test.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Adding Problem with LSTM and IndRNN #"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "from keras.models import Sequential\n",
    "from keras.layers import CuDNNLSTM, RNN, Dense\n",
    "from keras.optimizers import Adam\n",
    "\n",
    "from keras_tqdm.tqdm_callback import TQDMCallback\n",
    "\n",
    "import sys\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "\n",
    "from indrnn import IndRNNCell"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "def generator(timesteps, batchsize=128):\n",
    "    while True:\n",
    "        x = np.random.uniform(0, 1, size=(batchsize, timesteps, 1))\n",
    "        x = np.concatenate([x, np.zeros_like(x)], axis=-1)\n",
    "        x[:, np.random.randint(0, timesteps, size=2), 1] = 1\n",
    "\n",
    "        y = np.expand_dims(np.sum(np.prod(x, axis=-1), axis=-1), axis=-1)\n",
    "\n",
    "        yield x, y\n",
    "\n",
    "        \n",
    "batchsize = 128\n",
    "epochs = 100\n",
    "steps = 100\n",
    "\n",
    "gen_100 = generator(100, batchsize)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## LSTM Test ##"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "cu_dnnlstm_4 (CuDNNLSTM)     (None, 128)               67584     \n",
      "_________________________________________________________________\n",
      "dense_4 (Dense)              (None, 1)                 129       \n",
      "=================================================================\n",
      "Total params: 67,713\n",
      "Trainable params: 67,713\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "lstm = Sequential()\n",
    "lstm.add(CuDNNLSTM(128, input_shape=(100, 2)))\n",
    "lstm.add(Dense(1))\n",
    "\n",
    "lstm.compile(optimizer=Adam(0.002, amsgrad=False), loss='mse')\n",
    "lstm.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "lstm_hist = lstm.fit_generator(gen_100, \n",
    "                               steps_per_epoch=steps, \n",
    "                               epochs=epochs,\n",
    "                               callbacks=[\n",
    "                                   TQDMCallback(leave_inner=False,\n",
    "                                                output_file=sys.__stdout__)\n",
    "                               ],\n",
    "                               verbose=0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## IndRNN Test ##"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "rnn_1 (RNN)                  (None, 128)               17152     \n",
      "_________________________________________________________________\n",
      "dense_2 (Dense)              (None, 1)                 129       \n",
      "=================================================================\n",
      "Total params: 17,281\n",
      "Trainable params: 17,281\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "indrnn = Sequential()\n",
    "indrnn.add(RNN([\n",
    "        IndRNNCell(128),\n",
    "        IndRNNCell(128)\n",
    "    ], input_shape=(100, 2)))\n",
    "indrnn.add(Dense(1))\n",
    "\n",
    "indrnn.compile(optimizer=Adam(0.0002, amsgrad=False), loss='mse')\n",
    "indrnn.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "indrnn_hist = indrnn.fit_generator(gen_100, \n",
    "                                   steps_per_epoch=steps, \n",
    "                                   epochs=epochs,\n",
    "                                   callbacks=[\n",
    "                                       TQDMCallback(leave_inner=False,\n",
    "                                                    output_file=sys.__stdout__)\n",
    "                                   ],\n",
    "                                   verbose=0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Comparison ##"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "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>Loss</th>\n",
       "      <th>RNN</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Epoch</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.199443</td>\n",
       "      <td>LSTM</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.177778</td>\n",
       "      <td>LSTM</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.167373</td>\n",
       "      <td>LSTM</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.172410</td>\n",
       "      <td>LSTM</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.168577</td>\n",
       "      <td>LSTM</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           Loss   RNN\n",
       "Epoch                \n",
       "0      0.199443  LSTM\n",
       "1      0.177778  LSTM\n",
       "2      0.167373  LSTM\n",
       "3      0.172410  LSTM\n",
       "4      0.168577  LSTM"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "losses = []\n",
    "\n",
    "for hs in [lstm_hist, indrnn_hist]:\n",
    "    df = pd.DataFrame(index=range(epochs))\n",
    "    df.index.rename('Epoch', True)\n",
    "    df['Loss'] = hs.history['loss']\n",
    "    df['RNN'] = 'LSTM' if hs is lstm_hist else 'IndRNN'\n",
    "    losses.append(df)\n",
    "\n",
    "losses = pd.concat(losses)\n",
    "losses.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA7kAAAIaCAYAAADlWIo0AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4zLCBo\ndHRwOi8vbWF0cGxvdGxpYi5vcmcvIxREBQAAIABJREFUeJzs3Xd4VFXi//HPSZmUSSAzKbQQCaEX\nCRBAAQXFhr1gR0VUVn+6uuru2nYVXferu2KXVVlFXV1BUde2goqIIqg0qYZeQw3pvc39/TGBpQRI\nYGbuJHm/nifPZO7cO/kkIg+fnHPPMZZlCQAAAACApiDE7gAAAAAAAPgKJRcAAAAA0GRQcgEAAAAA\nTQYlFwAAAADQZFByAQAAAABNBiUXAAAAANBkUHIBAAAAAE0GJRcAAAAA0GRQcgEAAAAATQYlFwAA\nAADQZITZHcBXEhISrA4dOtgdAwAAAADgB4sWLdpjWVbi0c5rMiW3Q4cOWrhwod0xAAAAAAB+YIzZ\nXJ/zmK4MAAAAAGgyKLkAAAAAgCaDkgsAAAAAaDKazD25AAAAANAYVFVVKSsrS+Xl5XZHCUqRkZFK\nTk5WeHj4MV1PyQUAAACAAMrKylJsbKw6dOggY4zdcYKKZVnKyclRVlaWUlNTj+k9mK4MAAAAAAFU\nXl6u+Ph4Cm4djDGKj48/rlFuSi4AAAAABBgF9/CO92dDyQUAAACAZiY0NFTp6enq1auXLrjgAuXn\n50uSNm3aJGOMXnzxxX3n3nHHHXrzzTclSWPGjFG7du1UUVEhSdqzZ486dOgQ6PhHRMkFAAAAgGYm\nKipKS5Ys0YoVK+R2uzVx4sR9ryUlJen5559XZWVlndeGhoZq8uTJgYraYJRcAAAAAGjGTj75ZG3b\ntm3f88TERI0YMUJvvfVWnef/7ne/07PPPqvq6upARWwQSi4AAAAANFM1NTX65ptvdOGFFx5w/P77\n79fTTz+tmpqaQ65JSUnR0KFD9fbbbwcqZoNQcgEAAACgmSkrK1N6erri4+OVm5urM88884DXU1NT\nNXDgQL377rt1Xv/ggw/qqaeeksfjCUTcBqHkAgAAAEAzs/ee3M2bN6uysvKAe3L3evDBB/W3v/2t\nziLbqVMnpaen6/333w9E3Aah5AIAAABAM9WyZUu98MILmjBhgqqqqg54rVu3burRo4c+//zzOq99\n6KGHNGHChEDEbBBKLgAAAAA0Y3379lWfPn00derUQ1576KGHlJWVVed1PXv2VL9+/fwdr8GMZVl2\nZ/CJjIwMa+HChXbHAAAAAIAjyszMVPfu3e2OEdTq+hkZYxZZlpVxtGsZyQUAAAAANBmUXAAAAABA\nk0HJDaQmMjUcAAAAAIIVJTdQpl4rvXWB3SkAAAAAoEmj5AaKCZFKsu1OAQAAAABNGiU3UKJcUmmu\n3SkAAAAAoEmj5AZKtFsqy+O+XAAAAAC2i4mJOeTY6tWrNXz4cKWnp6t79+4aN26cvvzyS6Wnpys9\nPV0xMTHq2rWr0tPTdf3112v27Nkyxuj111/f9x6//PKLjDGaMGFCIL+dA4TZ9pWbmyiX5KmSKkuk\niEP/QAEAAACAne68807dfffduuiiiyRJy5cvV+/evXX22WdLkoYPH64JEyYoI8O7Ve3s2bPVu3dv\nvffee7rpppskSVOnTlWfPn3s+QZqMZIbKFEu72NZnr05AAAAAKAOO3bsUHJy8r7nvXv3Puo1KSkp\nKi8v165du2RZlmbMmKGRI0f6M+ZRMZIbKPtKbq4U197eLAAAAACCwqOfrdSv2wt9+p492rbQIxf0\nbPB1d999t04//XQNHjxYZ511lm688UbFxcUd9bpRo0Zp2rRp6tu3r/r166eIiIhjie0zjOQGSpTb\n+8hILgAAAIAgdOONNyozM1OXX365Zs+erZNOOkkVFRVHve6KK67QtGnTNGXKFF199dUBSHpkjOQG\nCtOVAQAAABzkWEZc/alt27YaO3asxo4dq169emnFihXq37//Ea9p3bq