03-DilatedCNN.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Dilated CNN\n",
    "Notebook for final predictions for February 2021"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "%matplotlib inline\n",
    "import datetime\n",
    "import os\n",
    "\n",
    "from sklearn.preprocessing import MinMaxScaler, RobustScaler, StandardScaler\n",
    "import tensorflow as tf\n",
    "from tqdm import tqdm\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Reading in Preprocessed datasets\n",
    "Concatting the the preprocessed datasets. I chose a work around by deleting the date column while concatting the individual dataframes AND then at the end set the index to the sample date index."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Work around for concatting muiltiple region datasets\n",
    "# Read the one of the dataframe sfor the Date columns\n",
    "df = pd.read_csv('data_regions/NSW_iso.csv')\n",
    "merge_index = pd.to_datetime(df['Date'])\n",
    "\n",
    "del df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Before Clipping: (5866, 63)\n",
      "\n",
      "\n",
      "After Clipping: (5862, 63)\n"
     ]
    }
   ],
   "source": [
    "# For loop for the 7 region datasets\n",
    "regions = ['NSW', 'NT', 'QL', 'SA', 'TA', 'VI', 'WA']\n",
    "\n",
    "# Instantiate a pandas dataframe with the length of the date column\n",
    "merge_df = pd.DataFrame(index = range(len(merge_index)))\n",
    "for each in regions:\n",
    "    df = pd.read_csv('data_regions/{}_iso.csv'.format(each))\n",
    "    df.drop(columns = ['Date'], inplace=True)\n",
    "    # for each region concatenate\n",
    "    merge_df = pd.concat([merge_df, df], axis = 1)\n",
    "# Finally reset the index with Datetime\n",
    "merge_df.index = merge_index\n",
    "print('Before Clipping: {}'.format(merge_df.shape))\n",
    "print('\\n')\n",
    "# Dropping values observations from 2021-01-19 to 2021-01-22\n",
    "# There seems to be more missing values and I will interpolate the missing values\n",
    "merge_df = merge_df['2005-01-01':'2021-01-18'].copy()\n",
    "\n",
    "# There are 3 missing values within the temperature statistics (01-10-2021, 01-11-21, 01-17-21)\n",
    "# Interpolate the missing values\n",
    "merge_df.interpolate(inplace=True)\n",
    "print('After Clipping: {}'.format(merge_df.shape))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Feature Engineering\n",
    "Transform the estimated fire area to log scale and exponentiate after predicting in log scale form. <br>\n",
    "\n",
    "Surface Area:\n",
    "- Assume the fire areas are conglomerated into one pixel and the surface area would be <br>\n",
    "4 * square root of the Area\n",
    "\n",
    "- Assume the fire area pixels are separated(non touching) and the areas of each pixel are the same.<br>\n",
    "The surface area would be the count of the pixels * square root of (area / count)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "fire_regions = ['Estimated_fire_area_{}'.format(i) for i in regions]\n",
    "count_regions = ['Count_{}'.format(i) for i in regions]\n",
    "sqrt_fire_regions = ['sqrt_{}'.format(i) for i in fire_regions]\n",
    "sqrt_count_regions = ['sqrt_Count_{}'.format(i) for i in regions]\n",
    "\n",
    "for i,x in zip(fire_regions, count_regions):\n",
    "    merge_df[i] = np.log1p(merge_df[i])\n",
    "    merge_df[x] = np.log1p(merge_df[i])\n",
    "    \n",
    "# merge_df = pd.concat([merge_df[fire_regions], merge_df.drop(columns = fire_regions)], axis =1)\n",
    "\n",
    "\n",
    "for i,x in zip(fire_regions, count_regions):\n",
    "    merge_df['sqrt_{}'.format(i)] = np.sqrt(merge_df[i])\n",
    "    merge_df['sqrt_{}'.format(x)] = np.sqrt(merge_df[i]/merge_df[x]) * merge_df[x]\n",
    "    \n",
    "# sqrt_count_regions has NaN when dividing by 0\n",
    "merge_df.fillna(0, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "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>Estimated_fire_area_NSW</th>\n",
       "      <th>Estimated_fire_area_NT</th>\n",
       "      <th>Estimated_fire_area_QL</th>\n",
       "      <th>Estimated_fire_area_SA</th>\n",
       "      <th>Estimated_fire_area_TA</th>\n",
       "      <th>Estimated_fire_area_VI</th>\n",
       "      <th>Estimated_fire_area_WA</th>\n",
       "      <th>Count_NSW</th>\n",
       "      <th>mean_Precipitation_NSW</th>\n",
       "      <th>mean_RelativeHumidity_NSW</th>\n",
       "      <th>...</th>\n",
       "      <th>sqrt_Estimated_fire_area_QL</th>\n",
       "      <th>sqrt_Count_QL</th>\n",
       "      <th>sqrt_Estimated_fire_area_SA</th>\n",
       "      <th>sqrt_Count_SA</th>\n",
       "      <th>sqrt_Estimated_fire_area_TA</th>\n",
       "      <th>sqrt_Count_TA</th>\n",
       "      <th>sqrt_Estimated_fire_area_VI</th>\n",
       "      <th>sqrt_Count_VI</th>\n",
       "      <th>sqrt_Estimated_fire_area_WA</th>\n",
       "      <th>sqrt_Count_WA</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",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2005-01-01</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.782551</td>\n",
       "      <td>3.684660</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>3.621203</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.044274</td>\n",
       "      <td>36.355567</td>\n",
       "      <td>...</td>\n",
       "      <td>1.919547</td>\n",
       "      <td>2.385413</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.902946</td>\n",
       "      <td>2.354318</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2005-01-02</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.841567</td>\n",
       "      <td>4.181505</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.335001</td>\n",
       "      <td>5.004661</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.173879</td>\n",
       "      <td>38.699607</td>\n",
       "      <td>...</td>\n",
       "      <td>2.044873</td>\n",
       "      <td>2.622780</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.155422</td>\n",
       "      <td>1.064001</td>\n",
       "      <td>2.237110</td>\n",
       "      <td>2.995168</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2005-01-03</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>4.496917</td>\n",
       "      <td>3.879223</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.693147</td>\n",
       "      <td>4.121185</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>3.011912</td>\n",
       "      <td>49.365152</td>\n",
       "      <td>...</td>\n",
       "      <td>1.969574</td>\n",
       "      <td>2.479620</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.832555</td>\n",
       "      <td>0.604155</td>\n",
       "      <td>2.030070</td>\n",
       "      <td>2.594511</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2005-01-04</th>\n",
       "      <td>2.270062</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>3.437208</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>4.470200</td>\n",
       "      <td>1.184809</td>\n",
       "      <td>2.808620</td>\n",
       "      <td>57.095628</td>\n",
       "      <td>...</td>\n",
       "      <td>1.853971</td>\n",
       "      <td>2.263079</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2.114285</td>\n",
       "      <td>2.756135</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2005-01-05</th>\n",
       "      <td>2.868538</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.609438</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>3.477508</td>\n",
       "      <td>1.352877</td>\n",
       "      <td>0.157935</td>\n",
       "      <td>47.170735</td>\n",
       "      <td>...</td>\n",
       "      <td>1.268636</td>\n",
       "      <td>1.242444</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.864808</td>\n",
       "      <td>2.283203</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 77 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            Estimated_fire_area_NSW  Estimated_fire_area_NT  \\\n",
       "Date                                                          \n",
       "2005-01-01                 0.000000                1.782551   \n",
       "2005-01-02                 0.000000                0.841567   \n",
       "2005-01-03                 0.000000                4.496917   \n",
       "2005-01-04                 2.270062                0.000000   \n",
       "2005-01-05                 2.868538                0.000000   \n",
       "\n",
       "            Estimated_fire_area_QL  Estimated_fire_area_SA  \\\n",
       "Date                                                         \n",
       "2005-01-01                3.684660                     0.0   \n",
       "2005-01-02                4.181505                     0.0   \n",
       "2005-01-03                3.879223                     0.0   \n",
       "2005-01-04                3.437208                     0.0   \n",
       "2005-01-05                1.609438                     0.0   \n",
       "\n",
       "            Estimated_fire_area_TA  Estimated_fire_area_VI  \\\n",
       "Date                                                         \n",
       "2005-01-01                     0.0                0.000000   \n",
       "2005-01-02                     0.0                1.335001   \n",
       "2005-01-03                     0.0                0.693147   \n",
       "2005-01-04                     0.0                0.000000   \n",
       "2005-01-05                     0.0                0.000000   \n",
       "\n",
       "            Estimated_fire_area_WA  Count_NSW  mean_Precipitation_NSW  \\\n",
       "Date                                                                    \n",
       "2005-01-01                3.621203   0.000000                0.044274   \n",
       "2005-01-02                5.004661   0.000000                0.173879   \n",
       "2005-01-03                4.121185   0.000000                3.011912   \n",
       "2005-01-04                4.470200   1.184809                2.808620   \n",
       "2005-01-05                3.477508   1.352877                0.157935   \n",
       "\n",
       "            mean_RelativeHumidity_NSW  ...  sqrt_Estimated_fire_area_QL  \\\n",
       "Date                                   ...                                \n",
       "2005-01-01                  36.355567  ...                     1.919547   \n",
       "2005-01-02                  38.699607  ...                     2.044873   \n",
       "2005-01-03                  49.365152  ...                     1.969574   \n",
       "2005-01-04                  57.095628  ...                     1.853971   \n",
       "2005-01-05                  47.170735  ...                     1.268636   \n",
       "\n",
       "            sqrt_Count_QL  sqrt_Estimated_fire_area_SA  sqrt_Count_SA  \\\n",
       "Date                                                                    \n",
       "2005-01-01       2.385413                          0.0            0.0   \n",
       "2005-01-02       2.622780                          0.0            0.0   \n",
       "2005-01-03       2.479620                          0.0            0.0   \n",
       "2005-01-04       2.263079                          0.0            0.0   \n",
       "2005-01-05       1.242444                          0.0            0.0   \n",
       "\n",
       "            sqrt_Estimated_fire_area_TA  sqrt_Count_TA  \\\n",
       "Date                                                     \n",
       "2005-01-01                          0.0            0.0   \n",
       "2005-01-02                          0.0            0.0   \n",
       "2005-01-03                          0.0            0.0   \n",
       "2005-01-04                          0.0            0.0   \n",
       "2005-01-05                          0.0            0.0   \n",
       "\n",
       "            sqrt_Estimated_fire_area_VI  sqrt_Count_VI  \\\n",
       "Date                                                     \n",
       "2005-01-01                     0.000000       0.000000   \n",
       "2005-01-02                     1.155422       1.064001   \n",
       "2005-01-03                     0.832555       0.604155   \n",
       "2005-01-04                     0.000000       0.000000   \n",
       "2005-01-05                     0.000000       0.000000   \n",
       "\n",
       "            sqrt_Estimated_fire_area_WA  sqrt_Count_WA  \n",
       "Date                                                    \n",
       "2005-01-01                     1.902946       2.354318  \n",
       "2005-01-02                     2.237110       2.995168  \n",
       "2005-01-03                     2.030070       2.594511  \n",
       "2005-01-04                     2.114285       2.756135  \n",
       "2005-01-05                     1.864808       2.283203  \n",
       "\n",
       "[5 rows x 77 columns]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# reassigning the columns so the 7 estimated fire areas are the first 7 columns of the dataframe\n",
    "merge_df = pd.concat([merge_df[fire_regions], merge_df.drop(columns = fire_regions)], axis =1)\n",
    "merge_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Train test split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training Shape: (5814, 77)\n",
      "\n",
      "\n",
      "Testing Shape: (49, 77)\n"
     ]
    }
   ],
   "source": [
    "# separating the train test split\n",
    "# the values of the test set will be validated from the input of the last 120 days of the train set\n",
    "# It's important to note that after windowing of the training set, the last input days will be 41 days before 2020-12-01(2020-06-23:2020-10-21)\n",
    "# and will be validated by the last 41 days before 2020-12-01(2020-10-22:2020-12-01)\n",
    "# Please look below at windowing for more explanation\n",
    "train = merge_df[:'2020-12-01']\n",
    "test = merge_df['2020-12-01':]\n",
    "\n",
    "print('Training Shape: {}'.format(train.shape))\n",
    "print('\\n')\n",
    "print('Testing Shape: {}'.format(test.shape))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Preprocessing\n",
    "I chose MinMaxScaler because I want to retain the original distribution shape while scaling the features to a range value of 0 and 1. <br>\n",
    "The dataset needs to get windowed, essentially chopping the (5814, 77) training dataframe to 2 matrices X and y:<br>\n",
    "X: <br>\n",
    "(number of sequences, input steps, input features) <br>\n",
    "y: <br>\n",
    "(number of sequences, output steps, output features) <br>\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "X shape: (5654, 120, 77)\n",
      "\n",
      "\n",
      "y shape: (5654, 41, 7)\n"
     ]
    }
   ],
   "source": [
    "# MinMax Scaled\n",
    "scaler = MinMaxScaler()\n",
    "scaler.fit(train)\n",
    "scaled_train = scaler.transform(train)\n",
    "\n",
    "# Sequence split used from machine learning mastery\n",
    "def split_sequence(sequences, n_steps_in, n_steps_out):\n",
    "    X = []\n",
    "    y = []\n",
    "    for i in range(0, len(sequences), 1):\n",
    "        end_ix = i + n_steps_in\n",
    "        out_end_ix = end_ix + n_steps_out\n",
    "        if out_end_ix > len(sequences):\n",
    "            break\n",
    "        seq_x = sequences[i:end_ix, :]\n",
    "        seq_y = sequences[end_ix:out_end_ix, :7]\n",
    "        X.append(seq_x)\n",
    "        y.append(seq_y)\n",
    "        \n",
    "    return np.array(X), np.array(y)\n",
    "\n",
    "\n",
    "n_steps_in, n_steps_out = 120, 41\n",
    "X, y = split_sequence(scaled_train, n_steps_in, n_steps_out)\n",
    "print('X shape: {}'.format(X.shape))\n",
    "print('\\n')\n",
    "print('y shape: {}'.format(y.shape))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "y shape after flatten: (5654, 287)\n"
     ]
    }
   ],
   "source": [
    "# The output should predict the next 41 days for the 7 regions\n",
    "# In order to fit into the model I will essentiall flatten the 41 output steps and 7 regions to a long array\n",
    "# From (5654 sequences, 41 output steps, 7 regions) to (5654 sequences, 287 outputs)\n",
    "n_output = y.shape[1] * y.shape[2]\n",
    "y = y.reshape((y.shape[0], n_output))\n",
    "n_features = X.shape[2]\n",
    "print('y shape after flatten: {}'.format(y.shape))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### DCNN Model\n",
    "I fit the model once with an increasing learning rate from 1e-6 to 1e-1 to find my a stable learning rate for the final model. <br>\n",
    "After I can visually locate the learnign rate, I refit the model over 1500 epochs."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# DCNN architecture taken from Joseph Eddy\n",
    "# https://github.com/JEddy92/TimeSeries_Seq2Seq\n",
    "n_filters = 32\n",
    "filter_width = 2\n",
    "dilation_rates = [2**i for i in range(8)] * 2\n",
    "\n",
    "tf.keras.backend.clear_session()\n",
    "history_seq = tf.keras.layers.Input(shape =(n_steps_in, n_features))\n",
    "x = history_seq\n",
    "\n",
    "skips = []\n",
    "for dilation_rate in dilation_rates:\n",
    "    \n",
    "    # preprocessing - equivalent to time-distributed dense\n",
    "    x = tf.keras.layers.Conv1D(16, 1, activation='relu')(x) \n",
    "    \n",
    "    # filter convolution\n",
    "    x_f = tf.keras.layers.Conv1D(filters=n_filters,\n",
    "                                kernel_size=filter_width, \n",
    "                               activation = 'tanh',\n",
    "                                padding='causal',\n",
    "                                dilation_rate=dilation_rate)(x)\n",
    "    \n",
    "    # gating convolution\n",
    "    x_g = tf.keras.layers.Conv1D(filters=n_filters,\n",
    "                                 kernel_size=filter_width, \n",
    "                                 activation = 'sigmoid',\n",
    "                                 padding='causal',\n",
    "                                 dilation_rate=dilation_rate)(x)\n",
    "    \n",
    "    # multiply filter and gating branches\n",
    "    z = tf.keras.layers.Multiply()([x_f, x_g])\n",
    "    \n",
    "    # postprocessing - equivalent to time-distributed dense\n",
    "    z = tf.keras.layers.Conv1D(16, 1, activation='relu')(z)\n",
    "\n",
    "    \n",
    "    # residual connection\n",
    "    x = tf.keras.layers.Add()([x, z])    \n",
    "\n",
    "\n",
    "    # collect skip connections\n",
    "    skips.append(z)\n",
    "\n",
    "# add all skip connection outputs \n",
    "out = tf.keras.layers.Add()(skips)\n",
    "out = tf.keras.layers.Activation('relu')(out)\n",
    "\n",
    "# final time-distributed dense layers \n",
    "out = tf.keras.layers.Conv1D(128, 1, activation = 'relu')(out)\n",
    "out = tf.keras.layers.Dropout(.2)(out)\n",
    "out = tf.keras.layers.Conv1D(1, 1,)(out)\n",
    "out = tf.keras.layers.Flatten()(out)\n",
    "pred_seq_train = tf.keras.layers.Dense(n_output, activation = 'relu')(out)\n",
    "\n",
    "model = tf.keras.Model(history_seq, pred_seq_train)\n",
    "\n",
    "lr_schedule = tf.keras.callbacks.LearningRateScheduler(\n",
    "    lambda epoch: 1e-6 * 10**(epoch / 20))\n",
    "optimizer = tf.keras.optimizers.Adam(lr = 1e-6)\n",
    "\n",
    "model.compile(optimizer, loss= tf.keras.losses.Poisson(), metrics = ['mae'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      "142/142 [==============================] - 14s 97ms/step - loss: 2.5325 - mae: 0.2612 - val_loss: 2.6717 - val_mae: 0.2818\n",
      "Epoch 2/100\n",
      "142/142 [==============================] - 11s 78ms/step - loss: 2.5211 - mae: 0.2609 - val_loss: 2.6527 - val_mae: 0.2813\n",
      "Epoch 3/100\n",
      "142/142 [==============================] - 12s 82ms/step - loss: 2.5071 - mae: 0.2605 - val_loss: 2.6346 - val_mae: 0.2809\n",
      "Epoch 4/100\n",
      "142/142 [==============================] - 12s 86ms/step - loss: 2.4933 - mae: 0.2602 - val_loss: 2.6177 - val_mae: 0.2804\n",
      "Epoch 5/100\n",
      "142/142 [==============================] - 12s 83ms/step - loss: 2.4757 - mae: 0.2598 - val_loss: 2.5991 - val_mae: 0.2799\n",
      "Epoch 6/100\n",
      "142/142 [==============================] - 12s 83ms/step - loss: 2.4630 - mae: 0.2593 - val_loss: 2.5802 - val_mae: 0.2793\n",
      "Epoch 7/100\n",
      "142/142 [==============================] - 12s 81ms/step - loss: 2.4490 - mae: 0.2589 - val_loss: 2.5584 - val_mae: 0.2784\n",
      "Epoch 8/100\n",
      "142/142 [==============================] - 12s 82ms/step - loss: 2.4270 - mae: 0.2583 - val_loss: 2.5396 - val_mae: 0.2779\n",
      "Epoch 9/100\n",
      "142/142 [==============================] - 12s 83ms/step - loss: 2.4136 - mae: 0.2579 - val_loss: 2.5198 - val_mae: 0.2773\n",
      "Epoch 10/100\n",
      "142/142 [==============================] - 12s 82ms/step - loss: 2.3963 - mae: 0.2575 - val_loss: 2.4948 - val_mae: 0.2766\n",
      "Epoch 11/100\n",
      "142/142 [==============================] - 12s 81ms/step - loss: 2.3752 - mae: 0.2568 - val_loss: 2.4689 - val_mae: 0.2757\n",
      "Epoch 12/100\n",
      "142/142 [==============================] - 12s 81ms/step - loss: 2.3570 - mae: 0.2564 - val_loss: 2.4461 - val_mae: 0.2753\n",
      "Epoch 13/100\n",
      "142/142 [==============================] - 11s 81ms/step - loss: 2.3340 - mae: 0.2557 - val_loss: 2.4129 - val_mae: 0.2739\n",
      "Epoch 14/100\n",
      "142/142 [==============================] - 12s 82ms/step - loss: 2.3037 - mae: 0.2546 - val_loss: 2.3799 - val_mae: 0.2724\n",
      "Epoch 15/100\n",
      "142/142 [==============================] - 12s 81ms/step - loss: 2.2786 - mae: 0.2536 - val_loss: 2.3483 - val_mae: 0.2710\n",
      "Epoch 16/100\n",
      "142/142 [==============================] - 12s 81ms/step - loss: 2.2499 - mae: 0.2525 - val_loss: 2.3093 - val_mae: 0.2694\n",
      "Epoch 17/100\n",
      "142/142 [==============================] - 12s 84ms/step - loss: 2.2159 - mae: 0.2512 - val_loss: 2.2672 - val_mae: 0.2675\n",
      "Epoch 18/100\n",
      "142/142 [==============================] - 12s 85ms/step - loss: 2.1799 - mae: 0.2498 - val_loss: 2.2284 - val_mae: 0.2657\n",
      "Epoch 19/100\n",
      "142/142 [==============================] - 12s 85ms/step - loss: 2.1389 - mae: 0.2481 - val_loss: 2.1848 - val_mae: 0.2630\n",
      "Epoch 20/100\n",
      "142/142 [==============================] - 12s 82ms/step - loss: 2.0993 - mae: 0.2462 - val_loss: 2.1420 - val_mae: 0.2607\n",
      "Epoch 21/100\n",
      "142/142 [==============================] - 12s 83ms/step - loss: 2.0596 - mae: 0.2446 - val_loss: 2.0945 - val_mae: 0.2580\n",
      "Epoch 22/100\n",
      "142/142 [==============================] - 12s 82ms/step - loss: 2.0132 - mae: 0.2427 - val_loss: 2.0417 - val_mae: 0.2551\n",
      "Epoch 23/100\n",
      "142/142 [==============================] - 12s 83ms/step - loss: 1.9613 - mae: 0.2405 - val_loss: 1.9851 - val_mae: 0.2519\n",
      "Epoch 24/100\n",
      "142/142 [==============================] - 12s 82ms/step - loss: 1.9051 - mae: 0.2384 - val_loss: 1.9305 - val_mae: 0.2487\n",
      "Epoch 25/100\n",
      "142/142 [==============================] - 12s 82ms/step - loss: 1.8455 - mae: 0.2359 - val_loss: 1.8763 - val_mae: 0.2446\n",
      "Epoch 26/100\n",
      "142/142 [==============================] - 12s 82ms/step - loss: 1.7818 - mae: 0.2334 - val_loss: 1.8179 - val_mae: 0.2408\n",
      "Epoch 27/100\n",
      "142/142 [==============================] - 12s 82ms/step - loss: 1.7100 - mae: 0.2307 - val_loss: 1.7449 - val_mae: 0.2357\n",
      "Epoch 28/100\n",
      "142/142 [==============================] - 12s 84ms/step - loss: 1.6281 - mae: 0.2277 - val_loss: 1.6389 - val_mae: 0.2314\n",
      "Epoch 29/100\n",
      "142/142 [==============================] - 12s 81ms/step - loss: 1.5484 - mae: 0.2249 - val_loss: 1.5631 - val_mae: 0.2267\n",
      "Epoch 30/100\n",
      "142/142 [==============================] - 12s 81ms/step - loss: 1.4763 - mae: 0.2227 - val_loss: 1.4831 - val_mae: 0.2233\n",
      "Epoch 31/100\n",
      "142/142 [==============================] - 12s 81ms/step - loss: 1.4044 - mae: 0.2204 - val_loss: 1.4019 - val_mae: 0.2193\n",
      "Epoch 32/100\n",
      "142/142 [==============================] - 12s 82ms/step - loss: 1.3417 - mae: 0.2181 - val_loss: 1.3469 - val_mae: 0.2155\n",
      "Epoch 33/100\n",
      "142/142 [==============================] - 12s 82ms/step - loss: 1.2932 - mae: 0.2158 - val_loss: 1.3131 - val_mae: 0.2117\n",
      "Epoch 34/100\n",
      "142/142 [==============================] - 12s 83ms/step - loss: 1.2559 - mae: 0.2131 - val_loss: 1.2713 - val_mae: 0.2085\n",
      "Epoch 35/100\n",
      "142/142 [==============================] - 12s 82ms/step - loss: 1.2161 - mae: 0.2107 - val_loss: 1.2550 - val_mae: 0.2051\n",
      "Epoch 36/100\n",
      "142/142 [==============================] - 12s 82ms/step - loss: 1.1968 - mae: 0.2084 - val_loss: 1.2485 - val_mae: 0.2019\n",
      "Epoch 37/100\n",
      "142/142 [==============================] - 12s 82ms/step - loss: 1.1866 - mae: 0.2054 - val_loss: 1.2423 - val_mae: 0.1988\n",
      "Epoch 38/100\n",
      "142/142 [==============================] - 12s 82ms/step - loss: 1.1665 - mae: 0.2027 - val_loss: 1.2152 - val_mae: 0.1967\n",
      "Epoch 39/100\n",
      "142/142 [==============================] - 12s 82ms/step - loss: 1.1350 - mae: 0.2001 - val_loss: 1.1904 - val_mae: 0.1936\n",
      "Epoch 40/100\n",
      "142/142 [==============================] - 12s 83ms/step - loss: 1.1289 - mae: 0.1978 - val_loss: 1.1888 - val_mae: 0.1919\n",
      "Epoch 41/100\n",
      "142/142 [==============================] - 12s 82ms/step - loss: 1.1254 - mae: 0.1947 - val_loss: 1.1861 - val_mae: 0.1889\n",
      "Epoch 42/100\n",
      "142/142 [==============================] - 12s 82ms/step - loss: 1.1226 - mae: 0.1926 - val_loss: 1.1851 - val_mae: 0.1876\n",
      "Epoch 43/100\n",
      "142/142 [==============================] - 12s 82ms/step - loss: 1.1196 - mae: 0.1907 - val_loss: 1.1834 - val_mae: 0.1854\n",
      "Epoch 44/100\n",
      "142/142 [==============================] - 12s 82ms/step - loss: 1.1172 - mae: 0.1884 - val_loss: 1.1827 - val_mae: 0.1840\n",
      "Epoch 45/100\n",
      "142/142 [==============================] - 12s 82ms/step - loss: 1.1144 - mae: 0.1853 - val_loss: 1.1806 - val_mae: 0.1806\n",
      "Epoch 46/100\n",
      "142/142 [==============================] - 12s 81ms/step - loss: 1.1125 - mae: 0.1835 - val_loss: 1.1809 - val_mae: 0.1805\n",
      "Epoch 47/100\n",
      "142/142 [==============================] - 12s 82ms/step - loss: 1.1118 - mae: 0.1825 - val_loss: 1.1805 - val_mae: 0.1799\n",
      "Epoch 48/100\n",
      "142/142 [==============================] - 12s 82ms/step - loss: 1.1090 - mae: 0.1794 - val_loss: 1.1787 - val_mae: 0.1771\n",
      "Epoch 49/100\n",
      "142/142 [==============================] - 11s 81ms/step - loss: 1.1078 - mae: 0.1778 - val_loss: 1.1784 - val_mae: 0.1762\n",
      "Epoch 50/100\n",
      "142/142 [==============================] - 11s 80ms/step - loss: 1.1064 - mae: 0.1759 - val_loss: 1.1776 - val_mae: 0.1747\n",
      "Epoch 51/100\n",
      "142/142 [==============================] - 11s 81ms/step - loss: 1.1051 - mae: 0.1741 - val_loss: 1.1768 - val_mae: 0.1726\n",
      "Epoch 52/100\n",
      "142/142 [==============================] - 12s 84ms/step - loss: 1.1054 - mae: 0.1744 - val_loss: 1.1769 - val_mae: 0.1725\n",
      "Epoch 53/100\n",
      "142/142 [==============================] - 12s 84ms/step - loss: 1.1029 - mae: 0.1712 - val_loss: 1.1762 - val_mae: 0.1710\n",
      "Epoch 54/100\n",
