909.4589 - val_MAE: 13.6351 - val_loss: 681.0530\n",
      "Epoch 6/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - MAE: 14.7439 - loss: 736.1569 - val_MAE: 13.4913 - val_loss: 576.8007\n",
      "Epoch 7/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - MAE: 14.3098 - loss: 640.4628 - val_MAE: 13.1203 - val_loss: 540.4563\n",
      "Epoch 8/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - MAE: 14.0543 - loss: 606.2888 - val_MAE: 12.7094 - val_loss: 515.8748\n",
      "Epoch 9/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - MAE: 13.2851 - loss: 563.5814 - val_MAE: 12.2683 - val_loss: 492.2311\n",
      "Epoch 10/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - MAE: 12.9097 - loss: 521.4202 - val_MAE: 11.9487 - val_loss: 469.5887\n",
      "Epoch 11/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - MAE: 12.8792 - loss: 535.7720 - val_MAE: 11.5843 - val_loss: 446.0460\n",
      "Epoch 12/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - MAE: 12.5625 - loss: 515.1829 - val_MAE: 11.2848 - val_loss: 424.2486\n",
      "Epoch 13/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - MAE: 12.1445 - loss: 481.4337 - val_MAE: 11.0334 - val_loss: 403.1581\n",
      "Epoch 14/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - MAE: 11.6975 - loss: 446.9361 - val_MAE: 10.7520 - val_loss: 382.7813\n",
      "Epoch 15/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - MAE: 11.4594 - loss: 421.3136 - val_MAE: 10.5109 - val_loss: 361.9933\n",
      "Epoch 16/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - MAE: 11.3654 - loss: 399.9904 - val_MAE: 10.2518 - val_loss: 340.9857\n",
      "Epoch 17/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - MAE: 11.0300 - loss: 378.5661 - val_MAE: 9.9901 - val_loss: 320.6636\n",
      "Epoch 18/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - MAE: 10.8471 - loss: 389.0188 - val_MAE: 9.6753 - val_loss: 300.1974\n",
      "Epoch 19/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - MAE: 10.5220 - loss: 347.7041 - val_MAE: 9.3929 - val_loss: 281.4189\n",
      "Epoch 20/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - MAE: 10.2178 - loss: 329.4752 - val_MAE: 9.1031 - val_loss: 261.1830\n",
      "Epoch 21/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - MAE: 10.0529 - loss: 326.3078 - val_MAE: 8.8164 - val_loss: 240.9214\n",
      "Epoch 22/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - MAE: 9.8703 - loss: 298.3856 - val_MAE: 8.5119 - val_loss: 223.7470\n",
      "Epoch 23/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - MAE: 9.3173 - loss: 263.1722 - val_MAE: 8.2387 - val_loss: 205.8522\n",
      "Epoch 24/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - MAE: 9.3828 - loss: 262.7557 - val_MAE: 7.9341 - val_loss: 187.8787\n",
      "Epoch 25/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - MAE: 8.9858 - loss: 245.0234 - val_MAE: 7.6745 - val_loss: 172.2486\n",
      "Epoch 26/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - MAE: 8.6810 - loss: 224.0077 - val_MAE: 7.3296 - val_loss: 158.3737\n",
      "Epoch 27/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - MAE: 8.5424 - loss: 213.3430 - val_MAE: 7.1275 - val_loss: 146.6829\n",
      "Epoch 28/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - MAE: 8.3252 - loss: 204.8928 - val_MAE: 6.8859 - val_loss: 137.0378\n",
      "Epoch 29/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - MAE: 8.1909 - loss: 198.2487 - val_MAE: 6.6924 - val_loss: 127.4621\n",
      "Epoch 30/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - MAE: 8.0167 - loss: 193.2352 - val_MAE: 6.5196 - val_loss: 120.5438\n",
      "Epoch 31/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - MAE: 7.8666 - loss: 183.9152 - val_MAE: 6.3064 - val_loss: 113.5595\n",
      "Epoch 32/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - MAE: 7.6106 - loss: 165.4146 - val_MAE: 6.2151 - val_loss: 108.0853\n",
      "Epoch 33/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 5ms/step - MAE: 7.4931 - loss: 159.5714 - val_MAE: 6.0811 - val_loss: 102.7872\n",
      "Epoch 34/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 3ms/step - MAE: 7.4790 - loss: 163.5011 - val_MAE: 5.9706 - val_loss: 99.8466\n",
      "Epoch 35/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - MAE: 7.3416 - loss: 154.7035 - val_MAE: 5.8715 - val_loss: 94.4943\n",
      "Epoch 36/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - MAE: 7.4507 - loss: 164.1558 - val_MAE: 5.7732 - val_loss: 89.9466\n",
      "Epoch 37/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - MAE: 7.2346 - loss: 145.1363 - val_MAE: 5.7063 - val_loss: 86.6159\n",
      "Epoch 38/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - MAE: 7.2972 - loss: 149.0172 - val_MAE: 5.5931 - val_loss: 84.2105\n",
      "Epoch 39/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - MAE: 6.9864 - loss: 136.6201 - val_MAE: 5.5006 - val_loss: 80.5902\n",
      "Epoch 40/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 5ms/step - MAE: 7.0157 - loss: 141.4220 - val_MAE: 5.4105 - val_loss: 78.1490\n",
      "Epoch 41/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - MAE: 6.8924 - loss: 132.0897 - val_MAE: 5.3607 - val_loss: 75.8688\n",
      "Epoch 42/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 6ms/step - MAE: 7.1026 - loss: 142.0486 - val_MAE: 5.2956 - val_loss: 74.0107\n",
      "Epoch 43/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - MAE: 6.8302 - loss: 131.0942 - val_MAE: 5.2134 - val_loss: 72.1025\n",
      "Epoch 44/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 9ms/step - MAE: 6.7791 - loss: 126.5920 - val_MAE: 5.1412 - val_loss: 69.2174\n",
      "Epoch 45/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 7ms/step - MAE: 6.7301 - loss: 129.7764 - val_MAE: 5.0443 - val_loss: 67.4178\n",
      "Epoch 46/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - MAE: 6.5521 - loss: 115.5165 - val_MAE: 4.9222 - val_loss: 64.9214\n",
      "Epoch 47/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 5ms/step - MAE: 6.6947 - loss: 133.4667 - val_MAE: 4.8235 - val_loss: 63.1912\n",
      "Epoch 48/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 5ms/step - MAE: 6.5288 - loss: 120.2014 - val_MAE: 4.6150 - val_loss: 59.7207\n",
      "Epoch 49/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 5ms/step - MAE: 6.4173 - loss: 121.4972 - val_MAE: 4.4260 - val_loss: 57.3760\n",
      "Epoch 50/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - MAE: 6.2508 - loss: 110.5670 - val_MAE: 4.2458 - val_loss: 55.8670\n",
      "Epoch 51/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 5ms/step - MAE: 6.2029 - loss: 113.0327 - val_MAE: 4.1460 - val_loss: 55.0462\n",
      "Epoch 52/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 10ms/step - MAE: 6.1251 - loss: 117.1150 - val_MAE: 3.9714 - val_loss: 52.8828\n",
      "Epoch 53/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 7ms/step - MAE: 6.1377 - loss: 112.6686 - val_MAE: 3.8395 - val_loss: 51.4333\n",
      "Epoch 54/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 10ms/step - MAE: 5.9226 - loss: 104.1353 - val_MAE: 3.7531 - val_loss: 49.3717\n",
      "Epoch 55/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 7ms/step - MAE: 5.8285 - loss: 104.0216 - val_MAE: 3.6808 - val_loss: 48.7270\n",
      "Epoch 56/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 8ms/step - MAE: 5.8890 - loss: 106.9323 - val_MAE: 3.6349 - val_loss: 48.0037\n",
      "Epoch 57/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 5ms/step - MAE: 5.8675 - loss: 109.6475 - val_MAE: 3.5818 - val_loss: 46.1918\n",
      "Epoch 58/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 5ms/step - MAE: 5.7442 - loss: 102.8653 - val_MAE: 3.5345 - val_loss: 45.3383\n",
      "Epoch 59/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 6ms/step - MAE: 5.6407 - loss: 102.9753 - val_MAE: 3.4946 - val_loss: 44.6815\n",
      "Epoch 60/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 5ms/step - MAE: 5.5727 - loss: 99.7347 - val_MAE: 3.4658 - val_loss: 43.9002\n",
      "Epoch 61/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 5ms/step - MAE: 5.9259 - loss: 116.0114 - val_MAE: 3.4282 - val_loss: 42.7656\n",
      "Epoch 62/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 6ms/step - MAE: 5.5048 - loss: 97.5376 - val_MAE: 3.4274 - val_loss: 42.8052\n",
      "Epoch 63/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - MAE: 5.6777 - loss: 104.7922 - val_MAE: 3.4032 - val_loss: 41.9105\n",
      "Epoch 64/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 7ms/step - MAE: 5.7192 - loss: 106.6130 - val_MAE: 3.3691 - val_loss: 40.8881\n",
      "Epoch 65/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - MAE: 5.3548 - loss: 92.1564 - val_MAE: 3.3662 - val_loss: 40.7042\n",
      "Epoch 66/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - MAE: 5.5456 - loss: 99.3733 - val_MAE: 3.3217 - val_loss: 39.3800\n",
      "Epoch 67/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 5ms/step - MAE: 5.4179 - loss: 95.5538 - val_MAE: 3.3101 - val_loss: 38.9041\n",
      "Epoch 68/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 4ms/step - MAE: 5.2462 - loss: 86.8064 - val_MAE: 3.2848 - val_loss: 38.1317\n",