1wsPD9fXXX+v555/XvHnz\nApS2bpTcQNlbctlGCAAAAEAQmjFjhkaMGKHw8HDt3LlTOTk5ateuXb2ufeyxx7R7926Fhob6OeXR\nUXIDhZFcAAAAAEGitLT0gEWm7rnnHmVlZemuu+5SZGSkJOmpp55S69at6/V+gwcP9kvOY0HJDZTw\nSCk8mpILAAAAwHYej6fO488888xhr5k9e/YBz4cPH67hw4cfct748eOPI9nxY+GpQIpyUXIBAAAA\nwI8ouYEU5abkAgAAAIAfUXIDKSqOkgsAAAAAfkTJDaQoF6srAwAAAIAfUXIDKZrpygAAAADgT5Tc\nQNq78JRl2Z0EAAAAAJokSm4gRbkkT5VUWWx3EgAAAADNWExMTIPOnz17ts4//3xJ0ptvvqnExESl\np6erW7duevbZZ/edN378eEVHR2v37t11fi1jjO699959zydMmODzLYcouYEU5fY+MmUZAAAAQCN2\n5ZVXasmSJZo7d67++te/auvWrfteS0hI0NNPP13ndREREfroo4+0Z88ev2Wj5AZSlMv7SMkFAAAA\nEARmz56t4cOHa9SoUerWrZuuvfZaWbW3V86YMUPdunXT0KFD9dFHH9V5fXx8vDp16qQdO3bsOzZ2\n7Fi99957ys09dNHdsLAwjRs37oDRX18L89s741B7Sy4rLAMAAACQpOn3SzuX+/Y9W/eWRj5Z79N/\n+eUXrVy5Um3bttWQIUM0d+5cZWRk6JZbbtGsWbPUqVMnXXnllXVeu2XLFpWXl+vEE0/cdywmJkZj\nx47V888/r0cfffSQa26//XadeOKJ+uMf/9jw760eGMkNpGimKwMAAAAILgMHDlRycrJCQkKUnp6u\nTZs2adWqVUpNTVXnzp1ljNHo0aMPuOa9995Tz5491bFjR911112KjIw84PU777xTb731lgoLCw/5\nei1atND111+vF154wS/fDyO5gcR0ZQAAAAD7a8CIq79ERETs+zw0NFTV1dWSvItEHc6VV16pl156\nST/++KPOO+88jRw5Uq1bt973elxcnK655hr94x//qPP63/3ud+rXr59uvPFGH30X/8NIbiDtK7lM\nVwYAAAAQvLp166aNGzdq/fr1kqQpU6bUed7JJ5+s6667Ts8///whr91zzz169dVX95Xm/bndbl1x\nxRV6/fXXfRtclNzACouQwp1SWb7dSQAAAADgsCIjIzVp0iSdd955Gjp0qE444YTDnnvffffpjTfe\nUFFR0QHHExISdMkll6iioqLO6+69916/rLJs9q6c1dhlZGRYCxcutDvG0T3TU+o4TLq47mF7AAAA\nAE1bZmamunfvbneMoFbXz8gYs8iyrIyjXctIbqBFubgnFwAAAAD8hJIbaFFxbCEEAAAAAH5CyQ20\naDcjuQAAAADgJ5TcQGO6MgAAANDsNZW1kfzheH82lNxAi3J5txDiDzUAAADQLEVGRionJ4eiWwfL\nspSTk6PIyMhjfo8wH+ZBfUS5JU+1VFksRcTanQYAAABAgCUnJysrK0vZ2dl2RwlKkZGRSk5OPubr\nKbmBFuXyPpblUXIBAACAZig8PFypqal2x2iymK4caHtLLissAwAAAIDPUXIDLdrtfWTxKQAAAADw\nOUpuoO0/XRkAAAAA4FOU3EDbV3KZrgwAAAAAvkbJDTRGcgEAAADAbyi5gRYWIYU7pbJ8u5MAAAAA\nQJNDybVDlIvVlQEAAADADyi5doh2MV0ZAAAAAPyAkmuHKEouAAAAAPgDJdcOUS5WVwYAAAAAP6Dk\n2iHKzUguAAAAAPgBJdcOe6crW5bdSQAAAACgSaHk2iHKJXmqpcpiu5MAAAAAQJNCybVDlMv7yDZC\nAAAAAOBTfi25xphzjDGrjTHrjDH31/H6PcaYX40xy4wx3xhjTtjvtRuMMWtrP27wZ86Ai3Z7H7kv\nFwAAAAB8ym8l1xgTKmmipJGSeki62hjT46DTfpGUYVnWiZI+kPT32mvdkh6RNEjSQEmPGGNc/soa\ncHtHcim5AAAAAOBT/hzJHShpnWVZGyzLqpQ0VdJF+59gWda3lmWV1j79SVJy7ednS/rasqxcy7Ly\nJH0t6Rw/Zg2sfSWX6coAAAAA4Ev+LLntJG3d73lW7bHDuUnS9GO8tnGJYroyAAAAAPhDmB/f29Rx\nrM49c4wxoyVlSBrWkGuNMeMkjZOklJSUY0sZIIu35KmkolqndE6UouK8Bym5AAAAAOBT/hzJzZLU\nfr/nyZK2H3ySMeYMSQ9JutCyrIqGXGtZ1iTLsjIsy8pITEz0WXB/ePbrNXrii1XeJ2ERUrhTKqXk\nAgAAAIAv+bPkLpDU2RiTaoxxSLpK0qf7n2CM6SvpVXkL7u79XvpS0lnGGFftglNn1R5rtHq2bam1\nu4tUWe3xHoh2M5ILAAAAAD7mt5JrWVa1pDvkLaeZkt63LGulMeYxY8yFtac9JSlG0jRjzBJjzKe1\n1+ZK+ou8RXmBpMdqjzVavdq1UFWNpTW7irwHouIouQAAAADgY/68J1eWZX0h6YuDjj283+dnHOHa\nyZIm+y9dYPVs21KStHJ7gXq1a+ldYZnVlQEAAADAp/w5XRn7OcEdrZiIMK3cXug9EMV0ZQAAAADw\nNUpugISEGHVvE7tfyXVRcgEAAADAxyi5AdSzbUtl7ihUjcf6X8m16txVCQAAAABwDCi5AdSzbQuV\nVtZoU06Jd3VlT7VUUWR3LAAAAABoMii5AfS/xacKvSO5ElOWAQAAAMCHKLkB1LlVjByhIVq5rWC/\nkssKywAAAADgK5TcAAoPDVGX1jG1I7lu70FGcgEAAADAZyi5AdazTUut3F4gKyrOe4CSCwAAAAA+\nQ8kNsJ7tWiivtEq7qqK9B0qZrgwAAAAAvkLJDbC9i0+tyK390Zfl25gGAAAAAJoWSm6AdW8TK2Ok\n5TvLJEcM05UBAAAAwIcouQEW7QhTxwTn/7YRouQCAAAAgM9Qcm3Qs21L/bq9QIqKYwshAAAAAPAh\nSq4NerZtoe0F5apyMJILAAAAAL5EybXB3sWn8sU9uQAAAADgS5RcG/Rs20KStLs6ii2EAAAAAMCH\nKLk2cDkdahcXpazyKO9IrmXZHQkAAAAAmgRKrk16tG2hDSXhklUjVRTZHQcAAAAAmgRKrk16tm2h\nDcXh3iessAwAAAAAPkHJtUnPti2VZ8V6n7D4FAAAAAD4BCXXJj3btlC+5fQ+oeQCAAAAgE9Qcm3S\npmWkrCiX9wkrLAMAAACAT1BybWKMUevWbb1PGMkFAAAAAJ+g5NoopZ235FaXUHIBAAAAwBcouTbq\nnpygYitSBTk77Y4CAAAAAE0CJddGPdu2UL5iVJSfbXcUAAAAAGgSKLk2So13qlAxqijMsTsKAAAA\nADQJlFwbhYQYVUfESWWsrgwAAAAAvkDJtVlItFvhlQXyeCy7owAAAABAo0fJtVlkiwS1ULE25ZTY\nHQUAAAAAGj1Krs1auJMUp2Kt3FZgdxQAAAAAaPQouTZzJ7RSmPFo7dbtdkcBAAAAgEaPkmuzMGe8\nJClr+zabkwAAAABA40fJtVuUS5K0a9dOWRaLTwEAAADA8aDk2q225IaU52lnYbnNYQAAAACgcaPk\n2i3aLUm1i08V2hwGAAAAABo3Sq7dakdy40yJVmxnhWUcv+KKaj30n+Wau26P3VEAAACAgAuzO0Cz\nFxknSUp1VujH7Yzk4vjkllTqxjfma2lWgTbllGhIpwS7IwEAAAABxUiu3cIckiNWqc5K9srFcdlR\nUKYrXv1RmTuLNCjVrZ835KqwvMruWAAAAEBAUXKDQZRLKVEV2l5QruVZFF003IbsYo16+UftLCjX\nWzcO1O/P7qpqj6U5a5iyDAAAgOaFkhsMouKUElWuyPAQvTt/i91p0Mis2Fagy1/5UWVVNZpyy0k6\nOS1efdvHKS46XN9k7rI7HgAAABBQlNxgEO1WeEWBzj+xrT5dsk3FFdV2J0IjMX9jrq6e9JMiwkI0\n7daT1Tu5pSQpLDREp3VN0rerd6vGw/7LAAAAaD4oucEgyiWV5emaQSkqqazRp0u2252oTp8u3a6x\nby5Qfmml3VHq5PFYeuar1br5rQXNotjNWrVL173+s5JaROiD2wYrLTHmgNdP75akvNIqLdmaZ1NC\nAAAAIPAoucEgyiWV5apv+zh1ax2rKUE4ZTmnuEJ/+s9yzVq1W2PfXKDSyuAabS6pqNat7yzSC7PW\naWbmbs1s4tN0P/5lm8b9a5G6tIrV+785WW3jog4559QuiQoLMZqZuduGhAAAAIA9KLnBIMotleXJ\nWJauGZSi5dsKgm4BqglfrVZpZY3uO6eblmzN163vLFZltcfuWJKkbfllGvXKj5qZuUt/Pr+H2sVF\nafIPG+2O5TcrthXo7veXKKODS+/eMkjxMRF1ntcyKlwDOrg1i5ILAACAZoSSGwyiXJLlkSqLdFF6\nu6BbgGpZVr6mLtiqMYM76LbhaXri0t76fk227nl/ie3Tgn/ZkqeLXpqrrNxSTR4zQDcNTdUNg0/Q\nzxtztaKJbsn03My1io0I06TrMxQbGX7Ec0d0T9LqXUXamlsaoHQAAACAvSi5wSDK5X0szVXLqPCg\nWoDK47E0/tOVindG6M4zOkuSrhyQogdGdtPny3bo4U9WyLLsKbqfLNmmKyf9pGhHqD76f4M1vGuS\nN19GiqIdoXpj7iZbcvnTim0Fmpm5SzcN7agWRym4kve+XEmatYrRXAAAADQPlNxgEO32PpZ5FwgK\npgWoPl6yTYu35Ou+c7oeUKp+MyxNtw5L079/3qKnv1oT0Ewej6Vnvl6ju6YuUXpynD6+fYg6t4rd\n93rL6HCN6p+sz5Zu1+6i8oBm87cXZ61VbGSYxgzpUK/zOybGqGOCU99QcgEAANBMUHKDwd6R3NqS\ne6wLUFXX+PYe2aLyKj0xfZXS28fpsn7Jh7x+3zlddfXA9nrp23V6bc4Gn37twymrrNFvp/6iF75Z\nq8v7J+udmwfJ7XQcct6YwR1UWePRv38KnmnfxytzR6G+XLlLNw5JVcuoo4/i7nV6tyT9tD4nKGYG\nAAAAAP5GyQ0GB5VcY4yuHtiwBagWbMpV+mNf6/++yPTZ9OGXZq1TdlGFHr2wp0JCzCGvG2P0+MW9\ndW7v1nr8v5n6YFGWT77u4ewuKtdVk37UF8t36MFzu+nvo06UI6zuP8IdE2N0erck/fvnzSqvqvFr\nrkB5cZb3XtybhqQ26LoR3VupssajH9bu8VMy/yivqgmaxc0AAADQeFByg0HUgdOVJenivvVfgCor\nr1S3vr1IlmVp0vcb9IcPlh33qO767GJNnrtRV2Qkq0/7uMOeFxpi9OyV6Tqlc4Lu+3CZvlq587i+\n7uGs3lmkSybO05pdxZp0XYbGnZomYw4t3vu7aWiq9hRX6rOl9k/7Pl6rdxbpi+U7NWZIB7WMrv8o\nriRldHApNjJMs1Y1nm2ViiuqdeFLP+iMZ77Thuxiu+MAAACgEaHkBoOo2hK5X8mt7wJUJRXVuvmt\nhaqs8ejT3w7VXSM664NFWbr1ncXHPIJpWZYe/exXRYaF6g9ndzvq+RFhoXpldH/1btdSd0z5RUu2\n5h/T1z2cOWuzNerleaqq8ej935ysM3u0qtd1g9Pi1bVVrCbP3WTb4li+8sKstXI6QjW2gaO4khQe\nGqJhXRI1a1W2PPVcDXt3YbnmrM1u8NfyBcuy9Pv3l2p9dokKy6t02cvztGhz3tEvBAAAAETJDQ6h\n4ZIj9oCSK0lXDzzyAlQej6V73l+iNbuK9NI1/ZSWGKO7z+yiRy/sqZmZu3TD5PkqLK9qcJyZmbv1\n/Zps/e7MLkqMrXsP1oM5I8L0xpgBatUiQre9s0h7iisa/HXrMmX+Fo15Y4HauaL08e1D1Du5Zb2v\nNcZo7NAOytxRqJ825NbrGsuy9P2abBUdw8/NX9buKtIXy3fohsEd5Krj/uP6OKN7K+0prtCyemyr\nVFXj0Y1vLtB1r8/XH6YtVVllYKd7/2P2es1YuVMPjOymj//fELWICtc1//xJX/pploCdFmzK1cjn\n5+iFb9Ye8yJpe4orfH4/PgAAQGNGyQ0WUS6p9MAi1i/lyAtQPTtzjb5cuUsPnddDw7ok7jt+w+AO\nev6qdC3anKerXv1J2UX1L5zlVTX6y+e/qnNSjK4/+YQGfQsup0OvjO6v3JJK3f7vxcf1D2+Px9KT\n01fpgY+Wa2inBE279WS1jYtq8PtclN5ObqdDk+duPOq5e0ewr588X2c+871mrAiOUvXirHWKCg/V\nzad0POb3GNYlUSFGmpV59CnLL81ap5XbC3Ve7zb6YHGWLvnH3IBNGf5+TbYmfLVaF/Rpq5uGpqpD\nglMf3TZY3dq00K3vLNK/ftwUkByB8tSM1dqQXaxnvl6jIU/O0p1TftGizblHnXmwNbdUr363Xhe9\n9IMyHp+pP36wLECJAQAAgh8lN1hEuw4ZyT3SAlSfLd2uF2et05UZ7TW2ju1kLkpvp3/ekKENe4p1\n+SvztDW3tF4xXpuzQVtyS/XIBT0VHtrwPx4927bUk5f11s8bc/Xk9FUNvl7yFu07pizWK9+t17WD\nUvT6DRmKrceesHWJDA/VtYNSNDNzlzbtKTnseXsL7pvzNmlU/2S5nA7d+s4ijfvXQu0oKDumr+0L\n63YX67Nl23X9yR3qXEW6vlxOhzJOcGtm5pG3ElqxrUATv12nS/q208Rr++mNMQO0q7BcF740V/9d\ntuOYv359bM0t1W+n/KKurWL1t8t677vnOj4mQlNvOUkjurXSw5+s1BPTM+s97TqYLdyUq/mbcnX/\nyG6ade8wXXdSB327ercue/lHnffCD3pvwZYDRtG35JTqle/W68KXftApf/9WT0xfJY8lnd2zlT76\nZZv+84t/F34DAABoLCi5wSLq0JIr1b0A1bKsfP1+2lIN6ODSXy7uddgFmE7rmqR/3zxIuSWVGvXK\nPK3eWXTECNvzyzTx2/U6p2drDe2ccMzfyiV9kzVmcAe99sNGfdrARZ/2FFfoqkk/afqKnfrTed31\n+MW9FHYMZXt/o086QWEhRm/O21Tn6/sX3JuGpuqpUSfq0zuG6P6R3fT92myd+cz3emveJtXYUKwm\nfrtOkWGhuuWUht+Le7DTuyfp1x2Fhy3tFdU1uuf9JXI7HRp/QU9J0vCuSfrvnaeoc6sY3f7uYo3/\ndKVfVjwuq6zRuNrF0169rr+iHWEHvB7lCNWr1/XX6JNS9Op3G/S795aootr/06i/WrnziL8cOR6v\nfLderuhwXTmgvTomxujhC3ropwdG6K+X9JLHsnTfh8t10hPf6P4Pl+n8F+fo1Ke+1ZPTV8lIemBk\nN83542n67LdD9Y9r+2tgB7f+/PFKbc7xT1YAAIDGxDT2BXn2ysjIsBYuXGh3jGM3bYy0c7n020WH\nvPT7aUs1ffkO/fzQGSqpXXU2LCREn9wxRAkxR79ndtXOQl3/+nxVVHt0Sd92Kq2sVklljUorah8r\nq1VaUaM9xRWqqPZo5j3D1N4dfVzfTmW1R9e+9pNWbCvUf24frG6tWxz1mu/XZOv+D5cpt7RSz1/V\nV2f3bH1cGfZ393tL9NXKnfrxwRFqsd+o8P4Fd+yQVP35/O4H/NJgS06pHvp4ueas3aP09nF64tLe\n6t7m6N+LL2zILtYZz3ynm0/pqAfP7X7c77d2V5HOfPZ7PX5xL40+6dCp6H+bsUovz16vN8YM0Gnd\nkg54rbLaoyenr9LkuRuV3j5OE6/tp3bHMH28LpZl6e73luiTpds1ecwAndY16YjnvvLdBv1txiqd\n1NGtV6/LaNCewQ3x/ZpsXT95vlpGhWvymAHqf4LLZ++9emeRzn7ue919RhfddUbnQ163LEvzN+bq\nXz9t1lcrd6pH25Y6r3drjezVps7/N7fll+mc575Xx8QYfXDrycc0CwMAACDYGWMWWZaVcbTz+JdQ\nsIhy1zmSK/1vAappC7dq3NuLVFRerX9en1GvgitJ3Vq30Ie3DVbbuCh9uChL36/Zo8wdhdpTXKkQ\nI7WKjVTPdi11bu82emV0/+MuuJLkCAvRxGv6KTYyTL95e5EKyg6/kFNBaZX+MG2prp88X9ERYZr2\nm8E+LbiSNHZIqkoqa/T+gq37jlmWpcc+P3zBlaSU+Gj9a+xAPXdlurbmluqCF3/Q32asCshiTC99\nu06OsBDdchz34u6vU1KMUtzRmrXq0CnLi7fk6dXv1uuKjORDCq7k/e/58AU99PK1/bRud7HOe2GO\nvsnc5ZNVq9+ct0kfL9mue87ocsSCK3mn8N82PE3PXem95/ySiXP9sv9vcUW1HvhouTomOOWKDtfo\n137W7NVHnurdEK98t17RjlDdMLju+96NMRrUMV4Tr+mnNY+P1Ce3D9G4U9MO+/9mu7goPXnpiVq6\nNV/PzVzjs5wAAACNUdjRT0FAxKdJpTnST69IJ916wEt7F6D6y+e/ymNJr4zurx5tGzaa2N4drel3\nneLLxEeV1CJSL4/up6sm/aS731ui167PUEjIgSXyy5U79aePV3gXqzotTb89vbMiw0N9nqV3cksN\n6ODSm/M26cYhqQox0mOf/6o35h6+4O5ljNHFfdtpWJdEPTE9Uy/PXq+Pf9mme8/qqkv7tjvke/KF\nTXtK9MmS7RozuEO9V7g+GmOMTu+WpCnzvfd6Rjm8P+fyqhr9ftpStW4RqT+d3+OI7zGydxt1a9NC\nt72zSDe9tVBtW0ZqWNckDe+aqCGdEhQT0bC/Un7ekKPH/5upM7q30u2ndar3dRf3bafWLSN134fL\nNPr1n3V2z1b603k9fPILGkl6cnqmtheU6YNbByvFHa0xb8zXzW8t1NNX9NFF6e2O67235pbq06Xb\ndePgDoqLPvp91kfbD3qv805so+/XtNc/Zq/XkE4JGpx27LccAAAANGah48ePtzuDT0yaNGn8uHHj\n7I5x7Nr2lbIzpZ8mSs5EqV2/fS8ZY2RZlmatytbvz+qiawY1bNVjO7WNi5IrOlyT526SMdJJHeMl\nSTnFFfrjB8v0zNdr1CHeqcljBuiSvsnHff/tkcRGhGnK/K3q0SZWU+Zv1RtzN+nGIR308Pk96lUk\nohyhOrNHaw1OS9DizXl6+6dk5HeAAAAgAElEQVTN+vrXXToh3qmU+KOXqxqPpe/WZOtv01fpma/X\n6Ls12Vq5vVDb88tUUV0jpyNsX8H/638ztXZ3sSZe20/OBhbHIwkLCdG0RVnq2z5OHRNjJElPTF+l\nWat26+XR/dW5VexR38MV7dCo/slq54pSaWWNZv66Sx/9sk3/nLNBP67PUU5xhWIjwxXvdBzx57qj\noEyjX/9ZrWIj9cbYAQ3+5UayK1rXDEpRtCNUHyzK0pvzNqmq2qM+7eOOa7ruvPV79Minv+rmoam6\namCKnBFhurBPWy3anKfX5mxUy6hw9U059qnLz3y1Wiu2F+jFa/oe84JqhzOkU7ymL9+pz5ft0GX9\nkvf9IgMAAKApePTRR3eMHz9+0tHO457cYFJdKU27QVr9hXTB81L/Mfte8ngsLd9WoBOTW9Z7ZCdY\nWJal309bpg8XZ+n1GzJUXFGt8Z+uVElFje4c0Um/GZYWkHsIq2s8GvbUbBWUVam4orpBBfdgHo+l\nz5fv0N9nrFJWXpmGdUnUA+d2q/Pe4235ZZq2cKumLczStvwyxTsd6n+CS1tyS7VhT8kBCzklxDiU\nlhijhZvzdP3JJ+iR2gWgfKWy2qO+j32lC9Pb6YlLe2v+xlxdOelHXTsoRY9f3PuY33Ph5lx9tyZb\n363O1qraBc6SYiMUFx0uIyNjvL+sMVLt59LuwgqVVFTr49uH1KtcH8nOgnI9OT1THy/ZrjYtI/Xg\nud11/oltGvzftrSyWuc8N0chRpp+16kHlMTyqhrdNfUXfblyl357eifdc2aXBr//nuIKDXlyli5K\nb6u/j+rToGvra8W2Al3yj7k6rWuSXr2uf6P7+wIAAOBw6ntPLiU32FRXSO+NltZ+JV00Ueo72u5E\nPlFeVaNRr8xT5o4i1XgspbeP01OjTjzuctNQr83ZoMf/m3lcBXd/FdU1+te8zXpx1loVV1Tr8v7t\ndc9ZXeR2OvRN5m5NXbBF363JliQN7ZSgqwem6IzureQI85b6Go+lrLxSrdtdrPXZxVq32/tRUlGj\nt28aqKQWkcf9PR/stncWafGWPM26d7hGPj9HkjT9rlN8NmK8o6BM363O1s8bc1VWWSNLlixL8i5O\n7f3ckmQkjRnSQad0TjzyGzbAgk25Gv/pSq3cXqiBqW6Nv6Bng6b2P/rZSr0xd5PeG3eSBtXOOthf\njcfSQ/9ZrqkLtuqaQSn6y0W9FNqA6eoTvlytibPXaeY9w5RWO5LuD//8foP++kWm/npJL13biGZ+\nAAAAHAkltzGrKpemXi2t/1a65BWpz1V2J/KJrLxS3TV1iUb2aq0bh6Q2qBz4So3H0pKt+eqXEufT\nEa780kq9NGud/vXjZoWGGDkjwrSnuEKtW0TqioxkXZ7R3mf3ix6vDxZl6ffTlmpwWrx+3JCjqbfU\nXegaqxqPpfcWbNVTX65SYXm17jmzi24blnbUe6cXbsrV5a/+qOtOOkGPXdTrsOdZlqUJX63WxG/X\na2Sv1nruqnRFhB19WnBReZWGPDlLQzol6OXR/Rv8fTWEx2Pphjfma8GmXH12x9CA/zIJAADAHyi5\njV1VmfTuldKmOdKl/5R6j7I7EephS06pnv9mrUoqqnV5RrKGdUn0633Gx2JPcYUG/HWmLMu76vTD\nFxx5sanGqqC0Sg99vFyfL9uhUzon6Jkr0g+7iFd5VY3OfX6OKqo9+uruU+s1qv36Dxv1l89/1cAO\nbk28tt9RFwib9P16/d8Xq/TJ7UPUp33cMX1PDbG7qFwjn5ujxNgIfXz7EL8s6AYAABBIbCHU2IVH\nSVdPlU4YIn10i7TyP3YnQj2kxEfr6Sv66JXr+mtE91ZBV3AlKSEmQgM7uNUx0ak/nN3V7jh+0zI6\nXC9e3XffvcfnvjBH89bVvd3Qs1+v0YY9JfrbZSfWe9r2TUNT9cLVfbVsW74ufOkHLd2af9hzK6pr\n9NqcjRrSKT4gBVeSkmIjNeHyPlq1s0j3vr9U+aWVAfm6AAAAdmMkN9hVFEv/HiVtnS9d9k+pbT+p\nNNe73dDBH9Xl0uA7pdaHn2oJSFJhuXff4hY+Xt03WK3aWajb/71YG/aU6Lend9ZdIzrvmy6/ZGu+\nLv3HXF05oL2euPTEBr/3yu0F+s3bi7S7qEKPX9RLVwxof8g5U+Zv0QMfLde/bx6kIZ0Cu7XPxG/X\n6Zmv1yguKlwPnNtdl/Vrx2JUAACgUWK6clNSUSS9famUNb/u10PCpOh47xTnUIc0doaU0DmwGYEg\nV1pZrYc/WakPFmVpUKpbz1/VVy5nuM5/4QcVV1Try7tPPebSn1dSqd9O+UU/rNuj0Sel6OHzex6w\nuNiIp2erRVS4Prl9iC0F89fthfrTx8u1eEu+Bqa69deLe3GfLgAAaHQouU1NeaG08iMpNMJbaKPd\ntR/xUkQL754se9ZJk8+WwiK9RTfu0BEloLn7cFGW/vzJCkWGh+qkjm59sXyn3rhxgE7rmnRc71td\n49FTX63Wq99tUP8TXHr52n5KahGp/y7bodvfXayXr+2nkb3b+Oi7aDiPx9J7C7fqyemrVFJRrVtO\n7ag7T+/MXroAAKDRoOQ2VzuWSW+eLzkTvEU35vj+4Q40Ret2F+uOdxdr1c4iXdYvWU9f4bs9az9f\ntl1/mLZMsZFhenl0Pz3y6UqVVtTo63uG2bKi+MFyiiv0xPRV+mBRltrFRemxi3pqRPdWdscCAAA4\nKkpuc7blJ+ntSyR3mjTmMynKZXciIOiUV9Xov8t26OxerRXjoz2C91q1s1C/eXuRtuaWymNJf7/s\nxDrv1bXT/I25+tPHy7VmV7FO6Zyg83q30endk5QU6/u9mQEAAHyBktvcrfvGuwVR277S9R9LDqfd\niYBmpaC0Sve8v0Rbckv1+Z1D67WXbqBV1Xg0+YeN+tePm7Utv0ySlN4+Tmd0T9IZPVqpa6tYFqkC\nAABBg5IL6ddPpGljpNRh0jXvSWFH3scTgO9ZlhX0RdGyLGXuKNI3mbs0M3OXlmYVSJLaxUXpzB6t\ndG7vNhqY6rY5JQAAaO4oufD65R3pk9ul7hdIo96UQn07LRNA07O7sFzfrNqtbzJ3ac7aPaqo9mju\n/aerXVyU3dEAAEAzVt+SGxKIMLBR39HSOU9KmZ9Jn90peTx2JwIQ5JJaROrqgSl67YYBevumQZKk\nNTuLbE4FAABQP5Tc5uCk26ThD0hL/i19/We70wBoRDonxUiS1mcX25wEAACgfpi72lwMu08qzZF+\nfEmKO0EaNM7uRAAaAZfTIbfTQckFAACNBiW3uTDGO225IEuacZ/Usp3U7Ty7UwFoBNISnVq/u8Tu\nGAAAAPXCdOXmJCRUuux1qU269MFNUtYiuxMBaATSEmMYyQUAAI0GJbe5cUR7txOKSZKmXCnlbbI7\nEYAgl5YYo5ySSuWVVNodBQAA4Kgouc1RTJJ07QdSTZX0ziipNNfuRACCWFqSU5K0YQ+juQAAIPhR\ncpurxC7S1VOk/M3S1GulqnK7EwEIUmmJtSssc18uAABoBCi5zdkJg6WLX5a2zJM++X/soQugTsmu\naDlCQ7gvFwAANAqsrtzc9R4lFWyVZo6X4lKkM8bbHAhAsAkNMUpNcFJyAQBAo0DJhTTkd1LeZumH\nZyVXB6n/GLsTAQgyaUlOZe4osjsGAADAUTFdGd49dM+dIHU8TZrxoFS0y+5EAIJMp8QYbcktVUV1\njd1RAAAAjoiSC6/QMOm8p6WaCmn2E3anARBk0pJiVOOxtDmn1O4oAAAAR0TJxf/Ep0kZN0mL35J2\nr7I7DYAg8r8VlrkvFwAABDdKLg407D7JESPNfMTuJACCSGqCd69cFp8CAADBjpKLAznjpVPukdbM\nkDZ+b3caAEHCGRGmti0jtT6bvXIBAEBwo+TiUINulVq2l776E3vnAtgnLSmGkVwAABD0/FpyjTHn\nGGNWG2PWGWPur+P1U40xi40x1caYUQe9VmOMWVL78ak/c+Ig4VHS6X+WdiyVlk+zOw2AIJGWGKP1\nu4tlWZbdUQAAAA7LbyXXGBMqaaKkkZJ6SLraGNPjoNO2SBoj6d063qLMsqz02o8L/ZUTh9H7cqlN\nH+mbx6SqMrvTAAgCaYlOlVTWaFdhhd1RAAAADsufI7kDJa2zLGuDZVmVkqZKumj/EyzL2mRZ1jJJ\nzIkNNiEh0lmPS4VZ0s+v2J0GQBDYt8IyU5YBAEAQ82fJbSdp637Ps2qP1VekMWahMeYnY8zFvo2G\nekk9VepyjjTnGalkj91pANgsLYmSCwAAgp8/S66p41hDbuRKsSwrQ9I1kp4zxqQd8gWMGVdbhBdm\nZ2cfa04cyZmPSZUl0nd/tzsJAJslxUYoJiKMvXIBAEBQ82fJzZLUfr/nyZK21/diy7K21z5ukDRb\nUt86zplkWVaGZVkZiYmJx5cWdUvsKvW7Xlr4urRnnd1pANjIGKO0RCfbCAEAgKDmz5K7QFJnY0yq\nMcYh6SpJ9Vol2RjjMsZE1H6eIGmIpF/9lhRHNvwBKSxS+ma83UkA2CwtkW2EAABAcPNbybUsq1rS\nHZK+lJQp6X3LslYaYx4zxlwoScaYAcaYLEmXS3rVGLOy9vLukhYaY5ZK+lbSk5ZlUXLtEttKGnKX\nlPmZtPlHu9MAsFFaUox2FJSruKLa7igAAAB1CvPnm1uW9YWkLw469vB+ny+QdxrzwdfNk9Tbn9nQ\nQCffLi2cLH39Z+mmryVT1y3XAJq6tESnJGljdol6J7e0OQ0AAMCh/DldGU2Jwymd+nspa4H3A0Cz\nxDZCAAAg2FFyUX8nXiU5YqUFr9udBIBNUuKjFRpiKLkAACBoUXJRfxExUp+rpJX/kUpy7E4DwAYR\nYaFKcUdTcgEAQNCi5KJhBtwk1VRIS96xOwkAm6QlOrV+N9sIAQCA4ETJRcMkdZdSBksL35A8HrvT\nALBBWmKMNu4pUXUNfwcAAIDgQ8lFww24ScrbKG2YZXcSADZIS4pRZY1HWXlldkcBAAA4BCUXDdf9\nQsmZyAJUQDPFCssAACCYUXLRcGEOqe910poZUv5Wu9MACLC9e+VScgEAQDCi5OLY9B8jWZa0+C27\nkwAIsLhohxJiHCw+BQAAghIlF8fGdYLU5Wxp0VtSdaXdaQAEWMfEGEZyAQBAUKLk4thl3CSV7JZW\nfW53EgABlkbJBQAAQYqSi2PXaYQUlyItnGx3EgABlpboVF5plXJLmMkBAACCCyUXxy4kVMoYK22a\nI2WvtjsNgABKS2KFZQAAEJwouTg+fa+TQh1sJwQ0M532biO0m5ILAACCCyUXx8eZIPW4SFo6Rapk\npVWguWgbF6WIsBBGcgEAQNCh5OL4DbhZqiiUln9gdxIAARIaYpSa4NT6bH65BQAAggslF8ev/SAp\nqae04DXv3rkAmoW0JFZYBgAAwYeSi+NnjDRgrLRzmbRtkd1pAARIWmKMtuaWqryqxu4oAAAA+1By\n4RsnXik5YliACmhG0hKd8ljS5pxSu6MAAADsQ8mFb0TEeovuig+l0ly70wAIgLREthECAADBh5IL\n3xlws1RTIS3+l91JAARAx0SnJLYRAgAAwYWSC99p1UNKPdW7AFVNtd1pAPhZtCNM7eKiGMkFAABB\nhZIL3xp0q1SwVVr9hd1JAARAx0S2EQIAAMGFkgvf6nKOFJci/fyq3UkABEBaoncbIY+H7cMAAEBw\noOTCt0JCpYHjpM0/SDuX250GgJ+lJcWotLJGOwvL7Y4CAAAgiZILf+g7WgqPZjQXaAY6scIyAAAI\nMpRc+F6US+pzlbR8mlSSY3caAH7UNi5SkrSrsMLmJAAAAF6UXPjHwN9I1eXS4rfsTgLAj1xOhyQp\nv7TS5iQAAABelFz4R1I3qeNwthMCmrjYiDCFhRjlllByAQBAcKDkwn8G3SoVbpNWfW53EgB+YoyR\ny+lQHiO5AAAgSFBy4T+dz5JcHViACmjiXNHhjOQCAICgQcmF/+zdTmjLPGnHUrvTAPATV7RDeSVV\ndscAAACQRMmFv6VfK4U7pZ8n2Z0EgJ+4nQ7lMl0ZAAAECUou/CsqTkq/unY7oT12pwHgBy6nQ3lM\nVwYAAEGCkgv/GzhOqqmQFr1pdxIAfuCOdii/rEoej2V3FAAAAEouAiCxq5R2urTgdamG+/aApsbl\ndKjGY6monO3CAACA/Si5CIxBt0pF26XMz+xOAsDH3M5wSeK+XAAAEBQouQiMTmdKrlS2EwKaoLho\nhySxjRAAAAgKlFwERkiINOg30tafpG2L7U4DwIfctSWXxacAAEAwoOQicNKvlSLjpG//z+4kAHzI\n7awdyWW6MgAACAKUXAROZAvplHuldV9LG+fYnQaAj7icjOQCAIDgQclFYA0cJ7VoJ818RLLYbgRo\nCpyOUDlCQ5RXyurpAADAfpRcBFZ4pHTag9K2RVLmp3anAeADxhi5nOGM5AIAgKBAyUXg9blaSuwm\nffOYVMO+mkBT4Ip2cE8uAAAICpRcBF5IqDTiESlnnfTL23anAeADrmgHI7kAACAoUHJhj64jpfYn\nSbOflCpL7U4D4Di5nYzkAgCA4EDJhT2Mkc58VCreKf38st1pABwn7skFAADBgpIL+6ScJHU9V/rh\nOak01+40AI6DO9qh/LIq1XhYNR0AANiLkgt7jXhYqiyW5jxtdxIAx8HldMiypIIythECAAD2ouTC\nXkndpT7XSPMnSflb7E4D4Bi5nQ5JUh735QIAAJtRcmG/0x6QZKRvn7A7CYBj5IquLbnclwsAAGxG\nyYX9WiZLg8ZJS6dIu1banQbAMdg7kptLyQUAADaj5CI4DL1HimghffOY3UkAHIO46HBJTFcGAAD2\no+QiOES7paG/k9bMkDbPszsNgAb630guC08BAAB7UXIRPAbdKkXHSwteszsJgAaKCg9VRFgII7kA\nAMB2lFwED0e01OlMaf23kqfG7jQAGsAYI7fTwT25AADAdpRcBJdOZ0hludKOJXYnAdBArmiH8hnJ\nBQAANqPkIriknSbJSOtm2Z0EQAMxkgsAAIJBvUquMSbNGBNR+/lwY8ydxpg4/0ZDs+RMkNr0kdZ/\nY3cSAA3kcjqUV8rCUwAAwF71Hcn9UFKNMaaTpNclpUp612+p0Lx1GiFtnS+VF9idBEADuKLDGckF\nAAC2q2/J9ViWVS3pEknPWZZ1t6Q2/ouFZi1thGTVSBu+szsJgAZwRTtUUFal6hqP3VEAAEAzVt+S\nW2WMuVrSDZI+rz0W7p9IaPbaD5QcsUxZBhqZvXvl5pcxZRkAANinviX3RkknS/qrZVkbjTGpkt7x\nXyw0a6HhUsdh3sWnLMvuNADqyVVbcvOYsgwAAGxUr5JrWdavlmXdaVnWFGOMS1KsZVlP+jkbmrO0\n06WCLVLOOruTAKgnd7S35HJfLgAAsFN9V1eebYxpYYxxS1oq6Q1jzDP+jYZmrdMI7+M6piwDjYXL\n6b2LhRWWAQCAneo7XbmlZVmFki6V9IZlWf0lneG/WGj2XB0kd5q0bqbdSQDU0957cvNKGckFAAD2\nqW/JDTPGtJF0hf638BTgX51GSJt+kKrK7U4CoB5cTFcGAABBoL4l9zFJX0pab1nWAmNMR0lr/RcL\nkNTpDKm6TNryo91JANRDZHioosJDWXgKAADYqr4LT02zLOtEy7Juq32+wbKsy/wbDc1eh6FSqIOt\nhIBGxO10KJfpygAAwEb1XXgq2RjzH2PMbmPMLmPMh8aYZH+HQzPncEopJ3m3EgLQKLic4YzkAgAA\nW9V3uvIbkj6V1FZSO0mf1R4D/CtthLR7pVS4w+4kAOrBFe1QLqsrAwAAG9W35CZalvWGZVnVtR9v\nSkr0Yy7Aa+9WQkxZBhoFt9OhfKYrAwAAG9W35O4xxow2xoTWfoyWlOPPYIAkqVUvKaYV++UCjYQr\n2sHqygAAwFb1Lblj5d0+aKekHZJGSbrRX6GAfYzxTlne8K3kqbE7DYCjcDsdKiqvVlWNx+4oAACg\nmarv6spbLMu60LKsRMuykizLuljSpX7OBnh1GiGV5Unbl9idBMBRuKLDJUl5TFkGAAA2qe9Ibl3u\n8VkK4Eg6nibJcF8u0Ai4nA5JUl4Ji08BAAB7HE/JNT5LARyJM15qmy6tm2l3EgBH4Y72llzuywUA\nAHY5npJr+SwFcDRpI6SshVJZvt1JABzBvpFcpisDAACbHLHkGmOKjDGFdXwUybtnLhAYnc6QrBpp\n43d2JwFwBG5KLgAAsNkRS65lWbGWZbWo4yPWsqywQIUElJwhRbRgKyEgyMXtXXiK6coAAMAmxzNd\nGQic0HAp9VRp/SzJYqY8EKwiwkIVExGmXBaeAgAANqHkovHoNEIq2CrtWWt3EgBHEBcdznRlAABg\nG0ouGo+0Ed5HVlkGgprb6WB1ZQAAYBtKLhoP1wlSfCdpw7d2JwFwBK5oByO5AADANpRcNC7t+ku7\nfrU7BYAjYCQXAADYiZKLxiWhs1SYJVUU250EwGG4oh2srgwAAGxDyUXjEt/Z+5izzt4cAA7L7QxX\nSWWNKqpr7I4CAACaIUouGpeELt5HSi4QtFxOhyQpv5RthAAAQOBRctG4uDtKMtKeNXYnAXAY7mhv\nyeW+XAAAYAe/llxjzDnGmNXGmHXGmPvreP1UY8xiY0y1MWbUQa/dYIxZW/txgz9zohEJj/Sussxe\nuUDQiqstudyXCwAA7OC3kmuMCZU0UdJIST0kXW2M6XHQaVskjZH07kHXuiU9ImmQpIGSHjHGuPyV\nFY1MfGdKLhDE3LXTlXPZRggAANjAnyO5AyWtsyxrg2VZlZKmSrpo/xMsy9pkWdYySZ6Drj1b0teW\nZeValpUn6WtJ5/gxKxqThC7ee3I9B/+xARAMXM5wSYzkAgAAe/iz5LaTtHW/51m1x/x9LZq6hM5S\ndZl3KyEAQce1755cFp4CAACB58+Sa+o4ZvnyWmPMOGPMQmPMwuzs7AaFQyOWULuNEFOWgaAUHhqi\n2Mgw5TFdGQAA2MCfJTdLUvv9nidL2u7Lay3LmmRZVoZlWRmJiYnHHBSNzN5thCi5QNByOx2UXAAA\nYAt/ltwFkjobY1KNMQ5JV0n6tJ7XfinpLGOMq3bBqbNqjwGSM1GKaCnlUHKBYOWKdrCFEAAAsIXf\nSq5lWdWS7pC3nGZKet+yrJXGmMeMMRdKkjFmgDEmS9Llkl41xqysvTZX0l/kLcoLJD1WewyQjPFO\nWWavXCBouaLDGckFAAC2CPPnm1uW9YWkLw469vB+ny+QdypyXddOljTZn/nQiCV0ljZ8Z3cKAIfh\ncjq0Zlex3TEAAEAz5M/pyoD/JHSWirZLFUV2JwFQBzfTlQEAgE0ouWic4mtXWM5ZZ28OAHVyOR0q\nq6pReVWN3VEAAEAzQ8lF48QKy0BQczu9e+VyXy4AAAg0Si4aJ3eqZEIouUCQckV7Sy5TlgEAQKBR\nctE4hUVIrg6ssAwEqX0juSVVNicBAADNDSUXjVd8Z+7JBYKU2xkuScplujIAAAgwSi4ar4Takuth\nYRsg2MRF7x3JpeQCAIDAouSi8UroIlWXSwVb7U4C4CBxUbUjuZRcAAAQYJRcNF4JtdsI7WHKMhBs\nwkJD1DIqnNWVAQBAwFFy0Xjt20aIxaeAYOR2OhjJBQAAAUfJReMVHS9Fxkk5bCMEBCNXdLjyS1ld\nGQAABBYlF42XMd7RXPbKBYISI7kAAMAOlFw0bgmdKblAkHJFO7gnFwAABBwlF41bQmepeKdUXmh3\nEgAHcdWO5FqWZXcUAADQjFBy0bjF166wzH25QNBxRTtUUe1RWRV7WQMAgMCh5KJx27fCMiUXCDZu\nJ3vlAgCAwKPkonFzdZBMKCUXCEKuaIckKa+EFZYBAEDgUHLRuIU5JHcqe+UCQcjtrC25LD4FAAAC\niJKLxi++s5Szzu4UAA7iouQCAAAbUHLR+CV0lnLWSx4WtwGCibt2ujL35AIAgECi5KLxS+gi1VRI\n+VvsTgJgPy2iwmWMlEfJBQDg/7N31+FR3dkfx983rkTRENwdglOoUqi7u7fbbn/brcvuVlYqW9u6\nu5eWemlLW4pToLgVd0mCxZOZ+/vjkGKRSZjJRD6v5+EZmNyZ+w0k4Z57zvccqUEKcqXuS90zRkjN\np0RqldAQh8TocLJVriwiIiI1SEGu1H2lY4Q0K1ek1kmKjVB3ZREREalRCnKl7otJhuhkdVgWqYWS\nYyK0J1dERERqlIJcqR9SO0GmOiyL1DZJsRHqriwiIiI1SkGu1A+pHZTJFamFkmMU5IqIiEjNUpAr\n9UNqJ8jdCvk7gr0SEdlH6Z5c13WDvRQRERFpIBTkSv2QsqfDcpZKlkVqk+TYcIo8XnKLNMdaRERE\naoaCXKkfSjssq2RZpFZJjIkANCtXREREao6CXKkfklpDSJhm5YrUMsl7glx1WBYREZGaoiBX6ofQ\ncEhup0yuSC2TFLsnyFXzKREREakhCnKl/kjpqD25IrVMcqzKlUVERKRmKciV+iO1I2StAE9JsFci\nInuUlitvzysO8kpERESkoVCQK/VHaifwFsOONcFeiYjsER8VRmiIQ3ZuYbCXIiIiIg2EglypP1L3\njBFS8ymRWiMkxCE5NoKsHJUri4iISM1QkCv1R0oHe8xSkCtSm6TERpCpIFdERERqiIJcqT9ikiEm\nVR2WRWqZxvGRZOaoXFlERERqhoJcqV9SO0GmOiyL1CYpsRFkaU+uiIiI1BAFuVK/pHaAbUvAdYO9\nEhHZIzUukszdKlcWERGRmqEgV+qXlgMhPxs2zw/2SkRkj5S4SPKLPeQVabyXiIiIBJ6CXKlfOo0C\nHFj2bbBXIiJ7pMbZrFxlc0VERKQmKMiV+iWuCbTsD0u/CfZKRGSP1LhIADK1L1dERERqgIJcqX86\njYaNs2H35mCvRETYJ8jdrSBXREREAk9BrtQ/nY+zR5Usi9QKKXvKlbNyVa4sIiIigacgV+qfJt0g\noRUsVZArUhskx5buyZmCpnEAACAASURBVFUmV0RERAJPQa7UP44DnUfDyp+hOD/YqxFp8KLCQ4mP\nClMmV0RERGqEglypnzofByX5sHJCsFciIti+3G05yuSKiIhI4CnIlfqp9WEQEQ/L1GVZpDZIjYsg\nS0GuiIiI1AAFuVI/hUVAh6NsX67XG+zViDR4KbGRZOWoXFlEREQCT0Gu1F+djoOczbBpTrBXItLg\npcZHkKlMroiIiNQABblSf3U8FpwQjRISqQVSYiPZnldMiUeVFSIiIhJYCnKl/opNgZYDYan25YoE\nW2p8JADZ6rAsIiIiAaYgV+q3zsfB5nmwc0OwVyLSoKWWzsrVvlwREREJMAW5Ur91Ps4e1WVZJKhK\nM7nalysiIiKBpiBX6rfUTpDU1rosi0jQpOzJ5GblKsgVERGRwFKQK/Wb41g2d9UvUJQb7NWINFh/\nZHJ3q1xZREREAktBrtR/nUaDpxBW/BTslYg0WPGRYUSEhpCpTK6IiIgEmIJcqf9aD4XIBO3LFQki\nx3FIjYtQJldEREQCTkGu1H+h4dDxGFg2Drya0SkSLClxkdqTKyIiIgGnIFcahk7HQe422DAr2CsR\nabBS4iLUXVlEREQCTkGuNAwdjwEnVCXLIkGUGhdJlubkioiISIApyJWGIToJWg3RKCGRIEqJiyAr\npwjXdYO9FBEREanHFORKw9F5NGxdCNvXBHslIg1S47hIijxedhWUBHspIiIiUo8pyJWGo/Px9rhM\n2VyRYEiJiwAgS/tyRUREJIAU5ErDkdIeUjrCUu3LFQmG1LhIADK1L1dEREQCSEGuNCydR8PqSVCw\nK9grEWlwUmItyFUmV0RERAJJQa40LB1GgrcY1k4L9kpEGpzUeCtX1hghERERCSQFudKwtBwAIeGw\nZnKwVyLS4CTHROA4KlcWERGRwFKQKw1LRAyk9YM1U4K9EpEGJyw0hKSYCGVyRUREJKAU5ErD03oo\nbJwNRXnBXolIg5MSa7NyRURERAJFQa40PK2HgbcE1s8I9kpEGpzUuEhlckVERCSgFORKw5M+EJwQ\nlSyLBEFKXARZucrkioiISOAoyJWGJyoBmvVUkCsSBKlxkWTuViZXREREAkdBrjRMrYfB+l+hRBfb\nIjUpNS6C3