      "142/142 [==============================] - 12s 82ms/step - loss: 1.1117 - mae: 0.1823 - val_loss: 1.1804 - val_mae: 0.1774\n",
      "Epoch 55/100\n",
      "142/142 [==============================] - 11s 80ms/step - loss: 1.1065 - mae: 0.1758 - val_loss: 1.1797 - val_mae: 0.1746\n",
      "Epoch 56/100\n",
      "142/142 [==============================] - 12s 81ms/step - loss: 1.1057 - mae: 0.1744 - val_loss: 1.1803 - val_mae: 0.1750\n",
      "Epoch 57/100\n",
      "142/142 [==============================] - 11s 81ms/step - loss: 1.1068 - mae: 0.1744 - val_loss: 1.1824 - val_mae: 0.1798\n",
      "Epoch 58/100\n",
      "142/142 [==============================] - 11s 81ms/step - loss: 1.0996 - mae: 0.1786 - val_loss: 1.1667 - val_mae: 0.1742\n",
      "Epoch 59/100\n",
      "142/142 [==============================] - 11s 80ms/step - loss: 1.0937 - mae: 0.1732 - val_loss: 1.1649 - val_mae: 0.1703\n",
      "Epoch 60/100\n",
      "142/142 [==============================] - 12s 86ms/step - loss: 1.0910 - mae: 0.1691 - val_loss: 1.1642 - val_mae: 0.1687\n",
      "Epoch 61/100\n",
      "142/142 [==============================] - 12s 86ms/step - loss: 1.0897 - mae: 0.1669 - val_loss: 1.1646 - val_mae: 0.1689\n",
      "Epoch 62/100\n",
      "142/142 [==============================] - 12s 83ms/step - loss: 1.0890 - mae: 0.1654 - val_loss: 1.1632 - val_mae: 0.1658\n",
      "Epoch 63/100\n",
      "142/142 [==============================] - 11s 81ms/step - loss: 1.0879 - mae: 0.1634 - val_loss: 1.1620 - val_mae: 0.1637\n",
      "Epoch 64/100\n",
      "142/142 [==============================] - 12s 83ms/step - loss: 1.0865 - mae: 0.1611 - val_loss: 1.1612 - val_mae: 0.1611\n",
      "Epoch 65/100\n",
      "142/142 [==============================] - 11s 81ms/step - loss: 1.0855 - mae: 0.1593 - val_loss: 1.1605 - val_mae: 0.1609\n",
      "Epoch 66/100\n",
      "142/142 [==============================] - 11s 81ms/step - loss: 1.0913 - mae: 0.1667 - val_loss: 1.1667 - val_mae: 0.1680\n",
      "Epoch 67/100\n",
      "142/142 [==============================] - 12s 83ms/step - loss: 1.0884 - mae: 0.1627 - val_loss: 1.1628 - val_mae: 0.1635\n",
      "Epoch 68/100\n",
      "142/142 [==============================] - 12s 83ms/step - loss: 1.0846 - mae: 0.1581 - val_loss: 1.1612 - val_mae: 0.1609\n",
      "Epoch 69/100\n",
      "142/142 [==============================] - 12s 83ms/step - loss: 1.2374 - mae: 0.2511 - val_loss: 1.2617 - val_mae: 0.1898\n",
      "Epoch 70/100\n",
      "142/142 [==============================] - 12s 85ms/step - loss: 1.1717 - mae: 0.1751 - val_loss: 1.2511 - val_mae: 0.1721\n",
      "Epoch 71/100\n",
      "142/142 [==============================] - 12s 81ms/step - loss: 1.1638 - mae: 0.1655 - val_loss: 1.2395 - val_mae: 0.1719\n",
      "Epoch 72/100\n",
      "142/142 [==============================] - 13s 88ms/step - loss: 1.1550 - mae: 0.1660 - val_loss: 1.2307 - val_mae: 0.1686\n",
      "Epoch 73/100\n",
      "142/142 [==============================] - 13s 89ms/step - loss: 1.1465 - mae: 0.1604 - val_loss: 1.2258 - val_mae: 0.1634\n",
      "Epoch 74/100\n",
      "142/142 [==============================] - 12s 88ms/step - loss: 1.1388 - mae: 0.1581 - val_loss: 1.2194 - val_mae: 0.1637\n",
      "Epoch 75/100\n",
      "142/142 [==============================] - 12s 86ms/step - loss: 1.1334 - mae: 0.1610 - val_loss: 1.1949 - val_mae: 0.1722\n",
      "Epoch 76/100\n",
      "142/142 [==============================] - 12s 83ms/step - loss: 1.0222 - mae: 0.1580 - val_loss: 1.0712 - val_mae: 0.1573\n",
      "Epoch 77/100\n",
      "142/142 [==============================] - 12s 83ms/step - loss: 1.0030 - mae: 0.1501 - val_loss: 1.0721 - val_mae: 0.1540\n",
      "Epoch 78/100\n",
      "142/142 [==============================] - 12s 82ms/step - loss: 0.9680 - mae: 0.1743 - val_loss: 0.9722 - val_mae: 0.1576\n",
      "Epoch 79/100\n",
      "142/142 [==============================] - 12s 82ms/step - loss: 0.9192 - mae: 0.1539 - val_loss: 0.9789 - val_mae: 0.1709\n",
      "Epoch 80/100\n",
      "142/142 [==============================] - 12s 82ms/step - loss: 0.9187 - mae: 0.1542 - val_loss: 0.9684 - val_mae: 0.1537\n",
      "Epoch 81/100\n",
      "142/142 [==============================] - 12s 82ms/step - loss: 0.9031 - mae: 0.1529 - val_loss: 0.9432 - val_mae: 0.1642\n",
      "Epoch 82/100\n",
      "142/142 [==============================] - 12s 86ms/step - loss: 0.8844 - mae: 0.1528 - val_loss: 0.9362 - val_mae: 0.1655\n",
      "Epoch 83/100\n",
      "142/142 [==============================] - 12s 83ms/step - loss: 0.8698 - mae: 0.1545 - val_loss: 0.9128 - val_mae: 0.1640\n",
      "Epoch 84/100\n",
      "142/142 [==============================] - 12s 85ms/step - loss: 0.8154 - mae: 0.1541 - val_loss: 0.8341 - val_mae: 0.1499\n",
      "Epoch 85/100\n",
      "142/142 [==============================] - 12s 82ms/step - loss: 0.7558 - mae: 0.1445 - val_loss: 0.8035 - val_mae: 0.1489\n",
      "Epoch 86/100\n",
      "142/142 [==============================] - 12s 82ms/step - loss: 0.7504 - mae: 0.1508 - val_loss: 0.7099 - val_mae: 0.1629\n",
      "Epoch 87/100\n",
      "142/142 [==============================] - 12s 81ms/step - loss: 0.6656 - mae: 0.1609 - val_loss: 0.6448 - val_mae: 0.1631\n",
      "Epoch 88/100\n",
      "142/142 [==============================] - 12s 86ms/step - loss: 1.2976 - mae: 0.4807 - val_loss: 3.2306 - val_mae: 1.1259\n",
      "Epoch 89/100\n",
      "142/142 [==============================] - 12s 82ms/step - loss: 1.4889 - mae: 0.3896 - val_loss: 0.7690 - val_mae: 0.1920\n",
      "Epoch 90/100\n",
      "142/142 [==============================] - 11s 81ms/step - loss: 0.8366 - mae: 0.2408 - val_loss: 0.7600 - val_mae: 0.1912\n",
      "Epoch 91/100\n",
      "142/142 [==============================] - 12s 82ms/step - loss: 0.7600 - mae: 0.2150 - val_loss: 0.7053 - val_mae: 0.1721\n",
      "Epoch 92/100\n",
      "142/142 [==============================] - 12s 83ms/step - loss: 0.7134 - mae: 0.1953 - val_loss: 0.7846 - val_mae: 0.1795\n",
      "Epoch 93/100\n",
      "142/142 [==============================] - 12s 86ms/step - loss: 0.6742 - mae: 0.1662 - val_loss: 0.6983 - val_mae: 0.1513\n",
      "Epoch 94/100\n",
      "142/142 [==============================] - 13s 90ms/step - loss: 0.6582 - mae: 0.1544 - val_loss: 0.6973 - val_mae: 0.1479\n",
      "Epoch 95/100\n",
      "142/142 [==============================] - 12s 82ms/step - loss: 0.6568 - mae: 0.1503 - val_loss: 0.6970 - val_mae: 0.1466\n",
      "Epoch 96/100\n",
      "142/142 [==============================] - 12s 82ms/step - loss: 0.6568 - mae: 0.1499 - val_loss: 0.6972 - val_mae: 0.1486\n",
      "Epoch 97/100\n",
      "142/142 [==============================] - 12s 83ms/step - loss: 0.6568 - mae: 0.1501 - val_loss: 0.6979 - val_mae: 0.1458\n",
      "Epoch 98/100\n",
      "142/142 [==============================] - 11s 80ms/step - loss: 0.6667 - mae: 0.1537 - val_loss: 1.0417 - val_mae: 0.3471\n",
      "Epoch 99/100\n",
      "142/142 [==============================] - 12s 81ms/step - loss: 1.3802 - mae: 0.4980 - val_loss: 1.3396 - val_mae: 0.3020\n",
      "Epoch 100/100\n",
      "142/142 [==============================] - 13s 90ms/step - loss: 2.2551 - mae: 0.8491 - val_loss: 2.9961 - val_mae: 0.8205\n"
     ]
    }
   ],
   "source": [
    "history = model.fit(X, y, epochs = 100, callbacks = [lr_schedule], validation_split = 0.2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYAAAAEMCAYAAADNtWEcAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3deXzU1b3/8ddnJvvCkhAW2QJFQcWViGitu1asSqvWulZbLdrWWrs8Wtt7f2313t7u7XWrikvdt1rbi1avy1VcAQ2KIiqW3YCYECBkT2bm8/tjBogxIROYJcm8n4/HPJzvd858v585hvnMOd9zztfcHRERyTyBdAcgIiLpoQQgIpKhlABERDKUEoCISIZSAhARyVBKACIiGUoJQAYEM3vSzC5MdFmRgcw0D0DSxcwaOmwWAK1AOLZ9qbvfl/qodp2ZHQ3c6+5j0h2LSDyy0h2AZC53L9r23MxWA5e4+7Ody5lZlruHUhmbSCZQF5D0OWZ2tJlVmdmPzWwD8BczG2pmj5tZjZltjj0f0+E988zsktjzi8zsZTP7fazsKjObuYtlJ5jZi2ZWb2bPmtmNZnbvLnymvWPn3WJmS83stA6vnWxm78bOsc7MfhjbPyz2ObeY2SYze8nM9G9WEkZ/TNJXjQRKgPHAbKJ/q3+JbY8DmoEbdvL+Q4FlwDDgt8DtZma7UPZ+4DWgFPgFcEFvP4iZZQOPAU8Dw4HvAPeZ2eRYkduJdnkVA1OB52L7fwBUAWXACOCngPpsJWGUAKSvigA/d/dWd29291p3/5u7N7l7PfBL4KidvH+Nu9/q7mHgLmAU0S/RuMua2TjgEOBn7t7m7i8Dc3fhs8wAioBfx47zHPA4cE7s9XZgHzMb5O6b3f2NDvtHAePdvd3dX3JdtJMEUgKQvqrG3Vu2bZhZgZndYmZrzGwr8CIwxMyC3bx/w7Yn7t4Ue1rUy7J7AJs67AP4sJefg9hxPnT3SId9a4DRsednACcDa8zsBTM7LLb/d8By4GkzW2lmV+3CuUW6pQQgfVXnX7o/ACYDh7r7IODI2P7uunUS4SOgxMwKOuwbuwvHWQ+M7dR/Pw5YB+Dur7v7LKLdQ/8AHo7tr3f3H7j7ROBU4PtmdtwunF+kS0oA0l8UE+3332JmJcDPk31Cd18DVAK/MLOc2C/zU3t6n5nldXwQvYbQCPzIzLJjw0VPBR6MHfc8Mxvs7u3AVmJDYc3sFDObFLsesW1/uMuTiuwCJQDpL/4byAc2AguA/03Rec8DDgNqgf8EHiI6X6E7o4kmqo6PscBpwEyi8f8Z+Kq7vx97zwXA6ljX1mXA+bH9ewLPAg3AfODP7j4vUR9MRBPBRHrBzB4C3nf3pLdARJJNLQCRnTCzQ8zsM2YWMLOTgFlE++lF+r24E4CZBc3sTTN7vIvXcs3sITNbbmYLzaw8kUGKpNFIYB7RbpjrgG+6+5tpjUgkQXqzFMR3gfeAQV28djGw2d0nmdnZwG+AryQgPpG0cvfHiE7iEhlw4moBxKbcfwG4rZsis4hOoAF4BDhuJ7MuRUSkD4i3C+i/gR8RnZ3ZldHEJsjEFu2qIzp1XkRE+qgeu4DM7BSg2t0XxcYvd1msi32fGl5kZrOJrutCYWHhtClTpvQiVBERWbRo0UZ3L0vEseK5BvBZ4DQzOxnIAwaZ2b3ufn6HMlVExzpXmVkWMBjY1PlA7j4HmANQUVHhlZWVuxu/iEhGMbM1iTpWj11A7v4Tdx/j7uXA2cBznb78IbpA1rY7LJ0ZK6MJBiIifdgu3xDGzK4BKt19LtHlbO8xs+VEf/mfnaD4REQkSXqVAGLT0OfFnv+sw/4W4MuJDExERJJLM4FFRDKUEoCISIZSAhARyVBKACIiGUoJQEQkQykBiIhkKCUAEZEMpQQgIpKhlABERDKUEoCISIZSAhARyVBKACIiGUoJQEQkQykBiIhkKCUAEZEMpQQgIpKhlABERDKUEoCISIbqMQGYWZ6ZvWZmb5nZUjO7uosyF5lZjZktjj0uSU64IiKSKPHcE7gVONbdG8wsG3jZzJ509wWdyj3k7pcnPkQREUmGHhOAuzvQENvMjj08mUGJiEjyxXUNwMyCZrYYqAaecfeFXRQ7w8zeNrNHzGxsQqMUEZGEiysBuHvY3Q8ExgDTzWxqpyKPAeXuvj/wLHBXV8cxs9lmVmlmlTU1NbsTt4iI7KZejQJy9y3APOCkTvtr3b01tnkrMK2b989x9wp3rygrK9uFcEVEJFHiGQVUZmZDYs/zgeOB9zuVGdVh8zTgvUQGKSIiiRfPKKBRwF1mFiSaMB5298fN7Bqg0t3nAleY2WlACNgEXJSsgEVEJDEsOsgn9SoqKryysjIt5xYR6a/MbJG7VyTiWJoJLCKSoZQAREQylBKAiEiGUgIQEclQSgAiIhlKCUBEJEMpAYiIZCglABGRDKUEICKSoZQAREQylBKAiEiGUgIQEclQSgAiIhlKCUBEJEMpAYiIZCglABGRDKUEICKSoZQAREQylBKAiEiG6jEBmFmemb1mZm+Z2VIzu7qLMrlm9pCZLTezhWZWnoxgRUQkceJpAbQCx7r7AcCBwElmNqNTmYuBze4+CfgT8JvEhikiIonWYwLwqIbYZnbs4Z2KzQLuij1/BDjOzCxhUYqISMLFdQ3AzIJmthioBp5x94WdiowGPgRw9xBQB5R2cZzZZlZpZpU1NTW7F7mIiOyWuBKAu4fd/UBgDDDdzKZ2KtLVr/3OrQTcfY67V7h7RVlZWe+jFRGRhOnVKCB33wLMA07q9FIVMBbAzLKAwcCmBMQnIiJJEs8ooDIzGxJ7ng8cD7zfqdhc4MLY8zOB59z9Uy0AERHpO7LiKDMKuMvMgkQTxsPu/riZXQNUuvtc4HbgHjNbTvSX/9lJi1hERBKixwTg7m8DB3Wx/2cdnrcAX05saCIikkyaCSwikqGUAEREMpQSgIhIhlICEBHJUEoAIiIZSglARCRDKQGIiGQoJQARkQylBCAikqGUAEREMpQSgIhIhlICEBHJUEoAIiIZSglARCRDKQGIiGQoJQARkQylBCAikqGUAEREMlQ8N4Ufa2bPm9l7ZrbUzL7bRZmjzazOzBbHHj/r6lgdbW5q4511dbSGwrsau4iI7IZ4bgofAn7g7m+YWTGwyMyecfd3O5V7yd1PiffEVZubOeX6l8kKGJOGF3HQuKFMGx99lJcWYGa9+RwiItJL8dwU/iPgo9jzejN7DxgNdE4AvbLXiGJ+de5BvLt+K++s38rjb6/ngdfWAlBSmMNBY4dw8PihVIwfysHjh5IdVG+ViEgimbvHX9isHHgRmOruWzvsPxr4G1AFrAd+6O5Lu3j/bGA2wLhx46atWbNm+2uRiLO8poHK1Zt5Y230sbKmEYDi3CyO2HMYR08uY9r4EiYOKyQQUAtBRDKPmS1y94qEHCveBGBmRcALwC/d/dFOrw0CIu7eYGYnA9e6+547O15FRYVXVlbu9JybG9tYuKqWectqmLeshg1bWwAoys1i6uhBHDg22mV08LghlBblxvU5RET6s5QnADPLBh4HnnL3P8ZRfjVQ4e4buysTTwLoyN1ZXt3A4g+3sGRdHW9V1fHu+jraw9H4Jw0v4pT9R/HFA0dTPqww7uOKiPQnKU0AFr0aexewyd2v7KbMSOBjd3czmw48Aoz3nRy8twmgKy3tYZasq2PRms3MW1bNwlWbcIcDxw7hiEnDOGjcEA4cq9aBiAwcqU4ARwAvAUuASGz3T4FxAO5+s5ldDnyT6IihZuD77v7qzo6biATQ2Ud1zcxdvJ5/LvmIpeu3Eo5EP9tnygo5aq/hHDW5jEMnlJCXHUzoeUVEUiUt1wASLRkJoKOmthDvrNvKG2s38+qKWhasrKUtFCE/O8ixew/n1P334OjJZUoGItKvKAHsgua2MAtX1fLMux/z5Dsb2NTYRlFuFkdNLuO4KcM5evJwSgpzUhaPiMiuUALYTaFwhPkra3n8rY94blk1NfWtmMEh5SWcefAYTt5/FEW58cyRExFJLSWABIpEnHfW1/Hse9U89tZ6Vm1sJD87yMz9RnLpkZ9h8sjidIcoIrKdEkCSuDtvrN3CI4uqmLt4HY1tYWZOHcnlx05i3z0Gpzs8ERElgFTY3NjGHa+s4s5XVlPfGuKIScM4f8Y4jt97BFlalkJE0kQJIIXqmtu5Z/5q7l+4lvV1LYwYlMt5h47nghnjGaqLxiKSYkoAaRAKR3h+WQ33LljDCx/UkJ8d5OzpY7nkcxMZPSQ/3eGJyADUForwxJKPOGjcEMaXRlc4SGQCUF9GnLKCAU7YZwR3fX06T115JDP3G8k989dw1G+f5xdzl7KpsS3dIYrIAFPb2MqVDy3m1RW1STm+EsAumDyymD+edSAv/OgYzjpkLHfPX81Rv32em+atoKVdN7gRkcTY0tQOwJD87KQcXwlgN4weks9/fWk/nrrySKZPKOE3//s+J/zpBZ5auoF0da2JyMCxLQEMLlAC6LP2HFHM7Rcdwn2XHEp+dpBL71nEBbe/xr8+rk93aCLSj9U1R7uWh+QnZ8CJEkACfXbSMJ644nNcfdq+vF21hZOve4kbnvsX7eFIz28WEelkexeQWgD9Q1YwwIWHl/P8D4/m8/uO5PdPf8CX/vwK72/Y2vObRUQ62NKsBNAvlRblcsO5B3PTeQezoa6FU69/mT898wGtIV0kFpH4bGlqJycYID9JqxYrASTZzP1G8fT3juIL+43i2v/7F6dc9zKL1mxOd1gi0g/UNbcxuCCb6H25Ek8JIAVKCnP477MP4i8XHUJja4gzb36VX8xdSkNrKN2hiUgftqWpPWlDQEEJIKWOmTKcp79/FF+dMZ675q/m8396keeXVac7LBHpo7Y0tSet/x+UAFKuKDeLq2dN5ZHLDiM/J8jX/vI63394sVoDIvIpW5rbGZykIaAQRwIws7Fm9ryZvWdmS83su12UMTO7zsyWm9nbZnZwcsIdOKaNL+GfVxzBFcdO4h9vruO061/WSCER+YS6pra0twBCwA/cfW9gBvBtM9unU5mZwJ6xx2zgpoRGOUDlZgX5/omTue+SGdS3hph1wys89PpazSIWESDaAkjrNQB3/8jd34g9rwfeA0Z3KjYLuNujFgBDzGxUwqMdoA77TClPXPE5KsqH8uO/LeFn/7OUkCaPiWS01lCYprZw2lsA25lZOXAQsLDTS6OBDztsV/HpJIGZzTazSjOrrKmp6V2kA1xZcS53f/1QLj1yIvcsWMPX7nydutgkEBHJPNv+/Q8uSOM1gG3MrAj4G3Clu3furO5qkOqn+jHcfY67V7h7RVlZWe8izQDBgPGTk/fmt2fsz/wVtZz+51dYU9uY7rBEJA3qkrwSKMSZAMwsm+iX/33u/mgXRaqAsR22xwDrdz+8zHTWIWO595JDqW1s44ybXuWddXXpDklEUizZy0BAfKOADLgdeM/d/9hNsbnAV2OjgWYAde7+UQLjzDgzJpbyyGWHkxMMcPacBcxP0g0hRKRv2nEvgPR2AX0WuAA41swWxx4nm9llZnZZrMwTwEpgOXAr8K3khJtZJg0v4m/fOpxRg/O48I7XeHKJcqpIptjSFFsKOoktgKyeCrj7y3Tdx9+xjAPfTlRQssOowfn89bLD+Pqdr/Pt+9/g918+gNMPHpPusEQkyXZcBO4jo4AkPYYU5HDvJYcyY2IpP/jrWzzw2tp0hyQiSbalqZ1gwCjO7fF3+i5TAugnCnKyuOOiQzhyzzJ+8ugS7np1dbpDEpEk2tLcxuD85K0ECkoA/UpedpA5X53GCfuM4Odzl/KXV1alOyQRSZJkrwQKSgD9Tm5WkD+fdzAn7jOCqx97l/sXqjtIZCCqa25Pav8/KAH0S9nBANefexBHTy7j3/6xhEffqEp3SCKSYGoBSLdys4LcfP40DptYyg//+hZPaIioyICypbmNIUlcBgKUAPq1vOwgt11YwcHjhnLlQ4tZtGZTukMSkQTZ0tTOYLUAZGcKcrK49asVjB6SzzfuXqS1g0QGgFA4Qn1LKKmTwEAJYEAYWpjDHRcdQsQ9uopok1YRFenPtrZE7xCoawASlwnDCplzQQVVm5q59N5K3U9ApB/bvH0ZCF0DkDhNn1DCr8/YjwUrN/GHZz5Idzgisou2LQSnYaDSK6cfPIZzpo/jpnkreP796nSHIyK7oK452gIYqhaA9NbPT92HvUcN4vsPL2b9luZ0hyMivbQlBTeDASWAASkvO8iN5x5EWyjCdx54k3ZdDxDpV7YnAHUBya6YWFbEr8/Yn0VrNvObJ99Pdzgi0gtbmtsxg+I8JQDZRacesAdfPWw8t728SjOFRfqRuqY2BuVlEwwkbyVQUAIY8P7tC3tz4Ngh/OiRt1lR05DucEQkDlua25Pe/QNKAAPettVDs4PGN+9dRFNbKN0hiUgPUrEQHMR3U/g7zKzazN7p5vWjzayuw/2Cf5b4MGV37DEkn+vOOYh/VTfwo0feJnoHTxHpq7Y0tzM4yUNAIb4WwJ3AST2UecndD4w9rtn9sCTRPrdnGT/6/BQef/sj/qRJYiJ9Wl1TW0paAPHcFP5FMytPeiSSdJcdNZFVGxu47rnljC8t5Ixpurm8SF/U364BHGZmb5nZk2a2b3eFzGy2mVWaWWVNTU2CTi3xMjP+84v7cdjEUq569G0WrqxNd0gifcotL6zg9dXpXVY9EnHqmvvINYA4vAGMd/cDgOuBf3RX0N3nuHuFu1eUlZUl4NTSWzlZAW4+fxpjSwq49N5FrNqo5aNFANZvaeZXT77Pw69/mNY46ltCuNNnrgHslLtvdfeG2PMngGwzG7bbkUnSDC7I5i8XHULAjK/f+TqbG9vSHZJI2j21dAMANQ2taY1jS2wdoH7RAjCzkWZmsefTY8dU30IfN760kDkXTGPd5mYuvWcRraFwukMSSav/fSeWAOrTnABStAwExDcM9AFgPjDZzKrM7GIzu8zMLosVORN4x8zeAq4DznaNM+wXKspL+N2X9+e11Zu46m9LNDxUMtbGhlZeX70Jsz6QAJpTlwDiGQV0Tg+v3wDckLCIJKVmHTiatbVN/OGZDxhenMtVM6cQa9CJZIxn3v2YiMPxew/nuferCUc86cswdGdL7GYwg/P7wTUA6f8uP3YSF8wYzy0vruT3Ty9TS0AyzpPvbGBcSQGf27OMiMOmNF4Xq+tLLQAZ+MyMq0/bl1Akwo3PryArEOB7J+yV7rBEUqKuuZ1Xl2/k4iMmMLw4F4h2A5XFnqdabcO2FoASgKRIIGD88ov7EQo71/7fvwhHnO+fsBeBNDWDRVLlufc/JhRxPj915PbWbzpHAtU2tjK0IJvsYPI7aJQAZLtAwPj1GfsTMOOG55fz/oat/OGsA1PyS0QkXZ5csoGRg/I4cMwQqjZH76CXzgvBtQ1tlBalpvWhawDyCcGA8esz9uPq0/Zl3rIaZt3wMss21Kc7LJGkaGoL8cIHNXx+3xEEAsaw4uiF17QngMLkXwAGJQDpgplx4eHlPDh7Bo1tYWbd+DI3Pr+ctpBuLSkDS+XqzbSGIhy/zwgACnKyKMrNorq+JW0xbWxsZZhaAJJuFeUl/PM7R3DUXmX87qllnHTti7yyfGO6wxJJmA110S/68tLC7fvKinPT3wIoUgtA+oDhg/K45YIK/nLRIYTCznm3LeR7Dy1O6zA5SaxMHva77Zd+xxE/ZUXpSwBtoQh1ze2UFqoFIH3IMVOG8/T3juSKYyfx2FvrOeGPL/A/i9dl9JfHQODufGXOAn7417cIRzLv/2V1fSuD8rLIyw5u31c2KDdto4A2xyaBbbsWkWxKABK3vOwg3z9xMo9fcQRjhubz3QcXc+bN87nj5VV8uKkp3eHJLpi/opbXVm3ikUVV/NvfM285kJr6VoYPyvvEvnS2ADbGEk+qWgAaBiq9NmXkIB791me5e/5qHnztQ655/F2uefxdJo8oZp89BjFpeBF7jShmenkJg1Mwm1F23X2vrWVwfjZnHzKWW15cSXFeFj89ee+MWQ6kur51++SvbcqKc6lvCdHSHv5EyyAVtk0CG5aiawBKALJLggHja5+dwNc+O4HVGxt55t2PeWn5RhasrOXvb64DICtgHD5pGCdPHcnn9x3J0BQNbZP4bGxo5emlG7hgRjlXzZxCS3uYW19aRXvYOWqvMsaW5DNmaEHKvwRTqbq+hWnjhn5iX1mH2cBjSwpSGs/2FkCKRgEpAchuKx9WyDeOnMg3jpwIQH1LO+9vqOfZ9z7mySUbuOrRJVz92LtceHg5lx45UYmgj/hrZRXtYefcQ8diZvz81H1pbg9z56urufPV1dvLjRyUR/mwAspLCxlfWsj40gLGlRQwvrSA4rz+28Jzd6q3dtEFFEsA1WlIANtaAKkaBaQEIAlXnJfNIeUlHFJewlUnTWHp+q3c+tJKbnlxBfcuWMN5M8axz6hBlBXnMmJQHkPysynOyyYnS5ekUiUScR54bS2HTihh0vBiIDoT/LdnHsAPT5zMh5ubqNrczNraJlbXNrG6NtrKq+00+mtwfjZjhuYzZmg+w4vzKC3KYVhRLsftPZxRg/PT8dHitrUlRGso8ukuoKIdLYBU29jYSk4wQHFuar6alQAkqcyMqaMHc+3ZB/HtYybxp2c+4JYXVnZZNicr+odfkBukMCeL/JwgRblZFOQEKYz9Ny8rSH5OkLzsIIU5QQo67M/NDpCbFSQraATNCJiRFTQKc6LHLMgJkh0MkBWw3e7jbmkPU7W5mfycIKWFOf2um+Tl5RtZu6mJH5z46UX/hg/KY/igPKaN//T76lvaWVPbxNpN0UdVLFGsrGlk4apN229mcttLBTz9vaP6dFKv6WIIKMDwQbEEkIaRQNvmAKTqGowSgKTMXiOKuen8adS3tPPx1haqt7ZSXd9KXXM79S3t1LeGaGgJ0dwWprEtRGNrmIbWENVbW2lojV6Ua2kP09QeZncHq2QFjOxggOygkZMVICsQIDvLyA4EyApG//FFPNpNkB0MkJ8TJD87SMSdNbVNfFT3yZmihbEklRUwAoFo8glHnFAkQjjilBTmsMeQfEYNzmfU4DzKinMZXpzLsKJcCnOD5OdkUZAdTWJZgWhcu/sl0NQWAiA/O/ipY92/cC0lhTmcNHVkr45ZnJfN1NGDmTp6cJevt4cjPPPux3zrvje4f+EaLvrshF0LPgWqt0a/4IcXf7ILqLQwl0CabgxT29Casu4fUAKQNCjOi3b5bOt66C13pzUUoaktTGNriKa2MK2hMK2hCC3tYUIRx90JR6JfSE1tYZraouXaQxHaI04oHCEUcdpCEdrDEdpC0e32cIRQ2DEj+sBoD0dobg/T3Ba9beZhE0spH1bImKH5tIYibGpso7ahjaa2EOGIE444EXeCHb7Iaxta+aiuhber6uKeRBcMxFoyAQiaEYwlraygkRUIkJMVPX5WLGltu4HJpsY2aupbaYrFawYF2dEENSg/m+K8LN6uquPiIyaQm5XYlkt2MMDMqSM5/DOlXPt//+JLB4/ps4sJVse+4Du3AIIBo6Qwd3sLIZVqG9tSNgQUlACkHzIz8rKj3UAl/fCCcmsozMaG6Jd0bUMrjW1hmmMtnvYOiSkcccLuRCJOKJZYtiWo9kiE9nA0kbWHI7HWhuMOY4cWUBZrXZhBU2uIxrYwDS0h6lvb2doc4pDyoVx4eHlSPp+Z8dOT9+bUG17mpnkruGrmlKScZ3dtmwW8rcuno3QtB1Hb0Mak4UUpO1+PCcDM7gBOAardfWoXrxtwLXAy0ARc5O5vJDpQkYEiNyvI6CH5jB7Sty+S7o6powfzpYNGc8crqzh/xjjGDE3taJp4VG9tJS+76wuu6UgA7s7GhtQtBAfxzQS+EzhpJ6/PBPaMPWYDN+1+WCLS3/3wxMkY8Ksn3qelPZzucD4lOgksr8trLemYDdzYFu3GTNUkMIjvpvAvmln5TorMAu726BzyBWY2xMxGuftHCYpRRPqhPYbkc+mRE7nuueU8v6yaoyeXceI+Izlp6sg+MWqqur7lU0NAtxkeWw/I3VM2Iqc2xctAQGKuAYwGPuywXRXbpwQgkuGuPH4vppWX8PTSDTzz7sc8sWQDo5/K5ycnT+EL+41K65IT1fWtTBnZ9UCEsqJc2sPOlqb2lE1c3JjiSWCQmMXguvo/2OUgPTObbWaVZlZZU1OTgFOLSF8WCBhH7VXGL7+0Hwt+chx3f306xXlZXH7/m3z55vm8s64ubbHVbG391BDQbbYvB5HCuQDbloHoa9cAelIFjO2wPQZY31VBd5/j7hXuXlFWVpaAU4tIfxEIGEfuVcY/r/gcvz59P1bXNnL6n1/lodfXpjyW5rYw9a2hTw0B3abjekCpkuplICAxCWAu8FWLmgHUqf9fRLoTDBhnTx/HM987iukTSvjx35bw078voTWUugvF24eA9qkEED1XKoc2xzMM9AHgaGCYmVUBPweyAdz9ZuAJokNAlxMdBvq1ZAUrIgPH0MIc7vr6dH731DJufmEFC1fWMrakgEBs0tuUkcXMmFjKtPFDE37ReNsksM4LwW2TlgTQ2EZxXlbCJ+ftTDyjgM7p4XUHvp2wiEQkYwQDxlUzp3DAmMHc9vIqNjW2EfHoRLjn3q/m+ueWkxMMUFE+lOP2HsEJe49gXOnuzynYsQxE1y2A4tws8rIDKb8GkMr+f9BMYBHpA2buN4qZ+436xL76lnYqV29m/spa5i2r5j8ef5f/ePxdJg0vYsbE6Gqz0yeU7NKqoz11AZkZZcW5VG9N3XIQtQ1tlKZ4ZrsSgIj0ScV52RwzZTjHTBnOT0/emzW1jTz7XjXzllXz9zfWce+C6MXjc6aP45pZ+5IdjP+SZnV9K1kBY2hB91+4ZUWpvTdwbWMrE4YVpux8oAQgIv3E+NJCLj5iAhcfMYFQOML7G+p59I113PHKKj6qa+bGcw+mMM519GvqWykrziUQ6H4eQllxLqs2NiYq/B7VNrRRUV6SsvOBbgovIv1QVjDA1NGD+dmp+/BfX9qPFz+o4Stz5m/v2ulJV/cC7mz0kAJW1zaxbEN9IkLeqXDE2VF+QLUAAAj+SURBVNTUxrAUdwEpAYhIv3buoeO47cIKVlQ3cuZN8/lwU1OP76ne2kJZN5PAtpl95EQG5WUz+55K6mI3ukmWzU1tuMOwHpJSoikBiEi/d+yUETwwewZbmto4e84CVvfQdVNT39rlMtAdjRycx83nH8z6Lc1c8eCbhCO7eReindiYhnWAQAlARAaIA8cO4f5vzKCpLcRX5sxneXVDl+XawxFqG9t67AICqCgv4erTpvLCBzX8/ulliQ55u3TMAgZdBBaRAWTq6ME8OPswzrttAaf/+RX22WMQpUW5lBXlcv6M8UwaXrT913Z3y0B0du6h41iyro6b5q0gaMaVx+9JVi9GHMVjxzpAugYgIrLLJo8s5qFLD+OoycMJR5z31m/lgdfW8rU7X2NrS3u39wLematP25ezKsZww/PLOf/2hQmfH7C9BZDiLiC1AERkwPlMWRHXn3PQ9u1FazZx1i0L+Pe/v8OpB+wBdD8JrCs5WQF+e+YBTJ9Qyr//YwknX/cyfzjrAI7aKzGLWtY2thIMWMrvn6wWgIgMeNPGl3DlcXsy96313PrSSqDrewH35MxpY5h7+REMKcjmwjte43sPLWZTY9tux1fb0EZJYc5O5yUkgxKAiGSEbx0ziUMnlPDaqk2Y7fq6+3uNKObx7xzBd46dxGNvree4P8zj3gVrdqtbaGMaloEAsOhabqlXUVHhlZWVaTm3iGSmj+qamXntSwTNWPT/Ttjt4y3bUM9PHn2bN9ZuAdi+TtHQghyygwGygwG6u17cHnbW1jbxr+p63lm/lUMnlHDPxYf2eE4zW+TuFbsdPLoGICIZZNTgfG6/8BDWb2lOyPEmjyzmkcsOZ+n6rby6YiOvrKjlH2+up6E1FNf7Swtz+MzwIs6cNoYzp41JSEy9oRaAiEiCuTvhiNMediLdfMcGzMjP6f3a/2oBiIj0YWZGVtBI4b1ddokuAouIZCglABGRDBVXAjCzk8xsmZktN7Orunj9IjOrMbPFsccliQ9VREQSKZ6bwgeBG4ETgCrgdTOb6+7vdir6kLtfnoQYRUQkCeJpAUwHlrv7SndvAx4EZiU3LBERSbZ4EsBo4MMO21WxfZ2dYWZvm9kjZjY2IdGJiEjSxJMAulqcovPA1seAcnffH3gWuKvLA5nNNrNKM6usqanpXaQiIpJQ8SSAKqDjL/oxwPqOBdy91t1bY5u3AtO6OpC7z3H3CnevKCtLzCp6IiKya+JJAK8De5rZBDPLAc4G5nYsYGajOmyeBryXuBBFRCQZehwF5O4hM7sceAoIAne4+1IzuwaodPe5wBVmdhoQAjYBFyUxZhERSQCtBSQi0o8kci0gzQQWEclQSgAiIhlKCUBEJEMpAYiIZCglABGRDKUEICKSoZQAREQylBKAiEiGUgIQEclQSgAiIhlKCUBEJEMpAYiIZCglABGRDKUEICKSoZQAREQylBKAiEiGUgIQEclQSgAiIhlKCUBEJEPFlQDM7CQzW2Zmy83sqi5ezzWzh2KvLzSz8kQHKiIiidVjAjCzIHAjMBPYBzjHzPbpVOxiYLO7TwL+BPwm0YGKiEhixdMCmA4sd/eV7t4GPAjM6lRmFnBX7PkjwHFmZokLU0REEi0rjjKjgQ87bFcBh3ZXxt1DZlYHlAIbOxYys9nA7Nhmq5m9sytBd2MwUJfg8t2V6Wp/PPs6bnd8PoxOdbWbEl0XO3tddbHz11QXvdunuuh6u+PzyT0FGzd33+kD+DJwW4ftC4DrO5VZCozpsL0CKO3huJU9nbs3D2BOost3V6ar/fHs67jd6Xmfroudva66UF2oLvpvXcTTBVQFjO2wPQZY310ZM8simq02xXHsRHosCeW7K9PV/nj2PbaT1xIp0XWxs9dVFzt/TXXRu32qi663k1IXFsso3ReIfqF/ABwHrANeB85196Udynwb2M/dLzOzs4HT3f2sHo5b6e4Vu/sBBgLVxQ6qix1UFzuoLnZIZF30eA3Ao336lwNPAUHgDndfambXEG2KzAVuB+4xs+VEf/mfHce55+xG3AON6mIH1cUOqosdVBc7JKwuemwBiIjIwKSZwCIiGUoJQEQkQykBiIhkqD6ZAMwsYGa/NLPrzezCdMeTTmZ2tJm9ZGY3m9nR6Y4n3cys0MwWmdkp6Y4lncxs79jfxCNm9s10x5NOZvZFM7vVzP7HzE5MdzzpZGYTzex2M3sknvIJTwBmdoeZVXee5dvTgnKdzCI6u7id6ByDfilBdeFAA5CH6gLgx8DDyYkyNRJRF+7+nrtfBpwF9NvhkQmqi3+4+zeAi4CvJDHcpEpQXax094vjPmeiRwGZ2ZFEv7DudvepsX1BonMJTiD6JfY6cA7RYaW/6nSIr8cem939FjN7xN3PTGiQKZKgutjo7hEzGwH80d3PS1X8iZSgutif6JIAeUTr5fHURJ9YiagLd682s9OAq4Ab3P3+VMWfSImqi9j7/gDc5+5vpCj8hEpwXcT1vRnPWkC94u4vdrEc9PYF5WLBPQjMcvdfAZ9qyptZFdAW2wwnOsZUSURddLAZyE1GnKmQoL+LY4BCoqvSNpvZE+4eSWrgSZCov4vYHJy5ZvZPoF8mgAT9XRjwa+DJ/vrlDwn/vohLwhNAN+JZUK6jR4HrzexzwIvJDCwNelUXZnY68HlgCHBDckNLuV7Vhbv/G4CZXUSsZZTU6FKrt38XRwOnE/1R8ERSI0u93n5ffAc4HhhsZpPc/eZkBpdivf27KAV+CRxkZj+JJYpupSoBdLU0dLd9T+7eRPQeAwNRb+viUaIJcSDqVV1sL+B+Z+JDSbve/l3MA+YlK5g0621dXAdcl7xw0qq3dVELXBbvwVM1CiieBeUyhepiB9XFDqqLHVQXOyS1LlKVAF4H9jSzCWaWQ3StoLkpOndfo7rYQXWxg+piB9XFDkmti2QMA30AmA9MNrMqM7vY3UPAtgXl3gMe7ria6ECluthBdbGD6mIH1cUO6agLLQYnIpKh+uRMYBERST4lABGRDKUEICKSoZQAREQylBKAiEiGUgIQEclQSgAiIhlKCUBEJEMpAYiIZKj/D5EDdm9yL5pDAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.semilogx(history.history[\"lr\"], history.history[\"loss\"])\n",
    "plt.axis([1e-6, 1e-1, 0, 4])\n",
    "plt.title('Training Loss');"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYAAAAEMCAYAAADNtWEcAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3dd5xU9b3/8ddnZhuwCwgsvUkHFduqWIMtwRJNsRujRkLQqMmNuUlu7k1MvKk3iRpb1Niwa9SHQaO/WII1tgUVpIqKAtLbUrZM+fz+mNnisuvOwpSzO+/n47EPZ8757pkP32zOZ771mLsjIiL5J5TrAEREJDeUAERE8pQSgIhInlICEBHJU0oAIiJ5SglARCRPKQFI4JmZm9mo5OubzexnqZTdhc8518ye2dU4RToaJQDJODP7p5ld1cLxU81stZkVpHotd5/u7v+bhpiGJ5NFw2e7+33u/sXdvXYLnzXZzFak+7oiu0sJQLLhLuA8M7Nmx88D7nP3aPZDEhElAMmGx4FewJH1B8xsD+Bk4G4zO9jMXjOzzWa2ysxuMLOili5kZneZ2a+avP/P5O98ambfalb2JDN728yqzGy5mf2iyemXkv/dbGbbzOxQM7vAzF5p8vuHmdlbZrYl+d/Dmpx7wcz+18xeNbOtZvaMmfVpb8WY2fjktTab2XwzO6XJuRPNbEHy+ivN7IfJ433M7Mnk72w0s5fNTP9flnbTH41knLtXAw8D32xy+Axgkbu/C8SA/wD6AIcCxwKXtHVdM5sC/BA4HhgNHNesyPbkZ/YETgIuNrOvJM8dlfxvT3cvdffXml27F/AP4DqgN3A18A8z692k2DnAhUBfoCgZS8rMrBB4AngmeY3LgPvMbGyyyO3Ad9y9DNgb+Ffy+BXACqAc6Af8FNCeLtJuSgCSLTOA082sS/L9N5PHcPfZ7v66u0fdfRlwC/CFFK55BnCnu7/n7tuBXzQ96e4vuPs8d4+7+1zggRSvC4mE8b6735OM6wFgEfDlJmXudPclTRLcfileu94koBT4nbvXufu/gCeBs5PnI8AEM+vu7pvcfU6T4wOAYe4ecfeXXZt6yS5QApCscPdXgHXAqWY2AjgIuB/AzMYkuzRWm1kV8BsSrYG2DASWN3n/cdOTZnaImc0ys3VmtgWYnuJ166/9cbNjHwODmrxf3eT1DhI38/YYCCx393grn/F14ETgYzN70cwOTR7/A7AUeMbMPjSzn7Tzc0UAJQDJrrtJfPM/D3jG3dckj/+FxLfr0e7enUSXRvMB45asAoY0eT+02fn7gZnAEHfvAdzc5LptfWP+FBjW7NhQYGUKcaXqU2BIs/77hs9w97fc/VQS3UOPk2hl4O5b3f0Kdx9BokXyAzM7No1xSZ5QApBsuptEP/23SXb/JJUBVcA2MxsHXJzi9R4GLjCzCWbWFbiy2fkyYKO715jZwST67OutA+LAiFau/RQwxszOMbMCMzsTmECii2aXmFlJ0x/gTRLjFD8ys0Izm0zihv6gmRUl1yX0cPcIifqJJa9zspmNSs6qqj8e29W4JH8pAUjWJPv3/w10I/HNvN4PSdyctwJ/BR5K8XpPA9eSGBxdSuMgab1LgKvMbCvwc5LfoJO/uwP4NfBqcjbNpGbX3kBiltIVwAbgR8DJ7r4+ldhaMAiobvYzBDgFOAFYD9wEfNPdFyV/5zxgWbJbbDrwjeTx0cBzwDbgNeAmd39hF+OSPGYaOxIRyU9qAYiI5KmUE4CZhZOLanbqAzWzYjN7yMyWmtkbZjY8nUGKiEj6tacF8D1gYSvnLgI2ufso4Brg97sbmIiIZFZKCcDMBpNYGHNbK0VOpXFWxyPAsS3s+yIiIgGSagvgWhKzIOKtnB9EckFOcmOvLSSWz4uISEC1uQ2vmZ0MrHX32cl5yi0Wa+HYTtOLzGwaMA2gW7duB44bN64doYqIyOzZs9e7e3k6rpXKPuyHA6eY2YlACdDdzO519280KbOCxJzmFcn91XsAG5tfyN1vBW4FqKio8MrKyt2NX0Qkr5hZ8y1KdlmbXUDu/l/uPtjdhwNnAf9qdvOHxKKe85OvT0uW0QIDEZEAS/lJTM0ln/BU6e4zSWxbe4+ZLSXxzf+sNMUnIiIZ0q4EkFxu/kLy9c+bHK8BTk9nYCIikllaCSwikqeUAERE8pQSgIhInlICEBHJU0oAIiJ5SglARCRPKQGIiOQpJQARkTylBCAikqeUAERE8pQSgIhInlICEBHJU0oAIiJ5SglARCRPKQGIiOQpJQARkTylBCAikqeUAERE8lSbCcDMSszsTTN718zmm9kvWyhzgZmtM7N3kj9TMxOuiIikSyrPBK4FjnH3bWZWCLxiZk+7++vNyj3k7pemP0QREcmENhOAuzuwLfm2MPnjmQxKREQyL6UxADMLm9k7wFrgWXd/o4ViXzezuWb2iJkNSWuUIiKSdiklAHePuft+wGDgYDPbu1mRJ4Dh7j4ReA6Y0dJ1zGyamVWaWeW6det2J24REdlN7ZoF5O6bgReAKc2Ob3D32uTbvwIHtvL7t7p7hbtXlJeX70K4IiKSLqnMAio3s57J112A44BFzcoMaPL2FGBhOoMUEZH0S2UW0ABghpmFSSSMh939STO7Cqh095nA5WZ2ChAFNgIXZCpgERFJD0tM8sm+iooKr6yszMlni4h0VGY2290r0nEtrQQWEclTSgAiInlKCUBEJE8pAYiI5CklABGRPKUEICKSp5QARETylBKAiEieUgIQkbzl7rz50UZytSA215QARCRvvbVsE2fc8hrzP63KdSg5oQQgInmrqjqS+G9NJMeR5IYSgIjkrbpYHIBITF1AIiJ5JZJMAHXReI4jyQ0lABHJW/U3/vpEkG+UAEQkb9V3/agFICKSZxq6gNQCEBHJL5GYuoBERPJSnQaBRUTyUySaGANQC6AVZlZiZm+a2btmNt/MftlCmWIze8jMlprZG2Y2PBPBioikk6aBtq0WOMbd9wX2A6aY2aRmZS4CNrn7KOAa4PfpDVNEJP0aB4G1EKxFnrAt+bYw+dO8tk4FZiRfPwIca2aWtihFRDKgNuDrALbWRPjBw+/w7w/WZ+T6KY0BmFnYzN4B1gLPuvsbzYoMApYDuHsU2AL0buE608ys0swq161bt3uRi4jspqB3AVXVRHlszkqWb9yRkeunlADcPebu+wGDgYPNbO9mRVr6tr9Tm8rdb3X3CnevKC8vb3+0IiJpFPRpoDWRGAAlheGMXL9ds4DcfTPwAjCl2akVwBAAMysAegAb0xCfiEjGBH0lcHVdjhOAmZWbWc/k6y7AccCiZsVmAucnX58G/Mvz9QkLItJh1AV8JXBtNLMJoCCFMgOAGWYWJpEwHnb3J83sKqDS3WcCtwP3mNlSEt/8z8pItCIiaRSJBns76Oq6RHxdcpUA3H0usH8Lx3/e5HUNcHp6QxMRyazGQeBYjiNpWeMYQGbW7GolsIjkrfpv/oFtASQTQKZaAEoAIpK3gr4XUKBmAYmIdCZB3w5aCUBEJEOC/kSwmkgiLo0BiIikWdBXAlerBSAikhmNg8DBTAA1kRgFIaMwrBaAiEha1X/zD3ILIFMzgEAJQETyWONeQMGcBloTiVOsBCAikn4dYRZQl6LM3aaVAEQkbwV9M7iaSIySArUARETSri7g20FXR2J0KeqECWDBqip++Ld3+ef81Q1bnoqIZIu7B34aaKZbAKnsBpoRZcUFPDN/NY/MXkFxQYjDRvbmmPH9OGZcXwb17JKrsEQkT8TijjuEQ0Y07sTjTigUrCfZVkfi9OhSmLHr5ywBDOnVldd+djxvfbSRZxasYdbitcx6/D1+BozuW8qRo8s5akwfDtmzd0abQCKSn+r7/7sVhamqiVIXi1MSCta9pjYSo6SsOGPXz1kCACgMhzhsVB8OG9WHK30CH67fzqxFa3lxyTrufeNj7nj1I4oKQhw+sjfHju/HseP7MqCHWgcisvvqu31KiwuoqokSicUztuJ2V2V6DCCnCaApM2NkeSkjy0uZeuQIaiIx3vxoI7MWr+X5hWuZtfg9/udx2GdQD44b34/jJ/Rj/IAyzILVZBORjqF+ALhrceI2GMS1AJ12DKAtJYVhjhpTzlFjyvn5yRNYunYbzy5cw3ML1nDt80u45rklDOvdlVP3Hcip+w9iZHlprkMWkQ6kfgC4W/IbdhAHgqvr8qQF8HnMjNH9yhjdr4xLJo9i3dZanl+4hifmfsr1s5Zy3b+Wsveg7nxxQn+OG6+WgYi0rSEBNLQAgpcAaqJxijO0EyikkADMbAhwN9AfiAO3uvufm5WZDPwd+Ch56DF3vyq9oTYqLyvmrIOHctbBQ1lTVcMT737KP+at4prnlnD1s0sY1LMLx0/oxwl796dieC/CARvZF5Hca54AagPWAojFnbpoPKN7AaXSAogCV7j7HDMrA2ab2bPuvqBZuZfd/eT0h/j5+nUvYeqRI5h65AjWbq1h1qK1PLtgDfe/+Ql3/XsZfUqLOGmfAZxzyDDG9i/LdngiElB10USff2lAWwC10cw+DhJSeyj8KmBV8vVWM1sIDAKaJ4Cc61tWwpkHDeXMg4ayrTbKC4vX8vS81Tzw1nJmvPYxFcP24NxJQzlxnwEUZ3BgRUSCr/6G3zXZxx60BFC/QDaTM5PaNQZgZsOB/YE3Wjh9qJm9C3wK/NDd57fw+9OAaQBDhw5tb6ztUlpcwMkTB3LyxIFs3F7Ho7NXcP+bn/AfD73Lr55cyJkHDeHcScO06EwkTzXvAgraIHBNMp5cdwEBYGalwKPA9929qtnpOcAwd99mZicCjwOjm1/D3W8FbgWoqKjI2pyrXt2K+PZRI5h65J68unQDd7+2jJtf/ICbX/yAo8aU8/UDBnP8hH6BmwMsIplT1zALqOAz74OivgWQ00FgADMrJHHzv8/dH2t+vmlCcPenzOwmM+vj7uvTF+ruMzOOGN2HI0b3YcWmHTz45nIenbOCyx54m7KSAk4/cAjfPXokvUszt/JORIKhYSVwcTCngdY/ED6nD4SxxHzK24GF7n51K2X6J8thZgcnr7shnYGm2+A9uvLDL43llR8fw31TD+HYcX2Z8doyjvq/Wfz5uffZXhvNdYgikkH1N/xuAV0IVpPh5wFDai2Aw4HzgHlm9k7y2E+BoQDufjNwGnCxmUWBauAsdw9WbbYiHDIOH9WHw0f14dJjRvPHfy7mmueWcM/ry7j82NGcddBQigq0a7ZIZxP0QeCaSHIMIJcLwdz9FeBzJ9K7+w3ADekKKldG9S3l5vMOZM4nm/jd04v4+d/nc9vLH3HFF8fw5YkDA7dToIjsukjzMYCAdQFV17cA9ECY7Dpg6B48NG0Sd15wEF2LwnzvwXc48bqXeXbBGjpIw0ZE2tC8Cyhwg8D1YwB6JGT2mRlHj+vLU5cfyZ/P2o+aSIxv313JV258lVeXBmpsW0R2QX2ff2lQp4EmE0Am1ywpAbQhFDJO3W8Qz/3gC/zfaRNZv62Oc297g/946B02bKvNdXgisosa1wEEdQygvgWgBJBzBeEQZ1QM4fkrvsDlx47mybmfcuzVL/LI7BXqFhLpgIK+GVw2ZgEpAbRTSWGYHxw/hqcuP5JR5aX88G/vctGMStZW1eQ6NBFph7pms4CC1gVUXZeIpySDsxCVAHbR6H5lPPydQ7nyyxN4del6vnjtSzw599NchyUiKYokN4Pr2rASOFgt+ZpojMKwURBWAgikUMi48PA9+cflRzKsdzcuvf9tLr53NmvUGhAJvEgsTjhkhENGUTgUwBZALOPb0ygBpMGovqU8Ov1Q/vNLY3l+0VqO+9OL3PP6x8TjwfpGISKN6mJxCsOJtT1FBaHAjQHURpUAOoyCcIjvHj2KZ75/FBOH9OBnj7/HGbe8xicbduQ6NBFpQV00TmGye6UwbIFLANV1sYzuAwRKAGk3vE837r3oEP50+r4sXrOVE697mUc1U0gkcCKxOEUNCSB4XUA1kTglGdwJFJQAMsLM+PqBg3n6e0cyYWB3rvjbu1z6wNts3lGX69BEJCkSa2wBFBWEArkSWC2ADmzwHl154NuT+NGUsfzzvdV86dqXePn9dbkOS0RIrAQuLEiOAQSyBRCjWAmgYwuHjEsmj+Lx7x5OWUkh593+Jlf+/b2Ghz2ISG7UNekCCuIgcI1aAJ3H3oN68ORlR3Dh4cOZ8drHfPWmV/lo/fZchyWStyKfGQQOBfB5ABoD6FRKCsNc+eW9uOvCg1hdVcMp17/C/3tvda7DEslLkVi84VkfhWELXBeQxgA6qclj+/LkZUcworwb0++dzW+fXkhMawZEsioS80APAtdEtA6g0xq8R1cenn4o5x4ylFte/JDp985mR50eQymSLU0XggVxGmi1EkDnVlwQ5tdf3YdffHkCzy9cwxm3vKZtJESypOlCsOIADgLXRuK5TwBmNsTMZpnZQjObb2bfa6GMmdl1ZrbUzOaa2QGZCbdzuuDwPbnt/Ao+XLedr9z4KotXb811SCKdXvOFYEFKALG4UxeLB2IMIApc4e7jgUnAd81sQrMyJwCjkz/TgL+kNco8cMy4fvxt+qHE4s4Zt7zG7I835TokkU6t6UKwoHUBNT4LIMezgNx9lbvPSb7eCiwEBjUrdipwtye8DvQ0swFpj7aT22tgDx69+DB6di3kG7e9wQuL1+Y6JJFOK7EQrOk6gOBMxKjOwtPAoJ1jAGY2HNgfeKPZqUHA8ibvV7BzksDMpplZpZlVrlunFbEtGdKrK49MP4zhfboxdUYlf39nZa5DEumUEmMAjYPAtUFsAWTwecDQjgRgZqXAo8D33b2q+ekWfmWndOrut7p7hbtXlJeXty/SPFJeVsxD35nEAcP24HsPvsOtL32gzeRE0qzpGEDQBoEbEkAQWgBmVkji5n+fuz/WQpEVwJAm7wcDejzWbuheUsjd3zqYkyYO4DdPLeKXTyzQWgGRNAryQrCaSCKWnA8Cm5kBtwML3f3qVorNBL6ZnA00Cdji7qvSGGdeKikMc/1Z+3PREXty17+Xcen9cxq+GYjI7mm6ECxos4CqszQIXJBCmcOB84B5ZvZO8thPgaEA7n4z8BRwIrAU2AFcmP5Q81MoZPzs5AkM6FHCr59ayNrb3uC2b1awR7eiXIcm0qHVNdsOOhp34nEnFGqpRzu76r/oZboF0GYCcPdXaLmPv2kZB76brqBkZ1OPHMHAnl34/kPv8PW//Ju7LjyYob275joskQ7J3ZNjAI2DwJBICiWhzN50U1G/W3DOF4JJcJy4zwDum3oIG3fU8dWbXuW9lVtyHZJIhxSLO+58ZiUwEJhuoJrkeIQSgHzGQcN78ejFh1FSGObsW1+nctnGXIck0uHUb/xWWNA4BgAEZiC4pi4gC8EkeEaWl/K36YdSXlbMebe/qaeMibRTJJqYUdd0EBgIzGKwmmh2xgCUADqogT278NB3DmVY765cdFclzy9ck+uQRDqM+hZA/RhAUcC6gDQGIG0qLyvmwWmTGDegjEvvf1tjAiIpqr/RN7YAEokgKKuB69cBKAHI5+rZtYjbzq+gV7ciLprxFqu3aDtpkbY0TwBBGwSujsQoCocIZ3hKqhJAJ9C3rITbzq9gW02UqXe/pQfLiLSh/kZfFNRB4Egs4wPAoATQaYwf0J3rz9mfBZ9WccXD72rvIJHPUdfqIHCQEkDm1yMoAXQix4zrx4+mjOPp91bzz/l62LxIaxpbAJ8dBA7Kc4FrIrGMbwUNSgCdztQj9mRsvzJ+9Y+F2jdIpBU7DwIHqwuoOhLL+FbQoATQ6RSEQ1z55Qms2FTNbS9/mOtwRAKprtVB4GB0ndZE4hnfChqUADqlw0b1Ycpe/blx1ges2lKd63BEAqf+m36wWwAaBJZd9N8njSfmzu+eXpTrUEQCp/6bflGzdQBBGQSu1RiA7I4hvbrynaNG8Pd3PuXVpetzHY5IoDSMAQR0EFhjALLbLp48klF9S7nsgbdZvnFHrsMRCYzmg8BFAesCqonE1QKQ3dO1qIC/frOCaCzOtHtma4GYSFL9jb6oyQNhIDhdQNVaCCbpsGefblx/zgEsXl3Ff/5trhaIidBkDCCoK4HrtBBM0uQLY8r5yQnj+Me8Vdz0wge5Dkck51pbBxCUFkBNNCAJwMzuMLO1ZvZeK+cnm9kWM3sn+fPz9Icpu+vbR47g1P0G8sdnFvPK+xoUlvzWmADsM/+tC8A6gGgsTiTmGX8WAKTWArgLmNJGmZfdfb/kz1W7H5akm5nx26/tw+i+pVz+4NtaHyB5rflCMDOjKBwKRBdQ4+MgAzAG4O4vAXruYCfQtaiAv3zjQGojMb5735xA/LGL5ELzJ4JBYjwgCF1A9Q+DCUoLIBWHmtm7Zva0me3VWiEzm2ZmlWZWuW6dHmOYCyPLS/m/0/Zlzieb+e3TC3MdjkhO1MVihEP2mf32C8MWiC9F9Xt4FXeQBDAHGObu+wLXA4+3VtDdb3X3CnevKC8vT8NHy644aeIALjx8OHe+uoyn5q3KdTgiWReJeUO/f73CcDBaAPUJoEO0ANy9yt23JV8/BRSaWZ/djkwy6r9OGM9+Q3ryo0fmsmz99lyHI5JVddH4Z7p/INEFFISVwNl6HCSkIQGYWX8zs+Trg5PX3LC715XMKioIceO5B1AQNi6+b462jpa8EonFGxaB1QvKIHB1kFoAZvYA8Bow1sxWmNlFZjbdzKYni5wGvGdm7wLXAWe5Vht1CIN6duHqM/Zl4aoqfvnE/FyHI5I1kVjLLYAgdQFlYxZQQVsF3P3sNs7fANyQtogkq44Z149LJo/kphc+YHjvbnznCyNzHZJIxkVi3rAKuF5hwFoA2egCajMBSOf3g+PHsGzDdn779CLWba3lpyeOJxSytn9RpIOqi8VbGAS2QDwQpkYJQLKpIBzi+rMPoLx0Pre98hFrt9byh9MnUpyF7WhFciES6EHg5BhAFnYDVQIQAMIh4xen7MWAnl343dOL+Gj9dv7npPEcMqJ3rkMTSbtILN5iF9DWmtzvmFs/CygQg8CSP8yM6V8YyV/OPYD122o589bXmTrjLd5fszXXoYmkVWIdwGdvf8UBGQRetaWGwrBRWpz57+dKALKTE/YZwKwfTubHU8bxxocbmfLnl7nm2SWB+D+HSDok1gHsvBAsCIPAc1dsZlz/7ju1UDJBCUBaVFIY5uLJI3nxR0dz6r4D+fPz7/PVm15liVoD0gnUtTANNAgrgeNxZ96KLUwc3CMrn6cEIJ+rV7cirj5zP27+xgGs2lzDyde9wiX3zea+Nz5m2frtesBMJ7BodRVvfrSRLTsiuQ4la1pcCFYQyvksoI82bGdrbZR9B/fMyudpEFhSMmXvAVQM78XVzy5h1qK1PDVvNQA9uhTSv3sJ/XqU0K+smP49SujbPfH6kBG96dGlMMeRy+dZubmar97474a55/26FzO2f3cmDOjOhIHdmTCgjKG9umWlOyKbWloIVhgOUZvjLqC5KzYDMHFIdloASgCSsj6lxfzmq/vg7ny4fjv/XrqeJWu2sbqqhrVVNSxeXcW6rbXEk1+iuhaFOf3AwVxw+J7s2adbboOXFv1iZmIF+A3n7M/KTdUsXrOVhau2cvsHHzZ8Gw4ZDN6jK8N6d2Vgjy7s0a2IXt0K6d0tkfD7dS+mf48uWRm0TJdIzCksCN4g8LvLt9ClMMyo8tKsfF7H+V9MAsPMGFleysgW/khjcWfDtlqWbdjBQ28t54E3l3P36x8zeUw535g0jMlj+35mC17JnecWrOHZBWv48ZRxnDxx4GfO1UXjLF27jYWrqli2YTvLNuxg2frtLFmzlY3b61rsKjl0RG/OOWQoX9yrX+DXkNRFd+4CCsJ20HNXbGbvQd0pCGenxaUEIGkVDhl9uye6gQ7esxc/PmEs977+CQ+8+QkXzahkYI8SvrL/IHqXFlNUEKI4HKKkKEzXwjBdi8IUF4YpCBkFYaMwHEr+GEUFIQpCoYY93AtCRnFBiOQ+hIHl7sSdwCW96roYv3hiPqP7lnLREXvudL6oIJToAhrYfadz7s622ijrt9WxpqqGNVU1fLB2G4+9vZLLHnibXt2K+O8Tx/P1Awdn45+ySxLrAIK1HXQkFmf+p1V8Y9KwrH2mEoBkVN+yEn5w/BguO2YUzy9cw31vfJLWB9MXF4QoTjbl3cGBxNa0ELJEsiguCFFUEKIoHGq4EYfMCIUgbEYoZISTZZs/JKReOFkm7s6W6ghbqiNU1URxT3SR1P9OLO7E3YnGndpInNpojLjDiPJuHDqiN4eO7M3YfmWUFIYbYgrVf64ZTiJhxN0JmVEYNgpDoXZvzbG9Nsq6rbWEzOjepYCyksLP/Luu/9f7rNhUzUPTJrW7f9/MKCsppKyk8DNde98/bgyvLF3Pn55dwpUz53Ps+L707FrUrmtnS2ubwUXjTjzuOdkK5f0126iNxrM2AwiUACRLCsMhpuw9gCl7D6AmEkvcHGOxhpvkjrrET00kRiyeuIFGYvHET9SpjcWJxeLEHGLxxEOzayMxaqPxhoE7M7DE7Z94cnZSJBanLhqnLvnfePIbef038/obdiye+KmLxok1mdlk0HBDjsUTN+UeXQoZ0KML3bsUYGbEk79rVp9YEjfzksIQxQVhQgbzVm7h8bdXct8bn+xS/dW3egrDIQrCn01YoSav3Z312+rYVrvzitYu9UmnIMTG7XV87YBBaV3pHQoZR40pp1/3Er507Uvc/spHXPHFsWm7fjq1tBCs/n1dLE5JKPtdWPUDwNmaAQRKAJIDJYXh5EZX+TVDKBqLM2/lFpZvqqYumkh8ddF4kySUaE2EzDBLJJ1IrDERRmOJ99F4vCFhxeJOzBPfWuu75fuUFtG3rIS+ZcU4UFUdoaomwvbaaDIZOl2Lwlx69KiM/DvH9i/jpH0GcOery7joiD0D2QpoaR1AfUsyEotnZSO25t5dsYXuJQUM6901a5+pBCCSJQXhEPsP3YP9h+6R61Ay7rJjR/GPeau445WP+EHAWgHunlwHsPMYAJCzgeC5KzYzcXDPrI5rda7JvSISCOP6d+eEvftz56vLArfALIqkIokAAA0WSURBVBp33GlxDADIyWKwmkiMxau3ZrX/H5QARCRDLj92NFtro9z+6ke4O9FYnGgA9pOqn+nTfB1ALlsAC1ZVEY07E7PY/w/qAhKRDBk/oDtT9urPdc+/z3XPvw8k5tr/4bR9+cr+g3IWVyTqyVh2XgcA5OSZAHOXJweAs7QCuF6bCcDM7gBOBta6+94tnDfgz8CJwA7gAnefk+5ARaTj+dmXJzCmfxkGFISMZxas4cqZ8zlydB96lxbnJKb6G3zzMYCmg8DZNnfFFsrLiunfvSSrn5tKF9BdwJTPOX8CMDr5Mw34y+6HJSKdwaCeXfjB8WP4j+PHcNmxo7nmzH3ZURflN08tyllM9Tf4lh4IA7npApq3cgsTB/XI+sLGNhOAu78EbPycIqcCd3vC60BPMxuQrgBFpPMY1beMaUeN4NE5K3jtgw05iaFhDKDVQeDsJoDaaIwP129vcdV1pqVjEHgQsLzJ+xXJYyIiO7n06NEM6dWF/3l8Xk6+bbeWAHLVAli6dhuxuDO2f1lWPxfSMwjcUpulxXlUZjaNRDcRQ4cOTcNHi0hH06UozFWn7M2Fd73F2X99HYDVW2rYVhtlzz7dGNuvjDH9yzh+fD+GZmBRVF2rg8CNK4GzadGqxEOWxvXvmC2AFcCQJu8HA5+2VNDdb3X3CnevKC8vT8NHi0hHdPS4vnxj0lA2ba+juCDEISN6cfLEAXQpDPPcwjX875ML+MIfZ3H+HW/yzPzVaZ0+2jgG0NogcHbXASxaXUVxQYjhWVwBXC8dLYCZwKVm9iBwCLDF3Vel4boi0on96iv7tHpu5eZq/la5nAfe/IRp98ymKByiX49iBvbowvDe3Thp4gAOH9Vnl3ZZDVoX0KLVWxnTryxrW0A3lco00AeAyUAfM1sBXElyExd3vxl4isQU0KUkpoFemKlgRSQ/DOrZhe8fN4ZLjx7F84vWMueTTazaXMOqLdU8/d4qHqpczsAeJXztgMFMPbJ9+w3V3+CDMgi8cNVWjh6bmx6RNhOAu5/dxnkHvpu2iEREkgrCIb60V3++tFf/hmO10RjPLVjLw5XLuemFpbz+4QYemDZppxt6a+pabQEkF4JlsQWwflst67fV5mQAGLQVhIh0MMUFYU6aOIAZ3zqYa8/an8qPN/Gbpxam/Pv1ffw7PRQ+B4PAi1cnBoDHD8j+ADAoAYhIB3bKvgO58PDh3PnqMv7+zsqUfqdxL6DPjh/kogto4aoqAMapBSAi0n4/PXE8Bw3fg588Oo9Fq6vaLN8wCygAg8CLVm+lvKw4Z9tiKAGISIdWGA5x4zkHUFpSwJm3vM4N/3qfrTWtb0EdpEHgRaurcvbtH5QARKQT6Nu9hPunHkLFsD344zNLOOL3s7hx1lJi8Z3n9DeMATTbC6gglN1B4GgszpI123LW/w9KACLSSYzuV8btFxzEzEsPp2LYHvzhn4u545WPdirX2joAM6MoHKIuhYVg7y7fzKot1bsV77IN26mLxhnbTy0AEZG0mDi4J7edX8Fx4/vxx2cW89H67Z8535gAdl5EVlQQ4p3lm/h/761i0eoqaqOxncrM/ngjX//Lvzn71tfZXhvd5TgXJWcAjRugBCAikjZmxq+/ujdFBSF+/Mhc4k26glpbBwCw35CevP7hRqbfO4cp177Mkb+fxXsrtzScX7e1lkvum0Pv0iI+3riDXz4xf5djXLRqK+GQMapv6S5fY3cpAYhIp9Svewk/O3kCby7byL1vfNxwvLVBYIB7px7CvF98kScuPYJrz9yPgpBx5i2v8fL764jG4lz2wBw274hw5wUHc8nkkTxcuYKn5u3azjeLVlcxsrwbxQXhXfsHpoESgIh0WqcfOJijxpTzu6cXsXzjDiDRBRQOWav7CJWVFLLP4B58Zf9BPHbJ4Qzp1ZUL73yLb82o5PUPN/Lrr+7DhIHd+f5xY9h3cA/+67F5uzQesHDV1pzsANqUngksIp2WmfHbr+3Dl655iakzKnlw2iQiMW+x/78l/XuU8PD0Q5l+z2xeWrKOcw4ZymkHDgYSLYg/n7U/J173Muf89Q1GlpdSEDLCYSNkhgH1OcYBd4i7E3cnFndWbq7m3Em53RZfCUBEOrVBPbtwy3kHcuFdb3H+nW8ypl9ZyvsGAXQvKeSuCw/mxSXr+MKYz27aNrxPN645cz9ueuEDPt1cTSzuROJxaLjZQ/1THhMJwQiFjLAZ+w7uwbHj+qXxX9p+ltjLLfsqKiq8srIyJ58tIvnnuQVrmH7vbKJxp09pEZX/c3yuQ9olZjbb3SvScS2NAYhIXjhuQj+uPnM/zKAgpFsfqAtIRPLIKfsOJGy224u4OgslABHJKydNHJDrEAJD7SARkTylBCAikqdSSgBmNsXMFpvZUjP7SQvnLzCzdWb2TvJnavpDFRGRdErlofBh4EbgeGAF8JaZzXT3Bc2KPuTul2YgRhERyYBUWgAHA0vd/UN3rwMeBE7NbFgiIpJpqSSAQcDyJu9XJI8193Uzm2tmj5jZkLREJyIiGZNKAmhp04zmy4efAIa7+0TgOWBGixcym2ZmlWZWuW7duvZFKiIiaZVKAlgBNP1GPxj4tGkBd9/g7rXJt38FDmzpQu5+q7tXuHtFeXl5S0VERCRLUkkAbwGjzWxPMysCzgJmNi1gZk1XVpwCLExfiCIikgltzgJy96iZXQr8EwgDd7j7fDO7Cqh095nA5WZ2ChAFNgIXZDBmERFJA+0GKiLSgWg3UBER2W1KACIieUoJQEQkTykBiIjkKSUAEZE8pQQgIpKnlABERPKUEoCISJ5SAhARyVNKACIieUoJQEQkTykBiIjkKSUAEZE8pQQgIpKnlABERPKUEoCISJ5SAhARyVNKACIieUoJQEQkT6WUAMxsipktNrOlZvaTFs4Xm9lDyfNvmNnwdAcqIiLp1WYCMLMwcCNwAjABONvMJjQrdhGwyd1HAdcAv093oCIikl6ptAAOBpa6+4fuXgc8CJzarMypwIzk60eAY83M0hemiIikW0EKZQYBy5u8XwEc0loZd4+a2RagN7C+aSEzmwZMS76tNbP3diXoVvQAtqS5fGtlWjqeyrGm75u+7kOzutpN6a6Lzzuvuvj8c6qL9h1TXbT8vunrsW0FmzJ3/9wf4HTgtibvzwOub1ZmPjC4yfsPgN5tXLeyrc9uzw9wa7rLt1ampeOpHGv6vtnrQNfF551XXaguVBcdty5S6QJaAQxp8n4w8GlrZcysgES22pjCtdPpiQyUb61MS8dTOfbE55xLp3TXxeedV118/jnVRfuOqS5afp+RurBkRmm9QOKGvgQ4FlgJvAWc4+7zm5T5LrCPu083s7OAr7n7GW1ct9LdK3b3H9AZqC4aqS4aqS4aqS4apbMu2hwD8ESf/qXAP4EwcIe7zzezq0g0RWYCtwP3mNlSEt/8z0rhs2/djbg7G9VFI9VFI9VFI9VFo7TVRZstABER6Zy0ElhEJE8pAYiI5CklABGRPBXIBGBmITP7tZldb2bn5zqeXDKzyWb2spndbGaTcx1PrplZNzObbWYn5zqWXDKz8cm/iUfM7OJcx5NLZvYVM/urmf3dzL6Y63hyycxGmNntZvZIKuXTngDM7A4zW9t8lW9bG8o1cyqJ1cUREmsMOqQ01YUD24ASVBcAPwYezkyU2ZGOunD3he4+HTgD6LDTI9NUF4+7+7eBC4AzMxhuRqWpLj5094tS/sx0zwIys6NI3LDudve9k8fCJNYSHE/iJvYWcDaJaaW/bXaJbyV/Nrn7LWb2iLufltYgsyRNdbHe3eNm1g+42t3PzVb86ZSmuphIYkuAEhL18mR2ok+vdNSFu681s1OAnwA3uPv92Yo/ndJVF8nf+xNwn7vPyVL4aZXmukjpvpnKXkDt4u4vtbAddMOGcsngHgROdfffAjs15c1sBVCXfBtLd4zZko66aGITUJyJOLMhTX8XRwPdSOxKW21mT7l7PKOBZ0C6/i6Sa3Bmmtk/gA6ZANL0d2HA74CnO+rNH9J+v0hJ2hNAK1LZUK6px4DrzexI4KVMBpYD7aoLM/sa8CWgJ3BDZkPLunbVhbv/N4CZXUCyZZTR6LKrvX8Xk4GvkfhS8FRGI8u+9t4vLgOOA3qY2Sh3vzmTwWVZe/8uegO/BvY3s/9KJopWZSsBtLQ1dKt9T+6+g8QzBjqj9tbFYyQSYmfUrrpoKOB+V/pDybn2/l28ALyQqWByrL11cR1wXebCyan21sUGYHqqF8/WLKBUNpTLF6qLRqqLRqqLRqqLRhmti2wlgLeA0Wa2p5kVkdgraGaWPjtoVBeNVBeNVBeNVBeNMloXmZgG+gDwGjDWzFaY2UXuHgXqN5RbCDzcdDfRzkp10Uh10Uh10Uh10SgXdaHN4ERE8lQgVwKLiEjmKQGIiOQpJQARkTylBCAikqeUAERE8pQSgIhInlICEBHJU0oAIiJ5SglARCRP/X+nzuNQTQmy4gAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.semilogx(history.history[\"lr\"], history.history[\"val_loss\"])\n",
    "plt.axis([1e-6, 1e-1, 0, 4])\n",
    "plt.title('Validation Loss');"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# A learning rate between 1e-5 and 1e-4 seemed appropriate as the loss began fluctuating between 1e-4 and 1e-3\n",
    "LEARNING_RATE = 1e-4\n",
    "\n",
    "n_filters = 32\n",
    "filter_width = 2\n",
    "dilation_rates = [2**i for i in range(8)] * 2\n",
    "\n",
    "tf.keras.backend.clear_session()\n",
    "history_seq = tf.keras.layers.Input(shape =(n_steps_in, n_features))\n",
    "x = history_seq\n",
    "\n",
    "skips = []\n",
    "for dilation_rate in dilation_rates:\n",
    "    \n",
    "    # preprocessing - equivalent to time-distributed dense\n",
    "    x = tf.keras.layers.Conv1D(16, 1, activation='relu')(x) \n",
    "    \n",
    "    # filter convolution\n",
    "    x_f = tf.keras.layers.Conv1D(filters=n_filters,\n",
    "                                kernel_size=filter_width, \n",
    "                               activation = 'tanh',\n",
    "                                padding='causal',\n",
    "                                dilation_rate=dilation_rate)(x)\n",
    "    \n",
    "    # gating convolution\n",
    "    x_g = tf.keras.layers.Conv1D(filters=n_filters,\n",
    "                                 kernel_size=filter_width, \n",
    "                                 activation = 'sigmoid',\n",
    "                                 padding='causal',\n",
    "                                 dilation_rate=dilation_rate)(x)\n",
    "    \n",
    "    # multiply filter and gating branches\n",
    "    # works better than activations.swish - x * sigmoid(x)\n",
    "    z = tf.keras.layers.Multiply()([x_f, x_g])\n",
    "    \n",
    "    # postprocessing - equivalent to time-distributed dense\n",
    "    z = tf.keras.layers.Conv1D(16, 1, activation='relu')(z)\n",
    "\n",
    "    \n",
    "    # residual connection\n",
    "    x = tf.keras.layers.Add()([x, z])    \n",
    "\n",
    "\n",
    "    # collect skip connections\n",
    "    skips.append(z)\n",
    "\n",
    "# add all skip connection outputs \n",
    "out = tf.keras.layers.Add()(skips)\n",
    "out = tf.keras.layers.Activation('relu')(out)\n",
    "\n",
    "# final time-distributed dense layers \n",
    "out = tf.keras.layers.Conv1D(128, 1, activation = 'relu')(out)\n",
    "out = tf.keras.layers.Dropout(.2)(out)\n",
    "out = tf.keras.layers.Conv1D(1, 1)(out)\n",
    "out = tf.keras.layers.Flatten()(out)\n",
    "pred_seq_train = tf.keras.layers.Dense(n_output, activation = 'relu')(out)\n",
    "\n",
    "model = tf.keras.Model(history_seq, pred_seq_train)\n",
    "\n",
    "\n",
    "optimizer = tf.keras.optimizers.Adam(lr = LEARNING_RATE\n",
    "                                    )\n",
    "\n",
    "# tf.keras.losses.Poisson()\n",
    "model.compile(optimizer, loss=tf.keras.losses.Poisson(), metrics = ['mae'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:Inconsistent references when loading the checkpoint into this object graph. Either the Trackable object references in the Python program have changed in an incompatible way, or the checkpoint was generated in an incompatible program.\n",
      "\n",
      "Two checkpoint references resolved to different objects (<tensorflow.python.keras.layers.convolutional.Conv1D object at 0x00000000361ECDC8> and <tensorflow.python.keras.layers.core.Activation object at 0x00000000361E3088>).\n",
      "WARNING:tensorflow:Inconsistent references when loading the checkpoint into this object graph. Either the Trackable object references in the Python program have changed in an incompatible way, or the checkpoint was generated in an incompatible program.\n",
      "\n",
      "Two checkpoint references resolved to different objects (<tensorflow.python.keras.layers.convolutional.Conv1D object at 0x00000000056828C8> and <tensorflow.python.keras.layers.core.Dropout object at 0x00000000361F0F08>).\n",
      "WARNING:tensorflow:Inconsistent references when loading the checkpoint into this object graph. Either the Trackable object references in the Python program have changed in an incompatible way, or the checkpoint was generated in an incompatible program.\n",
      "\n",
      "Two checkpoint references resolved to different objects (<tensorflow.python.keras.layers.core.Dense object at 0x00000000362519C8> and <tensorflow.python.keras.layers.core.Flatten object at 0x0000000036200F08>).\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<tensorflow.python.training.tracking.util.CheckpointLoadStatus at 0x36a9fa08>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.load_weights('DCNN_011820')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/10\n",
      "177/177 [==============================] - 14s 78ms/step - loss: 0.6580 - mae: 0.1102\n",
      "Epoch 2/10\n",
      "177/177 [==============================] - 14s 78ms/step - loss: 0.6580 - mae: 0.1100\n",
      "Epoch 3/10\n",
      "177/177 [==============================] - 13s 72ms/step - loss: 0.6603 - mae: 0.1129\n",
      "Epoch 4/10\n",
      "177/177 [==============================] - 14s 79ms/step - loss: 0.6581 - mae: 0.1102\n",
      "Epoch 5/10\n",
      "177/177 [==============================] - 15s 86ms/step - loss: 0.6602 - mae: 0.1133\n",
      "Epoch 6/10\n",
      "177/177 [==============================] - 14s 79ms/step - loss: 0.6580 - mae: 0.1102\n",
      "Epoch 7/10\n",
      "177/177 [==============================] - 14s 76ms/step - loss: 0.6576 - mae: 0.1095\n",
      "Epoch 8/10\n",
      "177/177 [==============================] - 13s 75ms/step - loss: 0.6573 - mae: 0.1092\n",
      "Epoch 9/10\n",
      "177/177 [==============================] - 13s 75ms/step - loss: 0.6577 - mae: 0.1095\n",
      "Epoch 10/10\n",
      "177/177 [==============================] - 15s 82ms/step - loss: 0.6621 - mae: 0.1153\n"
     ]
    }
   ],
   "source": [
    "history = model.fit(X, y, epochs = 10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "# model.save_weights('DCNN_011820')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "### Validate on the test-set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1, 287)"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_input = np.array(train.iloc[-120:].values)\n",
    "x_input = scaler.transform(x_input)\n",
    "\n",
    "x_input = x_input.reshape((1, n_steps_in, n_features))\n",
    "\n",
    "yhat = model.predict(x_input, verbose=0)\n",
    "yhat.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "yhat_df = pd.DataFrame(yhat.reshape(41,7), index = test[:41].index, columns = fire_regions)\n",
    "\n",
    "# create dummy variables to inverse transform\n",
    "for i in range(70):\n",
    "    yhat_df['dummy_{}'.format(i)] = 0\n",
    "    \n",
    "_yhat_df = scaler.inverse_transform(yhat_df)\n",
    "\n",
    "yhat_df = pd.DataFrame(data = _yhat_df,\n",
    "                      index = yhat_df.index,\n",
    "                      columns = yhat_df.columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAeAAAAF1CAYAAAAwfzllAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOzdd5ib1ZX48e+VNNLMSNPHbcbduGEwYKppgQAJJLS0TbIhS0IWkmzqL22z6WHTNrsJqYQlIYFsCBtSWBJCSGim2oDtgGm2weA24za9SSONdH9/3PfVyOPRqEuvNOfzPHpm1F9N0dE9995zlNYaIYQQQhSXq9QHIIQQQkxHEoCFEEKIEpAALIQQQpSABGAhhBCiBCQACyGEECUgAVgIIYQoAQnAQgghRAlIABZpUUotV0r9XSk1qJT6aKmPJx+UUh9USh1QSg0ppVpKfTw2pdRZSqltpT6OySilzlFK7U3jdof9vSilblBKfbEYxyhEuZAALNL1GWCd1rpOa/2DUh9MKkqpnUqp86e4vgr4LvA6rXVAa92dp+cMWgHdPv0ojftppdRR9nmt9SNa6+W5Hk+S57pZKfW1Qjz2BIf9vWitP6C1/vciPG/RWb+/Z5VSroTLvqaUujnh/PuUUlutDyQHlFJ/VkrVKaXWKqUGlFLuhNv+NMllNxTtRYmikAAs0rUAeD6bOyqlPHk+lnyYBVSTxWtSRrL/nUusgG6fPpzTUZavtP9eivX3kRjQCqANeEeS530N8A3gnVrrOmAlcLt19UbADaxJuMtZQOeEy84GHs7zMYsSkwAsUlJKPQCcC/zIGtUtU0o1KKV+qZQ6pJTapZT6gh2UlFLvUUo9ppS6TinVA3xFKeVTSv2XUmq3NQK4QSlVk/AclymlnrY++e9QSl1oXf5epdSL1sjhFaXU+xPu06qUuksp1aeU6lFKPaKUciml/geYD/zJOt7PTHg9ywA7xdtnvT6UUqcrpZ5SSvVbX09PuM86pdTXlVKPASPA4gx/hkcppR6yHrtLKfUb63L7TfUZ61jfPjHNa42sP62U2qKUGlZK3aSUmqWU+ov1c7lPKdWUcPvfKqX2W8/1sFJqlXX5NcC7gM9Yz/Un6/I2pdTvrd/lqyphikEpVWONmnuVUi8AJ6fxWif7e4mPvO3Xp5T6V6XUfuAX1u/ts9bvvlspdbtSqjmN55r0tVrX3ayU+olS6m6l1DBwborXeopSar3197RPKfUjpZQ31TFYvg18NcmHiZOB9VrrvwNorXu01rdorQe11hFgAybAopSaCXiB30y4bBkSgCuP1lpOckp5AtYB/5xw/pfAnUAdsBDYDrzPuu49wBjwEcAD1ADfA/4INFv3+RPwTev2pwD9wAWYD4XtwArrujcCSwAFvAYT/NZY130TuAGosk5nAcq6bidw/hSvZyGgAY91vhnoBd5tHfM7rfMtCa9/N7DKur5qksdM+pzAbcDnrddXDZyZcJ0Gjko4fw6wd8LjbsCM2tuBg8Bm4ATABzwAfDnh9ldZP2Of9XN/OuG6m4GvJZx3AZuAL2He+BcDrwCvt67/FvCI9fOZBzyXeGwZ/L3En9d6fWPAf1jHWAN83HqNc63L/hu4LY3nSfVa+4EzrNdZm+K1ngicZv1+FwIvAh9P4xg0sNR67H+2LvsacLP1/VlAEPiqdSy+Cff/MnCn9f1bMf9bF0y47JVSvwfIKf+nkh+AnMrjlPiGikmZjQJHJ1z/fsycH5gAvDvhOgUMA0sSLlsLvGp9/9/AdWkex/8BH7O+vxbzIeCoSW63k8wC8LuBJyfcZj3wnoTXf22KY9sJDAF9Caerret+CdwIzJ3kfukE4HclnP898JOE8x8B/i/JMTVaj99gnb+ZwwPwqYm/K+uyfwN+YX3/CnBhwnXXkJ8AHAaqE65/ETgv4fwcIGL/ftL825jstf4y3dc6yeN9HLgjjefVwFHAGzAf0nwkBGDrNhdhPnT2WX8j3wXcCT+Pbsz/yfeBq4EAcCDhskmPUU7lfZIUtMhGK2YEsSvhsl2Y0ZltT8L3M7BGH1Z6rw+4x7oczMhqx2RPpJS6SCm1wUox92He5Fqtq/8TeBn4m5We/mwOr6ltwutJ9ZqSuVxr3Zhw+ql1+Wcwb6ZPKqWeV0pdleHxHUj4PjjJ+QCYeU6l1LesVO4AJnjD+M9sogVAm/17sX7Gn8OMtsH8XBJf98SfUbYOaa1DE47jjoRjeBGIJhzHEdJ8rYnHPuVrtVLld1kp7QHMvG2yn9sRtNZ3YwLwNZNc9xet9SWYTMJlmA+p/2xdvQHz+zsGk3Z+RGs9ZB27fZmknyuQBGCRjS7M6GRBwmXzgY6E83rC7YPAqoTA1KC1DljX78GkmQ+jlPJhRnv/BczSWjcCd2MCGdrMoX1Sa70YuAT4hFLqvEmePx2dE15PqteUEa31fq311VrrNky24HqVsPI5j/4R8wZ/PtCAGemD9TPjyNewB5OJSPzQUKe1foN1/T7MByTb/Dwd52THcdGE46jWWndMdmdLqtc68XlSvdafAFuBpVrrekxwTnysdHwBM9VQO9mVWuuY1vp+zLTBMdZlIeAp4GJgjtZ6q3XzR6zLViMBuCJJABYZ01pHMas4v67MVooFwCeAXyW5fQz4KXCdtaAEpVS7Uur11k1uAt6rlDrPWozTrpRagRll+4BDwJhS6iLgdfbjKqUuthY3KWAAM2KKWlcfILOFUncDy5RS/6iU8iil3g4cDdyVwWMkpZR6m1JqrnW2FxMYsj3WqdRhpge6MUHgGxOun/hcTwID1oKoGmtUeYxSyl5sdTvwb0qpJuv4P5Kn45zoBszf0wIApdQMpdRlKe6T6rVOlOq11mH+joasv78PZvoitNbrgGeBK+3LlFlg+A7rZ6iUUqdg1jNsSLjrw5iU9+MJlz1qXbZfaz1phkiUNwnAIlsfwczrvoJ5o/g18PMpbv+vmHTxBiu9dx+wHEBr/STwXuA6zKKZh4AFWutB4KOYINCLGfH8MeExl1qPM4SZr73eegMEs0DrC1aq8VOpXow2+4AvBj6JeUP/DHCx1ror1X0nsFde26c7rMtPBp5QSg1Zr+FjWutXreu+AtxiHes/ZPh8E/0SkybuAF7g8Dd5MB92jrae6/+sD1OXAMcDr2KyFT/DjCjBLBzaZV33N+B/cjy+ZL6P+bn8TSk1aB33qSnuk+q1HiaN1/opzN/YIOYD42+yeSGYUXDiCu5ezLzuS5gA/yvgP7XWtybc5iFgJuZ/yfaodZmMfiuUvWJUCCGEEEUkI2AhhBCiBCQACyEyppSaPyHVnnjK10ItlFLvSvIcWVVly/IYzkr2Wot1DKIySQpaCCGEKAEZAQshhBAlUNQi+a2trXrhwoXFfEohhBCiZDZt2tSltZ4x2XVFDcALFy5k48aNxXxKIYQQomSUUkmrx0kKWgghhCgBCcBCCCFECUgAFkIIIUqgqHPAQgghKkskEmHv3r2EQqHUN65g1dXVzJ07l6qqqrTvIwFYCCFE1vbu3UtdXR0LFy7E9EWZfrTWdHd3s3fvXhYtWpT2/SQFLYQQImuhUIiWlpZpG3wBlFK0tLRknAWQACyEECIn0zn42rL5GUgAFkIIMW2sW7eOxx9/PPUNpxAIBPJyLBKAhRBCTBv5CMD5IgFYCCFE2bv88ss58cQTWbVqFTfeeCMA99xzD2vWrOG4447jvPPOY+fOndxwww1cd911HH/88TzyyCO85z3v4Xe/+138cezR7dDQEOeddx5r1qzh2GOP5c4778z7McsqaCGEEHnx1T89zwudA3l9zKPb6vnyJatS3u7nP/85zc3NBINBTj75ZC677DKuvvpqHn74YRYtWkRPTw/Nzc184AMfIBAI8KlPfQqAm266adLHq66u5o477qC+vp6uri5OO+00Lr300rzOd0sArkChSJSDA6PMb6kt9aEIIURR/OAHP+COO+4AYM+ePdx4442cffbZ8W1Bzc3NGT2e1prPfe5zPPzww7hcLjo6Ojhw4ACzZ8/O2zFLAK5Av9qwi+vu3c4zX34dHrfMMgghiiOdkWohrFu3jvvuu4/169dTW1vLOeecw3HHHce2bdtS3tfj8RCLxQATdMPhMAC33norhw4dYtOmTVRVVbFw4cK8FxuRd+cKdGholOFwlJFItNSHIoQQBdff309TUxO1tbVs3bqVDRs2MDo6ykMPPcSrr74KQE9PDwB1dXUMDg7G77tw4UI2bdoEwJ133kkkEok/5syZM6mqquLBBx9k166kTY2yJgG4AoXCJvAOj46V+EiEEKLwLrzwQsbGxli9ejVf/OIXOe2005gxYwY33ngjb37zmznuuON4+9vfDsAll1zCHXfcEV+EdfXVV/PQQw9xyimn8MQTT+D3+wF417vexcaNGznppJO49dZbWbFiRd6PW1LQFWhEArAQYhrx+Xz85S9/mfS6iy666LDzy5YtY8uWLYddtmHDhvj33/zmNwFobW1l/fr1kz7m0NBQLocbJyPgChSM2AFYUtBCCOFUKQOwUmqeUupBpdSLSqnnlVIfsy5vVkrdq5R6yfraVPjDFekIRWQELIQQTpfOCHgM+KTWeiVwGvAhpdTRwGeB+7XWS4H7rfPCAewR8JAEYCGEcKyUAVhrvU9rvdn6fhB4EWgHLgNusW52C3B5oQ5SZCZozwGHJQALIYRTZTQHrJRaCJwAPAHM0lrvAxOkgZlJ7nONUmqjUmrjoUOHcjtakZZgxOxpkzlgIYRwrrQDsFIqAPwe+LjWOu1aY1rrG7XWJ2mtT5oxY0Y2xygyJHPAQgjhfGkFYKVUFSb43qq1/oN18QGl1Bzr+jnAwcIcoshUULYhCSFEVtatW8fFF19clOdKZxW0Am4CXtRafzfhqj8CV1rfXwnkv1WEyEp8G1JYUtBCCAEQjTrv/TCdEfAZwLuB1yqlnrZObwC+BVyglHoJuMA6LxxARsBCiOlk586drFixgiuvvJLVq1fz1re+lZGRERYuXMi1117LmWeeyW9/+1v+9re/sXbtWtasWcPb3va2eEGNe+65hxUrVnDmmWfyhz/8IcWz5U/KSlha60eBZP2Xzsvv4YhcjUVjhKNmEZZsQxJCFNVfPgv7n83vY84+Fi5KPb7btm0bN910E2eccQZXXXUV119/PWDaCj766KN0dXXx5je/mfvuuw+/389//Md/8N3vfpfPfOYzXH311TzwwAMcddRR8ZKVxSCVsCpMaCwW/35EUtBCiGli3rx5nHHGGQBcccUVPProowDxgLphwwZeeOEFzjjjDI4//nhuueUWdu3axdatW1m0aBFLly5FKcUVV1xRtGOWWtAVJpgQdGUELIQoqjRGqoVilisded5urqC15oILLuC222477HZPP/30EfctFhkBV5hQQgtCmQMWQkwXu3fvjjdPuO222zjzzDMPu/60007jscce4+WXXwZgZGSE7du3s2LFCl599VV27NgRv2+xSACuMPYKaI9LSQpaCDFtrFy5kltuuYXVq1fT09PDBz/4wcOunzFjBjfffDPvfOc7Wb16Naeddhpbt26lurqaG2+8kTe+8Y2ceeaZLFiwoGjHLCnoCmOnoJv9XklBCyGmDZfLxQ033HDYZTt37jzs/Gtf+1qeeuqpI+574YUXsnXr1kIe3qRkBFxh7BFwa8AnKWghhHAwCcAVJh6A63yMhKPEYrrERySEEIW1cOFCnnvuuVIfRsYkAFcYOwXdGvACMBKReWAhhHAiCcAVxg7AMwI+QFZCCyEKT2vJtGXzM5AAXGES54BB9gILIQqrurqa7u7uaR2EtdZ0d3dTXV2d0f1kFXSFsfcBt9gpaOkJLIQooLlz57J3716me7/36upq5s6dm9F9JABXmPE5YBkBCyEKr6qqikWLFpX6MMqSpKArTDASxeNSNNZWATIHLIQQTiUBuMIEI1FqqtzUek1yYzgsAVgIIZxIAnCFCUWiVHvdBHxWAJY5YCGEcCQJwBUmGI5S63Xj97kBSUELIYRTSQCuMCNhSUELIUQ5kABcYYKRKNVVbtwuRU2VW0bAQgjhUBKAK0zIWoQF4Pd5GJI5YCGEcCQJwBUmGIlS47UDsJsRSUELIYQjSQCuMMFwwgjY65EUtBBCOJQE4AoTisSotgJwwOeRSlhCCOFQEoArjElBm19rrc/NSFjmgIUQwokkAFeYw1LQMgIWQgjHkgBcQbTW1gjY7AEOyBywEEI4lgTgChKKxAAOGwFLKUohhHAmCcAVJGj1Aq6pMr9Wv8/NcHhsWjfKFkIIp5IAXEHiAdg7PgLWevxyIYQQziEBuIIErRXP1QkpaEAWYgkhhANJAK4goXgK2i7EYb6OyDywEEI4jgTgCjJZChpkBCyEEE4kAbiC2CnomoRKWCA9gYUQwokkAFeQiSPgWjsFLdWwhBDCcSQAV5CJc8ABSUELIYRjSQCuIPZId+IcsKSghRDCeSQAV5CJc8B+qyTlsKSghRDCcSQAVxB7Dnh8H7D5KiNgIYRwHgnAFSQUiaIU+Dzm1+pxu/B5XBKAhRDCgSQAVxC7FaFSKn5ZwOdhOCwBWAghnEYCcAUJRsZ7AdtqfW7piCSEEA4kAbiCBCPR+Pyvze/1yDYkIYRwIAnAFSQUicaLb9gCPo/MAQshhANJAK4gwXA0vgfYVuvzyDYkIYRwIAnAFWQkfGQKOuBzywhYCCEcSAJwBQlNsgjL75UUtBBCOJEE4Aoy2Spov8wBCyGEI0kAriDByJFzwH6fm+FwFK11iY5KCCHEZCQAV5BgOHbkNiSfh2hMMzoWK9FRCSGEmIwE4AqSbA4YpB60EEI4jQTgCqG1tlLQh/9Kx1sSylYkIYRwEgnAFSIS1URjmlprxGsLWB2RpBqWEEI4iwTgCjGxFaHNDsgj0pBBCCEcRQJwhQha1a4m24YEMgIWQginkQBcIewR8MQ54IDMAQshhCNJAK4QyUfA5rz0BBZCCGeRAFwhks0ByzYkIYRwJgnAFSIUmXoOWAKwEEI4iwTgChFPQU8oRen1uPC6XQzJHLAQQjiKBOAKEUwyAgao9bllG5IQQjiMBOAKMb4K+sgA7Pd6ZBuSEEI4jATgCpFsDhjMViSZAxZCCGeRAFwhRpLMAYOdgpY5YCGEcBIJwBXCXoRV7Zl8BCwpaCGEcJaUAVgp9XOl1EGl1HMJl31FKdWhlHraOr2hsIcpUglFovg8LlwudcR1fq+koIUQwmnSGQHfDFw4yeXXaa2Pt0535/ewRKZMK8IjR79gUtBSilIIIZwlZQDWWj8M9BThWEQOguHopAuwwFqEJduQhBDCUXKZA/6wUmqLlaJuSnYjpdQ1SqmNSqmNhw4dyuHpxFSCkeQB2C+roIUQwnGyDcA/AZYAxwP7gO8ku6HW+kat9Ula65NmzJiR5dOJVEKR6BF1oG0Bn4dIVBMeixX5qIQQQiSTVQDWWh/QWke11jHgp8Ap+T0skalgJEptsjlg63IZBQshhHNkFYCVUnMSzr4JeC7ZbUVxBMPJF2HZDRlkK5IQQjiHJ9UNlFK3AecArUqpvcCXgXOUUscDGtgJvL+AxyjSMBKO0hLwTXpdwArAUoxDCCGcI2UA1lq/c5KLbyrAsYgchKZYhGWnoGUELIQQziGVsCrEVKugA9ITWAghHEcCcIVIZw5YWhIKIYRzSACuEKFILOk2JL/XXoQlc8BCCOEUEoArwFg0Rjgam6IQh2xDEkIIp5EAXAFCVoGNGu/kv07ZhiSEEM4jAbgCBOO9gCdf1O7zuHC7lMwBCyGEg0gArgChiBWAk6SglVL4vdIRSQghnEQCcAWwC2wkC8BgtiJJCloIIZxDAnAFCNoj4CRzwAC1Po+koIUQwkEkAFcAew442TYkMAuxZBuSEEI4hwTgCpBqDhgg4HPLNiQhhHAQCcAVYDwFPcUI2OuRACyEEA4iAbgCBNNYhOX3eRiWOWAhhHAMCcAVIK0RsE+2IQkhhJNIAK4A6cwB+32SghZCCCeRAFwB0loF7fUwOhZjLBor1mEJIYSYggTgCjASiVLlVlS5k/86/fGewJKGFkIIJ5AAXAGC4eiUo18w25AAWYglhBAOIQG4AoQi0SnnfwFqvfYIWAKwEEI4gQTgChCMRKdcAQ2mFjRIS0IhhHAKCcAVIBhOPQKWOWAhhHAWCcAVIBhJPQdc65U5YCGEcBIJwBUgFInGA2wyAZ/MAQshhJNIAK4AwTQWYfklAAshhKNIAK4AwXCU6hQjYH98G5LMAQshhBNIAK4A6SzCqqly41IyAhZCCKeQAFwB0klBK6Xwez2yDUkIIRxCAnAFSGcfMJh54BHZhiSEEI4gAbjMxWKaUCSWchsSQK3PzZBsQxJCCEeQAFzmRsdMd6NUKWgwW5FkDlgIIZxBAnCZC8Z7Aaf+Vfq9koIWQginkABc5uwAbDdbmIrf55ZFWEII4RASgMtc0NrXm2ofMJhFWFKKUgghnEECcJkLxVPQaQZgSUELIYQjSAAucyPhDAKw1y2LsIQQwiEkAJe5+CIsbxqLsHwegpEo0Zgu9GEJIYRIQQJwmYvPAae5DQmkJaEQQjiBBOAyl8kcsL1SWrYiCSFE6UkALnPjKeh0FmGZ28hWJCGEKD0JwGUumMEirID0BBZCCMeQAFzmMhsByxywEEI4hQTgMheKRHEp8LrTK0UJyF5gIYRwAAnAZS4YNr2AlVIpb2vPAUsKWgghSk8CcJkbSbMXMMg2JCGEcBIJwGUuFI6mtQcYoFYWYQkhhGNIAC5zwUg0rRXQALVV9jYkmQMWQohSkwBc5oIZpKBdLoXf62ZERsBCCFFyEoDLXDCDFDSYNLTMAQshROlJAC5zoQxS0GAWYkkKWgghSk8CcJkLRqLUppmCBrMVSVLQQghRehKAy1wmi7DANGSQWtBCCFF6EoDLXDAcpTqDEXBA5oCFEMIRJACXObsSVrr8Po+UohRCCAeQAFzGtNYZp6D9XrcU4hBCTCsbXulm2/7BUh/GESQAl7FwNEZMp9cJyWZGwBKAhRDTx6d/9wzX3bu91IdxBAnAZSwUjgFktA/Y7/MwHI4Si+lCHZYQQjhGNKbZ1xeiZzhc6kM5ggTgMhbvBZzRPmD3YfcVQohKdmhwlLGYpndEArDIo3gA9qb/a6z1SkMGIcT00dEXBKB3JFLiIzmSBOAyFgzbI2BP2vexWxLKXmAhxHSwr98E4L6RMFo7a+pNAnAZGx8BZzYHDDASlhS0EKLydVoj4LGYdtzAQwJwGRsfAWe2DQlkBCyEmB46+0Lx7/scloaWAFzGslmEZY+AZQ5YCDEd2CNgwHEroSUAl7FsFmHFA7CkoIUQ00Bnf5C6avO+57SV0CnfuZVSP1dKHVRKPZdwWbNS6l6l1EvW16bCHqaYTMgKopntAza3lRGwEGI62NcXYlVbPVCeKeibgQsnXPZZ4H6t9VLgfuu8KDJJQQshRHKhSJTu4TCr2hqAMhwBa60fBnomXHwZcIv1/S3A5Xk+LpGGrFZBx/cBSwpaCFHZ7PnfFbPrUMp5e4GznQOepbXeB2B9nZm/QxLpsldBV3vSD8Bul6K6yiUtCYUQFW9fv1kBPbeploaaKvrKbQScK6XUNUqpjUqpjYcOHSr0000roUiU6ioXLpfK6H4Bn0e2IQkhKp5dBautsZqmWm/FjIAPKKXmAFhfDya7odb6Rq31SVrrk2bMmJHl04nJZNqK0CYdkYQQ04Gdgp7dUE1jbeWMgP8IXGl9fyVwZ34OR2RiJJxlAPZ6ZA5YCFHx9vWFmFHnw+dxWyPgMgvASqnbgPXAcqXUXqXU+4BvARcopV4CLrDOiyILRqJUZ7AAy+b3uWUELISoeJ39QdoaqgForK2id9hZKeiUVfy11u9MctV5eT4WkaFQtiNgn8dxFWGEECLfOvuCLJtVB0BTrbdiUtDCAWQOWAghJqe1prMvRFtjDQBNtVUMh6OMjjln+k0CcBkLRqIZ7QG2+b1umQMWQlS0vpEIwUiUOfEUtDd+uVNIAC5jwXA0ozKUNhkBCyEqXafVB7g9PgI2AdhJC7EkAJexUCRKbRYj4IDPw3B4zHHNqYUQIl/sNoRz7ADsrwJw1EIsCcBlLNs54Fqvh5iGUCRWgKMSQojS29c/XoQDxkfATlqIJQG4jGWbgg5YHZGkGpYQolJ19AXxul20+n1AYgpaRsAiD7JehGV1RBqRetBCiArV2RdidkN1vFRvY62VgpYRsMhVJBojEtVZp6BBRsBCiMq1ry8YTz+D6ZteU+WWFLTIXSiLXsC2gE9aEgohKltnXzC+B9jWVFslKWiRO7sXcLalKAFpSSiEqEhj0Rj7B0K0NRwegBsdVg1LAnCZCoXNCuZsK2EBshdYCFGRDg6OEtMcOQL2ywhY5EEwhxS0BGAhRCXr7Dt8C5Kt0WEdkSQAlyk7AGdViCO+CEvmgIUQlaez3xThmHQO2EGNaCQAl6lg2JoDzmYVtDUHPCIjYCFEBbJHwHYdaFtTrZf+YIRYzBlVACUAl6n4KugsRsBVbhdej4shWYQlhKhAnX1B6qo91FVXHXZ5Y62XmIaBkDPmgSUAl6mRcPZzwGDVg5YRsBCiAnX2heJNGBI1xYtxSAAWOchlERaYrUgjMgcshKhAk+0BBmjyO6sjkgTgMjW+Dzi7X6Hf65FKWEKk6b4XDnD2tx+MT/0IZ+vsDx4x/wvOa8ggAbhMhXJMQfutloRCiNS27O1jd88IBwdGS30oIoWR8Bh9I5HJR8C1zmpJKAG4TMVHwLkEYElBC5GWbmvrSvewBGCns/sATzYH3FgrKWiRB8FIlCq3osqdbQraLYuwhEhT91D4sK/Cuew+wJOloOurPbhdij5ZhCVyEQxHs04/gz0ClgAsRDp6rBFwj4OKOIjJjVfBOnIErJSisaZKRsAiN6EsewHbAj4Pw2FJQQuRji4r9dwlKWjH6+gLoRTMnmQEDKYvsIyARU6CkdxGwLVWClprZ1SEEcLJ4iNgSUE73r6+IDPrfEmn55pqvY7JZEgALlMj4WjWC7DApKDHYprRsVgej0qIyjMWjcVHTE554xbJdfZPvgfY5qSGDBKAy1Q+UtAwXlFLCDG5nrYt9J0AACAASURBVIQ36y4JwI63r+/IPsCJmiQFLXKV6yIsu4uSLMQSYmqJo94emQN2NK01HX3BI9oQJmryywhY5CjXOWB7BCzVsISYmr31aF5zjWxDcrie4TCjY7EUKegqRsdi8Y5ypSQBuEwFI1Gqc0hB++MpaAnAQkzFLsKxbGYd3cNhWbjoYPusPsBzpkhBNzuoGIcE4DIVysM+YIAhqYYlxJS6h0zaeemsOsJjMckaOViHtQd4sipYNidVw5IAXKaCkWh8Hjcbfp/MAQuRjp7hMC4Fi1v98fPCmcaLcEwxB2zVg3bCQiwJwGUq1zlgv1fmgIVIR/dwmGa/lxl1vvh54Uz7+kP4PC6arbaDk3FSS0IJwGUoFtOEIrGc9gHHtyFJABZiSt1DozT7vfE3dVmI5VwdVh9gpVTS2zTaHZFkBCyyERqzWhHmkIKutVPQDlgJKIST9QyHafH7aAl4rfOyFcmp9vVN3gc4UWON1RPYAZkMCcBlKJhjL2AAn8dNlVtJClqIFLqHwjQHvLT4TQq6S0bAjtXZF5pyCxKA1+Mi4PMcVmClVCQAlyG7F3AuARjMSmhJQQsxte7hMC1+LzVeNzVVblmE5VCRaIwDg6kDMDinIYME4DIUsgJwLvuAwSzEkm1IQiQXicboD0bio9+WgHMK+YvDHRgIoTW0pUhBg2nIIIuwRFaCYdNAIfcRsFu2IQkxhV4r2DZb878tfi9dQzIH7ESdfaYIR7ojYFmEJbJip6Bz2QcMJgU9LJWwhEjK3nLUaq2Abgn4ZATsUPv67T3AqQNwU62XPhkBi2zYATiXbUhgUtAyAhYiOXvLkb0FqdnvlW1IDtWRRhEOW1NtVTy7UUoSgMtQPlZBg52CljlgIZLptrYctQQOnwOWetDO09kXpLG2ilqryNBUmvxeBkJjjEVL2w9dAnAZshdh5bIPGCQFLUQq9mi3xT8+BxyOSj1oJ0rVBzhRk1UPuj9Y2nlgCcBlaCRPI+CAT1LQQkylZziM26VoqDHVk5qt1dCShnaeVH2AEzmlGpYE4DKUr33AtV6PpKCFmEL38ChNtV5cLlPa0K6GJfWgnafTKkOZDnsEXOqFWBKAy9D4PuDcfn0Bn5twNEZ4rLTzIEI4VfdQOJ5+hvFUdLdsRXKUodExBkJjU/YBTtQUb0lY2hFw6tlq4TjBcBSXAq87twBs9wQeCY/h9STvHiKmMNwFW26Hhrkw71Som1XqIxJ51DMcjo96YXwxlmxFcpZ9GayAhsQUdGl/jxKAy5DdinCqjh/pSGxJaDepFmkaG4UnboCH/wtGB8Yvb1poArF9mrkSXLlNFSS14wFz8jVAdQNU14Ov3nytbhj/3ldfuGOocN3DYVa11cfPx0fAEoAdpbPfFOFoTzcFbbckLPHvUQJwGQpGotSksdQ+lVlWybadXSPMbarN+fHKQiwKoX6obc7u/lrDC3fCvV+Cvl2w9PVw3pcgEoQ9G2DPE7DjQdjyG3N7Xz3MPWk8IC86O3/B8P5/h87N6d12znHw9l9B4/z8PPc00T00Sqs16gWz997vdcsiLIfptEbAc9IMwH6vaUYjKWiRsVA4Sk2O878AJy5owuNSrH+lizOXtubhyBwsFoMX/wgPfh26tkPbCXD05bDqcjNqTUfHZvjr52D3epi5Ct59Byx57fj1804GPmKCdO9O2POkCch7noB13wI0XPRtOPX9+XlNA51wwhXwxuvMKDzUP/41NGB9PwDBHnjyRvjZ+fCu35pgLFIKj8UYCI0d0dy9OeCVloQO09kXxKVgVp0v9Y0BpRSNDqiGJQG4DNkp6FwFfB6Om9fI4zu683BUDqU1vHw/PHAt7HsGZqyAsz8DL98H933ZnFIF4/4OuP9a2PK/4J8BF38P1vxT8pGsUtC8yJyOe7u5LDQA162C7h35eV1jYRg6APVzweMFTyv4p/gQdcxb4da3wi/eAP9wCxx1fn6Oo4LZ84NHBGC/T1LQDtPZF2J2fTWeDNbFNNVWyRywyFy+AjDA6UtauH7dDoZGxwj4KuzPYdd6Ezh3P25Sr5ffAKv/wQTO134eeneZdPLzd0wejP0z4LHvw2M/AB2DM/8fnPkJM6+aqep6s1BroCM/r21oP6Chvi29289cAe+7F379Nrj1H+CS78Oad+fnWCqUnWZuDRwegFv9XvZZc47CGTr7gmmnn22NtV5JQYvMjYSjOdeBtq1d3MIPH3iZp17t4dwVM/PymOl4vrOfGXU+Ztalt2oxI/ueMfOjL98LgVnwhv+CNVeakWKipgVwxkfNabJg7K2D8CCsehOc/5X0U9XJ1LdB/97cHsM20Gk9ZnsGzz8H3vsXuP1K+OOHzbGc81kzYhdHsMtQ2sU3bM1+L8919pfikEQSnf1BVs9tzOg+TbVVvNo1XKAjSo8E4DIUikSPSItla82CJrweF4/v6CpqAL7y50/yulWz+cabjs3fg3a9ZOZ4n78Dqhvh/K/CKdeAN40FZpMF4wPPwUlXwfzT8nN89e3mw0E+2CPpdEfANl8d/ONv4E8fh4e+ZYLwJd8Dd1V+jquC2FuNWiaMgO2OSFrrnHciiNzFYpp9/SEuXJXZh/lmv5fNu/sKdFTpkQBchoLhKDWN+RkBV1e5OXF+U1HngYdGx+gaCrOnZyT5jXp3wUt/M/O3/XsBbeZzp/ra8wp4aswc79oPQU1mn4jj7GCcbw1zYfiQ2cLkSW+xSFL9VgBuyGAEbHNXwWU/gsZ5sO6bMNgJb7slu9R6BeuaUAfa1uL3EolqBkfHqK+WDy6l1j0cJjwWS7sKls1ehFXKD1ISgMtQPueAwcwDf/e+7fQOh+P74wrJ3jRvtw8DzKKi3Y/DS/eaU9c2c3nTQph5NKDGU6VKJZxPuHzlJXDahyAwo+CvISv2aHWg0yzQysVAJ3gDZptTNpQy6ef6dvjTx8zirHf91qSp86G/A3Y+Avu2wOkfyd/jFlHP8ChulzoiyDbHq2GFJQA7QCZ9gBM11VYRiWqGw9GSrX+RAFyGQpEo1Tl2Qkp0+lEtfOdeeOLVbi48pvBvlHbgjfXtQW/8Berl++CVdRAeArcXFp4JJ74Hlr4OWpZUzhylPV870JGHANxhAnquP5s17zbB8fYrzTalK35niodkauigCbivPgKvPgw9Cau9u7ab4F5mv8fuoTDN/vE60DY7Jd0zPMqiVn8pDk0kiO8BbsgsBW0XH+odDksAFukLhqPU5nEEvHpuI7VeN4/vKGAA1hq6X4Y9TzB30wPc693AUlcH3AU0zDOrk5e+zhSq8Fbom1rDXPO1Pw8roQc6MluANZWjzof33m1WR19/GtS2QN0cs4Ctbo4pr1k3B+pmQ2C2+eqpNoVH7IB76EXzWL56WHA6nPw+WHgW7HwU/vpv8Oxvze+4jHQPh49IPwO0WIuyuqQYhyN09GVWBcs23pAhwrws6/LkSgJwmdFaW5Ww8heAq9wuTl7YzPp8zgOHh03hij1PmIIUe5+EYC8A7e561uvF3B55Df/wzqtYuuqkshsdZSWegs5HAO6EJVmMVJOZcxxcfT88fZuZEx7cb04HXzT7jXWSrllVtTB/rdnvvOhsmH0cuBPeVmatMovi/vKvpmjJVHuVHWZiHWjb+AhYArAT7OsLUlPljtd3TleTdfueEu4FlgBcZsLRGDFN3rYh2U5f0sI3/7KVg4OhzLcG2ZWf9j5lTnuegP3Pjb9pty6HFRdb5RhP4Qv3D/P7p/cBcJKax9LpEHzBjOyrG3MPwNGICY6ZroBOpWEuvObTR14ei8JINwzug8ED5mt4CNpPhLY1R27vSuRyw6U/hP8+ywTht96U32MuoO6hUY6dZGtLs3REcpTO/iBzGqszXkjV6ICWhBKAy0wobFoH5nMRFsDpS8zIZP2Obi47PkVqc3TI1CDe+xTs3Wi+Dh8y11X5Ye6JpmjFvFNNHeQJdZf39q9nyQw/Ow4Nx+dvpo2GubmnoIcOkFERjly53BCYaU7ZzFDMXAFnfQrWfQOOfRssvzDvh1gIyVLQ1VVuAj6PVMNyiM6+UMbpZxgfAZeyIYME4DIzEhkDyGsKGuDotnrqqz2TB+DomEkj7n4c9jwFB583laEAWpbCURdYDQdOgRkrD09BTqKzP8ia+U3s7Q1OvwBc3577CDi+BWlu7sdTLGf+P3jh/+DPnzBzxA7f8jQ6FmUwNDZpAAYzCpYUtDN09gU5d3nmNQwaauyWhKWrhiUBuMwEwyatm+8RsNulOHVxy5H7gQc64XfvM8HXV2/Sjmd/GuaebL7PsKtQNKbZ328+sbY31tDZN81K+tW3QcfG3B4j2yIcpeTxwqU/gpvOh/u+Ahd/t9RHNKXeYfOm3DzJHDCYACwdkUovPBbj0NAoc9LsA5zI43ZRX+2RFLRIXzBiAnC+54DBzAPf+8IB9vZa7Qlfug/uuMa02rv8Blj9dnDl1oWpa2iUSFTT1lhDW2PN4XuBp4OGdjOfGglCVeZpMyChDGUZBWAwUxOnfhA2/BiOeQssPKPUR5RUlzW/2+KfvGBKa8AbX30rSufAQAitM98DbGvyl7YedE7vpkqpnUqpZ5VSTyulcvxYL9IRsgJwvlPQMD4PvOGlA2aUcutbzFaUax6C49+Zc/CF8T3A7Y01tDVWT8MUtJU2toNoNgY6zFx7dZaVvkrptZ+HxgXwp49CxLkBLFkZSpsZAcsirFKz30/aGrILwKYhQ+lGwLm/o8K5WuvjtdYn5eGxRArBAi3CAlg2K8DK2kHWPPhuePQ603Lvn++HGcvy9hx2wLVHwAcHRxkdS7LFpRI1JBTjyFa+inCUgtdvOjF1vwwP/UepjyapeABOMgfcEvDRa5UxFKUz/n6SXVOXptoq+sp1BDwdaa25feOeeIpqihuaPZTdO0wz+DyxU9C1U42A926Ee78EW/8Mo4NpP7Z66V5+y6eZFXwJ/eafmu0j6TQyyEDiP4ydNjrQP41GEnbxjFxWQg90ll/6OdGSc+H4K0yrx3w1p8izVCloux70QGismIclJrDbQmabgm4u8Qg41zlgDfxNKaWB/9Za35iHY3K0HYeG+czvtvCJC5bx0fOWHn5lLGqKTrz4J9j6J+jbbS73BmDWMTD72PHTzJVZzQGmnAPedg/89koYs9J7rirTzeeo80zFo1nHHDlyikbggX+Hx75PpG4Zb+m6hp/NeQOLMz661Dr7QtRVe6irropvHejoCzK/Jb+B3rHixThyaEvY32GCWDl7/ddMu8g7PwxXP5hy5Xyx9QyH8bgU9TWTH1fiXmB7Na0ovo6+IM1+b9ZrYkxDhvJdBX2G1rpTKTUTuFcptVVr/XDiDZRS1wDXAMyfPz/Hpyu9zbtMNadt+62R5VjYlOJ78Y+w7W6zH9bthcXnmr2PSsH+Z83pmf+Fp35q7qfc0LrMCsjHwLKL0kr1hsJTzAE/fRvc+SGYsxre8Wsz+n75PtNR6L6vmFNgthWMzzPHGBmB311limec+F76T/4Cr3zvCR7f0c3iGYE8/MQO19EXjAde+1PrtJoHrqoxpR6znQOOjsFQAYpwFFtNE7zhP+H2f4L1PzTblBzErgOdrLhDS8CMjHuGwyx2aO+P6WBfXzDr9DOYFPTQ6BjhsRheT/ETwjkFYK11p/X1oFLqDuAU4OEJt7kRuBHgpJNOKvsJk027eqklxMy998Dvfmxa5o0OmFHu0tfByovNvtjJ9jnGYtC3czwg738Odj0Oz94OD3wNXvtF00bPlfzTnD0CPmIOeP2P4a+fM+UA3/Fr0/e1vg0WnQUXfBUG9sGOB0xA3vpnePpWUC5T01e54C03wbFvZYHWzGmoZv2Obq44bUEef3JGZ18wHnjt4unTKgCDSUNnm4IeOmD2YJd7AAY4+jJTIW3dt2DlpabxhkN0D4en7Lltzw1LPejS6uwLsSCH7Fmjf7wa1sz67AN5trIOwEopP+DSWg9a378OuDZvR+ZQtTv+zAbf96kPBtGvtKCOvtS8eSx6DVSl+AW6XNC82JyOvmz88oFOuPvTcO8XYftf4U0/gcbJswUjE/cBaw33XwuPftccx1t+Nnmv2fo5cMK7zCkWhc6/m2DcuwvO/lT8zU8pxdolLTy07RCxmD6iE0yuOvuCnDDfrN6trnLTGvDS2T8dA/Ce7O4b34KUp0YMpfbG78CPToE/fhSu/FNeVtrnQ/fwKK2B5D2bpR506Wmt6ewLsnZJS9aPEa+GNRIpSQDO5a99FvCoUuoZ4Engz1rre/JzWA4UjTD653/jy8Fv0emZxzvCX+DFf9wIl/0Ylr0+dfCdSn0bvP1XcNn1ZlHKT84w6eRJVljaI2Cfx2UC6Z8+ZoLvie+Bt92cXqN3l9tUrjrnsybYTxh5nL6kle7hMNsPpr+AKx0j4TF6RyKHLZgwe4Gdux2lIBraoT/LOWB77rhSAnDdbDMfvOtRuOvj8YYdpdaTYgQs9aBLr7M/xODoGItnZN89ze6IVKqFWFkHYK31K1rr46zTKq311/N5YI4yuB9uuRTfU9dzy9gFPHXurWyIHc22QyP5ew6lzOj0g4+aDjL/9wGzmGqk57CbhSJRqqtcuGJh+O17YPMtcNYn4eLvTZm6zoT9ifLxl/PYHQniVa8S67a2NdRMzxR0qM90jMpUuRbhmMoJ74bTPgR//x/44Ymw+Zd53TmQjZ6hyTsh2XweN3VSD7qknt3bB5h2qtmyOyiVqhqWM/I9TrbzUbjhLNj3NHcvvZZrY1dx6YmLqHIrtu0fyv/zNS2E9/wZzv8KbL0brl9rKlJZguEoLZ4w3PpWs/Dr9d+A876U1z2h7Y01LGipPbIsZY4S9wDb2hpNAJ5W+ynt0Ws2C7EGOsFTYxYxVQql4MJvmIIvLUvhjx8xJSs7NpXkcEbHogyOJq8DbWsOSD3oUtqytx+PS7Fidl3WjzE+Ai7NSmgJwMlobfYp3nIpVDfA1Q9wa/BUVsyuo6GmiiUzAmzbP1CY53a5zarQqx8wb7S3vgX+/EkIj+AKdvMzvgo7HzPlIdd+qCCHcPqSFp54tZtoLH+BcfIAXM1IOEp/sHRbAYrOLsaRTRq6f6+5fzkW4Uhlzmq46h54043mdf70PDM3PJzfD4Kp2EG1OckeYFuz30v3sKSgS+XZjn6Wz67LqSxv2aagK1qoH35zhSlmsfJiuOZBoq0reHp3HycuMCOPZbPq2H6gACPgRHNWwzXrYO2H4amfwX+fxb+88iEW6z3wjltNecgCWbuklcHQGM939uftMTv7grgUzKobf2NL3As8bdTnUA2r3ItwpKIUHPd2+PBG8+Hy6Vvhh2vgyZ+aNQ9FYDdZmCoFDaZIhzRkKA2tNc929LN6bkNOj1PjdePzuEq2F1gC8ET7n4Mbz4Ht98DrvwlvuwV8dWzbP8hwOBoPwMtn19HRF2QgVOBfXFU1vP7r8E9/hEiIurEevlR/LSy/qKBPu3axNQ+cxzR0R1+I2fXVeNzjf3bje4Gn0UKseDGOLFPQlbIAayrV9ebv/gOPmQ+id38KbnwN7N5Q8KfuTlGG0tbi98occIns7Q3SNxLh2Pbc66E31XpL1hNYArAtFoO/3wo/Ox/CI3DlXbD2X+Kpvk27zerMNfOtADzLzDu8dCC/K4WTWvwa+PCT/L9Zv+CV2uMK/nQz6nwsmxXIawBO3ANsm5bFODw+8M/IPAUdi8LgvsoeAU80c4X58Pm2m82CxJ+/Hn7xRlj3H7BrvSmEk2c9Vlq5ZYptSOZ688Y9rdYvOMSWvSYzl+sIGMxCrFLNATur/luxjYVh58Pw4l2mitXQAVh4Frz15xA4vMHz5l29zKjzMbfJBIzl1sT/1v2DnLggs564WfP6ORgLEPDlvxHDZNYubuH2jXvzViWmsz/IcRNWLLb4vXg9rukVgMGMYjNNQQ8dAB2dHiPgRErBqjeZQjeP/wi23gXrvgnrvgFVtabU6sKzzF78OcflXNbSTitPtQ3Jvn4sphkIjtFQK+Uoi2lLRx9et4tls7JfgGVr9ntLtgp6+gXg0SFTg3brn2H732C037R2W3o+rLjE/KNP8g+8eXcvJ85vipema2+swe91s31/kUbAlmA4yowUn8zzZe2SVm5Zv4tn9vZx8sLcPmTEYpp9fSEuOubwEbDLpWhrqJ5ec8AADXOh55XM7lNpRTgy5fXDOf9qTiM9sOsxePURUwr2/q+a2/jqYcHpJiCveIMpepOh7uEwVW5FffXUb492oY6u4VEJwEX27N5+Vs6py8vAoKnWy9ZCLahNYXoE4OEuM8J98S54ZR1ER0093qMvMUF38WumbIxwaHCUXd0jvOvU8epULpdi2ew6thY7AEeiBekFPJnTFjejlNkPnGsA7hoeJRyN0T5J3VZ7K9K0Ut9mgkcm7BHzdEpBJ1PbDCsvMSeAoYOw85HxgLz9HlNZbvU74DWfzigQdw+NTlkH2maPkHuGwyyRetBFE4uZBViXHZ+f/wNJQRfKwD5TY/mZX5v6uQ3z4eT3mfqz805NO1W12Zr/tRdg2VbMruOe5/ajtU75z5ovwXC0IL2AJ9NY62VVWz3rX+niYyxNfYcp2IusJmsb1tZYw6MvdeX0+GWnvt1kX0YHTd3udNj1oxvmFu64ylVgJhzzFnMC6NsDT9xgdg9s+Y0pcnP2p5OWeE1kqmClzjJJNazS2NUzwmBojNV5WIAFZgTcNxIuSOndVCpzEVZ42CzS+OEa8893yvvh/Y/Ax7fAhd+EhWdkNE+0eXcvXreLVW2HT/gvm1VH70iEQ4PF+wcMRqI57XvL1OlLWtm8q49QJLctIJPtAba1NdZwYDBEJFra6kdFZQfRTJoyDHSY5hmVVISjUBrnmVXUH3sGTv5n04nsB2vgrk+k/Jl3DYVpTbEFCcZT0LISuri2WBWwjmnPfQEWmBFwTMNgCXo7V9YIOBaDLf9rmhMM7jMND87/SlbzQIk27+plVXv9EYHPXoi17cBg0Qp5h4qYggazEOvGh19h065ezjiqNevHmSoAtzdWozXs7w8xr3m69AVO2As8c0V697H3AFdiEY5CqZsNb/g2nPExeOS/TJnLv/8KTnqvKXZTN/uIu/QMjbK6ftisETnwLBx43mxPHOk23caWXQhLL6DJ32DdXgJwUsE+U99+39PQ+bT5Gz71Glj15qz/jp/d24/P42LprPy0S00sxlHsufzKCcCvPgx//Tzs3wLtJ8JbfwEL1ub8sOGxGFv29vPuSVrz2VuRtu0f5KylhZ8EikRjRKKa2iKOgE9e1IzbpXh8R1dOAbijL0jA55l0YUviVqTpE4DtvcAZjoCn6wKsXDW0w8XXwRkfh4f/0xT22HSzGR0ffTl0bYcDz8GB57hz5O807RiEHdZ9G+fDrGOh7QTT0vP5P4By4Zt7Ch+vXoy76yLQR8kHo4nBdt/Thy80bJhntuD97irzs3/Df8GM5Rk/zZaOfo5uq6fKnZ8EbpPf7ogUZiHZN3bIRvkH4K6XTMWqbXebX/BbbjKfrvLU1uyFfQOMjsWOmP8Fs0+wNeBjW5EWYtlp4GKOgAM+D8fNbch5P3Cn1Th7srnyeACeTm0J69sAlWEKuhMWnFGwQ5oWmhbAZT+Csz4BD30bNlwP639krvPUEJuxknuiJ9G+4mTOPutc0xilOiHVGYvBvr+btqHb7+Hj/Bpe/DV8b77pirbsQlh4JsQiZqX2SHfC124IJnwfHjFrUgpcVKegtIanf22yC4cF2/nQdhwc/y5oOx7mnAD+FrOXfdMvTJbyJ6ebamdnfwZ86Y1mozHN8x39vPXE/K2DaKy1ewIXfyFW+QbgkR7TyHvjTaY4/XlfgtP+ZcrVzNnYtMsqwDFJAAZYPjvAtiIV47BbERZzDhjMPPBPHtrB0OgYAV92fzKdfaFJ089gOiLZt5k23FUQmDXeXjCV6ViEo5CaF8ObboCzPgUHn4cZK6FlCfsHwvzbtx7gm8uOhQWTLNhyuUyGrf1EOPdzvO+Hf+TU6CaumfWSKZv51E+nfl7lgppmswsjEoTb3mGC1IXfPDzQl4PRQTOn/uztMPdkOOEKmHO8OfmT9Oh1ucezDvd92dTbf/Z3Zr7+6MtTZhFe7RpiOBzl2Bw6IE1UynrQ5RuAH/q2+WNfcyWc+7kjCmfky+bdvbQ31jAryRzv8ln1/PrJXUVZQRcKm0VKxVoFbVu7pIUfPfgyG3f2cM7y7H7OnX1Bjk1StabG66bZ752Ge4Hb0y9HOXQQYmMSgPOt9ShzssTrQKcowmFzNbTxh55GrnnnVyESMt3T9j4J3oAJsrVWsK1tMYvnqhvHs3NjYXj42/DId832yEt/CEedl+9XWBj7noHfvhd6X4VzP29aombSDtXfanqpn/BPcPcnTWvVxeeYtHRr8h0X+ayAZWuqtVPQEej8u1lBf/SleXv8qZRvAD77U3DilTBzZUGfZvOu3in3wC6fHSAUibG7Z4SFrYWdPxiJmFV6xUxBw/gf+3Md/VkF4FAkSvdw+LA+wBO1NVZPz73Ah7and1s7UMsWpILqjpehTC8At/i9/H23WZVLVbUp6LP0/PSezOOF134Bll1k+n//6s1w0lVwwb+nnZItOq3N/PnfPm8+VFx5l9lVkq35p8LV62Djz82W0evXwukfMe/v3iPfT7fs7aemys2SGfn7+dRXVzFT9XHKli/CfXdBy1Gw4o15668+lfINwP5Wcyqgzr4g+/pDrJmfPN2xfHY9YFZCFzoAB8PWHHCRR8B11VUsbKnluY7sqsWMr4BOvlK8raGGnd1ZNKgvZ/VzYceD5k0t1QIeKcJRFOMj4PSqzbUEvPTmuod07onw/odNAFr/Y7PQ67LrcwtshRDshTs/bEqBLn09XP6T5KnmTLg91sroy+HeL8Oj3zXbR+evNdM0gRnW15kM7tzHmbObcBMD8vA+ODaKa8NPeMD3LaoPReD0D5v94kUIeF3kHAAAHShJREFUvlDOAbgIxgtwJB8BL51pPolt2z/I61cduaUhn0o1Bwywqr2BZ/b0ZXXfeBGOhqlGwDU89nJXUYualFxDO4SHYHQg9fzfdC9DWSTxXsBpjoCb/T6iMc1AKBJfzJOVqhozD7rijfB/H4Sb32jWtJz3xbyva8nKnifN6uXB/fC6r5vFU/n+Pw3MhDf9BNb8Ezz0LejYaKZeIiPxm3zH/uZaZQZggVlm/nnlJab8qCfN34HWZuHuXz8Pva+yxX0yf5v3Ub7yuuKknm0SgKewaVcvNVVuVsxJXqnI7/Mwv7m2KAuxSrEK2nZMWwN/3rKP/pFIxnvlptoDbGtvrGE4HGUgNEZDzTSpq2uPZvs70gjAe8HtM2k/UTDdw2G8bhd1aS42tAt2dA2FcwvAtgWnmxaM930ZNvwYXvqbWSw296TcHzsbsRg89j0zOm+YC+/7q1mAVkgL1sI/3Tl+fnQIhg+ya9erfOO3D/MvJ9VxXFPYNCcZ6IQtt5uV1dUNJp2/8mJYch54k2xpPPAC/PXfzLz7jBVwxR/4zr01VI8Vvy6VBOApbN7Vy+q5DSn3my2bVVeUrUhBaxFWbQkC8Ko2k2p/vrOf0zPcD9zRF0QpmN0wRQo6YS/w9AnA1nzuQAfMOnrq20oRjqJItw60LbEedN74AvDG75iSuXd+GG66wGyxVK4pTsp8dXnA7TWr7N3ehO+rklzuO/wyj/fw6zf/0qTEV70JLvl+aVZq+wLgC/DEK1X8NRbiM2e9BhLngCNBE0xf/JMZ1W75X7MzZun5sPJS00WrptHsnHnwG2a+2VcHF33bzLm7q2h67Ck6SrALQwJwEqFIlOc7B7jm7NRVtFbMruPBbQcZHYvi8xQuONop6GLPAUNiAB7IOAB39gWZVVc95QcZe364sy/Iyjn12R9oOWmw0snp9AUe6JT0cxGYOtDpj2QLWg96ybnwL4/Do98zv38dS3LS1teo2a4WDcPYqNkmFItANGIui3+d8P1UPNVw8ffgxPeU/MPfcx39BHweFrVMWGtTVWP2Ui+/yLyuXY+ZYPziXearq8rMp3c+bX4mJ11lds7Ujk8tNtZ6eb6z+B2RJAAnsWVvP2MxPWkBjomWza4jGtPsODjM0W2FCx6lnANuCfhoa6jmuc7+jO/b2R+ccgEWEF8hPa1WQgdmm1FLOluR+jtM31tRUF3D4bRXQEMR6kFXN8D5Xy7MY4MJ3ocFZit425f5Wwu2xTNTW/b2c0x7/dSL3dxVZjvT4nPgov8088gv/hG23WPS+Bf8+6TZpqbaKtkH7CR2AY4T5qcOwCviNaEHChqAQ+HSzQEDHN3WwHMdWQTgvlB8BJ1Ma8BHlVuVJA1UMm6PCcKpylHGYjDYOT5iFgXTMzzKopb0y6HaRRzymoIuJqVM2jndxUslEonGeGHfAO85fWH6d3K5YN4p5vS6r01508ZaL6FIjFCRm91UZjekPNi8u5fFrf600lGLWv1UuRXb9g8V9JjiI+A8NKHOxjHt9bzSNcxIOP2uIVprOvqCUy7AAtNfeU7DNOwL3NCeOgU9fMgqwiEBuNC6h8K0BNLbggTg9bior/ZIS8IC235gkPBYLG8dkCYqVTUsCcCT0FqzeVdv0vKTE1W5XSyZEWDb/sLOIYyEo3jdLjx5KkKeqVVtDWgNL+5L/3V2D4cJj8Vom2IBlm16FuNIoxqW7AEuilAkykg4mtEcMJjpGWlJWFjP2hWwChaArWpYw8WtBy0BeBK7ukfoHg6zJo30s23ZrDq2HyjsCNikR0r3Kzum3aSRMynIkc4WJFtb4zQcAde3mwCrdfLbxAOwjIALyQ6i6fQCTtTi98YLeIjC2NLRT121hwUZTA9kYrwhg4yAS268AEf6AXj57Do6+oIMhAr3CSoYLm4v4Ilm11fT4vfyfAYLsTIJwO2NNewfCDEWjWV9jGWnod0UGgj2Jr+NFOEoCjuN3JxmFSxbs99bvnPAZeLZvf2snttQsCI94y0JZQRccpt29VLn88SrXKXDXoj1UgELcnT2B+NzFaWglOLotvqMRsD2oqqp6kDb2hpriGk4ODiN5tPsoDrVQqyBDrMvU4pwFJQ9As48Be2N15AW+Tc6FmXr/gGObc9fB6SJZA7YQTbt6uX4+Y0Z1XZdNssE4K0FKsgRHouxcWcvpy5KXhazGI5pb+Clg4OMjkXTun1nX5CaKjeNaVTPapuOW5Hs5gpTzQP3d0DdnLz1uBaTs9PImaegffSORIjFpphGEFnbtn+QSFTntQPSRPb7k6SgS2wwFGH7gcGM0s8Ac5tq8HvdbC9QAN6yt49gJMraJYVtQJHKMW0NRKKal9Kc7+7sM3uA00kdtVt7hadVW8J4OcopVkIPdEoXpCLoGbZT0JkF4Ga/l2hM0x8sfkP36cBuQXhsgRZgAfg8bvxet6SgS+2ZPf3EdGbzv2DSs8tm1xVsBPz4jm6UgtMWl3YEbO/nTXc/cGcaW5BscxrsEfA02gscmGXKB6ZKQcsK6ILrHg7j9bgIpFkH2mYX7pA0dGE8u7efptoq5jYVtilFY62X3iLP5UsAnmDTrl6UguPnZT7fsGJ2HdsPDKKnWtGapfU7ulk5uz4/Bd9zML+5ljqfJ+2ybR19obTmf8E0tmisrZpeKWiX26SX+5ME4FhsvA60KKjuoTAtGdSBttmtC2UldGFs6ejn2LmNBe+S1uQvfjUsCcATbN7dy/JZddRVZ94QYNmsOnpHIhzK8yKiUCTKpt29nL6k9ItwXC7Fyrb6tEpShiJRuoZG0x4Bg2lZOK0CMIxvRZrMSJep51svKehCy7QOtC1eD1pWQuddKBJl+4HBgu3/TdRU65UUdCnFYprNu9MvwDHR8nhJyvymoTfv7iU8FmOtAwIwmHngF/cNEE2x6GR/v9UHOJMA3FgzveaAwYxukwVgKcJRNN1DoxlVwbK1BiQAF8oL1vvMsQVcgGVrrPXKIqxSevnQEIOhsYwKcCRabq2Ezndrwg07unG7FKeUeAW07Zj2ekKRGK8cmnoh1vge4NRVsGzt07EaVoNVDWuyqYv4HmAJwIXWPWxS0JlqslsSSgo67+IVsIoQgE1DBhkBl8zmXZkX4EjUEvDRGvDlPQA/vqObY9obskqLF8KqNvPPkCoNbY9k050DBjMCHgiNMVjAgiaOUz8XxkKmX+lE/VIFq1jsOeBMVbldNNRUySKsAtiyt5/WgI/Z9el/iM9WY62XgVAkZWYvnyQAJ9i0q5dmv5eFOZQ7Wz47kNcU9Eh4jKf39Dli/te2ZIYfn8fF8ykKctirmWenUQfaZqer9/VPo5XQ9uh2YJKtSAMdpp+pf0Zxj2maCYajBCNRmjPcA2xr8XslBV0Az3b0FbQCVqKm2iq0pqjbySQAJ9i8u5c183Nbbbd8Vj3bDwzmbVP+Uzt7GYtp1i52TgD2uF2snJN6IVZnX5AZdT58nvTLZ9oBeFrNA9ttBidbCT3QCfVShKPQ7NFra4ZlKG0tAa90RMqzkfAYLx8cKuj+30SlqIYl/9WW3uEwOw4NZ70Ay7Z8doBQJMbunpG8HNf6Hd1UuRUnLcztuPJtVVs9z3cOTLnlqrM//T3AtvbpWA3LXuE82UKsgQ5JPxeBvYUom1XQ9v2kHnR+vdA5QEwXtgBHolJUw5IAbHlo+yGAnEs9Lp9tClXkKw29fkcXx89rpNabWXGAQjumvYHB0Bh7epIHyo6+YLy6Vbpm1PnwuNT0CsD+GSbNLAG4ZOzgmW0Kutnvk33AeRavgFWEBViQMAIuYktCCcCW32/ey7zmGk6Yl9tI027gkI+FWAOhCM929Dsq/WyLV8RKkobWWpsqWA2ZjYDdLsXshurpVQ3L5TJp5okpaK2lCEeRdA3lloJuDXjpHQlLPeg8erajn1n1PmYVYQEWSAq6ZA4MhHjs5S7edHx7Rg0YJuP3eZjfXJuXEfBTr/YQ05S8/vNkls2qw+NSSVsT9o5ECEViGaegYbruBZ575Ah4uAuiYRkBF0HuI2AvMQ19Ug86b7bs7StoB6SJGuMtCSUAF9WdT3cQ0/CmNfmpNrRsVl1eRsCP7+jG63Fxwvzi/RGmq7rKzdJZdUlbE2bSB3ii9sZpWA2rYZJqWPb5BgnAhdY9HMbnceHPst+2XcBDFmLlx2Aowitdw0XZ/2ur83nwuFRR9wJP+wCsteb3mzpYM7+RRa3+/9/euQfHVZ0H/PdJu1q9LcnCtla2sBxjG9kY4xoMFBKTBAwEQiAwkCb0MTCZPpwWUtqQSUv5J9MCySTNTJmGoQxJm6RAMY9Q82gpGaitmBjhJ8FYsl0sybasNZKsh/Xa0z/uvdJK3tVj9+7j6n6/mR1dnT33nKPvXJ3vnu985zuulLlqURlHOvtmfGRfIhpbIvxOXSWFweQGhXRjOWJ1x3XESmYPsEO4opAT3Wczuh8v65SHLXNzNDqepkE4MkaycaAd5ms4SlexHDwzt/4L1oE6mY6G5XsF/MHxHg6ePOPa7BeskJSjUUNLR1/SZXzSN8QHx3tyav/vZNaEy+nsHaIjTuzrZKJgOYQrihiJGtdjauc05Ystc3N/53hajwbhyBSn+5ILQ+kwdiKSOmK5wr4MHEEYj8rioDphZZIXmtoI5gs3r61xrUwnJvRHKawD7zwSAciZ+M/xWGP/c8Q7mrC9a4BQIC+pbR3+3gscE4yjp806qlCDcKSdSJIHMTg4957WaFiusLetm9qKIqpTeClKButABp0BZ4SR0Sgv7m7nc6sWunrMX311CcF8Sels4B0tEYoL8lm7OPfWfx0urClHhLhHE7bbxxAmY9Lz515gWwE7ZmfnuixsHVmopJVkw1A6OB60aoJ2h32tXRmf/YK1F7grg2vAubW5NMO809xJZ+8gt61318QXzM/jU+eVpjQDbmyJsGFpFQWB3H1HKgkFqK8uiTsDbuuafRAOhxo7dKU/FXCMI5ZuQcoYkb7BMTNyMgTz86goDqoJ2gUivYMcjfRzx4YlGa/7kS+vJZjBMTd3R/cMsLWpjcriIJtWLnC97JWLkveEPnVmkEMdvTm9/uuwOjwvwQx4IKn1X4CywiDlhQF/KeCSasgPTTRBd7eqAs4A/UMjnB2OUpXkHmAHjYblDk+8fRgRuLZhYcbrriwpoDSUuXmpbxXwmbPDvHHgBDdfHE7LLHPFwjLaugboSeJUn8bD9vpvDgbgmMyacDltXQN8EjPwDI6M0nFmMOkZMDh7gX0UjENk4rnAThAO3YKUdpxZayozYLCCeHTqNqSUON49wNM7jnLrJbWssI93ncv4VgG/uu8EgyNRbnPR+zmWVbYj1qEkzNCNLRHKQoGxaFO5jOOIFTsLPtltDUKpKGB/7gVePL4G3H8aRgfVAzoDOOu2qawBg86A3eBHbx4iagz3f35FtpuSEXyrgJ9vamVZdQkXp2mfmfP2lowjVmNLJxuXVRHIz/3ucV4SYiNipbIH2CFcUUR7t88UcHl4PBzl2BYkNUGnG8dzORUvaLCiaKkTVvK0nOrl2V2tfHXj+SypSv5IWC+R+yN8Gjh2up+dR05z2/ratJ0zubiyiPLCAK/tPzHliUGTae8a4Gikn8s9YH4G6xDr2ooi9sfMgFOJguUQriiiq3+YvsGRlNvoGcpr4Uw7REd1D3AG6bRN0Klueakusbaw+CqAjIt8/42DFAby2PLZ5dluSsbwpQJ+abc1uH3pkvQNbiLCA5tX8s6hTv5t58czvq+xxVr/vTIH4z8nYk1tOQdiPKEdBex4MyeD48B13E+z4Hm1EB2BvlOqgDPIWBxoF0zQxmT2OLu5wp5jXWzbd4J7r16W8b2/2cR3CtgYw9amNjbWV7G4Mr1mjrsvP5/PrDiP7/7nBzR39M7onsbDESqLg2NryF5gdXgeRyJ99Nqz1fbuAapLC1IKoVk7FozDR45YjrLtbrPWgiUfSt330FcmEukdpDCYR3GScaAdxuJBqxl61jz6+odUlRRw79X12W5KRvGdAt59rIvDnX18OU3OV7GICI/dsZbiggD3PfM+QyPRKfMbY2hsibCxfn7KpzJlkjW15RgDvz1umaHbus6mZH6GcfO1rxyxxvYCt1pKuKxGg3BkgEjfEPNLQikvR43Fg9a9wLPifw91sr05wp9ds5yywmC2m5NRfKeAX3i/jVAgjxsuWpSR+haUFfL3t13E/rYe/vHNj6bMe+z0AG1dA1y53Bvrvw6rw7YntG2GTuYc4MksKAuRnyf+UsDz7JfC7jbLBK1bkDLC6b6hlLcgQewMWLcizRRjDI+89iG1FUV87fK6bDcn4/hKAQ+NRHl5TzubVy/K6JvW5tWLuOvSJTz+qxbePXI6Yb7Gw1Ygfi/s/41lQVmI6tIQ+9t7MMbYQThSU8CB/DwWlRf6Kx50USUEiizlq1GwMkakN7U40A7j8aB1BjxTtu07wb62bu6/dgWhgP+sPb5SwG8d7KCrf5hbXQ49ORP+9qYG6qqKuf+Z3QmDc+xoiVBdGmL5gtIMty41RIQ1teXsb+ume2CY/qHRpKNgxRKuKPTXDDg2GEdPmzpgZYjTKR7E4FBZbL3Uqwl6ZgyPRvneGwe5YEEpt6bRITaX8ZUC3trUSnVpiKuXZ97DuCQU4Ad3ruNEz1kefvnAOd87679XfGp+2rZGpZPV4XKaO3o53GkdwZjKHmCHcEUR7X5ywgLL7HxiH4ycVQWcAYwxdPYOuuJ5G8jPo7I4qCboGfIf77VypLOPv9q8knwP+by4iW8UcFf/EP/zYQdfWhfOWoCL9XWVbLlmOVub2nhlb/uE71pO9dFxZtAT8Z/jsSY8j5Go4VcfdgCp7QF2CFcUcbx7gKif9lWWL4ZIs32tJuh00z80yuBI1JUZMGg0rJkyMDTKD//7I9bXVWQl5nOu4BsF/Mu9xxkeNVkxP8ey5bPLWbekgu+8sH/CHlcvxX+OhxOS8o0PTgLuKeDhUeOv+LqxSldnwGlnLA60Swp4fmloLLCHkpifNB7lZM8g37p+lSctfm7hGwW8tamVVYvKaKjJbnzlYH4eP7hzHcOjUR54bs/Y7O7XLRFq5hVy/nxvhmBzIn99eOIMBYE8Vwa0Wnsd2VeOWLGezzoDTjuOudgNL2iwFLnOgKemu3+Yx99qZtPK89jo0QmHW/hCAR/p7OP9j7vSGnpyNtRXl/DQTQ1sb47w1PYjRKOGxsPeXf8FyxHL2Y4Unlfoyj7m8b3APloHLre3Ikk+lGVmq5yfGY+C5U70JTVBT8+P326h5+wIf715VbabknVSUsAicr2IHBSRZhF50K1Guc0LTa3kCdyyLndMendeuoRrGxby6GsH+eXedk73DXnW/OzgHMzghvk5thxfeUI7M+CyRRqEIwOkwwSt8aATc7LnLE9tP8It68I0eOC0t3STtAIWkXzgn4AbgAbgKyLS4FbD3CIaNWx9v43fXV7NwvLUt8a4hYjwD7ddRHlRkL98dg8AV3jUAcvBWQd2SwGXFwYpCwX8ZYJ2zM5qfs4IY0cRumiCNgY+0XjQcfnRm4cYGTV881p/HDc4HYEU7r0MaDbGHAYQkX8HbgE+cKNh0/HznR/z3HvHps03PBql9ZMBHrhuZQZaNTvml4Z47Pa1/NHTv6GuqjjtsanTzZpad2fATlkv7W5jT2uXa2XmNMbwcymk6VSI7z2+PdutmfO0dw1QFMynuCCVoXAcR5H//r+8SyjoixW+WbG3tZvfu6yO8+eXZLspOUEqT10tEKsBW4GNkzOJyNeBrwPU1bkXaiwUyKM0NLPmf/HiMJtX5+Z62jWrFvA3X7iQymJ33sCzybLqUu65qp4vXFTjWpl/cOVSXt1/3LXyvMCL5XfzcUH9jJ9vJXlWLCzjkiUVrpV3WX0V1zUsZGB41LUy5xI3ra3hvs9fkO1m5Awym7NqJ9wocgew2Rhzr/373cBlxphvJLpnw4YNZteuXUnVpyiKoiheQ0TeM8ZsiPddKjaSVmBJzO+LgfYEeRVFURRFiSEVBfwb4AIRqReRAuAu4GV3mqUoiqIoc5ukF5mMMSMisgV4HcgHnjLGnBvkWFEURVGUc0jJy8MYsw3Y5lJbFEVRFMU3qJ+8oiiKomQBVcCKoiiKkgVUASuKoihKFlAFrCiKoihZQBWwoiiKomQBVcCKoiiKkgVUASuKoihKFlAFrCiKoihZQBWwoiiKomSBpE9DSqoykVPA/7lYZDXQ6WJ5cwmVTWJUNolR2eQm2i+JyXXZnG+MOS/eFxlVwG4jIrsSHfPkd1Q2iVHZJEZlk5tovyTGy7JRE7SiKIqiZAFVwIqiKIqSBbyugJ/IdgNyGJVNYlQ2iVHZ5CbaL4nxrGw8vQasKIqiKF7F6zNgRVEURfEkripgEVkiIm+JyG9F5ICI/IWdXiUi/yUih+yflXb6V0Vkr/3ZISIXx5R1vYgcFJFmEXlwijpfE5EuEXllUvrP7Pv3i8hTIhJMcH+9iOy02/aMiBTY6Z8WkSYRGRGR21U2E2RTZ/8t79vtu9GHstli12FEpDomfZOIdIvIbvvzkA9lEzffVG3zGi73y1Mi0iEi+6epM27/JXoW49yf6P85l8e6bMvG1bHuHIwxrn2AGmC9fV0GfAQ0AI8CD9rpDwKP2NdXApX29Q3ATvs6H2gBlgEFwB6gIUGdnwNuBl6ZlH4jIPbnF8CfJLj/WeAu+/qfnXzAUmAt8FPgdpXNBNk8EXPdABz1oWwusZ+Ro0B1TPqmyWX6UDZx8yVqmxc/bvWL/fungfXA/inqS9h/iZ7FOGV4aqzLEdm4OtadU2+aH9KXgGuBg0BNTOccjJO3Emizr68AXo/57tvAt6eoZxNTDHrA/cB346QL1gbuQLx67bSn3Xgo55JsgB8D34pJ3+En2UzKM+Efe7oy/SSbaZ6vsbbNhU+y/RKTtpSplcy0/Tf5WZz0nefGulyQDWke69K2BiwiS7HePnYCC40xxwHsnwvi3HIP8Kp9XQsci/mu1U5Lph1B4G7gtThfzwe6jDEjqdYzyzYtxduyeRj4moi0AtuAbyRTf4I2LSX3ZTMdV4jIHhF5VURWJ1N/gjYtxUOymSZfbNs8TYr9MlNS7T8vjnUzJZ2yeZg0jXUAATcLcxCRUuB54D5jTI+ITJf/GizBX+UkxclmkmzO48Dbxph34lXtYj0zYo7I5ivA08aY74vIFcC/isgaY0w0yXZYFXpHNlPRhBV6rtdeL3oRuCDJNozhUdnEzRenbZ7FhX6ZcVVx0mbTf14c62ZcVZw0t2STlrHOwfUZsP3W+zzwM2PMVjv5pIjU2N/XAB0x+dcCTwK3GGMidnIrsCSm2MVAu4hslHHnli/OoC1/B5wHfDMm7XX7/iexzA4VIuK8iCwG2mf/V8+MOSSbe7DWTDDGNAKFWPFYk8ZjskmIMabHGNNrX28DglM5f8wEL8omXr4p2uZJXOqXRGUviemXPyZB/01ThtfHukRlZ1I2ro91E3DZ1i9YC/k/nJT+GBMX3x+1r+uAZuDKSfkDwGGgnvFF9dVT1LuJcx1G7gV2AEXTtPk5Ji6+/+mk75/GHceEOSMbLBPRH9rXF2I9rOIn2cTkP8rENeBFjiyAy4CP/SabRPkStc2LH7f6Jea+pUy9zjlt/01+FuOU4amxLhdkg8tj3Tn1uvxQXoU1dd8L7LY/N2LZ2N8EDtk/q+z8TwKfxOTdFVPWjVjecy3Ad6ao8x3gFDCA9Sa02U4fse91yn4owf3LgHftB+A5IGSnX2qX1wdEgAMqmzHZNADb7Qd9N3CdD2Xz5/Z9I1j/lE/a6VuAA7Zsfk2Kysajsombb6q2ee3jcr/8AjgODNvyvidBnXH7L9GzOIv/51we67ItG1fHuskfjYSlKIqiKFlAI2EpiqIoShZQBawoiqIoWUAVsKIoiqJkAVXAiqIoipIFVAEriqIoShZQBawoiqIoWUAVsKIoiqJkAVXAiqIoipIF/h9yKqu0vH7+FwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAeMAAAF1CAYAAADbSIJmAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOy9eZxcZZX//366u6rX7OkkkH2DBJIQSMAAQQK4ALI4Ciqigiio47h81dHRGfWrP2fc5iuO4yCiKILIoCACgg5GSIAxCSRAWJJAVtIdQpYO6aTX2p7fH8+91dWVWu69dWu7dd6vV7+qq+5ST1VX17nnnM85R2mtEQRBEAShfNSVewGCIAiCUOuIMRYEQRCEMiPGWBAEQRDKjBhjQRAEQSgzYowFQRAEocyIMRYEQRCEMiPGWBAEQRDKjBhjwTeUUicqpZ5VSh1VSn263OvxA6XUJ5RS+5RSPUqpceVej41S6hyl1MvlXkcmlFIrlFKdDvYb9nlRSt2slPpqKdYoCJWGGGPBT74IrNJaj9Ba/6jci8mHUmqXUuotObaHgB8Ab9Nat2mtu3x6zn7LuNs/P3ZwnFZKzbHva62f0FqfWOh6sjzXbUqpbxXj3GkM+7xorT+utf7/SvC8JUcpdblS6jml1BGl1EGl1F+VUjPS9pmplEoopW4qzyqFciLGWPCT6cBLXg5USjX4vBY/mAg04eE1KUO2/69LLeNu//xDQausXhx/Xkr1+VBK1RfhnHOA24HPA6OAmcBNQCJt1w8BbwDvU0o1+r0OocLRWsuP/BT8AzwKxIEBoAc4AfPFcztwAHgV+Begztr/WuB/gRuBQ8C3gEbg34HdwD7gZqA55TkuB54DjgDbgQutxz8MbAaOAjuAj6UcMx74I3DYep4nMBehd2C+DPut9X4x7fWcAPQC2tr+qPX4WcDTQLd1e1bKMauAf7VeVz8wJ8P7tAt4S5b3cA6w2jr3QeBu6/HHrXX0Wmt5L7AC6Ew77z8Cz1v73Yq5mPiT9b6sBMak7P874HXruR4HTrYevwGIAhHruR60Hj8euNf6W+4EPp1yrmbgNowh2WStozPTa8zzebkN+Ja1fQXQCXzJWucd1t/tn6y/fRfwW2Csg89mxtdqbbsN+AnwsPW+vSXPaz0DWGN9nvYCPwbCeZ7/CuA5B+vcDnwC89m/otz/0/JT2p+yL0B+gvNjGaOPpty/HbgfGAHMAF4BPmJtuxaIAZ8CGqwv9B8CDwBjrWMeBL5t7X+G9WX6VutLeTIwz9r2DmA2oIBzgT7gNGvbtzFGPWT9nAMoa9sushhGa/sMjBFssO6PtQzOB601X2XdH5fy+ncDJ1vbQxnOmfU5gbuAf7ZeXxOwPGWbJsW4k9kYr8UY4MnAfuAZ4FTMRc6jwNdT9r/Oeo8brff9uZRtt2EZRet+HbAB+BoQBmZhLnrebm3/DuYiZywwFXiRPMY4y+cl+bzW64sB37XW2Ax81nqNU6zHfgrc5eB58r3WbuBs63W25HmtS4Bl1t93BuYi8LN5nn8W5qLjRuA8oC3DPucAg8AY4D+BB8r9/yw/pf0p+wLkJzg/qV+uQL315XJSyvaPYXKEYIzx7pRtCuOZzE557Exgp/X7T4EbHa7jD8BnrN+/ibkgcOWlWttnMNwYfxB4Km2fNcC1Ka//m3nWtgvjCR5O+bne2nY7cAswJcNxTozx1Sn37wV+knL/U8AfsqxptHX+Udb92xhujN+U+reyHvsy8Evr9x1YUQrr/g34Y4wjQFPK9s3ABSn3j8N48Q0uPqOZXuvtTl9rhvN9FrjPwfMuw3jyBzCG+TZSjDLwc/vvg/ncR4EJTl+X/FT/j+SMhWIxHuNZvJry2KsYr82mI+X3diyvRCl1WCl1GPiz9TgYj2t7pidSSl2klFqrlDpkHXex9fwA3we2AY8opXYopf6pgNd0fNrryfeasvFOrfXolJ+fWY9/EXNR8pRS6iWl1HUu17cv5ff+DPfbwORFlVLfUUptV0odwRhyGHrP0pkOHG//Xaz3+CsYLxzM+5L6utPfI68c0FoPpK3jvpQ1bMaEuidmPBrHrzV17Tlfq1LqBKXUH5VSr1vn+zeyv29JtNZrtdbv0Vq3Y7zgN2OiICilmoErgTutfddgIizvz3deITiIMRaKxUHM1f30lMemAXtS7uu0/fsx+TzbSI3SWrdZ2zswoehhWEKXezG55ola69GY/J8C0Fof1Vp/Xms9C7gU+JxS6oIMz++E19JeT77X5Aqt9eta6+u11sdjogg3pSqofeT9mPz7WzB5/RnW48peStr+HZgIReoFxAit9cXW9r2YiyWbaT6tM9M6LkpbR5PWek+mgy3yvdb058n3Wn8CbAHmaq1HYgx16rnyvyitnwZ+DyywHvo7YCTm7/26Uup1zAXeh9ycV6huxBgLRUFrHceE5f5VKTVCKTUd+Bzw6yz7J4CfATcqpSYAKKUmK6Xebu1yK/BhpdQFSqk6a9s8jPfdiAn/xZRSFwFvs8+rlLpEKTVHKaUwwq+49QPGc5zl4mU9DJyglHq/UqpBKfVe4CSMQKxglFJXKqWmWHffwBgJr2vNxQhMCqELE434t7Tt6c/1FHBEKfUlpVSz5W0uUEqdbm3/LfBlpdQYa/2f8mmd6dyM+TxNB1BKtSulLs9zTL7Xmk6+1zoC8znqsT5/n8i3aKXUcqXU9Smf63nAZZj8N8A1wC+AhcBi6+dsYLFSamG+8wvBQIyxUEw+hckD7wCeBH6D+dLJxpcwIeW1VghwJXAigNb6KYxq+kaM4GY1MF1rfRT4NMYgvIHxhB5IOedc6zw9mPzuTVrrVda2bwP/YoUjv5DvxWhTZ3wJpkSlCxNWvkRrfTDfsWk8mFZnfJ/1+OnAOqVUj/UaPqO13mlt+7/Ar6y1vsfl86VzOyaUvAejfl6btv1W4CTruf5gXVhdijESOzFRjJ9jPE2Ab1jn2wk8glE+F4P/wLwvjyiljlrrflOeY/K91mE4eK1fwHzGjmIuHu92sO7DGOP7gvW3/TNwH/A9pdRk4ALgh1ZkxP7ZYO13jYPzCwHAVpUKgiAIglAmxDMWBEEQhDIjxlgQhKKglJqWFo5P/fFL5IVS6uosz+GpG5zHNZyT7bWWag1CdSNhakEQBEEoM+IZC4IgCEKZKVtz/vHjx+sZM2aU6+kFQRAEoeRs2LDhoNX8ZRhlM8YzZsxg/fr15Xp6QRAEQSg5SqmMHeokTC0IgiAIZUaMsSAIgiCUGTHGgiAIglBmypYzzkQ0GqWzs5OBgYH8OweYpqYmpkyZQigUKvdSBEEQhBJQUca4s7OTESNGMGPGDExf/9pDa01XVxednZ3MnDmz3MsRBEEQSkBFhakHBgYYN25czRpiAKUU48aNq/nogCAIQi3h2Bhbo8SeVUodMy5OKdWolLpbKbVNKbVOKTXD64Jq2RDbyHsgCIJQW7jxjD8DbM6y7SPAG1rrOZgRd98tdGHVwKpVq/jb3/5W0Dna2tp8Wo0gCIJQrTgyxtbA8Hdg5npm4nLgV9bv9wAXqBpw7/wwxoIgCILg1DP+IWaQeiLL9slAB4DWOoYZ/j4ufSel1A1KqfVKqfUHDhzwsNzS8M53vpMlS5Zw8sknc8sttwDw5z//mdNOO41TTjmFCy64gF27dnHzzTdz4403snjxYp544gmuvfZa7rnnnuR5bK+3p6eHCy64gNNOO42FCxdy//33l+V1CYIgCJVJXjW1UuoSYL/WeoNSakW23TI8dsw4KK31LcAtAEuXLs05LuobD77EpteO5FueK046fiRfv/TkvPv94he/YOzYsfT393P66adz+eWXc/311/P4448zc+ZMDh06xNixY/n4xz9OW1sbX/jCFwC49dZbM56vqamJ++67j5EjR3Lw4EGWLVvGZZddJrlhQRAEAXBW2nQ2cJlS6mKgCRiplPq11voDKft0AlOBTqVUAzAKOOT7akvEj370I+677z4AOjo6uOWWW3jzm9+cLDUaO3asq/NprfnKV77C448/Tl1dHXv27GHfvn1MmjTJ97ULgiA4JjYI3Z0wbna5V1Lz5DXGWusvA18GsDzjL6QZYoAHgGuANcAVwKO6wEHJTjzYYrBq1SpWrlzJmjVraGlpYcWKFZxyyim8/PLLeY9taGggkTCRfK01kUgEgDvvvJMDBw6wYcMGQqEQM2bMkNIlQRDKz9M/h5XfgH/cCk2jyr2amsZznbFS6ptKqcusu7cC45RS24DPAf/kx+LKQXd3N2PGjKGlpYUtW7awdu1aBgcHWb16NTt37gTg0CHj9I8YMYKjR48mj50xYwYbNmwA4P777ycajSbPOWHCBEKhEI899hivvppxaIcgCEJpef0FiA/CgVfKvZKax5Ux1lqv0lpfYv3+Na31A9bvA1rrK7XWc7TWZ2itdxRjsaXgwgsvJBaLsWjRIr761a+ybNky2tvbueWWW3jXu97FKaecwnvf+14ALr30Uu67776kgOv6669n9erVnHHGGaxbt47W1lYArr76atavX8/SpUu58847mTdvXjlfoiAIguHgVnN7IFvVqlAqVIHRZM8sXbpUp88z3rx5M/Pnzy/LeioNeS8EQSgqWsN3p8NAN5z5D/D2fy33imoCpdQGrfXS9Mcrqh2mIAiCUCL6uowhBjiwpbxrEcQYC4Ig1CR2iLplPBzIL1AViosYY0EQhFqkyzLGJ14E3R0weDT3/kJREWMsCIJQixzcCvWNMPet5r4oqsuKGGNBEIRapGsbjJ0FE6yeDpI3LitijAVBOJa+Q3D75XB4d7lXIhSLg1th/BwYMwPqw2KMy4wY4yKyatUqLrnkknIvQxDcs3sN7FgF2x8t90qEYhCPwhs7YdxcqG+A8SeIiKvMiDH2QDweL/cSBKG42Epb+1YIFm+8CokYjJ9r7refKJ5xmRFjnMauXbuYN28e11xzDYsWLeKKK66gr6+PGTNm8M1vfpPly5fzu9/9jkceeYQzzzyT0047jSuvvJKenh7AjFqcN28ey5cv5/e//32ZX40geMRW2nZtK+86hOJg/33H2cZ4Hhx+FSK95VtTjeNkalN5+NM/mb6pfjJpIVz0nby7vfzyy9x6662cffbZXHfdddx0002AGYX45JNPcvDgQd71rnexcuVKWltb+e53v8sPfvADvvjFL3L99dfz6KOPMmfOnGTbTEGoOg5aRlg842Bi/13HzzG37Sdaj78Cx59anjXVOOIZZ2Dq1KmcffbZAHzgAx/gySefBEga17Vr17Jp0ybOPvtsFi9ezK9+9SteffVVtmzZwsyZM5k7dy5KKT7wgfThVoJQJdie0xu7IBYp61KEItC11TT7aB5j7rdb/fIlb1w2KtczduDBFgulVMb79uAHrTVvfetbueuuu4bt99xzzx1zrCBUHX2HTKvESYvg9eeN0Mf2nIRgcHDbUL4YTIlTXUjyxmVEPOMM7N69mzVr1gBw1113sXz58mHbly1bxv/+7/+ybZsJ5fX19fHKK68wb948du7cyfbt25PHCkLVYeeJT7zY3EqoOnh0bYVxc4bu14fM/f0FGuOe/fAfi6FzQ2HnqUHEGGdg/vz5/OpXv2LRokUcOnSIT3ziE8O2t7e3c9ttt3HVVVexaNEili1bxpYtW2hqauKWW27hHe94B8uXL2f69OllegWCUAC28T3xQnPbJcY4UPQfht4Dwz1j8EdRvWO1iaS8eG9h56lBKjdMXUbq6uq4+eabhz22a9euYffPP/98nn766WOOvfDCC9myRUI9QhVz8BUTspy4ENomDom5hGBgRz7GpRvjebDpfoj2Q6jZ27k7nzK3Ox7zvr4aRTxjQRCG07UNxs4cagYhnnGwSCqpM3jG6MLSEh3rzO3+TXD0de/nqUHEGKcxY8YMXnzxxXIvQxDKx8GtQ17TuDnGUxaCQ9c2UPWmDWYqE+abW6+K6kgvvP4izH27ub9jldcV1iRijAVBGCIeg0M7hupPx8+F/jegt6u86xL8o2ur1Y86NPzxsbONkT6w2dt59zwDOg6nf8SUTW2XULUbKs4Ya63LvYSyI++BUDYOvwqJqAlPw5CHLKHq4JBe1mTTEIZxs717xna+eMrpMOtckzeW7zLHVJQxbmpqoqurq6aNkdaarq4umpqayr0UoRZJF/fYHrKUNwWDRAIObR9e1pRKIYrqjqfNRVzLWJh1HvTsM7ljwREVpaaeMmUKnZ2dHDhwoNxLKStNTU1MmTKl3MsQapF0cc/o6Wa8nnjGwaC7A2IDmT1jMIrqLQ9BbBAaGp2fV2vjGZ9wkbk/+zxzu/0xmHhyYWuuESrKGIdCIWbOnFnuZQhC7dK1FZrHGu8GoK7e5BKlvCkYpA+ISKd9HuiEiZC4MaKHdpiubVPPMPdHTTFe8o7H4Kx/KGzNNUJFhakFQSgzmfKJ40VRHRjsi6pcnjHAfpciLrukyTbGYELVu/4XogPuzuWRv20/yBNbqzeqKsZYEIQhurYe6zWNm2u6KsWj5VmT4B9dW6FxFLS2Z94+bg6oOvciro6nzHnHp/Qwn30exPqHDHWR+c+/buP/PVK9F41ijAVBMAx0G9HN+DRxz/i5ZhD9G6+WZ12VRs9+uPVthfdxLgcHt5q/b7aBNqEmGDPTvYir82mYshTqUkzKjOVQ11Cyblx90TgD0XhJnqsYiDEWBMFwMEubRClvGs6m+4239+wd5V6Je7q2Zc8X27TPc+cZDxyBfS8ND1EDNI6AKWeUrN54IBKnX4yxIAhVT1eWNolS3jScLX80t5sfrK462kgvHNlzbOQjnQnzTPmT0znWezYA+lhjDCZUvXdjSZrGDMQC7hkrpZqUUk8ppTYqpV5SSn0jwz7XKqUOKKWes34+WpzlCoJQNA5utdokplU0NI8xOUYRcZmJR7uehFFTTYOU118o94qck21ARDrt80xa4tB2Z+fteApQMHnpsdtmnQdo2LnKxUK90R+J0x8JsDEGBoHztdanAIuBC5VSyzLsd7fWerH183NfVykIQvGx2yQ2hI/dNm7u0Jd5LbP1L8ZQXfQ9I3SyveRqINuAiHTaLRGW07xx51Mw4SRoGnnstuNPNcKuEoSqB6JxBmKJoj9PschrjLWhx7obsn6qKDYjCIIjsrVJBKu8ScLUvPwQtE6AEy6EaWeZUHW10LUNUDB2Vu79xs01+znJGycSpvPW1NMzb69vgJnnmKERRQ7pD0QTRGIJ4onqNE+OcsZKqXql1HPAfuAvWutMWvV3K6WeV0rdo5Sa6usqBUEoLvnaJI6bC30HzdCIWiU2CFtXwokXGdXw/EtNu8cuh+HccnNwK4yemn9WcbgFxkx35hkffAUGu41QKxuzzzedv4r4PsUTmkjceMXVmjd2ZIy11nGt9WJgCnCGUmpB2i4PAjO01ouAlcCvMp1HKXWDUmq9Ump9rbe8FISKIl+bRPvxWu7EtfMJiByFee8w9+3bavGOM9WQZ6N9vjPP2B4OMfVN2fexW2MWscQp1QAH2hjbaK0PA6uAC9Me79JaD1p3fwYsyXL8LVrrpVrrpe3tWYrOBUEoPfnaJNpTnGq5vOnlhyDUCjPPNfdHTzU50WowxlobzzRfvtim/UTjSedr9NKxzrRPHTc7+z5jZ5ke59sfdb5el6SWNFVreZMTNXW7Umq09Xsz8BZgS9o+x6XcvQzwOBBTEISykK9N4ujpUBeqXUV1IgEv/wnmXGAaY9jMuwT2rIcjr5VvbU44+jpEerKnIdJpn2dGaR7amXu/jqdNSVO2JiI2s88zkYUidXFLVVEH2TM+DnhMKfU88DQmZ/xHpdQ3lVKXWft82ip72gh8Gri2OMsVBKEo5GuTWN8AY2fWrojrtWfh6N6h0LTNfOsrcMtDpV+TG5KRD6fG2IGiuu8QHHzZzC/Ox6zzTIh/zwZnz++SwViqMa5ORXXeqU1a6+eBUzM8/rWU378MfNnfpQmCUDLytUmE2i5vevkhU4M9923DH28/wfRj3vwAnHF9edbmBKdlTTZ2WiJX3tg2rJmafaQz882mFGz7YzAtU2VsYfRHhgxwYMPUgiDUAAcdiHvGzzGj8hLV+WVXEFsegulnDY2WTGX+JWY6Ud+h0q/LKV3bINQCI453tn9jG4yeBgdyZBw7njIXKMeflv98LWNNfr1IIq6BWG2EqQVBCDKDPXD0tfxe07i5EI+YzlO1RNd2E65ND1HbzL8UdNzklCuVg1uNyKrOxVd+vh7VHevMzOPGNmfnm3UedK43A0l8JjVnXK1duMQYC0Kt05VHvGVjhy5rLW/88sPm9sSLM28/brFpj1nJqmo3ZU02SUV17NhtibgJU+cqaUpn9nnmomXnE+7W4YCaUFMLghBwnPYsTtYa15gx3vIQTFxoGmFkQimjqt7+qIkyVBqxQTi823m+2KZ9HsQHM0dC9m8y6mwn+WKbKWeY0rAihKpTQ9ODVSrgEmMsCLXOwa04apPYMtbUlNZSrXHvQROOzRaitpl/qTFc2/5SmnW54dAO0AkPnvE8c5tJUd1hNftwoqS2aQibGcdF6FM9IJ6xIAhVT9dWI9ZJrZ/Nxvi5tdWF65U/G0M2L0uI2mbaMmgZX5mh6qSS2mFZk41d3rQ/g4ir82nTo3vMDHfnnH2eabv6hr+6g2E5YzHGgiCUjUgfHPDYkOPgVuchzHFza8sz3vKQyQdPWpR7v7p6Y7BfecSEhSsJtzXGNo0jYOSUzCKujnXOmn2kM6s4rTFTpzWJmloQhPLx+Pfh5rNNWNUNdptEpyHM8XOgZ19RFLEVR6TPhFRPvNiZ0Zl/mWlssWO1++c6uNXMSi4GB7fBiOOMcXVL+4nHhql7D5rQt5t8cer5Rhzne6ja9oxD9Uo8Y0EQysgrfzZlRy/d5+64I69BtNd5CDOpqK6BUPX2RyHWnz9EbTPzzRAeAVtchqo7noKfnAW//ZD7NTqha6t7r9imfZ5pgZpaW57MF3swxkoZ73jnal/r1QdicRob6mgJN4iASxCEMtG9x6hbAV64x92x+QZEpGPvVwuh6pcfhqZRMP1sZ/s3NMIJbzehbaeGprsT/vtqQBkD9erfPC83I1q7S0Ok036imeZ1ePfQY51PmT7lxy/2ds7Z55tRnHs3ejs+AwOROE2heppCdVJnLAhCmdj+V3O78EroWOtOHOO2TeKYGabrUtDLm+Ix08Rj7tuhPuT8uPmXQF8X7F6Tf99IL9x1FUT74SP/YwRRq77jfc2Z6OuCgcPuldQ2E+ab29RQdcfTcNyi/HORszFrhbn1MW/cH43THKqnOVQvYWpBEMrEtpWmzeH5/2Luv3iv82MPboVwm8njOaEhbAxy0D3jjnXQf8h5iNpmzluhvhE2/zH3fokE/OET8PoLcMUvTKvI5Z/13zt22tAlG8ke1ZYxjkdNsw8vIWqbtnZTt+1j3nggmqA5XE9TqF4EXIIglIF4DLavMqP9xswwHZHchKrtfKIbVWwtlDe9/DDUh2HOW9wd19hm/habHzQh4mw8/j3YdD+89ZtwgjV8YsmH/feO7QhGrnnDuWgebS7UbEX1vhdNHt2LeCuV2Stg91oTHfCB/qjJGTeJZywIVUikF/78leKpWEvBnvUw2D1kNBZeCftfgn0vOTv+4Db3XtO4OaZWNKgDI7SGLX+Emed6UyDPuwSOdJqxi5nYdD+s+jacchWc9amhx8Mt/nvHXVvNRcXoLN3DnJCqqLbFW4Ua41nnmXnJz9wBg0cLOxemnKk5bMLU4hkLQrWx/TFY+1+V3eA/H9tWmhzurBXm/knvNPedeMfRfujucJ9PHH+CEfV0d7hdbXWwfzO8sSt/161snHiR+RtsyRCq3rsR7vu4CfNe8sNjIxJ+e8cHt5nOanX13s/RPs/UsCcSxhiPnAyjphS2rulnwahp8OcvwXemwU/PNRfGmx90X56HMcZNDUbAVa3zjMUYC+XnmTvgsX8r/fPaCuR9L5b+uf1i20rTkrB5tLnf1m4M8wv35A6TgqkvRrvvzJTsUR3QUPXLD5nbEy/ydnzLWNP2Mb0bV89+uOv90DwG3vvrzB3P/PaOCylrsmmfZ8rfujuMMXbTAjMboWb45Fr44B/gnC+YCMT6W+HuD8D3Z8OPz4AHPwvP/9YozvPQb3vGYQlTC4J31t0Mz99d+ue1jfDrL5T+uf2g54AJhabnNRdeCd27h0KK2XBb1mQT9PKmLQ/B5KUwYpL3c8y/1NTn2rnW2KApYeo/BFfdBSMmZj/WL+84HoNDO72Lt2zsHtU7V5vPVaEhaptwq2mPef4/w7V/hH/aDdc9Ahd83QzlePFe+P31cOPJ8Ni3c55qIJqgOSQCLkHwTv9hk9+M9JX+ufeleMb5vMhKxC4NmXPB8MfnvQMamuCF3+U+3vZs3Yp7Wseb+tsgljcdec1c4HgNUdvYx9tCrgc/a+pz3/kTOO6U3Mf65R0fftXkZb2WNdnYPaqfucPcuhmb6IaGRpj2Jjjnc3D17+BLu+BjT5j365U/5zy0PxKnMVQnxlgQPNP5NKAhWmJjHO03IqSW8aYW8+jrpX1+P9i2ElrGmXm6qTSNhBMuNN244tHsx3dtNb2Hw63unlep4PaotmcXF2qMRx5vvOvND8KaH8PG38CKL8PJ73R2vB/esdsa8my0jDVr6XzKlG3l69PtF3X1pp559gXmgjk6kHXXgdQ6Y2n6IQge2L3W3EZ6S+udHthipvEsvMLcr7a8cSIB2/5qvqjqMvwbL7wS+g7m7pN8cKv7fLHN+LnB9Iy3PGRyrHZ9bSHMvxT2PgePfBVOuhze/EXnx/rhHXsdEJEJ2zs+/lRTa15KJi+BRCxnOinVGA/EEugqjHSJMc7G6u9D5/pyryL42MYYbRS6pcIu/Vn0HnNbbXnj1zcaY5utDnbuW6FxVPZQtdamIYRXozN+Lhzd60tZSsUw0A07n3A+GCIf8y81t5MWmPB0poumXBTqHR/caiInLWO9HZ+K3Ylrqg/iLbdMXmJu92T+PtZa0x8daocZT2iicTHGwSDaD499C569o9wrCTaxiPkHa7Da6kX7S/fc+zaZ5z1usSmxqDbPeNtKczv7/MzbGxrhpMtMeU2mfHzPPsJfr3sAACAASURBVBg84j2fmBRxBUBRvX8L/PWb8JPlJsc6/zJ/zjtutlELf/AP7lMBULh33LWt8Hyxje0ZF9J5yysjjzPlVHs2ZNwciSdIaJIduKA6ZxqLMc6ELaVPbY4u+M/ejcYbnrHc3PepG48j9r0IE+aZvNSkBdXnGW/7q7mQaGvPvs/CKyHSk1n84nXgvE21lzcdfR3+9mO4+Ry46U3w5I3QfgK853Z/vb/Z5xnBm1cK8Y4LSUOkc8JF5vM0+zx/zueWyadlNcZ2XbHdgQtgUIxxQLCNsJuG+4J7OqwQte3dlVLEtX8TTDzZ/D5xgfEiSumZF0L/YVO2lK9V44zl0DYpc69qr2VNNmNngaqrLhHX4FF47jdw+zvhB/PhkX82F2MXfgc+twU+cK/J7VYSXr3jgW7o3e+fZzxqMrz75946kvnB5CVmhnLfoWM22eppuwMXiGccHOzOQt0dRigjFIfda2HMTBg709wvlWfcsx96D8AEyxhPWmDEXHYTkEpn52rQ8fzGuK4eFrwbtj5iRtalcnCbCdOPnOxtDQ2NMHpadYi43ngV7rkOvj/XDGc4tMM0mviH9XDDKlj2idx1v+XGi3d8sMABEZXG5KXmds8zx2xKGuOQafphHqu+720xxpk4bBnjeAR6qrDkpRrQ2oyZm3YmhFrMY6XyTG3xVqpnDPB6leSNt6004iwnnZAWXmE+x+ndoOzOTG5FRamMP6E6jPHzd5vowOL3m6YSn9loGk1Ui6Hy4h0XGvmoNI5fDKiMoWrbC7YFXKmPVRNijDOR2nNXQtXFoWubqe+d9qYhcUupwtTpxnjMTDNGsBpEXFqbfPGsc6G+If/+x58KY2cfq6r2I584bq75O1Z69GjwqKmPveQH5vPmh1K61DjxjiO9Rofxwj3m763qzSSvINA4wnQCy2SMI0OecVLAVYW1xg7+m2uQwx3mg9+733SxmX5muVcUPOySpmlnmhpCKF2Yev8maJs4JKypq4MJJ1WHZ3xgCxzZA+d+ydn+ShnhzervwpG9RpkaGzSf64VXFraW8XPMOL0je2D01MLOVUyifd7UzJWE7R3/z1dMM5dwm7mg6tpqLogOboOjr6UcoEyJVqlrgovJ5CXwyp/MBWnKBZUdkk71jAdiYoyDQXeHmSqy6Q+iqC4Wu9dC81gT6nxjl3msZJ7xi8b4pjJpAbxw7zH/6BWHXdKU3gIzFwuvgNXfgZd+D2d+0vQr1onCw7SpPaor2RhHAmCMwXjHT/4Qfnft0GNNo83fcda5VrOSueZ27CwzjCFITFkCz/3aXEimePwDyTD1kJp6IIiesVKqCXgcaLT2v0dr/fW0fRqB24ElQBfwXq31Lt9XWwriMdOf9pT3GSWqhKmLw+41MG2ZMXylDFPHY6Z5/+kfHf74xAWw/hfm4mtMAbNfi822ldA+390Iu/FzTRnUC78zxtivzkyp5U3Z6p0rgWjvkC6hmgm3wPvvNmkW2+i2jKvsi0c/STb/2DDMGPenqKmbGoKtph4EztdanwIsBi5USi1L2+cjwBta6znAjcB3/V1mCTn6mlGqjppq1KKHxRj7Ts9+0xfabjhvf1GWYljEoR2mttnOF9tMWmhuKzlvHOk1Ah43XrHNwivNAISD2/zrWdw2EcIjzHSiSibSZwxZEJh8Gpz2QXMh2zq+dgwxmGhWQxN0Ds8bJz3jhqGccSDV1NrQY90NWT/pvcYuB35l/X4PcIFSVfopsZXUo6caD0mMsf90rDO306xcvB1OK4VnvD9NvGUz4SRAVXbeeNeTRhmdr6QpEwveBSh48R5jjEccV3jNqFLGoFd6rXGkF0IBCFPXOvUhE+FJE3H111KdsVKqXin1HLAf+IvWel3aLpOBDgCtdQzoBsZlOM8NSqn1Sqn1Bw4cKGzlxcJWUo+aBqOnQ/ceE9oU/GP3WqNuPd6aNlRXb654SyHg2veSUZmOP3H4441tpt55XwV34tq20kQRpnkQFI483jQBeeF3/gyctxk/t/K7cEV7g5EzFkyoeu/GYdPIhgm4wpaAK6jGWGsd11ovBqYAZyilFqTtkskLPqZTt9b6Fq31Uq310vb2HG38yontGY+aYsLUOm7UooJ/7F5j/qkaGoceC7WUxjPet8kYolDTsdsmLqhsz3jbSphxTua1O2HhlUZ5u+cZ/2psx82FI52lbWXqliCFqWudyacZBX9Kg55UAVe4vg6lAmyMbbTWh4FVwIVpmzqBqQBKqQZgFHBs37JqoHu3KWsKNQ0JeSRU7R+RPnNlOy1tQHm4tTRNP/a9CBNPyrxt0kJ4Y2dlTiLq2m7y3V5C1DYnXQZ1IXOB6VczCLtWuWu7P+crBtE+CVMHhVQRl0V/JE6dwjLEyr+Zxq89B/d9PGMLzmKQ1xgrpdqVUqOt35uBtwBb0nZ7ALjG+v0K4FFdjQMlwXjGdpnGaMsYi6LaP/ZsMHXF6aHWUEvxvavBo+bCKj1fbJMUcVVgW8xtfzW3XsRbNs1jzGhF8NczBlj7E9P3eftjRq1eSRc0kV7xjIPCmBlGQZ5ijO1ZxrZMycw0LtAYJ+Lw4GfM/50qTW8sJ3XGxwG/UkrVY4z3b7XWf1RKfRNYr7V+ALgVuEMptQ3jEb+vaCsuNt0dQ+0RR00xfwipNfYPu9nH1LRRbKHm4oep9282txOyGGP7777vhWM993KzbaXpFDZudmHnWfJh2LEaJi3yZ112ic3G35ifVMIjTJOREceZnPWI42DKUlMGVcoa2EhASpsEIxqcvGRYj2p7lrFNU6ie/kiBauqnfgZ7n4MrfgHNows7l0PyGmOt9fPAqRke/1rK7wNAge18KgCtzfjEEy8y9+tDppG+hKn9Y/cao1xuHjP88XBr8Uub0ttgpjNqCjSNqry8cXQAdj0Bi68u/FwnvA2+3FlYT+pUGhrhUxuMwTv6uqnRP7r32NudT5g+74mYCRnPfYuZGzz3bdA00p+1ZCIeNTOKw23Few6htExeAlv/YqIvjSMyGOO6wnLG3Xvg0W+ZlNDJ7/Jhwc6QDlyp9B4wNaijpg09Nnq6hKn9IhGHzqdNR6h0Qi2mV3Ux2feS8dZGT8u8XSmYuLDyao13rzFRg0Lyxan4ZYhTCbcarz2X5x6PmouKzQ/C5j/CpvuhPgyzVsD8S037xkJm/2bCTn1ImDo4TF4KaJPTnXkOg9FEcloTGM+4IGP85y+ZC7iL/72kddwyKCKV1Bpjm9HTJEztF/s3weARmJreMwbzZVn0MPUmI97K9Q82aYEx2okKUmNuW2mM1ozl5V5JYdSHTIj6khvh81vguv+BM24w/bYf+BT8+1y47RJY91Po9enCzP5MSZg6OEw+zdxaeWPjGQ+ZsuZQvfc64y0Pm4vFc780NNq1RIgxTqXbMrqjUozxmOkmzBYbLM+agkRyOEQGYxxqKW6YWuvMPanTmbjAfIEf2lm8tbhl21+N4K0xQKHWunrzOXj7v8JnnoePPQ7nfN5Ep/70Rbj/k/48j/2Zkjrj4NAy1ugn9qwHhgRcNs1hj57xYA88/I/mO+KsT/m1WseIMU4lo2c8HdBD2wTv7F4DI47PHCYOtZjmDMXiyGsw0J09X2wzKUXEVQl0d8KBzf6FqCsRpeC4U+D8f4FProOZ50LfQX/OHbGaB4pnHCymLE2KuNJzxo0N9fR7aYe56tumZv6SH5ooTokRY5xKd4cZ2t40augx23CIiKtwdq8bGg6RTrjInnE+8ZZN+3zToatSRFzJkqYAG+N0mkb6V+YWFc84kExeYpoxHdlLf2S4MfbkGe/daMrzllxbtkoKMcappNYY20jjD3843GGuOjOFqMEobGP9xRtUb/ekzhemDjWZcp1KEXHtXmOa0EyYX+6VlI5w25BHWygSpg4mdvOP155hMJYYHqZ2q6ZOxOHBz5rw91v+r6/LdIMY41S6O4bni8HURtaFRFFdKLnyxTCkdo0VqQvXvpdg5BRnNYOV1BZzoNtMR6rSuSueCLf66Blb55EwdbCYtBDqGqBzveUZD5myJrcCrvW/gNeegQu/c2zJZQkRY5xKJs+4rt7Un4qiujB2rzFlRdkabhR7jOK+TflD1DaTFhgvvkRt8HIS6ak9r85PY5z0jMUYB4pQs7lo3rOB/nQBl5t2mEf2wspvwKzzYMG7i7RYZ4gxthnohsHuYz1jkFGKftCxDqaeDvVZStttY1wMEVcsYmbuZutJnc5Euy3mS/6vxS2RGpw4FG4z9f5+TEuzw93S9CN4TF4Crz3LYDRKU0qdcWOonsFYgkTCQUdmu6b4kh+UPfokxtgmk5LaZvQ0CVMXQv9hY9gy1Rfb2J5LMYZFdG01/3AT04eNZSGpqK6AUPVgT7BKmpxgX3z4cWEmdcbBZfISGDzC5PgemhqGe8YAg7E8+pNX/sc0nnnzP8LYWcVcqSPEGNukzjFOZ/R0U2pRyWPiKpnOpwGdPV8MQ1N1ihGmtgc/5BNv2bRNhJbxlZE3jvTWnldnG2M//t8iYowDy5SlACxW24d14GoOOZhpHOmFh74A7fPgrE8XdZlOEWNsk8szHjPD2kfyxp7YvcaUC1n/PBmxBwcUI0y970UjwnM6qUgpqxNXBdQa12TO2Lr48MMYR3uhobk4LUCF8jJuLjrcxil122lqGC7gAnKLuFZ9xzR5uuRGaAgXe6WOkE+oTfduaGiC1vZjt9m1xhKq9sbudaapQy6jEi6igGv/Jmg/0V0h/8QFsH+LP3nLQqjJnLHtGftQ3hTpE/FWUKmrY3DiYk6pS/OMw3mM8esvwJr/gtM+BNPPKsVKHSHG2OZwhzUyMUMS355rLJ6xe2IR07YuV4gahsLUxehPve8l50pqm0kLIT5o8s3lIhaxJg7VqjH2I0xdgxczNUR/+2Lmq1dpqRu6aG608scZw9SRXrj3elPC9JZvlGqZjhBjbHN4d2YlNUDbBOM1i6LaPXs3GmVsPmOcFHD5bIz73zCdepzmi21ssVc588a1qgT20xhHe4cu9ITAcWTcKYRVnPbeV5KP2Z7xMcZYa/jj/zGDSd79M9Pko4IQY2zTnaHG2EYpS1G9q6RLCgS715jbXEpqKF6dsS3ecqqkthl/gpmUVM68cc0aYztnLGFqITeHx5gyxPHdQxfNtpq6P5Kmpt7wS3j+bljxZTM9rMIQYwymnKb3QGYltc3o6RKm9kLHOlM2MGJi7v2KVWe83zbGLj3jhrDJM7v1jDc/CF3b3R2TjeQs3hrz7Hz1jPtESR1gjoba2avHMuaN55OPNWVSU+95Bv70JdPj/c3/WOplOkKMMZjJOJDdMwZp/OEFrY1nnM8rBktNrYrgGb9o8kMjjnN/7MSF7mqNX/4z3P0BWHez++fKRNIY15pn7GfOuAbV6DVEfzTOc4nZtB3cmHysOV1N3XcIfnuN6fH+rp9VrLK+MldVamyPN1vOGEyYeqDbNLAQnNG1Dfq68ueLwaQCQi3+54z3bTItOL1015m0AHr2Qc/+/PsefR3u/3vz+8AR98+ViWSYusaMSchvNXWNvX81xEA0zsbEbBqP7Eq2r7VLmwaicTN45r6PmZn077m94vLEqYgxhqGGH7k8Y1FUu8fOF08709n+YZ+NcSJhwtRuldQ2SRFXnrxxIgG/v8GkO1rbfZw4VKNh6oawyddLmFrIQ380zkY929yx5hsPM8ZP/j/Y+ghc+G2YsqRcy3SEGGMwZU2q3gy+z4aMUnTP7nXQPNZ5s41Qs79h6u7dxjC6zRfbTLJ7VOcJVf/tP2DnajP1ZczMIhjjGgtTg3/DIsQzDjQD0TgvJGaiUbBnAzCkph6z72/w2L/Bgivg9I+Wc5mOEGMMxjMeeXz2IQYgnrFbEnHY/qgpqncaIg61+ivgsgc9uFVS27SMNRdouURcnRvg0W/BSZebJgKNbaaftB8MHjW3tdabGqyZxgV+FrS2SpvEMw4qA9E4PbSgx5+YNMZNDXVM5BDnv/QVGDcXLv2Psg+BcIIYY7AafuQIUYMRAYVHSBcup2x/FI6+Bove4/yYcIu/gyLssqb2ed7PMWlBds948Cjc+xEjDrP/4cNtEqb2g3Br4e9jPAKJmJQ2BZhk+dKUJcYYa00DcW4K/4iG+AC8946quZgVYwy5a4xtlBJFtRuevQNaxsEJFzk/JtTib5h634smbFzIP+PEBWb8Ymzw2G0PfcF8Ht7986Gh5I0j/POMbWNci56dH2HqWg7z1wj90Tjh+jrqJi8xw3wOvwp/+TpL6l7h/mlfNuWJVYIY43gMjryW3zMGo6iWMHV+ertgy8Ow6H3umrCHWvwNUxci3rKZtMB4Vwe2DH98493w/H/DuV8arhYPt0HkaGHPaRPpMe9JXX3+fYOGH8ZYxicGnoFo3NQV20NoVn4D1v4X/113Mc+MPK+8i3OJGOOjr4GO5/eMweSN33jV5KKE7LzwW9NT+dQPuDsu7KNnHO03pVWFGuOJlogrNW98aAc89DmjEj/nC8P3D7caz9iPz0gt91X2I9xvf5Zq9T2sAYwxrjftbhua4KXfw+Sl3NL4YfojOaY2VSBijO3RiU484zHTjefW11XcNVUzWsMzd8Dxp7lXMYda/SttOvAy6IT7ntTpjJttRvDZeeN4FO79qPFW3/WzY0V/jW3m4i42UNjzQo0bYz884xoO89cIA9G4UU/Xh2DyElO9ceVthMJNDEQT+U9QQeSQD9cIyRrjHK0wbex9Dr8KreOLt6Zq5rVnYf9LZk6oW/ysMy5USW1TVw8T5g/VGj/2r0YocuWvMkdTwiPMbaR3aEazVyI9Q+erNXzNGYsxDir90ThN1pQm/u6nJiI3eipNoVdzzzOuQPJ6xkqpqUqpx5RSm5VSLymlPpNhnxVKqW6l1HPWz9eKs9wikPSMp+Tf1y5vEkV1dp79tQkXLXi3+2P9FHDt32Q82rEzCz+XrajesQqe/CGcdg2c/M7M+9pisUEf8sa13MrRj9KmZJhaBFxBpT+aoMmeZTx6qumDj2n8UW3G2IlnHAM+r7V+Rik1AtiglPqL1npT2n5PaK0v8X+JRaZ7t+ma5MSLSfWMhWOJ9sML95ia26ZR7o8PtZgZwol44aKlfS8aJaUf4qeJC+GZ2+F315oGJhd+O/u+vk4c6oWm0YWfpxqxPeNEwnsvYQlTB56BaJzm0LGfj6ZQPYf7ImVYkXfyfsq11nu11s9Yvx8FNgOTi72wkuGkxtimaaQpYRFFdWY2/xEGu90Lt2zscKIfnZf2bSo8RG0zyTpPpBeu+EVubzXpGftkjGvWM24FNMQKqDtPesZijINKUsCVRnMVesauLjmVUjOAU4F1GTafqZTaqJT6k1KqQAlrCXFSY5yKragWjuXZO8z7M325t+OTYxQLbPzRcwB693tvg5nOpEUwcrJpd2m3yMxGMmfslzGu0RCrH5ObkqVNNXpBUwP0R+LJKU2pNIerzxg7FnAppdqAe4HPaq3Tx9I8A0zXWvcopS4G/gAc05BYKXUDcAPAtGkOBFPFRmszPvGEC50fM3ra0IxcYYg3dpn+zOf9i/ewov0FXGit8X5bvOXTNWFjG/yfl5y11JOcsT8MC/dP8HaOWp16VUMMxDIb46ZQXdWpqR19ayqlQhhDfKfW+vfp27XWR7TWPdbvDwMhpdQxcmOt9S1a66Va66Xt7e0FLt0Heg+YEhQnSmqbMdNNaDtRXX/oovPcbwAFi6/yfg47b1+oiMuOXFhiDl9w2ts26dH54BkP9lRNKz/f8cMzjvQBqnBVu1Cx9EcSNGY0xvUMBK3OWCmlgFuBzVrrH2TZZ5K1H0qpM6zzVn4xrpsaY5vR043IqGdfcdZUjSTi8OydMPt8Z6r0bNjhxELLm2yv1IuIrFDCPuWMYxFTplGrXp1fYepQS1UMCRC8MRjNEqYO1TMQC5gxBs4GPgicn1K6dLFS6uNKqY9b+1wBvKiU2gj8CHif1lXQpqrbEmK5zRlD8BTViYRp8XjUw0XGztVwpNO7cMvGLwGXbYzLkW9tTKkzLoRkiLVWPWMfVOmRXhFvBZx+ux1mGk2heqJxTTRePRHMvDljrfWTQM5LS631j4Ef+7WokuHFMx6TMkoxtSdxtfPyw3DfDTDhZLjuT+68ymd/bVTm895R2Br8EnANHjVf5uXo6VwfgvrGwvtT1/LEJvDXMxYCSTSeIJbQWT1jMGrrUH11NJqsjlUWi+4OaBwJzS5qOe38cpAU1VrD49+D1glw8GX47YdMmNQJfYdMSdOi90JDY2HrCPsVpj4y5KGWAz9mGosxNrcF5YxrWI1eAwxYaunmcGYBl9mnejzj2jbGbmqMbULNxmgd3lWUJZWFrY/A3o3wlq/DZf9pOk09+Blnww5euMfk0AsNUUOKgMuHMHU5jbEvQw5sY1yr7TDtMHWhxlg846Bily5lE3DBkMGuBmq7N7XbGmObMdOD0/hDa1j9XePxL3qvCbMe7oBV/2YeO+/LuY9/9g447pT89bdOSIapfRBwldUz9mGmsR3mrnnPuID3UcLUgWYgYrzebHXGQFXVGotn7NYzhmA1/tj+qBl8sPxzxhADnPtFWPwBWP0dkw/Oxt6N8PrzcOoH/VmLb2HqSvCMJWdcEA1NoOoKL22q1fevBrDV0hnrjBuqzzOuXWM80G1aN3rxjEdPgyN7IB7zf12lxPaKR06Bxe8felwpuPSHMOs8E67e9tfMxz97pxErLbzCn/XUh0HVF15nXHZj3OpjzrhGc55KFT4sItIjnnGAsecVZ1JTJz3jKqo1rl1j7EVJbTNmOiRicPQ1f9dUanY+Dh3rYPlnjxVf1YfgPbdD+zz47TVDIwRtogPw/N0w/1KjpPYDpcyXpy+e8Uh/1uSFRj9yxtI9ygyLKDBMXcvvX8BJCrhy5YxjIuCqfNzMMU4nKKMUH/8+tE3KHmZuGglX/87c3vke6N4ztO3lh2DgsD/CrVTCLT7U6JbbMx7h4yzeGjYmhc40ljB1oLHzwU051NTiGVcDhXjGyVGKVSzievVvsOsJOPszEGrKvt/I4+H9vzXe5p1XmvA+mFzyqKkw81x/1xVqKazOWOvyh6mltMkfCjHGWpse5xKmDiy2Z2znh1NprkI1de0a4+7dJt/Z6qFH9qipgKruLlyrv2de+5Jr8+87aQG89/ahGuRDO2D7Y7D4au9DIbIRbi0sTB3tA52oDAFXIU3oBo8aQ1KOxiWVQiE549ig+RxIaVNgsWuIM9cZizGuHg53mD7KXoxJQ9iM1KvWMHXHU7DjMTjr086/rGafD5f+yNQg//Ji89ipV/u/tlBzYaFJuxVmuT1jnSjMw6/lWcY2heSM7c+QjE8MLMkwdSYBV0hKm6oHrzXGNqOnVW+YevX3oHksLL3O3XGnXg0rvgxH98Ksc73l2/NRqIAraYzLKODyra9yjRuSQsLUUQnzBx07HxyUOuPabfpxuANOeLv348dMN2rkamPPM7DtL3DB17yN5zv3S9A2Aaaf7f/awHx59h3yfvygNWq73E0/wFwYtHmdxSutHAsyxnZ5nISpA4tdZ9yUwRg3NlRfO8zaNMbRAejdX5hnN3o6HHnN5KYK7clcSh7/PjSNhtOv93a8Uu49ajf45hmXuc4YCvSMe8SrK6StaFTC1EFnIBJHqSHDm4pSiqZQneSMK57uTnPrRUltM3oaoIfOVQ3sfd5MZ1r296ZcqRIJB8EY+zDTWDzjIc/YixBOPOPAMxBL0NRQj8oyr7o5VC/GuOLxMsc4nTFVONf48e+bXOqbPlbulWQn1FJYB65KMMZ+zDQWz9i8/kQM4g4niKUipWGBpz8Sz6iktmkK1UudccVTSI2xTbU1/ti3CTY/AG/6uLuRkaUm1DIUYvRCRQm4CuhPLZ5xYZObJEwdePqjcZoyhKhtmkP1VSXgqk1j3N1hmtCPPN77OUYeD3UN1aOofvz75stt2SfKvZLchFssbyjq7XhbwFVOQ9boR5haPOOCcu8Spg48A9F4xu5bNo2h+qoScNWmMT7cASOOH5pS5IW6elOnXA1h6gOvwEv3wRnXQ8vYcq8mN7Yn4zXEO3jUTPxpCPu3JrdIaZM/hAv4LNi6A/GMA8tANJ6x+5ZNswi4qoBCa4xtRk+rjjD1E/9ummmc+Q/lXkl+Qs3m1quIa/Bo+cO7hQq4YhGTJy336yg3hYSpkzlj8YyDSn80d864OVxdYeraLG063AHTlhV+ntHT4ZX/Kfw8Xoj2G2X0QLf50o/0Gk8s0nPs/b0bjYK6dXx51uqGpDdUgDEup3gLoL4BGpq954xtj9pLHXiQKChM3WtSUQ05+q4LVc1ANJGx4YdNU0M9h/s8prvKQO0Z43jMzCL2wzMeM93UK0f6Sn8F/uQPYfV3hj8WajVfYI1t5jY8AlrGm6lM53y+tOvzit3Y36uIqxKMMRQ201iUwIZCw9ShVlMXLwSS/kicMS3ZU41N4hlXOEf3go4XpqS2sRXVh3fDhHmFn88p0QF4+ucw5y1w2Y+N8Q3KUAH7osZrX+dyzzK2KWSmsRhjQ6FhaglRB5qBWDxj9y2b5lA9gyLgqmCSc4x9Nsal5IXfQt9BM+hh5HHGEwyCIQYfBFxHKsQzLmCmcdIYS5ga8HZRE+2T8YkBZyASzx2mDtVVlWdce8Y4WWPsw5CDcjT+0BrW3AQTF8DMN5fueUuFHwKuSjDGjW1DNc9usY1PzXvGBVyYRfrk/Qs4/dH8nrE0/ahkbC921JTCz9U6wcxELqUx3vEYHNhsBFlBzIcFQcAFhfVVljC1wfZsPRljqdMOOgPRRN4OXAOxOLqQueIlpPaMcfduI2ryI59UVwdjZsDrLxZ+LqesuclcBCy8onTPWUqSAq4qN8aNbQUIuGzPuAJeRzmpq7fao0qYWhiO1jpvB66mUD1aw2CsOvLGtWeMD/tUY2xz0mWwY1Vp6o0PvGzGH57+0eqaFOWGcAHGODZotzB80wAAIABJREFU6nMrwRgX5BlLmDqJ1zGKEqYONLaBzdWBy84nV0vjj9ozxt0d/iipbZZca8LFG37p3zmzsfYnJixezBGG5SYZmvRgjCuhL7VN4wgpbfIDr8Y42iuecYCxDWxuAZdtjMUzrjy0NfKwkDnG6YyaAidcBM/cYTyzYtF3CDb+Nyx6D7S1F+95yk19COpC3uqM7b7UFeEZtxoP19P4PzHGScJtBXjGYoyDiq2SzingCtcN27fSyWuMlVJTlVKPKaU2K6VeUkp9JsM+Sin1I6XUNqXU80qp04qz3ALp2QexAX89Y4DTrzOlRpse8Pe8qaz/BcT6jXAr6IQ9jlGshPGJNuE2QHsXHzU0B6dcrRDsixq3yNSrQGN7u7k84yCGqWPA57XW84FlwCeVUiel7XMRMNf6uQH4ia+r9Ivda8zt5CX+nnfW+TBmJqy/1d/z2sQi8NTPYNZ5MDH9rQ8goVZvOeNKMsaNhTasEK8Y8Bam1loEXAHHLlnK5Rk3WtsC4xlrrfdqrZ+xfj8KbAYmp+12OXC7NqwFRiuljvN9tYWyY5XJJx5/qr/nrauD0z9ijP2+l/w9N5iJSz2vV8egBz8It1S/MbaV0F68usEe6Utt48UYR/sBLWHqADMUps49zxhMc5BqwFXOWCk1AzgVWJe2aTLQkXK/k2MNNkqpG5RS65VS6w8cOOBupX6wYxXMOMc08vebxVcbcdXTPnvHWsPa/4LxJ8KcC/w9d6USag6AgMue3OSh8YeEWIfwokqX8YmBZ9CNgCsWMGOslGoD7gU+q7U+kr45wyHHKFe01rdorZdqrZe2t5dYhHRoJ7yxC2atKM75W8bCgnfD83d777yUiVf/Zk1d+kQwm3xkItQaAAFXATONpWHFEF4842RpmHjGQcWRgMsOU0cCpKZWSoUwhvhOrfXvM+zSCaSqoqYArxW+PB/Zscrczj6veM9x+kfMF8Hzd/t3zrU3QfNYOOV9/p2z0gm3eBsUUUlh6sYCZhpLzngIT8a4b+hYIZDYxjjnPOOg5YyVUgq4Fdistf5Blt0eAD5kqaqXAd1a670+rrNwdqyCkZNh3JziPcfkJXDcKSZU7UcLtkM7YMtDpq7Y7tlcC4QKUFOr+sp4rwrJGYsxHiLcZiog4jHnx0iYOvA4UVPb+eQgqanPBj4InK+Ues76uVgp9XGl1MetfR4GdgDbgJ8BlVV/k0jAztUmRF3MUK9SsPQjsH/TkHK7ENb9FOoa4IzrCz9XNRFq8RimtlphVkI43zamkjMuDPt9dPN5SNZpS5g6qNjebmMOAZfdnatajHFeJZPW+kky54RT99HAJ/1alO+8/jz0v1G8fHEqC6+AR75qvOPpZ3k/z0A3PPtrk4ceMcm/9VUDhdQZV4J4C1JKm7x4xkfFGNukTm5qGuXsmKRnLMY4qDgRcAWxzrj6sfPFM88t/nOFW2HxVbDpfujZ7/08z9xuvsjPrKwgQ0kItXrPGVdCvhhSBFxSZ1wQXt5HmQcdeJzUGYfq66ivU8HJGQeCHatgwkkwYmJpnm/pRyARhWfv8HZ8PGZC1NOXmxx0rRG2wtRu8+6DRyrHGNsTh9yGqWMRM+xCDIkh6Rm7iDBImDrw9EfjNNQpQvW5TZiZaRwgNXVVEx0w+dtZK0r3nO0nmHrm9b+EhIerss0PmIEWtegVgxFg6YT7Xt+V5BmDxxpZ6Us9jNQwtVMkTB14BqKJnCFqm6ZQXfDqjKuWjnVGjTlrRWmf9/SPGoO69S/uj117E4ydBSdc6P+6qgFbBeu2C1elGWMvM41lSMRwvBhjeQ8DT380nnN8ok1TqD6YHbiqkh2PGUVyIWIqL8x7B7RNgqd/7u643eug82l40ydqd1CAHV50m2+tNGPsxTO2X7O0wzR4aZ4S7TMlbvXh4qxJKDsD0XjOVpg2zaF6yRlXDDtWwZTTS/8lXR+CJdfAtpWm+1c+tIZn74Q7r4SWcbD4/cVfY6VihxfdirgqzRh7mWls7y85Y4NXzzjcVhklbkJRGIjGHYap60VNXRH0HYLXnjPTjsrBadeAqoMNv8y935G98Jv3wv1/DxNPho+urG3PyEttaTxmPKJKKW0Cq3uUSwFXspWjhFiBAoyx5IuDTH80nlNJbSOecaWw6wlAlz5fbDNqMpx4ETxzhxGSpaM1bLwbbnoT7HwcLvwOXPuQyRfXMnYHLTe1xpEKaoVpE5acccGEPKipZXxi4BlwaIybwvXJbl2VTrCN8Y5Vpi3h5NPKt4bTPwr9h0zdcSo9++HuD8B9N0D7PPj4k2YYRF2w/ySO8CLgqqS+1DaNbR76KkuN7DAawib368oz7hPPOOD0O1VTN9RJmLoi2LEKZiw3+dtyMfNcGDt7SMilNbx4L/zXm4zS+m3fgg//CcYXsWd2tWF/kVa7MQ6P8CDgkjD1MbgdFhHtlb7UAWcg4lDAFZYwdfl541UzaGHWivKuo67OTHPqfAq2Pwq/uwbuuQ7GzjTe8Fmfql3VdDbsEKObMHUlGuNGS02dcBEmE8/4WMIuIwzSwSzwDMQcCrgaRMBVfnauNrezVpRzFYZTroKGJrjj7+DlP8EFX4frHjHNQYRjSaqpXXwBJ41xJQm4LIPqasiBeMbHEG512YFLwtRBpz8Szzk+0aY5XJ9snVnp5B0UUbXsWGXqfNtPLPdKoGUsnPkPphPYxf8OE08q94oqm7AXz/iIua00zxiMiMvpuiI90NAs0ZJUJEwtpNEfjdPY4LS0qToEXME0xomEMcZz3lo5tYYXfLXcK6geQgHKGYP7vsriFQ/HrTEWzzjwDEYTjjzjplAdkXiCeEJTX1chtiALwQxT73sR+roqI0QtuKeu3oT1q94Ye5hpLMb4WNx2MpPSpkATT2gi8QRNDjzjahqjGExjbI9MnLWijIsQCiLkcqaxbfAqSfjkZaax3T1KGMKNZ5xIGGMs72FgsQ1rc9iZmjr1mEomuMa4fR6MPK7cKxG8Empx7xmHR1RWnbanWbw9td19LRNujLH9mZEwdWCxS5WcqqlTj6lkKuibyydig/Dq38QrrnbCLe6MWCXNMrax1+OmC9dgj4Sp03FT2iTjEwOPrY5udNiBC8QzLg8dT0GsX4xxtRNqcTcootKGRECKZyw544KwPWMn9drSTjTwDMbceMbGxFWDojp4xnjHKjM+bfrZ5V6JUAjhVpdhahflQ6UitbTJKZIzPpZwK6DNRXY+bGMsnnFg6Y8Yw+rEGNs5YwlTl4Mdq2DKUmiqoOYPgntCzS7D1BXoGXsZchCRMPUxuJnclMwZywVNULENq9OpTUBVNP4IljHuPwyvPSMh6iDgRcBVaca4rs795CYJUx9L2IUqPRmmFs84qLhRUzdJaVOZ2PUE6IQY4yDgOkx9tLJaYdq4mWkcj0J8cKhZiGDw4hlLmDqw2J6x0w5cqcdUMsEyxjtWmdDg5KXlXolQKF7qjCvNMwZ3nrH0pc6MG2Nsf2bkPQwsQ56x85zxoAi4SsyOVTDjbDMDVahuwi7C1FpXZmkTuJtpLErgzLgKU8sFTdAZcFVnbEyceMal5HAHdG2DWeeVeyWCH9g5Y8flLLoyjbGbmcZijDMjYWohBVuM5UjAJWrqMlBJIxOFwrG/TJ2Us1RiX2qbxjbnvamTXp0ogYchYWohhYGY89ImuwOXCLhKyY5V0DoBJswv90oEP7C/TJ00/qhkY+xmyIF4xplx01Y02gt1IagPFXdNQtlIduBqyG++6uoU4Ya6YHjGSqlfKKX2K6VezLJ9hVKqWyn1nPXzNf+XmQetjTGetaJyRiYKhWF7xk6+gJPGuALV1I1uBFy9Q8cIQ4Rd1GvL+MTAMxCN09hQR53DkYjNoXoGqqDO2Mk849uAHwO359jnCa31Jb6syAv7N0HvAQlRB4lQs7l1IuIaPGJuq90zHpQwdUYamkDVOQxTSwezoDMQjTtSUts0heqC0Q5Ta/04cKgEa/FOIg4nvgNmnVvulQh+kfSGnBjjCg9TR/vMZzQfogTOjFLOh0VEe0W8FXD6o3FHs4xtmkP1wQhTO+RMpdRGpdSflFIn+3RO5xy3CK76DYyaUvKnFoqE/YXqyDO2jXEFekRuZhpLzjg74VYJUwuAGfrgzjOurxkB1zPAdK31KcB/An/ItqNS6gal1Hql1PoDBw748NRCYAl7McYVmDN2Iz5KDjkQY3wMTmcaR/vk/Qs4/dG4o7Imm6Za8Yy11ke01j3W7w8DIaXU+Cz73qK1Xqq1Xtre3l7oUwtBxouAqxJzhW5mGkd6TH603omUo8ZwaowjveIZB5yBaJymkHPT1VwrnrFSapJSRsKslDrDOmdXoecVahxXYeojxohVYuc1NzONRXyUHac5Yxm0EXgGonFHNcY21SLgynsJrpS6C1gBjFdKdQJfB0IAWuubgSuATyilYkA/8D6ttS7aioXawK2AqxLFW+BuprGMT8xOuNVUTORDwtSBpz8aZ2ST8zry5nB1hKnzGmOt9VV5tv8YU/okCP7hVsBVqcbY7fg/8YwzE26FN3bl30/C1IGnP+IhZ1wFdcbB6cAlBAtXdcYVbIzd5ozFM86MKwGXGOMgMxBNuDbGgzExxoLgDaWsMYoOBVyVqKQGDzljMcYZcdI8JRGH2IC8hwHHNP1wJ+ASz1gQCiHkcIxiJXvGtmFw5BmLMc6K7RnnkqNInXZNMOCh6cdALEGlS5nEGAuVS7jF4aCICp1lDJZhUA6VwD2V+zrKTbgVEjGIR7LvI+MTA4/Wmn4P7TDjCU00LsZYELwRcpgnrGTPONnK0YFnPCg546w4aZ4innHgicQTJLSzWcY29r6VrqgWYyxULqHm/GFqrSvbGIPzmcYSps6Ok8lN4hkHHrte2IsxHhRjLAgeCbfmrzOODUIiWtnG2IlnHI9CfFBKm7KRNMa5POO+4fsKgcPupOWm6UezeMaCUCChFjOFJxeV3JfaxslMYwmx5ibkxBjL1KugY6uiXbXDDIsxFoTCcCLgquRZxjZOPGMxxrmRMLUADMTce8a24a70lphijIXKJeQgTF3Js4xtGke48IwlTJ0RCVMLpHjGLkcoph5bqYgxFiqXULOLMHUFG+Nwa/6mH5EKnjxVCThRU9ufFfGMA0tSwOWyzhiGvOpKRYyxULmEW4LhGTuZOCRh6tw4CVPLexh4kgIuD57xgHjGguCRUKtRGCdy/BOJgKs2kDC1wJAIy+0849RjKxUxxkLlEnYwuakqBFwjINYP8Vj2fSRnnBsnxjjaC/WNUOfcaxKqCy+lTUnPWARcguARe3JTrlB1NYSpGx2MUbS3NYoxzkhdPTQ053kP+2R8YsDplzpjQSgDdm1pLhHX4FGoa4CGptKsyQtOZhpLmDo/+cYoRvuGPjNCILEV0Y1uPOOwXdokxlgQvJEMU+eoNbZbYSpVmjV5wfZ2c+WN7W1iTLKTzxjLPOjAMxgzoWY3nnG4vg6lxBgLgneSXZfyhKkrOUQNJmcM+cPUDU1Q31CaNVUj+VTpEqYOPP2ROPV1ilC984tvpVRVzDQWYyxULknPOE+YupKV1JAy0zhHrbEMichPuDV/By6JLAQaM8u4DuUyEmZmGosxFgRvOBJwVfAsY5tGh+P/xBjnJm+Yulc844DjdpaxTVOonv6IqKkFwRtJAVe1h6kdqqnDFf46yo0jAZcY4yDTH43T6KL7lk1TqE5yxoLgGUd1xlVgjO31SZi6MPIN3Ij0Sp12wBmMJjx7xmKMBcErtpdT9QIup56xGOOcSJi65umPxl0pqW2aQ/VSZywIngk5FXBVuDEONYOqy13aJJ5xfiRMXfP0R+KuWmHaNIfFGAuCdxoajRHL5hnHo6bNZKWrqZUy+eC8nrGEWHMSboPYQOa2ovEYxCNyQRNwBmLxZHtLNzQ21Es7TEHwjFJGxJWt6Uc1tMK0yTcsQjzj/IRzdGST8Yk1QX/EY5g6LDljQSiMcEv2MHU1GeN8M40jvdKXOh+5hkVIO9GaYDCW8OQZN4uaWhAKJNScPUxdVcY4R/eoeMyEXyVMnZtwjnptGZ9YE3j1jJtEwCUIBRJqzV7aVE3GOFeY2s4liyHJTdIzzvA+Spi6JuiPehRwBaEdplLqF0qp/UqpF7NsV0qpHymltimlnldKneb/MoWaJdyS3aNMGuMKF3BBbgGXhFidkTNMbXvGYoyDzEA0TpOHOuPGUD2DsQSJhC7CqvzBySXGbcCFObZfBMy1fm4AflL4sgTBItSSQ8B1xNxWjWecJWecNMYSps5JrjB1VN7DoJNIaAZjCc91xjA09akSyWuMtdaPA4dy7HI5cLs2rAVGK6WO82uBQo0Tbg2IgCtH9ygJUzsjV5g6ImHqoGMPevAq4AIqOm/sR854MtCRcr/TeuwYlFI3KKXWK6XWHzhwwIenFgJPLgGX/aVcDcY4Z85YvDpHSJi6prHrhL0KuMw5gm2MM82yyhiY11rforVeqrVe2t7e7sNTC4En1JJHwKWqY2xeeATEB02jknQkZ+yMXMY4KeCS9zCo2F6t1zrj1HNUIn4Y405gasr9KcBrPpxXEKwwdQ5j3DgC6qqgKCDXTONkmFo845zk6vEtnnHgsb3aRg9q6lrxjB8APmSpqpcB3VrrvT6cVxCMZ5y1zrgKZhnb5JppLDljZzSEoS6UxTO2PiOSMw4sdmlSUMPUDfl2UErdBawAxiulOoGvAyEArfXNwMPAxcA2oA/4cLEWK9QgoRZIRE14tz40fFs1DImwyenVSZjaMdmGRUR6oKEZ6tx/UQvVgW1IvQm4rDB1pHLV1HmNsdb6qjzbNfz/7Z1rjCTXVcd/Z6qrp2dnE2LHBhY/WEf4QzaRA2Gx5SgKRijYscAW2JEcgYWltSweTiB8YS0kx4rEhwTx+EAiMMYyCVFikqBkiTaxwkMiioPxEhzjTWSyNg5e1rE3fq1ntnv6MZcPVdVT01M1011Vs923+v+TWl19+3bVnVM19a9z7rn38tuVtUiINEnYsbsKS2/Y/J1PYjxc03g7MVaYekfyZjLrnlWIuuYME7gKrWe8EO9jdj1jDzrbxFwzXEYxI1TtkxgPPeOcPuNGC4Idn41FczlnBq6zSt6qOaUSuML5SOASYvcYrtSTMfGHT2Kc9BnnecYKUY9Hbph6VZ5xzdkIUyuBS4hzT5gKU4/ikxhv12e8tiIxHpdtxVg2rDPtEn3GEmMhyhIuRe+5YWoP5qWGHfqMV9RfPC55M5n1ziqTuuZ0NM5YiCmSN9HD+rpnnnHyd2T1GcurGxt5xnNLmWzqViNJ4JrdbGqJsZhthglcI33GvVXA+SPGjRZYICEpS54YyzOuPcmwpCJi3AgWCAOTZyxEYYYJXCNhap8WiQAwy5+furuqMPW4aGjT3NLuDWg2FggWsmZg3pnWjK9pLDEWs03SZzx6A07E2CcRy1vTWH3G45N4xusj4UY90NSeTm8wDDcXoRUGrPUlxkIUI2+c8dAz9iSBC/LXNO4qm3psmsuAg35Gt4XC1LWm0xsUmvAjYUmesRAlyA1Tn4nefQlTQ34msPqMxycroa/fhfW+wtQ1p9MbFOovTlgKAyVwCVGYIIwXB/C8zxiy+4wHfeh3FGIdl6zx2lo+cS5o9waFhjUltMIFJXAJUYpmxprGPopxlmfc0yIRE5HlGWv5xLmg3Vsv5Rm3wkBiLEQpwj35CVy+ifGoZ5z8XYvyjMciU4y10MY8EIWpi0vWUjNgTWIsRAnCmnjGixmesYRkMrYNU8szrjOdsmHqhjxjIcrR3LN10o+1M9H6taNrHM8yWWHq4RAthanHQmHquaXdLZlN3ZQYC1GOMGPWJZ+mwkxY3AuDbpT9m9BVn/FEZIlxEjVRAlet6fQHtBrlEriUTS1EGcKl7DC1b2LcjNub9o4lxpMxDFNn9RnLhnWm3V2nVcIzboUBHY0zFqIEzeXsoU2+ifFwTePUxB+JMKvPeDyGnnHWA43C1HVmrVfOM14KAzqagUuIEuQlcPkmxlnJR0rgmoxwCTCFqeeQdm/AUrPcdJi9gaM3mM1QtcRYzD5544x9mgoTUp6xwtSFMdu6WIQ849rTG6zTX3elsqmT33ZmNIlLYixmn3BPRpj6jMeecVaYWmI8Ns3lkaFN8bXRWJpOe8SuU2Yt44Skv3lWk7gkxmL2CfdEY0md2yjzOkyd9upWIFj0a4jWtBld07i7GoWoF3Q7qyvtKsQ4XvFJnrEQRWnuAbcO/bXos3N+inFemFpe8WRkibFC1LWm04282VJh6tgzntWxxhJjMfuEIys39TvRKj2+iXHe0CZNhTkZo33GvbOafavmJFnQ5Txj9RkLUY7myJrGPk6FCflDm5RJPRmjfcaKLtSeZB3iMtnUQ894RscaS4zF7JN4Pd1RMfYsm7qxGC8HKSEpxWiYWp5x7RkmcJWagSv2jPtK4BKiGMmNNlkQYO1M9O6bZwxb1zReW5EYT0rW0CbZsNYME7hKzcAVyZ08YyGKMgxTx4tF+BqmhqjfeItnrDD1RGwJU5+VGNecxDOe+3HGZnadmT1pZifM7HDG97eZ2Wkzeyx+3V59U8XckiRwbQlT+yjGyxl9xhKSiUjC1MlQt96qwtQ1JxkbXCqBa8bFuLFTBTMLgI8B7wZOAo+a2RHn3LdHqj7onLtzF9oo5p0wnsxhGKb2WIwXs0Ks8ownorkcZdMPulE/fPeshjbVnHaFnrHPQ5uuBE445552znWBzwA37m6zhEjRrEkCF2xd01j9nZMzOnlK76zmpa45w2zqOR9nfBHwbOrzybhslJvM7HEz+5yZXZK1IzO7w8yOmdmx06dPF2iumEtGxxnXJYFrfQD9tjzjSUmv3OScQv1zQDLOeDEsnua0OJyBy99sassocyOf/wHY75y7AvhH4G+yduScu9c5d9A5d/DCCy+crKVifskaZ7wQRiFK30gncGle6mIMxXg1mpXNrStMXXM63QFmG4JaBDOjFS7MbJ/xOH/ZSSDt6V4MnEpXcM696JyL5yrkr4CfrqZ5QpA9znjxddEKPr6xuHcjzK4Vm4qRDlNr+cS5oNNfp9UIsJL/80th4LUYPwpcbmaXmVkTuAU4kq5gZvtSH28AvlNdE8XcsxBEiymkE7h8DFHDRp+xc1rLuCjpMLWWT5wL2t3BsM+3DK0wmNlxxjtmUzvn+mZ2J/AQEAD3O+eOm9mHgWPOuSPAB8zsBqAPvATctottFvNIc8+IZ+xh8hZEnvF6PwqvJmFqzU09GekwdeIZK7pQa9q9QankrYSlMJjZBK4dxRjAOXcUODpSdndq+y7grmqbJkSKcHnzpB8+e8Yw4tVJSCYiHaZOHmgUpq41nd6gVPJWwmIYeJ3AJcT0ae7ZPB2mxHh+2RSmTjxjhanrTKcyz9jvBC4hpk+4tDWBy0fSaxoniVzqM56MrDC1PONaU1mYujm7YWqJsfCDcHnz0CZfxViecXnSYqwErrmg01svNRVmQqvhdza1ENOnuaceYpy0e01iXJiFABpL0QONErjmgnZ3UI0YyzMWoiRhnE3d70K/42829dAzfk1Dm8qQLBaR2FBh6lrT6VcztGkpDFhTApcQJWguRwlcw+FAvnrGqT7j7ko0fjoIp9smHxkVY4Wpa02nO6BVYvathFa4IM9YiFIkCVw+z0sNW/uMFV4tRnNvKoHLoNGadovELtLuVecZz+qkHxJj4Qdh3Gfs8/KJsCHGiWesEHUxmssbQ5uay35OjSrGprIErjCg0x/g3OjyCtNHYiz8oLlcDzFuNCFoxkKi1YYKk4Spe4ou1B3nHO1eRQlcYYBzsNafvX5jibHwg2SxiJUXondfE7hgY35qhamLk+4zDtVfXGcS4axqOkxgJoc3SYyFHwzF+Pno3VfPGDbWNO6ual7qogwfaM7qgabmJMLZqmA6zNZQjOUZC1GMJFv2te9H7z6LcbKmcXdVfcZFSYep5RnXmiT7uZoZuBY27XOWkBgLP9gSpvZYjJM1jdVnXJxhmPqshjXVnMSLrSqbOtqnxFiIYiSitfJ9wPwWsSTEuiYxLkxzbzT5y9oZRRdqTjIUabFRXowXYzGWZyxEUcKl6P2156PkLZ+HsqT7jCXGxRg+nL2gMHXNGYapq/SMZ3CsscRY+EGY8ox9DlFD5Ml1XoV+W15dURIxbr+kMHXNWUsSuCqZgSsW477EWIhiJDfc1R/UQ4xXT29si8lJ203zUtea3fCM211lUwtRjGEo0vkvxot7wcVP5gpTFyNtN3nGtabSbGr1GQtRkvTN13cxTnt18oyLsUmM9UBTZ5Js6mpm4FqI9ykxFqIYSQIX+C/G6fZLSIqhMPXc0B5O+lHNesYgMRaiOI20GHvuTW7yjCUkhVCYem5Y241saomxEAVZWNjoN/Z5XmrY/DChMHUx0mKsoU21JhlnXEU2dRgsECyY+oyFKMVQjD0PU6cF2Hcvf1ps8oxlwzrT7g0IA6MRVCNX0ZrGyqYWojh1FGOFqYuxyYbyjOtMp7dOq4LZtxJa4YLGGQtRimZNxHhRYlyaRhMWwmhbYepa0+4NholXVdAKA83AJUQp6ugZKxO4OMmDjB5oak2nN6hkjHHCUhioz1iIUiQ33bokcAXNyMMTxUgeauQZ15pOb1DJWsYJrTBQNrUQpaiNZxy3X4lH5ZBnPBe05RkLMWMkE3/4LsZBAxotiXFZJMZzQeQZV9hn3AyGs3rNEmOJsZldZ2ZPmtkJMzuc8f2imT0Yf/+Ime2vuqFCbISpPRdjiIRYIlKO5jJYEIX7RW1p99arFePGgp9hajMLgI8B7wEOAO8zswMj1Q4BLzvnfgL4U+AjVTdUiNqEqSHqN5YYlyN5oPF5bWuxI51uxWHq5myGqRtj1LkSOOGcexrAzD4D3Ah8O1XnRuCeePtzwJ+bmTnnXIVtzeXxk6/woSPHz8WhxBS5deVlfgW46f7/Yt2q++crnJCGAAAG30lEQVScBn/ymnFmpc89H//6tJviLR989Sxv7TU4JBvWmv95cZU376vuAbzVCDj1SptfHvO6+dtDV7G8OI5UlmOcI1wEPJv6fBK4Kq+Oc65vZq8CbwR+kK5kZncAdwBceumlBZu8lWDB2HsOjCWmy2P8LNZosae1OO2mlOarr7+JNWvpui3BN15/Ld/rHZANa85Vl53PDT/5Y5Xt7xffto/nznQY11c8V4EX26lBZvZe4Frn3O3x51uBK51z70/VOR7XORl/fiqu82Lefg8ePOiOHTtWwZ8ghBBC+IGZ/Ydz7uBo+TgJXCeBS1KfLwZO5dUxswbwQ8BLxZoqhBBCzBfjiPGjwOVmdpmZNYFbgCMjdY4Avx5v3wz887nqLxZCCCF8Z8fOlrgP+E7gISAA7nfOHTezDwPHnHNHgL8GPmlmJ4g84lt2s9FCCCFEnRgr88E5dxQ4OlJ2d2q7A7y32qYJIYQQ84Fm4BJCCCGmjMRYCCGEmDISYyGEEGLKSIyFEEKIKSMxFkIIIaaMxFgIIYSYMhJjIYQQYspIjIUQQogpIzEWQgghpsyOqzbt2oHNTgPfq3CXFzCyZKPYhOyTj2yTj2wze+ic5OODbX7cOXfhaOHUxLhqzOxY1rJUIkL2yUe2yUe2mT10TvLx2TYKUwshhBBTRmIshBBCTJk6ifG9027AjCP75CPb5CPbzB46J/l4a5va9BkLIYQQvlInz1gIIYTwkl0TYzO7xMz+xcy+Y2bHzex34vLzzeyrZvbd+P28uPxXzezx+PWwmb0tta/rzOxJMzthZoe3OeZXzOwVM/vSSPmn4t8/YWb3m1mY8/vLzOyRuG0PmlkzLn+XmX3TzPpmdrNss8k2l8Z/y3/G7bt+Dm1zZ3wMZ2YXpMqvMbNXzeyx+HX3HNoms952bfONis/L/Wb2gpk9scMxM89f3rWY8fu8/+dZvtdN2zaV3uu24JzblRewD3h7vP064L+BA8BHgcNx+WHgI/H2O4Dz4u33AI/E2wHwFPAmoAl8CziQc8yfB34J+NJI+fWAxa9PA7+Z8/u/A26Jt/8iqQfsB64APgHcLNtsss29qe0DwDNzaJufiq+RZ4ALUuXXjO5zDm2TWS+vbT6+qjov8ed3AW8HntjmeLnnL+9azNiHV/e6GbFNpfe6Lcc9hxfsF4F3A08C+1In6smMuucB/xdvXw08lPruLuCubY5zDdvcAIEPAn+YUW5Eg8UbWceNyx6o4gKtk22AvwR+P1X+8DzZZqTOpn/ynfY5T7bZ4foatq0Or6LnJVW2n+0FZ8fzN3otjnzn3b1uFmzDLt/rzkmfsZntJ3oqeQT4EefccwDx+w9n/OQQ8OV4+yLg2dR3J+OyIu0IgVuBr2R8/UbgFedcv+xxJmzTfvy2zT3Ar5nZSeAo8P4ix89p035m3zY7cbWZfcvMvmxmbyly/Jw27ccj2+xQL902ryl5Xsal7Pnz8V43Lrtpm3vYpXsdQKPKnWVhZnuBzwO/65w7Y2Y71f85opPwzqQoo5or2JyPA//qnPta1qErPM5Y1MQ27wMecM79sZldDXzSzN7qnFsv2I7ogP7YZju+STT13Urcv/QF4PKCbRjiqW0y62W0zVsqOC9jHyqjbJLz5+O9buxDZZRVZZtdudcl7KpnHD8Nfx74lHPu7+Pi581sX/z9PuCFVP0rgPuAG51zL8bFJ4FLUru9GDhlZlfZRmLMDWO05UPAhcDvpcoein9/H1Fo4g1mljygXAycmvyvHo8a2eYQUR8LzrlvAC2i+WEL45ltcnHOnXHOrcTbR4Fwu8SRcfDRNln1tmmbl1R0XvL2fUnqvPwGOedvh334fq/L2/e5tE3l97pNVN0nMBJ7/wTwZyPlf8TmjvuPxtuXAieAd4zUbwBPA5ex0SH/lm2Oew1bk01uBx4GlnZo82fZ3HH/WyPfP0A1SQ21sQ1RGOm2ePvNRBeuzZNtUvWfYXOf8Y8mtgCuBP533myTVy+vbT6+qjovqd/tZ/t+0R3P3+i1mLEPr+51s2AbKr7XbTnuLl6g7yRy7x8HHotf1xPF5P8J+G78fn5c/z7g5VTdY6l9XU+UhfcU8AfbHPNrwGmgTfSEdG1c3o9/m+z77pzfvwn49/hi+CywGJf/TLy/VeBF4LhsM7TNAeDr8UX/GPALc2ibD8S/6xP9g94Xl98JHI9t82+UFB5PbZNZb7u2+faq+Lx8GngO6MX2PpRzzMzzl3ctTvD/PMv3umnbptJ73ehLM3AJIYQQU0YzcAkhhBBTRmIshBBCTBmJsRBCCDFlJMZCCCHElJEYCyGEEFNGYiyEEEJMGYmxEEIIMWUkxkIIIcSU+X+hXHIb1DpvEwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "for i in fire_regions:\n",