      "Epoch 69/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - MAE: 5.3617 - loss: 95.2007 - val_MAE: 3.2715 - val_loss: 37.5769\n",
      "Epoch 70/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - MAE: 5.4045 - loss: 93.7264 - val_MAE: 3.2622 - val_loss: 36.9963\n",
      "Epoch 71/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 4ms/step - MAE: 5.2434 - loss: 83.9153 - val_MAE: 3.2575 - val_loss: 36.5683\n",
      "Epoch 72/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - MAE: 5.5160 - loss: 100.2933 - val_MAE: 3.2270 - val_loss: 35.9416\n",
      "Epoch 73/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - MAE: 5.3008 - loss: 89.5462 - val_MAE: 3.2151 - val_loss: 35.7225\n",
      "Epoch 74/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - MAE: 5.4559 - loss: 99.7541 - val_MAE: 3.1994 - val_loss: 35.7527\n",
      "Epoch 75/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 6ms/step - MAE: 5.1463 - loss: 83.8247 - val_MAE: 3.1805 - val_loss: 34.7261\n",
      "Epoch 76/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 6ms/step - MAE: 5.3629 - loss: 91.6417 - val_MAE: 3.1650 - val_loss: 34.3127\n",
      "Epoch 77/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - MAE: 5.3955 - loss: 92.0289 - val_MAE: 3.1503 - val_loss: 33.5066\n",
      "Epoch 78/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - MAE: 5.3138 - loss: 89.6922 - val_MAE: 3.1355 - val_loss: 33.1698\n",
      "Epoch 79/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - MAE: 5.2922 - loss: 90.6386 - val_MAE: 3.1332 - val_loss: 33.1518\n",
      "Epoch 80/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 7ms/step - MAE: 5.1617 - loss: 86.6792 - val_MAE: 3.1253 - val_loss: 33.1584\n",
      "Epoch 81/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 7ms/step - MAE: 5.2812 - loss: 87.4542 - val_MAE: 3.1287 - val_loss: 32.8321\n",
      "Epoch 82/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - MAE: 5.3886 - loss: 99.3839 - val_MAE: 3.1032 - val_loss: 32.2312\n",
      "Epoch 83/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 6ms/step - MAE: 5.2903 - loss: 89.9099 - val_MAE: 3.0795 - val_loss: 32.1639\n",
      "Epoch 84/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 8ms/step - MAE: 5.3033 - loss: 95.5413 - val_MAE: 3.0730 - val_loss: 31.2449\n",
      "Epoch 85/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 7ms/step - MAE: 5.1550 - loss: 83.5052 - val_MAE: 3.0661 - val_loss: 31.4878\n",
      "Epoch 86/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 5ms/step - MAE: 5.1943 - loss: 88.2070 - val_MAE: 3.0541 - val_loss: 31.1520\n",
      "Epoch 87/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 6ms/step - MAE: 5.2131 - loss: 85.7459 - val_MAE: 3.0363 - val_loss: 30.6804\n",
      "Epoch 88/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 5ms/step - MAE: 5.1001 - loss: 85.3221 - val_MAE: 3.0239 - val_loss: 30.3288\n",
      "Epoch 89/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 7ms/step - MAE: 5.0147 - loss: 79.6102 - val_MAE: 3.0147 - val_loss: 29.9305\n",
      "Epoch 90/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 9ms/step - MAE: 5.0741 - loss: 82.8962 - val_MAE: 3.0088 - val_loss: 29.7750\n",
      "Epoch 91/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 8ms/step - MAE: 5.1955 - loss: 90.8385 - val_MAE: 2.9957 - val_loss: 29.7932\n",
      "Epoch 92/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 9ms/step - MAE: 5.0171 - loss: 81.2627 - val_MAE: 2.9829 - val_loss: 29.4746\n",
      "Epoch 93/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 8ms/step - MAE: 5.1308 - loss: 87.6647 - val_MAE: 2.9577 - val_loss: 28.8187\n",
      "Epoch 94/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 4ms/step - MAE: 5.0583 - loss: 82.6275 - val_MAE: 2.9649 - val_loss: 29.0944\n",
      "Epoch 95/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 6ms/step - MAE: 5.0067 - loss: 81.2685 - val_MAE: 2.9474 - val_loss: 28.4651\n",
      "Epoch 96/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 7ms/step - MAE: 5.0206 - loss: 82.8147 - val_MAE: 2.9603 - val_loss: 28.6374\n",
      "Epoch 97/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 5ms/step - MAE: 5.1022 - loss: 84.2663 - val_MAE: 2.9283 - val_loss: 28.1656\n",
      "Epoch 98/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 4ms/step - MAE: 4.9747 - loss: 80.3162 - val_MAE: 2.9312 - val_loss: 28.5431\n",
      "Epoch 99/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 5ms/step - MAE: 5.0748 - loss: 83.2995 - val_MAE: 2.9168 - val_loss: 28.0942\n",
      "Epoch 100/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - MAE: 5.0340 - loss: 79.5591 - val_MAE: 2.9076 - val_loss: 27.7391\n",
      "Epoch 101/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 5ms/step - MAE: 4.9591 - loss: 78.3759 - val_MAE: 2.8941 - val_loss: 27.3996\n",
      "Epoch 102/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 5ms/step - MAE: 5.1986 - loss: 83.5380 - val_MAE: 2.8650 - val_loss: 26.9792\n",
      "Epoch 103/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 5ms/step - MAE: 5.1537 - loss: 85.3578 - val_MAE: 2.8506 - val_loss: 26.6566\n",
      "Epoch 104/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 5ms/step - MAE: 4.8782 - loss: 77.4095 - val_MAE: 2.8535 - val_loss: 26.4899\n",
      "Epoch 105/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - MAE: 4.9779 - loss: 82.9224 - val_MAE: 2.8538 - val_loss: 26.8362\n",
      "Epoch 106/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - MAE: 5.0166 - loss: 81.1771 - val_MAE: 2.8288 - val_loss: 26.2314\n",
      "Epoch 107/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 5ms/step - MAE: 4.9533 - loss: 79.4431 - val_MAE: 2.8230 - val_loss: 26.2455\n",
      "Epoch 108/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - MAE: 5.0186 - loss: 79.8300 - val_MAE: 2.8008 - val_loss: 25.9172\n",
      "Epoch 109/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - MAE: 5.0089 - loss: 80.6889 - val_MAE: 2.8149 - val_loss: 25.8451\n",
      "Epoch 110/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 5ms/step - MAE: 5.0195 - loss: 82.6719 - val_MAE: 2.7671 - val_loss: 25.5742\n",
      "Epoch 111/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 6ms/step - MAE: 4.9178 - loss: 77.7545 - val_MAE: 2.7855 - val_loss: 25.7585\n",
      "Epoch 112/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 5ms/step - MAE: 5.0872 - loss: 83.2765 - val_MAE: 2.7550 - val_loss: 25.1829\n",
      "Epoch 113/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - MAE: 4.8850 - loss: 77.9816 - val_MAE: 2.7987 - val_loss: 25.9937\n",
      "Epoch 114/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - MAE: 4.9326 - loss: 78.1332 - val_MAE: 2.7918 - val_loss: 25.4522\n",
      "Epoch 115/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 5ms/step - MAE: 4.9651 - loss: 80.4857 - val_MAE: 2.7772 - val_loss: 25.2881\n",
      "Epoch 116/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - MAE: 4.9171 - loss: 78.0268 - val_MAE: 2.7636 - val_loss: 24.9351\n",
      "Epoch 117/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - MAE: 4.8914 - loss: 75.7662 - val_MAE: 2.7389 - val_loss: 24.7223\n",
      "Epoch 118/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 5ms/step - MAE: 4.9639 - loss: 80.9398 - val_MAE: 2.7720 - val_loss: 24.9259\n",
      "Epoch 119/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 3ms/step - MAE: 4.9774 - loss: 80.6636 - val_MAE: 2.7327 - val_loss: 24.5800\n",
      "Epoch 120/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - MAE: 4.9330 - loss: 80.0601 - val_MAE: 2.7345 - val_loss: 24.4394\n",
      "Epoch 121/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - MAE: 4.9876 - loss: 80.9568 - val_MAE: 2.7217 - val_loss: 24.3103\n",
      "Epoch 122/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 6ms/step - MAE: 4.8257 - loss: 73.4428 - val_MAE: 2.7236 - val_loss: 24.4103\n",
      "Epoch 123/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - MAE: 5.0504 - loss: 80.6293 - val_MAE: 2.7041 - val_loss: 24.3843\n",
      "Epoch 124/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - MAE: 4.8144 - loss: 75.9588 - val_MAE: 2.7089 - val_loss: 24.2537\n",
      "Epoch 125/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - MAE: 4.8048 - loss: 76.1293 - val_MAE: 2.7101 - val_loss: 24.1524\n",
      "Epoch 126/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 5ms/step - MAE: 4.8174 - loss: 74.7071 - val_MAE: 2.6946 - val_loss: 24.1055\n",
      "Epoch 127/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - MAE: 4.7366 - loss: 71.7369 - val_MAE: 2.7027 - val_loss: 24.0472\n",
      "Epoch 128/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - MAE: 4.9359 - loss: 81.5684 - val_MAE: 2.7085 - val_loss: 23.9643\n",
      "Epoch 129/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 7ms/step - MAE: 4.8555 - loss: 77.3335 - val_MAE: 2.6729 - val_loss: 23.8410\n",
      "Epoch 130/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - MAE: 4.9062 - loss: 79.0801 - val_MAE: 2.6588 - val_loss: 23.6843\n",
      "Epoch 131/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - MAE: 5.0659 - loss: 85.4294 - val_MAE: 2.6446 - val_loss: 23.3428\n",
      "Epoch 132/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 6ms/step - MAE: 4.7251 - loss: 70.1966 - val_MAE: 2.6324 - val_loss: 23.3795\n",