YUlFBR7gr0UERERqacU5ErD1HoolBTAxt+CvRKRBiUlLhJAJcsiIiISMApypWFqNcQe\nNS9XpEallga5aj4lIiIiAaIgVxqm2FRo3EX7ckVqWEpcBIDGCImIiEjAKMiVhqv1UFg7HTwlwV6J\nSIPReE8md5syuSIiIhIgCnKl4Wo9DIp2w5b5wV6JSIOhTK6IiIgEmoJcabhaD7VHlSyL1JiYiDBi\nIkLJVCZXREREAkRBrjRcjVpAUlsFuSI1LCUuQo2nREREJGAU5ErD1nqYBbleb7BXItJgpMZFkqly\nZREREQkQBbnSsLUeCvnZkLk02CsRaTBSYiNVriwiIiIBoyBXGrY/9uVqXq5ITWkcH6FMroiIiASM\nglxp2JLaQHwL7csVqUEpsZFk5xbi9brBXoqIiIjUQwpypWFzHMvmrpkCri64RWpCalwEXhe25ymb\nKyIiIv6nIFek9VDYvQm2rwr2SkQahJS4SACychXkioiIiP8pyBVpPcweV2tfrkhNSN0T5GbuVvMp\nERER8T8FuSKNO0NMivblitSQ1LgIADKVyRUREZEAUJAr4jjQaog6LIvUkD/KlTVGSERERAIgoEGu\n4zijHcdZ6jjOcsdx7ijj45GO43yw5+PTHcdps+f5No7j5DuOM2fPr+cDuU4RWg+DHWtg5/pgr0Sk\n3kuMDic0xNGsXBEREQmIgAW5juOEAs8AxwHdgPMcx+l2wGFXANtd1+0APA48tM/HVriu22fPr2sD\ntU4RYJ95uVODuw6RBiAkxCE5NoIszcoVERGRAAhkJncgsNx13ZWu6xYB7wOnHHDMKcAbe37/MXC0\n4zhOANckUrZmPSEiXiXLIjUkNS5SmVwREREJiEAGuWnAun3+vH7Pc2Ue47puCbATSNnzsbaO4/zm\nOM4Ex3GGB3CdIhASCq0Gq/mUSA1JjYsgU5lcERERCYBABrllZWRdH4/ZBLRyXbcv8FfgXcdxGh10\nAse52nGcmY7jzNy2bdshL1gauNZDIXMp5OhrSSTQlMkVERGRQAlkkLseSN/nzy2BjeUd4zhOGJAA\nZLuuW+i6bhaA67qzgBVApwNP4Lrui67r9nddt3/jxo0D8ClIg1I6L3et9uWKBFqK9uSKiIhIgAQy\nyP0V6Og4TlvHcSKAc4HPDzjmc+CSPb8/E/jRdV3XcZzGexpX4ThOO6AjsDKAaxWBFn0hLEolyyI1\nIDU+kvxiD7mFJcFeioiIiNQzYYF6Y9d1SxzHuQEYB4QCr7quu9BxnPuBma7rfg68ArzlOM5yIBsL\nhAFGAPc7jlMCeIBrXdfNDtRaRQAIi4CWA9R8SqQGpMRGAJCVU0RsZMD+KxIREZEGKKBXFq7rfg18\nfcBzf9/n9wXAWWW8bgwwJpBrEylT62Ew4SEo2AlRCcFejUi9lRofCcC2nEJapcQEeTUiIiJSnwSy\nXFmk7mkzDHBh7bRgr0SkXkuNtSA3S82nRERExM8U5IrsKy0DnBBYPzPYKxGp11LjrVxZY4RERETE\n3xTkiuwrIhaadIMNs4K9EpF6LfmPPbnK5IqIiIh/KcgVOVBaPwty3QPHOouIv0SGhdIoKkyzckVE\nRMTvFOSKHCgtAwp2QLamVokEUmpcJJm5KlcWERER/1KQK3KgtAx73DA7uOsQqedS4iJUriwiIiJ+\npyBX5ECNu0JYtPbligRYalykGk+JiIiI3ynIFTlQaBi06KMgVyTAlMkVERGRQFCQK1KWtAzYNBc8\nxcFeiUi9lRoXyfa8Yoo93mAvRUREROoRBbkiZUnrB55C2LIw2CsRqbdS4iIB2K7mUyIiIuJHCnJF\nyvJH8ymVLIsESuM4m5W7TSXLIiIi4kcKckXKktgaYlLUYVkkgEozuVlqPiUiIiJ+pCBXpCyOY9lc\nZXJFAiZ1T5CbqUyuiIiI+JGCXJHypGXAtiVQuDvYKxGpl1L2lCsrkysiIiL+pCBXpDxpGYALG+cE\neyUi9VJ8ZBgRYSHK5IqIiIhfKcgVKU+LfvZYG0uWczOVYZY6z3EcUmMjyFQmV0RERPwoLNgLEKm1\nYlMgqU3tC3J3rocneoHrgehkSGxlv5JaW8OsxFb2mNoJQnQfS2q31PhIZXJFRETErxTkilQkLQPW\nTg/2Kva3ZaEFuAOuAtcLO9bY3uHfv4OSgr3HHXkPHH5r8NYp4oOU2AiNEBIRERG/UpArUpG0DFgw\nBnZvgfimwV6NyVpuj0fcadnmUq4LOVthx1oYcwVs/C046xOpgtS4SJZsVum9iIiI+I9qGUUqkpZh\njxtr0bzcrOUQlQgxyfs/7zgWiKcPgKY9IHtFcNYnUgUpcZFk5RThum6wlyIiIiL1hIJckYo06wVO\naO3al5u1AlI6WFBbnpT2kL0SvJ6aW5dINaTGRVDk8bKroCTYSxEREZF6QkGuSEUiYqBpt1oY5Lav\n+JiU9uApsiZVIrVYalwkgJpPiYiIiN8oyBWpTFqGBbm1oZyyKA92rbdMbkVKP166f1eklkqJiwAg\nS2OERERExE8U5IpUJi0DCnZa+W+wla6hskxucvv9jxeppZTJFREREX9TkCtSmdLmU7WhZLk0M1tZ\nJje+GYTHKpMrtd7eTK6CXBEREfEPBbkilWncxQLG2hTkJleSyXUcSGln+3dFarHkmAgcB7apXFlE\nRET8REGuSGVCQqFFn9oR5GavhPjmEBlX+bEpHZTJlVovLDSEpJgIZXJFRETEbxTkivgirR9smgcl\nQc42ZS2vPItbKrk97Fgb/DWLVCI1LoItuwqCvQwRERGpJxTkivgiLQM8hbB1YXDXkbW88qZTpVI6\ngOuBHWsCu6ZgWT8Tpj0X7FWIHwxqm8LPS7exJis32EsRERGRekBBrogvakPzqbxsyMuqvOlUqdJg\nuL7uy534KHx7J+TvCPZK5BDdcFQHwkIdHv1uWbCXIiIiIvWAglwRXySkQ0wqbJgdvDX8MT7I1yC3\nHs/K9Xpg9WTAhQ0zg70aOURNG0Vx+bC2fD53Iws37gz2ckRERKSOU5Ar4gvHsWxuMDO5pRlZX4Pc\nmGSISoTsepjJ3TQXCvcEQ+tmBHct4hfXHN6ehOhwHv52abCXIiIiInWcglwRX6VlwLalULArOOfP\nWg5OCCS18f019bXD8uqJ9tgoDdZND+5axC8SosO5/sj2TFi2jakrsoK9HBEREanDFOSK+CotA3Bh\n05zgnD9rOSS2grAI31+T0gGyVgZuTcGyaiKkdITOx1kDKq8n2CsSP7h4SBuaJ0Tx4LdLcF032MsR\nERGROkpBroiv0vrZY7BKlrOW+16qXCqlPexaD0V5gVlTMHiKYe1UaDsc0gdBUQ5sXRTsVYkfRIWH\nctMxnZi7bgfjFm4O9nJERESkjlKQK+KrmGRIahucINd1bU9udYJcgO2r/L+mYNk4xwLbNsMhfaA9\np5LleuP0fml0aBLHw+OWUuLxBns5IiIiUgcpyBWpirSMijssF+yCRZ/5vwvz7s1QnFv1IDe5dIxQ\nPdqXu/oXe2wzHBJbQ1wzWKsgt74ICw3hlmM7s3JbLh/PWh/s5YiIiEgdFBbsBYjUKWkZsOBj2LUJ\nGjW357JXwbJv7dfqyeAthpgUuHEORDXyz3lLOySXZmZ9VR9n5a6aCI27Qlxj+3P6QGVy65lR3ZvS\nt1UiT/zwO6f2TSMqPDTYSxIREZE6RJlckapIy7DH396C7/8OTw+E//WBb++wbOuQP8Epz0JeFkx+\n0n/nLc3EVjWTGxkPcU39H+RumgefXA352/37vpUpKbKAtu2Ivc+lD4Ida+zvX+oFx3G4fXQXNu8q\n4I0pq4O9HBEREaljlMkVqYrmvSAkDH76F4SEQ5th0P9y6HQsJLfbe9yKH2HqMzDgCmjU4tDPm7Uc\nQiOhUcuqvzalg39n5RbmwEeXQPZKy1iP/o//3rsyG2ZBcZ41nSqVPsge182AbifX3FokoAa3S+GI\nzo159ucVnDuwFQnR4cFekoiIiNQRyuSKVEV4NJz9Jpz1Oty2Ei7+DAZfu3+AC3D038D1WDDsD1kr\n7Bwh1fiWTW7n3z25391tJdqthsKMFyGzBvf7rp4IONB62N7nmveyGwAqWa53bhvVhV0FxTw/oR6V\n24uIiEjAKcgVqaouJ0D30yreb5vUBgZeDXPehS0LD/2cWcurvh+3VEoHyN0GBTsPfR1LvoZZr8Ow\nG+HsNyAsGr6759Df11erfoFmPazTdamwSGjR1zK5Uq90a9GIU3q34LXJq9iyqyDYyxEREZE6QkGu\nSKAMv9n2xH7/j0N7H6/HMqdV3Y9byl/Np3Zvgc9vgGY94ci7Ia4JjLgZln0DK346tPf2RXGBBbJt\nRhz8sfSBsGmOHSP1ys3HdsbjdXnih9+DvRQRERGpIxTkigRKTLIFusu/h5U/V/99dqy1js3VDnL3\nvC57ZfXX4LoW4BblwukvW/YUYNB1NsZn3F3gKan++/ti/QzwFO6/H7dU+iDwFFmgK/tbOBaWjQv2\nKqotPTmGMzPSGTN7PQXFnmAvR0REROoABbkigTTwGkhIt07MXm/13qM0A1vdIDepLeAc2r7cma/A\n79/ByPuhSZe9z4dHwbEPwNZF8Nub1X9/X6yaCE4ItB568Mf+aD6lfbn7KdgFn10P4+8P9koOybHd\nmlJU4mXm6hru5i0iIiJ1koJckUAKj4Kj/gab5sKCMdV7jz/GB1VzT254lAXa1S1Xzvwdxt0D7Y+G\nAVcd/PGuJ1sjqB//5Z99v+VZPRGa94aohIM/FtfYGmxpX+7+5r4HRTl2E6IoN9irqbaBbZMJC3GY\nvCIz2EsRERGROkBBrkig9TzL9rGOv796e0azlkNkI4htXP01pFSzw7KnGMZcaYHyKc+U3d3ZcWDU\nv2028C//rf4aK1KUB+tnQpsySpVLpQ+yTK7rBmYNdY3Xa92vw2PA9dqNljoqNjKMvq0SmbxcQa6I\niIhUTkGuSKCFhMDIB2DnWvj1paq/vrSzsuNUfw2ls3KrGgD+/KDtcz3pf9CoefnHtegDfS6Aac8d\neoOrsqybZvuS25bRdKpU+kDrIr19lf/PXxet/Mm+do7a0/16w6zgrucQDeuQyvwNO9mRVxTspYiI\niEgtpyBXpCa0PxI6HAO/PAJ52VV7bfaK6u/HLZXc3kqJ87J8f82aqTDpMehzIXQ7ufLjj/4bhEbY\n/mN/WzURQsKg1eDyj/ljX65KlgGY8ZJl/wdcCYmt6nyQe1iHVFwXpq6owtewiIiINEgKckVqyjH3\nWSOgiY/6/priAtix7tCD3NLX+5plLdgFn15twdFxD/r2mvhmMPwmWPKlBaX+tHoitOhnI5nK07iL\nlXWr+ZSNnFr2LWRcap2w0zLqfJDbOz2R2IhQJqlkWURERCqhIFekpjTrAX3Ot32S29f49prtqwDX\nD0Fu6axcH/flTvwv7FwPp71YcWB5oCE3WJOrcXfafF9/KNwNG2aXPTpoXyGh0LK/75ncnG3wztmw\nbemhr7G2mfmKdaLuf7n9OS3DRlHlbAvuug5BeGgIg9ulaF+uiIiIVEpBrkhNOvJuCz5+fMC340uD\n0uR2h3bexFZW7pvtQybX64G5H0Cn46DVoKqdJzwaRt4Hm+fDnHf2Pu+6FmQt/hJ++g+8dz481R+m\nPlP5e66ZCq6n4qZTpdIHwZaFvnV5/vF++H0cTP5f5cfWJUV5MPst6HoSNGphz6Vl2OPG2cFblx8M\n65DK6qw81mXnBXspIiIiUospyBWpSQlpMOgamP+xb9ncQx0fVCo0HBJb+5bJXfUL5GyGXmdX71zd\nT7dgc/wD8O2d8PqJ8FBreKInfHABTHgIsn6HiFgYd5cF1BVZ/QuEhO/dc1uR9EGAa52YK7JprgWC\nEfE22il/h8+fXq03/yMo2GFfZ6Wa97abK3W8ZPmwjqkATNEoIREREamAglyRmjbwauuUPOu1yo/N\nWgGxTcqeDVtVKR0ga2Xlx8370Pa2dhpdvfM4Doz+D+Rnw8zXoDjfAt8THoMrfoC7NsANv8IV31t2\n9rPrYfWk8t9v1URoOQAiYio/d1qGBXMVlSy7rgXf0Ulw7jtQkg/zKgm06wrXtXL4pj2g1ZC9z0fE\nQuOudT7I7dgkjsbxkUxaruZTIiIiUj4FuSI1LaEldD4eZr8JJYUVH5vlh87KpVLaVz5GqCgPFn9h\n3ZTDo6p/rrQMuOV3uHM9XDUeTnoCBlwB6QMs4AIIi4Bz3oLktvD+BZD5+8Hvk78DNs+rfD9uqahG\n0KR7xc2nFn0GaybbaJ12h0OLvhaM14f5umunwpYFe2+k7CutnwW5dfjzdByHwzqkMmV5Jl5v3f08\nREREJLAU5IoEw4ArbJzPos8qPq50Rq4/pLSH4jzYvan8Y5Z9A0W7oWc1S5X3FZMMoWEVHxOdBBd8\nZOXU75wJuQeUoa6ZAq7Xt/24pdIHWrlyWY2vigvg+79ZINzvEnuu/+WwbXH96Mo8/QWISoSeZx38\nsbQMyN8O21fX+LL8aViHVLJyi1i6ZXewlyIiIiK1lIJckWBoe4Q1k/r1lfKPKdgJuVv9l8lN9qHD\n8ryPIL4FtDnMP+f0RVIbOO992L0Z3jvXyptLrZ4IoZFWruyr9EEWqG9dfPDHpj5tDbBG/2dvAN7j\nDCvPnulD+fiBXLf2ZEZ3bbQsfL+Lyi7tLm0+VcdLlod1SAFQl2UREREpl4JckWAICYH+V8C6adaJ\nuCylM239lsmtZFZubhYs/x56nmHjeGpSy/5w+kuWgf30GvB67flVEy0zW5XS6fSB9nhgZnbXJpj4\nGHQ50cqUS0XEWpOthZ9CXrbv5/F64Z2z4MOLa0egO/NVy3r3v6LsjzfpCmHRNo6pDmueEE27xrGa\nlysiIiLlUpArEix9zoewqPKzuX8EuX7K5DZKs/OVl8ld9Cl4S6DXOf45X1V1OxmOfcBKuMffawHn\nlvnQdkTV3iepjTXrOrD51Pj7wFts5zhQxmXgKYS57/l+nrnv2k2BxZ/DsnFVW6O/lRTCrNetWVhy\n27KPCQ23Lst1PJMLcFiHVKavzKaoxBvspUh9lLOtfnVcFxFpgBTkigRLTDL0ONO6GZc11zV