    "    plt.figure(figsize=(8,6))\n",
    "    plt.plot(np.exp(test[:41][i]) - 1, label = 'actual')\n",
    "    plt.plot(np.exp(yhat_df[i]) - 1, label ='pred')\n",
    "    plt.title('forecast for {}'.format(i))\n",
    "    plt.legend()\n",
    "    plt.show();"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "21.072646229146606"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.metrics import mean_absolute_error\n",
    "pred_flat = np.exp(yhat_df[fire_regions].values.flatten()) - 1\n",
    "test_flat = np.exp(test[fire_regions].iloc[:41, :].values.flatten()) - 1\n",
    "mean_absolute_error(test_flat, pred_flat)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "66.46415643150922"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.metrics import mean_squared_error\n",
    "pred_flat = np.exp(yhat_df[fire_regions].values.flatten()) - 1\n",
    "test_flat = np.exp(test[fire_regions].iloc[:41, :].values.flatten()) - 1\n",
    "np.sqrt(mean_squared_error(test_flat, pred_flat))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1, 287)"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# submission\n",
    "x_input = np.array(merge_df.iloc[-120:].values)\n",
    "x_input = scaler.transform(x_input)\n",
    "\n",
    "x_input = x_input.reshape((1, n_steps_in, n_features))\n",
    "\n",
    "yhat = model.predict(x_input, verbose=0)\n",
    "yhat.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "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>Estimated_fire_area_NSW</th>\n",
       "      <th>Estimated_fire_area_NT</th>\n",
       "      <th>Estimated_fire_area_QL</th>\n",
       "      <th>Estimated_fire_area_SA</th>\n",
       "      <th>Estimated_fire_area_TA</th>\n",
       "      <th>Estimated_fire_area_VI</th>\n",
       "      <th>Estimated_fire_area_WA</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2021-01-19</th>\n",
       "      <td>1.277711</td>\n",
       "      <td>1.170450</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.313626</td>\n",
       "      <td>0.425741</td>\n",
       "      <td>0.715501</td>\n",
       "      <td>2.691141</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-01-20</th>\n",
       "      <td>1.235250</td>\n",
       "      <td>1.145159</td>\n",
       "      <td>2.586200</td>\n",
       "      <td>0.448582</td>\n",
       "      <td>0.465539</td>\n",
       "      <td>0.942946</td>\n",
       "      <td>2.760456</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-01-21</th>\n",
       "      <td>1.136686</td>\n",
       "      <td>1.147452</td>\n",
       "      <td>2.570229</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.513515</td>\n",
       "      <td>0.764036</td>\n",
       "      <td>2.738247</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-01-22</th>\n",
       "      <td>1.230877</td>\n",
       "      <td>1.307466</td>\n",
       "      <td>2.552538</td>\n",
       "      <td>0.421692</td>\n",
       "      <td>0.453735</td>\n",
       "      <td>0.884267</td>\n",
       "      <td>2.654948</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-01-23</th>\n",
       "      <td>1.299196</td>\n",
       "      <td>1.049376</td>\n",
       "      <td>2.425688</td>\n",
       "      <td>0.487095</td>\n",
       "      <td>0.490735</td>\n",
       "      <td>0.716635</td>\n",
       "      <td>2.803262</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-01-24</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.070312</td>\n",
       "      <td>2.438493</td>\n",
       "      <td>0.488286</td>\n",
       "      <td>0.559228</td>\n",
       "      <td>0.851720</td>\n",
       "      <td>2.639518</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-01-25</th>\n",
       "      <td>1.177327</td>\n",
       "      <td>1.121211</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.430117</td>\n",
       "      <td>0.606460</td>\n",
       "      <td>0.566472</td>\n",
       "      <td>2.715783</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-01-26</th>\n",
       "      <td>1.098833</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2.332636</td>\n",
       "      <td>0.471816</td>\n",
       "      <td>0.503438</td>\n",
       "      <td>0.709166</td>\n",
       "      <td>2.698100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-01-27</th>\n",
       "      <td>1.207018</td>\n",
       "      <td>1.132660</td>\n",
       "      <td>2.363791</td>\n",
       "      <td>0.398389</td>\n",
       "      <td>0.488981</td>\n",
       "      <td>0.826211</td>\n",
       "      <td>2.699170</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-01-28</th>\n",
       "      <td>1.329709</td>\n",
       "      <td>1.096810</td>\n",
       "      <td>2.420703</td>\n",
       "      <td>0.563184</td>\n",
       "      <td>0.463129</td>\n",
       "      <td>0.754954</td>\n",
       "      <td>2.497926</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-01-29</th>\n",
       "      <td>1.390766</td>\n",
       "      <td>1.088568</td>\n",
       "      <td>2.401286</td>\n",
       "      <td>0.463133</td>\n",
       "      <td>0.588192</td>\n",
       "      <td>0.702196</td>\n",
       "      <td>2.708305</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-01-30</th>\n",
       "      <td>1.299650</td>\n",
       "      <td>1.176784</td>\n",
       "      <td>2.444452</td>\n",
       "      <td>0.468201</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.870576</td>\n",
       "      <td>2.676711</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-01-31</th>\n",
       "      <td>1.279246</td>\n",
       "      <td>1.177087</td>\n",
       "      <td>2.444247</td>\n",
       "      <td>0.387157</td>\n",
       "      <td>0.488392</td>\n",
       "      <td>0.756739</td>\n",
       "      <td>2.707637</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-02-01</th>\n",
       "      <td>1.483171</td>\n",
       "      <td>1.074783</td>\n",
       "      <td>2.329539</td>\n",
       "      <td>0.557188</td>\n",
       "      <td>0.688913</td>\n",
       "      <td>0.789989</td>\n",
       "      <td>2.755858</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-02-02</th>\n",
       "      <td>1.419580</td>\n",
       "      <td>1.324003</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.440109</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.867065</td>\n",
       "      <td>2.743337</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-02-03</th>\n",
       "      <td>1.408136</td>\n",
       "      <td>1.076525</td>\n",
       "      <td>2.352344</td>\n",
       "      <td>0.508763</td>\n",
       "      <td>0.477047</td>\n",
       "      <td>0.803348</td>\n",
       "      <td>2.723411</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-02-04</th>\n",
       "      <td>1.282485</td>\n",
       "      <td>0.901542</td>\n",
       "      <td>2.438865</td>\n",
       "      <td>0.490257</td>\n",
       "      <td>0.538567</td>\n",
       "      <td>0.875495</td>\n",
       "      <td>2.596506</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-02-05</th>\n",
       "      <td>1.383366</td>\n",
       "      <td>0.982591</td>\n",
       "      <td>2.348564</td>\n",
       "      <td>0.145205</td>\n",
       "      <td>0.325560</td>\n",
       "      <td>0.894404</td>\n",
       "      <td>2.538219</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-02-06</th>\n",
       "      <td>1.301869</td>\n",
       "      <td>1.028274</td>\n",
       "      <td>2.446336</td>\n",
       "      <td>0.637284</td>\n",
       "      <td>0.465423</td>\n",
       "      <td>0.848755</td>\n",
       "      <td>2.490672</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-02-07</th>\n",
       "      <td>1.338688</td>\n",
       "      <td>0.940284</td>\n",
       "      <td>2.416660</td>\n",
       "      <td>0.512603</td>\n",
       "      <td>0.451310</td>\n",
       "      <td>0.687961</td>\n",
       "      <td>2.554772</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-02-08</th>\n",
       "      <td>1.444080</td>\n",
       "      <td>1.090775</td>\n",
       "      <td>2.279190</td>\n",
       "      <td>0.276679</td>\n",
       "      <td>0.541348</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2.348565</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-02-09</th>\n",
       "      <td>1.369038</td>\n",
       "      <td>0.876424</td>\n",
       "      <td>2.273729</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.939540</td>\n",
       "      <td>2.447760</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-02-10</th>\n",
       "      <td>1.297750</td>\n",
       "      <td>0.899707</td>\n",
       "      <td>2.258745</td>\n",
       "      <td>0.313340</td>\n",
       "      <td>0.473283</td>\n",
       "      <td>0.856095</td>\n",
       "      <td>2.483227</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-02-11</th>\n",
       "      <td>1.411862</td>\n",
       "      <td>1.141287</td>\n",
       "      <td>2.223997</td>\n",
       "      <td>0.364736</td>\n",
       "      <td>0.515678</td>\n",
       "      <td>0.680598</td>\n",
       "      <td>2.401755</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-02-12</th>\n",
       "      <td>1.502401</td>\n",
       "      <td>1.012040</td>\n",
       "      <td>2.098588</td>\n",
       "      <td>0.294853</td>\n",
       "      <td>0.473108</td>\n",
       "      <td>0.735516</td>\n",
       "      <td>2.588888</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-02-13</th>\n",
       "      <td>1.542311</td>\n",
       "      <td>0.827019</td>\n",
       "      <td>2.174872</td>\n",
       "      <td>0.423838</td>\n",
       "      <td>0.654752</td>\n",
       "      <td>1.063111</td>\n",
       "      <td>2.381424</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-02-14</th>\n",
       "      <td>1.352363</td>\n",
       "      <td>1.031140</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.370191</td>\n",
       "      <td>0.350365</td>\n",
       "      <td>1.009438</td>\n",
       "      <td>2.477999</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-02-15</th>\n",
       "      <td>1.561676</td>\n",
       "      <td>0.893991</td>\n",
       "      <td>2.008160</td>\n",
       "      <td>0.371687</td>\n",
       "      <td>0.419413</td>\n",
       "      <td>0.926736</td>\n",
       "      <td>2.455959</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-02-16</th>\n",
       "      <td>1.436063</td>\n",
       "      <td>0.794224</td>\n",
       "      <td>2.077960</td>\n",
       "      <td>0.448062</td>\n",
       "      <td>0.503306</td>\n",
       "      <td>0.848793</td>\n",
       "      <td>2.369667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-02-17</th>\n",
       "      <td>1.386202</td>\n",
       "      <td>1.032637</td>\n",
       "      <td>1.998737</td>\n",
       "      <td>0.367217</td>\n",
       "      <td>0.373905</td>\n",
       "      <td>0.869644</td>\n",
       "      <td>2.325793</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-02-18</th>\n",
       "      <td>1.409169</td>\n",
       "      <td>0.839159</td>\n",
       "      <td>1.920304</td>\n",
       "      <td>0.496600</td>\n",
       "      <td>0.437408</td>\n",
       "      <td>1.027101</td>\n",
       "      <td>2.540306</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-02-19</th>\n",
       "      <td>1.460102</td>\n",
       "      <td>0.917622</td>\n",
       "      <td>2.145368</td>\n",
       "      <td>0.354688</td>\n",
       "      <td>0.421339</td>\n",
       "      <td>0.956443</td>\n",
       "      <td>2.504966</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-02-20</th>\n",
       "      <td>1.680008</td>\n",
       "      <td>0.934984</td>\n",
       "      <td>1.944323</td>\n",
       "      <td>0.390791</td>\n",
       "      <td>0.693561</td>\n",
       "      <td>0.920785</td>\n",
       "      <td>2.512398</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-02-21</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.982039</td>\n",
       "      <td>0.361109</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.001109</td>\n",
       "      <td>2.462173</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-02-22</th>\n",
       "      <td>1.446575</td>\n",
       "      <td>0.821969</td>\n",
       "      <td>2.048856</td>\n",
       "      <td>0.292653</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.083013</td>\n",
       "      <td>2.548474</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-02-23</th>\n",
       "      <td>1.621282</td>\n",
       "      <td>0.898810</td>\n",
       "      <td>1.875493</td>\n",
       "      <td>0.321658</td>\n",
       "      <td>0.470358</td>\n",
       "      <td>0.801828</td>\n",
       "      <td>2.150457</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-02-24</th>\n",
       "      <td>1.546992</td>\n",
       "      <td>0.824173</td>\n",
       "      <td>1.783650</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.529834</td>\n",
       "      <td>0.863876</td>\n",
       "      <td>2.415357</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-02-25</th>\n",
       "      <td>1.436467</td>\n",
       "      <td>0.898403</td>\n",
       "      <td>1.780871</td>\n",
       "      <td>0.594314</td>\n",
       "      <td>0.565871</td>\n",
       "      <td>0.938489</td>\n",
       "      <td>2.295692</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-02-26</th>\n",
       "      <td>1.511240</td>\n",
       "      <td>0.990597</td>\n",
       "      <td>1.775682</td>\n",
       "      <td>0.377302</td>\n",
       "      <td>0.553075</td>\n",
       "      <td>0.783862</td>\n",
       "      <td>2.324946</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-02-27</th>\n",
       "      <td>1.536477</td>\n",
       "      <td>0.987932</td>\n",
       "      <td>1.788684</td>\n",
       "      <td>0.599497</td>\n",
       "      <td>0.537633</td>\n",
       "      <td>0.902503</td>\n",
       "      <td>2.409715</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-02-28</th>\n",
       "      <td>1.728261</td>\n",
       "      <td>1.348676</td>\n",
       "      <td>1.697415</td>\n",
       "      <td>0.323072</td>\n",
       "      <td>0.599202</td>\n",
       "      <td>0.991112</td>\n",
       "      <td>2.610270</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            Estimated_fire_area_NSW  Estimated_fire_area_NT  \\\n",
       "2021-01-19                 1.277711                1.170450   \n",
       "2021-01-20                 1.235250                1.145159   \n",
       "2021-01-21                 1.136686                1.147452   \n",
       "2021-01-22                 1.230877                1.307466   \n",
       "2021-01-23                 1.299196                1.049376   \n",
       "2021-01-24                 0.000000                1.070312   \n",
       "2021-01-25                 1.177327                1.121211   \n",
       "2021-01-26                 1.098833                0.000000   \n",
       "2021-01-27                 1.207018                1.132660   \n",
       "2021-01-28                 1.329709                1.096810   \n",
       "2021-01-29                 1.390766                1.088568   \n",
       "2021-01-30                 1.299650                1.176784   \n",
       "2021-01-31                 1.279246                1.177087   \n",
       "2021-02-01                 1.483171                1.074783   \n",
       "2021-02-02                 1.419580                1.324003   \n",
       "2021-02-03                 1.408136                1.076525   \n",
       "2021-02-04                 1.282485                0.901542   \n",
       "2021-02-05                 1.383366                0.982591   \n",
       "2021-02-06                 1.301869                1.028274   \n",
       "2021-02-07                 1.338688                0.940284   \n",
       "2021-02-08                 1.444080                1.090775   \n",
       "2021-02-09                 1.369038                0.876424   \n",
       "2021-02-10                 1.297750                0.899707   \n",
       "2021-02-11                 1.411862                1.141287   \n",
       "2021-02-12                 1.502401                1.012040   \n",
       "2021-02-13                 1.542311                0.827019   \n",
       "2021-02-14                 1.352363                1.031140   \n",
       "2021-02-15                 1.561676                0.893991   \n",
       "2021-02-16                 1.436063                0.794224   \n",
       "2021-02-17                 1.386202                1.032637   \n",
       "2021-02-18                 1.409169                0.839159   \n",
       "2021-02-19                 1.460102                0.917622   \n",
       "2021-02-20                 1.680008                0.934984   \n",
       "2021-02-21                 0.000000                0.000000   \n",
       "2021-02-22                 1.446575                0.821969   \n",
       "2021-02-23                 1.621282                0.898810   \n",
       "2021-02-24                 1.546992                0.824173   \n",
       "2021-02-25                 1.436467                0.898403   \n",
       "2021-02-26                 1.511240                0.990597   \n",
       "2021-02-27                 1.536477                0.987932   \n",
       "2021-02-28                 1.728261                1.348676   \n",
       "\n",
       "            Estimated_fire_area_QL  Estimated_fire_area_SA  \\\n",
       "2021-01-19                0.000000                0.313626   \n",
       "2021-01-20                2.586200                0.448582   \n",
       "2021-01-21                2.570229                0.000000   \n",
       "2021-01-22                2.552538                0.421692   \n",
       "2021-01-23                2.425688                0.487095   \n",
       "2021-01-24                2.438493                0.488286   \n",
       "2021-01-25                0.000000                0.430117   \n",
       "2021-01-26                2.332636                0.471816   \n",
       "2021-01-27                2.363791                0.398389   \n",
       "2021-01-28                2.420703                0.563184   \n",
       "2021-01-29                2.401286                0.463133   \n",
       "2021-01-30                2.444452                0.468201   \n",
       "2021-01-31                2.444247                0.387157   \n",
       "2021-02-01                2.329539                0.557188   \n",
       "2021-02-02                0.000000                0.440109   \n",
       "2021-02-03                2.352344                0.508763   \n",
       "2021-02-04                2.438865                0.490257   \n",
       "2021-02-05                2.348564                0.145205   \n",
       "2021-02-06                2.446336                0.637284   \n",
       "2021-02-07                2.416660                0.512603   \n",
       "2021-02-08                2.279190                0.276679   \n",
       "2021-02-09                2.273729                0.000000   \n",
       "2021-02-10                2.258745                0.313340   \n",
       "2021-02-11                2.223997                0.364736   \n",
       "2021-02-12                2.098588                0.294853   \n",
       "2021-02-13                2.174872                0.423838   \n",
       "2021-02-14                0.000000                0.370191   \n",
       "2021-02-15                2.008160                0.371687   \n",
       "2021-02-16                2.077960                0.448062   \n",
       "2021-02-17                1.998737                0.367217   \n",
       "2021-02-18                1.920304                0.496600   \n",
       "2021-02-19                2.145368                0.354688   \n",
       "2021-02-20                1.944323                0.390791   \n",
       "2021-02-21                1.982039                0.361109   \n",
       "2021-02-22                2.048856                0.292653   \n",
       "2021-02-23                1.875493                0.321658   \n",
       "2021-02-24                1.783650                0.000000   \n",
       "2021-02-25                1.780871                0.594314   \n",
       "2021-02-26                1.775682                0.377302   \n",
       "2021-02-27                1.788684                0.599497   \n",
       "2021-02-28                1.697415                0.323072   \n",
       "\n",
       "            Estimated_fire_area_TA  Estimated_fire_area_VI  \\\n",
       "2021-01-19                0.425741                0.715501   \n",
       "2021-01-20                0.465539                0.942946   \n",
       "2021-01-21                0.513515                0.764036   \n",
       "2021-01-22                0.453735                0.884267   \n",
       "2021-01-23                0.490735                0.716635   \n",
       "2021-01-24                0.559228                0.851720   \n",
       "2021-01-25                0.606460                0.566472   \n",
       "2021-01-26                0.503438                0.709166   \n",
       "2021-01-27                0.488981                0.826211   \n",
       "2021-01-28                0.463129                0.754954   \n",
       "2021-01-29                0.588192                0.702196   \n",
       "2021-01-30                0.000000                0.870576   \n",
       "2021-01-31                0.488392                0.756739   \n",
       "2021-02-01                0.688913                0.789989   \n",
       "2021-02-02                0.000000                0.867065   \n",
       "2021-02-03                0.477047                0.803348   \n",
       "2021-02-04                0.538567                0.875495   \n",
       "2021-02-05                0.325560                0.894404   \n",
       "2021-02-06                0.465423                0.848755   \n",
       "2021-02-07                0.451310                0.687961   \n",
       "2021-02-08                0.541348                0.000000   \n",
       "2021-02-09                0.000000                0.939540   \n",
       "2021-02-10                0.473283                0.856095   \n",
       "2021-02-11                0.515678                0.680598   \n",
       "2021-02-12                0.473108                0.735516   \n",
       "2021-02-13                0.654752                1.063111   \n",
       "2021-02-14                0.350365                1.009438   \n",
       "2021-02-15                0.419413                0.926736   \n",
       "2021-02-16                0.503306                0.848793   \n",
       "2021-02-17                0.373905                0.869644   \n",
       "2021-02-18                0.437408                1.027101   \n",
       "2021-02-19                0.421339                0.956443   \n",
       "2021-02-20                0.693561                0.920785   \n",
       "2021-02-21                0.000000                1.001109   \n",
       "2021-02-22                0.000000                1.083013   \n",
       "2021-02-23                0.470358                0.801828   \n",
       "2021-02-24                0.529834                0.863876   \n",
       "2021-02-25                0.565871                0.938489   \n",
       "2021-02-26                0.553075                0.783862   \n",
       "2021-02-27                0.537633                0.902503   \n",
       "2021-02-28                0.599202                0.991112   \n",
       "\n",
       "            Estimated_fire_area_WA  \n",
       "2021-01-19                2.691141  \n",
       "2021-01-20                2.760456  \n",
       "2021-01-21                2.738247  \n",
       "2021-01-22                2.654948  \n",
       "2021-01-23                2.803262  \n",
       "2021-01-24                2.639518  \n",
       "2021-01-25                2.715783  \n",
       "2021-01-26                2.698100  \n",
       "2021-01-27                2.699170  \n",
       "2021-01-28                2.497926  \n",
       "2021-01-29                2.708305  \n",
       "2021-01-30                2.676711  \n",
       "2021-01-31                2.707637  \n",
       "2021-02-01                2.755858  \n",
       "2021-02-02                2.743337  \n",
       "2021-02-03                2.723411  \n",
       "2021-02-04                2.596506  \n",
       "2021-02-05                2.538219  \n",
       "2021-02-06                2.490672  \n",
       "2021-02-07                2.554772  \n",
       "2021-02-08                2.348565  \n",
       "2021-02-09                2.447760  \n",
       "2021-02-10                2.483227  \n",
       "2021-02-11                2.401755  \n",
       "2021-02-12                2.588888  \n",
       "2021-02-13                2.381424  \n",
       "2021-02-14                2.477999  \n",
       "2021-02-15                2.455959  \n",
       "2021-02-16                2.369667  \n",
       "2021-02-17                2.325793  \n",
       "2021-02-18                2.540306  \n",
       "2021-02-19                2.504966  \n",
       "2021-02-20                2.512398  \n",
       "2021-02-21                2.462173  \n",
       "2021-02-22                2.548474  \n",
       "2021-02-23                2.150457  \n",
       "2021-02-24                2.415357  \n",
       "2021-02-25                2.295692  \n",
       "2021-02-26                2.324946  \n",
       "2021-02-27                2.409715  \n",
       "2021-02-28                2.610270  "
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "yhat_df = pd.DataFrame(yhat.reshape(41,7), index = pd.date_range(start = '2021-01-19', end = '2021-02-28'), columns = fire_regions)\n",
    "\n",
    "\n",
    "for i in range(70):\n",
    "    yhat_df['dummy_{}'.format(i)] = 0\n",
    "    \n",
    "_yhat_df = scaler.inverse_transform(yhat_df)\n",
    "\n",
    "yhat_df = pd.DataFrame(data = _yhat_df,\n",
    "                      index = yhat_df.index,\n",
    "                      columns = yhat_df.columns)\n",
    "\n",
    "yhat_df[fire_regions]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [],
   "source": [
    "submission = yhat_df['2021-02-01':'2021-02-28'][fire_regions]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [],
   "source": [
    "submission_df = pd.melt(submission.reset_index().rename(columns={'index':'Date'}), id_vars = 'Date')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [],
   "source": [
    "submission_df.rename(columns = {'variable': 'Region', 'value': 'Estimated_fire_area'}, inplace=True)\n",
    "submission_df = pd.concat([submission_df[['Region']], submission_df[['Date']], submission_df['Estimated_fire_area']], axis = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [],
   "source": [
    "for i,x in zip(fire_regions, regions):\n",
    "    submission_df['Region'].replace(to_replace = i, value = x, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [],
   "source": [