      "Epoch 133/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 3ms/step - MAE: 4.7432 - loss: 74.4950 - val_MAE: 2.7021 - val_loss: 23.8856\n",
      "Epoch 134/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - MAE: 4.7493 - loss: 74.9760 - val_MAE: 2.6144 - val_loss: 23.1510\n",
      "Epoch 135/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 5ms/step - MAE: 4.7575 - loss: 72.1437 - val_MAE: 2.6245 - val_loss: 22.9852\n",
      "Epoch 136/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 3ms/step - MAE: 4.9886 - loss: 81.1853 - val_MAE: 2.6160 - val_loss: 23.0618\n",
      "Epoch 137/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - MAE: 4.7898 - loss: 75.0333 - val_MAE: 2.6228 - val_loss: 22.8208\n",
      "Epoch 138/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - MAE: 4.7478 - loss: 72.7191 - val_MAE: 2.6274 - val_loss: 22.9533\n",
      "Epoch 139/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - MAE: 4.8865 - loss: 81.3245 - val_MAE: 2.6164 - val_loss: 22.8782\n",
      "Epoch 140/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - MAE: 4.9239 - loss: 79.5744 - val_MAE: 2.6121 - val_loss: 22.9906\n",
      "Epoch 141/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - MAE: 4.9597 - loss: 82.2193 - val_MAE: 2.5941 - val_loss: 22.4533\n",
      "Epoch 142/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - MAE: 4.9861 - loss: 82.5099 - val_MAE: 2.5987 - val_loss: 22.5358\n",
      "Epoch 143/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - MAE: 4.9277 - loss: 80.1518 - val_MAE: 2.5845 - val_loss: 22.4710\n",
      "Epoch 144/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 6ms/step - MAE: 4.7059 - loss: 71.3300 - val_MAE: 2.5859 - val_loss: 22.5037\n",
      "Epoch 145/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 11ms/step - MAE: 4.8213 - loss: 77.2711 - val_MAE: 2.6247 - val_loss: 22.5486\n",
      "Epoch 146/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 7ms/step - MAE: 4.8766 - loss: 77.0482 - val_MAE: 2.5887 - val_loss: 22.3435\n",
      "Epoch 147/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 10ms/step - MAE: 4.7806 - loss: 73.0544 - val_MAE: 2.5847 - val_loss: 22.4060\n",
      "Epoch 148/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 5ms/step - MAE: 4.8181 - loss: 78.2034 - val_MAE: 2.5823 - val_loss: 22.3920\n",
      "Epoch 149/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 9ms/step - MAE: 4.8836 - loss: 78.3467 - val_MAE: 2.5716 - val_loss: 21.9052\n",
      "Epoch 150/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 6ms/step - MAE: 4.8420 - loss: 77.3657 - val_MAE: 2.6050 - val_loss: 22.3361\n",
      "Epoch 151/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 6ms/step - MAE: 4.8981 - loss: 77.0327 - val_MAE: 2.5233 - val_loss: 22.0431\n",
      "Epoch 152/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 5ms/step - MAE: 4.7089 - loss: 74.1993 - val_MAE: 2.5537 - val_loss: 22.0447\n",
      "Epoch 153/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 7ms/step - MAE: 4.9683 - loss: 83.0249 - val_MAE: 2.5539 - val_loss: 21.8776\n",
      "Epoch 154/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - MAE: 4.7845 - loss: 74.9301 - val_MAE: 2.5336 - val_loss: 21.9973\n",
      "Epoch 155/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 6ms/step - MAE: 4.8338 - loss: 76.0924 - val_MAE: 2.5645 - val_loss: 22.0398\n",
      "Epoch 156/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 6ms/step - MAE: 4.7397 - loss: 71.4146 - val_MAE: 2.5577 - val_loss: 21.8866\n",
      "Epoch 157/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 8ms/step - MAE: 4.8100 - loss: 76.0041 - val_MAE: 2.5152 - val_loss: 21.4265\n",
      "Epoch 158/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 5ms/step - MAE: 4.8752 - loss: 78.2884 - val_MAE: 2.5458 - val_loss: 21.7469\n",
      "Epoch 159/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 5ms/step - MAE: 4.5425 - loss: 64.5808 - val_MAE: 2.5540 - val_loss: 21.5328\n",
      "Epoch 160/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step - MAE: 4.8676 - loss: 77.7115 - val_MAE: 2.5161 - val_loss: 21.7266\n",
      "Epoch 161/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 7ms/step - MAE: 4.7965 - loss: 75.7461 - val_MAE: 2.5114 - val_loss: 21.4202\n",
      "Epoch 162/10000\n",
      "\u001b[1m323/323\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 6ms/step - MAE: 4.6928 - loss: 70.6058 - val_MAE: 2.5316 - val_loss: 21.6164\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "input1 = Input(shape=(X_train.shape[1],), name='input')\n",
    "normalized_input = BatchNormalization()(input1)\n",
    "\n",
    "ann_layer1 = Dense(32, activation='relu')(normalized_input)\n",
    "dropout = Dropout(0.1)(ann_layer1)  \n",
    "ann_layer2 = Dense(32, activation='relu')(dropout)\n",
    "dropout2 = Dropout(0.1)(ann_layer2)  \n",
    "\n",
    "output_layer = Dense(1, activation='relu')(dropout2)\n",
    "\n",
    "model = tf.keras.Model(inputs=input1, outputs=output_layer)\n",
    "\n",
    "optimizer = tf.keras.optimizers.Adam(learning_rate=0.0001, clipvalue=1.0)\n",
    "model.compile(optimizer=optimizer, loss=\"MSE\", metrics=[\"MAE\"])\n",
    "\n",
    "early_stopping = EarlyStopping(monitor='val_loss', mode='min', patience=5, restore_best_weights=True, min_delta=0.01, start_from_epoch=100)\n",
    "\n",
    "history = model.fit(X_train, y_train, epochs=10000, batch_size=64, verbose=1, validation_data=(X_val, y_val), callbacks=[early_stopping], shuffle=True)\n",
    "\n",
    "plt.plot(history.history['loss'], label='train')\n",
    "plt.plot(history.history['val_loss'], label='val')\n",
    "plt.title('Loss Training History')\n",
    "plt.xlabel('Epochs')\n",
    "plt.ylabel('Loss')\n",
    "plt.legend()\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m81/81\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAdsAAAE8CAYAAACIOxsOAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8g+/7EAAAACXBIWXMAAA9hAAAPYQGoP6dpAACNmElEQVR4nO2dd3yT1ffHP0lHuveG0hbKngICZY/KRrYstSAqIqiAIOBgCVZBEUSG/mQqoGylMr4gUAQKSFmyymoppQso3SNtc39/HJ+0adOStEkHnPfrlVfJ89w8z7lp6Cfn3jNkQggBhmEYhmGMhryyDWAYhmGYZx0WW4ZhGIYxMiy2DMMwDGNkWGwZhmEYxsiw2DIMwzCMkWGxZRiGYRgjw2LLMAzDMEaGxZZhGIZhjAyLLcMwDMMYGRbbZwhfX1+MHTtW/fzYsWOQyWQ4duxYpdlUlKI2MoZh3rx5kMlklW1GiYwdOxa+vr6VbUaVIyoqCjKZDF9//XVlm2IQ+PdcMiy2BmLDhg2QyWTqh4WFBerVq4fJkycjISGhss3Ti3379mHevHmVbUalkZ+fDy8vL8hkMuzfv7/M19myZQuWLVtmOMMMSEhICHr37g1nZ2f1Z3X69Ol4/Phxma8ZGxuLefPm4eLFi4YztAIo/H/3xIkTxc4LIeDt7Q2ZTIb+/ftXgoW6o+tn7vz585DJZPj0009LHHPr1i3IZDJMmzbNgBY+v7DYGpgFCxbg559/xvfff4/27dtj9erVCAgIQGZmZoXb0rlzZ2RlZaFz5856vW7fvn2YP3++kayq+hw5cgRxcXHw9fXF5s2by3ydqiq206dPx4ABAxAfH4+ZM2fi+++/R2BgIL7//ns0b94cERERZbpubGws5s+fr1Vs/+///q/M160oLCwssGXLlmLHQ0NDERMTA4VCUQlW6Yeun7mWLVuiQYMG2Lp1a6nXAoBXX33VUOY917DYGpg+ffrg1VdfxZtvvokNGzZgypQpiIyMxO+//17iazIyMoxii1wuh4WFBeRy/jXrwy+//IKWLVti6tSp2LNnj9F+P5XB1q1b8c0332DEiBEIDw/HRx99hDfffBOrVq3C8ePH8eTJEwwfPhx5eXkGva+ZmVmVF6u+ffti+/btxea+ZcsWtGrVCh4eHga7l0qlQnZ2tsGuVxbGjBmDu3fv4vTp01rPb926FQ0aNEDLli0r2LJnE/4rbGS6d+8OAIiMjARAexo2Nja4c+cO+vbtC1tbW4wZMwYA/QdctmwZGjduDAsLC7i7u2PChAl48uSJxjWFEFi4cCFq1qwJKysrdOvWDVevXi1275L2bM+cOYO+ffvC0dER1tbWaNasGZYvX662b+XKlQCgsSwuYWgbi5KbmwsnJyeMGzeu2LnU1FRYWFhg+vTp6mMrVqxA48aNYWVlBUdHR7Ru3Vqrd6IrWVlZ2L17N0aOHIlXXnkFWVlZJX5R2r9/P7p06QJbW1vY2dnhxRdfVN+7a9eu+PPPP3Hv3j31eyjtZUnLllFRURrX0/b7+vvvvzF8+HDUqlULCoUC3t7emDp1KrKysso0v/nz58PR0RE//vgjTExMNM61adMGM2fOxL///osdO3aoj3ft2hVNmjRBeHg42rdvD0tLS/j5+WHNmjUatr/44osAgHHjxqnnvGHDBgDa9/IyMjLw4YcfwtvbGwqFAvXr18fXX3+Noo3IZDIZJk+ejD179qBJkyZQKBRo3LgxDhw4oDEuLS0NU6ZMga+vLxQKBdzc3PDSSy/h/PnzOr03o0aNwuPHj3Ho0CH1MaVSiR07dmD06NFaX/P111+jffv2cHZ2hqWlJVq1aqXx3hWdw+bNm9G4cWMoFIpi9ksIIfD222/D3Nwcu3btUh//5Zdf0KpVK1haWsLJyQkjR47E/fv31edL+8xpQ/q7o+3/S3h4OCIiItRjfv/9d/Tr1w9eXl5QKBSoU6cOPv/8c+Tn55d4faDkv0HSXrX0+ZC4ceMGhg0bBicnJ1hYWKB169b4448/NMbk5uZi/vz5qFu3LiwsLODs7IyOHTtq/N6qIiy2RubOnTsAAGdnZ/WxvLw89OrVC25ubvj6668xdOhQAMCECRMwY8YMdOjQAcuXL8e4ceOwefNm9OrVC7m5uerXz5kzB5999hmaN2+OJUuWoHbt2ujZs6dOHtihQ4fQuXNnXLt2DR988AG++eYbdOvWDSEhIWobXnrpJQDAzz//rH5IGNtGMzMzDB48GHv27IFSqdQ4t2fPHuTk5GDkyJEAaGny/fffR6NGjbBs2TLMnz8fLVq0wJkzZ576PpTEH3/8gfT0dIwcORIeHh7o2rWr1qXkDRs2oF+/fkhKSsLs2bPx5ZdfokWLFuo/oJ988glatGgBFxcX9XtYliXl7du3IzMzExMnTsSKFSvQq1cvrFixAq+//rre17p16xYiIiIwcOBA2NnZaR0jXVf6PEg8efIEffv2RatWrbB48WLUrFkTEydOxLp16wAADRs2xIIFCwAAb7/9tnrOJW1hCCHw8ssv49tvv0Xv3r2xdOlS1K9fHzNmzNC6R3jixAm8++67GDlyJBYvXozs7GwMHTpUY4/5nXfewerVqzF06FCsWrUK06dPh6WlJa5fv67T++Pr64uAgACNpdX9+/cjJSVF/ZkryvLly/HCCy9gwYIF+OKLL2Bqaorhw4fjzz//LDb2yJEjmDp1KkaMGIHly5drFcL8/HyMHTsWmzZtwu7duzFkyBAAwKJFi/D666+jbt26WLp0KaZMmYK//voLnTt3RnJyMgD9P3N+fn5o3749tm3bVkw0JQGWvmRs2LABNjY2mDZtGpYvX45WrVphzpw5mDVrVonX15erV6+iXbt2uH79OmbNmoVvvvkG1tbWGDRoEHbv3q0eN2/ePMyfPx/dunXD999/j08++QS1atXS+UtVpSEYg7B+/XoBQBw+fFg8fPhQ3L9/X/z666/C2dlZWFpaipiYGCGEEEFBQQKAmDVrlsbr//77bwFAbN68WeP4gQMHNI4nJiYKc3Nz0a9fP6FSqdTjPv74YwFABAUFqY8dPXpUABBHjx4VQgiRl5cn/Pz8hI+Pj3jy5InGfQpfa9KkSULbR8MYNmrj4MGDAoDYu3evxvG+ffuK2rVrq58PHDhQNG7cuNRr6Uv//v1Fhw4d1M9//PFHYWpqKhITE9XHkpOTha2trWjbtq3IysrSeH3h+fbr10/4+PgUu4f0WYmMjNQ4XvT3JYQQmZmZxV4fHBwsZDKZuHfvnvrY3Llztf7OCrNnzx4BQHz77beljrOzsxMtW7ZUP+/SpYsAIL755hv1sZycHNGiRQvh5uYmlEqlEEKIf/75RwAQ69evL3bNoKAgjfdCsmXhwoUa44YNGyZkMpm4ffu2+hgAYW5urnHs0qVLAoBYsWKF+pi9vb2YNGlSqXPThvT7+Oeff8T3338vbG1t1e/78OHDRbdu3YQQQvj4+Ih+/fppvLbo70epVIomTZqI7t27axwHIORyubh69arG8cjISAFALFmyROTm5ooRI0YIS0tLcfDgQfWYqKgoYWJiIhYtWqTx2n///VeYmppqHC/pM1cSK1euFAA07pefny9q1KghAgICSpynEEJMmDBBWFlZiezsbPWxor9nbZ/pwvMu/Fnp0aOHaNq0qcb1VCqVaN++vahbt676WPPmzYv9HqoD7NkamMDAQLi6usLb2xsjR46EjY0Ndu/ejRo1amiMmzhxosbz7du3w97eHi+99BIePXqkfrRq1Qo2NjY4evQoAODw4cNQKpV47733NJZ3p0yZ8lTbLly4gMjISEyZMgUODg4a53RJG6kIGwFaendxccFvv/2mPvbkyRMcOnQII0aMUB9zcHBATEwM/vnnH52u+zQeP36MgwcPYtSoUepjQ4cOhUwmw7Zt29THDh06hLS0NMyaNQsWFhYa1zB0+o2lpaX63xkZGXj06BHat28PIQQuXLig17XS0tIAALa2tqWOs7W1RWpqqsYxU1NTTJgwQf3c3NwcEyZMQGJiIsLDw/WyA6AgPBMTE7z//vsaxz/88EMIIYpFgQcGBqJOnTrq582aNYOdnR3u3r2rPubg4IAzZ84gNjZWb3skpK2DkJAQpKWlISQkpMQlZEDz9/PkyROkpKSgU6dOWr2sLl26oFGjRlqvo1QqMXz4cISEhGDfvn3o2bOn+tyuXbugUqnwyiuvaPy/8/DwQN26ddX/78rCiBEjYGZmprGUHBoaigcPHqiXkIvOMy0tDY8ePUKnTp2QmZmJGzdulPn+EklJSThy5AheeeUV9fUfPXqEx48fo1evXrh16xYePHgAgH7PV69exa1bt8p934rEtLINeNZYuXIl6tWrB1NTU7i7u6N+/frFApRMTU1Rs2ZNjWO3bt1CSkoK3NzctF43MTERAHDv3j0AQN26dTXOu7q6wtHRsVTbpCXtJk2a6D6hCrYRoPdn6NCh2LJlC3JycqBQKLBr1y7k5uZqiO3MmTNx+PBhtGnTBv7+/ujZsydGjx6NDh06lGl+v/32G3Jzc/HCCy/g9u3b6uNt27bF5s2bMWnSJADlfx/1ITo6GnPmzMEff/xRbF88JSVFr2tJIiuJbkmkpaUV+x17eXnB2tpa41i9evUA0P5bu3bt9LLl3r178PLyKib8DRs2VJ8vTK1atYpdw9HRUeM9Wbx4MYKCguDt7Y1WrVqhb9++eP3111G7dm2d7XJ1dUVgYCC2bNmCzMxM5OfnY9iwYSWODwkJwcKFC3Hx4kXk5OSoj2v70uXn51fidYKDg5Geno79+/eja9euGudu3boFIUSx/08SZmZmT5lVyTg7O6NXr17YvXs31qxZo47INjU1xSuvvKIed/XqVXz66ac4cuRIsS9i+n4OtXH79m0IIfDZZ5/hs88+0zomMTERNWrUwIIFCzBw4EDUq1cPTZo0Qe/evfHaa6+hWbNm5bbDmLDYGpg2bdqgdevWpY5RKBTFBFilUsHNza3EVBNXV1eD2VhWKtLGkSNH4ocffsD+/fsxaNAgbNu2DQ0aNEDz5s3VYxo2bIiIiAiEhITgwIED2LlzJ1atWoU5c+aUKXVJmldJYn337l29/nCXREneb9F9s/z8fLz00ktISkrCzJkz0aBBA1hbW+PBgwcYO3YsVCqVXveVhOzy5csljrl37x5SU1NL9MAqi6LBXBKiUDDVK6+8gk6dOmH37t343//+hyVLluCrr77Crl270KdPH53vNXr0aLz11luIj49Hnz59iq0CSfz99994+eWX0blzZ6xatQqenp4wMzPD+vXrtQYdFfYOi9KrVy8cOHAAixcvRteuXTVWTFQqlTrnW9v7YGNjo/PctPHqq68iJCQEISEhePnll7Fz50707NlT/f85OTkZXbp0gZ2dHRYsWIA6derAwsIC58+fx8yZM0v9HOr6WZeuMX36dPTq1Uvra/z9/QFQSuOdO3fw+++/43//+x9++uknfPvtt1izZg3efPNNvedfUbDYVhHq1KmDw4cPo0OHDqX+p/Tx8QFA33YL/+F/+PBhMc9H2z0A4MqVKwgMDCxxXEn/QSrCRonOnTvD09MTv/32Gzp27IgjR47gk08+KTbO2toaI0aMwIgRI6BUKjFkyBAsWrQIs2fPLrbEWxqRkZE4deoUJk+ejC5dumicU6lUeO2117BlyxZ8+umnGu+j9AdAGyW9j5J3LwW2SBT15v7991/cvHkTGzdu1AiIKmvUZb169VCvXj3s2bMHy5cv17qcvGnTJgAoVrwhNjYWGRkZGt7tzZs3AUAd6KPPErqPjw8OHz6MtLQ0DTukJUnpM6Qvnp6eePfdd/Huu+8iMTERLVu2xKJFi/QS28GDB2PChAk4ffq0xlZGUXbu3AkLCwscPHhQI61p/fr1etvdrl07vPPOO+jfvz+GDx+O3bt3w9SU/jzXqVMHQgj4+fmpVxNKoizbGC+//DJsbW2xZcsWmJmZ4cmTJxpLyMeOHcPjx4+xa9cujYA3KcOiNHT9rEt/J8zMzEr92yQhZSyMGzcO6enp6Ny5M+bNm1elxZb3bKsIr7zyCvLz8/H5558XO5eXl6f+sAYGBsLMzAwrVqzQ+FavayK7n58fli1bVuzDX/ha0h/UomMqwkYJuVyOYcOGYe/evfj555+Rl5ensYQMoFi1I3NzczRq1AhCCHVktLSn9OjRo1LvJ3m1H330EYYNG6bxeOWVV9ClSxf1mJ49e8LW1hbBwcHFciWLvo/altgksT5+/Lj6WH5+Pn788UeNcZIXU/iaQgh1mlZZmDNnDp48eYJ33nmnmHcRHh6Or776Ck2aNFFHyEvk5eXhhx9+UD9XKpX44Ycf4OrqilatWgEo+XOjjb59+yI/Px/ff/+9xvFvv/0WMplML3EE6P0r+l67ubnBy8tLY3lXF2xsbLB69WrMmzcPAwYMKHGciYkJZDKZxvsYFRWFPXv26HU/icDAQPz66684cOAAXnvtNbW3N2TIEJiYmGD+/PnF0qKEEBr/D0r6zJWGpaUlBg8ejH379mH16tWwtrbGwIED1ee1fQ6VSiVWrVr11Gv7+PjAxMRE47MOoNhr3dzc0LVrV/zwww+Ii4srdp2HDx+q/130/72NjQ38/f31/j1XNOzZVhG6dOmCCRMmIDg4GBcvXkTPnj1hZmaGW7duYfv27Vi+fDmGDRsGV1dXTJ8+HcHBwejfvz/69u2LCxcuYP/+/XBxcSn1HnK5HKtXr8aAAQPQokULjBs3Dp6enrhx4wauXr2KgwcPAoD6j+f777+PXr16wcTEBCNHjqwQGwszYsQIrFixAnPnzkXTpk3Vy6ASPXv2hIeHBzp06AB3d3dcv34d33//Pfr166f2ls6ePYtu3bph7ty5pZag3Lx5M1q0aAFvb2+t519++WW89957OH/+PFq2bIlvv/0Wb775Jl588UWMHj0ajo6OuHTpEjIzM7Fx40b1+/jbb79h2rRpePHFF2FjY4MBAwagcePGaNeuHWbPno2kpCQ4OTnh119/LVZMoUGDBqhTpw6mT5+OBw8ewM7ODjt37tR5dUAbY8aMwT///IPly5fj2rVrGDNmDBwdHXH+/HmsW7cOzs7O2LFjR7F9QC8vL3z11VeIiopCvXr18Ntvv+HixYv48ccf1WPr1KkDBwcHrFmzBra2trC2tkbbtm217lUOGDAA3bp1wyeffIKoqCg0b94c//vf//D7779jypQpGsFQupCWloaaNWti2LBhaN68OWxsbHD48GH8888/+Oabb/R+n4KCgp46pl+/fli6dCl69+6N0aNHIzExEStXroS/v3+pS/WlMWjQIKxfvx6vv/467Ozs8MMPP6BOnTpYuHAhZs+ejaioKAwaNAi2traIjIzE7t278fbbb6tzz0v6zD2NV199FZs2bcLBgwcxZswYjRWM9u3bw9HREUFBQXj//fchk8nw888/FxN+bdjb22P48OFYsWIFZDIZ6tSpg5CQEHV8R2FWrlyJjh07omnTpnjrrbdQu3ZtJCQkICwsDDExMbh06RIAoFGjRujatStatWoFJycnnDt3Djt27MDkyZN1fZsrhwqPf35GKZw+UBpBQUHC2tq6xPM//vijaNWqlbC0tBS2traiadOm4qOPPhKxsbHqMfn5+WL+/PnC09NTWFpaiq5du4orV64IHx+fUlN/JE6cOCFeeuklYWtrK6ytrUWzZs00Uijy8vLEe++9J1xdXYVMJiuWUmJIG0tDpVIJb29vrSkiQgjxww8/iM6dOwtnZ2ehUChEnTp1xIwZM0RKSkqx92Du3Lkl3ic8PFwAEJ999lmJY6KiogQAMXXqVPWxP/74Q7Rv315YWloKOzs70aZNG7F161b1+fT0dDF69Gjh4OAgAGikRNy5c0cEBgYKhUIh3N3dxccffywOHTpU7Pd17do1ERgYKGxsbISLi4t466231GkvhdMmdEn9KcyePXvESy+9JBwdHYVCoRD+/v7iww8/FA8fPiw2tkuXLqJx48bi3LlzIiAgQFhYWAgfHx/x/fffFxv7+++/i0aNGglTU1MNG4umhAghRFpampg6darw8vISZmZmom7dumLJkiUa6VNCUNqMtpSewp+lnJwcMWPGDNG8eXP157p58+Zi1apVT30vdP2/qy31Z+3ataJu3bpCoVCIBg0aiPXr12v9XZQ0h8KpP4VZtWqVACCmT5+uPrZz507RsWNHYW1tLaytrUWDBg3EpEmTREREhHpMaZ+50sjLyxOenp4CgNi3b1+x8ydPnhTt2rUTlpaWwsvLS3z00UfqFL3Cn1dtv+eHDx+KoUOHCisrK+Ho6CgmTJggrly5ojVN7M6dO+L1118XHh4ewszMTNSoUUP0799f7NixQz1m4cKFok2bNsLBwUFYWlqKBg0aiEWLFqlT0KoqMiF0+HrCMMxzS9euXfHo0SNcuXKlsk1hmGoL79kyDMMwjJFhsWUYhmEYI8NiyzAMwzBGplLFdvXq1eqya3Z2dggICNAo05adnY1JkybB2dkZNjY2GDp0aLFG7NHR0ejXrx+srKzg5uaGGTNmGLw9GMM8zxw7doz3axmmnFSq2NasWRNffvklwsPDce7cOXTv3h0DBw5Ut2KbOnUq9u7di+3btyM0NBSxsbHqLhgA5db169cPSqUSp06dwsaNG7FhwwbMmTOnsqbEMAzDMMWoctHITk5OWLJkiTpfc8uWLerapDdu3EDDhg0RFhaGdu3aYf/+/ejfvz9iY2Ph7u4OAFizZg1mzpyJhw8fwtzcvDKnwjAMwzAAqlBRi/z8fGzfvh0ZGRkICAhAeHg4cnNzNUp3NWjQALVq1VKLbVhYGJo2baoWWoBqjE6cOBFXr17FCy+8oPVeOTk5GtVGVCoVkpKS4OzsbPCuLQzDMEz1QAiBtLQ0eHl5FatfX14qXWz//fdfBAQEIDs7W92OrlGjRrh48SLMzc2LFQF3d3dHfHw8ACA+Pl5DaKXz0rmSCA4OLlOheoZhGObZ5/79+8U6s5WXShfb+vXr4+LFi0hJScGOHTsQFBSE0NBQo95z9uzZmDZtmvp5SkoKatWqhfv378POzs6o92YYhmGqJqmpqfD29n5qz+eyUOlia25uru6c0qpVK3XdVqmLS3JysoZ3m5CQAA8PDwCAh4cHzp49q3E9KVpZGqMNhUKh0aVDQoqKZhiGYZ5fjLGdWOXybFUqFXJyctCqVSuYmZnhr7/+Up+LiIhAdHQ0AgICAAABAQH4999/NYpaHzp0CHZ2dlWuHyfDMAzz/FKpnu3s2bPRp08f1KpVC2lpadiyZQuOHTuGgwcPwt7eHuPHj8e0adPg5OQEOzs7vPfeewgICEC7du0AUNeXRo0a4bXXXsPixYsRHx+PTz/9FJMmTdLquTIMwzBMZVCpYpuYmIjXX38dcXFxsLe3R7NmzXDw4EG89NJLAKi3pVwux9ChQ5GTk4NevXpp9EE0MTFBSEgIJk6ciICAAFhbWyMoKAgLFiyorCkxDMMwTDGqXJ5tZZCamgp7e3ukpKTwni3DMMxzijG1oMrt2TIMwzDMswaLLcMwDMMYGRZbhmEYplqTnJyM1NTUyjajVFhsGYZhmGqJSqXChg0bUL9+/SpfFZDFlmEYhql2nD9/Hh07dsS4cePQo0cPjaqAVZFKryDFMAzDMPoQGRmJNm3aoEGDBjh69Ci6du1a2SY9FfZsGYZhmCqPSqXCjh07oFKp4Ofnh5CQEFy4cKFaCC3AYsswDMNUcc6ePYt27dph+PDhOH78OACgd+/eMDMzq2TLdIfFlmEYhqmSPHr0CG+99RbatWsHpVKJv//+u9p4skXhPVuGYRimSrJr1y7s2LEDK1aswIQJE2BqWn0li8s1gss1MgzDVBVOnTqFv//+GzNnzkR+fj6SkpLg6upaIffmco0MwzDMM01CQgLGjh2LDh06YNeuXcjJyYGJiUmFCa2xYbFlGIZhKpXvv/8e9erVw969e/HDDz/g1KlTz1ybVBZbhmEYplKQdjGvXbuG0aNH4+bNm3j77bdhYmJSyZYZnuq728wwDMNUS2JjYzF9+nR07NgR7777LlauXAmZTFbZZhkV9mwZhmGYCiE3Nxdff/016tevj8OHD8PR0REAnnmhBdizZRiGYSqA2NhYBAYGIiIiApMmTcKCBQvg4OBQ2WZVGCy2DMMwjNF48uQJHB0d4eHhgR49euDXX39Fs2bNKtusCoeXkRmGYRiDk5OTg+DgYHh7e+PEiROQy+VYsWLFcym0AHu2DMMwjIE5ePAg3nvvPdy9excffPDBcyuwhWHPlmEYhjEYa9euRe/evVGzZk1cunQJ33zzDVfmA3u2DMMwTDnJzs7GhQsXEBAQgGHDhsHOzg7Dhg17LqKMdYU9W4ZhGKbM7N27F40bN0b//v2RmZkJe3t7DB8+nIW2CCy2DMMwjN7cuXMH/fv3x8svv4w6derg5MmTsLKyqmyzqiy8jMwwDMPozaRJk3D9+nXs3LkTgwcPZk/2KbDYMgzDME9FCIE9e/bA2dkZnTt3xtq1a+Ho6MjerI5U6jJycHAwXnzxRdja2sLNzQ2DBg1CRESExpiuXbtCJpNpPN555x2NMdHR0ejXrx+srKzg5uaGGTNmIC8vryKnwjAM88wSERGB3r17Y8iQIdixYwcAoEaNGiy0eqC32J4/fx7//vuv+vnvv/+OQYMG4eOPP4ZSqdTrWqGhoZg0aRJOnz6NQ4cOITc3Fz179kRGRobGuLfeegtxcXHqx+LFi9Xn8vPz0a9fPyiVSpw6dQobN27Ehg0bMGfOHH2nxjAMwxQiMzMTs2bNQtOmTXHr1i388ccfWL58eWWbVT0RetK6dWuxY8cOIYQQd+7cERYWFmLUqFHC399ffPDBB/peToPExEQBQISGhqqPdenSpdTr7tu3T8jlchEfH68+tnr1amFnZydycnJ0um9KSooAIFJSUspsO8MwzLNGWlqaqFu3rpg/f77IzMysbHOMjjG1QG/P9ubNm2jRogUAYPv27ejcuTO2bNmCDRs2YOfOneUS/pSUFACAk5OTxvHNmzfDxcUFTZo0wezZs5GZmak+FxYWhqZNm8Ld3V19rFevXkhNTcXVq1e13icnJwepqakaD4ZhGIZ6y/bv3x/37t2DjY0Nrl69ijlz5sDS0rKyTavW6C22QgioVCoAwOHDh9G3b18AgLe3Nx49elRmQ1QqFaZMmYIOHTqgSZMm6uOjR4/GL7/8gqNHj2L27Nn4+eef8eqrr6rPx8fHawgtAPXz+Ph4rfcKDg6Gvb29+uHt7V1muxmGYZ4FUlNT8eGHH6J58+aIiIhAYmIiAMDMzKySLXs20DsauXXr1li4cCECAwMRGhqK1atXAwAiIyOLiZ4+TJo0CVeuXMGJEyc0jr/99tvqfzdt2hSenp7o0aMH7ty5gzp16pTpXrNnz8a0adPUz1NTU1lwGYapUggBpKcDubmAmRlgYwMYK7vm6NGjGDNmDFJSUrBgwQJMmzYNCoXCODd7TtFbbJctW4YxY8Zgz549+OSTT+Dv7w8A2LFjB9q3b18mIyZPnoyQkBAcP34cNWvWLHVs27ZtAQC3b99GnTp14OHhgbNnz2qMSUhIAAB4eHhovYZCoeAPEsMwVZbkZODOHSAxsUBs3dyAOnUAQ7aAzc7OhoWFBWrWrIkuXbrgq6++Qq1atQx3A0aN3mLbrFkzjWhkiSVLlsDExESvawkh8N5772H37t04duwY/Pz8nvqaixcvAgA8PT0BAAEBAVi0aBESExPh5uYGADh06BDs7OzQqFEjvexhGIapbJKTgfBw8mqdnQGFAsjJAWJi6FyrVuUX3OTkZMydOxcHDhzApUuXULduXWzdutUA1jMlUaY82+TkZPz000+YPXs2kpKSANCmurTGryuTJk3CL7/8gi1btsDW1hbx8fGIj49HVlYWACoH9vnnnyM8PBxRUVH4448/8Prrr6Nz587qlk09e/ZEo0aN8Nprr+HSpUs4ePAgPv30U0yaNIm9V4ZhqhVCkEebng7UrAlYWgJyOf2sWZOO371L48qCSqXChg0bUL9+faxbtw5vvfWW3k4SUzZkQuj3a7t8+TJ69OgBBwcHREVFISIiArVr18ann36K6OhobNq0Sfebl7ABsX79eowdOxb379/Hq6++iitXriAjIwPe3t4YPHgwPv30U42WTffu3cPEiRNx7NgxWFtbIygoCF9++SVMTXVz3FNTU2Fvb4+UlBRuBcUwTKWRlgacOAHY2pLAFiUri8Z07Ehj9GX06NHYunUrRo0ahSVLlqBGjRrlN/oZwphaoLfYBgYGomXLlli8eDFsbW1x6dIl1K5dG6dOncLo0aMRFRVlUAMrAhZbhmGqAklJJLaenuTRFiU/H4iPJ7EtkiFZyjWTkJWVhRo1auDIkSMwMTFBly5dDGv4M4IxtUDvZeR//vkHEyZMKHa8Ro0aJabaMAzDME/HzIweOTnazyuVBWOehkqlwv/93/+hXr16mDp1KgCge/fuz6TQCkEef1IS/SzrMrsx0TtASqFQaC0CcfPmTbi6uhrEKIZhmOcRGxuKOo6JoT3aojx+DHh707jSOHv2LCZPnox//vlHva32rFJRkdvlRW/P9uWXX8aCBQuQm5sLgPZdo6OjMXPmTAwdOtTgBjIMwzwvyGQkEjY2JLhZWbR0nJVFz21sgNq1S8+3ffLkCbp164bc3FycOHECGzZsKDENsrojRW7HxNAetqcn/YyJoePJyZVtYQF6i+0333yD9PR0uLm5ISsrC126dIG/vz9sbW2xaNEiY9jIMAzz3ODgQOk9NWvSkmh8PP309i457Sc/Px9r165FRkYGHB