7BeBA\nUjlBS1WFhFiJdHY5HZbnfWgdeJv28M/5qmPIDZaJnPwkfP5ne66qQa7jWDZ33bS9z62fZQHs4D+V\nPXO4WQ9oOdD3BlR52fD93+01KR3gu3vAU1y1dfrTwrGQuw0GXlXxcWkZNj4pmGv1g2EdUskv9vDb\n2u3BXorUN14PvDISPr482CsREZFDoCBXJJgGXAHFuWXPis1aDonph9bl+EDJ7crO5GavsvLeXmcf\n3JW3JjkOHPcwdBgJS76E8Bho0a/q79NqsDVY2r1lz8igOyy7O+KW8l/T/zKb37tmcuXv/+M/rYnT\niY/Bsf+011W0vzrQZrwIKR2h3ZEVH5fWz0YmlbVfuQ4Z3C6FEMf3fbm7C4q574uFfLdws7oyS8WW\nfQvbV8GK8XW+SZuISEOmIFckmNL6WRD368sHZxCzlvuvVLlUSgcLaA/sPDz/Y3vseaZ/z1cdoWFw\n1mvQvA90HGmjhqoqfZA9rp9hn9v6GXD03yEyvvzXdD/N5hHPfLXi994w244ZeA0062klwm0Ph5//\nU7U9vf6yYRZsmGlZ3JBKfqTXk+ZTCdHh9GqZ6PO+3Ie/Xcprk1dz9VuzOPqxCbw1bQ35RWV03xaZ\n8SLEpAIO/PZOsFcjIiLVpCBXJNgGXAmZS2H1pL3Pua5/Z+SWSulg+1J3rN3/XPM/hNbDILGVf89X\nXZHxcOV4OPP16r2+eW8IjYAVP8EP/7A/97mg4teER0Pv82DR5wePMirl9cBXN0NcEzjyTnvOcWDU\nv6FwF/zySPXWeyhmvAQRcbb2yiS1gejkOh/kgu3Lnbt+J7sLKi69nr12O29PX8NFg1vz1Hl9aRQV\nxt/GLmDog+N59LulbN1dUEMrllpv21JY+TMMvg7aHwVz3il7FJmIiNR6CnJFgq3H6Tbb9NeX9z6X\nu82CpmQ/dVYuVdqpOXufDsub5kDmMitVrk1CwyrPTJYnLBJa9LWM664NMPoh394r4zK7CTCnnAzO\n7Ddh42wrUY5K2Pt8sx7Q9yLLAmVWMKLJ3+a8C/M+sAA+qlHlxzuOVQ/Utg7LOVth4xy7ueBjp+ph\nHVLxeF2mryw/e17s8XLXJ/NpGh/FbaM7c1LvFoy9fhgfXjOE/m2Sefqn5Rz24E/c+tFclm7W3F2K\ncq3xWx3fs11tM160kWUZl9oorl0b7EaZiIjUOQpyRYItPBr6Xmh7UHdvtuf83Vm5VFljhOZ9ZFnP\nbqf491zBlj4QcKH76dB6iG+vadIFWg21TsXeAzr35mZZh+bWh0HPsw5+7ZF3W/fq7/92qCv3zbTn\nYOx1Vip99N99f11aBmxbDIU5gVtbVb11Grx4ODzSHv7VDJ7KgDdOhrHXw0//tpsLB4y+6tc6kajw\nkApLll+dtIolm3dz3yndiY8KB8BxHAa2Teali/vz481HcPaAlnwxbyOjnviFFyaUM16roZj4mI3E\nemUkbF1S9dfv2gTf/a38sWi1WcFOmPOezc2OTYXOx1vVw29vBntlIiJSDQpyRWqD/pfb+J7Zey6o\nSptD+WtGbqnYxhARvzdg8HpgwcfQ8ViITvLvuYKty0mQ2hlG3l+11/W/zDpQr/5l/+fH3wuFu+GE\n/5bdnCu+KQz/Kyz9GlZOqPayK+W6Fvh9ewd0PRnO/wAi43x/fVqGzdPdNDdwa6yKLQthywIYeLVl\n3AdcaXudi/NgxY8w4WHrtP3C4fuVkUeGhTKwbUq5zafWZefx+A/LGNmtKaO6NyvzmLapsfzz1J5M\nveNoRnZryiPjljJvfQMdHeP12raF1M6wfQ28MAKmPO1buW5JIUx63G5OTPkfjLs78Ov1tznvWRPA\n0g7lYZE2Tm3J1+VvXxARkVpLQa5IbZDS3vaAzXwNPCUW5IaE+3+PrOPYuUqD6FUTIGdL7StV9odW\ng+CGGdahuiq6nmwZnH0bUK371W5ADL4OmnQt/7WDr4eEVnaRH4i9fF4vfHM7THjIsv9nvmYX41VR\n2q26tuzLXTAGnFA4/HYYfC2M+hec9Tpc+QPcvBj+tg2u+B6KciyA2sew9in8vjWHLbv231frui73\njF1AiONw38ndK11CUmwEj5zZi9S4SG76YE7tbUrlYyl3taybbnv1h98M10+HDkfDd3fD6ydas7ry\nLBsHzw6GH+6F9kfCgKvs50p1MsHB4vXCry9BywFWzl+q30W2fWFeGd3vRUSkVlOQK1JbDLgSdm+E\nZd9YEJrcDkJC/X+elPZ79+TO+xAiE6DjKP+fp64Kj4I+58OSr2wEkdcDX/0V4ptbIFbZa0feC1vm\nl7+vt7o8xVaePOMFmyd88tO2b7mq4hrbzZPaEOS6rgW57Y6wEtGyhIZb6XnPs6zJVs62Pz40rIO9\n5sBs7lfzNzFh2TZuPrYzLRKjfVpKYkwE/z2rNyu25fLgN7VwxFJRHjzdH768KTA3UOZ/aCO7upxg\njdXOfRdOfd6y7M8Ns5s++wbZmcvh7TPh3bMhJAwu/ATOfQeOuMP2tc540f9rDJSVP9rP3IFX7/98\n0+52U2j2W4G9wSAiIn6nIFektug4Chq1tAZUgeisXCqlg2Vs8nfA4i+g28n+ncVbH2RcZuXjc962\ni/vN86yDckUjiEp1Px1aDoTxD1h5sz8UF9heyXnvw1H3WOOrQ5lnnJZRO5pPbZhts0h7nFH5sYff\nDiUFMPmJP57q1rwRSTHh++3L3ZlfzH1fLKJnWgKXDm1TpeUc1jGVy4a14Y2pa5iwbFvlL6hJiz6z\nQGzmqzDmSv82xxI3NgAAIABJREFUhyopgoWf2j7U0tJ3x4E+58GfpkL6AAuu3z7dOhB/9zfL3q6b\nbt8X102xzC/YzYqeZ8Lc9+1nTF0w/UWbo93t1IM/1u8i28NeG24KiYiIzxTkitQWoWHQ/1IbYZG5\nDFLaBeY8ye1tT+a056wEtNc5gTlPXZbaAdoMh19ftWC13RE2R9cXjgOj/wO5W22f4qEq3A3vnGl7\nfY//L4y49dACXLAgd+fa/bKi5VozFd49JzAzgBeMsaZnXU6o/NjUDva1+usrlmEHQkIchnZIZfLy\nTNw9mbaHv11CVk4h/z6tJ6EhVf97un10Fzo2iePWj+ayPbeoyq8PmNlv2PfuyPth4SfwwYVQnO+f\n917+A+RvL/tnQUJLuGgsnPAorJ0Gzwy0svFe58CfZ8GQ6y3bvq+BV9v+Vn9XMwRC9kr4/TvrqFzW\nTO4eZ0BY9N5+CSIiUicoyBWpTfpebHtxXU9gM7kA056FRmk2H1cO1v8y2LXeGiAdX06zqfK07G/l\ntVOe3n8mcVUV5liX4TVT4PSX9jbFOVRpGfa4sZJsbnE+jL0Wln0LU5/2z7lLeT0WrHU8FqITfXvN\niFvBU7RfNvewDqls2VXIim05zFqznXemr+XSoW3p2TKhgjcqX1R4KI+f04fteUXcM3bBH8Ez2ass\n81zdUmFPsXXoro5tS2HtVMi4BIb9H5z4uO2Ffecs/3TJnvcBxKTYntqyOI5tp7h2Egy8Bq78EU59\nxsqay9KiD6QPsvLyA7uU1za/vmLbQvpfXvbHoxKg+6mw4BMbsSQ1x1NiNyFERKpBQa5IbRLf1MqH\nIYBB7p4MceEuKyus7iza+q7LSZDS0cpkUztW/fVH/8OCgx/urf4aJjxkgeg5b/m3OVjz3uCEVF6C\n+ct/rZy4aU+Y9nz1g7SyrJ0KuzfZnGhfpbSH3udZYLJrE2BBLsDPS7dx1yfzaZEQxc3HdjqkpfVI\nS+Avx3Tiq/mbGDtng3UPfv0EeOlIeLgtvH+BlbhuW1r+Xs3ifFg10bpDv3kKPNgKHusCmb9XfUGz\n37SbX73Ptz/3vxxOf9Fufrx1qmVhq6tgp93E6HHGwRnZA6W0h+MfhpYZlb/vwKth+ypY/n311xZo\nRbnw21vWbK5R8/KP63sRFO2GhWNrbm0CE/8LT/W3DuwiIlVUja4lIhJQw2+2vWzN+wTm/aOTLGuT\nlwU962FXZX8Ji4A/z6z+6xPTLes24SELzDqOrNrrtyyybHvfi3wr562KiFho3LXiIHfrEpj8pK39\nsJtsD+bkJ+DYB/yzhgVjrNFRp9FVe92IW2xv8qTH4PhHSE+OoVVyDI9+t4z8Yg8vXdyf2MhD/6/t\n2sPb89OSrfx97EKOyFlD0q4NcMSdsGuDjYha8qUdGNcM2h1u84pjG1vwvmaK/d16iwEHmvaAPhfA\nb2/bLNrTnvN9ISWFMOdd6HK8NQ0r1ets+/v7+DLrgHzRp+VnViuy+Avb6+zvnwXdToHv7oHpL0Cn\nWtrYbt6HFuQf2HDqQK2HWqn4b29D3wtqZm2BlptpW1byt1uwX5Sz5zF3759xrQFZs541v76SQutP\n4Xrgx3/Cee/V/BpEpE5TCkektmnaHS76pGqzT6uqSTe78G7WI3DnELth0bgLfH6jXUz7ynXh61us\n0dUx9wVmbWn9LBArKxPp9cKXf7GvwWP/CY0779PdeOuhn9tTbFmxzsdbwF0VyW2t+/Ws12HnBsC6\nLOcXexjVvSkjuzU99PUBoSEOj5/ThxC3mOKf/4vbIsOy+ic/BX+ZBzfOgZOetABo+Xj47E/w7lm2\nX9X1wJA/wfkfwu2r4bpJNl+5/2VWGrx9te8LWfwF5GdDv0sO/ljXE21OcvZKeO042Lm+6p/ovA8h\nqa2V2PtTaLhlnFeMr172OtBc176em/WEVoMrPtZxbGTX2inWVbo++PZOmPgoLBprn1f2SgtsI2Ls\ne6zVYCjYZTcqgmHhWMjdBu2OtH4Ea6cHZx0iUmcpyBVpiE573i7AJbDCIuGUZyFnc9UuFud9AGsm\nwzH3QmxKYNaWlmFZnO1lzECd87ZlJEc+sHe0z+G3237YSU8cfHxVrZxggZsvXZXLMvwWC1ImPgrA\nKX1a0KVZPPf6MBO3KtKTY3ix32qaeLbwXepF++/LTm5rzYrOeg1u+R2unQyXfAl3rLUZvyPvtwzm\nnv3GW3cVMKXpebghoVX7O5z9ho18alfOftn2R1kWN2crvDraOrP7atdGWPWLZYUPtZlZWTIutcZi\ntXGc0JrJsHWhZXF9+dz7nG/znH97K/BrC7TsVbDgYxh6A9y2Ev4y37poX/kDXPyZjYI6/UX7nl/5\nM6z4sebXOOMF27JzzlvW+Xr8fRrjFEiB3jvvuvr3kxqnIFekIUpoCQlpwV5Fw9AyA4b+2fZVLh9f\n+fH52y0gTutvjcgCpbT51IGjhHK22YiY1sMse1WqdD/szL37YattwRhr6FM6dqaqklrbaJfZb8KO\ntQxul8K3fxlB8wTfZuL6zOth4LrXWRfRnj/PasqijbvKPi4kxKoi2g4/KDO9M6+Yh75dwohHfuL8\nD9bxddjReH97548sdIWyV1oQ2vfiivfOtxoMl3xhTdLePdtm6vpiwRjADdy2hbgm1pV8zruWFaxN\npr9gWzd6nuXb8fHNrEna3PesIVJdNuV/Ntt48PUVHzfgCkhoBd//o2YbiK2fZVUmA6+2apbDb7Ob\nEr78/JSq8Xqt98J/Wto86ED54v/gheHV7+vgulYBU98D5bxs2LzAOtn//r01vJv9Jkx9Bn5+CMbd\nbePexCcKckVEAu2IuyC1k/1HX9ns3B//afulT3wssE3BmnS10SgH7sv97h7bk3fi4wdnuEbcYvOD\nJz1W/fMWF9h+1q4nWaa7uobfbOvbk80NiEWf4WT9TuKoO2kUHcEpz0zi2rdm8eOSLZR4Kr7ozysq\n4ZmflnPYwz/y/IQVjO7ejIfO6MmLnpPxeDxMf+decgsrCZZmv2kNwnzZB9qiD5z5ms3S/f7vvn1+\n8z6AFv1sPFOgDLzGymDn1qI9lTvXw5KvoN/FEF6FGyP9LoKcLTZyqK7avdn2Fvc5v+JmW2Dfn0fd\nY3PCF4ypmfWBZXEj4uymGlipfmJrGH9v7e/WXZfk74D3z4cfH7Cbjp/fYA31/G3jHKtI2Tx/z024\nKnYpd10L7p7sDT/92//rCybXhU1zYcIj8PIx8HA7eH4YvDrKRgd+fBl8/mcYdxf8/G+7OffRpTDv\no2CvvE5Q4ykRkUALj4JTnoFXjrUA5MRy5udumG2dgwddYx2QAyk03M6xb5C7coI1dRpxq+3DPVBy\nW2ugNOt1a6qV0LLq513+vXX2rm6pcqmElnbxO+s1a4yV1ObQ3u9ApRmO1M7E9z2DT9oU8ObU1Xz6\n2wa+XbiZJvGRnN6vJWf1b0n7xnv3zxeVeHlvxlqe+nE5mTmFHNO1CTcf25muzRsBcFzPs5j34hf0\n3vIpZz52HH89bRhHdSljH7GnGH57BzqOgkYtfFtzu8MtOzftGWvo1fGY8o/dutguOkc/VIW/lGpo\nmWFVCTNehAFX1Y5u7jNfBVzof0XVXtfxWCud/e0tawRWF019xm5UDfs/347veRZMeQp+vN86/x/K\njSlf5Gy17FX/yyDKvmcIi7Bg+5OrbOxYzzMDu4aGYMtCm7W9Yy0c97BtLfjoMvjmVqsIOewv/jvX\n+PutamLUv+Gz6y1IO/fdyru5gwWB391jP9NSO8EvD9v/p8Nv9t/6alrhbtsGsGycZWtzNtvzLfrB\nEXfYDeiIOKtiiIzf8/s4iIi3fg9vnWY9IBLSrCeElKsW/G8jItIApA+EIdfbBfbKCQd/3OuBr/5q\nJZ5H3lUza0rLsLvInmLLsH55kzUhqugCYsStduHxy3+rd84FY6wLcZsR1Xv9vob/1fZJVnctFVn2\nre3ZHH4zhITQKiWGe07sxrS7juaFizLo1TKRlyau5OhHJ3DGc1N4f8ZaPp61nqMe/Zl/fL6Q9o1j\nGXPdEF6+ZMAfAS5Ao6hwMs5/gMgQDxe5X3D56zO5/p3ZbN1VcPD5c7fabNyqOPrv1jn7s+ut9K08\n8z60v7uqjHCqrkHXWIa5sr2duZkw5kp446TAzaTNy7abNJ2Os7L3qggNhz7n2cXp7s0BWV5A5W+3\nnz/dT4fkdr69JiQERt5rwdDM1wK6PMD+bbzFB3e87nEmNOkOP/3Lfl7VJyVFNXu+eR/CS0fbtoZL\nv7Lvz7BIOPsNu/n4wz8sY+qP0uBVv1jzueE3W/XACY9aJcQXf6n8/V3X1jL1abtB9qdptrVi/P0w\n9dlDX1tNK8qz8XMPtbUbDIs+s60mpz5nfR2u/smC3G6n2Fae9IEW8Cam202C0DD7dzrnbatseP/8\n2tnUrxZRkCsiUlOOusdGkXx+AxTm7P+xWa/Bxt/sbndUQs2sJ62fjY/ZuggmPQ7ZK6xMuqISzsR0\nC7x+ewu2r6na+QpzYOm30O1U+w/7UDVqYRmfOe/a/lV/cV3LGCS1OSjjHB4awqjuzXj5kv5MvfMo\n7jq+Czvzi7njk/nc8tFcEmPCeePygbx/9WAyWieX/f6pHXC6n87ZfMc9Rzbh+8VbOPqxCbwzfc3e\nMuhZb0B8c+hQxdFT4VHWNCgvyzpkl9c9e/5H0P7I6o0dqqpup1oGdMYL5R+z6DN4ZpB11V09CT69\n1v+lqa4Ln91g+4OPuL1679H3Isum1Kbya1/NeMlKxw+7qWqva380tB1h3xOB3FvtKbYgvP1RB88m\nDwmxGzjZK6vW/Ku2B8SZy+G/HSzoC3QpdkkRfH2bZcTT+sE1v+zfWTw0HE5/yXoxTHjIMqiHEui6\nLvxwHzRKgwFX2nP9L4fD77DmhuPvr/i14++zMXb9r4DjH4GQUAsIu54M4+60qqe6ZMKDtlVn4NV2\nc+G2lXZjoc/5Vfs5HJMMF3xkNynfOdNuDkqZFOSKiNSU8GgrW96xDn64d+/zOVvhh/vtQvJQy3ir\norT51NwPbJ9tz7PsArMyw2/ek0F9pGrnW/oNlOT793M87Ca7OJtQxbVUZMV4u+Fw2F8rDMabxEdx\n9Yj2fH/TCD7901DeumIgX9xwGId3aoxTWcfeEbfgFOVwZcT3fPt/w+nRIoG7P13A4Y/8zNvjJuEu\n/8EuNqtzM6B5Lzjqbgsc55XRRX3dNNi5zu8Np3YXFPPMT8v5at4mvN59Lo7DIuxmxO/fH9z9OTfL\nyiQ/vNhK0K/5xTpTL/7cgip/+vVlWPoVjLyv+tsBUjtCqyEw4+XqN9HJWgEfXASb5lXv9dVRlGtz\ncTuNrvroOMexUWZ5Wda0KlAWfw67N9k+7rJ0GgXpg60Bjy/N1eZ+AA+2tv2ctZGnBMZea5/LrNes\nVDhQjZV2bYI3TrQbTUNusC7a8WVskwgJhZOeskBs6tPw1c3VD76XfAUbZlp2ct8bp0fcYeXRkx6z\nPaYHcl3rTTHpcci4DI7/797+EKFhcMYr9nX81V/tBmddsGkuTHnabpKN/je0Ocy3cu3yJLeF8963\nipL3zoXifP+ttR5RkCsiUpNaD4FB18KvL1nGCmyfbnEeHP9oYEa5lCepDUQn236n8GjLIvuiUQu7\nIz/n3aqNrFkwxu7qpw+q1nLLFN/M7vTPex/GPwAlhYf2fq5rAXOjtL2NbyrhOA59WyUxvKMPwW2p\nJl2t+da052kX7+HdqwbxwkUZtEqOIXPiK7guPLRlAIs3VTNzNvRGC8a+vsVuquxr3gcQHgNdTtjv\n6eVbd/Pu9LXszK969mvayixGPzGRR8Yt5fp3ZzP6yV/2D3YzLrML6F9f3vuixV/As4Ps8ch7bIRN\n0252Ed77fPj5Pxao+8Pm+RbsdBwFg/90aO818gGb4frBhVX/esvNsuzL4s/hzVNsb2RNmPWGje06\n7K/Ve31aPytznvpM4Eq1p79oP5M6llO94Dg2Vi1nc8VVAcUFlhn99Grb1zv16dpZ3jrlf7D+V8tO\nDv2zfW+Mu9v/ge6GWfDCCOvae+ZrMOpfFQdYISG2T3fY/1k3/c+ur3pHca/HMrUpHe17eV+OY//X\ndT4Bvrnd9mDv6+f/wMT/WmO4E8powBgWAWe9YWPVPru+ZpuiVYfXA5/fCDEpcOwD/nvf9AFWtbN+\nJnxytZqylUFBrohITTv6b7b39bPr4fcfrPRx2I3QuFPNrsNx7OIVLFNTlZKpw26yGai+ZnPzsmH5\nD7YH1N/Nh468C3qdaxdGzw+Hdb9W/73WTLZM57C/2MVUIA2/BQp3woyXcByHUd2b8d6VA7g+cRrL\n4wfw2iIvxz05kXNemMq3CzZV2tF5PyGhNg/b9cLY6/ZeAJUUWklwlxOsmQmWgf3nl4sY/cRE7vp0\nPoc/8hOvTV5FUUnl5yso9vCvrxZx3kvTCA91+PjaIfzvvL54Xbj+3dkc9+REvp6/CW9cMytb/u1t\n29855koLEuObwdU/w+G37r3wdhxrztZygJUtH2rGsyjXssXRSXDqs4d+Iyl9gL3P2inWMd3XoKQ4\n37IuuzbC6S9DWBS8cbI1AQukkiJrHtV6GLSyG0zz1+/kP98sZmdeFW5oHHWPzcqeEIBmZZvm2vfd\ngKvsa7c8rYdYA7BJj1Ock830lVn7f19sX2OdaWe9Zjd6/m+u3Uwad5f/bpj4w5aFtu+12ynWSGvk\nA3bzc9ozVuXjr0B36xJ4+wy7iXnVj77vwS/N3h95N8x9Fz65smql33Pfg8yl9n9dWdUooWFw5itW\nLv3pNbZ3F+DnB+3rq++FcOKT5f9fER5lzataDYExV8HiL31fW02b/jxsmgPHPWQ/g/yp2ylw7D/t\nptkPPnbVb0BC77333mCvwS9efPHFe6+++urKDxQRCbbQCCsZnPasdQtt1NLusB9K+VJ1hcdYtuPw\n26t28R8ZBwU77GKyxxl2l7oi8z+y/UjHPWx7Tf0pLAK6nmhdfBd/YReKBTvtAii0ioHq53+28sHT\nng/8v0d8M+uovegzGHiVrfX3Hwid9Sopp/ybC08aRXJsBBN/z+S9Gev4eNZ6cgo9tEyKplG0D2uL\nToS4pjD9OYhsZI1Mln5rF60j78eb1I4xszdw1ZuzmLwik3MGpHP38d1YnZXLW9PW8uW8TTRrFEX7\nxrFlZqgXbNjJpa/N4PtFW7lwcCuevyiD1imxdG4WzwWDWtOhSRxTV2Ty9rS1fLtgM+3bdyB9xbu2\nl27rYtubd9oLZY+yCQ2zrOu8D/Z01D3roBnEPvvyJmv2dt57lkH3h6bdgBD7uw0NswCyIl4vjLkC\nVv4EZ75qwUan0TD3fauI6DQaYiv5HqquOe/A/A/hpCcguR0Tlm3j0tdmMG1lNl/P38TAtsk0iY+q\n/H1ikm3/38zXbP2Vfc9XxfgHrDnZ6S9aAFORJl1h2nN8MHMD102JZ/ySrfRumUiTzRPg7dOtwdaZ\nr8Lga20ecOfjYNUE+7prM6J6XeH9qaQI3j3Lulxf8JF9XTsOdDjGms1Nfw5wbe72odixFl4/0caQ\nXfrVwfucK+M40GaYdfed9qxt4ehyYuU/F4sLrBw/pYNVB5X3/0pouN1sW/K1NRzbuc5uxvS5AE5+\nuvKboaHh1vF75QQre27R1+a5V9ey72DNFGjWy38VVdvX2FaM9kdZwB+ISq2WAyAv04Lp2MZ7b1zX\nY/fdd9+me++9t9J5V8rkiogEQ5vDbN+TtwSOexAiYoKzjq4nwUkV3DGvyLC/2Kzdnx+s/NgFY6yj\na/M+VT+PrzqOhOunwYAr7KLs2SE2qsFX636144f+uWrzUw/FiFutjLS0c+3sNyAmFTofT2JMBFeP\naM8vtx3Jixdl0K5xHI//sIzDHvqJ816cxphZ68krqqSMsM8FdmE6/j7YssiCnZhU5kf244znp3DL\nR3NJT47ms+uH8Z/TezGkfQpvXzGI1y4dQGiIw7Vvz+KcF6YxZ92OP96yxOPlmZ+Wc9qzk9mRV8zr\nlw3gn6f2JCZib8YmNMTh5N4t+O6mw3ny3D6UeL2c/62X38J6kRXTjiUnfUbBsFsqvmCObwrnvmOB\n1QcXVq8L7fyPLXs8/Gbb8+5Ph98Gvc6x/YMHllwe6Pu/WbZl1L8s+wI2n/iSL+z3b5xUtdJ/X3k9\nMOkJu3BvfzRjf9vAFa//SuuUWF66uD9FJV5Of3YKH81cV/l7AYy4zb43KmoaVFW5WXYTrNc5dmOm\nAsu27Obir/MZ6xnK6UWfc9+RKWTuyuOX52+Ed8/Gm5AO1/xsN71KhUfb/sVGLSyTHoi/56r45REr\nnz/pfxCbuvf50jLe0sZPVe15sK/cTBs1U5QLF35iezira+if4cQnrBLnrdNsvm5FZr4Cu9ZbaXll\nQV10Elz4sY3KmfmqbRE5+Snf/z+KjIcLx9iNjw8utL4P1TH1WZvh+/kNdqPTHw3LXNf2DTsh1lU6\nUFuRHMdGwXUcBV/fat3fBQDHDdQm9xrWv39/d+bMmcFehoiI7zwlsG0xNOsZ7JVU3w/32kX0sBst\nE5E++OAy391b4LEuVp57VA01gVk92S5WslfY3q6RD1R6Ac2758C66fCXBX+U8taIN06GbUvgiu/g\nf/1s1FQ5e7fWb8/j09kb+Hj2etZk5REbEcrxPZtzZkZLBrZNLntPcG4mPDsYYhvjZq1gauKJXLDx\nDFJiI7njuC6c3jeNkJCDX1fi8fLBzHU8/v0yMnOKOLl3C84b2IpHxi1h9todnNCrOf88pQdJsZVn\nyz1ely/nbeSp8b+zfJuNBwoLcejSPJ6eaYn0bplAr5aJdGwaR3joARe4C8bAx5db05aTn/L9YjF7\nlZWvN+0Gl37tn47eByoptH+/TXMsW9ay/8HHTH8BvrnNylFHP3jw+rcutoxbaARc9pXv4318sfBT\nm0t61hu8nN2Lf361mMHtknnx4v40igonM6eQG9/7jSkrsjhvYDr/OKk7UeEVlAuDNX76+d9wxQ9W\nun2oJj5mN2Gum7onQ36w7NwiHv9+Ge/OWEtMRCj3DIni7Omn43Q/jZKdmwlbO5H3S47g1UbXce8Z\n/RnaPvXgN8laAa+MtKqGK3/YP8CsKRtmwcsjLaA/7bmyj/F6bIvBvA/s59awG6t2jsLd9vW0bSlc\n9KmVePvDwrG2zaBxZwss45sdfEzBLniytzV2u3is7++dtcJGjPW/vOJy9fLkZsHbp1nZ+4jbrLmV\nL+/j9doNqKlP2w3f1E4w8VG7IXb2m4dWXjz/Y6veGP2QVRUEWmEOvHac3UBpNcRu9HQ5seqj0uoA\nx3Fmua5bxg/bA45TkCsiItWWvx0+vsLKAb0lEB5rWeoOR1uJVkoHG13yza3wp+nQpEvNra0435qY\nTHnKyrjaH23nb7znV0L63ozBprnWnOXIe2x/aE1aNdE6nzbuajc9bphlWb4KuK7LzDXb+Xjmer6a\nv4mcwhJaJcfQOz0R13VxAVxwcfF6oUfuFG7YfA8AZxQ/QN8hx3DjMR1pFFV52fPugmJemLCSlyau\npLDES6OoMB44tQcn927he6Otfda9aWcB89bvYO76ncxbv4N563eyu8Ay0pFhIaQlRePxupR4XEq8\nXko8Ltd43uVqPuEBzyV8GHICvdMTGdAmmQFtkujTKnG/LDJgWd9XR9lNjmsnQWKrKq2zSnIz4eWj\nLWt25fj9LyqXfGWzMbucYBfN5V14b14Ab5yENzyaX494mwlbY8gttL+TA/+OS//YtVkjTuzd/ODP\nvZTrwgvDcYvzeaj9Gzz/yxqO79mMx87us18g6/G6PPrdUp79eQU90xJ49oJ+pCdXUFlSmAP/62MB\nwaVfHVqGylNiQVFyW7j04H2VRSVe3py6mifH/05ekYcLBrXiL8d0Ijk2wjr//rpnb/MJjzIlfjR3\nfjqfNVl5nDsgnTuP70rCgWX962ZY1rxpD8ui12QFTXE+vHC4jXG6bkrFN908JbYPduGnVQuSigus\nFHr1ZCvP7zTKP2svteJHeP9CiGsMF409OEP8078tC33VTzVfNlucD1/dYuOJOhxj45BiyhnjBnaD\n6tNrbTvEwGtg9H/s+3POe3aDNKkNXPBh9W465WXD0wPsZ8EV31cvcK+O3Cxryrb4S5vzDnbDoctJ\nFsQ37lyzzS0DREGuiIjUnMLdFqytGG8XQqVzaxNagbfYujj/aUpw1rZhtl18bZ5vnVlLhcdas6/G\nXSDzd8hcBn+ZX3nG199cF14dbY13Wh9m2bwqyCsqYdzCzXwyewMbtufDnmsYBwuQ7BGuzX+ZXs5y\nuHwcHZo2qvIyN+3M59PfNnBa3zSaJ/ivnNvrdVmTnfdHwLt5ZwFhoQ5hISGEhzqEhTqEO3Duqrvo\nuGMSb7R/lA+zO7Jk8y5c1zLC3dMSGNgmif5tkunXKonEyQ8QPu0pis94A6fbyYQ4Do5zcMDoN9uW\nwcvHQEIaXD7O9rmvnwWvn2DZyUu+LDOg8nhdFmzYyaTlmaxbNJ07tt7KbjeGCzx/Z1dksz/6D5Ve\nq5VesXm9LrlFHuKjwjijX0suHNyKDk3i93/z33+Ad87gg+a3c/uq3lw4uBX3ndyD0DKy9gDfLdzM\nzR/NJcRxeOLcPhzZuexGdK7rkjvpeeLG30FJ38sI63EytBpa+V7asiz6HD68CM552y7CgV0FxcxZ\nu4Pf1u5g7JwNrMrM5fBOjbnnhK50bLrP55ibCT8+YDNY91TD5Bd5eGL8Ml6euIrk2AgeOKU7o7o3\n2//ffdHntk+yshsP/jbubssYXviJ3QSsjKfYsvBLvoTB19t88sadyz/e64GPLrG+BKe9CL3P8dvS\n97N+FrxzhlUeXPjJ3pFUOdvshkXHkTb/NRhc1/b3fnObZZrPfgtalLFFJn+H3XxaM8lGlg29cf/g\nb/Vk+OACwLEGV1XNho/9k2Xir55Q9ZFd/pK1wr52Fn8J62cAUJDQjsUJI2jedzTNuh8evG1Sh0hB\nroiIBE/b+bMnAAAUr0lEQVT2Kgt2V/xozTyOusf2ygZb/nYr49u62B63LbYOpDmb4Yi74Ijbg7Ou\nFT/afrczXrFuq3Kwwt3wyrHWoKhJV4qT2rM+tCXzC5owMTuJcVvi2FUSzoiQubwZ8RBvlxzNPSX7\nf805DoQ4DsmxEbRIiKJ5QjTNE6NoseexeUI0LRKjaBIfVW4wWJYSj5ctc8fR/IsLWZs4kI9Sr+e6\nVddTFBLL2z1epiQ6lciwECLDQogIC6HE4/Lr6mymrMj6Y2RT1+aNOKtFJhcv+zMhsSmEDL3BAonQ\nCNu7vM/v3dBwFmSH8tYSl7GLdlPk8TKobTIXDm7NqO7NiAgLwfPKaHZuXM6g3Ee54Zhu3Hh0h0qD\n/NWZuVz79iyWbtnNNSPak54czaYdBWzcmc+mHQVs