    "submission_df['Date'] = submission_df['Date'].dt.strftime('%b-%d')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "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>Region</th>\n",
       "      <th>Date</th>\n",
       "      <th>Estimated_fire_area</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>NSW</td>\n",
       "      <td>Feb-01</td>\n",
       "      <td>3.406900</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>NSW</td>\n",
       "      <td>Feb-02</td>\n",
       "      <td>3.135382</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>NSW</td>\n",
       "      <td>Feb-03</td>\n",
       "      <td>3.088329</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>NSW</td>\n",
       "      <td>Feb-04</td>\n",
       "      <td>2.605589</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>NSW</td>\n",
       "      <td>Feb-05</td>\n",
       "      <td>2.988302</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>191</th>\n",
       "      <td>WA</td>\n",
       "      <td>Feb-24</td>\n",
       "      <td>10.193762</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>192</th>\n",
       "      <td>WA</td>\n",
       "      <td>Feb-25</td>\n",
       "      <td>8.931304</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>193</th>\n",
       "      <td>WA</td>\n",
       "      <td>Feb-26</td>\n",
       "      <td>9.226124</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>194</th>\n",
       "      <td>WA</td>\n",
       "      <td>Feb-27</td>\n",
       "      <td>10.130787</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>195</th>\n",
       "      <td>WA</td>\n",
       "      <td>Feb-28</td>\n",
       "      <td>12.602717</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>196 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    Region    Date  Estimated_fire_area\n",
       "0      NSW  Feb-01             3.406900\n",
       "1      NSW  Feb-02             3.135382\n",
       "2      NSW  Feb-03             3.088329\n",
       "3      NSW  Feb-04             2.605589\n",
       "4      NSW  Feb-05             2.988302\n",
       "..     ...     ...                  ...\n",
       "191     WA  Feb-24            10.193762\n",
       "192     WA  Feb-25             8.931304\n",
       "193     WA  Feb-26             9.226124\n",
       "194     WA  Feb-27            10.130787\n",
       "195     WA  Feb-28            12.602717\n",
       "\n",
       "[196 rows x 3 columns]"
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "submission_df['Estimated_fire_area'] = np.exp(submission_df['Estimated_fire_area']) - 1\n",
    "submission_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [],
   "source": [
    "# submission_df.to_csv('submissions/final_dcnn70.csv', index= False, header=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x86ea4a08>"
      ]
     },
     "execution_count": 76,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYEAAAEGCAYAAACD7ClEAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOydd3iUVfqw7zMtM5OeSQGSQKihE4pIsYAodlBRFHVV0MUOi/izLKu4++mqKwK6qyLqWtlgQSyshQVsEIqUhN4JkEBIMumZmWTK+f4YEomkzWRKEt77uuaazFvOeWYy8z7veaqQUqKgoKCgcG6iCrYACgoKCgrBQ1ECCgoKCucwihJQUFBQOIdRlICCgoLCOYyiBBQUFBTOYTTBFqC5xMbGypSUlGCLoaCgoNCm2LJlS6GUMq6h/W1GCaSkpLB58+Zgi6GgoKDQphBCHG1sv2IOUlBQUDiHUZSAgoKCwjmMogQUFBQUzmHajE9AQUFBoQa73U5OTg42my3YorQa9Ho9SUlJaLVaj85TlICCgkKbIycnh/DwcFJSUhBCBFucoCOlxGw2k5OTQ9euXT06VzEHKSgotDlsNhsmk0lRAKcRQmAymbxaGSlKQEFBoU2iKIC6ePt5tHtzkCUrH2eFnZCUSLQdQhFq5YujoKCgUEO7VwLWHYVYd5oBEDo1us7hhKREoOsSga5zBKoQdZAlVFBQaIuo1WoGDBiAw+Gga9eufPjhh0RFRXk8zokTJ5gxYwafffaZH6RsmnavBEy398VRYqM6u4yqo2VUZ5dRtvoYSEAF2o5hhHSJQJcSQUhKBOqIkGCLrKCg0AYwGAxkZmYCcOedd/Laa68xZ84cj8fp1KlT0BQAnANKAEATpUeTpseYFg+Ay+ag+lg5VdmlVGeXUflrHhUZJ9zHmvQYBsVhHByPNs4YTLEVFBTaCCNHjmT79u0AHDp0iAcffJCCggKMRiNvvfUWvXv35tChQ9x22204nU6uvPJK5s+fT0VFBdnZ2VxzzTXs3LkTm83G/fffz+bNm9FoNMyfP5+xY8fy3nvv8dVXX2GxWDh06BDXX389//jHP3wi+zmhBH6PSq9B3ysafa9oAKTThf1EJVXZZdj2F1H+w3HK1xxHlxyOcUg8hoFxqEM9i71VUFA4N3A6naxevZq7774bgOnTp7No0SJ69uzJxo0beeCBB1izZg0zZ85k5syZTJkyhUWLFtU71muvvQbAjh072Lt3L+PHj2f//v0AZGZmsm3bNkJCQkhNTeXhhx8mOTm5xfKfk0rg9wi1Cl1yOLrkcMIvTMRZVoUlswDL1lOUfHmIkhWH0afGYBwcj6FPDEKjBFUpKJzrWK1W0tLSyM7OZujQoVx22WVUVFSQkZHBTTfdVHtcVVUVAOvXr+eLL74A4NZbb+XRRx89a8y1a9fy8MMPA9C7d2+6dOlSqwTGjRtHZGQkAH379uXo0aOKEvAX6ogQwi9KIvyiJKpPVGDZlo8lMx/bbjPCoME4MBbjkAR0ncOVMDUFhXOUGp9AaWkp11xzDa+99hp33XUXUVFRtb4CT5FSNrgvJOQ3f6VarcbhcHg1x+9RbmmbQNcpjKiru9HxifOJndYffWo0lq35FLyRxal5m7Gfqgy2iAoKCkEkMjKSV199lXnz5mEwGOjatSuffvop4L6oZ2VlATBixAiWLVsGwNKlS+sd66KLLmLJkiUA7N+/n2PHjpGamupX+f2qBIQQ/xZC5Ashdtaz71EhhBRCxPpTBl8h1AJ9r2hMt/Sm41/OJ/rGXjgtDkq/b7RUt4KCwjnA4MGDGTRoEEuXLmXJkiW88847DBo0iH79+vHll18CsHDhQubPn8/w4cM5efJkrWnnTB544AGcTicDBgzg5ptv5r333quzAvAHorHlR4sHF+IioAL4QErZ/4ztycDbQG9gqJSysKmxhg0bJltbU5nSldmU/3CchFlD0cYrkUQKCoFiz5499OnTJ9hieITFYsFgMCCEYOnSpaSnp9cqCF9R3+cihNgipRzW0Dl+XQlIKX8GiurZtQB4DHe0fpslbFQnUKso/zkn2KIoKCi0crZs2UJaWhoDBw7k9ddf5+WXXw62SEAQHMNCiAlArpQyqymnqhBiOjAdoHPnzgGQzjPUYTpChyVQ+WsekeO7KIlmCgoKDXLhhRfW+gdaEwF1DAshjMAc4OnmHC+lXCylHCalHBYX12Cf5KASfmEiuCTl604EWxQFBQUFjwl0dFB3oCuQJYTIBpKArUKIDgGWw2doTAYMA2Kp3HASl803IVsKCgoKgSKgSkBKuUNKGS+lTJFSpgA5wBApZV4g5fA14RcnI6ucVG48GWxRFBQUFDzC3yGi6cB6IFUIkSOEuNuf8wULXWIYIT2iKF97AulwBVscBQUFhWbj7+igKVLKjlJKrZQySUr5zu/2pzQnPLQtEH5xEq7yaizb8oMtioKCQgAQQjB79uza1/PmzeOZZ54BYN++fYwZM4a0tDT69OnD9OnTAXc+QU02scPhIDQ0lI8++qh2jKFDh7J169bAvQmUjGGfEdIjCm2nUMp/zkG62nTkq4KCQjMICQnh888/p7Dw7PvYGTNmMGvWLDIzM9mzZ09tPaBRo0aRkZEBQFZWFqmpqbWvKysrOXz4MIMGDQrcm0BRAj5DCEH4xUk4CqzY9piDLY6CgoKf0Wg0TJ8+nQULFpy17+TJkyQlJdW+HjBgAACjR4+uvehnZGRw33331a4MNm3axJAhQ1CrA9voSikg50MM/eNQxxyl/Kcc9H2VJtgKCoHgr1/vYveJMp+O2bdTBHOv7dfkcQ8++CADBw7kscceq7N91qxZXHLJJYwaNYrx48czdepUoqKiGDVqFH/5y18AtxKYO3cu6enplJeXk5GRwejRo336PpqDshLwIUItCL8wkepj5VRn+/ZLqaCg0PqIiIjgjjvu4NVXX62zferUqezZs4ebbrqJH3/8kREjRlBVVUVKSgrV1dXk5eWxd+9eUlNTOe+889i4cSMZGRmMGjUq4O9BWQn4GOPQBMpWuVcDIV3PLhCloKDgW5pzx+5P/vSnPzFkyBCmTp1aZ3unTp2YNm0a06ZNo3///uzcuZOhQ4cycuRIPvvsMzp27IgQghEjRrBu3To2bdrEiBEjAi6/shLwMSqdmrCRnbDtLcKep5SZVlBo78TExDB58mTeeee34MfvvvsOu90OQF5eHmazmcTERMDtF1iwYAEjR44E3K0pP/jgAzp06OBVo/qWoigBPxA6shNCqxSWU1A4V5g9e3adKKGVK1fSv39/Bg0axOWXX85LL71Ehw7uwgijR4/m8OHDtUqgY8eOOJ3OoJiCwM+lpH1Jaywl3RglXx2iYsNJOjx2HpoopbCcgoIvaYulpANBqyslfS4TdmEiIKlYmxtsURQUFBQaRFECfkITrcc4MI7KTXm4LPZgi6OgoKBQL4oS8CNhFycjq51UKIXlFBQUWimKEvAjuo6hhPSKpmLdCaTdGWxxFBQU2hhSSvztt1WUgJ8JvzgJV4Wdyq1KYTkFBQXPkNVO7CcrcVX77yZSUQJ+JqRbJNqkMCqUwnIKCgoe4rI4QILQ+O9SrSgBP+MuLJeMw2zDurNdVM1WUFCg4VLSzz33HGlpaaSlpaFWq2v//n1piaaQUuKyOlDp1QiV/+qQKWUjAoChnwlNvIHiZQdQhWrRdw98VqCCgoJvqSkl/eSTTxIbG1u7fc6cOcyZMweAsLCw2iqhniKrnOCSCIN/L9PKSiAACJUgdtoA1JEhFP57J5YdBcEWqd1SvPwABW/vCLYYCucAjZWS9gUuqwOEQKX3rxJQVgIBQhMVQvx9Ayl8fzdF/9mLa4KdsJGdgi1Wu8Jlc1C5JR8cLhxmKxqTIdgiKQSCb5+APB8r/g4D4MoXmjysoVLSLSVQpiDwf4/hfwsh8oUQO8/Y9pIQYq8QYrsQYrkQ4pyxjaiMWmLv7o++dwwlXx6idGW238O/ziWs2wvhdI9nyw7F/6LgfxoqJd1SakxBKqPWp+PWh79XAu8B/wI+OGPb/4AnpZQOIcSLwJPA436Wo9Wg0qkx3d6X4uUHKF9zHFeFnaiJPRBqpQFNS6ncegpNnAGVXoN1ewERY5KDLZJCIGjGHbs/aaiUdEtwWdymIKH3f5cxfzea/xko+t22lVJKx+mXG4Cks05s5wi1IHpST8LHJlO5KQ/zkj1KMlkLcZitVGeXYRySgGFgHPYTldgLrcEWS+EcoL5S0i1BSonL5kBl0ASkO2GwHcPTgG8b2imEmC6E2CyE2FxQ0L6cqUIIIi9PIWpCd2x7zBS8szMgNYbaq/mpcms+CDAOiccw0B2pYd3evr4zCq2X35eSbgnSdtoU5OeooBqC5hgWQswBHMCSho6RUi4GFoO7lHSARAsoYaM6oQrTUvTxPvLf3E7stP5oIv1TeroquxTzR3uIvLIroUMT/DJHMJAuiWVbPiHdo2o/O12XCKzbC4m4pHOQpVNor1RUVNT+nZCQgMViafSY5uKy2kEVGFMQBGklIIS4E7gGuE2211tTDzAOjCN2Wn+cJVUUvJ6FPf/sL1NLsR0opvCdnbgq7FRsaF8F7aqzy3AW2TAOia/dZhgYiz2v0i+fpYL/cFnsnPrntnO26KJ0SVxWJyp9YExBEAQlIIS4ArcjeIKUUvmFnkbfPYq46QORThcFi7KoOuq7RvXWXWYK39uFxmQg7KIk7MfLcbQje3nl1lMInRpD/98SdowDYkGc2yYhKSUV609gO1gSbFGahXRJzOl7sedWYNttDrY4QUFWOUBKVMbAGWn8HSKaDqwHUoUQOUKIu3FHC4UD/xNCZAohFvlThraELjGM+PsHoTJoKFi8nXIf1BuyZOZjXrIbXacw4qYPIGyUOzfBktU+Lo6uaifWHYUYBsSi0v22fFZHhKBLiTinQ0XtuRWUfHmIwrd3UPD2DqqO+e7Gwh+Urcym6kAJ6sgQqnMq2q3/qjFcFofbFBQSGFMQ+D86aIqUsqOUUiulTJJSviOl7CGlTJZSpp1+3OdPGdoaGpOBuAfS0PeOofSbIxS+vQNHic2rsSo35VH08T5CUiKJvac/KqMWTVQIuq4RWDLz28WPzLbbjKxy1jEF1WAcGIfjlAX7qcogSBZ8LNsLQCWIuCIF+8lKCl7PovCD3djzWt/nYdlRQPmPOYQO7+CuvFtpx1laHWyxAop0SVw2Z8CigmoIdnSQQj2oQ7WYbu9D9I09qc6p4NTCrVi2eXbRLl+bS/HnB9D3iiZ2aj9UIb8tL41p8TgKrNhPtr6LgadUbs1HHRVCSNfIs/YZ+rtNQpbt595qQLok1u2F6HtFEzEmmQ6PDSPisi5UHSrh1CtbKfp4Hw5z6zAJ2k9VUvzpfnSdw4ma0B1tYph7e255kCULLC7baVNQgKKCalCUQCtFCEHosA4kzByMNiGUoo/3UZS+t8kwUiklZauPUbriMIb+Jkx/6IvQ1l1aGvrHgkpgyWzbJiFnWRVVB4oxDomvN7VeHa4jpGsk1h0F7WLV4wnVx8txllTVhsuqQjREjOtMx8fPI/yiJKw7C8l7eQvFyw/gLK0KmpwuqwPzh3sQOjWm2/sgNCq0HUNBQHWu55E1bRmX1QHqwJqCQFECrR6NyUDcvQOJuLwL1p1mTi3ciu1gcb3HSikp/S6bsv8dxTgknpgpfeqtQ64O1aLvFY01K79N9ziwbMsHCcbBZ5uCajAMjMWRb8Vx6tyKQbBmFYBGYOhrqrNdZdQSeWVXOvzfeYSe34HKzac4+dJmSr45jLOy7g2Gy+bAfqoS2/5iKn/No2zVUYo/P0Dhuzs5tXAr+a9nUp3j/d26dEn3iqTIhun2Pqgj3OG9Kp0aTbwReytXAjk5OUycOJGePXvSrVs3HnroIaqqqvjxxx+55pprPBpLuiQygAliZ6IUkGsDCJUgYmxn9D2jKfp4H4Vv7yRsdCcir+iK0Lov8tIlKfnqEJUbThI6oiNRE7o3WnjKmBZH0d4iqo+W1WtKae1IKancko+uczjaOGODxxn6x1Ly5SEs2wuI7BAaQAmDh3RJLNsLMKTGNFiBUh2hI3piD8IvTKJs1VEqfsmlcmMeuqQwnGXVOEurkfV0s1KFaVFHhqCODsGeW0H+G1lEXp5C2AWJHhc6K1t9DNveIqImdickpe53UJcUjm1vEVLKgF8Um4OUkhtuuIH777+fL7/8EqfTyfTp03nssce4/vrrPR7PbQoClcH/tYJ+j6IE2hC6pHDiHx5M6bdHqFjnDv2LuTkVbUIoxcv2Y9maT9jFSURekdLkD0ffx4TQqrBk5rdJJWDPrcCRbyHq+h6NHqcO0xHSPcqdOHZZl1Z5QfE1VYdLcVXYMQyKa/JYTYyemMmphF+cRNma4ziLbGgTjOh7Rbsv9pE693NECOoIXZ2Vpctip3jZAUq/OYJtfzExk1NRR+iaJaN1t5ny1ccwDokndETHs/brEsOwbDmFs7QaTZR/kidbwpo1a9Dr9bX1gtRqNQsWLKBLly5cdtllHo/nsjhArULoAm+cUZRAG0OlUxM9sQeG3jEUfbaf/Ncy0SWGUX2snIjLuhB+SXKzLnSqEDX6viasOwrdqwZ127IMWrbmg0a48wGawDAglpLlB7GfrETXKSwA0gUX6/YChE6FvndMs8/RJoRimtLbo3lURi0xt/eh8tc8Sr8+zKlXthA9qddZJqjfYy+wUPTxPrSJYURf36Pe7+uZzuGmlMCLm15kb9Fej2Rvit4xvXl8eMN1LXft2sXQoUPrbIuIiCAlJYWDBw96NJd0upBVDlShuqDcpLStX75CLfrUGBL+NBRD7xiqj5UTeU03IsZ19uhLZBwUh8viwHYg8MlELquj6YMaQDpcWLLyMfQxNavUrtsRfrrUdDtHOl1Ydxai72uqkzfhL4QQhA3vSPzDg1FHhGD+YDfFXx5ssCCiq8qB+cPdCI3A9Ic+ZwUt1NDancMNmam8CUBw2ZxuU1AAE8TORFkJtGHUoe47MZfFgTrUc1uivlc0wqDBmpmPwYO7xpZQdbSMstXHqNpfTOQ13Qi/INHjMWz7inFVOjA2s/6ROlRLSPcoLDsKiLjce5NQ9fFy1DF6rz7rQGE7WILL4sA4sGlTkC/RxhuJfzCN0u+yqVibS9XhUkxTeqM9ww8jpaT4k/04Cq3E3j0ATZS+wfFUOjXaBCPVOU0rgcbu2P1Fv379WLZsWZ1tZWVlnDp1itTUVFatWtXssVxWB2hUtf69QKOsBNo4QgivL0pCo8I4IBbrbjOuepyAvqTqaBkF7+yg4I0s7Lnl6DqHU/rfw15lLlu2nkIVpkXfM7rZ5xgHxuE027Cf8C43onLLKfJfy+TksxvIfzOL8l9yWk2c/ZlYswoQejX6Xs3/bHyF0KiIuqYbsdP646q0c+pf26jIOFF7d1z+Yw7WXWYir+zarD7b2sRw7LmtM3N43LhxWCwWPvjA3SrF6XQye/ZsHnroIQyG5ne0k05X0KKCavBICQghooUQw4UQF9U8/CWYQmAwDIpDVruw7fFPrZaq7FIK3j598T9R6Q5PfGw4cX8ciK5LBEWf7MN2qPnmKGelHeveIoxp8R414tH3NYFKeFVLyJ5XSckXB9GlRBA+NhlpdVD63yPkvbSZvAVbKP0+m+rj5UEPt5V2F9ZdZgz9YusNDQ4U+l7RJPxpCPruUZR8dQjz+7uxbMunbGU2hkFxhDVz9adLDGu1mcNCCJYvX85nn31Gz549MZlMqFSq2gbzq1evJikpqfaxfv36esepMYsGOkHsTJo9sxDiHmAm7iYwmcAI3HWBLvGPaAqBIKRrJOoIHZbMAoyDGo6395Sqw6WUrT5K1aFSVGFaIq/qSuiIjnXs1LF39CX/ze2YP9hN3L0Dm+W0tW4vAKest0xEY6hDtYT0iMKyvYCIZkRP1eCqcmD+aA9Cr8Z0Wx/U4Toix6fgMFux7inCtttM+U/HKf/hOKoIHYY+MRj6xRLSLTLgF2LbviJ3CY1mRAX5G3WYDtNd/ajIOOGOHtpbhLZDKNGTejb7s/fEORwMkpOT+eqrrwDIyMhgypQpbNmyhTFjxmC1Nm+V6LI6EEE0BYFnPoGZwHnABinlWCFEb+Cv/hFLIVAIlcAwKI6KjBO4LPYW9zStOlxC2apjVB0+ffG/uiuh53es10mpMmqJndqfgjcyKXx3F/H3D0IT07CdGNxlIrQdQ72K8jEOjKX4swPYcyrQJYc3ebyUkuJlB3CYrcT9cQDq8N/CHzUmA+EXJBJ+QSLOSju2fW6FYNmWT+XGPESImohxnQm/KHCN8yzbC1CFaghphqklEAghCB+dSEi3KCrW5hIxrrNHzmptx1BQQXVOBYZ+TUeBBZNRo0Zx9OhRj85xRwU5UYUHJyqoBk/Uj01KaQMQQoRIKfcCqf4RS6G5OFwONudtptrp/ZLZmBYPTollp/fRM/YCCwWLt1OweAf2AguR13Sjw2PnEX5hUqM/fE1UCLHT+iPtLgrf3XlW1mqdOfIt2I+Xe7wKqMHQ1wRqgWVH80xClRkn3PkFl6cQ0q3hC6s6VEvokARMt/el01MjMd3Vj5CukZR+c4TKzXleyeoprmontj1FGAbEtbp+1bqOocTc1KtJBf97VDo12nhjq40Qaim1pqAgRQXV4IkSyBFCRAFf4C4D/SVwwj9itT8sdgsOl/dhkQ0xb/M8pn4/lXGfjmPer/M4UnrE4zG0nULRxBqwellLyGVzYH5vF/a8SiKv6UbHx84j/ILEZt/1aRNCib2zL45iG+b3dzXopLZsPQWq00rLC1RGtzPZur2wSWdj1bEySr45gr5PjEd380KrwtA7BtMf+hDSM4rizw9iO1B/mQ9fYttjRtpdAY8K8jet2TncUlwWB0KrQtVAmGygaLYSkFJeL6UskVI+AzwFvANc5y/B2hM55Tlcvfxqpn43lSqn74p1/ffwf1myZwlXdb2K8zqcx5I9S5jwxQSmfT+Nb4982+zVgRACY1ocVUdKPS4mJqWk6JP9OIptmO7oS/gFiQ3GfjdGSNdITLf0pvp4OUXpe5HOuj/6mhaS+l4xdcwynmIYEIuzpIrq4w3XvHFW2ilashd1hI6Ym3p5XA4BQKhVmG7rgzbeiPmjPVT7uWKrJasQVYQOXUqEX+cJNK3ZOdwSpMOFrHYG1SFcg6fRQRcIIaZKKX/C7RT2PMj7HKPEVsL9q+7H6rCSWZDJU+ue8sldzf7i/fx1/V8ZEj+EZy94lvlj5vO/m/7HjMEzOFFxgsd+foxLP72U+Zvnc6zsWJPjGQbFgTxdg94DKn7JxbbbHfb3+/ovnmLoH0vUhO7Y9hRR8sXBOp9T1aESnKXVXpuCaufo5zYJNZQ4VlPUzFlRjem2Pi3ykaj0GkxT+6EKUWN+d6ffqnW6rA5s+4owDoj1SmG1ZrRJ7bOsdI0pSLQlJSCEmIu7LeSTpzdpgY/8IVR7weaw8fCahzlRcYLXx73OzCEz+fbIt7yR9UaLxi2vLmfWD7MI1YYy7+J5aFXuC1WsIZY/Dvwj39zwDYsuXcSQhCF8sPsDrl5+NfesvIfvs7/H7qzf5q6NM6JNDPOovHTV4VJKvzuCob+p2WF/TRE2shPhY5NPV638TXlZtuYj9BoMfRovSdAUKr3GXUF1R0G9IZ3lPxynan8xURO6o0tq2nncFJrIEExT++OqclL43i53oTAfY91lBqdsVq2gtobuDOdwe8JlbR2mIPBsJXA9MAGoBJBSnsDdJlKhHpwuJ0/88gRZBVk8f+HzDEkYwt397+a6HtfxRtYbrDi8wqtxXdLFnLVzOFFxgpcvfpk449k/fJVQMTpxNAvHLmTljSt5KO0hjpUd49GfHuXSzy5l48mN9Y5tTIvDnluBvaDpssvO8mrM6XvQxBiIvrGXT6MbIsZ3wTg0gfLVx6jYeBJXlQPrzkKMg2J9EkpnHBiHs7T6LJOQ7UAxZauOYhwcT+jwDi2epwZdx1BMt/XBfqoS83/2Ip0un40N7tWbOjqkWRFPbQ2h9Z1zWDpcOMuqfeZfeO655+jXrx8DBw4kLS2NjRvdvyuHw0FsbCxPPvlkvefVmoJaGInnKzz5RVVL96cnAYQQTdblFUL8WwiRL4TYeca2GCHE/4QQB04/Bz610c9IKXnx1xdZfWw1j533GONTxgNu2/vTI55mWMIwnl73NNvyt3k89r93/psfjv/A7GGzGZIwpMnj443x3DvoXr694VteG/caYdownsl4pl7fhHFgnLs5exNZvNIp3RczmxPT7X0aLFfsLUIIom/ogT41mpIvDlL8+UG303NI88pENIW+Twxo6iaOOUqrKFq6F028kagGipq1aM5e0URf35Oq/cWUfHHIZxciZ0U1VQeLMQ6Ka7cVUn3lHHZW2HGWVeEqb7wxU3NYv349K1asYOvWrWzfvp1Vq1aRnJwMwMqVK0lNTeWTTz6pV2bn6cZQrcEfAJ4pgU+EEG8CUUKIPwKrgLeaOOc94IrfbXsCWC2l7AmsPv26XfHervdI35vOHX3v4Pa+t9fZp1VrWTBmAZ3COjFzzUyOlx9v9rgZJzL457Z/cmXKldzW5zaPZFKr1FyUdBFzRswhpyKHD3Z9cPYxke42jZbMxjtxla3MpvpIKVHX9ahTG8aXCLWKmNv6oE0Mw5pVgCbWgK6zb+503SahGCw7Ct3NPJwuiv6zF2mXbqXmp8Jroed1IPwSt6mr/Ifm/98bw7rLDC4wtLOooDPRJbXcOSylrLXDO8urcDVQ4K65nDx5ktjYWEJC3ElssbGxdOrUCYD09HRmzpxJ586d2bBhQ53znBY7rrJqhEET1KzuM2mWKhLuW4yPgd5AGe78gKellP9r7Dwp5c9CiJTfbZ4IjDn99/vAj7h9De2Cbw5/w/wt87k85XJmD5td7zFR+iheG/cat/73Vh5c/SAfXfUREbrGozpOVpzk8Z8fp1tkN54Z9YzXd32jOo3ikuRLeGvHW1zb/Vo6hNY1exjS4ij5/CD23Ip6beLW3WbKf3I3BA9tZgE3b1Hp1MTe1Q/zkj2EDuvg0ztd48BYbLvNVB8tw7rL/RwzpXejDWp8QcRlXW4gjlwAACAASURBVHAWV1G28iiaaH2jXdGagzWrAE2cwZ1Y1U6pzRzOqT9zOO/vf6dqT+OlpKWUyGoXQqNym+OEQGhVNPSNCunTmw5//nOD440fP56//e1v9OrVi0svvZSbb76Ziy++GKvVyurVq3nzzTcpKSkhPT2dkSNHAu5QamexDaFTo4n2LGfCnzRLFZ02A30hpfyflPL/pJSPNqUAGiFBSnny9LgngQZ/BUKI6UKIzUKIzQUFrb8f7qaTm5izbg5DE4by3AXPoRINf7xdIrqwcOxCjpcfZ/aPs7G7Gl6iVjmrmPXjLOwuOwvGLMCobdmF6tHzHsXpcrJgy4Kz9hn7x7oTquoxCTnMVoo+cdeBj7q2e4tkaC7qMB3x9w7yucLR9zGBRkXJV4eoWJtL2KhOASm3IIQgelJPQrpFUvTZfo/qJv0eZ1kVVUdKMQxsv6YgOMM53BK/QE0QgFq478BdEpzem5fCwsLYsmULixcvJi4ujptvvpn33nuPFStWMHbsWIxGI5MmTWL58uU4nU5cVU4cZhtCo0ITa2hVUVyeGKU2CCHOk1L+6jdpfoeUcjGwGGDYsGGtOltkf/F+Zv4wky7hXXhl7CuEqJuudXJeh/OYO3IuT617iuc3Ps9TI56q98f8wqYX2GXexcIxC0mJTGmxrMnhydzZ707e2vEWt/S+hcHxg2v3qYzu/sOWrAIir+xa+2WVdifmj/aAEJhu6xPUWie+QBWixtA7GutOM7rkcCKv6hqwuYVGhekPfcl/Iwvzh3uIf2AQ2njPFbtle6G7x3I7jAo6k6acw43dsYN7FeA4ZQG1QBtnREqJs8iGy+ZAE2/0OkJHrVYzZswYxowZw4ABA3j//ffRarWsW7eOlJQUAMxmM6tXrmLMoNEItWh1CgA88wmMBdYLIQ4JIbYLIXYIIbZ7MecpIURHgNPP+V6M0arIq8zjgVUPYNQYeePSN4gMaX68/HU9ruPu/nfz6f5P+WD32Xb65QeW89n+z7i7/92M6zLOZzLfM+Ae4o3xPL/xeZyuuvZRY1ocrrJqqo6U1m4r/vIQ9pOVxNyc6nH6f2sl7IJEdF0iiLmtd8DtsyqDhtip/RBaQeG/d+Is99zebd1egLZjqFcKpK3REuewdLiQDletI1YIgToqBITAWVzl1Zj79u3jwIEDta8zMzOJi4tj7dq1HDt2jOzsbLKzs/nXq/8k/YMlIHArgFbYwc8Tia4EuuOuGnotcM3pZ0/5Crjz9N93Al96MUaroby6nAdWP0CFvYLXL32djmFn90ttihlDZnBZl8t4efPL/HDsh9rtu827eXbDs5zf8XweGvyQL8XGqDUye+hs9hTtYfnB5XX26fuYEDpVbZRQ5eY8LJtPET42OWDNZwJBSEqku2hdI81N/IkmWk/snf1wVdopeGs7toPNNw05imxUHytvl7kB9fGbc9jzhDtZT7lmoVahjgpBVjtxVXgeLVRRUcGdd95J3759GThwILt376Zv375ccskltc5i6XRx1ejxrFj5Dc5wVatxBP+eZpuDpJRHAYQQ8UCzfjVCiHTcTuBYIUQOMBd4AXek0d3AMeAmD2VuNdiddv70w584UnKE1y99ndQY7+rpqYSK5y54zu38/eVx3r/ifTqGduSRHx8hxhDDPy76BxqV78PJrux6JR/v+5hXt77K+JTxtc5plU6Noa8Jy45CQs/rQPEXhwjpHknEZV18LsO5ji4pHNOdfSn+7ACFb+9AnxpN5JVdm4y6sp4ugtfeagU1xG/O4QqPlbbL6kDo1GfdhasMGqRBg7OsCqFXe2QWGjp0KBkZGQ3ul04XjkIrMRHRnDp5KiCtPr3Fk4zhCUKIA8AR4CcgG/i2sXOklFOklB2llFopZZKU8h0ppVlKOU5K2fP0c1GL3kGQkFLyVMZTbMrbxN9G/42RnUa2aDyDxsCrl7xKZEgkD615iNk/zSbfks/8i+cTo/fP3bcQgieGP0FJVQlvZNbNYjakxSOtDgre2oHKqCFmSu9WZ8tsL+h7RNNh9jAir+pK1dFyTr2ylaJP9+No5K7XklWANjm83ZjmmsJb57B0uJB2V70x+b4wC9U7p0viMNuQDhcak75VKwDwzBz0/3A3ktkvpewKjAPW+UWqNsDx8uP89/B/uWfAPVzb3Rur2NnEGeP41yX/oqK6gk15m3jy/CcZEDfAJ2M3RB9THyb1mkT63nQOlRyq3a7vGYXKqEE6XJhu7Y06zPuibQpNI7Qqwi9KouNjwwi7MBFLZj55L22m9LsjZ5WasBdYsJ+oPGdWAeB95nBTnbtaahb6PVJKHGYrstqJJkbv80RKf+CJErBLKc2ASgihklL+AKT5Sa5WT77F7c8e3mG4T8dNjUll0WWLmHP+HG7seaNPx26Ihwc/jFFr5IVNL9TeDQm1iujJqZju6NviwnAKzUdl1BJ1VTc6zB6GcUAs5T/mkPePXylfm4t0uMtNWLMKQLhzHc4lvHEOu2v0qBu1x6sMGlSnzUItSSKriTqSVU7U0XpUhtZRFqIpPFECJUKIMOBnYIkQ4hXA99WwfMyR0iPsKNjh83GLbG4rlj9MNYPjB3NL71sCFvsdo4/hwbQH2XByA2uOr6ndbugd064cwW0JTYyemJtTiX94MNpOYZSuOEze/C1YsvKxbC9AlxKBOrL1tVz0J546h38r19y4OcYXZiEpJc7iKlxWB+rIENShbUMBgGdKYCJgAWYB3wGH8C46KKC8svUVns542ufjmm3uxuwmQ8uqWrYWJqdOpkdUD1769SWf9jxQaBm6xDBi7+5P7LT+qHRqitL34ci3tvvcgPo40zncHGrMaM0p19wSs5DL7nQrAIsdVbiuRf0ugoEnTWUqpZQuKaVDSvm+lPLV0+YhAIQQ6/0jYssw6U2YreamD/QQs9WMQBAV0jr6ubYUrUrL48MfJ7cil/d3vR9scRTOQAiBvlc08TMGE31TLwwDYs8pf0ANnjqHa5q4NzfqR2XQIJppFpIOF87yauynKnGcsrgVQJgOdUTbUgDgYVOZJmiVYQomg4mSqhKft3YsshURrY/2S+hmsBjRcQSXdr6Ut3e8TV5lYHrjKjQfoRKEDk1ocbObtorbORzaLCVQ08Tdk6YtQgh3baIGzELS6cJZUU3e3uOkDUpjyIhhJA/oRtfhvRl+zYUMG3M++fn5aLVa3nzzTY/fX7DwpRJolWUdTHoTEkmxzbd9Xs1Ws99CN4PJ7GGzG6wrpKAQbLSJYc3qMuayue/kPS3X/HuzkHRJnBY79kIr9pOVOEuqMEXFsHX9r2RmZXLfA/cz65FZZGZmkpmZybJlyxgxYgTp6elevb9g0DpT2HxIjc2+xobvK4psRZj07cMfcCZJ4UlM7T+Vb458w9ZTW4MtjoJCHdzOYUe9XeHOxGV1gEblVY2rM81C9pOV7ogfuwtVuA5NvBFNghF1REi9EUfp6em8/PLL5OTkkJub6/HcwcCXtoxWmUkUa3CH0RVa6+8p6y1mm5n+sf19OmZrYVr/aXxx8Ate2PQC6Veno1a17mQXhXOHGucwjt+6s/3yyX4Kj9c1Eclqp7tiqJe1emKTQhlxSbLbp2DUuDOOm4jWO378OHl5eQwfPpzJkyfz8ccf88gjj3g1fyDxtNF8FyHEpaf/Ngghziw4/wefSuYjau7Wfe0cNlvN7XIlAKfrCg1z1xV6ftPzLD+wnNVHV7Pp5Cb2mPeQU55DWXUZLunbNokKCk1R4xyWjZSBli4JkpZluAt3xVFNtB5ViKZZ4dpLly5l8uTJANxyyy1txiTU7JXA6W5i04EY3IXkkoBFuDOHkVLubPjs4OEPc5DVYcXisLSb8ND6uCLlCr45/A0f7/u4wWMEgjBtGOG6cMJ14ZgMJrpHdadnVE96RPWge1T3Fvc+UFA4kxrn8Jl9mi+c3KvOMfbTGbvaDqEB7bOQnp7OqVOnWLJkCQAnTpzgwIED9OzZM2AyeIMn5qAHgeHARgAp5YHTxeRaNUatEYPG4NOVQE2iWHtdCYA7UuLVS16l3F5OefVvj7KqMsqqy9yv7b9tK68uJ9+az6f7PsXmtNWOkxiW6FYK0T3oEeV+dI3sik7d9kLpFFoH2sQwcFiQUp51kZcuibQ5UYU27+7dV+zbt4/Kyso6foC5c+eydOlSnnrqqYDJ4Q2eKIEqKWV1zQcrhNDQSiOCfk+MPsanK4EahdIeo4PORAhBhC6iydaXZ+J0OTlRcYIDJQc4WHKQg8UHOVBygLW5a3FId5iuWqjpGtmVy7pcxnU9rqNTWCd/vQWFdoguKQwpK92dwTR1L/QumwOkDHjNnvT0dK6//vo62yZNmsQtt9zSrpTAT0KIPwMGIcRlwAPA1/4Ry7fEGmL9sxJox+Ygb1Gr1CRHJJMckcwlnS+p3W532jladpSDJW6lkFWQxaKsRSzKWsTITiO5oecNjE0eq6wQFJpEmxgGOaeQdudZETrS6gCVQIQEJpjhmWeeaXBfTZ+B1o4nSuBx4B5gB3Av8A3wtj+E8jUmvYlj5cd8Nl6NQmnP5iBfo1Vr3Sah6B5cwRUAnKg4wRcHv2D5weU8+tOjRIVEcW33a7mhxw30iO4RZIkVWiu6jqGQA65qFyrDb9ullLhsTnc0TzvuuexrmqUEhBAqYLuUsj/wln9F8j0mg4nMgkyfjVdbPM7Qvs1B/qZTWCceSHuAewfey4aTG1h2YBnpe9P5cPeHDIwbyKSek7gi5QrFuaxQB6F1N4iR9rrRadLmdJuCPEwQO9dp1qclpXQJIbKEEJ2llL67pQ4QJoOJYlsxDpfDJ2UezDYzYdqwZjWTV2gatUrN6MTRjE4cTZGtiK8Pfc3nBz5nbsZcXtz0Ild0vYLpA6eTGJYYbFEVWglCLZDVzjrOYZfVASJwpqD2gidXxI7ALiHEJqCyZqOUcoI3EwshZuE2L0ncJqapUkpb42d5R03piJKqktrksZbQXktGtAZi9DHc2e9O7uh7B1kFWXx+4HO+PfItW09tZek1SwnVNt52UeEcQaMCl6x1DrtNQQ5UhqaTuhTq4okS+KuvJhVCJAIzgL5SSqsQ4hPgFuA9X81xJjUXfrPV7BMlUGQrUpzCfkYIQVp8GmnxaVzb/VruWXkPczPm8tJFLyk/cgWE2v0dqHEOyyonuBRTkDd4Ukr6p/oeLZhbgzvSSAMYgRMtGKtRai7Yviod0Z6zhVsj53U4j5lDZvJ99vcs2bMk2OIotAJqlICr2u0X+M0UpCgBT2lSCQgh1p5+LhdClJ3xKBdClHkzqZQyF5gHHANOAqVSypXejNUcaktH+ChXoMhWpJiDAszUflMZmzyWlze/TGa+75z8Cm0UIRBat3O41hSkV7esVIQHjBkzhu+//77OtoULF3LVVVfRv3/bqinWnJXAHQBSynApZcQZj3ApZfOziM5ACBGNu1NZV6ATECqEuL2e46YLITYLITYXFBR4MxVwRukIH+QKOFwOSqpKFHNQgBFC8OwFz9IxrCOzf5ztl0ZBCm0LoVW7ncPVTnBKj3oHtJQpU6awdOnSOtuWLl3Kk08+GTAZfEVzlMCnAEKI1T6c91LgiJSyQEppBz4HRv3+ICnlYinlMCnlsLg47zspGTVG9Gq9Ty4cxbZiJFIxBwWBCF0E88fMp7S6lMd/eRyny/um4AB5lXmsOrrKR9IpBBqhczuHXRV2EAQ0S/jGG29kxYoVVFW5W7FmZ2dz4sQJkpKSAiaDr2jOp6YSQswFegkhzqqLKqWc78W8x4ARQggjYMVdhG6zF+M0CyEEJoPJJ+YgJUcguPSO6c2c8+fwdMbTvJb5GjOGzPBqnHW563jilycoqSrh+0nfK6Ur2iDidNvIH9PfoTDvmFe9Axoivks3xt41vcH9JpOJ4cOH89133zFx4kSWLl3KzTff3CaDFprzqd0C2HArjPB6Hh4jpdwIfAZsxR0eqgIWezNWczEZTD5xDCvZwsHn+p7XM6nnJN7a8RY/HfcsNsHpcvJa5mvcv+p+5OnSV0orzbZJnYt+gHwBZ3KmSWjp0qVMmTIl4DL4giZXAlLKfcCLQojtUspvGzpOCHGnlLLZHcqllHOBuc09vqWY9CZyKnJaPE7NakJxDAeXJ89/kt3m3Ty59kk+ueYTksKbXoYX24p54pcnyDiRwYTuE5jSewpT/juFU5ZTAZC49VPTU7et3M0Klds5fNF1d6LtGOp1Axlvue6663jkkUfYunUrVquVIUOGkJ2dHVAZfIEnIaINKoDTzGyhLH7FZDD5xCegFI9rHYSoQ3h5zMsAPPLjI1Q5qxo9fnvBdiavmMzmvM3MHTmXZ0c/S5eILgDkW/L9Lm9rR0rJfavu4+E1DwdbFI9QhWpRhekCrgAAwsLCGDNmDNOmTWuzqwDwbY/hVn37YNKbKKkqabEz0Ww1o1PpCNOG+UgyBW9JDk/m7xf83d0BbePz9R4jpWTJniXc+d2dqIWaD676gBt73YgQ7oY4Bo1BMQcBm09tJuNEBj/l/ETGiYxgi9Ns1GE6NFHBK98yZcoUsrKyuOWWW4ImQ0vxpRJo1b0FTAYTLumiuKq4ReOYbWZiDDFtZsnc3hmTPIZ7BtzDsgPL+OLgF3X2WewWHv/5cV7Y9AKjOo3i42s+pp+pX+1+IQQJxgRlJQC8mfUmJr2JTqGdWLhlodI6tJlcf/31SCnp3bs3ACkpKezc2SqbLDbIObMSOLN0REsw25Rs4dbGg2kPcn6H83l2w7PsK9oHwOGSw0z57xS+P/o9M4fM5J+X/JPIkMizzk0wJpzzPoHM/Ew25m1kav+pPDT4IfYU7eHbI01ZfxXaC75UAut8OJbP8VXD+SKrki3c2tCoNLx40YtE6iKZ9eMsPj/wObf89xZKqkpYfNli7hlwDypR/1c9IVRRAou2LyI6JJqbet3E1d2uJjU6lX9u+yfVzupgi6YQAJpTNuKRxh41x0kpH/KvqC3DVw3nzTaz4hRuhZgMJuaNmcfJipPMzZhL75jefHLNJ5zf8fxGz0swJlBoKWyxr6itsrNwJ+ty13FHvzswao2ohIpZQ2eRW5HLp/s/DbZ4CgGgOcliNbkAqcB5wFenX18L/OwPofyBL1YCUkp3BVHFHNQqGRw/mOcvfJ5j5ceY2n8qWpW2yXPijfE4pIMiWxFxRu+z0tsqi7cvJkIXwS2pvzk2R3UaxfkdzufNrDeZ2H0iYTolCKI90+RKQEr5VynlX4FYYIiUcraUcjYwFGgzOdKh2lBC1CEtWgmUVZfhcDkUc1ArpqYBTXMUALhXAhDcMNGfc37mw90fYnfZAzrvvqJ9/HD8B27vc3udC70QgllDZ1FcVcy7u94NqEwKgccTn0Bn4EwjYTWQ4lNp/IgQglhDbIuyhmsUiGIOaj/Eh8YDkGcJfJhoobWQR396lAdXP8g/fv0Hd317FznlLU9obC6Lty8mVBvKrX1uPWtfv9h+XJFyBR/u/pACi/fFGxVaP54ogQ+BTUKIZ07XEtoIfOAfsfyDSd+yhLHakhGKEmg3BGMlIKVk+YHlTPxiImuOreGhtId48cIXOVJ6hJu+volvDn/jdxkOlxzmf0f/x629b603agrg4cEPY3faWZS1yO/ytDVmzZrFwoULa19ffvnl3HPPPbWvZ8+ezfz583E4HMTGxrbq6qKeZAw/B0wFioES3O0g/+4vwfxBjCGmReag2uJxijmo3RCjj0Gj0nCqMjARQsfLjvPH//2RpzOepkdUDz6b8Bn3DrqXq7pdxacTPqV7VHce/+Vx5qydQ6W9sukBvWTxjsXoNXr+0PcPDR7TOaIzN/a6kWUHlpFdmu03Wc7E7rSz/MByjpQeCch83jJq1CgyMtxJdS6Xi8LCQnbt2lW7PyMjg9GjR7Ny5UpSU1P55JNPastytDY8DRE1AmVSyleAHCFEVz/I5Dd8thJQHMPtBpVQEW+I93uYqMPl4N2d73LDVzews3AnT414ineveJdukd1qj0kMS+S9K97j3oH3suLwCiZ/PZld5l2NjOodR8uO8u2Rb5ncazLR+uhGj7130L2EqEN4ddurPpfj92w4uYFJX0/i6YynmbFmBlaH1e9zesvo0aNrlcCuXbvo378/4eHhFBcXU1VVxZ49exg8eDDp6enMnDmTzp07s2HDhiBLXT/NLsB92gQ0DHeU0LuAFvgIGO0f0XyPyWCiuKoYp8uJWqX2+HyzzYxKqIgKifKDdArBIiHUv1nDe8x7mJsxlz1FexibPJY5588hITSh3mM1Kg0PDX6IER1H8MQvT3D7N7czc/BM7uh3R4O5Dp7y9o630aq03NX/riaPjTXEcle/u3g963WyCrIYFDfIJzKcSV5lHi/9+hIrj64kOTyZGYNn8Oq2V3ll6ys8MfyJJs8v+foQ1Sd8u2rSdQol6truDe7v1KkTGo2GY8eOkZGRwciRI8nNzWX9+vVERkYycOBAnE4nq1ev5s0336SkpIT09HRGjhzpUzl9gSffquuBCUAlgJTyBF6Wkg4WsYZYXNJFSVWJV+cX2YqIConySoEotF7ijf5ZCVgdVuZvmc+U/04h35LPyxe/zCtjX2lQAZzJsA7DWDZhGRcnXczLW17m/lX3+6QUem5FLisOrWBSz0m1WfRNcUe/O4jRx7BgywKfmjSqndW8veNtJnwxgZ9zfuahtIdYPnE5fxz4R27tfStL9ixh08lNPpvP19SsBmqUwMiRI2tfjxo1ihUrVjB27FiMRiOTJk1i+fLlOJ2tLx/Fk1Y81VJKKYSQAEKIUD/J5DdqzDiF1kKvnLtmq5Io1h5JMCbwc87PSCl9VhNqW/425qydw/Hy49zQ8wYeGfpIgw7YhogMiWTBmAV8uv9T/vHrP5j01SSeHf0sFyZd6LVc7+x4ByEEU/tPbfY5odpQ7ht0H3/f+Hd+yf2Fi5Iu8nr+GtblruP5Tc9ztOwo4zqP47HzHqvT2OdPQ//EuhPreGrdUyybsKzRXIXG7tj9SY1fYMeOHfTv35/k5GRefvllIiIimDZtGu+++y7r1q0jJSUFALPZzA8//MCll14aFHkbwpOVwCdCiDeBKCHEH4FVwNv+Ecs/tDRr2GwzK07hdki8MR6rw0pZdZnPxpyzdg4Ol4O3x7/NX0f91WMFUIMQgsmpk1l69VJMBhMPrH6AFze92GTp7PrIq8zji4NfcF2P6+gQ2sGjc2/seSPJ4cks3LqwRdnVuRW5zFwzk/tW3QfAoksXsXDswrM6uxk0Bp4d/Sx5ljzmbZ7n9Xz+ZPTo0axYsYKYmBjUajUxMTGUlJSwfv16Bg0axNq1azl27BjZ2dlkZ2fz2muvkZ6eHmyxz8KT6KB5uLuBLcPtF3haSul/b5EPaWnWcJFVyRZuj9SYZ3zlF7A77eSU5zCxx8Qmy1Y0lx7RPfjPVf9hSu8pfLTnI276+iayCrI8GuPdne8ipeTuAXd7PL9WrWXGkBkcKD7AisMrPD6/ylnFoqxFTPxiIutPrmfmkJl8PuFzRic27FJMi0/jrn53sezAMn7OaX3FCQYMGEBhYSEjRoyosy0yMpI1a9ZwySWXEBLyW5nriRMn8tVXX9X2JW4teOIYflFK+Tjwv3q2tQlqVgI1oZ6eoqwE2ic1uQKnLKfoGd2zxeOdrDyJRJIU5tuEer1Gz5/P/zNjksYwd/1c7vj2Dv7Q5w88NPgh9Bp9o+cWWgtZdmAZ13S/hsSwRK/mH99lPO+Z3uNfmf/iiq5XEKJuvI6/lJKDJQdZdWwVXx78ktyKXMZ3Gc//nfd/zV6JPJj2ID/n/MwzGc+wfOJyr1dU/kCtVlNWVnf1+N5779X+fdddd9XZFxMTQ0FB60u888QcdFk92670dmIhRJQQ4jMhxF4hxB4hhN/d5mHaMHfpCC9WAha7BavDqvgE2iG1SsBHuQI1Wb/eXmybYlTiKJZPWM6knpN4f/f73PT1TWzL39boOe/veh+7y849A+5p9LjGqCkul1eZx9K9S+s9xiVdbC/YzoItC7j2i2u54asbeCPzDRKMCSy+bDEvj3nZI1OUTq3j7xf8nWJbMc9tfM5r2RUapsmVgBDifuABoJsQYvsZu8JpWfnoV4DvpJQ3CiF0uHMQ/IoQApPeu4bztW0lFXNQuyPOEIdA+MwcVNPLujl9j70lTBfG0yOfZnzKeOaum8ud397JbX1uY8aQGRg0hjrHFtuK+Xjfx1yRckVtS01vOb/j+YzuNJrF2xdzfc/ridBF4HA52HpqK6uOrWL1sdXkW/LRCA3DOw7njr53cEnnS5odiVQffUx9uHfQvbyW+RrjOo/j8pTLW/QeFOrSHHPQf4BvgeeBM4N2y6WUXtlVhBARwEXAXQBSymrq1iXyGyaDySvHsFI3qP2iVWuJ0cf4LEw0tyIXrUpLnMH/VUlHdBzB5xM/Z8GWBXy05yN+zvmZv43+G0MThtYe8+HuD7E5bEwfON0nc84aOoubvr6JZ9c/S4gmhB+P/0hJVQl6tZ7RiaMZ13kcFyVd5FPTzd0D7ubH4z/y7IZna9+bL6O52gPehu82p4poqZQyW0o5RUp5FLDibiUZJoTo7NWs0A0oAN4VQmwTQrxdX8ipEGK6EGKzEGKzr2xp3mYNF1mVlUB7xpfNZXLKc+gU1ilg+SSh2lD+MuIvvDP+HZzSydTvpvLCphew2C2UVpXyn73/4dIul9I9yjehlKkxqVzd7Wq+zf6W1UdXMzpxNAvGLOCnm39i4diFXNv9Wp/b7rUqLc9d8BwWu4W/rv8rer0es9ncaksxBBopJWazGb2+cd9QfXjiGL4WmA90AvKBLsAeoF9j5zUy7xDgYSnlRiHEK7hXGU+deZCUcjGwGGDYsGE++W+bDCZ2mj3vAaqsBNo38cZ4cityfTJWbkWu3/wBjTG843A+n/A5r2x9hSV7lvDT8Z8YEDuASnsl9w6816dzPTXiKW7s9ko6uQAAIABJREFUdSMDYweiVTevbHdL6R7VnRlDZjBv8zwuT76cfuX9WqWjNVjo9XqSkjw3QXqSLPYsMAJYJaUcLIQYC0zxeEY3OUCOlHLj6defUdfU5DdMBhPFNs9LR9T4BJqqtaLQNkkwJjTpXG0uORU5dRraBxKj1siT5z/JZV0u4+mMp/k2+1vGJI8hNSbV5/OcaXIKFLf3uZ01x9bw91//zucTP6draJsqX9Yq8SQ6yC6lNAMqIYRKSvkDkObNpFLKPOC4EKLmmzkO2O3NWJ5i0ptwSqfHpSPMVjPh2vAmw+IU2iYJxgRKq0pbXLSsorqC0qpSEsMDvxI4k5qyE08Mf4I5588Jqiy+RK1S8+zoZ3FIB0+ve1oxB/kAT5RAiRAiDHdLySWnTTiOFsz98OlxtuNWJgEpS+1t1rDZZibGoOQItFd8lTBWY1IKhjno9xg0Bm7rc5vH2cGtneSIZB4d9ijrT65X+iD7AE+UwETcTuFZwHfAIdx9hr1CSpkppRwmpRwopbxOSlns7Vie4G3WsNJbuH3jq+YyNTkC/gwPVYCbet3EyI4jmbd5HsfLjgdbnDaNJ2UjKqWUTtzx/F/jLiPd5tZiXq8ElOJx7Zp44+k2k5UtazNZmyPg42xhhboIIfjb6L+hERpm/DCDQyWHgi1Sm6XZSkAIca8Q4hSwHdgMbDn93KaoSVrxdCWglIxo3/hqJZBbkUu4NpwIXYQvxFJohA6hHXjp4pcotBYy+evJvL/r/RYVtztX8cQc9CjQT0qZIqXsJqXsKqXs1uRZrYwwbRg6lc4jJWB32SmtKlXMQe0Yo9ZIuDa8xbkCOeU5JIYnKklMAWJ04miWT1zOqMRRzNs8j2nfT+N4uWIe8gRPlMAhwOIvQQKFEMLjrOFim9tdoawE2je+6DAWrByBc5lYQyyvjn2VZ0c/y/7i/Uz6ahKf7Gu9PX1bG54ogSeBDCHEm0KIV2se/hLMn3iaNVxbN0jxCbRr4o3xLSoiJ6UktyJX8QcEASEEE3tMZPnE5QyKG8T/2/D/uG/VfS328ZwLeKIE3gTWABtw+wNqHm0OT1cCtQ3mFSXQrkkwtqx0RKG1kCpnVdBzBM5lOoR24M3L3mTO+XPYlr+NG768ga8Pfa2sChrBk4xhh5TyEb9JEkBiDbHsNjc/N61GYSjmoPZNQmgChdZC7C47WpXnpRBaU47AuYxKqLil9y2M6jSKv6z7C39e+2dWH1vNUyOeUm7k6sETJfCDEGI67vDQ2tY43lYSDSYx+hiKbEW4pAuVaHoxpBSPOzeIN8YjkZitZq8SrAJRQlqh+XSO6My7l/9/9s47PKoybeO/M72lTnqFJIQWSuhVqoqCCIJdV0XXtrvquuq6ll1X3U/XXtZde0HYtQEqKIgCSm9JgIQAIYG0Sc8kk0xv5/vjhEAgQBKSgJD7us51Zua098yceZ73affzEQtzF/Jm1pvM/WYuD454kDhDHAqZAqVMiUKmaLkIR18rZUrUcnWHiABFUcTqsVLrqKXGUUOts2ntqKXWWUuwOpgFaQvOiSY57VECNzSt/3LMZyISI+ivCkbtUeqItszua521qOVq9MoTiE57cB7h2A5jHVICXdxMpgfth1wm57a025gYO5HHNj7Gk5uePP1Bx0EpU6KRa1Ar1GjkGjQKDWq5Go1CI30uV6NWqHF4HZgd5mah31ofaLkgJ0QTgtlp5uv8r3lk5CNc3vvys5pN1mYlIIriecPU1Fww5mhb7v+R/XrS/s5vtOgw1oFWACariXBteA+/1DmIlJAUFs9czN6avTh9Trx+b4vF4/dIr8WWn7l8LpxeZ/Pa6XPi8rqktc9Fg7sBp8+J0+tEq9Bi1BhJDEzEqDUSpg0jVBNKmDas+X2wOhiZIGNf7T6e3vI0j254lG/yv+GJMU+QENhRZv4zQ1s6i00VRXGtIAhXtbZdFMWlnT+srkUzdYSzlj6cvqdsD2XEhYEzLRgzWU09rqBzGEqZkqERHeK87HT0N/Zn0eWL+CLvC97IfIO538zlt4N/y4K0Bajkqm4dS1uygyY1ra9oZZnVRePqUrS3ariHPO7CQJA6CJVM1eEModLG0h5XUA/aDLlMzvX9ruebOd8wJWEKb+16i3nfzmNHxY4W+9XZurbpYls6i/2t6eXToijeduwCPNOlo+siHHEHtbXXsNnRYwlcCBAEQeow1oFaAY/fQ6W9skcJ9KDdiNBF8NKkl/jP9P/g8XtY8MMCHt/4OHXOOn7YW8HEF9aRVdx1/JrtCQwvQeoGdiy+Arq/s8QZIkAZgFKmbFOtgF/0S+6gntSyCwIdrRWosFbgF/097qAedBgTYiew7MplvLvnXT7O+Zi1RT9Tb7qU1Igp9I/uOi6qtsQE+iG1kAw6Li4QCLS/oeU5gGbqiDa4gxpcDXhFb0+NwAWCCF0Eu6t3t/u4EqvEV9NjCfTgTKBVaLl/2P2MDJvGPaseQxH5JXpjPtWOXsQr47vkmm2xBPoi+f6Dadk/oBH4bVcMqjtg1LStariZMqLHHXRBIFIfSVVRFaIotisb7Eih2IVCGWG2udlf0cC45LCzPZTzDlaXl6eXmvHX3cM9s+v4+vDHXdrH+bRKQBTFb4BvBEEYK4rili4bSTcjTBvWpiyQngbzFxYidZF4/B7qXHXtsv5MjSYUMkVzX4LzGU6Pj1s/2s7esgZ2/fViAjTd02j+QoDX5+f3/83kYJWVj28bycQ+4dyZfg0KWXs89+1De7iD5gqCECgIglIQhDWCINQIgnBTl42si9FWd1APZcSFhRa1Au1AqbWUGH1Mh6pLf00QRZEnv85hT6kFn18kt6zhbA/pvIEoijy9IpefD1TzzJVpTOwjFat0pQKA9imBS0RRbEByDZUCqcDDZ3JxQRDkgiBkCYKw4kzO0xEccQf5Rf8p9+shj7uwcGQm395aAVPjhUEhvWhbMV9mlHL9KKmwaW+PEug0fLSpkIVbirjzoiRuGN19hWPtUQJHbL7Lgf91EmfQ/cC+TjhPu3GEOsLispxyv1pHLTJBRpDq7HN89KDrcSx1RHtwIRSKZRSZeXr5Xqb2i+Afc9IIM6jPuhKoanSSX9V4VsfQGfgpt5Jnvsvl0oGRPDqjX7deuz1KYLkgCPuBEcAaQRDCAWdHLywIQhwwE3i/o+c4E7S14bzZaSZEHXLem/k9kBCmDUMuyNulBGweG3WuuvPaEqhscHL3okxig7W8eu1QZDKBtNhA9padehLVlcivamTWGxuZ/sp6rn1nC6tyyvH6Tm3Zn4vIMVm477MsBsUG8dq16chk3UtP055G848CY4ERoih6kLqMXXkG134NeAQ46a8mCMKdgiDsFARhZ3V19Rlc6kS0teF8rbOnwfyFBLlMjlFrbFdM4Ahx3PlqCbi9fu5dnInN5eWdm0cQpJWcAgNjAjlYZcXp6f6+vrllDVz7zlZE4I/TUymtc3D3okwmvfgz7/xSQL29a6tsOwvlFge3f7KDYK2S938zAq2q+yebp1UCgiA8cszb6aIo+gBEUbQB93XkooIgzAKqRFE8ZVMaURTfFUVxhCiKI8LDO8DodQocSyJ3Kpgd5p6g8AWGKF1Uu2ICzRTS52l66NMr9pJRVMcL8wfTNyqg+fOBMUH4/CJ5ld3rjtlTWs/1721FpZDxxV1juX96H9Y/MoW3bxpOfKiW51buZ8xza3hsWXa3j609sLm83P7xTmwuHx/eNpKIwLNTdtUWS+C6Y17/5bhtMzp43fHAbEEQCoHPgKmCICzq4Lk6hCPuoNNRR/RYAhceInQR7XIHmRpPbCaTY7JQXPurb8nNFztLWLS1mLsuSmLW4JgW29JipDhZd8YFMorM3PjeNgK1Cr64ayy9wyR6d7lMYEZaFJ/dOZbv75vIlUNiWZJRyiWvruem97fxU24lPv+5013M5xe5739ZHKhs5F83pNMvqusqgk+HtigB4SSvW3vfJoii+BdRFONEUeyFpGTWiqLYremmgarANlFHmJ09lsD5Co/Pj83lPeHzSH37qCNMVhMGpaG5Qcj6vGrm/nsTF7/6C+9vOHROCZ/2YE9pPU98ncP4FCMPX9r3hO3xoVoCNApyTN0TF9hSUMvNH2wnLEDNF3eNJT5U1+p+A2IC+ef8wWz5yzQevrQv+VVW7li4kykv/cxHmw7j8na/++p4PPtdLmv2V/HU7IFM7nt2a0vakoAqnuR1a+9/NRAEgVBN6CndQXaPHYfX0VMtfJ6g3OJgV3E9WSX1ZBXXkW2Sct1fvXZoi1lupC4Sm8eG1W3FoDKc9rylVok9VBAEMorM3PVpBikRAcQGa3j2u32syqnghfmDSQo//bnOFdRYXdz9aQbhBjVvXj8MhfzE+aIgCAyIDuwWS2B9XjW/XbiThFAdi+8Y3SbXSahexe+mpHDnRUms3lvJR5sO8/fluXyyuZAnZw1gar+ILu8R4vH5Ka1zUFRro9hsp7DGTn61lfV51dw+oTc3j0ns0uu3BW1RAkMEQWhAmvVrm17T9P6MnViiKP4M/Hym5+kITtdwvqda+NcLh9tHtsnCrpI6sorrySqup6JBSmZTyWUMjA3khlGJ7Cmt577/ZeHy+Jk3XPLpH1sr0BYlYGo0kRiYyL7yBm77aAdRQRoWLhhFmEHFsiwTT327l8te38AjM/px27he3Z790V4cqVqttblZcs84QvUn57cfGBPEf7cX4fOLyLvovn7KreTexZmkRBj49PZRGA3ta9qjlMuYOTiamYOjWZ9Xzd+X7+X2T3YyuW84T84aQHInKOfSOjs5pgaKzTYKa+0U19opMtsw1Tk41hDUKuUkGnXcPqE3j13e/4yv2xloC23EeZsbGaYNo9p+8qyjI1ZCjzvo1wGzzc3CLYX8tK+SfeWNzW6YhFAdo3qHkp4QTHpCCP2jA1ArpMfa7vZy58IM/vTlbpxeHzeOTmxRK5AUfOruqaIoYrKaSAsdxc0fbEevVvDp7aMID5AE1VXD4hifEsZjS7N5ZkUuPzRZBb3Czt1Wpc+v3M/WQ2ZevnoIabGnro9Jiw3E6fFzqNpKn8iAU+7bEXy3p5z7P8tiYEwgCxeMJkh3ZhQVF6WGs+qBi/hkcyGv/3SQGa+t57bxvfnD1JR20184PT5+2FvB5ztK2FxwdDIZolOSYNSTHh/C3KGxJBj1JBp1JIbqCA9Qn3MdCru2HvkcwNLMUsw2NzPSoogLaelDNGqM7K/df9Jjm8njeiyBcxolZjvvbTjEFztLcHr8jO4dyt2TkkiPD2FoQjBhp5g56lQK3r9lBPcuzuTxZTk4PX4uGdz2grFaZy1On5MfdrnxiyKf3j7mhOcsMlDD+7eMYEmmib8v38uM19fz6Ix+/GbsuWcVfLPLxPsbD3PruF7NltGpMLApOJxTZul0JbAsq5Q/fbGb4YkhfHjryE7jKFLKZdwxMYkrh8by0g8HeG/DIZZmmnj0sn5clR572t8kt6yBz3cU8/WuMiwOD/GhWv50cSqT+oaTaNQ3p9D+WnDeK4F1B6pZvruMZ7/bx6DYIGakRXFZWhRJ4QaMWiNmpxm/6EcmnOjzbHYH9cQEzknkmCy8s/4Q3+0pQy4TmDM0lrsmJZES0T5hpFHKefum4dz/WRbPrMjF6pbaabclTTS3uhAAhz2I/902ipSI1l0LgiAwf3gcE1LCeHTpHp5ansvKnApenD+EBGPrAc7uRo7Jwp+X7GFUr1Aen9k2V0VyuB61QsZeUwNz0ztvLJ9tL+Yvy7IZm2Tk/VtGoFN1vqgKD1Dzz/mDuWF0Ak8t38tDX+7m061F/H32QIbGB7fYt8Hp4dtdZXy+o4RskwWVQsaMgVFcOzKesUnGc06ZtwfnvRJ48/p0/nRxKqv2Vkh/uh8O8OIPB0iNNBDfy4dX9GJxWgjRhpxw7IXuDiq3OPD6xJNmYZwNiKLI5oJa3v6lgA0HazCoFdwxMYkF43sTFdTxEJVKIePN69N56MvdvLr6MGEDAqiwVZzymEanh799tx508MysSQyKOz21SFSQho9uHcmXO0t5ZkUul762nr9c3o+bRid2uyCptbrYesjMlkM1bCmopaDaRmSgmn/dmI6ylUBwa1DIZfSLCujU4PCnW4t48uscJvcN5+2bhqNRdq1Hekh8MEvuHseyLBPPr9rPnLc2cfXwOB6e0ZfD1TY+31nC99nlOD1++kUF8NQVA5iTHkuwrnt7AXcVznslANArTM/dk5K5e1IyZfUOfmhSCOv3OdHEwhX/WcnM/ulclhbNkLigZp+d2WkmQBXQ7Y2fzzaKa+28ufYgS7NM+PwiwxNDmJsey6zB0Wftwff6/KzaW8E7vxwi22QhzKDmkRl9uXF0YqeZ3wq5jJevGYpGKefbGgObDhcgjmm9r4DT4+OOT3ZSbjOh1MFl/Qa0+TqCIHDNyHgm9Anj0aXZ/PWbvXy+o4TJfcMZ3dvIsMQQDOrO/2vW291sPWRm66FathTUcqCpkEqvkjOydyjXjoxn5uAYIgLap0wHxgaxYndZu3swtAa728szy3OZlBrOOzcPb47ddDVkMoF5w+O4NC2KN9ce5MONh1mSWYpfBINawVXD4rhuZDyDYoPOOZ/+mUIQxV9HlueIESPEnTt3duo5fzy0iQc33E2y7yFy8sPx+kVC9SpC9SoMagW1uvdxyUqZrH8Jg1pJgEZBgEaBQa0gQKMkWKdkZK9QVIr2UDCduygx2/nX2nyWZJYikwncODqBiAANy7JKyau0opLLmNovgrnDYpnSN6Jb7tvt9fNlRgnv/HKIYrOd3mF67rwoibnpsV02Q/T7RS797BZMjZXMi3yZv88e2GKW7vH5ufvTDNYeqOKicWspdmSy7pp1HbqWKIp8ubOUxduLyWlKWZXLBNJiAhmdZGRUr1BG9gptV0BUFEVqbW4qLE5K6+zsLKxjy6FacssbEEXQKGWM7BXKmCQjY5ONDIoNavPMvzUs3lbE48ty2PDIlDO2Gn/Jq+aWD7ezcMEoLkrtXJaA9uBQtZVFW4sZEBPI5YOiusQd1V0QBCFDFMURJ9v+672zTkByaDQAd04OZ/x1F/PTvkp2FJppcHpodHoxiRZ8Hj3r82qwurxYWyks6h2m5/HL+zOtf9fnHHcVSuvsvLUuny93SsL/pjGJ3DM5mcimXOy7JyWxt6yBpZkmvt1tYtXeCoJ1Sq4YHMPcYbGkxwd3+r37/CLf7DLx6k95lJgdDIkP5rHL+3HxgKguS0U8AplMYEKvZJbnF/Dp1iKcHh/PzxuMXCbg94s89OVu1uyv4pk5afxsWUqcvON0EUesgmtGxmNzecksrmP7YTPbDpn5eFMh764/hCBA38gAxiQZGdU7lLSYICwOD2UWBxUWJ+UWJ+UWB+UWJxVNi/sYIjWVQsawhGAemJbK2GQjQ+KDOnWGPbC5cthyxkpgc0ENSrnAiF4nume7E0nhBv56Rdutu18zLmglcCyJXJBOybzhcS0yImZ/7SUlOIVX7pgGSILJ5vZidXppdHo5XGPlpdV53LFwJxP7hPHkrAGknkGGhCiKbD9s5r0NhzHVO5g5KIq5w+KIDdae2Y2eBKZ6R5PwL0FA4IbRCdw7OeUE37ogCKTFBpEWG8Rjl/djQ34NSzNNfLGzhE+3FtE7TM/c9FjmDI094yCnKIr8mFvJy6vzOFDZyIDoQD66LY3JqeHdqmQj9ZG4xAb+MK0Xb64pxOn188o1Q3h6eS7f7Crj4Uv7cvOYRBZ9VUp6ZOdERPVqBRP7hDc3E3F6fOwuqZeUwmEzn+8o4ePNhSccp5QLRAVpiA7UMjQ+mOhBGqIDNUQHa4kO0pAaGdClfvV+UQHIZQJ7yxqYkRZ9RufaUlBLenzIr3rm/WvDBf1NB6oCUcgUJ60arnXUMipqVPN7uUwgUKMksClVrW9UANP6R7JoaxGv/XSQy17fwI2jE/jj9FRCTlFgczx8fpFVORW8u76A3aUWQvUqksL0vLQ6j5d/zGNskpH5w+OYkdY5Zmm5xcG/1xXw2Y5iAK4dGc+9k1OIaYOyUchlTOkbwZS+ETQ6PazMrmBpVimv/JjHKz/mkRYbyIyBUcxIiz5ppszJsLmghhd/OEBWcT29w/S8eX06MwdFn5XMiyO1AteOCUKv6sfzK/eTY7JwuMbGnRclce/kZDx+DxX2ii6jkNYo5YxOMjI6ycgfkNxQOSYL+ysaCdWriAnSEhWkwahXndXsFI1STkq44YzpIyx2D9kmC/dN7dNJIzsPIIpQvR8iuq6w7IJWAs3UEa1UDXt8HhrcDaetEVDKZdw2vjdzhsby6k95LN5WzDe7ynhgeh9uGpN4Sl+r3e3ly52lvL/xECVmB72MOp6dk8a8YXFoVXJKzHaWZppYklnKg1/s5smvc7hsUDTzhsUxundom//4dTY3+dVWDlZa2VVSx9dZZYiIXD0int9NSemwpRGgUTa7Mkz1DlbsLuOHvRW8tDqPl1bnkRJh4LK0KC4dGMXAmMCTzuT3lNbz4g8H2HCwhuggDc9fNYj5w+NapSroLhxRAlX2Ku6eNAytUs7fvt3LdSPj+ctl/RAEgQprBX7R323soUq5jPSEENITzq6rpDUMjAlkY/6pyRhPh62HaxFFGJ/S07y+GbsWw7d/gFtWQK/xXXKJC1oJgFQ13GwJ5H4LOz+EK17DrJIEY1trBEL0Kp6+Mo2bxiTyzIpc/r48l8XbinliZv8TCKKqG10s3FLIp1uLqLd7SE8I5vHL+5/g744P1XH/9D7cNy2FHYV1LMko5bvscr7KKCUuRMtV6bFcNSyOXmH65mDgwUor+VWNHKyShP7BKis1VlfzOXUqOfOGx/K7KSknFDWdCWKDtdw1KZm7JiVTYXGyOreCldkVvLUunzfX5hMfqm2yEKJIjw9BJhPIr2rkpR/yWLW3ghCdkidm9uemMYldnhLYFhyhjjhSMHbLuF5cOjCKyMCjFZ/nex+B9mBATCBLs0xUN7qaq6Xbiy0FtWiUshNy9C9Y1OTD949A4nhIGNNll7nglYBRY5TopHOWwJLfguiDj2ZinvNa8/b2IDUygIULRrFmXxXPfpfLrR/tYErfcJ6YJQWZ3t8gpZ55fH6m94/krouSGNHr1HUIgiAwqncoo3qH8tTsgazOreCrjFLeXJfPG2vzSY00UN3oos7uaT4mQK0gJdLA1H7h9IkIICXSQJ8IAzFB2i53HUQFafjN2F78ZmwvzDY3P+VWsjKnnI83F/LehsNEBKgZGBPIL3nV6FQKHpjeh9sn9O60itDOQKT+xIbzx8dKTNYTKaQvVByhl9hbZukwK+bmgprzKtvujOB1w5LbQaGCq96FLuxs2KMEtEYOVO2CJXdAwliY9lf43/XUrrgPghQdoowQBIHpAyK5KDWcTzYX8saag1zy6nr8oohSLmPesDjumNi7Q8RVWpWcK4fGcuXQWMotDpZlmdhSUMvwxFD6RBjoE2mgT0RAixnr2USoXtXsMmpweli3v4pVORVkFdezYHxv7p2SckqCsrMFg9KATqE7JXWEyWpCISiaXUcXMgbESHz4e8saOqQEqhqd5FVamZveY1UBsPYZKN8F1y6GwJjT738G6FECDVWY3Q2IieMQbvgCVHq4dQW1n88FFITa6zt8bpVCxm8vSmLusFjeW38IjVLOTWMSO2wuH4/oIC33Tk7h3skpnXK+rkagRtmswM51CIJw2uYypY2lRBuie/pPI/22CaG6Dvcc3tJEwDYuuYeihYJ1sPkNGLEA+s/q8std2HZX1mKM+77DKwg0zHtfUgAAkQMxj7odAONXd0Bl7hldJsyg5i+X9+ePF6d2mgLoQdcjUh95Sv4gk9V03raU7AikxvMdo4/YUlBLgEZxWtbS8x62Glh2F4T3g0v+0S2XvHCVQOan8M3vCDNKHZNqfY4Wm2vlcjQyFTpBAZ/MgorsszHKHpxFROpO3WHMZDURG3DuWzXdhYExQRTV2mlwek6/83HYXFDLmCRjlxcCntMQRfjmd+Coh3kfgKp7OLsuTCWw8yP49veQMg3jtKcATkgTrXXWEqoNQ7j1O1Bo4JMroGzXWRhsD84WInWR1Nhr8PlPbEdo99gxO809QeFjcCQukNtOa6DEbKfYbO9xBW1/D/JWwcVPQ1Rat132wlMCO96HFQ9An0vg2sUYDVLQ5fiG82anWQoKG5Ph1u9AZYCFs8GUcTZG3YOzgAhdBF7R29xX4liUWnvSQ4/HwGOCw+3B0XjABVwfULkXVj8hyaXRd3Xrpc+KEhAEIV4QhHWCIOwTBGGvIAj3d8uFt78H3/0JUi+DaxeBUtOcAnp81XCto/YohXRob7jte9AEw8I5ULKjW4bbgyYc/Akqcrr9ssd2GDsezTUCPTGBZkQEaIgIULc7OLy5oIYwg4rUyF9PD+ZOhccBX90OmiC48t/QzVl9Z8sS8AJ/EkWxPzAG+J0gCF3L1rT1P/D9Q9BvFlyzEBRSgDZQHYhCUJzgDmq2BI4gOEFSBPow+HQuFG3p0uH2AClXevkDsHgevD1eUsD5ayTfaTeguVagFSXQUyPQOgbGBLLX1HZL4Eh/iLHJYedESvNZweonoHofzH0bDN3PnHpWlIAoiuWiKGY2vW4E9gFd92/a/C9Y9Sj0vwKu/lgqwGiCTJARqg1tYQn4Rb+kBI4vFAuKk1xDAZGwaB4UbuyyIZ8TqNoH5sNn59qNFVJAPuMjGH8/TPubNJ5FV8F/xsOu/0pKogvRXDVsa10J6JV6gtU91a3HYmBMEPnVVpyeE+MoraGg2kZVo6tFPODXQm/fKdj/veSiHivFKM8GznpMQBCEXkA6sK2VbXcKgrBTEISd1dUnbwh/Smx6HVY/DgPmwPyPQH5iVapRY2xhCVhcFnyir/WOYoExcOv3kkJYNB92fAC+Eymmf9UozYDFV8O/x8C/RsKGl7v3Hku2wzuTpIzNDv/gAAAgAElEQVSsqz+WAmUTH4QH9sCc/wAifH0PvD4YNrwCjroOXcbv81F5KB+HtbHV7aGaUBQyRatpoqWNpcQaYi/c2etJMDAmEJ9f5EBF69/p8dhSUAOiSJrOQcZ337Dkub/x5q3XsOK1f+JxObt4tGcZDWVSNlDUYKlI9SzhrBaLCYJgAJYAD4iieIINKYriu8C7IDWV6dBFnBZImwdz3wV567dr1BpbBIZP22A+IFKyCL68Fb57ELa/K+X09pneoSGeMyjZDj8/DwVrQBsKU5+EyhxY8zQcWAlz3oawLi5My/gYvnsIgmLh5qUQOfDoNoUaht4AQ66Xxrj5X7Dm77D+JRh2M4y5B0J6tekyHreLb1/+Pwp3SYF+gzGM8IRehCf0IiyxN+EJvQiJjiVC23rBmMlqIj4gvhNuuAPwOEHZ8VaaXYmj9BENDDkFB5DTaqUoexcHl6zm9oqDrH5GUhoh0bEkpY/gwNaNWKormfPwk+iDzz3CvDOG3y/VA3idMP/DZvf02cBZUwKCICiRFMBiURSXdtmFpj4Jov+U3BtGjZGDdQeb3x9xDZ2SN8gQDreugP0r4Me/Sn7r5KlwybMtBVdnwmmB2nyoLWhaNy22GogbCSnTJZOyvWXmxVsl4X9oHeiMMP0pGHkHqJt6I/SbJQXU354gzcpH3gGyTjYivS5Y+WfJ/ZM8Dea9D7qTcCoJQtO9TpeshS1vSSb19neh/2yY8ADEnJzj3+108PULz1CSm82E636DTC6nuriQmqLDFO3Zhb/J6pErFEwM1OAKyWNn5n2EV/5EdEoflOPuwtRYytiYsZ37HZwOLiv88BfIWgTDbpGebX3rz6jP6yF3/Tp2//g9GkMAMan9iEntT3Sfvqh1+i4bYlyIlkCNgpzjgsN+n4/y/DwKd2dStDuTioKDiKIfrUyFP7oPF18+hcTB6QRFSHGYvjsu4vs3XuK/TzzEVY/+DWNcQpeN+axg8+tweD3MfhPCzi519llpLylINvQngFkUxQfackxXtJc8glczXuXT3E/JuCkDQRBYeXglj6x/hGWzl5ES0oaZr9ctCaFf/gmuBki/GaY8LlkMHUF9CZTvhtqDLYW+7RiXmCCTgtXGFClrqWgTNJZL2yIGSMogZbrEh3SyWUbhJvjleelh1IfDuPtg5O1HK6ePRUO5RGmb/yP0ngRXvgXBnTQTbqyAz2+G0u0w4Y+ScGsvFYPFBNvfgZ0fg8sCafMlEzskscVuLruNpc89RXn+AS773YP0nzC5xXaf14O5zERN0WGqiwvZkPEtQo0DtUuaLwmIhGkb+SnWz+SkFH4z/wUUQR0jTGsXTBkSwaH5EPS5WAqQqwNg6hMw/LZmK9fjdpGz7kd2fLOExtpqwhN7gyBQU1SIKPpBEAiLT2xWCjF9+xMcGd2pbq3r3t2Cw+3l3cujKcnNpiQ3G9O+vbjsNgRBRlRyHxKHDMMfk8pvllfwynXprXIGVRQcZNk//47P42H2nx4jIW1Ip42x0yCKkrVcsFaajAmy1hfZMa+9blj/AvSbCVd/ctJsIHuDheKc3RTt2cWUW+5Ape1Y8djp2kueLSUwAdgAZANH+uA9Jori9yc7piuVwKe5n/LCjhfYeN1GgtRBLN63mOe3P8/6a9cTommHKWo3w/oXpRmpQiPNSMf+HpSn4OsXRagvkgRy4UYo2gj1xUe3GyIlQW9Mblo3LSG9Wgp3UYSqXMj/SRIQxVvA5walDnpNPGolhCZB4Qb45QVprY+QAq8jFpy+QlEUIfMT+OFx6WG+7J+Sa+ZMBEjJdkkBuBpgzr9h4NyOnwvA2SDFgba8JTHCjr4LJv4JtCE4GhtY8n9/pbqokJn3P0zq6FPws1urYfUTvFi6ii8CA/llyLNUyRMozdnF/u0/UVdRj0wUUAh+YiL1JIyYTMKY6UQmpSCTdyKXkN8HG1+Fn58DQ5SUQdJ7ohQkX/mIpMAj03BPfYY9efXsXLEMW30dMX0HMOaqa+k1ZBiCIOB22CnPz6Msbx9lefspz9uPy24DQBsYJCmE1H6EJ/YmINSIwRiGWqdvs3Lw+31UFx6mJDebtes24y0rQO2XAvfBUdHEDxhE4uB0EgYNRWuQLMz31h/iH9/vY+tfpp3A0HoEDdVVLH3+KerKTVx85x9Im3wOuFwddRK/T/4a6f9mrZA+V2glr8OxCyeRr2GpcPtq0B6VLz6vh7ID+yjck0XRnl1UHs4HUUSt0zP/8WeISknt0HDPSSXQEXSlEvj+0Pf8ecOf+WbONyQFJfFG5ht8mPMhmTdnIhM64PaoLZBcRPtXQGCcNCMddLU0GxBFqDssCfzCTdIM3lIiHacNhcRxktCOHwnGPqAJ7NhNuazSNfJ/kvzn5kPS57owsNdIAmXCA5Jbob3l6ebDUkCraBP0nQlXvEZJUQWZK5cTlZJKXP80opJTkCtOQw298yP4/mEpyH7d4s51o1lMsO4fUhaRNhj7iPv5cmU+dRVlzH7wMZKGjWz9OL8fMj+Gn54Ct52Fg2fwomV38wQBpOflibWP8mbk3Xh2bKC4qJJqp/QdqtRK4gYOJSFtKHH9BxKe2LvjSqG+GJbeBcWbYeBVMOuVFkIDUcSV9RVZ/32NjDIDTp+ShP79GXPNb4jrn3ZKAS76/dSaSiSlcGA/ZXn7qCs3tdhHqVJhCDQQEKjDoFcToFdg0MgwaEQClF5EbSil6kEtZvoAipBwdnvCuHHuVEaNHUmAsfUisNs+2k6R2c7aP00+5dfgstv49pXnKM7exZirrmXcNTd1WUDeZbdRkX+Q8oP7qSkpImHQEPqPm4TSnHtU6JfukAS8JkhyAadcLK0DW2mtKYonKgbRDwoNoiDDbCqlaE8mhXuyKM3NweNyIshkxKT2I3FQOomD04lK7nNGE4ueRvNtQHOvYUctSUFJmJ1mQjQhHVMAIM3ar1ssCfkfHoNld8K2/0hCvXAjNJZJ++nCoNd4xLF/wBs7Brc+DrfLidvhABcEegU0otixB15tgL4zpAUkxVSwVrIQ4kfDsN+c2kI5FUJ7S52Otv4b1jyN760x/Fg0lsYGB/k7pPoJhUpNTGpf4voNJC65F1ExRpQ+m8SL4qyH/LWwa5Hk/5//QUvh1hkIipUsizH30Pjt43z18XIavBrm3jCTxPST/B8qsmHFH6U/ea+JMPNlImyH4ZfdVNorm5VAqbUUj1Jk2GW3ob3iXrDVYt/0HiW/LKG4xk9Jrp1DmVJBoUKtJjqlbwufvDagDYo9+ytY8aAkMOa+A4OvbWFx2RssZK38lqxVK3DZQ0hKDGG0YhMxygyoiQRvn1MGjwWZjLDIcMJ8pQxWixDSgP1QHnUWO1avmkaPGqtXRaNLjbVcRYlHjc2rwn9CQuF2gqOiSR0znvgBg4gbMIhyr5pXX13PrJhBJ1UAHp+f7YfNXDXs9MV2ap2eqx59ip/e/zdbl35OfWUFl959PwqfXXKT+dwQM6zd7le/34e5tISygwcoP3iA8oP7qTWVNNeh6PUaDmzZwIYPXmNwcBlDQyoISBwIFz0sWdYxw06abNIMQQBBDkhC3ONyUrhrNwWZ2ynak4XVLMUfQ6JjGDh5GomD0okfOBi1rnt4g6BHCQBSAFjwQ2nBfkJKfDSUlWNUnbrRS1vgTxhD3eULqVy/mKrtq7C6D+BW9cctH4VbVOGq8eE+4MDt+B6/b3mr51Dr9ARGRBIUHklQRCRBkVHSOjyKwIgIlKo2ZhUYk6Vl1G/P+L4AyaoZ93tImU7Ov35PXa2FOcMgOkSgtNJGaU0jpQX1bM7ZDQjIBT9RmkbidBbidBZidA2oLuqg/78daJBH8kV2BA5BwbxB1cRlPAaVX0sB/ITR0k6uRlj3HGx7W1JGxwjdKFEiFqyyV5EaIpnjJqsJo8aIVtGkRPVGdJc8St9pD9H3wHew7R0a8rdRZg+iTOhNWWU+2/dlI/ol4RIaE0d0k1KI7duf0Jg4hCOBdqcFvnsI7+4vcUeNwjX1WVzKEFw5u3Hb7bjsNqqLC8le8wMet4vUUeMYNfcaInsnQ12RlA699hnI+hQufQ76XiYJIlEESymUbJOUXMk2Sen5m1J/jSnoBk5HZ0wGdaBEk6IOkCYTqgBQByAqddidfhobHTTWVuP//hFiQgQC7v+mxW+o9flRK2TsLWvgqmGt/y57SuuxuX1t5guSC3DJ7IsI9payccMvNGb/yOzIDHSKY1KXA+MgNh1ih0tL9NBmS1oURRpra6guOkT5wTzKD+6noiBPmnABGq2a6DANfftAtO8QUfIK1DIfJn88mfZ+7DAlsrO+N30ixjMsdjYx8f3aNG6QEhEOZ+0kb+smDmXtwOtyodEbSEgbQuKQdBIHHQ2Inw1ckO4gURSxVFZQXpBHZUEeJXn7KCvYj8J/dJbjl0NkfBJh8YmExSdKqYMJvTCEGludmfv9PurKTFQeypeWw/lUHT7UnOusUKoICAtHpdWi0upQaXWotVpUOt0x73XN7/0+Lw1VlViqK7FUVmCpqqShugqvp2WBlD4klNDoWPpNmEz/8ZNQaro3ddDjdPLB/b8lWOPj2tTDCJoAKVCtDQZNME6ZAZPZT2mFjdLSWirLqxH9IjK5nH7jLmL4rLlE9ErqkrHVVZTx5dOP43bamffY00QnpUg9W9f+Q/Lj9p8tcbWs+z/JOht+m+S6OyYrqcxaxqVLLuWpsU8xL3UeAHf8cAdOn5NFly86+cUrc+HA95L1VbINj9dPhSecMvVAylyhlFVacdok94larycoIgpPYx0uSzUur4BPPLkVKggy+k2YxOg5V7eeNVOwFlY+CjUHIGmKJAhLth9NHFDqpFls/CjJKowbedIso1Ni79fw5S1SkkD6TS02zXlrExqljM/ubD2D6s01B3n5xzyynryYkNaaCllMkrIy7ZTqVsqywCsJ7P3O3qwqiiMgQMtVt19PSHQMmDLBlIFYmoGtupRat55al54aIZoaTxC1jX7cbklhCDKBcKOeaIObGKGEaKGYYKUTQSaXXJKxwyQlEjNMSrKQybBUVZC1agXZa1fjdtiJTunLsMtn02f0eOSKE+fSLrudQ5nbydu6icJdGXg9bnRBwfQZNY7UMeOJ65/WubGjU6AnJgDY6uuoKDhIRUGetM7Pw9lUIKRQqohISmaNewdpaWO5ZtQtPL78jyR5IhkgJlJTXNhssoH0hw2L70VYfCIh0bFYqiuoPFRAdeExAl+tJiIxicikFCKTUojonYwxNv6Mf3TR78dmqZeUQnUllipJOVTk51FbWoxap2fApKkMufhyjLHdk8O+dennbPr8U657+kVi+/Y/7f5up4OyA/s4lLmDnHU/4nE5SRg0lBGz5jYHMTsDtaXFfPnsE/i9XuY/8WxLReO2STUGm14Hjw0i02DWq5JQPA4en4fhi4Zzz5B7uGfoPQDMWDKDIeFD+OdF/2zbYJwNUhC+YK3kV647LIWGNEmUadMpc4Vira5AZTkoTQz6X4I6Mhm1Todap0el0ze/Vut0aAMCT58p4vNICQq/vCDN7I8I/PiR0v22UjTZbogivD9dUi5/yGjhXnx8WTbf7i5jz98uafU3vf7drVgcHr6/f2LLDW67VCl+hKhRroLoIRA7AuKaluBETHn7+ebFZxBFkZGz59FQXUVNSRG1pcXN/20AjQrCVHbClPUY1XbC1DYiNVaUMj+EJjdZDU1CP2rQaV2kboedvb+sIWvVcurKyzCEGhl6yUwGT5+BIJNRsHMbeds2UbQ7E5/XiyEklD6jx5M6ejwx/fojOwsNiC54JbD8tX+St2UDIM2gwuITiEpJJSo5laiUVIxxCcgVCqZ9MY0JcRN4auxTjFo8imv6XsPDIx8GwGFtpLa4iJqSImpKCqkuLqK2pAiX3YZSrSG8VxKRSclE9paEfmhsXLf+2KIoYjqQy+7V35O3dRN+n5eEtMEMufhykkeMaXWm0hmwN1j44L47SEgbwpUPPdHu451WK3vWrCJr5bdY68wY4xIYMWsu/SZMRqHsuJCqKjzEV88+gUwu5+on/3HyHPPGSsklkjT5lL7dyZ9PZnL8ZJ4a9xRev5cRi0Zw+6Db+UP6Hzo2QPMhSRkUrJWye9xW6fOhN8Flzx+tz+gMiGLXEpIVboKPL5dqSyb8sfnj/24r5rFl2ax/eAoJxpYKy+nxMfjvq7llbCKPzzyOMmzVY7D1Lcki6z1ZolQ+SYpzfUV5c+aQWqfHGJdAWHwixvimdVwCuqBgBJCSL0yZ0jpyoFRHcgZxKNHv5/CuDDJXfkvRniwUShV+vx+/z0uAMZzUMePoM3oCMX36HnX1nSVc8IHhlBGjiU7uQ1RKKpG9U07qLjlSNezwOnD6nC2qhbWGAOIGpBE34CjHtyiKOBosaAICzop2PxaCIEgB2H4DmXLLb8leu5o9a1ax/NXn0YeEMnjapQyadikBoZ1L1btt2Rd4nC4mXHdLh47XGAyMunI+w2deyf5N68lYsYwf3n6djZ8tJH3GFQy++LLmdMJTQRRFHI0NWKoqqC0t4eeF76HS6Lj6yWcJiT4FJVVAZJuCiZH6SCrsUhpgha0Cn+g7M/bQ0CQYlSTFZ3weyVUjyCCxC4rPuprWotd4SJ0BG16VMs2aXGlHaaUtJyiBjKI63F7/idTRJdulZIMRt0tpvadBcFQ0t7z0Fo4GC/qQ0FNbkcEJ0tJJEGQykoaNJGnYSGpKitizZhVyhZLUMeOJSk79VdGJnPdKIGn4BBwNboIjT20+HyGRa1O1MJLg1QWde+RhuqBgRs+9hpFXzuNwVga7V3/HliWfsXXp56SMGMOQSy4nIW3IGT+klqoKdv3wHWlTpmOMOzPXk1yhZOCkaQy4aCpF2bvIWLGMjZ8tZOuyz0mbPJ3hl8/BYAyjobpKcoFVVlDftJZcYhXNAT6QhMP8x5/ttGBbpC6yuX/AkXWnsYfKlZIg/ZXB7fRib3ATHKGTrID/jJM4pi6VWiL2jQpALhPYW9bAZYNapk5uLqhBLhMY2fuY5AuPU0o7DoqHi//e5nHIFQoMoWe3GU1YfCJTb+3eHgCdifNeCfzwbjb1VQ7m/3kEusBWAlBNMGqMFNQXNBPJtUoe9yuCTCYnefgokoePor6ygj0/rSR73Y8c3L6Z9BlXMPW2M3toN32xGJlMxtirb+ikEUuKtdfgdHoNTqe6uJCMFV+z56cf2LW6qYbwGNelQqlqzpSKG5BGcERU0/soQqJjTl+jAHjdPmwWF0Hhp54gROgiyKiUfNSmRimX/kJtJuP3i+zfUs62bw7hsHq44g9DiO/fX+J02v4ujLoTQhLRKOX0iTCcQB8BUivJIXFBGNTHiJ+fn4OaPLh5Wee6w3pwWpz3SmDMnGSWvZTJyrf3cOUf01EoW3fdhGnDWloCJyOPO8dg274d8wcfokpJRj92HLrhw5BpWwa3giOjuOjG2xh39Y2s/+9HZK1cTnBUDMMuu6JD16wqPMS+jT8zava8TncxHUF4Qi9m3PsAE667mZx1P+L3+wluEvJBkVHog4I75GsVRZGqwkb2bS7j4M4q3A4vyenhjL+6DwGhrbsKo/RRNLgbcHgdmKwmFIKiueHMhQTTgTo2fnWQmhIrkb0DUeuV/PBeDvMeGU7I5Mek2oZ1/4Cr3gWkdpMbDrbs2Nfo9LCn1MK9k5OPOXEGbH5DoltJntqdt9QDLgAlEJEYyPQFA1j1Tg5rF+7n4gUDWnWFGDVGPH4PhQ2Fze/PdViWr6D8sceQBQRg3bwZ8wcfIiiVaNPT0Y8dg37sWDRpaQhNgWGFSsXk39xBQ3U1P3/yHsFRUSSln6Ry9hTY8L9P0Oj0jLxyfmff0gkwhBoZM++6Mz6PvcHNgW0V7NtcTl25DYVSRtKwcAJCNOxeU0LR3lpGXN6LodMSkCtbKpcjfQWq7FWUNpYSpY9CfpbjQN0JS7WdzUsKOLSrGkOImktuH0jKiAgaa5189c+dfPfWHuY/OgLNmHtg42sSVUr0YNJigliaaaKqwUlEoKRgdxSa8flFxh6pD/C64OvfSRXsTa6kcwHe6mrKXngZ7eBBhMy5EkVA53Q9c2TnULdoEd46M8bb70A/WspIk+JaHhpqHTTWOGmoddBQ46ShxkFDjYPZ9w89rcXaUZz3SqDi2X8gz81lUO8ZZO+AwECBMVef2MTsyMw/ry4P6Bx3kN/nx1rnotHsxGp2EhShI7J34Bn740VRpPa996l+5RV0I0cS9683EZRK7BmZ2LZswbZ1C9Wvv0H1628gMxjQjRqFfswY9OPGokpOZuYfHuKzv/2ZFa+9wPVPvyCRjB0Dt8NLbZkNc5kVY6yBqKSg5m3FObsp3JXBpJsWoNGf2+0AfT4/Rdm17NtcTnFOLX6/SGTvQCbf2JeUEZGotdLjP2BiDJu+zGfr14fYv6WCi65NJX7A0d//yKy/yl6FyWq6YFxBLoeXnd8XsmdtCTKFjNGzkxg6PR6FSlKAgWFaLrtrEF+/lsWqd7O54rf3I8/4GH76G9y8rEXP4SNKYFN+LSqFjGEJTZk561+Sumrd8IVEw9BF8Pv82CxubPUunFYPTlvTcuxrmwen1YuzwYnD4sAvuwQ2AZu2I8OPXCVHrlIgV8iQKwRprZQhk0vvFUoZcqUcuUKGQiVtUyhkyOXgKy7EtScLX2kRCoUMvzoI+zPL8cTsxxWWiNUGXlfLRjzaACUBRi0RvQK7tJneea8ElLGxOLL3EPbti0QnXUvGmrF4Pnuf5L5adMPS0Q4bhqp37xZKIFAViPKYPGpRFPFbLLhLTXhKS3CXlOApNeEsr8ApGHCoQnAoAnHIAnCgw+7XYPeqcXgUiLQU+CHRegaMj6bvmCi0hpPHKE4G0eul4ulnqP/iCwJnzSL6//6BTCWdxzBxAoaJEwDw1tVh37YN2+Yt2LZuxbp2LQCK8HACZ85k5o0L+OLfL7Pk/55i3PWPYatXYDZZqTXZaDQfbeYhCHDRdamkTYpDFEXWL/6YAGM4Qy+d1e6xHw+P24ej0Y3T6sHR6MHR6JbWVjcOqwdnoxt7owen1Y1cKUdrUKINUKI1qKR1gAqNQVof+VxjUFJXYWP/5nIObKvA0ehBF6hiyLR4+o2LJjT6RIbUQKOWy+4eRFFOLes/z+PbN3aRPCyc8fMlF9ERS6DCVkGptZQp8VPO+N7dTi/1lXbqKuzUldvw+USik4KITglCG9D+5+J4iF4vzv0HcOZko4yJQTd8ODJ92yik/T4/uRvL2Lb8ME6bh35joxlzZRL6oBNTNaNTgpl6c39++iiX9V9XMnniwwirH4OCdQyIlZ7FvWUWpvSTvsPNBbWMSAxBo5RD+R7Y+AoMvg5SL+34vYoiLrsXa50Tq7lp0lXnpNHswmp20mh2YrO4myu2j4UgE9DoFWj0SjQGJQY9aPZlEGarxzj7UhBk2HZn4ywoxC8KyGJiUST3QRYehd8n4vOK+Lx+fB4/bqcPn9WDz+PH6/Hjc3vx2F34PH78MiVoRkLKUctbIfOjsVWhycsiIUJHxJQRhPZPINCoIcCoQaXpHvF83isB4223YrztVvxOJ/G7s1n5ZRU5XIZ663sELJXaGMiDgghOS+VKpZ/ysAJGeIKpLHkRT0kJblMpnpJS/I1HC1BEoCL5YvLi5uITlOABPCCIPtRuCxpnBYGOWiKcZjQuMxpnLWpXPQ2hfajSzGbTVza2LCsgaWg4A8bHENcvBEF2euvAb7NR+uCD2H5Zj37BvdinXs3GJYepKbWCQNMMRVpkCgG5IgF5ai/kA24Ehw1/ZRme4iIs263YcnbjVV+Ky/olP737IuqgawiNDiIqKZABE2MwxhoIjtCyeUk+v/wvj/oqB2GxFVQeOsiMe/+IQtUxQSWKIvs2lbNlWQFOm6fVfWQKoYWgDwrX4vP6cTS6MZfZcDTWn/RYBOkHkskEeg0Jo//YaBIGhiKTnz5+kJhm5Pq+o9j1YzE7VxZRlFPLyJm96XOR1Pe1sKEQs9PcZkvgiIlfV26jrsImCfymtbXOdXTIgoggwK4fpWcgJEpHdJ9gYlKCiU4JItB4eo4nv9OJY/ce7Bk7cWRkYs/Kwu0WcamDAQGZIKLrm4Jh+BAMI0dgGDYEhU59glVanFvLpq/yMZfZiOkTzISr+xCecOpAbd/RUdSV28hYVURIxGUMDX4bfvwrAXf+QqJRx94yqV+U2eZmX3kDD12SKqXGfnOv1L9ixnNt+j6Ph8vuYcPnBzm0qxrPcbNomVzAEKImIFRDbN8QAkI1GELU6IPV0uShSeirNPLm78BdWEjRbQvw22wkvP8e2sGDm842Bq/ZTP2XX1H32Wd4N5SjiI4m5LrrCL5uPorQll4D54EDmD/9lIbVKxBdLvTjxhJ8003oxk3EJwr4PH5kMgG1XoHfZsP8ySeYP3wX/7rXUM++goDf/x6VpvsaFp33xWLHw2n18NULO3Hbvcy+IQLFoWzsmVlYM3fiKzxK4SyoVCjj4lDGx6GKjUMZH48qPg5/WAybN7k4lF1HXL8Q+o2JIsCowRCqwRCsbhY2oseD3+VCdDrxO534Gxqo/fhjGr5djiN2AOapCyg0B+CyeQkwaug/Lpr+46IxhLQenLQUmMh54g2qbToaU8bSYG/y86vlhMcbkMmEo7OSYxZ/i89ERL+IPlBBgL8eTeFu3A17ORRsJSEyhjlPPYfS2DIW4veLbPrqILvXFOF3LSLQqOc3L77ZodoIp83Dz4v3U5BZTUyfYBIGhkqz+ONm88pj/pgng9/nx2nzStaDtaUVoTUo6TMistUZtbemBsfu3Th27cKxazfu4mLkoaEowsKkJTwcRVgYTn0YGQf0FBd7CQ7XsCTq3wSnqNhRsYMXL3qRGb0lYj6X3UOjWZp1NtZKs84ja0ulHZfjKLeNQubDQCM6RxW62sNozEXo7RVoHTWAQGNAPFDnK/wAABnpSURBVHXBKVhC+2IJTMIrk2beOo2fqDgNsWmRxA6WrBlXlZnqLbsx787Dkl9OY60DlzIQpzoEd0AkTlUQPvH0v5FM8CNXylGoFchkAjaLm8AwDePmpZA0NLzF7yB6PPisVvwNDfgaGhHdLmQ6HTK9HkGr48cvSzmcbWbmZRZ6Zd4KV73P77KTyTZZWP/IFL7PLufexZksvXccww6/D+uehWsXQ//2W5VlB+v58aO92Ovd9BsXTUiUDkOIRhL2oWp0Aao2TayOwJWfT/FtCxC9XhI+/ABN/9ar30WvF+vPP2NevBj7lq0ISiWBl19GyPXX462pwbzwU+zbtyNoNARdeSWhN92Ius/pm8Z46+qoff996hYtRvT5CL56PmF334My8sx7VVzwFcOtob7Szlf/3IkuSM28R4aj1irwi34mvZNOmNlH2oBJPD37zROyT8ry6/nxg73YLW5Gz0kifXpCux40AMeePVQ+9zyOrCwU/QbguvZBCip1lO6vQxAgIc3IgPExhCcEUJ5fjymvHtPeKix1kjBRKiC6r5HY1GBiUoMJTwhA3oZZ7hGIfrF5zKLHQ+OatWxd+AHZPjvJNQ2MGjGOkBtvOGYWJOH7fy1m34b/Yex1HfMfvRpDSPva4ZUdrOPHD3Nb/e78Tic+sxlvrRmfufa4tRlvnRmZVocyOhplTDSKqCiU0TEoY6KRh568SEh0u3Hu349j1+5mwe8xNdElKxRo+vdHnZSEz2LBW1PTvOA9KrhrQgdysM/VOLThBNbvQua1INdG41IG41QENQvqI5D5PWi8DWg9FjSN5egay9DbK9DZK1B7G1FFR6NKiEcZn4AqIQFlQjyqhARkGg3u4mLchwtxFxXhKiqmttxOrSuA+sBk6oOScTexmMr8Xvyy4414Ea1GICBcR0CYHkOIJAz1wWpkMkFyT3j9eKwOnIeKcBaV4Cwpw11nwS9TIKp0yCKiCAn0k+g7CNZ6/I1W/I2SwPc1NiLa7af8jX0yFZnpD2LXRTBq/8sEe01UBadSYIVxgxPY5VCww6bi4dlJqLf+DeXgqShuW9icuNAW+Hx+dn5XSMbKQgLCtFyyYCCRvTtIt94E5/79FC+4HeQyEj/8sE1CG8BVUEDdf/+H5euv8TfxQCliogm98UaC581DHtz+OiJPZRU1b/+H+i+/QpDLCbnpRox33IEipOPVzT1K4CQoPVDH8td3EdsvhFm/G4xMLmPqF1OpdlRzXd/reHzM4837+v0iGSsL2bHisPTg3T6QyF4df/BEUaRx1SqqXnwJT1kZhmnT0P72fvKLZOzfXI7NcpQkTqmCoKp9hDiL6Xff9cRNGtwm10Z7x7PqpX+Qu3MrQ/6/vTMPr6K6+/jnZCUJSSAQFoUI1YACAmpFbdUC7ra2ahFE0apFX63U9u3b9+lmq2+1rX2qrSIuoIJicUOlSl1AlMUNFDXsYVEEQhZCQgg3+733vH/8zphhuDMXKDGBez7PM8+9d+Y3537n/M6c35lzZs6U13B0RRWdhgyh61VXkXPxRYTRPHHbjWTm9qSx6RLSM1P57q1Dye8b/37uaCTKx+akzemewbnXFJK+YiG1r86lpaKCSFXVVyeQF5WeTkq3biR37Uq0ro6W8nJ0494vH1dpaaT07kVqr96k9u5NSu9e6IZGGlasoHHNGnSz5GVKr15kDBsmy/DhdBp0Akkxnh7X0agEhcpKIjt3Eq6spGlHFXM/2UqjPg10hOzkFrKSGslIaiAzqZFM1UBmciOZSY2kq2YJSkqR3KWLVPJ9C6TiP+oo1AFOh6Gbm2ku2U7Tl5up3lBK2eYQu/ckkdUjly4D+9B1aCE5PbPJ6pJOcsqBl4twdbWMHS1dRv3SpURqakjKzSU5O5uknGySs3O++kzOySapc7Z8Zueg0tOI1tcTrauTJVRHXU0TC7YUkhRt5uzNv6MlrQdrd6YyIAuaKyrIaPG8PD4piZT8fFJ79SKld2/57NUTlZQkadc3mM969oRgedMwalR3jm5YxwkV80iqqyGld2/yrplAzoUXHnD+NqxazdaJE0nKyKBgxnTS+/ePv5OHSKiOPfPeJCknh+xRow4oqPnRvG0bO6dMYferc0nKyqLgicfJGHZwb1azQSCAte+XsvDpYoacfTRnjx/A2H+Ppbi6mFuH38rNw24GILSrkbemr6V0Yw0DRvTkO+MHkpZxaIZSok1NVD/5FFVTpxJtaSHvqqvI+6+b2V4Spraqgc7bV9L419+SXtCXgmlTST36ED2lGoNIOMzLf7mDknWrueD0UaTPX0Dzps9Jyslhy6nDWLWzlCvv+hvpmX147aGVNNaHuWDiYPqd6P+cQO3OBt6avobyL2oZcFIXBte9R+iFZ4nU1JBeWEj6wIGkdMsjOa+b+cwjJS+P5G7dSMnLQ2Vm7t0doTWRmhpaSksJl5fTUlpGS1kZ4fKy1u+VlaiUFDoNGeKq9IeR2qvXf5Q/f3j/D8zZ+C8yUzNYetXSw2pagK+bHVtqmXPvp3TvtJ1Lut7JKbV/5sqzBvPYu5t54bj3GLp6GuGTf0lL+rG0lJcRLiunpbyccFmZBPqm1vESUlJQWVlU9DqNdT0vQqEZGl5Kn/RK6YrKzKD+089o/vxzUnr1Iu/aa+lyxRiSs+M3UOo//YxtN91Ecm4uBU89SVqfjnfXV9PGjVTPnEnP228nKf3gXkZvg0AcPnhpE5+9tZUzryjkkeifeL/0fX5/+u8ZO3AsXxRV8s7T64iENSPHD2Dg6THeHHQICFdWUjl5MjUvvkRybi7dJ00iGgpRef/9ZI4YQZ8HJ5Oc23a3zzk0hkI88/tf0lC7m/F33Uv6tu2Uv/Acc7cW0722nm+lZZN72eWkjLyQec+VsHPbHr59RSHDRu87iLXh43IWz1qPjkYZlvQZOQtmQCRC53NGk3fttWSeemqbVKS6RQaMD7RFGI+Hih7i0RWPUti1kJe///IhTftIZNMnO5j32GoGdFpMSdet3Be9kp7NW3gr83aSCs+Dcf+MOa+RE+gBkrOyaA4rFj+zno3Ld9D7uFzOu2HwPg/16WiU0JIlVM94kvply0jKyqLL2LHkXXsNqb1jn7N1yz5i2y23kJqfT8GTM3ztjgQ6bBBQSl0IPIC8cudxrfU9QfZtFQR0VPPG1FVsXrmTqu8UMbtpBn8/837SPurD6sXbyS/I5vwfD44799ChoLG4mIp7/kr90qUA+9wC+nVQU17GrNv/h4zO2Yy/+14+nP0MRfNf4/JzL4H5b9OwYgUkJ9Pp7NGs6v0DtpUqThzZhzOvOI6k5CSaG8MseXY965dV0DVcwfHLp5CV3ESXMT+k64QJpPX9+u56OJTM3jCbP374R0b2HcmDox9sbzmHBctf38yyVzdzSufn+EXKt3mk0xSGZlSifrJsvybuK91Uw4LpawnVNDHikv6cfMExJMUZg2tYvYbqGTOoffNNAHIuuoi8668jY3Drq0tD771Pya23ktq3DwXTp5Pa4z8ffO3IdMggoJRKBjYA5wElwMfAeK31Wr992vIdwy1NEebc9yk7SmtY0H8mP6idSH1FhGHn9uWMS489qL7Wg0VrTWjRIsLl5XQZN65dpqEtKV7Di3f9jh79jqVi8+cMGXku5900CZDBsN1z5lDzyiuEK6v4YvCVbMk/kz79O3HSeQUsfGoNocZk+m15g8LwKrpfO4Hcyy8nuXPHfrAsHktKlnDr27cy4YQJ/GrEr9pbzmGB1poFUz9mQ1GIU3MfYUTGfLhsGgwbF7ifexzpYAd/W0pLqZ75NDWzZxOtqyPztNPodsP16EiE7T/7OWnHHkvB9Cf2ub3zSKSjBoEzgDu11heY378B0Fr73jDclkEAILSriZl3L0bXpZCWlcz51w/hmCEdf+qItmLtuwt5Y8p9pKSl8+MHpu0zU6MOhwm99x67X55D8bomNnzjh2iVTHpjNSfrpRRecyGdR41CfU1vT2pr1levZ8zcMfx6xK+5+oSr21vOYUO4JcLzv53Dnj2dyUzfQ0peXyC4Nd/SGCa0q4njz+jFWeMG/EcPTUX27KHmhdlUz5xJuKICgE4nnkjBY9MO6u6dw5GOGgTGABdqrSea39cAp2mtJ3nsbgJuAigoKDhly5YtbaqreONmFsxbznUTvkfnLgf5EvYjiNUL3yK1UycGnnFWoF141y42PP0WJduaOW38cHJPHhJofzgSiUaY/Nlkrj7h6q+eILbsH1Xby5n/0Ot07j2E1PT9Oa8Ux53Sg+NOOXT5rFtaqH3zTRpWrCT/Z7ft18DxkUJHDQJXABd4gsAIrbXvq5ra+krAYrFYjkTiBYH2eu9ZCeAeIewDlLaTFovFYklY2isIfAwUKqX6K6XSgCuBV9tJi8VisSQs7TKBnNY6rJSaBMxDbhGdrrVe0x5aLBaLJZFpt1lEtdavA6+31/9bLBaLpf26gywWi8XSAbBBwGKxWBIYGwQsFoslgbFBwGKxWBKYw2YWUaVUJXCwjwx3B3a2gW1bpn2ka+4oOqzmxNKRCJq9HKO1zvfdqrU+4hdgeVvYtmXaR7rmjqLDak4sHYmg+UAX2x1ksVgsCYwNAhaLxZLAJEoQmNZGtm2Z9pGuuaPosJoTS0ciaD4gDpuBYYvFYrEcehLlSsBisVgsMbBBwGKxWBKZtrz16FAtQAQoci39Amw1sD6eLdAfqDX2u4AVxn4s8CkQBsa4bJcZ21qX7V3AWmAl8DZyP65jr4FmV9r9gJuBVWbf94BBATr6mbTGmG3fDNDxS6DSdcwTg3SYbWON9jXAMwGa/+FKdwNQE6D5W8BC4DOTJxcHaP62ybOVwCKgj9E1ydhqV171Q95JOBnYZPY5OcD2eOBDoAn4pSvdBo9tEfAzk95K4ANgmLHdZGw3uWxvNHZFwHLgzCDNrrJ2KlKGxwRoHgnsdv3XHwI0O/ZFxn+LAzT/2fV9tdGRF6AjF5hr/LQGuD5Ax1BgjsmTj4AhwCzk/NOe4+nn9Z/Jm1nGtgmoiuNDP9urY/hwFlBv7B3bmD4M0uz1XxwdsXzopyOWDx0ddcArQKr5v//18aFjvxqY7rLfx4eB9Wt7V/D7GQRCBxgwvrkfdi8glWA98Chwi1nfzxTumS6nv4C88yACvOSyHQVkmu+3AM+77KNIxe1OO8f1/98H3vTTYWyygSXAUpOWn47rgCk+xxhLRyFSUXc1v3v42XrS+6kpbH55N831fRDwZYDm2cCPzPfRwNPm+0nGthTo7vrvi4E3kGBwOhJY/Gx7ICfun2itQE4y+Rj12H7LlQ8XudLth5y457hsO9M6jjYUKA7SbLYlA+8gM+aOCdA8Evi3Z18/zV1M/he4jjemZk96lwDvxMnn3wJ/Nd/zgWqTl7F0/A24w3w/HgnqFxsfRZCp4h1/7+M/1/oIUu6eddnH8qGfbSwfXow0Luo9tvv40E9zLP/F0RHLh346YvnQ0bHIq8PHh4698qQdy4dpvnXhoaik23ohRhAATkGi5ycmw3qb9RHgfqRFsBoYEcsWeQJvMdLyXQnsQV5x6aRdDnxu7KuQabcjwHNI69eb9lqz3knb0fFV2h4dK0zhCtLxDNLy2oO0ivx0/B/wvM8xxtLxFNKC2B9bt+ZdwDg/zcBUpLW3GDm5dgVo/gI5YZ20wx4f1iKtSyefXzJ57GjeZHTHsnU0l5rjdNJdhFSSX9l6ylIR0Oiy10iQi1WWioFQHM2nGJ1bge20XqHFsr3J5JW3PO+jGfgJ8LjX3/uhuRppGQZpftBo/QRpgGxGuo1j6XgN+LFLRz1woivtxchbBFcDLyOVk2NbBwzznLObgR1xfBhk+wkSiMpcedfstvX4O5YP3ZpHAPcCG5EyXxRHR5APmz22QT5c5NXhqveeAW6MUR/+N/An8/03wMNIcOiPlMGkwz0IuLuD5gCpSCHPN9vHIS+mgb0vu5aYTPTazjIZswhp3axBLu1qXWk/i7TcJgK1rrR3IZfGe6UNTEEqOSdtR8capMA7tr9CgstO5G1qfjpONfvnm+13BegoMYVsDfCiKWBBOqqQk/19pIC/E0dzPnCM+c9nAjT3NfuVGtsgzTXISZEPXG62z3LZRpAA6ORzNfA9lw9LkZZYLFtH853A07SWjUW0dmlsZG9/5yPdam+zd1mqZu+yVGz2rQZu99h6dSw3aSeb/34nwNY5xhXIVdrLAZonA2XGf58AD8XR/AFQYNbfEEfzUqSrsgwJiPMDdNxj/JCPVGwR4BVX2mFju8TYr6L1PFzlsa1CgsiKOD70tTXbnwbWuzRHjW0Jrf4O8qFbczHSvdMTeBL4exwdQT706gjy4WKkfjgLOBtYbdZnmvTzPPVjKtKFfZb5nY10y5YhQe67gfVre1fw+xkEQp7fQ0xGOoFhlauwRoDRLtuyGLYLaa3IGpDKaxXSYjvD2FcjldRaoN6V9jhglSftLbRWmk7aEaSbw0nbq2OrWeen41OkhV1kHPlFgI5yl+02Yx+kI4xUwkVIQW806QdpLjUFMyjvHjBpOeMHTQGaK4AW5MSoNN/fcdmWY7oeXJo3unxYhbToYtk6msvNb3dFFnXZuv29yWhf67LXwKU+ZWmjyef5AZpbTD4UIQGwKMB2D6191V8CdQGaZ5v8WGnyogl4L47mL5EKzXuueHU0IK3VIqQFXg/k+Og4F2l8NCDnSz3wgdkeBV506WhEzhGn/IeAD106XgPu3w8fBtk6Plzoyruwy9bt71g+9GpucGmuRs7zIB1BPvTqCPJhqUfHVqT7aBwwN0b9+JiTtvk9BhnLU8BxSD2W493PWQ7Xu4MUsEZrPdwsJ2qtz3dt1x77YqTSAekvHI1kqjLrvC+6j5rPz4HBQL1SynkLW3ePbSnQDcnopa60HR3utN065iORPZYOBXzDbB8EdDJLs48ODawzaVcBaXF0VCPBAqTV9z4y0BSkuQvS4grKu0uQwgzSktoONPhojiKtpXeRE6hZaz0af5xKBsSHO9nbD27c+bzVUzb86Au8qrUetJ9lqc5suyogzYhJdxDSF32UUupSH9t1SD4ALAAqlVLdfWx3IAEmiuTFP5EAHKQ5AwlEy+LkRxNSlkF8sxTp749FHeKT9Uh3zw7gQqXUHUj5eNhlGzHH5/gljFwBYmxzgV94NMfyoZ9tHa0+HGXWH2Ps3bbQen4vRhogV/loTkW6Mgea/+zh2h5Lh58PY+mI6UOjI82jA8Snk4ETlVKPOyuNfb4n7euRqxCttd6E+NPPh4dtEFgP5CulzgBQSqUqpQa7to8z689EKsU84E6t9XBkAHcQ0qLNR5z8I6QVvxu5NNuJ9ME/arZ/gkRXkK6WV0zaIeBYZPR9iCdtZXQ4aVchl5WOjn8jFXEsHTVIgdsMfAepqC9DCkssHTXOMSLjA5/F0REGthkdtxvbxT62echg8A7gh3HybgPSErsFuA8JXMt9NO9yaZ4L/MPjwyyPD8NIhXMSMAMJGmU+tk66jwIvedJVLtvd5rg6AT/VWo+NUZYuc9nXutK+wfzuGaC5BLkqTUO60f6stf6Xj20+rWXjCSDdrI+leSpSKUwCbgNOQ1rksTRXIcE3A2nQ3BInnxVytTYcuMPs84WPjk3IuXILUjbfBX4OXIBUbmNd9pXIGM6dyF1yG4E8pdREk+46rXU0yIcBtj2QinYUcLXLNg+Iumx3I+eKo/lhpML9iY/mYuQcHY10M10HbAzQEcuHl/roiOXDs42OtR4du03+pwKDtdYTzbaJxn681toJbCBXDucYm55IEPsCP9qri+dAFmIPDA9HWpvObVA3ui7T/sLeA2P72CItbec2x51m/QikL77CpBNGKrXfIANiTov7Q5P2R8b5IeTSsdaVttM36057FtJicS4z7/DT4TnGEHJV4qfjSbM9ZJaNcXQMRyqoRrNMi2O7xeRJvLwbROulfgNySeyn+dcm3XqTh2tN2rcZ2yhSsVW7dJQiLdVG4O4A23OMXcQsu5Db7JpovQ3WsX3cHIujuQEZDyoxtnvMsa9GxlGcfK5DgnQ8zU652wU8EmD7F1rLRj3S335bLM2mbPzDZV8WoHkEcqviDvYu/346zqN17KbR5E9MHUjXiuPDGuPfiMkjbTRsdel43pXHnxsdYZdtnfn086Gf7ctIq9pJ+0tj22COz23rjMm5fRik2fGhc/UcpDmWD2Pq8PGhoyMUQ8d1wHOeOjBs7L+6JdWsPwrpaVhl9p8QVL/aaSMsFoslgTlcu4MsFovFcgiwQcBisVgSGBsELBaLJYGxQcBisVgSGBsELBaLJYGxQcBiiYFSKqKUKlJKrVFKrVBK/UIpFXi+KKX6KaWCHh6zWDocNghYLLFpME+jD0bunb8Yea4jiH4EP0FssXQ47HMCFksMlFIhrXVn1+9vAB8jT98egzxBmmU2T9Jaf6CUWgqcgDyA9BTymP89yBTD6cBDWuupX9tBWCz7gQ0CFksMvEHArNuFzMGyB5kGoFEpVQg8q7X+plJqJDL3/feM/U1AD6313UqpdGSOpiu01puxWDoIKfFNLBaLwZk0LxWYopQajjzqP8DH/nxgqFLKmTspF3mhjw0Clg6DDQIWy35guoMiyBw8dyBzuwxDxtUa/XZDJqab97WItFgOAjswbLHEQSmVj8xmOUVL/2ku8vaqKHAN8tIYkG6ibNeu85BZO1NNOgOUUllYLB0IeyVgscQmQylVhHT9hJGB4L+bbQ8j0xtfgUyrXWfWrwTCSqkVyMyuDyB3DH2qlFLIdMp+7xOwWNoFOzBssVgsCYztDrJYLJYExgYBi8ViSWBsELBYLJYExgYBi8ViSWBsELBYLJYExgYBi8ViSWBsELBYLJYE5v8BgdLskVMW/YwAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.lineplot(x = 'Date', y = 'Estimated_fire_area', hue = 'Region', data = submission_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}