0dERoainPnzqFDhw4VbX6FYezIbUOj9zKyvb09Dh06hBMnTuDy5ctIT09Hy5YtERgYaAz7GIZhnjscHICWLXWrIHXq1ClMmjQJFy9ehKOjI4YMGYLWrVtXuM0VTXo6LR07O2s/7+wMJCTQuLJEbhuaMjeP79ixIzp27GhIWxiGYZj/kMlKF4mEhATMnDkTGzduROvWrXHmzBm0adOm4gysZHJz6VFSOQVz84IxVQG9xXbBggWlnuc+sgzDMMbn8uXLCAkJwQ8//IDx48c/d8UpCkdua8tJ1idyuyLQO8/2hRde0Hiem5uLyMhImJqaok6dOjh//rxBDawIOM+WYZjqQGhoKH777TesXLkSMpkM6enpsHlaaPIzihDA+fMlR27HxNA+9wsv6N7AwZhaoLdne+HChWLHUlNTMXbsWAwePNggRjEMwzAFxMbGYvr06di6dSvatWuH5ORkODo6PrdCCxREbicnk7A6O9PSsVJJKVK6RG5XJGWqjVwUOzs7zJ8/H5999pkhLscwDMP8x8qVK1G/fn0cPnwY69atw8mTJ+Ho6FjZZlUJyhK5XVmUOUCqKCkpKUhJSTHU5RiGYZ5rVCoV5HI5lEolxo0bhwULFsChKqlHFUGfyO3KRG+x/e677zSeCyEQFxeHn3/+GX369DGYYQzDMM8j9+/fx4cffoiaNWti6dKl6lKLTMk8LXK7KqC32H777bcaz+VyOVxdXREUFITZs2cbzDCGYZjniZycHCxduhQLFy6EnZ0dli5dWtkmMQZEb7GNjIw0hh0MwzDPLWlpaWjdujXu3LmDDz74AHPnzuXMiGcMg+3ZMgzDMPoRExMDLy8v2NraYvz48ejXrx8aN25c2WYxRkCnPNshQ4bofMFdu3aVy6DKgPNsGYapSLKzs7F48WIEBwfjp59+wpgxYyrbJAZVIM/W3t7eoDdlGIZ5Xtm7dy+mTJmC+/fvY9q0aRg4cGBlm8RUADqJ7fr1641tB8MwzDPP77//jkGDBuGll17Cvn37UL9+/co2iakgDFLUgmEYhtFOZmYm9u7dCwDo168f9u3bh4MHD7LQPmeUKUBqx44d2LZtG6Kjo6FUKjXOVcfayAzDMIZGCIHdu3dj6tSpePjwIaKiouDm5sb1CJ5T9PZsv/vuO4wbNw7u7u64cOEC2rRpA2dnZ9y9e5c/RAzDMAAiIiLQu3dvDB06FE2aNMHly5fh5uZW2WYxlYjeYrtq1Sr8+OOPWLFiBczNzfHRRx/h0KFDeP/997lcI8Mw1QIhqIZuUhL91K/32dNZunQpbt26hT/++AN//vkn/P39DXsDptqhd4s9KysrXL9+HT4+PnBzc8OhQ4fQvHlz3Lp1C+3atcPjx4+NZavR4NQfhnl+SE4G7twBEhMLaum6uVEHmbKWHhZCYNu2bVCpVBg1ahRSUlJgbm4OS22NVpkqizG1QG/P1sPDA0lJSQCAWrVq4fTp0wCospSeus0wDFOhJCcD4eHUks3WFvD0pJ8xMXQ8OVn/a169ehU9evTAyJEjcfjwYQCULslCyxRGb7Ht3r07/vjjDwDAuHHjMHXqVLz00ksYMWIE97NlGKbKIgR5tOnp1JLN0hKQy+lnzZp0/O5d3ZeUs7Oz8eGHH6JFixaIiYnB/v37sXbtWuNOgqm26LyMHBISgr59+wKg1k+mphTI/Ouvv+LUqVOoW7cuJkyYAHNzc+NZayR4GZlhnn3S0oATJ8iT1eZ0ZmXRmI4dS+8gI4SATCaDSqVCYGAgevbsialTp0KhUBjPeKZCqBLLyIMGDYK3tzc+++wz3Lt3T3185MiR+O677/Dee+/pLbTBwcF48cUXYWtrCzc3NwwaNAgREREaY7KzszFp0iQ4OzvDxsYGQ4cORUJCgsaY6Oho9OvXD1ZWVnBzc8OMGTOQl5enly0Mwzzb5ObSoyRNNDcvGFMSly5dQteuXXHq1CnI5XL89ddfmDVrFgst81R0FtvIyEhMmDABv/76K+rVq4cuXbrg559/RlZWVplvHhoaikmTJuH06dM4dOgQcnNz0bNnT2RkZKjHTJ06FXv37sX27dsRGhqK2NhYjVrN+fn56NevH5RKJU6dOoWNGzdiw4YNmDNnTpntYhjm2cPMjB45OdrPK5UFY4qSnJyM999/Hy1btkRiYqL6uKyqdShnqi6iDBw5ckS8/vrrwtraWtjb24sJEyaIs2fPluVSGiQmJgoAIjQ0VAghRHJysjAzMxPbt29Xj7l+/boAIMLCwoQQQuzbt0/I5XIRHx+vHrN69WphZ2cncnJydLpvSkqKACBSUlLKPQeGYaomKpUQ584JsWcP/Sz62LNHiPBwGleYsLAw4ebmJmxsbMSSJUt0/rvCVD+MqQVlKtfYrVs3bNy4EXFxcViyZAn+/fdftGvXDs2bNy+X8Et5uk5OTgCA8PBw5ObmIjAwUD2mQYMGqFWrFsLCwgAAYWFhaNq0Kdzd3dVjevXqhdTUVFy9elXrfXJycpCamqrxYBjm2UYmo/QeGxuKPs7KAvLz6WdMDB2vXZvGAcCTJ08A0N+cwYMHIyIiAtOnT6+WcSlM5VOu2si2trbo0aMHunXrBgcHB1y7dq3M11KpVJgyZQo6dOiAJk2aAADi4+Nhbm4OhyLJb+7u7oiPj1ePKSy00nnpnDaCg4Nhb2+vfnh7e5fZboZhqg8ODkCrVhR9nJYGxMfTT29vOu7gACQlJWHixImoXbs2EhIS4ODggDVr1sDLy6uyzWeqMWWqjZyVlYXt27dj3bp1+Pvvv+Hn54dp06Zh7NixZTZk0qRJuHLlCk6cOFHma+jK7NmzMW3aNPXz1NRUFlyGeU5wcABatqRUH6mohY0NoFLl48cf1+Ljjz9Gbm4uFixYoF5lY5jyopfYnj59GuvWrcO2bdugVCoxZMgQHD58GN26dSuXEZMnT0ZISAiOHz+OmjVrqo97eHhAqVQiOTlZw7tNSEiAh4eHeszZs2c1ridFK0tjiqJQKDh6kGGeY2Sy4uk977//PlatWoWgoCB89dVXxVbMGKY86LyM3KhRI3To0AHnz59HcHAw4uLi8Msvv5RLaIUQmDx5Mnbv3o0jR47Az89P43yrVq1gZmaGv/76S30sIiIC0dHRCAgIAAAEBATg33//1YgQPHToEOzs7NCoUaMy28YwzLPPw4cPceXKFQC0unby5Els2LCBhZYxODp7toGBgdi6dWu5g6AKM2nSJGzZsgW///47bG1t1XusUqkze3t7jB8/HtOmTYOTkxPs7Ozw3nvvISAgAO3atQMA9OzZE40aNcJrr72GxYsXIz4+Hp9++ikmTZrE3ivDMFrJz8/HmjVr8Omnn6JJkyb4+++/+cs5Y1wMHt+sBwC0PtavX68ek5WVJd59913h6OgorKysxODBg0VcXJzGdaKiokSfPn2EpaWlcHFxER9++KHIzc3V2Q5O/WGY54cTJ06IFi1aCJlMJt58802RmJhY2SYxVQRjaoHeXX+eRbhcI8M8ewhRPAhKqcxB7dq14eXlhZUrV6JNmzaVbSZThTCmFpQpGplhGKYqU7iNXnZ2Hg4cWImXXx6CDh28cfz4cfj5+UEuL3vmozYh52JSTGmw2DIM80whtdFLTweio0OxfPlk3L17FebmtrCweAOtWtVBOXTWKP1wmWcfvT9ymzZtQo6W4qJKpRKbNm0yiFEMwzBlQWqjFx0dh9WrR+P997vCysoGP/98DmPHvqF3G72iGKMfLvN8oLfYjhs3Tl1WsTBpaWkYN26cQYxiGIYpC+np5HEqFKm4ePFvzJ27HmvXnkSDBi0BAM7OQEICjdMXQ/fDZZ4v9BZb8V8vx6LExMTA3t7eIEYxDMPoy+HDhzF8+ABkZGSjbt36+P33uxgwYKzG3qwubfRKQhJyZ2ft58sj5Myzj857ti+88AJkMhlkMhl69Oihbh4PUM5aZGQkevfubRQjGYZhSuL+/fuYNm0aduzYgfbtOyE7Owk5OV6wtCzeK6+0NnpPwxD9cJnnF53FdtCgQQCAixcvolevXrCxsVGfMzc3h6+vL4YOHWpwAxmGYUrixx9/xNSpU2FnZ4dffvkFo0aNxoULMsTE0NJuUR4/pqYDhf586UzhfriWlsXPl0fImWcfncV27ty5AABfX1+MGDECFhYWRjOKYRimNLKzs2FhYQF3d3dMnDgRc+bMUedF1qlDgUoxMbS0a25OQvj4cfE2evpgY0NRx8YQcubZp0xFLZKTk7Fjxw7cuXMHM2bMgJOTE86fPw93d3fUqFHDGHYaFS5qwTDVg6ioKEyZMgUqlQp//PFHieO0pee4u5PQlic9p3BakTYhl9r0MdWTKlXU4vLlywgMDIS9vT2ioqLw1ltvwcnJCbt27UJ0dDSn/zAMY3CysrKwZMkSBAcHw9nZGd98802JwZpAyW30ylt4QuqHW1TIvb3LL+TMs43eYjt16lSMHTsWixcvhm2hHlV9+/bF6NGjDWocwzDPH0WrM1lY5KF169a4desWpk2bhk8//VQjZqQktLXRMwTGEnLm2UZvsT137hx+/PHHYsdr1Kih7trDMMzzTVnLGRZe/o2OvgM3N0/UqmWFKVM+QefOrVC/fn2j264LxhJy5tlFb7FVKBRITU0tdvzmzZtwdXU1iFEMw1RfylrOUNoPvX8/EwcOBGPXrsUYN24++vadhdq1R4NbzDLVGb2LWrz88stYsGABcv9LJpPJZIiOjsbMmTM59YdhnnPKWs5QCODCBYGff96FadMaYseOxejW7SO0afM+7O25OhNT/dFbbL/55hukp6fDzc0NWVlZ6NKlC/z9/WFra4tFixYZw0aGYaoB5SlnGBMDbN9+Fhs3DoW3d1MEB1/FK698juRkK1y/TlG/XJ2Jqc7ovYxsb2+PQ4cO4eTJk7h06RLS09PRsmVLBAYGglvjMszziz7lDKX9zvT0dGzZshV16rwJR8e2mDPnNOrWbat+jbs7vebRI8DOjqszMdUXvT3bJUuWAAA6dOiAd999Fx999BECAwORn5/P0cgM8xyjTzlDIQR+++03NGjQAB988D6uXr0JR0fA27ttsdc5OADx8UBeHldnYqovZRLbtWvXahzLz8/HyJEjcfHiRUPZxTBMFUYIIC0NSEqin0JoljPUhlTO8M6dawgMDMTIkSPx4osvIizsOnx86sPNDdDSUAympsCTJ4CTE1dnYqovei8j//nnn+jZsyfs7e0xbNgw5OXl4ZVXXsGNGzdw9OhRY9jIMEwVoqRo49q1dStnePDgH7h//z7279+P3r17Iy0NiIsDXF2BjAxaNra3L6jOlJhIy87+/pzLylRfylSu8ciRIxg0aBB++eUXrF27Frdv38aRI0fgXk1j87lcI8PoRtFyhQoFebJSuUJ3d+D2bSAzk87L5YBKJbB//2akp8fg669nwcpKCSEEFP+tNwsBnD9PIm1vTz+TkmjZ2NSUfrZqBXTuzGLLGJcqVa4RALp3745NmzZh6NChaNiwIUJDQ+Hi4mJQwxiGqVoUjTaWsLSk4KWwMBJXKysgOprGZWZeQmjoZERHn8DAgaNhby8gk5lrXFcmK2gekJIC+PgANWoAWVlAairg4gI0b85Cy1RvdBLbIUOGaD3u6uoKBwcHvP322+pju3btMoxlDMMYjLJWdCpMSdHGcXHAiRMUxKRUAtbWgKmpCv/8MwUXLqyEl1c9zJt3CB07BiImhs4XtaGkmsP16nHNYebZQCextbe313q8V69eBjWGYRjDU9aKTkXRFm2cmgqcPEn7rF5ewN27KuTny2BjI0d+fjZ69vwKo0a9j1q1zHH+PHDtGlCrFu3HFrVBqjmclkYBUQDg6Fi1yiIa4ksL83yik9iuX78eAIXr379/H66urrDU1j2ZYZgqRUl7rDExdE6XlnCSwKSlkchkZ9NSsRC0P/voES0r37sXjkOHJqNdu0nw83sV3br9CBsb8nyTkui+QtC+rKmpdhtSUgzzxcAYGOpLC/N8olfqjxAC/v7+iImJMZY9DMMYiPJUdJJITqbgpRMngIsXgQcPgLNnSXgzM0lohUjCgQMT8fPPLyI/Px3u7r4AyOvLyADu3y+wgQKmtNtQ1lKPFUFVto2pHugltnK5HHXr1sXjx48NcvPjx49jwIAB8PLygkwmw549ezTOjx07FjKZTOPRu3dvjTFJSUkYM2YM7Ozs4ODggPHjxyOda7oxjF4VnbRRVGC8vIC6dUlAjx8njzUu7go2b66Hy5e3oEuXbzFmzAXUqNERAHmvGRkkyo6OtJ9rakqPojakpZX/i4GxMMSXFobRu6jFl19+iRkzZuDKlSvlvnlGRgaaN2+OlStXljimd+/eiIuLUz+2bt2qcX7MmDG4evUqDh06hJCQEBw/flwjYIthnlf0qehUlJIExs2NUnByc+8jOhqwtW2AJk0mICgoAiNHfgAPD1O1eOflkdBaW5NYp6RQYQorq+I2PHlSvi8GxqS8X1oYBihD6s/rr7+OzMxMNG/eHObm5sX2bpOSknS+Vp8+fdCnT59SxygUCnh4eGg9d/36dRw4cAD//PMPWrduDQBYsWIF+vbti6+//hpeXl4628IwzxqFKzppC7GQKjppK4FYksA8efIQ338/GyEhG7FkyQUEBDSBn98iKJXkxTo40GsTEsgzdnEhgX3wgLzbmjU1A4okG4CyfzEwNuX50sIwEnqL7bJly4xgRskcO3YMbm5ucHR0RPfu3bFw4UI4//cXICwsDA4ODmqhBYDAwEDI5XKcOXMGgwcP1nrNnJwc5BSqKaetPy/DVHdsbMgTvX+fBE8qEmFlRYInVXTSVgKxqMDk5+dj5841WL36UwDAtGnL4evbEI0bk5DHx5PopKfTax4/JnFt0oS827w8oGHD4pHFkg2OjmX/YmBsyvOlhWEk9BbboKAgY9ihld69e2PIkCHw8/PDnTt38PHHH6NPnz4ICwuDiYkJ4uPj4ebmpvEaU1NTODk5IT4+vsTrBgcHY/78+cY2n2EqFZmMPMszZyjIycoKsLAgcbW0BDw8KIdVW+pKUYH56acF+OmnzzFw4HhMmvQFLCxckZZG13d0pCVnae/V0xMIDAR8fQuCpCIiaBnZ1LSgDKNUdap2bRJhXUo9VkZtZOlLS1W0jak+lKmClER2djaUSqXGMUOWuBo5cqT6302bNkWzZs1Qp04dHDt2DD169CjzdWfPno1p06apn6empsLb27tctjJMVSM5mVJzpDrD6elUlSkpiYKdOnYsOWXFxgYwMUnAsWPX0adPVwwfPgkdOvRDkyZtAJDwSAIjk1F+bEn5p05OJKZF02a8vTULVkhVpGJiaPlamyhXRk5r4QpXVc02pvqgt9hmZGRg5syZ2LZtm9ao5Pz8fIMYpo3atWvDxcUFt2/fRo8ePeDh4YHExESNMXl5eUhKSipxnxegfWBFSRswDPMMUDjAqUEDei4t55qY0B7q3bsFRSMKC0VeXh5WrlyJzz6bAwcHTzRufA2urm5o2NANWVnaBUYmK734hFSworSCECVVkSoqypVBVbaNqR7oLbYfffQRjh49itWrV+O1117DypUr8eDBA/zwww/48ssvjWGjmpiYGDx+/Bienp4AgICAACQnJyM8PBytWrUCQE0SVCoV2rYt3heTYZ4XigY4yWQUFZyWRnu4cXG0tJuURMu9UmGG0NBQTJ48GVevXsU777yD6dM/x5MncoMIzNMEGdBNlCuLqmwbU/XRW2z37t2LTZs2oWvXrhg3bhw6deoEf39/+Pj4YPPmzRgzZozO10pPT8ft27fVzyMjI3Hx4kU4OTnByckJ8+fPx9ChQ+Hh4YE7d+7go48+gr+/v7pMZMOGDdG7d2+89dZbWLNmDXJzczF58mSMHDmSI5GZ5xptEbRpacD16+ThOjgU5IpKlZxathSYPn06bGxscO7cObRs2RJAQe9aQ5dQLKn0oS6iXFlUZduYqo3eYpuUlITatWsDoP1ZKdWnY8eOmDhxol7XOnfuHLp166Z+Lu2jBgUFYfXq1bh8+TI2btyI5ORkeHl5oWfPnvj88881loA3b96MyZMno0ePHpDL5Rg6dCi+++47fafFMNUGXerzFg5wsrCgIKUbN0gwvb3puJkZoFAo8b//fYcaNTrB0bEtQkL+hKurC+TyghR8qYRiQgL9W6kkj7lZM7pWWTw7Ln3IPG/oLba1a9dGZGQkatWqhQYNGmDbtm1o06YN9u7dCwc9/5d07doVpbXTPXjw4FOv4eTkhC1btuh1X4aprjxNpCQhlrrv3LtHx2JjgVu36Fh+PglkevphfPvte4iOvonJk5ciIaEt6tZ1QyGdVVeRio+nvdroaPqZmUmRyL16AS+8oL2TT2lzKG+9ZoapbugttuPGjcOlS5fQpUsXzJo1CwMGDMD333+P3NxcLF261Bg2MgyDp4uUvz/VKpaEOCkJuHCBlos9PWnJ2NwcuHo1AeHhk3H79g688EInBAf/htq1myE+XrMwgxRkFR9Pj1u3KLjKy4sE9dYtYONG4J9/KIfW0fHp3mlpPXFr1qS53L1LAs57ocyzhN5iO3XqVPW/AwMDcePGDYSHh8Pf3x/NmjUzqHEMwxBPE6kbN0j8PDzI4zQ3J9E1MaF/Z2UBmZkCcrkMfn5mOHYsAu+88wveeGM05HIZsrKKF2aQKkFlZpIIWlpSGg9A1zM3J0FPTCShl8loTzg2ltKKHB2Lz6Nw4FbhCGmp2Ebh0oe8N8o8S5QrzxYAfHx84OPjYwhbGIYpgdLq8wpB4hcbCzRqRKIoNQBo1Ehaet6P//1vLoYPD0HDhm74/PNLyMoikbW21l6YITeX+tUmJVGnHimFXgja+83NJUHNySGPu2ZNEs2bN2lvt1+/4oIrBW4plUBUFF1bElsnJ/LAufQh8yyis9hu2rRJp3Gvv/56mY1hGEY72qKLJc8wNZWWjy0sSLgA+pmXB8TFReKXX6bi5s3f4ePTHTExmcjOBvz9ZVCpSMSfPNFemMHMjJagMzLop9StR6ksKMsoFcnIzaXlY0dHsuPmTSqm0amT5pKyFLR19y7tHTs40LHcXPJoHz0CatTg0ofMs4fOYjt27FjY2NjA1NS0xKAmmUzGYsswRqBo+cS0NApWunePxC42lkQqK4tEzsQEOHFiE/74YwIUChe0a7cNLVoMQ16eDA8eUJ5trVqUY+vnpz1v1saGlqXDw8mzzcujpeP8fHrk5ZE9UjCWXE5ibWND10pJKb7/am1Nr0lMJK9bQqEA3N2Ba9cAV1caxzDPEjqLbcOGDZGQkIBXX30Vb7zxBu/PMoyBKS2lp3B9Xnt7auZ+9Sp5tkoliW1CAqBQCNStmwwTE0fk5DSDl9cU+Ph8AhMTG5ibk3h6eACXL9N1W7cm0dUWjCSTAU2bAv/+C5w7RwLr6krCm5VFYx4+JGFPTyfhT04m+0xNyd6i+68ZGQXCmpBQUEpSqSRxdnen5xkZvGfLPFvoLLZXr17FmTNnsG7dOnTu3Bn+/v4YP368unE7wzD6Iwnso0dU2Sk9nTzGoik9hevzHjpEjQVMTGgfVS6npdsHD25j1aoPoFJFY8yYC3B1bYFGjVogMZH2Q588ISHLy6OAJhMT8o5r1SrZPkdHoH9/IDsbuHiR7u/oSMFMUVEkuv7+gI8PCWxSEpWCfOEFEtKEBM39V2kpvGlT+oKQlETL4KamJLSeniS0vGfLPGvoFSDVtm1btG3bFsuWLcP27duxfv16TJ8+HYMGDcK6deu43jDD6EBRgY2Pp9KJubkkfLVrk9jeukXeYsOGJLzm5iS4GzaQ+Hl6kreZk5OJJ0++wJ07S2Bi4gEvr2XIyjLB48ckrO7uJMrJyXTvunXJQ01JoeXkp0X++vgAb7wBnDpFHnFKCj2srUlQa9TQ3C8G6MuBttZz0nNzc6rZXDQaOTtbs8ctwzwryERpVSWewvHjxzF37lwcP34cjx49gqO2WP9qQGpqKuzt7ZGSksJeOmNUpKIUd+9Suo5SSQJobk75qykpNM7GhjzRmzfJE/T3JyEWAvjtN4r8tbQEEhMF/vijPZKTL6Bhw49gZjYL6elWaNWK9kVlMioSYWZGXmhuLvWYlctJZD08gJ49C1J6SkMq23j/Pi0rW1vTHuv9+2SvqSmJur09jXdzA+rV09yzFYK88pLa1UndhDjPlqkMjKkFeqf+PHjwABs3bsT69euRkZGBV199FatXr662QsswFYVUlCItjUTVwoKWZC9epKVid3cSrQsXSGik/rMAeYAPHtDPpCTA1jYCSqUt0tO90KzZF3B0rAVLyzp48ICEtHZtErsbN0jgFAoS3ORk2nNNS6N72tnp7kXKZDTe05Ps9vSkVKSLF0m4nZzIfimtR2pYUFg0uV0d87yis9hu27YN69evR2hoKHr16oVvvvkG/fr1g4mJiTHtY5hngsJFKZycaK/UyYk8TVvbgtQXgMQyLY0Ex92dhMnVlfYyzc3Tcffu5wgN/RZ+fhNhZ7ccJibdkJ1NgvroEV3P3Z08xwcPSNTq1CGRValI1JycSMjd3cm2pCTdyy0Wjoy2syMvNCaGrvH4MS1te3hQ8JW2SlLcro55HtF5GVkul6NWrVoYM2YM3N3dSxz3/vvvG8y4ioKXkZny8rTmAGlpFEFsa0sideYMiUp+Pu3LmpjQfqW0pHznDgmPtI9Zt65AePg2bN36ITIyHsPB4WPUrj0DubkWsLAoSMVJTaU91NGjaVn3wQNajn7yhJaR3d2pgYCVFdlob1+wvKxrMwBtS8GFq0ElJNDyccuWpQu3Lg0VGKYiqRLLyLVq1YJMJiu16L9MJquWYssw5UGXDjZSUYqcHErZuXlT6rpT0LpOoSChlZZ1LSwK6iAnJETip59ehbv7ALz44lLExvqqg5RyckhYc3Jo37dmTQq4cnUlOxwc6B4yGYl6ZmZBcFVKSsFSbnLy08stAiUvBUseuTT3pwknt6tjnid0FtuoqCgjmsEw1RNdO9iYmdEya2goHU9OphxVKWdVKlbh7U1LsyYmwP37qbhzZwUGDpyB+PjaGDDgGpTKuvD1JWHLzaWlZSkKuHZtWr719aU9U4WC7DI1pUpOHh4kcCkptIxtb08RwWlpmqUTSyu3KMFLwQyjH+Wujcwwzyv6dLBRKilt5vZtqtiUnEyvk/Y9pepLd+8CQgioVL/gzp0ZUKnSYG7eA5aW7aBS1UVODqUKOThQBabYWBI5KysS7uRkuk7DhrScnJdHUcxWVgWepq0tBU4pFLTsfOMGebtSgYnSyi0WxsGBlop5KZhhng6LLcOUkdKaAwAFHWzu3wcOHiRvUiajaGNJZIUoqGvs5AQ8fPgv7t59F1lZJ+Du/grc3b/Gw4fesLOj5VmAXmtuTgLp4SHt6RY0DmjQoKBnrZQiVJi8vIKSj7dv03UKh2FYW5dcbrEovBTMMLrBYsswZURbc4DCSPugJ06Qp+jsXFBlSdpDtbIisRIC/xWhiIBc/hgvvHAYZmY9kJhI4ihFJvv6FiwFP3hAgUh5eeTVZmcXlF50dibh1WabqSmJbUoKLWG7umqel4pMuLpyuzuGMRTyyjaAYaorhVNgtJGTQ/uyGRkFwU5xcSSKUjF/c3MV8vLWIyrqbWRkAG5uQ9Gs2SU4OfWAuzulz3h6Fizl+vjQv62t6ToxMbRELdUoBkgY/f1J7LXZJkUip6cXRD8XJjmZxNrBgdvdMYyhYLFlmDIiNQd4/Fj7+ZgY8jLt7Sm9JyGBPFqVio7n5YXj8uUOuHjxDeTkZMDaWonMTBkyM81ga1uwtOzlRXu6OTm09OvrS8vDtrbkmebnk/jWqEF7ta1a0fmSbJPJyLOVugSlp0tlH8lGKys6J+3DculEhik/Oi0jp6am6nxBzlNlnhe0pcCYmdHyrFSAPzUVCAujzjlZWeSFUmP3qcjOXg4Tk8awtz8GK6suajG1tCTxlMkKer4+eUL/fvKEvFsXF9prbdaMgpSk2smFA5RKq9Tk4QF06FCwxO3gQEvHHh4ktLa2BaUTCzeUZximbOgktg4ODpDpGGKYn59fLoMYpqojFWNQKskb9PWlpgEREeTBPnpE55RK4NIlEkgHB8DePh8ZGflITTVHXp4HzMy+hZnZJJiZmUIup73cnBwam5pKIpeSQoFUkhedl0dLxkolCW3fvuVLz7G3p0dKCu3RSkvHMTFcOpFhDIlOYnv06FH1v6OiojBr1iyMHTsWAQEBAICwsDBs3LgRwcHBxrGSYaoAQpAI3bpFPx8+pNSbzEwSKKlaU34+LcVmZJAXKQSQm3sGd+5MhpVVP1hYzEN+/kzk5JCQZWeTiMrl5H3m5gKRkQX7tFFRlC7UtCld092dPNDSCk9IPC09x9GR0nskQU5I4HxZhjEGenf96dGjB958802MGjVK4/iWLVvw448/4tixY4a0r0Lgco1Vl8oq6Vf0vnl55KWGhZHIxsfTkm92Ni0PJyWRt2lhQcuxCgWNy85+iCdPZiM7ey3MzV+Are33ANojI4Puo1DQeGtrukdmJgmumxsJttSrtlYtWjr29QXat396ScXyzpfzZZnnkSpRrlEiLCwMa9asKXa8devWePPNNw1iFMMAupVB1AddBSU5mfJP790riBxOSysQ1MePya60NBJDW1v6t0pV0M2H7hWD1NSmAABLy5Wws5uArCwTCFGQBysEzcXKiuYpXS8vj7xOBwfKc334kLzZgQPJ6zS0EHK+LMMYF73F1tvbG//3f/+HxYsXaxz/6aef4O3tbTDDmGcTfQRPlzKIuvI04S7c0P3MGbqPSkX3lJaHFQp6SKUQTUzoeGQkeaQqFV0nL+9fuLg0QXZ2TZiZfQ5gBMzNXWFhQfeWy0m0pS48Jib0Oiurgo45Mhml/MhkNKZhQ9q7dXRkj5NhqiN6i+23336LoUOHYv/+/Wjbti0A4OzZs7h16xZ27txpcAOZZwddPVV9yiDqIjxPE25/fxLZhASq7nTjBn0JsLEhz/bBA9orjYsje/PzSQBjY+maUuqMShUPuXwmMjM3wczsfzA1fQlmZpORm0t2SsX6JYGVlogdHOi5VB85P5/m6etL/65fn9J/pPeNYZjqh955tn379sXNmzcxYMAAJCUlISkpCQMGDMDNmzfRt29fva51/PhxDBgwAF5eXpDJZNizZ4/GeSEE5syZA09PT1haWiIwMBC3bt3SGJOUlIQxY8bAzs4ODg4OGD9+PNLT0/WdFmNkJMGLiaHlSk/PgvSS8HA6L6FrGURdfs1FhdvSkkROEu74eGD/fiqpKJfT87w8Etg7d2g/NjOTfj58SB177t6lc48eScvMucjLWwaVqj7y8v6Eicn/IT+/hzqfFigIgJKWas3MKArY1LQgwEoupwd5x3Sf9HSy7coVEmPOeWWY6kmZyjV6e3vjiy++KPfNMzIy0Lx5c7zxxhsYMmRIsfOLFy/Gd999h40bN8LPzw+fffYZevXqhWvXrsHCwgIAMGbMGMTFxeHQoUPIzc3FuHHj8Pbbb5faCpCpWPT1VHUpg6hrZaOShFsI8lYfPSIP1dubPNqICFrizc0lrzIysmDZW6kk0ZX/9xVVWv5Vqf4PwDQA7wBYCMBJQxj/+6giJ4f+bWdHAVH5+QV7wkoljTc1JaHNzqYI5Hr16PidO2Sj1OGHYZjqRZnE9u+//8YPP/yAu3fvYvv27ahRowZ+/vln+Pn5oWPHjjpfp0+fPujTp4/Wc0IILFu2DJ9++ikGDhwIANi0aRPc3d2xZ88ejBw5EtevX8eBAwfwzz//oHXr1gCAFStWoG/fvvj666/h5eVVlukxBkYfT9XWVrMMYtEi+kCBMOni5UmibG5e0I5Oih6Oi6M0nowM4NAhOle4G48UEGVuTku/+flSGg8gk8UiL+9vmJiMAPAGgPaQy1sAKPBMVSqyQaGgaOLOncmbvXmTri3VivH3lwpdkEctkwGNG9Pyuqkp2SS13pNSgnjflmGqF3ovI+/cuRO9evWCpaUlzp8/j5z/iq+mpKQYxNuViIyMRHx8PAIDA9XH7O3t0bZtW4SFhQGgyGgHBwe10AJAYGAg5HI5zpw5U+K1c3JykJqaqvFgjIe+nurTyiA+fky5prpUNpJE+/Jl4OJFSt3Zt49+pqXRGJWKgp6uXSMRfPiQBC47u8AuSWwBJYRYApWqPlSq6cjPz4JcbgFT0xbqZWCZjDxXe3vad5UqO/n6koC2aEHPTUwoGMrCgvZk/fzonBRxLC1hu7tTgJSPj+7L5wzDVC309mwXLlyINWvW4PXXX8evv/6qPt6hQwcsXLjQYIbFx8cDANwL9/7677l0Lj4+Hm5S37H/MDU1hZOTk3qMNoKDgzF//nyD2cqUjr6eqrYyiIVLDepT2UiquCQtwz5+TM+lIhTJyQXBSqmpBTm1WVkF4l+wdPwXVKrJAG5CJpsMuXw+5HJLyOXkgZqYkCibmVFOrKVlQdMAe3sS9Lw8GvfCC3RNd3d6bmlJ80pOpnZ5UqELU9OCXrT5+dwYgGGqK3qLbUREBDp37lzsuL29PZILR7lUYWbPno1p06apn6empnLakhGRPNWYGM09W4nHj4vX4NWl1ODTEIIKUUhlEMPCKNjI0pLsSUsjMXv4kMba2tK/8/ML9kaFoAflxP4AmcwVZma/QYhm6j1bSfStrOgLgeThmpuTx+riQnY/eVKQO+vrS/MwNS2YmxDAyZN0DWvr4vPRZ/mcYZiqhd5i6+Hhgdu3b8PX11fj+IkTJ1C7dm1D2QUPDw8AQEJCAjw9PdXHExIS0KJFC/WYxMREjdfl5eUhKSlJ/XptKBQKKEpa02QMTlk91aeVGnwaMTEksHI5BRqlpdF1TEwoMMrDg/ZBk5LoeWoqPaS9ViAHQnwDoB6AYZDL18HU1BpWVjKkp9N1pI9RVlZBK7169chjtbKieUper40NzcffX/s8hND/SwnDMNUDvfds33rrLXzwwQc4c+YMZDIZYmNjsXnzZkyfPh0TJ040mGF+fn7w8PDAX3/9pT6WmpqKM2fOqGsyBwQEIDk5GeHh4eoxR44cgUqlUucAM1UDyVOtWZNETyp36O1deoEKqbKRkxP91FVohaDgp/R06mIj5bnWqEHBSg4O5OFaWEjt7khos7Pp9XL5fqhUTQDMhUx2+7+qTzbIy5OpK0dJ3qiUSqRQkFg6OBTk0zo40LKwhUVB/1mpepS2udapQ2IaE1PQPi8rixsDMEx1R2/PdtasWVCpVOjRowcyMzPRuXNnKBQKTJ8+He+9955e10pPT8ft27fVzyMjI3Hx4kU4OTmhVq1amDJlChYuXIi6deuqU3+8vLwwaNAgAEDDhg3Ru3dvvPXWW1izZg1yc3MxefJkjBw5kiORDYCh6+WW11N9mp1SFx6paEVSEv379m3yqqW2d9nZ1OEmNpaWqB8/pmOpqYAQjwGMhxC/QybrDhOT3yFEo0LVoeieJibksUp9Z62t6dpSk3gbGwpqcnWlcwkJVJwiP7/0ghyGWD5nGKbqoXcjAgmlUonbt28jPT0djRo1gk0Z1raOHTuGbt26FTseFBSEDRs2QAiBuXPn4scff0RycjI6duyIVatWoV69euqxSUlJmDx5Mvbu3Qu5XI6hQ4fiu+++08sebkRQHEPXJTYWUh3ja9eA6GjyAp2c6PHgAe3BpqSQzQ8e0Fwkr1pqhUd5rirk5soB5EKl6g+5fDxksuEASBEpnxbqYCgbG7qmhQW9L+3a0XGlklJ7WrYkD1pKJ7KyIvE1NaV7d+xYei1ibgzAMBWPMbVAb7F94403sHz5ctgW+UuRkZGB9957D+vWrTOogRUBi60mJZU3lPZX9a1LbCySk4HQUGrMfv8+2WhlReLk4EA5qdnZ5IWmpJC4Sik9Ug6tXC4gxF4oldMhl++EuXlTtRcrIQSJJEBL0QpFQSUqZ2cKgKpdm7rxmJkBhw/TlxJpv9bZuaAhe34+iX3HjvSFgGGYqoMxtUDvPduNGzciKyur2PGsrCxs2rTJIEYxlcfTyhump9MyaNnWQ/S3Req2k5ZWXAAvXqTHo0fkdUq5tzIZiW9aGnm2CQkFQpuWRlHBOTlAXt5tKJX9oFQOhIlJbZiaWmmUWDQxKUjLkbxZDw+KKDYxIRGV0nGUSjouBPWdbdGi4FG/foEXW1pEcWnzZRimeqPznm1qaiqEEBBCIC0tTV0uEQDy8/Oxb9++YjmvTNVG21KlvtWejMXTlrHT0iit58kTWh5WKgta1Dk4UGSwFCAlRRibmNC1qbrTTggxGoAnFIrdsLAYCJlMhtzcwqk+9DopmtjDgzzU1FTy8nNyyDYrK/JunzyhMR4e5Em7uhafV0kRxdVl2Z5hmLKhs9g6ODhAJpNBJpNp7JlKyGQyLhRRjSjpj7uTk+HqEpcFISjy9tw5ErMaNWhftGh7vfv3aak7Lo7+LYmpVFs5O5vmJrWoMzUFZDKB7OwYCOENoD1kso8hk81Afr6VOj82P78gR1aKVJau4eJCIikEFamQCmY4O5P36+NDXi1Q0HRBlzQnQ7cTZBim6qGz2B49ehRCCHTv3h07d+6EU6ENJ3Nzc/j4+HAEcDWhtD/usbH0b0PUJS6LXbdvA6dOkffs7k6CVrMmedFS04KLF0lgpSYBklDm5BTUP5aauQMkwnl5EcjLew9CXARwF4AngLkANOsYS0KrUFAOrpTCI5ORWEpL6o6O9F41aULBUY0bAx06FDQp0DWi2NDtBBmGqZroLLZdunQBQOk5tWrVgoz/51dLnvbHXQo0evSIxKEoxiqsIH0BSEyk+/v6krgkJJBwNmxIguvoCBw9SnY8fkxLuvn59JrsbBJeafmXSEd+/ufIz/8WgDeAdQA0jZfqGQMkvGZmtATctCldKzq6wFvOyiIhjo2lVYAXX6Q82qZNC4QW0D3Nqaos2zMMY1z0zrM9cuQIbGxsMHz4cI3j27dvR2ZmJoKCggxmHGN4nvbH3cWloMpSeesS60rhLwDu7rQ0rFAUNFePj6dGAjIZpdWcO0dilpNDr1GpCpoGFGcUgMMAPgMwAzKZhUbgkbT0LImztD9rb08iWqsWPY+IoBZ85uY0vnZt8l6bNy85/1UqyFEahmwnyDBM1UVvsQ0ODsYPP/xQ7LibmxvefvttFtsqji5/3BUKWh5NSqqYwgqFvwBI+6uSjY8eAWfOUHpPZiYtE6tUZEtKCr2WuvEU5ioAAaAJgC8ArADgC0BaUi4YKQTNTapzbGtLXyhcXOhYZibZ1bkzea9NmlBAlJMT2Vfe/FdDthNkGKbqorfYRkdHw8/Pr9hxHx8fREdHG8Qoxnjo+sfdxYWWcg1ZWKGkQg2FvwDIZCRkUrWnQ4cKes6ampLdSiUVsSgusqkA5gH4DsBwAFsBNC1mg4SlJT2X0pvMzAoauzs6UsCYuTnZcvcu0KULebKG9OrL0qSBYZjqh95i6+bmhsuXLxdrRHDp0iU4l7Q2yVQZ9PnjrssyqK4Ujn5WKsm7dHSkpVp7+4IvABYWJHDR0cDp0wVt6fLzNcVVU2gFgF8AzACQBmAhgKla7Sjc0N3ensRWElgnJ/KWLS3Je5WWmNV3MULeqyHbCTIMU3XRW2xHjRqF999/H7a2tupWe6Ghofjggw8wcuRIgxvI6Iau5f0q44974ehnc3MStMREOnb8ONCgQUHKT34+Hb93r6D8olQ4QloCLrwM/N8dAEwBEAjga1AglHakfWBHR7quhQWJrKcnib2rK70HUs6xmRnlzTo5kS3GCFTiesgM8+yjt9h+/vnniIqKQo8ePWD6Xw07lUqF119/HV988YXBDWSejr4FESryj7sU/JSWRve4epWEXcpXPXeOerhaW5OQpaTQeanSkxRdDBQNEkoGsAjATAAuAK4BcC/VFklopSXZ/HzycD08SEBzcihAq2lT+gIgVY6ysiKPOD7eeIFKxmrSwDBM1aDMjQhu3ryJS5cuwdLSEk2bNoWPj4+hbaswqnNt5PLUMa6IYvdpacCBA2TnzZtkl7U1/ZSat5ubk7BGRlKRiNzcggjj4qgAbAAwC0AWgJ0Aej7VDqmOsfReODqSyNapU+AxZ2VRvmzTpsW916ws3RoIMAxTfTGmFujt2UrUq1dPayUppuIob0EEQ+7JlsSjR5QyI5MV1A9OSKBlYoWCPMmHDym/Nz6e5lIy/wJ4C8AZAGMALAZQciEVKX9WoaBlYF9fskGKtvbyAjp1ospPANn05In294QDlRiGKQ86ie20adPw+eefw9raGtOmTSt17NKlSw1iGPN0Kroggr6esBAkorm59AXg0SPyIGNjKbo4PZ3SizIyqL6xlv4W0pVAre6UALIBhALoXOJ9peISMhndt0YNWhaWUnxsbIB69ciWR49o+dbWlrxefcosMgzD6IpOYnvhwgXk/rdZdeHChRLHcVWpiqVozqwQlBeal0fLswqF4QoilKVQfno6PVxcgPPnSbRUKqr6lJlJy7KpqSSO2oU2H8BPALYBOAigFYALkHrMFkYmK+jMY2FB4mlmVlCz2NKSzllbF4hwerpmMQwOVGIYxljoJLZHjx7V+m+mcimcM5uXRx5ZUlKB2FpZFaTVlIeyFsqXhD41lZZoExJovFR7OSurcFnFopwBMAlAOIAg0P6sLbQJrYlJQRP3Ll0oncjEhIQ0Pp4EVkopevKElpTlcvpyYmGh+f5woBLDMMagzHu2TOUj5czeuFHgLTo