2pnPpp0FeEqKeCL8WY4NmUmEU0JJaDRO2+GE\ndhxpAZyPDYDc106gJGsVnxz2JbPX5zB77XaWb8vBde3fqmdaAjeN7FRuwF2eBRt2ctvH81i0aRcR\nYSE0iY/c8yuKpo0iOWbXJwxf8SjrO11C9oj7iY0MIz4yjLioMKLDQ/1/M2TNFHjteCvHPfEx319X\nUmRZxfkfgeuxrFyvc63Z375zbl0XvrgRZr9pJfGDr9vvbbxel205hazLzmP99nzWb89jXXY+63fY\n4468Io7s0oSz+6czpF1KmdsX9rN1yd49vxd8aF2Sv7ndqnaun171Jlf+tn6W3cjIy7QxRH0voKjE\nS1iIQ8juDfD2mfZz5NTnoNdZZb9H1grbp7tjrTXCquSmgdfrsjIzl8Kl4+k+/mLmtL6cb5tdQ2ZO\nIdt2F5KZU8j23CIaRYfTMimatMRo0pKiSUuMIS0pmpZJ0aTERvj9ay+3sITvps1h44wx9Nw1kSEh\niwh3PJQQRnHzfkR3PMKqr9IH1lwvikNUK4Jcx3FGA08CocDLrus+eMDHI4E3gQwgCzjHdd3Vez52\nJ3AF4AFudF23wp3UCnJFRKTainKt03QwS7myVlhpXD0oJwuYnRssG7ZtCWQut46s7L2OKYxtgVO4\ni92RzRjT702KQyJxXRevC949jx6vl8zdRRa47Sxg0458cos8+50mxIGUuEiaNrLAqEl8JE0a2WPT\nRlF4XZfft+xm2ZYclm3ZzcptuRR5vJwb+iMPhr9MAREUEMFF7j9ZWtKMojLGP7VIiOKwjqkM62C/\nUuMi7QOlmbL87T79lXgjE8gOa8rigkSWFyaxI6IZvdKTOHrNE9xfchHtT7qVCwb5vi8vv8jD3WPn\n88nsDYA1EWsaH0nzxGiaJ0TRYs9jcmwEvy3fwPZFP9K3aBZHhc6llbPF1pTUlpCOI6FJN2sOFBnP\nDk8ka3JCWbELlmx32bl5NQ9l/pkHi8/lec/JJESH07dVIv1aJdG3VSK90xN9KqcvT7HH+//t3XmU\nXGWdxvHvr7q6qrt6X5KQpJPuhCSEgIAJBkScwwQdQVTwCLLNKI4ezyiOOmecAWeOjuMsDmc2h6Oj\nImFwYVRkcXJc0QRnBCEQgkoWhJBOJ52E3vettt/8cS+h00knxKSriurnc849de/bb1e/VXnzdj/1\n3vteHnx6P7s6h+kcHKdzaCLYBscZHE/zqeg3eH/0R/w48zq2ZVvY7fPZ7Qto4zSi8cSh0LugtpzX\nL23gDcsaWTW/+vgBcKqJYfjyG3CHPdc8ROsQtHaP0tYzQmv3CO19Y8SjERoqY9RXxGmoiAVbZZyG\nymC/PNlD1a4N1O96gMrebbhF6J57EXsXvYP2ub/P6Tu/zNmt63lk/k1saPhj+kdT9I+lGBxL0T+a\nonckeUQfbKyM01RXzqL6BGXRCA/t6GBgLEVTXTnXrFnE1ec3sbD2GKGnf28QdAf2Bwuq/egWOO8G\neMftR1TNZoNLKE7kg6OTMTCW4rndrSz42c0s7HuCH8Qu4xPDN3BOWRdfifwTFYyRuvrrJFYeZ0Z9\ntDcIy3t+EdyB4JJPHjY+D42neOT5bjY928nPn+ticGiIn8RuwYHLkreRLYnTWBlnTlXwWJeI0T+a\nZH//GPv7xhiaOHzhwLLSCCvmVbF6cR1rmutY3VzHgpqy3yn47jw4yD2b2/je0wcYnkiz8rQqbryw\nmYsXxXn85z9g+LcPs5btvCayhwjZ4AO0ptdByxvhrKtO3Ur0MyDvIdfMSoDngDcD7cCTwPXuvmNS\nnQ8D57j7n5jZdcA73f1aM1sFfAtYCywAfgascPfM1J/zEoVcERGRWSY1Fnw40P1cMDPT/XwwK3/5\nP7/i67/dncHxdDBLGc5adgwGYagjDEcdgxP0jEwccfvShbXlrJhXyYp5VYe2ldv/ldItXw0W52m+\nCAj+yE9mssGWzpJ1Z05lfPo/XtMTwSI+mWS4pV7ez6aDr490BSG/fx/078UH9pHpbSOaDhb26vFq\ntr7z57z5vBO/rYp7MCuViJUwpzJOdOpiYJObmsnyRGsvP3jmINueeZpzJ7awLvoMF0W2E/OJY/6c\nlMX44Zs3cfbyJSxtPPqtqmbCeCpDZ/8o8U1/TU3bQ5SNHjz0NccYiM2jI7aY/SVN7B6vomM4RZYI\n8ViUlsYqWuZUcfrcauory7CS0mAGrDTBsMfoGCvhwAjsGzbahuCCtq9wyfAPuT71aTZnXj7duCoe\npaWxgsX1CZKZLL0jSXqGJ+gZTh4RfiY73fbzzpJHuKrkUZqsm3EvpcxS3JO+lL+zD1BbHqemvJSa\nRCk15aXUlpdSXxGjqa6cproEi+qD2cPy2OGnaY+nMjy0o4N7n9zHoy90A3Dxskbeff4i/uCsecSj\nJYfq9Y+m6BtNMtxzkBU/u4ma/h2kI3H+8zX3sjddR/9okr6wTv9oiv7RJA7UlJdSn4hRVxGjLhGj\nvqKUuvC4sTLOksYESxorqUuUvqK+kExn2dMzwq7OYXZ1DrPjwCDbDw6wr3cMgBIyfCrxADdlH+TF\nxBnUjLczlI3xnolb2GXNrF1Sz7qVc7n0zHksaZzmFmXpJHz/4/Cre/Blb6Kv9hx2DJXxRFeUxztL\nOJipJhlv4HUrFnFz5puc+cJ69l/5HSrPuJTq8ugxX8fAWIr9fWNh6A1m2bcdGODX+wYYSwWR57Tq\nMtY0Bx/8rGmuo7mh4tBaBZmsk8pkSU9aw+C5jmH+e3MbW/f2E49GuOKc+dx4QTOrF9ce1pbu4Qnu\n/EUr9z+2g7PTO7hhbhsXR3dS3rM9uE/5mpuO+/7nSyGE3NcDn3H3t4THnwRw989NqvOTsM5jZhYF\nXgTmALdOrju53nQ/TyFXREREZko6k6V7OEnH4DgOLJtbSWV8mkWfUuO/2zWqJ8sdxvro2r+LsUg1\ni0/P4UJvBO/R5jDwbnxmH5XZAVbVwxl1xtIaaK7IsDCRoSYyhiWHg5neMy7PaRuPKjkSfFjS83xw\nhkBPeI1+zwvBQlEnaWPdu/n1mZ+guaGClsYKWhoS1B/j1NSJdCYMvUm6hyfIuofXqEeIRY3SkghR\ng+quJ6l5/kEisQQll/0jZfETvC/4NPb1jnL/1na+u6Wd/f1jVJVFqYhF6R9LMp46fEa4klE+V3on\nT2VX8O3IW6lLxKhNxKhLlIb7wWPEOBR8+0aT9I6k6BtJ0juaJJk+/Dmry6IsmVPJ0sYKljQG79lp\n1WW0940eCrS7uoZp6xklk305xyxprGDV/GpWLajmrAXVnLWghjlV8eA65Qc/BDULSV9/L1sHqtj0\nbCebnu3guY7g33dpYwWrFlST9SA8ZrJOOnzMZLK8dei7XDF0H3UMHP1NK62A9BicewNc9cWTev9T\nmSzPHhxi694+nmoLtv39Y6/4+5fOqeDGC5p51+qF1CaO3Sf6RpLc9Wgrdz+6h6GJNG9fUcGH1y3n\nzJaFJ/UaZlIhhNyrgcvc/QPh8R8BF7j7RybV2RbWaQ+PXwAuAD4DPO7u3wzL1wM/cvf7pvt5Crki\nIiIiheGlvy9zNTs7I9whNQqeDbZsBvcsbd3DbGnt4snWbg70DtFUaSyugqYKZ34iy7zyLI3xDOVM\nBKs/v+YaiMbz/WpOWDbr/PKFHr7/mwNksk5dRYya8pfDa22ilNryGHXhjOxxbz91FO7OaDJDx+A4\ne3pGaO0epbV7mNbuEfZ0jx4R7qIRo6WxgmVzKlk29+Vt6ZyK6Vcah+DU49LEER8+7esd5eHfdrJx\nZyf7ekcpidihLXroMUIkEsxEv/H0WtYtirCgZDA4m2K4E0Y6g8f0OFz66ZO79dA0OgbH2drWx4uD\n40RLIpSGbSstiRxaqC8aMRoqY5y3qPaE/98NjKW4+9E9rH9kN59++1lcvabplL+GU6UQQu41wFum\nhNy17v6nk+psD+tMDrlrgc8Cj00JuT909/un/IwPAh8EWLx48Zq2trYZeS0iIiIiIpJbY8kMbb0j\nvDgwTlNdguaGxJH30pZTZmg8RVlpSUG/x6805M7AndEPaQcWTTpuAg5MU6c9PF25Buh9hd+Lu98B\n3AHBTO4pa7mIiIiIiORVeayEladVs/K0E7/tmZy4qpNY7K3QzGRMfxJYbmZLzCwGXAdsmFJnA/De\ncP9qYJMHU8sbgOvMLG5mS4DlwBMz2FYREREREREpAjM2k+vuaTP7CPATglsI3eXu283ss8AWd98A\nrAe+YWa7CGZwrwu/d7uZ3QvsANLAzcdaWVlEREREREQEZvg+ubmkhadERERERESK1yu9JrdwryoW\nEREREREROUEKuSIiIiIiIlI0FHJFRERERESkaCjkioiIiIiISNFQyBUREREREZGioZArIiIiIiIi\nRUMhV0RERERERIqGQq6IiIiIiIgUDYVcERERERERKRoKuSIiIiIiIlI0FHJFRERERESkaCjkioiI\niIiISNFQyBUREREREZGioZArIiIiIiIiRUMhV0RERERERIqGQq6IiIiIiIgUDYVcERERERERKRrm\n7vluwylhZl1AW77bcRyNQHe+GyESUn+UQqL+KIVE/VEKifqjFJJ898dmd59zvEpFE3JfDcxsi7uf\nn+92iID6oxQW9UcpJOqPUkjUH6WQvFr6o05XFhERERERkaKhkCsiIiIiIiJFQyE3t+7IdwNEJlF/\nlEKi/iiFRP1RCon6oxSSV0V/1DW5IiIiIiIiUjQ0kysiIiIiIiJFQyE3R8zsMjP7rZntMrNb890e\nmT3MbJGZPWxmO81su5l9LCyvN7Ofmtnz4WNdvtsqs4eZlZjZ02b2/fB4iZltDvvjd8wslu82yuxg\nZrVmdp+ZPRuOk6/X+Cj5YmZ/Fv6u3mZm3zKzMo2PkktmdpeZdZrZtkllRx0TLXB7mG9+Y2ar89fy\nwynk5oCZlQBfBC4HVgHXm9mq/LZKZpE08OfufiZwIXBz2P9uBTa6+3JgY3gskisfA3ZOOr4N+Pew\nP/YB789Lq2Q2+g/gx+6+EjiXoF9qfJScM7OFwEeB8939bKAEuA6Nj5JbdwOXTSmbbky8HFgebh8E\nvpSjNh6XQm5urAV2uftud08C3wauzHObZJZw94PuvjXcHyL4A24hQR/8Wljta8BV+WmhzDZm1gRc\nAdwZHhuwDrgvrKL+KDlhZtXA7wHrAdw96e79aHyU/IkC5WYWBRLAQTQ+Sg65+/8BvVOKpxsTrwS+\n7oHHgVozm5+blh6bQm5uLAT2TTpuD8tEcsrMWoDXApuBee5+EIIgDMzNX8tklvk88JdANjxuAPrd\nPR0ea4yUXFkKdAH/FZ4+f6eZVaDxUfLA3fcD/wLsJQi3A8BTaHyU/JtuTCzYjKOQmxt2lDItay05\nZWaVwP3Ax919MN/tkdnJzN4GdLr7U5OLj1JVY6TkQhRYDXzJ3V8LjKBTkyVPwuscrwSWAAuACoLT\nQafS+CiFomB/fyvk5kY7sGjScRNwIE9tkVnIzEoJAu497v5AWNzx0ikl4WNnvtons8obgHeY2R6C\nSzfWEczs1oan54HGSMmddqDd3TeHx/cRhF6Nj5IPbwJa3b3L3VPAA8BFaHyU/JtuTCzYjKOQmxtP\nAsvD1fFiBIsIbMhzm2SWCK93XA/sdPd/m/SlDcB7w/33Av+T67bJ7OPun3T3JndvIRgLN7n7jcDD\nwNVhNfVHyQl3fxHYZ2ZnhEWXAjvQ+Cj5sRe40MwS4e/ul/qjxkfJt+nGxA3Ae8JVli8EBl46rTnf\nzL0gZpSLnpm9lWC2ogS4y93/Ic9NklnCzC4GfgE8w8vXQP4VwXW59wKLCX6xXuPuUxcaEJkxZnYJ\n8Al3f5uZLSWY2a0Hngb+0N0n8tk+mR3M7DyCRdBiwG7gfQSTABofJefM7G+BawnujPA08AGCaxw1\nPkpOmNm3gEuARqAD+BvgexxlTAw/jPkCwWrMo8D73H1LPto9lUKuiIiIiIiIFA2driwiIiIiIiJF\nQyFXREREREREioZCroiIiIiIiBQNhVwREREREREpGgq5IiIiIiIiUjQUckVERPLMzDJm9qtJ262n\n8LlbzGzbqXo+ERGRQhfNdwNERESEMXc/L9+NEBERKQaayRURESlQZrbHzG4zsyfCbVlY3mxmG83s\nN+Hj4rB8npk9aGa/DreLwqcqMbOvmtl2M3vIzMrz9qJERERmmEKuiIhI/pVPOV352klfG3T3tcAX\ngM+HZV8Avu7u5wD3ALeH5bcD/+vu5wKrge1h+XLgi+5+FtAPvGuGX4+IiEjemLvnuw0iIiKzmpkN\nu3vlUcr3AOvcfbeZlQIvunuDmXUD8909FZYfdPdGM+sCmtx9YtJztAA/dffl4fEtQKm7//3MvzIR\nEZHc00yuiIhIYfNp9qerczQTk/YzaE0OEREpYgq5IiIihe3aSY+Phfu/BK4L928EHgn3NwIfAjCz\nEjOrzlUjRURECoU+yRUREcm/cjP71aTjH7v7S7cRipvZZoIPpq8Pyz4K3GVmfwF0Ae8Lyz8G3GFm\n7yeYsf0QcHDGWy8iIlJAdE2uiIhIgQqvyT3f3bvz3RYREZFXC52uLCIiIiIiIkVDM7kiIiIiIiJS\nNDSTKyIiIiIiIkVDIVdERERERESKhkKuiIiIiIiIFA2FXBERERERESkaCrkiIiIiIiJSNBRyRURE\nREREpGj8P72Ine5z89e4AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 1152x648 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(figsize=(16, 9))\n",
    "\n",
    "sns.lineplot(data=losses, x=losses.index, y='Loss', hue='RNN', ax=ax)\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python3 (tf-gpu)",
   "language": "python",
   "name": "tf-gpu"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}