4kEDk5pKH5u2tLVVGd8qzL2xmRkvGp09T5K8QJLApKU/LWZ0F4CsALQCcBNBe6yhJZO3syJamTYEPPqA5nz9PDQtMTAqWsZ88KRBlyaP29S2+D1sRe9kMwzxf6CS2Q4YM0fmCu3btKrMxjH5INXrPnqW90dq1SeCUShI0b28SoshIEuGyeGdl2ReW9nazsijVJyqKvOGHD8mz1U4egEwAdgBeALASwAQAJlrnLZcXiK2vL9CsGdC+PeDnR3a0akXnU1Kok1ByMr0fLi60nJ2XR9Wg6tRhr5VhGOOjk9ja29ur/y2EwO7du2Fvb4/WrVsDAMLDw5GcnKyXKDOGwdSUBMTMjDzb1FQ65u5O3p6pqe5BUlIuaVYWeYIeHsX3hVWqgsbpCgWJeOF9YWlv984dWnoODSXvV+rKo50TACaDSiv+DGBEqXZK1Z8UCqBRI1r2dXKitB3JS3VwoCXlOnWACxfI+09PJ+/WxYXEuXlz3odlGKZi0Els169fr/73zJkz8corr2DNmjUwMSGvIz8/H++++261y1F9FpCEsFkzErPC+7UyWUH1pacFSd27B/z9Ny0JK5W0v1m7NtCiRcG+cHIyCdeDBwX7ma6ugL8//Vva2719m5Zxr18nj1vqMVuceAAfgQT2RQDvPXW+5ub0pcHenqo+1atH8/f3L+6lymTkzdasCXTtSkILUPqRrS17tAzDVBx679muW7cOJ06cUAstAJiYmGDatGlo3749lixZYlADmdKRgqSUSgoEKoouXWPu3QO2bSMxqlmThDozk4r9x8VRL9nr14ErV8hzdnen5dvsbBrz6BHlrd67R0vax4+TZ5uSQiKtnVQAjUEBT/8H4A0A8pIGw8qKlsTt7Mgbzc4msa1ThzzU0rxUmazgtQzDMJWB3mKbl5eHGzduoH79+hrHb9y4AVXJoaVMKZSnklN5u8aoVOTRPnlCS7ISdnb0/No18mTv3qWfDRuS55yXR2JqZ0fC+uGHdK87d0rLlwUo4OlF0N7sjwC6AXAqdY52diTwnp4k6vb2JLZDhpBHz14qwzBVHb3Fdty4cRg/fjzu3LmDNm3aAADOnDmDL7/8EuPGjTO4gc86ZclfLUx5GwvEx5OQahNqgI7fvk3iVqdOwVK1VN4wI4Nsj48vqGWsnQcApgP4FcBGAK8DGFrq3MzMyIO2tqYvJCYmFBjl6Un7yc2bc9QwwzDVA73F9uuvv4aHhwe++eYbxMXFAQA8PT0xY8YMfPjhhwY38FmmrPmrRSlPMYasLBJIKyvt5y0tSVABoH59KkTx6BEtL+fkUGeexMSCBgLFUQJYBmABAGtQXeNXnzonqUiFoyPZ36gRpfYIQbnEzZpx6USGYaoPZW5EAFDRZgDVPjCqMhoRCEFBRCUt/8bEkFiWVNe4pGvquhwtjY2OBrZsoTrBhYLO1aSmkueblkZe7f37JLZpafT6tLSCfrPa2QNgGCjaeD4ALTfRgq0tfQHx8qKfXbqQ6CYnk/h37kzHePmYYRhDUeUaEeTl5eHYsWO4c+cORo8eDQCIjY2FnZ0dbNjd0Alj1DXWtRhD0UbuSiW1umvXTjPISqWiwhD5+bR3GxFRUHkpLa20COdoADsATAMwEMB1AHX1moO/P3nSQMF7lJFh/N6yDMMwxkBvsb137x569+6N6Oho5OTk4KWXXoKtrS2++uor5OTkYM2aNQYzbt68eZg/f77Gsfr16+PGjRsAgOzsbHz44Yf49ddfkZOTg169emHVqlVwdy+9r2lVoGj+alHMzQ1X17gw2pau27YF9u4Fjh2j5WhnZ/JgQ0Ppp1JJBSmysmjfVKoGVZwcUPP2RQAcALwGwBWlCa2JCS1Vy2TkjTdpAnTvTve0tKRlZF/fiu0tyzAMY2j0FtsPPvgArVu3xqVLl+BcyC0bPHgw3nrrLYMaBwCNGzfG4cOH1c+lhvUAMHXqVPz555/Yvn077O3tMXnyZAwZMgQnT540uB2GpnBdY0vL4ud1SdnRF22lF4WgXNkuXYATJ6jik4kJ8O+/tC8rl9N+rBRhXHLpx4OgpeIoAB8AmAOKONaOiQlFGTs6Ukcer/865bVrR8vfEREFqT1FvVdjvDcMwzDGRG+x/fvvv3Hq1CmYm5trHPf19cWDBw8MZpiEqakpPDw8ih1PSUnB2rVrsWXLFnTv3h0AFd9o2LAhTp8+jXbt2hncFkNS3pSdslB06TotjURNKmeYlUU/U1IozScnh7zHjIyn1TIGqO1dLQC/A2j0lLHkudesSXuvL71EUcd5ebRHrVCQ0CYlcW9ZhmGeDUquIlACKpUK+VqiYWJiYmBrhA20W7duwcvLC7Vr18aYMWMQHR0NgEpE5ubmIjAwUD22QYMGqFWrFsLCwkq9Zk5ODlJTUzUeFY2UsmNjQ4KblUXLs1lZ9Lw8/VOlPdWkJPopCWXhpeu0NFpOvnCBCljY29O42FgqYCG9NiWlJG82CxTwJBUxmQ/gMHQRWgcHiix+5RVg6FASXalIhacnfRmQ8mcN/d4wDMNUBnqLbc+ePbFs2TL1c5lMhvT0dMydOxd9+/Y1pG1o27YtNmzYgAMHDmD16tWIjIxEp06dkJaWhvj4eJibm8OhSF6Lu7s74uPjS71ucHAw7O3t1Q9vb2+D2q0rUspOzZokbPHx9NPbW/e0n6IkJ1OU84kTBY/z5+m4qSmJbUICcPMm7cdaWlLBiMePSWjj4wu8XO3VnwSAP0DVnxaBRBcAzKCt/V1RnJwojWfwYCAggJaCtQmoMd4bhmGYykLv1J/79++jd+/eEELg1q1baN26NW7dugUXFxccP34cbm5uxrIVycnJ8PHxwdKlS2FpaYlx48Yhp4gitGnTBt26dcNXX31V4nVycnI0Xpeamgpvb+8KTf0pTGkpO/qk85SUt/v4MZ23s6O92Oho8lyzskjMhAD++YcE+OHD0tJ4MkFpPPsB9ALwHYB6Os3RzIz2Z0eMADp2pHnk5dFxd/eS84HLU12LYZ4HcnJyMHHiRBw+fBjJyclo1KgRvv32WwQEBFS2adWOKpX64+3tjUuXLuG3337DpUuXkJ6ejvHjx2PMmDGw1BbpY0AcHBxQr1493L59Gy+99BKUSiWSk5M1vNuEhASte7yFUSgUUJQUBlwJlJSyo091qdL6ztrbU71ia2sKRkpOpqXijIyCPdmoKKoApV1oswFYALAE4A1gNyil5+mqZ2VF+7M1agCdOgGffUZdd3QVUO4tyzClk5eXB19fX5w4cQI1a9bEtm3bMGDAAERFRXEqZhVCL882NzcXDRo0QEhICBo2bGhMu7SSnp6OWrVqYd68eQgKCoKrqyu2bt2KoUOp7F9ERAQaNGiAsLAwvQKkKqOoxdMozUu1sSm+lJqWRkvGtraa0c1CUHu56OgCcbt5k/ZqHz6kyF6VivZtiy8bCwC7AEwFsApAf51sl3rN2tgU7MF27Ur7tL17s3gyjLHx8vLC3r170apVq8o2pVphTC3Qa8/WzMwM2SU3JTU406dPR2hoKKKionDq1CkMHjwYJiYmGDVqFOzt7TF+/HhMmzYNR48eRXh4OMaNG4eAgIAqH4lcGkJQ1abz50kMa9Qg8ZTL6WfNmiTAd++SSEqBUElJJJyFHXYhqNrTzZvUkefoUSAkhApVxMXR9R8/Jo+2uNDeAC0VDwPQHIBuX64sLCiVqFYt+kLQsiVFG9erVxAQxjDM0zl48CBkMpn6YWZmBn9/f8ybNw/KkouQ49atW0hKSoK/v7/RbczJycHMmTPh5eUFS0tLtG3bFocOHdLpteHh4ejduzfs7Oxga2uLnj174uLFi1rH3rp1CyNHjkTNmjVhZWWFBg0aYMGCBcjMzDTgbIyL3svIkyZNwldffYWffvpJI+fVGMTExGDUqFF4/PgxXF1d0bFjR5w+fRqurq4AgG+//RZyuRxDhw7VKGpRXZGWjaOigMuXyQPMyyOBLewNOjsXLBlnZJDIpqVRwFNeHuDjQ+diYmiP9vhxEuP0dDqfnFxawwCAoor7gpaM90JXj9bCgsTU3Z0audeqRfu0NWtS5SeOIGYY3bl06RIAYOnSpXB1dUVmZia2b9+O+fPnIycnB8HBwcVek5WVhVdffRWzZ8+Gvbb6qwZm7Nix2LFjB6ZMmYK6detiw4YN6Nu3L44ePYqOHTuW+Lrz58+jY8eO8Pb2xty5c6FSqbBq1Sp06dIFZ8+e1egqd//+fbRp00ZdS8HJyQlhYWGYO3cuwsPD8fvvvxt9noZA7wCpwYMH46+//oKNjQ2aNm0K6yJNVHft2mVQAyuCil5G1hb0k5JSsGxsakp7qvb25OVaWpKASlWUlEogLAzw86Po3vv3KVr39m3ac23alERVqaQ82ps3ycOVlo1LsArAFQBNQRHGawBMBO3VPh07OypM4eJCZRbd3MjDrVmTPNqnNURgGEaTV199FTt37kR6erq6f7hSqYSPjw8sLS1x9+5djfG5ubkYPHgw7O3t8csvv0Bm5G+2Z8+eRdu2bbFkyRJMnz4dAFX1a9KkCdzc3HDq1KkSX9uvXz+EhYXh1q1b6uJIcXFxqFevHnr27ImdO3eqx37xxRf45JNPcOXKFTRu3Fh9PCgoCJs2bUJSUhIcHR0NMqcqFSDl4OCg3iNlCH0jhosGPbm60uul4KaMDAoqksvpWjduAJGR5B2ampJwZmfTmOPHaSlYoaDHtWtUBcrRkYKT7t+neyYklFb96SqA90C9ZiMBeIH2aZ+OpSUJ68svk6A2aEDLx1I+r7k5RxAzTFm4dOkSGjdurBZaADA3N4eXlxeioqI0xqpUKrz22muQyWTYuHGj0YUWAHbs2AETExO8/fbb6mMWFhYYP348Pv74Y9y/f7/EtMq///4bvXv31qhC6OnpiS5duiAkJATp6enq4C6pDkLRMryenp6Qy+XFCixVVfQW2/Xr1xvDjmqLPhHD2oKesrNpqffuXaBFCxJuKyvyWKOiCiKGhaBo4owMGu/qCpw8SZ6vtI8rLS0nJZEXm5FB1y/Zm00FMA+UwlMHlD/rpfPcPTyoYUCPHtRb1s6Oc2AZxhAolUpERETgtdde0zgeGxuLa9euoUuXLhrHJ0yYgLi4OBw8ePCp23u5ublISUnRyQ4nJyfI5dpDey5cuIB69eoV8wClPucXL14sUWxzcnK0Zq9YWVlBqVTiypUr6tibrl274quvvsL48eMxf/58ODs749SpU1i9ejXef//9YqurVRWdxValUmHJkiX4448/oFQq0aNHD8ydO9fo6T5VGX360WpLzUlLo7Hx8eS9ZmVRVLC3NwVGXb1K57y8SEBv3SroBCSl6bRrRx7rrVt0TxsbEunMzNLyZSWmAdgKYCHIk9U9HcrXF2jThr5UNG5My9y8VMwwhuHatWvIzc2Fn58fHj16hNzcXFy+fBkzZ86EiYkJFi5cqB577949/PTTT7CwsICLi4v6+P79+9GpU6di1z558iS6deumkx2RkZHw9fXVei4uLg6enp7FjkvHYmNjS7xu/fr1cfr0aeTn52sskZ85cwYANEr/9u7dG59//jm++OIL/PHHH+rjn3zyicb7UNXRWWwXLVqEefPmITAwEJaWlli+fDkSExOxbt06Y9pXZSktr7VmTRLRu3cL+tFKdYmdnMjjlDxilYqCnzw9SSCjo2mZ2MGB7uHiQs/v3yfP0cyMlpKlGsb//kv3jI0lcY2OJhEvmUsAHgPoDiqxOBcUCKUb1tZUAeqVV0hsvb3JRl4qZhjDcfnyZQDAZ599hs8++0x9vGvXrjhx4gRatGihPubj4wN9Qm+aN2+uc8RwaTULsrKytNYrsLCwUJ8viXfffRcTJ07E+PHj8dFHH0GlUmHhwoWIi4vT+lpfX1907twZQ4cOhbOzM/7880988cUX8PDwwOTJk3WaS2Wjs9hu2rQJq1atwoQJEwAAhw8fRr9+/fDTTz+VuMzwLKNvP9rcXBLIhw/JK42KIlH08SEBc3KiACeFgvZnlUp6jZsbibOjI/1b+j+VkkKvi48vqF+cmUn30E4ygM9A+bKBILGtofN8ra0pfefll4FBgwrSeFhgGcbwSJHIf/75J8zNzZGQkIDg4GCEh4eXO8rY0dFRo6Z8WbG0tCxWwQ+AOj20tFXPd955B/fv38eSJUuwceNGAEDr1q3x0UcfYdGiRRrFOH799Ve8/fbbuHnzJmr+59kMGTIEKpUKM2fOxKhRozT2fqsqOottdHS0Ru3jwMBAyGQyxMbGqt+A5wl9+9FKVZqEoChjuZwEWcpzNTMj8Y6NJWGzsCAPNiaGRNnTk465uZHXq1CQsKal0TVyc0taNlYB2ABgFqgS1BJQMJRuODgA/fpRUYpevWh5+zn8bsUwFcrly5fh4+Oj8Te3ZcuWaNSoEVatWoUlS5aU8urSUSqVSEpK0mmsq6urRoBWYTw9PbV2epO8Uy+v0uM/Fi1ahOnTp+Pq1auwt7dH06ZN8fHHHwMA6tUrKAO7atUqvPDCC8V05uWXX8aGDRtw4cIFg3x5MDY6i21eXp56eUDCzMwMuc9pB299+tEKQV6uFDRnYkLLxzIZBTAlJJBgqlQktJaWJKJ5efTIzaU93UePaMlWul5iIgmuSlWapXkgge3538/ieywlYWcHTJ5MQsuBTwxTcVy+fFkdaCTRsGFDtG7dGjt37iyX2J46dcoge7YtWrTA0aNHkZqaqhEkJe27Fl7qLglHR0eNfNzDhw+jZs2aaNCggfpYQkKC1tQeSXvySk6zqFLoLLZCCIwdO1ZjjT47OxvvvPOORjRYdcyzLQv69KNNT6fl40aNyLt9/JiCoaR0HCkn1smJxDY3l7xHT0+KNpYKUmRmkuebnU3HS44yfgzgU1Az98YAzgLQr0aikxMwfHjBkjELLcNUDPHx8UhMTESTJk2KnevVqxcWLVqE69evl7lkrqH2bIcNG4avv/4aP/74ozrPNicnB+vXr0fbtm3VkciZmZmIjo6Gi4uLRgBXUX777Tf8888/+PrrrzW2JuvVq4f//e9/uHnzpobHu3XrVsjlcjRr1kynuVQ2OottUFBQsWOvvvqqQY2pTkj9aJOTSXCdnclzVSoL6hdLFZOk5WRPT/Jao6PJU01IoEhjS0sSX0tLErlr10h0a9emohS5uSS2MhktFZdcMTMfwE8APgZ5tD1BYqu70NrYUODTsGFA//70RYL3ZRmm4pD2a5s2bVrsXM+ePbFo0SL8+eefZRZbQ+3Ztm3bFsOHD8fs2bORmJgIf39/bNy4EVFRUVi7dq163NmzZ9GtWzfMnTsX8+bNAwAcP34cCxYsQM+ePeHs7IzTp09j/fr16N27Nz744AON+8yYMUMdWT158mQ4OzsjJCQE+/fvx5tvvvnU5eqqgs5iy/m1xZF6rhbNs/X21kyDKbzkbGtLQVH+/rRMC5BIJyWRkCYm0lhra1o2josjz1eIAg9YO9cBvAYgHMBYAF8CcC9psAYyGd2vRQugTx+qZVy3LnuzDFMZSJHI2jzbgIAA2NraYt++fWpvsjLZtGkTPvvsM/z888948uQJmjVrhpCQEHTu3LnU19WoUQMmJiZYsmQJ0tLS4Ofnh4ULF2LatGnF8oQ7d+6MU6dOYd68eVi1ahUeP34MPz8/LFq0CB999JExp2dQ9C7X+CxS3hJdJVWQko4rlZQzm5REQpySQl135HIS15gY8lYzMigoSlpaTk2lBgKl95hVgfpJxAIYBSAYQHud7DYxoWCthg1J/Nu1o2IZ0pcBFluGYZ4nqlS5RqY42nquFq0slZNTUNXJwoL2XrOyaFxqKnm3WVnkySYmkkBL6T/ayQPVL14J4DSo8lOozjZLLfCaNyfv3M2NSi2amgIPHtAXAg6KYhiGMQwstkagpMpSSiXlxSYnU7pPSgpFHaelFeTtpqWR6JbOCVDw02UAb4K8W92R9oZdXUlQPTzoIXnk2opyMAzDMGWHMyYNSOFetAkJ5BVKDd9zckjEVCpaPm7TpiDtx8KCBDYlpbTgJ4kFADoBMAdwBsCPAHTreGFlVdAsICCAcmbNzel4jRqaolq4KAfDMAxTPtizNRCFe9GePUuiqVLR3qtUPtHCgjxahYKijG/donEPH5JXW/K+bC6AR6Ac2R6gJeM3oOt3JRMTCoBq0IBskDoKKRS0P1uvXvFl8KJFORiGYZiyw2JrAAovG+fn0/P8fApuevSooJRiZiZ5uCYmJGKpqQXlF0vmGGjJ2BHAcQAd/nvohkJBy8N2dnQvSWDlchLeGjWKCy2gWZSDYRiGKR8stuWkcEOCGjUoylilIm9WqlsstdJLTtZlmVjiAYDpAH4FEABqg6f75qmTE9mjUpGo2tqScNaoQcvZ9vZkc3g40LlzccEtXJSDYRiGKR8stuWkcEMCyXt1cADOnSNhlctJdPXb+8wFCWwOqK7xa9BnydjVlXJlGzQouP+TJ9TMICODxjRvTnu0x4/THvOLL9KXAm1FORiGYZjywWJbTgo3JEhNJVHNziaPNiuroEmAbhwG8AIAZwCbATQF4KCzLaam5LnWqkW5s+3/S7e9eJHyZk1NaXk7L69gH/fFF2nvWCqmoa0oB8MwDFM+WGzLiSRQDx8Cly4BR49S8YpHj57WV7Yw0aBG7jsBLAYwAxRxrDsmJhRV7OhInq1cTl5pXh4JrKMjHcvPpy8AUu1uZ2fyZlu0KFhq5tZ5DMMwhoXFtpzY2JBw/fQTNXJPTqal2pLLKhYmG8A3ABaBPNjNoCpQ+iEJK0D3dXYuiIQ2NaWHUkl7t4WPSePNzWmPV1ugFMMwDFN+WGzLSXIysGEDsH9/WdJkroLyZt8DMAdA2cqDyWQFbf7kcoo2zsoiQbWyIiGV8nmTk6mAhZUVjedAKIZhGOPDRS3KgRDAxx8Du3frI7SRoCjjfACtANwD8DXKKrQmJvSQqkJJjeU9PWk5Ozub/m1qSt2ETExIbLOzqUoUB0IxDMMYH/Zsy8G+feTV6tbKIQvAV/89XABMBFAHQMn9IkvDwqJgKdjBgWobW1nRkrCbG9ClC+0bS7WZvbwKSkdKy9wcCMUwDFMxsNiWkbw84JtvdM2bPQDgXQAxIK/2YwBlX7eVy8mLtbSkf9vYFAQ2eXpShLGPD0UlF+5GZG1NQlu0OxHDMAxjXJ6ZZeSVK1fC19cXFhYWaNu2Lc6ePWvU+0mRx6UjubyxAOoBuALgC5RHaE1NqSCFgwM9/Pyo3GL79tTMvnt3oH59Git1I5KCn+RyzecstAzDMBXDM+HZ/vbbb5g2bRrWrFmDtm3bYtmyZejVqxciIiLg5uZmlHu+8UZpZzNAfWUfgdrgjfvvUXZ1k8sBd3egb18KcpL2Wq2sSDQfPSKvtlMnGsswDMNUHZ6J5vFt27bFiy++iO+//x4AoFKp4O3tjffeew+zZs166uvL0jBYu1coAOwCMBVAIoBZAOaiPCILkLDWqQMMGwY0aULe7ZMnQGRkQepOnTpAx460fMwwDMPoDzePLwWlUonw8HDMnj1bfUwulyMwMBBhYWFaX5OTk4OcnBz189TUVANYkg+gP2h/tj+AZaAAqLIjk1FgU+vWtA/r60vLv61aUWOB+HhK8bG0pAhj9mgZhmGqJtVebB89eoT8/Hy4u7trHHd3d8eNGze0viY4OBjz5883sCUmALoAmAQS27Ijk5E327AhCW2dOhQ5XKeOZvSwl1c5TWYYhmEqhGovtmVh9uzZmDZtmvp5amoqvL29DXDlpy9Zl4apKaXt+PkBo0cDvXqR16pQ0FIxRw8zDMNUT6q92Lq4uMDExAQJCQkaxxMSEuDhoT2HVaFQQKFQlOu+q1YB775brkto4OVFwU3t2gGBgUDjxiysDMMwzwrVXmzNzc3RqlUr/PXXXxg0aBAACpD666+/MHnyZKPdd+LE8omtTEbRxU2bAj17At26UTQx770yDMM8e1R7sQWAadOmISgoCK1bt0abNm2wbNkyZGRkYNy4cUa9rxD6eZ8uLsArr5C4OjvT3qu3NwU7sRfLMAzz7PJMiO2IESPw8OFDzJkzB/Hx8WjRogUOHDhQLGjKGAgBrF5d3Mvt25fSdDw9qVpTrVpAhw7U6o6FlWEY5vnimcizLS/GzK1iGIZhqgfG1ALeHWQYhmEYI8NiyzAMwzBGhsWWYRiGYYzMMxEgVV6kbWvDlG1kGIZhqiOSBhgjlInFFkBaWhoAGKiKFMMwDFOdSUtLg729vUGvydHIoCIYsbGxsLW1hayMeTlSycf79+8/VxHNz+O8n8c5A8/nvJ/HOQPP57ylOV+7dg3169eH3MDVhdizBXUJqlmzpkGuZWdn99x8OAvzPM77eZwz8HzO+3mcM/B8zrtGjRoGF1qAA6QYhmEYxuiw2DIMwzCMkWGxNRAKhQJz584tdzeh6sbzOO/ncc7A8znv53HOwPM5b2PPmQOkGIZhGMbIsGfLMAzDMEaGxZZhGIZhjAyLLcMwDMMYGRZbhmEYhjEyLLYGYuXKlfD19YWFhQXatm2Ls2fPVrZJBmPevHmQyWQajwYNGqjPZ2dnY9KkSXB2doaNjQ2GDh2KhISESrS4bBw/fhwDBgyAl5cXZDIZ9uzZo3FeCIE5c+bA09MTlpaWCAwMxK1btzTGJCUlYcyYMbCzs4ODgwPGjx+P9PT0CpyFfjxtzmPHji32u+/du7fGmOo25+DgYLz44ouwtbWFm5sbBg0ahIiICI0xunymo6Oj0a9fP1hZWcHNzQ0zZsxAXl5eRU5FL3SZd9euXYv9vt955x2NMdVp3qtXr0azZs3UxTkCAgKwf/9+9fmK/D2z2BqA3377DdOmTcPcuXNx/vx5NG/eHL169UJiYmJlm2YwGjdujLi4OPXjxIkT6nNTp07F3r17sX37doSGhiI2NhZDhgypRGvLRkZGBpo3b46VK1dqPb948WJ89913WLNmDc6cOQNra2v06tUL2dnZ6jFjxozB1atXcejQIYSEhOD48eN4++23K2oKevO0OQNA7969NX73W7du1Thf3eYcGhqKSZMm4fTp0zh06BByc3PRs2dPZGRkqMc87TOdn5+Pfv36QalU4tSpU9i4cSM2bNiAOXPmVMaUdEKXeQPAW2+9pfH7Xrx4sfpcdZt3zZo18eWXXyI8PBznzp1D9+7dMXDgQFy9ehVABf+eBVNu2rRpIyZNmqR+np+fL7y8vERwcHAlWmU45s6dK5o3b671XHJysjAzMxPbt29XH7t+/boAIMLCwirIQsMDQOzevVv9XKVSCQ8PD7FkyRL1seTkZKFQKMTWrVuFEEJcu3ZNABD//POPesz+/fuFTCYTDx48qDDby0rROQshRFBQkBg4cGCJr6nucxZCiMTERAFAhIaGCiF0+0zv27dPyOVyER8frx6zevVqYWdnJ3Jycip2AmWk6LyFEKJLly7igw8+KPE1z8K8HR0dxU8//VThv2f2bMuJUqlEeHg4AgMD1cfkcjkCAwMRFhZWiZYZllu3bsHLywu1a9fGmDFjEB0dDQAIDw9Hbm6uxvwbNGiAWrVqPVPzj4yMRHx8vMY87e3t0bZtW/U8w8LC4ODggNatW6vHBAYGQi6X48yZMxVus6E4duwY3NzcUL9+fUycOBGPHz9Wn3sW5pySkgIAcHJyAqDbZzosLAxNmzaFu7u7ekyvXr2Qmpqq9pqqOkXnLbF582a4uLigSZMmmD17NjIzM9XnqvO88/Pz8euvvyIjIwMBAQEV/nvmRgTl5NGjR8jPz9f4ZQCAu7s7bty4UUlWGZa2bdtiw4YNqF+/PuLi4jB//nx06tQJV65cQXx8PMzNzeHg4KDxGnd3d8THx1eOwUZAmou237N0Lj4+Hm5ubhrnTU1N4eTkVG3fi969e2PIkCHw8/PDnTt38PHHH6NPnz4ICwuDiYlJtZ+zSqXClClT0KFDBzRp0gQAdPpMx8fHa/0sSOeqOtrmDQCjR4+Gj48PvLy8cPnyZcycORMRERHYtWsXgOo573///RcBAQHIzs6GjY0Ndu/ejUaNGuHixYsV+ntmsWWeSp8+fdT/btasGdq2bQsfHx9s27YNlpaWlWgZY2xGjhyp/nfTpk3RrFkz1KlTB8eOHUOPHj0q0TLDMGnSJFy5ckUjBuF5oKR5F95rb9q0KTw9PdGjRw/cuXMHderUqWgzDUL9+vVx8eJFpKSkYMeOHQgKCkJoaGiF28HLyOXExcUFJiYmxSLYEhIS4OHhUUlWGRcHBwfUq1cPt2/fhoeHB5RKJZKTkzXGPGvzl+ZS2u/Zw8OjWFBcXl4ekpKSnpn3onbt2nBxccHt27cBVO85T548GSEhITh69KhGi01dPtMeHh5aPwvSuapMSfPWRtu2bQFA4/dd3eZtbm4Of39/tGrVCsHBwWjevDmWL19e4b9nFttyYm5ujlatWuGvv/5SH1OpVPjrr78QEBBQiZYZj/T0dNy5cweenp5o1aoVzMzMNOYfERGB6OjoZ2r+fn5+8PDw0Jhnamoqzpw5o55nQEAAkpOTER4erh5z5MgRqFQq9R+t6k5MTAweP34MT09PANVzzkIITJ48Gbt378aRI0fg5+encV6Xz3RAQAD+/fdfjS8ahw4dgp2dHRo1alQxE9GTp81bGxcvXgQAjd93dZt3UVQqFXJycir+92yI6K7nnV9//VUoFAqxYcMGce3aNfH2228LBwcHjQi26syHH34ojh07JiIjI8XJkydFYGCgcHFxEYmJiUIIId555x1Rq1YtceTIEXHu3DkREBAgAgICKtlq/UlLSxMXLlwQFy5cEADE0qVLxYULF8S9e/eEEEJ8+eWXwsHBQfz+++/i8uXLYuDAgcLPz09kZWWpr9G7d2/xwgsviDNnzogTJ06IunXrilGjRlXWlJ5KaXNOS0sT06dPF2FhYSIyMlIcPnxYtGzZUtStW1dkZ2err1Hd5jxx4kRhb28vjh07JuLi4tSPzMxM9Zinfabz8vJEkyZNRM+ePcXFixfFgQMHhKurq5g9e3ZlTEknnjbv27dviwULFohz586JyMhI8fvvv4vatWuLzp07q69R3eY9a9YsERoaKiIjI8Xly5fFrFmzhEwmE//73/+EEBX7e2axNRArVqwQtWrVEubm5qJNmzbi9OnTlW2SwRgxYoTw9PQU5ubmokaNGmLEiBHi9u3b6vNZWVni3XffFY6OjsLKykoMHjxYxMXFVaLFZePo0aMCQLFHUFCQEILSfz777DPh7u4uFAqF6NGjh4iIiNC4xuPHj8WoUaOEjY2NsLOzE+PGjRNpaWmVMBvdKG3OmZmZomfPnsLV1VWYmZkJHx8f8dZbbxX7Elnd5qxtvgDE+vXr1WN0+UxHRUWJPn36CEtLS+Hi4iI+/PBDkZubW8Gz0Z2nzTs6Olp07txZODk5CYVCIfz9/cWMGTNESkqKxnWq07zfeOMN4ePjI8zNzYWrq6vo0aOHWmiFqNjfM7fYYxiGYRgjw3u2DMMwDGNkWGwZhmEYxsiw2DIMwzCMkWGxZRiGYRgjw2LLMAzDMEaGxZZhGIZhjAyLLcMwDMMYGRZbhmEYhjEyLLYMU4WQyWTYs2dPZZuBDRs2FGs9VlWpTrYyzy8stsxzidSPtV+/fnq/1tfXF8uWLTO8UTowduxYyGQyvPPOO8XOTZo0CTKZDGPHjq14w4pw7NgxyGSyYh1VCrNz506YmJjgwYMHWs/XrVsX06ZNM5KFDFOxsNgyzyVr167Fe++9h+PHjyM2NrayzdELb29v/Prrr8jKylIfy87OxpYtW1CrVq1yXz83N7fc19CFl19+Gc7Ozti4cWOxc8ePH8ft27cxfvz4CrGFYYwNiy3z3JGeno7ffvsNEydORL9+/bBhw4ZiY/bu3YsXX3wRFhYWcHFxweDBgwEAXbt2xb179zB16lTIZDLIZDIAwLx589CiRQuNayxbtgy+vr7q5//88w9eeukluLi4wN7eHl26dMH58+f1tr9ly5bw9vbGrl271Md27dqFWrVq4YUXXtAYe+DAAXTs2BEODg5wdnZG//79cefOHfX5qKgoyGQy/Pbbb+jSpQssLCywefPmYvd8+PAhWrdujcGDByMnJwcqlQrBwcHw8/ODpaUlmjdvjh07dqiv2a1bNwCAo6Njid62mZkZXnvtNa3v/7p169C2bVs0btwYS5cuRdOmTWFtbQ1vb2+8++67SE9PL/H9GTt2LAYNGqRxbMqUKejatav6eWn2A8CTJ08wZswYuLq6wtLSEnXr1sX69etLvCfDPA0WW+a5Y9u2bWjQoAHq16+PV199FevWrUPhfhx//vknBg8ejL59++LChQv466+/0KZNGwAkajVr1sSCBQsQFxeHuLg4ne+blpaGoKAgnDhxAqdPn0bdunXRt29fpKWl6T2HN954Q+OP/7p16zBu3Lhi4zIyMjBt2jScO3cOf/31F+RyOQYPHgyVSqUxbtasWfjggw9w/fp19OrVS+Pc/fv30alTJzRp0gQ7duyAQqFAcHAwNm3ahDVr1uDq1auYOnUqXn31VYSGhsLb2xs7d+4EQP1B4+LisHz5cq3zGD9+PG7duoXjx4+rj6Wnp2PHjh1qr1Yul+O7777D1atXsXHjRhw5cgQfffSR3u9ZYUqzHwA+++wzXLt2Dfv378f169exevVquLi4lOuezHNO+RoYMUz1o3379mLZsmVCCCFyc3OFi4uLOHr0qPp8QECAGDNmTImv9/HxEd9++63Gsblz54rmzZtrHPv222+Fj49PidfJz88Xtra2Yu/evepjAMTu3btLfE1QUJAYOHCgSExMFAqFQkRFRYmoqChhYWEhHj58KAYOHKhuCaiNhw8fCgDi33//FUIIERkZKQCo3w+J9evXC3t7e3Hjxg3h7e0t3n//faFSqYQQQmRnZwsrKytx6tQpjdeMHz9e3cdWat335MmTEm2RaNeunYbNa9euFVZWViI1NVXr+O3btwtnZ+ditkpI71FhPvjgA9GlSxed7R8wYIAYN27cU21nGF1hz5Z5roiIiMDZs2cxatQoAICpqSlGjBiBtWvXqsdcvHgRPXr0MPi9ExIS8NZbb6Fu3bqwt7eHnZ0d0tPTER0drfe1XF1d1Uvg69evR79+/bR6Xrdu3cKoUaNQu3Zt2NnZqZe1i96zdevWxV6blZWFTp06YciQIVi+fLl6yfz27dvIzMzESy+9BBsbG/Vj06ZNGkvUuvLGG29gx44dag9/3bp1GD58OGxtbQEAhw8fRo8ePVCjRg3Y2tritddew+PHj5GZman3vXS1f+LEifj111/RokULfPTRRzh16lSZ7sUwEqaVbQDDVCRr165FXl4evLy81MeEEFAoFPj+++9hb28PS0tLva8rl8s1lqKB4oFGQUFBePz4MZYvXw4fHx8oFAoEBARAqVSWaS5vvPEGJk+eDABYuXKl1jEDBgyAj48P/u///g9eXl5QqVRo0qRJsXtaW1sXe61CoUBgYCBCQkIwY8YM1KhRAwDU+6V//vmn+ljh1+jLyJEjMXXqVGzbtg2dO3fGyZMnERwcDID2f/v374+JEydi0aJFcHJywokTJzB+/HgolUpYWVkVu97Tfhe62N+nTx/cu3cP+/btw6FDh9CjRw9MmjQJX3/9td7zYxiAxZZ5jsjLy8OmTZvwzTffoGfPnhrnBg0ahK1bt+Kdd95Bs2bN8Ndff2ndAwUAc3Nz5OfnaxxzdXVFfHw8hBBqD/DixYsaY06ePIlVq1ahb9++AGgv9NGjR2WeT+/evaFUKiGTyYrtswLA48ePERERgf/7v/9Dp06dAAAnTpzQ+fpyuRw///wzRo8ejW7duuHYsWPw8vJCo0aNoFAoEB0djS5dumh9rbm5OQAUe5+0YWtri+HDh2PdunW4c+cO6tWrp7Y3PDwcKpUK33zzDeRyWojbtm1bqddzdXXFlStXNI5dvHgRZmZmAKCT/dJ1goKCEBQUhE6dOmHGjBkstkyZYbFlnhtCQkLw5MkTjB8/Hvb29hrnhg4dirVr1+Kdd97B3Llz0aNHD9SpUwcjR45EXl4e9u3bh5kzZwKgPNvjx49j5MiRUCgUcHFxQdeuXfHw4UMsXrwYw4YNw4EDB7B//37Y2dmp71G3bl38/PPPaN26NVJTUzFjxowyedESJiYmuH79uvrfRXF0dISzszN+/PFHeHp6Ijo6GrNmzdL7Hps3b8aoUaPQvXt3HDt2DB4eHpg+fTqmTp0KlUqFjh07IiUlBSdPnoSdnR2CgoLg4+MDmUyGkJAQ9O3bF5aWlrCxsSnxPuPHj0enTp1w/fp19fsMAP7+/sjNzcWKFSswYMAAnDx5EmvWrCnV5u7du2PJkiXYtGkTAgIC8Msvv+DKlSvqSG1bW9un2j9nzhy0atUKjRs3Rk5ODkJCQtCwYUO93juG0aByt4wZpuLo37+/6Nu3r9ZzZ86cEQDEpUuXhBBC7Ny5U7Ro0UKYm5sLFxcXMWTIEPXYsLAw0axZM6FQKETh/0KrV68W3t7ewtraWrz++uti0aJFGgFS58+fF61btxYWFhaibt26Yvv27cWCraBjgFRJFA2QOnTokGjYsKFQKBSiWbNm4tixYxr3kAKkLly4oHGdokFHubm5YsiQIaJhw4YiISFBqFQqsWzZMlG/fn1hZmYmXF1dRa9evURoaKj6NQsWLBAeHh5CJpOVGrQlUb9+fWFiYiJiY2M1ji9dulR4enoKS0tL0atXL7Fp0yaN4KuitgohxJw5c4S7u7uwt7cXU6dOFZMnT1YHSAkhnmr/559/Lho2bCgsLS2Fk5OTGDhwoLh79+5T58AwJSETosjmBsMwDMMwBoWjkRmGYRjGyLDYMgzDMIyRYbFlGIZhGCPDYsswDMMwRobFlmEYhmGMDIstwzAMwxgZFluGYRiGMTIstgzDMAxjZFhsGYZhGMbIsNgyDMMwjJFhsWUYhmEYI/P/JPCnzp8CdO0AAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 500x300 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "from sklearn.metrics import r2_score\n",
    "\n",
    "# Generate predictions for the test set\n",
    "y_pred = model.predict(X_test)\n",
    "\n",
    "# Calculate the R^2 value\n",
    "r_squared = r2_score(y_test, y_pred)\n",
    "\n",
    "# Visualizing the predictions against the actual Black-Scholes values\n",
    "plt.figure(figsize=(5, 3))\n",
    "plt.scatter(y_test, y_pred, c='b', alpha=0.2)\n",
    "plt.title('Predicted vs. Actual Options Market Values')\n",
    "plt.xlabel('Actual Market Values')\n",
    "plt.ylabel('Predicted Market Values')\n",
    "plt.plot([y_test.min(), y_test.max()], [y_test.min(), y_test.max()], 'k--', lw=1)  # Diagonal line for reference\n",
    "\n",
    "# Adding the R^2 value to the graph\n",
    "plt.text(0.75, 0.125, f'$R^2 = {r_squared:.2f}$', transform=plt.gca().transAxes, fontsize=12)\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_pred_flat = y_pred.flatten()\n",
    "\n",
    "X_test_df = X_test.copy()\n",
    "\n",
    "# Add y_test and y_pred_flat to the DataFrame\n",
    "X_test_df['y_test'] = y_test  # Reset index to ensure alignment\n",
    "X_test_df['y_pred'] = y_pred_flat  # Convert y_pred_flat to a Series for alignment\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_test_df.sort_values(by='y_test').sort_values(by='y_pred')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.save('neural_network_model.keras')\n",
    "model.save_weights('neural_network_model.weights.h5')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "QUANTTVenv",
   "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": "3.11.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}